diff --git a/Datasets/all_datasets.zip b/Datasets/all_datasets.zip new file mode 100644 index 0000000000000000000000000000000000000000..b1e9c0278a55f67c6ee4b4cc67e93282984896e6 --- /dev/null +++ b/Datasets/all_datasets.zip @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:03ed805a79236c45d2240d6e11f94e7e14910a3b2b84f06c0be89f788b55d1c2 +size 12648332 diff --git a/Finetuned Models/finetuning-final.ipynb b/Finetuned Models/finetuning-final.ipynb new file mode 100644 index 0000000000000000000000000000000000000000..ae0f7970b566c7532ebc0dd9f69bb8c42f2031f9 --- /dev/null +++ b/Finetuned Models/finetuning-final.ipynb @@ -0,0 +1 @@ +{"cells":[{"cell_type":"code","execution_count":null,"metadata":{"execution":{"iopub.execute_input":"2024-08-14T11:36:45.359507Z","iopub.status.busy":"2024-08-14T11:36:45.358807Z","iopub.status.idle":"2024-08-14T11:38:49.639719Z","shell.execute_reply":"2024-08-14T11:38:49.638484Z","shell.execute_reply.started":"2024-08-14T11:36:45.359475Z"},"trusted":true},"outputs":[],"source":["# For any HF basic activities like loading models\n","# and tokenizers for running inference\n","# upgrade is a must for the newest Gemma model\n","!pip install -q --upgrade datasets\n","!pip install -q --upgrade transformers\n","\n","# For doing efficient stuff - PEFT\n","!pip install -q --upgrade peft\n","!pip install -q --upgrade trl\n","!pip install -q bitsandbytes\n","!pip install -q accelerate\n","\n","# for logging and visualizing training progress\n","!pip install -q tensorboard\n","# If creating a new dataset, useful for creating *.jsonl files\n","!pip install -q jsonlines"]},{"cell_type":"code","execution_count":null,"metadata":{"execution":{"iopub.execute_input":"2024-08-14T09:14:14.730158Z","iopub.status.busy":"2024-08-14T09:14:14.729841Z","iopub.status.idle":"2024-08-14T09:15:58.023466Z","shell.execute_reply":"2024-08-14T09:15:58.022512Z","shell.execute_reply.started":"2024-08-14T09:14:14.730128Z"},"trusted":true},"outputs":[],"source":["! conda install -y gdown"]},{"cell_type":"code","execution_count":null,"metadata":{"execution":{"iopub.execute_input":"2024-08-14T09:17:54.015233Z","iopub.status.busy":"2024-08-14T09:17:54.014883Z","iopub.status.idle":"2024-08-14T09:19:23.458473Z","shell.execute_reply":"2024-08-14T09:19:23.457105Z","shell.execute_reply.started":"2024-08-14T09:17:54.015206Z"},"trusted":true},"outputs":[],"source":["import itertools\n","import time\n","import warnings\n","from peft import LoraConfig, get_peft_model\n","from transformers import BertForSequenceClassification, AutoTokenizer, TrainingArguments, Trainer\n","from matplotlib import pyplot as plt\n","from datasets import load_dataset\n","import torch\n","from tqdm import tqdm\n","import numpy as np\n","from sklearn.metrics import accuracy_score, recall_score, precision_score, f1_score, matthews_corrcoef, roc_auc_score\n","import huggingface_hub\n","\n","huggingface_hub.login(token=hf_token)\n","\n","# Suppress warnings\n","warnings.filterwarnings(\"ignore\")\n","\n","# Layer configurations\n","attention_plus_feed_forward = [\n"," \"bert.encoder.layer.0.attention.self.query\",\n"," \"bert.encoder.layer.0.attention.self.key\",\n"," \"bert.encoder.layer.0.attention.self.value\",\n"," \"bert.encoder.layer.0.attention.output.dense\",\n"," \"bert.encoder.layer.0.intermediate.dense\",\n"," \"bert.encoder.layer.0.output.dense\",\n"," \"bert.encoder.layer.1.attention.self.query\",\n"," \"bert.encoder.layer.1.attention.self.key\",\n"," \"bert.encoder.layer.1.attention.self.value\",\n"," \"bert.encoder.layer.1.attention.output.dense\",\n"," \"bert.encoder.layer.1.intermediate.dense\",\n"," \"bert.encoder.layer.1.output.dense\"\n","]\n","\n","\n","tokenizer = AutoTokenizer.from_pretrained('zhihan1996/DNA_bert_6')\n","# Function to preprocess the dataset\n","def preprocess_function(examples):\n"," try:\n"," return tokenizer(\n"," examples['sequence'],\n"," padding='max_length',\n"," truncation=True,\n"," max_length=512\n"," )\n"," except KeyError:\n"," return tokenizer(\n"," examples['Sequence'],\n"," padding='max_length',\n"," truncation=True,\n"," max_length=512\n"," )\n","\n","\n","def add_labels(examples):\n"," try:\n"," examples['labels'] = examples['label']\n"," return examples\n"," except KeyError:\n"," examples['labels'] = examples['Label']\n"," return examples\n","\n","def create_task_dataset(task_name):\n"," if task_name == 'tfbs':\n"," return load_dataset('csv', data_files='/kaggle/working/tfbs.csv', split='train[0:10000]'), load_dataset('csv', data_files='/kaggle/working/tfbs.csv', split='train[10001:13122]')\n","\n"," elif task_name == 'dnasplice':\n"," return load_dataset('csv', data_files='/kaggle/working/dnasplice.csv', split='train[0:10000]'), load_dataset('csv', data_files='/kaggle/working/dnasplice.csv', split='train[10001:13122]')\n","\n"," elif task_name == 'dnaprom':\n"," return load_dataset('csv', data_files='/kaggle/working/dnaprom.csv', split='train[0:10000]'), load_dataset('csv', data_files='/kaggle/working/dnaprom.csv', split='train[10001:13122]')\n","\n"," else:\n"," raise ValueError(f\"Unknown task: {task_name}\")\n","\n","def create_dataset_maps(train_dataset, test_dataset):\n"," train_dataset = train_dataset.map(preprocess_function, batched=True)\n"," train_dataset = train_dataset.map(add_labels)\n"," test_dataset = test_dataset.map(preprocess_function, batched=True)\n"," test_dataset = test_dataset.map(add_labels)\n"," return train_dataset, test_dataset\n","\n","def train_model(train_dataset, test_dataset, model, task, model_name, config_name):\n"," def specificity_score(y_true, y_pred):\n"," true_negatives = np.sum((y_pred == 0) & (y_true == 0))\n"," false_positives = np.sum((y_pred == 1) & (y_true == 0))\n"," specificity = true_negatives / (true_negatives + false_positives + np.finfo(float).eps)\n"," return specificity\n","\n"," def compute_metrics(eval_pred):\n"," logits, labels = eval_pred\n"," predictions = np.argmax(logits, axis=-1)\n"," y_pred = logits[:, 1]\n","\n"," accuracy = accuracy_score(labels, predictions)\n"," recall = recall_score(labels, predictions)\n"," specificity = specificity_score(labels, predictions)\n"," mcc = matthews_corrcoef(labels, predictions)\n"," roc_auc = roc_auc_score(labels, y_pred)\n"," precision = precision_score(labels, predictions)\n"," f1 = f1_score(labels, predictions)\n","\n"," true_pos = np.sum((predictions == 1) & (labels == 1))\n"," true_neg = np.sum((predictions == 0) & (labels == 0))\n"," false_pos = np.sum((predictions == 1) & (labels == 0))\n"," false_neg = np.sum((predictions == 0) & (labels == 1))\n","\n"," return {\n"," 'accuracy': accuracy,\n"," 'recall': recall,\n"," 'specificity': specificity,\n"," 'mcc': mcc,\n"," 'roc_auc': roc_auc,\n"," 'precision': precision,\n"," 'f1': f1,\n"," 'true_pos': true_pos,\n"," 'true_neg': true_neg,\n"," 'false_pos': false_pos,\n"," 'false_neg': false_neg\n"," }\n","\n"," # Define the training arguments\n"," training_arguments = TrainingArguments(\n"," output_dir=f\"outputs/{task}/{model_name}_{config_name}\",\n"," num_train_epochs=25,\n"," fp16=False,\n"," bf16=False,\n"," per_device_train_batch_size=20,\n"," per_device_eval_batch_size=10,\n"," gradient_accumulation_steps=2,\n"," gradient_checkpointing=True,\n"," max_grad_norm=0.3,\n"," learning_rate=4e-4,\n"," weight_decay=0.01,\n"," optim=\"paged_adamw_32bit\",\n"," lr_scheduler_type=\"linear\",\n"," max_steps=-1,\n"," warmup_ratio=0.03,\n"," group_by_length=True,\n"," save_steps=1000,\n"," logging_steps=25,\n"," dataloader_pin_memory=False,\n"," report_to='tensorboard',\n"," gradient_checkpointing_kwargs={'use_reentrant': False}\n"," )\n","\n"," trainer = Trainer(\n"," model=model,\n"," args=training_arguments,\n"," train_dataset=train_dataset,\n"," eval_dataset=test_dataset,\n"," tokenizer=tokenizer,\n"," compute_metrics=compute_metrics,\n"," )\n","\n"," start_time = time.time()\n"," trainer.train()\n"," end_time = time.time()\n","\n"," total_time = end_time - start_time\n"," metrics = trainer.evaluate()\n","\n"," return total_time, metrics\n","\n","# Task loop\n","task_list = ['dnasplice', 'tfbs', 'dnaprom']\n","log_file = \"training_log.txt\"\n","model_name = 'fabihamakhdoomi/TinyDNABERT'\n","for task in task_list:\n"," print(f\"Running TASK : {task}\")\n"," train_dataset, test_dataset = create_task_dataset(task)\n"," train_dataset, test_dataset = create_dataset_maps(train_dataset, test_dataset)\n"," train_dataset.set_format(type='torch', columns=['input_ids', 'attention_mask', 'labels'])\n"," test_dataset.set_format(type='torch', columns=['input_ids', 'attention_mask', 'labels'])\n","\n","\n"," # Train the base model first\n"," base_model = BertForSequenceClassification.from_pretrained(\n"," model_name,\n"," num_labels=2\n"," )\n"," config_name = \"base_model\"\n"," print(f\"Training MODEL : {config_name} for task : {task}\")\n"," training_time, metrics = train_model(train_dataset, test_dataset, base_model, task, model_name, config_name)\n"," with open(log_file, \"a\") as log:\n"," log.write(f\"Task: {task}, Model: {model_name}, Config: {config_name}, Training Time: {training_time}, Metrics: {metrics}\\n\")\n","\n"," # Train the LoRA models\n"," config_name = \"attention_plus_feed_forward\"\n"," base_model = BertForSequenceClassification.from_pretrained(\n"," model_name,\n"," num_labels=2\n"," )\n"," if task == 'dnasplice':\n"," r_value = 4\n"," print('Setting r value to 4 for dnasplice')\n"," else:\n"," r_value = 8\n"," peft_config = LoraConfig(\n"," lora_alpha=16,\n"," lora_dropout=0.2,\n"," r=r_value,\n"," bias=\"none\",\n"," task_type=\"SEQ_CLS\",\n"," target_modules=attention_plus_feed_forward\n"," )\n"," model = get_peft_model(base_model, peft_config)\n"," print(f\"Training MODEL : {config_name} for task : {task}\")\n"," training_time, metrics = train_model(train_dataset, test_dataset, model, task, model_name, config_name)\n"," with open(log_file, \"a\") as log:\n"," log.write(f\"Task: {task}, Model: {model_name}, Config: {config_name}, Training Time: {training_time}, Metrics: {metrics}\\n\")\n"]},{"cell_type":"code","execution_count":null,"metadata":{},"outputs":[],"source":[]}],"metadata":{"kaggle":{"accelerator":"gpu","dataSources":[],"dockerImageVersionId":30747,"isGpuEnabled":true,"isInternetEnabled":true,"language":"python","sourceType":"notebook"},"kernelspec":{"display_name":"Python 3","language":"python","name":"python3"},"language_info":{"codemirror_mode":{"name":"ipython","version":3},"file_extension":".py","mimetype":"text/x-python","name":"python","nbconvert_exporter":"python","pygments_lexer":"ipython3","version":"3.10.13"}},"nbformat":4,"nbformat_minor":4} diff --git a/Finetuned Models/finetuning_log.txt b/Finetuned Models/finetuning_log.txt new file mode 100644 index 0000000000000000000000000000000000000000..584cdd23e4bc6a77766520fdf53cd59a2615e297 --- /dev/null +++ b/Finetuned Models/finetuning_log.txt @@ -0,0 +1,6 @@ +Task: dnasplice, Model: fabihamakhdoomi/TinyDNABERT, Config: base_model, Training Time: 468.7715392112732, Metrics: {'eval_loss': 0.6981738209724426, 'eval_accuracy': 0.9202178788849728, 'eval_recall': 0.7544378698224852, 'eval_specificity': 0.9660531697341513, 'eval_mcc': 0.7565288289137028, 'eval_roc_auc': 0.9536779564622887, 'eval_precision': 0.8600337268128162, 'eval_f1': 0.8037825059101655, 'eval_true_pos': 510, 'eval_true_neg': 2362, 'eval_false_pos': 83, 'eval_false_neg': 166, 'eval_runtime': 2.8254, 'eval_samples_per_second': 1104.619, 'eval_steps_per_second': 110.78, 'epoch': 25.0} +Task: dnasplice, Model: fabihamakhdoomi/TinyDNABERT, Config: attention_plus_feed_forward, Training Time: 550.877524137497, Metrics: {'eval_loss': 0.3186633586883545, 'eval_accuracy': 0.8766421018904197, 'eval_recall': 0.4985207100591716, 'eval_specificity': 0.9811860940695296, 'eval_mcc': 0.6022440934663668, 'eval_roc_auc': 0.8810529882261832, 'eval_precision': 0.8798955613577023, 'eval_f1': 0.6364494806421152, 'eval_true_pos': 337, 'eval_true_neg': 2399, 'eval_false_pos': 46, 'eval_false_neg': 339, 'eval_runtime': 3.3308, 'eval_samples_per_second': 937.023, 'eval_steps_per_second': 93.973, 'epoch': 25.0} +Task: tfbs, Model: fabihamakhdoomi/TinyDNABERT, Config: base_model, Training Time: 469.97587299346924, Metrics: {'eval_loss': 2.653303384780884, 'eval_accuracy': 0.6578019865427748, 'eval_recall': 0.6830497794580971, 'eval_specificity': 0.6316818774445893, 'eval_mcc': 0.3151935621806032, 'eval_roc_auc': 0.6742539817897865, 'eval_precision': 0.6573681018799272, 'eval_f1': 0.6699629171817059, 'eval_true_pos': 1084, 'eval_true_neg': 969, 'eval_false_pos': 565, 'eval_false_neg': 503, 'eval_runtime': 2.7557, 'eval_samples_per_second': 1132.56, 'eval_steps_per_second': 113.583, 'epoch': 25.0} +Task: tfbs, Model: fabihamakhdoomi/TinyDNABERT, Config: attention_plus_feed_forward, Training Time: 549.6457896232605, Metrics: {'eval_loss': 0.635671079158783, 'eval_accuracy': 0.6536366549182954, 'eval_recall': 0.6962822936357907, 'eval_specificity': 0.6095176010430248, 'eval_mcc': 0.307056824567405, 'eval_roc_auc': 0.7121371574288815, 'eval_precision': 0.6484741784037559, 'eval_f1': 0.6715284108173807, 'eval_true_pos': 1105, 'eval_true_neg': 935, 'eval_false_pos': 599, 'eval_false_neg': 482, 'eval_runtime': 3.2385, 'eval_samples_per_second': 963.73, 'eval_steps_per_second': 96.651, 'epoch': 25.0} +Task: dnaprom, Model: fabihamakhdoomi/TinyDNABERT, Config: base_model, Training Time: 470.97279715538025, Metrics: {'eval_loss': 0.9456052184104919, 'eval_accuracy': 0.887536046139058, 'eval_recall': 0.9066073697585769, 'eval_specificity': 0.8681318681318682, 'eval_mcc': 0.7755020642874026, 'eval_roc_auc': 0.9237130684548279, 'eval_precision': 0.8749233599019006, 'eval_f1': 0.8904836193447738, 'eval_true_pos': 1427, 'eval_true_neg': 1343, 'eval_false_pos': 204, 'eval_false_neg': 147, 'eval_runtime': 2.8069, 'eval_samples_per_second': 1111.884, 'eval_steps_per_second': 111.509, 'epoch': 25.0} +Task: dnaprom, Model: fabihamakhdoomi/TinyDNABERT, Config: attention_plus_feed_forward, Training Time: 558.7053642272949, Metrics: {'eval_loss': 0.3336106240749359, 'eval_accuracy': 0.8583787247677026, 'eval_recall': 0.8875476493011436, 'eval_specificity': 0.8287007110536523, 'eval_mcc': 0.7177414140554846, 'eval_roc_auc': 0.9391871302327989, 'eval_precision': 0.8405535499398316, 'eval_f1': 0.8634116192830655, 'eval_true_pos': 1397, 'eval_true_neg': 1282, 'eval_false_pos': 265, 'eval_false_neg': 177, 'eval_runtime': 3.2789, 'eval_samples_per_second': 951.851, 'eval_steps_per_second': 95.46, 'epoch': 25.0} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/README.md b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..500fd976d92b217455f181bc98c77e612d9a18d9 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/README.md @@ -0,0 +1,202 @@ +--- +base_model: fabihamakhdoomi/TinyDNABERT +library_name: peft +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.12.0 \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/adapter_config.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..04c63e99bbc6b810449dece367c0afb165125b92 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/adapter_config.json @@ -0,0 +1,42 @@ +{ + "alpha_pattern": {}, + "auto_mapping": null, + "base_model_name_or_path": "fabihamakhdoomi/TinyDNABERT", + "bias": "none", + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 16, + "lora_dropout": 0.2, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": [ + "classifier", + "score" + ], + "peft_type": "LORA", + "r": 8, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "bert.encoder.layer.1.attention.self.query", + "bert.encoder.layer.0.output.dense", + "bert.encoder.layer.1.output.dense", + "bert.encoder.layer.0.intermediate.dense", + "bert.encoder.layer.1.intermediate.dense", + "bert.encoder.layer.0.attention.output.dense", + "bert.encoder.layer.1.attention.output.dense", + "bert.encoder.layer.0.attention.self.query", + "bert.encoder.layer.0.attention.self.value", + "bert.encoder.layer.1.attention.self.key", + "bert.encoder.layer.1.attention.self.value", + "bert.encoder.layer.0.attention.self.key" + ], + "task_type": "SEQ_CLS", + "use_dora": false, + "use_rslora": false +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/adapter_model.safetensors b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/adapter_model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..d529307aa360db0b4dcf5558cc7df55e65eb2195 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/adapter_model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:92ea0cf71ee3b3c8e6751f0820e5e8da96982438c6f6c28168fd940e9247344c +size 151952 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/optimizer.pt b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..9d14c6d34e82dbcc2185a6d36c25683aa695d1bb --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f5698b57e9b011fe13f493ec9f2d71ccd216389770806246d493da09194c1ee4 +size 312682 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/rng_state.pth b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..d769758d72ed276350b7475a3c6ab31d03d6d464 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:af365fb722ff3640b9c3cbda0c33c89309e74956fafe530249e5f878f489cae3 +size 14244 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/scheduler.pt b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..e21fdaa14c288938c4b2c1277194110b759d8724 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7480edaf7ecb5003556fedbf146cf26ec75a9ee19931ce031bbb3b509f9927ce +size 1064 diff --git a/special_tokens_map.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/special_tokens_map.json similarity index 100% rename from special_tokens_map.json rename to Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/special_tokens_map.json diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/tokenizer.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..ebd782de6963b9032ce9374cfe30f5275be1a30b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/tokenizer.json @@ -0,0 +1,4265 @@ +{ + "version": "1.0", + "truncation": { + "direction": "Right", + "max_length": 512, + "strategy": "LongestFirst", + "stride": 0 + }, + "padding": { + "strategy": { + "Fixed": 512 + }, + "direction": "Right", + "pad_to_multiple_of": null, + "pad_id": 0, + "pad_type_id": 0, + "pad_token": "[PAD]" + }, + "added_tokens": [ + { + "id": 0, + "content": "[PAD]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 1, + "content": "[UNK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 2, + "content": "[CLS]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 3, + "content": "[SEP]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 4, + "content": "[MASK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + } + ], + "normalizer": { + "type": "BertNormalizer", + "clean_text": true, + "handle_chinese_chars": true, + "strip_accents": null, + "lowercase": false + }, + "pre_tokenizer": { + "type": "BertPreTokenizer" + }, + "post_processor": { + "type": "TemplateProcessing", + "single": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + } + ], + "pair": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "B", + "type_id": 1 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 1 + } + } + ], + "special_tokens": { + "[CLS]": { + "id": "[CLS]", + "ids": [ + 2 + ], + "tokens": [ + "[CLS]" + ] + }, + "[SEP]": { + "id": "[SEP]", + "ids": [ + 3 + ], + "tokens": [ + "[SEP]" + ] + } + } + }, + "decoder": { + "type": "WordPiece", + "prefix": "##", + "cleanup": true + }, + "model": { + "type": "WordPiece", + "unk_token": "[UNK]", + "continuing_subword_prefix": "##", + "max_input_chars_per_word": 100, + "vocab": { + "[PAD]": 0, + "[UNK]": 1, + "[CLS]": 2, + "[SEP]": 3, + "[MASK]": 4, + "AAAAAA": 5, + "AAAAAT": 6, + "AAAAAC": 7, + "AAAAAG": 8, + "AAAATA": 9, + "AAAATT": 10, + "AAAATC": 11, + "AAAATG": 12, + "AAAACA": 13, + "AAAACT": 14, + "AAAACC": 15, + "AAAACG": 16, + "AAAAGA": 17, + "AAAAGT": 18, + "AAAAGC": 19, + "AAAAGG": 20, + "AAATAA": 21, + "AAATAT": 22, + "AAATAC": 23, + "AAATAG": 24, + "AAATTA": 25, + "AAATTT": 26, + "AAATTC": 27, + "AAATTG": 28, + "AAATCA": 29, + "AAATCT": 30, + "AAATCC": 31, + "AAATCG": 32, + "AAATGA": 33, + "AAATGT": 34, + "AAATGC": 35, + "AAATGG": 36, + "AAACAA": 37, + "AAACAT": 38, + "AAACAC": 39, + "AAACAG": 40, + "AAACTA": 41, + "AAACTT": 42, + "AAACTC": 43, + "AAACTG": 44, + "AAACCA": 45, + "AAACCT": 46, + "AAACCC": 47, + "AAACCG": 48, + "AAACGA": 49, + "AAACGT": 50, + "AAACGC": 51, + "AAACGG": 52, + "AAAGAA": 53, + "AAAGAT": 54, + "AAAGAC": 55, + "AAAGAG": 56, + "AAAGTA": 57, + "AAAGTT": 58, + "AAAGTC": 59, + "AAAGTG": 60, + "AAAGCA": 61, + "AAAGCT": 62, + "AAAGCC": 63, + "AAAGCG": 64, + "AAAGGA": 65, + "AAAGGT": 66, + "AAAGGC": 67, + "AAAGGG": 68, + "AATAAA": 69, + "AATAAT": 70, + "AATAAC": 71, + "AATAAG": 72, + "AATATA": 73, + "AATATT": 74, + "AATATC": 75, + "AATATG": 76, + "AATACA": 77, + "AATACT": 78, + "AATACC": 79, + "AATACG": 80, + "AATAGA": 81, + "AATAGT": 82, + "AATAGC": 83, + "AATAGG": 84, + "AATTAA": 85, + "AATTAT": 86, + "AATTAC": 87, + "AATTAG": 88, + "AATTTA": 89, + "AATTTT": 90, + "AATTTC": 91, + "AATTTG": 92, + "AATTCA": 93, + "AATTCT": 94, + "AATTCC": 95, + "AATTCG": 96, + "AATTGA": 97, + "AATTGT": 98, + "AATTGC": 99, + "AATTGG": 100, + "AATCAA": 101, + "AATCAT": 102, + "AATCAC": 103, + "AATCAG": 104, + "AATCTA": 105, + "AATCTT": 106, + "AATCTC": 107, + "AATCTG": 108, + "AATCCA": 109, + "AATCCT": 110, + "AATCCC": 111, + "AATCCG": 112, + "AATCGA": 113, + "AATCGT": 114, + "AATCGC": 115, + "AATCGG": 116, + "AATGAA": 117, + "AATGAT": 118, + "AATGAC": 119, + "AATGAG": 120, + "AATGTA": 121, + "AATGTT": 122, + "AATGTC": 123, + "AATGTG": 124, + "AATGCA": 125, + "AATGCT": 126, + "AATGCC": 127, + "AATGCG": 128, + "AATGGA": 129, + "AATGGT": 130, + "AATGGC": 131, + "AATGGG": 132, + "AACAAA": 133, + "AACAAT": 134, + "AACAAC": 135, + "AACAAG": 136, + "AACATA": 137, + "AACATT": 138, + "AACATC": 139, + "AACATG": 140, + "AACACA": 141, + "AACACT": 142, + "AACACC": 143, + "AACACG": 144, + "AACAGA": 145, + "AACAGT": 146, + "AACAGC": 147, + "AACAGG": 148, + "AACTAA": 149, + "AACTAT": 150, + "AACTAC": 151, + "AACTAG": 152, + "AACTTA": 153, + "AACTTT": 154, + "AACTTC": 155, + "AACTTG": 156, + "AACTCA": 157, + "AACTCT": 158, + "AACTCC": 159, + "AACTCG": 160, + "AACTGA": 161, + "AACTGT": 162, + "AACTGC": 163, + "AACTGG": 164, + "AACCAA": 165, + "AACCAT": 166, + "AACCAC": 167, + "AACCAG": 168, + "AACCTA": 169, + "AACCTT": 170, + "AACCTC": 171, + "AACCTG": 172, + "AACCCA": 173, + "AACCCT": 174, + "AACCCC": 175, + "AACCCG": 176, + "AACCGA": 177, + "AACCGT": 178, + "AACCGC": 179, + "AACCGG": 180, + "AACGAA": 181, + "AACGAT": 182, + "AACGAC": 183, + "AACGAG": 184, + "AACGTA": 185, + "AACGTT": 186, + "AACGTC": 187, + "AACGTG": 188, + "AACGCA": 189, + "AACGCT": 190, + "AACGCC": 191, + "AACGCG": 192, + "AACGGA": 193, + "AACGGT": 194, + "AACGGC": 195, + "AACGGG": 196, + "AAGAAA": 197, + "AAGAAT": 198, + "AAGAAC": 199, + "AAGAAG": 200, + "AAGATA": 201, + "AAGATT": 202, + "AAGATC": 203, + "AAGATG": 204, + "AAGACA": 205, + "AAGACT": 206, + "AAGACC": 207, + "AAGACG": 208, + "AAGAGA": 209, + "AAGAGT": 210, + "AAGAGC": 211, + "AAGAGG": 212, + "AAGTAA": 213, + "AAGTAT": 214, + "AAGTAC": 215, + "AAGTAG": 216, + "AAGTTA": 217, + "AAGTTT": 218, + "AAGTTC": 219, + "AAGTTG": 220, + "AAGTCA": 221, + "AAGTCT": 222, + "AAGTCC": 223, + "AAGTCG": 224, + "AAGTGA": 225, + "AAGTGT": 226, + "AAGTGC": 227, + "AAGTGG": 228, + "AAGCAA": 229, + "AAGCAT": 230, + "AAGCAC": 231, + "AAGCAG": 232, + "AAGCTA": 233, + "AAGCTT": 234, + "AAGCTC": 235, + "AAGCTG": 236, + "AAGCCA": 237, + "AAGCCT": 238, + "AAGCCC": 239, + "AAGCCG": 240, + "AAGCGA": 241, + "AAGCGT": 242, + "AAGCGC": 243, + "AAGCGG": 244, + "AAGGAA": 245, + "AAGGAT": 246, + "AAGGAC": 247, + "AAGGAG": 248, + "AAGGTA": 249, + "AAGGTT": 250, + "AAGGTC": 251, + "AAGGTG": 252, + "AAGGCA": 253, + "AAGGCT": 254, + "AAGGCC": 255, + "AAGGCG": 256, + "AAGGGA": 257, + "AAGGGT": 258, + "AAGGGC": 259, + "AAGGGG": 260, + "ATAAAA": 261, + "ATAAAT": 262, + "ATAAAC": 263, + "ATAAAG": 264, + "ATAATA": 265, + "ATAATT": 266, + "ATAATC": 267, + "ATAATG": 268, + "ATAACA": 269, + "ATAACT": 270, + "ATAACC": 271, + "ATAACG": 272, + "ATAAGA": 273, + "ATAAGT": 274, + "ATAAGC": 275, + "ATAAGG": 276, + "ATATAA": 277, + "ATATAT": 278, + "ATATAC": 279, + "ATATAG": 280, + "ATATTA": 281, + "ATATTT": 282, + "ATATTC": 283, + "ATATTG": 284, + "ATATCA": 285, + "ATATCT": 286, + "ATATCC": 287, + "ATATCG": 288, + "ATATGA": 289, + "ATATGT": 290, + "ATATGC": 291, + "ATATGG": 292, + "ATACAA": 293, + "ATACAT": 294, + "ATACAC": 295, + "ATACAG": 296, + "ATACTA": 297, + "ATACTT": 298, + "ATACTC": 299, + "ATACTG": 300, + "ATACCA": 301, + "ATACCT": 302, + "ATACCC": 303, + "ATACCG": 304, + "ATACGA": 305, + "ATACGT": 306, + "ATACGC": 307, + "ATACGG": 308, + "ATAGAA": 309, + "ATAGAT": 310, + "ATAGAC": 311, + "ATAGAG": 312, + "ATAGTA": 313, + "ATAGTT": 314, + "ATAGTC": 315, + "ATAGTG": 316, + "ATAGCA": 317, + "ATAGCT": 318, + "ATAGCC": 319, + "ATAGCG": 320, + "ATAGGA": 321, + "ATAGGT": 322, + "ATAGGC": 323, + "ATAGGG": 324, + "ATTAAA": 325, + "ATTAAT": 326, + "ATTAAC": 327, + "ATTAAG": 328, + "ATTATA": 329, + "ATTATT": 330, + "ATTATC": 331, + "ATTATG": 332, + "ATTACA": 333, + "ATTACT": 334, + "ATTACC": 335, + "ATTACG": 336, + "ATTAGA": 337, + "ATTAGT": 338, + "ATTAGC": 339, + "ATTAGG": 340, + "ATTTAA": 341, + "ATTTAT": 342, + "ATTTAC": 343, + "ATTTAG": 344, + "ATTTTA": 345, + "ATTTTT": 346, + "ATTTTC": 347, + "ATTTTG": 348, + "ATTTCA": 349, + "ATTTCT": 350, + "ATTTCC": 351, + "ATTTCG": 352, + "ATTTGA": 353, + "ATTTGT": 354, + "ATTTGC": 355, + "ATTTGG": 356, + "ATTCAA": 357, + "ATTCAT": 358, + "ATTCAC": 359, + "ATTCAG": 360, + "ATTCTA": 361, + "ATTCTT": 362, + "ATTCTC": 363, + "ATTCTG": 364, + "ATTCCA": 365, + "ATTCCT": 366, + "ATTCCC": 367, + "ATTCCG": 368, + "ATTCGA": 369, + "ATTCGT": 370, + "ATTCGC": 371, + "ATTCGG": 372, + "ATTGAA": 373, + "ATTGAT": 374, + "ATTGAC": 375, + "ATTGAG": 376, + "ATTGTA": 377, + "ATTGTT": 378, + "ATTGTC": 379, + "ATTGTG": 380, + "ATTGCA": 381, + "ATTGCT": 382, + "ATTGCC": 383, + "ATTGCG": 384, + "ATTGGA": 385, + "ATTGGT": 386, + "ATTGGC": 387, + "ATTGGG": 388, + "ATCAAA": 389, + "ATCAAT": 390, + "ATCAAC": 391, + "ATCAAG": 392, + "ATCATA": 393, + "ATCATT": 394, + "ATCATC": 395, + "ATCATG": 396, + "ATCACA": 397, + "ATCACT": 398, + "ATCACC": 399, + "ATCACG": 400, + "ATCAGA": 401, + "ATCAGT": 402, + "ATCAGC": 403, + "ATCAGG": 404, + "ATCTAA": 405, + "ATCTAT": 406, + "ATCTAC": 407, + "ATCTAG": 408, + "ATCTTA": 409, + "ATCTTT": 410, + "ATCTTC": 411, + "ATCTTG": 412, + "ATCTCA": 413, + "ATCTCT": 414, + "ATCTCC": 415, + "ATCTCG": 416, + "ATCTGA": 417, + "ATCTGT": 418, + "ATCTGC": 419, + "ATCTGG": 420, + "ATCCAA": 421, + "ATCCAT": 422, + "ATCCAC": 423, + "ATCCAG": 424, + "ATCCTA": 425, + "ATCCTT": 426, + "ATCCTC": 427, + "ATCCTG": 428, + "ATCCCA": 429, + "ATCCCT": 430, + "ATCCCC": 431, + "ATCCCG": 432, + "ATCCGA": 433, + "ATCCGT": 434, + "ATCCGC": 435, + "ATCCGG": 436, + "ATCGAA": 437, + "ATCGAT": 438, + "ATCGAC": 439, + "ATCGAG": 440, + "ATCGTA": 441, + "ATCGTT": 442, + "ATCGTC": 443, + "ATCGTG": 444, + "ATCGCA": 445, + "ATCGCT": 446, + "ATCGCC": 447, + "ATCGCG": 448, + "ATCGGA": 449, + "ATCGGT": 450, + "ATCGGC": 451, + "ATCGGG": 452, + "ATGAAA": 453, + "ATGAAT": 454, + "ATGAAC": 455, + "ATGAAG": 456, + "ATGATA": 457, + "ATGATT": 458, + "ATGATC": 459, + "ATGATG": 460, + "ATGACA": 461, + "ATGACT": 462, + "ATGACC": 463, + "ATGACG": 464, + "ATGAGA": 465, + "ATGAGT": 466, + "ATGAGC": 467, + "ATGAGG": 468, + "ATGTAA": 469, + "ATGTAT": 470, + "ATGTAC": 471, + "ATGTAG": 472, + "ATGTTA": 473, + "ATGTTT": 474, + "ATGTTC": 475, + "ATGTTG": 476, + "ATGTCA": 477, + "ATGTCT": 478, + "ATGTCC": 479, + "ATGTCG": 480, + "ATGTGA": 481, + "ATGTGT": 482, + "ATGTGC": 483, + "ATGTGG": 484, + "ATGCAA": 485, + "ATGCAT": 486, + "ATGCAC": 487, + "ATGCAG": 488, + "ATGCTA": 489, + "ATGCTT": 490, + "ATGCTC": 491, + "ATGCTG": 492, + "ATGCCA": 493, + "ATGCCT": 494, + "ATGCCC": 495, + "ATGCCG": 496, + "ATGCGA": 497, + "ATGCGT": 498, + "ATGCGC": 499, + "ATGCGG": 500, + "ATGGAA": 501, + "ATGGAT": 502, + "ATGGAC": 503, + "ATGGAG": 504, + "ATGGTA": 505, + "ATGGTT": 506, + "ATGGTC": 507, + "ATGGTG": 508, + "ATGGCA": 509, + "ATGGCT": 510, + "ATGGCC": 511, + "ATGGCG": 512, + "ATGGGA": 513, + "ATGGGT": 514, + "ATGGGC": 515, + "ATGGGG": 516, + "ACAAAA": 517, + "ACAAAT": 518, + "ACAAAC": 519, + "ACAAAG": 520, + "ACAATA": 521, + "ACAATT": 522, + "ACAATC": 523, + "ACAATG": 524, + "ACAACA": 525, + "ACAACT": 526, + "ACAACC": 527, + "ACAACG": 528, + "ACAAGA": 529, + "ACAAGT": 530, + "ACAAGC": 531, + "ACAAGG": 532, + "ACATAA": 533, + "ACATAT": 534, + "ACATAC": 535, + "ACATAG": 536, + "ACATTA": 537, + "ACATTT": 538, + "ACATTC": 539, + "ACATTG": 540, + "ACATCA": 541, + "ACATCT": 542, + "ACATCC": 543, + "ACATCG": 544, + "ACATGA": 545, + "ACATGT": 546, + "ACATGC": 547, + "ACATGG": 548, + "ACACAA": 549, + "ACACAT": 550, + "ACACAC": 551, + "ACACAG": 552, + "ACACTA": 553, + "ACACTT": 554, + "ACACTC": 555, + "ACACTG": 556, + "ACACCA": 557, + "ACACCT": 558, + "ACACCC": 559, + "ACACCG": 560, + "ACACGA": 561, + "ACACGT": 562, + "ACACGC": 563, + "ACACGG": 564, + "ACAGAA": 565, + "ACAGAT": 566, + "ACAGAC": 567, + "ACAGAG": 568, + "ACAGTA": 569, + "ACAGTT": 570, + "ACAGTC": 571, + "ACAGTG": 572, + "ACAGCA": 573, + "ACAGCT": 574, + "ACAGCC": 575, + "ACAGCG": 576, + "ACAGGA": 577, + "ACAGGT": 578, + "ACAGGC": 579, + "ACAGGG": 580, + "ACTAAA": 581, + "ACTAAT": 582, + "ACTAAC": 583, + "ACTAAG": 584, + "ACTATA": 585, + "ACTATT": 586, + "ACTATC": 587, + "ACTATG": 588, + "ACTACA": 589, + "ACTACT": 590, + "ACTACC": 591, + "ACTACG": 592, + "ACTAGA": 593, + "ACTAGT": 594, + "ACTAGC": 595, + "ACTAGG": 596, + "ACTTAA": 597, + "ACTTAT": 598, + "ACTTAC": 599, + "ACTTAG": 600, + "ACTTTA": 601, + "ACTTTT": 602, + "ACTTTC": 603, + "ACTTTG": 604, + "ACTTCA": 605, + "ACTTCT": 606, + "ACTTCC": 607, + "ACTTCG": 608, + "ACTTGA": 609, + "ACTTGT": 610, + "ACTTGC": 611, + "ACTTGG": 612, + "ACTCAA": 613, + "ACTCAT": 614, + "ACTCAC": 615, + "ACTCAG": 616, + "ACTCTA": 617, + "ACTCTT": 618, + "ACTCTC": 619, + "ACTCTG": 620, + "ACTCCA": 621, + "ACTCCT": 622, + "ACTCCC": 623, + "ACTCCG": 624, + "ACTCGA": 625, + "ACTCGT": 626, + "ACTCGC": 627, + "ACTCGG": 628, + "ACTGAA": 629, + "ACTGAT": 630, + "ACTGAC": 631, + "ACTGAG": 632, + "ACTGTA": 633, + "ACTGTT": 634, + "ACTGTC": 635, + "ACTGTG": 636, + "ACTGCA": 637, + "ACTGCT": 638, + "ACTGCC": 639, + "ACTGCG": 640, + "ACTGGA": 641, + "ACTGGT": 642, + "ACTGGC": 643, + "ACTGGG": 644, + "ACCAAA": 645, + "ACCAAT": 646, + "ACCAAC": 647, + "ACCAAG": 648, + "ACCATA": 649, + "ACCATT": 650, + "ACCATC": 651, + "ACCATG": 652, + "ACCACA": 653, + "ACCACT": 654, + "ACCACC": 655, + "ACCACG": 656, + "ACCAGA": 657, + "ACCAGT": 658, + "ACCAGC": 659, + "ACCAGG": 660, + "ACCTAA": 661, + "ACCTAT": 662, + "ACCTAC": 663, + "ACCTAG": 664, + "ACCTTA": 665, + "ACCTTT": 666, + "ACCTTC": 667, + "ACCTTG": 668, + "ACCTCA": 669, + "ACCTCT": 670, + "ACCTCC": 671, + "ACCTCG": 672, + "ACCTGA": 673, + "ACCTGT": 674, + "ACCTGC": 675, + "ACCTGG": 676, + "ACCCAA": 677, + "ACCCAT": 678, + "ACCCAC": 679, + "ACCCAG": 680, + "ACCCTA": 681, + "ACCCTT": 682, + "ACCCTC": 683, + "ACCCTG": 684, + "ACCCCA": 685, + "ACCCCT": 686, + "ACCCCC": 687, + "ACCCCG": 688, + "ACCCGA": 689, + "ACCCGT": 690, + "ACCCGC": 691, + "ACCCGG": 692, + "ACCGAA": 693, + "ACCGAT": 694, + "ACCGAC": 695, + "ACCGAG": 696, + "ACCGTA": 697, + "ACCGTT": 698, + "ACCGTC": 699, + "ACCGTG": 700, + "ACCGCA": 701, + "ACCGCT": 702, + "ACCGCC": 703, + "ACCGCG": 704, + "ACCGGA": 705, + "ACCGGT": 706, + "ACCGGC": 707, + "ACCGGG": 708, + "ACGAAA": 709, + "ACGAAT": 710, + "ACGAAC": 711, + "ACGAAG": 712, + "ACGATA": 713, + "ACGATT": 714, + "ACGATC": 715, + "ACGATG": 716, + "ACGACA": 717, + "ACGACT": 718, + "ACGACC": 719, + "ACGACG": 720, + "ACGAGA": 721, + "ACGAGT": 722, + "ACGAGC": 723, + "ACGAGG": 724, + "ACGTAA": 725, + "ACGTAT": 726, + "ACGTAC": 727, + "ACGTAG": 728, + "ACGTTA": 729, + "ACGTTT": 730, + "ACGTTC": 731, + "ACGTTG": 732, + "ACGTCA": 733, + "ACGTCT": 734, + "ACGTCC": 735, + "ACGTCG": 736, + "ACGTGA": 737, + "ACGTGT": 738, + "ACGTGC": 739, + "ACGTGG": 740, + "ACGCAA": 741, + "ACGCAT": 742, + "ACGCAC": 743, + "ACGCAG": 744, + "ACGCTA": 745, + "ACGCTT": 746, + "ACGCTC": 747, + "ACGCTG": 748, + "ACGCCA": 749, + "ACGCCT": 750, + "ACGCCC": 751, + "ACGCCG": 752, + "ACGCGA": 753, + "ACGCGT": 754, + "ACGCGC": 755, + "ACGCGG": 756, + "ACGGAA": 757, + "ACGGAT": 758, + "ACGGAC": 759, + "ACGGAG": 760, + "ACGGTA": 761, + "ACGGTT": 762, + "ACGGTC": 763, + "ACGGTG": 764, + "ACGGCA": 765, + "ACGGCT": 766, + "ACGGCC": 767, + "ACGGCG": 768, + "ACGGGA": 769, + "ACGGGT": 770, + "ACGGGC": 771, + "ACGGGG": 772, + "AGAAAA": 773, + "AGAAAT": 774, + "AGAAAC": 775, + "AGAAAG": 776, + "AGAATA": 777, + "AGAATT": 778, + "AGAATC": 779, + "AGAATG": 780, + "AGAACA": 781, + "AGAACT": 782, + "AGAACC": 783, + "AGAACG": 784, + "AGAAGA": 785, + "AGAAGT": 786, + "AGAAGC": 787, + "AGAAGG": 788, + "AGATAA": 789, + "AGATAT": 790, + "AGATAC": 791, + "AGATAG": 792, + "AGATTA": 793, + "AGATTT": 794, + "AGATTC": 795, + "AGATTG": 796, + "AGATCA": 797, + "AGATCT": 798, + "AGATCC": 799, + "AGATCG": 800, + "AGATGA": 801, + "AGATGT": 802, + "AGATGC": 803, + "AGATGG": 804, + "AGACAA": 805, + "AGACAT": 806, + "AGACAC": 807, + "AGACAG": 808, + "AGACTA": 809, + "AGACTT": 810, + "AGACTC": 811, + "AGACTG": 812, + "AGACCA": 813, + "AGACCT": 814, + "AGACCC": 815, + "AGACCG": 816, + "AGACGA": 817, + "AGACGT": 818, + "AGACGC": 819, + "AGACGG": 820, + "AGAGAA": 821, + "AGAGAT": 822, + "AGAGAC": 823, + "AGAGAG": 824, + "AGAGTA": 825, + "AGAGTT": 826, + "AGAGTC": 827, + "AGAGTG": 828, + "AGAGCA": 829, + "AGAGCT": 830, + "AGAGCC": 831, + "AGAGCG": 832, + "AGAGGA": 833, + "AGAGGT": 834, + "AGAGGC": 835, + "AGAGGG": 836, + "AGTAAA": 837, + "AGTAAT": 838, + "AGTAAC": 839, + "AGTAAG": 840, + "AGTATA": 841, + "AGTATT": 842, + "AGTATC": 843, + "AGTATG": 844, + "AGTACA": 845, + "AGTACT": 846, + "AGTACC": 847, + "AGTACG": 848, + "AGTAGA": 849, + "AGTAGT": 850, + "AGTAGC": 851, + "AGTAGG": 852, + "AGTTAA": 853, + "AGTTAT": 854, + "AGTTAC": 855, + "AGTTAG": 856, + "AGTTTA": 857, + "AGTTTT": 858, + "AGTTTC": 859, + "AGTTTG": 860, + "AGTTCA": 861, + "AGTTCT": 862, + "AGTTCC": 863, + "AGTTCG": 864, + "AGTTGA": 865, + "AGTTGT": 866, + "AGTTGC": 867, + "AGTTGG": 868, + "AGTCAA": 869, + "AGTCAT": 870, + "AGTCAC": 871, + "AGTCAG": 872, + "AGTCTA": 873, + "AGTCTT": 874, + "AGTCTC": 875, + "AGTCTG": 876, + "AGTCCA": 877, + "AGTCCT": 878, + "AGTCCC": 879, + "AGTCCG": 880, + "AGTCGA": 881, + "AGTCGT": 882, + "AGTCGC": 883, + "AGTCGG": 884, + "AGTGAA": 885, + "AGTGAT": 886, + "AGTGAC": 887, + "AGTGAG": 888, + "AGTGTA": 889, + "AGTGTT": 890, + "AGTGTC": 891, + "AGTGTG": 892, + "AGTGCA": 893, + "AGTGCT": 894, + "AGTGCC": 895, + "AGTGCG": 896, + "AGTGGA": 897, + "AGTGGT": 898, + "AGTGGC": 899, + "AGTGGG": 900, + "AGCAAA": 901, + "AGCAAT": 902, + "AGCAAC": 903, + "AGCAAG": 904, + "AGCATA": 905, + "AGCATT": 906, + "AGCATC": 907, + "AGCATG": 908, + "AGCACA": 909, + "AGCACT": 910, + "AGCACC": 911, + "AGCACG": 912, + "AGCAGA": 913, + "AGCAGT": 914, + "AGCAGC": 915, + "AGCAGG": 916, + "AGCTAA": 917, + "AGCTAT": 918, + "AGCTAC": 919, + "AGCTAG": 920, + "AGCTTA": 921, + "AGCTTT": 922, + "AGCTTC": 923, + "AGCTTG": 924, + "AGCTCA": 925, + "AGCTCT": 926, + "AGCTCC": 927, + "AGCTCG": 928, + "AGCTGA": 929, + "AGCTGT": 930, + "AGCTGC": 931, + "AGCTGG": 932, + "AGCCAA": 933, + "AGCCAT": 934, + "AGCCAC": 935, + "AGCCAG": 936, + "AGCCTA": 937, + "AGCCTT": 938, + "AGCCTC": 939, + "AGCCTG": 940, + "AGCCCA": 941, + "AGCCCT": 942, + "AGCCCC": 943, + "AGCCCG": 944, + "AGCCGA": 945, + "AGCCGT": 946, + "AGCCGC": 947, + "AGCCGG": 948, + "AGCGAA": 949, + "AGCGAT": 950, + "AGCGAC": 951, + "AGCGAG": 952, + "AGCGTA": 953, + "AGCGTT": 954, + "AGCGTC": 955, + "AGCGTG": 956, + "AGCGCA": 957, + "AGCGCT": 958, + "AGCGCC": 959, + "AGCGCG": 960, + "AGCGGA": 961, + "AGCGGT": 962, + "AGCGGC": 963, + "AGCGGG": 964, + "AGGAAA": 965, + "AGGAAT": 966, + "AGGAAC": 967, + "AGGAAG": 968, + "AGGATA": 969, + "AGGATT": 970, + "AGGATC": 971, + "AGGATG": 972, + "AGGACA": 973, + "AGGACT": 974, + "AGGACC": 975, + "AGGACG": 976, + "AGGAGA": 977, + "AGGAGT": 978, + "AGGAGC": 979, + "AGGAGG": 980, + "AGGTAA": 981, + "AGGTAT": 982, + "AGGTAC": 983, + "AGGTAG": 984, + "AGGTTA": 985, + "AGGTTT": 986, + "AGGTTC": 987, + "AGGTTG": 988, + "AGGTCA": 989, + "AGGTCT": 990, + "AGGTCC": 991, + "AGGTCG": 992, + "AGGTGA": 993, + "AGGTGT": 994, + "AGGTGC": 995, + "AGGTGG": 996, + "AGGCAA": 997, + "AGGCAT": 998, + "AGGCAC": 999, + "AGGCAG": 1000, + "AGGCTA": 1001, + "AGGCTT": 1002, + "AGGCTC": 1003, + "AGGCTG": 1004, + "AGGCCA": 1005, + "AGGCCT": 1006, + "AGGCCC": 1007, + "AGGCCG": 1008, + "AGGCGA": 1009, + "AGGCGT": 1010, + "AGGCGC": 1011, + "AGGCGG": 1012, + "AGGGAA": 1013, + "AGGGAT": 1014, + "AGGGAC": 1015, + "AGGGAG": 1016, + "AGGGTA": 1017, + "AGGGTT": 1018, + "AGGGTC": 1019, + "AGGGTG": 1020, + "AGGGCA": 1021, + "AGGGCT": 1022, + "AGGGCC": 1023, + "AGGGCG": 1024, + "AGGGGA": 1025, + "AGGGGT": 1026, + "AGGGGC": 1027, + "AGGGGG": 1028, + "TAAAAA": 1029, + "TAAAAT": 1030, + "TAAAAC": 1031, + "TAAAAG": 1032, + "TAAATA": 1033, + "TAAATT": 1034, + "TAAATC": 1035, + "TAAATG": 1036, + "TAAACA": 1037, + "TAAACT": 1038, + "TAAACC": 1039, + "TAAACG": 1040, + "TAAAGA": 1041, + "TAAAGT": 1042, + "TAAAGC": 1043, + "TAAAGG": 1044, + "TAATAA": 1045, + "TAATAT": 1046, + "TAATAC": 1047, + "TAATAG": 1048, + "TAATTA": 1049, + "TAATTT": 1050, + "TAATTC": 1051, + "TAATTG": 1052, + "TAATCA": 1053, + "TAATCT": 1054, + "TAATCC": 1055, + "TAATCG": 1056, + "TAATGA": 1057, + "TAATGT": 1058, + "TAATGC": 1059, + "TAATGG": 1060, + "TAACAA": 1061, + "TAACAT": 1062, + "TAACAC": 1063, + "TAACAG": 1064, + "TAACTA": 1065, + "TAACTT": 1066, + "TAACTC": 1067, + "TAACTG": 1068, + "TAACCA": 1069, + "TAACCT": 1070, + "TAACCC": 1071, + "TAACCG": 1072, + "TAACGA": 1073, + "TAACGT": 1074, + "TAACGC": 1075, + "TAACGG": 1076, + "TAAGAA": 1077, + "TAAGAT": 1078, + "TAAGAC": 1079, + "TAAGAG": 1080, + "TAAGTA": 1081, + "TAAGTT": 1082, + "TAAGTC": 1083, + "TAAGTG": 1084, + "TAAGCA": 1085, + "TAAGCT": 1086, + "TAAGCC": 1087, + "TAAGCG": 1088, + "TAAGGA": 1089, + "TAAGGT": 1090, + "TAAGGC": 1091, + "TAAGGG": 1092, + "TATAAA": 1093, + "TATAAT": 1094, + "TATAAC": 1095, + "TATAAG": 1096, + "TATATA": 1097, + "TATATT": 1098, + "TATATC": 1099, + "TATATG": 1100, + "TATACA": 1101, + "TATACT": 1102, + "TATACC": 1103, + "TATACG": 1104, + "TATAGA": 1105, + "TATAGT": 1106, + "TATAGC": 1107, + "TATAGG": 1108, + "TATTAA": 1109, + "TATTAT": 1110, + "TATTAC": 1111, + "TATTAG": 1112, + "TATTTA": 1113, + "TATTTT": 1114, + "TATTTC": 1115, + "TATTTG": 1116, + "TATTCA": 1117, + "TATTCT": 1118, + "TATTCC": 1119, + "TATTCG": 1120, + "TATTGA": 1121, + "TATTGT": 1122, + "TATTGC": 1123, + "TATTGG": 1124, + "TATCAA": 1125, + "TATCAT": 1126, + "TATCAC": 1127, + "TATCAG": 1128, + "TATCTA": 1129, + "TATCTT": 1130, + "TATCTC": 1131, + "TATCTG": 1132, + "TATCCA": 1133, + "TATCCT": 1134, + "TATCCC": 1135, + "TATCCG": 1136, + "TATCGA": 1137, + "TATCGT": 1138, + "TATCGC": 1139, + "TATCGG": 1140, + "TATGAA": 1141, + "TATGAT": 1142, + "TATGAC": 1143, + "TATGAG": 1144, + "TATGTA": 1145, + "TATGTT": 1146, + "TATGTC": 1147, + "TATGTG": 1148, + "TATGCA": 1149, + "TATGCT": 1150, + "TATGCC": 1151, + "TATGCG": 1152, + "TATGGA": 1153, + "TATGGT": 1154, + "TATGGC": 1155, + "TATGGG": 1156, + "TACAAA": 1157, + "TACAAT": 1158, + "TACAAC": 1159, + "TACAAG": 1160, + "TACATA": 1161, + "TACATT": 1162, + "TACATC": 1163, + "TACATG": 1164, + "TACACA": 1165, + "TACACT": 1166, + "TACACC": 1167, + "TACACG": 1168, + "TACAGA": 1169, + "TACAGT": 1170, + "TACAGC": 1171, + "TACAGG": 1172, + "TACTAA": 1173, + "TACTAT": 1174, + "TACTAC": 1175, + "TACTAG": 1176, + "TACTTA": 1177, + "TACTTT": 1178, + "TACTTC": 1179, + "TACTTG": 1180, + "TACTCA": 1181, + "TACTCT": 1182, + "TACTCC": 1183, + "TACTCG": 1184, + "TACTGA": 1185, + "TACTGT": 1186, + "TACTGC": 1187, + "TACTGG": 1188, + "TACCAA": 1189, + "TACCAT": 1190, + "TACCAC": 1191, + "TACCAG": 1192, + "TACCTA": 1193, + "TACCTT": 1194, + "TACCTC": 1195, + "TACCTG": 1196, + "TACCCA": 1197, + "TACCCT": 1198, + "TACCCC": 1199, + "TACCCG": 1200, + "TACCGA": 1201, + "TACCGT": 1202, + "TACCGC": 1203, + "TACCGG": 1204, + "TACGAA": 1205, + "TACGAT": 1206, + "TACGAC": 1207, + "TACGAG": 1208, + "TACGTA": 1209, + "TACGTT": 1210, + "TACGTC": 1211, + "TACGTG": 1212, + "TACGCA": 1213, + "TACGCT": 1214, + "TACGCC": 1215, + "TACGCG": 1216, + "TACGGA": 1217, + "TACGGT": 1218, + "TACGGC": 1219, + "TACGGG": 1220, + "TAGAAA": 1221, + "TAGAAT": 1222, + "TAGAAC": 1223, + "TAGAAG": 1224, + "TAGATA": 1225, + "TAGATT": 1226, + "TAGATC": 1227, + "TAGATG": 1228, + "TAGACA": 1229, + "TAGACT": 1230, + "TAGACC": 1231, + "TAGACG": 1232, + "TAGAGA": 1233, + "TAGAGT": 1234, + "TAGAGC": 1235, + "TAGAGG": 1236, + "TAGTAA": 1237, + "TAGTAT": 1238, + "TAGTAC": 1239, + "TAGTAG": 1240, + "TAGTTA": 1241, + "TAGTTT": 1242, + "TAGTTC": 1243, + "TAGTTG": 1244, + "TAGTCA": 1245, + "TAGTCT": 1246, + "TAGTCC": 1247, + "TAGTCG": 1248, + "TAGTGA": 1249, + "TAGTGT": 1250, + "TAGTGC": 1251, + "TAGTGG": 1252, + "TAGCAA": 1253, + "TAGCAT": 1254, + "TAGCAC": 1255, + "TAGCAG": 1256, + "TAGCTA": 1257, + "TAGCTT": 1258, + "TAGCTC": 1259, + "TAGCTG": 1260, + "TAGCCA": 1261, + "TAGCCT": 1262, + "TAGCCC": 1263, + "TAGCCG": 1264, + "TAGCGA": 1265, + "TAGCGT": 1266, + "TAGCGC": 1267, + "TAGCGG": 1268, + "TAGGAA": 1269, + "TAGGAT": 1270, + "TAGGAC": 1271, + "TAGGAG": 1272, + "TAGGTA": 1273, + "TAGGTT": 1274, + "TAGGTC": 1275, + "TAGGTG": 1276, + "TAGGCA": 1277, + "TAGGCT": 1278, + "TAGGCC": 1279, + "TAGGCG": 1280, + "TAGGGA": 1281, + "TAGGGT": 1282, + "TAGGGC": 1283, + "TAGGGG": 1284, + "TTAAAA": 1285, + "TTAAAT": 1286, + "TTAAAC": 1287, + "TTAAAG": 1288, + "TTAATA": 1289, + "TTAATT": 1290, + "TTAATC": 1291, + "TTAATG": 1292, + "TTAACA": 1293, + "TTAACT": 1294, + "TTAACC": 1295, + "TTAACG": 1296, + "TTAAGA": 1297, + "TTAAGT": 1298, + "TTAAGC": 1299, + "TTAAGG": 1300, + "TTATAA": 1301, + "TTATAT": 1302, + "TTATAC": 1303, + "TTATAG": 1304, + "TTATTA": 1305, + "TTATTT": 1306, + "TTATTC": 1307, + "TTATTG": 1308, + "TTATCA": 1309, + "TTATCT": 1310, + "TTATCC": 1311, + "TTATCG": 1312, + "TTATGA": 1313, + "TTATGT": 1314, + "TTATGC": 1315, + "TTATGG": 1316, + "TTACAA": 1317, + "TTACAT": 1318, + "TTACAC": 1319, + "TTACAG": 1320, + "TTACTA": 1321, + "TTACTT": 1322, + "TTACTC": 1323, + "TTACTG": 1324, + "TTACCA": 1325, + "TTACCT": 1326, + "TTACCC": 1327, + "TTACCG": 1328, + "TTACGA": 1329, + "TTACGT": 1330, + "TTACGC": 1331, + "TTACGG": 1332, + "TTAGAA": 1333, + "TTAGAT": 1334, + "TTAGAC": 1335, + "TTAGAG": 1336, + "TTAGTA": 1337, + "TTAGTT": 1338, + "TTAGTC": 1339, + "TTAGTG": 1340, + "TTAGCA": 1341, + "TTAGCT": 1342, + "TTAGCC": 1343, + "TTAGCG": 1344, + "TTAGGA": 1345, + "TTAGGT": 1346, + "TTAGGC": 1347, + "TTAGGG": 1348, + "TTTAAA": 1349, + "TTTAAT": 1350, + "TTTAAC": 1351, + "TTTAAG": 1352, + "TTTATA": 1353, + "TTTATT": 1354, + "TTTATC": 1355, + "TTTATG": 1356, + "TTTACA": 1357, + "TTTACT": 1358, + "TTTACC": 1359, + "TTTACG": 1360, + "TTTAGA": 1361, + "TTTAGT": 1362, + "TTTAGC": 1363, + "TTTAGG": 1364, + "TTTTAA": 1365, + "TTTTAT": 1366, + "TTTTAC": 1367, + "TTTTAG": 1368, + "TTTTTA": 1369, + "TTTTTT": 1370, + "TTTTTC": 1371, + "TTTTTG": 1372, + "TTTTCA": 1373, + "TTTTCT": 1374, + "TTTTCC": 1375, + "TTTTCG": 1376, + "TTTTGA": 1377, + "TTTTGT": 1378, + "TTTTGC": 1379, + "TTTTGG": 1380, + "TTTCAA": 1381, + "TTTCAT": 1382, + "TTTCAC": 1383, + "TTTCAG": 1384, + "TTTCTA": 1385, + "TTTCTT": 1386, + "TTTCTC": 1387, + "TTTCTG": 1388, + "TTTCCA": 1389, + "TTTCCT": 1390, + "TTTCCC": 1391, + "TTTCCG": 1392, + "TTTCGA": 1393, + "TTTCGT": 1394, + "TTTCGC": 1395, + "TTTCGG": 1396, + "TTTGAA": 1397, + "TTTGAT": 1398, + "TTTGAC": 1399, + "TTTGAG": 1400, + "TTTGTA": 1401, + "TTTGTT": 1402, + "TTTGTC": 1403, + "TTTGTG": 1404, + "TTTGCA": 1405, + "TTTGCT": 1406, + "TTTGCC": 1407, + "TTTGCG": 1408, + "TTTGGA": 1409, + "TTTGGT": 1410, + "TTTGGC": 1411, + "TTTGGG": 1412, + "TTCAAA": 1413, + "TTCAAT": 1414, + "TTCAAC": 1415, + "TTCAAG": 1416, + "TTCATA": 1417, + "TTCATT": 1418, + "TTCATC": 1419, + "TTCATG": 1420, + "TTCACA": 1421, + "TTCACT": 1422, + "TTCACC": 1423, + "TTCACG": 1424, + "TTCAGA": 1425, + "TTCAGT": 1426, + "TTCAGC": 1427, + "TTCAGG": 1428, + "TTCTAA": 1429, + "TTCTAT": 1430, + "TTCTAC": 1431, + "TTCTAG": 1432, + "TTCTTA": 1433, + "TTCTTT": 1434, + "TTCTTC": 1435, + "TTCTTG": 1436, + "TTCTCA": 1437, + "TTCTCT": 1438, + "TTCTCC": 1439, + "TTCTCG": 1440, + "TTCTGA": 1441, + "TTCTGT": 1442, + "TTCTGC": 1443, + "TTCTGG": 1444, + "TTCCAA": 1445, + "TTCCAT": 1446, + "TTCCAC": 1447, + "TTCCAG": 1448, + "TTCCTA": 1449, + "TTCCTT": 1450, + "TTCCTC": 1451, + "TTCCTG": 1452, + "TTCCCA": 1453, + "TTCCCT": 1454, + "TTCCCC": 1455, + "TTCCCG": 1456, + "TTCCGA": 1457, + "TTCCGT": 1458, + "TTCCGC": 1459, + "TTCCGG": 1460, + "TTCGAA": 1461, + "TTCGAT": 1462, + "TTCGAC": 1463, + "TTCGAG": 1464, + "TTCGTA": 1465, + "TTCGTT": 1466, + "TTCGTC": 1467, + "TTCGTG": 1468, + "TTCGCA": 1469, + "TTCGCT": 1470, + "TTCGCC": 1471, + "TTCGCG": 1472, + "TTCGGA": 1473, + "TTCGGT": 1474, + "TTCGGC": 1475, + "TTCGGG": 1476, + "TTGAAA": 1477, + "TTGAAT": 1478, + "TTGAAC": 1479, + "TTGAAG": 1480, + "TTGATA": 1481, + "TTGATT": 1482, + "TTGATC": 1483, + "TTGATG": 1484, + "TTGACA": 1485, + "TTGACT": 1486, + "TTGACC": 1487, + "TTGACG": 1488, + "TTGAGA": 1489, + "TTGAGT": 1490, + "TTGAGC": 1491, + "TTGAGG": 1492, + "TTGTAA": 1493, + "TTGTAT": 1494, + "TTGTAC": 1495, + "TTGTAG": 1496, + "TTGTTA": 1497, + "TTGTTT": 1498, + "TTGTTC": 1499, + "TTGTTG": 1500, + "TTGTCA": 1501, + "TTGTCT": 1502, + "TTGTCC": 1503, + "TTGTCG": 1504, + "TTGTGA": 1505, + "TTGTGT": 1506, + "TTGTGC": 1507, + "TTGTGG": 1508, + "TTGCAA": 1509, + "TTGCAT": 1510, + "TTGCAC": 1511, + "TTGCAG": 1512, + "TTGCTA": 1513, + "TTGCTT": 1514, + "TTGCTC": 1515, + "TTGCTG": 1516, + "TTGCCA": 1517, + "TTGCCT": 1518, + "TTGCCC": 1519, + "TTGCCG": 1520, + "TTGCGA": 1521, + "TTGCGT": 1522, + "TTGCGC": 1523, + "TTGCGG": 1524, + "TTGGAA": 1525, + "TTGGAT": 1526, + "TTGGAC": 1527, + "TTGGAG": 1528, + "TTGGTA": 1529, + "TTGGTT": 1530, + "TTGGTC": 1531, + "TTGGTG": 1532, + "TTGGCA": 1533, + "TTGGCT": 1534, + "TTGGCC": 1535, + "TTGGCG": 1536, + "TTGGGA": 1537, + "TTGGGT": 1538, + "TTGGGC": 1539, + "TTGGGG": 1540, + "TCAAAA": 1541, + "TCAAAT": 1542, + "TCAAAC": 1543, + "TCAAAG": 1544, + "TCAATA": 1545, + "TCAATT": 1546, + "TCAATC": 1547, + "TCAATG": 1548, + "TCAACA": 1549, + "TCAACT": 1550, + "TCAACC": 1551, + "TCAACG": 1552, + "TCAAGA": 1553, + "TCAAGT": 1554, + "TCAAGC": 1555, + "TCAAGG": 1556, + "TCATAA": 1557, + "TCATAT": 1558, + "TCATAC": 1559, + "TCATAG": 1560, + "TCATTA": 1561, + "TCATTT": 1562, + "TCATTC": 1563, + "TCATTG": 1564, + "TCATCA": 1565, + "TCATCT": 1566, + "TCATCC": 1567, + "TCATCG": 1568, + "TCATGA": 1569, + "TCATGT": 1570, + "TCATGC": 1571, + "TCATGG": 1572, + "TCACAA": 1573, + "TCACAT": 1574, + "TCACAC": 1575, + "TCACAG": 1576, + "TCACTA": 1577, + "TCACTT": 1578, + "TCACTC": 1579, + "TCACTG": 1580, + "TCACCA": 1581, + "TCACCT": 1582, + "TCACCC": 1583, + "TCACCG": 1584, + "TCACGA": 1585, + "TCACGT": 1586, + "TCACGC": 1587, + "TCACGG": 1588, + "TCAGAA": 1589, + "TCAGAT": 1590, + "TCAGAC": 1591, + "TCAGAG": 1592, + "TCAGTA": 1593, + "TCAGTT": 1594, + "TCAGTC": 1595, + "TCAGTG": 1596, + "TCAGCA": 1597, + "TCAGCT": 1598, + "TCAGCC": 1599, + "TCAGCG": 1600, + "TCAGGA": 1601, + "TCAGGT": 1602, + "TCAGGC": 1603, + "TCAGGG": 1604, + "TCTAAA": 1605, + "TCTAAT": 1606, + "TCTAAC": 1607, + "TCTAAG": 1608, + "TCTATA": 1609, + "TCTATT": 1610, + "TCTATC": 1611, + "TCTATG": 1612, + "TCTACA": 1613, + "TCTACT": 1614, + "TCTACC": 1615, + "TCTACG": 1616, + "TCTAGA": 1617, + "TCTAGT": 1618, + "TCTAGC": 1619, + "TCTAGG": 1620, + "TCTTAA": 1621, + "TCTTAT": 1622, + "TCTTAC": 1623, + "TCTTAG": 1624, + "TCTTTA": 1625, + "TCTTTT": 1626, + "TCTTTC": 1627, + "TCTTTG": 1628, + "TCTTCA": 1629, + "TCTTCT": 1630, + "TCTTCC": 1631, + "TCTTCG": 1632, + "TCTTGA": 1633, + "TCTTGT": 1634, + "TCTTGC": 1635, + "TCTTGG": 1636, + "TCTCAA": 1637, + "TCTCAT": 1638, + "TCTCAC": 1639, + "TCTCAG": 1640, + "TCTCTA": 1641, + "TCTCTT": 1642, + "TCTCTC": 1643, + "TCTCTG": 1644, + "TCTCCA": 1645, + "TCTCCT": 1646, + "TCTCCC": 1647, + "TCTCCG": 1648, + "TCTCGA": 1649, + "TCTCGT": 1650, + "TCTCGC": 1651, + "TCTCGG": 1652, + "TCTGAA": 1653, + "TCTGAT": 1654, + "TCTGAC": 1655, + "TCTGAG": 1656, + "TCTGTA": 1657, + "TCTGTT": 1658, + "TCTGTC": 1659, + "TCTGTG": 1660, + "TCTGCA": 1661, + "TCTGCT": 1662, + "TCTGCC": 1663, + "TCTGCG": 1664, + "TCTGGA": 1665, + "TCTGGT": 1666, + "TCTGGC": 1667, + "TCTGGG": 1668, + "TCCAAA": 1669, + "TCCAAT": 1670, + "TCCAAC": 1671, + "TCCAAG": 1672, + "TCCATA": 1673, + "TCCATT": 1674, + "TCCATC": 1675, + "TCCATG": 1676, + "TCCACA": 1677, + "TCCACT": 1678, + "TCCACC": 1679, + "TCCACG": 1680, + "TCCAGA": 1681, + "TCCAGT": 1682, + "TCCAGC": 1683, + "TCCAGG": 1684, + "TCCTAA": 1685, + "TCCTAT": 1686, + "TCCTAC": 1687, + "TCCTAG": 1688, + "TCCTTA": 1689, + "TCCTTT": 1690, + "TCCTTC": 1691, + "TCCTTG": 1692, + "TCCTCA": 1693, + "TCCTCT": 1694, + "TCCTCC": 1695, + "TCCTCG": 1696, + "TCCTGA": 1697, + "TCCTGT": 1698, + "TCCTGC": 1699, + "TCCTGG": 1700, + "TCCCAA": 1701, + "TCCCAT": 1702, + "TCCCAC": 1703, + "TCCCAG": 1704, + "TCCCTA": 1705, + "TCCCTT": 1706, + "TCCCTC": 1707, + "TCCCTG": 1708, + "TCCCCA": 1709, + "TCCCCT": 1710, + "TCCCCC": 1711, + "TCCCCG": 1712, + "TCCCGA": 1713, + "TCCCGT": 1714, + "TCCCGC": 1715, + "TCCCGG": 1716, + "TCCGAA": 1717, + "TCCGAT": 1718, + "TCCGAC": 1719, + "TCCGAG": 1720, + "TCCGTA": 1721, + "TCCGTT": 1722, + "TCCGTC": 1723, + "TCCGTG": 1724, + "TCCGCA": 1725, + "TCCGCT": 1726, + "TCCGCC": 1727, + "TCCGCG": 1728, + "TCCGGA": 1729, + "TCCGGT": 1730, + "TCCGGC": 1731, + "TCCGGG": 1732, + "TCGAAA": 1733, + "TCGAAT": 1734, + "TCGAAC": 1735, + "TCGAAG": 1736, + "TCGATA": 1737, + "TCGATT": 1738, + "TCGATC": 1739, + "TCGATG": 1740, + "TCGACA": 1741, + "TCGACT": 1742, + "TCGACC": 1743, + "TCGACG": 1744, + "TCGAGA": 1745, + "TCGAGT": 1746, + "TCGAGC": 1747, + "TCGAGG": 1748, + "TCGTAA": 1749, + "TCGTAT": 1750, + "TCGTAC": 1751, + "TCGTAG": 1752, + "TCGTTA": 1753, + "TCGTTT": 1754, + "TCGTTC": 1755, + "TCGTTG": 1756, + "TCGTCA": 1757, + "TCGTCT": 1758, + "TCGTCC": 1759, + "TCGTCG": 1760, + "TCGTGA": 1761, + "TCGTGT": 1762, + "TCGTGC": 1763, + "TCGTGG": 1764, + "TCGCAA": 1765, + "TCGCAT": 1766, + "TCGCAC": 1767, + "TCGCAG": 1768, + "TCGCTA": 1769, + "TCGCTT": 1770, + "TCGCTC": 1771, + "TCGCTG": 1772, + "TCGCCA": 1773, + "TCGCCT": 1774, + "TCGCCC": 1775, + "TCGCCG": 1776, + "TCGCGA": 1777, + "TCGCGT": 1778, + "TCGCGC": 1779, + "TCGCGG": 1780, + "TCGGAA": 1781, + "TCGGAT": 1782, + "TCGGAC": 1783, + "TCGGAG": 1784, + "TCGGTA": 1785, + "TCGGTT": 1786, + "TCGGTC": 1787, + "TCGGTG": 1788, + "TCGGCA": 1789, + "TCGGCT": 1790, + "TCGGCC": 1791, + "TCGGCG": 1792, + "TCGGGA": 1793, + "TCGGGT": 1794, + "TCGGGC": 1795, + "TCGGGG": 1796, + "TGAAAA": 1797, + "TGAAAT": 1798, + "TGAAAC": 1799, + "TGAAAG": 1800, + "TGAATA": 1801, + "TGAATT": 1802, + "TGAATC": 1803, + "TGAATG": 1804, + "TGAACA": 1805, + "TGAACT": 1806, + "TGAACC": 1807, + "TGAACG": 1808, + "TGAAGA": 1809, + "TGAAGT": 1810, + "TGAAGC": 1811, + "TGAAGG": 1812, + "TGATAA": 1813, + "TGATAT": 1814, + "TGATAC": 1815, + "TGATAG": 1816, + "TGATTA": 1817, + "TGATTT": 1818, + "TGATTC": 1819, + "TGATTG": 1820, + "TGATCA": 1821, + "TGATCT": 1822, + "TGATCC": 1823, + "TGATCG": 1824, + "TGATGA": 1825, + "TGATGT": 1826, + "TGATGC": 1827, + "TGATGG": 1828, + "TGACAA": 1829, + "TGACAT": 1830, + "TGACAC": 1831, + "TGACAG": 1832, + "TGACTA": 1833, + "TGACTT": 1834, + "TGACTC": 1835, + "TGACTG": 1836, + "TGACCA": 1837, + "TGACCT": 1838, + "TGACCC": 1839, + "TGACCG": 1840, + "TGACGA": 1841, + "TGACGT": 1842, + "TGACGC": 1843, + "TGACGG": 1844, + "TGAGAA": 1845, + "TGAGAT": 1846, + "TGAGAC": 1847, + "TGAGAG": 1848, + "TGAGTA": 1849, + "TGAGTT": 1850, + "TGAGTC": 1851, + "TGAGTG": 1852, + "TGAGCA": 1853, + "TGAGCT": 1854, + "TGAGCC": 1855, + "TGAGCG": 1856, + "TGAGGA": 1857, + "TGAGGT": 1858, + "TGAGGC": 1859, + "TGAGGG": 1860, + "TGTAAA": 1861, + "TGTAAT": 1862, + "TGTAAC": 1863, + "TGTAAG": 1864, + "TGTATA": 1865, + "TGTATT": 1866, + "TGTATC": 1867, + "TGTATG": 1868, + "TGTACA": 1869, + "TGTACT": 1870, + "TGTACC": 1871, + "TGTACG": 1872, + "TGTAGA": 1873, + "TGTAGT": 1874, + "TGTAGC": 1875, + "TGTAGG": 1876, + "TGTTAA": 1877, + "TGTTAT": 1878, + "TGTTAC": 1879, + "TGTTAG": 1880, + "TGTTTA": 1881, + "TGTTTT": 1882, + "TGTTTC": 1883, + "TGTTTG": 1884, + "TGTTCA": 1885, + "TGTTCT": 1886, + "TGTTCC": 1887, + "TGTTCG": 1888, + "TGTTGA": 1889, + "TGTTGT": 1890, + "TGTTGC": 1891, + "TGTTGG": 1892, + "TGTCAA": 1893, + "TGTCAT": 1894, + "TGTCAC": 1895, + "TGTCAG": 1896, + "TGTCTA": 1897, + "TGTCTT": 1898, + "TGTCTC": 1899, + "TGTCTG": 1900, + "TGTCCA": 1901, + "TGTCCT": 1902, + "TGTCCC": 1903, + "TGTCCG": 1904, + "TGTCGA": 1905, + "TGTCGT": 1906, + "TGTCGC": 1907, + "TGTCGG": 1908, + "TGTGAA": 1909, + "TGTGAT": 1910, + "TGTGAC": 1911, + "TGTGAG": 1912, + "TGTGTA": 1913, + "TGTGTT": 1914, + "TGTGTC": 1915, + "TGTGTG": 1916, + "TGTGCA": 1917, + "TGTGCT": 1918, + "TGTGCC": 1919, + "TGTGCG": 1920, + "TGTGGA": 1921, + "TGTGGT": 1922, + "TGTGGC": 1923, + "TGTGGG": 1924, + "TGCAAA": 1925, + "TGCAAT": 1926, + "TGCAAC": 1927, + "TGCAAG": 1928, + "TGCATA": 1929, + "TGCATT": 1930, + "TGCATC": 1931, + "TGCATG": 1932, + "TGCACA": 1933, + "TGCACT": 1934, + "TGCACC": 1935, + "TGCACG": 1936, + "TGCAGA": 1937, + "TGCAGT": 1938, + "TGCAGC": 1939, + "TGCAGG": 1940, + "TGCTAA": 1941, + "TGCTAT": 1942, + "TGCTAC": 1943, + "TGCTAG": 1944, + "TGCTTA": 1945, + "TGCTTT": 1946, + "TGCTTC": 1947, + "TGCTTG": 1948, + "TGCTCA": 1949, + "TGCTCT": 1950, + "TGCTCC": 1951, + "TGCTCG": 1952, + "TGCTGA": 1953, + "TGCTGT": 1954, + "TGCTGC": 1955, + "TGCTGG": 1956, + "TGCCAA": 1957, + "TGCCAT": 1958, + "TGCCAC": 1959, + "TGCCAG": 1960, + "TGCCTA": 1961, + "TGCCTT": 1962, + "TGCCTC": 1963, + "TGCCTG": 1964, + "TGCCCA": 1965, + "TGCCCT": 1966, + "TGCCCC": 1967, + "TGCCCG": 1968, + "TGCCGA": 1969, + "TGCCGT": 1970, + "TGCCGC": 1971, + "TGCCGG": 1972, + "TGCGAA": 1973, + "TGCGAT": 1974, + "TGCGAC": 1975, + "TGCGAG": 1976, + "TGCGTA": 1977, + "TGCGTT": 1978, + "TGCGTC": 1979, + "TGCGTG": 1980, + "TGCGCA": 1981, + "TGCGCT": 1982, + "TGCGCC": 1983, + "TGCGCG": 1984, + "TGCGGA": 1985, + "TGCGGT": 1986, + "TGCGGC": 1987, + "TGCGGG": 1988, + "TGGAAA": 1989, + "TGGAAT": 1990, + "TGGAAC": 1991, + "TGGAAG": 1992, + "TGGATA": 1993, + "TGGATT": 1994, + "TGGATC": 1995, + "TGGATG": 1996, + "TGGACA": 1997, + "TGGACT": 1998, + "TGGACC": 1999, + "TGGACG": 2000, + "TGGAGA": 2001, + "TGGAGT": 2002, + "TGGAGC": 2003, + "TGGAGG": 2004, + "TGGTAA": 2005, + "TGGTAT": 2006, + "TGGTAC": 2007, + "TGGTAG": 2008, + "TGGTTA": 2009, + "TGGTTT": 2010, + "TGGTTC": 2011, + "TGGTTG": 2012, + "TGGTCA": 2013, + "TGGTCT": 2014, + "TGGTCC": 2015, + "TGGTCG": 2016, + "TGGTGA": 2017, + "TGGTGT": 2018, + "TGGTGC": 2019, + "TGGTGG": 2020, + "TGGCAA": 2021, + "TGGCAT": 2022, + "TGGCAC": 2023, + "TGGCAG": 2024, + "TGGCTA": 2025, + "TGGCTT": 2026, + "TGGCTC": 2027, + "TGGCTG": 2028, + "TGGCCA": 2029, + "TGGCCT": 2030, + "TGGCCC": 2031, + "TGGCCG": 2032, + "TGGCGA": 2033, + "TGGCGT": 2034, + "TGGCGC": 2035, + "TGGCGG": 2036, + "TGGGAA": 2037, + "TGGGAT": 2038, + "TGGGAC": 2039, + "TGGGAG": 2040, + "TGGGTA": 2041, + "TGGGTT": 2042, + "TGGGTC": 2043, + "TGGGTG": 2044, + "TGGGCA": 2045, + "TGGGCT": 2046, + "TGGGCC": 2047, + "TGGGCG": 2048, + "TGGGGA": 2049, + "TGGGGT": 2050, + "TGGGGC": 2051, + "TGGGGG": 2052, + "CAAAAA": 2053, + "CAAAAT": 2054, + "CAAAAC": 2055, + "CAAAAG": 2056, + "CAAATA": 2057, + "CAAATT": 2058, + "CAAATC": 2059, + "CAAATG": 2060, + "CAAACA": 2061, + "CAAACT": 2062, + "CAAACC": 2063, + "CAAACG": 2064, + "CAAAGA": 2065, + "CAAAGT": 2066, + "CAAAGC": 2067, + "CAAAGG": 2068, + "CAATAA": 2069, + "CAATAT": 2070, + "CAATAC": 2071, + "CAATAG": 2072, + "CAATTA": 2073, + "CAATTT": 2074, + "CAATTC": 2075, + "CAATTG": 2076, + "CAATCA": 2077, + "CAATCT": 2078, + "CAATCC": 2079, + "CAATCG": 2080, + "CAATGA": 2081, + "CAATGT": 2082, + "CAATGC": 2083, + "CAATGG": 2084, + "CAACAA": 2085, + "CAACAT": 2086, + "CAACAC": 2087, + "CAACAG": 2088, + "CAACTA": 2089, + "CAACTT": 2090, + "CAACTC": 2091, + "CAACTG": 2092, + "CAACCA": 2093, + "CAACCT": 2094, + "CAACCC": 2095, + "CAACCG": 2096, + "CAACGA": 2097, + "CAACGT": 2098, + "CAACGC": 2099, + "CAACGG": 2100, + "CAAGAA": 2101, + "CAAGAT": 2102, + "CAAGAC": 2103, + "CAAGAG": 2104, + "CAAGTA": 2105, + "CAAGTT": 2106, + "CAAGTC": 2107, + "CAAGTG": 2108, + "CAAGCA": 2109, + "CAAGCT": 2110, + "CAAGCC": 2111, + "CAAGCG": 2112, + "CAAGGA": 2113, + "CAAGGT": 2114, + "CAAGGC": 2115, + "CAAGGG": 2116, + "CATAAA": 2117, + "CATAAT": 2118, + "CATAAC": 2119, + "CATAAG": 2120, + "CATATA": 2121, + "CATATT": 2122, + "CATATC": 2123, + "CATATG": 2124, + "CATACA": 2125, + "CATACT": 2126, + "CATACC": 2127, + "CATACG": 2128, + "CATAGA": 2129, + "CATAGT": 2130, + "CATAGC": 2131, + "CATAGG": 2132, + "CATTAA": 2133, + "CATTAT": 2134, + "CATTAC": 2135, + "CATTAG": 2136, + "CATTTA": 2137, + "CATTTT": 2138, + "CATTTC": 2139, + "CATTTG": 2140, + "CATTCA": 2141, + "CATTCT": 2142, + "CATTCC": 2143, + "CATTCG": 2144, + "CATTGA": 2145, + "CATTGT": 2146, + "CATTGC": 2147, + "CATTGG": 2148, + "CATCAA": 2149, + "CATCAT": 2150, + "CATCAC": 2151, + "CATCAG": 2152, + "CATCTA": 2153, + "CATCTT": 2154, + "CATCTC": 2155, + "CATCTG": 2156, + "CATCCA": 2157, + "CATCCT": 2158, + "CATCCC": 2159, + "CATCCG": 2160, + "CATCGA": 2161, + "CATCGT": 2162, + "CATCGC": 2163, + "CATCGG": 2164, + "CATGAA": 2165, + "CATGAT": 2166, + "CATGAC": 2167, + "CATGAG": 2168, + "CATGTA": 2169, + "CATGTT": 2170, + "CATGTC": 2171, + "CATGTG": 2172, + "CATGCA": 2173, + "CATGCT": 2174, + "CATGCC": 2175, + "CATGCG": 2176, + "CATGGA": 2177, + "CATGGT": 2178, + "CATGGC": 2179, + "CATGGG": 2180, + "CACAAA": 2181, + "CACAAT": 2182, + "CACAAC": 2183, + "CACAAG": 2184, + "CACATA": 2185, + "CACATT": 2186, + "CACATC": 2187, + "CACATG": 2188, + "CACACA": 2189, + "CACACT": 2190, + "CACACC": 2191, + "CACACG": 2192, + "CACAGA": 2193, + "CACAGT": 2194, + "CACAGC": 2195, + "CACAGG": 2196, + "CACTAA": 2197, + "CACTAT": 2198, + "CACTAC": 2199, + "CACTAG": 2200, + "CACTTA": 2201, + "CACTTT": 2202, + "CACTTC": 2203, + "CACTTG": 2204, + "CACTCA": 2205, + "CACTCT": 2206, + "CACTCC": 2207, + "CACTCG": 2208, + "CACTGA": 2209, + "CACTGT": 2210, + "CACTGC": 2211, + "CACTGG": 2212, + "CACCAA": 2213, + "CACCAT": 2214, + "CACCAC": 2215, + "CACCAG": 2216, + "CACCTA": 2217, + "CACCTT": 2218, + "CACCTC": 2219, + "CACCTG": 2220, + "CACCCA": 2221, + "CACCCT": 2222, + "CACCCC": 2223, + "CACCCG": 2224, + "CACCGA": 2225, + "CACCGT": 2226, + "CACCGC": 2227, + "CACCGG": 2228, + "CACGAA": 2229, + "CACGAT": 2230, + "CACGAC": 2231, + "CACGAG": 2232, + "CACGTA": 2233, + "CACGTT": 2234, + "CACGTC": 2235, + "CACGTG": 2236, + "CACGCA": 2237, + "CACGCT": 2238, + "CACGCC": 2239, + "CACGCG": 2240, + "CACGGA": 2241, + "CACGGT": 2242, + "CACGGC": 2243, + "CACGGG": 2244, + "CAGAAA": 2245, + "CAGAAT": 2246, + "CAGAAC": 2247, + "CAGAAG": 2248, + "CAGATA": 2249, + "CAGATT": 2250, + "CAGATC": 2251, + "CAGATG": 2252, + "CAGACA": 2253, + "CAGACT": 2254, + "CAGACC": 2255, + "CAGACG": 2256, + "CAGAGA": 2257, + "CAGAGT": 2258, + "CAGAGC": 2259, + "CAGAGG": 2260, + "CAGTAA": 2261, + "CAGTAT": 2262, + "CAGTAC": 2263, + "CAGTAG": 2264, + "CAGTTA": 2265, + "CAGTTT": 2266, + "CAGTTC": 2267, + "CAGTTG": 2268, + "CAGTCA": 2269, + "CAGTCT": 2270, + "CAGTCC": 2271, + "CAGTCG": 2272, + "CAGTGA": 2273, + "CAGTGT": 2274, + "CAGTGC": 2275, + "CAGTGG": 2276, + "CAGCAA": 2277, + "CAGCAT": 2278, + "CAGCAC": 2279, + "CAGCAG": 2280, + "CAGCTA": 2281, + "CAGCTT": 2282, + "CAGCTC": 2283, + "CAGCTG": 2284, + "CAGCCA": 2285, + "CAGCCT": 2286, + "CAGCCC": 2287, + "CAGCCG": 2288, + "CAGCGA": 2289, + "CAGCGT": 2290, + "CAGCGC": 2291, + "CAGCGG": 2292, + "CAGGAA": 2293, + "CAGGAT": 2294, + "CAGGAC": 2295, + "CAGGAG": 2296, + "CAGGTA": 2297, + "CAGGTT": 2298, + "CAGGTC": 2299, + "CAGGTG": 2300, + "CAGGCA": 2301, + "CAGGCT": 2302, + "CAGGCC": 2303, + "CAGGCG": 2304, + "CAGGGA": 2305, + "CAGGGT": 2306, + "CAGGGC": 2307, + "CAGGGG": 2308, + "CTAAAA": 2309, + "CTAAAT": 2310, + "CTAAAC": 2311, + "CTAAAG": 2312, + "CTAATA": 2313, + "CTAATT": 2314, + "CTAATC": 2315, + "CTAATG": 2316, + "CTAACA": 2317, + "CTAACT": 2318, + "CTAACC": 2319, + "CTAACG": 2320, + "CTAAGA": 2321, + "CTAAGT": 2322, + "CTAAGC": 2323, + "CTAAGG": 2324, + "CTATAA": 2325, + "CTATAT": 2326, + "CTATAC": 2327, + "CTATAG": 2328, + "CTATTA": 2329, + "CTATTT": 2330, + "CTATTC": 2331, + "CTATTG": 2332, + "CTATCA": 2333, + "CTATCT": 2334, + "CTATCC": 2335, + "CTATCG": 2336, + "CTATGA": 2337, + "CTATGT": 2338, + "CTATGC": 2339, + "CTATGG": 2340, + "CTACAA": 2341, + "CTACAT": 2342, + "CTACAC": 2343, + "CTACAG": 2344, + "CTACTA": 2345, + "CTACTT": 2346, + "CTACTC": 2347, + "CTACTG": 2348, + "CTACCA": 2349, + "CTACCT": 2350, + "CTACCC": 2351, + "CTACCG": 2352, + "CTACGA": 2353, + "CTACGT": 2354, + "CTACGC": 2355, + "CTACGG": 2356, + "CTAGAA": 2357, + "CTAGAT": 2358, + "CTAGAC": 2359, + "CTAGAG": 2360, + "CTAGTA": 2361, + "CTAGTT": 2362, + "CTAGTC": 2363, + "CTAGTG": 2364, + "CTAGCA": 2365, + "CTAGCT": 2366, + "CTAGCC": 2367, + "CTAGCG": 2368, + "CTAGGA": 2369, + "CTAGGT": 2370, + "CTAGGC": 2371, + "CTAGGG": 2372, + "CTTAAA": 2373, + "CTTAAT": 2374, + "CTTAAC": 2375, + "CTTAAG": 2376, + "CTTATA": 2377, + "CTTATT": 2378, + "CTTATC": 2379, + "CTTATG": 2380, + "CTTACA": 2381, + "CTTACT": 2382, + "CTTACC": 2383, + "CTTACG": 2384, + "CTTAGA": 2385, + "CTTAGT": 2386, + "CTTAGC": 2387, + "CTTAGG": 2388, + "CTTTAA": 2389, + "CTTTAT": 2390, + "CTTTAC": 2391, + "CTTTAG": 2392, + "CTTTTA": 2393, + "CTTTTT": 2394, + "CTTTTC": 2395, + "CTTTTG": 2396, + "CTTTCA": 2397, + "CTTTCT": 2398, + "CTTTCC": 2399, + "CTTTCG": 2400, + "CTTTGA": 2401, + "CTTTGT": 2402, + "CTTTGC": 2403, + "CTTTGG": 2404, + "CTTCAA": 2405, + "CTTCAT": 2406, + "CTTCAC": 2407, + "CTTCAG": 2408, + "CTTCTA": 2409, + "CTTCTT": 2410, + "CTTCTC": 2411, + "CTTCTG": 2412, + "CTTCCA": 2413, + "CTTCCT": 2414, + "CTTCCC": 2415, + "CTTCCG": 2416, + "CTTCGA": 2417, + "CTTCGT": 2418, + "CTTCGC": 2419, + "CTTCGG": 2420, + "CTTGAA": 2421, + "CTTGAT": 2422, + "CTTGAC": 2423, + "CTTGAG": 2424, + "CTTGTA": 2425, + "CTTGTT": 2426, + "CTTGTC": 2427, + "CTTGTG": 2428, + "CTTGCA": 2429, + "CTTGCT": 2430, + "CTTGCC": 2431, + "CTTGCG": 2432, + "CTTGGA": 2433, + "CTTGGT": 2434, + "CTTGGC": 2435, + "CTTGGG": 2436, + "CTCAAA": 2437, + "CTCAAT": 2438, + "CTCAAC": 2439, + "CTCAAG": 2440, + "CTCATA": 2441, + "CTCATT": 2442, + "CTCATC": 2443, + "CTCATG": 2444, + "CTCACA": 2445, + "CTCACT": 2446, + "CTCACC": 2447, + "CTCACG": 2448, + "CTCAGA": 2449, + "CTCAGT": 2450, + "CTCAGC": 2451, + "CTCAGG": 2452, + "CTCTAA": 2453, + "CTCTAT": 2454, + "CTCTAC": 2455, + "CTCTAG": 2456, + "CTCTTA": 2457, + "CTCTTT": 2458, + "CTCTTC": 2459, + "CTCTTG": 2460, + "CTCTCA": 2461, + "CTCTCT": 2462, + "CTCTCC": 2463, + "CTCTCG": 2464, + "CTCTGA": 2465, + "CTCTGT": 2466, + "CTCTGC": 2467, + "CTCTGG": 2468, + "CTCCAA": 2469, + "CTCCAT": 2470, + "CTCCAC": 2471, + "CTCCAG": 2472, + "CTCCTA": 2473, + "CTCCTT": 2474, + "CTCCTC": 2475, + "CTCCTG": 2476, + "CTCCCA": 2477, + "CTCCCT": 2478, + "CTCCCC": 2479, + "CTCCCG": 2480, + "CTCCGA": 2481, + "CTCCGT": 2482, + "CTCCGC": 2483, + "CTCCGG": 2484, + "CTCGAA": 2485, + "CTCGAT": 2486, + "CTCGAC": 2487, + "CTCGAG": 2488, + "CTCGTA": 2489, + "CTCGTT": 2490, + "CTCGTC": 2491, + "CTCGTG": 2492, + "CTCGCA": 2493, + "CTCGCT": 2494, + "CTCGCC": 2495, + "CTCGCG": 2496, + "CTCGGA": 2497, + "CTCGGT": 2498, + "CTCGGC": 2499, + "CTCGGG": 2500, + "CTGAAA": 2501, + "CTGAAT": 2502, + "CTGAAC": 2503, + "CTGAAG": 2504, + "CTGATA": 2505, + "CTGATT": 2506, + "CTGATC": 2507, + "CTGATG": 2508, + "CTGACA": 2509, + "CTGACT": 2510, + "CTGACC": 2511, + "CTGACG": 2512, + "CTGAGA": 2513, + "CTGAGT": 2514, + "CTGAGC": 2515, + "CTGAGG": 2516, + "CTGTAA": 2517, + "CTGTAT": 2518, + "CTGTAC": 2519, + "CTGTAG": 2520, + "CTGTTA": 2521, + "CTGTTT": 2522, + "CTGTTC": 2523, + "CTGTTG": 2524, + "CTGTCA": 2525, + "CTGTCT": 2526, + "CTGTCC": 2527, + "CTGTCG": 2528, + "CTGTGA": 2529, + "CTGTGT": 2530, + "CTGTGC": 2531, + "CTGTGG": 2532, + "CTGCAA": 2533, + "CTGCAT": 2534, + "CTGCAC": 2535, + "CTGCAG": 2536, + "CTGCTA": 2537, + "CTGCTT": 2538, + "CTGCTC": 2539, + "CTGCTG": 2540, + "CTGCCA": 2541, + "CTGCCT": 2542, + "CTGCCC": 2543, + "CTGCCG": 2544, + "CTGCGA": 2545, + "CTGCGT": 2546, + "CTGCGC": 2547, + "CTGCGG": 2548, + "CTGGAA": 2549, + "CTGGAT": 2550, + "CTGGAC": 2551, + "CTGGAG": 2552, + "CTGGTA": 2553, + "CTGGTT": 2554, + "CTGGTC": 2555, + "CTGGTG": 2556, + "CTGGCA": 2557, + "CTGGCT": 2558, + "CTGGCC": 2559, + "CTGGCG": 2560, + "CTGGGA": 2561, + "CTGGGT": 2562, + "CTGGGC": 2563, + "CTGGGG": 2564, + "CCAAAA": 2565, + "CCAAAT": 2566, + "CCAAAC": 2567, + "CCAAAG": 2568, + "CCAATA": 2569, + "CCAATT": 2570, + "CCAATC": 2571, + "CCAATG": 2572, + "CCAACA": 2573, + "CCAACT": 2574, + "CCAACC": 2575, + "CCAACG": 2576, + "CCAAGA": 2577, + "CCAAGT": 2578, + "CCAAGC": 2579, + "CCAAGG": 2580, + "CCATAA": 2581, + "CCATAT": 2582, + "CCATAC": 2583, + "CCATAG": 2584, + "CCATTA": 2585, + "CCATTT": 2586, + "CCATTC": 2587, + "CCATTG": 2588, + "CCATCA": 2589, + "CCATCT": 2590, + "CCATCC": 2591, + "CCATCG": 2592, + "CCATGA": 2593, + "CCATGT": 2594, + "CCATGC": 2595, + "CCATGG": 2596, + "CCACAA": 2597, + "CCACAT": 2598, + "CCACAC": 2599, + "CCACAG": 2600, + "CCACTA": 2601, + "CCACTT": 2602, + "CCACTC": 2603, + "CCACTG": 2604, + "CCACCA": 2605, + "CCACCT": 2606, + "CCACCC": 2607, + "CCACCG": 2608, + "CCACGA": 2609, + "CCACGT": 2610, + "CCACGC": 2611, + "CCACGG": 2612, + "CCAGAA": 2613, + "CCAGAT": 2614, + "CCAGAC": 2615, + "CCAGAG": 2616, + "CCAGTA": 2617, + "CCAGTT": 2618, + "CCAGTC": 2619, + "CCAGTG": 2620, + "CCAGCA": 2621, + "CCAGCT": 2622, + "CCAGCC": 2623, + "CCAGCG": 2624, + "CCAGGA": 2625, + "CCAGGT": 2626, + "CCAGGC": 2627, + "CCAGGG": 2628, + "CCTAAA": 2629, + "CCTAAT": 2630, + "CCTAAC": 2631, + "CCTAAG": 2632, + "CCTATA": 2633, + "CCTATT": 2634, + "CCTATC": 2635, + "CCTATG": 2636, + "CCTACA": 2637, + "CCTACT": 2638, + "CCTACC": 2639, + "CCTACG": 2640, + "CCTAGA": 2641, + "CCTAGT": 2642, + "CCTAGC": 2643, + "CCTAGG": 2644, + "CCTTAA": 2645, + "CCTTAT": 2646, + "CCTTAC": 2647, + "CCTTAG": 2648, + "CCTTTA": 2649, + "CCTTTT": 2650, + "CCTTTC": 2651, + "CCTTTG": 2652, + "CCTTCA": 2653, + "CCTTCT": 2654, + "CCTTCC": 2655, + "CCTTCG": 2656, + "CCTTGA": 2657, + "CCTTGT": 2658, + "CCTTGC": 2659, + "CCTTGG": 2660, + "CCTCAA": 2661, + "CCTCAT": 2662, + "CCTCAC": 2663, + "CCTCAG": 2664, + "CCTCTA": 2665, + "CCTCTT": 2666, + "CCTCTC": 2667, + "CCTCTG": 2668, + "CCTCCA": 2669, + "CCTCCT": 2670, + "CCTCCC": 2671, + "CCTCCG": 2672, + "CCTCGA": 2673, + "CCTCGT": 2674, + "CCTCGC": 2675, + "CCTCGG": 2676, + "CCTGAA": 2677, + "CCTGAT": 2678, + "CCTGAC": 2679, + "CCTGAG": 2680, + "CCTGTA": 2681, + "CCTGTT": 2682, + "CCTGTC": 2683, + "CCTGTG": 2684, + "CCTGCA": 2685, + "CCTGCT": 2686, + "CCTGCC": 2687, + "CCTGCG": 2688, + "CCTGGA": 2689, + "CCTGGT": 2690, + "CCTGGC": 2691, + "CCTGGG": 2692, + "CCCAAA": 2693, + "CCCAAT": 2694, + "CCCAAC": 2695, + "CCCAAG": 2696, + "CCCATA": 2697, + "CCCATT": 2698, + "CCCATC": 2699, + "CCCATG": 2700, + "CCCACA": 2701, + "CCCACT": 2702, + "CCCACC": 2703, + "CCCACG": 2704, + "CCCAGA": 2705, + "CCCAGT": 2706, + "CCCAGC": 2707, + "CCCAGG": 2708, + "CCCTAA": 2709, + "CCCTAT": 2710, + "CCCTAC": 2711, + "CCCTAG": 2712, + "CCCTTA": 2713, + "CCCTTT": 2714, + "CCCTTC": 2715, + "CCCTTG": 2716, + "CCCTCA": 2717, + "CCCTCT": 2718, + "CCCTCC": 2719, + "CCCTCG": 2720, + "CCCTGA": 2721, + "CCCTGT": 2722, + "CCCTGC": 2723, + "CCCTGG": 2724, + "CCCCAA": 2725, + "CCCCAT": 2726, + "CCCCAC": 2727, + "CCCCAG": 2728, + "CCCCTA": 2729, + "CCCCTT": 2730, + "CCCCTC": 2731, + "CCCCTG": 2732, + "CCCCCA": 2733, + "CCCCCT": 2734, + "CCCCCC": 2735, + "CCCCCG": 2736, + "CCCCGA": 2737, + "CCCCGT": 2738, + "CCCCGC": 2739, + "CCCCGG": 2740, + "CCCGAA": 2741, + "CCCGAT": 2742, + "CCCGAC": 2743, + "CCCGAG": 2744, + "CCCGTA": 2745, + "CCCGTT": 2746, + "CCCGTC": 2747, + "CCCGTG": 2748, + "CCCGCA": 2749, + "CCCGCT": 2750, + "CCCGCC": 2751, + "CCCGCG": 2752, + "CCCGGA": 2753, + "CCCGGT": 2754, + "CCCGGC": 2755, + "CCCGGG": 2756, + "CCGAAA": 2757, + "CCGAAT": 2758, + "CCGAAC": 2759, + "CCGAAG": 2760, + "CCGATA": 2761, + "CCGATT": 2762, + "CCGATC": 2763, + "CCGATG": 2764, + "CCGACA": 2765, + "CCGACT": 2766, + "CCGACC": 2767, + "CCGACG": 2768, + "CCGAGA": 2769, + "CCGAGT": 2770, + "CCGAGC": 2771, + "CCGAGG": 2772, + "CCGTAA": 2773, + "CCGTAT": 2774, + "CCGTAC": 2775, + "CCGTAG": 2776, + "CCGTTA": 2777, + "CCGTTT": 2778, + "CCGTTC": 2779, + "CCGTTG": 2780, + "CCGTCA": 2781, + "CCGTCT": 2782, + "CCGTCC": 2783, + "CCGTCG": 2784, + "CCGTGA": 2785, + "CCGTGT": 2786, + "CCGTGC": 2787, + "CCGTGG": 2788, + "CCGCAA": 2789, + "CCGCAT": 2790, + "CCGCAC": 2791, + "CCGCAG": 2792, + "CCGCTA": 2793, + "CCGCTT": 2794, + "CCGCTC": 2795, + "CCGCTG": 2796, + "CCGCCA": 2797, + "CCGCCT": 2798, + "CCGCCC": 2799, + "CCGCCG": 2800, + "CCGCGA": 2801, + "CCGCGT": 2802, + "CCGCGC": 2803, + "CCGCGG": 2804, + "CCGGAA": 2805, + "CCGGAT": 2806, + "CCGGAC": 2807, + "CCGGAG": 2808, + "CCGGTA": 2809, + "CCGGTT": 2810, + "CCGGTC": 2811, + "CCGGTG": 2812, + "CCGGCA": 2813, + "CCGGCT": 2814, + "CCGGCC": 2815, + "CCGGCG": 2816, + "CCGGGA": 2817, + "CCGGGT": 2818, + "CCGGGC": 2819, + "CCGGGG": 2820, + "CGAAAA": 2821, + "CGAAAT": 2822, + "CGAAAC": 2823, + "CGAAAG": 2824, + "CGAATA": 2825, + "CGAATT": 2826, + "CGAATC": 2827, + "CGAATG": 2828, + "CGAACA": 2829, + "CGAACT": 2830, + "CGAACC": 2831, + "CGAACG": 2832, + "CGAAGA": 2833, + "CGAAGT": 2834, + "CGAAGC": 2835, + "CGAAGG": 2836, + "CGATAA": 2837, + "CGATAT": 2838, + "CGATAC": 2839, + "CGATAG": 2840, + "CGATTA": 2841, + "CGATTT": 2842, + "CGATTC": 2843, + "CGATTG": 2844, + "CGATCA": 2845, + "CGATCT": 2846, + "CGATCC": 2847, + "CGATCG": 2848, + "CGATGA": 2849, + "CGATGT": 2850, + "CGATGC": 2851, + "CGATGG": 2852, + "CGACAA": 2853, + "CGACAT": 2854, + "CGACAC": 2855, + "CGACAG": 2856, + "CGACTA": 2857, + "CGACTT": 2858, + "CGACTC": 2859, + "CGACTG": 2860, + "CGACCA": 2861, + "CGACCT": 2862, + "CGACCC": 2863, + "CGACCG": 2864, + "CGACGA": 2865, + "CGACGT": 2866, + "CGACGC": 2867, + "CGACGG": 2868, + "CGAGAA": 2869, + "CGAGAT": 2870, + "CGAGAC": 2871, + "CGAGAG": 2872, + "CGAGTA": 2873, + "CGAGTT": 2874, + "CGAGTC": 2875, + "CGAGTG": 2876, + "CGAGCA": 2877, + "CGAGCT": 2878, + "CGAGCC": 2879, + "CGAGCG": 2880, + "CGAGGA": 2881, + "CGAGGT": 2882, + "CGAGGC": 2883, + "CGAGGG": 2884, + "CGTAAA": 2885, + "CGTAAT": 2886, + "CGTAAC": 2887, + "CGTAAG": 2888, + "CGTATA": 2889, + "CGTATT": 2890, + "CGTATC": 2891, + "CGTATG": 2892, + "CGTACA": 2893, + "CGTACT": 2894, + "CGTACC": 2895, + "CGTACG": 2896, + "CGTAGA": 2897, + "CGTAGT": 2898, + "CGTAGC": 2899, + "CGTAGG": 2900, + "CGTTAA": 2901, + "CGTTAT": 2902, + "CGTTAC": 2903, + "CGTTAG": 2904, + "CGTTTA": 2905, + "CGTTTT": 2906, + "CGTTTC": 2907, + "CGTTTG": 2908, + "CGTTCA": 2909, + "CGTTCT": 2910, + "CGTTCC": 2911, + "CGTTCG": 2912, + "CGTTGA": 2913, + "CGTTGT": 2914, + "CGTTGC": 2915, + "CGTTGG": 2916, + "CGTCAA": 2917, + "CGTCAT": 2918, + "CGTCAC": 2919, + "CGTCAG": 2920, + "CGTCTA": 2921, + "CGTCTT": 2922, + "CGTCTC": 2923, + "CGTCTG": 2924, + "CGTCCA": 2925, + "CGTCCT": 2926, + "CGTCCC": 2927, + "CGTCCG": 2928, + "CGTCGA": 2929, + "CGTCGT": 2930, + "CGTCGC": 2931, + "CGTCGG": 2932, + "CGTGAA": 2933, + "CGTGAT": 2934, + "CGTGAC": 2935, + "CGTGAG": 2936, + "CGTGTA": 2937, + "CGTGTT": 2938, + "CGTGTC": 2939, + "CGTGTG": 2940, + "CGTGCA": 2941, + "CGTGCT": 2942, + "CGTGCC": 2943, + "CGTGCG": 2944, + "CGTGGA": 2945, + "CGTGGT": 2946, + "CGTGGC": 2947, + "CGTGGG": 2948, + "CGCAAA": 2949, + "CGCAAT": 2950, + "CGCAAC": 2951, + "CGCAAG": 2952, + "CGCATA": 2953, + "CGCATT": 2954, + "CGCATC": 2955, + "CGCATG": 2956, + "CGCACA": 2957, + "CGCACT": 2958, + "CGCACC": 2959, + "CGCACG": 2960, + "CGCAGA": 2961, + "CGCAGT": 2962, + "CGCAGC": 2963, + "CGCAGG": 2964, + "CGCTAA": 2965, + "CGCTAT": 2966, + "CGCTAC": 2967, + "CGCTAG": 2968, + "CGCTTA": 2969, + "CGCTTT": 2970, + "CGCTTC": 2971, + "CGCTTG": 2972, + "CGCTCA": 2973, + "CGCTCT": 2974, + "CGCTCC": 2975, + "CGCTCG": 2976, + "CGCTGA": 2977, + "CGCTGT": 2978, + "CGCTGC": 2979, + "CGCTGG": 2980, + "CGCCAA": 2981, + "CGCCAT": 2982, + "CGCCAC": 2983, + "CGCCAG": 2984, + "CGCCTA": 2985, + "CGCCTT": 2986, + "CGCCTC": 2987, + "CGCCTG": 2988, + "CGCCCA": 2989, + "CGCCCT": 2990, + "CGCCCC": 2991, + "CGCCCG": 2992, + "CGCCGA": 2993, + "CGCCGT": 2994, + "CGCCGC": 2995, + "CGCCGG": 2996, + "CGCGAA": 2997, + "CGCGAT": 2998, + "CGCGAC": 2999, + "CGCGAG": 3000, + "CGCGTA": 3001, + "CGCGTT": 3002, + "CGCGTC": 3003, + "CGCGTG": 3004, + "CGCGCA": 3005, + "CGCGCT": 3006, + "CGCGCC": 3007, + "CGCGCG": 3008, + "CGCGGA": 3009, + "CGCGGT": 3010, + "CGCGGC": 3011, + "CGCGGG": 3012, + "CGGAAA": 3013, + "CGGAAT": 3014, + "CGGAAC": 3015, + "CGGAAG": 3016, + "CGGATA": 3017, + "CGGATT": 3018, + "CGGATC": 3019, + "CGGATG": 3020, + "CGGACA": 3021, + "CGGACT": 3022, + "CGGACC": 3023, + "CGGACG": 3024, + "CGGAGA": 3025, + "CGGAGT": 3026, + "CGGAGC": 3027, + "CGGAGG": 3028, + "CGGTAA": 3029, + "CGGTAT": 3030, + "CGGTAC": 3031, + "CGGTAG": 3032, + "CGGTTA": 3033, + "CGGTTT": 3034, + "CGGTTC": 3035, + "CGGTTG": 3036, + "CGGTCA": 3037, + "CGGTCT": 3038, + "CGGTCC": 3039, + "CGGTCG": 3040, + "CGGTGA": 3041, + "CGGTGT": 3042, + "CGGTGC": 3043, + "CGGTGG": 3044, + "CGGCAA": 3045, + "CGGCAT": 3046, + "CGGCAC": 3047, + "CGGCAG": 3048, + "CGGCTA": 3049, + "CGGCTT": 3050, + "CGGCTC": 3051, + "CGGCTG": 3052, + "CGGCCA": 3053, + "CGGCCT": 3054, + "CGGCCC": 3055, + "CGGCCG": 3056, + "CGGCGA": 3057, + "CGGCGT": 3058, + "CGGCGC": 3059, + "CGGCGG": 3060, + "CGGGAA": 3061, + "CGGGAT": 3062, + "CGGGAC": 3063, + "CGGGAG": 3064, + "CGGGTA": 3065, + "CGGGTT": 3066, + "CGGGTC": 3067, + "CGGGTG": 3068, + "CGGGCA": 3069, + "CGGGCT": 3070, + "CGGGCC": 3071, + "CGGGCG": 3072, + "CGGGGA": 3073, + "CGGGGT": 3074, + "CGGGGC": 3075, + "CGGGGG": 3076, + "GAAAAA": 3077, + "GAAAAT": 3078, + "GAAAAC": 3079, + "GAAAAG": 3080, + "GAAATA": 3081, + "GAAATT": 3082, + "GAAATC": 3083, + "GAAATG": 3084, + "GAAACA": 3085, + "GAAACT": 3086, + "GAAACC": 3087, + "GAAACG": 3088, + "GAAAGA": 3089, + "GAAAGT": 3090, + "GAAAGC": 3091, + "GAAAGG": 3092, + "GAATAA": 3093, + "GAATAT": 3094, + "GAATAC": 3095, + "GAATAG": 3096, + "GAATTA": 3097, + "GAATTT": 3098, + "GAATTC": 3099, + "GAATTG": 3100, + "GAATCA": 3101, + "GAATCT": 3102, + "GAATCC": 3103, + "GAATCG": 3104, + "GAATGA": 3105, + "GAATGT": 3106, + "GAATGC": 3107, + "GAATGG": 3108, + "GAACAA": 3109, + "GAACAT": 3110, + "GAACAC": 3111, + "GAACAG": 3112, + "GAACTA": 3113, + "GAACTT": 3114, + "GAACTC": 3115, + "GAACTG": 3116, + "GAACCA": 3117, + "GAACCT": 3118, + "GAACCC": 3119, + "GAACCG": 3120, + "GAACGA": 3121, + "GAACGT": 3122, + "GAACGC": 3123, + "GAACGG": 3124, + "GAAGAA": 3125, + "GAAGAT": 3126, + "GAAGAC": 3127, + "GAAGAG": 3128, + "GAAGTA": 3129, + "GAAGTT": 3130, + "GAAGTC": 3131, + "GAAGTG": 3132, + "GAAGCA": 3133, + "GAAGCT": 3134, + "GAAGCC": 3135, + "GAAGCG": 3136, + "GAAGGA": 3137, + "GAAGGT": 3138, + "GAAGGC": 3139, + "GAAGGG": 3140, + "GATAAA": 3141, + "GATAAT": 3142, + "GATAAC": 3143, + "GATAAG": 3144, + "GATATA": 3145, + "GATATT": 3146, + "GATATC": 3147, + "GATATG": 3148, + "GATACA": 3149, + "GATACT": 3150, + "GATACC": 3151, + "GATACG": 3152, + "GATAGA": 3153, + "GATAGT": 3154, + "GATAGC": 3155, + "GATAGG": 3156, + "GATTAA": 3157, + "GATTAT": 3158, + "GATTAC": 3159, + "GATTAG": 3160, + "GATTTA": 3161, + "GATTTT": 3162, + "GATTTC": 3163, + "GATTTG": 3164, + "GATTCA": 3165, + "GATTCT": 3166, + "GATTCC": 3167, + "GATTCG": 3168, + "GATTGA": 3169, + "GATTGT": 3170, + "GATTGC": 3171, + "GATTGG": 3172, + "GATCAA": 3173, + "GATCAT": 3174, + "GATCAC": 3175, + "GATCAG": 3176, + "GATCTA": 3177, + "GATCTT": 3178, + "GATCTC": 3179, + "GATCTG": 3180, + "GATCCA": 3181, + "GATCCT": 3182, + "GATCCC": 3183, + "GATCCG": 3184, + "GATCGA": 3185, + "GATCGT": 3186, + "GATCGC": 3187, + "GATCGG": 3188, + "GATGAA": 3189, + "GATGAT": 3190, + "GATGAC": 3191, + "GATGAG": 3192, + "GATGTA": 3193, + "GATGTT": 3194, + "GATGTC": 3195, + "GATGTG": 3196, + "GATGCA": 3197, + "GATGCT": 3198, + "GATGCC": 3199, + "GATGCG": 3200, + "GATGGA": 3201, + "GATGGT": 3202, + "GATGGC": 3203, + "GATGGG": 3204, + "GACAAA": 3205, + "GACAAT": 3206, + "GACAAC": 3207, + "GACAAG": 3208, + "GACATA": 3209, + "GACATT": 3210, + "GACATC": 3211, + "GACATG": 3212, + "GACACA": 3213, + "GACACT": 3214, + "GACACC": 3215, + "GACACG": 3216, + "GACAGA": 3217, + "GACAGT": 3218, + "GACAGC": 3219, + "GACAGG": 3220, + "GACTAA": 3221, + "GACTAT": 3222, + "GACTAC": 3223, + "GACTAG": 3224, + "GACTTA": 3225, + "GACTTT": 3226, + "GACTTC": 3227, + "GACTTG": 3228, + "GACTCA": 3229, + "GACTCT": 3230, + "GACTCC": 3231, + "GACTCG": 3232, + "GACTGA": 3233, + "GACTGT": 3234, + "GACTGC": 3235, + "GACTGG": 3236, + "GACCAA": 3237, + "GACCAT": 3238, + "GACCAC": 3239, + "GACCAG": 3240, + "GACCTA": 3241, + "GACCTT": 3242, + "GACCTC": 3243, + "GACCTG": 3244, + "GACCCA": 3245, + "GACCCT": 3246, + "GACCCC": 3247, + "GACCCG": 3248, + "GACCGA": 3249, + "GACCGT": 3250, + "GACCGC": 3251, + "GACCGG": 3252, + "GACGAA": 3253, + "GACGAT": 3254, + "GACGAC": 3255, + "GACGAG": 3256, + "GACGTA": 3257, + "GACGTT": 3258, + "GACGTC": 3259, + "GACGTG": 3260, + "GACGCA": 3261, + "GACGCT": 3262, + "GACGCC": 3263, + "GACGCG": 3264, + "GACGGA": 3265, + "GACGGT": 3266, + "GACGGC": 3267, + "GACGGG": 3268, + "GAGAAA": 3269, + "GAGAAT": 3270, + "GAGAAC": 3271, + "GAGAAG": 3272, + "GAGATA": 3273, + "GAGATT": 3274, + "GAGATC": 3275, + "GAGATG": 3276, + "GAGACA": 3277, + "GAGACT": 3278, + "GAGACC": 3279, + "GAGACG": 3280, + "GAGAGA": 3281, + "GAGAGT": 3282, + "GAGAGC": 3283, + "GAGAGG": 3284, + "GAGTAA": 3285, + "GAGTAT": 3286, + "GAGTAC": 3287, + "GAGTAG": 3288, + "GAGTTA": 3289, + "GAGTTT": 3290, + "GAGTTC": 3291, + "GAGTTG": 3292, + "GAGTCA": 3293, + "GAGTCT": 3294, + "GAGTCC": 3295, + "GAGTCG": 3296, + "GAGTGA": 3297, + "GAGTGT": 3298, + "GAGTGC": 3299, + "GAGTGG": 3300, + "GAGCAA": 3301, + "GAGCAT": 3302, + "GAGCAC": 3303, + "GAGCAG": 3304, + "GAGCTA": 3305, + "GAGCTT": 3306, + "GAGCTC": 3307, + "GAGCTG": 3308, + "GAGCCA": 3309, + "GAGCCT": 3310, + "GAGCCC": 3311, + "GAGCCG": 3312, + "GAGCGA": 3313, + "GAGCGT": 3314, + "GAGCGC": 3315, + "GAGCGG": 3316, + "GAGGAA": 3317, + "GAGGAT": 3318, + "GAGGAC": 3319, + "GAGGAG": 3320, + "GAGGTA": 3321, + "GAGGTT": 3322, + "GAGGTC": 3323, + "GAGGTG": 3324, + "GAGGCA": 3325, + "GAGGCT": 3326, + "GAGGCC": 3327, + "GAGGCG": 3328, + "GAGGGA": 3329, + "GAGGGT": 3330, + "GAGGGC": 3331, + "GAGGGG": 3332, + "GTAAAA": 3333, + "GTAAAT": 3334, + "GTAAAC": 3335, + "GTAAAG": 3336, + "GTAATA": 3337, + "GTAATT": 3338, + "GTAATC": 3339, + "GTAATG": 3340, + "GTAACA": 3341, + "GTAACT": 3342, + "GTAACC": 3343, + "GTAACG": 3344, + "GTAAGA": 3345, + "GTAAGT": 3346, + "GTAAGC": 3347, + "GTAAGG": 3348, + "GTATAA": 3349, + "GTATAT": 3350, + "GTATAC": 3351, + "GTATAG": 3352, + "GTATTA": 3353, + "GTATTT": 3354, + "GTATTC": 3355, + "GTATTG": 3356, + "GTATCA": 3357, + "GTATCT": 3358, + "GTATCC": 3359, + "GTATCG": 3360, + "GTATGA": 3361, + "GTATGT": 3362, + "GTATGC": 3363, + "GTATGG": 3364, + "GTACAA": 3365, + "GTACAT": 3366, + "GTACAC": 3367, + "GTACAG": 3368, + "GTACTA": 3369, + "GTACTT": 3370, + "GTACTC": 3371, + "GTACTG": 3372, + "GTACCA": 3373, + "GTACCT": 3374, + "GTACCC": 3375, + "GTACCG": 3376, + "GTACGA": 3377, + "GTACGT": 3378, + "GTACGC": 3379, + "GTACGG": 3380, + "GTAGAA": 3381, + "GTAGAT": 3382, + "GTAGAC": 3383, + "GTAGAG": 3384, + "GTAGTA": 3385, + "GTAGTT": 3386, + "GTAGTC": 3387, + "GTAGTG": 3388, + "GTAGCA": 3389, + "GTAGCT": 3390, + "GTAGCC": 3391, + "GTAGCG": 3392, + "GTAGGA": 3393, + "GTAGGT": 3394, + "GTAGGC": 3395, + "GTAGGG": 3396, + "GTTAAA": 3397, + "GTTAAT": 3398, + "GTTAAC": 3399, + "GTTAAG": 3400, + "GTTATA": 3401, + "GTTATT": 3402, + "GTTATC": 3403, + "GTTATG": 3404, + "GTTACA": 3405, + "GTTACT": 3406, + "GTTACC": 3407, + "GTTACG": 3408, + "GTTAGA": 3409, + "GTTAGT": 3410, + "GTTAGC": 3411, + "GTTAGG": 3412, + "GTTTAA": 3413, + "GTTTAT": 3414, + "GTTTAC": 3415, + "GTTTAG": 3416, + "GTTTTA": 3417, + "GTTTTT": 3418, + "GTTTTC": 3419, + "GTTTTG": 3420, + "GTTTCA": 3421, + "GTTTCT": 3422, + "GTTTCC": 3423, + "GTTTCG": 3424, + "GTTTGA": 3425, + "GTTTGT": 3426, + "GTTTGC": 3427, + "GTTTGG": 3428, + "GTTCAA": 3429, + "GTTCAT": 3430, + "GTTCAC": 3431, + "GTTCAG": 3432, + "GTTCTA": 3433, + "GTTCTT": 3434, + "GTTCTC": 3435, + "GTTCTG": 3436, + "GTTCCA": 3437, + "GTTCCT": 3438, + "GTTCCC": 3439, + "GTTCCG": 3440, + "GTTCGA": 3441, + "GTTCGT": 3442, + "GTTCGC": 3443, + "GTTCGG": 3444, + "GTTGAA": 3445, + "GTTGAT": 3446, + "GTTGAC": 3447, + "GTTGAG": 3448, + "GTTGTA": 3449, + "GTTGTT": 3450, + "GTTGTC": 3451, + "GTTGTG": 3452, + "GTTGCA": 3453, + "GTTGCT": 3454, + "GTTGCC": 3455, + "GTTGCG": 3456, + "GTTGGA": 3457, + "GTTGGT": 3458, + "GTTGGC": 3459, + "GTTGGG": 3460, + "GTCAAA": 3461, + "GTCAAT": 3462, + "GTCAAC": 3463, + "GTCAAG": 3464, + "GTCATA": 3465, + "GTCATT": 3466, + "GTCATC": 3467, + "GTCATG": 3468, + "GTCACA": 3469, + "GTCACT": 3470, + "GTCACC": 3471, + "GTCACG": 3472, + "GTCAGA": 3473, + "GTCAGT": 3474, + "GTCAGC": 3475, + "GTCAGG": 3476, + "GTCTAA": 3477, + "GTCTAT": 3478, + "GTCTAC": 3479, + "GTCTAG": 3480, + "GTCTTA": 3481, + "GTCTTT": 3482, + "GTCTTC": 3483, + "GTCTTG": 3484, + "GTCTCA": 3485, + "GTCTCT": 3486, + "GTCTCC": 3487, + "GTCTCG": 3488, + "GTCTGA": 3489, + "GTCTGT": 3490, + "GTCTGC": 3491, + "GTCTGG": 3492, + "GTCCAA": 3493, + "GTCCAT": 3494, + "GTCCAC": 3495, + "GTCCAG": 3496, + "GTCCTA": 3497, + "GTCCTT": 3498, + "GTCCTC": 3499, + "GTCCTG": 3500, + "GTCCCA": 3501, + "GTCCCT": 3502, + "GTCCCC": 3503, + "GTCCCG": 3504, + "GTCCGA": 3505, + "GTCCGT": 3506, + "GTCCGC": 3507, + "GTCCGG": 3508, + "GTCGAA": 3509, + "GTCGAT": 3510, + "GTCGAC": 3511, + "GTCGAG": 3512, + "GTCGTA": 3513, + "GTCGTT": 3514, + "GTCGTC": 3515, + "GTCGTG": 3516, + "GTCGCA": 3517, + "GTCGCT": 3518, + "GTCGCC": 3519, + "GTCGCG": 3520, + "GTCGGA": 3521, + "GTCGGT": 3522, + "GTCGGC": 3523, + "GTCGGG": 3524, + "GTGAAA": 3525, + "GTGAAT": 3526, + "GTGAAC": 3527, + "GTGAAG": 3528, + "GTGATA": 3529, + "GTGATT": 3530, + "GTGATC": 3531, + "GTGATG": 3532, + "GTGACA": 3533, + "GTGACT": 3534, + "GTGACC": 3535, + "GTGACG": 3536, + "GTGAGA": 3537, + "GTGAGT": 3538, + "GTGAGC": 3539, + "GTGAGG": 3540, + "GTGTAA": 3541, + "GTGTAT": 3542, + "GTGTAC": 3543, + "GTGTAG": 3544, + "GTGTTA": 3545, + "GTGTTT": 3546, + "GTGTTC": 3547, + "GTGTTG": 3548, + "GTGTCA": 3549, + "GTGTCT": 3550, + "GTGTCC": 3551, + "GTGTCG": 3552, + "GTGTGA": 3553, + "GTGTGT": 3554, + "GTGTGC": 3555, + "GTGTGG": 3556, + "GTGCAA": 3557, + "GTGCAT": 3558, + "GTGCAC": 3559, + "GTGCAG": 3560, + "GTGCTA": 3561, + "GTGCTT": 3562, + "GTGCTC": 3563, + "GTGCTG": 3564, + "GTGCCA": 3565, + "GTGCCT": 3566, + "GTGCCC": 3567, + "GTGCCG": 3568, + "GTGCGA": 3569, + "GTGCGT": 3570, + "GTGCGC": 3571, + "GTGCGG": 3572, + "GTGGAA": 3573, + "GTGGAT": 3574, + "GTGGAC": 3575, + "GTGGAG": 3576, + "GTGGTA": 3577, + "GTGGTT": 3578, + "GTGGTC": 3579, + "GTGGTG": 3580, + "GTGGCA": 3581, + "GTGGCT": 3582, + "GTGGCC": 3583, + "GTGGCG": 3584, + "GTGGGA": 3585, + "GTGGGT": 3586, + "GTGGGC": 3587, + "GTGGGG": 3588, + "GCAAAA": 3589, + "GCAAAT": 3590, + "GCAAAC": 3591, + "GCAAAG": 3592, + "GCAATA": 3593, + "GCAATT": 3594, + "GCAATC": 3595, + "GCAATG": 3596, + "GCAACA": 3597, + "GCAACT": 3598, + "GCAACC": 3599, + "GCAACG": 3600, + "GCAAGA": 3601, + "GCAAGT": 3602, + "GCAAGC": 3603, + "GCAAGG": 3604, + "GCATAA": 3605, + "GCATAT": 3606, + "GCATAC": 3607, + "GCATAG": 3608, + "GCATTA": 3609, + "GCATTT": 3610, + "GCATTC": 3611, + "GCATTG": 3612, + "GCATCA": 3613, + "GCATCT": 3614, + "GCATCC": 3615, + "GCATCG": 3616, + "GCATGA": 3617, + "GCATGT": 3618, + "GCATGC": 3619, + "GCATGG": 3620, + "GCACAA": 3621, + "GCACAT": 3622, + "GCACAC": 3623, + "GCACAG": 3624, + "GCACTA": 3625, + "GCACTT": 3626, + "GCACTC": 3627, + "GCACTG": 3628, + "GCACCA": 3629, + "GCACCT": 3630, + "GCACCC": 3631, + "GCACCG": 3632, + "GCACGA": 3633, + "GCACGT": 3634, + "GCACGC": 3635, + "GCACGG": 3636, + "GCAGAA": 3637, + "GCAGAT": 3638, + "GCAGAC": 3639, + "GCAGAG": 3640, + "GCAGTA": 3641, + "GCAGTT": 3642, + "GCAGTC": 3643, + "GCAGTG": 3644, + "GCAGCA": 3645, + "GCAGCT": 3646, + "GCAGCC": 3647, + "GCAGCG": 3648, + "GCAGGA": 3649, + "GCAGGT": 3650, + "GCAGGC": 3651, + "GCAGGG": 3652, + "GCTAAA": 3653, + "GCTAAT": 3654, + "GCTAAC": 3655, + "GCTAAG": 3656, + "GCTATA": 3657, + "GCTATT": 3658, + "GCTATC": 3659, + "GCTATG": 3660, + "GCTACA": 3661, + "GCTACT": 3662, + "GCTACC": 3663, + "GCTACG": 3664, + "GCTAGA": 3665, + "GCTAGT": 3666, + "GCTAGC": 3667, + "GCTAGG": 3668, + "GCTTAA": 3669, + "GCTTAT": 3670, + "GCTTAC": 3671, + "GCTTAG": 3672, + "GCTTTA": 3673, + "GCTTTT": 3674, + "GCTTTC": 3675, + "GCTTTG": 3676, + "GCTTCA": 3677, + "GCTTCT": 3678, + "GCTTCC": 3679, + "GCTTCG": 3680, + "GCTTGA": 3681, + "GCTTGT": 3682, + "GCTTGC": 3683, + "GCTTGG": 3684, + "GCTCAA": 3685, + "GCTCAT": 3686, + "GCTCAC": 3687, + "GCTCAG": 3688, + "GCTCTA": 3689, + "GCTCTT": 3690, + "GCTCTC": 3691, + "GCTCTG": 3692, + "GCTCCA": 3693, + "GCTCCT": 3694, + "GCTCCC": 3695, + "GCTCCG": 3696, + "GCTCGA": 3697, + "GCTCGT": 3698, + "GCTCGC": 3699, + "GCTCGG": 3700, + "GCTGAA": 3701, + "GCTGAT": 3702, + "GCTGAC": 3703, + "GCTGAG": 3704, + "GCTGTA": 3705, + "GCTGTT": 3706, + "GCTGTC": 3707, + "GCTGTG": 3708, + "GCTGCA": 3709, + "GCTGCT": 3710, + "GCTGCC": 3711, + "GCTGCG": 3712, + "GCTGGA": 3713, + "GCTGGT": 3714, + "GCTGGC": 3715, + "GCTGGG": 3716, + "GCCAAA": 3717, + "GCCAAT": 3718, + "GCCAAC": 3719, + "GCCAAG": 3720, + "GCCATA": 3721, + "GCCATT": 3722, + "GCCATC": 3723, + "GCCATG": 3724, + "GCCACA": 3725, + "GCCACT": 3726, + "GCCACC": 3727, + "GCCACG": 3728, + "GCCAGA": 3729, + "GCCAGT": 3730, + "GCCAGC": 3731, + "GCCAGG": 3732, + "GCCTAA": 3733, + "GCCTAT": 3734, + "GCCTAC": 3735, + "GCCTAG": 3736, + "GCCTTA": 3737, + "GCCTTT": 3738, + "GCCTTC": 3739, + "GCCTTG": 3740, + "GCCTCA": 3741, + "GCCTCT": 3742, + "GCCTCC": 3743, + "GCCTCG": 3744, + "GCCTGA": 3745, + "GCCTGT": 3746, + "GCCTGC": 3747, + "GCCTGG": 3748, + "GCCCAA": 3749, + "GCCCAT": 3750, + "GCCCAC": 3751, + "GCCCAG": 3752, + "GCCCTA": 3753, + "GCCCTT": 3754, + "GCCCTC": 3755, + "GCCCTG": 3756, + "GCCCCA": 3757, + "GCCCCT": 3758, + "GCCCCC": 3759, + "GCCCCG": 3760, + "GCCCGA": 3761, + "GCCCGT": 3762, + "GCCCGC": 3763, + "GCCCGG": 3764, + "GCCGAA": 3765, + "GCCGAT": 3766, + "GCCGAC": 3767, + "GCCGAG": 3768, + "GCCGTA": 3769, + "GCCGTT": 3770, + "GCCGTC": 3771, + "GCCGTG": 3772, + "GCCGCA": 3773, + "GCCGCT": 3774, + "GCCGCC": 3775, + "GCCGCG": 3776, + "GCCGGA": 3777, + "GCCGGT": 3778, + "GCCGGC": 3779, + "GCCGGG": 3780, + "GCGAAA": 3781, + "GCGAAT": 3782, + "GCGAAC": 3783, + "GCGAAG": 3784, + "GCGATA": 3785, + "GCGATT": 3786, + "GCGATC": 3787, + "GCGATG": 3788, + "GCGACA": 3789, + "GCGACT": 3790, + "GCGACC": 3791, + "GCGACG": 3792, + "GCGAGA": 3793, + "GCGAGT": 3794, + "GCGAGC": 3795, + "GCGAGG": 3796, + "GCGTAA": 3797, + "GCGTAT": 3798, + "GCGTAC": 3799, + "GCGTAG": 3800, + "GCGTTA": 3801, + "GCGTTT": 3802, + "GCGTTC": 3803, + "GCGTTG": 3804, + "GCGTCA": 3805, + "GCGTCT": 3806, + "GCGTCC": 3807, + "GCGTCG": 3808, + "GCGTGA": 3809, + "GCGTGT": 3810, + "GCGTGC": 3811, + "GCGTGG": 3812, + "GCGCAA": 3813, + "GCGCAT": 3814, + "GCGCAC": 3815, + "GCGCAG": 3816, + "GCGCTA": 3817, + "GCGCTT": 3818, + "GCGCTC": 3819, + "GCGCTG": 3820, + "GCGCCA": 3821, + "GCGCCT": 3822, + "GCGCCC": 3823, + "GCGCCG": 3824, + "GCGCGA": 3825, + "GCGCGT": 3826, + "GCGCGC": 3827, + "GCGCGG": 3828, + "GCGGAA": 3829, + "GCGGAT": 3830, + "GCGGAC": 3831, + "GCGGAG": 3832, + "GCGGTA": 3833, + "GCGGTT": 3834, + "GCGGTC": 3835, + "GCGGTG": 3836, + "GCGGCA": 3837, + "GCGGCT": 3838, + "GCGGCC": 3839, + "GCGGCG": 3840, + "GCGGGA": 3841, + "GCGGGT": 3842, + "GCGGGC": 3843, + "GCGGGG": 3844, + "GGAAAA": 3845, + "GGAAAT": 3846, + "GGAAAC": 3847, + "GGAAAG": 3848, + "GGAATA": 3849, + "GGAATT": 3850, + "GGAATC": 3851, + "GGAATG": 3852, + "GGAACA": 3853, + "GGAACT": 3854, + "GGAACC": 3855, + "GGAACG": 3856, + "GGAAGA": 3857, + "GGAAGT": 3858, + "GGAAGC": 3859, + "GGAAGG": 3860, + "GGATAA": 3861, + "GGATAT": 3862, + "GGATAC": 3863, + "GGATAG": 3864, + "GGATTA": 3865, + "GGATTT": 3866, + "GGATTC": 3867, + "GGATTG": 3868, + "GGATCA": 3869, + "GGATCT": 3870, + "GGATCC": 3871, + "GGATCG": 3872, + "GGATGA": 3873, + "GGATGT": 3874, + "GGATGC": 3875, + "GGATGG": 3876, + "GGACAA": 3877, + "GGACAT": 3878, + "GGACAC": 3879, + "GGACAG": 3880, + "GGACTA": 3881, + "GGACTT": 3882, + "GGACTC": 3883, + "GGACTG": 3884, + "GGACCA": 3885, + "GGACCT": 3886, + "GGACCC": 3887, + "GGACCG": 3888, + "GGACGA": 3889, + "GGACGT": 3890, + "GGACGC": 3891, + "GGACGG": 3892, + "GGAGAA": 3893, + "GGAGAT": 3894, + "GGAGAC": 3895, + "GGAGAG": 3896, + "GGAGTA": 3897, + "GGAGTT": 3898, + "GGAGTC": 3899, + "GGAGTG": 3900, + "GGAGCA": 3901, + "GGAGCT": 3902, + "GGAGCC": 3903, + "GGAGCG": 3904, + "GGAGGA": 3905, + "GGAGGT": 3906, + "GGAGGC": 3907, + "GGAGGG": 3908, + "GGTAAA": 3909, + "GGTAAT": 3910, + "GGTAAC": 3911, + "GGTAAG": 3912, + "GGTATA": 3913, + "GGTATT": 3914, + "GGTATC": 3915, + "GGTATG": 3916, + "GGTACA": 3917, + "GGTACT": 3918, + "GGTACC": 3919, + "GGTACG": 3920, + "GGTAGA": 3921, + "GGTAGT": 3922, + "GGTAGC": 3923, + "GGTAGG": 3924, + "GGTTAA": 3925, + "GGTTAT": 3926, + "GGTTAC": 3927, + "GGTTAG": 3928, + "GGTTTA": 3929, + "GGTTTT": 3930, + "GGTTTC": 3931, + "GGTTTG": 3932, + "GGTTCA": 3933, + "GGTTCT": 3934, + "GGTTCC": 3935, + "GGTTCG": 3936, + "GGTTGA": 3937, + "GGTTGT": 3938, + "GGTTGC": 3939, + "GGTTGG": 3940, + "GGTCAA": 3941, + "GGTCAT": 3942, + "GGTCAC": 3943, + "GGTCAG": 3944, + "GGTCTA": 3945, + "GGTCTT": 3946, + "GGTCTC": 3947, + "GGTCTG": 3948, + "GGTCCA": 3949, + "GGTCCT": 3950, + "GGTCCC": 3951, + "GGTCCG": 3952, + "GGTCGA": 3953, + "GGTCGT": 3954, + "GGTCGC": 3955, + "GGTCGG": 3956, + "GGTGAA": 3957, + "GGTGAT": 3958, + "GGTGAC": 3959, + "GGTGAG": 3960, + "GGTGTA": 3961, + "GGTGTT": 3962, + "GGTGTC": 3963, + "GGTGTG": 3964, + "GGTGCA": 3965, + "GGTGCT": 3966, + "GGTGCC": 3967, + "GGTGCG": 3968, + "GGTGGA": 3969, + "GGTGGT": 3970, + "GGTGGC": 3971, + "GGTGGG": 3972, + "GGCAAA": 3973, + "GGCAAT": 3974, + "GGCAAC": 3975, + "GGCAAG": 3976, + "GGCATA": 3977, + "GGCATT": 3978, + "GGCATC": 3979, + "GGCATG": 3980, + "GGCACA": 3981, + "GGCACT": 3982, + "GGCACC": 3983, + "GGCACG": 3984, + "GGCAGA": 3985, + "GGCAGT": 3986, + "GGCAGC": 3987, + "GGCAGG": 3988, + "GGCTAA": 3989, + "GGCTAT": 3990, + "GGCTAC": 3991, + "GGCTAG": 3992, + "GGCTTA": 3993, + "GGCTTT": 3994, + "GGCTTC": 3995, + "GGCTTG": 3996, + "GGCTCA": 3997, + "GGCTCT": 3998, + "GGCTCC": 3999, + "GGCTCG": 4000, + "GGCTGA": 4001, + "GGCTGT": 4002, + "GGCTGC": 4003, + "GGCTGG": 4004, + "GGCCAA": 4005, + "GGCCAT": 4006, + "GGCCAC": 4007, + "GGCCAG": 4008, + "GGCCTA": 4009, + "GGCCTT": 4010, + "GGCCTC": 4011, + "GGCCTG": 4012, + "GGCCCA": 4013, + "GGCCCT": 4014, + "GGCCCC": 4015, + "GGCCCG": 4016, + "GGCCGA": 4017, + "GGCCGT": 4018, + "GGCCGC": 4019, + "GGCCGG": 4020, + "GGCGAA": 4021, + "GGCGAT": 4022, + "GGCGAC": 4023, + "GGCGAG": 4024, + "GGCGTA": 4025, + "GGCGTT": 4026, + "GGCGTC": 4027, + "GGCGTG": 4028, + "GGCGCA": 4029, + "GGCGCT": 4030, + "GGCGCC": 4031, + "GGCGCG": 4032, + "GGCGGA": 4033, + "GGCGGT": 4034, + "GGCGGC": 4035, + "GGCGGG": 4036, + "GGGAAA": 4037, + "GGGAAT": 4038, + "GGGAAC": 4039, + "GGGAAG": 4040, + "GGGATA": 4041, + "GGGATT": 4042, + "GGGATC": 4043, + "GGGATG": 4044, + "GGGACA": 4045, + "GGGACT": 4046, + "GGGACC": 4047, + "GGGACG": 4048, + "GGGAGA": 4049, + "GGGAGT": 4050, + "GGGAGC": 4051, + "GGGAGG": 4052, + "GGGTAA": 4053, + "GGGTAT": 4054, + "GGGTAC": 4055, + "GGGTAG": 4056, + "GGGTTA": 4057, + "GGGTTT": 4058, + "GGGTTC": 4059, + "GGGTTG": 4060, + "GGGTCA": 4061, + "GGGTCT": 4062, + "GGGTCC": 4063, + "GGGTCG": 4064, + "GGGTGA": 4065, + "GGGTGT": 4066, + "GGGTGC": 4067, + "GGGTGG": 4068, + "GGGCAA": 4069, + "GGGCAT": 4070, + "GGGCAC": 4071, + "GGGCAG": 4072, + "GGGCTA": 4073, + "GGGCTT": 4074, + "GGGCTC": 4075, + "GGGCTG": 4076, + "GGGCCA": 4077, + "GGGCCT": 4078, + "GGGCCC": 4079, + "GGGCCG": 4080, + "GGGCGA": 4081, + "GGGCGT": 4082, + "GGGCGC": 4083, + "GGGCGG": 4084, + "GGGGAA": 4085, + "GGGGAT": 4086, + "GGGGAC": 4087, + "GGGGAG": 4088, + "GGGGTA": 4089, + "GGGGTT": 4090, + "GGGGTC": 4091, + "GGGGTG": 4092, + "GGGGCA": 4093, + "GGGGCT": 4094, + "GGGGCC": 4095, + "GGGGCG": 4096, + "GGGGGA": 4097, + "GGGGGT": 4098, + "GGGGGC": 4099, + "GGGGGG": 4100 + } + } +} \ No newline at end of file diff --git a/tokenizer_config.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/tokenizer_config.json similarity index 100% rename from tokenizer_config.json rename to Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/tokenizer_config.json diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/trainer_state.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..4b8c6cd357d1d58707715e9e6362313848ee93ca --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/trainer_state.json @@ -0,0 +1,313 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 4.0, + "eval_steps": 500, + "global_step": 1000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.1, + "grad_norm": 0.2192622423171997, + "learning_rate": 5.319148936170213e-05, + "loss": 0.6949, + "step": 25 + }, + { + "epoch": 0.2, + "grad_norm": 0.22659960389137268, + "learning_rate": 0.00010638297872340425, + "loss": 0.6926, + "step": 50 + }, + { + "epoch": 0.3, + "grad_norm": 0.33277687430381775, + "learning_rate": 0.00015957446808510637, + "loss": 0.693, + "step": 75 + }, + { + "epoch": 0.4, + "grad_norm": 0.5183415412902832, + "learning_rate": 0.0002127659574468085, + "loss": 0.6928, + "step": 100 + }, + { + "epoch": 0.5, + "grad_norm": 0.21227706968784332, + "learning_rate": 0.00026595744680851064, + "loss": 0.6938, + "step": 125 + }, + { + "epoch": 0.6, + "grad_norm": 0.1846485733985901, + "learning_rate": 0.00031914893617021275, + "loss": 0.6942, + "step": 150 + }, + { + "epoch": 0.7, + "grad_norm": 0.27858808636665344, + "learning_rate": 0.0003723404255319149, + "loss": 0.6931, + "step": 175 + }, + { + "epoch": 0.8, + "grad_norm": 0.26012367010116577, + "learning_rate": 0.0003992081821181128, + "loss": 0.6936, + "step": 200 + }, + { + "epoch": 0.9, + "grad_norm": 0.2534612715244293, + "learning_rate": 0.0003975585615308479, + "loss": 0.6939, + "step": 225 + }, + { + "epoch": 1.0, + "grad_norm": 0.42967700958251953, + "learning_rate": 0.00039590894094358297, + "loss": 0.6928, + "step": 250 + }, + { + "epoch": 1.1, + "grad_norm": 0.2613862454891205, + "learning_rate": 0.0003942593203563181, + "loss": 0.6918, + "step": 275 + }, + { + "epoch": 1.2, + "grad_norm": 0.5514339208602905, + "learning_rate": 0.0003926096997690532, + "loss": 0.6918, + "step": 300 + }, + { + "epoch": 1.3, + "grad_norm": 0.33801883459091187, + "learning_rate": 0.00039096007918178817, + "loss": 0.6906, + "step": 325 + }, + { + "epoch": 1.4, + "grad_norm": 0.2953580617904663, + "learning_rate": 0.0003893104585945233, + "loss": 0.6886, + "step": 350 + }, + { + "epoch": 1.5, + "grad_norm": 0.5340980887413025, + "learning_rate": 0.0003876608380072583, + "loss": 0.6877, + "step": 375 + }, + { + "epoch": 1.6, + "grad_norm": 0.3042701482772827, + "learning_rate": 0.00038601121741999343, + "loss": 0.6853, + "step": 400 + }, + { + "epoch": 1.7, + "grad_norm": 0.44497063755989075, + "learning_rate": 0.0003843615968327285, + "loss": 0.6772, + "step": 425 + }, + { + "epoch": 1.8, + "grad_norm": 0.6260074973106384, + "learning_rate": 0.0003827119762454636, + "loss": 0.6671, + "step": 450 + }, + { + "epoch": 1.9, + "grad_norm": 0.9574269652366638, + "learning_rate": 0.00038106235565819863, + "loss": 0.6517, + "step": 475 + }, + { + "epoch": 2.0, + "grad_norm": 0.9055777192115784, + "learning_rate": 0.0003794127350709337, + "loss": 0.6438, + "step": 500 + }, + { + "epoch": 2.1, + "grad_norm": 0.7860134840011597, + "learning_rate": 0.0003777631144836688, + "loss": 0.6379, + "step": 525 + }, + { + "epoch": 2.2, + "grad_norm": 1.5324610471725464, + "learning_rate": 0.00037611349389640383, + "loss": 0.6143, + "step": 550 + }, + { + "epoch": 2.3, + "grad_norm": 1.8535794019699097, + "learning_rate": 0.00037446387330913894, + "loss": 0.607, + "step": 575 + }, + { + "epoch": 2.4, + "grad_norm": 1.8850122690200806, + "learning_rate": 0.000372814252721874, + "loss": 0.5642, + "step": 600 + }, + { + "epoch": 2.5, + "grad_norm": 1.288278579711914, + "learning_rate": 0.00037116463213460903, + "loss": 0.5739, + "step": 625 + }, + { + "epoch": 2.6, + "grad_norm": 1.344173550605774, + "learning_rate": 0.00036951501154734414, + "loss": 0.5718, + "step": 650 + }, + { + "epoch": 2.7, + "grad_norm": 1.430834174156189, + "learning_rate": 0.0003678653909600792, + "loss": 0.543, + "step": 675 + }, + { + "epoch": 2.8, + "grad_norm": 1.5948325395584106, + "learning_rate": 0.0003662157703728143, + "loss": 0.5434, + "step": 700 + }, + { + "epoch": 2.9, + "grad_norm": 2.688519239425659, + "learning_rate": 0.00036456614978554934, + "loss": 0.5842, + "step": 725 + }, + { + "epoch": 3.0, + "grad_norm": 1.6625819206237793, + "learning_rate": 0.00036291652919828444, + "loss": 0.5526, + "step": 750 + }, + { + "epoch": 3.1, + "grad_norm": 1.869725227355957, + "learning_rate": 0.00036126690861101944, + "loss": 0.53, + "step": 775 + }, + { + "epoch": 3.2, + "grad_norm": 2.1625709533691406, + "learning_rate": 0.00035961728802375454, + "loss": 0.5341, + "step": 800 + }, + { + "epoch": 3.3, + "grad_norm": 1.9650202989578247, + "learning_rate": 0.0003579676674364896, + "loss": 0.5297, + "step": 825 + }, + { + "epoch": 3.4, + "grad_norm": 1.3358676433563232, + "learning_rate": 0.0003563180468492247, + "loss": 0.5273, + "step": 850 + }, + { + "epoch": 3.5, + "grad_norm": 2.1927096843719482, + "learning_rate": 0.0003546684262619598, + "loss": 0.5084, + "step": 875 + }, + { + "epoch": 3.6, + "grad_norm": 2.2377188205718994, + "learning_rate": 0.00035301880567469485, + "loss": 0.4972, + "step": 900 + }, + { + "epoch": 3.7, + "grad_norm": 1.8601839542388916, + "learning_rate": 0.0003513691850874299, + "loss": 0.4852, + "step": 925 + }, + { + "epoch": 3.8, + "grad_norm": 1.4995975494384766, + "learning_rate": 0.00034971956450016495, + "loss": 0.5309, + "step": 950 + }, + { + "epoch": 3.9, + "grad_norm": 2.0536930561065674, + "learning_rate": 0.00034806994391290005, + "loss": 0.5249, + "step": 975 + }, + { + "epoch": 4.0, + "grad_norm": 2.084728479385376, + "learning_rate": 0.0003464203233256351, + "loss": 0.4771, + "step": 1000 + } + ], + "logging_steps": 25, + "max_steps": 6250, + "num_input_tokens_seen": 0, + "num_train_epochs": 25, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 55381032960000.0, + "train_batch_size": 20, + "trial_name": null, + "trial_params": null +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/training_args.bin b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..10d16bb8e9df9c51aba38e7a6f04deeecf1ea29f --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:662a5f25d38ccae34f19d85e24456de6d3eaaabafca26fb7cc8816de0c990f20 +size 5304 diff --git a/vocab.txt b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/vocab.txt similarity index 100% rename from vocab.txt rename to Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/vocab.txt diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/README.md b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..500fd976d92b217455f181bc98c77e612d9a18d9 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/README.md @@ -0,0 +1,202 @@ +--- +base_model: fabihamakhdoomi/TinyDNABERT +library_name: peft +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.12.0 \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/adapter_config.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..04c63e99bbc6b810449dece367c0afb165125b92 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/adapter_config.json @@ -0,0 +1,42 @@ +{ + "alpha_pattern": {}, + "auto_mapping": null, + "base_model_name_or_path": "fabihamakhdoomi/TinyDNABERT", + "bias": "none", + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 16, + "lora_dropout": 0.2, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": [ + "classifier", + "score" + ], + "peft_type": "LORA", + "r": 8, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "bert.encoder.layer.1.attention.self.query", + "bert.encoder.layer.0.output.dense", + "bert.encoder.layer.1.output.dense", + "bert.encoder.layer.0.intermediate.dense", + "bert.encoder.layer.1.intermediate.dense", + "bert.encoder.layer.0.attention.output.dense", + "bert.encoder.layer.1.attention.output.dense", + "bert.encoder.layer.0.attention.self.query", + "bert.encoder.layer.0.attention.self.value", + "bert.encoder.layer.1.attention.self.key", + "bert.encoder.layer.1.attention.self.value", + "bert.encoder.layer.0.attention.self.key" + ], + "task_type": "SEQ_CLS", + "use_dora": false, + "use_rslora": false +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/adapter_model.safetensors b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/adapter_model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..ec2066e0a798b851a02313184fdc68229812c9b6 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/adapter_model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ecbce1e6c3e976a1c065c28896bedccee1d115d97bcfc8f4953120c3c07b8dba +size 151952 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/optimizer.pt b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..ed65154b6a44dc84a32089212693b1b0090bf853 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ae0177ec85ae9d18ccb3afeeb19215f9ab4f65a466fda22adfef27cc7e0228d0 +size 312682 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/rng_state.pth b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..f70e226d95e8a928fcaafa095fe05a0982a8afb3 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fed19b5a5643714b6623a6032bf908ff668a150f2ea673a45992e0ae4072b338 +size 14244 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/scheduler.pt b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..2287c4e5e02bffea6969638a53c3d5402c6dc2b0 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:da119a7607c7d2d96a92ebe57fbaf5fb22ebfce3399b1702b2209f460c0bfe35 +size 1064 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/special_tokens_map.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..a8b3208c2884c4efb86e49300fdd3dc877220cdf --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/special_tokens_map.json @@ -0,0 +1,7 @@ +{ + "cls_token": "[CLS]", + "mask_token": "[MASK]", + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/tokenizer.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..ebd782de6963b9032ce9374cfe30f5275be1a30b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/tokenizer.json @@ -0,0 +1,4265 @@ +{ + "version": "1.0", + "truncation": { + "direction": "Right", + "max_length": 512, + "strategy": "LongestFirst", + "stride": 0 + }, + "padding": { + "strategy": { + "Fixed": 512 + }, + "direction": "Right", + "pad_to_multiple_of": null, + "pad_id": 0, + "pad_type_id": 0, + "pad_token": "[PAD]" + }, + "added_tokens": [ + { + "id": 0, + "content": "[PAD]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 1, + "content": "[UNK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 2, + "content": "[CLS]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 3, + "content": "[SEP]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 4, + "content": "[MASK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + } + ], + "normalizer": { + "type": "BertNormalizer", + "clean_text": true, + "handle_chinese_chars": true, + "strip_accents": null, + "lowercase": false + }, + "pre_tokenizer": { + "type": "BertPreTokenizer" + }, + "post_processor": { + "type": "TemplateProcessing", + "single": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + } + ], + "pair": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "B", + "type_id": 1 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 1 + } + } + ], + "special_tokens": { + "[CLS]": { + "id": "[CLS]", + "ids": [ + 2 + ], + "tokens": [ + "[CLS]" + ] + }, + "[SEP]": { + "id": "[SEP]", + "ids": [ + 3 + ], + "tokens": [ + "[SEP]" + ] + } + } + }, + "decoder": { + "type": "WordPiece", + "prefix": "##", + "cleanup": true + }, + "model": { + "type": "WordPiece", + "unk_token": "[UNK]", + "continuing_subword_prefix": "##", + "max_input_chars_per_word": 100, + "vocab": { + "[PAD]": 0, + "[UNK]": 1, + "[CLS]": 2, + "[SEP]": 3, + "[MASK]": 4, + "AAAAAA": 5, + "AAAAAT": 6, + "AAAAAC": 7, + "AAAAAG": 8, + "AAAATA": 9, + "AAAATT": 10, + "AAAATC": 11, + "AAAATG": 12, + "AAAACA": 13, + "AAAACT": 14, + "AAAACC": 15, + "AAAACG": 16, + "AAAAGA": 17, + "AAAAGT": 18, + "AAAAGC": 19, + "AAAAGG": 20, + "AAATAA": 21, + "AAATAT": 22, + "AAATAC": 23, + "AAATAG": 24, + "AAATTA": 25, + "AAATTT": 26, + "AAATTC": 27, + "AAATTG": 28, + "AAATCA": 29, + "AAATCT": 30, + "AAATCC": 31, + "AAATCG": 32, + "AAATGA": 33, + "AAATGT": 34, + "AAATGC": 35, + "AAATGG": 36, + "AAACAA": 37, + "AAACAT": 38, + "AAACAC": 39, + "AAACAG": 40, + "AAACTA": 41, + "AAACTT": 42, + "AAACTC": 43, + "AAACTG": 44, + "AAACCA": 45, + "AAACCT": 46, + "AAACCC": 47, + "AAACCG": 48, + "AAACGA": 49, + "AAACGT": 50, + "AAACGC": 51, + "AAACGG": 52, + "AAAGAA": 53, + "AAAGAT": 54, + "AAAGAC": 55, + "AAAGAG": 56, + "AAAGTA": 57, + "AAAGTT": 58, + "AAAGTC": 59, + "AAAGTG": 60, + "AAAGCA": 61, + "AAAGCT": 62, + "AAAGCC": 63, + "AAAGCG": 64, + "AAAGGA": 65, + "AAAGGT": 66, + "AAAGGC": 67, + "AAAGGG": 68, + "AATAAA": 69, + "AATAAT": 70, + "AATAAC": 71, + "AATAAG": 72, + "AATATA": 73, + "AATATT": 74, + "AATATC": 75, + "AATATG": 76, + "AATACA": 77, + "AATACT": 78, + "AATACC": 79, + "AATACG": 80, + "AATAGA": 81, + "AATAGT": 82, + "AATAGC": 83, + "AATAGG": 84, + "AATTAA": 85, + "AATTAT": 86, + "AATTAC": 87, + "AATTAG": 88, + "AATTTA": 89, + "AATTTT": 90, + "AATTTC": 91, + "AATTTG": 92, + "AATTCA": 93, + "AATTCT": 94, + "AATTCC": 95, + "AATTCG": 96, + "AATTGA": 97, + "AATTGT": 98, + "AATTGC": 99, + "AATTGG": 100, + "AATCAA": 101, + "AATCAT": 102, + "AATCAC": 103, + "AATCAG": 104, + "AATCTA": 105, + "AATCTT": 106, + "AATCTC": 107, + "AATCTG": 108, + "AATCCA": 109, + "AATCCT": 110, + "AATCCC": 111, + "AATCCG": 112, + "AATCGA": 113, + "AATCGT": 114, + "AATCGC": 115, + "AATCGG": 116, + "AATGAA": 117, + "AATGAT": 118, + "AATGAC": 119, + "AATGAG": 120, + "AATGTA": 121, + "AATGTT": 122, + "AATGTC": 123, + "AATGTG": 124, + "AATGCA": 125, + "AATGCT": 126, + "AATGCC": 127, + "AATGCG": 128, + "AATGGA": 129, + "AATGGT": 130, + "AATGGC": 131, + "AATGGG": 132, + "AACAAA": 133, + "AACAAT": 134, + "AACAAC": 135, + "AACAAG": 136, + "AACATA": 137, + "AACATT": 138, + "AACATC": 139, + "AACATG": 140, + "AACACA": 141, + "AACACT": 142, + "AACACC": 143, + "AACACG": 144, + "AACAGA": 145, + "AACAGT": 146, + "AACAGC": 147, + "AACAGG": 148, + "AACTAA": 149, + "AACTAT": 150, + "AACTAC": 151, + "AACTAG": 152, + "AACTTA": 153, + "AACTTT": 154, + "AACTTC": 155, + "AACTTG": 156, + "AACTCA": 157, + "AACTCT": 158, + "AACTCC": 159, + "AACTCG": 160, + "AACTGA": 161, + "AACTGT": 162, + "AACTGC": 163, + "AACTGG": 164, + "AACCAA": 165, + "AACCAT": 166, + "AACCAC": 167, + "AACCAG": 168, + "AACCTA": 169, + "AACCTT": 170, + "AACCTC": 171, + "AACCTG": 172, + "AACCCA": 173, + "AACCCT": 174, + "AACCCC": 175, + "AACCCG": 176, + "AACCGA": 177, + "AACCGT": 178, + "AACCGC": 179, + "AACCGG": 180, + "AACGAA": 181, + "AACGAT": 182, + "AACGAC": 183, + "AACGAG": 184, + "AACGTA": 185, + "AACGTT": 186, + "AACGTC": 187, + "AACGTG": 188, + "AACGCA": 189, + "AACGCT": 190, + "AACGCC": 191, + "AACGCG": 192, + "AACGGA": 193, + "AACGGT": 194, + "AACGGC": 195, + "AACGGG": 196, + "AAGAAA": 197, + "AAGAAT": 198, + "AAGAAC": 199, + "AAGAAG": 200, + "AAGATA": 201, + "AAGATT": 202, + "AAGATC": 203, + "AAGATG": 204, + "AAGACA": 205, + "AAGACT": 206, + "AAGACC": 207, + "AAGACG": 208, + "AAGAGA": 209, + "AAGAGT": 210, + "AAGAGC": 211, + "AAGAGG": 212, + "AAGTAA": 213, + "AAGTAT": 214, + "AAGTAC": 215, + "AAGTAG": 216, + "AAGTTA": 217, + "AAGTTT": 218, + "AAGTTC": 219, + "AAGTTG": 220, + "AAGTCA": 221, + "AAGTCT": 222, + "AAGTCC": 223, + "AAGTCG": 224, + "AAGTGA": 225, + "AAGTGT": 226, + "AAGTGC": 227, + "AAGTGG": 228, + "AAGCAA": 229, + "AAGCAT": 230, + "AAGCAC": 231, + "AAGCAG": 232, + "AAGCTA": 233, + "AAGCTT": 234, + "AAGCTC": 235, + "AAGCTG": 236, + "AAGCCA": 237, + "AAGCCT": 238, + "AAGCCC": 239, + "AAGCCG": 240, + "AAGCGA": 241, + "AAGCGT": 242, + "AAGCGC": 243, + "AAGCGG": 244, + "AAGGAA": 245, + "AAGGAT": 246, + "AAGGAC": 247, + "AAGGAG": 248, + "AAGGTA": 249, + "AAGGTT": 250, + "AAGGTC": 251, + "AAGGTG": 252, + "AAGGCA": 253, + "AAGGCT": 254, + "AAGGCC": 255, + "AAGGCG": 256, + "AAGGGA": 257, + "AAGGGT": 258, + "AAGGGC": 259, + "AAGGGG": 260, + "ATAAAA": 261, + "ATAAAT": 262, + "ATAAAC": 263, + "ATAAAG": 264, + "ATAATA": 265, + "ATAATT": 266, + "ATAATC": 267, + "ATAATG": 268, + "ATAACA": 269, + "ATAACT": 270, + "ATAACC": 271, + "ATAACG": 272, + "ATAAGA": 273, + "ATAAGT": 274, + "ATAAGC": 275, + "ATAAGG": 276, + "ATATAA": 277, + "ATATAT": 278, + "ATATAC": 279, + "ATATAG": 280, + "ATATTA": 281, + "ATATTT": 282, + "ATATTC": 283, + "ATATTG": 284, + "ATATCA": 285, + "ATATCT": 286, + "ATATCC": 287, + "ATATCG": 288, + "ATATGA": 289, + "ATATGT": 290, + "ATATGC": 291, + "ATATGG": 292, + "ATACAA": 293, + "ATACAT": 294, + "ATACAC": 295, + "ATACAG": 296, + "ATACTA": 297, + "ATACTT": 298, + "ATACTC": 299, + "ATACTG": 300, + "ATACCA": 301, + "ATACCT": 302, + "ATACCC": 303, + "ATACCG": 304, + "ATACGA": 305, + "ATACGT": 306, + "ATACGC": 307, + "ATACGG": 308, + "ATAGAA": 309, + "ATAGAT": 310, + "ATAGAC": 311, + "ATAGAG": 312, + "ATAGTA": 313, + "ATAGTT": 314, + "ATAGTC": 315, + "ATAGTG": 316, + "ATAGCA": 317, + "ATAGCT": 318, + "ATAGCC": 319, + "ATAGCG": 320, + "ATAGGA": 321, + "ATAGGT": 322, + "ATAGGC": 323, + "ATAGGG": 324, + "ATTAAA": 325, + "ATTAAT": 326, + "ATTAAC": 327, + "ATTAAG": 328, + "ATTATA": 329, + "ATTATT": 330, + "ATTATC": 331, + "ATTATG": 332, + "ATTACA": 333, + "ATTACT": 334, + "ATTACC": 335, + "ATTACG": 336, + "ATTAGA": 337, + "ATTAGT": 338, + "ATTAGC": 339, + "ATTAGG": 340, + "ATTTAA": 341, + "ATTTAT": 342, + "ATTTAC": 343, + "ATTTAG": 344, + "ATTTTA": 345, + "ATTTTT": 346, + "ATTTTC": 347, + "ATTTTG": 348, + "ATTTCA": 349, + "ATTTCT": 350, + "ATTTCC": 351, + "ATTTCG": 352, + "ATTTGA": 353, + "ATTTGT": 354, + "ATTTGC": 355, + "ATTTGG": 356, + "ATTCAA": 357, + "ATTCAT": 358, + "ATTCAC": 359, + "ATTCAG": 360, + "ATTCTA": 361, + "ATTCTT": 362, + "ATTCTC": 363, + "ATTCTG": 364, + "ATTCCA": 365, + "ATTCCT": 366, + "ATTCCC": 367, + "ATTCCG": 368, + "ATTCGA": 369, + "ATTCGT": 370, + "ATTCGC": 371, + "ATTCGG": 372, + "ATTGAA": 373, + "ATTGAT": 374, + "ATTGAC": 375, + "ATTGAG": 376, + "ATTGTA": 377, + "ATTGTT": 378, + "ATTGTC": 379, + "ATTGTG": 380, + "ATTGCA": 381, + "ATTGCT": 382, + "ATTGCC": 383, + "ATTGCG": 384, + "ATTGGA": 385, + "ATTGGT": 386, + "ATTGGC": 387, + "ATTGGG": 388, + "ATCAAA": 389, + "ATCAAT": 390, + "ATCAAC": 391, + "ATCAAG": 392, + "ATCATA": 393, + "ATCATT": 394, + "ATCATC": 395, + "ATCATG": 396, + "ATCACA": 397, + "ATCACT": 398, + "ATCACC": 399, + "ATCACG": 400, + "ATCAGA": 401, + "ATCAGT": 402, + "ATCAGC": 403, + "ATCAGG": 404, + "ATCTAA": 405, + "ATCTAT": 406, + "ATCTAC": 407, + "ATCTAG": 408, + "ATCTTA": 409, + "ATCTTT": 410, + "ATCTTC": 411, + "ATCTTG": 412, + "ATCTCA": 413, + "ATCTCT": 414, + "ATCTCC": 415, + "ATCTCG": 416, + "ATCTGA": 417, + "ATCTGT": 418, + "ATCTGC": 419, + "ATCTGG": 420, + "ATCCAA": 421, + "ATCCAT": 422, + "ATCCAC": 423, + "ATCCAG": 424, + "ATCCTA": 425, + "ATCCTT": 426, + "ATCCTC": 427, + "ATCCTG": 428, + "ATCCCA": 429, + "ATCCCT": 430, + "ATCCCC": 431, + "ATCCCG": 432, + "ATCCGA": 433, + "ATCCGT": 434, + "ATCCGC": 435, + "ATCCGG": 436, + "ATCGAA": 437, + "ATCGAT": 438, + "ATCGAC": 439, + "ATCGAG": 440, + "ATCGTA": 441, + "ATCGTT": 442, + "ATCGTC": 443, + "ATCGTG": 444, + "ATCGCA": 445, + "ATCGCT": 446, + "ATCGCC": 447, + "ATCGCG": 448, + "ATCGGA": 449, + "ATCGGT": 450, + "ATCGGC": 451, + "ATCGGG": 452, + "ATGAAA": 453, + "ATGAAT": 454, + "ATGAAC": 455, + "ATGAAG": 456, + "ATGATA": 457, + "ATGATT": 458, + "ATGATC": 459, + "ATGATG": 460, + "ATGACA": 461, + "ATGACT": 462, + "ATGACC": 463, + "ATGACG": 464, + "ATGAGA": 465, + "ATGAGT": 466, + "ATGAGC": 467, + "ATGAGG": 468, + "ATGTAA": 469, + "ATGTAT": 470, + "ATGTAC": 471, + "ATGTAG": 472, + "ATGTTA": 473, + "ATGTTT": 474, + "ATGTTC": 475, + "ATGTTG": 476, + "ATGTCA": 477, + "ATGTCT": 478, + "ATGTCC": 479, + "ATGTCG": 480, + "ATGTGA": 481, + "ATGTGT": 482, + "ATGTGC": 483, + "ATGTGG": 484, + "ATGCAA": 485, + "ATGCAT": 486, + "ATGCAC": 487, + "ATGCAG": 488, + "ATGCTA": 489, + "ATGCTT": 490, + "ATGCTC": 491, + "ATGCTG": 492, + "ATGCCA": 493, + "ATGCCT": 494, + "ATGCCC": 495, + "ATGCCG": 496, + "ATGCGA": 497, + "ATGCGT": 498, + "ATGCGC": 499, + "ATGCGG": 500, + "ATGGAA": 501, + "ATGGAT": 502, + "ATGGAC": 503, + "ATGGAG": 504, + "ATGGTA": 505, + "ATGGTT": 506, + "ATGGTC": 507, + "ATGGTG": 508, + "ATGGCA": 509, + "ATGGCT": 510, + "ATGGCC": 511, + "ATGGCG": 512, + "ATGGGA": 513, + "ATGGGT": 514, + "ATGGGC": 515, + "ATGGGG": 516, + "ACAAAA": 517, + "ACAAAT": 518, + "ACAAAC": 519, + "ACAAAG": 520, + "ACAATA": 521, + "ACAATT": 522, + "ACAATC": 523, + "ACAATG": 524, + "ACAACA": 525, + "ACAACT": 526, + "ACAACC": 527, + "ACAACG": 528, + "ACAAGA": 529, + "ACAAGT": 530, + "ACAAGC": 531, + "ACAAGG": 532, + "ACATAA": 533, + "ACATAT": 534, + "ACATAC": 535, + "ACATAG": 536, + "ACATTA": 537, + "ACATTT": 538, + "ACATTC": 539, + "ACATTG": 540, + "ACATCA": 541, + "ACATCT": 542, + "ACATCC": 543, + "ACATCG": 544, + "ACATGA": 545, + "ACATGT": 546, + "ACATGC": 547, + "ACATGG": 548, + "ACACAA": 549, + "ACACAT": 550, + "ACACAC": 551, + "ACACAG": 552, + "ACACTA": 553, + "ACACTT": 554, + "ACACTC": 555, + "ACACTG": 556, + "ACACCA": 557, + "ACACCT": 558, + "ACACCC": 559, + "ACACCG": 560, + "ACACGA": 561, + "ACACGT": 562, + "ACACGC": 563, + "ACACGG": 564, + "ACAGAA": 565, + "ACAGAT": 566, + "ACAGAC": 567, + "ACAGAG": 568, + "ACAGTA": 569, + "ACAGTT": 570, + "ACAGTC": 571, + "ACAGTG": 572, + "ACAGCA": 573, + "ACAGCT": 574, + "ACAGCC": 575, + "ACAGCG": 576, + "ACAGGA": 577, + "ACAGGT": 578, + "ACAGGC": 579, + "ACAGGG": 580, + "ACTAAA": 581, + "ACTAAT": 582, + "ACTAAC": 583, + "ACTAAG": 584, + "ACTATA": 585, + "ACTATT": 586, + "ACTATC": 587, + "ACTATG": 588, + "ACTACA": 589, + "ACTACT": 590, + "ACTACC": 591, + "ACTACG": 592, + "ACTAGA": 593, + "ACTAGT": 594, + "ACTAGC": 595, + "ACTAGG": 596, + "ACTTAA": 597, + "ACTTAT": 598, + "ACTTAC": 599, + "ACTTAG": 600, + "ACTTTA": 601, + "ACTTTT": 602, + "ACTTTC": 603, + "ACTTTG": 604, + "ACTTCA": 605, + "ACTTCT": 606, + "ACTTCC": 607, + "ACTTCG": 608, + "ACTTGA": 609, + "ACTTGT": 610, + "ACTTGC": 611, + "ACTTGG": 612, + "ACTCAA": 613, + "ACTCAT": 614, + "ACTCAC": 615, + "ACTCAG": 616, + "ACTCTA": 617, + "ACTCTT": 618, + "ACTCTC": 619, + "ACTCTG": 620, + "ACTCCA": 621, + "ACTCCT": 622, + "ACTCCC": 623, + "ACTCCG": 624, + "ACTCGA": 625, + "ACTCGT": 626, + "ACTCGC": 627, + "ACTCGG": 628, + "ACTGAA": 629, + "ACTGAT": 630, + "ACTGAC": 631, + "ACTGAG": 632, + "ACTGTA": 633, + "ACTGTT": 634, + "ACTGTC": 635, + "ACTGTG": 636, + "ACTGCA": 637, + "ACTGCT": 638, + "ACTGCC": 639, + "ACTGCG": 640, + "ACTGGA": 641, + "ACTGGT": 642, + "ACTGGC": 643, + "ACTGGG": 644, + "ACCAAA": 645, + "ACCAAT": 646, + "ACCAAC": 647, + "ACCAAG": 648, + "ACCATA": 649, + "ACCATT": 650, + "ACCATC": 651, + "ACCATG": 652, + "ACCACA": 653, + "ACCACT": 654, + "ACCACC": 655, + "ACCACG": 656, + "ACCAGA": 657, + "ACCAGT": 658, + "ACCAGC": 659, + "ACCAGG": 660, + "ACCTAA": 661, + "ACCTAT": 662, + "ACCTAC": 663, + "ACCTAG": 664, + "ACCTTA": 665, + "ACCTTT": 666, + "ACCTTC": 667, + "ACCTTG": 668, + "ACCTCA": 669, + "ACCTCT": 670, + "ACCTCC": 671, + "ACCTCG": 672, + "ACCTGA": 673, + "ACCTGT": 674, + "ACCTGC": 675, + "ACCTGG": 676, + "ACCCAA": 677, + "ACCCAT": 678, + "ACCCAC": 679, + "ACCCAG": 680, + "ACCCTA": 681, + "ACCCTT": 682, + "ACCCTC": 683, + "ACCCTG": 684, + "ACCCCA": 685, + "ACCCCT": 686, + "ACCCCC": 687, + "ACCCCG": 688, + "ACCCGA": 689, + "ACCCGT": 690, + "ACCCGC": 691, + "ACCCGG": 692, + "ACCGAA": 693, + "ACCGAT": 694, + "ACCGAC": 695, + "ACCGAG": 696, + "ACCGTA": 697, + "ACCGTT": 698, + "ACCGTC": 699, + "ACCGTG": 700, + "ACCGCA": 701, + "ACCGCT": 702, + "ACCGCC": 703, + "ACCGCG": 704, + "ACCGGA": 705, + "ACCGGT": 706, + "ACCGGC": 707, + "ACCGGG": 708, + "ACGAAA": 709, + "ACGAAT": 710, + "ACGAAC": 711, + "ACGAAG": 712, + "ACGATA": 713, + "ACGATT": 714, + "ACGATC": 715, + "ACGATG": 716, + "ACGACA": 717, + "ACGACT": 718, + "ACGACC": 719, + "ACGACG": 720, + "ACGAGA": 721, + "ACGAGT": 722, + "ACGAGC": 723, + "ACGAGG": 724, + "ACGTAA": 725, + "ACGTAT": 726, + "ACGTAC": 727, + "ACGTAG": 728, + "ACGTTA": 729, + "ACGTTT": 730, + "ACGTTC": 731, + "ACGTTG": 732, + "ACGTCA": 733, + "ACGTCT": 734, + "ACGTCC": 735, + "ACGTCG": 736, + "ACGTGA": 737, + "ACGTGT": 738, + "ACGTGC": 739, + "ACGTGG": 740, + "ACGCAA": 741, + "ACGCAT": 742, + "ACGCAC": 743, + "ACGCAG": 744, + "ACGCTA": 745, + "ACGCTT": 746, + "ACGCTC": 747, + "ACGCTG": 748, + "ACGCCA": 749, + "ACGCCT": 750, + "ACGCCC": 751, + "ACGCCG": 752, + "ACGCGA": 753, + "ACGCGT": 754, + "ACGCGC": 755, + "ACGCGG": 756, + "ACGGAA": 757, + "ACGGAT": 758, + "ACGGAC": 759, + "ACGGAG": 760, + "ACGGTA": 761, + "ACGGTT": 762, + "ACGGTC": 763, + "ACGGTG": 764, + "ACGGCA": 765, + "ACGGCT": 766, + "ACGGCC": 767, + "ACGGCG": 768, + "ACGGGA": 769, + "ACGGGT": 770, + "ACGGGC": 771, + "ACGGGG": 772, + "AGAAAA": 773, + "AGAAAT": 774, + "AGAAAC": 775, + "AGAAAG": 776, + "AGAATA": 777, + "AGAATT": 778, + "AGAATC": 779, + "AGAATG": 780, + "AGAACA": 781, + "AGAACT": 782, + "AGAACC": 783, + "AGAACG": 784, + "AGAAGA": 785, + "AGAAGT": 786, + "AGAAGC": 787, + "AGAAGG": 788, + "AGATAA": 789, + "AGATAT": 790, + "AGATAC": 791, + "AGATAG": 792, + "AGATTA": 793, + "AGATTT": 794, + "AGATTC": 795, + "AGATTG": 796, + "AGATCA": 797, + "AGATCT": 798, + "AGATCC": 799, + "AGATCG": 800, + "AGATGA": 801, + "AGATGT": 802, + "AGATGC": 803, + "AGATGG": 804, + "AGACAA": 805, + "AGACAT": 806, + "AGACAC": 807, + "AGACAG": 808, + "AGACTA": 809, + "AGACTT": 810, + "AGACTC": 811, + "AGACTG": 812, + "AGACCA": 813, + "AGACCT": 814, + "AGACCC": 815, + "AGACCG": 816, + "AGACGA": 817, + "AGACGT": 818, + "AGACGC": 819, + "AGACGG": 820, + "AGAGAA": 821, + "AGAGAT": 822, + "AGAGAC": 823, + "AGAGAG": 824, + "AGAGTA": 825, + "AGAGTT": 826, + "AGAGTC": 827, + "AGAGTG": 828, + "AGAGCA": 829, + "AGAGCT": 830, + "AGAGCC": 831, + "AGAGCG": 832, + "AGAGGA": 833, + "AGAGGT": 834, + "AGAGGC": 835, + "AGAGGG": 836, + "AGTAAA": 837, + "AGTAAT": 838, + "AGTAAC": 839, + "AGTAAG": 840, + "AGTATA": 841, + "AGTATT": 842, + "AGTATC": 843, + "AGTATG": 844, + "AGTACA": 845, + "AGTACT": 846, + "AGTACC": 847, + "AGTACG": 848, + "AGTAGA": 849, + "AGTAGT": 850, + "AGTAGC": 851, + "AGTAGG": 852, + "AGTTAA": 853, + "AGTTAT": 854, + "AGTTAC": 855, + "AGTTAG": 856, + "AGTTTA": 857, + "AGTTTT": 858, + "AGTTTC": 859, + "AGTTTG": 860, + "AGTTCA": 861, + "AGTTCT": 862, + "AGTTCC": 863, + "AGTTCG": 864, + "AGTTGA": 865, + "AGTTGT": 866, + "AGTTGC": 867, + "AGTTGG": 868, + "AGTCAA": 869, + "AGTCAT": 870, + "AGTCAC": 871, + "AGTCAG": 872, + "AGTCTA": 873, + "AGTCTT": 874, + "AGTCTC": 875, + "AGTCTG": 876, + "AGTCCA": 877, + "AGTCCT": 878, + "AGTCCC": 879, + "AGTCCG": 880, + "AGTCGA": 881, + "AGTCGT": 882, + "AGTCGC": 883, + "AGTCGG": 884, + "AGTGAA": 885, + "AGTGAT": 886, + "AGTGAC": 887, + "AGTGAG": 888, + "AGTGTA": 889, + "AGTGTT": 890, + "AGTGTC": 891, + "AGTGTG": 892, + "AGTGCA": 893, + "AGTGCT": 894, + "AGTGCC": 895, + "AGTGCG": 896, + "AGTGGA": 897, + "AGTGGT": 898, + "AGTGGC": 899, + "AGTGGG": 900, + "AGCAAA": 901, + "AGCAAT": 902, + "AGCAAC": 903, + "AGCAAG": 904, + "AGCATA": 905, + "AGCATT": 906, + "AGCATC": 907, + "AGCATG": 908, + "AGCACA": 909, + "AGCACT": 910, + "AGCACC": 911, + "AGCACG": 912, + "AGCAGA": 913, + "AGCAGT": 914, + "AGCAGC": 915, + "AGCAGG": 916, + "AGCTAA": 917, + "AGCTAT": 918, + "AGCTAC": 919, + "AGCTAG": 920, + "AGCTTA": 921, + "AGCTTT": 922, + "AGCTTC": 923, + "AGCTTG": 924, + "AGCTCA": 925, + "AGCTCT": 926, + "AGCTCC": 927, + "AGCTCG": 928, + "AGCTGA": 929, + "AGCTGT": 930, + "AGCTGC": 931, + "AGCTGG": 932, + "AGCCAA": 933, + "AGCCAT": 934, + "AGCCAC": 935, + "AGCCAG": 936, + "AGCCTA": 937, + "AGCCTT": 938, + "AGCCTC": 939, + "AGCCTG": 940, + "AGCCCA": 941, + "AGCCCT": 942, + "AGCCCC": 943, + "AGCCCG": 944, + "AGCCGA": 945, + "AGCCGT": 946, + "AGCCGC": 947, + "AGCCGG": 948, + "AGCGAA": 949, + "AGCGAT": 950, + "AGCGAC": 951, + "AGCGAG": 952, + "AGCGTA": 953, + "AGCGTT": 954, + "AGCGTC": 955, + "AGCGTG": 956, + "AGCGCA": 957, + "AGCGCT": 958, + "AGCGCC": 959, + "AGCGCG": 960, + "AGCGGA": 961, + "AGCGGT": 962, + "AGCGGC": 963, + "AGCGGG": 964, + "AGGAAA": 965, + "AGGAAT": 966, + "AGGAAC": 967, + "AGGAAG": 968, + "AGGATA": 969, + "AGGATT": 970, + "AGGATC": 971, + "AGGATG": 972, + "AGGACA": 973, + "AGGACT": 974, + "AGGACC": 975, + "AGGACG": 976, + "AGGAGA": 977, + "AGGAGT": 978, + "AGGAGC": 979, + "AGGAGG": 980, + "AGGTAA": 981, + "AGGTAT": 982, + "AGGTAC": 983, + "AGGTAG": 984, + "AGGTTA": 985, + "AGGTTT": 986, + "AGGTTC": 987, + "AGGTTG": 988, + "AGGTCA": 989, + "AGGTCT": 990, + "AGGTCC": 991, + "AGGTCG": 992, + "AGGTGA": 993, + "AGGTGT": 994, + "AGGTGC": 995, + "AGGTGG": 996, + "AGGCAA": 997, + "AGGCAT": 998, + "AGGCAC": 999, + "AGGCAG": 1000, + "AGGCTA": 1001, + "AGGCTT": 1002, + "AGGCTC": 1003, + "AGGCTG": 1004, + "AGGCCA": 1005, + "AGGCCT": 1006, + "AGGCCC": 1007, + "AGGCCG": 1008, + "AGGCGA": 1009, + "AGGCGT": 1010, + "AGGCGC": 1011, + "AGGCGG": 1012, + "AGGGAA": 1013, + "AGGGAT": 1014, + "AGGGAC": 1015, + "AGGGAG": 1016, + "AGGGTA": 1017, + "AGGGTT": 1018, + "AGGGTC": 1019, + "AGGGTG": 1020, + "AGGGCA": 1021, + "AGGGCT": 1022, + "AGGGCC": 1023, + "AGGGCG": 1024, + "AGGGGA": 1025, + "AGGGGT": 1026, + "AGGGGC": 1027, + "AGGGGG": 1028, + "TAAAAA": 1029, + "TAAAAT": 1030, + "TAAAAC": 1031, + "TAAAAG": 1032, + "TAAATA": 1033, + "TAAATT": 1034, + "TAAATC": 1035, + "TAAATG": 1036, + "TAAACA": 1037, + "TAAACT": 1038, + "TAAACC": 1039, + "TAAACG": 1040, + "TAAAGA": 1041, + "TAAAGT": 1042, + "TAAAGC": 1043, + "TAAAGG": 1044, + "TAATAA": 1045, + "TAATAT": 1046, + "TAATAC": 1047, + "TAATAG": 1048, + "TAATTA": 1049, + "TAATTT": 1050, + "TAATTC": 1051, + "TAATTG": 1052, + "TAATCA": 1053, + "TAATCT": 1054, + "TAATCC": 1055, + "TAATCG": 1056, + "TAATGA": 1057, + "TAATGT": 1058, + "TAATGC": 1059, + "TAATGG": 1060, + "TAACAA": 1061, + "TAACAT": 1062, + "TAACAC": 1063, + "TAACAG": 1064, + "TAACTA": 1065, + "TAACTT": 1066, + "TAACTC": 1067, + "TAACTG": 1068, + "TAACCA": 1069, + "TAACCT": 1070, + "TAACCC": 1071, + "TAACCG": 1072, + "TAACGA": 1073, + "TAACGT": 1074, + "TAACGC": 1075, + "TAACGG": 1076, + "TAAGAA": 1077, + "TAAGAT": 1078, + "TAAGAC": 1079, + "TAAGAG": 1080, + "TAAGTA": 1081, + "TAAGTT": 1082, + "TAAGTC": 1083, + "TAAGTG": 1084, + "TAAGCA": 1085, + "TAAGCT": 1086, + "TAAGCC": 1087, + "TAAGCG": 1088, + "TAAGGA": 1089, + "TAAGGT": 1090, + "TAAGGC": 1091, + "TAAGGG": 1092, + "TATAAA": 1093, + "TATAAT": 1094, + "TATAAC": 1095, + "TATAAG": 1096, + "TATATA": 1097, + "TATATT": 1098, + "TATATC": 1099, + "TATATG": 1100, + "TATACA": 1101, + "TATACT": 1102, + "TATACC": 1103, + "TATACG": 1104, + "TATAGA": 1105, + "TATAGT": 1106, + "TATAGC": 1107, + "TATAGG": 1108, + "TATTAA": 1109, + "TATTAT": 1110, + "TATTAC": 1111, + "TATTAG": 1112, + "TATTTA": 1113, + "TATTTT": 1114, + "TATTTC": 1115, + "TATTTG": 1116, + "TATTCA": 1117, + "TATTCT": 1118, + "TATTCC": 1119, + "TATTCG": 1120, + "TATTGA": 1121, + "TATTGT": 1122, + "TATTGC": 1123, + "TATTGG": 1124, + "TATCAA": 1125, + "TATCAT": 1126, + "TATCAC": 1127, + "TATCAG": 1128, + "TATCTA": 1129, + "TATCTT": 1130, + "TATCTC": 1131, + "TATCTG": 1132, + "TATCCA": 1133, + "TATCCT": 1134, + "TATCCC": 1135, + "TATCCG": 1136, + "TATCGA": 1137, + "TATCGT": 1138, + "TATCGC": 1139, + "TATCGG": 1140, + "TATGAA": 1141, + "TATGAT": 1142, + "TATGAC": 1143, + "TATGAG": 1144, + "TATGTA": 1145, + "TATGTT": 1146, + "TATGTC": 1147, + "TATGTG": 1148, + "TATGCA": 1149, + "TATGCT": 1150, + "TATGCC": 1151, + "TATGCG": 1152, + "TATGGA": 1153, + "TATGGT": 1154, + "TATGGC": 1155, + "TATGGG": 1156, + "TACAAA": 1157, + "TACAAT": 1158, + "TACAAC": 1159, + "TACAAG": 1160, + "TACATA": 1161, + "TACATT": 1162, + "TACATC": 1163, + "TACATG": 1164, + "TACACA": 1165, + "TACACT": 1166, + "TACACC": 1167, + "TACACG": 1168, + "TACAGA": 1169, + "TACAGT": 1170, + "TACAGC": 1171, + "TACAGG": 1172, + "TACTAA": 1173, + "TACTAT": 1174, + "TACTAC": 1175, + "TACTAG": 1176, + "TACTTA": 1177, + "TACTTT": 1178, + "TACTTC": 1179, + "TACTTG": 1180, + "TACTCA": 1181, + "TACTCT": 1182, + "TACTCC": 1183, + "TACTCG": 1184, + "TACTGA": 1185, + "TACTGT": 1186, + "TACTGC": 1187, + "TACTGG": 1188, + "TACCAA": 1189, + "TACCAT": 1190, + "TACCAC": 1191, + "TACCAG": 1192, + "TACCTA": 1193, + "TACCTT": 1194, + "TACCTC": 1195, + "TACCTG": 1196, + "TACCCA": 1197, + "TACCCT": 1198, + "TACCCC": 1199, + "TACCCG": 1200, + "TACCGA": 1201, + "TACCGT": 1202, + "TACCGC": 1203, + "TACCGG": 1204, + "TACGAA": 1205, + "TACGAT": 1206, + "TACGAC": 1207, + "TACGAG": 1208, + "TACGTA": 1209, + "TACGTT": 1210, + "TACGTC": 1211, + "TACGTG": 1212, + "TACGCA": 1213, + "TACGCT": 1214, + "TACGCC": 1215, + "TACGCG": 1216, + "TACGGA": 1217, + "TACGGT": 1218, + "TACGGC": 1219, + "TACGGG": 1220, + "TAGAAA": 1221, + "TAGAAT": 1222, + "TAGAAC": 1223, + "TAGAAG": 1224, + "TAGATA": 1225, + "TAGATT": 1226, + "TAGATC": 1227, + "TAGATG": 1228, + "TAGACA": 1229, + "TAGACT": 1230, + "TAGACC": 1231, + "TAGACG": 1232, + "TAGAGA": 1233, + "TAGAGT": 1234, + "TAGAGC": 1235, + "TAGAGG": 1236, + "TAGTAA": 1237, + "TAGTAT": 1238, + "TAGTAC": 1239, + "TAGTAG": 1240, + "TAGTTA": 1241, + "TAGTTT": 1242, + "TAGTTC": 1243, + "TAGTTG": 1244, + "TAGTCA": 1245, + "TAGTCT": 1246, + "TAGTCC": 1247, + "TAGTCG": 1248, + "TAGTGA": 1249, + "TAGTGT": 1250, + "TAGTGC": 1251, + "TAGTGG": 1252, + "TAGCAA": 1253, + "TAGCAT": 1254, + "TAGCAC": 1255, + "TAGCAG": 1256, + "TAGCTA": 1257, + "TAGCTT": 1258, + "TAGCTC": 1259, + "TAGCTG": 1260, + "TAGCCA": 1261, + "TAGCCT": 1262, + "TAGCCC": 1263, + "TAGCCG": 1264, + "TAGCGA": 1265, + "TAGCGT": 1266, + "TAGCGC": 1267, + "TAGCGG": 1268, + "TAGGAA": 1269, + "TAGGAT": 1270, + "TAGGAC": 1271, + "TAGGAG": 1272, + "TAGGTA": 1273, + "TAGGTT": 1274, + "TAGGTC": 1275, + "TAGGTG": 1276, + "TAGGCA": 1277, + "TAGGCT": 1278, + "TAGGCC": 1279, + "TAGGCG": 1280, + "TAGGGA": 1281, + "TAGGGT": 1282, + "TAGGGC": 1283, + "TAGGGG": 1284, + "TTAAAA": 1285, + "TTAAAT": 1286, + "TTAAAC": 1287, + "TTAAAG": 1288, + "TTAATA": 1289, + "TTAATT": 1290, + "TTAATC": 1291, + "TTAATG": 1292, + "TTAACA": 1293, + "TTAACT": 1294, + "TTAACC": 1295, + "TTAACG": 1296, + "TTAAGA": 1297, + "TTAAGT": 1298, + "TTAAGC": 1299, + "TTAAGG": 1300, + "TTATAA": 1301, + "TTATAT": 1302, + "TTATAC": 1303, + "TTATAG": 1304, + "TTATTA": 1305, + "TTATTT": 1306, + "TTATTC": 1307, + "TTATTG": 1308, + "TTATCA": 1309, + "TTATCT": 1310, + "TTATCC": 1311, + "TTATCG": 1312, + "TTATGA": 1313, + "TTATGT": 1314, + "TTATGC": 1315, + "TTATGG": 1316, + "TTACAA": 1317, + "TTACAT": 1318, + "TTACAC": 1319, + "TTACAG": 1320, + "TTACTA": 1321, + "TTACTT": 1322, + "TTACTC": 1323, + "TTACTG": 1324, + "TTACCA": 1325, + "TTACCT": 1326, + "TTACCC": 1327, + "TTACCG": 1328, + "TTACGA": 1329, + "TTACGT": 1330, + "TTACGC": 1331, + "TTACGG": 1332, + "TTAGAA": 1333, + "TTAGAT": 1334, + "TTAGAC": 1335, + "TTAGAG": 1336, + "TTAGTA": 1337, + "TTAGTT": 1338, + "TTAGTC": 1339, + "TTAGTG": 1340, + "TTAGCA": 1341, + "TTAGCT": 1342, + "TTAGCC": 1343, + "TTAGCG": 1344, + "TTAGGA": 1345, + "TTAGGT": 1346, + "TTAGGC": 1347, + "TTAGGG": 1348, + "TTTAAA": 1349, + "TTTAAT": 1350, + "TTTAAC": 1351, + "TTTAAG": 1352, + "TTTATA": 1353, + "TTTATT": 1354, + "TTTATC": 1355, + "TTTATG": 1356, + "TTTACA": 1357, + "TTTACT": 1358, + "TTTACC": 1359, + "TTTACG": 1360, + "TTTAGA": 1361, + "TTTAGT": 1362, + "TTTAGC": 1363, + "TTTAGG": 1364, + "TTTTAA": 1365, + "TTTTAT": 1366, + "TTTTAC": 1367, + "TTTTAG": 1368, + "TTTTTA": 1369, + "TTTTTT": 1370, + "TTTTTC": 1371, + "TTTTTG": 1372, + "TTTTCA": 1373, + "TTTTCT": 1374, + "TTTTCC": 1375, + "TTTTCG": 1376, + "TTTTGA": 1377, + "TTTTGT": 1378, + "TTTTGC": 1379, + "TTTTGG": 1380, + "TTTCAA": 1381, + "TTTCAT": 1382, + "TTTCAC": 1383, + "TTTCAG": 1384, + "TTTCTA": 1385, + "TTTCTT": 1386, + "TTTCTC": 1387, + "TTTCTG": 1388, + "TTTCCA": 1389, + "TTTCCT": 1390, + "TTTCCC": 1391, + "TTTCCG": 1392, + "TTTCGA": 1393, + "TTTCGT": 1394, + "TTTCGC": 1395, + "TTTCGG": 1396, + "TTTGAA": 1397, + "TTTGAT": 1398, + "TTTGAC": 1399, + "TTTGAG": 1400, + "TTTGTA": 1401, + "TTTGTT": 1402, + "TTTGTC": 1403, + "TTTGTG": 1404, + "TTTGCA": 1405, + "TTTGCT": 1406, + "TTTGCC": 1407, + "TTTGCG": 1408, + "TTTGGA": 1409, + "TTTGGT": 1410, + "TTTGGC": 1411, + "TTTGGG": 1412, + "TTCAAA": 1413, + "TTCAAT": 1414, + "TTCAAC": 1415, + "TTCAAG": 1416, + "TTCATA": 1417, + "TTCATT": 1418, + "TTCATC": 1419, + "TTCATG": 1420, + "TTCACA": 1421, + "TTCACT": 1422, + "TTCACC": 1423, + "TTCACG": 1424, + "TTCAGA": 1425, + "TTCAGT": 1426, + "TTCAGC": 1427, + "TTCAGG": 1428, + "TTCTAA": 1429, + "TTCTAT": 1430, + "TTCTAC": 1431, + "TTCTAG": 1432, + "TTCTTA": 1433, + "TTCTTT": 1434, + "TTCTTC": 1435, + "TTCTTG": 1436, + "TTCTCA": 1437, + "TTCTCT": 1438, + "TTCTCC": 1439, + "TTCTCG": 1440, + "TTCTGA": 1441, + "TTCTGT": 1442, + "TTCTGC": 1443, + "TTCTGG": 1444, + "TTCCAA": 1445, + "TTCCAT": 1446, + "TTCCAC": 1447, + "TTCCAG": 1448, + "TTCCTA": 1449, + "TTCCTT": 1450, + "TTCCTC": 1451, + "TTCCTG": 1452, + "TTCCCA": 1453, + "TTCCCT": 1454, + "TTCCCC": 1455, + "TTCCCG": 1456, + "TTCCGA": 1457, + "TTCCGT": 1458, + "TTCCGC": 1459, + "TTCCGG": 1460, + "TTCGAA": 1461, + "TTCGAT": 1462, + "TTCGAC": 1463, + "TTCGAG": 1464, + "TTCGTA": 1465, + "TTCGTT": 1466, + "TTCGTC": 1467, + "TTCGTG": 1468, + "TTCGCA": 1469, + "TTCGCT": 1470, + "TTCGCC": 1471, + "TTCGCG": 1472, + "TTCGGA": 1473, + "TTCGGT": 1474, + "TTCGGC": 1475, + "TTCGGG": 1476, + "TTGAAA": 1477, + "TTGAAT": 1478, + "TTGAAC": 1479, + "TTGAAG": 1480, + "TTGATA": 1481, + "TTGATT": 1482, + "TTGATC": 1483, + "TTGATG": 1484, + "TTGACA": 1485, + "TTGACT": 1486, + "TTGACC": 1487, + "TTGACG": 1488, + "TTGAGA": 1489, + "TTGAGT": 1490, + "TTGAGC": 1491, + "TTGAGG": 1492, + "TTGTAA": 1493, + "TTGTAT": 1494, + "TTGTAC": 1495, + "TTGTAG": 1496, + "TTGTTA": 1497, + "TTGTTT": 1498, + "TTGTTC": 1499, + "TTGTTG": 1500, + "TTGTCA": 1501, + "TTGTCT": 1502, + "TTGTCC": 1503, + "TTGTCG": 1504, + "TTGTGA": 1505, + "TTGTGT": 1506, + "TTGTGC": 1507, + "TTGTGG": 1508, + "TTGCAA": 1509, + "TTGCAT": 1510, + "TTGCAC": 1511, + "TTGCAG": 1512, + "TTGCTA": 1513, + "TTGCTT": 1514, + "TTGCTC": 1515, + "TTGCTG": 1516, + "TTGCCA": 1517, + "TTGCCT": 1518, + "TTGCCC": 1519, + "TTGCCG": 1520, + "TTGCGA": 1521, + "TTGCGT": 1522, + "TTGCGC": 1523, + "TTGCGG": 1524, + "TTGGAA": 1525, + "TTGGAT": 1526, + "TTGGAC": 1527, + "TTGGAG": 1528, + "TTGGTA": 1529, + "TTGGTT": 1530, + "TTGGTC": 1531, + "TTGGTG": 1532, + "TTGGCA": 1533, + "TTGGCT": 1534, + "TTGGCC": 1535, + "TTGGCG": 1536, + "TTGGGA": 1537, + "TTGGGT": 1538, + "TTGGGC": 1539, + "TTGGGG": 1540, + "TCAAAA": 1541, + "TCAAAT": 1542, + "TCAAAC": 1543, + "TCAAAG": 1544, + "TCAATA": 1545, + "TCAATT": 1546, + "TCAATC": 1547, + "TCAATG": 1548, + "TCAACA": 1549, + "TCAACT": 1550, + "TCAACC": 1551, + "TCAACG": 1552, + "TCAAGA": 1553, + "TCAAGT": 1554, + "TCAAGC": 1555, + "TCAAGG": 1556, + "TCATAA": 1557, + "TCATAT": 1558, + "TCATAC": 1559, + "TCATAG": 1560, + "TCATTA": 1561, + "TCATTT": 1562, + "TCATTC": 1563, + "TCATTG": 1564, + "TCATCA": 1565, + "TCATCT": 1566, + "TCATCC": 1567, + "TCATCG": 1568, + "TCATGA": 1569, + "TCATGT": 1570, + "TCATGC": 1571, + "TCATGG": 1572, + "TCACAA": 1573, + "TCACAT": 1574, + "TCACAC": 1575, + "TCACAG": 1576, + "TCACTA": 1577, + "TCACTT": 1578, + "TCACTC": 1579, + "TCACTG": 1580, + "TCACCA": 1581, + "TCACCT": 1582, + "TCACCC": 1583, + "TCACCG": 1584, + "TCACGA": 1585, + "TCACGT": 1586, + "TCACGC": 1587, + "TCACGG": 1588, + "TCAGAA": 1589, + "TCAGAT": 1590, + "TCAGAC": 1591, + "TCAGAG": 1592, + "TCAGTA": 1593, + "TCAGTT": 1594, + "TCAGTC": 1595, + "TCAGTG": 1596, + "TCAGCA": 1597, + "TCAGCT": 1598, + "TCAGCC": 1599, + "TCAGCG": 1600, + "TCAGGA": 1601, + "TCAGGT": 1602, + "TCAGGC": 1603, + "TCAGGG": 1604, + "TCTAAA": 1605, + "TCTAAT": 1606, + "TCTAAC": 1607, + "TCTAAG": 1608, + "TCTATA": 1609, + "TCTATT": 1610, + "TCTATC": 1611, + "TCTATG": 1612, + "TCTACA": 1613, + "TCTACT": 1614, + "TCTACC": 1615, + "TCTACG": 1616, + "TCTAGA": 1617, + "TCTAGT": 1618, + "TCTAGC": 1619, + "TCTAGG": 1620, + "TCTTAA": 1621, + "TCTTAT": 1622, + "TCTTAC": 1623, + "TCTTAG": 1624, + "TCTTTA": 1625, + "TCTTTT": 1626, + "TCTTTC": 1627, + "TCTTTG": 1628, + "TCTTCA": 1629, + "TCTTCT": 1630, + "TCTTCC": 1631, + "TCTTCG": 1632, + "TCTTGA": 1633, + "TCTTGT": 1634, + "TCTTGC": 1635, + "TCTTGG": 1636, + "TCTCAA": 1637, + "TCTCAT": 1638, + "TCTCAC": 1639, + "TCTCAG": 1640, + "TCTCTA": 1641, + "TCTCTT": 1642, + "TCTCTC": 1643, + "TCTCTG": 1644, + "TCTCCA": 1645, + "TCTCCT": 1646, + "TCTCCC": 1647, + "TCTCCG": 1648, + "TCTCGA": 1649, + "TCTCGT": 1650, + "TCTCGC": 1651, + "TCTCGG": 1652, + "TCTGAA": 1653, + "TCTGAT": 1654, + "TCTGAC": 1655, + "TCTGAG": 1656, + "TCTGTA": 1657, + "TCTGTT": 1658, + "TCTGTC": 1659, + "TCTGTG": 1660, + "TCTGCA": 1661, + "TCTGCT": 1662, + "TCTGCC": 1663, + "TCTGCG": 1664, + "TCTGGA": 1665, + "TCTGGT": 1666, + "TCTGGC": 1667, + "TCTGGG": 1668, + "TCCAAA": 1669, + "TCCAAT": 1670, + "TCCAAC": 1671, + "TCCAAG": 1672, + "TCCATA": 1673, + "TCCATT": 1674, + "TCCATC": 1675, + "TCCATG": 1676, + "TCCACA": 1677, + "TCCACT": 1678, + "TCCACC": 1679, + "TCCACG": 1680, + "TCCAGA": 1681, + "TCCAGT": 1682, + "TCCAGC": 1683, + "TCCAGG": 1684, + "TCCTAA": 1685, + "TCCTAT": 1686, + "TCCTAC": 1687, + "TCCTAG": 1688, + "TCCTTA": 1689, + "TCCTTT": 1690, + "TCCTTC": 1691, + "TCCTTG": 1692, + "TCCTCA": 1693, + "TCCTCT": 1694, + "TCCTCC": 1695, + "TCCTCG": 1696, + "TCCTGA": 1697, + "TCCTGT": 1698, + "TCCTGC": 1699, + "TCCTGG": 1700, + "TCCCAA": 1701, + "TCCCAT": 1702, + "TCCCAC": 1703, + "TCCCAG": 1704, + "TCCCTA": 1705, + "TCCCTT": 1706, + "TCCCTC": 1707, + "TCCCTG": 1708, + "TCCCCA": 1709, + "TCCCCT": 1710, + "TCCCCC": 1711, + "TCCCCG": 1712, + "TCCCGA": 1713, + "TCCCGT": 1714, + "TCCCGC": 1715, + "TCCCGG": 1716, + "TCCGAA": 1717, + "TCCGAT": 1718, + "TCCGAC": 1719, + "TCCGAG": 1720, + "TCCGTA": 1721, + "TCCGTT": 1722, + "TCCGTC": 1723, + "TCCGTG": 1724, + "TCCGCA": 1725, + "TCCGCT": 1726, + "TCCGCC": 1727, + "TCCGCG": 1728, + "TCCGGA": 1729, + "TCCGGT": 1730, + "TCCGGC": 1731, + "TCCGGG": 1732, + "TCGAAA": 1733, + "TCGAAT": 1734, + "TCGAAC": 1735, + "TCGAAG": 1736, + "TCGATA": 1737, + "TCGATT": 1738, + "TCGATC": 1739, + "TCGATG": 1740, + "TCGACA": 1741, + "TCGACT": 1742, + "TCGACC": 1743, + "TCGACG": 1744, + "TCGAGA": 1745, + "TCGAGT": 1746, + "TCGAGC": 1747, + "TCGAGG": 1748, + "TCGTAA": 1749, + "TCGTAT": 1750, + "TCGTAC": 1751, + "TCGTAG": 1752, + "TCGTTA": 1753, + "TCGTTT": 1754, + "TCGTTC": 1755, + "TCGTTG": 1756, + "TCGTCA": 1757, + "TCGTCT": 1758, + "TCGTCC": 1759, + "TCGTCG": 1760, + "TCGTGA": 1761, + "TCGTGT": 1762, + "TCGTGC": 1763, + "TCGTGG": 1764, + "TCGCAA": 1765, + "TCGCAT": 1766, + "TCGCAC": 1767, + "TCGCAG": 1768, + "TCGCTA": 1769, + "TCGCTT": 1770, + "TCGCTC": 1771, + "TCGCTG": 1772, + "TCGCCA": 1773, + "TCGCCT": 1774, + "TCGCCC": 1775, + "TCGCCG": 1776, + "TCGCGA": 1777, + "TCGCGT": 1778, + "TCGCGC": 1779, + "TCGCGG": 1780, + "TCGGAA": 1781, + "TCGGAT": 1782, + "TCGGAC": 1783, + "TCGGAG": 1784, + "TCGGTA": 1785, + "TCGGTT": 1786, + "TCGGTC": 1787, + "TCGGTG": 1788, + "TCGGCA": 1789, + "TCGGCT": 1790, + "TCGGCC": 1791, + "TCGGCG": 1792, + "TCGGGA": 1793, + "TCGGGT": 1794, + "TCGGGC": 1795, + "TCGGGG": 1796, + "TGAAAA": 1797, + "TGAAAT": 1798, + "TGAAAC": 1799, + "TGAAAG": 1800, + "TGAATA": 1801, + "TGAATT": 1802, + "TGAATC": 1803, + "TGAATG": 1804, + "TGAACA": 1805, + "TGAACT": 1806, + "TGAACC": 1807, + "TGAACG": 1808, + "TGAAGA": 1809, + "TGAAGT": 1810, + "TGAAGC": 1811, + "TGAAGG": 1812, + "TGATAA": 1813, + "TGATAT": 1814, + "TGATAC": 1815, + "TGATAG": 1816, + "TGATTA": 1817, + "TGATTT": 1818, + "TGATTC": 1819, + "TGATTG": 1820, + "TGATCA": 1821, + "TGATCT": 1822, + "TGATCC": 1823, + "TGATCG": 1824, + "TGATGA": 1825, + "TGATGT": 1826, + "TGATGC": 1827, + "TGATGG": 1828, + "TGACAA": 1829, + "TGACAT": 1830, + "TGACAC": 1831, + "TGACAG": 1832, + "TGACTA": 1833, + "TGACTT": 1834, + "TGACTC": 1835, + "TGACTG": 1836, + "TGACCA": 1837, + "TGACCT": 1838, + "TGACCC": 1839, + "TGACCG": 1840, + "TGACGA": 1841, + "TGACGT": 1842, + "TGACGC": 1843, + "TGACGG": 1844, + "TGAGAA": 1845, + "TGAGAT": 1846, + "TGAGAC": 1847, + "TGAGAG": 1848, + "TGAGTA": 1849, + "TGAGTT": 1850, + "TGAGTC": 1851, + "TGAGTG": 1852, + "TGAGCA": 1853, + "TGAGCT": 1854, + "TGAGCC": 1855, + "TGAGCG": 1856, + "TGAGGA": 1857, + "TGAGGT": 1858, + "TGAGGC": 1859, + "TGAGGG": 1860, + "TGTAAA": 1861, + "TGTAAT": 1862, + "TGTAAC": 1863, + "TGTAAG": 1864, + "TGTATA": 1865, + "TGTATT": 1866, + "TGTATC": 1867, + "TGTATG": 1868, + "TGTACA": 1869, + "TGTACT": 1870, + "TGTACC": 1871, + "TGTACG": 1872, + "TGTAGA": 1873, + "TGTAGT": 1874, + "TGTAGC": 1875, + "TGTAGG": 1876, + "TGTTAA": 1877, + "TGTTAT": 1878, + "TGTTAC": 1879, + "TGTTAG": 1880, + "TGTTTA": 1881, + "TGTTTT": 1882, + "TGTTTC": 1883, + "TGTTTG": 1884, + "TGTTCA": 1885, + "TGTTCT": 1886, + "TGTTCC": 1887, + "TGTTCG": 1888, + "TGTTGA": 1889, + "TGTTGT": 1890, + "TGTTGC": 1891, + "TGTTGG": 1892, + "TGTCAA": 1893, + "TGTCAT": 1894, + "TGTCAC": 1895, + "TGTCAG": 1896, + "TGTCTA": 1897, + "TGTCTT": 1898, + "TGTCTC": 1899, + "TGTCTG": 1900, + "TGTCCA": 1901, + "TGTCCT": 1902, + "TGTCCC": 1903, + "TGTCCG": 1904, + "TGTCGA": 1905, + "TGTCGT": 1906, + "TGTCGC": 1907, + "TGTCGG": 1908, + "TGTGAA": 1909, + "TGTGAT": 1910, + "TGTGAC": 1911, + "TGTGAG": 1912, + "TGTGTA": 1913, + "TGTGTT": 1914, + "TGTGTC": 1915, + "TGTGTG": 1916, + "TGTGCA": 1917, + "TGTGCT": 1918, + "TGTGCC": 1919, + "TGTGCG": 1920, + "TGTGGA": 1921, + "TGTGGT": 1922, + "TGTGGC": 1923, + "TGTGGG": 1924, + "TGCAAA": 1925, + "TGCAAT": 1926, + "TGCAAC": 1927, + "TGCAAG": 1928, + "TGCATA": 1929, + "TGCATT": 1930, + "TGCATC": 1931, + "TGCATG": 1932, + "TGCACA": 1933, + "TGCACT": 1934, + "TGCACC": 1935, + "TGCACG": 1936, + "TGCAGA": 1937, + "TGCAGT": 1938, + "TGCAGC": 1939, + "TGCAGG": 1940, + "TGCTAA": 1941, + "TGCTAT": 1942, + "TGCTAC": 1943, + "TGCTAG": 1944, + "TGCTTA": 1945, + "TGCTTT": 1946, + "TGCTTC": 1947, + "TGCTTG": 1948, + "TGCTCA": 1949, + "TGCTCT": 1950, + "TGCTCC": 1951, + "TGCTCG": 1952, + "TGCTGA": 1953, + "TGCTGT": 1954, + "TGCTGC": 1955, + "TGCTGG": 1956, + "TGCCAA": 1957, + "TGCCAT": 1958, + "TGCCAC": 1959, + "TGCCAG": 1960, + "TGCCTA": 1961, + "TGCCTT": 1962, + "TGCCTC": 1963, + "TGCCTG": 1964, + "TGCCCA": 1965, + "TGCCCT": 1966, + "TGCCCC": 1967, + "TGCCCG": 1968, + "TGCCGA": 1969, + "TGCCGT": 1970, + "TGCCGC": 1971, + "TGCCGG": 1972, + "TGCGAA": 1973, + "TGCGAT": 1974, + "TGCGAC": 1975, + "TGCGAG": 1976, + "TGCGTA": 1977, + "TGCGTT": 1978, + "TGCGTC": 1979, + "TGCGTG": 1980, + "TGCGCA": 1981, + "TGCGCT": 1982, + "TGCGCC": 1983, + "TGCGCG": 1984, + "TGCGGA": 1985, + "TGCGGT": 1986, + "TGCGGC": 1987, + "TGCGGG": 1988, + "TGGAAA": 1989, + "TGGAAT": 1990, + "TGGAAC": 1991, + "TGGAAG": 1992, + "TGGATA": 1993, + "TGGATT": 1994, + "TGGATC": 1995, + "TGGATG": 1996, + "TGGACA": 1997, + "TGGACT": 1998, + "TGGACC": 1999, + "TGGACG": 2000, + "TGGAGA": 2001, + "TGGAGT": 2002, + "TGGAGC": 2003, + "TGGAGG": 2004, + "TGGTAA": 2005, + "TGGTAT": 2006, + "TGGTAC": 2007, + "TGGTAG": 2008, + "TGGTTA": 2009, + "TGGTTT": 2010, + "TGGTTC": 2011, + "TGGTTG": 2012, + "TGGTCA": 2013, + "TGGTCT": 2014, + "TGGTCC": 2015, + "TGGTCG": 2016, + "TGGTGA": 2017, + "TGGTGT": 2018, + "TGGTGC": 2019, + "TGGTGG": 2020, + "TGGCAA": 2021, + "TGGCAT": 2022, + "TGGCAC": 2023, + "TGGCAG": 2024, + "TGGCTA": 2025, + "TGGCTT": 2026, + "TGGCTC": 2027, + "TGGCTG": 2028, + "TGGCCA": 2029, + "TGGCCT": 2030, + "TGGCCC": 2031, + "TGGCCG": 2032, + "TGGCGA": 2033, + "TGGCGT": 2034, + "TGGCGC": 2035, + "TGGCGG": 2036, + "TGGGAA": 2037, + "TGGGAT": 2038, + "TGGGAC": 2039, + "TGGGAG": 2040, + "TGGGTA": 2041, + "TGGGTT": 2042, + "TGGGTC": 2043, + "TGGGTG": 2044, + "TGGGCA": 2045, + "TGGGCT": 2046, + "TGGGCC": 2047, + "TGGGCG": 2048, + "TGGGGA": 2049, + "TGGGGT": 2050, + "TGGGGC": 2051, + "TGGGGG": 2052, + "CAAAAA": 2053, + "CAAAAT": 2054, + "CAAAAC": 2055, + "CAAAAG": 2056, + "CAAATA": 2057, + "CAAATT": 2058, + "CAAATC": 2059, + "CAAATG": 2060, + "CAAACA": 2061, + "CAAACT": 2062, + "CAAACC": 2063, + "CAAACG": 2064, + "CAAAGA": 2065, + "CAAAGT": 2066, + "CAAAGC": 2067, + "CAAAGG": 2068, + "CAATAA": 2069, + "CAATAT": 2070, + "CAATAC": 2071, + "CAATAG": 2072, + "CAATTA": 2073, + "CAATTT": 2074, + "CAATTC": 2075, + "CAATTG": 2076, + "CAATCA": 2077, + "CAATCT": 2078, + "CAATCC": 2079, + "CAATCG": 2080, + "CAATGA": 2081, + "CAATGT": 2082, + "CAATGC": 2083, + "CAATGG": 2084, + "CAACAA": 2085, + "CAACAT": 2086, + "CAACAC": 2087, + "CAACAG": 2088, + "CAACTA": 2089, + "CAACTT": 2090, + "CAACTC": 2091, + "CAACTG": 2092, + "CAACCA": 2093, + "CAACCT": 2094, + "CAACCC": 2095, + "CAACCG": 2096, + "CAACGA": 2097, + "CAACGT": 2098, + "CAACGC": 2099, + "CAACGG": 2100, + "CAAGAA": 2101, + "CAAGAT": 2102, + "CAAGAC": 2103, + "CAAGAG": 2104, + "CAAGTA": 2105, + "CAAGTT": 2106, + "CAAGTC": 2107, + "CAAGTG": 2108, + "CAAGCA": 2109, + "CAAGCT": 2110, + "CAAGCC": 2111, + "CAAGCG": 2112, + "CAAGGA": 2113, + "CAAGGT": 2114, + "CAAGGC": 2115, + "CAAGGG": 2116, + "CATAAA": 2117, + "CATAAT": 2118, + "CATAAC": 2119, + "CATAAG": 2120, + "CATATA": 2121, + "CATATT": 2122, + "CATATC": 2123, + "CATATG": 2124, + "CATACA": 2125, + "CATACT": 2126, + "CATACC": 2127, + "CATACG": 2128, + "CATAGA": 2129, + "CATAGT": 2130, + "CATAGC": 2131, + "CATAGG": 2132, + "CATTAA": 2133, + "CATTAT": 2134, + "CATTAC": 2135, + "CATTAG": 2136, + "CATTTA": 2137, + "CATTTT": 2138, + "CATTTC": 2139, + "CATTTG": 2140, + "CATTCA": 2141, + "CATTCT": 2142, + "CATTCC": 2143, + "CATTCG": 2144, + "CATTGA": 2145, + "CATTGT": 2146, + "CATTGC": 2147, + "CATTGG": 2148, + "CATCAA": 2149, + "CATCAT": 2150, + "CATCAC": 2151, + "CATCAG": 2152, + "CATCTA": 2153, + "CATCTT": 2154, + "CATCTC": 2155, + "CATCTG": 2156, + "CATCCA": 2157, + "CATCCT": 2158, + "CATCCC": 2159, + "CATCCG": 2160, + "CATCGA": 2161, + "CATCGT": 2162, + "CATCGC": 2163, + "CATCGG": 2164, + "CATGAA": 2165, + "CATGAT": 2166, + "CATGAC": 2167, + "CATGAG": 2168, + "CATGTA": 2169, + "CATGTT": 2170, + "CATGTC": 2171, + "CATGTG": 2172, + "CATGCA": 2173, + "CATGCT": 2174, + "CATGCC": 2175, + "CATGCG": 2176, + "CATGGA": 2177, + "CATGGT": 2178, + "CATGGC": 2179, + "CATGGG": 2180, + "CACAAA": 2181, + "CACAAT": 2182, + "CACAAC": 2183, + "CACAAG": 2184, + "CACATA": 2185, + "CACATT": 2186, + "CACATC": 2187, + "CACATG": 2188, + "CACACA": 2189, + "CACACT": 2190, + "CACACC": 2191, + "CACACG": 2192, + "CACAGA": 2193, + "CACAGT": 2194, + "CACAGC": 2195, + "CACAGG": 2196, + "CACTAA": 2197, + "CACTAT": 2198, + "CACTAC": 2199, + "CACTAG": 2200, + "CACTTA": 2201, + "CACTTT": 2202, + "CACTTC": 2203, + "CACTTG": 2204, + "CACTCA": 2205, + "CACTCT": 2206, + "CACTCC": 2207, + "CACTCG": 2208, + "CACTGA": 2209, + "CACTGT": 2210, + "CACTGC": 2211, + "CACTGG": 2212, + "CACCAA": 2213, + "CACCAT": 2214, + "CACCAC": 2215, + "CACCAG": 2216, + "CACCTA": 2217, + "CACCTT": 2218, + "CACCTC": 2219, + "CACCTG": 2220, + "CACCCA": 2221, + "CACCCT": 2222, + "CACCCC": 2223, + "CACCCG": 2224, + "CACCGA": 2225, + "CACCGT": 2226, + "CACCGC": 2227, + "CACCGG": 2228, + "CACGAA": 2229, + "CACGAT": 2230, + "CACGAC": 2231, + "CACGAG": 2232, + "CACGTA": 2233, + "CACGTT": 2234, + "CACGTC": 2235, + "CACGTG": 2236, + "CACGCA": 2237, + "CACGCT": 2238, + "CACGCC": 2239, + "CACGCG": 2240, + "CACGGA": 2241, + "CACGGT": 2242, + "CACGGC": 2243, + "CACGGG": 2244, + "CAGAAA": 2245, + "CAGAAT": 2246, + "CAGAAC": 2247, + "CAGAAG": 2248, + "CAGATA": 2249, + "CAGATT": 2250, + "CAGATC": 2251, + "CAGATG": 2252, + "CAGACA": 2253, + "CAGACT": 2254, + "CAGACC": 2255, + "CAGACG": 2256, + "CAGAGA": 2257, + "CAGAGT": 2258, + "CAGAGC": 2259, + "CAGAGG": 2260, + "CAGTAA": 2261, + "CAGTAT": 2262, + "CAGTAC": 2263, + "CAGTAG": 2264, + "CAGTTA": 2265, + "CAGTTT": 2266, + "CAGTTC": 2267, + "CAGTTG": 2268, + "CAGTCA": 2269, + "CAGTCT": 2270, + "CAGTCC": 2271, + "CAGTCG": 2272, + "CAGTGA": 2273, + "CAGTGT": 2274, + "CAGTGC": 2275, + "CAGTGG": 2276, + "CAGCAA": 2277, + "CAGCAT": 2278, + "CAGCAC": 2279, + "CAGCAG": 2280, + "CAGCTA": 2281, + "CAGCTT": 2282, + "CAGCTC": 2283, + "CAGCTG": 2284, + "CAGCCA": 2285, + "CAGCCT": 2286, + "CAGCCC": 2287, + "CAGCCG": 2288, + "CAGCGA": 2289, + "CAGCGT": 2290, + "CAGCGC": 2291, + "CAGCGG": 2292, + "CAGGAA": 2293, + "CAGGAT": 2294, + "CAGGAC": 2295, + "CAGGAG": 2296, + "CAGGTA": 2297, + "CAGGTT": 2298, + "CAGGTC": 2299, + "CAGGTG": 2300, + "CAGGCA": 2301, + "CAGGCT": 2302, + "CAGGCC": 2303, + "CAGGCG": 2304, + "CAGGGA": 2305, + "CAGGGT": 2306, + "CAGGGC": 2307, + "CAGGGG": 2308, + "CTAAAA": 2309, + "CTAAAT": 2310, + "CTAAAC": 2311, + "CTAAAG": 2312, + "CTAATA": 2313, + "CTAATT": 2314, + "CTAATC": 2315, + "CTAATG": 2316, + "CTAACA": 2317, + "CTAACT": 2318, + "CTAACC": 2319, + "CTAACG": 2320, + "CTAAGA": 2321, + "CTAAGT": 2322, + "CTAAGC": 2323, + "CTAAGG": 2324, + "CTATAA": 2325, + "CTATAT": 2326, + "CTATAC": 2327, + "CTATAG": 2328, + "CTATTA": 2329, + "CTATTT": 2330, + "CTATTC": 2331, + "CTATTG": 2332, + "CTATCA": 2333, + "CTATCT": 2334, + "CTATCC": 2335, + "CTATCG": 2336, + "CTATGA": 2337, + "CTATGT": 2338, + "CTATGC": 2339, + "CTATGG": 2340, + "CTACAA": 2341, + "CTACAT": 2342, + "CTACAC": 2343, + "CTACAG": 2344, + "CTACTA": 2345, + "CTACTT": 2346, + "CTACTC": 2347, + "CTACTG": 2348, + "CTACCA": 2349, + "CTACCT": 2350, + "CTACCC": 2351, + "CTACCG": 2352, + "CTACGA": 2353, + "CTACGT": 2354, + "CTACGC": 2355, + "CTACGG": 2356, + "CTAGAA": 2357, + "CTAGAT": 2358, + "CTAGAC": 2359, + "CTAGAG": 2360, + "CTAGTA": 2361, + "CTAGTT": 2362, + "CTAGTC": 2363, + "CTAGTG": 2364, + "CTAGCA": 2365, + "CTAGCT": 2366, + "CTAGCC": 2367, + "CTAGCG": 2368, + "CTAGGA": 2369, + "CTAGGT": 2370, + "CTAGGC": 2371, + "CTAGGG": 2372, + "CTTAAA": 2373, + "CTTAAT": 2374, + "CTTAAC": 2375, + "CTTAAG": 2376, + "CTTATA": 2377, + "CTTATT": 2378, + "CTTATC": 2379, + "CTTATG": 2380, + "CTTACA": 2381, + "CTTACT": 2382, + "CTTACC": 2383, + "CTTACG": 2384, + "CTTAGA": 2385, + "CTTAGT": 2386, + "CTTAGC": 2387, + "CTTAGG": 2388, + "CTTTAA": 2389, + "CTTTAT": 2390, + "CTTTAC": 2391, + "CTTTAG": 2392, + "CTTTTA": 2393, + "CTTTTT": 2394, + "CTTTTC": 2395, + "CTTTTG": 2396, + "CTTTCA": 2397, + "CTTTCT": 2398, + "CTTTCC": 2399, + "CTTTCG": 2400, + "CTTTGA": 2401, + "CTTTGT": 2402, + "CTTTGC": 2403, + "CTTTGG": 2404, + "CTTCAA": 2405, + "CTTCAT": 2406, + "CTTCAC": 2407, + "CTTCAG": 2408, + "CTTCTA": 2409, + "CTTCTT": 2410, + "CTTCTC": 2411, + "CTTCTG": 2412, + "CTTCCA": 2413, + "CTTCCT": 2414, + "CTTCCC": 2415, + "CTTCCG": 2416, + "CTTCGA": 2417, + "CTTCGT": 2418, + "CTTCGC": 2419, + "CTTCGG": 2420, + "CTTGAA": 2421, + "CTTGAT": 2422, + "CTTGAC": 2423, + "CTTGAG": 2424, + "CTTGTA": 2425, + "CTTGTT": 2426, + "CTTGTC": 2427, + "CTTGTG": 2428, + "CTTGCA": 2429, + "CTTGCT": 2430, + "CTTGCC": 2431, + "CTTGCG": 2432, + "CTTGGA": 2433, + "CTTGGT": 2434, + "CTTGGC": 2435, + "CTTGGG": 2436, + "CTCAAA": 2437, + "CTCAAT": 2438, + "CTCAAC": 2439, + "CTCAAG": 2440, + "CTCATA": 2441, + "CTCATT": 2442, + "CTCATC": 2443, + "CTCATG": 2444, + "CTCACA": 2445, + "CTCACT": 2446, + "CTCACC": 2447, + "CTCACG": 2448, + "CTCAGA": 2449, + "CTCAGT": 2450, + "CTCAGC": 2451, + "CTCAGG": 2452, + "CTCTAA": 2453, + "CTCTAT": 2454, + "CTCTAC": 2455, + "CTCTAG": 2456, + "CTCTTA": 2457, + "CTCTTT": 2458, + "CTCTTC": 2459, + "CTCTTG": 2460, + "CTCTCA": 2461, + "CTCTCT": 2462, + "CTCTCC": 2463, + "CTCTCG": 2464, + "CTCTGA": 2465, + "CTCTGT": 2466, + "CTCTGC": 2467, + "CTCTGG": 2468, + "CTCCAA": 2469, + "CTCCAT": 2470, + "CTCCAC": 2471, + "CTCCAG": 2472, + "CTCCTA": 2473, + "CTCCTT": 2474, + "CTCCTC": 2475, + "CTCCTG": 2476, + "CTCCCA": 2477, + "CTCCCT": 2478, + "CTCCCC": 2479, + "CTCCCG": 2480, + "CTCCGA": 2481, + "CTCCGT": 2482, + "CTCCGC": 2483, + "CTCCGG": 2484, + "CTCGAA": 2485, + "CTCGAT": 2486, + "CTCGAC": 2487, + "CTCGAG": 2488, + "CTCGTA": 2489, + "CTCGTT": 2490, + "CTCGTC": 2491, + "CTCGTG": 2492, + "CTCGCA": 2493, + "CTCGCT": 2494, + "CTCGCC": 2495, + "CTCGCG": 2496, + "CTCGGA": 2497, + "CTCGGT": 2498, + "CTCGGC": 2499, + "CTCGGG": 2500, + "CTGAAA": 2501, + "CTGAAT": 2502, + "CTGAAC": 2503, + "CTGAAG": 2504, + "CTGATA": 2505, + "CTGATT": 2506, + "CTGATC": 2507, + "CTGATG": 2508, + "CTGACA": 2509, + "CTGACT": 2510, + "CTGACC": 2511, + "CTGACG": 2512, + "CTGAGA": 2513, + "CTGAGT": 2514, + "CTGAGC": 2515, + "CTGAGG": 2516, + "CTGTAA": 2517, + "CTGTAT": 2518, + "CTGTAC": 2519, + "CTGTAG": 2520, + "CTGTTA": 2521, + "CTGTTT": 2522, + "CTGTTC": 2523, + "CTGTTG": 2524, + "CTGTCA": 2525, + "CTGTCT": 2526, + "CTGTCC": 2527, + "CTGTCG": 2528, + "CTGTGA": 2529, + "CTGTGT": 2530, + "CTGTGC": 2531, + "CTGTGG": 2532, + "CTGCAA": 2533, + "CTGCAT": 2534, + "CTGCAC": 2535, + "CTGCAG": 2536, + "CTGCTA": 2537, + "CTGCTT": 2538, + "CTGCTC": 2539, + "CTGCTG": 2540, + "CTGCCA": 2541, + "CTGCCT": 2542, + "CTGCCC": 2543, + "CTGCCG": 2544, + "CTGCGA": 2545, + "CTGCGT": 2546, + "CTGCGC": 2547, + "CTGCGG": 2548, + "CTGGAA": 2549, + "CTGGAT": 2550, + "CTGGAC": 2551, + "CTGGAG": 2552, + "CTGGTA": 2553, + "CTGGTT": 2554, + "CTGGTC": 2555, + "CTGGTG": 2556, + "CTGGCA": 2557, + "CTGGCT": 2558, + "CTGGCC": 2559, + "CTGGCG": 2560, + "CTGGGA": 2561, + "CTGGGT": 2562, + "CTGGGC": 2563, + "CTGGGG": 2564, + "CCAAAA": 2565, + "CCAAAT": 2566, + "CCAAAC": 2567, + "CCAAAG": 2568, + "CCAATA": 2569, + "CCAATT": 2570, + "CCAATC": 2571, + "CCAATG": 2572, + "CCAACA": 2573, + "CCAACT": 2574, + "CCAACC": 2575, + "CCAACG": 2576, + "CCAAGA": 2577, + "CCAAGT": 2578, + "CCAAGC": 2579, + "CCAAGG": 2580, + "CCATAA": 2581, + "CCATAT": 2582, + "CCATAC": 2583, + "CCATAG": 2584, + "CCATTA": 2585, + "CCATTT": 2586, + "CCATTC": 2587, + "CCATTG": 2588, + "CCATCA": 2589, + "CCATCT": 2590, + "CCATCC": 2591, + "CCATCG": 2592, + "CCATGA": 2593, + "CCATGT": 2594, + "CCATGC": 2595, + "CCATGG": 2596, + "CCACAA": 2597, + "CCACAT": 2598, + "CCACAC": 2599, + "CCACAG": 2600, + "CCACTA": 2601, + "CCACTT": 2602, + "CCACTC": 2603, + "CCACTG": 2604, + "CCACCA": 2605, + "CCACCT": 2606, + "CCACCC": 2607, + "CCACCG": 2608, + "CCACGA": 2609, + "CCACGT": 2610, + "CCACGC": 2611, + "CCACGG": 2612, + "CCAGAA": 2613, + "CCAGAT": 2614, + "CCAGAC": 2615, + "CCAGAG": 2616, + "CCAGTA": 2617, + "CCAGTT": 2618, + "CCAGTC": 2619, + "CCAGTG": 2620, + "CCAGCA": 2621, + "CCAGCT": 2622, + "CCAGCC": 2623, + "CCAGCG": 2624, + "CCAGGA": 2625, + "CCAGGT": 2626, + "CCAGGC": 2627, + "CCAGGG": 2628, + "CCTAAA": 2629, + "CCTAAT": 2630, + "CCTAAC": 2631, + "CCTAAG": 2632, + "CCTATA": 2633, + "CCTATT": 2634, + "CCTATC": 2635, + "CCTATG": 2636, + "CCTACA": 2637, + "CCTACT": 2638, + "CCTACC": 2639, + "CCTACG": 2640, + "CCTAGA": 2641, + "CCTAGT": 2642, + "CCTAGC": 2643, + "CCTAGG": 2644, + "CCTTAA": 2645, + "CCTTAT": 2646, + "CCTTAC": 2647, + "CCTTAG": 2648, + "CCTTTA": 2649, + "CCTTTT": 2650, + "CCTTTC": 2651, + "CCTTTG": 2652, + "CCTTCA": 2653, + "CCTTCT": 2654, + "CCTTCC": 2655, + "CCTTCG": 2656, + "CCTTGA": 2657, + "CCTTGT": 2658, + "CCTTGC": 2659, + "CCTTGG": 2660, + "CCTCAA": 2661, + "CCTCAT": 2662, + "CCTCAC": 2663, + "CCTCAG": 2664, + "CCTCTA": 2665, + "CCTCTT": 2666, + "CCTCTC": 2667, + "CCTCTG": 2668, + "CCTCCA": 2669, + "CCTCCT": 2670, + "CCTCCC": 2671, + "CCTCCG": 2672, + "CCTCGA": 2673, + "CCTCGT": 2674, + "CCTCGC": 2675, + "CCTCGG": 2676, + "CCTGAA": 2677, + "CCTGAT": 2678, + "CCTGAC": 2679, + "CCTGAG": 2680, + "CCTGTA": 2681, + "CCTGTT": 2682, + "CCTGTC": 2683, + "CCTGTG": 2684, + "CCTGCA": 2685, + "CCTGCT": 2686, + "CCTGCC": 2687, + "CCTGCG": 2688, + "CCTGGA": 2689, + "CCTGGT": 2690, + "CCTGGC": 2691, + "CCTGGG": 2692, + "CCCAAA": 2693, + "CCCAAT": 2694, + "CCCAAC": 2695, + "CCCAAG": 2696, + "CCCATA": 2697, + "CCCATT": 2698, + "CCCATC": 2699, + "CCCATG": 2700, + "CCCACA": 2701, + "CCCACT": 2702, + "CCCACC": 2703, + "CCCACG": 2704, + "CCCAGA": 2705, + "CCCAGT": 2706, + "CCCAGC": 2707, + "CCCAGG": 2708, + "CCCTAA": 2709, + "CCCTAT": 2710, + "CCCTAC": 2711, + "CCCTAG": 2712, + "CCCTTA": 2713, + "CCCTTT": 2714, + "CCCTTC": 2715, + "CCCTTG": 2716, + "CCCTCA": 2717, + "CCCTCT": 2718, + "CCCTCC": 2719, + "CCCTCG": 2720, + "CCCTGA": 2721, + "CCCTGT": 2722, + "CCCTGC": 2723, + "CCCTGG": 2724, + "CCCCAA": 2725, + "CCCCAT": 2726, + "CCCCAC": 2727, + "CCCCAG": 2728, + "CCCCTA": 2729, + "CCCCTT": 2730, + "CCCCTC": 2731, + "CCCCTG": 2732, + "CCCCCA": 2733, + "CCCCCT": 2734, + "CCCCCC": 2735, + "CCCCCG": 2736, + "CCCCGA": 2737, + "CCCCGT": 2738, + "CCCCGC": 2739, + "CCCCGG": 2740, + "CCCGAA": 2741, + "CCCGAT": 2742, + "CCCGAC": 2743, + "CCCGAG": 2744, + "CCCGTA": 2745, + "CCCGTT": 2746, + "CCCGTC": 2747, + "CCCGTG": 2748, + "CCCGCA": 2749, + "CCCGCT": 2750, + "CCCGCC": 2751, + "CCCGCG": 2752, + "CCCGGA": 2753, + "CCCGGT": 2754, + "CCCGGC": 2755, + "CCCGGG": 2756, + "CCGAAA": 2757, + "CCGAAT": 2758, + "CCGAAC": 2759, + "CCGAAG": 2760, + "CCGATA": 2761, + "CCGATT": 2762, + "CCGATC": 2763, + "CCGATG": 2764, + "CCGACA": 2765, + "CCGACT": 2766, + "CCGACC": 2767, + "CCGACG": 2768, + "CCGAGA": 2769, + "CCGAGT": 2770, + "CCGAGC": 2771, + "CCGAGG": 2772, + "CCGTAA": 2773, + "CCGTAT": 2774, + "CCGTAC": 2775, + "CCGTAG": 2776, + "CCGTTA": 2777, + "CCGTTT": 2778, + "CCGTTC": 2779, + "CCGTTG": 2780, + "CCGTCA": 2781, + "CCGTCT": 2782, + "CCGTCC": 2783, + "CCGTCG": 2784, + "CCGTGA": 2785, + "CCGTGT": 2786, + "CCGTGC": 2787, + "CCGTGG": 2788, + "CCGCAA": 2789, + "CCGCAT": 2790, + "CCGCAC": 2791, + "CCGCAG": 2792, + "CCGCTA": 2793, + "CCGCTT": 2794, + "CCGCTC": 2795, + "CCGCTG": 2796, + "CCGCCA": 2797, + "CCGCCT": 2798, + "CCGCCC": 2799, + "CCGCCG": 2800, + "CCGCGA": 2801, + "CCGCGT": 2802, + "CCGCGC": 2803, + "CCGCGG": 2804, + "CCGGAA": 2805, + "CCGGAT": 2806, + "CCGGAC": 2807, + "CCGGAG": 2808, + "CCGGTA": 2809, + "CCGGTT": 2810, + "CCGGTC": 2811, + "CCGGTG": 2812, + "CCGGCA": 2813, + "CCGGCT": 2814, + "CCGGCC": 2815, + "CCGGCG": 2816, + "CCGGGA": 2817, + "CCGGGT": 2818, + "CCGGGC": 2819, + "CCGGGG": 2820, + "CGAAAA": 2821, + "CGAAAT": 2822, + "CGAAAC": 2823, + "CGAAAG": 2824, + "CGAATA": 2825, + "CGAATT": 2826, + "CGAATC": 2827, + "CGAATG": 2828, + "CGAACA": 2829, + "CGAACT": 2830, + "CGAACC": 2831, + "CGAACG": 2832, + "CGAAGA": 2833, + "CGAAGT": 2834, + "CGAAGC": 2835, + "CGAAGG": 2836, + "CGATAA": 2837, + "CGATAT": 2838, + "CGATAC": 2839, + "CGATAG": 2840, + "CGATTA": 2841, + "CGATTT": 2842, + "CGATTC": 2843, + "CGATTG": 2844, + "CGATCA": 2845, + "CGATCT": 2846, + "CGATCC": 2847, + "CGATCG": 2848, + "CGATGA": 2849, + "CGATGT": 2850, + "CGATGC": 2851, + "CGATGG": 2852, + "CGACAA": 2853, + "CGACAT": 2854, + "CGACAC": 2855, + "CGACAG": 2856, + "CGACTA": 2857, + "CGACTT": 2858, + "CGACTC": 2859, + "CGACTG": 2860, + "CGACCA": 2861, + "CGACCT": 2862, + "CGACCC": 2863, + "CGACCG": 2864, + "CGACGA": 2865, + "CGACGT": 2866, + "CGACGC": 2867, + "CGACGG": 2868, + "CGAGAA": 2869, + "CGAGAT": 2870, + "CGAGAC": 2871, + "CGAGAG": 2872, + "CGAGTA": 2873, + "CGAGTT": 2874, + "CGAGTC": 2875, + "CGAGTG": 2876, + "CGAGCA": 2877, + "CGAGCT": 2878, + "CGAGCC": 2879, + "CGAGCG": 2880, + "CGAGGA": 2881, + "CGAGGT": 2882, + "CGAGGC": 2883, + "CGAGGG": 2884, + "CGTAAA": 2885, + "CGTAAT": 2886, + "CGTAAC": 2887, + "CGTAAG": 2888, + "CGTATA": 2889, + "CGTATT": 2890, + "CGTATC": 2891, + "CGTATG": 2892, + "CGTACA": 2893, + "CGTACT": 2894, + "CGTACC": 2895, + "CGTACG": 2896, + "CGTAGA": 2897, + "CGTAGT": 2898, + "CGTAGC": 2899, + "CGTAGG": 2900, + "CGTTAA": 2901, + "CGTTAT": 2902, + "CGTTAC": 2903, + "CGTTAG": 2904, + "CGTTTA": 2905, + "CGTTTT": 2906, + "CGTTTC": 2907, + "CGTTTG": 2908, + "CGTTCA": 2909, + "CGTTCT": 2910, + "CGTTCC": 2911, + "CGTTCG": 2912, + "CGTTGA": 2913, + "CGTTGT": 2914, + "CGTTGC": 2915, + "CGTTGG": 2916, + "CGTCAA": 2917, + "CGTCAT": 2918, + "CGTCAC": 2919, + "CGTCAG": 2920, + "CGTCTA": 2921, + "CGTCTT": 2922, + "CGTCTC": 2923, + "CGTCTG": 2924, + "CGTCCA": 2925, + "CGTCCT": 2926, + "CGTCCC": 2927, + "CGTCCG": 2928, + "CGTCGA": 2929, + "CGTCGT": 2930, + "CGTCGC": 2931, + "CGTCGG": 2932, + "CGTGAA": 2933, + "CGTGAT": 2934, + "CGTGAC": 2935, + "CGTGAG": 2936, + "CGTGTA": 2937, + "CGTGTT": 2938, + "CGTGTC": 2939, + "CGTGTG": 2940, + "CGTGCA": 2941, + "CGTGCT": 2942, + "CGTGCC": 2943, + "CGTGCG": 2944, + "CGTGGA": 2945, + "CGTGGT": 2946, + "CGTGGC": 2947, + "CGTGGG": 2948, + "CGCAAA": 2949, + "CGCAAT": 2950, + "CGCAAC": 2951, + "CGCAAG": 2952, + "CGCATA": 2953, + "CGCATT": 2954, + "CGCATC": 2955, + "CGCATG": 2956, + "CGCACA": 2957, + "CGCACT": 2958, + "CGCACC": 2959, + "CGCACG": 2960, + "CGCAGA": 2961, + "CGCAGT": 2962, + "CGCAGC": 2963, + "CGCAGG": 2964, + "CGCTAA": 2965, + "CGCTAT": 2966, + "CGCTAC": 2967, + "CGCTAG": 2968, + "CGCTTA": 2969, + "CGCTTT": 2970, + "CGCTTC": 2971, + "CGCTTG": 2972, + "CGCTCA": 2973, + "CGCTCT": 2974, + "CGCTCC": 2975, + "CGCTCG": 2976, + "CGCTGA": 2977, + "CGCTGT": 2978, + "CGCTGC": 2979, + "CGCTGG": 2980, + "CGCCAA": 2981, + "CGCCAT": 2982, + "CGCCAC": 2983, + "CGCCAG": 2984, + "CGCCTA": 2985, + "CGCCTT": 2986, + "CGCCTC": 2987, + "CGCCTG": 2988, + "CGCCCA": 2989, + "CGCCCT": 2990, + "CGCCCC": 2991, + "CGCCCG": 2992, + "CGCCGA": 2993, + "CGCCGT": 2994, + "CGCCGC": 2995, + "CGCCGG": 2996, + "CGCGAA": 2997, + "CGCGAT": 2998, + "CGCGAC": 2999, + "CGCGAG": 3000, + "CGCGTA": 3001, + "CGCGTT": 3002, + "CGCGTC": 3003, + "CGCGTG": 3004, + "CGCGCA": 3005, + "CGCGCT": 3006, + "CGCGCC": 3007, + "CGCGCG": 3008, + "CGCGGA": 3009, + "CGCGGT": 3010, + "CGCGGC": 3011, + "CGCGGG": 3012, + "CGGAAA": 3013, + "CGGAAT": 3014, + "CGGAAC": 3015, + "CGGAAG": 3016, + "CGGATA": 3017, + "CGGATT": 3018, + "CGGATC": 3019, + "CGGATG": 3020, + "CGGACA": 3021, + "CGGACT": 3022, + "CGGACC": 3023, + "CGGACG": 3024, + "CGGAGA": 3025, + "CGGAGT": 3026, + "CGGAGC": 3027, + "CGGAGG": 3028, + "CGGTAA": 3029, + "CGGTAT": 3030, + "CGGTAC": 3031, + "CGGTAG": 3032, + "CGGTTA": 3033, + "CGGTTT": 3034, + "CGGTTC": 3035, + "CGGTTG": 3036, + "CGGTCA": 3037, + "CGGTCT": 3038, + "CGGTCC": 3039, + "CGGTCG": 3040, + "CGGTGA": 3041, + "CGGTGT": 3042, + "CGGTGC": 3043, + "CGGTGG": 3044, + "CGGCAA": 3045, + "CGGCAT": 3046, + "CGGCAC": 3047, + "CGGCAG": 3048, + "CGGCTA": 3049, + "CGGCTT": 3050, + "CGGCTC": 3051, + "CGGCTG": 3052, + "CGGCCA": 3053, + "CGGCCT": 3054, + "CGGCCC": 3055, + "CGGCCG": 3056, + "CGGCGA": 3057, + "CGGCGT": 3058, + "CGGCGC": 3059, + "CGGCGG": 3060, + "CGGGAA": 3061, + "CGGGAT": 3062, + "CGGGAC": 3063, + "CGGGAG": 3064, + "CGGGTA": 3065, + "CGGGTT": 3066, + "CGGGTC": 3067, + "CGGGTG": 3068, + "CGGGCA": 3069, + "CGGGCT": 3070, + "CGGGCC": 3071, + "CGGGCG": 3072, + "CGGGGA": 3073, + "CGGGGT": 3074, + "CGGGGC": 3075, + "CGGGGG": 3076, + "GAAAAA": 3077, + "GAAAAT": 3078, + "GAAAAC": 3079, + "GAAAAG": 3080, + "GAAATA": 3081, + "GAAATT": 3082, + "GAAATC": 3083, + "GAAATG": 3084, + "GAAACA": 3085, + "GAAACT": 3086, + "GAAACC": 3087, + "GAAACG": 3088, + "GAAAGA": 3089, + "GAAAGT": 3090, + "GAAAGC": 3091, + "GAAAGG": 3092, + "GAATAA": 3093, + "GAATAT": 3094, + "GAATAC": 3095, + "GAATAG": 3096, + "GAATTA": 3097, + "GAATTT": 3098, + "GAATTC": 3099, + "GAATTG": 3100, + "GAATCA": 3101, + "GAATCT": 3102, + "GAATCC": 3103, + "GAATCG": 3104, + "GAATGA": 3105, + "GAATGT": 3106, + "GAATGC": 3107, + "GAATGG": 3108, + "GAACAA": 3109, + "GAACAT": 3110, + "GAACAC": 3111, + "GAACAG": 3112, + "GAACTA": 3113, + "GAACTT": 3114, + "GAACTC": 3115, + "GAACTG": 3116, + "GAACCA": 3117, + "GAACCT": 3118, + "GAACCC": 3119, + "GAACCG": 3120, + "GAACGA": 3121, + "GAACGT": 3122, + "GAACGC": 3123, + "GAACGG": 3124, + "GAAGAA": 3125, + "GAAGAT": 3126, + "GAAGAC": 3127, + "GAAGAG": 3128, + "GAAGTA": 3129, + "GAAGTT": 3130, + "GAAGTC": 3131, + "GAAGTG": 3132, + "GAAGCA": 3133, + "GAAGCT": 3134, + "GAAGCC": 3135, + "GAAGCG": 3136, + "GAAGGA": 3137, + "GAAGGT": 3138, + "GAAGGC": 3139, + "GAAGGG": 3140, + "GATAAA": 3141, + "GATAAT": 3142, + "GATAAC": 3143, + "GATAAG": 3144, + "GATATA": 3145, + "GATATT": 3146, + "GATATC": 3147, + "GATATG": 3148, + "GATACA": 3149, + "GATACT": 3150, + "GATACC": 3151, + "GATACG": 3152, + "GATAGA": 3153, + "GATAGT": 3154, + "GATAGC": 3155, + "GATAGG": 3156, + "GATTAA": 3157, + "GATTAT": 3158, + "GATTAC": 3159, + "GATTAG": 3160, + "GATTTA": 3161, + "GATTTT": 3162, + "GATTTC": 3163, + "GATTTG": 3164, + "GATTCA": 3165, + "GATTCT": 3166, + "GATTCC": 3167, + "GATTCG": 3168, + "GATTGA": 3169, + "GATTGT": 3170, + "GATTGC": 3171, + "GATTGG": 3172, + "GATCAA": 3173, + "GATCAT": 3174, + "GATCAC": 3175, + "GATCAG": 3176, + "GATCTA": 3177, + "GATCTT": 3178, + "GATCTC": 3179, + "GATCTG": 3180, + "GATCCA": 3181, + "GATCCT": 3182, + "GATCCC": 3183, + "GATCCG": 3184, + "GATCGA": 3185, + "GATCGT": 3186, + "GATCGC": 3187, + "GATCGG": 3188, + "GATGAA": 3189, + "GATGAT": 3190, + "GATGAC": 3191, + "GATGAG": 3192, + "GATGTA": 3193, + "GATGTT": 3194, + "GATGTC": 3195, + "GATGTG": 3196, + "GATGCA": 3197, + "GATGCT": 3198, + "GATGCC": 3199, + "GATGCG": 3200, + "GATGGA": 3201, + "GATGGT": 3202, + "GATGGC": 3203, + "GATGGG": 3204, + "GACAAA": 3205, + "GACAAT": 3206, + "GACAAC": 3207, + "GACAAG": 3208, + "GACATA": 3209, + "GACATT": 3210, + "GACATC": 3211, + "GACATG": 3212, + "GACACA": 3213, + "GACACT": 3214, + "GACACC": 3215, + "GACACG": 3216, + "GACAGA": 3217, + "GACAGT": 3218, + "GACAGC": 3219, + "GACAGG": 3220, + "GACTAA": 3221, + "GACTAT": 3222, + "GACTAC": 3223, + "GACTAG": 3224, + "GACTTA": 3225, + "GACTTT": 3226, + "GACTTC": 3227, + "GACTTG": 3228, + "GACTCA": 3229, + "GACTCT": 3230, + "GACTCC": 3231, + "GACTCG": 3232, + "GACTGA": 3233, + "GACTGT": 3234, + "GACTGC": 3235, + "GACTGG": 3236, + "GACCAA": 3237, + "GACCAT": 3238, + "GACCAC": 3239, + "GACCAG": 3240, + "GACCTA": 3241, + "GACCTT": 3242, + "GACCTC": 3243, + "GACCTG": 3244, + "GACCCA": 3245, + "GACCCT": 3246, + "GACCCC": 3247, + "GACCCG": 3248, + "GACCGA": 3249, + "GACCGT": 3250, + "GACCGC": 3251, + "GACCGG": 3252, + "GACGAA": 3253, + "GACGAT": 3254, + "GACGAC": 3255, + "GACGAG": 3256, + "GACGTA": 3257, + "GACGTT": 3258, + "GACGTC": 3259, + "GACGTG": 3260, + "GACGCA": 3261, + "GACGCT": 3262, + "GACGCC": 3263, + "GACGCG": 3264, + "GACGGA": 3265, + "GACGGT": 3266, + "GACGGC": 3267, + "GACGGG": 3268, + "GAGAAA": 3269, + "GAGAAT": 3270, + "GAGAAC": 3271, + "GAGAAG": 3272, + "GAGATA": 3273, + "GAGATT": 3274, + "GAGATC": 3275, + "GAGATG": 3276, + "GAGACA": 3277, + "GAGACT": 3278, + "GAGACC": 3279, + "GAGACG": 3280, + "GAGAGA": 3281, + "GAGAGT": 3282, + "GAGAGC": 3283, + "GAGAGG": 3284, + "GAGTAA": 3285, + "GAGTAT": 3286, + "GAGTAC": 3287, + "GAGTAG": 3288, + "GAGTTA": 3289, + "GAGTTT": 3290, + "GAGTTC": 3291, + "GAGTTG": 3292, + "GAGTCA": 3293, + "GAGTCT": 3294, + "GAGTCC": 3295, + "GAGTCG": 3296, + "GAGTGA": 3297, + "GAGTGT": 3298, + "GAGTGC": 3299, + "GAGTGG": 3300, + "GAGCAA": 3301, + "GAGCAT": 3302, + "GAGCAC": 3303, + "GAGCAG": 3304, + "GAGCTA": 3305, + "GAGCTT": 3306, + "GAGCTC": 3307, + "GAGCTG": 3308, + "GAGCCA": 3309, + "GAGCCT": 3310, + "GAGCCC": 3311, + "GAGCCG": 3312, + "GAGCGA": 3313, + "GAGCGT": 3314, + "GAGCGC": 3315, + "GAGCGG": 3316, + "GAGGAA": 3317, + "GAGGAT": 3318, + "GAGGAC": 3319, + "GAGGAG": 3320, + "GAGGTA": 3321, + "GAGGTT": 3322, + "GAGGTC": 3323, + "GAGGTG": 3324, + "GAGGCA": 3325, + "GAGGCT": 3326, + "GAGGCC": 3327, + "GAGGCG": 3328, + "GAGGGA": 3329, + "GAGGGT": 3330, + "GAGGGC": 3331, + "GAGGGG": 3332, + "GTAAAA": 3333, + "GTAAAT": 3334, + "GTAAAC": 3335, + "GTAAAG": 3336, + "GTAATA": 3337, + "GTAATT": 3338, + "GTAATC": 3339, + "GTAATG": 3340, + "GTAACA": 3341, + "GTAACT": 3342, + "GTAACC": 3343, + "GTAACG": 3344, + "GTAAGA": 3345, + "GTAAGT": 3346, + "GTAAGC": 3347, + "GTAAGG": 3348, + "GTATAA": 3349, + "GTATAT": 3350, + "GTATAC": 3351, + "GTATAG": 3352, + "GTATTA": 3353, + "GTATTT": 3354, + "GTATTC": 3355, + "GTATTG": 3356, + "GTATCA": 3357, + "GTATCT": 3358, + "GTATCC": 3359, + "GTATCG": 3360, + "GTATGA": 3361, + "GTATGT": 3362, + "GTATGC": 3363, + "GTATGG": 3364, + "GTACAA": 3365, + "GTACAT": 3366, + "GTACAC": 3367, + "GTACAG": 3368, + "GTACTA": 3369, + "GTACTT": 3370, + "GTACTC": 3371, + "GTACTG": 3372, + "GTACCA": 3373, + "GTACCT": 3374, + "GTACCC": 3375, + "GTACCG": 3376, + "GTACGA": 3377, + "GTACGT": 3378, + "GTACGC": 3379, + "GTACGG": 3380, + "GTAGAA": 3381, + "GTAGAT": 3382, + "GTAGAC": 3383, + "GTAGAG": 3384, + "GTAGTA": 3385, + "GTAGTT": 3386, + "GTAGTC": 3387, + "GTAGTG": 3388, + "GTAGCA": 3389, + "GTAGCT": 3390, + "GTAGCC": 3391, + "GTAGCG": 3392, + "GTAGGA": 3393, + "GTAGGT": 3394, + "GTAGGC": 3395, + "GTAGGG": 3396, + "GTTAAA": 3397, + "GTTAAT": 3398, + "GTTAAC": 3399, + "GTTAAG": 3400, + "GTTATA": 3401, + "GTTATT": 3402, + "GTTATC": 3403, + "GTTATG": 3404, + "GTTACA": 3405, + "GTTACT": 3406, + "GTTACC": 3407, + "GTTACG": 3408, + "GTTAGA": 3409, + "GTTAGT": 3410, + "GTTAGC": 3411, + "GTTAGG": 3412, + "GTTTAA": 3413, + "GTTTAT": 3414, + "GTTTAC": 3415, + "GTTTAG": 3416, + "GTTTTA": 3417, + "GTTTTT": 3418, + "GTTTTC": 3419, + "GTTTTG": 3420, + "GTTTCA": 3421, + "GTTTCT": 3422, + "GTTTCC": 3423, + "GTTTCG": 3424, + "GTTTGA": 3425, + "GTTTGT": 3426, + "GTTTGC": 3427, + "GTTTGG": 3428, + "GTTCAA": 3429, + "GTTCAT": 3430, + "GTTCAC": 3431, + "GTTCAG": 3432, + "GTTCTA": 3433, + "GTTCTT": 3434, + "GTTCTC": 3435, + "GTTCTG": 3436, + "GTTCCA": 3437, + "GTTCCT": 3438, + "GTTCCC": 3439, + "GTTCCG": 3440, + "GTTCGA": 3441, + "GTTCGT": 3442, + "GTTCGC": 3443, + "GTTCGG": 3444, + "GTTGAA": 3445, + "GTTGAT": 3446, + "GTTGAC": 3447, + "GTTGAG": 3448, + "GTTGTA": 3449, + "GTTGTT": 3450, + "GTTGTC": 3451, + "GTTGTG": 3452, + "GTTGCA": 3453, + "GTTGCT": 3454, + "GTTGCC": 3455, + "GTTGCG": 3456, + "GTTGGA": 3457, + "GTTGGT": 3458, + "GTTGGC": 3459, + "GTTGGG": 3460, + "GTCAAA": 3461, + "GTCAAT": 3462, + "GTCAAC": 3463, + "GTCAAG": 3464, + "GTCATA": 3465, + "GTCATT": 3466, + "GTCATC": 3467, + "GTCATG": 3468, + "GTCACA": 3469, + "GTCACT": 3470, + "GTCACC": 3471, + "GTCACG": 3472, + "GTCAGA": 3473, + "GTCAGT": 3474, + "GTCAGC": 3475, + "GTCAGG": 3476, + "GTCTAA": 3477, + "GTCTAT": 3478, + "GTCTAC": 3479, + "GTCTAG": 3480, + "GTCTTA": 3481, + "GTCTTT": 3482, + "GTCTTC": 3483, + "GTCTTG": 3484, + "GTCTCA": 3485, + "GTCTCT": 3486, + "GTCTCC": 3487, + "GTCTCG": 3488, + "GTCTGA": 3489, + "GTCTGT": 3490, + "GTCTGC": 3491, + "GTCTGG": 3492, + "GTCCAA": 3493, + "GTCCAT": 3494, + "GTCCAC": 3495, + "GTCCAG": 3496, + "GTCCTA": 3497, + "GTCCTT": 3498, + "GTCCTC": 3499, + "GTCCTG": 3500, + "GTCCCA": 3501, + "GTCCCT": 3502, + "GTCCCC": 3503, + "GTCCCG": 3504, + "GTCCGA": 3505, + "GTCCGT": 3506, + "GTCCGC": 3507, + "GTCCGG": 3508, + "GTCGAA": 3509, + "GTCGAT": 3510, + "GTCGAC": 3511, + "GTCGAG": 3512, + "GTCGTA": 3513, + "GTCGTT": 3514, + "GTCGTC": 3515, + "GTCGTG": 3516, + "GTCGCA": 3517, + "GTCGCT": 3518, + "GTCGCC": 3519, + "GTCGCG": 3520, + "GTCGGA": 3521, + "GTCGGT": 3522, + "GTCGGC": 3523, + "GTCGGG": 3524, + "GTGAAA": 3525, + "GTGAAT": 3526, + "GTGAAC": 3527, + "GTGAAG": 3528, + "GTGATA": 3529, + "GTGATT": 3530, + "GTGATC": 3531, + "GTGATG": 3532, + "GTGACA": 3533, + "GTGACT": 3534, + "GTGACC": 3535, + "GTGACG": 3536, + "GTGAGA": 3537, + "GTGAGT": 3538, + "GTGAGC": 3539, + "GTGAGG": 3540, + "GTGTAA": 3541, + "GTGTAT": 3542, + "GTGTAC": 3543, + "GTGTAG": 3544, + "GTGTTA": 3545, + "GTGTTT": 3546, + "GTGTTC": 3547, + "GTGTTG": 3548, + "GTGTCA": 3549, + "GTGTCT": 3550, + "GTGTCC": 3551, + "GTGTCG": 3552, + "GTGTGA": 3553, + "GTGTGT": 3554, + "GTGTGC": 3555, + "GTGTGG": 3556, + "GTGCAA": 3557, + "GTGCAT": 3558, + "GTGCAC": 3559, + "GTGCAG": 3560, + "GTGCTA": 3561, + "GTGCTT": 3562, + "GTGCTC": 3563, + "GTGCTG": 3564, + "GTGCCA": 3565, + "GTGCCT": 3566, + "GTGCCC": 3567, + "GTGCCG": 3568, + "GTGCGA": 3569, + "GTGCGT": 3570, + "GTGCGC": 3571, + "GTGCGG": 3572, + "GTGGAA": 3573, + "GTGGAT": 3574, + "GTGGAC": 3575, + "GTGGAG": 3576, + "GTGGTA": 3577, + "GTGGTT": 3578, + "GTGGTC": 3579, + "GTGGTG": 3580, + "GTGGCA": 3581, + "GTGGCT": 3582, + "GTGGCC": 3583, + "GTGGCG": 3584, + "GTGGGA": 3585, + "GTGGGT": 3586, + "GTGGGC": 3587, + "GTGGGG": 3588, + "GCAAAA": 3589, + "GCAAAT": 3590, + "GCAAAC": 3591, + "GCAAAG": 3592, + "GCAATA": 3593, + "GCAATT": 3594, + "GCAATC": 3595, + "GCAATG": 3596, + "GCAACA": 3597, + "GCAACT": 3598, + "GCAACC": 3599, + "GCAACG": 3600, + "GCAAGA": 3601, + "GCAAGT": 3602, + "GCAAGC": 3603, + "GCAAGG": 3604, + "GCATAA": 3605, + "GCATAT": 3606, + "GCATAC": 3607, + "GCATAG": 3608, + "GCATTA": 3609, + "GCATTT": 3610, + "GCATTC": 3611, + "GCATTG": 3612, + "GCATCA": 3613, + "GCATCT": 3614, + "GCATCC": 3615, + "GCATCG": 3616, + "GCATGA": 3617, + "GCATGT": 3618, + "GCATGC": 3619, + "GCATGG": 3620, + "GCACAA": 3621, + "GCACAT": 3622, + "GCACAC": 3623, + "GCACAG": 3624, + "GCACTA": 3625, + "GCACTT": 3626, + "GCACTC": 3627, + "GCACTG": 3628, + "GCACCA": 3629, + "GCACCT": 3630, + "GCACCC": 3631, + "GCACCG": 3632, + "GCACGA": 3633, + "GCACGT": 3634, + "GCACGC": 3635, + "GCACGG": 3636, + "GCAGAA": 3637, + "GCAGAT": 3638, + "GCAGAC": 3639, + "GCAGAG": 3640, + "GCAGTA": 3641, + "GCAGTT": 3642, + "GCAGTC": 3643, + "GCAGTG": 3644, + "GCAGCA": 3645, + "GCAGCT": 3646, + "GCAGCC": 3647, + "GCAGCG": 3648, + "GCAGGA": 3649, + "GCAGGT": 3650, + "GCAGGC": 3651, + "GCAGGG": 3652, + "GCTAAA": 3653, + "GCTAAT": 3654, + "GCTAAC": 3655, + "GCTAAG": 3656, + "GCTATA": 3657, + "GCTATT": 3658, + "GCTATC": 3659, + "GCTATG": 3660, + "GCTACA": 3661, + "GCTACT": 3662, + "GCTACC": 3663, + "GCTACG": 3664, + "GCTAGA": 3665, + "GCTAGT": 3666, + "GCTAGC": 3667, + "GCTAGG": 3668, + "GCTTAA": 3669, + "GCTTAT": 3670, + "GCTTAC": 3671, + "GCTTAG": 3672, + "GCTTTA": 3673, + "GCTTTT": 3674, + "GCTTTC": 3675, + "GCTTTG": 3676, + "GCTTCA": 3677, + "GCTTCT": 3678, + "GCTTCC": 3679, + "GCTTCG": 3680, + "GCTTGA": 3681, + "GCTTGT": 3682, + "GCTTGC": 3683, + "GCTTGG": 3684, + "GCTCAA": 3685, + "GCTCAT": 3686, + "GCTCAC": 3687, + "GCTCAG": 3688, + "GCTCTA": 3689, + "GCTCTT": 3690, + "GCTCTC": 3691, + "GCTCTG": 3692, + "GCTCCA": 3693, + "GCTCCT": 3694, + "GCTCCC": 3695, + "GCTCCG": 3696, + "GCTCGA": 3697, + "GCTCGT": 3698, + "GCTCGC": 3699, + "GCTCGG": 3700, + "GCTGAA": 3701, + "GCTGAT": 3702, + "GCTGAC": 3703, + "GCTGAG": 3704, + "GCTGTA": 3705, + "GCTGTT": 3706, + "GCTGTC": 3707, + "GCTGTG": 3708, + "GCTGCA": 3709, + "GCTGCT": 3710, + "GCTGCC": 3711, + "GCTGCG": 3712, + "GCTGGA": 3713, + "GCTGGT": 3714, + "GCTGGC": 3715, + "GCTGGG": 3716, + "GCCAAA": 3717, + "GCCAAT": 3718, + "GCCAAC": 3719, + "GCCAAG": 3720, + "GCCATA": 3721, + "GCCATT": 3722, + "GCCATC": 3723, + "GCCATG": 3724, + "GCCACA": 3725, + "GCCACT": 3726, + "GCCACC": 3727, + "GCCACG": 3728, + "GCCAGA": 3729, + "GCCAGT": 3730, + "GCCAGC": 3731, + "GCCAGG": 3732, + "GCCTAA": 3733, + "GCCTAT": 3734, + "GCCTAC": 3735, + "GCCTAG": 3736, + "GCCTTA": 3737, + "GCCTTT": 3738, + "GCCTTC": 3739, + "GCCTTG": 3740, + "GCCTCA": 3741, + "GCCTCT": 3742, + "GCCTCC": 3743, + "GCCTCG": 3744, + "GCCTGA": 3745, + "GCCTGT": 3746, + "GCCTGC": 3747, + "GCCTGG": 3748, + "GCCCAA": 3749, + "GCCCAT": 3750, + "GCCCAC": 3751, + "GCCCAG": 3752, + "GCCCTA": 3753, + "GCCCTT": 3754, + "GCCCTC": 3755, + "GCCCTG": 3756, + "GCCCCA": 3757, + "GCCCCT": 3758, + "GCCCCC": 3759, + "GCCCCG": 3760, + "GCCCGA": 3761, + "GCCCGT": 3762, + "GCCCGC": 3763, + "GCCCGG": 3764, + "GCCGAA": 3765, + "GCCGAT": 3766, + "GCCGAC": 3767, + "GCCGAG": 3768, + "GCCGTA": 3769, + "GCCGTT": 3770, + "GCCGTC": 3771, + "GCCGTG": 3772, + "GCCGCA": 3773, + "GCCGCT": 3774, + "GCCGCC": 3775, + "GCCGCG": 3776, + "GCCGGA": 3777, + "GCCGGT": 3778, + "GCCGGC": 3779, + "GCCGGG": 3780, + "GCGAAA": 3781, + "GCGAAT": 3782, + "GCGAAC": 3783, + "GCGAAG": 3784, + "GCGATA": 3785, + "GCGATT": 3786, + "GCGATC": 3787, + "GCGATG": 3788, + "GCGACA": 3789, + "GCGACT": 3790, + "GCGACC": 3791, + "GCGACG": 3792, + "GCGAGA": 3793, + "GCGAGT": 3794, + "GCGAGC": 3795, + "GCGAGG": 3796, + "GCGTAA": 3797, + "GCGTAT": 3798, + "GCGTAC": 3799, + "GCGTAG": 3800, + "GCGTTA": 3801, + "GCGTTT": 3802, + "GCGTTC": 3803, + "GCGTTG": 3804, + "GCGTCA": 3805, + "GCGTCT": 3806, + "GCGTCC": 3807, + "GCGTCG": 3808, + "GCGTGA": 3809, + "GCGTGT": 3810, + "GCGTGC": 3811, + "GCGTGG": 3812, + "GCGCAA": 3813, + "GCGCAT": 3814, + "GCGCAC": 3815, + "GCGCAG": 3816, + "GCGCTA": 3817, + "GCGCTT": 3818, + "GCGCTC": 3819, + "GCGCTG": 3820, + "GCGCCA": 3821, + "GCGCCT": 3822, + "GCGCCC": 3823, + "GCGCCG": 3824, + "GCGCGA": 3825, + "GCGCGT": 3826, + "GCGCGC": 3827, + "GCGCGG": 3828, + "GCGGAA": 3829, + "GCGGAT": 3830, + "GCGGAC": 3831, + "GCGGAG": 3832, + "GCGGTA": 3833, + "GCGGTT": 3834, + "GCGGTC": 3835, + "GCGGTG": 3836, + "GCGGCA": 3837, + "GCGGCT": 3838, + "GCGGCC": 3839, + "GCGGCG": 3840, + "GCGGGA": 3841, + "GCGGGT": 3842, + "GCGGGC": 3843, + "GCGGGG": 3844, + "GGAAAA": 3845, + "GGAAAT": 3846, + "GGAAAC": 3847, + "GGAAAG": 3848, + "GGAATA": 3849, + "GGAATT": 3850, + "GGAATC": 3851, + "GGAATG": 3852, + "GGAACA": 3853, + "GGAACT": 3854, + "GGAACC": 3855, + "GGAACG": 3856, + "GGAAGA": 3857, + "GGAAGT": 3858, + "GGAAGC": 3859, + "GGAAGG": 3860, + "GGATAA": 3861, + "GGATAT": 3862, + "GGATAC": 3863, + "GGATAG": 3864, + "GGATTA": 3865, + "GGATTT": 3866, + "GGATTC": 3867, + "GGATTG": 3868, + "GGATCA": 3869, + "GGATCT": 3870, + "GGATCC": 3871, + "GGATCG": 3872, + "GGATGA": 3873, + "GGATGT": 3874, + "GGATGC": 3875, + "GGATGG": 3876, + "GGACAA": 3877, + "GGACAT": 3878, + "GGACAC": 3879, + "GGACAG": 3880, + "GGACTA": 3881, + "GGACTT": 3882, + "GGACTC": 3883, + "GGACTG": 3884, + "GGACCA": 3885, + "GGACCT": 3886, + "GGACCC": 3887, + "GGACCG": 3888, + "GGACGA": 3889, + "GGACGT": 3890, + "GGACGC": 3891, + "GGACGG": 3892, + "GGAGAA": 3893, + "GGAGAT": 3894, + "GGAGAC": 3895, + "GGAGAG": 3896, + "GGAGTA": 3897, + "GGAGTT": 3898, + "GGAGTC": 3899, + "GGAGTG": 3900, + "GGAGCA": 3901, + "GGAGCT": 3902, + "GGAGCC": 3903, + "GGAGCG": 3904, + "GGAGGA": 3905, + "GGAGGT": 3906, + "GGAGGC": 3907, + "GGAGGG": 3908, + "GGTAAA": 3909, + "GGTAAT": 3910, + "GGTAAC": 3911, + "GGTAAG": 3912, + "GGTATA": 3913, + "GGTATT": 3914, + "GGTATC": 3915, + "GGTATG": 3916, + "GGTACA": 3917, + "GGTACT": 3918, + "GGTACC": 3919, + "GGTACG": 3920, + "GGTAGA": 3921, + "GGTAGT": 3922, + "GGTAGC": 3923, + "GGTAGG": 3924, + "GGTTAA": 3925, + "GGTTAT": 3926, + "GGTTAC": 3927, + "GGTTAG": 3928, + "GGTTTA": 3929, + "GGTTTT": 3930, + "GGTTTC": 3931, + "GGTTTG": 3932, + "GGTTCA": 3933, + "GGTTCT": 3934, + "GGTTCC": 3935, + "GGTTCG": 3936, + "GGTTGA": 3937, + "GGTTGT": 3938, + "GGTTGC": 3939, + "GGTTGG": 3940, + "GGTCAA": 3941, + "GGTCAT": 3942, + "GGTCAC": 3943, + "GGTCAG": 3944, + "GGTCTA": 3945, + "GGTCTT": 3946, + "GGTCTC": 3947, + "GGTCTG": 3948, + "GGTCCA": 3949, + "GGTCCT": 3950, + "GGTCCC": 3951, + "GGTCCG": 3952, + "GGTCGA": 3953, + "GGTCGT": 3954, + "GGTCGC": 3955, + "GGTCGG": 3956, + "GGTGAA": 3957, + "GGTGAT": 3958, + "GGTGAC": 3959, + "GGTGAG": 3960, + "GGTGTA": 3961, + "GGTGTT": 3962, + "GGTGTC": 3963, + "GGTGTG": 3964, + "GGTGCA": 3965, + "GGTGCT": 3966, + "GGTGCC": 3967, + "GGTGCG": 3968, + "GGTGGA": 3969, + "GGTGGT": 3970, + "GGTGGC": 3971, + "GGTGGG": 3972, + "GGCAAA": 3973, + "GGCAAT": 3974, + "GGCAAC": 3975, + "GGCAAG": 3976, + "GGCATA": 3977, + "GGCATT": 3978, + "GGCATC": 3979, + "GGCATG": 3980, + "GGCACA": 3981, + "GGCACT": 3982, + "GGCACC": 3983, + "GGCACG": 3984, + "GGCAGA": 3985, + "GGCAGT": 3986, + "GGCAGC": 3987, + "GGCAGG": 3988, + "GGCTAA": 3989, + "GGCTAT": 3990, + "GGCTAC": 3991, + "GGCTAG": 3992, + "GGCTTA": 3993, + "GGCTTT": 3994, + "GGCTTC": 3995, + "GGCTTG": 3996, + "GGCTCA": 3997, + "GGCTCT": 3998, + "GGCTCC": 3999, + "GGCTCG": 4000, + "GGCTGA": 4001, + "GGCTGT": 4002, + "GGCTGC": 4003, + "GGCTGG": 4004, + "GGCCAA": 4005, + "GGCCAT": 4006, + "GGCCAC": 4007, + "GGCCAG": 4008, + "GGCCTA": 4009, + "GGCCTT": 4010, + "GGCCTC": 4011, + "GGCCTG": 4012, + "GGCCCA": 4013, + "GGCCCT": 4014, + "GGCCCC": 4015, + "GGCCCG": 4016, + "GGCCGA": 4017, + "GGCCGT": 4018, + "GGCCGC": 4019, + "GGCCGG": 4020, + "GGCGAA": 4021, + "GGCGAT": 4022, + "GGCGAC": 4023, + "GGCGAG": 4024, + "GGCGTA": 4025, + "GGCGTT": 4026, + "GGCGTC": 4027, + "GGCGTG": 4028, + "GGCGCA": 4029, + "GGCGCT": 4030, + "GGCGCC": 4031, + "GGCGCG": 4032, + "GGCGGA": 4033, + "GGCGGT": 4034, + "GGCGGC": 4035, + "GGCGGG": 4036, + "GGGAAA": 4037, + "GGGAAT": 4038, + "GGGAAC": 4039, + "GGGAAG": 4040, + "GGGATA": 4041, + "GGGATT": 4042, + "GGGATC": 4043, + "GGGATG": 4044, + "GGGACA": 4045, + "GGGACT": 4046, + "GGGACC": 4047, + "GGGACG": 4048, + "GGGAGA": 4049, + "GGGAGT": 4050, + "GGGAGC": 4051, + "GGGAGG": 4052, + "GGGTAA": 4053, + "GGGTAT": 4054, + "GGGTAC": 4055, + "GGGTAG": 4056, + "GGGTTA": 4057, + "GGGTTT": 4058, + "GGGTTC": 4059, + "GGGTTG": 4060, + "GGGTCA": 4061, + "GGGTCT": 4062, + "GGGTCC": 4063, + "GGGTCG": 4064, + "GGGTGA": 4065, + "GGGTGT": 4066, + "GGGTGC": 4067, + "GGGTGG": 4068, + "GGGCAA": 4069, + "GGGCAT": 4070, + "GGGCAC": 4071, + "GGGCAG": 4072, + "GGGCTA": 4073, + "GGGCTT": 4074, + "GGGCTC": 4075, + "GGGCTG": 4076, + "GGGCCA": 4077, + "GGGCCT": 4078, + "GGGCCC": 4079, + "GGGCCG": 4080, + "GGGCGA": 4081, + "GGGCGT": 4082, + "GGGCGC": 4083, + "GGGCGG": 4084, + "GGGGAA": 4085, + "GGGGAT": 4086, + "GGGGAC": 4087, + "GGGGAG": 4088, + "GGGGTA": 4089, + "GGGGTT": 4090, + "GGGGTC": 4091, + "GGGGTG": 4092, + "GGGGCA": 4093, + "GGGGCT": 4094, + "GGGGCC": 4095, + "GGGGCG": 4096, + "GGGGGA": 4097, + "GGGGGT": 4098, + "GGGGGC": 4099, + "GGGGGG": 4100 + } + } +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/tokenizer_config.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..604e2b1b29af1dcb3920d2828b6a90276e5e3d02 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/tokenizer_config.json @@ -0,0 +1,58 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "[UNK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "[CLS]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "[SEP]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "4": { + "content": "[MASK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "clean_up_tokenization_spaces": true, + "cls_token": "[CLS]", + "do_basic_tokenize": true, + "do_lower_case": false, + "mask_token": "[MASK]", + "max_len": 512, + "model_max_length": 512, + "never_split": null, + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "strip_accents": null, + "tokenize_chinese_chars": true, + "tokenizer_class": "BertTokenizer", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/trainer_state.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..1274d3eded92b1ae49cefa1819c87879d3b8cda6 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/trainer_state.json @@ -0,0 +1,593 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 8.0, + "eval_steps": 500, + "global_step": 2000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.1, + "grad_norm": 0.2192622423171997, + "learning_rate": 5.319148936170213e-05, + "loss": 0.6949, + "step": 25 + }, + { + "epoch": 0.2, + "grad_norm": 0.22659960389137268, + "learning_rate": 0.00010638297872340425, + "loss": 0.6926, + "step": 50 + }, + { + "epoch": 0.3, + "grad_norm": 0.33277687430381775, + "learning_rate": 0.00015957446808510637, + "loss": 0.693, + "step": 75 + }, + { + "epoch": 0.4, + "grad_norm": 0.5183415412902832, + "learning_rate": 0.0002127659574468085, + "loss": 0.6928, + "step": 100 + }, + { + "epoch": 0.5, + "grad_norm": 0.21227706968784332, + "learning_rate": 0.00026595744680851064, + "loss": 0.6938, + "step": 125 + }, + { + "epoch": 0.6, + "grad_norm": 0.1846485733985901, + "learning_rate": 0.00031914893617021275, + "loss": 0.6942, + "step": 150 + }, + { + "epoch": 0.7, + "grad_norm": 0.27858808636665344, + "learning_rate": 0.0003723404255319149, + "loss": 0.6931, + "step": 175 + }, + { + "epoch": 0.8, + "grad_norm": 0.26012367010116577, + "learning_rate": 0.0003992081821181128, + "loss": 0.6936, + "step": 200 + }, + { + "epoch": 0.9, + "grad_norm": 0.2534612715244293, + "learning_rate": 0.0003975585615308479, + "loss": 0.6939, + "step": 225 + }, + { + "epoch": 1.0, + "grad_norm": 0.42967700958251953, + "learning_rate": 0.00039590894094358297, + "loss": 0.6928, + "step": 250 + }, + { + "epoch": 1.1, + "grad_norm": 0.2613862454891205, + "learning_rate": 0.0003942593203563181, + "loss": 0.6918, + "step": 275 + }, + { + "epoch": 1.2, + "grad_norm": 0.5514339208602905, + "learning_rate": 0.0003926096997690532, + "loss": 0.6918, + "step": 300 + }, + { + "epoch": 1.3, + "grad_norm": 0.33801883459091187, + "learning_rate": 0.00039096007918178817, + "loss": 0.6906, + "step": 325 + }, + { + "epoch": 1.4, + "grad_norm": 0.2953580617904663, + "learning_rate": 0.0003893104585945233, + "loss": 0.6886, + "step": 350 + }, + { + "epoch": 1.5, + "grad_norm": 0.5340980887413025, + "learning_rate": 0.0003876608380072583, + "loss": 0.6877, + "step": 375 + }, + { + "epoch": 1.6, + "grad_norm": 0.3042701482772827, + "learning_rate": 0.00038601121741999343, + "loss": 0.6853, + "step": 400 + }, + { + "epoch": 1.7, + "grad_norm": 0.44497063755989075, + "learning_rate": 0.0003843615968327285, + "loss": 0.6772, + "step": 425 + }, + { + "epoch": 1.8, + "grad_norm": 0.6260074973106384, + "learning_rate": 0.0003827119762454636, + "loss": 0.6671, + "step": 450 + }, + { + "epoch": 1.9, + "grad_norm": 0.9574269652366638, + "learning_rate": 0.00038106235565819863, + "loss": 0.6517, + "step": 475 + }, + { + "epoch": 2.0, + "grad_norm": 0.9055777192115784, + "learning_rate": 0.0003794127350709337, + "loss": 0.6438, + "step": 500 + }, + { + "epoch": 2.1, + "grad_norm": 0.7860134840011597, + "learning_rate": 0.0003777631144836688, + "loss": 0.6379, + "step": 525 + }, + { + "epoch": 2.2, + "grad_norm": 1.5324610471725464, + "learning_rate": 0.00037611349389640383, + "loss": 0.6143, + "step": 550 + }, + { + "epoch": 2.3, + "grad_norm": 1.8535794019699097, + "learning_rate": 0.00037446387330913894, + "loss": 0.607, + "step": 575 + }, + { + "epoch": 2.4, + "grad_norm": 1.8850122690200806, + "learning_rate": 0.000372814252721874, + "loss": 0.5642, + "step": 600 + }, + { + "epoch": 2.5, + "grad_norm": 1.288278579711914, + "learning_rate": 0.00037116463213460903, + "loss": 0.5739, + "step": 625 + }, + { + "epoch": 2.6, + "grad_norm": 1.344173550605774, + "learning_rate": 0.00036951501154734414, + "loss": 0.5718, + "step": 650 + }, + { + "epoch": 2.7, + "grad_norm": 1.430834174156189, + "learning_rate": 0.0003678653909600792, + "loss": 0.543, + "step": 675 + }, + { + "epoch": 2.8, + "grad_norm": 1.5948325395584106, + "learning_rate": 0.0003662157703728143, + "loss": 0.5434, + "step": 700 + }, + { + "epoch": 2.9, + "grad_norm": 2.688519239425659, + "learning_rate": 0.00036456614978554934, + "loss": 0.5842, + "step": 725 + }, + { + "epoch": 3.0, + "grad_norm": 1.6625819206237793, + "learning_rate": 0.00036291652919828444, + "loss": 0.5526, + "step": 750 + }, + { + "epoch": 3.1, + "grad_norm": 1.869725227355957, + "learning_rate": 0.00036126690861101944, + "loss": 0.53, + "step": 775 + }, + { + "epoch": 3.2, + "grad_norm": 2.1625709533691406, + "learning_rate": 0.00035961728802375454, + "loss": 0.5341, + "step": 800 + }, + { + "epoch": 3.3, + "grad_norm": 1.9650202989578247, + "learning_rate": 0.0003579676674364896, + "loss": 0.5297, + "step": 825 + }, + { + "epoch": 3.4, + "grad_norm": 1.3358676433563232, + "learning_rate": 0.0003563180468492247, + "loss": 0.5273, + "step": 850 + }, + { + "epoch": 3.5, + "grad_norm": 2.1927096843719482, + "learning_rate": 0.0003546684262619598, + "loss": 0.5084, + "step": 875 + }, + { + "epoch": 3.6, + "grad_norm": 2.2377188205718994, + "learning_rate": 0.00035301880567469485, + "loss": 0.4972, + "step": 900 + }, + { + "epoch": 3.7, + "grad_norm": 1.8601839542388916, + "learning_rate": 0.0003513691850874299, + "loss": 0.4852, + "step": 925 + }, + { + "epoch": 3.8, + "grad_norm": 1.4995975494384766, + "learning_rate": 0.00034971956450016495, + "loss": 0.5309, + "step": 950 + }, + { + "epoch": 3.9, + "grad_norm": 2.0536930561065674, + "learning_rate": 0.00034806994391290005, + "loss": 0.5249, + "step": 975 + }, + { + "epoch": 4.0, + "grad_norm": 2.084728479385376, + "learning_rate": 0.0003464203233256351, + "loss": 0.4771, + "step": 1000 + }, + { + "epoch": 4.1, + "grad_norm": 1.705510139465332, + "learning_rate": 0.0003447707027383702, + "loss": 0.5181, + "step": 1025 + }, + { + "epoch": 4.2, + "grad_norm": 3.83380389213562, + "learning_rate": 0.00034312108215110525, + "loss": 0.4873, + "step": 1050 + }, + { + "epoch": 4.3, + "grad_norm": 2.0163533687591553, + "learning_rate": 0.0003414714615638403, + "loss": 0.4801, + "step": 1075 + }, + { + "epoch": 4.4, + "grad_norm": 1.9917407035827637, + "learning_rate": 0.0003398218409765754, + "loss": 0.4766, + "step": 1100 + }, + { + "epoch": 4.5, + "grad_norm": 2.111729860305786, + "learning_rate": 0.00033817222038931045, + "loss": 0.4877, + "step": 1125 + }, + { + "epoch": 4.6, + "grad_norm": 1.5487738847732544, + "learning_rate": 0.00033652259980204556, + "loss": 0.507, + "step": 1150 + }, + { + "epoch": 4.7, + "grad_norm": 1.8715145587921143, + "learning_rate": 0.0003348729792147806, + "loss": 0.452, + "step": 1175 + }, + { + "epoch": 4.8, + "grad_norm": 2.2618906497955322, + "learning_rate": 0.0003332233586275157, + "loss": 0.4799, + "step": 1200 + }, + { + "epoch": 4.9, + "grad_norm": 1.657012939453125, + "learning_rate": 0.00033157373804025076, + "loss": 0.4715, + "step": 1225 + }, + { + "epoch": 5.0, + "grad_norm": 2.8894591331481934, + "learning_rate": 0.0003299241174529858, + "loss": 0.4437, + "step": 1250 + }, + { + "epoch": 5.1, + "grad_norm": 1.9179915189743042, + "learning_rate": 0.0003282744968657209, + "loss": 0.4796, + "step": 1275 + }, + { + "epoch": 5.2, + "grad_norm": 2.1816720962524414, + "learning_rate": 0.00032662487627845596, + "loss": 0.4463, + "step": 1300 + }, + { + "epoch": 5.3, + "grad_norm": 2.2761082649230957, + "learning_rate": 0.00032497525569119106, + "loss": 0.4215, + "step": 1325 + }, + { + "epoch": 5.4, + "grad_norm": 1.7190684080123901, + "learning_rate": 0.0003233256351039261, + "loss": 0.4481, + "step": 1350 + }, + { + "epoch": 5.5, + "grad_norm": 1.4793736934661865, + "learning_rate": 0.00032167601451666116, + "loss": 0.44, + "step": 1375 + }, + { + "epoch": 5.6, + "grad_norm": 2.5064358711242676, + "learning_rate": 0.00032002639392939627, + "loss": 0.4583, + "step": 1400 + }, + { + "epoch": 5.7, + "grad_norm": 1.8364307880401611, + "learning_rate": 0.0003183767733421313, + "loss": 0.4516, + "step": 1425 + }, + { + "epoch": 5.8, + "grad_norm": 1.3747583627700806, + "learning_rate": 0.0003167271527548664, + "loss": 0.4535, + "step": 1450 + }, + { + "epoch": 5.9, + "grad_norm": 2.580183982849121, + "learning_rate": 0.00031507753216760147, + "loss": 0.4325, + "step": 1475 + }, + { + "epoch": 6.0, + "grad_norm": 1.562374472618103, + "learning_rate": 0.00031342791158033657, + "loss": 0.4616, + "step": 1500 + }, + { + "epoch": 6.1, + "grad_norm": 1.4970141649246216, + "learning_rate": 0.00031177829099307157, + "loss": 0.4607, + "step": 1525 + }, + { + "epoch": 6.2, + "grad_norm": 2.436326503753662, + "learning_rate": 0.00031012867040580667, + "loss": 0.4498, + "step": 1550 + }, + { + "epoch": 6.3, + "grad_norm": 1.9849016666412354, + "learning_rate": 0.0003084790498185417, + "loss": 0.4262, + "step": 1575 + }, + { + "epoch": 6.4, + "grad_norm": 1.6405338048934937, + "learning_rate": 0.0003068294292312768, + "loss": 0.4225, + "step": 1600 + }, + { + "epoch": 6.5, + "grad_norm": 1.8233951330184937, + "learning_rate": 0.0003051798086440119, + "loss": 0.4248, + "step": 1625 + }, + { + "epoch": 6.6, + "grad_norm": 1.396469235420227, + "learning_rate": 0.000303530188056747, + "loss": 0.412, + "step": 1650 + }, + { + "epoch": 6.7, + "grad_norm": 1.5438531637191772, + "learning_rate": 0.000301880567469482, + "loss": 0.4105, + "step": 1675 + }, + { + "epoch": 6.8, + "grad_norm": 2.49668288230896, + "learning_rate": 0.0003002309468822171, + "loss": 0.4318, + "step": 1700 + }, + { + "epoch": 6.9, + "grad_norm": 1.5911107063293457, + "learning_rate": 0.0002985813262949522, + "loss": 0.4454, + "step": 1725 + }, + { + "epoch": 7.0, + "grad_norm": 2.9934186935424805, + "learning_rate": 0.00029693170570768723, + "loss": 0.4306, + "step": 1750 + }, + { + "epoch": 7.1, + "grad_norm": 2.5476737022399902, + "learning_rate": 0.00029528208512042233, + "loss": 0.4066, + "step": 1775 + }, + { + "epoch": 7.2, + "grad_norm": 4.207254886627197, + "learning_rate": 0.0002936324645331574, + "loss": 0.4521, + "step": 1800 + }, + { + "epoch": 7.3, + "grad_norm": 1.8378102779388428, + "learning_rate": 0.00029198284394589243, + "loss": 0.4197, + "step": 1825 + }, + { + "epoch": 7.4, + "grad_norm": 2.2129015922546387, + "learning_rate": 0.00029033322335862753, + "loss": 0.4182, + "step": 1850 + }, + { + "epoch": 7.5, + "grad_norm": 1.4666684865951538, + "learning_rate": 0.0002886836027713626, + "loss": 0.4224, + "step": 1875 + }, + { + "epoch": 7.6, + "grad_norm": 2.0311567783355713, + "learning_rate": 0.0002870339821840977, + "loss": 0.4304, + "step": 1900 + }, + { + "epoch": 7.7, + "grad_norm": 2.346604585647583, + "learning_rate": 0.00028538436159683273, + "loss": 0.4012, + "step": 1925 + }, + { + "epoch": 7.8, + "grad_norm": 1.440136432647705, + "learning_rate": 0.00028373474100956784, + "loss": 0.4487, + "step": 1950 + }, + { + "epoch": 7.9, + "grad_norm": 2.426828384399414, + "learning_rate": 0.0002820851204223029, + "loss": 0.3577, + "step": 1975 + }, + { + "epoch": 8.0, + "grad_norm": 3.095980405807495, + "learning_rate": 0.00028043549983503794, + "loss": 0.4121, + "step": 2000 + } + ], + "logging_steps": 25, + "max_steps": 6250, + "num_input_tokens_seen": 0, + "num_train_epochs": 25, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 110762065920000.0, + "train_batch_size": 20, + "trial_name": null, + "trial_params": null +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/training_args.bin b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..10d16bb8e9df9c51aba38e7a6f04deeecf1ea29f --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:662a5f25d38ccae34f19d85e24456de6d3eaaabafca26fb7cc8816de0c990f20 +size 5304 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/vocab.txt b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/vocab.txt new file mode 100644 index 0000000000000000000000000000000000000000..cc9ad53732e21ea9180f18e1fb67a38cc5702d4c --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/vocab.txt @@ -0,0 +1,4101 @@ +[PAD] +[UNK] +[CLS] +[SEP] +[MASK] +AAAAAA +AAAAAT +AAAAAC +AAAAAG +AAAATA +AAAATT +AAAATC +AAAATG +AAAACA +AAAACT +AAAACC +AAAACG +AAAAGA +AAAAGT +AAAAGC +AAAAGG +AAATAA +AAATAT +AAATAC +AAATAG +AAATTA +AAATTT +AAATTC +AAATTG +AAATCA +AAATCT +AAATCC +AAATCG +AAATGA +AAATGT +AAATGC +AAATGG +AAACAA +AAACAT +AAACAC +AAACAG +AAACTA +AAACTT +AAACTC +AAACTG +AAACCA +AAACCT +AAACCC +AAACCG +AAACGA +AAACGT +AAACGC +AAACGG +AAAGAA +AAAGAT +AAAGAC +AAAGAG +AAAGTA +AAAGTT +AAAGTC +AAAGTG +AAAGCA +AAAGCT +AAAGCC +AAAGCG +AAAGGA +AAAGGT +AAAGGC +AAAGGG +AATAAA +AATAAT +AATAAC +AATAAG +AATATA +AATATT +AATATC +AATATG +AATACA +AATACT +AATACC +AATACG +AATAGA +AATAGT +AATAGC +AATAGG +AATTAA +AATTAT +AATTAC +AATTAG +AATTTA +AATTTT +AATTTC +AATTTG +AATTCA +AATTCT +AATTCC +AATTCG +AATTGA +AATTGT +AATTGC +AATTGG +AATCAA +AATCAT +AATCAC +AATCAG +AATCTA +AATCTT +AATCTC +AATCTG +AATCCA +AATCCT +AATCCC +AATCCG +AATCGA +AATCGT +AATCGC +AATCGG +AATGAA +AATGAT +AATGAC +AATGAG +AATGTA +AATGTT +AATGTC +AATGTG +AATGCA +AATGCT +AATGCC +AATGCG +AATGGA +AATGGT +AATGGC +AATGGG +AACAAA +AACAAT +AACAAC +AACAAG +AACATA +AACATT +AACATC +AACATG +AACACA +AACACT +AACACC +AACACG +AACAGA +AACAGT +AACAGC +AACAGG +AACTAA +AACTAT +AACTAC +AACTAG +AACTTA +AACTTT +AACTTC +AACTTG +AACTCA +AACTCT +AACTCC +AACTCG +AACTGA +AACTGT +AACTGC +AACTGG +AACCAA +AACCAT +AACCAC +AACCAG +AACCTA +AACCTT +AACCTC +AACCTG +AACCCA +AACCCT +AACCCC +AACCCG +AACCGA +AACCGT +AACCGC +AACCGG +AACGAA +AACGAT +AACGAC +AACGAG +AACGTA +AACGTT +AACGTC +AACGTG +AACGCA +AACGCT +AACGCC +AACGCG +AACGGA +AACGGT +AACGGC +AACGGG +AAGAAA +AAGAAT +AAGAAC +AAGAAG +AAGATA +AAGATT +AAGATC +AAGATG +AAGACA +AAGACT +AAGACC +AAGACG +AAGAGA +AAGAGT +AAGAGC +AAGAGG +AAGTAA +AAGTAT +AAGTAC +AAGTAG +AAGTTA +AAGTTT +AAGTTC +AAGTTG +AAGTCA +AAGTCT +AAGTCC +AAGTCG +AAGTGA +AAGTGT +AAGTGC +AAGTGG +AAGCAA +AAGCAT +AAGCAC +AAGCAG +AAGCTA +AAGCTT +AAGCTC +AAGCTG +AAGCCA +AAGCCT +AAGCCC +AAGCCG +AAGCGA +AAGCGT +AAGCGC +AAGCGG +AAGGAA +AAGGAT +AAGGAC +AAGGAG +AAGGTA +AAGGTT +AAGGTC +AAGGTG +AAGGCA +AAGGCT +AAGGCC +AAGGCG +AAGGGA +AAGGGT +AAGGGC +AAGGGG +ATAAAA +ATAAAT +ATAAAC +ATAAAG +ATAATA +ATAATT +ATAATC +ATAATG +ATAACA +ATAACT +ATAACC +ATAACG +ATAAGA +ATAAGT +ATAAGC +ATAAGG +ATATAA +ATATAT +ATATAC +ATATAG +ATATTA +ATATTT +ATATTC +ATATTG +ATATCA +ATATCT +ATATCC +ATATCG +ATATGA +ATATGT +ATATGC +ATATGG +ATACAA +ATACAT +ATACAC +ATACAG +ATACTA +ATACTT +ATACTC +ATACTG +ATACCA +ATACCT +ATACCC +ATACCG +ATACGA +ATACGT +ATACGC +ATACGG +ATAGAA +ATAGAT +ATAGAC +ATAGAG +ATAGTA +ATAGTT +ATAGTC +ATAGTG +ATAGCA +ATAGCT +ATAGCC +ATAGCG +ATAGGA +ATAGGT +ATAGGC +ATAGGG +ATTAAA +ATTAAT +ATTAAC +ATTAAG +ATTATA +ATTATT +ATTATC +ATTATG +ATTACA +ATTACT +ATTACC +ATTACG +ATTAGA +ATTAGT +ATTAGC +ATTAGG +ATTTAA +ATTTAT +ATTTAC +ATTTAG +ATTTTA +ATTTTT +ATTTTC +ATTTTG +ATTTCA +ATTTCT +ATTTCC +ATTTCG +ATTTGA +ATTTGT +ATTTGC +ATTTGG +ATTCAA +ATTCAT +ATTCAC +ATTCAG +ATTCTA +ATTCTT +ATTCTC +ATTCTG +ATTCCA +ATTCCT +ATTCCC +ATTCCG +ATTCGA +ATTCGT +ATTCGC +ATTCGG +ATTGAA +ATTGAT +ATTGAC +ATTGAG +ATTGTA +ATTGTT +ATTGTC +ATTGTG +ATTGCA +ATTGCT +ATTGCC +ATTGCG +ATTGGA +ATTGGT +ATTGGC +ATTGGG +ATCAAA +ATCAAT +ATCAAC +ATCAAG +ATCATA +ATCATT +ATCATC +ATCATG +ATCACA +ATCACT +ATCACC +ATCACG +ATCAGA +ATCAGT +ATCAGC +ATCAGG +ATCTAA +ATCTAT +ATCTAC +ATCTAG +ATCTTA +ATCTTT +ATCTTC +ATCTTG +ATCTCA +ATCTCT +ATCTCC +ATCTCG +ATCTGA +ATCTGT +ATCTGC +ATCTGG +ATCCAA +ATCCAT +ATCCAC +ATCCAG +ATCCTA +ATCCTT +ATCCTC +ATCCTG +ATCCCA +ATCCCT +ATCCCC +ATCCCG +ATCCGA +ATCCGT +ATCCGC +ATCCGG +ATCGAA +ATCGAT +ATCGAC +ATCGAG +ATCGTA +ATCGTT +ATCGTC +ATCGTG +ATCGCA +ATCGCT +ATCGCC +ATCGCG +ATCGGA +ATCGGT +ATCGGC +ATCGGG +ATGAAA +ATGAAT +ATGAAC +ATGAAG +ATGATA +ATGATT +ATGATC +ATGATG +ATGACA +ATGACT +ATGACC +ATGACG +ATGAGA +ATGAGT +ATGAGC +ATGAGG +ATGTAA +ATGTAT +ATGTAC +ATGTAG +ATGTTA +ATGTTT +ATGTTC +ATGTTG +ATGTCA +ATGTCT +ATGTCC +ATGTCG +ATGTGA +ATGTGT +ATGTGC +ATGTGG +ATGCAA +ATGCAT +ATGCAC +ATGCAG +ATGCTA +ATGCTT +ATGCTC +ATGCTG +ATGCCA +ATGCCT +ATGCCC +ATGCCG +ATGCGA +ATGCGT +ATGCGC +ATGCGG +ATGGAA +ATGGAT +ATGGAC +ATGGAG +ATGGTA +ATGGTT +ATGGTC +ATGGTG +ATGGCA +ATGGCT +ATGGCC +ATGGCG +ATGGGA +ATGGGT +ATGGGC +ATGGGG +ACAAAA +ACAAAT +ACAAAC +ACAAAG +ACAATA +ACAATT +ACAATC +ACAATG +ACAACA +ACAACT +ACAACC +ACAACG +ACAAGA +ACAAGT +ACAAGC +ACAAGG +ACATAA +ACATAT +ACATAC +ACATAG +ACATTA +ACATTT +ACATTC +ACATTG +ACATCA +ACATCT +ACATCC +ACATCG +ACATGA +ACATGT +ACATGC +ACATGG +ACACAA +ACACAT +ACACAC +ACACAG +ACACTA +ACACTT +ACACTC +ACACTG +ACACCA +ACACCT +ACACCC +ACACCG +ACACGA +ACACGT +ACACGC +ACACGG +ACAGAA +ACAGAT +ACAGAC +ACAGAG +ACAGTA +ACAGTT +ACAGTC +ACAGTG +ACAGCA +ACAGCT +ACAGCC +ACAGCG +ACAGGA +ACAGGT +ACAGGC +ACAGGG +ACTAAA +ACTAAT +ACTAAC +ACTAAG +ACTATA +ACTATT +ACTATC +ACTATG +ACTACA +ACTACT +ACTACC +ACTACG +ACTAGA +ACTAGT +ACTAGC +ACTAGG +ACTTAA +ACTTAT +ACTTAC +ACTTAG +ACTTTA +ACTTTT +ACTTTC +ACTTTG +ACTTCA +ACTTCT +ACTTCC +ACTTCG +ACTTGA +ACTTGT +ACTTGC +ACTTGG +ACTCAA +ACTCAT +ACTCAC +ACTCAG +ACTCTA +ACTCTT +ACTCTC +ACTCTG +ACTCCA +ACTCCT +ACTCCC +ACTCCG +ACTCGA +ACTCGT +ACTCGC +ACTCGG +ACTGAA +ACTGAT +ACTGAC +ACTGAG +ACTGTA +ACTGTT +ACTGTC +ACTGTG +ACTGCA +ACTGCT +ACTGCC +ACTGCG +ACTGGA +ACTGGT +ACTGGC +ACTGGG +ACCAAA +ACCAAT +ACCAAC +ACCAAG +ACCATA +ACCATT +ACCATC +ACCATG +ACCACA +ACCACT +ACCACC +ACCACG +ACCAGA +ACCAGT +ACCAGC +ACCAGG +ACCTAA +ACCTAT +ACCTAC +ACCTAG +ACCTTA +ACCTTT +ACCTTC +ACCTTG +ACCTCA +ACCTCT +ACCTCC +ACCTCG +ACCTGA +ACCTGT +ACCTGC +ACCTGG +ACCCAA +ACCCAT +ACCCAC +ACCCAG +ACCCTA +ACCCTT +ACCCTC +ACCCTG +ACCCCA +ACCCCT +ACCCCC +ACCCCG +ACCCGA +ACCCGT +ACCCGC +ACCCGG +ACCGAA +ACCGAT +ACCGAC +ACCGAG +ACCGTA +ACCGTT +ACCGTC +ACCGTG +ACCGCA +ACCGCT +ACCGCC +ACCGCG +ACCGGA +ACCGGT +ACCGGC +ACCGGG +ACGAAA +ACGAAT +ACGAAC +ACGAAG +ACGATA +ACGATT +ACGATC +ACGATG +ACGACA +ACGACT +ACGACC +ACGACG +ACGAGA +ACGAGT +ACGAGC +ACGAGG +ACGTAA +ACGTAT +ACGTAC +ACGTAG +ACGTTA +ACGTTT +ACGTTC +ACGTTG +ACGTCA +ACGTCT +ACGTCC +ACGTCG +ACGTGA +ACGTGT +ACGTGC +ACGTGG +ACGCAA +ACGCAT +ACGCAC +ACGCAG +ACGCTA +ACGCTT +ACGCTC +ACGCTG +ACGCCA +ACGCCT +ACGCCC +ACGCCG +ACGCGA +ACGCGT +ACGCGC +ACGCGG +ACGGAA +ACGGAT +ACGGAC +ACGGAG +ACGGTA +ACGGTT +ACGGTC +ACGGTG +ACGGCA +ACGGCT +ACGGCC +ACGGCG +ACGGGA +ACGGGT +ACGGGC +ACGGGG +AGAAAA +AGAAAT +AGAAAC +AGAAAG +AGAATA +AGAATT +AGAATC +AGAATG +AGAACA +AGAACT +AGAACC +AGAACG +AGAAGA +AGAAGT +AGAAGC +AGAAGG +AGATAA +AGATAT +AGATAC +AGATAG +AGATTA +AGATTT +AGATTC +AGATTG +AGATCA +AGATCT +AGATCC +AGATCG +AGATGA +AGATGT +AGATGC +AGATGG +AGACAA +AGACAT +AGACAC +AGACAG +AGACTA +AGACTT +AGACTC +AGACTG +AGACCA +AGACCT +AGACCC +AGACCG +AGACGA +AGACGT +AGACGC +AGACGG +AGAGAA +AGAGAT +AGAGAC +AGAGAG +AGAGTA +AGAGTT +AGAGTC +AGAGTG +AGAGCA +AGAGCT +AGAGCC +AGAGCG +AGAGGA +AGAGGT +AGAGGC +AGAGGG +AGTAAA +AGTAAT +AGTAAC +AGTAAG +AGTATA +AGTATT +AGTATC +AGTATG +AGTACA +AGTACT +AGTACC +AGTACG +AGTAGA +AGTAGT +AGTAGC +AGTAGG +AGTTAA +AGTTAT +AGTTAC +AGTTAG +AGTTTA +AGTTTT +AGTTTC +AGTTTG +AGTTCA +AGTTCT +AGTTCC +AGTTCG +AGTTGA +AGTTGT +AGTTGC +AGTTGG +AGTCAA +AGTCAT +AGTCAC +AGTCAG +AGTCTA +AGTCTT +AGTCTC +AGTCTG +AGTCCA +AGTCCT +AGTCCC +AGTCCG +AGTCGA +AGTCGT +AGTCGC +AGTCGG +AGTGAA +AGTGAT +AGTGAC +AGTGAG +AGTGTA +AGTGTT +AGTGTC +AGTGTG +AGTGCA +AGTGCT +AGTGCC +AGTGCG +AGTGGA +AGTGGT +AGTGGC +AGTGGG +AGCAAA +AGCAAT +AGCAAC +AGCAAG +AGCATA +AGCATT +AGCATC +AGCATG +AGCACA +AGCACT +AGCACC +AGCACG +AGCAGA +AGCAGT +AGCAGC +AGCAGG +AGCTAA +AGCTAT +AGCTAC +AGCTAG +AGCTTA +AGCTTT +AGCTTC +AGCTTG +AGCTCA +AGCTCT +AGCTCC +AGCTCG +AGCTGA +AGCTGT +AGCTGC +AGCTGG +AGCCAA +AGCCAT +AGCCAC +AGCCAG +AGCCTA +AGCCTT +AGCCTC +AGCCTG +AGCCCA +AGCCCT +AGCCCC +AGCCCG +AGCCGA +AGCCGT +AGCCGC +AGCCGG +AGCGAA +AGCGAT +AGCGAC +AGCGAG +AGCGTA +AGCGTT +AGCGTC +AGCGTG +AGCGCA +AGCGCT +AGCGCC +AGCGCG +AGCGGA +AGCGGT +AGCGGC +AGCGGG +AGGAAA +AGGAAT +AGGAAC +AGGAAG +AGGATA +AGGATT +AGGATC +AGGATG +AGGACA +AGGACT +AGGACC +AGGACG +AGGAGA +AGGAGT +AGGAGC +AGGAGG +AGGTAA +AGGTAT +AGGTAC +AGGTAG +AGGTTA +AGGTTT +AGGTTC +AGGTTG +AGGTCA +AGGTCT +AGGTCC +AGGTCG +AGGTGA +AGGTGT +AGGTGC +AGGTGG +AGGCAA +AGGCAT +AGGCAC +AGGCAG +AGGCTA +AGGCTT +AGGCTC +AGGCTG +AGGCCA +AGGCCT +AGGCCC +AGGCCG +AGGCGA +AGGCGT +AGGCGC +AGGCGG +AGGGAA +AGGGAT +AGGGAC +AGGGAG +AGGGTA +AGGGTT +AGGGTC +AGGGTG +AGGGCA +AGGGCT +AGGGCC +AGGGCG +AGGGGA +AGGGGT +AGGGGC +AGGGGG +TAAAAA +TAAAAT +TAAAAC +TAAAAG +TAAATA +TAAATT +TAAATC +TAAATG +TAAACA +TAAACT +TAAACC +TAAACG +TAAAGA +TAAAGT +TAAAGC +TAAAGG +TAATAA +TAATAT +TAATAC +TAATAG +TAATTA +TAATTT +TAATTC +TAATTG +TAATCA +TAATCT +TAATCC +TAATCG +TAATGA +TAATGT +TAATGC +TAATGG +TAACAA +TAACAT +TAACAC +TAACAG +TAACTA +TAACTT +TAACTC +TAACTG +TAACCA +TAACCT +TAACCC +TAACCG +TAACGA +TAACGT +TAACGC +TAACGG +TAAGAA +TAAGAT +TAAGAC +TAAGAG +TAAGTA +TAAGTT +TAAGTC +TAAGTG +TAAGCA +TAAGCT +TAAGCC +TAAGCG +TAAGGA +TAAGGT +TAAGGC +TAAGGG +TATAAA +TATAAT +TATAAC +TATAAG +TATATA +TATATT +TATATC +TATATG +TATACA +TATACT +TATACC +TATACG +TATAGA +TATAGT +TATAGC +TATAGG +TATTAA +TATTAT +TATTAC +TATTAG +TATTTA +TATTTT +TATTTC +TATTTG +TATTCA +TATTCT +TATTCC +TATTCG +TATTGA +TATTGT +TATTGC +TATTGG +TATCAA +TATCAT +TATCAC +TATCAG +TATCTA +TATCTT +TATCTC +TATCTG +TATCCA +TATCCT +TATCCC +TATCCG +TATCGA +TATCGT +TATCGC +TATCGG +TATGAA +TATGAT +TATGAC +TATGAG +TATGTA +TATGTT +TATGTC +TATGTG +TATGCA +TATGCT +TATGCC +TATGCG +TATGGA +TATGGT +TATGGC +TATGGG +TACAAA +TACAAT +TACAAC +TACAAG +TACATA +TACATT +TACATC +TACATG +TACACA +TACACT +TACACC +TACACG +TACAGA +TACAGT +TACAGC +TACAGG +TACTAA +TACTAT +TACTAC +TACTAG +TACTTA +TACTTT +TACTTC +TACTTG +TACTCA +TACTCT +TACTCC +TACTCG +TACTGA +TACTGT +TACTGC +TACTGG +TACCAA +TACCAT +TACCAC +TACCAG +TACCTA +TACCTT +TACCTC +TACCTG +TACCCA +TACCCT +TACCCC +TACCCG +TACCGA +TACCGT +TACCGC +TACCGG +TACGAA +TACGAT +TACGAC +TACGAG +TACGTA +TACGTT +TACGTC +TACGTG +TACGCA +TACGCT +TACGCC +TACGCG +TACGGA +TACGGT +TACGGC +TACGGG +TAGAAA +TAGAAT +TAGAAC +TAGAAG +TAGATA +TAGATT +TAGATC +TAGATG +TAGACA +TAGACT +TAGACC +TAGACG +TAGAGA +TAGAGT +TAGAGC +TAGAGG +TAGTAA +TAGTAT +TAGTAC +TAGTAG +TAGTTA +TAGTTT +TAGTTC +TAGTTG +TAGTCA +TAGTCT +TAGTCC +TAGTCG +TAGTGA +TAGTGT +TAGTGC +TAGTGG +TAGCAA +TAGCAT +TAGCAC +TAGCAG +TAGCTA +TAGCTT +TAGCTC +TAGCTG +TAGCCA +TAGCCT +TAGCCC +TAGCCG +TAGCGA +TAGCGT +TAGCGC +TAGCGG +TAGGAA +TAGGAT +TAGGAC +TAGGAG +TAGGTA +TAGGTT +TAGGTC +TAGGTG +TAGGCA +TAGGCT +TAGGCC +TAGGCG +TAGGGA +TAGGGT +TAGGGC +TAGGGG +TTAAAA +TTAAAT +TTAAAC +TTAAAG +TTAATA +TTAATT +TTAATC +TTAATG +TTAACA +TTAACT +TTAACC +TTAACG +TTAAGA +TTAAGT +TTAAGC +TTAAGG +TTATAA +TTATAT +TTATAC +TTATAG +TTATTA +TTATTT +TTATTC +TTATTG +TTATCA +TTATCT +TTATCC +TTATCG +TTATGA +TTATGT +TTATGC +TTATGG +TTACAA +TTACAT +TTACAC +TTACAG +TTACTA +TTACTT +TTACTC +TTACTG +TTACCA +TTACCT +TTACCC +TTACCG +TTACGA +TTACGT +TTACGC +TTACGG +TTAGAA +TTAGAT +TTAGAC +TTAGAG +TTAGTA +TTAGTT +TTAGTC +TTAGTG +TTAGCA +TTAGCT +TTAGCC +TTAGCG +TTAGGA +TTAGGT +TTAGGC +TTAGGG +TTTAAA +TTTAAT +TTTAAC +TTTAAG +TTTATA +TTTATT +TTTATC +TTTATG +TTTACA +TTTACT +TTTACC +TTTACG +TTTAGA +TTTAGT +TTTAGC +TTTAGG +TTTTAA +TTTTAT +TTTTAC +TTTTAG +TTTTTA +TTTTTT +TTTTTC +TTTTTG +TTTTCA +TTTTCT +TTTTCC +TTTTCG +TTTTGA +TTTTGT +TTTTGC +TTTTGG +TTTCAA +TTTCAT +TTTCAC +TTTCAG +TTTCTA +TTTCTT +TTTCTC +TTTCTG +TTTCCA +TTTCCT +TTTCCC +TTTCCG +TTTCGA +TTTCGT +TTTCGC +TTTCGG +TTTGAA +TTTGAT +TTTGAC +TTTGAG +TTTGTA +TTTGTT +TTTGTC +TTTGTG +TTTGCA +TTTGCT +TTTGCC +TTTGCG +TTTGGA +TTTGGT +TTTGGC +TTTGGG +TTCAAA +TTCAAT +TTCAAC +TTCAAG +TTCATA +TTCATT +TTCATC +TTCATG +TTCACA +TTCACT +TTCACC +TTCACG +TTCAGA +TTCAGT +TTCAGC +TTCAGG +TTCTAA +TTCTAT +TTCTAC +TTCTAG +TTCTTA +TTCTTT +TTCTTC +TTCTTG +TTCTCA +TTCTCT +TTCTCC +TTCTCG +TTCTGA +TTCTGT +TTCTGC +TTCTGG +TTCCAA +TTCCAT +TTCCAC +TTCCAG +TTCCTA +TTCCTT +TTCCTC +TTCCTG +TTCCCA +TTCCCT +TTCCCC +TTCCCG +TTCCGA +TTCCGT +TTCCGC +TTCCGG +TTCGAA +TTCGAT +TTCGAC +TTCGAG +TTCGTA +TTCGTT +TTCGTC +TTCGTG +TTCGCA +TTCGCT +TTCGCC +TTCGCG +TTCGGA +TTCGGT +TTCGGC +TTCGGG +TTGAAA +TTGAAT +TTGAAC +TTGAAG +TTGATA +TTGATT +TTGATC +TTGATG +TTGACA +TTGACT +TTGACC +TTGACG +TTGAGA +TTGAGT +TTGAGC +TTGAGG +TTGTAA +TTGTAT +TTGTAC +TTGTAG +TTGTTA +TTGTTT +TTGTTC +TTGTTG +TTGTCA +TTGTCT +TTGTCC +TTGTCG +TTGTGA +TTGTGT +TTGTGC +TTGTGG +TTGCAA +TTGCAT +TTGCAC +TTGCAG +TTGCTA +TTGCTT +TTGCTC +TTGCTG +TTGCCA +TTGCCT +TTGCCC +TTGCCG +TTGCGA +TTGCGT +TTGCGC +TTGCGG +TTGGAA +TTGGAT +TTGGAC +TTGGAG +TTGGTA +TTGGTT +TTGGTC +TTGGTG +TTGGCA +TTGGCT +TTGGCC +TTGGCG +TTGGGA +TTGGGT +TTGGGC +TTGGGG +TCAAAA +TCAAAT +TCAAAC +TCAAAG +TCAATA +TCAATT +TCAATC +TCAATG +TCAACA +TCAACT +TCAACC +TCAACG +TCAAGA +TCAAGT +TCAAGC +TCAAGG +TCATAA +TCATAT +TCATAC +TCATAG +TCATTA +TCATTT +TCATTC +TCATTG +TCATCA +TCATCT +TCATCC +TCATCG +TCATGA +TCATGT +TCATGC +TCATGG +TCACAA +TCACAT +TCACAC +TCACAG +TCACTA +TCACTT +TCACTC +TCACTG +TCACCA +TCACCT +TCACCC +TCACCG +TCACGA +TCACGT +TCACGC +TCACGG +TCAGAA +TCAGAT +TCAGAC +TCAGAG +TCAGTA +TCAGTT +TCAGTC +TCAGTG +TCAGCA +TCAGCT +TCAGCC +TCAGCG +TCAGGA +TCAGGT +TCAGGC +TCAGGG +TCTAAA +TCTAAT +TCTAAC +TCTAAG +TCTATA +TCTATT +TCTATC +TCTATG +TCTACA +TCTACT +TCTACC +TCTACG +TCTAGA +TCTAGT +TCTAGC +TCTAGG +TCTTAA +TCTTAT +TCTTAC +TCTTAG +TCTTTA +TCTTTT +TCTTTC +TCTTTG +TCTTCA +TCTTCT +TCTTCC +TCTTCG +TCTTGA +TCTTGT +TCTTGC +TCTTGG +TCTCAA +TCTCAT +TCTCAC +TCTCAG +TCTCTA +TCTCTT +TCTCTC +TCTCTG +TCTCCA +TCTCCT +TCTCCC +TCTCCG +TCTCGA +TCTCGT +TCTCGC +TCTCGG +TCTGAA +TCTGAT +TCTGAC +TCTGAG +TCTGTA +TCTGTT +TCTGTC +TCTGTG +TCTGCA +TCTGCT +TCTGCC +TCTGCG +TCTGGA +TCTGGT +TCTGGC +TCTGGG +TCCAAA +TCCAAT +TCCAAC +TCCAAG +TCCATA +TCCATT +TCCATC +TCCATG +TCCACA +TCCACT +TCCACC +TCCACG +TCCAGA +TCCAGT +TCCAGC +TCCAGG +TCCTAA +TCCTAT +TCCTAC +TCCTAG +TCCTTA +TCCTTT +TCCTTC +TCCTTG +TCCTCA +TCCTCT +TCCTCC +TCCTCG +TCCTGA +TCCTGT +TCCTGC +TCCTGG +TCCCAA +TCCCAT +TCCCAC +TCCCAG +TCCCTA +TCCCTT +TCCCTC +TCCCTG +TCCCCA +TCCCCT +TCCCCC +TCCCCG +TCCCGA +TCCCGT +TCCCGC +TCCCGG +TCCGAA +TCCGAT +TCCGAC +TCCGAG +TCCGTA +TCCGTT +TCCGTC +TCCGTG +TCCGCA +TCCGCT +TCCGCC +TCCGCG +TCCGGA +TCCGGT +TCCGGC +TCCGGG +TCGAAA +TCGAAT +TCGAAC +TCGAAG +TCGATA +TCGATT +TCGATC +TCGATG +TCGACA +TCGACT +TCGACC +TCGACG +TCGAGA +TCGAGT +TCGAGC +TCGAGG +TCGTAA +TCGTAT +TCGTAC +TCGTAG +TCGTTA +TCGTTT +TCGTTC +TCGTTG +TCGTCA +TCGTCT +TCGTCC +TCGTCG +TCGTGA +TCGTGT +TCGTGC +TCGTGG +TCGCAA +TCGCAT +TCGCAC +TCGCAG +TCGCTA +TCGCTT +TCGCTC +TCGCTG +TCGCCA +TCGCCT +TCGCCC +TCGCCG +TCGCGA +TCGCGT +TCGCGC +TCGCGG +TCGGAA +TCGGAT +TCGGAC +TCGGAG +TCGGTA +TCGGTT +TCGGTC +TCGGTG +TCGGCA +TCGGCT +TCGGCC +TCGGCG +TCGGGA +TCGGGT +TCGGGC +TCGGGG +TGAAAA +TGAAAT +TGAAAC +TGAAAG +TGAATA +TGAATT +TGAATC +TGAATG +TGAACA +TGAACT +TGAACC +TGAACG +TGAAGA +TGAAGT +TGAAGC +TGAAGG +TGATAA +TGATAT +TGATAC +TGATAG +TGATTA +TGATTT +TGATTC +TGATTG +TGATCA +TGATCT +TGATCC +TGATCG +TGATGA +TGATGT +TGATGC +TGATGG +TGACAA +TGACAT +TGACAC +TGACAG +TGACTA +TGACTT +TGACTC +TGACTG +TGACCA +TGACCT +TGACCC +TGACCG +TGACGA +TGACGT +TGACGC +TGACGG +TGAGAA +TGAGAT +TGAGAC +TGAGAG +TGAGTA +TGAGTT +TGAGTC +TGAGTG +TGAGCA +TGAGCT +TGAGCC +TGAGCG +TGAGGA +TGAGGT +TGAGGC +TGAGGG +TGTAAA +TGTAAT +TGTAAC +TGTAAG +TGTATA +TGTATT +TGTATC +TGTATG +TGTACA +TGTACT +TGTACC +TGTACG +TGTAGA +TGTAGT +TGTAGC +TGTAGG +TGTTAA +TGTTAT +TGTTAC +TGTTAG +TGTTTA +TGTTTT +TGTTTC +TGTTTG +TGTTCA +TGTTCT +TGTTCC +TGTTCG +TGTTGA +TGTTGT +TGTTGC +TGTTGG +TGTCAA +TGTCAT +TGTCAC +TGTCAG +TGTCTA +TGTCTT +TGTCTC +TGTCTG +TGTCCA +TGTCCT +TGTCCC +TGTCCG +TGTCGA +TGTCGT +TGTCGC +TGTCGG +TGTGAA +TGTGAT +TGTGAC +TGTGAG +TGTGTA +TGTGTT +TGTGTC +TGTGTG +TGTGCA +TGTGCT +TGTGCC +TGTGCG +TGTGGA +TGTGGT +TGTGGC +TGTGGG +TGCAAA +TGCAAT +TGCAAC +TGCAAG +TGCATA +TGCATT +TGCATC +TGCATG +TGCACA +TGCACT +TGCACC +TGCACG +TGCAGA +TGCAGT +TGCAGC +TGCAGG +TGCTAA +TGCTAT +TGCTAC +TGCTAG +TGCTTA +TGCTTT +TGCTTC +TGCTTG +TGCTCA +TGCTCT +TGCTCC +TGCTCG +TGCTGA +TGCTGT +TGCTGC +TGCTGG +TGCCAA +TGCCAT +TGCCAC +TGCCAG +TGCCTA +TGCCTT +TGCCTC +TGCCTG +TGCCCA +TGCCCT +TGCCCC +TGCCCG +TGCCGA +TGCCGT +TGCCGC +TGCCGG +TGCGAA +TGCGAT +TGCGAC +TGCGAG +TGCGTA +TGCGTT +TGCGTC +TGCGTG +TGCGCA +TGCGCT +TGCGCC +TGCGCG +TGCGGA +TGCGGT +TGCGGC +TGCGGG +TGGAAA +TGGAAT +TGGAAC +TGGAAG +TGGATA +TGGATT +TGGATC +TGGATG +TGGACA +TGGACT +TGGACC +TGGACG +TGGAGA +TGGAGT +TGGAGC +TGGAGG +TGGTAA +TGGTAT +TGGTAC +TGGTAG +TGGTTA +TGGTTT +TGGTTC +TGGTTG +TGGTCA +TGGTCT +TGGTCC +TGGTCG +TGGTGA +TGGTGT +TGGTGC +TGGTGG +TGGCAA +TGGCAT +TGGCAC +TGGCAG +TGGCTA +TGGCTT +TGGCTC +TGGCTG +TGGCCA +TGGCCT +TGGCCC +TGGCCG +TGGCGA +TGGCGT +TGGCGC +TGGCGG +TGGGAA +TGGGAT +TGGGAC +TGGGAG +TGGGTA +TGGGTT +TGGGTC +TGGGTG +TGGGCA +TGGGCT +TGGGCC +TGGGCG +TGGGGA +TGGGGT +TGGGGC +TGGGGG +CAAAAA +CAAAAT +CAAAAC +CAAAAG +CAAATA +CAAATT +CAAATC +CAAATG +CAAACA +CAAACT +CAAACC +CAAACG +CAAAGA +CAAAGT +CAAAGC +CAAAGG +CAATAA +CAATAT +CAATAC +CAATAG +CAATTA +CAATTT +CAATTC +CAATTG +CAATCA +CAATCT +CAATCC +CAATCG +CAATGA +CAATGT +CAATGC +CAATGG +CAACAA +CAACAT +CAACAC +CAACAG +CAACTA +CAACTT +CAACTC +CAACTG +CAACCA +CAACCT +CAACCC +CAACCG +CAACGA +CAACGT +CAACGC +CAACGG +CAAGAA +CAAGAT +CAAGAC +CAAGAG +CAAGTA +CAAGTT +CAAGTC +CAAGTG +CAAGCA +CAAGCT +CAAGCC +CAAGCG +CAAGGA +CAAGGT +CAAGGC +CAAGGG +CATAAA +CATAAT +CATAAC +CATAAG +CATATA +CATATT +CATATC +CATATG +CATACA +CATACT +CATACC +CATACG +CATAGA +CATAGT +CATAGC +CATAGG +CATTAA +CATTAT +CATTAC +CATTAG +CATTTA +CATTTT +CATTTC +CATTTG +CATTCA +CATTCT +CATTCC +CATTCG +CATTGA +CATTGT +CATTGC +CATTGG +CATCAA +CATCAT +CATCAC +CATCAG +CATCTA +CATCTT +CATCTC +CATCTG +CATCCA +CATCCT +CATCCC +CATCCG +CATCGA +CATCGT +CATCGC +CATCGG +CATGAA +CATGAT +CATGAC +CATGAG +CATGTA +CATGTT +CATGTC +CATGTG +CATGCA +CATGCT +CATGCC +CATGCG +CATGGA +CATGGT +CATGGC +CATGGG +CACAAA +CACAAT +CACAAC +CACAAG +CACATA +CACATT +CACATC +CACATG +CACACA +CACACT +CACACC +CACACG +CACAGA +CACAGT +CACAGC +CACAGG +CACTAA +CACTAT +CACTAC +CACTAG +CACTTA +CACTTT +CACTTC +CACTTG +CACTCA +CACTCT +CACTCC +CACTCG +CACTGA +CACTGT +CACTGC +CACTGG +CACCAA +CACCAT +CACCAC +CACCAG +CACCTA +CACCTT +CACCTC +CACCTG +CACCCA +CACCCT +CACCCC +CACCCG +CACCGA +CACCGT +CACCGC +CACCGG +CACGAA +CACGAT +CACGAC +CACGAG +CACGTA +CACGTT +CACGTC +CACGTG +CACGCA +CACGCT +CACGCC +CACGCG +CACGGA +CACGGT +CACGGC +CACGGG +CAGAAA +CAGAAT +CAGAAC +CAGAAG +CAGATA +CAGATT +CAGATC +CAGATG +CAGACA +CAGACT +CAGACC +CAGACG +CAGAGA +CAGAGT +CAGAGC +CAGAGG +CAGTAA +CAGTAT +CAGTAC +CAGTAG +CAGTTA +CAGTTT +CAGTTC +CAGTTG +CAGTCA +CAGTCT +CAGTCC +CAGTCG +CAGTGA +CAGTGT +CAGTGC +CAGTGG +CAGCAA +CAGCAT +CAGCAC +CAGCAG +CAGCTA +CAGCTT +CAGCTC +CAGCTG +CAGCCA +CAGCCT +CAGCCC +CAGCCG +CAGCGA +CAGCGT +CAGCGC +CAGCGG +CAGGAA +CAGGAT +CAGGAC +CAGGAG +CAGGTA +CAGGTT +CAGGTC +CAGGTG +CAGGCA +CAGGCT +CAGGCC +CAGGCG +CAGGGA +CAGGGT +CAGGGC +CAGGGG +CTAAAA +CTAAAT +CTAAAC +CTAAAG +CTAATA +CTAATT +CTAATC +CTAATG +CTAACA +CTAACT +CTAACC +CTAACG +CTAAGA +CTAAGT +CTAAGC +CTAAGG +CTATAA +CTATAT +CTATAC +CTATAG +CTATTA +CTATTT +CTATTC +CTATTG +CTATCA +CTATCT +CTATCC +CTATCG +CTATGA +CTATGT +CTATGC +CTATGG +CTACAA +CTACAT +CTACAC +CTACAG +CTACTA +CTACTT +CTACTC +CTACTG +CTACCA +CTACCT +CTACCC +CTACCG +CTACGA +CTACGT +CTACGC +CTACGG +CTAGAA +CTAGAT +CTAGAC +CTAGAG +CTAGTA +CTAGTT +CTAGTC +CTAGTG +CTAGCA +CTAGCT +CTAGCC +CTAGCG +CTAGGA +CTAGGT +CTAGGC +CTAGGG +CTTAAA +CTTAAT +CTTAAC +CTTAAG +CTTATA +CTTATT +CTTATC +CTTATG +CTTACA +CTTACT +CTTACC +CTTACG +CTTAGA +CTTAGT +CTTAGC +CTTAGG +CTTTAA +CTTTAT +CTTTAC +CTTTAG +CTTTTA +CTTTTT +CTTTTC +CTTTTG +CTTTCA +CTTTCT +CTTTCC +CTTTCG +CTTTGA +CTTTGT +CTTTGC +CTTTGG +CTTCAA +CTTCAT +CTTCAC +CTTCAG +CTTCTA +CTTCTT +CTTCTC +CTTCTG +CTTCCA +CTTCCT +CTTCCC +CTTCCG +CTTCGA +CTTCGT +CTTCGC +CTTCGG +CTTGAA +CTTGAT +CTTGAC +CTTGAG +CTTGTA +CTTGTT +CTTGTC +CTTGTG +CTTGCA +CTTGCT +CTTGCC +CTTGCG +CTTGGA +CTTGGT +CTTGGC +CTTGGG +CTCAAA +CTCAAT +CTCAAC +CTCAAG +CTCATA +CTCATT +CTCATC +CTCATG +CTCACA +CTCACT +CTCACC +CTCACG +CTCAGA +CTCAGT +CTCAGC +CTCAGG +CTCTAA +CTCTAT +CTCTAC +CTCTAG +CTCTTA +CTCTTT +CTCTTC +CTCTTG +CTCTCA +CTCTCT +CTCTCC +CTCTCG +CTCTGA +CTCTGT +CTCTGC +CTCTGG +CTCCAA +CTCCAT +CTCCAC +CTCCAG +CTCCTA +CTCCTT +CTCCTC +CTCCTG +CTCCCA +CTCCCT +CTCCCC +CTCCCG +CTCCGA +CTCCGT +CTCCGC +CTCCGG +CTCGAA +CTCGAT +CTCGAC +CTCGAG +CTCGTA +CTCGTT +CTCGTC +CTCGTG +CTCGCA +CTCGCT +CTCGCC +CTCGCG +CTCGGA +CTCGGT +CTCGGC +CTCGGG +CTGAAA +CTGAAT +CTGAAC +CTGAAG +CTGATA +CTGATT +CTGATC +CTGATG +CTGACA +CTGACT +CTGACC +CTGACG +CTGAGA +CTGAGT +CTGAGC +CTGAGG +CTGTAA +CTGTAT +CTGTAC +CTGTAG +CTGTTA +CTGTTT +CTGTTC +CTGTTG +CTGTCA +CTGTCT +CTGTCC +CTGTCG +CTGTGA +CTGTGT +CTGTGC +CTGTGG +CTGCAA +CTGCAT +CTGCAC +CTGCAG +CTGCTA +CTGCTT +CTGCTC +CTGCTG +CTGCCA +CTGCCT +CTGCCC +CTGCCG +CTGCGA +CTGCGT +CTGCGC +CTGCGG +CTGGAA +CTGGAT +CTGGAC +CTGGAG +CTGGTA +CTGGTT +CTGGTC +CTGGTG +CTGGCA +CTGGCT +CTGGCC +CTGGCG +CTGGGA +CTGGGT +CTGGGC +CTGGGG +CCAAAA +CCAAAT +CCAAAC +CCAAAG +CCAATA +CCAATT +CCAATC +CCAATG +CCAACA +CCAACT +CCAACC +CCAACG +CCAAGA +CCAAGT +CCAAGC +CCAAGG +CCATAA +CCATAT +CCATAC +CCATAG +CCATTA +CCATTT +CCATTC +CCATTG +CCATCA +CCATCT +CCATCC +CCATCG +CCATGA +CCATGT +CCATGC +CCATGG +CCACAA +CCACAT +CCACAC +CCACAG +CCACTA +CCACTT +CCACTC +CCACTG +CCACCA +CCACCT +CCACCC +CCACCG +CCACGA +CCACGT +CCACGC +CCACGG +CCAGAA +CCAGAT +CCAGAC +CCAGAG +CCAGTA +CCAGTT +CCAGTC +CCAGTG +CCAGCA +CCAGCT +CCAGCC +CCAGCG +CCAGGA +CCAGGT +CCAGGC +CCAGGG +CCTAAA +CCTAAT +CCTAAC +CCTAAG +CCTATA +CCTATT +CCTATC +CCTATG +CCTACA +CCTACT +CCTACC +CCTACG +CCTAGA +CCTAGT +CCTAGC +CCTAGG +CCTTAA +CCTTAT +CCTTAC +CCTTAG +CCTTTA +CCTTTT +CCTTTC +CCTTTG +CCTTCA +CCTTCT +CCTTCC +CCTTCG +CCTTGA +CCTTGT +CCTTGC +CCTTGG +CCTCAA +CCTCAT +CCTCAC +CCTCAG +CCTCTA +CCTCTT +CCTCTC +CCTCTG +CCTCCA +CCTCCT +CCTCCC +CCTCCG +CCTCGA +CCTCGT +CCTCGC +CCTCGG +CCTGAA +CCTGAT +CCTGAC +CCTGAG +CCTGTA +CCTGTT +CCTGTC +CCTGTG +CCTGCA +CCTGCT +CCTGCC +CCTGCG +CCTGGA +CCTGGT +CCTGGC +CCTGGG +CCCAAA +CCCAAT +CCCAAC +CCCAAG +CCCATA +CCCATT +CCCATC +CCCATG +CCCACA +CCCACT +CCCACC +CCCACG +CCCAGA +CCCAGT +CCCAGC +CCCAGG +CCCTAA +CCCTAT +CCCTAC +CCCTAG +CCCTTA +CCCTTT +CCCTTC +CCCTTG +CCCTCA +CCCTCT +CCCTCC +CCCTCG +CCCTGA +CCCTGT +CCCTGC +CCCTGG +CCCCAA +CCCCAT +CCCCAC +CCCCAG +CCCCTA +CCCCTT +CCCCTC +CCCCTG +CCCCCA +CCCCCT +CCCCCC +CCCCCG +CCCCGA +CCCCGT +CCCCGC +CCCCGG +CCCGAA +CCCGAT +CCCGAC +CCCGAG +CCCGTA +CCCGTT +CCCGTC +CCCGTG +CCCGCA +CCCGCT +CCCGCC +CCCGCG +CCCGGA +CCCGGT +CCCGGC +CCCGGG +CCGAAA +CCGAAT +CCGAAC +CCGAAG +CCGATA +CCGATT +CCGATC +CCGATG +CCGACA +CCGACT +CCGACC +CCGACG +CCGAGA +CCGAGT +CCGAGC +CCGAGG +CCGTAA +CCGTAT +CCGTAC +CCGTAG +CCGTTA +CCGTTT +CCGTTC +CCGTTG +CCGTCA +CCGTCT +CCGTCC +CCGTCG +CCGTGA +CCGTGT +CCGTGC +CCGTGG +CCGCAA +CCGCAT +CCGCAC +CCGCAG +CCGCTA +CCGCTT +CCGCTC +CCGCTG +CCGCCA +CCGCCT +CCGCCC +CCGCCG +CCGCGA +CCGCGT +CCGCGC +CCGCGG +CCGGAA +CCGGAT +CCGGAC +CCGGAG +CCGGTA +CCGGTT +CCGGTC +CCGGTG +CCGGCA +CCGGCT +CCGGCC +CCGGCG +CCGGGA +CCGGGT +CCGGGC +CCGGGG +CGAAAA +CGAAAT +CGAAAC +CGAAAG +CGAATA +CGAATT +CGAATC +CGAATG +CGAACA +CGAACT +CGAACC +CGAACG +CGAAGA +CGAAGT +CGAAGC +CGAAGG +CGATAA +CGATAT +CGATAC +CGATAG +CGATTA +CGATTT +CGATTC +CGATTG +CGATCA +CGATCT +CGATCC +CGATCG +CGATGA +CGATGT +CGATGC +CGATGG +CGACAA +CGACAT +CGACAC +CGACAG +CGACTA +CGACTT +CGACTC +CGACTG +CGACCA +CGACCT +CGACCC +CGACCG +CGACGA +CGACGT +CGACGC +CGACGG +CGAGAA +CGAGAT +CGAGAC +CGAGAG +CGAGTA +CGAGTT +CGAGTC +CGAGTG +CGAGCA +CGAGCT +CGAGCC +CGAGCG +CGAGGA +CGAGGT +CGAGGC +CGAGGG +CGTAAA +CGTAAT +CGTAAC +CGTAAG +CGTATA +CGTATT +CGTATC +CGTATG +CGTACA +CGTACT +CGTACC +CGTACG +CGTAGA +CGTAGT +CGTAGC +CGTAGG +CGTTAA +CGTTAT +CGTTAC +CGTTAG +CGTTTA +CGTTTT +CGTTTC +CGTTTG +CGTTCA +CGTTCT +CGTTCC +CGTTCG +CGTTGA +CGTTGT +CGTTGC +CGTTGG +CGTCAA +CGTCAT +CGTCAC +CGTCAG +CGTCTA +CGTCTT +CGTCTC +CGTCTG +CGTCCA +CGTCCT +CGTCCC +CGTCCG +CGTCGA +CGTCGT +CGTCGC +CGTCGG +CGTGAA +CGTGAT +CGTGAC +CGTGAG +CGTGTA +CGTGTT +CGTGTC +CGTGTG +CGTGCA +CGTGCT +CGTGCC +CGTGCG +CGTGGA +CGTGGT +CGTGGC +CGTGGG +CGCAAA +CGCAAT +CGCAAC +CGCAAG +CGCATA +CGCATT +CGCATC +CGCATG +CGCACA +CGCACT +CGCACC +CGCACG +CGCAGA +CGCAGT +CGCAGC +CGCAGG +CGCTAA +CGCTAT +CGCTAC +CGCTAG +CGCTTA +CGCTTT +CGCTTC +CGCTTG +CGCTCA +CGCTCT +CGCTCC +CGCTCG +CGCTGA +CGCTGT +CGCTGC +CGCTGG +CGCCAA +CGCCAT +CGCCAC +CGCCAG +CGCCTA +CGCCTT +CGCCTC +CGCCTG +CGCCCA +CGCCCT +CGCCCC +CGCCCG +CGCCGA +CGCCGT +CGCCGC +CGCCGG +CGCGAA +CGCGAT +CGCGAC +CGCGAG +CGCGTA +CGCGTT +CGCGTC +CGCGTG +CGCGCA +CGCGCT +CGCGCC +CGCGCG +CGCGGA +CGCGGT +CGCGGC +CGCGGG +CGGAAA +CGGAAT +CGGAAC +CGGAAG +CGGATA +CGGATT +CGGATC +CGGATG +CGGACA +CGGACT +CGGACC +CGGACG +CGGAGA +CGGAGT +CGGAGC +CGGAGG +CGGTAA +CGGTAT +CGGTAC +CGGTAG +CGGTTA +CGGTTT +CGGTTC +CGGTTG +CGGTCA +CGGTCT +CGGTCC +CGGTCG +CGGTGA +CGGTGT +CGGTGC +CGGTGG +CGGCAA +CGGCAT +CGGCAC +CGGCAG +CGGCTA +CGGCTT +CGGCTC +CGGCTG +CGGCCA +CGGCCT +CGGCCC +CGGCCG +CGGCGA +CGGCGT +CGGCGC +CGGCGG +CGGGAA +CGGGAT +CGGGAC +CGGGAG +CGGGTA +CGGGTT +CGGGTC +CGGGTG +CGGGCA +CGGGCT +CGGGCC +CGGGCG +CGGGGA +CGGGGT +CGGGGC +CGGGGG +GAAAAA +GAAAAT +GAAAAC +GAAAAG +GAAATA +GAAATT +GAAATC +GAAATG +GAAACA +GAAACT +GAAACC +GAAACG +GAAAGA +GAAAGT +GAAAGC +GAAAGG +GAATAA +GAATAT +GAATAC +GAATAG +GAATTA +GAATTT +GAATTC +GAATTG +GAATCA +GAATCT +GAATCC +GAATCG +GAATGA +GAATGT +GAATGC +GAATGG +GAACAA +GAACAT +GAACAC +GAACAG +GAACTA +GAACTT +GAACTC +GAACTG +GAACCA +GAACCT +GAACCC +GAACCG +GAACGA +GAACGT +GAACGC +GAACGG +GAAGAA +GAAGAT +GAAGAC +GAAGAG +GAAGTA +GAAGTT +GAAGTC +GAAGTG +GAAGCA +GAAGCT +GAAGCC +GAAGCG +GAAGGA +GAAGGT +GAAGGC +GAAGGG +GATAAA +GATAAT +GATAAC +GATAAG +GATATA +GATATT +GATATC +GATATG +GATACA +GATACT +GATACC +GATACG +GATAGA +GATAGT +GATAGC +GATAGG +GATTAA +GATTAT +GATTAC +GATTAG +GATTTA +GATTTT +GATTTC +GATTTG +GATTCA +GATTCT +GATTCC +GATTCG +GATTGA +GATTGT +GATTGC +GATTGG +GATCAA +GATCAT +GATCAC +GATCAG +GATCTA +GATCTT +GATCTC +GATCTG +GATCCA +GATCCT +GATCCC +GATCCG +GATCGA +GATCGT +GATCGC +GATCGG +GATGAA +GATGAT +GATGAC +GATGAG +GATGTA +GATGTT +GATGTC +GATGTG +GATGCA +GATGCT +GATGCC +GATGCG +GATGGA +GATGGT +GATGGC +GATGGG +GACAAA +GACAAT +GACAAC +GACAAG +GACATA +GACATT +GACATC +GACATG +GACACA +GACACT +GACACC +GACACG +GACAGA +GACAGT +GACAGC +GACAGG +GACTAA +GACTAT +GACTAC +GACTAG +GACTTA +GACTTT +GACTTC +GACTTG +GACTCA +GACTCT +GACTCC +GACTCG +GACTGA +GACTGT +GACTGC +GACTGG +GACCAA +GACCAT +GACCAC +GACCAG +GACCTA +GACCTT +GACCTC +GACCTG +GACCCA +GACCCT +GACCCC +GACCCG +GACCGA +GACCGT +GACCGC +GACCGG +GACGAA +GACGAT +GACGAC +GACGAG +GACGTA +GACGTT +GACGTC +GACGTG +GACGCA +GACGCT +GACGCC +GACGCG +GACGGA +GACGGT +GACGGC +GACGGG +GAGAAA +GAGAAT +GAGAAC +GAGAAG +GAGATA +GAGATT +GAGATC +GAGATG +GAGACA +GAGACT +GAGACC +GAGACG +GAGAGA +GAGAGT +GAGAGC +GAGAGG +GAGTAA +GAGTAT +GAGTAC +GAGTAG +GAGTTA +GAGTTT +GAGTTC +GAGTTG +GAGTCA +GAGTCT +GAGTCC +GAGTCG +GAGTGA +GAGTGT +GAGTGC +GAGTGG +GAGCAA +GAGCAT +GAGCAC +GAGCAG +GAGCTA +GAGCTT +GAGCTC +GAGCTG +GAGCCA +GAGCCT +GAGCCC +GAGCCG +GAGCGA +GAGCGT +GAGCGC +GAGCGG +GAGGAA +GAGGAT +GAGGAC +GAGGAG +GAGGTA +GAGGTT +GAGGTC +GAGGTG +GAGGCA +GAGGCT +GAGGCC +GAGGCG +GAGGGA +GAGGGT +GAGGGC +GAGGGG +GTAAAA +GTAAAT +GTAAAC +GTAAAG +GTAATA +GTAATT +GTAATC +GTAATG +GTAACA +GTAACT +GTAACC +GTAACG +GTAAGA +GTAAGT +GTAAGC +GTAAGG +GTATAA +GTATAT +GTATAC +GTATAG +GTATTA +GTATTT +GTATTC +GTATTG +GTATCA +GTATCT +GTATCC +GTATCG +GTATGA +GTATGT +GTATGC +GTATGG +GTACAA +GTACAT +GTACAC +GTACAG +GTACTA +GTACTT +GTACTC +GTACTG +GTACCA +GTACCT +GTACCC +GTACCG +GTACGA +GTACGT +GTACGC +GTACGG +GTAGAA +GTAGAT +GTAGAC +GTAGAG +GTAGTA +GTAGTT +GTAGTC +GTAGTG +GTAGCA +GTAGCT +GTAGCC +GTAGCG +GTAGGA +GTAGGT +GTAGGC +GTAGGG +GTTAAA +GTTAAT +GTTAAC +GTTAAG +GTTATA +GTTATT +GTTATC +GTTATG +GTTACA +GTTACT +GTTACC +GTTACG +GTTAGA +GTTAGT +GTTAGC +GTTAGG +GTTTAA +GTTTAT +GTTTAC +GTTTAG +GTTTTA +GTTTTT +GTTTTC +GTTTTG +GTTTCA +GTTTCT +GTTTCC +GTTTCG +GTTTGA +GTTTGT +GTTTGC +GTTTGG +GTTCAA +GTTCAT +GTTCAC +GTTCAG +GTTCTA +GTTCTT +GTTCTC +GTTCTG +GTTCCA +GTTCCT +GTTCCC +GTTCCG +GTTCGA +GTTCGT +GTTCGC +GTTCGG +GTTGAA +GTTGAT +GTTGAC +GTTGAG +GTTGTA +GTTGTT +GTTGTC +GTTGTG +GTTGCA +GTTGCT +GTTGCC +GTTGCG +GTTGGA +GTTGGT +GTTGGC +GTTGGG +GTCAAA +GTCAAT +GTCAAC +GTCAAG +GTCATA +GTCATT +GTCATC +GTCATG +GTCACA +GTCACT +GTCACC +GTCACG +GTCAGA +GTCAGT +GTCAGC +GTCAGG +GTCTAA +GTCTAT +GTCTAC +GTCTAG +GTCTTA +GTCTTT +GTCTTC +GTCTTG +GTCTCA +GTCTCT +GTCTCC +GTCTCG +GTCTGA +GTCTGT +GTCTGC +GTCTGG +GTCCAA +GTCCAT +GTCCAC +GTCCAG +GTCCTA +GTCCTT +GTCCTC +GTCCTG +GTCCCA +GTCCCT +GTCCCC +GTCCCG +GTCCGA +GTCCGT +GTCCGC +GTCCGG +GTCGAA +GTCGAT +GTCGAC +GTCGAG +GTCGTA +GTCGTT +GTCGTC +GTCGTG +GTCGCA +GTCGCT +GTCGCC +GTCGCG +GTCGGA +GTCGGT +GTCGGC +GTCGGG +GTGAAA +GTGAAT +GTGAAC +GTGAAG +GTGATA +GTGATT +GTGATC +GTGATG +GTGACA +GTGACT +GTGACC +GTGACG +GTGAGA +GTGAGT +GTGAGC +GTGAGG +GTGTAA +GTGTAT +GTGTAC +GTGTAG +GTGTTA +GTGTTT +GTGTTC +GTGTTG +GTGTCA +GTGTCT +GTGTCC +GTGTCG +GTGTGA +GTGTGT +GTGTGC +GTGTGG +GTGCAA +GTGCAT +GTGCAC +GTGCAG +GTGCTA +GTGCTT +GTGCTC +GTGCTG +GTGCCA +GTGCCT +GTGCCC +GTGCCG +GTGCGA +GTGCGT +GTGCGC +GTGCGG +GTGGAA +GTGGAT +GTGGAC +GTGGAG +GTGGTA +GTGGTT +GTGGTC +GTGGTG +GTGGCA +GTGGCT +GTGGCC +GTGGCG +GTGGGA +GTGGGT +GTGGGC +GTGGGG +GCAAAA +GCAAAT +GCAAAC +GCAAAG +GCAATA +GCAATT +GCAATC +GCAATG +GCAACA +GCAACT +GCAACC +GCAACG +GCAAGA +GCAAGT +GCAAGC +GCAAGG +GCATAA +GCATAT +GCATAC +GCATAG +GCATTA +GCATTT +GCATTC +GCATTG +GCATCA +GCATCT +GCATCC +GCATCG +GCATGA +GCATGT +GCATGC +GCATGG +GCACAA +GCACAT +GCACAC +GCACAG +GCACTA +GCACTT +GCACTC +GCACTG +GCACCA +GCACCT +GCACCC +GCACCG +GCACGA +GCACGT +GCACGC +GCACGG +GCAGAA +GCAGAT +GCAGAC +GCAGAG +GCAGTA +GCAGTT +GCAGTC +GCAGTG +GCAGCA +GCAGCT +GCAGCC +GCAGCG +GCAGGA +GCAGGT +GCAGGC +GCAGGG +GCTAAA +GCTAAT +GCTAAC +GCTAAG +GCTATA +GCTATT +GCTATC +GCTATG +GCTACA +GCTACT +GCTACC +GCTACG +GCTAGA +GCTAGT +GCTAGC +GCTAGG +GCTTAA +GCTTAT +GCTTAC +GCTTAG +GCTTTA +GCTTTT +GCTTTC +GCTTTG +GCTTCA +GCTTCT +GCTTCC +GCTTCG +GCTTGA +GCTTGT +GCTTGC +GCTTGG +GCTCAA +GCTCAT +GCTCAC +GCTCAG +GCTCTA +GCTCTT +GCTCTC +GCTCTG +GCTCCA +GCTCCT +GCTCCC +GCTCCG +GCTCGA +GCTCGT +GCTCGC +GCTCGG +GCTGAA +GCTGAT +GCTGAC +GCTGAG +GCTGTA +GCTGTT +GCTGTC +GCTGTG +GCTGCA +GCTGCT +GCTGCC +GCTGCG +GCTGGA +GCTGGT +GCTGGC +GCTGGG +GCCAAA +GCCAAT +GCCAAC +GCCAAG +GCCATA +GCCATT +GCCATC +GCCATG +GCCACA +GCCACT +GCCACC +GCCACG +GCCAGA +GCCAGT +GCCAGC +GCCAGG +GCCTAA +GCCTAT +GCCTAC +GCCTAG +GCCTTA +GCCTTT +GCCTTC +GCCTTG +GCCTCA +GCCTCT +GCCTCC +GCCTCG +GCCTGA +GCCTGT +GCCTGC +GCCTGG +GCCCAA +GCCCAT +GCCCAC +GCCCAG +GCCCTA +GCCCTT +GCCCTC +GCCCTG +GCCCCA +GCCCCT +GCCCCC +GCCCCG +GCCCGA +GCCCGT +GCCCGC +GCCCGG +GCCGAA +GCCGAT +GCCGAC +GCCGAG +GCCGTA +GCCGTT +GCCGTC +GCCGTG +GCCGCA +GCCGCT +GCCGCC +GCCGCG +GCCGGA +GCCGGT +GCCGGC +GCCGGG +GCGAAA +GCGAAT +GCGAAC +GCGAAG +GCGATA +GCGATT +GCGATC +GCGATG +GCGACA +GCGACT +GCGACC +GCGACG +GCGAGA +GCGAGT +GCGAGC +GCGAGG +GCGTAA +GCGTAT +GCGTAC +GCGTAG +GCGTTA +GCGTTT +GCGTTC +GCGTTG +GCGTCA +GCGTCT +GCGTCC +GCGTCG +GCGTGA +GCGTGT +GCGTGC +GCGTGG +GCGCAA +GCGCAT +GCGCAC +GCGCAG +GCGCTA +GCGCTT +GCGCTC +GCGCTG +GCGCCA +GCGCCT +GCGCCC +GCGCCG +GCGCGA +GCGCGT +GCGCGC +GCGCGG +GCGGAA +GCGGAT +GCGGAC +GCGGAG +GCGGTA +GCGGTT +GCGGTC +GCGGTG +GCGGCA +GCGGCT +GCGGCC +GCGGCG +GCGGGA +GCGGGT +GCGGGC +GCGGGG +GGAAAA +GGAAAT +GGAAAC +GGAAAG +GGAATA +GGAATT +GGAATC +GGAATG +GGAACA +GGAACT +GGAACC +GGAACG +GGAAGA +GGAAGT +GGAAGC +GGAAGG +GGATAA +GGATAT +GGATAC +GGATAG +GGATTA +GGATTT +GGATTC +GGATTG +GGATCA +GGATCT +GGATCC +GGATCG +GGATGA +GGATGT +GGATGC +GGATGG +GGACAA +GGACAT +GGACAC +GGACAG +GGACTA +GGACTT +GGACTC +GGACTG +GGACCA +GGACCT +GGACCC +GGACCG +GGACGA +GGACGT +GGACGC +GGACGG +GGAGAA +GGAGAT +GGAGAC +GGAGAG +GGAGTA +GGAGTT +GGAGTC +GGAGTG +GGAGCA +GGAGCT +GGAGCC +GGAGCG +GGAGGA +GGAGGT +GGAGGC +GGAGGG +GGTAAA +GGTAAT +GGTAAC +GGTAAG +GGTATA +GGTATT +GGTATC +GGTATG +GGTACA +GGTACT +GGTACC +GGTACG +GGTAGA +GGTAGT +GGTAGC +GGTAGG +GGTTAA +GGTTAT +GGTTAC +GGTTAG +GGTTTA +GGTTTT +GGTTTC +GGTTTG +GGTTCA +GGTTCT +GGTTCC +GGTTCG +GGTTGA +GGTTGT +GGTTGC +GGTTGG +GGTCAA +GGTCAT +GGTCAC +GGTCAG +GGTCTA +GGTCTT +GGTCTC +GGTCTG +GGTCCA +GGTCCT +GGTCCC +GGTCCG +GGTCGA +GGTCGT +GGTCGC +GGTCGG +GGTGAA +GGTGAT +GGTGAC +GGTGAG +GGTGTA +GGTGTT +GGTGTC +GGTGTG +GGTGCA +GGTGCT +GGTGCC +GGTGCG +GGTGGA +GGTGGT +GGTGGC +GGTGGG +GGCAAA +GGCAAT +GGCAAC +GGCAAG +GGCATA +GGCATT +GGCATC +GGCATG +GGCACA +GGCACT +GGCACC +GGCACG +GGCAGA +GGCAGT +GGCAGC +GGCAGG +GGCTAA +GGCTAT +GGCTAC +GGCTAG +GGCTTA +GGCTTT +GGCTTC +GGCTTG +GGCTCA +GGCTCT +GGCTCC +GGCTCG +GGCTGA +GGCTGT +GGCTGC +GGCTGG +GGCCAA +GGCCAT +GGCCAC +GGCCAG +GGCCTA +GGCCTT +GGCCTC +GGCCTG +GGCCCA +GGCCCT +GGCCCC +GGCCCG +GGCCGA +GGCCGT +GGCCGC +GGCCGG +GGCGAA +GGCGAT +GGCGAC +GGCGAG +GGCGTA +GGCGTT +GGCGTC +GGCGTG +GGCGCA +GGCGCT +GGCGCC +GGCGCG +GGCGGA +GGCGGT +GGCGGC +GGCGGG +GGGAAA +GGGAAT +GGGAAC +GGGAAG +GGGATA +GGGATT +GGGATC +GGGATG +GGGACA +GGGACT +GGGACC +GGGACG +GGGAGA +GGGAGT +GGGAGC +GGGAGG +GGGTAA +GGGTAT +GGGTAC +GGGTAG +GGGTTA +GGGTTT +GGGTTC +GGGTTG +GGGTCA +GGGTCT +GGGTCC +GGGTCG +GGGTGA +GGGTGT +GGGTGC +GGGTGG +GGGCAA +GGGCAT +GGGCAC +GGGCAG +GGGCTA +GGGCTT +GGGCTC +GGGCTG +GGGCCA +GGGCCT +GGGCCC +GGGCCG +GGGCGA +GGGCGT +GGGCGC +GGGCGG +GGGGAA +GGGGAT +GGGGAC +GGGGAG +GGGGTA +GGGGTT +GGGGTC +GGGGTG +GGGGCA +GGGGCT +GGGGCC +GGGGCG +GGGGGA +GGGGGT +GGGGGC +GGGGGG diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/README.md b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..500fd976d92b217455f181bc98c77e612d9a18d9 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/README.md @@ -0,0 +1,202 @@ +--- +base_model: fabihamakhdoomi/TinyDNABERT +library_name: peft +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.12.0 \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/adapter_config.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..04c63e99bbc6b810449dece367c0afb165125b92 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/adapter_config.json @@ -0,0 +1,42 @@ +{ + "alpha_pattern": {}, + "auto_mapping": null, + "base_model_name_or_path": "fabihamakhdoomi/TinyDNABERT", + "bias": "none", + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 16, + "lora_dropout": 0.2, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": [ + "classifier", + "score" + ], + "peft_type": "LORA", + "r": 8, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "bert.encoder.layer.1.attention.self.query", + "bert.encoder.layer.0.output.dense", + "bert.encoder.layer.1.output.dense", + "bert.encoder.layer.0.intermediate.dense", + "bert.encoder.layer.1.intermediate.dense", + "bert.encoder.layer.0.attention.output.dense", + "bert.encoder.layer.1.attention.output.dense", + "bert.encoder.layer.0.attention.self.query", + "bert.encoder.layer.0.attention.self.value", + "bert.encoder.layer.1.attention.self.key", + "bert.encoder.layer.1.attention.self.value", + "bert.encoder.layer.0.attention.self.key" + ], + "task_type": "SEQ_CLS", + "use_dora": false, + "use_rslora": false +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/adapter_model.safetensors b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/adapter_model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..2dde56106faba2b62dd8bf6e2b571ca94542c3f2 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/adapter_model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e8b02d1edfd67ef18e611b35267b496c6a7d5b511db067314b80238b8d8e817f +size 151952 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/optimizer.pt b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..2f88a11591cad8df36e7e8d95939090012b0e2d4 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f5d4527e7e68b8b5f74d8b343fb4adf2d99a54b191cc9c8ea0a2f143d371c69b +size 312682 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/rng_state.pth b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..b07d245718b57ab2fe28aa8fcd87fd566f5ca37a --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a7b08a9b6cab3ce1ea404a1560c73adefdab4cd5f0d8619baa241f06062472ab +size 14244 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/scheduler.pt b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..6c655a9f3dd5d4d532f6c0a84b19b9ae5b766437 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dbc2253b3cc1ddd11640ea56d3925f925017f9aa6e1577411dac42d629457af6 +size 1064 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/special_tokens_map.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..a8b3208c2884c4efb86e49300fdd3dc877220cdf --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/special_tokens_map.json @@ -0,0 +1,7 @@ +{ + "cls_token": "[CLS]", + "mask_token": "[MASK]", + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/tokenizer.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..ebd782de6963b9032ce9374cfe30f5275be1a30b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/tokenizer.json @@ -0,0 +1,4265 @@ +{ + "version": "1.0", + "truncation": { + "direction": "Right", + "max_length": 512, + "strategy": "LongestFirst", + "stride": 0 + }, + "padding": { + "strategy": { + "Fixed": 512 + }, + "direction": "Right", + "pad_to_multiple_of": null, + "pad_id": 0, + "pad_type_id": 0, + "pad_token": "[PAD]" + }, + "added_tokens": [ + { + "id": 0, + "content": "[PAD]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 1, + "content": "[UNK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 2, + "content": "[CLS]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 3, + "content": "[SEP]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 4, + "content": "[MASK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + } + ], + "normalizer": { + "type": "BertNormalizer", + "clean_text": true, + "handle_chinese_chars": true, + "strip_accents": null, + "lowercase": false + }, + "pre_tokenizer": { + "type": "BertPreTokenizer" + }, + "post_processor": { + "type": "TemplateProcessing", + "single": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + } + ], + "pair": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "B", + "type_id": 1 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 1 + } + } + ], + "special_tokens": { + "[CLS]": { + "id": "[CLS]", + "ids": [ + 2 + ], + "tokens": [ + "[CLS]" + ] + }, + "[SEP]": { + "id": "[SEP]", + "ids": [ + 3 + ], + "tokens": [ + "[SEP]" + ] + } + } + }, + "decoder": { + "type": "WordPiece", + "prefix": "##", + "cleanup": true + }, + "model": { + "type": "WordPiece", + "unk_token": "[UNK]", + "continuing_subword_prefix": "##", + "max_input_chars_per_word": 100, + "vocab": { + "[PAD]": 0, + "[UNK]": 1, + "[CLS]": 2, + "[SEP]": 3, + "[MASK]": 4, + "AAAAAA": 5, + "AAAAAT": 6, + "AAAAAC": 7, + "AAAAAG": 8, + "AAAATA": 9, + "AAAATT": 10, + "AAAATC": 11, + "AAAATG": 12, + "AAAACA": 13, + "AAAACT": 14, + "AAAACC": 15, + "AAAACG": 16, + "AAAAGA": 17, + "AAAAGT": 18, + "AAAAGC": 19, + "AAAAGG": 20, + "AAATAA": 21, + "AAATAT": 22, + "AAATAC": 23, + "AAATAG": 24, + "AAATTA": 25, + "AAATTT": 26, + "AAATTC": 27, + "AAATTG": 28, + "AAATCA": 29, + "AAATCT": 30, + "AAATCC": 31, + "AAATCG": 32, + "AAATGA": 33, + "AAATGT": 34, + "AAATGC": 35, + "AAATGG": 36, + "AAACAA": 37, + "AAACAT": 38, + "AAACAC": 39, + "AAACAG": 40, + "AAACTA": 41, + "AAACTT": 42, + "AAACTC": 43, + "AAACTG": 44, + "AAACCA": 45, + "AAACCT": 46, + "AAACCC": 47, + "AAACCG": 48, + "AAACGA": 49, + "AAACGT": 50, + "AAACGC": 51, + "AAACGG": 52, + "AAAGAA": 53, + "AAAGAT": 54, + "AAAGAC": 55, + "AAAGAG": 56, + "AAAGTA": 57, + "AAAGTT": 58, + "AAAGTC": 59, + "AAAGTG": 60, + "AAAGCA": 61, + "AAAGCT": 62, + "AAAGCC": 63, + "AAAGCG": 64, + "AAAGGA": 65, + "AAAGGT": 66, + "AAAGGC": 67, + "AAAGGG": 68, + "AATAAA": 69, + "AATAAT": 70, + "AATAAC": 71, + "AATAAG": 72, + "AATATA": 73, + "AATATT": 74, + "AATATC": 75, + "AATATG": 76, + "AATACA": 77, + "AATACT": 78, + "AATACC": 79, + "AATACG": 80, + "AATAGA": 81, + "AATAGT": 82, + "AATAGC": 83, + "AATAGG": 84, + "AATTAA": 85, + "AATTAT": 86, + "AATTAC": 87, + "AATTAG": 88, + "AATTTA": 89, + "AATTTT": 90, + "AATTTC": 91, + "AATTTG": 92, + "AATTCA": 93, + "AATTCT": 94, + "AATTCC": 95, + "AATTCG": 96, + "AATTGA": 97, + "AATTGT": 98, + "AATTGC": 99, + "AATTGG": 100, + "AATCAA": 101, + "AATCAT": 102, + "AATCAC": 103, + "AATCAG": 104, + "AATCTA": 105, + "AATCTT": 106, + "AATCTC": 107, + "AATCTG": 108, + "AATCCA": 109, + "AATCCT": 110, + "AATCCC": 111, + "AATCCG": 112, + "AATCGA": 113, + "AATCGT": 114, + "AATCGC": 115, + "AATCGG": 116, + "AATGAA": 117, + "AATGAT": 118, + "AATGAC": 119, + "AATGAG": 120, + "AATGTA": 121, + "AATGTT": 122, + "AATGTC": 123, + "AATGTG": 124, + "AATGCA": 125, + "AATGCT": 126, + "AATGCC": 127, + "AATGCG": 128, + "AATGGA": 129, + "AATGGT": 130, + "AATGGC": 131, + "AATGGG": 132, + "AACAAA": 133, + "AACAAT": 134, + "AACAAC": 135, + "AACAAG": 136, + "AACATA": 137, + "AACATT": 138, + "AACATC": 139, + "AACATG": 140, + "AACACA": 141, + "AACACT": 142, + "AACACC": 143, + "AACACG": 144, + "AACAGA": 145, + "AACAGT": 146, + "AACAGC": 147, + "AACAGG": 148, + "AACTAA": 149, + "AACTAT": 150, + "AACTAC": 151, + "AACTAG": 152, + "AACTTA": 153, + "AACTTT": 154, + "AACTTC": 155, + "AACTTG": 156, + "AACTCA": 157, + "AACTCT": 158, + "AACTCC": 159, + "AACTCG": 160, + "AACTGA": 161, + "AACTGT": 162, + "AACTGC": 163, + "AACTGG": 164, + "AACCAA": 165, + "AACCAT": 166, + "AACCAC": 167, + "AACCAG": 168, + "AACCTA": 169, + "AACCTT": 170, + "AACCTC": 171, + "AACCTG": 172, + "AACCCA": 173, + "AACCCT": 174, + "AACCCC": 175, + "AACCCG": 176, + "AACCGA": 177, + "AACCGT": 178, + "AACCGC": 179, + "AACCGG": 180, + "AACGAA": 181, + "AACGAT": 182, + "AACGAC": 183, + "AACGAG": 184, + "AACGTA": 185, + "AACGTT": 186, + "AACGTC": 187, + "AACGTG": 188, + "AACGCA": 189, + "AACGCT": 190, + "AACGCC": 191, + "AACGCG": 192, + "AACGGA": 193, + "AACGGT": 194, + "AACGGC": 195, + "AACGGG": 196, + "AAGAAA": 197, + "AAGAAT": 198, + "AAGAAC": 199, + "AAGAAG": 200, + "AAGATA": 201, + "AAGATT": 202, + "AAGATC": 203, + "AAGATG": 204, + "AAGACA": 205, + "AAGACT": 206, + "AAGACC": 207, + "AAGACG": 208, + "AAGAGA": 209, + "AAGAGT": 210, + "AAGAGC": 211, + "AAGAGG": 212, + "AAGTAA": 213, + "AAGTAT": 214, + "AAGTAC": 215, + "AAGTAG": 216, + "AAGTTA": 217, + "AAGTTT": 218, + "AAGTTC": 219, + "AAGTTG": 220, + "AAGTCA": 221, + "AAGTCT": 222, + "AAGTCC": 223, + "AAGTCG": 224, + "AAGTGA": 225, + "AAGTGT": 226, + "AAGTGC": 227, + "AAGTGG": 228, + "AAGCAA": 229, + "AAGCAT": 230, + "AAGCAC": 231, + "AAGCAG": 232, + "AAGCTA": 233, + "AAGCTT": 234, + "AAGCTC": 235, + "AAGCTG": 236, + "AAGCCA": 237, + "AAGCCT": 238, + "AAGCCC": 239, + "AAGCCG": 240, + "AAGCGA": 241, + "AAGCGT": 242, + "AAGCGC": 243, + "AAGCGG": 244, + "AAGGAA": 245, + "AAGGAT": 246, + "AAGGAC": 247, + "AAGGAG": 248, + "AAGGTA": 249, + "AAGGTT": 250, + "AAGGTC": 251, + "AAGGTG": 252, + "AAGGCA": 253, + "AAGGCT": 254, + "AAGGCC": 255, + "AAGGCG": 256, + "AAGGGA": 257, + "AAGGGT": 258, + "AAGGGC": 259, + "AAGGGG": 260, + "ATAAAA": 261, + "ATAAAT": 262, + "ATAAAC": 263, + "ATAAAG": 264, + "ATAATA": 265, + "ATAATT": 266, + "ATAATC": 267, + "ATAATG": 268, + "ATAACA": 269, + "ATAACT": 270, + "ATAACC": 271, + "ATAACG": 272, + "ATAAGA": 273, + "ATAAGT": 274, + "ATAAGC": 275, + "ATAAGG": 276, + "ATATAA": 277, + "ATATAT": 278, + "ATATAC": 279, + "ATATAG": 280, + "ATATTA": 281, + "ATATTT": 282, + "ATATTC": 283, + "ATATTG": 284, + "ATATCA": 285, + "ATATCT": 286, + "ATATCC": 287, + "ATATCG": 288, + "ATATGA": 289, + "ATATGT": 290, + "ATATGC": 291, + "ATATGG": 292, + "ATACAA": 293, + "ATACAT": 294, + "ATACAC": 295, + "ATACAG": 296, + "ATACTA": 297, + "ATACTT": 298, + "ATACTC": 299, + "ATACTG": 300, + "ATACCA": 301, + "ATACCT": 302, + "ATACCC": 303, + "ATACCG": 304, + "ATACGA": 305, + "ATACGT": 306, + "ATACGC": 307, + "ATACGG": 308, + "ATAGAA": 309, + "ATAGAT": 310, + "ATAGAC": 311, + "ATAGAG": 312, + "ATAGTA": 313, + "ATAGTT": 314, + "ATAGTC": 315, + "ATAGTG": 316, + "ATAGCA": 317, + "ATAGCT": 318, + "ATAGCC": 319, + "ATAGCG": 320, + "ATAGGA": 321, + "ATAGGT": 322, + "ATAGGC": 323, + "ATAGGG": 324, + "ATTAAA": 325, + "ATTAAT": 326, + "ATTAAC": 327, + "ATTAAG": 328, + "ATTATA": 329, + "ATTATT": 330, + "ATTATC": 331, + "ATTATG": 332, + "ATTACA": 333, + "ATTACT": 334, + "ATTACC": 335, + "ATTACG": 336, + "ATTAGA": 337, + "ATTAGT": 338, + "ATTAGC": 339, + "ATTAGG": 340, + "ATTTAA": 341, + "ATTTAT": 342, + "ATTTAC": 343, + "ATTTAG": 344, + "ATTTTA": 345, + "ATTTTT": 346, + "ATTTTC": 347, + "ATTTTG": 348, + "ATTTCA": 349, + "ATTTCT": 350, + "ATTTCC": 351, + "ATTTCG": 352, + "ATTTGA": 353, + "ATTTGT": 354, + "ATTTGC": 355, + "ATTTGG": 356, + "ATTCAA": 357, + "ATTCAT": 358, + "ATTCAC": 359, + "ATTCAG": 360, + "ATTCTA": 361, + "ATTCTT": 362, + "ATTCTC": 363, + "ATTCTG": 364, + "ATTCCA": 365, + "ATTCCT": 366, + "ATTCCC": 367, + "ATTCCG": 368, + "ATTCGA": 369, + "ATTCGT": 370, + "ATTCGC": 371, + "ATTCGG": 372, + "ATTGAA": 373, + "ATTGAT": 374, + "ATTGAC": 375, + "ATTGAG": 376, + "ATTGTA": 377, + "ATTGTT": 378, + "ATTGTC": 379, + "ATTGTG": 380, + "ATTGCA": 381, + "ATTGCT": 382, + "ATTGCC": 383, + "ATTGCG": 384, + "ATTGGA": 385, + "ATTGGT": 386, + "ATTGGC": 387, + "ATTGGG": 388, + "ATCAAA": 389, + "ATCAAT": 390, + "ATCAAC": 391, + "ATCAAG": 392, + "ATCATA": 393, + "ATCATT": 394, + "ATCATC": 395, + "ATCATG": 396, + "ATCACA": 397, + "ATCACT": 398, + "ATCACC": 399, + "ATCACG": 400, + "ATCAGA": 401, + "ATCAGT": 402, + "ATCAGC": 403, + "ATCAGG": 404, + "ATCTAA": 405, + "ATCTAT": 406, + "ATCTAC": 407, + "ATCTAG": 408, + "ATCTTA": 409, + "ATCTTT": 410, + "ATCTTC": 411, + "ATCTTG": 412, + "ATCTCA": 413, + "ATCTCT": 414, + "ATCTCC": 415, + "ATCTCG": 416, + "ATCTGA": 417, + "ATCTGT": 418, + "ATCTGC": 419, + "ATCTGG": 420, + "ATCCAA": 421, + "ATCCAT": 422, + "ATCCAC": 423, + "ATCCAG": 424, + "ATCCTA": 425, + "ATCCTT": 426, + "ATCCTC": 427, + "ATCCTG": 428, + "ATCCCA": 429, + "ATCCCT": 430, + "ATCCCC": 431, + "ATCCCG": 432, + "ATCCGA": 433, + "ATCCGT": 434, + "ATCCGC": 435, + "ATCCGG": 436, + "ATCGAA": 437, + "ATCGAT": 438, + "ATCGAC": 439, + "ATCGAG": 440, + "ATCGTA": 441, + "ATCGTT": 442, + "ATCGTC": 443, + "ATCGTG": 444, + "ATCGCA": 445, + "ATCGCT": 446, + "ATCGCC": 447, + "ATCGCG": 448, + "ATCGGA": 449, + "ATCGGT": 450, + "ATCGGC": 451, + "ATCGGG": 452, + "ATGAAA": 453, + "ATGAAT": 454, + "ATGAAC": 455, + "ATGAAG": 456, + "ATGATA": 457, + "ATGATT": 458, + "ATGATC": 459, + "ATGATG": 460, + "ATGACA": 461, + "ATGACT": 462, + "ATGACC": 463, + "ATGACG": 464, + "ATGAGA": 465, + "ATGAGT": 466, + "ATGAGC": 467, + "ATGAGG": 468, + "ATGTAA": 469, + "ATGTAT": 470, + "ATGTAC": 471, + "ATGTAG": 472, + "ATGTTA": 473, + "ATGTTT": 474, + "ATGTTC": 475, + "ATGTTG": 476, + "ATGTCA": 477, + "ATGTCT": 478, + "ATGTCC": 479, + "ATGTCG": 480, + "ATGTGA": 481, + "ATGTGT": 482, + "ATGTGC": 483, + "ATGTGG": 484, + "ATGCAA": 485, + "ATGCAT": 486, + "ATGCAC": 487, + "ATGCAG": 488, + "ATGCTA": 489, + "ATGCTT": 490, + "ATGCTC": 491, + "ATGCTG": 492, + "ATGCCA": 493, + "ATGCCT": 494, + "ATGCCC": 495, + "ATGCCG": 496, + "ATGCGA": 497, + "ATGCGT": 498, + "ATGCGC": 499, + "ATGCGG": 500, + "ATGGAA": 501, + "ATGGAT": 502, + "ATGGAC": 503, + "ATGGAG": 504, + "ATGGTA": 505, + "ATGGTT": 506, + "ATGGTC": 507, + "ATGGTG": 508, + "ATGGCA": 509, + "ATGGCT": 510, + "ATGGCC": 511, + "ATGGCG": 512, + "ATGGGA": 513, + "ATGGGT": 514, + "ATGGGC": 515, + "ATGGGG": 516, + "ACAAAA": 517, + "ACAAAT": 518, + "ACAAAC": 519, + "ACAAAG": 520, + "ACAATA": 521, + "ACAATT": 522, + "ACAATC": 523, + "ACAATG": 524, + "ACAACA": 525, + "ACAACT": 526, + "ACAACC": 527, + "ACAACG": 528, + "ACAAGA": 529, + "ACAAGT": 530, + "ACAAGC": 531, + "ACAAGG": 532, + "ACATAA": 533, + "ACATAT": 534, + "ACATAC": 535, + "ACATAG": 536, + "ACATTA": 537, + "ACATTT": 538, + "ACATTC": 539, + "ACATTG": 540, + "ACATCA": 541, + "ACATCT": 542, + "ACATCC": 543, + "ACATCG": 544, + "ACATGA": 545, + "ACATGT": 546, + "ACATGC": 547, + "ACATGG": 548, + "ACACAA": 549, + "ACACAT": 550, + "ACACAC": 551, + "ACACAG": 552, + "ACACTA": 553, + "ACACTT": 554, + "ACACTC": 555, + "ACACTG": 556, + "ACACCA": 557, + "ACACCT": 558, + "ACACCC": 559, + "ACACCG": 560, + "ACACGA": 561, + "ACACGT": 562, + "ACACGC": 563, + "ACACGG": 564, + "ACAGAA": 565, + "ACAGAT": 566, + "ACAGAC": 567, + "ACAGAG": 568, + "ACAGTA": 569, + "ACAGTT": 570, + "ACAGTC": 571, + "ACAGTG": 572, + "ACAGCA": 573, + "ACAGCT": 574, + "ACAGCC": 575, + "ACAGCG": 576, + "ACAGGA": 577, + "ACAGGT": 578, + "ACAGGC": 579, + "ACAGGG": 580, + "ACTAAA": 581, + "ACTAAT": 582, + "ACTAAC": 583, + "ACTAAG": 584, + "ACTATA": 585, + "ACTATT": 586, + "ACTATC": 587, + "ACTATG": 588, + "ACTACA": 589, + "ACTACT": 590, + "ACTACC": 591, + "ACTACG": 592, + "ACTAGA": 593, + "ACTAGT": 594, + "ACTAGC": 595, + "ACTAGG": 596, + "ACTTAA": 597, + "ACTTAT": 598, + "ACTTAC": 599, + "ACTTAG": 600, + "ACTTTA": 601, + "ACTTTT": 602, + "ACTTTC": 603, + "ACTTTG": 604, + "ACTTCA": 605, + "ACTTCT": 606, + "ACTTCC": 607, + "ACTTCG": 608, + "ACTTGA": 609, + "ACTTGT": 610, + "ACTTGC": 611, + "ACTTGG": 612, + "ACTCAA": 613, + "ACTCAT": 614, + "ACTCAC": 615, + "ACTCAG": 616, + "ACTCTA": 617, + "ACTCTT": 618, + "ACTCTC": 619, + "ACTCTG": 620, + "ACTCCA": 621, + "ACTCCT": 622, + "ACTCCC": 623, + "ACTCCG": 624, + "ACTCGA": 625, + "ACTCGT": 626, + "ACTCGC": 627, + "ACTCGG": 628, + "ACTGAA": 629, + "ACTGAT": 630, + "ACTGAC": 631, + "ACTGAG": 632, + "ACTGTA": 633, + "ACTGTT": 634, + "ACTGTC": 635, + "ACTGTG": 636, + "ACTGCA": 637, + "ACTGCT": 638, + "ACTGCC": 639, + "ACTGCG": 640, + "ACTGGA": 641, + "ACTGGT": 642, + "ACTGGC": 643, + "ACTGGG": 644, + "ACCAAA": 645, + "ACCAAT": 646, + "ACCAAC": 647, + "ACCAAG": 648, + "ACCATA": 649, + "ACCATT": 650, + "ACCATC": 651, + "ACCATG": 652, + "ACCACA": 653, + "ACCACT": 654, + "ACCACC": 655, + "ACCACG": 656, + "ACCAGA": 657, + "ACCAGT": 658, + "ACCAGC": 659, + "ACCAGG": 660, + "ACCTAA": 661, + "ACCTAT": 662, + "ACCTAC": 663, + "ACCTAG": 664, + "ACCTTA": 665, + "ACCTTT": 666, + "ACCTTC": 667, + "ACCTTG": 668, + "ACCTCA": 669, + "ACCTCT": 670, + "ACCTCC": 671, + "ACCTCG": 672, + "ACCTGA": 673, + "ACCTGT": 674, + "ACCTGC": 675, + "ACCTGG": 676, + "ACCCAA": 677, + "ACCCAT": 678, + "ACCCAC": 679, + "ACCCAG": 680, + "ACCCTA": 681, + "ACCCTT": 682, + "ACCCTC": 683, + "ACCCTG": 684, + "ACCCCA": 685, + "ACCCCT": 686, + "ACCCCC": 687, + "ACCCCG": 688, + "ACCCGA": 689, + "ACCCGT": 690, + "ACCCGC": 691, + "ACCCGG": 692, + "ACCGAA": 693, + "ACCGAT": 694, + "ACCGAC": 695, + "ACCGAG": 696, + "ACCGTA": 697, + "ACCGTT": 698, + "ACCGTC": 699, + "ACCGTG": 700, + "ACCGCA": 701, + "ACCGCT": 702, + "ACCGCC": 703, + "ACCGCG": 704, + "ACCGGA": 705, + "ACCGGT": 706, + "ACCGGC": 707, + "ACCGGG": 708, + "ACGAAA": 709, + "ACGAAT": 710, + "ACGAAC": 711, + "ACGAAG": 712, + "ACGATA": 713, + "ACGATT": 714, + "ACGATC": 715, + "ACGATG": 716, + "ACGACA": 717, + "ACGACT": 718, + "ACGACC": 719, + "ACGACG": 720, + "ACGAGA": 721, + "ACGAGT": 722, + "ACGAGC": 723, + "ACGAGG": 724, + "ACGTAA": 725, + "ACGTAT": 726, + "ACGTAC": 727, + "ACGTAG": 728, + "ACGTTA": 729, + "ACGTTT": 730, + "ACGTTC": 731, + "ACGTTG": 732, + "ACGTCA": 733, + "ACGTCT": 734, + "ACGTCC": 735, + "ACGTCG": 736, + "ACGTGA": 737, + "ACGTGT": 738, + "ACGTGC": 739, + "ACGTGG": 740, + "ACGCAA": 741, + "ACGCAT": 742, + "ACGCAC": 743, + "ACGCAG": 744, + "ACGCTA": 745, + "ACGCTT": 746, + "ACGCTC": 747, + "ACGCTG": 748, + "ACGCCA": 749, + "ACGCCT": 750, + "ACGCCC": 751, + "ACGCCG": 752, + "ACGCGA": 753, + "ACGCGT": 754, + "ACGCGC": 755, + "ACGCGG": 756, + "ACGGAA": 757, + "ACGGAT": 758, + "ACGGAC": 759, + "ACGGAG": 760, + "ACGGTA": 761, + "ACGGTT": 762, + "ACGGTC": 763, + "ACGGTG": 764, + "ACGGCA": 765, + "ACGGCT": 766, + "ACGGCC": 767, + "ACGGCG": 768, + "ACGGGA": 769, + "ACGGGT": 770, + "ACGGGC": 771, + "ACGGGG": 772, + "AGAAAA": 773, + "AGAAAT": 774, + "AGAAAC": 775, + "AGAAAG": 776, + "AGAATA": 777, + "AGAATT": 778, + "AGAATC": 779, + "AGAATG": 780, + "AGAACA": 781, + "AGAACT": 782, + "AGAACC": 783, + "AGAACG": 784, + "AGAAGA": 785, + "AGAAGT": 786, + "AGAAGC": 787, + "AGAAGG": 788, + "AGATAA": 789, + "AGATAT": 790, + "AGATAC": 791, + "AGATAG": 792, + "AGATTA": 793, + "AGATTT": 794, + "AGATTC": 795, + "AGATTG": 796, + "AGATCA": 797, + "AGATCT": 798, + "AGATCC": 799, + "AGATCG": 800, + "AGATGA": 801, + "AGATGT": 802, + "AGATGC": 803, + "AGATGG": 804, + "AGACAA": 805, + "AGACAT": 806, + "AGACAC": 807, + "AGACAG": 808, + "AGACTA": 809, + "AGACTT": 810, + "AGACTC": 811, + "AGACTG": 812, + "AGACCA": 813, + "AGACCT": 814, + "AGACCC": 815, + "AGACCG": 816, + "AGACGA": 817, + "AGACGT": 818, + "AGACGC": 819, + "AGACGG": 820, + "AGAGAA": 821, + "AGAGAT": 822, + "AGAGAC": 823, + "AGAGAG": 824, + "AGAGTA": 825, + "AGAGTT": 826, + "AGAGTC": 827, + "AGAGTG": 828, + "AGAGCA": 829, + "AGAGCT": 830, + "AGAGCC": 831, + "AGAGCG": 832, + "AGAGGA": 833, + "AGAGGT": 834, + "AGAGGC": 835, + "AGAGGG": 836, + "AGTAAA": 837, + "AGTAAT": 838, + "AGTAAC": 839, + "AGTAAG": 840, + "AGTATA": 841, + "AGTATT": 842, + "AGTATC": 843, + "AGTATG": 844, + "AGTACA": 845, + "AGTACT": 846, + "AGTACC": 847, + "AGTACG": 848, + "AGTAGA": 849, + "AGTAGT": 850, + "AGTAGC": 851, + "AGTAGG": 852, + "AGTTAA": 853, + "AGTTAT": 854, + "AGTTAC": 855, + "AGTTAG": 856, + "AGTTTA": 857, + "AGTTTT": 858, + "AGTTTC": 859, + "AGTTTG": 860, + "AGTTCA": 861, + "AGTTCT": 862, + "AGTTCC": 863, + "AGTTCG": 864, + "AGTTGA": 865, + "AGTTGT": 866, + "AGTTGC": 867, + "AGTTGG": 868, + "AGTCAA": 869, + "AGTCAT": 870, + "AGTCAC": 871, + "AGTCAG": 872, + "AGTCTA": 873, + "AGTCTT": 874, + "AGTCTC": 875, + "AGTCTG": 876, + "AGTCCA": 877, + "AGTCCT": 878, + "AGTCCC": 879, + "AGTCCG": 880, + "AGTCGA": 881, + "AGTCGT": 882, + "AGTCGC": 883, + "AGTCGG": 884, + "AGTGAA": 885, + "AGTGAT": 886, + "AGTGAC": 887, + "AGTGAG": 888, + "AGTGTA": 889, + "AGTGTT": 890, + "AGTGTC": 891, + "AGTGTG": 892, + "AGTGCA": 893, + "AGTGCT": 894, + "AGTGCC": 895, + "AGTGCG": 896, + "AGTGGA": 897, + "AGTGGT": 898, + "AGTGGC": 899, + "AGTGGG": 900, + "AGCAAA": 901, + "AGCAAT": 902, + "AGCAAC": 903, + "AGCAAG": 904, + "AGCATA": 905, + "AGCATT": 906, + "AGCATC": 907, + "AGCATG": 908, + "AGCACA": 909, + "AGCACT": 910, + "AGCACC": 911, + "AGCACG": 912, + "AGCAGA": 913, + "AGCAGT": 914, + "AGCAGC": 915, + "AGCAGG": 916, + "AGCTAA": 917, + "AGCTAT": 918, + "AGCTAC": 919, + "AGCTAG": 920, + "AGCTTA": 921, + "AGCTTT": 922, + "AGCTTC": 923, + "AGCTTG": 924, + "AGCTCA": 925, + "AGCTCT": 926, + "AGCTCC": 927, + "AGCTCG": 928, + "AGCTGA": 929, + "AGCTGT": 930, + "AGCTGC": 931, + "AGCTGG": 932, + "AGCCAA": 933, + "AGCCAT": 934, + "AGCCAC": 935, + "AGCCAG": 936, + "AGCCTA": 937, + "AGCCTT": 938, + "AGCCTC": 939, + "AGCCTG": 940, + "AGCCCA": 941, + "AGCCCT": 942, + "AGCCCC": 943, + "AGCCCG": 944, + "AGCCGA": 945, + "AGCCGT": 946, + "AGCCGC": 947, + "AGCCGG": 948, + "AGCGAA": 949, + "AGCGAT": 950, + "AGCGAC": 951, + "AGCGAG": 952, + "AGCGTA": 953, + "AGCGTT": 954, + "AGCGTC": 955, + "AGCGTG": 956, + "AGCGCA": 957, + "AGCGCT": 958, + "AGCGCC": 959, + "AGCGCG": 960, + "AGCGGA": 961, + "AGCGGT": 962, + "AGCGGC": 963, + "AGCGGG": 964, + "AGGAAA": 965, + "AGGAAT": 966, + "AGGAAC": 967, + "AGGAAG": 968, + "AGGATA": 969, + "AGGATT": 970, + "AGGATC": 971, + "AGGATG": 972, + "AGGACA": 973, + "AGGACT": 974, + "AGGACC": 975, + "AGGACG": 976, + "AGGAGA": 977, + "AGGAGT": 978, + "AGGAGC": 979, + "AGGAGG": 980, + "AGGTAA": 981, + "AGGTAT": 982, + "AGGTAC": 983, + "AGGTAG": 984, + "AGGTTA": 985, + "AGGTTT": 986, + "AGGTTC": 987, + "AGGTTG": 988, + "AGGTCA": 989, + "AGGTCT": 990, + "AGGTCC": 991, + "AGGTCG": 992, + "AGGTGA": 993, + "AGGTGT": 994, + "AGGTGC": 995, + "AGGTGG": 996, + "AGGCAA": 997, + "AGGCAT": 998, + "AGGCAC": 999, + "AGGCAG": 1000, + "AGGCTA": 1001, + "AGGCTT": 1002, + "AGGCTC": 1003, + "AGGCTG": 1004, + "AGGCCA": 1005, + "AGGCCT": 1006, + "AGGCCC": 1007, + "AGGCCG": 1008, + "AGGCGA": 1009, + "AGGCGT": 1010, + "AGGCGC": 1011, + "AGGCGG": 1012, + "AGGGAA": 1013, + "AGGGAT": 1014, + "AGGGAC": 1015, + "AGGGAG": 1016, + "AGGGTA": 1017, + "AGGGTT": 1018, + "AGGGTC": 1019, + "AGGGTG": 1020, + "AGGGCA": 1021, + "AGGGCT": 1022, + "AGGGCC": 1023, + "AGGGCG": 1024, + "AGGGGA": 1025, + "AGGGGT": 1026, + "AGGGGC": 1027, + "AGGGGG": 1028, + "TAAAAA": 1029, + "TAAAAT": 1030, + "TAAAAC": 1031, + "TAAAAG": 1032, + "TAAATA": 1033, + "TAAATT": 1034, + "TAAATC": 1035, + "TAAATG": 1036, + "TAAACA": 1037, + "TAAACT": 1038, + "TAAACC": 1039, + "TAAACG": 1040, + "TAAAGA": 1041, + "TAAAGT": 1042, + "TAAAGC": 1043, + "TAAAGG": 1044, + "TAATAA": 1045, + "TAATAT": 1046, + "TAATAC": 1047, + "TAATAG": 1048, + "TAATTA": 1049, + "TAATTT": 1050, + "TAATTC": 1051, + "TAATTG": 1052, + "TAATCA": 1053, + "TAATCT": 1054, + "TAATCC": 1055, + "TAATCG": 1056, + "TAATGA": 1057, + "TAATGT": 1058, + "TAATGC": 1059, + "TAATGG": 1060, + "TAACAA": 1061, + "TAACAT": 1062, + "TAACAC": 1063, + "TAACAG": 1064, + "TAACTA": 1065, + "TAACTT": 1066, + "TAACTC": 1067, + "TAACTG": 1068, + "TAACCA": 1069, + "TAACCT": 1070, + "TAACCC": 1071, + "TAACCG": 1072, + "TAACGA": 1073, + "TAACGT": 1074, + "TAACGC": 1075, + "TAACGG": 1076, + "TAAGAA": 1077, + "TAAGAT": 1078, + "TAAGAC": 1079, + "TAAGAG": 1080, + "TAAGTA": 1081, + "TAAGTT": 1082, + "TAAGTC": 1083, + "TAAGTG": 1084, + "TAAGCA": 1085, + "TAAGCT": 1086, + "TAAGCC": 1087, + "TAAGCG": 1088, + "TAAGGA": 1089, + "TAAGGT": 1090, + "TAAGGC": 1091, + "TAAGGG": 1092, + "TATAAA": 1093, + "TATAAT": 1094, + "TATAAC": 1095, + "TATAAG": 1096, + "TATATA": 1097, + "TATATT": 1098, + "TATATC": 1099, + "TATATG": 1100, + "TATACA": 1101, + "TATACT": 1102, + "TATACC": 1103, + "TATACG": 1104, + "TATAGA": 1105, + "TATAGT": 1106, + "TATAGC": 1107, + "TATAGG": 1108, + "TATTAA": 1109, + "TATTAT": 1110, + "TATTAC": 1111, + "TATTAG": 1112, + "TATTTA": 1113, + "TATTTT": 1114, + "TATTTC": 1115, + "TATTTG": 1116, + "TATTCA": 1117, + "TATTCT": 1118, + "TATTCC": 1119, + "TATTCG": 1120, + "TATTGA": 1121, + "TATTGT": 1122, + "TATTGC": 1123, + "TATTGG": 1124, + "TATCAA": 1125, + "TATCAT": 1126, + "TATCAC": 1127, + "TATCAG": 1128, + "TATCTA": 1129, + "TATCTT": 1130, + "TATCTC": 1131, + "TATCTG": 1132, + "TATCCA": 1133, + "TATCCT": 1134, + "TATCCC": 1135, + "TATCCG": 1136, + "TATCGA": 1137, + "TATCGT": 1138, + "TATCGC": 1139, + "TATCGG": 1140, + "TATGAA": 1141, + "TATGAT": 1142, + "TATGAC": 1143, + "TATGAG": 1144, + "TATGTA": 1145, + "TATGTT": 1146, + "TATGTC": 1147, + "TATGTG": 1148, + "TATGCA": 1149, + "TATGCT": 1150, + "TATGCC": 1151, + "TATGCG": 1152, + "TATGGA": 1153, + "TATGGT": 1154, + "TATGGC": 1155, + "TATGGG": 1156, + "TACAAA": 1157, + "TACAAT": 1158, + "TACAAC": 1159, + "TACAAG": 1160, + "TACATA": 1161, + "TACATT": 1162, + "TACATC": 1163, + "TACATG": 1164, + "TACACA": 1165, + "TACACT": 1166, + "TACACC": 1167, + "TACACG": 1168, + "TACAGA": 1169, + "TACAGT": 1170, + "TACAGC": 1171, + "TACAGG": 1172, + "TACTAA": 1173, + "TACTAT": 1174, + "TACTAC": 1175, + "TACTAG": 1176, + "TACTTA": 1177, + "TACTTT": 1178, + "TACTTC": 1179, + "TACTTG": 1180, + "TACTCA": 1181, + "TACTCT": 1182, + "TACTCC": 1183, + "TACTCG": 1184, + "TACTGA": 1185, + "TACTGT": 1186, + "TACTGC": 1187, + "TACTGG": 1188, + "TACCAA": 1189, + "TACCAT": 1190, + "TACCAC": 1191, + "TACCAG": 1192, + "TACCTA": 1193, + "TACCTT": 1194, + "TACCTC": 1195, + "TACCTG": 1196, + "TACCCA": 1197, + "TACCCT": 1198, + "TACCCC": 1199, + "TACCCG": 1200, + "TACCGA": 1201, + "TACCGT": 1202, + "TACCGC": 1203, + "TACCGG": 1204, + "TACGAA": 1205, + "TACGAT": 1206, + "TACGAC": 1207, + "TACGAG": 1208, + "TACGTA": 1209, + "TACGTT": 1210, + "TACGTC": 1211, + "TACGTG": 1212, + "TACGCA": 1213, + "TACGCT": 1214, + "TACGCC": 1215, + "TACGCG": 1216, + "TACGGA": 1217, + "TACGGT": 1218, + "TACGGC": 1219, + "TACGGG": 1220, + "TAGAAA": 1221, + "TAGAAT": 1222, + "TAGAAC": 1223, + "TAGAAG": 1224, + "TAGATA": 1225, + "TAGATT": 1226, + "TAGATC": 1227, + "TAGATG": 1228, + "TAGACA": 1229, + "TAGACT": 1230, + "TAGACC": 1231, + "TAGACG": 1232, + "TAGAGA": 1233, + "TAGAGT": 1234, + "TAGAGC": 1235, + "TAGAGG": 1236, + "TAGTAA": 1237, + "TAGTAT": 1238, + "TAGTAC": 1239, + "TAGTAG": 1240, + "TAGTTA": 1241, + "TAGTTT": 1242, + "TAGTTC": 1243, + "TAGTTG": 1244, + "TAGTCA": 1245, + "TAGTCT": 1246, + "TAGTCC": 1247, + "TAGTCG": 1248, + "TAGTGA": 1249, + "TAGTGT": 1250, + "TAGTGC": 1251, + "TAGTGG": 1252, + "TAGCAA": 1253, + "TAGCAT": 1254, + "TAGCAC": 1255, + "TAGCAG": 1256, + "TAGCTA": 1257, + "TAGCTT": 1258, + "TAGCTC": 1259, + "TAGCTG": 1260, + "TAGCCA": 1261, + "TAGCCT": 1262, + "TAGCCC": 1263, + "TAGCCG": 1264, + "TAGCGA": 1265, + "TAGCGT": 1266, + "TAGCGC": 1267, + "TAGCGG": 1268, + "TAGGAA": 1269, + "TAGGAT": 1270, + "TAGGAC": 1271, + "TAGGAG": 1272, + "TAGGTA": 1273, + "TAGGTT": 1274, + "TAGGTC": 1275, + "TAGGTG": 1276, + "TAGGCA": 1277, + "TAGGCT": 1278, + "TAGGCC": 1279, + "TAGGCG": 1280, + "TAGGGA": 1281, + "TAGGGT": 1282, + "TAGGGC": 1283, + "TAGGGG": 1284, + "TTAAAA": 1285, + "TTAAAT": 1286, + "TTAAAC": 1287, + "TTAAAG": 1288, + "TTAATA": 1289, + "TTAATT": 1290, + "TTAATC": 1291, + "TTAATG": 1292, + "TTAACA": 1293, + "TTAACT": 1294, + "TTAACC": 1295, + "TTAACG": 1296, + "TTAAGA": 1297, + "TTAAGT": 1298, + "TTAAGC": 1299, + "TTAAGG": 1300, + "TTATAA": 1301, + "TTATAT": 1302, + "TTATAC": 1303, + "TTATAG": 1304, + "TTATTA": 1305, + "TTATTT": 1306, + "TTATTC": 1307, + "TTATTG": 1308, + "TTATCA": 1309, + "TTATCT": 1310, + "TTATCC": 1311, + "TTATCG": 1312, + "TTATGA": 1313, + "TTATGT": 1314, + "TTATGC": 1315, + "TTATGG": 1316, + "TTACAA": 1317, + "TTACAT": 1318, + "TTACAC": 1319, + "TTACAG": 1320, + "TTACTA": 1321, + "TTACTT": 1322, + "TTACTC": 1323, + "TTACTG": 1324, + "TTACCA": 1325, + "TTACCT": 1326, + "TTACCC": 1327, + "TTACCG": 1328, + "TTACGA": 1329, + "TTACGT": 1330, + "TTACGC": 1331, + "TTACGG": 1332, + "TTAGAA": 1333, + "TTAGAT": 1334, + "TTAGAC": 1335, + "TTAGAG": 1336, + "TTAGTA": 1337, + "TTAGTT": 1338, + "TTAGTC": 1339, + "TTAGTG": 1340, + "TTAGCA": 1341, + "TTAGCT": 1342, + "TTAGCC": 1343, + "TTAGCG": 1344, + "TTAGGA": 1345, + "TTAGGT": 1346, + "TTAGGC": 1347, + "TTAGGG": 1348, + "TTTAAA": 1349, + "TTTAAT": 1350, + "TTTAAC": 1351, + "TTTAAG": 1352, + "TTTATA": 1353, + "TTTATT": 1354, + "TTTATC": 1355, + "TTTATG": 1356, + "TTTACA": 1357, + "TTTACT": 1358, + "TTTACC": 1359, + "TTTACG": 1360, + "TTTAGA": 1361, + "TTTAGT": 1362, + "TTTAGC": 1363, + "TTTAGG": 1364, + "TTTTAA": 1365, + "TTTTAT": 1366, + "TTTTAC": 1367, + "TTTTAG": 1368, + "TTTTTA": 1369, + "TTTTTT": 1370, + "TTTTTC": 1371, + "TTTTTG": 1372, + "TTTTCA": 1373, + "TTTTCT": 1374, + "TTTTCC": 1375, + "TTTTCG": 1376, + "TTTTGA": 1377, + "TTTTGT": 1378, + "TTTTGC": 1379, + "TTTTGG": 1380, + "TTTCAA": 1381, + "TTTCAT": 1382, + "TTTCAC": 1383, + "TTTCAG": 1384, + "TTTCTA": 1385, + "TTTCTT": 1386, + "TTTCTC": 1387, + "TTTCTG": 1388, + "TTTCCA": 1389, + "TTTCCT": 1390, + "TTTCCC": 1391, + "TTTCCG": 1392, + "TTTCGA": 1393, + "TTTCGT": 1394, + "TTTCGC": 1395, + "TTTCGG": 1396, + "TTTGAA": 1397, + "TTTGAT": 1398, + "TTTGAC": 1399, + "TTTGAG": 1400, + "TTTGTA": 1401, + "TTTGTT": 1402, + "TTTGTC": 1403, + "TTTGTG": 1404, + "TTTGCA": 1405, + "TTTGCT": 1406, + "TTTGCC": 1407, + "TTTGCG": 1408, + "TTTGGA": 1409, + "TTTGGT": 1410, + "TTTGGC": 1411, + "TTTGGG": 1412, + "TTCAAA": 1413, + "TTCAAT": 1414, + "TTCAAC": 1415, + "TTCAAG": 1416, + "TTCATA": 1417, + "TTCATT": 1418, + "TTCATC": 1419, + "TTCATG": 1420, + "TTCACA": 1421, + "TTCACT": 1422, + "TTCACC": 1423, + "TTCACG": 1424, + "TTCAGA": 1425, + "TTCAGT": 1426, + "TTCAGC": 1427, + "TTCAGG": 1428, + "TTCTAA": 1429, + "TTCTAT": 1430, + "TTCTAC": 1431, + "TTCTAG": 1432, + "TTCTTA": 1433, + "TTCTTT": 1434, + "TTCTTC": 1435, + "TTCTTG": 1436, + "TTCTCA": 1437, + "TTCTCT": 1438, + "TTCTCC": 1439, + "TTCTCG": 1440, + "TTCTGA": 1441, + "TTCTGT": 1442, + "TTCTGC": 1443, + "TTCTGG": 1444, + "TTCCAA": 1445, + "TTCCAT": 1446, + "TTCCAC": 1447, + "TTCCAG": 1448, + "TTCCTA": 1449, + "TTCCTT": 1450, + "TTCCTC": 1451, + "TTCCTG": 1452, + "TTCCCA": 1453, + "TTCCCT": 1454, + "TTCCCC": 1455, + "TTCCCG": 1456, + "TTCCGA": 1457, + "TTCCGT": 1458, + "TTCCGC": 1459, + "TTCCGG": 1460, + "TTCGAA": 1461, + "TTCGAT": 1462, + "TTCGAC": 1463, + "TTCGAG": 1464, + "TTCGTA": 1465, + "TTCGTT": 1466, + "TTCGTC": 1467, + "TTCGTG": 1468, + "TTCGCA": 1469, + "TTCGCT": 1470, + "TTCGCC": 1471, + "TTCGCG": 1472, + "TTCGGA": 1473, + "TTCGGT": 1474, + "TTCGGC": 1475, + "TTCGGG": 1476, + "TTGAAA": 1477, + "TTGAAT": 1478, + "TTGAAC": 1479, + "TTGAAG": 1480, + "TTGATA": 1481, + "TTGATT": 1482, + "TTGATC": 1483, + "TTGATG": 1484, + "TTGACA": 1485, + "TTGACT": 1486, + "TTGACC": 1487, + "TTGACG": 1488, + "TTGAGA": 1489, + "TTGAGT": 1490, + "TTGAGC": 1491, + "TTGAGG": 1492, + "TTGTAA": 1493, + "TTGTAT": 1494, + "TTGTAC": 1495, + "TTGTAG": 1496, + "TTGTTA": 1497, + "TTGTTT": 1498, + "TTGTTC": 1499, + "TTGTTG": 1500, + "TTGTCA": 1501, + "TTGTCT": 1502, + "TTGTCC": 1503, + "TTGTCG": 1504, + "TTGTGA": 1505, + "TTGTGT": 1506, + "TTGTGC": 1507, + "TTGTGG": 1508, + "TTGCAA": 1509, + "TTGCAT": 1510, + "TTGCAC": 1511, + "TTGCAG": 1512, + "TTGCTA": 1513, + "TTGCTT": 1514, + "TTGCTC": 1515, + "TTGCTG": 1516, + "TTGCCA": 1517, + "TTGCCT": 1518, + "TTGCCC": 1519, + "TTGCCG": 1520, + "TTGCGA": 1521, + "TTGCGT": 1522, + "TTGCGC": 1523, + "TTGCGG": 1524, + "TTGGAA": 1525, + "TTGGAT": 1526, + "TTGGAC": 1527, + "TTGGAG": 1528, + "TTGGTA": 1529, + "TTGGTT": 1530, + "TTGGTC": 1531, + "TTGGTG": 1532, + "TTGGCA": 1533, + "TTGGCT": 1534, + "TTGGCC": 1535, + "TTGGCG": 1536, + "TTGGGA": 1537, + "TTGGGT": 1538, + "TTGGGC": 1539, + "TTGGGG": 1540, + "TCAAAA": 1541, + "TCAAAT": 1542, + "TCAAAC": 1543, + "TCAAAG": 1544, + "TCAATA": 1545, + "TCAATT": 1546, + "TCAATC": 1547, + "TCAATG": 1548, + "TCAACA": 1549, + "TCAACT": 1550, + "TCAACC": 1551, + "TCAACG": 1552, + "TCAAGA": 1553, + "TCAAGT": 1554, + "TCAAGC": 1555, + "TCAAGG": 1556, + "TCATAA": 1557, + "TCATAT": 1558, + "TCATAC": 1559, + "TCATAG": 1560, + "TCATTA": 1561, + "TCATTT": 1562, + "TCATTC": 1563, + "TCATTG": 1564, + "TCATCA": 1565, + "TCATCT": 1566, + "TCATCC": 1567, + "TCATCG": 1568, + "TCATGA": 1569, + "TCATGT": 1570, + "TCATGC": 1571, + "TCATGG": 1572, + "TCACAA": 1573, + "TCACAT": 1574, + "TCACAC": 1575, + "TCACAG": 1576, + "TCACTA": 1577, + "TCACTT": 1578, + "TCACTC": 1579, + "TCACTG": 1580, + "TCACCA": 1581, + "TCACCT": 1582, + "TCACCC": 1583, + "TCACCG": 1584, + "TCACGA": 1585, + "TCACGT": 1586, + "TCACGC": 1587, + "TCACGG": 1588, + "TCAGAA": 1589, + "TCAGAT": 1590, + "TCAGAC": 1591, + "TCAGAG": 1592, + "TCAGTA": 1593, + "TCAGTT": 1594, + "TCAGTC": 1595, + "TCAGTG": 1596, + "TCAGCA": 1597, + "TCAGCT": 1598, + "TCAGCC": 1599, + "TCAGCG": 1600, + "TCAGGA": 1601, + "TCAGGT": 1602, + "TCAGGC": 1603, + "TCAGGG": 1604, + "TCTAAA": 1605, + "TCTAAT": 1606, + "TCTAAC": 1607, + "TCTAAG": 1608, + "TCTATA": 1609, + "TCTATT": 1610, + "TCTATC": 1611, + "TCTATG": 1612, + "TCTACA": 1613, + "TCTACT": 1614, + "TCTACC": 1615, + "TCTACG": 1616, + "TCTAGA": 1617, + "TCTAGT": 1618, + "TCTAGC": 1619, + "TCTAGG": 1620, + "TCTTAA": 1621, + "TCTTAT": 1622, + "TCTTAC": 1623, + "TCTTAG": 1624, + "TCTTTA": 1625, + "TCTTTT": 1626, + "TCTTTC": 1627, + "TCTTTG": 1628, + "TCTTCA": 1629, + "TCTTCT": 1630, + "TCTTCC": 1631, + "TCTTCG": 1632, + "TCTTGA": 1633, + "TCTTGT": 1634, + "TCTTGC": 1635, + "TCTTGG": 1636, + "TCTCAA": 1637, + "TCTCAT": 1638, + "TCTCAC": 1639, + "TCTCAG": 1640, + "TCTCTA": 1641, + "TCTCTT": 1642, + "TCTCTC": 1643, + "TCTCTG": 1644, + "TCTCCA": 1645, + "TCTCCT": 1646, + "TCTCCC": 1647, + "TCTCCG": 1648, + "TCTCGA": 1649, + "TCTCGT": 1650, + "TCTCGC": 1651, + "TCTCGG": 1652, + "TCTGAA": 1653, + "TCTGAT": 1654, + "TCTGAC": 1655, + "TCTGAG": 1656, + "TCTGTA": 1657, + "TCTGTT": 1658, + "TCTGTC": 1659, + "TCTGTG": 1660, + "TCTGCA": 1661, + "TCTGCT": 1662, + "TCTGCC": 1663, + "TCTGCG": 1664, + "TCTGGA": 1665, + "TCTGGT": 1666, + "TCTGGC": 1667, + "TCTGGG": 1668, + "TCCAAA": 1669, + "TCCAAT": 1670, + "TCCAAC": 1671, + "TCCAAG": 1672, + "TCCATA": 1673, + "TCCATT": 1674, + "TCCATC": 1675, + "TCCATG": 1676, + "TCCACA": 1677, + "TCCACT": 1678, + "TCCACC": 1679, + "TCCACG": 1680, + "TCCAGA": 1681, + "TCCAGT": 1682, + "TCCAGC": 1683, + "TCCAGG": 1684, + "TCCTAA": 1685, + "TCCTAT": 1686, + "TCCTAC": 1687, + "TCCTAG": 1688, + "TCCTTA": 1689, + "TCCTTT": 1690, + "TCCTTC": 1691, + "TCCTTG": 1692, + "TCCTCA": 1693, + "TCCTCT": 1694, + "TCCTCC": 1695, + "TCCTCG": 1696, + "TCCTGA": 1697, + "TCCTGT": 1698, + "TCCTGC": 1699, + "TCCTGG": 1700, + "TCCCAA": 1701, + "TCCCAT": 1702, + "TCCCAC": 1703, + "TCCCAG": 1704, + "TCCCTA": 1705, + "TCCCTT": 1706, + "TCCCTC": 1707, + "TCCCTG": 1708, + "TCCCCA": 1709, + "TCCCCT": 1710, + "TCCCCC": 1711, + "TCCCCG": 1712, + "TCCCGA": 1713, + "TCCCGT": 1714, + "TCCCGC": 1715, + "TCCCGG": 1716, + "TCCGAA": 1717, + "TCCGAT": 1718, + "TCCGAC": 1719, + "TCCGAG": 1720, + "TCCGTA": 1721, + "TCCGTT": 1722, + "TCCGTC": 1723, + "TCCGTG": 1724, + "TCCGCA": 1725, + "TCCGCT": 1726, + "TCCGCC": 1727, + "TCCGCG": 1728, + "TCCGGA": 1729, + "TCCGGT": 1730, + "TCCGGC": 1731, + "TCCGGG": 1732, + "TCGAAA": 1733, + "TCGAAT": 1734, + "TCGAAC": 1735, + "TCGAAG": 1736, + "TCGATA": 1737, + "TCGATT": 1738, + "TCGATC": 1739, + "TCGATG": 1740, + "TCGACA": 1741, + "TCGACT": 1742, + "TCGACC": 1743, + "TCGACG": 1744, + "TCGAGA": 1745, + "TCGAGT": 1746, + "TCGAGC": 1747, + "TCGAGG": 1748, + "TCGTAA": 1749, + "TCGTAT": 1750, + "TCGTAC": 1751, + "TCGTAG": 1752, + "TCGTTA": 1753, + "TCGTTT": 1754, + "TCGTTC": 1755, + "TCGTTG": 1756, + "TCGTCA": 1757, + "TCGTCT": 1758, + "TCGTCC": 1759, + "TCGTCG": 1760, + "TCGTGA": 1761, + "TCGTGT": 1762, + "TCGTGC": 1763, + "TCGTGG": 1764, + "TCGCAA": 1765, + "TCGCAT": 1766, + "TCGCAC": 1767, + "TCGCAG": 1768, + "TCGCTA": 1769, + "TCGCTT": 1770, + "TCGCTC": 1771, + "TCGCTG": 1772, + "TCGCCA": 1773, + "TCGCCT": 1774, + "TCGCCC": 1775, + "TCGCCG": 1776, + "TCGCGA": 1777, + "TCGCGT": 1778, + "TCGCGC": 1779, + "TCGCGG": 1780, + "TCGGAA": 1781, + "TCGGAT": 1782, + "TCGGAC": 1783, + "TCGGAG": 1784, + "TCGGTA": 1785, + "TCGGTT": 1786, + "TCGGTC": 1787, + "TCGGTG": 1788, + "TCGGCA": 1789, + "TCGGCT": 1790, + "TCGGCC": 1791, + "TCGGCG": 1792, + "TCGGGA": 1793, + "TCGGGT": 1794, + "TCGGGC": 1795, + "TCGGGG": 1796, + "TGAAAA": 1797, + "TGAAAT": 1798, + "TGAAAC": 1799, + "TGAAAG": 1800, + "TGAATA": 1801, + "TGAATT": 1802, + "TGAATC": 1803, + "TGAATG": 1804, + "TGAACA": 1805, + "TGAACT": 1806, + "TGAACC": 1807, + "TGAACG": 1808, + "TGAAGA": 1809, + "TGAAGT": 1810, + "TGAAGC": 1811, + "TGAAGG": 1812, + "TGATAA": 1813, + "TGATAT": 1814, + "TGATAC": 1815, + "TGATAG": 1816, + "TGATTA": 1817, + "TGATTT": 1818, + "TGATTC": 1819, + "TGATTG": 1820, + "TGATCA": 1821, + "TGATCT": 1822, + "TGATCC": 1823, + "TGATCG": 1824, + "TGATGA": 1825, + "TGATGT": 1826, + "TGATGC": 1827, + "TGATGG": 1828, + "TGACAA": 1829, + "TGACAT": 1830, + "TGACAC": 1831, + "TGACAG": 1832, + "TGACTA": 1833, + "TGACTT": 1834, + "TGACTC": 1835, + "TGACTG": 1836, + "TGACCA": 1837, + "TGACCT": 1838, + "TGACCC": 1839, + "TGACCG": 1840, + "TGACGA": 1841, + "TGACGT": 1842, + "TGACGC": 1843, + "TGACGG": 1844, + "TGAGAA": 1845, + "TGAGAT": 1846, + "TGAGAC": 1847, + "TGAGAG": 1848, + "TGAGTA": 1849, + "TGAGTT": 1850, + "TGAGTC": 1851, + "TGAGTG": 1852, + "TGAGCA": 1853, + "TGAGCT": 1854, + "TGAGCC": 1855, + "TGAGCG": 1856, + "TGAGGA": 1857, + "TGAGGT": 1858, + "TGAGGC": 1859, + "TGAGGG": 1860, + "TGTAAA": 1861, + "TGTAAT": 1862, + "TGTAAC": 1863, + "TGTAAG": 1864, + "TGTATA": 1865, + "TGTATT": 1866, + "TGTATC": 1867, + "TGTATG": 1868, + "TGTACA": 1869, + "TGTACT": 1870, + "TGTACC": 1871, + "TGTACG": 1872, + "TGTAGA": 1873, + "TGTAGT": 1874, + "TGTAGC": 1875, + "TGTAGG": 1876, + "TGTTAA": 1877, + "TGTTAT": 1878, + "TGTTAC": 1879, + "TGTTAG": 1880, + "TGTTTA": 1881, + "TGTTTT": 1882, + "TGTTTC": 1883, + "TGTTTG": 1884, + "TGTTCA": 1885, + "TGTTCT": 1886, + "TGTTCC": 1887, + "TGTTCG": 1888, + "TGTTGA": 1889, + "TGTTGT": 1890, + "TGTTGC": 1891, + "TGTTGG": 1892, + "TGTCAA": 1893, + "TGTCAT": 1894, + "TGTCAC": 1895, + "TGTCAG": 1896, + "TGTCTA": 1897, + "TGTCTT": 1898, + "TGTCTC": 1899, + "TGTCTG": 1900, + "TGTCCA": 1901, + "TGTCCT": 1902, + "TGTCCC": 1903, + "TGTCCG": 1904, + "TGTCGA": 1905, + "TGTCGT": 1906, + "TGTCGC": 1907, + "TGTCGG": 1908, + "TGTGAA": 1909, + "TGTGAT": 1910, + "TGTGAC": 1911, + "TGTGAG": 1912, + "TGTGTA": 1913, + "TGTGTT": 1914, + "TGTGTC": 1915, + "TGTGTG": 1916, + "TGTGCA": 1917, + "TGTGCT": 1918, + "TGTGCC": 1919, + "TGTGCG": 1920, + "TGTGGA": 1921, + "TGTGGT": 1922, + "TGTGGC": 1923, + "TGTGGG": 1924, + "TGCAAA": 1925, + "TGCAAT": 1926, + "TGCAAC": 1927, + "TGCAAG": 1928, + "TGCATA": 1929, + "TGCATT": 1930, + "TGCATC": 1931, + "TGCATG": 1932, + "TGCACA": 1933, + "TGCACT": 1934, + "TGCACC": 1935, + "TGCACG": 1936, + "TGCAGA": 1937, + "TGCAGT": 1938, + "TGCAGC": 1939, + "TGCAGG": 1940, + "TGCTAA": 1941, + "TGCTAT": 1942, + "TGCTAC": 1943, + "TGCTAG": 1944, + "TGCTTA": 1945, + "TGCTTT": 1946, + "TGCTTC": 1947, + "TGCTTG": 1948, + "TGCTCA": 1949, + "TGCTCT": 1950, + "TGCTCC": 1951, + "TGCTCG": 1952, + "TGCTGA": 1953, + "TGCTGT": 1954, + "TGCTGC": 1955, + "TGCTGG": 1956, + "TGCCAA": 1957, + "TGCCAT": 1958, + "TGCCAC": 1959, + "TGCCAG": 1960, + "TGCCTA": 1961, + "TGCCTT": 1962, + "TGCCTC": 1963, + "TGCCTG": 1964, + "TGCCCA": 1965, + "TGCCCT": 1966, + "TGCCCC": 1967, + "TGCCCG": 1968, + "TGCCGA": 1969, + "TGCCGT": 1970, + "TGCCGC": 1971, + "TGCCGG": 1972, + "TGCGAA": 1973, + "TGCGAT": 1974, + "TGCGAC": 1975, + "TGCGAG": 1976, + "TGCGTA": 1977, + "TGCGTT": 1978, + "TGCGTC": 1979, + "TGCGTG": 1980, + "TGCGCA": 1981, + "TGCGCT": 1982, + "TGCGCC": 1983, + "TGCGCG": 1984, + "TGCGGA": 1985, + "TGCGGT": 1986, + "TGCGGC": 1987, + "TGCGGG": 1988, + "TGGAAA": 1989, + "TGGAAT": 1990, + "TGGAAC": 1991, + "TGGAAG": 1992, + "TGGATA": 1993, + "TGGATT": 1994, + "TGGATC": 1995, + "TGGATG": 1996, + "TGGACA": 1997, + "TGGACT": 1998, + "TGGACC": 1999, + "TGGACG": 2000, + "TGGAGA": 2001, + "TGGAGT": 2002, + "TGGAGC": 2003, + "TGGAGG": 2004, + "TGGTAA": 2005, + "TGGTAT": 2006, + "TGGTAC": 2007, + "TGGTAG": 2008, + "TGGTTA": 2009, + "TGGTTT": 2010, + "TGGTTC": 2011, + "TGGTTG": 2012, + "TGGTCA": 2013, + "TGGTCT": 2014, + "TGGTCC": 2015, + "TGGTCG": 2016, + "TGGTGA": 2017, + "TGGTGT": 2018, + "TGGTGC": 2019, + "TGGTGG": 2020, + "TGGCAA": 2021, + "TGGCAT": 2022, + "TGGCAC": 2023, + "TGGCAG": 2024, + "TGGCTA": 2025, + "TGGCTT": 2026, + "TGGCTC": 2027, + "TGGCTG": 2028, + "TGGCCA": 2029, + "TGGCCT": 2030, + "TGGCCC": 2031, + "TGGCCG": 2032, + "TGGCGA": 2033, + "TGGCGT": 2034, + "TGGCGC": 2035, + "TGGCGG": 2036, + "TGGGAA": 2037, + "TGGGAT": 2038, + "TGGGAC": 2039, + "TGGGAG": 2040, + "TGGGTA": 2041, + "TGGGTT": 2042, + "TGGGTC": 2043, + "TGGGTG": 2044, + "TGGGCA": 2045, + "TGGGCT": 2046, + "TGGGCC": 2047, + "TGGGCG": 2048, + "TGGGGA": 2049, + "TGGGGT": 2050, + "TGGGGC": 2051, + "TGGGGG": 2052, + "CAAAAA": 2053, + "CAAAAT": 2054, + "CAAAAC": 2055, + "CAAAAG": 2056, + "CAAATA": 2057, + "CAAATT": 2058, + "CAAATC": 2059, + "CAAATG": 2060, + "CAAACA": 2061, + "CAAACT": 2062, + "CAAACC": 2063, + "CAAACG": 2064, + "CAAAGA": 2065, + "CAAAGT": 2066, + "CAAAGC": 2067, + "CAAAGG": 2068, + "CAATAA": 2069, + "CAATAT": 2070, + "CAATAC": 2071, + "CAATAG": 2072, + "CAATTA": 2073, + "CAATTT": 2074, + "CAATTC": 2075, + "CAATTG": 2076, + "CAATCA": 2077, + "CAATCT": 2078, + "CAATCC": 2079, + "CAATCG": 2080, + "CAATGA": 2081, + "CAATGT": 2082, + "CAATGC": 2083, + "CAATGG": 2084, + "CAACAA": 2085, + "CAACAT": 2086, + "CAACAC": 2087, + "CAACAG": 2088, + "CAACTA": 2089, + "CAACTT": 2090, + "CAACTC": 2091, + "CAACTG": 2092, + "CAACCA": 2093, + "CAACCT": 2094, + "CAACCC": 2095, + "CAACCG": 2096, + "CAACGA": 2097, + "CAACGT": 2098, + "CAACGC": 2099, + "CAACGG": 2100, + "CAAGAA": 2101, + "CAAGAT": 2102, + "CAAGAC": 2103, + "CAAGAG": 2104, + "CAAGTA": 2105, + "CAAGTT": 2106, + "CAAGTC": 2107, + "CAAGTG": 2108, + "CAAGCA": 2109, + "CAAGCT": 2110, + "CAAGCC": 2111, + "CAAGCG": 2112, + "CAAGGA": 2113, + "CAAGGT": 2114, + "CAAGGC": 2115, + "CAAGGG": 2116, + "CATAAA": 2117, + "CATAAT": 2118, + "CATAAC": 2119, + "CATAAG": 2120, + "CATATA": 2121, + "CATATT": 2122, + "CATATC": 2123, + "CATATG": 2124, + "CATACA": 2125, + "CATACT": 2126, + "CATACC": 2127, + "CATACG": 2128, + "CATAGA": 2129, + "CATAGT": 2130, + "CATAGC": 2131, + "CATAGG": 2132, + "CATTAA": 2133, + "CATTAT": 2134, + "CATTAC": 2135, + "CATTAG": 2136, + "CATTTA": 2137, + "CATTTT": 2138, + "CATTTC": 2139, + "CATTTG": 2140, + "CATTCA": 2141, + "CATTCT": 2142, + "CATTCC": 2143, + "CATTCG": 2144, + "CATTGA": 2145, + "CATTGT": 2146, + "CATTGC": 2147, + "CATTGG": 2148, + "CATCAA": 2149, + "CATCAT": 2150, + "CATCAC": 2151, + "CATCAG": 2152, + "CATCTA": 2153, + "CATCTT": 2154, + "CATCTC": 2155, + "CATCTG": 2156, + "CATCCA": 2157, + "CATCCT": 2158, + "CATCCC": 2159, + "CATCCG": 2160, + "CATCGA": 2161, + "CATCGT": 2162, + "CATCGC": 2163, + "CATCGG": 2164, + "CATGAA": 2165, + "CATGAT": 2166, + "CATGAC": 2167, + "CATGAG": 2168, + "CATGTA": 2169, + "CATGTT": 2170, + "CATGTC": 2171, + "CATGTG": 2172, + "CATGCA": 2173, + "CATGCT": 2174, + "CATGCC": 2175, + "CATGCG": 2176, + "CATGGA": 2177, + "CATGGT": 2178, + "CATGGC": 2179, + "CATGGG": 2180, + "CACAAA": 2181, + "CACAAT": 2182, + "CACAAC": 2183, + "CACAAG": 2184, + "CACATA": 2185, + "CACATT": 2186, + "CACATC": 2187, + "CACATG": 2188, + "CACACA": 2189, + "CACACT": 2190, + "CACACC": 2191, + "CACACG": 2192, + "CACAGA": 2193, + "CACAGT": 2194, + "CACAGC": 2195, + "CACAGG": 2196, + "CACTAA": 2197, + "CACTAT": 2198, + "CACTAC": 2199, + "CACTAG": 2200, + "CACTTA": 2201, + "CACTTT": 2202, + "CACTTC": 2203, + "CACTTG": 2204, + "CACTCA": 2205, + "CACTCT": 2206, + "CACTCC": 2207, + "CACTCG": 2208, + "CACTGA": 2209, + "CACTGT": 2210, + "CACTGC": 2211, + "CACTGG": 2212, + "CACCAA": 2213, + "CACCAT": 2214, + "CACCAC": 2215, + "CACCAG": 2216, + "CACCTA": 2217, + "CACCTT": 2218, + "CACCTC": 2219, + "CACCTG": 2220, + "CACCCA": 2221, + "CACCCT": 2222, + "CACCCC": 2223, + "CACCCG": 2224, + "CACCGA": 2225, + "CACCGT": 2226, + "CACCGC": 2227, + "CACCGG": 2228, + "CACGAA": 2229, + "CACGAT": 2230, + "CACGAC": 2231, + "CACGAG": 2232, + "CACGTA": 2233, + "CACGTT": 2234, + "CACGTC": 2235, + "CACGTG": 2236, + "CACGCA": 2237, + "CACGCT": 2238, + "CACGCC": 2239, + "CACGCG": 2240, + "CACGGA": 2241, + "CACGGT": 2242, + "CACGGC": 2243, + "CACGGG": 2244, + "CAGAAA": 2245, + "CAGAAT": 2246, + "CAGAAC": 2247, + "CAGAAG": 2248, + "CAGATA": 2249, + "CAGATT": 2250, + "CAGATC": 2251, + "CAGATG": 2252, + "CAGACA": 2253, + "CAGACT": 2254, + "CAGACC": 2255, + "CAGACG": 2256, + "CAGAGA": 2257, + "CAGAGT": 2258, + "CAGAGC": 2259, + "CAGAGG": 2260, + "CAGTAA": 2261, + "CAGTAT": 2262, + "CAGTAC": 2263, + "CAGTAG": 2264, + "CAGTTA": 2265, + "CAGTTT": 2266, + "CAGTTC": 2267, + "CAGTTG": 2268, + "CAGTCA": 2269, + "CAGTCT": 2270, + "CAGTCC": 2271, + "CAGTCG": 2272, + "CAGTGA": 2273, + "CAGTGT": 2274, + "CAGTGC": 2275, + "CAGTGG": 2276, + "CAGCAA": 2277, + "CAGCAT": 2278, + "CAGCAC": 2279, + "CAGCAG": 2280, + "CAGCTA": 2281, + "CAGCTT": 2282, + "CAGCTC": 2283, + "CAGCTG": 2284, + "CAGCCA": 2285, + "CAGCCT": 2286, + "CAGCCC": 2287, + "CAGCCG": 2288, + "CAGCGA": 2289, + "CAGCGT": 2290, + "CAGCGC": 2291, + "CAGCGG": 2292, + "CAGGAA": 2293, + "CAGGAT": 2294, + "CAGGAC": 2295, + "CAGGAG": 2296, + "CAGGTA": 2297, + "CAGGTT": 2298, + "CAGGTC": 2299, + "CAGGTG": 2300, + "CAGGCA": 2301, + "CAGGCT": 2302, + "CAGGCC": 2303, + "CAGGCG": 2304, + "CAGGGA": 2305, + "CAGGGT": 2306, + "CAGGGC": 2307, + "CAGGGG": 2308, + "CTAAAA": 2309, + "CTAAAT": 2310, + "CTAAAC": 2311, + "CTAAAG": 2312, + "CTAATA": 2313, + "CTAATT": 2314, + "CTAATC": 2315, + "CTAATG": 2316, + "CTAACA": 2317, + "CTAACT": 2318, + "CTAACC": 2319, + "CTAACG": 2320, + "CTAAGA": 2321, + "CTAAGT": 2322, + "CTAAGC": 2323, + "CTAAGG": 2324, + "CTATAA": 2325, + "CTATAT": 2326, + "CTATAC": 2327, + "CTATAG": 2328, + "CTATTA": 2329, + "CTATTT": 2330, + "CTATTC": 2331, + "CTATTG": 2332, + "CTATCA": 2333, + "CTATCT": 2334, + "CTATCC": 2335, + "CTATCG": 2336, + "CTATGA": 2337, + "CTATGT": 2338, + "CTATGC": 2339, + "CTATGG": 2340, + "CTACAA": 2341, + "CTACAT": 2342, + "CTACAC": 2343, + "CTACAG": 2344, + "CTACTA": 2345, + "CTACTT": 2346, + "CTACTC": 2347, + "CTACTG": 2348, + "CTACCA": 2349, + "CTACCT": 2350, + "CTACCC": 2351, + "CTACCG": 2352, + "CTACGA": 2353, + "CTACGT": 2354, + "CTACGC": 2355, + "CTACGG": 2356, + "CTAGAA": 2357, + "CTAGAT": 2358, + "CTAGAC": 2359, + "CTAGAG": 2360, + "CTAGTA": 2361, + "CTAGTT": 2362, + "CTAGTC": 2363, + "CTAGTG": 2364, + "CTAGCA": 2365, + "CTAGCT": 2366, + "CTAGCC": 2367, + "CTAGCG": 2368, + "CTAGGA": 2369, + "CTAGGT": 2370, + "CTAGGC": 2371, + "CTAGGG": 2372, + "CTTAAA": 2373, + "CTTAAT": 2374, + "CTTAAC": 2375, + "CTTAAG": 2376, + "CTTATA": 2377, + "CTTATT": 2378, + "CTTATC": 2379, + "CTTATG": 2380, + "CTTACA": 2381, + "CTTACT": 2382, + "CTTACC": 2383, + "CTTACG": 2384, + "CTTAGA": 2385, + "CTTAGT": 2386, + "CTTAGC": 2387, + "CTTAGG": 2388, + "CTTTAA": 2389, + "CTTTAT": 2390, + "CTTTAC": 2391, + "CTTTAG": 2392, + "CTTTTA": 2393, + "CTTTTT": 2394, + "CTTTTC": 2395, + "CTTTTG": 2396, + "CTTTCA": 2397, + "CTTTCT": 2398, + "CTTTCC": 2399, + "CTTTCG": 2400, + "CTTTGA": 2401, + "CTTTGT": 2402, + "CTTTGC": 2403, + "CTTTGG": 2404, + "CTTCAA": 2405, + "CTTCAT": 2406, + "CTTCAC": 2407, + "CTTCAG": 2408, + "CTTCTA": 2409, + "CTTCTT": 2410, + "CTTCTC": 2411, + "CTTCTG": 2412, + "CTTCCA": 2413, + "CTTCCT": 2414, + "CTTCCC": 2415, + "CTTCCG": 2416, + "CTTCGA": 2417, + "CTTCGT": 2418, + "CTTCGC": 2419, + "CTTCGG": 2420, + "CTTGAA": 2421, + "CTTGAT": 2422, + "CTTGAC": 2423, + "CTTGAG": 2424, + "CTTGTA": 2425, + "CTTGTT": 2426, + "CTTGTC": 2427, + "CTTGTG": 2428, + "CTTGCA": 2429, + "CTTGCT": 2430, + "CTTGCC": 2431, + "CTTGCG": 2432, + "CTTGGA": 2433, + "CTTGGT": 2434, + "CTTGGC": 2435, + "CTTGGG": 2436, + "CTCAAA": 2437, + "CTCAAT": 2438, + "CTCAAC": 2439, + "CTCAAG": 2440, + "CTCATA": 2441, + "CTCATT": 2442, + "CTCATC": 2443, + "CTCATG": 2444, + "CTCACA": 2445, + "CTCACT": 2446, + "CTCACC": 2447, + "CTCACG": 2448, + "CTCAGA": 2449, + "CTCAGT": 2450, + "CTCAGC": 2451, + "CTCAGG": 2452, + "CTCTAA": 2453, + "CTCTAT": 2454, + "CTCTAC": 2455, + "CTCTAG": 2456, + "CTCTTA": 2457, + "CTCTTT": 2458, + "CTCTTC": 2459, + "CTCTTG": 2460, + "CTCTCA": 2461, + "CTCTCT": 2462, + "CTCTCC": 2463, + "CTCTCG": 2464, + "CTCTGA": 2465, + "CTCTGT": 2466, + "CTCTGC": 2467, + "CTCTGG": 2468, + "CTCCAA": 2469, + "CTCCAT": 2470, + "CTCCAC": 2471, + "CTCCAG": 2472, + "CTCCTA": 2473, + "CTCCTT": 2474, + "CTCCTC": 2475, + "CTCCTG": 2476, + "CTCCCA": 2477, + "CTCCCT": 2478, + "CTCCCC": 2479, + "CTCCCG": 2480, + "CTCCGA": 2481, + "CTCCGT": 2482, + "CTCCGC": 2483, + "CTCCGG": 2484, + "CTCGAA": 2485, + "CTCGAT": 2486, + "CTCGAC": 2487, + "CTCGAG": 2488, + "CTCGTA": 2489, + "CTCGTT": 2490, + "CTCGTC": 2491, + "CTCGTG": 2492, + "CTCGCA": 2493, + "CTCGCT": 2494, + "CTCGCC": 2495, + "CTCGCG": 2496, + "CTCGGA": 2497, + "CTCGGT": 2498, + "CTCGGC": 2499, + "CTCGGG": 2500, + "CTGAAA": 2501, + "CTGAAT": 2502, + "CTGAAC": 2503, + "CTGAAG": 2504, + "CTGATA": 2505, + "CTGATT": 2506, + "CTGATC": 2507, + "CTGATG": 2508, + "CTGACA": 2509, + "CTGACT": 2510, + "CTGACC": 2511, + "CTGACG": 2512, + "CTGAGA": 2513, + "CTGAGT": 2514, + "CTGAGC": 2515, + "CTGAGG": 2516, + "CTGTAA": 2517, + "CTGTAT": 2518, + "CTGTAC": 2519, + "CTGTAG": 2520, + "CTGTTA": 2521, + "CTGTTT": 2522, + "CTGTTC": 2523, + "CTGTTG": 2524, + "CTGTCA": 2525, + "CTGTCT": 2526, + "CTGTCC": 2527, + "CTGTCG": 2528, + "CTGTGA": 2529, + "CTGTGT": 2530, + "CTGTGC": 2531, + "CTGTGG": 2532, + "CTGCAA": 2533, + "CTGCAT": 2534, + "CTGCAC": 2535, + "CTGCAG": 2536, + "CTGCTA": 2537, + "CTGCTT": 2538, + "CTGCTC": 2539, + "CTGCTG": 2540, + "CTGCCA": 2541, + "CTGCCT": 2542, + "CTGCCC": 2543, + "CTGCCG": 2544, + "CTGCGA": 2545, + "CTGCGT": 2546, + "CTGCGC": 2547, + "CTGCGG": 2548, + "CTGGAA": 2549, + "CTGGAT": 2550, + "CTGGAC": 2551, + "CTGGAG": 2552, + "CTGGTA": 2553, + "CTGGTT": 2554, + "CTGGTC": 2555, + "CTGGTG": 2556, + "CTGGCA": 2557, + "CTGGCT": 2558, + "CTGGCC": 2559, + "CTGGCG": 2560, + "CTGGGA": 2561, + "CTGGGT": 2562, + "CTGGGC": 2563, + "CTGGGG": 2564, + "CCAAAA": 2565, + "CCAAAT": 2566, + "CCAAAC": 2567, + "CCAAAG": 2568, + "CCAATA": 2569, + "CCAATT": 2570, + "CCAATC": 2571, + "CCAATG": 2572, + "CCAACA": 2573, + "CCAACT": 2574, + "CCAACC": 2575, + "CCAACG": 2576, + "CCAAGA": 2577, + "CCAAGT": 2578, + "CCAAGC": 2579, + "CCAAGG": 2580, + "CCATAA": 2581, + "CCATAT": 2582, + "CCATAC": 2583, + "CCATAG": 2584, + "CCATTA": 2585, + "CCATTT": 2586, + "CCATTC": 2587, + "CCATTG": 2588, + "CCATCA": 2589, + "CCATCT": 2590, + "CCATCC": 2591, + "CCATCG": 2592, + "CCATGA": 2593, + "CCATGT": 2594, + "CCATGC": 2595, + "CCATGG": 2596, + "CCACAA": 2597, + "CCACAT": 2598, + "CCACAC": 2599, + "CCACAG": 2600, + "CCACTA": 2601, + "CCACTT": 2602, + "CCACTC": 2603, + "CCACTG": 2604, + "CCACCA": 2605, + "CCACCT": 2606, + "CCACCC": 2607, + "CCACCG": 2608, + "CCACGA": 2609, + "CCACGT": 2610, + "CCACGC": 2611, + "CCACGG": 2612, + "CCAGAA": 2613, + "CCAGAT": 2614, + "CCAGAC": 2615, + "CCAGAG": 2616, + "CCAGTA": 2617, + "CCAGTT": 2618, + "CCAGTC": 2619, + "CCAGTG": 2620, + "CCAGCA": 2621, + "CCAGCT": 2622, + "CCAGCC": 2623, + "CCAGCG": 2624, + "CCAGGA": 2625, + "CCAGGT": 2626, + "CCAGGC": 2627, + "CCAGGG": 2628, + "CCTAAA": 2629, + "CCTAAT": 2630, + "CCTAAC": 2631, + "CCTAAG": 2632, + "CCTATA": 2633, + "CCTATT": 2634, + "CCTATC": 2635, + "CCTATG": 2636, + "CCTACA": 2637, + "CCTACT": 2638, + "CCTACC": 2639, + "CCTACG": 2640, + "CCTAGA": 2641, + "CCTAGT": 2642, + "CCTAGC": 2643, + "CCTAGG": 2644, + "CCTTAA": 2645, + "CCTTAT": 2646, + "CCTTAC": 2647, + "CCTTAG": 2648, + "CCTTTA": 2649, + "CCTTTT": 2650, + "CCTTTC": 2651, + "CCTTTG": 2652, + "CCTTCA": 2653, + "CCTTCT": 2654, + "CCTTCC": 2655, + "CCTTCG": 2656, + "CCTTGA": 2657, + "CCTTGT": 2658, + "CCTTGC": 2659, + "CCTTGG": 2660, + "CCTCAA": 2661, + "CCTCAT": 2662, + "CCTCAC": 2663, + "CCTCAG": 2664, + "CCTCTA": 2665, + "CCTCTT": 2666, + "CCTCTC": 2667, + "CCTCTG": 2668, + "CCTCCA": 2669, + "CCTCCT": 2670, + "CCTCCC": 2671, + "CCTCCG": 2672, + "CCTCGA": 2673, + "CCTCGT": 2674, + "CCTCGC": 2675, + "CCTCGG": 2676, + "CCTGAA": 2677, + "CCTGAT": 2678, + "CCTGAC": 2679, + "CCTGAG": 2680, + "CCTGTA": 2681, + "CCTGTT": 2682, + "CCTGTC": 2683, + "CCTGTG": 2684, + "CCTGCA": 2685, + "CCTGCT": 2686, + "CCTGCC": 2687, + "CCTGCG": 2688, + "CCTGGA": 2689, + "CCTGGT": 2690, + "CCTGGC": 2691, + "CCTGGG": 2692, + "CCCAAA": 2693, + "CCCAAT": 2694, + "CCCAAC": 2695, + "CCCAAG": 2696, + "CCCATA": 2697, + "CCCATT": 2698, + "CCCATC": 2699, + "CCCATG": 2700, + "CCCACA": 2701, + "CCCACT": 2702, + "CCCACC": 2703, + "CCCACG": 2704, + "CCCAGA": 2705, + "CCCAGT": 2706, + "CCCAGC": 2707, + "CCCAGG": 2708, + "CCCTAA": 2709, + "CCCTAT": 2710, + "CCCTAC": 2711, + "CCCTAG": 2712, + "CCCTTA": 2713, + "CCCTTT": 2714, + "CCCTTC": 2715, + "CCCTTG": 2716, + "CCCTCA": 2717, + "CCCTCT": 2718, + "CCCTCC": 2719, + "CCCTCG": 2720, + "CCCTGA": 2721, + "CCCTGT": 2722, + "CCCTGC": 2723, + "CCCTGG": 2724, + "CCCCAA": 2725, + "CCCCAT": 2726, + "CCCCAC": 2727, + "CCCCAG": 2728, + "CCCCTA": 2729, + "CCCCTT": 2730, + "CCCCTC": 2731, + "CCCCTG": 2732, + "CCCCCA": 2733, + "CCCCCT": 2734, + "CCCCCC": 2735, + "CCCCCG": 2736, + "CCCCGA": 2737, + "CCCCGT": 2738, + "CCCCGC": 2739, + "CCCCGG": 2740, + "CCCGAA": 2741, + "CCCGAT": 2742, + "CCCGAC": 2743, + "CCCGAG": 2744, + "CCCGTA": 2745, + "CCCGTT": 2746, + "CCCGTC": 2747, + "CCCGTG": 2748, + "CCCGCA": 2749, + "CCCGCT": 2750, + "CCCGCC": 2751, + "CCCGCG": 2752, + "CCCGGA": 2753, + "CCCGGT": 2754, + "CCCGGC": 2755, + "CCCGGG": 2756, + "CCGAAA": 2757, + "CCGAAT": 2758, + "CCGAAC": 2759, + "CCGAAG": 2760, + "CCGATA": 2761, + "CCGATT": 2762, + "CCGATC": 2763, + "CCGATG": 2764, + "CCGACA": 2765, + "CCGACT": 2766, + "CCGACC": 2767, + "CCGACG": 2768, + "CCGAGA": 2769, + "CCGAGT": 2770, + "CCGAGC": 2771, + "CCGAGG": 2772, + "CCGTAA": 2773, + "CCGTAT": 2774, + "CCGTAC": 2775, + "CCGTAG": 2776, + "CCGTTA": 2777, + "CCGTTT": 2778, + "CCGTTC": 2779, + "CCGTTG": 2780, + "CCGTCA": 2781, + "CCGTCT": 2782, + "CCGTCC": 2783, + "CCGTCG": 2784, + "CCGTGA": 2785, + "CCGTGT": 2786, + "CCGTGC": 2787, + "CCGTGG": 2788, + "CCGCAA": 2789, + "CCGCAT": 2790, + "CCGCAC": 2791, + "CCGCAG": 2792, + "CCGCTA": 2793, + "CCGCTT": 2794, + "CCGCTC": 2795, + "CCGCTG": 2796, + "CCGCCA": 2797, + "CCGCCT": 2798, + "CCGCCC": 2799, + "CCGCCG": 2800, + "CCGCGA": 2801, + "CCGCGT": 2802, + "CCGCGC": 2803, + "CCGCGG": 2804, + "CCGGAA": 2805, + "CCGGAT": 2806, + "CCGGAC": 2807, + "CCGGAG": 2808, + "CCGGTA": 2809, + "CCGGTT": 2810, + "CCGGTC": 2811, + "CCGGTG": 2812, + "CCGGCA": 2813, + "CCGGCT": 2814, + "CCGGCC": 2815, + "CCGGCG": 2816, + "CCGGGA": 2817, + "CCGGGT": 2818, + "CCGGGC": 2819, + "CCGGGG": 2820, + "CGAAAA": 2821, + "CGAAAT": 2822, + "CGAAAC": 2823, + "CGAAAG": 2824, + "CGAATA": 2825, + "CGAATT": 2826, + "CGAATC": 2827, + "CGAATG": 2828, + "CGAACA": 2829, + "CGAACT": 2830, + "CGAACC": 2831, + "CGAACG": 2832, + "CGAAGA": 2833, + "CGAAGT": 2834, + "CGAAGC": 2835, + "CGAAGG": 2836, + "CGATAA": 2837, + "CGATAT": 2838, + "CGATAC": 2839, + "CGATAG": 2840, + "CGATTA": 2841, + "CGATTT": 2842, + "CGATTC": 2843, + "CGATTG": 2844, + "CGATCA": 2845, + "CGATCT": 2846, + "CGATCC": 2847, + "CGATCG": 2848, + "CGATGA": 2849, + "CGATGT": 2850, + "CGATGC": 2851, + "CGATGG": 2852, + "CGACAA": 2853, + "CGACAT": 2854, + "CGACAC": 2855, + "CGACAG": 2856, + "CGACTA": 2857, + "CGACTT": 2858, + "CGACTC": 2859, + "CGACTG": 2860, + "CGACCA": 2861, + "CGACCT": 2862, + "CGACCC": 2863, + "CGACCG": 2864, + "CGACGA": 2865, + "CGACGT": 2866, + "CGACGC": 2867, + "CGACGG": 2868, + "CGAGAA": 2869, + "CGAGAT": 2870, + "CGAGAC": 2871, + "CGAGAG": 2872, + "CGAGTA": 2873, + "CGAGTT": 2874, + "CGAGTC": 2875, + "CGAGTG": 2876, + "CGAGCA": 2877, + "CGAGCT": 2878, + "CGAGCC": 2879, + "CGAGCG": 2880, + "CGAGGA": 2881, + "CGAGGT": 2882, + "CGAGGC": 2883, + "CGAGGG": 2884, + "CGTAAA": 2885, + "CGTAAT": 2886, + "CGTAAC": 2887, + "CGTAAG": 2888, + "CGTATA": 2889, + "CGTATT": 2890, + "CGTATC": 2891, + "CGTATG": 2892, + "CGTACA": 2893, + "CGTACT": 2894, + "CGTACC": 2895, + "CGTACG": 2896, + "CGTAGA": 2897, + "CGTAGT": 2898, + "CGTAGC": 2899, + "CGTAGG": 2900, + "CGTTAA": 2901, + "CGTTAT": 2902, + "CGTTAC": 2903, + "CGTTAG": 2904, + "CGTTTA": 2905, + "CGTTTT": 2906, + "CGTTTC": 2907, + "CGTTTG": 2908, + "CGTTCA": 2909, + "CGTTCT": 2910, + "CGTTCC": 2911, + "CGTTCG": 2912, + "CGTTGA": 2913, + "CGTTGT": 2914, + "CGTTGC": 2915, + "CGTTGG": 2916, + "CGTCAA": 2917, + "CGTCAT": 2918, + "CGTCAC": 2919, + "CGTCAG": 2920, + "CGTCTA": 2921, + "CGTCTT": 2922, + "CGTCTC": 2923, + "CGTCTG": 2924, + "CGTCCA": 2925, + "CGTCCT": 2926, + "CGTCCC": 2927, + "CGTCCG": 2928, + "CGTCGA": 2929, + "CGTCGT": 2930, + "CGTCGC": 2931, + "CGTCGG": 2932, + "CGTGAA": 2933, + "CGTGAT": 2934, + "CGTGAC": 2935, + "CGTGAG": 2936, + "CGTGTA": 2937, + "CGTGTT": 2938, + "CGTGTC": 2939, + "CGTGTG": 2940, + "CGTGCA": 2941, + "CGTGCT": 2942, + "CGTGCC": 2943, + "CGTGCG": 2944, + "CGTGGA": 2945, + "CGTGGT": 2946, + "CGTGGC": 2947, + "CGTGGG": 2948, + "CGCAAA": 2949, + "CGCAAT": 2950, + "CGCAAC": 2951, + "CGCAAG": 2952, + "CGCATA": 2953, + "CGCATT": 2954, + "CGCATC": 2955, + "CGCATG": 2956, + "CGCACA": 2957, + "CGCACT": 2958, + "CGCACC": 2959, + "CGCACG": 2960, + "CGCAGA": 2961, + "CGCAGT": 2962, + "CGCAGC": 2963, + "CGCAGG": 2964, + "CGCTAA": 2965, + "CGCTAT": 2966, + "CGCTAC": 2967, + "CGCTAG": 2968, + "CGCTTA": 2969, + "CGCTTT": 2970, + "CGCTTC": 2971, + "CGCTTG": 2972, + "CGCTCA": 2973, + "CGCTCT": 2974, + "CGCTCC": 2975, + "CGCTCG": 2976, + "CGCTGA": 2977, + "CGCTGT": 2978, + "CGCTGC": 2979, + "CGCTGG": 2980, + "CGCCAA": 2981, + "CGCCAT": 2982, + "CGCCAC": 2983, + "CGCCAG": 2984, + "CGCCTA": 2985, + "CGCCTT": 2986, + "CGCCTC": 2987, + "CGCCTG": 2988, + "CGCCCA": 2989, + "CGCCCT": 2990, + "CGCCCC": 2991, + "CGCCCG": 2992, + "CGCCGA": 2993, + "CGCCGT": 2994, + "CGCCGC": 2995, + "CGCCGG": 2996, + "CGCGAA": 2997, + "CGCGAT": 2998, + "CGCGAC": 2999, + "CGCGAG": 3000, + "CGCGTA": 3001, + "CGCGTT": 3002, + "CGCGTC": 3003, + "CGCGTG": 3004, + "CGCGCA": 3005, + "CGCGCT": 3006, + "CGCGCC": 3007, + "CGCGCG": 3008, + "CGCGGA": 3009, + "CGCGGT": 3010, + "CGCGGC": 3011, + "CGCGGG": 3012, + "CGGAAA": 3013, + "CGGAAT": 3014, + "CGGAAC": 3015, + "CGGAAG": 3016, + "CGGATA": 3017, + "CGGATT": 3018, + "CGGATC": 3019, + "CGGATG": 3020, + "CGGACA": 3021, + "CGGACT": 3022, + "CGGACC": 3023, + "CGGACG": 3024, + "CGGAGA": 3025, + "CGGAGT": 3026, + "CGGAGC": 3027, + "CGGAGG": 3028, + "CGGTAA": 3029, + "CGGTAT": 3030, + "CGGTAC": 3031, + "CGGTAG": 3032, + "CGGTTA": 3033, + "CGGTTT": 3034, + "CGGTTC": 3035, + "CGGTTG": 3036, + "CGGTCA": 3037, + "CGGTCT": 3038, + "CGGTCC": 3039, + "CGGTCG": 3040, + "CGGTGA": 3041, + "CGGTGT": 3042, + "CGGTGC": 3043, + "CGGTGG": 3044, + "CGGCAA": 3045, + "CGGCAT": 3046, + "CGGCAC": 3047, + "CGGCAG": 3048, + "CGGCTA": 3049, + "CGGCTT": 3050, + "CGGCTC": 3051, + "CGGCTG": 3052, + "CGGCCA": 3053, + "CGGCCT": 3054, + "CGGCCC": 3055, + "CGGCCG": 3056, + "CGGCGA": 3057, + "CGGCGT": 3058, + "CGGCGC": 3059, + "CGGCGG": 3060, + "CGGGAA": 3061, + "CGGGAT": 3062, + "CGGGAC": 3063, + "CGGGAG": 3064, + "CGGGTA": 3065, + "CGGGTT": 3066, + "CGGGTC": 3067, + "CGGGTG": 3068, + "CGGGCA": 3069, + "CGGGCT": 3070, + "CGGGCC": 3071, + "CGGGCG": 3072, + "CGGGGA": 3073, + "CGGGGT": 3074, + "CGGGGC": 3075, + "CGGGGG": 3076, + "GAAAAA": 3077, + "GAAAAT": 3078, + "GAAAAC": 3079, + "GAAAAG": 3080, + "GAAATA": 3081, + "GAAATT": 3082, + "GAAATC": 3083, + "GAAATG": 3084, + "GAAACA": 3085, + "GAAACT": 3086, + "GAAACC": 3087, + "GAAACG": 3088, + "GAAAGA": 3089, + "GAAAGT": 3090, + "GAAAGC": 3091, + "GAAAGG": 3092, + "GAATAA": 3093, + "GAATAT": 3094, + "GAATAC": 3095, + "GAATAG": 3096, + "GAATTA": 3097, + "GAATTT": 3098, + "GAATTC": 3099, + "GAATTG": 3100, + "GAATCA": 3101, + "GAATCT": 3102, + "GAATCC": 3103, + "GAATCG": 3104, + "GAATGA": 3105, + "GAATGT": 3106, + "GAATGC": 3107, + "GAATGG": 3108, + "GAACAA": 3109, + "GAACAT": 3110, + "GAACAC": 3111, + "GAACAG": 3112, + "GAACTA": 3113, + "GAACTT": 3114, + "GAACTC": 3115, + "GAACTG": 3116, + "GAACCA": 3117, + "GAACCT": 3118, + "GAACCC": 3119, + "GAACCG": 3120, + "GAACGA": 3121, + "GAACGT": 3122, + "GAACGC": 3123, + "GAACGG": 3124, + "GAAGAA": 3125, + "GAAGAT": 3126, + "GAAGAC": 3127, + "GAAGAG": 3128, + "GAAGTA": 3129, + "GAAGTT": 3130, + "GAAGTC": 3131, + "GAAGTG": 3132, + "GAAGCA": 3133, + "GAAGCT": 3134, + "GAAGCC": 3135, + "GAAGCG": 3136, + "GAAGGA": 3137, + "GAAGGT": 3138, + "GAAGGC": 3139, + "GAAGGG": 3140, + "GATAAA": 3141, + "GATAAT": 3142, + "GATAAC": 3143, + "GATAAG": 3144, + "GATATA": 3145, + "GATATT": 3146, + "GATATC": 3147, + "GATATG": 3148, + "GATACA": 3149, + "GATACT": 3150, + "GATACC": 3151, + "GATACG": 3152, + "GATAGA": 3153, + "GATAGT": 3154, + "GATAGC": 3155, + "GATAGG": 3156, + "GATTAA": 3157, + "GATTAT": 3158, + "GATTAC": 3159, + "GATTAG": 3160, + "GATTTA": 3161, + "GATTTT": 3162, + "GATTTC": 3163, + "GATTTG": 3164, + "GATTCA": 3165, + "GATTCT": 3166, + "GATTCC": 3167, + "GATTCG": 3168, + "GATTGA": 3169, + "GATTGT": 3170, + "GATTGC": 3171, + "GATTGG": 3172, + "GATCAA": 3173, + "GATCAT": 3174, + "GATCAC": 3175, + "GATCAG": 3176, + "GATCTA": 3177, + "GATCTT": 3178, + "GATCTC": 3179, + "GATCTG": 3180, + "GATCCA": 3181, + "GATCCT": 3182, + "GATCCC": 3183, + "GATCCG": 3184, + "GATCGA": 3185, + "GATCGT": 3186, + "GATCGC": 3187, + "GATCGG": 3188, + "GATGAA": 3189, + "GATGAT": 3190, + "GATGAC": 3191, + "GATGAG": 3192, + "GATGTA": 3193, + "GATGTT": 3194, + "GATGTC": 3195, + "GATGTG": 3196, + "GATGCA": 3197, + "GATGCT": 3198, + "GATGCC": 3199, + "GATGCG": 3200, + "GATGGA": 3201, + "GATGGT": 3202, + "GATGGC": 3203, + "GATGGG": 3204, + "GACAAA": 3205, + "GACAAT": 3206, + "GACAAC": 3207, + "GACAAG": 3208, + "GACATA": 3209, + "GACATT": 3210, + "GACATC": 3211, + "GACATG": 3212, + "GACACA": 3213, + "GACACT": 3214, + "GACACC": 3215, + "GACACG": 3216, + "GACAGA": 3217, + "GACAGT": 3218, + "GACAGC": 3219, + "GACAGG": 3220, + "GACTAA": 3221, + "GACTAT": 3222, + "GACTAC": 3223, + "GACTAG": 3224, + "GACTTA": 3225, + "GACTTT": 3226, + "GACTTC": 3227, + "GACTTG": 3228, + "GACTCA": 3229, + "GACTCT": 3230, + "GACTCC": 3231, + "GACTCG": 3232, + "GACTGA": 3233, + "GACTGT": 3234, + "GACTGC": 3235, + "GACTGG": 3236, + "GACCAA": 3237, + "GACCAT": 3238, + "GACCAC": 3239, + "GACCAG": 3240, + "GACCTA": 3241, + "GACCTT": 3242, + "GACCTC": 3243, + "GACCTG": 3244, + "GACCCA": 3245, + "GACCCT": 3246, + "GACCCC": 3247, + "GACCCG": 3248, + "GACCGA": 3249, + "GACCGT": 3250, + "GACCGC": 3251, + "GACCGG": 3252, + "GACGAA": 3253, + "GACGAT": 3254, + "GACGAC": 3255, + "GACGAG": 3256, + "GACGTA": 3257, + "GACGTT": 3258, + "GACGTC": 3259, + "GACGTG": 3260, + "GACGCA": 3261, + "GACGCT": 3262, + "GACGCC": 3263, + "GACGCG": 3264, + "GACGGA": 3265, + "GACGGT": 3266, + "GACGGC": 3267, + "GACGGG": 3268, + "GAGAAA": 3269, + "GAGAAT": 3270, + "GAGAAC": 3271, + "GAGAAG": 3272, + "GAGATA": 3273, + "GAGATT": 3274, + "GAGATC": 3275, + "GAGATG": 3276, + "GAGACA": 3277, + "GAGACT": 3278, + "GAGACC": 3279, + "GAGACG": 3280, + "GAGAGA": 3281, + "GAGAGT": 3282, + "GAGAGC": 3283, + "GAGAGG": 3284, + "GAGTAA": 3285, + "GAGTAT": 3286, + "GAGTAC": 3287, + "GAGTAG": 3288, + "GAGTTA": 3289, + "GAGTTT": 3290, + "GAGTTC": 3291, + "GAGTTG": 3292, + "GAGTCA": 3293, + "GAGTCT": 3294, + "GAGTCC": 3295, + "GAGTCG": 3296, + "GAGTGA": 3297, + "GAGTGT": 3298, + "GAGTGC": 3299, + "GAGTGG": 3300, + "GAGCAA": 3301, + "GAGCAT": 3302, + "GAGCAC": 3303, + "GAGCAG": 3304, + "GAGCTA": 3305, + "GAGCTT": 3306, + "GAGCTC": 3307, + "GAGCTG": 3308, + "GAGCCA": 3309, + "GAGCCT": 3310, + "GAGCCC": 3311, + "GAGCCG": 3312, + "GAGCGA": 3313, + "GAGCGT": 3314, + "GAGCGC": 3315, + "GAGCGG": 3316, + "GAGGAA": 3317, + "GAGGAT": 3318, + "GAGGAC": 3319, + "GAGGAG": 3320, + "GAGGTA": 3321, + "GAGGTT": 3322, + "GAGGTC": 3323, + "GAGGTG": 3324, + "GAGGCA": 3325, + "GAGGCT": 3326, + "GAGGCC": 3327, + "GAGGCG": 3328, + "GAGGGA": 3329, + "GAGGGT": 3330, + "GAGGGC": 3331, + "GAGGGG": 3332, + "GTAAAA": 3333, + "GTAAAT": 3334, + "GTAAAC": 3335, + "GTAAAG": 3336, + "GTAATA": 3337, + "GTAATT": 3338, + "GTAATC": 3339, + "GTAATG": 3340, + "GTAACA": 3341, + "GTAACT": 3342, + "GTAACC": 3343, + "GTAACG": 3344, + "GTAAGA": 3345, + "GTAAGT": 3346, + "GTAAGC": 3347, + "GTAAGG": 3348, + "GTATAA": 3349, + "GTATAT": 3350, + "GTATAC": 3351, + "GTATAG": 3352, + "GTATTA": 3353, + "GTATTT": 3354, + "GTATTC": 3355, + "GTATTG": 3356, + "GTATCA": 3357, + "GTATCT": 3358, + "GTATCC": 3359, + "GTATCG": 3360, + "GTATGA": 3361, + "GTATGT": 3362, + "GTATGC": 3363, + "GTATGG": 3364, + "GTACAA": 3365, + "GTACAT": 3366, + "GTACAC": 3367, + "GTACAG": 3368, + "GTACTA": 3369, + "GTACTT": 3370, + "GTACTC": 3371, + "GTACTG": 3372, + "GTACCA": 3373, + "GTACCT": 3374, + "GTACCC": 3375, + "GTACCG": 3376, + "GTACGA": 3377, + "GTACGT": 3378, + "GTACGC": 3379, + "GTACGG": 3380, + "GTAGAA": 3381, + "GTAGAT": 3382, + "GTAGAC": 3383, + "GTAGAG": 3384, + "GTAGTA": 3385, + "GTAGTT": 3386, + "GTAGTC": 3387, + "GTAGTG": 3388, + "GTAGCA": 3389, + "GTAGCT": 3390, + "GTAGCC": 3391, + "GTAGCG": 3392, + "GTAGGA": 3393, + "GTAGGT": 3394, + "GTAGGC": 3395, + "GTAGGG": 3396, + "GTTAAA": 3397, + "GTTAAT": 3398, + "GTTAAC": 3399, + "GTTAAG": 3400, + "GTTATA": 3401, + "GTTATT": 3402, + "GTTATC": 3403, + "GTTATG": 3404, + "GTTACA": 3405, + "GTTACT": 3406, + "GTTACC": 3407, + "GTTACG": 3408, + "GTTAGA": 3409, + "GTTAGT": 3410, + "GTTAGC": 3411, + "GTTAGG": 3412, + "GTTTAA": 3413, + "GTTTAT": 3414, + "GTTTAC": 3415, + "GTTTAG": 3416, + "GTTTTA": 3417, + "GTTTTT": 3418, + "GTTTTC": 3419, + "GTTTTG": 3420, + "GTTTCA": 3421, + "GTTTCT": 3422, + "GTTTCC": 3423, + "GTTTCG": 3424, + "GTTTGA": 3425, + "GTTTGT": 3426, + "GTTTGC": 3427, + "GTTTGG": 3428, + "GTTCAA": 3429, + "GTTCAT": 3430, + "GTTCAC": 3431, + "GTTCAG": 3432, + "GTTCTA": 3433, + "GTTCTT": 3434, + "GTTCTC": 3435, + "GTTCTG": 3436, + "GTTCCA": 3437, + "GTTCCT": 3438, + "GTTCCC": 3439, + "GTTCCG": 3440, + "GTTCGA": 3441, + "GTTCGT": 3442, + "GTTCGC": 3443, + "GTTCGG": 3444, + "GTTGAA": 3445, + "GTTGAT": 3446, + "GTTGAC": 3447, + "GTTGAG": 3448, + "GTTGTA": 3449, + "GTTGTT": 3450, + "GTTGTC": 3451, + "GTTGTG": 3452, + "GTTGCA": 3453, + "GTTGCT": 3454, + "GTTGCC": 3455, + "GTTGCG": 3456, + "GTTGGA": 3457, + "GTTGGT": 3458, + "GTTGGC": 3459, + "GTTGGG": 3460, + "GTCAAA": 3461, + "GTCAAT": 3462, + "GTCAAC": 3463, + "GTCAAG": 3464, + "GTCATA": 3465, + "GTCATT": 3466, + "GTCATC": 3467, + "GTCATG": 3468, + "GTCACA": 3469, + "GTCACT": 3470, + "GTCACC": 3471, + "GTCACG": 3472, + "GTCAGA": 3473, + "GTCAGT": 3474, + "GTCAGC": 3475, + "GTCAGG": 3476, + "GTCTAA": 3477, + "GTCTAT": 3478, + "GTCTAC": 3479, + "GTCTAG": 3480, + "GTCTTA": 3481, + "GTCTTT": 3482, + "GTCTTC": 3483, + "GTCTTG": 3484, + "GTCTCA": 3485, + "GTCTCT": 3486, + "GTCTCC": 3487, + "GTCTCG": 3488, + "GTCTGA": 3489, + "GTCTGT": 3490, + "GTCTGC": 3491, + "GTCTGG": 3492, + "GTCCAA": 3493, + "GTCCAT": 3494, + "GTCCAC": 3495, + "GTCCAG": 3496, + "GTCCTA": 3497, + "GTCCTT": 3498, + "GTCCTC": 3499, + "GTCCTG": 3500, + "GTCCCA": 3501, + "GTCCCT": 3502, + "GTCCCC": 3503, + "GTCCCG": 3504, + "GTCCGA": 3505, + "GTCCGT": 3506, + "GTCCGC": 3507, + "GTCCGG": 3508, + "GTCGAA": 3509, + "GTCGAT": 3510, + "GTCGAC": 3511, + "GTCGAG": 3512, + "GTCGTA": 3513, + "GTCGTT": 3514, + "GTCGTC": 3515, + "GTCGTG": 3516, + "GTCGCA": 3517, + "GTCGCT": 3518, + "GTCGCC": 3519, + "GTCGCG": 3520, + "GTCGGA": 3521, + "GTCGGT": 3522, + "GTCGGC": 3523, + "GTCGGG": 3524, + "GTGAAA": 3525, + "GTGAAT": 3526, + "GTGAAC": 3527, + "GTGAAG": 3528, + "GTGATA": 3529, + "GTGATT": 3530, + "GTGATC": 3531, + "GTGATG": 3532, + "GTGACA": 3533, + "GTGACT": 3534, + "GTGACC": 3535, + "GTGACG": 3536, + "GTGAGA": 3537, + "GTGAGT": 3538, + "GTGAGC": 3539, + "GTGAGG": 3540, + "GTGTAA": 3541, + "GTGTAT": 3542, + "GTGTAC": 3543, + "GTGTAG": 3544, + "GTGTTA": 3545, + "GTGTTT": 3546, + "GTGTTC": 3547, + "GTGTTG": 3548, + "GTGTCA": 3549, + "GTGTCT": 3550, + "GTGTCC": 3551, + "GTGTCG": 3552, + "GTGTGA": 3553, + "GTGTGT": 3554, + "GTGTGC": 3555, + "GTGTGG": 3556, + "GTGCAA": 3557, + "GTGCAT": 3558, + "GTGCAC": 3559, + "GTGCAG": 3560, + "GTGCTA": 3561, + "GTGCTT": 3562, + "GTGCTC": 3563, + "GTGCTG": 3564, + "GTGCCA": 3565, + "GTGCCT": 3566, + "GTGCCC": 3567, + "GTGCCG": 3568, + "GTGCGA": 3569, + "GTGCGT": 3570, + "GTGCGC": 3571, + "GTGCGG": 3572, + "GTGGAA": 3573, + "GTGGAT": 3574, + "GTGGAC": 3575, + "GTGGAG": 3576, + "GTGGTA": 3577, + "GTGGTT": 3578, + "GTGGTC": 3579, + "GTGGTG": 3580, + "GTGGCA": 3581, + "GTGGCT": 3582, + "GTGGCC": 3583, + "GTGGCG": 3584, + "GTGGGA": 3585, + "GTGGGT": 3586, + "GTGGGC": 3587, + "GTGGGG": 3588, + "GCAAAA": 3589, + "GCAAAT": 3590, + "GCAAAC": 3591, + "GCAAAG": 3592, + "GCAATA": 3593, + "GCAATT": 3594, + "GCAATC": 3595, + "GCAATG": 3596, + "GCAACA": 3597, + "GCAACT": 3598, + "GCAACC": 3599, + "GCAACG": 3600, + "GCAAGA": 3601, + "GCAAGT": 3602, + "GCAAGC": 3603, + "GCAAGG": 3604, + "GCATAA": 3605, + "GCATAT": 3606, + "GCATAC": 3607, + "GCATAG": 3608, + "GCATTA": 3609, + "GCATTT": 3610, + "GCATTC": 3611, + "GCATTG": 3612, + "GCATCA": 3613, + "GCATCT": 3614, + "GCATCC": 3615, + "GCATCG": 3616, + "GCATGA": 3617, + "GCATGT": 3618, + "GCATGC": 3619, + "GCATGG": 3620, + "GCACAA": 3621, + "GCACAT": 3622, + "GCACAC": 3623, + "GCACAG": 3624, + "GCACTA": 3625, + "GCACTT": 3626, + "GCACTC": 3627, + "GCACTG": 3628, + "GCACCA": 3629, + "GCACCT": 3630, + "GCACCC": 3631, + "GCACCG": 3632, + "GCACGA": 3633, + "GCACGT": 3634, + "GCACGC": 3635, + "GCACGG": 3636, + "GCAGAA": 3637, + "GCAGAT": 3638, + "GCAGAC": 3639, + "GCAGAG": 3640, + "GCAGTA": 3641, + "GCAGTT": 3642, + "GCAGTC": 3643, + "GCAGTG": 3644, + "GCAGCA": 3645, + "GCAGCT": 3646, + "GCAGCC": 3647, + "GCAGCG": 3648, + "GCAGGA": 3649, + "GCAGGT": 3650, + "GCAGGC": 3651, + "GCAGGG": 3652, + "GCTAAA": 3653, + "GCTAAT": 3654, + "GCTAAC": 3655, + "GCTAAG": 3656, + "GCTATA": 3657, + "GCTATT": 3658, + "GCTATC": 3659, + "GCTATG": 3660, + "GCTACA": 3661, + "GCTACT": 3662, + "GCTACC": 3663, + "GCTACG": 3664, + "GCTAGA": 3665, + "GCTAGT": 3666, + "GCTAGC": 3667, + "GCTAGG": 3668, + "GCTTAA": 3669, + "GCTTAT": 3670, + "GCTTAC": 3671, + "GCTTAG": 3672, + "GCTTTA": 3673, + "GCTTTT": 3674, + "GCTTTC": 3675, + "GCTTTG": 3676, + "GCTTCA": 3677, + "GCTTCT": 3678, + "GCTTCC": 3679, + "GCTTCG": 3680, + "GCTTGA": 3681, + "GCTTGT": 3682, + "GCTTGC": 3683, + "GCTTGG": 3684, + "GCTCAA": 3685, + "GCTCAT": 3686, + "GCTCAC": 3687, + "GCTCAG": 3688, + "GCTCTA": 3689, + "GCTCTT": 3690, + "GCTCTC": 3691, + "GCTCTG": 3692, + "GCTCCA": 3693, + "GCTCCT": 3694, + "GCTCCC": 3695, + "GCTCCG": 3696, + "GCTCGA": 3697, + "GCTCGT": 3698, + "GCTCGC": 3699, + "GCTCGG": 3700, + "GCTGAA": 3701, + "GCTGAT": 3702, + "GCTGAC": 3703, + "GCTGAG": 3704, + "GCTGTA": 3705, + "GCTGTT": 3706, + "GCTGTC": 3707, + "GCTGTG": 3708, + "GCTGCA": 3709, + "GCTGCT": 3710, + "GCTGCC": 3711, + "GCTGCG": 3712, + "GCTGGA": 3713, + "GCTGGT": 3714, + "GCTGGC": 3715, + "GCTGGG": 3716, + "GCCAAA": 3717, + "GCCAAT": 3718, + "GCCAAC": 3719, + "GCCAAG": 3720, + "GCCATA": 3721, + "GCCATT": 3722, + "GCCATC": 3723, + "GCCATG": 3724, + "GCCACA": 3725, + "GCCACT": 3726, + "GCCACC": 3727, + "GCCACG": 3728, + "GCCAGA": 3729, + "GCCAGT": 3730, + "GCCAGC": 3731, + "GCCAGG": 3732, + "GCCTAA": 3733, + "GCCTAT": 3734, + "GCCTAC": 3735, + "GCCTAG": 3736, + "GCCTTA": 3737, + "GCCTTT": 3738, + "GCCTTC": 3739, + "GCCTTG": 3740, + "GCCTCA": 3741, + "GCCTCT": 3742, + "GCCTCC": 3743, + "GCCTCG": 3744, + "GCCTGA": 3745, + "GCCTGT": 3746, + "GCCTGC": 3747, + "GCCTGG": 3748, + "GCCCAA": 3749, + "GCCCAT": 3750, + "GCCCAC": 3751, + "GCCCAG": 3752, + "GCCCTA": 3753, + "GCCCTT": 3754, + "GCCCTC": 3755, + "GCCCTG": 3756, + "GCCCCA": 3757, + "GCCCCT": 3758, + "GCCCCC": 3759, + "GCCCCG": 3760, + "GCCCGA": 3761, + "GCCCGT": 3762, + "GCCCGC": 3763, + "GCCCGG": 3764, + "GCCGAA": 3765, + "GCCGAT": 3766, + "GCCGAC": 3767, + "GCCGAG": 3768, + "GCCGTA": 3769, + "GCCGTT": 3770, + "GCCGTC": 3771, + "GCCGTG": 3772, + "GCCGCA": 3773, + "GCCGCT": 3774, + "GCCGCC": 3775, + "GCCGCG": 3776, + "GCCGGA": 3777, + "GCCGGT": 3778, + "GCCGGC": 3779, + "GCCGGG": 3780, + "GCGAAA": 3781, + "GCGAAT": 3782, + "GCGAAC": 3783, + "GCGAAG": 3784, + "GCGATA": 3785, + "GCGATT": 3786, + "GCGATC": 3787, + "GCGATG": 3788, + "GCGACA": 3789, + "GCGACT": 3790, + "GCGACC": 3791, + "GCGACG": 3792, + "GCGAGA": 3793, + "GCGAGT": 3794, + "GCGAGC": 3795, + "GCGAGG": 3796, + "GCGTAA": 3797, + "GCGTAT": 3798, + "GCGTAC": 3799, + "GCGTAG": 3800, + "GCGTTA": 3801, + "GCGTTT": 3802, + "GCGTTC": 3803, + "GCGTTG": 3804, + "GCGTCA": 3805, + "GCGTCT": 3806, + "GCGTCC": 3807, + "GCGTCG": 3808, + "GCGTGA": 3809, + "GCGTGT": 3810, + "GCGTGC": 3811, + "GCGTGG": 3812, + "GCGCAA": 3813, + "GCGCAT": 3814, + "GCGCAC": 3815, + "GCGCAG": 3816, + "GCGCTA": 3817, + "GCGCTT": 3818, + "GCGCTC": 3819, + "GCGCTG": 3820, + "GCGCCA": 3821, + "GCGCCT": 3822, + "GCGCCC": 3823, + "GCGCCG": 3824, + "GCGCGA": 3825, + "GCGCGT": 3826, + "GCGCGC": 3827, + "GCGCGG": 3828, + "GCGGAA": 3829, + "GCGGAT": 3830, + "GCGGAC": 3831, + "GCGGAG": 3832, + "GCGGTA": 3833, + "GCGGTT": 3834, + "GCGGTC": 3835, + "GCGGTG": 3836, + "GCGGCA": 3837, + "GCGGCT": 3838, + "GCGGCC": 3839, + "GCGGCG": 3840, + "GCGGGA": 3841, + "GCGGGT": 3842, + "GCGGGC": 3843, + "GCGGGG": 3844, + "GGAAAA": 3845, + "GGAAAT": 3846, + "GGAAAC": 3847, + "GGAAAG": 3848, + "GGAATA": 3849, + "GGAATT": 3850, + "GGAATC": 3851, + "GGAATG": 3852, + "GGAACA": 3853, + "GGAACT": 3854, + "GGAACC": 3855, + "GGAACG": 3856, + "GGAAGA": 3857, + "GGAAGT": 3858, + "GGAAGC": 3859, + "GGAAGG": 3860, + "GGATAA": 3861, + "GGATAT": 3862, + "GGATAC": 3863, + "GGATAG": 3864, + "GGATTA": 3865, + "GGATTT": 3866, + "GGATTC": 3867, + "GGATTG": 3868, + "GGATCA": 3869, + "GGATCT": 3870, + "GGATCC": 3871, + "GGATCG": 3872, + "GGATGA": 3873, + "GGATGT": 3874, + "GGATGC": 3875, + "GGATGG": 3876, + "GGACAA": 3877, + "GGACAT": 3878, + "GGACAC": 3879, + "GGACAG": 3880, + "GGACTA": 3881, + "GGACTT": 3882, + "GGACTC": 3883, + "GGACTG": 3884, + "GGACCA": 3885, + "GGACCT": 3886, + "GGACCC": 3887, + "GGACCG": 3888, + "GGACGA": 3889, + "GGACGT": 3890, + "GGACGC": 3891, + "GGACGG": 3892, + "GGAGAA": 3893, + "GGAGAT": 3894, + "GGAGAC": 3895, + "GGAGAG": 3896, + "GGAGTA": 3897, + "GGAGTT": 3898, + "GGAGTC": 3899, + "GGAGTG": 3900, + "GGAGCA": 3901, + "GGAGCT": 3902, + "GGAGCC": 3903, + "GGAGCG": 3904, + "GGAGGA": 3905, + "GGAGGT": 3906, + "GGAGGC": 3907, + "GGAGGG": 3908, + "GGTAAA": 3909, + "GGTAAT": 3910, + "GGTAAC": 3911, + "GGTAAG": 3912, + "GGTATA": 3913, + "GGTATT": 3914, + "GGTATC": 3915, + "GGTATG": 3916, + "GGTACA": 3917, + "GGTACT": 3918, + "GGTACC": 3919, + "GGTACG": 3920, + "GGTAGA": 3921, + "GGTAGT": 3922, + "GGTAGC": 3923, + "GGTAGG": 3924, + "GGTTAA": 3925, + "GGTTAT": 3926, + "GGTTAC": 3927, + "GGTTAG": 3928, + "GGTTTA": 3929, + "GGTTTT": 3930, + "GGTTTC": 3931, + "GGTTTG": 3932, + "GGTTCA": 3933, + "GGTTCT": 3934, + "GGTTCC": 3935, + "GGTTCG": 3936, + "GGTTGA": 3937, + "GGTTGT": 3938, + "GGTTGC": 3939, + "GGTTGG": 3940, + "GGTCAA": 3941, + "GGTCAT": 3942, + "GGTCAC": 3943, + "GGTCAG": 3944, + "GGTCTA": 3945, + "GGTCTT": 3946, + "GGTCTC": 3947, + "GGTCTG": 3948, + "GGTCCA": 3949, + "GGTCCT": 3950, + "GGTCCC": 3951, + "GGTCCG": 3952, + "GGTCGA": 3953, + "GGTCGT": 3954, + "GGTCGC": 3955, + "GGTCGG": 3956, + "GGTGAA": 3957, + "GGTGAT": 3958, + "GGTGAC": 3959, + "GGTGAG": 3960, + "GGTGTA": 3961, + "GGTGTT": 3962, + "GGTGTC": 3963, + "GGTGTG": 3964, + "GGTGCA": 3965, + "GGTGCT": 3966, + "GGTGCC": 3967, + "GGTGCG": 3968, + "GGTGGA": 3969, + "GGTGGT": 3970, + "GGTGGC": 3971, + "GGTGGG": 3972, + "GGCAAA": 3973, + "GGCAAT": 3974, + "GGCAAC": 3975, + "GGCAAG": 3976, + "GGCATA": 3977, + "GGCATT": 3978, + "GGCATC": 3979, + "GGCATG": 3980, + "GGCACA": 3981, + "GGCACT": 3982, + "GGCACC": 3983, + "GGCACG": 3984, + "GGCAGA": 3985, + "GGCAGT": 3986, + "GGCAGC": 3987, + "GGCAGG": 3988, + "GGCTAA": 3989, + "GGCTAT": 3990, + "GGCTAC": 3991, + "GGCTAG": 3992, + "GGCTTA": 3993, + "GGCTTT": 3994, + "GGCTTC": 3995, + "GGCTTG": 3996, + "GGCTCA": 3997, + "GGCTCT": 3998, + "GGCTCC": 3999, + "GGCTCG": 4000, + "GGCTGA": 4001, + "GGCTGT": 4002, + "GGCTGC": 4003, + "GGCTGG": 4004, + "GGCCAA": 4005, + "GGCCAT": 4006, + "GGCCAC": 4007, + "GGCCAG": 4008, + "GGCCTA": 4009, + "GGCCTT": 4010, + "GGCCTC": 4011, + "GGCCTG": 4012, + "GGCCCA": 4013, + "GGCCCT": 4014, + "GGCCCC": 4015, + "GGCCCG": 4016, + "GGCCGA": 4017, + "GGCCGT": 4018, + "GGCCGC": 4019, + "GGCCGG": 4020, + "GGCGAA": 4021, + "GGCGAT": 4022, + "GGCGAC": 4023, + "GGCGAG": 4024, + "GGCGTA": 4025, + "GGCGTT": 4026, + "GGCGTC": 4027, + "GGCGTG": 4028, + "GGCGCA": 4029, + "GGCGCT": 4030, + "GGCGCC": 4031, + "GGCGCG": 4032, + "GGCGGA": 4033, + "GGCGGT": 4034, + "GGCGGC": 4035, + "GGCGGG": 4036, + "GGGAAA": 4037, + "GGGAAT": 4038, + "GGGAAC": 4039, + "GGGAAG": 4040, + "GGGATA": 4041, + "GGGATT": 4042, + "GGGATC": 4043, + "GGGATG": 4044, + "GGGACA": 4045, + "GGGACT": 4046, + "GGGACC": 4047, + "GGGACG": 4048, + "GGGAGA": 4049, + "GGGAGT": 4050, + "GGGAGC": 4051, + "GGGAGG": 4052, + "GGGTAA": 4053, + "GGGTAT": 4054, + "GGGTAC": 4055, + "GGGTAG": 4056, + "GGGTTA": 4057, + "GGGTTT": 4058, + "GGGTTC": 4059, + "GGGTTG": 4060, + "GGGTCA": 4061, + "GGGTCT": 4062, + "GGGTCC": 4063, + "GGGTCG": 4064, + "GGGTGA": 4065, + "GGGTGT": 4066, + "GGGTGC": 4067, + "GGGTGG": 4068, + "GGGCAA": 4069, + "GGGCAT": 4070, + "GGGCAC": 4071, + "GGGCAG": 4072, + "GGGCTA": 4073, + "GGGCTT": 4074, + "GGGCTC": 4075, + "GGGCTG": 4076, + "GGGCCA": 4077, + "GGGCCT": 4078, + "GGGCCC": 4079, + "GGGCCG": 4080, + "GGGCGA": 4081, + "GGGCGT": 4082, + "GGGCGC": 4083, + "GGGCGG": 4084, + "GGGGAA": 4085, + "GGGGAT": 4086, + "GGGGAC": 4087, + "GGGGAG": 4088, + "GGGGTA": 4089, + "GGGGTT": 4090, + "GGGGTC": 4091, + "GGGGTG": 4092, + "GGGGCA": 4093, + "GGGGCT": 4094, + "GGGGCC": 4095, + "GGGGCG": 4096, + "GGGGGA": 4097, + "GGGGGT": 4098, + "GGGGGC": 4099, + "GGGGGG": 4100 + } + } +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/tokenizer_config.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..604e2b1b29af1dcb3920d2828b6a90276e5e3d02 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/tokenizer_config.json @@ -0,0 +1,58 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "[UNK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "[CLS]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "[SEP]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "4": { + "content": "[MASK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "clean_up_tokenization_spaces": true, + "cls_token": "[CLS]", + "do_basic_tokenize": true, + "do_lower_case": false, + "mask_token": "[MASK]", + "max_len": 512, + "model_max_length": 512, + "never_split": null, + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "strip_accents": null, + "tokenize_chinese_chars": true, + "tokenizer_class": "BertTokenizer", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/trainer_state.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..14e1fffedd05380f85346cda3744de36d1cf06c5 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/trainer_state.json @@ -0,0 +1,873 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 12.0, + "eval_steps": 500, + "global_step": 3000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.1, + "grad_norm": 0.2192622423171997, + "learning_rate": 5.319148936170213e-05, + "loss": 0.6949, + "step": 25 + }, + { + "epoch": 0.2, + "grad_norm": 0.22659960389137268, + "learning_rate": 0.00010638297872340425, + "loss": 0.6926, + "step": 50 + }, + { + "epoch": 0.3, + "grad_norm": 0.33277687430381775, + "learning_rate": 0.00015957446808510637, + "loss": 0.693, + "step": 75 + }, + { + "epoch": 0.4, + "grad_norm": 0.5183415412902832, + "learning_rate": 0.0002127659574468085, + "loss": 0.6928, + "step": 100 + }, + { + "epoch": 0.5, + "grad_norm": 0.21227706968784332, + "learning_rate": 0.00026595744680851064, + "loss": 0.6938, + "step": 125 + }, + { + "epoch": 0.6, + "grad_norm": 0.1846485733985901, + "learning_rate": 0.00031914893617021275, + "loss": 0.6942, + "step": 150 + }, + { + "epoch": 0.7, + "grad_norm": 0.27858808636665344, + "learning_rate": 0.0003723404255319149, + "loss": 0.6931, + "step": 175 + }, + { + "epoch": 0.8, + "grad_norm": 0.26012367010116577, + "learning_rate": 0.0003992081821181128, + "loss": 0.6936, + "step": 200 + }, + { + "epoch": 0.9, + "grad_norm": 0.2534612715244293, + "learning_rate": 0.0003975585615308479, + "loss": 0.6939, + "step": 225 + }, + { + "epoch": 1.0, + "grad_norm": 0.42967700958251953, + "learning_rate": 0.00039590894094358297, + "loss": 0.6928, + "step": 250 + }, + { + "epoch": 1.1, + "grad_norm": 0.2613862454891205, + "learning_rate": 0.0003942593203563181, + "loss": 0.6918, + "step": 275 + }, + { + "epoch": 1.2, + "grad_norm": 0.5514339208602905, + "learning_rate": 0.0003926096997690532, + "loss": 0.6918, + "step": 300 + }, + { + "epoch": 1.3, + "grad_norm": 0.33801883459091187, + "learning_rate": 0.00039096007918178817, + "loss": 0.6906, + "step": 325 + }, + { + "epoch": 1.4, + "grad_norm": 0.2953580617904663, + "learning_rate": 0.0003893104585945233, + "loss": 0.6886, + "step": 350 + }, + { + "epoch": 1.5, + "grad_norm": 0.5340980887413025, + "learning_rate": 0.0003876608380072583, + "loss": 0.6877, + "step": 375 + }, + { + "epoch": 1.6, + "grad_norm": 0.3042701482772827, + "learning_rate": 0.00038601121741999343, + "loss": 0.6853, + "step": 400 + }, + { + "epoch": 1.7, + "grad_norm": 0.44497063755989075, + "learning_rate": 0.0003843615968327285, + "loss": 0.6772, + "step": 425 + }, + { + "epoch": 1.8, + "grad_norm": 0.6260074973106384, + "learning_rate": 0.0003827119762454636, + "loss": 0.6671, + "step": 450 + }, + { + "epoch": 1.9, + "grad_norm": 0.9574269652366638, + "learning_rate": 0.00038106235565819863, + "loss": 0.6517, + "step": 475 + }, + { + "epoch": 2.0, + "grad_norm": 0.9055777192115784, + "learning_rate": 0.0003794127350709337, + "loss": 0.6438, + "step": 500 + }, + { + "epoch": 2.1, + "grad_norm": 0.7860134840011597, + "learning_rate": 0.0003777631144836688, + "loss": 0.6379, + "step": 525 + }, + { + "epoch": 2.2, + "grad_norm": 1.5324610471725464, + "learning_rate": 0.00037611349389640383, + "loss": 0.6143, + "step": 550 + }, + { + "epoch": 2.3, + "grad_norm": 1.8535794019699097, + "learning_rate": 0.00037446387330913894, + "loss": 0.607, + "step": 575 + }, + { + "epoch": 2.4, + "grad_norm": 1.8850122690200806, + "learning_rate": 0.000372814252721874, + "loss": 0.5642, + "step": 600 + }, + { + "epoch": 2.5, + "grad_norm": 1.288278579711914, + "learning_rate": 0.00037116463213460903, + "loss": 0.5739, + "step": 625 + }, + { + "epoch": 2.6, + "grad_norm": 1.344173550605774, + "learning_rate": 0.00036951501154734414, + "loss": 0.5718, + "step": 650 + }, + { + "epoch": 2.7, + "grad_norm": 1.430834174156189, + "learning_rate": 0.0003678653909600792, + "loss": 0.543, + "step": 675 + }, + { + "epoch": 2.8, + "grad_norm": 1.5948325395584106, + "learning_rate": 0.0003662157703728143, + "loss": 0.5434, + "step": 700 + }, + { + "epoch": 2.9, + "grad_norm": 2.688519239425659, + "learning_rate": 0.00036456614978554934, + "loss": 0.5842, + "step": 725 + }, + { + "epoch": 3.0, + "grad_norm": 1.6625819206237793, + "learning_rate": 0.00036291652919828444, + "loss": 0.5526, + "step": 750 + }, + { + "epoch": 3.1, + "grad_norm": 1.869725227355957, + "learning_rate": 0.00036126690861101944, + "loss": 0.53, + "step": 775 + }, + { + "epoch": 3.2, + "grad_norm": 2.1625709533691406, + "learning_rate": 0.00035961728802375454, + "loss": 0.5341, + "step": 800 + }, + { + "epoch": 3.3, + "grad_norm": 1.9650202989578247, + "learning_rate": 0.0003579676674364896, + "loss": 0.5297, + "step": 825 + }, + { + "epoch": 3.4, + "grad_norm": 1.3358676433563232, + "learning_rate": 0.0003563180468492247, + "loss": 0.5273, + "step": 850 + }, + { + "epoch": 3.5, + "grad_norm": 2.1927096843719482, + "learning_rate": 0.0003546684262619598, + "loss": 0.5084, + "step": 875 + }, + { + "epoch": 3.6, + "grad_norm": 2.2377188205718994, + "learning_rate": 0.00035301880567469485, + "loss": 0.4972, + "step": 900 + }, + { + "epoch": 3.7, + "grad_norm": 1.8601839542388916, + "learning_rate": 0.0003513691850874299, + "loss": 0.4852, + "step": 925 + }, + { + "epoch": 3.8, + "grad_norm": 1.4995975494384766, + "learning_rate": 0.00034971956450016495, + "loss": 0.5309, + "step": 950 + }, + { + "epoch": 3.9, + "grad_norm": 2.0536930561065674, + "learning_rate": 0.00034806994391290005, + "loss": 0.5249, + "step": 975 + }, + { + "epoch": 4.0, + "grad_norm": 2.084728479385376, + "learning_rate": 0.0003464203233256351, + "loss": 0.4771, + "step": 1000 + }, + { + "epoch": 4.1, + "grad_norm": 1.705510139465332, + "learning_rate": 0.0003447707027383702, + "loss": 0.5181, + "step": 1025 + }, + { + "epoch": 4.2, + "grad_norm": 3.83380389213562, + "learning_rate": 0.00034312108215110525, + "loss": 0.4873, + "step": 1050 + }, + { + "epoch": 4.3, + "grad_norm": 2.0163533687591553, + "learning_rate": 0.0003414714615638403, + "loss": 0.4801, + "step": 1075 + }, + { + "epoch": 4.4, + "grad_norm": 1.9917407035827637, + "learning_rate": 0.0003398218409765754, + "loss": 0.4766, + "step": 1100 + }, + { + "epoch": 4.5, + "grad_norm": 2.111729860305786, + "learning_rate": 0.00033817222038931045, + "loss": 0.4877, + "step": 1125 + }, + { + "epoch": 4.6, + "grad_norm": 1.5487738847732544, + "learning_rate": 0.00033652259980204556, + "loss": 0.507, + "step": 1150 + }, + { + "epoch": 4.7, + "grad_norm": 1.8715145587921143, + "learning_rate": 0.0003348729792147806, + "loss": 0.452, + "step": 1175 + }, + { + "epoch": 4.8, + "grad_norm": 2.2618906497955322, + "learning_rate": 0.0003332233586275157, + "loss": 0.4799, + "step": 1200 + }, + { + "epoch": 4.9, + "grad_norm": 1.657012939453125, + "learning_rate": 0.00033157373804025076, + "loss": 0.4715, + "step": 1225 + }, + { + "epoch": 5.0, + "grad_norm": 2.8894591331481934, + "learning_rate": 0.0003299241174529858, + "loss": 0.4437, + "step": 1250 + }, + { + "epoch": 5.1, + "grad_norm": 1.9179915189743042, + "learning_rate": 0.0003282744968657209, + "loss": 0.4796, + "step": 1275 + }, + { + "epoch": 5.2, + "grad_norm": 2.1816720962524414, + "learning_rate": 0.00032662487627845596, + "loss": 0.4463, + "step": 1300 + }, + { + "epoch": 5.3, + "grad_norm": 2.2761082649230957, + "learning_rate": 0.00032497525569119106, + "loss": 0.4215, + "step": 1325 + }, + { + "epoch": 5.4, + "grad_norm": 1.7190684080123901, + "learning_rate": 0.0003233256351039261, + "loss": 0.4481, + "step": 1350 + }, + { + "epoch": 5.5, + "grad_norm": 1.4793736934661865, + "learning_rate": 0.00032167601451666116, + "loss": 0.44, + "step": 1375 + }, + { + "epoch": 5.6, + "grad_norm": 2.5064358711242676, + "learning_rate": 0.00032002639392939627, + "loss": 0.4583, + "step": 1400 + }, + { + "epoch": 5.7, + "grad_norm": 1.8364307880401611, + "learning_rate": 0.0003183767733421313, + "loss": 0.4516, + "step": 1425 + }, + { + "epoch": 5.8, + "grad_norm": 1.3747583627700806, + "learning_rate": 0.0003167271527548664, + "loss": 0.4535, + "step": 1450 + }, + { + "epoch": 5.9, + "grad_norm": 2.580183982849121, + "learning_rate": 0.00031507753216760147, + "loss": 0.4325, + "step": 1475 + }, + { + "epoch": 6.0, + "grad_norm": 1.562374472618103, + "learning_rate": 0.00031342791158033657, + "loss": 0.4616, + "step": 1500 + }, + { + "epoch": 6.1, + "grad_norm": 1.4970141649246216, + "learning_rate": 0.00031177829099307157, + "loss": 0.4607, + "step": 1525 + }, + { + "epoch": 6.2, + "grad_norm": 2.436326503753662, + "learning_rate": 0.00031012867040580667, + "loss": 0.4498, + "step": 1550 + }, + { + "epoch": 6.3, + "grad_norm": 1.9849016666412354, + "learning_rate": 0.0003084790498185417, + "loss": 0.4262, + "step": 1575 + }, + { + "epoch": 6.4, + "grad_norm": 1.6405338048934937, + "learning_rate": 0.0003068294292312768, + "loss": 0.4225, + "step": 1600 + }, + { + "epoch": 6.5, + "grad_norm": 1.8233951330184937, + "learning_rate": 0.0003051798086440119, + "loss": 0.4248, + "step": 1625 + }, + { + "epoch": 6.6, + "grad_norm": 1.396469235420227, + "learning_rate": 0.000303530188056747, + "loss": 0.412, + "step": 1650 + }, + { + "epoch": 6.7, + "grad_norm": 1.5438531637191772, + "learning_rate": 0.000301880567469482, + "loss": 0.4105, + "step": 1675 + }, + { + "epoch": 6.8, + "grad_norm": 2.49668288230896, + "learning_rate": 0.0003002309468822171, + "loss": 0.4318, + "step": 1700 + }, + { + "epoch": 6.9, + "grad_norm": 1.5911107063293457, + "learning_rate": 0.0002985813262949522, + "loss": 0.4454, + "step": 1725 + }, + { + "epoch": 7.0, + "grad_norm": 2.9934186935424805, + "learning_rate": 0.00029693170570768723, + "loss": 0.4306, + "step": 1750 + }, + { + "epoch": 7.1, + "grad_norm": 2.5476737022399902, + "learning_rate": 0.00029528208512042233, + "loss": 0.4066, + "step": 1775 + }, + { + "epoch": 7.2, + "grad_norm": 4.207254886627197, + "learning_rate": 0.0002936324645331574, + "loss": 0.4521, + "step": 1800 + }, + { + "epoch": 7.3, + "grad_norm": 1.8378102779388428, + "learning_rate": 0.00029198284394589243, + "loss": 0.4197, + "step": 1825 + }, + { + "epoch": 7.4, + "grad_norm": 2.2129015922546387, + "learning_rate": 0.00029033322335862753, + "loss": 0.4182, + "step": 1850 + }, + { + "epoch": 7.5, + "grad_norm": 1.4666684865951538, + "learning_rate": 0.0002886836027713626, + "loss": 0.4224, + "step": 1875 + }, + { + "epoch": 7.6, + "grad_norm": 2.0311567783355713, + "learning_rate": 0.0002870339821840977, + "loss": 0.4304, + "step": 1900 + }, + { + "epoch": 7.7, + "grad_norm": 2.346604585647583, + "learning_rate": 0.00028538436159683273, + "loss": 0.4012, + "step": 1925 + }, + { + "epoch": 7.8, + "grad_norm": 1.440136432647705, + "learning_rate": 0.00028373474100956784, + "loss": 0.4487, + "step": 1950 + }, + { + "epoch": 7.9, + "grad_norm": 2.426828384399414, + "learning_rate": 0.0002820851204223029, + "loss": 0.3577, + "step": 1975 + }, + { + "epoch": 8.0, + "grad_norm": 3.095980405807495, + "learning_rate": 0.00028043549983503794, + "loss": 0.4121, + "step": 2000 + }, + { + "epoch": 8.1, + "grad_norm": 1.8646572828292847, + "learning_rate": 0.00027878587924777304, + "loss": 0.4213, + "step": 2025 + }, + { + "epoch": 8.2, + "grad_norm": 2.3382062911987305, + "learning_rate": 0.0002771362586605081, + "loss": 0.4119, + "step": 2050 + }, + { + "epoch": 8.3, + "grad_norm": 1.7999392747879028, + "learning_rate": 0.0002754866380732432, + "loss": 0.4165, + "step": 2075 + }, + { + "epoch": 8.4, + "grad_norm": 2.262005567550659, + "learning_rate": 0.00027383701748597824, + "loss": 0.4043, + "step": 2100 + }, + { + "epoch": 8.5, + "grad_norm": 1.8611185550689697, + "learning_rate": 0.0002721873968987133, + "loss": 0.4022, + "step": 2125 + }, + { + "epoch": 8.6, + "grad_norm": 1.3322739601135254, + "learning_rate": 0.00027053777631144834, + "loss": 0.3835, + "step": 2150 + }, + { + "epoch": 8.7, + "grad_norm": 1.335750937461853, + "learning_rate": 0.00026888815572418344, + "loss": 0.424, + "step": 2175 + }, + { + "epoch": 8.8, + "grad_norm": 1.4633246660232544, + "learning_rate": 0.00026723853513691855, + "loss": 0.4124, + "step": 2200 + }, + { + "epoch": 8.9, + "grad_norm": 2.061042308807373, + "learning_rate": 0.0002655889145496536, + "loss": 0.4016, + "step": 2225 + }, + { + "epoch": 9.0, + "grad_norm": 1.9621604681015015, + "learning_rate": 0.0002639392939623887, + "loss": 0.4256, + "step": 2250 + }, + { + "epoch": 9.1, + "grad_norm": 1.4014378786087036, + "learning_rate": 0.0002622896733751237, + "loss": 0.3806, + "step": 2275 + }, + { + "epoch": 9.2, + "grad_norm": 2.6264309883117676, + "learning_rate": 0.0002606400527878588, + "loss": 0.3636, + "step": 2300 + }, + { + "epoch": 9.3, + "grad_norm": 1.8708316087722778, + "learning_rate": 0.00025899043220059385, + "loss": 0.4101, + "step": 2325 + }, + { + "epoch": 9.4, + "grad_norm": 2.156869649887085, + "learning_rate": 0.00025734081161332895, + "loss": 0.4114, + "step": 2350 + }, + { + "epoch": 9.5, + "grad_norm": 2.483182907104492, + "learning_rate": 0.000255691191026064, + "loss": 0.421, + "step": 2375 + }, + { + "epoch": 9.6, + "grad_norm": 2.1955626010894775, + "learning_rate": 0.0002540415704387991, + "loss": 0.3918, + "step": 2400 + }, + { + "epoch": 9.7, + "grad_norm": 4.115762233734131, + "learning_rate": 0.00025239194985153415, + "loss": 0.4172, + "step": 2425 + }, + { + "epoch": 9.8, + "grad_norm": 2.144148349761963, + "learning_rate": 0.0002507423292642692, + "loss": 0.3636, + "step": 2450 + }, + { + "epoch": 9.9, + "grad_norm": 2.340376377105713, + "learning_rate": 0.0002490927086770043, + "loss": 0.4059, + "step": 2475 + }, + { + "epoch": 10.0, + "grad_norm": 1.5438555479049683, + "learning_rate": 0.00024744308808973936, + "loss": 0.4232, + "step": 2500 + }, + { + "epoch": 10.1, + "grad_norm": 2.485889196395874, + "learning_rate": 0.00024579346750247446, + "loss": 0.3861, + "step": 2525 + }, + { + "epoch": 10.2, + "grad_norm": 2.4561119079589844, + "learning_rate": 0.0002441438469152095, + "loss": 0.3565, + "step": 2550 + }, + { + "epoch": 10.3, + "grad_norm": 2.460170269012451, + "learning_rate": 0.00024249422632794456, + "loss": 0.4003, + "step": 2575 + }, + { + "epoch": 10.4, + "grad_norm": 2.774445056915283, + "learning_rate": 0.00024084460574067963, + "loss": 0.4088, + "step": 2600 + }, + { + "epoch": 10.5, + "grad_norm": 2.678745985031128, + "learning_rate": 0.0002391949851534147, + "loss": 0.4021, + "step": 2625 + }, + { + "epoch": 10.6, + "grad_norm": 1.4933565855026245, + "learning_rate": 0.0002375453645661498, + "loss": 0.3744, + "step": 2650 + }, + { + "epoch": 10.7, + "grad_norm": 1.367900013923645, + "learning_rate": 0.00023589574397888486, + "loss": 0.3925, + "step": 2675 + }, + { + "epoch": 10.8, + "grad_norm": 1.653441309928894, + "learning_rate": 0.00023424612339161997, + "loss": 0.4049, + "step": 2700 + }, + { + "epoch": 10.9, + "grad_norm": 1.5892399549484253, + "learning_rate": 0.000232596502804355, + "loss": 0.3927, + "step": 2725 + }, + { + "epoch": 11.0, + "grad_norm": 1.535090684890747, + "learning_rate": 0.00023094688221709007, + "loss": 0.3739, + "step": 2750 + }, + { + "epoch": 11.1, + "grad_norm": 2.247271776199341, + "learning_rate": 0.00022929726162982514, + "loss": 0.3362, + "step": 2775 + }, + { + "epoch": 11.2, + "grad_norm": 2.408069372177124, + "learning_rate": 0.00022764764104256022, + "loss": 0.3933, + "step": 2800 + }, + { + "epoch": 11.3, + "grad_norm": 2.219688653945923, + "learning_rate": 0.0002259980204552953, + "loss": 0.4062, + "step": 2825 + }, + { + "epoch": 11.4, + "grad_norm": 3.0188546180725098, + "learning_rate": 0.00022434839986803037, + "loss": 0.3926, + "step": 2850 + }, + { + "epoch": 11.5, + "grad_norm": 2.8080244064331055, + "learning_rate": 0.00022269877928076542, + "loss": 0.3942, + "step": 2875 + }, + { + "epoch": 11.6, + "grad_norm": 2.263610601425171, + "learning_rate": 0.0002210491586935005, + "loss": 0.3887, + "step": 2900 + }, + { + "epoch": 11.7, + "grad_norm": 2.1239607334136963, + "learning_rate": 0.00021939953810623557, + "loss": 0.3989, + "step": 2925 + }, + { + "epoch": 11.8, + "grad_norm": 3.435351848602295, + "learning_rate": 0.00021774991751897065, + "loss": 0.4185, + "step": 2950 + }, + { + "epoch": 11.9, + "grad_norm": 4.1570515632629395, + "learning_rate": 0.00021610029693170573, + "loss": 0.3597, + "step": 2975 + }, + { + "epoch": 12.0, + "grad_norm": 1.7937941551208496, + "learning_rate": 0.0002144506763444408, + "loss": 0.3719, + "step": 3000 + } + ], + "logging_steps": 25, + "max_steps": 6250, + "num_input_tokens_seen": 0, + "num_train_epochs": 25, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 166143098880000.0, + "train_batch_size": 20, + "trial_name": null, + "trial_params": null +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/training_args.bin b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..10d16bb8e9df9c51aba38e7a6f04deeecf1ea29f --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:662a5f25d38ccae34f19d85e24456de6d3eaaabafca26fb7cc8816de0c990f20 +size 5304 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/vocab.txt b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/vocab.txt new file mode 100644 index 0000000000000000000000000000000000000000..cc9ad53732e21ea9180f18e1fb67a38cc5702d4c --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/vocab.txt @@ -0,0 +1,4101 @@ +[PAD] +[UNK] +[CLS] +[SEP] +[MASK] +AAAAAA +AAAAAT +AAAAAC +AAAAAG +AAAATA +AAAATT +AAAATC +AAAATG +AAAACA +AAAACT +AAAACC +AAAACG +AAAAGA +AAAAGT +AAAAGC +AAAAGG +AAATAA +AAATAT +AAATAC +AAATAG +AAATTA +AAATTT +AAATTC +AAATTG +AAATCA +AAATCT +AAATCC +AAATCG +AAATGA +AAATGT +AAATGC +AAATGG +AAACAA +AAACAT +AAACAC +AAACAG +AAACTA +AAACTT +AAACTC +AAACTG +AAACCA +AAACCT +AAACCC +AAACCG +AAACGA +AAACGT +AAACGC +AAACGG +AAAGAA +AAAGAT +AAAGAC +AAAGAG +AAAGTA +AAAGTT +AAAGTC +AAAGTG +AAAGCA +AAAGCT +AAAGCC +AAAGCG +AAAGGA +AAAGGT +AAAGGC +AAAGGG +AATAAA +AATAAT +AATAAC +AATAAG +AATATA +AATATT +AATATC +AATATG +AATACA +AATACT +AATACC +AATACG +AATAGA +AATAGT +AATAGC +AATAGG +AATTAA +AATTAT +AATTAC +AATTAG +AATTTA +AATTTT +AATTTC +AATTTG +AATTCA +AATTCT +AATTCC +AATTCG +AATTGA +AATTGT +AATTGC +AATTGG +AATCAA +AATCAT +AATCAC +AATCAG +AATCTA +AATCTT +AATCTC +AATCTG +AATCCA +AATCCT +AATCCC +AATCCG +AATCGA +AATCGT +AATCGC +AATCGG +AATGAA +AATGAT +AATGAC +AATGAG +AATGTA +AATGTT +AATGTC +AATGTG +AATGCA +AATGCT +AATGCC +AATGCG +AATGGA +AATGGT +AATGGC +AATGGG +AACAAA +AACAAT +AACAAC +AACAAG +AACATA +AACATT +AACATC +AACATG +AACACA +AACACT +AACACC +AACACG +AACAGA +AACAGT +AACAGC +AACAGG +AACTAA +AACTAT +AACTAC +AACTAG +AACTTA +AACTTT +AACTTC +AACTTG +AACTCA +AACTCT +AACTCC +AACTCG +AACTGA +AACTGT +AACTGC +AACTGG +AACCAA +AACCAT +AACCAC +AACCAG +AACCTA +AACCTT +AACCTC +AACCTG +AACCCA +AACCCT +AACCCC +AACCCG +AACCGA +AACCGT +AACCGC +AACCGG +AACGAA +AACGAT +AACGAC +AACGAG +AACGTA +AACGTT +AACGTC +AACGTG +AACGCA +AACGCT +AACGCC +AACGCG +AACGGA +AACGGT +AACGGC +AACGGG +AAGAAA +AAGAAT +AAGAAC +AAGAAG +AAGATA +AAGATT +AAGATC +AAGATG +AAGACA +AAGACT +AAGACC +AAGACG +AAGAGA +AAGAGT +AAGAGC +AAGAGG +AAGTAA +AAGTAT +AAGTAC +AAGTAG +AAGTTA +AAGTTT +AAGTTC +AAGTTG +AAGTCA +AAGTCT +AAGTCC +AAGTCG +AAGTGA +AAGTGT +AAGTGC +AAGTGG +AAGCAA +AAGCAT +AAGCAC +AAGCAG +AAGCTA +AAGCTT +AAGCTC +AAGCTG +AAGCCA +AAGCCT +AAGCCC +AAGCCG +AAGCGA +AAGCGT +AAGCGC +AAGCGG +AAGGAA +AAGGAT +AAGGAC +AAGGAG +AAGGTA +AAGGTT +AAGGTC +AAGGTG +AAGGCA +AAGGCT +AAGGCC +AAGGCG +AAGGGA +AAGGGT +AAGGGC +AAGGGG +ATAAAA +ATAAAT +ATAAAC +ATAAAG +ATAATA +ATAATT +ATAATC +ATAATG +ATAACA +ATAACT +ATAACC +ATAACG +ATAAGA +ATAAGT +ATAAGC +ATAAGG +ATATAA +ATATAT +ATATAC +ATATAG +ATATTA +ATATTT +ATATTC +ATATTG +ATATCA +ATATCT +ATATCC +ATATCG +ATATGA +ATATGT +ATATGC +ATATGG +ATACAA +ATACAT +ATACAC +ATACAG +ATACTA +ATACTT +ATACTC +ATACTG +ATACCA +ATACCT +ATACCC +ATACCG +ATACGA +ATACGT +ATACGC +ATACGG +ATAGAA +ATAGAT +ATAGAC +ATAGAG +ATAGTA +ATAGTT +ATAGTC +ATAGTG +ATAGCA +ATAGCT +ATAGCC +ATAGCG +ATAGGA +ATAGGT +ATAGGC +ATAGGG +ATTAAA +ATTAAT +ATTAAC +ATTAAG +ATTATA +ATTATT +ATTATC +ATTATG +ATTACA +ATTACT +ATTACC +ATTACG +ATTAGA +ATTAGT +ATTAGC +ATTAGG +ATTTAA +ATTTAT +ATTTAC +ATTTAG +ATTTTA +ATTTTT +ATTTTC +ATTTTG +ATTTCA +ATTTCT +ATTTCC +ATTTCG +ATTTGA +ATTTGT +ATTTGC +ATTTGG +ATTCAA +ATTCAT +ATTCAC +ATTCAG +ATTCTA +ATTCTT +ATTCTC +ATTCTG +ATTCCA +ATTCCT +ATTCCC +ATTCCG +ATTCGA +ATTCGT +ATTCGC +ATTCGG +ATTGAA +ATTGAT +ATTGAC +ATTGAG +ATTGTA +ATTGTT +ATTGTC +ATTGTG +ATTGCA +ATTGCT +ATTGCC +ATTGCG +ATTGGA +ATTGGT +ATTGGC +ATTGGG +ATCAAA +ATCAAT +ATCAAC +ATCAAG +ATCATA +ATCATT +ATCATC +ATCATG +ATCACA +ATCACT +ATCACC +ATCACG +ATCAGA +ATCAGT +ATCAGC +ATCAGG +ATCTAA +ATCTAT +ATCTAC +ATCTAG +ATCTTA +ATCTTT +ATCTTC +ATCTTG +ATCTCA +ATCTCT +ATCTCC +ATCTCG +ATCTGA +ATCTGT +ATCTGC +ATCTGG +ATCCAA +ATCCAT +ATCCAC +ATCCAG +ATCCTA +ATCCTT +ATCCTC +ATCCTG +ATCCCA +ATCCCT +ATCCCC +ATCCCG +ATCCGA +ATCCGT +ATCCGC +ATCCGG +ATCGAA +ATCGAT +ATCGAC +ATCGAG +ATCGTA +ATCGTT +ATCGTC +ATCGTG +ATCGCA +ATCGCT +ATCGCC +ATCGCG +ATCGGA +ATCGGT +ATCGGC +ATCGGG +ATGAAA +ATGAAT +ATGAAC +ATGAAG +ATGATA +ATGATT +ATGATC +ATGATG +ATGACA +ATGACT +ATGACC +ATGACG +ATGAGA +ATGAGT +ATGAGC +ATGAGG +ATGTAA +ATGTAT +ATGTAC +ATGTAG +ATGTTA +ATGTTT +ATGTTC +ATGTTG +ATGTCA +ATGTCT +ATGTCC +ATGTCG +ATGTGA +ATGTGT +ATGTGC +ATGTGG +ATGCAA +ATGCAT +ATGCAC +ATGCAG +ATGCTA +ATGCTT +ATGCTC +ATGCTG +ATGCCA +ATGCCT +ATGCCC +ATGCCG +ATGCGA +ATGCGT +ATGCGC +ATGCGG +ATGGAA +ATGGAT +ATGGAC +ATGGAG +ATGGTA +ATGGTT +ATGGTC +ATGGTG +ATGGCA +ATGGCT +ATGGCC +ATGGCG +ATGGGA +ATGGGT +ATGGGC +ATGGGG +ACAAAA +ACAAAT +ACAAAC +ACAAAG +ACAATA +ACAATT +ACAATC +ACAATG +ACAACA +ACAACT +ACAACC +ACAACG +ACAAGA +ACAAGT +ACAAGC +ACAAGG +ACATAA +ACATAT +ACATAC +ACATAG +ACATTA +ACATTT +ACATTC +ACATTG +ACATCA +ACATCT +ACATCC +ACATCG +ACATGA +ACATGT +ACATGC +ACATGG +ACACAA +ACACAT +ACACAC +ACACAG +ACACTA +ACACTT +ACACTC +ACACTG +ACACCA +ACACCT +ACACCC +ACACCG +ACACGA +ACACGT +ACACGC +ACACGG +ACAGAA +ACAGAT +ACAGAC +ACAGAG +ACAGTA +ACAGTT +ACAGTC +ACAGTG +ACAGCA +ACAGCT +ACAGCC +ACAGCG +ACAGGA +ACAGGT +ACAGGC +ACAGGG +ACTAAA +ACTAAT +ACTAAC +ACTAAG +ACTATA +ACTATT +ACTATC +ACTATG +ACTACA +ACTACT +ACTACC +ACTACG +ACTAGA +ACTAGT +ACTAGC +ACTAGG +ACTTAA +ACTTAT +ACTTAC +ACTTAG +ACTTTA +ACTTTT +ACTTTC +ACTTTG +ACTTCA +ACTTCT +ACTTCC +ACTTCG +ACTTGA +ACTTGT +ACTTGC +ACTTGG +ACTCAA +ACTCAT +ACTCAC +ACTCAG +ACTCTA +ACTCTT +ACTCTC +ACTCTG +ACTCCA +ACTCCT +ACTCCC +ACTCCG +ACTCGA +ACTCGT +ACTCGC +ACTCGG +ACTGAA +ACTGAT +ACTGAC +ACTGAG +ACTGTA +ACTGTT +ACTGTC +ACTGTG +ACTGCA +ACTGCT +ACTGCC +ACTGCG +ACTGGA +ACTGGT +ACTGGC +ACTGGG +ACCAAA +ACCAAT +ACCAAC +ACCAAG +ACCATA +ACCATT +ACCATC +ACCATG +ACCACA +ACCACT +ACCACC +ACCACG +ACCAGA +ACCAGT +ACCAGC +ACCAGG +ACCTAA +ACCTAT +ACCTAC +ACCTAG +ACCTTA +ACCTTT +ACCTTC +ACCTTG +ACCTCA +ACCTCT +ACCTCC +ACCTCG +ACCTGA +ACCTGT +ACCTGC +ACCTGG +ACCCAA +ACCCAT +ACCCAC +ACCCAG +ACCCTA +ACCCTT +ACCCTC +ACCCTG +ACCCCA +ACCCCT +ACCCCC +ACCCCG +ACCCGA +ACCCGT +ACCCGC +ACCCGG +ACCGAA +ACCGAT +ACCGAC +ACCGAG +ACCGTA +ACCGTT +ACCGTC +ACCGTG +ACCGCA +ACCGCT +ACCGCC +ACCGCG +ACCGGA +ACCGGT +ACCGGC +ACCGGG +ACGAAA +ACGAAT +ACGAAC +ACGAAG +ACGATA +ACGATT +ACGATC +ACGATG +ACGACA +ACGACT +ACGACC +ACGACG +ACGAGA +ACGAGT +ACGAGC +ACGAGG +ACGTAA +ACGTAT +ACGTAC +ACGTAG +ACGTTA +ACGTTT +ACGTTC +ACGTTG +ACGTCA +ACGTCT +ACGTCC +ACGTCG +ACGTGA +ACGTGT +ACGTGC +ACGTGG +ACGCAA +ACGCAT +ACGCAC +ACGCAG +ACGCTA +ACGCTT +ACGCTC +ACGCTG +ACGCCA +ACGCCT +ACGCCC +ACGCCG +ACGCGA +ACGCGT +ACGCGC +ACGCGG +ACGGAA +ACGGAT +ACGGAC +ACGGAG +ACGGTA +ACGGTT +ACGGTC +ACGGTG +ACGGCA +ACGGCT +ACGGCC +ACGGCG +ACGGGA +ACGGGT +ACGGGC +ACGGGG +AGAAAA +AGAAAT +AGAAAC +AGAAAG +AGAATA +AGAATT +AGAATC +AGAATG +AGAACA +AGAACT +AGAACC +AGAACG +AGAAGA +AGAAGT +AGAAGC +AGAAGG +AGATAA +AGATAT +AGATAC +AGATAG +AGATTA +AGATTT +AGATTC +AGATTG +AGATCA +AGATCT +AGATCC +AGATCG +AGATGA +AGATGT +AGATGC +AGATGG +AGACAA +AGACAT +AGACAC +AGACAG +AGACTA +AGACTT +AGACTC +AGACTG +AGACCA +AGACCT +AGACCC +AGACCG +AGACGA +AGACGT +AGACGC +AGACGG +AGAGAA +AGAGAT +AGAGAC +AGAGAG +AGAGTA +AGAGTT +AGAGTC +AGAGTG +AGAGCA +AGAGCT +AGAGCC +AGAGCG +AGAGGA +AGAGGT +AGAGGC +AGAGGG +AGTAAA +AGTAAT +AGTAAC +AGTAAG +AGTATA +AGTATT +AGTATC +AGTATG +AGTACA +AGTACT +AGTACC +AGTACG +AGTAGA +AGTAGT +AGTAGC +AGTAGG +AGTTAA +AGTTAT +AGTTAC +AGTTAG +AGTTTA +AGTTTT +AGTTTC +AGTTTG +AGTTCA +AGTTCT +AGTTCC +AGTTCG +AGTTGA +AGTTGT +AGTTGC +AGTTGG +AGTCAA +AGTCAT +AGTCAC +AGTCAG +AGTCTA +AGTCTT +AGTCTC +AGTCTG +AGTCCA +AGTCCT +AGTCCC +AGTCCG +AGTCGA +AGTCGT +AGTCGC +AGTCGG +AGTGAA +AGTGAT +AGTGAC +AGTGAG +AGTGTA +AGTGTT +AGTGTC +AGTGTG +AGTGCA +AGTGCT +AGTGCC +AGTGCG +AGTGGA +AGTGGT +AGTGGC +AGTGGG +AGCAAA +AGCAAT +AGCAAC +AGCAAG +AGCATA +AGCATT +AGCATC +AGCATG +AGCACA +AGCACT +AGCACC +AGCACG +AGCAGA +AGCAGT +AGCAGC +AGCAGG +AGCTAA +AGCTAT +AGCTAC +AGCTAG +AGCTTA +AGCTTT +AGCTTC +AGCTTG +AGCTCA +AGCTCT +AGCTCC +AGCTCG +AGCTGA +AGCTGT +AGCTGC +AGCTGG +AGCCAA +AGCCAT +AGCCAC +AGCCAG +AGCCTA +AGCCTT +AGCCTC +AGCCTG +AGCCCA +AGCCCT +AGCCCC +AGCCCG +AGCCGA +AGCCGT +AGCCGC +AGCCGG +AGCGAA +AGCGAT +AGCGAC +AGCGAG +AGCGTA +AGCGTT +AGCGTC +AGCGTG +AGCGCA +AGCGCT +AGCGCC +AGCGCG +AGCGGA +AGCGGT +AGCGGC +AGCGGG +AGGAAA +AGGAAT +AGGAAC +AGGAAG +AGGATA +AGGATT +AGGATC +AGGATG +AGGACA +AGGACT +AGGACC +AGGACG +AGGAGA +AGGAGT +AGGAGC +AGGAGG +AGGTAA +AGGTAT +AGGTAC +AGGTAG +AGGTTA +AGGTTT +AGGTTC +AGGTTG +AGGTCA +AGGTCT +AGGTCC +AGGTCG +AGGTGA +AGGTGT +AGGTGC +AGGTGG +AGGCAA +AGGCAT +AGGCAC +AGGCAG +AGGCTA +AGGCTT +AGGCTC +AGGCTG +AGGCCA +AGGCCT +AGGCCC +AGGCCG +AGGCGA +AGGCGT +AGGCGC +AGGCGG +AGGGAA +AGGGAT +AGGGAC +AGGGAG +AGGGTA +AGGGTT +AGGGTC +AGGGTG +AGGGCA +AGGGCT +AGGGCC +AGGGCG +AGGGGA +AGGGGT +AGGGGC +AGGGGG +TAAAAA +TAAAAT +TAAAAC +TAAAAG +TAAATA +TAAATT +TAAATC +TAAATG +TAAACA +TAAACT +TAAACC +TAAACG +TAAAGA +TAAAGT +TAAAGC +TAAAGG +TAATAA +TAATAT +TAATAC +TAATAG +TAATTA +TAATTT +TAATTC +TAATTG +TAATCA +TAATCT +TAATCC +TAATCG +TAATGA +TAATGT +TAATGC +TAATGG +TAACAA +TAACAT +TAACAC +TAACAG +TAACTA +TAACTT +TAACTC +TAACTG +TAACCA +TAACCT +TAACCC +TAACCG +TAACGA +TAACGT +TAACGC +TAACGG +TAAGAA +TAAGAT +TAAGAC +TAAGAG +TAAGTA +TAAGTT +TAAGTC +TAAGTG +TAAGCA +TAAGCT +TAAGCC +TAAGCG +TAAGGA +TAAGGT +TAAGGC +TAAGGG +TATAAA +TATAAT +TATAAC +TATAAG +TATATA +TATATT +TATATC +TATATG +TATACA +TATACT +TATACC +TATACG +TATAGA +TATAGT +TATAGC +TATAGG +TATTAA +TATTAT +TATTAC +TATTAG +TATTTA +TATTTT +TATTTC +TATTTG +TATTCA +TATTCT +TATTCC +TATTCG +TATTGA +TATTGT +TATTGC +TATTGG +TATCAA +TATCAT +TATCAC +TATCAG +TATCTA +TATCTT +TATCTC +TATCTG +TATCCA +TATCCT +TATCCC +TATCCG +TATCGA +TATCGT +TATCGC +TATCGG +TATGAA +TATGAT +TATGAC +TATGAG +TATGTA +TATGTT +TATGTC +TATGTG +TATGCA +TATGCT +TATGCC +TATGCG +TATGGA +TATGGT +TATGGC +TATGGG +TACAAA +TACAAT +TACAAC +TACAAG +TACATA +TACATT +TACATC +TACATG +TACACA +TACACT +TACACC +TACACG +TACAGA +TACAGT +TACAGC +TACAGG +TACTAA +TACTAT +TACTAC +TACTAG +TACTTA +TACTTT +TACTTC +TACTTG +TACTCA +TACTCT +TACTCC +TACTCG +TACTGA +TACTGT +TACTGC +TACTGG +TACCAA +TACCAT +TACCAC +TACCAG +TACCTA +TACCTT +TACCTC +TACCTG +TACCCA +TACCCT +TACCCC +TACCCG +TACCGA +TACCGT +TACCGC +TACCGG +TACGAA +TACGAT +TACGAC +TACGAG +TACGTA +TACGTT +TACGTC +TACGTG +TACGCA +TACGCT +TACGCC +TACGCG +TACGGA +TACGGT +TACGGC +TACGGG +TAGAAA +TAGAAT +TAGAAC +TAGAAG +TAGATA +TAGATT +TAGATC +TAGATG +TAGACA +TAGACT +TAGACC +TAGACG +TAGAGA +TAGAGT +TAGAGC +TAGAGG +TAGTAA +TAGTAT +TAGTAC +TAGTAG +TAGTTA +TAGTTT +TAGTTC +TAGTTG +TAGTCA +TAGTCT +TAGTCC +TAGTCG +TAGTGA +TAGTGT +TAGTGC +TAGTGG +TAGCAA +TAGCAT +TAGCAC +TAGCAG +TAGCTA +TAGCTT +TAGCTC +TAGCTG +TAGCCA +TAGCCT +TAGCCC +TAGCCG +TAGCGA +TAGCGT +TAGCGC +TAGCGG +TAGGAA +TAGGAT +TAGGAC +TAGGAG +TAGGTA +TAGGTT +TAGGTC +TAGGTG +TAGGCA +TAGGCT +TAGGCC +TAGGCG +TAGGGA +TAGGGT +TAGGGC +TAGGGG +TTAAAA +TTAAAT +TTAAAC +TTAAAG +TTAATA +TTAATT +TTAATC +TTAATG +TTAACA +TTAACT +TTAACC +TTAACG +TTAAGA +TTAAGT +TTAAGC +TTAAGG +TTATAA +TTATAT +TTATAC +TTATAG +TTATTA +TTATTT +TTATTC +TTATTG +TTATCA +TTATCT +TTATCC +TTATCG +TTATGA +TTATGT +TTATGC +TTATGG +TTACAA +TTACAT +TTACAC +TTACAG +TTACTA +TTACTT +TTACTC +TTACTG +TTACCA +TTACCT +TTACCC +TTACCG +TTACGA +TTACGT +TTACGC +TTACGG +TTAGAA +TTAGAT +TTAGAC +TTAGAG +TTAGTA +TTAGTT +TTAGTC +TTAGTG +TTAGCA +TTAGCT +TTAGCC +TTAGCG +TTAGGA +TTAGGT +TTAGGC +TTAGGG +TTTAAA +TTTAAT +TTTAAC +TTTAAG +TTTATA +TTTATT +TTTATC +TTTATG +TTTACA +TTTACT +TTTACC +TTTACG +TTTAGA +TTTAGT +TTTAGC +TTTAGG +TTTTAA +TTTTAT +TTTTAC +TTTTAG +TTTTTA +TTTTTT +TTTTTC +TTTTTG +TTTTCA +TTTTCT +TTTTCC +TTTTCG +TTTTGA +TTTTGT +TTTTGC +TTTTGG +TTTCAA +TTTCAT +TTTCAC +TTTCAG +TTTCTA +TTTCTT +TTTCTC +TTTCTG +TTTCCA +TTTCCT +TTTCCC +TTTCCG +TTTCGA +TTTCGT +TTTCGC +TTTCGG +TTTGAA +TTTGAT +TTTGAC +TTTGAG +TTTGTA +TTTGTT +TTTGTC +TTTGTG +TTTGCA +TTTGCT +TTTGCC +TTTGCG +TTTGGA +TTTGGT +TTTGGC +TTTGGG +TTCAAA +TTCAAT +TTCAAC +TTCAAG +TTCATA +TTCATT +TTCATC +TTCATG +TTCACA +TTCACT +TTCACC +TTCACG +TTCAGA +TTCAGT +TTCAGC +TTCAGG +TTCTAA +TTCTAT +TTCTAC +TTCTAG +TTCTTA +TTCTTT +TTCTTC +TTCTTG +TTCTCA +TTCTCT +TTCTCC +TTCTCG +TTCTGA +TTCTGT +TTCTGC +TTCTGG +TTCCAA +TTCCAT +TTCCAC +TTCCAG +TTCCTA +TTCCTT +TTCCTC +TTCCTG +TTCCCA +TTCCCT +TTCCCC +TTCCCG +TTCCGA +TTCCGT +TTCCGC +TTCCGG +TTCGAA +TTCGAT +TTCGAC +TTCGAG +TTCGTA +TTCGTT +TTCGTC +TTCGTG +TTCGCA +TTCGCT +TTCGCC +TTCGCG +TTCGGA +TTCGGT +TTCGGC +TTCGGG +TTGAAA +TTGAAT +TTGAAC +TTGAAG +TTGATA +TTGATT +TTGATC +TTGATG +TTGACA +TTGACT +TTGACC +TTGACG +TTGAGA +TTGAGT +TTGAGC +TTGAGG +TTGTAA +TTGTAT +TTGTAC +TTGTAG +TTGTTA +TTGTTT +TTGTTC +TTGTTG +TTGTCA +TTGTCT +TTGTCC +TTGTCG +TTGTGA +TTGTGT +TTGTGC +TTGTGG +TTGCAA +TTGCAT +TTGCAC +TTGCAG +TTGCTA +TTGCTT +TTGCTC +TTGCTG +TTGCCA +TTGCCT +TTGCCC +TTGCCG +TTGCGA +TTGCGT +TTGCGC +TTGCGG +TTGGAA +TTGGAT +TTGGAC +TTGGAG +TTGGTA +TTGGTT +TTGGTC +TTGGTG +TTGGCA +TTGGCT +TTGGCC +TTGGCG +TTGGGA +TTGGGT +TTGGGC +TTGGGG +TCAAAA +TCAAAT +TCAAAC +TCAAAG +TCAATA +TCAATT +TCAATC +TCAATG +TCAACA +TCAACT +TCAACC +TCAACG +TCAAGA +TCAAGT +TCAAGC +TCAAGG +TCATAA +TCATAT +TCATAC +TCATAG +TCATTA +TCATTT +TCATTC +TCATTG +TCATCA +TCATCT +TCATCC +TCATCG +TCATGA +TCATGT +TCATGC +TCATGG +TCACAA +TCACAT +TCACAC +TCACAG +TCACTA +TCACTT +TCACTC +TCACTG +TCACCA +TCACCT +TCACCC +TCACCG +TCACGA +TCACGT +TCACGC +TCACGG +TCAGAA +TCAGAT +TCAGAC +TCAGAG +TCAGTA +TCAGTT +TCAGTC +TCAGTG +TCAGCA +TCAGCT +TCAGCC +TCAGCG +TCAGGA +TCAGGT +TCAGGC +TCAGGG +TCTAAA +TCTAAT +TCTAAC +TCTAAG +TCTATA +TCTATT +TCTATC +TCTATG +TCTACA +TCTACT +TCTACC +TCTACG +TCTAGA +TCTAGT +TCTAGC +TCTAGG +TCTTAA +TCTTAT +TCTTAC +TCTTAG +TCTTTA +TCTTTT +TCTTTC +TCTTTG +TCTTCA +TCTTCT +TCTTCC +TCTTCG +TCTTGA +TCTTGT +TCTTGC +TCTTGG +TCTCAA +TCTCAT +TCTCAC +TCTCAG +TCTCTA +TCTCTT +TCTCTC +TCTCTG +TCTCCA +TCTCCT +TCTCCC +TCTCCG +TCTCGA +TCTCGT +TCTCGC +TCTCGG +TCTGAA +TCTGAT +TCTGAC +TCTGAG +TCTGTA +TCTGTT +TCTGTC +TCTGTG +TCTGCA +TCTGCT +TCTGCC +TCTGCG +TCTGGA +TCTGGT +TCTGGC +TCTGGG +TCCAAA +TCCAAT +TCCAAC +TCCAAG +TCCATA +TCCATT +TCCATC +TCCATG +TCCACA +TCCACT +TCCACC +TCCACG +TCCAGA +TCCAGT +TCCAGC +TCCAGG +TCCTAA +TCCTAT +TCCTAC +TCCTAG +TCCTTA +TCCTTT +TCCTTC +TCCTTG +TCCTCA +TCCTCT +TCCTCC +TCCTCG +TCCTGA +TCCTGT +TCCTGC +TCCTGG +TCCCAA +TCCCAT +TCCCAC +TCCCAG +TCCCTA +TCCCTT +TCCCTC +TCCCTG +TCCCCA +TCCCCT +TCCCCC +TCCCCG +TCCCGA +TCCCGT +TCCCGC +TCCCGG +TCCGAA +TCCGAT +TCCGAC +TCCGAG +TCCGTA +TCCGTT +TCCGTC +TCCGTG +TCCGCA +TCCGCT +TCCGCC +TCCGCG +TCCGGA +TCCGGT +TCCGGC +TCCGGG +TCGAAA +TCGAAT +TCGAAC +TCGAAG +TCGATA +TCGATT +TCGATC +TCGATG +TCGACA +TCGACT +TCGACC +TCGACG +TCGAGA +TCGAGT +TCGAGC +TCGAGG +TCGTAA +TCGTAT +TCGTAC +TCGTAG +TCGTTA +TCGTTT +TCGTTC +TCGTTG +TCGTCA +TCGTCT +TCGTCC +TCGTCG +TCGTGA +TCGTGT +TCGTGC +TCGTGG +TCGCAA +TCGCAT +TCGCAC +TCGCAG +TCGCTA +TCGCTT +TCGCTC +TCGCTG +TCGCCA +TCGCCT +TCGCCC +TCGCCG +TCGCGA +TCGCGT +TCGCGC +TCGCGG +TCGGAA +TCGGAT +TCGGAC +TCGGAG +TCGGTA +TCGGTT +TCGGTC +TCGGTG +TCGGCA +TCGGCT +TCGGCC +TCGGCG +TCGGGA +TCGGGT +TCGGGC +TCGGGG +TGAAAA +TGAAAT +TGAAAC +TGAAAG +TGAATA +TGAATT +TGAATC +TGAATG +TGAACA +TGAACT +TGAACC +TGAACG +TGAAGA +TGAAGT +TGAAGC +TGAAGG +TGATAA +TGATAT +TGATAC +TGATAG +TGATTA +TGATTT +TGATTC +TGATTG +TGATCA +TGATCT +TGATCC +TGATCG +TGATGA +TGATGT +TGATGC +TGATGG +TGACAA +TGACAT +TGACAC +TGACAG +TGACTA +TGACTT +TGACTC +TGACTG +TGACCA +TGACCT +TGACCC +TGACCG +TGACGA +TGACGT +TGACGC +TGACGG +TGAGAA +TGAGAT +TGAGAC +TGAGAG +TGAGTA +TGAGTT +TGAGTC +TGAGTG +TGAGCA +TGAGCT +TGAGCC +TGAGCG +TGAGGA +TGAGGT +TGAGGC +TGAGGG +TGTAAA +TGTAAT +TGTAAC +TGTAAG +TGTATA +TGTATT +TGTATC +TGTATG +TGTACA +TGTACT +TGTACC +TGTACG +TGTAGA +TGTAGT +TGTAGC +TGTAGG +TGTTAA +TGTTAT +TGTTAC +TGTTAG +TGTTTA +TGTTTT +TGTTTC +TGTTTG +TGTTCA +TGTTCT +TGTTCC +TGTTCG +TGTTGA +TGTTGT +TGTTGC +TGTTGG +TGTCAA +TGTCAT +TGTCAC +TGTCAG +TGTCTA +TGTCTT +TGTCTC +TGTCTG +TGTCCA +TGTCCT +TGTCCC +TGTCCG +TGTCGA +TGTCGT +TGTCGC +TGTCGG +TGTGAA +TGTGAT +TGTGAC +TGTGAG +TGTGTA +TGTGTT +TGTGTC +TGTGTG +TGTGCA +TGTGCT +TGTGCC +TGTGCG +TGTGGA +TGTGGT +TGTGGC +TGTGGG +TGCAAA +TGCAAT +TGCAAC +TGCAAG +TGCATA +TGCATT +TGCATC +TGCATG +TGCACA +TGCACT +TGCACC +TGCACG +TGCAGA +TGCAGT +TGCAGC +TGCAGG +TGCTAA +TGCTAT +TGCTAC +TGCTAG +TGCTTA +TGCTTT +TGCTTC +TGCTTG +TGCTCA +TGCTCT +TGCTCC +TGCTCG +TGCTGA +TGCTGT +TGCTGC +TGCTGG +TGCCAA +TGCCAT +TGCCAC +TGCCAG +TGCCTA +TGCCTT +TGCCTC +TGCCTG +TGCCCA +TGCCCT +TGCCCC +TGCCCG +TGCCGA +TGCCGT +TGCCGC +TGCCGG +TGCGAA +TGCGAT +TGCGAC +TGCGAG +TGCGTA +TGCGTT +TGCGTC +TGCGTG +TGCGCA +TGCGCT +TGCGCC +TGCGCG +TGCGGA +TGCGGT +TGCGGC +TGCGGG +TGGAAA +TGGAAT +TGGAAC +TGGAAG +TGGATA +TGGATT +TGGATC +TGGATG +TGGACA +TGGACT +TGGACC +TGGACG +TGGAGA +TGGAGT +TGGAGC +TGGAGG +TGGTAA +TGGTAT +TGGTAC +TGGTAG +TGGTTA +TGGTTT +TGGTTC +TGGTTG +TGGTCA +TGGTCT +TGGTCC +TGGTCG +TGGTGA +TGGTGT +TGGTGC +TGGTGG +TGGCAA +TGGCAT +TGGCAC +TGGCAG +TGGCTA +TGGCTT +TGGCTC +TGGCTG +TGGCCA +TGGCCT +TGGCCC +TGGCCG +TGGCGA +TGGCGT +TGGCGC +TGGCGG +TGGGAA +TGGGAT +TGGGAC +TGGGAG +TGGGTA +TGGGTT +TGGGTC +TGGGTG +TGGGCA +TGGGCT +TGGGCC +TGGGCG +TGGGGA +TGGGGT +TGGGGC +TGGGGG +CAAAAA +CAAAAT +CAAAAC +CAAAAG +CAAATA +CAAATT +CAAATC +CAAATG +CAAACA +CAAACT +CAAACC +CAAACG +CAAAGA +CAAAGT +CAAAGC +CAAAGG +CAATAA +CAATAT +CAATAC +CAATAG +CAATTA +CAATTT +CAATTC +CAATTG +CAATCA +CAATCT +CAATCC +CAATCG +CAATGA +CAATGT +CAATGC +CAATGG +CAACAA +CAACAT +CAACAC +CAACAG +CAACTA +CAACTT +CAACTC +CAACTG +CAACCA +CAACCT +CAACCC +CAACCG +CAACGA +CAACGT +CAACGC +CAACGG +CAAGAA +CAAGAT +CAAGAC +CAAGAG +CAAGTA +CAAGTT +CAAGTC +CAAGTG +CAAGCA +CAAGCT +CAAGCC +CAAGCG +CAAGGA +CAAGGT +CAAGGC +CAAGGG +CATAAA +CATAAT +CATAAC +CATAAG +CATATA +CATATT +CATATC +CATATG +CATACA +CATACT +CATACC +CATACG +CATAGA +CATAGT +CATAGC +CATAGG +CATTAA +CATTAT +CATTAC +CATTAG +CATTTA +CATTTT +CATTTC +CATTTG +CATTCA +CATTCT +CATTCC +CATTCG +CATTGA +CATTGT +CATTGC +CATTGG +CATCAA +CATCAT +CATCAC +CATCAG +CATCTA +CATCTT +CATCTC +CATCTG +CATCCA +CATCCT +CATCCC +CATCCG +CATCGA +CATCGT +CATCGC +CATCGG +CATGAA +CATGAT +CATGAC +CATGAG +CATGTA +CATGTT +CATGTC +CATGTG +CATGCA +CATGCT +CATGCC +CATGCG +CATGGA +CATGGT +CATGGC +CATGGG +CACAAA +CACAAT +CACAAC +CACAAG +CACATA +CACATT +CACATC +CACATG +CACACA +CACACT +CACACC +CACACG +CACAGA +CACAGT +CACAGC +CACAGG +CACTAA +CACTAT +CACTAC +CACTAG +CACTTA +CACTTT +CACTTC +CACTTG +CACTCA +CACTCT +CACTCC +CACTCG +CACTGA +CACTGT +CACTGC +CACTGG +CACCAA +CACCAT +CACCAC +CACCAG +CACCTA +CACCTT +CACCTC +CACCTG +CACCCA +CACCCT +CACCCC +CACCCG +CACCGA +CACCGT +CACCGC +CACCGG +CACGAA +CACGAT +CACGAC +CACGAG +CACGTA +CACGTT +CACGTC +CACGTG +CACGCA +CACGCT +CACGCC +CACGCG +CACGGA +CACGGT +CACGGC +CACGGG +CAGAAA +CAGAAT +CAGAAC +CAGAAG +CAGATA +CAGATT +CAGATC +CAGATG +CAGACA +CAGACT +CAGACC +CAGACG +CAGAGA +CAGAGT +CAGAGC +CAGAGG +CAGTAA +CAGTAT +CAGTAC +CAGTAG +CAGTTA +CAGTTT +CAGTTC +CAGTTG +CAGTCA +CAGTCT +CAGTCC +CAGTCG +CAGTGA +CAGTGT +CAGTGC +CAGTGG +CAGCAA +CAGCAT +CAGCAC +CAGCAG +CAGCTA +CAGCTT +CAGCTC +CAGCTG +CAGCCA +CAGCCT +CAGCCC +CAGCCG +CAGCGA +CAGCGT +CAGCGC +CAGCGG +CAGGAA +CAGGAT +CAGGAC +CAGGAG +CAGGTA +CAGGTT +CAGGTC +CAGGTG +CAGGCA +CAGGCT +CAGGCC +CAGGCG +CAGGGA +CAGGGT +CAGGGC +CAGGGG +CTAAAA +CTAAAT +CTAAAC +CTAAAG +CTAATA +CTAATT +CTAATC +CTAATG +CTAACA +CTAACT +CTAACC +CTAACG +CTAAGA +CTAAGT +CTAAGC +CTAAGG +CTATAA +CTATAT +CTATAC +CTATAG +CTATTA +CTATTT +CTATTC +CTATTG +CTATCA +CTATCT +CTATCC +CTATCG +CTATGA +CTATGT +CTATGC +CTATGG +CTACAA +CTACAT +CTACAC +CTACAG +CTACTA +CTACTT +CTACTC +CTACTG +CTACCA +CTACCT +CTACCC +CTACCG +CTACGA +CTACGT +CTACGC +CTACGG +CTAGAA +CTAGAT +CTAGAC +CTAGAG +CTAGTA +CTAGTT +CTAGTC +CTAGTG +CTAGCA +CTAGCT +CTAGCC +CTAGCG +CTAGGA +CTAGGT +CTAGGC +CTAGGG +CTTAAA +CTTAAT +CTTAAC +CTTAAG +CTTATA +CTTATT +CTTATC +CTTATG +CTTACA +CTTACT +CTTACC +CTTACG +CTTAGA +CTTAGT +CTTAGC +CTTAGG +CTTTAA +CTTTAT +CTTTAC +CTTTAG +CTTTTA +CTTTTT +CTTTTC +CTTTTG +CTTTCA +CTTTCT +CTTTCC +CTTTCG +CTTTGA +CTTTGT +CTTTGC +CTTTGG +CTTCAA +CTTCAT +CTTCAC +CTTCAG +CTTCTA +CTTCTT +CTTCTC +CTTCTG +CTTCCA +CTTCCT +CTTCCC +CTTCCG +CTTCGA +CTTCGT +CTTCGC +CTTCGG +CTTGAA +CTTGAT +CTTGAC +CTTGAG +CTTGTA +CTTGTT +CTTGTC +CTTGTG +CTTGCA +CTTGCT +CTTGCC +CTTGCG +CTTGGA +CTTGGT +CTTGGC +CTTGGG +CTCAAA +CTCAAT +CTCAAC +CTCAAG +CTCATA +CTCATT +CTCATC +CTCATG +CTCACA +CTCACT +CTCACC +CTCACG +CTCAGA +CTCAGT +CTCAGC +CTCAGG +CTCTAA +CTCTAT +CTCTAC +CTCTAG +CTCTTA +CTCTTT +CTCTTC +CTCTTG +CTCTCA +CTCTCT +CTCTCC +CTCTCG +CTCTGA +CTCTGT +CTCTGC +CTCTGG +CTCCAA +CTCCAT +CTCCAC +CTCCAG +CTCCTA +CTCCTT +CTCCTC +CTCCTG +CTCCCA +CTCCCT +CTCCCC +CTCCCG +CTCCGA +CTCCGT +CTCCGC +CTCCGG +CTCGAA +CTCGAT +CTCGAC +CTCGAG +CTCGTA +CTCGTT +CTCGTC +CTCGTG +CTCGCA +CTCGCT +CTCGCC +CTCGCG +CTCGGA +CTCGGT +CTCGGC +CTCGGG +CTGAAA +CTGAAT +CTGAAC +CTGAAG +CTGATA +CTGATT +CTGATC +CTGATG +CTGACA +CTGACT +CTGACC +CTGACG +CTGAGA +CTGAGT +CTGAGC +CTGAGG +CTGTAA +CTGTAT +CTGTAC +CTGTAG +CTGTTA +CTGTTT +CTGTTC +CTGTTG +CTGTCA +CTGTCT +CTGTCC +CTGTCG +CTGTGA +CTGTGT +CTGTGC +CTGTGG +CTGCAA +CTGCAT +CTGCAC +CTGCAG +CTGCTA +CTGCTT +CTGCTC +CTGCTG +CTGCCA +CTGCCT +CTGCCC +CTGCCG +CTGCGA +CTGCGT +CTGCGC +CTGCGG +CTGGAA +CTGGAT +CTGGAC +CTGGAG +CTGGTA +CTGGTT +CTGGTC +CTGGTG +CTGGCA +CTGGCT +CTGGCC +CTGGCG +CTGGGA +CTGGGT +CTGGGC +CTGGGG +CCAAAA +CCAAAT +CCAAAC +CCAAAG +CCAATA +CCAATT +CCAATC +CCAATG +CCAACA +CCAACT +CCAACC +CCAACG +CCAAGA +CCAAGT +CCAAGC +CCAAGG +CCATAA +CCATAT +CCATAC +CCATAG +CCATTA +CCATTT +CCATTC +CCATTG +CCATCA +CCATCT +CCATCC +CCATCG +CCATGA +CCATGT +CCATGC +CCATGG +CCACAA +CCACAT +CCACAC +CCACAG +CCACTA +CCACTT +CCACTC +CCACTG +CCACCA +CCACCT +CCACCC +CCACCG +CCACGA +CCACGT +CCACGC +CCACGG +CCAGAA +CCAGAT +CCAGAC +CCAGAG +CCAGTA +CCAGTT +CCAGTC +CCAGTG +CCAGCA +CCAGCT +CCAGCC +CCAGCG +CCAGGA +CCAGGT +CCAGGC +CCAGGG +CCTAAA +CCTAAT +CCTAAC +CCTAAG +CCTATA +CCTATT +CCTATC +CCTATG +CCTACA +CCTACT +CCTACC +CCTACG +CCTAGA +CCTAGT +CCTAGC +CCTAGG +CCTTAA +CCTTAT +CCTTAC +CCTTAG +CCTTTA +CCTTTT +CCTTTC +CCTTTG +CCTTCA +CCTTCT +CCTTCC +CCTTCG +CCTTGA +CCTTGT +CCTTGC +CCTTGG +CCTCAA +CCTCAT +CCTCAC +CCTCAG +CCTCTA +CCTCTT +CCTCTC +CCTCTG +CCTCCA +CCTCCT +CCTCCC +CCTCCG +CCTCGA +CCTCGT +CCTCGC +CCTCGG +CCTGAA +CCTGAT +CCTGAC +CCTGAG +CCTGTA +CCTGTT +CCTGTC +CCTGTG +CCTGCA +CCTGCT +CCTGCC +CCTGCG +CCTGGA +CCTGGT +CCTGGC +CCTGGG +CCCAAA +CCCAAT +CCCAAC +CCCAAG +CCCATA +CCCATT +CCCATC +CCCATG +CCCACA +CCCACT +CCCACC +CCCACG +CCCAGA +CCCAGT +CCCAGC +CCCAGG +CCCTAA +CCCTAT +CCCTAC +CCCTAG +CCCTTA +CCCTTT +CCCTTC +CCCTTG +CCCTCA +CCCTCT +CCCTCC +CCCTCG +CCCTGA +CCCTGT +CCCTGC +CCCTGG +CCCCAA +CCCCAT +CCCCAC +CCCCAG +CCCCTA +CCCCTT +CCCCTC +CCCCTG +CCCCCA +CCCCCT +CCCCCC +CCCCCG +CCCCGA +CCCCGT +CCCCGC +CCCCGG +CCCGAA +CCCGAT +CCCGAC +CCCGAG +CCCGTA +CCCGTT +CCCGTC +CCCGTG +CCCGCA +CCCGCT +CCCGCC +CCCGCG +CCCGGA +CCCGGT +CCCGGC +CCCGGG +CCGAAA +CCGAAT +CCGAAC +CCGAAG +CCGATA +CCGATT +CCGATC +CCGATG +CCGACA +CCGACT +CCGACC +CCGACG +CCGAGA +CCGAGT +CCGAGC +CCGAGG +CCGTAA +CCGTAT +CCGTAC +CCGTAG +CCGTTA +CCGTTT +CCGTTC +CCGTTG +CCGTCA +CCGTCT +CCGTCC +CCGTCG +CCGTGA +CCGTGT +CCGTGC +CCGTGG +CCGCAA +CCGCAT +CCGCAC +CCGCAG +CCGCTA +CCGCTT +CCGCTC +CCGCTG +CCGCCA +CCGCCT +CCGCCC +CCGCCG +CCGCGA +CCGCGT +CCGCGC +CCGCGG +CCGGAA +CCGGAT +CCGGAC +CCGGAG +CCGGTA +CCGGTT +CCGGTC +CCGGTG +CCGGCA +CCGGCT +CCGGCC +CCGGCG +CCGGGA +CCGGGT +CCGGGC +CCGGGG +CGAAAA +CGAAAT +CGAAAC +CGAAAG +CGAATA +CGAATT +CGAATC +CGAATG +CGAACA +CGAACT +CGAACC +CGAACG +CGAAGA +CGAAGT +CGAAGC +CGAAGG +CGATAA +CGATAT +CGATAC +CGATAG +CGATTA +CGATTT +CGATTC +CGATTG +CGATCA +CGATCT +CGATCC +CGATCG +CGATGA +CGATGT +CGATGC +CGATGG +CGACAA +CGACAT +CGACAC +CGACAG +CGACTA +CGACTT +CGACTC +CGACTG +CGACCA +CGACCT +CGACCC +CGACCG +CGACGA +CGACGT +CGACGC +CGACGG +CGAGAA +CGAGAT +CGAGAC +CGAGAG +CGAGTA +CGAGTT +CGAGTC +CGAGTG +CGAGCA +CGAGCT +CGAGCC +CGAGCG +CGAGGA +CGAGGT +CGAGGC +CGAGGG +CGTAAA +CGTAAT +CGTAAC +CGTAAG +CGTATA +CGTATT +CGTATC +CGTATG +CGTACA +CGTACT +CGTACC +CGTACG +CGTAGA +CGTAGT +CGTAGC +CGTAGG +CGTTAA +CGTTAT +CGTTAC +CGTTAG +CGTTTA +CGTTTT +CGTTTC +CGTTTG +CGTTCA +CGTTCT +CGTTCC +CGTTCG +CGTTGA +CGTTGT +CGTTGC +CGTTGG +CGTCAA +CGTCAT +CGTCAC +CGTCAG +CGTCTA +CGTCTT +CGTCTC +CGTCTG +CGTCCA +CGTCCT +CGTCCC +CGTCCG +CGTCGA +CGTCGT +CGTCGC +CGTCGG +CGTGAA +CGTGAT +CGTGAC +CGTGAG +CGTGTA +CGTGTT +CGTGTC +CGTGTG +CGTGCA +CGTGCT +CGTGCC +CGTGCG +CGTGGA +CGTGGT +CGTGGC +CGTGGG +CGCAAA +CGCAAT +CGCAAC +CGCAAG +CGCATA +CGCATT +CGCATC +CGCATG +CGCACA +CGCACT +CGCACC +CGCACG +CGCAGA +CGCAGT +CGCAGC +CGCAGG +CGCTAA +CGCTAT +CGCTAC +CGCTAG +CGCTTA +CGCTTT +CGCTTC +CGCTTG +CGCTCA +CGCTCT +CGCTCC +CGCTCG +CGCTGA +CGCTGT +CGCTGC +CGCTGG +CGCCAA +CGCCAT +CGCCAC +CGCCAG +CGCCTA +CGCCTT +CGCCTC +CGCCTG +CGCCCA +CGCCCT +CGCCCC +CGCCCG +CGCCGA +CGCCGT +CGCCGC +CGCCGG +CGCGAA +CGCGAT +CGCGAC +CGCGAG +CGCGTA +CGCGTT +CGCGTC +CGCGTG +CGCGCA +CGCGCT +CGCGCC +CGCGCG +CGCGGA +CGCGGT +CGCGGC +CGCGGG +CGGAAA +CGGAAT +CGGAAC +CGGAAG +CGGATA +CGGATT +CGGATC +CGGATG +CGGACA +CGGACT +CGGACC +CGGACG +CGGAGA +CGGAGT +CGGAGC +CGGAGG +CGGTAA +CGGTAT +CGGTAC +CGGTAG +CGGTTA +CGGTTT +CGGTTC +CGGTTG +CGGTCA +CGGTCT +CGGTCC +CGGTCG +CGGTGA +CGGTGT +CGGTGC +CGGTGG +CGGCAA +CGGCAT +CGGCAC +CGGCAG +CGGCTA +CGGCTT +CGGCTC +CGGCTG +CGGCCA +CGGCCT +CGGCCC +CGGCCG +CGGCGA +CGGCGT +CGGCGC +CGGCGG +CGGGAA +CGGGAT +CGGGAC +CGGGAG +CGGGTA +CGGGTT +CGGGTC +CGGGTG +CGGGCA +CGGGCT +CGGGCC +CGGGCG +CGGGGA +CGGGGT +CGGGGC +CGGGGG +GAAAAA +GAAAAT +GAAAAC +GAAAAG +GAAATA +GAAATT +GAAATC +GAAATG +GAAACA +GAAACT +GAAACC +GAAACG +GAAAGA +GAAAGT +GAAAGC +GAAAGG +GAATAA +GAATAT +GAATAC +GAATAG +GAATTA +GAATTT +GAATTC +GAATTG +GAATCA +GAATCT +GAATCC +GAATCG +GAATGA +GAATGT +GAATGC +GAATGG +GAACAA +GAACAT +GAACAC +GAACAG +GAACTA +GAACTT +GAACTC +GAACTG +GAACCA +GAACCT +GAACCC +GAACCG +GAACGA +GAACGT +GAACGC +GAACGG +GAAGAA +GAAGAT +GAAGAC +GAAGAG +GAAGTA +GAAGTT +GAAGTC +GAAGTG +GAAGCA +GAAGCT +GAAGCC +GAAGCG +GAAGGA +GAAGGT +GAAGGC +GAAGGG +GATAAA +GATAAT +GATAAC +GATAAG +GATATA +GATATT +GATATC +GATATG +GATACA +GATACT +GATACC +GATACG +GATAGA +GATAGT +GATAGC +GATAGG +GATTAA +GATTAT +GATTAC +GATTAG +GATTTA +GATTTT +GATTTC +GATTTG +GATTCA +GATTCT +GATTCC +GATTCG +GATTGA +GATTGT +GATTGC +GATTGG +GATCAA +GATCAT +GATCAC +GATCAG +GATCTA +GATCTT +GATCTC +GATCTG +GATCCA +GATCCT +GATCCC +GATCCG +GATCGA +GATCGT +GATCGC +GATCGG +GATGAA +GATGAT +GATGAC +GATGAG +GATGTA +GATGTT +GATGTC +GATGTG +GATGCA +GATGCT +GATGCC +GATGCG +GATGGA +GATGGT +GATGGC +GATGGG +GACAAA +GACAAT +GACAAC +GACAAG +GACATA +GACATT +GACATC +GACATG +GACACA +GACACT +GACACC +GACACG +GACAGA +GACAGT +GACAGC +GACAGG +GACTAA +GACTAT +GACTAC +GACTAG +GACTTA +GACTTT +GACTTC +GACTTG +GACTCA +GACTCT +GACTCC +GACTCG +GACTGA +GACTGT +GACTGC +GACTGG +GACCAA +GACCAT +GACCAC +GACCAG +GACCTA +GACCTT +GACCTC +GACCTG +GACCCA +GACCCT +GACCCC +GACCCG +GACCGA +GACCGT +GACCGC +GACCGG +GACGAA +GACGAT +GACGAC +GACGAG +GACGTA +GACGTT +GACGTC +GACGTG +GACGCA +GACGCT +GACGCC +GACGCG +GACGGA +GACGGT +GACGGC +GACGGG +GAGAAA +GAGAAT +GAGAAC +GAGAAG +GAGATA +GAGATT +GAGATC +GAGATG +GAGACA +GAGACT +GAGACC +GAGACG +GAGAGA +GAGAGT +GAGAGC +GAGAGG +GAGTAA +GAGTAT +GAGTAC +GAGTAG +GAGTTA +GAGTTT +GAGTTC +GAGTTG +GAGTCA +GAGTCT +GAGTCC +GAGTCG +GAGTGA +GAGTGT +GAGTGC +GAGTGG +GAGCAA +GAGCAT +GAGCAC +GAGCAG +GAGCTA +GAGCTT +GAGCTC +GAGCTG +GAGCCA +GAGCCT +GAGCCC +GAGCCG +GAGCGA +GAGCGT +GAGCGC +GAGCGG +GAGGAA +GAGGAT +GAGGAC +GAGGAG +GAGGTA +GAGGTT +GAGGTC +GAGGTG +GAGGCA +GAGGCT +GAGGCC +GAGGCG +GAGGGA +GAGGGT +GAGGGC +GAGGGG +GTAAAA +GTAAAT +GTAAAC +GTAAAG +GTAATA +GTAATT +GTAATC +GTAATG +GTAACA +GTAACT +GTAACC +GTAACG +GTAAGA +GTAAGT +GTAAGC +GTAAGG +GTATAA +GTATAT +GTATAC +GTATAG +GTATTA +GTATTT +GTATTC +GTATTG +GTATCA +GTATCT +GTATCC +GTATCG +GTATGA +GTATGT +GTATGC +GTATGG +GTACAA +GTACAT +GTACAC +GTACAG +GTACTA +GTACTT +GTACTC +GTACTG +GTACCA +GTACCT +GTACCC +GTACCG +GTACGA +GTACGT +GTACGC +GTACGG +GTAGAA +GTAGAT +GTAGAC +GTAGAG +GTAGTA +GTAGTT +GTAGTC +GTAGTG +GTAGCA +GTAGCT +GTAGCC +GTAGCG +GTAGGA +GTAGGT +GTAGGC +GTAGGG +GTTAAA +GTTAAT +GTTAAC +GTTAAG +GTTATA +GTTATT +GTTATC +GTTATG +GTTACA +GTTACT +GTTACC +GTTACG +GTTAGA +GTTAGT +GTTAGC +GTTAGG +GTTTAA +GTTTAT +GTTTAC +GTTTAG +GTTTTA +GTTTTT +GTTTTC +GTTTTG +GTTTCA +GTTTCT +GTTTCC +GTTTCG +GTTTGA +GTTTGT +GTTTGC +GTTTGG +GTTCAA +GTTCAT +GTTCAC +GTTCAG +GTTCTA +GTTCTT +GTTCTC +GTTCTG +GTTCCA +GTTCCT +GTTCCC +GTTCCG +GTTCGA +GTTCGT +GTTCGC +GTTCGG +GTTGAA +GTTGAT +GTTGAC +GTTGAG +GTTGTA +GTTGTT +GTTGTC +GTTGTG +GTTGCA +GTTGCT +GTTGCC +GTTGCG +GTTGGA +GTTGGT +GTTGGC +GTTGGG +GTCAAA +GTCAAT +GTCAAC +GTCAAG +GTCATA +GTCATT +GTCATC +GTCATG +GTCACA +GTCACT +GTCACC +GTCACG +GTCAGA +GTCAGT +GTCAGC +GTCAGG +GTCTAA +GTCTAT +GTCTAC +GTCTAG +GTCTTA +GTCTTT +GTCTTC +GTCTTG +GTCTCA +GTCTCT +GTCTCC +GTCTCG +GTCTGA +GTCTGT +GTCTGC +GTCTGG +GTCCAA +GTCCAT +GTCCAC +GTCCAG +GTCCTA +GTCCTT +GTCCTC +GTCCTG +GTCCCA +GTCCCT +GTCCCC +GTCCCG +GTCCGA +GTCCGT +GTCCGC +GTCCGG +GTCGAA +GTCGAT +GTCGAC +GTCGAG +GTCGTA +GTCGTT +GTCGTC +GTCGTG +GTCGCA +GTCGCT +GTCGCC +GTCGCG +GTCGGA +GTCGGT +GTCGGC +GTCGGG +GTGAAA +GTGAAT +GTGAAC +GTGAAG +GTGATA +GTGATT +GTGATC +GTGATG +GTGACA +GTGACT +GTGACC +GTGACG +GTGAGA +GTGAGT +GTGAGC +GTGAGG +GTGTAA +GTGTAT +GTGTAC +GTGTAG +GTGTTA +GTGTTT +GTGTTC +GTGTTG +GTGTCA +GTGTCT +GTGTCC +GTGTCG +GTGTGA +GTGTGT +GTGTGC +GTGTGG +GTGCAA +GTGCAT +GTGCAC +GTGCAG +GTGCTA +GTGCTT +GTGCTC +GTGCTG +GTGCCA +GTGCCT +GTGCCC +GTGCCG +GTGCGA +GTGCGT +GTGCGC +GTGCGG +GTGGAA +GTGGAT +GTGGAC +GTGGAG +GTGGTA +GTGGTT +GTGGTC +GTGGTG +GTGGCA +GTGGCT +GTGGCC +GTGGCG +GTGGGA +GTGGGT +GTGGGC +GTGGGG +GCAAAA +GCAAAT +GCAAAC +GCAAAG +GCAATA +GCAATT +GCAATC +GCAATG +GCAACA +GCAACT +GCAACC +GCAACG +GCAAGA +GCAAGT +GCAAGC +GCAAGG +GCATAA +GCATAT +GCATAC +GCATAG +GCATTA +GCATTT +GCATTC +GCATTG +GCATCA +GCATCT +GCATCC +GCATCG +GCATGA +GCATGT +GCATGC +GCATGG +GCACAA +GCACAT +GCACAC +GCACAG +GCACTA +GCACTT +GCACTC +GCACTG +GCACCA +GCACCT +GCACCC +GCACCG +GCACGA +GCACGT +GCACGC +GCACGG +GCAGAA +GCAGAT +GCAGAC +GCAGAG +GCAGTA +GCAGTT +GCAGTC +GCAGTG +GCAGCA +GCAGCT +GCAGCC +GCAGCG +GCAGGA +GCAGGT +GCAGGC +GCAGGG +GCTAAA +GCTAAT +GCTAAC +GCTAAG +GCTATA +GCTATT +GCTATC +GCTATG +GCTACA +GCTACT +GCTACC +GCTACG +GCTAGA +GCTAGT +GCTAGC +GCTAGG +GCTTAA +GCTTAT +GCTTAC +GCTTAG +GCTTTA +GCTTTT +GCTTTC +GCTTTG +GCTTCA +GCTTCT +GCTTCC +GCTTCG +GCTTGA +GCTTGT +GCTTGC +GCTTGG +GCTCAA +GCTCAT +GCTCAC +GCTCAG +GCTCTA +GCTCTT +GCTCTC +GCTCTG +GCTCCA +GCTCCT +GCTCCC +GCTCCG +GCTCGA +GCTCGT +GCTCGC +GCTCGG +GCTGAA +GCTGAT +GCTGAC +GCTGAG +GCTGTA +GCTGTT +GCTGTC +GCTGTG +GCTGCA +GCTGCT +GCTGCC +GCTGCG +GCTGGA +GCTGGT +GCTGGC +GCTGGG +GCCAAA +GCCAAT +GCCAAC +GCCAAG +GCCATA +GCCATT +GCCATC +GCCATG +GCCACA +GCCACT +GCCACC +GCCACG +GCCAGA +GCCAGT +GCCAGC +GCCAGG +GCCTAA +GCCTAT +GCCTAC +GCCTAG +GCCTTA +GCCTTT +GCCTTC +GCCTTG +GCCTCA +GCCTCT +GCCTCC +GCCTCG +GCCTGA +GCCTGT +GCCTGC +GCCTGG +GCCCAA +GCCCAT +GCCCAC +GCCCAG +GCCCTA +GCCCTT +GCCCTC +GCCCTG +GCCCCA +GCCCCT +GCCCCC +GCCCCG +GCCCGA +GCCCGT +GCCCGC +GCCCGG +GCCGAA +GCCGAT +GCCGAC +GCCGAG +GCCGTA +GCCGTT +GCCGTC +GCCGTG +GCCGCA +GCCGCT +GCCGCC +GCCGCG +GCCGGA +GCCGGT +GCCGGC +GCCGGG +GCGAAA +GCGAAT +GCGAAC +GCGAAG +GCGATA +GCGATT +GCGATC +GCGATG +GCGACA +GCGACT +GCGACC +GCGACG +GCGAGA +GCGAGT +GCGAGC +GCGAGG +GCGTAA +GCGTAT +GCGTAC +GCGTAG +GCGTTA +GCGTTT +GCGTTC +GCGTTG +GCGTCA +GCGTCT +GCGTCC +GCGTCG +GCGTGA +GCGTGT +GCGTGC +GCGTGG +GCGCAA +GCGCAT +GCGCAC +GCGCAG +GCGCTA +GCGCTT +GCGCTC +GCGCTG +GCGCCA +GCGCCT +GCGCCC +GCGCCG +GCGCGA +GCGCGT +GCGCGC +GCGCGG +GCGGAA +GCGGAT +GCGGAC +GCGGAG +GCGGTA +GCGGTT +GCGGTC +GCGGTG +GCGGCA +GCGGCT +GCGGCC +GCGGCG +GCGGGA +GCGGGT +GCGGGC +GCGGGG +GGAAAA +GGAAAT +GGAAAC +GGAAAG +GGAATA +GGAATT +GGAATC +GGAATG +GGAACA +GGAACT +GGAACC +GGAACG +GGAAGA +GGAAGT +GGAAGC +GGAAGG +GGATAA +GGATAT +GGATAC +GGATAG +GGATTA +GGATTT +GGATTC +GGATTG +GGATCA +GGATCT +GGATCC +GGATCG +GGATGA +GGATGT +GGATGC +GGATGG +GGACAA +GGACAT +GGACAC +GGACAG +GGACTA +GGACTT +GGACTC +GGACTG +GGACCA +GGACCT +GGACCC +GGACCG +GGACGA +GGACGT +GGACGC +GGACGG +GGAGAA +GGAGAT +GGAGAC +GGAGAG +GGAGTA +GGAGTT +GGAGTC +GGAGTG +GGAGCA +GGAGCT +GGAGCC +GGAGCG +GGAGGA +GGAGGT +GGAGGC +GGAGGG +GGTAAA +GGTAAT +GGTAAC +GGTAAG +GGTATA +GGTATT +GGTATC +GGTATG +GGTACA +GGTACT +GGTACC +GGTACG +GGTAGA +GGTAGT +GGTAGC +GGTAGG +GGTTAA +GGTTAT +GGTTAC +GGTTAG +GGTTTA +GGTTTT +GGTTTC +GGTTTG +GGTTCA +GGTTCT +GGTTCC +GGTTCG +GGTTGA +GGTTGT +GGTTGC +GGTTGG +GGTCAA +GGTCAT +GGTCAC +GGTCAG +GGTCTA +GGTCTT +GGTCTC +GGTCTG +GGTCCA +GGTCCT +GGTCCC +GGTCCG +GGTCGA +GGTCGT +GGTCGC +GGTCGG +GGTGAA +GGTGAT +GGTGAC +GGTGAG +GGTGTA +GGTGTT +GGTGTC +GGTGTG +GGTGCA +GGTGCT +GGTGCC +GGTGCG +GGTGGA +GGTGGT +GGTGGC +GGTGGG +GGCAAA +GGCAAT +GGCAAC +GGCAAG +GGCATA +GGCATT +GGCATC +GGCATG +GGCACA +GGCACT +GGCACC +GGCACG +GGCAGA +GGCAGT +GGCAGC +GGCAGG +GGCTAA +GGCTAT +GGCTAC +GGCTAG +GGCTTA +GGCTTT +GGCTTC +GGCTTG +GGCTCA +GGCTCT +GGCTCC +GGCTCG +GGCTGA +GGCTGT +GGCTGC +GGCTGG +GGCCAA +GGCCAT +GGCCAC +GGCCAG +GGCCTA +GGCCTT +GGCCTC +GGCCTG +GGCCCA +GGCCCT +GGCCCC +GGCCCG +GGCCGA +GGCCGT +GGCCGC +GGCCGG +GGCGAA +GGCGAT +GGCGAC +GGCGAG +GGCGTA +GGCGTT +GGCGTC +GGCGTG +GGCGCA +GGCGCT +GGCGCC +GGCGCG +GGCGGA +GGCGGT +GGCGGC +GGCGGG +GGGAAA +GGGAAT +GGGAAC +GGGAAG +GGGATA +GGGATT +GGGATC +GGGATG +GGGACA +GGGACT +GGGACC +GGGACG +GGGAGA +GGGAGT +GGGAGC +GGGAGG +GGGTAA +GGGTAT +GGGTAC +GGGTAG +GGGTTA +GGGTTT +GGGTTC +GGGTTG +GGGTCA +GGGTCT +GGGTCC +GGGTCG +GGGTGA +GGGTGT +GGGTGC +GGGTGG +GGGCAA +GGGCAT +GGGCAC +GGGCAG +GGGCTA +GGGCTT +GGGCTC +GGGCTG +GGGCCA +GGGCCT +GGGCCC +GGGCCG +GGGCGA +GGGCGT +GGGCGC +GGGCGG +GGGGAA +GGGGAT +GGGGAC +GGGGAG +GGGGTA +GGGGTT +GGGGTC +GGGGTG +GGGGCA +GGGGCT +GGGGCC +GGGGCG +GGGGGA +GGGGGT +GGGGGC +GGGGGG diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/README.md b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..500fd976d92b217455f181bc98c77e612d9a18d9 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/README.md @@ -0,0 +1,202 @@ +--- +base_model: fabihamakhdoomi/TinyDNABERT +library_name: peft +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.12.0 \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/adapter_config.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..04c63e99bbc6b810449dece367c0afb165125b92 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/adapter_config.json @@ -0,0 +1,42 @@ +{ + "alpha_pattern": {}, + "auto_mapping": null, + "base_model_name_or_path": "fabihamakhdoomi/TinyDNABERT", + "bias": "none", + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 16, + "lora_dropout": 0.2, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": [ + "classifier", + "score" + ], + "peft_type": "LORA", + "r": 8, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "bert.encoder.layer.1.attention.self.query", + "bert.encoder.layer.0.output.dense", + "bert.encoder.layer.1.output.dense", + "bert.encoder.layer.0.intermediate.dense", + "bert.encoder.layer.1.intermediate.dense", + "bert.encoder.layer.0.attention.output.dense", + "bert.encoder.layer.1.attention.output.dense", + "bert.encoder.layer.0.attention.self.query", + "bert.encoder.layer.0.attention.self.value", + "bert.encoder.layer.1.attention.self.key", + "bert.encoder.layer.1.attention.self.value", + "bert.encoder.layer.0.attention.self.key" + ], + "task_type": "SEQ_CLS", + "use_dora": false, + "use_rslora": false +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/adapter_model.safetensors b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/adapter_model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..f171b29c98b5de9d58b3212fc3791071633de64b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/adapter_model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:35b94c7dbf49946a1012c5dad3f8501904a456ba9040ab32f83dff85302a60ce +size 151952 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/optimizer.pt b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..7d24b02b30ddabae1e12180be1d7090e492a8444 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6dda63930ad5b447a561f6cf9374ec9b47b804fd681c5db83663431a933d2e00 +size 312682 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/rng_state.pth b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..ab07d6382b5d2f619b3927fc83d6571e1bcccc54 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f0f0d7ddc5f130ac1babf491a193bd5c163e7a1b572c7a4eac20de52cd6b2bd7 +size 14244 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/scheduler.pt b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..8c9a6adac55ac4d1afd6e1b61889589acfe31330 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f3e454de3c0f52ddb8e5884c9783308658a6f61ede2ed6e377ca9a68d47c4515 +size 1064 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/special_tokens_map.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..a8b3208c2884c4efb86e49300fdd3dc877220cdf --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/special_tokens_map.json @@ -0,0 +1,7 @@ +{ + "cls_token": "[CLS]", + "mask_token": "[MASK]", + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/tokenizer.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..ebd782de6963b9032ce9374cfe30f5275be1a30b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/tokenizer.json @@ -0,0 +1,4265 @@ +{ + "version": "1.0", + "truncation": { + "direction": "Right", + "max_length": 512, + "strategy": "LongestFirst", + "stride": 0 + }, + "padding": { + "strategy": { + "Fixed": 512 + }, + "direction": "Right", + "pad_to_multiple_of": null, + "pad_id": 0, + "pad_type_id": 0, + "pad_token": "[PAD]" + }, + "added_tokens": [ + { + "id": 0, + "content": "[PAD]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 1, + "content": "[UNK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 2, + "content": "[CLS]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 3, + "content": "[SEP]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 4, + "content": "[MASK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + } + ], + "normalizer": { + "type": "BertNormalizer", + "clean_text": true, + "handle_chinese_chars": true, + "strip_accents": null, + "lowercase": false + }, + "pre_tokenizer": { + "type": "BertPreTokenizer" + }, + "post_processor": { + "type": "TemplateProcessing", + "single": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + } + ], + "pair": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "B", + "type_id": 1 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 1 + } + } + ], + "special_tokens": { + "[CLS]": { + "id": "[CLS]", + "ids": [ + 2 + ], + "tokens": [ + "[CLS]" + ] + }, + "[SEP]": { + "id": "[SEP]", + "ids": [ + 3 + ], + "tokens": [ + "[SEP]" + ] + } + } + }, + "decoder": { + "type": "WordPiece", + "prefix": "##", + "cleanup": true + }, + "model": { + "type": "WordPiece", + "unk_token": "[UNK]", + "continuing_subword_prefix": "##", + "max_input_chars_per_word": 100, + "vocab": { + "[PAD]": 0, + "[UNK]": 1, + "[CLS]": 2, + "[SEP]": 3, + "[MASK]": 4, + "AAAAAA": 5, + "AAAAAT": 6, + "AAAAAC": 7, + "AAAAAG": 8, + "AAAATA": 9, + "AAAATT": 10, + "AAAATC": 11, + "AAAATG": 12, + "AAAACA": 13, + "AAAACT": 14, + "AAAACC": 15, + "AAAACG": 16, + "AAAAGA": 17, + "AAAAGT": 18, + "AAAAGC": 19, + "AAAAGG": 20, + "AAATAA": 21, + "AAATAT": 22, + "AAATAC": 23, + "AAATAG": 24, + "AAATTA": 25, + "AAATTT": 26, + "AAATTC": 27, + "AAATTG": 28, + "AAATCA": 29, + "AAATCT": 30, + "AAATCC": 31, + "AAATCG": 32, + "AAATGA": 33, + "AAATGT": 34, + "AAATGC": 35, + "AAATGG": 36, + "AAACAA": 37, + "AAACAT": 38, + "AAACAC": 39, + "AAACAG": 40, + "AAACTA": 41, + "AAACTT": 42, + "AAACTC": 43, + "AAACTG": 44, + "AAACCA": 45, + "AAACCT": 46, + "AAACCC": 47, + "AAACCG": 48, + "AAACGA": 49, + "AAACGT": 50, + "AAACGC": 51, + "AAACGG": 52, + "AAAGAA": 53, + "AAAGAT": 54, + "AAAGAC": 55, + "AAAGAG": 56, + "AAAGTA": 57, + "AAAGTT": 58, + "AAAGTC": 59, + "AAAGTG": 60, + "AAAGCA": 61, + "AAAGCT": 62, + "AAAGCC": 63, + "AAAGCG": 64, + "AAAGGA": 65, + "AAAGGT": 66, + "AAAGGC": 67, + "AAAGGG": 68, + "AATAAA": 69, + "AATAAT": 70, + "AATAAC": 71, + "AATAAG": 72, + "AATATA": 73, + "AATATT": 74, + "AATATC": 75, + "AATATG": 76, + "AATACA": 77, + "AATACT": 78, + "AATACC": 79, + "AATACG": 80, + "AATAGA": 81, + "AATAGT": 82, + "AATAGC": 83, + "AATAGG": 84, + "AATTAA": 85, + "AATTAT": 86, + "AATTAC": 87, + "AATTAG": 88, + "AATTTA": 89, + "AATTTT": 90, + "AATTTC": 91, + "AATTTG": 92, + "AATTCA": 93, + "AATTCT": 94, + "AATTCC": 95, + "AATTCG": 96, + "AATTGA": 97, + "AATTGT": 98, + "AATTGC": 99, + "AATTGG": 100, + "AATCAA": 101, + "AATCAT": 102, + "AATCAC": 103, + "AATCAG": 104, + "AATCTA": 105, + "AATCTT": 106, + "AATCTC": 107, + "AATCTG": 108, + "AATCCA": 109, + "AATCCT": 110, + "AATCCC": 111, + "AATCCG": 112, + "AATCGA": 113, + "AATCGT": 114, + "AATCGC": 115, + "AATCGG": 116, + "AATGAA": 117, + "AATGAT": 118, + "AATGAC": 119, + "AATGAG": 120, + "AATGTA": 121, + "AATGTT": 122, + "AATGTC": 123, + "AATGTG": 124, + "AATGCA": 125, + "AATGCT": 126, + "AATGCC": 127, + "AATGCG": 128, + "AATGGA": 129, + "AATGGT": 130, + "AATGGC": 131, + "AATGGG": 132, + "AACAAA": 133, + "AACAAT": 134, + "AACAAC": 135, + "AACAAG": 136, + "AACATA": 137, + "AACATT": 138, + "AACATC": 139, + "AACATG": 140, + "AACACA": 141, + "AACACT": 142, + "AACACC": 143, + "AACACG": 144, + "AACAGA": 145, + "AACAGT": 146, + "AACAGC": 147, + "AACAGG": 148, + "AACTAA": 149, + "AACTAT": 150, + "AACTAC": 151, + "AACTAG": 152, + "AACTTA": 153, + "AACTTT": 154, + "AACTTC": 155, + "AACTTG": 156, + "AACTCA": 157, + "AACTCT": 158, + "AACTCC": 159, + "AACTCG": 160, + "AACTGA": 161, + "AACTGT": 162, + "AACTGC": 163, + "AACTGG": 164, + "AACCAA": 165, + "AACCAT": 166, + "AACCAC": 167, + "AACCAG": 168, + "AACCTA": 169, + "AACCTT": 170, + "AACCTC": 171, + "AACCTG": 172, + "AACCCA": 173, + "AACCCT": 174, + "AACCCC": 175, + "AACCCG": 176, + "AACCGA": 177, + "AACCGT": 178, + "AACCGC": 179, + "AACCGG": 180, + "AACGAA": 181, + "AACGAT": 182, + "AACGAC": 183, + "AACGAG": 184, + "AACGTA": 185, + "AACGTT": 186, + "AACGTC": 187, + "AACGTG": 188, + "AACGCA": 189, + "AACGCT": 190, + "AACGCC": 191, + "AACGCG": 192, + "AACGGA": 193, + "AACGGT": 194, + "AACGGC": 195, + "AACGGG": 196, + "AAGAAA": 197, + "AAGAAT": 198, + "AAGAAC": 199, + "AAGAAG": 200, + "AAGATA": 201, + "AAGATT": 202, + "AAGATC": 203, + "AAGATG": 204, + "AAGACA": 205, + "AAGACT": 206, + "AAGACC": 207, + "AAGACG": 208, + "AAGAGA": 209, + "AAGAGT": 210, + "AAGAGC": 211, + "AAGAGG": 212, + "AAGTAA": 213, + "AAGTAT": 214, + "AAGTAC": 215, + "AAGTAG": 216, + "AAGTTA": 217, + "AAGTTT": 218, + "AAGTTC": 219, + "AAGTTG": 220, + "AAGTCA": 221, + "AAGTCT": 222, + "AAGTCC": 223, + "AAGTCG": 224, + "AAGTGA": 225, + "AAGTGT": 226, + "AAGTGC": 227, + "AAGTGG": 228, + "AAGCAA": 229, + "AAGCAT": 230, + "AAGCAC": 231, + "AAGCAG": 232, + "AAGCTA": 233, + "AAGCTT": 234, + "AAGCTC": 235, + "AAGCTG": 236, + "AAGCCA": 237, + "AAGCCT": 238, + "AAGCCC": 239, + "AAGCCG": 240, + "AAGCGA": 241, + "AAGCGT": 242, + "AAGCGC": 243, + "AAGCGG": 244, + "AAGGAA": 245, + "AAGGAT": 246, + "AAGGAC": 247, + "AAGGAG": 248, + "AAGGTA": 249, + "AAGGTT": 250, + "AAGGTC": 251, + "AAGGTG": 252, + "AAGGCA": 253, + "AAGGCT": 254, + "AAGGCC": 255, + "AAGGCG": 256, + "AAGGGA": 257, + "AAGGGT": 258, + "AAGGGC": 259, + "AAGGGG": 260, + "ATAAAA": 261, + "ATAAAT": 262, + "ATAAAC": 263, + "ATAAAG": 264, + "ATAATA": 265, + "ATAATT": 266, + "ATAATC": 267, + "ATAATG": 268, + "ATAACA": 269, + "ATAACT": 270, + "ATAACC": 271, + "ATAACG": 272, + "ATAAGA": 273, + "ATAAGT": 274, + "ATAAGC": 275, + "ATAAGG": 276, + "ATATAA": 277, + "ATATAT": 278, + "ATATAC": 279, + "ATATAG": 280, + "ATATTA": 281, + "ATATTT": 282, + "ATATTC": 283, + "ATATTG": 284, + "ATATCA": 285, + "ATATCT": 286, + "ATATCC": 287, + "ATATCG": 288, + "ATATGA": 289, + "ATATGT": 290, + "ATATGC": 291, + "ATATGG": 292, + "ATACAA": 293, + "ATACAT": 294, + "ATACAC": 295, + "ATACAG": 296, + "ATACTA": 297, + "ATACTT": 298, + "ATACTC": 299, + "ATACTG": 300, + "ATACCA": 301, + "ATACCT": 302, + "ATACCC": 303, + "ATACCG": 304, + "ATACGA": 305, + "ATACGT": 306, + "ATACGC": 307, + "ATACGG": 308, + "ATAGAA": 309, + "ATAGAT": 310, + "ATAGAC": 311, + "ATAGAG": 312, + "ATAGTA": 313, + "ATAGTT": 314, + "ATAGTC": 315, + "ATAGTG": 316, + "ATAGCA": 317, + "ATAGCT": 318, + "ATAGCC": 319, + "ATAGCG": 320, + "ATAGGA": 321, + "ATAGGT": 322, + "ATAGGC": 323, + "ATAGGG": 324, + "ATTAAA": 325, + "ATTAAT": 326, + "ATTAAC": 327, + "ATTAAG": 328, + "ATTATA": 329, + "ATTATT": 330, + "ATTATC": 331, + "ATTATG": 332, + "ATTACA": 333, + "ATTACT": 334, + "ATTACC": 335, + "ATTACG": 336, + "ATTAGA": 337, + "ATTAGT": 338, + "ATTAGC": 339, + "ATTAGG": 340, + "ATTTAA": 341, + "ATTTAT": 342, + "ATTTAC": 343, + "ATTTAG": 344, + "ATTTTA": 345, + "ATTTTT": 346, + "ATTTTC": 347, + "ATTTTG": 348, + "ATTTCA": 349, + "ATTTCT": 350, + "ATTTCC": 351, + "ATTTCG": 352, + "ATTTGA": 353, + "ATTTGT": 354, + "ATTTGC": 355, + "ATTTGG": 356, + "ATTCAA": 357, + "ATTCAT": 358, + "ATTCAC": 359, + "ATTCAG": 360, + "ATTCTA": 361, + "ATTCTT": 362, + "ATTCTC": 363, + "ATTCTG": 364, + "ATTCCA": 365, + "ATTCCT": 366, + "ATTCCC": 367, + "ATTCCG": 368, + "ATTCGA": 369, + "ATTCGT": 370, + "ATTCGC": 371, + "ATTCGG": 372, + "ATTGAA": 373, + "ATTGAT": 374, + "ATTGAC": 375, + "ATTGAG": 376, + "ATTGTA": 377, + "ATTGTT": 378, + "ATTGTC": 379, + "ATTGTG": 380, + "ATTGCA": 381, + "ATTGCT": 382, + "ATTGCC": 383, + "ATTGCG": 384, + "ATTGGA": 385, + "ATTGGT": 386, + "ATTGGC": 387, + "ATTGGG": 388, + "ATCAAA": 389, + "ATCAAT": 390, + "ATCAAC": 391, + "ATCAAG": 392, + "ATCATA": 393, + "ATCATT": 394, + "ATCATC": 395, + "ATCATG": 396, + "ATCACA": 397, + "ATCACT": 398, + "ATCACC": 399, + "ATCACG": 400, + "ATCAGA": 401, + "ATCAGT": 402, + "ATCAGC": 403, + "ATCAGG": 404, + "ATCTAA": 405, + "ATCTAT": 406, + "ATCTAC": 407, + "ATCTAG": 408, + "ATCTTA": 409, + "ATCTTT": 410, + "ATCTTC": 411, + "ATCTTG": 412, + "ATCTCA": 413, + "ATCTCT": 414, + "ATCTCC": 415, + "ATCTCG": 416, + "ATCTGA": 417, + "ATCTGT": 418, + "ATCTGC": 419, + "ATCTGG": 420, + "ATCCAA": 421, + "ATCCAT": 422, + "ATCCAC": 423, + "ATCCAG": 424, + "ATCCTA": 425, + "ATCCTT": 426, + "ATCCTC": 427, + "ATCCTG": 428, + "ATCCCA": 429, + "ATCCCT": 430, + "ATCCCC": 431, + "ATCCCG": 432, + "ATCCGA": 433, + "ATCCGT": 434, + "ATCCGC": 435, + "ATCCGG": 436, + "ATCGAA": 437, + "ATCGAT": 438, + "ATCGAC": 439, + "ATCGAG": 440, + "ATCGTA": 441, + "ATCGTT": 442, + "ATCGTC": 443, + "ATCGTG": 444, + "ATCGCA": 445, + "ATCGCT": 446, + "ATCGCC": 447, + "ATCGCG": 448, + "ATCGGA": 449, + "ATCGGT": 450, + "ATCGGC": 451, + "ATCGGG": 452, + "ATGAAA": 453, + "ATGAAT": 454, + "ATGAAC": 455, + "ATGAAG": 456, + "ATGATA": 457, + "ATGATT": 458, + "ATGATC": 459, + "ATGATG": 460, + "ATGACA": 461, + "ATGACT": 462, + "ATGACC": 463, + "ATGACG": 464, + "ATGAGA": 465, + "ATGAGT": 466, + "ATGAGC": 467, + "ATGAGG": 468, + "ATGTAA": 469, + "ATGTAT": 470, + "ATGTAC": 471, + "ATGTAG": 472, + "ATGTTA": 473, + "ATGTTT": 474, + "ATGTTC": 475, + "ATGTTG": 476, + "ATGTCA": 477, + "ATGTCT": 478, + "ATGTCC": 479, + "ATGTCG": 480, + "ATGTGA": 481, + "ATGTGT": 482, + "ATGTGC": 483, + "ATGTGG": 484, + "ATGCAA": 485, + "ATGCAT": 486, + "ATGCAC": 487, + "ATGCAG": 488, + "ATGCTA": 489, + "ATGCTT": 490, + "ATGCTC": 491, + "ATGCTG": 492, + "ATGCCA": 493, + "ATGCCT": 494, + "ATGCCC": 495, + "ATGCCG": 496, + "ATGCGA": 497, + "ATGCGT": 498, + "ATGCGC": 499, + "ATGCGG": 500, + "ATGGAA": 501, + "ATGGAT": 502, + "ATGGAC": 503, + "ATGGAG": 504, + "ATGGTA": 505, + "ATGGTT": 506, + "ATGGTC": 507, + "ATGGTG": 508, + "ATGGCA": 509, + "ATGGCT": 510, + "ATGGCC": 511, + "ATGGCG": 512, + "ATGGGA": 513, + "ATGGGT": 514, + "ATGGGC": 515, + "ATGGGG": 516, + "ACAAAA": 517, + "ACAAAT": 518, + "ACAAAC": 519, + "ACAAAG": 520, + "ACAATA": 521, + "ACAATT": 522, + "ACAATC": 523, + "ACAATG": 524, + "ACAACA": 525, + "ACAACT": 526, + "ACAACC": 527, + "ACAACG": 528, + "ACAAGA": 529, + "ACAAGT": 530, + "ACAAGC": 531, + "ACAAGG": 532, + "ACATAA": 533, + "ACATAT": 534, + "ACATAC": 535, + "ACATAG": 536, + "ACATTA": 537, + "ACATTT": 538, + "ACATTC": 539, + "ACATTG": 540, + "ACATCA": 541, + "ACATCT": 542, + "ACATCC": 543, + "ACATCG": 544, + "ACATGA": 545, + "ACATGT": 546, + "ACATGC": 547, + "ACATGG": 548, + "ACACAA": 549, + "ACACAT": 550, + "ACACAC": 551, + "ACACAG": 552, + "ACACTA": 553, + "ACACTT": 554, + "ACACTC": 555, + "ACACTG": 556, + "ACACCA": 557, + "ACACCT": 558, + "ACACCC": 559, + "ACACCG": 560, + "ACACGA": 561, + "ACACGT": 562, + "ACACGC": 563, + "ACACGG": 564, + "ACAGAA": 565, + "ACAGAT": 566, + "ACAGAC": 567, + "ACAGAG": 568, + "ACAGTA": 569, + "ACAGTT": 570, + "ACAGTC": 571, + "ACAGTG": 572, + "ACAGCA": 573, + "ACAGCT": 574, + "ACAGCC": 575, + "ACAGCG": 576, + "ACAGGA": 577, + "ACAGGT": 578, + "ACAGGC": 579, + "ACAGGG": 580, + "ACTAAA": 581, + "ACTAAT": 582, + "ACTAAC": 583, + "ACTAAG": 584, + "ACTATA": 585, + "ACTATT": 586, + "ACTATC": 587, + "ACTATG": 588, + "ACTACA": 589, + "ACTACT": 590, + "ACTACC": 591, + "ACTACG": 592, + "ACTAGA": 593, + "ACTAGT": 594, + "ACTAGC": 595, + "ACTAGG": 596, + "ACTTAA": 597, + "ACTTAT": 598, + "ACTTAC": 599, + "ACTTAG": 600, + "ACTTTA": 601, + "ACTTTT": 602, + "ACTTTC": 603, + "ACTTTG": 604, + "ACTTCA": 605, + "ACTTCT": 606, + "ACTTCC": 607, + "ACTTCG": 608, + "ACTTGA": 609, + "ACTTGT": 610, + "ACTTGC": 611, + "ACTTGG": 612, + "ACTCAA": 613, + "ACTCAT": 614, + "ACTCAC": 615, + "ACTCAG": 616, + "ACTCTA": 617, + "ACTCTT": 618, + "ACTCTC": 619, + "ACTCTG": 620, + "ACTCCA": 621, + "ACTCCT": 622, + "ACTCCC": 623, + "ACTCCG": 624, + "ACTCGA": 625, + "ACTCGT": 626, + "ACTCGC": 627, + "ACTCGG": 628, + "ACTGAA": 629, + "ACTGAT": 630, + "ACTGAC": 631, + "ACTGAG": 632, + "ACTGTA": 633, + "ACTGTT": 634, + "ACTGTC": 635, + "ACTGTG": 636, + "ACTGCA": 637, + "ACTGCT": 638, + "ACTGCC": 639, + "ACTGCG": 640, + "ACTGGA": 641, + "ACTGGT": 642, + "ACTGGC": 643, + "ACTGGG": 644, + "ACCAAA": 645, + "ACCAAT": 646, + "ACCAAC": 647, + "ACCAAG": 648, + "ACCATA": 649, + "ACCATT": 650, + "ACCATC": 651, + "ACCATG": 652, + "ACCACA": 653, + "ACCACT": 654, + "ACCACC": 655, + "ACCACG": 656, + "ACCAGA": 657, + "ACCAGT": 658, + "ACCAGC": 659, + "ACCAGG": 660, + "ACCTAA": 661, + "ACCTAT": 662, + "ACCTAC": 663, + "ACCTAG": 664, + "ACCTTA": 665, + "ACCTTT": 666, + "ACCTTC": 667, + "ACCTTG": 668, + "ACCTCA": 669, + "ACCTCT": 670, + "ACCTCC": 671, + "ACCTCG": 672, + "ACCTGA": 673, + "ACCTGT": 674, + "ACCTGC": 675, + "ACCTGG": 676, + "ACCCAA": 677, + "ACCCAT": 678, + "ACCCAC": 679, + "ACCCAG": 680, + "ACCCTA": 681, + "ACCCTT": 682, + "ACCCTC": 683, + "ACCCTG": 684, + "ACCCCA": 685, + "ACCCCT": 686, + "ACCCCC": 687, + "ACCCCG": 688, + "ACCCGA": 689, + "ACCCGT": 690, + "ACCCGC": 691, + "ACCCGG": 692, + "ACCGAA": 693, + "ACCGAT": 694, + "ACCGAC": 695, + "ACCGAG": 696, + "ACCGTA": 697, + "ACCGTT": 698, + "ACCGTC": 699, + "ACCGTG": 700, + "ACCGCA": 701, + "ACCGCT": 702, + "ACCGCC": 703, + "ACCGCG": 704, + "ACCGGA": 705, + "ACCGGT": 706, + "ACCGGC": 707, + "ACCGGG": 708, + "ACGAAA": 709, + "ACGAAT": 710, + "ACGAAC": 711, + "ACGAAG": 712, + "ACGATA": 713, + "ACGATT": 714, + "ACGATC": 715, + "ACGATG": 716, + "ACGACA": 717, + "ACGACT": 718, + "ACGACC": 719, + "ACGACG": 720, + "ACGAGA": 721, + "ACGAGT": 722, + "ACGAGC": 723, + "ACGAGG": 724, + "ACGTAA": 725, + "ACGTAT": 726, + "ACGTAC": 727, + "ACGTAG": 728, + "ACGTTA": 729, + "ACGTTT": 730, + "ACGTTC": 731, + "ACGTTG": 732, + "ACGTCA": 733, + "ACGTCT": 734, + "ACGTCC": 735, + "ACGTCG": 736, + "ACGTGA": 737, + "ACGTGT": 738, + "ACGTGC": 739, + "ACGTGG": 740, + "ACGCAA": 741, + "ACGCAT": 742, + "ACGCAC": 743, + "ACGCAG": 744, + "ACGCTA": 745, + "ACGCTT": 746, + "ACGCTC": 747, + "ACGCTG": 748, + "ACGCCA": 749, + "ACGCCT": 750, + "ACGCCC": 751, + "ACGCCG": 752, + "ACGCGA": 753, + "ACGCGT": 754, + "ACGCGC": 755, + "ACGCGG": 756, + "ACGGAA": 757, + "ACGGAT": 758, + "ACGGAC": 759, + "ACGGAG": 760, + "ACGGTA": 761, + "ACGGTT": 762, + "ACGGTC": 763, + "ACGGTG": 764, + "ACGGCA": 765, + "ACGGCT": 766, + "ACGGCC": 767, + "ACGGCG": 768, + "ACGGGA": 769, + "ACGGGT": 770, + "ACGGGC": 771, + "ACGGGG": 772, + "AGAAAA": 773, + "AGAAAT": 774, + "AGAAAC": 775, + "AGAAAG": 776, + "AGAATA": 777, + "AGAATT": 778, + "AGAATC": 779, + "AGAATG": 780, + "AGAACA": 781, + "AGAACT": 782, + "AGAACC": 783, + "AGAACG": 784, + "AGAAGA": 785, + "AGAAGT": 786, + "AGAAGC": 787, + "AGAAGG": 788, + "AGATAA": 789, + "AGATAT": 790, + "AGATAC": 791, + "AGATAG": 792, + "AGATTA": 793, + "AGATTT": 794, + "AGATTC": 795, + "AGATTG": 796, + "AGATCA": 797, + "AGATCT": 798, + "AGATCC": 799, + "AGATCG": 800, + "AGATGA": 801, + "AGATGT": 802, + "AGATGC": 803, + "AGATGG": 804, + "AGACAA": 805, + "AGACAT": 806, + "AGACAC": 807, + "AGACAG": 808, + "AGACTA": 809, + "AGACTT": 810, + "AGACTC": 811, + "AGACTG": 812, + "AGACCA": 813, + "AGACCT": 814, + "AGACCC": 815, + "AGACCG": 816, + "AGACGA": 817, + "AGACGT": 818, + "AGACGC": 819, + "AGACGG": 820, + "AGAGAA": 821, + "AGAGAT": 822, + "AGAGAC": 823, + "AGAGAG": 824, + "AGAGTA": 825, + "AGAGTT": 826, + "AGAGTC": 827, + "AGAGTG": 828, + "AGAGCA": 829, + "AGAGCT": 830, + "AGAGCC": 831, + "AGAGCG": 832, + "AGAGGA": 833, + "AGAGGT": 834, + "AGAGGC": 835, + "AGAGGG": 836, + "AGTAAA": 837, + "AGTAAT": 838, + "AGTAAC": 839, + "AGTAAG": 840, + "AGTATA": 841, + "AGTATT": 842, + "AGTATC": 843, + "AGTATG": 844, + "AGTACA": 845, + "AGTACT": 846, + "AGTACC": 847, + "AGTACG": 848, + "AGTAGA": 849, + "AGTAGT": 850, + "AGTAGC": 851, + "AGTAGG": 852, + "AGTTAA": 853, + "AGTTAT": 854, + "AGTTAC": 855, + "AGTTAG": 856, + "AGTTTA": 857, + "AGTTTT": 858, + "AGTTTC": 859, + "AGTTTG": 860, + "AGTTCA": 861, + "AGTTCT": 862, + "AGTTCC": 863, + "AGTTCG": 864, + "AGTTGA": 865, + "AGTTGT": 866, + "AGTTGC": 867, + "AGTTGG": 868, + "AGTCAA": 869, + "AGTCAT": 870, + "AGTCAC": 871, + "AGTCAG": 872, + "AGTCTA": 873, + "AGTCTT": 874, + "AGTCTC": 875, + "AGTCTG": 876, + "AGTCCA": 877, + "AGTCCT": 878, + "AGTCCC": 879, + "AGTCCG": 880, + "AGTCGA": 881, + "AGTCGT": 882, + "AGTCGC": 883, + "AGTCGG": 884, + "AGTGAA": 885, + "AGTGAT": 886, + "AGTGAC": 887, + "AGTGAG": 888, + "AGTGTA": 889, + "AGTGTT": 890, + "AGTGTC": 891, + "AGTGTG": 892, + "AGTGCA": 893, + "AGTGCT": 894, + "AGTGCC": 895, + "AGTGCG": 896, + "AGTGGA": 897, + "AGTGGT": 898, + "AGTGGC": 899, + "AGTGGG": 900, + "AGCAAA": 901, + "AGCAAT": 902, + "AGCAAC": 903, + "AGCAAG": 904, + "AGCATA": 905, + "AGCATT": 906, + "AGCATC": 907, + "AGCATG": 908, + "AGCACA": 909, + "AGCACT": 910, + "AGCACC": 911, + "AGCACG": 912, + "AGCAGA": 913, + "AGCAGT": 914, + "AGCAGC": 915, + "AGCAGG": 916, + "AGCTAA": 917, + "AGCTAT": 918, + "AGCTAC": 919, + "AGCTAG": 920, + "AGCTTA": 921, + "AGCTTT": 922, + "AGCTTC": 923, + "AGCTTG": 924, + "AGCTCA": 925, + "AGCTCT": 926, + "AGCTCC": 927, + "AGCTCG": 928, + "AGCTGA": 929, + "AGCTGT": 930, + "AGCTGC": 931, + "AGCTGG": 932, + "AGCCAA": 933, + "AGCCAT": 934, + "AGCCAC": 935, + "AGCCAG": 936, + "AGCCTA": 937, + "AGCCTT": 938, + "AGCCTC": 939, + "AGCCTG": 940, + "AGCCCA": 941, + "AGCCCT": 942, + "AGCCCC": 943, + "AGCCCG": 944, + "AGCCGA": 945, + "AGCCGT": 946, + "AGCCGC": 947, + "AGCCGG": 948, + "AGCGAA": 949, + "AGCGAT": 950, + "AGCGAC": 951, + "AGCGAG": 952, + "AGCGTA": 953, + "AGCGTT": 954, + "AGCGTC": 955, + "AGCGTG": 956, + "AGCGCA": 957, + "AGCGCT": 958, + "AGCGCC": 959, + "AGCGCG": 960, + "AGCGGA": 961, + "AGCGGT": 962, + "AGCGGC": 963, + "AGCGGG": 964, + "AGGAAA": 965, + "AGGAAT": 966, + "AGGAAC": 967, + "AGGAAG": 968, + "AGGATA": 969, + "AGGATT": 970, + "AGGATC": 971, + "AGGATG": 972, + "AGGACA": 973, + "AGGACT": 974, + "AGGACC": 975, + "AGGACG": 976, + "AGGAGA": 977, + "AGGAGT": 978, + "AGGAGC": 979, + "AGGAGG": 980, + "AGGTAA": 981, + "AGGTAT": 982, + "AGGTAC": 983, + "AGGTAG": 984, + "AGGTTA": 985, + "AGGTTT": 986, + "AGGTTC": 987, + "AGGTTG": 988, + "AGGTCA": 989, + "AGGTCT": 990, + "AGGTCC": 991, + "AGGTCG": 992, + "AGGTGA": 993, + "AGGTGT": 994, + "AGGTGC": 995, + "AGGTGG": 996, + "AGGCAA": 997, + "AGGCAT": 998, + "AGGCAC": 999, + "AGGCAG": 1000, + "AGGCTA": 1001, + "AGGCTT": 1002, + "AGGCTC": 1003, + "AGGCTG": 1004, + "AGGCCA": 1005, + "AGGCCT": 1006, + "AGGCCC": 1007, + "AGGCCG": 1008, + "AGGCGA": 1009, + "AGGCGT": 1010, + "AGGCGC": 1011, + "AGGCGG": 1012, + "AGGGAA": 1013, + "AGGGAT": 1014, + "AGGGAC": 1015, + "AGGGAG": 1016, + "AGGGTA": 1017, + "AGGGTT": 1018, + "AGGGTC": 1019, + "AGGGTG": 1020, + "AGGGCA": 1021, + "AGGGCT": 1022, + "AGGGCC": 1023, + "AGGGCG": 1024, + "AGGGGA": 1025, + "AGGGGT": 1026, + "AGGGGC": 1027, + "AGGGGG": 1028, + "TAAAAA": 1029, + "TAAAAT": 1030, + "TAAAAC": 1031, + "TAAAAG": 1032, + "TAAATA": 1033, + "TAAATT": 1034, + "TAAATC": 1035, + "TAAATG": 1036, + "TAAACA": 1037, + "TAAACT": 1038, + "TAAACC": 1039, + "TAAACG": 1040, + "TAAAGA": 1041, + "TAAAGT": 1042, + "TAAAGC": 1043, + "TAAAGG": 1044, + "TAATAA": 1045, + "TAATAT": 1046, + "TAATAC": 1047, + "TAATAG": 1048, + "TAATTA": 1049, + "TAATTT": 1050, + "TAATTC": 1051, + "TAATTG": 1052, + "TAATCA": 1053, + "TAATCT": 1054, + "TAATCC": 1055, + "TAATCG": 1056, + "TAATGA": 1057, + "TAATGT": 1058, + "TAATGC": 1059, + "TAATGG": 1060, + "TAACAA": 1061, + "TAACAT": 1062, + "TAACAC": 1063, + "TAACAG": 1064, + "TAACTA": 1065, + "TAACTT": 1066, + "TAACTC": 1067, + "TAACTG": 1068, + "TAACCA": 1069, + "TAACCT": 1070, + "TAACCC": 1071, + "TAACCG": 1072, + "TAACGA": 1073, + "TAACGT": 1074, + "TAACGC": 1075, + "TAACGG": 1076, + "TAAGAA": 1077, + "TAAGAT": 1078, + "TAAGAC": 1079, + "TAAGAG": 1080, + "TAAGTA": 1081, + "TAAGTT": 1082, + "TAAGTC": 1083, + "TAAGTG": 1084, + "TAAGCA": 1085, + "TAAGCT": 1086, + "TAAGCC": 1087, + "TAAGCG": 1088, + "TAAGGA": 1089, + "TAAGGT": 1090, + "TAAGGC": 1091, + "TAAGGG": 1092, + "TATAAA": 1093, + "TATAAT": 1094, + "TATAAC": 1095, + "TATAAG": 1096, + "TATATA": 1097, + "TATATT": 1098, + "TATATC": 1099, + "TATATG": 1100, + "TATACA": 1101, + "TATACT": 1102, + "TATACC": 1103, + "TATACG": 1104, + "TATAGA": 1105, + "TATAGT": 1106, + "TATAGC": 1107, + "TATAGG": 1108, + "TATTAA": 1109, + "TATTAT": 1110, + "TATTAC": 1111, + "TATTAG": 1112, + "TATTTA": 1113, + "TATTTT": 1114, + "TATTTC": 1115, + "TATTTG": 1116, + "TATTCA": 1117, + "TATTCT": 1118, + "TATTCC": 1119, + "TATTCG": 1120, + "TATTGA": 1121, + "TATTGT": 1122, + "TATTGC": 1123, + "TATTGG": 1124, + "TATCAA": 1125, + "TATCAT": 1126, + "TATCAC": 1127, + "TATCAG": 1128, + "TATCTA": 1129, + "TATCTT": 1130, + "TATCTC": 1131, + "TATCTG": 1132, + "TATCCA": 1133, + "TATCCT": 1134, + "TATCCC": 1135, + "TATCCG": 1136, + "TATCGA": 1137, + "TATCGT": 1138, + "TATCGC": 1139, + "TATCGG": 1140, + "TATGAA": 1141, + "TATGAT": 1142, + "TATGAC": 1143, + "TATGAG": 1144, + "TATGTA": 1145, + "TATGTT": 1146, + "TATGTC": 1147, + "TATGTG": 1148, + "TATGCA": 1149, + "TATGCT": 1150, + "TATGCC": 1151, + "TATGCG": 1152, + "TATGGA": 1153, + "TATGGT": 1154, + "TATGGC": 1155, + "TATGGG": 1156, + "TACAAA": 1157, + "TACAAT": 1158, + "TACAAC": 1159, + "TACAAG": 1160, + "TACATA": 1161, + "TACATT": 1162, + "TACATC": 1163, + "TACATG": 1164, + "TACACA": 1165, + "TACACT": 1166, + "TACACC": 1167, + "TACACG": 1168, + "TACAGA": 1169, + "TACAGT": 1170, + "TACAGC": 1171, + "TACAGG": 1172, + "TACTAA": 1173, + "TACTAT": 1174, + "TACTAC": 1175, + "TACTAG": 1176, + "TACTTA": 1177, + "TACTTT": 1178, + "TACTTC": 1179, + "TACTTG": 1180, + "TACTCA": 1181, + "TACTCT": 1182, + "TACTCC": 1183, + "TACTCG": 1184, + "TACTGA": 1185, + "TACTGT": 1186, + "TACTGC": 1187, + "TACTGG": 1188, + "TACCAA": 1189, + "TACCAT": 1190, + "TACCAC": 1191, + "TACCAG": 1192, + "TACCTA": 1193, + "TACCTT": 1194, + "TACCTC": 1195, + "TACCTG": 1196, + "TACCCA": 1197, + "TACCCT": 1198, + "TACCCC": 1199, + "TACCCG": 1200, + "TACCGA": 1201, + "TACCGT": 1202, + "TACCGC": 1203, + "TACCGG": 1204, + "TACGAA": 1205, + "TACGAT": 1206, + "TACGAC": 1207, + "TACGAG": 1208, + "TACGTA": 1209, + "TACGTT": 1210, + "TACGTC": 1211, + "TACGTG": 1212, + "TACGCA": 1213, + "TACGCT": 1214, + "TACGCC": 1215, + "TACGCG": 1216, + "TACGGA": 1217, + "TACGGT": 1218, + "TACGGC": 1219, + "TACGGG": 1220, + "TAGAAA": 1221, + "TAGAAT": 1222, + "TAGAAC": 1223, + "TAGAAG": 1224, + "TAGATA": 1225, + "TAGATT": 1226, + "TAGATC": 1227, + "TAGATG": 1228, + "TAGACA": 1229, + "TAGACT": 1230, + "TAGACC": 1231, + "TAGACG": 1232, + "TAGAGA": 1233, + "TAGAGT": 1234, + "TAGAGC": 1235, + "TAGAGG": 1236, + "TAGTAA": 1237, + "TAGTAT": 1238, + "TAGTAC": 1239, + "TAGTAG": 1240, + "TAGTTA": 1241, + "TAGTTT": 1242, + "TAGTTC": 1243, + "TAGTTG": 1244, + "TAGTCA": 1245, + "TAGTCT": 1246, + "TAGTCC": 1247, + "TAGTCG": 1248, + "TAGTGA": 1249, + "TAGTGT": 1250, + "TAGTGC": 1251, + "TAGTGG": 1252, + "TAGCAA": 1253, + "TAGCAT": 1254, + "TAGCAC": 1255, + "TAGCAG": 1256, + "TAGCTA": 1257, + "TAGCTT": 1258, + "TAGCTC": 1259, + "TAGCTG": 1260, + "TAGCCA": 1261, + "TAGCCT": 1262, + "TAGCCC": 1263, + "TAGCCG": 1264, + "TAGCGA": 1265, + "TAGCGT": 1266, + "TAGCGC": 1267, + "TAGCGG": 1268, + "TAGGAA": 1269, + "TAGGAT": 1270, + "TAGGAC": 1271, + "TAGGAG": 1272, + "TAGGTA": 1273, + "TAGGTT": 1274, + "TAGGTC": 1275, + "TAGGTG": 1276, + "TAGGCA": 1277, + "TAGGCT": 1278, + "TAGGCC": 1279, + "TAGGCG": 1280, + "TAGGGA": 1281, + "TAGGGT": 1282, + "TAGGGC": 1283, + "TAGGGG": 1284, + "TTAAAA": 1285, + "TTAAAT": 1286, + "TTAAAC": 1287, + "TTAAAG": 1288, + "TTAATA": 1289, + "TTAATT": 1290, + "TTAATC": 1291, + "TTAATG": 1292, + "TTAACA": 1293, + "TTAACT": 1294, + "TTAACC": 1295, + "TTAACG": 1296, + "TTAAGA": 1297, + "TTAAGT": 1298, + "TTAAGC": 1299, + "TTAAGG": 1300, + "TTATAA": 1301, + "TTATAT": 1302, + "TTATAC": 1303, + "TTATAG": 1304, + "TTATTA": 1305, + "TTATTT": 1306, + "TTATTC": 1307, + "TTATTG": 1308, + "TTATCA": 1309, + "TTATCT": 1310, + "TTATCC": 1311, + "TTATCG": 1312, + "TTATGA": 1313, + "TTATGT": 1314, + "TTATGC": 1315, + "TTATGG": 1316, + "TTACAA": 1317, + "TTACAT": 1318, + "TTACAC": 1319, + "TTACAG": 1320, + "TTACTA": 1321, + "TTACTT": 1322, + "TTACTC": 1323, + "TTACTG": 1324, + "TTACCA": 1325, + "TTACCT": 1326, + "TTACCC": 1327, + "TTACCG": 1328, + "TTACGA": 1329, + "TTACGT": 1330, + "TTACGC": 1331, + "TTACGG": 1332, + "TTAGAA": 1333, + "TTAGAT": 1334, + "TTAGAC": 1335, + "TTAGAG": 1336, + "TTAGTA": 1337, + "TTAGTT": 1338, + "TTAGTC": 1339, + "TTAGTG": 1340, + "TTAGCA": 1341, + "TTAGCT": 1342, + "TTAGCC": 1343, + "TTAGCG": 1344, + "TTAGGA": 1345, + "TTAGGT": 1346, + "TTAGGC": 1347, + "TTAGGG": 1348, + "TTTAAA": 1349, + "TTTAAT": 1350, + "TTTAAC": 1351, + "TTTAAG": 1352, + "TTTATA": 1353, + "TTTATT": 1354, + "TTTATC": 1355, + "TTTATG": 1356, + "TTTACA": 1357, + "TTTACT": 1358, + "TTTACC": 1359, + "TTTACG": 1360, + "TTTAGA": 1361, + "TTTAGT": 1362, + "TTTAGC": 1363, + "TTTAGG": 1364, + "TTTTAA": 1365, + "TTTTAT": 1366, + "TTTTAC": 1367, + "TTTTAG": 1368, + "TTTTTA": 1369, + "TTTTTT": 1370, + "TTTTTC": 1371, + "TTTTTG": 1372, + "TTTTCA": 1373, + "TTTTCT": 1374, + "TTTTCC": 1375, + "TTTTCG": 1376, + "TTTTGA": 1377, + "TTTTGT": 1378, + "TTTTGC": 1379, + "TTTTGG": 1380, + "TTTCAA": 1381, + "TTTCAT": 1382, + "TTTCAC": 1383, + "TTTCAG": 1384, + "TTTCTA": 1385, + "TTTCTT": 1386, + "TTTCTC": 1387, + "TTTCTG": 1388, + "TTTCCA": 1389, + "TTTCCT": 1390, + "TTTCCC": 1391, + "TTTCCG": 1392, + "TTTCGA": 1393, + "TTTCGT": 1394, + "TTTCGC": 1395, + "TTTCGG": 1396, + "TTTGAA": 1397, + "TTTGAT": 1398, + "TTTGAC": 1399, + "TTTGAG": 1400, + "TTTGTA": 1401, + "TTTGTT": 1402, + "TTTGTC": 1403, + "TTTGTG": 1404, + "TTTGCA": 1405, + "TTTGCT": 1406, + "TTTGCC": 1407, + "TTTGCG": 1408, + "TTTGGA": 1409, + "TTTGGT": 1410, + "TTTGGC": 1411, + "TTTGGG": 1412, + "TTCAAA": 1413, + "TTCAAT": 1414, + "TTCAAC": 1415, + "TTCAAG": 1416, + "TTCATA": 1417, + "TTCATT": 1418, + "TTCATC": 1419, + "TTCATG": 1420, + "TTCACA": 1421, + "TTCACT": 1422, + "TTCACC": 1423, + "TTCACG": 1424, + "TTCAGA": 1425, + "TTCAGT": 1426, + "TTCAGC": 1427, + "TTCAGG": 1428, + "TTCTAA": 1429, + "TTCTAT": 1430, + "TTCTAC": 1431, + "TTCTAG": 1432, + "TTCTTA": 1433, + "TTCTTT": 1434, + "TTCTTC": 1435, + "TTCTTG": 1436, + "TTCTCA": 1437, + "TTCTCT": 1438, + "TTCTCC": 1439, + "TTCTCG": 1440, + "TTCTGA": 1441, + "TTCTGT": 1442, + "TTCTGC": 1443, + "TTCTGG": 1444, + "TTCCAA": 1445, + "TTCCAT": 1446, + "TTCCAC": 1447, + "TTCCAG": 1448, + "TTCCTA": 1449, + "TTCCTT": 1450, + "TTCCTC": 1451, + "TTCCTG": 1452, + "TTCCCA": 1453, + "TTCCCT": 1454, + "TTCCCC": 1455, + "TTCCCG": 1456, + "TTCCGA": 1457, + "TTCCGT": 1458, + "TTCCGC": 1459, + "TTCCGG": 1460, + "TTCGAA": 1461, + "TTCGAT": 1462, + "TTCGAC": 1463, + "TTCGAG": 1464, + "TTCGTA": 1465, + "TTCGTT": 1466, + "TTCGTC": 1467, + "TTCGTG": 1468, + "TTCGCA": 1469, + "TTCGCT": 1470, + "TTCGCC": 1471, + "TTCGCG": 1472, + "TTCGGA": 1473, + "TTCGGT": 1474, + "TTCGGC": 1475, + "TTCGGG": 1476, + "TTGAAA": 1477, + "TTGAAT": 1478, + "TTGAAC": 1479, + "TTGAAG": 1480, + "TTGATA": 1481, + "TTGATT": 1482, + "TTGATC": 1483, + "TTGATG": 1484, + "TTGACA": 1485, + "TTGACT": 1486, + "TTGACC": 1487, + "TTGACG": 1488, + "TTGAGA": 1489, + "TTGAGT": 1490, + "TTGAGC": 1491, + "TTGAGG": 1492, + "TTGTAA": 1493, + "TTGTAT": 1494, + "TTGTAC": 1495, + "TTGTAG": 1496, + "TTGTTA": 1497, + "TTGTTT": 1498, + "TTGTTC": 1499, + "TTGTTG": 1500, + "TTGTCA": 1501, + "TTGTCT": 1502, + "TTGTCC": 1503, + "TTGTCG": 1504, + "TTGTGA": 1505, + "TTGTGT": 1506, + "TTGTGC": 1507, + "TTGTGG": 1508, + "TTGCAA": 1509, + "TTGCAT": 1510, + "TTGCAC": 1511, + "TTGCAG": 1512, + "TTGCTA": 1513, + "TTGCTT": 1514, + "TTGCTC": 1515, + "TTGCTG": 1516, + "TTGCCA": 1517, + "TTGCCT": 1518, + "TTGCCC": 1519, + "TTGCCG": 1520, + "TTGCGA": 1521, + "TTGCGT": 1522, + "TTGCGC": 1523, + "TTGCGG": 1524, + "TTGGAA": 1525, + "TTGGAT": 1526, + "TTGGAC": 1527, + "TTGGAG": 1528, + "TTGGTA": 1529, + "TTGGTT": 1530, + "TTGGTC": 1531, + "TTGGTG": 1532, + "TTGGCA": 1533, + "TTGGCT": 1534, + "TTGGCC": 1535, + "TTGGCG": 1536, + "TTGGGA": 1537, + "TTGGGT": 1538, + "TTGGGC": 1539, + "TTGGGG": 1540, + "TCAAAA": 1541, + "TCAAAT": 1542, + "TCAAAC": 1543, + "TCAAAG": 1544, + "TCAATA": 1545, + "TCAATT": 1546, + "TCAATC": 1547, + "TCAATG": 1548, + "TCAACA": 1549, + "TCAACT": 1550, + "TCAACC": 1551, + "TCAACG": 1552, + "TCAAGA": 1553, + "TCAAGT": 1554, + "TCAAGC": 1555, + "TCAAGG": 1556, + "TCATAA": 1557, + "TCATAT": 1558, + "TCATAC": 1559, + "TCATAG": 1560, + "TCATTA": 1561, + "TCATTT": 1562, + "TCATTC": 1563, + "TCATTG": 1564, + "TCATCA": 1565, + "TCATCT": 1566, + "TCATCC": 1567, + "TCATCG": 1568, + "TCATGA": 1569, + "TCATGT": 1570, + "TCATGC": 1571, + "TCATGG": 1572, + "TCACAA": 1573, + "TCACAT": 1574, + "TCACAC": 1575, + "TCACAG": 1576, + "TCACTA": 1577, + "TCACTT": 1578, + "TCACTC": 1579, + "TCACTG": 1580, + "TCACCA": 1581, + "TCACCT": 1582, + "TCACCC": 1583, + "TCACCG": 1584, + "TCACGA": 1585, + "TCACGT": 1586, + "TCACGC": 1587, + "TCACGG": 1588, + "TCAGAA": 1589, + "TCAGAT": 1590, + "TCAGAC": 1591, + "TCAGAG": 1592, + "TCAGTA": 1593, + "TCAGTT": 1594, + "TCAGTC": 1595, + "TCAGTG": 1596, + "TCAGCA": 1597, + "TCAGCT": 1598, + "TCAGCC": 1599, + "TCAGCG": 1600, + "TCAGGA": 1601, + "TCAGGT": 1602, + "TCAGGC": 1603, + "TCAGGG": 1604, + "TCTAAA": 1605, + "TCTAAT": 1606, + "TCTAAC": 1607, + "TCTAAG": 1608, + "TCTATA": 1609, + "TCTATT": 1610, + "TCTATC": 1611, + "TCTATG": 1612, + "TCTACA": 1613, + "TCTACT": 1614, + "TCTACC": 1615, + "TCTACG": 1616, + "TCTAGA": 1617, + "TCTAGT": 1618, + "TCTAGC": 1619, + "TCTAGG": 1620, + "TCTTAA": 1621, + "TCTTAT": 1622, + "TCTTAC": 1623, + "TCTTAG": 1624, + "TCTTTA": 1625, + "TCTTTT": 1626, + "TCTTTC": 1627, + "TCTTTG": 1628, + "TCTTCA": 1629, + "TCTTCT": 1630, + "TCTTCC": 1631, + "TCTTCG": 1632, + "TCTTGA": 1633, + "TCTTGT": 1634, + "TCTTGC": 1635, + "TCTTGG": 1636, + "TCTCAA": 1637, + "TCTCAT": 1638, + "TCTCAC": 1639, + "TCTCAG": 1640, + "TCTCTA": 1641, + "TCTCTT": 1642, + "TCTCTC": 1643, + "TCTCTG": 1644, + "TCTCCA": 1645, + "TCTCCT": 1646, + "TCTCCC": 1647, + "TCTCCG": 1648, + "TCTCGA": 1649, + "TCTCGT": 1650, + "TCTCGC": 1651, + "TCTCGG": 1652, + "TCTGAA": 1653, + "TCTGAT": 1654, + "TCTGAC": 1655, + "TCTGAG": 1656, + "TCTGTA": 1657, + "TCTGTT": 1658, + "TCTGTC": 1659, + "TCTGTG": 1660, + "TCTGCA": 1661, + "TCTGCT": 1662, + "TCTGCC": 1663, + "TCTGCG": 1664, + "TCTGGA": 1665, + "TCTGGT": 1666, + "TCTGGC": 1667, + "TCTGGG": 1668, + "TCCAAA": 1669, + "TCCAAT": 1670, + "TCCAAC": 1671, + "TCCAAG": 1672, + "TCCATA": 1673, + "TCCATT": 1674, + "TCCATC": 1675, + "TCCATG": 1676, + "TCCACA": 1677, + "TCCACT": 1678, + "TCCACC": 1679, + "TCCACG": 1680, + "TCCAGA": 1681, + "TCCAGT": 1682, + "TCCAGC": 1683, + "TCCAGG": 1684, + "TCCTAA": 1685, + "TCCTAT": 1686, + "TCCTAC": 1687, + "TCCTAG": 1688, + "TCCTTA": 1689, + "TCCTTT": 1690, + "TCCTTC": 1691, + "TCCTTG": 1692, + "TCCTCA": 1693, + "TCCTCT": 1694, + "TCCTCC": 1695, + "TCCTCG": 1696, + "TCCTGA": 1697, + "TCCTGT": 1698, + "TCCTGC": 1699, + "TCCTGG": 1700, + "TCCCAA": 1701, + "TCCCAT": 1702, + "TCCCAC": 1703, + "TCCCAG": 1704, + "TCCCTA": 1705, + "TCCCTT": 1706, + "TCCCTC": 1707, + "TCCCTG": 1708, + "TCCCCA": 1709, + "TCCCCT": 1710, + "TCCCCC": 1711, + "TCCCCG": 1712, + "TCCCGA": 1713, + "TCCCGT": 1714, + "TCCCGC": 1715, + "TCCCGG": 1716, + "TCCGAA": 1717, + "TCCGAT": 1718, + "TCCGAC": 1719, + "TCCGAG": 1720, + "TCCGTA": 1721, + "TCCGTT": 1722, + "TCCGTC": 1723, + "TCCGTG": 1724, + "TCCGCA": 1725, + "TCCGCT": 1726, + "TCCGCC": 1727, + "TCCGCG": 1728, + "TCCGGA": 1729, + "TCCGGT": 1730, + "TCCGGC": 1731, + "TCCGGG": 1732, + "TCGAAA": 1733, + "TCGAAT": 1734, + "TCGAAC": 1735, + "TCGAAG": 1736, + "TCGATA": 1737, + "TCGATT": 1738, + "TCGATC": 1739, + "TCGATG": 1740, + "TCGACA": 1741, + "TCGACT": 1742, + "TCGACC": 1743, + "TCGACG": 1744, + "TCGAGA": 1745, + "TCGAGT": 1746, + "TCGAGC": 1747, + "TCGAGG": 1748, + "TCGTAA": 1749, + "TCGTAT": 1750, + "TCGTAC": 1751, + "TCGTAG": 1752, + "TCGTTA": 1753, + "TCGTTT": 1754, + "TCGTTC": 1755, + "TCGTTG": 1756, + "TCGTCA": 1757, + "TCGTCT": 1758, + "TCGTCC": 1759, + "TCGTCG": 1760, + "TCGTGA": 1761, + "TCGTGT": 1762, + "TCGTGC": 1763, + "TCGTGG": 1764, + "TCGCAA": 1765, + "TCGCAT": 1766, + "TCGCAC": 1767, + "TCGCAG": 1768, + "TCGCTA": 1769, + "TCGCTT": 1770, + "TCGCTC": 1771, + "TCGCTG": 1772, + "TCGCCA": 1773, + "TCGCCT": 1774, + "TCGCCC": 1775, + "TCGCCG": 1776, + "TCGCGA": 1777, + "TCGCGT": 1778, + "TCGCGC": 1779, + "TCGCGG": 1780, + "TCGGAA": 1781, + "TCGGAT": 1782, + "TCGGAC": 1783, + "TCGGAG": 1784, + "TCGGTA": 1785, + "TCGGTT": 1786, + "TCGGTC": 1787, + "TCGGTG": 1788, + "TCGGCA": 1789, + "TCGGCT": 1790, + "TCGGCC": 1791, + "TCGGCG": 1792, + "TCGGGA": 1793, + "TCGGGT": 1794, + "TCGGGC": 1795, + "TCGGGG": 1796, + "TGAAAA": 1797, + "TGAAAT": 1798, + "TGAAAC": 1799, + "TGAAAG": 1800, + "TGAATA": 1801, + "TGAATT": 1802, + "TGAATC": 1803, + "TGAATG": 1804, + "TGAACA": 1805, + "TGAACT": 1806, + "TGAACC": 1807, + "TGAACG": 1808, + "TGAAGA": 1809, + "TGAAGT": 1810, + "TGAAGC": 1811, + "TGAAGG": 1812, + "TGATAA": 1813, + "TGATAT": 1814, + "TGATAC": 1815, + "TGATAG": 1816, + "TGATTA": 1817, + "TGATTT": 1818, + "TGATTC": 1819, + "TGATTG": 1820, + "TGATCA": 1821, + "TGATCT": 1822, + "TGATCC": 1823, + "TGATCG": 1824, + "TGATGA": 1825, + "TGATGT": 1826, + "TGATGC": 1827, + "TGATGG": 1828, + "TGACAA": 1829, + "TGACAT": 1830, + "TGACAC": 1831, + "TGACAG": 1832, + "TGACTA": 1833, + "TGACTT": 1834, + "TGACTC": 1835, + "TGACTG": 1836, + "TGACCA": 1837, + "TGACCT": 1838, + "TGACCC": 1839, + "TGACCG": 1840, + "TGACGA": 1841, + "TGACGT": 1842, + "TGACGC": 1843, + "TGACGG": 1844, + "TGAGAA": 1845, + "TGAGAT": 1846, + "TGAGAC": 1847, + "TGAGAG": 1848, + "TGAGTA": 1849, + "TGAGTT": 1850, + "TGAGTC": 1851, + "TGAGTG": 1852, + "TGAGCA": 1853, + "TGAGCT": 1854, + "TGAGCC": 1855, + "TGAGCG": 1856, + "TGAGGA": 1857, + "TGAGGT": 1858, + "TGAGGC": 1859, + "TGAGGG": 1860, + "TGTAAA": 1861, + "TGTAAT": 1862, + "TGTAAC": 1863, + "TGTAAG": 1864, + "TGTATA": 1865, + "TGTATT": 1866, + "TGTATC": 1867, + "TGTATG": 1868, + "TGTACA": 1869, + "TGTACT": 1870, + "TGTACC": 1871, + "TGTACG": 1872, + "TGTAGA": 1873, + "TGTAGT": 1874, + "TGTAGC": 1875, + "TGTAGG": 1876, + "TGTTAA": 1877, + "TGTTAT": 1878, + "TGTTAC": 1879, + "TGTTAG": 1880, + "TGTTTA": 1881, + "TGTTTT": 1882, + "TGTTTC": 1883, + "TGTTTG": 1884, + "TGTTCA": 1885, + "TGTTCT": 1886, + "TGTTCC": 1887, + "TGTTCG": 1888, + "TGTTGA": 1889, + "TGTTGT": 1890, + "TGTTGC": 1891, + "TGTTGG": 1892, + "TGTCAA": 1893, + "TGTCAT": 1894, + "TGTCAC": 1895, + "TGTCAG": 1896, + "TGTCTA": 1897, + "TGTCTT": 1898, + "TGTCTC": 1899, + "TGTCTG": 1900, + "TGTCCA": 1901, + "TGTCCT": 1902, + "TGTCCC": 1903, + "TGTCCG": 1904, + "TGTCGA": 1905, + "TGTCGT": 1906, + "TGTCGC": 1907, + "TGTCGG": 1908, + "TGTGAA": 1909, + "TGTGAT": 1910, + "TGTGAC": 1911, + "TGTGAG": 1912, + "TGTGTA": 1913, + "TGTGTT": 1914, + "TGTGTC": 1915, + "TGTGTG": 1916, + "TGTGCA": 1917, + "TGTGCT": 1918, + "TGTGCC": 1919, + "TGTGCG": 1920, + "TGTGGA": 1921, + "TGTGGT": 1922, + "TGTGGC": 1923, + "TGTGGG": 1924, + "TGCAAA": 1925, + "TGCAAT": 1926, + "TGCAAC": 1927, + "TGCAAG": 1928, + "TGCATA": 1929, + "TGCATT": 1930, + "TGCATC": 1931, + "TGCATG": 1932, + "TGCACA": 1933, + "TGCACT": 1934, + "TGCACC": 1935, + "TGCACG": 1936, + "TGCAGA": 1937, + "TGCAGT": 1938, + "TGCAGC": 1939, + "TGCAGG": 1940, + "TGCTAA": 1941, + "TGCTAT": 1942, + "TGCTAC": 1943, + "TGCTAG": 1944, + "TGCTTA": 1945, + "TGCTTT": 1946, + "TGCTTC": 1947, + "TGCTTG": 1948, + "TGCTCA": 1949, + "TGCTCT": 1950, + "TGCTCC": 1951, + "TGCTCG": 1952, + "TGCTGA": 1953, + "TGCTGT": 1954, + "TGCTGC": 1955, + "TGCTGG": 1956, + "TGCCAA": 1957, + "TGCCAT": 1958, + "TGCCAC": 1959, + "TGCCAG": 1960, + "TGCCTA": 1961, + "TGCCTT": 1962, + "TGCCTC": 1963, + "TGCCTG": 1964, + "TGCCCA": 1965, + "TGCCCT": 1966, + "TGCCCC": 1967, + "TGCCCG": 1968, + "TGCCGA": 1969, + "TGCCGT": 1970, + "TGCCGC": 1971, + "TGCCGG": 1972, + "TGCGAA": 1973, + "TGCGAT": 1974, + "TGCGAC": 1975, + "TGCGAG": 1976, + "TGCGTA": 1977, + "TGCGTT": 1978, + "TGCGTC": 1979, + "TGCGTG": 1980, + "TGCGCA": 1981, + "TGCGCT": 1982, + "TGCGCC": 1983, + "TGCGCG": 1984, + "TGCGGA": 1985, + "TGCGGT": 1986, + "TGCGGC": 1987, + "TGCGGG": 1988, + "TGGAAA": 1989, + "TGGAAT": 1990, + "TGGAAC": 1991, + "TGGAAG": 1992, + "TGGATA": 1993, + "TGGATT": 1994, + "TGGATC": 1995, + "TGGATG": 1996, + "TGGACA": 1997, + "TGGACT": 1998, + "TGGACC": 1999, + "TGGACG": 2000, + "TGGAGA": 2001, + "TGGAGT": 2002, + "TGGAGC": 2003, + "TGGAGG": 2004, + "TGGTAA": 2005, + "TGGTAT": 2006, + "TGGTAC": 2007, + "TGGTAG": 2008, + "TGGTTA": 2009, + "TGGTTT": 2010, + "TGGTTC": 2011, + "TGGTTG": 2012, + "TGGTCA": 2013, + "TGGTCT": 2014, + "TGGTCC": 2015, + "TGGTCG": 2016, + "TGGTGA": 2017, + "TGGTGT": 2018, + "TGGTGC": 2019, + "TGGTGG": 2020, + "TGGCAA": 2021, + "TGGCAT": 2022, + "TGGCAC": 2023, + "TGGCAG": 2024, + "TGGCTA": 2025, + "TGGCTT": 2026, + "TGGCTC": 2027, + "TGGCTG": 2028, + "TGGCCA": 2029, + "TGGCCT": 2030, + "TGGCCC": 2031, + "TGGCCG": 2032, + "TGGCGA": 2033, + "TGGCGT": 2034, + "TGGCGC": 2035, + "TGGCGG": 2036, + "TGGGAA": 2037, + "TGGGAT": 2038, + "TGGGAC": 2039, + "TGGGAG": 2040, + "TGGGTA": 2041, + "TGGGTT": 2042, + "TGGGTC": 2043, + "TGGGTG": 2044, + "TGGGCA": 2045, + "TGGGCT": 2046, + "TGGGCC": 2047, + "TGGGCG": 2048, + "TGGGGA": 2049, + "TGGGGT": 2050, + "TGGGGC": 2051, + "TGGGGG": 2052, + "CAAAAA": 2053, + "CAAAAT": 2054, + "CAAAAC": 2055, + "CAAAAG": 2056, + "CAAATA": 2057, + "CAAATT": 2058, + "CAAATC": 2059, + "CAAATG": 2060, + "CAAACA": 2061, + "CAAACT": 2062, + "CAAACC": 2063, + "CAAACG": 2064, + "CAAAGA": 2065, + "CAAAGT": 2066, + "CAAAGC": 2067, + "CAAAGG": 2068, + "CAATAA": 2069, + "CAATAT": 2070, + "CAATAC": 2071, + "CAATAG": 2072, + "CAATTA": 2073, + "CAATTT": 2074, + "CAATTC": 2075, + "CAATTG": 2076, + "CAATCA": 2077, + "CAATCT": 2078, + "CAATCC": 2079, + "CAATCG": 2080, + "CAATGA": 2081, + "CAATGT": 2082, + "CAATGC": 2083, + "CAATGG": 2084, + "CAACAA": 2085, + "CAACAT": 2086, + "CAACAC": 2087, + "CAACAG": 2088, + "CAACTA": 2089, + "CAACTT": 2090, + "CAACTC": 2091, + "CAACTG": 2092, + "CAACCA": 2093, + "CAACCT": 2094, + "CAACCC": 2095, + "CAACCG": 2096, + "CAACGA": 2097, + "CAACGT": 2098, + "CAACGC": 2099, + "CAACGG": 2100, + "CAAGAA": 2101, + "CAAGAT": 2102, + "CAAGAC": 2103, + "CAAGAG": 2104, + "CAAGTA": 2105, + "CAAGTT": 2106, + "CAAGTC": 2107, + "CAAGTG": 2108, + "CAAGCA": 2109, + "CAAGCT": 2110, + "CAAGCC": 2111, + "CAAGCG": 2112, + "CAAGGA": 2113, + "CAAGGT": 2114, + "CAAGGC": 2115, + "CAAGGG": 2116, + "CATAAA": 2117, + "CATAAT": 2118, + "CATAAC": 2119, + "CATAAG": 2120, + "CATATA": 2121, + "CATATT": 2122, + "CATATC": 2123, + "CATATG": 2124, + "CATACA": 2125, + "CATACT": 2126, + "CATACC": 2127, + "CATACG": 2128, + "CATAGA": 2129, + "CATAGT": 2130, + "CATAGC": 2131, + "CATAGG": 2132, + "CATTAA": 2133, + "CATTAT": 2134, + "CATTAC": 2135, + "CATTAG": 2136, + "CATTTA": 2137, + "CATTTT": 2138, + "CATTTC": 2139, + "CATTTG": 2140, + "CATTCA": 2141, + "CATTCT": 2142, + "CATTCC": 2143, + "CATTCG": 2144, + "CATTGA": 2145, + "CATTGT": 2146, + "CATTGC": 2147, + "CATTGG": 2148, + "CATCAA": 2149, + "CATCAT": 2150, + "CATCAC": 2151, + "CATCAG": 2152, + "CATCTA": 2153, + "CATCTT": 2154, + "CATCTC": 2155, + "CATCTG": 2156, + "CATCCA": 2157, + "CATCCT": 2158, + "CATCCC": 2159, + "CATCCG": 2160, + "CATCGA": 2161, + "CATCGT": 2162, + "CATCGC": 2163, + "CATCGG": 2164, + "CATGAA": 2165, + "CATGAT": 2166, + "CATGAC": 2167, + "CATGAG": 2168, + "CATGTA": 2169, + "CATGTT": 2170, + "CATGTC": 2171, + "CATGTG": 2172, + "CATGCA": 2173, + "CATGCT": 2174, + "CATGCC": 2175, + "CATGCG": 2176, + "CATGGA": 2177, + "CATGGT": 2178, + "CATGGC": 2179, + "CATGGG": 2180, + "CACAAA": 2181, + "CACAAT": 2182, + "CACAAC": 2183, + "CACAAG": 2184, + "CACATA": 2185, + "CACATT": 2186, + "CACATC": 2187, + "CACATG": 2188, + "CACACA": 2189, + "CACACT": 2190, + "CACACC": 2191, + "CACACG": 2192, + "CACAGA": 2193, + "CACAGT": 2194, + "CACAGC": 2195, + "CACAGG": 2196, + "CACTAA": 2197, + "CACTAT": 2198, + "CACTAC": 2199, + "CACTAG": 2200, + "CACTTA": 2201, + "CACTTT": 2202, + "CACTTC": 2203, + "CACTTG": 2204, + "CACTCA": 2205, + "CACTCT": 2206, + "CACTCC": 2207, + "CACTCG": 2208, + "CACTGA": 2209, + "CACTGT": 2210, + "CACTGC": 2211, + "CACTGG": 2212, + "CACCAA": 2213, + "CACCAT": 2214, + "CACCAC": 2215, + "CACCAG": 2216, + "CACCTA": 2217, + "CACCTT": 2218, + "CACCTC": 2219, + "CACCTG": 2220, + "CACCCA": 2221, + "CACCCT": 2222, + "CACCCC": 2223, + "CACCCG": 2224, + "CACCGA": 2225, + "CACCGT": 2226, + "CACCGC": 2227, + "CACCGG": 2228, + "CACGAA": 2229, + "CACGAT": 2230, + "CACGAC": 2231, + "CACGAG": 2232, + "CACGTA": 2233, + "CACGTT": 2234, + "CACGTC": 2235, + "CACGTG": 2236, + "CACGCA": 2237, + "CACGCT": 2238, + "CACGCC": 2239, + "CACGCG": 2240, + "CACGGA": 2241, + "CACGGT": 2242, + "CACGGC": 2243, + "CACGGG": 2244, + "CAGAAA": 2245, + "CAGAAT": 2246, + "CAGAAC": 2247, + "CAGAAG": 2248, + "CAGATA": 2249, + "CAGATT": 2250, + "CAGATC": 2251, + "CAGATG": 2252, + "CAGACA": 2253, + "CAGACT": 2254, + "CAGACC": 2255, + "CAGACG": 2256, + "CAGAGA": 2257, + "CAGAGT": 2258, + "CAGAGC": 2259, + "CAGAGG": 2260, + "CAGTAA": 2261, + "CAGTAT": 2262, + "CAGTAC": 2263, + "CAGTAG": 2264, + "CAGTTA": 2265, + "CAGTTT": 2266, + "CAGTTC": 2267, + "CAGTTG": 2268, + "CAGTCA": 2269, + "CAGTCT": 2270, + "CAGTCC": 2271, + "CAGTCG": 2272, + "CAGTGA": 2273, + "CAGTGT": 2274, + "CAGTGC": 2275, + "CAGTGG": 2276, + "CAGCAA": 2277, + "CAGCAT": 2278, + "CAGCAC": 2279, + "CAGCAG": 2280, + "CAGCTA": 2281, + "CAGCTT": 2282, + "CAGCTC": 2283, + "CAGCTG": 2284, + "CAGCCA": 2285, + "CAGCCT": 2286, + "CAGCCC": 2287, + "CAGCCG": 2288, + "CAGCGA": 2289, + "CAGCGT": 2290, + "CAGCGC": 2291, + "CAGCGG": 2292, + "CAGGAA": 2293, + "CAGGAT": 2294, + "CAGGAC": 2295, + "CAGGAG": 2296, + "CAGGTA": 2297, + "CAGGTT": 2298, + "CAGGTC": 2299, + "CAGGTG": 2300, + "CAGGCA": 2301, + "CAGGCT": 2302, + "CAGGCC": 2303, + "CAGGCG": 2304, + "CAGGGA": 2305, + "CAGGGT": 2306, + "CAGGGC": 2307, + "CAGGGG": 2308, + "CTAAAA": 2309, + "CTAAAT": 2310, + "CTAAAC": 2311, + "CTAAAG": 2312, + "CTAATA": 2313, + "CTAATT": 2314, + "CTAATC": 2315, + "CTAATG": 2316, + "CTAACA": 2317, + "CTAACT": 2318, + "CTAACC": 2319, + "CTAACG": 2320, + "CTAAGA": 2321, + "CTAAGT": 2322, + "CTAAGC": 2323, + "CTAAGG": 2324, + "CTATAA": 2325, + "CTATAT": 2326, + "CTATAC": 2327, + "CTATAG": 2328, + "CTATTA": 2329, + "CTATTT": 2330, + "CTATTC": 2331, + "CTATTG": 2332, + "CTATCA": 2333, + "CTATCT": 2334, + "CTATCC": 2335, + "CTATCG": 2336, + "CTATGA": 2337, + "CTATGT": 2338, + "CTATGC": 2339, + "CTATGG": 2340, + "CTACAA": 2341, + "CTACAT": 2342, + "CTACAC": 2343, + "CTACAG": 2344, + "CTACTA": 2345, + "CTACTT": 2346, + "CTACTC": 2347, + "CTACTG": 2348, + "CTACCA": 2349, + "CTACCT": 2350, + "CTACCC": 2351, + "CTACCG": 2352, + "CTACGA": 2353, + "CTACGT": 2354, + "CTACGC": 2355, + "CTACGG": 2356, + "CTAGAA": 2357, + "CTAGAT": 2358, + "CTAGAC": 2359, + "CTAGAG": 2360, + "CTAGTA": 2361, + "CTAGTT": 2362, + "CTAGTC": 2363, + "CTAGTG": 2364, + "CTAGCA": 2365, + "CTAGCT": 2366, + "CTAGCC": 2367, + "CTAGCG": 2368, + "CTAGGA": 2369, + "CTAGGT": 2370, + "CTAGGC": 2371, + "CTAGGG": 2372, + "CTTAAA": 2373, + "CTTAAT": 2374, + "CTTAAC": 2375, + "CTTAAG": 2376, + "CTTATA": 2377, + "CTTATT": 2378, + "CTTATC": 2379, + "CTTATG": 2380, + "CTTACA": 2381, + "CTTACT": 2382, + "CTTACC": 2383, + "CTTACG": 2384, + "CTTAGA": 2385, + "CTTAGT": 2386, + "CTTAGC": 2387, + "CTTAGG": 2388, + "CTTTAA": 2389, + "CTTTAT": 2390, + "CTTTAC": 2391, + "CTTTAG": 2392, + "CTTTTA": 2393, + "CTTTTT": 2394, + "CTTTTC": 2395, + "CTTTTG": 2396, + "CTTTCA": 2397, + "CTTTCT": 2398, + "CTTTCC": 2399, + "CTTTCG": 2400, + "CTTTGA": 2401, + "CTTTGT": 2402, + "CTTTGC": 2403, + "CTTTGG": 2404, + "CTTCAA": 2405, + "CTTCAT": 2406, + "CTTCAC": 2407, + "CTTCAG": 2408, + "CTTCTA": 2409, + "CTTCTT": 2410, + "CTTCTC": 2411, + "CTTCTG": 2412, + "CTTCCA": 2413, + "CTTCCT": 2414, + "CTTCCC": 2415, + "CTTCCG": 2416, + "CTTCGA": 2417, + "CTTCGT": 2418, + "CTTCGC": 2419, + "CTTCGG": 2420, + "CTTGAA": 2421, + "CTTGAT": 2422, + "CTTGAC": 2423, + "CTTGAG": 2424, + "CTTGTA": 2425, + "CTTGTT": 2426, + "CTTGTC": 2427, + "CTTGTG": 2428, + "CTTGCA": 2429, + "CTTGCT": 2430, + "CTTGCC": 2431, + "CTTGCG": 2432, + "CTTGGA": 2433, + "CTTGGT": 2434, + "CTTGGC": 2435, + "CTTGGG": 2436, + "CTCAAA": 2437, + "CTCAAT": 2438, + "CTCAAC": 2439, + "CTCAAG": 2440, + "CTCATA": 2441, + "CTCATT": 2442, + "CTCATC": 2443, + "CTCATG": 2444, + "CTCACA": 2445, + "CTCACT": 2446, + "CTCACC": 2447, + "CTCACG": 2448, + "CTCAGA": 2449, + "CTCAGT": 2450, + "CTCAGC": 2451, + "CTCAGG": 2452, + "CTCTAA": 2453, + "CTCTAT": 2454, + "CTCTAC": 2455, + "CTCTAG": 2456, + "CTCTTA": 2457, + "CTCTTT": 2458, + "CTCTTC": 2459, + "CTCTTG": 2460, + "CTCTCA": 2461, + "CTCTCT": 2462, + "CTCTCC": 2463, + "CTCTCG": 2464, + "CTCTGA": 2465, + "CTCTGT": 2466, + "CTCTGC": 2467, + "CTCTGG": 2468, + "CTCCAA": 2469, + "CTCCAT": 2470, + "CTCCAC": 2471, + "CTCCAG": 2472, + "CTCCTA": 2473, + "CTCCTT": 2474, + "CTCCTC": 2475, + "CTCCTG": 2476, + "CTCCCA": 2477, + "CTCCCT": 2478, + "CTCCCC": 2479, + "CTCCCG": 2480, + "CTCCGA": 2481, + "CTCCGT": 2482, + "CTCCGC": 2483, + "CTCCGG": 2484, + "CTCGAA": 2485, + "CTCGAT": 2486, + "CTCGAC": 2487, + "CTCGAG": 2488, + "CTCGTA": 2489, + "CTCGTT": 2490, + "CTCGTC": 2491, + "CTCGTG": 2492, + "CTCGCA": 2493, + "CTCGCT": 2494, + "CTCGCC": 2495, + "CTCGCG": 2496, + "CTCGGA": 2497, + "CTCGGT": 2498, + "CTCGGC": 2499, + "CTCGGG": 2500, + "CTGAAA": 2501, + "CTGAAT": 2502, + "CTGAAC": 2503, + "CTGAAG": 2504, + "CTGATA": 2505, + "CTGATT": 2506, + "CTGATC": 2507, + "CTGATG": 2508, + "CTGACA": 2509, + "CTGACT": 2510, + "CTGACC": 2511, + "CTGACG": 2512, + "CTGAGA": 2513, + "CTGAGT": 2514, + "CTGAGC": 2515, + "CTGAGG": 2516, + "CTGTAA": 2517, + "CTGTAT": 2518, + "CTGTAC": 2519, + "CTGTAG": 2520, + "CTGTTA": 2521, + "CTGTTT": 2522, + "CTGTTC": 2523, + "CTGTTG": 2524, + "CTGTCA": 2525, + "CTGTCT": 2526, + "CTGTCC": 2527, + "CTGTCG": 2528, + "CTGTGA": 2529, + "CTGTGT": 2530, + "CTGTGC": 2531, + "CTGTGG": 2532, + "CTGCAA": 2533, + "CTGCAT": 2534, + "CTGCAC": 2535, + "CTGCAG": 2536, + "CTGCTA": 2537, + "CTGCTT": 2538, + "CTGCTC": 2539, + "CTGCTG": 2540, + "CTGCCA": 2541, + "CTGCCT": 2542, + "CTGCCC": 2543, + "CTGCCG": 2544, + "CTGCGA": 2545, + "CTGCGT": 2546, + "CTGCGC": 2547, + "CTGCGG": 2548, + "CTGGAA": 2549, + "CTGGAT": 2550, + "CTGGAC": 2551, + "CTGGAG": 2552, + "CTGGTA": 2553, + "CTGGTT": 2554, + "CTGGTC": 2555, + "CTGGTG": 2556, + "CTGGCA": 2557, + "CTGGCT": 2558, + "CTGGCC": 2559, + "CTGGCG": 2560, + "CTGGGA": 2561, + "CTGGGT": 2562, + "CTGGGC": 2563, + "CTGGGG": 2564, + "CCAAAA": 2565, + "CCAAAT": 2566, + "CCAAAC": 2567, + "CCAAAG": 2568, + "CCAATA": 2569, + "CCAATT": 2570, + "CCAATC": 2571, + "CCAATG": 2572, + "CCAACA": 2573, + "CCAACT": 2574, + "CCAACC": 2575, + "CCAACG": 2576, + "CCAAGA": 2577, + "CCAAGT": 2578, + "CCAAGC": 2579, + "CCAAGG": 2580, + "CCATAA": 2581, + "CCATAT": 2582, + "CCATAC": 2583, + "CCATAG": 2584, + "CCATTA": 2585, + "CCATTT": 2586, + "CCATTC": 2587, + "CCATTG": 2588, + "CCATCA": 2589, + "CCATCT": 2590, + "CCATCC": 2591, + "CCATCG": 2592, + "CCATGA": 2593, + "CCATGT": 2594, + "CCATGC": 2595, + "CCATGG": 2596, + "CCACAA": 2597, + "CCACAT": 2598, + "CCACAC": 2599, + "CCACAG": 2600, + "CCACTA": 2601, + "CCACTT": 2602, + "CCACTC": 2603, + "CCACTG": 2604, + "CCACCA": 2605, + "CCACCT": 2606, + "CCACCC": 2607, + "CCACCG": 2608, + "CCACGA": 2609, + "CCACGT": 2610, + "CCACGC": 2611, + "CCACGG": 2612, + "CCAGAA": 2613, + "CCAGAT": 2614, + "CCAGAC": 2615, + "CCAGAG": 2616, + "CCAGTA": 2617, + "CCAGTT": 2618, + "CCAGTC": 2619, + "CCAGTG": 2620, + "CCAGCA": 2621, + "CCAGCT": 2622, + "CCAGCC": 2623, + "CCAGCG": 2624, + "CCAGGA": 2625, + "CCAGGT": 2626, + "CCAGGC": 2627, + "CCAGGG": 2628, + "CCTAAA": 2629, + "CCTAAT": 2630, + "CCTAAC": 2631, + "CCTAAG": 2632, + "CCTATA": 2633, + "CCTATT": 2634, + "CCTATC": 2635, + "CCTATG": 2636, + "CCTACA": 2637, + "CCTACT": 2638, + "CCTACC": 2639, + "CCTACG": 2640, + "CCTAGA": 2641, + "CCTAGT": 2642, + "CCTAGC": 2643, + "CCTAGG": 2644, + "CCTTAA": 2645, + "CCTTAT": 2646, + "CCTTAC": 2647, + "CCTTAG": 2648, + "CCTTTA": 2649, + "CCTTTT": 2650, + "CCTTTC": 2651, + "CCTTTG": 2652, + "CCTTCA": 2653, + "CCTTCT": 2654, + "CCTTCC": 2655, + "CCTTCG": 2656, + "CCTTGA": 2657, + "CCTTGT": 2658, + "CCTTGC": 2659, + "CCTTGG": 2660, + "CCTCAA": 2661, + "CCTCAT": 2662, + "CCTCAC": 2663, + "CCTCAG": 2664, + "CCTCTA": 2665, + "CCTCTT": 2666, + "CCTCTC": 2667, + "CCTCTG": 2668, + "CCTCCA": 2669, + "CCTCCT": 2670, + "CCTCCC": 2671, + "CCTCCG": 2672, + "CCTCGA": 2673, + "CCTCGT": 2674, + "CCTCGC": 2675, + "CCTCGG": 2676, + "CCTGAA": 2677, + "CCTGAT": 2678, + "CCTGAC": 2679, + "CCTGAG": 2680, + "CCTGTA": 2681, + "CCTGTT": 2682, + "CCTGTC": 2683, + "CCTGTG": 2684, + "CCTGCA": 2685, + "CCTGCT": 2686, + "CCTGCC": 2687, + "CCTGCG": 2688, + "CCTGGA": 2689, + "CCTGGT": 2690, + "CCTGGC": 2691, + "CCTGGG": 2692, + "CCCAAA": 2693, + "CCCAAT": 2694, + "CCCAAC": 2695, + "CCCAAG": 2696, + "CCCATA": 2697, + "CCCATT": 2698, + "CCCATC": 2699, + "CCCATG": 2700, + "CCCACA": 2701, + "CCCACT": 2702, + "CCCACC": 2703, + "CCCACG": 2704, + "CCCAGA": 2705, + "CCCAGT": 2706, + "CCCAGC": 2707, + "CCCAGG": 2708, + "CCCTAA": 2709, + "CCCTAT": 2710, + "CCCTAC": 2711, + "CCCTAG": 2712, + "CCCTTA": 2713, + "CCCTTT": 2714, + "CCCTTC": 2715, + "CCCTTG": 2716, + "CCCTCA": 2717, + "CCCTCT": 2718, + "CCCTCC": 2719, + "CCCTCG": 2720, + "CCCTGA": 2721, + "CCCTGT": 2722, + "CCCTGC": 2723, + "CCCTGG": 2724, + "CCCCAA": 2725, + "CCCCAT": 2726, + "CCCCAC": 2727, + "CCCCAG": 2728, + "CCCCTA": 2729, + "CCCCTT": 2730, + "CCCCTC": 2731, + "CCCCTG": 2732, + "CCCCCA": 2733, + "CCCCCT": 2734, + "CCCCCC": 2735, + "CCCCCG": 2736, + "CCCCGA": 2737, + "CCCCGT": 2738, + "CCCCGC": 2739, + "CCCCGG": 2740, + "CCCGAA": 2741, + "CCCGAT": 2742, + "CCCGAC": 2743, + "CCCGAG": 2744, + "CCCGTA": 2745, + "CCCGTT": 2746, + "CCCGTC": 2747, + "CCCGTG": 2748, + "CCCGCA": 2749, + "CCCGCT": 2750, + "CCCGCC": 2751, + "CCCGCG": 2752, + "CCCGGA": 2753, + "CCCGGT": 2754, + "CCCGGC": 2755, + "CCCGGG": 2756, + "CCGAAA": 2757, + "CCGAAT": 2758, + "CCGAAC": 2759, + "CCGAAG": 2760, + "CCGATA": 2761, + "CCGATT": 2762, + "CCGATC": 2763, + "CCGATG": 2764, + "CCGACA": 2765, + "CCGACT": 2766, + "CCGACC": 2767, + "CCGACG": 2768, + "CCGAGA": 2769, + "CCGAGT": 2770, + "CCGAGC": 2771, + "CCGAGG": 2772, + "CCGTAA": 2773, + "CCGTAT": 2774, + "CCGTAC": 2775, + "CCGTAG": 2776, + "CCGTTA": 2777, + "CCGTTT": 2778, + "CCGTTC": 2779, + "CCGTTG": 2780, + "CCGTCA": 2781, + "CCGTCT": 2782, + "CCGTCC": 2783, + "CCGTCG": 2784, + "CCGTGA": 2785, + "CCGTGT": 2786, + "CCGTGC": 2787, + "CCGTGG": 2788, + "CCGCAA": 2789, + "CCGCAT": 2790, + "CCGCAC": 2791, + "CCGCAG": 2792, + "CCGCTA": 2793, + "CCGCTT": 2794, + "CCGCTC": 2795, + "CCGCTG": 2796, + "CCGCCA": 2797, + "CCGCCT": 2798, + "CCGCCC": 2799, + "CCGCCG": 2800, + "CCGCGA": 2801, + "CCGCGT": 2802, + "CCGCGC": 2803, + "CCGCGG": 2804, + "CCGGAA": 2805, + "CCGGAT": 2806, + "CCGGAC": 2807, + "CCGGAG": 2808, + "CCGGTA": 2809, + "CCGGTT": 2810, + "CCGGTC": 2811, + "CCGGTG": 2812, + "CCGGCA": 2813, + "CCGGCT": 2814, + "CCGGCC": 2815, + "CCGGCG": 2816, + "CCGGGA": 2817, + "CCGGGT": 2818, + "CCGGGC": 2819, + "CCGGGG": 2820, + "CGAAAA": 2821, + "CGAAAT": 2822, + "CGAAAC": 2823, + "CGAAAG": 2824, + "CGAATA": 2825, + "CGAATT": 2826, + "CGAATC": 2827, + "CGAATG": 2828, + "CGAACA": 2829, + "CGAACT": 2830, + "CGAACC": 2831, + "CGAACG": 2832, + "CGAAGA": 2833, + "CGAAGT": 2834, + "CGAAGC": 2835, + "CGAAGG": 2836, + "CGATAA": 2837, + "CGATAT": 2838, + "CGATAC": 2839, + "CGATAG": 2840, + "CGATTA": 2841, + "CGATTT": 2842, + "CGATTC": 2843, + "CGATTG": 2844, + "CGATCA": 2845, + "CGATCT": 2846, + "CGATCC": 2847, + "CGATCG": 2848, + "CGATGA": 2849, + "CGATGT": 2850, + "CGATGC": 2851, + "CGATGG": 2852, + "CGACAA": 2853, + "CGACAT": 2854, + "CGACAC": 2855, + "CGACAG": 2856, + "CGACTA": 2857, + "CGACTT": 2858, + "CGACTC": 2859, + "CGACTG": 2860, + "CGACCA": 2861, + "CGACCT": 2862, + "CGACCC": 2863, + "CGACCG": 2864, + "CGACGA": 2865, + "CGACGT": 2866, + "CGACGC": 2867, + "CGACGG": 2868, + "CGAGAA": 2869, + "CGAGAT": 2870, + "CGAGAC": 2871, + "CGAGAG": 2872, + "CGAGTA": 2873, + "CGAGTT": 2874, + "CGAGTC": 2875, + "CGAGTG": 2876, + "CGAGCA": 2877, + "CGAGCT": 2878, + "CGAGCC": 2879, + "CGAGCG": 2880, + "CGAGGA": 2881, + "CGAGGT": 2882, + "CGAGGC": 2883, + "CGAGGG": 2884, + "CGTAAA": 2885, + "CGTAAT": 2886, + "CGTAAC": 2887, + "CGTAAG": 2888, + "CGTATA": 2889, + "CGTATT": 2890, + "CGTATC": 2891, + "CGTATG": 2892, + "CGTACA": 2893, + "CGTACT": 2894, + "CGTACC": 2895, + "CGTACG": 2896, + "CGTAGA": 2897, + "CGTAGT": 2898, + "CGTAGC": 2899, + "CGTAGG": 2900, + "CGTTAA": 2901, + "CGTTAT": 2902, + "CGTTAC": 2903, + "CGTTAG": 2904, + "CGTTTA": 2905, + "CGTTTT": 2906, + "CGTTTC": 2907, + "CGTTTG": 2908, + "CGTTCA": 2909, + "CGTTCT": 2910, + "CGTTCC": 2911, + "CGTTCG": 2912, + "CGTTGA": 2913, + "CGTTGT": 2914, + "CGTTGC": 2915, + "CGTTGG": 2916, + "CGTCAA": 2917, + "CGTCAT": 2918, + "CGTCAC": 2919, + "CGTCAG": 2920, + "CGTCTA": 2921, + "CGTCTT": 2922, + "CGTCTC": 2923, + "CGTCTG": 2924, + "CGTCCA": 2925, + "CGTCCT": 2926, + "CGTCCC": 2927, + "CGTCCG": 2928, + "CGTCGA": 2929, + "CGTCGT": 2930, + "CGTCGC": 2931, + "CGTCGG": 2932, + "CGTGAA": 2933, + "CGTGAT": 2934, + "CGTGAC": 2935, + "CGTGAG": 2936, + "CGTGTA": 2937, + "CGTGTT": 2938, + "CGTGTC": 2939, + "CGTGTG": 2940, + "CGTGCA": 2941, + "CGTGCT": 2942, + "CGTGCC": 2943, + "CGTGCG": 2944, + "CGTGGA": 2945, + "CGTGGT": 2946, + "CGTGGC": 2947, + "CGTGGG": 2948, + "CGCAAA": 2949, + "CGCAAT": 2950, + "CGCAAC": 2951, + "CGCAAG": 2952, + "CGCATA": 2953, + "CGCATT": 2954, + "CGCATC": 2955, + "CGCATG": 2956, + "CGCACA": 2957, + "CGCACT": 2958, + "CGCACC": 2959, + "CGCACG": 2960, + "CGCAGA": 2961, + "CGCAGT": 2962, + "CGCAGC": 2963, + "CGCAGG": 2964, + "CGCTAA": 2965, + "CGCTAT": 2966, + "CGCTAC": 2967, + "CGCTAG": 2968, + "CGCTTA": 2969, + "CGCTTT": 2970, + "CGCTTC": 2971, + "CGCTTG": 2972, + "CGCTCA": 2973, + "CGCTCT": 2974, + "CGCTCC": 2975, + "CGCTCG": 2976, + "CGCTGA": 2977, + "CGCTGT": 2978, + "CGCTGC": 2979, + "CGCTGG": 2980, + "CGCCAA": 2981, + "CGCCAT": 2982, + "CGCCAC": 2983, + "CGCCAG": 2984, + "CGCCTA": 2985, + "CGCCTT": 2986, + "CGCCTC": 2987, + "CGCCTG": 2988, + "CGCCCA": 2989, + "CGCCCT": 2990, + "CGCCCC": 2991, + "CGCCCG": 2992, + "CGCCGA": 2993, + "CGCCGT": 2994, + "CGCCGC": 2995, + "CGCCGG": 2996, + "CGCGAA": 2997, + "CGCGAT": 2998, + "CGCGAC": 2999, + "CGCGAG": 3000, + "CGCGTA": 3001, + "CGCGTT": 3002, + "CGCGTC": 3003, + "CGCGTG": 3004, + "CGCGCA": 3005, + "CGCGCT": 3006, + "CGCGCC": 3007, + "CGCGCG": 3008, + "CGCGGA": 3009, + "CGCGGT": 3010, + "CGCGGC": 3011, + "CGCGGG": 3012, + "CGGAAA": 3013, + "CGGAAT": 3014, + "CGGAAC": 3015, + "CGGAAG": 3016, + "CGGATA": 3017, + "CGGATT": 3018, + "CGGATC": 3019, + "CGGATG": 3020, + "CGGACA": 3021, + "CGGACT": 3022, + "CGGACC": 3023, + "CGGACG": 3024, + "CGGAGA": 3025, + "CGGAGT": 3026, + "CGGAGC": 3027, + "CGGAGG": 3028, + "CGGTAA": 3029, + "CGGTAT": 3030, + "CGGTAC": 3031, + "CGGTAG": 3032, + "CGGTTA": 3033, + "CGGTTT": 3034, + "CGGTTC": 3035, + "CGGTTG": 3036, + "CGGTCA": 3037, + "CGGTCT": 3038, + "CGGTCC": 3039, + "CGGTCG": 3040, + "CGGTGA": 3041, + "CGGTGT": 3042, + "CGGTGC": 3043, + "CGGTGG": 3044, + "CGGCAA": 3045, + "CGGCAT": 3046, + "CGGCAC": 3047, + "CGGCAG": 3048, + "CGGCTA": 3049, + "CGGCTT": 3050, + "CGGCTC": 3051, + "CGGCTG": 3052, + "CGGCCA": 3053, + "CGGCCT": 3054, + "CGGCCC": 3055, + "CGGCCG": 3056, + "CGGCGA": 3057, + "CGGCGT": 3058, + "CGGCGC": 3059, + "CGGCGG": 3060, + "CGGGAA": 3061, + "CGGGAT": 3062, + "CGGGAC": 3063, + "CGGGAG": 3064, + "CGGGTA": 3065, + "CGGGTT": 3066, + "CGGGTC": 3067, + "CGGGTG": 3068, + "CGGGCA": 3069, + "CGGGCT": 3070, + "CGGGCC": 3071, + "CGGGCG": 3072, + "CGGGGA": 3073, + "CGGGGT": 3074, + "CGGGGC": 3075, + "CGGGGG": 3076, + "GAAAAA": 3077, + "GAAAAT": 3078, + "GAAAAC": 3079, + "GAAAAG": 3080, + "GAAATA": 3081, + "GAAATT": 3082, + "GAAATC": 3083, + "GAAATG": 3084, + "GAAACA": 3085, + "GAAACT": 3086, + "GAAACC": 3087, + "GAAACG": 3088, + "GAAAGA": 3089, + "GAAAGT": 3090, + "GAAAGC": 3091, + "GAAAGG": 3092, + "GAATAA": 3093, + "GAATAT": 3094, + "GAATAC": 3095, + "GAATAG": 3096, + "GAATTA": 3097, + "GAATTT": 3098, + "GAATTC": 3099, + "GAATTG": 3100, + "GAATCA": 3101, + "GAATCT": 3102, + "GAATCC": 3103, + "GAATCG": 3104, + "GAATGA": 3105, + "GAATGT": 3106, + "GAATGC": 3107, + "GAATGG": 3108, + "GAACAA": 3109, + "GAACAT": 3110, + "GAACAC": 3111, + "GAACAG": 3112, + "GAACTA": 3113, + "GAACTT": 3114, + "GAACTC": 3115, + "GAACTG": 3116, + "GAACCA": 3117, + "GAACCT": 3118, + "GAACCC": 3119, + "GAACCG": 3120, + "GAACGA": 3121, + "GAACGT": 3122, + "GAACGC": 3123, + "GAACGG": 3124, + "GAAGAA": 3125, + "GAAGAT": 3126, + "GAAGAC": 3127, + "GAAGAG": 3128, + "GAAGTA": 3129, + "GAAGTT": 3130, + "GAAGTC": 3131, + "GAAGTG": 3132, + "GAAGCA": 3133, + "GAAGCT": 3134, + "GAAGCC": 3135, + "GAAGCG": 3136, + "GAAGGA": 3137, + "GAAGGT": 3138, + "GAAGGC": 3139, + "GAAGGG": 3140, + "GATAAA": 3141, + "GATAAT": 3142, + "GATAAC": 3143, + "GATAAG": 3144, + "GATATA": 3145, + "GATATT": 3146, + "GATATC": 3147, + "GATATG": 3148, + "GATACA": 3149, + "GATACT": 3150, + "GATACC": 3151, + "GATACG": 3152, + "GATAGA": 3153, + "GATAGT": 3154, + "GATAGC": 3155, + "GATAGG": 3156, + "GATTAA": 3157, + "GATTAT": 3158, + "GATTAC": 3159, + "GATTAG": 3160, + "GATTTA": 3161, + "GATTTT": 3162, + "GATTTC": 3163, + "GATTTG": 3164, + "GATTCA": 3165, + "GATTCT": 3166, + "GATTCC": 3167, + "GATTCG": 3168, + "GATTGA": 3169, + "GATTGT": 3170, + "GATTGC": 3171, + "GATTGG": 3172, + "GATCAA": 3173, + "GATCAT": 3174, + "GATCAC": 3175, + "GATCAG": 3176, + "GATCTA": 3177, + "GATCTT": 3178, + "GATCTC": 3179, + "GATCTG": 3180, + "GATCCA": 3181, + "GATCCT": 3182, + "GATCCC": 3183, + "GATCCG": 3184, + "GATCGA": 3185, + "GATCGT": 3186, + "GATCGC": 3187, + "GATCGG": 3188, + "GATGAA": 3189, + "GATGAT": 3190, + "GATGAC": 3191, + "GATGAG": 3192, + "GATGTA": 3193, + "GATGTT": 3194, + "GATGTC": 3195, + "GATGTG": 3196, + "GATGCA": 3197, + "GATGCT": 3198, + "GATGCC": 3199, + "GATGCG": 3200, + "GATGGA": 3201, + "GATGGT": 3202, + "GATGGC": 3203, + "GATGGG": 3204, + "GACAAA": 3205, + "GACAAT": 3206, + "GACAAC": 3207, + "GACAAG": 3208, + "GACATA": 3209, + "GACATT": 3210, + "GACATC": 3211, + "GACATG": 3212, + "GACACA": 3213, + "GACACT": 3214, + "GACACC": 3215, + "GACACG": 3216, + "GACAGA": 3217, + "GACAGT": 3218, + "GACAGC": 3219, + "GACAGG": 3220, + "GACTAA": 3221, + "GACTAT": 3222, + "GACTAC": 3223, + "GACTAG": 3224, + "GACTTA": 3225, + "GACTTT": 3226, + "GACTTC": 3227, + "GACTTG": 3228, + "GACTCA": 3229, + "GACTCT": 3230, + "GACTCC": 3231, + "GACTCG": 3232, + "GACTGA": 3233, + "GACTGT": 3234, + "GACTGC": 3235, + "GACTGG": 3236, + "GACCAA": 3237, + "GACCAT": 3238, + "GACCAC": 3239, + "GACCAG": 3240, + "GACCTA": 3241, + "GACCTT": 3242, + "GACCTC": 3243, + "GACCTG": 3244, + "GACCCA": 3245, + "GACCCT": 3246, + "GACCCC": 3247, + "GACCCG": 3248, + "GACCGA": 3249, + "GACCGT": 3250, + "GACCGC": 3251, + "GACCGG": 3252, + "GACGAA": 3253, + "GACGAT": 3254, + "GACGAC": 3255, + "GACGAG": 3256, + "GACGTA": 3257, + "GACGTT": 3258, + "GACGTC": 3259, + "GACGTG": 3260, + "GACGCA": 3261, + "GACGCT": 3262, + "GACGCC": 3263, + "GACGCG": 3264, + "GACGGA": 3265, + "GACGGT": 3266, + "GACGGC": 3267, + "GACGGG": 3268, + "GAGAAA": 3269, + "GAGAAT": 3270, + "GAGAAC": 3271, + "GAGAAG": 3272, + "GAGATA": 3273, + "GAGATT": 3274, + "GAGATC": 3275, + "GAGATG": 3276, + "GAGACA": 3277, + "GAGACT": 3278, + "GAGACC": 3279, + "GAGACG": 3280, + "GAGAGA": 3281, + "GAGAGT": 3282, + "GAGAGC": 3283, + "GAGAGG": 3284, + "GAGTAA": 3285, + "GAGTAT": 3286, + "GAGTAC": 3287, + "GAGTAG": 3288, + "GAGTTA": 3289, + "GAGTTT": 3290, + "GAGTTC": 3291, + "GAGTTG": 3292, + "GAGTCA": 3293, + "GAGTCT": 3294, + "GAGTCC": 3295, + "GAGTCG": 3296, + "GAGTGA": 3297, + "GAGTGT": 3298, + "GAGTGC": 3299, + "GAGTGG": 3300, + "GAGCAA": 3301, + "GAGCAT": 3302, + "GAGCAC": 3303, + "GAGCAG": 3304, + "GAGCTA": 3305, + "GAGCTT": 3306, + "GAGCTC": 3307, + "GAGCTG": 3308, + "GAGCCA": 3309, + "GAGCCT": 3310, + "GAGCCC": 3311, + "GAGCCG": 3312, + "GAGCGA": 3313, + "GAGCGT": 3314, + "GAGCGC": 3315, + "GAGCGG": 3316, + "GAGGAA": 3317, + "GAGGAT": 3318, + "GAGGAC": 3319, + "GAGGAG": 3320, + "GAGGTA": 3321, + "GAGGTT": 3322, + "GAGGTC": 3323, + "GAGGTG": 3324, + "GAGGCA": 3325, + "GAGGCT": 3326, + "GAGGCC": 3327, + "GAGGCG": 3328, + "GAGGGA": 3329, + "GAGGGT": 3330, + "GAGGGC": 3331, + "GAGGGG": 3332, + "GTAAAA": 3333, + "GTAAAT": 3334, + "GTAAAC": 3335, + "GTAAAG": 3336, + "GTAATA": 3337, + "GTAATT": 3338, + "GTAATC": 3339, + "GTAATG": 3340, + "GTAACA": 3341, + "GTAACT": 3342, + "GTAACC": 3343, + "GTAACG": 3344, + "GTAAGA": 3345, + "GTAAGT": 3346, + "GTAAGC": 3347, + "GTAAGG": 3348, + "GTATAA": 3349, + "GTATAT": 3350, + "GTATAC": 3351, + "GTATAG": 3352, + "GTATTA": 3353, + "GTATTT": 3354, + "GTATTC": 3355, + "GTATTG": 3356, + "GTATCA": 3357, + "GTATCT": 3358, + "GTATCC": 3359, + "GTATCG": 3360, + "GTATGA": 3361, + "GTATGT": 3362, + "GTATGC": 3363, + "GTATGG": 3364, + "GTACAA": 3365, + "GTACAT": 3366, + "GTACAC": 3367, + "GTACAG": 3368, + "GTACTA": 3369, + "GTACTT": 3370, + "GTACTC": 3371, + "GTACTG": 3372, + "GTACCA": 3373, + "GTACCT": 3374, + "GTACCC": 3375, + "GTACCG": 3376, + "GTACGA": 3377, + "GTACGT": 3378, + "GTACGC": 3379, + "GTACGG": 3380, + "GTAGAA": 3381, + "GTAGAT": 3382, + "GTAGAC": 3383, + "GTAGAG": 3384, + "GTAGTA": 3385, + "GTAGTT": 3386, + "GTAGTC": 3387, + "GTAGTG": 3388, + "GTAGCA": 3389, + "GTAGCT": 3390, + "GTAGCC": 3391, + "GTAGCG": 3392, + "GTAGGA": 3393, + "GTAGGT": 3394, + "GTAGGC": 3395, + "GTAGGG": 3396, + "GTTAAA": 3397, + "GTTAAT": 3398, + "GTTAAC": 3399, + "GTTAAG": 3400, + "GTTATA": 3401, + "GTTATT": 3402, + "GTTATC": 3403, + "GTTATG": 3404, + "GTTACA": 3405, + "GTTACT": 3406, + "GTTACC": 3407, + "GTTACG": 3408, + "GTTAGA": 3409, + "GTTAGT": 3410, + "GTTAGC": 3411, + "GTTAGG": 3412, + "GTTTAA": 3413, + "GTTTAT": 3414, + "GTTTAC": 3415, + "GTTTAG": 3416, + "GTTTTA": 3417, + "GTTTTT": 3418, + "GTTTTC": 3419, + "GTTTTG": 3420, + "GTTTCA": 3421, + "GTTTCT": 3422, + "GTTTCC": 3423, + "GTTTCG": 3424, + "GTTTGA": 3425, + "GTTTGT": 3426, + "GTTTGC": 3427, + "GTTTGG": 3428, + "GTTCAA": 3429, + "GTTCAT": 3430, + "GTTCAC": 3431, + "GTTCAG": 3432, + "GTTCTA": 3433, + "GTTCTT": 3434, + "GTTCTC": 3435, + "GTTCTG": 3436, + "GTTCCA": 3437, + "GTTCCT": 3438, + "GTTCCC": 3439, + "GTTCCG": 3440, + "GTTCGA": 3441, + "GTTCGT": 3442, + "GTTCGC": 3443, + "GTTCGG": 3444, + "GTTGAA": 3445, + "GTTGAT": 3446, + "GTTGAC": 3447, + "GTTGAG": 3448, + "GTTGTA": 3449, + "GTTGTT": 3450, + "GTTGTC": 3451, + "GTTGTG": 3452, + "GTTGCA": 3453, + "GTTGCT": 3454, + "GTTGCC": 3455, + "GTTGCG": 3456, + "GTTGGA": 3457, + "GTTGGT": 3458, + "GTTGGC": 3459, + "GTTGGG": 3460, + "GTCAAA": 3461, + "GTCAAT": 3462, + "GTCAAC": 3463, + "GTCAAG": 3464, + "GTCATA": 3465, + "GTCATT": 3466, + "GTCATC": 3467, + "GTCATG": 3468, + "GTCACA": 3469, + "GTCACT": 3470, + "GTCACC": 3471, + "GTCACG": 3472, + "GTCAGA": 3473, + "GTCAGT": 3474, + "GTCAGC": 3475, + "GTCAGG": 3476, + "GTCTAA": 3477, + "GTCTAT": 3478, + "GTCTAC": 3479, + "GTCTAG": 3480, + "GTCTTA": 3481, + "GTCTTT": 3482, + "GTCTTC": 3483, + "GTCTTG": 3484, + "GTCTCA": 3485, + "GTCTCT": 3486, + "GTCTCC": 3487, + "GTCTCG": 3488, + "GTCTGA": 3489, + "GTCTGT": 3490, + "GTCTGC": 3491, + "GTCTGG": 3492, + "GTCCAA": 3493, + "GTCCAT": 3494, + "GTCCAC": 3495, + "GTCCAG": 3496, + "GTCCTA": 3497, + "GTCCTT": 3498, + "GTCCTC": 3499, + "GTCCTG": 3500, + "GTCCCA": 3501, + "GTCCCT": 3502, + "GTCCCC": 3503, + "GTCCCG": 3504, + "GTCCGA": 3505, + "GTCCGT": 3506, + "GTCCGC": 3507, + "GTCCGG": 3508, + "GTCGAA": 3509, + "GTCGAT": 3510, + "GTCGAC": 3511, + "GTCGAG": 3512, + "GTCGTA": 3513, + "GTCGTT": 3514, + "GTCGTC": 3515, + "GTCGTG": 3516, + "GTCGCA": 3517, + "GTCGCT": 3518, + "GTCGCC": 3519, + "GTCGCG": 3520, + "GTCGGA": 3521, + "GTCGGT": 3522, + "GTCGGC": 3523, + "GTCGGG": 3524, + "GTGAAA": 3525, + "GTGAAT": 3526, + "GTGAAC": 3527, + "GTGAAG": 3528, + "GTGATA": 3529, + "GTGATT": 3530, + "GTGATC": 3531, + "GTGATG": 3532, + "GTGACA": 3533, + "GTGACT": 3534, + "GTGACC": 3535, + "GTGACG": 3536, + "GTGAGA": 3537, + "GTGAGT": 3538, + "GTGAGC": 3539, + "GTGAGG": 3540, + "GTGTAA": 3541, + "GTGTAT": 3542, + "GTGTAC": 3543, + "GTGTAG": 3544, + "GTGTTA": 3545, + "GTGTTT": 3546, + "GTGTTC": 3547, + "GTGTTG": 3548, + "GTGTCA": 3549, + "GTGTCT": 3550, + "GTGTCC": 3551, + "GTGTCG": 3552, + "GTGTGA": 3553, + "GTGTGT": 3554, + "GTGTGC": 3555, + "GTGTGG": 3556, + "GTGCAA": 3557, + "GTGCAT": 3558, + "GTGCAC": 3559, + "GTGCAG": 3560, + "GTGCTA": 3561, + "GTGCTT": 3562, + "GTGCTC": 3563, + "GTGCTG": 3564, + "GTGCCA": 3565, + "GTGCCT": 3566, + "GTGCCC": 3567, + "GTGCCG": 3568, + "GTGCGA": 3569, + "GTGCGT": 3570, + "GTGCGC": 3571, + "GTGCGG": 3572, + "GTGGAA": 3573, + "GTGGAT": 3574, + "GTGGAC": 3575, + "GTGGAG": 3576, + "GTGGTA": 3577, + "GTGGTT": 3578, + "GTGGTC": 3579, + "GTGGTG": 3580, + "GTGGCA": 3581, + "GTGGCT": 3582, + "GTGGCC": 3583, + "GTGGCG": 3584, + "GTGGGA": 3585, + "GTGGGT": 3586, + "GTGGGC": 3587, + "GTGGGG": 3588, + "GCAAAA": 3589, + "GCAAAT": 3590, + "GCAAAC": 3591, + "GCAAAG": 3592, + "GCAATA": 3593, + "GCAATT": 3594, + "GCAATC": 3595, + "GCAATG": 3596, + "GCAACA": 3597, + "GCAACT": 3598, + "GCAACC": 3599, + "GCAACG": 3600, + "GCAAGA": 3601, + "GCAAGT": 3602, + "GCAAGC": 3603, + "GCAAGG": 3604, + "GCATAA": 3605, + "GCATAT": 3606, + "GCATAC": 3607, + "GCATAG": 3608, + "GCATTA": 3609, + "GCATTT": 3610, + "GCATTC": 3611, + "GCATTG": 3612, + "GCATCA": 3613, + "GCATCT": 3614, + "GCATCC": 3615, + "GCATCG": 3616, + "GCATGA": 3617, + "GCATGT": 3618, + "GCATGC": 3619, + "GCATGG": 3620, + "GCACAA": 3621, + "GCACAT": 3622, + "GCACAC": 3623, + "GCACAG": 3624, + "GCACTA": 3625, + "GCACTT": 3626, + "GCACTC": 3627, + "GCACTG": 3628, + "GCACCA": 3629, + "GCACCT": 3630, + "GCACCC": 3631, + "GCACCG": 3632, + "GCACGA": 3633, + "GCACGT": 3634, + "GCACGC": 3635, + "GCACGG": 3636, + "GCAGAA": 3637, + "GCAGAT": 3638, + "GCAGAC": 3639, + "GCAGAG": 3640, + "GCAGTA": 3641, + "GCAGTT": 3642, + "GCAGTC": 3643, + "GCAGTG": 3644, + "GCAGCA": 3645, + "GCAGCT": 3646, + "GCAGCC": 3647, + "GCAGCG": 3648, + "GCAGGA": 3649, + "GCAGGT": 3650, + "GCAGGC": 3651, + "GCAGGG": 3652, + "GCTAAA": 3653, + "GCTAAT": 3654, + "GCTAAC": 3655, + "GCTAAG": 3656, + "GCTATA": 3657, + "GCTATT": 3658, + "GCTATC": 3659, + "GCTATG": 3660, + "GCTACA": 3661, + "GCTACT": 3662, + "GCTACC": 3663, + "GCTACG": 3664, + "GCTAGA": 3665, + "GCTAGT": 3666, + "GCTAGC": 3667, + "GCTAGG": 3668, + "GCTTAA": 3669, + "GCTTAT": 3670, + "GCTTAC": 3671, + "GCTTAG": 3672, + "GCTTTA": 3673, + "GCTTTT": 3674, + "GCTTTC": 3675, + "GCTTTG": 3676, + "GCTTCA": 3677, + "GCTTCT": 3678, + "GCTTCC": 3679, + "GCTTCG": 3680, + "GCTTGA": 3681, + "GCTTGT": 3682, + "GCTTGC": 3683, + "GCTTGG": 3684, + "GCTCAA": 3685, + "GCTCAT": 3686, + "GCTCAC": 3687, + "GCTCAG": 3688, + "GCTCTA": 3689, + "GCTCTT": 3690, + "GCTCTC": 3691, + "GCTCTG": 3692, + "GCTCCA": 3693, + "GCTCCT": 3694, + "GCTCCC": 3695, + "GCTCCG": 3696, + "GCTCGA": 3697, + "GCTCGT": 3698, + "GCTCGC": 3699, + "GCTCGG": 3700, + "GCTGAA": 3701, + "GCTGAT": 3702, + "GCTGAC": 3703, + "GCTGAG": 3704, + "GCTGTA": 3705, + "GCTGTT": 3706, + "GCTGTC": 3707, + "GCTGTG": 3708, + "GCTGCA": 3709, + "GCTGCT": 3710, + "GCTGCC": 3711, + "GCTGCG": 3712, + "GCTGGA": 3713, + "GCTGGT": 3714, + "GCTGGC": 3715, + "GCTGGG": 3716, + "GCCAAA": 3717, + "GCCAAT": 3718, + "GCCAAC": 3719, + "GCCAAG": 3720, + "GCCATA": 3721, + "GCCATT": 3722, + "GCCATC": 3723, + "GCCATG": 3724, + "GCCACA": 3725, + "GCCACT": 3726, + "GCCACC": 3727, + "GCCACG": 3728, + "GCCAGA": 3729, + "GCCAGT": 3730, + "GCCAGC": 3731, + "GCCAGG": 3732, + "GCCTAA": 3733, + "GCCTAT": 3734, + "GCCTAC": 3735, + "GCCTAG": 3736, + "GCCTTA": 3737, + "GCCTTT": 3738, + "GCCTTC": 3739, + "GCCTTG": 3740, + "GCCTCA": 3741, + "GCCTCT": 3742, + "GCCTCC": 3743, + "GCCTCG": 3744, + "GCCTGA": 3745, + "GCCTGT": 3746, + "GCCTGC": 3747, + "GCCTGG": 3748, + "GCCCAA": 3749, + "GCCCAT": 3750, + "GCCCAC": 3751, + "GCCCAG": 3752, + "GCCCTA": 3753, + "GCCCTT": 3754, + "GCCCTC": 3755, + "GCCCTG": 3756, + "GCCCCA": 3757, + "GCCCCT": 3758, + "GCCCCC": 3759, + "GCCCCG": 3760, + "GCCCGA": 3761, + "GCCCGT": 3762, + "GCCCGC": 3763, + "GCCCGG": 3764, + "GCCGAA": 3765, + "GCCGAT": 3766, + "GCCGAC": 3767, + "GCCGAG": 3768, + "GCCGTA": 3769, + "GCCGTT": 3770, + "GCCGTC": 3771, + "GCCGTG": 3772, + "GCCGCA": 3773, + "GCCGCT": 3774, + "GCCGCC": 3775, + "GCCGCG": 3776, + "GCCGGA": 3777, + "GCCGGT": 3778, + "GCCGGC": 3779, + "GCCGGG": 3780, + "GCGAAA": 3781, + "GCGAAT": 3782, + "GCGAAC": 3783, + "GCGAAG": 3784, + "GCGATA": 3785, + "GCGATT": 3786, + "GCGATC": 3787, + "GCGATG": 3788, + "GCGACA": 3789, + "GCGACT": 3790, + "GCGACC": 3791, + "GCGACG": 3792, + "GCGAGA": 3793, + "GCGAGT": 3794, + "GCGAGC": 3795, + "GCGAGG": 3796, + "GCGTAA": 3797, + "GCGTAT": 3798, + "GCGTAC": 3799, + "GCGTAG": 3800, + "GCGTTA": 3801, + "GCGTTT": 3802, + "GCGTTC": 3803, + "GCGTTG": 3804, + "GCGTCA": 3805, + "GCGTCT": 3806, + "GCGTCC": 3807, + "GCGTCG": 3808, + "GCGTGA": 3809, + "GCGTGT": 3810, + "GCGTGC": 3811, + "GCGTGG": 3812, + "GCGCAA": 3813, + "GCGCAT": 3814, + "GCGCAC": 3815, + "GCGCAG": 3816, + "GCGCTA": 3817, + "GCGCTT": 3818, + "GCGCTC": 3819, + "GCGCTG": 3820, + "GCGCCA": 3821, + "GCGCCT": 3822, + "GCGCCC": 3823, + "GCGCCG": 3824, + "GCGCGA": 3825, + "GCGCGT": 3826, + "GCGCGC": 3827, + "GCGCGG": 3828, + "GCGGAA": 3829, + "GCGGAT": 3830, + "GCGGAC": 3831, + "GCGGAG": 3832, + "GCGGTA": 3833, + "GCGGTT": 3834, + "GCGGTC": 3835, + "GCGGTG": 3836, + "GCGGCA": 3837, + "GCGGCT": 3838, + "GCGGCC": 3839, + "GCGGCG": 3840, + "GCGGGA": 3841, + "GCGGGT": 3842, + "GCGGGC": 3843, + "GCGGGG": 3844, + "GGAAAA": 3845, + "GGAAAT": 3846, + "GGAAAC": 3847, + "GGAAAG": 3848, + "GGAATA": 3849, + "GGAATT": 3850, + "GGAATC": 3851, + "GGAATG": 3852, + "GGAACA": 3853, + "GGAACT": 3854, + "GGAACC": 3855, + "GGAACG": 3856, + "GGAAGA": 3857, + "GGAAGT": 3858, + "GGAAGC": 3859, + "GGAAGG": 3860, + "GGATAA": 3861, + "GGATAT": 3862, + "GGATAC": 3863, + "GGATAG": 3864, + "GGATTA": 3865, + "GGATTT": 3866, + "GGATTC": 3867, + "GGATTG": 3868, + "GGATCA": 3869, + "GGATCT": 3870, + "GGATCC": 3871, + "GGATCG": 3872, + "GGATGA": 3873, + "GGATGT": 3874, + "GGATGC": 3875, + "GGATGG": 3876, + "GGACAA": 3877, + "GGACAT": 3878, + "GGACAC": 3879, + "GGACAG": 3880, + "GGACTA": 3881, + "GGACTT": 3882, + "GGACTC": 3883, + "GGACTG": 3884, + "GGACCA": 3885, + "GGACCT": 3886, + "GGACCC": 3887, + "GGACCG": 3888, + "GGACGA": 3889, + "GGACGT": 3890, + "GGACGC": 3891, + "GGACGG": 3892, + "GGAGAA": 3893, + "GGAGAT": 3894, + "GGAGAC": 3895, + "GGAGAG": 3896, + "GGAGTA": 3897, + "GGAGTT": 3898, + "GGAGTC": 3899, + "GGAGTG": 3900, + "GGAGCA": 3901, + "GGAGCT": 3902, + "GGAGCC": 3903, + "GGAGCG": 3904, + "GGAGGA": 3905, + "GGAGGT": 3906, + "GGAGGC": 3907, + "GGAGGG": 3908, + "GGTAAA": 3909, + "GGTAAT": 3910, + "GGTAAC": 3911, + "GGTAAG": 3912, + "GGTATA": 3913, + "GGTATT": 3914, + "GGTATC": 3915, + "GGTATG": 3916, + "GGTACA": 3917, + "GGTACT": 3918, + "GGTACC": 3919, + "GGTACG": 3920, + "GGTAGA": 3921, + "GGTAGT": 3922, + "GGTAGC": 3923, + "GGTAGG": 3924, + "GGTTAA": 3925, + "GGTTAT": 3926, + "GGTTAC": 3927, + "GGTTAG": 3928, + "GGTTTA": 3929, + "GGTTTT": 3930, + "GGTTTC": 3931, + "GGTTTG": 3932, + "GGTTCA": 3933, + "GGTTCT": 3934, + "GGTTCC": 3935, + "GGTTCG": 3936, + "GGTTGA": 3937, + "GGTTGT": 3938, + "GGTTGC": 3939, + "GGTTGG": 3940, + "GGTCAA": 3941, + "GGTCAT": 3942, + "GGTCAC": 3943, + "GGTCAG": 3944, + "GGTCTA": 3945, + "GGTCTT": 3946, + "GGTCTC": 3947, + "GGTCTG": 3948, + "GGTCCA": 3949, + "GGTCCT": 3950, + "GGTCCC": 3951, + "GGTCCG": 3952, + "GGTCGA": 3953, + "GGTCGT": 3954, + "GGTCGC": 3955, + "GGTCGG": 3956, + "GGTGAA": 3957, + "GGTGAT": 3958, + "GGTGAC": 3959, + "GGTGAG": 3960, + "GGTGTA": 3961, + "GGTGTT": 3962, + "GGTGTC": 3963, + "GGTGTG": 3964, + "GGTGCA": 3965, + "GGTGCT": 3966, + "GGTGCC": 3967, + "GGTGCG": 3968, + "GGTGGA": 3969, + "GGTGGT": 3970, + "GGTGGC": 3971, + "GGTGGG": 3972, + "GGCAAA": 3973, + "GGCAAT": 3974, + "GGCAAC": 3975, + "GGCAAG": 3976, + "GGCATA": 3977, + "GGCATT": 3978, + "GGCATC": 3979, + "GGCATG": 3980, + "GGCACA": 3981, + "GGCACT": 3982, + "GGCACC": 3983, + "GGCACG": 3984, + "GGCAGA": 3985, + "GGCAGT": 3986, + "GGCAGC": 3987, + "GGCAGG": 3988, + "GGCTAA": 3989, + "GGCTAT": 3990, + "GGCTAC": 3991, + "GGCTAG": 3992, + "GGCTTA": 3993, + "GGCTTT": 3994, + "GGCTTC": 3995, + "GGCTTG": 3996, + "GGCTCA": 3997, + "GGCTCT": 3998, + "GGCTCC": 3999, + "GGCTCG": 4000, + "GGCTGA": 4001, + "GGCTGT": 4002, + "GGCTGC": 4003, + "GGCTGG": 4004, + "GGCCAA": 4005, + "GGCCAT": 4006, + "GGCCAC": 4007, + "GGCCAG": 4008, + "GGCCTA": 4009, + "GGCCTT": 4010, + "GGCCTC": 4011, + "GGCCTG": 4012, + "GGCCCA": 4013, + "GGCCCT": 4014, + "GGCCCC": 4015, + "GGCCCG": 4016, + "GGCCGA": 4017, + "GGCCGT": 4018, + "GGCCGC": 4019, + "GGCCGG": 4020, + "GGCGAA": 4021, + "GGCGAT": 4022, + "GGCGAC": 4023, + "GGCGAG": 4024, + "GGCGTA": 4025, + "GGCGTT": 4026, + "GGCGTC": 4027, + "GGCGTG": 4028, + "GGCGCA": 4029, + "GGCGCT": 4030, + "GGCGCC": 4031, + "GGCGCG": 4032, + "GGCGGA": 4033, + "GGCGGT": 4034, + "GGCGGC": 4035, + "GGCGGG": 4036, + "GGGAAA": 4037, + "GGGAAT": 4038, + "GGGAAC": 4039, + "GGGAAG": 4040, + "GGGATA": 4041, + "GGGATT": 4042, + "GGGATC": 4043, + "GGGATG": 4044, + "GGGACA": 4045, + "GGGACT": 4046, + "GGGACC": 4047, + "GGGACG": 4048, + "GGGAGA": 4049, + "GGGAGT": 4050, + "GGGAGC": 4051, + "GGGAGG": 4052, + "GGGTAA": 4053, + "GGGTAT": 4054, + "GGGTAC": 4055, + "GGGTAG": 4056, + "GGGTTA": 4057, + "GGGTTT": 4058, + "GGGTTC": 4059, + "GGGTTG": 4060, + "GGGTCA": 4061, + "GGGTCT": 4062, + "GGGTCC": 4063, + "GGGTCG": 4064, + "GGGTGA": 4065, + "GGGTGT": 4066, + "GGGTGC": 4067, + "GGGTGG": 4068, + "GGGCAA": 4069, + "GGGCAT": 4070, + "GGGCAC": 4071, + "GGGCAG": 4072, + "GGGCTA": 4073, + "GGGCTT": 4074, + "GGGCTC": 4075, + "GGGCTG": 4076, + "GGGCCA": 4077, + "GGGCCT": 4078, + "GGGCCC": 4079, + "GGGCCG": 4080, + "GGGCGA": 4081, + "GGGCGT": 4082, + "GGGCGC": 4083, + "GGGCGG": 4084, + "GGGGAA": 4085, + "GGGGAT": 4086, + "GGGGAC": 4087, + "GGGGAG": 4088, + "GGGGTA": 4089, + "GGGGTT": 4090, + "GGGGTC": 4091, + "GGGGTG": 4092, + "GGGGCA": 4093, + "GGGGCT": 4094, + "GGGGCC": 4095, + "GGGGCG": 4096, + "GGGGGA": 4097, + "GGGGGT": 4098, + "GGGGGC": 4099, + "GGGGGG": 4100 + } + } +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/tokenizer_config.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..604e2b1b29af1dcb3920d2828b6a90276e5e3d02 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/tokenizer_config.json @@ -0,0 +1,58 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "[UNK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "[CLS]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "[SEP]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "4": { + "content": "[MASK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "clean_up_tokenization_spaces": true, + "cls_token": "[CLS]", + "do_basic_tokenize": true, + "do_lower_case": false, + "mask_token": "[MASK]", + "max_len": 512, + "model_max_length": 512, + "never_split": null, + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "strip_accents": null, + "tokenize_chinese_chars": true, + "tokenizer_class": "BertTokenizer", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/trainer_state.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..583c9a1a4737d47c8a75693391d5b530699df0fe --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/trainer_state.json @@ -0,0 +1,1153 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 16.0, + "eval_steps": 500, + "global_step": 4000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.1, + "grad_norm": 0.2192622423171997, + "learning_rate": 5.319148936170213e-05, + "loss": 0.6949, + "step": 25 + }, + { + "epoch": 0.2, + "grad_norm": 0.22659960389137268, + "learning_rate": 0.00010638297872340425, + "loss": 0.6926, + "step": 50 + }, + { + "epoch": 0.3, + "grad_norm": 0.33277687430381775, + "learning_rate": 0.00015957446808510637, + "loss": 0.693, + "step": 75 + }, + { + "epoch": 0.4, + "grad_norm": 0.5183415412902832, + "learning_rate": 0.0002127659574468085, + "loss": 0.6928, + "step": 100 + }, + { + "epoch": 0.5, + "grad_norm": 0.21227706968784332, + "learning_rate": 0.00026595744680851064, + "loss": 0.6938, + "step": 125 + }, + { + "epoch": 0.6, + "grad_norm": 0.1846485733985901, + "learning_rate": 0.00031914893617021275, + "loss": 0.6942, + "step": 150 + }, + { + "epoch": 0.7, + "grad_norm": 0.27858808636665344, + "learning_rate": 0.0003723404255319149, + "loss": 0.6931, + "step": 175 + }, + { + "epoch": 0.8, + "grad_norm": 0.26012367010116577, + "learning_rate": 0.0003992081821181128, + "loss": 0.6936, + "step": 200 + }, + { + "epoch": 0.9, + "grad_norm": 0.2534612715244293, + "learning_rate": 0.0003975585615308479, + "loss": 0.6939, + "step": 225 + }, + { + "epoch": 1.0, + "grad_norm": 0.42967700958251953, + "learning_rate": 0.00039590894094358297, + "loss": 0.6928, + "step": 250 + }, + { + "epoch": 1.1, + "grad_norm": 0.2613862454891205, + "learning_rate": 0.0003942593203563181, + "loss": 0.6918, + "step": 275 + }, + { + "epoch": 1.2, + "grad_norm": 0.5514339208602905, + "learning_rate": 0.0003926096997690532, + "loss": 0.6918, + "step": 300 + }, + { + "epoch": 1.3, + "grad_norm": 0.33801883459091187, + "learning_rate": 0.00039096007918178817, + "loss": 0.6906, + "step": 325 + }, + { + "epoch": 1.4, + "grad_norm": 0.2953580617904663, + "learning_rate": 0.0003893104585945233, + "loss": 0.6886, + "step": 350 + }, + { + "epoch": 1.5, + "grad_norm": 0.5340980887413025, + "learning_rate": 0.0003876608380072583, + "loss": 0.6877, + "step": 375 + }, + { + "epoch": 1.6, + "grad_norm": 0.3042701482772827, + "learning_rate": 0.00038601121741999343, + "loss": 0.6853, + "step": 400 + }, + { + "epoch": 1.7, + "grad_norm": 0.44497063755989075, + "learning_rate": 0.0003843615968327285, + "loss": 0.6772, + "step": 425 + }, + { + "epoch": 1.8, + "grad_norm": 0.6260074973106384, + "learning_rate": 0.0003827119762454636, + "loss": 0.6671, + "step": 450 + }, + { + "epoch": 1.9, + "grad_norm": 0.9574269652366638, + "learning_rate": 0.00038106235565819863, + "loss": 0.6517, + "step": 475 + }, + { + "epoch": 2.0, + "grad_norm": 0.9055777192115784, + "learning_rate": 0.0003794127350709337, + "loss": 0.6438, + "step": 500 + }, + { + "epoch": 2.1, + "grad_norm": 0.7860134840011597, + "learning_rate": 0.0003777631144836688, + "loss": 0.6379, + "step": 525 + }, + { + "epoch": 2.2, + "grad_norm": 1.5324610471725464, + "learning_rate": 0.00037611349389640383, + "loss": 0.6143, + "step": 550 + }, + { + "epoch": 2.3, + "grad_norm": 1.8535794019699097, + "learning_rate": 0.00037446387330913894, + "loss": 0.607, + "step": 575 + }, + { + "epoch": 2.4, + "grad_norm": 1.8850122690200806, + "learning_rate": 0.000372814252721874, + "loss": 0.5642, + "step": 600 + }, + { + "epoch": 2.5, + "grad_norm": 1.288278579711914, + "learning_rate": 0.00037116463213460903, + "loss": 0.5739, + "step": 625 + }, + { + "epoch": 2.6, + "grad_norm": 1.344173550605774, + "learning_rate": 0.00036951501154734414, + "loss": 0.5718, + "step": 650 + }, + { + "epoch": 2.7, + "grad_norm": 1.430834174156189, + "learning_rate": 0.0003678653909600792, + "loss": 0.543, + "step": 675 + }, + { + "epoch": 2.8, + "grad_norm": 1.5948325395584106, + "learning_rate": 0.0003662157703728143, + "loss": 0.5434, + "step": 700 + }, + { + "epoch": 2.9, + "grad_norm": 2.688519239425659, + "learning_rate": 0.00036456614978554934, + "loss": 0.5842, + "step": 725 + }, + { + "epoch": 3.0, + "grad_norm": 1.6625819206237793, + "learning_rate": 0.00036291652919828444, + "loss": 0.5526, + "step": 750 + }, + { + "epoch": 3.1, + "grad_norm": 1.869725227355957, + "learning_rate": 0.00036126690861101944, + "loss": 0.53, + "step": 775 + }, + { + "epoch": 3.2, + "grad_norm": 2.1625709533691406, + "learning_rate": 0.00035961728802375454, + "loss": 0.5341, + "step": 800 + }, + { + "epoch": 3.3, + "grad_norm": 1.9650202989578247, + "learning_rate": 0.0003579676674364896, + "loss": 0.5297, + "step": 825 + }, + { + "epoch": 3.4, + "grad_norm": 1.3358676433563232, + "learning_rate": 0.0003563180468492247, + "loss": 0.5273, + "step": 850 + }, + { + "epoch": 3.5, + "grad_norm": 2.1927096843719482, + "learning_rate": 0.0003546684262619598, + "loss": 0.5084, + "step": 875 + }, + { + "epoch": 3.6, + "grad_norm": 2.2377188205718994, + "learning_rate": 0.00035301880567469485, + "loss": 0.4972, + "step": 900 + }, + { + "epoch": 3.7, + "grad_norm": 1.8601839542388916, + "learning_rate": 0.0003513691850874299, + "loss": 0.4852, + "step": 925 + }, + { + "epoch": 3.8, + "grad_norm": 1.4995975494384766, + "learning_rate": 0.00034971956450016495, + "loss": 0.5309, + "step": 950 + }, + { + "epoch": 3.9, + "grad_norm": 2.0536930561065674, + "learning_rate": 0.00034806994391290005, + "loss": 0.5249, + "step": 975 + }, + { + "epoch": 4.0, + "grad_norm": 2.084728479385376, + "learning_rate": 0.0003464203233256351, + "loss": 0.4771, + "step": 1000 + }, + { + "epoch": 4.1, + "grad_norm": 1.705510139465332, + "learning_rate": 0.0003447707027383702, + "loss": 0.5181, + "step": 1025 + }, + { + "epoch": 4.2, + "grad_norm": 3.83380389213562, + "learning_rate": 0.00034312108215110525, + "loss": 0.4873, + "step": 1050 + }, + { + "epoch": 4.3, + "grad_norm": 2.0163533687591553, + "learning_rate": 0.0003414714615638403, + "loss": 0.4801, + "step": 1075 + }, + { + "epoch": 4.4, + "grad_norm": 1.9917407035827637, + "learning_rate": 0.0003398218409765754, + "loss": 0.4766, + "step": 1100 + }, + { + "epoch": 4.5, + "grad_norm": 2.111729860305786, + "learning_rate": 0.00033817222038931045, + "loss": 0.4877, + "step": 1125 + }, + { + "epoch": 4.6, + "grad_norm": 1.5487738847732544, + "learning_rate": 0.00033652259980204556, + "loss": 0.507, + "step": 1150 + }, + { + "epoch": 4.7, + "grad_norm": 1.8715145587921143, + "learning_rate": 0.0003348729792147806, + "loss": 0.452, + "step": 1175 + }, + { + "epoch": 4.8, + "grad_norm": 2.2618906497955322, + "learning_rate": 0.0003332233586275157, + "loss": 0.4799, + "step": 1200 + }, + { + "epoch": 4.9, + "grad_norm": 1.657012939453125, + "learning_rate": 0.00033157373804025076, + "loss": 0.4715, + "step": 1225 + }, + { + "epoch": 5.0, + "grad_norm": 2.8894591331481934, + "learning_rate": 0.0003299241174529858, + "loss": 0.4437, + "step": 1250 + }, + { + "epoch": 5.1, + "grad_norm": 1.9179915189743042, + "learning_rate": 0.0003282744968657209, + "loss": 0.4796, + "step": 1275 + }, + { + "epoch": 5.2, + "grad_norm": 2.1816720962524414, + "learning_rate": 0.00032662487627845596, + "loss": 0.4463, + "step": 1300 + }, + { + "epoch": 5.3, + "grad_norm": 2.2761082649230957, + "learning_rate": 0.00032497525569119106, + "loss": 0.4215, + "step": 1325 + }, + { + "epoch": 5.4, + "grad_norm": 1.7190684080123901, + "learning_rate": 0.0003233256351039261, + "loss": 0.4481, + "step": 1350 + }, + { + "epoch": 5.5, + "grad_norm": 1.4793736934661865, + "learning_rate": 0.00032167601451666116, + "loss": 0.44, + "step": 1375 + }, + { + "epoch": 5.6, + "grad_norm": 2.5064358711242676, + "learning_rate": 0.00032002639392939627, + "loss": 0.4583, + "step": 1400 + }, + { + "epoch": 5.7, + "grad_norm": 1.8364307880401611, + "learning_rate": 0.0003183767733421313, + "loss": 0.4516, + "step": 1425 + }, + { + "epoch": 5.8, + "grad_norm": 1.3747583627700806, + "learning_rate": 0.0003167271527548664, + "loss": 0.4535, + "step": 1450 + }, + { + "epoch": 5.9, + "grad_norm": 2.580183982849121, + "learning_rate": 0.00031507753216760147, + "loss": 0.4325, + "step": 1475 + }, + { + "epoch": 6.0, + "grad_norm": 1.562374472618103, + "learning_rate": 0.00031342791158033657, + "loss": 0.4616, + "step": 1500 + }, + { + "epoch": 6.1, + "grad_norm": 1.4970141649246216, + "learning_rate": 0.00031177829099307157, + "loss": 0.4607, + "step": 1525 + }, + { + "epoch": 6.2, + "grad_norm": 2.436326503753662, + "learning_rate": 0.00031012867040580667, + "loss": 0.4498, + "step": 1550 + }, + { + "epoch": 6.3, + "grad_norm": 1.9849016666412354, + "learning_rate": 0.0003084790498185417, + "loss": 0.4262, + "step": 1575 + }, + { + "epoch": 6.4, + "grad_norm": 1.6405338048934937, + "learning_rate": 0.0003068294292312768, + "loss": 0.4225, + "step": 1600 + }, + { + "epoch": 6.5, + "grad_norm": 1.8233951330184937, + "learning_rate": 0.0003051798086440119, + "loss": 0.4248, + "step": 1625 + }, + { + "epoch": 6.6, + "grad_norm": 1.396469235420227, + "learning_rate": 0.000303530188056747, + "loss": 0.412, + "step": 1650 + }, + { + "epoch": 6.7, + "grad_norm": 1.5438531637191772, + "learning_rate": 0.000301880567469482, + "loss": 0.4105, + "step": 1675 + }, + { + "epoch": 6.8, + "grad_norm": 2.49668288230896, + "learning_rate": 0.0003002309468822171, + "loss": 0.4318, + "step": 1700 + }, + { + "epoch": 6.9, + "grad_norm": 1.5911107063293457, + "learning_rate": 0.0002985813262949522, + "loss": 0.4454, + "step": 1725 + }, + { + "epoch": 7.0, + "grad_norm": 2.9934186935424805, + "learning_rate": 0.00029693170570768723, + "loss": 0.4306, + "step": 1750 + }, + { + "epoch": 7.1, + "grad_norm": 2.5476737022399902, + "learning_rate": 0.00029528208512042233, + "loss": 0.4066, + "step": 1775 + }, + { + "epoch": 7.2, + "grad_norm": 4.207254886627197, + "learning_rate": 0.0002936324645331574, + "loss": 0.4521, + "step": 1800 + }, + { + "epoch": 7.3, + "grad_norm": 1.8378102779388428, + "learning_rate": 0.00029198284394589243, + "loss": 0.4197, + "step": 1825 + }, + { + "epoch": 7.4, + "grad_norm": 2.2129015922546387, + "learning_rate": 0.00029033322335862753, + "loss": 0.4182, + "step": 1850 + }, + { + "epoch": 7.5, + "grad_norm": 1.4666684865951538, + "learning_rate": 0.0002886836027713626, + "loss": 0.4224, + "step": 1875 + }, + { + "epoch": 7.6, + "grad_norm": 2.0311567783355713, + "learning_rate": 0.0002870339821840977, + "loss": 0.4304, + "step": 1900 + }, + { + "epoch": 7.7, + "grad_norm": 2.346604585647583, + "learning_rate": 0.00028538436159683273, + "loss": 0.4012, + "step": 1925 + }, + { + "epoch": 7.8, + "grad_norm": 1.440136432647705, + "learning_rate": 0.00028373474100956784, + "loss": 0.4487, + "step": 1950 + }, + { + "epoch": 7.9, + "grad_norm": 2.426828384399414, + "learning_rate": 0.0002820851204223029, + "loss": 0.3577, + "step": 1975 + }, + { + "epoch": 8.0, + "grad_norm": 3.095980405807495, + "learning_rate": 0.00028043549983503794, + "loss": 0.4121, + "step": 2000 + }, + { + "epoch": 8.1, + "grad_norm": 1.8646572828292847, + "learning_rate": 0.00027878587924777304, + "loss": 0.4213, + "step": 2025 + }, + { + "epoch": 8.2, + "grad_norm": 2.3382062911987305, + "learning_rate": 0.0002771362586605081, + "loss": 0.4119, + "step": 2050 + }, + { + "epoch": 8.3, + "grad_norm": 1.7999392747879028, + "learning_rate": 0.0002754866380732432, + "loss": 0.4165, + "step": 2075 + }, + { + "epoch": 8.4, + "grad_norm": 2.262005567550659, + "learning_rate": 0.00027383701748597824, + "loss": 0.4043, + "step": 2100 + }, + { + "epoch": 8.5, + "grad_norm": 1.8611185550689697, + "learning_rate": 0.0002721873968987133, + "loss": 0.4022, + "step": 2125 + }, + { + "epoch": 8.6, + "grad_norm": 1.3322739601135254, + "learning_rate": 0.00027053777631144834, + "loss": 0.3835, + "step": 2150 + }, + { + "epoch": 8.7, + "grad_norm": 1.335750937461853, + "learning_rate": 0.00026888815572418344, + "loss": 0.424, + "step": 2175 + }, + { + "epoch": 8.8, + "grad_norm": 1.4633246660232544, + "learning_rate": 0.00026723853513691855, + "loss": 0.4124, + "step": 2200 + }, + { + "epoch": 8.9, + "grad_norm": 2.061042308807373, + "learning_rate": 0.0002655889145496536, + "loss": 0.4016, + "step": 2225 + }, + { + "epoch": 9.0, + "grad_norm": 1.9621604681015015, + "learning_rate": 0.0002639392939623887, + "loss": 0.4256, + "step": 2250 + }, + { + "epoch": 9.1, + "grad_norm": 1.4014378786087036, + "learning_rate": 0.0002622896733751237, + "loss": 0.3806, + "step": 2275 + }, + { + "epoch": 9.2, + "grad_norm": 2.6264309883117676, + "learning_rate": 0.0002606400527878588, + "loss": 0.3636, + "step": 2300 + }, + { + "epoch": 9.3, + "grad_norm": 1.8708316087722778, + "learning_rate": 0.00025899043220059385, + "loss": 0.4101, + "step": 2325 + }, + { + "epoch": 9.4, + "grad_norm": 2.156869649887085, + "learning_rate": 0.00025734081161332895, + "loss": 0.4114, + "step": 2350 + }, + { + "epoch": 9.5, + "grad_norm": 2.483182907104492, + "learning_rate": 0.000255691191026064, + "loss": 0.421, + "step": 2375 + }, + { + "epoch": 9.6, + "grad_norm": 2.1955626010894775, + "learning_rate": 0.0002540415704387991, + "loss": 0.3918, + "step": 2400 + }, + { + "epoch": 9.7, + "grad_norm": 4.115762233734131, + "learning_rate": 0.00025239194985153415, + "loss": 0.4172, + "step": 2425 + }, + { + "epoch": 9.8, + "grad_norm": 2.144148349761963, + "learning_rate": 0.0002507423292642692, + "loss": 0.3636, + "step": 2450 + }, + { + "epoch": 9.9, + "grad_norm": 2.340376377105713, + "learning_rate": 0.0002490927086770043, + "loss": 0.4059, + "step": 2475 + }, + { + "epoch": 10.0, + "grad_norm": 1.5438555479049683, + "learning_rate": 0.00024744308808973936, + "loss": 0.4232, + "step": 2500 + }, + { + "epoch": 10.1, + "grad_norm": 2.485889196395874, + "learning_rate": 0.00024579346750247446, + "loss": 0.3861, + "step": 2525 + }, + { + "epoch": 10.2, + "grad_norm": 2.4561119079589844, + "learning_rate": 0.0002441438469152095, + "loss": 0.3565, + "step": 2550 + }, + { + "epoch": 10.3, + "grad_norm": 2.460170269012451, + "learning_rate": 0.00024249422632794456, + "loss": 0.4003, + "step": 2575 + }, + { + "epoch": 10.4, + "grad_norm": 2.774445056915283, + "learning_rate": 0.00024084460574067963, + "loss": 0.4088, + "step": 2600 + }, + { + "epoch": 10.5, + "grad_norm": 2.678745985031128, + "learning_rate": 0.0002391949851534147, + "loss": 0.4021, + "step": 2625 + }, + { + "epoch": 10.6, + "grad_norm": 1.4933565855026245, + "learning_rate": 0.0002375453645661498, + "loss": 0.3744, + "step": 2650 + }, + { + "epoch": 10.7, + "grad_norm": 1.367900013923645, + "learning_rate": 0.00023589574397888486, + "loss": 0.3925, + "step": 2675 + }, + { + "epoch": 10.8, + "grad_norm": 1.653441309928894, + "learning_rate": 0.00023424612339161997, + "loss": 0.4049, + "step": 2700 + }, + { + "epoch": 10.9, + "grad_norm": 1.5892399549484253, + "learning_rate": 0.000232596502804355, + "loss": 0.3927, + "step": 2725 + }, + { + "epoch": 11.0, + "grad_norm": 1.535090684890747, + "learning_rate": 0.00023094688221709007, + "loss": 0.3739, + "step": 2750 + }, + { + "epoch": 11.1, + "grad_norm": 2.247271776199341, + "learning_rate": 0.00022929726162982514, + "loss": 0.3362, + "step": 2775 + }, + { + "epoch": 11.2, + "grad_norm": 2.408069372177124, + "learning_rate": 0.00022764764104256022, + "loss": 0.3933, + "step": 2800 + }, + { + "epoch": 11.3, + "grad_norm": 2.219688653945923, + "learning_rate": 0.0002259980204552953, + "loss": 0.4062, + "step": 2825 + }, + { + "epoch": 11.4, + "grad_norm": 3.0188546180725098, + "learning_rate": 0.00022434839986803037, + "loss": 0.3926, + "step": 2850 + }, + { + "epoch": 11.5, + "grad_norm": 2.8080244064331055, + "learning_rate": 0.00022269877928076542, + "loss": 0.3942, + "step": 2875 + }, + { + "epoch": 11.6, + "grad_norm": 2.263610601425171, + "learning_rate": 0.0002210491586935005, + "loss": 0.3887, + "step": 2900 + }, + { + "epoch": 11.7, + "grad_norm": 2.1239607334136963, + "learning_rate": 0.00021939953810623557, + "loss": 0.3989, + "step": 2925 + }, + { + "epoch": 11.8, + "grad_norm": 3.435351848602295, + "learning_rate": 0.00021774991751897065, + "loss": 0.4185, + "step": 2950 + }, + { + "epoch": 11.9, + "grad_norm": 4.1570515632629395, + "learning_rate": 0.00021610029693170573, + "loss": 0.3597, + "step": 2975 + }, + { + "epoch": 12.0, + "grad_norm": 1.7937941551208496, + "learning_rate": 0.0002144506763444408, + "loss": 0.3719, + "step": 3000 + }, + { + "epoch": 12.1, + "grad_norm": 2.3080639839172363, + "learning_rate": 0.00021280105575717582, + "loss": 0.3856, + "step": 3025 + }, + { + "epoch": 12.2, + "grad_norm": 1.9331200122833252, + "learning_rate": 0.00021115143516991093, + "loss": 0.3798, + "step": 3050 + }, + { + "epoch": 12.3, + "grad_norm": 2.0342001914978027, + "learning_rate": 0.000209501814582646, + "loss": 0.3789, + "step": 3075 + }, + { + "epoch": 12.4, + "grad_norm": 1.7878581285476685, + "learning_rate": 0.00020785219399538108, + "loss": 0.3546, + "step": 3100 + }, + { + "epoch": 12.5, + "grad_norm": 3.657317638397217, + "learning_rate": 0.00020620257340811616, + "loss": 0.3582, + "step": 3125 + }, + { + "epoch": 12.6, + "grad_norm": 2.246429681777954, + "learning_rate": 0.00020455295282085123, + "loss": 0.3991, + "step": 3150 + }, + { + "epoch": 12.7, + "grad_norm": 1.7742749452590942, + "learning_rate": 0.00020290333223358626, + "loss": 0.3914, + "step": 3175 + }, + { + "epoch": 12.8, + "grad_norm": 2.3869900703430176, + "learning_rate": 0.00020125371164632133, + "loss": 0.3435, + "step": 3200 + }, + { + "epoch": 12.9, + "grad_norm": 2.444920539855957, + "learning_rate": 0.0001996040910590564, + "loss": 0.393, + "step": 3225 + }, + { + "epoch": 13.0, + "grad_norm": 2.26476788520813, + "learning_rate": 0.00019795447047179148, + "loss": 0.3828, + "step": 3250 + }, + { + "epoch": 13.1, + "grad_norm": 2.4492876529693604, + "learning_rate": 0.0001963048498845266, + "loss": 0.3856, + "step": 3275 + }, + { + "epoch": 13.2, + "grad_norm": 4.578381061553955, + "learning_rate": 0.00019465522929726164, + "loss": 0.3676, + "step": 3300 + }, + { + "epoch": 13.3, + "grad_norm": 2.0352611541748047, + "learning_rate": 0.00019300560870999671, + "loss": 0.369, + "step": 3325 + }, + { + "epoch": 13.4, + "grad_norm": 2.151038646697998, + "learning_rate": 0.0001913559881227318, + "loss": 0.3713, + "step": 3350 + }, + { + "epoch": 13.5, + "grad_norm": 2.4661877155303955, + "learning_rate": 0.00018970636753546684, + "loss": 0.3859, + "step": 3375 + }, + { + "epoch": 13.6, + "grad_norm": 1.7616342306137085, + "learning_rate": 0.00018805674694820192, + "loss": 0.363, + "step": 3400 + }, + { + "epoch": 13.7, + "grad_norm": 2.26426362991333, + "learning_rate": 0.000186407126360937, + "loss": 0.3696, + "step": 3425 + }, + { + "epoch": 13.8, + "grad_norm": 2.355781078338623, + "learning_rate": 0.00018475750577367207, + "loss": 0.372, + "step": 3450 + }, + { + "epoch": 13.9, + "grad_norm": 1.961585283279419, + "learning_rate": 0.00018310788518640715, + "loss": 0.3418, + "step": 3475 + }, + { + "epoch": 14.0, + "grad_norm": 2.54199481010437, + "learning_rate": 0.00018145826459914222, + "loss": 0.3648, + "step": 3500 + }, + { + "epoch": 14.1, + "grad_norm": 1.5017932653427124, + "learning_rate": 0.00017980864401187727, + "loss": 0.3914, + "step": 3525 + }, + { + "epoch": 14.2, + "grad_norm": 3.051287889480591, + "learning_rate": 0.00017815902342461235, + "loss": 0.3887, + "step": 3550 + }, + { + "epoch": 14.3, + "grad_norm": 2.697949171066284, + "learning_rate": 0.00017650940283734742, + "loss": 0.3524, + "step": 3575 + }, + { + "epoch": 14.4, + "grad_norm": 2.470728635787964, + "learning_rate": 0.00017485978225008247, + "loss": 0.3767, + "step": 3600 + }, + { + "epoch": 14.5, + "grad_norm": 2.5166375637054443, + "learning_rate": 0.00017321016166281755, + "loss": 0.362, + "step": 3625 + }, + { + "epoch": 14.6, + "grad_norm": 1.7630703449249268, + "learning_rate": 0.00017156054107555263, + "loss": 0.3936, + "step": 3650 + }, + { + "epoch": 14.7, + "grad_norm": 2.083225727081299, + "learning_rate": 0.0001699109204882877, + "loss": 0.36, + "step": 3675 + }, + { + "epoch": 14.8, + "grad_norm": 1.7796965837478638, + "learning_rate": 0.00016826129990102278, + "loss": 0.3646, + "step": 3700 + }, + { + "epoch": 14.9, + "grad_norm": 2.127898693084717, + "learning_rate": 0.00016661167931375785, + "loss": 0.3722, + "step": 3725 + }, + { + "epoch": 15.0, + "grad_norm": 2.666217088699341, + "learning_rate": 0.0001649620587264929, + "loss": 0.3329, + "step": 3750 + }, + { + "epoch": 15.1, + "grad_norm": 2.521188497543335, + "learning_rate": 0.00016331243813922798, + "loss": 0.331, + "step": 3775 + }, + { + "epoch": 15.2, + "grad_norm": 1.528865098953247, + "learning_rate": 0.00016166281755196306, + "loss": 0.3513, + "step": 3800 + }, + { + "epoch": 15.3, + "grad_norm": 3.452749490737915, + "learning_rate": 0.00016001319696469813, + "loss": 0.3777, + "step": 3825 + }, + { + "epoch": 15.4, + "grad_norm": 3.0441641807556152, + "learning_rate": 0.0001583635763774332, + "loss": 0.3536, + "step": 3850 + }, + { + "epoch": 15.5, + "grad_norm": 1.9194542169570923, + "learning_rate": 0.00015671395579016829, + "loss": 0.3635, + "step": 3875 + }, + { + "epoch": 15.6, + "grad_norm": 2.011859655380249, + "learning_rate": 0.00015506433520290334, + "loss": 0.3467, + "step": 3900 + }, + { + "epoch": 15.7, + "grad_norm": 1.5169134140014648, + "learning_rate": 0.0001534147146156384, + "loss": 0.3639, + "step": 3925 + }, + { + "epoch": 15.8, + "grad_norm": 2.715104103088379, + "learning_rate": 0.0001517650940283735, + "loss": 0.3473, + "step": 3950 + }, + { + "epoch": 15.9, + "grad_norm": 1.8220083713531494, + "learning_rate": 0.00015011547344110854, + "loss": 0.3902, + "step": 3975 + }, + { + "epoch": 16.0, + "grad_norm": 3.227846384048462, + "learning_rate": 0.00014846585285384361, + "loss": 0.3576, + "step": 4000 + } + ], + "logging_steps": 25, + "max_steps": 6250, + "num_input_tokens_seen": 0, + "num_train_epochs": 25, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 221524131840000.0, + "train_batch_size": 20, + "trial_name": null, + "trial_params": null +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/training_args.bin b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..10d16bb8e9df9c51aba38e7a6f04deeecf1ea29f --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:662a5f25d38ccae34f19d85e24456de6d3eaaabafca26fb7cc8816de0c990f20 +size 5304 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/vocab.txt b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/vocab.txt new file mode 100644 index 0000000000000000000000000000000000000000..cc9ad53732e21ea9180f18e1fb67a38cc5702d4c --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/vocab.txt @@ -0,0 +1,4101 @@ +[PAD] +[UNK] +[CLS] +[SEP] +[MASK] +AAAAAA +AAAAAT +AAAAAC +AAAAAG +AAAATA +AAAATT +AAAATC +AAAATG +AAAACA +AAAACT +AAAACC +AAAACG +AAAAGA +AAAAGT +AAAAGC +AAAAGG +AAATAA +AAATAT +AAATAC +AAATAG +AAATTA +AAATTT +AAATTC +AAATTG +AAATCA +AAATCT +AAATCC +AAATCG +AAATGA +AAATGT +AAATGC +AAATGG +AAACAA +AAACAT +AAACAC +AAACAG +AAACTA +AAACTT +AAACTC +AAACTG +AAACCA +AAACCT +AAACCC +AAACCG +AAACGA +AAACGT +AAACGC +AAACGG +AAAGAA +AAAGAT +AAAGAC +AAAGAG +AAAGTA +AAAGTT +AAAGTC +AAAGTG +AAAGCA +AAAGCT +AAAGCC +AAAGCG +AAAGGA +AAAGGT +AAAGGC +AAAGGG +AATAAA +AATAAT +AATAAC +AATAAG +AATATA +AATATT +AATATC +AATATG +AATACA +AATACT +AATACC +AATACG +AATAGA +AATAGT +AATAGC +AATAGG +AATTAA +AATTAT +AATTAC +AATTAG +AATTTA +AATTTT +AATTTC +AATTTG +AATTCA +AATTCT +AATTCC +AATTCG +AATTGA +AATTGT +AATTGC +AATTGG +AATCAA +AATCAT +AATCAC +AATCAG +AATCTA +AATCTT +AATCTC +AATCTG +AATCCA +AATCCT +AATCCC +AATCCG +AATCGA +AATCGT +AATCGC +AATCGG +AATGAA +AATGAT +AATGAC +AATGAG +AATGTA +AATGTT +AATGTC +AATGTG +AATGCA +AATGCT +AATGCC +AATGCG +AATGGA +AATGGT +AATGGC +AATGGG +AACAAA +AACAAT +AACAAC +AACAAG +AACATA +AACATT +AACATC +AACATG +AACACA +AACACT +AACACC +AACACG +AACAGA +AACAGT +AACAGC +AACAGG +AACTAA +AACTAT +AACTAC +AACTAG +AACTTA +AACTTT +AACTTC +AACTTG +AACTCA +AACTCT +AACTCC +AACTCG +AACTGA +AACTGT +AACTGC +AACTGG +AACCAA +AACCAT +AACCAC +AACCAG +AACCTA +AACCTT +AACCTC +AACCTG +AACCCA +AACCCT +AACCCC +AACCCG +AACCGA +AACCGT +AACCGC +AACCGG +AACGAA +AACGAT +AACGAC +AACGAG +AACGTA +AACGTT +AACGTC +AACGTG +AACGCA +AACGCT +AACGCC +AACGCG +AACGGA +AACGGT +AACGGC +AACGGG +AAGAAA +AAGAAT +AAGAAC +AAGAAG +AAGATA +AAGATT +AAGATC +AAGATG +AAGACA +AAGACT +AAGACC +AAGACG +AAGAGA +AAGAGT +AAGAGC +AAGAGG +AAGTAA +AAGTAT +AAGTAC +AAGTAG +AAGTTA +AAGTTT +AAGTTC +AAGTTG +AAGTCA +AAGTCT +AAGTCC +AAGTCG +AAGTGA +AAGTGT +AAGTGC +AAGTGG +AAGCAA +AAGCAT +AAGCAC +AAGCAG +AAGCTA +AAGCTT +AAGCTC +AAGCTG +AAGCCA +AAGCCT +AAGCCC +AAGCCG +AAGCGA +AAGCGT +AAGCGC +AAGCGG +AAGGAA +AAGGAT +AAGGAC +AAGGAG +AAGGTA +AAGGTT +AAGGTC +AAGGTG +AAGGCA +AAGGCT +AAGGCC +AAGGCG +AAGGGA +AAGGGT +AAGGGC +AAGGGG +ATAAAA +ATAAAT +ATAAAC +ATAAAG +ATAATA +ATAATT +ATAATC +ATAATG +ATAACA +ATAACT +ATAACC +ATAACG +ATAAGA +ATAAGT +ATAAGC +ATAAGG +ATATAA +ATATAT +ATATAC +ATATAG +ATATTA +ATATTT +ATATTC +ATATTG +ATATCA +ATATCT +ATATCC +ATATCG +ATATGA +ATATGT +ATATGC +ATATGG +ATACAA +ATACAT +ATACAC +ATACAG +ATACTA +ATACTT +ATACTC +ATACTG +ATACCA +ATACCT +ATACCC +ATACCG +ATACGA +ATACGT +ATACGC +ATACGG +ATAGAA +ATAGAT +ATAGAC +ATAGAG +ATAGTA +ATAGTT +ATAGTC +ATAGTG +ATAGCA +ATAGCT +ATAGCC +ATAGCG +ATAGGA +ATAGGT +ATAGGC +ATAGGG +ATTAAA +ATTAAT +ATTAAC +ATTAAG +ATTATA +ATTATT +ATTATC +ATTATG +ATTACA +ATTACT +ATTACC +ATTACG +ATTAGA +ATTAGT +ATTAGC +ATTAGG +ATTTAA +ATTTAT +ATTTAC +ATTTAG +ATTTTA +ATTTTT +ATTTTC +ATTTTG +ATTTCA +ATTTCT +ATTTCC +ATTTCG +ATTTGA +ATTTGT +ATTTGC +ATTTGG +ATTCAA +ATTCAT +ATTCAC +ATTCAG +ATTCTA +ATTCTT +ATTCTC +ATTCTG +ATTCCA +ATTCCT +ATTCCC +ATTCCG +ATTCGA +ATTCGT +ATTCGC +ATTCGG +ATTGAA +ATTGAT +ATTGAC +ATTGAG +ATTGTA +ATTGTT +ATTGTC +ATTGTG +ATTGCA +ATTGCT +ATTGCC +ATTGCG +ATTGGA +ATTGGT +ATTGGC +ATTGGG +ATCAAA +ATCAAT +ATCAAC +ATCAAG +ATCATA +ATCATT +ATCATC +ATCATG +ATCACA +ATCACT +ATCACC +ATCACG +ATCAGA +ATCAGT +ATCAGC +ATCAGG +ATCTAA +ATCTAT +ATCTAC +ATCTAG +ATCTTA +ATCTTT +ATCTTC +ATCTTG +ATCTCA +ATCTCT +ATCTCC +ATCTCG +ATCTGA +ATCTGT +ATCTGC +ATCTGG +ATCCAA +ATCCAT +ATCCAC +ATCCAG +ATCCTA +ATCCTT +ATCCTC +ATCCTG +ATCCCA +ATCCCT +ATCCCC +ATCCCG +ATCCGA +ATCCGT +ATCCGC +ATCCGG +ATCGAA +ATCGAT +ATCGAC +ATCGAG +ATCGTA +ATCGTT +ATCGTC +ATCGTG +ATCGCA +ATCGCT +ATCGCC +ATCGCG +ATCGGA +ATCGGT +ATCGGC +ATCGGG +ATGAAA +ATGAAT +ATGAAC +ATGAAG +ATGATA +ATGATT +ATGATC +ATGATG +ATGACA +ATGACT +ATGACC +ATGACG +ATGAGA +ATGAGT +ATGAGC +ATGAGG +ATGTAA +ATGTAT +ATGTAC +ATGTAG +ATGTTA +ATGTTT +ATGTTC +ATGTTG +ATGTCA +ATGTCT +ATGTCC +ATGTCG +ATGTGA +ATGTGT +ATGTGC +ATGTGG +ATGCAA +ATGCAT +ATGCAC +ATGCAG +ATGCTA +ATGCTT +ATGCTC +ATGCTG +ATGCCA +ATGCCT +ATGCCC +ATGCCG +ATGCGA +ATGCGT +ATGCGC +ATGCGG +ATGGAA +ATGGAT +ATGGAC +ATGGAG +ATGGTA +ATGGTT +ATGGTC +ATGGTG +ATGGCA +ATGGCT +ATGGCC +ATGGCG +ATGGGA +ATGGGT +ATGGGC +ATGGGG +ACAAAA +ACAAAT +ACAAAC +ACAAAG +ACAATA +ACAATT +ACAATC +ACAATG +ACAACA +ACAACT +ACAACC +ACAACG +ACAAGA +ACAAGT +ACAAGC +ACAAGG +ACATAA +ACATAT +ACATAC +ACATAG +ACATTA +ACATTT +ACATTC +ACATTG +ACATCA +ACATCT +ACATCC +ACATCG +ACATGA +ACATGT +ACATGC +ACATGG +ACACAA +ACACAT +ACACAC +ACACAG +ACACTA +ACACTT +ACACTC +ACACTG +ACACCA +ACACCT +ACACCC +ACACCG +ACACGA +ACACGT +ACACGC +ACACGG +ACAGAA +ACAGAT +ACAGAC +ACAGAG +ACAGTA +ACAGTT +ACAGTC +ACAGTG +ACAGCA +ACAGCT +ACAGCC +ACAGCG +ACAGGA +ACAGGT +ACAGGC +ACAGGG +ACTAAA +ACTAAT +ACTAAC +ACTAAG +ACTATA +ACTATT +ACTATC +ACTATG +ACTACA +ACTACT +ACTACC +ACTACG +ACTAGA +ACTAGT +ACTAGC +ACTAGG +ACTTAA +ACTTAT +ACTTAC +ACTTAG +ACTTTA +ACTTTT +ACTTTC +ACTTTG +ACTTCA +ACTTCT +ACTTCC +ACTTCG +ACTTGA +ACTTGT +ACTTGC +ACTTGG +ACTCAA +ACTCAT +ACTCAC +ACTCAG +ACTCTA +ACTCTT +ACTCTC +ACTCTG +ACTCCA +ACTCCT +ACTCCC +ACTCCG +ACTCGA +ACTCGT +ACTCGC +ACTCGG +ACTGAA +ACTGAT +ACTGAC +ACTGAG +ACTGTA +ACTGTT +ACTGTC +ACTGTG +ACTGCA +ACTGCT +ACTGCC +ACTGCG +ACTGGA +ACTGGT +ACTGGC +ACTGGG +ACCAAA +ACCAAT +ACCAAC +ACCAAG +ACCATA +ACCATT +ACCATC +ACCATG +ACCACA +ACCACT +ACCACC +ACCACG +ACCAGA +ACCAGT +ACCAGC +ACCAGG +ACCTAA +ACCTAT +ACCTAC +ACCTAG +ACCTTA +ACCTTT +ACCTTC +ACCTTG +ACCTCA +ACCTCT +ACCTCC +ACCTCG +ACCTGA +ACCTGT +ACCTGC +ACCTGG +ACCCAA +ACCCAT +ACCCAC +ACCCAG +ACCCTA +ACCCTT +ACCCTC +ACCCTG +ACCCCA +ACCCCT +ACCCCC +ACCCCG +ACCCGA +ACCCGT +ACCCGC +ACCCGG +ACCGAA +ACCGAT +ACCGAC +ACCGAG +ACCGTA +ACCGTT +ACCGTC +ACCGTG +ACCGCA +ACCGCT +ACCGCC +ACCGCG +ACCGGA +ACCGGT +ACCGGC +ACCGGG +ACGAAA +ACGAAT +ACGAAC +ACGAAG +ACGATA +ACGATT +ACGATC +ACGATG +ACGACA +ACGACT +ACGACC +ACGACG +ACGAGA +ACGAGT +ACGAGC +ACGAGG +ACGTAA +ACGTAT +ACGTAC +ACGTAG +ACGTTA +ACGTTT +ACGTTC +ACGTTG +ACGTCA +ACGTCT +ACGTCC +ACGTCG +ACGTGA +ACGTGT +ACGTGC +ACGTGG +ACGCAA +ACGCAT +ACGCAC +ACGCAG +ACGCTA +ACGCTT +ACGCTC +ACGCTG +ACGCCA +ACGCCT +ACGCCC +ACGCCG +ACGCGA +ACGCGT +ACGCGC +ACGCGG +ACGGAA +ACGGAT +ACGGAC +ACGGAG +ACGGTA +ACGGTT +ACGGTC +ACGGTG +ACGGCA +ACGGCT +ACGGCC +ACGGCG +ACGGGA +ACGGGT +ACGGGC +ACGGGG +AGAAAA +AGAAAT +AGAAAC +AGAAAG +AGAATA +AGAATT +AGAATC +AGAATG +AGAACA +AGAACT +AGAACC +AGAACG +AGAAGA +AGAAGT +AGAAGC +AGAAGG +AGATAA +AGATAT +AGATAC +AGATAG +AGATTA +AGATTT +AGATTC +AGATTG +AGATCA +AGATCT +AGATCC +AGATCG +AGATGA +AGATGT +AGATGC +AGATGG +AGACAA +AGACAT +AGACAC +AGACAG +AGACTA +AGACTT +AGACTC +AGACTG +AGACCA +AGACCT +AGACCC +AGACCG +AGACGA +AGACGT +AGACGC +AGACGG +AGAGAA +AGAGAT +AGAGAC +AGAGAG +AGAGTA +AGAGTT +AGAGTC +AGAGTG +AGAGCA +AGAGCT +AGAGCC +AGAGCG +AGAGGA +AGAGGT +AGAGGC +AGAGGG +AGTAAA +AGTAAT +AGTAAC +AGTAAG +AGTATA +AGTATT +AGTATC +AGTATG +AGTACA +AGTACT +AGTACC +AGTACG +AGTAGA +AGTAGT +AGTAGC +AGTAGG +AGTTAA +AGTTAT +AGTTAC +AGTTAG +AGTTTA +AGTTTT +AGTTTC +AGTTTG +AGTTCA +AGTTCT +AGTTCC +AGTTCG +AGTTGA +AGTTGT +AGTTGC +AGTTGG +AGTCAA +AGTCAT +AGTCAC +AGTCAG +AGTCTA +AGTCTT +AGTCTC +AGTCTG +AGTCCA +AGTCCT +AGTCCC +AGTCCG +AGTCGA +AGTCGT +AGTCGC +AGTCGG +AGTGAA +AGTGAT +AGTGAC +AGTGAG +AGTGTA +AGTGTT +AGTGTC +AGTGTG +AGTGCA +AGTGCT +AGTGCC +AGTGCG +AGTGGA +AGTGGT +AGTGGC +AGTGGG +AGCAAA +AGCAAT +AGCAAC +AGCAAG +AGCATA +AGCATT +AGCATC +AGCATG +AGCACA +AGCACT +AGCACC +AGCACG +AGCAGA +AGCAGT +AGCAGC +AGCAGG +AGCTAA +AGCTAT +AGCTAC +AGCTAG +AGCTTA +AGCTTT +AGCTTC +AGCTTG +AGCTCA +AGCTCT +AGCTCC +AGCTCG +AGCTGA +AGCTGT +AGCTGC +AGCTGG +AGCCAA +AGCCAT +AGCCAC +AGCCAG +AGCCTA +AGCCTT +AGCCTC +AGCCTG +AGCCCA +AGCCCT +AGCCCC +AGCCCG +AGCCGA +AGCCGT +AGCCGC +AGCCGG +AGCGAA +AGCGAT +AGCGAC +AGCGAG +AGCGTA +AGCGTT +AGCGTC +AGCGTG +AGCGCA +AGCGCT +AGCGCC +AGCGCG +AGCGGA +AGCGGT +AGCGGC +AGCGGG +AGGAAA +AGGAAT +AGGAAC +AGGAAG +AGGATA +AGGATT +AGGATC +AGGATG +AGGACA +AGGACT +AGGACC +AGGACG +AGGAGA +AGGAGT +AGGAGC +AGGAGG +AGGTAA +AGGTAT +AGGTAC +AGGTAG +AGGTTA +AGGTTT +AGGTTC +AGGTTG +AGGTCA +AGGTCT +AGGTCC +AGGTCG +AGGTGA +AGGTGT +AGGTGC +AGGTGG +AGGCAA +AGGCAT +AGGCAC +AGGCAG +AGGCTA +AGGCTT +AGGCTC +AGGCTG +AGGCCA +AGGCCT +AGGCCC +AGGCCG +AGGCGA +AGGCGT +AGGCGC +AGGCGG +AGGGAA +AGGGAT +AGGGAC +AGGGAG +AGGGTA +AGGGTT +AGGGTC +AGGGTG +AGGGCA +AGGGCT +AGGGCC +AGGGCG +AGGGGA +AGGGGT +AGGGGC +AGGGGG +TAAAAA +TAAAAT +TAAAAC +TAAAAG +TAAATA +TAAATT +TAAATC +TAAATG +TAAACA +TAAACT +TAAACC +TAAACG +TAAAGA +TAAAGT +TAAAGC +TAAAGG +TAATAA +TAATAT +TAATAC +TAATAG +TAATTA +TAATTT +TAATTC +TAATTG +TAATCA +TAATCT +TAATCC +TAATCG +TAATGA +TAATGT +TAATGC +TAATGG +TAACAA +TAACAT +TAACAC +TAACAG +TAACTA +TAACTT +TAACTC +TAACTG +TAACCA +TAACCT +TAACCC +TAACCG +TAACGA +TAACGT +TAACGC +TAACGG +TAAGAA +TAAGAT +TAAGAC +TAAGAG +TAAGTA +TAAGTT +TAAGTC +TAAGTG +TAAGCA +TAAGCT +TAAGCC +TAAGCG +TAAGGA +TAAGGT +TAAGGC +TAAGGG +TATAAA +TATAAT +TATAAC +TATAAG +TATATA +TATATT +TATATC +TATATG +TATACA +TATACT +TATACC +TATACG +TATAGA +TATAGT +TATAGC +TATAGG +TATTAA +TATTAT +TATTAC +TATTAG +TATTTA +TATTTT +TATTTC +TATTTG +TATTCA +TATTCT +TATTCC +TATTCG +TATTGA +TATTGT +TATTGC +TATTGG +TATCAA +TATCAT +TATCAC +TATCAG +TATCTA +TATCTT +TATCTC +TATCTG +TATCCA +TATCCT +TATCCC +TATCCG +TATCGA +TATCGT +TATCGC +TATCGG +TATGAA +TATGAT +TATGAC +TATGAG +TATGTA +TATGTT +TATGTC +TATGTG +TATGCA +TATGCT +TATGCC +TATGCG +TATGGA +TATGGT +TATGGC +TATGGG +TACAAA +TACAAT +TACAAC +TACAAG +TACATA +TACATT +TACATC +TACATG +TACACA +TACACT +TACACC +TACACG +TACAGA +TACAGT +TACAGC +TACAGG +TACTAA +TACTAT +TACTAC +TACTAG +TACTTA +TACTTT +TACTTC +TACTTG +TACTCA +TACTCT +TACTCC +TACTCG +TACTGA +TACTGT +TACTGC +TACTGG +TACCAA +TACCAT +TACCAC +TACCAG +TACCTA +TACCTT +TACCTC +TACCTG +TACCCA +TACCCT +TACCCC +TACCCG +TACCGA +TACCGT +TACCGC +TACCGG +TACGAA +TACGAT +TACGAC +TACGAG +TACGTA +TACGTT +TACGTC +TACGTG +TACGCA +TACGCT +TACGCC +TACGCG +TACGGA +TACGGT +TACGGC +TACGGG +TAGAAA +TAGAAT +TAGAAC +TAGAAG +TAGATA +TAGATT +TAGATC +TAGATG +TAGACA +TAGACT +TAGACC +TAGACG +TAGAGA +TAGAGT +TAGAGC +TAGAGG +TAGTAA +TAGTAT +TAGTAC +TAGTAG +TAGTTA +TAGTTT +TAGTTC +TAGTTG +TAGTCA +TAGTCT +TAGTCC +TAGTCG +TAGTGA +TAGTGT +TAGTGC +TAGTGG +TAGCAA +TAGCAT +TAGCAC +TAGCAG +TAGCTA +TAGCTT +TAGCTC +TAGCTG +TAGCCA +TAGCCT +TAGCCC +TAGCCG +TAGCGA +TAGCGT +TAGCGC +TAGCGG +TAGGAA +TAGGAT +TAGGAC +TAGGAG +TAGGTA +TAGGTT +TAGGTC +TAGGTG +TAGGCA +TAGGCT +TAGGCC +TAGGCG +TAGGGA +TAGGGT +TAGGGC +TAGGGG +TTAAAA +TTAAAT +TTAAAC +TTAAAG +TTAATA +TTAATT +TTAATC +TTAATG +TTAACA +TTAACT +TTAACC +TTAACG +TTAAGA +TTAAGT +TTAAGC +TTAAGG +TTATAA +TTATAT +TTATAC +TTATAG +TTATTA +TTATTT +TTATTC +TTATTG +TTATCA +TTATCT +TTATCC +TTATCG +TTATGA +TTATGT +TTATGC +TTATGG +TTACAA +TTACAT +TTACAC +TTACAG +TTACTA +TTACTT +TTACTC +TTACTG +TTACCA +TTACCT +TTACCC +TTACCG +TTACGA +TTACGT +TTACGC +TTACGG +TTAGAA +TTAGAT +TTAGAC +TTAGAG +TTAGTA +TTAGTT +TTAGTC +TTAGTG +TTAGCA +TTAGCT +TTAGCC +TTAGCG +TTAGGA +TTAGGT +TTAGGC +TTAGGG +TTTAAA +TTTAAT +TTTAAC +TTTAAG +TTTATA +TTTATT +TTTATC +TTTATG +TTTACA +TTTACT +TTTACC +TTTACG +TTTAGA +TTTAGT +TTTAGC +TTTAGG +TTTTAA +TTTTAT +TTTTAC +TTTTAG +TTTTTA +TTTTTT +TTTTTC +TTTTTG +TTTTCA +TTTTCT +TTTTCC +TTTTCG +TTTTGA +TTTTGT +TTTTGC +TTTTGG +TTTCAA +TTTCAT +TTTCAC +TTTCAG +TTTCTA +TTTCTT +TTTCTC +TTTCTG +TTTCCA +TTTCCT +TTTCCC +TTTCCG +TTTCGA +TTTCGT +TTTCGC +TTTCGG +TTTGAA +TTTGAT +TTTGAC +TTTGAG +TTTGTA +TTTGTT +TTTGTC +TTTGTG +TTTGCA +TTTGCT +TTTGCC +TTTGCG +TTTGGA +TTTGGT +TTTGGC +TTTGGG +TTCAAA +TTCAAT +TTCAAC +TTCAAG +TTCATA +TTCATT +TTCATC +TTCATG +TTCACA +TTCACT +TTCACC +TTCACG +TTCAGA +TTCAGT +TTCAGC +TTCAGG +TTCTAA +TTCTAT +TTCTAC +TTCTAG +TTCTTA +TTCTTT +TTCTTC +TTCTTG +TTCTCA +TTCTCT +TTCTCC +TTCTCG +TTCTGA +TTCTGT +TTCTGC +TTCTGG +TTCCAA +TTCCAT +TTCCAC +TTCCAG +TTCCTA +TTCCTT +TTCCTC +TTCCTG +TTCCCA +TTCCCT +TTCCCC +TTCCCG +TTCCGA +TTCCGT +TTCCGC +TTCCGG +TTCGAA +TTCGAT +TTCGAC +TTCGAG +TTCGTA +TTCGTT +TTCGTC +TTCGTG +TTCGCA +TTCGCT +TTCGCC +TTCGCG +TTCGGA +TTCGGT +TTCGGC +TTCGGG +TTGAAA +TTGAAT +TTGAAC +TTGAAG +TTGATA +TTGATT +TTGATC +TTGATG +TTGACA +TTGACT +TTGACC +TTGACG +TTGAGA +TTGAGT +TTGAGC +TTGAGG +TTGTAA +TTGTAT +TTGTAC +TTGTAG +TTGTTA +TTGTTT +TTGTTC +TTGTTG +TTGTCA +TTGTCT +TTGTCC +TTGTCG +TTGTGA +TTGTGT +TTGTGC +TTGTGG +TTGCAA +TTGCAT +TTGCAC +TTGCAG +TTGCTA +TTGCTT +TTGCTC +TTGCTG +TTGCCA +TTGCCT +TTGCCC +TTGCCG +TTGCGA +TTGCGT +TTGCGC +TTGCGG +TTGGAA +TTGGAT +TTGGAC +TTGGAG +TTGGTA +TTGGTT +TTGGTC +TTGGTG +TTGGCA +TTGGCT +TTGGCC +TTGGCG +TTGGGA +TTGGGT +TTGGGC +TTGGGG +TCAAAA +TCAAAT +TCAAAC +TCAAAG +TCAATA +TCAATT +TCAATC +TCAATG +TCAACA +TCAACT +TCAACC +TCAACG +TCAAGA +TCAAGT +TCAAGC +TCAAGG +TCATAA +TCATAT +TCATAC +TCATAG +TCATTA +TCATTT +TCATTC +TCATTG +TCATCA +TCATCT +TCATCC +TCATCG +TCATGA +TCATGT +TCATGC +TCATGG +TCACAA +TCACAT +TCACAC +TCACAG +TCACTA +TCACTT +TCACTC +TCACTG +TCACCA +TCACCT +TCACCC +TCACCG +TCACGA +TCACGT +TCACGC +TCACGG +TCAGAA +TCAGAT +TCAGAC +TCAGAG +TCAGTA +TCAGTT +TCAGTC +TCAGTG +TCAGCA +TCAGCT +TCAGCC +TCAGCG +TCAGGA +TCAGGT +TCAGGC +TCAGGG +TCTAAA +TCTAAT +TCTAAC +TCTAAG +TCTATA +TCTATT +TCTATC +TCTATG +TCTACA +TCTACT +TCTACC +TCTACG +TCTAGA +TCTAGT +TCTAGC +TCTAGG +TCTTAA +TCTTAT +TCTTAC +TCTTAG +TCTTTA +TCTTTT +TCTTTC +TCTTTG +TCTTCA +TCTTCT +TCTTCC +TCTTCG +TCTTGA +TCTTGT +TCTTGC +TCTTGG +TCTCAA +TCTCAT +TCTCAC +TCTCAG +TCTCTA +TCTCTT +TCTCTC +TCTCTG +TCTCCA +TCTCCT +TCTCCC +TCTCCG +TCTCGA +TCTCGT +TCTCGC +TCTCGG +TCTGAA +TCTGAT +TCTGAC +TCTGAG +TCTGTA +TCTGTT +TCTGTC +TCTGTG +TCTGCA +TCTGCT +TCTGCC +TCTGCG +TCTGGA +TCTGGT +TCTGGC +TCTGGG +TCCAAA +TCCAAT +TCCAAC +TCCAAG +TCCATA +TCCATT +TCCATC +TCCATG +TCCACA +TCCACT +TCCACC +TCCACG +TCCAGA +TCCAGT +TCCAGC +TCCAGG +TCCTAA +TCCTAT +TCCTAC +TCCTAG +TCCTTA +TCCTTT +TCCTTC +TCCTTG +TCCTCA +TCCTCT +TCCTCC +TCCTCG +TCCTGA +TCCTGT +TCCTGC +TCCTGG +TCCCAA +TCCCAT +TCCCAC +TCCCAG +TCCCTA +TCCCTT +TCCCTC +TCCCTG +TCCCCA +TCCCCT +TCCCCC +TCCCCG +TCCCGA +TCCCGT +TCCCGC +TCCCGG +TCCGAA +TCCGAT +TCCGAC +TCCGAG +TCCGTA +TCCGTT +TCCGTC +TCCGTG +TCCGCA +TCCGCT +TCCGCC +TCCGCG +TCCGGA +TCCGGT +TCCGGC +TCCGGG +TCGAAA +TCGAAT +TCGAAC +TCGAAG +TCGATA +TCGATT +TCGATC +TCGATG +TCGACA +TCGACT +TCGACC +TCGACG +TCGAGA +TCGAGT +TCGAGC +TCGAGG +TCGTAA +TCGTAT +TCGTAC +TCGTAG +TCGTTA +TCGTTT +TCGTTC +TCGTTG +TCGTCA +TCGTCT +TCGTCC +TCGTCG +TCGTGA +TCGTGT +TCGTGC +TCGTGG +TCGCAA +TCGCAT +TCGCAC +TCGCAG +TCGCTA +TCGCTT +TCGCTC +TCGCTG +TCGCCA +TCGCCT +TCGCCC +TCGCCG +TCGCGA +TCGCGT +TCGCGC +TCGCGG +TCGGAA +TCGGAT +TCGGAC +TCGGAG +TCGGTA +TCGGTT +TCGGTC +TCGGTG +TCGGCA +TCGGCT +TCGGCC +TCGGCG +TCGGGA +TCGGGT +TCGGGC +TCGGGG +TGAAAA +TGAAAT +TGAAAC +TGAAAG +TGAATA +TGAATT +TGAATC +TGAATG +TGAACA +TGAACT +TGAACC +TGAACG +TGAAGA +TGAAGT +TGAAGC +TGAAGG +TGATAA +TGATAT +TGATAC +TGATAG +TGATTA +TGATTT +TGATTC +TGATTG +TGATCA +TGATCT +TGATCC +TGATCG +TGATGA +TGATGT +TGATGC +TGATGG +TGACAA +TGACAT +TGACAC +TGACAG +TGACTA +TGACTT +TGACTC +TGACTG +TGACCA +TGACCT +TGACCC +TGACCG +TGACGA +TGACGT +TGACGC +TGACGG +TGAGAA +TGAGAT +TGAGAC +TGAGAG +TGAGTA +TGAGTT +TGAGTC +TGAGTG +TGAGCA +TGAGCT +TGAGCC +TGAGCG +TGAGGA +TGAGGT +TGAGGC +TGAGGG +TGTAAA +TGTAAT +TGTAAC +TGTAAG +TGTATA +TGTATT +TGTATC +TGTATG +TGTACA +TGTACT +TGTACC +TGTACG +TGTAGA +TGTAGT +TGTAGC +TGTAGG +TGTTAA +TGTTAT +TGTTAC +TGTTAG +TGTTTA +TGTTTT +TGTTTC +TGTTTG +TGTTCA +TGTTCT +TGTTCC +TGTTCG +TGTTGA +TGTTGT +TGTTGC +TGTTGG +TGTCAA +TGTCAT +TGTCAC +TGTCAG +TGTCTA +TGTCTT +TGTCTC +TGTCTG +TGTCCA +TGTCCT +TGTCCC +TGTCCG +TGTCGA +TGTCGT +TGTCGC +TGTCGG +TGTGAA +TGTGAT +TGTGAC +TGTGAG +TGTGTA +TGTGTT +TGTGTC +TGTGTG +TGTGCA +TGTGCT +TGTGCC +TGTGCG +TGTGGA +TGTGGT +TGTGGC +TGTGGG +TGCAAA +TGCAAT +TGCAAC +TGCAAG +TGCATA +TGCATT +TGCATC +TGCATG +TGCACA +TGCACT +TGCACC +TGCACG +TGCAGA +TGCAGT +TGCAGC +TGCAGG +TGCTAA +TGCTAT +TGCTAC +TGCTAG +TGCTTA +TGCTTT +TGCTTC +TGCTTG +TGCTCA +TGCTCT +TGCTCC +TGCTCG +TGCTGA +TGCTGT +TGCTGC +TGCTGG +TGCCAA +TGCCAT +TGCCAC +TGCCAG +TGCCTA +TGCCTT +TGCCTC +TGCCTG +TGCCCA +TGCCCT +TGCCCC +TGCCCG +TGCCGA +TGCCGT +TGCCGC +TGCCGG +TGCGAA +TGCGAT +TGCGAC +TGCGAG +TGCGTA +TGCGTT +TGCGTC +TGCGTG +TGCGCA +TGCGCT +TGCGCC +TGCGCG +TGCGGA +TGCGGT +TGCGGC +TGCGGG +TGGAAA +TGGAAT +TGGAAC +TGGAAG +TGGATA +TGGATT +TGGATC +TGGATG +TGGACA +TGGACT +TGGACC +TGGACG +TGGAGA +TGGAGT +TGGAGC +TGGAGG +TGGTAA +TGGTAT +TGGTAC +TGGTAG +TGGTTA +TGGTTT +TGGTTC +TGGTTG +TGGTCA +TGGTCT +TGGTCC +TGGTCG +TGGTGA +TGGTGT +TGGTGC +TGGTGG +TGGCAA +TGGCAT +TGGCAC +TGGCAG +TGGCTA +TGGCTT +TGGCTC +TGGCTG +TGGCCA +TGGCCT +TGGCCC +TGGCCG +TGGCGA +TGGCGT +TGGCGC +TGGCGG +TGGGAA +TGGGAT +TGGGAC +TGGGAG +TGGGTA +TGGGTT +TGGGTC +TGGGTG +TGGGCA +TGGGCT +TGGGCC +TGGGCG +TGGGGA +TGGGGT +TGGGGC +TGGGGG +CAAAAA +CAAAAT +CAAAAC +CAAAAG +CAAATA +CAAATT +CAAATC +CAAATG +CAAACA +CAAACT +CAAACC +CAAACG +CAAAGA +CAAAGT +CAAAGC +CAAAGG +CAATAA +CAATAT +CAATAC +CAATAG +CAATTA +CAATTT +CAATTC +CAATTG +CAATCA +CAATCT +CAATCC +CAATCG +CAATGA +CAATGT +CAATGC +CAATGG +CAACAA +CAACAT +CAACAC +CAACAG +CAACTA +CAACTT +CAACTC +CAACTG +CAACCA +CAACCT +CAACCC +CAACCG +CAACGA +CAACGT +CAACGC +CAACGG +CAAGAA +CAAGAT +CAAGAC +CAAGAG +CAAGTA +CAAGTT +CAAGTC +CAAGTG +CAAGCA +CAAGCT +CAAGCC +CAAGCG +CAAGGA +CAAGGT +CAAGGC +CAAGGG +CATAAA +CATAAT +CATAAC +CATAAG +CATATA +CATATT +CATATC +CATATG +CATACA +CATACT +CATACC +CATACG +CATAGA +CATAGT +CATAGC +CATAGG +CATTAA +CATTAT +CATTAC +CATTAG +CATTTA +CATTTT +CATTTC +CATTTG +CATTCA +CATTCT +CATTCC +CATTCG +CATTGA +CATTGT +CATTGC +CATTGG +CATCAA +CATCAT +CATCAC +CATCAG +CATCTA +CATCTT +CATCTC +CATCTG +CATCCA +CATCCT +CATCCC +CATCCG +CATCGA +CATCGT +CATCGC +CATCGG +CATGAA +CATGAT +CATGAC +CATGAG +CATGTA +CATGTT +CATGTC +CATGTG +CATGCA +CATGCT +CATGCC +CATGCG +CATGGA +CATGGT +CATGGC +CATGGG +CACAAA +CACAAT +CACAAC +CACAAG +CACATA +CACATT +CACATC +CACATG +CACACA +CACACT +CACACC +CACACG +CACAGA +CACAGT +CACAGC +CACAGG +CACTAA +CACTAT +CACTAC +CACTAG +CACTTA +CACTTT +CACTTC +CACTTG +CACTCA +CACTCT +CACTCC +CACTCG +CACTGA +CACTGT +CACTGC +CACTGG +CACCAA +CACCAT +CACCAC +CACCAG +CACCTA +CACCTT +CACCTC +CACCTG +CACCCA +CACCCT +CACCCC +CACCCG +CACCGA +CACCGT +CACCGC +CACCGG +CACGAA +CACGAT +CACGAC +CACGAG +CACGTA +CACGTT +CACGTC +CACGTG +CACGCA +CACGCT +CACGCC +CACGCG +CACGGA +CACGGT +CACGGC +CACGGG +CAGAAA +CAGAAT +CAGAAC +CAGAAG +CAGATA +CAGATT +CAGATC +CAGATG +CAGACA +CAGACT +CAGACC +CAGACG +CAGAGA +CAGAGT +CAGAGC +CAGAGG +CAGTAA +CAGTAT +CAGTAC +CAGTAG +CAGTTA +CAGTTT +CAGTTC +CAGTTG +CAGTCA +CAGTCT +CAGTCC +CAGTCG +CAGTGA +CAGTGT +CAGTGC +CAGTGG +CAGCAA +CAGCAT +CAGCAC +CAGCAG +CAGCTA +CAGCTT +CAGCTC +CAGCTG +CAGCCA +CAGCCT +CAGCCC +CAGCCG +CAGCGA +CAGCGT +CAGCGC +CAGCGG +CAGGAA +CAGGAT +CAGGAC +CAGGAG +CAGGTA +CAGGTT +CAGGTC +CAGGTG +CAGGCA +CAGGCT +CAGGCC +CAGGCG +CAGGGA +CAGGGT +CAGGGC +CAGGGG +CTAAAA +CTAAAT +CTAAAC +CTAAAG +CTAATA +CTAATT +CTAATC +CTAATG +CTAACA +CTAACT +CTAACC +CTAACG +CTAAGA +CTAAGT +CTAAGC +CTAAGG +CTATAA +CTATAT +CTATAC +CTATAG +CTATTA +CTATTT +CTATTC +CTATTG +CTATCA +CTATCT +CTATCC +CTATCG +CTATGA +CTATGT +CTATGC +CTATGG +CTACAA +CTACAT +CTACAC +CTACAG +CTACTA +CTACTT +CTACTC +CTACTG +CTACCA +CTACCT +CTACCC +CTACCG +CTACGA +CTACGT +CTACGC +CTACGG +CTAGAA +CTAGAT +CTAGAC +CTAGAG +CTAGTA +CTAGTT +CTAGTC +CTAGTG +CTAGCA +CTAGCT +CTAGCC +CTAGCG +CTAGGA +CTAGGT +CTAGGC +CTAGGG +CTTAAA +CTTAAT +CTTAAC +CTTAAG +CTTATA +CTTATT +CTTATC +CTTATG +CTTACA +CTTACT +CTTACC +CTTACG +CTTAGA +CTTAGT +CTTAGC +CTTAGG +CTTTAA +CTTTAT +CTTTAC +CTTTAG +CTTTTA +CTTTTT +CTTTTC +CTTTTG +CTTTCA +CTTTCT +CTTTCC +CTTTCG +CTTTGA +CTTTGT +CTTTGC +CTTTGG +CTTCAA +CTTCAT +CTTCAC +CTTCAG +CTTCTA +CTTCTT +CTTCTC +CTTCTG +CTTCCA +CTTCCT +CTTCCC +CTTCCG +CTTCGA +CTTCGT +CTTCGC +CTTCGG +CTTGAA +CTTGAT +CTTGAC +CTTGAG +CTTGTA +CTTGTT +CTTGTC +CTTGTG +CTTGCA +CTTGCT +CTTGCC +CTTGCG +CTTGGA +CTTGGT +CTTGGC +CTTGGG +CTCAAA +CTCAAT +CTCAAC +CTCAAG +CTCATA +CTCATT +CTCATC +CTCATG +CTCACA +CTCACT +CTCACC +CTCACG +CTCAGA +CTCAGT +CTCAGC +CTCAGG +CTCTAA +CTCTAT +CTCTAC +CTCTAG +CTCTTA +CTCTTT +CTCTTC +CTCTTG +CTCTCA +CTCTCT +CTCTCC +CTCTCG +CTCTGA +CTCTGT +CTCTGC +CTCTGG +CTCCAA +CTCCAT +CTCCAC +CTCCAG +CTCCTA +CTCCTT +CTCCTC +CTCCTG +CTCCCA +CTCCCT +CTCCCC +CTCCCG +CTCCGA +CTCCGT +CTCCGC +CTCCGG +CTCGAA +CTCGAT +CTCGAC +CTCGAG +CTCGTA +CTCGTT +CTCGTC +CTCGTG +CTCGCA +CTCGCT +CTCGCC +CTCGCG +CTCGGA +CTCGGT +CTCGGC +CTCGGG +CTGAAA +CTGAAT +CTGAAC +CTGAAG +CTGATA +CTGATT +CTGATC +CTGATG +CTGACA +CTGACT +CTGACC +CTGACG +CTGAGA +CTGAGT +CTGAGC +CTGAGG +CTGTAA +CTGTAT +CTGTAC +CTGTAG +CTGTTA +CTGTTT +CTGTTC +CTGTTG +CTGTCA +CTGTCT +CTGTCC +CTGTCG +CTGTGA +CTGTGT +CTGTGC +CTGTGG +CTGCAA +CTGCAT +CTGCAC +CTGCAG +CTGCTA +CTGCTT +CTGCTC +CTGCTG +CTGCCA +CTGCCT +CTGCCC +CTGCCG +CTGCGA +CTGCGT +CTGCGC +CTGCGG +CTGGAA +CTGGAT +CTGGAC +CTGGAG +CTGGTA +CTGGTT +CTGGTC +CTGGTG +CTGGCA +CTGGCT +CTGGCC +CTGGCG +CTGGGA +CTGGGT +CTGGGC +CTGGGG +CCAAAA +CCAAAT +CCAAAC +CCAAAG +CCAATA +CCAATT +CCAATC +CCAATG +CCAACA +CCAACT +CCAACC +CCAACG +CCAAGA +CCAAGT +CCAAGC +CCAAGG +CCATAA +CCATAT +CCATAC +CCATAG +CCATTA +CCATTT +CCATTC +CCATTG +CCATCA +CCATCT +CCATCC +CCATCG +CCATGA +CCATGT +CCATGC +CCATGG +CCACAA +CCACAT +CCACAC +CCACAG +CCACTA +CCACTT +CCACTC +CCACTG +CCACCA +CCACCT +CCACCC +CCACCG +CCACGA +CCACGT +CCACGC +CCACGG +CCAGAA +CCAGAT +CCAGAC +CCAGAG +CCAGTA +CCAGTT +CCAGTC +CCAGTG +CCAGCA +CCAGCT +CCAGCC +CCAGCG +CCAGGA +CCAGGT +CCAGGC +CCAGGG +CCTAAA +CCTAAT +CCTAAC +CCTAAG +CCTATA +CCTATT +CCTATC +CCTATG +CCTACA +CCTACT +CCTACC +CCTACG +CCTAGA +CCTAGT +CCTAGC +CCTAGG +CCTTAA +CCTTAT +CCTTAC +CCTTAG +CCTTTA +CCTTTT +CCTTTC +CCTTTG +CCTTCA +CCTTCT +CCTTCC +CCTTCG +CCTTGA +CCTTGT +CCTTGC +CCTTGG +CCTCAA +CCTCAT +CCTCAC +CCTCAG +CCTCTA +CCTCTT +CCTCTC +CCTCTG +CCTCCA +CCTCCT +CCTCCC +CCTCCG +CCTCGA +CCTCGT +CCTCGC +CCTCGG +CCTGAA +CCTGAT +CCTGAC +CCTGAG +CCTGTA +CCTGTT +CCTGTC +CCTGTG +CCTGCA +CCTGCT +CCTGCC +CCTGCG +CCTGGA +CCTGGT +CCTGGC +CCTGGG +CCCAAA +CCCAAT +CCCAAC +CCCAAG +CCCATA +CCCATT +CCCATC +CCCATG +CCCACA +CCCACT +CCCACC +CCCACG +CCCAGA +CCCAGT +CCCAGC +CCCAGG +CCCTAA +CCCTAT +CCCTAC +CCCTAG +CCCTTA +CCCTTT +CCCTTC +CCCTTG +CCCTCA +CCCTCT +CCCTCC +CCCTCG +CCCTGA +CCCTGT +CCCTGC +CCCTGG +CCCCAA +CCCCAT +CCCCAC +CCCCAG +CCCCTA +CCCCTT +CCCCTC +CCCCTG +CCCCCA +CCCCCT +CCCCCC +CCCCCG +CCCCGA +CCCCGT +CCCCGC +CCCCGG +CCCGAA +CCCGAT +CCCGAC +CCCGAG +CCCGTA +CCCGTT +CCCGTC +CCCGTG +CCCGCA +CCCGCT +CCCGCC +CCCGCG +CCCGGA +CCCGGT +CCCGGC +CCCGGG +CCGAAA +CCGAAT +CCGAAC +CCGAAG +CCGATA +CCGATT +CCGATC +CCGATG +CCGACA +CCGACT +CCGACC +CCGACG +CCGAGA +CCGAGT +CCGAGC +CCGAGG +CCGTAA +CCGTAT +CCGTAC +CCGTAG +CCGTTA +CCGTTT +CCGTTC +CCGTTG +CCGTCA +CCGTCT +CCGTCC +CCGTCG +CCGTGA +CCGTGT +CCGTGC +CCGTGG +CCGCAA +CCGCAT +CCGCAC +CCGCAG +CCGCTA +CCGCTT +CCGCTC +CCGCTG +CCGCCA +CCGCCT +CCGCCC +CCGCCG +CCGCGA +CCGCGT +CCGCGC +CCGCGG +CCGGAA +CCGGAT +CCGGAC +CCGGAG +CCGGTA +CCGGTT +CCGGTC +CCGGTG +CCGGCA +CCGGCT +CCGGCC +CCGGCG +CCGGGA +CCGGGT +CCGGGC +CCGGGG +CGAAAA +CGAAAT +CGAAAC +CGAAAG +CGAATA +CGAATT +CGAATC +CGAATG +CGAACA +CGAACT +CGAACC +CGAACG +CGAAGA +CGAAGT +CGAAGC +CGAAGG +CGATAA +CGATAT +CGATAC +CGATAG +CGATTA +CGATTT +CGATTC +CGATTG +CGATCA +CGATCT +CGATCC +CGATCG +CGATGA +CGATGT +CGATGC +CGATGG +CGACAA +CGACAT +CGACAC +CGACAG +CGACTA +CGACTT +CGACTC +CGACTG +CGACCA +CGACCT +CGACCC +CGACCG +CGACGA +CGACGT +CGACGC +CGACGG +CGAGAA +CGAGAT +CGAGAC +CGAGAG +CGAGTA +CGAGTT +CGAGTC +CGAGTG +CGAGCA +CGAGCT +CGAGCC +CGAGCG +CGAGGA +CGAGGT +CGAGGC +CGAGGG +CGTAAA +CGTAAT +CGTAAC +CGTAAG +CGTATA +CGTATT +CGTATC +CGTATG +CGTACA +CGTACT +CGTACC +CGTACG +CGTAGA +CGTAGT +CGTAGC +CGTAGG +CGTTAA +CGTTAT +CGTTAC +CGTTAG +CGTTTA +CGTTTT +CGTTTC +CGTTTG +CGTTCA +CGTTCT +CGTTCC +CGTTCG +CGTTGA +CGTTGT +CGTTGC +CGTTGG +CGTCAA +CGTCAT +CGTCAC +CGTCAG +CGTCTA +CGTCTT +CGTCTC +CGTCTG +CGTCCA +CGTCCT +CGTCCC +CGTCCG +CGTCGA +CGTCGT +CGTCGC +CGTCGG +CGTGAA +CGTGAT +CGTGAC +CGTGAG +CGTGTA +CGTGTT +CGTGTC +CGTGTG +CGTGCA +CGTGCT +CGTGCC +CGTGCG +CGTGGA +CGTGGT +CGTGGC +CGTGGG +CGCAAA +CGCAAT +CGCAAC +CGCAAG +CGCATA +CGCATT +CGCATC +CGCATG +CGCACA +CGCACT +CGCACC +CGCACG +CGCAGA +CGCAGT +CGCAGC +CGCAGG +CGCTAA +CGCTAT +CGCTAC +CGCTAG +CGCTTA +CGCTTT +CGCTTC +CGCTTG +CGCTCA +CGCTCT +CGCTCC +CGCTCG +CGCTGA +CGCTGT +CGCTGC +CGCTGG +CGCCAA +CGCCAT +CGCCAC +CGCCAG +CGCCTA +CGCCTT +CGCCTC +CGCCTG +CGCCCA +CGCCCT +CGCCCC +CGCCCG +CGCCGA +CGCCGT +CGCCGC +CGCCGG +CGCGAA +CGCGAT +CGCGAC +CGCGAG +CGCGTA +CGCGTT +CGCGTC +CGCGTG +CGCGCA +CGCGCT +CGCGCC +CGCGCG +CGCGGA +CGCGGT +CGCGGC +CGCGGG +CGGAAA +CGGAAT +CGGAAC +CGGAAG +CGGATA +CGGATT +CGGATC +CGGATG +CGGACA +CGGACT +CGGACC +CGGACG +CGGAGA +CGGAGT +CGGAGC +CGGAGG +CGGTAA +CGGTAT +CGGTAC +CGGTAG +CGGTTA +CGGTTT +CGGTTC +CGGTTG +CGGTCA +CGGTCT +CGGTCC +CGGTCG +CGGTGA +CGGTGT +CGGTGC +CGGTGG +CGGCAA +CGGCAT +CGGCAC +CGGCAG +CGGCTA +CGGCTT +CGGCTC +CGGCTG +CGGCCA +CGGCCT +CGGCCC +CGGCCG +CGGCGA +CGGCGT +CGGCGC +CGGCGG +CGGGAA +CGGGAT +CGGGAC +CGGGAG +CGGGTA +CGGGTT +CGGGTC +CGGGTG +CGGGCA +CGGGCT +CGGGCC +CGGGCG +CGGGGA +CGGGGT +CGGGGC +CGGGGG +GAAAAA +GAAAAT +GAAAAC +GAAAAG +GAAATA +GAAATT +GAAATC +GAAATG +GAAACA +GAAACT +GAAACC +GAAACG +GAAAGA +GAAAGT +GAAAGC +GAAAGG +GAATAA +GAATAT +GAATAC +GAATAG +GAATTA +GAATTT +GAATTC +GAATTG +GAATCA +GAATCT +GAATCC +GAATCG +GAATGA +GAATGT +GAATGC +GAATGG +GAACAA +GAACAT +GAACAC +GAACAG +GAACTA +GAACTT +GAACTC +GAACTG +GAACCA +GAACCT +GAACCC +GAACCG +GAACGA +GAACGT +GAACGC +GAACGG +GAAGAA +GAAGAT +GAAGAC +GAAGAG +GAAGTA +GAAGTT +GAAGTC +GAAGTG +GAAGCA +GAAGCT +GAAGCC +GAAGCG +GAAGGA +GAAGGT +GAAGGC +GAAGGG +GATAAA +GATAAT +GATAAC +GATAAG +GATATA +GATATT +GATATC +GATATG +GATACA +GATACT +GATACC +GATACG +GATAGA +GATAGT +GATAGC +GATAGG +GATTAA +GATTAT +GATTAC +GATTAG +GATTTA +GATTTT +GATTTC +GATTTG +GATTCA +GATTCT +GATTCC +GATTCG +GATTGA +GATTGT +GATTGC +GATTGG +GATCAA +GATCAT +GATCAC +GATCAG +GATCTA +GATCTT +GATCTC +GATCTG +GATCCA +GATCCT +GATCCC +GATCCG +GATCGA +GATCGT +GATCGC +GATCGG +GATGAA +GATGAT +GATGAC +GATGAG +GATGTA +GATGTT +GATGTC +GATGTG +GATGCA +GATGCT +GATGCC +GATGCG +GATGGA +GATGGT +GATGGC +GATGGG +GACAAA +GACAAT +GACAAC +GACAAG +GACATA +GACATT +GACATC +GACATG +GACACA +GACACT +GACACC +GACACG +GACAGA +GACAGT +GACAGC +GACAGG +GACTAA +GACTAT +GACTAC +GACTAG +GACTTA +GACTTT +GACTTC +GACTTG +GACTCA +GACTCT +GACTCC +GACTCG +GACTGA +GACTGT +GACTGC +GACTGG +GACCAA +GACCAT +GACCAC +GACCAG +GACCTA +GACCTT +GACCTC +GACCTG +GACCCA +GACCCT +GACCCC +GACCCG +GACCGA +GACCGT +GACCGC +GACCGG +GACGAA +GACGAT +GACGAC +GACGAG +GACGTA +GACGTT +GACGTC +GACGTG +GACGCA +GACGCT +GACGCC +GACGCG +GACGGA +GACGGT +GACGGC +GACGGG +GAGAAA +GAGAAT +GAGAAC +GAGAAG +GAGATA +GAGATT +GAGATC +GAGATG +GAGACA +GAGACT +GAGACC +GAGACG +GAGAGA +GAGAGT +GAGAGC +GAGAGG +GAGTAA +GAGTAT +GAGTAC +GAGTAG +GAGTTA +GAGTTT +GAGTTC +GAGTTG +GAGTCA +GAGTCT +GAGTCC +GAGTCG +GAGTGA +GAGTGT +GAGTGC +GAGTGG +GAGCAA +GAGCAT +GAGCAC +GAGCAG +GAGCTA +GAGCTT +GAGCTC +GAGCTG +GAGCCA +GAGCCT +GAGCCC +GAGCCG +GAGCGA +GAGCGT +GAGCGC +GAGCGG +GAGGAA +GAGGAT +GAGGAC +GAGGAG +GAGGTA +GAGGTT +GAGGTC +GAGGTG +GAGGCA +GAGGCT +GAGGCC +GAGGCG +GAGGGA +GAGGGT +GAGGGC +GAGGGG +GTAAAA +GTAAAT +GTAAAC +GTAAAG +GTAATA +GTAATT +GTAATC +GTAATG +GTAACA +GTAACT +GTAACC +GTAACG +GTAAGA +GTAAGT +GTAAGC +GTAAGG +GTATAA +GTATAT +GTATAC +GTATAG +GTATTA +GTATTT +GTATTC +GTATTG +GTATCA +GTATCT +GTATCC +GTATCG +GTATGA +GTATGT +GTATGC +GTATGG +GTACAA +GTACAT +GTACAC +GTACAG +GTACTA +GTACTT +GTACTC +GTACTG +GTACCA +GTACCT +GTACCC +GTACCG +GTACGA +GTACGT +GTACGC +GTACGG +GTAGAA +GTAGAT +GTAGAC +GTAGAG +GTAGTA +GTAGTT +GTAGTC +GTAGTG +GTAGCA +GTAGCT +GTAGCC +GTAGCG +GTAGGA +GTAGGT +GTAGGC +GTAGGG +GTTAAA +GTTAAT +GTTAAC +GTTAAG +GTTATA +GTTATT +GTTATC +GTTATG +GTTACA +GTTACT +GTTACC +GTTACG +GTTAGA +GTTAGT +GTTAGC +GTTAGG +GTTTAA +GTTTAT +GTTTAC +GTTTAG +GTTTTA +GTTTTT +GTTTTC +GTTTTG +GTTTCA +GTTTCT +GTTTCC +GTTTCG +GTTTGA +GTTTGT +GTTTGC +GTTTGG +GTTCAA +GTTCAT +GTTCAC +GTTCAG +GTTCTA +GTTCTT +GTTCTC +GTTCTG +GTTCCA +GTTCCT +GTTCCC +GTTCCG +GTTCGA +GTTCGT +GTTCGC +GTTCGG +GTTGAA +GTTGAT +GTTGAC +GTTGAG +GTTGTA +GTTGTT +GTTGTC +GTTGTG +GTTGCA +GTTGCT +GTTGCC +GTTGCG +GTTGGA +GTTGGT +GTTGGC +GTTGGG +GTCAAA +GTCAAT +GTCAAC +GTCAAG +GTCATA +GTCATT +GTCATC +GTCATG +GTCACA +GTCACT +GTCACC +GTCACG +GTCAGA +GTCAGT +GTCAGC +GTCAGG +GTCTAA +GTCTAT +GTCTAC +GTCTAG +GTCTTA +GTCTTT +GTCTTC +GTCTTG +GTCTCA +GTCTCT +GTCTCC +GTCTCG +GTCTGA +GTCTGT +GTCTGC +GTCTGG +GTCCAA +GTCCAT +GTCCAC +GTCCAG +GTCCTA +GTCCTT +GTCCTC +GTCCTG +GTCCCA +GTCCCT +GTCCCC +GTCCCG +GTCCGA +GTCCGT +GTCCGC +GTCCGG +GTCGAA +GTCGAT +GTCGAC +GTCGAG +GTCGTA +GTCGTT +GTCGTC +GTCGTG +GTCGCA +GTCGCT +GTCGCC +GTCGCG +GTCGGA +GTCGGT +GTCGGC +GTCGGG +GTGAAA +GTGAAT +GTGAAC +GTGAAG +GTGATA +GTGATT +GTGATC +GTGATG +GTGACA +GTGACT +GTGACC +GTGACG +GTGAGA +GTGAGT +GTGAGC +GTGAGG +GTGTAA +GTGTAT +GTGTAC +GTGTAG +GTGTTA +GTGTTT +GTGTTC +GTGTTG +GTGTCA +GTGTCT +GTGTCC +GTGTCG +GTGTGA +GTGTGT +GTGTGC +GTGTGG +GTGCAA +GTGCAT +GTGCAC +GTGCAG +GTGCTA +GTGCTT +GTGCTC +GTGCTG +GTGCCA +GTGCCT +GTGCCC +GTGCCG +GTGCGA +GTGCGT +GTGCGC +GTGCGG +GTGGAA +GTGGAT +GTGGAC +GTGGAG +GTGGTA +GTGGTT +GTGGTC +GTGGTG +GTGGCA +GTGGCT +GTGGCC +GTGGCG +GTGGGA +GTGGGT +GTGGGC +GTGGGG +GCAAAA +GCAAAT +GCAAAC +GCAAAG +GCAATA +GCAATT +GCAATC +GCAATG +GCAACA +GCAACT +GCAACC +GCAACG +GCAAGA +GCAAGT +GCAAGC +GCAAGG +GCATAA +GCATAT +GCATAC +GCATAG +GCATTA +GCATTT +GCATTC +GCATTG +GCATCA +GCATCT +GCATCC +GCATCG +GCATGA +GCATGT +GCATGC +GCATGG +GCACAA +GCACAT +GCACAC +GCACAG +GCACTA +GCACTT +GCACTC +GCACTG +GCACCA +GCACCT +GCACCC +GCACCG +GCACGA +GCACGT +GCACGC +GCACGG +GCAGAA +GCAGAT +GCAGAC +GCAGAG +GCAGTA +GCAGTT +GCAGTC +GCAGTG +GCAGCA +GCAGCT +GCAGCC +GCAGCG +GCAGGA +GCAGGT +GCAGGC +GCAGGG +GCTAAA +GCTAAT +GCTAAC +GCTAAG +GCTATA +GCTATT +GCTATC +GCTATG +GCTACA +GCTACT +GCTACC +GCTACG +GCTAGA +GCTAGT +GCTAGC +GCTAGG +GCTTAA +GCTTAT +GCTTAC +GCTTAG +GCTTTA +GCTTTT +GCTTTC +GCTTTG +GCTTCA +GCTTCT +GCTTCC +GCTTCG +GCTTGA +GCTTGT +GCTTGC +GCTTGG +GCTCAA +GCTCAT +GCTCAC +GCTCAG +GCTCTA +GCTCTT +GCTCTC +GCTCTG +GCTCCA +GCTCCT +GCTCCC +GCTCCG +GCTCGA +GCTCGT +GCTCGC +GCTCGG +GCTGAA +GCTGAT +GCTGAC +GCTGAG +GCTGTA +GCTGTT +GCTGTC +GCTGTG +GCTGCA +GCTGCT +GCTGCC +GCTGCG +GCTGGA +GCTGGT +GCTGGC +GCTGGG +GCCAAA +GCCAAT +GCCAAC +GCCAAG +GCCATA +GCCATT +GCCATC +GCCATG +GCCACA +GCCACT +GCCACC +GCCACG +GCCAGA +GCCAGT +GCCAGC +GCCAGG +GCCTAA +GCCTAT +GCCTAC +GCCTAG +GCCTTA +GCCTTT +GCCTTC +GCCTTG +GCCTCA +GCCTCT +GCCTCC +GCCTCG +GCCTGA +GCCTGT +GCCTGC +GCCTGG +GCCCAA +GCCCAT +GCCCAC +GCCCAG +GCCCTA +GCCCTT +GCCCTC +GCCCTG +GCCCCA +GCCCCT +GCCCCC +GCCCCG +GCCCGA +GCCCGT +GCCCGC +GCCCGG +GCCGAA +GCCGAT +GCCGAC +GCCGAG +GCCGTA +GCCGTT +GCCGTC +GCCGTG +GCCGCA +GCCGCT +GCCGCC +GCCGCG +GCCGGA +GCCGGT +GCCGGC +GCCGGG +GCGAAA +GCGAAT +GCGAAC +GCGAAG +GCGATA +GCGATT +GCGATC +GCGATG +GCGACA +GCGACT +GCGACC +GCGACG +GCGAGA +GCGAGT +GCGAGC +GCGAGG +GCGTAA +GCGTAT +GCGTAC +GCGTAG +GCGTTA +GCGTTT +GCGTTC +GCGTTG +GCGTCA +GCGTCT +GCGTCC +GCGTCG +GCGTGA +GCGTGT +GCGTGC +GCGTGG +GCGCAA +GCGCAT +GCGCAC +GCGCAG +GCGCTA +GCGCTT +GCGCTC +GCGCTG +GCGCCA +GCGCCT +GCGCCC +GCGCCG +GCGCGA +GCGCGT +GCGCGC +GCGCGG +GCGGAA +GCGGAT +GCGGAC +GCGGAG +GCGGTA +GCGGTT +GCGGTC +GCGGTG +GCGGCA +GCGGCT +GCGGCC +GCGGCG +GCGGGA +GCGGGT +GCGGGC +GCGGGG +GGAAAA +GGAAAT +GGAAAC +GGAAAG +GGAATA +GGAATT +GGAATC +GGAATG +GGAACA +GGAACT +GGAACC +GGAACG +GGAAGA +GGAAGT +GGAAGC +GGAAGG +GGATAA +GGATAT +GGATAC +GGATAG +GGATTA +GGATTT +GGATTC +GGATTG +GGATCA +GGATCT +GGATCC +GGATCG +GGATGA +GGATGT +GGATGC +GGATGG +GGACAA +GGACAT +GGACAC +GGACAG +GGACTA +GGACTT +GGACTC +GGACTG +GGACCA +GGACCT +GGACCC +GGACCG +GGACGA +GGACGT +GGACGC +GGACGG +GGAGAA +GGAGAT +GGAGAC +GGAGAG +GGAGTA +GGAGTT +GGAGTC +GGAGTG +GGAGCA +GGAGCT +GGAGCC +GGAGCG +GGAGGA +GGAGGT +GGAGGC +GGAGGG +GGTAAA +GGTAAT +GGTAAC +GGTAAG +GGTATA +GGTATT +GGTATC +GGTATG +GGTACA +GGTACT +GGTACC +GGTACG +GGTAGA +GGTAGT +GGTAGC +GGTAGG +GGTTAA +GGTTAT +GGTTAC +GGTTAG +GGTTTA +GGTTTT +GGTTTC +GGTTTG +GGTTCA +GGTTCT +GGTTCC +GGTTCG +GGTTGA +GGTTGT +GGTTGC +GGTTGG +GGTCAA +GGTCAT +GGTCAC +GGTCAG +GGTCTA +GGTCTT +GGTCTC +GGTCTG +GGTCCA +GGTCCT +GGTCCC +GGTCCG +GGTCGA +GGTCGT +GGTCGC +GGTCGG +GGTGAA +GGTGAT +GGTGAC +GGTGAG +GGTGTA +GGTGTT +GGTGTC +GGTGTG +GGTGCA +GGTGCT +GGTGCC +GGTGCG +GGTGGA +GGTGGT +GGTGGC +GGTGGG +GGCAAA +GGCAAT +GGCAAC +GGCAAG +GGCATA +GGCATT +GGCATC +GGCATG +GGCACA +GGCACT +GGCACC +GGCACG +GGCAGA +GGCAGT +GGCAGC +GGCAGG +GGCTAA +GGCTAT +GGCTAC +GGCTAG +GGCTTA +GGCTTT +GGCTTC +GGCTTG +GGCTCA +GGCTCT +GGCTCC +GGCTCG +GGCTGA +GGCTGT +GGCTGC +GGCTGG +GGCCAA +GGCCAT +GGCCAC +GGCCAG +GGCCTA +GGCCTT +GGCCTC +GGCCTG +GGCCCA +GGCCCT +GGCCCC +GGCCCG +GGCCGA +GGCCGT +GGCCGC +GGCCGG +GGCGAA +GGCGAT +GGCGAC +GGCGAG +GGCGTA +GGCGTT +GGCGTC +GGCGTG +GGCGCA +GGCGCT +GGCGCC +GGCGCG +GGCGGA +GGCGGT +GGCGGC +GGCGGG +GGGAAA +GGGAAT +GGGAAC +GGGAAG +GGGATA +GGGATT +GGGATC +GGGATG +GGGACA +GGGACT +GGGACC +GGGACG +GGGAGA +GGGAGT +GGGAGC +GGGAGG +GGGTAA +GGGTAT +GGGTAC +GGGTAG +GGGTTA +GGGTTT +GGGTTC +GGGTTG +GGGTCA +GGGTCT +GGGTCC +GGGTCG +GGGTGA +GGGTGT +GGGTGC +GGGTGG +GGGCAA +GGGCAT +GGGCAC +GGGCAG +GGGCTA +GGGCTT +GGGCTC +GGGCTG +GGGCCA +GGGCCT +GGGCCC +GGGCCG +GGGCGA +GGGCGT +GGGCGC +GGGCGG +GGGGAA +GGGGAT +GGGGAC +GGGGAG +GGGGTA +GGGGTT +GGGGTC +GGGGTG +GGGGCA +GGGGCT +GGGGCC +GGGGCG +GGGGGA +GGGGGT +GGGGGC +GGGGGG diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/README.md b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..500fd976d92b217455f181bc98c77e612d9a18d9 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/README.md @@ -0,0 +1,202 @@ +--- +base_model: fabihamakhdoomi/TinyDNABERT +library_name: peft +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.12.0 \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/adapter_config.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..04c63e99bbc6b810449dece367c0afb165125b92 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/adapter_config.json @@ -0,0 +1,42 @@ +{ + "alpha_pattern": {}, + "auto_mapping": null, + "base_model_name_or_path": "fabihamakhdoomi/TinyDNABERT", + "bias": "none", + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 16, + "lora_dropout": 0.2, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": [ + "classifier", + "score" + ], + "peft_type": "LORA", + "r": 8, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "bert.encoder.layer.1.attention.self.query", + "bert.encoder.layer.0.output.dense", + "bert.encoder.layer.1.output.dense", + "bert.encoder.layer.0.intermediate.dense", + "bert.encoder.layer.1.intermediate.dense", + "bert.encoder.layer.0.attention.output.dense", + "bert.encoder.layer.1.attention.output.dense", + "bert.encoder.layer.0.attention.self.query", + "bert.encoder.layer.0.attention.self.value", + "bert.encoder.layer.1.attention.self.key", + "bert.encoder.layer.1.attention.self.value", + "bert.encoder.layer.0.attention.self.key" + ], + "task_type": "SEQ_CLS", + "use_dora": false, + "use_rslora": false +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/adapter_model.safetensors b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/adapter_model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..8fcfd6d62a3a716278289f9b22a1db3d6981e291 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/adapter_model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:01ae74eee26320c4069117d7149afba823150e8604e1188a46408989813e907c +size 151952 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/optimizer.pt b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..d3f04f12fbfe30599686ad73ff323f4b08bca34e --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0eeb7af704a9c3048e8393abab5ddb4cc79638cfdea092c8b0c7e3425014c586 +size 312682 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/rng_state.pth b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..bd3d1cad75d68826b1d075775d8e5141764e5725 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:855054099544841bdc3a4d2a62131897c2be69ce849e4324026c75100489f631 +size 14244 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/scheduler.pt b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..14aa799e6f5c93344d1b233b91d0bda9d8eb59e7 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:eca4da5723e14faf16c3be96d23919390da5288d0d9db40021d332ee1df25b93 +size 1064 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/special_tokens_map.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..a8b3208c2884c4efb86e49300fdd3dc877220cdf --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/special_tokens_map.json @@ -0,0 +1,7 @@ +{ + "cls_token": "[CLS]", + "mask_token": "[MASK]", + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/tokenizer.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..ebd782de6963b9032ce9374cfe30f5275be1a30b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/tokenizer.json @@ -0,0 +1,4265 @@ +{ + "version": "1.0", + "truncation": { + "direction": "Right", + "max_length": 512, + "strategy": "LongestFirst", + "stride": 0 + }, + "padding": { + "strategy": { + "Fixed": 512 + }, + "direction": "Right", + "pad_to_multiple_of": null, + "pad_id": 0, + "pad_type_id": 0, + "pad_token": "[PAD]" + }, + "added_tokens": [ + { + "id": 0, + "content": "[PAD]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 1, + "content": "[UNK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 2, + "content": "[CLS]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 3, + "content": "[SEP]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 4, + "content": "[MASK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + } + ], + "normalizer": { + "type": "BertNormalizer", + "clean_text": true, + "handle_chinese_chars": true, + "strip_accents": null, + "lowercase": false + }, + "pre_tokenizer": { + "type": "BertPreTokenizer" + }, + "post_processor": { + "type": "TemplateProcessing", + "single": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + } + ], + "pair": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "B", + "type_id": 1 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 1 + } + } + ], + "special_tokens": { + "[CLS]": { + "id": "[CLS]", + "ids": [ + 2 + ], + "tokens": [ + "[CLS]" + ] + }, + "[SEP]": { + "id": "[SEP]", + "ids": [ + 3 + ], + "tokens": [ + "[SEP]" + ] + } + } + }, + "decoder": { + "type": "WordPiece", + "prefix": "##", + "cleanup": true + }, + "model": { + "type": "WordPiece", + "unk_token": "[UNK]", + "continuing_subword_prefix": "##", + "max_input_chars_per_word": 100, + "vocab": { + "[PAD]": 0, + "[UNK]": 1, + "[CLS]": 2, + "[SEP]": 3, + "[MASK]": 4, + "AAAAAA": 5, + "AAAAAT": 6, + "AAAAAC": 7, + "AAAAAG": 8, + "AAAATA": 9, + "AAAATT": 10, + "AAAATC": 11, + "AAAATG": 12, + "AAAACA": 13, + "AAAACT": 14, + "AAAACC": 15, + "AAAACG": 16, + "AAAAGA": 17, + "AAAAGT": 18, + "AAAAGC": 19, + "AAAAGG": 20, + "AAATAA": 21, + "AAATAT": 22, + "AAATAC": 23, + "AAATAG": 24, + "AAATTA": 25, + "AAATTT": 26, + "AAATTC": 27, + "AAATTG": 28, + "AAATCA": 29, + "AAATCT": 30, + "AAATCC": 31, + "AAATCG": 32, + "AAATGA": 33, + "AAATGT": 34, + "AAATGC": 35, + "AAATGG": 36, + "AAACAA": 37, + "AAACAT": 38, + "AAACAC": 39, + "AAACAG": 40, + "AAACTA": 41, + "AAACTT": 42, + "AAACTC": 43, + "AAACTG": 44, + "AAACCA": 45, + "AAACCT": 46, + "AAACCC": 47, + "AAACCG": 48, + "AAACGA": 49, + "AAACGT": 50, + "AAACGC": 51, + "AAACGG": 52, + "AAAGAA": 53, + "AAAGAT": 54, + "AAAGAC": 55, + "AAAGAG": 56, + "AAAGTA": 57, + "AAAGTT": 58, + "AAAGTC": 59, + "AAAGTG": 60, + "AAAGCA": 61, + "AAAGCT": 62, + "AAAGCC": 63, + "AAAGCG": 64, + "AAAGGA": 65, + "AAAGGT": 66, + "AAAGGC": 67, + "AAAGGG": 68, + "AATAAA": 69, + "AATAAT": 70, + "AATAAC": 71, + "AATAAG": 72, + "AATATA": 73, + "AATATT": 74, + "AATATC": 75, + "AATATG": 76, + "AATACA": 77, + "AATACT": 78, + "AATACC": 79, + "AATACG": 80, + "AATAGA": 81, + "AATAGT": 82, + "AATAGC": 83, + "AATAGG": 84, + "AATTAA": 85, + "AATTAT": 86, + "AATTAC": 87, + "AATTAG": 88, + "AATTTA": 89, + "AATTTT": 90, + "AATTTC": 91, + "AATTTG": 92, + "AATTCA": 93, + "AATTCT": 94, + "AATTCC": 95, + "AATTCG": 96, + "AATTGA": 97, + "AATTGT": 98, + "AATTGC": 99, + "AATTGG": 100, + "AATCAA": 101, + "AATCAT": 102, + "AATCAC": 103, + "AATCAG": 104, + "AATCTA": 105, + "AATCTT": 106, + "AATCTC": 107, + "AATCTG": 108, + "AATCCA": 109, + "AATCCT": 110, + "AATCCC": 111, + "AATCCG": 112, + "AATCGA": 113, + "AATCGT": 114, + "AATCGC": 115, + "AATCGG": 116, + "AATGAA": 117, + "AATGAT": 118, + "AATGAC": 119, + "AATGAG": 120, + "AATGTA": 121, + "AATGTT": 122, + "AATGTC": 123, + "AATGTG": 124, + "AATGCA": 125, + "AATGCT": 126, + "AATGCC": 127, + "AATGCG": 128, + "AATGGA": 129, + "AATGGT": 130, + "AATGGC": 131, + "AATGGG": 132, + "AACAAA": 133, + "AACAAT": 134, + "AACAAC": 135, + "AACAAG": 136, + "AACATA": 137, + "AACATT": 138, + "AACATC": 139, + "AACATG": 140, + "AACACA": 141, + "AACACT": 142, + "AACACC": 143, + "AACACG": 144, + "AACAGA": 145, + "AACAGT": 146, + "AACAGC": 147, + "AACAGG": 148, + "AACTAA": 149, + "AACTAT": 150, + "AACTAC": 151, + "AACTAG": 152, + "AACTTA": 153, + "AACTTT": 154, + "AACTTC": 155, + "AACTTG": 156, + "AACTCA": 157, + "AACTCT": 158, + "AACTCC": 159, + "AACTCG": 160, + "AACTGA": 161, + "AACTGT": 162, + "AACTGC": 163, + "AACTGG": 164, + "AACCAA": 165, + "AACCAT": 166, + "AACCAC": 167, + "AACCAG": 168, + "AACCTA": 169, + "AACCTT": 170, + "AACCTC": 171, + "AACCTG": 172, + "AACCCA": 173, + "AACCCT": 174, + "AACCCC": 175, + "AACCCG": 176, + "AACCGA": 177, + "AACCGT": 178, + "AACCGC": 179, + "AACCGG": 180, + "AACGAA": 181, + "AACGAT": 182, + "AACGAC": 183, + "AACGAG": 184, + "AACGTA": 185, + "AACGTT": 186, + "AACGTC": 187, + "AACGTG": 188, + "AACGCA": 189, + "AACGCT": 190, + "AACGCC": 191, + "AACGCG": 192, + "AACGGA": 193, + "AACGGT": 194, + "AACGGC": 195, + "AACGGG": 196, + "AAGAAA": 197, + "AAGAAT": 198, + "AAGAAC": 199, + "AAGAAG": 200, + "AAGATA": 201, + "AAGATT": 202, + "AAGATC": 203, + "AAGATG": 204, + "AAGACA": 205, + "AAGACT": 206, + "AAGACC": 207, + "AAGACG": 208, + "AAGAGA": 209, + "AAGAGT": 210, + "AAGAGC": 211, + "AAGAGG": 212, + "AAGTAA": 213, + "AAGTAT": 214, + "AAGTAC": 215, + "AAGTAG": 216, + "AAGTTA": 217, + "AAGTTT": 218, + "AAGTTC": 219, + "AAGTTG": 220, + "AAGTCA": 221, + "AAGTCT": 222, + "AAGTCC": 223, + "AAGTCG": 224, + "AAGTGA": 225, + "AAGTGT": 226, + "AAGTGC": 227, + "AAGTGG": 228, + "AAGCAA": 229, + "AAGCAT": 230, + "AAGCAC": 231, + "AAGCAG": 232, + "AAGCTA": 233, + "AAGCTT": 234, + "AAGCTC": 235, + "AAGCTG": 236, + "AAGCCA": 237, + "AAGCCT": 238, + "AAGCCC": 239, + "AAGCCG": 240, + "AAGCGA": 241, + "AAGCGT": 242, + "AAGCGC": 243, + "AAGCGG": 244, + "AAGGAA": 245, + "AAGGAT": 246, + "AAGGAC": 247, + "AAGGAG": 248, + "AAGGTA": 249, + "AAGGTT": 250, + "AAGGTC": 251, + "AAGGTG": 252, + "AAGGCA": 253, + "AAGGCT": 254, + "AAGGCC": 255, + "AAGGCG": 256, + "AAGGGA": 257, + "AAGGGT": 258, + "AAGGGC": 259, + "AAGGGG": 260, + "ATAAAA": 261, + "ATAAAT": 262, + "ATAAAC": 263, + "ATAAAG": 264, + "ATAATA": 265, + "ATAATT": 266, + "ATAATC": 267, + "ATAATG": 268, + "ATAACA": 269, + "ATAACT": 270, + "ATAACC": 271, + "ATAACG": 272, + "ATAAGA": 273, + "ATAAGT": 274, + "ATAAGC": 275, + "ATAAGG": 276, + "ATATAA": 277, + "ATATAT": 278, + "ATATAC": 279, + "ATATAG": 280, + "ATATTA": 281, + "ATATTT": 282, + "ATATTC": 283, + "ATATTG": 284, + "ATATCA": 285, + "ATATCT": 286, + "ATATCC": 287, + "ATATCG": 288, + "ATATGA": 289, + "ATATGT": 290, + "ATATGC": 291, + "ATATGG": 292, + "ATACAA": 293, + "ATACAT": 294, + "ATACAC": 295, + "ATACAG": 296, + "ATACTA": 297, + "ATACTT": 298, + "ATACTC": 299, + "ATACTG": 300, + "ATACCA": 301, + "ATACCT": 302, + "ATACCC": 303, + "ATACCG": 304, + "ATACGA": 305, + "ATACGT": 306, + "ATACGC": 307, + "ATACGG": 308, + "ATAGAA": 309, + "ATAGAT": 310, + "ATAGAC": 311, + "ATAGAG": 312, + "ATAGTA": 313, + "ATAGTT": 314, + "ATAGTC": 315, + "ATAGTG": 316, + "ATAGCA": 317, + "ATAGCT": 318, + "ATAGCC": 319, + "ATAGCG": 320, + "ATAGGA": 321, + "ATAGGT": 322, + "ATAGGC": 323, + "ATAGGG": 324, + "ATTAAA": 325, + "ATTAAT": 326, + "ATTAAC": 327, + "ATTAAG": 328, + "ATTATA": 329, + "ATTATT": 330, + "ATTATC": 331, + "ATTATG": 332, + "ATTACA": 333, + "ATTACT": 334, + "ATTACC": 335, + "ATTACG": 336, + "ATTAGA": 337, + "ATTAGT": 338, + "ATTAGC": 339, + "ATTAGG": 340, + "ATTTAA": 341, + "ATTTAT": 342, + "ATTTAC": 343, + "ATTTAG": 344, + "ATTTTA": 345, + "ATTTTT": 346, + "ATTTTC": 347, + "ATTTTG": 348, + "ATTTCA": 349, + "ATTTCT": 350, + "ATTTCC": 351, + "ATTTCG": 352, + "ATTTGA": 353, + "ATTTGT": 354, + "ATTTGC": 355, + "ATTTGG": 356, + "ATTCAA": 357, + "ATTCAT": 358, + "ATTCAC": 359, + "ATTCAG": 360, + "ATTCTA": 361, + "ATTCTT": 362, + "ATTCTC": 363, + "ATTCTG": 364, + "ATTCCA": 365, + "ATTCCT": 366, + "ATTCCC": 367, + "ATTCCG": 368, + "ATTCGA": 369, + "ATTCGT": 370, + "ATTCGC": 371, + "ATTCGG": 372, + "ATTGAA": 373, + "ATTGAT": 374, + "ATTGAC": 375, + "ATTGAG": 376, + "ATTGTA": 377, + "ATTGTT": 378, + "ATTGTC": 379, + "ATTGTG": 380, + "ATTGCA": 381, + "ATTGCT": 382, + "ATTGCC": 383, + "ATTGCG": 384, + "ATTGGA": 385, + "ATTGGT": 386, + "ATTGGC": 387, + "ATTGGG": 388, + "ATCAAA": 389, + "ATCAAT": 390, + "ATCAAC": 391, + "ATCAAG": 392, + "ATCATA": 393, + "ATCATT": 394, + "ATCATC": 395, + "ATCATG": 396, + "ATCACA": 397, + "ATCACT": 398, + "ATCACC": 399, + "ATCACG": 400, + "ATCAGA": 401, + "ATCAGT": 402, + "ATCAGC": 403, + "ATCAGG": 404, + "ATCTAA": 405, + "ATCTAT": 406, + "ATCTAC": 407, + "ATCTAG": 408, + "ATCTTA": 409, + "ATCTTT": 410, + "ATCTTC": 411, + "ATCTTG": 412, + "ATCTCA": 413, + "ATCTCT": 414, + "ATCTCC": 415, + "ATCTCG": 416, + "ATCTGA": 417, + "ATCTGT": 418, + "ATCTGC": 419, + "ATCTGG": 420, + "ATCCAA": 421, + "ATCCAT": 422, + "ATCCAC": 423, + "ATCCAG": 424, + "ATCCTA": 425, + "ATCCTT": 426, + "ATCCTC": 427, + "ATCCTG": 428, + "ATCCCA": 429, + "ATCCCT": 430, + "ATCCCC": 431, + "ATCCCG": 432, + "ATCCGA": 433, + "ATCCGT": 434, + "ATCCGC": 435, + "ATCCGG": 436, + "ATCGAA": 437, + "ATCGAT": 438, + "ATCGAC": 439, + "ATCGAG": 440, + "ATCGTA": 441, + "ATCGTT": 442, + "ATCGTC": 443, + "ATCGTG": 444, + "ATCGCA": 445, + "ATCGCT": 446, + "ATCGCC": 447, + "ATCGCG": 448, + "ATCGGA": 449, + "ATCGGT": 450, + "ATCGGC": 451, + "ATCGGG": 452, + "ATGAAA": 453, + "ATGAAT": 454, + "ATGAAC": 455, + "ATGAAG": 456, + "ATGATA": 457, + "ATGATT": 458, + "ATGATC": 459, + "ATGATG": 460, + "ATGACA": 461, + "ATGACT": 462, + "ATGACC": 463, + "ATGACG": 464, + "ATGAGA": 465, + "ATGAGT": 466, + "ATGAGC": 467, + "ATGAGG": 468, + "ATGTAA": 469, + "ATGTAT": 470, + "ATGTAC": 471, + "ATGTAG": 472, + "ATGTTA": 473, + "ATGTTT": 474, + "ATGTTC": 475, + "ATGTTG": 476, + "ATGTCA": 477, + "ATGTCT": 478, + "ATGTCC": 479, + "ATGTCG": 480, + "ATGTGA": 481, + "ATGTGT": 482, + "ATGTGC": 483, + "ATGTGG": 484, + "ATGCAA": 485, + "ATGCAT": 486, + "ATGCAC": 487, + "ATGCAG": 488, + "ATGCTA": 489, + "ATGCTT": 490, + "ATGCTC": 491, + "ATGCTG": 492, + "ATGCCA": 493, + "ATGCCT": 494, + "ATGCCC": 495, + "ATGCCG": 496, + "ATGCGA": 497, + "ATGCGT": 498, + "ATGCGC": 499, + "ATGCGG": 500, + "ATGGAA": 501, + "ATGGAT": 502, + "ATGGAC": 503, + "ATGGAG": 504, + "ATGGTA": 505, + "ATGGTT": 506, + "ATGGTC": 507, + "ATGGTG": 508, + "ATGGCA": 509, + "ATGGCT": 510, + "ATGGCC": 511, + "ATGGCG": 512, + "ATGGGA": 513, + "ATGGGT": 514, + "ATGGGC": 515, + "ATGGGG": 516, + "ACAAAA": 517, + "ACAAAT": 518, + "ACAAAC": 519, + "ACAAAG": 520, + "ACAATA": 521, + "ACAATT": 522, + "ACAATC": 523, + "ACAATG": 524, + "ACAACA": 525, + "ACAACT": 526, + "ACAACC": 527, + "ACAACG": 528, + "ACAAGA": 529, + "ACAAGT": 530, + "ACAAGC": 531, + "ACAAGG": 532, + "ACATAA": 533, + "ACATAT": 534, + "ACATAC": 535, + "ACATAG": 536, + "ACATTA": 537, + "ACATTT": 538, + "ACATTC": 539, + "ACATTG": 540, + "ACATCA": 541, + "ACATCT": 542, + "ACATCC": 543, + "ACATCG": 544, + "ACATGA": 545, + "ACATGT": 546, + "ACATGC": 547, + "ACATGG": 548, + "ACACAA": 549, + "ACACAT": 550, + "ACACAC": 551, + "ACACAG": 552, + "ACACTA": 553, + "ACACTT": 554, + "ACACTC": 555, + "ACACTG": 556, + "ACACCA": 557, + "ACACCT": 558, + "ACACCC": 559, + "ACACCG": 560, + "ACACGA": 561, + "ACACGT": 562, + "ACACGC": 563, + "ACACGG": 564, + "ACAGAA": 565, + "ACAGAT": 566, + "ACAGAC": 567, + "ACAGAG": 568, + "ACAGTA": 569, + "ACAGTT": 570, + "ACAGTC": 571, + "ACAGTG": 572, + "ACAGCA": 573, + "ACAGCT": 574, + "ACAGCC": 575, + "ACAGCG": 576, + "ACAGGA": 577, + "ACAGGT": 578, + "ACAGGC": 579, + "ACAGGG": 580, + "ACTAAA": 581, + "ACTAAT": 582, + "ACTAAC": 583, + "ACTAAG": 584, + "ACTATA": 585, + "ACTATT": 586, + "ACTATC": 587, + "ACTATG": 588, + "ACTACA": 589, + "ACTACT": 590, + "ACTACC": 591, + "ACTACG": 592, + "ACTAGA": 593, + "ACTAGT": 594, + "ACTAGC": 595, + "ACTAGG": 596, + "ACTTAA": 597, + "ACTTAT": 598, + "ACTTAC": 599, + "ACTTAG": 600, + "ACTTTA": 601, + "ACTTTT": 602, + "ACTTTC": 603, + "ACTTTG": 604, + "ACTTCA": 605, + "ACTTCT": 606, + "ACTTCC": 607, + "ACTTCG": 608, + "ACTTGA": 609, + "ACTTGT": 610, + "ACTTGC": 611, + "ACTTGG": 612, + "ACTCAA": 613, + "ACTCAT": 614, + "ACTCAC": 615, + "ACTCAG": 616, + "ACTCTA": 617, + "ACTCTT": 618, + "ACTCTC": 619, + "ACTCTG": 620, + "ACTCCA": 621, + "ACTCCT": 622, + "ACTCCC": 623, + "ACTCCG": 624, + "ACTCGA": 625, + "ACTCGT": 626, + "ACTCGC": 627, + "ACTCGG": 628, + "ACTGAA": 629, + "ACTGAT": 630, + "ACTGAC": 631, + "ACTGAG": 632, + "ACTGTA": 633, + "ACTGTT": 634, + "ACTGTC": 635, + "ACTGTG": 636, + "ACTGCA": 637, + "ACTGCT": 638, + "ACTGCC": 639, + "ACTGCG": 640, + "ACTGGA": 641, + "ACTGGT": 642, + "ACTGGC": 643, + "ACTGGG": 644, + "ACCAAA": 645, + "ACCAAT": 646, + "ACCAAC": 647, + "ACCAAG": 648, + "ACCATA": 649, + "ACCATT": 650, + "ACCATC": 651, + "ACCATG": 652, + "ACCACA": 653, + "ACCACT": 654, + "ACCACC": 655, + "ACCACG": 656, + "ACCAGA": 657, + "ACCAGT": 658, + "ACCAGC": 659, + "ACCAGG": 660, + "ACCTAA": 661, + "ACCTAT": 662, + "ACCTAC": 663, + "ACCTAG": 664, + "ACCTTA": 665, + "ACCTTT": 666, + "ACCTTC": 667, + "ACCTTG": 668, + "ACCTCA": 669, + "ACCTCT": 670, + "ACCTCC": 671, + "ACCTCG": 672, + "ACCTGA": 673, + "ACCTGT": 674, + "ACCTGC": 675, + "ACCTGG": 676, + "ACCCAA": 677, + "ACCCAT": 678, + "ACCCAC": 679, + "ACCCAG": 680, + "ACCCTA": 681, + "ACCCTT": 682, + "ACCCTC": 683, + "ACCCTG": 684, + "ACCCCA": 685, + "ACCCCT": 686, + "ACCCCC": 687, + "ACCCCG": 688, + "ACCCGA": 689, + "ACCCGT": 690, + "ACCCGC": 691, + "ACCCGG": 692, + "ACCGAA": 693, + "ACCGAT": 694, + "ACCGAC": 695, + "ACCGAG": 696, + "ACCGTA": 697, + "ACCGTT": 698, + "ACCGTC": 699, + "ACCGTG": 700, + "ACCGCA": 701, + "ACCGCT": 702, + "ACCGCC": 703, + "ACCGCG": 704, + "ACCGGA": 705, + "ACCGGT": 706, + "ACCGGC": 707, + "ACCGGG": 708, + "ACGAAA": 709, + "ACGAAT": 710, + "ACGAAC": 711, + "ACGAAG": 712, + "ACGATA": 713, + "ACGATT": 714, + "ACGATC": 715, + "ACGATG": 716, + "ACGACA": 717, + "ACGACT": 718, + "ACGACC": 719, + "ACGACG": 720, + "ACGAGA": 721, + "ACGAGT": 722, + "ACGAGC": 723, + "ACGAGG": 724, + "ACGTAA": 725, + "ACGTAT": 726, + "ACGTAC": 727, + "ACGTAG": 728, + "ACGTTA": 729, + "ACGTTT": 730, + "ACGTTC": 731, + "ACGTTG": 732, + "ACGTCA": 733, + "ACGTCT": 734, + "ACGTCC": 735, + "ACGTCG": 736, + "ACGTGA": 737, + "ACGTGT": 738, + "ACGTGC": 739, + "ACGTGG": 740, + "ACGCAA": 741, + "ACGCAT": 742, + "ACGCAC": 743, + "ACGCAG": 744, + "ACGCTA": 745, + "ACGCTT": 746, + "ACGCTC": 747, + "ACGCTG": 748, + "ACGCCA": 749, + "ACGCCT": 750, + "ACGCCC": 751, + "ACGCCG": 752, + "ACGCGA": 753, + "ACGCGT": 754, + "ACGCGC": 755, + "ACGCGG": 756, + "ACGGAA": 757, + "ACGGAT": 758, + "ACGGAC": 759, + "ACGGAG": 760, + "ACGGTA": 761, + "ACGGTT": 762, + "ACGGTC": 763, + "ACGGTG": 764, + "ACGGCA": 765, + "ACGGCT": 766, + "ACGGCC": 767, + "ACGGCG": 768, + "ACGGGA": 769, + "ACGGGT": 770, + "ACGGGC": 771, + "ACGGGG": 772, + "AGAAAA": 773, + "AGAAAT": 774, + "AGAAAC": 775, + "AGAAAG": 776, + "AGAATA": 777, + "AGAATT": 778, + "AGAATC": 779, + "AGAATG": 780, + "AGAACA": 781, + "AGAACT": 782, + "AGAACC": 783, + "AGAACG": 784, + "AGAAGA": 785, + "AGAAGT": 786, + "AGAAGC": 787, + "AGAAGG": 788, + "AGATAA": 789, + "AGATAT": 790, + "AGATAC": 791, + "AGATAG": 792, + "AGATTA": 793, + "AGATTT": 794, + "AGATTC": 795, + "AGATTG": 796, + "AGATCA": 797, + "AGATCT": 798, + "AGATCC": 799, + "AGATCG": 800, + "AGATGA": 801, + "AGATGT": 802, + "AGATGC": 803, + "AGATGG": 804, + "AGACAA": 805, + "AGACAT": 806, + "AGACAC": 807, + "AGACAG": 808, + "AGACTA": 809, + "AGACTT": 810, + "AGACTC": 811, + "AGACTG": 812, + "AGACCA": 813, + "AGACCT": 814, + "AGACCC": 815, + "AGACCG": 816, + "AGACGA": 817, + "AGACGT": 818, + "AGACGC": 819, + "AGACGG": 820, + "AGAGAA": 821, + "AGAGAT": 822, + "AGAGAC": 823, + "AGAGAG": 824, + "AGAGTA": 825, + "AGAGTT": 826, + "AGAGTC": 827, + "AGAGTG": 828, + "AGAGCA": 829, + "AGAGCT": 830, + "AGAGCC": 831, + "AGAGCG": 832, + "AGAGGA": 833, + "AGAGGT": 834, + "AGAGGC": 835, + "AGAGGG": 836, + "AGTAAA": 837, + "AGTAAT": 838, + "AGTAAC": 839, + "AGTAAG": 840, + "AGTATA": 841, + "AGTATT": 842, + "AGTATC": 843, + "AGTATG": 844, + "AGTACA": 845, + "AGTACT": 846, + "AGTACC": 847, + "AGTACG": 848, + "AGTAGA": 849, + "AGTAGT": 850, + "AGTAGC": 851, + "AGTAGG": 852, + "AGTTAA": 853, + "AGTTAT": 854, + "AGTTAC": 855, + "AGTTAG": 856, + "AGTTTA": 857, + "AGTTTT": 858, + "AGTTTC": 859, + "AGTTTG": 860, + "AGTTCA": 861, + "AGTTCT": 862, + "AGTTCC": 863, + "AGTTCG": 864, + "AGTTGA": 865, + "AGTTGT": 866, + "AGTTGC": 867, + "AGTTGG": 868, + "AGTCAA": 869, + "AGTCAT": 870, + "AGTCAC": 871, + "AGTCAG": 872, + "AGTCTA": 873, + "AGTCTT": 874, + "AGTCTC": 875, + "AGTCTG": 876, + "AGTCCA": 877, + "AGTCCT": 878, + "AGTCCC": 879, + "AGTCCG": 880, + "AGTCGA": 881, + "AGTCGT": 882, + "AGTCGC": 883, + "AGTCGG": 884, + "AGTGAA": 885, + "AGTGAT": 886, + "AGTGAC": 887, + "AGTGAG": 888, + "AGTGTA": 889, + "AGTGTT": 890, + "AGTGTC": 891, + "AGTGTG": 892, + "AGTGCA": 893, + "AGTGCT": 894, + "AGTGCC": 895, + "AGTGCG": 896, + "AGTGGA": 897, + "AGTGGT": 898, + "AGTGGC": 899, + "AGTGGG": 900, + "AGCAAA": 901, + "AGCAAT": 902, + "AGCAAC": 903, + "AGCAAG": 904, + "AGCATA": 905, + "AGCATT": 906, + "AGCATC": 907, + "AGCATG": 908, + "AGCACA": 909, + "AGCACT": 910, + "AGCACC": 911, + "AGCACG": 912, + "AGCAGA": 913, + "AGCAGT": 914, + "AGCAGC": 915, + "AGCAGG": 916, + "AGCTAA": 917, + "AGCTAT": 918, + "AGCTAC": 919, + "AGCTAG": 920, + "AGCTTA": 921, + "AGCTTT": 922, + "AGCTTC": 923, + "AGCTTG": 924, + "AGCTCA": 925, + "AGCTCT": 926, + "AGCTCC": 927, + "AGCTCG": 928, + "AGCTGA": 929, + "AGCTGT": 930, + "AGCTGC": 931, + "AGCTGG": 932, + "AGCCAA": 933, + "AGCCAT": 934, + "AGCCAC": 935, + "AGCCAG": 936, + "AGCCTA": 937, + "AGCCTT": 938, + "AGCCTC": 939, + "AGCCTG": 940, + "AGCCCA": 941, + "AGCCCT": 942, + "AGCCCC": 943, + "AGCCCG": 944, + "AGCCGA": 945, + "AGCCGT": 946, + "AGCCGC": 947, + "AGCCGG": 948, + "AGCGAA": 949, + "AGCGAT": 950, + "AGCGAC": 951, + "AGCGAG": 952, + "AGCGTA": 953, + "AGCGTT": 954, + "AGCGTC": 955, + "AGCGTG": 956, + "AGCGCA": 957, + "AGCGCT": 958, + "AGCGCC": 959, + "AGCGCG": 960, + "AGCGGA": 961, + "AGCGGT": 962, + "AGCGGC": 963, + "AGCGGG": 964, + "AGGAAA": 965, + "AGGAAT": 966, + "AGGAAC": 967, + "AGGAAG": 968, + "AGGATA": 969, + "AGGATT": 970, + "AGGATC": 971, + "AGGATG": 972, + "AGGACA": 973, + "AGGACT": 974, + "AGGACC": 975, + "AGGACG": 976, + "AGGAGA": 977, + "AGGAGT": 978, + "AGGAGC": 979, + "AGGAGG": 980, + "AGGTAA": 981, + "AGGTAT": 982, + "AGGTAC": 983, + "AGGTAG": 984, + "AGGTTA": 985, + "AGGTTT": 986, + "AGGTTC": 987, + "AGGTTG": 988, + "AGGTCA": 989, + "AGGTCT": 990, + "AGGTCC": 991, + "AGGTCG": 992, + "AGGTGA": 993, + "AGGTGT": 994, + "AGGTGC": 995, + "AGGTGG": 996, + "AGGCAA": 997, + "AGGCAT": 998, + "AGGCAC": 999, + "AGGCAG": 1000, + "AGGCTA": 1001, + "AGGCTT": 1002, + "AGGCTC": 1003, + "AGGCTG": 1004, + "AGGCCA": 1005, + "AGGCCT": 1006, + "AGGCCC": 1007, + "AGGCCG": 1008, + "AGGCGA": 1009, + "AGGCGT": 1010, + "AGGCGC": 1011, + "AGGCGG": 1012, + "AGGGAA": 1013, + "AGGGAT": 1014, + "AGGGAC": 1015, + "AGGGAG": 1016, + "AGGGTA": 1017, + "AGGGTT": 1018, + "AGGGTC": 1019, + "AGGGTG": 1020, + "AGGGCA": 1021, + "AGGGCT": 1022, + "AGGGCC": 1023, + "AGGGCG": 1024, + "AGGGGA": 1025, + "AGGGGT": 1026, + "AGGGGC": 1027, + "AGGGGG": 1028, + "TAAAAA": 1029, + "TAAAAT": 1030, + "TAAAAC": 1031, + "TAAAAG": 1032, + "TAAATA": 1033, + "TAAATT": 1034, + "TAAATC": 1035, + "TAAATG": 1036, + "TAAACA": 1037, + "TAAACT": 1038, + "TAAACC": 1039, + "TAAACG": 1040, + "TAAAGA": 1041, + "TAAAGT": 1042, + "TAAAGC": 1043, + "TAAAGG": 1044, + "TAATAA": 1045, + "TAATAT": 1046, + "TAATAC": 1047, + "TAATAG": 1048, + "TAATTA": 1049, + "TAATTT": 1050, + "TAATTC": 1051, + "TAATTG": 1052, + "TAATCA": 1053, + "TAATCT": 1054, + "TAATCC": 1055, + "TAATCG": 1056, + "TAATGA": 1057, + "TAATGT": 1058, + "TAATGC": 1059, + "TAATGG": 1060, + "TAACAA": 1061, + "TAACAT": 1062, + "TAACAC": 1063, + "TAACAG": 1064, + "TAACTA": 1065, + "TAACTT": 1066, + "TAACTC": 1067, + "TAACTG": 1068, + "TAACCA": 1069, + "TAACCT": 1070, + "TAACCC": 1071, + "TAACCG": 1072, + "TAACGA": 1073, + "TAACGT": 1074, + "TAACGC": 1075, + "TAACGG": 1076, + "TAAGAA": 1077, + "TAAGAT": 1078, + "TAAGAC": 1079, + "TAAGAG": 1080, + "TAAGTA": 1081, + "TAAGTT": 1082, + "TAAGTC": 1083, + "TAAGTG": 1084, + "TAAGCA": 1085, + "TAAGCT": 1086, + "TAAGCC": 1087, + "TAAGCG": 1088, + "TAAGGA": 1089, + "TAAGGT": 1090, + "TAAGGC": 1091, + "TAAGGG": 1092, + "TATAAA": 1093, + "TATAAT": 1094, + "TATAAC": 1095, + "TATAAG": 1096, + "TATATA": 1097, + "TATATT": 1098, + "TATATC": 1099, + "TATATG": 1100, + "TATACA": 1101, + "TATACT": 1102, + "TATACC": 1103, + "TATACG": 1104, + "TATAGA": 1105, + "TATAGT": 1106, + "TATAGC": 1107, + "TATAGG": 1108, + "TATTAA": 1109, + "TATTAT": 1110, + "TATTAC": 1111, + "TATTAG": 1112, + "TATTTA": 1113, + "TATTTT": 1114, + "TATTTC": 1115, + "TATTTG": 1116, + "TATTCA": 1117, + "TATTCT": 1118, + "TATTCC": 1119, + "TATTCG": 1120, + "TATTGA": 1121, + "TATTGT": 1122, + "TATTGC": 1123, + "TATTGG": 1124, + "TATCAA": 1125, + "TATCAT": 1126, + "TATCAC": 1127, + "TATCAG": 1128, + "TATCTA": 1129, + "TATCTT": 1130, + "TATCTC": 1131, + "TATCTG": 1132, + "TATCCA": 1133, + "TATCCT": 1134, + "TATCCC": 1135, + "TATCCG": 1136, + "TATCGA": 1137, + "TATCGT": 1138, + "TATCGC": 1139, + "TATCGG": 1140, + "TATGAA": 1141, + "TATGAT": 1142, + "TATGAC": 1143, + "TATGAG": 1144, + "TATGTA": 1145, + "TATGTT": 1146, + "TATGTC": 1147, + "TATGTG": 1148, + "TATGCA": 1149, + "TATGCT": 1150, + "TATGCC": 1151, + "TATGCG": 1152, + "TATGGA": 1153, + "TATGGT": 1154, + "TATGGC": 1155, + "TATGGG": 1156, + "TACAAA": 1157, + "TACAAT": 1158, + "TACAAC": 1159, + "TACAAG": 1160, + "TACATA": 1161, + "TACATT": 1162, + "TACATC": 1163, + "TACATG": 1164, + "TACACA": 1165, + "TACACT": 1166, + "TACACC": 1167, + "TACACG": 1168, + "TACAGA": 1169, + "TACAGT": 1170, + "TACAGC": 1171, + "TACAGG": 1172, + "TACTAA": 1173, + "TACTAT": 1174, + "TACTAC": 1175, + "TACTAG": 1176, + "TACTTA": 1177, + "TACTTT": 1178, + "TACTTC": 1179, + "TACTTG": 1180, + "TACTCA": 1181, + "TACTCT": 1182, + "TACTCC": 1183, + "TACTCG": 1184, + "TACTGA": 1185, + "TACTGT": 1186, + "TACTGC": 1187, + "TACTGG": 1188, + "TACCAA": 1189, + "TACCAT": 1190, + "TACCAC": 1191, + "TACCAG": 1192, + "TACCTA": 1193, + "TACCTT": 1194, + "TACCTC": 1195, + "TACCTG": 1196, + "TACCCA": 1197, + "TACCCT": 1198, + "TACCCC": 1199, + "TACCCG": 1200, + "TACCGA": 1201, + "TACCGT": 1202, + "TACCGC": 1203, + "TACCGG": 1204, + "TACGAA": 1205, + "TACGAT": 1206, + "TACGAC": 1207, + "TACGAG": 1208, + "TACGTA": 1209, + "TACGTT": 1210, + "TACGTC": 1211, + "TACGTG": 1212, + "TACGCA": 1213, + "TACGCT": 1214, + "TACGCC": 1215, + "TACGCG": 1216, + "TACGGA": 1217, + "TACGGT": 1218, + "TACGGC": 1219, + "TACGGG": 1220, + "TAGAAA": 1221, + "TAGAAT": 1222, + "TAGAAC": 1223, + "TAGAAG": 1224, + "TAGATA": 1225, + "TAGATT": 1226, + "TAGATC": 1227, + "TAGATG": 1228, + "TAGACA": 1229, + "TAGACT": 1230, + "TAGACC": 1231, + "TAGACG": 1232, + "TAGAGA": 1233, + "TAGAGT": 1234, + "TAGAGC": 1235, + "TAGAGG": 1236, + "TAGTAA": 1237, + "TAGTAT": 1238, + "TAGTAC": 1239, + "TAGTAG": 1240, + "TAGTTA": 1241, + "TAGTTT": 1242, + "TAGTTC": 1243, + "TAGTTG": 1244, + "TAGTCA": 1245, + "TAGTCT": 1246, + "TAGTCC": 1247, + "TAGTCG": 1248, + "TAGTGA": 1249, + "TAGTGT": 1250, + "TAGTGC": 1251, + "TAGTGG": 1252, + "TAGCAA": 1253, + "TAGCAT": 1254, + "TAGCAC": 1255, + "TAGCAG": 1256, + "TAGCTA": 1257, + "TAGCTT": 1258, + "TAGCTC": 1259, + "TAGCTG": 1260, + "TAGCCA": 1261, + "TAGCCT": 1262, + "TAGCCC": 1263, + "TAGCCG": 1264, + "TAGCGA": 1265, + "TAGCGT": 1266, + "TAGCGC": 1267, + "TAGCGG": 1268, + "TAGGAA": 1269, + "TAGGAT": 1270, + "TAGGAC": 1271, + "TAGGAG": 1272, + "TAGGTA": 1273, + "TAGGTT": 1274, + "TAGGTC": 1275, + "TAGGTG": 1276, + "TAGGCA": 1277, + "TAGGCT": 1278, + "TAGGCC": 1279, + "TAGGCG": 1280, + "TAGGGA": 1281, + "TAGGGT": 1282, + "TAGGGC": 1283, + "TAGGGG": 1284, + "TTAAAA": 1285, + "TTAAAT": 1286, + "TTAAAC": 1287, + "TTAAAG": 1288, + "TTAATA": 1289, + "TTAATT": 1290, + "TTAATC": 1291, + "TTAATG": 1292, + "TTAACA": 1293, + "TTAACT": 1294, + "TTAACC": 1295, + "TTAACG": 1296, + "TTAAGA": 1297, + "TTAAGT": 1298, + "TTAAGC": 1299, + "TTAAGG": 1300, + "TTATAA": 1301, + "TTATAT": 1302, + "TTATAC": 1303, + "TTATAG": 1304, + "TTATTA": 1305, + "TTATTT": 1306, + "TTATTC": 1307, + "TTATTG": 1308, + "TTATCA": 1309, + "TTATCT": 1310, + "TTATCC": 1311, + "TTATCG": 1312, + "TTATGA": 1313, + "TTATGT": 1314, + "TTATGC": 1315, + "TTATGG": 1316, + "TTACAA": 1317, + "TTACAT": 1318, + "TTACAC": 1319, + "TTACAG": 1320, + "TTACTA": 1321, + "TTACTT": 1322, + "TTACTC": 1323, + "TTACTG": 1324, + "TTACCA": 1325, + "TTACCT": 1326, + "TTACCC": 1327, + "TTACCG": 1328, + "TTACGA": 1329, + "TTACGT": 1330, + "TTACGC": 1331, + "TTACGG": 1332, + "TTAGAA": 1333, + "TTAGAT": 1334, + "TTAGAC": 1335, + "TTAGAG": 1336, + "TTAGTA": 1337, + "TTAGTT": 1338, + "TTAGTC": 1339, + "TTAGTG": 1340, + "TTAGCA": 1341, + "TTAGCT": 1342, + "TTAGCC": 1343, + "TTAGCG": 1344, + "TTAGGA": 1345, + "TTAGGT": 1346, + "TTAGGC": 1347, + "TTAGGG": 1348, + "TTTAAA": 1349, + "TTTAAT": 1350, + "TTTAAC": 1351, + "TTTAAG": 1352, + "TTTATA": 1353, + "TTTATT": 1354, + "TTTATC": 1355, + "TTTATG": 1356, + "TTTACA": 1357, + "TTTACT": 1358, + "TTTACC": 1359, + "TTTACG": 1360, + "TTTAGA": 1361, + "TTTAGT": 1362, + "TTTAGC": 1363, + "TTTAGG": 1364, + "TTTTAA": 1365, + "TTTTAT": 1366, + "TTTTAC": 1367, + "TTTTAG": 1368, + "TTTTTA": 1369, + "TTTTTT": 1370, + "TTTTTC": 1371, + "TTTTTG": 1372, + "TTTTCA": 1373, + "TTTTCT": 1374, + "TTTTCC": 1375, + "TTTTCG": 1376, + "TTTTGA": 1377, + "TTTTGT": 1378, + "TTTTGC": 1379, + "TTTTGG": 1380, + "TTTCAA": 1381, + "TTTCAT": 1382, + "TTTCAC": 1383, + "TTTCAG": 1384, + "TTTCTA": 1385, + "TTTCTT": 1386, + "TTTCTC": 1387, + "TTTCTG": 1388, + "TTTCCA": 1389, + "TTTCCT": 1390, + "TTTCCC": 1391, + "TTTCCG": 1392, + "TTTCGA": 1393, + "TTTCGT": 1394, + "TTTCGC": 1395, + "TTTCGG": 1396, + "TTTGAA": 1397, + "TTTGAT": 1398, + "TTTGAC": 1399, + "TTTGAG": 1400, + "TTTGTA": 1401, + "TTTGTT": 1402, + "TTTGTC": 1403, + "TTTGTG": 1404, + "TTTGCA": 1405, + "TTTGCT": 1406, + "TTTGCC": 1407, + "TTTGCG": 1408, + "TTTGGA": 1409, + "TTTGGT": 1410, + "TTTGGC": 1411, + "TTTGGG": 1412, + "TTCAAA": 1413, + "TTCAAT": 1414, + "TTCAAC": 1415, + "TTCAAG": 1416, + "TTCATA": 1417, + "TTCATT": 1418, + "TTCATC": 1419, + "TTCATG": 1420, + "TTCACA": 1421, + "TTCACT": 1422, + "TTCACC": 1423, + "TTCACG": 1424, + "TTCAGA": 1425, + "TTCAGT": 1426, + "TTCAGC": 1427, + "TTCAGG": 1428, + "TTCTAA": 1429, + "TTCTAT": 1430, + "TTCTAC": 1431, + "TTCTAG": 1432, + "TTCTTA": 1433, + "TTCTTT": 1434, + "TTCTTC": 1435, + "TTCTTG": 1436, + "TTCTCA": 1437, + "TTCTCT": 1438, + "TTCTCC": 1439, + "TTCTCG": 1440, + "TTCTGA": 1441, + "TTCTGT": 1442, + "TTCTGC": 1443, + "TTCTGG": 1444, + "TTCCAA": 1445, + "TTCCAT": 1446, + "TTCCAC": 1447, + "TTCCAG": 1448, + "TTCCTA": 1449, + "TTCCTT": 1450, + "TTCCTC": 1451, + "TTCCTG": 1452, + "TTCCCA": 1453, + "TTCCCT": 1454, + "TTCCCC": 1455, + "TTCCCG": 1456, + "TTCCGA": 1457, + "TTCCGT": 1458, + "TTCCGC": 1459, + "TTCCGG": 1460, + "TTCGAA": 1461, + "TTCGAT": 1462, + "TTCGAC": 1463, + "TTCGAG": 1464, + "TTCGTA": 1465, + "TTCGTT": 1466, + "TTCGTC": 1467, + "TTCGTG": 1468, + "TTCGCA": 1469, + "TTCGCT": 1470, + "TTCGCC": 1471, + "TTCGCG": 1472, + "TTCGGA": 1473, + "TTCGGT": 1474, + "TTCGGC": 1475, + "TTCGGG": 1476, + "TTGAAA": 1477, + "TTGAAT": 1478, + "TTGAAC": 1479, + "TTGAAG": 1480, + "TTGATA": 1481, + "TTGATT": 1482, + "TTGATC": 1483, + "TTGATG": 1484, + "TTGACA": 1485, + "TTGACT": 1486, + "TTGACC": 1487, + "TTGACG": 1488, + "TTGAGA": 1489, + "TTGAGT": 1490, + "TTGAGC": 1491, + "TTGAGG": 1492, + "TTGTAA": 1493, + "TTGTAT": 1494, + "TTGTAC": 1495, + "TTGTAG": 1496, + "TTGTTA": 1497, + "TTGTTT": 1498, + "TTGTTC": 1499, + "TTGTTG": 1500, + "TTGTCA": 1501, + "TTGTCT": 1502, + "TTGTCC": 1503, + "TTGTCG": 1504, + "TTGTGA": 1505, + "TTGTGT": 1506, + "TTGTGC": 1507, + "TTGTGG": 1508, + "TTGCAA": 1509, + "TTGCAT": 1510, + "TTGCAC": 1511, + "TTGCAG": 1512, + "TTGCTA": 1513, + "TTGCTT": 1514, + "TTGCTC": 1515, + "TTGCTG": 1516, + "TTGCCA": 1517, + "TTGCCT": 1518, + "TTGCCC": 1519, + "TTGCCG": 1520, + "TTGCGA": 1521, + "TTGCGT": 1522, + "TTGCGC": 1523, + "TTGCGG": 1524, + "TTGGAA": 1525, + "TTGGAT": 1526, + "TTGGAC": 1527, + "TTGGAG": 1528, + "TTGGTA": 1529, + "TTGGTT": 1530, + "TTGGTC": 1531, + "TTGGTG": 1532, + "TTGGCA": 1533, + "TTGGCT": 1534, + "TTGGCC": 1535, + "TTGGCG": 1536, + "TTGGGA": 1537, + "TTGGGT": 1538, + "TTGGGC": 1539, + "TTGGGG": 1540, + "TCAAAA": 1541, + "TCAAAT": 1542, + "TCAAAC": 1543, + "TCAAAG": 1544, + "TCAATA": 1545, + "TCAATT": 1546, + "TCAATC": 1547, + "TCAATG": 1548, + "TCAACA": 1549, + "TCAACT": 1550, + "TCAACC": 1551, + "TCAACG": 1552, + "TCAAGA": 1553, + "TCAAGT": 1554, + "TCAAGC": 1555, + "TCAAGG": 1556, + "TCATAA": 1557, + "TCATAT": 1558, + "TCATAC": 1559, + "TCATAG": 1560, + "TCATTA": 1561, + "TCATTT": 1562, + "TCATTC": 1563, + "TCATTG": 1564, + "TCATCA": 1565, + "TCATCT": 1566, + "TCATCC": 1567, + "TCATCG": 1568, + "TCATGA": 1569, + "TCATGT": 1570, + "TCATGC": 1571, + "TCATGG": 1572, + "TCACAA": 1573, + "TCACAT": 1574, + "TCACAC": 1575, + "TCACAG": 1576, + "TCACTA": 1577, + "TCACTT": 1578, + "TCACTC": 1579, + "TCACTG": 1580, + "TCACCA": 1581, + "TCACCT": 1582, + "TCACCC": 1583, + "TCACCG": 1584, + "TCACGA": 1585, + "TCACGT": 1586, + "TCACGC": 1587, + "TCACGG": 1588, + "TCAGAA": 1589, + "TCAGAT": 1590, + "TCAGAC": 1591, + "TCAGAG": 1592, + "TCAGTA": 1593, + "TCAGTT": 1594, + "TCAGTC": 1595, + "TCAGTG": 1596, + "TCAGCA": 1597, + "TCAGCT": 1598, + "TCAGCC": 1599, + "TCAGCG": 1600, + "TCAGGA": 1601, + "TCAGGT": 1602, + "TCAGGC": 1603, + "TCAGGG": 1604, + "TCTAAA": 1605, + "TCTAAT": 1606, + "TCTAAC": 1607, + "TCTAAG": 1608, + "TCTATA": 1609, + "TCTATT": 1610, + "TCTATC": 1611, + "TCTATG": 1612, + "TCTACA": 1613, + "TCTACT": 1614, + "TCTACC": 1615, + "TCTACG": 1616, + "TCTAGA": 1617, + "TCTAGT": 1618, + "TCTAGC": 1619, + "TCTAGG": 1620, + "TCTTAA": 1621, + "TCTTAT": 1622, + "TCTTAC": 1623, + "TCTTAG": 1624, + "TCTTTA": 1625, + "TCTTTT": 1626, + "TCTTTC": 1627, + "TCTTTG": 1628, + "TCTTCA": 1629, + "TCTTCT": 1630, + "TCTTCC": 1631, + "TCTTCG": 1632, + "TCTTGA": 1633, + "TCTTGT": 1634, + "TCTTGC": 1635, + "TCTTGG": 1636, + "TCTCAA": 1637, + "TCTCAT": 1638, + "TCTCAC": 1639, + "TCTCAG": 1640, + "TCTCTA": 1641, + "TCTCTT": 1642, + "TCTCTC": 1643, + "TCTCTG": 1644, + "TCTCCA": 1645, + "TCTCCT": 1646, + "TCTCCC": 1647, + "TCTCCG": 1648, + "TCTCGA": 1649, + "TCTCGT": 1650, + "TCTCGC": 1651, + "TCTCGG": 1652, + "TCTGAA": 1653, + "TCTGAT": 1654, + "TCTGAC": 1655, + "TCTGAG": 1656, + "TCTGTA": 1657, + "TCTGTT": 1658, + "TCTGTC": 1659, + "TCTGTG": 1660, + "TCTGCA": 1661, + "TCTGCT": 1662, + "TCTGCC": 1663, + "TCTGCG": 1664, + "TCTGGA": 1665, + "TCTGGT": 1666, + "TCTGGC": 1667, + "TCTGGG": 1668, + "TCCAAA": 1669, + "TCCAAT": 1670, + "TCCAAC": 1671, + "TCCAAG": 1672, + "TCCATA": 1673, + "TCCATT": 1674, + "TCCATC": 1675, + "TCCATG": 1676, + "TCCACA": 1677, + "TCCACT": 1678, + "TCCACC": 1679, + "TCCACG": 1680, + "TCCAGA": 1681, + "TCCAGT": 1682, + "TCCAGC": 1683, + "TCCAGG": 1684, + "TCCTAA": 1685, + "TCCTAT": 1686, + "TCCTAC": 1687, + "TCCTAG": 1688, + "TCCTTA": 1689, + "TCCTTT": 1690, + "TCCTTC": 1691, + "TCCTTG": 1692, + "TCCTCA": 1693, + "TCCTCT": 1694, + "TCCTCC": 1695, + "TCCTCG": 1696, + "TCCTGA": 1697, + "TCCTGT": 1698, + "TCCTGC": 1699, + "TCCTGG": 1700, + "TCCCAA": 1701, + "TCCCAT": 1702, + "TCCCAC": 1703, + "TCCCAG": 1704, + "TCCCTA": 1705, + "TCCCTT": 1706, + "TCCCTC": 1707, + "TCCCTG": 1708, + "TCCCCA": 1709, + "TCCCCT": 1710, + "TCCCCC": 1711, + "TCCCCG": 1712, + "TCCCGA": 1713, + "TCCCGT": 1714, + "TCCCGC": 1715, + "TCCCGG": 1716, + "TCCGAA": 1717, + "TCCGAT": 1718, + "TCCGAC": 1719, + "TCCGAG": 1720, + "TCCGTA": 1721, + "TCCGTT": 1722, + "TCCGTC": 1723, + "TCCGTG": 1724, + "TCCGCA": 1725, + "TCCGCT": 1726, + "TCCGCC": 1727, + "TCCGCG": 1728, + "TCCGGA": 1729, + "TCCGGT": 1730, + "TCCGGC": 1731, + "TCCGGG": 1732, + "TCGAAA": 1733, + "TCGAAT": 1734, + "TCGAAC": 1735, + "TCGAAG": 1736, + "TCGATA": 1737, + "TCGATT": 1738, + "TCGATC": 1739, + "TCGATG": 1740, + "TCGACA": 1741, + "TCGACT": 1742, + "TCGACC": 1743, + "TCGACG": 1744, + "TCGAGA": 1745, + "TCGAGT": 1746, + "TCGAGC": 1747, + "TCGAGG": 1748, + "TCGTAA": 1749, + "TCGTAT": 1750, + "TCGTAC": 1751, + "TCGTAG": 1752, + "TCGTTA": 1753, + "TCGTTT": 1754, + "TCGTTC": 1755, + "TCGTTG": 1756, + "TCGTCA": 1757, + "TCGTCT": 1758, + "TCGTCC": 1759, + "TCGTCG": 1760, + "TCGTGA": 1761, + "TCGTGT": 1762, + "TCGTGC": 1763, + "TCGTGG": 1764, + "TCGCAA": 1765, + "TCGCAT": 1766, + "TCGCAC": 1767, + "TCGCAG": 1768, + "TCGCTA": 1769, + "TCGCTT": 1770, + "TCGCTC": 1771, + "TCGCTG": 1772, + "TCGCCA": 1773, + "TCGCCT": 1774, + "TCGCCC": 1775, + "TCGCCG": 1776, + "TCGCGA": 1777, + "TCGCGT": 1778, + "TCGCGC": 1779, + "TCGCGG": 1780, + "TCGGAA": 1781, + "TCGGAT": 1782, + "TCGGAC": 1783, + "TCGGAG": 1784, + "TCGGTA": 1785, + "TCGGTT": 1786, + "TCGGTC": 1787, + "TCGGTG": 1788, + "TCGGCA": 1789, + "TCGGCT": 1790, + "TCGGCC": 1791, + "TCGGCG": 1792, + "TCGGGA": 1793, + "TCGGGT": 1794, + "TCGGGC": 1795, + "TCGGGG": 1796, + "TGAAAA": 1797, + "TGAAAT": 1798, + "TGAAAC": 1799, + "TGAAAG": 1800, + "TGAATA": 1801, + "TGAATT": 1802, + "TGAATC": 1803, + "TGAATG": 1804, + "TGAACA": 1805, + "TGAACT": 1806, + "TGAACC": 1807, + "TGAACG": 1808, + "TGAAGA": 1809, + "TGAAGT": 1810, + "TGAAGC": 1811, + "TGAAGG": 1812, + "TGATAA": 1813, + "TGATAT": 1814, + "TGATAC": 1815, + "TGATAG": 1816, + "TGATTA": 1817, + "TGATTT": 1818, + "TGATTC": 1819, + "TGATTG": 1820, + "TGATCA": 1821, + "TGATCT": 1822, + "TGATCC": 1823, + "TGATCG": 1824, + "TGATGA": 1825, + "TGATGT": 1826, + "TGATGC": 1827, + "TGATGG": 1828, + "TGACAA": 1829, + "TGACAT": 1830, + "TGACAC": 1831, + "TGACAG": 1832, + "TGACTA": 1833, + "TGACTT": 1834, + "TGACTC": 1835, + "TGACTG": 1836, + "TGACCA": 1837, + "TGACCT": 1838, + "TGACCC": 1839, + "TGACCG": 1840, + "TGACGA": 1841, + "TGACGT": 1842, + "TGACGC": 1843, + "TGACGG": 1844, + "TGAGAA": 1845, + "TGAGAT": 1846, + "TGAGAC": 1847, + "TGAGAG": 1848, + "TGAGTA": 1849, + "TGAGTT": 1850, + "TGAGTC": 1851, + "TGAGTG": 1852, + "TGAGCA": 1853, + "TGAGCT": 1854, + "TGAGCC": 1855, + "TGAGCG": 1856, + "TGAGGA": 1857, + "TGAGGT": 1858, + "TGAGGC": 1859, + "TGAGGG": 1860, + "TGTAAA": 1861, + "TGTAAT": 1862, + "TGTAAC": 1863, + "TGTAAG": 1864, + "TGTATA": 1865, + "TGTATT": 1866, + "TGTATC": 1867, + "TGTATG": 1868, + "TGTACA": 1869, + "TGTACT": 1870, + "TGTACC": 1871, + "TGTACG": 1872, + "TGTAGA": 1873, + "TGTAGT": 1874, + "TGTAGC": 1875, + "TGTAGG": 1876, + "TGTTAA": 1877, + "TGTTAT": 1878, + "TGTTAC": 1879, + "TGTTAG": 1880, + "TGTTTA": 1881, + "TGTTTT": 1882, + "TGTTTC": 1883, + "TGTTTG": 1884, + "TGTTCA": 1885, + "TGTTCT": 1886, + "TGTTCC": 1887, + "TGTTCG": 1888, + "TGTTGA": 1889, + "TGTTGT": 1890, + "TGTTGC": 1891, + "TGTTGG": 1892, + "TGTCAA": 1893, + "TGTCAT": 1894, + "TGTCAC": 1895, + "TGTCAG": 1896, + "TGTCTA": 1897, + "TGTCTT": 1898, + "TGTCTC": 1899, + "TGTCTG": 1900, + "TGTCCA": 1901, + "TGTCCT": 1902, + "TGTCCC": 1903, + "TGTCCG": 1904, + "TGTCGA": 1905, + "TGTCGT": 1906, + "TGTCGC": 1907, + "TGTCGG": 1908, + "TGTGAA": 1909, + "TGTGAT": 1910, + "TGTGAC": 1911, + "TGTGAG": 1912, + "TGTGTA": 1913, + "TGTGTT": 1914, + "TGTGTC": 1915, + "TGTGTG": 1916, + "TGTGCA": 1917, + "TGTGCT": 1918, + "TGTGCC": 1919, + "TGTGCG": 1920, + "TGTGGA": 1921, + "TGTGGT": 1922, + "TGTGGC": 1923, + "TGTGGG": 1924, + "TGCAAA": 1925, + "TGCAAT": 1926, + "TGCAAC": 1927, + "TGCAAG": 1928, + "TGCATA": 1929, + "TGCATT": 1930, + "TGCATC": 1931, + "TGCATG": 1932, + "TGCACA": 1933, + "TGCACT": 1934, + "TGCACC": 1935, + "TGCACG": 1936, + "TGCAGA": 1937, + "TGCAGT": 1938, + "TGCAGC": 1939, + "TGCAGG": 1940, + "TGCTAA": 1941, + "TGCTAT": 1942, + "TGCTAC": 1943, + "TGCTAG": 1944, + "TGCTTA": 1945, + "TGCTTT": 1946, + "TGCTTC": 1947, + "TGCTTG": 1948, + "TGCTCA": 1949, + "TGCTCT": 1950, + "TGCTCC": 1951, + "TGCTCG": 1952, + "TGCTGA": 1953, + "TGCTGT": 1954, + "TGCTGC": 1955, + "TGCTGG": 1956, + "TGCCAA": 1957, + "TGCCAT": 1958, + "TGCCAC": 1959, + "TGCCAG": 1960, + "TGCCTA": 1961, + "TGCCTT": 1962, + "TGCCTC": 1963, + "TGCCTG": 1964, + "TGCCCA": 1965, + "TGCCCT": 1966, + "TGCCCC": 1967, + "TGCCCG": 1968, + "TGCCGA": 1969, + "TGCCGT": 1970, + "TGCCGC": 1971, + "TGCCGG": 1972, + "TGCGAA": 1973, + "TGCGAT": 1974, + "TGCGAC": 1975, + "TGCGAG": 1976, + "TGCGTA": 1977, + "TGCGTT": 1978, + "TGCGTC": 1979, + "TGCGTG": 1980, + "TGCGCA": 1981, + "TGCGCT": 1982, + "TGCGCC": 1983, + "TGCGCG": 1984, + "TGCGGA": 1985, + "TGCGGT": 1986, + "TGCGGC": 1987, + "TGCGGG": 1988, + "TGGAAA": 1989, + "TGGAAT": 1990, + "TGGAAC": 1991, + "TGGAAG": 1992, + "TGGATA": 1993, + "TGGATT": 1994, + "TGGATC": 1995, + "TGGATG": 1996, + "TGGACA": 1997, + "TGGACT": 1998, + "TGGACC": 1999, + "TGGACG": 2000, + "TGGAGA": 2001, + "TGGAGT": 2002, + "TGGAGC": 2003, + "TGGAGG": 2004, + "TGGTAA": 2005, + "TGGTAT": 2006, + "TGGTAC": 2007, + "TGGTAG": 2008, + "TGGTTA": 2009, + "TGGTTT": 2010, + "TGGTTC": 2011, + "TGGTTG": 2012, + "TGGTCA": 2013, + "TGGTCT": 2014, + "TGGTCC": 2015, + "TGGTCG": 2016, + "TGGTGA": 2017, + "TGGTGT": 2018, + "TGGTGC": 2019, + "TGGTGG": 2020, + "TGGCAA": 2021, + "TGGCAT": 2022, + "TGGCAC": 2023, + "TGGCAG": 2024, + "TGGCTA": 2025, + "TGGCTT": 2026, + "TGGCTC": 2027, + "TGGCTG": 2028, + "TGGCCA": 2029, + "TGGCCT": 2030, + "TGGCCC": 2031, + "TGGCCG": 2032, + "TGGCGA": 2033, + "TGGCGT": 2034, + "TGGCGC": 2035, + "TGGCGG": 2036, + "TGGGAA": 2037, + "TGGGAT": 2038, + "TGGGAC": 2039, + "TGGGAG": 2040, + "TGGGTA": 2041, + "TGGGTT": 2042, + "TGGGTC": 2043, + "TGGGTG": 2044, + "TGGGCA": 2045, + "TGGGCT": 2046, + "TGGGCC": 2047, + "TGGGCG": 2048, + "TGGGGA": 2049, + "TGGGGT": 2050, + "TGGGGC": 2051, + "TGGGGG": 2052, + "CAAAAA": 2053, + "CAAAAT": 2054, + "CAAAAC": 2055, + "CAAAAG": 2056, + "CAAATA": 2057, + "CAAATT": 2058, + "CAAATC": 2059, + "CAAATG": 2060, + "CAAACA": 2061, + "CAAACT": 2062, + "CAAACC": 2063, + "CAAACG": 2064, + "CAAAGA": 2065, + "CAAAGT": 2066, + "CAAAGC": 2067, + "CAAAGG": 2068, + "CAATAA": 2069, + "CAATAT": 2070, + "CAATAC": 2071, + "CAATAG": 2072, + "CAATTA": 2073, + "CAATTT": 2074, + "CAATTC": 2075, + "CAATTG": 2076, + "CAATCA": 2077, + "CAATCT": 2078, + "CAATCC": 2079, + "CAATCG": 2080, + "CAATGA": 2081, + "CAATGT": 2082, + "CAATGC": 2083, + "CAATGG": 2084, + "CAACAA": 2085, + "CAACAT": 2086, + "CAACAC": 2087, + "CAACAG": 2088, + "CAACTA": 2089, + "CAACTT": 2090, + "CAACTC": 2091, + "CAACTG": 2092, + "CAACCA": 2093, + "CAACCT": 2094, + "CAACCC": 2095, + "CAACCG": 2096, + "CAACGA": 2097, + "CAACGT": 2098, + "CAACGC": 2099, + "CAACGG": 2100, + "CAAGAA": 2101, + "CAAGAT": 2102, + "CAAGAC": 2103, + "CAAGAG": 2104, + "CAAGTA": 2105, + "CAAGTT": 2106, + "CAAGTC": 2107, + "CAAGTG": 2108, + "CAAGCA": 2109, + "CAAGCT": 2110, + "CAAGCC": 2111, + "CAAGCG": 2112, + "CAAGGA": 2113, + "CAAGGT": 2114, + "CAAGGC": 2115, + "CAAGGG": 2116, + "CATAAA": 2117, + "CATAAT": 2118, + "CATAAC": 2119, + "CATAAG": 2120, + "CATATA": 2121, + "CATATT": 2122, + "CATATC": 2123, + "CATATG": 2124, + "CATACA": 2125, + "CATACT": 2126, + "CATACC": 2127, + "CATACG": 2128, + "CATAGA": 2129, + "CATAGT": 2130, + "CATAGC": 2131, + "CATAGG": 2132, + "CATTAA": 2133, + "CATTAT": 2134, + "CATTAC": 2135, + "CATTAG": 2136, + "CATTTA": 2137, + "CATTTT": 2138, + "CATTTC": 2139, + "CATTTG": 2140, + "CATTCA": 2141, + "CATTCT": 2142, + "CATTCC": 2143, + "CATTCG": 2144, + "CATTGA": 2145, + "CATTGT": 2146, + "CATTGC": 2147, + "CATTGG": 2148, + "CATCAA": 2149, + "CATCAT": 2150, + "CATCAC": 2151, + "CATCAG": 2152, + "CATCTA": 2153, + "CATCTT": 2154, + "CATCTC": 2155, + "CATCTG": 2156, + "CATCCA": 2157, + "CATCCT": 2158, + "CATCCC": 2159, + "CATCCG": 2160, + "CATCGA": 2161, + "CATCGT": 2162, + "CATCGC": 2163, + "CATCGG": 2164, + "CATGAA": 2165, + "CATGAT": 2166, + "CATGAC": 2167, + "CATGAG": 2168, + "CATGTA": 2169, + "CATGTT": 2170, + "CATGTC": 2171, + "CATGTG": 2172, + "CATGCA": 2173, + "CATGCT": 2174, + "CATGCC": 2175, + "CATGCG": 2176, + "CATGGA": 2177, + "CATGGT": 2178, + "CATGGC": 2179, + "CATGGG": 2180, + "CACAAA": 2181, + "CACAAT": 2182, + "CACAAC": 2183, + "CACAAG": 2184, + "CACATA": 2185, + "CACATT": 2186, + "CACATC": 2187, + "CACATG": 2188, + "CACACA": 2189, + "CACACT": 2190, + "CACACC": 2191, + "CACACG": 2192, + "CACAGA": 2193, + "CACAGT": 2194, + "CACAGC": 2195, + "CACAGG": 2196, + "CACTAA": 2197, + "CACTAT": 2198, + "CACTAC": 2199, + "CACTAG": 2200, + "CACTTA": 2201, + "CACTTT": 2202, + "CACTTC": 2203, + "CACTTG": 2204, + "CACTCA": 2205, + "CACTCT": 2206, + "CACTCC": 2207, + "CACTCG": 2208, + "CACTGA": 2209, + "CACTGT": 2210, + "CACTGC": 2211, + "CACTGG": 2212, + "CACCAA": 2213, + "CACCAT": 2214, + "CACCAC": 2215, + "CACCAG": 2216, + "CACCTA": 2217, + "CACCTT": 2218, + "CACCTC": 2219, + "CACCTG": 2220, + "CACCCA": 2221, + "CACCCT": 2222, + "CACCCC": 2223, + "CACCCG": 2224, + "CACCGA": 2225, + "CACCGT": 2226, + "CACCGC": 2227, + "CACCGG": 2228, + "CACGAA": 2229, + "CACGAT": 2230, + "CACGAC": 2231, + "CACGAG": 2232, + "CACGTA": 2233, + "CACGTT": 2234, + "CACGTC": 2235, + "CACGTG": 2236, + "CACGCA": 2237, + "CACGCT": 2238, + "CACGCC": 2239, + "CACGCG": 2240, + "CACGGA": 2241, + "CACGGT": 2242, + "CACGGC": 2243, + "CACGGG": 2244, + "CAGAAA": 2245, + "CAGAAT": 2246, + "CAGAAC": 2247, + "CAGAAG": 2248, + "CAGATA": 2249, + "CAGATT": 2250, + "CAGATC": 2251, + "CAGATG": 2252, + "CAGACA": 2253, + "CAGACT": 2254, + "CAGACC": 2255, + "CAGACG": 2256, + "CAGAGA": 2257, + "CAGAGT": 2258, + "CAGAGC": 2259, + "CAGAGG": 2260, + "CAGTAA": 2261, + "CAGTAT": 2262, + "CAGTAC": 2263, + "CAGTAG": 2264, + "CAGTTA": 2265, + "CAGTTT": 2266, + "CAGTTC": 2267, + "CAGTTG": 2268, + "CAGTCA": 2269, + "CAGTCT": 2270, + "CAGTCC": 2271, + "CAGTCG": 2272, + "CAGTGA": 2273, + "CAGTGT": 2274, + "CAGTGC": 2275, + "CAGTGG": 2276, + "CAGCAA": 2277, + "CAGCAT": 2278, + "CAGCAC": 2279, + "CAGCAG": 2280, + "CAGCTA": 2281, + "CAGCTT": 2282, + "CAGCTC": 2283, + "CAGCTG": 2284, + "CAGCCA": 2285, + "CAGCCT": 2286, + "CAGCCC": 2287, + "CAGCCG": 2288, + "CAGCGA": 2289, + "CAGCGT": 2290, + "CAGCGC": 2291, + "CAGCGG": 2292, + "CAGGAA": 2293, + "CAGGAT": 2294, + "CAGGAC": 2295, + "CAGGAG": 2296, + "CAGGTA": 2297, + "CAGGTT": 2298, + "CAGGTC": 2299, + "CAGGTG": 2300, + "CAGGCA": 2301, + "CAGGCT": 2302, + "CAGGCC": 2303, + "CAGGCG": 2304, + "CAGGGA": 2305, + "CAGGGT": 2306, + "CAGGGC": 2307, + "CAGGGG": 2308, + "CTAAAA": 2309, + "CTAAAT": 2310, + "CTAAAC": 2311, + "CTAAAG": 2312, + "CTAATA": 2313, + "CTAATT": 2314, + "CTAATC": 2315, + "CTAATG": 2316, + "CTAACA": 2317, + "CTAACT": 2318, + "CTAACC": 2319, + "CTAACG": 2320, + "CTAAGA": 2321, + "CTAAGT": 2322, + "CTAAGC": 2323, + "CTAAGG": 2324, + "CTATAA": 2325, + "CTATAT": 2326, + "CTATAC": 2327, + "CTATAG": 2328, + "CTATTA": 2329, + "CTATTT": 2330, + "CTATTC": 2331, + "CTATTG": 2332, + "CTATCA": 2333, + "CTATCT": 2334, + "CTATCC": 2335, + "CTATCG": 2336, + "CTATGA": 2337, + "CTATGT": 2338, + "CTATGC": 2339, + "CTATGG": 2340, + "CTACAA": 2341, + "CTACAT": 2342, + "CTACAC": 2343, + "CTACAG": 2344, + "CTACTA": 2345, + "CTACTT": 2346, + "CTACTC": 2347, + "CTACTG": 2348, + "CTACCA": 2349, + "CTACCT": 2350, + "CTACCC": 2351, + "CTACCG": 2352, + "CTACGA": 2353, + "CTACGT": 2354, + "CTACGC": 2355, + "CTACGG": 2356, + "CTAGAA": 2357, + "CTAGAT": 2358, + "CTAGAC": 2359, + "CTAGAG": 2360, + "CTAGTA": 2361, + "CTAGTT": 2362, + "CTAGTC": 2363, + "CTAGTG": 2364, + "CTAGCA": 2365, + "CTAGCT": 2366, + "CTAGCC": 2367, + "CTAGCG": 2368, + "CTAGGA": 2369, + "CTAGGT": 2370, + "CTAGGC": 2371, + "CTAGGG": 2372, + "CTTAAA": 2373, + "CTTAAT": 2374, + "CTTAAC": 2375, + "CTTAAG": 2376, + "CTTATA": 2377, + "CTTATT": 2378, + "CTTATC": 2379, + "CTTATG": 2380, + "CTTACA": 2381, + "CTTACT": 2382, + "CTTACC": 2383, + "CTTACG": 2384, + "CTTAGA": 2385, + "CTTAGT": 2386, + "CTTAGC": 2387, + "CTTAGG": 2388, + "CTTTAA": 2389, + "CTTTAT": 2390, + "CTTTAC": 2391, + "CTTTAG": 2392, + "CTTTTA": 2393, + "CTTTTT": 2394, + "CTTTTC": 2395, + "CTTTTG": 2396, + "CTTTCA": 2397, + "CTTTCT": 2398, + "CTTTCC": 2399, + "CTTTCG": 2400, + "CTTTGA": 2401, + "CTTTGT": 2402, + "CTTTGC": 2403, + "CTTTGG": 2404, + "CTTCAA": 2405, + "CTTCAT": 2406, + "CTTCAC": 2407, + "CTTCAG": 2408, + "CTTCTA": 2409, + "CTTCTT": 2410, + "CTTCTC": 2411, + "CTTCTG": 2412, + "CTTCCA": 2413, + "CTTCCT": 2414, + "CTTCCC": 2415, + "CTTCCG": 2416, + "CTTCGA": 2417, + "CTTCGT": 2418, + "CTTCGC": 2419, + "CTTCGG": 2420, + "CTTGAA": 2421, + "CTTGAT": 2422, + "CTTGAC": 2423, + "CTTGAG": 2424, + "CTTGTA": 2425, + "CTTGTT": 2426, + "CTTGTC": 2427, + "CTTGTG": 2428, + "CTTGCA": 2429, + "CTTGCT": 2430, + "CTTGCC": 2431, + "CTTGCG": 2432, + "CTTGGA": 2433, + "CTTGGT": 2434, + "CTTGGC": 2435, + "CTTGGG": 2436, + "CTCAAA": 2437, + "CTCAAT": 2438, + "CTCAAC": 2439, + "CTCAAG": 2440, + "CTCATA": 2441, + "CTCATT": 2442, + "CTCATC": 2443, + "CTCATG": 2444, + "CTCACA": 2445, + "CTCACT": 2446, + "CTCACC": 2447, + "CTCACG": 2448, + "CTCAGA": 2449, + "CTCAGT": 2450, + "CTCAGC": 2451, + "CTCAGG": 2452, + "CTCTAA": 2453, + "CTCTAT": 2454, + "CTCTAC": 2455, + "CTCTAG": 2456, + "CTCTTA": 2457, + "CTCTTT": 2458, + "CTCTTC": 2459, + "CTCTTG": 2460, + "CTCTCA": 2461, + "CTCTCT": 2462, + "CTCTCC": 2463, + "CTCTCG": 2464, + "CTCTGA": 2465, + "CTCTGT": 2466, + "CTCTGC": 2467, + "CTCTGG": 2468, + "CTCCAA": 2469, + "CTCCAT": 2470, + "CTCCAC": 2471, + "CTCCAG": 2472, + "CTCCTA": 2473, + "CTCCTT": 2474, + "CTCCTC": 2475, + "CTCCTG": 2476, + "CTCCCA": 2477, + "CTCCCT": 2478, + "CTCCCC": 2479, + "CTCCCG": 2480, + "CTCCGA": 2481, + "CTCCGT": 2482, + "CTCCGC": 2483, + "CTCCGG": 2484, + "CTCGAA": 2485, + "CTCGAT": 2486, + "CTCGAC": 2487, + "CTCGAG": 2488, + "CTCGTA": 2489, + "CTCGTT": 2490, + "CTCGTC": 2491, + "CTCGTG": 2492, + "CTCGCA": 2493, + "CTCGCT": 2494, + "CTCGCC": 2495, + "CTCGCG": 2496, + "CTCGGA": 2497, + "CTCGGT": 2498, + "CTCGGC": 2499, + "CTCGGG": 2500, + "CTGAAA": 2501, + "CTGAAT": 2502, + "CTGAAC": 2503, + "CTGAAG": 2504, + "CTGATA": 2505, + "CTGATT": 2506, + "CTGATC": 2507, + "CTGATG": 2508, + "CTGACA": 2509, + "CTGACT": 2510, + "CTGACC": 2511, + "CTGACG": 2512, + "CTGAGA": 2513, + "CTGAGT": 2514, + "CTGAGC": 2515, + "CTGAGG": 2516, + "CTGTAA": 2517, + "CTGTAT": 2518, + "CTGTAC": 2519, + "CTGTAG": 2520, + "CTGTTA": 2521, + "CTGTTT": 2522, + "CTGTTC": 2523, + "CTGTTG": 2524, + "CTGTCA": 2525, + "CTGTCT": 2526, + "CTGTCC": 2527, + "CTGTCG": 2528, + "CTGTGA": 2529, + "CTGTGT": 2530, + "CTGTGC": 2531, + "CTGTGG": 2532, + "CTGCAA": 2533, + "CTGCAT": 2534, + "CTGCAC": 2535, + "CTGCAG": 2536, + "CTGCTA": 2537, + "CTGCTT": 2538, + "CTGCTC": 2539, + "CTGCTG": 2540, + "CTGCCA": 2541, + "CTGCCT": 2542, + "CTGCCC": 2543, + "CTGCCG": 2544, + "CTGCGA": 2545, + "CTGCGT": 2546, + "CTGCGC": 2547, + "CTGCGG": 2548, + "CTGGAA": 2549, + "CTGGAT": 2550, + "CTGGAC": 2551, + "CTGGAG": 2552, + "CTGGTA": 2553, + "CTGGTT": 2554, + "CTGGTC": 2555, + "CTGGTG": 2556, + "CTGGCA": 2557, + "CTGGCT": 2558, + "CTGGCC": 2559, + "CTGGCG": 2560, + "CTGGGA": 2561, + "CTGGGT": 2562, + "CTGGGC": 2563, + "CTGGGG": 2564, + "CCAAAA": 2565, + "CCAAAT": 2566, + "CCAAAC": 2567, + "CCAAAG": 2568, + "CCAATA": 2569, + "CCAATT": 2570, + "CCAATC": 2571, + "CCAATG": 2572, + "CCAACA": 2573, + "CCAACT": 2574, + "CCAACC": 2575, + "CCAACG": 2576, + "CCAAGA": 2577, + "CCAAGT": 2578, + "CCAAGC": 2579, + "CCAAGG": 2580, + "CCATAA": 2581, + "CCATAT": 2582, + "CCATAC": 2583, + "CCATAG": 2584, + "CCATTA": 2585, + "CCATTT": 2586, + "CCATTC": 2587, + "CCATTG": 2588, + "CCATCA": 2589, + "CCATCT": 2590, + "CCATCC": 2591, + "CCATCG": 2592, + "CCATGA": 2593, + "CCATGT": 2594, + "CCATGC": 2595, + "CCATGG": 2596, + "CCACAA": 2597, + "CCACAT": 2598, + "CCACAC": 2599, + "CCACAG": 2600, + "CCACTA": 2601, + "CCACTT": 2602, + "CCACTC": 2603, + "CCACTG": 2604, + "CCACCA": 2605, + "CCACCT": 2606, + "CCACCC": 2607, + "CCACCG": 2608, + "CCACGA": 2609, + "CCACGT": 2610, + "CCACGC": 2611, + "CCACGG": 2612, + "CCAGAA": 2613, + "CCAGAT": 2614, + "CCAGAC": 2615, + "CCAGAG": 2616, + "CCAGTA": 2617, + "CCAGTT": 2618, + "CCAGTC": 2619, + "CCAGTG": 2620, + "CCAGCA": 2621, + "CCAGCT": 2622, + "CCAGCC": 2623, + "CCAGCG": 2624, + "CCAGGA": 2625, + "CCAGGT": 2626, + "CCAGGC": 2627, + "CCAGGG": 2628, + "CCTAAA": 2629, + "CCTAAT": 2630, + "CCTAAC": 2631, + "CCTAAG": 2632, + "CCTATA": 2633, + "CCTATT": 2634, + "CCTATC": 2635, + "CCTATG": 2636, + "CCTACA": 2637, + "CCTACT": 2638, + "CCTACC": 2639, + "CCTACG": 2640, + "CCTAGA": 2641, + "CCTAGT": 2642, + "CCTAGC": 2643, + "CCTAGG": 2644, + "CCTTAA": 2645, + "CCTTAT": 2646, + "CCTTAC": 2647, + "CCTTAG": 2648, + "CCTTTA": 2649, + "CCTTTT": 2650, + "CCTTTC": 2651, + "CCTTTG": 2652, + "CCTTCA": 2653, + "CCTTCT": 2654, + "CCTTCC": 2655, + "CCTTCG": 2656, + "CCTTGA": 2657, + "CCTTGT": 2658, + "CCTTGC": 2659, + "CCTTGG": 2660, + "CCTCAA": 2661, + "CCTCAT": 2662, + "CCTCAC": 2663, + "CCTCAG": 2664, + "CCTCTA": 2665, + "CCTCTT": 2666, + "CCTCTC": 2667, + "CCTCTG": 2668, + "CCTCCA": 2669, + "CCTCCT": 2670, + "CCTCCC": 2671, + "CCTCCG": 2672, + "CCTCGA": 2673, + "CCTCGT": 2674, + "CCTCGC": 2675, + "CCTCGG": 2676, + "CCTGAA": 2677, + "CCTGAT": 2678, + "CCTGAC": 2679, + "CCTGAG": 2680, + "CCTGTA": 2681, + "CCTGTT": 2682, + "CCTGTC": 2683, + "CCTGTG": 2684, + "CCTGCA": 2685, + "CCTGCT": 2686, + "CCTGCC": 2687, + "CCTGCG": 2688, + "CCTGGA": 2689, + "CCTGGT": 2690, + "CCTGGC": 2691, + "CCTGGG": 2692, + "CCCAAA": 2693, + "CCCAAT": 2694, + "CCCAAC": 2695, + "CCCAAG": 2696, + "CCCATA": 2697, + "CCCATT": 2698, + "CCCATC": 2699, + "CCCATG": 2700, + "CCCACA": 2701, + "CCCACT": 2702, + "CCCACC": 2703, + "CCCACG": 2704, + "CCCAGA": 2705, + "CCCAGT": 2706, + "CCCAGC": 2707, + "CCCAGG": 2708, + "CCCTAA": 2709, + "CCCTAT": 2710, + "CCCTAC": 2711, + "CCCTAG": 2712, + "CCCTTA": 2713, + "CCCTTT": 2714, + "CCCTTC": 2715, + "CCCTTG": 2716, + "CCCTCA": 2717, + "CCCTCT": 2718, + "CCCTCC": 2719, + "CCCTCG": 2720, + "CCCTGA": 2721, + "CCCTGT": 2722, + "CCCTGC": 2723, + "CCCTGG": 2724, + "CCCCAA": 2725, + "CCCCAT": 2726, + "CCCCAC": 2727, + "CCCCAG": 2728, + "CCCCTA": 2729, + "CCCCTT": 2730, + "CCCCTC": 2731, + "CCCCTG": 2732, + "CCCCCA": 2733, + "CCCCCT": 2734, + "CCCCCC": 2735, + "CCCCCG": 2736, + "CCCCGA": 2737, + "CCCCGT": 2738, + "CCCCGC": 2739, + "CCCCGG": 2740, + "CCCGAA": 2741, + "CCCGAT": 2742, + "CCCGAC": 2743, + "CCCGAG": 2744, + "CCCGTA": 2745, + "CCCGTT": 2746, + "CCCGTC": 2747, + "CCCGTG": 2748, + "CCCGCA": 2749, + "CCCGCT": 2750, + "CCCGCC": 2751, + "CCCGCG": 2752, + "CCCGGA": 2753, + "CCCGGT": 2754, + "CCCGGC": 2755, + "CCCGGG": 2756, + "CCGAAA": 2757, + "CCGAAT": 2758, + "CCGAAC": 2759, + "CCGAAG": 2760, + "CCGATA": 2761, + "CCGATT": 2762, + "CCGATC": 2763, + "CCGATG": 2764, + "CCGACA": 2765, + "CCGACT": 2766, + "CCGACC": 2767, + "CCGACG": 2768, + "CCGAGA": 2769, + "CCGAGT": 2770, + "CCGAGC": 2771, + "CCGAGG": 2772, + "CCGTAA": 2773, + "CCGTAT": 2774, + "CCGTAC": 2775, + "CCGTAG": 2776, + "CCGTTA": 2777, + "CCGTTT": 2778, + "CCGTTC": 2779, + "CCGTTG": 2780, + "CCGTCA": 2781, + "CCGTCT": 2782, + "CCGTCC": 2783, + "CCGTCG": 2784, + "CCGTGA": 2785, + "CCGTGT": 2786, + "CCGTGC": 2787, + "CCGTGG": 2788, + "CCGCAA": 2789, + "CCGCAT": 2790, + "CCGCAC": 2791, + "CCGCAG": 2792, + "CCGCTA": 2793, + "CCGCTT": 2794, + "CCGCTC": 2795, + "CCGCTG": 2796, + "CCGCCA": 2797, + "CCGCCT": 2798, + "CCGCCC": 2799, + "CCGCCG": 2800, + "CCGCGA": 2801, + "CCGCGT": 2802, + "CCGCGC": 2803, + "CCGCGG": 2804, + "CCGGAA": 2805, + "CCGGAT": 2806, + "CCGGAC": 2807, + "CCGGAG": 2808, + "CCGGTA": 2809, + "CCGGTT": 2810, + "CCGGTC": 2811, + "CCGGTG": 2812, + "CCGGCA": 2813, + "CCGGCT": 2814, + "CCGGCC": 2815, + "CCGGCG": 2816, + "CCGGGA": 2817, + "CCGGGT": 2818, + "CCGGGC": 2819, + "CCGGGG": 2820, + "CGAAAA": 2821, + "CGAAAT": 2822, + "CGAAAC": 2823, + "CGAAAG": 2824, + "CGAATA": 2825, + "CGAATT": 2826, + "CGAATC": 2827, + "CGAATG": 2828, + "CGAACA": 2829, + "CGAACT": 2830, + "CGAACC": 2831, + "CGAACG": 2832, + "CGAAGA": 2833, + "CGAAGT": 2834, + "CGAAGC": 2835, + "CGAAGG": 2836, + "CGATAA": 2837, + "CGATAT": 2838, + "CGATAC": 2839, + "CGATAG": 2840, + "CGATTA": 2841, + "CGATTT": 2842, + "CGATTC": 2843, + "CGATTG": 2844, + "CGATCA": 2845, + "CGATCT": 2846, + "CGATCC": 2847, + "CGATCG": 2848, + "CGATGA": 2849, + "CGATGT": 2850, + "CGATGC": 2851, + "CGATGG": 2852, + "CGACAA": 2853, + "CGACAT": 2854, + "CGACAC": 2855, + "CGACAG": 2856, + "CGACTA": 2857, + "CGACTT": 2858, + "CGACTC": 2859, + "CGACTG": 2860, + "CGACCA": 2861, + "CGACCT": 2862, + "CGACCC": 2863, + "CGACCG": 2864, + "CGACGA": 2865, + "CGACGT": 2866, + "CGACGC": 2867, + "CGACGG": 2868, + "CGAGAA": 2869, + "CGAGAT": 2870, + "CGAGAC": 2871, + "CGAGAG": 2872, + "CGAGTA": 2873, + "CGAGTT": 2874, + "CGAGTC": 2875, + "CGAGTG": 2876, + "CGAGCA": 2877, + "CGAGCT": 2878, + "CGAGCC": 2879, + "CGAGCG": 2880, + "CGAGGA": 2881, + "CGAGGT": 2882, + "CGAGGC": 2883, + "CGAGGG": 2884, + "CGTAAA": 2885, + "CGTAAT": 2886, + "CGTAAC": 2887, + "CGTAAG": 2888, + "CGTATA": 2889, + "CGTATT": 2890, + "CGTATC": 2891, + "CGTATG": 2892, + "CGTACA": 2893, + "CGTACT": 2894, + "CGTACC": 2895, + "CGTACG": 2896, + "CGTAGA": 2897, + "CGTAGT": 2898, + "CGTAGC": 2899, + "CGTAGG": 2900, + "CGTTAA": 2901, + "CGTTAT": 2902, + "CGTTAC": 2903, + "CGTTAG": 2904, + "CGTTTA": 2905, + "CGTTTT": 2906, + "CGTTTC": 2907, + "CGTTTG": 2908, + "CGTTCA": 2909, + "CGTTCT": 2910, + "CGTTCC": 2911, + "CGTTCG": 2912, + "CGTTGA": 2913, + "CGTTGT": 2914, + "CGTTGC": 2915, + "CGTTGG": 2916, + "CGTCAA": 2917, + "CGTCAT": 2918, + "CGTCAC": 2919, + "CGTCAG": 2920, + "CGTCTA": 2921, + "CGTCTT": 2922, + "CGTCTC": 2923, + "CGTCTG": 2924, + "CGTCCA": 2925, + "CGTCCT": 2926, + "CGTCCC": 2927, + "CGTCCG": 2928, + "CGTCGA": 2929, + "CGTCGT": 2930, + "CGTCGC": 2931, + "CGTCGG": 2932, + "CGTGAA": 2933, + "CGTGAT": 2934, + "CGTGAC": 2935, + "CGTGAG": 2936, + "CGTGTA": 2937, + "CGTGTT": 2938, + "CGTGTC": 2939, + "CGTGTG": 2940, + "CGTGCA": 2941, + "CGTGCT": 2942, + "CGTGCC": 2943, + "CGTGCG": 2944, + "CGTGGA": 2945, + "CGTGGT": 2946, + "CGTGGC": 2947, + "CGTGGG": 2948, + "CGCAAA": 2949, + "CGCAAT": 2950, + "CGCAAC": 2951, + "CGCAAG": 2952, + "CGCATA": 2953, + "CGCATT": 2954, + "CGCATC": 2955, + "CGCATG": 2956, + "CGCACA": 2957, + "CGCACT": 2958, + "CGCACC": 2959, + "CGCACG": 2960, + "CGCAGA": 2961, + "CGCAGT": 2962, + "CGCAGC": 2963, + "CGCAGG": 2964, + "CGCTAA": 2965, + "CGCTAT": 2966, + "CGCTAC": 2967, + "CGCTAG": 2968, + "CGCTTA": 2969, + "CGCTTT": 2970, + "CGCTTC": 2971, + "CGCTTG": 2972, + "CGCTCA": 2973, + "CGCTCT": 2974, + "CGCTCC": 2975, + "CGCTCG": 2976, + "CGCTGA": 2977, + "CGCTGT": 2978, + "CGCTGC": 2979, + "CGCTGG": 2980, + "CGCCAA": 2981, + "CGCCAT": 2982, + "CGCCAC": 2983, + "CGCCAG": 2984, + "CGCCTA": 2985, + "CGCCTT": 2986, + "CGCCTC": 2987, + "CGCCTG": 2988, + "CGCCCA": 2989, + "CGCCCT": 2990, + "CGCCCC": 2991, + "CGCCCG": 2992, + "CGCCGA": 2993, + "CGCCGT": 2994, + "CGCCGC": 2995, + "CGCCGG": 2996, + "CGCGAA": 2997, + "CGCGAT": 2998, + "CGCGAC": 2999, + "CGCGAG": 3000, + "CGCGTA": 3001, + "CGCGTT": 3002, + "CGCGTC": 3003, + "CGCGTG": 3004, + "CGCGCA": 3005, + "CGCGCT": 3006, + "CGCGCC": 3007, + "CGCGCG": 3008, + "CGCGGA": 3009, + "CGCGGT": 3010, + "CGCGGC": 3011, + "CGCGGG": 3012, + "CGGAAA": 3013, + "CGGAAT": 3014, + "CGGAAC": 3015, + "CGGAAG": 3016, + "CGGATA": 3017, + "CGGATT": 3018, + "CGGATC": 3019, + "CGGATG": 3020, + "CGGACA": 3021, + "CGGACT": 3022, + "CGGACC": 3023, + "CGGACG": 3024, + "CGGAGA": 3025, + "CGGAGT": 3026, + "CGGAGC": 3027, + "CGGAGG": 3028, + "CGGTAA": 3029, + "CGGTAT": 3030, + "CGGTAC": 3031, + "CGGTAG": 3032, + "CGGTTA": 3033, + "CGGTTT": 3034, + "CGGTTC": 3035, + "CGGTTG": 3036, + "CGGTCA": 3037, + "CGGTCT": 3038, + "CGGTCC": 3039, + "CGGTCG": 3040, + "CGGTGA": 3041, + "CGGTGT": 3042, + "CGGTGC": 3043, + "CGGTGG": 3044, + "CGGCAA": 3045, + "CGGCAT": 3046, + "CGGCAC": 3047, + "CGGCAG": 3048, + "CGGCTA": 3049, + "CGGCTT": 3050, + "CGGCTC": 3051, + "CGGCTG": 3052, + "CGGCCA": 3053, + "CGGCCT": 3054, + "CGGCCC": 3055, + "CGGCCG": 3056, + "CGGCGA": 3057, + "CGGCGT": 3058, + "CGGCGC": 3059, + "CGGCGG": 3060, + "CGGGAA": 3061, + "CGGGAT": 3062, + "CGGGAC": 3063, + "CGGGAG": 3064, + "CGGGTA": 3065, + "CGGGTT": 3066, + "CGGGTC": 3067, + "CGGGTG": 3068, + "CGGGCA": 3069, + "CGGGCT": 3070, + "CGGGCC": 3071, + "CGGGCG": 3072, + "CGGGGA": 3073, + "CGGGGT": 3074, + "CGGGGC": 3075, + "CGGGGG": 3076, + "GAAAAA": 3077, + "GAAAAT": 3078, + "GAAAAC": 3079, + "GAAAAG": 3080, + "GAAATA": 3081, + "GAAATT": 3082, + "GAAATC": 3083, + "GAAATG": 3084, + "GAAACA": 3085, + "GAAACT": 3086, + "GAAACC": 3087, + "GAAACG": 3088, + "GAAAGA": 3089, + "GAAAGT": 3090, + "GAAAGC": 3091, + "GAAAGG": 3092, + "GAATAA": 3093, + "GAATAT": 3094, + "GAATAC": 3095, + "GAATAG": 3096, + "GAATTA": 3097, + "GAATTT": 3098, + "GAATTC": 3099, + "GAATTG": 3100, + "GAATCA": 3101, + "GAATCT": 3102, + "GAATCC": 3103, + "GAATCG": 3104, + "GAATGA": 3105, + "GAATGT": 3106, + "GAATGC": 3107, + "GAATGG": 3108, + "GAACAA": 3109, + "GAACAT": 3110, + "GAACAC": 3111, + "GAACAG": 3112, + "GAACTA": 3113, + "GAACTT": 3114, + "GAACTC": 3115, + "GAACTG": 3116, + "GAACCA": 3117, + "GAACCT": 3118, + "GAACCC": 3119, + "GAACCG": 3120, + "GAACGA": 3121, + "GAACGT": 3122, + "GAACGC": 3123, + "GAACGG": 3124, + "GAAGAA": 3125, + "GAAGAT": 3126, + "GAAGAC": 3127, + "GAAGAG": 3128, + "GAAGTA": 3129, + "GAAGTT": 3130, + "GAAGTC": 3131, + "GAAGTG": 3132, + "GAAGCA": 3133, + "GAAGCT": 3134, + "GAAGCC": 3135, + "GAAGCG": 3136, + "GAAGGA": 3137, + "GAAGGT": 3138, + "GAAGGC": 3139, + "GAAGGG": 3140, + "GATAAA": 3141, + "GATAAT": 3142, + "GATAAC": 3143, + "GATAAG": 3144, + "GATATA": 3145, + "GATATT": 3146, + "GATATC": 3147, + "GATATG": 3148, + "GATACA": 3149, + "GATACT": 3150, + "GATACC": 3151, + "GATACG": 3152, + "GATAGA": 3153, + "GATAGT": 3154, + "GATAGC": 3155, + "GATAGG": 3156, + "GATTAA": 3157, + "GATTAT": 3158, + "GATTAC": 3159, + "GATTAG": 3160, + "GATTTA": 3161, + "GATTTT": 3162, + "GATTTC": 3163, + "GATTTG": 3164, + "GATTCA": 3165, + "GATTCT": 3166, + "GATTCC": 3167, + "GATTCG": 3168, + "GATTGA": 3169, + "GATTGT": 3170, + "GATTGC": 3171, + "GATTGG": 3172, + "GATCAA": 3173, + "GATCAT": 3174, + "GATCAC": 3175, + "GATCAG": 3176, + "GATCTA": 3177, + "GATCTT": 3178, + "GATCTC": 3179, + "GATCTG": 3180, + "GATCCA": 3181, + "GATCCT": 3182, + "GATCCC": 3183, + "GATCCG": 3184, + "GATCGA": 3185, + "GATCGT": 3186, + "GATCGC": 3187, + "GATCGG": 3188, + "GATGAA": 3189, + "GATGAT": 3190, + "GATGAC": 3191, + "GATGAG": 3192, + "GATGTA": 3193, + "GATGTT": 3194, + "GATGTC": 3195, + "GATGTG": 3196, + "GATGCA": 3197, + "GATGCT": 3198, + "GATGCC": 3199, + "GATGCG": 3200, + "GATGGA": 3201, + "GATGGT": 3202, + "GATGGC": 3203, + "GATGGG": 3204, + "GACAAA": 3205, + "GACAAT": 3206, + "GACAAC": 3207, + "GACAAG": 3208, + "GACATA": 3209, + "GACATT": 3210, + "GACATC": 3211, + "GACATG": 3212, + "GACACA": 3213, + "GACACT": 3214, + "GACACC": 3215, + "GACACG": 3216, + "GACAGA": 3217, + "GACAGT": 3218, + "GACAGC": 3219, + "GACAGG": 3220, + "GACTAA": 3221, + "GACTAT": 3222, + "GACTAC": 3223, + "GACTAG": 3224, + "GACTTA": 3225, + "GACTTT": 3226, + "GACTTC": 3227, + "GACTTG": 3228, + "GACTCA": 3229, + "GACTCT": 3230, + "GACTCC": 3231, + "GACTCG": 3232, + "GACTGA": 3233, + "GACTGT": 3234, + "GACTGC": 3235, + "GACTGG": 3236, + "GACCAA": 3237, + "GACCAT": 3238, + "GACCAC": 3239, + "GACCAG": 3240, + "GACCTA": 3241, + "GACCTT": 3242, + "GACCTC": 3243, + "GACCTG": 3244, + "GACCCA": 3245, + "GACCCT": 3246, + "GACCCC": 3247, + "GACCCG": 3248, + "GACCGA": 3249, + "GACCGT": 3250, + "GACCGC": 3251, + "GACCGG": 3252, + "GACGAA": 3253, + "GACGAT": 3254, + "GACGAC": 3255, + "GACGAG": 3256, + "GACGTA": 3257, + "GACGTT": 3258, + "GACGTC": 3259, + "GACGTG": 3260, + "GACGCA": 3261, + "GACGCT": 3262, + "GACGCC": 3263, + "GACGCG": 3264, + "GACGGA": 3265, + "GACGGT": 3266, + "GACGGC": 3267, + "GACGGG": 3268, + "GAGAAA": 3269, + "GAGAAT": 3270, + "GAGAAC": 3271, + "GAGAAG": 3272, + "GAGATA": 3273, + "GAGATT": 3274, + "GAGATC": 3275, + "GAGATG": 3276, + "GAGACA": 3277, + "GAGACT": 3278, + "GAGACC": 3279, + "GAGACG": 3280, + "GAGAGA": 3281, + "GAGAGT": 3282, + "GAGAGC": 3283, + "GAGAGG": 3284, + "GAGTAA": 3285, + "GAGTAT": 3286, + "GAGTAC": 3287, + "GAGTAG": 3288, + "GAGTTA": 3289, + "GAGTTT": 3290, + "GAGTTC": 3291, + "GAGTTG": 3292, + "GAGTCA": 3293, + "GAGTCT": 3294, + "GAGTCC": 3295, + "GAGTCG": 3296, + "GAGTGA": 3297, + "GAGTGT": 3298, + "GAGTGC": 3299, + "GAGTGG": 3300, + "GAGCAA": 3301, + "GAGCAT": 3302, + "GAGCAC": 3303, + "GAGCAG": 3304, + "GAGCTA": 3305, + "GAGCTT": 3306, + "GAGCTC": 3307, + "GAGCTG": 3308, + "GAGCCA": 3309, + "GAGCCT": 3310, + "GAGCCC": 3311, + "GAGCCG": 3312, + "GAGCGA": 3313, + "GAGCGT": 3314, + "GAGCGC": 3315, + "GAGCGG": 3316, + "GAGGAA": 3317, + "GAGGAT": 3318, + "GAGGAC": 3319, + "GAGGAG": 3320, + "GAGGTA": 3321, + "GAGGTT": 3322, + "GAGGTC": 3323, + "GAGGTG": 3324, + "GAGGCA": 3325, + "GAGGCT": 3326, + "GAGGCC": 3327, + "GAGGCG": 3328, + "GAGGGA": 3329, + "GAGGGT": 3330, + "GAGGGC": 3331, + "GAGGGG": 3332, + "GTAAAA": 3333, + "GTAAAT": 3334, + "GTAAAC": 3335, + "GTAAAG": 3336, + "GTAATA": 3337, + "GTAATT": 3338, + "GTAATC": 3339, + "GTAATG": 3340, + "GTAACA": 3341, + "GTAACT": 3342, + "GTAACC": 3343, + "GTAACG": 3344, + "GTAAGA": 3345, + "GTAAGT": 3346, + "GTAAGC": 3347, + "GTAAGG": 3348, + "GTATAA": 3349, + "GTATAT": 3350, + "GTATAC": 3351, + "GTATAG": 3352, + "GTATTA": 3353, + "GTATTT": 3354, + "GTATTC": 3355, + "GTATTG": 3356, + "GTATCA": 3357, + "GTATCT": 3358, + "GTATCC": 3359, + "GTATCG": 3360, + "GTATGA": 3361, + "GTATGT": 3362, + "GTATGC": 3363, + "GTATGG": 3364, + "GTACAA": 3365, + "GTACAT": 3366, + "GTACAC": 3367, + "GTACAG": 3368, + "GTACTA": 3369, + "GTACTT": 3370, + "GTACTC": 3371, + "GTACTG": 3372, + "GTACCA": 3373, + "GTACCT": 3374, + "GTACCC": 3375, + "GTACCG": 3376, + "GTACGA": 3377, + "GTACGT": 3378, + "GTACGC": 3379, + "GTACGG": 3380, + "GTAGAA": 3381, + "GTAGAT": 3382, + "GTAGAC": 3383, + "GTAGAG": 3384, + "GTAGTA": 3385, + "GTAGTT": 3386, + "GTAGTC": 3387, + "GTAGTG": 3388, + "GTAGCA": 3389, + "GTAGCT": 3390, + "GTAGCC": 3391, + "GTAGCG": 3392, + "GTAGGA": 3393, + "GTAGGT": 3394, + "GTAGGC": 3395, + "GTAGGG": 3396, + "GTTAAA": 3397, + "GTTAAT": 3398, + "GTTAAC": 3399, + "GTTAAG": 3400, + "GTTATA": 3401, + "GTTATT": 3402, + "GTTATC": 3403, + "GTTATG": 3404, + "GTTACA": 3405, + "GTTACT": 3406, + "GTTACC": 3407, + "GTTACG": 3408, + "GTTAGA": 3409, + "GTTAGT": 3410, + "GTTAGC": 3411, + "GTTAGG": 3412, + "GTTTAA": 3413, + "GTTTAT": 3414, + "GTTTAC": 3415, + "GTTTAG": 3416, + "GTTTTA": 3417, + "GTTTTT": 3418, + "GTTTTC": 3419, + "GTTTTG": 3420, + "GTTTCA": 3421, + "GTTTCT": 3422, + "GTTTCC": 3423, + "GTTTCG": 3424, + "GTTTGA": 3425, + "GTTTGT": 3426, + "GTTTGC": 3427, + "GTTTGG": 3428, + "GTTCAA": 3429, + "GTTCAT": 3430, + "GTTCAC": 3431, + "GTTCAG": 3432, + "GTTCTA": 3433, + "GTTCTT": 3434, + "GTTCTC": 3435, + "GTTCTG": 3436, + "GTTCCA": 3437, + "GTTCCT": 3438, + "GTTCCC": 3439, + "GTTCCG": 3440, + "GTTCGA": 3441, + "GTTCGT": 3442, + "GTTCGC": 3443, + "GTTCGG": 3444, + "GTTGAA": 3445, + "GTTGAT": 3446, + "GTTGAC": 3447, + "GTTGAG": 3448, + "GTTGTA": 3449, + "GTTGTT": 3450, + "GTTGTC": 3451, + "GTTGTG": 3452, + "GTTGCA": 3453, + "GTTGCT": 3454, + "GTTGCC": 3455, + "GTTGCG": 3456, + "GTTGGA": 3457, + "GTTGGT": 3458, + "GTTGGC": 3459, + "GTTGGG": 3460, + "GTCAAA": 3461, + "GTCAAT": 3462, + "GTCAAC": 3463, + "GTCAAG": 3464, + "GTCATA": 3465, + "GTCATT": 3466, + "GTCATC": 3467, + "GTCATG": 3468, + "GTCACA": 3469, + "GTCACT": 3470, + "GTCACC": 3471, + "GTCACG": 3472, + "GTCAGA": 3473, + "GTCAGT": 3474, + "GTCAGC": 3475, + "GTCAGG": 3476, + "GTCTAA": 3477, + "GTCTAT": 3478, + "GTCTAC": 3479, + "GTCTAG": 3480, + "GTCTTA": 3481, + "GTCTTT": 3482, + "GTCTTC": 3483, + "GTCTTG": 3484, + "GTCTCA": 3485, + "GTCTCT": 3486, + "GTCTCC": 3487, + "GTCTCG": 3488, + "GTCTGA": 3489, + "GTCTGT": 3490, + "GTCTGC": 3491, + "GTCTGG": 3492, + "GTCCAA": 3493, + "GTCCAT": 3494, + "GTCCAC": 3495, + "GTCCAG": 3496, + "GTCCTA": 3497, + "GTCCTT": 3498, + "GTCCTC": 3499, + "GTCCTG": 3500, + "GTCCCA": 3501, + "GTCCCT": 3502, + "GTCCCC": 3503, + "GTCCCG": 3504, + "GTCCGA": 3505, + "GTCCGT": 3506, + "GTCCGC": 3507, + "GTCCGG": 3508, + "GTCGAA": 3509, + "GTCGAT": 3510, + "GTCGAC": 3511, + "GTCGAG": 3512, + "GTCGTA": 3513, + "GTCGTT": 3514, + "GTCGTC": 3515, + "GTCGTG": 3516, + "GTCGCA": 3517, + "GTCGCT": 3518, + "GTCGCC": 3519, + "GTCGCG": 3520, + "GTCGGA": 3521, + "GTCGGT": 3522, + "GTCGGC": 3523, + "GTCGGG": 3524, + "GTGAAA": 3525, + "GTGAAT": 3526, + "GTGAAC": 3527, + "GTGAAG": 3528, + "GTGATA": 3529, + "GTGATT": 3530, + "GTGATC": 3531, + "GTGATG": 3532, + "GTGACA": 3533, + "GTGACT": 3534, + "GTGACC": 3535, + "GTGACG": 3536, + "GTGAGA": 3537, + "GTGAGT": 3538, + "GTGAGC": 3539, + "GTGAGG": 3540, + "GTGTAA": 3541, + "GTGTAT": 3542, + "GTGTAC": 3543, + "GTGTAG": 3544, + "GTGTTA": 3545, + "GTGTTT": 3546, + "GTGTTC": 3547, + "GTGTTG": 3548, + "GTGTCA": 3549, + "GTGTCT": 3550, + "GTGTCC": 3551, + "GTGTCG": 3552, + "GTGTGA": 3553, + "GTGTGT": 3554, + "GTGTGC": 3555, + "GTGTGG": 3556, + "GTGCAA": 3557, + "GTGCAT": 3558, + "GTGCAC": 3559, + "GTGCAG": 3560, + "GTGCTA": 3561, + "GTGCTT": 3562, + "GTGCTC": 3563, + "GTGCTG": 3564, + "GTGCCA": 3565, + "GTGCCT": 3566, + "GTGCCC": 3567, + "GTGCCG": 3568, + "GTGCGA": 3569, + "GTGCGT": 3570, + "GTGCGC": 3571, + "GTGCGG": 3572, + "GTGGAA": 3573, + "GTGGAT": 3574, + "GTGGAC": 3575, + "GTGGAG": 3576, + "GTGGTA": 3577, + "GTGGTT": 3578, + "GTGGTC": 3579, + "GTGGTG": 3580, + "GTGGCA": 3581, + "GTGGCT": 3582, + "GTGGCC": 3583, + "GTGGCG": 3584, + "GTGGGA": 3585, + "GTGGGT": 3586, + "GTGGGC": 3587, + "GTGGGG": 3588, + "GCAAAA": 3589, + "GCAAAT": 3590, + "GCAAAC": 3591, + "GCAAAG": 3592, + "GCAATA": 3593, + "GCAATT": 3594, + "GCAATC": 3595, + "GCAATG": 3596, + "GCAACA": 3597, + "GCAACT": 3598, + "GCAACC": 3599, + "GCAACG": 3600, + "GCAAGA": 3601, + "GCAAGT": 3602, + "GCAAGC": 3603, + "GCAAGG": 3604, + "GCATAA": 3605, + "GCATAT": 3606, + "GCATAC": 3607, + "GCATAG": 3608, + "GCATTA": 3609, + "GCATTT": 3610, + "GCATTC": 3611, + "GCATTG": 3612, + "GCATCA": 3613, + "GCATCT": 3614, + "GCATCC": 3615, + "GCATCG": 3616, + "GCATGA": 3617, + "GCATGT": 3618, + "GCATGC": 3619, + "GCATGG": 3620, + "GCACAA": 3621, + "GCACAT": 3622, + "GCACAC": 3623, + "GCACAG": 3624, + "GCACTA": 3625, + "GCACTT": 3626, + "GCACTC": 3627, + "GCACTG": 3628, + "GCACCA": 3629, + "GCACCT": 3630, + "GCACCC": 3631, + "GCACCG": 3632, + "GCACGA": 3633, + "GCACGT": 3634, + "GCACGC": 3635, + "GCACGG": 3636, + "GCAGAA": 3637, + "GCAGAT": 3638, + "GCAGAC": 3639, + "GCAGAG": 3640, + "GCAGTA": 3641, + "GCAGTT": 3642, + "GCAGTC": 3643, + "GCAGTG": 3644, + "GCAGCA": 3645, + "GCAGCT": 3646, + "GCAGCC": 3647, + "GCAGCG": 3648, + "GCAGGA": 3649, + "GCAGGT": 3650, + "GCAGGC": 3651, + "GCAGGG": 3652, + "GCTAAA": 3653, + "GCTAAT": 3654, + "GCTAAC": 3655, + "GCTAAG": 3656, + "GCTATA": 3657, + "GCTATT": 3658, + "GCTATC": 3659, + "GCTATG": 3660, + "GCTACA": 3661, + "GCTACT": 3662, + "GCTACC": 3663, + "GCTACG": 3664, + "GCTAGA": 3665, + "GCTAGT": 3666, + "GCTAGC": 3667, + "GCTAGG": 3668, + "GCTTAA": 3669, + "GCTTAT": 3670, + "GCTTAC": 3671, + "GCTTAG": 3672, + "GCTTTA": 3673, + "GCTTTT": 3674, + "GCTTTC": 3675, + "GCTTTG": 3676, + "GCTTCA": 3677, + "GCTTCT": 3678, + "GCTTCC": 3679, + "GCTTCG": 3680, + "GCTTGA": 3681, + "GCTTGT": 3682, + "GCTTGC": 3683, + "GCTTGG": 3684, + "GCTCAA": 3685, + "GCTCAT": 3686, + "GCTCAC": 3687, + "GCTCAG": 3688, + "GCTCTA": 3689, + "GCTCTT": 3690, + "GCTCTC": 3691, + "GCTCTG": 3692, + "GCTCCA": 3693, + "GCTCCT": 3694, + "GCTCCC": 3695, + "GCTCCG": 3696, + "GCTCGA": 3697, + "GCTCGT": 3698, + "GCTCGC": 3699, + "GCTCGG": 3700, + "GCTGAA": 3701, + "GCTGAT": 3702, + "GCTGAC": 3703, + "GCTGAG": 3704, + "GCTGTA": 3705, + "GCTGTT": 3706, + "GCTGTC": 3707, + "GCTGTG": 3708, + "GCTGCA": 3709, + "GCTGCT": 3710, + "GCTGCC": 3711, + "GCTGCG": 3712, + "GCTGGA": 3713, + "GCTGGT": 3714, + "GCTGGC": 3715, + "GCTGGG": 3716, + "GCCAAA": 3717, + "GCCAAT": 3718, + "GCCAAC": 3719, + "GCCAAG": 3720, + "GCCATA": 3721, + "GCCATT": 3722, + "GCCATC": 3723, + "GCCATG": 3724, + "GCCACA": 3725, + "GCCACT": 3726, + "GCCACC": 3727, + "GCCACG": 3728, + "GCCAGA": 3729, + "GCCAGT": 3730, + "GCCAGC": 3731, + "GCCAGG": 3732, + "GCCTAA": 3733, + "GCCTAT": 3734, + "GCCTAC": 3735, + "GCCTAG": 3736, + "GCCTTA": 3737, + "GCCTTT": 3738, + "GCCTTC": 3739, + "GCCTTG": 3740, + "GCCTCA": 3741, + "GCCTCT": 3742, + "GCCTCC": 3743, + "GCCTCG": 3744, + "GCCTGA": 3745, + "GCCTGT": 3746, + "GCCTGC": 3747, + "GCCTGG": 3748, + "GCCCAA": 3749, + "GCCCAT": 3750, + "GCCCAC": 3751, + "GCCCAG": 3752, + "GCCCTA": 3753, + "GCCCTT": 3754, + "GCCCTC": 3755, + "GCCCTG": 3756, + "GCCCCA": 3757, + "GCCCCT": 3758, + "GCCCCC": 3759, + "GCCCCG": 3760, + "GCCCGA": 3761, + "GCCCGT": 3762, + "GCCCGC": 3763, + "GCCCGG": 3764, + "GCCGAA": 3765, + "GCCGAT": 3766, + "GCCGAC": 3767, + "GCCGAG": 3768, + "GCCGTA": 3769, + "GCCGTT": 3770, + "GCCGTC": 3771, + "GCCGTG": 3772, + "GCCGCA": 3773, + "GCCGCT": 3774, + "GCCGCC": 3775, + "GCCGCG": 3776, + "GCCGGA": 3777, + "GCCGGT": 3778, + "GCCGGC": 3779, + "GCCGGG": 3780, + "GCGAAA": 3781, + "GCGAAT": 3782, + "GCGAAC": 3783, + "GCGAAG": 3784, + "GCGATA": 3785, + "GCGATT": 3786, + "GCGATC": 3787, + "GCGATG": 3788, + "GCGACA": 3789, + "GCGACT": 3790, + "GCGACC": 3791, + "GCGACG": 3792, + "GCGAGA": 3793, + "GCGAGT": 3794, + "GCGAGC": 3795, + "GCGAGG": 3796, + "GCGTAA": 3797, + "GCGTAT": 3798, + "GCGTAC": 3799, + "GCGTAG": 3800, + "GCGTTA": 3801, + "GCGTTT": 3802, + "GCGTTC": 3803, + "GCGTTG": 3804, + "GCGTCA": 3805, + "GCGTCT": 3806, + "GCGTCC": 3807, + "GCGTCG": 3808, + "GCGTGA": 3809, + "GCGTGT": 3810, + "GCGTGC": 3811, + "GCGTGG": 3812, + "GCGCAA": 3813, + "GCGCAT": 3814, + "GCGCAC": 3815, + "GCGCAG": 3816, + "GCGCTA": 3817, + "GCGCTT": 3818, + "GCGCTC": 3819, + "GCGCTG": 3820, + "GCGCCA": 3821, + "GCGCCT": 3822, + "GCGCCC": 3823, + "GCGCCG": 3824, + "GCGCGA": 3825, + "GCGCGT": 3826, + "GCGCGC": 3827, + "GCGCGG": 3828, + "GCGGAA": 3829, + "GCGGAT": 3830, + "GCGGAC": 3831, + "GCGGAG": 3832, + "GCGGTA": 3833, + "GCGGTT": 3834, + "GCGGTC": 3835, + "GCGGTG": 3836, + "GCGGCA": 3837, + "GCGGCT": 3838, + "GCGGCC": 3839, + "GCGGCG": 3840, + "GCGGGA": 3841, + "GCGGGT": 3842, + "GCGGGC": 3843, + "GCGGGG": 3844, + "GGAAAA": 3845, + "GGAAAT": 3846, + "GGAAAC": 3847, + "GGAAAG": 3848, + "GGAATA": 3849, + "GGAATT": 3850, + "GGAATC": 3851, + "GGAATG": 3852, + "GGAACA": 3853, + "GGAACT": 3854, + "GGAACC": 3855, + "GGAACG": 3856, + "GGAAGA": 3857, + "GGAAGT": 3858, + "GGAAGC": 3859, + "GGAAGG": 3860, + "GGATAA": 3861, + "GGATAT": 3862, + "GGATAC": 3863, + "GGATAG": 3864, + "GGATTA": 3865, + "GGATTT": 3866, + "GGATTC": 3867, + "GGATTG": 3868, + "GGATCA": 3869, + "GGATCT": 3870, + "GGATCC": 3871, + "GGATCG": 3872, + "GGATGA": 3873, + "GGATGT": 3874, + "GGATGC": 3875, + "GGATGG": 3876, + "GGACAA": 3877, + "GGACAT": 3878, + "GGACAC": 3879, + "GGACAG": 3880, + "GGACTA": 3881, + "GGACTT": 3882, + "GGACTC": 3883, + "GGACTG": 3884, + "GGACCA": 3885, + "GGACCT": 3886, + "GGACCC": 3887, + "GGACCG": 3888, + "GGACGA": 3889, + "GGACGT": 3890, + "GGACGC": 3891, + "GGACGG": 3892, + "GGAGAA": 3893, + "GGAGAT": 3894, + "GGAGAC": 3895, + "GGAGAG": 3896, + "GGAGTA": 3897, + "GGAGTT": 3898, + "GGAGTC": 3899, + "GGAGTG": 3900, + "GGAGCA": 3901, + "GGAGCT": 3902, + "GGAGCC": 3903, + "GGAGCG": 3904, + "GGAGGA": 3905, + "GGAGGT": 3906, + "GGAGGC": 3907, + "GGAGGG": 3908, + "GGTAAA": 3909, + "GGTAAT": 3910, + "GGTAAC": 3911, + "GGTAAG": 3912, + "GGTATA": 3913, + "GGTATT": 3914, + "GGTATC": 3915, + "GGTATG": 3916, + "GGTACA": 3917, + "GGTACT": 3918, + "GGTACC": 3919, + "GGTACG": 3920, + "GGTAGA": 3921, + "GGTAGT": 3922, + "GGTAGC": 3923, + "GGTAGG": 3924, + "GGTTAA": 3925, + "GGTTAT": 3926, + "GGTTAC": 3927, + "GGTTAG": 3928, + "GGTTTA": 3929, + "GGTTTT": 3930, + "GGTTTC": 3931, + "GGTTTG": 3932, + "GGTTCA": 3933, + "GGTTCT": 3934, + "GGTTCC": 3935, + "GGTTCG": 3936, + "GGTTGA": 3937, + "GGTTGT": 3938, + "GGTTGC": 3939, + "GGTTGG": 3940, + "GGTCAA": 3941, + "GGTCAT": 3942, + "GGTCAC": 3943, + "GGTCAG": 3944, + "GGTCTA": 3945, + "GGTCTT": 3946, + "GGTCTC": 3947, + "GGTCTG": 3948, + "GGTCCA": 3949, + "GGTCCT": 3950, + "GGTCCC": 3951, + "GGTCCG": 3952, + "GGTCGA": 3953, + "GGTCGT": 3954, + "GGTCGC": 3955, + "GGTCGG": 3956, + "GGTGAA": 3957, + "GGTGAT": 3958, + "GGTGAC": 3959, + "GGTGAG": 3960, + "GGTGTA": 3961, + "GGTGTT": 3962, + "GGTGTC": 3963, + "GGTGTG": 3964, + "GGTGCA": 3965, + "GGTGCT": 3966, + "GGTGCC": 3967, + "GGTGCG": 3968, + "GGTGGA": 3969, + "GGTGGT": 3970, + "GGTGGC": 3971, + "GGTGGG": 3972, + "GGCAAA": 3973, + "GGCAAT": 3974, + "GGCAAC": 3975, + "GGCAAG": 3976, + "GGCATA": 3977, + "GGCATT": 3978, + "GGCATC": 3979, + "GGCATG": 3980, + "GGCACA": 3981, + "GGCACT": 3982, + "GGCACC": 3983, + "GGCACG": 3984, + "GGCAGA": 3985, + "GGCAGT": 3986, + "GGCAGC": 3987, + "GGCAGG": 3988, + "GGCTAA": 3989, + "GGCTAT": 3990, + "GGCTAC": 3991, + "GGCTAG": 3992, + "GGCTTA": 3993, + "GGCTTT": 3994, + "GGCTTC": 3995, + "GGCTTG": 3996, + "GGCTCA": 3997, + "GGCTCT": 3998, + "GGCTCC": 3999, + "GGCTCG": 4000, + "GGCTGA": 4001, + "GGCTGT": 4002, + "GGCTGC": 4003, + "GGCTGG": 4004, + "GGCCAA": 4005, + "GGCCAT": 4006, + "GGCCAC": 4007, + "GGCCAG": 4008, + "GGCCTA": 4009, + "GGCCTT": 4010, + "GGCCTC": 4011, + "GGCCTG": 4012, + "GGCCCA": 4013, + "GGCCCT": 4014, + "GGCCCC": 4015, + "GGCCCG": 4016, + "GGCCGA": 4017, + "GGCCGT": 4018, + "GGCCGC": 4019, + "GGCCGG": 4020, + "GGCGAA": 4021, + "GGCGAT": 4022, + "GGCGAC": 4023, + "GGCGAG": 4024, + "GGCGTA": 4025, + "GGCGTT": 4026, + "GGCGTC": 4027, + "GGCGTG": 4028, + "GGCGCA": 4029, + "GGCGCT": 4030, + "GGCGCC": 4031, + "GGCGCG": 4032, + "GGCGGA": 4033, + "GGCGGT": 4034, + "GGCGGC": 4035, + "GGCGGG": 4036, + "GGGAAA": 4037, + "GGGAAT": 4038, + "GGGAAC": 4039, + "GGGAAG": 4040, + "GGGATA": 4041, + "GGGATT": 4042, + "GGGATC": 4043, + "GGGATG": 4044, + "GGGACA": 4045, + "GGGACT": 4046, + "GGGACC": 4047, + "GGGACG": 4048, + "GGGAGA": 4049, + "GGGAGT": 4050, + "GGGAGC": 4051, + "GGGAGG": 4052, + "GGGTAA": 4053, + "GGGTAT": 4054, + "GGGTAC": 4055, + "GGGTAG": 4056, + "GGGTTA": 4057, + "GGGTTT": 4058, + "GGGTTC": 4059, + "GGGTTG": 4060, + "GGGTCA": 4061, + "GGGTCT": 4062, + "GGGTCC": 4063, + "GGGTCG": 4064, + "GGGTGA": 4065, + "GGGTGT": 4066, + "GGGTGC": 4067, + "GGGTGG": 4068, + "GGGCAA": 4069, + "GGGCAT": 4070, + "GGGCAC": 4071, + "GGGCAG": 4072, + "GGGCTA": 4073, + "GGGCTT": 4074, + "GGGCTC": 4075, + "GGGCTG": 4076, + "GGGCCA": 4077, + "GGGCCT": 4078, + "GGGCCC": 4079, + "GGGCCG": 4080, + "GGGCGA": 4081, + "GGGCGT": 4082, + "GGGCGC": 4083, + "GGGCGG": 4084, + "GGGGAA": 4085, + "GGGGAT": 4086, + "GGGGAC": 4087, + "GGGGAG": 4088, + "GGGGTA": 4089, + "GGGGTT": 4090, + "GGGGTC": 4091, + "GGGGTG": 4092, + "GGGGCA": 4093, + "GGGGCT": 4094, + "GGGGCC": 4095, + "GGGGCG": 4096, + "GGGGGA": 4097, + "GGGGGT": 4098, + "GGGGGC": 4099, + "GGGGGG": 4100 + } + } +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/tokenizer_config.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..604e2b1b29af1dcb3920d2828b6a90276e5e3d02 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/tokenizer_config.json @@ -0,0 +1,58 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "[UNK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "[CLS]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "[SEP]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "4": { + "content": "[MASK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "clean_up_tokenization_spaces": true, + "cls_token": "[CLS]", + "do_basic_tokenize": true, + "do_lower_case": false, + "mask_token": "[MASK]", + "max_len": 512, + "model_max_length": 512, + "never_split": null, + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "strip_accents": null, + "tokenize_chinese_chars": true, + "tokenizer_class": "BertTokenizer", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/trainer_state.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..c62ffe9d9a2b5f1bcd2e068e0f27167bd5364039 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/trainer_state.json @@ -0,0 +1,1433 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 20.0, + "eval_steps": 500, + "global_step": 5000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.1, + "grad_norm": 0.2192622423171997, + "learning_rate": 5.319148936170213e-05, + "loss": 0.6949, + "step": 25 + }, + { + "epoch": 0.2, + "grad_norm": 0.22659960389137268, + "learning_rate": 0.00010638297872340425, + "loss": 0.6926, + "step": 50 + }, + { + "epoch": 0.3, + "grad_norm": 0.33277687430381775, + "learning_rate": 0.00015957446808510637, + "loss": 0.693, + "step": 75 + }, + { + "epoch": 0.4, + "grad_norm": 0.5183415412902832, + "learning_rate": 0.0002127659574468085, + "loss": 0.6928, + "step": 100 + }, + { + "epoch": 0.5, + "grad_norm": 0.21227706968784332, + "learning_rate": 0.00026595744680851064, + "loss": 0.6938, + "step": 125 + }, + { + "epoch": 0.6, + "grad_norm": 0.1846485733985901, + "learning_rate": 0.00031914893617021275, + "loss": 0.6942, + "step": 150 + }, + { + "epoch": 0.7, + "grad_norm": 0.27858808636665344, + "learning_rate": 0.0003723404255319149, + "loss": 0.6931, + "step": 175 + }, + { + "epoch": 0.8, + "grad_norm": 0.26012367010116577, + "learning_rate": 0.0003992081821181128, + "loss": 0.6936, + "step": 200 + }, + { + "epoch": 0.9, + "grad_norm": 0.2534612715244293, + "learning_rate": 0.0003975585615308479, + "loss": 0.6939, + "step": 225 + }, + { + "epoch": 1.0, + "grad_norm": 0.42967700958251953, + "learning_rate": 0.00039590894094358297, + "loss": 0.6928, + "step": 250 + }, + { + "epoch": 1.1, + "grad_norm": 0.2613862454891205, + "learning_rate": 0.0003942593203563181, + "loss": 0.6918, + "step": 275 + }, + { + "epoch": 1.2, + "grad_norm": 0.5514339208602905, + "learning_rate": 0.0003926096997690532, + "loss": 0.6918, + "step": 300 + }, + { + "epoch": 1.3, + "grad_norm": 0.33801883459091187, + "learning_rate": 0.00039096007918178817, + "loss": 0.6906, + "step": 325 + }, + { + "epoch": 1.4, + "grad_norm": 0.2953580617904663, + "learning_rate": 0.0003893104585945233, + "loss": 0.6886, + "step": 350 + }, + { + "epoch": 1.5, + "grad_norm": 0.5340980887413025, + "learning_rate": 0.0003876608380072583, + "loss": 0.6877, + "step": 375 + }, + { + "epoch": 1.6, + "grad_norm": 0.3042701482772827, + "learning_rate": 0.00038601121741999343, + "loss": 0.6853, + "step": 400 + }, + { + "epoch": 1.7, + "grad_norm": 0.44497063755989075, + "learning_rate": 0.0003843615968327285, + "loss": 0.6772, + "step": 425 + }, + { + "epoch": 1.8, + "grad_norm": 0.6260074973106384, + "learning_rate": 0.0003827119762454636, + "loss": 0.6671, + "step": 450 + }, + { + "epoch": 1.9, + "grad_norm": 0.9574269652366638, + "learning_rate": 0.00038106235565819863, + "loss": 0.6517, + "step": 475 + }, + { + "epoch": 2.0, + "grad_norm": 0.9055777192115784, + "learning_rate": 0.0003794127350709337, + "loss": 0.6438, + "step": 500 + }, + { + "epoch": 2.1, + "grad_norm": 0.7860134840011597, + "learning_rate": 0.0003777631144836688, + "loss": 0.6379, + "step": 525 + }, + { + "epoch": 2.2, + "grad_norm": 1.5324610471725464, + "learning_rate": 0.00037611349389640383, + "loss": 0.6143, + "step": 550 + }, + { + "epoch": 2.3, + "grad_norm": 1.8535794019699097, + "learning_rate": 0.00037446387330913894, + "loss": 0.607, + "step": 575 + }, + { + "epoch": 2.4, + "grad_norm": 1.8850122690200806, + "learning_rate": 0.000372814252721874, + "loss": 0.5642, + "step": 600 + }, + { + "epoch": 2.5, + "grad_norm": 1.288278579711914, + "learning_rate": 0.00037116463213460903, + "loss": 0.5739, + "step": 625 + }, + { + "epoch": 2.6, + "grad_norm": 1.344173550605774, + "learning_rate": 0.00036951501154734414, + "loss": 0.5718, + "step": 650 + }, + { + "epoch": 2.7, + "grad_norm": 1.430834174156189, + "learning_rate": 0.0003678653909600792, + "loss": 0.543, + "step": 675 + }, + { + "epoch": 2.8, + "grad_norm": 1.5948325395584106, + "learning_rate": 0.0003662157703728143, + "loss": 0.5434, + "step": 700 + }, + { + "epoch": 2.9, + "grad_norm": 2.688519239425659, + "learning_rate": 0.00036456614978554934, + "loss": 0.5842, + "step": 725 + }, + { + "epoch": 3.0, + "grad_norm": 1.6625819206237793, + "learning_rate": 0.00036291652919828444, + "loss": 0.5526, + "step": 750 + }, + { + "epoch": 3.1, + "grad_norm": 1.869725227355957, + "learning_rate": 0.00036126690861101944, + "loss": 0.53, + "step": 775 + }, + { + "epoch": 3.2, + "grad_norm": 2.1625709533691406, + "learning_rate": 0.00035961728802375454, + "loss": 0.5341, + "step": 800 + }, + { + "epoch": 3.3, + "grad_norm": 1.9650202989578247, + "learning_rate": 0.0003579676674364896, + "loss": 0.5297, + "step": 825 + }, + { + "epoch": 3.4, + "grad_norm": 1.3358676433563232, + "learning_rate": 0.0003563180468492247, + "loss": 0.5273, + "step": 850 + }, + { + "epoch": 3.5, + "grad_norm": 2.1927096843719482, + "learning_rate": 0.0003546684262619598, + "loss": 0.5084, + "step": 875 + }, + { + "epoch": 3.6, + "grad_norm": 2.2377188205718994, + "learning_rate": 0.00035301880567469485, + "loss": 0.4972, + "step": 900 + }, + { + "epoch": 3.7, + "grad_norm": 1.8601839542388916, + "learning_rate": 0.0003513691850874299, + "loss": 0.4852, + "step": 925 + }, + { + "epoch": 3.8, + "grad_norm": 1.4995975494384766, + "learning_rate": 0.00034971956450016495, + "loss": 0.5309, + "step": 950 + }, + { + "epoch": 3.9, + "grad_norm": 2.0536930561065674, + "learning_rate": 0.00034806994391290005, + "loss": 0.5249, + "step": 975 + }, + { + "epoch": 4.0, + "grad_norm": 2.084728479385376, + "learning_rate": 0.0003464203233256351, + "loss": 0.4771, + "step": 1000 + }, + { + "epoch": 4.1, + "grad_norm": 1.705510139465332, + "learning_rate": 0.0003447707027383702, + "loss": 0.5181, + "step": 1025 + }, + { + "epoch": 4.2, + "grad_norm": 3.83380389213562, + "learning_rate": 0.00034312108215110525, + "loss": 0.4873, + "step": 1050 + }, + { + "epoch": 4.3, + "grad_norm": 2.0163533687591553, + "learning_rate": 0.0003414714615638403, + "loss": 0.4801, + "step": 1075 + }, + { + "epoch": 4.4, + "grad_norm": 1.9917407035827637, + "learning_rate": 0.0003398218409765754, + "loss": 0.4766, + "step": 1100 + }, + { + "epoch": 4.5, + "grad_norm": 2.111729860305786, + "learning_rate": 0.00033817222038931045, + "loss": 0.4877, + "step": 1125 + }, + { + "epoch": 4.6, + "grad_norm": 1.5487738847732544, + "learning_rate": 0.00033652259980204556, + "loss": 0.507, + "step": 1150 + }, + { + "epoch": 4.7, + "grad_norm": 1.8715145587921143, + "learning_rate": 0.0003348729792147806, + "loss": 0.452, + "step": 1175 + }, + { + "epoch": 4.8, + "grad_norm": 2.2618906497955322, + "learning_rate": 0.0003332233586275157, + "loss": 0.4799, + "step": 1200 + }, + { + "epoch": 4.9, + "grad_norm": 1.657012939453125, + "learning_rate": 0.00033157373804025076, + "loss": 0.4715, + "step": 1225 + }, + { + "epoch": 5.0, + "grad_norm": 2.8894591331481934, + "learning_rate": 0.0003299241174529858, + "loss": 0.4437, + "step": 1250 + }, + { + "epoch": 5.1, + "grad_norm": 1.9179915189743042, + "learning_rate": 0.0003282744968657209, + "loss": 0.4796, + "step": 1275 + }, + { + "epoch": 5.2, + "grad_norm": 2.1816720962524414, + "learning_rate": 0.00032662487627845596, + "loss": 0.4463, + "step": 1300 + }, + { + "epoch": 5.3, + "grad_norm": 2.2761082649230957, + "learning_rate": 0.00032497525569119106, + "loss": 0.4215, + "step": 1325 + }, + { + "epoch": 5.4, + "grad_norm": 1.7190684080123901, + "learning_rate": 0.0003233256351039261, + "loss": 0.4481, + "step": 1350 + }, + { + "epoch": 5.5, + "grad_norm": 1.4793736934661865, + "learning_rate": 0.00032167601451666116, + "loss": 0.44, + "step": 1375 + }, + { + "epoch": 5.6, + "grad_norm": 2.5064358711242676, + "learning_rate": 0.00032002639392939627, + "loss": 0.4583, + "step": 1400 + }, + { + "epoch": 5.7, + "grad_norm": 1.8364307880401611, + "learning_rate": 0.0003183767733421313, + "loss": 0.4516, + "step": 1425 + }, + { + "epoch": 5.8, + "grad_norm": 1.3747583627700806, + "learning_rate": 0.0003167271527548664, + "loss": 0.4535, + "step": 1450 + }, + { + "epoch": 5.9, + "grad_norm": 2.580183982849121, + "learning_rate": 0.00031507753216760147, + "loss": 0.4325, + "step": 1475 + }, + { + "epoch": 6.0, + "grad_norm": 1.562374472618103, + "learning_rate": 0.00031342791158033657, + "loss": 0.4616, + "step": 1500 + }, + { + "epoch": 6.1, + "grad_norm": 1.4970141649246216, + "learning_rate": 0.00031177829099307157, + "loss": 0.4607, + "step": 1525 + }, + { + "epoch": 6.2, + "grad_norm": 2.436326503753662, + "learning_rate": 0.00031012867040580667, + "loss": 0.4498, + "step": 1550 + }, + { + "epoch": 6.3, + "grad_norm": 1.9849016666412354, + "learning_rate": 0.0003084790498185417, + "loss": 0.4262, + "step": 1575 + }, + { + "epoch": 6.4, + "grad_norm": 1.6405338048934937, + "learning_rate": 0.0003068294292312768, + "loss": 0.4225, + "step": 1600 + }, + { + "epoch": 6.5, + "grad_norm": 1.8233951330184937, + "learning_rate": 0.0003051798086440119, + "loss": 0.4248, + "step": 1625 + }, + { + "epoch": 6.6, + "grad_norm": 1.396469235420227, + "learning_rate": 0.000303530188056747, + "loss": 0.412, + "step": 1650 + }, + { + "epoch": 6.7, + "grad_norm": 1.5438531637191772, + "learning_rate": 0.000301880567469482, + "loss": 0.4105, + "step": 1675 + }, + { + "epoch": 6.8, + "grad_norm": 2.49668288230896, + "learning_rate": 0.0003002309468822171, + "loss": 0.4318, + "step": 1700 + }, + { + "epoch": 6.9, + "grad_norm": 1.5911107063293457, + "learning_rate": 0.0002985813262949522, + "loss": 0.4454, + "step": 1725 + }, + { + "epoch": 7.0, + "grad_norm": 2.9934186935424805, + "learning_rate": 0.00029693170570768723, + "loss": 0.4306, + "step": 1750 + }, + { + "epoch": 7.1, + "grad_norm": 2.5476737022399902, + "learning_rate": 0.00029528208512042233, + "loss": 0.4066, + "step": 1775 + }, + { + "epoch": 7.2, + "grad_norm": 4.207254886627197, + "learning_rate": 0.0002936324645331574, + "loss": 0.4521, + "step": 1800 + }, + { + "epoch": 7.3, + "grad_norm": 1.8378102779388428, + "learning_rate": 0.00029198284394589243, + "loss": 0.4197, + "step": 1825 + }, + { + "epoch": 7.4, + "grad_norm": 2.2129015922546387, + "learning_rate": 0.00029033322335862753, + "loss": 0.4182, + "step": 1850 + }, + { + "epoch": 7.5, + "grad_norm": 1.4666684865951538, + "learning_rate": 0.0002886836027713626, + "loss": 0.4224, + "step": 1875 + }, + { + "epoch": 7.6, + "grad_norm": 2.0311567783355713, + "learning_rate": 0.0002870339821840977, + "loss": 0.4304, + "step": 1900 + }, + { + "epoch": 7.7, + "grad_norm": 2.346604585647583, + "learning_rate": 0.00028538436159683273, + "loss": 0.4012, + "step": 1925 + }, + { + "epoch": 7.8, + "grad_norm": 1.440136432647705, + "learning_rate": 0.00028373474100956784, + "loss": 0.4487, + "step": 1950 + }, + { + "epoch": 7.9, + "grad_norm": 2.426828384399414, + "learning_rate": 0.0002820851204223029, + "loss": 0.3577, + "step": 1975 + }, + { + "epoch": 8.0, + "grad_norm": 3.095980405807495, + "learning_rate": 0.00028043549983503794, + "loss": 0.4121, + "step": 2000 + }, + { + "epoch": 8.1, + "grad_norm": 1.8646572828292847, + "learning_rate": 0.00027878587924777304, + "loss": 0.4213, + "step": 2025 + }, + { + "epoch": 8.2, + "grad_norm": 2.3382062911987305, + "learning_rate": 0.0002771362586605081, + "loss": 0.4119, + "step": 2050 + }, + { + "epoch": 8.3, + "grad_norm": 1.7999392747879028, + "learning_rate": 0.0002754866380732432, + "loss": 0.4165, + "step": 2075 + }, + { + "epoch": 8.4, + "grad_norm": 2.262005567550659, + "learning_rate": 0.00027383701748597824, + "loss": 0.4043, + "step": 2100 + }, + { + "epoch": 8.5, + "grad_norm": 1.8611185550689697, + "learning_rate": 0.0002721873968987133, + "loss": 0.4022, + "step": 2125 + }, + { + "epoch": 8.6, + "grad_norm": 1.3322739601135254, + "learning_rate": 0.00027053777631144834, + "loss": 0.3835, + "step": 2150 + }, + { + "epoch": 8.7, + "grad_norm": 1.335750937461853, + "learning_rate": 0.00026888815572418344, + "loss": 0.424, + "step": 2175 + }, + { + "epoch": 8.8, + "grad_norm": 1.4633246660232544, + "learning_rate": 0.00026723853513691855, + "loss": 0.4124, + "step": 2200 + }, + { + "epoch": 8.9, + "grad_norm": 2.061042308807373, + "learning_rate": 0.0002655889145496536, + "loss": 0.4016, + "step": 2225 + }, + { + "epoch": 9.0, + "grad_norm": 1.9621604681015015, + "learning_rate": 0.0002639392939623887, + "loss": 0.4256, + "step": 2250 + }, + { + "epoch": 9.1, + "grad_norm": 1.4014378786087036, + "learning_rate": 0.0002622896733751237, + "loss": 0.3806, + "step": 2275 + }, + { + "epoch": 9.2, + "grad_norm": 2.6264309883117676, + "learning_rate": 0.0002606400527878588, + "loss": 0.3636, + "step": 2300 + }, + { + "epoch": 9.3, + "grad_norm": 1.8708316087722778, + "learning_rate": 0.00025899043220059385, + "loss": 0.4101, + "step": 2325 + }, + { + "epoch": 9.4, + "grad_norm": 2.156869649887085, + "learning_rate": 0.00025734081161332895, + "loss": 0.4114, + "step": 2350 + }, + { + "epoch": 9.5, + "grad_norm": 2.483182907104492, + "learning_rate": 0.000255691191026064, + "loss": 0.421, + "step": 2375 + }, + { + "epoch": 9.6, + "grad_norm": 2.1955626010894775, + "learning_rate": 0.0002540415704387991, + "loss": 0.3918, + "step": 2400 + }, + { + "epoch": 9.7, + "grad_norm": 4.115762233734131, + "learning_rate": 0.00025239194985153415, + "loss": 0.4172, + "step": 2425 + }, + { + "epoch": 9.8, + "grad_norm": 2.144148349761963, + "learning_rate": 0.0002507423292642692, + "loss": 0.3636, + "step": 2450 + }, + { + "epoch": 9.9, + "grad_norm": 2.340376377105713, + "learning_rate": 0.0002490927086770043, + "loss": 0.4059, + "step": 2475 + }, + { + "epoch": 10.0, + "grad_norm": 1.5438555479049683, + "learning_rate": 0.00024744308808973936, + "loss": 0.4232, + "step": 2500 + }, + { + "epoch": 10.1, + "grad_norm": 2.485889196395874, + "learning_rate": 0.00024579346750247446, + "loss": 0.3861, + "step": 2525 + }, + { + "epoch": 10.2, + "grad_norm": 2.4561119079589844, + "learning_rate": 0.0002441438469152095, + "loss": 0.3565, + "step": 2550 + }, + { + "epoch": 10.3, + "grad_norm": 2.460170269012451, + "learning_rate": 0.00024249422632794456, + "loss": 0.4003, + "step": 2575 + }, + { + "epoch": 10.4, + "grad_norm": 2.774445056915283, + "learning_rate": 0.00024084460574067963, + "loss": 0.4088, + "step": 2600 + }, + { + "epoch": 10.5, + "grad_norm": 2.678745985031128, + "learning_rate": 0.0002391949851534147, + "loss": 0.4021, + "step": 2625 + }, + { + "epoch": 10.6, + "grad_norm": 1.4933565855026245, + "learning_rate": 0.0002375453645661498, + "loss": 0.3744, + "step": 2650 + }, + { + "epoch": 10.7, + "grad_norm": 1.367900013923645, + "learning_rate": 0.00023589574397888486, + "loss": 0.3925, + "step": 2675 + }, + { + "epoch": 10.8, + "grad_norm": 1.653441309928894, + "learning_rate": 0.00023424612339161997, + "loss": 0.4049, + "step": 2700 + }, + { + "epoch": 10.9, + "grad_norm": 1.5892399549484253, + "learning_rate": 0.000232596502804355, + "loss": 0.3927, + "step": 2725 + }, + { + "epoch": 11.0, + "grad_norm": 1.535090684890747, + "learning_rate": 0.00023094688221709007, + "loss": 0.3739, + "step": 2750 + }, + { + "epoch": 11.1, + "grad_norm": 2.247271776199341, + "learning_rate": 0.00022929726162982514, + "loss": 0.3362, + "step": 2775 + }, + { + "epoch": 11.2, + "grad_norm": 2.408069372177124, + "learning_rate": 0.00022764764104256022, + "loss": 0.3933, + "step": 2800 + }, + { + "epoch": 11.3, + "grad_norm": 2.219688653945923, + "learning_rate": 0.0002259980204552953, + "loss": 0.4062, + "step": 2825 + }, + { + "epoch": 11.4, + "grad_norm": 3.0188546180725098, + "learning_rate": 0.00022434839986803037, + "loss": 0.3926, + "step": 2850 + }, + { + "epoch": 11.5, + "grad_norm": 2.8080244064331055, + "learning_rate": 0.00022269877928076542, + "loss": 0.3942, + "step": 2875 + }, + { + "epoch": 11.6, + "grad_norm": 2.263610601425171, + "learning_rate": 0.0002210491586935005, + "loss": 0.3887, + "step": 2900 + }, + { + "epoch": 11.7, + "grad_norm": 2.1239607334136963, + "learning_rate": 0.00021939953810623557, + "loss": 0.3989, + "step": 2925 + }, + { + "epoch": 11.8, + "grad_norm": 3.435351848602295, + "learning_rate": 0.00021774991751897065, + "loss": 0.4185, + "step": 2950 + }, + { + "epoch": 11.9, + "grad_norm": 4.1570515632629395, + "learning_rate": 0.00021610029693170573, + "loss": 0.3597, + "step": 2975 + }, + { + "epoch": 12.0, + "grad_norm": 1.7937941551208496, + "learning_rate": 0.0002144506763444408, + "loss": 0.3719, + "step": 3000 + }, + { + "epoch": 12.1, + "grad_norm": 2.3080639839172363, + "learning_rate": 0.00021280105575717582, + "loss": 0.3856, + "step": 3025 + }, + { + "epoch": 12.2, + "grad_norm": 1.9331200122833252, + "learning_rate": 0.00021115143516991093, + "loss": 0.3798, + "step": 3050 + }, + { + "epoch": 12.3, + "grad_norm": 2.0342001914978027, + "learning_rate": 0.000209501814582646, + "loss": 0.3789, + "step": 3075 + }, + { + "epoch": 12.4, + "grad_norm": 1.7878581285476685, + "learning_rate": 0.00020785219399538108, + "loss": 0.3546, + "step": 3100 + }, + { + "epoch": 12.5, + "grad_norm": 3.657317638397217, + "learning_rate": 0.00020620257340811616, + "loss": 0.3582, + "step": 3125 + }, + { + "epoch": 12.6, + "grad_norm": 2.246429681777954, + "learning_rate": 0.00020455295282085123, + "loss": 0.3991, + "step": 3150 + }, + { + "epoch": 12.7, + "grad_norm": 1.7742749452590942, + "learning_rate": 0.00020290333223358626, + "loss": 0.3914, + "step": 3175 + }, + { + "epoch": 12.8, + "grad_norm": 2.3869900703430176, + "learning_rate": 0.00020125371164632133, + "loss": 0.3435, + "step": 3200 + }, + { + "epoch": 12.9, + "grad_norm": 2.444920539855957, + "learning_rate": 0.0001996040910590564, + "loss": 0.393, + "step": 3225 + }, + { + "epoch": 13.0, + "grad_norm": 2.26476788520813, + "learning_rate": 0.00019795447047179148, + "loss": 0.3828, + "step": 3250 + }, + { + "epoch": 13.1, + "grad_norm": 2.4492876529693604, + "learning_rate": 0.0001963048498845266, + "loss": 0.3856, + "step": 3275 + }, + { + "epoch": 13.2, + "grad_norm": 4.578381061553955, + "learning_rate": 0.00019465522929726164, + "loss": 0.3676, + "step": 3300 + }, + { + "epoch": 13.3, + "grad_norm": 2.0352611541748047, + "learning_rate": 0.00019300560870999671, + "loss": 0.369, + "step": 3325 + }, + { + "epoch": 13.4, + "grad_norm": 2.151038646697998, + "learning_rate": 0.0001913559881227318, + "loss": 0.3713, + "step": 3350 + }, + { + "epoch": 13.5, + "grad_norm": 2.4661877155303955, + "learning_rate": 0.00018970636753546684, + "loss": 0.3859, + "step": 3375 + }, + { + "epoch": 13.6, + "grad_norm": 1.7616342306137085, + "learning_rate": 0.00018805674694820192, + "loss": 0.363, + "step": 3400 + }, + { + "epoch": 13.7, + "grad_norm": 2.26426362991333, + "learning_rate": 0.000186407126360937, + "loss": 0.3696, + "step": 3425 + }, + { + "epoch": 13.8, + "grad_norm": 2.355781078338623, + "learning_rate": 0.00018475750577367207, + "loss": 0.372, + "step": 3450 + }, + { + "epoch": 13.9, + "grad_norm": 1.961585283279419, + "learning_rate": 0.00018310788518640715, + "loss": 0.3418, + "step": 3475 + }, + { + "epoch": 14.0, + "grad_norm": 2.54199481010437, + "learning_rate": 0.00018145826459914222, + "loss": 0.3648, + "step": 3500 + }, + { + "epoch": 14.1, + "grad_norm": 1.5017932653427124, + "learning_rate": 0.00017980864401187727, + "loss": 0.3914, + "step": 3525 + }, + { + "epoch": 14.2, + "grad_norm": 3.051287889480591, + "learning_rate": 0.00017815902342461235, + "loss": 0.3887, + "step": 3550 + }, + { + "epoch": 14.3, + "grad_norm": 2.697949171066284, + "learning_rate": 0.00017650940283734742, + "loss": 0.3524, + "step": 3575 + }, + { + "epoch": 14.4, + "grad_norm": 2.470728635787964, + "learning_rate": 0.00017485978225008247, + "loss": 0.3767, + "step": 3600 + }, + { + "epoch": 14.5, + "grad_norm": 2.5166375637054443, + "learning_rate": 0.00017321016166281755, + "loss": 0.362, + "step": 3625 + }, + { + "epoch": 14.6, + "grad_norm": 1.7630703449249268, + "learning_rate": 0.00017156054107555263, + "loss": 0.3936, + "step": 3650 + }, + { + "epoch": 14.7, + "grad_norm": 2.083225727081299, + "learning_rate": 0.0001699109204882877, + "loss": 0.36, + "step": 3675 + }, + { + "epoch": 14.8, + "grad_norm": 1.7796965837478638, + "learning_rate": 0.00016826129990102278, + "loss": 0.3646, + "step": 3700 + }, + { + "epoch": 14.9, + "grad_norm": 2.127898693084717, + "learning_rate": 0.00016661167931375785, + "loss": 0.3722, + "step": 3725 + }, + { + "epoch": 15.0, + "grad_norm": 2.666217088699341, + "learning_rate": 0.0001649620587264929, + "loss": 0.3329, + "step": 3750 + }, + { + "epoch": 15.1, + "grad_norm": 2.521188497543335, + "learning_rate": 0.00016331243813922798, + "loss": 0.331, + "step": 3775 + }, + { + "epoch": 15.2, + "grad_norm": 1.528865098953247, + "learning_rate": 0.00016166281755196306, + "loss": 0.3513, + "step": 3800 + }, + { + "epoch": 15.3, + "grad_norm": 3.452749490737915, + "learning_rate": 0.00016001319696469813, + "loss": 0.3777, + "step": 3825 + }, + { + "epoch": 15.4, + "grad_norm": 3.0441641807556152, + "learning_rate": 0.0001583635763774332, + "loss": 0.3536, + "step": 3850 + }, + { + "epoch": 15.5, + "grad_norm": 1.9194542169570923, + "learning_rate": 0.00015671395579016829, + "loss": 0.3635, + "step": 3875 + }, + { + "epoch": 15.6, + "grad_norm": 2.011859655380249, + "learning_rate": 0.00015506433520290334, + "loss": 0.3467, + "step": 3900 + }, + { + "epoch": 15.7, + "grad_norm": 1.5169134140014648, + "learning_rate": 0.0001534147146156384, + "loss": 0.3639, + "step": 3925 + }, + { + "epoch": 15.8, + "grad_norm": 2.715104103088379, + "learning_rate": 0.0001517650940283735, + "loss": 0.3473, + "step": 3950 + }, + { + "epoch": 15.9, + "grad_norm": 1.8220083713531494, + "learning_rate": 0.00015011547344110854, + "loss": 0.3902, + "step": 3975 + }, + { + "epoch": 16.0, + "grad_norm": 3.227846384048462, + "learning_rate": 0.00014846585285384361, + "loss": 0.3576, + "step": 4000 + }, + { + "epoch": 16.1, + "grad_norm": 1.9271538257598877, + "learning_rate": 0.0001468162322665787, + "loss": 0.3426, + "step": 4025 + }, + { + "epoch": 16.2, + "grad_norm": 2.58815336227417, + "learning_rate": 0.00014516661167931377, + "loss": 0.3714, + "step": 4050 + }, + { + "epoch": 16.3, + "grad_norm": 1.7768404483795166, + "learning_rate": 0.00014351699109204884, + "loss": 0.3561, + "step": 4075 + }, + { + "epoch": 16.4, + "grad_norm": 1.622894525527954, + "learning_rate": 0.00014186737050478392, + "loss": 0.3384, + "step": 4100 + }, + { + "epoch": 16.5, + "grad_norm": 1.8811430931091309, + "learning_rate": 0.00014021774991751897, + "loss": 0.3603, + "step": 4125 + }, + { + "epoch": 16.6, + "grad_norm": 2.1986522674560547, + "learning_rate": 0.00013856812933025404, + "loss": 0.34, + "step": 4150 + }, + { + "epoch": 16.7, + "grad_norm": 2.3372039794921875, + "learning_rate": 0.00013691850874298912, + "loss": 0.368, + "step": 4175 + }, + { + "epoch": 16.8, + "grad_norm": 1.5386862754821777, + "learning_rate": 0.00013526888815572417, + "loss": 0.3441, + "step": 4200 + }, + { + "epoch": 16.9, + "grad_norm": 1.9392023086547852, + "learning_rate": 0.00013361926756845927, + "loss": 0.3704, + "step": 4225 + }, + { + "epoch": 17.0, + "grad_norm": 3.1489596366882324, + "learning_rate": 0.00013196964698119435, + "loss": 0.3466, + "step": 4250 + }, + { + "epoch": 17.1, + "grad_norm": 3.081573724746704, + "learning_rate": 0.0001303200263939294, + "loss": 0.3352, + "step": 4275 + }, + { + "epoch": 17.2, + "grad_norm": 1.7880629301071167, + "learning_rate": 0.00012867040580666448, + "loss": 0.3569, + "step": 4300 + }, + { + "epoch": 17.3, + "grad_norm": 2.12654972076416, + "learning_rate": 0.00012702078521939955, + "loss": 0.3493, + "step": 4325 + }, + { + "epoch": 17.4, + "grad_norm": 2.646132230758667, + "learning_rate": 0.0001253711646321346, + "loss": 0.346, + "step": 4350 + }, + { + "epoch": 17.5, + "grad_norm": 1.6917959451675415, + "learning_rate": 0.00012372154404486968, + "loss": 0.3544, + "step": 4375 + }, + { + "epoch": 17.6, + "grad_norm": 1.7370896339416504, + "learning_rate": 0.00012207192345760475, + "loss": 0.3392, + "step": 4400 + }, + { + "epoch": 17.7, + "grad_norm": 2.7857675552368164, + "learning_rate": 0.00012042230287033982, + "loss": 0.3479, + "step": 4425 + }, + { + "epoch": 17.8, + "grad_norm": 2.1937146186828613, + "learning_rate": 0.0001187726822830749, + "loss": 0.4039, + "step": 4450 + }, + { + "epoch": 17.9, + "grad_norm": 2.3681347370147705, + "learning_rate": 0.00011712306169580998, + "loss": 0.3534, + "step": 4475 + }, + { + "epoch": 18.0, + "grad_norm": 1.5830249786376953, + "learning_rate": 0.00011547344110854503, + "loss": 0.3599, + "step": 4500 + }, + { + "epoch": 18.1, + "grad_norm": 2.077099084854126, + "learning_rate": 0.00011382382052128011, + "loss": 0.3369, + "step": 4525 + }, + { + "epoch": 18.2, + "grad_norm": 2.1928341388702393, + "learning_rate": 0.00011217419993401519, + "loss": 0.3525, + "step": 4550 + }, + { + "epoch": 18.3, + "grad_norm": 1.9083303213119507, + "learning_rate": 0.00011052457934675025, + "loss": 0.3781, + "step": 4575 + }, + { + "epoch": 18.4, + "grad_norm": 1.9660146236419678, + "learning_rate": 0.00010887495875948532, + "loss": 0.3589, + "step": 4600 + }, + { + "epoch": 18.5, + "grad_norm": 1.6620488166809082, + "learning_rate": 0.0001072253381722204, + "loss": 0.3571, + "step": 4625 + }, + { + "epoch": 18.6, + "grad_norm": 2.4148569107055664, + "learning_rate": 0.00010557571758495546, + "loss": 0.3718, + "step": 4650 + }, + { + "epoch": 18.7, + "grad_norm": 2.578110933303833, + "learning_rate": 0.00010392609699769054, + "loss": 0.3491, + "step": 4675 + }, + { + "epoch": 18.8, + "grad_norm": 2.4224448204040527, + "learning_rate": 0.00010227647641042562, + "loss": 0.3241, + "step": 4700 + }, + { + "epoch": 18.9, + "grad_norm": 1.9691663980484009, + "learning_rate": 0.00010062685582316067, + "loss": 0.3407, + "step": 4725 + }, + { + "epoch": 19.0, + "grad_norm": 1.8373006582260132, + "learning_rate": 9.897723523589574e-05, + "loss": 0.3415, + "step": 4750 + }, + { + "epoch": 19.1, + "grad_norm": 2.405202865600586, + "learning_rate": 9.732761464863082e-05, + "loss": 0.3259, + "step": 4775 + }, + { + "epoch": 19.2, + "grad_norm": 2.754739761352539, + "learning_rate": 9.56779940613659e-05, + "loss": 0.3361, + "step": 4800 + }, + { + "epoch": 19.3, + "grad_norm": 2.4953665733337402, + "learning_rate": 9.402837347410096e-05, + "loss": 0.3534, + "step": 4825 + }, + { + "epoch": 19.4, + "grad_norm": 2.974691390991211, + "learning_rate": 9.237875288683603e-05, + "loss": 0.3742, + "step": 4850 + }, + { + "epoch": 19.5, + "grad_norm": 1.4064937829971313, + "learning_rate": 9.072913229957111e-05, + "loss": 0.3271, + "step": 4875 + }, + { + "epoch": 19.6, + "grad_norm": 3.1546576023101807, + "learning_rate": 8.907951171230617e-05, + "loss": 0.3549, + "step": 4900 + }, + { + "epoch": 19.7, + "grad_norm": 1.948807716369629, + "learning_rate": 8.742989112504124e-05, + "loss": 0.3107, + "step": 4925 + }, + { + "epoch": 19.8, + "grad_norm": 1.1512947082519531, + "learning_rate": 8.578027053777631e-05, + "loss": 0.3448, + "step": 4950 + }, + { + "epoch": 19.9, + "grad_norm": 3.7861363887786865, + "learning_rate": 8.413064995051139e-05, + "loss": 0.3312, + "step": 4975 + }, + { + "epoch": 20.0, + "grad_norm": 1.6949777603149414, + "learning_rate": 8.248102936324645e-05, + "loss": 0.3532, + "step": 5000 + } + ], + "logging_steps": 25, + "max_steps": 6250, + "num_input_tokens_seen": 0, + "num_train_epochs": 25, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 276905164800000.0, + "train_batch_size": 20, + "trial_name": null, + "trial_params": null +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/training_args.bin b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..10d16bb8e9df9c51aba38e7a6f04deeecf1ea29f --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:662a5f25d38ccae34f19d85e24456de6d3eaaabafca26fb7cc8816de0c990f20 +size 5304 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/vocab.txt b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/vocab.txt new file mode 100644 index 0000000000000000000000000000000000000000..cc9ad53732e21ea9180f18e1fb67a38cc5702d4c --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/vocab.txt @@ -0,0 +1,4101 @@ +[PAD] +[UNK] +[CLS] +[SEP] +[MASK] +AAAAAA +AAAAAT +AAAAAC +AAAAAG +AAAATA +AAAATT +AAAATC +AAAATG +AAAACA +AAAACT +AAAACC +AAAACG +AAAAGA +AAAAGT +AAAAGC +AAAAGG +AAATAA +AAATAT +AAATAC +AAATAG +AAATTA +AAATTT +AAATTC +AAATTG +AAATCA +AAATCT +AAATCC +AAATCG +AAATGA +AAATGT +AAATGC +AAATGG +AAACAA +AAACAT +AAACAC +AAACAG +AAACTA +AAACTT +AAACTC +AAACTG +AAACCA +AAACCT +AAACCC +AAACCG +AAACGA +AAACGT +AAACGC +AAACGG +AAAGAA +AAAGAT +AAAGAC +AAAGAG +AAAGTA +AAAGTT +AAAGTC +AAAGTG +AAAGCA +AAAGCT +AAAGCC +AAAGCG +AAAGGA +AAAGGT +AAAGGC +AAAGGG +AATAAA +AATAAT +AATAAC +AATAAG +AATATA +AATATT +AATATC +AATATG +AATACA +AATACT +AATACC +AATACG +AATAGA +AATAGT +AATAGC +AATAGG +AATTAA +AATTAT +AATTAC +AATTAG +AATTTA +AATTTT +AATTTC +AATTTG +AATTCA +AATTCT +AATTCC +AATTCG +AATTGA +AATTGT +AATTGC +AATTGG +AATCAA +AATCAT +AATCAC +AATCAG +AATCTA +AATCTT +AATCTC +AATCTG +AATCCA +AATCCT +AATCCC +AATCCG +AATCGA +AATCGT +AATCGC +AATCGG +AATGAA +AATGAT +AATGAC +AATGAG +AATGTA +AATGTT +AATGTC +AATGTG +AATGCA +AATGCT +AATGCC +AATGCG +AATGGA +AATGGT +AATGGC +AATGGG +AACAAA +AACAAT +AACAAC +AACAAG +AACATA +AACATT +AACATC +AACATG +AACACA +AACACT +AACACC +AACACG +AACAGA +AACAGT +AACAGC +AACAGG +AACTAA +AACTAT +AACTAC +AACTAG +AACTTA +AACTTT +AACTTC +AACTTG +AACTCA +AACTCT +AACTCC +AACTCG +AACTGA +AACTGT +AACTGC +AACTGG +AACCAA +AACCAT +AACCAC +AACCAG +AACCTA +AACCTT +AACCTC +AACCTG +AACCCA +AACCCT +AACCCC +AACCCG +AACCGA +AACCGT +AACCGC +AACCGG +AACGAA +AACGAT +AACGAC +AACGAG +AACGTA +AACGTT +AACGTC +AACGTG +AACGCA +AACGCT +AACGCC +AACGCG +AACGGA +AACGGT +AACGGC +AACGGG +AAGAAA +AAGAAT +AAGAAC +AAGAAG +AAGATA +AAGATT +AAGATC +AAGATG +AAGACA +AAGACT +AAGACC +AAGACG +AAGAGA +AAGAGT +AAGAGC +AAGAGG +AAGTAA +AAGTAT +AAGTAC +AAGTAG +AAGTTA +AAGTTT +AAGTTC +AAGTTG +AAGTCA +AAGTCT +AAGTCC +AAGTCG +AAGTGA +AAGTGT +AAGTGC +AAGTGG +AAGCAA +AAGCAT +AAGCAC +AAGCAG +AAGCTA +AAGCTT +AAGCTC +AAGCTG +AAGCCA +AAGCCT +AAGCCC +AAGCCG +AAGCGA +AAGCGT +AAGCGC +AAGCGG +AAGGAA +AAGGAT +AAGGAC +AAGGAG +AAGGTA +AAGGTT +AAGGTC +AAGGTG +AAGGCA +AAGGCT +AAGGCC +AAGGCG +AAGGGA +AAGGGT +AAGGGC +AAGGGG +ATAAAA +ATAAAT +ATAAAC +ATAAAG +ATAATA +ATAATT +ATAATC +ATAATG +ATAACA +ATAACT +ATAACC +ATAACG +ATAAGA +ATAAGT +ATAAGC +ATAAGG +ATATAA +ATATAT +ATATAC +ATATAG +ATATTA +ATATTT +ATATTC +ATATTG +ATATCA +ATATCT +ATATCC +ATATCG +ATATGA +ATATGT +ATATGC +ATATGG +ATACAA +ATACAT +ATACAC +ATACAG +ATACTA +ATACTT +ATACTC +ATACTG +ATACCA +ATACCT +ATACCC +ATACCG +ATACGA +ATACGT +ATACGC +ATACGG +ATAGAA +ATAGAT +ATAGAC +ATAGAG +ATAGTA +ATAGTT +ATAGTC +ATAGTG +ATAGCA +ATAGCT +ATAGCC +ATAGCG +ATAGGA +ATAGGT +ATAGGC +ATAGGG +ATTAAA +ATTAAT +ATTAAC +ATTAAG +ATTATA +ATTATT +ATTATC +ATTATG +ATTACA +ATTACT +ATTACC +ATTACG +ATTAGA +ATTAGT +ATTAGC +ATTAGG +ATTTAA +ATTTAT +ATTTAC +ATTTAG +ATTTTA +ATTTTT +ATTTTC +ATTTTG +ATTTCA +ATTTCT +ATTTCC +ATTTCG +ATTTGA +ATTTGT +ATTTGC +ATTTGG +ATTCAA +ATTCAT +ATTCAC +ATTCAG +ATTCTA +ATTCTT +ATTCTC +ATTCTG +ATTCCA +ATTCCT +ATTCCC +ATTCCG +ATTCGA +ATTCGT +ATTCGC +ATTCGG +ATTGAA +ATTGAT +ATTGAC +ATTGAG +ATTGTA +ATTGTT +ATTGTC +ATTGTG +ATTGCA +ATTGCT +ATTGCC +ATTGCG +ATTGGA +ATTGGT +ATTGGC +ATTGGG +ATCAAA +ATCAAT +ATCAAC +ATCAAG +ATCATA +ATCATT +ATCATC +ATCATG +ATCACA +ATCACT +ATCACC +ATCACG +ATCAGA +ATCAGT +ATCAGC +ATCAGG +ATCTAA +ATCTAT +ATCTAC +ATCTAG +ATCTTA +ATCTTT +ATCTTC +ATCTTG +ATCTCA +ATCTCT +ATCTCC +ATCTCG +ATCTGA +ATCTGT +ATCTGC +ATCTGG +ATCCAA +ATCCAT +ATCCAC +ATCCAG +ATCCTA +ATCCTT +ATCCTC +ATCCTG +ATCCCA +ATCCCT +ATCCCC +ATCCCG +ATCCGA +ATCCGT +ATCCGC +ATCCGG +ATCGAA +ATCGAT +ATCGAC +ATCGAG +ATCGTA +ATCGTT +ATCGTC +ATCGTG +ATCGCA +ATCGCT +ATCGCC +ATCGCG +ATCGGA +ATCGGT +ATCGGC +ATCGGG +ATGAAA +ATGAAT +ATGAAC +ATGAAG +ATGATA +ATGATT +ATGATC +ATGATG +ATGACA +ATGACT +ATGACC +ATGACG +ATGAGA +ATGAGT +ATGAGC +ATGAGG +ATGTAA +ATGTAT +ATGTAC +ATGTAG +ATGTTA +ATGTTT +ATGTTC +ATGTTG +ATGTCA +ATGTCT +ATGTCC +ATGTCG +ATGTGA +ATGTGT +ATGTGC +ATGTGG +ATGCAA +ATGCAT +ATGCAC +ATGCAG +ATGCTA +ATGCTT +ATGCTC +ATGCTG +ATGCCA +ATGCCT +ATGCCC +ATGCCG +ATGCGA +ATGCGT +ATGCGC +ATGCGG +ATGGAA +ATGGAT +ATGGAC +ATGGAG +ATGGTA +ATGGTT +ATGGTC +ATGGTG +ATGGCA +ATGGCT +ATGGCC +ATGGCG +ATGGGA +ATGGGT +ATGGGC +ATGGGG +ACAAAA +ACAAAT +ACAAAC +ACAAAG +ACAATA +ACAATT +ACAATC +ACAATG +ACAACA +ACAACT +ACAACC +ACAACG +ACAAGA +ACAAGT +ACAAGC +ACAAGG +ACATAA +ACATAT +ACATAC +ACATAG +ACATTA +ACATTT +ACATTC +ACATTG +ACATCA +ACATCT +ACATCC +ACATCG +ACATGA +ACATGT +ACATGC +ACATGG +ACACAA +ACACAT +ACACAC +ACACAG +ACACTA +ACACTT +ACACTC +ACACTG +ACACCA +ACACCT +ACACCC +ACACCG +ACACGA +ACACGT +ACACGC +ACACGG +ACAGAA +ACAGAT +ACAGAC +ACAGAG +ACAGTA +ACAGTT +ACAGTC +ACAGTG +ACAGCA +ACAGCT +ACAGCC +ACAGCG +ACAGGA +ACAGGT +ACAGGC +ACAGGG +ACTAAA +ACTAAT +ACTAAC +ACTAAG +ACTATA +ACTATT +ACTATC +ACTATG +ACTACA +ACTACT +ACTACC +ACTACG +ACTAGA +ACTAGT +ACTAGC +ACTAGG +ACTTAA +ACTTAT +ACTTAC +ACTTAG +ACTTTA +ACTTTT +ACTTTC +ACTTTG +ACTTCA +ACTTCT +ACTTCC +ACTTCG +ACTTGA +ACTTGT +ACTTGC +ACTTGG +ACTCAA +ACTCAT +ACTCAC +ACTCAG +ACTCTA +ACTCTT +ACTCTC +ACTCTG +ACTCCA +ACTCCT +ACTCCC +ACTCCG +ACTCGA +ACTCGT +ACTCGC +ACTCGG +ACTGAA +ACTGAT +ACTGAC +ACTGAG +ACTGTA +ACTGTT +ACTGTC +ACTGTG +ACTGCA +ACTGCT +ACTGCC +ACTGCG +ACTGGA +ACTGGT +ACTGGC +ACTGGG +ACCAAA +ACCAAT +ACCAAC +ACCAAG +ACCATA +ACCATT +ACCATC +ACCATG +ACCACA +ACCACT +ACCACC +ACCACG +ACCAGA +ACCAGT +ACCAGC +ACCAGG +ACCTAA +ACCTAT +ACCTAC +ACCTAG +ACCTTA +ACCTTT +ACCTTC +ACCTTG +ACCTCA +ACCTCT +ACCTCC +ACCTCG +ACCTGA +ACCTGT +ACCTGC +ACCTGG +ACCCAA +ACCCAT +ACCCAC +ACCCAG +ACCCTA +ACCCTT +ACCCTC +ACCCTG +ACCCCA +ACCCCT +ACCCCC +ACCCCG +ACCCGA +ACCCGT +ACCCGC +ACCCGG +ACCGAA +ACCGAT +ACCGAC +ACCGAG +ACCGTA +ACCGTT +ACCGTC +ACCGTG +ACCGCA +ACCGCT +ACCGCC +ACCGCG +ACCGGA +ACCGGT +ACCGGC +ACCGGG +ACGAAA +ACGAAT +ACGAAC +ACGAAG +ACGATA +ACGATT +ACGATC +ACGATG +ACGACA +ACGACT +ACGACC +ACGACG +ACGAGA +ACGAGT +ACGAGC +ACGAGG +ACGTAA +ACGTAT +ACGTAC +ACGTAG +ACGTTA +ACGTTT +ACGTTC +ACGTTG +ACGTCA +ACGTCT +ACGTCC +ACGTCG +ACGTGA +ACGTGT +ACGTGC +ACGTGG +ACGCAA +ACGCAT +ACGCAC +ACGCAG +ACGCTA +ACGCTT +ACGCTC +ACGCTG +ACGCCA +ACGCCT +ACGCCC +ACGCCG +ACGCGA +ACGCGT +ACGCGC +ACGCGG +ACGGAA +ACGGAT +ACGGAC +ACGGAG +ACGGTA +ACGGTT +ACGGTC +ACGGTG +ACGGCA +ACGGCT +ACGGCC +ACGGCG +ACGGGA +ACGGGT +ACGGGC +ACGGGG +AGAAAA +AGAAAT +AGAAAC +AGAAAG +AGAATA +AGAATT +AGAATC +AGAATG +AGAACA +AGAACT +AGAACC +AGAACG +AGAAGA +AGAAGT +AGAAGC +AGAAGG +AGATAA +AGATAT +AGATAC +AGATAG +AGATTA +AGATTT +AGATTC +AGATTG +AGATCA +AGATCT +AGATCC +AGATCG +AGATGA +AGATGT +AGATGC +AGATGG +AGACAA +AGACAT +AGACAC +AGACAG +AGACTA +AGACTT +AGACTC +AGACTG +AGACCA +AGACCT +AGACCC +AGACCG +AGACGA +AGACGT +AGACGC +AGACGG +AGAGAA +AGAGAT +AGAGAC +AGAGAG +AGAGTA +AGAGTT +AGAGTC +AGAGTG +AGAGCA +AGAGCT +AGAGCC +AGAGCG +AGAGGA +AGAGGT +AGAGGC +AGAGGG +AGTAAA +AGTAAT +AGTAAC +AGTAAG +AGTATA +AGTATT +AGTATC +AGTATG +AGTACA +AGTACT +AGTACC +AGTACG +AGTAGA +AGTAGT +AGTAGC +AGTAGG +AGTTAA +AGTTAT +AGTTAC +AGTTAG +AGTTTA +AGTTTT +AGTTTC +AGTTTG +AGTTCA +AGTTCT +AGTTCC +AGTTCG +AGTTGA +AGTTGT +AGTTGC +AGTTGG +AGTCAA +AGTCAT +AGTCAC +AGTCAG +AGTCTA +AGTCTT +AGTCTC +AGTCTG +AGTCCA +AGTCCT +AGTCCC +AGTCCG +AGTCGA +AGTCGT +AGTCGC +AGTCGG +AGTGAA +AGTGAT +AGTGAC +AGTGAG +AGTGTA +AGTGTT +AGTGTC +AGTGTG +AGTGCA +AGTGCT +AGTGCC +AGTGCG +AGTGGA +AGTGGT +AGTGGC +AGTGGG +AGCAAA +AGCAAT +AGCAAC +AGCAAG +AGCATA +AGCATT +AGCATC +AGCATG +AGCACA +AGCACT +AGCACC +AGCACG +AGCAGA +AGCAGT +AGCAGC +AGCAGG +AGCTAA +AGCTAT +AGCTAC +AGCTAG +AGCTTA +AGCTTT +AGCTTC +AGCTTG +AGCTCA +AGCTCT +AGCTCC +AGCTCG +AGCTGA +AGCTGT +AGCTGC +AGCTGG +AGCCAA +AGCCAT +AGCCAC +AGCCAG +AGCCTA +AGCCTT +AGCCTC +AGCCTG +AGCCCA +AGCCCT +AGCCCC +AGCCCG +AGCCGA +AGCCGT +AGCCGC +AGCCGG +AGCGAA +AGCGAT +AGCGAC +AGCGAG +AGCGTA +AGCGTT +AGCGTC +AGCGTG +AGCGCA +AGCGCT +AGCGCC +AGCGCG +AGCGGA +AGCGGT +AGCGGC +AGCGGG +AGGAAA +AGGAAT +AGGAAC +AGGAAG +AGGATA +AGGATT +AGGATC +AGGATG +AGGACA +AGGACT +AGGACC +AGGACG +AGGAGA +AGGAGT +AGGAGC +AGGAGG +AGGTAA +AGGTAT +AGGTAC +AGGTAG +AGGTTA +AGGTTT +AGGTTC +AGGTTG +AGGTCA +AGGTCT +AGGTCC +AGGTCG +AGGTGA +AGGTGT +AGGTGC +AGGTGG +AGGCAA +AGGCAT +AGGCAC +AGGCAG +AGGCTA +AGGCTT +AGGCTC +AGGCTG +AGGCCA +AGGCCT +AGGCCC +AGGCCG +AGGCGA +AGGCGT +AGGCGC +AGGCGG +AGGGAA +AGGGAT +AGGGAC +AGGGAG +AGGGTA +AGGGTT +AGGGTC +AGGGTG +AGGGCA +AGGGCT +AGGGCC +AGGGCG +AGGGGA +AGGGGT +AGGGGC +AGGGGG +TAAAAA +TAAAAT +TAAAAC +TAAAAG +TAAATA +TAAATT +TAAATC +TAAATG +TAAACA +TAAACT +TAAACC +TAAACG +TAAAGA +TAAAGT +TAAAGC +TAAAGG +TAATAA +TAATAT +TAATAC +TAATAG +TAATTA +TAATTT +TAATTC +TAATTG +TAATCA +TAATCT +TAATCC +TAATCG +TAATGA +TAATGT +TAATGC +TAATGG +TAACAA +TAACAT +TAACAC +TAACAG +TAACTA +TAACTT +TAACTC +TAACTG +TAACCA +TAACCT +TAACCC +TAACCG +TAACGA +TAACGT +TAACGC +TAACGG +TAAGAA +TAAGAT +TAAGAC +TAAGAG +TAAGTA +TAAGTT +TAAGTC +TAAGTG +TAAGCA +TAAGCT +TAAGCC +TAAGCG +TAAGGA +TAAGGT +TAAGGC +TAAGGG +TATAAA +TATAAT +TATAAC +TATAAG +TATATA +TATATT +TATATC +TATATG +TATACA +TATACT +TATACC +TATACG +TATAGA +TATAGT +TATAGC +TATAGG +TATTAA +TATTAT +TATTAC +TATTAG +TATTTA +TATTTT +TATTTC +TATTTG +TATTCA +TATTCT +TATTCC +TATTCG +TATTGA +TATTGT +TATTGC +TATTGG +TATCAA +TATCAT +TATCAC +TATCAG +TATCTA +TATCTT +TATCTC +TATCTG +TATCCA +TATCCT +TATCCC +TATCCG +TATCGA +TATCGT +TATCGC +TATCGG +TATGAA +TATGAT +TATGAC +TATGAG +TATGTA +TATGTT +TATGTC +TATGTG +TATGCA +TATGCT +TATGCC +TATGCG +TATGGA +TATGGT +TATGGC +TATGGG +TACAAA +TACAAT +TACAAC +TACAAG +TACATA +TACATT +TACATC +TACATG +TACACA +TACACT +TACACC +TACACG +TACAGA +TACAGT +TACAGC +TACAGG +TACTAA +TACTAT +TACTAC +TACTAG +TACTTA +TACTTT +TACTTC +TACTTG +TACTCA +TACTCT +TACTCC +TACTCG +TACTGA +TACTGT +TACTGC +TACTGG +TACCAA +TACCAT +TACCAC +TACCAG +TACCTA +TACCTT +TACCTC +TACCTG +TACCCA +TACCCT +TACCCC +TACCCG +TACCGA +TACCGT +TACCGC +TACCGG +TACGAA +TACGAT +TACGAC +TACGAG +TACGTA +TACGTT +TACGTC +TACGTG +TACGCA +TACGCT +TACGCC +TACGCG +TACGGA +TACGGT +TACGGC +TACGGG +TAGAAA +TAGAAT +TAGAAC +TAGAAG +TAGATA +TAGATT +TAGATC +TAGATG +TAGACA +TAGACT +TAGACC +TAGACG +TAGAGA +TAGAGT +TAGAGC +TAGAGG +TAGTAA +TAGTAT +TAGTAC +TAGTAG +TAGTTA +TAGTTT +TAGTTC +TAGTTG +TAGTCA +TAGTCT +TAGTCC +TAGTCG +TAGTGA +TAGTGT +TAGTGC +TAGTGG +TAGCAA +TAGCAT +TAGCAC +TAGCAG +TAGCTA +TAGCTT +TAGCTC +TAGCTG +TAGCCA +TAGCCT +TAGCCC +TAGCCG +TAGCGA +TAGCGT +TAGCGC +TAGCGG +TAGGAA +TAGGAT +TAGGAC +TAGGAG +TAGGTA +TAGGTT +TAGGTC +TAGGTG +TAGGCA +TAGGCT +TAGGCC +TAGGCG +TAGGGA +TAGGGT +TAGGGC +TAGGGG +TTAAAA +TTAAAT +TTAAAC +TTAAAG +TTAATA +TTAATT +TTAATC +TTAATG +TTAACA +TTAACT +TTAACC +TTAACG +TTAAGA +TTAAGT +TTAAGC +TTAAGG +TTATAA +TTATAT +TTATAC +TTATAG +TTATTA +TTATTT +TTATTC +TTATTG +TTATCA +TTATCT +TTATCC +TTATCG +TTATGA +TTATGT +TTATGC +TTATGG +TTACAA +TTACAT +TTACAC +TTACAG +TTACTA +TTACTT +TTACTC +TTACTG +TTACCA +TTACCT +TTACCC +TTACCG +TTACGA +TTACGT +TTACGC +TTACGG +TTAGAA +TTAGAT +TTAGAC +TTAGAG +TTAGTA +TTAGTT +TTAGTC +TTAGTG +TTAGCA +TTAGCT +TTAGCC +TTAGCG +TTAGGA +TTAGGT +TTAGGC +TTAGGG +TTTAAA +TTTAAT +TTTAAC +TTTAAG +TTTATA +TTTATT +TTTATC +TTTATG +TTTACA +TTTACT +TTTACC +TTTACG +TTTAGA +TTTAGT +TTTAGC +TTTAGG +TTTTAA +TTTTAT +TTTTAC +TTTTAG +TTTTTA +TTTTTT +TTTTTC +TTTTTG +TTTTCA +TTTTCT +TTTTCC +TTTTCG +TTTTGA +TTTTGT +TTTTGC +TTTTGG +TTTCAA +TTTCAT +TTTCAC +TTTCAG +TTTCTA +TTTCTT +TTTCTC +TTTCTG +TTTCCA +TTTCCT +TTTCCC +TTTCCG +TTTCGA +TTTCGT +TTTCGC +TTTCGG +TTTGAA +TTTGAT +TTTGAC +TTTGAG +TTTGTA +TTTGTT +TTTGTC +TTTGTG +TTTGCA +TTTGCT +TTTGCC +TTTGCG +TTTGGA +TTTGGT +TTTGGC +TTTGGG +TTCAAA +TTCAAT +TTCAAC +TTCAAG +TTCATA +TTCATT +TTCATC +TTCATG +TTCACA +TTCACT +TTCACC +TTCACG +TTCAGA +TTCAGT +TTCAGC +TTCAGG +TTCTAA +TTCTAT +TTCTAC +TTCTAG +TTCTTA +TTCTTT +TTCTTC +TTCTTG +TTCTCA +TTCTCT +TTCTCC +TTCTCG +TTCTGA +TTCTGT +TTCTGC +TTCTGG +TTCCAA +TTCCAT +TTCCAC +TTCCAG +TTCCTA +TTCCTT +TTCCTC +TTCCTG +TTCCCA +TTCCCT +TTCCCC +TTCCCG +TTCCGA +TTCCGT +TTCCGC +TTCCGG +TTCGAA +TTCGAT +TTCGAC +TTCGAG +TTCGTA +TTCGTT +TTCGTC +TTCGTG +TTCGCA +TTCGCT +TTCGCC +TTCGCG +TTCGGA +TTCGGT +TTCGGC +TTCGGG +TTGAAA +TTGAAT +TTGAAC +TTGAAG +TTGATA +TTGATT +TTGATC +TTGATG +TTGACA +TTGACT +TTGACC +TTGACG +TTGAGA +TTGAGT +TTGAGC +TTGAGG +TTGTAA +TTGTAT +TTGTAC +TTGTAG +TTGTTA +TTGTTT +TTGTTC +TTGTTG +TTGTCA +TTGTCT +TTGTCC +TTGTCG +TTGTGA +TTGTGT +TTGTGC +TTGTGG +TTGCAA +TTGCAT +TTGCAC +TTGCAG +TTGCTA +TTGCTT +TTGCTC +TTGCTG +TTGCCA +TTGCCT +TTGCCC +TTGCCG +TTGCGA +TTGCGT +TTGCGC +TTGCGG +TTGGAA +TTGGAT +TTGGAC +TTGGAG +TTGGTA +TTGGTT +TTGGTC +TTGGTG +TTGGCA +TTGGCT +TTGGCC +TTGGCG +TTGGGA +TTGGGT +TTGGGC +TTGGGG +TCAAAA +TCAAAT +TCAAAC +TCAAAG +TCAATA +TCAATT +TCAATC +TCAATG +TCAACA +TCAACT +TCAACC +TCAACG +TCAAGA +TCAAGT +TCAAGC +TCAAGG +TCATAA +TCATAT +TCATAC +TCATAG +TCATTA +TCATTT +TCATTC +TCATTG +TCATCA +TCATCT +TCATCC +TCATCG +TCATGA +TCATGT +TCATGC +TCATGG +TCACAA +TCACAT +TCACAC +TCACAG +TCACTA +TCACTT +TCACTC +TCACTG +TCACCA +TCACCT +TCACCC +TCACCG +TCACGA +TCACGT +TCACGC +TCACGG +TCAGAA +TCAGAT +TCAGAC +TCAGAG +TCAGTA +TCAGTT +TCAGTC +TCAGTG +TCAGCA +TCAGCT +TCAGCC +TCAGCG +TCAGGA +TCAGGT +TCAGGC +TCAGGG +TCTAAA +TCTAAT +TCTAAC +TCTAAG +TCTATA +TCTATT +TCTATC +TCTATG +TCTACA +TCTACT +TCTACC +TCTACG +TCTAGA +TCTAGT +TCTAGC +TCTAGG +TCTTAA +TCTTAT +TCTTAC +TCTTAG +TCTTTA +TCTTTT +TCTTTC +TCTTTG +TCTTCA +TCTTCT +TCTTCC +TCTTCG +TCTTGA +TCTTGT +TCTTGC +TCTTGG +TCTCAA +TCTCAT +TCTCAC +TCTCAG +TCTCTA +TCTCTT +TCTCTC +TCTCTG +TCTCCA +TCTCCT +TCTCCC +TCTCCG +TCTCGA +TCTCGT +TCTCGC +TCTCGG +TCTGAA +TCTGAT +TCTGAC +TCTGAG +TCTGTA +TCTGTT +TCTGTC +TCTGTG +TCTGCA +TCTGCT +TCTGCC +TCTGCG +TCTGGA +TCTGGT +TCTGGC +TCTGGG +TCCAAA +TCCAAT +TCCAAC +TCCAAG +TCCATA +TCCATT +TCCATC +TCCATG +TCCACA +TCCACT +TCCACC +TCCACG +TCCAGA +TCCAGT +TCCAGC +TCCAGG +TCCTAA +TCCTAT +TCCTAC +TCCTAG +TCCTTA +TCCTTT +TCCTTC +TCCTTG +TCCTCA +TCCTCT +TCCTCC +TCCTCG +TCCTGA +TCCTGT +TCCTGC +TCCTGG +TCCCAA +TCCCAT +TCCCAC +TCCCAG +TCCCTA +TCCCTT +TCCCTC +TCCCTG +TCCCCA +TCCCCT +TCCCCC +TCCCCG +TCCCGA +TCCCGT +TCCCGC +TCCCGG +TCCGAA +TCCGAT +TCCGAC +TCCGAG +TCCGTA +TCCGTT +TCCGTC +TCCGTG +TCCGCA +TCCGCT +TCCGCC +TCCGCG +TCCGGA +TCCGGT +TCCGGC +TCCGGG +TCGAAA +TCGAAT +TCGAAC +TCGAAG +TCGATA +TCGATT +TCGATC +TCGATG +TCGACA +TCGACT +TCGACC +TCGACG +TCGAGA +TCGAGT +TCGAGC +TCGAGG +TCGTAA +TCGTAT +TCGTAC +TCGTAG +TCGTTA +TCGTTT +TCGTTC +TCGTTG +TCGTCA +TCGTCT +TCGTCC +TCGTCG +TCGTGA +TCGTGT +TCGTGC +TCGTGG +TCGCAA +TCGCAT +TCGCAC +TCGCAG +TCGCTA +TCGCTT +TCGCTC +TCGCTG +TCGCCA +TCGCCT +TCGCCC +TCGCCG +TCGCGA +TCGCGT +TCGCGC +TCGCGG +TCGGAA +TCGGAT +TCGGAC +TCGGAG +TCGGTA +TCGGTT +TCGGTC +TCGGTG +TCGGCA +TCGGCT +TCGGCC +TCGGCG +TCGGGA +TCGGGT +TCGGGC +TCGGGG +TGAAAA +TGAAAT +TGAAAC +TGAAAG +TGAATA +TGAATT +TGAATC +TGAATG +TGAACA +TGAACT +TGAACC +TGAACG +TGAAGA +TGAAGT +TGAAGC +TGAAGG +TGATAA +TGATAT +TGATAC +TGATAG +TGATTA +TGATTT +TGATTC +TGATTG +TGATCA +TGATCT +TGATCC +TGATCG +TGATGA +TGATGT +TGATGC +TGATGG +TGACAA +TGACAT +TGACAC +TGACAG +TGACTA +TGACTT +TGACTC +TGACTG +TGACCA +TGACCT +TGACCC +TGACCG +TGACGA +TGACGT +TGACGC +TGACGG +TGAGAA +TGAGAT +TGAGAC +TGAGAG +TGAGTA +TGAGTT +TGAGTC +TGAGTG +TGAGCA +TGAGCT +TGAGCC +TGAGCG +TGAGGA +TGAGGT +TGAGGC +TGAGGG +TGTAAA +TGTAAT +TGTAAC +TGTAAG +TGTATA +TGTATT +TGTATC +TGTATG +TGTACA +TGTACT +TGTACC +TGTACG +TGTAGA +TGTAGT +TGTAGC +TGTAGG +TGTTAA +TGTTAT +TGTTAC +TGTTAG +TGTTTA +TGTTTT +TGTTTC +TGTTTG +TGTTCA +TGTTCT +TGTTCC +TGTTCG +TGTTGA +TGTTGT +TGTTGC +TGTTGG +TGTCAA +TGTCAT +TGTCAC +TGTCAG +TGTCTA +TGTCTT +TGTCTC +TGTCTG +TGTCCA +TGTCCT +TGTCCC +TGTCCG +TGTCGA +TGTCGT +TGTCGC +TGTCGG +TGTGAA +TGTGAT +TGTGAC +TGTGAG +TGTGTA +TGTGTT +TGTGTC +TGTGTG +TGTGCA +TGTGCT +TGTGCC +TGTGCG +TGTGGA +TGTGGT +TGTGGC +TGTGGG +TGCAAA +TGCAAT +TGCAAC +TGCAAG +TGCATA +TGCATT +TGCATC +TGCATG +TGCACA +TGCACT +TGCACC +TGCACG +TGCAGA +TGCAGT +TGCAGC +TGCAGG +TGCTAA +TGCTAT +TGCTAC +TGCTAG +TGCTTA +TGCTTT +TGCTTC +TGCTTG +TGCTCA +TGCTCT +TGCTCC +TGCTCG +TGCTGA +TGCTGT +TGCTGC +TGCTGG +TGCCAA +TGCCAT +TGCCAC +TGCCAG +TGCCTA +TGCCTT +TGCCTC +TGCCTG +TGCCCA +TGCCCT +TGCCCC +TGCCCG +TGCCGA +TGCCGT +TGCCGC +TGCCGG +TGCGAA +TGCGAT +TGCGAC +TGCGAG +TGCGTA +TGCGTT +TGCGTC +TGCGTG +TGCGCA +TGCGCT +TGCGCC +TGCGCG +TGCGGA +TGCGGT +TGCGGC +TGCGGG +TGGAAA +TGGAAT +TGGAAC +TGGAAG +TGGATA +TGGATT +TGGATC +TGGATG +TGGACA +TGGACT +TGGACC +TGGACG +TGGAGA +TGGAGT +TGGAGC +TGGAGG +TGGTAA +TGGTAT +TGGTAC +TGGTAG +TGGTTA +TGGTTT +TGGTTC +TGGTTG +TGGTCA +TGGTCT +TGGTCC +TGGTCG +TGGTGA +TGGTGT +TGGTGC +TGGTGG +TGGCAA +TGGCAT +TGGCAC +TGGCAG +TGGCTA +TGGCTT +TGGCTC +TGGCTG +TGGCCA +TGGCCT +TGGCCC +TGGCCG +TGGCGA +TGGCGT +TGGCGC +TGGCGG +TGGGAA +TGGGAT +TGGGAC +TGGGAG +TGGGTA +TGGGTT +TGGGTC +TGGGTG +TGGGCA +TGGGCT +TGGGCC +TGGGCG +TGGGGA +TGGGGT +TGGGGC +TGGGGG +CAAAAA +CAAAAT +CAAAAC +CAAAAG +CAAATA +CAAATT +CAAATC +CAAATG +CAAACA +CAAACT +CAAACC +CAAACG +CAAAGA +CAAAGT +CAAAGC +CAAAGG +CAATAA +CAATAT +CAATAC +CAATAG +CAATTA +CAATTT +CAATTC +CAATTG +CAATCA +CAATCT +CAATCC +CAATCG +CAATGA +CAATGT +CAATGC +CAATGG +CAACAA +CAACAT +CAACAC +CAACAG +CAACTA +CAACTT +CAACTC +CAACTG +CAACCA +CAACCT +CAACCC +CAACCG +CAACGA +CAACGT +CAACGC +CAACGG +CAAGAA +CAAGAT +CAAGAC +CAAGAG +CAAGTA +CAAGTT +CAAGTC +CAAGTG +CAAGCA +CAAGCT +CAAGCC +CAAGCG +CAAGGA +CAAGGT +CAAGGC +CAAGGG +CATAAA +CATAAT +CATAAC +CATAAG +CATATA +CATATT +CATATC +CATATG +CATACA +CATACT +CATACC +CATACG +CATAGA +CATAGT +CATAGC +CATAGG +CATTAA +CATTAT +CATTAC +CATTAG +CATTTA +CATTTT +CATTTC +CATTTG +CATTCA +CATTCT +CATTCC +CATTCG +CATTGA +CATTGT +CATTGC +CATTGG +CATCAA +CATCAT +CATCAC +CATCAG +CATCTA +CATCTT +CATCTC +CATCTG +CATCCA +CATCCT +CATCCC +CATCCG +CATCGA +CATCGT +CATCGC +CATCGG +CATGAA +CATGAT +CATGAC +CATGAG +CATGTA +CATGTT +CATGTC +CATGTG +CATGCA +CATGCT +CATGCC +CATGCG +CATGGA +CATGGT +CATGGC +CATGGG +CACAAA +CACAAT +CACAAC +CACAAG +CACATA +CACATT +CACATC +CACATG +CACACA +CACACT +CACACC +CACACG +CACAGA +CACAGT +CACAGC +CACAGG +CACTAA +CACTAT +CACTAC +CACTAG +CACTTA +CACTTT +CACTTC +CACTTG +CACTCA +CACTCT +CACTCC +CACTCG +CACTGA +CACTGT +CACTGC +CACTGG +CACCAA +CACCAT +CACCAC +CACCAG +CACCTA +CACCTT +CACCTC +CACCTG +CACCCA +CACCCT +CACCCC +CACCCG +CACCGA +CACCGT +CACCGC +CACCGG +CACGAA +CACGAT +CACGAC +CACGAG +CACGTA +CACGTT +CACGTC +CACGTG +CACGCA +CACGCT +CACGCC +CACGCG +CACGGA +CACGGT +CACGGC +CACGGG +CAGAAA +CAGAAT +CAGAAC +CAGAAG +CAGATA +CAGATT +CAGATC +CAGATG +CAGACA +CAGACT +CAGACC +CAGACG +CAGAGA +CAGAGT +CAGAGC +CAGAGG +CAGTAA +CAGTAT +CAGTAC +CAGTAG +CAGTTA +CAGTTT +CAGTTC +CAGTTG +CAGTCA +CAGTCT +CAGTCC +CAGTCG +CAGTGA +CAGTGT +CAGTGC +CAGTGG +CAGCAA +CAGCAT +CAGCAC +CAGCAG +CAGCTA +CAGCTT +CAGCTC +CAGCTG +CAGCCA +CAGCCT +CAGCCC +CAGCCG +CAGCGA +CAGCGT +CAGCGC +CAGCGG +CAGGAA +CAGGAT +CAGGAC +CAGGAG +CAGGTA +CAGGTT +CAGGTC +CAGGTG +CAGGCA +CAGGCT +CAGGCC +CAGGCG +CAGGGA +CAGGGT +CAGGGC +CAGGGG +CTAAAA +CTAAAT +CTAAAC +CTAAAG +CTAATA +CTAATT +CTAATC +CTAATG +CTAACA +CTAACT +CTAACC +CTAACG +CTAAGA +CTAAGT +CTAAGC +CTAAGG +CTATAA +CTATAT +CTATAC +CTATAG +CTATTA +CTATTT +CTATTC +CTATTG +CTATCA +CTATCT +CTATCC +CTATCG +CTATGA +CTATGT +CTATGC +CTATGG +CTACAA +CTACAT +CTACAC +CTACAG +CTACTA +CTACTT +CTACTC +CTACTG +CTACCA +CTACCT +CTACCC +CTACCG +CTACGA +CTACGT +CTACGC +CTACGG +CTAGAA +CTAGAT +CTAGAC +CTAGAG +CTAGTA +CTAGTT +CTAGTC +CTAGTG +CTAGCA +CTAGCT +CTAGCC +CTAGCG +CTAGGA +CTAGGT +CTAGGC +CTAGGG +CTTAAA +CTTAAT +CTTAAC +CTTAAG +CTTATA +CTTATT +CTTATC +CTTATG +CTTACA +CTTACT +CTTACC +CTTACG +CTTAGA +CTTAGT +CTTAGC +CTTAGG +CTTTAA +CTTTAT +CTTTAC +CTTTAG +CTTTTA +CTTTTT +CTTTTC +CTTTTG +CTTTCA +CTTTCT +CTTTCC +CTTTCG +CTTTGA +CTTTGT +CTTTGC +CTTTGG +CTTCAA +CTTCAT +CTTCAC +CTTCAG +CTTCTA +CTTCTT +CTTCTC +CTTCTG +CTTCCA +CTTCCT +CTTCCC +CTTCCG +CTTCGA +CTTCGT +CTTCGC +CTTCGG +CTTGAA +CTTGAT +CTTGAC +CTTGAG +CTTGTA +CTTGTT +CTTGTC +CTTGTG +CTTGCA +CTTGCT +CTTGCC +CTTGCG +CTTGGA +CTTGGT +CTTGGC +CTTGGG +CTCAAA +CTCAAT +CTCAAC +CTCAAG +CTCATA +CTCATT +CTCATC +CTCATG +CTCACA +CTCACT +CTCACC +CTCACG +CTCAGA +CTCAGT +CTCAGC +CTCAGG +CTCTAA +CTCTAT +CTCTAC +CTCTAG +CTCTTA +CTCTTT +CTCTTC +CTCTTG +CTCTCA +CTCTCT +CTCTCC +CTCTCG +CTCTGA +CTCTGT +CTCTGC +CTCTGG +CTCCAA +CTCCAT +CTCCAC +CTCCAG +CTCCTA +CTCCTT +CTCCTC +CTCCTG +CTCCCA +CTCCCT +CTCCCC +CTCCCG +CTCCGA +CTCCGT +CTCCGC +CTCCGG +CTCGAA +CTCGAT +CTCGAC +CTCGAG +CTCGTA +CTCGTT +CTCGTC +CTCGTG +CTCGCA +CTCGCT +CTCGCC +CTCGCG +CTCGGA +CTCGGT +CTCGGC +CTCGGG +CTGAAA +CTGAAT +CTGAAC +CTGAAG +CTGATA +CTGATT +CTGATC +CTGATG +CTGACA +CTGACT +CTGACC +CTGACG +CTGAGA +CTGAGT +CTGAGC +CTGAGG +CTGTAA +CTGTAT +CTGTAC +CTGTAG +CTGTTA +CTGTTT +CTGTTC +CTGTTG +CTGTCA +CTGTCT +CTGTCC +CTGTCG +CTGTGA +CTGTGT +CTGTGC +CTGTGG +CTGCAA +CTGCAT +CTGCAC +CTGCAG +CTGCTA +CTGCTT +CTGCTC +CTGCTG +CTGCCA +CTGCCT +CTGCCC +CTGCCG +CTGCGA +CTGCGT +CTGCGC +CTGCGG +CTGGAA +CTGGAT +CTGGAC +CTGGAG +CTGGTA +CTGGTT +CTGGTC +CTGGTG +CTGGCA +CTGGCT +CTGGCC +CTGGCG +CTGGGA +CTGGGT +CTGGGC +CTGGGG +CCAAAA +CCAAAT +CCAAAC +CCAAAG +CCAATA +CCAATT +CCAATC +CCAATG +CCAACA +CCAACT +CCAACC +CCAACG +CCAAGA +CCAAGT +CCAAGC +CCAAGG +CCATAA +CCATAT +CCATAC +CCATAG +CCATTA +CCATTT +CCATTC +CCATTG +CCATCA +CCATCT +CCATCC +CCATCG +CCATGA +CCATGT +CCATGC +CCATGG +CCACAA +CCACAT +CCACAC +CCACAG +CCACTA +CCACTT +CCACTC +CCACTG +CCACCA +CCACCT +CCACCC +CCACCG +CCACGA +CCACGT +CCACGC +CCACGG +CCAGAA +CCAGAT +CCAGAC +CCAGAG +CCAGTA +CCAGTT +CCAGTC +CCAGTG +CCAGCA +CCAGCT +CCAGCC +CCAGCG +CCAGGA +CCAGGT +CCAGGC +CCAGGG +CCTAAA +CCTAAT +CCTAAC +CCTAAG +CCTATA +CCTATT +CCTATC +CCTATG +CCTACA +CCTACT +CCTACC +CCTACG +CCTAGA +CCTAGT +CCTAGC +CCTAGG +CCTTAA +CCTTAT +CCTTAC +CCTTAG +CCTTTA +CCTTTT +CCTTTC +CCTTTG +CCTTCA +CCTTCT +CCTTCC +CCTTCG +CCTTGA +CCTTGT +CCTTGC +CCTTGG +CCTCAA +CCTCAT +CCTCAC +CCTCAG +CCTCTA +CCTCTT +CCTCTC +CCTCTG +CCTCCA +CCTCCT +CCTCCC +CCTCCG +CCTCGA +CCTCGT +CCTCGC +CCTCGG +CCTGAA +CCTGAT +CCTGAC +CCTGAG +CCTGTA +CCTGTT +CCTGTC +CCTGTG +CCTGCA +CCTGCT +CCTGCC +CCTGCG +CCTGGA +CCTGGT +CCTGGC +CCTGGG +CCCAAA +CCCAAT +CCCAAC +CCCAAG +CCCATA +CCCATT +CCCATC +CCCATG +CCCACA +CCCACT +CCCACC +CCCACG +CCCAGA +CCCAGT +CCCAGC +CCCAGG +CCCTAA +CCCTAT +CCCTAC +CCCTAG +CCCTTA +CCCTTT +CCCTTC +CCCTTG +CCCTCA +CCCTCT +CCCTCC +CCCTCG +CCCTGA +CCCTGT +CCCTGC +CCCTGG +CCCCAA +CCCCAT +CCCCAC +CCCCAG +CCCCTA +CCCCTT +CCCCTC +CCCCTG +CCCCCA +CCCCCT +CCCCCC +CCCCCG +CCCCGA +CCCCGT +CCCCGC +CCCCGG +CCCGAA +CCCGAT +CCCGAC +CCCGAG +CCCGTA +CCCGTT +CCCGTC +CCCGTG +CCCGCA +CCCGCT +CCCGCC +CCCGCG +CCCGGA +CCCGGT +CCCGGC +CCCGGG +CCGAAA +CCGAAT +CCGAAC +CCGAAG +CCGATA +CCGATT +CCGATC +CCGATG +CCGACA +CCGACT +CCGACC +CCGACG +CCGAGA +CCGAGT +CCGAGC +CCGAGG +CCGTAA +CCGTAT +CCGTAC +CCGTAG +CCGTTA +CCGTTT +CCGTTC +CCGTTG +CCGTCA +CCGTCT +CCGTCC +CCGTCG +CCGTGA +CCGTGT +CCGTGC +CCGTGG +CCGCAA +CCGCAT +CCGCAC +CCGCAG +CCGCTA +CCGCTT +CCGCTC +CCGCTG +CCGCCA +CCGCCT +CCGCCC +CCGCCG +CCGCGA +CCGCGT +CCGCGC +CCGCGG +CCGGAA +CCGGAT +CCGGAC +CCGGAG +CCGGTA +CCGGTT +CCGGTC +CCGGTG +CCGGCA +CCGGCT +CCGGCC +CCGGCG +CCGGGA +CCGGGT +CCGGGC +CCGGGG +CGAAAA +CGAAAT +CGAAAC +CGAAAG +CGAATA +CGAATT +CGAATC +CGAATG +CGAACA +CGAACT +CGAACC +CGAACG +CGAAGA +CGAAGT +CGAAGC +CGAAGG +CGATAA +CGATAT +CGATAC +CGATAG +CGATTA +CGATTT +CGATTC +CGATTG +CGATCA +CGATCT +CGATCC +CGATCG +CGATGA +CGATGT +CGATGC +CGATGG +CGACAA +CGACAT +CGACAC +CGACAG +CGACTA +CGACTT +CGACTC +CGACTG +CGACCA +CGACCT +CGACCC +CGACCG +CGACGA +CGACGT +CGACGC +CGACGG +CGAGAA +CGAGAT +CGAGAC +CGAGAG +CGAGTA +CGAGTT +CGAGTC +CGAGTG +CGAGCA +CGAGCT +CGAGCC +CGAGCG +CGAGGA +CGAGGT +CGAGGC +CGAGGG +CGTAAA +CGTAAT +CGTAAC +CGTAAG +CGTATA +CGTATT +CGTATC +CGTATG +CGTACA +CGTACT +CGTACC +CGTACG +CGTAGA +CGTAGT +CGTAGC +CGTAGG +CGTTAA +CGTTAT +CGTTAC +CGTTAG +CGTTTA +CGTTTT +CGTTTC +CGTTTG +CGTTCA +CGTTCT +CGTTCC +CGTTCG +CGTTGA +CGTTGT +CGTTGC +CGTTGG +CGTCAA +CGTCAT +CGTCAC +CGTCAG +CGTCTA +CGTCTT +CGTCTC +CGTCTG +CGTCCA +CGTCCT +CGTCCC +CGTCCG +CGTCGA +CGTCGT +CGTCGC +CGTCGG +CGTGAA +CGTGAT +CGTGAC +CGTGAG +CGTGTA +CGTGTT +CGTGTC +CGTGTG +CGTGCA +CGTGCT +CGTGCC +CGTGCG +CGTGGA +CGTGGT +CGTGGC +CGTGGG +CGCAAA +CGCAAT +CGCAAC +CGCAAG +CGCATA +CGCATT +CGCATC +CGCATG +CGCACA +CGCACT +CGCACC +CGCACG +CGCAGA +CGCAGT +CGCAGC +CGCAGG +CGCTAA +CGCTAT +CGCTAC +CGCTAG +CGCTTA +CGCTTT +CGCTTC +CGCTTG +CGCTCA +CGCTCT +CGCTCC +CGCTCG +CGCTGA +CGCTGT +CGCTGC +CGCTGG +CGCCAA +CGCCAT +CGCCAC +CGCCAG +CGCCTA +CGCCTT +CGCCTC +CGCCTG +CGCCCA +CGCCCT +CGCCCC +CGCCCG +CGCCGA +CGCCGT +CGCCGC +CGCCGG +CGCGAA +CGCGAT +CGCGAC +CGCGAG +CGCGTA +CGCGTT +CGCGTC +CGCGTG +CGCGCA +CGCGCT +CGCGCC +CGCGCG +CGCGGA +CGCGGT +CGCGGC +CGCGGG +CGGAAA +CGGAAT +CGGAAC +CGGAAG +CGGATA +CGGATT +CGGATC +CGGATG +CGGACA +CGGACT +CGGACC +CGGACG +CGGAGA +CGGAGT +CGGAGC +CGGAGG +CGGTAA +CGGTAT +CGGTAC +CGGTAG +CGGTTA +CGGTTT +CGGTTC +CGGTTG +CGGTCA +CGGTCT +CGGTCC +CGGTCG +CGGTGA +CGGTGT +CGGTGC +CGGTGG +CGGCAA +CGGCAT +CGGCAC +CGGCAG +CGGCTA +CGGCTT +CGGCTC +CGGCTG +CGGCCA +CGGCCT +CGGCCC +CGGCCG +CGGCGA +CGGCGT +CGGCGC +CGGCGG +CGGGAA +CGGGAT +CGGGAC +CGGGAG +CGGGTA +CGGGTT +CGGGTC +CGGGTG +CGGGCA +CGGGCT +CGGGCC +CGGGCG +CGGGGA +CGGGGT +CGGGGC +CGGGGG +GAAAAA +GAAAAT +GAAAAC +GAAAAG +GAAATA +GAAATT +GAAATC +GAAATG +GAAACA +GAAACT +GAAACC +GAAACG +GAAAGA +GAAAGT +GAAAGC +GAAAGG +GAATAA +GAATAT +GAATAC +GAATAG +GAATTA +GAATTT +GAATTC +GAATTG +GAATCA +GAATCT +GAATCC +GAATCG +GAATGA +GAATGT +GAATGC +GAATGG +GAACAA +GAACAT +GAACAC +GAACAG +GAACTA +GAACTT +GAACTC +GAACTG +GAACCA +GAACCT +GAACCC +GAACCG +GAACGA +GAACGT +GAACGC +GAACGG +GAAGAA +GAAGAT +GAAGAC +GAAGAG +GAAGTA +GAAGTT +GAAGTC +GAAGTG +GAAGCA +GAAGCT +GAAGCC +GAAGCG +GAAGGA +GAAGGT +GAAGGC +GAAGGG +GATAAA +GATAAT +GATAAC +GATAAG +GATATA +GATATT +GATATC +GATATG +GATACA +GATACT +GATACC +GATACG +GATAGA +GATAGT +GATAGC +GATAGG +GATTAA +GATTAT +GATTAC +GATTAG +GATTTA +GATTTT +GATTTC +GATTTG +GATTCA +GATTCT +GATTCC +GATTCG +GATTGA +GATTGT +GATTGC +GATTGG +GATCAA +GATCAT +GATCAC +GATCAG +GATCTA +GATCTT +GATCTC +GATCTG +GATCCA +GATCCT +GATCCC +GATCCG +GATCGA +GATCGT +GATCGC +GATCGG +GATGAA +GATGAT +GATGAC +GATGAG +GATGTA +GATGTT +GATGTC +GATGTG +GATGCA +GATGCT +GATGCC +GATGCG +GATGGA +GATGGT +GATGGC +GATGGG +GACAAA +GACAAT +GACAAC +GACAAG +GACATA +GACATT +GACATC +GACATG +GACACA +GACACT +GACACC +GACACG +GACAGA +GACAGT +GACAGC +GACAGG +GACTAA +GACTAT +GACTAC +GACTAG +GACTTA +GACTTT +GACTTC +GACTTG +GACTCA +GACTCT +GACTCC +GACTCG +GACTGA +GACTGT +GACTGC +GACTGG +GACCAA +GACCAT +GACCAC +GACCAG +GACCTA +GACCTT +GACCTC +GACCTG +GACCCA +GACCCT +GACCCC +GACCCG +GACCGA +GACCGT +GACCGC +GACCGG +GACGAA +GACGAT +GACGAC +GACGAG +GACGTA +GACGTT +GACGTC +GACGTG +GACGCA +GACGCT +GACGCC +GACGCG +GACGGA +GACGGT +GACGGC +GACGGG +GAGAAA +GAGAAT +GAGAAC +GAGAAG +GAGATA +GAGATT +GAGATC +GAGATG +GAGACA +GAGACT +GAGACC +GAGACG +GAGAGA +GAGAGT +GAGAGC +GAGAGG +GAGTAA +GAGTAT +GAGTAC +GAGTAG +GAGTTA +GAGTTT +GAGTTC +GAGTTG +GAGTCA +GAGTCT +GAGTCC +GAGTCG +GAGTGA +GAGTGT +GAGTGC +GAGTGG +GAGCAA +GAGCAT +GAGCAC +GAGCAG +GAGCTA +GAGCTT +GAGCTC +GAGCTG +GAGCCA +GAGCCT +GAGCCC +GAGCCG +GAGCGA +GAGCGT +GAGCGC +GAGCGG +GAGGAA +GAGGAT +GAGGAC +GAGGAG +GAGGTA +GAGGTT +GAGGTC +GAGGTG +GAGGCA +GAGGCT +GAGGCC +GAGGCG +GAGGGA +GAGGGT +GAGGGC +GAGGGG +GTAAAA +GTAAAT +GTAAAC +GTAAAG +GTAATA +GTAATT +GTAATC +GTAATG +GTAACA +GTAACT +GTAACC +GTAACG +GTAAGA +GTAAGT +GTAAGC +GTAAGG +GTATAA +GTATAT +GTATAC +GTATAG +GTATTA +GTATTT +GTATTC +GTATTG +GTATCA +GTATCT +GTATCC +GTATCG +GTATGA +GTATGT +GTATGC +GTATGG +GTACAA +GTACAT +GTACAC +GTACAG +GTACTA +GTACTT +GTACTC +GTACTG +GTACCA +GTACCT +GTACCC +GTACCG +GTACGA +GTACGT +GTACGC +GTACGG +GTAGAA +GTAGAT +GTAGAC +GTAGAG +GTAGTA +GTAGTT +GTAGTC +GTAGTG +GTAGCA +GTAGCT +GTAGCC +GTAGCG +GTAGGA +GTAGGT +GTAGGC +GTAGGG +GTTAAA +GTTAAT +GTTAAC +GTTAAG +GTTATA +GTTATT +GTTATC +GTTATG +GTTACA +GTTACT +GTTACC +GTTACG +GTTAGA +GTTAGT +GTTAGC +GTTAGG +GTTTAA +GTTTAT +GTTTAC +GTTTAG +GTTTTA +GTTTTT +GTTTTC +GTTTTG +GTTTCA +GTTTCT +GTTTCC +GTTTCG +GTTTGA +GTTTGT +GTTTGC +GTTTGG +GTTCAA +GTTCAT +GTTCAC +GTTCAG +GTTCTA +GTTCTT +GTTCTC +GTTCTG +GTTCCA +GTTCCT +GTTCCC +GTTCCG +GTTCGA +GTTCGT +GTTCGC +GTTCGG +GTTGAA +GTTGAT +GTTGAC +GTTGAG +GTTGTA +GTTGTT +GTTGTC +GTTGTG +GTTGCA +GTTGCT +GTTGCC +GTTGCG +GTTGGA +GTTGGT +GTTGGC +GTTGGG +GTCAAA +GTCAAT +GTCAAC +GTCAAG +GTCATA +GTCATT +GTCATC +GTCATG +GTCACA +GTCACT +GTCACC +GTCACG +GTCAGA +GTCAGT +GTCAGC +GTCAGG +GTCTAA +GTCTAT +GTCTAC +GTCTAG +GTCTTA +GTCTTT +GTCTTC +GTCTTG +GTCTCA +GTCTCT +GTCTCC +GTCTCG +GTCTGA +GTCTGT +GTCTGC +GTCTGG +GTCCAA +GTCCAT +GTCCAC +GTCCAG +GTCCTA +GTCCTT +GTCCTC +GTCCTG +GTCCCA +GTCCCT +GTCCCC +GTCCCG +GTCCGA +GTCCGT +GTCCGC +GTCCGG +GTCGAA +GTCGAT +GTCGAC +GTCGAG +GTCGTA +GTCGTT +GTCGTC +GTCGTG +GTCGCA +GTCGCT +GTCGCC +GTCGCG +GTCGGA +GTCGGT +GTCGGC +GTCGGG +GTGAAA +GTGAAT +GTGAAC +GTGAAG +GTGATA +GTGATT +GTGATC +GTGATG +GTGACA +GTGACT +GTGACC +GTGACG +GTGAGA +GTGAGT +GTGAGC +GTGAGG +GTGTAA +GTGTAT +GTGTAC +GTGTAG +GTGTTA +GTGTTT +GTGTTC +GTGTTG +GTGTCA +GTGTCT +GTGTCC +GTGTCG +GTGTGA +GTGTGT +GTGTGC +GTGTGG +GTGCAA +GTGCAT +GTGCAC +GTGCAG +GTGCTA +GTGCTT +GTGCTC +GTGCTG +GTGCCA +GTGCCT +GTGCCC +GTGCCG +GTGCGA +GTGCGT +GTGCGC +GTGCGG +GTGGAA +GTGGAT +GTGGAC +GTGGAG +GTGGTA +GTGGTT +GTGGTC +GTGGTG +GTGGCA +GTGGCT +GTGGCC +GTGGCG +GTGGGA +GTGGGT +GTGGGC +GTGGGG +GCAAAA +GCAAAT +GCAAAC +GCAAAG +GCAATA +GCAATT +GCAATC +GCAATG +GCAACA +GCAACT +GCAACC +GCAACG +GCAAGA +GCAAGT +GCAAGC +GCAAGG +GCATAA +GCATAT +GCATAC +GCATAG +GCATTA +GCATTT +GCATTC +GCATTG +GCATCA +GCATCT +GCATCC +GCATCG +GCATGA +GCATGT +GCATGC +GCATGG +GCACAA +GCACAT +GCACAC +GCACAG +GCACTA +GCACTT +GCACTC +GCACTG +GCACCA +GCACCT +GCACCC +GCACCG +GCACGA +GCACGT +GCACGC +GCACGG +GCAGAA +GCAGAT +GCAGAC +GCAGAG +GCAGTA +GCAGTT +GCAGTC +GCAGTG +GCAGCA +GCAGCT +GCAGCC +GCAGCG +GCAGGA +GCAGGT +GCAGGC +GCAGGG +GCTAAA +GCTAAT +GCTAAC +GCTAAG +GCTATA +GCTATT +GCTATC +GCTATG +GCTACA +GCTACT +GCTACC +GCTACG +GCTAGA +GCTAGT +GCTAGC +GCTAGG +GCTTAA +GCTTAT +GCTTAC +GCTTAG +GCTTTA +GCTTTT +GCTTTC +GCTTTG +GCTTCA +GCTTCT +GCTTCC +GCTTCG +GCTTGA +GCTTGT +GCTTGC +GCTTGG +GCTCAA +GCTCAT +GCTCAC +GCTCAG +GCTCTA +GCTCTT +GCTCTC +GCTCTG +GCTCCA +GCTCCT +GCTCCC +GCTCCG +GCTCGA +GCTCGT +GCTCGC +GCTCGG +GCTGAA +GCTGAT +GCTGAC +GCTGAG +GCTGTA +GCTGTT +GCTGTC +GCTGTG +GCTGCA +GCTGCT +GCTGCC +GCTGCG +GCTGGA +GCTGGT +GCTGGC +GCTGGG +GCCAAA +GCCAAT +GCCAAC +GCCAAG +GCCATA +GCCATT +GCCATC +GCCATG +GCCACA +GCCACT +GCCACC +GCCACG +GCCAGA +GCCAGT +GCCAGC +GCCAGG +GCCTAA +GCCTAT +GCCTAC +GCCTAG +GCCTTA +GCCTTT +GCCTTC +GCCTTG +GCCTCA +GCCTCT +GCCTCC +GCCTCG +GCCTGA +GCCTGT +GCCTGC +GCCTGG +GCCCAA +GCCCAT +GCCCAC +GCCCAG +GCCCTA +GCCCTT +GCCCTC +GCCCTG +GCCCCA +GCCCCT +GCCCCC +GCCCCG +GCCCGA +GCCCGT +GCCCGC +GCCCGG +GCCGAA +GCCGAT +GCCGAC +GCCGAG +GCCGTA +GCCGTT +GCCGTC +GCCGTG +GCCGCA +GCCGCT +GCCGCC +GCCGCG +GCCGGA +GCCGGT +GCCGGC +GCCGGG +GCGAAA +GCGAAT +GCGAAC +GCGAAG +GCGATA +GCGATT +GCGATC +GCGATG +GCGACA +GCGACT +GCGACC +GCGACG +GCGAGA +GCGAGT +GCGAGC +GCGAGG +GCGTAA +GCGTAT +GCGTAC +GCGTAG +GCGTTA +GCGTTT +GCGTTC +GCGTTG +GCGTCA +GCGTCT +GCGTCC +GCGTCG +GCGTGA +GCGTGT +GCGTGC +GCGTGG +GCGCAA +GCGCAT +GCGCAC +GCGCAG +GCGCTA +GCGCTT +GCGCTC +GCGCTG +GCGCCA +GCGCCT +GCGCCC +GCGCCG +GCGCGA +GCGCGT +GCGCGC +GCGCGG +GCGGAA +GCGGAT +GCGGAC +GCGGAG +GCGGTA +GCGGTT +GCGGTC +GCGGTG +GCGGCA +GCGGCT +GCGGCC +GCGGCG +GCGGGA +GCGGGT +GCGGGC +GCGGGG +GGAAAA +GGAAAT +GGAAAC +GGAAAG +GGAATA +GGAATT +GGAATC +GGAATG +GGAACA +GGAACT +GGAACC +GGAACG +GGAAGA +GGAAGT +GGAAGC +GGAAGG +GGATAA +GGATAT +GGATAC +GGATAG +GGATTA +GGATTT +GGATTC +GGATTG +GGATCA +GGATCT +GGATCC +GGATCG +GGATGA +GGATGT +GGATGC +GGATGG +GGACAA +GGACAT +GGACAC +GGACAG +GGACTA +GGACTT +GGACTC +GGACTG +GGACCA +GGACCT +GGACCC +GGACCG +GGACGA +GGACGT +GGACGC +GGACGG +GGAGAA +GGAGAT +GGAGAC +GGAGAG +GGAGTA +GGAGTT +GGAGTC +GGAGTG +GGAGCA +GGAGCT +GGAGCC +GGAGCG +GGAGGA +GGAGGT +GGAGGC +GGAGGG +GGTAAA +GGTAAT +GGTAAC +GGTAAG +GGTATA +GGTATT +GGTATC +GGTATG +GGTACA +GGTACT +GGTACC +GGTACG +GGTAGA +GGTAGT +GGTAGC +GGTAGG +GGTTAA +GGTTAT +GGTTAC +GGTTAG +GGTTTA +GGTTTT +GGTTTC +GGTTTG +GGTTCA +GGTTCT +GGTTCC +GGTTCG +GGTTGA +GGTTGT +GGTTGC +GGTTGG +GGTCAA +GGTCAT +GGTCAC +GGTCAG +GGTCTA +GGTCTT +GGTCTC +GGTCTG +GGTCCA +GGTCCT +GGTCCC +GGTCCG +GGTCGA +GGTCGT +GGTCGC +GGTCGG +GGTGAA +GGTGAT +GGTGAC +GGTGAG +GGTGTA +GGTGTT +GGTGTC +GGTGTG +GGTGCA +GGTGCT +GGTGCC +GGTGCG +GGTGGA +GGTGGT +GGTGGC +GGTGGG +GGCAAA +GGCAAT +GGCAAC +GGCAAG +GGCATA +GGCATT +GGCATC +GGCATG +GGCACA +GGCACT +GGCACC +GGCACG +GGCAGA +GGCAGT +GGCAGC +GGCAGG +GGCTAA +GGCTAT +GGCTAC +GGCTAG +GGCTTA +GGCTTT +GGCTTC +GGCTTG +GGCTCA +GGCTCT +GGCTCC +GGCTCG +GGCTGA +GGCTGT +GGCTGC +GGCTGG +GGCCAA +GGCCAT +GGCCAC +GGCCAG +GGCCTA +GGCCTT +GGCCTC +GGCCTG +GGCCCA +GGCCCT +GGCCCC +GGCCCG +GGCCGA +GGCCGT +GGCCGC +GGCCGG +GGCGAA +GGCGAT +GGCGAC +GGCGAG +GGCGTA +GGCGTT +GGCGTC +GGCGTG +GGCGCA +GGCGCT +GGCGCC +GGCGCG +GGCGGA +GGCGGT +GGCGGC +GGCGGG +GGGAAA +GGGAAT +GGGAAC +GGGAAG +GGGATA +GGGATT +GGGATC +GGGATG +GGGACA +GGGACT +GGGACC +GGGACG +GGGAGA +GGGAGT +GGGAGC +GGGAGG +GGGTAA +GGGTAT +GGGTAC +GGGTAG +GGGTTA +GGGTTT +GGGTTC +GGGTTG +GGGTCA +GGGTCT +GGGTCC +GGGTCG +GGGTGA +GGGTGT +GGGTGC +GGGTGG +GGGCAA +GGGCAT +GGGCAC +GGGCAG +GGGCTA +GGGCTT +GGGCTC +GGGCTG +GGGCCA +GGGCCT +GGGCCC +GGGCCG +GGGCGA +GGGCGT +GGGCGC +GGGCGG +GGGGAA +GGGGAT +GGGGAC +GGGGAG +GGGGTA +GGGGTT +GGGGTC +GGGGTG +GGGGCA +GGGGCT +GGGGCC +GGGGCG +GGGGGA +GGGGGT +GGGGGC +GGGGGG diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/README.md b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..500fd976d92b217455f181bc98c77e612d9a18d9 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/README.md @@ -0,0 +1,202 @@ +--- +base_model: fabihamakhdoomi/TinyDNABERT +library_name: peft +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.12.0 \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/adapter_config.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..04c63e99bbc6b810449dece367c0afb165125b92 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/adapter_config.json @@ -0,0 +1,42 @@ +{ + "alpha_pattern": {}, + "auto_mapping": null, + "base_model_name_or_path": "fabihamakhdoomi/TinyDNABERT", + "bias": "none", + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 16, + "lora_dropout": 0.2, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": [ + "classifier", + "score" + ], + "peft_type": "LORA", + "r": 8, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "bert.encoder.layer.1.attention.self.query", + "bert.encoder.layer.0.output.dense", + "bert.encoder.layer.1.output.dense", + "bert.encoder.layer.0.intermediate.dense", + "bert.encoder.layer.1.intermediate.dense", + "bert.encoder.layer.0.attention.output.dense", + "bert.encoder.layer.1.attention.output.dense", + "bert.encoder.layer.0.attention.self.query", + "bert.encoder.layer.0.attention.self.value", + "bert.encoder.layer.1.attention.self.key", + "bert.encoder.layer.1.attention.self.value", + "bert.encoder.layer.0.attention.self.key" + ], + "task_type": "SEQ_CLS", + "use_dora": false, + "use_rslora": false +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/adapter_model.safetensors b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/adapter_model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..6eb94619eb78b3ab06f91426c756498b3b0358e0 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/adapter_model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1378bc9e9a7a4961257e0fa628108afbd64adc255c6b9d0657f7b85a508abd92 +size 151952 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/optimizer.pt b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..35d9af3213dd32b1cda67f52c06156479d1e42cd --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:651699c326cfdb8505691a5904ece6dc9b86804fd0425a3c6c96a7f0821ab64f +size 312682 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/rng_state.pth b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..7d59ecb6c9b7d6ec50b060f3843e47cd8ba9b667 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:84106ec8b7c96df034e207cd31f8dd32759188fe95dd21501975d40030684bf4 +size 14244 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/scheduler.pt b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..f3ebc22eebe7fac9f20742118527e91760861e6d --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7e153e480933b4bdcc7d98226c2fdfde72db931809685729c8748251ae3ecae9 +size 1064 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/special_tokens_map.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..a8b3208c2884c4efb86e49300fdd3dc877220cdf --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/special_tokens_map.json @@ -0,0 +1,7 @@ +{ + "cls_token": "[CLS]", + "mask_token": "[MASK]", + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/tokenizer.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..ebd782de6963b9032ce9374cfe30f5275be1a30b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/tokenizer.json @@ -0,0 +1,4265 @@ +{ + "version": "1.0", + "truncation": { + "direction": "Right", + "max_length": 512, + "strategy": "LongestFirst", + "stride": 0 + }, + "padding": { + "strategy": { + "Fixed": 512 + }, + "direction": "Right", + "pad_to_multiple_of": null, + "pad_id": 0, + "pad_type_id": 0, + "pad_token": "[PAD]" + }, + "added_tokens": [ + { + "id": 0, + "content": "[PAD]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 1, + "content": "[UNK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 2, + "content": "[CLS]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 3, + "content": "[SEP]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 4, + "content": "[MASK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + } + ], + "normalizer": { + "type": "BertNormalizer", + "clean_text": true, + "handle_chinese_chars": true, + "strip_accents": null, + "lowercase": false + }, + "pre_tokenizer": { + "type": "BertPreTokenizer" + }, + "post_processor": { + "type": "TemplateProcessing", + "single": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + } + ], + "pair": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "B", + "type_id": 1 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 1 + } + } + ], + "special_tokens": { + "[CLS]": { + "id": "[CLS]", + "ids": [ + 2 + ], + "tokens": [ + "[CLS]" + ] + }, + "[SEP]": { + "id": "[SEP]", + "ids": [ + 3 + ], + "tokens": [ + "[SEP]" + ] + } + } + }, + "decoder": { + "type": "WordPiece", + "prefix": "##", + "cleanup": true + }, + "model": { + "type": "WordPiece", + "unk_token": "[UNK]", + "continuing_subword_prefix": "##", + "max_input_chars_per_word": 100, + "vocab": { + "[PAD]": 0, + "[UNK]": 1, + "[CLS]": 2, + "[SEP]": 3, + "[MASK]": 4, + "AAAAAA": 5, + "AAAAAT": 6, + "AAAAAC": 7, + "AAAAAG": 8, + "AAAATA": 9, + "AAAATT": 10, + "AAAATC": 11, + "AAAATG": 12, + "AAAACA": 13, + "AAAACT": 14, + "AAAACC": 15, + "AAAACG": 16, + "AAAAGA": 17, + "AAAAGT": 18, + "AAAAGC": 19, + "AAAAGG": 20, + "AAATAA": 21, + "AAATAT": 22, + "AAATAC": 23, + "AAATAG": 24, + "AAATTA": 25, + "AAATTT": 26, + "AAATTC": 27, + "AAATTG": 28, + "AAATCA": 29, + "AAATCT": 30, + "AAATCC": 31, + "AAATCG": 32, + "AAATGA": 33, + "AAATGT": 34, + "AAATGC": 35, + "AAATGG": 36, + "AAACAA": 37, + "AAACAT": 38, + "AAACAC": 39, + "AAACAG": 40, + "AAACTA": 41, + "AAACTT": 42, + "AAACTC": 43, + "AAACTG": 44, + "AAACCA": 45, + "AAACCT": 46, + "AAACCC": 47, + "AAACCG": 48, + "AAACGA": 49, + "AAACGT": 50, + "AAACGC": 51, + "AAACGG": 52, + "AAAGAA": 53, + "AAAGAT": 54, + "AAAGAC": 55, + "AAAGAG": 56, + "AAAGTA": 57, + "AAAGTT": 58, + "AAAGTC": 59, + "AAAGTG": 60, + "AAAGCA": 61, + "AAAGCT": 62, + "AAAGCC": 63, + "AAAGCG": 64, + "AAAGGA": 65, + "AAAGGT": 66, + "AAAGGC": 67, + "AAAGGG": 68, + "AATAAA": 69, + "AATAAT": 70, + "AATAAC": 71, + "AATAAG": 72, + "AATATA": 73, + "AATATT": 74, + "AATATC": 75, + "AATATG": 76, + "AATACA": 77, + "AATACT": 78, + "AATACC": 79, + "AATACG": 80, + "AATAGA": 81, + "AATAGT": 82, + "AATAGC": 83, + "AATAGG": 84, + "AATTAA": 85, + "AATTAT": 86, + "AATTAC": 87, + "AATTAG": 88, + "AATTTA": 89, + "AATTTT": 90, + "AATTTC": 91, + "AATTTG": 92, + "AATTCA": 93, + "AATTCT": 94, + "AATTCC": 95, + "AATTCG": 96, + "AATTGA": 97, + "AATTGT": 98, + "AATTGC": 99, + "AATTGG": 100, + "AATCAA": 101, + "AATCAT": 102, + "AATCAC": 103, + "AATCAG": 104, + "AATCTA": 105, + "AATCTT": 106, + "AATCTC": 107, + "AATCTG": 108, + "AATCCA": 109, + "AATCCT": 110, + "AATCCC": 111, + "AATCCG": 112, + "AATCGA": 113, + "AATCGT": 114, + "AATCGC": 115, + "AATCGG": 116, + "AATGAA": 117, + "AATGAT": 118, + "AATGAC": 119, + "AATGAG": 120, + "AATGTA": 121, + "AATGTT": 122, + "AATGTC": 123, + "AATGTG": 124, + "AATGCA": 125, + "AATGCT": 126, + "AATGCC": 127, + "AATGCG": 128, + "AATGGA": 129, + "AATGGT": 130, + "AATGGC": 131, + "AATGGG": 132, + "AACAAA": 133, + "AACAAT": 134, + "AACAAC": 135, + "AACAAG": 136, + "AACATA": 137, + "AACATT": 138, + "AACATC": 139, + "AACATG": 140, + "AACACA": 141, + "AACACT": 142, + "AACACC": 143, + "AACACG": 144, + "AACAGA": 145, + "AACAGT": 146, + "AACAGC": 147, + "AACAGG": 148, + "AACTAA": 149, + "AACTAT": 150, + "AACTAC": 151, + "AACTAG": 152, + "AACTTA": 153, + "AACTTT": 154, + "AACTTC": 155, + "AACTTG": 156, + "AACTCA": 157, + "AACTCT": 158, + "AACTCC": 159, + "AACTCG": 160, + "AACTGA": 161, + "AACTGT": 162, + "AACTGC": 163, + "AACTGG": 164, + "AACCAA": 165, + "AACCAT": 166, + "AACCAC": 167, + "AACCAG": 168, + "AACCTA": 169, + "AACCTT": 170, + "AACCTC": 171, + "AACCTG": 172, + "AACCCA": 173, + "AACCCT": 174, + "AACCCC": 175, + "AACCCG": 176, + "AACCGA": 177, + "AACCGT": 178, + "AACCGC": 179, + "AACCGG": 180, + "AACGAA": 181, + "AACGAT": 182, + "AACGAC": 183, + "AACGAG": 184, + "AACGTA": 185, + "AACGTT": 186, + "AACGTC": 187, + "AACGTG": 188, + "AACGCA": 189, + "AACGCT": 190, + "AACGCC": 191, + "AACGCG": 192, + "AACGGA": 193, + "AACGGT": 194, + "AACGGC": 195, + "AACGGG": 196, + "AAGAAA": 197, + "AAGAAT": 198, + "AAGAAC": 199, + "AAGAAG": 200, + "AAGATA": 201, + "AAGATT": 202, + "AAGATC": 203, + "AAGATG": 204, + "AAGACA": 205, + "AAGACT": 206, + "AAGACC": 207, + "AAGACG": 208, + "AAGAGA": 209, + "AAGAGT": 210, + "AAGAGC": 211, + "AAGAGG": 212, + "AAGTAA": 213, + "AAGTAT": 214, + "AAGTAC": 215, + "AAGTAG": 216, + "AAGTTA": 217, + "AAGTTT": 218, + "AAGTTC": 219, + "AAGTTG": 220, + "AAGTCA": 221, + "AAGTCT": 222, + "AAGTCC": 223, + "AAGTCG": 224, + "AAGTGA": 225, + "AAGTGT": 226, + "AAGTGC": 227, + "AAGTGG": 228, + "AAGCAA": 229, + "AAGCAT": 230, + "AAGCAC": 231, + "AAGCAG": 232, + "AAGCTA": 233, + "AAGCTT": 234, + "AAGCTC": 235, + "AAGCTG": 236, + "AAGCCA": 237, + "AAGCCT": 238, + "AAGCCC": 239, + "AAGCCG": 240, + "AAGCGA": 241, + "AAGCGT": 242, + "AAGCGC": 243, + "AAGCGG": 244, + "AAGGAA": 245, + "AAGGAT": 246, + "AAGGAC": 247, + "AAGGAG": 248, + "AAGGTA": 249, + "AAGGTT": 250, + "AAGGTC": 251, + "AAGGTG": 252, + "AAGGCA": 253, + "AAGGCT": 254, + "AAGGCC": 255, + "AAGGCG": 256, + "AAGGGA": 257, + "AAGGGT": 258, + "AAGGGC": 259, + "AAGGGG": 260, + "ATAAAA": 261, + "ATAAAT": 262, + "ATAAAC": 263, + "ATAAAG": 264, + "ATAATA": 265, + "ATAATT": 266, + "ATAATC": 267, + "ATAATG": 268, + "ATAACA": 269, + "ATAACT": 270, + "ATAACC": 271, + "ATAACG": 272, + "ATAAGA": 273, + "ATAAGT": 274, + "ATAAGC": 275, + "ATAAGG": 276, + "ATATAA": 277, + "ATATAT": 278, + "ATATAC": 279, + "ATATAG": 280, + "ATATTA": 281, + "ATATTT": 282, + "ATATTC": 283, + "ATATTG": 284, + "ATATCA": 285, + "ATATCT": 286, + "ATATCC": 287, + "ATATCG": 288, + "ATATGA": 289, + "ATATGT": 290, + "ATATGC": 291, + "ATATGG": 292, + "ATACAA": 293, + "ATACAT": 294, + "ATACAC": 295, + "ATACAG": 296, + "ATACTA": 297, + "ATACTT": 298, + "ATACTC": 299, + "ATACTG": 300, + "ATACCA": 301, + "ATACCT": 302, + "ATACCC": 303, + "ATACCG": 304, + "ATACGA": 305, + "ATACGT": 306, + "ATACGC": 307, + "ATACGG": 308, + "ATAGAA": 309, + "ATAGAT": 310, + "ATAGAC": 311, + "ATAGAG": 312, + "ATAGTA": 313, + "ATAGTT": 314, + "ATAGTC": 315, + "ATAGTG": 316, + "ATAGCA": 317, + "ATAGCT": 318, + "ATAGCC": 319, + "ATAGCG": 320, + "ATAGGA": 321, + "ATAGGT": 322, + "ATAGGC": 323, + "ATAGGG": 324, + "ATTAAA": 325, + "ATTAAT": 326, + "ATTAAC": 327, + "ATTAAG": 328, + "ATTATA": 329, + "ATTATT": 330, + "ATTATC": 331, + "ATTATG": 332, + "ATTACA": 333, + "ATTACT": 334, + "ATTACC": 335, + "ATTACG": 336, + "ATTAGA": 337, + "ATTAGT": 338, + "ATTAGC": 339, + "ATTAGG": 340, + "ATTTAA": 341, + "ATTTAT": 342, + "ATTTAC": 343, + "ATTTAG": 344, + "ATTTTA": 345, + "ATTTTT": 346, + "ATTTTC": 347, + "ATTTTG": 348, + "ATTTCA": 349, + "ATTTCT": 350, + "ATTTCC": 351, + "ATTTCG": 352, + "ATTTGA": 353, + "ATTTGT": 354, + "ATTTGC": 355, + "ATTTGG": 356, + "ATTCAA": 357, + "ATTCAT": 358, + "ATTCAC": 359, + "ATTCAG": 360, + "ATTCTA": 361, + "ATTCTT": 362, + "ATTCTC": 363, + "ATTCTG": 364, + "ATTCCA": 365, + "ATTCCT": 366, + "ATTCCC": 367, + "ATTCCG": 368, + "ATTCGA": 369, + "ATTCGT": 370, + "ATTCGC": 371, + "ATTCGG": 372, + "ATTGAA": 373, + "ATTGAT": 374, + "ATTGAC": 375, + "ATTGAG": 376, + "ATTGTA": 377, + "ATTGTT": 378, + "ATTGTC": 379, + "ATTGTG": 380, + "ATTGCA": 381, + "ATTGCT": 382, + "ATTGCC": 383, + "ATTGCG": 384, + "ATTGGA": 385, + "ATTGGT": 386, + "ATTGGC": 387, + "ATTGGG": 388, + "ATCAAA": 389, + "ATCAAT": 390, + "ATCAAC": 391, + "ATCAAG": 392, + "ATCATA": 393, + "ATCATT": 394, + "ATCATC": 395, + "ATCATG": 396, + "ATCACA": 397, + "ATCACT": 398, + "ATCACC": 399, + "ATCACG": 400, + "ATCAGA": 401, + "ATCAGT": 402, + "ATCAGC": 403, + "ATCAGG": 404, + "ATCTAA": 405, + "ATCTAT": 406, + "ATCTAC": 407, + "ATCTAG": 408, + "ATCTTA": 409, + "ATCTTT": 410, + "ATCTTC": 411, + "ATCTTG": 412, + "ATCTCA": 413, + "ATCTCT": 414, + "ATCTCC": 415, + "ATCTCG": 416, + "ATCTGA": 417, + "ATCTGT": 418, + "ATCTGC": 419, + "ATCTGG": 420, + "ATCCAA": 421, + "ATCCAT": 422, + "ATCCAC": 423, + "ATCCAG": 424, + "ATCCTA": 425, + "ATCCTT": 426, + "ATCCTC": 427, + "ATCCTG": 428, + "ATCCCA": 429, + "ATCCCT": 430, + "ATCCCC": 431, + "ATCCCG": 432, + "ATCCGA": 433, + "ATCCGT": 434, + "ATCCGC": 435, + "ATCCGG": 436, + "ATCGAA": 437, + "ATCGAT": 438, + "ATCGAC": 439, + "ATCGAG": 440, + "ATCGTA": 441, + "ATCGTT": 442, + "ATCGTC": 443, + "ATCGTG": 444, + "ATCGCA": 445, + "ATCGCT": 446, + "ATCGCC": 447, + "ATCGCG": 448, + "ATCGGA": 449, + "ATCGGT": 450, + "ATCGGC": 451, + "ATCGGG": 452, + "ATGAAA": 453, + "ATGAAT": 454, + "ATGAAC": 455, + "ATGAAG": 456, + "ATGATA": 457, + "ATGATT": 458, + "ATGATC": 459, + "ATGATG": 460, + "ATGACA": 461, + "ATGACT": 462, + "ATGACC": 463, + "ATGACG": 464, + "ATGAGA": 465, + "ATGAGT": 466, + "ATGAGC": 467, + "ATGAGG": 468, + "ATGTAA": 469, + "ATGTAT": 470, + "ATGTAC": 471, + "ATGTAG": 472, + "ATGTTA": 473, + "ATGTTT": 474, + "ATGTTC": 475, + "ATGTTG": 476, + "ATGTCA": 477, + "ATGTCT": 478, + "ATGTCC": 479, + "ATGTCG": 480, + "ATGTGA": 481, + "ATGTGT": 482, + "ATGTGC": 483, + "ATGTGG": 484, + "ATGCAA": 485, + "ATGCAT": 486, + "ATGCAC": 487, + "ATGCAG": 488, + "ATGCTA": 489, + "ATGCTT": 490, + "ATGCTC": 491, + "ATGCTG": 492, + "ATGCCA": 493, + "ATGCCT": 494, + "ATGCCC": 495, + "ATGCCG": 496, + "ATGCGA": 497, + "ATGCGT": 498, + "ATGCGC": 499, + "ATGCGG": 500, + "ATGGAA": 501, + "ATGGAT": 502, + "ATGGAC": 503, + "ATGGAG": 504, + "ATGGTA": 505, + "ATGGTT": 506, + "ATGGTC": 507, + "ATGGTG": 508, + "ATGGCA": 509, + "ATGGCT": 510, + "ATGGCC": 511, + "ATGGCG": 512, + "ATGGGA": 513, + "ATGGGT": 514, + "ATGGGC": 515, + "ATGGGG": 516, + "ACAAAA": 517, + "ACAAAT": 518, + "ACAAAC": 519, + "ACAAAG": 520, + "ACAATA": 521, + "ACAATT": 522, + "ACAATC": 523, + "ACAATG": 524, + "ACAACA": 525, + "ACAACT": 526, + "ACAACC": 527, + "ACAACG": 528, + "ACAAGA": 529, + "ACAAGT": 530, + "ACAAGC": 531, + "ACAAGG": 532, + "ACATAA": 533, + "ACATAT": 534, + "ACATAC": 535, + "ACATAG": 536, + "ACATTA": 537, + "ACATTT": 538, + "ACATTC": 539, + "ACATTG": 540, + "ACATCA": 541, + "ACATCT": 542, + "ACATCC": 543, + "ACATCG": 544, + "ACATGA": 545, + "ACATGT": 546, + "ACATGC": 547, + "ACATGG": 548, + "ACACAA": 549, + "ACACAT": 550, + "ACACAC": 551, + "ACACAG": 552, + "ACACTA": 553, + "ACACTT": 554, + "ACACTC": 555, + "ACACTG": 556, + "ACACCA": 557, + "ACACCT": 558, + "ACACCC": 559, + "ACACCG": 560, + "ACACGA": 561, + "ACACGT": 562, + "ACACGC": 563, + "ACACGG": 564, + "ACAGAA": 565, + "ACAGAT": 566, + "ACAGAC": 567, + "ACAGAG": 568, + "ACAGTA": 569, + "ACAGTT": 570, + "ACAGTC": 571, + "ACAGTG": 572, + "ACAGCA": 573, + "ACAGCT": 574, + "ACAGCC": 575, + "ACAGCG": 576, + "ACAGGA": 577, + "ACAGGT": 578, + "ACAGGC": 579, + "ACAGGG": 580, + "ACTAAA": 581, + "ACTAAT": 582, + "ACTAAC": 583, + "ACTAAG": 584, + "ACTATA": 585, + "ACTATT": 586, + "ACTATC": 587, + "ACTATG": 588, + "ACTACA": 589, + "ACTACT": 590, + "ACTACC": 591, + "ACTACG": 592, + "ACTAGA": 593, + "ACTAGT": 594, + "ACTAGC": 595, + "ACTAGG": 596, + "ACTTAA": 597, + "ACTTAT": 598, + "ACTTAC": 599, + "ACTTAG": 600, + "ACTTTA": 601, + "ACTTTT": 602, + "ACTTTC": 603, + "ACTTTG": 604, + "ACTTCA": 605, + "ACTTCT": 606, + "ACTTCC": 607, + "ACTTCG": 608, + "ACTTGA": 609, + "ACTTGT": 610, + "ACTTGC": 611, + "ACTTGG": 612, + "ACTCAA": 613, + "ACTCAT": 614, + "ACTCAC": 615, + "ACTCAG": 616, + "ACTCTA": 617, + "ACTCTT": 618, + "ACTCTC": 619, + "ACTCTG": 620, + "ACTCCA": 621, + "ACTCCT": 622, + "ACTCCC": 623, + "ACTCCG": 624, + "ACTCGA": 625, + "ACTCGT": 626, + "ACTCGC": 627, + "ACTCGG": 628, + "ACTGAA": 629, + "ACTGAT": 630, + "ACTGAC": 631, + "ACTGAG": 632, + "ACTGTA": 633, + "ACTGTT": 634, + "ACTGTC": 635, + "ACTGTG": 636, + "ACTGCA": 637, + "ACTGCT": 638, + "ACTGCC": 639, + "ACTGCG": 640, + "ACTGGA": 641, + "ACTGGT": 642, + "ACTGGC": 643, + "ACTGGG": 644, + "ACCAAA": 645, + "ACCAAT": 646, + "ACCAAC": 647, + "ACCAAG": 648, + "ACCATA": 649, + "ACCATT": 650, + "ACCATC": 651, + "ACCATG": 652, + "ACCACA": 653, + "ACCACT": 654, + "ACCACC": 655, + "ACCACG": 656, + "ACCAGA": 657, + "ACCAGT": 658, + "ACCAGC": 659, + "ACCAGG": 660, + "ACCTAA": 661, + "ACCTAT": 662, + "ACCTAC": 663, + "ACCTAG": 664, + "ACCTTA": 665, + "ACCTTT": 666, + "ACCTTC": 667, + "ACCTTG": 668, + "ACCTCA": 669, + "ACCTCT": 670, + "ACCTCC": 671, + "ACCTCG": 672, + "ACCTGA": 673, + "ACCTGT": 674, + "ACCTGC": 675, + "ACCTGG": 676, + "ACCCAA": 677, + "ACCCAT": 678, + "ACCCAC": 679, + "ACCCAG": 680, + "ACCCTA": 681, + "ACCCTT": 682, + "ACCCTC": 683, + "ACCCTG": 684, + "ACCCCA": 685, + "ACCCCT": 686, + "ACCCCC": 687, + "ACCCCG": 688, + "ACCCGA": 689, + "ACCCGT": 690, + "ACCCGC": 691, + "ACCCGG": 692, + "ACCGAA": 693, + "ACCGAT": 694, + "ACCGAC": 695, + "ACCGAG": 696, + "ACCGTA": 697, + "ACCGTT": 698, + "ACCGTC": 699, + "ACCGTG": 700, + "ACCGCA": 701, + "ACCGCT": 702, + "ACCGCC": 703, + "ACCGCG": 704, + "ACCGGA": 705, + "ACCGGT": 706, + "ACCGGC": 707, + "ACCGGG": 708, + "ACGAAA": 709, + "ACGAAT": 710, + "ACGAAC": 711, + "ACGAAG": 712, + "ACGATA": 713, + "ACGATT": 714, + "ACGATC": 715, + "ACGATG": 716, + "ACGACA": 717, + "ACGACT": 718, + "ACGACC": 719, + "ACGACG": 720, + "ACGAGA": 721, + "ACGAGT": 722, + "ACGAGC": 723, + "ACGAGG": 724, + "ACGTAA": 725, + "ACGTAT": 726, + "ACGTAC": 727, + "ACGTAG": 728, + "ACGTTA": 729, + "ACGTTT": 730, + "ACGTTC": 731, + "ACGTTG": 732, + "ACGTCA": 733, + "ACGTCT": 734, + "ACGTCC": 735, + "ACGTCG": 736, + "ACGTGA": 737, + "ACGTGT": 738, + "ACGTGC": 739, + "ACGTGG": 740, + "ACGCAA": 741, + "ACGCAT": 742, + "ACGCAC": 743, + "ACGCAG": 744, + "ACGCTA": 745, + "ACGCTT": 746, + "ACGCTC": 747, + "ACGCTG": 748, + "ACGCCA": 749, + "ACGCCT": 750, + "ACGCCC": 751, + "ACGCCG": 752, + "ACGCGA": 753, + "ACGCGT": 754, + "ACGCGC": 755, + "ACGCGG": 756, + "ACGGAA": 757, + "ACGGAT": 758, + "ACGGAC": 759, + "ACGGAG": 760, + "ACGGTA": 761, + "ACGGTT": 762, + "ACGGTC": 763, + "ACGGTG": 764, + "ACGGCA": 765, + "ACGGCT": 766, + "ACGGCC": 767, + "ACGGCG": 768, + "ACGGGA": 769, + "ACGGGT": 770, + "ACGGGC": 771, + "ACGGGG": 772, + "AGAAAA": 773, + "AGAAAT": 774, + "AGAAAC": 775, + "AGAAAG": 776, + "AGAATA": 777, + "AGAATT": 778, + "AGAATC": 779, + "AGAATG": 780, + "AGAACA": 781, + "AGAACT": 782, + "AGAACC": 783, + "AGAACG": 784, + "AGAAGA": 785, + "AGAAGT": 786, + "AGAAGC": 787, + "AGAAGG": 788, + "AGATAA": 789, + "AGATAT": 790, + "AGATAC": 791, + "AGATAG": 792, + "AGATTA": 793, + "AGATTT": 794, + "AGATTC": 795, + "AGATTG": 796, + "AGATCA": 797, + "AGATCT": 798, + "AGATCC": 799, + "AGATCG": 800, + "AGATGA": 801, + "AGATGT": 802, + "AGATGC": 803, + "AGATGG": 804, + "AGACAA": 805, + "AGACAT": 806, + "AGACAC": 807, + "AGACAG": 808, + "AGACTA": 809, + "AGACTT": 810, + "AGACTC": 811, + "AGACTG": 812, + "AGACCA": 813, + "AGACCT": 814, + "AGACCC": 815, + "AGACCG": 816, + "AGACGA": 817, + "AGACGT": 818, + "AGACGC": 819, + "AGACGG": 820, + "AGAGAA": 821, + "AGAGAT": 822, + "AGAGAC": 823, + "AGAGAG": 824, + "AGAGTA": 825, + "AGAGTT": 826, + "AGAGTC": 827, + "AGAGTG": 828, + "AGAGCA": 829, + "AGAGCT": 830, + "AGAGCC": 831, + "AGAGCG": 832, + "AGAGGA": 833, + "AGAGGT": 834, + "AGAGGC": 835, + "AGAGGG": 836, + "AGTAAA": 837, + "AGTAAT": 838, + "AGTAAC": 839, + "AGTAAG": 840, + "AGTATA": 841, + "AGTATT": 842, + "AGTATC": 843, + "AGTATG": 844, + "AGTACA": 845, + "AGTACT": 846, + "AGTACC": 847, + "AGTACG": 848, + "AGTAGA": 849, + "AGTAGT": 850, + "AGTAGC": 851, + "AGTAGG": 852, + "AGTTAA": 853, + "AGTTAT": 854, + "AGTTAC": 855, + "AGTTAG": 856, + "AGTTTA": 857, + "AGTTTT": 858, + "AGTTTC": 859, + "AGTTTG": 860, + "AGTTCA": 861, + "AGTTCT": 862, + "AGTTCC": 863, + "AGTTCG": 864, + "AGTTGA": 865, + "AGTTGT": 866, + "AGTTGC": 867, + "AGTTGG": 868, + "AGTCAA": 869, + "AGTCAT": 870, + "AGTCAC": 871, + "AGTCAG": 872, + "AGTCTA": 873, + "AGTCTT": 874, + "AGTCTC": 875, + "AGTCTG": 876, + "AGTCCA": 877, + "AGTCCT": 878, + "AGTCCC": 879, + "AGTCCG": 880, + "AGTCGA": 881, + "AGTCGT": 882, + "AGTCGC": 883, + "AGTCGG": 884, + "AGTGAA": 885, + "AGTGAT": 886, + "AGTGAC": 887, + "AGTGAG": 888, + "AGTGTA": 889, + "AGTGTT": 890, + "AGTGTC": 891, + "AGTGTG": 892, + "AGTGCA": 893, + "AGTGCT": 894, + "AGTGCC": 895, + "AGTGCG": 896, + "AGTGGA": 897, + "AGTGGT": 898, + "AGTGGC": 899, + "AGTGGG": 900, + "AGCAAA": 901, + "AGCAAT": 902, + "AGCAAC": 903, + "AGCAAG": 904, + "AGCATA": 905, + "AGCATT": 906, + "AGCATC": 907, + "AGCATG": 908, + "AGCACA": 909, + "AGCACT": 910, + "AGCACC": 911, + "AGCACG": 912, + "AGCAGA": 913, + "AGCAGT": 914, + "AGCAGC": 915, + "AGCAGG": 916, + "AGCTAA": 917, + "AGCTAT": 918, + "AGCTAC": 919, + "AGCTAG": 920, + "AGCTTA": 921, + "AGCTTT": 922, + "AGCTTC": 923, + "AGCTTG": 924, + "AGCTCA": 925, + "AGCTCT": 926, + "AGCTCC": 927, + "AGCTCG": 928, + "AGCTGA": 929, + "AGCTGT": 930, + "AGCTGC": 931, + "AGCTGG": 932, + "AGCCAA": 933, + "AGCCAT": 934, + "AGCCAC": 935, + "AGCCAG": 936, + "AGCCTA": 937, + "AGCCTT": 938, + "AGCCTC": 939, + "AGCCTG": 940, + "AGCCCA": 941, + "AGCCCT": 942, + "AGCCCC": 943, + "AGCCCG": 944, + "AGCCGA": 945, + "AGCCGT": 946, + "AGCCGC": 947, + "AGCCGG": 948, + "AGCGAA": 949, + "AGCGAT": 950, + "AGCGAC": 951, + "AGCGAG": 952, + "AGCGTA": 953, + "AGCGTT": 954, + "AGCGTC": 955, + "AGCGTG": 956, + "AGCGCA": 957, + "AGCGCT": 958, + "AGCGCC": 959, + "AGCGCG": 960, + "AGCGGA": 961, + "AGCGGT": 962, + "AGCGGC": 963, + "AGCGGG": 964, + "AGGAAA": 965, + "AGGAAT": 966, + "AGGAAC": 967, + "AGGAAG": 968, + "AGGATA": 969, + "AGGATT": 970, + "AGGATC": 971, + "AGGATG": 972, + "AGGACA": 973, + "AGGACT": 974, + "AGGACC": 975, + "AGGACG": 976, + "AGGAGA": 977, + "AGGAGT": 978, + "AGGAGC": 979, + "AGGAGG": 980, + "AGGTAA": 981, + "AGGTAT": 982, + "AGGTAC": 983, + "AGGTAG": 984, + "AGGTTA": 985, + "AGGTTT": 986, + "AGGTTC": 987, + "AGGTTG": 988, + "AGGTCA": 989, + "AGGTCT": 990, + "AGGTCC": 991, + "AGGTCG": 992, + "AGGTGA": 993, + "AGGTGT": 994, + "AGGTGC": 995, + "AGGTGG": 996, + "AGGCAA": 997, + "AGGCAT": 998, + "AGGCAC": 999, + "AGGCAG": 1000, + "AGGCTA": 1001, + "AGGCTT": 1002, + "AGGCTC": 1003, + "AGGCTG": 1004, + "AGGCCA": 1005, + "AGGCCT": 1006, + "AGGCCC": 1007, + "AGGCCG": 1008, + "AGGCGA": 1009, + "AGGCGT": 1010, + "AGGCGC": 1011, + "AGGCGG": 1012, + "AGGGAA": 1013, + "AGGGAT": 1014, + "AGGGAC": 1015, + "AGGGAG": 1016, + "AGGGTA": 1017, + "AGGGTT": 1018, + "AGGGTC": 1019, + "AGGGTG": 1020, + "AGGGCA": 1021, + "AGGGCT": 1022, + "AGGGCC": 1023, + "AGGGCG": 1024, + "AGGGGA": 1025, + "AGGGGT": 1026, + "AGGGGC": 1027, + "AGGGGG": 1028, + "TAAAAA": 1029, + "TAAAAT": 1030, + "TAAAAC": 1031, + "TAAAAG": 1032, + "TAAATA": 1033, + "TAAATT": 1034, + "TAAATC": 1035, + "TAAATG": 1036, + "TAAACA": 1037, + "TAAACT": 1038, + "TAAACC": 1039, + "TAAACG": 1040, + "TAAAGA": 1041, + "TAAAGT": 1042, + "TAAAGC": 1043, + "TAAAGG": 1044, + "TAATAA": 1045, + "TAATAT": 1046, + "TAATAC": 1047, + "TAATAG": 1048, + "TAATTA": 1049, + "TAATTT": 1050, + "TAATTC": 1051, + "TAATTG": 1052, + "TAATCA": 1053, + "TAATCT": 1054, + "TAATCC": 1055, + "TAATCG": 1056, + "TAATGA": 1057, + "TAATGT": 1058, + "TAATGC": 1059, + "TAATGG": 1060, + "TAACAA": 1061, + "TAACAT": 1062, + "TAACAC": 1063, + "TAACAG": 1064, + "TAACTA": 1065, + "TAACTT": 1066, + "TAACTC": 1067, + "TAACTG": 1068, + "TAACCA": 1069, + "TAACCT": 1070, + "TAACCC": 1071, + "TAACCG": 1072, + "TAACGA": 1073, + "TAACGT": 1074, + "TAACGC": 1075, + "TAACGG": 1076, + "TAAGAA": 1077, + "TAAGAT": 1078, + "TAAGAC": 1079, + "TAAGAG": 1080, + "TAAGTA": 1081, + "TAAGTT": 1082, + "TAAGTC": 1083, + "TAAGTG": 1084, + "TAAGCA": 1085, + "TAAGCT": 1086, + "TAAGCC": 1087, + "TAAGCG": 1088, + "TAAGGA": 1089, + "TAAGGT": 1090, + "TAAGGC": 1091, + "TAAGGG": 1092, + "TATAAA": 1093, + "TATAAT": 1094, + "TATAAC": 1095, + "TATAAG": 1096, + "TATATA": 1097, + "TATATT": 1098, + "TATATC": 1099, + "TATATG": 1100, + "TATACA": 1101, + "TATACT": 1102, + "TATACC": 1103, + "TATACG": 1104, + "TATAGA": 1105, + "TATAGT": 1106, + "TATAGC": 1107, + "TATAGG": 1108, + "TATTAA": 1109, + "TATTAT": 1110, + "TATTAC": 1111, + "TATTAG": 1112, + "TATTTA": 1113, + "TATTTT": 1114, + "TATTTC": 1115, + "TATTTG": 1116, + "TATTCA": 1117, + "TATTCT": 1118, + "TATTCC": 1119, + "TATTCG": 1120, + "TATTGA": 1121, + "TATTGT": 1122, + "TATTGC": 1123, + "TATTGG": 1124, + "TATCAA": 1125, + "TATCAT": 1126, + "TATCAC": 1127, + "TATCAG": 1128, + "TATCTA": 1129, + "TATCTT": 1130, + "TATCTC": 1131, + "TATCTG": 1132, + "TATCCA": 1133, + "TATCCT": 1134, + "TATCCC": 1135, + "TATCCG": 1136, + "TATCGA": 1137, + "TATCGT": 1138, + "TATCGC": 1139, + "TATCGG": 1140, + "TATGAA": 1141, + "TATGAT": 1142, + "TATGAC": 1143, + "TATGAG": 1144, + "TATGTA": 1145, + "TATGTT": 1146, + "TATGTC": 1147, + "TATGTG": 1148, + "TATGCA": 1149, + "TATGCT": 1150, + "TATGCC": 1151, + "TATGCG": 1152, + "TATGGA": 1153, + "TATGGT": 1154, + "TATGGC": 1155, + "TATGGG": 1156, + "TACAAA": 1157, + "TACAAT": 1158, + "TACAAC": 1159, + "TACAAG": 1160, + "TACATA": 1161, + "TACATT": 1162, + "TACATC": 1163, + "TACATG": 1164, + "TACACA": 1165, + "TACACT": 1166, + "TACACC": 1167, + "TACACG": 1168, + "TACAGA": 1169, + "TACAGT": 1170, + "TACAGC": 1171, + "TACAGG": 1172, + "TACTAA": 1173, + "TACTAT": 1174, + "TACTAC": 1175, + "TACTAG": 1176, + "TACTTA": 1177, + "TACTTT": 1178, + "TACTTC": 1179, + "TACTTG": 1180, + "TACTCA": 1181, + "TACTCT": 1182, + "TACTCC": 1183, + "TACTCG": 1184, + "TACTGA": 1185, + "TACTGT": 1186, + "TACTGC": 1187, + "TACTGG": 1188, + "TACCAA": 1189, + "TACCAT": 1190, + "TACCAC": 1191, + "TACCAG": 1192, + "TACCTA": 1193, + "TACCTT": 1194, + "TACCTC": 1195, + "TACCTG": 1196, + "TACCCA": 1197, + "TACCCT": 1198, + "TACCCC": 1199, + "TACCCG": 1200, + "TACCGA": 1201, + "TACCGT": 1202, + "TACCGC": 1203, + "TACCGG": 1204, + "TACGAA": 1205, + "TACGAT": 1206, + "TACGAC": 1207, + "TACGAG": 1208, + "TACGTA": 1209, + "TACGTT": 1210, + "TACGTC": 1211, + "TACGTG": 1212, + "TACGCA": 1213, + "TACGCT": 1214, + "TACGCC": 1215, + "TACGCG": 1216, + "TACGGA": 1217, + "TACGGT": 1218, + "TACGGC": 1219, + "TACGGG": 1220, + "TAGAAA": 1221, + "TAGAAT": 1222, + "TAGAAC": 1223, + "TAGAAG": 1224, + "TAGATA": 1225, + "TAGATT": 1226, + "TAGATC": 1227, + "TAGATG": 1228, + "TAGACA": 1229, + "TAGACT": 1230, + "TAGACC": 1231, + "TAGACG": 1232, + "TAGAGA": 1233, + "TAGAGT": 1234, + "TAGAGC": 1235, + "TAGAGG": 1236, + "TAGTAA": 1237, + "TAGTAT": 1238, + "TAGTAC": 1239, + "TAGTAG": 1240, + "TAGTTA": 1241, + "TAGTTT": 1242, + "TAGTTC": 1243, + "TAGTTG": 1244, + "TAGTCA": 1245, + "TAGTCT": 1246, + "TAGTCC": 1247, + "TAGTCG": 1248, + "TAGTGA": 1249, + "TAGTGT": 1250, + "TAGTGC": 1251, + "TAGTGG": 1252, + "TAGCAA": 1253, + "TAGCAT": 1254, + "TAGCAC": 1255, + "TAGCAG": 1256, + "TAGCTA": 1257, + "TAGCTT": 1258, + "TAGCTC": 1259, + "TAGCTG": 1260, + "TAGCCA": 1261, + "TAGCCT": 1262, + "TAGCCC": 1263, + "TAGCCG": 1264, + "TAGCGA": 1265, + "TAGCGT": 1266, + "TAGCGC": 1267, + "TAGCGG": 1268, + "TAGGAA": 1269, + "TAGGAT": 1270, + "TAGGAC": 1271, + "TAGGAG": 1272, + "TAGGTA": 1273, + "TAGGTT": 1274, + "TAGGTC": 1275, + "TAGGTG": 1276, + "TAGGCA": 1277, + "TAGGCT": 1278, + "TAGGCC": 1279, + "TAGGCG": 1280, + "TAGGGA": 1281, + "TAGGGT": 1282, + "TAGGGC": 1283, + "TAGGGG": 1284, + "TTAAAA": 1285, + "TTAAAT": 1286, + "TTAAAC": 1287, + "TTAAAG": 1288, + "TTAATA": 1289, + "TTAATT": 1290, + "TTAATC": 1291, + "TTAATG": 1292, + "TTAACA": 1293, + "TTAACT": 1294, + "TTAACC": 1295, + "TTAACG": 1296, + "TTAAGA": 1297, + "TTAAGT": 1298, + "TTAAGC": 1299, + "TTAAGG": 1300, + "TTATAA": 1301, + "TTATAT": 1302, + "TTATAC": 1303, + "TTATAG": 1304, + "TTATTA": 1305, + "TTATTT": 1306, + "TTATTC": 1307, + "TTATTG": 1308, + "TTATCA": 1309, + "TTATCT": 1310, + "TTATCC": 1311, + "TTATCG": 1312, + "TTATGA": 1313, + "TTATGT": 1314, + "TTATGC": 1315, + "TTATGG": 1316, + "TTACAA": 1317, + "TTACAT": 1318, + "TTACAC": 1319, + "TTACAG": 1320, + "TTACTA": 1321, + "TTACTT": 1322, + "TTACTC": 1323, + "TTACTG": 1324, + "TTACCA": 1325, + "TTACCT": 1326, + "TTACCC": 1327, + "TTACCG": 1328, + "TTACGA": 1329, + "TTACGT": 1330, + "TTACGC": 1331, + "TTACGG": 1332, + "TTAGAA": 1333, + "TTAGAT": 1334, + "TTAGAC": 1335, + "TTAGAG": 1336, + "TTAGTA": 1337, + "TTAGTT": 1338, + "TTAGTC": 1339, + "TTAGTG": 1340, + "TTAGCA": 1341, + "TTAGCT": 1342, + "TTAGCC": 1343, + "TTAGCG": 1344, + "TTAGGA": 1345, + "TTAGGT": 1346, + "TTAGGC": 1347, + "TTAGGG": 1348, + "TTTAAA": 1349, + "TTTAAT": 1350, + "TTTAAC": 1351, + "TTTAAG": 1352, + "TTTATA": 1353, + "TTTATT": 1354, + "TTTATC": 1355, + "TTTATG": 1356, + "TTTACA": 1357, + "TTTACT": 1358, + "TTTACC": 1359, + "TTTACG": 1360, + "TTTAGA": 1361, + "TTTAGT": 1362, + "TTTAGC": 1363, + "TTTAGG": 1364, + "TTTTAA": 1365, + "TTTTAT": 1366, + "TTTTAC": 1367, + "TTTTAG": 1368, + "TTTTTA": 1369, + "TTTTTT": 1370, + "TTTTTC": 1371, + "TTTTTG": 1372, + "TTTTCA": 1373, + "TTTTCT": 1374, + "TTTTCC": 1375, + "TTTTCG": 1376, + "TTTTGA": 1377, + "TTTTGT": 1378, + "TTTTGC": 1379, + "TTTTGG": 1380, + "TTTCAA": 1381, + "TTTCAT": 1382, + "TTTCAC": 1383, + "TTTCAG": 1384, + "TTTCTA": 1385, + "TTTCTT": 1386, + "TTTCTC": 1387, + "TTTCTG": 1388, + "TTTCCA": 1389, + "TTTCCT": 1390, + "TTTCCC": 1391, + "TTTCCG": 1392, + "TTTCGA": 1393, + "TTTCGT": 1394, + "TTTCGC": 1395, + "TTTCGG": 1396, + "TTTGAA": 1397, + "TTTGAT": 1398, + "TTTGAC": 1399, + "TTTGAG": 1400, + "TTTGTA": 1401, + "TTTGTT": 1402, + "TTTGTC": 1403, + "TTTGTG": 1404, + "TTTGCA": 1405, + "TTTGCT": 1406, + "TTTGCC": 1407, + "TTTGCG": 1408, + "TTTGGA": 1409, + "TTTGGT": 1410, + "TTTGGC": 1411, + "TTTGGG": 1412, + "TTCAAA": 1413, + "TTCAAT": 1414, + "TTCAAC": 1415, + "TTCAAG": 1416, + "TTCATA": 1417, + "TTCATT": 1418, + "TTCATC": 1419, + "TTCATG": 1420, + "TTCACA": 1421, + "TTCACT": 1422, + "TTCACC": 1423, + "TTCACG": 1424, + "TTCAGA": 1425, + "TTCAGT": 1426, + "TTCAGC": 1427, + "TTCAGG": 1428, + "TTCTAA": 1429, + "TTCTAT": 1430, + "TTCTAC": 1431, + "TTCTAG": 1432, + "TTCTTA": 1433, + "TTCTTT": 1434, + "TTCTTC": 1435, + "TTCTTG": 1436, + "TTCTCA": 1437, + "TTCTCT": 1438, + "TTCTCC": 1439, + "TTCTCG": 1440, + "TTCTGA": 1441, + "TTCTGT": 1442, + "TTCTGC": 1443, + "TTCTGG": 1444, + "TTCCAA": 1445, + "TTCCAT": 1446, + "TTCCAC": 1447, + "TTCCAG": 1448, + "TTCCTA": 1449, + "TTCCTT": 1450, + "TTCCTC": 1451, + "TTCCTG": 1452, + "TTCCCA": 1453, + "TTCCCT": 1454, + "TTCCCC": 1455, + "TTCCCG": 1456, + "TTCCGA": 1457, + "TTCCGT": 1458, + "TTCCGC": 1459, + "TTCCGG": 1460, + "TTCGAA": 1461, + "TTCGAT": 1462, + "TTCGAC": 1463, + "TTCGAG": 1464, + "TTCGTA": 1465, + "TTCGTT": 1466, + "TTCGTC": 1467, + "TTCGTG": 1468, + "TTCGCA": 1469, + "TTCGCT": 1470, + "TTCGCC": 1471, + "TTCGCG": 1472, + "TTCGGA": 1473, + "TTCGGT": 1474, + "TTCGGC": 1475, + "TTCGGG": 1476, + "TTGAAA": 1477, + "TTGAAT": 1478, + "TTGAAC": 1479, + "TTGAAG": 1480, + "TTGATA": 1481, + "TTGATT": 1482, + "TTGATC": 1483, + "TTGATG": 1484, + "TTGACA": 1485, + "TTGACT": 1486, + "TTGACC": 1487, + "TTGACG": 1488, + "TTGAGA": 1489, + "TTGAGT": 1490, + "TTGAGC": 1491, + "TTGAGG": 1492, + "TTGTAA": 1493, + "TTGTAT": 1494, + "TTGTAC": 1495, + "TTGTAG": 1496, + "TTGTTA": 1497, + "TTGTTT": 1498, + "TTGTTC": 1499, + "TTGTTG": 1500, + "TTGTCA": 1501, + "TTGTCT": 1502, + "TTGTCC": 1503, + "TTGTCG": 1504, + "TTGTGA": 1505, + "TTGTGT": 1506, + "TTGTGC": 1507, + "TTGTGG": 1508, + "TTGCAA": 1509, + "TTGCAT": 1510, + "TTGCAC": 1511, + "TTGCAG": 1512, + "TTGCTA": 1513, + "TTGCTT": 1514, + "TTGCTC": 1515, + "TTGCTG": 1516, + "TTGCCA": 1517, + "TTGCCT": 1518, + "TTGCCC": 1519, + "TTGCCG": 1520, + "TTGCGA": 1521, + "TTGCGT": 1522, + "TTGCGC": 1523, + "TTGCGG": 1524, + "TTGGAA": 1525, + "TTGGAT": 1526, + "TTGGAC": 1527, + "TTGGAG": 1528, + "TTGGTA": 1529, + "TTGGTT": 1530, + "TTGGTC": 1531, + "TTGGTG": 1532, + "TTGGCA": 1533, + "TTGGCT": 1534, + "TTGGCC": 1535, + "TTGGCG": 1536, + "TTGGGA": 1537, + "TTGGGT": 1538, + "TTGGGC": 1539, + "TTGGGG": 1540, + "TCAAAA": 1541, + "TCAAAT": 1542, + "TCAAAC": 1543, + "TCAAAG": 1544, + "TCAATA": 1545, + "TCAATT": 1546, + "TCAATC": 1547, + "TCAATG": 1548, + "TCAACA": 1549, + "TCAACT": 1550, + "TCAACC": 1551, + "TCAACG": 1552, + "TCAAGA": 1553, + "TCAAGT": 1554, + "TCAAGC": 1555, + "TCAAGG": 1556, + "TCATAA": 1557, + "TCATAT": 1558, + "TCATAC": 1559, + "TCATAG": 1560, + "TCATTA": 1561, + "TCATTT": 1562, + "TCATTC": 1563, + "TCATTG": 1564, + "TCATCA": 1565, + "TCATCT": 1566, + "TCATCC": 1567, + "TCATCG": 1568, + "TCATGA": 1569, + "TCATGT": 1570, + "TCATGC": 1571, + "TCATGG": 1572, + "TCACAA": 1573, + "TCACAT": 1574, + "TCACAC": 1575, + "TCACAG": 1576, + "TCACTA": 1577, + "TCACTT": 1578, + "TCACTC": 1579, + "TCACTG": 1580, + "TCACCA": 1581, + "TCACCT": 1582, + "TCACCC": 1583, + "TCACCG": 1584, + "TCACGA": 1585, + "TCACGT": 1586, + "TCACGC": 1587, + "TCACGG": 1588, + "TCAGAA": 1589, + "TCAGAT": 1590, + "TCAGAC": 1591, + "TCAGAG": 1592, + "TCAGTA": 1593, + "TCAGTT": 1594, + "TCAGTC": 1595, + "TCAGTG": 1596, + "TCAGCA": 1597, + "TCAGCT": 1598, + "TCAGCC": 1599, + "TCAGCG": 1600, + "TCAGGA": 1601, + "TCAGGT": 1602, + "TCAGGC": 1603, + "TCAGGG": 1604, + "TCTAAA": 1605, + "TCTAAT": 1606, + "TCTAAC": 1607, + "TCTAAG": 1608, + "TCTATA": 1609, + "TCTATT": 1610, + "TCTATC": 1611, + "TCTATG": 1612, + "TCTACA": 1613, + "TCTACT": 1614, + "TCTACC": 1615, + "TCTACG": 1616, + "TCTAGA": 1617, + "TCTAGT": 1618, + "TCTAGC": 1619, + "TCTAGG": 1620, + "TCTTAA": 1621, + "TCTTAT": 1622, + "TCTTAC": 1623, + "TCTTAG": 1624, + "TCTTTA": 1625, + "TCTTTT": 1626, + "TCTTTC": 1627, + "TCTTTG": 1628, + "TCTTCA": 1629, + "TCTTCT": 1630, + "TCTTCC": 1631, + "TCTTCG": 1632, + "TCTTGA": 1633, + "TCTTGT": 1634, + "TCTTGC": 1635, + "TCTTGG": 1636, + "TCTCAA": 1637, + "TCTCAT": 1638, + "TCTCAC": 1639, + "TCTCAG": 1640, + "TCTCTA": 1641, + "TCTCTT": 1642, + "TCTCTC": 1643, + "TCTCTG": 1644, + "TCTCCA": 1645, + "TCTCCT": 1646, + "TCTCCC": 1647, + "TCTCCG": 1648, + "TCTCGA": 1649, + "TCTCGT": 1650, + "TCTCGC": 1651, + "TCTCGG": 1652, + "TCTGAA": 1653, + "TCTGAT": 1654, + "TCTGAC": 1655, + "TCTGAG": 1656, + "TCTGTA": 1657, + "TCTGTT": 1658, + "TCTGTC": 1659, + "TCTGTG": 1660, + "TCTGCA": 1661, + "TCTGCT": 1662, + "TCTGCC": 1663, + "TCTGCG": 1664, + "TCTGGA": 1665, + "TCTGGT": 1666, + "TCTGGC": 1667, + "TCTGGG": 1668, + "TCCAAA": 1669, + "TCCAAT": 1670, + "TCCAAC": 1671, + "TCCAAG": 1672, + "TCCATA": 1673, + "TCCATT": 1674, + "TCCATC": 1675, + "TCCATG": 1676, + "TCCACA": 1677, + "TCCACT": 1678, + "TCCACC": 1679, + "TCCACG": 1680, + "TCCAGA": 1681, + "TCCAGT": 1682, + "TCCAGC": 1683, + "TCCAGG": 1684, + "TCCTAA": 1685, + "TCCTAT": 1686, + "TCCTAC": 1687, + "TCCTAG": 1688, + "TCCTTA": 1689, + "TCCTTT": 1690, + "TCCTTC": 1691, + "TCCTTG": 1692, + "TCCTCA": 1693, + "TCCTCT": 1694, + "TCCTCC": 1695, + "TCCTCG": 1696, + "TCCTGA": 1697, + "TCCTGT": 1698, + "TCCTGC": 1699, + "TCCTGG": 1700, + "TCCCAA": 1701, + "TCCCAT": 1702, + "TCCCAC": 1703, + "TCCCAG": 1704, + "TCCCTA": 1705, + "TCCCTT": 1706, + "TCCCTC": 1707, + "TCCCTG": 1708, + "TCCCCA": 1709, + "TCCCCT": 1710, + "TCCCCC": 1711, + "TCCCCG": 1712, + "TCCCGA": 1713, + "TCCCGT": 1714, + "TCCCGC": 1715, + "TCCCGG": 1716, + "TCCGAA": 1717, + "TCCGAT": 1718, + "TCCGAC": 1719, + "TCCGAG": 1720, + "TCCGTA": 1721, + "TCCGTT": 1722, + "TCCGTC": 1723, + "TCCGTG": 1724, + "TCCGCA": 1725, + "TCCGCT": 1726, + "TCCGCC": 1727, + "TCCGCG": 1728, + "TCCGGA": 1729, + "TCCGGT": 1730, + "TCCGGC": 1731, + "TCCGGG": 1732, + "TCGAAA": 1733, + "TCGAAT": 1734, + "TCGAAC": 1735, + "TCGAAG": 1736, + "TCGATA": 1737, + "TCGATT": 1738, + "TCGATC": 1739, + "TCGATG": 1740, + "TCGACA": 1741, + "TCGACT": 1742, + "TCGACC": 1743, + "TCGACG": 1744, + "TCGAGA": 1745, + "TCGAGT": 1746, + "TCGAGC": 1747, + "TCGAGG": 1748, + "TCGTAA": 1749, + "TCGTAT": 1750, + "TCGTAC": 1751, + "TCGTAG": 1752, + "TCGTTA": 1753, + "TCGTTT": 1754, + "TCGTTC": 1755, + "TCGTTG": 1756, + "TCGTCA": 1757, + "TCGTCT": 1758, + "TCGTCC": 1759, + "TCGTCG": 1760, + "TCGTGA": 1761, + "TCGTGT": 1762, + "TCGTGC": 1763, + "TCGTGG": 1764, + "TCGCAA": 1765, + "TCGCAT": 1766, + "TCGCAC": 1767, + "TCGCAG": 1768, + "TCGCTA": 1769, + "TCGCTT": 1770, + "TCGCTC": 1771, + "TCGCTG": 1772, + "TCGCCA": 1773, + "TCGCCT": 1774, + "TCGCCC": 1775, + "TCGCCG": 1776, + "TCGCGA": 1777, + "TCGCGT": 1778, + "TCGCGC": 1779, + "TCGCGG": 1780, + "TCGGAA": 1781, + "TCGGAT": 1782, + "TCGGAC": 1783, + "TCGGAG": 1784, + "TCGGTA": 1785, + "TCGGTT": 1786, + "TCGGTC": 1787, + "TCGGTG": 1788, + "TCGGCA": 1789, + "TCGGCT": 1790, + "TCGGCC": 1791, + "TCGGCG": 1792, + "TCGGGA": 1793, + "TCGGGT": 1794, + "TCGGGC": 1795, + "TCGGGG": 1796, + "TGAAAA": 1797, + "TGAAAT": 1798, + "TGAAAC": 1799, + "TGAAAG": 1800, + "TGAATA": 1801, + "TGAATT": 1802, + "TGAATC": 1803, + "TGAATG": 1804, + "TGAACA": 1805, + "TGAACT": 1806, + "TGAACC": 1807, + "TGAACG": 1808, + "TGAAGA": 1809, + "TGAAGT": 1810, + "TGAAGC": 1811, + "TGAAGG": 1812, + "TGATAA": 1813, + "TGATAT": 1814, + "TGATAC": 1815, + "TGATAG": 1816, + "TGATTA": 1817, + "TGATTT": 1818, + "TGATTC": 1819, + "TGATTG": 1820, + "TGATCA": 1821, + "TGATCT": 1822, + "TGATCC": 1823, + "TGATCG": 1824, + "TGATGA": 1825, + "TGATGT": 1826, + "TGATGC": 1827, + "TGATGG": 1828, + "TGACAA": 1829, + "TGACAT": 1830, + "TGACAC": 1831, + "TGACAG": 1832, + "TGACTA": 1833, + "TGACTT": 1834, + "TGACTC": 1835, + "TGACTG": 1836, + "TGACCA": 1837, + "TGACCT": 1838, + "TGACCC": 1839, + "TGACCG": 1840, + "TGACGA": 1841, + "TGACGT": 1842, + "TGACGC": 1843, + "TGACGG": 1844, + "TGAGAA": 1845, + "TGAGAT": 1846, + "TGAGAC": 1847, + "TGAGAG": 1848, + "TGAGTA": 1849, + "TGAGTT": 1850, + "TGAGTC": 1851, + "TGAGTG": 1852, + "TGAGCA": 1853, + "TGAGCT": 1854, + "TGAGCC": 1855, + "TGAGCG": 1856, + "TGAGGA": 1857, + "TGAGGT": 1858, + "TGAGGC": 1859, + "TGAGGG": 1860, + "TGTAAA": 1861, + "TGTAAT": 1862, + "TGTAAC": 1863, + "TGTAAG": 1864, + "TGTATA": 1865, + "TGTATT": 1866, + "TGTATC": 1867, + "TGTATG": 1868, + "TGTACA": 1869, + "TGTACT": 1870, + "TGTACC": 1871, + "TGTACG": 1872, + "TGTAGA": 1873, + "TGTAGT": 1874, + "TGTAGC": 1875, + "TGTAGG": 1876, + "TGTTAA": 1877, + "TGTTAT": 1878, + "TGTTAC": 1879, + "TGTTAG": 1880, + "TGTTTA": 1881, + "TGTTTT": 1882, + "TGTTTC": 1883, + "TGTTTG": 1884, + "TGTTCA": 1885, + "TGTTCT": 1886, + "TGTTCC": 1887, + "TGTTCG": 1888, + "TGTTGA": 1889, + "TGTTGT": 1890, + "TGTTGC": 1891, + "TGTTGG": 1892, + "TGTCAA": 1893, + "TGTCAT": 1894, + "TGTCAC": 1895, + "TGTCAG": 1896, + "TGTCTA": 1897, + "TGTCTT": 1898, + "TGTCTC": 1899, + "TGTCTG": 1900, + "TGTCCA": 1901, + "TGTCCT": 1902, + "TGTCCC": 1903, + "TGTCCG": 1904, + "TGTCGA": 1905, + "TGTCGT": 1906, + "TGTCGC": 1907, + "TGTCGG": 1908, + "TGTGAA": 1909, + "TGTGAT": 1910, + "TGTGAC": 1911, + "TGTGAG": 1912, + "TGTGTA": 1913, + "TGTGTT": 1914, + "TGTGTC": 1915, + "TGTGTG": 1916, + "TGTGCA": 1917, + "TGTGCT": 1918, + "TGTGCC": 1919, + "TGTGCG": 1920, + "TGTGGA": 1921, + "TGTGGT": 1922, + "TGTGGC": 1923, + "TGTGGG": 1924, + "TGCAAA": 1925, + "TGCAAT": 1926, + "TGCAAC": 1927, + "TGCAAG": 1928, + "TGCATA": 1929, + "TGCATT": 1930, + "TGCATC": 1931, + "TGCATG": 1932, + "TGCACA": 1933, + "TGCACT": 1934, + "TGCACC": 1935, + "TGCACG": 1936, + "TGCAGA": 1937, + "TGCAGT": 1938, + "TGCAGC": 1939, + "TGCAGG": 1940, + "TGCTAA": 1941, + "TGCTAT": 1942, + "TGCTAC": 1943, + "TGCTAG": 1944, + "TGCTTA": 1945, + "TGCTTT": 1946, + "TGCTTC": 1947, + "TGCTTG": 1948, + "TGCTCA": 1949, + "TGCTCT": 1950, + "TGCTCC": 1951, + "TGCTCG": 1952, + "TGCTGA": 1953, + "TGCTGT": 1954, + "TGCTGC": 1955, + "TGCTGG": 1956, + "TGCCAA": 1957, + "TGCCAT": 1958, + "TGCCAC": 1959, + "TGCCAG": 1960, + "TGCCTA": 1961, + "TGCCTT": 1962, + "TGCCTC": 1963, + "TGCCTG": 1964, + "TGCCCA": 1965, + "TGCCCT": 1966, + "TGCCCC": 1967, + "TGCCCG": 1968, + "TGCCGA": 1969, + "TGCCGT": 1970, + "TGCCGC": 1971, + "TGCCGG": 1972, + "TGCGAA": 1973, + "TGCGAT": 1974, + "TGCGAC": 1975, + "TGCGAG": 1976, + "TGCGTA": 1977, + "TGCGTT": 1978, + "TGCGTC": 1979, + "TGCGTG": 1980, + "TGCGCA": 1981, + "TGCGCT": 1982, + "TGCGCC": 1983, + "TGCGCG": 1984, + "TGCGGA": 1985, + "TGCGGT": 1986, + "TGCGGC": 1987, + "TGCGGG": 1988, + "TGGAAA": 1989, + "TGGAAT": 1990, + "TGGAAC": 1991, + "TGGAAG": 1992, + "TGGATA": 1993, + "TGGATT": 1994, + "TGGATC": 1995, + "TGGATG": 1996, + "TGGACA": 1997, + "TGGACT": 1998, + "TGGACC": 1999, + "TGGACG": 2000, + "TGGAGA": 2001, + "TGGAGT": 2002, + "TGGAGC": 2003, + "TGGAGG": 2004, + "TGGTAA": 2005, + "TGGTAT": 2006, + "TGGTAC": 2007, + "TGGTAG": 2008, + "TGGTTA": 2009, + "TGGTTT": 2010, + "TGGTTC": 2011, + "TGGTTG": 2012, + "TGGTCA": 2013, + "TGGTCT": 2014, + "TGGTCC": 2015, + "TGGTCG": 2016, + "TGGTGA": 2017, + "TGGTGT": 2018, + "TGGTGC": 2019, + "TGGTGG": 2020, + "TGGCAA": 2021, + "TGGCAT": 2022, + "TGGCAC": 2023, + "TGGCAG": 2024, + "TGGCTA": 2025, + "TGGCTT": 2026, + "TGGCTC": 2027, + "TGGCTG": 2028, + "TGGCCA": 2029, + "TGGCCT": 2030, + "TGGCCC": 2031, + "TGGCCG": 2032, + "TGGCGA": 2033, + "TGGCGT": 2034, + "TGGCGC": 2035, + "TGGCGG": 2036, + "TGGGAA": 2037, + "TGGGAT": 2038, + "TGGGAC": 2039, + "TGGGAG": 2040, + "TGGGTA": 2041, + "TGGGTT": 2042, + "TGGGTC": 2043, + "TGGGTG": 2044, + "TGGGCA": 2045, + "TGGGCT": 2046, + "TGGGCC": 2047, + "TGGGCG": 2048, + "TGGGGA": 2049, + "TGGGGT": 2050, + "TGGGGC": 2051, + "TGGGGG": 2052, + "CAAAAA": 2053, + "CAAAAT": 2054, + "CAAAAC": 2055, + "CAAAAG": 2056, + "CAAATA": 2057, + "CAAATT": 2058, + "CAAATC": 2059, + "CAAATG": 2060, + "CAAACA": 2061, + "CAAACT": 2062, + "CAAACC": 2063, + "CAAACG": 2064, + "CAAAGA": 2065, + "CAAAGT": 2066, + "CAAAGC": 2067, + "CAAAGG": 2068, + "CAATAA": 2069, + "CAATAT": 2070, + "CAATAC": 2071, + "CAATAG": 2072, + "CAATTA": 2073, + "CAATTT": 2074, + "CAATTC": 2075, + "CAATTG": 2076, + "CAATCA": 2077, + "CAATCT": 2078, + "CAATCC": 2079, + "CAATCG": 2080, + "CAATGA": 2081, + "CAATGT": 2082, + "CAATGC": 2083, + "CAATGG": 2084, + "CAACAA": 2085, + "CAACAT": 2086, + "CAACAC": 2087, + "CAACAG": 2088, + "CAACTA": 2089, + "CAACTT": 2090, + "CAACTC": 2091, + "CAACTG": 2092, + "CAACCA": 2093, + "CAACCT": 2094, + "CAACCC": 2095, + "CAACCG": 2096, + "CAACGA": 2097, + "CAACGT": 2098, + "CAACGC": 2099, + "CAACGG": 2100, + "CAAGAA": 2101, + "CAAGAT": 2102, + "CAAGAC": 2103, + "CAAGAG": 2104, + "CAAGTA": 2105, + "CAAGTT": 2106, + "CAAGTC": 2107, + "CAAGTG": 2108, + "CAAGCA": 2109, + "CAAGCT": 2110, + "CAAGCC": 2111, + "CAAGCG": 2112, + "CAAGGA": 2113, + "CAAGGT": 2114, + "CAAGGC": 2115, + "CAAGGG": 2116, + "CATAAA": 2117, + "CATAAT": 2118, + "CATAAC": 2119, + "CATAAG": 2120, + "CATATA": 2121, + "CATATT": 2122, + "CATATC": 2123, + "CATATG": 2124, + "CATACA": 2125, + "CATACT": 2126, + "CATACC": 2127, + "CATACG": 2128, + "CATAGA": 2129, + "CATAGT": 2130, + "CATAGC": 2131, + "CATAGG": 2132, + "CATTAA": 2133, + "CATTAT": 2134, + "CATTAC": 2135, + "CATTAG": 2136, + "CATTTA": 2137, + "CATTTT": 2138, + "CATTTC": 2139, + "CATTTG": 2140, + "CATTCA": 2141, + "CATTCT": 2142, + "CATTCC": 2143, + "CATTCG": 2144, + "CATTGA": 2145, + "CATTGT": 2146, + "CATTGC": 2147, + "CATTGG": 2148, + "CATCAA": 2149, + "CATCAT": 2150, + "CATCAC": 2151, + "CATCAG": 2152, + "CATCTA": 2153, + "CATCTT": 2154, + "CATCTC": 2155, + "CATCTG": 2156, + "CATCCA": 2157, + "CATCCT": 2158, + "CATCCC": 2159, + "CATCCG": 2160, + "CATCGA": 2161, + "CATCGT": 2162, + "CATCGC": 2163, + "CATCGG": 2164, + "CATGAA": 2165, + "CATGAT": 2166, + "CATGAC": 2167, + "CATGAG": 2168, + "CATGTA": 2169, + "CATGTT": 2170, + "CATGTC": 2171, + "CATGTG": 2172, + "CATGCA": 2173, + "CATGCT": 2174, + "CATGCC": 2175, + "CATGCG": 2176, + "CATGGA": 2177, + "CATGGT": 2178, + "CATGGC": 2179, + "CATGGG": 2180, + "CACAAA": 2181, + "CACAAT": 2182, + "CACAAC": 2183, + "CACAAG": 2184, + "CACATA": 2185, + "CACATT": 2186, + "CACATC": 2187, + "CACATG": 2188, + "CACACA": 2189, + "CACACT": 2190, + "CACACC": 2191, + "CACACG": 2192, + "CACAGA": 2193, + "CACAGT": 2194, + "CACAGC": 2195, + "CACAGG": 2196, + "CACTAA": 2197, + "CACTAT": 2198, + "CACTAC": 2199, + "CACTAG": 2200, + "CACTTA": 2201, + "CACTTT": 2202, + "CACTTC": 2203, + "CACTTG": 2204, + "CACTCA": 2205, + "CACTCT": 2206, + "CACTCC": 2207, + "CACTCG": 2208, + "CACTGA": 2209, + "CACTGT": 2210, + "CACTGC": 2211, + "CACTGG": 2212, + "CACCAA": 2213, + "CACCAT": 2214, + "CACCAC": 2215, + "CACCAG": 2216, + "CACCTA": 2217, + "CACCTT": 2218, + "CACCTC": 2219, + "CACCTG": 2220, + "CACCCA": 2221, + "CACCCT": 2222, + "CACCCC": 2223, + "CACCCG": 2224, + "CACCGA": 2225, + "CACCGT": 2226, + "CACCGC": 2227, + "CACCGG": 2228, + "CACGAA": 2229, + "CACGAT": 2230, + "CACGAC": 2231, + "CACGAG": 2232, + "CACGTA": 2233, + "CACGTT": 2234, + "CACGTC": 2235, + "CACGTG": 2236, + "CACGCA": 2237, + "CACGCT": 2238, + "CACGCC": 2239, + "CACGCG": 2240, + "CACGGA": 2241, + "CACGGT": 2242, + "CACGGC": 2243, + "CACGGG": 2244, + "CAGAAA": 2245, + "CAGAAT": 2246, + "CAGAAC": 2247, + "CAGAAG": 2248, + "CAGATA": 2249, + "CAGATT": 2250, + "CAGATC": 2251, + "CAGATG": 2252, + "CAGACA": 2253, + "CAGACT": 2254, + "CAGACC": 2255, + "CAGACG": 2256, + "CAGAGA": 2257, + "CAGAGT": 2258, + "CAGAGC": 2259, + "CAGAGG": 2260, + "CAGTAA": 2261, + "CAGTAT": 2262, + "CAGTAC": 2263, + "CAGTAG": 2264, + "CAGTTA": 2265, + "CAGTTT": 2266, + "CAGTTC": 2267, + "CAGTTG": 2268, + "CAGTCA": 2269, + "CAGTCT": 2270, + "CAGTCC": 2271, + "CAGTCG": 2272, + "CAGTGA": 2273, + "CAGTGT": 2274, + "CAGTGC": 2275, + "CAGTGG": 2276, + "CAGCAA": 2277, + "CAGCAT": 2278, + "CAGCAC": 2279, + "CAGCAG": 2280, + "CAGCTA": 2281, + "CAGCTT": 2282, + "CAGCTC": 2283, + "CAGCTG": 2284, + "CAGCCA": 2285, + "CAGCCT": 2286, + "CAGCCC": 2287, + "CAGCCG": 2288, + "CAGCGA": 2289, + "CAGCGT": 2290, + "CAGCGC": 2291, + "CAGCGG": 2292, + "CAGGAA": 2293, + "CAGGAT": 2294, + "CAGGAC": 2295, + "CAGGAG": 2296, + "CAGGTA": 2297, + "CAGGTT": 2298, + "CAGGTC": 2299, + "CAGGTG": 2300, + "CAGGCA": 2301, + "CAGGCT": 2302, + "CAGGCC": 2303, + "CAGGCG": 2304, + "CAGGGA": 2305, + "CAGGGT": 2306, + "CAGGGC": 2307, + "CAGGGG": 2308, + "CTAAAA": 2309, + "CTAAAT": 2310, + "CTAAAC": 2311, + "CTAAAG": 2312, + "CTAATA": 2313, + "CTAATT": 2314, + "CTAATC": 2315, + "CTAATG": 2316, + "CTAACA": 2317, + "CTAACT": 2318, + "CTAACC": 2319, + "CTAACG": 2320, + "CTAAGA": 2321, + "CTAAGT": 2322, + "CTAAGC": 2323, + "CTAAGG": 2324, + "CTATAA": 2325, + "CTATAT": 2326, + "CTATAC": 2327, + "CTATAG": 2328, + "CTATTA": 2329, + "CTATTT": 2330, + "CTATTC": 2331, + "CTATTG": 2332, + "CTATCA": 2333, + "CTATCT": 2334, + "CTATCC": 2335, + "CTATCG": 2336, + "CTATGA": 2337, + "CTATGT": 2338, + "CTATGC": 2339, + "CTATGG": 2340, + "CTACAA": 2341, + "CTACAT": 2342, + "CTACAC": 2343, + "CTACAG": 2344, + "CTACTA": 2345, + "CTACTT": 2346, + "CTACTC": 2347, + "CTACTG": 2348, + "CTACCA": 2349, + "CTACCT": 2350, + "CTACCC": 2351, + "CTACCG": 2352, + "CTACGA": 2353, + "CTACGT": 2354, + "CTACGC": 2355, + "CTACGG": 2356, + "CTAGAA": 2357, + "CTAGAT": 2358, + "CTAGAC": 2359, + "CTAGAG": 2360, + "CTAGTA": 2361, + "CTAGTT": 2362, + "CTAGTC": 2363, + "CTAGTG": 2364, + "CTAGCA": 2365, + "CTAGCT": 2366, + "CTAGCC": 2367, + "CTAGCG": 2368, + "CTAGGA": 2369, + "CTAGGT": 2370, + "CTAGGC": 2371, + "CTAGGG": 2372, + "CTTAAA": 2373, + "CTTAAT": 2374, + "CTTAAC": 2375, + "CTTAAG": 2376, + "CTTATA": 2377, + "CTTATT": 2378, + "CTTATC": 2379, + "CTTATG": 2380, + "CTTACA": 2381, + "CTTACT": 2382, + "CTTACC": 2383, + "CTTACG": 2384, + "CTTAGA": 2385, + "CTTAGT": 2386, + "CTTAGC": 2387, + "CTTAGG": 2388, + "CTTTAA": 2389, + "CTTTAT": 2390, + "CTTTAC": 2391, + "CTTTAG": 2392, + "CTTTTA": 2393, + "CTTTTT": 2394, + "CTTTTC": 2395, + "CTTTTG": 2396, + "CTTTCA": 2397, + "CTTTCT": 2398, + "CTTTCC": 2399, + "CTTTCG": 2400, + "CTTTGA": 2401, + "CTTTGT": 2402, + "CTTTGC": 2403, + "CTTTGG": 2404, + "CTTCAA": 2405, + "CTTCAT": 2406, + "CTTCAC": 2407, + "CTTCAG": 2408, + "CTTCTA": 2409, + "CTTCTT": 2410, + "CTTCTC": 2411, + "CTTCTG": 2412, + "CTTCCA": 2413, + "CTTCCT": 2414, + "CTTCCC": 2415, + "CTTCCG": 2416, + "CTTCGA": 2417, + "CTTCGT": 2418, + "CTTCGC": 2419, + "CTTCGG": 2420, + "CTTGAA": 2421, + "CTTGAT": 2422, + "CTTGAC": 2423, + "CTTGAG": 2424, + "CTTGTA": 2425, + "CTTGTT": 2426, + "CTTGTC": 2427, + "CTTGTG": 2428, + "CTTGCA": 2429, + "CTTGCT": 2430, + "CTTGCC": 2431, + "CTTGCG": 2432, + "CTTGGA": 2433, + "CTTGGT": 2434, + "CTTGGC": 2435, + "CTTGGG": 2436, + "CTCAAA": 2437, + "CTCAAT": 2438, + "CTCAAC": 2439, + "CTCAAG": 2440, + "CTCATA": 2441, + "CTCATT": 2442, + "CTCATC": 2443, + "CTCATG": 2444, + "CTCACA": 2445, + "CTCACT": 2446, + "CTCACC": 2447, + "CTCACG": 2448, + "CTCAGA": 2449, + "CTCAGT": 2450, + "CTCAGC": 2451, + "CTCAGG": 2452, + "CTCTAA": 2453, + "CTCTAT": 2454, + "CTCTAC": 2455, + "CTCTAG": 2456, + "CTCTTA": 2457, + "CTCTTT": 2458, + "CTCTTC": 2459, + "CTCTTG": 2460, + "CTCTCA": 2461, + "CTCTCT": 2462, + "CTCTCC": 2463, + "CTCTCG": 2464, + "CTCTGA": 2465, + "CTCTGT": 2466, + "CTCTGC": 2467, + "CTCTGG": 2468, + "CTCCAA": 2469, + "CTCCAT": 2470, + "CTCCAC": 2471, + "CTCCAG": 2472, + "CTCCTA": 2473, + "CTCCTT": 2474, + "CTCCTC": 2475, + "CTCCTG": 2476, + "CTCCCA": 2477, + "CTCCCT": 2478, + "CTCCCC": 2479, + "CTCCCG": 2480, + "CTCCGA": 2481, + "CTCCGT": 2482, + "CTCCGC": 2483, + "CTCCGG": 2484, + "CTCGAA": 2485, + "CTCGAT": 2486, + "CTCGAC": 2487, + "CTCGAG": 2488, + "CTCGTA": 2489, + "CTCGTT": 2490, + "CTCGTC": 2491, + "CTCGTG": 2492, + "CTCGCA": 2493, + "CTCGCT": 2494, + "CTCGCC": 2495, + "CTCGCG": 2496, + "CTCGGA": 2497, + "CTCGGT": 2498, + "CTCGGC": 2499, + "CTCGGG": 2500, + "CTGAAA": 2501, + "CTGAAT": 2502, + "CTGAAC": 2503, + "CTGAAG": 2504, + "CTGATA": 2505, + "CTGATT": 2506, + "CTGATC": 2507, + "CTGATG": 2508, + "CTGACA": 2509, + "CTGACT": 2510, + "CTGACC": 2511, + "CTGACG": 2512, + "CTGAGA": 2513, + "CTGAGT": 2514, + "CTGAGC": 2515, + "CTGAGG": 2516, + "CTGTAA": 2517, + "CTGTAT": 2518, + "CTGTAC": 2519, + "CTGTAG": 2520, + "CTGTTA": 2521, + "CTGTTT": 2522, + "CTGTTC": 2523, + "CTGTTG": 2524, + "CTGTCA": 2525, + "CTGTCT": 2526, + "CTGTCC": 2527, + "CTGTCG": 2528, + "CTGTGA": 2529, + "CTGTGT": 2530, + "CTGTGC": 2531, + "CTGTGG": 2532, + "CTGCAA": 2533, + "CTGCAT": 2534, + "CTGCAC": 2535, + "CTGCAG": 2536, + "CTGCTA": 2537, + "CTGCTT": 2538, + "CTGCTC": 2539, + "CTGCTG": 2540, + "CTGCCA": 2541, + "CTGCCT": 2542, + "CTGCCC": 2543, + "CTGCCG": 2544, + "CTGCGA": 2545, + "CTGCGT": 2546, + "CTGCGC": 2547, + "CTGCGG": 2548, + "CTGGAA": 2549, + "CTGGAT": 2550, + "CTGGAC": 2551, + "CTGGAG": 2552, + "CTGGTA": 2553, + "CTGGTT": 2554, + "CTGGTC": 2555, + "CTGGTG": 2556, + "CTGGCA": 2557, + "CTGGCT": 2558, + "CTGGCC": 2559, + "CTGGCG": 2560, + "CTGGGA": 2561, + "CTGGGT": 2562, + "CTGGGC": 2563, + "CTGGGG": 2564, + "CCAAAA": 2565, + "CCAAAT": 2566, + "CCAAAC": 2567, + "CCAAAG": 2568, + "CCAATA": 2569, + "CCAATT": 2570, + "CCAATC": 2571, + "CCAATG": 2572, + "CCAACA": 2573, + "CCAACT": 2574, + "CCAACC": 2575, + "CCAACG": 2576, + "CCAAGA": 2577, + "CCAAGT": 2578, + "CCAAGC": 2579, + "CCAAGG": 2580, + "CCATAA": 2581, + "CCATAT": 2582, + "CCATAC": 2583, + "CCATAG": 2584, + "CCATTA": 2585, + "CCATTT": 2586, + "CCATTC": 2587, + "CCATTG": 2588, + "CCATCA": 2589, + "CCATCT": 2590, + "CCATCC": 2591, + "CCATCG": 2592, + "CCATGA": 2593, + "CCATGT": 2594, + "CCATGC": 2595, + "CCATGG": 2596, + "CCACAA": 2597, + "CCACAT": 2598, + "CCACAC": 2599, + "CCACAG": 2600, + "CCACTA": 2601, + "CCACTT": 2602, + "CCACTC": 2603, + "CCACTG": 2604, + "CCACCA": 2605, + "CCACCT": 2606, + "CCACCC": 2607, + "CCACCG": 2608, + "CCACGA": 2609, + "CCACGT": 2610, + "CCACGC": 2611, + "CCACGG": 2612, + "CCAGAA": 2613, + "CCAGAT": 2614, + "CCAGAC": 2615, + "CCAGAG": 2616, + "CCAGTA": 2617, + "CCAGTT": 2618, + "CCAGTC": 2619, + "CCAGTG": 2620, + "CCAGCA": 2621, + "CCAGCT": 2622, + "CCAGCC": 2623, + "CCAGCG": 2624, + "CCAGGA": 2625, + "CCAGGT": 2626, + "CCAGGC": 2627, + "CCAGGG": 2628, + "CCTAAA": 2629, + "CCTAAT": 2630, + "CCTAAC": 2631, + "CCTAAG": 2632, + "CCTATA": 2633, + "CCTATT": 2634, + "CCTATC": 2635, + "CCTATG": 2636, + "CCTACA": 2637, + "CCTACT": 2638, + "CCTACC": 2639, + "CCTACG": 2640, + "CCTAGA": 2641, + "CCTAGT": 2642, + "CCTAGC": 2643, + "CCTAGG": 2644, + "CCTTAA": 2645, + "CCTTAT": 2646, + "CCTTAC": 2647, + "CCTTAG": 2648, + "CCTTTA": 2649, + "CCTTTT": 2650, + "CCTTTC": 2651, + "CCTTTG": 2652, + "CCTTCA": 2653, + "CCTTCT": 2654, + "CCTTCC": 2655, + "CCTTCG": 2656, + "CCTTGA": 2657, + "CCTTGT": 2658, + "CCTTGC": 2659, + "CCTTGG": 2660, + "CCTCAA": 2661, + "CCTCAT": 2662, + "CCTCAC": 2663, + "CCTCAG": 2664, + "CCTCTA": 2665, + "CCTCTT": 2666, + "CCTCTC": 2667, + "CCTCTG": 2668, + "CCTCCA": 2669, + "CCTCCT": 2670, + "CCTCCC": 2671, + "CCTCCG": 2672, + "CCTCGA": 2673, + "CCTCGT": 2674, + "CCTCGC": 2675, + "CCTCGG": 2676, + "CCTGAA": 2677, + "CCTGAT": 2678, + "CCTGAC": 2679, + "CCTGAG": 2680, + "CCTGTA": 2681, + "CCTGTT": 2682, + "CCTGTC": 2683, + "CCTGTG": 2684, + "CCTGCA": 2685, + "CCTGCT": 2686, + "CCTGCC": 2687, + "CCTGCG": 2688, + "CCTGGA": 2689, + "CCTGGT": 2690, + "CCTGGC": 2691, + "CCTGGG": 2692, + "CCCAAA": 2693, + "CCCAAT": 2694, + "CCCAAC": 2695, + "CCCAAG": 2696, + "CCCATA": 2697, + "CCCATT": 2698, + "CCCATC": 2699, + "CCCATG": 2700, + "CCCACA": 2701, + "CCCACT": 2702, + "CCCACC": 2703, + "CCCACG": 2704, + "CCCAGA": 2705, + "CCCAGT": 2706, + "CCCAGC": 2707, + "CCCAGG": 2708, + "CCCTAA": 2709, + "CCCTAT": 2710, + "CCCTAC": 2711, + "CCCTAG": 2712, + "CCCTTA": 2713, + "CCCTTT": 2714, + "CCCTTC": 2715, + "CCCTTG": 2716, + "CCCTCA": 2717, + "CCCTCT": 2718, + "CCCTCC": 2719, + "CCCTCG": 2720, + "CCCTGA": 2721, + "CCCTGT": 2722, + "CCCTGC": 2723, + "CCCTGG": 2724, + "CCCCAA": 2725, + "CCCCAT": 2726, + "CCCCAC": 2727, + "CCCCAG": 2728, + "CCCCTA": 2729, + "CCCCTT": 2730, + "CCCCTC": 2731, + "CCCCTG": 2732, + "CCCCCA": 2733, + "CCCCCT": 2734, + "CCCCCC": 2735, + "CCCCCG": 2736, + "CCCCGA": 2737, + "CCCCGT": 2738, + "CCCCGC": 2739, + "CCCCGG": 2740, + "CCCGAA": 2741, + "CCCGAT": 2742, + "CCCGAC": 2743, + "CCCGAG": 2744, + "CCCGTA": 2745, + "CCCGTT": 2746, + "CCCGTC": 2747, + "CCCGTG": 2748, + "CCCGCA": 2749, + "CCCGCT": 2750, + "CCCGCC": 2751, + "CCCGCG": 2752, + "CCCGGA": 2753, + "CCCGGT": 2754, + "CCCGGC": 2755, + "CCCGGG": 2756, + "CCGAAA": 2757, + "CCGAAT": 2758, + "CCGAAC": 2759, + "CCGAAG": 2760, + "CCGATA": 2761, + "CCGATT": 2762, + "CCGATC": 2763, + "CCGATG": 2764, + "CCGACA": 2765, + "CCGACT": 2766, + "CCGACC": 2767, + "CCGACG": 2768, + "CCGAGA": 2769, + "CCGAGT": 2770, + "CCGAGC": 2771, + "CCGAGG": 2772, + "CCGTAA": 2773, + "CCGTAT": 2774, + "CCGTAC": 2775, + "CCGTAG": 2776, + "CCGTTA": 2777, + "CCGTTT": 2778, + "CCGTTC": 2779, + "CCGTTG": 2780, + "CCGTCA": 2781, + "CCGTCT": 2782, + "CCGTCC": 2783, + "CCGTCG": 2784, + "CCGTGA": 2785, + "CCGTGT": 2786, + "CCGTGC": 2787, + "CCGTGG": 2788, + "CCGCAA": 2789, + "CCGCAT": 2790, + "CCGCAC": 2791, + "CCGCAG": 2792, + "CCGCTA": 2793, + "CCGCTT": 2794, + "CCGCTC": 2795, + "CCGCTG": 2796, + "CCGCCA": 2797, + "CCGCCT": 2798, + "CCGCCC": 2799, + "CCGCCG": 2800, + "CCGCGA": 2801, + "CCGCGT": 2802, + "CCGCGC": 2803, + "CCGCGG": 2804, + "CCGGAA": 2805, + "CCGGAT": 2806, + "CCGGAC": 2807, + "CCGGAG": 2808, + "CCGGTA": 2809, + "CCGGTT": 2810, + "CCGGTC": 2811, + "CCGGTG": 2812, + "CCGGCA": 2813, + "CCGGCT": 2814, + "CCGGCC": 2815, + "CCGGCG": 2816, + "CCGGGA": 2817, + "CCGGGT": 2818, + "CCGGGC": 2819, + "CCGGGG": 2820, + "CGAAAA": 2821, + "CGAAAT": 2822, + "CGAAAC": 2823, + "CGAAAG": 2824, + "CGAATA": 2825, + "CGAATT": 2826, + "CGAATC": 2827, + "CGAATG": 2828, + "CGAACA": 2829, + "CGAACT": 2830, + "CGAACC": 2831, + "CGAACG": 2832, + "CGAAGA": 2833, + "CGAAGT": 2834, + "CGAAGC": 2835, + "CGAAGG": 2836, + "CGATAA": 2837, + "CGATAT": 2838, + "CGATAC": 2839, + "CGATAG": 2840, + "CGATTA": 2841, + "CGATTT": 2842, + "CGATTC": 2843, + "CGATTG": 2844, + "CGATCA": 2845, + "CGATCT": 2846, + "CGATCC": 2847, + "CGATCG": 2848, + "CGATGA": 2849, + "CGATGT": 2850, + "CGATGC": 2851, + "CGATGG": 2852, + "CGACAA": 2853, + "CGACAT": 2854, + "CGACAC": 2855, + "CGACAG": 2856, + "CGACTA": 2857, + "CGACTT": 2858, + "CGACTC": 2859, + "CGACTG": 2860, + "CGACCA": 2861, + "CGACCT": 2862, + "CGACCC": 2863, + "CGACCG": 2864, + "CGACGA": 2865, + "CGACGT": 2866, + "CGACGC": 2867, + "CGACGG": 2868, + "CGAGAA": 2869, + "CGAGAT": 2870, + "CGAGAC": 2871, + "CGAGAG": 2872, + "CGAGTA": 2873, + "CGAGTT": 2874, + "CGAGTC": 2875, + "CGAGTG": 2876, + "CGAGCA": 2877, + "CGAGCT": 2878, + "CGAGCC": 2879, + "CGAGCG": 2880, + "CGAGGA": 2881, + "CGAGGT": 2882, + "CGAGGC": 2883, + "CGAGGG": 2884, + "CGTAAA": 2885, + "CGTAAT": 2886, + "CGTAAC": 2887, + "CGTAAG": 2888, + "CGTATA": 2889, + "CGTATT": 2890, + "CGTATC": 2891, + "CGTATG": 2892, + "CGTACA": 2893, + "CGTACT": 2894, + "CGTACC": 2895, + "CGTACG": 2896, + "CGTAGA": 2897, + "CGTAGT": 2898, + "CGTAGC": 2899, + "CGTAGG": 2900, + "CGTTAA": 2901, + "CGTTAT": 2902, + "CGTTAC": 2903, + "CGTTAG": 2904, + "CGTTTA": 2905, + "CGTTTT": 2906, + "CGTTTC": 2907, + "CGTTTG": 2908, + "CGTTCA": 2909, + "CGTTCT": 2910, + "CGTTCC": 2911, + "CGTTCG": 2912, + "CGTTGA": 2913, + "CGTTGT": 2914, + "CGTTGC": 2915, + "CGTTGG": 2916, + "CGTCAA": 2917, + "CGTCAT": 2918, + "CGTCAC": 2919, + "CGTCAG": 2920, + "CGTCTA": 2921, + "CGTCTT": 2922, + "CGTCTC": 2923, + "CGTCTG": 2924, + "CGTCCA": 2925, + "CGTCCT": 2926, + "CGTCCC": 2927, + "CGTCCG": 2928, + "CGTCGA": 2929, + "CGTCGT": 2930, + "CGTCGC": 2931, + "CGTCGG": 2932, + "CGTGAA": 2933, + "CGTGAT": 2934, + "CGTGAC": 2935, + "CGTGAG": 2936, + "CGTGTA": 2937, + "CGTGTT": 2938, + "CGTGTC": 2939, + "CGTGTG": 2940, + "CGTGCA": 2941, + "CGTGCT": 2942, + "CGTGCC": 2943, + "CGTGCG": 2944, + "CGTGGA": 2945, + "CGTGGT": 2946, + "CGTGGC": 2947, + "CGTGGG": 2948, + "CGCAAA": 2949, + "CGCAAT": 2950, + "CGCAAC": 2951, + "CGCAAG": 2952, + "CGCATA": 2953, + "CGCATT": 2954, + "CGCATC": 2955, + "CGCATG": 2956, + "CGCACA": 2957, + "CGCACT": 2958, + "CGCACC": 2959, + "CGCACG": 2960, + "CGCAGA": 2961, + "CGCAGT": 2962, + "CGCAGC": 2963, + "CGCAGG": 2964, + "CGCTAA": 2965, + "CGCTAT": 2966, + "CGCTAC": 2967, + "CGCTAG": 2968, + "CGCTTA": 2969, + "CGCTTT": 2970, + "CGCTTC": 2971, + "CGCTTG": 2972, + "CGCTCA": 2973, + "CGCTCT": 2974, + "CGCTCC": 2975, + "CGCTCG": 2976, + "CGCTGA": 2977, + "CGCTGT": 2978, + "CGCTGC": 2979, + "CGCTGG": 2980, + "CGCCAA": 2981, + "CGCCAT": 2982, + "CGCCAC": 2983, + "CGCCAG": 2984, + "CGCCTA": 2985, + "CGCCTT": 2986, + "CGCCTC": 2987, + "CGCCTG": 2988, + "CGCCCA": 2989, + "CGCCCT": 2990, + "CGCCCC": 2991, + "CGCCCG": 2992, + "CGCCGA": 2993, + "CGCCGT": 2994, + "CGCCGC": 2995, + "CGCCGG": 2996, + "CGCGAA": 2997, + "CGCGAT": 2998, + "CGCGAC": 2999, + "CGCGAG": 3000, + "CGCGTA": 3001, + "CGCGTT": 3002, + "CGCGTC": 3003, + "CGCGTG": 3004, + "CGCGCA": 3005, + "CGCGCT": 3006, + "CGCGCC": 3007, + "CGCGCG": 3008, + "CGCGGA": 3009, + "CGCGGT": 3010, + "CGCGGC": 3011, + "CGCGGG": 3012, + "CGGAAA": 3013, + "CGGAAT": 3014, + "CGGAAC": 3015, + "CGGAAG": 3016, + "CGGATA": 3017, + "CGGATT": 3018, + "CGGATC": 3019, + "CGGATG": 3020, + "CGGACA": 3021, + "CGGACT": 3022, + "CGGACC": 3023, + "CGGACG": 3024, + "CGGAGA": 3025, + "CGGAGT": 3026, + "CGGAGC": 3027, + "CGGAGG": 3028, + "CGGTAA": 3029, + "CGGTAT": 3030, + "CGGTAC": 3031, + "CGGTAG": 3032, + "CGGTTA": 3033, + "CGGTTT": 3034, + "CGGTTC": 3035, + "CGGTTG": 3036, + "CGGTCA": 3037, + "CGGTCT": 3038, + "CGGTCC": 3039, + "CGGTCG": 3040, + "CGGTGA": 3041, + "CGGTGT": 3042, + "CGGTGC": 3043, + "CGGTGG": 3044, + "CGGCAA": 3045, + "CGGCAT": 3046, + "CGGCAC": 3047, + "CGGCAG": 3048, + "CGGCTA": 3049, + "CGGCTT": 3050, + "CGGCTC": 3051, + "CGGCTG": 3052, + "CGGCCA": 3053, + "CGGCCT": 3054, + "CGGCCC": 3055, + "CGGCCG": 3056, + "CGGCGA": 3057, + "CGGCGT": 3058, + "CGGCGC": 3059, + "CGGCGG": 3060, + "CGGGAA": 3061, + "CGGGAT": 3062, + "CGGGAC": 3063, + "CGGGAG": 3064, + "CGGGTA": 3065, + "CGGGTT": 3066, + "CGGGTC": 3067, + "CGGGTG": 3068, + "CGGGCA": 3069, + "CGGGCT": 3070, + "CGGGCC": 3071, + "CGGGCG": 3072, + "CGGGGA": 3073, + "CGGGGT": 3074, + "CGGGGC": 3075, + "CGGGGG": 3076, + "GAAAAA": 3077, + "GAAAAT": 3078, + "GAAAAC": 3079, + "GAAAAG": 3080, + "GAAATA": 3081, + "GAAATT": 3082, + "GAAATC": 3083, + "GAAATG": 3084, + "GAAACA": 3085, + "GAAACT": 3086, + "GAAACC": 3087, + "GAAACG": 3088, + "GAAAGA": 3089, + "GAAAGT": 3090, + "GAAAGC": 3091, + "GAAAGG": 3092, + "GAATAA": 3093, + "GAATAT": 3094, + "GAATAC": 3095, + "GAATAG": 3096, + "GAATTA": 3097, + "GAATTT": 3098, + "GAATTC": 3099, + "GAATTG": 3100, + "GAATCA": 3101, + "GAATCT": 3102, + "GAATCC": 3103, + "GAATCG": 3104, + "GAATGA": 3105, + "GAATGT": 3106, + "GAATGC": 3107, + "GAATGG": 3108, + "GAACAA": 3109, + "GAACAT": 3110, + "GAACAC": 3111, + "GAACAG": 3112, + "GAACTA": 3113, + "GAACTT": 3114, + "GAACTC": 3115, + "GAACTG": 3116, + "GAACCA": 3117, + "GAACCT": 3118, + "GAACCC": 3119, + "GAACCG": 3120, + "GAACGA": 3121, + "GAACGT": 3122, + "GAACGC": 3123, + "GAACGG": 3124, + "GAAGAA": 3125, + "GAAGAT": 3126, + "GAAGAC": 3127, + "GAAGAG": 3128, + "GAAGTA": 3129, + "GAAGTT": 3130, + "GAAGTC": 3131, + "GAAGTG": 3132, + "GAAGCA": 3133, + "GAAGCT": 3134, + "GAAGCC": 3135, + "GAAGCG": 3136, + "GAAGGA": 3137, + "GAAGGT": 3138, + "GAAGGC": 3139, + "GAAGGG": 3140, + "GATAAA": 3141, + "GATAAT": 3142, + "GATAAC": 3143, + "GATAAG": 3144, + "GATATA": 3145, + "GATATT": 3146, + "GATATC": 3147, + "GATATG": 3148, + "GATACA": 3149, + "GATACT": 3150, + "GATACC": 3151, + "GATACG": 3152, + "GATAGA": 3153, + "GATAGT": 3154, + "GATAGC": 3155, + "GATAGG": 3156, + "GATTAA": 3157, + "GATTAT": 3158, + "GATTAC": 3159, + "GATTAG": 3160, + "GATTTA": 3161, + "GATTTT": 3162, + "GATTTC": 3163, + "GATTTG": 3164, + "GATTCA": 3165, + "GATTCT": 3166, + "GATTCC": 3167, + "GATTCG": 3168, + "GATTGA": 3169, + "GATTGT": 3170, + "GATTGC": 3171, + "GATTGG": 3172, + "GATCAA": 3173, + "GATCAT": 3174, + "GATCAC": 3175, + "GATCAG": 3176, + "GATCTA": 3177, + "GATCTT": 3178, + "GATCTC": 3179, + "GATCTG": 3180, + "GATCCA": 3181, + "GATCCT": 3182, + "GATCCC": 3183, + "GATCCG": 3184, + "GATCGA": 3185, + "GATCGT": 3186, + "GATCGC": 3187, + "GATCGG": 3188, + "GATGAA": 3189, + "GATGAT": 3190, + "GATGAC": 3191, + "GATGAG": 3192, + "GATGTA": 3193, + "GATGTT": 3194, + "GATGTC": 3195, + "GATGTG": 3196, + "GATGCA": 3197, + "GATGCT": 3198, + "GATGCC": 3199, + "GATGCG": 3200, + "GATGGA": 3201, + "GATGGT": 3202, + "GATGGC": 3203, + "GATGGG": 3204, + "GACAAA": 3205, + "GACAAT": 3206, + "GACAAC": 3207, + "GACAAG": 3208, + "GACATA": 3209, + "GACATT": 3210, + "GACATC": 3211, + "GACATG": 3212, + "GACACA": 3213, + "GACACT": 3214, + "GACACC": 3215, + "GACACG": 3216, + "GACAGA": 3217, + "GACAGT": 3218, + "GACAGC": 3219, + "GACAGG": 3220, + "GACTAA": 3221, + "GACTAT": 3222, + "GACTAC": 3223, + "GACTAG": 3224, + "GACTTA": 3225, + "GACTTT": 3226, + "GACTTC": 3227, + "GACTTG": 3228, + "GACTCA": 3229, + "GACTCT": 3230, + "GACTCC": 3231, + "GACTCG": 3232, + "GACTGA": 3233, + "GACTGT": 3234, + "GACTGC": 3235, + "GACTGG": 3236, + "GACCAA": 3237, + "GACCAT": 3238, + "GACCAC": 3239, + "GACCAG": 3240, + "GACCTA": 3241, + "GACCTT": 3242, + "GACCTC": 3243, + "GACCTG": 3244, + "GACCCA": 3245, + "GACCCT": 3246, + "GACCCC": 3247, + "GACCCG": 3248, + "GACCGA": 3249, + "GACCGT": 3250, + "GACCGC": 3251, + "GACCGG": 3252, + "GACGAA": 3253, + "GACGAT": 3254, + "GACGAC": 3255, + "GACGAG": 3256, + "GACGTA": 3257, + "GACGTT": 3258, + "GACGTC": 3259, + "GACGTG": 3260, + "GACGCA": 3261, + "GACGCT": 3262, + "GACGCC": 3263, + "GACGCG": 3264, + "GACGGA": 3265, + "GACGGT": 3266, + "GACGGC": 3267, + "GACGGG": 3268, + "GAGAAA": 3269, + "GAGAAT": 3270, + "GAGAAC": 3271, + "GAGAAG": 3272, + "GAGATA": 3273, + "GAGATT": 3274, + "GAGATC": 3275, + "GAGATG": 3276, + "GAGACA": 3277, + "GAGACT": 3278, + "GAGACC": 3279, + "GAGACG": 3280, + "GAGAGA": 3281, + "GAGAGT": 3282, + "GAGAGC": 3283, + "GAGAGG": 3284, + "GAGTAA": 3285, + "GAGTAT": 3286, + "GAGTAC": 3287, + "GAGTAG": 3288, + "GAGTTA": 3289, + "GAGTTT": 3290, + "GAGTTC": 3291, + "GAGTTG": 3292, + "GAGTCA": 3293, + "GAGTCT": 3294, + "GAGTCC": 3295, + "GAGTCG": 3296, + "GAGTGA": 3297, + "GAGTGT": 3298, + "GAGTGC": 3299, + "GAGTGG": 3300, + "GAGCAA": 3301, + "GAGCAT": 3302, + "GAGCAC": 3303, + "GAGCAG": 3304, + "GAGCTA": 3305, + "GAGCTT": 3306, + "GAGCTC": 3307, + "GAGCTG": 3308, + "GAGCCA": 3309, + "GAGCCT": 3310, + "GAGCCC": 3311, + "GAGCCG": 3312, + "GAGCGA": 3313, + "GAGCGT": 3314, + "GAGCGC": 3315, + "GAGCGG": 3316, + "GAGGAA": 3317, + "GAGGAT": 3318, + "GAGGAC": 3319, + "GAGGAG": 3320, + "GAGGTA": 3321, + "GAGGTT": 3322, + "GAGGTC": 3323, + "GAGGTG": 3324, + "GAGGCA": 3325, + "GAGGCT": 3326, + "GAGGCC": 3327, + "GAGGCG": 3328, + "GAGGGA": 3329, + "GAGGGT": 3330, + "GAGGGC": 3331, + "GAGGGG": 3332, + "GTAAAA": 3333, + "GTAAAT": 3334, + "GTAAAC": 3335, + "GTAAAG": 3336, + "GTAATA": 3337, + "GTAATT": 3338, + "GTAATC": 3339, + "GTAATG": 3340, + "GTAACA": 3341, + "GTAACT": 3342, + "GTAACC": 3343, + "GTAACG": 3344, + "GTAAGA": 3345, + "GTAAGT": 3346, + "GTAAGC": 3347, + "GTAAGG": 3348, + "GTATAA": 3349, + "GTATAT": 3350, + "GTATAC": 3351, + "GTATAG": 3352, + "GTATTA": 3353, + "GTATTT": 3354, + "GTATTC": 3355, + "GTATTG": 3356, + "GTATCA": 3357, + "GTATCT": 3358, + "GTATCC": 3359, + "GTATCG": 3360, + "GTATGA": 3361, + "GTATGT": 3362, + "GTATGC": 3363, + "GTATGG": 3364, + "GTACAA": 3365, + "GTACAT": 3366, + "GTACAC": 3367, + "GTACAG": 3368, + "GTACTA": 3369, + "GTACTT": 3370, + "GTACTC": 3371, + "GTACTG": 3372, + "GTACCA": 3373, + "GTACCT": 3374, + "GTACCC": 3375, + "GTACCG": 3376, + "GTACGA": 3377, + "GTACGT": 3378, + "GTACGC": 3379, + "GTACGG": 3380, + "GTAGAA": 3381, + "GTAGAT": 3382, + "GTAGAC": 3383, + "GTAGAG": 3384, + "GTAGTA": 3385, + "GTAGTT": 3386, + "GTAGTC": 3387, + "GTAGTG": 3388, + "GTAGCA": 3389, + "GTAGCT": 3390, + "GTAGCC": 3391, + "GTAGCG": 3392, + "GTAGGA": 3393, + "GTAGGT": 3394, + "GTAGGC": 3395, + "GTAGGG": 3396, + "GTTAAA": 3397, + "GTTAAT": 3398, + "GTTAAC": 3399, + "GTTAAG": 3400, + "GTTATA": 3401, + "GTTATT": 3402, + "GTTATC": 3403, + "GTTATG": 3404, + "GTTACA": 3405, + "GTTACT": 3406, + "GTTACC": 3407, + "GTTACG": 3408, + "GTTAGA": 3409, + "GTTAGT": 3410, + "GTTAGC": 3411, + "GTTAGG": 3412, + "GTTTAA": 3413, + "GTTTAT": 3414, + "GTTTAC": 3415, + "GTTTAG": 3416, + "GTTTTA": 3417, + "GTTTTT": 3418, + "GTTTTC": 3419, + "GTTTTG": 3420, + "GTTTCA": 3421, + "GTTTCT": 3422, + "GTTTCC": 3423, + "GTTTCG": 3424, + "GTTTGA": 3425, + "GTTTGT": 3426, + "GTTTGC": 3427, + "GTTTGG": 3428, + "GTTCAA": 3429, + "GTTCAT": 3430, + "GTTCAC": 3431, + "GTTCAG": 3432, + "GTTCTA": 3433, + "GTTCTT": 3434, + "GTTCTC": 3435, + "GTTCTG": 3436, + "GTTCCA": 3437, + "GTTCCT": 3438, + "GTTCCC": 3439, + "GTTCCG": 3440, + "GTTCGA": 3441, + "GTTCGT": 3442, + "GTTCGC": 3443, + "GTTCGG": 3444, + "GTTGAA": 3445, + "GTTGAT": 3446, + "GTTGAC": 3447, + "GTTGAG": 3448, + "GTTGTA": 3449, + "GTTGTT": 3450, + "GTTGTC": 3451, + "GTTGTG": 3452, + "GTTGCA": 3453, + "GTTGCT": 3454, + "GTTGCC": 3455, + "GTTGCG": 3456, + "GTTGGA": 3457, + "GTTGGT": 3458, + "GTTGGC": 3459, + "GTTGGG": 3460, + "GTCAAA": 3461, + "GTCAAT": 3462, + "GTCAAC": 3463, + "GTCAAG": 3464, + "GTCATA": 3465, + "GTCATT": 3466, + "GTCATC": 3467, + "GTCATG": 3468, + "GTCACA": 3469, + "GTCACT": 3470, + "GTCACC": 3471, + "GTCACG": 3472, + "GTCAGA": 3473, + "GTCAGT": 3474, + "GTCAGC": 3475, + "GTCAGG": 3476, + "GTCTAA": 3477, + "GTCTAT": 3478, + "GTCTAC": 3479, + "GTCTAG": 3480, + "GTCTTA": 3481, + "GTCTTT": 3482, + "GTCTTC": 3483, + "GTCTTG": 3484, + "GTCTCA": 3485, + "GTCTCT": 3486, + "GTCTCC": 3487, + "GTCTCG": 3488, + "GTCTGA": 3489, + "GTCTGT": 3490, + "GTCTGC": 3491, + "GTCTGG": 3492, + "GTCCAA": 3493, + "GTCCAT": 3494, + "GTCCAC": 3495, + "GTCCAG": 3496, + "GTCCTA": 3497, + "GTCCTT": 3498, + "GTCCTC": 3499, + "GTCCTG": 3500, + "GTCCCA": 3501, + "GTCCCT": 3502, + "GTCCCC": 3503, + "GTCCCG": 3504, + "GTCCGA": 3505, + "GTCCGT": 3506, + "GTCCGC": 3507, + "GTCCGG": 3508, + "GTCGAA": 3509, + "GTCGAT": 3510, + "GTCGAC": 3511, + "GTCGAG": 3512, + "GTCGTA": 3513, + "GTCGTT": 3514, + "GTCGTC": 3515, + "GTCGTG": 3516, + "GTCGCA": 3517, + "GTCGCT": 3518, + "GTCGCC": 3519, + "GTCGCG": 3520, + "GTCGGA": 3521, + "GTCGGT": 3522, + "GTCGGC": 3523, + "GTCGGG": 3524, + "GTGAAA": 3525, + "GTGAAT": 3526, + "GTGAAC": 3527, + "GTGAAG": 3528, + "GTGATA": 3529, + "GTGATT": 3530, + "GTGATC": 3531, + "GTGATG": 3532, + "GTGACA": 3533, + "GTGACT": 3534, + "GTGACC": 3535, + "GTGACG": 3536, + "GTGAGA": 3537, + "GTGAGT": 3538, + "GTGAGC": 3539, + "GTGAGG": 3540, + "GTGTAA": 3541, + "GTGTAT": 3542, + "GTGTAC": 3543, + "GTGTAG": 3544, + "GTGTTA": 3545, + "GTGTTT": 3546, + "GTGTTC": 3547, + "GTGTTG": 3548, + "GTGTCA": 3549, + "GTGTCT": 3550, + "GTGTCC": 3551, + "GTGTCG": 3552, + "GTGTGA": 3553, + "GTGTGT": 3554, + "GTGTGC": 3555, + "GTGTGG": 3556, + "GTGCAA": 3557, + "GTGCAT": 3558, + "GTGCAC": 3559, + "GTGCAG": 3560, + "GTGCTA": 3561, + "GTGCTT": 3562, + "GTGCTC": 3563, + "GTGCTG": 3564, + "GTGCCA": 3565, + "GTGCCT": 3566, + "GTGCCC": 3567, + "GTGCCG": 3568, + "GTGCGA": 3569, + "GTGCGT": 3570, + "GTGCGC": 3571, + "GTGCGG": 3572, + "GTGGAA": 3573, + "GTGGAT": 3574, + "GTGGAC": 3575, + "GTGGAG": 3576, + "GTGGTA": 3577, + "GTGGTT": 3578, + "GTGGTC": 3579, + "GTGGTG": 3580, + "GTGGCA": 3581, + "GTGGCT": 3582, + "GTGGCC": 3583, + "GTGGCG": 3584, + "GTGGGA": 3585, + "GTGGGT": 3586, + "GTGGGC": 3587, + "GTGGGG": 3588, + "GCAAAA": 3589, + "GCAAAT": 3590, + "GCAAAC": 3591, + "GCAAAG": 3592, + "GCAATA": 3593, + "GCAATT": 3594, + "GCAATC": 3595, + "GCAATG": 3596, + "GCAACA": 3597, + "GCAACT": 3598, + "GCAACC": 3599, + "GCAACG": 3600, + "GCAAGA": 3601, + "GCAAGT": 3602, + "GCAAGC": 3603, + "GCAAGG": 3604, + "GCATAA": 3605, + "GCATAT": 3606, + "GCATAC": 3607, + "GCATAG": 3608, + "GCATTA": 3609, + "GCATTT": 3610, + "GCATTC": 3611, + "GCATTG": 3612, + "GCATCA": 3613, + "GCATCT": 3614, + "GCATCC": 3615, + "GCATCG": 3616, + "GCATGA": 3617, + "GCATGT": 3618, + "GCATGC": 3619, + "GCATGG": 3620, + "GCACAA": 3621, + "GCACAT": 3622, + "GCACAC": 3623, + "GCACAG": 3624, + "GCACTA": 3625, + "GCACTT": 3626, + "GCACTC": 3627, + "GCACTG": 3628, + "GCACCA": 3629, + "GCACCT": 3630, + "GCACCC": 3631, + "GCACCG": 3632, + "GCACGA": 3633, + "GCACGT": 3634, + "GCACGC": 3635, + "GCACGG": 3636, + "GCAGAA": 3637, + "GCAGAT": 3638, + "GCAGAC": 3639, + "GCAGAG": 3640, + "GCAGTA": 3641, + "GCAGTT": 3642, + "GCAGTC": 3643, + "GCAGTG": 3644, + "GCAGCA": 3645, + "GCAGCT": 3646, + "GCAGCC": 3647, + "GCAGCG": 3648, + "GCAGGA": 3649, + "GCAGGT": 3650, + "GCAGGC": 3651, + "GCAGGG": 3652, + "GCTAAA": 3653, + "GCTAAT": 3654, + "GCTAAC": 3655, + "GCTAAG": 3656, + "GCTATA": 3657, + "GCTATT": 3658, + "GCTATC": 3659, + "GCTATG": 3660, + "GCTACA": 3661, + "GCTACT": 3662, + "GCTACC": 3663, + "GCTACG": 3664, + "GCTAGA": 3665, + "GCTAGT": 3666, + "GCTAGC": 3667, + "GCTAGG": 3668, + "GCTTAA": 3669, + "GCTTAT": 3670, + "GCTTAC": 3671, + "GCTTAG": 3672, + "GCTTTA": 3673, + "GCTTTT": 3674, + "GCTTTC": 3675, + "GCTTTG": 3676, + "GCTTCA": 3677, + "GCTTCT": 3678, + "GCTTCC": 3679, + "GCTTCG": 3680, + "GCTTGA": 3681, + "GCTTGT": 3682, + "GCTTGC": 3683, + "GCTTGG": 3684, + "GCTCAA": 3685, + "GCTCAT": 3686, + "GCTCAC": 3687, + "GCTCAG": 3688, + "GCTCTA": 3689, + "GCTCTT": 3690, + "GCTCTC": 3691, + "GCTCTG": 3692, + "GCTCCA": 3693, + "GCTCCT": 3694, + "GCTCCC": 3695, + "GCTCCG": 3696, + "GCTCGA": 3697, + "GCTCGT": 3698, + "GCTCGC": 3699, + "GCTCGG": 3700, + "GCTGAA": 3701, + "GCTGAT": 3702, + "GCTGAC": 3703, + "GCTGAG": 3704, + "GCTGTA": 3705, + "GCTGTT": 3706, + "GCTGTC": 3707, + "GCTGTG": 3708, + "GCTGCA": 3709, + "GCTGCT": 3710, + "GCTGCC": 3711, + "GCTGCG": 3712, + "GCTGGA": 3713, + "GCTGGT": 3714, + "GCTGGC": 3715, + "GCTGGG": 3716, + "GCCAAA": 3717, + "GCCAAT": 3718, + "GCCAAC": 3719, + "GCCAAG": 3720, + "GCCATA": 3721, + "GCCATT": 3722, + "GCCATC": 3723, + "GCCATG": 3724, + "GCCACA": 3725, + "GCCACT": 3726, + "GCCACC": 3727, + "GCCACG": 3728, + "GCCAGA": 3729, + "GCCAGT": 3730, + "GCCAGC": 3731, + "GCCAGG": 3732, + "GCCTAA": 3733, + "GCCTAT": 3734, + "GCCTAC": 3735, + "GCCTAG": 3736, + "GCCTTA": 3737, + "GCCTTT": 3738, + "GCCTTC": 3739, + "GCCTTG": 3740, + "GCCTCA": 3741, + "GCCTCT": 3742, + "GCCTCC": 3743, + "GCCTCG": 3744, + "GCCTGA": 3745, + "GCCTGT": 3746, + "GCCTGC": 3747, + "GCCTGG": 3748, + "GCCCAA": 3749, + "GCCCAT": 3750, + "GCCCAC": 3751, + "GCCCAG": 3752, + "GCCCTA": 3753, + "GCCCTT": 3754, + "GCCCTC": 3755, + "GCCCTG": 3756, + "GCCCCA": 3757, + "GCCCCT": 3758, + "GCCCCC": 3759, + "GCCCCG": 3760, + "GCCCGA": 3761, + "GCCCGT": 3762, + "GCCCGC": 3763, + "GCCCGG": 3764, + "GCCGAA": 3765, + "GCCGAT": 3766, + "GCCGAC": 3767, + "GCCGAG": 3768, + "GCCGTA": 3769, + "GCCGTT": 3770, + "GCCGTC": 3771, + "GCCGTG": 3772, + "GCCGCA": 3773, + "GCCGCT": 3774, + "GCCGCC": 3775, + "GCCGCG": 3776, + "GCCGGA": 3777, + "GCCGGT": 3778, + "GCCGGC": 3779, + "GCCGGG": 3780, + "GCGAAA": 3781, + "GCGAAT": 3782, + "GCGAAC": 3783, + "GCGAAG": 3784, + "GCGATA": 3785, + "GCGATT": 3786, + "GCGATC": 3787, + "GCGATG": 3788, + "GCGACA": 3789, + "GCGACT": 3790, + "GCGACC": 3791, + "GCGACG": 3792, + "GCGAGA": 3793, + "GCGAGT": 3794, + "GCGAGC": 3795, + "GCGAGG": 3796, + "GCGTAA": 3797, + "GCGTAT": 3798, + "GCGTAC": 3799, + "GCGTAG": 3800, + "GCGTTA": 3801, + "GCGTTT": 3802, + "GCGTTC": 3803, + "GCGTTG": 3804, + "GCGTCA": 3805, + "GCGTCT": 3806, + "GCGTCC": 3807, + "GCGTCG": 3808, + "GCGTGA": 3809, + "GCGTGT": 3810, + "GCGTGC": 3811, + "GCGTGG": 3812, + "GCGCAA": 3813, + "GCGCAT": 3814, + "GCGCAC": 3815, + "GCGCAG": 3816, + "GCGCTA": 3817, + "GCGCTT": 3818, + "GCGCTC": 3819, + "GCGCTG": 3820, + "GCGCCA": 3821, + "GCGCCT": 3822, + "GCGCCC": 3823, + "GCGCCG": 3824, + "GCGCGA": 3825, + "GCGCGT": 3826, + "GCGCGC": 3827, + "GCGCGG": 3828, + "GCGGAA": 3829, + "GCGGAT": 3830, + "GCGGAC": 3831, + "GCGGAG": 3832, + "GCGGTA": 3833, + "GCGGTT": 3834, + "GCGGTC": 3835, + "GCGGTG": 3836, + "GCGGCA": 3837, + "GCGGCT": 3838, + "GCGGCC": 3839, + "GCGGCG": 3840, + "GCGGGA": 3841, + "GCGGGT": 3842, + "GCGGGC": 3843, + "GCGGGG": 3844, + "GGAAAA": 3845, + "GGAAAT": 3846, + "GGAAAC": 3847, + "GGAAAG": 3848, + "GGAATA": 3849, + "GGAATT": 3850, + "GGAATC": 3851, + "GGAATG": 3852, + "GGAACA": 3853, + "GGAACT": 3854, + "GGAACC": 3855, + "GGAACG": 3856, + "GGAAGA": 3857, + "GGAAGT": 3858, + "GGAAGC": 3859, + "GGAAGG": 3860, + "GGATAA": 3861, + "GGATAT": 3862, + "GGATAC": 3863, + "GGATAG": 3864, + "GGATTA": 3865, + "GGATTT": 3866, + "GGATTC": 3867, + "GGATTG": 3868, + "GGATCA": 3869, + "GGATCT": 3870, + "GGATCC": 3871, + "GGATCG": 3872, + "GGATGA": 3873, + "GGATGT": 3874, + "GGATGC": 3875, + "GGATGG": 3876, + "GGACAA": 3877, + "GGACAT": 3878, + "GGACAC": 3879, + "GGACAG": 3880, + "GGACTA": 3881, + "GGACTT": 3882, + "GGACTC": 3883, + "GGACTG": 3884, + "GGACCA": 3885, + "GGACCT": 3886, + "GGACCC": 3887, + "GGACCG": 3888, + "GGACGA": 3889, + "GGACGT": 3890, + "GGACGC": 3891, + "GGACGG": 3892, + "GGAGAA": 3893, + "GGAGAT": 3894, + "GGAGAC": 3895, + "GGAGAG": 3896, + "GGAGTA": 3897, + "GGAGTT": 3898, + "GGAGTC": 3899, + "GGAGTG": 3900, + "GGAGCA": 3901, + "GGAGCT": 3902, + "GGAGCC": 3903, + "GGAGCG": 3904, + "GGAGGA": 3905, + "GGAGGT": 3906, + "GGAGGC": 3907, + "GGAGGG": 3908, + "GGTAAA": 3909, + "GGTAAT": 3910, + "GGTAAC": 3911, + "GGTAAG": 3912, + "GGTATA": 3913, + "GGTATT": 3914, + "GGTATC": 3915, + "GGTATG": 3916, + "GGTACA": 3917, + "GGTACT": 3918, + "GGTACC": 3919, + "GGTACG": 3920, + "GGTAGA": 3921, + "GGTAGT": 3922, + "GGTAGC": 3923, + "GGTAGG": 3924, + "GGTTAA": 3925, + "GGTTAT": 3926, + "GGTTAC": 3927, + "GGTTAG": 3928, + "GGTTTA": 3929, + "GGTTTT": 3930, + "GGTTTC": 3931, + "GGTTTG": 3932, + "GGTTCA": 3933, + "GGTTCT": 3934, + "GGTTCC": 3935, + "GGTTCG": 3936, + "GGTTGA": 3937, + "GGTTGT": 3938, + "GGTTGC": 3939, + "GGTTGG": 3940, + "GGTCAA": 3941, + "GGTCAT": 3942, + "GGTCAC": 3943, + "GGTCAG": 3944, + "GGTCTA": 3945, + "GGTCTT": 3946, + "GGTCTC": 3947, + "GGTCTG": 3948, + "GGTCCA": 3949, + "GGTCCT": 3950, + "GGTCCC": 3951, + "GGTCCG": 3952, + "GGTCGA": 3953, + "GGTCGT": 3954, + "GGTCGC": 3955, + "GGTCGG": 3956, + "GGTGAA": 3957, + "GGTGAT": 3958, + "GGTGAC": 3959, + "GGTGAG": 3960, + "GGTGTA": 3961, + "GGTGTT": 3962, + "GGTGTC": 3963, + "GGTGTG": 3964, + "GGTGCA": 3965, + "GGTGCT": 3966, + "GGTGCC": 3967, + "GGTGCG": 3968, + "GGTGGA": 3969, + "GGTGGT": 3970, + "GGTGGC": 3971, + "GGTGGG": 3972, + "GGCAAA": 3973, + "GGCAAT": 3974, + "GGCAAC": 3975, + "GGCAAG": 3976, + "GGCATA": 3977, + "GGCATT": 3978, + "GGCATC": 3979, + "GGCATG": 3980, + "GGCACA": 3981, + "GGCACT": 3982, + "GGCACC": 3983, + "GGCACG": 3984, + "GGCAGA": 3985, + "GGCAGT": 3986, + "GGCAGC": 3987, + "GGCAGG": 3988, + "GGCTAA": 3989, + "GGCTAT": 3990, + "GGCTAC": 3991, + "GGCTAG": 3992, + "GGCTTA": 3993, + "GGCTTT": 3994, + "GGCTTC": 3995, + "GGCTTG": 3996, + "GGCTCA": 3997, + "GGCTCT": 3998, + "GGCTCC": 3999, + "GGCTCG": 4000, + "GGCTGA": 4001, + "GGCTGT": 4002, + "GGCTGC": 4003, + "GGCTGG": 4004, + "GGCCAA": 4005, + "GGCCAT": 4006, + "GGCCAC": 4007, + "GGCCAG": 4008, + "GGCCTA": 4009, + "GGCCTT": 4010, + "GGCCTC": 4011, + "GGCCTG": 4012, + "GGCCCA": 4013, + "GGCCCT": 4014, + "GGCCCC": 4015, + "GGCCCG": 4016, + "GGCCGA": 4017, + "GGCCGT": 4018, + "GGCCGC": 4019, + "GGCCGG": 4020, + "GGCGAA": 4021, + "GGCGAT": 4022, + "GGCGAC": 4023, + "GGCGAG": 4024, + "GGCGTA": 4025, + "GGCGTT": 4026, + "GGCGTC": 4027, + "GGCGTG": 4028, + "GGCGCA": 4029, + "GGCGCT": 4030, + "GGCGCC": 4031, + "GGCGCG": 4032, + "GGCGGA": 4033, + "GGCGGT": 4034, + "GGCGGC": 4035, + "GGCGGG": 4036, + "GGGAAA": 4037, + "GGGAAT": 4038, + "GGGAAC": 4039, + "GGGAAG": 4040, + "GGGATA": 4041, + "GGGATT": 4042, + "GGGATC": 4043, + "GGGATG": 4044, + "GGGACA": 4045, + "GGGACT": 4046, + "GGGACC": 4047, + "GGGACG": 4048, + "GGGAGA": 4049, + "GGGAGT": 4050, + "GGGAGC": 4051, + "GGGAGG": 4052, + "GGGTAA": 4053, + "GGGTAT": 4054, + "GGGTAC": 4055, + "GGGTAG": 4056, + "GGGTTA": 4057, + "GGGTTT": 4058, + "GGGTTC": 4059, + "GGGTTG": 4060, + "GGGTCA": 4061, + "GGGTCT": 4062, + "GGGTCC": 4063, + "GGGTCG": 4064, + "GGGTGA": 4065, + "GGGTGT": 4066, + "GGGTGC": 4067, + "GGGTGG": 4068, + "GGGCAA": 4069, + "GGGCAT": 4070, + "GGGCAC": 4071, + "GGGCAG": 4072, + "GGGCTA": 4073, + "GGGCTT": 4074, + "GGGCTC": 4075, + "GGGCTG": 4076, + "GGGCCA": 4077, + "GGGCCT": 4078, + "GGGCCC": 4079, + "GGGCCG": 4080, + "GGGCGA": 4081, + "GGGCGT": 4082, + "GGGCGC": 4083, + "GGGCGG": 4084, + "GGGGAA": 4085, + "GGGGAT": 4086, + "GGGGAC": 4087, + "GGGGAG": 4088, + "GGGGTA": 4089, + "GGGGTT": 4090, + "GGGGTC": 4091, + "GGGGTG": 4092, + "GGGGCA": 4093, + "GGGGCT": 4094, + "GGGGCC": 4095, + "GGGGCG": 4096, + "GGGGGA": 4097, + "GGGGGT": 4098, + "GGGGGC": 4099, + "GGGGGG": 4100 + } + } +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/tokenizer_config.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..604e2b1b29af1dcb3920d2828b6a90276e5e3d02 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/tokenizer_config.json @@ -0,0 +1,58 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "[UNK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "[CLS]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "[SEP]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "4": { + "content": "[MASK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "clean_up_tokenization_spaces": true, + "cls_token": "[CLS]", + "do_basic_tokenize": true, + "do_lower_case": false, + "mask_token": "[MASK]", + "max_len": 512, + "model_max_length": 512, + "never_split": null, + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "strip_accents": null, + "tokenize_chinese_chars": true, + "tokenizer_class": "BertTokenizer", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/trainer_state.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..e456b2205877f1a066f6dd6366a08d67ebc8ee6f --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/trainer_state.json @@ -0,0 +1,1713 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 24.0, + "eval_steps": 500, + "global_step": 6000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.1, + "grad_norm": 0.2192622423171997, + "learning_rate": 5.319148936170213e-05, + "loss": 0.6949, + "step": 25 + }, + { + "epoch": 0.2, + "grad_norm": 0.22659960389137268, + "learning_rate": 0.00010638297872340425, + "loss": 0.6926, + "step": 50 + }, + { + "epoch": 0.3, + "grad_norm": 0.33277687430381775, + "learning_rate": 0.00015957446808510637, + "loss": 0.693, + "step": 75 + }, + { + "epoch": 0.4, + "grad_norm": 0.5183415412902832, + "learning_rate": 0.0002127659574468085, + "loss": 0.6928, + "step": 100 + }, + { + "epoch": 0.5, + "grad_norm": 0.21227706968784332, + "learning_rate": 0.00026595744680851064, + "loss": 0.6938, + "step": 125 + }, + { + "epoch": 0.6, + "grad_norm": 0.1846485733985901, + "learning_rate": 0.00031914893617021275, + "loss": 0.6942, + "step": 150 + }, + { + "epoch": 0.7, + "grad_norm": 0.27858808636665344, + "learning_rate": 0.0003723404255319149, + "loss": 0.6931, + "step": 175 + }, + { + "epoch": 0.8, + "grad_norm": 0.26012367010116577, + "learning_rate": 0.0003992081821181128, + "loss": 0.6936, + "step": 200 + }, + { + "epoch": 0.9, + "grad_norm": 0.2534612715244293, + "learning_rate": 0.0003975585615308479, + "loss": 0.6939, + "step": 225 + }, + { + "epoch": 1.0, + "grad_norm": 0.42967700958251953, + "learning_rate": 0.00039590894094358297, + "loss": 0.6928, + "step": 250 + }, + { + "epoch": 1.1, + "grad_norm": 0.2613862454891205, + "learning_rate": 0.0003942593203563181, + "loss": 0.6918, + "step": 275 + }, + { + "epoch": 1.2, + "grad_norm": 0.5514339208602905, + "learning_rate": 0.0003926096997690532, + "loss": 0.6918, + "step": 300 + }, + { + "epoch": 1.3, + "grad_norm": 0.33801883459091187, + "learning_rate": 0.00039096007918178817, + "loss": 0.6906, + "step": 325 + }, + { + "epoch": 1.4, + "grad_norm": 0.2953580617904663, + "learning_rate": 0.0003893104585945233, + "loss": 0.6886, + "step": 350 + }, + { + "epoch": 1.5, + "grad_norm": 0.5340980887413025, + "learning_rate": 0.0003876608380072583, + "loss": 0.6877, + "step": 375 + }, + { + "epoch": 1.6, + "grad_norm": 0.3042701482772827, + "learning_rate": 0.00038601121741999343, + "loss": 0.6853, + "step": 400 + }, + { + "epoch": 1.7, + "grad_norm": 0.44497063755989075, + "learning_rate": 0.0003843615968327285, + "loss": 0.6772, + "step": 425 + }, + { + "epoch": 1.8, + "grad_norm": 0.6260074973106384, + "learning_rate": 0.0003827119762454636, + "loss": 0.6671, + "step": 450 + }, + { + "epoch": 1.9, + "grad_norm": 0.9574269652366638, + "learning_rate": 0.00038106235565819863, + "loss": 0.6517, + "step": 475 + }, + { + "epoch": 2.0, + "grad_norm": 0.9055777192115784, + "learning_rate": 0.0003794127350709337, + "loss": 0.6438, + "step": 500 + }, + { + "epoch": 2.1, + "grad_norm": 0.7860134840011597, + "learning_rate": 0.0003777631144836688, + "loss": 0.6379, + "step": 525 + }, + { + "epoch": 2.2, + "grad_norm": 1.5324610471725464, + "learning_rate": 0.00037611349389640383, + "loss": 0.6143, + "step": 550 + }, + { + "epoch": 2.3, + "grad_norm": 1.8535794019699097, + "learning_rate": 0.00037446387330913894, + "loss": 0.607, + "step": 575 + }, + { + "epoch": 2.4, + "grad_norm": 1.8850122690200806, + "learning_rate": 0.000372814252721874, + "loss": 0.5642, + "step": 600 + }, + { + "epoch": 2.5, + "grad_norm": 1.288278579711914, + "learning_rate": 0.00037116463213460903, + "loss": 0.5739, + "step": 625 + }, + { + "epoch": 2.6, + "grad_norm": 1.344173550605774, + "learning_rate": 0.00036951501154734414, + "loss": 0.5718, + "step": 650 + }, + { + "epoch": 2.7, + "grad_norm": 1.430834174156189, + "learning_rate": 0.0003678653909600792, + "loss": 0.543, + "step": 675 + }, + { + "epoch": 2.8, + "grad_norm": 1.5948325395584106, + "learning_rate": 0.0003662157703728143, + "loss": 0.5434, + "step": 700 + }, + { + "epoch": 2.9, + "grad_norm": 2.688519239425659, + "learning_rate": 0.00036456614978554934, + "loss": 0.5842, + "step": 725 + }, + { + "epoch": 3.0, + "grad_norm": 1.6625819206237793, + "learning_rate": 0.00036291652919828444, + "loss": 0.5526, + "step": 750 + }, + { + "epoch": 3.1, + "grad_norm": 1.869725227355957, + "learning_rate": 0.00036126690861101944, + "loss": 0.53, + "step": 775 + }, + { + "epoch": 3.2, + "grad_norm": 2.1625709533691406, + "learning_rate": 0.00035961728802375454, + "loss": 0.5341, + "step": 800 + }, + { + "epoch": 3.3, + "grad_norm": 1.9650202989578247, + "learning_rate": 0.0003579676674364896, + "loss": 0.5297, + "step": 825 + }, + { + "epoch": 3.4, + "grad_norm": 1.3358676433563232, + "learning_rate": 0.0003563180468492247, + "loss": 0.5273, + "step": 850 + }, + { + "epoch": 3.5, + "grad_norm": 2.1927096843719482, + "learning_rate": 0.0003546684262619598, + "loss": 0.5084, + "step": 875 + }, + { + "epoch": 3.6, + "grad_norm": 2.2377188205718994, + "learning_rate": 0.00035301880567469485, + "loss": 0.4972, + "step": 900 + }, + { + "epoch": 3.7, + "grad_norm": 1.8601839542388916, + "learning_rate": 0.0003513691850874299, + "loss": 0.4852, + "step": 925 + }, + { + "epoch": 3.8, + "grad_norm": 1.4995975494384766, + "learning_rate": 0.00034971956450016495, + "loss": 0.5309, + "step": 950 + }, + { + "epoch": 3.9, + "grad_norm": 2.0536930561065674, + "learning_rate": 0.00034806994391290005, + "loss": 0.5249, + "step": 975 + }, + { + "epoch": 4.0, + "grad_norm": 2.084728479385376, + "learning_rate": 0.0003464203233256351, + "loss": 0.4771, + "step": 1000 + }, + { + "epoch": 4.1, + "grad_norm": 1.705510139465332, + "learning_rate": 0.0003447707027383702, + "loss": 0.5181, + "step": 1025 + }, + { + "epoch": 4.2, + "grad_norm": 3.83380389213562, + "learning_rate": 0.00034312108215110525, + "loss": 0.4873, + "step": 1050 + }, + { + "epoch": 4.3, + "grad_norm": 2.0163533687591553, + "learning_rate": 0.0003414714615638403, + "loss": 0.4801, + "step": 1075 + }, + { + "epoch": 4.4, + "grad_norm": 1.9917407035827637, + "learning_rate": 0.0003398218409765754, + "loss": 0.4766, + "step": 1100 + }, + { + "epoch": 4.5, + "grad_norm": 2.111729860305786, + "learning_rate": 0.00033817222038931045, + "loss": 0.4877, + "step": 1125 + }, + { + "epoch": 4.6, + "grad_norm": 1.5487738847732544, + "learning_rate": 0.00033652259980204556, + "loss": 0.507, + "step": 1150 + }, + { + "epoch": 4.7, + "grad_norm": 1.8715145587921143, + "learning_rate": 0.0003348729792147806, + "loss": 0.452, + "step": 1175 + }, + { + "epoch": 4.8, + "grad_norm": 2.2618906497955322, + "learning_rate": 0.0003332233586275157, + "loss": 0.4799, + "step": 1200 + }, + { + "epoch": 4.9, + "grad_norm": 1.657012939453125, + "learning_rate": 0.00033157373804025076, + "loss": 0.4715, + "step": 1225 + }, + { + "epoch": 5.0, + "grad_norm": 2.8894591331481934, + "learning_rate": 0.0003299241174529858, + "loss": 0.4437, + "step": 1250 + }, + { + "epoch": 5.1, + "grad_norm": 1.9179915189743042, + "learning_rate": 0.0003282744968657209, + "loss": 0.4796, + "step": 1275 + }, + { + "epoch": 5.2, + "grad_norm": 2.1816720962524414, + "learning_rate": 0.00032662487627845596, + "loss": 0.4463, + "step": 1300 + }, + { + "epoch": 5.3, + "grad_norm": 2.2761082649230957, + "learning_rate": 0.00032497525569119106, + "loss": 0.4215, + "step": 1325 + }, + { + "epoch": 5.4, + "grad_norm": 1.7190684080123901, + "learning_rate": 0.0003233256351039261, + "loss": 0.4481, + "step": 1350 + }, + { + "epoch": 5.5, + "grad_norm": 1.4793736934661865, + "learning_rate": 0.00032167601451666116, + "loss": 0.44, + "step": 1375 + }, + { + "epoch": 5.6, + "grad_norm": 2.5064358711242676, + "learning_rate": 0.00032002639392939627, + "loss": 0.4583, + "step": 1400 + }, + { + "epoch": 5.7, + "grad_norm": 1.8364307880401611, + "learning_rate": 0.0003183767733421313, + "loss": 0.4516, + "step": 1425 + }, + { + "epoch": 5.8, + "grad_norm": 1.3747583627700806, + "learning_rate": 0.0003167271527548664, + "loss": 0.4535, + "step": 1450 + }, + { + "epoch": 5.9, + "grad_norm": 2.580183982849121, + "learning_rate": 0.00031507753216760147, + "loss": 0.4325, + "step": 1475 + }, + { + "epoch": 6.0, + "grad_norm": 1.562374472618103, + "learning_rate": 0.00031342791158033657, + "loss": 0.4616, + "step": 1500 + }, + { + "epoch": 6.1, + "grad_norm": 1.4970141649246216, + "learning_rate": 0.00031177829099307157, + "loss": 0.4607, + "step": 1525 + }, + { + "epoch": 6.2, + "grad_norm": 2.436326503753662, + "learning_rate": 0.00031012867040580667, + "loss": 0.4498, + "step": 1550 + }, + { + "epoch": 6.3, + "grad_norm": 1.9849016666412354, + "learning_rate": 0.0003084790498185417, + "loss": 0.4262, + "step": 1575 + }, + { + "epoch": 6.4, + "grad_norm": 1.6405338048934937, + "learning_rate": 0.0003068294292312768, + "loss": 0.4225, + "step": 1600 + }, + { + "epoch": 6.5, + "grad_norm": 1.8233951330184937, + "learning_rate": 0.0003051798086440119, + "loss": 0.4248, + "step": 1625 + }, + { + "epoch": 6.6, + "grad_norm": 1.396469235420227, + "learning_rate": 0.000303530188056747, + "loss": 0.412, + "step": 1650 + }, + { + "epoch": 6.7, + "grad_norm": 1.5438531637191772, + "learning_rate": 0.000301880567469482, + "loss": 0.4105, + "step": 1675 + }, + { + "epoch": 6.8, + "grad_norm": 2.49668288230896, + "learning_rate": 0.0003002309468822171, + "loss": 0.4318, + "step": 1700 + }, + { + "epoch": 6.9, + "grad_norm": 1.5911107063293457, + "learning_rate": 0.0002985813262949522, + "loss": 0.4454, + "step": 1725 + }, + { + "epoch": 7.0, + "grad_norm": 2.9934186935424805, + "learning_rate": 0.00029693170570768723, + "loss": 0.4306, + "step": 1750 + }, + { + "epoch": 7.1, + "grad_norm": 2.5476737022399902, + "learning_rate": 0.00029528208512042233, + "loss": 0.4066, + "step": 1775 + }, + { + "epoch": 7.2, + "grad_norm": 4.207254886627197, + "learning_rate": 0.0002936324645331574, + "loss": 0.4521, + "step": 1800 + }, + { + "epoch": 7.3, + "grad_norm": 1.8378102779388428, + "learning_rate": 0.00029198284394589243, + "loss": 0.4197, + "step": 1825 + }, + { + "epoch": 7.4, + "grad_norm": 2.2129015922546387, + "learning_rate": 0.00029033322335862753, + "loss": 0.4182, + "step": 1850 + }, + { + "epoch": 7.5, + "grad_norm": 1.4666684865951538, + "learning_rate": 0.0002886836027713626, + "loss": 0.4224, + "step": 1875 + }, + { + "epoch": 7.6, + "grad_norm": 2.0311567783355713, + "learning_rate": 0.0002870339821840977, + "loss": 0.4304, + "step": 1900 + }, + { + "epoch": 7.7, + "grad_norm": 2.346604585647583, + "learning_rate": 0.00028538436159683273, + "loss": 0.4012, + "step": 1925 + }, + { + "epoch": 7.8, + "grad_norm": 1.440136432647705, + "learning_rate": 0.00028373474100956784, + "loss": 0.4487, + "step": 1950 + }, + { + "epoch": 7.9, + "grad_norm": 2.426828384399414, + "learning_rate": 0.0002820851204223029, + "loss": 0.3577, + "step": 1975 + }, + { + "epoch": 8.0, + "grad_norm": 3.095980405807495, + "learning_rate": 0.00028043549983503794, + "loss": 0.4121, + "step": 2000 + }, + { + "epoch": 8.1, + "grad_norm": 1.8646572828292847, + "learning_rate": 0.00027878587924777304, + "loss": 0.4213, + "step": 2025 + }, + { + "epoch": 8.2, + "grad_norm": 2.3382062911987305, + "learning_rate": 0.0002771362586605081, + "loss": 0.4119, + "step": 2050 + }, + { + "epoch": 8.3, + "grad_norm": 1.7999392747879028, + "learning_rate": 0.0002754866380732432, + "loss": 0.4165, + "step": 2075 + }, + { + "epoch": 8.4, + "grad_norm": 2.262005567550659, + "learning_rate": 0.00027383701748597824, + "loss": 0.4043, + "step": 2100 + }, + { + "epoch": 8.5, + "grad_norm": 1.8611185550689697, + "learning_rate": 0.0002721873968987133, + "loss": 0.4022, + "step": 2125 + }, + { + "epoch": 8.6, + "grad_norm": 1.3322739601135254, + "learning_rate": 0.00027053777631144834, + "loss": 0.3835, + "step": 2150 + }, + { + "epoch": 8.7, + "grad_norm": 1.335750937461853, + "learning_rate": 0.00026888815572418344, + "loss": 0.424, + "step": 2175 + }, + { + "epoch": 8.8, + "grad_norm": 1.4633246660232544, + "learning_rate": 0.00026723853513691855, + "loss": 0.4124, + "step": 2200 + }, + { + "epoch": 8.9, + "grad_norm": 2.061042308807373, + "learning_rate": 0.0002655889145496536, + "loss": 0.4016, + "step": 2225 + }, + { + "epoch": 9.0, + "grad_norm": 1.9621604681015015, + "learning_rate": 0.0002639392939623887, + "loss": 0.4256, + "step": 2250 + }, + { + "epoch": 9.1, + "grad_norm": 1.4014378786087036, + "learning_rate": 0.0002622896733751237, + "loss": 0.3806, + "step": 2275 + }, + { + "epoch": 9.2, + "grad_norm": 2.6264309883117676, + "learning_rate": 0.0002606400527878588, + "loss": 0.3636, + "step": 2300 + }, + { + "epoch": 9.3, + "grad_norm": 1.8708316087722778, + "learning_rate": 0.00025899043220059385, + "loss": 0.4101, + "step": 2325 + }, + { + "epoch": 9.4, + "grad_norm": 2.156869649887085, + "learning_rate": 0.00025734081161332895, + "loss": 0.4114, + "step": 2350 + }, + { + "epoch": 9.5, + "grad_norm": 2.483182907104492, + "learning_rate": 0.000255691191026064, + "loss": 0.421, + "step": 2375 + }, + { + "epoch": 9.6, + "grad_norm": 2.1955626010894775, + "learning_rate": 0.0002540415704387991, + "loss": 0.3918, + "step": 2400 + }, + { + "epoch": 9.7, + "grad_norm": 4.115762233734131, + "learning_rate": 0.00025239194985153415, + "loss": 0.4172, + "step": 2425 + }, + { + "epoch": 9.8, + "grad_norm": 2.144148349761963, + "learning_rate": 0.0002507423292642692, + "loss": 0.3636, + "step": 2450 + }, + { + "epoch": 9.9, + "grad_norm": 2.340376377105713, + "learning_rate": 0.0002490927086770043, + "loss": 0.4059, + "step": 2475 + }, + { + "epoch": 10.0, + "grad_norm": 1.5438555479049683, + "learning_rate": 0.00024744308808973936, + "loss": 0.4232, + "step": 2500 + }, + { + "epoch": 10.1, + "grad_norm": 2.485889196395874, + "learning_rate": 0.00024579346750247446, + "loss": 0.3861, + "step": 2525 + }, + { + "epoch": 10.2, + "grad_norm": 2.4561119079589844, + "learning_rate": 0.0002441438469152095, + "loss": 0.3565, + "step": 2550 + }, + { + "epoch": 10.3, + "grad_norm": 2.460170269012451, + "learning_rate": 0.00024249422632794456, + "loss": 0.4003, + "step": 2575 + }, + { + "epoch": 10.4, + "grad_norm": 2.774445056915283, + "learning_rate": 0.00024084460574067963, + "loss": 0.4088, + "step": 2600 + }, + { + "epoch": 10.5, + "grad_norm": 2.678745985031128, + "learning_rate": 0.0002391949851534147, + "loss": 0.4021, + "step": 2625 + }, + { + "epoch": 10.6, + "grad_norm": 1.4933565855026245, + "learning_rate": 0.0002375453645661498, + "loss": 0.3744, + "step": 2650 + }, + { + "epoch": 10.7, + "grad_norm": 1.367900013923645, + "learning_rate": 0.00023589574397888486, + "loss": 0.3925, + "step": 2675 + }, + { + "epoch": 10.8, + "grad_norm": 1.653441309928894, + "learning_rate": 0.00023424612339161997, + "loss": 0.4049, + "step": 2700 + }, + { + "epoch": 10.9, + "grad_norm": 1.5892399549484253, + "learning_rate": 0.000232596502804355, + "loss": 0.3927, + "step": 2725 + }, + { + "epoch": 11.0, + "grad_norm": 1.535090684890747, + "learning_rate": 0.00023094688221709007, + "loss": 0.3739, + "step": 2750 + }, + { + "epoch": 11.1, + "grad_norm": 2.247271776199341, + "learning_rate": 0.00022929726162982514, + "loss": 0.3362, + "step": 2775 + }, + { + "epoch": 11.2, + "grad_norm": 2.408069372177124, + "learning_rate": 0.00022764764104256022, + "loss": 0.3933, + "step": 2800 + }, + { + "epoch": 11.3, + "grad_norm": 2.219688653945923, + "learning_rate": 0.0002259980204552953, + "loss": 0.4062, + "step": 2825 + }, + { + "epoch": 11.4, + "grad_norm": 3.0188546180725098, + "learning_rate": 0.00022434839986803037, + "loss": 0.3926, + "step": 2850 + }, + { + "epoch": 11.5, + "grad_norm": 2.8080244064331055, + "learning_rate": 0.00022269877928076542, + "loss": 0.3942, + "step": 2875 + }, + { + "epoch": 11.6, + "grad_norm": 2.263610601425171, + "learning_rate": 0.0002210491586935005, + "loss": 0.3887, + "step": 2900 + }, + { + "epoch": 11.7, + "grad_norm": 2.1239607334136963, + "learning_rate": 0.00021939953810623557, + "loss": 0.3989, + "step": 2925 + }, + { + "epoch": 11.8, + "grad_norm": 3.435351848602295, + "learning_rate": 0.00021774991751897065, + "loss": 0.4185, + "step": 2950 + }, + { + "epoch": 11.9, + "grad_norm": 4.1570515632629395, + "learning_rate": 0.00021610029693170573, + "loss": 0.3597, + "step": 2975 + }, + { + "epoch": 12.0, + "grad_norm": 1.7937941551208496, + "learning_rate": 0.0002144506763444408, + "loss": 0.3719, + "step": 3000 + }, + { + "epoch": 12.1, + "grad_norm": 2.3080639839172363, + "learning_rate": 0.00021280105575717582, + "loss": 0.3856, + "step": 3025 + }, + { + "epoch": 12.2, + "grad_norm": 1.9331200122833252, + "learning_rate": 0.00021115143516991093, + "loss": 0.3798, + "step": 3050 + }, + { + "epoch": 12.3, + "grad_norm": 2.0342001914978027, + "learning_rate": 0.000209501814582646, + "loss": 0.3789, + "step": 3075 + }, + { + "epoch": 12.4, + "grad_norm": 1.7878581285476685, + "learning_rate": 0.00020785219399538108, + "loss": 0.3546, + "step": 3100 + }, + { + "epoch": 12.5, + "grad_norm": 3.657317638397217, + "learning_rate": 0.00020620257340811616, + "loss": 0.3582, + "step": 3125 + }, + { + "epoch": 12.6, + "grad_norm": 2.246429681777954, + "learning_rate": 0.00020455295282085123, + "loss": 0.3991, + "step": 3150 + }, + { + "epoch": 12.7, + "grad_norm": 1.7742749452590942, + "learning_rate": 0.00020290333223358626, + "loss": 0.3914, + "step": 3175 + }, + { + "epoch": 12.8, + "grad_norm": 2.3869900703430176, + "learning_rate": 0.00020125371164632133, + "loss": 0.3435, + "step": 3200 + }, + { + "epoch": 12.9, + "grad_norm": 2.444920539855957, + "learning_rate": 0.0001996040910590564, + "loss": 0.393, + "step": 3225 + }, + { + "epoch": 13.0, + "grad_norm": 2.26476788520813, + "learning_rate": 0.00019795447047179148, + "loss": 0.3828, + "step": 3250 + }, + { + "epoch": 13.1, + "grad_norm": 2.4492876529693604, + "learning_rate": 0.0001963048498845266, + "loss": 0.3856, + "step": 3275 + }, + { + "epoch": 13.2, + "grad_norm": 4.578381061553955, + "learning_rate": 0.00019465522929726164, + "loss": 0.3676, + "step": 3300 + }, + { + "epoch": 13.3, + "grad_norm": 2.0352611541748047, + "learning_rate": 0.00019300560870999671, + "loss": 0.369, + "step": 3325 + }, + { + "epoch": 13.4, + "grad_norm": 2.151038646697998, + "learning_rate": 0.0001913559881227318, + "loss": 0.3713, + "step": 3350 + }, + { + "epoch": 13.5, + "grad_norm": 2.4661877155303955, + "learning_rate": 0.00018970636753546684, + "loss": 0.3859, + "step": 3375 + }, + { + "epoch": 13.6, + "grad_norm": 1.7616342306137085, + "learning_rate": 0.00018805674694820192, + "loss": 0.363, + "step": 3400 + }, + { + "epoch": 13.7, + "grad_norm": 2.26426362991333, + "learning_rate": 0.000186407126360937, + "loss": 0.3696, + "step": 3425 + }, + { + "epoch": 13.8, + "grad_norm": 2.355781078338623, + "learning_rate": 0.00018475750577367207, + "loss": 0.372, + "step": 3450 + }, + { + "epoch": 13.9, + "grad_norm": 1.961585283279419, + "learning_rate": 0.00018310788518640715, + "loss": 0.3418, + "step": 3475 + }, + { + "epoch": 14.0, + "grad_norm": 2.54199481010437, + "learning_rate": 0.00018145826459914222, + "loss": 0.3648, + "step": 3500 + }, + { + "epoch": 14.1, + "grad_norm": 1.5017932653427124, + "learning_rate": 0.00017980864401187727, + "loss": 0.3914, + "step": 3525 + }, + { + "epoch": 14.2, + "grad_norm": 3.051287889480591, + "learning_rate": 0.00017815902342461235, + "loss": 0.3887, + "step": 3550 + }, + { + "epoch": 14.3, + "grad_norm": 2.697949171066284, + "learning_rate": 0.00017650940283734742, + "loss": 0.3524, + "step": 3575 + }, + { + "epoch": 14.4, + "grad_norm": 2.470728635787964, + "learning_rate": 0.00017485978225008247, + "loss": 0.3767, + "step": 3600 + }, + { + "epoch": 14.5, + "grad_norm": 2.5166375637054443, + "learning_rate": 0.00017321016166281755, + "loss": 0.362, + "step": 3625 + }, + { + "epoch": 14.6, + "grad_norm": 1.7630703449249268, + "learning_rate": 0.00017156054107555263, + "loss": 0.3936, + "step": 3650 + }, + { + "epoch": 14.7, + "grad_norm": 2.083225727081299, + "learning_rate": 0.0001699109204882877, + "loss": 0.36, + "step": 3675 + }, + { + "epoch": 14.8, + "grad_norm": 1.7796965837478638, + "learning_rate": 0.00016826129990102278, + "loss": 0.3646, + "step": 3700 + }, + { + "epoch": 14.9, + "grad_norm": 2.127898693084717, + "learning_rate": 0.00016661167931375785, + "loss": 0.3722, + "step": 3725 + }, + { + "epoch": 15.0, + "grad_norm": 2.666217088699341, + "learning_rate": 0.0001649620587264929, + "loss": 0.3329, + "step": 3750 + }, + { + "epoch": 15.1, + "grad_norm": 2.521188497543335, + "learning_rate": 0.00016331243813922798, + "loss": 0.331, + "step": 3775 + }, + { + "epoch": 15.2, + "grad_norm": 1.528865098953247, + "learning_rate": 0.00016166281755196306, + "loss": 0.3513, + "step": 3800 + }, + { + "epoch": 15.3, + "grad_norm": 3.452749490737915, + "learning_rate": 0.00016001319696469813, + "loss": 0.3777, + "step": 3825 + }, + { + "epoch": 15.4, + "grad_norm": 3.0441641807556152, + "learning_rate": 0.0001583635763774332, + "loss": 0.3536, + "step": 3850 + }, + { + "epoch": 15.5, + "grad_norm": 1.9194542169570923, + "learning_rate": 0.00015671395579016829, + "loss": 0.3635, + "step": 3875 + }, + { + "epoch": 15.6, + "grad_norm": 2.011859655380249, + "learning_rate": 0.00015506433520290334, + "loss": 0.3467, + "step": 3900 + }, + { + "epoch": 15.7, + "grad_norm": 1.5169134140014648, + "learning_rate": 0.0001534147146156384, + "loss": 0.3639, + "step": 3925 + }, + { + "epoch": 15.8, + "grad_norm": 2.715104103088379, + "learning_rate": 0.0001517650940283735, + "loss": 0.3473, + "step": 3950 + }, + { + "epoch": 15.9, + "grad_norm": 1.8220083713531494, + "learning_rate": 0.00015011547344110854, + "loss": 0.3902, + "step": 3975 + }, + { + "epoch": 16.0, + "grad_norm": 3.227846384048462, + "learning_rate": 0.00014846585285384361, + "loss": 0.3576, + "step": 4000 + }, + { + "epoch": 16.1, + "grad_norm": 1.9271538257598877, + "learning_rate": 0.0001468162322665787, + "loss": 0.3426, + "step": 4025 + }, + { + "epoch": 16.2, + "grad_norm": 2.58815336227417, + "learning_rate": 0.00014516661167931377, + "loss": 0.3714, + "step": 4050 + }, + { + "epoch": 16.3, + "grad_norm": 1.7768404483795166, + "learning_rate": 0.00014351699109204884, + "loss": 0.3561, + "step": 4075 + }, + { + "epoch": 16.4, + "grad_norm": 1.622894525527954, + "learning_rate": 0.00014186737050478392, + "loss": 0.3384, + "step": 4100 + }, + { + "epoch": 16.5, + "grad_norm": 1.8811430931091309, + "learning_rate": 0.00014021774991751897, + "loss": 0.3603, + "step": 4125 + }, + { + "epoch": 16.6, + "grad_norm": 2.1986522674560547, + "learning_rate": 0.00013856812933025404, + "loss": 0.34, + "step": 4150 + }, + { + "epoch": 16.7, + "grad_norm": 2.3372039794921875, + "learning_rate": 0.00013691850874298912, + "loss": 0.368, + "step": 4175 + }, + { + "epoch": 16.8, + "grad_norm": 1.5386862754821777, + "learning_rate": 0.00013526888815572417, + "loss": 0.3441, + "step": 4200 + }, + { + "epoch": 16.9, + "grad_norm": 1.9392023086547852, + "learning_rate": 0.00013361926756845927, + "loss": 0.3704, + "step": 4225 + }, + { + "epoch": 17.0, + "grad_norm": 3.1489596366882324, + "learning_rate": 0.00013196964698119435, + "loss": 0.3466, + "step": 4250 + }, + { + "epoch": 17.1, + "grad_norm": 3.081573724746704, + "learning_rate": 0.0001303200263939294, + "loss": 0.3352, + "step": 4275 + }, + { + "epoch": 17.2, + "grad_norm": 1.7880629301071167, + "learning_rate": 0.00012867040580666448, + "loss": 0.3569, + "step": 4300 + }, + { + "epoch": 17.3, + "grad_norm": 2.12654972076416, + "learning_rate": 0.00012702078521939955, + "loss": 0.3493, + "step": 4325 + }, + { + "epoch": 17.4, + "grad_norm": 2.646132230758667, + "learning_rate": 0.0001253711646321346, + "loss": 0.346, + "step": 4350 + }, + { + "epoch": 17.5, + "grad_norm": 1.6917959451675415, + "learning_rate": 0.00012372154404486968, + "loss": 0.3544, + "step": 4375 + }, + { + "epoch": 17.6, + "grad_norm": 1.7370896339416504, + "learning_rate": 0.00012207192345760475, + "loss": 0.3392, + "step": 4400 + }, + { + "epoch": 17.7, + "grad_norm": 2.7857675552368164, + "learning_rate": 0.00012042230287033982, + "loss": 0.3479, + "step": 4425 + }, + { + "epoch": 17.8, + "grad_norm": 2.1937146186828613, + "learning_rate": 0.0001187726822830749, + "loss": 0.4039, + "step": 4450 + }, + { + "epoch": 17.9, + "grad_norm": 2.3681347370147705, + "learning_rate": 0.00011712306169580998, + "loss": 0.3534, + "step": 4475 + }, + { + "epoch": 18.0, + "grad_norm": 1.5830249786376953, + "learning_rate": 0.00011547344110854503, + "loss": 0.3599, + "step": 4500 + }, + { + "epoch": 18.1, + "grad_norm": 2.077099084854126, + "learning_rate": 0.00011382382052128011, + "loss": 0.3369, + "step": 4525 + }, + { + "epoch": 18.2, + "grad_norm": 2.1928341388702393, + "learning_rate": 0.00011217419993401519, + "loss": 0.3525, + "step": 4550 + }, + { + "epoch": 18.3, + "grad_norm": 1.9083303213119507, + "learning_rate": 0.00011052457934675025, + "loss": 0.3781, + "step": 4575 + }, + { + "epoch": 18.4, + "grad_norm": 1.9660146236419678, + "learning_rate": 0.00010887495875948532, + "loss": 0.3589, + "step": 4600 + }, + { + "epoch": 18.5, + "grad_norm": 1.6620488166809082, + "learning_rate": 0.0001072253381722204, + "loss": 0.3571, + "step": 4625 + }, + { + "epoch": 18.6, + "grad_norm": 2.4148569107055664, + "learning_rate": 0.00010557571758495546, + "loss": 0.3718, + "step": 4650 + }, + { + "epoch": 18.7, + "grad_norm": 2.578110933303833, + "learning_rate": 0.00010392609699769054, + "loss": 0.3491, + "step": 4675 + }, + { + "epoch": 18.8, + "grad_norm": 2.4224448204040527, + "learning_rate": 0.00010227647641042562, + "loss": 0.3241, + "step": 4700 + }, + { + "epoch": 18.9, + "grad_norm": 1.9691663980484009, + "learning_rate": 0.00010062685582316067, + "loss": 0.3407, + "step": 4725 + }, + { + "epoch": 19.0, + "grad_norm": 1.8373006582260132, + "learning_rate": 9.897723523589574e-05, + "loss": 0.3415, + "step": 4750 + }, + { + "epoch": 19.1, + "grad_norm": 2.405202865600586, + "learning_rate": 9.732761464863082e-05, + "loss": 0.3259, + "step": 4775 + }, + { + "epoch": 19.2, + "grad_norm": 2.754739761352539, + "learning_rate": 9.56779940613659e-05, + "loss": 0.3361, + "step": 4800 + }, + { + "epoch": 19.3, + "grad_norm": 2.4953665733337402, + "learning_rate": 9.402837347410096e-05, + "loss": 0.3534, + "step": 4825 + }, + { + "epoch": 19.4, + "grad_norm": 2.974691390991211, + "learning_rate": 9.237875288683603e-05, + "loss": 0.3742, + "step": 4850 + }, + { + "epoch": 19.5, + "grad_norm": 1.4064937829971313, + "learning_rate": 9.072913229957111e-05, + "loss": 0.3271, + "step": 4875 + }, + { + "epoch": 19.6, + "grad_norm": 3.1546576023101807, + "learning_rate": 8.907951171230617e-05, + "loss": 0.3549, + "step": 4900 + }, + { + "epoch": 19.7, + "grad_norm": 1.948807716369629, + "learning_rate": 8.742989112504124e-05, + "loss": 0.3107, + "step": 4925 + }, + { + "epoch": 19.8, + "grad_norm": 1.1512947082519531, + "learning_rate": 8.578027053777631e-05, + "loss": 0.3448, + "step": 4950 + }, + { + "epoch": 19.9, + "grad_norm": 3.7861363887786865, + "learning_rate": 8.413064995051139e-05, + "loss": 0.3312, + "step": 4975 + }, + { + "epoch": 20.0, + "grad_norm": 1.6949777603149414, + "learning_rate": 8.248102936324645e-05, + "loss": 0.3532, + "step": 5000 + }, + { + "epoch": 20.1, + "grad_norm": 2.2876861095428467, + "learning_rate": 8.083140877598153e-05, + "loss": 0.3628, + "step": 5025 + }, + { + "epoch": 20.2, + "grad_norm": 1.8852239847183228, + "learning_rate": 7.91817881887166e-05, + "loss": 0.3595, + "step": 5050 + }, + { + "epoch": 20.3, + "grad_norm": 2.6525633335113525, + "learning_rate": 7.753216760145167e-05, + "loss": 0.345, + "step": 5075 + }, + { + "epoch": 20.4, + "grad_norm": 3.196200370788574, + "learning_rate": 7.588254701418674e-05, + "loss": 0.3442, + "step": 5100 + }, + { + "epoch": 20.5, + "grad_norm": 4.056946277618408, + "learning_rate": 7.423292642692181e-05, + "loss": 0.3232, + "step": 5125 + }, + { + "epoch": 20.6, + "grad_norm": 2.9283547401428223, + "learning_rate": 7.258330583965688e-05, + "loss": 0.3606, + "step": 5150 + }, + { + "epoch": 20.7, + "grad_norm": 2.2991044521331787, + "learning_rate": 7.093368525239196e-05, + "loss": 0.3595, + "step": 5175 + }, + { + "epoch": 20.8, + "grad_norm": 2.3510475158691406, + "learning_rate": 6.928406466512702e-05, + "loss": 0.3406, + "step": 5200 + }, + { + "epoch": 20.9, + "grad_norm": 1.8012481927871704, + "learning_rate": 6.763444407786209e-05, + "loss": 0.3304, + "step": 5225 + }, + { + "epoch": 21.0, + "grad_norm": 2.620314121246338, + "learning_rate": 6.598482349059718e-05, + "loss": 0.3572, + "step": 5250 + }, + { + "epoch": 21.1, + "grad_norm": 2.1940581798553467, + "learning_rate": 6.433520290333224e-05, + "loss": 0.332, + "step": 5275 + }, + { + "epoch": 21.2, + "grad_norm": 1.6455153226852417, + "learning_rate": 6.26855823160673e-05, + "loss": 0.3329, + "step": 5300 + }, + { + "epoch": 21.3, + "grad_norm": 2.369469165802002, + "learning_rate": 6.103596172880238e-05, + "loss": 0.3513, + "step": 5325 + }, + { + "epoch": 21.4, + "grad_norm": 2.4541404247283936, + "learning_rate": 5.938634114153745e-05, + "loss": 0.321, + "step": 5350 + }, + { + "epoch": 21.5, + "grad_norm": 2.259453296661377, + "learning_rate": 5.7736720554272516e-05, + "loss": 0.3589, + "step": 5375 + }, + { + "epoch": 21.6, + "grad_norm": 2.1575701236724854, + "learning_rate": 5.608709996700759e-05, + "loss": 0.3474, + "step": 5400 + }, + { + "epoch": 21.7, + "grad_norm": 1.968269944190979, + "learning_rate": 5.443747937974266e-05, + "loss": 0.3738, + "step": 5425 + }, + { + "epoch": 21.8, + "grad_norm": 1.7775858640670776, + "learning_rate": 5.278785879247773e-05, + "loss": 0.3471, + "step": 5450 + }, + { + "epoch": 21.9, + "grad_norm": 2.118597984313965, + "learning_rate": 5.113823820521281e-05, + "loss": 0.338, + "step": 5475 + }, + { + "epoch": 22.0, + "grad_norm": 2.509803056716919, + "learning_rate": 4.948861761794787e-05, + "loss": 0.3561, + "step": 5500 + }, + { + "epoch": 22.1, + "grad_norm": 1.1670511960983276, + "learning_rate": 4.783899703068295e-05, + "loss": 0.3304, + "step": 5525 + }, + { + "epoch": 22.2, + "grad_norm": 2.318145751953125, + "learning_rate": 4.618937644341802e-05, + "loss": 0.3776, + "step": 5550 + }, + { + "epoch": 22.3, + "grad_norm": 1.859100103378296, + "learning_rate": 4.453975585615309e-05, + "loss": 0.3171, + "step": 5575 + }, + { + "epoch": 22.4, + "grad_norm": 3.5954763889312744, + "learning_rate": 4.2890135268888156e-05, + "loss": 0.3686, + "step": 5600 + }, + { + "epoch": 22.5, + "grad_norm": 2.928518533706665, + "learning_rate": 4.1240514681623226e-05, + "loss": 0.4016, + "step": 5625 + }, + { + "epoch": 22.6, + "grad_norm": 2.325444221496582, + "learning_rate": 3.95908940943583e-05, + "loss": 0.3316, + "step": 5650 + }, + { + "epoch": 22.7, + "grad_norm": 2.2691650390625, + "learning_rate": 3.794127350709337e-05, + "loss": 0.3249, + "step": 5675 + }, + { + "epoch": 22.8, + "grad_norm": 2.954089879989624, + "learning_rate": 3.629165291982844e-05, + "loss": 0.3015, + "step": 5700 + }, + { + "epoch": 22.9, + "grad_norm": 1.7389692068099976, + "learning_rate": 3.464203233256351e-05, + "loss": 0.342, + "step": 5725 + }, + { + "epoch": 23.0, + "grad_norm": 2.889202356338501, + "learning_rate": 3.299241174529859e-05, + "loss": 0.3314, + "step": 5750 + }, + { + "epoch": 23.1, + "grad_norm": 2.633880138397217, + "learning_rate": 3.134279115803365e-05, + "loss": 0.307, + "step": 5775 + }, + { + "epoch": 23.2, + "grad_norm": 3.491842746734619, + "learning_rate": 2.9693170570768723e-05, + "loss": 0.3791, + "step": 5800 + }, + { + "epoch": 23.3, + "grad_norm": 2.2318499088287354, + "learning_rate": 2.8043549983503796e-05, + "loss": 0.3381, + "step": 5825 + }, + { + "epoch": 23.4, + "grad_norm": 1.6060622930526733, + "learning_rate": 2.6393929396238866e-05, + "loss": 0.3374, + "step": 5850 + }, + { + "epoch": 23.5, + "grad_norm": 4.242088317871094, + "learning_rate": 2.4744308808973936e-05, + "loss": 0.3234, + "step": 5875 + }, + { + "epoch": 23.6, + "grad_norm": 1.861912727355957, + "learning_rate": 2.309468822170901e-05, + "loss": 0.3507, + "step": 5900 + }, + { + "epoch": 23.7, + "grad_norm": 2.4759881496429443, + "learning_rate": 2.1445067634444078e-05, + "loss": 0.3562, + "step": 5925 + }, + { + "epoch": 23.8, + "grad_norm": 2.1233272552490234, + "learning_rate": 1.979544704717915e-05, + "loss": 0.3397, + "step": 5950 + }, + { + "epoch": 23.9, + "grad_norm": 2.3019630908966064, + "learning_rate": 1.814582645991422e-05, + "loss": 0.3103, + "step": 5975 + }, + { + "epoch": 24.0, + "grad_norm": 2.005906105041504, + "learning_rate": 1.6496205872649294e-05, + "loss": 0.3524, + "step": 6000 + } + ], + "logging_steps": 25, + "max_steps": 6250, + "num_input_tokens_seen": 0, + "num_train_epochs": 25, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 332286197760000.0, + "train_batch_size": 20, + "trial_name": null, + "trial_params": null +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/training_args.bin b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..10d16bb8e9df9c51aba38e7a6f04deeecf1ea29f --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:662a5f25d38ccae34f19d85e24456de6d3eaaabafca26fb7cc8816de0c990f20 +size 5304 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/vocab.txt b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/vocab.txt new file mode 100644 index 0000000000000000000000000000000000000000..cc9ad53732e21ea9180f18e1fb67a38cc5702d4c --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/vocab.txt @@ -0,0 +1,4101 @@ +[PAD] +[UNK] +[CLS] +[SEP] +[MASK] +AAAAAA +AAAAAT +AAAAAC +AAAAAG +AAAATA +AAAATT +AAAATC +AAAATG +AAAACA +AAAACT +AAAACC +AAAACG +AAAAGA +AAAAGT +AAAAGC +AAAAGG +AAATAA +AAATAT +AAATAC +AAATAG +AAATTA +AAATTT +AAATTC +AAATTG +AAATCA +AAATCT +AAATCC +AAATCG +AAATGA +AAATGT +AAATGC +AAATGG +AAACAA +AAACAT +AAACAC +AAACAG +AAACTA +AAACTT +AAACTC +AAACTG +AAACCA +AAACCT +AAACCC +AAACCG +AAACGA +AAACGT +AAACGC +AAACGG +AAAGAA +AAAGAT +AAAGAC +AAAGAG +AAAGTA +AAAGTT +AAAGTC +AAAGTG +AAAGCA +AAAGCT +AAAGCC +AAAGCG +AAAGGA +AAAGGT +AAAGGC +AAAGGG +AATAAA +AATAAT +AATAAC +AATAAG +AATATA +AATATT +AATATC +AATATG +AATACA +AATACT +AATACC +AATACG +AATAGA +AATAGT +AATAGC +AATAGG +AATTAA +AATTAT +AATTAC +AATTAG +AATTTA +AATTTT +AATTTC +AATTTG +AATTCA +AATTCT +AATTCC +AATTCG +AATTGA +AATTGT +AATTGC +AATTGG +AATCAA +AATCAT +AATCAC +AATCAG +AATCTA +AATCTT +AATCTC +AATCTG +AATCCA +AATCCT +AATCCC +AATCCG +AATCGA +AATCGT +AATCGC +AATCGG +AATGAA +AATGAT +AATGAC +AATGAG +AATGTA +AATGTT +AATGTC +AATGTG +AATGCA +AATGCT +AATGCC +AATGCG +AATGGA +AATGGT +AATGGC +AATGGG +AACAAA +AACAAT +AACAAC +AACAAG +AACATA +AACATT +AACATC +AACATG +AACACA +AACACT +AACACC +AACACG +AACAGA +AACAGT +AACAGC +AACAGG +AACTAA +AACTAT +AACTAC +AACTAG +AACTTA +AACTTT +AACTTC +AACTTG +AACTCA +AACTCT +AACTCC +AACTCG +AACTGA +AACTGT +AACTGC +AACTGG +AACCAA +AACCAT +AACCAC +AACCAG +AACCTA +AACCTT +AACCTC +AACCTG +AACCCA +AACCCT +AACCCC +AACCCG +AACCGA +AACCGT +AACCGC +AACCGG +AACGAA +AACGAT +AACGAC +AACGAG +AACGTA +AACGTT +AACGTC +AACGTG +AACGCA +AACGCT +AACGCC +AACGCG +AACGGA +AACGGT +AACGGC +AACGGG +AAGAAA +AAGAAT +AAGAAC +AAGAAG +AAGATA +AAGATT +AAGATC +AAGATG +AAGACA +AAGACT +AAGACC +AAGACG +AAGAGA +AAGAGT +AAGAGC +AAGAGG +AAGTAA +AAGTAT +AAGTAC +AAGTAG +AAGTTA +AAGTTT +AAGTTC +AAGTTG +AAGTCA +AAGTCT +AAGTCC +AAGTCG +AAGTGA +AAGTGT +AAGTGC +AAGTGG +AAGCAA +AAGCAT +AAGCAC +AAGCAG +AAGCTA +AAGCTT +AAGCTC +AAGCTG +AAGCCA +AAGCCT +AAGCCC +AAGCCG +AAGCGA +AAGCGT +AAGCGC +AAGCGG +AAGGAA +AAGGAT +AAGGAC +AAGGAG +AAGGTA +AAGGTT +AAGGTC +AAGGTG +AAGGCA +AAGGCT +AAGGCC +AAGGCG +AAGGGA +AAGGGT +AAGGGC +AAGGGG +ATAAAA +ATAAAT +ATAAAC +ATAAAG +ATAATA +ATAATT +ATAATC +ATAATG +ATAACA +ATAACT +ATAACC +ATAACG +ATAAGA +ATAAGT +ATAAGC +ATAAGG +ATATAA +ATATAT +ATATAC +ATATAG +ATATTA +ATATTT +ATATTC +ATATTG +ATATCA +ATATCT +ATATCC +ATATCG +ATATGA +ATATGT +ATATGC +ATATGG +ATACAA +ATACAT +ATACAC +ATACAG +ATACTA +ATACTT +ATACTC +ATACTG +ATACCA +ATACCT +ATACCC +ATACCG +ATACGA +ATACGT +ATACGC +ATACGG +ATAGAA +ATAGAT +ATAGAC +ATAGAG +ATAGTA +ATAGTT +ATAGTC +ATAGTG +ATAGCA +ATAGCT +ATAGCC +ATAGCG +ATAGGA +ATAGGT +ATAGGC +ATAGGG +ATTAAA +ATTAAT +ATTAAC +ATTAAG +ATTATA +ATTATT +ATTATC +ATTATG +ATTACA +ATTACT +ATTACC +ATTACG +ATTAGA +ATTAGT +ATTAGC +ATTAGG +ATTTAA +ATTTAT +ATTTAC +ATTTAG +ATTTTA +ATTTTT +ATTTTC +ATTTTG +ATTTCA +ATTTCT +ATTTCC +ATTTCG +ATTTGA +ATTTGT +ATTTGC +ATTTGG +ATTCAA +ATTCAT +ATTCAC +ATTCAG +ATTCTA +ATTCTT +ATTCTC +ATTCTG +ATTCCA +ATTCCT +ATTCCC +ATTCCG +ATTCGA +ATTCGT +ATTCGC +ATTCGG +ATTGAA +ATTGAT +ATTGAC +ATTGAG +ATTGTA +ATTGTT +ATTGTC +ATTGTG +ATTGCA +ATTGCT +ATTGCC +ATTGCG +ATTGGA +ATTGGT +ATTGGC +ATTGGG +ATCAAA +ATCAAT +ATCAAC +ATCAAG +ATCATA +ATCATT +ATCATC +ATCATG +ATCACA +ATCACT +ATCACC +ATCACG +ATCAGA +ATCAGT +ATCAGC +ATCAGG +ATCTAA +ATCTAT +ATCTAC +ATCTAG +ATCTTA +ATCTTT +ATCTTC +ATCTTG +ATCTCA +ATCTCT +ATCTCC +ATCTCG +ATCTGA +ATCTGT +ATCTGC +ATCTGG +ATCCAA +ATCCAT +ATCCAC +ATCCAG +ATCCTA +ATCCTT +ATCCTC +ATCCTG +ATCCCA +ATCCCT +ATCCCC +ATCCCG +ATCCGA +ATCCGT +ATCCGC +ATCCGG +ATCGAA +ATCGAT +ATCGAC +ATCGAG +ATCGTA +ATCGTT +ATCGTC +ATCGTG +ATCGCA +ATCGCT +ATCGCC +ATCGCG +ATCGGA +ATCGGT +ATCGGC +ATCGGG +ATGAAA +ATGAAT +ATGAAC +ATGAAG +ATGATA +ATGATT +ATGATC +ATGATG +ATGACA +ATGACT +ATGACC +ATGACG +ATGAGA +ATGAGT +ATGAGC +ATGAGG +ATGTAA +ATGTAT +ATGTAC +ATGTAG +ATGTTA +ATGTTT +ATGTTC +ATGTTG +ATGTCA +ATGTCT +ATGTCC +ATGTCG +ATGTGA +ATGTGT +ATGTGC +ATGTGG +ATGCAA +ATGCAT +ATGCAC +ATGCAG +ATGCTA +ATGCTT +ATGCTC +ATGCTG +ATGCCA +ATGCCT +ATGCCC +ATGCCG +ATGCGA +ATGCGT +ATGCGC +ATGCGG +ATGGAA +ATGGAT +ATGGAC +ATGGAG +ATGGTA +ATGGTT +ATGGTC +ATGGTG +ATGGCA +ATGGCT +ATGGCC +ATGGCG +ATGGGA +ATGGGT +ATGGGC +ATGGGG +ACAAAA +ACAAAT +ACAAAC +ACAAAG +ACAATA +ACAATT +ACAATC +ACAATG +ACAACA +ACAACT +ACAACC +ACAACG +ACAAGA +ACAAGT +ACAAGC +ACAAGG +ACATAA +ACATAT +ACATAC +ACATAG +ACATTA +ACATTT +ACATTC +ACATTG +ACATCA +ACATCT +ACATCC +ACATCG +ACATGA +ACATGT +ACATGC +ACATGG +ACACAA +ACACAT +ACACAC +ACACAG +ACACTA +ACACTT +ACACTC +ACACTG +ACACCA +ACACCT +ACACCC +ACACCG +ACACGA +ACACGT +ACACGC +ACACGG +ACAGAA +ACAGAT +ACAGAC +ACAGAG +ACAGTA +ACAGTT +ACAGTC +ACAGTG +ACAGCA +ACAGCT +ACAGCC +ACAGCG +ACAGGA +ACAGGT +ACAGGC +ACAGGG +ACTAAA +ACTAAT +ACTAAC +ACTAAG +ACTATA +ACTATT +ACTATC +ACTATG +ACTACA +ACTACT +ACTACC +ACTACG +ACTAGA +ACTAGT +ACTAGC +ACTAGG +ACTTAA +ACTTAT +ACTTAC +ACTTAG +ACTTTA +ACTTTT +ACTTTC +ACTTTG +ACTTCA +ACTTCT +ACTTCC +ACTTCG +ACTTGA +ACTTGT +ACTTGC +ACTTGG +ACTCAA +ACTCAT +ACTCAC +ACTCAG +ACTCTA +ACTCTT +ACTCTC +ACTCTG +ACTCCA +ACTCCT +ACTCCC +ACTCCG +ACTCGA +ACTCGT +ACTCGC +ACTCGG +ACTGAA +ACTGAT +ACTGAC +ACTGAG +ACTGTA +ACTGTT +ACTGTC +ACTGTG +ACTGCA +ACTGCT +ACTGCC +ACTGCG +ACTGGA +ACTGGT +ACTGGC +ACTGGG +ACCAAA +ACCAAT +ACCAAC +ACCAAG +ACCATA +ACCATT +ACCATC +ACCATG +ACCACA +ACCACT +ACCACC +ACCACG +ACCAGA +ACCAGT +ACCAGC +ACCAGG +ACCTAA +ACCTAT +ACCTAC +ACCTAG +ACCTTA +ACCTTT +ACCTTC +ACCTTG +ACCTCA +ACCTCT +ACCTCC +ACCTCG +ACCTGA +ACCTGT +ACCTGC +ACCTGG +ACCCAA +ACCCAT +ACCCAC +ACCCAG +ACCCTA +ACCCTT +ACCCTC +ACCCTG +ACCCCA +ACCCCT +ACCCCC +ACCCCG +ACCCGA +ACCCGT +ACCCGC +ACCCGG +ACCGAA +ACCGAT +ACCGAC +ACCGAG +ACCGTA +ACCGTT +ACCGTC +ACCGTG +ACCGCA +ACCGCT +ACCGCC +ACCGCG +ACCGGA +ACCGGT +ACCGGC +ACCGGG +ACGAAA +ACGAAT +ACGAAC +ACGAAG +ACGATA +ACGATT +ACGATC +ACGATG +ACGACA +ACGACT +ACGACC +ACGACG +ACGAGA +ACGAGT +ACGAGC +ACGAGG +ACGTAA +ACGTAT +ACGTAC +ACGTAG +ACGTTA +ACGTTT +ACGTTC +ACGTTG +ACGTCA +ACGTCT +ACGTCC +ACGTCG +ACGTGA +ACGTGT +ACGTGC +ACGTGG +ACGCAA +ACGCAT +ACGCAC +ACGCAG +ACGCTA +ACGCTT +ACGCTC +ACGCTG +ACGCCA +ACGCCT +ACGCCC +ACGCCG +ACGCGA +ACGCGT +ACGCGC +ACGCGG +ACGGAA +ACGGAT +ACGGAC +ACGGAG +ACGGTA +ACGGTT +ACGGTC +ACGGTG +ACGGCA +ACGGCT +ACGGCC +ACGGCG +ACGGGA +ACGGGT +ACGGGC +ACGGGG +AGAAAA +AGAAAT +AGAAAC +AGAAAG +AGAATA +AGAATT +AGAATC +AGAATG +AGAACA +AGAACT +AGAACC +AGAACG +AGAAGA +AGAAGT +AGAAGC +AGAAGG +AGATAA +AGATAT +AGATAC +AGATAG +AGATTA +AGATTT +AGATTC +AGATTG +AGATCA +AGATCT +AGATCC +AGATCG +AGATGA +AGATGT +AGATGC +AGATGG +AGACAA +AGACAT +AGACAC +AGACAG +AGACTA +AGACTT +AGACTC +AGACTG +AGACCA +AGACCT +AGACCC +AGACCG +AGACGA +AGACGT +AGACGC +AGACGG +AGAGAA +AGAGAT +AGAGAC +AGAGAG +AGAGTA +AGAGTT +AGAGTC +AGAGTG +AGAGCA +AGAGCT +AGAGCC +AGAGCG +AGAGGA +AGAGGT +AGAGGC +AGAGGG +AGTAAA +AGTAAT +AGTAAC +AGTAAG +AGTATA +AGTATT +AGTATC +AGTATG +AGTACA +AGTACT +AGTACC +AGTACG +AGTAGA +AGTAGT +AGTAGC +AGTAGG +AGTTAA +AGTTAT +AGTTAC +AGTTAG +AGTTTA +AGTTTT +AGTTTC +AGTTTG +AGTTCA +AGTTCT +AGTTCC +AGTTCG +AGTTGA +AGTTGT +AGTTGC +AGTTGG +AGTCAA +AGTCAT +AGTCAC +AGTCAG +AGTCTA +AGTCTT +AGTCTC +AGTCTG +AGTCCA +AGTCCT +AGTCCC +AGTCCG +AGTCGA +AGTCGT +AGTCGC +AGTCGG +AGTGAA +AGTGAT +AGTGAC +AGTGAG +AGTGTA +AGTGTT +AGTGTC +AGTGTG +AGTGCA +AGTGCT +AGTGCC +AGTGCG +AGTGGA +AGTGGT +AGTGGC +AGTGGG +AGCAAA +AGCAAT +AGCAAC +AGCAAG +AGCATA +AGCATT +AGCATC +AGCATG +AGCACA +AGCACT +AGCACC +AGCACG +AGCAGA +AGCAGT +AGCAGC +AGCAGG +AGCTAA +AGCTAT +AGCTAC +AGCTAG +AGCTTA +AGCTTT +AGCTTC +AGCTTG +AGCTCA +AGCTCT +AGCTCC +AGCTCG +AGCTGA +AGCTGT +AGCTGC +AGCTGG +AGCCAA +AGCCAT +AGCCAC +AGCCAG +AGCCTA +AGCCTT +AGCCTC +AGCCTG +AGCCCA +AGCCCT +AGCCCC +AGCCCG +AGCCGA +AGCCGT +AGCCGC +AGCCGG +AGCGAA +AGCGAT +AGCGAC +AGCGAG +AGCGTA +AGCGTT +AGCGTC +AGCGTG +AGCGCA +AGCGCT +AGCGCC +AGCGCG +AGCGGA +AGCGGT +AGCGGC +AGCGGG +AGGAAA +AGGAAT +AGGAAC +AGGAAG +AGGATA +AGGATT +AGGATC +AGGATG +AGGACA +AGGACT +AGGACC +AGGACG +AGGAGA +AGGAGT +AGGAGC +AGGAGG +AGGTAA +AGGTAT +AGGTAC +AGGTAG +AGGTTA +AGGTTT +AGGTTC +AGGTTG +AGGTCA +AGGTCT +AGGTCC +AGGTCG +AGGTGA +AGGTGT +AGGTGC +AGGTGG +AGGCAA +AGGCAT +AGGCAC +AGGCAG +AGGCTA +AGGCTT +AGGCTC +AGGCTG +AGGCCA +AGGCCT +AGGCCC +AGGCCG +AGGCGA +AGGCGT +AGGCGC +AGGCGG +AGGGAA +AGGGAT +AGGGAC +AGGGAG +AGGGTA +AGGGTT +AGGGTC +AGGGTG +AGGGCA +AGGGCT +AGGGCC +AGGGCG +AGGGGA +AGGGGT +AGGGGC +AGGGGG +TAAAAA +TAAAAT +TAAAAC +TAAAAG +TAAATA +TAAATT +TAAATC +TAAATG +TAAACA +TAAACT +TAAACC +TAAACG +TAAAGA +TAAAGT +TAAAGC +TAAAGG +TAATAA +TAATAT +TAATAC +TAATAG +TAATTA +TAATTT +TAATTC +TAATTG +TAATCA +TAATCT +TAATCC +TAATCG +TAATGA +TAATGT +TAATGC +TAATGG +TAACAA +TAACAT +TAACAC +TAACAG +TAACTA +TAACTT +TAACTC +TAACTG +TAACCA +TAACCT +TAACCC +TAACCG +TAACGA +TAACGT +TAACGC +TAACGG +TAAGAA +TAAGAT +TAAGAC +TAAGAG +TAAGTA +TAAGTT +TAAGTC +TAAGTG +TAAGCA +TAAGCT +TAAGCC +TAAGCG +TAAGGA +TAAGGT +TAAGGC +TAAGGG +TATAAA +TATAAT +TATAAC +TATAAG +TATATA +TATATT +TATATC +TATATG +TATACA +TATACT +TATACC +TATACG +TATAGA +TATAGT +TATAGC +TATAGG +TATTAA +TATTAT +TATTAC +TATTAG +TATTTA +TATTTT +TATTTC +TATTTG +TATTCA +TATTCT +TATTCC +TATTCG +TATTGA +TATTGT +TATTGC +TATTGG +TATCAA +TATCAT +TATCAC +TATCAG +TATCTA +TATCTT +TATCTC +TATCTG +TATCCA +TATCCT +TATCCC +TATCCG +TATCGA +TATCGT +TATCGC +TATCGG +TATGAA +TATGAT +TATGAC +TATGAG +TATGTA +TATGTT +TATGTC +TATGTG +TATGCA +TATGCT +TATGCC +TATGCG +TATGGA +TATGGT +TATGGC +TATGGG +TACAAA +TACAAT +TACAAC +TACAAG +TACATA +TACATT +TACATC +TACATG +TACACA +TACACT +TACACC +TACACG +TACAGA +TACAGT +TACAGC +TACAGG +TACTAA +TACTAT +TACTAC +TACTAG +TACTTA +TACTTT +TACTTC +TACTTG +TACTCA +TACTCT +TACTCC +TACTCG +TACTGA +TACTGT +TACTGC +TACTGG +TACCAA +TACCAT +TACCAC +TACCAG +TACCTA +TACCTT +TACCTC +TACCTG +TACCCA +TACCCT +TACCCC +TACCCG +TACCGA +TACCGT +TACCGC +TACCGG +TACGAA +TACGAT +TACGAC +TACGAG +TACGTA +TACGTT +TACGTC +TACGTG +TACGCA +TACGCT +TACGCC +TACGCG +TACGGA +TACGGT +TACGGC +TACGGG +TAGAAA +TAGAAT +TAGAAC +TAGAAG +TAGATA +TAGATT +TAGATC +TAGATG +TAGACA +TAGACT +TAGACC +TAGACG +TAGAGA +TAGAGT +TAGAGC +TAGAGG +TAGTAA +TAGTAT +TAGTAC +TAGTAG +TAGTTA +TAGTTT +TAGTTC +TAGTTG +TAGTCA +TAGTCT +TAGTCC +TAGTCG +TAGTGA +TAGTGT +TAGTGC +TAGTGG +TAGCAA +TAGCAT +TAGCAC +TAGCAG +TAGCTA +TAGCTT +TAGCTC +TAGCTG +TAGCCA +TAGCCT +TAGCCC +TAGCCG +TAGCGA +TAGCGT +TAGCGC +TAGCGG +TAGGAA +TAGGAT +TAGGAC +TAGGAG +TAGGTA +TAGGTT +TAGGTC +TAGGTG +TAGGCA +TAGGCT +TAGGCC +TAGGCG +TAGGGA +TAGGGT +TAGGGC +TAGGGG +TTAAAA +TTAAAT +TTAAAC +TTAAAG +TTAATA +TTAATT +TTAATC +TTAATG +TTAACA +TTAACT +TTAACC +TTAACG +TTAAGA +TTAAGT +TTAAGC +TTAAGG +TTATAA +TTATAT +TTATAC +TTATAG +TTATTA +TTATTT +TTATTC +TTATTG +TTATCA +TTATCT +TTATCC +TTATCG +TTATGA +TTATGT +TTATGC +TTATGG +TTACAA +TTACAT +TTACAC +TTACAG +TTACTA +TTACTT +TTACTC +TTACTG +TTACCA +TTACCT +TTACCC +TTACCG +TTACGA +TTACGT +TTACGC +TTACGG +TTAGAA +TTAGAT +TTAGAC +TTAGAG +TTAGTA +TTAGTT +TTAGTC +TTAGTG +TTAGCA +TTAGCT +TTAGCC +TTAGCG +TTAGGA +TTAGGT +TTAGGC +TTAGGG +TTTAAA +TTTAAT +TTTAAC +TTTAAG +TTTATA +TTTATT +TTTATC +TTTATG +TTTACA +TTTACT +TTTACC +TTTACG +TTTAGA +TTTAGT +TTTAGC +TTTAGG +TTTTAA +TTTTAT +TTTTAC +TTTTAG +TTTTTA +TTTTTT +TTTTTC +TTTTTG +TTTTCA +TTTTCT +TTTTCC +TTTTCG +TTTTGA +TTTTGT +TTTTGC +TTTTGG +TTTCAA +TTTCAT +TTTCAC +TTTCAG +TTTCTA +TTTCTT +TTTCTC +TTTCTG +TTTCCA +TTTCCT +TTTCCC +TTTCCG +TTTCGA +TTTCGT +TTTCGC +TTTCGG +TTTGAA +TTTGAT +TTTGAC +TTTGAG +TTTGTA +TTTGTT +TTTGTC +TTTGTG +TTTGCA +TTTGCT +TTTGCC +TTTGCG +TTTGGA +TTTGGT +TTTGGC +TTTGGG +TTCAAA +TTCAAT +TTCAAC +TTCAAG +TTCATA +TTCATT +TTCATC +TTCATG +TTCACA +TTCACT +TTCACC +TTCACG +TTCAGA +TTCAGT +TTCAGC +TTCAGG +TTCTAA +TTCTAT +TTCTAC +TTCTAG +TTCTTA +TTCTTT +TTCTTC +TTCTTG +TTCTCA +TTCTCT +TTCTCC +TTCTCG +TTCTGA +TTCTGT +TTCTGC +TTCTGG +TTCCAA +TTCCAT +TTCCAC +TTCCAG +TTCCTA +TTCCTT +TTCCTC +TTCCTG +TTCCCA +TTCCCT +TTCCCC +TTCCCG +TTCCGA +TTCCGT +TTCCGC +TTCCGG +TTCGAA +TTCGAT +TTCGAC +TTCGAG +TTCGTA +TTCGTT +TTCGTC +TTCGTG +TTCGCA +TTCGCT +TTCGCC +TTCGCG +TTCGGA +TTCGGT +TTCGGC +TTCGGG +TTGAAA +TTGAAT +TTGAAC +TTGAAG +TTGATA +TTGATT +TTGATC +TTGATG +TTGACA +TTGACT +TTGACC +TTGACG +TTGAGA +TTGAGT +TTGAGC +TTGAGG +TTGTAA +TTGTAT +TTGTAC +TTGTAG +TTGTTA +TTGTTT +TTGTTC +TTGTTG +TTGTCA +TTGTCT +TTGTCC +TTGTCG +TTGTGA +TTGTGT +TTGTGC +TTGTGG +TTGCAA +TTGCAT +TTGCAC +TTGCAG +TTGCTA +TTGCTT +TTGCTC +TTGCTG +TTGCCA +TTGCCT +TTGCCC +TTGCCG +TTGCGA +TTGCGT +TTGCGC +TTGCGG +TTGGAA +TTGGAT +TTGGAC +TTGGAG +TTGGTA +TTGGTT +TTGGTC +TTGGTG +TTGGCA +TTGGCT +TTGGCC +TTGGCG +TTGGGA +TTGGGT +TTGGGC +TTGGGG +TCAAAA +TCAAAT +TCAAAC +TCAAAG +TCAATA +TCAATT +TCAATC +TCAATG +TCAACA +TCAACT +TCAACC +TCAACG +TCAAGA +TCAAGT +TCAAGC +TCAAGG +TCATAA +TCATAT +TCATAC +TCATAG +TCATTA +TCATTT +TCATTC +TCATTG +TCATCA +TCATCT +TCATCC +TCATCG +TCATGA +TCATGT +TCATGC +TCATGG +TCACAA +TCACAT +TCACAC +TCACAG +TCACTA +TCACTT +TCACTC +TCACTG +TCACCA +TCACCT +TCACCC +TCACCG +TCACGA +TCACGT +TCACGC +TCACGG +TCAGAA +TCAGAT +TCAGAC +TCAGAG +TCAGTA +TCAGTT +TCAGTC +TCAGTG +TCAGCA +TCAGCT +TCAGCC +TCAGCG +TCAGGA +TCAGGT +TCAGGC +TCAGGG +TCTAAA +TCTAAT +TCTAAC +TCTAAG +TCTATA +TCTATT +TCTATC +TCTATG +TCTACA +TCTACT +TCTACC +TCTACG +TCTAGA +TCTAGT +TCTAGC +TCTAGG +TCTTAA +TCTTAT +TCTTAC +TCTTAG +TCTTTA +TCTTTT +TCTTTC +TCTTTG +TCTTCA +TCTTCT +TCTTCC +TCTTCG +TCTTGA +TCTTGT +TCTTGC +TCTTGG +TCTCAA +TCTCAT +TCTCAC +TCTCAG +TCTCTA +TCTCTT +TCTCTC +TCTCTG +TCTCCA +TCTCCT +TCTCCC +TCTCCG +TCTCGA +TCTCGT +TCTCGC +TCTCGG +TCTGAA +TCTGAT +TCTGAC +TCTGAG +TCTGTA +TCTGTT +TCTGTC +TCTGTG +TCTGCA +TCTGCT +TCTGCC +TCTGCG +TCTGGA +TCTGGT +TCTGGC +TCTGGG +TCCAAA +TCCAAT +TCCAAC +TCCAAG +TCCATA +TCCATT +TCCATC +TCCATG +TCCACA +TCCACT +TCCACC +TCCACG +TCCAGA +TCCAGT +TCCAGC +TCCAGG +TCCTAA +TCCTAT +TCCTAC +TCCTAG +TCCTTA +TCCTTT +TCCTTC +TCCTTG +TCCTCA +TCCTCT +TCCTCC +TCCTCG +TCCTGA +TCCTGT +TCCTGC +TCCTGG +TCCCAA +TCCCAT +TCCCAC +TCCCAG +TCCCTA +TCCCTT +TCCCTC +TCCCTG +TCCCCA +TCCCCT +TCCCCC +TCCCCG +TCCCGA +TCCCGT +TCCCGC +TCCCGG +TCCGAA +TCCGAT +TCCGAC +TCCGAG +TCCGTA +TCCGTT +TCCGTC +TCCGTG +TCCGCA +TCCGCT +TCCGCC +TCCGCG +TCCGGA +TCCGGT +TCCGGC +TCCGGG +TCGAAA +TCGAAT +TCGAAC +TCGAAG +TCGATA +TCGATT +TCGATC +TCGATG +TCGACA +TCGACT +TCGACC +TCGACG +TCGAGA +TCGAGT +TCGAGC +TCGAGG +TCGTAA +TCGTAT +TCGTAC +TCGTAG +TCGTTA +TCGTTT +TCGTTC +TCGTTG +TCGTCA +TCGTCT +TCGTCC +TCGTCG +TCGTGA +TCGTGT +TCGTGC +TCGTGG +TCGCAA +TCGCAT +TCGCAC +TCGCAG +TCGCTA +TCGCTT +TCGCTC +TCGCTG +TCGCCA +TCGCCT +TCGCCC +TCGCCG +TCGCGA +TCGCGT +TCGCGC +TCGCGG +TCGGAA +TCGGAT +TCGGAC +TCGGAG +TCGGTA +TCGGTT +TCGGTC +TCGGTG +TCGGCA +TCGGCT +TCGGCC +TCGGCG +TCGGGA +TCGGGT +TCGGGC +TCGGGG +TGAAAA +TGAAAT +TGAAAC +TGAAAG +TGAATA +TGAATT +TGAATC +TGAATG +TGAACA +TGAACT +TGAACC +TGAACG +TGAAGA +TGAAGT +TGAAGC +TGAAGG +TGATAA +TGATAT +TGATAC +TGATAG +TGATTA +TGATTT +TGATTC +TGATTG +TGATCA +TGATCT +TGATCC +TGATCG +TGATGA +TGATGT +TGATGC +TGATGG +TGACAA +TGACAT +TGACAC +TGACAG +TGACTA +TGACTT +TGACTC +TGACTG +TGACCA +TGACCT +TGACCC +TGACCG +TGACGA +TGACGT +TGACGC +TGACGG +TGAGAA +TGAGAT +TGAGAC +TGAGAG +TGAGTA +TGAGTT +TGAGTC +TGAGTG +TGAGCA +TGAGCT +TGAGCC +TGAGCG +TGAGGA +TGAGGT +TGAGGC +TGAGGG +TGTAAA +TGTAAT +TGTAAC +TGTAAG +TGTATA +TGTATT +TGTATC +TGTATG +TGTACA +TGTACT +TGTACC +TGTACG +TGTAGA +TGTAGT +TGTAGC +TGTAGG +TGTTAA +TGTTAT +TGTTAC +TGTTAG +TGTTTA +TGTTTT +TGTTTC +TGTTTG +TGTTCA +TGTTCT +TGTTCC +TGTTCG +TGTTGA +TGTTGT +TGTTGC +TGTTGG +TGTCAA +TGTCAT +TGTCAC +TGTCAG +TGTCTA +TGTCTT +TGTCTC +TGTCTG +TGTCCA +TGTCCT +TGTCCC +TGTCCG +TGTCGA +TGTCGT +TGTCGC +TGTCGG +TGTGAA +TGTGAT +TGTGAC +TGTGAG +TGTGTA +TGTGTT +TGTGTC +TGTGTG +TGTGCA +TGTGCT +TGTGCC +TGTGCG +TGTGGA +TGTGGT +TGTGGC +TGTGGG +TGCAAA +TGCAAT +TGCAAC +TGCAAG +TGCATA +TGCATT +TGCATC +TGCATG +TGCACA +TGCACT +TGCACC +TGCACG +TGCAGA +TGCAGT +TGCAGC +TGCAGG +TGCTAA +TGCTAT +TGCTAC +TGCTAG +TGCTTA +TGCTTT +TGCTTC +TGCTTG +TGCTCA +TGCTCT +TGCTCC +TGCTCG +TGCTGA +TGCTGT +TGCTGC +TGCTGG +TGCCAA +TGCCAT +TGCCAC +TGCCAG +TGCCTA +TGCCTT +TGCCTC +TGCCTG +TGCCCA +TGCCCT +TGCCCC +TGCCCG +TGCCGA +TGCCGT +TGCCGC +TGCCGG +TGCGAA +TGCGAT +TGCGAC +TGCGAG +TGCGTA +TGCGTT +TGCGTC +TGCGTG +TGCGCA +TGCGCT +TGCGCC +TGCGCG +TGCGGA +TGCGGT +TGCGGC +TGCGGG +TGGAAA +TGGAAT +TGGAAC +TGGAAG +TGGATA +TGGATT +TGGATC +TGGATG +TGGACA +TGGACT +TGGACC +TGGACG +TGGAGA +TGGAGT +TGGAGC +TGGAGG +TGGTAA +TGGTAT +TGGTAC +TGGTAG +TGGTTA +TGGTTT +TGGTTC +TGGTTG +TGGTCA +TGGTCT +TGGTCC +TGGTCG +TGGTGA +TGGTGT +TGGTGC +TGGTGG +TGGCAA +TGGCAT +TGGCAC +TGGCAG +TGGCTA +TGGCTT +TGGCTC +TGGCTG +TGGCCA +TGGCCT +TGGCCC +TGGCCG +TGGCGA +TGGCGT +TGGCGC +TGGCGG +TGGGAA +TGGGAT +TGGGAC +TGGGAG +TGGGTA +TGGGTT +TGGGTC +TGGGTG +TGGGCA +TGGGCT +TGGGCC +TGGGCG +TGGGGA +TGGGGT +TGGGGC +TGGGGG +CAAAAA +CAAAAT +CAAAAC +CAAAAG +CAAATA +CAAATT +CAAATC +CAAATG +CAAACA +CAAACT +CAAACC +CAAACG +CAAAGA +CAAAGT +CAAAGC +CAAAGG +CAATAA +CAATAT +CAATAC +CAATAG +CAATTA +CAATTT +CAATTC +CAATTG +CAATCA +CAATCT +CAATCC +CAATCG +CAATGA +CAATGT +CAATGC +CAATGG +CAACAA +CAACAT +CAACAC +CAACAG +CAACTA +CAACTT +CAACTC +CAACTG +CAACCA +CAACCT +CAACCC +CAACCG +CAACGA +CAACGT +CAACGC +CAACGG +CAAGAA +CAAGAT +CAAGAC +CAAGAG +CAAGTA +CAAGTT +CAAGTC +CAAGTG +CAAGCA +CAAGCT +CAAGCC +CAAGCG +CAAGGA +CAAGGT +CAAGGC +CAAGGG +CATAAA +CATAAT +CATAAC +CATAAG +CATATA +CATATT +CATATC +CATATG +CATACA +CATACT +CATACC +CATACG +CATAGA +CATAGT +CATAGC +CATAGG +CATTAA +CATTAT +CATTAC +CATTAG +CATTTA +CATTTT +CATTTC +CATTTG +CATTCA +CATTCT +CATTCC +CATTCG +CATTGA +CATTGT +CATTGC +CATTGG +CATCAA +CATCAT +CATCAC +CATCAG +CATCTA +CATCTT +CATCTC +CATCTG +CATCCA +CATCCT +CATCCC +CATCCG +CATCGA +CATCGT +CATCGC +CATCGG +CATGAA +CATGAT +CATGAC +CATGAG +CATGTA +CATGTT +CATGTC +CATGTG +CATGCA +CATGCT +CATGCC +CATGCG +CATGGA +CATGGT +CATGGC +CATGGG +CACAAA +CACAAT +CACAAC +CACAAG +CACATA +CACATT +CACATC +CACATG +CACACA +CACACT +CACACC +CACACG +CACAGA +CACAGT +CACAGC +CACAGG +CACTAA +CACTAT +CACTAC +CACTAG +CACTTA +CACTTT +CACTTC +CACTTG +CACTCA +CACTCT +CACTCC +CACTCG +CACTGA +CACTGT +CACTGC +CACTGG +CACCAA +CACCAT +CACCAC +CACCAG +CACCTA +CACCTT +CACCTC +CACCTG +CACCCA +CACCCT +CACCCC +CACCCG +CACCGA +CACCGT +CACCGC +CACCGG +CACGAA +CACGAT +CACGAC +CACGAG +CACGTA +CACGTT +CACGTC +CACGTG +CACGCA +CACGCT +CACGCC +CACGCG +CACGGA +CACGGT +CACGGC +CACGGG +CAGAAA +CAGAAT +CAGAAC +CAGAAG +CAGATA +CAGATT +CAGATC +CAGATG +CAGACA +CAGACT +CAGACC +CAGACG +CAGAGA +CAGAGT +CAGAGC +CAGAGG +CAGTAA +CAGTAT +CAGTAC +CAGTAG +CAGTTA +CAGTTT +CAGTTC +CAGTTG +CAGTCA +CAGTCT +CAGTCC +CAGTCG +CAGTGA +CAGTGT +CAGTGC +CAGTGG +CAGCAA +CAGCAT +CAGCAC +CAGCAG +CAGCTA +CAGCTT +CAGCTC +CAGCTG +CAGCCA +CAGCCT +CAGCCC +CAGCCG +CAGCGA +CAGCGT +CAGCGC +CAGCGG +CAGGAA +CAGGAT +CAGGAC +CAGGAG +CAGGTA +CAGGTT +CAGGTC +CAGGTG +CAGGCA +CAGGCT +CAGGCC +CAGGCG +CAGGGA +CAGGGT +CAGGGC +CAGGGG +CTAAAA +CTAAAT +CTAAAC +CTAAAG +CTAATA +CTAATT +CTAATC +CTAATG +CTAACA +CTAACT +CTAACC +CTAACG +CTAAGA +CTAAGT +CTAAGC +CTAAGG +CTATAA +CTATAT +CTATAC +CTATAG +CTATTA +CTATTT +CTATTC +CTATTG +CTATCA +CTATCT +CTATCC +CTATCG +CTATGA +CTATGT +CTATGC +CTATGG +CTACAA +CTACAT +CTACAC +CTACAG +CTACTA +CTACTT +CTACTC +CTACTG +CTACCA +CTACCT +CTACCC +CTACCG +CTACGA +CTACGT +CTACGC +CTACGG +CTAGAA +CTAGAT +CTAGAC +CTAGAG +CTAGTA +CTAGTT +CTAGTC +CTAGTG +CTAGCA +CTAGCT +CTAGCC +CTAGCG +CTAGGA +CTAGGT +CTAGGC +CTAGGG +CTTAAA +CTTAAT +CTTAAC +CTTAAG +CTTATA +CTTATT +CTTATC +CTTATG +CTTACA +CTTACT +CTTACC +CTTACG +CTTAGA +CTTAGT +CTTAGC +CTTAGG +CTTTAA +CTTTAT +CTTTAC +CTTTAG +CTTTTA +CTTTTT +CTTTTC +CTTTTG +CTTTCA +CTTTCT +CTTTCC +CTTTCG +CTTTGA +CTTTGT +CTTTGC +CTTTGG +CTTCAA +CTTCAT +CTTCAC +CTTCAG +CTTCTA +CTTCTT +CTTCTC +CTTCTG +CTTCCA +CTTCCT +CTTCCC +CTTCCG +CTTCGA +CTTCGT +CTTCGC +CTTCGG +CTTGAA +CTTGAT +CTTGAC +CTTGAG +CTTGTA +CTTGTT +CTTGTC +CTTGTG +CTTGCA +CTTGCT +CTTGCC +CTTGCG +CTTGGA +CTTGGT +CTTGGC +CTTGGG +CTCAAA +CTCAAT +CTCAAC +CTCAAG +CTCATA +CTCATT +CTCATC +CTCATG +CTCACA +CTCACT +CTCACC +CTCACG +CTCAGA +CTCAGT +CTCAGC +CTCAGG +CTCTAA +CTCTAT +CTCTAC +CTCTAG +CTCTTA +CTCTTT +CTCTTC +CTCTTG +CTCTCA +CTCTCT +CTCTCC +CTCTCG +CTCTGA +CTCTGT +CTCTGC +CTCTGG +CTCCAA +CTCCAT +CTCCAC +CTCCAG +CTCCTA +CTCCTT +CTCCTC +CTCCTG +CTCCCA +CTCCCT +CTCCCC +CTCCCG +CTCCGA +CTCCGT +CTCCGC +CTCCGG +CTCGAA +CTCGAT +CTCGAC +CTCGAG +CTCGTA +CTCGTT +CTCGTC +CTCGTG +CTCGCA +CTCGCT +CTCGCC +CTCGCG +CTCGGA +CTCGGT +CTCGGC +CTCGGG +CTGAAA +CTGAAT +CTGAAC +CTGAAG +CTGATA +CTGATT +CTGATC +CTGATG +CTGACA +CTGACT +CTGACC +CTGACG +CTGAGA +CTGAGT +CTGAGC +CTGAGG +CTGTAA +CTGTAT +CTGTAC +CTGTAG +CTGTTA +CTGTTT +CTGTTC +CTGTTG +CTGTCA +CTGTCT +CTGTCC +CTGTCG +CTGTGA +CTGTGT +CTGTGC +CTGTGG +CTGCAA +CTGCAT +CTGCAC +CTGCAG +CTGCTA +CTGCTT +CTGCTC +CTGCTG +CTGCCA +CTGCCT +CTGCCC +CTGCCG +CTGCGA +CTGCGT +CTGCGC +CTGCGG +CTGGAA +CTGGAT +CTGGAC +CTGGAG +CTGGTA +CTGGTT +CTGGTC +CTGGTG +CTGGCA +CTGGCT +CTGGCC +CTGGCG +CTGGGA +CTGGGT +CTGGGC +CTGGGG +CCAAAA +CCAAAT +CCAAAC +CCAAAG +CCAATA +CCAATT +CCAATC +CCAATG +CCAACA +CCAACT +CCAACC +CCAACG +CCAAGA +CCAAGT +CCAAGC +CCAAGG +CCATAA +CCATAT +CCATAC +CCATAG +CCATTA +CCATTT +CCATTC +CCATTG +CCATCA +CCATCT +CCATCC +CCATCG +CCATGA +CCATGT +CCATGC +CCATGG +CCACAA +CCACAT +CCACAC +CCACAG +CCACTA +CCACTT +CCACTC +CCACTG +CCACCA +CCACCT +CCACCC +CCACCG +CCACGA +CCACGT +CCACGC +CCACGG +CCAGAA +CCAGAT +CCAGAC +CCAGAG +CCAGTA +CCAGTT +CCAGTC +CCAGTG +CCAGCA +CCAGCT +CCAGCC +CCAGCG +CCAGGA +CCAGGT +CCAGGC +CCAGGG +CCTAAA +CCTAAT +CCTAAC +CCTAAG +CCTATA +CCTATT +CCTATC +CCTATG +CCTACA +CCTACT +CCTACC +CCTACG +CCTAGA +CCTAGT +CCTAGC +CCTAGG +CCTTAA +CCTTAT +CCTTAC +CCTTAG +CCTTTA +CCTTTT +CCTTTC +CCTTTG +CCTTCA +CCTTCT +CCTTCC +CCTTCG +CCTTGA +CCTTGT +CCTTGC +CCTTGG +CCTCAA +CCTCAT +CCTCAC +CCTCAG +CCTCTA +CCTCTT +CCTCTC +CCTCTG +CCTCCA +CCTCCT +CCTCCC +CCTCCG +CCTCGA +CCTCGT +CCTCGC +CCTCGG +CCTGAA +CCTGAT +CCTGAC +CCTGAG +CCTGTA +CCTGTT +CCTGTC +CCTGTG +CCTGCA +CCTGCT +CCTGCC +CCTGCG +CCTGGA +CCTGGT +CCTGGC +CCTGGG +CCCAAA +CCCAAT +CCCAAC +CCCAAG +CCCATA +CCCATT +CCCATC +CCCATG +CCCACA +CCCACT +CCCACC +CCCACG +CCCAGA +CCCAGT +CCCAGC +CCCAGG +CCCTAA +CCCTAT +CCCTAC +CCCTAG +CCCTTA +CCCTTT +CCCTTC +CCCTTG +CCCTCA +CCCTCT +CCCTCC +CCCTCG +CCCTGA +CCCTGT +CCCTGC +CCCTGG +CCCCAA +CCCCAT +CCCCAC +CCCCAG +CCCCTA +CCCCTT +CCCCTC +CCCCTG +CCCCCA +CCCCCT +CCCCCC +CCCCCG +CCCCGA +CCCCGT +CCCCGC +CCCCGG +CCCGAA +CCCGAT +CCCGAC +CCCGAG +CCCGTA +CCCGTT +CCCGTC +CCCGTG +CCCGCA +CCCGCT +CCCGCC +CCCGCG +CCCGGA +CCCGGT +CCCGGC +CCCGGG +CCGAAA +CCGAAT +CCGAAC +CCGAAG +CCGATA +CCGATT +CCGATC +CCGATG +CCGACA +CCGACT +CCGACC +CCGACG +CCGAGA +CCGAGT +CCGAGC +CCGAGG +CCGTAA +CCGTAT +CCGTAC +CCGTAG +CCGTTA +CCGTTT +CCGTTC +CCGTTG +CCGTCA +CCGTCT +CCGTCC +CCGTCG +CCGTGA +CCGTGT +CCGTGC +CCGTGG +CCGCAA +CCGCAT +CCGCAC +CCGCAG +CCGCTA +CCGCTT +CCGCTC +CCGCTG +CCGCCA +CCGCCT +CCGCCC +CCGCCG +CCGCGA +CCGCGT +CCGCGC +CCGCGG +CCGGAA +CCGGAT +CCGGAC +CCGGAG +CCGGTA +CCGGTT +CCGGTC +CCGGTG +CCGGCA +CCGGCT +CCGGCC +CCGGCG +CCGGGA +CCGGGT +CCGGGC +CCGGGG +CGAAAA +CGAAAT +CGAAAC +CGAAAG +CGAATA +CGAATT +CGAATC +CGAATG +CGAACA +CGAACT +CGAACC +CGAACG +CGAAGA +CGAAGT +CGAAGC +CGAAGG +CGATAA +CGATAT +CGATAC +CGATAG +CGATTA +CGATTT +CGATTC +CGATTG +CGATCA +CGATCT +CGATCC +CGATCG +CGATGA +CGATGT +CGATGC +CGATGG +CGACAA +CGACAT +CGACAC +CGACAG +CGACTA +CGACTT +CGACTC +CGACTG +CGACCA +CGACCT +CGACCC +CGACCG +CGACGA +CGACGT +CGACGC +CGACGG +CGAGAA +CGAGAT +CGAGAC +CGAGAG +CGAGTA +CGAGTT +CGAGTC +CGAGTG +CGAGCA +CGAGCT +CGAGCC +CGAGCG +CGAGGA +CGAGGT +CGAGGC +CGAGGG +CGTAAA +CGTAAT +CGTAAC +CGTAAG +CGTATA +CGTATT +CGTATC +CGTATG +CGTACA +CGTACT +CGTACC +CGTACG +CGTAGA +CGTAGT +CGTAGC +CGTAGG +CGTTAA +CGTTAT +CGTTAC +CGTTAG +CGTTTA +CGTTTT +CGTTTC +CGTTTG +CGTTCA +CGTTCT +CGTTCC +CGTTCG +CGTTGA +CGTTGT +CGTTGC +CGTTGG +CGTCAA +CGTCAT +CGTCAC +CGTCAG +CGTCTA +CGTCTT +CGTCTC +CGTCTG +CGTCCA +CGTCCT +CGTCCC +CGTCCG +CGTCGA +CGTCGT +CGTCGC +CGTCGG +CGTGAA +CGTGAT +CGTGAC +CGTGAG +CGTGTA +CGTGTT +CGTGTC +CGTGTG +CGTGCA +CGTGCT +CGTGCC +CGTGCG +CGTGGA +CGTGGT +CGTGGC +CGTGGG +CGCAAA +CGCAAT +CGCAAC +CGCAAG +CGCATA +CGCATT +CGCATC +CGCATG +CGCACA +CGCACT +CGCACC +CGCACG +CGCAGA +CGCAGT +CGCAGC +CGCAGG +CGCTAA +CGCTAT +CGCTAC +CGCTAG +CGCTTA +CGCTTT +CGCTTC +CGCTTG +CGCTCA +CGCTCT +CGCTCC +CGCTCG +CGCTGA +CGCTGT +CGCTGC +CGCTGG +CGCCAA +CGCCAT +CGCCAC +CGCCAG +CGCCTA +CGCCTT +CGCCTC +CGCCTG +CGCCCA +CGCCCT +CGCCCC +CGCCCG +CGCCGA +CGCCGT +CGCCGC +CGCCGG +CGCGAA +CGCGAT +CGCGAC +CGCGAG +CGCGTA +CGCGTT +CGCGTC +CGCGTG +CGCGCA +CGCGCT +CGCGCC +CGCGCG +CGCGGA +CGCGGT +CGCGGC +CGCGGG +CGGAAA +CGGAAT +CGGAAC +CGGAAG +CGGATA +CGGATT +CGGATC +CGGATG +CGGACA +CGGACT +CGGACC +CGGACG +CGGAGA +CGGAGT +CGGAGC +CGGAGG +CGGTAA +CGGTAT +CGGTAC +CGGTAG +CGGTTA +CGGTTT +CGGTTC +CGGTTG +CGGTCA +CGGTCT +CGGTCC +CGGTCG +CGGTGA +CGGTGT +CGGTGC +CGGTGG +CGGCAA +CGGCAT +CGGCAC +CGGCAG +CGGCTA +CGGCTT +CGGCTC +CGGCTG +CGGCCA +CGGCCT +CGGCCC +CGGCCG +CGGCGA +CGGCGT +CGGCGC +CGGCGG +CGGGAA +CGGGAT +CGGGAC +CGGGAG +CGGGTA +CGGGTT +CGGGTC +CGGGTG +CGGGCA +CGGGCT +CGGGCC +CGGGCG +CGGGGA +CGGGGT +CGGGGC +CGGGGG +GAAAAA +GAAAAT +GAAAAC +GAAAAG +GAAATA +GAAATT +GAAATC +GAAATG +GAAACA +GAAACT +GAAACC +GAAACG +GAAAGA +GAAAGT +GAAAGC +GAAAGG +GAATAA +GAATAT +GAATAC +GAATAG +GAATTA +GAATTT +GAATTC +GAATTG +GAATCA +GAATCT +GAATCC +GAATCG +GAATGA +GAATGT +GAATGC +GAATGG +GAACAA +GAACAT +GAACAC +GAACAG +GAACTA +GAACTT +GAACTC +GAACTG +GAACCA +GAACCT +GAACCC +GAACCG +GAACGA +GAACGT +GAACGC +GAACGG +GAAGAA +GAAGAT +GAAGAC +GAAGAG +GAAGTA +GAAGTT +GAAGTC +GAAGTG +GAAGCA +GAAGCT +GAAGCC +GAAGCG +GAAGGA +GAAGGT +GAAGGC +GAAGGG +GATAAA +GATAAT +GATAAC +GATAAG +GATATA +GATATT +GATATC +GATATG +GATACA +GATACT +GATACC +GATACG +GATAGA +GATAGT +GATAGC +GATAGG +GATTAA +GATTAT +GATTAC +GATTAG +GATTTA +GATTTT +GATTTC +GATTTG +GATTCA +GATTCT +GATTCC +GATTCG +GATTGA +GATTGT +GATTGC +GATTGG +GATCAA +GATCAT +GATCAC +GATCAG +GATCTA +GATCTT +GATCTC +GATCTG +GATCCA +GATCCT +GATCCC +GATCCG +GATCGA +GATCGT +GATCGC +GATCGG +GATGAA +GATGAT +GATGAC +GATGAG +GATGTA +GATGTT +GATGTC +GATGTG +GATGCA +GATGCT +GATGCC +GATGCG +GATGGA +GATGGT +GATGGC +GATGGG +GACAAA +GACAAT +GACAAC +GACAAG +GACATA +GACATT +GACATC +GACATG +GACACA +GACACT +GACACC +GACACG +GACAGA +GACAGT +GACAGC +GACAGG +GACTAA +GACTAT +GACTAC +GACTAG +GACTTA +GACTTT +GACTTC +GACTTG +GACTCA +GACTCT +GACTCC +GACTCG +GACTGA +GACTGT +GACTGC +GACTGG +GACCAA +GACCAT +GACCAC +GACCAG +GACCTA +GACCTT +GACCTC +GACCTG +GACCCA +GACCCT +GACCCC +GACCCG +GACCGA +GACCGT +GACCGC +GACCGG +GACGAA +GACGAT +GACGAC +GACGAG +GACGTA +GACGTT +GACGTC +GACGTG +GACGCA +GACGCT +GACGCC +GACGCG +GACGGA +GACGGT +GACGGC +GACGGG +GAGAAA +GAGAAT +GAGAAC +GAGAAG +GAGATA +GAGATT +GAGATC +GAGATG +GAGACA +GAGACT +GAGACC +GAGACG +GAGAGA +GAGAGT +GAGAGC +GAGAGG +GAGTAA +GAGTAT +GAGTAC +GAGTAG +GAGTTA +GAGTTT +GAGTTC +GAGTTG +GAGTCA +GAGTCT +GAGTCC +GAGTCG +GAGTGA +GAGTGT +GAGTGC +GAGTGG +GAGCAA +GAGCAT +GAGCAC +GAGCAG +GAGCTA +GAGCTT +GAGCTC +GAGCTG +GAGCCA +GAGCCT +GAGCCC +GAGCCG +GAGCGA +GAGCGT +GAGCGC +GAGCGG +GAGGAA +GAGGAT +GAGGAC +GAGGAG +GAGGTA +GAGGTT +GAGGTC +GAGGTG +GAGGCA +GAGGCT +GAGGCC +GAGGCG +GAGGGA +GAGGGT +GAGGGC +GAGGGG +GTAAAA +GTAAAT +GTAAAC +GTAAAG +GTAATA +GTAATT +GTAATC +GTAATG +GTAACA +GTAACT +GTAACC +GTAACG +GTAAGA +GTAAGT +GTAAGC +GTAAGG +GTATAA +GTATAT +GTATAC +GTATAG +GTATTA +GTATTT +GTATTC +GTATTG +GTATCA +GTATCT +GTATCC +GTATCG +GTATGA +GTATGT +GTATGC +GTATGG +GTACAA +GTACAT +GTACAC +GTACAG +GTACTA +GTACTT +GTACTC +GTACTG +GTACCA +GTACCT +GTACCC +GTACCG +GTACGA +GTACGT +GTACGC +GTACGG +GTAGAA +GTAGAT +GTAGAC +GTAGAG +GTAGTA +GTAGTT +GTAGTC +GTAGTG +GTAGCA +GTAGCT +GTAGCC +GTAGCG +GTAGGA +GTAGGT +GTAGGC +GTAGGG +GTTAAA +GTTAAT +GTTAAC +GTTAAG +GTTATA +GTTATT +GTTATC +GTTATG +GTTACA +GTTACT +GTTACC +GTTACG +GTTAGA +GTTAGT +GTTAGC +GTTAGG +GTTTAA +GTTTAT +GTTTAC +GTTTAG +GTTTTA +GTTTTT +GTTTTC +GTTTTG +GTTTCA +GTTTCT +GTTTCC +GTTTCG +GTTTGA +GTTTGT +GTTTGC +GTTTGG +GTTCAA +GTTCAT +GTTCAC +GTTCAG +GTTCTA +GTTCTT +GTTCTC +GTTCTG +GTTCCA +GTTCCT +GTTCCC +GTTCCG +GTTCGA +GTTCGT +GTTCGC +GTTCGG +GTTGAA +GTTGAT +GTTGAC +GTTGAG +GTTGTA +GTTGTT +GTTGTC +GTTGTG +GTTGCA +GTTGCT +GTTGCC +GTTGCG +GTTGGA +GTTGGT +GTTGGC +GTTGGG +GTCAAA +GTCAAT +GTCAAC +GTCAAG +GTCATA +GTCATT +GTCATC +GTCATG +GTCACA +GTCACT +GTCACC +GTCACG +GTCAGA +GTCAGT +GTCAGC +GTCAGG +GTCTAA +GTCTAT +GTCTAC +GTCTAG +GTCTTA +GTCTTT +GTCTTC +GTCTTG +GTCTCA +GTCTCT +GTCTCC +GTCTCG +GTCTGA +GTCTGT +GTCTGC +GTCTGG +GTCCAA +GTCCAT +GTCCAC +GTCCAG +GTCCTA +GTCCTT +GTCCTC +GTCCTG +GTCCCA +GTCCCT +GTCCCC +GTCCCG +GTCCGA +GTCCGT +GTCCGC +GTCCGG +GTCGAA +GTCGAT +GTCGAC +GTCGAG +GTCGTA +GTCGTT +GTCGTC +GTCGTG +GTCGCA +GTCGCT +GTCGCC +GTCGCG +GTCGGA +GTCGGT +GTCGGC +GTCGGG +GTGAAA +GTGAAT +GTGAAC +GTGAAG +GTGATA +GTGATT +GTGATC +GTGATG +GTGACA +GTGACT +GTGACC +GTGACG +GTGAGA +GTGAGT +GTGAGC +GTGAGG +GTGTAA +GTGTAT +GTGTAC +GTGTAG +GTGTTA +GTGTTT +GTGTTC +GTGTTG +GTGTCA +GTGTCT +GTGTCC +GTGTCG +GTGTGA +GTGTGT +GTGTGC +GTGTGG +GTGCAA +GTGCAT +GTGCAC +GTGCAG +GTGCTA +GTGCTT +GTGCTC +GTGCTG +GTGCCA +GTGCCT +GTGCCC +GTGCCG +GTGCGA +GTGCGT +GTGCGC +GTGCGG +GTGGAA +GTGGAT +GTGGAC +GTGGAG +GTGGTA +GTGGTT +GTGGTC +GTGGTG +GTGGCA +GTGGCT +GTGGCC +GTGGCG +GTGGGA +GTGGGT +GTGGGC +GTGGGG +GCAAAA +GCAAAT +GCAAAC +GCAAAG +GCAATA +GCAATT +GCAATC +GCAATG +GCAACA +GCAACT +GCAACC +GCAACG +GCAAGA +GCAAGT +GCAAGC +GCAAGG +GCATAA +GCATAT +GCATAC +GCATAG +GCATTA +GCATTT +GCATTC +GCATTG +GCATCA +GCATCT +GCATCC +GCATCG +GCATGA +GCATGT +GCATGC +GCATGG +GCACAA +GCACAT +GCACAC +GCACAG +GCACTA +GCACTT +GCACTC +GCACTG +GCACCA +GCACCT +GCACCC +GCACCG +GCACGA +GCACGT +GCACGC +GCACGG +GCAGAA +GCAGAT +GCAGAC +GCAGAG +GCAGTA +GCAGTT +GCAGTC +GCAGTG +GCAGCA +GCAGCT +GCAGCC +GCAGCG +GCAGGA +GCAGGT +GCAGGC +GCAGGG +GCTAAA +GCTAAT +GCTAAC +GCTAAG +GCTATA +GCTATT +GCTATC +GCTATG +GCTACA +GCTACT +GCTACC +GCTACG +GCTAGA +GCTAGT +GCTAGC +GCTAGG +GCTTAA +GCTTAT +GCTTAC +GCTTAG +GCTTTA +GCTTTT +GCTTTC +GCTTTG +GCTTCA +GCTTCT +GCTTCC +GCTTCG +GCTTGA +GCTTGT +GCTTGC +GCTTGG +GCTCAA +GCTCAT +GCTCAC +GCTCAG +GCTCTA +GCTCTT +GCTCTC +GCTCTG +GCTCCA +GCTCCT +GCTCCC +GCTCCG +GCTCGA +GCTCGT +GCTCGC +GCTCGG +GCTGAA +GCTGAT +GCTGAC +GCTGAG +GCTGTA +GCTGTT +GCTGTC +GCTGTG +GCTGCA +GCTGCT +GCTGCC +GCTGCG +GCTGGA +GCTGGT +GCTGGC +GCTGGG +GCCAAA +GCCAAT +GCCAAC +GCCAAG +GCCATA +GCCATT +GCCATC +GCCATG +GCCACA +GCCACT +GCCACC +GCCACG +GCCAGA +GCCAGT +GCCAGC +GCCAGG +GCCTAA +GCCTAT +GCCTAC +GCCTAG +GCCTTA +GCCTTT +GCCTTC +GCCTTG +GCCTCA +GCCTCT +GCCTCC +GCCTCG +GCCTGA +GCCTGT +GCCTGC +GCCTGG +GCCCAA +GCCCAT +GCCCAC +GCCCAG +GCCCTA +GCCCTT +GCCCTC +GCCCTG +GCCCCA +GCCCCT +GCCCCC +GCCCCG +GCCCGA +GCCCGT +GCCCGC +GCCCGG +GCCGAA +GCCGAT +GCCGAC +GCCGAG +GCCGTA +GCCGTT +GCCGTC +GCCGTG +GCCGCA +GCCGCT +GCCGCC +GCCGCG +GCCGGA +GCCGGT +GCCGGC +GCCGGG +GCGAAA +GCGAAT +GCGAAC +GCGAAG +GCGATA +GCGATT +GCGATC +GCGATG +GCGACA +GCGACT +GCGACC +GCGACG +GCGAGA +GCGAGT +GCGAGC +GCGAGG +GCGTAA +GCGTAT +GCGTAC +GCGTAG +GCGTTA +GCGTTT +GCGTTC +GCGTTG +GCGTCA +GCGTCT +GCGTCC +GCGTCG +GCGTGA +GCGTGT +GCGTGC +GCGTGG +GCGCAA +GCGCAT +GCGCAC +GCGCAG +GCGCTA +GCGCTT +GCGCTC +GCGCTG +GCGCCA +GCGCCT +GCGCCC +GCGCCG +GCGCGA +GCGCGT +GCGCGC +GCGCGG +GCGGAA +GCGGAT +GCGGAC +GCGGAG +GCGGTA +GCGGTT +GCGGTC +GCGGTG +GCGGCA +GCGGCT +GCGGCC +GCGGCG +GCGGGA +GCGGGT +GCGGGC +GCGGGG +GGAAAA +GGAAAT +GGAAAC +GGAAAG +GGAATA +GGAATT +GGAATC +GGAATG +GGAACA +GGAACT +GGAACC +GGAACG +GGAAGA +GGAAGT +GGAAGC +GGAAGG +GGATAA +GGATAT +GGATAC +GGATAG +GGATTA +GGATTT +GGATTC +GGATTG +GGATCA +GGATCT +GGATCC +GGATCG +GGATGA +GGATGT +GGATGC +GGATGG +GGACAA +GGACAT +GGACAC +GGACAG +GGACTA +GGACTT +GGACTC +GGACTG +GGACCA +GGACCT +GGACCC +GGACCG +GGACGA +GGACGT +GGACGC +GGACGG +GGAGAA +GGAGAT +GGAGAC +GGAGAG +GGAGTA +GGAGTT +GGAGTC +GGAGTG +GGAGCA +GGAGCT +GGAGCC +GGAGCG +GGAGGA +GGAGGT +GGAGGC +GGAGGG +GGTAAA +GGTAAT +GGTAAC +GGTAAG +GGTATA +GGTATT +GGTATC +GGTATG +GGTACA +GGTACT +GGTACC +GGTACG +GGTAGA +GGTAGT +GGTAGC +GGTAGG +GGTTAA +GGTTAT +GGTTAC +GGTTAG +GGTTTA +GGTTTT +GGTTTC +GGTTTG +GGTTCA +GGTTCT +GGTTCC +GGTTCG +GGTTGA +GGTTGT +GGTTGC +GGTTGG +GGTCAA +GGTCAT +GGTCAC +GGTCAG +GGTCTA +GGTCTT +GGTCTC +GGTCTG +GGTCCA +GGTCCT +GGTCCC +GGTCCG +GGTCGA +GGTCGT +GGTCGC +GGTCGG +GGTGAA +GGTGAT +GGTGAC +GGTGAG +GGTGTA +GGTGTT +GGTGTC +GGTGTG +GGTGCA +GGTGCT +GGTGCC +GGTGCG +GGTGGA +GGTGGT +GGTGGC +GGTGGG +GGCAAA +GGCAAT +GGCAAC +GGCAAG +GGCATA +GGCATT +GGCATC +GGCATG +GGCACA +GGCACT +GGCACC +GGCACG +GGCAGA +GGCAGT +GGCAGC +GGCAGG +GGCTAA +GGCTAT +GGCTAC +GGCTAG +GGCTTA +GGCTTT +GGCTTC +GGCTTG +GGCTCA +GGCTCT +GGCTCC +GGCTCG +GGCTGA +GGCTGT +GGCTGC +GGCTGG +GGCCAA +GGCCAT +GGCCAC +GGCCAG +GGCCTA +GGCCTT +GGCCTC +GGCCTG +GGCCCA +GGCCCT +GGCCCC +GGCCCG +GGCCGA +GGCCGT +GGCCGC +GGCCGG +GGCGAA +GGCGAT +GGCGAC +GGCGAG +GGCGTA +GGCGTT +GGCGTC +GGCGTG +GGCGCA +GGCGCT +GGCGCC +GGCGCG +GGCGGA +GGCGGT +GGCGGC +GGCGGG +GGGAAA +GGGAAT +GGGAAC +GGGAAG +GGGATA +GGGATT +GGGATC +GGGATG +GGGACA +GGGACT +GGGACC +GGGACG +GGGAGA +GGGAGT +GGGAGC +GGGAGG +GGGTAA +GGGTAT +GGGTAC +GGGTAG +GGGTTA +GGGTTT +GGGTTC +GGGTTG +GGGTCA +GGGTCT +GGGTCC +GGGTCG +GGGTGA +GGGTGT +GGGTGC +GGGTGG +GGGCAA +GGGCAT +GGGCAC +GGGCAG +GGGCTA +GGGCTT +GGGCTC +GGGCTG +GGGCCA +GGGCCT +GGGCCC +GGGCCG +GGGCGA +GGGCGT +GGGCGC +GGGCGG +GGGGAA +GGGGAT +GGGGAC +GGGGAG +GGGGTA +GGGGTT +GGGGTC +GGGGTG +GGGGCA +GGGGCT +GGGGCC +GGGGCG +GGGGGA +GGGGGT +GGGGGC +GGGGGG diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/README.md b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/README.md new file mode 100644 index 0000000000000000000000000000000000000000..500fd976d92b217455f181bc98c77e612d9a18d9 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/README.md @@ -0,0 +1,202 @@ +--- +base_model: fabihamakhdoomi/TinyDNABERT +library_name: peft +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.12.0 \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/adapter_config.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..04c63e99bbc6b810449dece367c0afb165125b92 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/adapter_config.json @@ -0,0 +1,42 @@ +{ + "alpha_pattern": {}, + "auto_mapping": null, + "base_model_name_or_path": "fabihamakhdoomi/TinyDNABERT", + "bias": "none", + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 16, + "lora_dropout": 0.2, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": [ + "classifier", + "score" + ], + "peft_type": "LORA", + "r": 8, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "bert.encoder.layer.1.attention.self.query", + "bert.encoder.layer.0.output.dense", + "bert.encoder.layer.1.output.dense", + "bert.encoder.layer.0.intermediate.dense", + "bert.encoder.layer.1.intermediate.dense", + "bert.encoder.layer.0.attention.output.dense", + "bert.encoder.layer.1.attention.output.dense", + "bert.encoder.layer.0.attention.self.query", + "bert.encoder.layer.0.attention.self.value", + "bert.encoder.layer.1.attention.self.key", + "bert.encoder.layer.1.attention.self.value", + "bert.encoder.layer.0.attention.self.key" + ], + "task_type": "SEQ_CLS", + "use_dora": false, + "use_rslora": false +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/adapter_model.safetensors b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/adapter_model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..3c9344a5bf908d67f4d459d3410bc5e3d583a370 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/adapter_model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a71c9d93546be6204db043675c7431a05738bb42c64d40c0bc612e01da989d8c +size 151952 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/optimizer.pt b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..c70b8e39effb3c0f2ade7702a6344eaa43cc88ac --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7b7b062724392af6146fd49588472fa52a40c8d258a394af21ed409adf334bc4 +size 312682 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/rng_state.pth b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..506b93da132609b3f461ebc7645f04db448a3371 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:77b9489769cad6c0f8fc553b5b129c47588998775afee5582baf3b7ef2be1cfa +size 14244 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/scheduler.pt b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..4c34d9dcafc0eb1507a84da1af7c2c5228be082e --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:20611d193eaf6658e6192005c679e5aaa561ac33254f734047a2032edc3e8f9b +size 1064 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/special_tokens_map.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..a8b3208c2884c4efb86e49300fdd3dc877220cdf --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/special_tokens_map.json @@ -0,0 +1,7 @@ +{ + "cls_token": "[CLS]", + "mask_token": "[MASK]", + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/tokenizer.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..ebd782de6963b9032ce9374cfe30f5275be1a30b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/tokenizer.json @@ -0,0 +1,4265 @@ +{ + "version": "1.0", + "truncation": { + "direction": "Right", + "max_length": 512, + "strategy": "LongestFirst", + "stride": 0 + }, + "padding": { + "strategy": { + "Fixed": 512 + }, + "direction": "Right", + "pad_to_multiple_of": null, + "pad_id": 0, + "pad_type_id": 0, + "pad_token": "[PAD]" + }, + "added_tokens": [ + { + "id": 0, + "content": "[PAD]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 1, + "content": "[UNK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 2, + "content": "[CLS]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 3, + "content": "[SEP]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 4, + "content": "[MASK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + } + ], + "normalizer": { + "type": "BertNormalizer", + "clean_text": true, + "handle_chinese_chars": true, + "strip_accents": null, + "lowercase": false + }, + "pre_tokenizer": { + "type": "BertPreTokenizer" + }, + "post_processor": { + "type": "TemplateProcessing", + "single": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + } + ], + "pair": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "B", + "type_id": 1 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 1 + } + } + ], + "special_tokens": { + "[CLS]": { + "id": "[CLS]", + "ids": [ + 2 + ], + "tokens": [ + "[CLS]" + ] + }, + "[SEP]": { + "id": "[SEP]", + "ids": [ + 3 + ], + "tokens": [ + "[SEP]" + ] + } + } + }, + "decoder": { + "type": "WordPiece", + "prefix": "##", + "cleanup": true + }, + "model": { + "type": "WordPiece", + "unk_token": "[UNK]", + "continuing_subword_prefix": "##", + "max_input_chars_per_word": 100, + "vocab": { + "[PAD]": 0, + "[UNK]": 1, + "[CLS]": 2, + "[SEP]": 3, + "[MASK]": 4, + "AAAAAA": 5, + "AAAAAT": 6, + "AAAAAC": 7, + "AAAAAG": 8, + "AAAATA": 9, + "AAAATT": 10, + "AAAATC": 11, + "AAAATG": 12, + "AAAACA": 13, + "AAAACT": 14, + "AAAACC": 15, + "AAAACG": 16, + "AAAAGA": 17, + "AAAAGT": 18, + "AAAAGC": 19, + "AAAAGG": 20, + "AAATAA": 21, + "AAATAT": 22, + "AAATAC": 23, + "AAATAG": 24, + "AAATTA": 25, + "AAATTT": 26, + "AAATTC": 27, + "AAATTG": 28, + "AAATCA": 29, + "AAATCT": 30, + "AAATCC": 31, + "AAATCG": 32, + "AAATGA": 33, + "AAATGT": 34, + "AAATGC": 35, + "AAATGG": 36, + "AAACAA": 37, + "AAACAT": 38, + "AAACAC": 39, + "AAACAG": 40, + "AAACTA": 41, + "AAACTT": 42, + "AAACTC": 43, + "AAACTG": 44, + "AAACCA": 45, + "AAACCT": 46, + "AAACCC": 47, + "AAACCG": 48, + "AAACGA": 49, + "AAACGT": 50, + "AAACGC": 51, + "AAACGG": 52, + "AAAGAA": 53, + "AAAGAT": 54, + "AAAGAC": 55, + "AAAGAG": 56, + "AAAGTA": 57, + "AAAGTT": 58, + "AAAGTC": 59, + "AAAGTG": 60, + "AAAGCA": 61, + "AAAGCT": 62, + "AAAGCC": 63, + "AAAGCG": 64, + "AAAGGA": 65, + "AAAGGT": 66, + "AAAGGC": 67, + "AAAGGG": 68, + "AATAAA": 69, + "AATAAT": 70, + "AATAAC": 71, + "AATAAG": 72, + "AATATA": 73, + "AATATT": 74, + "AATATC": 75, + "AATATG": 76, + "AATACA": 77, + "AATACT": 78, + "AATACC": 79, + "AATACG": 80, + "AATAGA": 81, + "AATAGT": 82, + "AATAGC": 83, + "AATAGG": 84, + "AATTAA": 85, + "AATTAT": 86, + "AATTAC": 87, + "AATTAG": 88, + "AATTTA": 89, + "AATTTT": 90, + "AATTTC": 91, + "AATTTG": 92, + "AATTCA": 93, + "AATTCT": 94, + "AATTCC": 95, + "AATTCG": 96, + "AATTGA": 97, + "AATTGT": 98, + "AATTGC": 99, + "AATTGG": 100, + "AATCAA": 101, + "AATCAT": 102, + "AATCAC": 103, + "AATCAG": 104, + "AATCTA": 105, + "AATCTT": 106, + "AATCTC": 107, + "AATCTG": 108, + "AATCCA": 109, + "AATCCT": 110, + "AATCCC": 111, + "AATCCG": 112, + "AATCGA": 113, + "AATCGT": 114, + "AATCGC": 115, + "AATCGG": 116, + "AATGAA": 117, + "AATGAT": 118, + "AATGAC": 119, + "AATGAG": 120, + "AATGTA": 121, + "AATGTT": 122, + "AATGTC": 123, + "AATGTG": 124, + "AATGCA": 125, + "AATGCT": 126, + "AATGCC": 127, + "AATGCG": 128, + "AATGGA": 129, + "AATGGT": 130, + "AATGGC": 131, + "AATGGG": 132, + "AACAAA": 133, + "AACAAT": 134, + "AACAAC": 135, + "AACAAG": 136, + "AACATA": 137, + "AACATT": 138, + "AACATC": 139, + "AACATG": 140, + "AACACA": 141, + "AACACT": 142, + "AACACC": 143, + "AACACG": 144, + "AACAGA": 145, + "AACAGT": 146, + "AACAGC": 147, + "AACAGG": 148, + "AACTAA": 149, + "AACTAT": 150, + "AACTAC": 151, + "AACTAG": 152, + "AACTTA": 153, + "AACTTT": 154, + "AACTTC": 155, + "AACTTG": 156, + "AACTCA": 157, + "AACTCT": 158, + "AACTCC": 159, + "AACTCG": 160, + "AACTGA": 161, + "AACTGT": 162, + "AACTGC": 163, + "AACTGG": 164, + "AACCAA": 165, + "AACCAT": 166, + "AACCAC": 167, + "AACCAG": 168, + "AACCTA": 169, + "AACCTT": 170, + "AACCTC": 171, + "AACCTG": 172, + "AACCCA": 173, + "AACCCT": 174, + "AACCCC": 175, + "AACCCG": 176, + "AACCGA": 177, + "AACCGT": 178, + "AACCGC": 179, + "AACCGG": 180, + "AACGAA": 181, + "AACGAT": 182, + "AACGAC": 183, + "AACGAG": 184, + "AACGTA": 185, + "AACGTT": 186, + "AACGTC": 187, + "AACGTG": 188, + "AACGCA": 189, + "AACGCT": 190, + "AACGCC": 191, + "AACGCG": 192, + "AACGGA": 193, + "AACGGT": 194, + "AACGGC": 195, + "AACGGG": 196, + "AAGAAA": 197, + "AAGAAT": 198, + "AAGAAC": 199, + "AAGAAG": 200, + "AAGATA": 201, + "AAGATT": 202, + "AAGATC": 203, + "AAGATG": 204, + "AAGACA": 205, + "AAGACT": 206, + "AAGACC": 207, + "AAGACG": 208, + "AAGAGA": 209, + "AAGAGT": 210, + "AAGAGC": 211, + "AAGAGG": 212, + "AAGTAA": 213, + "AAGTAT": 214, + "AAGTAC": 215, + "AAGTAG": 216, + "AAGTTA": 217, + "AAGTTT": 218, + "AAGTTC": 219, + "AAGTTG": 220, + "AAGTCA": 221, + "AAGTCT": 222, + "AAGTCC": 223, + "AAGTCG": 224, + "AAGTGA": 225, + "AAGTGT": 226, + "AAGTGC": 227, + "AAGTGG": 228, + "AAGCAA": 229, + "AAGCAT": 230, + "AAGCAC": 231, + "AAGCAG": 232, + "AAGCTA": 233, + "AAGCTT": 234, + "AAGCTC": 235, + "AAGCTG": 236, + "AAGCCA": 237, + "AAGCCT": 238, + "AAGCCC": 239, + "AAGCCG": 240, + "AAGCGA": 241, + "AAGCGT": 242, + "AAGCGC": 243, + "AAGCGG": 244, + "AAGGAA": 245, + "AAGGAT": 246, + "AAGGAC": 247, + "AAGGAG": 248, + "AAGGTA": 249, + "AAGGTT": 250, + "AAGGTC": 251, + "AAGGTG": 252, + "AAGGCA": 253, + "AAGGCT": 254, + "AAGGCC": 255, + "AAGGCG": 256, + "AAGGGA": 257, + "AAGGGT": 258, + "AAGGGC": 259, + "AAGGGG": 260, + "ATAAAA": 261, + "ATAAAT": 262, + "ATAAAC": 263, + "ATAAAG": 264, + "ATAATA": 265, + "ATAATT": 266, + "ATAATC": 267, + "ATAATG": 268, + "ATAACA": 269, + "ATAACT": 270, + "ATAACC": 271, + "ATAACG": 272, + "ATAAGA": 273, + "ATAAGT": 274, + "ATAAGC": 275, + "ATAAGG": 276, + "ATATAA": 277, + "ATATAT": 278, + "ATATAC": 279, + "ATATAG": 280, + "ATATTA": 281, + "ATATTT": 282, + "ATATTC": 283, + "ATATTG": 284, + "ATATCA": 285, + "ATATCT": 286, + "ATATCC": 287, + "ATATCG": 288, + "ATATGA": 289, + "ATATGT": 290, + "ATATGC": 291, + "ATATGG": 292, + "ATACAA": 293, + "ATACAT": 294, + "ATACAC": 295, + "ATACAG": 296, + "ATACTA": 297, + "ATACTT": 298, + "ATACTC": 299, + "ATACTG": 300, + "ATACCA": 301, + "ATACCT": 302, + "ATACCC": 303, + "ATACCG": 304, + "ATACGA": 305, + "ATACGT": 306, + "ATACGC": 307, + "ATACGG": 308, + "ATAGAA": 309, + "ATAGAT": 310, + "ATAGAC": 311, + "ATAGAG": 312, + "ATAGTA": 313, + "ATAGTT": 314, + "ATAGTC": 315, + "ATAGTG": 316, + "ATAGCA": 317, + "ATAGCT": 318, + "ATAGCC": 319, + "ATAGCG": 320, + "ATAGGA": 321, + "ATAGGT": 322, + "ATAGGC": 323, + "ATAGGG": 324, + "ATTAAA": 325, + "ATTAAT": 326, + "ATTAAC": 327, + "ATTAAG": 328, + "ATTATA": 329, + "ATTATT": 330, + "ATTATC": 331, + "ATTATG": 332, + "ATTACA": 333, + "ATTACT": 334, + "ATTACC": 335, + "ATTACG": 336, + "ATTAGA": 337, + "ATTAGT": 338, + "ATTAGC": 339, + "ATTAGG": 340, + "ATTTAA": 341, + "ATTTAT": 342, + "ATTTAC": 343, + "ATTTAG": 344, + "ATTTTA": 345, + "ATTTTT": 346, + "ATTTTC": 347, + "ATTTTG": 348, + "ATTTCA": 349, + "ATTTCT": 350, + "ATTTCC": 351, + "ATTTCG": 352, + "ATTTGA": 353, + "ATTTGT": 354, + "ATTTGC": 355, + "ATTTGG": 356, + "ATTCAA": 357, + "ATTCAT": 358, + "ATTCAC": 359, + "ATTCAG": 360, + "ATTCTA": 361, + "ATTCTT": 362, + "ATTCTC": 363, + "ATTCTG": 364, + "ATTCCA": 365, + "ATTCCT": 366, + "ATTCCC": 367, + "ATTCCG": 368, + "ATTCGA": 369, + "ATTCGT": 370, + "ATTCGC": 371, + "ATTCGG": 372, + "ATTGAA": 373, + "ATTGAT": 374, + "ATTGAC": 375, + "ATTGAG": 376, + "ATTGTA": 377, + "ATTGTT": 378, + "ATTGTC": 379, + "ATTGTG": 380, + "ATTGCA": 381, + "ATTGCT": 382, + "ATTGCC": 383, + "ATTGCG": 384, + "ATTGGA": 385, + "ATTGGT": 386, + "ATTGGC": 387, + "ATTGGG": 388, + "ATCAAA": 389, + "ATCAAT": 390, + "ATCAAC": 391, + "ATCAAG": 392, + "ATCATA": 393, + "ATCATT": 394, + "ATCATC": 395, + "ATCATG": 396, + "ATCACA": 397, + "ATCACT": 398, + "ATCACC": 399, + "ATCACG": 400, + "ATCAGA": 401, + "ATCAGT": 402, + "ATCAGC": 403, + "ATCAGG": 404, + "ATCTAA": 405, + "ATCTAT": 406, + "ATCTAC": 407, + "ATCTAG": 408, + "ATCTTA": 409, + "ATCTTT": 410, + "ATCTTC": 411, + "ATCTTG": 412, + "ATCTCA": 413, + "ATCTCT": 414, + "ATCTCC": 415, + "ATCTCG": 416, + "ATCTGA": 417, + "ATCTGT": 418, + "ATCTGC": 419, + "ATCTGG": 420, + "ATCCAA": 421, + "ATCCAT": 422, + "ATCCAC": 423, + "ATCCAG": 424, + "ATCCTA": 425, + "ATCCTT": 426, + "ATCCTC": 427, + "ATCCTG": 428, + "ATCCCA": 429, + "ATCCCT": 430, + "ATCCCC": 431, + "ATCCCG": 432, + "ATCCGA": 433, + "ATCCGT": 434, + "ATCCGC": 435, + "ATCCGG": 436, + "ATCGAA": 437, + "ATCGAT": 438, + "ATCGAC": 439, + "ATCGAG": 440, + "ATCGTA": 441, + "ATCGTT": 442, + "ATCGTC": 443, + "ATCGTG": 444, + "ATCGCA": 445, + "ATCGCT": 446, + "ATCGCC": 447, + "ATCGCG": 448, + "ATCGGA": 449, + "ATCGGT": 450, + "ATCGGC": 451, + "ATCGGG": 452, + "ATGAAA": 453, + "ATGAAT": 454, + "ATGAAC": 455, + "ATGAAG": 456, + "ATGATA": 457, + "ATGATT": 458, + "ATGATC": 459, + "ATGATG": 460, + "ATGACA": 461, + "ATGACT": 462, + "ATGACC": 463, + "ATGACG": 464, + "ATGAGA": 465, + "ATGAGT": 466, + "ATGAGC": 467, + "ATGAGG": 468, + "ATGTAA": 469, + "ATGTAT": 470, + "ATGTAC": 471, + "ATGTAG": 472, + "ATGTTA": 473, + "ATGTTT": 474, + "ATGTTC": 475, + "ATGTTG": 476, + "ATGTCA": 477, + "ATGTCT": 478, + "ATGTCC": 479, + "ATGTCG": 480, + "ATGTGA": 481, + "ATGTGT": 482, + "ATGTGC": 483, + "ATGTGG": 484, + "ATGCAA": 485, + "ATGCAT": 486, + "ATGCAC": 487, + "ATGCAG": 488, + "ATGCTA": 489, + "ATGCTT": 490, + "ATGCTC": 491, + "ATGCTG": 492, + "ATGCCA": 493, + "ATGCCT": 494, + "ATGCCC": 495, + "ATGCCG": 496, + "ATGCGA": 497, + "ATGCGT": 498, + "ATGCGC": 499, + "ATGCGG": 500, + "ATGGAA": 501, + "ATGGAT": 502, + "ATGGAC": 503, + "ATGGAG": 504, + "ATGGTA": 505, + "ATGGTT": 506, + "ATGGTC": 507, + "ATGGTG": 508, + "ATGGCA": 509, + "ATGGCT": 510, + "ATGGCC": 511, + "ATGGCG": 512, + "ATGGGA": 513, + "ATGGGT": 514, + "ATGGGC": 515, + "ATGGGG": 516, + "ACAAAA": 517, + "ACAAAT": 518, + "ACAAAC": 519, + "ACAAAG": 520, + "ACAATA": 521, + "ACAATT": 522, + "ACAATC": 523, + "ACAATG": 524, + "ACAACA": 525, + "ACAACT": 526, + "ACAACC": 527, + "ACAACG": 528, + "ACAAGA": 529, + "ACAAGT": 530, + "ACAAGC": 531, + "ACAAGG": 532, + "ACATAA": 533, + "ACATAT": 534, + "ACATAC": 535, + "ACATAG": 536, + "ACATTA": 537, + "ACATTT": 538, + "ACATTC": 539, + "ACATTG": 540, + "ACATCA": 541, + "ACATCT": 542, + "ACATCC": 543, + "ACATCG": 544, + "ACATGA": 545, + "ACATGT": 546, + "ACATGC": 547, + "ACATGG": 548, + "ACACAA": 549, + "ACACAT": 550, + "ACACAC": 551, + "ACACAG": 552, + "ACACTA": 553, + "ACACTT": 554, + "ACACTC": 555, + "ACACTG": 556, + "ACACCA": 557, + "ACACCT": 558, + "ACACCC": 559, + "ACACCG": 560, + "ACACGA": 561, + "ACACGT": 562, + "ACACGC": 563, + "ACACGG": 564, + "ACAGAA": 565, + "ACAGAT": 566, + "ACAGAC": 567, + "ACAGAG": 568, + "ACAGTA": 569, + "ACAGTT": 570, + "ACAGTC": 571, + "ACAGTG": 572, + "ACAGCA": 573, + "ACAGCT": 574, + "ACAGCC": 575, + "ACAGCG": 576, + "ACAGGA": 577, + "ACAGGT": 578, + "ACAGGC": 579, + "ACAGGG": 580, + "ACTAAA": 581, + "ACTAAT": 582, + "ACTAAC": 583, + "ACTAAG": 584, + "ACTATA": 585, + "ACTATT": 586, + "ACTATC": 587, + "ACTATG": 588, + "ACTACA": 589, + "ACTACT": 590, + "ACTACC": 591, + "ACTACG": 592, + "ACTAGA": 593, + "ACTAGT": 594, + "ACTAGC": 595, + "ACTAGG": 596, + "ACTTAA": 597, + "ACTTAT": 598, + "ACTTAC": 599, + "ACTTAG": 600, + "ACTTTA": 601, + "ACTTTT": 602, + "ACTTTC": 603, + "ACTTTG": 604, + "ACTTCA": 605, + "ACTTCT": 606, + "ACTTCC": 607, + "ACTTCG": 608, + "ACTTGA": 609, + "ACTTGT": 610, + "ACTTGC": 611, + "ACTTGG": 612, + "ACTCAA": 613, + "ACTCAT": 614, + "ACTCAC": 615, + "ACTCAG": 616, + "ACTCTA": 617, + "ACTCTT": 618, + "ACTCTC": 619, + "ACTCTG": 620, + "ACTCCA": 621, + "ACTCCT": 622, + "ACTCCC": 623, + "ACTCCG": 624, + "ACTCGA": 625, + "ACTCGT": 626, + "ACTCGC": 627, + "ACTCGG": 628, + "ACTGAA": 629, + "ACTGAT": 630, + "ACTGAC": 631, + "ACTGAG": 632, + "ACTGTA": 633, + "ACTGTT": 634, + "ACTGTC": 635, + "ACTGTG": 636, + "ACTGCA": 637, + "ACTGCT": 638, + "ACTGCC": 639, + "ACTGCG": 640, + "ACTGGA": 641, + "ACTGGT": 642, + "ACTGGC": 643, + "ACTGGG": 644, + "ACCAAA": 645, + "ACCAAT": 646, + "ACCAAC": 647, + "ACCAAG": 648, + "ACCATA": 649, + "ACCATT": 650, + "ACCATC": 651, + "ACCATG": 652, + "ACCACA": 653, + "ACCACT": 654, + "ACCACC": 655, + "ACCACG": 656, + "ACCAGA": 657, + "ACCAGT": 658, + "ACCAGC": 659, + "ACCAGG": 660, + "ACCTAA": 661, + "ACCTAT": 662, + "ACCTAC": 663, + "ACCTAG": 664, + "ACCTTA": 665, + "ACCTTT": 666, + "ACCTTC": 667, + "ACCTTG": 668, + "ACCTCA": 669, + "ACCTCT": 670, + "ACCTCC": 671, + "ACCTCG": 672, + "ACCTGA": 673, + "ACCTGT": 674, + "ACCTGC": 675, + "ACCTGG": 676, + "ACCCAA": 677, + "ACCCAT": 678, + "ACCCAC": 679, + "ACCCAG": 680, + "ACCCTA": 681, + "ACCCTT": 682, + "ACCCTC": 683, + "ACCCTG": 684, + "ACCCCA": 685, + "ACCCCT": 686, + "ACCCCC": 687, + "ACCCCG": 688, + "ACCCGA": 689, + "ACCCGT": 690, + "ACCCGC": 691, + "ACCCGG": 692, + "ACCGAA": 693, + "ACCGAT": 694, + "ACCGAC": 695, + "ACCGAG": 696, + "ACCGTA": 697, + "ACCGTT": 698, + "ACCGTC": 699, + "ACCGTG": 700, + "ACCGCA": 701, + "ACCGCT": 702, + "ACCGCC": 703, + "ACCGCG": 704, + "ACCGGA": 705, + "ACCGGT": 706, + "ACCGGC": 707, + "ACCGGG": 708, + "ACGAAA": 709, + "ACGAAT": 710, + "ACGAAC": 711, + "ACGAAG": 712, + "ACGATA": 713, + "ACGATT": 714, + "ACGATC": 715, + "ACGATG": 716, + "ACGACA": 717, + "ACGACT": 718, + "ACGACC": 719, + "ACGACG": 720, + "ACGAGA": 721, + "ACGAGT": 722, + "ACGAGC": 723, + "ACGAGG": 724, + "ACGTAA": 725, + "ACGTAT": 726, + "ACGTAC": 727, + "ACGTAG": 728, + "ACGTTA": 729, + "ACGTTT": 730, + "ACGTTC": 731, + "ACGTTG": 732, + "ACGTCA": 733, + "ACGTCT": 734, + "ACGTCC": 735, + "ACGTCG": 736, + "ACGTGA": 737, + "ACGTGT": 738, + "ACGTGC": 739, + "ACGTGG": 740, + "ACGCAA": 741, + "ACGCAT": 742, + "ACGCAC": 743, + "ACGCAG": 744, + "ACGCTA": 745, + "ACGCTT": 746, + "ACGCTC": 747, + "ACGCTG": 748, + "ACGCCA": 749, + "ACGCCT": 750, + "ACGCCC": 751, + "ACGCCG": 752, + "ACGCGA": 753, + "ACGCGT": 754, + "ACGCGC": 755, + "ACGCGG": 756, + "ACGGAA": 757, + "ACGGAT": 758, + "ACGGAC": 759, + "ACGGAG": 760, + "ACGGTA": 761, + "ACGGTT": 762, + "ACGGTC": 763, + "ACGGTG": 764, + "ACGGCA": 765, + "ACGGCT": 766, + "ACGGCC": 767, + "ACGGCG": 768, + "ACGGGA": 769, + "ACGGGT": 770, + "ACGGGC": 771, + "ACGGGG": 772, + "AGAAAA": 773, + "AGAAAT": 774, + "AGAAAC": 775, + "AGAAAG": 776, + "AGAATA": 777, + "AGAATT": 778, + "AGAATC": 779, + "AGAATG": 780, + "AGAACA": 781, + "AGAACT": 782, + "AGAACC": 783, + "AGAACG": 784, + "AGAAGA": 785, + "AGAAGT": 786, + "AGAAGC": 787, + "AGAAGG": 788, + "AGATAA": 789, + "AGATAT": 790, + "AGATAC": 791, + "AGATAG": 792, + "AGATTA": 793, + "AGATTT": 794, + "AGATTC": 795, + "AGATTG": 796, + "AGATCA": 797, + "AGATCT": 798, + "AGATCC": 799, + "AGATCG": 800, + "AGATGA": 801, + "AGATGT": 802, + "AGATGC": 803, + "AGATGG": 804, + "AGACAA": 805, + "AGACAT": 806, + "AGACAC": 807, + "AGACAG": 808, + "AGACTA": 809, + "AGACTT": 810, + "AGACTC": 811, + "AGACTG": 812, + "AGACCA": 813, + "AGACCT": 814, + "AGACCC": 815, + "AGACCG": 816, + "AGACGA": 817, + "AGACGT": 818, + "AGACGC": 819, + "AGACGG": 820, + "AGAGAA": 821, + "AGAGAT": 822, + "AGAGAC": 823, + "AGAGAG": 824, + "AGAGTA": 825, + "AGAGTT": 826, + "AGAGTC": 827, + "AGAGTG": 828, + "AGAGCA": 829, + "AGAGCT": 830, + "AGAGCC": 831, + "AGAGCG": 832, + "AGAGGA": 833, + "AGAGGT": 834, + "AGAGGC": 835, + "AGAGGG": 836, + "AGTAAA": 837, + "AGTAAT": 838, + "AGTAAC": 839, + "AGTAAG": 840, + "AGTATA": 841, + "AGTATT": 842, + "AGTATC": 843, + "AGTATG": 844, + "AGTACA": 845, + "AGTACT": 846, + "AGTACC": 847, + "AGTACG": 848, + "AGTAGA": 849, + "AGTAGT": 850, + "AGTAGC": 851, + "AGTAGG": 852, + "AGTTAA": 853, + "AGTTAT": 854, + "AGTTAC": 855, + "AGTTAG": 856, + "AGTTTA": 857, + "AGTTTT": 858, + "AGTTTC": 859, + "AGTTTG": 860, + "AGTTCA": 861, + "AGTTCT": 862, + "AGTTCC": 863, + "AGTTCG": 864, + "AGTTGA": 865, + "AGTTGT": 866, + "AGTTGC": 867, + "AGTTGG": 868, + "AGTCAA": 869, + "AGTCAT": 870, + "AGTCAC": 871, + "AGTCAG": 872, + "AGTCTA": 873, + "AGTCTT": 874, + "AGTCTC": 875, + "AGTCTG": 876, + "AGTCCA": 877, + "AGTCCT": 878, + "AGTCCC": 879, + "AGTCCG": 880, + "AGTCGA": 881, + "AGTCGT": 882, + "AGTCGC": 883, + "AGTCGG": 884, + "AGTGAA": 885, + "AGTGAT": 886, + "AGTGAC": 887, + "AGTGAG": 888, + "AGTGTA": 889, + "AGTGTT": 890, + "AGTGTC": 891, + "AGTGTG": 892, + "AGTGCA": 893, + "AGTGCT": 894, + "AGTGCC": 895, + "AGTGCG": 896, + "AGTGGA": 897, + "AGTGGT": 898, + "AGTGGC": 899, + "AGTGGG": 900, + "AGCAAA": 901, + "AGCAAT": 902, + "AGCAAC": 903, + "AGCAAG": 904, + "AGCATA": 905, + "AGCATT": 906, + "AGCATC": 907, + "AGCATG": 908, + "AGCACA": 909, + "AGCACT": 910, + "AGCACC": 911, + "AGCACG": 912, + "AGCAGA": 913, + "AGCAGT": 914, + "AGCAGC": 915, + "AGCAGG": 916, + "AGCTAA": 917, + "AGCTAT": 918, + "AGCTAC": 919, + "AGCTAG": 920, + "AGCTTA": 921, + "AGCTTT": 922, + "AGCTTC": 923, + "AGCTTG": 924, + "AGCTCA": 925, + "AGCTCT": 926, + "AGCTCC": 927, + "AGCTCG": 928, + "AGCTGA": 929, + "AGCTGT": 930, + "AGCTGC": 931, + "AGCTGG": 932, + "AGCCAA": 933, + "AGCCAT": 934, + "AGCCAC": 935, + "AGCCAG": 936, + "AGCCTA": 937, + "AGCCTT": 938, + "AGCCTC": 939, + "AGCCTG": 940, + "AGCCCA": 941, + "AGCCCT": 942, + "AGCCCC": 943, + "AGCCCG": 944, + "AGCCGA": 945, + "AGCCGT": 946, + "AGCCGC": 947, + "AGCCGG": 948, + "AGCGAA": 949, + "AGCGAT": 950, + "AGCGAC": 951, + "AGCGAG": 952, + "AGCGTA": 953, + "AGCGTT": 954, + "AGCGTC": 955, + "AGCGTG": 956, + "AGCGCA": 957, + "AGCGCT": 958, + "AGCGCC": 959, + "AGCGCG": 960, + "AGCGGA": 961, + "AGCGGT": 962, + "AGCGGC": 963, + "AGCGGG": 964, + "AGGAAA": 965, + "AGGAAT": 966, + "AGGAAC": 967, + "AGGAAG": 968, + "AGGATA": 969, + "AGGATT": 970, + "AGGATC": 971, + "AGGATG": 972, + "AGGACA": 973, + "AGGACT": 974, + "AGGACC": 975, + "AGGACG": 976, + "AGGAGA": 977, + "AGGAGT": 978, + "AGGAGC": 979, + "AGGAGG": 980, + "AGGTAA": 981, + "AGGTAT": 982, + "AGGTAC": 983, + "AGGTAG": 984, + "AGGTTA": 985, + "AGGTTT": 986, + "AGGTTC": 987, + "AGGTTG": 988, + "AGGTCA": 989, + "AGGTCT": 990, + "AGGTCC": 991, + "AGGTCG": 992, + "AGGTGA": 993, + "AGGTGT": 994, + "AGGTGC": 995, + "AGGTGG": 996, + "AGGCAA": 997, + "AGGCAT": 998, + "AGGCAC": 999, + "AGGCAG": 1000, + "AGGCTA": 1001, + "AGGCTT": 1002, + "AGGCTC": 1003, + "AGGCTG": 1004, + "AGGCCA": 1005, + "AGGCCT": 1006, + "AGGCCC": 1007, + "AGGCCG": 1008, + "AGGCGA": 1009, + "AGGCGT": 1010, + "AGGCGC": 1011, + "AGGCGG": 1012, + "AGGGAA": 1013, + "AGGGAT": 1014, + "AGGGAC": 1015, + "AGGGAG": 1016, + "AGGGTA": 1017, + "AGGGTT": 1018, + "AGGGTC": 1019, + "AGGGTG": 1020, + "AGGGCA": 1021, + "AGGGCT": 1022, + "AGGGCC": 1023, + "AGGGCG": 1024, + "AGGGGA": 1025, + "AGGGGT": 1026, + "AGGGGC": 1027, + "AGGGGG": 1028, + "TAAAAA": 1029, + "TAAAAT": 1030, + "TAAAAC": 1031, + "TAAAAG": 1032, + "TAAATA": 1033, + "TAAATT": 1034, + "TAAATC": 1035, + "TAAATG": 1036, + "TAAACA": 1037, + "TAAACT": 1038, + "TAAACC": 1039, + "TAAACG": 1040, + "TAAAGA": 1041, + "TAAAGT": 1042, + "TAAAGC": 1043, + "TAAAGG": 1044, + "TAATAA": 1045, + "TAATAT": 1046, + "TAATAC": 1047, + "TAATAG": 1048, + "TAATTA": 1049, + "TAATTT": 1050, + "TAATTC": 1051, + "TAATTG": 1052, + "TAATCA": 1053, + "TAATCT": 1054, + "TAATCC": 1055, + "TAATCG": 1056, + "TAATGA": 1057, + "TAATGT": 1058, + "TAATGC": 1059, + "TAATGG": 1060, + "TAACAA": 1061, + "TAACAT": 1062, + "TAACAC": 1063, + "TAACAG": 1064, + "TAACTA": 1065, + "TAACTT": 1066, + "TAACTC": 1067, + "TAACTG": 1068, + "TAACCA": 1069, + "TAACCT": 1070, + "TAACCC": 1071, + "TAACCG": 1072, + "TAACGA": 1073, + "TAACGT": 1074, + "TAACGC": 1075, + "TAACGG": 1076, + "TAAGAA": 1077, + "TAAGAT": 1078, + "TAAGAC": 1079, + "TAAGAG": 1080, + "TAAGTA": 1081, + "TAAGTT": 1082, + "TAAGTC": 1083, + "TAAGTG": 1084, + "TAAGCA": 1085, + "TAAGCT": 1086, + "TAAGCC": 1087, + "TAAGCG": 1088, + "TAAGGA": 1089, + "TAAGGT": 1090, + "TAAGGC": 1091, + "TAAGGG": 1092, + "TATAAA": 1093, + "TATAAT": 1094, + "TATAAC": 1095, + "TATAAG": 1096, + "TATATA": 1097, + "TATATT": 1098, + "TATATC": 1099, + "TATATG": 1100, + "TATACA": 1101, + "TATACT": 1102, + "TATACC": 1103, + "TATACG": 1104, + "TATAGA": 1105, + "TATAGT": 1106, + "TATAGC": 1107, + "TATAGG": 1108, + "TATTAA": 1109, + "TATTAT": 1110, + "TATTAC": 1111, + "TATTAG": 1112, + "TATTTA": 1113, + "TATTTT": 1114, + "TATTTC": 1115, + "TATTTG": 1116, + "TATTCA": 1117, + "TATTCT": 1118, + "TATTCC": 1119, + "TATTCG": 1120, + "TATTGA": 1121, + "TATTGT": 1122, + "TATTGC": 1123, + "TATTGG": 1124, + "TATCAA": 1125, + "TATCAT": 1126, + "TATCAC": 1127, + "TATCAG": 1128, + "TATCTA": 1129, + "TATCTT": 1130, + "TATCTC": 1131, + "TATCTG": 1132, + "TATCCA": 1133, + "TATCCT": 1134, + "TATCCC": 1135, + "TATCCG": 1136, + "TATCGA": 1137, + "TATCGT": 1138, + "TATCGC": 1139, + "TATCGG": 1140, + "TATGAA": 1141, + "TATGAT": 1142, + "TATGAC": 1143, + "TATGAG": 1144, + "TATGTA": 1145, + "TATGTT": 1146, + "TATGTC": 1147, + "TATGTG": 1148, + "TATGCA": 1149, + "TATGCT": 1150, + "TATGCC": 1151, + "TATGCG": 1152, + "TATGGA": 1153, + "TATGGT": 1154, + "TATGGC": 1155, + "TATGGG": 1156, + "TACAAA": 1157, + "TACAAT": 1158, + "TACAAC": 1159, + "TACAAG": 1160, + "TACATA": 1161, + "TACATT": 1162, + "TACATC": 1163, + "TACATG": 1164, + "TACACA": 1165, + "TACACT": 1166, + "TACACC": 1167, + "TACACG": 1168, + "TACAGA": 1169, + "TACAGT": 1170, + "TACAGC": 1171, + "TACAGG": 1172, + "TACTAA": 1173, + "TACTAT": 1174, + "TACTAC": 1175, + "TACTAG": 1176, + "TACTTA": 1177, + "TACTTT": 1178, + "TACTTC": 1179, + "TACTTG": 1180, + "TACTCA": 1181, + "TACTCT": 1182, + "TACTCC": 1183, + "TACTCG": 1184, + "TACTGA": 1185, + "TACTGT": 1186, + "TACTGC": 1187, + "TACTGG": 1188, + "TACCAA": 1189, + "TACCAT": 1190, + "TACCAC": 1191, + "TACCAG": 1192, + "TACCTA": 1193, + "TACCTT": 1194, + "TACCTC": 1195, + "TACCTG": 1196, + "TACCCA": 1197, + "TACCCT": 1198, + "TACCCC": 1199, + "TACCCG": 1200, + "TACCGA": 1201, + "TACCGT": 1202, + "TACCGC": 1203, + "TACCGG": 1204, + "TACGAA": 1205, + "TACGAT": 1206, + "TACGAC": 1207, + "TACGAG": 1208, + "TACGTA": 1209, + "TACGTT": 1210, + "TACGTC": 1211, + "TACGTG": 1212, + "TACGCA": 1213, + "TACGCT": 1214, + "TACGCC": 1215, + "TACGCG": 1216, + "TACGGA": 1217, + "TACGGT": 1218, + "TACGGC": 1219, + "TACGGG": 1220, + "TAGAAA": 1221, + "TAGAAT": 1222, + "TAGAAC": 1223, + "TAGAAG": 1224, + "TAGATA": 1225, + "TAGATT": 1226, + "TAGATC": 1227, + "TAGATG": 1228, + "TAGACA": 1229, + "TAGACT": 1230, + "TAGACC": 1231, + "TAGACG": 1232, + "TAGAGA": 1233, + "TAGAGT": 1234, + "TAGAGC": 1235, + "TAGAGG": 1236, + "TAGTAA": 1237, + "TAGTAT": 1238, + "TAGTAC": 1239, + "TAGTAG": 1240, + "TAGTTA": 1241, + "TAGTTT": 1242, + "TAGTTC": 1243, + "TAGTTG": 1244, + "TAGTCA": 1245, + "TAGTCT": 1246, + "TAGTCC": 1247, + "TAGTCG": 1248, + "TAGTGA": 1249, + "TAGTGT": 1250, + "TAGTGC": 1251, + "TAGTGG": 1252, + "TAGCAA": 1253, + "TAGCAT": 1254, + "TAGCAC": 1255, + "TAGCAG": 1256, + "TAGCTA": 1257, + "TAGCTT": 1258, + "TAGCTC": 1259, + "TAGCTG": 1260, + "TAGCCA": 1261, + "TAGCCT": 1262, + "TAGCCC": 1263, + "TAGCCG": 1264, + "TAGCGA": 1265, + "TAGCGT": 1266, + "TAGCGC": 1267, + "TAGCGG": 1268, + "TAGGAA": 1269, + "TAGGAT": 1270, + "TAGGAC": 1271, + "TAGGAG": 1272, + "TAGGTA": 1273, + "TAGGTT": 1274, + "TAGGTC": 1275, + "TAGGTG": 1276, + "TAGGCA": 1277, + "TAGGCT": 1278, + "TAGGCC": 1279, + "TAGGCG": 1280, + "TAGGGA": 1281, + "TAGGGT": 1282, + "TAGGGC": 1283, + "TAGGGG": 1284, + "TTAAAA": 1285, + "TTAAAT": 1286, + "TTAAAC": 1287, + "TTAAAG": 1288, + "TTAATA": 1289, + "TTAATT": 1290, + "TTAATC": 1291, + "TTAATG": 1292, + "TTAACA": 1293, + "TTAACT": 1294, + "TTAACC": 1295, + "TTAACG": 1296, + "TTAAGA": 1297, + "TTAAGT": 1298, + "TTAAGC": 1299, + "TTAAGG": 1300, + "TTATAA": 1301, + "TTATAT": 1302, + "TTATAC": 1303, + "TTATAG": 1304, + "TTATTA": 1305, + "TTATTT": 1306, + "TTATTC": 1307, + "TTATTG": 1308, + "TTATCA": 1309, + "TTATCT": 1310, + "TTATCC": 1311, + "TTATCG": 1312, + "TTATGA": 1313, + "TTATGT": 1314, + "TTATGC": 1315, + "TTATGG": 1316, + "TTACAA": 1317, + "TTACAT": 1318, + "TTACAC": 1319, + "TTACAG": 1320, + "TTACTA": 1321, + "TTACTT": 1322, + "TTACTC": 1323, + "TTACTG": 1324, + "TTACCA": 1325, + "TTACCT": 1326, + "TTACCC": 1327, + "TTACCG": 1328, + "TTACGA": 1329, + "TTACGT": 1330, + "TTACGC": 1331, + "TTACGG": 1332, + "TTAGAA": 1333, + "TTAGAT": 1334, + "TTAGAC": 1335, + "TTAGAG": 1336, + "TTAGTA": 1337, + "TTAGTT": 1338, + "TTAGTC": 1339, + "TTAGTG": 1340, + "TTAGCA": 1341, + "TTAGCT": 1342, + "TTAGCC": 1343, + "TTAGCG": 1344, + "TTAGGA": 1345, + "TTAGGT": 1346, + "TTAGGC": 1347, + "TTAGGG": 1348, + "TTTAAA": 1349, + "TTTAAT": 1350, + "TTTAAC": 1351, + "TTTAAG": 1352, + "TTTATA": 1353, + "TTTATT": 1354, + "TTTATC": 1355, + "TTTATG": 1356, + "TTTACA": 1357, + "TTTACT": 1358, + "TTTACC": 1359, + "TTTACG": 1360, + "TTTAGA": 1361, + "TTTAGT": 1362, + "TTTAGC": 1363, + "TTTAGG": 1364, + "TTTTAA": 1365, + "TTTTAT": 1366, + "TTTTAC": 1367, + "TTTTAG": 1368, + "TTTTTA": 1369, + "TTTTTT": 1370, + "TTTTTC": 1371, + "TTTTTG": 1372, + "TTTTCA": 1373, + "TTTTCT": 1374, + "TTTTCC": 1375, + "TTTTCG": 1376, + "TTTTGA": 1377, + "TTTTGT": 1378, + "TTTTGC": 1379, + "TTTTGG": 1380, + "TTTCAA": 1381, + "TTTCAT": 1382, + "TTTCAC": 1383, + "TTTCAG": 1384, + "TTTCTA": 1385, + "TTTCTT": 1386, + "TTTCTC": 1387, + "TTTCTG": 1388, + "TTTCCA": 1389, + "TTTCCT": 1390, + "TTTCCC": 1391, + "TTTCCG": 1392, + "TTTCGA": 1393, + "TTTCGT": 1394, + "TTTCGC": 1395, + "TTTCGG": 1396, + "TTTGAA": 1397, + "TTTGAT": 1398, + "TTTGAC": 1399, + "TTTGAG": 1400, + "TTTGTA": 1401, + "TTTGTT": 1402, + "TTTGTC": 1403, + "TTTGTG": 1404, + "TTTGCA": 1405, + "TTTGCT": 1406, + "TTTGCC": 1407, + "TTTGCG": 1408, + "TTTGGA": 1409, + "TTTGGT": 1410, + "TTTGGC": 1411, + "TTTGGG": 1412, + "TTCAAA": 1413, + "TTCAAT": 1414, + "TTCAAC": 1415, + "TTCAAG": 1416, + "TTCATA": 1417, + "TTCATT": 1418, + "TTCATC": 1419, + "TTCATG": 1420, + "TTCACA": 1421, + "TTCACT": 1422, + "TTCACC": 1423, + "TTCACG": 1424, + "TTCAGA": 1425, + "TTCAGT": 1426, + "TTCAGC": 1427, + "TTCAGG": 1428, + "TTCTAA": 1429, + "TTCTAT": 1430, + "TTCTAC": 1431, + "TTCTAG": 1432, + "TTCTTA": 1433, + "TTCTTT": 1434, + "TTCTTC": 1435, + "TTCTTG": 1436, + "TTCTCA": 1437, + "TTCTCT": 1438, + "TTCTCC": 1439, + "TTCTCG": 1440, + "TTCTGA": 1441, + "TTCTGT": 1442, + "TTCTGC": 1443, + "TTCTGG": 1444, + "TTCCAA": 1445, + "TTCCAT": 1446, + "TTCCAC": 1447, + "TTCCAG": 1448, + "TTCCTA": 1449, + "TTCCTT": 1450, + "TTCCTC": 1451, + "TTCCTG": 1452, + "TTCCCA": 1453, + "TTCCCT": 1454, + "TTCCCC": 1455, + "TTCCCG": 1456, + "TTCCGA": 1457, + "TTCCGT": 1458, + "TTCCGC": 1459, + "TTCCGG": 1460, + "TTCGAA": 1461, + "TTCGAT": 1462, + "TTCGAC": 1463, + "TTCGAG": 1464, + "TTCGTA": 1465, + "TTCGTT": 1466, + "TTCGTC": 1467, + "TTCGTG": 1468, + "TTCGCA": 1469, + "TTCGCT": 1470, + "TTCGCC": 1471, + "TTCGCG": 1472, + "TTCGGA": 1473, + "TTCGGT": 1474, + "TTCGGC": 1475, + "TTCGGG": 1476, + "TTGAAA": 1477, + "TTGAAT": 1478, + "TTGAAC": 1479, + "TTGAAG": 1480, + "TTGATA": 1481, + "TTGATT": 1482, + "TTGATC": 1483, + "TTGATG": 1484, + "TTGACA": 1485, + "TTGACT": 1486, + "TTGACC": 1487, + "TTGACG": 1488, + "TTGAGA": 1489, + "TTGAGT": 1490, + "TTGAGC": 1491, + "TTGAGG": 1492, + "TTGTAA": 1493, + "TTGTAT": 1494, + "TTGTAC": 1495, + "TTGTAG": 1496, + "TTGTTA": 1497, + "TTGTTT": 1498, + "TTGTTC": 1499, + "TTGTTG": 1500, + "TTGTCA": 1501, + "TTGTCT": 1502, + "TTGTCC": 1503, + "TTGTCG": 1504, + "TTGTGA": 1505, + "TTGTGT": 1506, + "TTGTGC": 1507, + "TTGTGG": 1508, + "TTGCAA": 1509, + "TTGCAT": 1510, + "TTGCAC": 1511, + "TTGCAG": 1512, + "TTGCTA": 1513, + "TTGCTT": 1514, + "TTGCTC": 1515, + "TTGCTG": 1516, + "TTGCCA": 1517, + "TTGCCT": 1518, + "TTGCCC": 1519, + "TTGCCG": 1520, + "TTGCGA": 1521, + "TTGCGT": 1522, + "TTGCGC": 1523, + "TTGCGG": 1524, + "TTGGAA": 1525, + "TTGGAT": 1526, + "TTGGAC": 1527, + "TTGGAG": 1528, + "TTGGTA": 1529, + "TTGGTT": 1530, + "TTGGTC": 1531, + "TTGGTG": 1532, + "TTGGCA": 1533, + "TTGGCT": 1534, + "TTGGCC": 1535, + "TTGGCG": 1536, + "TTGGGA": 1537, + "TTGGGT": 1538, + "TTGGGC": 1539, + "TTGGGG": 1540, + "TCAAAA": 1541, + "TCAAAT": 1542, + "TCAAAC": 1543, + "TCAAAG": 1544, + "TCAATA": 1545, + "TCAATT": 1546, + "TCAATC": 1547, + "TCAATG": 1548, + "TCAACA": 1549, + "TCAACT": 1550, + "TCAACC": 1551, + "TCAACG": 1552, + "TCAAGA": 1553, + "TCAAGT": 1554, + "TCAAGC": 1555, + "TCAAGG": 1556, + "TCATAA": 1557, + "TCATAT": 1558, + "TCATAC": 1559, + "TCATAG": 1560, + "TCATTA": 1561, + "TCATTT": 1562, + "TCATTC": 1563, + "TCATTG": 1564, + "TCATCA": 1565, + "TCATCT": 1566, + "TCATCC": 1567, + "TCATCG": 1568, + "TCATGA": 1569, + "TCATGT": 1570, + "TCATGC": 1571, + "TCATGG": 1572, + "TCACAA": 1573, + "TCACAT": 1574, + "TCACAC": 1575, + "TCACAG": 1576, + "TCACTA": 1577, + "TCACTT": 1578, + "TCACTC": 1579, + "TCACTG": 1580, + "TCACCA": 1581, + "TCACCT": 1582, + "TCACCC": 1583, + "TCACCG": 1584, + "TCACGA": 1585, + "TCACGT": 1586, + "TCACGC": 1587, + "TCACGG": 1588, + "TCAGAA": 1589, + "TCAGAT": 1590, + "TCAGAC": 1591, + "TCAGAG": 1592, + "TCAGTA": 1593, + "TCAGTT": 1594, + "TCAGTC": 1595, + "TCAGTG": 1596, + "TCAGCA": 1597, + "TCAGCT": 1598, + "TCAGCC": 1599, + "TCAGCG": 1600, + "TCAGGA": 1601, + "TCAGGT": 1602, + "TCAGGC": 1603, + "TCAGGG": 1604, + "TCTAAA": 1605, + "TCTAAT": 1606, + "TCTAAC": 1607, + "TCTAAG": 1608, + "TCTATA": 1609, + "TCTATT": 1610, + "TCTATC": 1611, + "TCTATG": 1612, + "TCTACA": 1613, + "TCTACT": 1614, + "TCTACC": 1615, + "TCTACG": 1616, + "TCTAGA": 1617, + "TCTAGT": 1618, + "TCTAGC": 1619, + "TCTAGG": 1620, + "TCTTAA": 1621, + "TCTTAT": 1622, + "TCTTAC": 1623, + "TCTTAG": 1624, + "TCTTTA": 1625, + "TCTTTT": 1626, + "TCTTTC": 1627, + "TCTTTG": 1628, + "TCTTCA": 1629, + "TCTTCT": 1630, + "TCTTCC": 1631, + "TCTTCG": 1632, + "TCTTGA": 1633, + "TCTTGT": 1634, + "TCTTGC": 1635, + "TCTTGG": 1636, + "TCTCAA": 1637, + "TCTCAT": 1638, + "TCTCAC": 1639, + "TCTCAG": 1640, + "TCTCTA": 1641, + "TCTCTT": 1642, + "TCTCTC": 1643, + "TCTCTG": 1644, + "TCTCCA": 1645, + "TCTCCT": 1646, + "TCTCCC": 1647, + "TCTCCG": 1648, + "TCTCGA": 1649, + "TCTCGT": 1650, + "TCTCGC": 1651, + "TCTCGG": 1652, + "TCTGAA": 1653, + "TCTGAT": 1654, + "TCTGAC": 1655, + "TCTGAG": 1656, + "TCTGTA": 1657, + "TCTGTT": 1658, + "TCTGTC": 1659, + "TCTGTG": 1660, + "TCTGCA": 1661, + "TCTGCT": 1662, + "TCTGCC": 1663, + "TCTGCG": 1664, + "TCTGGA": 1665, + "TCTGGT": 1666, + "TCTGGC": 1667, + "TCTGGG": 1668, + "TCCAAA": 1669, + "TCCAAT": 1670, + "TCCAAC": 1671, + "TCCAAG": 1672, + "TCCATA": 1673, + "TCCATT": 1674, + "TCCATC": 1675, + "TCCATG": 1676, + "TCCACA": 1677, + "TCCACT": 1678, + "TCCACC": 1679, + "TCCACG": 1680, + "TCCAGA": 1681, + "TCCAGT": 1682, + "TCCAGC": 1683, + "TCCAGG": 1684, + "TCCTAA": 1685, + "TCCTAT": 1686, + "TCCTAC": 1687, + "TCCTAG": 1688, + "TCCTTA": 1689, + "TCCTTT": 1690, + "TCCTTC": 1691, + "TCCTTG": 1692, + "TCCTCA": 1693, + "TCCTCT": 1694, + "TCCTCC": 1695, + "TCCTCG": 1696, + "TCCTGA": 1697, + "TCCTGT": 1698, + "TCCTGC": 1699, + "TCCTGG": 1700, + "TCCCAA": 1701, + "TCCCAT": 1702, + "TCCCAC": 1703, + "TCCCAG": 1704, + "TCCCTA": 1705, + "TCCCTT": 1706, + "TCCCTC": 1707, + "TCCCTG": 1708, + "TCCCCA": 1709, + "TCCCCT": 1710, + "TCCCCC": 1711, + "TCCCCG": 1712, + "TCCCGA": 1713, + "TCCCGT": 1714, + "TCCCGC": 1715, + "TCCCGG": 1716, + "TCCGAA": 1717, + "TCCGAT": 1718, + "TCCGAC": 1719, + "TCCGAG": 1720, + "TCCGTA": 1721, + "TCCGTT": 1722, + "TCCGTC": 1723, + "TCCGTG": 1724, + "TCCGCA": 1725, + "TCCGCT": 1726, + "TCCGCC": 1727, + "TCCGCG": 1728, + "TCCGGA": 1729, + "TCCGGT": 1730, + "TCCGGC": 1731, + "TCCGGG": 1732, + "TCGAAA": 1733, + "TCGAAT": 1734, + "TCGAAC": 1735, + "TCGAAG": 1736, + "TCGATA": 1737, + "TCGATT": 1738, + "TCGATC": 1739, + "TCGATG": 1740, + "TCGACA": 1741, + "TCGACT": 1742, + "TCGACC": 1743, + "TCGACG": 1744, + "TCGAGA": 1745, + "TCGAGT": 1746, + "TCGAGC": 1747, + "TCGAGG": 1748, + "TCGTAA": 1749, + "TCGTAT": 1750, + "TCGTAC": 1751, + "TCGTAG": 1752, + "TCGTTA": 1753, + "TCGTTT": 1754, + "TCGTTC": 1755, + "TCGTTG": 1756, + "TCGTCA": 1757, + "TCGTCT": 1758, + "TCGTCC": 1759, + "TCGTCG": 1760, + "TCGTGA": 1761, + "TCGTGT": 1762, + "TCGTGC": 1763, + "TCGTGG": 1764, + "TCGCAA": 1765, + "TCGCAT": 1766, + "TCGCAC": 1767, + "TCGCAG": 1768, + "TCGCTA": 1769, + "TCGCTT": 1770, + "TCGCTC": 1771, + "TCGCTG": 1772, + "TCGCCA": 1773, + "TCGCCT": 1774, + "TCGCCC": 1775, + "TCGCCG": 1776, + "TCGCGA": 1777, + "TCGCGT": 1778, + "TCGCGC": 1779, + "TCGCGG": 1780, + "TCGGAA": 1781, + "TCGGAT": 1782, + "TCGGAC": 1783, + "TCGGAG": 1784, + "TCGGTA": 1785, + "TCGGTT": 1786, + "TCGGTC": 1787, + "TCGGTG": 1788, + "TCGGCA": 1789, + "TCGGCT": 1790, + "TCGGCC": 1791, + "TCGGCG": 1792, + "TCGGGA": 1793, + "TCGGGT": 1794, + "TCGGGC": 1795, + "TCGGGG": 1796, + "TGAAAA": 1797, + "TGAAAT": 1798, + "TGAAAC": 1799, + "TGAAAG": 1800, + "TGAATA": 1801, + "TGAATT": 1802, + "TGAATC": 1803, + "TGAATG": 1804, + "TGAACA": 1805, + "TGAACT": 1806, + "TGAACC": 1807, + "TGAACG": 1808, + "TGAAGA": 1809, + "TGAAGT": 1810, + "TGAAGC": 1811, + "TGAAGG": 1812, + "TGATAA": 1813, + "TGATAT": 1814, + "TGATAC": 1815, + "TGATAG": 1816, + "TGATTA": 1817, + "TGATTT": 1818, + "TGATTC": 1819, + "TGATTG": 1820, + "TGATCA": 1821, + "TGATCT": 1822, + "TGATCC": 1823, + "TGATCG": 1824, + "TGATGA": 1825, + "TGATGT": 1826, + "TGATGC": 1827, + "TGATGG": 1828, + "TGACAA": 1829, + "TGACAT": 1830, + "TGACAC": 1831, + "TGACAG": 1832, + "TGACTA": 1833, + "TGACTT": 1834, + "TGACTC": 1835, + "TGACTG": 1836, + "TGACCA": 1837, + "TGACCT": 1838, + "TGACCC": 1839, + "TGACCG": 1840, + "TGACGA": 1841, + "TGACGT": 1842, + "TGACGC": 1843, + "TGACGG": 1844, + "TGAGAA": 1845, + "TGAGAT": 1846, + "TGAGAC": 1847, + "TGAGAG": 1848, + "TGAGTA": 1849, + "TGAGTT": 1850, + "TGAGTC": 1851, + "TGAGTG": 1852, + "TGAGCA": 1853, + "TGAGCT": 1854, + "TGAGCC": 1855, + "TGAGCG": 1856, + "TGAGGA": 1857, + "TGAGGT": 1858, + "TGAGGC": 1859, + "TGAGGG": 1860, + "TGTAAA": 1861, + "TGTAAT": 1862, + "TGTAAC": 1863, + "TGTAAG": 1864, + "TGTATA": 1865, + "TGTATT": 1866, + "TGTATC": 1867, + "TGTATG": 1868, + "TGTACA": 1869, + "TGTACT": 1870, + "TGTACC": 1871, + "TGTACG": 1872, + "TGTAGA": 1873, + "TGTAGT": 1874, + "TGTAGC": 1875, + "TGTAGG": 1876, + "TGTTAA": 1877, + "TGTTAT": 1878, + "TGTTAC": 1879, + "TGTTAG": 1880, + "TGTTTA": 1881, + "TGTTTT": 1882, + "TGTTTC": 1883, + "TGTTTG": 1884, + "TGTTCA": 1885, + "TGTTCT": 1886, + "TGTTCC": 1887, + "TGTTCG": 1888, + "TGTTGA": 1889, + "TGTTGT": 1890, + "TGTTGC": 1891, + "TGTTGG": 1892, + "TGTCAA": 1893, + "TGTCAT": 1894, + "TGTCAC": 1895, + "TGTCAG": 1896, + "TGTCTA": 1897, + "TGTCTT": 1898, + "TGTCTC": 1899, + "TGTCTG": 1900, + "TGTCCA": 1901, + "TGTCCT": 1902, + "TGTCCC": 1903, + "TGTCCG": 1904, + "TGTCGA": 1905, + "TGTCGT": 1906, + "TGTCGC": 1907, + "TGTCGG": 1908, + "TGTGAA": 1909, + "TGTGAT": 1910, + "TGTGAC": 1911, + "TGTGAG": 1912, + "TGTGTA": 1913, + "TGTGTT": 1914, + "TGTGTC": 1915, + "TGTGTG": 1916, + "TGTGCA": 1917, + "TGTGCT": 1918, + "TGTGCC": 1919, + "TGTGCG": 1920, + "TGTGGA": 1921, + "TGTGGT": 1922, + "TGTGGC": 1923, + "TGTGGG": 1924, + "TGCAAA": 1925, + "TGCAAT": 1926, + "TGCAAC": 1927, + "TGCAAG": 1928, + "TGCATA": 1929, + "TGCATT": 1930, + "TGCATC": 1931, + "TGCATG": 1932, + "TGCACA": 1933, + "TGCACT": 1934, + "TGCACC": 1935, + "TGCACG": 1936, + "TGCAGA": 1937, + "TGCAGT": 1938, + "TGCAGC": 1939, + "TGCAGG": 1940, + "TGCTAA": 1941, + "TGCTAT": 1942, + "TGCTAC": 1943, + "TGCTAG": 1944, + "TGCTTA": 1945, + "TGCTTT": 1946, + "TGCTTC": 1947, + "TGCTTG": 1948, + "TGCTCA": 1949, + "TGCTCT": 1950, + "TGCTCC": 1951, + "TGCTCG": 1952, + "TGCTGA": 1953, + "TGCTGT": 1954, + "TGCTGC": 1955, + "TGCTGG": 1956, + "TGCCAA": 1957, + "TGCCAT": 1958, + "TGCCAC": 1959, + "TGCCAG": 1960, + "TGCCTA": 1961, + "TGCCTT": 1962, + "TGCCTC": 1963, + "TGCCTG": 1964, + "TGCCCA": 1965, + "TGCCCT": 1966, + "TGCCCC": 1967, + "TGCCCG": 1968, + "TGCCGA": 1969, + "TGCCGT": 1970, + "TGCCGC": 1971, + "TGCCGG": 1972, + "TGCGAA": 1973, + "TGCGAT": 1974, + "TGCGAC": 1975, + "TGCGAG": 1976, + "TGCGTA": 1977, + "TGCGTT": 1978, + "TGCGTC": 1979, + "TGCGTG": 1980, + "TGCGCA": 1981, + "TGCGCT": 1982, + "TGCGCC": 1983, + "TGCGCG": 1984, + "TGCGGA": 1985, + "TGCGGT": 1986, + "TGCGGC": 1987, + "TGCGGG": 1988, + "TGGAAA": 1989, + "TGGAAT": 1990, + "TGGAAC": 1991, + "TGGAAG": 1992, + "TGGATA": 1993, + "TGGATT": 1994, + "TGGATC": 1995, + "TGGATG": 1996, + "TGGACA": 1997, + "TGGACT": 1998, + "TGGACC": 1999, + "TGGACG": 2000, + "TGGAGA": 2001, + "TGGAGT": 2002, + "TGGAGC": 2003, + "TGGAGG": 2004, + "TGGTAA": 2005, + "TGGTAT": 2006, + "TGGTAC": 2007, + "TGGTAG": 2008, + "TGGTTA": 2009, + "TGGTTT": 2010, + "TGGTTC": 2011, + "TGGTTG": 2012, + "TGGTCA": 2013, + "TGGTCT": 2014, + "TGGTCC": 2015, + "TGGTCG": 2016, + "TGGTGA": 2017, + "TGGTGT": 2018, + "TGGTGC": 2019, + "TGGTGG": 2020, + "TGGCAA": 2021, + "TGGCAT": 2022, + "TGGCAC": 2023, + "TGGCAG": 2024, + "TGGCTA": 2025, + "TGGCTT": 2026, + "TGGCTC": 2027, + "TGGCTG": 2028, + "TGGCCA": 2029, + "TGGCCT": 2030, + "TGGCCC": 2031, + "TGGCCG": 2032, + "TGGCGA": 2033, + "TGGCGT": 2034, + "TGGCGC": 2035, + "TGGCGG": 2036, + "TGGGAA": 2037, + "TGGGAT": 2038, + "TGGGAC": 2039, + "TGGGAG": 2040, + "TGGGTA": 2041, + "TGGGTT": 2042, + "TGGGTC": 2043, + "TGGGTG": 2044, + "TGGGCA": 2045, + "TGGGCT": 2046, + "TGGGCC": 2047, + "TGGGCG": 2048, + "TGGGGA": 2049, + "TGGGGT": 2050, + "TGGGGC": 2051, + "TGGGGG": 2052, + "CAAAAA": 2053, + "CAAAAT": 2054, + "CAAAAC": 2055, + "CAAAAG": 2056, + "CAAATA": 2057, + "CAAATT": 2058, + "CAAATC": 2059, + "CAAATG": 2060, + "CAAACA": 2061, + "CAAACT": 2062, + "CAAACC": 2063, + "CAAACG": 2064, + "CAAAGA": 2065, + "CAAAGT": 2066, + "CAAAGC": 2067, + "CAAAGG": 2068, + "CAATAA": 2069, + "CAATAT": 2070, + "CAATAC": 2071, + "CAATAG": 2072, + "CAATTA": 2073, + "CAATTT": 2074, + "CAATTC": 2075, + "CAATTG": 2076, + "CAATCA": 2077, + "CAATCT": 2078, + "CAATCC": 2079, + "CAATCG": 2080, + "CAATGA": 2081, + "CAATGT": 2082, + "CAATGC": 2083, + "CAATGG": 2084, + "CAACAA": 2085, + "CAACAT": 2086, + "CAACAC": 2087, + "CAACAG": 2088, + "CAACTA": 2089, + "CAACTT": 2090, + "CAACTC": 2091, + "CAACTG": 2092, + "CAACCA": 2093, + "CAACCT": 2094, + "CAACCC": 2095, + "CAACCG": 2096, + "CAACGA": 2097, + "CAACGT": 2098, + "CAACGC": 2099, + "CAACGG": 2100, + "CAAGAA": 2101, + "CAAGAT": 2102, + "CAAGAC": 2103, + "CAAGAG": 2104, + "CAAGTA": 2105, + "CAAGTT": 2106, + "CAAGTC": 2107, + "CAAGTG": 2108, + "CAAGCA": 2109, + "CAAGCT": 2110, + "CAAGCC": 2111, + "CAAGCG": 2112, + "CAAGGA": 2113, + "CAAGGT": 2114, + "CAAGGC": 2115, + "CAAGGG": 2116, + "CATAAA": 2117, + "CATAAT": 2118, + "CATAAC": 2119, + "CATAAG": 2120, + "CATATA": 2121, + "CATATT": 2122, + "CATATC": 2123, + "CATATG": 2124, + "CATACA": 2125, + "CATACT": 2126, + "CATACC": 2127, + "CATACG": 2128, + "CATAGA": 2129, + "CATAGT": 2130, + "CATAGC": 2131, + "CATAGG": 2132, + "CATTAA": 2133, + "CATTAT": 2134, + "CATTAC": 2135, + "CATTAG": 2136, + "CATTTA": 2137, + "CATTTT": 2138, + "CATTTC": 2139, + "CATTTG": 2140, + "CATTCA": 2141, + "CATTCT": 2142, + "CATTCC": 2143, + "CATTCG": 2144, + "CATTGA": 2145, + "CATTGT": 2146, + "CATTGC": 2147, + "CATTGG": 2148, + "CATCAA": 2149, + "CATCAT": 2150, + "CATCAC": 2151, + "CATCAG": 2152, + "CATCTA": 2153, + "CATCTT": 2154, + "CATCTC": 2155, + "CATCTG": 2156, + "CATCCA": 2157, + "CATCCT": 2158, + "CATCCC": 2159, + "CATCCG": 2160, + "CATCGA": 2161, + "CATCGT": 2162, + "CATCGC": 2163, + "CATCGG": 2164, + "CATGAA": 2165, + "CATGAT": 2166, + "CATGAC": 2167, + "CATGAG": 2168, + "CATGTA": 2169, + "CATGTT": 2170, + "CATGTC": 2171, + "CATGTG": 2172, + "CATGCA": 2173, + "CATGCT": 2174, + "CATGCC": 2175, + "CATGCG": 2176, + "CATGGA": 2177, + "CATGGT": 2178, + "CATGGC": 2179, + "CATGGG": 2180, + "CACAAA": 2181, + "CACAAT": 2182, + "CACAAC": 2183, + "CACAAG": 2184, + "CACATA": 2185, + "CACATT": 2186, + "CACATC": 2187, + "CACATG": 2188, + "CACACA": 2189, + "CACACT": 2190, + "CACACC": 2191, + "CACACG": 2192, + "CACAGA": 2193, + "CACAGT": 2194, + "CACAGC": 2195, + "CACAGG": 2196, + "CACTAA": 2197, + "CACTAT": 2198, + "CACTAC": 2199, + "CACTAG": 2200, + "CACTTA": 2201, + "CACTTT": 2202, + "CACTTC": 2203, + "CACTTG": 2204, + "CACTCA": 2205, + "CACTCT": 2206, + "CACTCC": 2207, + "CACTCG": 2208, + "CACTGA": 2209, + "CACTGT": 2210, + "CACTGC": 2211, + "CACTGG": 2212, + "CACCAA": 2213, + "CACCAT": 2214, + "CACCAC": 2215, + "CACCAG": 2216, + "CACCTA": 2217, + "CACCTT": 2218, + "CACCTC": 2219, + "CACCTG": 2220, + "CACCCA": 2221, + "CACCCT": 2222, + "CACCCC": 2223, + "CACCCG": 2224, + "CACCGA": 2225, + "CACCGT": 2226, + "CACCGC": 2227, + "CACCGG": 2228, + "CACGAA": 2229, + "CACGAT": 2230, + "CACGAC": 2231, + "CACGAG": 2232, + "CACGTA": 2233, + "CACGTT": 2234, + "CACGTC": 2235, + "CACGTG": 2236, + "CACGCA": 2237, + "CACGCT": 2238, + "CACGCC": 2239, + "CACGCG": 2240, + "CACGGA": 2241, + "CACGGT": 2242, + "CACGGC": 2243, + "CACGGG": 2244, + "CAGAAA": 2245, + "CAGAAT": 2246, + "CAGAAC": 2247, + "CAGAAG": 2248, + "CAGATA": 2249, + "CAGATT": 2250, + "CAGATC": 2251, + "CAGATG": 2252, + "CAGACA": 2253, + "CAGACT": 2254, + "CAGACC": 2255, + "CAGACG": 2256, + "CAGAGA": 2257, + "CAGAGT": 2258, + "CAGAGC": 2259, + "CAGAGG": 2260, + "CAGTAA": 2261, + "CAGTAT": 2262, + "CAGTAC": 2263, + "CAGTAG": 2264, + "CAGTTA": 2265, + "CAGTTT": 2266, + "CAGTTC": 2267, + "CAGTTG": 2268, + "CAGTCA": 2269, + "CAGTCT": 2270, + "CAGTCC": 2271, + "CAGTCG": 2272, + "CAGTGA": 2273, + "CAGTGT": 2274, + "CAGTGC": 2275, + "CAGTGG": 2276, + "CAGCAA": 2277, + "CAGCAT": 2278, + "CAGCAC": 2279, + "CAGCAG": 2280, + "CAGCTA": 2281, + "CAGCTT": 2282, + "CAGCTC": 2283, + "CAGCTG": 2284, + "CAGCCA": 2285, + "CAGCCT": 2286, + "CAGCCC": 2287, + "CAGCCG": 2288, + "CAGCGA": 2289, + "CAGCGT": 2290, + "CAGCGC": 2291, + "CAGCGG": 2292, + "CAGGAA": 2293, + "CAGGAT": 2294, + "CAGGAC": 2295, + "CAGGAG": 2296, + "CAGGTA": 2297, + "CAGGTT": 2298, + "CAGGTC": 2299, + "CAGGTG": 2300, + "CAGGCA": 2301, + "CAGGCT": 2302, + "CAGGCC": 2303, + "CAGGCG": 2304, + "CAGGGA": 2305, + "CAGGGT": 2306, + "CAGGGC": 2307, + "CAGGGG": 2308, + "CTAAAA": 2309, + "CTAAAT": 2310, + "CTAAAC": 2311, + "CTAAAG": 2312, + "CTAATA": 2313, + "CTAATT": 2314, + "CTAATC": 2315, + "CTAATG": 2316, + "CTAACA": 2317, + "CTAACT": 2318, + "CTAACC": 2319, + "CTAACG": 2320, + "CTAAGA": 2321, + "CTAAGT": 2322, + "CTAAGC": 2323, + "CTAAGG": 2324, + "CTATAA": 2325, + "CTATAT": 2326, + "CTATAC": 2327, + "CTATAG": 2328, + "CTATTA": 2329, + "CTATTT": 2330, + "CTATTC": 2331, + "CTATTG": 2332, + "CTATCA": 2333, + "CTATCT": 2334, + "CTATCC": 2335, + "CTATCG": 2336, + "CTATGA": 2337, + "CTATGT": 2338, + "CTATGC": 2339, + "CTATGG": 2340, + "CTACAA": 2341, + "CTACAT": 2342, + "CTACAC": 2343, + "CTACAG": 2344, + "CTACTA": 2345, + "CTACTT": 2346, + "CTACTC": 2347, + "CTACTG": 2348, + "CTACCA": 2349, + "CTACCT": 2350, + "CTACCC": 2351, + "CTACCG": 2352, + "CTACGA": 2353, + "CTACGT": 2354, + "CTACGC": 2355, + "CTACGG": 2356, + "CTAGAA": 2357, + "CTAGAT": 2358, + "CTAGAC": 2359, + "CTAGAG": 2360, + "CTAGTA": 2361, + "CTAGTT": 2362, + "CTAGTC": 2363, + "CTAGTG": 2364, + "CTAGCA": 2365, + "CTAGCT": 2366, + "CTAGCC": 2367, + "CTAGCG": 2368, + "CTAGGA": 2369, + "CTAGGT": 2370, + "CTAGGC": 2371, + "CTAGGG": 2372, + "CTTAAA": 2373, + "CTTAAT": 2374, + "CTTAAC": 2375, + "CTTAAG": 2376, + "CTTATA": 2377, + "CTTATT": 2378, + "CTTATC": 2379, + "CTTATG": 2380, + "CTTACA": 2381, + "CTTACT": 2382, + "CTTACC": 2383, + "CTTACG": 2384, + "CTTAGA": 2385, + "CTTAGT": 2386, + "CTTAGC": 2387, + "CTTAGG": 2388, + "CTTTAA": 2389, + "CTTTAT": 2390, + "CTTTAC": 2391, + "CTTTAG": 2392, + "CTTTTA": 2393, + "CTTTTT": 2394, + "CTTTTC": 2395, + "CTTTTG": 2396, + "CTTTCA": 2397, + "CTTTCT": 2398, + "CTTTCC": 2399, + "CTTTCG": 2400, + "CTTTGA": 2401, + "CTTTGT": 2402, + "CTTTGC": 2403, + "CTTTGG": 2404, + "CTTCAA": 2405, + "CTTCAT": 2406, + "CTTCAC": 2407, + "CTTCAG": 2408, + "CTTCTA": 2409, + "CTTCTT": 2410, + "CTTCTC": 2411, + "CTTCTG": 2412, + "CTTCCA": 2413, + "CTTCCT": 2414, + "CTTCCC": 2415, + "CTTCCG": 2416, + "CTTCGA": 2417, + "CTTCGT": 2418, + "CTTCGC": 2419, + "CTTCGG": 2420, + "CTTGAA": 2421, + "CTTGAT": 2422, + "CTTGAC": 2423, + "CTTGAG": 2424, + "CTTGTA": 2425, + "CTTGTT": 2426, + "CTTGTC": 2427, + "CTTGTG": 2428, + "CTTGCA": 2429, + "CTTGCT": 2430, + "CTTGCC": 2431, + "CTTGCG": 2432, + "CTTGGA": 2433, + "CTTGGT": 2434, + "CTTGGC": 2435, + "CTTGGG": 2436, + "CTCAAA": 2437, + "CTCAAT": 2438, + "CTCAAC": 2439, + "CTCAAG": 2440, + "CTCATA": 2441, + "CTCATT": 2442, + "CTCATC": 2443, + "CTCATG": 2444, + "CTCACA": 2445, + "CTCACT": 2446, + "CTCACC": 2447, + "CTCACG": 2448, + "CTCAGA": 2449, + "CTCAGT": 2450, + "CTCAGC": 2451, + "CTCAGG": 2452, + "CTCTAA": 2453, + "CTCTAT": 2454, + "CTCTAC": 2455, + "CTCTAG": 2456, + "CTCTTA": 2457, + "CTCTTT": 2458, + "CTCTTC": 2459, + "CTCTTG": 2460, + "CTCTCA": 2461, + "CTCTCT": 2462, + "CTCTCC": 2463, + "CTCTCG": 2464, + "CTCTGA": 2465, + "CTCTGT": 2466, + "CTCTGC": 2467, + "CTCTGG": 2468, + "CTCCAA": 2469, + "CTCCAT": 2470, + "CTCCAC": 2471, + "CTCCAG": 2472, + "CTCCTA": 2473, + "CTCCTT": 2474, + "CTCCTC": 2475, + "CTCCTG": 2476, + "CTCCCA": 2477, + "CTCCCT": 2478, + "CTCCCC": 2479, + "CTCCCG": 2480, + "CTCCGA": 2481, + "CTCCGT": 2482, + "CTCCGC": 2483, + "CTCCGG": 2484, + "CTCGAA": 2485, + "CTCGAT": 2486, + "CTCGAC": 2487, + "CTCGAG": 2488, + "CTCGTA": 2489, + "CTCGTT": 2490, + "CTCGTC": 2491, + "CTCGTG": 2492, + "CTCGCA": 2493, + "CTCGCT": 2494, + "CTCGCC": 2495, + "CTCGCG": 2496, + "CTCGGA": 2497, + "CTCGGT": 2498, + "CTCGGC": 2499, + "CTCGGG": 2500, + "CTGAAA": 2501, + "CTGAAT": 2502, + "CTGAAC": 2503, + "CTGAAG": 2504, + "CTGATA": 2505, + "CTGATT": 2506, + "CTGATC": 2507, + "CTGATG": 2508, + "CTGACA": 2509, + "CTGACT": 2510, + "CTGACC": 2511, + "CTGACG": 2512, + "CTGAGA": 2513, + "CTGAGT": 2514, + "CTGAGC": 2515, + "CTGAGG": 2516, + "CTGTAA": 2517, + "CTGTAT": 2518, + "CTGTAC": 2519, + "CTGTAG": 2520, + "CTGTTA": 2521, + "CTGTTT": 2522, + "CTGTTC": 2523, + "CTGTTG": 2524, + "CTGTCA": 2525, + "CTGTCT": 2526, + "CTGTCC": 2527, + "CTGTCG": 2528, + "CTGTGA": 2529, + "CTGTGT": 2530, + "CTGTGC": 2531, + "CTGTGG": 2532, + "CTGCAA": 2533, + "CTGCAT": 2534, + "CTGCAC": 2535, + "CTGCAG": 2536, + "CTGCTA": 2537, + "CTGCTT": 2538, + "CTGCTC": 2539, + "CTGCTG": 2540, + "CTGCCA": 2541, + "CTGCCT": 2542, + "CTGCCC": 2543, + "CTGCCG": 2544, + "CTGCGA": 2545, + "CTGCGT": 2546, + "CTGCGC": 2547, + "CTGCGG": 2548, + "CTGGAA": 2549, + "CTGGAT": 2550, + "CTGGAC": 2551, + "CTGGAG": 2552, + "CTGGTA": 2553, + "CTGGTT": 2554, + "CTGGTC": 2555, + "CTGGTG": 2556, + "CTGGCA": 2557, + "CTGGCT": 2558, + "CTGGCC": 2559, + "CTGGCG": 2560, + "CTGGGA": 2561, + "CTGGGT": 2562, + "CTGGGC": 2563, + "CTGGGG": 2564, + "CCAAAA": 2565, + "CCAAAT": 2566, + "CCAAAC": 2567, + "CCAAAG": 2568, + "CCAATA": 2569, + "CCAATT": 2570, + "CCAATC": 2571, + "CCAATG": 2572, + "CCAACA": 2573, + "CCAACT": 2574, + "CCAACC": 2575, + "CCAACG": 2576, + "CCAAGA": 2577, + "CCAAGT": 2578, + "CCAAGC": 2579, + "CCAAGG": 2580, + "CCATAA": 2581, + "CCATAT": 2582, + "CCATAC": 2583, + "CCATAG": 2584, + "CCATTA": 2585, + "CCATTT": 2586, + "CCATTC": 2587, + "CCATTG": 2588, + "CCATCA": 2589, + "CCATCT": 2590, + "CCATCC": 2591, + "CCATCG": 2592, + "CCATGA": 2593, + "CCATGT": 2594, + "CCATGC": 2595, + "CCATGG": 2596, + "CCACAA": 2597, + "CCACAT": 2598, + "CCACAC": 2599, + "CCACAG": 2600, + "CCACTA": 2601, + "CCACTT": 2602, + "CCACTC": 2603, + "CCACTG": 2604, + "CCACCA": 2605, + "CCACCT": 2606, + "CCACCC": 2607, + "CCACCG": 2608, + "CCACGA": 2609, + "CCACGT": 2610, + "CCACGC": 2611, + "CCACGG": 2612, + "CCAGAA": 2613, + "CCAGAT": 2614, + "CCAGAC": 2615, + "CCAGAG": 2616, + "CCAGTA": 2617, + "CCAGTT": 2618, + "CCAGTC": 2619, + "CCAGTG": 2620, + "CCAGCA": 2621, + "CCAGCT": 2622, + "CCAGCC": 2623, + "CCAGCG": 2624, + "CCAGGA": 2625, + "CCAGGT": 2626, + "CCAGGC": 2627, + "CCAGGG": 2628, + "CCTAAA": 2629, + "CCTAAT": 2630, + "CCTAAC": 2631, + "CCTAAG": 2632, + "CCTATA": 2633, + "CCTATT": 2634, + "CCTATC": 2635, + "CCTATG": 2636, + "CCTACA": 2637, + "CCTACT": 2638, + "CCTACC": 2639, + "CCTACG": 2640, + "CCTAGA": 2641, + "CCTAGT": 2642, + "CCTAGC": 2643, + "CCTAGG": 2644, + "CCTTAA": 2645, + "CCTTAT": 2646, + "CCTTAC": 2647, + "CCTTAG": 2648, + "CCTTTA": 2649, + "CCTTTT": 2650, + "CCTTTC": 2651, + "CCTTTG": 2652, + "CCTTCA": 2653, + "CCTTCT": 2654, + "CCTTCC": 2655, + "CCTTCG": 2656, + "CCTTGA": 2657, + "CCTTGT": 2658, + "CCTTGC": 2659, + "CCTTGG": 2660, + "CCTCAA": 2661, + "CCTCAT": 2662, + "CCTCAC": 2663, + "CCTCAG": 2664, + "CCTCTA": 2665, + "CCTCTT": 2666, + "CCTCTC": 2667, + "CCTCTG": 2668, + "CCTCCA": 2669, + "CCTCCT": 2670, + "CCTCCC": 2671, + "CCTCCG": 2672, + "CCTCGA": 2673, + "CCTCGT": 2674, + "CCTCGC": 2675, + "CCTCGG": 2676, + "CCTGAA": 2677, + "CCTGAT": 2678, + "CCTGAC": 2679, + "CCTGAG": 2680, + "CCTGTA": 2681, + "CCTGTT": 2682, + "CCTGTC": 2683, + "CCTGTG": 2684, + "CCTGCA": 2685, + "CCTGCT": 2686, + "CCTGCC": 2687, + "CCTGCG": 2688, + "CCTGGA": 2689, + "CCTGGT": 2690, + "CCTGGC": 2691, + "CCTGGG": 2692, + "CCCAAA": 2693, + "CCCAAT": 2694, + "CCCAAC": 2695, + "CCCAAG": 2696, + "CCCATA": 2697, + "CCCATT": 2698, + "CCCATC": 2699, + "CCCATG": 2700, + "CCCACA": 2701, + "CCCACT": 2702, + "CCCACC": 2703, + "CCCACG": 2704, + "CCCAGA": 2705, + "CCCAGT": 2706, + "CCCAGC": 2707, + "CCCAGG": 2708, + "CCCTAA": 2709, + "CCCTAT": 2710, + "CCCTAC": 2711, + "CCCTAG": 2712, + "CCCTTA": 2713, + "CCCTTT": 2714, + "CCCTTC": 2715, + "CCCTTG": 2716, + "CCCTCA": 2717, + "CCCTCT": 2718, + "CCCTCC": 2719, + "CCCTCG": 2720, + "CCCTGA": 2721, + "CCCTGT": 2722, + "CCCTGC": 2723, + "CCCTGG": 2724, + "CCCCAA": 2725, + "CCCCAT": 2726, + "CCCCAC": 2727, + "CCCCAG": 2728, + "CCCCTA": 2729, + "CCCCTT": 2730, + "CCCCTC": 2731, + "CCCCTG": 2732, + "CCCCCA": 2733, + "CCCCCT": 2734, + "CCCCCC": 2735, + "CCCCCG": 2736, + "CCCCGA": 2737, + "CCCCGT": 2738, + "CCCCGC": 2739, + "CCCCGG": 2740, + "CCCGAA": 2741, + "CCCGAT": 2742, + "CCCGAC": 2743, + "CCCGAG": 2744, + "CCCGTA": 2745, + "CCCGTT": 2746, + "CCCGTC": 2747, + "CCCGTG": 2748, + "CCCGCA": 2749, + "CCCGCT": 2750, + "CCCGCC": 2751, + "CCCGCG": 2752, + "CCCGGA": 2753, + "CCCGGT": 2754, + "CCCGGC": 2755, + "CCCGGG": 2756, + "CCGAAA": 2757, + "CCGAAT": 2758, + "CCGAAC": 2759, + "CCGAAG": 2760, + "CCGATA": 2761, + "CCGATT": 2762, + "CCGATC": 2763, + "CCGATG": 2764, + "CCGACA": 2765, + "CCGACT": 2766, + "CCGACC": 2767, + "CCGACG": 2768, + "CCGAGA": 2769, + "CCGAGT": 2770, + "CCGAGC": 2771, + "CCGAGG": 2772, + "CCGTAA": 2773, + "CCGTAT": 2774, + "CCGTAC": 2775, + "CCGTAG": 2776, + "CCGTTA": 2777, + "CCGTTT": 2778, + "CCGTTC": 2779, + "CCGTTG": 2780, + "CCGTCA": 2781, + "CCGTCT": 2782, + "CCGTCC": 2783, + "CCGTCG": 2784, + "CCGTGA": 2785, + "CCGTGT": 2786, + "CCGTGC": 2787, + "CCGTGG": 2788, + "CCGCAA": 2789, + "CCGCAT": 2790, + "CCGCAC": 2791, + "CCGCAG": 2792, + "CCGCTA": 2793, + "CCGCTT": 2794, + "CCGCTC": 2795, + "CCGCTG": 2796, + "CCGCCA": 2797, + "CCGCCT": 2798, + "CCGCCC": 2799, + "CCGCCG": 2800, + "CCGCGA": 2801, + "CCGCGT": 2802, + "CCGCGC": 2803, + "CCGCGG": 2804, + "CCGGAA": 2805, + "CCGGAT": 2806, + "CCGGAC": 2807, + "CCGGAG": 2808, + "CCGGTA": 2809, + "CCGGTT": 2810, + "CCGGTC": 2811, + "CCGGTG": 2812, + "CCGGCA": 2813, + "CCGGCT": 2814, + "CCGGCC": 2815, + "CCGGCG": 2816, + "CCGGGA": 2817, + "CCGGGT": 2818, + "CCGGGC": 2819, + "CCGGGG": 2820, + "CGAAAA": 2821, + "CGAAAT": 2822, + "CGAAAC": 2823, + "CGAAAG": 2824, + "CGAATA": 2825, + "CGAATT": 2826, + "CGAATC": 2827, + "CGAATG": 2828, + "CGAACA": 2829, + "CGAACT": 2830, + "CGAACC": 2831, + "CGAACG": 2832, + "CGAAGA": 2833, + "CGAAGT": 2834, + "CGAAGC": 2835, + "CGAAGG": 2836, + "CGATAA": 2837, + "CGATAT": 2838, + "CGATAC": 2839, + "CGATAG": 2840, + "CGATTA": 2841, + "CGATTT": 2842, + "CGATTC": 2843, + "CGATTG": 2844, + "CGATCA": 2845, + "CGATCT": 2846, + "CGATCC": 2847, + "CGATCG": 2848, + "CGATGA": 2849, + "CGATGT": 2850, + "CGATGC": 2851, + "CGATGG": 2852, + "CGACAA": 2853, + "CGACAT": 2854, + "CGACAC": 2855, + "CGACAG": 2856, + "CGACTA": 2857, + "CGACTT": 2858, + "CGACTC": 2859, + "CGACTG": 2860, + "CGACCA": 2861, + "CGACCT": 2862, + "CGACCC": 2863, + "CGACCG": 2864, + "CGACGA": 2865, + "CGACGT": 2866, + "CGACGC": 2867, + "CGACGG": 2868, + "CGAGAA": 2869, + "CGAGAT": 2870, + "CGAGAC": 2871, + "CGAGAG": 2872, + "CGAGTA": 2873, + "CGAGTT": 2874, + "CGAGTC": 2875, + "CGAGTG": 2876, + "CGAGCA": 2877, + "CGAGCT": 2878, + "CGAGCC": 2879, + "CGAGCG": 2880, + "CGAGGA": 2881, + "CGAGGT": 2882, + "CGAGGC": 2883, + "CGAGGG": 2884, + "CGTAAA": 2885, + "CGTAAT": 2886, + "CGTAAC": 2887, + "CGTAAG": 2888, + "CGTATA": 2889, + "CGTATT": 2890, + "CGTATC": 2891, + "CGTATG": 2892, + "CGTACA": 2893, + "CGTACT": 2894, + "CGTACC": 2895, + "CGTACG": 2896, + "CGTAGA": 2897, + "CGTAGT": 2898, + "CGTAGC": 2899, + "CGTAGG": 2900, + "CGTTAA": 2901, + "CGTTAT": 2902, + "CGTTAC": 2903, + "CGTTAG": 2904, + "CGTTTA": 2905, + "CGTTTT": 2906, + "CGTTTC": 2907, + "CGTTTG": 2908, + "CGTTCA": 2909, + "CGTTCT": 2910, + "CGTTCC": 2911, + "CGTTCG": 2912, + "CGTTGA": 2913, + "CGTTGT": 2914, + "CGTTGC": 2915, + "CGTTGG": 2916, + "CGTCAA": 2917, + "CGTCAT": 2918, + "CGTCAC": 2919, + "CGTCAG": 2920, + "CGTCTA": 2921, + "CGTCTT": 2922, + "CGTCTC": 2923, + "CGTCTG": 2924, + "CGTCCA": 2925, + "CGTCCT": 2926, + "CGTCCC": 2927, + "CGTCCG": 2928, + "CGTCGA": 2929, + "CGTCGT": 2930, + "CGTCGC": 2931, + "CGTCGG": 2932, + "CGTGAA": 2933, + "CGTGAT": 2934, + "CGTGAC": 2935, + "CGTGAG": 2936, + "CGTGTA": 2937, + "CGTGTT": 2938, + "CGTGTC": 2939, + "CGTGTG": 2940, + "CGTGCA": 2941, + "CGTGCT": 2942, + "CGTGCC": 2943, + "CGTGCG": 2944, + "CGTGGA": 2945, + "CGTGGT": 2946, + "CGTGGC": 2947, + "CGTGGG": 2948, + "CGCAAA": 2949, + "CGCAAT": 2950, + "CGCAAC": 2951, + "CGCAAG": 2952, + "CGCATA": 2953, + "CGCATT": 2954, + "CGCATC": 2955, + "CGCATG": 2956, + "CGCACA": 2957, + "CGCACT": 2958, + "CGCACC": 2959, + "CGCACG": 2960, + "CGCAGA": 2961, + "CGCAGT": 2962, + "CGCAGC": 2963, + "CGCAGG": 2964, + "CGCTAA": 2965, + "CGCTAT": 2966, + "CGCTAC": 2967, + "CGCTAG": 2968, + "CGCTTA": 2969, + "CGCTTT": 2970, + "CGCTTC": 2971, + "CGCTTG": 2972, + "CGCTCA": 2973, + "CGCTCT": 2974, + "CGCTCC": 2975, + "CGCTCG": 2976, + "CGCTGA": 2977, + "CGCTGT": 2978, + "CGCTGC": 2979, + "CGCTGG": 2980, + "CGCCAA": 2981, + "CGCCAT": 2982, + "CGCCAC": 2983, + "CGCCAG": 2984, + "CGCCTA": 2985, + "CGCCTT": 2986, + "CGCCTC": 2987, + "CGCCTG": 2988, + "CGCCCA": 2989, + "CGCCCT": 2990, + "CGCCCC": 2991, + "CGCCCG": 2992, + "CGCCGA": 2993, + "CGCCGT": 2994, + "CGCCGC": 2995, + "CGCCGG": 2996, + "CGCGAA": 2997, + "CGCGAT": 2998, + "CGCGAC": 2999, + "CGCGAG": 3000, + "CGCGTA": 3001, + "CGCGTT": 3002, + "CGCGTC": 3003, + "CGCGTG": 3004, + "CGCGCA": 3005, + "CGCGCT": 3006, + "CGCGCC": 3007, + "CGCGCG": 3008, + "CGCGGA": 3009, + "CGCGGT": 3010, + "CGCGGC": 3011, + "CGCGGG": 3012, + "CGGAAA": 3013, + "CGGAAT": 3014, + "CGGAAC": 3015, + "CGGAAG": 3016, + "CGGATA": 3017, + "CGGATT": 3018, + "CGGATC": 3019, + "CGGATG": 3020, + "CGGACA": 3021, + "CGGACT": 3022, + "CGGACC": 3023, + "CGGACG": 3024, + "CGGAGA": 3025, + "CGGAGT": 3026, + "CGGAGC": 3027, + "CGGAGG": 3028, + "CGGTAA": 3029, + "CGGTAT": 3030, + "CGGTAC": 3031, + "CGGTAG": 3032, + "CGGTTA": 3033, + "CGGTTT": 3034, + "CGGTTC": 3035, + "CGGTTG": 3036, + "CGGTCA": 3037, + "CGGTCT": 3038, + "CGGTCC": 3039, + "CGGTCG": 3040, + "CGGTGA": 3041, + "CGGTGT": 3042, + "CGGTGC": 3043, + "CGGTGG": 3044, + "CGGCAA": 3045, + "CGGCAT": 3046, + "CGGCAC": 3047, + "CGGCAG": 3048, + "CGGCTA": 3049, + "CGGCTT": 3050, + "CGGCTC": 3051, + "CGGCTG": 3052, + "CGGCCA": 3053, + "CGGCCT": 3054, + "CGGCCC": 3055, + "CGGCCG": 3056, + "CGGCGA": 3057, + "CGGCGT": 3058, + "CGGCGC": 3059, + "CGGCGG": 3060, + "CGGGAA": 3061, + "CGGGAT": 3062, + "CGGGAC": 3063, + "CGGGAG": 3064, + "CGGGTA": 3065, + "CGGGTT": 3066, + "CGGGTC": 3067, + "CGGGTG": 3068, + "CGGGCA": 3069, + "CGGGCT": 3070, + "CGGGCC": 3071, + "CGGGCG": 3072, + "CGGGGA": 3073, + "CGGGGT": 3074, + "CGGGGC": 3075, + "CGGGGG": 3076, + "GAAAAA": 3077, + "GAAAAT": 3078, + "GAAAAC": 3079, + "GAAAAG": 3080, + "GAAATA": 3081, + "GAAATT": 3082, + "GAAATC": 3083, + "GAAATG": 3084, + "GAAACA": 3085, + "GAAACT": 3086, + "GAAACC": 3087, + "GAAACG": 3088, + "GAAAGA": 3089, + "GAAAGT": 3090, + "GAAAGC": 3091, + "GAAAGG": 3092, + "GAATAA": 3093, + "GAATAT": 3094, + "GAATAC": 3095, + "GAATAG": 3096, + "GAATTA": 3097, + "GAATTT": 3098, + "GAATTC": 3099, + "GAATTG": 3100, + "GAATCA": 3101, + "GAATCT": 3102, + "GAATCC": 3103, + "GAATCG": 3104, + "GAATGA": 3105, + "GAATGT": 3106, + "GAATGC": 3107, + "GAATGG": 3108, + "GAACAA": 3109, + "GAACAT": 3110, + "GAACAC": 3111, + "GAACAG": 3112, + "GAACTA": 3113, + "GAACTT": 3114, + "GAACTC": 3115, + "GAACTG": 3116, + "GAACCA": 3117, + "GAACCT": 3118, + "GAACCC": 3119, + "GAACCG": 3120, + "GAACGA": 3121, + "GAACGT": 3122, + "GAACGC": 3123, + "GAACGG": 3124, + "GAAGAA": 3125, + "GAAGAT": 3126, + "GAAGAC": 3127, + "GAAGAG": 3128, + "GAAGTA": 3129, + "GAAGTT": 3130, + "GAAGTC": 3131, + "GAAGTG": 3132, + "GAAGCA": 3133, + "GAAGCT": 3134, + "GAAGCC": 3135, + "GAAGCG": 3136, + "GAAGGA": 3137, + "GAAGGT": 3138, + "GAAGGC": 3139, + "GAAGGG": 3140, + "GATAAA": 3141, + "GATAAT": 3142, + "GATAAC": 3143, + "GATAAG": 3144, + "GATATA": 3145, + "GATATT": 3146, + "GATATC": 3147, + "GATATG": 3148, + "GATACA": 3149, + "GATACT": 3150, + "GATACC": 3151, + "GATACG": 3152, + "GATAGA": 3153, + "GATAGT": 3154, + "GATAGC": 3155, + "GATAGG": 3156, + "GATTAA": 3157, + "GATTAT": 3158, + "GATTAC": 3159, + "GATTAG": 3160, + "GATTTA": 3161, + "GATTTT": 3162, + "GATTTC": 3163, + "GATTTG": 3164, + "GATTCA": 3165, + "GATTCT": 3166, + "GATTCC": 3167, + "GATTCG": 3168, + "GATTGA": 3169, + "GATTGT": 3170, + "GATTGC": 3171, + "GATTGG": 3172, + "GATCAA": 3173, + "GATCAT": 3174, + "GATCAC": 3175, + "GATCAG": 3176, + "GATCTA": 3177, + "GATCTT": 3178, + "GATCTC": 3179, + "GATCTG": 3180, + "GATCCA": 3181, + "GATCCT": 3182, + "GATCCC": 3183, + "GATCCG": 3184, + "GATCGA": 3185, + "GATCGT": 3186, + "GATCGC": 3187, + "GATCGG": 3188, + "GATGAA": 3189, + "GATGAT": 3190, + "GATGAC": 3191, + "GATGAG": 3192, + "GATGTA": 3193, + "GATGTT": 3194, + "GATGTC": 3195, + "GATGTG": 3196, + "GATGCA": 3197, + "GATGCT": 3198, + "GATGCC": 3199, + "GATGCG": 3200, + "GATGGA": 3201, + "GATGGT": 3202, + "GATGGC": 3203, + "GATGGG": 3204, + "GACAAA": 3205, + "GACAAT": 3206, + "GACAAC": 3207, + "GACAAG": 3208, + "GACATA": 3209, + "GACATT": 3210, + "GACATC": 3211, + "GACATG": 3212, + "GACACA": 3213, + "GACACT": 3214, + "GACACC": 3215, + "GACACG": 3216, + "GACAGA": 3217, + "GACAGT": 3218, + "GACAGC": 3219, + "GACAGG": 3220, + "GACTAA": 3221, + "GACTAT": 3222, + "GACTAC": 3223, + "GACTAG": 3224, + "GACTTA": 3225, + "GACTTT": 3226, + "GACTTC": 3227, + "GACTTG": 3228, + "GACTCA": 3229, + "GACTCT": 3230, + "GACTCC": 3231, + "GACTCG": 3232, + "GACTGA": 3233, + "GACTGT": 3234, + "GACTGC": 3235, + "GACTGG": 3236, + "GACCAA": 3237, + "GACCAT": 3238, + "GACCAC": 3239, + "GACCAG": 3240, + "GACCTA": 3241, + "GACCTT": 3242, + "GACCTC": 3243, + "GACCTG": 3244, + "GACCCA": 3245, + "GACCCT": 3246, + "GACCCC": 3247, + "GACCCG": 3248, + "GACCGA": 3249, + "GACCGT": 3250, + "GACCGC": 3251, + "GACCGG": 3252, + "GACGAA": 3253, + "GACGAT": 3254, + "GACGAC": 3255, + "GACGAG": 3256, + "GACGTA": 3257, + "GACGTT": 3258, + "GACGTC": 3259, + "GACGTG": 3260, + "GACGCA": 3261, + "GACGCT": 3262, + "GACGCC": 3263, + "GACGCG": 3264, + "GACGGA": 3265, + "GACGGT": 3266, + "GACGGC": 3267, + "GACGGG": 3268, + "GAGAAA": 3269, + "GAGAAT": 3270, + "GAGAAC": 3271, + "GAGAAG": 3272, + "GAGATA": 3273, + "GAGATT": 3274, + "GAGATC": 3275, + "GAGATG": 3276, + "GAGACA": 3277, + "GAGACT": 3278, + "GAGACC": 3279, + "GAGACG": 3280, + "GAGAGA": 3281, + "GAGAGT": 3282, + "GAGAGC": 3283, + "GAGAGG": 3284, + "GAGTAA": 3285, + "GAGTAT": 3286, + "GAGTAC": 3287, + "GAGTAG": 3288, + "GAGTTA": 3289, + "GAGTTT": 3290, + "GAGTTC": 3291, + "GAGTTG": 3292, + "GAGTCA": 3293, + "GAGTCT": 3294, + "GAGTCC": 3295, + "GAGTCG": 3296, + "GAGTGA": 3297, + "GAGTGT": 3298, + "GAGTGC": 3299, + "GAGTGG": 3300, + "GAGCAA": 3301, + "GAGCAT": 3302, + "GAGCAC": 3303, + "GAGCAG": 3304, + "GAGCTA": 3305, + "GAGCTT": 3306, + "GAGCTC": 3307, + "GAGCTG": 3308, + "GAGCCA": 3309, + "GAGCCT": 3310, + "GAGCCC": 3311, + "GAGCCG": 3312, + "GAGCGA": 3313, + "GAGCGT": 3314, + "GAGCGC": 3315, + "GAGCGG": 3316, + "GAGGAA": 3317, + "GAGGAT": 3318, + "GAGGAC": 3319, + "GAGGAG": 3320, + "GAGGTA": 3321, + "GAGGTT": 3322, + "GAGGTC": 3323, + "GAGGTG": 3324, + "GAGGCA": 3325, + "GAGGCT": 3326, + "GAGGCC": 3327, + "GAGGCG": 3328, + "GAGGGA": 3329, + "GAGGGT": 3330, + "GAGGGC": 3331, + "GAGGGG": 3332, + "GTAAAA": 3333, + "GTAAAT": 3334, + "GTAAAC": 3335, + "GTAAAG": 3336, + "GTAATA": 3337, + "GTAATT": 3338, + "GTAATC": 3339, + "GTAATG": 3340, + "GTAACA": 3341, + "GTAACT": 3342, + "GTAACC": 3343, + "GTAACG": 3344, + "GTAAGA": 3345, + "GTAAGT": 3346, + "GTAAGC": 3347, + "GTAAGG": 3348, + "GTATAA": 3349, + "GTATAT": 3350, + "GTATAC": 3351, + "GTATAG": 3352, + "GTATTA": 3353, + "GTATTT": 3354, + "GTATTC": 3355, + "GTATTG": 3356, + "GTATCA": 3357, + "GTATCT": 3358, + "GTATCC": 3359, + "GTATCG": 3360, + "GTATGA": 3361, + "GTATGT": 3362, + "GTATGC": 3363, + "GTATGG": 3364, + "GTACAA": 3365, + "GTACAT": 3366, + "GTACAC": 3367, + "GTACAG": 3368, + "GTACTA": 3369, + "GTACTT": 3370, + "GTACTC": 3371, + "GTACTG": 3372, + "GTACCA": 3373, + "GTACCT": 3374, + "GTACCC": 3375, + "GTACCG": 3376, + "GTACGA": 3377, + "GTACGT": 3378, + "GTACGC": 3379, + "GTACGG": 3380, + "GTAGAA": 3381, + "GTAGAT": 3382, + "GTAGAC": 3383, + "GTAGAG": 3384, + "GTAGTA": 3385, + "GTAGTT": 3386, + "GTAGTC": 3387, + "GTAGTG": 3388, + "GTAGCA": 3389, + "GTAGCT": 3390, + "GTAGCC": 3391, + "GTAGCG": 3392, + "GTAGGA": 3393, + "GTAGGT": 3394, + "GTAGGC": 3395, + "GTAGGG": 3396, + "GTTAAA": 3397, + "GTTAAT": 3398, + "GTTAAC": 3399, + "GTTAAG": 3400, + "GTTATA": 3401, + "GTTATT": 3402, + "GTTATC": 3403, + "GTTATG": 3404, + "GTTACA": 3405, + "GTTACT": 3406, + "GTTACC": 3407, + "GTTACG": 3408, + "GTTAGA": 3409, + "GTTAGT": 3410, + "GTTAGC": 3411, + "GTTAGG": 3412, + "GTTTAA": 3413, + "GTTTAT": 3414, + "GTTTAC": 3415, + "GTTTAG": 3416, + "GTTTTA": 3417, + "GTTTTT": 3418, + "GTTTTC": 3419, + "GTTTTG": 3420, + "GTTTCA": 3421, + "GTTTCT": 3422, + "GTTTCC": 3423, + "GTTTCG": 3424, + "GTTTGA": 3425, + "GTTTGT": 3426, + "GTTTGC": 3427, + "GTTTGG": 3428, + "GTTCAA": 3429, + "GTTCAT": 3430, + "GTTCAC": 3431, + "GTTCAG": 3432, + "GTTCTA": 3433, + "GTTCTT": 3434, + "GTTCTC": 3435, + "GTTCTG": 3436, + "GTTCCA": 3437, + "GTTCCT": 3438, + "GTTCCC": 3439, + "GTTCCG": 3440, + "GTTCGA": 3441, + "GTTCGT": 3442, + "GTTCGC": 3443, + "GTTCGG": 3444, + "GTTGAA": 3445, + "GTTGAT": 3446, + "GTTGAC": 3447, + "GTTGAG": 3448, + "GTTGTA": 3449, + "GTTGTT": 3450, + "GTTGTC": 3451, + "GTTGTG": 3452, + "GTTGCA": 3453, + "GTTGCT": 3454, + "GTTGCC": 3455, + "GTTGCG": 3456, + "GTTGGA": 3457, + "GTTGGT": 3458, + "GTTGGC": 3459, + "GTTGGG": 3460, + "GTCAAA": 3461, + "GTCAAT": 3462, + "GTCAAC": 3463, + "GTCAAG": 3464, + "GTCATA": 3465, + "GTCATT": 3466, + "GTCATC": 3467, + "GTCATG": 3468, + "GTCACA": 3469, + "GTCACT": 3470, + "GTCACC": 3471, + "GTCACG": 3472, + "GTCAGA": 3473, + "GTCAGT": 3474, + "GTCAGC": 3475, + "GTCAGG": 3476, + "GTCTAA": 3477, + "GTCTAT": 3478, + "GTCTAC": 3479, + "GTCTAG": 3480, + "GTCTTA": 3481, + "GTCTTT": 3482, + "GTCTTC": 3483, + "GTCTTG": 3484, + "GTCTCA": 3485, + "GTCTCT": 3486, + "GTCTCC": 3487, + "GTCTCG": 3488, + "GTCTGA": 3489, + "GTCTGT": 3490, + "GTCTGC": 3491, + "GTCTGG": 3492, + "GTCCAA": 3493, + "GTCCAT": 3494, + "GTCCAC": 3495, + "GTCCAG": 3496, + "GTCCTA": 3497, + "GTCCTT": 3498, + "GTCCTC": 3499, + "GTCCTG": 3500, + "GTCCCA": 3501, + "GTCCCT": 3502, + "GTCCCC": 3503, + "GTCCCG": 3504, + "GTCCGA": 3505, + "GTCCGT": 3506, + "GTCCGC": 3507, + "GTCCGG": 3508, + "GTCGAA": 3509, + "GTCGAT": 3510, + "GTCGAC": 3511, + "GTCGAG": 3512, + "GTCGTA": 3513, + "GTCGTT": 3514, + "GTCGTC": 3515, + "GTCGTG": 3516, + "GTCGCA": 3517, + "GTCGCT": 3518, + "GTCGCC": 3519, + "GTCGCG": 3520, + "GTCGGA": 3521, + "GTCGGT": 3522, + "GTCGGC": 3523, + "GTCGGG": 3524, + "GTGAAA": 3525, + "GTGAAT": 3526, + "GTGAAC": 3527, + "GTGAAG": 3528, + "GTGATA": 3529, + "GTGATT": 3530, + "GTGATC": 3531, + "GTGATG": 3532, + "GTGACA": 3533, + "GTGACT": 3534, + "GTGACC": 3535, + "GTGACG": 3536, + "GTGAGA": 3537, + "GTGAGT": 3538, + "GTGAGC": 3539, + "GTGAGG": 3540, + "GTGTAA": 3541, + "GTGTAT": 3542, + "GTGTAC": 3543, + "GTGTAG": 3544, + "GTGTTA": 3545, + "GTGTTT": 3546, + "GTGTTC": 3547, + "GTGTTG": 3548, + "GTGTCA": 3549, + "GTGTCT": 3550, + "GTGTCC": 3551, + "GTGTCG": 3552, + "GTGTGA": 3553, + "GTGTGT": 3554, + "GTGTGC": 3555, + "GTGTGG": 3556, + "GTGCAA": 3557, + "GTGCAT": 3558, + "GTGCAC": 3559, + "GTGCAG": 3560, + "GTGCTA": 3561, + "GTGCTT": 3562, + "GTGCTC": 3563, + "GTGCTG": 3564, + "GTGCCA": 3565, + "GTGCCT": 3566, + "GTGCCC": 3567, + "GTGCCG": 3568, + "GTGCGA": 3569, + "GTGCGT": 3570, + "GTGCGC": 3571, + "GTGCGG": 3572, + "GTGGAA": 3573, + "GTGGAT": 3574, + "GTGGAC": 3575, + "GTGGAG": 3576, + "GTGGTA": 3577, + "GTGGTT": 3578, + "GTGGTC": 3579, + "GTGGTG": 3580, + "GTGGCA": 3581, + "GTGGCT": 3582, + "GTGGCC": 3583, + "GTGGCG": 3584, + "GTGGGA": 3585, + "GTGGGT": 3586, + "GTGGGC": 3587, + "GTGGGG": 3588, + "GCAAAA": 3589, + "GCAAAT": 3590, + "GCAAAC": 3591, + "GCAAAG": 3592, + "GCAATA": 3593, + "GCAATT": 3594, + "GCAATC": 3595, + "GCAATG": 3596, + "GCAACA": 3597, + "GCAACT": 3598, + "GCAACC": 3599, + "GCAACG": 3600, + "GCAAGA": 3601, + "GCAAGT": 3602, + "GCAAGC": 3603, + "GCAAGG": 3604, + "GCATAA": 3605, + "GCATAT": 3606, + "GCATAC": 3607, + "GCATAG": 3608, + "GCATTA": 3609, + "GCATTT": 3610, + "GCATTC": 3611, + "GCATTG": 3612, + "GCATCA": 3613, + "GCATCT": 3614, + "GCATCC": 3615, + "GCATCG": 3616, + "GCATGA": 3617, + "GCATGT": 3618, + "GCATGC": 3619, + "GCATGG": 3620, + "GCACAA": 3621, + "GCACAT": 3622, + "GCACAC": 3623, + "GCACAG": 3624, + "GCACTA": 3625, + "GCACTT": 3626, + "GCACTC": 3627, + "GCACTG": 3628, + "GCACCA": 3629, + "GCACCT": 3630, + "GCACCC": 3631, + "GCACCG": 3632, + "GCACGA": 3633, + "GCACGT": 3634, + "GCACGC": 3635, + "GCACGG": 3636, + "GCAGAA": 3637, + "GCAGAT": 3638, + "GCAGAC": 3639, + "GCAGAG": 3640, + "GCAGTA": 3641, + "GCAGTT": 3642, + "GCAGTC": 3643, + "GCAGTG": 3644, + "GCAGCA": 3645, + "GCAGCT": 3646, + "GCAGCC": 3647, + "GCAGCG": 3648, + "GCAGGA": 3649, + "GCAGGT": 3650, + "GCAGGC": 3651, + "GCAGGG": 3652, + "GCTAAA": 3653, + "GCTAAT": 3654, + "GCTAAC": 3655, + "GCTAAG": 3656, + "GCTATA": 3657, + "GCTATT": 3658, + "GCTATC": 3659, + "GCTATG": 3660, + "GCTACA": 3661, + "GCTACT": 3662, + "GCTACC": 3663, + "GCTACG": 3664, + "GCTAGA": 3665, + "GCTAGT": 3666, + "GCTAGC": 3667, + "GCTAGG": 3668, + "GCTTAA": 3669, + "GCTTAT": 3670, + "GCTTAC": 3671, + "GCTTAG": 3672, + "GCTTTA": 3673, + "GCTTTT": 3674, + "GCTTTC": 3675, + "GCTTTG": 3676, + "GCTTCA": 3677, + "GCTTCT": 3678, + "GCTTCC": 3679, + "GCTTCG": 3680, + "GCTTGA": 3681, + "GCTTGT": 3682, + "GCTTGC": 3683, + "GCTTGG": 3684, + "GCTCAA": 3685, + "GCTCAT": 3686, + "GCTCAC": 3687, + "GCTCAG": 3688, + "GCTCTA": 3689, + "GCTCTT": 3690, + "GCTCTC": 3691, + "GCTCTG": 3692, + "GCTCCA": 3693, + "GCTCCT": 3694, + "GCTCCC": 3695, + "GCTCCG": 3696, + "GCTCGA": 3697, + "GCTCGT": 3698, + "GCTCGC": 3699, + "GCTCGG": 3700, + "GCTGAA": 3701, + "GCTGAT": 3702, + "GCTGAC": 3703, + "GCTGAG": 3704, + "GCTGTA": 3705, + "GCTGTT": 3706, + "GCTGTC": 3707, + "GCTGTG": 3708, + "GCTGCA": 3709, + "GCTGCT": 3710, + "GCTGCC": 3711, + "GCTGCG": 3712, + "GCTGGA": 3713, + "GCTGGT": 3714, + "GCTGGC": 3715, + "GCTGGG": 3716, + "GCCAAA": 3717, + "GCCAAT": 3718, + "GCCAAC": 3719, + "GCCAAG": 3720, + "GCCATA": 3721, + "GCCATT": 3722, + "GCCATC": 3723, + "GCCATG": 3724, + "GCCACA": 3725, + "GCCACT": 3726, + "GCCACC": 3727, + "GCCACG": 3728, + "GCCAGA": 3729, + "GCCAGT": 3730, + "GCCAGC": 3731, + "GCCAGG": 3732, + "GCCTAA": 3733, + "GCCTAT": 3734, + "GCCTAC": 3735, + "GCCTAG": 3736, + "GCCTTA": 3737, + "GCCTTT": 3738, + "GCCTTC": 3739, + "GCCTTG": 3740, + "GCCTCA": 3741, + "GCCTCT": 3742, + "GCCTCC": 3743, + "GCCTCG": 3744, + "GCCTGA": 3745, + "GCCTGT": 3746, + "GCCTGC": 3747, + "GCCTGG": 3748, + "GCCCAA": 3749, + "GCCCAT": 3750, + "GCCCAC": 3751, + "GCCCAG": 3752, + "GCCCTA": 3753, + "GCCCTT": 3754, + "GCCCTC": 3755, + "GCCCTG": 3756, + "GCCCCA": 3757, + "GCCCCT": 3758, + "GCCCCC": 3759, + "GCCCCG": 3760, + "GCCCGA": 3761, + "GCCCGT": 3762, + "GCCCGC": 3763, + "GCCCGG": 3764, + "GCCGAA": 3765, + "GCCGAT": 3766, + "GCCGAC": 3767, + "GCCGAG": 3768, + "GCCGTA": 3769, + "GCCGTT": 3770, + "GCCGTC": 3771, + "GCCGTG": 3772, + "GCCGCA": 3773, + "GCCGCT": 3774, + "GCCGCC": 3775, + "GCCGCG": 3776, + "GCCGGA": 3777, + "GCCGGT": 3778, + "GCCGGC": 3779, + "GCCGGG": 3780, + "GCGAAA": 3781, + "GCGAAT": 3782, + "GCGAAC": 3783, + "GCGAAG": 3784, + "GCGATA": 3785, + "GCGATT": 3786, + "GCGATC": 3787, + "GCGATG": 3788, + "GCGACA": 3789, + "GCGACT": 3790, + "GCGACC": 3791, + "GCGACG": 3792, + "GCGAGA": 3793, + "GCGAGT": 3794, + "GCGAGC": 3795, + "GCGAGG": 3796, + "GCGTAA": 3797, + "GCGTAT": 3798, + "GCGTAC": 3799, + "GCGTAG": 3800, + "GCGTTA": 3801, + "GCGTTT": 3802, + "GCGTTC": 3803, + "GCGTTG": 3804, + "GCGTCA": 3805, + "GCGTCT": 3806, + "GCGTCC": 3807, + "GCGTCG": 3808, + "GCGTGA": 3809, + "GCGTGT": 3810, + "GCGTGC": 3811, + "GCGTGG": 3812, + "GCGCAA": 3813, + "GCGCAT": 3814, + "GCGCAC": 3815, + "GCGCAG": 3816, + "GCGCTA": 3817, + "GCGCTT": 3818, + "GCGCTC": 3819, + "GCGCTG": 3820, + "GCGCCA": 3821, + "GCGCCT": 3822, + "GCGCCC": 3823, + "GCGCCG": 3824, + "GCGCGA": 3825, + "GCGCGT": 3826, + "GCGCGC": 3827, + "GCGCGG": 3828, + "GCGGAA": 3829, + "GCGGAT": 3830, + "GCGGAC": 3831, + "GCGGAG": 3832, + "GCGGTA": 3833, + "GCGGTT": 3834, + "GCGGTC": 3835, + "GCGGTG": 3836, + "GCGGCA": 3837, + "GCGGCT": 3838, + "GCGGCC": 3839, + "GCGGCG": 3840, + "GCGGGA": 3841, + "GCGGGT": 3842, + "GCGGGC": 3843, + "GCGGGG": 3844, + "GGAAAA": 3845, + "GGAAAT": 3846, + "GGAAAC": 3847, + "GGAAAG": 3848, + "GGAATA": 3849, + "GGAATT": 3850, + "GGAATC": 3851, + "GGAATG": 3852, + "GGAACA": 3853, + "GGAACT": 3854, + "GGAACC": 3855, + "GGAACG": 3856, + "GGAAGA": 3857, + "GGAAGT": 3858, + "GGAAGC": 3859, + "GGAAGG": 3860, + "GGATAA": 3861, + "GGATAT": 3862, + "GGATAC": 3863, + "GGATAG": 3864, + "GGATTA": 3865, + "GGATTT": 3866, + "GGATTC": 3867, + "GGATTG": 3868, + "GGATCA": 3869, + "GGATCT": 3870, + "GGATCC": 3871, + "GGATCG": 3872, + "GGATGA": 3873, + "GGATGT": 3874, + "GGATGC": 3875, + "GGATGG": 3876, + "GGACAA": 3877, + "GGACAT": 3878, + "GGACAC": 3879, + "GGACAG": 3880, + "GGACTA": 3881, + "GGACTT": 3882, + "GGACTC": 3883, + "GGACTG": 3884, + "GGACCA": 3885, + "GGACCT": 3886, + "GGACCC": 3887, + "GGACCG": 3888, + "GGACGA": 3889, + "GGACGT": 3890, + "GGACGC": 3891, + "GGACGG": 3892, + "GGAGAA": 3893, + "GGAGAT": 3894, + "GGAGAC": 3895, + "GGAGAG": 3896, + "GGAGTA": 3897, + "GGAGTT": 3898, + "GGAGTC": 3899, + "GGAGTG": 3900, + "GGAGCA": 3901, + "GGAGCT": 3902, + "GGAGCC": 3903, + "GGAGCG": 3904, + "GGAGGA": 3905, + "GGAGGT": 3906, + "GGAGGC": 3907, + "GGAGGG": 3908, + "GGTAAA": 3909, + "GGTAAT": 3910, + "GGTAAC": 3911, + "GGTAAG": 3912, + "GGTATA": 3913, + "GGTATT": 3914, + "GGTATC": 3915, + "GGTATG": 3916, + "GGTACA": 3917, + "GGTACT": 3918, + "GGTACC": 3919, + "GGTACG": 3920, + "GGTAGA": 3921, + "GGTAGT": 3922, + "GGTAGC": 3923, + "GGTAGG": 3924, + "GGTTAA": 3925, + "GGTTAT": 3926, + "GGTTAC": 3927, + "GGTTAG": 3928, + "GGTTTA": 3929, + "GGTTTT": 3930, + "GGTTTC": 3931, + "GGTTTG": 3932, + "GGTTCA": 3933, + "GGTTCT": 3934, + "GGTTCC": 3935, + "GGTTCG": 3936, + "GGTTGA": 3937, + "GGTTGT": 3938, + "GGTTGC": 3939, + "GGTTGG": 3940, + "GGTCAA": 3941, + "GGTCAT": 3942, + "GGTCAC": 3943, + "GGTCAG": 3944, + "GGTCTA": 3945, + "GGTCTT": 3946, + "GGTCTC": 3947, + "GGTCTG": 3948, + "GGTCCA": 3949, + "GGTCCT": 3950, + "GGTCCC": 3951, + "GGTCCG": 3952, + "GGTCGA": 3953, + "GGTCGT": 3954, + "GGTCGC": 3955, + "GGTCGG": 3956, + "GGTGAA": 3957, + "GGTGAT": 3958, + "GGTGAC": 3959, + "GGTGAG": 3960, + "GGTGTA": 3961, + "GGTGTT": 3962, + "GGTGTC": 3963, + "GGTGTG": 3964, + "GGTGCA": 3965, + "GGTGCT": 3966, + "GGTGCC": 3967, + "GGTGCG": 3968, + "GGTGGA": 3969, + "GGTGGT": 3970, + "GGTGGC": 3971, + "GGTGGG": 3972, + "GGCAAA": 3973, + "GGCAAT": 3974, + "GGCAAC": 3975, + "GGCAAG": 3976, + "GGCATA": 3977, + "GGCATT": 3978, + "GGCATC": 3979, + "GGCATG": 3980, + "GGCACA": 3981, + "GGCACT": 3982, + "GGCACC": 3983, + "GGCACG": 3984, + "GGCAGA": 3985, + "GGCAGT": 3986, + "GGCAGC": 3987, + "GGCAGG": 3988, + "GGCTAA": 3989, + "GGCTAT": 3990, + "GGCTAC": 3991, + "GGCTAG": 3992, + "GGCTTA": 3993, + "GGCTTT": 3994, + "GGCTTC": 3995, + "GGCTTG": 3996, + "GGCTCA": 3997, + "GGCTCT": 3998, + "GGCTCC": 3999, + "GGCTCG": 4000, + "GGCTGA": 4001, + "GGCTGT": 4002, + "GGCTGC": 4003, + "GGCTGG": 4004, + "GGCCAA": 4005, + "GGCCAT": 4006, + "GGCCAC": 4007, + "GGCCAG": 4008, + "GGCCTA": 4009, + "GGCCTT": 4010, + "GGCCTC": 4011, + "GGCCTG": 4012, + "GGCCCA": 4013, + "GGCCCT": 4014, + "GGCCCC": 4015, + "GGCCCG": 4016, + "GGCCGA": 4017, + "GGCCGT": 4018, + "GGCCGC": 4019, + "GGCCGG": 4020, + "GGCGAA": 4021, + "GGCGAT": 4022, + "GGCGAC": 4023, + "GGCGAG": 4024, + "GGCGTA": 4025, + "GGCGTT": 4026, + "GGCGTC": 4027, + "GGCGTG": 4028, + "GGCGCA": 4029, + "GGCGCT": 4030, + "GGCGCC": 4031, + "GGCGCG": 4032, + "GGCGGA": 4033, + "GGCGGT": 4034, + "GGCGGC": 4035, + "GGCGGG": 4036, + "GGGAAA": 4037, + "GGGAAT": 4038, + "GGGAAC": 4039, + "GGGAAG": 4040, + "GGGATA": 4041, + "GGGATT": 4042, + "GGGATC": 4043, + "GGGATG": 4044, + "GGGACA": 4045, + "GGGACT": 4046, + "GGGACC": 4047, + "GGGACG": 4048, + "GGGAGA": 4049, + "GGGAGT": 4050, + "GGGAGC": 4051, + "GGGAGG": 4052, + "GGGTAA": 4053, + "GGGTAT": 4054, + "GGGTAC": 4055, + "GGGTAG": 4056, + "GGGTTA": 4057, + "GGGTTT": 4058, + "GGGTTC": 4059, + "GGGTTG": 4060, + "GGGTCA": 4061, + "GGGTCT": 4062, + "GGGTCC": 4063, + "GGGTCG": 4064, + "GGGTGA": 4065, + "GGGTGT": 4066, + "GGGTGC": 4067, + "GGGTGG": 4068, + "GGGCAA": 4069, + "GGGCAT": 4070, + "GGGCAC": 4071, + "GGGCAG": 4072, + "GGGCTA": 4073, + "GGGCTT": 4074, + "GGGCTC": 4075, + "GGGCTG": 4076, + "GGGCCA": 4077, + "GGGCCT": 4078, + "GGGCCC": 4079, + "GGGCCG": 4080, + "GGGCGA": 4081, + "GGGCGT": 4082, + "GGGCGC": 4083, + "GGGCGG": 4084, + "GGGGAA": 4085, + "GGGGAT": 4086, + "GGGGAC": 4087, + "GGGGAG": 4088, + "GGGGTA": 4089, + "GGGGTT": 4090, + "GGGGTC": 4091, + "GGGGTG": 4092, + "GGGGCA": 4093, + "GGGGCT": 4094, + "GGGGCC": 4095, + "GGGGCG": 4096, + "GGGGGA": 4097, + "GGGGGT": 4098, + "GGGGGC": 4099, + "GGGGGG": 4100 + } + } +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/tokenizer_config.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..604e2b1b29af1dcb3920d2828b6a90276e5e3d02 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/tokenizer_config.json @@ -0,0 +1,58 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "[UNK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "[CLS]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "[SEP]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "4": { + "content": "[MASK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "clean_up_tokenization_spaces": true, + "cls_token": "[CLS]", + "do_basic_tokenize": true, + "do_lower_case": false, + "mask_token": "[MASK]", + "max_len": 512, + "model_max_length": 512, + "never_split": null, + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "strip_accents": null, + "tokenize_chinese_chars": true, + "tokenizer_class": "BertTokenizer", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/trainer_state.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..87b4d20df7415de48fe92c788c50784addd179ff --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/trainer_state.json @@ -0,0 +1,1783 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 25.0, + "eval_steps": 500, + "global_step": 6250, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.1, + "grad_norm": 0.2192622423171997, + "learning_rate": 5.319148936170213e-05, + "loss": 0.6949, + "step": 25 + }, + { + "epoch": 0.2, + "grad_norm": 0.22659960389137268, + "learning_rate": 0.00010638297872340425, + "loss": 0.6926, + "step": 50 + }, + { + "epoch": 0.3, + "grad_norm": 0.33277687430381775, + "learning_rate": 0.00015957446808510637, + "loss": 0.693, + "step": 75 + }, + { + "epoch": 0.4, + "grad_norm": 0.5183415412902832, + "learning_rate": 0.0002127659574468085, + "loss": 0.6928, + "step": 100 + }, + { + "epoch": 0.5, + "grad_norm": 0.21227706968784332, + "learning_rate": 0.00026595744680851064, + "loss": 0.6938, + "step": 125 + }, + { + "epoch": 0.6, + "grad_norm": 0.1846485733985901, + "learning_rate": 0.00031914893617021275, + "loss": 0.6942, + "step": 150 + }, + { + "epoch": 0.7, + "grad_norm": 0.27858808636665344, + "learning_rate": 0.0003723404255319149, + "loss": 0.6931, + "step": 175 + }, + { + "epoch": 0.8, + "grad_norm": 0.26012367010116577, + "learning_rate": 0.0003992081821181128, + "loss": 0.6936, + "step": 200 + }, + { + "epoch": 0.9, + "grad_norm": 0.2534612715244293, + "learning_rate": 0.0003975585615308479, + "loss": 0.6939, + "step": 225 + }, + { + "epoch": 1.0, + "grad_norm": 0.42967700958251953, + "learning_rate": 0.00039590894094358297, + "loss": 0.6928, + "step": 250 + }, + { + "epoch": 1.1, + "grad_norm": 0.2613862454891205, + "learning_rate": 0.0003942593203563181, + "loss": 0.6918, + "step": 275 + }, + { + "epoch": 1.2, + "grad_norm": 0.5514339208602905, + "learning_rate": 0.0003926096997690532, + "loss": 0.6918, + "step": 300 + }, + { + "epoch": 1.3, + "grad_norm": 0.33801883459091187, + "learning_rate": 0.00039096007918178817, + "loss": 0.6906, + "step": 325 + }, + { + "epoch": 1.4, + "grad_norm": 0.2953580617904663, + "learning_rate": 0.0003893104585945233, + "loss": 0.6886, + "step": 350 + }, + { + "epoch": 1.5, + "grad_norm": 0.5340980887413025, + "learning_rate": 0.0003876608380072583, + "loss": 0.6877, + "step": 375 + }, + { + "epoch": 1.6, + "grad_norm": 0.3042701482772827, + "learning_rate": 0.00038601121741999343, + "loss": 0.6853, + "step": 400 + }, + { + "epoch": 1.7, + "grad_norm": 0.44497063755989075, + "learning_rate": 0.0003843615968327285, + "loss": 0.6772, + "step": 425 + }, + { + "epoch": 1.8, + "grad_norm": 0.6260074973106384, + "learning_rate": 0.0003827119762454636, + "loss": 0.6671, + "step": 450 + }, + { + "epoch": 1.9, + "grad_norm": 0.9574269652366638, + "learning_rate": 0.00038106235565819863, + "loss": 0.6517, + "step": 475 + }, + { + "epoch": 2.0, + "grad_norm": 0.9055777192115784, + "learning_rate": 0.0003794127350709337, + "loss": 0.6438, + "step": 500 + }, + { + "epoch": 2.1, + "grad_norm": 0.7860134840011597, + "learning_rate": 0.0003777631144836688, + "loss": 0.6379, + "step": 525 + }, + { + "epoch": 2.2, + "grad_norm": 1.5324610471725464, + "learning_rate": 0.00037611349389640383, + "loss": 0.6143, + "step": 550 + }, + { + "epoch": 2.3, + "grad_norm": 1.8535794019699097, + "learning_rate": 0.00037446387330913894, + "loss": 0.607, + "step": 575 + }, + { + "epoch": 2.4, + "grad_norm": 1.8850122690200806, + "learning_rate": 0.000372814252721874, + "loss": 0.5642, + "step": 600 + }, + { + "epoch": 2.5, + "grad_norm": 1.288278579711914, + "learning_rate": 0.00037116463213460903, + "loss": 0.5739, + "step": 625 + }, + { + "epoch": 2.6, + "grad_norm": 1.344173550605774, + "learning_rate": 0.00036951501154734414, + "loss": 0.5718, + "step": 650 + }, + { + "epoch": 2.7, + "grad_norm": 1.430834174156189, + "learning_rate": 0.0003678653909600792, + "loss": 0.543, + "step": 675 + }, + { + "epoch": 2.8, + "grad_norm": 1.5948325395584106, + "learning_rate": 0.0003662157703728143, + "loss": 0.5434, + "step": 700 + }, + { + "epoch": 2.9, + "grad_norm": 2.688519239425659, + "learning_rate": 0.00036456614978554934, + "loss": 0.5842, + "step": 725 + }, + { + "epoch": 3.0, + "grad_norm": 1.6625819206237793, + "learning_rate": 0.00036291652919828444, + "loss": 0.5526, + "step": 750 + }, + { + "epoch": 3.1, + "grad_norm": 1.869725227355957, + "learning_rate": 0.00036126690861101944, + "loss": 0.53, + "step": 775 + }, + { + "epoch": 3.2, + "grad_norm": 2.1625709533691406, + "learning_rate": 0.00035961728802375454, + "loss": 0.5341, + "step": 800 + }, + { + "epoch": 3.3, + "grad_norm": 1.9650202989578247, + "learning_rate": 0.0003579676674364896, + "loss": 0.5297, + "step": 825 + }, + { + "epoch": 3.4, + "grad_norm": 1.3358676433563232, + "learning_rate": 0.0003563180468492247, + "loss": 0.5273, + "step": 850 + }, + { + "epoch": 3.5, + "grad_norm": 2.1927096843719482, + "learning_rate": 0.0003546684262619598, + "loss": 0.5084, + "step": 875 + }, + { + "epoch": 3.6, + "grad_norm": 2.2377188205718994, + "learning_rate": 0.00035301880567469485, + "loss": 0.4972, + "step": 900 + }, + { + "epoch": 3.7, + "grad_norm": 1.8601839542388916, + "learning_rate": 0.0003513691850874299, + "loss": 0.4852, + "step": 925 + }, + { + "epoch": 3.8, + "grad_norm": 1.4995975494384766, + "learning_rate": 0.00034971956450016495, + "loss": 0.5309, + "step": 950 + }, + { + "epoch": 3.9, + "grad_norm": 2.0536930561065674, + "learning_rate": 0.00034806994391290005, + "loss": 0.5249, + "step": 975 + }, + { + "epoch": 4.0, + "grad_norm": 2.084728479385376, + "learning_rate": 0.0003464203233256351, + "loss": 0.4771, + "step": 1000 + }, + { + "epoch": 4.1, + "grad_norm": 1.705510139465332, + "learning_rate": 0.0003447707027383702, + "loss": 0.5181, + "step": 1025 + }, + { + "epoch": 4.2, + "grad_norm": 3.83380389213562, + "learning_rate": 0.00034312108215110525, + "loss": 0.4873, + "step": 1050 + }, + { + "epoch": 4.3, + "grad_norm": 2.0163533687591553, + "learning_rate": 0.0003414714615638403, + "loss": 0.4801, + "step": 1075 + }, + { + "epoch": 4.4, + "grad_norm": 1.9917407035827637, + "learning_rate": 0.0003398218409765754, + "loss": 0.4766, + "step": 1100 + }, + { + "epoch": 4.5, + "grad_norm": 2.111729860305786, + "learning_rate": 0.00033817222038931045, + "loss": 0.4877, + "step": 1125 + }, + { + "epoch": 4.6, + "grad_norm": 1.5487738847732544, + "learning_rate": 0.00033652259980204556, + "loss": 0.507, + "step": 1150 + }, + { + "epoch": 4.7, + "grad_norm": 1.8715145587921143, + "learning_rate": 0.0003348729792147806, + "loss": 0.452, + "step": 1175 + }, + { + "epoch": 4.8, + "grad_norm": 2.2618906497955322, + "learning_rate": 0.0003332233586275157, + "loss": 0.4799, + "step": 1200 + }, + { + "epoch": 4.9, + "grad_norm": 1.657012939453125, + "learning_rate": 0.00033157373804025076, + "loss": 0.4715, + "step": 1225 + }, + { + "epoch": 5.0, + "grad_norm": 2.8894591331481934, + "learning_rate": 0.0003299241174529858, + "loss": 0.4437, + "step": 1250 + }, + { + "epoch": 5.1, + "grad_norm": 1.9179915189743042, + "learning_rate": 0.0003282744968657209, + "loss": 0.4796, + "step": 1275 + }, + { + "epoch": 5.2, + "grad_norm": 2.1816720962524414, + "learning_rate": 0.00032662487627845596, + "loss": 0.4463, + "step": 1300 + }, + { + "epoch": 5.3, + "grad_norm": 2.2761082649230957, + "learning_rate": 0.00032497525569119106, + "loss": 0.4215, + "step": 1325 + }, + { + "epoch": 5.4, + "grad_norm": 1.7190684080123901, + "learning_rate": 0.0003233256351039261, + "loss": 0.4481, + "step": 1350 + }, + { + "epoch": 5.5, + "grad_norm": 1.4793736934661865, + "learning_rate": 0.00032167601451666116, + "loss": 0.44, + "step": 1375 + }, + { + "epoch": 5.6, + "grad_norm": 2.5064358711242676, + "learning_rate": 0.00032002639392939627, + "loss": 0.4583, + "step": 1400 + }, + { + "epoch": 5.7, + "grad_norm": 1.8364307880401611, + "learning_rate": 0.0003183767733421313, + "loss": 0.4516, + "step": 1425 + }, + { + "epoch": 5.8, + "grad_norm": 1.3747583627700806, + "learning_rate": 0.0003167271527548664, + "loss": 0.4535, + "step": 1450 + }, + { + "epoch": 5.9, + "grad_norm": 2.580183982849121, + "learning_rate": 0.00031507753216760147, + "loss": 0.4325, + "step": 1475 + }, + { + "epoch": 6.0, + "grad_norm": 1.562374472618103, + "learning_rate": 0.00031342791158033657, + "loss": 0.4616, + "step": 1500 + }, + { + "epoch": 6.1, + "grad_norm": 1.4970141649246216, + "learning_rate": 0.00031177829099307157, + "loss": 0.4607, + "step": 1525 + }, + { + "epoch": 6.2, + "grad_norm": 2.436326503753662, + "learning_rate": 0.00031012867040580667, + "loss": 0.4498, + "step": 1550 + }, + { + "epoch": 6.3, + "grad_norm": 1.9849016666412354, + "learning_rate": 0.0003084790498185417, + "loss": 0.4262, + "step": 1575 + }, + { + "epoch": 6.4, + "grad_norm": 1.6405338048934937, + "learning_rate": 0.0003068294292312768, + "loss": 0.4225, + "step": 1600 + }, + { + "epoch": 6.5, + "grad_norm": 1.8233951330184937, + "learning_rate": 0.0003051798086440119, + "loss": 0.4248, + "step": 1625 + }, + { + "epoch": 6.6, + "grad_norm": 1.396469235420227, + "learning_rate": 0.000303530188056747, + "loss": 0.412, + "step": 1650 + }, + { + "epoch": 6.7, + "grad_norm": 1.5438531637191772, + "learning_rate": 0.000301880567469482, + "loss": 0.4105, + "step": 1675 + }, + { + "epoch": 6.8, + "grad_norm": 2.49668288230896, + "learning_rate": 0.0003002309468822171, + "loss": 0.4318, + "step": 1700 + }, + { + "epoch": 6.9, + "grad_norm": 1.5911107063293457, + "learning_rate": 0.0002985813262949522, + "loss": 0.4454, + "step": 1725 + }, + { + "epoch": 7.0, + "grad_norm": 2.9934186935424805, + "learning_rate": 0.00029693170570768723, + "loss": 0.4306, + "step": 1750 + }, + { + "epoch": 7.1, + "grad_norm": 2.5476737022399902, + "learning_rate": 0.00029528208512042233, + "loss": 0.4066, + "step": 1775 + }, + { + "epoch": 7.2, + "grad_norm": 4.207254886627197, + "learning_rate": 0.0002936324645331574, + "loss": 0.4521, + "step": 1800 + }, + { + "epoch": 7.3, + "grad_norm": 1.8378102779388428, + "learning_rate": 0.00029198284394589243, + "loss": 0.4197, + "step": 1825 + }, + { + "epoch": 7.4, + "grad_norm": 2.2129015922546387, + "learning_rate": 0.00029033322335862753, + "loss": 0.4182, + "step": 1850 + }, + { + "epoch": 7.5, + "grad_norm": 1.4666684865951538, + "learning_rate": 0.0002886836027713626, + "loss": 0.4224, + "step": 1875 + }, + { + "epoch": 7.6, + "grad_norm": 2.0311567783355713, + "learning_rate": 0.0002870339821840977, + "loss": 0.4304, + "step": 1900 + }, + { + "epoch": 7.7, + "grad_norm": 2.346604585647583, + "learning_rate": 0.00028538436159683273, + "loss": 0.4012, + "step": 1925 + }, + { + "epoch": 7.8, + "grad_norm": 1.440136432647705, + "learning_rate": 0.00028373474100956784, + "loss": 0.4487, + "step": 1950 + }, + { + "epoch": 7.9, + "grad_norm": 2.426828384399414, + "learning_rate": 0.0002820851204223029, + "loss": 0.3577, + "step": 1975 + }, + { + "epoch": 8.0, + "grad_norm": 3.095980405807495, + "learning_rate": 0.00028043549983503794, + "loss": 0.4121, + "step": 2000 + }, + { + "epoch": 8.1, + "grad_norm": 1.8646572828292847, + "learning_rate": 0.00027878587924777304, + "loss": 0.4213, + "step": 2025 + }, + { + "epoch": 8.2, + "grad_norm": 2.3382062911987305, + "learning_rate": 0.0002771362586605081, + "loss": 0.4119, + "step": 2050 + }, + { + "epoch": 8.3, + "grad_norm": 1.7999392747879028, + "learning_rate": 0.0002754866380732432, + "loss": 0.4165, + "step": 2075 + }, + { + "epoch": 8.4, + "grad_norm": 2.262005567550659, + "learning_rate": 0.00027383701748597824, + "loss": 0.4043, + "step": 2100 + }, + { + "epoch": 8.5, + "grad_norm": 1.8611185550689697, + "learning_rate": 0.0002721873968987133, + "loss": 0.4022, + "step": 2125 + }, + { + "epoch": 8.6, + "grad_norm": 1.3322739601135254, + "learning_rate": 0.00027053777631144834, + "loss": 0.3835, + "step": 2150 + }, + { + "epoch": 8.7, + "grad_norm": 1.335750937461853, + "learning_rate": 0.00026888815572418344, + "loss": 0.424, + "step": 2175 + }, + { + "epoch": 8.8, + "grad_norm": 1.4633246660232544, + "learning_rate": 0.00026723853513691855, + "loss": 0.4124, + "step": 2200 + }, + { + "epoch": 8.9, + "grad_norm": 2.061042308807373, + "learning_rate": 0.0002655889145496536, + "loss": 0.4016, + "step": 2225 + }, + { + "epoch": 9.0, + "grad_norm": 1.9621604681015015, + "learning_rate": 0.0002639392939623887, + "loss": 0.4256, + "step": 2250 + }, + { + "epoch": 9.1, + "grad_norm": 1.4014378786087036, + "learning_rate": 0.0002622896733751237, + "loss": 0.3806, + "step": 2275 + }, + { + "epoch": 9.2, + "grad_norm": 2.6264309883117676, + "learning_rate": 0.0002606400527878588, + "loss": 0.3636, + "step": 2300 + }, + { + "epoch": 9.3, + "grad_norm": 1.8708316087722778, + "learning_rate": 0.00025899043220059385, + "loss": 0.4101, + "step": 2325 + }, + { + "epoch": 9.4, + "grad_norm": 2.156869649887085, + "learning_rate": 0.00025734081161332895, + "loss": 0.4114, + "step": 2350 + }, + { + "epoch": 9.5, + "grad_norm": 2.483182907104492, + "learning_rate": 0.000255691191026064, + "loss": 0.421, + "step": 2375 + }, + { + "epoch": 9.6, + "grad_norm": 2.1955626010894775, + "learning_rate": 0.0002540415704387991, + "loss": 0.3918, + "step": 2400 + }, + { + "epoch": 9.7, + "grad_norm": 4.115762233734131, + "learning_rate": 0.00025239194985153415, + "loss": 0.4172, + "step": 2425 + }, + { + "epoch": 9.8, + "grad_norm": 2.144148349761963, + "learning_rate": 0.0002507423292642692, + "loss": 0.3636, + "step": 2450 + }, + { + "epoch": 9.9, + "grad_norm": 2.340376377105713, + "learning_rate": 0.0002490927086770043, + "loss": 0.4059, + "step": 2475 + }, + { + "epoch": 10.0, + "grad_norm": 1.5438555479049683, + "learning_rate": 0.00024744308808973936, + "loss": 0.4232, + "step": 2500 + }, + { + "epoch": 10.1, + "grad_norm": 2.485889196395874, + "learning_rate": 0.00024579346750247446, + "loss": 0.3861, + "step": 2525 + }, + { + "epoch": 10.2, + "grad_norm": 2.4561119079589844, + "learning_rate": 0.0002441438469152095, + "loss": 0.3565, + "step": 2550 + }, + { + "epoch": 10.3, + "grad_norm": 2.460170269012451, + "learning_rate": 0.00024249422632794456, + "loss": 0.4003, + "step": 2575 + }, + { + "epoch": 10.4, + "grad_norm": 2.774445056915283, + "learning_rate": 0.00024084460574067963, + "loss": 0.4088, + "step": 2600 + }, + { + "epoch": 10.5, + "grad_norm": 2.678745985031128, + "learning_rate": 0.0002391949851534147, + "loss": 0.4021, + "step": 2625 + }, + { + "epoch": 10.6, + "grad_norm": 1.4933565855026245, + "learning_rate": 0.0002375453645661498, + "loss": 0.3744, + "step": 2650 + }, + { + "epoch": 10.7, + "grad_norm": 1.367900013923645, + "learning_rate": 0.00023589574397888486, + "loss": 0.3925, + "step": 2675 + }, + { + "epoch": 10.8, + "grad_norm": 1.653441309928894, + "learning_rate": 0.00023424612339161997, + "loss": 0.4049, + "step": 2700 + }, + { + "epoch": 10.9, + "grad_norm": 1.5892399549484253, + "learning_rate": 0.000232596502804355, + "loss": 0.3927, + "step": 2725 + }, + { + "epoch": 11.0, + "grad_norm": 1.535090684890747, + "learning_rate": 0.00023094688221709007, + "loss": 0.3739, + "step": 2750 + }, + { + "epoch": 11.1, + "grad_norm": 2.247271776199341, + "learning_rate": 0.00022929726162982514, + "loss": 0.3362, + "step": 2775 + }, + { + "epoch": 11.2, + "grad_norm": 2.408069372177124, + "learning_rate": 0.00022764764104256022, + "loss": 0.3933, + "step": 2800 + }, + { + "epoch": 11.3, + "grad_norm": 2.219688653945923, + "learning_rate": 0.0002259980204552953, + "loss": 0.4062, + "step": 2825 + }, + { + "epoch": 11.4, + "grad_norm": 3.0188546180725098, + "learning_rate": 0.00022434839986803037, + "loss": 0.3926, + "step": 2850 + }, + { + "epoch": 11.5, + "grad_norm": 2.8080244064331055, + "learning_rate": 0.00022269877928076542, + "loss": 0.3942, + "step": 2875 + }, + { + "epoch": 11.6, + "grad_norm": 2.263610601425171, + "learning_rate": 0.0002210491586935005, + "loss": 0.3887, + "step": 2900 + }, + { + "epoch": 11.7, + "grad_norm": 2.1239607334136963, + "learning_rate": 0.00021939953810623557, + "loss": 0.3989, + "step": 2925 + }, + { + "epoch": 11.8, + "grad_norm": 3.435351848602295, + "learning_rate": 0.00021774991751897065, + "loss": 0.4185, + "step": 2950 + }, + { + "epoch": 11.9, + "grad_norm": 4.1570515632629395, + "learning_rate": 0.00021610029693170573, + "loss": 0.3597, + "step": 2975 + }, + { + "epoch": 12.0, + "grad_norm": 1.7937941551208496, + "learning_rate": 0.0002144506763444408, + "loss": 0.3719, + "step": 3000 + }, + { + "epoch": 12.1, + "grad_norm": 2.3080639839172363, + "learning_rate": 0.00021280105575717582, + "loss": 0.3856, + "step": 3025 + }, + { + "epoch": 12.2, + "grad_norm": 1.9331200122833252, + "learning_rate": 0.00021115143516991093, + "loss": 0.3798, + "step": 3050 + }, + { + "epoch": 12.3, + "grad_norm": 2.0342001914978027, + "learning_rate": 0.000209501814582646, + "loss": 0.3789, + "step": 3075 + }, + { + "epoch": 12.4, + "grad_norm": 1.7878581285476685, + "learning_rate": 0.00020785219399538108, + "loss": 0.3546, + "step": 3100 + }, + { + "epoch": 12.5, + "grad_norm": 3.657317638397217, + "learning_rate": 0.00020620257340811616, + "loss": 0.3582, + "step": 3125 + }, + { + "epoch": 12.6, + "grad_norm": 2.246429681777954, + "learning_rate": 0.00020455295282085123, + "loss": 0.3991, + "step": 3150 + }, + { + "epoch": 12.7, + "grad_norm": 1.7742749452590942, + "learning_rate": 0.00020290333223358626, + "loss": 0.3914, + "step": 3175 + }, + { + "epoch": 12.8, + "grad_norm": 2.3869900703430176, + "learning_rate": 0.00020125371164632133, + "loss": 0.3435, + "step": 3200 + }, + { + "epoch": 12.9, + "grad_norm": 2.444920539855957, + "learning_rate": 0.0001996040910590564, + "loss": 0.393, + "step": 3225 + }, + { + "epoch": 13.0, + "grad_norm": 2.26476788520813, + "learning_rate": 0.00019795447047179148, + "loss": 0.3828, + "step": 3250 + }, + { + "epoch": 13.1, + "grad_norm": 2.4492876529693604, + "learning_rate": 0.0001963048498845266, + "loss": 0.3856, + "step": 3275 + }, + { + "epoch": 13.2, + "grad_norm": 4.578381061553955, + "learning_rate": 0.00019465522929726164, + "loss": 0.3676, + "step": 3300 + }, + { + "epoch": 13.3, + "grad_norm": 2.0352611541748047, + "learning_rate": 0.00019300560870999671, + "loss": 0.369, + "step": 3325 + }, + { + "epoch": 13.4, + "grad_norm": 2.151038646697998, + "learning_rate": 0.0001913559881227318, + "loss": 0.3713, + "step": 3350 + }, + { + "epoch": 13.5, + "grad_norm": 2.4661877155303955, + "learning_rate": 0.00018970636753546684, + "loss": 0.3859, + "step": 3375 + }, + { + "epoch": 13.6, + "grad_norm": 1.7616342306137085, + "learning_rate": 0.00018805674694820192, + "loss": 0.363, + "step": 3400 + }, + { + "epoch": 13.7, + "grad_norm": 2.26426362991333, + "learning_rate": 0.000186407126360937, + "loss": 0.3696, + "step": 3425 + }, + { + "epoch": 13.8, + "grad_norm": 2.355781078338623, + "learning_rate": 0.00018475750577367207, + "loss": 0.372, + "step": 3450 + }, + { + "epoch": 13.9, + "grad_norm": 1.961585283279419, + "learning_rate": 0.00018310788518640715, + "loss": 0.3418, + "step": 3475 + }, + { + "epoch": 14.0, + "grad_norm": 2.54199481010437, + "learning_rate": 0.00018145826459914222, + "loss": 0.3648, + "step": 3500 + }, + { + "epoch": 14.1, + "grad_norm": 1.5017932653427124, + "learning_rate": 0.00017980864401187727, + "loss": 0.3914, + "step": 3525 + }, + { + "epoch": 14.2, + "grad_norm": 3.051287889480591, + "learning_rate": 0.00017815902342461235, + "loss": 0.3887, + "step": 3550 + }, + { + "epoch": 14.3, + "grad_norm": 2.697949171066284, + "learning_rate": 0.00017650940283734742, + "loss": 0.3524, + "step": 3575 + }, + { + "epoch": 14.4, + "grad_norm": 2.470728635787964, + "learning_rate": 0.00017485978225008247, + "loss": 0.3767, + "step": 3600 + }, + { + "epoch": 14.5, + "grad_norm": 2.5166375637054443, + "learning_rate": 0.00017321016166281755, + "loss": 0.362, + "step": 3625 + }, + { + "epoch": 14.6, + "grad_norm": 1.7630703449249268, + "learning_rate": 0.00017156054107555263, + "loss": 0.3936, + "step": 3650 + }, + { + "epoch": 14.7, + "grad_norm": 2.083225727081299, + "learning_rate": 0.0001699109204882877, + "loss": 0.36, + "step": 3675 + }, + { + "epoch": 14.8, + "grad_norm": 1.7796965837478638, + "learning_rate": 0.00016826129990102278, + "loss": 0.3646, + "step": 3700 + }, + { + "epoch": 14.9, + "grad_norm": 2.127898693084717, + "learning_rate": 0.00016661167931375785, + "loss": 0.3722, + "step": 3725 + }, + { + "epoch": 15.0, + "grad_norm": 2.666217088699341, + "learning_rate": 0.0001649620587264929, + "loss": 0.3329, + "step": 3750 + }, + { + "epoch": 15.1, + "grad_norm": 2.521188497543335, + "learning_rate": 0.00016331243813922798, + "loss": 0.331, + "step": 3775 + }, + { + "epoch": 15.2, + "grad_norm": 1.528865098953247, + "learning_rate": 0.00016166281755196306, + "loss": 0.3513, + "step": 3800 + }, + { + "epoch": 15.3, + "grad_norm": 3.452749490737915, + "learning_rate": 0.00016001319696469813, + "loss": 0.3777, + "step": 3825 + }, + { + "epoch": 15.4, + "grad_norm": 3.0441641807556152, + "learning_rate": 0.0001583635763774332, + "loss": 0.3536, + "step": 3850 + }, + { + "epoch": 15.5, + "grad_norm": 1.9194542169570923, + "learning_rate": 0.00015671395579016829, + "loss": 0.3635, + "step": 3875 + }, + { + "epoch": 15.6, + "grad_norm": 2.011859655380249, + "learning_rate": 0.00015506433520290334, + "loss": 0.3467, + "step": 3900 + }, + { + "epoch": 15.7, + "grad_norm": 1.5169134140014648, + "learning_rate": 0.0001534147146156384, + "loss": 0.3639, + "step": 3925 + }, + { + "epoch": 15.8, + "grad_norm": 2.715104103088379, + "learning_rate": 0.0001517650940283735, + "loss": 0.3473, + "step": 3950 + }, + { + "epoch": 15.9, + "grad_norm": 1.8220083713531494, + "learning_rate": 0.00015011547344110854, + "loss": 0.3902, + "step": 3975 + }, + { + "epoch": 16.0, + "grad_norm": 3.227846384048462, + "learning_rate": 0.00014846585285384361, + "loss": 0.3576, + "step": 4000 + }, + { + "epoch": 16.1, + "grad_norm": 1.9271538257598877, + "learning_rate": 0.0001468162322665787, + "loss": 0.3426, + "step": 4025 + }, + { + "epoch": 16.2, + "grad_norm": 2.58815336227417, + "learning_rate": 0.00014516661167931377, + "loss": 0.3714, + "step": 4050 + }, + { + "epoch": 16.3, + "grad_norm": 1.7768404483795166, + "learning_rate": 0.00014351699109204884, + "loss": 0.3561, + "step": 4075 + }, + { + "epoch": 16.4, + "grad_norm": 1.622894525527954, + "learning_rate": 0.00014186737050478392, + "loss": 0.3384, + "step": 4100 + }, + { + "epoch": 16.5, + "grad_norm": 1.8811430931091309, + "learning_rate": 0.00014021774991751897, + "loss": 0.3603, + "step": 4125 + }, + { + "epoch": 16.6, + "grad_norm": 2.1986522674560547, + "learning_rate": 0.00013856812933025404, + "loss": 0.34, + "step": 4150 + }, + { + "epoch": 16.7, + "grad_norm": 2.3372039794921875, + "learning_rate": 0.00013691850874298912, + "loss": 0.368, + "step": 4175 + }, + { + "epoch": 16.8, + "grad_norm": 1.5386862754821777, + "learning_rate": 0.00013526888815572417, + "loss": 0.3441, + "step": 4200 + }, + { + "epoch": 16.9, + "grad_norm": 1.9392023086547852, + "learning_rate": 0.00013361926756845927, + "loss": 0.3704, + "step": 4225 + }, + { + "epoch": 17.0, + "grad_norm": 3.1489596366882324, + "learning_rate": 0.00013196964698119435, + "loss": 0.3466, + "step": 4250 + }, + { + "epoch": 17.1, + "grad_norm": 3.081573724746704, + "learning_rate": 0.0001303200263939294, + "loss": 0.3352, + "step": 4275 + }, + { + "epoch": 17.2, + "grad_norm": 1.7880629301071167, + "learning_rate": 0.00012867040580666448, + "loss": 0.3569, + "step": 4300 + }, + { + "epoch": 17.3, + "grad_norm": 2.12654972076416, + "learning_rate": 0.00012702078521939955, + "loss": 0.3493, + "step": 4325 + }, + { + "epoch": 17.4, + "grad_norm": 2.646132230758667, + "learning_rate": 0.0001253711646321346, + "loss": 0.346, + "step": 4350 + }, + { + "epoch": 17.5, + "grad_norm": 1.6917959451675415, + "learning_rate": 0.00012372154404486968, + "loss": 0.3544, + "step": 4375 + }, + { + "epoch": 17.6, + "grad_norm": 1.7370896339416504, + "learning_rate": 0.00012207192345760475, + "loss": 0.3392, + "step": 4400 + }, + { + "epoch": 17.7, + "grad_norm": 2.7857675552368164, + "learning_rate": 0.00012042230287033982, + "loss": 0.3479, + "step": 4425 + }, + { + "epoch": 17.8, + "grad_norm": 2.1937146186828613, + "learning_rate": 0.0001187726822830749, + "loss": 0.4039, + "step": 4450 + }, + { + "epoch": 17.9, + "grad_norm": 2.3681347370147705, + "learning_rate": 0.00011712306169580998, + "loss": 0.3534, + "step": 4475 + }, + { + "epoch": 18.0, + "grad_norm": 1.5830249786376953, + "learning_rate": 0.00011547344110854503, + "loss": 0.3599, + "step": 4500 + }, + { + "epoch": 18.1, + "grad_norm": 2.077099084854126, + "learning_rate": 0.00011382382052128011, + "loss": 0.3369, + "step": 4525 + }, + { + "epoch": 18.2, + "grad_norm": 2.1928341388702393, + "learning_rate": 0.00011217419993401519, + "loss": 0.3525, + "step": 4550 + }, + { + "epoch": 18.3, + "grad_norm": 1.9083303213119507, + "learning_rate": 0.00011052457934675025, + "loss": 0.3781, + "step": 4575 + }, + { + "epoch": 18.4, + "grad_norm": 1.9660146236419678, + "learning_rate": 0.00010887495875948532, + "loss": 0.3589, + "step": 4600 + }, + { + "epoch": 18.5, + "grad_norm": 1.6620488166809082, + "learning_rate": 0.0001072253381722204, + "loss": 0.3571, + "step": 4625 + }, + { + "epoch": 18.6, + "grad_norm": 2.4148569107055664, + "learning_rate": 0.00010557571758495546, + "loss": 0.3718, + "step": 4650 + }, + { + "epoch": 18.7, + "grad_norm": 2.578110933303833, + "learning_rate": 0.00010392609699769054, + "loss": 0.3491, + "step": 4675 + }, + { + "epoch": 18.8, + "grad_norm": 2.4224448204040527, + "learning_rate": 0.00010227647641042562, + "loss": 0.3241, + "step": 4700 + }, + { + "epoch": 18.9, + "grad_norm": 1.9691663980484009, + "learning_rate": 0.00010062685582316067, + "loss": 0.3407, + "step": 4725 + }, + { + "epoch": 19.0, + "grad_norm": 1.8373006582260132, + "learning_rate": 9.897723523589574e-05, + "loss": 0.3415, + "step": 4750 + }, + { + "epoch": 19.1, + "grad_norm": 2.405202865600586, + "learning_rate": 9.732761464863082e-05, + "loss": 0.3259, + "step": 4775 + }, + { + "epoch": 19.2, + "grad_norm": 2.754739761352539, + "learning_rate": 9.56779940613659e-05, + "loss": 0.3361, + "step": 4800 + }, + { + "epoch": 19.3, + "grad_norm": 2.4953665733337402, + "learning_rate": 9.402837347410096e-05, + "loss": 0.3534, + "step": 4825 + }, + { + "epoch": 19.4, + "grad_norm": 2.974691390991211, + "learning_rate": 9.237875288683603e-05, + "loss": 0.3742, + "step": 4850 + }, + { + "epoch": 19.5, + "grad_norm": 1.4064937829971313, + "learning_rate": 9.072913229957111e-05, + "loss": 0.3271, + "step": 4875 + }, + { + "epoch": 19.6, + "grad_norm": 3.1546576023101807, + "learning_rate": 8.907951171230617e-05, + "loss": 0.3549, + "step": 4900 + }, + { + "epoch": 19.7, + "grad_norm": 1.948807716369629, + "learning_rate": 8.742989112504124e-05, + "loss": 0.3107, + "step": 4925 + }, + { + "epoch": 19.8, + "grad_norm": 1.1512947082519531, + "learning_rate": 8.578027053777631e-05, + "loss": 0.3448, + "step": 4950 + }, + { + "epoch": 19.9, + "grad_norm": 3.7861363887786865, + "learning_rate": 8.413064995051139e-05, + "loss": 0.3312, + "step": 4975 + }, + { + "epoch": 20.0, + "grad_norm": 1.6949777603149414, + "learning_rate": 8.248102936324645e-05, + "loss": 0.3532, + "step": 5000 + }, + { + "epoch": 20.1, + "grad_norm": 2.2876861095428467, + "learning_rate": 8.083140877598153e-05, + "loss": 0.3628, + "step": 5025 + }, + { + "epoch": 20.2, + "grad_norm": 1.8852239847183228, + "learning_rate": 7.91817881887166e-05, + "loss": 0.3595, + "step": 5050 + }, + { + "epoch": 20.3, + "grad_norm": 2.6525633335113525, + "learning_rate": 7.753216760145167e-05, + "loss": 0.345, + "step": 5075 + }, + { + "epoch": 20.4, + "grad_norm": 3.196200370788574, + "learning_rate": 7.588254701418674e-05, + "loss": 0.3442, + "step": 5100 + }, + { + "epoch": 20.5, + "grad_norm": 4.056946277618408, + "learning_rate": 7.423292642692181e-05, + "loss": 0.3232, + "step": 5125 + }, + { + "epoch": 20.6, + "grad_norm": 2.9283547401428223, + "learning_rate": 7.258330583965688e-05, + "loss": 0.3606, + "step": 5150 + }, + { + "epoch": 20.7, + "grad_norm": 2.2991044521331787, + "learning_rate": 7.093368525239196e-05, + "loss": 0.3595, + "step": 5175 + }, + { + "epoch": 20.8, + "grad_norm": 2.3510475158691406, + "learning_rate": 6.928406466512702e-05, + "loss": 0.3406, + "step": 5200 + }, + { + "epoch": 20.9, + "grad_norm": 1.8012481927871704, + "learning_rate": 6.763444407786209e-05, + "loss": 0.3304, + "step": 5225 + }, + { + "epoch": 21.0, + "grad_norm": 2.620314121246338, + "learning_rate": 6.598482349059718e-05, + "loss": 0.3572, + "step": 5250 + }, + { + "epoch": 21.1, + "grad_norm": 2.1940581798553467, + "learning_rate": 6.433520290333224e-05, + "loss": 0.332, + "step": 5275 + }, + { + "epoch": 21.2, + "grad_norm": 1.6455153226852417, + "learning_rate": 6.26855823160673e-05, + "loss": 0.3329, + "step": 5300 + }, + { + "epoch": 21.3, + "grad_norm": 2.369469165802002, + "learning_rate": 6.103596172880238e-05, + "loss": 0.3513, + "step": 5325 + }, + { + "epoch": 21.4, + "grad_norm": 2.4541404247283936, + "learning_rate": 5.938634114153745e-05, + "loss": 0.321, + "step": 5350 + }, + { + "epoch": 21.5, + "grad_norm": 2.259453296661377, + "learning_rate": 5.7736720554272516e-05, + "loss": 0.3589, + "step": 5375 + }, + { + "epoch": 21.6, + "grad_norm": 2.1575701236724854, + "learning_rate": 5.608709996700759e-05, + "loss": 0.3474, + "step": 5400 + }, + { + "epoch": 21.7, + "grad_norm": 1.968269944190979, + "learning_rate": 5.443747937974266e-05, + "loss": 0.3738, + "step": 5425 + }, + { + "epoch": 21.8, + "grad_norm": 1.7775858640670776, + "learning_rate": 5.278785879247773e-05, + "loss": 0.3471, + "step": 5450 + }, + { + "epoch": 21.9, + "grad_norm": 2.118597984313965, + "learning_rate": 5.113823820521281e-05, + "loss": 0.338, + "step": 5475 + }, + { + "epoch": 22.0, + "grad_norm": 2.509803056716919, + "learning_rate": 4.948861761794787e-05, + "loss": 0.3561, + "step": 5500 + }, + { + "epoch": 22.1, + "grad_norm": 1.1670511960983276, + "learning_rate": 4.783899703068295e-05, + "loss": 0.3304, + "step": 5525 + }, + { + "epoch": 22.2, + "grad_norm": 2.318145751953125, + "learning_rate": 4.618937644341802e-05, + "loss": 0.3776, + "step": 5550 + }, + { + "epoch": 22.3, + "grad_norm": 1.859100103378296, + "learning_rate": 4.453975585615309e-05, + "loss": 0.3171, + "step": 5575 + }, + { + "epoch": 22.4, + "grad_norm": 3.5954763889312744, + "learning_rate": 4.2890135268888156e-05, + "loss": 0.3686, + "step": 5600 + }, + { + "epoch": 22.5, + "grad_norm": 2.928518533706665, + "learning_rate": 4.1240514681623226e-05, + "loss": 0.4016, + "step": 5625 + }, + { + "epoch": 22.6, + "grad_norm": 2.325444221496582, + "learning_rate": 3.95908940943583e-05, + "loss": 0.3316, + "step": 5650 + }, + { + "epoch": 22.7, + "grad_norm": 2.2691650390625, + "learning_rate": 3.794127350709337e-05, + "loss": 0.3249, + "step": 5675 + }, + { + "epoch": 22.8, + "grad_norm": 2.954089879989624, + "learning_rate": 3.629165291982844e-05, + "loss": 0.3015, + "step": 5700 + }, + { + "epoch": 22.9, + "grad_norm": 1.7389692068099976, + "learning_rate": 3.464203233256351e-05, + "loss": 0.342, + "step": 5725 + }, + { + "epoch": 23.0, + "grad_norm": 2.889202356338501, + "learning_rate": 3.299241174529859e-05, + "loss": 0.3314, + "step": 5750 + }, + { + "epoch": 23.1, + "grad_norm": 2.633880138397217, + "learning_rate": 3.134279115803365e-05, + "loss": 0.307, + "step": 5775 + }, + { + "epoch": 23.2, + "grad_norm": 3.491842746734619, + "learning_rate": 2.9693170570768723e-05, + "loss": 0.3791, + "step": 5800 + }, + { + "epoch": 23.3, + "grad_norm": 2.2318499088287354, + "learning_rate": 2.8043549983503796e-05, + "loss": 0.3381, + "step": 5825 + }, + { + "epoch": 23.4, + "grad_norm": 1.6060622930526733, + "learning_rate": 2.6393929396238866e-05, + "loss": 0.3374, + "step": 5850 + }, + { + "epoch": 23.5, + "grad_norm": 4.242088317871094, + "learning_rate": 2.4744308808973936e-05, + "loss": 0.3234, + "step": 5875 + }, + { + "epoch": 23.6, + "grad_norm": 1.861912727355957, + "learning_rate": 2.309468822170901e-05, + "loss": 0.3507, + "step": 5900 + }, + { + "epoch": 23.7, + "grad_norm": 2.4759881496429443, + "learning_rate": 2.1445067634444078e-05, + "loss": 0.3562, + "step": 5925 + }, + { + "epoch": 23.8, + "grad_norm": 2.1233272552490234, + "learning_rate": 1.979544704717915e-05, + "loss": 0.3397, + "step": 5950 + }, + { + "epoch": 23.9, + "grad_norm": 2.3019630908966064, + "learning_rate": 1.814582645991422e-05, + "loss": 0.3103, + "step": 5975 + }, + { + "epoch": 24.0, + "grad_norm": 2.005906105041504, + "learning_rate": 1.6496205872649294e-05, + "loss": 0.3524, + "step": 6000 + }, + { + "epoch": 24.1, + "grad_norm": 3.350837230682373, + "learning_rate": 1.4846585285384362e-05, + "loss": 0.3272, + "step": 6025 + }, + { + "epoch": 24.2, + "grad_norm": 1.622349500656128, + "learning_rate": 1.3196964698119433e-05, + "loss": 0.31, + "step": 6050 + }, + { + "epoch": 24.3, + "grad_norm": 2.5132482051849365, + "learning_rate": 1.1547344110854504e-05, + "loss": 0.3683, + "step": 6075 + }, + { + "epoch": 24.4, + "grad_norm": 3.2578160762786865, + "learning_rate": 9.897723523589576e-06, + "loss": 0.3238, + "step": 6100 + }, + { + "epoch": 24.5, + "grad_norm": 2.5393242835998535, + "learning_rate": 8.248102936324647e-06, + "loss": 0.3392, + "step": 6125 + }, + { + "epoch": 24.6, + "grad_norm": 3.6766817569732666, + "learning_rate": 6.5984823490597165e-06, + "loss": 0.3343, + "step": 6150 + }, + { + "epoch": 24.7, + "grad_norm": 1.9477918148040771, + "learning_rate": 4.948861761794788e-06, + "loss": 0.3423, + "step": 6175 + }, + { + "epoch": 24.8, + "grad_norm": 3.434939384460449, + "learning_rate": 3.2992411745298582e-06, + "loss": 0.3603, + "step": 6200 + }, + { + "epoch": 24.9, + "grad_norm": 2.9802677631378174, + "learning_rate": 1.6496205872649291e-06, + "loss": 0.3046, + "step": 6225 + }, + { + "epoch": 25.0, + "grad_norm": 1.9037508964538574, + "learning_rate": 0.0, + "loss": 0.3437, + "step": 6250 + } + ], + "logging_steps": 25, + "max_steps": 6250, + "num_input_tokens_seen": 0, + "num_train_epochs": 25, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 346131456000000.0, + "train_batch_size": 20, + "trial_name": null, + "trial_params": null +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/training_args.bin b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..10d16bb8e9df9c51aba38e7a6f04deeecf1ea29f --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:662a5f25d38ccae34f19d85e24456de6d3eaaabafca26fb7cc8816de0c990f20 +size 5304 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/vocab.txt b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/vocab.txt new file mode 100644 index 0000000000000000000000000000000000000000..cc9ad53732e21ea9180f18e1fb67a38cc5702d4c --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/vocab.txt @@ -0,0 +1,4101 @@ +[PAD] +[UNK] +[CLS] +[SEP] +[MASK] +AAAAAA +AAAAAT +AAAAAC +AAAAAG +AAAATA +AAAATT +AAAATC +AAAATG +AAAACA +AAAACT +AAAACC +AAAACG +AAAAGA +AAAAGT +AAAAGC +AAAAGG +AAATAA +AAATAT +AAATAC +AAATAG +AAATTA +AAATTT +AAATTC +AAATTG +AAATCA +AAATCT +AAATCC +AAATCG +AAATGA +AAATGT +AAATGC +AAATGG +AAACAA +AAACAT +AAACAC +AAACAG +AAACTA +AAACTT +AAACTC +AAACTG +AAACCA +AAACCT +AAACCC +AAACCG +AAACGA +AAACGT +AAACGC +AAACGG +AAAGAA +AAAGAT +AAAGAC +AAAGAG +AAAGTA +AAAGTT +AAAGTC +AAAGTG +AAAGCA +AAAGCT +AAAGCC +AAAGCG +AAAGGA +AAAGGT +AAAGGC +AAAGGG +AATAAA +AATAAT +AATAAC +AATAAG +AATATA +AATATT +AATATC +AATATG +AATACA +AATACT +AATACC +AATACG +AATAGA +AATAGT +AATAGC +AATAGG +AATTAA +AATTAT +AATTAC +AATTAG +AATTTA +AATTTT +AATTTC +AATTTG +AATTCA +AATTCT +AATTCC +AATTCG +AATTGA +AATTGT +AATTGC +AATTGG +AATCAA +AATCAT +AATCAC +AATCAG +AATCTA +AATCTT +AATCTC +AATCTG +AATCCA +AATCCT +AATCCC +AATCCG +AATCGA +AATCGT +AATCGC +AATCGG +AATGAA +AATGAT +AATGAC +AATGAG +AATGTA +AATGTT +AATGTC +AATGTG +AATGCA +AATGCT +AATGCC +AATGCG +AATGGA +AATGGT +AATGGC +AATGGG +AACAAA +AACAAT +AACAAC +AACAAG +AACATA +AACATT +AACATC +AACATG +AACACA +AACACT +AACACC +AACACG +AACAGA +AACAGT +AACAGC +AACAGG +AACTAA +AACTAT +AACTAC +AACTAG +AACTTA +AACTTT +AACTTC +AACTTG +AACTCA +AACTCT +AACTCC +AACTCG +AACTGA +AACTGT +AACTGC +AACTGG +AACCAA +AACCAT +AACCAC +AACCAG +AACCTA +AACCTT +AACCTC +AACCTG +AACCCA +AACCCT +AACCCC +AACCCG +AACCGA +AACCGT +AACCGC +AACCGG +AACGAA +AACGAT +AACGAC +AACGAG +AACGTA +AACGTT +AACGTC +AACGTG +AACGCA +AACGCT +AACGCC +AACGCG +AACGGA +AACGGT +AACGGC +AACGGG +AAGAAA +AAGAAT +AAGAAC +AAGAAG +AAGATA +AAGATT +AAGATC +AAGATG +AAGACA +AAGACT +AAGACC +AAGACG +AAGAGA +AAGAGT +AAGAGC +AAGAGG +AAGTAA +AAGTAT +AAGTAC +AAGTAG +AAGTTA +AAGTTT +AAGTTC +AAGTTG +AAGTCA +AAGTCT +AAGTCC +AAGTCG +AAGTGA +AAGTGT +AAGTGC +AAGTGG +AAGCAA +AAGCAT +AAGCAC +AAGCAG +AAGCTA +AAGCTT +AAGCTC +AAGCTG +AAGCCA +AAGCCT +AAGCCC +AAGCCG +AAGCGA +AAGCGT +AAGCGC +AAGCGG +AAGGAA +AAGGAT +AAGGAC +AAGGAG +AAGGTA +AAGGTT +AAGGTC +AAGGTG +AAGGCA +AAGGCT +AAGGCC +AAGGCG +AAGGGA +AAGGGT +AAGGGC +AAGGGG +ATAAAA +ATAAAT +ATAAAC +ATAAAG +ATAATA +ATAATT +ATAATC +ATAATG +ATAACA +ATAACT +ATAACC +ATAACG +ATAAGA +ATAAGT +ATAAGC +ATAAGG +ATATAA +ATATAT +ATATAC +ATATAG +ATATTA +ATATTT +ATATTC +ATATTG +ATATCA +ATATCT +ATATCC +ATATCG +ATATGA +ATATGT +ATATGC +ATATGG +ATACAA +ATACAT +ATACAC +ATACAG +ATACTA +ATACTT +ATACTC +ATACTG +ATACCA +ATACCT +ATACCC +ATACCG +ATACGA +ATACGT +ATACGC +ATACGG +ATAGAA +ATAGAT +ATAGAC +ATAGAG +ATAGTA +ATAGTT +ATAGTC +ATAGTG +ATAGCA +ATAGCT +ATAGCC +ATAGCG +ATAGGA +ATAGGT +ATAGGC +ATAGGG +ATTAAA +ATTAAT +ATTAAC +ATTAAG +ATTATA +ATTATT +ATTATC +ATTATG +ATTACA +ATTACT +ATTACC +ATTACG +ATTAGA +ATTAGT +ATTAGC +ATTAGG +ATTTAA +ATTTAT +ATTTAC +ATTTAG +ATTTTA +ATTTTT +ATTTTC +ATTTTG +ATTTCA +ATTTCT +ATTTCC +ATTTCG +ATTTGA +ATTTGT +ATTTGC +ATTTGG +ATTCAA +ATTCAT +ATTCAC +ATTCAG +ATTCTA +ATTCTT +ATTCTC +ATTCTG +ATTCCA +ATTCCT +ATTCCC +ATTCCG +ATTCGA +ATTCGT +ATTCGC +ATTCGG +ATTGAA +ATTGAT +ATTGAC +ATTGAG +ATTGTA +ATTGTT +ATTGTC +ATTGTG +ATTGCA +ATTGCT +ATTGCC +ATTGCG +ATTGGA +ATTGGT +ATTGGC +ATTGGG +ATCAAA +ATCAAT +ATCAAC +ATCAAG +ATCATA +ATCATT +ATCATC +ATCATG +ATCACA +ATCACT +ATCACC +ATCACG +ATCAGA +ATCAGT +ATCAGC +ATCAGG +ATCTAA +ATCTAT +ATCTAC +ATCTAG +ATCTTA +ATCTTT +ATCTTC +ATCTTG +ATCTCA +ATCTCT +ATCTCC +ATCTCG +ATCTGA +ATCTGT +ATCTGC +ATCTGG +ATCCAA +ATCCAT +ATCCAC +ATCCAG +ATCCTA +ATCCTT +ATCCTC +ATCCTG +ATCCCA +ATCCCT +ATCCCC +ATCCCG +ATCCGA +ATCCGT +ATCCGC +ATCCGG +ATCGAA +ATCGAT +ATCGAC +ATCGAG +ATCGTA +ATCGTT +ATCGTC +ATCGTG +ATCGCA +ATCGCT +ATCGCC +ATCGCG +ATCGGA +ATCGGT +ATCGGC +ATCGGG +ATGAAA +ATGAAT +ATGAAC +ATGAAG +ATGATA +ATGATT +ATGATC +ATGATG +ATGACA +ATGACT +ATGACC +ATGACG +ATGAGA +ATGAGT +ATGAGC +ATGAGG +ATGTAA +ATGTAT +ATGTAC +ATGTAG +ATGTTA +ATGTTT +ATGTTC +ATGTTG +ATGTCA +ATGTCT +ATGTCC +ATGTCG +ATGTGA +ATGTGT +ATGTGC +ATGTGG +ATGCAA +ATGCAT +ATGCAC +ATGCAG +ATGCTA +ATGCTT +ATGCTC +ATGCTG +ATGCCA +ATGCCT +ATGCCC +ATGCCG +ATGCGA +ATGCGT +ATGCGC +ATGCGG +ATGGAA +ATGGAT +ATGGAC +ATGGAG +ATGGTA +ATGGTT +ATGGTC +ATGGTG +ATGGCA +ATGGCT +ATGGCC +ATGGCG +ATGGGA +ATGGGT +ATGGGC +ATGGGG +ACAAAA +ACAAAT +ACAAAC +ACAAAG +ACAATA +ACAATT +ACAATC +ACAATG +ACAACA +ACAACT +ACAACC +ACAACG +ACAAGA +ACAAGT +ACAAGC +ACAAGG +ACATAA +ACATAT +ACATAC +ACATAG +ACATTA +ACATTT +ACATTC +ACATTG +ACATCA +ACATCT +ACATCC +ACATCG +ACATGA +ACATGT +ACATGC +ACATGG +ACACAA +ACACAT +ACACAC +ACACAG +ACACTA +ACACTT +ACACTC +ACACTG +ACACCA +ACACCT +ACACCC +ACACCG +ACACGA +ACACGT +ACACGC +ACACGG +ACAGAA +ACAGAT +ACAGAC +ACAGAG +ACAGTA +ACAGTT +ACAGTC +ACAGTG +ACAGCA +ACAGCT +ACAGCC +ACAGCG +ACAGGA +ACAGGT +ACAGGC +ACAGGG +ACTAAA +ACTAAT +ACTAAC +ACTAAG +ACTATA +ACTATT +ACTATC +ACTATG +ACTACA +ACTACT +ACTACC +ACTACG +ACTAGA +ACTAGT +ACTAGC +ACTAGG +ACTTAA +ACTTAT +ACTTAC +ACTTAG +ACTTTA +ACTTTT +ACTTTC +ACTTTG +ACTTCA +ACTTCT +ACTTCC +ACTTCG +ACTTGA +ACTTGT +ACTTGC +ACTTGG +ACTCAA +ACTCAT +ACTCAC +ACTCAG +ACTCTA +ACTCTT +ACTCTC +ACTCTG +ACTCCA +ACTCCT +ACTCCC +ACTCCG +ACTCGA +ACTCGT +ACTCGC +ACTCGG +ACTGAA +ACTGAT +ACTGAC +ACTGAG +ACTGTA +ACTGTT +ACTGTC +ACTGTG +ACTGCA +ACTGCT +ACTGCC +ACTGCG +ACTGGA +ACTGGT +ACTGGC +ACTGGG +ACCAAA +ACCAAT +ACCAAC +ACCAAG +ACCATA +ACCATT +ACCATC +ACCATG +ACCACA +ACCACT +ACCACC +ACCACG +ACCAGA +ACCAGT +ACCAGC +ACCAGG +ACCTAA +ACCTAT +ACCTAC +ACCTAG +ACCTTA +ACCTTT +ACCTTC +ACCTTG +ACCTCA +ACCTCT +ACCTCC +ACCTCG +ACCTGA +ACCTGT +ACCTGC +ACCTGG +ACCCAA +ACCCAT +ACCCAC +ACCCAG +ACCCTA +ACCCTT +ACCCTC +ACCCTG +ACCCCA +ACCCCT +ACCCCC +ACCCCG +ACCCGA +ACCCGT +ACCCGC +ACCCGG +ACCGAA +ACCGAT +ACCGAC +ACCGAG +ACCGTA +ACCGTT +ACCGTC +ACCGTG +ACCGCA +ACCGCT +ACCGCC +ACCGCG +ACCGGA +ACCGGT +ACCGGC +ACCGGG +ACGAAA +ACGAAT +ACGAAC +ACGAAG +ACGATA +ACGATT +ACGATC +ACGATG +ACGACA +ACGACT +ACGACC +ACGACG +ACGAGA +ACGAGT +ACGAGC +ACGAGG +ACGTAA +ACGTAT +ACGTAC +ACGTAG +ACGTTA +ACGTTT +ACGTTC +ACGTTG +ACGTCA +ACGTCT +ACGTCC +ACGTCG +ACGTGA +ACGTGT +ACGTGC +ACGTGG +ACGCAA +ACGCAT +ACGCAC +ACGCAG +ACGCTA +ACGCTT +ACGCTC +ACGCTG +ACGCCA +ACGCCT +ACGCCC +ACGCCG +ACGCGA +ACGCGT +ACGCGC +ACGCGG +ACGGAA +ACGGAT +ACGGAC +ACGGAG +ACGGTA +ACGGTT +ACGGTC +ACGGTG +ACGGCA +ACGGCT +ACGGCC +ACGGCG +ACGGGA +ACGGGT +ACGGGC +ACGGGG +AGAAAA +AGAAAT +AGAAAC +AGAAAG +AGAATA +AGAATT +AGAATC +AGAATG +AGAACA +AGAACT +AGAACC +AGAACG +AGAAGA +AGAAGT +AGAAGC +AGAAGG +AGATAA +AGATAT +AGATAC +AGATAG +AGATTA +AGATTT +AGATTC +AGATTG +AGATCA +AGATCT +AGATCC +AGATCG +AGATGA +AGATGT +AGATGC +AGATGG +AGACAA +AGACAT +AGACAC +AGACAG +AGACTA +AGACTT +AGACTC +AGACTG +AGACCA +AGACCT +AGACCC +AGACCG +AGACGA +AGACGT +AGACGC +AGACGG +AGAGAA +AGAGAT +AGAGAC +AGAGAG +AGAGTA +AGAGTT +AGAGTC +AGAGTG +AGAGCA +AGAGCT +AGAGCC +AGAGCG +AGAGGA +AGAGGT +AGAGGC +AGAGGG +AGTAAA +AGTAAT +AGTAAC +AGTAAG +AGTATA +AGTATT +AGTATC +AGTATG +AGTACA +AGTACT +AGTACC +AGTACG +AGTAGA +AGTAGT +AGTAGC +AGTAGG +AGTTAA +AGTTAT +AGTTAC +AGTTAG +AGTTTA +AGTTTT +AGTTTC +AGTTTG +AGTTCA +AGTTCT +AGTTCC +AGTTCG +AGTTGA +AGTTGT +AGTTGC +AGTTGG +AGTCAA +AGTCAT +AGTCAC +AGTCAG +AGTCTA +AGTCTT +AGTCTC +AGTCTG +AGTCCA +AGTCCT +AGTCCC +AGTCCG +AGTCGA +AGTCGT +AGTCGC +AGTCGG +AGTGAA +AGTGAT +AGTGAC +AGTGAG +AGTGTA +AGTGTT +AGTGTC +AGTGTG +AGTGCA +AGTGCT +AGTGCC +AGTGCG +AGTGGA +AGTGGT +AGTGGC +AGTGGG +AGCAAA +AGCAAT +AGCAAC +AGCAAG +AGCATA +AGCATT +AGCATC +AGCATG +AGCACA +AGCACT +AGCACC +AGCACG +AGCAGA +AGCAGT +AGCAGC +AGCAGG +AGCTAA +AGCTAT +AGCTAC +AGCTAG +AGCTTA +AGCTTT +AGCTTC +AGCTTG +AGCTCA +AGCTCT +AGCTCC +AGCTCG +AGCTGA +AGCTGT +AGCTGC +AGCTGG +AGCCAA +AGCCAT +AGCCAC +AGCCAG +AGCCTA +AGCCTT +AGCCTC +AGCCTG +AGCCCA +AGCCCT +AGCCCC +AGCCCG +AGCCGA +AGCCGT +AGCCGC +AGCCGG +AGCGAA +AGCGAT +AGCGAC +AGCGAG +AGCGTA +AGCGTT +AGCGTC +AGCGTG +AGCGCA +AGCGCT +AGCGCC +AGCGCG +AGCGGA +AGCGGT +AGCGGC +AGCGGG +AGGAAA +AGGAAT +AGGAAC +AGGAAG +AGGATA +AGGATT +AGGATC +AGGATG +AGGACA +AGGACT +AGGACC +AGGACG +AGGAGA +AGGAGT +AGGAGC +AGGAGG +AGGTAA +AGGTAT +AGGTAC +AGGTAG +AGGTTA +AGGTTT +AGGTTC +AGGTTG +AGGTCA +AGGTCT +AGGTCC +AGGTCG +AGGTGA +AGGTGT +AGGTGC +AGGTGG +AGGCAA +AGGCAT +AGGCAC +AGGCAG +AGGCTA +AGGCTT +AGGCTC +AGGCTG +AGGCCA +AGGCCT +AGGCCC +AGGCCG +AGGCGA +AGGCGT +AGGCGC +AGGCGG +AGGGAA +AGGGAT +AGGGAC +AGGGAG +AGGGTA +AGGGTT +AGGGTC +AGGGTG +AGGGCA +AGGGCT +AGGGCC +AGGGCG +AGGGGA +AGGGGT +AGGGGC +AGGGGG +TAAAAA +TAAAAT +TAAAAC +TAAAAG +TAAATA +TAAATT +TAAATC +TAAATG +TAAACA +TAAACT +TAAACC +TAAACG +TAAAGA +TAAAGT +TAAAGC +TAAAGG +TAATAA +TAATAT +TAATAC +TAATAG +TAATTA +TAATTT +TAATTC +TAATTG +TAATCA +TAATCT +TAATCC +TAATCG +TAATGA +TAATGT +TAATGC +TAATGG +TAACAA +TAACAT +TAACAC +TAACAG +TAACTA +TAACTT +TAACTC +TAACTG +TAACCA +TAACCT +TAACCC +TAACCG +TAACGA +TAACGT +TAACGC +TAACGG +TAAGAA +TAAGAT +TAAGAC +TAAGAG +TAAGTA +TAAGTT +TAAGTC +TAAGTG +TAAGCA +TAAGCT +TAAGCC +TAAGCG +TAAGGA +TAAGGT +TAAGGC +TAAGGG +TATAAA +TATAAT +TATAAC +TATAAG +TATATA +TATATT +TATATC +TATATG +TATACA +TATACT +TATACC +TATACG +TATAGA +TATAGT +TATAGC +TATAGG +TATTAA +TATTAT +TATTAC +TATTAG +TATTTA +TATTTT +TATTTC +TATTTG +TATTCA +TATTCT +TATTCC +TATTCG +TATTGA +TATTGT +TATTGC +TATTGG +TATCAA +TATCAT +TATCAC +TATCAG +TATCTA +TATCTT +TATCTC +TATCTG +TATCCA +TATCCT +TATCCC +TATCCG +TATCGA +TATCGT +TATCGC +TATCGG +TATGAA +TATGAT +TATGAC +TATGAG +TATGTA +TATGTT +TATGTC +TATGTG +TATGCA +TATGCT +TATGCC +TATGCG +TATGGA +TATGGT +TATGGC +TATGGG +TACAAA +TACAAT +TACAAC +TACAAG +TACATA +TACATT +TACATC +TACATG +TACACA +TACACT +TACACC +TACACG +TACAGA +TACAGT +TACAGC +TACAGG +TACTAA +TACTAT +TACTAC +TACTAG +TACTTA +TACTTT +TACTTC +TACTTG +TACTCA +TACTCT +TACTCC +TACTCG +TACTGA +TACTGT +TACTGC +TACTGG +TACCAA +TACCAT +TACCAC +TACCAG +TACCTA +TACCTT +TACCTC +TACCTG +TACCCA +TACCCT +TACCCC +TACCCG +TACCGA +TACCGT +TACCGC +TACCGG +TACGAA +TACGAT +TACGAC +TACGAG +TACGTA +TACGTT +TACGTC +TACGTG +TACGCA +TACGCT +TACGCC +TACGCG +TACGGA +TACGGT +TACGGC +TACGGG +TAGAAA +TAGAAT +TAGAAC +TAGAAG +TAGATA +TAGATT +TAGATC +TAGATG +TAGACA +TAGACT +TAGACC +TAGACG +TAGAGA +TAGAGT +TAGAGC +TAGAGG +TAGTAA +TAGTAT +TAGTAC +TAGTAG +TAGTTA +TAGTTT +TAGTTC +TAGTTG +TAGTCA +TAGTCT +TAGTCC +TAGTCG +TAGTGA +TAGTGT +TAGTGC +TAGTGG +TAGCAA +TAGCAT +TAGCAC +TAGCAG +TAGCTA +TAGCTT +TAGCTC +TAGCTG +TAGCCA +TAGCCT +TAGCCC +TAGCCG +TAGCGA +TAGCGT +TAGCGC +TAGCGG +TAGGAA +TAGGAT +TAGGAC +TAGGAG +TAGGTA +TAGGTT +TAGGTC +TAGGTG +TAGGCA +TAGGCT +TAGGCC +TAGGCG +TAGGGA +TAGGGT +TAGGGC +TAGGGG +TTAAAA +TTAAAT +TTAAAC +TTAAAG +TTAATA +TTAATT +TTAATC +TTAATG +TTAACA +TTAACT +TTAACC +TTAACG +TTAAGA +TTAAGT +TTAAGC +TTAAGG +TTATAA +TTATAT +TTATAC +TTATAG +TTATTA +TTATTT +TTATTC +TTATTG +TTATCA +TTATCT +TTATCC +TTATCG +TTATGA +TTATGT +TTATGC +TTATGG +TTACAA +TTACAT +TTACAC +TTACAG +TTACTA +TTACTT +TTACTC +TTACTG +TTACCA +TTACCT +TTACCC +TTACCG +TTACGA +TTACGT +TTACGC +TTACGG +TTAGAA +TTAGAT +TTAGAC +TTAGAG +TTAGTA +TTAGTT +TTAGTC +TTAGTG +TTAGCA +TTAGCT +TTAGCC +TTAGCG +TTAGGA +TTAGGT +TTAGGC +TTAGGG +TTTAAA +TTTAAT +TTTAAC +TTTAAG +TTTATA +TTTATT +TTTATC +TTTATG +TTTACA +TTTACT +TTTACC +TTTACG +TTTAGA +TTTAGT +TTTAGC +TTTAGG +TTTTAA +TTTTAT +TTTTAC +TTTTAG +TTTTTA +TTTTTT +TTTTTC +TTTTTG +TTTTCA +TTTTCT +TTTTCC +TTTTCG +TTTTGA +TTTTGT +TTTTGC +TTTTGG +TTTCAA +TTTCAT +TTTCAC +TTTCAG +TTTCTA +TTTCTT +TTTCTC +TTTCTG +TTTCCA +TTTCCT +TTTCCC +TTTCCG +TTTCGA +TTTCGT +TTTCGC +TTTCGG +TTTGAA +TTTGAT +TTTGAC +TTTGAG +TTTGTA +TTTGTT +TTTGTC +TTTGTG +TTTGCA +TTTGCT +TTTGCC +TTTGCG +TTTGGA +TTTGGT +TTTGGC +TTTGGG +TTCAAA +TTCAAT +TTCAAC +TTCAAG +TTCATA +TTCATT +TTCATC +TTCATG +TTCACA +TTCACT +TTCACC +TTCACG +TTCAGA +TTCAGT +TTCAGC +TTCAGG +TTCTAA +TTCTAT +TTCTAC +TTCTAG +TTCTTA +TTCTTT +TTCTTC +TTCTTG +TTCTCA +TTCTCT +TTCTCC +TTCTCG +TTCTGA +TTCTGT +TTCTGC +TTCTGG +TTCCAA +TTCCAT +TTCCAC +TTCCAG +TTCCTA +TTCCTT +TTCCTC +TTCCTG +TTCCCA +TTCCCT +TTCCCC +TTCCCG +TTCCGA +TTCCGT +TTCCGC +TTCCGG +TTCGAA +TTCGAT +TTCGAC +TTCGAG +TTCGTA +TTCGTT +TTCGTC +TTCGTG +TTCGCA +TTCGCT +TTCGCC +TTCGCG +TTCGGA +TTCGGT +TTCGGC +TTCGGG +TTGAAA +TTGAAT +TTGAAC +TTGAAG +TTGATA +TTGATT +TTGATC +TTGATG +TTGACA +TTGACT +TTGACC +TTGACG +TTGAGA +TTGAGT +TTGAGC +TTGAGG +TTGTAA +TTGTAT +TTGTAC +TTGTAG +TTGTTA +TTGTTT +TTGTTC +TTGTTG +TTGTCA +TTGTCT +TTGTCC +TTGTCG +TTGTGA +TTGTGT +TTGTGC +TTGTGG +TTGCAA +TTGCAT +TTGCAC +TTGCAG +TTGCTA +TTGCTT +TTGCTC +TTGCTG +TTGCCA +TTGCCT +TTGCCC +TTGCCG +TTGCGA +TTGCGT +TTGCGC +TTGCGG +TTGGAA +TTGGAT +TTGGAC +TTGGAG +TTGGTA +TTGGTT +TTGGTC +TTGGTG +TTGGCA +TTGGCT +TTGGCC +TTGGCG +TTGGGA +TTGGGT +TTGGGC +TTGGGG +TCAAAA +TCAAAT +TCAAAC +TCAAAG +TCAATA +TCAATT +TCAATC +TCAATG +TCAACA +TCAACT +TCAACC +TCAACG +TCAAGA +TCAAGT +TCAAGC +TCAAGG +TCATAA +TCATAT +TCATAC +TCATAG +TCATTA +TCATTT +TCATTC +TCATTG +TCATCA +TCATCT +TCATCC +TCATCG +TCATGA +TCATGT +TCATGC +TCATGG +TCACAA +TCACAT +TCACAC +TCACAG +TCACTA +TCACTT +TCACTC +TCACTG +TCACCA +TCACCT +TCACCC +TCACCG +TCACGA +TCACGT +TCACGC +TCACGG +TCAGAA +TCAGAT +TCAGAC +TCAGAG +TCAGTA +TCAGTT +TCAGTC +TCAGTG +TCAGCA +TCAGCT +TCAGCC +TCAGCG +TCAGGA +TCAGGT +TCAGGC +TCAGGG +TCTAAA +TCTAAT +TCTAAC +TCTAAG +TCTATA +TCTATT +TCTATC +TCTATG +TCTACA +TCTACT +TCTACC +TCTACG +TCTAGA +TCTAGT +TCTAGC +TCTAGG +TCTTAA +TCTTAT +TCTTAC +TCTTAG +TCTTTA +TCTTTT +TCTTTC +TCTTTG +TCTTCA +TCTTCT +TCTTCC +TCTTCG +TCTTGA +TCTTGT +TCTTGC +TCTTGG +TCTCAA +TCTCAT +TCTCAC +TCTCAG +TCTCTA +TCTCTT +TCTCTC +TCTCTG +TCTCCA +TCTCCT +TCTCCC +TCTCCG +TCTCGA +TCTCGT +TCTCGC +TCTCGG +TCTGAA +TCTGAT +TCTGAC +TCTGAG +TCTGTA +TCTGTT +TCTGTC +TCTGTG +TCTGCA +TCTGCT +TCTGCC +TCTGCG +TCTGGA +TCTGGT +TCTGGC +TCTGGG +TCCAAA +TCCAAT +TCCAAC +TCCAAG +TCCATA +TCCATT +TCCATC +TCCATG +TCCACA +TCCACT +TCCACC +TCCACG +TCCAGA +TCCAGT +TCCAGC +TCCAGG +TCCTAA +TCCTAT +TCCTAC +TCCTAG +TCCTTA +TCCTTT +TCCTTC +TCCTTG +TCCTCA +TCCTCT +TCCTCC +TCCTCG +TCCTGA +TCCTGT +TCCTGC +TCCTGG +TCCCAA +TCCCAT +TCCCAC +TCCCAG +TCCCTA +TCCCTT +TCCCTC +TCCCTG +TCCCCA +TCCCCT +TCCCCC +TCCCCG +TCCCGA +TCCCGT +TCCCGC +TCCCGG +TCCGAA +TCCGAT +TCCGAC +TCCGAG +TCCGTA +TCCGTT +TCCGTC +TCCGTG +TCCGCA +TCCGCT +TCCGCC +TCCGCG +TCCGGA +TCCGGT +TCCGGC +TCCGGG +TCGAAA +TCGAAT +TCGAAC +TCGAAG +TCGATA +TCGATT +TCGATC +TCGATG +TCGACA +TCGACT +TCGACC +TCGACG +TCGAGA +TCGAGT +TCGAGC +TCGAGG +TCGTAA +TCGTAT +TCGTAC +TCGTAG +TCGTTA +TCGTTT +TCGTTC +TCGTTG +TCGTCA +TCGTCT +TCGTCC +TCGTCG +TCGTGA +TCGTGT +TCGTGC +TCGTGG +TCGCAA +TCGCAT +TCGCAC +TCGCAG +TCGCTA +TCGCTT +TCGCTC +TCGCTG +TCGCCA +TCGCCT +TCGCCC +TCGCCG +TCGCGA +TCGCGT +TCGCGC +TCGCGG +TCGGAA +TCGGAT +TCGGAC +TCGGAG +TCGGTA +TCGGTT +TCGGTC +TCGGTG +TCGGCA +TCGGCT +TCGGCC +TCGGCG +TCGGGA +TCGGGT +TCGGGC +TCGGGG +TGAAAA +TGAAAT +TGAAAC +TGAAAG +TGAATA +TGAATT +TGAATC +TGAATG +TGAACA +TGAACT +TGAACC +TGAACG +TGAAGA +TGAAGT +TGAAGC +TGAAGG +TGATAA +TGATAT +TGATAC +TGATAG +TGATTA +TGATTT +TGATTC +TGATTG +TGATCA +TGATCT +TGATCC +TGATCG +TGATGA +TGATGT +TGATGC +TGATGG +TGACAA +TGACAT +TGACAC +TGACAG +TGACTA +TGACTT +TGACTC +TGACTG +TGACCA +TGACCT +TGACCC +TGACCG +TGACGA +TGACGT +TGACGC +TGACGG +TGAGAA +TGAGAT +TGAGAC +TGAGAG +TGAGTA +TGAGTT +TGAGTC +TGAGTG +TGAGCA +TGAGCT +TGAGCC +TGAGCG +TGAGGA +TGAGGT +TGAGGC +TGAGGG +TGTAAA +TGTAAT +TGTAAC +TGTAAG +TGTATA +TGTATT +TGTATC +TGTATG +TGTACA +TGTACT +TGTACC +TGTACG +TGTAGA +TGTAGT +TGTAGC +TGTAGG +TGTTAA +TGTTAT +TGTTAC +TGTTAG +TGTTTA +TGTTTT +TGTTTC +TGTTTG +TGTTCA +TGTTCT +TGTTCC +TGTTCG +TGTTGA +TGTTGT +TGTTGC +TGTTGG +TGTCAA +TGTCAT +TGTCAC +TGTCAG +TGTCTA +TGTCTT +TGTCTC +TGTCTG +TGTCCA +TGTCCT +TGTCCC +TGTCCG +TGTCGA +TGTCGT +TGTCGC +TGTCGG +TGTGAA +TGTGAT +TGTGAC +TGTGAG +TGTGTA +TGTGTT +TGTGTC +TGTGTG +TGTGCA +TGTGCT +TGTGCC +TGTGCG +TGTGGA +TGTGGT +TGTGGC +TGTGGG +TGCAAA +TGCAAT +TGCAAC +TGCAAG +TGCATA +TGCATT +TGCATC +TGCATG +TGCACA +TGCACT +TGCACC +TGCACG +TGCAGA +TGCAGT +TGCAGC +TGCAGG +TGCTAA +TGCTAT +TGCTAC +TGCTAG +TGCTTA +TGCTTT +TGCTTC +TGCTTG +TGCTCA +TGCTCT +TGCTCC +TGCTCG +TGCTGA +TGCTGT +TGCTGC +TGCTGG +TGCCAA +TGCCAT +TGCCAC +TGCCAG +TGCCTA +TGCCTT +TGCCTC +TGCCTG +TGCCCA +TGCCCT +TGCCCC +TGCCCG +TGCCGA +TGCCGT +TGCCGC +TGCCGG +TGCGAA +TGCGAT +TGCGAC +TGCGAG +TGCGTA +TGCGTT +TGCGTC +TGCGTG +TGCGCA +TGCGCT +TGCGCC +TGCGCG +TGCGGA +TGCGGT +TGCGGC +TGCGGG +TGGAAA +TGGAAT +TGGAAC +TGGAAG +TGGATA +TGGATT +TGGATC +TGGATG +TGGACA +TGGACT +TGGACC +TGGACG +TGGAGA +TGGAGT +TGGAGC +TGGAGG +TGGTAA +TGGTAT +TGGTAC +TGGTAG +TGGTTA +TGGTTT +TGGTTC +TGGTTG +TGGTCA +TGGTCT +TGGTCC +TGGTCG +TGGTGA +TGGTGT +TGGTGC +TGGTGG +TGGCAA +TGGCAT +TGGCAC +TGGCAG +TGGCTA +TGGCTT +TGGCTC +TGGCTG +TGGCCA +TGGCCT +TGGCCC +TGGCCG +TGGCGA +TGGCGT +TGGCGC +TGGCGG +TGGGAA +TGGGAT +TGGGAC +TGGGAG +TGGGTA +TGGGTT +TGGGTC +TGGGTG +TGGGCA +TGGGCT +TGGGCC +TGGGCG +TGGGGA +TGGGGT +TGGGGC +TGGGGG +CAAAAA +CAAAAT +CAAAAC +CAAAAG +CAAATA +CAAATT +CAAATC +CAAATG +CAAACA +CAAACT +CAAACC +CAAACG +CAAAGA +CAAAGT +CAAAGC +CAAAGG +CAATAA +CAATAT +CAATAC +CAATAG +CAATTA +CAATTT +CAATTC +CAATTG +CAATCA +CAATCT +CAATCC +CAATCG +CAATGA +CAATGT +CAATGC +CAATGG +CAACAA +CAACAT +CAACAC +CAACAG +CAACTA +CAACTT +CAACTC +CAACTG +CAACCA +CAACCT +CAACCC +CAACCG +CAACGA +CAACGT +CAACGC +CAACGG +CAAGAA +CAAGAT +CAAGAC +CAAGAG +CAAGTA +CAAGTT +CAAGTC +CAAGTG +CAAGCA +CAAGCT +CAAGCC +CAAGCG +CAAGGA +CAAGGT +CAAGGC +CAAGGG +CATAAA +CATAAT +CATAAC +CATAAG +CATATA +CATATT +CATATC +CATATG +CATACA +CATACT +CATACC +CATACG +CATAGA +CATAGT +CATAGC +CATAGG +CATTAA +CATTAT +CATTAC +CATTAG +CATTTA +CATTTT +CATTTC +CATTTG +CATTCA +CATTCT +CATTCC +CATTCG +CATTGA +CATTGT +CATTGC +CATTGG +CATCAA +CATCAT +CATCAC +CATCAG +CATCTA +CATCTT +CATCTC +CATCTG +CATCCA +CATCCT +CATCCC +CATCCG +CATCGA +CATCGT +CATCGC +CATCGG +CATGAA +CATGAT +CATGAC +CATGAG +CATGTA +CATGTT +CATGTC +CATGTG +CATGCA +CATGCT +CATGCC +CATGCG +CATGGA +CATGGT +CATGGC +CATGGG +CACAAA +CACAAT +CACAAC +CACAAG +CACATA +CACATT +CACATC +CACATG +CACACA +CACACT +CACACC +CACACG +CACAGA +CACAGT +CACAGC +CACAGG +CACTAA +CACTAT +CACTAC +CACTAG +CACTTA +CACTTT +CACTTC +CACTTG +CACTCA +CACTCT +CACTCC +CACTCG +CACTGA +CACTGT +CACTGC +CACTGG +CACCAA +CACCAT +CACCAC +CACCAG +CACCTA +CACCTT +CACCTC +CACCTG +CACCCA +CACCCT +CACCCC +CACCCG +CACCGA +CACCGT +CACCGC +CACCGG +CACGAA +CACGAT +CACGAC +CACGAG +CACGTA +CACGTT +CACGTC +CACGTG +CACGCA +CACGCT +CACGCC +CACGCG +CACGGA +CACGGT +CACGGC +CACGGG +CAGAAA +CAGAAT +CAGAAC +CAGAAG +CAGATA +CAGATT +CAGATC +CAGATG +CAGACA +CAGACT +CAGACC +CAGACG +CAGAGA +CAGAGT +CAGAGC +CAGAGG +CAGTAA +CAGTAT +CAGTAC +CAGTAG +CAGTTA +CAGTTT +CAGTTC +CAGTTG +CAGTCA +CAGTCT +CAGTCC +CAGTCG +CAGTGA +CAGTGT +CAGTGC +CAGTGG +CAGCAA +CAGCAT +CAGCAC +CAGCAG +CAGCTA +CAGCTT +CAGCTC +CAGCTG +CAGCCA +CAGCCT +CAGCCC +CAGCCG +CAGCGA +CAGCGT +CAGCGC +CAGCGG +CAGGAA +CAGGAT +CAGGAC +CAGGAG +CAGGTA +CAGGTT +CAGGTC +CAGGTG +CAGGCA +CAGGCT +CAGGCC +CAGGCG +CAGGGA +CAGGGT +CAGGGC +CAGGGG +CTAAAA +CTAAAT +CTAAAC +CTAAAG +CTAATA +CTAATT +CTAATC +CTAATG +CTAACA +CTAACT +CTAACC +CTAACG +CTAAGA +CTAAGT +CTAAGC +CTAAGG +CTATAA +CTATAT +CTATAC +CTATAG +CTATTA +CTATTT +CTATTC +CTATTG +CTATCA +CTATCT +CTATCC +CTATCG +CTATGA +CTATGT +CTATGC +CTATGG +CTACAA +CTACAT +CTACAC +CTACAG +CTACTA +CTACTT +CTACTC +CTACTG +CTACCA +CTACCT +CTACCC +CTACCG +CTACGA +CTACGT +CTACGC +CTACGG +CTAGAA +CTAGAT +CTAGAC +CTAGAG +CTAGTA +CTAGTT +CTAGTC +CTAGTG +CTAGCA +CTAGCT +CTAGCC +CTAGCG +CTAGGA +CTAGGT +CTAGGC +CTAGGG +CTTAAA +CTTAAT +CTTAAC +CTTAAG +CTTATA +CTTATT +CTTATC +CTTATG +CTTACA +CTTACT +CTTACC +CTTACG +CTTAGA +CTTAGT +CTTAGC +CTTAGG +CTTTAA +CTTTAT +CTTTAC +CTTTAG +CTTTTA +CTTTTT +CTTTTC +CTTTTG +CTTTCA +CTTTCT +CTTTCC +CTTTCG +CTTTGA +CTTTGT +CTTTGC +CTTTGG +CTTCAA +CTTCAT +CTTCAC +CTTCAG +CTTCTA +CTTCTT +CTTCTC +CTTCTG +CTTCCA +CTTCCT +CTTCCC +CTTCCG +CTTCGA +CTTCGT +CTTCGC +CTTCGG +CTTGAA +CTTGAT +CTTGAC +CTTGAG +CTTGTA +CTTGTT +CTTGTC +CTTGTG +CTTGCA +CTTGCT +CTTGCC +CTTGCG +CTTGGA +CTTGGT +CTTGGC +CTTGGG +CTCAAA +CTCAAT +CTCAAC +CTCAAG +CTCATA +CTCATT +CTCATC +CTCATG +CTCACA +CTCACT +CTCACC +CTCACG +CTCAGA +CTCAGT +CTCAGC +CTCAGG +CTCTAA +CTCTAT +CTCTAC +CTCTAG +CTCTTA +CTCTTT +CTCTTC +CTCTTG +CTCTCA +CTCTCT +CTCTCC +CTCTCG +CTCTGA +CTCTGT +CTCTGC +CTCTGG +CTCCAA +CTCCAT +CTCCAC +CTCCAG +CTCCTA +CTCCTT +CTCCTC +CTCCTG +CTCCCA +CTCCCT +CTCCCC +CTCCCG +CTCCGA +CTCCGT +CTCCGC +CTCCGG +CTCGAA +CTCGAT +CTCGAC +CTCGAG +CTCGTA +CTCGTT +CTCGTC +CTCGTG +CTCGCA +CTCGCT +CTCGCC +CTCGCG +CTCGGA +CTCGGT +CTCGGC +CTCGGG +CTGAAA +CTGAAT +CTGAAC +CTGAAG +CTGATA +CTGATT +CTGATC +CTGATG +CTGACA +CTGACT +CTGACC +CTGACG +CTGAGA +CTGAGT +CTGAGC +CTGAGG +CTGTAA +CTGTAT +CTGTAC +CTGTAG +CTGTTA +CTGTTT +CTGTTC +CTGTTG +CTGTCA +CTGTCT +CTGTCC +CTGTCG +CTGTGA +CTGTGT +CTGTGC +CTGTGG +CTGCAA +CTGCAT +CTGCAC +CTGCAG +CTGCTA +CTGCTT +CTGCTC +CTGCTG +CTGCCA +CTGCCT +CTGCCC +CTGCCG +CTGCGA +CTGCGT +CTGCGC +CTGCGG +CTGGAA +CTGGAT +CTGGAC +CTGGAG +CTGGTA +CTGGTT +CTGGTC +CTGGTG +CTGGCA +CTGGCT +CTGGCC +CTGGCG +CTGGGA +CTGGGT +CTGGGC +CTGGGG +CCAAAA +CCAAAT +CCAAAC +CCAAAG +CCAATA +CCAATT +CCAATC +CCAATG +CCAACA +CCAACT +CCAACC +CCAACG +CCAAGA +CCAAGT +CCAAGC +CCAAGG +CCATAA +CCATAT +CCATAC +CCATAG +CCATTA +CCATTT +CCATTC +CCATTG +CCATCA +CCATCT +CCATCC +CCATCG +CCATGA +CCATGT +CCATGC +CCATGG +CCACAA +CCACAT +CCACAC +CCACAG +CCACTA +CCACTT +CCACTC +CCACTG +CCACCA +CCACCT +CCACCC +CCACCG +CCACGA +CCACGT +CCACGC +CCACGG +CCAGAA +CCAGAT +CCAGAC +CCAGAG +CCAGTA +CCAGTT +CCAGTC +CCAGTG +CCAGCA +CCAGCT +CCAGCC +CCAGCG +CCAGGA +CCAGGT +CCAGGC +CCAGGG +CCTAAA +CCTAAT +CCTAAC +CCTAAG +CCTATA +CCTATT +CCTATC +CCTATG +CCTACA +CCTACT +CCTACC +CCTACG +CCTAGA +CCTAGT +CCTAGC +CCTAGG +CCTTAA +CCTTAT +CCTTAC +CCTTAG +CCTTTA +CCTTTT +CCTTTC +CCTTTG +CCTTCA +CCTTCT +CCTTCC +CCTTCG +CCTTGA +CCTTGT +CCTTGC +CCTTGG +CCTCAA +CCTCAT +CCTCAC +CCTCAG +CCTCTA +CCTCTT +CCTCTC +CCTCTG +CCTCCA +CCTCCT +CCTCCC +CCTCCG +CCTCGA +CCTCGT +CCTCGC +CCTCGG +CCTGAA +CCTGAT +CCTGAC +CCTGAG +CCTGTA +CCTGTT +CCTGTC +CCTGTG +CCTGCA +CCTGCT +CCTGCC +CCTGCG +CCTGGA +CCTGGT +CCTGGC +CCTGGG +CCCAAA +CCCAAT +CCCAAC +CCCAAG +CCCATA +CCCATT +CCCATC +CCCATG +CCCACA +CCCACT +CCCACC +CCCACG +CCCAGA +CCCAGT +CCCAGC +CCCAGG +CCCTAA +CCCTAT +CCCTAC +CCCTAG +CCCTTA +CCCTTT +CCCTTC +CCCTTG +CCCTCA +CCCTCT +CCCTCC +CCCTCG +CCCTGA +CCCTGT +CCCTGC +CCCTGG +CCCCAA +CCCCAT +CCCCAC +CCCCAG +CCCCTA +CCCCTT +CCCCTC +CCCCTG +CCCCCA +CCCCCT +CCCCCC +CCCCCG +CCCCGA +CCCCGT +CCCCGC +CCCCGG +CCCGAA +CCCGAT +CCCGAC +CCCGAG +CCCGTA +CCCGTT +CCCGTC +CCCGTG +CCCGCA +CCCGCT +CCCGCC +CCCGCG +CCCGGA +CCCGGT +CCCGGC +CCCGGG +CCGAAA +CCGAAT +CCGAAC +CCGAAG +CCGATA +CCGATT +CCGATC +CCGATG +CCGACA +CCGACT +CCGACC +CCGACG +CCGAGA +CCGAGT +CCGAGC +CCGAGG +CCGTAA +CCGTAT +CCGTAC +CCGTAG +CCGTTA +CCGTTT +CCGTTC +CCGTTG +CCGTCA +CCGTCT +CCGTCC +CCGTCG +CCGTGA +CCGTGT +CCGTGC +CCGTGG +CCGCAA +CCGCAT +CCGCAC +CCGCAG +CCGCTA +CCGCTT +CCGCTC +CCGCTG +CCGCCA +CCGCCT +CCGCCC +CCGCCG +CCGCGA +CCGCGT +CCGCGC +CCGCGG +CCGGAA +CCGGAT +CCGGAC +CCGGAG +CCGGTA +CCGGTT +CCGGTC +CCGGTG +CCGGCA +CCGGCT +CCGGCC +CCGGCG +CCGGGA +CCGGGT +CCGGGC +CCGGGG +CGAAAA +CGAAAT +CGAAAC +CGAAAG +CGAATA +CGAATT +CGAATC +CGAATG +CGAACA +CGAACT +CGAACC +CGAACG +CGAAGA +CGAAGT +CGAAGC +CGAAGG +CGATAA +CGATAT +CGATAC +CGATAG +CGATTA +CGATTT +CGATTC +CGATTG +CGATCA +CGATCT +CGATCC +CGATCG +CGATGA +CGATGT +CGATGC +CGATGG +CGACAA +CGACAT +CGACAC +CGACAG +CGACTA +CGACTT +CGACTC +CGACTG +CGACCA +CGACCT +CGACCC +CGACCG +CGACGA +CGACGT +CGACGC +CGACGG +CGAGAA +CGAGAT +CGAGAC +CGAGAG +CGAGTA +CGAGTT +CGAGTC +CGAGTG +CGAGCA +CGAGCT +CGAGCC +CGAGCG +CGAGGA +CGAGGT +CGAGGC +CGAGGG +CGTAAA +CGTAAT +CGTAAC +CGTAAG +CGTATA +CGTATT +CGTATC +CGTATG +CGTACA +CGTACT +CGTACC +CGTACG +CGTAGA +CGTAGT +CGTAGC +CGTAGG +CGTTAA +CGTTAT +CGTTAC +CGTTAG +CGTTTA +CGTTTT +CGTTTC +CGTTTG +CGTTCA +CGTTCT +CGTTCC +CGTTCG +CGTTGA +CGTTGT +CGTTGC +CGTTGG +CGTCAA +CGTCAT +CGTCAC +CGTCAG +CGTCTA +CGTCTT +CGTCTC +CGTCTG +CGTCCA +CGTCCT +CGTCCC +CGTCCG +CGTCGA +CGTCGT +CGTCGC +CGTCGG +CGTGAA +CGTGAT +CGTGAC +CGTGAG +CGTGTA +CGTGTT +CGTGTC +CGTGTG +CGTGCA +CGTGCT +CGTGCC +CGTGCG +CGTGGA +CGTGGT +CGTGGC +CGTGGG +CGCAAA +CGCAAT +CGCAAC +CGCAAG +CGCATA +CGCATT +CGCATC +CGCATG +CGCACA +CGCACT +CGCACC +CGCACG +CGCAGA +CGCAGT +CGCAGC +CGCAGG +CGCTAA +CGCTAT +CGCTAC +CGCTAG +CGCTTA +CGCTTT +CGCTTC +CGCTTG +CGCTCA +CGCTCT +CGCTCC +CGCTCG +CGCTGA +CGCTGT +CGCTGC +CGCTGG +CGCCAA +CGCCAT +CGCCAC +CGCCAG +CGCCTA +CGCCTT +CGCCTC +CGCCTG +CGCCCA +CGCCCT +CGCCCC +CGCCCG +CGCCGA +CGCCGT +CGCCGC +CGCCGG +CGCGAA +CGCGAT +CGCGAC +CGCGAG +CGCGTA +CGCGTT +CGCGTC +CGCGTG +CGCGCA +CGCGCT +CGCGCC +CGCGCG +CGCGGA +CGCGGT +CGCGGC +CGCGGG +CGGAAA +CGGAAT +CGGAAC +CGGAAG +CGGATA +CGGATT +CGGATC +CGGATG +CGGACA +CGGACT +CGGACC +CGGACG +CGGAGA +CGGAGT +CGGAGC +CGGAGG +CGGTAA +CGGTAT +CGGTAC +CGGTAG +CGGTTA +CGGTTT +CGGTTC +CGGTTG +CGGTCA +CGGTCT +CGGTCC +CGGTCG +CGGTGA +CGGTGT +CGGTGC +CGGTGG +CGGCAA +CGGCAT +CGGCAC +CGGCAG +CGGCTA +CGGCTT +CGGCTC +CGGCTG +CGGCCA +CGGCCT +CGGCCC +CGGCCG +CGGCGA +CGGCGT +CGGCGC +CGGCGG +CGGGAA +CGGGAT +CGGGAC +CGGGAG +CGGGTA +CGGGTT +CGGGTC +CGGGTG +CGGGCA +CGGGCT +CGGGCC +CGGGCG +CGGGGA +CGGGGT +CGGGGC +CGGGGG +GAAAAA +GAAAAT +GAAAAC +GAAAAG +GAAATA +GAAATT +GAAATC +GAAATG +GAAACA +GAAACT +GAAACC +GAAACG +GAAAGA +GAAAGT +GAAAGC +GAAAGG +GAATAA +GAATAT +GAATAC +GAATAG +GAATTA +GAATTT +GAATTC +GAATTG +GAATCA +GAATCT +GAATCC +GAATCG +GAATGA +GAATGT +GAATGC +GAATGG +GAACAA +GAACAT +GAACAC +GAACAG +GAACTA +GAACTT +GAACTC +GAACTG +GAACCA +GAACCT +GAACCC +GAACCG +GAACGA +GAACGT +GAACGC +GAACGG +GAAGAA +GAAGAT +GAAGAC +GAAGAG +GAAGTA +GAAGTT +GAAGTC +GAAGTG +GAAGCA +GAAGCT +GAAGCC +GAAGCG +GAAGGA +GAAGGT +GAAGGC +GAAGGG +GATAAA +GATAAT +GATAAC +GATAAG +GATATA +GATATT +GATATC +GATATG +GATACA +GATACT +GATACC +GATACG +GATAGA +GATAGT +GATAGC +GATAGG +GATTAA +GATTAT +GATTAC +GATTAG +GATTTA +GATTTT +GATTTC +GATTTG +GATTCA +GATTCT +GATTCC +GATTCG +GATTGA +GATTGT +GATTGC +GATTGG +GATCAA +GATCAT +GATCAC +GATCAG +GATCTA +GATCTT +GATCTC +GATCTG +GATCCA +GATCCT +GATCCC +GATCCG +GATCGA +GATCGT +GATCGC +GATCGG +GATGAA +GATGAT +GATGAC +GATGAG +GATGTA +GATGTT +GATGTC +GATGTG +GATGCA +GATGCT +GATGCC +GATGCG +GATGGA +GATGGT +GATGGC +GATGGG +GACAAA +GACAAT +GACAAC +GACAAG +GACATA +GACATT +GACATC +GACATG +GACACA +GACACT +GACACC +GACACG +GACAGA +GACAGT +GACAGC +GACAGG +GACTAA +GACTAT +GACTAC +GACTAG +GACTTA +GACTTT +GACTTC +GACTTG +GACTCA +GACTCT +GACTCC +GACTCG +GACTGA +GACTGT +GACTGC +GACTGG +GACCAA +GACCAT +GACCAC +GACCAG +GACCTA +GACCTT +GACCTC +GACCTG +GACCCA +GACCCT +GACCCC +GACCCG +GACCGA +GACCGT +GACCGC +GACCGG +GACGAA +GACGAT +GACGAC +GACGAG +GACGTA +GACGTT +GACGTC +GACGTG +GACGCA +GACGCT +GACGCC +GACGCG +GACGGA +GACGGT +GACGGC +GACGGG +GAGAAA +GAGAAT +GAGAAC +GAGAAG +GAGATA +GAGATT +GAGATC +GAGATG +GAGACA +GAGACT +GAGACC +GAGACG +GAGAGA +GAGAGT +GAGAGC +GAGAGG +GAGTAA +GAGTAT +GAGTAC +GAGTAG +GAGTTA +GAGTTT +GAGTTC +GAGTTG +GAGTCA +GAGTCT +GAGTCC +GAGTCG +GAGTGA +GAGTGT +GAGTGC +GAGTGG +GAGCAA +GAGCAT +GAGCAC +GAGCAG +GAGCTA +GAGCTT +GAGCTC +GAGCTG +GAGCCA +GAGCCT +GAGCCC +GAGCCG +GAGCGA +GAGCGT +GAGCGC +GAGCGG +GAGGAA +GAGGAT +GAGGAC +GAGGAG +GAGGTA +GAGGTT +GAGGTC +GAGGTG +GAGGCA +GAGGCT +GAGGCC +GAGGCG +GAGGGA +GAGGGT +GAGGGC +GAGGGG +GTAAAA +GTAAAT +GTAAAC +GTAAAG +GTAATA +GTAATT +GTAATC +GTAATG +GTAACA +GTAACT +GTAACC +GTAACG +GTAAGA +GTAAGT +GTAAGC +GTAAGG +GTATAA +GTATAT +GTATAC +GTATAG +GTATTA +GTATTT +GTATTC +GTATTG +GTATCA +GTATCT +GTATCC +GTATCG +GTATGA +GTATGT +GTATGC +GTATGG +GTACAA +GTACAT +GTACAC +GTACAG +GTACTA +GTACTT +GTACTC +GTACTG +GTACCA +GTACCT +GTACCC +GTACCG +GTACGA +GTACGT +GTACGC +GTACGG +GTAGAA +GTAGAT +GTAGAC +GTAGAG +GTAGTA +GTAGTT +GTAGTC +GTAGTG +GTAGCA +GTAGCT +GTAGCC +GTAGCG +GTAGGA +GTAGGT +GTAGGC +GTAGGG +GTTAAA +GTTAAT +GTTAAC +GTTAAG +GTTATA +GTTATT +GTTATC +GTTATG +GTTACA +GTTACT +GTTACC +GTTACG +GTTAGA +GTTAGT +GTTAGC +GTTAGG +GTTTAA +GTTTAT +GTTTAC +GTTTAG +GTTTTA +GTTTTT +GTTTTC +GTTTTG +GTTTCA +GTTTCT +GTTTCC +GTTTCG +GTTTGA +GTTTGT +GTTTGC +GTTTGG +GTTCAA +GTTCAT +GTTCAC +GTTCAG +GTTCTA +GTTCTT +GTTCTC +GTTCTG +GTTCCA +GTTCCT +GTTCCC +GTTCCG +GTTCGA +GTTCGT +GTTCGC +GTTCGG +GTTGAA +GTTGAT +GTTGAC +GTTGAG +GTTGTA +GTTGTT +GTTGTC +GTTGTG +GTTGCA +GTTGCT +GTTGCC +GTTGCG +GTTGGA +GTTGGT +GTTGGC +GTTGGG +GTCAAA +GTCAAT +GTCAAC +GTCAAG +GTCATA +GTCATT +GTCATC +GTCATG +GTCACA +GTCACT +GTCACC +GTCACG +GTCAGA +GTCAGT +GTCAGC +GTCAGG +GTCTAA +GTCTAT +GTCTAC +GTCTAG +GTCTTA +GTCTTT +GTCTTC +GTCTTG +GTCTCA +GTCTCT +GTCTCC +GTCTCG +GTCTGA +GTCTGT +GTCTGC +GTCTGG +GTCCAA +GTCCAT +GTCCAC +GTCCAG +GTCCTA +GTCCTT +GTCCTC +GTCCTG +GTCCCA +GTCCCT +GTCCCC +GTCCCG +GTCCGA +GTCCGT +GTCCGC +GTCCGG +GTCGAA +GTCGAT +GTCGAC +GTCGAG +GTCGTA +GTCGTT +GTCGTC +GTCGTG +GTCGCA +GTCGCT +GTCGCC +GTCGCG +GTCGGA +GTCGGT +GTCGGC +GTCGGG +GTGAAA +GTGAAT +GTGAAC +GTGAAG +GTGATA +GTGATT +GTGATC +GTGATG +GTGACA +GTGACT +GTGACC +GTGACG +GTGAGA +GTGAGT +GTGAGC +GTGAGG +GTGTAA +GTGTAT +GTGTAC +GTGTAG +GTGTTA +GTGTTT +GTGTTC +GTGTTG +GTGTCA +GTGTCT +GTGTCC +GTGTCG +GTGTGA +GTGTGT +GTGTGC +GTGTGG +GTGCAA +GTGCAT +GTGCAC +GTGCAG +GTGCTA +GTGCTT +GTGCTC +GTGCTG +GTGCCA +GTGCCT +GTGCCC +GTGCCG +GTGCGA +GTGCGT +GTGCGC +GTGCGG +GTGGAA +GTGGAT +GTGGAC +GTGGAG +GTGGTA +GTGGTT +GTGGTC +GTGGTG +GTGGCA +GTGGCT +GTGGCC +GTGGCG +GTGGGA +GTGGGT +GTGGGC +GTGGGG +GCAAAA +GCAAAT +GCAAAC +GCAAAG +GCAATA +GCAATT +GCAATC +GCAATG +GCAACA +GCAACT +GCAACC +GCAACG +GCAAGA +GCAAGT +GCAAGC +GCAAGG +GCATAA +GCATAT +GCATAC +GCATAG +GCATTA +GCATTT +GCATTC +GCATTG +GCATCA +GCATCT +GCATCC +GCATCG +GCATGA +GCATGT +GCATGC +GCATGG +GCACAA +GCACAT +GCACAC +GCACAG +GCACTA +GCACTT +GCACTC +GCACTG +GCACCA +GCACCT +GCACCC +GCACCG +GCACGA +GCACGT +GCACGC +GCACGG +GCAGAA +GCAGAT +GCAGAC +GCAGAG +GCAGTA +GCAGTT +GCAGTC +GCAGTG +GCAGCA +GCAGCT +GCAGCC +GCAGCG +GCAGGA +GCAGGT +GCAGGC +GCAGGG +GCTAAA +GCTAAT +GCTAAC +GCTAAG +GCTATA +GCTATT +GCTATC +GCTATG +GCTACA +GCTACT +GCTACC +GCTACG +GCTAGA +GCTAGT +GCTAGC +GCTAGG +GCTTAA +GCTTAT +GCTTAC +GCTTAG +GCTTTA +GCTTTT +GCTTTC +GCTTTG +GCTTCA +GCTTCT +GCTTCC +GCTTCG +GCTTGA +GCTTGT +GCTTGC +GCTTGG +GCTCAA +GCTCAT +GCTCAC +GCTCAG +GCTCTA +GCTCTT +GCTCTC +GCTCTG +GCTCCA +GCTCCT +GCTCCC +GCTCCG +GCTCGA +GCTCGT +GCTCGC +GCTCGG +GCTGAA +GCTGAT +GCTGAC +GCTGAG +GCTGTA +GCTGTT +GCTGTC +GCTGTG +GCTGCA +GCTGCT +GCTGCC +GCTGCG +GCTGGA +GCTGGT +GCTGGC +GCTGGG +GCCAAA +GCCAAT +GCCAAC +GCCAAG +GCCATA +GCCATT +GCCATC +GCCATG +GCCACA +GCCACT +GCCACC +GCCACG +GCCAGA +GCCAGT +GCCAGC +GCCAGG +GCCTAA +GCCTAT +GCCTAC +GCCTAG +GCCTTA +GCCTTT +GCCTTC +GCCTTG +GCCTCA +GCCTCT +GCCTCC +GCCTCG +GCCTGA +GCCTGT +GCCTGC +GCCTGG +GCCCAA +GCCCAT +GCCCAC +GCCCAG +GCCCTA +GCCCTT +GCCCTC +GCCCTG +GCCCCA +GCCCCT +GCCCCC +GCCCCG +GCCCGA +GCCCGT +GCCCGC +GCCCGG +GCCGAA +GCCGAT +GCCGAC +GCCGAG +GCCGTA +GCCGTT +GCCGTC +GCCGTG +GCCGCA +GCCGCT +GCCGCC +GCCGCG +GCCGGA +GCCGGT +GCCGGC +GCCGGG +GCGAAA +GCGAAT +GCGAAC +GCGAAG +GCGATA +GCGATT +GCGATC +GCGATG +GCGACA +GCGACT +GCGACC +GCGACG +GCGAGA +GCGAGT +GCGAGC +GCGAGG +GCGTAA +GCGTAT +GCGTAC +GCGTAG +GCGTTA +GCGTTT +GCGTTC +GCGTTG +GCGTCA +GCGTCT +GCGTCC +GCGTCG +GCGTGA +GCGTGT +GCGTGC +GCGTGG +GCGCAA +GCGCAT +GCGCAC +GCGCAG +GCGCTA +GCGCTT +GCGCTC +GCGCTG +GCGCCA +GCGCCT +GCGCCC +GCGCCG +GCGCGA +GCGCGT +GCGCGC +GCGCGG +GCGGAA +GCGGAT +GCGGAC +GCGGAG +GCGGTA +GCGGTT +GCGGTC +GCGGTG +GCGGCA +GCGGCT +GCGGCC +GCGGCG +GCGGGA +GCGGGT +GCGGGC +GCGGGG +GGAAAA +GGAAAT +GGAAAC +GGAAAG +GGAATA +GGAATT +GGAATC +GGAATG +GGAACA +GGAACT +GGAACC +GGAACG +GGAAGA +GGAAGT +GGAAGC +GGAAGG +GGATAA +GGATAT +GGATAC +GGATAG +GGATTA +GGATTT +GGATTC +GGATTG +GGATCA +GGATCT +GGATCC +GGATCG +GGATGA +GGATGT +GGATGC +GGATGG +GGACAA +GGACAT +GGACAC +GGACAG +GGACTA +GGACTT +GGACTC +GGACTG +GGACCA +GGACCT +GGACCC +GGACCG +GGACGA +GGACGT +GGACGC +GGACGG +GGAGAA +GGAGAT +GGAGAC +GGAGAG +GGAGTA +GGAGTT +GGAGTC +GGAGTG +GGAGCA +GGAGCT +GGAGCC +GGAGCG +GGAGGA +GGAGGT +GGAGGC +GGAGGG +GGTAAA +GGTAAT +GGTAAC +GGTAAG +GGTATA +GGTATT +GGTATC +GGTATG +GGTACA +GGTACT +GGTACC +GGTACG +GGTAGA +GGTAGT +GGTAGC +GGTAGG +GGTTAA +GGTTAT +GGTTAC +GGTTAG +GGTTTA +GGTTTT +GGTTTC +GGTTTG +GGTTCA +GGTTCT +GGTTCC +GGTTCG +GGTTGA +GGTTGT +GGTTGC +GGTTGG +GGTCAA +GGTCAT +GGTCAC +GGTCAG +GGTCTA +GGTCTT +GGTCTC +GGTCTG +GGTCCA +GGTCCT +GGTCCC +GGTCCG +GGTCGA +GGTCGT +GGTCGC +GGTCGG +GGTGAA +GGTGAT +GGTGAC +GGTGAG +GGTGTA +GGTGTT +GGTGTC +GGTGTG +GGTGCA +GGTGCT +GGTGCC +GGTGCG +GGTGGA +GGTGGT +GGTGGC +GGTGGG +GGCAAA +GGCAAT +GGCAAC +GGCAAG +GGCATA +GGCATT +GGCATC +GGCATG +GGCACA +GGCACT +GGCACC +GGCACG +GGCAGA +GGCAGT +GGCAGC +GGCAGG +GGCTAA +GGCTAT +GGCTAC +GGCTAG +GGCTTA +GGCTTT +GGCTTC +GGCTTG +GGCTCA +GGCTCT +GGCTCC +GGCTCG +GGCTGA +GGCTGT +GGCTGC +GGCTGG +GGCCAA +GGCCAT +GGCCAC +GGCCAG +GGCCTA +GGCCTT +GGCCTC +GGCCTG +GGCCCA +GGCCCT +GGCCCC +GGCCCG +GGCCGA +GGCCGT +GGCCGC +GGCCGG +GGCGAA +GGCGAT +GGCGAC +GGCGAG +GGCGTA +GGCGTT +GGCGTC +GGCGTG +GGCGCA +GGCGCT +GGCGCC +GGCGCG +GGCGGA +GGCGGT +GGCGGC +GGCGGG +GGGAAA +GGGAAT +GGGAAC +GGGAAG +GGGATA +GGGATT +GGGATC +GGGATG +GGGACA +GGGACT +GGGACC +GGGACG +GGGAGA +GGGAGT +GGGAGC +GGGAGG +GGGTAA +GGGTAT +GGGTAC +GGGTAG +GGGTTA +GGGTTT +GGGTTC +GGGTTG +GGGTCA +GGGTCT +GGGTCC +GGGTCG +GGGTGA +GGGTGT +GGGTGC +GGGTGG +GGGCAA +GGGCAT +GGGCAC +GGGCAG +GGGCTA +GGGCTT +GGGCTC +GGGCTG +GGGCCA +GGGCCT +GGGCCC +GGGCCG +GGGCGA +GGGCGT +GGGCGC +GGGCGG +GGGGAA +GGGGAT +GGGGAC +GGGGAG +GGGGTA +GGGGTT +GGGGTC +GGGGTG +GGGGCA +GGGGCT +GGGGCC +GGGGCG +GGGGGA +GGGGGT +GGGGGC +GGGGGG diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-1000/config.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..9b14e61d327bf586fcff6e817cf9d365bae18ad2 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-1000/config.json @@ -0,0 +1,26 @@ +{ + "_name_or_path": "fabihamakhdoomi/TinyDNABERT", + "architectures": [ + "BertForSequenceClassification" + ], + "attention_probs_dropout_prob": 0.1, + "classifier_dropout": null, + "hidden_act": "gelu", + "hidden_dropout_prob": 0.1, + "hidden_size": 128, + "initializer_range": 0.02, + "intermediate_size": 512, + "layer_norm_eps": 1e-12, + "max_position_embeddings": 512, + "model_type": "bert", + "num_attention_heads": 2, + "num_hidden_layers": 2, + "pad_token_id": 0, + "position_embedding_type": "absolute", + "problem_type": "single_label_classification", + "torch_dtype": "float32", + "transformers_version": "4.44.0", + "type_vocab_size": 2, + "use_cache": true, + "vocab_size": 4101 +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-1000/model.safetensors b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-1000/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..9a982d22fcd6ed4d6fb15fb14aad9ca44114de0a --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-1000/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f730c474539ed980f94115af278c0c1df519a8653f69752a516766cffa107790 +size 4021688 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-1000/optimizer.pt b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-1000/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..e2f252df5b43a381f67bff2baed6e2d803c66c73 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-1000/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bf8e442c007796468d28537362e222a91db7578edf93805cbecca8fd7d78b157 +size 8058700 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-1000/rng_state.pth b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-1000/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..031c7b1895cd8940e942badcb833c16894800a59 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-1000/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e570743f570544924d9d30f7a79905faa2dba99076a5c64ec39e4f6a500cf60e +size 14244 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-1000/scheduler.pt b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-1000/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..e21fdaa14c288938c4b2c1277194110b759d8724 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-1000/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7480edaf7ecb5003556fedbf146cf26ec75a9ee19931ce031bbb3b509f9927ce +size 1064 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-1000/special_tokens_map.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-1000/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..a8b3208c2884c4efb86e49300fdd3dc877220cdf --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-1000/special_tokens_map.json @@ -0,0 +1,7 @@ +{ + "cls_token": "[CLS]", + "mask_token": "[MASK]", + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-1000/tokenizer.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-1000/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..ebd782de6963b9032ce9374cfe30f5275be1a30b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-1000/tokenizer.json @@ -0,0 +1,4265 @@ +{ + "version": "1.0", + "truncation": { + "direction": "Right", + "max_length": 512, + "strategy": "LongestFirst", + "stride": 0 + }, + "padding": { + "strategy": { + "Fixed": 512 + }, + "direction": "Right", + "pad_to_multiple_of": null, + "pad_id": 0, + "pad_type_id": 0, + "pad_token": "[PAD]" + }, + "added_tokens": [ + { + "id": 0, + "content": "[PAD]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 1, + "content": "[UNK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 2, + "content": "[CLS]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 3, + "content": "[SEP]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 4, + "content": "[MASK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + } + ], + "normalizer": { + "type": "BertNormalizer", + "clean_text": true, + "handle_chinese_chars": true, + "strip_accents": null, + "lowercase": false + }, + "pre_tokenizer": { + "type": "BertPreTokenizer" + }, + "post_processor": { + "type": "TemplateProcessing", + "single": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + } + ], + "pair": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "B", + "type_id": 1 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 1 + } + } + ], + "special_tokens": { + "[CLS]": { + "id": "[CLS]", + "ids": [ + 2 + ], + "tokens": [ + "[CLS]" + ] + }, + "[SEP]": { + "id": "[SEP]", + "ids": [ + 3 + ], + "tokens": [ + "[SEP]" + ] + } + } + }, + "decoder": { + "type": "WordPiece", + "prefix": "##", + "cleanup": true + }, + "model": { + "type": "WordPiece", + "unk_token": "[UNK]", + "continuing_subword_prefix": "##", + "max_input_chars_per_word": 100, + "vocab": { + "[PAD]": 0, + "[UNK]": 1, + "[CLS]": 2, + "[SEP]": 3, + "[MASK]": 4, + "AAAAAA": 5, + "AAAAAT": 6, + "AAAAAC": 7, + "AAAAAG": 8, + "AAAATA": 9, + "AAAATT": 10, + "AAAATC": 11, + "AAAATG": 12, + "AAAACA": 13, + "AAAACT": 14, + "AAAACC": 15, + "AAAACG": 16, + "AAAAGA": 17, + "AAAAGT": 18, + "AAAAGC": 19, + "AAAAGG": 20, + "AAATAA": 21, + "AAATAT": 22, + "AAATAC": 23, + "AAATAG": 24, + "AAATTA": 25, + "AAATTT": 26, + "AAATTC": 27, + "AAATTG": 28, + "AAATCA": 29, + "AAATCT": 30, + "AAATCC": 31, + "AAATCG": 32, + "AAATGA": 33, + "AAATGT": 34, + "AAATGC": 35, + "AAATGG": 36, + "AAACAA": 37, + "AAACAT": 38, + "AAACAC": 39, + "AAACAG": 40, + "AAACTA": 41, + "AAACTT": 42, + "AAACTC": 43, + "AAACTG": 44, + "AAACCA": 45, + "AAACCT": 46, + "AAACCC": 47, + "AAACCG": 48, + "AAACGA": 49, + "AAACGT": 50, + "AAACGC": 51, + "AAACGG": 52, + "AAAGAA": 53, + "AAAGAT": 54, + "AAAGAC": 55, + "AAAGAG": 56, + "AAAGTA": 57, + "AAAGTT": 58, + "AAAGTC": 59, + "AAAGTG": 60, + "AAAGCA": 61, + "AAAGCT": 62, + "AAAGCC": 63, + "AAAGCG": 64, + "AAAGGA": 65, + "AAAGGT": 66, + "AAAGGC": 67, + "AAAGGG": 68, + "AATAAA": 69, + "AATAAT": 70, + "AATAAC": 71, + "AATAAG": 72, + "AATATA": 73, + "AATATT": 74, + "AATATC": 75, + "AATATG": 76, + "AATACA": 77, + "AATACT": 78, + "AATACC": 79, + "AATACG": 80, + "AATAGA": 81, + "AATAGT": 82, + "AATAGC": 83, + "AATAGG": 84, + "AATTAA": 85, + "AATTAT": 86, + "AATTAC": 87, + "AATTAG": 88, + "AATTTA": 89, + "AATTTT": 90, + "AATTTC": 91, + "AATTTG": 92, + "AATTCA": 93, + "AATTCT": 94, + "AATTCC": 95, + "AATTCG": 96, + "AATTGA": 97, + "AATTGT": 98, + "AATTGC": 99, + "AATTGG": 100, + "AATCAA": 101, + "AATCAT": 102, + "AATCAC": 103, + "AATCAG": 104, + "AATCTA": 105, + "AATCTT": 106, + "AATCTC": 107, + "AATCTG": 108, + "AATCCA": 109, + "AATCCT": 110, + "AATCCC": 111, + "AATCCG": 112, + "AATCGA": 113, + "AATCGT": 114, + "AATCGC": 115, + "AATCGG": 116, + "AATGAA": 117, + "AATGAT": 118, + "AATGAC": 119, + "AATGAG": 120, + "AATGTA": 121, + "AATGTT": 122, + "AATGTC": 123, + "AATGTG": 124, + "AATGCA": 125, + "AATGCT": 126, + "AATGCC": 127, + "AATGCG": 128, + "AATGGA": 129, + "AATGGT": 130, + "AATGGC": 131, + "AATGGG": 132, + "AACAAA": 133, + "AACAAT": 134, + "AACAAC": 135, + "AACAAG": 136, + "AACATA": 137, + "AACATT": 138, + "AACATC": 139, + "AACATG": 140, + "AACACA": 141, + "AACACT": 142, + "AACACC": 143, + "AACACG": 144, + "AACAGA": 145, + "AACAGT": 146, + "AACAGC": 147, + "AACAGG": 148, + "AACTAA": 149, + "AACTAT": 150, + "AACTAC": 151, + "AACTAG": 152, + "AACTTA": 153, + "AACTTT": 154, + "AACTTC": 155, + "AACTTG": 156, + "AACTCA": 157, + "AACTCT": 158, + "AACTCC": 159, + "AACTCG": 160, + "AACTGA": 161, + "AACTGT": 162, + "AACTGC": 163, + "AACTGG": 164, + "AACCAA": 165, + "AACCAT": 166, + "AACCAC": 167, + "AACCAG": 168, + "AACCTA": 169, + "AACCTT": 170, + "AACCTC": 171, + "AACCTG": 172, + "AACCCA": 173, + "AACCCT": 174, + "AACCCC": 175, + "AACCCG": 176, + "AACCGA": 177, + "AACCGT": 178, + "AACCGC": 179, + "AACCGG": 180, + "AACGAA": 181, + "AACGAT": 182, + "AACGAC": 183, + "AACGAG": 184, + "AACGTA": 185, + "AACGTT": 186, + "AACGTC": 187, + "AACGTG": 188, + "AACGCA": 189, + "AACGCT": 190, + "AACGCC": 191, + "AACGCG": 192, + "AACGGA": 193, + "AACGGT": 194, + "AACGGC": 195, + "AACGGG": 196, + "AAGAAA": 197, + "AAGAAT": 198, + "AAGAAC": 199, + "AAGAAG": 200, + "AAGATA": 201, + "AAGATT": 202, + "AAGATC": 203, + "AAGATG": 204, + "AAGACA": 205, + "AAGACT": 206, + "AAGACC": 207, + "AAGACG": 208, + "AAGAGA": 209, + "AAGAGT": 210, + "AAGAGC": 211, + "AAGAGG": 212, + "AAGTAA": 213, + "AAGTAT": 214, + "AAGTAC": 215, + "AAGTAG": 216, + "AAGTTA": 217, + "AAGTTT": 218, + "AAGTTC": 219, + "AAGTTG": 220, + "AAGTCA": 221, + "AAGTCT": 222, + "AAGTCC": 223, + "AAGTCG": 224, + "AAGTGA": 225, + "AAGTGT": 226, + "AAGTGC": 227, + "AAGTGG": 228, + "AAGCAA": 229, + "AAGCAT": 230, + "AAGCAC": 231, + "AAGCAG": 232, + "AAGCTA": 233, + "AAGCTT": 234, + "AAGCTC": 235, + "AAGCTG": 236, + "AAGCCA": 237, + "AAGCCT": 238, + "AAGCCC": 239, + "AAGCCG": 240, + "AAGCGA": 241, + "AAGCGT": 242, + "AAGCGC": 243, + "AAGCGG": 244, + "AAGGAA": 245, + "AAGGAT": 246, + "AAGGAC": 247, + "AAGGAG": 248, + "AAGGTA": 249, + "AAGGTT": 250, + "AAGGTC": 251, + "AAGGTG": 252, + "AAGGCA": 253, + "AAGGCT": 254, + "AAGGCC": 255, + "AAGGCG": 256, + "AAGGGA": 257, + "AAGGGT": 258, + "AAGGGC": 259, + "AAGGGG": 260, + "ATAAAA": 261, + "ATAAAT": 262, + "ATAAAC": 263, + "ATAAAG": 264, + "ATAATA": 265, + "ATAATT": 266, + "ATAATC": 267, + "ATAATG": 268, + "ATAACA": 269, + "ATAACT": 270, + "ATAACC": 271, + "ATAACG": 272, + "ATAAGA": 273, + "ATAAGT": 274, + "ATAAGC": 275, + "ATAAGG": 276, + "ATATAA": 277, + "ATATAT": 278, + "ATATAC": 279, + "ATATAG": 280, + "ATATTA": 281, + "ATATTT": 282, + "ATATTC": 283, + "ATATTG": 284, + "ATATCA": 285, + "ATATCT": 286, + "ATATCC": 287, + "ATATCG": 288, + "ATATGA": 289, + "ATATGT": 290, + "ATATGC": 291, + "ATATGG": 292, + "ATACAA": 293, + "ATACAT": 294, + "ATACAC": 295, + "ATACAG": 296, + "ATACTA": 297, + "ATACTT": 298, + "ATACTC": 299, + "ATACTG": 300, + "ATACCA": 301, + "ATACCT": 302, + "ATACCC": 303, + "ATACCG": 304, + "ATACGA": 305, + "ATACGT": 306, + "ATACGC": 307, + "ATACGG": 308, + "ATAGAA": 309, + "ATAGAT": 310, + "ATAGAC": 311, + "ATAGAG": 312, + "ATAGTA": 313, + "ATAGTT": 314, + "ATAGTC": 315, + "ATAGTG": 316, + "ATAGCA": 317, + "ATAGCT": 318, + "ATAGCC": 319, + "ATAGCG": 320, + "ATAGGA": 321, + "ATAGGT": 322, + "ATAGGC": 323, + "ATAGGG": 324, + "ATTAAA": 325, + "ATTAAT": 326, + "ATTAAC": 327, + "ATTAAG": 328, + "ATTATA": 329, + "ATTATT": 330, + "ATTATC": 331, + "ATTATG": 332, + "ATTACA": 333, + "ATTACT": 334, + "ATTACC": 335, + "ATTACG": 336, + "ATTAGA": 337, + "ATTAGT": 338, + "ATTAGC": 339, + "ATTAGG": 340, + "ATTTAA": 341, + "ATTTAT": 342, + "ATTTAC": 343, + "ATTTAG": 344, + "ATTTTA": 345, + "ATTTTT": 346, + "ATTTTC": 347, + "ATTTTG": 348, + "ATTTCA": 349, + "ATTTCT": 350, + "ATTTCC": 351, + "ATTTCG": 352, + "ATTTGA": 353, + "ATTTGT": 354, + "ATTTGC": 355, + "ATTTGG": 356, + "ATTCAA": 357, + "ATTCAT": 358, + "ATTCAC": 359, + "ATTCAG": 360, + "ATTCTA": 361, + "ATTCTT": 362, + "ATTCTC": 363, + "ATTCTG": 364, + "ATTCCA": 365, + "ATTCCT": 366, + "ATTCCC": 367, + "ATTCCG": 368, + "ATTCGA": 369, + "ATTCGT": 370, + "ATTCGC": 371, + "ATTCGG": 372, + "ATTGAA": 373, + "ATTGAT": 374, + "ATTGAC": 375, + "ATTGAG": 376, + "ATTGTA": 377, + "ATTGTT": 378, + "ATTGTC": 379, + "ATTGTG": 380, + "ATTGCA": 381, + "ATTGCT": 382, + "ATTGCC": 383, + "ATTGCG": 384, + "ATTGGA": 385, + "ATTGGT": 386, + "ATTGGC": 387, + "ATTGGG": 388, + "ATCAAA": 389, + "ATCAAT": 390, + "ATCAAC": 391, + "ATCAAG": 392, + "ATCATA": 393, + "ATCATT": 394, + "ATCATC": 395, + "ATCATG": 396, + "ATCACA": 397, + "ATCACT": 398, + "ATCACC": 399, + "ATCACG": 400, + "ATCAGA": 401, + "ATCAGT": 402, + "ATCAGC": 403, + "ATCAGG": 404, + "ATCTAA": 405, + "ATCTAT": 406, + "ATCTAC": 407, + "ATCTAG": 408, + "ATCTTA": 409, + "ATCTTT": 410, + "ATCTTC": 411, + "ATCTTG": 412, + "ATCTCA": 413, + "ATCTCT": 414, + "ATCTCC": 415, + "ATCTCG": 416, + "ATCTGA": 417, + "ATCTGT": 418, + "ATCTGC": 419, + "ATCTGG": 420, + "ATCCAA": 421, + "ATCCAT": 422, + "ATCCAC": 423, + "ATCCAG": 424, + "ATCCTA": 425, + "ATCCTT": 426, + "ATCCTC": 427, + "ATCCTG": 428, + "ATCCCA": 429, + "ATCCCT": 430, + "ATCCCC": 431, + "ATCCCG": 432, + "ATCCGA": 433, + "ATCCGT": 434, + "ATCCGC": 435, + "ATCCGG": 436, + "ATCGAA": 437, + "ATCGAT": 438, + "ATCGAC": 439, + "ATCGAG": 440, + "ATCGTA": 441, + "ATCGTT": 442, + "ATCGTC": 443, + "ATCGTG": 444, + "ATCGCA": 445, + "ATCGCT": 446, + "ATCGCC": 447, + "ATCGCG": 448, + "ATCGGA": 449, + "ATCGGT": 450, + "ATCGGC": 451, + "ATCGGG": 452, + "ATGAAA": 453, + "ATGAAT": 454, + "ATGAAC": 455, + "ATGAAG": 456, + "ATGATA": 457, + "ATGATT": 458, + "ATGATC": 459, + "ATGATG": 460, + "ATGACA": 461, + "ATGACT": 462, + "ATGACC": 463, + "ATGACG": 464, + "ATGAGA": 465, + "ATGAGT": 466, + "ATGAGC": 467, + "ATGAGG": 468, + "ATGTAA": 469, + "ATGTAT": 470, + "ATGTAC": 471, + "ATGTAG": 472, + "ATGTTA": 473, + "ATGTTT": 474, + "ATGTTC": 475, + "ATGTTG": 476, + "ATGTCA": 477, + "ATGTCT": 478, + "ATGTCC": 479, + "ATGTCG": 480, + "ATGTGA": 481, + "ATGTGT": 482, + "ATGTGC": 483, + "ATGTGG": 484, + "ATGCAA": 485, + "ATGCAT": 486, + "ATGCAC": 487, + "ATGCAG": 488, + "ATGCTA": 489, + "ATGCTT": 490, + "ATGCTC": 491, + "ATGCTG": 492, + "ATGCCA": 493, + "ATGCCT": 494, + "ATGCCC": 495, + "ATGCCG": 496, + "ATGCGA": 497, + "ATGCGT": 498, + "ATGCGC": 499, + "ATGCGG": 500, + "ATGGAA": 501, + "ATGGAT": 502, + "ATGGAC": 503, + "ATGGAG": 504, + "ATGGTA": 505, + "ATGGTT": 506, + "ATGGTC": 507, + "ATGGTG": 508, + "ATGGCA": 509, + "ATGGCT": 510, + "ATGGCC": 511, + "ATGGCG": 512, + "ATGGGA": 513, + "ATGGGT": 514, + "ATGGGC": 515, + "ATGGGG": 516, + "ACAAAA": 517, + "ACAAAT": 518, + "ACAAAC": 519, + "ACAAAG": 520, + "ACAATA": 521, + "ACAATT": 522, + "ACAATC": 523, + "ACAATG": 524, + "ACAACA": 525, + "ACAACT": 526, + "ACAACC": 527, + "ACAACG": 528, + "ACAAGA": 529, + "ACAAGT": 530, + "ACAAGC": 531, + "ACAAGG": 532, + "ACATAA": 533, + "ACATAT": 534, + "ACATAC": 535, + "ACATAG": 536, + "ACATTA": 537, + "ACATTT": 538, + "ACATTC": 539, + "ACATTG": 540, + "ACATCA": 541, + "ACATCT": 542, + "ACATCC": 543, + "ACATCG": 544, + "ACATGA": 545, + "ACATGT": 546, + "ACATGC": 547, + "ACATGG": 548, + "ACACAA": 549, + "ACACAT": 550, + "ACACAC": 551, + "ACACAG": 552, + "ACACTA": 553, + "ACACTT": 554, + "ACACTC": 555, + "ACACTG": 556, + "ACACCA": 557, + "ACACCT": 558, + "ACACCC": 559, + "ACACCG": 560, + "ACACGA": 561, + "ACACGT": 562, + "ACACGC": 563, + "ACACGG": 564, + "ACAGAA": 565, + "ACAGAT": 566, + "ACAGAC": 567, + "ACAGAG": 568, + "ACAGTA": 569, + "ACAGTT": 570, + "ACAGTC": 571, + "ACAGTG": 572, + "ACAGCA": 573, + "ACAGCT": 574, + "ACAGCC": 575, + "ACAGCG": 576, + "ACAGGA": 577, + "ACAGGT": 578, + "ACAGGC": 579, + "ACAGGG": 580, + "ACTAAA": 581, + "ACTAAT": 582, + "ACTAAC": 583, + "ACTAAG": 584, + "ACTATA": 585, + "ACTATT": 586, + "ACTATC": 587, + "ACTATG": 588, + "ACTACA": 589, + "ACTACT": 590, + "ACTACC": 591, + "ACTACG": 592, + "ACTAGA": 593, + "ACTAGT": 594, + "ACTAGC": 595, + "ACTAGG": 596, + "ACTTAA": 597, + "ACTTAT": 598, + "ACTTAC": 599, + "ACTTAG": 600, + "ACTTTA": 601, + "ACTTTT": 602, + "ACTTTC": 603, + "ACTTTG": 604, + "ACTTCA": 605, + "ACTTCT": 606, + "ACTTCC": 607, + "ACTTCG": 608, + "ACTTGA": 609, + "ACTTGT": 610, + "ACTTGC": 611, + "ACTTGG": 612, + "ACTCAA": 613, + "ACTCAT": 614, + "ACTCAC": 615, + "ACTCAG": 616, + "ACTCTA": 617, + "ACTCTT": 618, + "ACTCTC": 619, + "ACTCTG": 620, + "ACTCCA": 621, + "ACTCCT": 622, + "ACTCCC": 623, + "ACTCCG": 624, + "ACTCGA": 625, + "ACTCGT": 626, + "ACTCGC": 627, + "ACTCGG": 628, + "ACTGAA": 629, + "ACTGAT": 630, + "ACTGAC": 631, + "ACTGAG": 632, + "ACTGTA": 633, + "ACTGTT": 634, + "ACTGTC": 635, + "ACTGTG": 636, + "ACTGCA": 637, + "ACTGCT": 638, + "ACTGCC": 639, + "ACTGCG": 640, + "ACTGGA": 641, + "ACTGGT": 642, + "ACTGGC": 643, + "ACTGGG": 644, + "ACCAAA": 645, + "ACCAAT": 646, + "ACCAAC": 647, + "ACCAAG": 648, + "ACCATA": 649, + "ACCATT": 650, + "ACCATC": 651, + "ACCATG": 652, + "ACCACA": 653, + "ACCACT": 654, + "ACCACC": 655, + "ACCACG": 656, + "ACCAGA": 657, + "ACCAGT": 658, + "ACCAGC": 659, + "ACCAGG": 660, + "ACCTAA": 661, + "ACCTAT": 662, + "ACCTAC": 663, + "ACCTAG": 664, + "ACCTTA": 665, + "ACCTTT": 666, + "ACCTTC": 667, + "ACCTTG": 668, + "ACCTCA": 669, + "ACCTCT": 670, + "ACCTCC": 671, + "ACCTCG": 672, + "ACCTGA": 673, + "ACCTGT": 674, + "ACCTGC": 675, + "ACCTGG": 676, + "ACCCAA": 677, + "ACCCAT": 678, + "ACCCAC": 679, + "ACCCAG": 680, + "ACCCTA": 681, + "ACCCTT": 682, + "ACCCTC": 683, + "ACCCTG": 684, + "ACCCCA": 685, + "ACCCCT": 686, + "ACCCCC": 687, + "ACCCCG": 688, + "ACCCGA": 689, + "ACCCGT": 690, + "ACCCGC": 691, + "ACCCGG": 692, + "ACCGAA": 693, + "ACCGAT": 694, + "ACCGAC": 695, + "ACCGAG": 696, + "ACCGTA": 697, + "ACCGTT": 698, + "ACCGTC": 699, + "ACCGTG": 700, + "ACCGCA": 701, + "ACCGCT": 702, + "ACCGCC": 703, + "ACCGCG": 704, + "ACCGGA": 705, + "ACCGGT": 706, + "ACCGGC": 707, + "ACCGGG": 708, + "ACGAAA": 709, + "ACGAAT": 710, + "ACGAAC": 711, + "ACGAAG": 712, + "ACGATA": 713, + "ACGATT": 714, + "ACGATC": 715, + "ACGATG": 716, + "ACGACA": 717, + "ACGACT": 718, + "ACGACC": 719, + "ACGACG": 720, + "ACGAGA": 721, + "ACGAGT": 722, + "ACGAGC": 723, + "ACGAGG": 724, + "ACGTAA": 725, + "ACGTAT": 726, + "ACGTAC": 727, + "ACGTAG": 728, + "ACGTTA": 729, + "ACGTTT": 730, + "ACGTTC": 731, + "ACGTTG": 732, + "ACGTCA": 733, + "ACGTCT": 734, + "ACGTCC": 735, + "ACGTCG": 736, + "ACGTGA": 737, + "ACGTGT": 738, + "ACGTGC": 739, + "ACGTGG": 740, + "ACGCAA": 741, + "ACGCAT": 742, + "ACGCAC": 743, + "ACGCAG": 744, + "ACGCTA": 745, + "ACGCTT": 746, + "ACGCTC": 747, + "ACGCTG": 748, + "ACGCCA": 749, + "ACGCCT": 750, + "ACGCCC": 751, + "ACGCCG": 752, + "ACGCGA": 753, + "ACGCGT": 754, + "ACGCGC": 755, + "ACGCGG": 756, + "ACGGAA": 757, + "ACGGAT": 758, + "ACGGAC": 759, + "ACGGAG": 760, + "ACGGTA": 761, + "ACGGTT": 762, + "ACGGTC": 763, + "ACGGTG": 764, + "ACGGCA": 765, + "ACGGCT": 766, + "ACGGCC": 767, + "ACGGCG": 768, + "ACGGGA": 769, + "ACGGGT": 770, + "ACGGGC": 771, + "ACGGGG": 772, + "AGAAAA": 773, + "AGAAAT": 774, + "AGAAAC": 775, + "AGAAAG": 776, + "AGAATA": 777, + "AGAATT": 778, + "AGAATC": 779, + "AGAATG": 780, + "AGAACA": 781, + "AGAACT": 782, + "AGAACC": 783, + "AGAACG": 784, + "AGAAGA": 785, + "AGAAGT": 786, + "AGAAGC": 787, + "AGAAGG": 788, + "AGATAA": 789, + "AGATAT": 790, + "AGATAC": 791, + "AGATAG": 792, + "AGATTA": 793, + "AGATTT": 794, + "AGATTC": 795, + "AGATTG": 796, + "AGATCA": 797, + "AGATCT": 798, + "AGATCC": 799, + "AGATCG": 800, + "AGATGA": 801, + "AGATGT": 802, + "AGATGC": 803, + "AGATGG": 804, + "AGACAA": 805, + "AGACAT": 806, + "AGACAC": 807, + "AGACAG": 808, + "AGACTA": 809, + "AGACTT": 810, + "AGACTC": 811, + "AGACTG": 812, + "AGACCA": 813, + "AGACCT": 814, + "AGACCC": 815, + "AGACCG": 816, + "AGACGA": 817, + "AGACGT": 818, + "AGACGC": 819, + "AGACGG": 820, + "AGAGAA": 821, + "AGAGAT": 822, + "AGAGAC": 823, + "AGAGAG": 824, + "AGAGTA": 825, + "AGAGTT": 826, + "AGAGTC": 827, + "AGAGTG": 828, + "AGAGCA": 829, + "AGAGCT": 830, + "AGAGCC": 831, + "AGAGCG": 832, + "AGAGGA": 833, + "AGAGGT": 834, + "AGAGGC": 835, + "AGAGGG": 836, + "AGTAAA": 837, + "AGTAAT": 838, + "AGTAAC": 839, + "AGTAAG": 840, + "AGTATA": 841, + "AGTATT": 842, + "AGTATC": 843, + "AGTATG": 844, + "AGTACA": 845, + "AGTACT": 846, + "AGTACC": 847, + "AGTACG": 848, + "AGTAGA": 849, + "AGTAGT": 850, + "AGTAGC": 851, + "AGTAGG": 852, + "AGTTAA": 853, + "AGTTAT": 854, + "AGTTAC": 855, + "AGTTAG": 856, + "AGTTTA": 857, + "AGTTTT": 858, + "AGTTTC": 859, + "AGTTTG": 860, + "AGTTCA": 861, + "AGTTCT": 862, + "AGTTCC": 863, + "AGTTCG": 864, + "AGTTGA": 865, + "AGTTGT": 866, + "AGTTGC": 867, + "AGTTGG": 868, + "AGTCAA": 869, + "AGTCAT": 870, + "AGTCAC": 871, + "AGTCAG": 872, + "AGTCTA": 873, + "AGTCTT": 874, + "AGTCTC": 875, + "AGTCTG": 876, + "AGTCCA": 877, + "AGTCCT": 878, + "AGTCCC": 879, + "AGTCCG": 880, + "AGTCGA": 881, + "AGTCGT": 882, + "AGTCGC": 883, + "AGTCGG": 884, + "AGTGAA": 885, + "AGTGAT": 886, + "AGTGAC": 887, + "AGTGAG": 888, + "AGTGTA": 889, + "AGTGTT": 890, + "AGTGTC": 891, + "AGTGTG": 892, + "AGTGCA": 893, + "AGTGCT": 894, + "AGTGCC": 895, + "AGTGCG": 896, + "AGTGGA": 897, + "AGTGGT": 898, + "AGTGGC": 899, + "AGTGGG": 900, + "AGCAAA": 901, + "AGCAAT": 902, + "AGCAAC": 903, + "AGCAAG": 904, + "AGCATA": 905, + "AGCATT": 906, + "AGCATC": 907, + "AGCATG": 908, + "AGCACA": 909, + "AGCACT": 910, + "AGCACC": 911, + "AGCACG": 912, + "AGCAGA": 913, + "AGCAGT": 914, + "AGCAGC": 915, + "AGCAGG": 916, + "AGCTAA": 917, + "AGCTAT": 918, + "AGCTAC": 919, + "AGCTAG": 920, + "AGCTTA": 921, + "AGCTTT": 922, + "AGCTTC": 923, + "AGCTTG": 924, + "AGCTCA": 925, + "AGCTCT": 926, + "AGCTCC": 927, + "AGCTCG": 928, + "AGCTGA": 929, + "AGCTGT": 930, + "AGCTGC": 931, + "AGCTGG": 932, + "AGCCAA": 933, + "AGCCAT": 934, + "AGCCAC": 935, + "AGCCAG": 936, + "AGCCTA": 937, + "AGCCTT": 938, + "AGCCTC": 939, + "AGCCTG": 940, + "AGCCCA": 941, + "AGCCCT": 942, + "AGCCCC": 943, + "AGCCCG": 944, + "AGCCGA": 945, + "AGCCGT": 946, + "AGCCGC": 947, + "AGCCGG": 948, + "AGCGAA": 949, + "AGCGAT": 950, + "AGCGAC": 951, + "AGCGAG": 952, + "AGCGTA": 953, + "AGCGTT": 954, + "AGCGTC": 955, + "AGCGTG": 956, + "AGCGCA": 957, + "AGCGCT": 958, + "AGCGCC": 959, + "AGCGCG": 960, + "AGCGGA": 961, + "AGCGGT": 962, + "AGCGGC": 963, + "AGCGGG": 964, + "AGGAAA": 965, + "AGGAAT": 966, + "AGGAAC": 967, + "AGGAAG": 968, + "AGGATA": 969, + "AGGATT": 970, + "AGGATC": 971, + "AGGATG": 972, + "AGGACA": 973, + "AGGACT": 974, + "AGGACC": 975, + "AGGACG": 976, + "AGGAGA": 977, + "AGGAGT": 978, + "AGGAGC": 979, + "AGGAGG": 980, + "AGGTAA": 981, + "AGGTAT": 982, + "AGGTAC": 983, + "AGGTAG": 984, + "AGGTTA": 985, + "AGGTTT": 986, + "AGGTTC": 987, + "AGGTTG": 988, + "AGGTCA": 989, + "AGGTCT": 990, + "AGGTCC": 991, + "AGGTCG": 992, + "AGGTGA": 993, + "AGGTGT": 994, + "AGGTGC": 995, + "AGGTGG": 996, + "AGGCAA": 997, + "AGGCAT": 998, + "AGGCAC": 999, + "AGGCAG": 1000, + "AGGCTA": 1001, + "AGGCTT": 1002, + "AGGCTC": 1003, + "AGGCTG": 1004, + "AGGCCA": 1005, + "AGGCCT": 1006, + "AGGCCC": 1007, + "AGGCCG": 1008, + "AGGCGA": 1009, + "AGGCGT": 1010, + "AGGCGC": 1011, + "AGGCGG": 1012, + "AGGGAA": 1013, + "AGGGAT": 1014, + "AGGGAC": 1015, + "AGGGAG": 1016, + "AGGGTA": 1017, + "AGGGTT": 1018, + "AGGGTC": 1019, + "AGGGTG": 1020, + "AGGGCA": 1021, + "AGGGCT": 1022, + "AGGGCC": 1023, + "AGGGCG": 1024, + "AGGGGA": 1025, + "AGGGGT": 1026, + "AGGGGC": 1027, + "AGGGGG": 1028, + "TAAAAA": 1029, + "TAAAAT": 1030, + "TAAAAC": 1031, + "TAAAAG": 1032, + "TAAATA": 1033, + "TAAATT": 1034, + "TAAATC": 1035, + "TAAATG": 1036, + "TAAACA": 1037, + "TAAACT": 1038, + "TAAACC": 1039, + "TAAACG": 1040, + "TAAAGA": 1041, + "TAAAGT": 1042, + "TAAAGC": 1043, + "TAAAGG": 1044, + "TAATAA": 1045, + "TAATAT": 1046, + "TAATAC": 1047, + "TAATAG": 1048, + "TAATTA": 1049, + "TAATTT": 1050, + "TAATTC": 1051, + "TAATTG": 1052, + "TAATCA": 1053, + "TAATCT": 1054, + "TAATCC": 1055, + "TAATCG": 1056, + "TAATGA": 1057, + "TAATGT": 1058, + "TAATGC": 1059, + "TAATGG": 1060, + "TAACAA": 1061, + "TAACAT": 1062, + "TAACAC": 1063, + "TAACAG": 1064, + "TAACTA": 1065, + "TAACTT": 1066, + "TAACTC": 1067, + "TAACTG": 1068, + "TAACCA": 1069, + "TAACCT": 1070, + "TAACCC": 1071, + "TAACCG": 1072, + "TAACGA": 1073, + "TAACGT": 1074, + "TAACGC": 1075, + "TAACGG": 1076, + "TAAGAA": 1077, + "TAAGAT": 1078, + "TAAGAC": 1079, + "TAAGAG": 1080, + "TAAGTA": 1081, + "TAAGTT": 1082, + "TAAGTC": 1083, + "TAAGTG": 1084, + "TAAGCA": 1085, + "TAAGCT": 1086, + "TAAGCC": 1087, + "TAAGCG": 1088, + "TAAGGA": 1089, + "TAAGGT": 1090, + "TAAGGC": 1091, + "TAAGGG": 1092, + "TATAAA": 1093, + "TATAAT": 1094, + "TATAAC": 1095, + "TATAAG": 1096, + "TATATA": 1097, + "TATATT": 1098, + "TATATC": 1099, + "TATATG": 1100, + "TATACA": 1101, + "TATACT": 1102, + "TATACC": 1103, + "TATACG": 1104, + "TATAGA": 1105, + "TATAGT": 1106, + "TATAGC": 1107, + "TATAGG": 1108, + "TATTAA": 1109, + "TATTAT": 1110, + "TATTAC": 1111, + "TATTAG": 1112, + "TATTTA": 1113, + "TATTTT": 1114, + "TATTTC": 1115, + "TATTTG": 1116, + "TATTCA": 1117, + "TATTCT": 1118, + "TATTCC": 1119, + "TATTCG": 1120, + "TATTGA": 1121, + "TATTGT": 1122, + "TATTGC": 1123, + "TATTGG": 1124, + "TATCAA": 1125, + "TATCAT": 1126, + "TATCAC": 1127, + "TATCAG": 1128, + "TATCTA": 1129, + "TATCTT": 1130, + "TATCTC": 1131, + "TATCTG": 1132, + "TATCCA": 1133, + "TATCCT": 1134, + "TATCCC": 1135, + "TATCCG": 1136, + "TATCGA": 1137, + "TATCGT": 1138, + "TATCGC": 1139, + "TATCGG": 1140, + "TATGAA": 1141, + "TATGAT": 1142, + "TATGAC": 1143, + "TATGAG": 1144, + "TATGTA": 1145, + "TATGTT": 1146, + "TATGTC": 1147, + "TATGTG": 1148, + "TATGCA": 1149, + "TATGCT": 1150, + "TATGCC": 1151, + "TATGCG": 1152, + "TATGGA": 1153, + "TATGGT": 1154, + "TATGGC": 1155, + "TATGGG": 1156, + "TACAAA": 1157, + "TACAAT": 1158, + "TACAAC": 1159, + "TACAAG": 1160, + "TACATA": 1161, + "TACATT": 1162, + "TACATC": 1163, + "TACATG": 1164, + "TACACA": 1165, + "TACACT": 1166, + "TACACC": 1167, + "TACACG": 1168, + "TACAGA": 1169, + "TACAGT": 1170, + "TACAGC": 1171, + "TACAGG": 1172, + "TACTAA": 1173, + "TACTAT": 1174, + "TACTAC": 1175, + "TACTAG": 1176, + "TACTTA": 1177, + "TACTTT": 1178, + "TACTTC": 1179, + "TACTTG": 1180, + "TACTCA": 1181, + "TACTCT": 1182, + "TACTCC": 1183, + "TACTCG": 1184, + "TACTGA": 1185, + "TACTGT": 1186, + "TACTGC": 1187, + "TACTGG": 1188, + "TACCAA": 1189, + "TACCAT": 1190, + "TACCAC": 1191, + "TACCAG": 1192, + "TACCTA": 1193, + "TACCTT": 1194, + "TACCTC": 1195, + "TACCTG": 1196, + "TACCCA": 1197, + "TACCCT": 1198, + "TACCCC": 1199, + "TACCCG": 1200, + "TACCGA": 1201, + "TACCGT": 1202, + "TACCGC": 1203, + "TACCGG": 1204, + "TACGAA": 1205, + "TACGAT": 1206, + "TACGAC": 1207, + "TACGAG": 1208, + "TACGTA": 1209, + "TACGTT": 1210, + "TACGTC": 1211, + "TACGTG": 1212, + "TACGCA": 1213, + "TACGCT": 1214, + "TACGCC": 1215, + "TACGCG": 1216, + "TACGGA": 1217, + "TACGGT": 1218, + "TACGGC": 1219, + "TACGGG": 1220, + "TAGAAA": 1221, + "TAGAAT": 1222, + "TAGAAC": 1223, + "TAGAAG": 1224, + "TAGATA": 1225, + "TAGATT": 1226, + "TAGATC": 1227, + "TAGATG": 1228, + "TAGACA": 1229, + "TAGACT": 1230, + "TAGACC": 1231, + "TAGACG": 1232, + "TAGAGA": 1233, + "TAGAGT": 1234, + "TAGAGC": 1235, + "TAGAGG": 1236, + "TAGTAA": 1237, + "TAGTAT": 1238, + "TAGTAC": 1239, + "TAGTAG": 1240, + "TAGTTA": 1241, + "TAGTTT": 1242, + "TAGTTC": 1243, + "TAGTTG": 1244, + "TAGTCA": 1245, + "TAGTCT": 1246, + "TAGTCC": 1247, + "TAGTCG": 1248, + "TAGTGA": 1249, + "TAGTGT": 1250, + "TAGTGC": 1251, + "TAGTGG": 1252, + "TAGCAA": 1253, + "TAGCAT": 1254, + "TAGCAC": 1255, + "TAGCAG": 1256, + "TAGCTA": 1257, + "TAGCTT": 1258, + "TAGCTC": 1259, + "TAGCTG": 1260, + "TAGCCA": 1261, + "TAGCCT": 1262, + "TAGCCC": 1263, + "TAGCCG": 1264, + "TAGCGA": 1265, + "TAGCGT": 1266, + "TAGCGC": 1267, + "TAGCGG": 1268, + "TAGGAA": 1269, + "TAGGAT": 1270, + "TAGGAC": 1271, + "TAGGAG": 1272, + "TAGGTA": 1273, + "TAGGTT": 1274, + "TAGGTC": 1275, + "TAGGTG": 1276, + "TAGGCA": 1277, + "TAGGCT": 1278, + "TAGGCC": 1279, + "TAGGCG": 1280, + "TAGGGA": 1281, + "TAGGGT": 1282, + "TAGGGC": 1283, + "TAGGGG": 1284, + "TTAAAA": 1285, + "TTAAAT": 1286, + "TTAAAC": 1287, + "TTAAAG": 1288, + "TTAATA": 1289, + "TTAATT": 1290, + "TTAATC": 1291, + "TTAATG": 1292, + "TTAACA": 1293, + "TTAACT": 1294, + "TTAACC": 1295, + "TTAACG": 1296, + "TTAAGA": 1297, + "TTAAGT": 1298, + "TTAAGC": 1299, + "TTAAGG": 1300, + "TTATAA": 1301, + "TTATAT": 1302, + "TTATAC": 1303, + "TTATAG": 1304, + "TTATTA": 1305, + "TTATTT": 1306, + "TTATTC": 1307, + "TTATTG": 1308, + "TTATCA": 1309, + "TTATCT": 1310, + "TTATCC": 1311, + "TTATCG": 1312, + "TTATGA": 1313, + "TTATGT": 1314, + "TTATGC": 1315, + "TTATGG": 1316, + "TTACAA": 1317, + "TTACAT": 1318, + "TTACAC": 1319, + "TTACAG": 1320, + "TTACTA": 1321, + "TTACTT": 1322, + "TTACTC": 1323, + "TTACTG": 1324, + "TTACCA": 1325, + "TTACCT": 1326, + "TTACCC": 1327, + "TTACCG": 1328, + "TTACGA": 1329, + "TTACGT": 1330, + "TTACGC": 1331, + "TTACGG": 1332, + "TTAGAA": 1333, + "TTAGAT": 1334, + "TTAGAC": 1335, + "TTAGAG": 1336, + "TTAGTA": 1337, + "TTAGTT": 1338, + "TTAGTC": 1339, + "TTAGTG": 1340, + "TTAGCA": 1341, + "TTAGCT": 1342, + "TTAGCC": 1343, + "TTAGCG": 1344, + "TTAGGA": 1345, + "TTAGGT": 1346, + "TTAGGC": 1347, + "TTAGGG": 1348, + "TTTAAA": 1349, + "TTTAAT": 1350, + "TTTAAC": 1351, + "TTTAAG": 1352, + "TTTATA": 1353, + "TTTATT": 1354, + "TTTATC": 1355, + "TTTATG": 1356, + "TTTACA": 1357, + "TTTACT": 1358, + "TTTACC": 1359, + "TTTACG": 1360, + "TTTAGA": 1361, + "TTTAGT": 1362, + "TTTAGC": 1363, + "TTTAGG": 1364, + "TTTTAA": 1365, + "TTTTAT": 1366, + "TTTTAC": 1367, + "TTTTAG": 1368, + "TTTTTA": 1369, + "TTTTTT": 1370, + "TTTTTC": 1371, + "TTTTTG": 1372, + "TTTTCA": 1373, + "TTTTCT": 1374, + "TTTTCC": 1375, + "TTTTCG": 1376, + "TTTTGA": 1377, + "TTTTGT": 1378, + "TTTTGC": 1379, + "TTTTGG": 1380, + "TTTCAA": 1381, + "TTTCAT": 1382, + "TTTCAC": 1383, + "TTTCAG": 1384, + "TTTCTA": 1385, + "TTTCTT": 1386, + "TTTCTC": 1387, + "TTTCTG": 1388, + "TTTCCA": 1389, + "TTTCCT": 1390, + "TTTCCC": 1391, + "TTTCCG": 1392, + "TTTCGA": 1393, + "TTTCGT": 1394, + "TTTCGC": 1395, + "TTTCGG": 1396, + "TTTGAA": 1397, + "TTTGAT": 1398, + "TTTGAC": 1399, + "TTTGAG": 1400, + "TTTGTA": 1401, + "TTTGTT": 1402, + "TTTGTC": 1403, + "TTTGTG": 1404, + "TTTGCA": 1405, + "TTTGCT": 1406, + "TTTGCC": 1407, + "TTTGCG": 1408, + "TTTGGA": 1409, + "TTTGGT": 1410, + "TTTGGC": 1411, + "TTTGGG": 1412, + "TTCAAA": 1413, + "TTCAAT": 1414, + "TTCAAC": 1415, + "TTCAAG": 1416, + "TTCATA": 1417, + "TTCATT": 1418, + "TTCATC": 1419, + "TTCATG": 1420, + "TTCACA": 1421, + "TTCACT": 1422, + "TTCACC": 1423, + "TTCACG": 1424, + "TTCAGA": 1425, + "TTCAGT": 1426, + "TTCAGC": 1427, + "TTCAGG": 1428, + "TTCTAA": 1429, + "TTCTAT": 1430, + "TTCTAC": 1431, + "TTCTAG": 1432, + "TTCTTA": 1433, + "TTCTTT": 1434, + "TTCTTC": 1435, + "TTCTTG": 1436, + "TTCTCA": 1437, + "TTCTCT": 1438, + "TTCTCC": 1439, + "TTCTCG": 1440, + "TTCTGA": 1441, + "TTCTGT": 1442, + "TTCTGC": 1443, + "TTCTGG": 1444, + "TTCCAA": 1445, + "TTCCAT": 1446, + "TTCCAC": 1447, + "TTCCAG": 1448, + "TTCCTA": 1449, + "TTCCTT": 1450, + "TTCCTC": 1451, + "TTCCTG": 1452, + "TTCCCA": 1453, + "TTCCCT": 1454, + "TTCCCC": 1455, + "TTCCCG": 1456, + "TTCCGA": 1457, + "TTCCGT": 1458, + "TTCCGC": 1459, + "TTCCGG": 1460, + "TTCGAA": 1461, + "TTCGAT": 1462, + "TTCGAC": 1463, + "TTCGAG": 1464, + "TTCGTA": 1465, + "TTCGTT": 1466, + "TTCGTC": 1467, + "TTCGTG": 1468, + "TTCGCA": 1469, + "TTCGCT": 1470, + "TTCGCC": 1471, + "TTCGCG": 1472, + "TTCGGA": 1473, + "TTCGGT": 1474, + "TTCGGC": 1475, + "TTCGGG": 1476, + "TTGAAA": 1477, + "TTGAAT": 1478, + "TTGAAC": 1479, + "TTGAAG": 1480, + "TTGATA": 1481, + "TTGATT": 1482, + "TTGATC": 1483, + "TTGATG": 1484, + "TTGACA": 1485, + "TTGACT": 1486, + "TTGACC": 1487, + "TTGACG": 1488, + "TTGAGA": 1489, + "TTGAGT": 1490, + "TTGAGC": 1491, + "TTGAGG": 1492, + "TTGTAA": 1493, + "TTGTAT": 1494, + "TTGTAC": 1495, + "TTGTAG": 1496, + "TTGTTA": 1497, + "TTGTTT": 1498, + "TTGTTC": 1499, + "TTGTTG": 1500, + "TTGTCA": 1501, + "TTGTCT": 1502, + "TTGTCC": 1503, + "TTGTCG": 1504, + "TTGTGA": 1505, + "TTGTGT": 1506, + "TTGTGC": 1507, + "TTGTGG": 1508, + "TTGCAA": 1509, + "TTGCAT": 1510, + "TTGCAC": 1511, + "TTGCAG": 1512, + "TTGCTA": 1513, + "TTGCTT": 1514, + "TTGCTC": 1515, + "TTGCTG": 1516, + "TTGCCA": 1517, + "TTGCCT": 1518, + "TTGCCC": 1519, + "TTGCCG": 1520, + "TTGCGA": 1521, + "TTGCGT": 1522, + "TTGCGC": 1523, + "TTGCGG": 1524, + "TTGGAA": 1525, + "TTGGAT": 1526, + "TTGGAC": 1527, + "TTGGAG": 1528, + "TTGGTA": 1529, + "TTGGTT": 1530, + "TTGGTC": 1531, + "TTGGTG": 1532, + "TTGGCA": 1533, + "TTGGCT": 1534, + "TTGGCC": 1535, + "TTGGCG": 1536, + "TTGGGA": 1537, + "TTGGGT": 1538, + "TTGGGC": 1539, + "TTGGGG": 1540, + "TCAAAA": 1541, + "TCAAAT": 1542, + "TCAAAC": 1543, + "TCAAAG": 1544, + "TCAATA": 1545, + "TCAATT": 1546, + "TCAATC": 1547, + "TCAATG": 1548, + "TCAACA": 1549, + "TCAACT": 1550, + "TCAACC": 1551, + "TCAACG": 1552, + "TCAAGA": 1553, + "TCAAGT": 1554, + "TCAAGC": 1555, + "TCAAGG": 1556, + "TCATAA": 1557, + "TCATAT": 1558, + "TCATAC": 1559, + "TCATAG": 1560, + "TCATTA": 1561, + "TCATTT": 1562, + "TCATTC": 1563, + "TCATTG": 1564, + "TCATCA": 1565, + "TCATCT": 1566, + "TCATCC": 1567, + "TCATCG": 1568, + "TCATGA": 1569, + "TCATGT": 1570, + "TCATGC": 1571, + "TCATGG": 1572, + "TCACAA": 1573, + "TCACAT": 1574, + "TCACAC": 1575, + "TCACAG": 1576, + "TCACTA": 1577, + "TCACTT": 1578, + "TCACTC": 1579, + "TCACTG": 1580, + "TCACCA": 1581, + "TCACCT": 1582, + "TCACCC": 1583, + "TCACCG": 1584, + "TCACGA": 1585, + "TCACGT": 1586, + "TCACGC": 1587, + "TCACGG": 1588, + "TCAGAA": 1589, + "TCAGAT": 1590, + "TCAGAC": 1591, + "TCAGAG": 1592, + "TCAGTA": 1593, + "TCAGTT": 1594, + "TCAGTC": 1595, + "TCAGTG": 1596, + "TCAGCA": 1597, + "TCAGCT": 1598, + "TCAGCC": 1599, + "TCAGCG": 1600, + "TCAGGA": 1601, + "TCAGGT": 1602, + "TCAGGC": 1603, + "TCAGGG": 1604, + "TCTAAA": 1605, + "TCTAAT": 1606, + "TCTAAC": 1607, + "TCTAAG": 1608, + "TCTATA": 1609, + "TCTATT": 1610, + "TCTATC": 1611, + "TCTATG": 1612, + "TCTACA": 1613, + "TCTACT": 1614, + "TCTACC": 1615, + "TCTACG": 1616, + "TCTAGA": 1617, + "TCTAGT": 1618, + "TCTAGC": 1619, + "TCTAGG": 1620, + "TCTTAA": 1621, + "TCTTAT": 1622, + "TCTTAC": 1623, + "TCTTAG": 1624, + "TCTTTA": 1625, + "TCTTTT": 1626, + "TCTTTC": 1627, + "TCTTTG": 1628, + "TCTTCA": 1629, + "TCTTCT": 1630, + "TCTTCC": 1631, + "TCTTCG": 1632, + "TCTTGA": 1633, + "TCTTGT": 1634, + "TCTTGC": 1635, + "TCTTGG": 1636, + "TCTCAA": 1637, + "TCTCAT": 1638, + "TCTCAC": 1639, + "TCTCAG": 1640, + "TCTCTA": 1641, + "TCTCTT": 1642, + "TCTCTC": 1643, + "TCTCTG": 1644, + "TCTCCA": 1645, + "TCTCCT": 1646, + "TCTCCC": 1647, + "TCTCCG": 1648, + "TCTCGA": 1649, + "TCTCGT": 1650, + "TCTCGC": 1651, + "TCTCGG": 1652, + "TCTGAA": 1653, + "TCTGAT": 1654, + "TCTGAC": 1655, + "TCTGAG": 1656, + "TCTGTA": 1657, + "TCTGTT": 1658, + "TCTGTC": 1659, + "TCTGTG": 1660, + "TCTGCA": 1661, + "TCTGCT": 1662, + "TCTGCC": 1663, + "TCTGCG": 1664, + "TCTGGA": 1665, + "TCTGGT": 1666, + "TCTGGC": 1667, + "TCTGGG": 1668, + "TCCAAA": 1669, + "TCCAAT": 1670, + "TCCAAC": 1671, + "TCCAAG": 1672, + "TCCATA": 1673, + "TCCATT": 1674, + "TCCATC": 1675, + "TCCATG": 1676, + "TCCACA": 1677, + "TCCACT": 1678, + "TCCACC": 1679, + "TCCACG": 1680, + "TCCAGA": 1681, + "TCCAGT": 1682, + "TCCAGC": 1683, + "TCCAGG": 1684, + "TCCTAA": 1685, + "TCCTAT": 1686, + "TCCTAC": 1687, + "TCCTAG": 1688, + "TCCTTA": 1689, + "TCCTTT": 1690, + "TCCTTC": 1691, + "TCCTTG": 1692, + "TCCTCA": 1693, + "TCCTCT": 1694, + "TCCTCC": 1695, + "TCCTCG": 1696, + "TCCTGA": 1697, + "TCCTGT": 1698, + "TCCTGC": 1699, + "TCCTGG": 1700, + "TCCCAA": 1701, + "TCCCAT": 1702, + "TCCCAC": 1703, + "TCCCAG": 1704, + "TCCCTA": 1705, + "TCCCTT": 1706, + "TCCCTC": 1707, + "TCCCTG": 1708, + "TCCCCA": 1709, + "TCCCCT": 1710, + "TCCCCC": 1711, + "TCCCCG": 1712, + "TCCCGA": 1713, + "TCCCGT": 1714, + "TCCCGC": 1715, + "TCCCGG": 1716, + "TCCGAA": 1717, + "TCCGAT": 1718, + "TCCGAC": 1719, + "TCCGAG": 1720, + "TCCGTA": 1721, + "TCCGTT": 1722, + "TCCGTC": 1723, + "TCCGTG": 1724, + "TCCGCA": 1725, + "TCCGCT": 1726, + "TCCGCC": 1727, + "TCCGCG": 1728, + "TCCGGA": 1729, + "TCCGGT": 1730, + "TCCGGC": 1731, + "TCCGGG": 1732, + "TCGAAA": 1733, + "TCGAAT": 1734, + "TCGAAC": 1735, + "TCGAAG": 1736, + "TCGATA": 1737, + "TCGATT": 1738, + "TCGATC": 1739, + "TCGATG": 1740, + "TCGACA": 1741, + "TCGACT": 1742, + "TCGACC": 1743, + "TCGACG": 1744, + "TCGAGA": 1745, + "TCGAGT": 1746, + "TCGAGC": 1747, + "TCGAGG": 1748, + "TCGTAA": 1749, + "TCGTAT": 1750, + "TCGTAC": 1751, + "TCGTAG": 1752, + "TCGTTA": 1753, + "TCGTTT": 1754, + "TCGTTC": 1755, + "TCGTTG": 1756, + "TCGTCA": 1757, + "TCGTCT": 1758, + "TCGTCC": 1759, + "TCGTCG": 1760, + "TCGTGA": 1761, + "TCGTGT": 1762, + "TCGTGC": 1763, + "TCGTGG": 1764, + "TCGCAA": 1765, + "TCGCAT": 1766, + "TCGCAC": 1767, + "TCGCAG": 1768, + "TCGCTA": 1769, + "TCGCTT": 1770, + "TCGCTC": 1771, + "TCGCTG": 1772, + "TCGCCA": 1773, + "TCGCCT": 1774, + "TCGCCC": 1775, + "TCGCCG": 1776, + "TCGCGA": 1777, + "TCGCGT": 1778, + "TCGCGC": 1779, + "TCGCGG": 1780, + "TCGGAA": 1781, + "TCGGAT": 1782, + "TCGGAC": 1783, + "TCGGAG": 1784, + "TCGGTA": 1785, + "TCGGTT": 1786, + "TCGGTC": 1787, + "TCGGTG": 1788, + "TCGGCA": 1789, + "TCGGCT": 1790, + "TCGGCC": 1791, + "TCGGCG": 1792, + "TCGGGA": 1793, + "TCGGGT": 1794, + "TCGGGC": 1795, + "TCGGGG": 1796, + "TGAAAA": 1797, + "TGAAAT": 1798, + "TGAAAC": 1799, + "TGAAAG": 1800, + "TGAATA": 1801, + "TGAATT": 1802, + "TGAATC": 1803, + "TGAATG": 1804, + "TGAACA": 1805, + "TGAACT": 1806, + "TGAACC": 1807, + "TGAACG": 1808, + "TGAAGA": 1809, + "TGAAGT": 1810, + "TGAAGC": 1811, + "TGAAGG": 1812, + "TGATAA": 1813, + "TGATAT": 1814, + "TGATAC": 1815, + "TGATAG": 1816, + "TGATTA": 1817, + "TGATTT": 1818, + "TGATTC": 1819, + "TGATTG": 1820, + "TGATCA": 1821, + "TGATCT": 1822, + "TGATCC": 1823, + "TGATCG": 1824, + "TGATGA": 1825, + "TGATGT": 1826, + "TGATGC": 1827, + "TGATGG": 1828, + "TGACAA": 1829, + "TGACAT": 1830, + "TGACAC": 1831, + "TGACAG": 1832, + "TGACTA": 1833, + "TGACTT": 1834, + "TGACTC": 1835, + "TGACTG": 1836, + "TGACCA": 1837, + "TGACCT": 1838, + "TGACCC": 1839, + "TGACCG": 1840, + "TGACGA": 1841, + "TGACGT": 1842, + "TGACGC": 1843, + "TGACGG": 1844, + "TGAGAA": 1845, + "TGAGAT": 1846, + "TGAGAC": 1847, + "TGAGAG": 1848, + "TGAGTA": 1849, + "TGAGTT": 1850, + "TGAGTC": 1851, + "TGAGTG": 1852, + "TGAGCA": 1853, + "TGAGCT": 1854, + "TGAGCC": 1855, + "TGAGCG": 1856, + "TGAGGA": 1857, + "TGAGGT": 1858, + "TGAGGC": 1859, + "TGAGGG": 1860, + "TGTAAA": 1861, + "TGTAAT": 1862, + "TGTAAC": 1863, + "TGTAAG": 1864, + "TGTATA": 1865, + "TGTATT": 1866, + "TGTATC": 1867, + "TGTATG": 1868, + "TGTACA": 1869, + "TGTACT": 1870, + "TGTACC": 1871, + "TGTACG": 1872, + "TGTAGA": 1873, + "TGTAGT": 1874, + "TGTAGC": 1875, + "TGTAGG": 1876, + "TGTTAA": 1877, + "TGTTAT": 1878, + "TGTTAC": 1879, + "TGTTAG": 1880, + "TGTTTA": 1881, + "TGTTTT": 1882, + "TGTTTC": 1883, + "TGTTTG": 1884, + "TGTTCA": 1885, + "TGTTCT": 1886, + "TGTTCC": 1887, + "TGTTCG": 1888, + "TGTTGA": 1889, + "TGTTGT": 1890, + "TGTTGC": 1891, + "TGTTGG": 1892, + "TGTCAA": 1893, + "TGTCAT": 1894, + "TGTCAC": 1895, + "TGTCAG": 1896, + "TGTCTA": 1897, + "TGTCTT": 1898, + "TGTCTC": 1899, + "TGTCTG": 1900, + "TGTCCA": 1901, + "TGTCCT": 1902, + "TGTCCC": 1903, + "TGTCCG": 1904, + "TGTCGA": 1905, + "TGTCGT": 1906, + "TGTCGC": 1907, + "TGTCGG": 1908, + "TGTGAA": 1909, + "TGTGAT": 1910, + "TGTGAC": 1911, + "TGTGAG": 1912, + "TGTGTA": 1913, + "TGTGTT": 1914, + "TGTGTC": 1915, + "TGTGTG": 1916, + "TGTGCA": 1917, + "TGTGCT": 1918, + "TGTGCC": 1919, + "TGTGCG": 1920, + "TGTGGA": 1921, + "TGTGGT": 1922, + "TGTGGC": 1923, + "TGTGGG": 1924, + "TGCAAA": 1925, + "TGCAAT": 1926, + "TGCAAC": 1927, + "TGCAAG": 1928, + "TGCATA": 1929, + "TGCATT": 1930, + "TGCATC": 1931, + "TGCATG": 1932, + "TGCACA": 1933, + "TGCACT": 1934, + "TGCACC": 1935, + "TGCACG": 1936, + "TGCAGA": 1937, + "TGCAGT": 1938, + "TGCAGC": 1939, + "TGCAGG": 1940, + "TGCTAA": 1941, + "TGCTAT": 1942, + "TGCTAC": 1943, + "TGCTAG": 1944, + "TGCTTA": 1945, + "TGCTTT": 1946, + "TGCTTC": 1947, + "TGCTTG": 1948, + "TGCTCA": 1949, + "TGCTCT": 1950, + "TGCTCC": 1951, + "TGCTCG": 1952, + "TGCTGA": 1953, + "TGCTGT": 1954, + "TGCTGC": 1955, + "TGCTGG": 1956, + "TGCCAA": 1957, + "TGCCAT": 1958, + "TGCCAC": 1959, + "TGCCAG": 1960, + "TGCCTA": 1961, + "TGCCTT": 1962, + "TGCCTC": 1963, + "TGCCTG": 1964, + "TGCCCA": 1965, + "TGCCCT": 1966, + "TGCCCC": 1967, + "TGCCCG": 1968, + "TGCCGA": 1969, + "TGCCGT": 1970, + "TGCCGC": 1971, + "TGCCGG": 1972, + "TGCGAA": 1973, + "TGCGAT": 1974, + "TGCGAC": 1975, + "TGCGAG": 1976, + "TGCGTA": 1977, + "TGCGTT": 1978, + "TGCGTC": 1979, + "TGCGTG": 1980, + "TGCGCA": 1981, + "TGCGCT": 1982, + "TGCGCC": 1983, + "TGCGCG": 1984, + "TGCGGA": 1985, + "TGCGGT": 1986, + "TGCGGC": 1987, + "TGCGGG": 1988, + "TGGAAA": 1989, + "TGGAAT": 1990, + "TGGAAC": 1991, + "TGGAAG": 1992, + "TGGATA": 1993, + "TGGATT": 1994, + "TGGATC": 1995, + "TGGATG": 1996, + "TGGACA": 1997, + "TGGACT": 1998, + "TGGACC": 1999, + "TGGACG": 2000, + "TGGAGA": 2001, + "TGGAGT": 2002, + "TGGAGC": 2003, + "TGGAGG": 2004, + "TGGTAA": 2005, + "TGGTAT": 2006, + "TGGTAC": 2007, + "TGGTAG": 2008, + "TGGTTA": 2009, + "TGGTTT": 2010, + "TGGTTC": 2011, + "TGGTTG": 2012, + "TGGTCA": 2013, + "TGGTCT": 2014, + "TGGTCC": 2015, + "TGGTCG": 2016, + "TGGTGA": 2017, + "TGGTGT": 2018, + "TGGTGC": 2019, + "TGGTGG": 2020, + "TGGCAA": 2021, + "TGGCAT": 2022, + "TGGCAC": 2023, + "TGGCAG": 2024, + "TGGCTA": 2025, + "TGGCTT": 2026, + "TGGCTC": 2027, + "TGGCTG": 2028, + "TGGCCA": 2029, + "TGGCCT": 2030, + "TGGCCC": 2031, + "TGGCCG": 2032, + "TGGCGA": 2033, + "TGGCGT": 2034, + "TGGCGC": 2035, + "TGGCGG": 2036, + "TGGGAA": 2037, + "TGGGAT": 2038, + "TGGGAC": 2039, + "TGGGAG": 2040, + "TGGGTA": 2041, + "TGGGTT": 2042, + "TGGGTC": 2043, + "TGGGTG": 2044, + "TGGGCA": 2045, + "TGGGCT": 2046, + "TGGGCC": 2047, + "TGGGCG": 2048, + "TGGGGA": 2049, + "TGGGGT": 2050, + "TGGGGC": 2051, + "TGGGGG": 2052, + "CAAAAA": 2053, + "CAAAAT": 2054, + "CAAAAC": 2055, + "CAAAAG": 2056, + "CAAATA": 2057, + "CAAATT": 2058, + "CAAATC": 2059, + "CAAATG": 2060, + "CAAACA": 2061, + "CAAACT": 2062, + "CAAACC": 2063, + "CAAACG": 2064, + "CAAAGA": 2065, + "CAAAGT": 2066, + "CAAAGC": 2067, + "CAAAGG": 2068, + "CAATAA": 2069, + "CAATAT": 2070, + "CAATAC": 2071, + "CAATAG": 2072, + "CAATTA": 2073, + "CAATTT": 2074, + "CAATTC": 2075, + "CAATTG": 2076, + "CAATCA": 2077, + "CAATCT": 2078, + "CAATCC": 2079, + "CAATCG": 2080, + "CAATGA": 2081, + "CAATGT": 2082, + "CAATGC": 2083, + "CAATGG": 2084, + "CAACAA": 2085, + "CAACAT": 2086, + "CAACAC": 2087, + "CAACAG": 2088, + "CAACTA": 2089, + "CAACTT": 2090, + "CAACTC": 2091, + "CAACTG": 2092, + "CAACCA": 2093, + "CAACCT": 2094, + "CAACCC": 2095, + "CAACCG": 2096, + "CAACGA": 2097, + "CAACGT": 2098, + "CAACGC": 2099, + "CAACGG": 2100, + "CAAGAA": 2101, + "CAAGAT": 2102, + "CAAGAC": 2103, + "CAAGAG": 2104, + "CAAGTA": 2105, + "CAAGTT": 2106, + "CAAGTC": 2107, + "CAAGTG": 2108, + "CAAGCA": 2109, + "CAAGCT": 2110, + "CAAGCC": 2111, + "CAAGCG": 2112, + "CAAGGA": 2113, + "CAAGGT": 2114, + "CAAGGC": 2115, + "CAAGGG": 2116, + "CATAAA": 2117, + "CATAAT": 2118, + "CATAAC": 2119, + "CATAAG": 2120, + "CATATA": 2121, + "CATATT": 2122, + "CATATC": 2123, + "CATATG": 2124, + "CATACA": 2125, + "CATACT": 2126, + "CATACC": 2127, + "CATACG": 2128, + "CATAGA": 2129, + "CATAGT": 2130, + "CATAGC": 2131, + "CATAGG": 2132, + "CATTAA": 2133, + "CATTAT": 2134, + "CATTAC": 2135, + "CATTAG": 2136, + "CATTTA": 2137, + "CATTTT": 2138, + "CATTTC": 2139, + "CATTTG": 2140, + "CATTCA": 2141, + "CATTCT": 2142, + "CATTCC": 2143, + "CATTCG": 2144, + "CATTGA": 2145, + "CATTGT": 2146, + "CATTGC": 2147, + "CATTGG": 2148, + "CATCAA": 2149, + "CATCAT": 2150, + "CATCAC": 2151, + "CATCAG": 2152, + "CATCTA": 2153, + "CATCTT": 2154, + "CATCTC": 2155, + "CATCTG": 2156, + "CATCCA": 2157, + "CATCCT": 2158, + "CATCCC": 2159, + "CATCCG": 2160, + "CATCGA": 2161, + "CATCGT": 2162, + "CATCGC": 2163, + "CATCGG": 2164, + "CATGAA": 2165, + "CATGAT": 2166, + "CATGAC": 2167, + "CATGAG": 2168, + "CATGTA": 2169, + "CATGTT": 2170, + "CATGTC": 2171, + "CATGTG": 2172, + "CATGCA": 2173, + "CATGCT": 2174, + "CATGCC": 2175, + "CATGCG": 2176, + "CATGGA": 2177, + "CATGGT": 2178, + "CATGGC": 2179, + "CATGGG": 2180, + "CACAAA": 2181, + "CACAAT": 2182, + "CACAAC": 2183, + "CACAAG": 2184, + "CACATA": 2185, + "CACATT": 2186, + "CACATC": 2187, + "CACATG": 2188, + "CACACA": 2189, + "CACACT": 2190, + "CACACC": 2191, + "CACACG": 2192, + "CACAGA": 2193, + "CACAGT": 2194, + "CACAGC": 2195, + "CACAGG": 2196, + "CACTAA": 2197, + "CACTAT": 2198, + "CACTAC": 2199, + "CACTAG": 2200, + "CACTTA": 2201, + "CACTTT": 2202, + "CACTTC": 2203, + "CACTTG": 2204, + "CACTCA": 2205, + "CACTCT": 2206, + "CACTCC": 2207, + "CACTCG": 2208, + "CACTGA": 2209, + "CACTGT": 2210, + "CACTGC": 2211, + "CACTGG": 2212, + "CACCAA": 2213, + "CACCAT": 2214, + "CACCAC": 2215, + "CACCAG": 2216, + "CACCTA": 2217, + "CACCTT": 2218, + "CACCTC": 2219, + "CACCTG": 2220, + "CACCCA": 2221, + "CACCCT": 2222, + "CACCCC": 2223, + "CACCCG": 2224, + "CACCGA": 2225, + "CACCGT": 2226, + "CACCGC": 2227, + "CACCGG": 2228, + "CACGAA": 2229, + "CACGAT": 2230, + "CACGAC": 2231, + "CACGAG": 2232, + "CACGTA": 2233, + "CACGTT": 2234, + "CACGTC": 2235, + "CACGTG": 2236, + "CACGCA": 2237, + "CACGCT": 2238, + "CACGCC": 2239, + "CACGCG": 2240, + "CACGGA": 2241, + "CACGGT": 2242, + "CACGGC": 2243, + "CACGGG": 2244, + "CAGAAA": 2245, + "CAGAAT": 2246, + "CAGAAC": 2247, + "CAGAAG": 2248, + "CAGATA": 2249, + "CAGATT": 2250, + "CAGATC": 2251, + "CAGATG": 2252, + "CAGACA": 2253, + "CAGACT": 2254, + "CAGACC": 2255, + "CAGACG": 2256, + "CAGAGA": 2257, + "CAGAGT": 2258, + "CAGAGC": 2259, + "CAGAGG": 2260, + "CAGTAA": 2261, + "CAGTAT": 2262, + "CAGTAC": 2263, + "CAGTAG": 2264, + "CAGTTA": 2265, + "CAGTTT": 2266, + "CAGTTC": 2267, + "CAGTTG": 2268, + "CAGTCA": 2269, + "CAGTCT": 2270, + "CAGTCC": 2271, + "CAGTCG": 2272, + "CAGTGA": 2273, + "CAGTGT": 2274, + "CAGTGC": 2275, + "CAGTGG": 2276, + "CAGCAA": 2277, + "CAGCAT": 2278, + "CAGCAC": 2279, + "CAGCAG": 2280, + "CAGCTA": 2281, + "CAGCTT": 2282, + "CAGCTC": 2283, + "CAGCTG": 2284, + "CAGCCA": 2285, + "CAGCCT": 2286, + "CAGCCC": 2287, + "CAGCCG": 2288, + "CAGCGA": 2289, + "CAGCGT": 2290, + "CAGCGC": 2291, + "CAGCGG": 2292, + "CAGGAA": 2293, + "CAGGAT": 2294, + "CAGGAC": 2295, + "CAGGAG": 2296, + "CAGGTA": 2297, + "CAGGTT": 2298, + "CAGGTC": 2299, + "CAGGTG": 2300, + "CAGGCA": 2301, + "CAGGCT": 2302, + "CAGGCC": 2303, + "CAGGCG": 2304, + "CAGGGA": 2305, + "CAGGGT": 2306, + "CAGGGC": 2307, + "CAGGGG": 2308, + "CTAAAA": 2309, + "CTAAAT": 2310, + "CTAAAC": 2311, + "CTAAAG": 2312, + "CTAATA": 2313, + "CTAATT": 2314, + "CTAATC": 2315, + "CTAATG": 2316, + "CTAACA": 2317, + "CTAACT": 2318, + "CTAACC": 2319, + "CTAACG": 2320, + "CTAAGA": 2321, + "CTAAGT": 2322, + "CTAAGC": 2323, + "CTAAGG": 2324, + "CTATAA": 2325, + "CTATAT": 2326, + "CTATAC": 2327, + "CTATAG": 2328, + "CTATTA": 2329, + "CTATTT": 2330, + "CTATTC": 2331, + "CTATTG": 2332, + "CTATCA": 2333, + "CTATCT": 2334, + "CTATCC": 2335, + "CTATCG": 2336, + "CTATGA": 2337, + "CTATGT": 2338, + "CTATGC": 2339, + "CTATGG": 2340, + "CTACAA": 2341, + "CTACAT": 2342, + "CTACAC": 2343, + "CTACAG": 2344, + "CTACTA": 2345, + "CTACTT": 2346, + "CTACTC": 2347, + "CTACTG": 2348, + "CTACCA": 2349, + "CTACCT": 2350, + "CTACCC": 2351, + "CTACCG": 2352, + "CTACGA": 2353, + "CTACGT": 2354, + "CTACGC": 2355, + "CTACGG": 2356, + "CTAGAA": 2357, + "CTAGAT": 2358, + "CTAGAC": 2359, + "CTAGAG": 2360, + "CTAGTA": 2361, + "CTAGTT": 2362, + "CTAGTC": 2363, + "CTAGTG": 2364, + "CTAGCA": 2365, + "CTAGCT": 2366, + "CTAGCC": 2367, + "CTAGCG": 2368, + "CTAGGA": 2369, + "CTAGGT": 2370, + "CTAGGC": 2371, + "CTAGGG": 2372, + "CTTAAA": 2373, + "CTTAAT": 2374, + "CTTAAC": 2375, + "CTTAAG": 2376, + "CTTATA": 2377, + "CTTATT": 2378, + "CTTATC": 2379, + "CTTATG": 2380, + "CTTACA": 2381, + "CTTACT": 2382, + "CTTACC": 2383, + "CTTACG": 2384, + "CTTAGA": 2385, + "CTTAGT": 2386, + "CTTAGC": 2387, + "CTTAGG": 2388, + "CTTTAA": 2389, + "CTTTAT": 2390, + "CTTTAC": 2391, + "CTTTAG": 2392, + "CTTTTA": 2393, + "CTTTTT": 2394, + "CTTTTC": 2395, + "CTTTTG": 2396, + "CTTTCA": 2397, + "CTTTCT": 2398, + "CTTTCC": 2399, + "CTTTCG": 2400, + "CTTTGA": 2401, + "CTTTGT": 2402, + "CTTTGC": 2403, + "CTTTGG": 2404, + "CTTCAA": 2405, + "CTTCAT": 2406, + "CTTCAC": 2407, + "CTTCAG": 2408, + "CTTCTA": 2409, + "CTTCTT": 2410, + "CTTCTC": 2411, + "CTTCTG": 2412, + "CTTCCA": 2413, + "CTTCCT": 2414, + "CTTCCC": 2415, + "CTTCCG": 2416, + "CTTCGA": 2417, + "CTTCGT": 2418, + "CTTCGC": 2419, + "CTTCGG": 2420, + "CTTGAA": 2421, + "CTTGAT": 2422, + "CTTGAC": 2423, + "CTTGAG": 2424, + "CTTGTA": 2425, + "CTTGTT": 2426, + "CTTGTC": 2427, + "CTTGTG": 2428, + "CTTGCA": 2429, + "CTTGCT": 2430, + "CTTGCC": 2431, + "CTTGCG": 2432, + "CTTGGA": 2433, + "CTTGGT": 2434, + "CTTGGC": 2435, + "CTTGGG": 2436, + "CTCAAA": 2437, + "CTCAAT": 2438, + "CTCAAC": 2439, + "CTCAAG": 2440, + "CTCATA": 2441, + "CTCATT": 2442, + "CTCATC": 2443, + "CTCATG": 2444, + "CTCACA": 2445, + "CTCACT": 2446, + "CTCACC": 2447, + "CTCACG": 2448, + "CTCAGA": 2449, + "CTCAGT": 2450, + "CTCAGC": 2451, + "CTCAGG": 2452, + "CTCTAA": 2453, + "CTCTAT": 2454, + "CTCTAC": 2455, + "CTCTAG": 2456, + "CTCTTA": 2457, + "CTCTTT": 2458, + "CTCTTC": 2459, + "CTCTTG": 2460, + "CTCTCA": 2461, + "CTCTCT": 2462, + "CTCTCC": 2463, + "CTCTCG": 2464, + "CTCTGA": 2465, + "CTCTGT": 2466, + "CTCTGC": 2467, + "CTCTGG": 2468, + "CTCCAA": 2469, + "CTCCAT": 2470, + "CTCCAC": 2471, + "CTCCAG": 2472, + "CTCCTA": 2473, + "CTCCTT": 2474, + "CTCCTC": 2475, + "CTCCTG": 2476, + "CTCCCA": 2477, + "CTCCCT": 2478, + "CTCCCC": 2479, + "CTCCCG": 2480, + "CTCCGA": 2481, + "CTCCGT": 2482, + "CTCCGC": 2483, + "CTCCGG": 2484, + "CTCGAA": 2485, + "CTCGAT": 2486, + "CTCGAC": 2487, + "CTCGAG": 2488, + "CTCGTA": 2489, + "CTCGTT": 2490, + "CTCGTC": 2491, + "CTCGTG": 2492, + "CTCGCA": 2493, + "CTCGCT": 2494, + "CTCGCC": 2495, + "CTCGCG": 2496, + "CTCGGA": 2497, + "CTCGGT": 2498, + "CTCGGC": 2499, + "CTCGGG": 2500, + "CTGAAA": 2501, + "CTGAAT": 2502, + "CTGAAC": 2503, + "CTGAAG": 2504, + "CTGATA": 2505, + "CTGATT": 2506, + "CTGATC": 2507, + "CTGATG": 2508, + "CTGACA": 2509, + "CTGACT": 2510, + "CTGACC": 2511, + "CTGACG": 2512, + "CTGAGA": 2513, + "CTGAGT": 2514, + "CTGAGC": 2515, + "CTGAGG": 2516, + "CTGTAA": 2517, + "CTGTAT": 2518, + "CTGTAC": 2519, + "CTGTAG": 2520, + "CTGTTA": 2521, + "CTGTTT": 2522, + "CTGTTC": 2523, + "CTGTTG": 2524, + "CTGTCA": 2525, + "CTGTCT": 2526, + "CTGTCC": 2527, + "CTGTCG": 2528, + "CTGTGA": 2529, + "CTGTGT": 2530, + "CTGTGC": 2531, + "CTGTGG": 2532, + "CTGCAA": 2533, + "CTGCAT": 2534, + "CTGCAC": 2535, + "CTGCAG": 2536, + "CTGCTA": 2537, + "CTGCTT": 2538, + "CTGCTC": 2539, + "CTGCTG": 2540, + "CTGCCA": 2541, + "CTGCCT": 2542, + "CTGCCC": 2543, + "CTGCCG": 2544, + "CTGCGA": 2545, + "CTGCGT": 2546, + "CTGCGC": 2547, + "CTGCGG": 2548, + "CTGGAA": 2549, + "CTGGAT": 2550, + "CTGGAC": 2551, + "CTGGAG": 2552, + "CTGGTA": 2553, + "CTGGTT": 2554, + "CTGGTC": 2555, + "CTGGTG": 2556, + "CTGGCA": 2557, + "CTGGCT": 2558, + "CTGGCC": 2559, + "CTGGCG": 2560, + "CTGGGA": 2561, + "CTGGGT": 2562, + "CTGGGC": 2563, + "CTGGGG": 2564, + "CCAAAA": 2565, + "CCAAAT": 2566, + "CCAAAC": 2567, + "CCAAAG": 2568, + "CCAATA": 2569, + "CCAATT": 2570, + "CCAATC": 2571, + "CCAATG": 2572, + "CCAACA": 2573, + "CCAACT": 2574, + "CCAACC": 2575, + "CCAACG": 2576, + "CCAAGA": 2577, + "CCAAGT": 2578, + "CCAAGC": 2579, + "CCAAGG": 2580, + "CCATAA": 2581, + "CCATAT": 2582, + "CCATAC": 2583, + "CCATAG": 2584, + "CCATTA": 2585, + "CCATTT": 2586, + "CCATTC": 2587, + "CCATTG": 2588, + "CCATCA": 2589, + "CCATCT": 2590, + "CCATCC": 2591, + "CCATCG": 2592, + "CCATGA": 2593, + "CCATGT": 2594, + "CCATGC": 2595, + "CCATGG": 2596, + "CCACAA": 2597, + "CCACAT": 2598, + "CCACAC": 2599, + "CCACAG": 2600, + "CCACTA": 2601, + "CCACTT": 2602, + "CCACTC": 2603, + "CCACTG": 2604, + "CCACCA": 2605, + "CCACCT": 2606, + "CCACCC": 2607, + "CCACCG": 2608, + "CCACGA": 2609, + "CCACGT": 2610, + "CCACGC": 2611, + "CCACGG": 2612, + "CCAGAA": 2613, + "CCAGAT": 2614, + "CCAGAC": 2615, + "CCAGAG": 2616, + "CCAGTA": 2617, + "CCAGTT": 2618, + "CCAGTC": 2619, + "CCAGTG": 2620, + "CCAGCA": 2621, + "CCAGCT": 2622, + "CCAGCC": 2623, + "CCAGCG": 2624, + "CCAGGA": 2625, + "CCAGGT": 2626, + "CCAGGC": 2627, + "CCAGGG": 2628, + "CCTAAA": 2629, + "CCTAAT": 2630, + "CCTAAC": 2631, + "CCTAAG": 2632, + "CCTATA": 2633, + "CCTATT": 2634, + "CCTATC": 2635, + "CCTATG": 2636, + "CCTACA": 2637, + "CCTACT": 2638, + "CCTACC": 2639, + "CCTACG": 2640, + "CCTAGA": 2641, + "CCTAGT": 2642, + "CCTAGC": 2643, + "CCTAGG": 2644, + "CCTTAA": 2645, + "CCTTAT": 2646, + "CCTTAC": 2647, + "CCTTAG": 2648, + "CCTTTA": 2649, + "CCTTTT": 2650, + "CCTTTC": 2651, + "CCTTTG": 2652, + "CCTTCA": 2653, + "CCTTCT": 2654, + "CCTTCC": 2655, + "CCTTCG": 2656, + "CCTTGA": 2657, + "CCTTGT": 2658, + "CCTTGC": 2659, + "CCTTGG": 2660, + "CCTCAA": 2661, + "CCTCAT": 2662, + "CCTCAC": 2663, + "CCTCAG": 2664, + "CCTCTA": 2665, + "CCTCTT": 2666, + "CCTCTC": 2667, + "CCTCTG": 2668, + "CCTCCA": 2669, + "CCTCCT": 2670, + "CCTCCC": 2671, + "CCTCCG": 2672, + "CCTCGA": 2673, + "CCTCGT": 2674, + "CCTCGC": 2675, + "CCTCGG": 2676, + "CCTGAA": 2677, + "CCTGAT": 2678, + "CCTGAC": 2679, + "CCTGAG": 2680, + "CCTGTA": 2681, + "CCTGTT": 2682, + "CCTGTC": 2683, + "CCTGTG": 2684, + "CCTGCA": 2685, + "CCTGCT": 2686, + "CCTGCC": 2687, + "CCTGCG": 2688, + "CCTGGA": 2689, + "CCTGGT": 2690, + "CCTGGC": 2691, + "CCTGGG": 2692, + "CCCAAA": 2693, + "CCCAAT": 2694, + "CCCAAC": 2695, + "CCCAAG": 2696, + "CCCATA": 2697, + "CCCATT": 2698, + "CCCATC": 2699, + "CCCATG": 2700, + "CCCACA": 2701, + "CCCACT": 2702, + "CCCACC": 2703, + "CCCACG": 2704, + "CCCAGA": 2705, + "CCCAGT": 2706, + "CCCAGC": 2707, + "CCCAGG": 2708, + "CCCTAA": 2709, + "CCCTAT": 2710, + "CCCTAC": 2711, + "CCCTAG": 2712, + "CCCTTA": 2713, + "CCCTTT": 2714, + "CCCTTC": 2715, + "CCCTTG": 2716, + "CCCTCA": 2717, + "CCCTCT": 2718, + "CCCTCC": 2719, + "CCCTCG": 2720, + "CCCTGA": 2721, + "CCCTGT": 2722, + "CCCTGC": 2723, + "CCCTGG": 2724, + "CCCCAA": 2725, + "CCCCAT": 2726, + "CCCCAC": 2727, + "CCCCAG": 2728, + "CCCCTA": 2729, + "CCCCTT": 2730, + "CCCCTC": 2731, + "CCCCTG": 2732, + "CCCCCA": 2733, + "CCCCCT": 2734, + "CCCCCC": 2735, + "CCCCCG": 2736, + "CCCCGA": 2737, + "CCCCGT": 2738, + "CCCCGC": 2739, + "CCCCGG": 2740, + "CCCGAA": 2741, + "CCCGAT": 2742, + "CCCGAC": 2743, + "CCCGAG": 2744, + "CCCGTA": 2745, + "CCCGTT": 2746, + "CCCGTC": 2747, + "CCCGTG": 2748, + "CCCGCA": 2749, + "CCCGCT": 2750, + "CCCGCC": 2751, + "CCCGCG": 2752, + "CCCGGA": 2753, + "CCCGGT": 2754, + "CCCGGC": 2755, + "CCCGGG": 2756, + "CCGAAA": 2757, + "CCGAAT": 2758, + "CCGAAC": 2759, + "CCGAAG": 2760, + "CCGATA": 2761, + "CCGATT": 2762, + "CCGATC": 2763, + "CCGATG": 2764, + "CCGACA": 2765, + "CCGACT": 2766, + "CCGACC": 2767, + "CCGACG": 2768, + "CCGAGA": 2769, + "CCGAGT": 2770, + "CCGAGC": 2771, + "CCGAGG": 2772, + "CCGTAA": 2773, + "CCGTAT": 2774, + "CCGTAC": 2775, + "CCGTAG": 2776, + "CCGTTA": 2777, + "CCGTTT": 2778, + "CCGTTC": 2779, + "CCGTTG": 2780, + "CCGTCA": 2781, + "CCGTCT": 2782, + "CCGTCC": 2783, + "CCGTCG": 2784, + "CCGTGA": 2785, + "CCGTGT": 2786, + "CCGTGC": 2787, + "CCGTGG": 2788, + "CCGCAA": 2789, + "CCGCAT": 2790, + "CCGCAC": 2791, + "CCGCAG": 2792, + "CCGCTA": 2793, + "CCGCTT": 2794, + "CCGCTC": 2795, + "CCGCTG": 2796, + "CCGCCA": 2797, + "CCGCCT": 2798, + "CCGCCC": 2799, + "CCGCCG": 2800, + "CCGCGA": 2801, + "CCGCGT": 2802, + "CCGCGC": 2803, + "CCGCGG": 2804, + "CCGGAA": 2805, + "CCGGAT": 2806, + "CCGGAC": 2807, + "CCGGAG": 2808, + "CCGGTA": 2809, + "CCGGTT": 2810, + "CCGGTC": 2811, + "CCGGTG": 2812, + "CCGGCA": 2813, + "CCGGCT": 2814, + "CCGGCC": 2815, + "CCGGCG": 2816, + "CCGGGA": 2817, + "CCGGGT": 2818, + "CCGGGC": 2819, + "CCGGGG": 2820, + "CGAAAA": 2821, + "CGAAAT": 2822, + "CGAAAC": 2823, + "CGAAAG": 2824, + "CGAATA": 2825, + "CGAATT": 2826, + "CGAATC": 2827, + "CGAATG": 2828, + "CGAACA": 2829, + "CGAACT": 2830, + "CGAACC": 2831, + "CGAACG": 2832, + "CGAAGA": 2833, + "CGAAGT": 2834, + "CGAAGC": 2835, + "CGAAGG": 2836, + "CGATAA": 2837, + "CGATAT": 2838, + "CGATAC": 2839, + "CGATAG": 2840, + "CGATTA": 2841, + "CGATTT": 2842, + "CGATTC": 2843, + "CGATTG": 2844, + "CGATCA": 2845, + "CGATCT": 2846, + "CGATCC": 2847, + "CGATCG": 2848, + "CGATGA": 2849, + "CGATGT": 2850, + "CGATGC": 2851, + "CGATGG": 2852, + "CGACAA": 2853, + "CGACAT": 2854, + "CGACAC": 2855, + "CGACAG": 2856, + "CGACTA": 2857, + "CGACTT": 2858, + "CGACTC": 2859, + "CGACTG": 2860, + "CGACCA": 2861, + "CGACCT": 2862, + "CGACCC": 2863, + "CGACCG": 2864, + "CGACGA": 2865, + "CGACGT": 2866, + "CGACGC": 2867, + "CGACGG": 2868, + "CGAGAA": 2869, + "CGAGAT": 2870, + "CGAGAC": 2871, + "CGAGAG": 2872, + "CGAGTA": 2873, + "CGAGTT": 2874, + "CGAGTC": 2875, + "CGAGTG": 2876, + "CGAGCA": 2877, + "CGAGCT": 2878, + "CGAGCC": 2879, + "CGAGCG": 2880, + "CGAGGA": 2881, + "CGAGGT": 2882, + "CGAGGC": 2883, + "CGAGGG": 2884, + "CGTAAA": 2885, + "CGTAAT": 2886, + "CGTAAC": 2887, + "CGTAAG": 2888, + "CGTATA": 2889, + "CGTATT": 2890, + "CGTATC": 2891, + "CGTATG": 2892, + "CGTACA": 2893, + "CGTACT": 2894, + "CGTACC": 2895, + "CGTACG": 2896, + "CGTAGA": 2897, + "CGTAGT": 2898, + "CGTAGC": 2899, + "CGTAGG": 2900, + "CGTTAA": 2901, + "CGTTAT": 2902, + "CGTTAC": 2903, + "CGTTAG": 2904, + "CGTTTA": 2905, + "CGTTTT": 2906, + "CGTTTC": 2907, + "CGTTTG": 2908, + "CGTTCA": 2909, + "CGTTCT": 2910, + "CGTTCC": 2911, + "CGTTCG": 2912, + "CGTTGA": 2913, + "CGTTGT": 2914, + "CGTTGC": 2915, + "CGTTGG": 2916, + "CGTCAA": 2917, + "CGTCAT": 2918, + "CGTCAC": 2919, + "CGTCAG": 2920, + "CGTCTA": 2921, + "CGTCTT": 2922, + "CGTCTC": 2923, + "CGTCTG": 2924, + "CGTCCA": 2925, + "CGTCCT": 2926, + "CGTCCC": 2927, + "CGTCCG": 2928, + "CGTCGA": 2929, + "CGTCGT": 2930, + "CGTCGC": 2931, + "CGTCGG": 2932, + "CGTGAA": 2933, + "CGTGAT": 2934, + "CGTGAC": 2935, + "CGTGAG": 2936, + "CGTGTA": 2937, + "CGTGTT": 2938, + "CGTGTC": 2939, + "CGTGTG": 2940, + "CGTGCA": 2941, + "CGTGCT": 2942, + "CGTGCC": 2943, + "CGTGCG": 2944, + "CGTGGA": 2945, + "CGTGGT": 2946, + "CGTGGC": 2947, + "CGTGGG": 2948, + "CGCAAA": 2949, + "CGCAAT": 2950, + "CGCAAC": 2951, + "CGCAAG": 2952, + "CGCATA": 2953, + "CGCATT": 2954, + "CGCATC": 2955, + "CGCATG": 2956, + "CGCACA": 2957, + "CGCACT": 2958, + "CGCACC": 2959, + "CGCACG": 2960, + "CGCAGA": 2961, + "CGCAGT": 2962, + "CGCAGC": 2963, + "CGCAGG": 2964, + "CGCTAA": 2965, + "CGCTAT": 2966, + "CGCTAC": 2967, + "CGCTAG": 2968, + "CGCTTA": 2969, + "CGCTTT": 2970, + "CGCTTC": 2971, + "CGCTTG": 2972, + "CGCTCA": 2973, + "CGCTCT": 2974, + "CGCTCC": 2975, + "CGCTCG": 2976, + "CGCTGA": 2977, + "CGCTGT": 2978, + "CGCTGC": 2979, + "CGCTGG": 2980, + "CGCCAA": 2981, + "CGCCAT": 2982, + "CGCCAC": 2983, + "CGCCAG": 2984, + "CGCCTA": 2985, + "CGCCTT": 2986, + "CGCCTC": 2987, + "CGCCTG": 2988, + "CGCCCA": 2989, + "CGCCCT": 2990, + "CGCCCC": 2991, + "CGCCCG": 2992, + "CGCCGA": 2993, + "CGCCGT": 2994, + "CGCCGC": 2995, + "CGCCGG": 2996, + "CGCGAA": 2997, + "CGCGAT": 2998, + "CGCGAC": 2999, + "CGCGAG": 3000, + "CGCGTA": 3001, + "CGCGTT": 3002, + "CGCGTC": 3003, + "CGCGTG": 3004, + "CGCGCA": 3005, + "CGCGCT": 3006, + "CGCGCC": 3007, + "CGCGCG": 3008, + "CGCGGA": 3009, + "CGCGGT": 3010, + "CGCGGC": 3011, + "CGCGGG": 3012, + "CGGAAA": 3013, + "CGGAAT": 3014, + "CGGAAC": 3015, + "CGGAAG": 3016, + "CGGATA": 3017, + "CGGATT": 3018, + "CGGATC": 3019, + "CGGATG": 3020, + "CGGACA": 3021, + "CGGACT": 3022, + "CGGACC": 3023, + "CGGACG": 3024, + "CGGAGA": 3025, + "CGGAGT": 3026, + "CGGAGC": 3027, + "CGGAGG": 3028, + "CGGTAA": 3029, + "CGGTAT": 3030, + "CGGTAC": 3031, + "CGGTAG": 3032, + "CGGTTA": 3033, + "CGGTTT": 3034, + "CGGTTC": 3035, + "CGGTTG": 3036, + "CGGTCA": 3037, + "CGGTCT": 3038, + "CGGTCC": 3039, + "CGGTCG": 3040, + "CGGTGA": 3041, + "CGGTGT": 3042, + "CGGTGC": 3043, + "CGGTGG": 3044, + "CGGCAA": 3045, + "CGGCAT": 3046, + "CGGCAC": 3047, + "CGGCAG": 3048, + "CGGCTA": 3049, + "CGGCTT": 3050, + "CGGCTC": 3051, + "CGGCTG": 3052, + "CGGCCA": 3053, + "CGGCCT": 3054, + "CGGCCC": 3055, + "CGGCCG": 3056, + "CGGCGA": 3057, + "CGGCGT": 3058, + "CGGCGC": 3059, + "CGGCGG": 3060, + "CGGGAA": 3061, + "CGGGAT": 3062, + "CGGGAC": 3063, + "CGGGAG": 3064, + "CGGGTA": 3065, + "CGGGTT": 3066, + "CGGGTC": 3067, + "CGGGTG": 3068, + "CGGGCA": 3069, + "CGGGCT": 3070, + "CGGGCC": 3071, + "CGGGCG": 3072, + "CGGGGA": 3073, + "CGGGGT": 3074, + "CGGGGC": 3075, + "CGGGGG": 3076, + "GAAAAA": 3077, + "GAAAAT": 3078, + "GAAAAC": 3079, + "GAAAAG": 3080, + "GAAATA": 3081, + "GAAATT": 3082, + "GAAATC": 3083, + "GAAATG": 3084, + "GAAACA": 3085, + "GAAACT": 3086, + "GAAACC": 3087, + "GAAACG": 3088, + "GAAAGA": 3089, + "GAAAGT": 3090, + "GAAAGC": 3091, + "GAAAGG": 3092, + "GAATAA": 3093, + "GAATAT": 3094, + "GAATAC": 3095, + "GAATAG": 3096, + "GAATTA": 3097, + "GAATTT": 3098, + "GAATTC": 3099, + "GAATTG": 3100, + "GAATCA": 3101, + "GAATCT": 3102, + "GAATCC": 3103, + "GAATCG": 3104, + "GAATGA": 3105, + "GAATGT": 3106, + "GAATGC": 3107, + "GAATGG": 3108, + "GAACAA": 3109, + "GAACAT": 3110, + "GAACAC": 3111, + "GAACAG": 3112, + "GAACTA": 3113, + "GAACTT": 3114, + "GAACTC": 3115, + "GAACTG": 3116, + "GAACCA": 3117, + "GAACCT": 3118, + "GAACCC": 3119, + "GAACCG": 3120, + "GAACGA": 3121, + "GAACGT": 3122, + "GAACGC": 3123, + "GAACGG": 3124, + "GAAGAA": 3125, + "GAAGAT": 3126, + "GAAGAC": 3127, + "GAAGAG": 3128, + "GAAGTA": 3129, + "GAAGTT": 3130, + "GAAGTC": 3131, + "GAAGTG": 3132, + "GAAGCA": 3133, + "GAAGCT": 3134, + "GAAGCC": 3135, + "GAAGCG": 3136, + "GAAGGA": 3137, + "GAAGGT": 3138, + "GAAGGC": 3139, + "GAAGGG": 3140, + "GATAAA": 3141, + "GATAAT": 3142, + "GATAAC": 3143, + "GATAAG": 3144, + "GATATA": 3145, + "GATATT": 3146, + "GATATC": 3147, + "GATATG": 3148, + "GATACA": 3149, + "GATACT": 3150, + "GATACC": 3151, + "GATACG": 3152, + "GATAGA": 3153, + "GATAGT": 3154, + "GATAGC": 3155, + "GATAGG": 3156, + "GATTAA": 3157, + "GATTAT": 3158, + "GATTAC": 3159, + "GATTAG": 3160, + "GATTTA": 3161, + "GATTTT": 3162, + "GATTTC": 3163, + "GATTTG": 3164, + "GATTCA": 3165, + "GATTCT": 3166, + "GATTCC": 3167, + "GATTCG": 3168, + "GATTGA": 3169, + "GATTGT": 3170, + "GATTGC": 3171, + "GATTGG": 3172, + "GATCAA": 3173, + "GATCAT": 3174, + "GATCAC": 3175, + "GATCAG": 3176, + "GATCTA": 3177, + "GATCTT": 3178, + "GATCTC": 3179, + "GATCTG": 3180, + "GATCCA": 3181, + "GATCCT": 3182, + "GATCCC": 3183, + "GATCCG": 3184, + "GATCGA": 3185, + "GATCGT": 3186, + "GATCGC": 3187, + "GATCGG": 3188, + "GATGAA": 3189, + "GATGAT": 3190, + "GATGAC": 3191, + "GATGAG": 3192, + "GATGTA": 3193, + "GATGTT": 3194, + "GATGTC": 3195, + "GATGTG": 3196, + "GATGCA": 3197, + "GATGCT": 3198, + "GATGCC": 3199, + "GATGCG": 3200, + "GATGGA": 3201, + "GATGGT": 3202, + "GATGGC": 3203, + "GATGGG": 3204, + "GACAAA": 3205, + "GACAAT": 3206, + "GACAAC": 3207, + "GACAAG": 3208, + "GACATA": 3209, + "GACATT": 3210, + "GACATC": 3211, + "GACATG": 3212, + "GACACA": 3213, + "GACACT": 3214, + "GACACC": 3215, + "GACACG": 3216, + "GACAGA": 3217, + "GACAGT": 3218, + "GACAGC": 3219, + "GACAGG": 3220, + "GACTAA": 3221, + "GACTAT": 3222, + "GACTAC": 3223, + "GACTAG": 3224, + "GACTTA": 3225, + "GACTTT": 3226, + "GACTTC": 3227, + "GACTTG": 3228, + "GACTCA": 3229, + "GACTCT": 3230, + "GACTCC": 3231, + "GACTCG": 3232, + "GACTGA": 3233, + "GACTGT": 3234, + "GACTGC": 3235, + "GACTGG": 3236, + "GACCAA": 3237, + "GACCAT": 3238, + "GACCAC": 3239, + "GACCAG": 3240, + "GACCTA": 3241, + "GACCTT": 3242, + "GACCTC": 3243, + "GACCTG": 3244, + "GACCCA": 3245, + "GACCCT": 3246, + "GACCCC": 3247, + "GACCCG": 3248, + "GACCGA": 3249, + "GACCGT": 3250, + "GACCGC": 3251, + "GACCGG": 3252, + "GACGAA": 3253, + "GACGAT": 3254, + "GACGAC": 3255, + "GACGAG": 3256, + "GACGTA": 3257, + "GACGTT": 3258, + "GACGTC": 3259, + "GACGTG": 3260, + "GACGCA": 3261, + "GACGCT": 3262, + "GACGCC": 3263, + "GACGCG": 3264, + "GACGGA": 3265, + "GACGGT": 3266, + "GACGGC": 3267, + "GACGGG": 3268, + "GAGAAA": 3269, + "GAGAAT": 3270, + "GAGAAC": 3271, + "GAGAAG": 3272, + "GAGATA": 3273, + "GAGATT": 3274, + "GAGATC": 3275, + "GAGATG": 3276, + "GAGACA": 3277, + "GAGACT": 3278, + "GAGACC": 3279, + "GAGACG": 3280, + "GAGAGA": 3281, + "GAGAGT": 3282, + "GAGAGC": 3283, + "GAGAGG": 3284, + "GAGTAA": 3285, + "GAGTAT": 3286, + "GAGTAC": 3287, + "GAGTAG": 3288, + "GAGTTA": 3289, + "GAGTTT": 3290, + "GAGTTC": 3291, + "GAGTTG": 3292, + "GAGTCA": 3293, + "GAGTCT": 3294, + "GAGTCC": 3295, + "GAGTCG": 3296, + "GAGTGA": 3297, + "GAGTGT": 3298, + "GAGTGC": 3299, + "GAGTGG": 3300, + "GAGCAA": 3301, + "GAGCAT": 3302, + "GAGCAC": 3303, + "GAGCAG": 3304, + "GAGCTA": 3305, + "GAGCTT": 3306, + "GAGCTC": 3307, + "GAGCTG": 3308, + "GAGCCA": 3309, + "GAGCCT": 3310, + "GAGCCC": 3311, + "GAGCCG": 3312, + "GAGCGA": 3313, + "GAGCGT": 3314, + "GAGCGC": 3315, + "GAGCGG": 3316, + "GAGGAA": 3317, + "GAGGAT": 3318, + "GAGGAC": 3319, + "GAGGAG": 3320, + "GAGGTA": 3321, + "GAGGTT": 3322, + "GAGGTC": 3323, + "GAGGTG": 3324, + "GAGGCA": 3325, + "GAGGCT": 3326, + "GAGGCC": 3327, + "GAGGCG": 3328, + "GAGGGA": 3329, + "GAGGGT": 3330, + "GAGGGC": 3331, + "GAGGGG": 3332, + "GTAAAA": 3333, + "GTAAAT": 3334, + "GTAAAC": 3335, + "GTAAAG": 3336, + "GTAATA": 3337, + "GTAATT": 3338, + "GTAATC": 3339, + "GTAATG": 3340, + "GTAACA": 3341, + "GTAACT": 3342, + "GTAACC": 3343, + "GTAACG": 3344, + "GTAAGA": 3345, + "GTAAGT": 3346, + "GTAAGC": 3347, + "GTAAGG": 3348, + "GTATAA": 3349, + "GTATAT": 3350, + "GTATAC": 3351, + "GTATAG": 3352, + "GTATTA": 3353, + "GTATTT": 3354, + "GTATTC": 3355, + "GTATTG": 3356, + "GTATCA": 3357, + "GTATCT": 3358, + "GTATCC": 3359, + "GTATCG": 3360, + "GTATGA": 3361, + "GTATGT": 3362, + "GTATGC": 3363, + "GTATGG": 3364, + "GTACAA": 3365, + "GTACAT": 3366, + "GTACAC": 3367, + "GTACAG": 3368, + "GTACTA": 3369, + "GTACTT": 3370, + "GTACTC": 3371, + "GTACTG": 3372, + "GTACCA": 3373, + "GTACCT": 3374, + "GTACCC": 3375, + "GTACCG": 3376, + "GTACGA": 3377, + "GTACGT": 3378, + "GTACGC": 3379, + "GTACGG": 3380, + "GTAGAA": 3381, + "GTAGAT": 3382, + "GTAGAC": 3383, + "GTAGAG": 3384, + "GTAGTA": 3385, + "GTAGTT": 3386, + "GTAGTC": 3387, + "GTAGTG": 3388, + "GTAGCA": 3389, + "GTAGCT": 3390, + "GTAGCC": 3391, + "GTAGCG": 3392, + "GTAGGA": 3393, + "GTAGGT": 3394, + "GTAGGC": 3395, + "GTAGGG": 3396, + "GTTAAA": 3397, + "GTTAAT": 3398, + "GTTAAC": 3399, + "GTTAAG": 3400, + "GTTATA": 3401, + "GTTATT": 3402, + "GTTATC": 3403, + "GTTATG": 3404, + "GTTACA": 3405, + "GTTACT": 3406, + "GTTACC": 3407, + "GTTACG": 3408, + "GTTAGA": 3409, + "GTTAGT": 3410, + "GTTAGC": 3411, + "GTTAGG": 3412, + "GTTTAA": 3413, + "GTTTAT": 3414, + "GTTTAC": 3415, + "GTTTAG": 3416, + "GTTTTA": 3417, + "GTTTTT": 3418, + "GTTTTC": 3419, + "GTTTTG": 3420, + "GTTTCA": 3421, + "GTTTCT": 3422, + "GTTTCC": 3423, + "GTTTCG": 3424, + "GTTTGA": 3425, + "GTTTGT": 3426, + "GTTTGC": 3427, + "GTTTGG": 3428, + "GTTCAA": 3429, + "GTTCAT": 3430, + "GTTCAC": 3431, + "GTTCAG": 3432, + "GTTCTA": 3433, + "GTTCTT": 3434, + "GTTCTC": 3435, + "GTTCTG": 3436, + "GTTCCA": 3437, + "GTTCCT": 3438, + "GTTCCC": 3439, + "GTTCCG": 3440, + "GTTCGA": 3441, + "GTTCGT": 3442, + "GTTCGC": 3443, + "GTTCGG": 3444, + "GTTGAA": 3445, + "GTTGAT": 3446, + "GTTGAC": 3447, + "GTTGAG": 3448, + "GTTGTA": 3449, + "GTTGTT": 3450, + "GTTGTC": 3451, + "GTTGTG": 3452, + "GTTGCA": 3453, + "GTTGCT": 3454, + "GTTGCC": 3455, + "GTTGCG": 3456, + "GTTGGA": 3457, + "GTTGGT": 3458, + "GTTGGC": 3459, + "GTTGGG": 3460, + "GTCAAA": 3461, + "GTCAAT": 3462, + "GTCAAC": 3463, + "GTCAAG": 3464, + "GTCATA": 3465, + "GTCATT": 3466, + "GTCATC": 3467, + "GTCATG": 3468, + "GTCACA": 3469, + "GTCACT": 3470, + "GTCACC": 3471, + "GTCACG": 3472, + "GTCAGA": 3473, + "GTCAGT": 3474, + "GTCAGC": 3475, + "GTCAGG": 3476, + "GTCTAA": 3477, + "GTCTAT": 3478, + "GTCTAC": 3479, + "GTCTAG": 3480, + "GTCTTA": 3481, + "GTCTTT": 3482, + "GTCTTC": 3483, + "GTCTTG": 3484, + "GTCTCA": 3485, + "GTCTCT": 3486, + "GTCTCC": 3487, + "GTCTCG": 3488, + "GTCTGA": 3489, + "GTCTGT": 3490, + "GTCTGC": 3491, + "GTCTGG": 3492, + "GTCCAA": 3493, + "GTCCAT": 3494, + "GTCCAC": 3495, + "GTCCAG": 3496, + "GTCCTA": 3497, + "GTCCTT": 3498, + "GTCCTC": 3499, + "GTCCTG": 3500, + "GTCCCA": 3501, + "GTCCCT": 3502, + "GTCCCC": 3503, + "GTCCCG": 3504, + "GTCCGA": 3505, + "GTCCGT": 3506, + "GTCCGC": 3507, + "GTCCGG": 3508, + "GTCGAA": 3509, + "GTCGAT": 3510, + "GTCGAC": 3511, + "GTCGAG": 3512, + "GTCGTA": 3513, + "GTCGTT": 3514, + "GTCGTC": 3515, + "GTCGTG": 3516, + "GTCGCA": 3517, + "GTCGCT": 3518, + "GTCGCC": 3519, + "GTCGCG": 3520, + "GTCGGA": 3521, + "GTCGGT": 3522, + "GTCGGC": 3523, + "GTCGGG": 3524, + "GTGAAA": 3525, + "GTGAAT": 3526, + "GTGAAC": 3527, + "GTGAAG": 3528, + "GTGATA": 3529, + "GTGATT": 3530, + "GTGATC": 3531, + "GTGATG": 3532, + "GTGACA": 3533, + "GTGACT": 3534, + "GTGACC": 3535, + "GTGACG": 3536, + "GTGAGA": 3537, + "GTGAGT": 3538, + "GTGAGC": 3539, + "GTGAGG": 3540, + "GTGTAA": 3541, + "GTGTAT": 3542, + "GTGTAC": 3543, + "GTGTAG": 3544, + "GTGTTA": 3545, + "GTGTTT": 3546, + "GTGTTC": 3547, + "GTGTTG": 3548, + "GTGTCA": 3549, + "GTGTCT": 3550, + "GTGTCC": 3551, + "GTGTCG": 3552, + "GTGTGA": 3553, + "GTGTGT": 3554, + "GTGTGC": 3555, + "GTGTGG": 3556, + "GTGCAA": 3557, + "GTGCAT": 3558, + "GTGCAC": 3559, + "GTGCAG": 3560, + "GTGCTA": 3561, + "GTGCTT": 3562, + "GTGCTC": 3563, + "GTGCTG": 3564, + "GTGCCA": 3565, + "GTGCCT": 3566, + "GTGCCC": 3567, + "GTGCCG": 3568, + "GTGCGA": 3569, + "GTGCGT": 3570, + "GTGCGC": 3571, + "GTGCGG": 3572, + "GTGGAA": 3573, + "GTGGAT": 3574, + "GTGGAC": 3575, + "GTGGAG": 3576, + "GTGGTA": 3577, + "GTGGTT": 3578, + "GTGGTC": 3579, + "GTGGTG": 3580, + "GTGGCA": 3581, + "GTGGCT": 3582, + "GTGGCC": 3583, + "GTGGCG": 3584, + "GTGGGA": 3585, + "GTGGGT": 3586, + "GTGGGC": 3587, + "GTGGGG": 3588, + "GCAAAA": 3589, + "GCAAAT": 3590, + "GCAAAC": 3591, + "GCAAAG": 3592, + "GCAATA": 3593, + "GCAATT": 3594, + "GCAATC": 3595, + "GCAATG": 3596, + "GCAACA": 3597, + "GCAACT": 3598, + "GCAACC": 3599, + "GCAACG": 3600, + "GCAAGA": 3601, + "GCAAGT": 3602, + "GCAAGC": 3603, + "GCAAGG": 3604, + "GCATAA": 3605, + "GCATAT": 3606, + "GCATAC": 3607, + "GCATAG": 3608, + "GCATTA": 3609, + "GCATTT": 3610, + "GCATTC": 3611, + "GCATTG": 3612, + "GCATCA": 3613, + "GCATCT": 3614, + "GCATCC": 3615, + "GCATCG": 3616, + "GCATGA": 3617, + "GCATGT": 3618, + "GCATGC": 3619, + "GCATGG": 3620, + "GCACAA": 3621, + "GCACAT": 3622, + "GCACAC": 3623, + "GCACAG": 3624, + "GCACTA": 3625, + "GCACTT": 3626, + "GCACTC": 3627, + "GCACTG": 3628, + "GCACCA": 3629, + "GCACCT": 3630, + "GCACCC": 3631, + "GCACCG": 3632, + "GCACGA": 3633, + "GCACGT": 3634, + "GCACGC": 3635, + "GCACGG": 3636, + "GCAGAA": 3637, + "GCAGAT": 3638, + "GCAGAC": 3639, + "GCAGAG": 3640, + "GCAGTA": 3641, + "GCAGTT": 3642, + "GCAGTC": 3643, + "GCAGTG": 3644, + "GCAGCA": 3645, + "GCAGCT": 3646, + "GCAGCC": 3647, + "GCAGCG": 3648, + "GCAGGA": 3649, + "GCAGGT": 3650, + "GCAGGC": 3651, + "GCAGGG": 3652, + "GCTAAA": 3653, + "GCTAAT": 3654, + "GCTAAC": 3655, + "GCTAAG": 3656, + "GCTATA": 3657, + "GCTATT": 3658, + "GCTATC": 3659, + "GCTATG": 3660, + "GCTACA": 3661, + "GCTACT": 3662, + "GCTACC": 3663, + "GCTACG": 3664, + "GCTAGA": 3665, + "GCTAGT": 3666, + "GCTAGC": 3667, + "GCTAGG": 3668, + "GCTTAA": 3669, + "GCTTAT": 3670, + "GCTTAC": 3671, + "GCTTAG": 3672, + "GCTTTA": 3673, + "GCTTTT": 3674, + "GCTTTC": 3675, + "GCTTTG": 3676, + "GCTTCA": 3677, + "GCTTCT": 3678, + "GCTTCC": 3679, + "GCTTCG": 3680, + "GCTTGA": 3681, + "GCTTGT": 3682, + "GCTTGC": 3683, + "GCTTGG": 3684, + "GCTCAA": 3685, + "GCTCAT": 3686, + "GCTCAC": 3687, + "GCTCAG": 3688, + "GCTCTA": 3689, + "GCTCTT": 3690, + "GCTCTC": 3691, + "GCTCTG": 3692, + "GCTCCA": 3693, + "GCTCCT": 3694, + "GCTCCC": 3695, + "GCTCCG": 3696, + "GCTCGA": 3697, + "GCTCGT": 3698, + "GCTCGC": 3699, + "GCTCGG": 3700, + "GCTGAA": 3701, + "GCTGAT": 3702, + "GCTGAC": 3703, + "GCTGAG": 3704, + "GCTGTA": 3705, + "GCTGTT": 3706, + "GCTGTC": 3707, + "GCTGTG": 3708, + "GCTGCA": 3709, + "GCTGCT": 3710, + "GCTGCC": 3711, + "GCTGCG": 3712, + "GCTGGA": 3713, + "GCTGGT": 3714, + "GCTGGC": 3715, + "GCTGGG": 3716, + "GCCAAA": 3717, + "GCCAAT": 3718, + "GCCAAC": 3719, + "GCCAAG": 3720, + "GCCATA": 3721, + "GCCATT": 3722, + "GCCATC": 3723, + "GCCATG": 3724, + "GCCACA": 3725, + "GCCACT": 3726, + "GCCACC": 3727, + "GCCACG": 3728, + "GCCAGA": 3729, + "GCCAGT": 3730, + "GCCAGC": 3731, + "GCCAGG": 3732, + "GCCTAA": 3733, + "GCCTAT": 3734, + "GCCTAC": 3735, + "GCCTAG": 3736, + "GCCTTA": 3737, + "GCCTTT": 3738, + "GCCTTC": 3739, + "GCCTTG": 3740, + "GCCTCA": 3741, + "GCCTCT": 3742, + "GCCTCC": 3743, + "GCCTCG": 3744, + "GCCTGA": 3745, + "GCCTGT": 3746, + "GCCTGC": 3747, + "GCCTGG": 3748, + "GCCCAA": 3749, + "GCCCAT": 3750, + "GCCCAC": 3751, + "GCCCAG": 3752, + "GCCCTA": 3753, + "GCCCTT": 3754, + "GCCCTC": 3755, + "GCCCTG": 3756, + "GCCCCA": 3757, + "GCCCCT": 3758, + "GCCCCC": 3759, + "GCCCCG": 3760, + "GCCCGA": 3761, + "GCCCGT": 3762, + "GCCCGC": 3763, + "GCCCGG": 3764, + "GCCGAA": 3765, + "GCCGAT": 3766, + "GCCGAC": 3767, + "GCCGAG": 3768, + "GCCGTA": 3769, + "GCCGTT": 3770, + "GCCGTC": 3771, + "GCCGTG": 3772, + "GCCGCA": 3773, + "GCCGCT": 3774, + "GCCGCC": 3775, + "GCCGCG": 3776, + "GCCGGA": 3777, + "GCCGGT": 3778, + "GCCGGC": 3779, + "GCCGGG": 3780, + "GCGAAA": 3781, + "GCGAAT": 3782, + "GCGAAC": 3783, + "GCGAAG": 3784, + "GCGATA": 3785, + "GCGATT": 3786, + "GCGATC": 3787, + "GCGATG": 3788, + "GCGACA": 3789, + "GCGACT": 3790, + "GCGACC": 3791, + "GCGACG": 3792, + "GCGAGA": 3793, + "GCGAGT": 3794, + "GCGAGC": 3795, + "GCGAGG": 3796, + "GCGTAA": 3797, + "GCGTAT": 3798, + "GCGTAC": 3799, + "GCGTAG": 3800, + "GCGTTA": 3801, + "GCGTTT": 3802, + "GCGTTC": 3803, + "GCGTTG": 3804, + "GCGTCA": 3805, + "GCGTCT": 3806, + "GCGTCC": 3807, + "GCGTCG": 3808, + "GCGTGA": 3809, + "GCGTGT": 3810, + "GCGTGC": 3811, + "GCGTGG": 3812, + "GCGCAA": 3813, + "GCGCAT": 3814, + "GCGCAC": 3815, + "GCGCAG": 3816, + "GCGCTA": 3817, + "GCGCTT": 3818, + "GCGCTC": 3819, + "GCGCTG": 3820, + "GCGCCA": 3821, + "GCGCCT": 3822, + "GCGCCC": 3823, + "GCGCCG": 3824, + "GCGCGA": 3825, + "GCGCGT": 3826, + "GCGCGC": 3827, + "GCGCGG": 3828, + "GCGGAA": 3829, + "GCGGAT": 3830, + "GCGGAC": 3831, + "GCGGAG": 3832, + "GCGGTA": 3833, + "GCGGTT": 3834, + "GCGGTC": 3835, + "GCGGTG": 3836, + "GCGGCA": 3837, + "GCGGCT": 3838, + "GCGGCC": 3839, + "GCGGCG": 3840, + "GCGGGA": 3841, + "GCGGGT": 3842, + "GCGGGC": 3843, + "GCGGGG": 3844, + "GGAAAA": 3845, + "GGAAAT": 3846, + "GGAAAC": 3847, + "GGAAAG": 3848, + "GGAATA": 3849, + "GGAATT": 3850, + "GGAATC": 3851, + "GGAATG": 3852, + "GGAACA": 3853, + "GGAACT": 3854, + "GGAACC": 3855, + "GGAACG": 3856, + "GGAAGA": 3857, + "GGAAGT": 3858, + "GGAAGC": 3859, + "GGAAGG": 3860, + "GGATAA": 3861, + "GGATAT": 3862, + "GGATAC": 3863, + "GGATAG": 3864, + "GGATTA": 3865, + "GGATTT": 3866, + "GGATTC": 3867, + "GGATTG": 3868, + "GGATCA": 3869, + "GGATCT": 3870, + "GGATCC": 3871, + "GGATCG": 3872, + "GGATGA": 3873, + "GGATGT": 3874, + "GGATGC": 3875, + "GGATGG": 3876, + "GGACAA": 3877, + "GGACAT": 3878, + "GGACAC": 3879, + "GGACAG": 3880, + "GGACTA": 3881, + "GGACTT": 3882, + "GGACTC": 3883, + "GGACTG": 3884, + "GGACCA": 3885, + "GGACCT": 3886, + "GGACCC": 3887, + "GGACCG": 3888, + "GGACGA": 3889, + "GGACGT": 3890, + "GGACGC": 3891, + "GGACGG": 3892, + "GGAGAA": 3893, + "GGAGAT": 3894, + "GGAGAC": 3895, + "GGAGAG": 3896, + "GGAGTA": 3897, + "GGAGTT": 3898, + "GGAGTC": 3899, + "GGAGTG": 3900, + "GGAGCA": 3901, + "GGAGCT": 3902, + "GGAGCC": 3903, + "GGAGCG": 3904, + "GGAGGA": 3905, + "GGAGGT": 3906, + "GGAGGC": 3907, + "GGAGGG": 3908, + "GGTAAA": 3909, + "GGTAAT": 3910, + "GGTAAC": 3911, + "GGTAAG": 3912, + "GGTATA": 3913, + "GGTATT": 3914, + "GGTATC": 3915, + "GGTATG": 3916, + "GGTACA": 3917, + "GGTACT": 3918, + "GGTACC": 3919, + "GGTACG": 3920, + "GGTAGA": 3921, + "GGTAGT": 3922, + "GGTAGC": 3923, + "GGTAGG": 3924, + "GGTTAA": 3925, + "GGTTAT": 3926, + "GGTTAC": 3927, + "GGTTAG": 3928, + "GGTTTA": 3929, + "GGTTTT": 3930, + "GGTTTC": 3931, + "GGTTTG": 3932, + "GGTTCA": 3933, + "GGTTCT": 3934, + "GGTTCC": 3935, + "GGTTCG": 3936, + "GGTTGA": 3937, + "GGTTGT": 3938, + "GGTTGC": 3939, + "GGTTGG": 3940, + "GGTCAA": 3941, + "GGTCAT": 3942, + "GGTCAC": 3943, + "GGTCAG": 3944, + "GGTCTA": 3945, + "GGTCTT": 3946, + "GGTCTC": 3947, + "GGTCTG": 3948, + "GGTCCA": 3949, + "GGTCCT": 3950, + "GGTCCC": 3951, + "GGTCCG": 3952, + "GGTCGA": 3953, + "GGTCGT": 3954, + "GGTCGC": 3955, + "GGTCGG": 3956, + "GGTGAA": 3957, + "GGTGAT": 3958, + "GGTGAC": 3959, + "GGTGAG": 3960, + "GGTGTA": 3961, + "GGTGTT": 3962, + "GGTGTC": 3963, + "GGTGTG": 3964, + "GGTGCA": 3965, + "GGTGCT": 3966, + "GGTGCC": 3967, + "GGTGCG": 3968, + "GGTGGA": 3969, + "GGTGGT": 3970, + "GGTGGC": 3971, + "GGTGGG": 3972, + "GGCAAA": 3973, + "GGCAAT": 3974, + "GGCAAC": 3975, + "GGCAAG": 3976, + "GGCATA": 3977, + "GGCATT": 3978, + "GGCATC": 3979, + "GGCATG": 3980, + "GGCACA": 3981, + "GGCACT": 3982, + "GGCACC": 3983, + "GGCACG": 3984, + "GGCAGA": 3985, + "GGCAGT": 3986, + "GGCAGC": 3987, + "GGCAGG": 3988, + "GGCTAA": 3989, + "GGCTAT": 3990, + "GGCTAC": 3991, + "GGCTAG": 3992, + "GGCTTA": 3993, + "GGCTTT": 3994, + "GGCTTC": 3995, + "GGCTTG": 3996, + "GGCTCA": 3997, + "GGCTCT": 3998, + "GGCTCC": 3999, + "GGCTCG": 4000, + "GGCTGA": 4001, + "GGCTGT": 4002, + "GGCTGC": 4003, + "GGCTGG": 4004, + "GGCCAA": 4005, + "GGCCAT": 4006, + "GGCCAC": 4007, + "GGCCAG": 4008, + "GGCCTA": 4009, + "GGCCTT": 4010, + "GGCCTC": 4011, + "GGCCTG": 4012, + "GGCCCA": 4013, + "GGCCCT": 4014, + "GGCCCC": 4015, + "GGCCCG": 4016, + "GGCCGA": 4017, + "GGCCGT": 4018, + "GGCCGC": 4019, + "GGCCGG": 4020, + "GGCGAA": 4021, + "GGCGAT": 4022, + "GGCGAC": 4023, + "GGCGAG": 4024, + "GGCGTA": 4025, + "GGCGTT": 4026, + "GGCGTC": 4027, + "GGCGTG": 4028, + "GGCGCA": 4029, + "GGCGCT": 4030, + "GGCGCC": 4031, + "GGCGCG": 4032, + "GGCGGA": 4033, + "GGCGGT": 4034, + "GGCGGC": 4035, + "GGCGGG": 4036, + "GGGAAA": 4037, + "GGGAAT": 4038, + "GGGAAC": 4039, + "GGGAAG": 4040, + "GGGATA": 4041, + "GGGATT": 4042, + "GGGATC": 4043, + "GGGATG": 4044, + "GGGACA": 4045, + "GGGACT": 4046, + "GGGACC": 4047, + "GGGACG": 4048, + "GGGAGA": 4049, + "GGGAGT": 4050, + "GGGAGC": 4051, + "GGGAGG": 4052, + "GGGTAA": 4053, + "GGGTAT": 4054, + "GGGTAC": 4055, + "GGGTAG": 4056, + "GGGTTA": 4057, + "GGGTTT": 4058, + "GGGTTC": 4059, + "GGGTTG": 4060, + "GGGTCA": 4061, + "GGGTCT": 4062, + "GGGTCC": 4063, + "GGGTCG": 4064, + "GGGTGA": 4065, + "GGGTGT": 4066, + "GGGTGC": 4067, + "GGGTGG": 4068, + "GGGCAA": 4069, + "GGGCAT": 4070, + "GGGCAC": 4071, + "GGGCAG": 4072, + "GGGCTA": 4073, + "GGGCTT": 4074, + "GGGCTC": 4075, + "GGGCTG": 4076, + "GGGCCA": 4077, + "GGGCCT": 4078, + "GGGCCC": 4079, + "GGGCCG": 4080, + "GGGCGA": 4081, + "GGGCGT": 4082, + "GGGCGC": 4083, + "GGGCGG": 4084, + "GGGGAA": 4085, + "GGGGAT": 4086, + "GGGGAC": 4087, + "GGGGAG": 4088, + "GGGGTA": 4089, + "GGGGTT": 4090, + "GGGGTC": 4091, + "GGGGTG": 4092, + "GGGGCA": 4093, + "GGGGCT": 4094, + "GGGGCC": 4095, + "GGGGCG": 4096, + "GGGGGA": 4097, + "GGGGGT": 4098, + "GGGGGC": 4099, + "GGGGGG": 4100 + } + } +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-1000/tokenizer_config.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-1000/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..604e2b1b29af1dcb3920d2828b6a90276e5e3d02 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-1000/tokenizer_config.json @@ -0,0 +1,58 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "[UNK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "[CLS]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "[SEP]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "4": { + "content": "[MASK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "clean_up_tokenization_spaces": true, + "cls_token": "[CLS]", + "do_basic_tokenize": true, + "do_lower_case": false, + "mask_token": "[MASK]", + "max_len": 512, + "model_max_length": 512, + "never_split": null, + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "strip_accents": null, + "tokenize_chinese_chars": true, + "tokenizer_class": "BertTokenizer", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-1000/trainer_state.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-1000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..1eb1b6e271b93b6cfc5452aa2b3e54ac6ddfdfe3 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-1000/trainer_state.json @@ -0,0 +1,313 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 4.0, + "eval_steps": 500, + "global_step": 1000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.1, + "grad_norm": 0.6038870215415955, + "learning_rate": 5.319148936170213e-05, + "loss": 0.6945, + "step": 25 + }, + { + "epoch": 0.2, + "grad_norm": 0.6668091416358948, + "learning_rate": 0.00010638297872340425, + "loss": 0.6929, + "step": 50 + }, + { + "epoch": 0.3, + "grad_norm": 0.8282511234283447, + "learning_rate": 0.00015957446808510637, + "loss": 0.6925, + "step": 75 + }, + { + "epoch": 0.4, + "grad_norm": 1.3149235248565674, + "learning_rate": 0.0002127659574468085, + "loss": 0.6919, + "step": 100 + }, + { + "epoch": 0.5, + "grad_norm": 1.8841172456741333, + "learning_rate": 0.00026595744680851064, + "loss": 0.6788, + "step": 125 + }, + { + "epoch": 0.6, + "grad_norm": 4.435206413269043, + "learning_rate": 0.00031914893617021275, + "loss": 0.6285, + "step": 150 + }, + { + "epoch": 0.7, + "grad_norm": 8.858148574829102, + "learning_rate": 0.0003723404255319149, + "loss": 0.5386, + "step": 175 + }, + { + "epoch": 0.8, + "grad_norm": 5.627648830413818, + "learning_rate": 0.0003992081821181128, + "loss": 0.5537, + "step": 200 + }, + { + "epoch": 0.9, + "grad_norm": 5.086580753326416, + "learning_rate": 0.0003975585615308479, + "loss": 0.5595, + "step": 225 + }, + { + "epoch": 1.0, + "grad_norm": 5.627719402313232, + "learning_rate": 0.00039590894094358297, + "loss": 0.4824, + "step": 250 + }, + { + "epoch": 1.1, + "grad_norm": 5.888038158416748, + "learning_rate": 0.0003942593203563181, + "loss": 0.4404, + "step": 275 + }, + { + "epoch": 1.2, + "grad_norm": 5.646231651306152, + "learning_rate": 0.0003926096997690532, + "loss": 0.4354, + "step": 300 + }, + { + "epoch": 1.3, + "grad_norm": 5.191990852355957, + "learning_rate": 0.00039096007918178817, + "loss": 0.4469, + "step": 325 + }, + { + "epoch": 1.4, + "grad_norm": 13.90101432800293, + "learning_rate": 0.0003893104585945233, + "loss": 0.3858, + "step": 350 + }, + { + "epoch": 1.5, + "grad_norm": 17.75942039489746, + "learning_rate": 0.0003876608380072583, + "loss": 0.4131, + "step": 375 + }, + { + "epoch": 1.6, + "grad_norm": 7.398252487182617, + "learning_rate": 0.00038601121741999343, + "loss": 0.3842, + "step": 400 + }, + { + "epoch": 1.7, + "grad_norm": 11.010066986083984, + "learning_rate": 0.0003843615968327285, + "loss": 0.3959, + "step": 425 + }, + { + "epoch": 1.8, + "grad_norm": 9.277341842651367, + "learning_rate": 0.0003827119762454636, + "loss": 0.3552, + "step": 450 + }, + { + "epoch": 1.9, + "grad_norm": 26.011449813842773, + "learning_rate": 0.00038106235565819863, + "loss": 0.3961, + "step": 475 + }, + { + "epoch": 2.0, + "grad_norm": 11.123054504394531, + "learning_rate": 0.0003794127350709337, + "loss": 0.3548, + "step": 500 + }, + { + "epoch": 2.1, + "grad_norm": 20.128774642944336, + "learning_rate": 0.0003777631144836688, + "loss": 0.302, + "step": 525 + }, + { + "epoch": 2.2, + "grad_norm": 4.64390230178833, + "learning_rate": 0.00037611349389640383, + "loss": 0.2853, + "step": 550 + }, + { + "epoch": 2.3, + "grad_norm": 4.199934482574463, + "learning_rate": 0.00037446387330913894, + "loss": 0.2724, + "step": 575 + }, + { + "epoch": 2.4, + "grad_norm": 6.909411430358887, + "learning_rate": 0.000372814252721874, + "loss": 0.2976, + "step": 600 + }, + { + "epoch": 2.5, + "grad_norm": 9.234820365905762, + "learning_rate": 0.00037116463213460903, + "loss": 0.2798, + "step": 625 + }, + { + "epoch": 2.6, + "grad_norm": 10.472943305969238, + "learning_rate": 0.00036951501154734414, + "loss": 0.3007, + "step": 650 + }, + { + "epoch": 2.7, + "grad_norm": 19.46732521057129, + "learning_rate": 0.0003678653909600792, + "loss": 0.2701, + "step": 675 + }, + { + "epoch": 2.8, + "grad_norm": 7.501255512237549, + "learning_rate": 0.0003662157703728143, + "loss": 0.2896, + "step": 700 + }, + { + "epoch": 2.9, + "grad_norm": 10.645585060119629, + "learning_rate": 0.00036456614978554934, + "loss": 0.2748, + "step": 725 + }, + { + "epoch": 3.0, + "grad_norm": 30.26616859436035, + "learning_rate": 0.00036291652919828444, + "loss": 0.268, + "step": 750 + }, + { + "epoch": 3.1, + "grad_norm": 14.13516616821289, + "learning_rate": 0.00036126690861101944, + "loss": 0.2159, + "step": 775 + }, + { + "epoch": 3.2, + "grad_norm": 10.091023445129395, + "learning_rate": 0.00035961728802375454, + "loss": 0.2071, + "step": 800 + }, + { + "epoch": 3.3, + "grad_norm": 22.201452255249023, + "learning_rate": 0.0003579676674364896, + "loss": 0.2146, + "step": 825 + }, + { + "epoch": 3.4, + "grad_norm": 10.044412612915039, + "learning_rate": 0.0003563180468492247, + "loss": 0.1827, + "step": 850 + }, + { + "epoch": 3.5, + "grad_norm": 21.04941177368164, + "learning_rate": 0.0003546684262619598, + "loss": 0.202, + "step": 875 + }, + { + "epoch": 3.6, + "grad_norm": 11.754891395568848, + "learning_rate": 0.00035301880567469485, + "loss": 0.2021, + "step": 900 + }, + { + "epoch": 3.7, + "grad_norm": 9.113747596740723, + "learning_rate": 0.0003513691850874299, + "loss": 0.2067, + "step": 925 + }, + { + "epoch": 3.8, + "grad_norm": 9.075196266174316, + "learning_rate": 0.00034971956450016495, + "loss": 0.3285, + "step": 950 + }, + { + "epoch": 3.9, + "grad_norm": 8.568796157836914, + "learning_rate": 0.00034806994391290005, + "loss": 0.2254, + "step": 975 + }, + { + "epoch": 4.0, + "grad_norm": 3.9587976932525635, + "learning_rate": 0.0003464203233256351, + "loss": 0.2245, + "step": 1000 + } + ], + "logging_steps": 25, + "max_steps": 6250, + "num_input_tokens_seen": 0, + "num_train_epochs": 25, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 50819481600000.0, + "train_batch_size": 20, + "trial_name": null, + "trial_params": null +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-1000/training_args.bin b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-1000/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..77692b6666e83a688b38e799c964e4694a57eec9 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-1000/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f73b41315068a430d2439f5707c62152a34210df62454b9549f9642446dd90b2 +size 5304 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-1000/vocab.txt b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-1000/vocab.txt new file mode 100644 index 0000000000000000000000000000000000000000..cc9ad53732e21ea9180f18e1fb67a38cc5702d4c --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-1000/vocab.txt @@ -0,0 +1,4101 @@ +[PAD] +[UNK] +[CLS] +[SEP] +[MASK] +AAAAAA +AAAAAT +AAAAAC +AAAAAG +AAAATA +AAAATT +AAAATC +AAAATG +AAAACA +AAAACT +AAAACC +AAAACG +AAAAGA +AAAAGT +AAAAGC +AAAAGG +AAATAA +AAATAT +AAATAC +AAATAG +AAATTA +AAATTT +AAATTC +AAATTG +AAATCA +AAATCT +AAATCC +AAATCG +AAATGA +AAATGT +AAATGC +AAATGG +AAACAA +AAACAT +AAACAC +AAACAG +AAACTA +AAACTT +AAACTC +AAACTG +AAACCA +AAACCT +AAACCC +AAACCG +AAACGA +AAACGT +AAACGC +AAACGG +AAAGAA +AAAGAT +AAAGAC +AAAGAG +AAAGTA +AAAGTT +AAAGTC +AAAGTG +AAAGCA +AAAGCT +AAAGCC +AAAGCG +AAAGGA +AAAGGT +AAAGGC +AAAGGG +AATAAA +AATAAT +AATAAC +AATAAG +AATATA +AATATT +AATATC +AATATG +AATACA +AATACT +AATACC +AATACG +AATAGA +AATAGT +AATAGC +AATAGG +AATTAA +AATTAT +AATTAC +AATTAG +AATTTA +AATTTT +AATTTC +AATTTG +AATTCA +AATTCT +AATTCC +AATTCG +AATTGA +AATTGT +AATTGC +AATTGG +AATCAA +AATCAT +AATCAC +AATCAG +AATCTA +AATCTT +AATCTC +AATCTG +AATCCA +AATCCT +AATCCC +AATCCG +AATCGA +AATCGT +AATCGC +AATCGG +AATGAA +AATGAT +AATGAC +AATGAG +AATGTA +AATGTT +AATGTC +AATGTG +AATGCA +AATGCT +AATGCC +AATGCG +AATGGA +AATGGT +AATGGC +AATGGG +AACAAA +AACAAT +AACAAC +AACAAG +AACATA +AACATT +AACATC +AACATG +AACACA +AACACT +AACACC +AACACG +AACAGA +AACAGT +AACAGC +AACAGG +AACTAA +AACTAT +AACTAC +AACTAG +AACTTA +AACTTT +AACTTC +AACTTG +AACTCA +AACTCT +AACTCC +AACTCG +AACTGA +AACTGT +AACTGC +AACTGG +AACCAA +AACCAT +AACCAC +AACCAG +AACCTA +AACCTT +AACCTC +AACCTG +AACCCA +AACCCT +AACCCC +AACCCG +AACCGA +AACCGT +AACCGC +AACCGG +AACGAA +AACGAT +AACGAC +AACGAG +AACGTA +AACGTT +AACGTC +AACGTG +AACGCA +AACGCT +AACGCC +AACGCG +AACGGA +AACGGT +AACGGC +AACGGG +AAGAAA +AAGAAT +AAGAAC +AAGAAG +AAGATA +AAGATT +AAGATC +AAGATG +AAGACA +AAGACT +AAGACC +AAGACG +AAGAGA +AAGAGT +AAGAGC +AAGAGG +AAGTAA +AAGTAT +AAGTAC +AAGTAG +AAGTTA +AAGTTT +AAGTTC +AAGTTG +AAGTCA +AAGTCT +AAGTCC +AAGTCG +AAGTGA +AAGTGT +AAGTGC +AAGTGG +AAGCAA +AAGCAT +AAGCAC +AAGCAG +AAGCTA +AAGCTT +AAGCTC +AAGCTG +AAGCCA +AAGCCT +AAGCCC +AAGCCG +AAGCGA +AAGCGT +AAGCGC +AAGCGG +AAGGAA +AAGGAT +AAGGAC +AAGGAG +AAGGTA +AAGGTT +AAGGTC +AAGGTG +AAGGCA +AAGGCT +AAGGCC +AAGGCG +AAGGGA +AAGGGT +AAGGGC +AAGGGG +ATAAAA +ATAAAT +ATAAAC +ATAAAG +ATAATA +ATAATT +ATAATC +ATAATG +ATAACA +ATAACT +ATAACC +ATAACG +ATAAGA +ATAAGT +ATAAGC +ATAAGG +ATATAA +ATATAT +ATATAC +ATATAG +ATATTA +ATATTT +ATATTC +ATATTG +ATATCA +ATATCT +ATATCC +ATATCG +ATATGA +ATATGT +ATATGC +ATATGG +ATACAA +ATACAT +ATACAC +ATACAG +ATACTA +ATACTT +ATACTC +ATACTG +ATACCA +ATACCT +ATACCC +ATACCG +ATACGA +ATACGT +ATACGC +ATACGG +ATAGAA +ATAGAT +ATAGAC +ATAGAG +ATAGTA +ATAGTT +ATAGTC +ATAGTG +ATAGCA +ATAGCT +ATAGCC +ATAGCG +ATAGGA +ATAGGT +ATAGGC +ATAGGG +ATTAAA +ATTAAT +ATTAAC +ATTAAG +ATTATA +ATTATT +ATTATC +ATTATG +ATTACA +ATTACT +ATTACC +ATTACG +ATTAGA +ATTAGT +ATTAGC +ATTAGG +ATTTAA +ATTTAT +ATTTAC +ATTTAG +ATTTTA +ATTTTT +ATTTTC +ATTTTG +ATTTCA +ATTTCT +ATTTCC +ATTTCG +ATTTGA +ATTTGT +ATTTGC +ATTTGG +ATTCAA +ATTCAT +ATTCAC +ATTCAG +ATTCTA +ATTCTT +ATTCTC +ATTCTG +ATTCCA +ATTCCT +ATTCCC +ATTCCG +ATTCGA +ATTCGT +ATTCGC +ATTCGG +ATTGAA +ATTGAT +ATTGAC +ATTGAG +ATTGTA +ATTGTT +ATTGTC +ATTGTG +ATTGCA +ATTGCT +ATTGCC +ATTGCG +ATTGGA +ATTGGT +ATTGGC +ATTGGG +ATCAAA +ATCAAT +ATCAAC +ATCAAG +ATCATA +ATCATT +ATCATC +ATCATG +ATCACA +ATCACT +ATCACC +ATCACG +ATCAGA +ATCAGT +ATCAGC +ATCAGG +ATCTAA +ATCTAT +ATCTAC +ATCTAG +ATCTTA +ATCTTT +ATCTTC +ATCTTG +ATCTCA +ATCTCT +ATCTCC +ATCTCG +ATCTGA +ATCTGT +ATCTGC +ATCTGG +ATCCAA +ATCCAT +ATCCAC +ATCCAG +ATCCTA +ATCCTT +ATCCTC +ATCCTG +ATCCCA +ATCCCT +ATCCCC +ATCCCG +ATCCGA +ATCCGT +ATCCGC +ATCCGG +ATCGAA +ATCGAT +ATCGAC +ATCGAG +ATCGTA +ATCGTT +ATCGTC +ATCGTG +ATCGCA +ATCGCT +ATCGCC +ATCGCG +ATCGGA +ATCGGT +ATCGGC +ATCGGG +ATGAAA +ATGAAT +ATGAAC +ATGAAG +ATGATA +ATGATT +ATGATC +ATGATG +ATGACA +ATGACT +ATGACC +ATGACG +ATGAGA +ATGAGT +ATGAGC +ATGAGG +ATGTAA +ATGTAT +ATGTAC +ATGTAG +ATGTTA +ATGTTT +ATGTTC +ATGTTG +ATGTCA +ATGTCT +ATGTCC +ATGTCG +ATGTGA +ATGTGT +ATGTGC +ATGTGG +ATGCAA +ATGCAT +ATGCAC +ATGCAG +ATGCTA +ATGCTT +ATGCTC +ATGCTG +ATGCCA +ATGCCT +ATGCCC +ATGCCG +ATGCGA +ATGCGT +ATGCGC +ATGCGG +ATGGAA +ATGGAT +ATGGAC +ATGGAG +ATGGTA +ATGGTT +ATGGTC +ATGGTG +ATGGCA +ATGGCT +ATGGCC +ATGGCG +ATGGGA +ATGGGT +ATGGGC +ATGGGG +ACAAAA +ACAAAT +ACAAAC +ACAAAG +ACAATA +ACAATT +ACAATC +ACAATG +ACAACA +ACAACT +ACAACC +ACAACG +ACAAGA +ACAAGT +ACAAGC +ACAAGG +ACATAA +ACATAT +ACATAC +ACATAG +ACATTA +ACATTT +ACATTC +ACATTG +ACATCA +ACATCT +ACATCC +ACATCG +ACATGA +ACATGT +ACATGC +ACATGG +ACACAA +ACACAT +ACACAC +ACACAG +ACACTA +ACACTT +ACACTC +ACACTG +ACACCA +ACACCT +ACACCC +ACACCG +ACACGA +ACACGT +ACACGC +ACACGG +ACAGAA +ACAGAT +ACAGAC +ACAGAG +ACAGTA +ACAGTT +ACAGTC +ACAGTG +ACAGCA +ACAGCT +ACAGCC +ACAGCG +ACAGGA +ACAGGT +ACAGGC +ACAGGG +ACTAAA +ACTAAT +ACTAAC +ACTAAG +ACTATA +ACTATT +ACTATC +ACTATG +ACTACA +ACTACT +ACTACC +ACTACG +ACTAGA +ACTAGT +ACTAGC +ACTAGG +ACTTAA +ACTTAT +ACTTAC +ACTTAG +ACTTTA +ACTTTT +ACTTTC +ACTTTG +ACTTCA +ACTTCT +ACTTCC +ACTTCG +ACTTGA +ACTTGT +ACTTGC +ACTTGG +ACTCAA +ACTCAT +ACTCAC +ACTCAG +ACTCTA +ACTCTT +ACTCTC +ACTCTG +ACTCCA +ACTCCT +ACTCCC +ACTCCG +ACTCGA +ACTCGT +ACTCGC +ACTCGG +ACTGAA +ACTGAT +ACTGAC +ACTGAG +ACTGTA +ACTGTT +ACTGTC +ACTGTG +ACTGCA +ACTGCT +ACTGCC +ACTGCG +ACTGGA +ACTGGT +ACTGGC +ACTGGG +ACCAAA +ACCAAT +ACCAAC +ACCAAG +ACCATA +ACCATT +ACCATC +ACCATG +ACCACA +ACCACT +ACCACC +ACCACG +ACCAGA +ACCAGT +ACCAGC +ACCAGG +ACCTAA +ACCTAT +ACCTAC +ACCTAG +ACCTTA +ACCTTT +ACCTTC +ACCTTG +ACCTCA +ACCTCT +ACCTCC +ACCTCG +ACCTGA +ACCTGT +ACCTGC +ACCTGG +ACCCAA +ACCCAT +ACCCAC +ACCCAG +ACCCTA +ACCCTT +ACCCTC +ACCCTG +ACCCCA +ACCCCT +ACCCCC +ACCCCG +ACCCGA +ACCCGT +ACCCGC +ACCCGG +ACCGAA +ACCGAT +ACCGAC +ACCGAG +ACCGTA +ACCGTT +ACCGTC +ACCGTG +ACCGCA +ACCGCT +ACCGCC +ACCGCG +ACCGGA +ACCGGT +ACCGGC +ACCGGG +ACGAAA +ACGAAT +ACGAAC +ACGAAG +ACGATA +ACGATT +ACGATC +ACGATG +ACGACA +ACGACT +ACGACC +ACGACG +ACGAGA +ACGAGT +ACGAGC +ACGAGG +ACGTAA +ACGTAT +ACGTAC +ACGTAG +ACGTTA +ACGTTT +ACGTTC +ACGTTG +ACGTCA +ACGTCT +ACGTCC +ACGTCG +ACGTGA +ACGTGT +ACGTGC +ACGTGG +ACGCAA +ACGCAT +ACGCAC +ACGCAG +ACGCTA +ACGCTT +ACGCTC +ACGCTG +ACGCCA +ACGCCT +ACGCCC +ACGCCG +ACGCGA +ACGCGT +ACGCGC +ACGCGG +ACGGAA +ACGGAT +ACGGAC +ACGGAG +ACGGTA +ACGGTT +ACGGTC +ACGGTG +ACGGCA +ACGGCT +ACGGCC +ACGGCG +ACGGGA +ACGGGT +ACGGGC +ACGGGG +AGAAAA +AGAAAT +AGAAAC +AGAAAG +AGAATA +AGAATT +AGAATC +AGAATG +AGAACA +AGAACT +AGAACC +AGAACG +AGAAGA +AGAAGT +AGAAGC +AGAAGG +AGATAA +AGATAT +AGATAC +AGATAG +AGATTA +AGATTT +AGATTC +AGATTG +AGATCA +AGATCT +AGATCC +AGATCG +AGATGA +AGATGT +AGATGC +AGATGG +AGACAA +AGACAT +AGACAC +AGACAG +AGACTA +AGACTT +AGACTC +AGACTG +AGACCA +AGACCT +AGACCC +AGACCG +AGACGA +AGACGT +AGACGC +AGACGG +AGAGAA +AGAGAT +AGAGAC +AGAGAG +AGAGTA +AGAGTT +AGAGTC +AGAGTG +AGAGCA +AGAGCT +AGAGCC +AGAGCG +AGAGGA +AGAGGT +AGAGGC +AGAGGG +AGTAAA +AGTAAT +AGTAAC +AGTAAG +AGTATA +AGTATT +AGTATC +AGTATG +AGTACA +AGTACT +AGTACC +AGTACG +AGTAGA +AGTAGT +AGTAGC +AGTAGG +AGTTAA +AGTTAT +AGTTAC +AGTTAG +AGTTTA +AGTTTT +AGTTTC +AGTTTG +AGTTCA +AGTTCT +AGTTCC +AGTTCG +AGTTGA +AGTTGT +AGTTGC +AGTTGG +AGTCAA +AGTCAT +AGTCAC +AGTCAG +AGTCTA +AGTCTT +AGTCTC +AGTCTG +AGTCCA +AGTCCT +AGTCCC +AGTCCG +AGTCGA +AGTCGT +AGTCGC +AGTCGG +AGTGAA +AGTGAT +AGTGAC +AGTGAG +AGTGTA +AGTGTT +AGTGTC +AGTGTG +AGTGCA +AGTGCT +AGTGCC +AGTGCG +AGTGGA +AGTGGT +AGTGGC +AGTGGG +AGCAAA +AGCAAT +AGCAAC +AGCAAG +AGCATA +AGCATT +AGCATC +AGCATG +AGCACA +AGCACT +AGCACC +AGCACG +AGCAGA +AGCAGT +AGCAGC +AGCAGG +AGCTAA +AGCTAT +AGCTAC +AGCTAG +AGCTTA +AGCTTT +AGCTTC +AGCTTG +AGCTCA +AGCTCT +AGCTCC +AGCTCG +AGCTGA +AGCTGT +AGCTGC +AGCTGG +AGCCAA +AGCCAT +AGCCAC +AGCCAG +AGCCTA +AGCCTT +AGCCTC +AGCCTG +AGCCCA +AGCCCT +AGCCCC +AGCCCG +AGCCGA +AGCCGT +AGCCGC +AGCCGG +AGCGAA +AGCGAT +AGCGAC +AGCGAG +AGCGTA +AGCGTT +AGCGTC +AGCGTG +AGCGCA +AGCGCT +AGCGCC +AGCGCG +AGCGGA +AGCGGT +AGCGGC +AGCGGG +AGGAAA +AGGAAT +AGGAAC +AGGAAG +AGGATA +AGGATT +AGGATC +AGGATG +AGGACA +AGGACT +AGGACC +AGGACG +AGGAGA +AGGAGT +AGGAGC +AGGAGG +AGGTAA +AGGTAT +AGGTAC +AGGTAG +AGGTTA +AGGTTT +AGGTTC +AGGTTG +AGGTCA +AGGTCT +AGGTCC +AGGTCG +AGGTGA +AGGTGT +AGGTGC +AGGTGG +AGGCAA +AGGCAT +AGGCAC +AGGCAG +AGGCTA +AGGCTT +AGGCTC +AGGCTG +AGGCCA +AGGCCT +AGGCCC +AGGCCG +AGGCGA +AGGCGT +AGGCGC +AGGCGG +AGGGAA +AGGGAT +AGGGAC +AGGGAG +AGGGTA +AGGGTT +AGGGTC +AGGGTG +AGGGCA +AGGGCT +AGGGCC +AGGGCG +AGGGGA +AGGGGT +AGGGGC +AGGGGG +TAAAAA +TAAAAT +TAAAAC +TAAAAG +TAAATA +TAAATT +TAAATC +TAAATG +TAAACA +TAAACT +TAAACC +TAAACG +TAAAGA +TAAAGT +TAAAGC +TAAAGG +TAATAA +TAATAT +TAATAC +TAATAG +TAATTA +TAATTT +TAATTC +TAATTG +TAATCA +TAATCT +TAATCC +TAATCG +TAATGA +TAATGT +TAATGC +TAATGG +TAACAA +TAACAT +TAACAC +TAACAG +TAACTA +TAACTT +TAACTC +TAACTG +TAACCA +TAACCT +TAACCC +TAACCG +TAACGA +TAACGT +TAACGC +TAACGG +TAAGAA +TAAGAT +TAAGAC +TAAGAG +TAAGTA +TAAGTT +TAAGTC +TAAGTG +TAAGCA +TAAGCT +TAAGCC +TAAGCG +TAAGGA +TAAGGT +TAAGGC +TAAGGG +TATAAA +TATAAT +TATAAC +TATAAG +TATATA +TATATT +TATATC +TATATG +TATACA +TATACT +TATACC +TATACG +TATAGA +TATAGT +TATAGC +TATAGG +TATTAA +TATTAT +TATTAC +TATTAG +TATTTA +TATTTT +TATTTC +TATTTG +TATTCA +TATTCT +TATTCC +TATTCG +TATTGA +TATTGT +TATTGC +TATTGG +TATCAA +TATCAT +TATCAC +TATCAG +TATCTA +TATCTT +TATCTC +TATCTG +TATCCA +TATCCT +TATCCC +TATCCG +TATCGA +TATCGT +TATCGC +TATCGG +TATGAA +TATGAT +TATGAC +TATGAG +TATGTA +TATGTT +TATGTC +TATGTG +TATGCA +TATGCT +TATGCC +TATGCG +TATGGA +TATGGT +TATGGC +TATGGG +TACAAA +TACAAT +TACAAC +TACAAG +TACATA +TACATT +TACATC +TACATG +TACACA +TACACT +TACACC +TACACG +TACAGA +TACAGT +TACAGC +TACAGG +TACTAA +TACTAT +TACTAC +TACTAG +TACTTA +TACTTT +TACTTC +TACTTG +TACTCA +TACTCT +TACTCC +TACTCG +TACTGA +TACTGT +TACTGC +TACTGG +TACCAA +TACCAT +TACCAC +TACCAG +TACCTA +TACCTT +TACCTC +TACCTG +TACCCA +TACCCT +TACCCC +TACCCG +TACCGA +TACCGT +TACCGC +TACCGG +TACGAA +TACGAT +TACGAC +TACGAG +TACGTA +TACGTT +TACGTC +TACGTG +TACGCA +TACGCT +TACGCC +TACGCG +TACGGA +TACGGT +TACGGC +TACGGG +TAGAAA +TAGAAT +TAGAAC +TAGAAG +TAGATA +TAGATT +TAGATC +TAGATG +TAGACA +TAGACT +TAGACC +TAGACG +TAGAGA +TAGAGT +TAGAGC +TAGAGG +TAGTAA +TAGTAT +TAGTAC +TAGTAG +TAGTTA +TAGTTT +TAGTTC +TAGTTG +TAGTCA +TAGTCT +TAGTCC +TAGTCG +TAGTGA +TAGTGT +TAGTGC +TAGTGG +TAGCAA +TAGCAT +TAGCAC +TAGCAG +TAGCTA +TAGCTT +TAGCTC +TAGCTG +TAGCCA +TAGCCT +TAGCCC +TAGCCG +TAGCGA +TAGCGT +TAGCGC +TAGCGG +TAGGAA +TAGGAT +TAGGAC +TAGGAG +TAGGTA +TAGGTT +TAGGTC +TAGGTG +TAGGCA +TAGGCT +TAGGCC +TAGGCG +TAGGGA +TAGGGT +TAGGGC +TAGGGG +TTAAAA +TTAAAT +TTAAAC +TTAAAG +TTAATA +TTAATT +TTAATC +TTAATG +TTAACA +TTAACT +TTAACC +TTAACG +TTAAGA +TTAAGT +TTAAGC +TTAAGG +TTATAA +TTATAT +TTATAC +TTATAG +TTATTA +TTATTT +TTATTC +TTATTG +TTATCA +TTATCT +TTATCC +TTATCG +TTATGA +TTATGT +TTATGC +TTATGG +TTACAA +TTACAT +TTACAC +TTACAG +TTACTA +TTACTT +TTACTC +TTACTG +TTACCA +TTACCT +TTACCC +TTACCG +TTACGA +TTACGT +TTACGC +TTACGG +TTAGAA +TTAGAT +TTAGAC +TTAGAG +TTAGTA +TTAGTT +TTAGTC +TTAGTG +TTAGCA +TTAGCT +TTAGCC +TTAGCG +TTAGGA +TTAGGT +TTAGGC +TTAGGG +TTTAAA +TTTAAT +TTTAAC +TTTAAG +TTTATA +TTTATT +TTTATC +TTTATG +TTTACA +TTTACT +TTTACC +TTTACG +TTTAGA +TTTAGT +TTTAGC +TTTAGG +TTTTAA +TTTTAT +TTTTAC +TTTTAG +TTTTTA +TTTTTT +TTTTTC +TTTTTG +TTTTCA +TTTTCT +TTTTCC +TTTTCG +TTTTGA +TTTTGT +TTTTGC +TTTTGG +TTTCAA +TTTCAT +TTTCAC +TTTCAG +TTTCTA +TTTCTT +TTTCTC +TTTCTG +TTTCCA +TTTCCT +TTTCCC +TTTCCG +TTTCGA +TTTCGT +TTTCGC +TTTCGG +TTTGAA +TTTGAT +TTTGAC +TTTGAG +TTTGTA +TTTGTT +TTTGTC +TTTGTG +TTTGCA +TTTGCT +TTTGCC +TTTGCG +TTTGGA +TTTGGT +TTTGGC +TTTGGG +TTCAAA +TTCAAT +TTCAAC +TTCAAG +TTCATA +TTCATT +TTCATC +TTCATG +TTCACA +TTCACT +TTCACC +TTCACG +TTCAGA +TTCAGT +TTCAGC +TTCAGG +TTCTAA +TTCTAT +TTCTAC +TTCTAG +TTCTTA +TTCTTT +TTCTTC +TTCTTG +TTCTCA +TTCTCT +TTCTCC +TTCTCG +TTCTGA +TTCTGT +TTCTGC +TTCTGG +TTCCAA +TTCCAT +TTCCAC +TTCCAG +TTCCTA +TTCCTT +TTCCTC +TTCCTG +TTCCCA +TTCCCT +TTCCCC +TTCCCG +TTCCGA +TTCCGT +TTCCGC +TTCCGG +TTCGAA +TTCGAT +TTCGAC +TTCGAG +TTCGTA +TTCGTT +TTCGTC +TTCGTG +TTCGCA +TTCGCT +TTCGCC +TTCGCG +TTCGGA +TTCGGT +TTCGGC +TTCGGG +TTGAAA +TTGAAT +TTGAAC +TTGAAG +TTGATA +TTGATT +TTGATC +TTGATG +TTGACA +TTGACT +TTGACC +TTGACG +TTGAGA +TTGAGT +TTGAGC +TTGAGG +TTGTAA +TTGTAT +TTGTAC +TTGTAG +TTGTTA +TTGTTT +TTGTTC +TTGTTG +TTGTCA +TTGTCT +TTGTCC +TTGTCG +TTGTGA +TTGTGT +TTGTGC +TTGTGG +TTGCAA +TTGCAT +TTGCAC +TTGCAG +TTGCTA +TTGCTT +TTGCTC +TTGCTG +TTGCCA +TTGCCT +TTGCCC +TTGCCG +TTGCGA +TTGCGT +TTGCGC +TTGCGG +TTGGAA +TTGGAT +TTGGAC +TTGGAG +TTGGTA +TTGGTT +TTGGTC +TTGGTG +TTGGCA +TTGGCT +TTGGCC +TTGGCG +TTGGGA +TTGGGT +TTGGGC +TTGGGG +TCAAAA +TCAAAT +TCAAAC +TCAAAG +TCAATA +TCAATT +TCAATC +TCAATG +TCAACA +TCAACT +TCAACC +TCAACG +TCAAGA +TCAAGT +TCAAGC +TCAAGG +TCATAA +TCATAT +TCATAC +TCATAG +TCATTA +TCATTT +TCATTC +TCATTG +TCATCA +TCATCT +TCATCC +TCATCG +TCATGA +TCATGT +TCATGC +TCATGG +TCACAA +TCACAT +TCACAC +TCACAG +TCACTA +TCACTT +TCACTC +TCACTG +TCACCA +TCACCT +TCACCC +TCACCG +TCACGA +TCACGT +TCACGC +TCACGG +TCAGAA +TCAGAT +TCAGAC +TCAGAG +TCAGTA +TCAGTT +TCAGTC +TCAGTG +TCAGCA +TCAGCT +TCAGCC +TCAGCG +TCAGGA +TCAGGT +TCAGGC +TCAGGG +TCTAAA +TCTAAT +TCTAAC +TCTAAG +TCTATA +TCTATT +TCTATC +TCTATG +TCTACA +TCTACT +TCTACC +TCTACG +TCTAGA +TCTAGT +TCTAGC +TCTAGG +TCTTAA +TCTTAT +TCTTAC +TCTTAG +TCTTTA +TCTTTT +TCTTTC +TCTTTG +TCTTCA +TCTTCT +TCTTCC +TCTTCG +TCTTGA +TCTTGT +TCTTGC +TCTTGG +TCTCAA +TCTCAT +TCTCAC +TCTCAG +TCTCTA +TCTCTT +TCTCTC +TCTCTG +TCTCCA +TCTCCT +TCTCCC +TCTCCG +TCTCGA +TCTCGT +TCTCGC +TCTCGG +TCTGAA +TCTGAT +TCTGAC +TCTGAG +TCTGTA +TCTGTT +TCTGTC +TCTGTG +TCTGCA +TCTGCT +TCTGCC +TCTGCG +TCTGGA +TCTGGT +TCTGGC +TCTGGG +TCCAAA +TCCAAT +TCCAAC +TCCAAG +TCCATA +TCCATT +TCCATC +TCCATG +TCCACA +TCCACT +TCCACC +TCCACG +TCCAGA +TCCAGT +TCCAGC +TCCAGG +TCCTAA +TCCTAT +TCCTAC +TCCTAG +TCCTTA +TCCTTT +TCCTTC +TCCTTG +TCCTCA +TCCTCT +TCCTCC +TCCTCG +TCCTGA +TCCTGT +TCCTGC +TCCTGG +TCCCAA +TCCCAT +TCCCAC +TCCCAG +TCCCTA +TCCCTT +TCCCTC +TCCCTG +TCCCCA +TCCCCT +TCCCCC +TCCCCG +TCCCGA +TCCCGT +TCCCGC +TCCCGG +TCCGAA +TCCGAT +TCCGAC +TCCGAG +TCCGTA +TCCGTT +TCCGTC +TCCGTG +TCCGCA +TCCGCT +TCCGCC +TCCGCG +TCCGGA +TCCGGT +TCCGGC +TCCGGG +TCGAAA +TCGAAT +TCGAAC +TCGAAG +TCGATA +TCGATT +TCGATC +TCGATG +TCGACA +TCGACT +TCGACC +TCGACG +TCGAGA +TCGAGT +TCGAGC +TCGAGG +TCGTAA +TCGTAT +TCGTAC +TCGTAG +TCGTTA +TCGTTT +TCGTTC +TCGTTG +TCGTCA +TCGTCT +TCGTCC +TCGTCG +TCGTGA +TCGTGT +TCGTGC +TCGTGG +TCGCAA +TCGCAT +TCGCAC +TCGCAG +TCGCTA +TCGCTT +TCGCTC +TCGCTG +TCGCCA +TCGCCT +TCGCCC +TCGCCG +TCGCGA +TCGCGT +TCGCGC +TCGCGG +TCGGAA +TCGGAT +TCGGAC +TCGGAG +TCGGTA +TCGGTT +TCGGTC +TCGGTG +TCGGCA +TCGGCT +TCGGCC +TCGGCG +TCGGGA +TCGGGT +TCGGGC +TCGGGG +TGAAAA +TGAAAT +TGAAAC +TGAAAG +TGAATA +TGAATT +TGAATC +TGAATG +TGAACA +TGAACT +TGAACC +TGAACG +TGAAGA +TGAAGT +TGAAGC +TGAAGG +TGATAA +TGATAT +TGATAC +TGATAG +TGATTA +TGATTT +TGATTC +TGATTG +TGATCA +TGATCT +TGATCC +TGATCG +TGATGA +TGATGT +TGATGC +TGATGG +TGACAA +TGACAT +TGACAC +TGACAG +TGACTA +TGACTT +TGACTC +TGACTG +TGACCA +TGACCT +TGACCC +TGACCG +TGACGA +TGACGT +TGACGC +TGACGG +TGAGAA +TGAGAT +TGAGAC +TGAGAG +TGAGTA +TGAGTT +TGAGTC +TGAGTG +TGAGCA +TGAGCT +TGAGCC +TGAGCG +TGAGGA +TGAGGT +TGAGGC +TGAGGG +TGTAAA +TGTAAT +TGTAAC +TGTAAG +TGTATA +TGTATT +TGTATC +TGTATG +TGTACA +TGTACT +TGTACC +TGTACG +TGTAGA +TGTAGT +TGTAGC +TGTAGG +TGTTAA +TGTTAT +TGTTAC +TGTTAG +TGTTTA +TGTTTT +TGTTTC +TGTTTG +TGTTCA +TGTTCT +TGTTCC +TGTTCG +TGTTGA +TGTTGT +TGTTGC +TGTTGG +TGTCAA +TGTCAT +TGTCAC +TGTCAG +TGTCTA +TGTCTT +TGTCTC +TGTCTG +TGTCCA +TGTCCT +TGTCCC +TGTCCG +TGTCGA +TGTCGT +TGTCGC +TGTCGG +TGTGAA +TGTGAT +TGTGAC +TGTGAG +TGTGTA +TGTGTT +TGTGTC +TGTGTG +TGTGCA +TGTGCT +TGTGCC +TGTGCG +TGTGGA +TGTGGT +TGTGGC +TGTGGG +TGCAAA +TGCAAT +TGCAAC +TGCAAG +TGCATA +TGCATT +TGCATC +TGCATG +TGCACA +TGCACT +TGCACC +TGCACG +TGCAGA +TGCAGT +TGCAGC +TGCAGG +TGCTAA +TGCTAT +TGCTAC +TGCTAG +TGCTTA +TGCTTT +TGCTTC +TGCTTG +TGCTCA +TGCTCT +TGCTCC +TGCTCG +TGCTGA +TGCTGT +TGCTGC +TGCTGG +TGCCAA +TGCCAT +TGCCAC +TGCCAG +TGCCTA +TGCCTT +TGCCTC +TGCCTG +TGCCCA +TGCCCT +TGCCCC +TGCCCG +TGCCGA +TGCCGT +TGCCGC +TGCCGG +TGCGAA +TGCGAT +TGCGAC +TGCGAG +TGCGTA +TGCGTT +TGCGTC +TGCGTG +TGCGCA +TGCGCT +TGCGCC +TGCGCG +TGCGGA +TGCGGT +TGCGGC +TGCGGG +TGGAAA +TGGAAT +TGGAAC +TGGAAG +TGGATA +TGGATT +TGGATC +TGGATG +TGGACA +TGGACT +TGGACC +TGGACG +TGGAGA +TGGAGT +TGGAGC +TGGAGG +TGGTAA +TGGTAT +TGGTAC +TGGTAG +TGGTTA +TGGTTT +TGGTTC +TGGTTG +TGGTCA +TGGTCT +TGGTCC +TGGTCG +TGGTGA +TGGTGT +TGGTGC +TGGTGG +TGGCAA +TGGCAT +TGGCAC +TGGCAG +TGGCTA +TGGCTT +TGGCTC +TGGCTG +TGGCCA +TGGCCT +TGGCCC +TGGCCG +TGGCGA +TGGCGT +TGGCGC +TGGCGG +TGGGAA +TGGGAT +TGGGAC +TGGGAG +TGGGTA +TGGGTT +TGGGTC +TGGGTG +TGGGCA +TGGGCT +TGGGCC +TGGGCG +TGGGGA +TGGGGT +TGGGGC +TGGGGG +CAAAAA +CAAAAT +CAAAAC +CAAAAG +CAAATA +CAAATT +CAAATC +CAAATG +CAAACA +CAAACT +CAAACC +CAAACG +CAAAGA +CAAAGT +CAAAGC +CAAAGG +CAATAA +CAATAT +CAATAC +CAATAG +CAATTA +CAATTT +CAATTC +CAATTG +CAATCA +CAATCT +CAATCC +CAATCG +CAATGA +CAATGT +CAATGC +CAATGG +CAACAA +CAACAT +CAACAC +CAACAG +CAACTA +CAACTT +CAACTC +CAACTG +CAACCA +CAACCT +CAACCC +CAACCG +CAACGA +CAACGT +CAACGC +CAACGG +CAAGAA +CAAGAT +CAAGAC +CAAGAG +CAAGTA +CAAGTT +CAAGTC +CAAGTG +CAAGCA +CAAGCT +CAAGCC +CAAGCG +CAAGGA +CAAGGT +CAAGGC +CAAGGG +CATAAA +CATAAT +CATAAC +CATAAG +CATATA +CATATT +CATATC +CATATG +CATACA +CATACT +CATACC +CATACG +CATAGA +CATAGT +CATAGC +CATAGG +CATTAA +CATTAT +CATTAC +CATTAG +CATTTA +CATTTT +CATTTC +CATTTG +CATTCA +CATTCT +CATTCC +CATTCG +CATTGA +CATTGT +CATTGC +CATTGG +CATCAA +CATCAT +CATCAC +CATCAG +CATCTA +CATCTT +CATCTC +CATCTG +CATCCA +CATCCT +CATCCC +CATCCG +CATCGA +CATCGT +CATCGC +CATCGG +CATGAA +CATGAT +CATGAC +CATGAG +CATGTA +CATGTT +CATGTC +CATGTG +CATGCA +CATGCT +CATGCC +CATGCG +CATGGA +CATGGT +CATGGC +CATGGG +CACAAA +CACAAT +CACAAC +CACAAG +CACATA +CACATT +CACATC +CACATG +CACACA +CACACT +CACACC +CACACG +CACAGA +CACAGT +CACAGC +CACAGG +CACTAA +CACTAT +CACTAC +CACTAG +CACTTA +CACTTT +CACTTC +CACTTG +CACTCA +CACTCT +CACTCC +CACTCG +CACTGA +CACTGT +CACTGC +CACTGG +CACCAA +CACCAT +CACCAC +CACCAG +CACCTA +CACCTT +CACCTC +CACCTG +CACCCA +CACCCT +CACCCC +CACCCG +CACCGA +CACCGT +CACCGC +CACCGG +CACGAA +CACGAT +CACGAC +CACGAG +CACGTA +CACGTT +CACGTC +CACGTG +CACGCA +CACGCT +CACGCC +CACGCG +CACGGA +CACGGT +CACGGC +CACGGG +CAGAAA +CAGAAT +CAGAAC +CAGAAG +CAGATA +CAGATT +CAGATC +CAGATG +CAGACA +CAGACT +CAGACC +CAGACG +CAGAGA +CAGAGT +CAGAGC +CAGAGG +CAGTAA +CAGTAT +CAGTAC +CAGTAG +CAGTTA +CAGTTT +CAGTTC +CAGTTG +CAGTCA +CAGTCT +CAGTCC +CAGTCG +CAGTGA +CAGTGT +CAGTGC +CAGTGG +CAGCAA +CAGCAT +CAGCAC +CAGCAG +CAGCTA +CAGCTT +CAGCTC +CAGCTG +CAGCCA +CAGCCT +CAGCCC +CAGCCG +CAGCGA +CAGCGT +CAGCGC +CAGCGG +CAGGAA +CAGGAT +CAGGAC +CAGGAG +CAGGTA +CAGGTT +CAGGTC +CAGGTG +CAGGCA +CAGGCT +CAGGCC +CAGGCG +CAGGGA +CAGGGT +CAGGGC +CAGGGG +CTAAAA +CTAAAT +CTAAAC +CTAAAG +CTAATA +CTAATT +CTAATC +CTAATG +CTAACA +CTAACT +CTAACC +CTAACG +CTAAGA +CTAAGT +CTAAGC +CTAAGG +CTATAA +CTATAT +CTATAC +CTATAG +CTATTA +CTATTT +CTATTC +CTATTG +CTATCA +CTATCT +CTATCC +CTATCG +CTATGA +CTATGT +CTATGC +CTATGG +CTACAA +CTACAT +CTACAC +CTACAG +CTACTA +CTACTT +CTACTC +CTACTG +CTACCA +CTACCT +CTACCC +CTACCG +CTACGA +CTACGT +CTACGC +CTACGG +CTAGAA +CTAGAT +CTAGAC +CTAGAG +CTAGTA +CTAGTT +CTAGTC +CTAGTG +CTAGCA +CTAGCT +CTAGCC +CTAGCG +CTAGGA +CTAGGT +CTAGGC +CTAGGG +CTTAAA +CTTAAT +CTTAAC +CTTAAG +CTTATA +CTTATT +CTTATC +CTTATG +CTTACA +CTTACT +CTTACC +CTTACG +CTTAGA +CTTAGT +CTTAGC +CTTAGG +CTTTAA +CTTTAT +CTTTAC +CTTTAG +CTTTTA +CTTTTT +CTTTTC +CTTTTG +CTTTCA +CTTTCT +CTTTCC +CTTTCG +CTTTGA +CTTTGT +CTTTGC +CTTTGG +CTTCAA +CTTCAT +CTTCAC +CTTCAG +CTTCTA +CTTCTT +CTTCTC +CTTCTG +CTTCCA +CTTCCT +CTTCCC +CTTCCG +CTTCGA +CTTCGT +CTTCGC +CTTCGG +CTTGAA +CTTGAT +CTTGAC +CTTGAG +CTTGTA +CTTGTT +CTTGTC +CTTGTG +CTTGCA +CTTGCT +CTTGCC +CTTGCG +CTTGGA +CTTGGT +CTTGGC +CTTGGG +CTCAAA +CTCAAT +CTCAAC +CTCAAG +CTCATA +CTCATT +CTCATC +CTCATG +CTCACA +CTCACT +CTCACC +CTCACG +CTCAGA +CTCAGT +CTCAGC +CTCAGG +CTCTAA +CTCTAT +CTCTAC +CTCTAG +CTCTTA +CTCTTT +CTCTTC +CTCTTG +CTCTCA +CTCTCT +CTCTCC +CTCTCG +CTCTGA +CTCTGT +CTCTGC +CTCTGG +CTCCAA +CTCCAT +CTCCAC +CTCCAG +CTCCTA +CTCCTT +CTCCTC +CTCCTG +CTCCCA +CTCCCT +CTCCCC +CTCCCG +CTCCGA +CTCCGT +CTCCGC +CTCCGG +CTCGAA +CTCGAT +CTCGAC +CTCGAG +CTCGTA +CTCGTT +CTCGTC +CTCGTG +CTCGCA +CTCGCT +CTCGCC +CTCGCG +CTCGGA +CTCGGT +CTCGGC +CTCGGG +CTGAAA +CTGAAT +CTGAAC +CTGAAG +CTGATA +CTGATT +CTGATC +CTGATG +CTGACA +CTGACT +CTGACC +CTGACG +CTGAGA +CTGAGT +CTGAGC +CTGAGG +CTGTAA +CTGTAT +CTGTAC +CTGTAG +CTGTTA +CTGTTT +CTGTTC +CTGTTG +CTGTCA +CTGTCT +CTGTCC +CTGTCG +CTGTGA +CTGTGT +CTGTGC +CTGTGG +CTGCAA +CTGCAT +CTGCAC +CTGCAG +CTGCTA +CTGCTT +CTGCTC +CTGCTG +CTGCCA +CTGCCT +CTGCCC +CTGCCG +CTGCGA +CTGCGT +CTGCGC +CTGCGG +CTGGAA +CTGGAT +CTGGAC +CTGGAG +CTGGTA +CTGGTT +CTGGTC +CTGGTG +CTGGCA +CTGGCT +CTGGCC +CTGGCG +CTGGGA +CTGGGT +CTGGGC +CTGGGG +CCAAAA +CCAAAT +CCAAAC +CCAAAG +CCAATA +CCAATT +CCAATC +CCAATG +CCAACA +CCAACT +CCAACC +CCAACG +CCAAGA +CCAAGT +CCAAGC +CCAAGG +CCATAA +CCATAT +CCATAC +CCATAG +CCATTA +CCATTT +CCATTC +CCATTG +CCATCA +CCATCT +CCATCC +CCATCG +CCATGA +CCATGT +CCATGC +CCATGG +CCACAA +CCACAT +CCACAC +CCACAG +CCACTA +CCACTT +CCACTC +CCACTG +CCACCA +CCACCT +CCACCC +CCACCG +CCACGA +CCACGT +CCACGC +CCACGG +CCAGAA +CCAGAT +CCAGAC +CCAGAG +CCAGTA +CCAGTT +CCAGTC +CCAGTG +CCAGCA +CCAGCT +CCAGCC +CCAGCG +CCAGGA +CCAGGT +CCAGGC +CCAGGG +CCTAAA +CCTAAT +CCTAAC +CCTAAG +CCTATA +CCTATT +CCTATC +CCTATG +CCTACA +CCTACT +CCTACC +CCTACG +CCTAGA +CCTAGT +CCTAGC +CCTAGG +CCTTAA +CCTTAT +CCTTAC +CCTTAG +CCTTTA +CCTTTT +CCTTTC +CCTTTG +CCTTCA +CCTTCT +CCTTCC +CCTTCG +CCTTGA +CCTTGT +CCTTGC +CCTTGG +CCTCAA +CCTCAT +CCTCAC +CCTCAG +CCTCTA +CCTCTT +CCTCTC +CCTCTG +CCTCCA +CCTCCT +CCTCCC +CCTCCG +CCTCGA +CCTCGT +CCTCGC +CCTCGG +CCTGAA +CCTGAT +CCTGAC +CCTGAG +CCTGTA +CCTGTT +CCTGTC +CCTGTG +CCTGCA +CCTGCT +CCTGCC +CCTGCG +CCTGGA +CCTGGT +CCTGGC +CCTGGG +CCCAAA +CCCAAT +CCCAAC +CCCAAG +CCCATA +CCCATT +CCCATC +CCCATG +CCCACA +CCCACT +CCCACC +CCCACG +CCCAGA +CCCAGT +CCCAGC +CCCAGG +CCCTAA +CCCTAT +CCCTAC +CCCTAG +CCCTTA +CCCTTT +CCCTTC +CCCTTG +CCCTCA +CCCTCT +CCCTCC +CCCTCG +CCCTGA +CCCTGT +CCCTGC +CCCTGG +CCCCAA +CCCCAT +CCCCAC +CCCCAG +CCCCTA +CCCCTT +CCCCTC +CCCCTG +CCCCCA +CCCCCT +CCCCCC +CCCCCG +CCCCGA +CCCCGT +CCCCGC +CCCCGG +CCCGAA +CCCGAT +CCCGAC +CCCGAG +CCCGTA +CCCGTT +CCCGTC +CCCGTG +CCCGCA +CCCGCT +CCCGCC +CCCGCG +CCCGGA +CCCGGT +CCCGGC +CCCGGG +CCGAAA +CCGAAT +CCGAAC +CCGAAG +CCGATA +CCGATT +CCGATC +CCGATG +CCGACA +CCGACT +CCGACC +CCGACG +CCGAGA +CCGAGT +CCGAGC +CCGAGG +CCGTAA +CCGTAT +CCGTAC +CCGTAG +CCGTTA +CCGTTT +CCGTTC +CCGTTG +CCGTCA +CCGTCT +CCGTCC +CCGTCG +CCGTGA +CCGTGT +CCGTGC +CCGTGG +CCGCAA +CCGCAT +CCGCAC +CCGCAG +CCGCTA +CCGCTT +CCGCTC +CCGCTG +CCGCCA +CCGCCT +CCGCCC +CCGCCG +CCGCGA +CCGCGT +CCGCGC +CCGCGG +CCGGAA +CCGGAT +CCGGAC +CCGGAG +CCGGTA +CCGGTT +CCGGTC +CCGGTG +CCGGCA +CCGGCT +CCGGCC +CCGGCG +CCGGGA +CCGGGT +CCGGGC +CCGGGG +CGAAAA +CGAAAT +CGAAAC +CGAAAG +CGAATA +CGAATT +CGAATC +CGAATG +CGAACA +CGAACT +CGAACC +CGAACG +CGAAGA +CGAAGT +CGAAGC +CGAAGG +CGATAA +CGATAT +CGATAC +CGATAG +CGATTA +CGATTT +CGATTC +CGATTG +CGATCA +CGATCT +CGATCC +CGATCG +CGATGA +CGATGT +CGATGC +CGATGG +CGACAA +CGACAT +CGACAC +CGACAG +CGACTA +CGACTT +CGACTC +CGACTG +CGACCA +CGACCT +CGACCC +CGACCG +CGACGA +CGACGT +CGACGC +CGACGG +CGAGAA +CGAGAT +CGAGAC +CGAGAG +CGAGTA +CGAGTT +CGAGTC +CGAGTG +CGAGCA +CGAGCT +CGAGCC +CGAGCG +CGAGGA +CGAGGT +CGAGGC +CGAGGG +CGTAAA +CGTAAT +CGTAAC +CGTAAG +CGTATA +CGTATT +CGTATC +CGTATG +CGTACA +CGTACT +CGTACC +CGTACG +CGTAGA +CGTAGT +CGTAGC +CGTAGG +CGTTAA +CGTTAT +CGTTAC +CGTTAG +CGTTTA +CGTTTT +CGTTTC +CGTTTG +CGTTCA +CGTTCT +CGTTCC +CGTTCG +CGTTGA +CGTTGT +CGTTGC +CGTTGG +CGTCAA +CGTCAT +CGTCAC +CGTCAG +CGTCTA +CGTCTT +CGTCTC +CGTCTG +CGTCCA +CGTCCT +CGTCCC +CGTCCG +CGTCGA +CGTCGT +CGTCGC +CGTCGG +CGTGAA +CGTGAT +CGTGAC +CGTGAG +CGTGTA +CGTGTT +CGTGTC +CGTGTG +CGTGCA +CGTGCT +CGTGCC +CGTGCG +CGTGGA +CGTGGT +CGTGGC +CGTGGG +CGCAAA +CGCAAT +CGCAAC +CGCAAG +CGCATA +CGCATT +CGCATC +CGCATG +CGCACA +CGCACT +CGCACC +CGCACG +CGCAGA +CGCAGT +CGCAGC +CGCAGG +CGCTAA +CGCTAT +CGCTAC +CGCTAG +CGCTTA +CGCTTT +CGCTTC +CGCTTG +CGCTCA +CGCTCT +CGCTCC +CGCTCG +CGCTGA +CGCTGT +CGCTGC +CGCTGG +CGCCAA +CGCCAT +CGCCAC +CGCCAG +CGCCTA +CGCCTT +CGCCTC +CGCCTG +CGCCCA +CGCCCT +CGCCCC +CGCCCG +CGCCGA +CGCCGT +CGCCGC +CGCCGG +CGCGAA +CGCGAT +CGCGAC +CGCGAG +CGCGTA +CGCGTT +CGCGTC +CGCGTG +CGCGCA +CGCGCT +CGCGCC +CGCGCG +CGCGGA +CGCGGT +CGCGGC +CGCGGG +CGGAAA +CGGAAT +CGGAAC +CGGAAG +CGGATA +CGGATT +CGGATC +CGGATG +CGGACA +CGGACT +CGGACC +CGGACG +CGGAGA +CGGAGT +CGGAGC +CGGAGG +CGGTAA +CGGTAT +CGGTAC +CGGTAG +CGGTTA +CGGTTT +CGGTTC +CGGTTG +CGGTCA +CGGTCT +CGGTCC +CGGTCG +CGGTGA +CGGTGT +CGGTGC +CGGTGG +CGGCAA +CGGCAT +CGGCAC +CGGCAG +CGGCTA +CGGCTT +CGGCTC +CGGCTG +CGGCCA +CGGCCT +CGGCCC +CGGCCG +CGGCGA +CGGCGT +CGGCGC +CGGCGG +CGGGAA +CGGGAT +CGGGAC +CGGGAG +CGGGTA +CGGGTT +CGGGTC +CGGGTG +CGGGCA +CGGGCT +CGGGCC +CGGGCG +CGGGGA +CGGGGT +CGGGGC +CGGGGG +GAAAAA +GAAAAT +GAAAAC +GAAAAG +GAAATA +GAAATT +GAAATC +GAAATG +GAAACA +GAAACT +GAAACC +GAAACG +GAAAGA +GAAAGT +GAAAGC +GAAAGG +GAATAA +GAATAT +GAATAC +GAATAG +GAATTA +GAATTT +GAATTC +GAATTG +GAATCA +GAATCT +GAATCC +GAATCG +GAATGA +GAATGT +GAATGC +GAATGG +GAACAA +GAACAT +GAACAC +GAACAG +GAACTA +GAACTT +GAACTC +GAACTG +GAACCA +GAACCT +GAACCC +GAACCG +GAACGA +GAACGT +GAACGC +GAACGG +GAAGAA +GAAGAT +GAAGAC +GAAGAG +GAAGTA +GAAGTT +GAAGTC +GAAGTG +GAAGCA +GAAGCT +GAAGCC +GAAGCG +GAAGGA +GAAGGT +GAAGGC +GAAGGG +GATAAA +GATAAT +GATAAC +GATAAG +GATATA +GATATT +GATATC +GATATG +GATACA +GATACT +GATACC +GATACG +GATAGA +GATAGT +GATAGC +GATAGG +GATTAA +GATTAT +GATTAC +GATTAG +GATTTA +GATTTT +GATTTC +GATTTG +GATTCA +GATTCT +GATTCC +GATTCG +GATTGA +GATTGT +GATTGC +GATTGG +GATCAA +GATCAT +GATCAC +GATCAG +GATCTA +GATCTT +GATCTC +GATCTG +GATCCA +GATCCT +GATCCC +GATCCG +GATCGA +GATCGT +GATCGC +GATCGG +GATGAA +GATGAT +GATGAC +GATGAG +GATGTA +GATGTT +GATGTC +GATGTG +GATGCA +GATGCT +GATGCC +GATGCG +GATGGA +GATGGT +GATGGC +GATGGG +GACAAA +GACAAT +GACAAC +GACAAG +GACATA +GACATT +GACATC +GACATG +GACACA +GACACT +GACACC +GACACG +GACAGA +GACAGT +GACAGC +GACAGG +GACTAA +GACTAT +GACTAC +GACTAG +GACTTA +GACTTT +GACTTC +GACTTG +GACTCA +GACTCT +GACTCC +GACTCG +GACTGA +GACTGT +GACTGC +GACTGG +GACCAA +GACCAT +GACCAC +GACCAG +GACCTA +GACCTT +GACCTC +GACCTG +GACCCA +GACCCT +GACCCC +GACCCG +GACCGA +GACCGT +GACCGC +GACCGG +GACGAA +GACGAT +GACGAC +GACGAG +GACGTA +GACGTT +GACGTC +GACGTG +GACGCA +GACGCT +GACGCC +GACGCG +GACGGA +GACGGT +GACGGC +GACGGG +GAGAAA +GAGAAT +GAGAAC +GAGAAG +GAGATA +GAGATT +GAGATC +GAGATG +GAGACA +GAGACT +GAGACC +GAGACG +GAGAGA +GAGAGT +GAGAGC +GAGAGG +GAGTAA +GAGTAT +GAGTAC +GAGTAG +GAGTTA +GAGTTT +GAGTTC +GAGTTG +GAGTCA +GAGTCT +GAGTCC +GAGTCG +GAGTGA +GAGTGT +GAGTGC +GAGTGG +GAGCAA +GAGCAT +GAGCAC +GAGCAG +GAGCTA +GAGCTT +GAGCTC +GAGCTG +GAGCCA +GAGCCT +GAGCCC +GAGCCG +GAGCGA +GAGCGT +GAGCGC +GAGCGG +GAGGAA +GAGGAT +GAGGAC +GAGGAG +GAGGTA +GAGGTT +GAGGTC +GAGGTG +GAGGCA +GAGGCT +GAGGCC +GAGGCG +GAGGGA +GAGGGT +GAGGGC +GAGGGG +GTAAAA +GTAAAT +GTAAAC +GTAAAG +GTAATA +GTAATT +GTAATC +GTAATG +GTAACA +GTAACT +GTAACC +GTAACG +GTAAGA +GTAAGT +GTAAGC +GTAAGG +GTATAA +GTATAT +GTATAC +GTATAG +GTATTA +GTATTT +GTATTC +GTATTG +GTATCA +GTATCT +GTATCC +GTATCG +GTATGA +GTATGT +GTATGC +GTATGG +GTACAA +GTACAT +GTACAC +GTACAG +GTACTA +GTACTT +GTACTC +GTACTG +GTACCA +GTACCT +GTACCC +GTACCG +GTACGA +GTACGT +GTACGC +GTACGG +GTAGAA +GTAGAT +GTAGAC +GTAGAG +GTAGTA +GTAGTT +GTAGTC +GTAGTG +GTAGCA +GTAGCT +GTAGCC +GTAGCG +GTAGGA +GTAGGT +GTAGGC +GTAGGG +GTTAAA +GTTAAT +GTTAAC +GTTAAG +GTTATA +GTTATT +GTTATC +GTTATG +GTTACA +GTTACT +GTTACC +GTTACG +GTTAGA +GTTAGT +GTTAGC +GTTAGG +GTTTAA +GTTTAT +GTTTAC +GTTTAG +GTTTTA +GTTTTT +GTTTTC +GTTTTG +GTTTCA +GTTTCT +GTTTCC +GTTTCG +GTTTGA +GTTTGT +GTTTGC +GTTTGG +GTTCAA +GTTCAT +GTTCAC +GTTCAG +GTTCTA +GTTCTT +GTTCTC +GTTCTG +GTTCCA +GTTCCT +GTTCCC +GTTCCG +GTTCGA +GTTCGT +GTTCGC +GTTCGG +GTTGAA +GTTGAT +GTTGAC +GTTGAG +GTTGTA +GTTGTT +GTTGTC +GTTGTG +GTTGCA +GTTGCT +GTTGCC +GTTGCG +GTTGGA +GTTGGT +GTTGGC +GTTGGG +GTCAAA +GTCAAT +GTCAAC +GTCAAG +GTCATA +GTCATT +GTCATC +GTCATG +GTCACA +GTCACT +GTCACC +GTCACG +GTCAGA +GTCAGT +GTCAGC +GTCAGG +GTCTAA +GTCTAT +GTCTAC +GTCTAG +GTCTTA +GTCTTT +GTCTTC +GTCTTG +GTCTCA +GTCTCT +GTCTCC +GTCTCG +GTCTGA +GTCTGT +GTCTGC +GTCTGG +GTCCAA +GTCCAT +GTCCAC +GTCCAG +GTCCTA +GTCCTT +GTCCTC +GTCCTG +GTCCCA +GTCCCT +GTCCCC +GTCCCG +GTCCGA +GTCCGT +GTCCGC +GTCCGG +GTCGAA +GTCGAT +GTCGAC +GTCGAG +GTCGTA +GTCGTT +GTCGTC +GTCGTG +GTCGCA +GTCGCT +GTCGCC +GTCGCG +GTCGGA +GTCGGT +GTCGGC +GTCGGG +GTGAAA +GTGAAT +GTGAAC +GTGAAG +GTGATA +GTGATT +GTGATC +GTGATG +GTGACA +GTGACT +GTGACC +GTGACG +GTGAGA +GTGAGT +GTGAGC +GTGAGG +GTGTAA +GTGTAT +GTGTAC +GTGTAG +GTGTTA +GTGTTT +GTGTTC +GTGTTG +GTGTCA +GTGTCT +GTGTCC +GTGTCG +GTGTGA +GTGTGT +GTGTGC +GTGTGG +GTGCAA +GTGCAT +GTGCAC +GTGCAG +GTGCTA +GTGCTT +GTGCTC +GTGCTG +GTGCCA +GTGCCT +GTGCCC +GTGCCG +GTGCGA +GTGCGT +GTGCGC +GTGCGG +GTGGAA +GTGGAT +GTGGAC +GTGGAG +GTGGTA +GTGGTT +GTGGTC +GTGGTG +GTGGCA +GTGGCT +GTGGCC +GTGGCG +GTGGGA +GTGGGT +GTGGGC +GTGGGG +GCAAAA +GCAAAT +GCAAAC +GCAAAG +GCAATA +GCAATT +GCAATC +GCAATG +GCAACA +GCAACT +GCAACC +GCAACG +GCAAGA +GCAAGT +GCAAGC +GCAAGG +GCATAA +GCATAT +GCATAC +GCATAG +GCATTA +GCATTT +GCATTC +GCATTG +GCATCA +GCATCT +GCATCC +GCATCG +GCATGA +GCATGT +GCATGC +GCATGG +GCACAA +GCACAT +GCACAC +GCACAG +GCACTA +GCACTT +GCACTC +GCACTG +GCACCA +GCACCT +GCACCC +GCACCG +GCACGA +GCACGT +GCACGC +GCACGG +GCAGAA +GCAGAT +GCAGAC +GCAGAG +GCAGTA +GCAGTT +GCAGTC +GCAGTG +GCAGCA +GCAGCT +GCAGCC +GCAGCG +GCAGGA +GCAGGT +GCAGGC +GCAGGG +GCTAAA +GCTAAT +GCTAAC +GCTAAG +GCTATA +GCTATT +GCTATC +GCTATG +GCTACA +GCTACT +GCTACC +GCTACG +GCTAGA +GCTAGT +GCTAGC +GCTAGG +GCTTAA +GCTTAT +GCTTAC +GCTTAG +GCTTTA +GCTTTT +GCTTTC +GCTTTG +GCTTCA +GCTTCT +GCTTCC +GCTTCG +GCTTGA +GCTTGT +GCTTGC +GCTTGG +GCTCAA +GCTCAT +GCTCAC +GCTCAG +GCTCTA +GCTCTT +GCTCTC +GCTCTG +GCTCCA +GCTCCT +GCTCCC +GCTCCG +GCTCGA +GCTCGT +GCTCGC +GCTCGG +GCTGAA +GCTGAT +GCTGAC +GCTGAG +GCTGTA +GCTGTT +GCTGTC +GCTGTG +GCTGCA +GCTGCT +GCTGCC +GCTGCG +GCTGGA +GCTGGT +GCTGGC +GCTGGG +GCCAAA +GCCAAT +GCCAAC +GCCAAG +GCCATA +GCCATT +GCCATC +GCCATG +GCCACA +GCCACT +GCCACC +GCCACG +GCCAGA +GCCAGT +GCCAGC +GCCAGG +GCCTAA +GCCTAT +GCCTAC +GCCTAG +GCCTTA +GCCTTT +GCCTTC +GCCTTG +GCCTCA +GCCTCT +GCCTCC +GCCTCG +GCCTGA +GCCTGT +GCCTGC +GCCTGG +GCCCAA +GCCCAT +GCCCAC +GCCCAG +GCCCTA +GCCCTT +GCCCTC +GCCCTG +GCCCCA +GCCCCT +GCCCCC +GCCCCG +GCCCGA +GCCCGT +GCCCGC +GCCCGG +GCCGAA +GCCGAT +GCCGAC +GCCGAG +GCCGTA +GCCGTT +GCCGTC +GCCGTG +GCCGCA +GCCGCT +GCCGCC +GCCGCG +GCCGGA +GCCGGT +GCCGGC +GCCGGG +GCGAAA +GCGAAT +GCGAAC +GCGAAG +GCGATA +GCGATT +GCGATC +GCGATG +GCGACA +GCGACT +GCGACC +GCGACG +GCGAGA +GCGAGT +GCGAGC +GCGAGG +GCGTAA +GCGTAT +GCGTAC +GCGTAG +GCGTTA +GCGTTT +GCGTTC +GCGTTG +GCGTCA +GCGTCT +GCGTCC +GCGTCG +GCGTGA +GCGTGT +GCGTGC +GCGTGG +GCGCAA +GCGCAT +GCGCAC +GCGCAG +GCGCTA +GCGCTT +GCGCTC +GCGCTG +GCGCCA +GCGCCT +GCGCCC +GCGCCG +GCGCGA +GCGCGT +GCGCGC +GCGCGG +GCGGAA +GCGGAT +GCGGAC +GCGGAG +GCGGTA +GCGGTT +GCGGTC +GCGGTG +GCGGCA +GCGGCT +GCGGCC +GCGGCG +GCGGGA +GCGGGT +GCGGGC +GCGGGG +GGAAAA +GGAAAT +GGAAAC +GGAAAG +GGAATA +GGAATT +GGAATC +GGAATG +GGAACA +GGAACT +GGAACC +GGAACG +GGAAGA +GGAAGT +GGAAGC +GGAAGG +GGATAA +GGATAT +GGATAC +GGATAG +GGATTA +GGATTT +GGATTC +GGATTG +GGATCA +GGATCT +GGATCC +GGATCG +GGATGA +GGATGT +GGATGC +GGATGG +GGACAA +GGACAT +GGACAC +GGACAG +GGACTA +GGACTT +GGACTC +GGACTG +GGACCA +GGACCT +GGACCC +GGACCG +GGACGA +GGACGT +GGACGC +GGACGG +GGAGAA +GGAGAT +GGAGAC +GGAGAG +GGAGTA +GGAGTT +GGAGTC +GGAGTG +GGAGCA +GGAGCT +GGAGCC +GGAGCG +GGAGGA +GGAGGT +GGAGGC +GGAGGG +GGTAAA +GGTAAT +GGTAAC +GGTAAG +GGTATA +GGTATT +GGTATC +GGTATG +GGTACA +GGTACT +GGTACC +GGTACG +GGTAGA +GGTAGT +GGTAGC +GGTAGG +GGTTAA +GGTTAT +GGTTAC +GGTTAG +GGTTTA +GGTTTT +GGTTTC +GGTTTG +GGTTCA +GGTTCT +GGTTCC +GGTTCG +GGTTGA +GGTTGT +GGTTGC +GGTTGG +GGTCAA +GGTCAT +GGTCAC +GGTCAG +GGTCTA +GGTCTT +GGTCTC +GGTCTG +GGTCCA +GGTCCT +GGTCCC +GGTCCG +GGTCGA +GGTCGT +GGTCGC +GGTCGG +GGTGAA +GGTGAT +GGTGAC +GGTGAG +GGTGTA +GGTGTT +GGTGTC +GGTGTG +GGTGCA +GGTGCT +GGTGCC +GGTGCG +GGTGGA +GGTGGT +GGTGGC +GGTGGG +GGCAAA +GGCAAT +GGCAAC +GGCAAG +GGCATA +GGCATT +GGCATC +GGCATG +GGCACA +GGCACT +GGCACC +GGCACG +GGCAGA +GGCAGT +GGCAGC +GGCAGG +GGCTAA +GGCTAT +GGCTAC +GGCTAG +GGCTTA +GGCTTT +GGCTTC +GGCTTG +GGCTCA +GGCTCT +GGCTCC +GGCTCG +GGCTGA +GGCTGT +GGCTGC +GGCTGG +GGCCAA +GGCCAT +GGCCAC +GGCCAG +GGCCTA +GGCCTT +GGCCTC +GGCCTG +GGCCCA +GGCCCT +GGCCCC +GGCCCG +GGCCGA +GGCCGT +GGCCGC +GGCCGG +GGCGAA +GGCGAT +GGCGAC +GGCGAG +GGCGTA +GGCGTT +GGCGTC +GGCGTG +GGCGCA +GGCGCT +GGCGCC +GGCGCG +GGCGGA +GGCGGT +GGCGGC +GGCGGG +GGGAAA +GGGAAT +GGGAAC +GGGAAG +GGGATA +GGGATT +GGGATC +GGGATG +GGGACA +GGGACT +GGGACC +GGGACG +GGGAGA +GGGAGT +GGGAGC +GGGAGG +GGGTAA +GGGTAT +GGGTAC +GGGTAG +GGGTTA +GGGTTT +GGGTTC +GGGTTG +GGGTCA +GGGTCT +GGGTCC +GGGTCG +GGGTGA +GGGTGT +GGGTGC +GGGTGG +GGGCAA +GGGCAT +GGGCAC +GGGCAG +GGGCTA +GGGCTT +GGGCTC +GGGCTG +GGGCCA +GGGCCT +GGGCCC +GGGCCG +GGGCGA +GGGCGT +GGGCGC +GGGCGG +GGGGAA +GGGGAT +GGGGAC +GGGGAG +GGGGTA +GGGGTT +GGGGTC +GGGGTG +GGGGCA +GGGGCT +GGGGCC +GGGGCG +GGGGGA +GGGGGT +GGGGGC +GGGGGG diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-2000/config.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..9b14e61d327bf586fcff6e817cf9d365bae18ad2 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-2000/config.json @@ -0,0 +1,26 @@ +{ + "_name_or_path": "fabihamakhdoomi/TinyDNABERT", + "architectures": [ + "BertForSequenceClassification" + ], + "attention_probs_dropout_prob": 0.1, + "classifier_dropout": null, + "hidden_act": "gelu", + "hidden_dropout_prob": 0.1, + "hidden_size": 128, + "initializer_range": 0.02, + "intermediate_size": 512, + "layer_norm_eps": 1e-12, + "max_position_embeddings": 512, + "model_type": "bert", + "num_attention_heads": 2, + "num_hidden_layers": 2, + "pad_token_id": 0, + "position_embedding_type": "absolute", + "problem_type": "single_label_classification", + "torch_dtype": "float32", + "transformers_version": "4.44.0", + "type_vocab_size": 2, + "use_cache": true, + "vocab_size": 4101 +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-2000/model.safetensors b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-2000/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..018d1a0d784982931133922b93c26160aa542100 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-2000/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:70c00ba45510b61387ecdb345180bc76f9dcac3f3e268e0c09690ebba6ef9b10 +size 4021688 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-2000/optimizer.pt b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-2000/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..e6fe814ef15c2b54b116ead849a043d03d3c3e2a --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-2000/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2dc36824c0fdb046202ebd190ba36b779ab1eef3d9d51e1f65ec27a8606cc731 +size 8058700 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-2000/rng_state.pth b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-2000/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..dc62b940907551fbd6412f1b948afb4ecb3c061f --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-2000/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b1753d15a4935b9bbb100d316f8edf3879525b0e6b5b0c8edeffb1787132e3cd +size 14244 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-2000/scheduler.pt b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-2000/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..2287c4e5e02bffea6969638a53c3d5402c6dc2b0 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-2000/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:da119a7607c7d2d96a92ebe57fbaf5fb22ebfce3399b1702b2209f460c0bfe35 +size 1064 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-2000/special_tokens_map.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-2000/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..a8b3208c2884c4efb86e49300fdd3dc877220cdf --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-2000/special_tokens_map.json @@ -0,0 +1,7 @@ +{ + "cls_token": "[CLS]", + "mask_token": "[MASK]", + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-2000/tokenizer.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-2000/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..ebd782de6963b9032ce9374cfe30f5275be1a30b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-2000/tokenizer.json @@ -0,0 +1,4265 @@ +{ + "version": "1.0", + "truncation": { + "direction": "Right", + "max_length": 512, + "strategy": "LongestFirst", + "stride": 0 + }, + "padding": { + "strategy": { + "Fixed": 512 + }, + "direction": "Right", + "pad_to_multiple_of": null, + "pad_id": 0, + "pad_type_id": 0, + "pad_token": "[PAD]" + }, + "added_tokens": [ + { + "id": 0, + "content": "[PAD]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 1, + "content": "[UNK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 2, + "content": "[CLS]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 3, + "content": "[SEP]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 4, + "content": "[MASK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + } + ], + "normalizer": { + "type": "BertNormalizer", + "clean_text": true, + "handle_chinese_chars": true, + "strip_accents": null, + "lowercase": false + }, + "pre_tokenizer": { + "type": "BertPreTokenizer" + }, + "post_processor": { + "type": "TemplateProcessing", + "single": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + } + ], + "pair": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "B", + "type_id": 1 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 1 + } + } + ], + "special_tokens": { + "[CLS]": { + "id": "[CLS]", + "ids": [ + 2 + ], + "tokens": [ + "[CLS]" + ] + }, + "[SEP]": { + "id": "[SEP]", + "ids": [ + 3 + ], + "tokens": [ + "[SEP]" + ] + } + } + }, + "decoder": { + "type": "WordPiece", + "prefix": "##", + "cleanup": true + }, + "model": { + "type": "WordPiece", + "unk_token": "[UNK]", + "continuing_subword_prefix": "##", + "max_input_chars_per_word": 100, + "vocab": { + "[PAD]": 0, + "[UNK]": 1, + "[CLS]": 2, + "[SEP]": 3, + "[MASK]": 4, + "AAAAAA": 5, + "AAAAAT": 6, + "AAAAAC": 7, + "AAAAAG": 8, + "AAAATA": 9, + "AAAATT": 10, + "AAAATC": 11, + "AAAATG": 12, + "AAAACA": 13, + "AAAACT": 14, + "AAAACC": 15, + "AAAACG": 16, + "AAAAGA": 17, + "AAAAGT": 18, + "AAAAGC": 19, + "AAAAGG": 20, + "AAATAA": 21, + "AAATAT": 22, + "AAATAC": 23, + "AAATAG": 24, + "AAATTA": 25, + "AAATTT": 26, + "AAATTC": 27, + "AAATTG": 28, + "AAATCA": 29, + "AAATCT": 30, + "AAATCC": 31, + "AAATCG": 32, + "AAATGA": 33, + "AAATGT": 34, + "AAATGC": 35, + "AAATGG": 36, + "AAACAA": 37, + "AAACAT": 38, + "AAACAC": 39, + "AAACAG": 40, + "AAACTA": 41, + "AAACTT": 42, + "AAACTC": 43, + "AAACTG": 44, + "AAACCA": 45, + "AAACCT": 46, + "AAACCC": 47, + "AAACCG": 48, + "AAACGA": 49, + "AAACGT": 50, + "AAACGC": 51, + "AAACGG": 52, + "AAAGAA": 53, + "AAAGAT": 54, + "AAAGAC": 55, + "AAAGAG": 56, + "AAAGTA": 57, + "AAAGTT": 58, + "AAAGTC": 59, + "AAAGTG": 60, + "AAAGCA": 61, + "AAAGCT": 62, + "AAAGCC": 63, + "AAAGCG": 64, + "AAAGGA": 65, + "AAAGGT": 66, + "AAAGGC": 67, + "AAAGGG": 68, + "AATAAA": 69, + "AATAAT": 70, + "AATAAC": 71, + "AATAAG": 72, + "AATATA": 73, + "AATATT": 74, + "AATATC": 75, + "AATATG": 76, + "AATACA": 77, + "AATACT": 78, + "AATACC": 79, + "AATACG": 80, + "AATAGA": 81, + "AATAGT": 82, + "AATAGC": 83, + "AATAGG": 84, + "AATTAA": 85, + "AATTAT": 86, + "AATTAC": 87, + "AATTAG": 88, + "AATTTA": 89, + "AATTTT": 90, + "AATTTC": 91, + "AATTTG": 92, + "AATTCA": 93, + "AATTCT": 94, + "AATTCC": 95, + "AATTCG": 96, + "AATTGA": 97, + "AATTGT": 98, + "AATTGC": 99, + "AATTGG": 100, + "AATCAA": 101, + "AATCAT": 102, + "AATCAC": 103, + "AATCAG": 104, + "AATCTA": 105, + "AATCTT": 106, + "AATCTC": 107, + "AATCTG": 108, + "AATCCA": 109, + "AATCCT": 110, + "AATCCC": 111, + "AATCCG": 112, + "AATCGA": 113, + "AATCGT": 114, + "AATCGC": 115, + "AATCGG": 116, + "AATGAA": 117, + "AATGAT": 118, + "AATGAC": 119, + "AATGAG": 120, + "AATGTA": 121, + "AATGTT": 122, + "AATGTC": 123, + "AATGTG": 124, + "AATGCA": 125, + "AATGCT": 126, + "AATGCC": 127, + "AATGCG": 128, + "AATGGA": 129, + "AATGGT": 130, + "AATGGC": 131, + "AATGGG": 132, + "AACAAA": 133, + "AACAAT": 134, + "AACAAC": 135, + "AACAAG": 136, + "AACATA": 137, + "AACATT": 138, + "AACATC": 139, + "AACATG": 140, + "AACACA": 141, + "AACACT": 142, + "AACACC": 143, + "AACACG": 144, + "AACAGA": 145, + "AACAGT": 146, + "AACAGC": 147, + "AACAGG": 148, + "AACTAA": 149, + "AACTAT": 150, + "AACTAC": 151, + "AACTAG": 152, + "AACTTA": 153, + "AACTTT": 154, + "AACTTC": 155, + "AACTTG": 156, + "AACTCA": 157, + "AACTCT": 158, + "AACTCC": 159, + "AACTCG": 160, + "AACTGA": 161, + "AACTGT": 162, + "AACTGC": 163, + "AACTGG": 164, + "AACCAA": 165, + "AACCAT": 166, + "AACCAC": 167, + "AACCAG": 168, + "AACCTA": 169, + "AACCTT": 170, + "AACCTC": 171, + "AACCTG": 172, + "AACCCA": 173, + "AACCCT": 174, + "AACCCC": 175, + "AACCCG": 176, + "AACCGA": 177, + "AACCGT": 178, + "AACCGC": 179, + "AACCGG": 180, + "AACGAA": 181, + "AACGAT": 182, + "AACGAC": 183, + "AACGAG": 184, + "AACGTA": 185, + "AACGTT": 186, + "AACGTC": 187, + "AACGTG": 188, + "AACGCA": 189, + "AACGCT": 190, + "AACGCC": 191, + "AACGCG": 192, + "AACGGA": 193, + "AACGGT": 194, + "AACGGC": 195, + "AACGGG": 196, + "AAGAAA": 197, + "AAGAAT": 198, + "AAGAAC": 199, + "AAGAAG": 200, + "AAGATA": 201, + "AAGATT": 202, + "AAGATC": 203, + "AAGATG": 204, + "AAGACA": 205, + "AAGACT": 206, + "AAGACC": 207, + "AAGACG": 208, + "AAGAGA": 209, + "AAGAGT": 210, + "AAGAGC": 211, + "AAGAGG": 212, + "AAGTAA": 213, + "AAGTAT": 214, + "AAGTAC": 215, + "AAGTAG": 216, + "AAGTTA": 217, + "AAGTTT": 218, + "AAGTTC": 219, + "AAGTTG": 220, + "AAGTCA": 221, + "AAGTCT": 222, + "AAGTCC": 223, + "AAGTCG": 224, + "AAGTGA": 225, + "AAGTGT": 226, + "AAGTGC": 227, + "AAGTGG": 228, + "AAGCAA": 229, + "AAGCAT": 230, + "AAGCAC": 231, + "AAGCAG": 232, + "AAGCTA": 233, + "AAGCTT": 234, + "AAGCTC": 235, + "AAGCTG": 236, + "AAGCCA": 237, + "AAGCCT": 238, + "AAGCCC": 239, + "AAGCCG": 240, + "AAGCGA": 241, + "AAGCGT": 242, + "AAGCGC": 243, + "AAGCGG": 244, + "AAGGAA": 245, + "AAGGAT": 246, + "AAGGAC": 247, + "AAGGAG": 248, + "AAGGTA": 249, + "AAGGTT": 250, + "AAGGTC": 251, + "AAGGTG": 252, + "AAGGCA": 253, + "AAGGCT": 254, + "AAGGCC": 255, + "AAGGCG": 256, + "AAGGGA": 257, + "AAGGGT": 258, + "AAGGGC": 259, + "AAGGGG": 260, + "ATAAAA": 261, + "ATAAAT": 262, + "ATAAAC": 263, + "ATAAAG": 264, + "ATAATA": 265, + "ATAATT": 266, + "ATAATC": 267, + "ATAATG": 268, + "ATAACA": 269, + "ATAACT": 270, + "ATAACC": 271, + "ATAACG": 272, + "ATAAGA": 273, + "ATAAGT": 274, + "ATAAGC": 275, + "ATAAGG": 276, + "ATATAA": 277, + "ATATAT": 278, + "ATATAC": 279, + "ATATAG": 280, + "ATATTA": 281, + "ATATTT": 282, + "ATATTC": 283, + "ATATTG": 284, + "ATATCA": 285, + "ATATCT": 286, + "ATATCC": 287, + "ATATCG": 288, + "ATATGA": 289, + "ATATGT": 290, + "ATATGC": 291, + "ATATGG": 292, + "ATACAA": 293, + "ATACAT": 294, + "ATACAC": 295, + "ATACAG": 296, + "ATACTA": 297, + "ATACTT": 298, + "ATACTC": 299, + "ATACTG": 300, + "ATACCA": 301, + "ATACCT": 302, + "ATACCC": 303, + "ATACCG": 304, + "ATACGA": 305, + "ATACGT": 306, + "ATACGC": 307, + "ATACGG": 308, + "ATAGAA": 309, + "ATAGAT": 310, + "ATAGAC": 311, + "ATAGAG": 312, + "ATAGTA": 313, + "ATAGTT": 314, + "ATAGTC": 315, + "ATAGTG": 316, + "ATAGCA": 317, + "ATAGCT": 318, + "ATAGCC": 319, + "ATAGCG": 320, + "ATAGGA": 321, + "ATAGGT": 322, + "ATAGGC": 323, + "ATAGGG": 324, + "ATTAAA": 325, + "ATTAAT": 326, + "ATTAAC": 327, + "ATTAAG": 328, + "ATTATA": 329, + "ATTATT": 330, + "ATTATC": 331, + "ATTATG": 332, + "ATTACA": 333, + "ATTACT": 334, + "ATTACC": 335, + "ATTACG": 336, + "ATTAGA": 337, + "ATTAGT": 338, + "ATTAGC": 339, + "ATTAGG": 340, + "ATTTAA": 341, + "ATTTAT": 342, + "ATTTAC": 343, + "ATTTAG": 344, + "ATTTTA": 345, + "ATTTTT": 346, + "ATTTTC": 347, + "ATTTTG": 348, + "ATTTCA": 349, + "ATTTCT": 350, + "ATTTCC": 351, + "ATTTCG": 352, + "ATTTGA": 353, + "ATTTGT": 354, + "ATTTGC": 355, + "ATTTGG": 356, + "ATTCAA": 357, + "ATTCAT": 358, + "ATTCAC": 359, + "ATTCAG": 360, + "ATTCTA": 361, + "ATTCTT": 362, + "ATTCTC": 363, + "ATTCTG": 364, + "ATTCCA": 365, + "ATTCCT": 366, + "ATTCCC": 367, + "ATTCCG": 368, + "ATTCGA": 369, + "ATTCGT": 370, + "ATTCGC": 371, + "ATTCGG": 372, + "ATTGAA": 373, + "ATTGAT": 374, + "ATTGAC": 375, + "ATTGAG": 376, + "ATTGTA": 377, + "ATTGTT": 378, + "ATTGTC": 379, + "ATTGTG": 380, + "ATTGCA": 381, + "ATTGCT": 382, + "ATTGCC": 383, + "ATTGCG": 384, + "ATTGGA": 385, + "ATTGGT": 386, + "ATTGGC": 387, + "ATTGGG": 388, + "ATCAAA": 389, + "ATCAAT": 390, + "ATCAAC": 391, + "ATCAAG": 392, + "ATCATA": 393, + "ATCATT": 394, + "ATCATC": 395, + "ATCATG": 396, + "ATCACA": 397, + "ATCACT": 398, + "ATCACC": 399, + "ATCACG": 400, + "ATCAGA": 401, + "ATCAGT": 402, + "ATCAGC": 403, + "ATCAGG": 404, + "ATCTAA": 405, + "ATCTAT": 406, + "ATCTAC": 407, + "ATCTAG": 408, + "ATCTTA": 409, + "ATCTTT": 410, + "ATCTTC": 411, + "ATCTTG": 412, + "ATCTCA": 413, + "ATCTCT": 414, + "ATCTCC": 415, + "ATCTCG": 416, + "ATCTGA": 417, + "ATCTGT": 418, + "ATCTGC": 419, + "ATCTGG": 420, + "ATCCAA": 421, + "ATCCAT": 422, + "ATCCAC": 423, + "ATCCAG": 424, + "ATCCTA": 425, + "ATCCTT": 426, + "ATCCTC": 427, + "ATCCTG": 428, + "ATCCCA": 429, + "ATCCCT": 430, + "ATCCCC": 431, + "ATCCCG": 432, + "ATCCGA": 433, + "ATCCGT": 434, + "ATCCGC": 435, + "ATCCGG": 436, + "ATCGAA": 437, + "ATCGAT": 438, + "ATCGAC": 439, + "ATCGAG": 440, + "ATCGTA": 441, + "ATCGTT": 442, + "ATCGTC": 443, + "ATCGTG": 444, + "ATCGCA": 445, + "ATCGCT": 446, + "ATCGCC": 447, + "ATCGCG": 448, + "ATCGGA": 449, + "ATCGGT": 450, + "ATCGGC": 451, + "ATCGGG": 452, + "ATGAAA": 453, + "ATGAAT": 454, + "ATGAAC": 455, + "ATGAAG": 456, + "ATGATA": 457, + "ATGATT": 458, + "ATGATC": 459, + "ATGATG": 460, + "ATGACA": 461, + "ATGACT": 462, + "ATGACC": 463, + "ATGACG": 464, + "ATGAGA": 465, + "ATGAGT": 466, + "ATGAGC": 467, + "ATGAGG": 468, + "ATGTAA": 469, + "ATGTAT": 470, + "ATGTAC": 471, + "ATGTAG": 472, + "ATGTTA": 473, + "ATGTTT": 474, + "ATGTTC": 475, + "ATGTTG": 476, + "ATGTCA": 477, + "ATGTCT": 478, + "ATGTCC": 479, + "ATGTCG": 480, + "ATGTGA": 481, + "ATGTGT": 482, + "ATGTGC": 483, + "ATGTGG": 484, + "ATGCAA": 485, + "ATGCAT": 486, + "ATGCAC": 487, + "ATGCAG": 488, + "ATGCTA": 489, + "ATGCTT": 490, + "ATGCTC": 491, + "ATGCTG": 492, + "ATGCCA": 493, + "ATGCCT": 494, + "ATGCCC": 495, + "ATGCCG": 496, + "ATGCGA": 497, + "ATGCGT": 498, + "ATGCGC": 499, + "ATGCGG": 500, + "ATGGAA": 501, + "ATGGAT": 502, + "ATGGAC": 503, + "ATGGAG": 504, + "ATGGTA": 505, + "ATGGTT": 506, + "ATGGTC": 507, + "ATGGTG": 508, + "ATGGCA": 509, + "ATGGCT": 510, + "ATGGCC": 511, + "ATGGCG": 512, + "ATGGGA": 513, + "ATGGGT": 514, + "ATGGGC": 515, + "ATGGGG": 516, + "ACAAAA": 517, + "ACAAAT": 518, + "ACAAAC": 519, + "ACAAAG": 520, + "ACAATA": 521, + "ACAATT": 522, + "ACAATC": 523, + "ACAATG": 524, + "ACAACA": 525, + "ACAACT": 526, + "ACAACC": 527, + "ACAACG": 528, + "ACAAGA": 529, + "ACAAGT": 530, + "ACAAGC": 531, + "ACAAGG": 532, + "ACATAA": 533, + "ACATAT": 534, + "ACATAC": 535, + "ACATAG": 536, + "ACATTA": 537, + "ACATTT": 538, + "ACATTC": 539, + "ACATTG": 540, + "ACATCA": 541, + "ACATCT": 542, + "ACATCC": 543, + "ACATCG": 544, + "ACATGA": 545, + "ACATGT": 546, + "ACATGC": 547, + "ACATGG": 548, + "ACACAA": 549, + "ACACAT": 550, + "ACACAC": 551, + "ACACAG": 552, + "ACACTA": 553, + "ACACTT": 554, + "ACACTC": 555, + "ACACTG": 556, + "ACACCA": 557, + "ACACCT": 558, + "ACACCC": 559, + "ACACCG": 560, + "ACACGA": 561, + "ACACGT": 562, + "ACACGC": 563, + "ACACGG": 564, + "ACAGAA": 565, + "ACAGAT": 566, + "ACAGAC": 567, + "ACAGAG": 568, + "ACAGTA": 569, + "ACAGTT": 570, + "ACAGTC": 571, + "ACAGTG": 572, + "ACAGCA": 573, + "ACAGCT": 574, + "ACAGCC": 575, + "ACAGCG": 576, + "ACAGGA": 577, + "ACAGGT": 578, + "ACAGGC": 579, + "ACAGGG": 580, + "ACTAAA": 581, + "ACTAAT": 582, + "ACTAAC": 583, + "ACTAAG": 584, + "ACTATA": 585, + "ACTATT": 586, + "ACTATC": 587, + "ACTATG": 588, + "ACTACA": 589, + "ACTACT": 590, + "ACTACC": 591, + "ACTACG": 592, + "ACTAGA": 593, + "ACTAGT": 594, + "ACTAGC": 595, + "ACTAGG": 596, + "ACTTAA": 597, + "ACTTAT": 598, + "ACTTAC": 599, + "ACTTAG": 600, + "ACTTTA": 601, + "ACTTTT": 602, + "ACTTTC": 603, + "ACTTTG": 604, + "ACTTCA": 605, + "ACTTCT": 606, + "ACTTCC": 607, + "ACTTCG": 608, + "ACTTGA": 609, + "ACTTGT": 610, + "ACTTGC": 611, + "ACTTGG": 612, + "ACTCAA": 613, + "ACTCAT": 614, + "ACTCAC": 615, + "ACTCAG": 616, + "ACTCTA": 617, + "ACTCTT": 618, + "ACTCTC": 619, + "ACTCTG": 620, + "ACTCCA": 621, + "ACTCCT": 622, + "ACTCCC": 623, + "ACTCCG": 624, + "ACTCGA": 625, + "ACTCGT": 626, + "ACTCGC": 627, + "ACTCGG": 628, + "ACTGAA": 629, + "ACTGAT": 630, + "ACTGAC": 631, + "ACTGAG": 632, + "ACTGTA": 633, + "ACTGTT": 634, + "ACTGTC": 635, + "ACTGTG": 636, + "ACTGCA": 637, + "ACTGCT": 638, + "ACTGCC": 639, + "ACTGCG": 640, + "ACTGGA": 641, + "ACTGGT": 642, + "ACTGGC": 643, + "ACTGGG": 644, + "ACCAAA": 645, + "ACCAAT": 646, + "ACCAAC": 647, + "ACCAAG": 648, + "ACCATA": 649, + "ACCATT": 650, + "ACCATC": 651, + "ACCATG": 652, + "ACCACA": 653, + "ACCACT": 654, + "ACCACC": 655, + "ACCACG": 656, + "ACCAGA": 657, + "ACCAGT": 658, + "ACCAGC": 659, + "ACCAGG": 660, + "ACCTAA": 661, + "ACCTAT": 662, + "ACCTAC": 663, + "ACCTAG": 664, + "ACCTTA": 665, + "ACCTTT": 666, + "ACCTTC": 667, + "ACCTTG": 668, + "ACCTCA": 669, + "ACCTCT": 670, + "ACCTCC": 671, + "ACCTCG": 672, + "ACCTGA": 673, + "ACCTGT": 674, + "ACCTGC": 675, + "ACCTGG": 676, + "ACCCAA": 677, + "ACCCAT": 678, + "ACCCAC": 679, + "ACCCAG": 680, + "ACCCTA": 681, + "ACCCTT": 682, + "ACCCTC": 683, + "ACCCTG": 684, + "ACCCCA": 685, + "ACCCCT": 686, + "ACCCCC": 687, + "ACCCCG": 688, + "ACCCGA": 689, + "ACCCGT": 690, + "ACCCGC": 691, + "ACCCGG": 692, + "ACCGAA": 693, + "ACCGAT": 694, + "ACCGAC": 695, + "ACCGAG": 696, + "ACCGTA": 697, + "ACCGTT": 698, + "ACCGTC": 699, + "ACCGTG": 700, + "ACCGCA": 701, + "ACCGCT": 702, + "ACCGCC": 703, + "ACCGCG": 704, + "ACCGGA": 705, + "ACCGGT": 706, + "ACCGGC": 707, + "ACCGGG": 708, + "ACGAAA": 709, + "ACGAAT": 710, + "ACGAAC": 711, + "ACGAAG": 712, + "ACGATA": 713, + "ACGATT": 714, + "ACGATC": 715, + "ACGATG": 716, + "ACGACA": 717, + "ACGACT": 718, + "ACGACC": 719, + "ACGACG": 720, + "ACGAGA": 721, + "ACGAGT": 722, + "ACGAGC": 723, + "ACGAGG": 724, + "ACGTAA": 725, + "ACGTAT": 726, + "ACGTAC": 727, + "ACGTAG": 728, + "ACGTTA": 729, + "ACGTTT": 730, + "ACGTTC": 731, + "ACGTTG": 732, + "ACGTCA": 733, + "ACGTCT": 734, + "ACGTCC": 735, + "ACGTCG": 736, + "ACGTGA": 737, + "ACGTGT": 738, + "ACGTGC": 739, + "ACGTGG": 740, + "ACGCAA": 741, + "ACGCAT": 742, + "ACGCAC": 743, + "ACGCAG": 744, + "ACGCTA": 745, + "ACGCTT": 746, + "ACGCTC": 747, + "ACGCTG": 748, + "ACGCCA": 749, + "ACGCCT": 750, + "ACGCCC": 751, + "ACGCCG": 752, + "ACGCGA": 753, + "ACGCGT": 754, + "ACGCGC": 755, + "ACGCGG": 756, + "ACGGAA": 757, + "ACGGAT": 758, + "ACGGAC": 759, + "ACGGAG": 760, + "ACGGTA": 761, + "ACGGTT": 762, + "ACGGTC": 763, + "ACGGTG": 764, + "ACGGCA": 765, + "ACGGCT": 766, + "ACGGCC": 767, + "ACGGCG": 768, + "ACGGGA": 769, + "ACGGGT": 770, + "ACGGGC": 771, + "ACGGGG": 772, + "AGAAAA": 773, + "AGAAAT": 774, + "AGAAAC": 775, + "AGAAAG": 776, + "AGAATA": 777, + "AGAATT": 778, + "AGAATC": 779, + "AGAATG": 780, + "AGAACA": 781, + "AGAACT": 782, + "AGAACC": 783, + "AGAACG": 784, + "AGAAGA": 785, + "AGAAGT": 786, + "AGAAGC": 787, + "AGAAGG": 788, + "AGATAA": 789, + "AGATAT": 790, + "AGATAC": 791, + "AGATAG": 792, + "AGATTA": 793, + "AGATTT": 794, + "AGATTC": 795, + "AGATTG": 796, + "AGATCA": 797, + "AGATCT": 798, + "AGATCC": 799, + "AGATCG": 800, + "AGATGA": 801, + "AGATGT": 802, + "AGATGC": 803, + "AGATGG": 804, + "AGACAA": 805, + "AGACAT": 806, + "AGACAC": 807, + "AGACAG": 808, + "AGACTA": 809, + "AGACTT": 810, + "AGACTC": 811, + "AGACTG": 812, + "AGACCA": 813, + "AGACCT": 814, + "AGACCC": 815, + "AGACCG": 816, + "AGACGA": 817, + "AGACGT": 818, + "AGACGC": 819, + "AGACGG": 820, + "AGAGAA": 821, + "AGAGAT": 822, + "AGAGAC": 823, + "AGAGAG": 824, + "AGAGTA": 825, + "AGAGTT": 826, + "AGAGTC": 827, + "AGAGTG": 828, + "AGAGCA": 829, + "AGAGCT": 830, + "AGAGCC": 831, + "AGAGCG": 832, + "AGAGGA": 833, + "AGAGGT": 834, + "AGAGGC": 835, + "AGAGGG": 836, + "AGTAAA": 837, + "AGTAAT": 838, + "AGTAAC": 839, + "AGTAAG": 840, + "AGTATA": 841, + "AGTATT": 842, + "AGTATC": 843, + "AGTATG": 844, + "AGTACA": 845, + "AGTACT": 846, + "AGTACC": 847, + "AGTACG": 848, + "AGTAGA": 849, + "AGTAGT": 850, + "AGTAGC": 851, + "AGTAGG": 852, + "AGTTAA": 853, + "AGTTAT": 854, + "AGTTAC": 855, + "AGTTAG": 856, + "AGTTTA": 857, + "AGTTTT": 858, + "AGTTTC": 859, + "AGTTTG": 860, + "AGTTCA": 861, + "AGTTCT": 862, + "AGTTCC": 863, + "AGTTCG": 864, + "AGTTGA": 865, + "AGTTGT": 866, + "AGTTGC": 867, + "AGTTGG": 868, + "AGTCAA": 869, + "AGTCAT": 870, + "AGTCAC": 871, + "AGTCAG": 872, + "AGTCTA": 873, + "AGTCTT": 874, + "AGTCTC": 875, + "AGTCTG": 876, + "AGTCCA": 877, + "AGTCCT": 878, + "AGTCCC": 879, + "AGTCCG": 880, + "AGTCGA": 881, + "AGTCGT": 882, + "AGTCGC": 883, + "AGTCGG": 884, + "AGTGAA": 885, + "AGTGAT": 886, + "AGTGAC": 887, + "AGTGAG": 888, + "AGTGTA": 889, + "AGTGTT": 890, + "AGTGTC": 891, + "AGTGTG": 892, + "AGTGCA": 893, + "AGTGCT": 894, + "AGTGCC": 895, + "AGTGCG": 896, + "AGTGGA": 897, + "AGTGGT": 898, + "AGTGGC": 899, + "AGTGGG": 900, + "AGCAAA": 901, + "AGCAAT": 902, + "AGCAAC": 903, + "AGCAAG": 904, + "AGCATA": 905, + "AGCATT": 906, + "AGCATC": 907, + "AGCATG": 908, + "AGCACA": 909, + "AGCACT": 910, + "AGCACC": 911, + "AGCACG": 912, + "AGCAGA": 913, + "AGCAGT": 914, + "AGCAGC": 915, + "AGCAGG": 916, + "AGCTAA": 917, + "AGCTAT": 918, + "AGCTAC": 919, + "AGCTAG": 920, + "AGCTTA": 921, + "AGCTTT": 922, + "AGCTTC": 923, + "AGCTTG": 924, + "AGCTCA": 925, + "AGCTCT": 926, + "AGCTCC": 927, + "AGCTCG": 928, + "AGCTGA": 929, + "AGCTGT": 930, + "AGCTGC": 931, + "AGCTGG": 932, + "AGCCAA": 933, + "AGCCAT": 934, + "AGCCAC": 935, + "AGCCAG": 936, + "AGCCTA": 937, + "AGCCTT": 938, + "AGCCTC": 939, + "AGCCTG": 940, + "AGCCCA": 941, + "AGCCCT": 942, + "AGCCCC": 943, + "AGCCCG": 944, + "AGCCGA": 945, + "AGCCGT": 946, + "AGCCGC": 947, + "AGCCGG": 948, + "AGCGAA": 949, + "AGCGAT": 950, + "AGCGAC": 951, + "AGCGAG": 952, + "AGCGTA": 953, + "AGCGTT": 954, + "AGCGTC": 955, + "AGCGTG": 956, + "AGCGCA": 957, + "AGCGCT": 958, + "AGCGCC": 959, + "AGCGCG": 960, + "AGCGGA": 961, + "AGCGGT": 962, + "AGCGGC": 963, + "AGCGGG": 964, + "AGGAAA": 965, + "AGGAAT": 966, + "AGGAAC": 967, + "AGGAAG": 968, + "AGGATA": 969, + "AGGATT": 970, + "AGGATC": 971, + "AGGATG": 972, + "AGGACA": 973, + "AGGACT": 974, + "AGGACC": 975, + "AGGACG": 976, + "AGGAGA": 977, + "AGGAGT": 978, + "AGGAGC": 979, + "AGGAGG": 980, + "AGGTAA": 981, + "AGGTAT": 982, + "AGGTAC": 983, + "AGGTAG": 984, + "AGGTTA": 985, + "AGGTTT": 986, + "AGGTTC": 987, + "AGGTTG": 988, + "AGGTCA": 989, + "AGGTCT": 990, + "AGGTCC": 991, + "AGGTCG": 992, + "AGGTGA": 993, + "AGGTGT": 994, + "AGGTGC": 995, + "AGGTGG": 996, + "AGGCAA": 997, + "AGGCAT": 998, + "AGGCAC": 999, + "AGGCAG": 1000, + "AGGCTA": 1001, + "AGGCTT": 1002, + "AGGCTC": 1003, + "AGGCTG": 1004, + "AGGCCA": 1005, + "AGGCCT": 1006, + "AGGCCC": 1007, + "AGGCCG": 1008, + "AGGCGA": 1009, + "AGGCGT": 1010, + "AGGCGC": 1011, + "AGGCGG": 1012, + "AGGGAA": 1013, + "AGGGAT": 1014, + "AGGGAC": 1015, + "AGGGAG": 1016, + "AGGGTA": 1017, + "AGGGTT": 1018, + "AGGGTC": 1019, + "AGGGTG": 1020, + "AGGGCA": 1021, + "AGGGCT": 1022, + "AGGGCC": 1023, + "AGGGCG": 1024, + "AGGGGA": 1025, + "AGGGGT": 1026, + "AGGGGC": 1027, + "AGGGGG": 1028, + "TAAAAA": 1029, + "TAAAAT": 1030, + "TAAAAC": 1031, + "TAAAAG": 1032, + "TAAATA": 1033, + "TAAATT": 1034, + "TAAATC": 1035, + "TAAATG": 1036, + "TAAACA": 1037, + "TAAACT": 1038, + "TAAACC": 1039, + "TAAACG": 1040, + "TAAAGA": 1041, + "TAAAGT": 1042, + "TAAAGC": 1043, + "TAAAGG": 1044, + "TAATAA": 1045, + "TAATAT": 1046, + "TAATAC": 1047, + "TAATAG": 1048, + "TAATTA": 1049, + "TAATTT": 1050, + "TAATTC": 1051, + "TAATTG": 1052, + "TAATCA": 1053, + "TAATCT": 1054, + "TAATCC": 1055, + "TAATCG": 1056, + "TAATGA": 1057, + "TAATGT": 1058, + "TAATGC": 1059, + "TAATGG": 1060, + "TAACAA": 1061, + "TAACAT": 1062, + "TAACAC": 1063, + "TAACAG": 1064, + "TAACTA": 1065, + "TAACTT": 1066, + "TAACTC": 1067, + "TAACTG": 1068, + "TAACCA": 1069, + "TAACCT": 1070, + "TAACCC": 1071, + "TAACCG": 1072, + "TAACGA": 1073, + "TAACGT": 1074, + "TAACGC": 1075, + "TAACGG": 1076, + "TAAGAA": 1077, + "TAAGAT": 1078, + "TAAGAC": 1079, + "TAAGAG": 1080, + "TAAGTA": 1081, + "TAAGTT": 1082, + "TAAGTC": 1083, + "TAAGTG": 1084, + "TAAGCA": 1085, + "TAAGCT": 1086, + "TAAGCC": 1087, + "TAAGCG": 1088, + "TAAGGA": 1089, + "TAAGGT": 1090, + "TAAGGC": 1091, + "TAAGGG": 1092, + "TATAAA": 1093, + "TATAAT": 1094, + "TATAAC": 1095, + "TATAAG": 1096, + "TATATA": 1097, + "TATATT": 1098, + "TATATC": 1099, + "TATATG": 1100, + "TATACA": 1101, + "TATACT": 1102, + "TATACC": 1103, + "TATACG": 1104, + "TATAGA": 1105, + "TATAGT": 1106, + "TATAGC": 1107, + "TATAGG": 1108, + "TATTAA": 1109, + "TATTAT": 1110, + "TATTAC": 1111, + "TATTAG": 1112, + "TATTTA": 1113, + "TATTTT": 1114, + "TATTTC": 1115, + "TATTTG": 1116, + "TATTCA": 1117, + "TATTCT": 1118, + "TATTCC": 1119, + "TATTCG": 1120, + "TATTGA": 1121, + "TATTGT": 1122, + "TATTGC": 1123, + "TATTGG": 1124, + "TATCAA": 1125, + "TATCAT": 1126, + "TATCAC": 1127, + "TATCAG": 1128, + "TATCTA": 1129, + "TATCTT": 1130, + "TATCTC": 1131, + "TATCTG": 1132, + "TATCCA": 1133, + "TATCCT": 1134, + "TATCCC": 1135, + "TATCCG": 1136, + "TATCGA": 1137, + "TATCGT": 1138, + "TATCGC": 1139, + "TATCGG": 1140, + "TATGAA": 1141, + "TATGAT": 1142, + "TATGAC": 1143, + "TATGAG": 1144, + "TATGTA": 1145, + "TATGTT": 1146, + "TATGTC": 1147, + "TATGTG": 1148, + "TATGCA": 1149, + "TATGCT": 1150, + "TATGCC": 1151, + "TATGCG": 1152, + "TATGGA": 1153, + "TATGGT": 1154, + "TATGGC": 1155, + "TATGGG": 1156, + "TACAAA": 1157, + "TACAAT": 1158, + "TACAAC": 1159, + "TACAAG": 1160, + "TACATA": 1161, + "TACATT": 1162, + "TACATC": 1163, + "TACATG": 1164, + "TACACA": 1165, + "TACACT": 1166, + "TACACC": 1167, + "TACACG": 1168, + "TACAGA": 1169, + "TACAGT": 1170, + "TACAGC": 1171, + "TACAGG": 1172, + "TACTAA": 1173, + "TACTAT": 1174, + "TACTAC": 1175, + "TACTAG": 1176, + "TACTTA": 1177, + "TACTTT": 1178, + "TACTTC": 1179, + "TACTTG": 1180, + "TACTCA": 1181, + "TACTCT": 1182, + "TACTCC": 1183, + "TACTCG": 1184, + "TACTGA": 1185, + "TACTGT": 1186, + "TACTGC": 1187, + "TACTGG": 1188, + "TACCAA": 1189, + "TACCAT": 1190, + "TACCAC": 1191, + "TACCAG": 1192, + "TACCTA": 1193, + "TACCTT": 1194, + "TACCTC": 1195, + "TACCTG": 1196, + "TACCCA": 1197, + "TACCCT": 1198, + "TACCCC": 1199, + "TACCCG": 1200, + "TACCGA": 1201, + "TACCGT": 1202, + "TACCGC": 1203, + "TACCGG": 1204, + "TACGAA": 1205, + "TACGAT": 1206, + "TACGAC": 1207, + "TACGAG": 1208, + "TACGTA": 1209, + "TACGTT": 1210, + "TACGTC": 1211, + "TACGTG": 1212, + "TACGCA": 1213, + "TACGCT": 1214, + "TACGCC": 1215, + "TACGCG": 1216, + "TACGGA": 1217, + "TACGGT": 1218, + "TACGGC": 1219, + "TACGGG": 1220, + "TAGAAA": 1221, + "TAGAAT": 1222, + "TAGAAC": 1223, + "TAGAAG": 1224, + "TAGATA": 1225, + "TAGATT": 1226, + "TAGATC": 1227, + "TAGATG": 1228, + "TAGACA": 1229, + "TAGACT": 1230, + "TAGACC": 1231, + "TAGACG": 1232, + "TAGAGA": 1233, + "TAGAGT": 1234, + "TAGAGC": 1235, + "TAGAGG": 1236, + "TAGTAA": 1237, + "TAGTAT": 1238, + "TAGTAC": 1239, + "TAGTAG": 1240, + "TAGTTA": 1241, + "TAGTTT": 1242, + "TAGTTC": 1243, + "TAGTTG": 1244, + "TAGTCA": 1245, + "TAGTCT": 1246, + "TAGTCC": 1247, + "TAGTCG": 1248, + "TAGTGA": 1249, + "TAGTGT": 1250, + "TAGTGC": 1251, + "TAGTGG": 1252, + "TAGCAA": 1253, + "TAGCAT": 1254, + "TAGCAC": 1255, + "TAGCAG": 1256, + "TAGCTA": 1257, + "TAGCTT": 1258, + "TAGCTC": 1259, + "TAGCTG": 1260, + "TAGCCA": 1261, + "TAGCCT": 1262, + "TAGCCC": 1263, + "TAGCCG": 1264, + "TAGCGA": 1265, + "TAGCGT": 1266, + "TAGCGC": 1267, + "TAGCGG": 1268, + "TAGGAA": 1269, + "TAGGAT": 1270, + "TAGGAC": 1271, + "TAGGAG": 1272, + "TAGGTA": 1273, + "TAGGTT": 1274, + "TAGGTC": 1275, + "TAGGTG": 1276, + "TAGGCA": 1277, + "TAGGCT": 1278, + "TAGGCC": 1279, + "TAGGCG": 1280, + "TAGGGA": 1281, + "TAGGGT": 1282, + "TAGGGC": 1283, + "TAGGGG": 1284, + "TTAAAA": 1285, + "TTAAAT": 1286, + "TTAAAC": 1287, + "TTAAAG": 1288, + "TTAATA": 1289, + "TTAATT": 1290, + "TTAATC": 1291, + "TTAATG": 1292, + "TTAACA": 1293, + "TTAACT": 1294, + "TTAACC": 1295, + "TTAACG": 1296, + "TTAAGA": 1297, + "TTAAGT": 1298, + "TTAAGC": 1299, + "TTAAGG": 1300, + "TTATAA": 1301, + "TTATAT": 1302, + "TTATAC": 1303, + "TTATAG": 1304, + "TTATTA": 1305, + "TTATTT": 1306, + "TTATTC": 1307, + "TTATTG": 1308, + "TTATCA": 1309, + "TTATCT": 1310, + "TTATCC": 1311, + "TTATCG": 1312, + "TTATGA": 1313, + "TTATGT": 1314, + "TTATGC": 1315, + "TTATGG": 1316, + "TTACAA": 1317, + "TTACAT": 1318, + "TTACAC": 1319, + "TTACAG": 1320, + "TTACTA": 1321, + "TTACTT": 1322, + "TTACTC": 1323, + "TTACTG": 1324, + "TTACCA": 1325, + "TTACCT": 1326, + "TTACCC": 1327, + "TTACCG": 1328, + "TTACGA": 1329, + "TTACGT": 1330, + "TTACGC": 1331, + "TTACGG": 1332, + "TTAGAA": 1333, + "TTAGAT": 1334, + "TTAGAC": 1335, + "TTAGAG": 1336, + "TTAGTA": 1337, + "TTAGTT": 1338, + "TTAGTC": 1339, + "TTAGTG": 1340, + "TTAGCA": 1341, + "TTAGCT": 1342, + "TTAGCC": 1343, + "TTAGCG": 1344, + "TTAGGA": 1345, + "TTAGGT": 1346, + "TTAGGC": 1347, + "TTAGGG": 1348, + "TTTAAA": 1349, + "TTTAAT": 1350, + "TTTAAC": 1351, + "TTTAAG": 1352, + "TTTATA": 1353, + "TTTATT": 1354, + "TTTATC": 1355, + "TTTATG": 1356, + "TTTACA": 1357, + "TTTACT": 1358, + "TTTACC": 1359, + "TTTACG": 1360, + "TTTAGA": 1361, + "TTTAGT": 1362, + "TTTAGC": 1363, + "TTTAGG": 1364, + "TTTTAA": 1365, + "TTTTAT": 1366, + "TTTTAC": 1367, + "TTTTAG": 1368, + "TTTTTA": 1369, + "TTTTTT": 1370, + "TTTTTC": 1371, + "TTTTTG": 1372, + "TTTTCA": 1373, + "TTTTCT": 1374, + "TTTTCC": 1375, + "TTTTCG": 1376, + "TTTTGA": 1377, + "TTTTGT": 1378, + "TTTTGC": 1379, + "TTTTGG": 1380, + "TTTCAA": 1381, + "TTTCAT": 1382, + "TTTCAC": 1383, + "TTTCAG": 1384, + "TTTCTA": 1385, + "TTTCTT": 1386, + "TTTCTC": 1387, + "TTTCTG": 1388, + "TTTCCA": 1389, + "TTTCCT": 1390, + "TTTCCC": 1391, + "TTTCCG": 1392, + "TTTCGA": 1393, + "TTTCGT": 1394, + "TTTCGC": 1395, + "TTTCGG": 1396, + "TTTGAA": 1397, + "TTTGAT": 1398, + "TTTGAC": 1399, + "TTTGAG": 1400, + "TTTGTA": 1401, + "TTTGTT": 1402, + "TTTGTC": 1403, + "TTTGTG": 1404, + "TTTGCA": 1405, + "TTTGCT": 1406, + "TTTGCC": 1407, + "TTTGCG": 1408, + "TTTGGA": 1409, + "TTTGGT": 1410, + "TTTGGC": 1411, + "TTTGGG": 1412, + "TTCAAA": 1413, + "TTCAAT": 1414, + "TTCAAC": 1415, + "TTCAAG": 1416, + "TTCATA": 1417, + "TTCATT": 1418, + "TTCATC": 1419, + "TTCATG": 1420, + "TTCACA": 1421, + "TTCACT": 1422, + "TTCACC": 1423, + "TTCACG": 1424, + "TTCAGA": 1425, + "TTCAGT": 1426, + "TTCAGC": 1427, + "TTCAGG": 1428, + "TTCTAA": 1429, + "TTCTAT": 1430, + "TTCTAC": 1431, + "TTCTAG": 1432, + "TTCTTA": 1433, + "TTCTTT": 1434, + "TTCTTC": 1435, + "TTCTTG": 1436, + "TTCTCA": 1437, + "TTCTCT": 1438, + "TTCTCC": 1439, + "TTCTCG": 1440, + "TTCTGA": 1441, + "TTCTGT": 1442, + "TTCTGC": 1443, + "TTCTGG": 1444, + "TTCCAA": 1445, + "TTCCAT": 1446, + "TTCCAC": 1447, + "TTCCAG": 1448, + "TTCCTA": 1449, + "TTCCTT": 1450, + "TTCCTC": 1451, + "TTCCTG": 1452, + "TTCCCA": 1453, + "TTCCCT": 1454, + "TTCCCC": 1455, + "TTCCCG": 1456, + "TTCCGA": 1457, + "TTCCGT": 1458, + "TTCCGC": 1459, + "TTCCGG": 1460, + "TTCGAA": 1461, + "TTCGAT": 1462, + "TTCGAC": 1463, + "TTCGAG": 1464, + "TTCGTA": 1465, + "TTCGTT": 1466, + "TTCGTC": 1467, + "TTCGTG": 1468, + "TTCGCA": 1469, + "TTCGCT": 1470, + "TTCGCC": 1471, + "TTCGCG": 1472, + "TTCGGA": 1473, + "TTCGGT": 1474, + "TTCGGC": 1475, + "TTCGGG": 1476, + "TTGAAA": 1477, + "TTGAAT": 1478, + "TTGAAC": 1479, + "TTGAAG": 1480, + "TTGATA": 1481, + "TTGATT": 1482, + "TTGATC": 1483, + "TTGATG": 1484, + "TTGACA": 1485, + "TTGACT": 1486, + "TTGACC": 1487, + "TTGACG": 1488, + "TTGAGA": 1489, + "TTGAGT": 1490, + "TTGAGC": 1491, + "TTGAGG": 1492, + "TTGTAA": 1493, + "TTGTAT": 1494, + "TTGTAC": 1495, + "TTGTAG": 1496, + "TTGTTA": 1497, + "TTGTTT": 1498, + "TTGTTC": 1499, + "TTGTTG": 1500, + "TTGTCA": 1501, + "TTGTCT": 1502, + "TTGTCC": 1503, + "TTGTCG": 1504, + "TTGTGA": 1505, + "TTGTGT": 1506, + "TTGTGC": 1507, + "TTGTGG": 1508, + "TTGCAA": 1509, + "TTGCAT": 1510, + "TTGCAC": 1511, + "TTGCAG": 1512, + "TTGCTA": 1513, + "TTGCTT": 1514, + "TTGCTC": 1515, + "TTGCTG": 1516, + "TTGCCA": 1517, + "TTGCCT": 1518, + "TTGCCC": 1519, + "TTGCCG": 1520, + "TTGCGA": 1521, + "TTGCGT": 1522, + "TTGCGC": 1523, + "TTGCGG": 1524, + "TTGGAA": 1525, + "TTGGAT": 1526, + "TTGGAC": 1527, + "TTGGAG": 1528, + "TTGGTA": 1529, + "TTGGTT": 1530, + "TTGGTC": 1531, + "TTGGTG": 1532, + "TTGGCA": 1533, + "TTGGCT": 1534, + "TTGGCC": 1535, + "TTGGCG": 1536, + "TTGGGA": 1537, + "TTGGGT": 1538, + "TTGGGC": 1539, + "TTGGGG": 1540, + "TCAAAA": 1541, + "TCAAAT": 1542, + "TCAAAC": 1543, + "TCAAAG": 1544, + "TCAATA": 1545, + "TCAATT": 1546, + "TCAATC": 1547, + "TCAATG": 1548, + "TCAACA": 1549, + "TCAACT": 1550, + "TCAACC": 1551, + "TCAACG": 1552, + "TCAAGA": 1553, + "TCAAGT": 1554, + "TCAAGC": 1555, + "TCAAGG": 1556, + "TCATAA": 1557, + "TCATAT": 1558, + "TCATAC": 1559, + "TCATAG": 1560, + "TCATTA": 1561, + "TCATTT": 1562, + "TCATTC": 1563, + "TCATTG": 1564, + "TCATCA": 1565, + "TCATCT": 1566, + "TCATCC": 1567, + "TCATCG": 1568, + "TCATGA": 1569, + "TCATGT": 1570, + "TCATGC": 1571, + "TCATGG": 1572, + "TCACAA": 1573, + "TCACAT": 1574, + "TCACAC": 1575, + "TCACAG": 1576, + "TCACTA": 1577, + "TCACTT": 1578, + "TCACTC": 1579, + "TCACTG": 1580, + "TCACCA": 1581, + "TCACCT": 1582, + "TCACCC": 1583, + "TCACCG": 1584, + "TCACGA": 1585, + "TCACGT": 1586, + "TCACGC": 1587, + "TCACGG": 1588, + "TCAGAA": 1589, + "TCAGAT": 1590, + "TCAGAC": 1591, + "TCAGAG": 1592, + "TCAGTA": 1593, + "TCAGTT": 1594, + "TCAGTC": 1595, + "TCAGTG": 1596, + "TCAGCA": 1597, + "TCAGCT": 1598, + "TCAGCC": 1599, + "TCAGCG": 1600, + "TCAGGA": 1601, + "TCAGGT": 1602, + "TCAGGC": 1603, + "TCAGGG": 1604, + "TCTAAA": 1605, + "TCTAAT": 1606, + "TCTAAC": 1607, + "TCTAAG": 1608, + "TCTATA": 1609, + "TCTATT": 1610, + "TCTATC": 1611, + "TCTATG": 1612, + "TCTACA": 1613, + "TCTACT": 1614, + "TCTACC": 1615, + "TCTACG": 1616, + "TCTAGA": 1617, + "TCTAGT": 1618, + "TCTAGC": 1619, + "TCTAGG": 1620, + "TCTTAA": 1621, + "TCTTAT": 1622, + "TCTTAC": 1623, + "TCTTAG": 1624, + "TCTTTA": 1625, + "TCTTTT": 1626, + "TCTTTC": 1627, + "TCTTTG": 1628, + "TCTTCA": 1629, + "TCTTCT": 1630, + "TCTTCC": 1631, + "TCTTCG": 1632, + "TCTTGA": 1633, + "TCTTGT": 1634, + "TCTTGC": 1635, + "TCTTGG": 1636, + "TCTCAA": 1637, + "TCTCAT": 1638, + "TCTCAC": 1639, + "TCTCAG": 1640, + "TCTCTA": 1641, + "TCTCTT": 1642, + "TCTCTC": 1643, + "TCTCTG": 1644, + "TCTCCA": 1645, + "TCTCCT": 1646, + "TCTCCC": 1647, + "TCTCCG": 1648, + "TCTCGA": 1649, + "TCTCGT": 1650, + "TCTCGC": 1651, + "TCTCGG": 1652, + "TCTGAA": 1653, + "TCTGAT": 1654, + "TCTGAC": 1655, + "TCTGAG": 1656, + "TCTGTA": 1657, + "TCTGTT": 1658, + "TCTGTC": 1659, + "TCTGTG": 1660, + "TCTGCA": 1661, + "TCTGCT": 1662, + "TCTGCC": 1663, + "TCTGCG": 1664, + "TCTGGA": 1665, + "TCTGGT": 1666, + "TCTGGC": 1667, + "TCTGGG": 1668, + "TCCAAA": 1669, + "TCCAAT": 1670, + "TCCAAC": 1671, + "TCCAAG": 1672, + "TCCATA": 1673, + "TCCATT": 1674, + "TCCATC": 1675, + "TCCATG": 1676, + "TCCACA": 1677, + "TCCACT": 1678, + "TCCACC": 1679, + "TCCACG": 1680, + "TCCAGA": 1681, + "TCCAGT": 1682, + "TCCAGC": 1683, + "TCCAGG": 1684, + "TCCTAA": 1685, + "TCCTAT": 1686, + "TCCTAC": 1687, + "TCCTAG": 1688, + "TCCTTA": 1689, + "TCCTTT": 1690, + "TCCTTC": 1691, + "TCCTTG": 1692, + "TCCTCA": 1693, + "TCCTCT": 1694, + "TCCTCC": 1695, + "TCCTCG": 1696, + "TCCTGA": 1697, + "TCCTGT": 1698, + "TCCTGC": 1699, + "TCCTGG": 1700, + "TCCCAA": 1701, + "TCCCAT": 1702, + "TCCCAC": 1703, + "TCCCAG": 1704, + "TCCCTA": 1705, + "TCCCTT": 1706, + "TCCCTC": 1707, + "TCCCTG": 1708, + "TCCCCA": 1709, + "TCCCCT": 1710, + "TCCCCC": 1711, + "TCCCCG": 1712, + "TCCCGA": 1713, + "TCCCGT": 1714, + "TCCCGC": 1715, + "TCCCGG": 1716, + "TCCGAA": 1717, + "TCCGAT": 1718, + "TCCGAC": 1719, + "TCCGAG": 1720, + "TCCGTA": 1721, + "TCCGTT": 1722, + "TCCGTC": 1723, + "TCCGTG": 1724, + "TCCGCA": 1725, + "TCCGCT": 1726, + "TCCGCC": 1727, + "TCCGCG": 1728, + "TCCGGA": 1729, + "TCCGGT": 1730, + "TCCGGC": 1731, + "TCCGGG": 1732, + "TCGAAA": 1733, + "TCGAAT": 1734, + "TCGAAC": 1735, + "TCGAAG": 1736, + "TCGATA": 1737, + "TCGATT": 1738, + "TCGATC": 1739, + "TCGATG": 1740, + "TCGACA": 1741, + "TCGACT": 1742, + "TCGACC": 1743, + "TCGACG": 1744, + "TCGAGA": 1745, + "TCGAGT": 1746, + "TCGAGC": 1747, + "TCGAGG": 1748, + "TCGTAA": 1749, + "TCGTAT": 1750, + "TCGTAC": 1751, + "TCGTAG": 1752, + "TCGTTA": 1753, + "TCGTTT": 1754, + "TCGTTC": 1755, + "TCGTTG": 1756, + "TCGTCA": 1757, + "TCGTCT": 1758, + "TCGTCC": 1759, + "TCGTCG": 1760, + "TCGTGA": 1761, + "TCGTGT": 1762, + "TCGTGC": 1763, + "TCGTGG": 1764, + "TCGCAA": 1765, + "TCGCAT": 1766, + "TCGCAC": 1767, + "TCGCAG": 1768, + "TCGCTA": 1769, + "TCGCTT": 1770, + "TCGCTC": 1771, + "TCGCTG": 1772, + "TCGCCA": 1773, + "TCGCCT": 1774, + "TCGCCC": 1775, + "TCGCCG": 1776, + "TCGCGA": 1777, + "TCGCGT": 1778, + "TCGCGC": 1779, + "TCGCGG": 1780, + "TCGGAA": 1781, + "TCGGAT": 1782, + "TCGGAC": 1783, + "TCGGAG": 1784, + "TCGGTA": 1785, + "TCGGTT": 1786, + "TCGGTC": 1787, + "TCGGTG": 1788, + "TCGGCA": 1789, + "TCGGCT": 1790, + "TCGGCC": 1791, + "TCGGCG": 1792, + "TCGGGA": 1793, + "TCGGGT": 1794, + "TCGGGC": 1795, + "TCGGGG": 1796, + "TGAAAA": 1797, + "TGAAAT": 1798, + "TGAAAC": 1799, + "TGAAAG": 1800, + "TGAATA": 1801, + "TGAATT": 1802, + "TGAATC": 1803, + "TGAATG": 1804, + "TGAACA": 1805, + "TGAACT": 1806, + "TGAACC": 1807, + "TGAACG": 1808, + "TGAAGA": 1809, + "TGAAGT": 1810, + "TGAAGC": 1811, + "TGAAGG": 1812, + "TGATAA": 1813, + "TGATAT": 1814, + "TGATAC": 1815, + "TGATAG": 1816, + "TGATTA": 1817, + "TGATTT": 1818, + "TGATTC": 1819, + "TGATTG": 1820, + "TGATCA": 1821, + "TGATCT": 1822, + "TGATCC": 1823, + "TGATCG": 1824, + "TGATGA": 1825, + "TGATGT": 1826, + "TGATGC": 1827, + "TGATGG": 1828, + "TGACAA": 1829, + "TGACAT": 1830, + "TGACAC": 1831, + "TGACAG": 1832, + "TGACTA": 1833, + "TGACTT": 1834, + "TGACTC": 1835, + "TGACTG": 1836, + "TGACCA": 1837, + "TGACCT": 1838, + "TGACCC": 1839, + "TGACCG": 1840, + "TGACGA": 1841, + "TGACGT": 1842, + "TGACGC": 1843, + "TGACGG": 1844, + "TGAGAA": 1845, + "TGAGAT": 1846, + "TGAGAC": 1847, + "TGAGAG": 1848, + "TGAGTA": 1849, + "TGAGTT": 1850, + "TGAGTC": 1851, + "TGAGTG": 1852, + "TGAGCA": 1853, + "TGAGCT": 1854, + "TGAGCC": 1855, + "TGAGCG": 1856, + "TGAGGA": 1857, + "TGAGGT": 1858, + "TGAGGC": 1859, + "TGAGGG": 1860, + "TGTAAA": 1861, + "TGTAAT": 1862, + "TGTAAC": 1863, + "TGTAAG": 1864, + "TGTATA": 1865, + "TGTATT": 1866, + "TGTATC": 1867, + "TGTATG": 1868, + "TGTACA": 1869, + "TGTACT": 1870, + "TGTACC": 1871, + "TGTACG": 1872, + "TGTAGA": 1873, + "TGTAGT": 1874, + "TGTAGC": 1875, + "TGTAGG": 1876, + "TGTTAA": 1877, + "TGTTAT": 1878, + "TGTTAC": 1879, + "TGTTAG": 1880, + "TGTTTA": 1881, + "TGTTTT": 1882, + "TGTTTC": 1883, + "TGTTTG": 1884, + "TGTTCA": 1885, + "TGTTCT": 1886, + "TGTTCC": 1887, + "TGTTCG": 1888, + "TGTTGA": 1889, + "TGTTGT": 1890, + "TGTTGC": 1891, + "TGTTGG": 1892, + "TGTCAA": 1893, + "TGTCAT": 1894, + "TGTCAC": 1895, + "TGTCAG": 1896, + "TGTCTA": 1897, + "TGTCTT": 1898, + "TGTCTC": 1899, + "TGTCTG": 1900, + "TGTCCA": 1901, + "TGTCCT": 1902, + "TGTCCC": 1903, + "TGTCCG": 1904, + "TGTCGA": 1905, + "TGTCGT": 1906, + "TGTCGC": 1907, + "TGTCGG": 1908, + "TGTGAA": 1909, + "TGTGAT": 1910, + "TGTGAC": 1911, + "TGTGAG": 1912, + "TGTGTA": 1913, + "TGTGTT": 1914, + "TGTGTC": 1915, + "TGTGTG": 1916, + "TGTGCA": 1917, + "TGTGCT": 1918, + "TGTGCC": 1919, + "TGTGCG": 1920, + "TGTGGA": 1921, + "TGTGGT": 1922, + "TGTGGC": 1923, + "TGTGGG": 1924, + "TGCAAA": 1925, + "TGCAAT": 1926, + "TGCAAC": 1927, + "TGCAAG": 1928, + "TGCATA": 1929, + "TGCATT": 1930, + "TGCATC": 1931, + "TGCATG": 1932, + "TGCACA": 1933, + "TGCACT": 1934, + "TGCACC": 1935, + "TGCACG": 1936, + "TGCAGA": 1937, + "TGCAGT": 1938, + "TGCAGC": 1939, + "TGCAGG": 1940, + "TGCTAA": 1941, + "TGCTAT": 1942, + "TGCTAC": 1943, + "TGCTAG": 1944, + "TGCTTA": 1945, + "TGCTTT": 1946, + "TGCTTC": 1947, + "TGCTTG": 1948, + "TGCTCA": 1949, + "TGCTCT": 1950, + "TGCTCC": 1951, + "TGCTCG": 1952, + "TGCTGA": 1953, + "TGCTGT": 1954, + "TGCTGC": 1955, + "TGCTGG": 1956, + "TGCCAA": 1957, + "TGCCAT": 1958, + "TGCCAC": 1959, + "TGCCAG": 1960, + "TGCCTA": 1961, + "TGCCTT": 1962, + "TGCCTC": 1963, + "TGCCTG": 1964, + "TGCCCA": 1965, + "TGCCCT": 1966, + "TGCCCC": 1967, + "TGCCCG": 1968, + "TGCCGA": 1969, + "TGCCGT": 1970, + "TGCCGC": 1971, + "TGCCGG": 1972, + "TGCGAA": 1973, + "TGCGAT": 1974, + "TGCGAC": 1975, + "TGCGAG": 1976, + "TGCGTA": 1977, + "TGCGTT": 1978, + "TGCGTC": 1979, + "TGCGTG": 1980, + "TGCGCA": 1981, + "TGCGCT": 1982, + "TGCGCC": 1983, + "TGCGCG": 1984, + "TGCGGA": 1985, + "TGCGGT": 1986, + "TGCGGC": 1987, + "TGCGGG": 1988, + "TGGAAA": 1989, + "TGGAAT": 1990, + "TGGAAC": 1991, + "TGGAAG": 1992, + "TGGATA": 1993, + "TGGATT": 1994, + "TGGATC": 1995, + "TGGATG": 1996, + "TGGACA": 1997, + "TGGACT": 1998, + "TGGACC": 1999, + "TGGACG": 2000, + "TGGAGA": 2001, + "TGGAGT": 2002, + "TGGAGC": 2003, + "TGGAGG": 2004, + "TGGTAA": 2005, + "TGGTAT": 2006, + "TGGTAC": 2007, + "TGGTAG": 2008, + "TGGTTA": 2009, + "TGGTTT": 2010, + "TGGTTC": 2011, + "TGGTTG": 2012, + "TGGTCA": 2013, + "TGGTCT": 2014, + "TGGTCC": 2015, + "TGGTCG": 2016, + "TGGTGA": 2017, + "TGGTGT": 2018, + "TGGTGC": 2019, + "TGGTGG": 2020, + "TGGCAA": 2021, + "TGGCAT": 2022, + "TGGCAC": 2023, + "TGGCAG": 2024, + "TGGCTA": 2025, + "TGGCTT": 2026, + "TGGCTC": 2027, + "TGGCTG": 2028, + "TGGCCA": 2029, + "TGGCCT": 2030, + "TGGCCC": 2031, + "TGGCCG": 2032, + "TGGCGA": 2033, + "TGGCGT": 2034, + "TGGCGC": 2035, + "TGGCGG": 2036, + "TGGGAA": 2037, + "TGGGAT": 2038, + "TGGGAC": 2039, + "TGGGAG": 2040, + "TGGGTA": 2041, + "TGGGTT": 2042, + "TGGGTC": 2043, + "TGGGTG": 2044, + "TGGGCA": 2045, + "TGGGCT": 2046, + "TGGGCC": 2047, + "TGGGCG": 2048, + "TGGGGA": 2049, + "TGGGGT": 2050, + "TGGGGC": 2051, + "TGGGGG": 2052, + "CAAAAA": 2053, + "CAAAAT": 2054, + "CAAAAC": 2055, + "CAAAAG": 2056, + "CAAATA": 2057, + "CAAATT": 2058, + "CAAATC": 2059, + "CAAATG": 2060, + "CAAACA": 2061, + "CAAACT": 2062, + "CAAACC": 2063, + "CAAACG": 2064, + "CAAAGA": 2065, + "CAAAGT": 2066, + "CAAAGC": 2067, + "CAAAGG": 2068, + "CAATAA": 2069, + "CAATAT": 2070, + "CAATAC": 2071, + "CAATAG": 2072, + "CAATTA": 2073, + "CAATTT": 2074, + "CAATTC": 2075, + "CAATTG": 2076, + "CAATCA": 2077, + "CAATCT": 2078, + "CAATCC": 2079, + "CAATCG": 2080, + "CAATGA": 2081, + "CAATGT": 2082, + "CAATGC": 2083, + "CAATGG": 2084, + "CAACAA": 2085, + "CAACAT": 2086, + "CAACAC": 2087, + "CAACAG": 2088, + "CAACTA": 2089, + "CAACTT": 2090, + "CAACTC": 2091, + "CAACTG": 2092, + "CAACCA": 2093, + "CAACCT": 2094, + "CAACCC": 2095, + "CAACCG": 2096, + "CAACGA": 2097, + "CAACGT": 2098, + "CAACGC": 2099, + "CAACGG": 2100, + "CAAGAA": 2101, + "CAAGAT": 2102, + "CAAGAC": 2103, + "CAAGAG": 2104, + "CAAGTA": 2105, + "CAAGTT": 2106, + "CAAGTC": 2107, + "CAAGTG": 2108, + "CAAGCA": 2109, + "CAAGCT": 2110, + "CAAGCC": 2111, + "CAAGCG": 2112, + "CAAGGA": 2113, + "CAAGGT": 2114, + "CAAGGC": 2115, + "CAAGGG": 2116, + "CATAAA": 2117, + "CATAAT": 2118, + "CATAAC": 2119, + "CATAAG": 2120, + "CATATA": 2121, + "CATATT": 2122, + "CATATC": 2123, + "CATATG": 2124, + "CATACA": 2125, + "CATACT": 2126, + "CATACC": 2127, + "CATACG": 2128, + "CATAGA": 2129, + "CATAGT": 2130, + "CATAGC": 2131, + "CATAGG": 2132, + "CATTAA": 2133, + "CATTAT": 2134, + "CATTAC": 2135, + "CATTAG": 2136, + "CATTTA": 2137, + "CATTTT": 2138, + "CATTTC": 2139, + "CATTTG": 2140, + "CATTCA": 2141, + "CATTCT": 2142, + "CATTCC": 2143, + "CATTCG": 2144, + "CATTGA": 2145, + "CATTGT": 2146, + "CATTGC": 2147, + "CATTGG": 2148, + "CATCAA": 2149, + "CATCAT": 2150, + "CATCAC": 2151, + "CATCAG": 2152, + "CATCTA": 2153, + "CATCTT": 2154, + "CATCTC": 2155, + "CATCTG": 2156, + "CATCCA": 2157, + "CATCCT": 2158, + "CATCCC": 2159, + "CATCCG": 2160, + "CATCGA": 2161, + "CATCGT": 2162, + "CATCGC": 2163, + "CATCGG": 2164, + "CATGAA": 2165, + "CATGAT": 2166, + "CATGAC": 2167, + "CATGAG": 2168, + "CATGTA": 2169, + "CATGTT": 2170, + "CATGTC": 2171, + "CATGTG": 2172, + "CATGCA": 2173, + "CATGCT": 2174, + "CATGCC": 2175, + "CATGCG": 2176, + "CATGGA": 2177, + "CATGGT": 2178, + "CATGGC": 2179, + "CATGGG": 2180, + "CACAAA": 2181, + "CACAAT": 2182, + "CACAAC": 2183, + "CACAAG": 2184, + "CACATA": 2185, + "CACATT": 2186, + "CACATC": 2187, + "CACATG": 2188, + "CACACA": 2189, + "CACACT": 2190, + "CACACC": 2191, + "CACACG": 2192, + "CACAGA": 2193, + "CACAGT": 2194, + "CACAGC": 2195, + "CACAGG": 2196, + "CACTAA": 2197, + "CACTAT": 2198, + "CACTAC": 2199, + "CACTAG": 2200, + "CACTTA": 2201, + "CACTTT": 2202, + "CACTTC": 2203, + "CACTTG": 2204, + "CACTCA": 2205, + "CACTCT": 2206, + "CACTCC": 2207, + "CACTCG": 2208, + "CACTGA": 2209, + "CACTGT": 2210, + "CACTGC": 2211, + "CACTGG": 2212, + "CACCAA": 2213, + "CACCAT": 2214, + "CACCAC": 2215, + "CACCAG": 2216, + "CACCTA": 2217, + "CACCTT": 2218, + "CACCTC": 2219, + "CACCTG": 2220, + "CACCCA": 2221, + "CACCCT": 2222, + "CACCCC": 2223, + "CACCCG": 2224, + "CACCGA": 2225, + "CACCGT": 2226, + "CACCGC": 2227, + "CACCGG": 2228, + "CACGAA": 2229, + "CACGAT": 2230, + "CACGAC": 2231, + "CACGAG": 2232, + "CACGTA": 2233, + "CACGTT": 2234, + "CACGTC": 2235, + "CACGTG": 2236, + "CACGCA": 2237, + "CACGCT": 2238, + "CACGCC": 2239, + "CACGCG": 2240, + "CACGGA": 2241, + "CACGGT": 2242, + "CACGGC": 2243, + "CACGGG": 2244, + "CAGAAA": 2245, + "CAGAAT": 2246, + "CAGAAC": 2247, + "CAGAAG": 2248, + "CAGATA": 2249, + "CAGATT": 2250, + "CAGATC": 2251, + "CAGATG": 2252, + "CAGACA": 2253, + "CAGACT": 2254, + "CAGACC": 2255, + "CAGACG": 2256, + "CAGAGA": 2257, + "CAGAGT": 2258, + "CAGAGC": 2259, + "CAGAGG": 2260, + "CAGTAA": 2261, + "CAGTAT": 2262, + "CAGTAC": 2263, + "CAGTAG": 2264, + "CAGTTA": 2265, + "CAGTTT": 2266, + "CAGTTC": 2267, + "CAGTTG": 2268, + "CAGTCA": 2269, + "CAGTCT": 2270, + "CAGTCC": 2271, + "CAGTCG": 2272, + "CAGTGA": 2273, + "CAGTGT": 2274, + "CAGTGC": 2275, + "CAGTGG": 2276, + "CAGCAA": 2277, + "CAGCAT": 2278, + "CAGCAC": 2279, + "CAGCAG": 2280, + "CAGCTA": 2281, + "CAGCTT": 2282, + "CAGCTC": 2283, + "CAGCTG": 2284, + "CAGCCA": 2285, + "CAGCCT": 2286, + "CAGCCC": 2287, + "CAGCCG": 2288, + "CAGCGA": 2289, + "CAGCGT": 2290, + "CAGCGC": 2291, + "CAGCGG": 2292, + "CAGGAA": 2293, + "CAGGAT": 2294, + "CAGGAC": 2295, + "CAGGAG": 2296, + "CAGGTA": 2297, + "CAGGTT": 2298, + "CAGGTC": 2299, + "CAGGTG": 2300, + "CAGGCA": 2301, + "CAGGCT": 2302, + "CAGGCC": 2303, + "CAGGCG": 2304, + "CAGGGA": 2305, + "CAGGGT": 2306, + "CAGGGC": 2307, + "CAGGGG": 2308, + "CTAAAA": 2309, + "CTAAAT": 2310, + "CTAAAC": 2311, + "CTAAAG": 2312, + "CTAATA": 2313, + "CTAATT": 2314, + "CTAATC": 2315, + "CTAATG": 2316, + "CTAACA": 2317, + "CTAACT": 2318, + "CTAACC": 2319, + "CTAACG": 2320, + "CTAAGA": 2321, + "CTAAGT": 2322, + "CTAAGC": 2323, + "CTAAGG": 2324, + "CTATAA": 2325, + "CTATAT": 2326, + "CTATAC": 2327, + "CTATAG": 2328, + "CTATTA": 2329, + "CTATTT": 2330, + "CTATTC": 2331, + "CTATTG": 2332, + "CTATCA": 2333, + "CTATCT": 2334, + "CTATCC": 2335, + "CTATCG": 2336, + "CTATGA": 2337, + "CTATGT": 2338, + "CTATGC": 2339, + "CTATGG": 2340, + "CTACAA": 2341, + "CTACAT": 2342, + "CTACAC": 2343, + "CTACAG": 2344, + "CTACTA": 2345, + "CTACTT": 2346, + "CTACTC": 2347, + "CTACTG": 2348, + "CTACCA": 2349, + "CTACCT": 2350, + "CTACCC": 2351, + "CTACCG": 2352, + "CTACGA": 2353, + "CTACGT": 2354, + "CTACGC": 2355, + "CTACGG": 2356, + "CTAGAA": 2357, + "CTAGAT": 2358, + "CTAGAC": 2359, + "CTAGAG": 2360, + "CTAGTA": 2361, + "CTAGTT": 2362, + "CTAGTC": 2363, + "CTAGTG": 2364, + "CTAGCA": 2365, + "CTAGCT": 2366, + "CTAGCC": 2367, + "CTAGCG": 2368, + "CTAGGA": 2369, + "CTAGGT": 2370, + "CTAGGC": 2371, + "CTAGGG": 2372, + "CTTAAA": 2373, + "CTTAAT": 2374, + "CTTAAC": 2375, + "CTTAAG": 2376, + "CTTATA": 2377, + "CTTATT": 2378, + "CTTATC": 2379, + "CTTATG": 2380, + "CTTACA": 2381, + "CTTACT": 2382, + "CTTACC": 2383, + "CTTACG": 2384, + "CTTAGA": 2385, + "CTTAGT": 2386, + "CTTAGC": 2387, + "CTTAGG": 2388, + "CTTTAA": 2389, + "CTTTAT": 2390, + "CTTTAC": 2391, + "CTTTAG": 2392, + "CTTTTA": 2393, + "CTTTTT": 2394, + "CTTTTC": 2395, + "CTTTTG": 2396, + "CTTTCA": 2397, + "CTTTCT": 2398, + "CTTTCC": 2399, + "CTTTCG": 2400, + "CTTTGA": 2401, + "CTTTGT": 2402, + "CTTTGC": 2403, + "CTTTGG": 2404, + "CTTCAA": 2405, + "CTTCAT": 2406, + "CTTCAC": 2407, + "CTTCAG": 2408, + "CTTCTA": 2409, + "CTTCTT": 2410, + "CTTCTC": 2411, + "CTTCTG": 2412, + "CTTCCA": 2413, + "CTTCCT": 2414, + "CTTCCC": 2415, + "CTTCCG": 2416, + "CTTCGA": 2417, + "CTTCGT": 2418, + "CTTCGC": 2419, + "CTTCGG": 2420, + "CTTGAA": 2421, + "CTTGAT": 2422, + "CTTGAC": 2423, + "CTTGAG": 2424, + "CTTGTA": 2425, + "CTTGTT": 2426, + "CTTGTC": 2427, + "CTTGTG": 2428, + "CTTGCA": 2429, + "CTTGCT": 2430, + "CTTGCC": 2431, + "CTTGCG": 2432, + "CTTGGA": 2433, + "CTTGGT": 2434, + "CTTGGC": 2435, + "CTTGGG": 2436, + "CTCAAA": 2437, + "CTCAAT": 2438, + "CTCAAC": 2439, + "CTCAAG": 2440, + "CTCATA": 2441, + "CTCATT": 2442, + "CTCATC": 2443, + "CTCATG": 2444, + "CTCACA": 2445, + "CTCACT": 2446, + "CTCACC": 2447, + "CTCACG": 2448, + "CTCAGA": 2449, + "CTCAGT": 2450, + "CTCAGC": 2451, + "CTCAGG": 2452, + "CTCTAA": 2453, + "CTCTAT": 2454, + "CTCTAC": 2455, + "CTCTAG": 2456, + "CTCTTA": 2457, + "CTCTTT": 2458, + "CTCTTC": 2459, + "CTCTTG": 2460, + "CTCTCA": 2461, + "CTCTCT": 2462, + "CTCTCC": 2463, + "CTCTCG": 2464, + "CTCTGA": 2465, + "CTCTGT": 2466, + "CTCTGC": 2467, + "CTCTGG": 2468, + "CTCCAA": 2469, + "CTCCAT": 2470, + "CTCCAC": 2471, + "CTCCAG": 2472, + "CTCCTA": 2473, + "CTCCTT": 2474, + "CTCCTC": 2475, + "CTCCTG": 2476, + "CTCCCA": 2477, + "CTCCCT": 2478, + "CTCCCC": 2479, + "CTCCCG": 2480, + "CTCCGA": 2481, + "CTCCGT": 2482, + "CTCCGC": 2483, + "CTCCGG": 2484, + "CTCGAA": 2485, + "CTCGAT": 2486, + "CTCGAC": 2487, + "CTCGAG": 2488, + "CTCGTA": 2489, + "CTCGTT": 2490, + "CTCGTC": 2491, + "CTCGTG": 2492, + "CTCGCA": 2493, + "CTCGCT": 2494, + "CTCGCC": 2495, + "CTCGCG": 2496, + "CTCGGA": 2497, + "CTCGGT": 2498, + "CTCGGC": 2499, + "CTCGGG": 2500, + "CTGAAA": 2501, + "CTGAAT": 2502, + "CTGAAC": 2503, + "CTGAAG": 2504, + "CTGATA": 2505, + "CTGATT": 2506, + "CTGATC": 2507, + "CTGATG": 2508, + "CTGACA": 2509, + "CTGACT": 2510, + "CTGACC": 2511, + "CTGACG": 2512, + "CTGAGA": 2513, + "CTGAGT": 2514, + "CTGAGC": 2515, + "CTGAGG": 2516, + "CTGTAA": 2517, + "CTGTAT": 2518, + "CTGTAC": 2519, + "CTGTAG": 2520, + "CTGTTA": 2521, + "CTGTTT": 2522, + "CTGTTC": 2523, + "CTGTTG": 2524, + "CTGTCA": 2525, + "CTGTCT": 2526, + "CTGTCC": 2527, + "CTGTCG": 2528, + "CTGTGA": 2529, + "CTGTGT": 2530, + "CTGTGC": 2531, + "CTGTGG": 2532, + "CTGCAA": 2533, + "CTGCAT": 2534, + "CTGCAC": 2535, + "CTGCAG": 2536, + "CTGCTA": 2537, + "CTGCTT": 2538, + "CTGCTC": 2539, + "CTGCTG": 2540, + "CTGCCA": 2541, + "CTGCCT": 2542, + "CTGCCC": 2543, + "CTGCCG": 2544, + "CTGCGA": 2545, + "CTGCGT": 2546, + "CTGCGC": 2547, + "CTGCGG": 2548, + "CTGGAA": 2549, + "CTGGAT": 2550, + "CTGGAC": 2551, + "CTGGAG": 2552, + "CTGGTA": 2553, + "CTGGTT": 2554, + "CTGGTC": 2555, + "CTGGTG": 2556, + "CTGGCA": 2557, + "CTGGCT": 2558, + "CTGGCC": 2559, + "CTGGCG": 2560, + "CTGGGA": 2561, + "CTGGGT": 2562, + "CTGGGC": 2563, + "CTGGGG": 2564, + "CCAAAA": 2565, + "CCAAAT": 2566, + "CCAAAC": 2567, + "CCAAAG": 2568, + "CCAATA": 2569, + "CCAATT": 2570, + "CCAATC": 2571, + "CCAATG": 2572, + "CCAACA": 2573, + "CCAACT": 2574, + "CCAACC": 2575, + "CCAACG": 2576, + "CCAAGA": 2577, + "CCAAGT": 2578, + "CCAAGC": 2579, + "CCAAGG": 2580, + "CCATAA": 2581, + "CCATAT": 2582, + "CCATAC": 2583, + "CCATAG": 2584, + "CCATTA": 2585, + "CCATTT": 2586, + "CCATTC": 2587, + "CCATTG": 2588, + "CCATCA": 2589, + "CCATCT": 2590, + "CCATCC": 2591, + "CCATCG": 2592, + "CCATGA": 2593, + "CCATGT": 2594, + "CCATGC": 2595, + "CCATGG": 2596, + "CCACAA": 2597, + "CCACAT": 2598, + "CCACAC": 2599, + "CCACAG": 2600, + "CCACTA": 2601, + "CCACTT": 2602, + "CCACTC": 2603, + "CCACTG": 2604, + "CCACCA": 2605, + "CCACCT": 2606, + "CCACCC": 2607, + "CCACCG": 2608, + "CCACGA": 2609, + "CCACGT": 2610, + "CCACGC": 2611, + "CCACGG": 2612, + "CCAGAA": 2613, + "CCAGAT": 2614, + "CCAGAC": 2615, + "CCAGAG": 2616, + "CCAGTA": 2617, + "CCAGTT": 2618, + "CCAGTC": 2619, + "CCAGTG": 2620, + "CCAGCA": 2621, + "CCAGCT": 2622, + "CCAGCC": 2623, + "CCAGCG": 2624, + "CCAGGA": 2625, + "CCAGGT": 2626, + "CCAGGC": 2627, + "CCAGGG": 2628, + "CCTAAA": 2629, + "CCTAAT": 2630, + "CCTAAC": 2631, + "CCTAAG": 2632, + "CCTATA": 2633, + "CCTATT": 2634, + "CCTATC": 2635, + "CCTATG": 2636, + "CCTACA": 2637, + "CCTACT": 2638, + "CCTACC": 2639, + "CCTACG": 2640, + "CCTAGA": 2641, + "CCTAGT": 2642, + "CCTAGC": 2643, + "CCTAGG": 2644, + "CCTTAA": 2645, + "CCTTAT": 2646, + "CCTTAC": 2647, + "CCTTAG": 2648, + "CCTTTA": 2649, + "CCTTTT": 2650, + "CCTTTC": 2651, + "CCTTTG": 2652, + "CCTTCA": 2653, + "CCTTCT": 2654, + "CCTTCC": 2655, + "CCTTCG": 2656, + "CCTTGA": 2657, + "CCTTGT": 2658, + "CCTTGC": 2659, + "CCTTGG": 2660, + "CCTCAA": 2661, + "CCTCAT": 2662, + "CCTCAC": 2663, + "CCTCAG": 2664, + "CCTCTA": 2665, + "CCTCTT": 2666, + "CCTCTC": 2667, + "CCTCTG": 2668, + "CCTCCA": 2669, + "CCTCCT": 2670, + "CCTCCC": 2671, + "CCTCCG": 2672, + "CCTCGA": 2673, + "CCTCGT": 2674, + "CCTCGC": 2675, + "CCTCGG": 2676, + "CCTGAA": 2677, + "CCTGAT": 2678, + "CCTGAC": 2679, + "CCTGAG": 2680, + "CCTGTA": 2681, + "CCTGTT": 2682, + "CCTGTC": 2683, + "CCTGTG": 2684, + "CCTGCA": 2685, + "CCTGCT": 2686, + "CCTGCC": 2687, + "CCTGCG": 2688, + "CCTGGA": 2689, + "CCTGGT": 2690, + "CCTGGC": 2691, + "CCTGGG": 2692, + "CCCAAA": 2693, + "CCCAAT": 2694, + "CCCAAC": 2695, + "CCCAAG": 2696, + "CCCATA": 2697, + "CCCATT": 2698, + "CCCATC": 2699, + "CCCATG": 2700, + "CCCACA": 2701, + "CCCACT": 2702, + "CCCACC": 2703, + "CCCACG": 2704, + "CCCAGA": 2705, + "CCCAGT": 2706, + "CCCAGC": 2707, + "CCCAGG": 2708, + "CCCTAA": 2709, + "CCCTAT": 2710, + "CCCTAC": 2711, + "CCCTAG": 2712, + "CCCTTA": 2713, + "CCCTTT": 2714, + "CCCTTC": 2715, + "CCCTTG": 2716, + "CCCTCA": 2717, + "CCCTCT": 2718, + "CCCTCC": 2719, + "CCCTCG": 2720, + "CCCTGA": 2721, + "CCCTGT": 2722, + "CCCTGC": 2723, + "CCCTGG": 2724, + "CCCCAA": 2725, + "CCCCAT": 2726, + "CCCCAC": 2727, + "CCCCAG": 2728, + "CCCCTA": 2729, + "CCCCTT": 2730, + "CCCCTC": 2731, + "CCCCTG": 2732, + "CCCCCA": 2733, + "CCCCCT": 2734, + "CCCCCC": 2735, + "CCCCCG": 2736, + "CCCCGA": 2737, + "CCCCGT": 2738, + "CCCCGC": 2739, + "CCCCGG": 2740, + "CCCGAA": 2741, + "CCCGAT": 2742, + "CCCGAC": 2743, + "CCCGAG": 2744, + "CCCGTA": 2745, + "CCCGTT": 2746, + "CCCGTC": 2747, + "CCCGTG": 2748, + "CCCGCA": 2749, + "CCCGCT": 2750, + "CCCGCC": 2751, + "CCCGCG": 2752, + "CCCGGA": 2753, + "CCCGGT": 2754, + "CCCGGC": 2755, + "CCCGGG": 2756, + "CCGAAA": 2757, + "CCGAAT": 2758, + "CCGAAC": 2759, + "CCGAAG": 2760, + "CCGATA": 2761, + "CCGATT": 2762, + "CCGATC": 2763, + "CCGATG": 2764, + "CCGACA": 2765, + "CCGACT": 2766, + "CCGACC": 2767, + "CCGACG": 2768, + "CCGAGA": 2769, + "CCGAGT": 2770, + "CCGAGC": 2771, + "CCGAGG": 2772, + "CCGTAA": 2773, + "CCGTAT": 2774, + "CCGTAC": 2775, + "CCGTAG": 2776, + "CCGTTA": 2777, + "CCGTTT": 2778, + "CCGTTC": 2779, + "CCGTTG": 2780, + "CCGTCA": 2781, + "CCGTCT": 2782, + "CCGTCC": 2783, + "CCGTCG": 2784, + "CCGTGA": 2785, + "CCGTGT": 2786, + "CCGTGC": 2787, + "CCGTGG": 2788, + "CCGCAA": 2789, + "CCGCAT": 2790, + "CCGCAC": 2791, + "CCGCAG": 2792, + "CCGCTA": 2793, + "CCGCTT": 2794, + "CCGCTC": 2795, + "CCGCTG": 2796, + "CCGCCA": 2797, + "CCGCCT": 2798, + "CCGCCC": 2799, + "CCGCCG": 2800, + "CCGCGA": 2801, + "CCGCGT": 2802, + "CCGCGC": 2803, + "CCGCGG": 2804, + "CCGGAA": 2805, + "CCGGAT": 2806, + "CCGGAC": 2807, + "CCGGAG": 2808, + "CCGGTA": 2809, + "CCGGTT": 2810, + "CCGGTC": 2811, + "CCGGTG": 2812, + "CCGGCA": 2813, + "CCGGCT": 2814, + "CCGGCC": 2815, + "CCGGCG": 2816, + "CCGGGA": 2817, + "CCGGGT": 2818, + "CCGGGC": 2819, + "CCGGGG": 2820, + "CGAAAA": 2821, + "CGAAAT": 2822, + "CGAAAC": 2823, + "CGAAAG": 2824, + "CGAATA": 2825, + "CGAATT": 2826, + "CGAATC": 2827, + "CGAATG": 2828, + "CGAACA": 2829, + "CGAACT": 2830, + "CGAACC": 2831, + "CGAACG": 2832, + "CGAAGA": 2833, + "CGAAGT": 2834, + "CGAAGC": 2835, + "CGAAGG": 2836, + "CGATAA": 2837, + "CGATAT": 2838, + "CGATAC": 2839, + "CGATAG": 2840, + "CGATTA": 2841, + "CGATTT": 2842, + "CGATTC": 2843, + "CGATTG": 2844, + "CGATCA": 2845, + "CGATCT": 2846, + "CGATCC": 2847, + "CGATCG": 2848, + "CGATGA": 2849, + "CGATGT": 2850, + "CGATGC": 2851, + "CGATGG": 2852, + "CGACAA": 2853, + "CGACAT": 2854, + "CGACAC": 2855, + "CGACAG": 2856, + "CGACTA": 2857, + "CGACTT": 2858, + "CGACTC": 2859, + "CGACTG": 2860, + "CGACCA": 2861, + "CGACCT": 2862, + "CGACCC": 2863, + "CGACCG": 2864, + "CGACGA": 2865, + "CGACGT": 2866, + "CGACGC": 2867, + "CGACGG": 2868, + "CGAGAA": 2869, + "CGAGAT": 2870, + "CGAGAC": 2871, + "CGAGAG": 2872, + "CGAGTA": 2873, + "CGAGTT": 2874, + "CGAGTC": 2875, + "CGAGTG": 2876, + "CGAGCA": 2877, + "CGAGCT": 2878, + "CGAGCC": 2879, + "CGAGCG": 2880, + "CGAGGA": 2881, + "CGAGGT": 2882, + "CGAGGC": 2883, + "CGAGGG": 2884, + "CGTAAA": 2885, + "CGTAAT": 2886, + "CGTAAC": 2887, + "CGTAAG": 2888, + "CGTATA": 2889, + "CGTATT": 2890, + "CGTATC": 2891, + "CGTATG": 2892, + "CGTACA": 2893, + "CGTACT": 2894, + "CGTACC": 2895, + "CGTACG": 2896, + "CGTAGA": 2897, + "CGTAGT": 2898, + "CGTAGC": 2899, + "CGTAGG": 2900, + "CGTTAA": 2901, + "CGTTAT": 2902, + "CGTTAC": 2903, + "CGTTAG": 2904, + "CGTTTA": 2905, + "CGTTTT": 2906, + "CGTTTC": 2907, + "CGTTTG": 2908, + "CGTTCA": 2909, + "CGTTCT": 2910, + "CGTTCC": 2911, + "CGTTCG": 2912, + "CGTTGA": 2913, + "CGTTGT": 2914, + "CGTTGC": 2915, + "CGTTGG": 2916, + "CGTCAA": 2917, + "CGTCAT": 2918, + "CGTCAC": 2919, + "CGTCAG": 2920, + "CGTCTA": 2921, + "CGTCTT": 2922, + "CGTCTC": 2923, + "CGTCTG": 2924, + "CGTCCA": 2925, + "CGTCCT": 2926, + "CGTCCC": 2927, + "CGTCCG": 2928, + "CGTCGA": 2929, + "CGTCGT": 2930, + "CGTCGC": 2931, + "CGTCGG": 2932, + "CGTGAA": 2933, + "CGTGAT": 2934, + "CGTGAC": 2935, + "CGTGAG": 2936, + "CGTGTA": 2937, + "CGTGTT": 2938, + "CGTGTC": 2939, + "CGTGTG": 2940, + "CGTGCA": 2941, + "CGTGCT": 2942, + "CGTGCC": 2943, + "CGTGCG": 2944, + "CGTGGA": 2945, + "CGTGGT": 2946, + "CGTGGC": 2947, + "CGTGGG": 2948, + "CGCAAA": 2949, + "CGCAAT": 2950, + "CGCAAC": 2951, + "CGCAAG": 2952, + "CGCATA": 2953, + "CGCATT": 2954, + "CGCATC": 2955, + "CGCATG": 2956, + "CGCACA": 2957, + "CGCACT": 2958, + "CGCACC": 2959, + "CGCACG": 2960, + "CGCAGA": 2961, + "CGCAGT": 2962, + "CGCAGC": 2963, + "CGCAGG": 2964, + "CGCTAA": 2965, + "CGCTAT": 2966, + "CGCTAC": 2967, + "CGCTAG": 2968, + "CGCTTA": 2969, + "CGCTTT": 2970, + "CGCTTC": 2971, + "CGCTTG": 2972, + "CGCTCA": 2973, + "CGCTCT": 2974, + "CGCTCC": 2975, + "CGCTCG": 2976, + "CGCTGA": 2977, + "CGCTGT": 2978, + "CGCTGC": 2979, + "CGCTGG": 2980, + "CGCCAA": 2981, + "CGCCAT": 2982, + "CGCCAC": 2983, + "CGCCAG": 2984, + "CGCCTA": 2985, + "CGCCTT": 2986, + "CGCCTC": 2987, + "CGCCTG": 2988, + "CGCCCA": 2989, + "CGCCCT": 2990, + "CGCCCC": 2991, + "CGCCCG": 2992, + "CGCCGA": 2993, + "CGCCGT": 2994, + "CGCCGC": 2995, + "CGCCGG": 2996, + "CGCGAA": 2997, + "CGCGAT": 2998, + "CGCGAC": 2999, + "CGCGAG": 3000, + "CGCGTA": 3001, + "CGCGTT": 3002, + "CGCGTC": 3003, + "CGCGTG": 3004, + "CGCGCA": 3005, + "CGCGCT": 3006, + "CGCGCC": 3007, + "CGCGCG": 3008, + "CGCGGA": 3009, + "CGCGGT": 3010, + "CGCGGC": 3011, + "CGCGGG": 3012, + "CGGAAA": 3013, + "CGGAAT": 3014, + "CGGAAC": 3015, + "CGGAAG": 3016, + "CGGATA": 3017, + "CGGATT": 3018, + "CGGATC": 3019, + "CGGATG": 3020, + "CGGACA": 3021, + "CGGACT": 3022, + "CGGACC": 3023, + "CGGACG": 3024, + "CGGAGA": 3025, + "CGGAGT": 3026, + "CGGAGC": 3027, + "CGGAGG": 3028, + "CGGTAA": 3029, + "CGGTAT": 3030, + "CGGTAC": 3031, + "CGGTAG": 3032, + "CGGTTA": 3033, + "CGGTTT": 3034, + "CGGTTC": 3035, + "CGGTTG": 3036, + "CGGTCA": 3037, + "CGGTCT": 3038, + "CGGTCC": 3039, + "CGGTCG": 3040, + "CGGTGA": 3041, + "CGGTGT": 3042, + "CGGTGC": 3043, + "CGGTGG": 3044, + "CGGCAA": 3045, + "CGGCAT": 3046, + "CGGCAC": 3047, + "CGGCAG": 3048, + "CGGCTA": 3049, + "CGGCTT": 3050, + "CGGCTC": 3051, + "CGGCTG": 3052, + "CGGCCA": 3053, + "CGGCCT": 3054, + "CGGCCC": 3055, + "CGGCCG": 3056, + "CGGCGA": 3057, + "CGGCGT": 3058, + "CGGCGC": 3059, + "CGGCGG": 3060, + "CGGGAA": 3061, + "CGGGAT": 3062, + "CGGGAC": 3063, + "CGGGAG": 3064, + "CGGGTA": 3065, + "CGGGTT": 3066, + "CGGGTC": 3067, + "CGGGTG": 3068, + "CGGGCA": 3069, + "CGGGCT": 3070, + "CGGGCC": 3071, + "CGGGCG": 3072, + "CGGGGA": 3073, + "CGGGGT": 3074, + "CGGGGC": 3075, + "CGGGGG": 3076, + "GAAAAA": 3077, + "GAAAAT": 3078, + "GAAAAC": 3079, + "GAAAAG": 3080, + "GAAATA": 3081, + "GAAATT": 3082, + "GAAATC": 3083, + "GAAATG": 3084, + "GAAACA": 3085, + "GAAACT": 3086, + "GAAACC": 3087, + "GAAACG": 3088, + "GAAAGA": 3089, + "GAAAGT": 3090, + "GAAAGC": 3091, + "GAAAGG": 3092, + "GAATAA": 3093, + "GAATAT": 3094, + "GAATAC": 3095, + "GAATAG": 3096, + "GAATTA": 3097, + "GAATTT": 3098, + "GAATTC": 3099, + "GAATTG": 3100, + "GAATCA": 3101, + "GAATCT": 3102, + "GAATCC": 3103, + "GAATCG": 3104, + "GAATGA": 3105, + "GAATGT": 3106, + "GAATGC": 3107, + "GAATGG": 3108, + "GAACAA": 3109, + "GAACAT": 3110, + "GAACAC": 3111, + "GAACAG": 3112, + "GAACTA": 3113, + "GAACTT": 3114, + "GAACTC": 3115, + "GAACTG": 3116, + "GAACCA": 3117, + "GAACCT": 3118, + "GAACCC": 3119, + "GAACCG": 3120, + "GAACGA": 3121, + "GAACGT": 3122, + "GAACGC": 3123, + "GAACGG": 3124, + "GAAGAA": 3125, + "GAAGAT": 3126, + "GAAGAC": 3127, + "GAAGAG": 3128, + "GAAGTA": 3129, + "GAAGTT": 3130, + "GAAGTC": 3131, + "GAAGTG": 3132, + "GAAGCA": 3133, + "GAAGCT": 3134, + "GAAGCC": 3135, + "GAAGCG": 3136, + "GAAGGA": 3137, + "GAAGGT": 3138, + "GAAGGC": 3139, + "GAAGGG": 3140, + "GATAAA": 3141, + "GATAAT": 3142, + "GATAAC": 3143, + "GATAAG": 3144, + "GATATA": 3145, + "GATATT": 3146, + "GATATC": 3147, + "GATATG": 3148, + "GATACA": 3149, + "GATACT": 3150, + "GATACC": 3151, + "GATACG": 3152, + "GATAGA": 3153, + "GATAGT": 3154, + "GATAGC": 3155, + "GATAGG": 3156, + "GATTAA": 3157, + "GATTAT": 3158, + "GATTAC": 3159, + "GATTAG": 3160, + "GATTTA": 3161, + "GATTTT": 3162, + "GATTTC": 3163, + "GATTTG": 3164, + "GATTCA": 3165, + "GATTCT": 3166, + "GATTCC": 3167, + "GATTCG": 3168, + "GATTGA": 3169, + "GATTGT": 3170, + "GATTGC": 3171, + "GATTGG": 3172, + "GATCAA": 3173, + "GATCAT": 3174, + "GATCAC": 3175, + "GATCAG": 3176, + "GATCTA": 3177, + "GATCTT": 3178, + "GATCTC": 3179, + "GATCTG": 3180, + "GATCCA": 3181, + "GATCCT": 3182, + "GATCCC": 3183, + "GATCCG": 3184, + "GATCGA": 3185, + "GATCGT": 3186, + "GATCGC": 3187, + "GATCGG": 3188, + "GATGAA": 3189, + "GATGAT": 3190, + "GATGAC": 3191, + "GATGAG": 3192, + "GATGTA": 3193, + "GATGTT": 3194, + "GATGTC": 3195, + "GATGTG": 3196, + "GATGCA": 3197, + "GATGCT": 3198, + "GATGCC": 3199, + "GATGCG": 3200, + "GATGGA": 3201, + "GATGGT": 3202, + "GATGGC": 3203, + "GATGGG": 3204, + "GACAAA": 3205, + "GACAAT": 3206, + "GACAAC": 3207, + "GACAAG": 3208, + "GACATA": 3209, + "GACATT": 3210, + "GACATC": 3211, + "GACATG": 3212, + "GACACA": 3213, + "GACACT": 3214, + "GACACC": 3215, + "GACACG": 3216, + "GACAGA": 3217, + "GACAGT": 3218, + "GACAGC": 3219, + "GACAGG": 3220, + "GACTAA": 3221, + "GACTAT": 3222, + "GACTAC": 3223, + "GACTAG": 3224, + "GACTTA": 3225, + "GACTTT": 3226, + "GACTTC": 3227, + "GACTTG": 3228, + "GACTCA": 3229, + "GACTCT": 3230, + "GACTCC": 3231, + "GACTCG": 3232, + "GACTGA": 3233, + "GACTGT": 3234, + "GACTGC": 3235, + "GACTGG": 3236, + "GACCAA": 3237, + "GACCAT": 3238, + "GACCAC": 3239, + "GACCAG": 3240, + "GACCTA": 3241, + "GACCTT": 3242, + "GACCTC": 3243, + "GACCTG": 3244, + "GACCCA": 3245, + "GACCCT": 3246, + "GACCCC": 3247, + "GACCCG": 3248, + "GACCGA": 3249, + "GACCGT": 3250, + "GACCGC": 3251, + "GACCGG": 3252, + "GACGAA": 3253, + "GACGAT": 3254, + "GACGAC": 3255, + "GACGAG": 3256, + "GACGTA": 3257, + "GACGTT": 3258, + "GACGTC": 3259, + "GACGTG": 3260, + "GACGCA": 3261, + "GACGCT": 3262, + "GACGCC": 3263, + "GACGCG": 3264, + "GACGGA": 3265, + "GACGGT": 3266, + "GACGGC": 3267, + "GACGGG": 3268, + "GAGAAA": 3269, + "GAGAAT": 3270, + "GAGAAC": 3271, + "GAGAAG": 3272, + "GAGATA": 3273, + "GAGATT": 3274, + "GAGATC": 3275, + "GAGATG": 3276, + "GAGACA": 3277, + "GAGACT": 3278, + "GAGACC": 3279, + "GAGACG": 3280, + "GAGAGA": 3281, + "GAGAGT": 3282, + "GAGAGC": 3283, + "GAGAGG": 3284, + "GAGTAA": 3285, + "GAGTAT": 3286, + "GAGTAC": 3287, + "GAGTAG": 3288, + "GAGTTA": 3289, + "GAGTTT": 3290, + "GAGTTC": 3291, + "GAGTTG": 3292, + "GAGTCA": 3293, + "GAGTCT": 3294, + "GAGTCC": 3295, + "GAGTCG": 3296, + "GAGTGA": 3297, + "GAGTGT": 3298, + "GAGTGC": 3299, + "GAGTGG": 3300, + "GAGCAA": 3301, + "GAGCAT": 3302, + "GAGCAC": 3303, + "GAGCAG": 3304, + "GAGCTA": 3305, + "GAGCTT": 3306, + "GAGCTC": 3307, + "GAGCTG": 3308, + "GAGCCA": 3309, + "GAGCCT": 3310, + "GAGCCC": 3311, + "GAGCCG": 3312, + "GAGCGA": 3313, + "GAGCGT": 3314, + "GAGCGC": 3315, + "GAGCGG": 3316, + "GAGGAA": 3317, + "GAGGAT": 3318, + "GAGGAC": 3319, + "GAGGAG": 3320, + "GAGGTA": 3321, + "GAGGTT": 3322, + "GAGGTC": 3323, + "GAGGTG": 3324, + "GAGGCA": 3325, + "GAGGCT": 3326, + "GAGGCC": 3327, + "GAGGCG": 3328, + "GAGGGA": 3329, + "GAGGGT": 3330, + "GAGGGC": 3331, + "GAGGGG": 3332, + "GTAAAA": 3333, + "GTAAAT": 3334, + "GTAAAC": 3335, + "GTAAAG": 3336, + "GTAATA": 3337, + "GTAATT": 3338, + "GTAATC": 3339, + "GTAATG": 3340, + "GTAACA": 3341, + "GTAACT": 3342, + "GTAACC": 3343, + "GTAACG": 3344, + "GTAAGA": 3345, + "GTAAGT": 3346, + "GTAAGC": 3347, + "GTAAGG": 3348, + "GTATAA": 3349, + "GTATAT": 3350, + "GTATAC": 3351, + "GTATAG": 3352, + "GTATTA": 3353, + "GTATTT": 3354, + "GTATTC": 3355, + "GTATTG": 3356, + "GTATCA": 3357, + "GTATCT": 3358, + "GTATCC": 3359, + "GTATCG": 3360, + "GTATGA": 3361, + "GTATGT": 3362, + "GTATGC": 3363, + "GTATGG": 3364, + "GTACAA": 3365, + "GTACAT": 3366, + "GTACAC": 3367, + "GTACAG": 3368, + "GTACTA": 3369, + "GTACTT": 3370, + "GTACTC": 3371, + "GTACTG": 3372, + "GTACCA": 3373, + "GTACCT": 3374, + "GTACCC": 3375, + "GTACCG": 3376, + "GTACGA": 3377, + "GTACGT": 3378, + "GTACGC": 3379, + "GTACGG": 3380, + "GTAGAA": 3381, + "GTAGAT": 3382, + "GTAGAC": 3383, + "GTAGAG": 3384, + "GTAGTA": 3385, + "GTAGTT": 3386, + "GTAGTC": 3387, + "GTAGTG": 3388, + "GTAGCA": 3389, + "GTAGCT": 3390, + "GTAGCC": 3391, + "GTAGCG": 3392, + "GTAGGA": 3393, + "GTAGGT": 3394, + "GTAGGC": 3395, + "GTAGGG": 3396, + "GTTAAA": 3397, + "GTTAAT": 3398, + "GTTAAC": 3399, + "GTTAAG": 3400, + "GTTATA": 3401, + "GTTATT": 3402, + "GTTATC": 3403, + "GTTATG": 3404, + "GTTACA": 3405, + "GTTACT": 3406, + "GTTACC": 3407, + "GTTACG": 3408, + "GTTAGA": 3409, + "GTTAGT": 3410, + "GTTAGC": 3411, + "GTTAGG": 3412, + "GTTTAA": 3413, + "GTTTAT": 3414, + "GTTTAC": 3415, + "GTTTAG": 3416, + "GTTTTA": 3417, + "GTTTTT": 3418, + "GTTTTC": 3419, + "GTTTTG": 3420, + "GTTTCA": 3421, + "GTTTCT": 3422, + "GTTTCC": 3423, + "GTTTCG": 3424, + "GTTTGA": 3425, + "GTTTGT": 3426, + "GTTTGC": 3427, + "GTTTGG": 3428, + "GTTCAA": 3429, + "GTTCAT": 3430, + "GTTCAC": 3431, + "GTTCAG": 3432, + "GTTCTA": 3433, + "GTTCTT": 3434, + "GTTCTC": 3435, + "GTTCTG": 3436, + "GTTCCA": 3437, + "GTTCCT": 3438, + "GTTCCC": 3439, + "GTTCCG": 3440, + "GTTCGA": 3441, + "GTTCGT": 3442, + "GTTCGC": 3443, + "GTTCGG": 3444, + "GTTGAA": 3445, + "GTTGAT": 3446, + "GTTGAC": 3447, + "GTTGAG": 3448, + "GTTGTA": 3449, + "GTTGTT": 3450, + "GTTGTC": 3451, + "GTTGTG": 3452, + "GTTGCA": 3453, + "GTTGCT": 3454, + "GTTGCC": 3455, + "GTTGCG": 3456, + "GTTGGA": 3457, + "GTTGGT": 3458, + "GTTGGC": 3459, + "GTTGGG": 3460, + "GTCAAA": 3461, + "GTCAAT": 3462, + "GTCAAC": 3463, + "GTCAAG": 3464, + "GTCATA": 3465, + "GTCATT": 3466, + "GTCATC": 3467, + "GTCATG": 3468, + "GTCACA": 3469, + "GTCACT": 3470, + "GTCACC": 3471, + "GTCACG": 3472, + "GTCAGA": 3473, + "GTCAGT": 3474, + "GTCAGC": 3475, + "GTCAGG": 3476, + "GTCTAA": 3477, + "GTCTAT": 3478, + "GTCTAC": 3479, + "GTCTAG": 3480, + "GTCTTA": 3481, + "GTCTTT": 3482, + "GTCTTC": 3483, + "GTCTTG": 3484, + "GTCTCA": 3485, + "GTCTCT": 3486, + "GTCTCC": 3487, + "GTCTCG": 3488, + "GTCTGA": 3489, + "GTCTGT": 3490, + "GTCTGC": 3491, + "GTCTGG": 3492, + "GTCCAA": 3493, + "GTCCAT": 3494, + "GTCCAC": 3495, + "GTCCAG": 3496, + "GTCCTA": 3497, + "GTCCTT": 3498, + "GTCCTC": 3499, + "GTCCTG": 3500, + "GTCCCA": 3501, + "GTCCCT": 3502, + "GTCCCC": 3503, + "GTCCCG": 3504, + "GTCCGA": 3505, + "GTCCGT": 3506, + "GTCCGC": 3507, + "GTCCGG": 3508, + "GTCGAA": 3509, + "GTCGAT": 3510, + "GTCGAC": 3511, + "GTCGAG": 3512, + "GTCGTA": 3513, + "GTCGTT": 3514, + "GTCGTC": 3515, + "GTCGTG": 3516, + "GTCGCA": 3517, + "GTCGCT": 3518, + "GTCGCC": 3519, + "GTCGCG": 3520, + "GTCGGA": 3521, + "GTCGGT": 3522, + "GTCGGC": 3523, + "GTCGGG": 3524, + "GTGAAA": 3525, + "GTGAAT": 3526, + "GTGAAC": 3527, + "GTGAAG": 3528, + "GTGATA": 3529, + "GTGATT": 3530, + "GTGATC": 3531, + "GTGATG": 3532, + "GTGACA": 3533, + "GTGACT": 3534, + "GTGACC": 3535, + "GTGACG": 3536, + "GTGAGA": 3537, + "GTGAGT": 3538, + "GTGAGC": 3539, + "GTGAGG": 3540, + "GTGTAA": 3541, + "GTGTAT": 3542, + "GTGTAC": 3543, + "GTGTAG": 3544, + "GTGTTA": 3545, + "GTGTTT": 3546, + "GTGTTC": 3547, + "GTGTTG": 3548, + "GTGTCA": 3549, + "GTGTCT": 3550, + "GTGTCC": 3551, + "GTGTCG": 3552, + "GTGTGA": 3553, + "GTGTGT": 3554, + "GTGTGC": 3555, + "GTGTGG": 3556, + "GTGCAA": 3557, + "GTGCAT": 3558, + "GTGCAC": 3559, + "GTGCAG": 3560, + "GTGCTA": 3561, + "GTGCTT": 3562, + "GTGCTC": 3563, + "GTGCTG": 3564, + "GTGCCA": 3565, + "GTGCCT": 3566, + "GTGCCC": 3567, + "GTGCCG": 3568, + "GTGCGA": 3569, + "GTGCGT": 3570, + "GTGCGC": 3571, + "GTGCGG": 3572, + "GTGGAA": 3573, + "GTGGAT": 3574, + "GTGGAC": 3575, + "GTGGAG": 3576, + "GTGGTA": 3577, + "GTGGTT": 3578, + "GTGGTC": 3579, + "GTGGTG": 3580, + "GTGGCA": 3581, + "GTGGCT": 3582, + "GTGGCC": 3583, + "GTGGCG": 3584, + "GTGGGA": 3585, + "GTGGGT": 3586, + "GTGGGC": 3587, + "GTGGGG": 3588, + "GCAAAA": 3589, + "GCAAAT": 3590, + "GCAAAC": 3591, + "GCAAAG": 3592, + "GCAATA": 3593, + "GCAATT": 3594, + "GCAATC": 3595, + "GCAATG": 3596, + "GCAACA": 3597, + "GCAACT": 3598, + "GCAACC": 3599, + "GCAACG": 3600, + "GCAAGA": 3601, + "GCAAGT": 3602, + "GCAAGC": 3603, + "GCAAGG": 3604, + "GCATAA": 3605, + "GCATAT": 3606, + "GCATAC": 3607, + "GCATAG": 3608, + "GCATTA": 3609, + "GCATTT": 3610, + "GCATTC": 3611, + "GCATTG": 3612, + "GCATCA": 3613, + "GCATCT": 3614, + "GCATCC": 3615, + "GCATCG": 3616, + "GCATGA": 3617, + "GCATGT": 3618, + "GCATGC": 3619, + "GCATGG": 3620, + "GCACAA": 3621, + "GCACAT": 3622, + "GCACAC": 3623, + "GCACAG": 3624, + "GCACTA": 3625, + "GCACTT": 3626, + "GCACTC": 3627, + "GCACTG": 3628, + "GCACCA": 3629, + "GCACCT": 3630, + "GCACCC": 3631, + "GCACCG": 3632, + "GCACGA": 3633, + "GCACGT": 3634, + "GCACGC": 3635, + "GCACGG": 3636, + "GCAGAA": 3637, + "GCAGAT": 3638, + "GCAGAC": 3639, + "GCAGAG": 3640, + "GCAGTA": 3641, + "GCAGTT": 3642, + "GCAGTC": 3643, + "GCAGTG": 3644, + "GCAGCA": 3645, + "GCAGCT": 3646, + "GCAGCC": 3647, + "GCAGCG": 3648, + "GCAGGA": 3649, + "GCAGGT": 3650, + "GCAGGC": 3651, + "GCAGGG": 3652, + "GCTAAA": 3653, + "GCTAAT": 3654, + "GCTAAC": 3655, + "GCTAAG": 3656, + "GCTATA": 3657, + "GCTATT": 3658, + "GCTATC": 3659, + "GCTATG": 3660, + "GCTACA": 3661, + "GCTACT": 3662, + "GCTACC": 3663, + "GCTACG": 3664, + "GCTAGA": 3665, + "GCTAGT": 3666, + "GCTAGC": 3667, + "GCTAGG": 3668, + "GCTTAA": 3669, + "GCTTAT": 3670, + "GCTTAC": 3671, + "GCTTAG": 3672, + "GCTTTA": 3673, + "GCTTTT": 3674, + "GCTTTC": 3675, + "GCTTTG": 3676, + "GCTTCA": 3677, + "GCTTCT": 3678, + "GCTTCC": 3679, + "GCTTCG": 3680, + "GCTTGA": 3681, + "GCTTGT": 3682, + "GCTTGC": 3683, + "GCTTGG": 3684, + "GCTCAA": 3685, + "GCTCAT": 3686, + "GCTCAC": 3687, + "GCTCAG": 3688, + "GCTCTA": 3689, + "GCTCTT": 3690, + "GCTCTC": 3691, + "GCTCTG": 3692, + "GCTCCA": 3693, + "GCTCCT": 3694, + "GCTCCC": 3695, + "GCTCCG": 3696, + "GCTCGA": 3697, + "GCTCGT": 3698, + "GCTCGC": 3699, + "GCTCGG": 3700, + "GCTGAA": 3701, + "GCTGAT": 3702, + "GCTGAC": 3703, + "GCTGAG": 3704, + "GCTGTA": 3705, + "GCTGTT": 3706, + "GCTGTC": 3707, + "GCTGTG": 3708, + "GCTGCA": 3709, + "GCTGCT": 3710, + "GCTGCC": 3711, + "GCTGCG": 3712, + "GCTGGA": 3713, + "GCTGGT": 3714, + "GCTGGC": 3715, + "GCTGGG": 3716, + "GCCAAA": 3717, + "GCCAAT": 3718, + "GCCAAC": 3719, + "GCCAAG": 3720, + "GCCATA": 3721, + "GCCATT": 3722, + "GCCATC": 3723, + "GCCATG": 3724, + "GCCACA": 3725, + "GCCACT": 3726, + "GCCACC": 3727, + "GCCACG": 3728, + "GCCAGA": 3729, + "GCCAGT": 3730, + "GCCAGC": 3731, + "GCCAGG": 3732, + "GCCTAA": 3733, + "GCCTAT": 3734, + "GCCTAC": 3735, + "GCCTAG": 3736, + "GCCTTA": 3737, + "GCCTTT": 3738, + "GCCTTC": 3739, + "GCCTTG": 3740, + "GCCTCA": 3741, + "GCCTCT": 3742, + "GCCTCC": 3743, + "GCCTCG": 3744, + "GCCTGA": 3745, + "GCCTGT": 3746, + "GCCTGC": 3747, + "GCCTGG": 3748, + "GCCCAA": 3749, + "GCCCAT": 3750, + "GCCCAC": 3751, + "GCCCAG": 3752, + "GCCCTA": 3753, + "GCCCTT": 3754, + "GCCCTC": 3755, + "GCCCTG": 3756, + "GCCCCA": 3757, + "GCCCCT": 3758, + "GCCCCC": 3759, + "GCCCCG": 3760, + "GCCCGA": 3761, + "GCCCGT": 3762, + "GCCCGC": 3763, + "GCCCGG": 3764, + "GCCGAA": 3765, + "GCCGAT": 3766, + "GCCGAC": 3767, + "GCCGAG": 3768, + "GCCGTA": 3769, + "GCCGTT": 3770, + "GCCGTC": 3771, + "GCCGTG": 3772, + "GCCGCA": 3773, + "GCCGCT": 3774, + "GCCGCC": 3775, + "GCCGCG": 3776, + "GCCGGA": 3777, + "GCCGGT": 3778, + "GCCGGC": 3779, + "GCCGGG": 3780, + "GCGAAA": 3781, + "GCGAAT": 3782, + "GCGAAC": 3783, + "GCGAAG": 3784, + "GCGATA": 3785, + "GCGATT": 3786, + "GCGATC": 3787, + "GCGATG": 3788, + "GCGACA": 3789, + "GCGACT": 3790, + "GCGACC": 3791, + "GCGACG": 3792, + "GCGAGA": 3793, + "GCGAGT": 3794, + "GCGAGC": 3795, + "GCGAGG": 3796, + "GCGTAA": 3797, + "GCGTAT": 3798, + "GCGTAC": 3799, + "GCGTAG": 3800, + "GCGTTA": 3801, + "GCGTTT": 3802, + "GCGTTC": 3803, + "GCGTTG": 3804, + "GCGTCA": 3805, + "GCGTCT": 3806, + "GCGTCC": 3807, + "GCGTCG": 3808, + "GCGTGA": 3809, + "GCGTGT": 3810, + "GCGTGC": 3811, + "GCGTGG": 3812, + "GCGCAA": 3813, + "GCGCAT": 3814, + "GCGCAC": 3815, + "GCGCAG": 3816, + "GCGCTA": 3817, + "GCGCTT": 3818, + "GCGCTC": 3819, + "GCGCTG": 3820, + "GCGCCA": 3821, + "GCGCCT": 3822, + "GCGCCC": 3823, + "GCGCCG": 3824, + "GCGCGA": 3825, + "GCGCGT": 3826, + "GCGCGC": 3827, + "GCGCGG": 3828, + "GCGGAA": 3829, + "GCGGAT": 3830, + "GCGGAC": 3831, + "GCGGAG": 3832, + "GCGGTA": 3833, + "GCGGTT": 3834, + "GCGGTC": 3835, + "GCGGTG": 3836, + "GCGGCA": 3837, + "GCGGCT": 3838, + "GCGGCC": 3839, + "GCGGCG": 3840, + "GCGGGA": 3841, + "GCGGGT": 3842, + "GCGGGC": 3843, + "GCGGGG": 3844, + "GGAAAA": 3845, + "GGAAAT": 3846, + "GGAAAC": 3847, + "GGAAAG": 3848, + "GGAATA": 3849, + "GGAATT": 3850, + "GGAATC": 3851, + "GGAATG": 3852, + "GGAACA": 3853, + "GGAACT": 3854, + "GGAACC": 3855, + "GGAACG": 3856, + "GGAAGA": 3857, + "GGAAGT": 3858, + "GGAAGC": 3859, + "GGAAGG": 3860, + "GGATAA": 3861, + "GGATAT": 3862, + "GGATAC": 3863, + "GGATAG": 3864, + "GGATTA": 3865, + "GGATTT": 3866, + "GGATTC": 3867, + "GGATTG": 3868, + "GGATCA": 3869, + "GGATCT": 3870, + "GGATCC": 3871, + "GGATCG": 3872, + "GGATGA": 3873, + "GGATGT": 3874, + "GGATGC": 3875, + "GGATGG": 3876, + "GGACAA": 3877, + "GGACAT": 3878, + "GGACAC": 3879, + "GGACAG": 3880, + "GGACTA": 3881, + "GGACTT": 3882, + "GGACTC": 3883, + "GGACTG": 3884, + "GGACCA": 3885, + "GGACCT": 3886, + "GGACCC": 3887, + "GGACCG": 3888, + "GGACGA": 3889, + "GGACGT": 3890, + "GGACGC": 3891, + "GGACGG": 3892, + "GGAGAA": 3893, + "GGAGAT": 3894, + "GGAGAC": 3895, + "GGAGAG": 3896, + "GGAGTA": 3897, + "GGAGTT": 3898, + "GGAGTC": 3899, + "GGAGTG": 3900, + "GGAGCA": 3901, + "GGAGCT": 3902, + "GGAGCC": 3903, + "GGAGCG": 3904, + "GGAGGA": 3905, + "GGAGGT": 3906, + "GGAGGC": 3907, + "GGAGGG": 3908, + "GGTAAA": 3909, + "GGTAAT": 3910, + "GGTAAC": 3911, + "GGTAAG": 3912, + "GGTATA": 3913, + "GGTATT": 3914, + "GGTATC": 3915, + "GGTATG": 3916, + "GGTACA": 3917, + "GGTACT": 3918, + "GGTACC": 3919, + "GGTACG": 3920, + "GGTAGA": 3921, + "GGTAGT": 3922, + "GGTAGC": 3923, + "GGTAGG": 3924, + "GGTTAA": 3925, + "GGTTAT": 3926, + "GGTTAC": 3927, + "GGTTAG": 3928, + "GGTTTA": 3929, + "GGTTTT": 3930, + "GGTTTC": 3931, + "GGTTTG": 3932, + "GGTTCA": 3933, + "GGTTCT": 3934, + "GGTTCC": 3935, + "GGTTCG": 3936, + "GGTTGA": 3937, + "GGTTGT": 3938, + "GGTTGC": 3939, + "GGTTGG": 3940, + "GGTCAA": 3941, + "GGTCAT": 3942, + "GGTCAC": 3943, + "GGTCAG": 3944, + "GGTCTA": 3945, + "GGTCTT": 3946, + "GGTCTC": 3947, + "GGTCTG": 3948, + "GGTCCA": 3949, + "GGTCCT": 3950, + "GGTCCC": 3951, + "GGTCCG": 3952, + "GGTCGA": 3953, + "GGTCGT": 3954, + "GGTCGC": 3955, + "GGTCGG": 3956, + "GGTGAA": 3957, + "GGTGAT": 3958, + "GGTGAC": 3959, + "GGTGAG": 3960, + "GGTGTA": 3961, + "GGTGTT": 3962, + "GGTGTC": 3963, + "GGTGTG": 3964, + "GGTGCA": 3965, + "GGTGCT": 3966, + "GGTGCC": 3967, + "GGTGCG": 3968, + "GGTGGA": 3969, + "GGTGGT": 3970, + "GGTGGC": 3971, + "GGTGGG": 3972, + "GGCAAA": 3973, + "GGCAAT": 3974, + "GGCAAC": 3975, + "GGCAAG": 3976, + "GGCATA": 3977, + "GGCATT": 3978, + "GGCATC": 3979, + "GGCATG": 3980, + "GGCACA": 3981, + "GGCACT": 3982, + "GGCACC": 3983, + "GGCACG": 3984, + "GGCAGA": 3985, + "GGCAGT": 3986, + "GGCAGC": 3987, + "GGCAGG": 3988, + "GGCTAA": 3989, + "GGCTAT": 3990, + "GGCTAC": 3991, + "GGCTAG": 3992, + "GGCTTA": 3993, + "GGCTTT": 3994, + "GGCTTC": 3995, + "GGCTTG": 3996, + "GGCTCA": 3997, + "GGCTCT": 3998, + "GGCTCC": 3999, + "GGCTCG": 4000, + "GGCTGA": 4001, + "GGCTGT": 4002, + "GGCTGC": 4003, + "GGCTGG": 4004, + "GGCCAA": 4005, + "GGCCAT": 4006, + "GGCCAC": 4007, + "GGCCAG": 4008, + "GGCCTA": 4009, + "GGCCTT": 4010, + "GGCCTC": 4011, + "GGCCTG": 4012, + "GGCCCA": 4013, + "GGCCCT": 4014, + "GGCCCC": 4015, + "GGCCCG": 4016, + "GGCCGA": 4017, + "GGCCGT": 4018, + "GGCCGC": 4019, + "GGCCGG": 4020, + "GGCGAA": 4021, + "GGCGAT": 4022, + "GGCGAC": 4023, + "GGCGAG": 4024, + "GGCGTA": 4025, + "GGCGTT": 4026, + "GGCGTC": 4027, + "GGCGTG": 4028, + "GGCGCA": 4029, + "GGCGCT": 4030, + "GGCGCC": 4031, + "GGCGCG": 4032, + "GGCGGA": 4033, + "GGCGGT": 4034, + "GGCGGC": 4035, + "GGCGGG": 4036, + "GGGAAA": 4037, + "GGGAAT": 4038, + "GGGAAC": 4039, + "GGGAAG": 4040, + "GGGATA": 4041, + "GGGATT": 4042, + "GGGATC": 4043, + "GGGATG": 4044, + "GGGACA": 4045, + "GGGACT": 4046, + "GGGACC": 4047, + "GGGACG": 4048, + "GGGAGA": 4049, + "GGGAGT": 4050, + "GGGAGC": 4051, + "GGGAGG": 4052, + "GGGTAA": 4053, + "GGGTAT": 4054, + "GGGTAC": 4055, + "GGGTAG": 4056, + "GGGTTA": 4057, + "GGGTTT": 4058, + "GGGTTC": 4059, + "GGGTTG": 4060, + "GGGTCA": 4061, + "GGGTCT": 4062, + "GGGTCC": 4063, + "GGGTCG": 4064, + "GGGTGA": 4065, + "GGGTGT": 4066, + "GGGTGC": 4067, + "GGGTGG": 4068, + "GGGCAA": 4069, + "GGGCAT": 4070, + "GGGCAC": 4071, + "GGGCAG": 4072, + "GGGCTA": 4073, + "GGGCTT": 4074, + "GGGCTC": 4075, + "GGGCTG": 4076, + "GGGCCA": 4077, + "GGGCCT": 4078, + "GGGCCC": 4079, + "GGGCCG": 4080, + "GGGCGA": 4081, + "GGGCGT": 4082, + "GGGCGC": 4083, + "GGGCGG": 4084, + "GGGGAA": 4085, + "GGGGAT": 4086, + "GGGGAC": 4087, + "GGGGAG": 4088, + "GGGGTA": 4089, + "GGGGTT": 4090, + "GGGGTC": 4091, + "GGGGTG": 4092, + "GGGGCA": 4093, + "GGGGCT": 4094, + "GGGGCC": 4095, + "GGGGCG": 4096, + "GGGGGA": 4097, + "GGGGGT": 4098, + "GGGGGC": 4099, + "GGGGGG": 4100 + } + } +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-2000/tokenizer_config.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-2000/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..604e2b1b29af1dcb3920d2828b6a90276e5e3d02 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-2000/tokenizer_config.json @@ -0,0 +1,58 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "[UNK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "[CLS]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "[SEP]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "4": { + "content": "[MASK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "clean_up_tokenization_spaces": true, + "cls_token": "[CLS]", + "do_basic_tokenize": true, + "do_lower_case": false, + "mask_token": "[MASK]", + "max_len": 512, + "model_max_length": 512, + "never_split": null, + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "strip_accents": null, + "tokenize_chinese_chars": true, + "tokenizer_class": "BertTokenizer", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-2000/trainer_state.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-2000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..316553f708196d5f7771f0462f988eb2f85f4600 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-2000/trainer_state.json @@ -0,0 +1,593 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 8.0, + "eval_steps": 500, + "global_step": 2000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.1, + "grad_norm": 0.6038870215415955, + "learning_rate": 5.319148936170213e-05, + "loss": 0.6945, + "step": 25 + }, + { + "epoch": 0.2, + "grad_norm": 0.6668091416358948, + "learning_rate": 0.00010638297872340425, + "loss": 0.6929, + "step": 50 + }, + { + "epoch": 0.3, + "grad_norm": 0.8282511234283447, + "learning_rate": 0.00015957446808510637, + "loss": 0.6925, + "step": 75 + }, + { + "epoch": 0.4, + "grad_norm": 1.3149235248565674, + "learning_rate": 0.0002127659574468085, + "loss": 0.6919, + "step": 100 + }, + { + "epoch": 0.5, + "grad_norm": 1.8841172456741333, + "learning_rate": 0.00026595744680851064, + "loss": 0.6788, + "step": 125 + }, + { + "epoch": 0.6, + "grad_norm": 4.435206413269043, + "learning_rate": 0.00031914893617021275, + "loss": 0.6285, + "step": 150 + }, + { + "epoch": 0.7, + "grad_norm": 8.858148574829102, + "learning_rate": 0.0003723404255319149, + "loss": 0.5386, + "step": 175 + }, + { + "epoch": 0.8, + "grad_norm": 5.627648830413818, + "learning_rate": 0.0003992081821181128, + "loss": 0.5537, + "step": 200 + }, + { + "epoch": 0.9, + "grad_norm": 5.086580753326416, + "learning_rate": 0.0003975585615308479, + "loss": 0.5595, + "step": 225 + }, + { + "epoch": 1.0, + "grad_norm": 5.627719402313232, + "learning_rate": 0.00039590894094358297, + "loss": 0.4824, + "step": 250 + }, + { + "epoch": 1.1, + "grad_norm": 5.888038158416748, + "learning_rate": 0.0003942593203563181, + "loss": 0.4404, + "step": 275 + }, + { + "epoch": 1.2, + "grad_norm": 5.646231651306152, + "learning_rate": 0.0003926096997690532, + "loss": 0.4354, + "step": 300 + }, + { + "epoch": 1.3, + "grad_norm": 5.191990852355957, + "learning_rate": 0.00039096007918178817, + "loss": 0.4469, + "step": 325 + }, + { + "epoch": 1.4, + "grad_norm": 13.90101432800293, + "learning_rate": 0.0003893104585945233, + "loss": 0.3858, + "step": 350 + }, + { + "epoch": 1.5, + "grad_norm": 17.75942039489746, + "learning_rate": 0.0003876608380072583, + "loss": 0.4131, + "step": 375 + }, + { + "epoch": 1.6, + "grad_norm": 7.398252487182617, + "learning_rate": 0.00038601121741999343, + "loss": 0.3842, + "step": 400 + }, + { + "epoch": 1.7, + "grad_norm": 11.010066986083984, + "learning_rate": 0.0003843615968327285, + "loss": 0.3959, + "step": 425 + }, + { + "epoch": 1.8, + "grad_norm": 9.277341842651367, + "learning_rate": 0.0003827119762454636, + "loss": 0.3552, + "step": 450 + }, + { + "epoch": 1.9, + "grad_norm": 26.011449813842773, + "learning_rate": 0.00038106235565819863, + "loss": 0.3961, + "step": 475 + }, + { + "epoch": 2.0, + "grad_norm": 11.123054504394531, + "learning_rate": 0.0003794127350709337, + "loss": 0.3548, + "step": 500 + }, + { + "epoch": 2.1, + "grad_norm": 20.128774642944336, + "learning_rate": 0.0003777631144836688, + "loss": 0.302, + "step": 525 + }, + { + "epoch": 2.2, + "grad_norm": 4.64390230178833, + "learning_rate": 0.00037611349389640383, + "loss": 0.2853, + "step": 550 + }, + { + "epoch": 2.3, + "grad_norm": 4.199934482574463, + "learning_rate": 0.00037446387330913894, + "loss": 0.2724, + "step": 575 + }, + { + "epoch": 2.4, + "grad_norm": 6.909411430358887, + "learning_rate": 0.000372814252721874, + "loss": 0.2976, + "step": 600 + }, + { + "epoch": 2.5, + "grad_norm": 9.234820365905762, + "learning_rate": 0.00037116463213460903, + "loss": 0.2798, + "step": 625 + }, + { + "epoch": 2.6, + "grad_norm": 10.472943305969238, + "learning_rate": 0.00036951501154734414, + "loss": 0.3007, + "step": 650 + }, + { + "epoch": 2.7, + "grad_norm": 19.46732521057129, + "learning_rate": 0.0003678653909600792, + "loss": 0.2701, + "step": 675 + }, + { + "epoch": 2.8, + "grad_norm": 7.501255512237549, + "learning_rate": 0.0003662157703728143, + "loss": 0.2896, + "step": 700 + }, + { + "epoch": 2.9, + "grad_norm": 10.645585060119629, + "learning_rate": 0.00036456614978554934, + "loss": 0.2748, + "step": 725 + }, + { + "epoch": 3.0, + "grad_norm": 30.26616859436035, + "learning_rate": 0.00036291652919828444, + "loss": 0.268, + "step": 750 + }, + { + "epoch": 3.1, + "grad_norm": 14.13516616821289, + "learning_rate": 0.00036126690861101944, + "loss": 0.2159, + "step": 775 + }, + { + "epoch": 3.2, + "grad_norm": 10.091023445129395, + "learning_rate": 0.00035961728802375454, + "loss": 0.2071, + "step": 800 + }, + { + "epoch": 3.3, + "grad_norm": 22.201452255249023, + "learning_rate": 0.0003579676674364896, + "loss": 0.2146, + "step": 825 + }, + { + "epoch": 3.4, + "grad_norm": 10.044412612915039, + "learning_rate": 0.0003563180468492247, + "loss": 0.1827, + "step": 850 + }, + { + "epoch": 3.5, + "grad_norm": 21.04941177368164, + "learning_rate": 0.0003546684262619598, + "loss": 0.202, + "step": 875 + }, + { + "epoch": 3.6, + "grad_norm": 11.754891395568848, + "learning_rate": 0.00035301880567469485, + "loss": 0.2021, + "step": 900 + }, + { + "epoch": 3.7, + "grad_norm": 9.113747596740723, + "learning_rate": 0.0003513691850874299, + "loss": 0.2067, + "step": 925 + }, + { + "epoch": 3.8, + "grad_norm": 9.075196266174316, + "learning_rate": 0.00034971956450016495, + "loss": 0.3285, + "step": 950 + }, + { + "epoch": 3.9, + "grad_norm": 8.568796157836914, + "learning_rate": 0.00034806994391290005, + "loss": 0.2254, + "step": 975 + }, + { + "epoch": 4.0, + "grad_norm": 3.9587976932525635, + "learning_rate": 0.0003464203233256351, + "loss": 0.2245, + "step": 1000 + }, + { + "epoch": 4.1, + "grad_norm": 11.143651008605957, + "learning_rate": 0.0003447707027383702, + "loss": 0.1755, + "step": 1025 + }, + { + "epoch": 4.2, + "grad_norm": 12.53118896484375, + "learning_rate": 0.00034312108215110525, + "loss": 0.1808, + "step": 1050 + }, + { + "epoch": 4.3, + "grad_norm": 9.880820274353027, + "learning_rate": 0.0003414714615638403, + "loss": 0.1387, + "step": 1075 + }, + { + "epoch": 4.4, + "grad_norm": 5.729593753814697, + "learning_rate": 0.0003398218409765754, + "loss": 0.1958, + "step": 1100 + }, + { + "epoch": 4.5, + "grad_norm": 8.633047103881836, + "learning_rate": 0.00033817222038931045, + "loss": 0.1515, + "step": 1125 + }, + { + "epoch": 4.6, + "grad_norm": 19.237892150878906, + "learning_rate": 0.00033652259980204556, + "loss": 0.1452, + "step": 1150 + }, + { + "epoch": 4.7, + "grad_norm": 6.129818439483643, + "learning_rate": 0.0003348729792147806, + "loss": 0.1617, + "step": 1175 + }, + { + "epoch": 4.8, + "grad_norm": 5.073274612426758, + "learning_rate": 0.0003332233586275157, + "loss": 0.1967, + "step": 1200 + }, + { + "epoch": 4.9, + "grad_norm": 5.729922771453857, + "learning_rate": 0.00033157373804025076, + "loss": 0.1482, + "step": 1225 + }, + { + "epoch": 5.0, + "grad_norm": 12.314518928527832, + "learning_rate": 0.0003299241174529858, + "loss": 0.1588, + "step": 1250 + }, + { + "epoch": 5.1, + "grad_norm": 3.2636606693267822, + "learning_rate": 0.0003282744968657209, + "loss": 0.1588, + "step": 1275 + }, + { + "epoch": 5.2, + "grad_norm": 5.914486408233643, + "learning_rate": 0.00032662487627845596, + "loss": 0.1288, + "step": 1300 + }, + { + "epoch": 5.3, + "grad_norm": 13.849296569824219, + "learning_rate": 0.00032497525569119106, + "loss": 0.0961, + "step": 1325 + }, + { + "epoch": 5.4, + "grad_norm": 7.7317328453063965, + "learning_rate": 0.0003233256351039261, + "loss": 0.1148, + "step": 1350 + }, + { + "epoch": 5.5, + "grad_norm": 1.768608808517456, + "learning_rate": 0.00032167601451666116, + "loss": 0.121, + "step": 1375 + }, + { + "epoch": 5.6, + "grad_norm": 10.802785873413086, + "learning_rate": 0.00032002639392939627, + "loss": 0.1786, + "step": 1400 + }, + { + "epoch": 5.7, + "grad_norm": 4.158301830291748, + "learning_rate": 0.0003183767733421313, + "loss": 0.1078, + "step": 1425 + }, + { + "epoch": 5.8, + "grad_norm": 10.65804672241211, + "learning_rate": 0.0003167271527548664, + "loss": 0.1597, + "step": 1450 + }, + { + "epoch": 5.9, + "grad_norm": 10.880438804626465, + "learning_rate": 0.00031507753216760147, + "loss": 0.1225, + "step": 1475 + }, + { + "epoch": 6.0, + "grad_norm": 9.95138168334961, + "learning_rate": 0.00031342791158033657, + "loss": 0.1254, + "step": 1500 + }, + { + "epoch": 6.1, + "grad_norm": 1.5018162727355957, + "learning_rate": 0.00031177829099307157, + "loss": 0.0902, + "step": 1525 + }, + { + "epoch": 6.2, + "grad_norm": 14.281820297241211, + "learning_rate": 0.00031012867040580667, + "loss": 0.1026, + "step": 1550 + }, + { + "epoch": 6.3, + "grad_norm": 4.7540435791015625, + "learning_rate": 0.0003084790498185417, + "loss": 0.0876, + "step": 1575 + }, + { + "epoch": 6.4, + "grad_norm": 4.304183006286621, + "learning_rate": 0.0003068294292312768, + "loss": 0.1355, + "step": 1600 + }, + { + "epoch": 6.5, + "grad_norm": 14.653755187988281, + "learning_rate": 0.0003051798086440119, + "loss": 0.0878, + "step": 1625 + }, + { + "epoch": 6.6, + "grad_norm": 11.13028335571289, + "learning_rate": 0.000303530188056747, + "loss": 0.1203, + "step": 1650 + }, + { + "epoch": 6.7, + "grad_norm": 9.729938507080078, + "learning_rate": 0.000301880567469482, + "loss": 0.0775, + "step": 1675 + }, + { + "epoch": 6.8, + "grad_norm": 10.12938117980957, + "learning_rate": 0.0003002309468822171, + "loss": 0.1015, + "step": 1700 + }, + { + "epoch": 6.9, + "grad_norm": 8.506614685058594, + "learning_rate": 0.0002985813262949522, + "loss": 0.1362, + "step": 1725 + }, + { + "epoch": 7.0, + "grad_norm": 25.618762969970703, + "learning_rate": 0.00029693170570768723, + "loss": 0.1269, + "step": 1750 + }, + { + "epoch": 7.1, + "grad_norm": 27.155860900878906, + "learning_rate": 0.00029528208512042233, + "loss": 0.0555, + "step": 1775 + }, + { + "epoch": 7.2, + "grad_norm": 9.412527084350586, + "learning_rate": 0.0002936324645331574, + "loss": 0.0695, + "step": 1800 + }, + { + "epoch": 7.3, + "grad_norm": 6.4328293800354, + "learning_rate": 0.00029198284394589243, + "loss": 0.0808, + "step": 1825 + }, + { + "epoch": 7.4, + "grad_norm": 15.267293930053711, + "learning_rate": 0.00029033322335862753, + "loss": 0.0871, + "step": 1850 + }, + { + "epoch": 7.5, + "grad_norm": 6.774560928344727, + "learning_rate": 0.0002886836027713626, + "loss": 0.0819, + "step": 1875 + }, + { + "epoch": 7.6, + "grad_norm": 18.67180633544922, + "learning_rate": 0.0002870339821840977, + "loss": 0.0565, + "step": 1900 + }, + { + "epoch": 7.7, + "grad_norm": 16.39004898071289, + "learning_rate": 0.00028538436159683273, + "loss": 0.0907, + "step": 1925 + }, + { + "epoch": 7.8, + "grad_norm": 6.351583003997803, + "learning_rate": 0.00028373474100956784, + "loss": 0.1289, + "step": 1950 + }, + { + "epoch": 7.9, + "grad_norm": 11.396740913391113, + "learning_rate": 0.0002820851204223029, + "loss": 0.1149, + "step": 1975 + }, + { + "epoch": 8.0, + "grad_norm": 6.491570949554443, + "learning_rate": 0.00028043549983503794, + "loss": 0.0891, + "step": 2000 + } + ], + "logging_steps": 25, + "max_steps": 6250, + "num_input_tokens_seen": 0, + "num_train_epochs": 25, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 101638963200000.0, + "train_batch_size": 20, + "trial_name": null, + "trial_params": null +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-2000/training_args.bin b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-2000/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..77692b6666e83a688b38e799c964e4694a57eec9 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-2000/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f73b41315068a430d2439f5707c62152a34210df62454b9549f9642446dd90b2 +size 5304 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-2000/vocab.txt b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-2000/vocab.txt new file mode 100644 index 0000000000000000000000000000000000000000..cc9ad53732e21ea9180f18e1fb67a38cc5702d4c --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-2000/vocab.txt @@ -0,0 +1,4101 @@ +[PAD] +[UNK] +[CLS] +[SEP] +[MASK] +AAAAAA +AAAAAT +AAAAAC +AAAAAG +AAAATA +AAAATT +AAAATC +AAAATG +AAAACA +AAAACT +AAAACC +AAAACG +AAAAGA +AAAAGT +AAAAGC +AAAAGG +AAATAA +AAATAT +AAATAC +AAATAG +AAATTA +AAATTT +AAATTC +AAATTG +AAATCA +AAATCT +AAATCC +AAATCG +AAATGA +AAATGT +AAATGC +AAATGG +AAACAA +AAACAT +AAACAC +AAACAG +AAACTA +AAACTT +AAACTC +AAACTG +AAACCA +AAACCT +AAACCC +AAACCG +AAACGA +AAACGT +AAACGC +AAACGG +AAAGAA +AAAGAT +AAAGAC +AAAGAG +AAAGTA +AAAGTT +AAAGTC +AAAGTG +AAAGCA +AAAGCT +AAAGCC +AAAGCG +AAAGGA +AAAGGT +AAAGGC +AAAGGG +AATAAA +AATAAT +AATAAC +AATAAG +AATATA +AATATT +AATATC +AATATG +AATACA +AATACT +AATACC +AATACG +AATAGA +AATAGT +AATAGC +AATAGG +AATTAA +AATTAT +AATTAC +AATTAG +AATTTA +AATTTT +AATTTC +AATTTG +AATTCA +AATTCT +AATTCC +AATTCG +AATTGA +AATTGT +AATTGC +AATTGG +AATCAA +AATCAT +AATCAC +AATCAG +AATCTA +AATCTT +AATCTC +AATCTG +AATCCA +AATCCT +AATCCC +AATCCG +AATCGA +AATCGT +AATCGC +AATCGG +AATGAA +AATGAT +AATGAC +AATGAG +AATGTA +AATGTT +AATGTC +AATGTG +AATGCA +AATGCT +AATGCC +AATGCG +AATGGA +AATGGT +AATGGC +AATGGG +AACAAA +AACAAT +AACAAC +AACAAG +AACATA +AACATT +AACATC +AACATG +AACACA +AACACT +AACACC +AACACG +AACAGA +AACAGT +AACAGC +AACAGG +AACTAA +AACTAT +AACTAC +AACTAG +AACTTA +AACTTT +AACTTC +AACTTG +AACTCA +AACTCT +AACTCC +AACTCG +AACTGA +AACTGT +AACTGC +AACTGG +AACCAA +AACCAT +AACCAC +AACCAG +AACCTA +AACCTT +AACCTC +AACCTG +AACCCA +AACCCT +AACCCC +AACCCG +AACCGA +AACCGT +AACCGC +AACCGG +AACGAA +AACGAT +AACGAC +AACGAG +AACGTA +AACGTT +AACGTC +AACGTG +AACGCA +AACGCT +AACGCC +AACGCG +AACGGA +AACGGT +AACGGC +AACGGG +AAGAAA +AAGAAT +AAGAAC +AAGAAG +AAGATA +AAGATT +AAGATC +AAGATG +AAGACA +AAGACT +AAGACC +AAGACG +AAGAGA +AAGAGT +AAGAGC +AAGAGG +AAGTAA +AAGTAT +AAGTAC +AAGTAG +AAGTTA +AAGTTT +AAGTTC +AAGTTG +AAGTCA +AAGTCT +AAGTCC +AAGTCG +AAGTGA +AAGTGT +AAGTGC +AAGTGG +AAGCAA +AAGCAT +AAGCAC +AAGCAG +AAGCTA +AAGCTT +AAGCTC +AAGCTG +AAGCCA +AAGCCT +AAGCCC +AAGCCG +AAGCGA +AAGCGT +AAGCGC +AAGCGG +AAGGAA +AAGGAT +AAGGAC +AAGGAG +AAGGTA +AAGGTT +AAGGTC +AAGGTG +AAGGCA +AAGGCT +AAGGCC +AAGGCG +AAGGGA +AAGGGT +AAGGGC +AAGGGG +ATAAAA +ATAAAT +ATAAAC +ATAAAG +ATAATA +ATAATT +ATAATC +ATAATG +ATAACA +ATAACT +ATAACC +ATAACG +ATAAGA +ATAAGT +ATAAGC +ATAAGG +ATATAA +ATATAT +ATATAC +ATATAG +ATATTA +ATATTT +ATATTC +ATATTG +ATATCA +ATATCT +ATATCC +ATATCG +ATATGA +ATATGT +ATATGC +ATATGG +ATACAA +ATACAT +ATACAC +ATACAG +ATACTA +ATACTT +ATACTC +ATACTG +ATACCA +ATACCT +ATACCC +ATACCG +ATACGA +ATACGT +ATACGC +ATACGG +ATAGAA +ATAGAT +ATAGAC +ATAGAG +ATAGTA +ATAGTT +ATAGTC +ATAGTG +ATAGCA +ATAGCT +ATAGCC +ATAGCG +ATAGGA +ATAGGT +ATAGGC +ATAGGG +ATTAAA +ATTAAT +ATTAAC +ATTAAG +ATTATA +ATTATT +ATTATC +ATTATG +ATTACA +ATTACT +ATTACC +ATTACG +ATTAGA +ATTAGT +ATTAGC +ATTAGG +ATTTAA +ATTTAT +ATTTAC +ATTTAG +ATTTTA +ATTTTT +ATTTTC +ATTTTG +ATTTCA +ATTTCT +ATTTCC +ATTTCG +ATTTGA +ATTTGT +ATTTGC +ATTTGG +ATTCAA +ATTCAT +ATTCAC +ATTCAG +ATTCTA +ATTCTT +ATTCTC +ATTCTG +ATTCCA +ATTCCT +ATTCCC +ATTCCG +ATTCGA +ATTCGT +ATTCGC +ATTCGG +ATTGAA +ATTGAT +ATTGAC +ATTGAG +ATTGTA +ATTGTT +ATTGTC +ATTGTG +ATTGCA +ATTGCT +ATTGCC +ATTGCG +ATTGGA +ATTGGT +ATTGGC +ATTGGG +ATCAAA +ATCAAT +ATCAAC +ATCAAG +ATCATA +ATCATT +ATCATC +ATCATG +ATCACA +ATCACT +ATCACC +ATCACG +ATCAGA +ATCAGT +ATCAGC +ATCAGG +ATCTAA +ATCTAT +ATCTAC +ATCTAG +ATCTTA +ATCTTT +ATCTTC +ATCTTG +ATCTCA +ATCTCT +ATCTCC +ATCTCG +ATCTGA +ATCTGT +ATCTGC +ATCTGG +ATCCAA +ATCCAT +ATCCAC +ATCCAG +ATCCTA +ATCCTT +ATCCTC +ATCCTG +ATCCCA +ATCCCT +ATCCCC +ATCCCG +ATCCGA +ATCCGT +ATCCGC +ATCCGG +ATCGAA +ATCGAT +ATCGAC +ATCGAG +ATCGTA +ATCGTT +ATCGTC +ATCGTG +ATCGCA +ATCGCT +ATCGCC +ATCGCG +ATCGGA +ATCGGT +ATCGGC +ATCGGG +ATGAAA +ATGAAT +ATGAAC +ATGAAG +ATGATA +ATGATT +ATGATC +ATGATG +ATGACA +ATGACT +ATGACC +ATGACG +ATGAGA +ATGAGT +ATGAGC +ATGAGG +ATGTAA +ATGTAT +ATGTAC +ATGTAG +ATGTTA +ATGTTT +ATGTTC +ATGTTG +ATGTCA +ATGTCT +ATGTCC +ATGTCG +ATGTGA +ATGTGT +ATGTGC +ATGTGG +ATGCAA +ATGCAT +ATGCAC +ATGCAG +ATGCTA +ATGCTT +ATGCTC +ATGCTG +ATGCCA +ATGCCT +ATGCCC +ATGCCG +ATGCGA +ATGCGT +ATGCGC +ATGCGG +ATGGAA +ATGGAT +ATGGAC +ATGGAG +ATGGTA +ATGGTT +ATGGTC +ATGGTG +ATGGCA +ATGGCT +ATGGCC +ATGGCG +ATGGGA +ATGGGT +ATGGGC +ATGGGG +ACAAAA +ACAAAT +ACAAAC +ACAAAG +ACAATA +ACAATT +ACAATC +ACAATG +ACAACA +ACAACT +ACAACC +ACAACG +ACAAGA +ACAAGT +ACAAGC +ACAAGG +ACATAA +ACATAT +ACATAC +ACATAG +ACATTA +ACATTT +ACATTC +ACATTG +ACATCA +ACATCT +ACATCC +ACATCG +ACATGA +ACATGT +ACATGC +ACATGG +ACACAA +ACACAT +ACACAC +ACACAG +ACACTA +ACACTT +ACACTC +ACACTG +ACACCA +ACACCT +ACACCC +ACACCG +ACACGA +ACACGT +ACACGC +ACACGG +ACAGAA +ACAGAT +ACAGAC +ACAGAG +ACAGTA +ACAGTT +ACAGTC +ACAGTG +ACAGCA +ACAGCT +ACAGCC +ACAGCG +ACAGGA +ACAGGT +ACAGGC +ACAGGG +ACTAAA +ACTAAT +ACTAAC +ACTAAG +ACTATA +ACTATT +ACTATC +ACTATG +ACTACA +ACTACT +ACTACC +ACTACG +ACTAGA +ACTAGT +ACTAGC +ACTAGG +ACTTAA +ACTTAT +ACTTAC +ACTTAG +ACTTTA +ACTTTT +ACTTTC +ACTTTG +ACTTCA +ACTTCT +ACTTCC +ACTTCG +ACTTGA +ACTTGT +ACTTGC +ACTTGG +ACTCAA +ACTCAT +ACTCAC +ACTCAG +ACTCTA +ACTCTT +ACTCTC +ACTCTG +ACTCCA +ACTCCT +ACTCCC +ACTCCG +ACTCGA +ACTCGT +ACTCGC +ACTCGG +ACTGAA +ACTGAT +ACTGAC +ACTGAG +ACTGTA +ACTGTT +ACTGTC +ACTGTG +ACTGCA +ACTGCT +ACTGCC +ACTGCG +ACTGGA +ACTGGT +ACTGGC +ACTGGG +ACCAAA +ACCAAT +ACCAAC +ACCAAG +ACCATA +ACCATT +ACCATC +ACCATG +ACCACA +ACCACT +ACCACC +ACCACG +ACCAGA +ACCAGT +ACCAGC +ACCAGG +ACCTAA +ACCTAT +ACCTAC +ACCTAG +ACCTTA +ACCTTT +ACCTTC +ACCTTG +ACCTCA +ACCTCT +ACCTCC +ACCTCG +ACCTGA +ACCTGT +ACCTGC +ACCTGG +ACCCAA +ACCCAT +ACCCAC +ACCCAG +ACCCTA +ACCCTT +ACCCTC +ACCCTG +ACCCCA +ACCCCT +ACCCCC +ACCCCG +ACCCGA +ACCCGT +ACCCGC +ACCCGG +ACCGAA +ACCGAT +ACCGAC +ACCGAG +ACCGTA +ACCGTT +ACCGTC +ACCGTG +ACCGCA +ACCGCT +ACCGCC +ACCGCG +ACCGGA +ACCGGT +ACCGGC +ACCGGG +ACGAAA +ACGAAT +ACGAAC +ACGAAG +ACGATA +ACGATT +ACGATC +ACGATG +ACGACA +ACGACT +ACGACC +ACGACG +ACGAGA +ACGAGT +ACGAGC +ACGAGG +ACGTAA +ACGTAT +ACGTAC +ACGTAG +ACGTTA +ACGTTT +ACGTTC +ACGTTG +ACGTCA +ACGTCT +ACGTCC +ACGTCG +ACGTGA +ACGTGT +ACGTGC +ACGTGG +ACGCAA +ACGCAT +ACGCAC +ACGCAG +ACGCTA +ACGCTT +ACGCTC +ACGCTG +ACGCCA +ACGCCT +ACGCCC +ACGCCG +ACGCGA +ACGCGT +ACGCGC +ACGCGG +ACGGAA +ACGGAT +ACGGAC +ACGGAG +ACGGTA +ACGGTT +ACGGTC +ACGGTG +ACGGCA +ACGGCT +ACGGCC +ACGGCG +ACGGGA +ACGGGT +ACGGGC +ACGGGG +AGAAAA +AGAAAT +AGAAAC +AGAAAG +AGAATA +AGAATT +AGAATC +AGAATG +AGAACA +AGAACT +AGAACC +AGAACG +AGAAGA +AGAAGT +AGAAGC +AGAAGG +AGATAA +AGATAT +AGATAC +AGATAG +AGATTA +AGATTT +AGATTC +AGATTG +AGATCA +AGATCT +AGATCC +AGATCG +AGATGA +AGATGT +AGATGC +AGATGG +AGACAA +AGACAT +AGACAC +AGACAG +AGACTA +AGACTT +AGACTC +AGACTG +AGACCA +AGACCT +AGACCC +AGACCG +AGACGA +AGACGT +AGACGC +AGACGG +AGAGAA +AGAGAT +AGAGAC +AGAGAG +AGAGTA +AGAGTT +AGAGTC +AGAGTG +AGAGCA +AGAGCT +AGAGCC +AGAGCG +AGAGGA +AGAGGT +AGAGGC +AGAGGG +AGTAAA +AGTAAT +AGTAAC +AGTAAG +AGTATA +AGTATT +AGTATC +AGTATG +AGTACA +AGTACT +AGTACC +AGTACG +AGTAGA +AGTAGT +AGTAGC +AGTAGG +AGTTAA +AGTTAT +AGTTAC +AGTTAG +AGTTTA +AGTTTT +AGTTTC +AGTTTG +AGTTCA +AGTTCT +AGTTCC +AGTTCG +AGTTGA +AGTTGT +AGTTGC +AGTTGG +AGTCAA +AGTCAT +AGTCAC +AGTCAG +AGTCTA +AGTCTT +AGTCTC +AGTCTG +AGTCCA +AGTCCT +AGTCCC +AGTCCG +AGTCGA +AGTCGT +AGTCGC +AGTCGG +AGTGAA +AGTGAT +AGTGAC +AGTGAG +AGTGTA +AGTGTT +AGTGTC +AGTGTG +AGTGCA +AGTGCT +AGTGCC +AGTGCG +AGTGGA +AGTGGT +AGTGGC +AGTGGG +AGCAAA +AGCAAT +AGCAAC +AGCAAG +AGCATA +AGCATT +AGCATC +AGCATG +AGCACA +AGCACT +AGCACC +AGCACG +AGCAGA +AGCAGT +AGCAGC +AGCAGG +AGCTAA +AGCTAT +AGCTAC +AGCTAG +AGCTTA +AGCTTT +AGCTTC +AGCTTG +AGCTCA +AGCTCT +AGCTCC +AGCTCG +AGCTGA +AGCTGT +AGCTGC +AGCTGG +AGCCAA +AGCCAT +AGCCAC +AGCCAG +AGCCTA +AGCCTT +AGCCTC +AGCCTG +AGCCCA +AGCCCT +AGCCCC +AGCCCG +AGCCGA +AGCCGT +AGCCGC +AGCCGG +AGCGAA +AGCGAT +AGCGAC +AGCGAG +AGCGTA +AGCGTT +AGCGTC +AGCGTG +AGCGCA +AGCGCT +AGCGCC +AGCGCG +AGCGGA +AGCGGT +AGCGGC +AGCGGG +AGGAAA +AGGAAT +AGGAAC +AGGAAG +AGGATA +AGGATT +AGGATC +AGGATG +AGGACA +AGGACT +AGGACC +AGGACG +AGGAGA +AGGAGT +AGGAGC +AGGAGG +AGGTAA +AGGTAT +AGGTAC +AGGTAG +AGGTTA +AGGTTT +AGGTTC +AGGTTG +AGGTCA +AGGTCT +AGGTCC +AGGTCG +AGGTGA +AGGTGT +AGGTGC +AGGTGG +AGGCAA +AGGCAT +AGGCAC +AGGCAG +AGGCTA +AGGCTT +AGGCTC +AGGCTG +AGGCCA +AGGCCT +AGGCCC +AGGCCG +AGGCGA +AGGCGT +AGGCGC +AGGCGG +AGGGAA +AGGGAT +AGGGAC +AGGGAG +AGGGTA +AGGGTT +AGGGTC +AGGGTG +AGGGCA +AGGGCT +AGGGCC +AGGGCG +AGGGGA +AGGGGT +AGGGGC +AGGGGG +TAAAAA +TAAAAT +TAAAAC +TAAAAG +TAAATA +TAAATT +TAAATC +TAAATG +TAAACA +TAAACT +TAAACC +TAAACG +TAAAGA +TAAAGT +TAAAGC +TAAAGG +TAATAA +TAATAT +TAATAC +TAATAG +TAATTA +TAATTT +TAATTC +TAATTG +TAATCA +TAATCT +TAATCC +TAATCG +TAATGA +TAATGT +TAATGC +TAATGG +TAACAA +TAACAT +TAACAC +TAACAG +TAACTA +TAACTT +TAACTC +TAACTG +TAACCA +TAACCT +TAACCC +TAACCG +TAACGA +TAACGT +TAACGC +TAACGG +TAAGAA +TAAGAT +TAAGAC +TAAGAG +TAAGTA +TAAGTT +TAAGTC +TAAGTG +TAAGCA +TAAGCT +TAAGCC +TAAGCG +TAAGGA +TAAGGT +TAAGGC +TAAGGG +TATAAA +TATAAT +TATAAC +TATAAG +TATATA +TATATT +TATATC +TATATG +TATACA +TATACT +TATACC +TATACG +TATAGA +TATAGT +TATAGC +TATAGG +TATTAA +TATTAT +TATTAC +TATTAG +TATTTA +TATTTT +TATTTC +TATTTG +TATTCA +TATTCT +TATTCC +TATTCG +TATTGA +TATTGT +TATTGC +TATTGG +TATCAA +TATCAT +TATCAC +TATCAG +TATCTA +TATCTT +TATCTC +TATCTG +TATCCA +TATCCT +TATCCC +TATCCG +TATCGA +TATCGT +TATCGC +TATCGG +TATGAA +TATGAT +TATGAC +TATGAG +TATGTA +TATGTT +TATGTC +TATGTG +TATGCA +TATGCT +TATGCC +TATGCG +TATGGA +TATGGT +TATGGC +TATGGG +TACAAA +TACAAT +TACAAC +TACAAG +TACATA +TACATT +TACATC +TACATG +TACACA +TACACT +TACACC +TACACG +TACAGA +TACAGT +TACAGC +TACAGG +TACTAA +TACTAT +TACTAC +TACTAG +TACTTA +TACTTT +TACTTC +TACTTG +TACTCA +TACTCT +TACTCC +TACTCG +TACTGA +TACTGT +TACTGC +TACTGG +TACCAA +TACCAT +TACCAC +TACCAG +TACCTA +TACCTT +TACCTC +TACCTG +TACCCA +TACCCT +TACCCC +TACCCG +TACCGA +TACCGT +TACCGC +TACCGG +TACGAA +TACGAT +TACGAC +TACGAG +TACGTA +TACGTT +TACGTC +TACGTG +TACGCA +TACGCT +TACGCC +TACGCG +TACGGA +TACGGT +TACGGC +TACGGG +TAGAAA +TAGAAT +TAGAAC +TAGAAG +TAGATA +TAGATT +TAGATC +TAGATG +TAGACA +TAGACT +TAGACC +TAGACG +TAGAGA +TAGAGT +TAGAGC +TAGAGG +TAGTAA +TAGTAT +TAGTAC +TAGTAG +TAGTTA +TAGTTT +TAGTTC +TAGTTG +TAGTCA +TAGTCT +TAGTCC +TAGTCG +TAGTGA +TAGTGT +TAGTGC +TAGTGG +TAGCAA +TAGCAT +TAGCAC +TAGCAG +TAGCTA +TAGCTT +TAGCTC +TAGCTG +TAGCCA +TAGCCT +TAGCCC +TAGCCG +TAGCGA +TAGCGT +TAGCGC +TAGCGG +TAGGAA +TAGGAT +TAGGAC +TAGGAG +TAGGTA +TAGGTT +TAGGTC +TAGGTG +TAGGCA +TAGGCT +TAGGCC +TAGGCG +TAGGGA +TAGGGT +TAGGGC +TAGGGG +TTAAAA +TTAAAT +TTAAAC +TTAAAG +TTAATA +TTAATT +TTAATC +TTAATG +TTAACA +TTAACT +TTAACC +TTAACG +TTAAGA +TTAAGT +TTAAGC +TTAAGG +TTATAA +TTATAT +TTATAC +TTATAG +TTATTA +TTATTT +TTATTC +TTATTG +TTATCA +TTATCT +TTATCC +TTATCG +TTATGA +TTATGT +TTATGC +TTATGG +TTACAA +TTACAT +TTACAC +TTACAG +TTACTA +TTACTT +TTACTC +TTACTG +TTACCA +TTACCT +TTACCC +TTACCG +TTACGA +TTACGT +TTACGC +TTACGG +TTAGAA +TTAGAT +TTAGAC +TTAGAG +TTAGTA +TTAGTT +TTAGTC +TTAGTG +TTAGCA +TTAGCT +TTAGCC +TTAGCG +TTAGGA +TTAGGT +TTAGGC +TTAGGG +TTTAAA +TTTAAT +TTTAAC +TTTAAG +TTTATA +TTTATT +TTTATC +TTTATG +TTTACA +TTTACT +TTTACC +TTTACG +TTTAGA +TTTAGT +TTTAGC +TTTAGG +TTTTAA +TTTTAT +TTTTAC +TTTTAG +TTTTTA +TTTTTT +TTTTTC +TTTTTG +TTTTCA +TTTTCT +TTTTCC +TTTTCG +TTTTGA +TTTTGT +TTTTGC +TTTTGG +TTTCAA +TTTCAT +TTTCAC +TTTCAG +TTTCTA +TTTCTT +TTTCTC +TTTCTG +TTTCCA +TTTCCT +TTTCCC +TTTCCG +TTTCGA +TTTCGT +TTTCGC +TTTCGG +TTTGAA +TTTGAT +TTTGAC +TTTGAG +TTTGTA +TTTGTT +TTTGTC +TTTGTG +TTTGCA +TTTGCT +TTTGCC +TTTGCG +TTTGGA +TTTGGT +TTTGGC +TTTGGG +TTCAAA +TTCAAT +TTCAAC +TTCAAG +TTCATA +TTCATT +TTCATC +TTCATG +TTCACA +TTCACT +TTCACC +TTCACG +TTCAGA +TTCAGT +TTCAGC +TTCAGG +TTCTAA +TTCTAT +TTCTAC +TTCTAG +TTCTTA +TTCTTT +TTCTTC +TTCTTG +TTCTCA +TTCTCT +TTCTCC +TTCTCG +TTCTGA +TTCTGT +TTCTGC +TTCTGG +TTCCAA +TTCCAT +TTCCAC +TTCCAG +TTCCTA +TTCCTT +TTCCTC +TTCCTG +TTCCCA +TTCCCT +TTCCCC +TTCCCG +TTCCGA +TTCCGT +TTCCGC +TTCCGG +TTCGAA +TTCGAT +TTCGAC +TTCGAG +TTCGTA +TTCGTT +TTCGTC +TTCGTG +TTCGCA +TTCGCT +TTCGCC +TTCGCG +TTCGGA +TTCGGT +TTCGGC +TTCGGG +TTGAAA +TTGAAT +TTGAAC +TTGAAG +TTGATA +TTGATT +TTGATC +TTGATG +TTGACA +TTGACT +TTGACC +TTGACG +TTGAGA +TTGAGT +TTGAGC +TTGAGG +TTGTAA +TTGTAT +TTGTAC +TTGTAG +TTGTTA +TTGTTT +TTGTTC +TTGTTG +TTGTCA +TTGTCT +TTGTCC +TTGTCG +TTGTGA +TTGTGT +TTGTGC +TTGTGG +TTGCAA +TTGCAT +TTGCAC +TTGCAG +TTGCTA +TTGCTT +TTGCTC +TTGCTG +TTGCCA +TTGCCT +TTGCCC +TTGCCG +TTGCGA +TTGCGT +TTGCGC +TTGCGG +TTGGAA +TTGGAT +TTGGAC +TTGGAG +TTGGTA +TTGGTT +TTGGTC +TTGGTG +TTGGCA +TTGGCT +TTGGCC +TTGGCG +TTGGGA +TTGGGT +TTGGGC +TTGGGG +TCAAAA +TCAAAT +TCAAAC +TCAAAG +TCAATA +TCAATT +TCAATC +TCAATG +TCAACA +TCAACT +TCAACC +TCAACG +TCAAGA +TCAAGT +TCAAGC +TCAAGG +TCATAA +TCATAT +TCATAC +TCATAG +TCATTA +TCATTT +TCATTC +TCATTG +TCATCA +TCATCT +TCATCC +TCATCG +TCATGA +TCATGT +TCATGC +TCATGG +TCACAA +TCACAT +TCACAC +TCACAG +TCACTA +TCACTT +TCACTC +TCACTG +TCACCA +TCACCT +TCACCC +TCACCG +TCACGA +TCACGT +TCACGC +TCACGG +TCAGAA +TCAGAT +TCAGAC +TCAGAG +TCAGTA +TCAGTT +TCAGTC +TCAGTG +TCAGCA +TCAGCT +TCAGCC +TCAGCG +TCAGGA +TCAGGT +TCAGGC +TCAGGG +TCTAAA +TCTAAT +TCTAAC +TCTAAG +TCTATA +TCTATT +TCTATC +TCTATG +TCTACA +TCTACT +TCTACC +TCTACG +TCTAGA +TCTAGT +TCTAGC +TCTAGG +TCTTAA +TCTTAT +TCTTAC +TCTTAG +TCTTTA +TCTTTT +TCTTTC +TCTTTG +TCTTCA +TCTTCT +TCTTCC +TCTTCG +TCTTGA +TCTTGT +TCTTGC +TCTTGG +TCTCAA +TCTCAT +TCTCAC +TCTCAG +TCTCTA +TCTCTT +TCTCTC +TCTCTG +TCTCCA +TCTCCT +TCTCCC +TCTCCG +TCTCGA +TCTCGT +TCTCGC +TCTCGG +TCTGAA +TCTGAT +TCTGAC +TCTGAG +TCTGTA +TCTGTT +TCTGTC +TCTGTG +TCTGCA +TCTGCT +TCTGCC +TCTGCG +TCTGGA +TCTGGT +TCTGGC +TCTGGG +TCCAAA +TCCAAT +TCCAAC +TCCAAG +TCCATA +TCCATT +TCCATC +TCCATG +TCCACA +TCCACT +TCCACC +TCCACG +TCCAGA +TCCAGT +TCCAGC +TCCAGG +TCCTAA +TCCTAT +TCCTAC +TCCTAG +TCCTTA +TCCTTT +TCCTTC +TCCTTG +TCCTCA +TCCTCT +TCCTCC +TCCTCG +TCCTGA +TCCTGT +TCCTGC +TCCTGG +TCCCAA +TCCCAT +TCCCAC +TCCCAG +TCCCTA +TCCCTT +TCCCTC +TCCCTG +TCCCCA +TCCCCT +TCCCCC +TCCCCG +TCCCGA +TCCCGT +TCCCGC +TCCCGG +TCCGAA +TCCGAT +TCCGAC +TCCGAG +TCCGTA +TCCGTT +TCCGTC +TCCGTG +TCCGCA +TCCGCT +TCCGCC +TCCGCG +TCCGGA +TCCGGT +TCCGGC +TCCGGG +TCGAAA +TCGAAT +TCGAAC +TCGAAG +TCGATA +TCGATT +TCGATC +TCGATG +TCGACA +TCGACT +TCGACC +TCGACG +TCGAGA +TCGAGT +TCGAGC +TCGAGG +TCGTAA +TCGTAT +TCGTAC +TCGTAG +TCGTTA +TCGTTT +TCGTTC +TCGTTG +TCGTCA +TCGTCT +TCGTCC +TCGTCG +TCGTGA +TCGTGT +TCGTGC +TCGTGG +TCGCAA +TCGCAT +TCGCAC +TCGCAG +TCGCTA +TCGCTT +TCGCTC +TCGCTG +TCGCCA +TCGCCT +TCGCCC +TCGCCG +TCGCGA +TCGCGT +TCGCGC +TCGCGG +TCGGAA +TCGGAT +TCGGAC +TCGGAG +TCGGTA +TCGGTT +TCGGTC +TCGGTG +TCGGCA +TCGGCT +TCGGCC +TCGGCG +TCGGGA +TCGGGT +TCGGGC +TCGGGG +TGAAAA +TGAAAT +TGAAAC +TGAAAG +TGAATA +TGAATT +TGAATC +TGAATG +TGAACA +TGAACT +TGAACC +TGAACG +TGAAGA +TGAAGT +TGAAGC +TGAAGG +TGATAA +TGATAT +TGATAC +TGATAG +TGATTA +TGATTT +TGATTC +TGATTG +TGATCA +TGATCT +TGATCC +TGATCG +TGATGA +TGATGT +TGATGC +TGATGG +TGACAA +TGACAT +TGACAC +TGACAG +TGACTA +TGACTT +TGACTC +TGACTG +TGACCA +TGACCT +TGACCC +TGACCG +TGACGA +TGACGT +TGACGC +TGACGG +TGAGAA +TGAGAT +TGAGAC +TGAGAG +TGAGTA +TGAGTT +TGAGTC +TGAGTG +TGAGCA +TGAGCT +TGAGCC +TGAGCG +TGAGGA +TGAGGT +TGAGGC +TGAGGG +TGTAAA +TGTAAT +TGTAAC +TGTAAG +TGTATA +TGTATT +TGTATC +TGTATG +TGTACA +TGTACT +TGTACC +TGTACG +TGTAGA +TGTAGT +TGTAGC +TGTAGG +TGTTAA +TGTTAT +TGTTAC +TGTTAG +TGTTTA +TGTTTT +TGTTTC +TGTTTG +TGTTCA +TGTTCT +TGTTCC +TGTTCG +TGTTGA +TGTTGT +TGTTGC +TGTTGG +TGTCAA +TGTCAT +TGTCAC +TGTCAG +TGTCTA +TGTCTT +TGTCTC +TGTCTG +TGTCCA +TGTCCT +TGTCCC +TGTCCG +TGTCGA +TGTCGT +TGTCGC +TGTCGG +TGTGAA +TGTGAT +TGTGAC +TGTGAG +TGTGTA +TGTGTT +TGTGTC +TGTGTG +TGTGCA +TGTGCT +TGTGCC +TGTGCG +TGTGGA +TGTGGT +TGTGGC +TGTGGG +TGCAAA +TGCAAT +TGCAAC +TGCAAG +TGCATA +TGCATT +TGCATC +TGCATG +TGCACA +TGCACT +TGCACC +TGCACG +TGCAGA +TGCAGT +TGCAGC +TGCAGG +TGCTAA +TGCTAT +TGCTAC +TGCTAG +TGCTTA +TGCTTT +TGCTTC +TGCTTG +TGCTCA +TGCTCT +TGCTCC +TGCTCG +TGCTGA +TGCTGT +TGCTGC +TGCTGG +TGCCAA +TGCCAT +TGCCAC +TGCCAG +TGCCTA +TGCCTT +TGCCTC +TGCCTG +TGCCCA +TGCCCT +TGCCCC +TGCCCG +TGCCGA +TGCCGT +TGCCGC +TGCCGG +TGCGAA +TGCGAT +TGCGAC +TGCGAG +TGCGTA +TGCGTT +TGCGTC +TGCGTG +TGCGCA +TGCGCT +TGCGCC +TGCGCG +TGCGGA +TGCGGT +TGCGGC +TGCGGG +TGGAAA +TGGAAT +TGGAAC +TGGAAG +TGGATA +TGGATT +TGGATC +TGGATG +TGGACA +TGGACT +TGGACC +TGGACG +TGGAGA +TGGAGT +TGGAGC +TGGAGG +TGGTAA +TGGTAT +TGGTAC +TGGTAG +TGGTTA +TGGTTT +TGGTTC +TGGTTG +TGGTCA +TGGTCT +TGGTCC +TGGTCG +TGGTGA +TGGTGT +TGGTGC +TGGTGG +TGGCAA +TGGCAT +TGGCAC +TGGCAG +TGGCTA +TGGCTT +TGGCTC +TGGCTG +TGGCCA +TGGCCT +TGGCCC +TGGCCG +TGGCGA +TGGCGT +TGGCGC +TGGCGG +TGGGAA +TGGGAT +TGGGAC +TGGGAG +TGGGTA +TGGGTT +TGGGTC +TGGGTG +TGGGCA +TGGGCT +TGGGCC +TGGGCG +TGGGGA +TGGGGT +TGGGGC +TGGGGG +CAAAAA +CAAAAT +CAAAAC +CAAAAG +CAAATA +CAAATT +CAAATC +CAAATG +CAAACA +CAAACT +CAAACC +CAAACG +CAAAGA +CAAAGT +CAAAGC +CAAAGG +CAATAA +CAATAT +CAATAC +CAATAG +CAATTA +CAATTT +CAATTC +CAATTG +CAATCA +CAATCT +CAATCC +CAATCG +CAATGA +CAATGT +CAATGC +CAATGG +CAACAA +CAACAT +CAACAC +CAACAG +CAACTA +CAACTT +CAACTC +CAACTG +CAACCA +CAACCT +CAACCC +CAACCG +CAACGA +CAACGT +CAACGC +CAACGG +CAAGAA +CAAGAT +CAAGAC +CAAGAG +CAAGTA +CAAGTT +CAAGTC +CAAGTG +CAAGCA +CAAGCT +CAAGCC +CAAGCG +CAAGGA +CAAGGT +CAAGGC +CAAGGG +CATAAA +CATAAT +CATAAC +CATAAG +CATATA +CATATT +CATATC +CATATG +CATACA +CATACT +CATACC +CATACG +CATAGA +CATAGT +CATAGC +CATAGG +CATTAA +CATTAT +CATTAC +CATTAG +CATTTA +CATTTT +CATTTC +CATTTG +CATTCA +CATTCT +CATTCC +CATTCG +CATTGA +CATTGT +CATTGC +CATTGG +CATCAA +CATCAT +CATCAC +CATCAG +CATCTA +CATCTT +CATCTC +CATCTG +CATCCA +CATCCT +CATCCC +CATCCG +CATCGA +CATCGT +CATCGC +CATCGG +CATGAA +CATGAT +CATGAC +CATGAG +CATGTA +CATGTT +CATGTC +CATGTG +CATGCA +CATGCT +CATGCC +CATGCG +CATGGA +CATGGT +CATGGC +CATGGG +CACAAA +CACAAT +CACAAC +CACAAG +CACATA +CACATT +CACATC +CACATG +CACACA +CACACT +CACACC +CACACG +CACAGA +CACAGT +CACAGC +CACAGG +CACTAA +CACTAT +CACTAC +CACTAG +CACTTA +CACTTT +CACTTC +CACTTG +CACTCA +CACTCT +CACTCC +CACTCG +CACTGA +CACTGT +CACTGC +CACTGG +CACCAA +CACCAT +CACCAC +CACCAG +CACCTA +CACCTT +CACCTC +CACCTG +CACCCA +CACCCT +CACCCC +CACCCG +CACCGA +CACCGT +CACCGC +CACCGG +CACGAA +CACGAT +CACGAC +CACGAG +CACGTA +CACGTT +CACGTC +CACGTG +CACGCA +CACGCT +CACGCC +CACGCG +CACGGA +CACGGT +CACGGC +CACGGG +CAGAAA +CAGAAT +CAGAAC +CAGAAG +CAGATA +CAGATT +CAGATC +CAGATG +CAGACA +CAGACT +CAGACC +CAGACG +CAGAGA +CAGAGT +CAGAGC +CAGAGG +CAGTAA +CAGTAT +CAGTAC +CAGTAG +CAGTTA +CAGTTT +CAGTTC +CAGTTG +CAGTCA +CAGTCT +CAGTCC +CAGTCG +CAGTGA +CAGTGT +CAGTGC +CAGTGG +CAGCAA +CAGCAT +CAGCAC +CAGCAG +CAGCTA +CAGCTT +CAGCTC +CAGCTG +CAGCCA +CAGCCT +CAGCCC +CAGCCG +CAGCGA +CAGCGT +CAGCGC +CAGCGG +CAGGAA +CAGGAT +CAGGAC +CAGGAG +CAGGTA +CAGGTT +CAGGTC +CAGGTG +CAGGCA +CAGGCT +CAGGCC +CAGGCG +CAGGGA +CAGGGT +CAGGGC +CAGGGG +CTAAAA +CTAAAT +CTAAAC +CTAAAG +CTAATA +CTAATT +CTAATC +CTAATG +CTAACA +CTAACT +CTAACC +CTAACG +CTAAGA +CTAAGT +CTAAGC +CTAAGG +CTATAA +CTATAT +CTATAC +CTATAG +CTATTA +CTATTT +CTATTC +CTATTG +CTATCA +CTATCT +CTATCC +CTATCG +CTATGA +CTATGT +CTATGC +CTATGG +CTACAA +CTACAT +CTACAC +CTACAG +CTACTA +CTACTT +CTACTC +CTACTG +CTACCA +CTACCT +CTACCC +CTACCG +CTACGA +CTACGT +CTACGC +CTACGG +CTAGAA +CTAGAT +CTAGAC +CTAGAG +CTAGTA +CTAGTT +CTAGTC +CTAGTG +CTAGCA +CTAGCT +CTAGCC +CTAGCG +CTAGGA +CTAGGT +CTAGGC +CTAGGG +CTTAAA +CTTAAT +CTTAAC +CTTAAG +CTTATA +CTTATT +CTTATC +CTTATG +CTTACA +CTTACT +CTTACC +CTTACG +CTTAGA +CTTAGT +CTTAGC +CTTAGG +CTTTAA +CTTTAT +CTTTAC +CTTTAG +CTTTTA +CTTTTT +CTTTTC +CTTTTG +CTTTCA +CTTTCT +CTTTCC +CTTTCG +CTTTGA +CTTTGT +CTTTGC +CTTTGG +CTTCAA +CTTCAT +CTTCAC +CTTCAG +CTTCTA +CTTCTT +CTTCTC +CTTCTG +CTTCCA +CTTCCT +CTTCCC +CTTCCG +CTTCGA +CTTCGT +CTTCGC +CTTCGG +CTTGAA +CTTGAT +CTTGAC +CTTGAG +CTTGTA +CTTGTT +CTTGTC +CTTGTG +CTTGCA +CTTGCT +CTTGCC +CTTGCG +CTTGGA +CTTGGT +CTTGGC +CTTGGG +CTCAAA +CTCAAT +CTCAAC +CTCAAG +CTCATA +CTCATT +CTCATC +CTCATG +CTCACA +CTCACT +CTCACC +CTCACG +CTCAGA +CTCAGT +CTCAGC +CTCAGG +CTCTAA +CTCTAT +CTCTAC +CTCTAG +CTCTTA +CTCTTT +CTCTTC +CTCTTG +CTCTCA +CTCTCT +CTCTCC +CTCTCG +CTCTGA +CTCTGT +CTCTGC +CTCTGG +CTCCAA +CTCCAT +CTCCAC +CTCCAG +CTCCTA +CTCCTT +CTCCTC +CTCCTG +CTCCCA +CTCCCT +CTCCCC +CTCCCG +CTCCGA +CTCCGT +CTCCGC +CTCCGG +CTCGAA +CTCGAT +CTCGAC +CTCGAG +CTCGTA +CTCGTT +CTCGTC +CTCGTG +CTCGCA +CTCGCT +CTCGCC +CTCGCG +CTCGGA +CTCGGT +CTCGGC +CTCGGG +CTGAAA +CTGAAT +CTGAAC +CTGAAG +CTGATA +CTGATT +CTGATC +CTGATG +CTGACA +CTGACT +CTGACC +CTGACG +CTGAGA +CTGAGT +CTGAGC +CTGAGG +CTGTAA +CTGTAT +CTGTAC +CTGTAG +CTGTTA +CTGTTT +CTGTTC +CTGTTG +CTGTCA +CTGTCT +CTGTCC +CTGTCG +CTGTGA +CTGTGT +CTGTGC +CTGTGG +CTGCAA +CTGCAT +CTGCAC +CTGCAG +CTGCTA +CTGCTT +CTGCTC +CTGCTG +CTGCCA +CTGCCT +CTGCCC +CTGCCG +CTGCGA +CTGCGT +CTGCGC +CTGCGG +CTGGAA +CTGGAT +CTGGAC +CTGGAG +CTGGTA +CTGGTT +CTGGTC +CTGGTG +CTGGCA +CTGGCT +CTGGCC +CTGGCG +CTGGGA +CTGGGT +CTGGGC +CTGGGG +CCAAAA +CCAAAT +CCAAAC +CCAAAG +CCAATA +CCAATT +CCAATC +CCAATG +CCAACA +CCAACT +CCAACC +CCAACG +CCAAGA +CCAAGT +CCAAGC +CCAAGG +CCATAA +CCATAT +CCATAC +CCATAG +CCATTA +CCATTT +CCATTC +CCATTG +CCATCA +CCATCT +CCATCC +CCATCG +CCATGA +CCATGT +CCATGC +CCATGG +CCACAA +CCACAT +CCACAC +CCACAG +CCACTA +CCACTT +CCACTC +CCACTG +CCACCA +CCACCT +CCACCC +CCACCG +CCACGA +CCACGT +CCACGC +CCACGG +CCAGAA +CCAGAT +CCAGAC +CCAGAG +CCAGTA +CCAGTT +CCAGTC +CCAGTG +CCAGCA +CCAGCT +CCAGCC +CCAGCG +CCAGGA +CCAGGT +CCAGGC +CCAGGG +CCTAAA +CCTAAT +CCTAAC +CCTAAG +CCTATA +CCTATT +CCTATC +CCTATG +CCTACA +CCTACT +CCTACC +CCTACG +CCTAGA +CCTAGT +CCTAGC +CCTAGG +CCTTAA +CCTTAT +CCTTAC +CCTTAG +CCTTTA +CCTTTT +CCTTTC +CCTTTG +CCTTCA +CCTTCT +CCTTCC +CCTTCG +CCTTGA +CCTTGT +CCTTGC +CCTTGG +CCTCAA +CCTCAT +CCTCAC +CCTCAG +CCTCTA +CCTCTT +CCTCTC +CCTCTG +CCTCCA +CCTCCT +CCTCCC +CCTCCG +CCTCGA +CCTCGT +CCTCGC +CCTCGG +CCTGAA +CCTGAT +CCTGAC +CCTGAG +CCTGTA +CCTGTT +CCTGTC +CCTGTG +CCTGCA +CCTGCT +CCTGCC +CCTGCG +CCTGGA +CCTGGT +CCTGGC +CCTGGG +CCCAAA +CCCAAT +CCCAAC +CCCAAG +CCCATA +CCCATT +CCCATC +CCCATG +CCCACA +CCCACT +CCCACC +CCCACG +CCCAGA +CCCAGT +CCCAGC +CCCAGG +CCCTAA +CCCTAT +CCCTAC +CCCTAG +CCCTTA +CCCTTT +CCCTTC +CCCTTG +CCCTCA +CCCTCT +CCCTCC +CCCTCG +CCCTGA +CCCTGT +CCCTGC +CCCTGG +CCCCAA +CCCCAT +CCCCAC +CCCCAG +CCCCTA +CCCCTT +CCCCTC +CCCCTG +CCCCCA +CCCCCT +CCCCCC +CCCCCG +CCCCGA +CCCCGT +CCCCGC +CCCCGG +CCCGAA +CCCGAT +CCCGAC +CCCGAG +CCCGTA +CCCGTT +CCCGTC +CCCGTG +CCCGCA +CCCGCT +CCCGCC +CCCGCG +CCCGGA +CCCGGT +CCCGGC +CCCGGG +CCGAAA +CCGAAT +CCGAAC +CCGAAG +CCGATA +CCGATT +CCGATC +CCGATG +CCGACA +CCGACT +CCGACC +CCGACG +CCGAGA +CCGAGT +CCGAGC +CCGAGG +CCGTAA +CCGTAT +CCGTAC +CCGTAG +CCGTTA +CCGTTT +CCGTTC +CCGTTG +CCGTCA +CCGTCT +CCGTCC +CCGTCG +CCGTGA +CCGTGT +CCGTGC +CCGTGG +CCGCAA +CCGCAT +CCGCAC +CCGCAG +CCGCTA +CCGCTT +CCGCTC +CCGCTG +CCGCCA +CCGCCT +CCGCCC +CCGCCG +CCGCGA +CCGCGT +CCGCGC +CCGCGG +CCGGAA +CCGGAT +CCGGAC +CCGGAG +CCGGTA +CCGGTT +CCGGTC +CCGGTG +CCGGCA +CCGGCT +CCGGCC +CCGGCG +CCGGGA +CCGGGT +CCGGGC +CCGGGG +CGAAAA +CGAAAT +CGAAAC +CGAAAG +CGAATA +CGAATT +CGAATC +CGAATG +CGAACA +CGAACT +CGAACC +CGAACG +CGAAGA +CGAAGT +CGAAGC +CGAAGG +CGATAA +CGATAT +CGATAC +CGATAG +CGATTA +CGATTT +CGATTC +CGATTG +CGATCA +CGATCT +CGATCC +CGATCG +CGATGA +CGATGT +CGATGC +CGATGG +CGACAA +CGACAT +CGACAC +CGACAG +CGACTA +CGACTT +CGACTC +CGACTG +CGACCA +CGACCT +CGACCC +CGACCG +CGACGA +CGACGT +CGACGC +CGACGG +CGAGAA +CGAGAT +CGAGAC +CGAGAG +CGAGTA +CGAGTT +CGAGTC +CGAGTG +CGAGCA +CGAGCT +CGAGCC +CGAGCG +CGAGGA +CGAGGT +CGAGGC +CGAGGG +CGTAAA +CGTAAT +CGTAAC +CGTAAG +CGTATA +CGTATT +CGTATC +CGTATG +CGTACA +CGTACT +CGTACC +CGTACG +CGTAGA +CGTAGT +CGTAGC +CGTAGG +CGTTAA +CGTTAT +CGTTAC +CGTTAG +CGTTTA +CGTTTT +CGTTTC +CGTTTG +CGTTCA +CGTTCT +CGTTCC +CGTTCG +CGTTGA +CGTTGT +CGTTGC +CGTTGG +CGTCAA +CGTCAT +CGTCAC +CGTCAG +CGTCTA +CGTCTT +CGTCTC +CGTCTG +CGTCCA +CGTCCT +CGTCCC +CGTCCG +CGTCGA +CGTCGT +CGTCGC +CGTCGG +CGTGAA +CGTGAT +CGTGAC +CGTGAG +CGTGTA +CGTGTT +CGTGTC +CGTGTG +CGTGCA +CGTGCT +CGTGCC +CGTGCG +CGTGGA +CGTGGT +CGTGGC +CGTGGG +CGCAAA +CGCAAT +CGCAAC +CGCAAG +CGCATA +CGCATT +CGCATC +CGCATG +CGCACA +CGCACT +CGCACC +CGCACG +CGCAGA +CGCAGT +CGCAGC +CGCAGG +CGCTAA +CGCTAT +CGCTAC +CGCTAG +CGCTTA +CGCTTT +CGCTTC +CGCTTG +CGCTCA +CGCTCT +CGCTCC +CGCTCG +CGCTGA +CGCTGT +CGCTGC +CGCTGG +CGCCAA +CGCCAT +CGCCAC +CGCCAG +CGCCTA +CGCCTT +CGCCTC +CGCCTG +CGCCCA +CGCCCT +CGCCCC +CGCCCG +CGCCGA +CGCCGT +CGCCGC +CGCCGG +CGCGAA +CGCGAT +CGCGAC +CGCGAG +CGCGTA +CGCGTT +CGCGTC +CGCGTG +CGCGCA +CGCGCT +CGCGCC +CGCGCG +CGCGGA +CGCGGT +CGCGGC +CGCGGG +CGGAAA +CGGAAT +CGGAAC +CGGAAG +CGGATA +CGGATT +CGGATC +CGGATG +CGGACA +CGGACT +CGGACC +CGGACG +CGGAGA +CGGAGT +CGGAGC +CGGAGG +CGGTAA +CGGTAT +CGGTAC +CGGTAG +CGGTTA +CGGTTT +CGGTTC +CGGTTG +CGGTCA +CGGTCT +CGGTCC +CGGTCG +CGGTGA +CGGTGT +CGGTGC +CGGTGG +CGGCAA +CGGCAT +CGGCAC +CGGCAG +CGGCTA +CGGCTT +CGGCTC +CGGCTG +CGGCCA +CGGCCT +CGGCCC +CGGCCG +CGGCGA +CGGCGT +CGGCGC +CGGCGG +CGGGAA +CGGGAT +CGGGAC +CGGGAG +CGGGTA +CGGGTT +CGGGTC +CGGGTG +CGGGCA +CGGGCT +CGGGCC +CGGGCG +CGGGGA +CGGGGT +CGGGGC +CGGGGG +GAAAAA +GAAAAT +GAAAAC +GAAAAG +GAAATA +GAAATT +GAAATC +GAAATG +GAAACA +GAAACT +GAAACC +GAAACG +GAAAGA +GAAAGT +GAAAGC +GAAAGG +GAATAA +GAATAT +GAATAC +GAATAG +GAATTA +GAATTT +GAATTC +GAATTG +GAATCA +GAATCT +GAATCC +GAATCG +GAATGA +GAATGT +GAATGC +GAATGG +GAACAA +GAACAT +GAACAC +GAACAG +GAACTA +GAACTT +GAACTC +GAACTG +GAACCA +GAACCT +GAACCC +GAACCG +GAACGA +GAACGT +GAACGC +GAACGG +GAAGAA +GAAGAT +GAAGAC +GAAGAG +GAAGTA +GAAGTT +GAAGTC +GAAGTG +GAAGCA +GAAGCT +GAAGCC +GAAGCG +GAAGGA +GAAGGT +GAAGGC +GAAGGG +GATAAA +GATAAT +GATAAC +GATAAG +GATATA +GATATT +GATATC +GATATG +GATACA +GATACT +GATACC +GATACG +GATAGA +GATAGT +GATAGC +GATAGG +GATTAA +GATTAT +GATTAC +GATTAG +GATTTA +GATTTT +GATTTC +GATTTG +GATTCA +GATTCT +GATTCC +GATTCG +GATTGA +GATTGT +GATTGC +GATTGG +GATCAA +GATCAT +GATCAC +GATCAG +GATCTA +GATCTT +GATCTC +GATCTG +GATCCA +GATCCT +GATCCC +GATCCG +GATCGA +GATCGT +GATCGC +GATCGG +GATGAA +GATGAT +GATGAC +GATGAG +GATGTA +GATGTT +GATGTC +GATGTG +GATGCA +GATGCT +GATGCC +GATGCG +GATGGA +GATGGT +GATGGC +GATGGG +GACAAA +GACAAT +GACAAC +GACAAG +GACATA +GACATT +GACATC +GACATG +GACACA +GACACT +GACACC +GACACG +GACAGA +GACAGT +GACAGC +GACAGG +GACTAA +GACTAT +GACTAC +GACTAG +GACTTA +GACTTT +GACTTC +GACTTG +GACTCA +GACTCT +GACTCC +GACTCG +GACTGA +GACTGT +GACTGC +GACTGG +GACCAA +GACCAT +GACCAC +GACCAG +GACCTA +GACCTT +GACCTC +GACCTG +GACCCA +GACCCT +GACCCC +GACCCG +GACCGA +GACCGT +GACCGC +GACCGG +GACGAA +GACGAT +GACGAC +GACGAG +GACGTA +GACGTT +GACGTC +GACGTG +GACGCA +GACGCT +GACGCC +GACGCG +GACGGA +GACGGT +GACGGC +GACGGG +GAGAAA +GAGAAT +GAGAAC +GAGAAG +GAGATA +GAGATT +GAGATC +GAGATG +GAGACA +GAGACT +GAGACC +GAGACG +GAGAGA +GAGAGT +GAGAGC +GAGAGG +GAGTAA +GAGTAT +GAGTAC +GAGTAG +GAGTTA +GAGTTT +GAGTTC +GAGTTG +GAGTCA +GAGTCT +GAGTCC +GAGTCG +GAGTGA +GAGTGT +GAGTGC +GAGTGG +GAGCAA +GAGCAT +GAGCAC +GAGCAG +GAGCTA +GAGCTT +GAGCTC +GAGCTG +GAGCCA +GAGCCT +GAGCCC +GAGCCG +GAGCGA +GAGCGT +GAGCGC +GAGCGG +GAGGAA +GAGGAT +GAGGAC +GAGGAG +GAGGTA +GAGGTT +GAGGTC +GAGGTG +GAGGCA +GAGGCT +GAGGCC +GAGGCG +GAGGGA +GAGGGT +GAGGGC +GAGGGG +GTAAAA +GTAAAT +GTAAAC +GTAAAG +GTAATA +GTAATT +GTAATC +GTAATG +GTAACA +GTAACT +GTAACC +GTAACG +GTAAGA +GTAAGT +GTAAGC +GTAAGG +GTATAA +GTATAT +GTATAC +GTATAG +GTATTA +GTATTT +GTATTC +GTATTG +GTATCA +GTATCT +GTATCC +GTATCG +GTATGA +GTATGT +GTATGC +GTATGG +GTACAA +GTACAT +GTACAC +GTACAG +GTACTA +GTACTT +GTACTC +GTACTG +GTACCA +GTACCT +GTACCC +GTACCG +GTACGA +GTACGT +GTACGC +GTACGG +GTAGAA +GTAGAT +GTAGAC +GTAGAG +GTAGTA +GTAGTT +GTAGTC +GTAGTG +GTAGCA +GTAGCT +GTAGCC +GTAGCG +GTAGGA +GTAGGT +GTAGGC +GTAGGG +GTTAAA +GTTAAT +GTTAAC +GTTAAG +GTTATA +GTTATT +GTTATC +GTTATG +GTTACA +GTTACT +GTTACC +GTTACG +GTTAGA +GTTAGT +GTTAGC +GTTAGG +GTTTAA +GTTTAT +GTTTAC +GTTTAG +GTTTTA +GTTTTT +GTTTTC +GTTTTG +GTTTCA +GTTTCT +GTTTCC +GTTTCG +GTTTGA +GTTTGT +GTTTGC +GTTTGG +GTTCAA +GTTCAT +GTTCAC +GTTCAG +GTTCTA +GTTCTT +GTTCTC +GTTCTG +GTTCCA +GTTCCT +GTTCCC +GTTCCG +GTTCGA +GTTCGT +GTTCGC +GTTCGG +GTTGAA +GTTGAT +GTTGAC +GTTGAG +GTTGTA +GTTGTT +GTTGTC +GTTGTG +GTTGCA +GTTGCT +GTTGCC +GTTGCG +GTTGGA +GTTGGT +GTTGGC +GTTGGG +GTCAAA +GTCAAT +GTCAAC +GTCAAG +GTCATA +GTCATT +GTCATC +GTCATG +GTCACA +GTCACT +GTCACC +GTCACG +GTCAGA +GTCAGT +GTCAGC +GTCAGG +GTCTAA +GTCTAT +GTCTAC +GTCTAG +GTCTTA +GTCTTT +GTCTTC +GTCTTG +GTCTCA +GTCTCT +GTCTCC +GTCTCG +GTCTGA +GTCTGT +GTCTGC +GTCTGG +GTCCAA +GTCCAT +GTCCAC +GTCCAG +GTCCTA +GTCCTT +GTCCTC +GTCCTG +GTCCCA +GTCCCT +GTCCCC +GTCCCG +GTCCGA +GTCCGT +GTCCGC +GTCCGG +GTCGAA +GTCGAT +GTCGAC +GTCGAG +GTCGTA +GTCGTT +GTCGTC +GTCGTG +GTCGCA +GTCGCT +GTCGCC +GTCGCG +GTCGGA +GTCGGT +GTCGGC +GTCGGG +GTGAAA +GTGAAT +GTGAAC +GTGAAG +GTGATA +GTGATT +GTGATC +GTGATG +GTGACA +GTGACT +GTGACC +GTGACG +GTGAGA +GTGAGT +GTGAGC +GTGAGG +GTGTAA +GTGTAT +GTGTAC +GTGTAG +GTGTTA +GTGTTT +GTGTTC +GTGTTG +GTGTCA +GTGTCT +GTGTCC +GTGTCG +GTGTGA +GTGTGT +GTGTGC +GTGTGG +GTGCAA +GTGCAT +GTGCAC +GTGCAG +GTGCTA +GTGCTT +GTGCTC +GTGCTG +GTGCCA +GTGCCT +GTGCCC +GTGCCG +GTGCGA +GTGCGT +GTGCGC +GTGCGG +GTGGAA +GTGGAT +GTGGAC +GTGGAG +GTGGTA +GTGGTT +GTGGTC +GTGGTG +GTGGCA +GTGGCT +GTGGCC +GTGGCG +GTGGGA +GTGGGT +GTGGGC +GTGGGG +GCAAAA +GCAAAT +GCAAAC +GCAAAG +GCAATA +GCAATT +GCAATC +GCAATG +GCAACA +GCAACT +GCAACC +GCAACG +GCAAGA +GCAAGT +GCAAGC +GCAAGG +GCATAA +GCATAT +GCATAC +GCATAG +GCATTA +GCATTT +GCATTC +GCATTG +GCATCA +GCATCT +GCATCC +GCATCG +GCATGA +GCATGT +GCATGC +GCATGG +GCACAA +GCACAT +GCACAC +GCACAG +GCACTA +GCACTT +GCACTC +GCACTG +GCACCA +GCACCT +GCACCC +GCACCG +GCACGA +GCACGT +GCACGC +GCACGG +GCAGAA +GCAGAT +GCAGAC +GCAGAG +GCAGTA +GCAGTT +GCAGTC +GCAGTG +GCAGCA +GCAGCT +GCAGCC +GCAGCG +GCAGGA +GCAGGT +GCAGGC +GCAGGG +GCTAAA +GCTAAT +GCTAAC +GCTAAG +GCTATA +GCTATT +GCTATC +GCTATG +GCTACA +GCTACT +GCTACC +GCTACG +GCTAGA +GCTAGT +GCTAGC +GCTAGG +GCTTAA +GCTTAT +GCTTAC +GCTTAG +GCTTTA +GCTTTT +GCTTTC +GCTTTG +GCTTCA +GCTTCT +GCTTCC +GCTTCG +GCTTGA +GCTTGT +GCTTGC +GCTTGG +GCTCAA +GCTCAT +GCTCAC +GCTCAG +GCTCTA +GCTCTT +GCTCTC +GCTCTG +GCTCCA +GCTCCT +GCTCCC +GCTCCG +GCTCGA +GCTCGT +GCTCGC +GCTCGG +GCTGAA +GCTGAT +GCTGAC +GCTGAG +GCTGTA +GCTGTT +GCTGTC +GCTGTG +GCTGCA +GCTGCT +GCTGCC +GCTGCG +GCTGGA +GCTGGT +GCTGGC +GCTGGG +GCCAAA +GCCAAT +GCCAAC +GCCAAG +GCCATA +GCCATT +GCCATC +GCCATG +GCCACA +GCCACT +GCCACC +GCCACG +GCCAGA +GCCAGT +GCCAGC +GCCAGG +GCCTAA +GCCTAT +GCCTAC +GCCTAG +GCCTTA +GCCTTT +GCCTTC +GCCTTG +GCCTCA +GCCTCT +GCCTCC +GCCTCG +GCCTGA +GCCTGT +GCCTGC +GCCTGG +GCCCAA +GCCCAT +GCCCAC +GCCCAG +GCCCTA +GCCCTT +GCCCTC +GCCCTG +GCCCCA +GCCCCT +GCCCCC +GCCCCG +GCCCGA +GCCCGT +GCCCGC +GCCCGG +GCCGAA +GCCGAT +GCCGAC +GCCGAG +GCCGTA +GCCGTT +GCCGTC +GCCGTG +GCCGCA +GCCGCT +GCCGCC +GCCGCG +GCCGGA +GCCGGT +GCCGGC +GCCGGG +GCGAAA +GCGAAT +GCGAAC +GCGAAG +GCGATA +GCGATT +GCGATC +GCGATG +GCGACA +GCGACT +GCGACC +GCGACG +GCGAGA +GCGAGT +GCGAGC +GCGAGG +GCGTAA +GCGTAT +GCGTAC +GCGTAG +GCGTTA +GCGTTT +GCGTTC +GCGTTG +GCGTCA +GCGTCT +GCGTCC +GCGTCG +GCGTGA +GCGTGT +GCGTGC +GCGTGG +GCGCAA +GCGCAT +GCGCAC +GCGCAG +GCGCTA +GCGCTT +GCGCTC +GCGCTG +GCGCCA +GCGCCT +GCGCCC +GCGCCG +GCGCGA +GCGCGT +GCGCGC +GCGCGG +GCGGAA +GCGGAT +GCGGAC +GCGGAG +GCGGTA +GCGGTT +GCGGTC +GCGGTG +GCGGCA +GCGGCT +GCGGCC +GCGGCG +GCGGGA +GCGGGT +GCGGGC +GCGGGG +GGAAAA +GGAAAT +GGAAAC +GGAAAG +GGAATA +GGAATT +GGAATC +GGAATG +GGAACA +GGAACT +GGAACC +GGAACG +GGAAGA +GGAAGT +GGAAGC +GGAAGG +GGATAA +GGATAT +GGATAC +GGATAG +GGATTA +GGATTT +GGATTC +GGATTG +GGATCA +GGATCT +GGATCC +GGATCG +GGATGA +GGATGT +GGATGC +GGATGG +GGACAA +GGACAT +GGACAC +GGACAG +GGACTA +GGACTT +GGACTC +GGACTG +GGACCA +GGACCT +GGACCC +GGACCG +GGACGA +GGACGT +GGACGC +GGACGG +GGAGAA +GGAGAT +GGAGAC +GGAGAG +GGAGTA +GGAGTT +GGAGTC +GGAGTG +GGAGCA +GGAGCT +GGAGCC +GGAGCG +GGAGGA +GGAGGT +GGAGGC +GGAGGG +GGTAAA +GGTAAT +GGTAAC +GGTAAG +GGTATA +GGTATT +GGTATC +GGTATG +GGTACA +GGTACT +GGTACC +GGTACG +GGTAGA +GGTAGT +GGTAGC +GGTAGG +GGTTAA +GGTTAT +GGTTAC +GGTTAG +GGTTTA +GGTTTT +GGTTTC +GGTTTG +GGTTCA +GGTTCT +GGTTCC +GGTTCG +GGTTGA +GGTTGT +GGTTGC +GGTTGG +GGTCAA +GGTCAT +GGTCAC +GGTCAG +GGTCTA +GGTCTT +GGTCTC +GGTCTG +GGTCCA +GGTCCT +GGTCCC +GGTCCG +GGTCGA +GGTCGT +GGTCGC +GGTCGG +GGTGAA +GGTGAT +GGTGAC +GGTGAG +GGTGTA +GGTGTT +GGTGTC +GGTGTG +GGTGCA +GGTGCT +GGTGCC +GGTGCG +GGTGGA +GGTGGT +GGTGGC +GGTGGG +GGCAAA +GGCAAT +GGCAAC +GGCAAG +GGCATA +GGCATT +GGCATC +GGCATG +GGCACA +GGCACT +GGCACC +GGCACG +GGCAGA +GGCAGT +GGCAGC +GGCAGG +GGCTAA +GGCTAT +GGCTAC +GGCTAG +GGCTTA +GGCTTT +GGCTTC +GGCTTG +GGCTCA +GGCTCT +GGCTCC +GGCTCG +GGCTGA +GGCTGT +GGCTGC +GGCTGG +GGCCAA +GGCCAT +GGCCAC +GGCCAG +GGCCTA +GGCCTT +GGCCTC +GGCCTG +GGCCCA +GGCCCT +GGCCCC +GGCCCG +GGCCGA +GGCCGT +GGCCGC +GGCCGG +GGCGAA +GGCGAT +GGCGAC +GGCGAG +GGCGTA +GGCGTT +GGCGTC +GGCGTG +GGCGCA +GGCGCT +GGCGCC +GGCGCG +GGCGGA +GGCGGT +GGCGGC +GGCGGG +GGGAAA +GGGAAT +GGGAAC +GGGAAG +GGGATA +GGGATT +GGGATC +GGGATG +GGGACA +GGGACT +GGGACC +GGGACG +GGGAGA +GGGAGT +GGGAGC +GGGAGG +GGGTAA +GGGTAT +GGGTAC +GGGTAG +GGGTTA +GGGTTT +GGGTTC +GGGTTG +GGGTCA +GGGTCT +GGGTCC +GGGTCG +GGGTGA +GGGTGT +GGGTGC +GGGTGG +GGGCAA +GGGCAT +GGGCAC +GGGCAG +GGGCTA +GGGCTT +GGGCTC +GGGCTG +GGGCCA +GGGCCT +GGGCCC +GGGCCG +GGGCGA +GGGCGT +GGGCGC +GGGCGG +GGGGAA +GGGGAT +GGGGAC +GGGGAG +GGGGTA +GGGGTT +GGGGTC +GGGGTG +GGGGCA +GGGGCT +GGGGCC +GGGGCG +GGGGGA +GGGGGT +GGGGGC +GGGGGG diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-3000/config.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..9b14e61d327bf586fcff6e817cf9d365bae18ad2 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-3000/config.json @@ -0,0 +1,26 @@ +{ + "_name_or_path": "fabihamakhdoomi/TinyDNABERT", + "architectures": [ + "BertForSequenceClassification" + ], + "attention_probs_dropout_prob": 0.1, + "classifier_dropout": null, + "hidden_act": "gelu", + "hidden_dropout_prob": 0.1, + "hidden_size": 128, + "initializer_range": 0.02, + "intermediate_size": 512, + "layer_norm_eps": 1e-12, + "max_position_embeddings": 512, + "model_type": "bert", + "num_attention_heads": 2, + "num_hidden_layers": 2, + "pad_token_id": 0, + "position_embedding_type": "absolute", + "problem_type": "single_label_classification", + "torch_dtype": "float32", + "transformers_version": "4.44.0", + "type_vocab_size": 2, + "use_cache": true, + "vocab_size": 4101 +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-3000/model.safetensors b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-3000/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..493c994f5799f8a1687248a1b94bdf73811f45f4 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-3000/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a76ff49450947e0812b095b2592e11a330460c79c486e177e508f21c4ea9f7a0 +size 4021688 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-3000/optimizer.pt b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-3000/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..8a503c8f32f1d44329c5099e3f7d6d7a485810ef --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-3000/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f544cc8208641b7395e3d6810e616fbcdac7c77f297bf6c098001f5c512e9491 +size 8058700 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-3000/rng_state.pth b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-3000/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..6a219fb332ae976c7a91fd53a363fd52b1ca4044 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-3000/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9a88049e3e15cef9ae24c0067d91d8e8d16ed125ffa58bfc970554dbe3e0b701 +size 14244 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-3000/scheduler.pt b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-3000/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..6c655a9f3dd5d4d532f6c0a84b19b9ae5b766437 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-3000/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dbc2253b3cc1ddd11640ea56d3925f925017f9aa6e1577411dac42d629457af6 +size 1064 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-3000/special_tokens_map.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-3000/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..a8b3208c2884c4efb86e49300fdd3dc877220cdf --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-3000/special_tokens_map.json @@ -0,0 +1,7 @@ +{ + "cls_token": "[CLS]", + "mask_token": "[MASK]", + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-3000/tokenizer.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-3000/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..ebd782de6963b9032ce9374cfe30f5275be1a30b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-3000/tokenizer.json @@ -0,0 +1,4265 @@ +{ + "version": "1.0", + "truncation": { + "direction": "Right", + "max_length": 512, + "strategy": "LongestFirst", + "stride": 0 + }, + "padding": { + "strategy": { + "Fixed": 512 + }, + "direction": "Right", + "pad_to_multiple_of": null, + "pad_id": 0, + "pad_type_id": 0, + "pad_token": "[PAD]" + }, + "added_tokens": [ + { + "id": 0, + "content": "[PAD]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 1, + "content": "[UNK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 2, + "content": "[CLS]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 3, + "content": "[SEP]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 4, + "content": "[MASK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + } + ], + "normalizer": { + "type": "BertNormalizer", + "clean_text": true, + "handle_chinese_chars": true, + "strip_accents": null, + "lowercase": false + }, + "pre_tokenizer": { + "type": "BertPreTokenizer" + }, + "post_processor": { + "type": "TemplateProcessing", + "single": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + } + ], + "pair": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "B", + "type_id": 1 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 1 + } + } + ], + "special_tokens": { + "[CLS]": { + "id": "[CLS]", + "ids": [ + 2 + ], + "tokens": [ + "[CLS]" + ] + }, + "[SEP]": { + "id": "[SEP]", + "ids": [ + 3 + ], + "tokens": [ + "[SEP]" + ] + } + } + }, + "decoder": { + "type": "WordPiece", + "prefix": "##", + "cleanup": true + }, + "model": { + "type": "WordPiece", + "unk_token": "[UNK]", + "continuing_subword_prefix": "##", + "max_input_chars_per_word": 100, + "vocab": { + "[PAD]": 0, + "[UNK]": 1, + "[CLS]": 2, + "[SEP]": 3, + "[MASK]": 4, + "AAAAAA": 5, + "AAAAAT": 6, + "AAAAAC": 7, + "AAAAAG": 8, + "AAAATA": 9, + "AAAATT": 10, + "AAAATC": 11, + "AAAATG": 12, + "AAAACA": 13, + "AAAACT": 14, + "AAAACC": 15, + "AAAACG": 16, + "AAAAGA": 17, + "AAAAGT": 18, + "AAAAGC": 19, + "AAAAGG": 20, + "AAATAA": 21, + "AAATAT": 22, + "AAATAC": 23, + "AAATAG": 24, + "AAATTA": 25, + "AAATTT": 26, + "AAATTC": 27, + "AAATTG": 28, + "AAATCA": 29, + "AAATCT": 30, + "AAATCC": 31, + "AAATCG": 32, + "AAATGA": 33, + "AAATGT": 34, + "AAATGC": 35, + "AAATGG": 36, + "AAACAA": 37, + "AAACAT": 38, + "AAACAC": 39, + "AAACAG": 40, + "AAACTA": 41, + "AAACTT": 42, + "AAACTC": 43, + "AAACTG": 44, + "AAACCA": 45, + "AAACCT": 46, + "AAACCC": 47, + "AAACCG": 48, + "AAACGA": 49, + "AAACGT": 50, + "AAACGC": 51, + "AAACGG": 52, + "AAAGAA": 53, + "AAAGAT": 54, + "AAAGAC": 55, + "AAAGAG": 56, + "AAAGTA": 57, + "AAAGTT": 58, + "AAAGTC": 59, + "AAAGTG": 60, + "AAAGCA": 61, + "AAAGCT": 62, + "AAAGCC": 63, + "AAAGCG": 64, + "AAAGGA": 65, + "AAAGGT": 66, + "AAAGGC": 67, + "AAAGGG": 68, + "AATAAA": 69, + "AATAAT": 70, + "AATAAC": 71, + "AATAAG": 72, + "AATATA": 73, + "AATATT": 74, + "AATATC": 75, + "AATATG": 76, + "AATACA": 77, + "AATACT": 78, + "AATACC": 79, + "AATACG": 80, + "AATAGA": 81, + "AATAGT": 82, + "AATAGC": 83, + "AATAGG": 84, + "AATTAA": 85, + "AATTAT": 86, + "AATTAC": 87, + "AATTAG": 88, + "AATTTA": 89, + "AATTTT": 90, + "AATTTC": 91, + "AATTTG": 92, + "AATTCA": 93, + "AATTCT": 94, + "AATTCC": 95, + "AATTCG": 96, + "AATTGA": 97, + "AATTGT": 98, + "AATTGC": 99, + "AATTGG": 100, + "AATCAA": 101, + "AATCAT": 102, + "AATCAC": 103, + "AATCAG": 104, + "AATCTA": 105, + "AATCTT": 106, + "AATCTC": 107, + "AATCTG": 108, + "AATCCA": 109, + "AATCCT": 110, + "AATCCC": 111, + "AATCCG": 112, + "AATCGA": 113, + "AATCGT": 114, + "AATCGC": 115, + "AATCGG": 116, + "AATGAA": 117, + "AATGAT": 118, + "AATGAC": 119, + "AATGAG": 120, + "AATGTA": 121, + "AATGTT": 122, + "AATGTC": 123, + "AATGTG": 124, + "AATGCA": 125, + "AATGCT": 126, + "AATGCC": 127, + "AATGCG": 128, + "AATGGA": 129, + "AATGGT": 130, + "AATGGC": 131, + "AATGGG": 132, + "AACAAA": 133, + "AACAAT": 134, + "AACAAC": 135, + "AACAAG": 136, + "AACATA": 137, + "AACATT": 138, + "AACATC": 139, + "AACATG": 140, + "AACACA": 141, + "AACACT": 142, + "AACACC": 143, + "AACACG": 144, + "AACAGA": 145, + "AACAGT": 146, + "AACAGC": 147, + "AACAGG": 148, + "AACTAA": 149, + "AACTAT": 150, + "AACTAC": 151, + "AACTAG": 152, + "AACTTA": 153, + "AACTTT": 154, + "AACTTC": 155, + "AACTTG": 156, + "AACTCA": 157, + "AACTCT": 158, + "AACTCC": 159, + "AACTCG": 160, + "AACTGA": 161, + "AACTGT": 162, + "AACTGC": 163, + "AACTGG": 164, + "AACCAA": 165, + "AACCAT": 166, + "AACCAC": 167, + "AACCAG": 168, + "AACCTA": 169, + "AACCTT": 170, + "AACCTC": 171, + "AACCTG": 172, + "AACCCA": 173, + "AACCCT": 174, + "AACCCC": 175, + "AACCCG": 176, + "AACCGA": 177, + "AACCGT": 178, + "AACCGC": 179, + "AACCGG": 180, + "AACGAA": 181, + "AACGAT": 182, + "AACGAC": 183, + "AACGAG": 184, + "AACGTA": 185, + "AACGTT": 186, + "AACGTC": 187, + "AACGTG": 188, + "AACGCA": 189, + "AACGCT": 190, + "AACGCC": 191, + "AACGCG": 192, + "AACGGA": 193, + "AACGGT": 194, + "AACGGC": 195, + "AACGGG": 196, + "AAGAAA": 197, + "AAGAAT": 198, + "AAGAAC": 199, + "AAGAAG": 200, + "AAGATA": 201, + "AAGATT": 202, + "AAGATC": 203, + "AAGATG": 204, + "AAGACA": 205, + "AAGACT": 206, + "AAGACC": 207, + "AAGACG": 208, + "AAGAGA": 209, + "AAGAGT": 210, + "AAGAGC": 211, + "AAGAGG": 212, + "AAGTAA": 213, + "AAGTAT": 214, + "AAGTAC": 215, + "AAGTAG": 216, + "AAGTTA": 217, + "AAGTTT": 218, + "AAGTTC": 219, + "AAGTTG": 220, + "AAGTCA": 221, + "AAGTCT": 222, + "AAGTCC": 223, + "AAGTCG": 224, + "AAGTGA": 225, + "AAGTGT": 226, + "AAGTGC": 227, + "AAGTGG": 228, + "AAGCAA": 229, + "AAGCAT": 230, + "AAGCAC": 231, + "AAGCAG": 232, + "AAGCTA": 233, + "AAGCTT": 234, + "AAGCTC": 235, + "AAGCTG": 236, + "AAGCCA": 237, + "AAGCCT": 238, + "AAGCCC": 239, + "AAGCCG": 240, + "AAGCGA": 241, + "AAGCGT": 242, + "AAGCGC": 243, + "AAGCGG": 244, + "AAGGAA": 245, + "AAGGAT": 246, + "AAGGAC": 247, + "AAGGAG": 248, + "AAGGTA": 249, + "AAGGTT": 250, + "AAGGTC": 251, + "AAGGTG": 252, + "AAGGCA": 253, + "AAGGCT": 254, + "AAGGCC": 255, + "AAGGCG": 256, + "AAGGGA": 257, + "AAGGGT": 258, + "AAGGGC": 259, + "AAGGGG": 260, + "ATAAAA": 261, + "ATAAAT": 262, + "ATAAAC": 263, + "ATAAAG": 264, + "ATAATA": 265, + "ATAATT": 266, + "ATAATC": 267, + "ATAATG": 268, + "ATAACA": 269, + "ATAACT": 270, + "ATAACC": 271, + "ATAACG": 272, + "ATAAGA": 273, + "ATAAGT": 274, + "ATAAGC": 275, + "ATAAGG": 276, + "ATATAA": 277, + "ATATAT": 278, + "ATATAC": 279, + "ATATAG": 280, + "ATATTA": 281, + "ATATTT": 282, + "ATATTC": 283, + "ATATTG": 284, + "ATATCA": 285, + "ATATCT": 286, + "ATATCC": 287, + "ATATCG": 288, + "ATATGA": 289, + "ATATGT": 290, + "ATATGC": 291, + "ATATGG": 292, + "ATACAA": 293, + "ATACAT": 294, + "ATACAC": 295, + "ATACAG": 296, + "ATACTA": 297, + "ATACTT": 298, + "ATACTC": 299, + "ATACTG": 300, + "ATACCA": 301, + "ATACCT": 302, + "ATACCC": 303, + "ATACCG": 304, + "ATACGA": 305, + "ATACGT": 306, + "ATACGC": 307, + "ATACGG": 308, + "ATAGAA": 309, + "ATAGAT": 310, + "ATAGAC": 311, + "ATAGAG": 312, + "ATAGTA": 313, + "ATAGTT": 314, + "ATAGTC": 315, + "ATAGTG": 316, + "ATAGCA": 317, + "ATAGCT": 318, + "ATAGCC": 319, + "ATAGCG": 320, + "ATAGGA": 321, + "ATAGGT": 322, + "ATAGGC": 323, + "ATAGGG": 324, + "ATTAAA": 325, + "ATTAAT": 326, + "ATTAAC": 327, + "ATTAAG": 328, + "ATTATA": 329, + "ATTATT": 330, + "ATTATC": 331, + "ATTATG": 332, + "ATTACA": 333, + "ATTACT": 334, + "ATTACC": 335, + "ATTACG": 336, + "ATTAGA": 337, + "ATTAGT": 338, + "ATTAGC": 339, + "ATTAGG": 340, + "ATTTAA": 341, + "ATTTAT": 342, + "ATTTAC": 343, + "ATTTAG": 344, + "ATTTTA": 345, + "ATTTTT": 346, + "ATTTTC": 347, + "ATTTTG": 348, + "ATTTCA": 349, + "ATTTCT": 350, + "ATTTCC": 351, + "ATTTCG": 352, + "ATTTGA": 353, + "ATTTGT": 354, + "ATTTGC": 355, + "ATTTGG": 356, + "ATTCAA": 357, + "ATTCAT": 358, + "ATTCAC": 359, + "ATTCAG": 360, + "ATTCTA": 361, + "ATTCTT": 362, + "ATTCTC": 363, + "ATTCTG": 364, + "ATTCCA": 365, + "ATTCCT": 366, + "ATTCCC": 367, + "ATTCCG": 368, + "ATTCGA": 369, + "ATTCGT": 370, + "ATTCGC": 371, + "ATTCGG": 372, + "ATTGAA": 373, + "ATTGAT": 374, + "ATTGAC": 375, + "ATTGAG": 376, + "ATTGTA": 377, + "ATTGTT": 378, + "ATTGTC": 379, + "ATTGTG": 380, + "ATTGCA": 381, + "ATTGCT": 382, + "ATTGCC": 383, + "ATTGCG": 384, + "ATTGGA": 385, + "ATTGGT": 386, + "ATTGGC": 387, + "ATTGGG": 388, + "ATCAAA": 389, + "ATCAAT": 390, + "ATCAAC": 391, + "ATCAAG": 392, + "ATCATA": 393, + "ATCATT": 394, + "ATCATC": 395, + "ATCATG": 396, + "ATCACA": 397, + "ATCACT": 398, + "ATCACC": 399, + "ATCACG": 400, + "ATCAGA": 401, + "ATCAGT": 402, + "ATCAGC": 403, + "ATCAGG": 404, + "ATCTAA": 405, + "ATCTAT": 406, + "ATCTAC": 407, + "ATCTAG": 408, + "ATCTTA": 409, + "ATCTTT": 410, + "ATCTTC": 411, + "ATCTTG": 412, + "ATCTCA": 413, + "ATCTCT": 414, + "ATCTCC": 415, + "ATCTCG": 416, + "ATCTGA": 417, + "ATCTGT": 418, + "ATCTGC": 419, + "ATCTGG": 420, + "ATCCAA": 421, + "ATCCAT": 422, + "ATCCAC": 423, + "ATCCAG": 424, + "ATCCTA": 425, + "ATCCTT": 426, + "ATCCTC": 427, + "ATCCTG": 428, + "ATCCCA": 429, + "ATCCCT": 430, + "ATCCCC": 431, + "ATCCCG": 432, + "ATCCGA": 433, + "ATCCGT": 434, + "ATCCGC": 435, + "ATCCGG": 436, + "ATCGAA": 437, + "ATCGAT": 438, + "ATCGAC": 439, + "ATCGAG": 440, + "ATCGTA": 441, + "ATCGTT": 442, + "ATCGTC": 443, + "ATCGTG": 444, + "ATCGCA": 445, + "ATCGCT": 446, + "ATCGCC": 447, + "ATCGCG": 448, + "ATCGGA": 449, + "ATCGGT": 450, + "ATCGGC": 451, + "ATCGGG": 452, + "ATGAAA": 453, + "ATGAAT": 454, + "ATGAAC": 455, + "ATGAAG": 456, + "ATGATA": 457, + "ATGATT": 458, + "ATGATC": 459, + "ATGATG": 460, + "ATGACA": 461, + "ATGACT": 462, + "ATGACC": 463, + "ATGACG": 464, + "ATGAGA": 465, + "ATGAGT": 466, + "ATGAGC": 467, + "ATGAGG": 468, + "ATGTAA": 469, + "ATGTAT": 470, + "ATGTAC": 471, + "ATGTAG": 472, + "ATGTTA": 473, + "ATGTTT": 474, + "ATGTTC": 475, + "ATGTTG": 476, + "ATGTCA": 477, + "ATGTCT": 478, + "ATGTCC": 479, + "ATGTCG": 480, + "ATGTGA": 481, + "ATGTGT": 482, + "ATGTGC": 483, + "ATGTGG": 484, + "ATGCAA": 485, + "ATGCAT": 486, + "ATGCAC": 487, + "ATGCAG": 488, + "ATGCTA": 489, + "ATGCTT": 490, + "ATGCTC": 491, + "ATGCTG": 492, + "ATGCCA": 493, + "ATGCCT": 494, + "ATGCCC": 495, + "ATGCCG": 496, + "ATGCGA": 497, + "ATGCGT": 498, + "ATGCGC": 499, + "ATGCGG": 500, + "ATGGAA": 501, + "ATGGAT": 502, + "ATGGAC": 503, + "ATGGAG": 504, + "ATGGTA": 505, + "ATGGTT": 506, + "ATGGTC": 507, + "ATGGTG": 508, + "ATGGCA": 509, + "ATGGCT": 510, + "ATGGCC": 511, + "ATGGCG": 512, + "ATGGGA": 513, + "ATGGGT": 514, + "ATGGGC": 515, + "ATGGGG": 516, + "ACAAAA": 517, + "ACAAAT": 518, + "ACAAAC": 519, + "ACAAAG": 520, + "ACAATA": 521, + "ACAATT": 522, + "ACAATC": 523, + "ACAATG": 524, + "ACAACA": 525, + "ACAACT": 526, + "ACAACC": 527, + "ACAACG": 528, + "ACAAGA": 529, + "ACAAGT": 530, + "ACAAGC": 531, + "ACAAGG": 532, + "ACATAA": 533, + "ACATAT": 534, + "ACATAC": 535, + "ACATAG": 536, + "ACATTA": 537, + "ACATTT": 538, + "ACATTC": 539, + "ACATTG": 540, + "ACATCA": 541, + "ACATCT": 542, + "ACATCC": 543, + "ACATCG": 544, + "ACATGA": 545, + "ACATGT": 546, + "ACATGC": 547, + "ACATGG": 548, + "ACACAA": 549, + "ACACAT": 550, + "ACACAC": 551, + "ACACAG": 552, + "ACACTA": 553, + "ACACTT": 554, + "ACACTC": 555, + "ACACTG": 556, + "ACACCA": 557, + "ACACCT": 558, + "ACACCC": 559, + "ACACCG": 560, + "ACACGA": 561, + "ACACGT": 562, + "ACACGC": 563, + "ACACGG": 564, + "ACAGAA": 565, + "ACAGAT": 566, + "ACAGAC": 567, + "ACAGAG": 568, + "ACAGTA": 569, + "ACAGTT": 570, + "ACAGTC": 571, + "ACAGTG": 572, + "ACAGCA": 573, + "ACAGCT": 574, + "ACAGCC": 575, + "ACAGCG": 576, + "ACAGGA": 577, + "ACAGGT": 578, + "ACAGGC": 579, + "ACAGGG": 580, + "ACTAAA": 581, + "ACTAAT": 582, + "ACTAAC": 583, + "ACTAAG": 584, + "ACTATA": 585, + "ACTATT": 586, + "ACTATC": 587, + "ACTATG": 588, + "ACTACA": 589, + "ACTACT": 590, + "ACTACC": 591, + "ACTACG": 592, + "ACTAGA": 593, + "ACTAGT": 594, + "ACTAGC": 595, + "ACTAGG": 596, + "ACTTAA": 597, + "ACTTAT": 598, + "ACTTAC": 599, + "ACTTAG": 600, + "ACTTTA": 601, + "ACTTTT": 602, + "ACTTTC": 603, + "ACTTTG": 604, + "ACTTCA": 605, + "ACTTCT": 606, + "ACTTCC": 607, + "ACTTCG": 608, + "ACTTGA": 609, + "ACTTGT": 610, + "ACTTGC": 611, + "ACTTGG": 612, + "ACTCAA": 613, + "ACTCAT": 614, + "ACTCAC": 615, + "ACTCAG": 616, + "ACTCTA": 617, + "ACTCTT": 618, + "ACTCTC": 619, + "ACTCTG": 620, + "ACTCCA": 621, + "ACTCCT": 622, + "ACTCCC": 623, + "ACTCCG": 624, + "ACTCGA": 625, + "ACTCGT": 626, + "ACTCGC": 627, + "ACTCGG": 628, + "ACTGAA": 629, + "ACTGAT": 630, + "ACTGAC": 631, + "ACTGAG": 632, + "ACTGTA": 633, + "ACTGTT": 634, + "ACTGTC": 635, + "ACTGTG": 636, + "ACTGCA": 637, + "ACTGCT": 638, + "ACTGCC": 639, + "ACTGCG": 640, + "ACTGGA": 641, + "ACTGGT": 642, + "ACTGGC": 643, + "ACTGGG": 644, + "ACCAAA": 645, + "ACCAAT": 646, + "ACCAAC": 647, + "ACCAAG": 648, + "ACCATA": 649, + "ACCATT": 650, + "ACCATC": 651, + "ACCATG": 652, + "ACCACA": 653, + "ACCACT": 654, + "ACCACC": 655, + "ACCACG": 656, + "ACCAGA": 657, + "ACCAGT": 658, + "ACCAGC": 659, + "ACCAGG": 660, + "ACCTAA": 661, + "ACCTAT": 662, + "ACCTAC": 663, + "ACCTAG": 664, + "ACCTTA": 665, + "ACCTTT": 666, + "ACCTTC": 667, + "ACCTTG": 668, + "ACCTCA": 669, + "ACCTCT": 670, + "ACCTCC": 671, + "ACCTCG": 672, + "ACCTGA": 673, + "ACCTGT": 674, + "ACCTGC": 675, + "ACCTGG": 676, + "ACCCAA": 677, + "ACCCAT": 678, + "ACCCAC": 679, + "ACCCAG": 680, + "ACCCTA": 681, + "ACCCTT": 682, + "ACCCTC": 683, + "ACCCTG": 684, + "ACCCCA": 685, + "ACCCCT": 686, + "ACCCCC": 687, + "ACCCCG": 688, + "ACCCGA": 689, + "ACCCGT": 690, + "ACCCGC": 691, + "ACCCGG": 692, + "ACCGAA": 693, + "ACCGAT": 694, + "ACCGAC": 695, + "ACCGAG": 696, + "ACCGTA": 697, + "ACCGTT": 698, + "ACCGTC": 699, + "ACCGTG": 700, + "ACCGCA": 701, + "ACCGCT": 702, + "ACCGCC": 703, + "ACCGCG": 704, + "ACCGGA": 705, + "ACCGGT": 706, + "ACCGGC": 707, + "ACCGGG": 708, + "ACGAAA": 709, + "ACGAAT": 710, + "ACGAAC": 711, + "ACGAAG": 712, + "ACGATA": 713, + "ACGATT": 714, + "ACGATC": 715, + "ACGATG": 716, + "ACGACA": 717, + "ACGACT": 718, + "ACGACC": 719, + "ACGACG": 720, + "ACGAGA": 721, + "ACGAGT": 722, + "ACGAGC": 723, + "ACGAGG": 724, + "ACGTAA": 725, + "ACGTAT": 726, + "ACGTAC": 727, + "ACGTAG": 728, + "ACGTTA": 729, + "ACGTTT": 730, + "ACGTTC": 731, + "ACGTTG": 732, + "ACGTCA": 733, + "ACGTCT": 734, + "ACGTCC": 735, + "ACGTCG": 736, + "ACGTGA": 737, + "ACGTGT": 738, + "ACGTGC": 739, + "ACGTGG": 740, + "ACGCAA": 741, + "ACGCAT": 742, + "ACGCAC": 743, + "ACGCAG": 744, + "ACGCTA": 745, + "ACGCTT": 746, + "ACGCTC": 747, + "ACGCTG": 748, + "ACGCCA": 749, + "ACGCCT": 750, + "ACGCCC": 751, + "ACGCCG": 752, + "ACGCGA": 753, + "ACGCGT": 754, + "ACGCGC": 755, + "ACGCGG": 756, + "ACGGAA": 757, + "ACGGAT": 758, + "ACGGAC": 759, + "ACGGAG": 760, + "ACGGTA": 761, + "ACGGTT": 762, + "ACGGTC": 763, + "ACGGTG": 764, + "ACGGCA": 765, + "ACGGCT": 766, + "ACGGCC": 767, + "ACGGCG": 768, + "ACGGGA": 769, + "ACGGGT": 770, + "ACGGGC": 771, + "ACGGGG": 772, + "AGAAAA": 773, + "AGAAAT": 774, + "AGAAAC": 775, + "AGAAAG": 776, + "AGAATA": 777, + "AGAATT": 778, + "AGAATC": 779, + "AGAATG": 780, + "AGAACA": 781, + "AGAACT": 782, + "AGAACC": 783, + "AGAACG": 784, + "AGAAGA": 785, + "AGAAGT": 786, + "AGAAGC": 787, + "AGAAGG": 788, + "AGATAA": 789, + "AGATAT": 790, + "AGATAC": 791, + "AGATAG": 792, + "AGATTA": 793, + "AGATTT": 794, + "AGATTC": 795, + "AGATTG": 796, + "AGATCA": 797, + "AGATCT": 798, + "AGATCC": 799, + "AGATCG": 800, + "AGATGA": 801, + "AGATGT": 802, + "AGATGC": 803, + "AGATGG": 804, + "AGACAA": 805, + "AGACAT": 806, + "AGACAC": 807, + "AGACAG": 808, + "AGACTA": 809, + "AGACTT": 810, + "AGACTC": 811, + "AGACTG": 812, + "AGACCA": 813, + "AGACCT": 814, + "AGACCC": 815, + "AGACCG": 816, + "AGACGA": 817, + "AGACGT": 818, + "AGACGC": 819, + "AGACGG": 820, + "AGAGAA": 821, + "AGAGAT": 822, + "AGAGAC": 823, + "AGAGAG": 824, + "AGAGTA": 825, + "AGAGTT": 826, + "AGAGTC": 827, + "AGAGTG": 828, + "AGAGCA": 829, + "AGAGCT": 830, + "AGAGCC": 831, + "AGAGCG": 832, + "AGAGGA": 833, + "AGAGGT": 834, + "AGAGGC": 835, + "AGAGGG": 836, + "AGTAAA": 837, + "AGTAAT": 838, + "AGTAAC": 839, + "AGTAAG": 840, + "AGTATA": 841, + "AGTATT": 842, + "AGTATC": 843, + "AGTATG": 844, + "AGTACA": 845, + "AGTACT": 846, + "AGTACC": 847, + "AGTACG": 848, + "AGTAGA": 849, + "AGTAGT": 850, + "AGTAGC": 851, + "AGTAGG": 852, + "AGTTAA": 853, + "AGTTAT": 854, + "AGTTAC": 855, + "AGTTAG": 856, + "AGTTTA": 857, + "AGTTTT": 858, + "AGTTTC": 859, + "AGTTTG": 860, + "AGTTCA": 861, + "AGTTCT": 862, + "AGTTCC": 863, + "AGTTCG": 864, + "AGTTGA": 865, + "AGTTGT": 866, + "AGTTGC": 867, + "AGTTGG": 868, + "AGTCAA": 869, + "AGTCAT": 870, + "AGTCAC": 871, + "AGTCAG": 872, + "AGTCTA": 873, + "AGTCTT": 874, + "AGTCTC": 875, + "AGTCTG": 876, + "AGTCCA": 877, + "AGTCCT": 878, + "AGTCCC": 879, + "AGTCCG": 880, + "AGTCGA": 881, + "AGTCGT": 882, + "AGTCGC": 883, + "AGTCGG": 884, + "AGTGAA": 885, + "AGTGAT": 886, + "AGTGAC": 887, + "AGTGAG": 888, + "AGTGTA": 889, + "AGTGTT": 890, + "AGTGTC": 891, + "AGTGTG": 892, + "AGTGCA": 893, + "AGTGCT": 894, + "AGTGCC": 895, + "AGTGCG": 896, + "AGTGGA": 897, + "AGTGGT": 898, + "AGTGGC": 899, + "AGTGGG": 900, + "AGCAAA": 901, + "AGCAAT": 902, + "AGCAAC": 903, + "AGCAAG": 904, + "AGCATA": 905, + "AGCATT": 906, + "AGCATC": 907, + "AGCATG": 908, + "AGCACA": 909, + "AGCACT": 910, + "AGCACC": 911, + "AGCACG": 912, + "AGCAGA": 913, + "AGCAGT": 914, + "AGCAGC": 915, + "AGCAGG": 916, + "AGCTAA": 917, + "AGCTAT": 918, + "AGCTAC": 919, + "AGCTAG": 920, + "AGCTTA": 921, + "AGCTTT": 922, + "AGCTTC": 923, + "AGCTTG": 924, + "AGCTCA": 925, + "AGCTCT": 926, + "AGCTCC": 927, + "AGCTCG": 928, + "AGCTGA": 929, + "AGCTGT": 930, + "AGCTGC": 931, + "AGCTGG": 932, + "AGCCAA": 933, + "AGCCAT": 934, + "AGCCAC": 935, + "AGCCAG": 936, + "AGCCTA": 937, + "AGCCTT": 938, + "AGCCTC": 939, + "AGCCTG": 940, + "AGCCCA": 941, + "AGCCCT": 942, + "AGCCCC": 943, + "AGCCCG": 944, + "AGCCGA": 945, + "AGCCGT": 946, + "AGCCGC": 947, + "AGCCGG": 948, + "AGCGAA": 949, + "AGCGAT": 950, + "AGCGAC": 951, + "AGCGAG": 952, + "AGCGTA": 953, + "AGCGTT": 954, + "AGCGTC": 955, + "AGCGTG": 956, + "AGCGCA": 957, + "AGCGCT": 958, + "AGCGCC": 959, + "AGCGCG": 960, + "AGCGGA": 961, + "AGCGGT": 962, + "AGCGGC": 963, + "AGCGGG": 964, + "AGGAAA": 965, + "AGGAAT": 966, + "AGGAAC": 967, + "AGGAAG": 968, + "AGGATA": 969, + "AGGATT": 970, + "AGGATC": 971, + "AGGATG": 972, + "AGGACA": 973, + "AGGACT": 974, + "AGGACC": 975, + "AGGACG": 976, + "AGGAGA": 977, + "AGGAGT": 978, + "AGGAGC": 979, + "AGGAGG": 980, + "AGGTAA": 981, + "AGGTAT": 982, + "AGGTAC": 983, + "AGGTAG": 984, + "AGGTTA": 985, + "AGGTTT": 986, + "AGGTTC": 987, + "AGGTTG": 988, + "AGGTCA": 989, + "AGGTCT": 990, + "AGGTCC": 991, + "AGGTCG": 992, + "AGGTGA": 993, + "AGGTGT": 994, + "AGGTGC": 995, + "AGGTGG": 996, + "AGGCAA": 997, + "AGGCAT": 998, + "AGGCAC": 999, + "AGGCAG": 1000, + "AGGCTA": 1001, + "AGGCTT": 1002, + "AGGCTC": 1003, + "AGGCTG": 1004, + "AGGCCA": 1005, + "AGGCCT": 1006, + "AGGCCC": 1007, + "AGGCCG": 1008, + "AGGCGA": 1009, + "AGGCGT": 1010, + "AGGCGC": 1011, + "AGGCGG": 1012, + "AGGGAA": 1013, + "AGGGAT": 1014, + "AGGGAC": 1015, + "AGGGAG": 1016, + "AGGGTA": 1017, + "AGGGTT": 1018, + "AGGGTC": 1019, + "AGGGTG": 1020, + "AGGGCA": 1021, + "AGGGCT": 1022, + "AGGGCC": 1023, + "AGGGCG": 1024, + "AGGGGA": 1025, + "AGGGGT": 1026, + "AGGGGC": 1027, + "AGGGGG": 1028, + "TAAAAA": 1029, + "TAAAAT": 1030, + "TAAAAC": 1031, + "TAAAAG": 1032, + "TAAATA": 1033, + "TAAATT": 1034, + "TAAATC": 1035, + "TAAATG": 1036, + "TAAACA": 1037, + "TAAACT": 1038, + "TAAACC": 1039, + "TAAACG": 1040, + "TAAAGA": 1041, + "TAAAGT": 1042, + "TAAAGC": 1043, + "TAAAGG": 1044, + "TAATAA": 1045, + "TAATAT": 1046, + "TAATAC": 1047, + "TAATAG": 1048, + "TAATTA": 1049, + "TAATTT": 1050, + "TAATTC": 1051, + "TAATTG": 1052, + "TAATCA": 1053, + "TAATCT": 1054, + "TAATCC": 1055, + "TAATCG": 1056, + "TAATGA": 1057, + "TAATGT": 1058, + "TAATGC": 1059, + "TAATGG": 1060, + "TAACAA": 1061, + "TAACAT": 1062, + "TAACAC": 1063, + "TAACAG": 1064, + "TAACTA": 1065, + "TAACTT": 1066, + "TAACTC": 1067, + "TAACTG": 1068, + "TAACCA": 1069, + "TAACCT": 1070, + "TAACCC": 1071, + "TAACCG": 1072, + "TAACGA": 1073, + "TAACGT": 1074, + "TAACGC": 1075, + "TAACGG": 1076, + "TAAGAA": 1077, + "TAAGAT": 1078, + "TAAGAC": 1079, + "TAAGAG": 1080, + "TAAGTA": 1081, + "TAAGTT": 1082, + "TAAGTC": 1083, + "TAAGTG": 1084, + "TAAGCA": 1085, + "TAAGCT": 1086, + "TAAGCC": 1087, + "TAAGCG": 1088, + "TAAGGA": 1089, + "TAAGGT": 1090, + "TAAGGC": 1091, + "TAAGGG": 1092, + "TATAAA": 1093, + "TATAAT": 1094, + "TATAAC": 1095, + "TATAAG": 1096, + "TATATA": 1097, + "TATATT": 1098, + "TATATC": 1099, + "TATATG": 1100, + "TATACA": 1101, + "TATACT": 1102, + "TATACC": 1103, + "TATACG": 1104, + "TATAGA": 1105, + "TATAGT": 1106, + "TATAGC": 1107, + "TATAGG": 1108, + "TATTAA": 1109, + "TATTAT": 1110, + "TATTAC": 1111, + "TATTAG": 1112, + "TATTTA": 1113, + "TATTTT": 1114, + "TATTTC": 1115, + "TATTTG": 1116, + "TATTCA": 1117, + "TATTCT": 1118, + "TATTCC": 1119, + "TATTCG": 1120, + "TATTGA": 1121, + "TATTGT": 1122, + "TATTGC": 1123, + "TATTGG": 1124, + "TATCAA": 1125, + "TATCAT": 1126, + "TATCAC": 1127, + "TATCAG": 1128, + "TATCTA": 1129, + "TATCTT": 1130, + "TATCTC": 1131, + "TATCTG": 1132, + "TATCCA": 1133, + "TATCCT": 1134, + "TATCCC": 1135, + "TATCCG": 1136, + "TATCGA": 1137, + "TATCGT": 1138, + "TATCGC": 1139, + "TATCGG": 1140, + "TATGAA": 1141, + "TATGAT": 1142, + "TATGAC": 1143, + "TATGAG": 1144, + "TATGTA": 1145, + "TATGTT": 1146, + "TATGTC": 1147, + "TATGTG": 1148, + "TATGCA": 1149, + "TATGCT": 1150, + "TATGCC": 1151, + "TATGCG": 1152, + "TATGGA": 1153, + "TATGGT": 1154, + "TATGGC": 1155, + "TATGGG": 1156, + "TACAAA": 1157, + "TACAAT": 1158, + "TACAAC": 1159, + "TACAAG": 1160, + "TACATA": 1161, + "TACATT": 1162, + "TACATC": 1163, + "TACATG": 1164, + "TACACA": 1165, + "TACACT": 1166, + "TACACC": 1167, + "TACACG": 1168, + "TACAGA": 1169, + "TACAGT": 1170, + "TACAGC": 1171, + "TACAGG": 1172, + "TACTAA": 1173, + "TACTAT": 1174, + "TACTAC": 1175, + "TACTAG": 1176, + "TACTTA": 1177, + "TACTTT": 1178, + "TACTTC": 1179, + "TACTTG": 1180, + "TACTCA": 1181, + "TACTCT": 1182, + "TACTCC": 1183, + "TACTCG": 1184, + "TACTGA": 1185, + "TACTGT": 1186, + "TACTGC": 1187, + "TACTGG": 1188, + "TACCAA": 1189, + "TACCAT": 1190, + "TACCAC": 1191, + "TACCAG": 1192, + "TACCTA": 1193, + "TACCTT": 1194, + "TACCTC": 1195, + "TACCTG": 1196, + "TACCCA": 1197, + "TACCCT": 1198, + "TACCCC": 1199, + "TACCCG": 1200, + "TACCGA": 1201, + "TACCGT": 1202, + "TACCGC": 1203, + "TACCGG": 1204, + "TACGAA": 1205, + "TACGAT": 1206, + "TACGAC": 1207, + "TACGAG": 1208, + "TACGTA": 1209, + "TACGTT": 1210, + "TACGTC": 1211, + "TACGTG": 1212, + "TACGCA": 1213, + "TACGCT": 1214, + "TACGCC": 1215, + "TACGCG": 1216, + "TACGGA": 1217, + "TACGGT": 1218, + "TACGGC": 1219, + "TACGGG": 1220, + "TAGAAA": 1221, + "TAGAAT": 1222, + "TAGAAC": 1223, + "TAGAAG": 1224, + "TAGATA": 1225, + "TAGATT": 1226, + "TAGATC": 1227, + "TAGATG": 1228, + "TAGACA": 1229, + "TAGACT": 1230, + "TAGACC": 1231, + "TAGACG": 1232, + "TAGAGA": 1233, + "TAGAGT": 1234, + "TAGAGC": 1235, + "TAGAGG": 1236, + "TAGTAA": 1237, + "TAGTAT": 1238, + "TAGTAC": 1239, + "TAGTAG": 1240, + "TAGTTA": 1241, + "TAGTTT": 1242, + "TAGTTC": 1243, + "TAGTTG": 1244, + "TAGTCA": 1245, + "TAGTCT": 1246, + "TAGTCC": 1247, + "TAGTCG": 1248, + "TAGTGA": 1249, + "TAGTGT": 1250, + "TAGTGC": 1251, + "TAGTGG": 1252, + "TAGCAA": 1253, + "TAGCAT": 1254, + "TAGCAC": 1255, + "TAGCAG": 1256, + "TAGCTA": 1257, + "TAGCTT": 1258, + "TAGCTC": 1259, + "TAGCTG": 1260, + "TAGCCA": 1261, + "TAGCCT": 1262, + "TAGCCC": 1263, + "TAGCCG": 1264, + "TAGCGA": 1265, + "TAGCGT": 1266, + "TAGCGC": 1267, + "TAGCGG": 1268, + "TAGGAA": 1269, + "TAGGAT": 1270, + "TAGGAC": 1271, + "TAGGAG": 1272, + "TAGGTA": 1273, + "TAGGTT": 1274, + "TAGGTC": 1275, + "TAGGTG": 1276, + "TAGGCA": 1277, + "TAGGCT": 1278, + "TAGGCC": 1279, + "TAGGCG": 1280, + "TAGGGA": 1281, + "TAGGGT": 1282, + "TAGGGC": 1283, + "TAGGGG": 1284, + "TTAAAA": 1285, + "TTAAAT": 1286, + "TTAAAC": 1287, + "TTAAAG": 1288, + "TTAATA": 1289, + "TTAATT": 1290, + "TTAATC": 1291, + "TTAATG": 1292, + "TTAACA": 1293, + "TTAACT": 1294, + "TTAACC": 1295, + "TTAACG": 1296, + "TTAAGA": 1297, + "TTAAGT": 1298, + "TTAAGC": 1299, + "TTAAGG": 1300, + "TTATAA": 1301, + "TTATAT": 1302, + "TTATAC": 1303, + "TTATAG": 1304, + "TTATTA": 1305, + "TTATTT": 1306, + "TTATTC": 1307, + "TTATTG": 1308, + "TTATCA": 1309, + "TTATCT": 1310, + "TTATCC": 1311, + "TTATCG": 1312, + "TTATGA": 1313, + "TTATGT": 1314, + "TTATGC": 1315, + "TTATGG": 1316, + "TTACAA": 1317, + "TTACAT": 1318, + "TTACAC": 1319, + "TTACAG": 1320, + "TTACTA": 1321, + "TTACTT": 1322, + "TTACTC": 1323, + "TTACTG": 1324, + "TTACCA": 1325, + "TTACCT": 1326, + "TTACCC": 1327, + "TTACCG": 1328, + "TTACGA": 1329, + "TTACGT": 1330, + "TTACGC": 1331, + "TTACGG": 1332, + "TTAGAA": 1333, + "TTAGAT": 1334, + "TTAGAC": 1335, + "TTAGAG": 1336, + "TTAGTA": 1337, + "TTAGTT": 1338, + "TTAGTC": 1339, + "TTAGTG": 1340, + "TTAGCA": 1341, + "TTAGCT": 1342, + "TTAGCC": 1343, + "TTAGCG": 1344, + "TTAGGA": 1345, + "TTAGGT": 1346, + "TTAGGC": 1347, + "TTAGGG": 1348, + "TTTAAA": 1349, + "TTTAAT": 1350, + "TTTAAC": 1351, + "TTTAAG": 1352, + "TTTATA": 1353, + "TTTATT": 1354, + "TTTATC": 1355, + "TTTATG": 1356, + "TTTACA": 1357, + "TTTACT": 1358, + "TTTACC": 1359, + "TTTACG": 1360, + "TTTAGA": 1361, + "TTTAGT": 1362, + "TTTAGC": 1363, + "TTTAGG": 1364, + "TTTTAA": 1365, + "TTTTAT": 1366, + "TTTTAC": 1367, + "TTTTAG": 1368, + "TTTTTA": 1369, + "TTTTTT": 1370, + "TTTTTC": 1371, + "TTTTTG": 1372, + "TTTTCA": 1373, + "TTTTCT": 1374, + "TTTTCC": 1375, + "TTTTCG": 1376, + "TTTTGA": 1377, + "TTTTGT": 1378, + "TTTTGC": 1379, + "TTTTGG": 1380, + "TTTCAA": 1381, + "TTTCAT": 1382, + "TTTCAC": 1383, + "TTTCAG": 1384, + "TTTCTA": 1385, + "TTTCTT": 1386, + "TTTCTC": 1387, + "TTTCTG": 1388, + "TTTCCA": 1389, + "TTTCCT": 1390, + "TTTCCC": 1391, + "TTTCCG": 1392, + "TTTCGA": 1393, + "TTTCGT": 1394, + "TTTCGC": 1395, + "TTTCGG": 1396, + "TTTGAA": 1397, + "TTTGAT": 1398, + "TTTGAC": 1399, + "TTTGAG": 1400, + "TTTGTA": 1401, + "TTTGTT": 1402, + "TTTGTC": 1403, + "TTTGTG": 1404, + "TTTGCA": 1405, + "TTTGCT": 1406, + "TTTGCC": 1407, + "TTTGCG": 1408, + "TTTGGA": 1409, + "TTTGGT": 1410, + "TTTGGC": 1411, + "TTTGGG": 1412, + "TTCAAA": 1413, + "TTCAAT": 1414, + "TTCAAC": 1415, + "TTCAAG": 1416, + "TTCATA": 1417, + "TTCATT": 1418, + "TTCATC": 1419, + "TTCATG": 1420, + "TTCACA": 1421, + "TTCACT": 1422, + "TTCACC": 1423, + "TTCACG": 1424, + "TTCAGA": 1425, + "TTCAGT": 1426, + "TTCAGC": 1427, + "TTCAGG": 1428, + "TTCTAA": 1429, + "TTCTAT": 1430, + "TTCTAC": 1431, + "TTCTAG": 1432, + "TTCTTA": 1433, + "TTCTTT": 1434, + "TTCTTC": 1435, + "TTCTTG": 1436, + "TTCTCA": 1437, + "TTCTCT": 1438, + "TTCTCC": 1439, + "TTCTCG": 1440, + "TTCTGA": 1441, + "TTCTGT": 1442, + "TTCTGC": 1443, + "TTCTGG": 1444, + "TTCCAA": 1445, + "TTCCAT": 1446, + "TTCCAC": 1447, + "TTCCAG": 1448, + "TTCCTA": 1449, + "TTCCTT": 1450, + "TTCCTC": 1451, + "TTCCTG": 1452, + "TTCCCA": 1453, + "TTCCCT": 1454, + "TTCCCC": 1455, + "TTCCCG": 1456, + "TTCCGA": 1457, + "TTCCGT": 1458, + "TTCCGC": 1459, + "TTCCGG": 1460, + "TTCGAA": 1461, + "TTCGAT": 1462, + "TTCGAC": 1463, + "TTCGAG": 1464, + "TTCGTA": 1465, + "TTCGTT": 1466, + "TTCGTC": 1467, + "TTCGTG": 1468, + "TTCGCA": 1469, + "TTCGCT": 1470, + "TTCGCC": 1471, + "TTCGCG": 1472, + "TTCGGA": 1473, + "TTCGGT": 1474, + "TTCGGC": 1475, + "TTCGGG": 1476, + "TTGAAA": 1477, + "TTGAAT": 1478, + "TTGAAC": 1479, + "TTGAAG": 1480, + "TTGATA": 1481, + "TTGATT": 1482, + "TTGATC": 1483, + "TTGATG": 1484, + "TTGACA": 1485, + "TTGACT": 1486, + "TTGACC": 1487, + "TTGACG": 1488, + "TTGAGA": 1489, + "TTGAGT": 1490, + "TTGAGC": 1491, + "TTGAGG": 1492, + "TTGTAA": 1493, + "TTGTAT": 1494, + "TTGTAC": 1495, + "TTGTAG": 1496, + "TTGTTA": 1497, + "TTGTTT": 1498, + "TTGTTC": 1499, + "TTGTTG": 1500, + "TTGTCA": 1501, + "TTGTCT": 1502, + "TTGTCC": 1503, + "TTGTCG": 1504, + "TTGTGA": 1505, + "TTGTGT": 1506, + "TTGTGC": 1507, + "TTGTGG": 1508, + "TTGCAA": 1509, + "TTGCAT": 1510, + "TTGCAC": 1511, + "TTGCAG": 1512, + "TTGCTA": 1513, + "TTGCTT": 1514, + "TTGCTC": 1515, + "TTGCTG": 1516, + "TTGCCA": 1517, + "TTGCCT": 1518, + "TTGCCC": 1519, + "TTGCCG": 1520, + "TTGCGA": 1521, + "TTGCGT": 1522, + "TTGCGC": 1523, + "TTGCGG": 1524, + "TTGGAA": 1525, + "TTGGAT": 1526, + "TTGGAC": 1527, + "TTGGAG": 1528, + "TTGGTA": 1529, + "TTGGTT": 1530, + "TTGGTC": 1531, + "TTGGTG": 1532, + "TTGGCA": 1533, + "TTGGCT": 1534, + "TTGGCC": 1535, + "TTGGCG": 1536, + "TTGGGA": 1537, + "TTGGGT": 1538, + "TTGGGC": 1539, + "TTGGGG": 1540, + "TCAAAA": 1541, + "TCAAAT": 1542, + "TCAAAC": 1543, + "TCAAAG": 1544, + "TCAATA": 1545, + "TCAATT": 1546, + "TCAATC": 1547, + "TCAATG": 1548, + "TCAACA": 1549, + "TCAACT": 1550, + "TCAACC": 1551, + "TCAACG": 1552, + "TCAAGA": 1553, + "TCAAGT": 1554, + "TCAAGC": 1555, + "TCAAGG": 1556, + "TCATAA": 1557, + "TCATAT": 1558, + "TCATAC": 1559, + "TCATAG": 1560, + "TCATTA": 1561, + "TCATTT": 1562, + "TCATTC": 1563, + "TCATTG": 1564, + "TCATCA": 1565, + "TCATCT": 1566, + "TCATCC": 1567, + "TCATCG": 1568, + "TCATGA": 1569, + "TCATGT": 1570, + "TCATGC": 1571, + "TCATGG": 1572, + "TCACAA": 1573, + "TCACAT": 1574, + "TCACAC": 1575, + "TCACAG": 1576, + "TCACTA": 1577, + "TCACTT": 1578, + "TCACTC": 1579, + "TCACTG": 1580, + "TCACCA": 1581, + "TCACCT": 1582, + "TCACCC": 1583, + "TCACCG": 1584, + "TCACGA": 1585, + "TCACGT": 1586, + "TCACGC": 1587, + "TCACGG": 1588, + "TCAGAA": 1589, + "TCAGAT": 1590, + "TCAGAC": 1591, + "TCAGAG": 1592, + "TCAGTA": 1593, + "TCAGTT": 1594, + "TCAGTC": 1595, + "TCAGTG": 1596, + "TCAGCA": 1597, + "TCAGCT": 1598, + "TCAGCC": 1599, + "TCAGCG": 1600, + "TCAGGA": 1601, + "TCAGGT": 1602, + "TCAGGC": 1603, + "TCAGGG": 1604, + "TCTAAA": 1605, + "TCTAAT": 1606, + "TCTAAC": 1607, + "TCTAAG": 1608, + "TCTATA": 1609, + "TCTATT": 1610, + "TCTATC": 1611, + "TCTATG": 1612, + "TCTACA": 1613, + "TCTACT": 1614, + "TCTACC": 1615, + "TCTACG": 1616, + "TCTAGA": 1617, + "TCTAGT": 1618, + "TCTAGC": 1619, + "TCTAGG": 1620, + "TCTTAA": 1621, + "TCTTAT": 1622, + "TCTTAC": 1623, + "TCTTAG": 1624, + "TCTTTA": 1625, + "TCTTTT": 1626, + "TCTTTC": 1627, + "TCTTTG": 1628, + "TCTTCA": 1629, + "TCTTCT": 1630, + "TCTTCC": 1631, + "TCTTCG": 1632, + "TCTTGA": 1633, + "TCTTGT": 1634, + "TCTTGC": 1635, + "TCTTGG": 1636, + "TCTCAA": 1637, + "TCTCAT": 1638, + "TCTCAC": 1639, + "TCTCAG": 1640, + "TCTCTA": 1641, + "TCTCTT": 1642, + "TCTCTC": 1643, + "TCTCTG": 1644, + "TCTCCA": 1645, + "TCTCCT": 1646, + "TCTCCC": 1647, + "TCTCCG": 1648, + "TCTCGA": 1649, + "TCTCGT": 1650, + "TCTCGC": 1651, + "TCTCGG": 1652, + "TCTGAA": 1653, + "TCTGAT": 1654, + "TCTGAC": 1655, + "TCTGAG": 1656, + "TCTGTA": 1657, + "TCTGTT": 1658, + "TCTGTC": 1659, + "TCTGTG": 1660, + "TCTGCA": 1661, + "TCTGCT": 1662, + "TCTGCC": 1663, + "TCTGCG": 1664, + "TCTGGA": 1665, + "TCTGGT": 1666, + "TCTGGC": 1667, + "TCTGGG": 1668, + "TCCAAA": 1669, + "TCCAAT": 1670, + "TCCAAC": 1671, + "TCCAAG": 1672, + "TCCATA": 1673, + "TCCATT": 1674, + "TCCATC": 1675, + "TCCATG": 1676, + "TCCACA": 1677, + "TCCACT": 1678, + "TCCACC": 1679, + "TCCACG": 1680, + "TCCAGA": 1681, + "TCCAGT": 1682, + "TCCAGC": 1683, + "TCCAGG": 1684, + "TCCTAA": 1685, + "TCCTAT": 1686, + "TCCTAC": 1687, + "TCCTAG": 1688, + "TCCTTA": 1689, + "TCCTTT": 1690, + "TCCTTC": 1691, + "TCCTTG": 1692, + "TCCTCA": 1693, + "TCCTCT": 1694, + "TCCTCC": 1695, + "TCCTCG": 1696, + "TCCTGA": 1697, + "TCCTGT": 1698, + "TCCTGC": 1699, + "TCCTGG": 1700, + "TCCCAA": 1701, + "TCCCAT": 1702, + "TCCCAC": 1703, + "TCCCAG": 1704, + "TCCCTA": 1705, + "TCCCTT": 1706, + "TCCCTC": 1707, + "TCCCTG": 1708, + "TCCCCA": 1709, + "TCCCCT": 1710, + "TCCCCC": 1711, + "TCCCCG": 1712, + "TCCCGA": 1713, + "TCCCGT": 1714, + "TCCCGC": 1715, + "TCCCGG": 1716, + "TCCGAA": 1717, + "TCCGAT": 1718, + "TCCGAC": 1719, + "TCCGAG": 1720, + "TCCGTA": 1721, + "TCCGTT": 1722, + "TCCGTC": 1723, + "TCCGTG": 1724, + "TCCGCA": 1725, + "TCCGCT": 1726, + "TCCGCC": 1727, + "TCCGCG": 1728, + "TCCGGA": 1729, + "TCCGGT": 1730, + "TCCGGC": 1731, + "TCCGGG": 1732, + "TCGAAA": 1733, + "TCGAAT": 1734, + "TCGAAC": 1735, + "TCGAAG": 1736, + "TCGATA": 1737, + "TCGATT": 1738, + "TCGATC": 1739, + "TCGATG": 1740, + "TCGACA": 1741, + "TCGACT": 1742, + "TCGACC": 1743, + "TCGACG": 1744, + "TCGAGA": 1745, + "TCGAGT": 1746, + "TCGAGC": 1747, + "TCGAGG": 1748, + "TCGTAA": 1749, + "TCGTAT": 1750, + "TCGTAC": 1751, + "TCGTAG": 1752, + "TCGTTA": 1753, + "TCGTTT": 1754, + "TCGTTC": 1755, + "TCGTTG": 1756, + "TCGTCA": 1757, + "TCGTCT": 1758, + "TCGTCC": 1759, + "TCGTCG": 1760, + "TCGTGA": 1761, + "TCGTGT": 1762, + "TCGTGC": 1763, + "TCGTGG": 1764, + "TCGCAA": 1765, + "TCGCAT": 1766, + "TCGCAC": 1767, + "TCGCAG": 1768, + "TCGCTA": 1769, + "TCGCTT": 1770, + "TCGCTC": 1771, + "TCGCTG": 1772, + "TCGCCA": 1773, + "TCGCCT": 1774, + "TCGCCC": 1775, + "TCGCCG": 1776, + "TCGCGA": 1777, + "TCGCGT": 1778, + "TCGCGC": 1779, + "TCGCGG": 1780, + "TCGGAA": 1781, + "TCGGAT": 1782, + "TCGGAC": 1783, + "TCGGAG": 1784, + "TCGGTA": 1785, + "TCGGTT": 1786, + "TCGGTC": 1787, + "TCGGTG": 1788, + "TCGGCA": 1789, + "TCGGCT": 1790, + "TCGGCC": 1791, + "TCGGCG": 1792, + "TCGGGA": 1793, + "TCGGGT": 1794, + "TCGGGC": 1795, + "TCGGGG": 1796, + "TGAAAA": 1797, + "TGAAAT": 1798, + "TGAAAC": 1799, + "TGAAAG": 1800, + "TGAATA": 1801, + "TGAATT": 1802, + "TGAATC": 1803, + "TGAATG": 1804, + "TGAACA": 1805, + "TGAACT": 1806, + "TGAACC": 1807, + "TGAACG": 1808, + "TGAAGA": 1809, + "TGAAGT": 1810, + "TGAAGC": 1811, + "TGAAGG": 1812, + "TGATAA": 1813, + "TGATAT": 1814, + "TGATAC": 1815, + "TGATAG": 1816, + "TGATTA": 1817, + "TGATTT": 1818, + "TGATTC": 1819, + "TGATTG": 1820, + "TGATCA": 1821, + "TGATCT": 1822, + "TGATCC": 1823, + "TGATCG": 1824, + "TGATGA": 1825, + "TGATGT": 1826, + "TGATGC": 1827, + "TGATGG": 1828, + "TGACAA": 1829, + "TGACAT": 1830, + "TGACAC": 1831, + "TGACAG": 1832, + "TGACTA": 1833, + "TGACTT": 1834, + "TGACTC": 1835, + "TGACTG": 1836, + "TGACCA": 1837, + "TGACCT": 1838, + "TGACCC": 1839, + "TGACCG": 1840, + "TGACGA": 1841, + "TGACGT": 1842, + "TGACGC": 1843, + "TGACGG": 1844, + "TGAGAA": 1845, + "TGAGAT": 1846, + "TGAGAC": 1847, + "TGAGAG": 1848, + "TGAGTA": 1849, + "TGAGTT": 1850, + "TGAGTC": 1851, + "TGAGTG": 1852, + "TGAGCA": 1853, + "TGAGCT": 1854, + "TGAGCC": 1855, + "TGAGCG": 1856, + "TGAGGA": 1857, + "TGAGGT": 1858, + "TGAGGC": 1859, + "TGAGGG": 1860, + "TGTAAA": 1861, + "TGTAAT": 1862, + "TGTAAC": 1863, + "TGTAAG": 1864, + "TGTATA": 1865, + "TGTATT": 1866, + "TGTATC": 1867, + "TGTATG": 1868, + "TGTACA": 1869, + "TGTACT": 1870, + "TGTACC": 1871, + "TGTACG": 1872, + "TGTAGA": 1873, + "TGTAGT": 1874, + "TGTAGC": 1875, + "TGTAGG": 1876, + "TGTTAA": 1877, + "TGTTAT": 1878, + "TGTTAC": 1879, + "TGTTAG": 1880, + "TGTTTA": 1881, + "TGTTTT": 1882, + "TGTTTC": 1883, + "TGTTTG": 1884, + "TGTTCA": 1885, + "TGTTCT": 1886, + "TGTTCC": 1887, + "TGTTCG": 1888, + "TGTTGA": 1889, + "TGTTGT": 1890, + "TGTTGC": 1891, + "TGTTGG": 1892, + "TGTCAA": 1893, + "TGTCAT": 1894, + "TGTCAC": 1895, + "TGTCAG": 1896, + "TGTCTA": 1897, + "TGTCTT": 1898, + "TGTCTC": 1899, + "TGTCTG": 1900, + "TGTCCA": 1901, + "TGTCCT": 1902, + "TGTCCC": 1903, + "TGTCCG": 1904, + "TGTCGA": 1905, + "TGTCGT": 1906, + "TGTCGC": 1907, + "TGTCGG": 1908, + "TGTGAA": 1909, + "TGTGAT": 1910, + "TGTGAC": 1911, + "TGTGAG": 1912, + "TGTGTA": 1913, + "TGTGTT": 1914, + "TGTGTC": 1915, + "TGTGTG": 1916, + "TGTGCA": 1917, + "TGTGCT": 1918, + "TGTGCC": 1919, + "TGTGCG": 1920, + "TGTGGA": 1921, + "TGTGGT": 1922, + "TGTGGC": 1923, + "TGTGGG": 1924, + "TGCAAA": 1925, + "TGCAAT": 1926, + "TGCAAC": 1927, + "TGCAAG": 1928, + "TGCATA": 1929, + "TGCATT": 1930, + "TGCATC": 1931, + "TGCATG": 1932, + "TGCACA": 1933, + "TGCACT": 1934, + "TGCACC": 1935, + "TGCACG": 1936, + "TGCAGA": 1937, + "TGCAGT": 1938, + "TGCAGC": 1939, + "TGCAGG": 1940, + "TGCTAA": 1941, + "TGCTAT": 1942, + "TGCTAC": 1943, + "TGCTAG": 1944, + "TGCTTA": 1945, + "TGCTTT": 1946, + "TGCTTC": 1947, + "TGCTTG": 1948, + "TGCTCA": 1949, + "TGCTCT": 1950, + "TGCTCC": 1951, + "TGCTCG": 1952, + "TGCTGA": 1953, + "TGCTGT": 1954, + "TGCTGC": 1955, + "TGCTGG": 1956, + "TGCCAA": 1957, + "TGCCAT": 1958, + "TGCCAC": 1959, + "TGCCAG": 1960, + "TGCCTA": 1961, + "TGCCTT": 1962, + "TGCCTC": 1963, + "TGCCTG": 1964, + "TGCCCA": 1965, + "TGCCCT": 1966, + "TGCCCC": 1967, + "TGCCCG": 1968, + "TGCCGA": 1969, + "TGCCGT": 1970, + "TGCCGC": 1971, + "TGCCGG": 1972, + "TGCGAA": 1973, + "TGCGAT": 1974, + "TGCGAC": 1975, + "TGCGAG": 1976, + "TGCGTA": 1977, + "TGCGTT": 1978, + "TGCGTC": 1979, + "TGCGTG": 1980, + "TGCGCA": 1981, + "TGCGCT": 1982, + "TGCGCC": 1983, + "TGCGCG": 1984, + "TGCGGA": 1985, + "TGCGGT": 1986, + "TGCGGC": 1987, + "TGCGGG": 1988, + "TGGAAA": 1989, + "TGGAAT": 1990, + "TGGAAC": 1991, + "TGGAAG": 1992, + "TGGATA": 1993, + "TGGATT": 1994, + "TGGATC": 1995, + "TGGATG": 1996, + "TGGACA": 1997, + "TGGACT": 1998, + "TGGACC": 1999, + "TGGACG": 2000, + "TGGAGA": 2001, + "TGGAGT": 2002, + "TGGAGC": 2003, + "TGGAGG": 2004, + "TGGTAA": 2005, + "TGGTAT": 2006, + "TGGTAC": 2007, + "TGGTAG": 2008, + "TGGTTA": 2009, + "TGGTTT": 2010, + "TGGTTC": 2011, + "TGGTTG": 2012, + "TGGTCA": 2013, + "TGGTCT": 2014, + "TGGTCC": 2015, + "TGGTCG": 2016, + "TGGTGA": 2017, + "TGGTGT": 2018, + "TGGTGC": 2019, + "TGGTGG": 2020, + "TGGCAA": 2021, + "TGGCAT": 2022, + "TGGCAC": 2023, + "TGGCAG": 2024, + "TGGCTA": 2025, + "TGGCTT": 2026, + "TGGCTC": 2027, + "TGGCTG": 2028, + "TGGCCA": 2029, + "TGGCCT": 2030, + "TGGCCC": 2031, + "TGGCCG": 2032, + "TGGCGA": 2033, + "TGGCGT": 2034, + "TGGCGC": 2035, + "TGGCGG": 2036, + "TGGGAA": 2037, + "TGGGAT": 2038, + "TGGGAC": 2039, + "TGGGAG": 2040, + "TGGGTA": 2041, + "TGGGTT": 2042, + "TGGGTC": 2043, + "TGGGTG": 2044, + "TGGGCA": 2045, + "TGGGCT": 2046, + "TGGGCC": 2047, + "TGGGCG": 2048, + "TGGGGA": 2049, + "TGGGGT": 2050, + "TGGGGC": 2051, + "TGGGGG": 2052, + "CAAAAA": 2053, + "CAAAAT": 2054, + "CAAAAC": 2055, + "CAAAAG": 2056, + "CAAATA": 2057, + "CAAATT": 2058, + "CAAATC": 2059, + "CAAATG": 2060, + "CAAACA": 2061, + "CAAACT": 2062, + "CAAACC": 2063, + "CAAACG": 2064, + "CAAAGA": 2065, + "CAAAGT": 2066, + "CAAAGC": 2067, + "CAAAGG": 2068, + "CAATAA": 2069, + "CAATAT": 2070, + "CAATAC": 2071, + "CAATAG": 2072, + "CAATTA": 2073, + "CAATTT": 2074, + "CAATTC": 2075, + "CAATTG": 2076, + "CAATCA": 2077, + "CAATCT": 2078, + "CAATCC": 2079, + "CAATCG": 2080, + "CAATGA": 2081, + "CAATGT": 2082, + "CAATGC": 2083, + "CAATGG": 2084, + "CAACAA": 2085, + "CAACAT": 2086, + "CAACAC": 2087, + "CAACAG": 2088, + "CAACTA": 2089, + "CAACTT": 2090, + "CAACTC": 2091, + "CAACTG": 2092, + "CAACCA": 2093, + "CAACCT": 2094, + "CAACCC": 2095, + "CAACCG": 2096, + "CAACGA": 2097, + "CAACGT": 2098, + "CAACGC": 2099, + "CAACGG": 2100, + "CAAGAA": 2101, + "CAAGAT": 2102, + "CAAGAC": 2103, + "CAAGAG": 2104, + "CAAGTA": 2105, + "CAAGTT": 2106, + "CAAGTC": 2107, + "CAAGTG": 2108, + "CAAGCA": 2109, + "CAAGCT": 2110, + "CAAGCC": 2111, + "CAAGCG": 2112, + "CAAGGA": 2113, + "CAAGGT": 2114, + "CAAGGC": 2115, + "CAAGGG": 2116, + "CATAAA": 2117, + "CATAAT": 2118, + "CATAAC": 2119, + "CATAAG": 2120, + "CATATA": 2121, + "CATATT": 2122, + "CATATC": 2123, + "CATATG": 2124, + "CATACA": 2125, + "CATACT": 2126, + "CATACC": 2127, + "CATACG": 2128, + "CATAGA": 2129, + "CATAGT": 2130, + "CATAGC": 2131, + "CATAGG": 2132, + "CATTAA": 2133, + "CATTAT": 2134, + "CATTAC": 2135, + "CATTAG": 2136, + "CATTTA": 2137, + "CATTTT": 2138, + "CATTTC": 2139, + "CATTTG": 2140, + "CATTCA": 2141, + "CATTCT": 2142, + "CATTCC": 2143, + "CATTCG": 2144, + "CATTGA": 2145, + "CATTGT": 2146, + "CATTGC": 2147, + "CATTGG": 2148, + "CATCAA": 2149, + "CATCAT": 2150, + "CATCAC": 2151, + "CATCAG": 2152, + "CATCTA": 2153, + "CATCTT": 2154, + "CATCTC": 2155, + "CATCTG": 2156, + "CATCCA": 2157, + "CATCCT": 2158, + "CATCCC": 2159, + "CATCCG": 2160, + "CATCGA": 2161, + "CATCGT": 2162, + "CATCGC": 2163, + "CATCGG": 2164, + "CATGAA": 2165, + "CATGAT": 2166, + "CATGAC": 2167, + "CATGAG": 2168, + "CATGTA": 2169, + "CATGTT": 2170, + "CATGTC": 2171, + "CATGTG": 2172, + "CATGCA": 2173, + "CATGCT": 2174, + "CATGCC": 2175, + "CATGCG": 2176, + "CATGGA": 2177, + "CATGGT": 2178, + "CATGGC": 2179, + "CATGGG": 2180, + "CACAAA": 2181, + "CACAAT": 2182, + "CACAAC": 2183, + "CACAAG": 2184, + "CACATA": 2185, + "CACATT": 2186, + "CACATC": 2187, + "CACATG": 2188, + "CACACA": 2189, + "CACACT": 2190, + "CACACC": 2191, + "CACACG": 2192, + "CACAGA": 2193, + "CACAGT": 2194, + "CACAGC": 2195, + "CACAGG": 2196, + "CACTAA": 2197, + "CACTAT": 2198, + "CACTAC": 2199, + "CACTAG": 2200, + "CACTTA": 2201, + "CACTTT": 2202, + "CACTTC": 2203, + "CACTTG": 2204, + "CACTCA": 2205, + "CACTCT": 2206, + "CACTCC": 2207, + "CACTCG": 2208, + "CACTGA": 2209, + "CACTGT": 2210, + "CACTGC": 2211, + "CACTGG": 2212, + "CACCAA": 2213, + "CACCAT": 2214, + "CACCAC": 2215, + "CACCAG": 2216, + "CACCTA": 2217, + "CACCTT": 2218, + "CACCTC": 2219, + "CACCTG": 2220, + "CACCCA": 2221, + "CACCCT": 2222, + "CACCCC": 2223, + "CACCCG": 2224, + "CACCGA": 2225, + "CACCGT": 2226, + "CACCGC": 2227, + "CACCGG": 2228, + "CACGAA": 2229, + "CACGAT": 2230, + "CACGAC": 2231, + "CACGAG": 2232, + "CACGTA": 2233, + "CACGTT": 2234, + "CACGTC": 2235, + "CACGTG": 2236, + "CACGCA": 2237, + "CACGCT": 2238, + "CACGCC": 2239, + "CACGCG": 2240, + "CACGGA": 2241, + "CACGGT": 2242, + "CACGGC": 2243, + "CACGGG": 2244, + "CAGAAA": 2245, + "CAGAAT": 2246, + "CAGAAC": 2247, + "CAGAAG": 2248, + "CAGATA": 2249, + "CAGATT": 2250, + "CAGATC": 2251, + "CAGATG": 2252, + "CAGACA": 2253, + "CAGACT": 2254, + "CAGACC": 2255, + "CAGACG": 2256, + "CAGAGA": 2257, + "CAGAGT": 2258, + "CAGAGC": 2259, + "CAGAGG": 2260, + "CAGTAA": 2261, + "CAGTAT": 2262, + "CAGTAC": 2263, + "CAGTAG": 2264, + "CAGTTA": 2265, + "CAGTTT": 2266, + "CAGTTC": 2267, + "CAGTTG": 2268, + "CAGTCA": 2269, + "CAGTCT": 2270, + "CAGTCC": 2271, + "CAGTCG": 2272, + "CAGTGA": 2273, + "CAGTGT": 2274, + "CAGTGC": 2275, + "CAGTGG": 2276, + "CAGCAA": 2277, + "CAGCAT": 2278, + "CAGCAC": 2279, + "CAGCAG": 2280, + "CAGCTA": 2281, + "CAGCTT": 2282, + "CAGCTC": 2283, + "CAGCTG": 2284, + "CAGCCA": 2285, + "CAGCCT": 2286, + "CAGCCC": 2287, + "CAGCCG": 2288, + "CAGCGA": 2289, + "CAGCGT": 2290, + "CAGCGC": 2291, + "CAGCGG": 2292, + "CAGGAA": 2293, + "CAGGAT": 2294, + "CAGGAC": 2295, + "CAGGAG": 2296, + "CAGGTA": 2297, + "CAGGTT": 2298, + "CAGGTC": 2299, + "CAGGTG": 2300, + "CAGGCA": 2301, + "CAGGCT": 2302, + "CAGGCC": 2303, + "CAGGCG": 2304, + "CAGGGA": 2305, + "CAGGGT": 2306, + "CAGGGC": 2307, + "CAGGGG": 2308, + "CTAAAA": 2309, + "CTAAAT": 2310, + "CTAAAC": 2311, + "CTAAAG": 2312, + "CTAATA": 2313, + "CTAATT": 2314, + "CTAATC": 2315, + "CTAATG": 2316, + "CTAACA": 2317, + "CTAACT": 2318, + "CTAACC": 2319, + "CTAACG": 2320, + "CTAAGA": 2321, + "CTAAGT": 2322, + "CTAAGC": 2323, + "CTAAGG": 2324, + "CTATAA": 2325, + "CTATAT": 2326, + "CTATAC": 2327, + "CTATAG": 2328, + "CTATTA": 2329, + "CTATTT": 2330, + "CTATTC": 2331, + "CTATTG": 2332, + "CTATCA": 2333, + "CTATCT": 2334, + "CTATCC": 2335, + "CTATCG": 2336, + "CTATGA": 2337, + "CTATGT": 2338, + "CTATGC": 2339, + "CTATGG": 2340, + "CTACAA": 2341, + "CTACAT": 2342, + "CTACAC": 2343, + "CTACAG": 2344, + "CTACTA": 2345, + "CTACTT": 2346, + "CTACTC": 2347, + "CTACTG": 2348, + "CTACCA": 2349, + "CTACCT": 2350, + "CTACCC": 2351, + "CTACCG": 2352, + "CTACGA": 2353, + "CTACGT": 2354, + "CTACGC": 2355, + "CTACGG": 2356, + "CTAGAA": 2357, + "CTAGAT": 2358, + "CTAGAC": 2359, + "CTAGAG": 2360, + "CTAGTA": 2361, + "CTAGTT": 2362, + "CTAGTC": 2363, + "CTAGTG": 2364, + "CTAGCA": 2365, + "CTAGCT": 2366, + "CTAGCC": 2367, + "CTAGCG": 2368, + "CTAGGA": 2369, + "CTAGGT": 2370, + "CTAGGC": 2371, + "CTAGGG": 2372, + "CTTAAA": 2373, + "CTTAAT": 2374, + "CTTAAC": 2375, + "CTTAAG": 2376, + "CTTATA": 2377, + "CTTATT": 2378, + "CTTATC": 2379, + "CTTATG": 2380, + "CTTACA": 2381, + "CTTACT": 2382, + "CTTACC": 2383, + "CTTACG": 2384, + "CTTAGA": 2385, + "CTTAGT": 2386, + "CTTAGC": 2387, + "CTTAGG": 2388, + "CTTTAA": 2389, + "CTTTAT": 2390, + "CTTTAC": 2391, + "CTTTAG": 2392, + "CTTTTA": 2393, + "CTTTTT": 2394, + "CTTTTC": 2395, + "CTTTTG": 2396, + "CTTTCA": 2397, + "CTTTCT": 2398, + "CTTTCC": 2399, + "CTTTCG": 2400, + "CTTTGA": 2401, + "CTTTGT": 2402, + "CTTTGC": 2403, + "CTTTGG": 2404, + "CTTCAA": 2405, + "CTTCAT": 2406, + "CTTCAC": 2407, + "CTTCAG": 2408, + "CTTCTA": 2409, + "CTTCTT": 2410, + "CTTCTC": 2411, + "CTTCTG": 2412, + "CTTCCA": 2413, + "CTTCCT": 2414, + "CTTCCC": 2415, + "CTTCCG": 2416, + "CTTCGA": 2417, + "CTTCGT": 2418, + "CTTCGC": 2419, + "CTTCGG": 2420, + "CTTGAA": 2421, + "CTTGAT": 2422, + "CTTGAC": 2423, + "CTTGAG": 2424, + "CTTGTA": 2425, + "CTTGTT": 2426, + "CTTGTC": 2427, + "CTTGTG": 2428, + "CTTGCA": 2429, + "CTTGCT": 2430, + "CTTGCC": 2431, + "CTTGCG": 2432, + "CTTGGA": 2433, + "CTTGGT": 2434, + "CTTGGC": 2435, + "CTTGGG": 2436, + "CTCAAA": 2437, + "CTCAAT": 2438, + "CTCAAC": 2439, + "CTCAAG": 2440, + "CTCATA": 2441, + "CTCATT": 2442, + "CTCATC": 2443, + "CTCATG": 2444, + "CTCACA": 2445, + "CTCACT": 2446, + "CTCACC": 2447, + "CTCACG": 2448, + "CTCAGA": 2449, + "CTCAGT": 2450, + "CTCAGC": 2451, + "CTCAGG": 2452, + "CTCTAA": 2453, + "CTCTAT": 2454, + "CTCTAC": 2455, + "CTCTAG": 2456, + "CTCTTA": 2457, + "CTCTTT": 2458, + "CTCTTC": 2459, + "CTCTTG": 2460, + "CTCTCA": 2461, + "CTCTCT": 2462, + "CTCTCC": 2463, + "CTCTCG": 2464, + "CTCTGA": 2465, + "CTCTGT": 2466, + "CTCTGC": 2467, + "CTCTGG": 2468, + "CTCCAA": 2469, + "CTCCAT": 2470, + "CTCCAC": 2471, + "CTCCAG": 2472, + "CTCCTA": 2473, + "CTCCTT": 2474, + "CTCCTC": 2475, + "CTCCTG": 2476, + "CTCCCA": 2477, + "CTCCCT": 2478, + "CTCCCC": 2479, + "CTCCCG": 2480, + "CTCCGA": 2481, + "CTCCGT": 2482, + "CTCCGC": 2483, + "CTCCGG": 2484, + "CTCGAA": 2485, + "CTCGAT": 2486, + "CTCGAC": 2487, + "CTCGAG": 2488, + "CTCGTA": 2489, + "CTCGTT": 2490, + "CTCGTC": 2491, + "CTCGTG": 2492, + "CTCGCA": 2493, + "CTCGCT": 2494, + "CTCGCC": 2495, + "CTCGCG": 2496, + "CTCGGA": 2497, + "CTCGGT": 2498, + "CTCGGC": 2499, + "CTCGGG": 2500, + "CTGAAA": 2501, + "CTGAAT": 2502, + "CTGAAC": 2503, + "CTGAAG": 2504, + "CTGATA": 2505, + "CTGATT": 2506, + "CTGATC": 2507, + "CTGATG": 2508, + "CTGACA": 2509, + "CTGACT": 2510, + "CTGACC": 2511, + "CTGACG": 2512, + "CTGAGA": 2513, + "CTGAGT": 2514, + "CTGAGC": 2515, + "CTGAGG": 2516, + "CTGTAA": 2517, + "CTGTAT": 2518, + "CTGTAC": 2519, + "CTGTAG": 2520, + "CTGTTA": 2521, + "CTGTTT": 2522, + "CTGTTC": 2523, + "CTGTTG": 2524, + "CTGTCA": 2525, + "CTGTCT": 2526, + "CTGTCC": 2527, + "CTGTCG": 2528, + "CTGTGA": 2529, + "CTGTGT": 2530, + "CTGTGC": 2531, + "CTGTGG": 2532, + "CTGCAA": 2533, + "CTGCAT": 2534, + "CTGCAC": 2535, + "CTGCAG": 2536, + "CTGCTA": 2537, + "CTGCTT": 2538, + "CTGCTC": 2539, + "CTGCTG": 2540, + "CTGCCA": 2541, + "CTGCCT": 2542, + "CTGCCC": 2543, + "CTGCCG": 2544, + "CTGCGA": 2545, + "CTGCGT": 2546, + "CTGCGC": 2547, + "CTGCGG": 2548, + "CTGGAA": 2549, + "CTGGAT": 2550, + "CTGGAC": 2551, + "CTGGAG": 2552, + "CTGGTA": 2553, + "CTGGTT": 2554, + "CTGGTC": 2555, + "CTGGTG": 2556, + "CTGGCA": 2557, + "CTGGCT": 2558, + "CTGGCC": 2559, + "CTGGCG": 2560, + "CTGGGA": 2561, + "CTGGGT": 2562, + "CTGGGC": 2563, + "CTGGGG": 2564, + "CCAAAA": 2565, + "CCAAAT": 2566, + "CCAAAC": 2567, + "CCAAAG": 2568, + "CCAATA": 2569, + "CCAATT": 2570, + "CCAATC": 2571, + "CCAATG": 2572, + "CCAACA": 2573, + "CCAACT": 2574, + "CCAACC": 2575, + "CCAACG": 2576, + "CCAAGA": 2577, + "CCAAGT": 2578, + "CCAAGC": 2579, + "CCAAGG": 2580, + "CCATAA": 2581, + "CCATAT": 2582, + "CCATAC": 2583, + "CCATAG": 2584, + "CCATTA": 2585, + "CCATTT": 2586, + "CCATTC": 2587, + "CCATTG": 2588, + "CCATCA": 2589, + "CCATCT": 2590, + "CCATCC": 2591, + "CCATCG": 2592, + "CCATGA": 2593, + "CCATGT": 2594, + "CCATGC": 2595, + "CCATGG": 2596, + "CCACAA": 2597, + "CCACAT": 2598, + "CCACAC": 2599, + "CCACAG": 2600, + "CCACTA": 2601, + "CCACTT": 2602, + "CCACTC": 2603, + "CCACTG": 2604, + "CCACCA": 2605, + "CCACCT": 2606, + "CCACCC": 2607, + "CCACCG": 2608, + "CCACGA": 2609, + "CCACGT": 2610, + "CCACGC": 2611, + "CCACGG": 2612, + "CCAGAA": 2613, + "CCAGAT": 2614, + "CCAGAC": 2615, + "CCAGAG": 2616, + "CCAGTA": 2617, + "CCAGTT": 2618, + "CCAGTC": 2619, + "CCAGTG": 2620, + "CCAGCA": 2621, + "CCAGCT": 2622, + "CCAGCC": 2623, + "CCAGCG": 2624, + "CCAGGA": 2625, + "CCAGGT": 2626, + "CCAGGC": 2627, + "CCAGGG": 2628, + "CCTAAA": 2629, + "CCTAAT": 2630, + "CCTAAC": 2631, + "CCTAAG": 2632, + "CCTATA": 2633, + "CCTATT": 2634, + "CCTATC": 2635, + "CCTATG": 2636, + "CCTACA": 2637, + "CCTACT": 2638, + "CCTACC": 2639, + "CCTACG": 2640, + "CCTAGA": 2641, + "CCTAGT": 2642, + "CCTAGC": 2643, + "CCTAGG": 2644, + "CCTTAA": 2645, + "CCTTAT": 2646, + "CCTTAC": 2647, + "CCTTAG": 2648, + "CCTTTA": 2649, + "CCTTTT": 2650, + "CCTTTC": 2651, + "CCTTTG": 2652, + "CCTTCA": 2653, + "CCTTCT": 2654, + "CCTTCC": 2655, + "CCTTCG": 2656, + "CCTTGA": 2657, + "CCTTGT": 2658, + "CCTTGC": 2659, + "CCTTGG": 2660, + "CCTCAA": 2661, + "CCTCAT": 2662, + "CCTCAC": 2663, + "CCTCAG": 2664, + "CCTCTA": 2665, + "CCTCTT": 2666, + "CCTCTC": 2667, + "CCTCTG": 2668, + "CCTCCA": 2669, + "CCTCCT": 2670, + "CCTCCC": 2671, + "CCTCCG": 2672, + "CCTCGA": 2673, + "CCTCGT": 2674, + "CCTCGC": 2675, + "CCTCGG": 2676, + "CCTGAA": 2677, + "CCTGAT": 2678, + "CCTGAC": 2679, + "CCTGAG": 2680, + "CCTGTA": 2681, + "CCTGTT": 2682, + "CCTGTC": 2683, + "CCTGTG": 2684, + "CCTGCA": 2685, + "CCTGCT": 2686, + "CCTGCC": 2687, + "CCTGCG": 2688, + "CCTGGA": 2689, + "CCTGGT": 2690, + "CCTGGC": 2691, + "CCTGGG": 2692, + "CCCAAA": 2693, + "CCCAAT": 2694, + "CCCAAC": 2695, + "CCCAAG": 2696, + "CCCATA": 2697, + "CCCATT": 2698, + "CCCATC": 2699, + "CCCATG": 2700, + "CCCACA": 2701, + "CCCACT": 2702, + "CCCACC": 2703, + "CCCACG": 2704, + "CCCAGA": 2705, + "CCCAGT": 2706, + "CCCAGC": 2707, + "CCCAGG": 2708, + "CCCTAA": 2709, + "CCCTAT": 2710, + "CCCTAC": 2711, + "CCCTAG": 2712, + "CCCTTA": 2713, + "CCCTTT": 2714, + "CCCTTC": 2715, + "CCCTTG": 2716, + "CCCTCA": 2717, + "CCCTCT": 2718, + "CCCTCC": 2719, + "CCCTCG": 2720, + "CCCTGA": 2721, + "CCCTGT": 2722, + "CCCTGC": 2723, + "CCCTGG": 2724, + "CCCCAA": 2725, + "CCCCAT": 2726, + "CCCCAC": 2727, + "CCCCAG": 2728, + "CCCCTA": 2729, + "CCCCTT": 2730, + "CCCCTC": 2731, + "CCCCTG": 2732, + "CCCCCA": 2733, + "CCCCCT": 2734, + "CCCCCC": 2735, + "CCCCCG": 2736, + "CCCCGA": 2737, + "CCCCGT": 2738, + "CCCCGC": 2739, + "CCCCGG": 2740, + "CCCGAA": 2741, + "CCCGAT": 2742, + "CCCGAC": 2743, + "CCCGAG": 2744, + "CCCGTA": 2745, + "CCCGTT": 2746, + "CCCGTC": 2747, + "CCCGTG": 2748, + "CCCGCA": 2749, + "CCCGCT": 2750, + "CCCGCC": 2751, + "CCCGCG": 2752, + "CCCGGA": 2753, + "CCCGGT": 2754, + "CCCGGC": 2755, + "CCCGGG": 2756, + "CCGAAA": 2757, + "CCGAAT": 2758, + "CCGAAC": 2759, + "CCGAAG": 2760, + "CCGATA": 2761, + "CCGATT": 2762, + "CCGATC": 2763, + "CCGATG": 2764, + "CCGACA": 2765, + "CCGACT": 2766, + "CCGACC": 2767, + "CCGACG": 2768, + "CCGAGA": 2769, + "CCGAGT": 2770, + "CCGAGC": 2771, + "CCGAGG": 2772, + "CCGTAA": 2773, + "CCGTAT": 2774, + "CCGTAC": 2775, + "CCGTAG": 2776, + "CCGTTA": 2777, + "CCGTTT": 2778, + "CCGTTC": 2779, + "CCGTTG": 2780, + "CCGTCA": 2781, + "CCGTCT": 2782, + "CCGTCC": 2783, + "CCGTCG": 2784, + "CCGTGA": 2785, + "CCGTGT": 2786, + "CCGTGC": 2787, + "CCGTGG": 2788, + "CCGCAA": 2789, + "CCGCAT": 2790, + "CCGCAC": 2791, + "CCGCAG": 2792, + "CCGCTA": 2793, + "CCGCTT": 2794, + "CCGCTC": 2795, + "CCGCTG": 2796, + "CCGCCA": 2797, + "CCGCCT": 2798, + "CCGCCC": 2799, + "CCGCCG": 2800, + "CCGCGA": 2801, + "CCGCGT": 2802, + "CCGCGC": 2803, + "CCGCGG": 2804, + "CCGGAA": 2805, + "CCGGAT": 2806, + "CCGGAC": 2807, + "CCGGAG": 2808, + "CCGGTA": 2809, + "CCGGTT": 2810, + "CCGGTC": 2811, + "CCGGTG": 2812, + "CCGGCA": 2813, + "CCGGCT": 2814, + "CCGGCC": 2815, + "CCGGCG": 2816, + "CCGGGA": 2817, + "CCGGGT": 2818, + "CCGGGC": 2819, + "CCGGGG": 2820, + "CGAAAA": 2821, + "CGAAAT": 2822, + "CGAAAC": 2823, + "CGAAAG": 2824, + "CGAATA": 2825, + "CGAATT": 2826, + "CGAATC": 2827, + "CGAATG": 2828, + "CGAACA": 2829, + "CGAACT": 2830, + "CGAACC": 2831, + "CGAACG": 2832, + "CGAAGA": 2833, + "CGAAGT": 2834, + "CGAAGC": 2835, + "CGAAGG": 2836, + "CGATAA": 2837, + "CGATAT": 2838, + "CGATAC": 2839, + "CGATAG": 2840, + "CGATTA": 2841, + "CGATTT": 2842, + "CGATTC": 2843, + "CGATTG": 2844, + "CGATCA": 2845, + "CGATCT": 2846, + "CGATCC": 2847, + "CGATCG": 2848, + "CGATGA": 2849, + "CGATGT": 2850, + "CGATGC": 2851, + "CGATGG": 2852, + "CGACAA": 2853, + "CGACAT": 2854, + "CGACAC": 2855, + "CGACAG": 2856, + "CGACTA": 2857, + "CGACTT": 2858, + "CGACTC": 2859, + "CGACTG": 2860, + "CGACCA": 2861, + "CGACCT": 2862, + "CGACCC": 2863, + "CGACCG": 2864, + "CGACGA": 2865, + "CGACGT": 2866, + "CGACGC": 2867, + "CGACGG": 2868, + "CGAGAA": 2869, + "CGAGAT": 2870, + "CGAGAC": 2871, + "CGAGAG": 2872, + "CGAGTA": 2873, + "CGAGTT": 2874, + "CGAGTC": 2875, + "CGAGTG": 2876, + "CGAGCA": 2877, + "CGAGCT": 2878, + "CGAGCC": 2879, + "CGAGCG": 2880, + "CGAGGA": 2881, + "CGAGGT": 2882, + "CGAGGC": 2883, + "CGAGGG": 2884, + "CGTAAA": 2885, + "CGTAAT": 2886, + "CGTAAC": 2887, + "CGTAAG": 2888, + "CGTATA": 2889, + "CGTATT": 2890, + "CGTATC": 2891, + "CGTATG": 2892, + "CGTACA": 2893, + "CGTACT": 2894, + "CGTACC": 2895, + "CGTACG": 2896, + "CGTAGA": 2897, + "CGTAGT": 2898, + "CGTAGC": 2899, + "CGTAGG": 2900, + "CGTTAA": 2901, + "CGTTAT": 2902, + "CGTTAC": 2903, + "CGTTAG": 2904, + "CGTTTA": 2905, + "CGTTTT": 2906, + "CGTTTC": 2907, + "CGTTTG": 2908, + "CGTTCA": 2909, + "CGTTCT": 2910, + "CGTTCC": 2911, + "CGTTCG": 2912, + "CGTTGA": 2913, + "CGTTGT": 2914, + "CGTTGC": 2915, + "CGTTGG": 2916, + "CGTCAA": 2917, + "CGTCAT": 2918, + "CGTCAC": 2919, + "CGTCAG": 2920, + "CGTCTA": 2921, + "CGTCTT": 2922, + "CGTCTC": 2923, + "CGTCTG": 2924, + "CGTCCA": 2925, + "CGTCCT": 2926, + "CGTCCC": 2927, + "CGTCCG": 2928, + "CGTCGA": 2929, + "CGTCGT": 2930, + "CGTCGC": 2931, + "CGTCGG": 2932, + "CGTGAA": 2933, + "CGTGAT": 2934, + "CGTGAC": 2935, + "CGTGAG": 2936, + "CGTGTA": 2937, + "CGTGTT": 2938, + "CGTGTC": 2939, + "CGTGTG": 2940, + "CGTGCA": 2941, + "CGTGCT": 2942, + "CGTGCC": 2943, + "CGTGCG": 2944, + "CGTGGA": 2945, + "CGTGGT": 2946, + "CGTGGC": 2947, + "CGTGGG": 2948, + "CGCAAA": 2949, + "CGCAAT": 2950, + "CGCAAC": 2951, + "CGCAAG": 2952, + "CGCATA": 2953, + "CGCATT": 2954, + "CGCATC": 2955, + "CGCATG": 2956, + "CGCACA": 2957, + "CGCACT": 2958, + "CGCACC": 2959, + "CGCACG": 2960, + "CGCAGA": 2961, + "CGCAGT": 2962, + "CGCAGC": 2963, + "CGCAGG": 2964, + "CGCTAA": 2965, + "CGCTAT": 2966, + "CGCTAC": 2967, + "CGCTAG": 2968, + "CGCTTA": 2969, + "CGCTTT": 2970, + "CGCTTC": 2971, + "CGCTTG": 2972, + "CGCTCA": 2973, + "CGCTCT": 2974, + "CGCTCC": 2975, + "CGCTCG": 2976, + "CGCTGA": 2977, + "CGCTGT": 2978, + "CGCTGC": 2979, + "CGCTGG": 2980, + "CGCCAA": 2981, + "CGCCAT": 2982, + "CGCCAC": 2983, + "CGCCAG": 2984, + "CGCCTA": 2985, + "CGCCTT": 2986, + "CGCCTC": 2987, + "CGCCTG": 2988, + "CGCCCA": 2989, + "CGCCCT": 2990, + "CGCCCC": 2991, + "CGCCCG": 2992, + "CGCCGA": 2993, + "CGCCGT": 2994, + "CGCCGC": 2995, + "CGCCGG": 2996, + "CGCGAA": 2997, + "CGCGAT": 2998, + "CGCGAC": 2999, + "CGCGAG": 3000, + "CGCGTA": 3001, + "CGCGTT": 3002, + "CGCGTC": 3003, + "CGCGTG": 3004, + "CGCGCA": 3005, + "CGCGCT": 3006, + "CGCGCC": 3007, + "CGCGCG": 3008, + "CGCGGA": 3009, + "CGCGGT": 3010, + "CGCGGC": 3011, + "CGCGGG": 3012, + "CGGAAA": 3013, + "CGGAAT": 3014, + "CGGAAC": 3015, + "CGGAAG": 3016, + "CGGATA": 3017, + "CGGATT": 3018, + "CGGATC": 3019, + "CGGATG": 3020, + "CGGACA": 3021, + "CGGACT": 3022, + "CGGACC": 3023, + "CGGACG": 3024, + "CGGAGA": 3025, + "CGGAGT": 3026, + "CGGAGC": 3027, + "CGGAGG": 3028, + "CGGTAA": 3029, + "CGGTAT": 3030, + "CGGTAC": 3031, + "CGGTAG": 3032, + "CGGTTA": 3033, + "CGGTTT": 3034, + "CGGTTC": 3035, + "CGGTTG": 3036, + "CGGTCA": 3037, + "CGGTCT": 3038, + "CGGTCC": 3039, + "CGGTCG": 3040, + "CGGTGA": 3041, + "CGGTGT": 3042, + "CGGTGC": 3043, + "CGGTGG": 3044, + "CGGCAA": 3045, + "CGGCAT": 3046, + "CGGCAC": 3047, + "CGGCAG": 3048, + "CGGCTA": 3049, + "CGGCTT": 3050, + "CGGCTC": 3051, + "CGGCTG": 3052, + "CGGCCA": 3053, + "CGGCCT": 3054, + "CGGCCC": 3055, + "CGGCCG": 3056, + "CGGCGA": 3057, + "CGGCGT": 3058, + "CGGCGC": 3059, + "CGGCGG": 3060, + "CGGGAA": 3061, + "CGGGAT": 3062, + "CGGGAC": 3063, + "CGGGAG": 3064, + "CGGGTA": 3065, + "CGGGTT": 3066, + "CGGGTC": 3067, + "CGGGTG": 3068, + "CGGGCA": 3069, + "CGGGCT": 3070, + "CGGGCC": 3071, + "CGGGCG": 3072, + "CGGGGA": 3073, + "CGGGGT": 3074, + "CGGGGC": 3075, + "CGGGGG": 3076, + "GAAAAA": 3077, + "GAAAAT": 3078, + "GAAAAC": 3079, + "GAAAAG": 3080, + "GAAATA": 3081, + "GAAATT": 3082, + "GAAATC": 3083, + "GAAATG": 3084, + "GAAACA": 3085, + "GAAACT": 3086, + "GAAACC": 3087, + "GAAACG": 3088, + "GAAAGA": 3089, + "GAAAGT": 3090, + "GAAAGC": 3091, + "GAAAGG": 3092, + "GAATAA": 3093, + "GAATAT": 3094, + "GAATAC": 3095, + "GAATAG": 3096, + "GAATTA": 3097, + "GAATTT": 3098, + "GAATTC": 3099, + "GAATTG": 3100, + "GAATCA": 3101, + "GAATCT": 3102, + "GAATCC": 3103, + "GAATCG": 3104, + "GAATGA": 3105, + "GAATGT": 3106, + "GAATGC": 3107, + "GAATGG": 3108, + "GAACAA": 3109, + "GAACAT": 3110, + "GAACAC": 3111, + "GAACAG": 3112, + "GAACTA": 3113, + "GAACTT": 3114, + "GAACTC": 3115, + "GAACTG": 3116, + "GAACCA": 3117, + "GAACCT": 3118, + "GAACCC": 3119, + "GAACCG": 3120, + "GAACGA": 3121, + "GAACGT": 3122, + "GAACGC": 3123, + "GAACGG": 3124, + "GAAGAA": 3125, + "GAAGAT": 3126, + "GAAGAC": 3127, + "GAAGAG": 3128, + "GAAGTA": 3129, + "GAAGTT": 3130, + "GAAGTC": 3131, + "GAAGTG": 3132, + "GAAGCA": 3133, + "GAAGCT": 3134, + "GAAGCC": 3135, + "GAAGCG": 3136, + "GAAGGA": 3137, + "GAAGGT": 3138, + "GAAGGC": 3139, + "GAAGGG": 3140, + "GATAAA": 3141, + "GATAAT": 3142, + "GATAAC": 3143, + "GATAAG": 3144, + "GATATA": 3145, + "GATATT": 3146, + "GATATC": 3147, + "GATATG": 3148, + "GATACA": 3149, + "GATACT": 3150, + "GATACC": 3151, + "GATACG": 3152, + "GATAGA": 3153, + "GATAGT": 3154, + "GATAGC": 3155, + "GATAGG": 3156, + "GATTAA": 3157, + "GATTAT": 3158, + "GATTAC": 3159, + "GATTAG": 3160, + "GATTTA": 3161, + "GATTTT": 3162, + "GATTTC": 3163, + "GATTTG": 3164, + "GATTCA": 3165, + "GATTCT": 3166, + "GATTCC": 3167, + "GATTCG": 3168, + "GATTGA": 3169, + "GATTGT": 3170, + "GATTGC": 3171, + "GATTGG": 3172, + "GATCAA": 3173, + "GATCAT": 3174, + "GATCAC": 3175, + "GATCAG": 3176, + "GATCTA": 3177, + "GATCTT": 3178, + "GATCTC": 3179, + "GATCTG": 3180, + "GATCCA": 3181, + "GATCCT": 3182, + "GATCCC": 3183, + "GATCCG": 3184, + "GATCGA": 3185, + "GATCGT": 3186, + "GATCGC": 3187, + "GATCGG": 3188, + "GATGAA": 3189, + "GATGAT": 3190, + "GATGAC": 3191, + "GATGAG": 3192, + "GATGTA": 3193, + "GATGTT": 3194, + "GATGTC": 3195, + "GATGTG": 3196, + "GATGCA": 3197, + "GATGCT": 3198, + "GATGCC": 3199, + "GATGCG": 3200, + "GATGGA": 3201, + "GATGGT": 3202, + "GATGGC": 3203, + "GATGGG": 3204, + "GACAAA": 3205, + "GACAAT": 3206, + "GACAAC": 3207, + "GACAAG": 3208, + "GACATA": 3209, + "GACATT": 3210, + "GACATC": 3211, + "GACATG": 3212, + "GACACA": 3213, + "GACACT": 3214, + "GACACC": 3215, + "GACACG": 3216, + "GACAGA": 3217, + "GACAGT": 3218, + "GACAGC": 3219, + "GACAGG": 3220, + "GACTAA": 3221, + "GACTAT": 3222, + "GACTAC": 3223, + "GACTAG": 3224, + "GACTTA": 3225, + "GACTTT": 3226, + "GACTTC": 3227, + "GACTTG": 3228, + "GACTCA": 3229, + "GACTCT": 3230, + "GACTCC": 3231, + "GACTCG": 3232, + "GACTGA": 3233, + "GACTGT": 3234, + "GACTGC": 3235, + "GACTGG": 3236, + "GACCAA": 3237, + "GACCAT": 3238, + "GACCAC": 3239, + "GACCAG": 3240, + "GACCTA": 3241, + "GACCTT": 3242, + "GACCTC": 3243, + "GACCTG": 3244, + "GACCCA": 3245, + "GACCCT": 3246, + "GACCCC": 3247, + "GACCCG": 3248, + "GACCGA": 3249, + "GACCGT": 3250, + "GACCGC": 3251, + "GACCGG": 3252, + "GACGAA": 3253, + "GACGAT": 3254, + "GACGAC": 3255, + "GACGAG": 3256, + "GACGTA": 3257, + "GACGTT": 3258, + "GACGTC": 3259, + "GACGTG": 3260, + "GACGCA": 3261, + "GACGCT": 3262, + "GACGCC": 3263, + "GACGCG": 3264, + "GACGGA": 3265, + "GACGGT": 3266, + "GACGGC": 3267, + "GACGGG": 3268, + "GAGAAA": 3269, + "GAGAAT": 3270, + "GAGAAC": 3271, + "GAGAAG": 3272, + "GAGATA": 3273, + "GAGATT": 3274, + "GAGATC": 3275, + "GAGATG": 3276, + "GAGACA": 3277, + "GAGACT": 3278, + "GAGACC": 3279, + "GAGACG": 3280, + "GAGAGA": 3281, + "GAGAGT": 3282, + "GAGAGC": 3283, + "GAGAGG": 3284, + "GAGTAA": 3285, + "GAGTAT": 3286, + "GAGTAC": 3287, + "GAGTAG": 3288, + "GAGTTA": 3289, + "GAGTTT": 3290, + "GAGTTC": 3291, + "GAGTTG": 3292, + "GAGTCA": 3293, + "GAGTCT": 3294, + "GAGTCC": 3295, + "GAGTCG": 3296, + "GAGTGA": 3297, + "GAGTGT": 3298, + "GAGTGC": 3299, + "GAGTGG": 3300, + "GAGCAA": 3301, + "GAGCAT": 3302, + "GAGCAC": 3303, + "GAGCAG": 3304, + "GAGCTA": 3305, + "GAGCTT": 3306, + "GAGCTC": 3307, + "GAGCTG": 3308, + "GAGCCA": 3309, + "GAGCCT": 3310, + "GAGCCC": 3311, + "GAGCCG": 3312, + "GAGCGA": 3313, + "GAGCGT": 3314, + "GAGCGC": 3315, + "GAGCGG": 3316, + "GAGGAA": 3317, + "GAGGAT": 3318, + "GAGGAC": 3319, + "GAGGAG": 3320, + "GAGGTA": 3321, + "GAGGTT": 3322, + "GAGGTC": 3323, + "GAGGTG": 3324, + "GAGGCA": 3325, + "GAGGCT": 3326, + "GAGGCC": 3327, + "GAGGCG": 3328, + "GAGGGA": 3329, + "GAGGGT": 3330, + "GAGGGC": 3331, + "GAGGGG": 3332, + "GTAAAA": 3333, + "GTAAAT": 3334, + "GTAAAC": 3335, + "GTAAAG": 3336, + "GTAATA": 3337, + "GTAATT": 3338, + "GTAATC": 3339, + "GTAATG": 3340, + "GTAACA": 3341, + "GTAACT": 3342, + "GTAACC": 3343, + "GTAACG": 3344, + "GTAAGA": 3345, + "GTAAGT": 3346, + "GTAAGC": 3347, + "GTAAGG": 3348, + "GTATAA": 3349, + "GTATAT": 3350, + "GTATAC": 3351, + "GTATAG": 3352, + "GTATTA": 3353, + "GTATTT": 3354, + "GTATTC": 3355, + "GTATTG": 3356, + "GTATCA": 3357, + "GTATCT": 3358, + "GTATCC": 3359, + "GTATCG": 3360, + "GTATGA": 3361, + "GTATGT": 3362, + "GTATGC": 3363, + "GTATGG": 3364, + "GTACAA": 3365, + "GTACAT": 3366, + "GTACAC": 3367, + "GTACAG": 3368, + "GTACTA": 3369, + "GTACTT": 3370, + "GTACTC": 3371, + "GTACTG": 3372, + "GTACCA": 3373, + "GTACCT": 3374, + "GTACCC": 3375, + "GTACCG": 3376, + "GTACGA": 3377, + "GTACGT": 3378, + "GTACGC": 3379, + "GTACGG": 3380, + "GTAGAA": 3381, + "GTAGAT": 3382, + "GTAGAC": 3383, + "GTAGAG": 3384, + "GTAGTA": 3385, + "GTAGTT": 3386, + "GTAGTC": 3387, + "GTAGTG": 3388, + "GTAGCA": 3389, + "GTAGCT": 3390, + "GTAGCC": 3391, + "GTAGCG": 3392, + "GTAGGA": 3393, + "GTAGGT": 3394, + "GTAGGC": 3395, + "GTAGGG": 3396, + "GTTAAA": 3397, + "GTTAAT": 3398, + "GTTAAC": 3399, + "GTTAAG": 3400, + "GTTATA": 3401, + "GTTATT": 3402, + "GTTATC": 3403, + "GTTATG": 3404, + "GTTACA": 3405, + "GTTACT": 3406, + "GTTACC": 3407, + "GTTACG": 3408, + "GTTAGA": 3409, + "GTTAGT": 3410, + "GTTAGC": 3411, + "GTTAGG": 3412, + "GTTTAA": 3413, + "GTTTAT": 3414, + "GTTTAC": 3415, + "GTTTAG": 3416, + "GTTTTA": 3417, + "GTTTTT": 3418, + "GTTTTC": 3419, + "GTTTTG": 3420, + "GTTTCA": 3421, + "GTTTCT": 3422, + "GTTTCC": 3423, + "GTTTCG": 3424, + "GTTTGA": 3425, + "GTTTGT": 3426, + "GTTTGC": 3427, + "GTTTGG": 3428, + "GTTCAA": 3429, + "GTTCAT": 3430, + "GTTCAC": 3431, + "GTTCAG": 3432, + "GTTCTA": 3433, + "GTTCTT": 3434, + "GTTCTC": 3435, + "GTTCTG": 3436, + "GTTCCA": 3437, + "GTTCCT": 3438, + "GTTCCC": 3439, + "GTTCCG": 3440, + "GTTCGA": 3441, + "GTTCGT": 3442, + "GTTCGC": 3443, + "GTTCGG": 3444, + "GTTGAA": 3445, + "GTTGAT": 3446, + "GTTGAC": 3447, + "GTTGAG": 3448, + "GTTGTA": 3449, + "GTTGTT": 3450, + "GTTGTC": 3451, + "GTTGTG": 3452, + "GTTGCA": 3453, + "GTTGCT": 3454, + "GTTGCC": 3455, + "GTTGCG": 3456, + "GTTGGA": 3457, + "GTTGGT": 3458, + "GTTGGC": 3459, + "GTTGGG": 3460, + "GTCAAA": 3461, + "GTCAAT": 3462, + "GTCAAC": 3463, + "GTCAAG": 3464, + "GTCATA": 3465, + "GTCATT": 3466, + "GTCATC": 3467, + "GTCATG": 3468, + "GTCACA": 3469, + "GTCACT": 3470, + "GTCACC": 3471, + "GTCACG": 3472, + "GTCAGA": 3473, + "GTCAGT": 3474, + "GTCAGC": 3475, + "GTCAGG": 3476, + "GTCTAA": 3477, + "GTCTAT": 3478, + "GTCTAC": 3479, + "GTCTAG": 3480, + "GTCTTA": 3481, + "GTCTTT": 3482, + "GTCTTC": 3483, + "GTCTTG": 3484, + "GTCTCA": 3485, + "GTCTCT": 3486, + "GTCTCC": 3487, + "GTCTCG": 3488, + "GTCTGA": 3489, + "GTCTGT": 3490, + "GTCTGC": 3491, + "GTCTGG": 3492, + "GTCCAA": 3493, + "GTCCAT": 3494, + "GTCCAC": 3495, + "GTCCAG": 3496, + "GTCCTA": 3497, + "GTCCTT": 3498, + "GTCCTC": 3499, + "GTCCTG": 3500, + "GTCCCA": 3501, + "GTCCCT": 3502, + "GTCCCC": 3503, + "GTCCCG": 3504, + "GTCCGA": 3505, + "GTCCGT": 3506, + "GTCCGC": 3507, + "GTCCGG": 3508, + "GTCGAA": 3509, + "GTCGAT": 3510, + "GTCGAC": 3511, + "GTCGAG": 3512, + "GTCGTA": 3513, + "GTCGTT": 3514, + "GTCGTC": 3515, + "GTCGTG": 3516, + "GTCGCA": 3517, + "GTCGCT": 3518, + "GTCGCC": 3519, + "GTCGCG": 3520, + "GTCGGA": 3521, + "GTCGGT": 3522, + "GTCGGC": 3523, + "GTCGGG": 3524, + "GTGAAA": 3525, + "GTGAAT": 3526, + "GTGAAC": 3527, + "GTGAAG": 3528, + "GTGATA": 3529, + "GTGATT": 3530, + "GTGATC": 3531, + "GTGATG": 3532, + "GTGACA": 3533, + "GTGACT": 3534, + "GTGACC": 3535, + "GTGACG": 3536, + "GTGAGA": 3537, + "GTGAGT": 3538, + "GTGAGC": 3539, + "GTGAGG": 3540, + "GTGTAA": 3541, + "GTGTAT": 3542, + "GTGTAC": 3543, + "GTGTAG": 3544, + "GTGTTA": 3545, + "GTGTTT": 3546, + "GTGTTC": 3547, + "GTGTTG": 3548, + "GTGTCA": 3549, + "GTGTCT": 3550, + "GTGTCC": 3551, + "GTGTCG": 3552, + "GTGTGA": 3553, + "GTGTGT": 3554, + "GTGTGC": 3555, + "GTGTGG": 3556, + "GTGCAA": 3557, + "GTGCAT": 3558, + "GTGCAC": 3559, + "GTGCAG": 3560, + "GTGCTA": 3561, + "GTGCTT": 3562, + "GTGCTC": 3563, + "GTGCTG": 3564, + "GTGCCA": 3565, + "GTGCCT": 3566, + "GTGCCC": 3567, + "GTGCCG": 3568, + "GTGCGA": 3569, + "GTGCGT": 3570, + "GTGCGC": 3571, + "GTGCGG": 3572, + "GTGGAA": 3573, + "GTGGAT": 3574, + "GTGGAC": 3575, + "GTGGAG": 3576, + "GTGGTA": 3577, + "GTGGTT": 3578, + "GTGGTC": 3579, + "GTGGTG": 3580, + "GTGGCA": 3581, + "GTGGCT": 3582, + "GTGGCC": 3583, + "GTGGCG": 3584, + "GTGGGA": 3585, + "GTGGGT": 3586, + "GTGGGC": 3587, + "GTGGGG": 3588, + "GCAAAA": 3589, + "GCAAAT": 3590, + "GCAAAC": 3591, + "GCAAAG": 3592, + "GCAATA": 3593, + "GCAATT": 3594, + "GCAATC": 3595, + "GCAATG": 3596, + "GCAACA": 3597, + "GCAACT": 3598, + "GCAACC": 3599, + "GCAACG": 3600, + "GCAAGA": 3601, + "GCAAGT": 3602, + "GCAAGC": 3603, + "GCAAGG": 3604, + "GCATAA": 3605, + "GCATAT": 3606, + "GCATAC": 3607, + "GCATAG": 3608, + "GCATTA": 3609, + "GCATTT": 3610, + "GCATTC": 3611, + "GCATTG": 3612, + "GCATCA": 3613, + "GCATCT": 3614, + "GCATCC": 3615, + "GCATCG": 3616, + "GCATGA": 3617, + "GCATGT": 3618, + "GCATGC": 3619, + "GCATGG": 3620, + "GCACAA": 3621, + "GCACAT": 3622, + "GCACAC": 3623, + "GCACAG": 3624, + "GCACTA": 3625, + "GCACTT": 3626, + "GCACTC": 3627, + "GCACTG": 3628, + "GCACCA": 3629, + "GCACCT": 3630, + "GCACCC": 3631, + "GCACCG": 3632, + "GCACGA": 3633, + "GCACGT": 3634, + "GCACGC": 3635, + "GCACGG": 3636, + "GCAGAA": 3637, + "GCAGAT": 3638, + "GCAGAC": 3639, + "GCAGAG": 3640, + "GCAGTA": 3641, + "GCAGTT": 3642, + "GCAGTC": 3643, + "GCAGTG": 3644, + "GCAGCA": 3645, + "GCAGCT": 3646, + "GCAGCC": 3647, + "GCAGCG": 3648, + "GCAGGA": 3649, + "GCAGGT": 3650, + "GCAGGC": 3651, + "GCAGGG": 3652, + "GCTAAA": 3653, + "GCTAAT": 3654, + "GCTAAC": 3655, + "GCTAAG": 3656, + "GCTATA": 3657, + "GCTATT": 3658, + "GCTATC": 3659, + "GCTATG": 3660, + "GCTACA": 3661, + "GCTACT": 3662, + "GCTACC": 3663, + "GCTACG": 3664, + "GCTAGA": 3665, + "GCTAGT": 3666, + "GCTAGC": 3667, + "GCTAGG": 3668, + "GCTTAA": 3669, + "GCTTAT": 3670, + "GCTTAC": 3671, + "GCTTAG": 3672, + "GCTTTA": 3673, + "GCTTTT": 3674, + "GCTTTC": 3675, + "GCTTTG": 3676, + "GCTTCA": 3677, + "GCTTCT": 3678, + "GCTTCC": 3679, + "GCTTCG": 3680, + "GCTTGA": 3681, + "GCTTGT": 3682, + "GCTTGC": 3683, + "GCTTGG": 3684, + "GCTCAA": 3685, + "GCTCAT": 3686, + "GCTCAC": 3687, + "GCTCAG": 3688, + "GCTCTA": 3689, + "GCTCTT": 3690, + "GCTCTC": 3691, + "GCTCTG": 3692, + "GCTCCA": 3693, + "GCTCCT": 3694, + "GCTCCC": 3695, + "GCTCCG": 3696, + "GCTCGA": 3697, + "GCTCGT": 3698, + "GCTCGC": 3699, + "GCTCGG": 3700, + "GCTGAA": 3701, + "GCTGAT": 3702, + "GCTGAC": 3703, + "GCTGAG": 3704, + "GCTGTA": 3705, + "GCTGTT": 3706, + "GCTGTC": 3707, + "GCTGTG": 3708, + "GCTGCA": 3709, + "GCTGCT": 3710, + "GCTGCC": 3711, + "GCTGCG": 3712, + "GCTGGA": 3713, + "GCTGGT": 3714, + "GCTGGC": 3715, + "GCTGGG": 3716, + "GCCAAA": 3717, + "GCCAAT": 3718, + "GCCAAC": 3719, + "GCCAAG": 3720, + "GCCATA": 3721, + "GCCATT": 3722, + "GCCATC": 3723, + "GCCATG": 3724, + "GCCACA": 3725, + "GCCACT": 3726, + "GCCACC": 3727, + "GCCACG": 3728, + "GCCAGA": 3729, + "GCCAGT": 3730, + "GCCAGC": 3731, + "GCCAGG": 3732, + "GCCTAA": 3733, + "GCCTAT": 3734, + "GCCTAC": 3735, + "GCCTAG": 3736, + "GCCTTA": 3737, + "GCCTTT": 3738, + "GCCTTC": 3739, + "GCCTTG": 3740, + "GCCTCA": 3741, + "GCCTCT": 3742, + "GCCTCC": 3743, + "GCCTCG": 3744, + "GCCTGA": 3745, + "GCCTGT": 3746, + "GCCTGC": 3747, + "GCCTGG": 3748, + "GCCCAA": 3749, + "GCCCAT": 3750, + "GCCCAC": 3751, + "GCCCAG": 3752, + "GCCCTA": 3753, + "GCCCTT": 3754, + "GCCCTC": 3755, + "GCCCTG": 3756, + "GCCCCA": 3757, + "GCCCCT": 3758, + "GCCCCC": 3759, + "GCCCCG": 3760, + "GCCCGA": 3761, + "GCCCGT": 3762, + "GCCCGC": 3763, + "GCCCGG": 3764, + "GCCGAA": 3765, + "GCCGAT": 3766, + "GCCGAC": 3767, + "GCCGAG": 3768, + "GCCGTA": 3769, + "GCCGTT": 3770, + "GCCGTC": 3771, + "GCCGTG": 3772, + "GCCGCA": 3773, + "GCCGCT": 3774, + "GCCGCC": 3775, + "GCCGCG": 3776, + "GCCGGA": 3777, + "GCCGGT": 3778, + "GCCGGC": 3779, + "GCCGGG": 3780, + "GCGAAA": 3781, + "GCGAAT": 3782, + "GCGAAC": 3783, + "GCGAAG": 3784, + "GCGATA": 3785, + "GCGATT": 3786, + "GCGATC": 3787, + "GCGATG": 3788, + "GCGACA": 3789, + "GCGACT": 3790, + "GCGACC": 3791, + "GCGACG": 3792, + "GCGAGA": 3793, + "GCGAGT": 3794, + "GCGAGC": 3795, + "GCGAGG": 3796, + "GCGTAA": 3797, + "GCGTAT": 3798, + "GCGTAC": 3799, + "GCGTAG": 3800, + "GCGTTA": 3801, + "GCGTTT": 3802, + "GCGTTC": 3803, + "GCGTTG": 3804, + "GCGTCA": 3805, + "GCGTCT": 3806, + "GCGTCC": 3807, + "GCGTCG": 3808, + "GCGTGA": 3809, + "GCGTGT": 3810, + "GCGTGC": 3811, + "GCGTGG": 3812, + "GCGCAA": 3813, + "GCGCAT": 3814, + "GCGCAC": 3815, + "GCGCAG": 3816, + "GCGCTA": 3817, + "GCGCTT": 3818, + "GCGCTC": 3819, + "GCGCTG": 3820, + "GCGCCA": 3821, + "GCGCCT": 3822, + "GCGCCC": 3823, + "GCGCCG": 3824, + "GCGCGA": 3825, + "GCGCGT": 3826, + "GCGCGC": 3827, + "GCGCGG": 3828, + "GCGGAA": 3829, + "GCGGAT": 3830, + "GCGGAC": 3831, + "GCGGAG": 3832, + "GCGGTA": 3833, + "GCGGTT": 3834, + "GCGGTC": 3835, + "GCGGTG": 3836, + "GCGGCA": 3837, + "GCGGCT": 3838, + "GCGGCC": 3839, + "GCGGCG": 3840, + "GCGGGA": 3841, + "GCGGGT": 3842, + "GCGGGC": 3843, + "GCGGGG": 3844, + "GGAAAA": 3845, + "GGAAAT": 3846, + "GGAAAC": 3847, + "GGAAAG": 3848, + "GGAATA": 3849, + "GGAATT": 3850, + "GGAATC": 3851, + "GGAATG": 3852, + "GGAACA": 3853, + "GGAACT": 3854, + "GGAACC": 3855, + "GGAACG": 3856, + "GGAAGA": 3857, + "GGAAGT": 3858, + "GGAAGC": 3859, + "GGAAGG": 3860, + "GGATAA": 3861, + "GGATAT": 3862, + "GGATAC": 3863, + "GGATAG": 3864, + "GGATTA": 3865, + "GGATTT": 3866, + "GGATTC": 3867, + "GGATTG": 3868, + "GGATCA": 3869, + "GGATCT": 3870, + "GGATCC": 3871, + "GGATCG": 3872, + "GGATGA": 3873, + "GGATGT": 3874, + "GGATGC": 3875, + "GGATGG": 3876, + "GGACAA": 3877, + "GGACAT": 3878, + "GGACAC": 3879, + "GGACAG": 3880, + "GGACTA": 3881, + "GGACTT": 3882, + "GGACTC": 3883, + "GGACTG": 3884, + "GGACCA": 3885, + "GGACCT": 3886, + "GGACCC": 3887, + "GGACCG": 3888, + "GGACGA": 3889, + "GGACGT": 3890, + "GGACGC": 3891, + "GGACGG": 3892, + "GGAGAA": 3893, + "GGAGAT": 3894, + "GGAGAC": 3895, + "GGAGAG": 3896, + "GGAGTA": 3897, + "GGAGTT": 3898, + "GGAGTC": 3899, + "GGAGTG": 3900, + "GGAGCA": 3901, + "GGAGCT": 3902, + "GGAGCC": 3903, + "GGAGCG": 3904, + "GGAGGA": 3905, + "GGAGGT": 3906, + "GGAGGC": 3907, + "GGAGGG": 3908, + "GGTAAA": 3909, + "GGTAAT": 3910, + "GGTAAC": 3911, + "GGTAAG": 3912, + "GGTATA": 3913, + "GGTATT": 3914, + "GGTATC": 3915, + "GGTATG": 3916, + "GGTACA": 3917, + "GGTACT": 3918, + "GGTACC": 3919, + "GGTACG": 3920, + "GGTAGA": 3921, + "GGTAGT": 3922, + "GGTAGC": 3923, + "GGTAGG": 3924, + "GGTTAA": 3925, + "GGTTAT": 3926, + "GGTTAC": 3927, + "GGTTAG": 3928, + "GGTTTA": 3929, + "GGTTTT": 3930, + "GGTTTC": 3931, + "GGTTTG": 3932, + "GGTTCA": 3933, + "GGTTCT": 3934, + "GGTTCC": 3935, + "GGTTCG": 3936, + "GGTTGA": 3937, + "GGTTGT": 3938, + "GGTTGC": 3939, + "GGTTGG": 3940, + "GGTCAA": 3941, + "GGTCAT": 3942, + "GGTCAC": 3943, + "GGTCAG": 3944, + "GGTCTA": 3945, + "GGTCTT": 3946, + "GGTCTC": 3947, + "GGTCTG": 3948, + "GGTCCA": 3949, + "GGTCCT": 3950, + "GGTCCC": 3951, + "GGTCCG": 3952, + "GGTCGA": 3953, + "GGTCGT": 3954, + "GGTCGC": 3955, + "GGTCGG": 3956, + "GGTGAA": 3957, + "GGTGAT": 3958, + "GGTGAC": 3959, + "GGTGAG": 3960, + "GGTGTA": 3961, + "GGTGTT": 3962, + "GGTGTC": 3963, + "GGTGTG": 3964, + "GGTGCA": 3965, + "GGTGCT": 3966, + "GGTGCC": 3967, + "GGTGCG": 3968, + "GGTGGA": 3969, + "GGTGGT": 3970, + "GGTGGC": 3971, + "GGTGGG": 3972, + "GGCAAA": 3973, + "GGCAAT": 3974, + "GGCAAC": 3975, + "GGCAAG": 3976, + "GGCATA": 3977, + "GGCATT": 3978, + "GGCATC": 3979, + "GGCATG": 3980, + "GGCACA": 3981, + "GGCACT": 3982, + "GGCACC": 3983, + "GGCACG": 3984, + "GGCAGA": 3985, + "GGCAGT": 3986, + "GGCAGC": 3987, + "GGCAGG": 3988, + "GGCTAA": 3989, + "GGCTAT": 3990, + "GGCTAC": 3991, + "GGCTAG": 3992, + "GGCTTA": 3993, + "GGCTTT": 3994, + "GGCTTC": 3995, + "GGCTTG": 3996, + "GGCTCA": 3997, + "GGCTCT": 3998, + "GGCTCC": 3999, + "GGCTCG": 4000, + "GGCTGA": 4001, + "GGCTGT": 4002, + "GGCTGC": 4003, + "GGCTGG": 4004, + "GGCCAA": 4005, + "GGCCAT": 4006, + "GGCCAC": 4007, + "GGCCAG": 4008, + "GGCCTA": 4009, + "GGCCTT": 4010, + "GGCCTC": 4011, + "GGCCTG": 4012, + "GGCCCA": 4013, + "GGCCCT": 4014, + "GGCCCC": 4015, + "GGCCCG": 4016, + "GGCCGA": 4017, + "GGCCGT": 4018, + "GGCCGC": 4019, + "GGCCGG": 4020, + "GGCGAA": 4021, + "GGCGAT": 4022, + "GGCGAC": 4023, + "GGCGAG": 4024, + "GGCGTA": 4025, + "GGCGTT": 4026, + "GGCGTC": 4027, + "GGCGTG": 4028, + "GGCGCA": 4029, + "GGCGCT": 4030, + "GGCGCC": 4031, + "GGCGCG": 4032, + "GGCGGA": 4033, + "GGCGGT": 4034, + "GGCGGC": 4035, + "GGCGGG": 4036, + "GGGAAA": 4037, + "GGGAAT": 4038, + "GGGAAC": 4039, + "GGGAAG": 4040, + "GGGATA": 4041, + "GGGATT": 4042, + "GGGATC": 4043, + "GGGATG": 4044, + "GGGACA": 4045, + "GGGACT": 4046, + "GGGACC": 4047, + "GGGACG": 4048, + "GGGAGA": 4049, + "GGGAGT": 4050, + "GGGAGC": 4051, + "GGGAGG": 4052, + "GGGTAA": 4053, + "GGGTAT": 4054, + "GGGTAC": 4055, + "GGGTAG": 4056, + "GGGTTA": 4057, + "GGGTTT": 4058, + "GGGTTC": 4059, + "GGGTTG": 4060, + "GGGTCA": 4061, + "GGGTCT": 4062, + "GGGTCC": 4063, + "GGGTCG": 4064, + "GGGTGA": 4065, + "GGGTGT": 4066, + "GGGTGC": 4067, + "GGGTGG": 4068, + "GGGCAA": 4069, + "GGGCAT": 4070, + "GGGCAC": 4071, + "GGGCAG": 4072, + "GGGCTA": 4073, + "GGGCTT": 4074, + "GGGCTC": 4075, + "GGGCTG": 4076, + "GGGCCA": 4077, + "GGGCCT": 4078, + "GGGCCC": 4079, + "GGGCCG": 4080, + "GGGCGA": 4081, + "GGGCGT": 4082, + "GGGCGC": 4083, + "GGGCGG": 4084, + "GGGGAA": 4085, + "GGGGAT": 4086, + "GGGGAC": 4087, + "GGGGAG": 4088, + "GGGGTA": 4089, + "GGGGTT": 4090, + "GGGGTC": 4091, + "GGGGTG": 4092, + "GGGGCA": 4093, + "GGGGCT": 4094, + "GGGGCC": 4095, + "GGGGCG": 4096, + "GGGGGA": 4097, + "GGGGGT": 4098, + "GGGGGC": 4099, + "GGGGGG": 4100 + } + } +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-3000/tokenizer_config.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-3000/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..604e2b1b29af1dcb3920d2828b6a90276e5e3d02 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-3000/tokenizer_config.json @@ -0,0 +1,58 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "[UNK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "[CLS]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "[SEP]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "4": { + "content": "[MASK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "clean_up_tokenization_spaces": true, + "cls_token": "[CLS]", + "do_basic_tokenize": true, + "do_lower_case": false, + "mask_token": "[MASK]", + "max_len": 512, + "model_max_length": 512, + "never_split": null, + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "strip_accents": null, + "tokenize_chinese_chars": true, + "tokenizer_class": "BertTokenizer", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-3000/trainer_state.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-3000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..57fec5519a3992240118c78b1ea46044b3ce0d7b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-3000/trainer_state.json @@ -0,0 +1,873 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 12.0, + "eval_steps": 500, + "global_step": 3000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.1, + "grad_norm": 0.6038870215415955, + "learning_rate": 5.319148936170213e-05, + "loss": 0.6945, + "step": 25 + }, + { + "epoch": 0.2, + "grad_norm": 0.6668091416358948, + "learning_rate": 0.00010638297872340425, + "loss": 0.6929, + "step": 50 + }, + { + "epoch": 0.3, + "grad_norm": 0.8282511234283447, + "learning_rate": 0.00015957446808510637, + "loss": 0.6925, + "step": 75 + }, + { + "epoch": 0.4, + "grad_norm": 1.3149235248565674, + "learning_rate": 0.0002127659574468085, + "loss": 0.6919, + "step": 100 + }, + { + "epoch": 0.5, + "grad_norm": 1.8841172456741333, + "learning_rate": 0.00026595744680851064, + "loss": 0.6788, + "step": 125 + }, + { + "epoch": 0.6, + "grad_norm": 4.435206413269043, + "learning_rate": 0.00031914893617021275, + "loss": 0.6285, + "step": 150 + }, + { + "epoch": 0.7, + "grad_norm": 8.858148574829102, + "learning_rate": 0.0003723404255319149, + "loss": 0.5386, + "step": 175 + }, + { + "epoch": 0.8, + "grad_norm": 5.627648830413818, + "learning_rate": 0.0003992081821181128, + "loss": 0.5537, + "step": 200 + }, + { + "epoch": 0.9, + "grad_norm": 5.086580753326416, + "learning_rate": 0.0003975585615308479, + "loss": 0.5595, + "step": 225 + }, + { + "epoch": 1.0, + "grad_norm": 5.627719402313232, + "learning_rate": 0.00039590894094358297, + "loss": 0.4824, + "step": 250 + }, + { + "epoch": 1.1, + "grad_norm": 5.888038158416748, + "learning_rate": 0.0003942593203563181, + "loss": 0.4404, + "step": 275 + }, + { + "epoch": 1.2, + "grad_norm": 5.646231651306152, + "learning_rate": 0.0003926096997690532, + "loss": 0.4354, + "step": 300 + }, + { + "epoch": 1.3, + "grad_norm": 5.191990852355957, + "learning_rate": 0.00039096007918178817, + "loss": 0.4469, + "step": 325 + }, + { + "epoch": 1.4, + "grad_norm": 13.90101432800293, + "learning_rate": 0.0003893104585945233, + "loss": 0.3858, + "step": 350 + }, + { + "epoch": 1.5, + "grad_norm": 17.75942039489746, + "learning_rate": 0.0003876608380072583, + "loss": 0.4131, + "step": 375 + }, + { + "epoch": 1.6, + "grad_norm": 7.398252487182617, + "learning_rate": 0.00038601121741999343, + "loss": 0.3842, + "step": 400 + }, + { + "epoch": 1.7, + "grad_norm": 11.010066986083984, + "learning_rate": 0.0003843615968327285, + "loss": 0.3959, + "step": 425 + }, + { + "epoch": 1.8, + "grad_norm": 9.277341842651367, + "learning_rate": 0.0003827119762454636, + "loss": 0.3552, + "step": 450 + }, + { + "epoch": 1.9, + "grad_norm": 26.011449813842773, + "learning_rate": 0.00038106235565819863, + "loss": 0.3961, + "step": 475 + }, + { + "epoch": 2.0, + "grad_norm": 11.123054504394531, + "learning_rate": 0.0003794127350709337, + "loss": 0.3548, + "step": 500 + }, + { + "epoch": 2.1, + "grad_norm": 20.128774642944336, + "learning_rate": 0.0003777631144836688, + "loss": 0.302, + "step": 525 + }, + { + "epoch": 2.2, + "grad_norm": 4.64390230178833, + "learning_rate": 0.00037611349389640383, + "loss": 0.2853, + "step": 550 + }, + { + "epoch": 2.3, + "grad_norm": 4.199934482574463, + "learning_rate": 0.00037446387330913894, + "loss": 0.2724, + "step": 575 + }, + { + "epoch": 2.4, + "grad_norm": 6.909411430358887, + "learning_rate": 0.000372814252721874, + "loss": 0.2976, + "step": 600 + }, + { + "epoch": 2.5, + "grad_norm": 9.234820365905762, + "learning_rate": 0.00037116463213460903, + "loss": 0.2798, + "step": 625 + }, + { + "epoch": 2.6, + "grad_norm": 10.472943305969238, + "learning_rate": 0.00036951501154734414, + "loss": 0.3007, + "step": 650 + }, + { + "epoch": 2.7, + "grad_norm": 19.46732521057129, + "learning_rate": 0.0003678653909600792, + "loss": 0.2701, + "step": 675 + }, + { + "epoch": 2.8, + "grad_norm": 7.501255512237549, + "learning_rate": 0.0003662157703728143, + "loss": 0.2896, + "step": 700 + }, + { + "epoch": 2.9, + "grad_norm": 10.645585060119629, + "learning_rate": 0.00036456614978554934, + "loss": 0.2748, + "step": 725 + }, + { + "epoch": 3.0, + "grad_norm": 30.26616859436035, + "learning_rate": 0.00036291652919828444, + "loss": 0.268, + "step": 750 + }, + { + "epoch": 3.1, + "grad_norm": 14.13516616821289, + "learning_rate": 0.00036126690861101944, + "loss": 0.2159, + "step": 775 + }, + { + "epoch": 3.2, + "grad_norm": 10.091023445129395, + "learning_rate": 0.00035961728802375454, + "loss": 0.2071, + "step": 800 + }, + { + "epoch": 3.3, + "grad_norm": 22.201452255249023, + "learning_rate": 0.0003579676674364896, + "loss": 0.2146, + "step": 825 + }, + { + "epoch": 3.4, + "grad_norm": 10.044412612915039, + "learning_rate": 0.0003563180468492247, + "loss": 0.1827, + "step": 850 + }, + { + "epoch": 3.5, + "grad_norm": 21.04941177368164, + "learning_rate": 0.0003546684262619598, + "loss": 0.202, + "step": 875 + }, + { + "epoch": 3.6, + "grad_norm": 11.754891395568848, + "learning_rate": 0.00035301880567469485, + "loss": 0.2021, + "step": 900 + }, + { + "epoch": 3.7, + "grad_norm": 9.113747596740723, + "learning_rate": 0.0003513691850874299, + "loss": 0.2067, + "step": 925 + }, + { + "epoch": 3.8, + "grad_norm": 9.075196266174316, + "learning_rate": 0.00034971956450016495, + "loss": 0.3285, + "step": 950 + }, + { + "epoch": 3.9, + "grad_norm": 8.568796157836914, + "learning_rate": 0.00034806994391290005, + "loss": 0.2254, + "step": 975 + }, + { + "epoch": 4.0, + "grad_norm": 3.9587976932525635, + "learning_rate": 0.0003464203233256351, + "loss": 0.2245, + "step": 1000 + }, + { + "epoch": 4.1, + "grad_norm": 11.143651008605957, + "learning_rate": 0.0003447707027383702, + "loss": 0.1755, + "step": 1025 + }, + { + "epoch": 4.2, + "grad_norm": 12.53118896484375, + "learning_rate": 0.00034312108215110525, + "loss": 0.1808, + "step": 1050 + }, + { + "epoch": 4.3, + "grad_norm": 9.880820274353027, + "learning_rate": 0.0003414714615638403, + "loss": 0.1387, + "step": 1075 + }, + { + "epoch": 4.4, + "grad_norm": 5.729593753814697, + "learning_rate": 0.0003398218409765754, + "loss": 0.1958, + "step": 1100 + }, + { + "epoch": 4.5, + "grad_norm": 8.633047103881836, + "learning_rate": 0.00033817222038931045, + "loss": 0.1515, + "step": 1125 + }, + { + "epoch": 4.6, + "grad_norm": 19.237892150878906, + "learning_rate": 0.00033652259980204556, + "loss": 0.1452, + "step": 1150 + }, + { + "epoch": 4.7, + "grad_norm": 6.129818439483643, + "learning_rate": 0.0003348729792147806, + "loss": 0.1617, + "step": 1175 + }, + { + "epoch": 4.8, + "grad_norm": 5.073274612426758, + "learning_rate": 0.0003332233586275157, + "loss": 0.1967, + "step": 1200 + }, + { + "epoch": 4.9, + "grad_norm": 5.729922771453857, + "learning_rate": 0.00033157373804025076, + "loss": 0.1482, + "step": 1225 + }, + { + "epoch": 5.0, + "grad_norm": 12.314518928527832, + "learning_rate": 0.0003299241174529858, + "loss": 0.1588, + "step": 1250 + }, + { + "epoch": 5.1, + "grad_norm": 3.2636606693267822, + "learning_rate": 0.0003282744968657209, + "loss": 0.1588, + "step": 1275 + }, + { + "epoch": 5.2, + "grad_norm": 5.914486408233643, + "learning_rate": 0.00032662487627845596, + "loss": 0.1288, + "step": 1300 + }, + { + "epoch": 5.3, + "grad_norm": 13.849296569824219, + "learning_rate": 0.00032497525569119106, + "loss": 0.0961, + "step": 1325 + }, + { + "epoch": 5.4, + "grad_norm": 7.7317328453063965, + "learning_rate": 0.0003233256351039261, + "loss": 0.1148, + "step": 1350 + }, + { + "epoch": 5.5, + "grad_norm": 1.768608808517456, + "learning_rate": 0.00032167601451666116, + "loss": 0.121, + "step": 1375 + }, + { + "epoch": 5.6, + "grad_norm": 10.802785873413086, + "learning_rate": 0.00032002639392939627, + "loss": 0.1786, + "step": 1400 + }, + { + "epoch": 5.7, + "grad_norm": 4.158301830291748, + "learning_rate": 0.0003183767733421313, + "loss": 0.1078, + "step": 1425 + }, + { + "epoch": 5.8, + "grad_norm": 10.65804672241211, + "learning_rate": 0.0003167271527548664, + "loss": 0.1597, + "step": 1450 + }, + { + "epoch": 5.9, + "grad_norm": 10.880438804626465, + "learning_rate": 0.00031507753216760147, + "loss": 0.1225, + "step": 1475 + }, + { + "epoch": 6.0, + "grad_norm": 9.95138168334961, + "learning_rate": 0.00031342791158033657, + "loss": 0.1254, + "step": 1500 + }, + { + "epoch": 6.1, + "grad_norm": 1.5018162727355957, + "learning_rate": 0.00031177829099307157, + "loss": 0.0902, + "step": 1525 + }, + { + "epoch": 6.2, + "grad_norm": 14.281820297241211, + "learning_rate": 0.00031012867040580667, + "loss": 0.1026, + "step": 1550 + }, + { + "epoch": 6.3, + "grad_norm": 4.7540435791015625, + "learning_rate": 0.0003084790498185417, + "loss": 0.0876, + "step": 1575 + }, + { + "epoch": 6.4, + "grad_norm": 4.304183006286621, + "learning_rate": 0.0003068294292312768, + "loss": 0.1355, + "step": 1600 + }, + { + "epoch": 6.5, + "grad_norm": 14.653755187988281, + "learning_rate": 0.0003051798086440119, + "loss": 0.0878, + "step": 1625 + }, + { + "epoch": 6.6, + "grad_norm": 11.13028335571289, + "learning_rate": 0.000303530188056747, + "loss": 0.1203, + "step": 1650 + }, + { + "epoch": 6.7, + "grad_norm": 9.729938507080078, + "learning_rate": 0.000301880567469482, + "loss": 0.0775, + "step": 1675 + }, + { + "epoch": 6.8, + "grad_norm": 10.12938117980957, + "learning_rate": 0.0003002309468822171, + "loss": 0.1015, + "step": 1700 + }, + { + "epoch": 6.9, + "grad_norm": 8.506614685058594, + "learning_rate": 0.0002985813262949522, + "loss": 0.1362, + "step": 1725 + }, + { + "epoch": 7.0, + "grad_norm": 25.618762969970703, + "learning_rate": 0.00029693170570768723, + "loss": 0.1269, + "step": 1750 + }, + { + "epoch": 7.1, + "grad_norm": 27.155860900878906, + "learning_rate": 0.00029528208512042233, + "loss": 0.0555, + "step": 1775 + }, + { + "epoch": 7.2, + "grad_norm": 9.412527084350586, + "learning_rate": 0.0002936324645331574, + "loss": 0.0695, + "step": 1800 + }, + { + "epoch": 7.3, + "grad_norm": 6.4328293800354, + "learning_rate": 0.00029198284394589243, + "loss": 0.0808, + "step": 1825 + }, + { + "epoch": 7.4, + "grad_norm": 15.267293930053711, + "learning_rate": 0.00029033322335862753, + "loss": 0.0871, + "step": 1850 + }, + { + "epoch": 7.5, + "grad_norm": 6.774560928344727, + "learning_rate": 0.0002886836027713626, + "loss": 0.0819, + "step": 1875 + }, + { + "epoch": 7.6, + "grad_norm": 18.67180633544922, + "learning_rate": 0.0002870339821840977, + "loss": 0.0565, + "step": 1900 + }, + { + "epoch": 7.7, + "grad_norm": 16.39004898071289, + "learning_rate": 0.00028538436159683273, + "loss": 0.0907, + "step": 1925 + }, + { + "epoch": 7.8, + "grad_norm": 6.351583003997803, + "learning_rate": 0.00028373474100956784, + "loss": 0.1289, + "step": 1950 + }, + { + "epoch": 7.9, + "grad_norm": 11.396740913391113, + "learning_rate": 0.0002820851204223029, + "loss": 0.1149, + "step": 1975 + }, + { + "epoch": 8.0, + "grad_norm": 6.491570949554443, + "learning_rate": 0.00028043549983503794, + "loss": 0.0891, + "step": 2000 + }, + { + "epoch": 8.1, + "grad_norm": 0.22946956753730774, + "learning_rate": 0.00027878587924777304, + "loss": 0.0413, + "step": 2025 + }, + { + "epoch": 8.2, + "grad_norm": 42.31606674194336, + "learning_rate": 0.0002771362586605081, + "loss": 0.0881, + "step": 2050 + }, + { + "epoch": 8.3, + "grad_norm": 18.490406036376953, + "learning_rate": 0.0002754866380732432, + "loss": 0.1238, + "step": 2075 + }, + { + "epoch": 8.4, + "grad_norm": 17.950319290161133, + "learning_rate": 0.00027383701748597824, + "loss": 0.0642, + "step": 2100 + }, + { + "epoch": 8.5, + "grad_norm": 37.16359329223633, + "learning_rate": 0.0002721873968987133, + "loss": 0.0916, + "step": 2125 + }, + { + "epoch": 8.6, + "grad_norm": 39.75199890136719, + "learning_rate": 0.00027053777631144834, + "loss": 0.1087, + "step": 2150 + }, + { + "epoch": 8.7, + "grad_norm": 0.08794238418340683, + "learning_rate": 0.00026888815572418344, + "loss": 0.0825, + "step": 2175 + }, + { + "epoch": 8.8, + "grad_norm": 14.681429862976074, + "learning_rate": 0.00026723853513691855, + "loss": 0.0925, + "step": 2200 + }, + { + "epoch": 8.9, + "grad_norm": 17.90182876586914, + "learning_rate": 0.0002655889145496536, + "loss": 0.0933, + "step": 2225 + }, + { + "epoch": 9.0, + "grad_norm": 0.9108687043190002, + "learning_rate": 0.0002639392939623887, + "loss": 0.0602, + "step": 2250 + }, + { + "epoch": 9.1, + "grad_norm": 9.671142578125, + "learning_rate": 0.0002622896733751237, + "loss": 0.0438, + "step": 2275 + }, + { + "epoch": 9.2, + "grad_norm": 12.797035217285156, + "learning_rate": 0.0002606400527878588, + "loss": 0.0495, + "step": 2300 + }, + { + "epoch": 9.3, + "grad_norm": 0.4585617184638977, + "learning_rate": 0.00025899043220059385, + "loss": 0.032, + "step": 2325 + }, + { + "epoch": 9.4, + "grad_norm": 0.05073042958974838, + "learning_rate": 0.00025734081161332895, + "loss": 0.067, + "step": 2350 + }, + { + "epoch": 9.5, + "grad_norm": 0.013653877191245556, + "learning_rate": 0.000255691191026064, + "loss": 0.0898, + "step": 2375 + }, + { + "epoch": 9.6, + "grad_norm": 18.198965072631836, + "learning_rate": 0.0002540415704387991, + "loss": 0.0804, + "step": 2400 + }, + { + "epoch": 9.7, + "grad_norm": 0.1363959163427353, + "learning_rate": 0.00025239194985153415, + "loss": 0.0584, + "step": 2425 + }, + { + "epoch": 9.8, + "grad_norm": 17.61290740966797, + "learning_rate": 0.0002507423292642692, + "loss": 0.0778, + "step": 2450 + }, + { + "epoch": 9.9, + "grad_norm": 0.10548887401819229, + "learning_rate": 0.0002490927086770043, + "loss": 0.0667, + "step": 2475 + }, + { + "epoch": 10.0, + "grad_norm": 0.1262211948633194, + "learning_rate": 0.00024744308808973936, + "loss": 0.0625, + "step": 2500 + }, + { + "epoch": 10.1, + "grad_norm": 0.019293297082185745, + "learning_rate": 0.00024579346750247446, + "loss": 0.0762, + "step": 2525 + }, + { + "epoch": 10.2, + "grad_norm": 0.9273566603660583, + "learning_rate": 0.0002441438469152095, + "loss": 0.0357, + "step": 2550 + }, + { + "epoch": 10.3, + "grad_norm": 10.720550537109375, + "learning_rate": 0.00024249422632794456, + "loss": 0.046, + "step": 2575 + }, + { + "epoch": 10.4, + "grad_norm": 1.8441611528396606, + "learning_rate": 0.00024084460574067963, + "loss": 0.0369, + "step": 2600 + }, + { + "epoch": 10.5, + "grad_norm": 0.8760197758674622, + "learning_rate": 0.0002391949851534147, + "loss": 0.046, + "step": 2625 + }, + { + "epoch": 10.6, + "grad_norm": 8.899499893188477, + "learning_rate": 0.0002375453645661498, + "loss": 0.08, + "step": 2650 + }, + { + "epoch": 10.7, + "grad_norm": 0.2664288580417633, + "learning_rate": 0.00023589574397888486, + "loss": 0.0616, + "step": 2675 + }, + { + "epoch": 10.8, + "grad_norm": 0.06609692424535751, + "learning_rate": 0.00023424612339161997, + "loss": 0.05, + "step": 2700 + }, + { + "epoch": 10.9, + "grad_norm": 32.44829177856445, + "learning_rate": 0.000232596502804355, + "loss": 0.0487, + "step": 2725 + }, + { + "epoch": 11.0, + "grad_norm": 7.25365686416626, + "learning_rate": 0.00023094688221709007, + "loss": 0.0456, + "step": 2750 + }, + { + "epoch": 11.1, + "grad_norm": 0.10101272910833359, + "learning_rate": 0.00022929726162982514, + "loss": 0.0328, + "step": 2775 + }, + { + "epoch": 11.2, + "grad_norm": 1.3087996244430542, + "learning_rate": 0.00022764764104256022, + "loss": 0.0606, + "step": 2800 + }, + { + "epoch": 11.3, + "grad_norm": 1.247656226158142, + "learning_rate": 0.0002259980204552953, + "loss": 0.0838, + "step": 2825 + }, + { + "epoch": 11.4, + "grad_norm": 1.7708797454833984, + "learning_rate": 0.00022434839986803037, + "loss": 0.0416, + "step": 2850 + }, + { + "epoch": 11.5, + "grad_norm": 11.131721496582031, + "learning_rate": 0.00022269877928076542, + "loss": 0.0528, + "step": 2875 + }, + { + "epoch": 11.6, + "grad_norm": 3.824920892715454, + "learning_rate": 0.0002210491586935005, + "loss": 0.0526, + "step": 2900 + }, + { + "epoch": 11.7, + "grad_norm": 0.8507682085037231, + "learning_rate": 0.00021939953810623557, + "loss": 0.0965, + "step": 2925 + }, + { + "epoch": 11.8, + "grad_norm": 0.011997719295322895, + "learning_rate": 0.00021774991751897065, + "loss": 0.0446, + "step": 2950 + }, + { + "epoch": 11.9, + "grad_norm": 0.0792650505900383, + "learning_rate": 0.00021610029693170573, + "loss": 0.0556, + "step": 2975 + }, + { + "epoch": 12.0, + "grad_norm": 0.007047992665320635, + "learning_rate": 0.0002144506763444408, + "loss": 0.0759, + "step": 3000 + } + ], + "logging_steps": 25, + "max_steps": 6250, + "num_input_tokens_seen": 0, + "num_train_epochs": 25, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 152458444800000.0, + "train_batch_size": 20, + "trial_name": null, + "trial_params": null +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-3000/training_args.bin b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-3000/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..77692b6666e83a688b38e799c964e4694a57eec9 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-3000/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f73b41315068a430d2439f5707c62152a34210df62454b9549f9642446dd90b2 +size 5304 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-3000/vocab.txt b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-3000/vocab.txt new file mode 100644 index 0000000000000000000000000000000000000000..cc9ad53732e21ea9180f18e1fb67a38cc5702d4c --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-3000/vocab.txt @@ -0,0 +1,4101 @@ +[PAD] +[UNK] +[CLS] +[SEP] +[MASK] +AAAAAA +AAAAAT +AAAAAC +AAAAAG +AAAATA +AAAATT +AAAATC +AAAATG +AAAACA +AAAACT +AAAACC +AAAACG +AAAAGA +AAAAGT +AAAAGC +AAAAGG +AAATAA +AAATAT +AAATAC +AAATAG +AAATTA +AAATTT +AAATTC +AAATTG +AAATCA +AAATCT +AAATCC +AAATCG +AAATGA +AAATGT +AAATGC +AAATGG +AAACAA +AAACAT +AAACAC +AAACAG +AAACTA +AAACTT +AAACTC +AAACTG +AAACCA +AAACCT +AAACCC +AAACCG +AAACGA +AAACGT +AAACGC +AAACGG +AAAGAA +AAAGAT +AAAGAC +AAAGAG +AAAGTA +AAAGTT +AAAGTC +AAAGTG +AAAGCA +AAAGCT +AAAGCC +AAAGCG +AAAGGA +AAAGGT +AAAGGC +AAAGGG +AATAAA +AATAAT +AATAAC +AATAAG +AATATA +AATATT +AATATC +AATATG +AATACA +AATACT +AATACC +AATACG +AATAGA +AATAGT +AATAGC +AATAGG +AATTAA +AATTAT +AATTAC +AATTAG +AATTTA +AATTTT +AATTTC +AATTTG +AATTCA +AATTCT +AATTCC +AATTCG +AATTGA +AATTGT +AATTGC +AATTGG +AATCAA +AATCAT +AATCAC +AATCAG +AATCTA +AATCTT +AATCTC +AATCTG +AATCCA +AATCCT +AATCCC +AATCCG +AATCGA +AATCGT +AATCGC +AATCGG +AATGAA +AATGAT +AATGAC +AATGAG +AATGTA +AATGTT +AATGTC +AATGTG +AATGCA +AATGCT +AATGCC +AATGCG +AATGGA +AATGGT +AATGGC +AATGGG +AACAAA +AACAAT +AACAAC +AACAAG +AACATA +AACATT +AACATC +AACATG +AACACA +AACACT +AACACC +AACACG +AACAGA +AACAGT +AACAGC +AACAGG +AACTAA +AACTAT +AACTAC +AACTAG +AACTTA +AACTTT +AACTTC +AACTTG +AACTCA +AACTCT +AACTCC +AACTCG +AACTGA +AACTGT +AACTGC +AACTGG +AACCAA +AACCAT +AACCAC +AACCAG +AACCTA +AACCTT +AACCTC +AACCTG +AACCCA +AACCCT +AACCCC +AACCCG +AACCGA +AACCGT +AACCGC +AACCGG +AACGAA +AACGAT +AACGAC +AACGAG +AACGTA +AACGTT +AACGTC +AACGTG +AACGCA +AACGCT +AACGCC +AACGCG +AACGGA +AACGGT +AACGGC +AACGGG +AAGAAA +AAGAAT +AAGAAC +AAGAAG +AAGATA +AAGATT +AAGATC +AAGATG +AAGACA +AAGACT +AAGACC +AAGACG +AAGAGA +AAGAGT +AAGAGC +AAGAGG +AAGTAA +AAGTAT +AAGTAC +AAGTAG +AAGTTA +AAGTTT +AAGTTC +AAGTTG +AAGTCA +AAGTCT +AAGTCC +AAGTCG +AAGTGA +AAGTGT +AAGTGC +AAGTGG +AAGCAA +AAGCAT +AAGCAC +AAGCAG +AAGCTA +AAGCTT +AAGCTC +AAGCTG +AAGCCA +AAGCCT +AAGCCC +AAGCCG +AAGCGA +AAGCGT +AAGCGC +AAGCGG +AAGGAA +AAGGAT +AAGGAC +AAGGAG +AAGGTA +AAGGTT +AAGGTC +AAGGTG +AAGGCA +AAGGCT +AAGGCC +AAGGCG +AAGGGA +AAGGGT +AAGGGC +AAGGGG +ATAAAA +ATAAAT +ATAAAC +ATAAAG +ATAATA +ATAATT +ATAATC +ATAATG +ATAACA +ATAACT +ATAACC +ATAACG +ATAAGA +ATAAGT +ATAAGC +ATAAGG +ATATAA +ATATAT +ATATAC +ATATAG +ATATTA +ATATTT +ATATTC +ATATTG +ATATCA +ATATCT +ATATCC +ATATCG +ATATGA +ATATGT +ATATGC +ATATGG +ATACAA +ATACAT +ATACAC +ATACAG +ATACTA +ATACTT +ATACTC +ATACTG +ATACCA +ATACCT +ATACCC +ATACCG +ATACGA +ATACGT +ATACGC +ATACGG +ATAGAA +ATAGAT +ATAGAC +ATAGAG +ATAGTA +ATAGTT +ATAGTC +ATAGTG +ATAGCA +ATAGCT +ATAGCC +ATAGCG +ATAGGA +ATAGGT +ATAGGC +ATAGGG +ATTAAA +ATTAAT +ATTAAC +ATTAAG +ATTATA +ATTATT +ATTATC +ATTATG +ATTACA +ATTACT +ATTACC +ATTACG +ATTAGA +ATTAGT +ATTAGC +ATTAGG +ATTTAA +ATTTAT +ATTTAC +ATTTAG +ATTTTA +ATTTTT +ATTTTC +ATTTTG +ATTTCA +ATTTCT +ATTTCC +ATTTCG +ATTTGA +ATTTGT +ATTTGC +ATTTGG +ATTCAA +ATTCAT +ATTCAC +ATTCAG +ATTCTA +ATTCTT +ATTCTC +ATTCTG +ATTCCA +ATTCCT +ATTCCC +ATTCCG +ATTCGA +ATTCGT +ATTCGC +ATTCGG +ATTGAA +ATTGAT +ATTGAC +ATTGAG +ATTGTA +ATTGTT +ATTGTC +ATTGTG +ATTGCA +ATTGCT +ATTGCC +ATTGCG +ATTGGA +ATTGGT +ATTGGC +ATTGGG +ATCAAA +ATCAAT +ATCAAC +ATCAAG +ATCATA +ATCATT +ATCATC +ATCATG +ATCACA +ATCACT +ATCACC +ATCACG +ATCAGA +ATCAGT +ATCAGC +ATCAGG +ATCTAA +ATCTAT +ATCTAC +ATCTAG +ATCTTA +ATCTTT +ATCTTC +ATCTTG +ATCTCA +ATCTCT +ATCTCC +ATCTCG +ATCTGA +ATCTGT +ATCTGC +ATCTGG +ATCCAA +ATCCAT +ATCCAC +ATCCAG +ATCCTA +ATCCTT +ATCCTC +ATCCTG +ATCCCA +ATCCCT +ATCCCC +ATCCCG +ATCCGA +ATCCGT +ATCCGC +ATCCGG +ATCGAA +ATCGAT +ATCGAC +ATCGAG +ATCGTA +ATCGTT +ATCGTC +ATCGTG +ATCGCA +ATCGCT +ATCGCC +ATCGCG +ATCGGA +ATCGGT +ATCGGC +ATCGGG +ATGAAA +ATGAAT +ATGAAC +ATGAAG +ATGATA +ATGATT +ATGATC +ATGATG +ATGACA +ATGACT +ATGACC +ATGACG +ATGAGA +ATGAGT +ATGAGC +ATGAGG +ATGTAA +ATGTAT +ATGTAC +ATGTAG +ATGTTA +ATGTTT +ATGTTC +ATGTTG +ATGTCA +ATGTCT +ATGTCC +ATGTCG +ATGTGA +ATGTGT +ATGTGC +ATGTGG +ATGCAA +ATGCAT +ATGCAC +ATGCAG +ATGCTA +ATGCTT +ATGCTC +ATGCTG +ATGCCA +ATGCCT +ATGCCC +ATGCCG +ATGCGA +ATGCGT +ATGCGC +ATGCGG +ATGGAA +ATGGAT +ATGGAC +ATGGAG +ATGGTA +ATGGTT +ATGGTC +ATGGTG +ATGGCA +ATGGCT +ATGGCC +ATGGCG +ATGGGA +ATGGGT +ATGGGC +ATGGGG +ACAAAA +ACAAAT +ACAAAC +ACAAAG +ACAATA +ACAATT +ACAATC +ACAATG +ACAACA +ACAACT +ACAACC +ACAACG +ACAAGA +ACAAGT +ACAAGC +ACAAGG +ACATAA +ACATAT +ACATAC +ACATAG +ACATTA +ACATTT +ACATTC +ACATTG +ACATCA +ACATCT +ACATCC +ACATCG +ACATGA +ACATGT +ACATGC +ACATGG +ACACAA +ACACAT +ACACAC +ACACAG +ACACTA +ACACTT +ACACTC +ACACTG +ACACCA +ACACCT +ACACCC +ACACCG +ACACGA +ACACGT +ACACGC +ACACGG +ACAGAA +ACAGAT +ACAGAC +ACAGAG +ACAGTA +ACAGTT +ACAGTC +ACAGTG +ACAGCA +ACAGCT +ACAGCC +ACAGCG +ACAGGA +ACAGGT +ACAGGC +ACAGGG +ACTAAA +ACTAAT +ACTAAC +ACTAAG +ACTATA +ACTATT +ACTATC +ACTATG +ACTACA +ACTACT +ACTACC +ACTACG +ACTAGA +ACTAGT +ACTAGC +ACTAGG +ACTTAA +ACTTAT +ACTTAC +ACTTAG +ACTTTA +ACTTTT +ACTTTC +ACTTTG +ACTTCA +ACTTCT +ACTTCC +ACTTCG +ACTTGA +ACTTGT +ACTTGC +ACTTGG +ACTCAA +ACTCAT +ACTCAC +ACTCAG +ACTCTA +ACTCTT +ACTCTC +ACTCTG +ACTCCA +ACTCCT +ACTCCC +ACTCCG +ACTCGA +ACTCGT +ACTCGC +ACTCGG +ACTGAA +ACTGAT +ACTGAC +ACTGAG +ACTGTA +ACTGTT +ACTGTC +ACTGTG +ACTGCA +ACTGCT +ACTGCC +ACTGCG +ACTGGA +ACTGGT +ACTGGC +ACTGGG +ACCAAA +ACCAAT +ACCAAC +ACCAAG +ACCATA +ACCATT +ACCATC +ACCATG +ACCACA +ACCACT +ACCACC +ACCACG +ACCAGA +ACCAGT +ACCAGC +ACCAGG +ACCTAA +ACCTAT +ACCTAC +ACCTAG +ACCTTA +ACCTTT +ACCTTC +ACCTTG +ACCTCA +ACCTCT +ACCTCC +ACCTCG +ACCTGA +ACCTGT +ACCTGC +ACCTGG +ACCCAA +ACCCAT +ACCCAC +ACCCAG +ACCCTA +ACCCTT +ACCCTC +ACCCTG +ACCCCA +ACCCCT +ACCCCC +ACCCCG +ACCCGA +ACCCGT +ACCCGC +ACCCGG +ACCGAA +ACCGAT +ACCGAC +ACCGAG +ACCGTA +ACCGTT +ACCGTC +ACCGTG +ACCGCA +ACCGCT +ACCGCC +ACCGCG +ACCGGA +ACCGGT +ACCGGC +ACCGGG +ACGAAA +ACGAAT +ACGAAC +ACGAAG +ACGATA +ACGATT +ACGATC +ACGATG +ACGACA +ACGACT +ACGACC +ACGACG +ACGAGA +ACGAGT +ACGAGC +ACGAGG +ACGTAA +ACGTAT +ACGTAC +ACGTAG +ACGTTA +ACGTTT +ACGTTC +ACGTTG +ACGTCA +ACGTCT +ACGTCC +ACGTCG +ACGTGA +ACGTGT +ACGTGC +ACGTGG +ACGCAA +ACGCAT +ACGCAC +ACGCAG +ACGCTA +ACGCTT +ACGCTC +ACGCTG +ACGCCA +ACGCCT +ACGCCC +ACGCCG +ACGCGA +ACGCGT +ACGCGC +ACGCGG +ACGGAA +ACGGAT +ACGGAC +ACGGAG +ACGGTA +ACGGTT +ACGGTC +ACGGTG +ACGGCA +ACGGCT +ACGGCC +ACGGCG +ACGGGA +ACGGGT +ACGGGC +ACGGGG +AGAAAA +AGAAAT +AGAAAC +AGAAAG +AGAATA +AGAATT +AGAATC +AGAATG +AGAACA +AGAACT +AGAACC +AGAACG +AGAAGA +AGAAGT +AGAAGC +AGAAGG +AGATAA +AGATAT +AGATAC +AGATAG +AGATTA +AGATTT +AGATTC +AGATTG +AGATCA +AGATCT +AGATCC +AGATCG +AGATGA +AGATGT +AGATGC +AGATGG +AGACAA +AGACAT +AGACAC +AGACAG +AGACTA +AGACTT +AGACTC +AGACTG +AGACCA +AGACCT +AGACCC +AGACCG +AGACGA +AGACGT +AGACGC +AGACGG +AGAGAA +AGAGAT +AGAGAC +AGAGAG +AGAGTA +AGAGTT +AGAGTC +AGAGTG +AGAGCA +AGAGCT +AGAGCC +AGAGCG +AGAGGA +AGAGGT +AGAGGC +AGAGGG +AGTAAA +AGTAAT +AGTAAC +AGTAAG +AGTATA +AGTATT +AGTATC +AGTATG +AGTACA +AGTACT +AGTACC +AGTACG +AGTAGA +AGTAGT +AGTAGC +AGTAGG +AGTTAA +AGTTAT +AGTTAC +AGTTAG +AGTTTA +AGTTTT +AGTTTC +AGTTTG +AGTTCA +AGTTCT +AGTTCC +AGTTCG +AGTTGA +AGTTGT +AGTTGC +AGTTGG +AGTCAA +AGTCAT +AGTCAC +AGTCAG +AGTCTA +AGTCTT +AGTCTC +AGTCTG +AGTCCA +AGTCCT +AGTCCC +AGTCCG +AGTCGA +AGTCGT +AGTCGC +AGTCGG +AGTGAA +AGTGAT +AGTGAC +AGTGAG +AGTGTA +AGTGTT +AGTGTC +AGTGTG +AGTGCA +AGTGCT +AGTGCC +AGTGCG +AGTGGA +AGTGGT +AGTGGC +AGTGGG +AGCAAA +AGCAAT +AGCAAC +AGCAAG +AGCATA +AGCATT +AGCATC +AGCATG +AGCACA +AGCACT +AGCACC +AGCACG +AGCAGA +AGCAGT +AGCAGC +AGCAGG +AGCTAA +AGCTAT +AGCTAC +AGCTAG +AGCTTA +AGCTTT +AGCTTC +AGCTTG +AGCTCA +AGCTCT +AGCTCC +AGCTCG +AGCTGA +AGCTGT +AGCTGC +AGCTGG +AGCCAA +AGCCAT +AGCCAC +AGCCAG +AGCCTA +AGCCTT +AGCCTC +AGCCTG +AGCCCA +AGCCCT +AGCCCC +AGCCCG +AGCCGA +AGCCGT +AGCCGC +AGCCGG +AGCGAA +AGCGAT +AGCGAC +AGCGAG +AGCGTA +AGCGTT +AGCGTC +AGCGTG +AGCGCA +AGCGCT +AGCGCC +AGCGCG +AGCGGA +AGCGGT +AGCGGC +AGCGGG +AGGAAA +AGGAAT +AGGAAC +AGGAAG +AGGATA +AGGATT +AGGATC +AGGATG +AGGACA +AGGACT +AGGACC +AGGACG +AGGAGA +AGGAGT +AGGAGC +AGGAGG +AGGTAA +AGGTAT +AGGTAC +AGGTAG +AGGTTA +AGGTTT +AGGTTC +AGGTTG +AGGTCA +AGGTCT +AGGTCC +AGGTCG +AGGTGA +AGGTGT +AGGTGC +AGGTGG +AGGCAA +AGGCAT +AGGCAC +AGGCAG +AGGCTA +AGGCTT +AGGCTC +AGGCTG +AGGCCA +AGGCCT +AGGCCC +AGGCCG +AGGCGA +AGGCGT +AGGCGC +AGGCGG +AGGGAA +AGGGAT +AGGGAC +AGGGAG +AGGGTA +AGGGTT +AGGGTC +AGGGTG +AGGGCA +AGGGCT +AGGGCC +AGGGCG +AGGGGA +AGGGGT +AGGGGC +AGGGGG +TAAAAA +TAAAAT +TAAAAC +TAAAAG +TAAATA +TAAATT +TAAATC +TAAATG +TAAACA +TAAACT +TAAACC +TAAACG +TAAAGA +TAAAGT +TAAAGC +TAAAGG +TAATAA +TAATAT +TAATAC +TAATAG +TAATTA +TAATTT +TAATTC +TAATTG +TAATCA +TAATCT +TAATCC +TAATCG +TAATGA +TAATGT +TAATGC +TAATGG +TAACAA +TAACAT +TAACAC +TAACAG +TAACTA +TAACTT +TAACTC +TAACTG +TAACCA +TAACCT +TAACCC +TAACCG +TAACGA +TAACGT +TAACGC +TAACGG +TAAGAA +TAAGAT +TAAGAC +TAAGAG +TAAGTA +TAAGTT +TAAGTC +TAAGTG +TAAGCA +TAAGCT +TAAGCC +TAAGCG +TAAGGA +TAAGGT +TAAGGC +TAAGGG +TATAAA +TATAAT +TATAAC +TATAAG +TATATA +TATATT +TATATC +TATATG +TATACA +TATACT +TATACC +TATACG +TATAGA +TATAGT +TATAGC +TATAGG +TATTAA +TATTAT +TATTAC +TATTAG +TATTTA +TATTTT +TATTTC +TATTTG +TATTCA +TATTCT +TATTCC +TATTCG +TATTGA +TATTGT +TATTGC +TATTGG +TATCAA +TATCAT +TATCAC +TATCAG +TATCTA +TATCTT +TATCTC +TATCTG +TATCCA +TATCCT +TATCCC +TATCCG +TATCGA +TATCGT +TATCGC +TATCGG +TATGAA +TATGAT +TATGAC +TATGAG +TATGTA +TATGTT +TATGTC +TATGTG +TATGCA +TATGCT +TATGCC +TATGCG +TATGGA +TATGGT +TATGGC +TATGGG +TACAAA +TACAAT +TACAAC +TACAAG +TACATA +TACATT +TACATC +TACATG +TACACA +TACACT +TACACC +TACACG +TACAGA +TACAGT +TACAGC +TACAGG +TACTAA +TACTAT +TACTAC +TACTAG +TACTTA +TACTTT +TACTTC +TACTTG +TACTCA +TACTCT +TACTCC +TACTCG +TACTGA +TACTGT +TACTGC +TACTGG +TACCAA +TACCAT +TACCAC +TACCAG +TACCTA +TACCTT +TACCTC +TACCTG +TACCCA +TACCCT +TACCCC +TACCCG +TACCGA +TACCGT +TACCGC +TACCGG +TACGAA +TACGAT +TACGAC +TACGAG +TACGTA +TACGTT +TACGTC +TACGTG +TACGCA +TACGCT +TACGCC +TACGCG +TACGGA +TACGGT +TACGGC +TACGGG +TAGAAA +TAGAAT +TAGAAC +TAGAAG +TAGATA +TAGATT +TAGATC +TAGATG +TAGACA +TAGACT +TAGACC +TAGACG +TAGAGA +TAGAGT +TAGAGC +TAGAGG +TAGTAA +TAGTAT +TAGTAC +TAGTAG +TAGTTA +TAGTTT +TAGTTC +TAGTTG +TAGTCA +TAGTCT +TAGTCC +TAGTCG +TAGTGA +TAGTGT +TAGTGC +TAGTGG +TAGCAA +TAGCAT +TAGCAC +TAGCAG +TAGCTA +TAGCTT +TAGCTC +TAGCTG +TAGCCA +TAGCCT +TAGCCC +TAGCCG +TAGCGA +TAGCGT +TAGCGC +TAGCGG +TAGGAA +TAGGAT +TAGGAC +TAGGAG +TAGGTA +TAGGTT +TAGGTC +TAGGTG +TAGGCA +TAGGCT +TAGGCC +TAGGCG +TAGGGA +TAGGGT +TAGGGC +TAGGGG +TTAAAA +TTAAAT +TTAAAC +TTAAAG +TTAATA +TTAATT +TTAATC +TTAATG +TTAACA +TTAACT +TTAACC +TTAACG +TTAAGA +TTAAGT +TTAAGC +TTAAGG +TTATAA +TTATAT +TTATAC +TTATAG +TTATTA +TTATTT +TTATTC +TTATTG +TTATCA +TTATCT +TTATCC +TTATCG +TTATGA +TTATGT +TTATGC +TTATGG +TTACAA +TTACAT +TTACAC +TTACAG +TTACTA +TTACTT +TTACTC +TTACTG +TTACCA +TTACCT +TTACCC +TTACCG +TTACGA +TTACGT +TTACGC +TTACGG +TTAGAA +TTAGAT +TTAGAC +TTAGAG +TTAGTA +TTAGTT +TTAGTC +TTAGTG +TTAGCA +TTAGCT +TTAGCC +TTAGCG +TTAGGA +TTAGGT +TTAGGC +TTAGGG +TTTAAA +TTTAAT +TTTAAC +TTTAAG +TTTATA +TTTATT +TTTATC +TTTATG +TTTACA +TTTACT +TTTACC +TTTACG +TTTAGA +TTTAGT +TTTAGC +TTTAGG +TTTTAA +TTTTAT +TTTTAC +TTTTAG +TTTTTA +TTTTTT +TTTTTC +TTTTTG +TTTTCA +TTTTCT +TTTTCC +TTTTCG +TTTTGA +TTTTGT +TTTTGC +TTTTGG +TTTCAA +TTTCAT +TTTCAC +TTTCAG +TTTCTA +TTTCTT +TTTCTC +TTTCTG +TTTCCA +TTTCCT +TTTCCC +TTTCCG +TTTCGA +TTTCGT +TTTCGC +TTTCGG +TTTGAA +TTTGAT +TTTGAC +TTTGAG +TTTGTA +TTTGTT +TTTGTC +TTTGTG +TTTGCA +TTTGCT +TTTGCC +TTTGCG +TTTGGA +TTTGGT +TTTGGC +TTTGGG +TTCAAA +TTCAAT +TTCAAC +TTCAAG +TTCATA +TTCATT +TTCATC +TTCATG +TTCACA +TTCACT +TTCACC +TTCACG +TTCAGA +TTCAGT +TTCAGC +TTCAGG +TTCTAA +TTCTAT +TTCTAC +TTCTAG +TTCTTA +TTCTTT +TTCTTC +TTCTTG +TTCTCA +TTCTCT +TTCTCC +TTCTCG +TTCTGA +TTCTGT +TTCTGC +TTCTGG +TTCCAA +TTCCAT +TTCCAC +TTCCAG +TTCCTA +TTCCTT +TTCCTC +TTCCTG +TTCCCA +TTCCCT +TTCCCC +TTCCCG +TTCCGA +TTCCGT +TTCCGC +TTCCGG +TTCGAA +TTCGAT +TTCGAC +TTCGAG +TTCGTA +TTCGTT +TTCGTC +TTCGTG +TTCGCA +TTCGCT +TTCGCC +TTCGCG +TTCGGA +TTCGGT +TTCGGC +TTCGGG +TTGAAA +TTGAAT +TTGAAC +TTGAAG +TTGATA +TTGATT +TTGATC +TTGATG +TTGACA +TTGACT +TTGACC +TTGACG +TTGAGA +TTGAGT +TTGAGC +TTGAGG +TTGTAA +TTGTAT +TTGTAC +TTGTAG +TTGTTA +TTGTTT +TTGTTC +TTGTTG +TTGTCA +TTGTCT +TTGTCC +TTGTCG +TTGTGA +TTGTGT +TTGTGC +TTGTGG +TTGCAA +TTGCAT +TTGCAC +TTGCAG +TTGCTA +TTGCTT +TTGCTC +TTGCTG +TTGCCA +TTGCCT +TTGCCC +TTGCCG +TTGCGA +TTGCGT +TTGCGC +TTGCGG +TTGGAA +TTGGAT +TTGGAC +TTGGAG +TTGGTA +TTGGTT +TTGGTC +TTGGTG +TTGGCA +TTGGCT +TTGGCC +TTGGCG +TTGGGA +TTGGGT +TTGGGC +TTGGGG +TCAAAA +TCAAAT +TCAAAC +TCAAAG +TCAATA +TCAATT +TCAATC +TCAATG +TCAACA +TCAACT +TCAACC +TCAACG +TCAAGA +TCAAGT +TCAAGC +TCAAGG +TCATAA +TCATAT +TCATAC +TCATAG +TCATTA +TCATTT +TCATTC +TCATTG +TCATCA +TCATCT +TCATCC +TCATCG +TCATGA +TCATGT +TCATGC +TCATGG +TCACAA +TCACAT +TCACAC +TCACAG +TCACTA +TCACTT +TCACTC +TCACTG +TCACCA +TCACCT +TCACCC +TCACCG +TCACGA +TCACGT +TCACGC +TCACGG +TCAGAA +TCAGAT +TCAGAC +TCAGAG +TCAGTA +TCAGTT +TCAGTC +TCAGTG +TCAGCA +TCAGCT +TCAGCC +TCAGCG +TCAGGA +TCAGGT +TCAGGC +TCAGGG +TCTAAA +TCTAAT +TCTAAC +TCTAAG +TCTATA +TCTATT +TCTATC +TCTATG +TCTACA +TCTACT +TCTACC +TCTACG +TCTAGA +TCTAGT +TCTAGC +TCTAGG +TCTTAA +TCTTAT +TCTTAC +TCTTAG +TCTTTA +TCTTTT +TCTTTC +TCTTTG +TCTTCA +TCTTCT +TCTTCC +TCTTCG +TCTTGA +TCTTGT +TCTTGC +TCTTGG +TCTCAA +TCTCAT +TCTCAC +TCTCAG +TCTCTA +TCTCTT +TCTCTC +TCTCTG +TCTCCA +TCTCCT +TCTCCC +TCTCCG +TCTCGA +TCTCGT +TCTCGC +TCTCGG +TCTGAA +TCTGAT +TCTGAC +TCTGAG +TCTGTA +TCTGTT +TCTGTC +TCTGTG +TCTGCA +TCTGCT +TCTGCC +TCTGCG +TCTGGA +TCTGGT +TCTGGC +TCTGGG +TCCAAA +TCCAAT +TCCAAC +TCCAAG +TCCATA +TCCATT +TCCATC +TCCATG +TCCACA +TCCACT +TCCACC +TCCACG +TCCAGA +TCCAGT +TCCAGC +TCCAGG +TCCTAA +TCCTAT +TCCTAC +TCCTAG +TCCTTA +TCCTTT +TCCTTC +TCCTTG +TCCTCA +TCCTCT +TCCTCC +TCCTCG +TCCTGA +TCCTGT +TCCTGC +TCCTGG +TCCCAA +TCCCAT +TCCCAC +TCCCAG +TCCCTA +TCCCTT +TCCCTC +TCCCTG +TCCCCA +TCCCCT +TCCCCC +TCCCCG +TCCCGA +TCCCGT +TCCCGC +TCCCGG +TCCGAA +TCCGAT +TCCGAC +TCCGAG +TCCGTA +TCCGTT +TCCGTC +TCCGTG +TCCGCA +TCCGCT +TCCGCC +TCCGCG +TCCGGA +TCCGGT +TCCGGC +TCCGGG +TCGAAA +TCGAAT +TCGAAC +TCGAAG +TCGATA +TCGATT +TCGATC +TCGATG +TCGACA +TCGACT +TCGACC +TCGACG +TCGAGA +TCGAGT +TCGAGC +TCGAGG +TCGTAA +TCGTAT +TCGTAC +TCGTAG +TCGTTA +TCGTTT +TCGTTC +TCGTTG +TCGTCA +TCGTCT +TCGTCC +TCGTCG +TCGTGA +TCGTGT +TCGTGC +TCGTGG +TCGCAA +TCGCAT +TCGCAC +TCGCAG +TCGCTA +TCGCTT +TCGCTC +TCGCTG +TCGCCA +TCGCCT +TCGCCC +TCGCCG +TCGCGA +TCGCGT +TCGCGC +TCGCGG +TCGGAA +TCGGAT +TCGGAC +TCGGAG +TCGGTA +TCGGTT +TCGGTC +TCGGTG +TCGGCA +TCGGCT +TCGGCC +TCGGCG +TCGGGA +TCGGGT +TCGGGC +TCGGGG +TGAAAA +TGAAAT +TGAAAC +TGAAAG +TGAATA +TGAATT +TGAATC +TGAATG +TGAACA +TGAACT +TGAACC +TGAACG +TGAAGA +TGAAGT +TGAAGC +TGAAGG +TGATAA +TGATAT +TGATAC +TGATAG +TGATTA +TGATTT +TGATTC +TGATTG +TGATCA +TGATCT +TGATCC +TGATCG +TGATGA +TGATGT +TGATGC +TGATGG +TGACAA +TGACAT +TGACAC +TGACAG +TGACTA +TGACTT +TGACTC +TGACTG +TGACCA +TGACCT +TGACCC +TGACCG +TGACGA +TGACGT +TGACGC +TGACGG +TGAGAA +TGAGAT +TGAGAC +TGAGAG +TGAGTA +TGAGTT +TGAGTC +TGAGTG +TGAGCA +TGAGCT +TGAGCC +TGAGCG +TGAGGA +TGAGGT +TGAGGC +TGAGGG +TGTAAA +TGTAAT +TGTAAC +TGTAAG +TGTATA +TGTATT +TGTATC +TGTATG +TGTACA +TGTACT +TGTACC +TGTACG +TGTAGA +TGTAGT +TGTAGC +TGTAGG +TGTTAA +TGTTAT +TGTTAC +TGTTAG +TGTTTA +TGTTTT +TGTTTC +TGTTTG +TGTTCA +TGTTCT +TGTTCC +TGTTCG +TGTTGA +TGTTGT +TGTTGC +TGTTGG +TGTCAA +TGTCAT +TGTCAC +TGTCAG +TGTCTA +TGTCTT +TGTCTC +TGTCTG +TGTCCA +TGTCCT +TGTCCC +TGTCCG +TGTCGA +TGTCGT +TGTCGC +TGTCGG +TGTGAA +TGTGAT +TGTGAC +TGTGAG +TGTGTA +TGTGTT +TGTGTC +TGTGTG +TGTGCA +TGTGCT +TGTGCC +TGTGCG +TGTGGA +TGTGGT +TGTGGC +TGTGGG +TGCAAA +TGCAAT +TGCAAC +TGCAAG +TGCATA +TGCATT +TGCATC +TGCATG +TGCACA +TGCACT +TGCACC +TGCACG +TGCAGA +TGCAGT +TGCAGC +TGCAGG +TGCTAA +TGCTAT +TGCTAC +TGCTAG +TGCTTA +TGCTTT +TGCTTC +TGCTTG +TGCTCA +TGCTCT +TGCTCC +TGCTCG +TGCTGA +TGCTGT +TGCTGC +TGCTGG +TGCCAA +TGCCAT +TGCCAC +TGCCAG +TGCCTA +TGCCTT +TGCCTC +TGCCTG +TGCCCA +TGCCCT +TGCCCC +TGCCCG +TGCCGA +TGCCGT +TGCCGC +TGCCGG +TGCGAA +TGCGAT +TGCGAC +TGCGAG +TGCGTA +TGCGTT +TGCGTC +TGCGTG +TGCGCA +TGCGCT +TGCGCC +TGCGCG +TGCGGA +TGCGGT +TGCGGC +TGCGGG +TGGAAA +TGGAAT +TGGAAC +TGGAAG +TGGATA +TGGATT +TGGATC +TGGATG +TGGACA +TGGACT +TGGACC +TGGACG +TGGAGA +TGGAGT +TGGAGC +TGGAGG +TGGTAA +TGGTAT +TGGTAC +TGGTAG +TGGTTA +TGGTTT +TGGTTC +TGGTTG +TGGTCA +TGGTCT +TGGTCC +TGGTCG +TGGTGA +TGGTGT +TGGTGC +TGGTGG +TGGCAA +TGGCAT +TGGCAC +TGGCAG +TGGCTA +TGGCTT +TGGCTC +TGGCTG +TGGCCA +TGGCCT +TGGCCC +TGGCCG +TGGCGA +TGGCGT +TGGCGC +TGGCGG +TGGGAA +TGGGAT +TGGGAC +TGGGAG +TGGGTA +TGGGTT +TGGGTC +TGGGTG +TGGGCA +TGGGCT +TGGGCC +TGGGCG +TGGGGA +TGGGGT +TGGGGC +TGGGGG +CAAAAA +CAAAAT +CAAAAC +CAAAAG +CAAATA +CAAATT +CAAATC +CAAATG +CAAACA +CAAACT +CAAACC +CAAACG +CAAAGA +CAAAGT +CAAAGC +CAAAGG +CAATAA +CAATAT +CAATAC +CAATAG +CAATTA +CAATTT +CAATTC +CAATTG +CAATCA +CAATCT +CAATCC +CAATCG +CAATGA +CAATGT +CAATGC +CAATGG +CAACAA +CAACAT +CAACAC +CAACAG +CAACTA +CAACTT +CAACTC +CAACTG +CAACCA +CAACCT +CAACCC +CAACCG +CAACGA +CAACGT +CAACGC +CAACGG +CAAGAA +CAAGAT +CAAGAC +CAAGAG +CAAGTA +CAAGTT +CAAGTC +CAAGTG +CAAGCA +CAAGCT +CAAGCC +CAAGCG +CAAGGA +CAAGGT +CAAGGC +CAAGGG +CATAAA +CATAAT +CATAAC +CATAAG +CATATA +CATATT +CATATC +CATATG +CATACA +CATACT +CATACC +CATACG +CATAGA +CATAGT +CATAGC +CATAGG +CATTAA +CATTAT +CATTAC +CATTAG +CATTTA +CATTTT +CATTTC +CATTTG +CATTCA +CATTCT +CATTCC +CATTCG +CATTGA +CATTGT +CATTGC +CATTGG +CATCAA +CATCAT +CATCAC +CATCAG +CATCTA +CATCTT +CATCTC +CATCTG +CATCCA +CATCCT +CATCCC +CATCCG +CATCGA +CATCGT +CATCGC +CATCGG +CATGAA +CATGAT +CATGAC +CATGAG +CATGTA +CATGTT +CATGTC +CATGTG +CATGCA +CATGCT +CATGCC +CATGCG +CATGGA +CATGGT +CATGGC +CATGGG +CACAAA +CACAAT +CACAAC +CACAAG +CACATA +CACATT +CACATC +CACATG +CACACA +CACACT +CACACC +CACACG +CACAGA +CACAGT +CACAGC +CACAGG +CACTAA +CACTAT +CACTAC +CACTAG +CACTTA +CACTTT +CACTTC +CACTTG +CACTCA +CACTCT +CACTCC +CACTCG +CACTGA +CACTGT +CACTGC +CACTGG +CACCAA +CACCAT +CACCAC +CACCAG +CACCTA +CACCTT +CACCTC +CACCTG +CACCCA +CACCCT +CACCCC +CACCCG +CACCGA +CACCGT +CACCGC +CACCGG +CACGAA +CACGAT +CACGAC +CACGAG +CACGTA +CACGTT +CACGTC +CACGTG +CACGCA +CACGCT +CACGCC +CACGCG +CACGGA +CACGGT +CACGGC +CACGGG +CAGAAA +CAGAAT +CAGAAC +CAGAAG +CAGATA +CAGATT +CAGATC +CAGATG +CAGACA +CAGACT +CAGACC +CAGACG +CAGAGA +CAGAGT +CAGAGC +CAGAGG +CAGTAA +CAGTAT +CAGTAC +CAGTAG +CAGTTA +CAGTTT +CAGTTC +CAGTTG +CAGTCA +CAGTCT +CAGTCC +CAGTCG +CAGTGA +CAGTGT +CAGTGC +CAGTGG +CAGCAA +CAGCAT +CAGCAC +CAGCAG +CAGCTA +CAGCTT +CAGCTC +CAGCTG +CAGCCA +CAGCCT +CAGCCC +CAGCCG +CAGCGA +CAGCGT +CAGCGC +CAGCGG +CAGGAA +CAGGAT +CAGGAC +CAGGAG +CAGGTA +CAGGTT +CAGGTC +CAGGTG +CAGGCA +CAGGCT +CAGGCC +CAGGCG +CAGGGA +CAGGGT +CAGGGC +CAGGGG +CTAAAA +CTAAAT +CTAAAC +CTAAAG +CTAATA +CTAATT +CTAATC +CTAATG +CTAACA +CTAACT +CTAACC +CTAACG +CTAAGA +CTAAGT +CTAAGC +CTAAGG +CTATAA +CTATAT +CTATAC +CTATAG +CTATTA +CTATTT +CTATTC +CTATTG +CTATCA +CTATCT +CTATCC +CTATCG +CTATGA +CTATGT +CTATGC +CTATGG +CTACAA +CTACAT +CTACAC +CTACAG +CTACTA +CTACTT +CTACTC +CTACTG +CTACCA +CTACCT +CTACCC +CTACCG +CTACGA +CTACGT +CTACGC +CTACGG +CTAGAA +CTAGAT +CTAGAC +CTAGAG +CTAGTA +CTAGTT +CTAGTC +CTAGTG +CTAGCA +CTAGCT +CTAGCC +CTAGCG +CTAGGA +CTAGGT +CTAGGC +CTAGGG +CTTAAA +CTTAAT +CTTAAC +CTTAAG +CTTATA +CTTATT +CTTATC +CTTATG +CTTACA +CTTACT +CTTACC +CTTACG +CTTAGA +CTTAGT +CTTAGC +CTTAGG +CTTTAA +CTTTAT +CTTTAC +CTTTAG +CTTTTA +CTTTTT +CTTTTC +CTTTTG +CTTTCA +CTTTCT +CTTTCC +CTTTCG +CTTTGA +CTTTGT +CTTTGC +CTTTGG +CTTCAA +CTTCAT +CTTCAC +CTTCAG +CTTCTA +CTTCTT +CTTCTC +CTTCTG +CTTCCA +CTTCCT +CTTCCC +CTTCCG +CTTCGA +CTTCGT +CTTCGC +CTTCGG +CTTGAA +CTTGAT +CTTGAC +CTTGAG +CTTGTA +CTTGTT +CTTGTC +CTTGTG +CTTGCA +CTTGCT +CTTGCC +CTTGCG +CTTGGA +CTTGGT +CTTGGC +CTTGGG +CTCAAA +CTCAAT +CTCAAC +CTCAAG +CTCATA +CTCATT +CTCATC +CTCATG +CTCACA +CTCACT +CTCACC +CTCACG +CTCAGA +CTCAGT +CTCAGC +CTCAGG +CTCTAA +CTCTAT +CTCTAC +CTCTAG +CTCTTA +CTCTTT +CTCTTC +CTCTTG +CTCTCA +CTCTCT +CTCTCC +CTCTCG +CTCTGA +CTCTGT +CTCTGC +CTCTGG +CTCCAA +CTCCAT +CTCCAC +CTCCAG +CTCCTA +CTCCTT +CTCCTC +CTCCTG +CTCCCA +CTCCCT +CTCCCC +CTCCCG +CTCCGA +CTCCGT +CTCCGC +CTCCGG +CTCGAA +CTCGAT +CTCGAC +CTCGAG +CTCGTA +CTCGTT +CTCGTC +CTCGTG +CTCGCA +CTCGCT +CTCGCC +CTCGCG +CTCGGA +CTCGGT +CTCGGC +CTCGGG +CTGAAA +CTGAAT +CTGAAC +CTGAAG +CTGATA +CTGATT +CTGATC +CTGATG +CTGACA +CTGACT +CTGACC +CTGACG +CTGAGA +CTGAGT +CTGAGC +CTGAGG +CTGTAA +CTGTAT +CTGTAC +CTGTAG +CTGTTA +CTGTTT +CTGTTC +CTGTTG +CTGTCA +CTGTCT +CTGTCC +CTGTCG +CTGTGA +CTGTGT +CTGTGC +CTGTGG +CTGCAA +CTGCAT +CTGCAC +CTGCAG +CTGCTA +CTGCTT +CTGCTC +CTGCTG +CTGCCA +CTGCCT +CTGCCC +CTGCCG +CTGCGA +CTGCGT +CTGCGC +CTGCGG +CTGGAA +CTGGAT +CTGGAC +CTGGAG +CTGGTA +CTGGTT +CTGGTC +CTGGTG +CTGGCA +CTGGCT +CTGGCC +CTGGCG +CTGGGA +CTGGGT +CTGGGC +CTGGGG +CCAAAA +CCAAAT +CCAAAC +CCAAAG +CCAATA +CCAATT +CCAATC +CCAATG +CCAACA +CCAACT +CCAACC +CCAACG +CCAAGA +CCAAGT +CCAAGC +CCAAGG +CCATAA +CCATAT +CCATAC +CCATAG +CCATTA +CCATTT +CCATTC +CCATTG +CCATCA +CCATCT +CCATCC +CCATCG +CCATGA +CCATGT +CCATGC +CCATGG +CCACAA +CCACAT +CCACAC +CCACAG +CCACTA +CCACTT +CCACTC +CCACTG +CCACCA +CCACCT +CCACCC +CCACCG +CCACGA +CCACGT +CCACGC +CCACGG +CCAGAA +CCAGAT +CCAGAC +CCAGAG +CCAGTA +CCAGTT +CCAGTC +CCAGTG +CCAGCA +CCAGCT +CCAGCC +CCAGCG +CCAGGA +CCAGGT +CCAGGC +CCAGGG +CCTAAA +CCTAAT +CCTAAC +CCTAAG +CCTATA +CCTATT +CCTATC +CCTATG +CCTACA +CCTACT +CCTACC +CCTACG +CCTAGA +CCTAGT +CCTAGC +CCTAGG +CCTTAA +CCTTAT +CCTTAC +CCTTAG +CCTTTA +CCTTTT +CCTTTC +CCTTTG +CCTTCA +CCTTCT +CCTTCC +CCTTCG +CCTTGA +CCTTGT +CCTTGC +CCTTGG +CCTCAA +CCTCAT +CCTCAC +CCTCAG +CCTCTA +CCTCTT +CCTCTC +CCTCTG +CCTCCA +CCTCCT +CCTCCC +CCTCCG +CCTCGA +CCTCGT +CCTCGC +CCTCGG +CCTGAA +CCTGAT +CCTGAC +CCTGAG +CCTGTA +CCTGTT +CCTGTC +CCTGTG +CCTGCA +CCTGCT +CCTGCC +CCTGCG +CCTGGA +CCTGGT +CCTGGC +CCTGGG +CCCAAA +CCCAAT +CCCAAC +CCCAAG +CCCATA +CCCATT +CCCATC +CCCATG +CCCACA +CCCACT +CCCACC +CCCACG +CCCAGA +CCCAGT +CCCAGC +CCCAGG +CCCTAA +CCCTAT +CCCTAC +CCCTAG +CCCTTA +CCCTTT +CCCTTC +CCCTTG +CCCTCA +CCCTCT +CCCTCC +CCCTCG +CCCTGA +CCCTGT +CCCTGC +CCCTGG +CCCCAA +CCCCAT +CCCCAC +CCCCAG +CCCCTA +CCCCTT +CCCCTC +CCCCTG +CCCCCA +CCCCCT +CCCCCC +CCCCCG +CCCCGA +CCCCGT +CCCCGC +CCCCGG +CCCGAA +CCCGAT +CCCGAC +CCCGAG +CCCGTA +CCCGTT +CCCGTC +CCCGTG +CCCGCA +CCCGCT +CCCGCC +CCCGCG +CCCGGA +CCCGGT +CCCGGC +CCCGGG +CCGAAA +CCGAAT +CCGAAC +CCGAAG +CCGATA +CCGATT +CCGATC +CCGATG +CCGACA +CCGACT +CCGACC +CCGACG +CCGAGA +CCGAGT +CCGAGC +CCGAGG +CCGTAA +CCGTAT +CCGTAC +CCGTAG +CCGTTA +CCGTTT +CCGTTC +CCGTTG +CCGTCA +CCGTCT +CCGTCC +CCGTCG +CCGTGA +CCGTGT +CCGTGC +CCGTGG +CCGCAA +CCGCAT +CCGCAC +CCGCAG +CCGCTA +CCGCTT +CCGCTC +CCGCTG +CCGCCA +CCGCCT +CCGCCC +CCGCCG +CCGCGA +CCGCGT +CCGCGC +CCGCGG +CCGGAA +CCGGAT +CCGGAC +CCGGAG +CCGGTA +CCGGTT +CCGGTC +CCGGTG +CCGGCA +CCGGCT +CCGGCC +CCGGCG +CCGGGA +CCGGGT +CCGGGC +CCGGGG +CGAAAA +CGAAAT +CGAAAC +CGAAAG +CGAATA +CGAATT +CGAATC +CGAATG +CGAACA +CGAACT +CGAACC +CGAACG +CGAAGA +CGAAGT +CGAAGC +CGAAGG +CGATAA +CGATAT +CGATAC +CGATAG +CGATTA +CGATTT +CGATTC +CGATTG +CGATCA +CGATCT +CGATCC +CGATCG +CGATGA +CGATGT +CGATGC +CGATGG +CGACAA +CGACAT +CGACAC +CGACAG +CGACTA +CGACTT +CGACTC +CGACTG +CGACCA +CGACCT +CGACCC +CGACCG +CGACGA +CGACGT +CGACGC +CGACGG +CGAGAA +CGAGAT +CGAGAC +CGAGAG +CGAGTA +CGAGTT +CGAGTC +CGAGTG +CGAGCA +CGAGCT +CGAGCC +CGAGCG +CGAGGA +CGAGGT +CGAGGC +CGAGGG +CGTAAA +CGTAAT +CGTAAC +CGTAAG +CGTATA +CGTATT +CGTATC +CGTATG +CGTACA +CGTACT +CGTACC +CGTACG +CGTAGA +CGTAGT +CGTAGC +CGTAGG +CGTTAA +CGTTAT +CGTTAC +CGTTAG +CGTTTA +CGTTTT +CGTTTC +CGTTTG +CGTTCA +CGTTCT +CGTTCC +CGTTCG +CGTTGA +CGTTGT +CGTTGC +CGTTGG +CGTCAA +CGTCAT +CGTCAC +CGTCAG +CGTCTA +CGTCTT +CGTCTC +CGTCTG +CGTCCA +CGTCCT +CGTCCC +CGTCCG +CGTCGA +CGTCGT +CGTCGC +CGTCGG +CGTGAA +CGTGAT +CGTGAC +CGTGAG +CGTGTA +CGTGTT +CGTGTC +CGTGTG +CGTGCA +CGTGCT +CGTGCC +CGTGCG +CGTGGA +CGTGGT +CGTGGC +CGTGGG +CGCAAA +CGCAAT +CGCAAC +CGCAAG +CGCATA +CGCATT +CGCATC +CGCATG +CGCACA +CGCACT +CGCACC +CGCACG +CGCAGA +CGCAGT +CGCAGC +CGCAGG +CGCTAA +CGCTAT +CGCTAC +CGCTAG +CGCTTA +CGCTTT +CGCTTC +CGCTTG +CGCTCA +CGCTCT +CGCTCC +CGCTCG +CGCTGA +CGCTGT +CGCTGC +CGCTGG +CGCCAA +CGCCAT +CGCCAC +CGCCAG +CGCCTA +CGCCTT +CGCCTC +CGCCTG +CGCCCA +CGCCCT +CGCCCC +CGCCCG +CGCCGA +CGCCGT +CGCCGC +CGCCGG +CGCGAA +CGCGAT +CGCGAC +CGCGAG +CGCGTA +CGCGTT +CGCGTC +CGCGTG +CGCGCA +CGCGCT +CGCGCC +CGCGCG +CGCGGA +CGCGGT +CGCGGC +CGCGGG +CGGAAA +CGGAAT +CGGAAC +CGGAAG +CGGATA +CGGATT +CGGATC +CGGATG +CGGACA +CGGACT +CGGACC +CGGACG +CGGAGA +CGGAGT +CGGAGC +CGGAGG +CGGTAA +CGGTAT +CGGTAC +CGGTAG +CGGTTA +CGGTTT +CGGTTC +CGGTTG +CGGTCA +CGGTCT +CGGTCC +CGGTCG +CGGTGA +CGGTGT +CGGTGC +CGGTGG +CGGCAA +CGGCAT +CGGCAC +CGGCAG +CGGCTA +CGGCTT +CGGCTC +CGGCTG +CGGCCA +CGGCCT +CGGCCC +CGGCCG +CGGCGA +CGGCGT +CGGCGC +CGGCGG +CGGGAA +CGGGAT +CGGGAC +CGGGAG +CGGGTA +CGGGTT +CGGGTC +CGGGTG +CGGGCA +CGGGCT +CGGGCC +CGGGCG +CGGGGA +CGGGGT +CGGGGC +CGGGGG +GAAAAA +GAAAAT +GAAAAC +GAAAAG +GAAATA +GAAATT +GAAATC +GAAATG +GAAACA +GAAACT +GAAACC +GAAACG +GAAAGA +GAAAGT +GAAAGC +GAAAGG +GAATAA +GAATAT +GAATAC +GAATAG +GAATTA +GAATTT +GAATTC +GAATTG +GAATCA +GAATCT +GAATCC +GAATCG +GAATGA +GAATGT +GAATGC +GAATGG +GAACAA +GAACAT +GAACAC +GAACAG +GAACTA +GAACTT +GAACTC +GAACTG +GAACCA +GAACCT +GAACCC +GAACCG +GAACGA +GAACGT +GAACGC +GAACGG +GAAGAA +GAAGAT +GAAGAC +GAAGAG +GAAGTA +GAAGTT +GAAGTC +GAAGTG +GAAGCA +GAAGCT +GAAGCC +GAAGCG +GAAGGA +GAAGGT +GAAGGC +GAAGGG +GATAAA +GATAAT +GATAAC +GATAAG +GATATA +GATATT +GATATC +GATATG +GATACA +GATACT +GATACC +GATACG +GATAGA +GATAGT +GATAGC +GATAGG +GATTAA +GATTAT +GATTAC +GATTAG +GATTTA +GATTTT +GATTTC +GATTTG +GATTCA +GATTCT +GATTCC +GATTCG +GATTGA +GATTGT +GATTGC +GATTGG +GATCAA +GATCAT +GATCAC +GATCAG +GATCTA +GATCTT +GATCTC +GATCTG +GATCCA +GATCCT +GATCCC +GATCCG +GATCGA +GATCGT +GATCGC +GATCGG +GATGAA +GATGAT +GATGAC +GATGAG +GATGTA +GATGTT +GATGTC +GATGTG +GATGCA +GATGCT +GATGCC +GATGCG +GATGGA +GATGGT +GATGGC +GATGGG +GACAAA +GACAAT +GACAAC +GACAAG +GACATA +GACATT +GACATC +GACATG +GACACA +GACACT +GACACC +GACACG +GACAGA +GACAGT +GACAGC +GACAGG +GACTAA +GACTAT +GACTAC +GACTAG +GACTTA +GACTTT +GACTTC +GACTTG +GACTCA +GACTCT +GACTCC +GACTCG +GACTGA +GACTGT +GACTGC +GACTGG +GACCAA +GACCAT +GACCAC +GACCAG +GACCTA +GACCTT +GACCTC +GACCTG +GACCCA +GACCCT +GACCCC +GACCCG +GACCGA +GACCGT +GACCGC +GACCGG +GACGAA +GACGAT +GACGAC +GACGAG +GACGTA +GACGTT +GACGTC +GACGTG +GACGCA +GACGCT +GACGCC +GACGCG +GACGGA +GACGGT +GACGGC +GACGGG +GAGAAA +GAGAAT +GAGAAC +GAGAAG +GAGATA +GAGATT +GAGATC +GAGATG +GAGACA +GAGACT +GAGACC +GAGACG +GAGAGA +GAGAGT +GAGAGC +GAGAGG +GAGTAA +GAGTAT +GAGTAC +GAGTAG +GAGTTA +GAGTTT +GAGTTC +GAGTTG +GAGTCA +GAGTCT +GAGTCC +GAGTCG +GAGTGA +GAGTGT +GAGTGC +GAGTGG +GAGCAA +GAGCAT +GAGCAC +GAGCAG +GAGCTA +GAGCTT +GAGCTC +GAGCTG +GAGCCA +GAGCCT +GAGCCC +GAGCCG +GAGCGA +GAGCGT +GAGCGC +GAGCGG +GAGGAA +GAGGAT +GAGGAC +GAGGAG +GAGGTA +GAGGTT +GAGGTC +GAGGTG +GAGGCA +GAGGCT +GAGGCC +GAGGCG +GAGGGA +GAGGGT +GAGGGC +GAGGGG +GTAAAA +GTAAAT +GTAAAC +GTAAAG +GTAATA +GTAATT +GTAATC +GTAATG +GTAACA +GTAACT +GTAACC +GTAACG +GTAAGA +GTAAGT +GTAAGC +GTAAGG +GTATAA +GTATAT +GTATAC +GTATAG +GTATTA +GTATTT +GTATTC +GTATTG +GTATCA +GTATCT +GTATCC +GTATCG +GTATGA +GTATGT +GTATGC +GTATGG +GTACAA +GTACAT +GTACAC +GTACAG +GTACTA +GTACTT +GTACTC +GTACTG +GTACCA +GTACCT +GTACCC +GTACCG +GTACGA +GTACGT +GTACGC +GTACGG +GTAGAA +GTAGAT +GTAGAC +GTAGAG +GTAGTA +GTAGTT +GTAGTC +GTAGTG +GTAGCA +GTAGCT +GTAGCC +GTAGCG +GTAGGA +GTAGGT +GTAGGC +GTAGGG +GTTAAA +GTTAAT +GTTAAC +GTTAAG +GTTATA +GTTATT +GTTATC +GTTATG +GTTACA +GTTACT +GTTACC +GTTACG +GTTAGA +GTTAGT +GTTAGC +GTTAGG +GTTTAA +GTTTAT +GTTTAC +GTTTAG +GTTTTA +GTTTTT +GTTTTC +GTTTTG +GTTTCA +GTTTCT +GTTTCC +GTTTCG +GTTTGA +GTTTGT +GTTTGC +GTTTGG +GTTCAA +GTTCAT +GTTCAC +GTTCAG +GTTCTA +GTTCTT +GTTCTC +GTTCTG +GTTCCA +GTTCCT +GTTCCC +GTTCCG +GTTCGA +GTTCGT +GTTCGC +GTTCGG +GTTGAA +GTTGAT +GTTGAC +GTTGAG +GTTGTA +GTTGTT +GTTGTC +GTTGTG +GTTGCA +GTTGCT +GTTGCC +GTTGCG +GTTGGA +GTTGGT +GTTGGC +GTTGGG +GTCAAA +GTCAAT +GTCAAC +GTCAAG +GTCATA +GTCATT +GTCATC +GTCATG +GTCACA +GTCACT +GTCACC +GTCACG +GTCAGA +GTCAGT +GTCAGC +GTCAGG +GTCTAA +GTCTAT +GTCTAC +GTCTAG +GTCTTA +GTCTTT +GTCTTC +GTCTTG +GTCTCA +GTCTCT +GTCTCC +GTCTCG +GTCTGA +GTCTGT +GTCTGC +GTCTGG +GTCCAA +GTCCAT +GTCCAC +GTCCAG +GTCCTA +GTCCTT +GTCCTC +GTCCTG +GTCCCA +GTCCCT +GTCCCC +GTCCCG +GTCCGA +GTCCGT +GTCCGC +GTCCGG +GTCGAA +GTCGAT +GTCGAC +GTCGAG +GTCGTA +GTCGTT +GTCGTC +GTCGTG +GTCGCA +GTCGCT +GTCGCC +GTCGCG +GTCGGA +GTCGGT +GTCGGC +GTCGGG +GTGAAA +GTGAAT +GTGAAC +GTGAAG +GTGATA +GTGATT +GTGATC +GTGATG +GTGACA +GTGACT +GTGACC +GTGACG +GTGAGA +GTGAGT +GTGAGC +GTGAGG +GTGTAA +GTGTAT +GTGTAC +GTGTAG +GTGTTA +GTGTTT +GTGTTC +GTGTTG +GTGTCA +GTGTCT +GTGTCC +GTGTCG +GTGTGA +GTGTGT +GTGTGC +GTGTGG +GTGCAA +GTGCAT +GTGCAC +GTGCAG +GTGCTA +GTGCTT +GTGCTC +GTGCTG +GTGCCA +GTGCCT +GTGCCC +GTGCCG +GTGCGA +GTGCGT +GTGCGC +GTGCGG +GTGGAA +GTGGAT +GTGGAC +GTGGAG +GTGGTA +GTGGTT +GTGGTC +GTGGTG +GTGGCA +GTGGCT +GTGGCC +GTGGCG +GTGGGA +GTGGGT +GTGGGC +GTGGGG +GCAAAA +GCAAAT +GCAAAC +GCAAAG +GCAATA +GCAATT +GCAATC +GCAATG +GCAACA +GCAACT +GCAACC +GCAACG +GCAAGA +GCAAGT +GCAAGC +GCAAGG +GCATAA +GCATAT +GCATAC +GCATAG +GCATTA +GCATTT +GCATTC +GCATTG +GCATCA +GCATCT +GCATCC +GCATCG +GCATGA +GCATGT +GCATGC +GCATGG +GCACAA +GCACAT +GCACAC +GCACAG +GCACTA +GCACTT +GCACTC +GCACTG +GCACCA +GCACCT +GCACCC +GCACCG +GCACGA +GCACGT +GCACGC +GCACGG +GCAGAA +GCAGAT +GCAGAC +GCAGAG +GCAGTA +GCAGTT +GCAGTC +GCAGTG +GCAGCA +GCAGCT +GCAGCC +GCAGCG +GCAGGA +GCAGGT +GCAGGC +GCAGGG +GCTAAA +GCTAAT +GCTAAC +GCTAAG +GCTATA +GCTATT +GCTATC +GCTATG +GCTACA +GCTACT +GCTACC +GCTACG +GCTAGA +GCTAGT +GCTAGC +GCTAGG +GCTTAA +GCTTAT +GCTTAC +GCTTAG +GCTTTA +GCTTTT +GCTTTC +GCTTTG +GCTTCA +GCTTCT +GCTTCC +GCTTCG +GCTTGA +GCTTGT +GCTTGC +GCTTGG +GCTCAA +GCTCAT +GCTCAC +GCTCAG +GCTCTA +GCTCTT +GCTCTC +GCTCTG +GCTCCA +GCTCCT +GCTCCC +GCTCCG +GCTCGA +GCTCGT +GCTCGC +GCTCGG +GCTGAA +GCTGAT +GCTGAC +GCTGAG +GCTGTA +GCTGTT +GCTGTC +GCTGTG +GCTGCA +GCTGCT +GCTGCC +GCTGCG +GCTGGA +GCTGGT +GCTGGC +GCTGGG +GCCAAA +GCCAAT +GCCAAC +GCCAAG +GCCATA +GCCATT +GCCATC +GCCATG +GCCACA +GCCACT +GCCACC +GCCACG +GCCAGA +GCCAGT +GCCAGC +GCCAGG +GCCTAA +GCCTAT +GCCTAC +GCCTAG +GCCTTA +GCCTTT +GCCTTC +GCCTTG +GCCTCA +GCCTCT +GCCTCC +GCCTCG +GCCTGA +GCCTGT +GCCTGC +GCCTGG +GCCCAA +GCCCAT +GCCCAC +GCCCAG +GCCCTA +GCCCTT +GCCCTC +GCCCTG +GCCCCA +GCCCCT +GCCCCC +GCCCCG +GCCCGA +GCCCGT +GCCCGC +GCCCGG +GCCGAA +GCCGAT +GCCGAC +GCCGAG +GCCGTA +GCCGTT +GCCGTC +GCCGTG +GCCGCA +GCCGCT +GCCGCC +GCCGCG +GCCGGA +GCCGGT +GCCGGC +GCCGGG +GCGAAA +GCGAAT +GCGAAC +GCGAAG +GCGATA +GCGATT +GCGATC +GCGATG +GCGACA +GCGACT +GCGACC +GCGACG +GCGAGA +GCGAGT +GCGAGC +GCGAGG +GCGTAA +GCGTAT +GCGTAC +GCGTAG +GCGTTA +GCGTTT +GCGTTC +GCGTTG +GCGTCA +GCGTCT +GCGTCC +GCGTCG +GCGTGA +GCGTGT +GCGTGC +GCGTGG +GCGCAA +GCGCAT +GCGCAC +GCGCAG +GCGCTA +GCGCTT +GCGCTC +GCGCTG +GCGCCA +GCGCCT +GCGCCC +GCGCCG +GCGCGA +GCGCGT +GCGCGC +GCGCGG +GCGGAA +GCGGAT +GCGGAC +GCGGAG +GCGGTA +GCGGTT +GCGGTC +GCGGTG +GCGGCA +GCGGCT +GCGGCC +GCGGCG +GCGGGA +GCGGGT +GCGGGC +GCGGGG +GGAAAA +GGAAAT +GGAAAC +GGAAAG +GGAATA +GGAATT +GGAATC +GGAATG +GGAACA +GGAACT +GGAACC +GGAACG +GGAAGA +GGAAGT +GGAAGC +GGAAGG +GGATAA +GGATAT +GGATAC +GGATAG +GGATTA +GGATTT +GGATTC +GGATTG +GGATCA +GGATCT +GGATCC +GGATCG +GGATGA +GGATGT +GGATGC +GGATGG +GGACAA +GGACAT +GGACAC +GGACAG +GGACTA +GGACTT +GGACTC +GGACTG +GGACCA +GGACCT +GGACCC +GGACCG +GGACGA +GGACGT +GGACGC +GGACGG +GGAGAA +GGAGAT +GGAGAC +GGAGAG +GGAGTA +GGAGTT +GGAGTC +GGAGTG +GGAGCA +GGAGCT +GGAGCC +GGAGCG +GGAGGA +GGAGGT +GGAGGC +GGAGGG +GGTAAA +GGTAAT +GGTAAC +GGTAAG +GGTATA +GGTATT +GGTATC +GGTATG +GGTACA +GGTACT +GGTACC +GGTACG +GGTAGA +GGTAGT +GGTAGC +GGTAGG +GGTTAA +GGTTAT +GGTTAC +GGTTAG +GGTTTA +GGTTTT +GGTTTC +GGTTTG +GGTTCA +GGTTCT +GGTTCC +GGTTCG +GGTTGA +GGTTGT +GGTTGC +GGTTGG +GGTCAA +GGTCAT +GGTCAC +GGTCAG +GGTCTA +GGTCTT +GGTCTC +GGTCTG +GGTCCA +GGTCCT +GGTCCC +GGTCCG +GGTCGA +GGTCGT +GGTCGC +GGTCGG +GGTGAA +GGTGAT +GGTGAC +GGTGAG +GGTGTA +GGTGTT +GGTGTC +GGTGTG +GGTGCA +GGTGCT +GGTGCC +GGTGCG +GGTGGA +GGTGGT +GGTGGC +GGTGGG +GGCAAA +GGCAAT +GGCAAC +GGCAAG +GGCATA +GGCATT +GGCATC +GGCATG +GGCACA +GGCACT +GGCACC +GGCACG +GGCAGA +GGCAGT +GGCAGC +GGCAGG +GGCTAA +GGCTAT +GGCTAC +GGCTAG +GGCTTA +GGCTTT +GGCTTC +GGCTTG +GGCTCA +GGCTCT +GGCTCC +GGCTCG +GGCTGA +GGCTGT +GGCTGC +GGCTGG +GGCCAA +GGCCAT +GGCCAC +GGCCAG +GGCCTA +GGCCTT +GGCCTC +GGCCTG +GGCCCA +GGCCCT +GGCCCC +GGCCCG +GGCCGA +GGCCGT +GGCCGC +GGCCGG +GGCGAA +GGCGAT +GGCGAC +GGCGAG +GGCGTA +GGCGTT +GGCGTC +GGCGTG +GGCGCA +GGCGCT +GGCGCC +GGCGCG +GGCGGA +GGCGGT +GGCGGC +GGCGGG +GGGAAA +GGGAAT +GGGAAC +GGGAAG +GGGATA +GGGATT +GGGATC +GGGATG +GGGACA +GGGACT +GGGACC +GGGACG +GGGAGA +GGGAGT +GGGAGC +GGGAGG +GGGTAA +GGGTAT +GGGTAC +GGGTAG +GGGTTA +GGGTTT +GGGTTC +GGGTTG +GGGTCA +GGGTCT +GGGTCC +GGGTCG +GGGTGA +GGGTGT +GGGTGC +GGGTGG +GGGCAA +GGGCAT +GGGCAC +GGGCAG +GGGCTA +GGGCTT +GGGCTC +GGGCTG +GGGCCA +GGGCCT +GGGCCC +GGGCCG +GGGCGA +GGGCGT +GGGCGC +GGGCGG +GGGGAA +GGGGAT +GGGGAC +GGGGAG +GGGGTA +GGGGTT +GGGGTC +GGGGTG +GGGGCA +GGGGCT +GGGGCC +GGGGCG +GGGGGA +GGGGGT +GGGGGC +GGGGGG diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-4000/config.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..9b14e61d327bf586fcff6e817cf9d365bae18ad2 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-4000/config.json @@ -0,0 +1,26 @@ +{ + "_name_or_path": "fabihamakhdoomi/TinyDNABERT", + "architectures": [ + "BertForSequenceClassification" + ], + "attention_probs_dropout_prob": 0.1, + "classifier_dropout": null, + "hidden_act": "gelu", + "hidden_dropout_prob": 0.1, + "hidden_size": 128, + "initializer_range": 0.02, + "intermediate_size": 512, + "layer_norm_eps": 1e-12, + "max_position_embeddings": 512, + "model_type": "bert", + "num_attention_heads": 2, + "num_hidden_layers": 2, + "pad_token_id": 0, + "position_embedding_type": "absolute", + "problem_type": "single_label_classification", + "torch_dtype": "float32", + "transformers_version": "4.44.0", + "type_vocab_size": 2, + "use_cache": true, + "vocab_size": 4101 +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-4000/model.safetensors b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-4000/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..e6589f5d3f821ba0c083819346fee1bb5135d075 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-4000/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a398c92f86aca880b963e6288b9526f277da0d7b10e57f6653e1b8f20954be09 +size 4021688 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-4000/optimizer.pt b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-4000/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..175e097c3084370fc834207afaf8ef07e2a173c4 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-4000/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bdad34f1f0ab531135a08d7c242065c1a81d1aef75496b008527cc7129c2871f +size 8058700 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-4000/rng_state.pth b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-4000/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..e341a4a808ae24c21407bcc648bed6b2d6a16c6e --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-4000/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bc0eaa426d9fdf3c0f9bd6754ed553c3bff648a3431f959a858959a437717ed6 +size 14244 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-4000/scheduler.pt b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-4000/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..8c9a6adac55ac4d1afd6e1b61889589acfe31330 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-4000/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f3e454de3c0f52ddb8e5884c9783308658a6f61ede2ed6e377ca9a68d47c4515 +size 1064 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-4000/special_tokens_map.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-4000/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..a8b3208c2884c4efb86e49300fdd3dc877220cdf --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-4000/special_tokens_map.json @@ -0,0 +1,7 @@ +{ + "cls_token": "[CLS]", + "mask_token": "[MASK]", + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-4000/tokenizer.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-4000/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..ebd782de6963b9032ce9374cfe30f5275be1a30b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-4000/tokenizer.json @@ -0,0 +1,4265 @@ +{ + "version": "1.0", + "truncation": { + "direction": "Right", + "max_length": 512, + "strategy": "LongestFirst", + "stride": 0 + }, + "padding": { + "strategy": { + "Fixed": 512 + }, + "direction": "Right", + "pad_to_multiple_of": null, + "pad_id": 0, + "pad_type_id": 0, + "pad_token": "[PAD]" + }, + "added_tokens": [ + { + "id": 0, + "content": "[PAD]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 1, + "content": "[UNK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 2, + "content": "[CLS]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 3, + "content": "[SEP]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 4, + "content": "[MASK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + } + ], + "normalizer": { + "type": "BertNormalizer", + "clean_text": true, + "handle_chinese_chars": true, + "strip_accents": null, + "lowercase": false + }, + "pre_tokenizer": { + "type": "BertPreTokenizer" + }, + "post_processor": { + "type": "TemplateProcessing", + "single": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + } + ], + "pair": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "B", + "type_id": 1 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 1 + } + } + ], + "special_tokens": { + "[CLS]": { + "id": "[CLS]", + "ids": [ + 2 + ], + "tokens": [ + "[CLS]" + ] + }, + "[SEP]": { + "id": "[SEP]", + "ids": [ + 3 + ], + "tokens": [ + "[SEP]" + ] + } + } + }, + "decoder": { + "type": "WordPiece", + "prefix": "##", + "cleanup": true + }, + "model": { + "type": "WordPiece", + "unk_token": "[UNK]", + "continuing_subword_prefix": "##", + "max_input_chars_per_word": 100, + "vocab": { + "[PAD]": 0, + "[UNK]": 1, + "[CLS]": 2, + "[SEP]": 3, + "[MASK]": 4, + "AAAAAA": 5, + "AAAAAT": 6, + "AAAAAC": 7, + "AAAAAG": 8, + "AAAATA": 9, + "AAAATT": 10, + "AAAATC": 11, + "AAAATG": 12, + "AAAACA": 13, + "AAAACT": 14, + "AAAACC": 15, + "AAAACG": 16, + "AAAAGA": 17, + "AAAAGT": 18, + "AAAAGC": 19, + "AAAAGG": 20, + "AAATAA": 21, + "AAATAT": 22, + "AAATAC": 23, + "AAATAG": 24, + "AAATTA": 25, + "AAATTT": 26, + "AAATTC": 27, + "AAATTG": 28, + "AAATCA": 29, + "AAATCT": 30, + "AAATCC": 31, + "AAATCG": 32, + "AAATGA": 33, + "AAATGT": 34, + "AAATGC": 35, + "AAATGG": 36, + "AAACAA": 37, + "AAACAT": 38, + "AAACAC": 39, + "AAACAG": 40, + "AAACTA": 41, + "AAACTT": 42, + "AAACTC": 43, + "AAACTG": 44, + "AAACCA": 45, + "AAACCT": 46, + "AAACCC": 47, + "AAACCG": 48, + "AAACGA": 49, + "AAACGT": 50, + "AAACGC": 51, + "AAACGG": 52, + "AAAGAA": 53, + "AAAGAT": 54, + "AAAGAC": 55, + "AAAGAG": 56, + "AAAGTA": 57, + "AAAGTT": 58, + "AAAGTC": 59, + "AAAGTG": 60, + "AAAGCA": 61, + "AAAGCT": 62, + "AAAGCC": 63, + "AAAGCG": 64, + "AAAGGA": 65, + "AAAGGT": 66, + "AAAGGC": 67, + "AAAGGG": 68, + "AATAAA": 69, + "AATAAT": 70, + "AATAAC": 71, + "AATAAG": 72, + "AATATA": 73, + "AATATT": 74, + "AATATC": 75, + "AATATG": 76, + "AATACA": 77, + "AATACT": 78, + "AATACC": 79, + "AATACG": 80, + "AATAGA": 81, + "AATAGT": 82, + "AATAGC": 83, + "AATAGG": 84, + "AATTAA": 85, + "AATTAT": 86, + "AATTAC": 87, + "AATTAG": 88, + "AATTTA": 89, + "AATTTT": 90, + "AATTTC": 91, + "AATTTG": 92, + "AATTCA": 93, + "AATTCT": 94, + "AATTCC": 95, + "AATTCG": 96, + "AATTGA": 97, + "AATTGT": 98, + "AATTGC": 99, + "AATTGG": 100, + "AATCAA": 101, + "AATCAT": 102, + "AATCAC": 103, + "AATCAG": 104, + "AATCTA": 105, + "AATCTT": 106, + "AATCTC": 107, + "AATCTG": 108, + "AATCCA": 109, + "AATCCT": 110, + "AATCCC": 111, + "AATCCG": 112, + "AATCGA": 113, + "AATCGT": 114, + "AATCGC": 115, + "AATCGG": 116, + "AATGAA": 117, + "AATGAT": 118, + "AATGAC": 119, + "AATGAG": 120, + "AATGTA": 121, + "AATGTT": 122, + "AATGTC": 123, + "AATGTG": 124, + "AATGCA": 125, + "AATGCT": 126, + "AATGCC": 127, + "AATGCG": 128, + "AATGGA": 129, + "AATGGT": 130, + "AATGGC": 131, + "AATGGG": 132, + "AACAAA": 133, + "AACAAT": 134, + "AACAAC": 135, + "AACAAG": 136, + "AACATA": 137, + "AACATT": 138, + "AACATC": 139, + "AACATG": 140, + "AACACA": 141, + "AACACT": 142, + "AACACC": 143, + "AACACG": 144, + "AACAGA": 145, + "AACAGT": 146, + "AACAGC": 147, + "AACAGG": 148, + "AACTAA": 149, + "AACTAT": 150, + "AACTAC": 151, + "AACTAG": 152, + "AACTTA": 153, + "AACTTT": 154, + "AACTTC": 155, + "AACTTG": 156, + "AACTCA": 157, + "AACTCT": 158, + "AACTCC": 159, + "AACTCG": 160, + "AACTGA": 161, + "AACTGT": 162, + "AACTGC": 163, + "AACTGG": 164, + "AACCAA": 165, + "AACCAT": 166, + "AACCAC": 167, + "AACCAG": 168, + "AACCTA": 169, + "AACCTT": 170, + "AACCTC": 171, + "AACCTG": 172, + "AACCCA": 173, + "AACCCT": 174, + "AACCCC": 175, + "AACCCG": 176, + "AACCGA": 177, + "AACCGT": 178, + "AACCGC": 179, + "AACCGG": 180, + "AACGAA": 181, + "AACGAT": 182, + "AACGAC": 183, + "AACGAG": 184, + "AACGTA": 185, + "AACGTT": 186, + "AACGTC": 187, + "AACGTG": 188, + "AACGCA": 189, + "AACGCT": 190, + "AACGCC": 191, + "AACGCG": 192, + "AACGGA": 193, + "AACGGT": 194, + "AACGGC": 195, + "AACGGG": 196, + "AAGAAA": 197, + "AAGAAT": 198, + "AAGAAC": 199, + "AAGAAG": 200, + "AAGATA": 201, + "AAGATT": 202, + "AAGATC": 203, + "AAGATG": 204, + "AAGACA": 205, + "AAGACT": 206, + "AAGACC": 207, + "AAGACG": 208, + "AAGAGA": 209, + "AAGAGT": 210, + "AAGAGC": 211, + "AAGAGG": 212, + "AAGTAA": 213, + "AAGTAT": 214, + "AAGTAC": 215, + "AAGTAG": 216, + "AAGTTA": 217, + "AAGTTT": 218, + "AAGTTC": 219, + "AAGTTG": 220, + "AAGTCA": 221, + "AAGTCT": 222, + "AAGTCC": 223, + "AAGTCG": 224, + "AAGTGA": 225, + "AAGTGT": 226, + "AAGTGC": 227, + "AAGTGG": 228, + "AAGCAA": 229, + "AAGCAT": 230, + "AAGCAC": 231, + "AAGCAG": 232, + "AAGCTA": 233, + "AAGCTT": 234, + "AAGCTC": 235, + "AAGCTG": 236, + "AAGCCA": 237, + "AAGCCT": 238, + "AAGCCC": 239, + "AAGCCG": 240, + "AAGCGA": 241, + "AAGCGT": 242, + "AAGCGC": 243, + "AAGCGG": 244, + "AAGGAA": 245, + "AAGGAT": 246, + "AAGGAC": 247, + "AAGGAG": 248, + "AAGGTA": 249, + "AAGGTT": 250, + "AAGGTC": 251, + "AAGGTG": 252, + "AAGGCA": 253, + "AAGGCT": 254, + "AAGGCC": 255, + "AAGGCG": 256, + "AAGGGA": 257, + "AAGGGT": 258, + "AAGGGC": 259, + "AAGGGG": 260, + "ATAAAA": 261, + "ATAAAT": 262, + "ATAAAC": 263, + "ATAAAG": 264, + "ATAATA": 265, + "ATAATT": 266, + "ATAATC": 267, + "ATAATG": 268, + "ATAACA": 269, + "ATAACT": 270, + "ATAACC": 271, + "ATAACG": 272, + "ATAAGA": 273, + "ATAAGT": 274, + "ATAAGC": 275, + "ATAAGG": 276, + "ATATAA": 277, + "ATATAT": 278, + "ATATAC": 279, + "ATATAG": 280, + "ATATTA": 281, + "ATATTT": 282, + "ATATTC": 283, + "ATATTG": 284, + "ATATCA": 285, + "ATATCT": 286, + "ATATCC": 287, + "ATATCG": 288, + "ATATGA": 289, + "ATATGT": 290, + "ATATGC": 291, + "ATATGG": 292, + "ATACAA": 293, + "ATACAT": 294, + "ATACAC": 295, + "ATACAG": 296, + "ATACTA": 297, + "ATACTT": 298, + "ATACTC": 299, + "ATACTG": 300, + "ATACCA": 301, + "ATACCT": 302, + "ATACCC": 303, + "ATACCG": 304, + "ATACGA": 305, + "ATACGT": 306, + "ATACGC": 307, + "ATACGG": 308, + "ATAGAA": 309, + "ATAGAT": 310, + "ATAGAC": 311, + "ATAGAG": 312, + "ATAGTA": 313, + "ATAGTT": 314, + "ATAGTC": 315, + "ATAGTG": 316, + "ATAGCA": 317, + "ATAGCT": 318, + "ATAGCC": 319, + "ATAGCG": 320, + "ATAGGA": 321, + "ATAGGT": 322, + "ATAGGC": 323, + "ATAGGG": 324, + "ATTAAA": 325, + "ATTAAT": 326, + "ATTAAC": 327, + "ATTAAG": 328, + "ATTATA": 329, + "ATTATT": 330, + "ATTATC": 331, + "ATTATG": 332, + "ATTACA": 333, + "ATTACT": 334, + "ATTACC": 335, + "ATTACG": 336, + "ATTAGA": 337, + "ATTAGT": 338, + "ATTAGC": 339, + "ATTAGG": 340, + "ATTTAA": 341, + "ATTTAT": 342, + "ATTTAC": 343, + "ATTTAG": 344, + "ATTTTA": 345, + "ATTTTT": 346, + "ATTTTC": 347, + "ATTTTG": 348, + "ATTTCA": 349, + "ATTTCT": 350, + "ATTTCC": 351, + "ATTTCG": 352, + "ATTTGA": 353, + "ATTTGT": 354, + "ATTTGC": 355, + "ATTTGG": 356, + "ATTCAA": 357, + "ATTCAT": 358, + "ATTCAC": 359, + "ATTCAG": 360, + "ATTCTA": 361, + "ATTCTT": 362, + "ATTCTC": 363, + "ATTCTG": 364, + "ATTCCA": 365, + "ATTCCT": 366, + "ATTCCC": 367, + "ATTCCG": 368, + "ATTCGA": 369, + "ATTCGT": 370, + "ATTCGC": 371, + "ATTCGG": 372, + "ATTGAA": 373, + "ATTGAT": 374, + "ATTGAC": 375, + "ATTGAG": 376, + "ATTGTA": 377, + "ATTGTT": 378, + "ATTGTC": 379, + "ATTGTG": 380, + "ATTGCA": 381, + "ATTGCT": 382, + "ATTGCC": 383, + "ATTGCG": 384, + "ATTGGA": 385, + "ATTGGT": 386, + "ATTGGC": 387, + "ATTGGG": 388, + "ATCAAA": 389, + "ATCAAT": 390, + "ATCAAC": 391, + "ATCAAG": 392, + "ATCATA": 393, + "ATCATT": 394, + "ATCATC": 395, + "ATCATG": 396, + "ATCACA": 397, + "ATCACT": 398, + "ATCACC": 399, + "ATCACG": 400, + "ATCAGA": 401, + "ATCAGT": 402, + "ATCAGC": 403, + "ATCAGG": 404, + "ATCTAA": 405, + "ATCTAT": 406, + "ATCTAC": 407, + "ATCTAG": 408, + "ATCTTA": 409, + "ATCTTT": 410, + "ATCTTC": 411, + "ATCTTG": 412, + "ATCTCA": 413, + "ATCTCT": 414, + "ATCTCC": 415, + "ATCTCG": 416, + "ATCTGA": 417, + "ATCTGT": 418, + "ATCTGC": 419, + "ATCTGG": 420, + "ATCCAA": 421, + "ATCCAT": 422, + "ATCCAC": 423, + "ATCCAG": 424, + "ATCCTA": 425, + "ATCCTT": 426, + "ATCCTC": 427, + "ATCCTG": 428, + "ATCCCA": 429, + "ATCCCT": 430, + "ATCCCC": 431, + "ATCCCG": 432, + "ATCCGA": 433, + "ATCCGT": 434, + "ATCCGC": 435, + "ATCCGG": 436, + "ATCGAA": 437, + "ATCGAT": 438, + "ATCGAC": 439, + "ATCGAG": 440, + "ATCGTA": 441, + "ATCGTT": 442, + "ATCGTC": 443, + "ATCGTG": 444, + "ATCGCA": 445, + "ATCGCT": 446, + "ATCGCC": 447, + "ATCGCG": 448, + "ATCGGA": 449, + "ATCGGT": 450, + "ATCGGC": 451, + "ATCGGG": 452, + "ATGAAA": 453, + "ATGAAT": 454, + "ATGAAC": 455, + "ATGAAG": 456, + "ATGATA": 457, + "ATGATT": 458, + "ATGATC": 459, + "ATGATG": 460, + "ATGACA": 461, + "ATGACT": 462, + "ATGACC": 463, + "ATGACG": 464, + "ATGAGA": 465, + "ATGAGT": 466, + "ATGAGC": 467, + "ATGAGG": 468, + "ATGTAA": 469, + "ATGTAT": 470, + "ATGTAC": 471, + "ATGTAG": 472, + "ATGTTA": 473, + "ATGTTT": 474, + "ATGTTC": 475, + "ATGTTG": 476, + "ATGTCA": 477, + "ATGTCT": 478, + "ATGTCC": 479, + "ATGTCG": 480, + "ATGTGA": 481, + "ATGTGT": 482, + "ATGTGC": 483, + "ATGTGG": 484, + "ATGCAA": 485, + "ATGCAT": 486, + "ATGCAC": 487, + "ATGCAG": 488, + "ATGCTA": 489, + "ATGCTT": 490, + "ATGCTC": 491, + "ATGCTG": 492, + "ATGCCA": 493, + "ATGCCT": 494, + "ATGCCC": 495, + "ATGCCG": 496, + "ATGCGA": 497, + "ATGCGT": 498, + "ATGCGC": 499, + "ATGCGG": 500, + "ATGGAA": 501, + "ATGGAT": 502, + "ATGGAC": 503, + "ATGGAG": 504, + "ATGGTA": 505, + "ATGGTT": 506, + "ATGGTC": 507, + "ATGGTG": 508, + "ATGGCA": 509, + "ATGGCT": 510, + "ATGGCC": 511, + "ATGGCG": 512, + "ATGGGA": 513, + "ATGGGT": 514, + "ATGGGC": 515, + "ATGGGG": 516, + "ACAAAA": 517, + "ACAAAT": 518, + "ACAAAC": 519, + "ACAAAG": 520, + "ACAATA": 521, + "ACAATT": 522, + "ACAATC": 523, + "ACAATG": 524, + "ACAACA": 525, + "ACAACT": 526, + "ACAACC": 527, + "ACAACG": 528, + "ACAAGA": 529, + "ACAAGT": 530, + "ACAAGC": 531, + "ACAAGG": 532, + "ACATAA": 533, + "ACATAT": 534, + "ACATAC": 535, + "ACATAG": 536, + "ACATTA": 537, + "ACATTT": 538, + "ACATTC": 539, + "ACATTG": 540, + "ACATCA": 541, + "ACATCT": 542, + "ACATCC": 543, + "ACATCG": 544, + "ACATGA": 545, + "ACATGT": 546, + "ACATGC": 547, + "ACATGG": 548, + "ACACAA": 549, + "ACACAT": 550, + "ACACAC": 551, + "ACACAG": 552, + "ACACTA": 553, + "ACACTT": 554, + "ACACTC": 555, + "ACACTG": 556, + "ACACCA": 557, + "ACACCT": 558, + "ACACCC": 559, + "ACACCG": 560, + "ACACGA": 561, + "ACACGT": 562, + "ACACGC": 563, + "ACACGG": 564, + "ACAGAA": 565, + "ACAGAT": 566, + "ACAGAC": 567, + "ACAGAG": 568, + "ACAGTA": 569, + "ACAGTT": 570, + "ACAGTC": 571, + "ACAGTG": 572, + "ACAGCA": 573, + "ACAGCT": 574, + "ACAGCC": 575, + "ACAGCG": 576, + "ACAGGA": 577, + "ACAGGT": 578, + "ACAGGC": 579, + "ACAGGG": 580, + "ACTAAA": 581, + "ACTAAT": 582, + "ACTAAC": 583, + "ACTAAG": 584, + "ACTATA": 585, + "ACTATT": 586, + "ACTATC": 587, + "ACTATG": 588, + "ACTACA": 589, + "ACTACT": 590, + "ACTACC": 591, + "ACTACG": 592, + "ACTAGA": 593, + "ACTAGT": 594, + "ACTAGC": 595, + "ACTAGG": 596, + "ACTTAA": 597, + "ACTTAT": 598, + "ACTTAC": 599, + "ACTTAG": 600, + "ACTTTA": 601, + "ACTTTT": 602, + "ACTTTC": 603, + "ACTTTG": 604, + "ACTTCA": 605, + "ACTTCT": 606, + "ACTTCC": 607, + "ACTTCG": 608, + "ACTTGA": 609, + "ACTTGT": 610, + "ACTTGC": 611, + "ACTTGG": 612, + "ACTCAA": 613, + "ACTCAT": 614, + "ACTCAC": 615, + "ACTCAG": 616, + "ACTCTA": 617, + "ACTCTT": 618, + "ACTCTC": 619, + "ACTCTG": 620, + "ACTCCA": 621, + "ACTCCT": 622, + "ACTCCC": 623, + "ACTCCG": 624, + "ACTCGA": 625, + "ACTCGT": 626, + "ACTCGC": 627, + "ACTCGG": 628, + "ACTGAA": 629, + "ACTGAT": 630, + "ACTGAC": 631, + "ACTGAG": 632, + "ACTGTA": 633, + "ACTGTT": 634, + "ACTGTC": 635, + "ACTGTG": 636, + "ACTGCA": 637, + "ACTGCT": 638, + "ACTGCC": 639, + "ACTGCG": 640, + "ACTGGA": 641, + "ACTGGT": 642, + "ACTGGC": 643, + "ACTGGG": 644, + "ACCAAA": 645, + "ACCAAT": 646, + "ACCAAC": 647, + "ACCAAG": 648, + "ACCATA": 649, + "ACCATT": 650, + "ACCATC": 651, + "ACCATG": 652, + "ACCACA": 653, + "ACCACT": 654, + "ACCACC": 655, + "ACCACG": 656, + "ACCAGA": 657, + "ACCAGT": 658, + "ACCAGC": 659, + "ACCAGG": 660, + "ACCTAA": 661, + "ACCTAT": 662, + "ACCTAC": 663, + "ACCTAG": 664, + "ACCTTA": 665, + "ACCTTT": 666, + "ACCTTC": 667, + "ACCTTG": 668, + "ACCTCA": 669, + "ACCTCT": 670, + "ACCTCC": 671, + "ACCTCG": 672, + "ACCTGA": 673, + "ACCTGT": 674, + "ACCTGC": 675, + "ACCTGG": 676, + "ACCCAA": 677, + "ACCCAT": 678, + "ACCCAC": 679, + "ACCCAG": 680, + "ACCCTA": 681, + "ACCCTT": 682, + "ACCCTC": 683, + "ACCCTG": 684, + "ACCCCA": 685, + "ACCCCT": 686, + "ACCCCC": 687, + "ACCCCG": 688, + "ACCCGA": 689, + "ACCCGT": 690, + "ACCCGC": 691, + "ACCCGG": 692, + "ACCGAA": 693, + "ACCGAT": 694, + "ACCGAC": 695, + "ACCGAG": 696, + "ACCGTA": 697, + "ACCGTT": 698, + "ACCGTC": 699, + "ACCGTG": 700, + "ACCGCA": 701, + "ACCGCT": 702, + "ACCGCC": 703, + "ACCGCG": 704, + "ACCGGA": 705, + "ACCGGT": 706, + "ACCGGC": 707, + "ACCGGG": 708, + "ACGAAA": 709, + "ACGAAT": 710, + "ACGAAC": 711, + "ACGAAG": 712, + "ACGATA": 713, + "ACGATT": 714, + "ACGATC": 715, + "ACGATG": 716, + "ACGACA": 717, + "ACGACT": 718, + "ACGACC": 719, + "ACGACG": 720, + "ACGAGA": 721, + "ACGAGT": 722, + "ACGAGC": 723, + "ACGAGG": 724, + "ACGTAA": 725, + "ACGTAT": 726, + "ACGTAC": 727, + "ACGTAG": 728, + "ACGTTA": 729, + "ACGTTT": 730, + "ACGTTC": 731, + "ACGTTG": 732, + "ACGTCA": 733, + "ACGTCT": 734, + "ACGTCC": 735, + "ACGTCG": 736, + "ACGTGA": 737, + "ACGTGT": 738, + "ACGTGC": 739, + "ACGTGG": 740, + "ACGCAA": 741, + "ACGCAT": 742, + "ACGCAC": 743, + "ACGCAG": 744, + "ACGCTA": 745, + "ACGCTT": 746, + "ACGCTC": 747, + "ACGCTG": 748, + "ACGCCA": 749, + "ACGCCT": 750, + "ACGCCC": 751, + "ACGCCG": 752, + "ACGCGA": 753, + "ACGCGT": 754, + "ACGCGC": 755, + "ACGCGG": 756, + "ACGGAA": 757, + "ACGGAT": 758, + "ACGGAC": 759, + "ACGGAG": 760, + "ACGGTA": 761, + "ACGGTT": 762, + "ACGGTC": 763, + "ACGGTG": 764, + "ACGGCA": 765, + "ACGGCT": 766, + "ACGGCC": 767, + "ACGGCG": 768, + "ACGGGA": 769, + "ACGGGT": 770, + "ACGGGC": 771, + "ACGGGG": 772, + "AGAAAA": 773, + "AGAAAT": 774, + "AGAAAC": 775, + "AGAAAG": 776, + "AGAATA": 777, + "AGAATT": 778, + "AGAATC": 779, + "AGAATG": 780, + "AGAACA": 781, + "AGAACT": 782, + "AGAACC": 783, + "AGAACG": 784, + "AGAAGA": 785, + "AGAAGT": 786, + "AGAAGC": 787, + "AGAAGG": 788, + "AGATAA": 789, + "AGATAT": 790, + "AGATAC": 791, + "AGATAG": 792, + "AGATTA": 793, + "AGATTT": 794, + "AGATTC": 795, + "AGATTG": 796, + "AGATCA": 797, + "AGATCT": 798, + "AGATCC": 799, + "AGATCG": 800, + "AGATGA": 801, + "AGATGT": 802, + "AGATGC": 803, + "AGATGG": 804, + "AGACAA": 805, + "AGACAT": 806, + "AGACAC": 807, + "AGACAG": 808, + "AGACTA": 809, + "AGACTT": 810, + "AGACTC": 811, + "AGACTG": 812, + "AGACCA": 813, + "AGACCT": 814, + "AGACCC": 815, + "AGACCG": 816, + "AGACGA": 817, + "AGACGT": 818, + "AGACGC": 819, + "AGACGG": 820, + "AGAGAA": 821, + "AGAGAT": 822, + "AGAGAC": 823, + "AGAGAG": 824, + "AGAGTA": 825, + "AGAGTT": 826, + "AGAGTC": 827, + "AGAGTG": 828, + "AGAGCA": 829, + "AGAGCT": 830, + "AGAGCC": 831, + "AGAGCG": 832, + "AGAGGA": 833, + "AGAGGT": 834, + "AGAGGC": 835, + "AGAGGG": 836, + "AGTAAA": 837, + "AGTAAT": 838, + "AGTAAC": 839, + "AGTAAG": 840, + "AGTATA": 841, + "AGTATT": 842, + "AGTATC": 843, + "AGTATG": 844, + "AGTACA": 845, + "AGTACT": 846, + "AGTACC": 847, + "AGTACG": 848, + "AGTAGA": 849, + "AGTAGT": 850, + "AGTAGC": 851, + "AGTAGG": 852, + "AGTTAA": 853, + "AGTTAT": 854, + "AGTTAC": 855, + "AGTTAG": 856, + "AGTTTA": 857, + "AGTTTT": 858, + "AGTTTC": 859, + "AGTTTG": 860, + "AGTTCA": 861, + "AGTTCT": 862, + "AGTTCC": 863, + "AGTTCG": 864, + "AGTTGA": 865, + "AGTTGT": 866, + "AGTTGC": 867, + "AGTTGG": 868, + "AGTCAA": 869, + "AGTCAT": 870, + "AGTCAC": 871, + "AGTCAG": 872, + "AGTCTA": 873, + "AGTCTT": 874, + "AGTCTC": 875, + "AGTCTG": 876, + "AGTCCA": 877, + "AGTCCT": 878, + "AGTCCC": 879, + "AGTCCG": 880, + "AGTCGA": 881, + "AGTCGT": 882, + "AGTCGC": 883, + "AGTCGG": 884, + "AGTGAA": 885, + "AGTGAT": 886, + "AGTGAC": 887, + "AGTGAG": 888, + "AGTGTA": 889, + "AGTGTT": 890, + "AGTGTC": 891, + "AGTGTG": 892, + "AGTGCA": 893, + "AGTGCT": 894, + "AGTGCC": 895, + "AGTGCG": 896, + "AGTGGA": 897, + "AGTGGT": 898, + "AGTGGC": 899, + "AGTGGG": 900, + "AGCAAA": 901, + "AGCAAT": 902, + "AGCAAC": 903, + "AGCAAG": 904, + "AGCATA": 905, + "AGCATT": 906, + "AGCATC": 907, + "AGCATG": 908, + "AGCACA": 909, + "AGCACT": 910, + "AGCACC": 911, + "AGCACG": 912, + "AGCAGA": 913, + "AGCAGT": 914, + "AGCAGC": 915, + "AGCAGG": 916, + "AGCTAA": 917, + "AGCTAT": 918, + "AGCTAC": 919, + "AGCTAG": 920, + "AGCTTA": 921, + "AGCTTT": 922, + "AGCTTC": 923, + "AGCTTG": 924, + "AGCTCA": 925, + "AGCTCT": 926, + "AGCTCC": 927, + "AGCTCG": 928, + "AGCTGA": 929, + "AGCTGT": 930, + "AGCTGC": 931, + "AGCTGG": 932, + "AGCCAA": 933, + "AGCCAT": 934, + "AGCCAC": 935, + "AGCCAG": 936, + "AGCCTA": 937, + "AGCCTT": 938, + "AGCCTC": 939, + "AGCCTG": 940, + "AGCCCA": 941, + "AGCCCT": 942, + "AGCCCC": 943, + "AGCCCG": 944, + "AGCCGA": 945, + "AGCCGT": 946, + "AGCCGC": 947, + "AGCCGG": 948, + "AGCGAA": 949, + "AGCGAT": 950, + "AGCGAC": 951, + "AGCGAG": 952, + "AGCGTA": 953, + "AGCGTT": 954, + "AGCGTC": 955, + "AGCGTG": 956, + "AGCGCA": 957, + "AGCGCT": 958, + "AGCGCC": 959, + "AGCGCG": 960, + "AGCGGA": 961, + "AGCGGT": 962, + "AGCGGC": 963, + "AGCGGG": 964, + "AGGAAA": 965, + "AGGAAT": 966, + "AGGAAC": 967, + "AGGAAG": 968, + "AGGATA": 969, + "AGGATT": 970, + "AGGATC": 971, + "AGGATG": 972, + "AGGACA": 973, + "AGGACT": 974, + "AGGACC": 975, + "AGGACG": 976, + "AGGAGA": 977, + "AGGAGT": 978, + "AGGAGC": 979, + "AGGAGG": 980, + "AGGTAA": 981, + "AGGTAT": 982, + "AGGTAC": 983, + "AGGTAG": 984, + "AGGTTA": 985, + "AGGTTT": 986, + "AGGTTC": 987, + "AGGTTG": 988, + "AGGTCA": 989, + "AGGTCT": 990, + "AGGTCC": 991, + "AGGTCG": 992, + "AGGTGA": 993, + "AGGTGT": 994, + "AGGTGC": 995, + "AGGTGG": 996, + "AGGCAA": 997, + "AGGCAT": 998, + "AGGCAC": 999, + "AGGCAG": 1000, + "AGGCTA": 1001, + "AGGCTT": 1002, + "AGGCTC": 1003, + "AGGCTG": 1004, + "AGGCCA": 1005, + "AGGCCT": 1006, + "AGGCCC": 1007, + "AGGCCG": 1008, + "AGGCGA": 1009, + "AGGCGT": 1010, + "AGGCGC": 1011, + "AGGCGG": 1012, + "AGGGAA": 1013, + "AGGGAT": 1014, + "AGGGAC": 1015, + "AGGGAG": 1016, + "AGGGTA": 1017, + "AGGGTT": 1018, + "AGGGTC": 1019, + "AGGGTG": 1020, + "AGGGCA": 1021, + "AGGGCT": 1022, + "AGGGCC": 1023, + "AGGGCG": 1024, + "AGGGGA": 1025, + "AGGGGT": 1026, + "AGGGGC": 1027, + "AGGGGG": 1028, + "TAAAAA": 1029, + "TAAAAT": 1030, + "TAAAAC": 1031, + "TAAAAG": 1032, + "TAAATA": 1033, + "TAAATT": 1034, + "TAAATC": 1035, + "TAAATG": 1036, + "TAAACA": 1037, + "TAAACT": 1038, + "TAAACC": 1039, + "TAAACG": 1040, + "TAAAGA": 1041, + "TAAAGT": 1042, + "TAAAGC": 1043, + "TAAAGG": 1044, + "TAATAA": 1045, + "TAATAT": 1046, + "TAATAC": 1047, + "TAATAG": 1048, + "TAATTA": 1049, + "TAATTT": 1050, + "TAATTC": 1051, + "TAATTG": 1052, + "TAATCA": 1053, + "TAATCT": 1054, + "TAATCC": 1055, + "TAATCG": 1056, + "TAATGA": 1057, + "TAATGT": 1058, + "TAATGC": 1059, + "TAATGG": 1060, + "TAACAA": 1061, + "TAACAT": 1062, + "TAACAC": 1063, + "TAACAG": 1064, + "TAACTA": 1065, + "TAACTT": 1066, + "TAACTC": 1067, + "TAACTG": 1068, + "TAACCA": 1069, + "TAACCT": 1070, + "TAACCC": 1071, + "TAACCG": 1072, + "TAACGA": 1073, + "TAACGT": 1074, + "TAACGC": 1075, + "TAACGG": 1076, + "TAAGAA": 1077, + "TAAGAT": 1078, + "TAAGAC": 1079, + "TAAGAG": 1080, + "TAAGTA": 1081, + "TAAGTT": 1082, + "TAAGTC": 1083, + "TAAGTG": 1084, + "TAAGCA": 1085, + "TAAGCT": 1086, + "TAAGCC": 1087, + "TAAGCG": 1088, + "TAAGGA": 1089, + "TAAGGT": 1090, + "TAAGGC": 1091, + "TAAGGG": 1092, + "TATAAA": 1093, + "TATAAT": 1094, + "TATAAC": 1095, + "TATAAG": 1096, + "TATATA": 1097, + "TATATT": 1098, + "TATATC": 1099, + "TATATG": 1100, + "TATACA": 1101, + "TATACT": 1102, + "TATACC": 1103, + "TATACG": 1104, + "TATAGA": 1105, + "TATAGT": 1106, + "TATAGC": 1107, + "TATAGG": 1108, + "TATTAA": 1109, + "TATTAT": 1110, + "TATTAC": 1111, + "TATTAG": 1112, + "TATTTA": 1113, + "TATTTT": 1114, + "TATTTC": 1115, + "TATTTG": 1116, + "TATTCA": 1117, + "TATTCT": 1118, + "TATTCC": 1119, + "TATTCG": 1120, + "TATTGA": 1121, + "TATTGT": 1122, + "TATTGC": 1123, + "TATTGG": 1124, + "TATCAA": 1125, + "TATCAT": 1126, + "TATCAC": 1127, + "TATCAG": 1128, + "TATCTA": 1129, + "TATCTT": 1130, + "TATCTC": 1131, + "TATCTG": 1132, + "TATCCA": 1133, + "TATCCT": 1134, + "TATCCC": 1135, + "TATCCG": 1136, + "TATCGA": 1137, + "TATCGT": 1138, + "TATCGC": 1139, + "TATCGG": 1140, + "TATGAA": 1141, + "TATGAT": 1142, + "TATGAC": 1143, + "TATGAG": 1144, + "TATGTA": 1145, + "TATGTT": 1146, + "TATGTC": 1147, + "TATGTG": 1148, + "TATGCA": 1149, + "TATGCT": 1150, + "TATGCC": 1151, + "TATGCG": 1152, + "TATGGA": 1153, + "TATGGT": 1154, + "TATGGC": 1155, + "TATGGG": 1156, + "TACAAA": 1157, + "TACAAT": 1158, + "TACAAC": 1159, + "TACAAG": 1160, + "TACATA": 1161, + "TACATT": 1162, + "TACATC": 1163, + "TACATG": 1164, + "TACACA": 1165, + "TACACT": 1166, + "TACACC": 1167, + "TACACG": 1168, + "TACAGA": 1169, + "TACAGT": 1170, + "TACAGC": 1171, + "TACAGG": 1172, + "TACTAA": 1173, + "TACTAT": 1174, + "TACTAC": 1175, + "TACTAG": 1176, + "TACTTA": 1177, + "TACTTT": 1178, + "TACTTC": 1179, + "TACTTG": 1180, + "TACTCA": 1181, + "TACTCT": 1182, + "TACTCC": 1183, + "TACTCG": 1184, + "TACTGA": 1185, + "TACTGT": 1186, + "TACTGC": 1187, + "TACTGG": 1188, + "TACCAA": 1189, + "TACCAT": 1190, + "TACCAC": 1191, + "TACCAG": 1192, + "TACCTA": 1193, + "TACCTT": 1194, + "TACCTC": 1195, + "TACCTG": 1196, + "TACCCA": 1197, + "TACCCT": 1198, + "TACCCC": 1199, + "TACCCG": 1200, + "TACCGA": 1201, + "TACCGT": 1202, + "TACCGC": 1203, + "TACCGG": 1204, + "TACGAA": 1205, + "TACGAT": 1206, + "TACGAC": 1207, + "TACGAG": 1208, + "TACGTA": 1209, + "TACGTT": 1210, + "TACGTC": 1211, + "TACGTG": 1212, + "TACGCA": 1213, + "TACGCT": 1214, + "TACGCC": 1215, + "TACGCG": 1216, + "TACGGA": 1217, + "TACGGT": 1218, + "TACGGC": 1219, + "TACGGG": 1220, + "TAGAAA": 1221, + "TAGAAT": 1222, + "TAGAAC": 1223, + "TAGAAG": 1224, + "TAGATA": 1225, + "TAGATT": 1226, + "TAGATC": 1227, + "TAGATG": 1228, + "TAGACA": 1229, + "TAGACT": 1230, + "TAGACC": 1231, + "TAGACG": 1232, + "TAGAGA": 1233, + "TAGAGT": 1234, + "TAGAGC": 1235, + "TAGAGG": 1236, + "TAGTAA": 1237, + "TAGTAT": 1238, + "TAGTAC": 1239, + "TAGTAG": 1240, + "TAGTTA": 1241, + "TAGTTT": 1242, + "TAGTTC": 1243, + "TAGTTG": 1244, + "TAGTCA": 1245, + "TAGTCT": 1246, + "TAGTCC": 1247, + "TAGTCG": 1248, + "TAGTGA": 1249, + "TAGTGT": 1250, + "TAGTGC": 1251, + "TAGTGG": 1252, + "TAGCAA": 1253, + "TAGCAT": 1254, + "TAGCAC": 1255, + "TAGCAG": 1256, + "TAGCTA": 1257, + "TAGCTT": 1258, + "TAGCTC": 1259, + "TAGCTG": 1260, + "TAGCCA": 1261, + "TAGCCT": 1262, + "TAGCCC": 1263, + "TAGCCG": 1264, + "TAGCGA": 1265, + "TAGCGT": 1266, + "TAGCGC": 1267, + "TAGCGG": 1268, + "TAGGAA": 1269, + "TAGGAT": 1270, + "TAGGAC": 1271, + "TAGGAG": 1272, + "TAGGTA": 1273, + "TAGGTT": 1274, + "TAGGTC": 1275, + "TAGGTG": 1276, + "TAGGCA": 1277, + "TAGGCT": 1278, + "TAGGCC": 1279, + "TAGGCG": 1280, + "TAGGGA": 1281, + "TAGGGT": 1282, + "TAGGGC": 1283, + "TAGGGG": 1284, + "TTAAAA": 1285, + "TTAAAT": 1286, + "TTAAAC": 1287, + "TTAAAG": 1288, + "TTAATA": 1289, + "TTAATT": 1290, + "TTAATC": 1291, + "TTAATG": 1292, + "TTAACA": 1293, + "TTAACT": 1294, + "TTAACC": 1295, + "TTAACG": 1296, + "TTAAGA": 1297, + "TTAAGT": 1298, + "TTAAGC": 1299, + "TTAAGG": 1300, + "TTATAA": 1301, + "TTATAT": 1302, + "TTATAC": 1303, + "TTATAG": 1304, + "TTATTA": 1305, + "TTATTT": 1306, + "TTATTC": 1307, + "TTATTG": 1308, + "TTATCA": 1309, + "TTATCT": 1310, + "TTATCC": 1311, + "TTATCG": 1312, + "TTATGA": 1313, + "TTATGT": 1314, + "TTATGC": 1315, + "TTATGG": 1316, + "TTACAA": 1317, + "TTACAT": 1318, + "TTACAC": 1319, + "TTACAG": 1320, + "TTACTA": 1321, + "TTACTT": 1322, + "TTACTC": 1323, + "TTACTG": 1324, + "TTACCA": 1325, + "TTACCT": 1326, + "TTACCC": 1327, + "TTACCG": 1328, + "TTACGA": 1329, + "TTACGT": 1330, + "TTACGC": 1331, + "TTACGG": 1332, + "TTAGAA": 1333, + "TTAGAT": 1334, + "TTAGAC": 1335, + "TTAGAG": 1336, + "TTAGTA": 1337, + "TTAGTT": 1338, + "TTAGTC": 1339, + "TTAGTG": 1340, + "TTAGCA": 1341, + "TTAGCT": 1342, + "TTAGCC": 1343, + "TTAGCG": 1344, + "TTAGGA": 1345, + "TTAGGT": 1346, + "TTAGGC": 1347, + "TTAGGG": 1348, + "TTTAAA": 1349, + "TTTAAT": 1350, + "TTTAAC": 1351, + "TTTAAG": 1352, + "TTTATA": 1353, + "TTTATT": 1354, + "TTTATC": 1355, + "TTTATG": 1356, + "TTTACA": 1357, + "TTTACT": 1358, + "TTTACC": 1359, + "TTTACG": 1360, + "TTTAGA": 1361, + "TTTAGT": 1362, + "TTTAGC": 1363, + "TTTAGG": 1364, + "TTTTAA": 1365, + "TTTTAT": 1366, + "TTTTAC": 1367, + "TTTTAG": 1368, + "TTTTTA": 1369, + "TTTTTT": 1370, + "TTTTTC": 1371, + "TTTTTG": 1372, + "TTTTCA": 1373, + "TTTTCT": 1374, + "TTTTCC": 1375, + "TTTTCG": 1376, + "TTTTGA": 1377, + "TTTTGT": 1378, + "TTTTGC": 1379, + "TTTTGG": 1380, + "TTTCAA": 1381, + "TTTCAT": 1382, + "TTTCAC": 1383, + "TTTCAG": 1384, + "TTTCTA": 1385, + "TTTCTT": 1386, + "TTTCTC": 1387, + "TTTCTG": 1388, + "TTTCCA": 1389, + "TTTCCT": 1390, + "TTTCCC": 1391, + "TTTCCG": 1392, + "TTTCGA": 1393, + "TTTCGT": 1394, + "TTTCGC": 1395, + "TTTCGG": 1396, + "TTTGAA": 1397, + "TTTGAT": 1398, + "TTTGAC": 1399, + "TTTGAG": 1400, + "TTTGTA": 1401, + "TTTGTT": 1402, + "TTTGTC": 1403, + "TTTGTG": 1404, + "TTTGCA": 1405, + "TTTGCT": 1406, + "TTTGCC": 1407, + "TTTGCG": 1408, + "TTTGGA": 1409, + "TTTGGT": 1410, + "TTTGGC": 1411, + "TTTGGG": 1412, + "TTCAAA": 1413, + "TTCAAT": 1414, + "TTCAAC": 1415, + "TTCAAG": 1416, + "TTCATA": 1417, + "TTCATT": 1418, + "TTCATC": 1419, + "TTCATG": 1420, + "TTCACA": 1421, + "TTCACT": 1422, + "TTCACC": 1423, + "TTCACG": 1424, + "TTCAGA": 1425, + "TTCAGT": 1426, + "TTCAGC": 1427, + "TTCAGG": 1428, + "TTCTAA": 1429, + "TTCTAT": 1430, + "TTCTAC": 1431, + "TTCTAG": 1432, + "TTCTTA": 1433, + "TTCTTT": 1434, + "TTCTTC": 1435, + "TTCTTG": 1436, + "TTCTCA": 1437, + "TTCTCT": 1438, + "TTCTCC": 1439, + "TTCTCG": 1440, + "TTCTGA": 1441, + "TTCTGT": 1442, + "TTCTGC": 1443, + "TTCTGG": 1444, + "TTCCAA": 1445, + "TTCCAT": 1446, + "TTCCAC": 1447, + "TTCCAG": 1448, + "TTCCTA": 1449, + "TTCCTT": 1450, + "TTCCTC": 1451, + "TTCCTG": 1452, + "TTCCCA": 1453, + "TTCCCT": 1454, + "TTCCCC": 1455, + "TTCCCG": 1456, + "TTCCGA": 1457, + "TTCCGT": 1458, + "TTCCGC": 1459, + "TTCCGG": 1460, + "TTCGAA": 1461, + "TTCGAT": 1462, + "TTCGAC": 1463, + "TTCGAG": 1464, + "TTCGTA": 1465, + "TTCGTT": 1466, + "TTCGTC": 1467, + "TTCGTG": 1468, + "TTCGCA": 1469, + "TTCGCT": 1470, + "TTCGCC": 1471, + "TTCGCG": 1472, + "TTCGGA": 1473, + "TTCGGT": 1474, + "TTCGGC": 1475, + "TTCGGG": 1476, + "TTGAAA": 1477, + "TTGAAT": 1478, + "TTGAAC": 1479, + "TTGAAG": 1480, + "TTGATA": 1481, + "TTGATT": 1482, + "TTGATC": 1483, + "TTGATG": 1484, + "TTGACA": 1485, + "TTGACT": 1486, + "TTGACC": 1487, + "TTGACG": 1488, + "TTGAGA": 1489, + "TTGAGT": 1490, + "TTGAGC": 1491, + "TTGAGG": 1492, + "TTGTAA": 1493, + "TTGTAT": 1494, + "TTGTAC": 1495, + "TTGTAG": 1496, + "TTGTTA": 1497, + "TTGTTT": 1498, + "TTGTTC": 1499, + "TTGTTG": 1500, + "TTGTCA": 1501, + "TTGTCT": 1502, + "TTGTCC": 1503, + "TTGTCG": 1504, + "TTGTGA": 1505, + "TTGTGT": 1506, + "TTGTGC": 1507, + "TTGTGG": 1508, + "TTGCAA": 1509, + "TTGCAT": 1510, + "TTGCAC": 1511, + "TTGCAG": 1512, + "TTGCTA": 1513, + "TTGCTT": 1514, + "TTGCTC": 1515, + "TTGCTG": 1516, + "TTGCCA": 1517, + "TTGCCT": 1518, + "TTGCCC": 1519, + "TTGCCG": 1520, + "TTGCGA": 1521, + "TTGCGT": 1522, + "TTGCGC": 1523, + "TTGCGG": 1524, + "TTGGAA": 1525, + "TTGGAT": 1526, + "TTGGAC": 1527, + "TTGGAG": 1528, + "TTGGTA": 1529, + "TTGGTT": 1530, + "TTGGTC": 1531, + "TTGGTG": 1532, + "TTGGCA": 1533, + "TTGGCT": 1534, + "TTGGCC": 1535, + "TTGGCG": 1536, + "TTGGGA": 1537, + "TTGGGT": 1538, + "TTGGGC": 1539, + "TTGGGG": 1540, + "TCAAAA": 1541, + "TCAAAT": 1542, + "TCAAAC": 1543, + "TCAAAG": 1544, + "TCAATA": 1545, + "TCAATT": 1546, + "TCAATC": 1547, + "TCAATG": 1548, + "TCAACA": 1549, + "TCAACT": 1550, + "TCAACC": 1551, + "TCAACG": 1552, + "TCAAGA": 1553, + "TCAAGT": 1554, + "TCAAGC": 1555, + "TCAAGG": 1556, + "TCATAA": 1557, + "TCATAT": 1558, + "TCATAC": 1559, + "TCATAG": 1560, + "TCATTA": 1561, + "TCATTT": 1562, + "TCATTC": 1563, + "TCATTG": 1564, + "TCATCA": 1565, + "TCATCT": 1566, + "TCATCC": 1567, + "TCATCG": 1568, + "TCATGA": 1569, + "TCATGT": 1570, + "TCATGC": 1571, + "TCATGG": 1572, + "TCACAA": 1573, + "TCACAT": 1574, + "TCACAC": 1575, + "TCACAG": 1576, + "TCACTA": 1577, + "TCACTT": 1578, + "TCACTC": 1579, + "TCACTG": 1580, + "TCACCA": 1581, + "TCACCT": 1582, + "TCACCC": 1583, + "TCACCG": 1584, + "TCACGA": 1585, + "TCACGT": 1586, + "TCACGC": 1587, + "TCACGG": 1588, + "TCAGAA": 1589, + "TCAGAT": 1590, + "TCAGAC": 1591, + "TCAGAG": 1592, + "TCAGTA": 1593, + "TCAGTT": 1594, + "TCAGTC": 1595, + "TCAGTG": 1596, + "TCAGCA": 1597, + "TCAGCT": 1598, + "TCAGCC": 1599, + "TCAGCG": 1600, + "TCAGGA": 1601, + "TCAGGT": 1602, + "TCAGGC": 1603, + "TCAGGG": 1604, + "TCTAAA": 1605, + "TCTAAT": 1606, + "TCTAAC": 1607, + "TCTAAG": 1608, + "TCTATA": 1609, + "TCTATT": 1610, + "TCTATC": 1611, + "TCTATG": 1612, + "TCTACA": 1613, + "TCTACT": 1614, + "TCTACC": 1615, + "TCTACG": 1616, + "TCTAGA": 1617, + "TCTAGT": 1618, + "TCTAGC": 1619, + "TCTAGG": 1620, + "TCTTAA": 1621, + "TCTTAT": 1622, + "TCTTAC": 1623, + "TCTTAG": 1624, + "TCTTTA": 1625, + "TCTTTT": 1626, + "TCTTTC": 1627, + "TCTTTG": 1628, + "TCTTCA": 1629, + "TCTTCT": 1630, + "TCTTCC": 1631, + "TCTTCG": 1632, + "TCTTGA": 1633, + "TCTTGT": 1634, + "TCTTGC": 1635, + "TCTTGG": 1636, + "TCTCAA": 1637, + "TCTCAT": 1638, + "TCTCAC": 1639, + "TCTCAG": 1640, + "TCTCTA": 1641, + "TCTCTT": 1642, + "TCTCTC": 1643, + "TCTCTG": 1644, + "TCTCCA": 1645, + "TCTCCT": 1646, + "TCTCCC": 1647, + "TCTCCG": 1648, + "TCTCGA": 1649, + "TCTCGT": 1650, + "TCTCGC": 1651, + "TCTCGG": 1652, + "TCTGAA": 1653, + "TCTGAT": 1654, + "TCTGAC": 1655, + "TCTGAG": 1656, + "TCTGTA": 1657, + "TCTGTT": 1658, + "TCTGTC": 1659, + "TCTGTG": 1660, + "TCTGCA": 1661, + "TCTGCT": 1662, + "TCTGCC": 1663, + "TCTGCG": 1664, + "TCTGGA": 1665, + "TCTGGT": 1666, + "TCTGGC": 1667, + "TCTGGG": 1668, + "TCCAAA": 1669, + "TCCAAT": 1670, + "TCCAAC": 1671, + "TCCAAG": 1672, + "TCCATA": 1673, + "TCCATT": 1674, + "TCCATC": 1675, + "TCCATG": 1676, + "TCCACA": 1677, + "TCCACT": 1678, + "TCCACC": 1679, + "TCCACG": 1680, + "TCCAGA": 1681, + "TCCAGT": 1682, + "TCCAGC": 1683, + "TCCAGG": 1684, + "TCCTAA": 1685, + "TCCTAT": 1686, + "TCCTAC": 1687, + "TCCTAG": 1688, + "TCCTTA": 1689, + "TCCTTT": 1690, + "TCCTTC": 1691, + "TCCTTG": 1692, + "TCCTCA": 1693, + "TCCTCT": 1694, + "TCCTCC": 1695, + "TCCTCG": 1696, + "TCCTGA": 1697, + "TCCTGT": 1698, + "TCCTGC": 1699, + "TCCTGG": 1700, + "TCCCAA": 1701, + "TCCCAT": 1702, + "TCCCAC": 1703, + "TCCCAG": 1704, + "TCCCTA": 1705, + "TCCCTT": 1706, + "TCCCTC": 1707, + "TCCCTG": 1708, + "TCCCCA": 1709, + "TCCCCT": 1710, + "TCCCCC": 1711, + "TCCCCG": 1712, + "TCCCGA": 1713, + "TCCCGT": 1714, + "TCCCGC": 1715, + "TCCCGG": 1716, + "TCCGAA": 1717, + "TCCGAT": 1718, + "TCCGAC": 1719, + "TCCGAG": 1720, + "TCCGTA": 1721, + "TCCGTT": 1722, + "TCCGTC": 1723, + "TCCGTG": 1724, + "TCCGCA": 1725, + "TCCGCT": 1726, + "TCCGCC": 1727, + "TCCGCG": 1728, + "TCCGGA": 1729, + "TCCGGT": 1730, + "TCCGGC": 1731, + "TCCGGG": 1732, + "TCGAAA": 1733, + "TCGAAT": 1734, + "TCGAAC": 1735, + "TCGAAG": 1736, + "TCGATA": 1737, + "TCGATT": 1738, + "TCGATC": 1739, + "TCGATG": 1740, + "TCGACA": 1741, + "TCGACT": 1742, + "TCGACC": 1743, + "TCGACG": 1744, + "TCGAGA": 1745, + "TCGAGT": 1746, + "TCGAGC": 1747, + "TCGAGG": 1748, + "TCGTAA": 1749, + "TCGTAT": 1750, + "TCGTAC": 1751, + "TCGTAG": 1752, + "TCGTTA": 1753, + "TCGTTT": 1754, + "TCGTTC": 1755, + "TCGTTG": 1756, + "TCGTCA": 1757, + "TCGTCT": 1758, + "TCGTCC": 1759, + "TCGTCG": 1760, + "TCGTGA": 1761, + "TCGTGT": 1762, + "TCGTGC": 1763, + "TCGTGG": 1764, + "TCGCAA": 1765, + "TCGCAT": 1766, + "TCGCAC": 1767, + "TCGCAG": 1768, + "TCGCTA": 1769, + "TCGCTT": 1770, + "TCGCTC": 1771, + "TCGCTG": 1772, + "TCGCCA": 1773, + "TCGCCT": 1774, + "TCGCCC": 1775, + "TCGCCG": 1776, + "TCGCGA": 1777, + "TCGCGT": 1778, + "TCGCGC": 1779, + "TCGCGG": 1780, + "TCGGAA": 1781, + "TCGGAT": 1782, + "TCGGAC": 1783, + "TCGGAG": 1784, + "TCGGTA": 1785, + "TCGGTT": 1786, + "TCGGTC": 1787, + "TCGGTG": 1788, + "TCGGCA": 1789, + "TCGGCT": 1790, + "TCGGCC": 1791, + "TCGGCG": 1792, + "TCGGGA": 1793, + "TCGGGT": 1794, + "TCGGGC": 1795, + "TCGGGG": 1796, + "TGAAAA": 1797, + "TGAAAT": 1798, + "TGAAAC": 1799, + "TGAAAG": 1800, + "TGAATA": 1801, + "TGAATT": 1802, + "TGAATC": 1803, + "TGAATG": 1804, + "TGAACA": 1805, + "TGAACT": 1806, + "TGAACC": 1807, + "TGAACG": 1808, + "TGAAGA": 1809, + "TGAAGT": 1810, + "TGAAGC": 1811, + "TGAAGG": 1812, + "TGATAA": 1813, + "TGATAT": 1814, + "TGATAC": 1815, + "TGATAG": 1816, + "TGATTA": 1817, + "TGATTT": 1818, + "TGATTC": 1819, + "TGATTG": 1820, + "TGATCA": 1821, + "TGATCT": 1822, + "TGATCC": 1823, + "TGATCG": 1824, + "TGATGA": 1825, + "TGATGT": 1826, + "TGATGC": 1827, + "TGATGG": 1828, + "TGACAA": 1829, + "TGACAT": 1830, + "TGACAC": 1831, + "TGACAG": 1832, + "TGACTA": 1833, + "TGACTT": 1834, + "TGACTC": 1835, + "TGACTG": 1836, + "TGACCA": 1837, + "TGACCT": 1838, + "TGACCC": 1839, + "TGACCG": 1840, + "TGACGA": 1841, + "TGACGT": 1842, + "TGACGC": 1843, + "TGACGG": 1844, + "TGAGAA": 1845, + "TGAGAT": 1846, + "TGAGAC": 1847, + "TGAGAG": 1848, + "TGAGTA": 1849, + "TGAGTT": 1850, + "TGAGTC": 1851, + "TGAGTG": 1852, + "TGAGCA": 1853, + "TGAGCT": 1854, + "TGAGCC": 1855, + "TGAGCG": 1856, + "TGAGGA": 1857, + "TGAGGT": 1858, + "TGAGGC": 1859, + "TGAGGG": 1860, + "TGTAAA": 1861, + "TGTAAT": 1862, + "TGTAAC": 1863, + "TGTAAG": 1864, + "TGTATA": 1865, + "TGTATT": 1866, + "TGTATC": 1867, + "TGTATG": 1868, + "TGTACA": 1869, + "TGTACT": 1870, + "TGTACC": 1871, + "TGTACG": 1872, + "TGTAGA": 1873, + "TGTAGT": 1874, + "TGTAGC": 1875, + "TGTAGG": 1876, + "TGTTAA": 1877, + "TGTTAT": 1878, + "TGTTAC": 1879, + "TGTTAG": 1880, + "TGTTTA": 1881, + "TGTTTT": 1882, + "TGTTTC": 1883, + "TGTTTG": 1884, + "TGTTCA": 1885, + "TGTTCT": 1886, + "TGTTCC": 1887, + "TGTTCG": 1888, + "TGTTGA": 1889, + "TGTTGT": 1890, + "TGTTGC": 1891, + "TGTTGG": 1892, + "TGTCAA": 1893, + "TGTCAT": 1894, + "TGTCAC": 1895, + "TGTCAG": 1896, + "TGTCTA": 1897, + "TGTCTT": 1898, + "TGTCTC": 1899, + "TGTCTG": 1900, + "TGTCCA": 1901, + "TGTCCT": 1902, + "TGTCCC": 1903, + "TGTCCG": 1904, + "TGTCGA": 1905, + "TGTCGT": 1906, + "TGTCGC": 1907, + "TGTCGG": 1908, + "TGTGAA": 1909, + "TGTGAT": 1910, + "TGTGAC": 1911, + "TGTGAG": 1912, + "TGTGTA": 1913, + "TGTGTT": 1914, + "TGTGTC": 1915, + "TGTGTG": 1916, + "TGTGCA": 1917, + "TGTGCT": 1918, + "TGTGCC": 1919, + "TGTGCG": 1920, + "TGTGGA": 1921, + "TGTGGT": 1922, + "TGTGGC": 1923, + "TGTGGG": 1924, + "TGCAAA": 1925, + "TGCAAT": 1926, + "TGCAAC": 1927, + "TGCAAG": 1928, + "TGCATA": 1929, + "TGCATT": 1930, + "TGCATC": 1931, + "TGCATG": 1932, + "TGCACA": 1933, + "TGCACT": 1934, + "TGCACC": 1935, + "TGCACG": 1936, + "TGCAGA": 1937, + "TGCAGT": 1938, + "TGCAGC": 1939, + "TGCAGG": 1940, + "TGCTAA": 1941, + "TGCTAT": 1942, + "TGCTAC": 1943, + "TGCTAG": 1944, + "TGCTTA": 1945, + "TGCTTT": 1946, + "TGCTTC": 1947, + "TGCTTG": 1948, + "TGCTCA": 1949, + "TGCTCT": 1950, + "TGCTCC": 1951, + "TGCTCG": 1952, + "TGCTGA": 1953, + "TGCTGT": 1954, + "TGCTGC": 1955, + "TGCTGG": 1956, + "TGCCAA": 1957, + "TGCCAT": 1958, + "TGCCAC": 1959, + "TGCCAG": 1960, + "TGCCTA": 1961, + "TGCCTT": 1962, + "TGCCTC": 1963, + "TGCCTG": 1964, + "TGCCCA": 1965, + "TGCCCT": 1966, + "TGCCCC": 1967, + "TGCCCG": 1968, + "TGCCGA": 1969, + "TGCCGT": 1970, + "TGCCGC": 1971, + "TGCCGG": 1972, + "TGCGAA": 1973, + "TGCGAT": 1974, + "TGCGAC": 1975, + "TGCGAG": 1976, + "TGCGTA": 1977, + "TGCGTT": 1978, + "TGCGTC": 1979, + "TGCGTG": 1980, + "TGCGCA": 1981, + "TGCGCT": 1982, + "TGCGCC": 1983, + "TGCGCG": 1984, + "TGCGGA": 1985, + "TGCGGT": 1986, + "TGCGGC": 1987, + "TGCGGG": 1988, + "TGGAAA": 1989, + "TGGAAT": 1990, + "TGGAAC": 1991, + "TGGAAG": 1992, + "TGGATA": 1993, + "TGGATT": 1994, + "TGGATC": 1995, + "TGGATG": 1996, + "TGGACA": 1997, + "TGGACT": 1998, + "TGGACC": 1999, + "TGGACG": 2000, + "TGGAGA": 2001, + "TGGAGT": 2002, + "TGGAGC": 2003, + "TGGAGG": 2004, + "TGGTAA": 2005, + "TGGTAT": 2006, + "TGGTAC": 2007, + "TGGTAG": 2008, + "TGGTTA": 2009, + "TGGTTT": 2010, + "TGGTTC": 2011, + "TGGTTG": 2012, + "TGGTCA": 2013, + "TGGTCT": 2014, + "TGGTCC": 2015, + "TGGTCG": 2016, + "TGGTGA": 2017, + "TGGTGT": 2018, + "TGGTGC": 2019, + "TGGTGG": 2020, + "TGGCAA": 2021, + "TGGCAT": 2022, + "TGGCAC": 2023, + "TGGCAG": 2024, + "TGGCTA": 2025, + "TGGCTT": 2026, + "TGGCTC": 2027, + "TGGCTG": 2028, + "TGGCCA": 2029, + "TGGCCT": 2030, + "TGGCCC": 2031, + "TGGCCG": 2032, + "TGGCGA": 2033, + "TGGCGT": 2034, + "TGGCGC": 2035, + "TGGCGG": 2036, + "TGGGAA": 2037, + "TGGGAT": 2038, + "TGGGAC": 2039, + "TGGGAG": 2040, + "TGGGTA": 2041, + "TGGGTT": 2042, + "TGGGTC": 2043, + "TGGGTG": 2044, + "TGGGCA": 2045, + "TGGGCT": 2046, + "TGGGCC": 2047, + "TGGGCG": 2048, + "TGGGGA": 2049, + "TGGGGT": 2050, + "TGGGGC": 2051, + "TGGGGG": 2052, + "CAAAAA": 2053, + "CAAAAT": 2054, + "CAAAAC": 2055, + "CAAAAG": 2056, + "CAAATA": 2057, + "CAAATT": 2058, + "CAAATC": 2059, + "CAAATG": 2060, + "CAAACA": 2061, + "CAAACT": 2062, + "CAAACC": 2063, + "CAAACG": 2064, + "CAAAGA": 2065, + "CAAAGT": 2066, + "CAAAGC": 2067, + "CAAAGG": 2068, + "CAATAA": 2069, + "CAATAT": 2070, + "CAATAC": 2071, + "CAATAG": 2072, + "CAATTA": 2073, + "CAATTT": 2074, + "CAATTC": 2075, + "CAATTG": 2076, + "CAATCA": 2077, + "CAATCT": 2078, + "CAATCC": 2079, + "CAATCG": 2080, + "CAATGA": 2081, + "CAATGT": 2082, + "CAATGC": 2083, + "CAATGG": 2084, + "CAACAA": 2085, + "CAACAT": 2086, + "CAACAC": 2087, + "CAACAG": 2088, + "CAACTA": 2089, + "CAACTT": 2090, + "CAACTC": 2091, + "CAACTG": 2092, + "CAACCA": 2093, + "CAACCT": 2094, + "CAACCC": 2095, + "CAACCG": 2096, + "CAACGA": 2097, + "CAACGT": 2098, + "CAACGC": 2099, + "CAACGG": 2100, + "CAAGAA": 2101, + "CAAGAT": 2102, + "CAAGAC": 2103, + "CAAGAG": 2104, + "CAAGTA": 2105, + "CAAGTT": 2106, + "CAAGTC": 2107, + "CAAGTG": 2108, + "CAAGCA": 2109, + "CAAGCT": 2110, + "CAAGCC": 2111, + "CAAGCG": 2112, + "CAAGGA": 2113, + "CAAGGT": 2114, + "CAAGGC": 2115, + "CAAGGG": 2116, + "CATAAA": 2117, + "CATAAT": 2118, + "CATAAC": 2119, + "CATAAG": 2120, + "CATATA": 2121, + "CATATT": 2122, + "CATATC": 2123, + "CATATG": 2124, + "CATACA": 2125, + "CATACT": 2126, + "CATACC": 2127, + "CATACG": 2128, + "CATAGA": 2129, + "CATAGT": 2130, + "CATAGC": 2131, + "CATAGG": 2132, + "CATTAA": 2133, + "CATTAT": 2134, + "CATTAC": 2135, + "CATTAG": 2136, + "CATTTA": 2137, + "CATTTT": 2138, + "CATTTC": 2139, + "CATTTG": 2140, + "CATTCA": 2141, + "CATTCT": 2142, + "CATTCC": 2143, + "CATTCG": 2144, + "CATTGA": 2145, + "CATTGT": 2146, + "CATTGC": 2147, + "CATTGG": 2148, + "CATCAA": 2149, + "CATCAT": 2150, + "CATCAC": 2151, + "CATCAG": 2152, + "CATCTA": 2153, + "CATCTT": 2154, + "CATCTC": 2155, + "CATCTG": 2156, + "CATCCA": 2157, + "CATCCT": 2158, + "CATCCC": 2159, + "CATCCG": 2160, + "CATCGA": 2161, + "CATCGT": 2162, + "CATCGC": 2163, + "CATCGG": 2164, + "CATGAA": 2165, + "CATGAT": 2166, + "CATGAC": 2167, + "CATGAG": 2168, + "CATGTA": 2169, + "CATGTT": 2170, + "CATGTC": 2171, + "CATGTG": 2172, + "CATGCA": 2173, + "CATGCT": 2174, + "CATGCC": 2175, + "CATGCG": 2176, + "CATGGA": 2177, + "CATGGT": 2178, + "CATGGC": 2179, + "CATGGG": 2180, + "CACAAA": 2181, + "CACAAT": 2182, + "CACAAC": 2183, + "CACAAG": 2184, + "CACATA": 2185, + "CACATT": 2186, + "CACATC": 2187, + "CACATG": 2188, + "CACACA": 2189, + "CACACT": 2190, + "CACACC": 2191, + "CACACG": 2192, + "CACAGA": 2193, + "CACAGT": 2194, + "CACAGC": 2195, + "CACAGG": 2196, + "CACTAA": 2197, + "CACTAT": 2198, + "CACTAC": 2199, + "CACTAG": 2200, + "CACTTA": 2201, + "CACTTT": 2202, + "CACTTC": 2203, + "CACTTG": 2204, + "CACTCA": 2205, + "CACTCT": 2206, + "CACTCC": 2207, + "CACTCG": 2208, + "CACTGA": 2209, + "CACTGT": 2210, + "CACTGC": 2211, + "CACTGG": 2212, + "CACCAA": 2213, + "CACCAT": 2214, + "CACCAC": 2215, + "CACCAG": 2216, + "CACCTA": 2217, + "CACCTT": 2218, + "CACCTC": 2219, + "CACCTG": 2220, + "CACCCA": 2221, + "CACCCT": 2222, + "CACCCC": 2223, + "CACCCG": 2224, + "CACCGA": 2225, + "CACCGT": 2226, + "CACCGC": 2227, + "CACCGG": 2228, + "CACGAA": 2229, + "CACGAT": 2230, + "CACGAC": 2231, + "CACGAG": 2232, + "CACGTA": 2233, + "CACGTT": 2234, + "CACGTC": 2235, + "CACGTG": 2236, + "CACGCA": 2237, + "CACGCT": 2238, + "CACGCC": 2239, + "CACGCG": 2240, + "CACGGA": 2241, + "CACGGT": 2242, + "CACGGC": 2243, + "CACGGG": 2244, + "CAGAAA": 2245, + "CAGAAT": 2246, + "CAGAAC": 2247, + "CAGAAG": 2248, + "CAGATA": 2249, + "CAGATT": 2250, + "CAGATC": 2251, + "CAGATG": 2252, + "CAGACA": 2253, + "CAGACT": 2254, + "CAGACC": 2255, + "CAGACG": 2256, + "CAGAGA": 2257, + "CAGAGT": 2258, + "CAGAGC": 2259, + "CAGAGG": 2260, + "CAGTAA": 2261, + "CAGTAT": 2262, + "CAGTAC": 2263, + "CAGTAG": 2264, + "CAGTTA": 2265, + "CAGTTT": 2266, + "CAGTTC": 2267, + "CAGTTG": 2268, + "CAGTCA": 2269, + "CAGTCT": 2270, + "CAGTCC": 2271, + "CAGTCG": 2272, + "CAGTGA": 2273, + "CAGTGT": 2274, + "CAGTGC": 2275, + "CAGTGG": 2276, + "CAGCAA": 2277, + "CAGCAT": 2278, + "CAGCAC": 2279, + "CAGCAG": 2280, + "CAGCTA": 2281, + "CAGCTT": 2282, + "CAGCTC": 2283, + "CAGCTG": 2284, + "CAGCCA": 2285, + "CAGCCT": 2286, + "CAGCCC": 2287, + "CAGCCG": 2288, + "CAGCGA": 2289, + "CAGCGT": 2290, + "CAGCGC": 2291, + "CAGCGG": 2292, + "CAGGAA": 2293, + "CAGGAT": 2294, + "CAGGAC": 2295, + "CAGGAG": 2296, + "CAGGTA": 2297, + "CAGGTT": 2298, + "CAGGTC": 2299, + "CAGGTG": 2300, + "CAGGCA": 2301, + "CAGGCT": 2302, + "CAGGCC": 2303, + "CAGGCG": 2304, + "CAGGGA": 2305, + "CAGGGT": 2306, + "CAGGGC": 2307, + "CAGGGG": 2308, + "CTAAAA": 2309, + "CTAAAT": 2310, + "CTAAAC": 2311, + "CTAAAG": 2312, + "CTAATA": 2313, + "CTAATT": 2314, + "CTAATC": 2315, + "CTAATG": 2316, + "CTAACA": 2317, + "CTAACT": 2318, + "CTAACC": 2319, + "CTAACG": 2320, + "CTAAGA": 2321, + "CTAAGT": 2322, + "CTAAGC": 2323, + "CTAAGG": 2324, + "CTATAA": 2325, + "CTATAT": 2326, + "CTATAC": 2327, + "CTATAG": 2328, + "CTATTA": 2329, + "CTATTT": 2330, + "CTATTC": 2331, + "CTATTG": 2332, + "CTATCA": 2333, + "CTATCT": 2334, + "CTATCC": 2335, + "CTATCG": 2336, + "CTATGA": 2337, + "CTATGT": 2338, + "CTATGC": 2339, + "CTATGG": 2340, + "CTACAA": 2341, + "CTACAT": 2342, + "CTACAC": 2343, + "CTACAG": 2344, + "CTACTA": 2345, + "CTACTT": 2346, + "CTACTC": 2347, + "CTACTG": 2348, + "CTACCA": 2349, + "CTACCT": 2350, + "CTACCC": 2351, + "CTACCG": 2352, + "CTACGA": 2353, + "CTACGT": 2354, + "CTACGC": 2355, + "CTACGG": 2356, + "CTAGAA": 2357, + "CTAGAT": 2358, + "CTAGAC": 2359, + "CTAGAG": 2360, + "CTAGTA": 2361, + "CTAGTT": 2362, + "CTAGTC": 2363, + "CTAGTG": 2364, + "CTAGCA": 2365, + "CTAGCT": 2366, + "CTAGCC": 2367, + "CTAGCG": 2368, + "CTAGGA": 2369, + "CTAGGT": 2370, + "CTAGGC": 2371, + "CTAGGG": 2372, + "CTTAAA": 2373, + "CTTAAT": 2374, + "CTTAAC": 2375, + "CTTAAG": 2376, + "CTTATA": 2377, + "CTTATT": 2378, + "CTTATC": 2379, + "CTTATG": 2380, + "CTTACA": 2381, + "CTTACT": 2382, + "CTTACC": 2383, + "CTTACG": 2384, + "CTTAGA": 2385, + "CTTAGT": 2386, + "CTTAGC": 2387, + "CTTAGG": 2388, + "CTTTAA": 2389, + "CTTTAT": 2390, + "CTTTAC": 2391, + "CTTTAG": 2392, + "CTTTTA": 2393, + "CTTTTT": 2394, + "CTTTTC": 2395, + "CTTTTG": 2396, + "CTTTCA": 2397, + "CTTTCT": 2398, + "CTTTCC": 2399, + "CTTTCG": 2400, + "CTTTGA": 2401, + "CTTTGT": 2402, + "CTTTGC": 2403, + "CTTTGG": 2404, + "CTTCAA": 2405, + "CTTCAT": 2406, + "CTTCAC": 2407, + "CTTCAG": 2408, + "CTTCTA": 2409, + "CTTCTT": 2410, + "CTTCTC": 2411, + "CTTCTG": 2412, + "CTTCCA": 2413, + "CTTCCT": 2414, + "CTTCCC": 2415, + "CTTCCG": 2416, + "CTTCGA": 2417, + "CTTCGT": 2418, + "CTTCGC": 2419, + "CTTCGG": 2420, + "CTTGAA": 2421, + "CTTGAT": 2422, + "CTTGAC": 2423, + "CTTGAG": 2424, + "CTTGTA": 2425, + "CTTGTT": 2426, + "CTTGTC": 2427, + "CTTGTG": 2428, + "CTTGCA": 2429, + "CTTGCT": 2430, + "CTTGCC": 2431, + "CTTGCG": 2432, + "CTTGGA": 2433, + "CTTGGT": 2434, + "CTTGGC": 2435, + "CTTGGG": 2436, + "CTCAAA": 2437, + "CTCAAT": 2438, + "CTCAAC": 2439, + "CTCAAG": 2440, + "CTCATA": 2441, + "CTCATT": 2442, + "CTCATC": 2443, + "CTCATG": 2444, + "CTCACA": 2445, + "CTCACT": 2446, + "CTCACC": 2447, + "CTCACG": 2448, + "CTCAGA": 2449, + "CTCAGT": 2450, + "CTCAGC": 2451, + "CTCAGG": 2452, + "CTCTAA": 2453, + "CTCTAT": 2454, + "CTCTAC": 2455, + "CTCTAG": 2456, + "CTCTTA": 2457, + "CTCTTT": 2458, + "CTCTTC": 2459, + "CTCTTG": 2460, + "CTCTCA": 2461, + "CTCTCT": 2462, + "CTCTCC": 2463, + "CTCTCG": 2464, + "CTCTGA": 2465, + "CTCTGT": 2466, + "CTCTGC": 2467, + "CTCTGG": 2468, + "CTCCAA": 2469, + "CTCCAT": 2470, + "CTCCAC": 2471, + "CTCCAG": 2472, + "CTCCTA": 2473, + "CTCCTT": 2474, + "CTCCTC": 2475, + "CTCCTG": 2476, + "CTCCCA": 2477, + "CTCCCT": 2478, + "CTCCCC": 2479, + "CTCCCG": 2480, + "CTCCGA": 2481, + "CTCCGT": 2482, + "CTCCGC": 2483, + "CTCCGG": 2484, + "CTCGAA": 2485, + "CTCGAT": 2486, + "CTCGAC": 2487, + "CTCGAG": 2488, + "CTCGTA": 2489, + "CTCGTT": 2490, + "CTCGTC": 2491, + "CTCGTG": 2492, + "CTCGCA": 2493, + "CTCGCT": 2494, + "CTCGCC": 2495, + "CTCGCG": 2496, + "CTCGGA": 2497, + "CTCGGT": 2498, + "CTCGGC": 2499, + "CTCGGG": 2500, + "CTGAAA": 2501, + "CTGAAT": 2502, + "CTGAAC": 2503, + "CTGAAG": 2504, + "CTGATA": 2505, + "CTGATT": 2506, + "CTGATC": 2507, + "CTGATG": 2508, + "CTGACA": 2509, + "CTGACT": 2510, + "CTGACC": 2511, + "CTGACG": 2512, + "CTGAGA": 2513, + "CTGAGT": 2514, + "CTGAGC": 2515, + "CTGAGG": 2516, + "CTGTAA": 2517, + "CTGTAT": 2518, + "CTGTAC": 2519, + "CTGTAG": 2520, + "CTGTTA": 2521, + "CTGTTT": 2522, + "CTGTTC": 2523, + "CTGTTG": 2524, + "CTGTCA": 2525, + "CTGTCT": 2526, + "CTGTCC": 2527, + "CTGTCG": 2528, + "CTGTGA": 2529, + "CTGTGT": 2530, + "CTGTGC": 2531, + "CTGTGG": 2532, + "CTGCAA": 2533, + "CTGCAT": 2534, + "CTGCAC": 2535, + "CTGCAG": 2536, + "CTGCTA": 2537, + "CTGCTT": 2538, + "CTGCTC": 2539, + "CTGCTG": 2540, + "CTGCCA": 2541, + "CTGCCT": 2542, + "CTGCCC": 2543, + "CTGCCG": 2544, + "CTGCGA": 2545, + "CTGCGT": 2546, + "CTGCGC": 2547, + "CTGCGG": 2548, + "CTGGAA": 2549, + "CTGGAT": 2550, + "CTGGAC": 2551, + "CTGGAG": 2552, + "CTGGTA": 2553, + "CTGGTT": 2554, + "CTGGTC": 2555, + "CTGGTG": 2556, + "CTGGCA": 2557, + "CTGGCT": 2558, + "CTGGCC": 2559, + "CTGGCG": 2560, + "CTGGGA": 2561, + "CTGGGT": 2562, + "CTGGGC": 2563, + "CTGGGG": 2564, + "CCAAAA": 2565, + "CCAAAT": 2566, + "CCAAAC": 2567, + "CCAAAG": 2568, + "CCAATA": 2569, + "CCAATT": 2570, + "CCAATC": 2571, + "CCAATG": 2572, + "CCAACA": 2573, + "CCAACT": 2574, + "CCAACC": 2575, + "CCAACG": 2576, + "CCAAGA": 2577, + "CCAAGT": 2578, + "CCAAGC": 2579, + "CCAAGG": 2580, + "CCATAA": 2581, + "CCATAT": 2582, + "CCATAC": 2583, + "CCATAG": 2584, + "CCATTA": 2585, + "CCATTT": 2586, + "CCATTC": 2587, + "CCATTG": 2588, + "CCATCA": 2589, + "CCATCT": 2590, + "CCATCC": 2591, + "CCATCG": 2592, + "CCATGA": 2593, + "CCATGT": 2594, + "CCATGC": 2595, + "CCATGG": 2596, + "CCACAA": 2597, + "CCACAT": 2598, + "CCACAC": 2599, + "CCACAG": 2600, + "CCACTA": 2601, + "CCACTT": 2602, + "CCACTC": 2603, + "CCACTG": 2604, + "CCACCA": 2605, + "CCACCT": 2606, + "CCACCC": 2607, + "CCACCG": 2608, + "CCACGA": 2609, + "CCACGT": 2610, + "CCACGC": 2611, + "CCACGG": 2612, + "CCAGAA": 2613, + "CCAGAT": 2614, + "CCAGAC": 2615, + "CCAGAG": 2616, + "CCAGTA": 2617, + "CCAGTT": 2618, + "CCAGTC": 2619, + "CCAGTG": 2620, + "CCAGCA": 2621, + "CCAGCT": 2622, + "CCAGCC": 2623, + "CCAGCG": 2624, + "CCAGGA": 2625, + "CCAGGT": 2626, + "CCAGGC": 2627, + "CCAGGG": 2628, + "CCTAAA": 2629, + "CCTAAT": 2630, + "CCTAAC": 2631, + "CCTAAG": 2632, + "CCTATA": 2633, + "CCTATT": 2634, + "CCTATC": 2635, + "CCTATG": 2636, + "CCTACA": 2637, + "CCTACT": 2638, + "CCTACC": 2639, + "CCTACG": 2640, + "CCTAGA": 2641, + "CCTAGT": 2642, + "CCTAGC": 2643, + "CCTAGG": 2644, + "CCTTAA": 2645, + "CCTTAT": 2646, + "CCTTAC": 2647, + "CCTTAG": 2648, + "CCTTTA": 2649, + "CCTTTT": 2650, + "CCTTTC": 2651, + "CCTTTG": 2652, + "CCTTCA": 2653, + "CCTTCT": 2654, + "CCTTCC": 2655, + "CCTTCG": 2656, + "CCTTGA": 2657, + "CCTTGT": 2658, + "CCTTGC": 2659, + "CCTTGG": 2660, + "CCTCAA": 2661, + "CCTCAT": 2662, + "CCTCAC": 2663, + "CCTCAG": 2664, + "CCTCTA": 2665, + "CCTCTT": 2666, + "CCTCTC": 2667, + "CCTCTG": 2668, + "CCTCCA": 2669, + "CCTCCT": 2670, + "CCTCCC": 2671, + "CCTCCG": 2672, + "CCTCGA": 2673, + "CCTCGT": 2674, + "CCTCGC": 2675, + "CCTCGG": 2676, + "CCTGAA": 2677, + "CCTGAT": 2678, + "CCTGAC": 2679, + "CCTGAG": 2680, + "CCTGTA": 2681, + "CCTGTT": 2682, + "CCTGTC": 2683, + "CCTGTG": 2684, + "CCTGCA": 2685, + "CCTGCT": 2686, + "CCTGCC": 2687, + "CCTGCG": 2688, + "CCTGGA": 2689, + "CCTGGT": 2690, + "CCTGGC": 2691, + "CCTGGG": 2692, + "CCCAAA": 2693, + "CCCAAT": 2694, + "CCCAAC": 2695, + "CCCAAG": 2696, + "CCCATA": 2697, + "CCCATT": 2698, + "CCCATC": 2699, + "CCCATG": 2700, + "CCCACA": 2701, + "CCCACT": 2702, + "CCCACC": 2703, + "CCCACG": 2704, + "CCCAGA": 2705, + "CCCAGT": 2706, + "CCCAGC": 2707, + "CCCAGG": 2708, + "CCCTAA": 2709, + "CCCTAT": 2710, + "CCCTAC": 2711, + "CCCTAG": 2712, + "CCCTTA": 2713, + "CCCTTT": 2714, + "CCCTTC": 2715, + "CCCTTG": 2716, + "CCCTCA": 2717, + "CCCTCT": 2718, + "CCCTCC": 2719, + "CCCTCG": 2720, + "CCCTGA": 2721, + "CCCTGT": 2722, + "CCCTGC": 2723, + "CCCTGG": 2724, + "CCCCAA": 2725, + "CCCCAT": 2726, + "CCCCAC": 2727, + "CCCCAG": 2728, + "CCCCTA": 2729, + "CCCCTT": 2730, + "CCCCTC": 2731, + "CCCCTG": 2732, + "CCCCCA": 2733, + "CCCCCT": 2734, + "CCCCCC": 2735, + "CCCCCG": 2736, + "CCCCGA": 2737, + "CCCCGT": 2738, + "CCCCGC": 2739, + "CCCCGG": 2740, + "CCCGAA": 2741, + "CCCGAT": 2742, + "CCCGAC": 2743, + "CCCGAG": 2744, + "CCCGTA": 2745, + "CCCGTT": 2746, + "CCCGTC": 2747, + "CCCGTG": 2748, + "CCCGCA": 2749, + "CCCGCT": 2750, + "CCCGCC": 2751, + "CCCGCG": 2752, + "CCCGGA": 2753, + "CCCGGT": 2754, + "CCCGGC": 2755, + "CCCGGG": 2756, + "CCGAAA": 2757, + "CCGAAT": 2758, + "CCGAAC": 2759, + "CCGAAG": 2760, + "CCGATA": 2761, + "CCGATT": 2762, + "CCGATC": 2763, + "CCGATG": 2764, + "CCGACA": 2765, + "CCGACT": 2766, + "CCGACC": 2767, + "CCGACG": 2768, + "CCGAGA": 2769, + "CCGAGT": 2770, + "CCGAGC": 2771, + "CCGAGG": 2772, + "CCGTAA": 2773, + "CCGTAT": 2774, + "CCGTAC": 2775, + "CCGTAG": 2776, + "CCGTTA": 2777, + "CCGTTT": 2778, + "CCGTTC": 2779, + "CCGTTG": 2780, + "CCGTCA": 2781, + "CCGTCT": 2782, + "CCGTCC": 2783, + "CCGTCG": 2784, + "CCGTGA": 2785, + "CCGTGT": 2786, + "CCGTGC": 2787, + "CCGTGG": 2788, + "CCGCAA": 2789, + "CCGCAT": 2790, + "CCGCAC": 2791, + "CCGCAG": 2792, + "CCGCTA": 2793, + "CCGCTT": 2794, + "CCGCTC": 2795, + "CCGCTG": 2796, + "CCGCCA": 2797, + "CCGCCT": 2798, + "CCGCCC": 2799, + "CCGCCG": 2800, + "CCGCGA": 2801, + "CCGCGT": 2802, + "CCGCGC": 2803, + "CCGCGG": 2804, + "CCGGAA": 2805, + "CCGGAT": 2806, + "CCGGAC": 2807, + "CCGGAG": 2808, + "CCGGTA": 2809, + "CCGGTT": 2810, + "CCGGTC": 2811, + "CCGGTG": 2812, + "CCGGCA": 2813, + "CCGGCT": 2814, + "CCGGCC": 2815, + "CCGGCG": 2816, + "CCGGGA": 2817, + "CCGGGT": 2818, + "CCGGGC": 2819, + "CCGGGG": 2820, + "CGAAAA": 2821, + "CGAAAT": 2822, + "CGAAAC": 2823, + "CGAAAG": 2824, + "CGAATA": 2825, + "CGAATT": 2826, + "CGAATC": 2827, + "CGAATG": 2828, + "CGAACA": 2829, + "CGAACT": 2830, + "CGAACC": 2831, + "CGAACG": 2832, + "CGAAGA": 2833, + "CGAAGT": 2834, + "CGAAGC": 2835, + "CGAAGG": 2836, + "CGATAA": 2837, + "CGATAT": 2838, + "CGATAC": 2839, + "CGATAG": 2840, + "CGATTA": 2841, + "CGATTT": 2842, + "CGATTC": 2843, + "CGATTG": 2844, + "CGATCA": 2845, + "CGATCT": 2846, + "CGATCC": 2847, + "CGATCG": 2848, + "CGATGA": 2849, + "CGATGT": 2850, + "CGATGC": 2851, + "CGATGG": 2852, + "CGACAA": 2853, + "CGACAT": 2854, + "CGACAC": 2855, + "CGACAG": 2856, + "CGACTA": 2857, + "CGACTT": 2858, + "CGACTC": 2859, + "CGACTG": 2860, + "CGACCA": 2861, + "CGACCT": 2862, + "CGACCC": 2863, + "CGACCG": 2864, + "CGACGA": 2865, + "CGACGT": 2866, + "CGACGC": 2867, + "CGACGG": 2868, + "CGAGAA": 2869, + "CGAGAT": 2870, + "CGAGAC": 2871, + "CGAGAG": 2872, + "CGAGTA": 2873, + "CGAGTT": 2874, + "CGAGTC": 2875, + "CGAGTG": 2876, + "CGAGCA": 2877, + "CGAGCT": 2878, + "CGAGCC": 2879, + "CGAGCG": 2880, + "CGAGGA": 2881, + "CGAGGT": 2882, + "CGAGGC": 2883, + "CGAGGG": 2884, + "CGTAAA": 2885, + "CGTAAT": 2886, + "CGTAAC": 2887, + "CGTAAG": 2888, + "CGTATA": 2889, + "CGTATT": 2890, + "CGTATC": 2891, + "CGTATG": 2892, + "CGTACA": 2893, + "CGTACT": 2894, + "CGTACC": 2895, + "CGTACG": 2896, + "CGTAGA": 2897, + "CGTAGT": 2898, + "CGTAGC": 2899, + "CGTAGG": 2900, + "CGTTAA": 2901, + "CGTTAT": 2902, + "CGTTAC": 2903, + "CGTTAG": 2904, + "CGTTTA": 2905, + "CGTTTT": 2906, + "CGTTTC": 2907, + "CGTTTG": 2908, + "CGTTCA": 2909, + "CGTTCT": 2910, + "CGTTCC": 2911, + "CGTTCG": 2912, + "CGTTGA": 2913, + "CGTTGT": 2914, + "CGTTGC": 2915, + "CGTTGG": 2916, + "CGTCAA": 2917, + "CGTCAT": 2918, + "CGTCAC": 2919, + "CGTCAG": 2920, + "CGTCTA": 2921, + "CGTCTT": 2922, + "CGTCTC": 2923, + "CGTCTG": 2924, + "CGTCCA": 2925, + "CGTCCT": 2926, + "CGTCCC": 2927, + "CGTCCG": 2928, + "CGTCGA": 2929, + "CGTCGT": 2930, + "CGTCGC": 2931, + "CGTCGG": 2932, + "CGTGAA": 2933, + "CGTGAT": 2934, + "CGTGAC": 2935, + "CGTGAG": 2936, + "CGTGTA": 2937, + "CGTGTT": 2938, + "CGTGTC": 2939, + "CGTGTG": 2940, + "CGTGCA": 2941, + "CGTGCT": 2942, + "CGTGCC": 2943, + "CGTGCG": 2944, + "CGTGGA": 2945, + "CGTGGT": 2946, + "CGTGGC": 2947, + "CGTGGG": 2948, + "CGCAAA": 2949, + "CGCAAT": 2950, + "CGCAAC": 2951, + "CGCAAG": 2952, + "CGCATA": 2953, + "CGCATT": 2954, + "CGCATC": 2955, + "CGCATG": 2956, + "CGCACA": 2957, + "CGCACT": 2958, + "CGCACC": 2959, + "CGCACG": 2960, + "CGCAGA": 2961, + "CGCAGT": 2962, + "CGCAGC": 2963, + "CGCAGG": 2964, + "CGCTAA": 2965, + "CGCTAT": 2966, + "CGCTAC": 2967, + "CGCTAG": 2968, + "CGCTTA": 2969, + "CGCTTT": 2970, + "CGCTTC": 2971, + "CGCTTG": 2972, + "CGCTCA": 2973, + "CGCTCT": 2974, + "CGCTCC": 2975, + "CGCTCG": 2976, + "CGCTGA": 2977, + "CGCTGT": 2978, + "CGCTGC": 2979, + "CGCTGG": 2980, + "CGCCAA": 2981, + "CGCCAT": 2982, + "CGCCAC": 2983, + "CGCCAG": 2984, + "CGCCTA": 2985, + "CGCCTT": 2986, + "CGCCTC": 2987, + "CGCCTG": 2988, + "CGCCCA": 2989, + "CGCCCT": 2990, + "CGCCCC": 2991, + "CGCCCG": 2992, + "CGCCGA": 2993, + "CGCCGT": 2994, + "CGCCGC": 2995, + "CGCCGG": 2996, + "CGCGAA": 2997, + "CGCGAT": 2998, + "CGCGAC": 2999, + "CGCGAG": 3000, + "CGCGTA": 3001, + "CGCGTT": 3002, + "CGCGTC": 3003, + "CGCGTG": 3004, + "CGCGCA": 3005, + "CGCGCT": 3006, + "CGCGCC": 3007, + "CGCGCG": 3008, + "CGCGGA": 3009, + "CGCGGT": 3010, + "CGCGGC": 3011, + "CGCGGG": 3012, + "CGGAAA": 3013, + "CGGAAT": 3014, + "CGGAAC": 3015, + "CGGAAG": 3016, + "CGGATA": 3017, + "CGGATT": 3018, + "CGGATC": 3019, + "CGGATG": 3020, + "CGGACA": 3021, + "CGGACT": 3022, + "CGGACC": 3023, + "CGGACG": 3024, + "CGGAGA": 3025, + "CGGAGT": 3026, + "CGGAGC": 3027, + "CGGAGG": 3028, + "CGGTAA": 3029, + "CGGTAT": 3030, + "CGGTAC": 3031, + "CGGTAG": 3032, + "CGGTTA": 3033, + "CGGTTT": 3034, + "CGGTTC": 3035, + "CGGTTG": 3036, + "CGGTCA": 3037, + "CGGTCT": 3038, + "CGGTCC": 3039, + "CGGTCG": 3040, + "CGGTGA": 3041, + "CGGTGT": 3042, + "CGGTGC": 3043, + "CGGTGG": 3044, + "CGGCAA": 3045, + "CGGCAT": 3046, + "CGGCAC": 3047, + "CGGCAG": 3048, + "CGGCTA": 3049, + "CGGCTT": 3050, + "CGGCTC": 3051, + "CGGCTG": 3052, + "CGGCCA": 3053, + "CGGCCT": 3054, + "CGGCCC": 3055, + "CGGCCG": 3056, + "CGGCGA": 3057, + "CGGCGT": 3058, + "CGGCGC": 3059, + "CGGCGG": 3060, + "CGGGAA": 3061, + "CGGGAT": 3062, + "CGGGAC": 3063, + "CGGGAG": 3064, + "CGGGTA": 3065, + "CGGGTT": 3066, + "CGGGTC": 3067, + "CGGGTG": 3068, + "CGGGCA": 3069, + "CGGGCT": 3070, + "CGGGCC": 3071, + "CGGGCG": 3072, + "CGGGGA": 3073, + "CGGGGT": 3074, + "CGGGGC": 3075, + "CGGGGG": 3076, + "GAAAAA": 3077, + "GAAAAT": 3078, + "GAAAAC": 3079, + "GAAAAG": 3080, + "GAAATA": 3081, + "GAAATT": 3082, + "GAAATC": 3083, + "GAAATG": 3084, + "GAAACA": 3085, + "GAAACT": 3086, + "GAAACC": 3087, + "GAAACG": 3088, + "GAAAGA": 3089, + "GAAAGT": 3090, + "GAAAGC": 3091, + "GAAAGG": 3092, + "GAATAA": 3093, + "GAATAT": 3094, + "GAATAC": 3095, + "GAATAG": 3096, + "GAATTA": 3097, + "GAATTT": 3098, + "GAATTC": 3099, + "GAATTG": 3100, + "GAATCA": 3101, + "GAATCT": 3102, + "GAATCC": 3103, + "GAATCG": 3104, + "GAATGA": 3105, + "GAATGT": 3106, + "GAATGC": 3107, + "GAATGG": 3108, + "GAACAA": 3109, + "GAACAT": 3110, + "GAACAC": 3111, + "GAACAG": 3112, + "GAACTA": 3113, + "GAACTT": 3114, + "GAACTC": 3115, + "GAACTG": 3116, + "GAACCA": 3117, + "GAACCT": 3118, + "GAACCC": 3119, + "GAACCG": 3120, + "GAACGA": 3121, + "GAACGT": 3122, + "GAACGC": 3123, + "GAACGG": 3124, + "GAAGAA": 3125, + "GAAGAT": 3126, + "GAAGAC": 3127, + "GAAGAG": 3128, + "GAAGTA": 3129, + "GAAGTT": 3130, + "GAAGTC": 3131, + "GAAGTG": 3132, + "GAAGCA": 3133, + "GAAGCT": 3134, + "GAAGCC": 3135, + "GAAGCG": 3136, + "GAAGGA": 3137, + "GAAGGT": 3138, + "GAAGGC": 3139, + "GAAGGG": 3140, + "GATAAA": 3141, + "GATAAT": 3142, + "GATAAC": 3143, + "GATAAG": 3144, + "GATATA": 3145, + "GATATT": 3146, + "GATATC": 3147, + "GATATG": 3148, + "GATACA": 3149, + "GATACT": 3150, + "GATACC": 3151, + "GATACG": 3152, + "GATAGA": 3153, + "GATAGT": 3154, + "GATAGC": 3155, + "GATAGG": 3156, + "GATTAA": 3157, + "GATTAT": 3158, + "GATTAC": 3159, + "GATTAG": 3160, + "GATTTA": 3161, + "GATTTT": 3162, + "GATTTC": 3163, + "GATTTG": 3164, + "GATTCA": 3165, + "GATTCT": 3166, + "GATTCC": 3167, + "GATTCG": 3168, + "GATTGA": 3169, + "GATTGT": 3170, + "GATTGC": 3171, + "GATTGG": 3172, + "GATCAA": 3173, + "GATCAT": 3174, + "GATCAC": 3175, + "GATCAG": 3176, + "GATCTA": 3177, + "GATCTT": 3178, + "GATCTC": 3179, + "GATCTG": 3180, + "GATCCA": 3181, + "GATCCT": 3182, + "GATCCC": 3183, + "GATCCG": 3184, + "GATCGA": 3185, + "GATCGT": 3186, + "GATCGC": 3187, + "GATCGG": 3188, + "GATGAA": 3189, + "GATGAT": 3190, + "GATGAC": 3191, + "GATGAG": 3192, + "GATGTA": 3193, + "GATGTT": 3194, + "GATGTC": 3195, + "GATGTG": 3196, + "GATGCA": 3197, + "GATGCT": 3198, + "GATGCC": 3199, + "GATGCG": 3200, + "GATGGA": 3201, + "GATGGT": 3202, + "GATGGC": 3203, + "GATGGG": 3204, + "GACAAA": 3205, + "GACAAT": 3206, + "GACAAC": 3207, + "GACAAG": 3208, + "GACATA": 3209, + "GACATT": 3210, + "GACATC": 3211, + "GACATG": 3212, + "GACACA": 3213, + "GACACT": 3214, + "GACACC": 3215, + "GACACG": 3216, + "GACAGA": 3217, + "GACAGT": 3218, + "GACAGC": 3219, + "GACAGG": 3220, + "GACTAA": 3221, + "GACTAT": 3222, + "GACTAC": 3223, + "GACTAG": 3224, + "GACTTA": 3225, + "GACTTT": 3226, + "GACTTC": 3227, + "GACTTG": 3228, + "GACTCA": 3229, + "GACTCT": 3230, + "GACTCC": 3231, + "GACTCG": 3232, + "GACTGA": 3233, + "GACTGT": 3234, + "GACTGC": 3235, + "GACTGG": 3236, + "GACCAA": 3237, + "GACCAT": 3238, + "GACCAC": 3239, + "GACCAG": 3240, + "GACCTA": 3241, + "GACCTT": 3242, + "GACCTC": 3243, + "GACCTG": 3244, + "GACCCA": 3245, + "GACCCT": 3246, + "GACCCC": 3247, + "GACCCG": 3248, + "GACCGA": 3249, + "GACCGT": 3250, + "GACCGC": 3251, + "GACCGG": 3252, + "GACGAA": 3253, + "GACGAT": 3254, + "GACGAC": 3255, + "GACGAG": 3256, + "GACGTA": 3257, + "GACGTT": 3258, + "GACGTC": 3259, + "GACGTG": 3260, + "GACGCA": 3261, + "GACGCT": 3262, + "GACGCC": 3263, + "GACGCG": 3264, + "GACGGA": 3265, + "GACGGT": 3266, + "GACGGC": 3267, + "GACGGG": 3268, + "GAGAAA": 3269, + "GAGAAT": 3270, + "GAGAAC": 3271, + "GAGAAG": 3272, + "GAGATA": 3273, + "GAGATT": 3274, + "GAGATC": 3275, + "GAGATG": 3276, + "GAGACA": 3277, + "GAGACT": 3278, + "GAGACC": 3279, + "GAGACG": 3280, + "GAGAGA": 3281, + "GAGAGT": 3282, + "GAGAGC": 3283, + "GAGAGG": 3284, + "GAGTAA": 3285, + "GAGTAT": 3286, + "GAGTAC": 3287, + "GAGTAG": 3288, + "GAGTTA": 3289, + "GAGTTT": 3290, + "GAGTTC": 3291, + "GAGTTG": 3292, + "GAGTCA": 3293, + "GAGTCT": 3294, + "GAGTCC": 3295, + "GAGTCG": 3296, + "GAGTGA": 3297, + "GAGTGT": 3298, + "GAGTGC": 3299, + "GAGTGG": 3300, + "GAGCAA": 3301, + "GAGCAT": 3302, + "GAGCAC": 3303, + "GAGCAG": 3304, + "GAGCTA": 3305, + "GAGCTT": 3306, + "GAGCTC": 3307, + "GAGCTG": 3308, + "GAGCCA": 3309, + "GAGCCT": 3310, + "GAGCCC": 3311, + "GAGCCG": 3312, + "GAGCGA": 3313, + "GAGCGT": 3314, + "GAGCGC": 3315, + "GAGCGG": 3316, + "GAGGAA": 3317, + "GAGGAT": 3318, + "GAGGAC": 3319, + "GAGGAG": 3320, + "GAGGTA": 3321, + "GAGGTT": 3322, + "GAGGTC": 3323, + "GAGGTG": 3324, + "GAGGCA": 3325, + "GAGGCT": 3326, + "GAGGCC": 3327, + "GAGGCG": 3328, + "GAGGGA": 3329, + "GAGGGT": 3330, + "GAGGGC": 3331, + "GAGGGG": 3332, + "GTAAAA": 3333, + "GTAAAT": 3334, + "GTAAAC": 3335, + "GTAAAG": 3336, + "GTAATA": 3337, + "GTAATT": 3338, + "GTAATC": 3339, + "GTAATG": 3340, + "GTAACA": 3341, + "GTAACT": 3342, + "GTAACC": 3343, + "GTAACG": 3344, + "GTAAGA": 3345, + "GTAAGT": 3346, + "GTAAGC": 3347, + "GTAAGG": 3348, + "GTATAA": 3349, + "GTATAT": 3350, + "GTATAC": 3351, + "GTATAG": 3352, + "GTATTA": 3353, + "GTATTT": 3354, + "GTATTC": 3355, + "GTATTG": 3356, + "GTATCA": 3357, + "GTATCT": 3358, + "GTATCC": 3359, + "GTATCG": 3360, + "GTATGA": 3361, + "GTATGT": 3362, + "GTATGC": 3363, + "GTATGG": 3364, + "GTACAA": 3365, + "GTACAT": 3366, + "GTACAC": 3367, + "GTACAG": 3368, + "GTACTA": 3369, + "GTACTT": 3370, + "GTACTC": 3371, + "GTACTG": 3372, + "GTACCA": 3373, + "GTACCT": 3374, + "GTACCC": 3375, + "GTACCG": 3376, + "GTACGA": 3377, + "GTACGT": 3378, + "GTACGC": 3379, + "GTACGG": 3380, + "GTAGAA": 3381, + "GTAGAT": 3382, + "GTAGAC": 3383, + "GTAGAG": 3384, + "GTAGTA": 3385, + "GTAGTT": 3386, + "GTAGTC": 3387, + "GTAGTG": 3388, + "GTAGCA": 3389, + "GTAGCT": 3390, + "GTAGCC": 3391, + "GTAGCG": 3392, + "GTAGGA": 3393, + "GTAGGT": 3394, + "GTAGGC": 3395, + "GTAGGG": 3396, + "GTTAAA": 3397, + "GTTAAT": 3398, + "GTTAAC": 3399, + "GTTAAG": 3400, + "GTTATA": 3401, + "GTTATT": 3402, + "GTTATC": 3403, + "GTTATG": 3404, + "GTTACA": 3405, + "GTTACT": 3406, + "GTTACC": 3407, + "GTTACG": 3408, + "GTTAGA": 3409, + "GTTAGT": 3410, + "GTTAGC": 3411, + "GTTAGG": 3412, + "GTTTAA": 3413, + "GTTTAT": 3414, + "GTTTAC": 3415, + "GTTTAG": 3416, + "GTTTTA": 3417, + "GTTTTT": 3418, + "GTTTTC": 3419, + "GTTTTG": 3420, + "GTTTCA": 3421, + "GTTTCT": 3422, + "GTTTCC": 3423, + "GTTTCG": 3424, + "GTTTGA": 3425, + "GTTTGT": 3426, + "GTTTGC": 3427, + "GTTTGG": 3428, + "GTTCAA": 3429, + "GTTCAT": 3430, + "GTTCAC": 3431, + "GTTCAG": 3432, + "GTTCTA": 3433, + "GTTCTT": 3434, + "GTTCTC": 3435, + "GTTCTG": 3436, + "GTTCCA": 3437, + "GTTCCT": 3438, + "GTTCCC": 3439, + "GTTCCG": 3440, + "GTTCGA": 3441, + "GTTCGT": 3442, + "GTTCGC": 3443, + "GTTCGG": 3444, + "GTTGAA": 3445, + "GTTGAT": 3446, + "GTTGAC": 3447, + "GTTGAG": 3448, + "GTTGTA": 3449, + "GTTGTT": 3450, + "GTTGTC": 3451, + "GTTGTG": 3452, + "GTTGCA": 3453, + "GTTGCT": 3454, + "GTTGCC": 3455, + "GTTGCG": 3456, + "GTTGGA": 3457, + "GTTGGT": 3458, + "GTTGGC": 3459, + "GTTGGG": 3460, + "GTCAAA": 3461, + "GTCAAT": 3462, + "GTCAAC": 3463, + "GTCAAG": 3464, + "GTCATA": 3465, + "GTCATT": 3466, + "GTCATC": 3467, + "GTCATG": 3468, + "GTCACA": 3469, + "GTCACT": 3470, + "GTCACC": 3471, + "GTCACG": 3472, + "GTCAGA": 3473, + "GTCAGT": 3474, + "GTCAGC": 3475, + "GTCAGG": 3476, + "GTCTAA": 3477, + "GTCTAT": 3478, + "GTCTAC": 3479, + "GTCTAG": 3480, + "GTCTTA": 3481, + "GTCTTT": 3482, + "GTCTTC": 3483, + "GTCTTG": 3484, + "GTCTCA": 3485, + "GTCTCT": 3486, + "GTCTCC": 3487, + "GTCTCG": 3488, + "GTCTGA": 3489, + "GTCTGT": 3490, + "GTCTGC": 3491, + "GTCTGG": 3492, + "GTCCAA": 3493, + "GTCCAT": 3494, + "GTCCAC": 3495, + "GTCCAG": 3496, + "GTCCTA": 3497, + "GTCCTT": 3498, + "GTCCTC": 3499, + "GTCCTG": 3500, + "GTCCCA": 3501, + "GTCCCT": 3502, + "GTCCCC": 3503, + "GTCCCG": 3504, + "GTCCGA": 3505, + "GTCCGT": 3506, + "GTCCGC": 3507, + "GTCCGG": 3508, + "GTCGAA": 3509, + "GTCGAT": 3510, + "GTCGAC": 3511, + "GTCGAG": 3512, + "GTCGTA": 3513, + "GTCGTT": 3514, + "GTCGTC": 3515, + "GTCGTG": 3516, + "GTCGCA": 3517, + "GTCGCT": 3518, + "GTCGCC": 3519, + "GTCGCG": 3520, + "GTCGGA": 3521, + "GTCGGT": 3522, + "GTCGGC": 3523, + "GTCGGG": 3524, + "GTGAAA": 3525, + "GTGAAT": 3526, + "GTGAAC": 3527, + "GTGAAG": 3528, + "GTGATA": 3529, + "GTGATT": 3530, + "GTGATC": 3531, + "GTGATG": 3532, + "GTGACA": 3533, + "GTGACT": 3534, + "GTGACC": 3535, + "GTGACG": 3536, + "GTGAGA": 3537, + "GTGAGT": 3538, + "GTGAGC": 3539, + "GTGAGG": 3540, + "GTGTAA": 3541, + "GTGTAT": 3542, + "GTGTAC": 3543, + "GTGTAG": 3544, + "GTGTTA": 3545, + "GTGTTT": 3546, + "GTGTTC": 3547, + "GTGTTG": 3548, + "GTGTCA": 3549, + "GTGTCT": 3550, + "GTGTCC": 3551, + "GTGTCG": 3552, + "GTGTGA": 3553, + "GTGTGT": 3554, + "GTGTGC": 3555, + "GTGTGG": 3556, + "GTGCAA": 3557, + "GTGCAT": 3558, + "GTGCAC": 3559, + "GTGCAG": 3560, + "GTGCTA": 3561, + "GTGCTT": 3562, + "GTGCTC": 3563, + "GTGCTG": 3564, + "GTGCCA": 3565, + "GTGCCT": 3566, + "GTGCCC": 3567, + "GTGCCG": 3568, + "GTGCGA": 3569, + "GTGCGT": 3570, + "GTGCGC": 3571, + "GTGCGG": 3572, + "GTGGAA": 3573, + "GTGGAT": 3574, + "GTGGAC": 3575, + "GTGGAG": 3576, + "GTGGTA": 3577, + "GTGGTT": 3578, + "GTGGTC": 3579, + "GTGGTG": 3580, + "GTGGCA": 3581, + "GTGGCT": 3582, + "GTGGCC": 3583, + "GTGGCG": 3584, + "GTGGGA": 3585, + "GTGGGT": 3586, + "GTGGGC": 3587, + "GTGGGG": 3588, + "GCAAAA": 3589, + "GCAAAT": 3590, + "GCAAAC": 3591, + "GCAAAG": 3592, + "GCAATA": 3593, + "GCAATT": 3594, + "GCAATC": 3595, + "GCAATG": 3596, + "GCAACA": 3597, + "GCAACT": 3598, + "GCAACC": 3599, + "GCAACG": 3600, + "GCAAGA": 3601, + "GCAAGT": 3602, + "GCAAGC": 3603, + "GCAAGG": 3604, + "GCATAA": 3605, + "GCATAT": 3606, + "GCATAC": 3607, + "GCATAG": 3608, + "GCATTA": 3609, + "GCATTT": 3610, + "GCATTC": 3611, + "GCATTG": 3612, + "GCATCA": 3613, + "GCATCT": 3614, + "GCATCC": 3615, + "GCATCG": 3616, + "GCATGA": 3617, + "GCATGT": 3618, + "GCATGC": 3619, + "GCATGG": 3620, + "GCACAA": 3621, + "GCACAT": 3622, + "GCACAC": 3623, + "GCACAG": 3624, + "GCACTA": 3625, + "GCACTT": 3626, + "GCACTC": 3627, + "GCACTG": 3628, + "GCACCA": 3629, + "GCACCT": 3630, + "GCACCC": 3631, + "GCACCG": 3632, + "GCACGA": 3633, + "GCACGT": 3634, + "GCACGC": 3635, + "GCACGG": 3636, + "GCAGAA": 3637, + "GCAGAT": 3638, + "GCAGAC": 3639, + "GCAGAG": 3640, + "GCAGTA": 3641, + "GCAGTT": 3642, + "GCAGTC": 3643, + "GCAGTG": 3644, + "GCAGCA": 3645, + "GCAGCT": 3646, + "GCAGCC": 3647, + "GCAGCG": 3648, + "GCAGGA": 3649, + "GCAGGT": 3650, + "GCAGGC": 3651, + "GCAGGG": 3652, + "GCTAAA": 3653, + "GCTAAT": 3654, + "GCTAAC": 3655, + "GCTAAG": 3656, + "GCTATA": 3657, + "GCTATT": 3658, + "GCTATC": 3659, + "GCTATG": 3660, + "GCTACA": 3661, + "GCTACT": 3662, + "GCTACC": 3663, + "GCTACG": 3664, + "GCTAGA": 3665, + "GCTAGT": 3666, + "GCTAGC": 3667, + "GCTAGG": 3668, + "GCTTAA": 3669, + "GCTTAT": 3670, + "GCTTAC": 3671, + "GCTTAG": 3672, + "GCTTTA": 3673, + "GCTTTT": 3674, + "GCTTTC": 3675, + "GCTTTG": 3676, + "GCTTCA": 3677, + "GCTTCT": 3678, + "GCTTCC": 3679, + "GCTTCG": 3680, + "GCTTGA": 3681, + "GCTTGT": 3682, + "GCTTGC": 3683, + "GCTTGG": 3684, + "GCTCAA": 3685, + "GCTCAT": 3686, + "GCTCAC": 3687, + "GCTCAG": 3688, + "GCTCTA": 3689, + "GCTCTT": 3690, + "GCTCTC": 3691, + "GCTCTG": 3692, + "GCTCCA": 3693, + "GCTCCT": 3694, + "GCTCCC": 3695, + "GCTCCG": 3696, + "GCTCGA": 3697, + "GCTCGT": 3698, + "GCTCGC": 3699, + "GCTCGG": 3700, + "GCTGAA": 3701, + "GCTGAT": 3702, + "GCTGAC": 3703, + "GCTGAG": 3704, + "GCTGTA": 3705, + "GCTGTT": 3706, + "GCTGTC": 3707, + "GCTGTG": 3708, + "GCTGCA": 3709, + "GCTGCT": 3710, + "GCTGCC": 3711, + "GCTGCG": 3712, + "GCTGGA": 3713, + "GCTGGT": 3714, + "GCTGGC": 3715, + "GCTGGG": 3716, + "GCCAAA": 3717, + "GCCAAT": 3718, + "GCCAAC": 3719, + "GCCAAG": 3720, + "GCCATA": 3721, + "GCCATT": 3722, + "GCCATC": 3723, + "GCCATG": 3724, + "GCCACA": 3725, + "GCCACT": 3726, + "GCCACC": 3727, + "GCCACG": 3728, + "GCCAGA": 3729, + "GCCAGT": 3730, + "GCCAGC": 3731, + "GCCAGG": 3732, + "GCCTAA": 3733, + "GCCTAT": 3734, + "GCCTAC": 3735, + "GCCTAG": 3736, + "GCCTTA": 3737, + "GCCTTT": 3738, + "GCCTTC": 3739, + "GCCTTG": 3740, + "GCCTCA": 3741, + "GCCTCT": 3742, + "GCCTCC": 3743, + "GCCTCG": 3744, + "GCCTGA": 3745, + "GCCTGT": 3746, + "GCCTGC": 3747, + "GCCTGG": 3748, + "GCCCAA": 3749, + "GCCCAT": 3750, + "GCCCAC": 3751, + "GCCCAG": 3752, + "GCCCTA": 3753, + "GCCCTT": 3754, + "GCCCTC": 3755, + "GCCCTG": 3756, + "GCCCCA": 3757, + "GCCCCT": 3758, + "GCCCCC": 3759, + "GCCCCG": 3760, + "GCCCGA": 3761, + "GCCCGT": 3762, + "GCCCGC": 3763, + "GCCCGG": 3764, + "GCCGAA": 3765, + "GCCGAT": 3766, + "GCCGAC": 3767, + "GCCGAG": 3768, + "GCCGTA": 3769, + "GCCGTT": 3770, + "GCCGTC": 3771, + "GCCGTG": 3772, + "GCCGCA": 3773, + "GCCGCT": 3774, + "GCCGCC": 3775, + "GCCGCG": 3776, + "GCCGGA": 3777, + "GCCGGT": 3778, + "GCCGGC": 3779, + "GCCGGG": 3780, + "GCGAAA": 3781, + "GCGAAT": 3782, + "GCGAAC": 3783, + "GCGAAG": 3784, + "GCGATA": 3785, + "GCGATT": 3786, + "GCGATC": 3787, + "GCGATG": 3788, + "GCGACA": 3789, + "GCGACT": 3790, + "GCGACC": 3791, + "GCGACG": 3792, + "GCGAGA": 3793, + "GCGAGT": 3794, + "GCGAGC": 3795, + "GCGAGG": 3796, + "GCGTAA": 3797, + "GCGTAT": 3798, + "GCGTAC": 3799, + "GCGTAG": 3800, + "GCGTTA": 3801, + "GCGTTT": 3802, + "GCGTTC": 3803, + "GCGTTG": 3804, + "GCGTCA": 3805, + "GCGTCT": 3806, + "GCGTCC": 3807, + "GCGTCG": 3808, + "GCGTGA": 3809, + "GCGTGT": 3810, + "GCGTGC": 3811, + "GCGTGG": 3812, + "GCGCAA": 3813, + "GCGCAT": 3814, + "GCGCAC": 3815, + "GCGCAG": 3816, + "GCGCTA": 3817, + "GCGCTT": 3818, + "GCGCTC": 3819, + "GCGCTG": 3820, + "GCGCCA": 3821, + "GCGCCT": 3822, + "GCGCCC": 3823, + "GCGCCG": 3824, + "GCGCGA": 3825, + "GCGCGT": 3826, + "GCGCGC": 3827, + "GCGCGG": 3828, + "GCGGAA": 3829, + "GCGGAT": 3830, + "GCGGAC": 3831, + "GCGGAG": 3832, + "GCGGTA": 3833, + "GCGGTT": 3834, + "GCGGTC": 3835, + "GCGGTG": 3836, + "GCGGCA": 3837, + "GCGGCT": 3838, + "GCGGCC": 3839, + "GCGGCG": 3840, + "GCGGGA": 3841, + "GCGGGT": 3842, + "GCGGGC": 3843, + "GCGGGG": 3844, + "GGAAAA": 3845, + "GGAAAT": 3846, + "GGAAAC": 3847, + "GGAAAG": 3848, + "GGAATA": 3849, + "GGAATT": 3850, + "GGAATC": 3851, + "GGAATG": 3852, + "GGAACA": 3853, + "GGAACT": 3854, + "GGAACC": 3855, + "GGAACG": 3856, + "GGAAGA": 3857, + "GGAAGT": 3858, + "GGAAGC": 3859, + "GGAAGG": 3860, + "GGATAA": 3861, + "GGATAT": 3862, + "GGATAC": 3863, + "GGATAG": 3864, + "GGATTA": 3865, + "GGATTT": 3866, + "GGATTC": 3867, + "GGATTG": 3868, + "GGATCA": 3869, + "GGATCT": 3870, + "GGATCC": 3871, + "GGATCG": 3872, + "GGATGA": 3873, + "GGATGT": 3874, + "GGATGC": 3875, + "GGATGG": 3876, + "GGACAA": 3877, + "GGACAT": 3878, + "GGACAC": 3879, + "GGACAG": 3880, + "GGACTA": 3881, + "GGACTT": 3882, + "GGACTC": 3883, + "GGACTG": 3884, + "GGACCA": 3885, + "GGACCT": 3886, + "GGACCC": 3887, + "GGACCG": 3888, + "GGACGA": 3889, + "GGACGT": 3890, + "GGACGC": 3891, + "GGACGG": 3892, + "GGAGAA": 3893, + "GGAGAT": 3894, + "GGAGAC": 3895, + "GGAGAG": 3896, + "GGAGTA": 3897, + "GGAGTT": 3898, + "GGAGTC": 3899, + "GGAGTG": 3900, + "GGAGCA": 3901, + "GGAGCT": 3902, + "GGAGCC": 3903, + "GGAGCG": 3904, + "GGAGGA": 3905, + "GGAGGT": 3906, + "GGAGGC": 3907, + "GGAGGG": 3908, + "GGTAAA": 3909, + "GGTAAT": 3910, + "GGTAAC": 3911, + "GGTAAG": 3912, + "GGTATA": 3913, + "GGTATT": 3914, + "GGTATC": 3915, + "GGTATG": 3916, + "GGTACA": 3917, + "GGTACT": 3918, + "GGTACC": 3919, + "GGTACG": 3920, + "GGTAGA": 3921, + "GGTAGT": 3922, + "GGTAGC": 3923, + "GGTAGG": 3924, + "GGTTAA": 3925, + "GGTTAT": 3926, + "GGTTAC": 3927, + "GGTTAG": 3928, + "GGTTTA": 3929, + "GGTTTT": 3930, + "GGTTTC": 3931, + "GGTTTG": 3932, + "GGTTCA": 3933, + "GGTTCT": 3934, + "GGTTCC": 3935, + "GGTTCG": 3936, + "GGTTGA": 3937, + "GGTTGT": 3938, + "GGTTGC": 3939, + "GGTTGG": 3940, + "GGTCAA": 3941, + "GGTCAT": 3942, + "GGTCAC": 3943, + "GGTCAG": 3944, + "GGTCTA": 3945, + "GGTCTT": 3946, + "GGTCTC": 3947, + "GGTCTG": 3948, + "GGTCCA": 3949, + "GGTCCT": 3950, + "GGTCCC": 3951, + "GGTCCG": 3952, + "GGTCGA": 3953, + "GGTCGT": 3954, + "GGTCGC": 3955, + "GGTCGG": 3956, + "GGTGAA": 3957, + "GGTGAT": 3958, + "GGTGAC": 3959, + "GGTGAG": 3960, + "GGTGTA": 3961, + "GGTGTT": 3962, + "GGTGTC": 3963, + "GGTGTG": 3964, + "GGTGCA": 3965, + "GGTGCT": 3966, + "GGTGCC": 3967, + "GGTGCG": 3968, + "GGTGGA": 3969, + "GGTGGT": 3970, + "GGTGGC": 3971, + "GGTGGG": 3972, + "GGCAAA": 3973, + "GGCAAT": 3974, + "GGCAAC": 3975, + "GGCAAG": 3976, + "GGCATA": 3977, + "GGCATT": 3978, + "GGCATC": 3979, + "GGCATG": 3980, + "GGCACA": 3981, + "GGCACT": 3982, + "GGCACC": 3983, + "GGCACG": 3984, + "GGCAGA": 3985, + "GGCAGT": 3986, + "GGCAGC": 3987, + "GGCAGG": 3988, + "GGCTAA": 3989, + "GGCTAT": 3990, + "GGCTAC": 3991, + "GGCTAG": 3992, + "GGCTTA": 3993, + "GGCTTT": 3994, + "GGCTTC": 3995, + "GGCTTG": 3996, + "GGCTCA": 3997, + "GGCTCT": 3998, + "GGCTCC": 3999, + "GGCTCG": 4000, + "GGCTGA": 4001, + "GGCTGT": 4002, + "GGCTGC": 4003, + "GGCTGG": 4004, + "GGCCAA": 4005, + "GGCCAT": 4006, + "GGCCAC": 4007, + "GGCCAG": 4008, + "GGCCTA": 4009, + "GGCCTT": 4010, + "GGCCTC": 4011, + "GGCCTG": 4012, + "GGCCCA": 4013, + "GGCCCT": 4014, + "GGCCCC": 4015, + "GGCCCG": 4016, + "GGCCGA": 4017, + "GGCCGT": 4018, + "GGCCGC": 4019, + "GGCCGG": 4020, + "GGCGAA": 4021, + "GGCGAT": 4022, + "GGCGAC": 4023, + "GGCGAG": 4024, + "GGCGTA": 4025, + "GGCGTT": 4026, + "GGCGTC": 4027, + "GGCGTG": 4028, + "GGCGCA": 4029, + "GGCGCT": 4030, + "GGCGCC": 4031, + "GGCGCG": 4032, + "GGCGGA": 4033, + "GGCGGT": 4034, + "GGCGGC": 4035, + "GGCGGG": 4036, + "GGGAAA": 4037, + "GGGAAT": 4038, + "GGGAAC": 4039, + "GGGAAG": 4040, + "GGGATA": 4041, + "GGGATT": 4042, + "GGGATC": 4043, + "GGGATG": 4044, + "GGGACA": 4045, + "GGGACT": 4046, + "GGGACC": 4047, + "GGGACG": 4048, + "GGGAGA": 4049, + "GGGAGT": 4050, + "GGGAGC": 4051, + "GGGAGG": 4052, + "GGGTAA": 4053, + "GGGTAT": 4054, + "GGGTAC": 4055, + "GGGTAG": 4056, + "GGGTTA": 4057, + "GGGTTT": 4058, + "GGGTTC": 4059, + "GGGTTG": 4060, + "GGGTCA": 4061, + "GGGTCT": 4062, + "GGGTCC": 4063, + "GGGTCG": 4064, + "GGGTGA": 4065, + "GGGTGT": 4066, + "GGGTGC": 4067, + "GGGTGG": 4068, + "GGGCAA": 4069, + "GGGCAT": 4070, + "GGGCAC": 4071, + "GGGCAG": 4072, + "GGGCTA": 4073, + "GGGCTT": 4074, + "GGGCTC": 4075, + "GGGCTG": 4076, + "GGGCCA": 4077, + "GGGCCT": 4078, + "GGGCCC": 4079, + "GGGCCG": 4080, + "GGGCGA": 4081, + "GGGCGT": 4082, + "GGGCGC": 4083, + "GGGCGG": 4084, + "GGGGAA": 4085, + "GGGGAT": 4086, + "GGGGAC": 4087, + "GGGGAG": 4088, + "GGGGTA": 4089, + "GGGGTT": 4090, + "GGGGTC": 4091, + "GGGGTG": 4092, + "GGGGCA": 4093, + "GGGGCT": 4094, + "GGGGCC": 4095, + "GGGGCG": 4096, + "GGGGGA": 4097, + "GGGGGT": 4098, + "GGGGGC": 4099, + "GGGGGG": 4100 + } + } +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-4000/tokenizer_config.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-4000/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..604e2b1b29af1dcb3920d2828b6a90276e5e3d02 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-4000/tokenizer_config.json @@ -0,0 +1,58 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "[UNK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "[CLS]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "[SEP]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "4": { + "content": "[MASK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "clean_up_tokenization_spaces": true, + "cls_token": "[CLS]", + "do_basic_tokenize": true, + "do_lower_case": false, + "mask_token": "[MASK]", + "max_len": 512, + "model_max_length": 512, + "never_split": null, + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "strip_accents": null, + "tokenize_chinese_chars": true, + "tokenizer_class": "BertTokenizer", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-4000/trainer_state.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-4000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..4b246e92f2e7e541c6ccde10e67a13b9f1134022 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-4000/trainer_state.json @@ -0,0 +1,1153 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 16.0, + "eval_steps": 500, + "global_step": 4000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.1, + "grad_norm": 0.6038870215415955, + "learning_rate": 5.319148936170213e-05, + "loss": 0.6945, + "step": 25 + }, + { + "epoch": 0.2, + "grad_norm": 0.6668091416358948, + "learning_rate": 0.00010638297872340425, + "loss": 0.6929, + "step": 50 + }, + { + "epoch": 0.3, + "grad_norm": 0.8282511234283447, + "learning_rate": 0.00015957446808510637, + "loss": 0.6925, + "step": 75 + }, + { + "epoch": 0.4, + "grad_norm": 1.3149235248565674, + "learning_rate": 0.0002127659574468085, + "loss": 0.6919, + "step": 100 + }, + { + "epoch": 0.5, + "grad_norm": 1.8841172456741333, + "learning_rate": 0.00026595744680851064, + "loss": 0.6788, + "step": 125 + }, + { + "epoch": 0.6, + "grad_norm": 4.435206413269043, + "learning_rate": 0.00031914893617021275, + "loss": 0.6285, + "step": 150 + }, + { + "epoch": 0.7, + "grad_norm": 8.858148574829102, + "learning_rate": 0.0003723404255319149, + "loss": 0.5386, + "step": 175 + }, + { + "epoch": 0.8, + "grad_norm": 5.627648830413818, + "learning_rate": 0.0003992081821181128, + "loss": 0.5537, + "step": 200 + }, + { + "epoch": 0.9, + "grad_norm": 5.086580753326416, + "learning_rate": 0.0003975585615308479, + "loss": 0.5595, + "step": 225 + }, + { + "epoch": 1.0, + "grad_norm": 5.627719402313232, + "learning_rate": 0.00039590894094358297, + "loss": 0.4824, + "step": 250 + }, + { + "epoch": 1.1, + "grad_norm": 5.888038158416748, + "learning_rate": 0.0003942593203563181, + "loss": 0.4404, + "step": 275 + }, + { + "epoch": 1.2, + "grad_norm": 5.646231651306152, + "learning_rate": 0.0003926096997690532, + "loss": 0.4354, + "step": 300 + }, + { + "epoch": 1.3, + "grad_norm": 5.191990852355957, + "learning_rate": 0.00039096007918178817, + "loss": 0.4469, + "step": 325 + }, + { + "epoch": 1.4, + "grad_norm": 13.90101432800293, + "learning_rate": 0.0003893104585945233, + "loss": 0.3858, + "step": 350 + }, + { + "epoch": 1.5, + "grad_norm": 17.75942039489746, + "learning_rate": 0.0003876608380072583, + "loss": 0.4131, + "step": 375 + }, + { + "epoch": 1.6, + "grad_norm": 7.398252487182617, + "learning_rate": 0.00038601121741999343, + "loss": 0.3842, + "step": 400 + }, + { + "epoch": 1.7, + "grad_norm": 11.010066986083984, + "learning_rate": 0.0003843615968327285, + "loss": 0.3959, + "step": 425 + }, + { + "epoch": 1.8, + "grad_norm": 9.277341842651367, + "learning_rate": 0.0003827119762454636, + "loss": 0.3552, + "step": 450 + }, + { + "epoch": 1.9, + "grad_norm": 26.011449813842773, + "learning_rate": 0.00038106235565819863, + "loss": 0.3961, + "step": 475 + }, + { + "epoch": 2.0, + "grad_norm": 11.123054504394531, + "learning_rate": 0.0003794127350709337, + "loss": 0.3548, + "step": 500 + }, + { + "epoch": 2.1, + "grad_norm": 20.128774642944336, + "learning_rate": 0.0003777631144836688, + "loss": 0.302, + "step": 525 + }, + { + "epoch": 2.2, + "grad_norm": 4.64390230178833, + "learning_rate": 0.00037611349389640383, + "loss": 0.2853, + "step": 550 + }, + { + "epoch": 2.3, + "grad_norm": 4.199934482574463, + "learning_rate": 0.00037446387330913894, + "loss": 0.2724, + "step": 575 + }, + { + "epoch": 2.4, + "grad_norm": 6.909411430358887, + "learning_rate": 0.000372814252721874, + "loss": 0.2976, + "step": 600 + }, + { + "epoch": 2.5, + "grad_norm": 9.234820365905762, + "learning_rate": 0.00037116463213460903, + "loss": 0.2798, + "step": 625 + }, + { + "epoch": 2.6, + "grad_norm": 10.472943305969238, + "learning_rate": 0.00036951501154734414, + "loss": 0.3007, + "step": 650 + }, + { + "epoch": 2.7, + "grad_norm": 19.46732521057129, + "learning_rate": 0.0003678653909600792, + "loss": 0.2701, + "step": 675 + }, + { + "epoch": 2.8, + "grad_norm": 7.501255512237549, + "learning_rate": 0.0003662157703728143, + "loss": 0.2896, + "step": 700 + }, + { + "epoch": 2.9, + "grad_norm": 10.645585060119629, + "learning_rate": 0.00036456614978554934, + "loss": 0.2748, + "step": 725 + }, + { + "epoch": 3.0, + "grad_norm": 30.26616859436035, + "learning_rate": 0.00036291652919828444, + "loss": 0.268, + "step": 750 + }, + { + "epoch": 3.1, + "grad_norm": 14.13516616821289, + "learning_rate": 0.00036126690861101944, + "loss": 0.2159, + "step": 775 + }, + { + "epoch": 3.2, + "grad_norm": 10.091023445129395, + "learning_rate": 0.00035961728802375454, + "loss": 0.2071, + "step": 800 + }, + { + "epoch": 3.3, + "grad_norm": 22.201452255249023, + "learning_rate": 0.0003579676674364896, + "loss": 0.2146, + "step": 825 + }, + { + "epoch": 3.4, + "grad_norm": 10.044412612915039, + "learning_rate": 0.0003563180468492247, + "loss": 0.1827, + "step": 850 + }, + { + "epoch": 3.5, + "grad_norm": 21.04941177368164, + "learning_rate": 0.0003546684262619598, + "loss": 0.202, + "step": 875 + }, + { + "epoch": 3.6, + "grad_norm": 11.754891395568848, + "learning_rate": 0.00035301880567469485, + "loss": 0.2021, + "step": 900 + }, + { + "epoch": 3.7, + "grad_norm": 9.113747596740723, + "learning_rate": 0.0003513691850874299, + "loss": 0.2067, + "step": 925 + }, + { + "epoch": 3.8, + "grad_norm": 9.075196266174316, + "learning_rate": 0.00034971956450016495, + "loss": 0.3285, + "step": 950 + }, + { + "epoch": 3.9, + "grad_norm": 8.568796157836914, + "learning_rate": 0.00034806994391290005, + "loss": 0.2254, + "step": 975 + }, + { + "epoch": 4.0, + "grad_norm": 3.9587976932525635, + "learning_rate": 0.0003464203233256351, + "loss": 0.2245, + "step": 1000 + }, + { + "epoch": 4.1, + "grad_norm": 11.143651008605957, + "learning_rate": 0.0003447707027383702, + "loss": 0.1755, + "step": 1025 + }, + { + "epoch": 4.2, + "grad_norm": 12.53118896484375, + "learning_rate": 0.00034312108215110525, + "loss": 0.1808, + "step": 1050 + }, + { + "epoch": 4.3, + "grad_norm": 9.880820274353027, + "learning_rate": 0.0003414714615638403, + "loss": 0.1387, + "step": 1075 + }, + { + "epoch": 4.4, + "grad_norm": 5.729593753814697, + "learning_rate": 0.0003398218409765754, + "loss": 0.1958, + "step": 1100 + }, + { + "epoch": 4.5, + "grad_norm": 8.633047103881836, + "learning_rate": 0.00033817222038931045, + "loss": 0.1515, + "step": 1125 + }, + { + "epoch": 4.6, + "grad_norm": 19.237892150878906, + "learning_rate": 0.00033652259980204556, + "loss": 0.1452, + "step": 1150 + }, + { + "epoch": 4.7, + "grad_norm": 6.129818439483643, + "learning_rate": 0.0003348729792147806, + "loss": 0.1617, + "step": 1175 + }, + { + "epoch": 4.8, + "grad_norm": 5.073274612426758, + "learning_rate": 0.0003332233586275157, + "loss": 0.1967, + "step": 1200 + }, + { + "epoch": 4.9, + "grad_norm": 5.729922771453857, + "learning_rate": 0.00033157373804025076, + "loss": 0.1482, + "step": 1225 + }, + { + "epoch": 5.0, + "grad_norm": 12.314518928527832, + "learning_rate": 0.0003299241174529858, + "loss": 0.1588, + "step": 1250 + }, + { + "epoch": 5.1, + "grad_norm": 3.2636606693267822, + "learning_rate": 0.0003282744968657209, + "loss": 0.1588, + "step": 1275 + }, + { + "epoch": 5.2, + "grad_norm": 5.914486408233643, + "learning_rate": 0.00032662487627845596, + "loss": 0.1288, + "step": 1300 + }, + { + "epoch": 5.3, + "grad_norm": 13.849296569824219, + "learning_rate": 0.00032497525569119106, + "loss": 0.0961, + "step": 1325 + }, + { + "epoch": 5.4, + "grad_norm": 7.7317328453063965, + "learning_rate": 0.0003233256351039261, + "loss": 0.1148, + "step": 1350 + }, + { + "epoch": 5.5, + "grad_norm": 1.768608808517456, + "learning_rate": 0.00032167601451666116, + "loss": 0.121, + "step": 1375 + }, + { + "epoch": 5.6, + "grad_norm": 10.802785873413086, + "learning_rate": 0.00032002639392939627, + "loss": 0.1786, + "step": 1400 + }, + { + "epoch": 5.7, + "grad_norm": 4.158301830291748, + "learning_rate": 0.0003183767733421313, + "loss": 0.1078, + "step": 1425 + }, + { + "epoch": 5.8, + "grad_norm": 10.65804672241211, + "learning_rate": 0.0003167271527548664, + "loss": 0.1597, + "step": 1450 + }, + { + "epoch": 5.9, + "grad_norm": 10.880438804626465, + "learning_rate": 0.00031507753216760147, + "loss": 0.1225, + "step": 1475 + }, + { + "epoch": 6.0, + "grad_norm": 9.95138168334961, + "learning_rate": 0.00031342791158033657, + "loss": 0.1254, + "step": 1500 + }, + { + "epoch": 6.1, + "grad_norm": 1.5018162727355957, + "learning_rate": 0.00031177829099307157, + "loss": 0.0902, + "step": 1525 + }, + { + "epoch": 6.2, + "grad_norm": 14.281820297241211, + "learning_rate": 0.00031012867040580667, + "loss": 0.1026, + "step": 1550 + }, + { + "epoch": 6.3, + "grad_norm": 4.7540435791015625, + "learning_rate": 0.0003084790498185417, + "loss": 0.0876, + "step": 1575 + }, + { + "epoch": 6.4, + "grad_norm": 4.304183006286621, + "learning_rate": 0.0003068294292312768, + "loss": 0.1355, + "step": 1600 + }, + { + "epoch": 6.5, + "grad_norm": 14.653755187988281, + "learning_rate": 0.0003051798086440119, + "loss": 0.0878, + "step": 1625 + }, + { + "epoch": 6.6, + "grad_norm": 11.13028335571289, + "learning_rate": 0.000303530188056747, + "loss": 0.1203, + "step": 1650 + }, + { + "epoch": 6.7, + "grad_norm": 9.729938507080078, + "learning_rate": 0.000301880567469482, + "loss": 0.0775, + "step": 1675 + }, + { + "epoch": 6.8, + "grad_norm": 10.12938117980957, + "learning_rate": 0.0003002309468822171, + "loss": 0.1015, + "step": 1700 + }, + { + "epoch": 6.9, + "grad_norm": 8.506614685058594, + "learning_rate": 0.0002985813262949522, + "loss": 0.1362, + "step": 1725 + }, + { + "epoch": 7.0, + "grad_norm": 25.618762969970703, + "learning_rate": 0.00029693170570768723, + "loss": 0.1269, + "step": 1750 + }, + { + "epoch": 7.1, + "grad_norm": 27.155860900878906, + "learning_rate": 0.00029528208512042233, + "loss": 0.0555, + "step": 1775 + }, + { + "epoch": 7.2, + "grad_norm": 9.412527084350586, + "learning_rate": 0.0002936324645331574, + "loss": 0.0695, + "step": 1800 + }, + { + "epoch": 7.3, + "grad_norm": 6.4328293800354, + "learning_rate": 0.00029198284394589243, + "loss": 0.0808, + "step": 1825 + }, + { + "epoch": 7.4, + "grad_norm": 15.267293930053711, + "learning_rate": 0.00029033322335862753, + "loss": 0.0871, + "step": 1850 + }, + { + "epoch": 7.5, + "grad_norm": 6.774560928344727, + "learning_rate": 0.0002886836027713626, + "loss": 0.0819, + "step": 1875 + }, + { + "epoch": 7.6, + "grad_norm": 18.67180633544922, + "learning_rate": 0.0002870339821840977, + "loss": 0.0565, + "step": 1900 + }, + { + "epoch": 7.7, + "grad_norm": 16.39004898071289, + "learning_rate": 0.00028538436159683273, + "loss": 0.0907, + "step": 1925 + }, + { + "epoch": 7.8, + "grad_norm": 6.351583003997803, + "learning_rate": 0.00028373474100956784, + "loss": 0.1289, + "step": 1950 + }, + { + "epoch": 7.9, + "grad_norm": 11.396740913391113, + "learning_rate": 0.0002820851204223029, + "loss": 0.1149, + "step": 1975 + }, + { + "epoch": 8.0, + "grad_norm": 6.491570949554443, + "learning_rate": 0.00028043549983503794, + "loss": 0.0891, + "step": 2000 + }, + { + "epoch": 8.1, + "grad_norm": 0.22946956753730774, + "learning_rate": 0.00027878587924777304, + "loss": 0.0413, + "step": 2025 + }, + { + "epoch": 8.2, + "grad_norm": 42.31606674194336, + "learning_rate": 0.0002771362586605081, + "loss": 0.0881, + "step": 2050 + }, + { + "epoch": 8.3, + "grad_norm": 18.490406036376953, + "learning_rate": 0.0002754866380732432, + "loss": 0.1238, + "step": 2075 + }, + { + "epoch": 8.4, + "grad_norm": 17.950319290161133, + "learning_rate": 0.00027383701748597824, + "loss": 0.0642, + "step": 2100 + }, + { + "epoch": 8.5, + "grad_norm": 37.16359329223633, + "learning_rate": 0.0002721873968987133, + "loss": 0.0916, + "step": 2125 + }, + { + "epoch": 8.6, + "grad_norm": 39.75199890136719, + "learning_rate": 0.00027053777631144834, + "loss": 0.1087, + "step": 2150 + }, + { + "epoch": 8.7, + "grad_norm": 0.08794238418340683, + "learning_rate": 0.00026888815572418344, + "loss": 0.0825, + "step": 2175 + }, + { + "epoch": 8.8, + "grad_norm": 14.681429862976074, + "learning_rate": 0.00026723853513691855, + "loss": 0.0925, + "step": 2200 + }, + { + "epoch": 8.9, + "grad_norm": 17.90182876586914, + "learning_rate": 0.0002655889145496536, + "loss": 0.0933, + "step": 2225 + }, + { + "epoch": 9.0, + "grad_norm": 0.9108687043190002, + "learning_rate": 0.0002639392939623887, + "loss": 0.0602, + "step": 2250 + }, + { + "epoch": 9.1, + "grad_norm": 9.671142578125, + "learning_rate": 0.0002622896733751237, + "loss": 0.0438, + "step": 2275 + }, + { + "epoch": 9.2, + "grad_norm": 12.797035217285156, + "learning_rate": 0.0002606400527878588, + "loss": 0.0495, + "step": 2300 + }, + { + "epoch": 9.3, + "grad_norm": 0.4585617184638977, + "learning_rate": 0.00025899043220059385, + "loss": 0.032, + "step": 2325 + }, + { + "epoch": 9.4, + "grad_norm": 0.05073042958974838, + "learning_rate": 0.00025734081161332895, + "loss": 0.067, + "step": 2350 + }, + { + "epoch": 9.5, + "grad_norm": 0.013653877191245556, + "learning_rate": 0.000255691191026064, + "loss": 0.0898, + "step": 2375 + }, + { + "epoch": 9.6, + "grad_norm": 18.198965072631836, + "learning_rate": 0.0002540415704387991, + "loss": 0.0804, + "step": 2400 + }, + { + "epoch": 9.7, + "grad_norm": 0.1363959163427353, + "learning_rate": 0.00025239194985153415, + "loss": 0.0584, + "step": 2425 + }, + { + "epoch": 9.8, + "grad_norm": 17.61290740966797, + "learning_rate": 0.0002507423292642692, + "loss": 0.0778, + "step": 2450 + }, + { + "epoch": 9.9, + "grad_norm": 0.10548887401819229, + "learning_rate": 0.0002490927086770043, + "loss": 0.0667, + "step": 2475 + }, + { + "epoch": 10.0, + "grad_norm": 0.1262211948633194, + "learning_rate": 0.00024744308808973936, + "loss": 0.0625, + "step": 2500 + }, + { + "epoch": 10.1, + "grad_norm": 0.019293297082185745, + "learning_rate": 0.00024579346750247446, + "loss": 0.0762, + "step": 2525 + }, + { + "epoch": 10.2, + "grad_norm": 0.9273566603660583, + "learning_rate": 0.0002441438469152095, + "loss": 0.0357, + "step": 2550 + }, + { + "epoch": 10.3, + "grad_norm": 10.720550537109375, + "learning_rate": 0.00024249422632794456, + "loss": 0.046, + "step": 2575 + }, + { + "epoch": 10.4, + "grad_norm": 1.8441611528396606, + "learning_rate": 0.00024084460574067963, + "loss": 0.0369, + "step": 2600 + }, + { + "epoch": 10.5, + "grad_norm": 0.8760197758674622, + "learning_rate": 0.0002391949851534147, + "loss": 0.046, + "step": 2625 + }, + { + "epoch": 10.6, + "grad_norm": 8.899499893188477, + "learning_rate": 0.0002375453645661498, + "loss": 0.08, + "step": 2650 + }, + { + "epoch": 10.7, + "grad_norm": 0.2664288580417633, + "learning_rate": 0.00023589574397888486, + "loss": 0.0616, + "step": 2675 + }, + { + "epoch": 10.8, + "grad_norm": 0.06609692424535751, + "learning_rate": 0.00023424612339161997, + "loss": 0.05, + "step": 2700 + }, + { + "epoch": 10.9, + "grad_norm": 32.44829177856445, + "learning_rate": 0.000232596502804355, + "loss": 0.0487, + "step": 2725 + }, + { + "epoch": 11.0, + "grad_norm": 7.25365686416626, + "learning_rate": 0.00023094688221709007, + "loss": 0.0456, + "step": 2750 + }, + { + "epoch": 11.1, + "grad_norm": 0.10101272910833359, + "learning_rate": 0.00022929726162982514, + "loss": 0.0328, + "step": 2775 + }, + { + "epoch": 11.2, + "grad_norm": 1.3087996244430542, + "learning_rate": 0.00022764764104256022, + "loss": 0.0606, + "step": 2800 + }, + { + "epoch": 11.3, + "grad_norm": 1.247656226158142, + "learning_rate": 0.0002259980204552953, + "loss": 0.0838, + "step": 2825 + }, + { + "epoch": 11.4, + "grad_norm": 1.7708797454833984, + "learning_rate": 0.00022434839986803037, + "loss": 0.0416, + "step": 2850 + }, + { + "epoch": 11.5, + "grad_norm": 11.131721496582031, + "learning_rate": 0.00022269877928076542, + "loss": 0.0528, + "step": 2875 + }, + { + "epoch": 11.6, + "grad_norm": 3.824920892715454, + "learning_rate": 0.0002210491586935005, + "loss": 0.0526, + "step": 2900 + }, + { + "epoch": 11.7, + "grad_norm": 0.8507682085037231, + "learning_rate": 0.00021939953810623557, + "loss": 0.0965, + "step": 2925 + }, + { + "epoch": 11.8, + "grad_norm": 0.011997719295322895, + "learning_rate": 0.00021774991751897065, + "loss": 0.0446, + "step": 2950 + }, + { + "epoch": 11.9, + "grad_norm": 0.0792650505900383, + "learning_rate": 0.00021610029693170573, + "loss": 0.0556, + "step": 2975 + }, + { + "epoch": 12.0, + "grad_norm": 0.007047992665320635, + "learning_rate": 0.0002144506763444408, + "loss": 0.0759, + "step": 3000 + }, + { + "epoch": 12.1, + "grad_norm": 0.7784481644630432, + "learning_rate": 0.00021280105575717582, + "loss": 0.0505, + "step": 3025 + }, + { + "epoch": 12.2, + "grad_norm": 9.635031700134277, + "learning_rate": 0.00021115143516991093, + "loss": 0.0284, + "step": 3050 + }, + { + "epoch": 12.3, + "grad_norm": 12.473057746887207, + "learning_rate": 0.000209501814582646, + "loss": 0.0531, + "step": 3075 + }, + { + "epoch": 12.4, + "grad_norm": 10.904062271118164, + "learning_rate": 0.00020785219399538108, + "loss": 0.0212, + "step": 3100 + }, + { + "epoch": 12.5, + "grad_norm": 9.258865356445312, + "learning_rate": 0.00020620257340811616, + "loss": 0.0367, + "step": 3125 + }, + { + "epoch": 12.6, + "grad_norm": 2.6826488971710205, + "learning_rate": 0.00020455295282085123, + "loss": 0.0457, + "step": 3150 + }, + { + "epoch": 12.7, + "grad_norm": 8.800697326660156, + "learning_rate": 0.00020290333223358626, + "loss": 0.0504, + "step": 3175 + }, + { + "epoch": 12.8, + "grad_norm": 0.023783667013049126, + "learning_rate": 0.00020125371164632133, + "loss": 0.0255, + "step": 3200 + }, + { + "epoch": 12.9, + "grad_norm": 5.948949337005615, + "learning_rate": 0.0001996040910590564, + "loss": 0.0413, + "step": 3225 + }, + { + "epoch": 13.0, + "grad_norm": 0.026645230129361153, + "learning_rate": 0.00019795447047179148, + "loss": 0.0249, + "step": 3250 + }, + { + "epoch": 13.1, + "grad_norm": 60.826637268066406, + "learning_rate": 0.0001963048498845266, + "loss": 0.0293, + "step": 3275 + }, + { + "epoch": 13.2, + "grad_norm": 26.056686401367188, + "learning_rate": 0.00019465522929726164, + "loss": 0.0638, + "step": 3300 + }, + { + "epoch": 13.3, + "grad_norm": 18.143638610839844, + "learning_rate": 0.00019300560870999671, + "loss": 0.0206, + "step": 3325 + }, + { + "epoch": 13.4, + "grad_norm": 27.899967193603516, + "learning_rate": 0.0001913559881227318, + "loss": 0.0423, + "step": 3350 + }, + { + "epoch": 13.5, + "grad_norm": 51.9879264831543, + "learning_rate": 0.00018970636753546684, + "loss": 0.0274, + "step": 3375 + }, + { + "epoch": 13.6, + "grad_norm": 0.028627146035432816, + "learning_rate": 0.00018805674694820192, + "loss": 0.0513, + "step": 3400 + }, + { + "epoch": 13.7, + "grad_norm": 33.211944580078125, + "learning_rate": 0.000186407126360937, + "loss": 0.0446, + "step": 3425 + }, + { + "epoch": 13.8, + "grad_norm": 32.37785339355469, + "learning_rate": 0.00018475750577367207, + "loss": 0.0538, + "step": 3450 + }, + { + "epoch": 13.9, + "grad_norm": 30.558673858642578, + "learning_rate": 0.00018310788518640715, + "loss": 0.0339, + "step": 3475 + }, + { + "epoch": 14.0, + "grad_norm": 19.316938400268555, + "learning_rate": 0.00018145826459914222, + "loss": 0.071, + "step": 3500 + }, + { + "epoch": 14.1, + "grad_norm": 0.1271055042743683, + "learning_rate": 0.00017980864401187727, + "loss": 0.0159, + "step": 3525 + }, + { + "epoch": 14.2, + "grad_norm": 51.1268196105957, + "learning_rate": 0.00017815902342461235, + "loss": 0.028, + "step": 3550 + }, + { + "epoch": 14.3, + "grad_norm": 0.003214583033695817, + "learning_rate": 0.00017650940283734742, + "loss": 0.0342, + "step": 3575 + }, + { + "epoch": 14.4, + "grad_norm": 94.75691223144531, + "learning_rate": 0.00017485978225008247, + "loss": 0.0243, + "step": 3600 + }, + { + "epoch": 14.5, + "grad_norm": 0.019844651222229004, + "learning_rate": 0.00017321016166281755, + "loss": 0.0205, + "step": 3625 + }, + { + "epoch": 14.6, + "grad_norm": 0.10069809854030609, + "learning_rate": 0.00017156054107555263, + "loss": 0.0402, + "step": 3650 + }, + { + "epoch": 14.7, + "grad_norm": 0.0023937481455504894, + "learning_rate": 0.0001699109204882877, + "loss": 0.0173, + "step": 3675 + }, + { + "epoch": 14.8, + "grad_norm": 0.2573629319667816, + "learning_rate": 0.00016826129990102278, + "loss": 0.0312, + "step": 3700 + }, + { + "epoch": 14.9, + "grad_norm": 0.0067031364887952805, + "learning_rate": 0.00016661167931375785, + "loss": 0.0418, + "step": 3725 + }, + { + "epoch": 15.0, + "grad_norm": 0.08329538255929947, + "learning_rate": 0.0001649620587264929, + "loss": 0.0164, + "step": 3750 + }, + { + "epoch": 15.1, + "grad_norm": 0.04175519943237305, + "learning_rate": 0.00016331243813922798, + "loss": 0.044, + "step": 3775 + }, + { + "epoch": 15.2, + "grad_norm": 0.005420651752501726, + "learning_rate": 0.00016166281755196306, + "loss": 0.0599, + "step": 3800 + }, + { + "epoch": 15.3, + "grad_norm": 20.68973731994629, + "learning_rate": 0.00016001319696469813, + "loss": 0.047, + "step": 3825 + }, + { + "epoch": 15.4, + "grad_norm": 1.1241929531097412, + "learning_rate": 0.0001583635763774332, + "loss": 0.0184, + "step": 3850 + }, + { + "epoch": 15.5, + "grad_norm": 0.00928444042801857, + "learning_rate": 0.00015671395579016829, + "loss": 0.0146, + "step": 3875 + }, + { + "epoch": 15.6, + "grad_norm": 3.2419674396514893, + "learning_rate": 0.00015506433520290334, + "loss": 0.0047, + "step": 3900 + }, + { + "epoch": 15.7, + "grad_norm": 0.0019528602715581656, + "learning_rate": 0.0001534147146156384, + "loss": 0.0177, + "step": 3925 + }, + { + "epoch": 15.8, + "grad_norm": 0.003142274683341384, + "learning_rate": 0.0001517650940283735, + "loss": 0.0319, + "step": 3950 + }, + { + "epoch": 15.9, + "grad_norm": 45.58081817626953, + "learning_rate": 0.00015011547344110854, + "loss": 0.0353, + "step": 3975 + }, + { + "epoch": 16.0, + "grad_norm": 0.003515947377309203, + "learning_rate": 0.00014846585285384361, + "loss": 0.0208, + "step": 4000 + } + ], + "logging_steps": 25, + "max_steps": 6250, + "num_input_tokens_seen": 0, + "num_train_epochs": 25, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 203277926400000.0, + "train_batch_size": 20, + "trial_name": null, + "trial_params": null +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-4000/training_args.bin b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-4000/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..77692b6666e83a688b38e799c964e4694a57eec9 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-4000/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f73b41315068a430d2439f5707c62152a34210df62454b9549f9642446dd90b2 +size 5304 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-4000/vocab.txt b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-4000/vocab.txt new file mode 100644 index 0000000000000000000000000000000000000000..cc9ad53732e21ea9180f18e1fb67a38cc5702d4c --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-4000/vocab.txt @@ -0,0 +1,4101 @@ +[PAD] +[UNK] +[CLS] +[SEP] +[MASK] +AAAAAA +AAAAAT +AAAAAC +AAAAAG +AAAATA +AAAATT +AAAATC +AAAATG +AAAACA +AAAACT +AAAACC +AAAACG +AAAAGA +AAAAGT +AAAAGC +AAAAGG +AAATAA +AAATAT +AAATAC +AAATAG +AAATTA +AAATTT +AAATTC +AAATTG +AAATCA +AAATCT +AAATCC +AAATCG +AAATGA +AAATGT +AAATGC +AAATGG +AAACAA +AAACAT +AAACAC +AAACAG +AAACTA +AAACTT +AAACTC +AAACTG +AAACCA +AAACCT +AAACCC +AAACCG +AAACGA +AAACGT +AAACGC +AAACGG +AAAGAA +AAAGAT +AAAGAC +AAAGAG +AAAGTA +AAAGTT +AAAGTC +AAAGTG +AAAGCA +AAAGCT +AAAGCC +AAAGCG +AAAGGA +AAAGGT +AAAGGC +AAAGGG +AATAAA +AATAAT +AATAAC +AATAAG +AATATA +AATATT +AATATC +AATATG +AATACA +AATACT +AATACC +AATACG +AATAGA +AATAGT +AATAGC +AATAGG +AATTAA +AATTAT +AATTAC +AATTAG +AATTTA +AATTTT +AATTTC +AATTTG +AATTCA +AATTCT +AATTCC +AATTCG +AATTGA +AATTGT +AATTGC +AATTGG +AATCAA +AATCAT +AATCAC +AATCAG +AATCTA +AATCTT +AATCTC +AATCTG +AATCCA +AATCCT +AATCCC +AATCCG +AATCGA +AATCGT +AATCGC +AATCGG +AATGAA +AATGAT +AATGAC +AATGAG +AATGTA +AATGTT +AATGTC +AATGTG +AATGCA +AATGCT +AATGCC +AATGCG +AATGGA +AATGGT +AATGGC +AATGGG +AACAAA +AACAAT +AACAAC +AACAAG +AACATA +AACATT +AACATC +AACATG +AACACA +AACACT +AACACC +AACACG +AACAGA +AACAGT +AACAGC +AACAGG +AACTAA +AACTAT +AACTAC +AACTAG +AACTTA +AACTTT +AACTTC +AACTTG +AACTCA +AACTCT +AACTCC +AACTCG +AACTGA +AACTGT +AACTGC +AACTGG +AACCAA +AACCAT +AACCAC +AACCAG +AACCTA +AACCTT +AACCTC +AACCTG +AACCCA +AACCCT +AACCCC +AACCCG +AACCGA +AACCGT +AACCGC +AACCGG +AACGAA +AACGAT +AACGAC +AACGAG +AACGTA +AACGTT +AACGTC +AACGTG +AACGCA +AACGCT +AACGCC +AACGCG +AACGGA +AACGGT +AACGGC +AACGGG +AAGAAA +AAGAAT +AAGAAC +AAGAAG +AAGATA +AAGATT +AAGATC +AAGATG +AAGACA +AAGACT +AAGACC +AAGACG +AAGAGA +AAGAGT +AAGAGC +AAGAGG +AAGTAA +AAGTAT +AAGTAC +AAGTAG +AAGTTA +AAGTTT +AAGTTC +AAGTTG +AAGTCA +AAGTCT +AAGTCC +AAGTCG +AAGTGA +AAGTGT +AAGTGC +AAGTGG +AAGCAA +AAGCAT +AAGCAC +AAGCAG +AAGCTA +AAGCTT +AAGCTC +AAGCTG +AAGCCA +AAGCCT +AAGCCC +AAGCCG +AAGCGA +AAGCGT +AAGCGC +AAGCGG +AAGGAA +AAGGAT +AAGGAC +AAGGAG +AAGGTA +AAGGTT +AAGGTC +AAGGTG +AAGGCA +AAGGCT +AAGGCC +AAGGCG +AAGGGA +AAGGGT +AAGGGC +AAGGGG +ATAAAA +ATAAAT +ATAAAC +ATAAAG +ATAATA +ATAATT +ATAATC +ATAATG +ATAACA +ATAACT +ATAACC +ATAACG +ATAAGA +ATAAGT +ATAAGC +ATAAGG +ATATAA +ATATAT +ATATAC +ATATAG +ATATTA +ATATTT +ATATTC +ATATTG +ATATCA +ATATCT +ATATCC +ATATCG +ATATGA +ATATGT +ATATGC +ATATGG +ATACAA +ATACAT +ATACAC +ATACAG +ATACTA +ATACTT +ATACTC +ATACTG +ATACCA +ATACCT +ATACCC +ATACCG +ATACGA +ATACGT +ATACGC +ATACGG +ATAGAA +ATAGAT +ATAGAC +ATAGAG +ATAGTA +ATAGTT +ATAGTC +ATAGTG +ATAGCA +ATAGCT +ATAGCC +ATAGCG +ATAGGA +ATAGGT +ATAGGC +ATAGGG +ATTAAA +ATTAAT +ATTAAC +ATTAAG +ATTATA +ATTATT +ATTATC +ATTATG +ATTACA +ATTACT +ATTACC +ATTACG +ATTAGA +ATTAGT +ATTAGC +ATTAGG +ATTTAA +ATTTAT +ATTTAC +ATTTAG +ATTTTA +ATTTTT +ATTTTC +ATTTTG +ATTTCA +ATTTCT +ATTTCC +ATTTCG +ATTTGA +ATTTGT +ATTTGC +ATTTGG +ATTCAA +ATTCAT +ATTCAC +ATTCAG +ATTCTA +ATTCTT +ATTCTC +ATTCTG +ATTCCA +ATTCCT +ATTCCC +ATTCCG +ATTCGA +ATTCGT +ATTCGC +ATTCGG +ATTGAA +ATTGAT +ATTGAC +ATTGAG +ATTGTA +ATTGTT +ATTGTC +ATTGTG +ATTGCA +ATTGCT +ATTGCC +ATTGCG +ATTGGA +ATTGGT +ATTGGC +ATTGGG +ATCAAA +ATCAAT +ATCAAC +ATCAAG +ATCATA +ATCATT +ATCATC +ATCATG +ATCACA +ATCACT +ATCACC +ATCACG +ATCAGA +ATCAGT +ATCAGC +ATCAGG +ATCTAA +ATCTAT +ATCTAC +ATCTAG +ATCTTA +ATCTTT +ATCTTC +ATCTTG +ATCTCA +ATCTCT +ATCTCC +ATCTCG +ATCTGA +ATCTGT +ATCTGC +ATCTGG +ATCCAA +ATCCAT +ATCCAC +ATCCAG +ATCCTA +ATCCTT +ATCCTC +ATCCTG +ATCCCA +ATCCCT +ATCCCC +ATCCCG +ATCCGA +ATCCGT +ATCCGC +ATCCGG +ATCGAA +ATCGAT +ATCGAC +ATCGAG +ATCGTA +ATCGTT +ATCGTC +ATCGTG +ATCGCA +ATCGCT +ATCGCC +ATCGCG +ATCGGA +ATCGGT +ATCGGC +ATCGGG +ATGAAA +ATGAAT +ATGAAC +ATGAAG +ATGATA +ATGATT +ATGATC +ATGATG +ATGACA +ATGACT +ATGACC +ATGACG +ATGAGA +ATGAGT +ATGAGC +ATGAGG +ATGTAA +ATGTAT +ATGTAC +ATGTAG +ATGTTA +ATGTTT +ATGTTC +ATGTTG +ATGTCA +ATGTCT +ATGTCC +ATGTCG +ATGTGA +ATGTGT +ATGTGC +ATGTGG +ATGCAA +ATGCAT +ATGCAC +ATGCAG +ATGCTA +ATGCTT +ATGCTC +ATGCTG +ATGCCA +ATGCCT +ATGCCC +ATGCCG +ATGCGA +ATGCGT +ATGCGC +ATGCGG +ATGGAA +ATGGAT +ATGGAC +ATGGAG +ATGGTA +ATGGTT +ATGGTC +ATGGTG +ATGGCA +ATGGCT +ATGGCC +ATGGCG +ATGGGA +ATGGGT +ATGGGC +ATGGGG +ACAAAA +ACAAAT +ACAAAC +ACAAAG +ACAATA +ACAATT +ACAATC +ACAATG +ACAACA +ACAACT +ACAACC +ACAACG +ACAAGA +ACAAGT +ACAAGC +ACAAGG +ACATAA +ACATAT +ACATAC +ACATAG +ACATTA +ACATTT +ACATTC +ACATTG +ACATCA +ACATCT +ACATCC +ACATCG +ACATGA +ACATGT +ACATGC +ACATGG +ACACAA +ACACAT +ACACAC +ACACAG +ACACTA +ACACTT +ACACTC +ACACTG +ACACCA +ACACCT +ACACCC +ACACCG +ACACGA +ACACGT +ACACGC +ACACGG +ACAGAA +ACAGAT +ACAGAC +ACAGAG +ACAGTA +ACAGTT +ACAGTC +ACAGTG +ACAGCA +ACAGCT +ACAGCC +ACAGCG +ACAGGA +ACAGGT +ACAGGC +ACAGGG +ACTAAA +ACTAAT +ACTAAC +ACTAAG +ACTATA +ACTATT +ACTATC +ACTATG +ACTACA +ACTACT +ACTACC +ACTACG +ACTAGA +ACTAGT +ACTAGC +ACTAGG +ACTTAA +ACTTAT +ACTTAC +ACTTAG +ACTTTA +ACTTTT +ACTTTC +ACTTTG +ACTTCA +ACTTCT +ACTTCC +ACTTCG +ACTTGA +ACTTGT +ACTTGC +ACTTGG +ACTCAA +ACTCAT +ACTCAC +ACTCAG +ACTCTA +ACTCTT +ACTCTC +ACTCTG +ACTCCA +ACTCCT +ACTCCC +ACTCCG +ACTCGA +ACTCGT +ACTCGC +ACTCGG +ACTGAA +ACTGAT +ACTGAC +ACTGAG +ACTGTA +ACTGTT +ACTGTC +ACTGTG +ACTGCA +ACTGCT +ACTGCC +ACTGCG +ACTGGA +ACTGGT +ACTGGC +ACTGGG +ACCAAA +ACCAAT +ACCAAC +ACCAAG +ACCATA +ACCATT +ACCATC +ACCATG +ACCACA +ACCACT +ACCACC +ACCACG +ACCAGA +ACCAGT +ACCAGC +ACCAGG +ACCTAA +ACCTAT +ACCTAC +ACCTAG +ACCTTA +ACCTTT +ACCTTC +ACCTTG +ACCTCA +ACCTCT +ACCTCC +ACCTCG +ACCTGA +ACCTGT +ACCTGC +ACCTGG +ACCCAA +ACCCAT +ACCCAC +ACCCAG +ACCCTA +ACCCTT +ACCCTC +ACCCTG +ACCCCA +ACCCCT +ACCCCC +ACCCCG +ACCCGA +ACCCGT +ACCCGC +ACCCGG +ACCGAA +ACCGAT +ACCGAC +ACCGAG +ACCGTA +ACCGTT +ACCGTC +ACCGTG +ACCGCA +ACCGCT +ACCGCC +ACCGCG +ACCGGA +ACCGGT +ACCGGC +ACCGGG +ACGAAA +ACGAAT +ACGAAC +ACGAAG +ACGATA +ACGATT +ACGATC +ACGATG +ACGACA +ACGACT +ACGACC +ACGACG +ACGAGA +ACGAGT +ACGAGC +ACGAGG +ACGTAA +ACGTAT +ACGTAC +ACGTAG +ACGTTA +ACGTTT +ACGTTC +ACGTTG +ACGTCA +ACGTCT +ACGTCC +ACGTCG +ACGTGA +ACGTGT +ACGTGC +ACGTGG +ACGCAA +ACGCAT +ACGCAC +ACGCAG +ACGCTA +ACGCTT +ACGCTC +ACGCTG +ACGCCA +ACGCCT +ACGCCC +ACGCCG +ACGCGA +ACGCGT +ACGCGC +ACGCGG +ACGGAA +ACGGAT +ACGGAC +ACGGAG +ACGGTA +ACGGTT +ACGGTC +ACGGTG +ACGGCA +ACGGCT +ACGGCC +ACGGCG +ACGGGA +ACGGGT +ACGGGC +ACGGGG +AGAAAA +AGAAAT +AGAAAC +AGAAAG +AGAATA +AGAATT +AGAATC +AGAATG +AGAACA +AGAACT +AGAACC +AGAACG +AGAAGA +AGAAGT +AGAAGC +AGAAGG +AGATAA +AGATAT +AGATAC +AGATAG +AGATTA +AGATTT +AGATTC +AGATTG +AGATCA +AGATCT +AGATCC +AGATCG +AGATGA +AGATGT +AGATGC +AGATGG +AGACAA +AGACAT +AGACAC +AGACAG +AGACTA +AGACTT +AGACTC +AGACTG +AGACCA +AGACCT +AGACCC +AGACCG +AGACGA +AGACGT +AGACGC +AGACGG +AGAGAA +AGAGAT +AGAGAC +AGAGAG +AGAGTA +AGAGTT +AGAGTC +AGAGTG +AGAGCA +AGAGCT +AGAGCC +AGAGCG +AGAGGA +AGAGGT +AGAGGC +AGAGGG +AGTAAA +AGTAAT +AGTAAC +AGTAAG +AGTATA +AGTATT +AGTATC +AGTATG +AGTACA +AGTACT +AGTACC +AGTACG +AGTAGA +AGTAGT +AGTAGC +AGTAGG +AGTTAA +AGTTAT +AGTTAC +AGTTAG +AGTTTA +AGTTTT +AGTTTC +AGTTTG +AGTTCA +AGTTCT +AGTTCC +AGTTCG +AGTTGA +AGTTGT +AGTTGC +AGTTGG +AGTCAA +AGTCAT +AGTCAC +AGTCAG +AGTCTA +AGTCTT +AGTCTC +AGTCTG +AGTCCA +AGTCCT +AGTCCC +AGTCCG +AGTCGA +AGTCGT +AGTCGC +AGTCGG +AGTGAA +AGTGAT +AGTGAC +AGTGAG +AGTGTA +AGTGTT +AGTGTC +AGTGTG +AGTGCA +AGTGCT +AGTGCC +AGTGCG +AGTGGA +AGTGGT +AGTGGC +AGTGGG +AGCAAA +AGCAAT +AGCAAC +AGCAAG +AGCATA +AGCATT +AGCATC +AGCATG +AGCACA +AGCACT +AGCACC +AGCACG +AGCAGA +AGCAGT +AGCAGC +AGCAGG +AGCTAA +AGCTAT +AGCTAC +AGCTAG +AGCTTA +AGCTTT +AGCTTC +AGCTTG +AGCTCA +AGCTCT +AGCTCC +AGCTCG +AGCTGA +AGCTGT +AGCTGC +AGCTGG +AGCCAA +AGCCAT +AGCCAC +AGCCAG +AGCCTA +AGCCTT +AGCCTC +AGCCTG +AGCCCA +AGCCCT +AGCCCC +AGCCCG +AGCCGA +AGCCGT +AGCCGC +AGCCGG +AGCGAA +AGCGAT +AGCGAC +AGCGAG +AGCGTA +AGCGTT +AGCGTC +AGCGTG +AGCGCA +AGCGCT +AGCGCC +AGCGCG +AGCGGA +AGCGGT +AGCGGC +AGCGGG +AGGAAA +AGGAAT +AGGAAC +AGGAAG +AGGATA +AGGATT +AGGATC +AGGATG +AGGACA +AGGACT +AGGACC +AGGACG +AGGAGA +AGGAGT +AGGAGC +AGGAGG +AGGTAA +AGGTAT +AGGTAC +AGGTAG +AGGTTA +AGGTTT +AGGTTC +AGGTTG +AGGTCA +AGGTCT +AGGTCC +AGGTCG +AGGTGA +AGGTGT +AGGTGC +AGGTGG +AGGCAA +AGGCAT +AGGCAC +AGGCAG +AGGCTA +AGGCTT +AGGCTC +AGGCTG +AGGCCA +AGGCCT +AGGCCC +AGGCCG +AGGCGA +AGGCGT +AGGCGC +AGGCGG +AGGGAA +AGGGAT +AGGGAC +AGGGAG +AGGGTA +AGGGTT +AGGGTC +AGGGTG +AGGGCA +AGGGCT +AGGGCC +AGGGCG +AGGGGA +AGGGGT +AGGGGC +AGGGGG +TAAAAA +TAAAAT +TAAAAC +TAAAAG +TAAATA +TAAATT +TAAATC +TAAATG +TAAACA +TAAACT +TAAACC +TAAACG +TAAAGA +TAAAGT +TAAAGC +TAAAGG +TAATAA +TAATAT +TAATAC +TAATAG +TAATTA +TAATTT +TAATTC +TAATTG +TAATCA +TAATCT +TAATCC +TAATCG +TAATGA +TAATGT +TAATGC +TAATGG +TAACAA +TAACAT +TAACAC +TAACAG +TAACTA +TAACTT +TAACTC +TAACTG +TAACCA +TAACCT +TAACCC +TAACCG +TAACGA +TAACGT +TAACGC +TAACGG +TAAGAA +TAAGAT +TAAGAC +TAAGAG +TAAGTA +TAAGTT +TAAGTC +TAAGTG +TAAGCA +TAAGCT +TAAGCC +TAAGCG +TAAGGA +TAAGGT +TAAGGC +TAAGGG +TATAAA +TATAAT +TATAAC +TATAAG +TATATA +TATATT +TATATC +TATATG +TATACA +TATACT +TATACC +TATACG +TATAGA +TATAGT +TATAGC +TATAGG +TATTAA +TATTAT +TATTAC +TATTAG +TATTTA +TATTTT +TATTTC +TATTTG +TATTCA +TATTCT +TATTCC +TATTCG +TATTGA +TATTGT +TATTGC +TATTGG +TATCAA +TATCAT +TATCAC +TATCAG +TATCTA +TATCTT +TATCTC +TATCTG +TATCCA +TATCCT +TATCCC +TATCCG +TATCGA +TATCGT +TATCGC +TATCGG +TATGAA +TATGAT +TATGAC +TATGAG +TATGTA +TATGTT +TATGTC +TATGTG +TATGCA +TATGCT +TATGCC +TATGCG +TATGGA +TATGGT +TATGGC +TATGGG +TACAAA +TACAAT +TACAAC +TACAAG +TACATA +TACATT +TACATC +TACATG +TACACA +TACACT +TACACC +TACACG +TACAGA +TACAGT +TACAGC +TACAGG +TACTAA +TACTAT +TACTAC +TACTAG +TACTTA +TACTTT +TACTTC +TACTTG +TACTCA +TACTCT +TACTCC +TACTCG +TACTGA +TACTGT +TACTGC +TACTGG +TACCAA +TACCAT +TACCAC +TACCAG +TACCTA +TACCTT +TACCTC +TACCTG +TACCCA +TACCCT +TACCCC +TACCCG +TACCGA +TACCGT +TACCGC +TACCGG +TACGAA +TACGAT +TACGAC +TACGAG +TACGTA +TACGTT +TACGTC +TACGTG +TACGCA +TACGCT +TACGCC +TACGCG +TACGGA +TACGGT +TACGGC +TACGGG +TAGAAA +TAGAAT +TAGAAC +TAGAAG +TAGATA +TAGATT +TAGATC +TAGATG +TAGACA +TAGACT +TAGACC +TAGACG +TAGAGA +TAGAGT +TAGAGC +TAGAGG +TAGTAA +TAGTAT +TAGTAC +TAGTAG +TAGTTA +TAGTTT +TAGTTC +TAGTTG +TAGTCA +TAGTCT +TAGTCC +TAGTCG +TAGTGA +TAGTGT +TAGTGC +TAGTGG +TAGCAA +TAGCAT +TAGCAC +TAGCAG +TAGCTA +TAGCTT +TAGCTC +TAGCTG +TAGCCA +TAGCCT +TAGCCC +TAGCCG +TAGCGA +TAGCGT +TAGCGC +TAGCGG +TAGGAA +TAGGAT +TAGGAC +TAGGAG +TAGGTA +TAGGTT +TAGGTC +TAGGTG +TAGGCA +TAGGCT +TAGGCC +TAGGCG +TAGGGA +TAGGGT +TAGGGC +TAGGGG +TTAAAA +TTAAAT +TTAAAC +TTAAAG +TTAATA +TTAATT +TTAATC +TTAATG +TTAACA +TTAACT +TTAACC +TTAACG +TTAAGA +TTAAGT +TTAAGC +TTAAGG +TTATAA +TTATAT +TTATAC +TTATAG +TTATTA +TTATTT +TTATTC +TTATTG +TTATCA +TTATCT +TTATCC +TTATCG +TTATGA +TTATGT +TTATGC +TTATGG +TTACAA +TTACAT +TTACAC +TTACAG +TTACTA +TTACTT +TTACTC +TTACTG +TTACCA +TTACCT +TTACCC +TTACCG +TTACGA +TTACGT +TTACGC +TTACGG +TTAGAA +TTAGAT +TTAGAC +TTAGAG +TTAGTA +TTAGTT +TTAGTC +TTAGTG +TTAGCA +TTAGCT +TTAGCC +TTAGCG +TTAGGA +TTAGGT +TTAGGC +TTAGGG +TTTAAA +TTTAAT +TTTAAC +TTTAAG +TTTATA +TTTATT +TTTATC +TTTATG +TTTACA +TTTACT +TTTACC +TTTACG +TTTAGA +TTTAGT +TTTAGC +TTTAGG +TTTTAA +TTTTAT +TTTTAC +TTTTAG +TTTTTA +TTTTTT +TTTTTC +TTTTTG +TTTTCA +TTTTCT +TTTTCC +TTTTCG +TTTTGA +TTTTGT +TTTTGC +TTTTGG +TTTCAA +TTTCAT +TTTCAC +TTTCAG +TTTCTA +TTTCTT +TTTCTC +TTTCTG +TTTCCA +TTTCCT +TTTCCC +TTTCCG +TTTCGA +TTTCGT +TTTCGC +TTTCGG +TTTGAA +TTTGAT +TTTGAC +TTTGAG +TTTGTA +TTTGTT +TTTGTC +TTTGTG +TTTGCA +TTTGCT +TTTGCC +TTTGCG +TTTGGA +TTTGGT +TTTGGC +TTTGGG +TTCAAA +TTCAAT +TTCAAC +TTCAAG +TTCATA +TTCATT +TTCATC +TTCATG +TTCACA +TTCACT +TTCACC +TTCACG +TTCAGA +TTCAGT +TTCAGC +TTCAGG +TTCTAA +TTCTAT +TTCTAC +TTCTAG +TTCTTA +TTCTTT +TTCTTC +TTCTTG +TTCTCA +TTCTCT +TTCTCC +TTCTCG +TTCTGA +TTCTGT +TTCTGC +TTCTGG +TTCCAA +TTCCAT +TTCCAC +TTCCAG +TTCCTA +TTCCTT +TTCCTC +TTCCTG +TTCCCA +TTCCCT +TTCCCC +TTCCCG +TTCCGA +TTCCGT +TTCCGC +TTCCGG +TTCGAA +TTCGAT +TTCGAC +TTCGAG +TTCGTA +TTCGTT +TTCGTC +TTCGTG +TTCGCA +TTCGCT +TTCGCC +TTCGCG +TTCGGA +TTCGGT +TTCGGC +TTCGGG +TTGAAA +TTGAAT +TTGAAC +TTGAAG +TTGATA +TTGATT +TTGATC +TTGATG +TTGACA +TTGACT +TTGACC +TTGACG +TTGAGA +TTGAGT +TTGAGC +TTGAGG +TTGTAA +TTGTAT +TTGTAC +TTGTAG +TTGTTA +TTGTTT +TTGTTC +TTGTTG +TTGTCA +TTGTCT +TTGTCC +TTGTCG +TTGTGA +TTGTGT +TTGTGC +TTGTGG +TTGCAA +TTGCAT +TTGCAC +TTGCAG +TTGCTA +TTGCTT +TTGCTC +TTGCTG +TTGCCA +TTGCCT +TTGCCC +TTGCCG +TTGCGA +TTGCGT +TTGCGC +TTGCGG +TTGGAA +TTGGAT +TTGGAC +TTGGAG +TTGGTA +TTGGTT +TTGGTC +TTGGTG +TTGGCA +TTGGCT +TTGGCC +TTGGCG +TTGGGA +TTGGGT +TTGGGC +TTGGGG +TCAAAA +TCAAAT +TCAAAC +TCAAAG +TCAATA +TCAATT +TCAATC +TCAATG +TCAACA +TCAACT +TCAACC +TCAACG +TCAAGA +TCAAGT +TCAAGC +TCAAGG +TCATAA +TCATAT +TCATAC +TCATAG +TCATTA +TCATTT +TCATTC +TCATTG +TCATCA +TCATCT +TCATCC +TCATCG +TCATGA +TCATGT +TCATGC +TCATGG +TCACAA +TCACAT +TCACAC +TCACAG +TCACTA +TCACTT +TCACTC +TCACTG +TCACCA +TCACCT +TCACCC +TCACCG +TCACGA +TCACGT +TCACGC +TCACGG +TCAGAA +TCAGAT +TCAGAC +TCAGAG +TCAGTA +TCAGTT +TCAGTC +TCAGTG +TCAGCA +TCAGCT +TCAGCC +TCAGCG +TCAGGA +TCAGGT +TCAGGC +TCAGGG +TCTAAA +TCTAAT +TCTAAC +TCTAAG +TCTATA +TCTATT +TCTATC +TCTATG +TCTACA +TCTACT +TCTACC +TCTACG +TCTAGA +TCTAGT +TCTAGC +TCTAGG +TCTTAA +TCTTAT +TCTTAC +TCTTAG +TCTTTA +TCTTTT +TCTTTC +TCTTTG +TCTTCA +TCTTCT +TCTTCC +TCTTCG +TCTTGA +TCTTGT +TCTTGC +TCTTGG +TCTCAA +TCTCAT +TCTCAC +TCTCAG +TCTCTA +TCTCTT +TCTCTC +TCTCTG +TCTCCA +TCTCCT +TCTCCC +TCTCCG +TCTCGA +TCTCGT +TCTCGC +TCTCGG +TCTGAA +TCTGAT +TCTGAC +TCTGAG +TCTGTA +TCTGTT +TCTGTC +TCTGTG +TCTGCA +TCTGCT +TCTGCC +TCTGCG +TCTGGA +TCTGGT +TCTGGC +TCTGGG +TCCAAA +TCCAAT +TCCAAC +TCCAAG +TCCATA +TCCATT +TCCATC +TCCATG +TCCACA +TCCACT +TCCACC +TCCACG +TCCAGA +TCCAGT +TCCAGC +TCCAGG +TCCTAA +TCCTAT +TCCTAC +TCCTAG +TCCTTA +TCCTTT +TCCTTC +TCCTTG +TCCTCA +TCCTCT +TCCTCC +TCCTCG +TCCTGA +TCCTGT +TCCTGC +TCCTGG +TCCCAA +TCCCAT +TCCCAC +TCCCAG +TCCCTA +TCCCTT +TCCCTC +TCCCTG +TCCCCA +TCCCCT +TCCCCC +TCCCCG +TCCCGA +TCCCGT +TCCCGC +TCCCGG +TCCGAA +TCCGAT +TCCGAC +TCCGAG +TCCGTA +TCCGTT +TCCGTC +TCCGTG +TCCGCA +TCCGCT +TCCGCC +TCCGCG +TCCGGA +TCCGGT +TCCGGC +TCCGGG +TCGAAA +TCGAAT +TCGAAC +TCGAAG +TCGATA +TCGATT +TCGATC +TCGATG +TCGACA +TCGACT +TCGACC +TCGACG +TCGAGA +TCGAGT +TCGAGC +TCGAGG +TCGTAA +TCGTAT +TCGTAC +TCGTAG +TCGTTA +TCGTTT +TCGTTC +TCGTTG +TCGTCA +TCGTCT +TCGTCC +TCGTCG +TCGTGA +TCGTGT +TCGTGC +TCGTGG +TCGCAA +TCGCAT +TCGCAC +TCGCAG +TCGCTA +TCGCTT +TCGCTC +TCGCTG +TCGCCA +TCGCCT +TCGCCC +TCGCCG +TCGCGA +TCGCGT +TCGCGC +TCGCGG +TCGGAA +TCGGAT +TCGGAC +TCGGAG +TCGGTA +TCGGTT +TCGGTC +TCGGTG +TCGGCA +TCGGCT +TCGGCC +TCGGCG +TCGGGA +TCGGGT +TCGGGC +TCGGGG +TGAAAA +TGAAAT +TGAAAC +TGAAAG +TGAATA +TGAATT +TGAATC +TGAATG +TGAACA +TGAACT +TGAACC +TGAACG +TGAAGA +TGAAGT +TGAAGC +TGAAGG +TGATAA +TGATAT +TGATAC +TGATAG +TGATTA +TGATTT +TGATTC +TGATTG +TGATCA +TGATCT +TGATCC +TGATCG +TGATGA +TGATGT +TGATGC +TGATGG +TGACAA +TGACAT +TGACAC +TGACAG +TGACTA +TGACTT +TGACTC +TGACTG +TGACCA +TGACCT +TGACCC +TGACCG +TGACGA +TGACGT +TGACGC +TGACGG +TGAGAA +TGAGAT +TGAGAC +TGAGAG +TGAGTA +TGAGTT +TGAGTC +TGAGTG +TGAGCA +TGAGCT +TGAGCC +TGAGCG +TGAGGA +TGAGGT +TGAGGC +TGAGGG +TGTAAA +TGTAAT +TGTAAC +TGTAAG +TGTATA +TGTATT +TGTATC +TGTATG +TGTACA +TGTACT +TGTACC +TGTACG +TGTAGA +TGTAGT +TGTAGC +TGTAGG +TGTTAA +TGTTAT +TGTTAC +TGTTAG +TGTTTA +TGTTTT +TGTTTC +TGTTTG +TGTTCA +TGTTCT +TGTTCC +TGTTCG +TGTTGA +TGTTGT +TGTTGC +TGTTGG +TGTCAA +TGTCAT +TGTCAC +TGTCAG +TGTCTA +TGTCTT +TGTCTC +TGTCTG +TGTCCA +TGTCCT +TGTCCC +TGTCCG +TGTCGA +TGTCGT +TGTCGC +TGTCGG +TGTGAA +TGTGAT +TGTGAC +TGTGAG +TGTGTA +TGTGTT +TGTGTC +TGTGTG +TGTGCA +TGTGCT +TGTGCC +TGTGCG +TGTGGA +TGTGGT +TGTGGC +TGTGGG +TGCAAA +TGCAAT +TGCAAC +TGCAAG +TGCATA +TGCATT +TGCATC +TGCATG +TGCACA +TGCACT +TGCACC +TGCACG +TGCAGA +TGCAGT +TGCAGC +TGCAGG +TGCTAA +TGCTAT +TGCTAC +TGCTAG +TGCTTA +TGCTTT +TGCTTC +TGCTTG +TGCTCA +TGCTCT +TGCTCC +TGCTCG +TGCTGA +TGCTGT +TGCTGC +TGCTGG +TGCCAA +TGCCAT +TGCCAC +TGCCAG +TGCCTA +TGCCTT +TGCCTC +TGCCTG +TGCCCA +TGCCCT +TGCCCC +TGCCCG +TGCCGA +TGCCGT +TGCCGC +TGCCGG +TGCGAA +TGCGAT +TGCGAC +TGCGAG +TGCGTA +TGCGTT +TGCGTC +TGCGTG +TGCGCA +TGCGCT +TGCGCC +TGCGCG +TGCGGA +TGCGGT +TGCGGC +TGCGGG +TGGAAA +TGGAAT +TGGAAC +TGGAAG +TGGATA +TGGATT +TGGATC +TGGATG +TGGACA +TGGACT +TGGACC +TGGACG +TGGAGA +TGGAGT +TGGAGC +TGGAGG +TGGTAA +TGGTAT +TGGTAC +TGGTAG +TGGTTA +TGGTTT +TGGTTC +TGGTTG +TGGTCA +TGGTCT +TGGTCC +TGGTCG +TGGTGA +TGGTGT +TGGTGC +TGGTGG +TGGCAA +TGGCAT +TGGCAC +TGGCAG +TGGCTA +TGGCTT +TGGCTC +TGGCTG +TGGCCA +TGGCCT +TGGCCC +TGGCCG +TGGCGA +TGGCGT +TGGCGC +TGGCGG +TGGGAA +TGGGAT +TGGGAC +TGGGAG +TGGGTA +TGGGTT +TGGGTC +TGGGTG +TGGGCA +TGGGCT +TGGGCC +TGGGCG +TGGGGA +TGGGGT +TGGGGC +TGGGGG +CAAAAA +CAAAAT +CAAAAC +CAAAAG +CAAATA +CAAATT +CAAATC +CAAATG +CAAACA +CAAACT +CAAACC +CAAACG +CAAAGA +CAAAGT +CAAAGC +CAAAGG +CAATAA +CAATAT +CAATAC +CAATAG +CAATTA +CAATTT +CAATTC +CAATTG +CAATCA +CAATCT +CAATCC +CAATCG +CAATGA +CAATGT +CAATGC +CAATGG +CAACAA +CAACAT +CAACAC +CAACAG +CAACTA +CAACTT +CAACTC +CAACTG +CAACCA +CAACCT +CAACCC +CAACCG +CAACGA +CAACGT +CAACGC +CAACGG +CAAGAA +CAAGAT +CAAGAC +CAAGAG +CAAGTA +CAAGTT +CAAGTC +CAAGTG +CAAGCA +CAAGCT +CAAGCC +CAAGCG +CAAGGA +CAAGGT +CAAGGC +CAAGGG +CATAAA +CATAAT +CATAAC +CATAAG +CATATA +CATATT +CATATC +CATATG +CATACA +CATACT +CATACC +CATACG +CATAGA +CATAGT +CATAGC +CATAGG +CATTAA +CATTAT +CATTAC +CATTAG +CATTTA +CATTTT +CATTTC +CATTTG +CATTCA +CATTCT +CATTCC +CATTCG +CATTGA +CATTGT +CATTGC +CATTGG +CATCAA +CATCAT +CATCAC +CATCAG +CATCTA +CATCTT +CATCTC +CATCTG +CATCCA +CATCCT +CATCCC +CATCCG +CATCGA +CATCGT +CATCGC +CATCGG +CATGAA +CATGAT +CATGAC +CATGAG +CATGTA +CATGTT +CATGTC +CATGTG +CATGCA +CATGCT +CATGCC +CATGCG +CATGGA +CATGGT +CATGGC +CATGGG +CACAAA +CACAAT +CACAAC +CACAAG +CACATA +CACATT +CACATC +CACATG +CACACA +CACACT +CACACC +CACACG +CACAGA +CACAGT +CACAGC +CACAGG +CACTAA +CACTAT +CACTAC +CACTAG +CACTTA +CACTTT +CACTTC +CACTTG +CACTCA +CACTCT +CACTCC +CACTCG +CACTGA +CACTGT +CACTGC +CACTGG +CACCAA +CACCAT +CACCAC +CACCAG +CACCTA +CACCTT +CACCTC +CACCTG +CACCCA +CACCCT +CACCCC +CACCCG +CACCGA +CACCGT +CACCGC +CACCGG +CACGAA +CACGAT +CACGAC +CACGAG +CACGTA +CACGTT +CACGTC +CACGTG +CACGCA +CACGCT +CACGCC +CACGCG +CACGGA +CACGGT +CACGGC +CACGGG +CAGAAA +CAGAAT +CAGAAC +CAGAAG +CAGATA +CAGATT +CAGATC +CAGATG +CAGACA +CAGACT +CAGACC +CAGACG +CAGAGA +CAGAGT +CAGAGC +CAGAGG +CAGTAA +CAGTAT +CAGTAC +CAGTAG +CAGTTA +CAGTTT +CAGTTC +CAGTTG +CAGTCA +CAGTCT +CAGTCC +CAGTCG +CAGTGA +CAGTGT +CAGTGC +CAGTGG +CAGCAA +CAGCAT +CAGCAC +CAGCAG +CAGCTA +CAGCTT +CAGCTC +CAGCTG +CAGCCA +CAGCCT +CAGCCC +CAGCCG +CAGCGA +CAGCGT +CAGCGC +CAGCGG +CAGGAA +CAGGAT +CAGGAC +CAGGAG +CAGGTA +CAGGTT +CAGGTC +CAGGTG +CAGGCA +CAGGCT +CAGGCC +CAGGCG +CAGGGA +CAGGGT +CAGGGC +CAGGGG +CTAAAA +CTAAAT +CTAAAC +CTAAAG +CTAATA +CTAATT +CTAATC +CTAATG +CTAACA +CTAACT +CTAACC +CTAACG +CTAAGA +CTAAGT +CTAAGC +CTAAGG +CTATAA +CTATAT +CTATAC +CTATAG +CTATTA +CTATTT +CTATTC +CTATTG +CTATCA +CTATCT +CTATCC +CTATCG +CTATGA +CTATGT +CTATGC +CTATGG +CTACAA +CTACAT +CTACAC +CTACAG +CTACTA +CTACTT +CTACTC +CTACTG +CTACCA +CTACCT +CTACCC +CTACCG +CTACGA +CTACGT +CTACGC +CTACGG +CTAGAA +CTAGAT +CTAGAC +CTAGAG +CTAGTA +CTAGTT +CTAGTC +CTAGTG +CTAGCA +CTAGCT +CTAGCC +CTAGCG +CTAGGA +CTAGGT +CTAGGC +CTAGGG +CTTAAA +CTTAAT +CTTAAC +CTTAAG +CTTATA +CTTATT +CTTATC +CTTATG +CTTACA +CTTACT +CTTACC +CTTACG +CTTAGA +CTTAGT +CTTAGC +CTTAGG +CTTTAA +CTTTAT +CTTTAC +CTTTAG +CTTTTA +CTTTTT +CTTTTC +CTTTTG +CTTTCA +CTTTCT +CTTTCC +CTTTCG +CTTTGA +CTTTGT +CTTTGC +CTTTGG +CTTCAA +CTTCAT +CTTCAC +CTTCAG +CTTCTA +CTTCTT +CTTCTC +CTTCTG +CTTCCA +CTTCCT +CTTCCC +CTTCCG +CTTCGA +CTTCGT +CTTCGC +CTTCGG +CTTGAA +CTTGAT +CTTGAC +CTTGAG +CTTGTA +CTTGTT +CTTGTC +CTTGTG +CTTGCA +CTTGCT +CTTGCC +CTTGCG +CTTGGA +CTTGGT +CTTGGC +CTTGGG +CTCAAA +CTCAAT +CTCAAC +CTCAAG +CTCATA +CTCATT +CTCATC +CTCATG +CTCACA +CTCACT +CTCACC +CTCACG +CTCAGA +CTCAGT +CTCAGC +CTCAGG +CTCTAA +CTCTAT +CTCTAC +CTCTAG +CTCTTA +CTCTTT +CTCTTC +CTCTTG +CTCTCA +CTCTCT +CTCTCC +CTCTCG +CTCTGA +CTCTGT +CTCTGC +CTCTGG +CTCCAA +CTCCAT +CTCCAC +CTCCAG +CTCCTA +CTCCTT +CTCCTC +CTCCTG +CTCCCA +CTCCCT +CTCCCC +CTCCCG +CTCCGA +CTCCGT +CTCCGC +CTCCGG +CTCGAA +CTCGAT +CTCGAC +CTCGAG +CTCGTA +CTCGTT +CTCGTC +CTCGTG +CTCGCA +CTCGCT +CTCGCC +CTCGCG +CTCGGA +CTCGGT +CTCGGC +CTCGGG +CTGAAA +CTGAAT +CTGAAC +CTGAAG +CTGATA +CTGATT +CTGATC +CTGATG +CTGACA +CTGACT +CTGACC +CTGACG +CTGAGA +CTGAGT +CTGAGC +CTGAGG +CTGTAA +CTGTAT +CTGTAC +CTGTAG +CTGTTA +CTGTTT +CTGTTC +CTGTTG +CTGTCA +CTGTCT +CTGTCC +CTGTCG +CTGTGA +CTGTGT +CTGTGC +CTGTGG +CTGCAA +CTGCAT +CTGCAC +CTGCAG +CTGCTA +CTGCTT +CTGCTC +CTGCTG +CTGCCA +CTGCCT +CTGCCC +CTGCCG +CTGCGA +CTGCGT +CTGCGC +CTGCGG +CTGGAA +CTGGAT +CTGGAC +CTGGAG +CTGGTA +CTGGTT +CTGGTC +CTGGTG +CTGGCA +CTGGCT +CTGGCC +CTGGCG +CTGGGA +CTGGGT +CTGGGC +CTGGGG +CCAAAA +CCAAAT +CCAAAC +CCAAAG +CCAATA +CCAATT +CCAATC +CCAATG +CCAACA +CCAACT +CCAACC +CCAACG +CCAAGA +CCAAGT +CCAAGC +CCAAGG +CCATAA +CCATAT +CCATAC +CCATAG +CCATTA +CCATTT +CCATTC +CCATTG +CCATCA +CCATCT +CCATCC +CCATCG +CCATGA +CCATGT +CCATGC +CCATGG +CCACAA +CCACAT +CCACAC +CCACAG +CCACTA +CCACTT +CCACTC +CCACTG +CCACCA +CCACCT +CCACCC +CCACCG +CCACGA +CCACGT +CCACGC +CCACGG +CCAGAA +CCAGAT +CCAGAC +CCAGAG +CCAGTA +CCAGTT +CCAGTC +CCAGTG +CCAGCA +CCAGCT +CCAGCC +CCAGCG +CCAGGA +CCAGGT +CCAGGC +CCAGGG +CCTAAA +CCTAAT +CCTAAC +CCTAAG +CCTATA +CCTATT +CCTATC +CCTATG +CCTACA +CCTACT +CCTACC +CCTACG +CCTAGA +CCTAGT +CCTAGC +CCTAGG +CCTTAA +CCTTAT +CCTTAC +CCTTAG +CCTTTA +CCTTTT +CCTTTC +CCTTTG +CCTTCA +CCTTCT +CCTTCC +CCTTCG +CCTTGA +CCTTGT +CCTTGC +CCTTGG +CCTCAA +CCTCAT +CCTCAC +CCTCAG +CCTCTA +CCTCTT +CCTCTC +CCTCTG +CCTCCA +CCTCCT +CCTCCC +CCTCCG +CCTCGA +CCTCGT +CCTCGC +CCTCGG +CCTGAA +CCTGAT +CCTGAC +CCTGAG +CCTGTA +CCTGTT +CCTGTC +CCTGTG +CCTGCA +CCTGCT +CCTGCC +CCTGCG +CCTGGA +CCTGGT +CCTGGC +CCTGGG +CCCAAA +CCCAAT +CCCAAC +CCCAAG +CCCATA +CCCATT +CCCATC +CCCATG +CCCACA +CCCACT +CCCACC +CCCACG +CCCAGA +CCCAGT +CCCAGC +CCCAGG +CCCTAA +CCCTAT +CCCTAC +CCCTAG +CCCTTA +CCCTTT +CCCTTC +CCCTTG +CCCTCA +CCCTCT +CCCTCC +CCCTCG +CCCTGA +CCCTGT +CCCTGC +CCCTGG +CCCCAA +CCCCAT +CCCCAC +CCCCAG +CCCCTA +CCCCTT +CCCCTC +CCCCTG +CCCCCA +CCCCCT +CCCCCC +CCCCCG +CCCCGA +CCCCGT +CCCCGC +CCCCGG +CCCGAA +CCCGAT +CCCGAC +CCCGAG +CCCGTA +CCCGTT +CCCGTC +CCCGTG +CCCGCA +CCCGCT +CCCGCC +CCCGCG +CCCGGA +CCCGGT +CCCGGC +CCCGGG +CCGAAA +CCGAAT +CCGAAC +CCGAAG +CCGATA +CCGATT +CCGATC +CCGATG +CCGACA +CCGACT +CCGACC +CCGACG +CCGAGA +CCGAGT +CCGAGC +CCGAGG +CCGTAA +CCGTAT +CCGTAC +CCGTAG +CCGTTA +CCGTTT +CCGTTC +CCGTTG +CCGTCA +CCGTCT +CCGTCC +CCGTCG +CCGTGA +CCGTGT +CCGTGC +CCGTGG +CCGCAA +CCGCAT +CCGCAC +CCGCAG +CCGCTA +CCGCTT +CCGCTC +CCGCTG +CCGCCA +CCGCCT +CCGCCC +CCGCCG +CCGCGA +CCGCGT +CCGCGC +CCGCGG +CCGGAA +CCGGAT +CCGGAC +CCGGAG +CCGGTA +CCGGTT +CCGGTC +CCGGTG +CCGGCA +CCGGCT +CCGGCC +CCGGCG +CCGGGA +CCGGGT +CCGGGC +CCGGGG +CGAAAA +CGAAAT +CGAAAC +CGAAAG +CGAATA +CGAATT +CGAATC +CGAATG +CGAACA +CGAACT +CGAACC +CGAACG +CGAAGA +CGAAGT +CGAAGC +CGAAGG +CGATAA +CGATAT +CGATAC +CGATAG +CGATTA +CGATTT +CGATTC +CGATTG +CGATCA +CGATCT +CGATCC +CGATCG +CGATGA +CGATGT +CGATGC +CGATGG +CGACAA +CGACAT +CGACAC +CGACAG +CGACTA +CGACTT +CGACTC +CGACTG +CGACCA +CGACCT +CGACCC +CGACCG +CGACGA +CGACGT +CGACGC +CGACGG +CGAGAA +CGAGAT +CGAGAC +CGAGAG +CGAGTA +CGAGTT +CGAGTC +CGAGTG +CGAGCA +CGAGCT +CGAGCC +CGAGCG +CGAGGA +CGAGGT +CGAGGC +CGAGGG +CGTAAA +CGTAAT +CGTAAC +CGTAAG +CGTATA +CGTATT +CGTATC +CGTATG +CGTACA +CGTACT +CGTACC +CGTACG +CGTAGA +CGTAGT +CGTAGC +CGTAGG +CGTTAA +CGTTAT +CGTTAC +CGTTAG +CGTTTA +CGTTTT +CGTTTC +CGTTTG +CGTTCA +CGTTCT +CGTTCC +CGTTCG +CGTTGA +CGTTGT +CGTTGC +CGTTGG +CGTCAA +CGTCAT +CGTCAC +CGTCAG +CGTCTA +CGTCTT +CGTCTC +CGTCTG +CGTCCA +CGTCCT +CGTCCC +CGTCCG +CGTCGA +CGTCGT +CGTCGC +CGTCGG +CGTGAA +CGTGAT +CGTGAC +CGTGAG +CGTGTA +CGTGTT +CGTGTC +CGTGTG +CGTGCA +CGTGCT +CGTGCC +CGTGCG +CGTGGA +CGTGGT +CGTGGC +CGTGGG +CGCAAA +CGCAAT +CGCAAC +CGCAAG +CGCATA +CGCATT +CGCATC +CGCATG +CGCACA +CGCACT +CGCACC +CGCACG +CGCAGA +CGCAGT +CGCAGC +CGCAGG +CGCTAA +CGCTAT +CGCTAC +CGCTAG +CGCTTA +CGCTTT +CGCTTC +CGCTTG +CGCTCA +CGCTCT +CGCTCC +CGCTCG +CGCTGA +CGCTGT +CGCTGC +CGCTGG +CGCCAA +CGCCAT +CGCCAC +CGCCAG +CGCCTA +CGCCTT +CGCCTC +CGCCTG +CGCCCA +CGCCCT +CGCCCC +CGCCCG +CGCCGA +CGCCGT +CGCCGC +CGCCGG +CGCGAA +CGCGAT +CGCGAC +CGCGAG +CGCGTA +CGCGTT +CGCGTC +CGCGTG +CGCGCA +CGCGCT +CGCGCC +CGCGCG +CGCGGA +CGCGGT +CGCGGC +CGCGGG +CGGAAA +CGGAAT +CGGAAC +CGGAAG +CGGATA +CGGATT +CGGATC +CGGATG +CGGACA +CGGACT +CGGACC +CGGACG +CGGAGA +CGGAGT +CGGAGC +CGGAGG +CGGTAA +CGGTAT +CGGTAC +CGGTAG +CGGTTA +CGGTTT +CGGTTC +CGGTTG +CGGTCA +CGGTCT +CGGTCC +CGGTCG +CGGTGA +CGGTGT +CGGTGC +CGGTGG +CGGCAA +CGGCAT +CGGCAC +CGGCAG +CGGCTA +CGGCTT +CGGCTC +CGGCTG +CGGCCA +CGGCCT +CGGCCC +CGGCCG +CGGCGA +CGGCGT +CGGCGC +CGGCGG +CGGGAA +CGGGAT +CGGGAC +CGGGAG +CGGGTA +CGGGTT +CGGGTC +CGGGTG +CGGGCA +CGGGCT +CGGGCC +CGGGCG +CGGGGA +CGGGGT +CGGGGC +CGGGGG +GAAAAA +GAAAAT +GAAAAC +GAAAAG +GAAATA +GAAATT +GAAATC +GAAATG +GAAACA +GAAACT +GAAACC +GAAACG +GAAAGA +GAAAGT +GAAAGC +GAAAGG +GAATAA +GAATAT +GAATAC +GAATAG +GAATTA +GAATTT +GAATTC +GAATTG +GAATCA +GAATCT +GAATCC +GAATCG +GAATGA +GAATGT +GAATGC +GAATGG +GAACAA +GAACAT +GAACAC +GAACAG +GAACTA +GAACTT +GAACTC +GAACTG +GAACCA +GAACCT +GAACCC +GAACCG +GAACGA +GAACGT +GAACGC +GAACGG +GAAGAA +GAAGAT +GAAGAC +GAAGAG +GAAGTA +GAAGTT +GAAGTC +GAAGTG +GAAGCA +GAAGCT +GAAGCC +GAAGCG +GAAGGA +GAAGGT +GAAGGC +GAAGGG +GATAAA +GATAAT +GATAAC +GATAAG +GATATA +GATATT +GATATC +GATATG +GATACA +GATACT +GATACC +GATACG +GATAGA +GATAGT +GATAGC +GATAGG +GATTAA +GATTAT +GATTAC +GATTAG +GATTTA +GATTTT +GATTTC +GATTTG +GATTCA +GATTCT +GATTCC +GATTCG +GATTGA +GATTGT +GATTGC +GATTGG +GATCAA +GATCAT +GATCAC +GATCAG +GATCTA +GATCTT +GATCTC +GATCTG +GATCCA +GATCCT +GATCCC +GATCCG +GATCGA +GATCGT +GATCGC +GATCGG +GATGAA +GATGAT +GATGAC +GATGAG +GATGTA +GATGTT +GATGTC +GATGTG +GATGCA +GATGCT +GATGCC +GATGCG +GATGGA +GATGGT +GATGGC +GATGGG +GACAAA +GACAAT +GACAAC +GACAAG +GACATA +GACATT +GACATC +GACATG +GACACA +GACACT +GACACC +GACACG +GACAGA +GACAGT +GACAGC +GACAGG +GACTAA +GACTAT +GACTAC +GACTAG +GACTTA +GACTTT +GACTTC +GACTTG +GACTCA +GACTCT +GACTCC +GACTCG +GACTGA +GACTGT +GACTGC +GACTGG +GACCAA +GACCAT +GACCAC +GACCAG +GACCTA +GACCTT +GACCTC +GACCTG +GACCCA +GACCCT +GACCCC +GACCCG +GACCGA +GACCGT +GACCGC +GACCGG +GACGAA +GACGAT +GACGAC +GACGAG +GACGTA +GACGTT +GACGTC +GACGTG +GACGCA +GACGCT +GACGCC +GACGCG +GACGGA +GACGGT +GACGGC +GACGGG +GAGAAA +GAGAAT +GAGAAC +GAGAAG +GAGATA +GAGATT +GAGATC +GAGATG +GAGACA +GAGACT +GAGACC +GAGACG +GAGAGA +GAGAGT +GAGAGC +GAGAGG +GAGTAA +GAGTAT +GAGTAC +GAGTAG +GAGTTA +GAGTTT +GAGTTC +GAGTTG +GAGTCA +GAGTCT +GAGTCC +GAGTCG +GAGTGA +GAGTGT +GAGTGC +GAGTGG +GAGCAA +GAGCAT +GAGCAC +GAGCAG +GAGCTA +GAGCTT +GAGCTC +GAGCTG +GAGCCA +GAGCCT +GAGCCC +GAGCCG +GAGCGA +GAGCGT +GAGCGC +GAGCGG +GAGGAA +GAGGAT +GAGGAC +GAGGAG +GAGGTA +GAGGTT +GAGGTC +GAGGTG +GAGGCA +GAGGCT +GAGGCC +GAGGCG +GAGGGA +GAGGGT +GAGGGC +GAGGGG +GTAAAA +GTAAAT +GTAAAC +GTAAAG +GTAATA +GTAATT +GTAATC +GTAATG +GTAACA +GTAACT +GTAACC +GTAACG +GTAAGA +GTAAGT +GTAAGC +GTAAGG +GTATAA +GTATAT +GTATAC +GTATAG +GTATTA +GTATTT +GTATTC +GTATTG +GTATCA +GTATCT +GTATCC +GTATCG +GTATGA +GTATGT +GTATGC +GTATGG +GTACAA +GTACAT +GTACAC +GTACAG +GTACTA +GTACTT +GTACTC +GTACTG +GTACCA +GTACCT +GTACCC +GTACCG +GTACGA +GTACGT +GTACGC +GTACGG +GTAGAA +GTAGAT +GTAGAC +GTAGAG +GTAGTA +GTAGTT +GTAGTC +GTAGTG +GTAGCA +GTAGCT +GTAGCC +GTAGCG +GTAGGA +GTAGGT +GTAGGC +GTAGGG +GTTAAA +GTTAAT +GTTAAC +GTTAAG +GTTATA +GTTATT +GTTATC +GTTATG +GTTACA +GTTACT +GTTACC +GTTACG +GTTAGA +GTTAGT +GTTAGC +GTTAGG +GTTTAA +GTTTAT +GTTTAC +GTTTAG +GTTTTA +GTTTTT +GTTTTC +GTTTTG +GTTTCA +GTTTCT +GTTTCC +GTTTCG +GTTTGA +GTTTGT +GTTTGC +GTTTGG +GTTCAA +GTTCAT +GTTCAC +GTTCAG +GTTCTA +GTTCTT +GTTCTC +GTTCTG +GTTCCA +GTTCCT +GTTCCC +GTTCCG +GTTCGA +GTTCGT +GTTCGC +GTTCGG +GTTGAA +GTTGAT +GTTGAC +GTTGAG +GTTGTA +GTTGTT +GTTGTC +GTTGTG +GTTGCA +GTTGCT +GTTGCC +GTTGCG +GTTGGA +GTTGGT +GTTGGC +GTTGGG +GTCAAA +GTCAAT +GTCAAC +GTCAAG +GTCATA +GTCATT +GTCATC +GTCATG +GTCACA +GTCACT +GTCACC +GTCACG +GTCAGA +GTCAGT +GTCAGC +GTCAGG +GTCTAA +GTCTAT +GTCTAC +GTCTAG +GTCTTA +GTCTTT +GTCTTC +GTCTTG +GTCTCA +GTCTCT +GTCTCC +GTCTCG +GTCTGA +GTCTGT +GTCTGC +GTCTGG +GTCCAA +GTCCAT +GTCCAC +GTCCAG +GTCCTA +GTCCTT +GTCCTC +GTCCTG +GTCCCA +GTCCCT +GTCCCC +GTCCCG +GTCCGA +GTCCGT +GTCCGC +GTCCGG +GTCGAA +GTCGAT +GTCGAC +GTCGAG +GTCGTA +GTCGTT +GTCGTC +GTCGTG +GTCGCA +GTCGCT +GTCGCC +GTCGCG +GTCGGA +GTCGGT +GTCGGC +GTCGGG +GTGAAA +GTGAAT +GTGAAC +GTGAAG +GTGATA +GTGATT +GTGATC +GTGATG +GTGACA +GTGACT +GTGACC +GTGACG +GTGAGA +GTGAGT +GTGAGC +GTGAGG +GTGTAA +GTGTAT +GTGTAC +GTGTAG +GTGTTA +GTGTTT +GTGTTC +GTGTTG +GTGTCA +GTGTCT +GTGTCC +GTGTCG +GTGTGA +GTGTGT +GTGTGC +GTGTGG +GTGCAA +GTGCAT +GTGCAC +GTGCAG +GTGCTA +GTGCTT +GTGCTC +GTGCTG +GTGCCA +GTGCCT +GTGCCC +GTGCCG +GTGCGA +GTGCGT +GTGCGC +GTGCGG +GTGGAA +GTGGAT +GTGGAC +GTGGAG +GTGGTA +GTGGTT +GTGGTC +GTGGTG +GTGGCA +GTGGCT +GTGGCC +GTGGCG +GTGGGA +GTGGGT +GTGGGC +GTGGGG +GCAAAA +GCAAAT +GCAAAC +GCAAAG +GCAATA +GCAATT +GCAATC +GCAATG +GCAACA +GCAACT +GCAACC +GCAACG +GCAAGA +GCAAGT +GCAAGC +GCAAGG +GCATAA +GCATAT +GCATAC +GCATAG +GCATTA +GCATTT +GCATTC +GCATTG +GCATCA +GCATCT +GCATCC +GCATCG +GCATGA +GCATGT +GCATGC +GCATGG +GCACAA +GCACAT +GCACAC +GCACAG +GCACTA +GCACTT +GCACTC +GCACTG +GCACCA +GCACCT +GCACCC +GCACCG +GCACGA +GCACGT +GCACGC +GCACGG +GCAGAA +GCAGAT +GCAGAC +GCAGAG +GCAGTA +GCAGTT +GCAGTC +GCAGTG +GCAGCA +GCAGCT +GCAGCC +GCAGCG +GCAGGA +GCAGGT +GCAGGC +GCAGGG +GCTAAA +GCTAAT +GCTAAC +GCTAAG +GCTATA +GCTATT +GCTATC +GCTATG +GCTACA +GCTACT +GCTACC +GCTACG +GCTAGA +GCTAGT +GCTAGC +GCTAGG +GCTTAA +GCTTAT +GCTTAC +GCTTAG +GCTTTA +GCTTTT +GCTTTC +GCTTTG +GCTTCA +GCTTCT +GCTTCC +GCTTCG +GCTTGA +GCTTGT +GCTTGC +GCTTGG +GCTCAA +GCTCAT +GCTCAC +GCTCAG +GCTCTA +GCTCTT +GCTCTC +GCTCTG +GCTCCA +GCTCCT +GCTCCC +GCTCCG +GCTCGA +GCTCGT +GCTCGC +GCTCGG +GCTGAA +GCTGAT +GCTGAC +GCTGAG +GCTGTA +GCTGTT +GCTGTC +GCTGTG +GCTGCA +GCTGCT +GCTGCC +GCTGCG +GCTGGA +GCTGGT +GCTGGC +GCTGGG +GCCAAA +GCCAAT +GCCAAC +GCCAAG +GCCATA +GCCATT +GCCATC +GCCATG +GCCACA +GCCACT +GCCACC +GCCACG +GCCAGA +GCCAGT +GCCAGC +GCCAGG +GCCTAA +GCCTAT +GCCTAC +GCCTAG +GCCTTA +GCCTTT +GCCTTC +GCCTTG +GCCTCA +GCCTCT +GCCTCC +GCCTCG +GCCTGA +GCCTGT +GCCTGC +GCCTGG +GCCCAA +GCCCAT +GCCCAC +GCCCAG +GCCCTA +GCCCTT +GCCCTC +GCCCTG +GCCCCA +GCCCCT +GCCCCC +GCCCCG +GCCCGA +GCCCGT +GCCCGC +GCCCGG +GCCGAA +GCCGAT +GCCGAC +GCCGAG +GCCGTA +GCCGTT +GCCGTC +GCCGTG +GCCGCA +GCCGCT +GCCGCC +GCCGCG +GCCGGA +GCCGGT +GCCGGC +GCCGGG +GCGAAA +GCGAAT +GCGAAC +GCGAAG +GCGATA +GCGATT +GCGATC +GCGATG +GCGACA +GCGACT +GCGACC +GCGACG +GCGAGA +GCGAGT +GCGAGC +GCGAGG +GCGTAA +GCGTAT +GCGTAC +GCGTAG +GCGTTA +GCGTTT +GCGTTC +GCGTTG +GCGTCA +GCGTCT +GCGTCC +GCGTCG +GCGTGA +GCGTGT +GCGTGC +GCGTGG +GCGCAA +GCGCAT +GCGCAC +GCGCAG +GCGCTA +GCGCTT +GCGCTC +GCGCTG +GCGCCA +GCGCCT +GCGCCC +GCGCCG +GCGCGA +GCGCGT +GCGCGC +GCGCGG +GCGGAA +GCGGAT +GCGGAC +GCGGAG +GCGGTA +GCGGTT +GCGGTC +GCGGTG +GCGGCA +GCGGCT +GCGGCC +GCGGCG +GCGGGA +GCGGGT +GCGGGC +GCGGGG +GGAAAA +GGAAAT +GGAAAC +GGAAAG +GGAATA +GGAATT +GGAATC +GGAATG +GGAACA +GGAACT +GGAACC +GGAACG +GGAAGA +GGAAGT +GGAAGC +GGAAGG +GGATAA +GGATAT +GGATAC +GGATAG +GGATTA +GGATTT +GGATTC +GGATTG +GGATCA +GGATCT +GGATCC +GGATCG +GGATGA +GGATGT +GGATGC +GGATGG +GGACAA +GGACAT +GGACAC +GGACAG +GGACTA +GGACTT +GGACTC +GGACTG +GGACCA +GGACCT +GGACCC +GGACCG +GGACGA +GGACGT +GGACGC +GGACGG +GGAGAA +GGAGAT +GGAGAC +GGAGAG +GGAGTA +GGAGTT +GGAGTC +GGAGTG +GGAGCA +GGAGCT +GGAGCC +GGAGCG +GGAGGA +GGAGGT +GGAGGC +GGAGGG +GGTAAA +GGTAAT +GGTAAC +GGTAAG +GGTATA +GGTATT +GGTATC +GGTATG +GGTACA +GGTACT +GGTACC +GGTACG +GGTAGA +GGTAGT +GGTAGC +GGTAGG +GGTTAA +GGTTAT +GGTTAC +GGTTAG +GGTTTA +GGTTTT +GGTTTC +GGTTTG +GGTTCA +GGTTCT +GGTTCC +GGTTCG +GGTTGA +GGTTGT +GGTTGC +GGTTGG +GGTCAA +GGTCAT +GGTCAC +GGTCAG +GGTCTA +GGTCTT +GGTCTC +GGTCTG +GGTCCA +GGTCCT +GGTCCC +GGTCCG +GGTCGA +GGTCGT +GGTCGC +GGTCGG +GGTGAA +GGTGAT +GGTGAC +GGTGAG +GGTGTA +GGTGTT +GGTGTC +GGTGTG +GGTGCA +GGTGCT +GGTGCC +GGTGCG +GGTGGA +GGTGGT +GGTGGC +GGTGGG +GGCAAA +GGCAAT +GGCAAC +GGCAAG +GGCATA +GGCATT +GGCATC +GGCATG +GGCACA +GGCACT +GGCACC +GGCACG +GGCAGA +GGCAGT +GGCAGC +GGCAGG +GGCTAA +GGCTAT +GGCTAC +GGCTAG +GGCTTA +GGCTTT +GGCTTC +GGCTTG +GGCTCA +GGCTCT +GGCTCC +GGCTCG +GGCTGA +GGCTGT +GGCTGC +GGCTGG +GGCCAA +GGCCAT +GGCCAC +GGCCAG +GGCCTA +GGCCTT +GGCCTC +GGCCTG +GGCCCA +GGCCCT +GGCCCC +GGCCCG +GGCCGA +GGCCGT +GGCCGC +GGCCGG +GGCGAA +GGCGAT +GGCGAC +GGCGAG +GGCGTA +GGCGTT +GGCGTC +GGCGTG +GGCGCA +GGCGCT +GGCGCC +GGCGCG +GGCGGA +GGCGGT +GGCGGC +GGCGGG +GGGAAA +GGGAAT +GGGAAC +GGGAAG +GGGATA +GGGATT +GGGATC +GGGATG +GGGACA +GGGACT +GGGACC +GGGACG +GGGAGA +GGGAGT +GGGAGC +GGGAGG +GGGTAA +GGGTAT +GGGTAC +GGGTAG +GGGTTA +GGGTTT +GGGTTC +GGGTTG +GGGTCA +GGGTCT +GGGTCC +GGGTCG +GGGTGA +GGGTGT +GGGTGC +GGGTGG +GGGCAA +GGGCAT +GGGCAC +GGGCAG +GGGCTA +GGGCTT +GGGCTC +GGGCTG +GGGCCA +GGGCCT +GGGCCC +GGGCCG +GGGCGA +GGGCGT +GGGCGC +GGGCGG +GGGGAA +GGGGAT +GGGGAC +GGGGAG +GGGGTA +GGGGTT +GGGGTC +GGGGTG +GGGGCA +GGGGCT +GGGGCC +GGGGCG +GGGGGA +GGGGGT +GGGGGC +GGGGGG diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-5000/config.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-5000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..9b14e61d327bf586fcff6e817cf9d365bae18ad2 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-5000/config.json @@ -0,0 +1,26 @@ +{ + "_name_or_path": "fabihamakhdoomi/TinyDNABERT", + "architectures": [ + "BertForSequenceClassification" + ], + "attention_probs_dropout_prob": 0.1, + "classifier_dropout": null, + "hidden_act": "gelu", + "hidden_dropout_prob": 0.1, + "hidden_size": 128, + "initializer_range": 0.02, + "intermediate_size": 512, + "layer_norm_eps": 1e-12, + "max_position_embeddings": 512, + "model_type": "bert", + "num_attention_heads": 2, + "num_hidden_layers": 2, + "pad_token_id": 0, + "position_embedding_type": "absolute", + "problem_type": "single_label_classification", + "torch_dtype": "float32", + "transformers_version": "4.44.0", + "type_vocab_size": 2, + "use_cache": true, + "vocab_size": 4101 +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-5000/model.safetensors b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-5000/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..8c6974fc5d837cb5358cb219c89755cdbefbf52b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-5000/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6a6f9f76e0215f8dc1065bfef289c54d32c716151680567be9575d5bec774ef5 +size 4021688 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-5000/optimizer.pt b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-5000/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..7e47905c0917183e4d6ddeb02389ce21efa68a0f --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-5000/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:16abbd5a2b31dc8e26fc714978e583848e82589758afb2bac11bbfb0be1a0aa1 +size 8058700 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-5000/rng_state.pth b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-5000/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..053218cb7d47f7e442d909ea8ed9bc9b286adfe8 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-5000/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:104246ae2bc9d8b31d442f0b88ec2d93b536dbfb9e32b9ff5076aee3a117c13d +size 14244 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-5000/scheduler.pt b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-5000/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..14aa799e6f5c93344d1b233b91d0bda9d8eb59e7 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-5000/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:eca4da5723e14faf16c3be96d23919390da5288d0d9db40021d332ee1df25b93 +size 1064 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-5000/special_tokens_map.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-5000/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..a8b3208c2884c4efb86e49300fdd3dc877220cdf --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-5000/special_tokens_map.json @@ -0,0 +1,7 @@ +{ + "cls_token": "[CLS]", + "mask_token": "[MASK]", + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-5000/tokenizer.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-5000/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..ebd782de6963b9032ce9374cfe30f5275be1a30b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-5000/tokenizer.json @@ -0,0 +1,4265 @@ +{ + "version": "1.0", + "truncation": { + "direction": "Right", + "max_length": 512, + "strategy": "LongestFirst", + "stride": 0 + }, + "padding": { + "strategy": { + "Fixed": 512 + }, + "direction": "Right", + "pad_to_multiple_of": null, + "pad_id": 0, + "pad_type_id": 0, + "pad_token": "[PAD]" + }, + "added_tokens": [ + { + "id": 0, + "content": "[PAD]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 1, + "content": "[UNK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 2, + "content": "[CLS]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 3, + "content": "[SEP]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 4, + "content": "[MASK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + } + ], + "normalizer": { + "type": "BertNormalizer", + "clean_text": true, + "handle_chinese_chars": true, + "strip_accents": null, + "lowercase": false + }, + "pre_tokenizer": { + "type": "BertPreTokenizer" + }, + "post_processor": { + "type": "TemplateProcessing", + "single": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + } + ], + "pair": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "B", + "type_id": 1 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 1 + } + } + ], + "special_tokens": { + "[CLS]": { + "id": "[CLS]", + "ids": [ + 2 + ], + "tokens": [ + "[CLS]" + ] + }, + "[SEP]": { + "id": "[SEP]", + "ids": [ + 3 + ], + "tokens": [ + "[SEP]" + ] + } + } + }, + "decoder": { + "type": "WordPiece", + "prefix": "##", + "cleanup": true + }, + "model": { + "type": "WordPiece", + "unk_token": "[UNK]", + "continuing_subword_prefix": "##", + "max_input_chars_per_word": 100, + "vocab": { + "[PAD]": 0, + "[UNK]": 1, + "[CLS]": 2, + "[SEP]": 3, + "[MASK]": 4, + "AAAAAA": 5, + "AAAAAT": 6, + "AAAAAC": 7, + "AAAAAG": 8, + "AAAATA": 9, + "AAAATT": 10, + "AAAATC": 11, + "AAAATG": 12, + "AAAACA": 13, + "AAAACT": 14, + "AAAACC": 15, + "AAAACG": 16, + "AAAAGA": 17, + "AAAAGT": 18, + "AAAAGC": 19, + "AAAAGG": 20, + "AAATAA": 21, + "AAATAT": 22, + "AAATAC": 23, + "AAATAG": 24, + "AAATTA": 25, + "AAATTT": 26, + "AAATTC": 27, + "AAATTG": 28, + "AAATCA": 29, + "AAATCT": 30, + "AAATCC": 31, + "AAATCG": 32, + "AAATGA": 33, + "AAATGT": 34, + "AAATGC": 35, + "AAATGG": 36, + "AAACAA": 37, + "AAACAT": 38, + "AAACAC": 39, + "AAACAG": 40, + "AAACTA": 41, + "AAACTT": 42, + "AAACTC": 43, + "AAACTG": 44, + "AAACCA": 45, + "AAACCT": 46, + "AAACCC": 47, + "AAACCG": 48, + "AAACGA": 49, + "AAACGT": 50, + "AAACGC": 51, + "AAACGG": 52, + "AAAGAA": 53, + "AAAGAT": 54, + "AAAGAC": 55, + "AAAGAG": 56, + "AAAGTA": 57, + "AAAGTT": 58, + "AAAGTC": 59, + "AAAGTG": 60, + "AAAGCA": 61, + "AAAGCT": 62, + "AAAGCC": 63, + "AAAGCG": 64, + "AAAGGA": 65, + "AAAGGT": 66, + "AAAGGC": 67, + "AAAGGG": 68, + "AATAAA": 69, + "AATAAT": 70, + "AATAAC": 71, + "AATAAG": 72, + "AATATA": 73, + "AATATT": 74, + "AATATC": 75, + "AATATG": 76, + "AATACA": 77, + "AATACT": 78, + "AATACC": 79, + "AATACG": 80, + "AATAGA": 81, + "AATAGT": 82, + "AATAGC": 83, + "AATAGG": 84, + "AATTAA": 85, + "AATTAT": 86, + "AATTAC": 87, + "AATTAG": 88, + "AATTTA": 89, + "AATTTT": 90, + "AATTTC": 91, + "AATTTG": 92, + "AATTCA": 93, + "AATTCT": 94, + "AATTCC": 95, + "AATTCG": 96, + "AATTGA": 97, + "AATTGT": 98, + "AATTGC": 99, + "AATTGG": 100, + "AATCAA": 101, + "AATCAT": 102, + "AATCAC": 103, + "AATCAG": 104, + "AATCTA": 105, + "AATCTT": 106, + "AATCTC": 107, + "AATCTG": 108, + "AATCCA": 109, + "AATCCT": 110, + "AATCCC": 111, + "AATCCG": 112, + "AATCGA": 113, + "AATCGT": 114, + "AATCGC": 115, + "AATCGG": 116, + "AATGAA": 117, + "AATGAT": 118, + "AATGAC": 119, + "AATGAG": 120, + "AATGTA": 121, + "AATGTT": 122, + "AATGTC": 123, + "AATGTG": 124, + "AATGCA": 125, + "AATGCT": 126, + "AATGCC": 127, + "AATGCG": 128, + "AATGGA": 129, + "AATGGT": 130, + "AATGGC": 131, + "AATGGG": 132, + "AACAAA": 133, + "AACAAT": 134, + "AACAAC": 135, + "AACAAG": 136, + "AACATA": 137, + "AACATT": 138, + "AACATC": 139, + "AACATG": 140, + "AACACA": 141, + "AACACT": 142, + "AACACC": 143, + "AACACG": 144, + "AACAGA": 145, + "AACAGT": 146, + "AACAGC": 147, + "AACAGG": 148, + "AACTAA": 149, + "AACTAT": 150, + "AACTAC": 151, + "AACTAG": 152, + "AACTTA": 153, + "AACTTT": 154, + "AACTTC": 155, + "AACTTG": 156, + "AACTCA": 157, + "AACTCT": 158, + "AACTCC": 159, + "AACTCG": 160, + "AACTGA": 161, + "AACTGT": 162, + "AACTGC": 163, + "AACTGG": 164, + "AACCAA": 165, + "AACCAT": 166, + "AACCAC": 167, + "AACCAG": 168, + "AACCTA": 169, + "AACCTT": 170, + "AACCTC": 171, + "AACCTG": 172, + "AACCCA": 173, + "AACCCT": 174, + "AACCCC": 175, + "AACCCG": 176, + "AACCGA": 177, + "AACCGT": 178, + "AACCGC": 179, + "AACCGG": 180, + "AACGAA": 181, + "AACGAT": 182, + "AACGAC": 183, + "AACGAG": 184, + "AACGTA": 185, + "AACGTT": 186, + "AACGTC": 187, + "AACGTG": 188, + "AACGCA": 189, + "AACGCT": 190, + "AACGCC": 191, + "AACGCG": 192, + "AACGGA": 193, + "AACGGT": 194, + "AACGGC": 195, + "AACGGG": 196, + "AAGAAA": 197, + "AAGAAT": 198, + "AAGAAC": 199, + "AAGAAG": 200, + "AAGATA": 201, + "AAGATT": 202, + "AAGATC": 203, + "AAGATG": 204, + "AAGACA": 205, + "AAGACT": 206, + "AAGACC": 207, + "AAGACG": 208, + "AAGAGA": 209, + "AAGAGT": 210, + "AAGAGC": 211, + "AAGAGG": 212, + "AAGTAA": 213, + "AAGTAT": 214, + "AAGTAC": 215, + "AAGTAG": 216, + "AAGTTA": 217, + "AAGTTT": 218, + "AAGTTC": 219, + "AAGTTG": 220, + "AAGTCA": 221, + "AAGTCT": 222, + "AAGTCC": 223, + "AAGTCG": 224, + "AAGTGA": 225, + "AAGTGT": 226, + "AAGTGC": 227, + "AAGTGG": 228, + "AAGCAA": 229, + "AAGCAT": 230, + "AAGCAC": 231, + "AAGCAG": 232, + "AAGCTA": 233, + "AAGCTT": 234, + "AAGCTC": 235, + "AAGCTG": 236, + "AAGCCA": 237, + "AAGCCT": 238, + "AAGCCC": 239, + "AAGCCG": 240, + "AAGCGA": 241, + "AAGCGT": 242, + "AAGCGC": 243, + "AAGCGG": 244, + "AAGGAA": 245, + "AAGGAT": 246, + "AAGGAC": 247, + "AAGGAG": 248, + "AAGGTA": 249, + "AAGGTT": 250, + "AAGGTC": 251, + "AAGGTG": 252, + "AAGGCA": 253, + "AAGGCT": 254, + "AAGGCC": 255, + "AAGGCG": 256, + "AAGGGA": 257, + "AAGGGT": 258, + "AAGGGC": 259, + "AAGGGG": 260, + "ATAAAA": 261, + "ATAAAT": 262, + "ATAAAC": 263, + "ATAAAG": 264, + "ATAATA": 265, + "ATAATT": 266, + "ATAATC": 267, + "ATAATG": 268, + "ATAACA": 269, + "ATAACT": 270, + "ATAACC": 271, + "ATAACG": 272, + "ATAAGA": 273, + "ATAAGT": 274, + "ATAAGC": 275, + "ATAAGG": 276, + "ATATAA": 277, + "ATATAT": 278, + "ATATAC": 279, + "ATATAG": 280, + "ATATTA": 281, + "ATATTT": 282, + "ATATTC": 283, + "ATATTG": 284, + "ATATCA": 285, + "ATATCT": 286, + "ATATCC": 287, + "ATATCG": 288, + "ATATGA": 289, + "ATATGT": 290, + "ATATGC": 291, + "ATATGG": 292, + "ATACAA": 293, + "ATACAT": 294, + "ATACAC": 295, + "ATACAG": 296, + "ATACTA": 297, + "ATACTT": 298, + "ATACTC": 299, + "ATACTG": 300, + "ATACCA": 301, + "ATACCT": 302, + "ATACCC": 303, + "ATACCG": 304, + "ATACGA": 305, + "ATACGT": 306, + "ATACGC": 307, + "ATACGG": 308, + "ATAGAA": 309, + "ATAGAT": 310, + "ATAGAC": 311, + "ATAGAG": 312, + "ATAGTA": 313, + "ATAGTT": 314, + "ATAGTC": 315, + "ATAGTG": 316, + "ATAGCA": 317, + "ATAGCT": 318, + "ATAGCC": 319, + "ATAGCG": 320, + "ATAGGA": 321, + "ATAGGT": 322, + "ATAGGC": 323, + "ATAGGG": 324, + "ATTAAA": 325, + "ATTAAT": 326, + "ATTAAC": 327, + "ATTAAG": 328, + "ATTATA": 329, + "ATTATT": 330, + "ATTATC": 331, + "ATTATG": 332, + "ATTACA": 333, + "ATTACT": 334, + "ATTACC": 335, + "ATTACG": 336, + "ATTAGA": 337, + "ATTAGT": 338, + "ATTAGC": 339, + "ATTAGG": 340, + "ATTTAA": 341, + "ATTTAT": 342, + "ATTTAC": 343, + "ATTTAG": 344, + "ATTTTA": 345, + "ATTTTT": 346, + "ATTTTC": 347, + "ATTTTG": 348, + "ATTTCA": 349, + "ATTTCT": 350, + "ATTTCC": 351, + "ATTTCG": 352, + "ATTTGA": 353, + "ATTTGT": 354, + "ATTTGC": 355, + "ATTTGG": 356, + "ATTCAA": 357, + "ATTCAT": 358, + "ATTCAC": 359, + "ATTCAG": 360, + "ATTCTA": 361, + "ATTCTT": 362, + "ATTCTC": 363, + "ATTCTG": 364, + "ATTCCA": 365, + "ATTCCT": 366, + "ATTCCC": 367, + "ATTCCG": 368, + "ATTCGA": 369, + "ATTCGT": 370, + "ATTCGC": 371, + "ATTCGG": 372, + "ATTGAA": 373, + "ATTGAT": 374, + "ATTGAC": 375, + "ATTGAG": 376, + "ATTGTA": 377, + "ATTGTT": 378, + "ATTGTC": 379, + "ATTGTG": 380, + "ATTGCA": 381, + "ATTGCT": 382, + "ATTGCC": 383, + "ATTGCG": 384, + "ATTGGA": 385, + "ATTGGT": 386, + "ATTGGC": 387, + "ATTGGG": 388, + "ATCAAA": 389, + "ATCAAT": 390, + "ATCAAC": 391, + "ATCAAG": 392, + "ATCATA": 393, + "ATCATT": 394, + "ATCATC": 395, + "ATCATG": 396, + "ATCACA": 397, + "ATCACT": 398, + "ATCACC": 399, + "ATCACG": 400, + "ATCAGA": 401, + "ATCAGT": 402, + "ATCAGC": 403, + "ATCAGG": 404, + "ATCTAA": 405, + "ATCTAT": 406, + "ATCTAC": 407, + "ATCTAG": 408, + "ATCTTA": 409, + "ATCTTT": 410, + "ATCTTC": 411, + "ATCTTG": 412, + "ATCTCA": 413, + "ATCTCT": 414, + "ATCTCC": 415, + "ATCTCG": 416, + "ATCTGA": 417, + "ATCTGT": 418, + "ATCTGC": 419, + "ATCTGG": 420, + "ATCCAA": 421, + "ATCCAT": 422, + "ATCCAC": 423, + "ATCCAG": 424, + "ATCCTA": 425, + "ATCCTT": 426, + "ATCCTC": 427, + "ATCCTG": 428, + "ATCCCA": 429, + "ATCCCT": 430, + "ATCCCC": 431, + "ATCCCG": 432, + "ATCCGA": 433, + "ATCCGT": 434, + "ATCCGC": 435, + "ATCCGG": 436, + "ATCGAA": 437, + "ATCGAT": 438, + "ATCGAC": 439, + "ATCGAG": 440, + "ATCGTA": 441, + "ATCGTT": 442, + "ATCGTC": 443, + "ATCGTG": 444, + "ATCGCA": 445, + "ATCGCT": 446, + "ATCGCC": 447, + "ATCGCG": 448, + "ATCGGA": 449, + "ATCGGT": 450, + "ATCGGC": 451, + "ATCGGG": 452, + "ATGAAA": 453, + "ATGAAT": 454, + "ATGAAC": 455, + "ATGAAG": 456, + "ATGATA": 457, + "ATGATT": 458, + "ATGATC": 459, + "ATGATG": 460, + "ATGACA": 461, + "ATGACT": 462, + "ATGACC": 463, + "ATGACG": 464, + "ATGAGA": 465, + "ATGAGT": 466, + "ATGAGC": 467, + "ATGAGG": 468, + "ATGTAA": 469, + "ATGTAT": 470, + "ATGTAC": 471, + "ATGTAG": 472, + "ATGTTA": 473, + "ATGTTT": 474, + "ATGTTC": 475, + "ATGTTG": 476, + "ATGTCA": 477, + "ATGTCT": 478, + "ATGTCC": 479, + "ATGTCG": 480, + "ATGTGA": 481, + "ATGTGT": 482, + "ATGTGC": 483, + "ATGTGG": 484, + "ATGCAA": 485, + "ATGCAT": 486, + "ATGCAC": 487, + "ATGCAG": 488, + "ATGCTA": 489, + "ATGCTT": 490, + "ATGCTC": 491, + "ATGCTG": 492, + "ATGCCA": 493, + "ATGCCT": 494, + "ATGCCC": 495, + "ATGCCG": 496, + "ATGCGA": 497, + "ATGCGT": 498, + "ATGCGC": 499, + "ATGCGG": 500, + "ATGGAA": 501, + "ATGGAT": 502, + "ATGGAC": 503, + "ATGGAG": 504, + "ATGGTA": 505, + "ATGGTT": 506, + "ATGGTC": 507, + "ATGGTG": 508, + "ATGGCA": 509, + "ATGGCT": 510, + "ATGGCC": 511, + "ATGGCG": 512, + "ATGGGA": 513, + "ATGGGT": 514, + "ATGGGC": 515, + "ATGGGG": 516, + "ACAAAA": 517, + "ACAAAT": 518, + "ACAAAC": 519, + "ACAAAG": 520, + "ACAATA": 521, + "ACAATT": 522, + "ACAATC": 523, + "ACAATG": 524, + "ACAACA": 525, + "ACAACT": 526, + "ACAACC": 527, + "ACAACG": 528, + "ACAAGA": 529, + "ACAAGT": 530, + "ACAAGC": 531, + "ACAAGG": 532, + "ACATAA": 533, + "ACATAT": 534, + "ACATAC": 535, + "ACATAG": 536, + "ACATTA": 537, + "ACATTT": 538, + "ACATTC": 539, + "ACATTG": 540, + "ACATCA": 541, + "ACATCT": 542, + "ACATCC": 543, + "ACATCG": 544, + "ACATGA": 545, + "ACATGT": 546, + "ACATGC": 547, + "ACATGG": 548, + "ACACAA": 549, + "ACACAT": 550, + "ACACAC": 551, + "ACACAG": 552, + "ACACTA": 553, + "ACACTT": 554, + "ACACTC": 555, + "ACACTG": 556, + "ACACCA": 557, + "ACACCT": 558, + "ACACCC": 559, + "ACACCG": 560, + "ACACGA": 561, + "ACACGT": 562, + "ACACGC": 563, + "ACACGG": 564, + "ACAGAA": 565, + "ACAGAT": 566, + "ACAGAC": 567, + "ACAGAG": 568, + "ACAGTA": 569, + "ACAGTT": 570, + "ACAGTC": 571, + "ACAGTG": 572, + "ACAGCA": 573, + "ACAGCT": 574, + "ACAGCC": 575, + "ACAGCG": 576, + "ACAGGA": 577, + "ACAGGT": 578, + "ACAGGC": 579, + "ACAGGG": 580, + "ACTAAA": 581, + "ACTAAT": 582, + "ACTAAC": 583, + "ACTAAG": 584, + "ACTATA": 585, + "ACTATT": 586, + "ACTATC": 587, + "ACTATG": 588, + "ACTACA": 589, + "ACTACT": 590, + "ACTACC": 591, + "ACTACG": 592, + "ACTAGA": 593, + "ACTAGT": 594, + "ACTAGC": 595, + "ACTAGG": 596, + "ACTTAA": 597, + "ACTTAT": 598, + "ACTTAC": 599, + "ACTTAG": 600, + "ACTTTA": 601, + "ACTTTT": 602, + "ACTTTC": 603, + "ACTTTG": 604, + "ACTTCA": 605, + "ACTTCT": 606, + "ACTTCC": 607, + "ACTTCG": 608, + "ACTTGA": 609, + "ACTTGT": 610, + "ACTTGC": 611, + "ACTTGG": 612, + "ACTCAA": 613, + "ACTCAT": 614, + "ACTCAC": 615, + "ACTCAG": 616, + "ACTCTA": 617, + "ACTCTT": 618, + "ACTCTC": 619, + "ACTCTG": 620, + "ACTCCA": 621, + "ACTCCT": 622, + "ACTCCC": 623, + "ACTCCG": 624, + "ACTCGA": 625, + "ACTCGT": 626, + "ACTCGC": 627, + "ACTCGG": 628, + "ACTGAA": 629, + "ACTGAT": 630, + "ACTGAC": 631, + "ACTGAG": 632, + "ACTGTA": 633, + "ACTGTT": 634, + "ACTGTC": 635, + "ACTGTG": 636, + "ACTGCA": 637, + "ACTGCT": 638, + "ACTGCC": 639, + "ACTGCG": 640, + "ACTGGA": 641, + "ACTGGT": 642, + "ACTGGC": 643, + "ACTGGG": 644, + "ACCAAA": 645, + "ACCAAT": 646, + "ACCAAC": 647, + "ACCAAG": 648, + "ACCATA": 649, + "ACCATT": 650, + "ACCATC": 651, + "ACCATG": 652, + "ACCACA": 653, + "ACCACT": 654, + "ACCACC": 655, + "ACCACG": 656, + "ACCAGA": 657, + "ACCAGT": 658, + "ACCAGC": 659, + "ACCAGG": 660, + "ACCTAA": 661, + "ACCTAT": 662, + "ACCTAC": 663, + "ACCTAG": 664, + "ACCTTA": 665, + "ACCTTT": 666, + "ACCTTC": 667, + "ACCTTG": 668, + "ACCTCA": 669, + "ACCTCT": 670, + "ACCTCC": 671, + "ACCTCG": 672, + "ACCTGA": 673, + "ACCTGT": 674, + "ACCTGC": 675, + "ACCTGG": 676, + "ACCCAA": 677, + "ACCCAT": 678, + "ACCCAC": 679, + "ACCCAG": 680, + "ACCCTA": 681, + "ACCCTT": 682, + "ACCCTC": 683, + "ACCCTG": 684, + "ACCCCA": 685, + "ACCCCT": 686, + "ACCCCC": 687, + "ACCCCG": 688, + "ACCCGA": 689, + "ACCCGT": 690, + "ACCCGC": 691, + "ACCCGG": 692, + "ACCGAA": 693, + "ACCGAT": 694, + "ACCGAC": 695, + "ACCGAG": 696, + "ACCGTA": 697, + "ACCGTT": 698, + "ACCGTC": 699, + "ACCGTG": 700, + "ACCGCA": 701, + "ACCGCT": 702, + "ACCGCC": 703, + "ACCGCG": 704, + "ACCGGA": 705, + "ACCGGT": 706, + "ACCGGC": 707, + "ACCGGG": 708, + "ACGAAA": 709, + "ACGAAT": 710, + "ACGAAC": 711, + "ACGAAG": 712, + "ACGATA": 713, + "ACGATT": 714, + "ACGATC": 715, + "ACGATG": 716, + "ACGACA": 717, + "ACGACT": 718, + "ACGACC": 719, + "ACGACG": 720, + "ACGAGA": 721, + "ACGAGT": 722, + "ACGAGC": 723, + "ACGAGG": 724, + "ACGTAA": 725, + "ACGTAT": 726, + "ACGTAC": 727, + "ACGTAG": 728, + "ACGTTA": 729, + "ACGTTT": 730, + "ACGTTC": 731, + "ACGTTG": 732, + "ACGTCA": 733, + "ACGTCT": 734, + "ACGTCC": 735, + "ACGTCG": 736, + "ACGTGA": 737, + "ACGTGT": 738, + "ACGTGC": 739, + "ACGTGG": 740, + "ACGCAA": 741, + "ACGCAT": 742, + "ACGCAC": 743, + "ACGCAG": 744, + "ACGCTA": 745, + "ACGCTT": 746, + "ACGCTC": 747, + "ACGCTG": 748, + "ACGCCA": 749, + "ACGCCT": 750, + "ACGCCC": 751, + "ACGCCG": 752, + "ACGCGA": 753, + "ACGCGT": 754, + "ACGCGC": 755, + "ACGCGG": 756, + "ACGGAA": 757, + "ACGGAT": 758, + "ACGGAC": 759, + "ACGGAG": 760, + "ACGGTA": 761, + "ACGGTT": 762, + "ACGGTC": 763, + "ACGGTG": 764, + "ACGGCA": 765, + "ACGGCT": 766, + "ACGGCC": 767, + "ACGGCG": 768, + "ACGGGA": 769, + "ACGGGT": 770, + "ACGGGC": 771, + "ACGGGG": 772, + "AGAAAA": 773, + "AGAAAT": 774, + "AGAAAC": 775, + "AGAAAG": 776, + "AGAATA": 777, + "AGAATT": 778, + "AGAATC": 779, + "AGAATG": 780, + "AGAACA": 781, + "AGAACT": 782, + "AGAACC": 783, + "AGAACG": 784, + "AGAAGA": 785, + "AGAAGT": 786, + "AGAAGC": 787, + "AGAAGG": 788, + "AGATAA": 789, + "AGATAT": 790, + "AGATAC": 791, + "AGATAG": 792, + "AGATTA": 793, + "AGATTT": 794, + "AGATTC": 795, + "AGATTG": 796, + "AGATCA": 797, + "AGATCT": 798, + "AGATCC": 799, + "AGATCG": 800, + "AGATGA": 801, + "AGATGT": 802, + "AGATGC": 803, + "AGATGG": 804, + "AGACAA": 805, + "AGACAT": 806, + "AGACAC": 807, + "AGACAG": 808, + "AGACTA": 809, + "AGACTT": 810, + "AGACTC": 811, + "AGACTG": 812, + "AGACCA": 813, + "AGACCT": 814, + "AGACCC": 815, + "AGACCG": 816, + "AGACGA": 817, + "AGACGT": 818, + "AGACGC": 819, + "AGACGG": 820, + "AGAGAA": 821, + "AGAGAT": 822, + "AGAGAC": 823, + "AGAGAG": 824, + "AGAGTA": 825, + "AGAGTT": 826, + "AGAGTC": 827, + "AGAGTG": 828, + "AGAGCA": 829, + "AGAGCT": 830, + "AGAGCC": 831, + "AGAGCG": 832, + "AGAGGA": 833, + "AGAGGT": 834, + "AGAGGC": 835, + "AGAGGG": 836, + "AGTAAA": 837, + "AGTAAT": 838, + "AGTAAC": 839, + "AGTAAG": 840, + "AGTATA": 841, + "AGTATT": 842, + "AGTATC": 843, + "AGTATG": 844, + "AGTACA": 845, + "AGTACT": 846, + "AGTACC": 847, + "AGTACG": 848, + "AGTAGA": 849, + "AGTAGT": 850, + "AGTAGC": 851, + "AGTAGG": 852, + "AGTTAA": 853, + "AGTTAT": 854, + "AGTTAC": 855, + "AGTTAG": 856, + "AGTTTA": 857, + "AGTTTT": 858, + "AGTTTC": 859, + "AGTTTG": 860, + "AGTTCA": 861, + "AGTTCT": 862, + "AGTTCC": 863, + "AGTTCG": 864, + "AGTTGA": 865, + "AGTTGT": 866, + "AGTTGC": 867, + "AGTTGG": 868, + "AGTCAA": 869, + "AGTCAT": 870, + "AGTCAC": 871, + "AGTCAG": 872, + "AGTCTA": 873, + "AGTCTT": 874, + "AGTCTC": 875, + "AGTCTG": 876, + "AGTCCA": 877, + "AGTCCT": 878, + "AGTCCC": 879, + "AGTCCG": 880, + "AGTCGA": 881, + "AGTCGT": 882, + "AGTCGC": 883, + "AGTCGG": 884, + "AGTGAA": 885, + "AGTGAT": 886, + "AGTGAC": 887, + "AGTGAG": 888, + "AGTGTA": 889, + "AGTGTT": 890, + "AGTGTC": 891, + "AGTGTG": 892, + "AGTGCA": 893, + "AGTGCT": 894, + "AGTGCC": 895, + "AGTGCG": 896, + "AGTGGA": 897, + "AGTGGT": 898, + "AGTGGC": 899, + "AGTGGG": 900, + "AGCAAA": 901, + "AGCAAT": 902, + "AGCAAC": 903, + "AGCAAG": 904, + "AGCATA": 905, + "AGCATT": 906, + "AGCATC": 907, + "AGCATG": 908, + "AGCACA": 909, + "AGCACT": 910, + "AGCACC": 911, + "AGCACG": 912, + "AGCAGA": 913, + "AGCAGT": 914, + "AGCAGC": 915, + "AGCAGG": 916, + "AGCTAA": 917, + "AGCTAT": 918, + "AGCTAC": 919, + "AGCTAG": 920, + "AGCTTA": 921, + "AGCTTT": 922, + "AGCTTC": 923, + "AGCTTG": 924, + "AGCTCA": 925, + "AGCTCT": 926, + "AGCTCC": 927, + "AGCTCG": 928, + "AGCTGA": 929, + "AGCTGT": 930, + "AGCTGC": 931, + "AGCTGG": 932, + "AGCCAA": 933, + "AGCCAT": 934, + "AGCCAC": 935, + "AGCCAG": 936, + "AGCCTA": 937, + "AGCCTT": 938, + "AGCCTC": 939, + "AGCCTG": 940, + "AGCCCA": 941, + "AGCCCT": 942, + "AGCCCC": 943, + "AGCCCG": 944, + "AGCCGA": 945, + "AGCCGT": 946, + "AGCCGC": 947, + "AGCCGG": 948, + "AGCGAA": 949, + "AGCGAT": 950, + "AGCGAC": 951, + "AGCGAG": 952, + "AGCGTA": 953, + "AGCGTT": 954, + "AGCGTC": 955, + "AGCGTG": 956, + "AGCGCA": 957, + "AGCGCT": 958, + "AGCGCC": 959, + "AGCGCG": 960, + "AGCGGA": 961, + "AGCGGT": 962, + "AGCGGC": 963, + "AGCGGG": 964, + "AGGAAA": 965, + "AGGAAT": 966, + "AGGAAC": 967, + "AGGAAG": 968, + "AGGATA": 969, + "AGGATT": 970, + "AGGATC": 971, + "AGGATG": 972, + "AGGACA": 973, + "AGGACT": 974, + "AGGACC": 975, + "AGGACG": 976, + "AGGAGA": 977, + "AGGAGT": 978, + "AGGAGC": 979, + "AGGAGG": 980, + "AGGTAA": 981, + "AGGTAT": 982, + "AGGTAC": 983, + "AGGTAG": 984, + "AGGTTA": 985, + "AGGTTT": 986, + "AGGTTC": 987, + "AGGTTG": 988, + "AGGTCA": 989, + "AGGTCT": 990, + "AGGTCC": 991, + "AGGTCG": 992, + "AGGTGA": 993, + "AGGTGT": 994, + "AGGTGC": 995, + "AGGTGG": 996, + "AGGCAA": 997, + "AGGCAT": 998, + "AGGCAC": 999, + "AGGCAG": 1000, + "AGGCTA": 1001, + "AGGCTT": 1002, + "AGGCTC": 1003, + "AGGCTG": 1004, + "AGGCCA": 1005, + "AGGCCT": 1006, + "AGGCCC": 1007, + "AGGCCG": 1008, + "AGGCGA": 1009, + "AGGCGT": 1010, + "AGGCGC": 1011, + "AGGCGG": 1012, + "AGGGAA": 1013, + "AGGGAT": 1014, + "AGGGAC": 1015, + "AGGGAG": 1016, + "AGGGTA": 1017, + "AGGGTT": 1018, + "AGGGTC": 1019, + "AGGGTG": 1020, + "AGGGCA": 1021, + "AGGGCT": 1022, + "AGGGCC": 1023, + "AGGGCG": 1024, + "AGGGGA": 1025, + "AGGGGT": 1026, + "AGGGGC": 1027, + "AGGGGG": 1028, + "TAAAAA": 1029, + "TAAAAT": 1030, + "TAAAAC": 1031, + "TAAAAG": 1032, + "TAAATA": 1033, + "TAAATT": 1034, + "TAAATC": 1035, + "TAAATG": 1036, + "TAAACA": 1037, + "TAAACT": 1038, + "TAAACC": 1039, + "TAAACG": 1040, + "TAAAGA": 1041, + "TAAAGT": 1042, + "TAAAGC": 1043, + "TAAAGG": 1044, + "TAATAA": 1045, + "TAATAT": 1046, + "TAATAC": 1047, + "TAATAG": 1048, + "TAATTA": 1049, + "TAATTT": 1050, + "TAATTC": 1051, + "TAATTG": 1052, + "TAATCA": 1053, + "TAATCT": 1054, + "TAATCC": 1055, + "TAATCG": 1056, + "TAATGA": 1057, + "TAATGT": 1058, + "TAATGC": 1059, + "TAATGG": 1060, + "TAACAA": 1061, + "TAACAT": 1062, + "TAACAC": 1063, + "TAACAG": 1064, + "TAACTA": 1065, + "TAACTT": 1066, + "TAACTC": 1067, + "TAACTG": 1068, + "TAACCA": 1069, + "TAACCT": 1070, + "TAACCC": 1071, + "TAACCG": 1072, + "TAACGA": 1073, + "TAACGT": 1074, + "TAACGC": 1075, + "TAACGG": 1076, + "TAAGAA": 1077, + "TAAGAT": 1078, + "TAAGAC": 1079, + "TAAGAG": 1080, + "TAAGTA": 1081, + "TAAGTT": 1082, + "TAAGTC": 1083, + "TAAGTG": 1084, + "TAAGCA": 1085, + "TAAGCT": 1086, + "TAAGCC": 1087, + "TAAGCG": 1088, + "TAAGGA": 1089, + "TAAGGT": 1090, + "TAAGGC": 1091, + "TAAGGG": 1092, + "TATAAA": 1093, + "TATAAT": 1094, + "TATAAC": 1095, + "TATAAG": 1096, + "TATATA": 1097, + "TATATT": 1098, + "TATATC": 1099, + "TATATG": 1100, + "TATACA": 1101, + "TATACT": 1102, + "TATACC": 1103, + "TATACG": 1104, + "TATAGA": 1105, + "TATAGT": 1106, + "TATAGC": 1107, + "TATAGG": 1108, + "TATTAA": 1109, + "TATTAT": 1110, + "TATTAC": 1111, + "TATTAG": 1112, + "TATTTA": 1113, + "TATTTT": 1114, + "TATTTC": 1115, + "TATTTG": 1116, + "TATTCA": 1117, + "TATTCT": 1118, + "TATTCC": 1119, + "TATTCG": 1120, + "TATTGA": 1121, + "TATTGT": 1122, + "TATTGC": 1123, + "TATTGG": 1124, + "TATCAA": 1125, + "TATCAT": 1126, + "TATCAC": 1127, + "TATCAG": 1128, + "TATCTA": 1129, + "TATCTT": 1130, + "TATCTC": 1131, + "TATCTG": 1132, + "TATCCA": 1133, + "TATCCT": 1134, + "TATCCC": 1135, + "TATCCG": 1136, + "TATCGA": 1137, + "TATCGT": 1138, + "TATCGC": 1139, + "TATCGG": 1140, + "TATGAA": 1141, + "TATGAT": 1142, + "TATGAC": 1143, + "TATGAG": 1144, + "TATGTA": 1145, + "TATGTT": 1146, + "TATGTC": 1147, + "TATGTG": 1148, + "TATGCA": 1149, + "TATGCT": 1150, + "TATGCC": 1151, + "TATGCG": 1152, + "TATGGA": 1153, + "TATGGT": 1154, + "TATGGC": 1155, + "TATGGG": 1156, + "TACAAA": 1157, + "TACAAT": 1158, + "TACAAC": 1159, + "TACAAG": 1160, + "TACATA": 1161, + "TACATT": 1162, + "TACATC": 1163, + "TACATG": 1164, + "TACACA": 1165, + "TACACT": 1166, + "TACACC": 1167, + "TACACG": 1168, + "TACAGA": 1169, + "TACAGT": 1170, + "TACAGC": 1171, + "TACAGG": 1172, + "TACTAA": 1173, + "TACTAT": 1174, + "TACTAC": 1175, + "TACTAG": 1176, + "TACTTA": 1177, + "TACTTT": 1178, + "TACTTC": 1179, + "TACTTG": 1180, + "TACTCA": 1181, + "TACTCT": 1182, + "TACTCC": 1183, + "TACTCG": 1184, + "TACTGA": 1185, + "TACTGT": 1186, + "TACTGC": 1187, + "TACTGG": 1188, + "TACCAA": 1189, + "TACCAT": 1190, + "TACCAC": 1191, + "TACCAG": 1192, + "TACCTA": 1193, + "TACCTT": 1194, + "TACCTC": 1195, + "TACCTG": 1196, + "TACCCA": 1197, + "TACCCT": 1198, + "TACCCC": 1199, + "TACCCG": 1200, + "TACCGA": 1201, + "TACCGT": 1202, + "TACCGC": 1203, + "TACCGG": 1204, + "TACGAA": 1205, + "TACGAT": 1206, + "TACGAC": 1207, + "TACGAG": 1208, + "TACGTA": 1209, + "TACGTT": 1210, + "TACGTC": 1211, + "TACGTG": 1212, + "TACGCA": 1213, + "TACGCT": 1214, + "TACGCC": 1215, + "TACGCG": 1216, + "TACGGA": 1217, + "TACGGT": 1218, + "TACGGC": 1219, + "TACGGG": 1220, + "TAGAAA": 1221, + "TAGAAT": 1222, + "TAGAAC": 1223, + "TAGAAG": 1224, + "TAGATA": 1225, + "TAGATT": 1226, + "TAGATC": 1227, + "TAGATG": 1228, + "TAGACA": 1229, + "TAGACT": 1230, + "TAGACC": 1231, + "TAGACG": 1232, + "TAGAGA": 1233, + "TAGAGT": 1234, + "TAGAGC": 1235, + "TAGAGG": 1236, + "TAGTAA": 1237, + "TAGTAT": 1238, + "TAGTAC": 1239, + "TAGTAG": 1240, + "TAGTTA": 1241, + "TAGTTT": 1242, + "TAGTTC": 1243, + "TAGTTG": 1244, + "TAGTCA": 1245, + "TAGTCT": 1246, + "TAGTCC": 1247, + "TAGTCG": 1248, + "TAGTGA": 1249, + "TAGTGT": 1250, + "TAGTGC": 1251, + "TAGTGG": 1252, + "TAGCAA": 1253, + "TAGCAT": 1254, + "TAGCAC": 1255, + "TAGCAG": 1256, + "TAGCTA": 1257, + "TAGCTT": 1258, + "TAGCTC": 1259, + "TAGCTG": 1260, + "TAGCCA": 1261, + "TAGCCT": 1262, + "TAGCCC": 1263, + "TAGCCG": 1264, + "TAGCGA": 1265, + "TAGCGT": 1266, + "TAGCGC": 1267, + "TAGCGG": 1268, + "TAGGAA": 1269, + "TAGGAT": 1270, + "TAGGAC": 1271, + "TAGGAG": 1272, + "TAGGTA": 1273, + "TAGGTT": 1274, + "TAGGTC": 1275, + "TAGGTG": 1276, + "TAGGCA": 1277, + "TAGGCT": 1278, + "TAGGCC": 1279, + "TAGGCG": 1280, + "TAGGGA": 1281, + "TAGGGT": 1282, + "TAGGGC": 1283, + "TAGGGG": 1284, + "TTAAAA": 1285, + "TTAAAT": 1286, + "TTAAAC": 1287, + "TTAAAG": 1288, + "TTAATA": 1289, + "TTAATT": 1290, + "TTAATC": 1291, + "TTAATG": 1292, + "TTAACA": 1293, + "TTAACT": 1294, + "TTAACC": 1295, + "TTAACG": 1296, + "TTAAGA": 1297, + "TTAAGT": 1298, + "TTAAGC": 1299, + "TTAAGG": 1300, + "TTATAA": 1301, + "TTATAT": 1302, + "TTATAC": 1303, + "TTATAG": 1304, + "TTATTA": 1305, + "TTATTT": 1306, + "TTATTC": 1307, + "TTATTG": 1308, + "TTATCA": 1309, + "TTATCT": 1310, + "TTATCC": 1311, + "TTATCG": 1312, + "TTATGA": 1313, + "TTATGT": 1314, + "TTATGC": 1315, + "TTATGG": 1316, + "TTACAA": 1317, + "TTACAT": 1318, + "TTACAC": 1319, + "TTACAG": 1320, + "TTACTA": 1321, + "TTACTT": 1322, + "TTACTC": 1323, + "TTACTG": 1324, + "TTACCA": 1325, + "TTACCT": 1326, + "TTACCC": 1327, + "TTACCG": 1328, + "TTACGA": 1329, + "TTACGT": 1330, + "TTACGC": 1331, + "TTACGG": 1332, + "TTAGAA": 1333, + "TTAGAT": 1334, + "TTAGAC": 1335, + "TTAGAG": 1336, + "TTAGTA": 1337, + "TTAGTT": 1338, + "TTAGTC": 1339, + "TTAGTG": 1340, + "TTAGCA": 1341, + "TTAGCT": 1342, + "TTAGCC": 1343, + "TTAGCG": 1344, + "TTAGGA": 1345, + "TTAGGT": 1346, + "TTAGGC": 1347, + "TTAGGG": 1348, + "TTTAAA": 1349, + "TTTAAT": 1350, + "TTTAAC": 1351, + "TTTAAG": 1352, + "TTTATA": 1353, + "TTTATT": 1354, + "TTTATC": 1355, + "TTTATG": 1356, + "TTTACA": 1357, + "TTTACT": 1358, + "TTTACC": 1359, + "TTTACG": 1360, + "TTTAGA": 1361, + "TTTAGT": 1362, + "TTTAGC": 1363, + "TTTAGG": 1364, + "TTTTAA": 1365, + "TTTTAT": 1366, + "TTTTAC": 1367, + "TTTTAG": 1368, + "TTTTTA": 1369, + "TTTTTT": 1370, + "TTTTTC": 1371, + "TTTTTG": 1372, + "TTTTCA": 1373, + "TTTTCT": 1374, + "TTTTCC": 1375, + "TTTTCG": 1376, + "TTTTGA": 1377, + "TTTTGT": 1378, + "TTTTGC": 1379, + "TTTTGG": 1380, + "TTTCAA": 1381, + "TTTCAT": 1382, + "TTTCAC": 1383, + "TTTCAG": 1384, + "TTTCTA": 1385, + "TTTCTT": 1386, + "TTTCTC": 1387, + "TTTCTG": 1388, + "TTTCCA": 1389, + "TTTCCT": 1390, + "TTTCCC": 1391, + "TTTCCG": 1392, + "TTTCGA": 1393, + "TTTCGT": 1394, + "TTTCGC": 1395, + "TTTCGG": 1396, + "TTTGAA": 1397, + "TTTGAT": 1398, + "TTTGAC": 1399, + "TTTGAG": 1400, + "TTTGTA": 1401, + "TTTGTT": 1402, + "TTTGTC": 1403, + "TTTGTG": 1404, + "TTTGCA": 1405, + "TTTGCT": 1406, + "TTTGCC": 1407, + "TTTGCG": 1408, + "TTTGGA": 1409, + "TTTGGT": 1410, + "TTTGGC": 1411, + "TTTGGG": 1412, + "TTCAAA": 1413, + "TTCAAT": 1414, + "TTCAAC": 1415, + "TTCAAG": 1416, + "TTCATA": 1417, + "TTCATT": 1418, + "TTCATC": 1419, + "TTCATG": 1420, + "TTCACA": 1421, + "TTCACT": 1422, + "TTCACC": 1423, + "TTCACG": 1424, + "TTCAGA": 1425, + "TTCAGT": 1426, + "TTCAGC": 1427, + "TTCAGG": 1428, + "TTCTAA": 1429, + "TTCTAT": 1430, + "TTCTAC": 1431, + "TTCTAG": 1432, + "TTCTTA": 1433, + "TTCTTT": 1434, + "TTCTTC": 1435, + "TTCTTG": 1436, + "TTCTCA": 1437, + "TTCTCT": 1438, + "TTCTCC": 1439, + "TTCTCG": 1440, + "TTCTGA": 1441, + "TTCTGT": 1442, + "TTCTGC": 1443, + "TTCTGG": 1444, + "TTCCAA": 1445, + "TTCCAT": 1446, + "TTCCAC": 1447, + "TTCCAG": 1448, + "TTCCTA": 1449, + "TTCCTT": 1450, + "TTCCTC": 1451, + "TTCCTG": 1452, + "TTCCCA": 1453, + "TTCCCT": 1454, + "TTCCCC": 1455, + "TTCCCG": 1456, + "TTCCGA": 1457, + "TTCCGT": 1458, + "TTCCGC": 1459, + "TTCCGG": 1460, + "TTCGAA": 1461, + "TTCGAT": 1462, + "TTCGAC": 1463, + "TTCGAG": 1464, + "TTCGTA": 1465, + "TTCGTT": 1466, + "TTCGTC": 1467, + "TTCGTG": 1468, + "TTCGCA": 1469, + "TTCGCT": 1470, + "TTCGCC": 1471, + "TTCGCG": 1472, + "TTCGGA": 1473, + "TTCGGT": 1474, + "TTCGGC": 1475, + "TTCGGG": 1476, + "TTGAAA": 1477, + "TTGAAT": 1478, + "TTGAAC": 1479, + "TTGAAG": 1480, + "TTGATA": 1481, + "TTGATT": 1482, + "TTGATC": 1483, + "TTGATG": 1484, + "TTGACA": 1485, + "TTGACT": 1486, + "TTGACC": 1487, + "TTGACG": 1488, + "TTGAGA": 1489, + "TTGAGT": 1490, + "TTGAGC": 1491, + "TTGAGG": 1492, + "TTGTAA": 1493, + "TTGTAT": 1494, + "TTGTAC": 1495, + "TTGTAG": 1496, + "TTGTTA": 1497, + "TTGTTT": 1498, + "TTGTTC": 1499, + "TTGTTG": 1500, + "TTGTCA": 1501, + "TTGTCT": 1502, + "TTGTCC": 1503, + "TTGTCG": 1504, + "TTGTGA": 1505, + "TTGTGT": 1506, + "TTGTGC": 1507, + "TTGTGG": 1508, + "TTGCAA": 1509, + "TTGCAT": 1510, + "TTGCAC": 1511, + "TTGCAG": 1512, + "TTGCTA": 1513, + "TTGCTT": 1514, + "TTGCTC": 1515, + "TTGCTG": 1516, + "TTGCCA": 1517, + "TTGCCT": 1518, + "TTGCCC": 1519, + "TTGCCG": 1520, + "TTGCGA": 1521, + "TTGCGT": 1522, + "TTGCGC": 1523, + "TTGCGG": 1524, + "TTGGAA": 1525, + "TTGGAT": 1526, + "TTGGAC": 1527, + "TTGGAG": 1528, + "TTGGTA": 1529, + "TTGGTT": 1530, + "TTGGTC": 1531, + "TTGGTG": 1532, + "TTGGCA": 1533, + "TTGGCT": 1534, + "TTGGCC": 1535, + "TTGGCG": 1536, + "TTGGGA": 1537, + "TTGGGT": 1538, + "TTGGGC": 1539, + "TTGGGG": 1540, + "TCAAAA": 1541, + "TCAAAT": 1542, + "TCAAAC": 1543, + "TCAAAG": 1544, + "TCAATA": 1545, + "TCAATT": 1546, + "TCAATC": 1547, + "TCAATG": 1548, + "TCAACA": 1549, + "TCAACT": 1550, + "TCAACC": 1551, + "TCAACG": 1552, + "TCAAGA": 1553, + "TCAAGT": 1554, + "TCAAGC": 1555, + "TCAAGG": 1556, + "TCATAA": 1557, + "TCATAT": 1558, + "TCATAC": 1559, + "TCATAG": 1560, + "TCATTA": 1561, + "TCATTT": 1562, + "TCATTC": 1563, + "TCATTG": 1564, + "TCATCA": 1565, + "TCATCT": 1566, + "TCATCC": 1567, + "TCATCG": 1568, + "TCATGA": 1569, + "TCATGT": 1570, + "TCATGC": 1571, + "TCATGG": 1572, + "TCACAA": 1573, + "TCACAT": 1574, + "TCACAC": 1575, + "TCACAG": 1576, + "TCACTA": 1577, + "TCACTT": 1578, + "TCACTC": 1579, + "TCACTG": 1580, + "TCACCA": 1581, + "TCACCT": 1582, + "TCACCC": 1583, + "TCACCG": 1584, + "TCACGA": 1585, + "TCACGT": 1586, + "TCACGC": 1587, + "TCACGG": 1588, + "TCAGAA": 1589, + "TCAGAT": 1590, + "TCAGAC": 1591, + "TCAGAG": 1592, + "TCAGTA": 1593, + "TCAGTT": 1594, + "TCAGTC": 1595, + "TCAGTG": 1596, + "TCAGCA": 1597, + "TCAGCT": 1598, + "TCAGCC": 1599, + "TCAGCG": 1600, + "TCAGGA": 1601, + "TCAGGT": 1602, + "TCAGGC": 1603, + "TCAGGG": 1604, + "TCTAAA": 1605, + "TCTAAT": 1606, + "TCTAAC": 1607, + "TCTAAG": 1608, + "TCTATA": 1609, + "TCTATT": 1610, + "TCTATC": 1611, + "TCTATG": 1612, + "TCTACA": 1613, + "TCTACT": 1614, + "TCTACC": 1615, + "TCTACG": 1616, + "TCTAGA": 1617, + "TCTAGT": 1618, + "TCTAGC": 1619, + "TCTAGG": 1620, + "TCTTAA": 1621, + "TCTTAT": 1622, + "TCTTAC": 1623, + "TCTTAG": 1624, + "TCTTTA": 1625, + "TCTTTT": 1626, + "TCTTTC": 1627, + "TCTTTG": 1628, + "TCTTCA": 1629, + "TCTTCT": 1630, + "TCTTCC": 1631, + "TCTTCG": 1632, + "TCTTGA": 1633, + "TCTTGT": 1634, + "TCTTGC": 1635, + "TCTTGG": 1636, + "TCTCAA": 1637, + "TCTCAT": 1638, + "TCTCAC": 1639, + "TCTCAG": 1640, + "TCTCTA": 1641, + "TCTCTT": 1642, + "TCTCTC": 1643, + "TCTCTG": 1644, + "TCTCCA": 1645, + "TCTCCT": 1646, + "TCTCCC": 1647, + "TCTCCG": 1648, + "TCTCGA": 1649, + "TCTCGT": 1650, + "TCTCGC": 1651, + "TCTCGG": 1652, + "TCTGAA": 1653, + "TCTGAT": 1654, + "TCTGAC": 1655, + "TCTGAG": 1656, + "TCTGTA": 1657, + "TCTGTT": 1658, + "TCTGTC": 1659, + "TCTGTG": 1660, + "TCTGCA": 1661, + "TCTGCT": 1662, + "TCTGCC": 1663, + "TCTGCG": 1664, + "TCTGGA": 1665, + "TCTGGT": 1666, + "TCTGGC": 1667, + "TCTGGG": 1668, + "TCCAAA": 1669, + "TCCAAT": 1670, + "TCCAAC": 1671, + "TCCAAG": 1672, + "TCCATA": 1673, + "TCCATT": 1674, + "TCCATC": 1675, + "TCCATG": 1676, + "TCCACA": 1677, + "TCCACT": 1678, + "TCCACC": 1679, + "TCCACG": 1680, + "TCCAGA": 1681, + "TCCAGT": 1682, + "TCCAGC": 1683, + "TCCAGG": 1684, + "TCCTAA": 1685, + "TCCTAT": 1686, + "TCCTAC": 1687, + "TCCTAG": 1688, + "TCCTTA": 1689, + "TCCTTT": 1690, + "TCCTTC": 1691, + "TCCTTG": 1692, + "TCCTCA": 1693, + "TCCTCT": 1694, + "TCCTCC": 1695, + "TCCTCG": 1696, + "TCCTGA": 1697, + "TCCTGT": 1698, + "TCCTGC": 1699, + "TCCTGG": 1700, + "TCCCAA": 1701, + "TCCCAT": 1702, + "TCCCAC": 1703, + "TCCCAG": 1704, + "TCCCTA": 1705, + "TCCCTT": 1706, + "TCCCTC": 1707, + "TCCCTG": 1708, + "TCCCCA": 1709, + "TCCCCT": 1710, + "TCCCCC": 1711, + "TCCCCG": 1712, + "TCCCGA": 1713, + "TCCCGT": 1714, + "TCCCGC": 1715, + "TCCCGG": 1716, + "TCCGAA": 1717, + "TCCGAT": 1718, + "TCCGAC": 1719, + "TCCGAG": 1720, + "TCCGTA": 1721, + "TCCGTT": 1722, + "TCCGTC": 1723, + "TCCGTG": 1724, + "TCCGCA": 1725, + "TCCGCT": 1726, + "TCCGCC": 1727, + "TCCGCG": 1728, + "TCCGGA": 1729, + "TCCGGT": 1730, + "TCCGGC": 1731, + "TCCGGG": 1732, + "TCGAAA": 1733, + "TCGAAT": 1734, + "TCGAAC": 1735, + "TCGAAG": 1736, + "TCGATA": 1737, + "TCGATT": 1738, + "TCGATC": 1739, + "TCGATG": 1740, + "TCGACA": 1741, + "TCGACT": 1742, + "TCGACC": 1743, + "TCGACG": 1744, + "TCGAGA": 1745, + "TCGAGT": 1746, + "TCGAGC": 1747, + "TCGAGG": 1748, + "TCGTAA": 1749, + "TCGTAT": 1750, + "TCGTAC": 1751, + "TCGTAG": 1752, + "TCGTTA": 1753, + "TCGTTT": 1754, + "TCGTTC": 1755, + "TCGTTG": 1756, + "TCGTCA": 1757, + "TCGTCT": 1758, + "TCGTCC": 1759, + "TCGTCG": 1760, + "TCGTGA": 1761, + "TCGTGT": 1762, + "TCGTGC": 1763, + "TCGTGG": 1764, + "TCGCAA": 1765, + "TCGCAT": 1766, + "TCGCAC": 1767, + "TCGCAG": 1768, + "TCGCTA": 1769, + "TCGCTT": 1770, + "TCGCTC": 1771, + "TCGCTG": 1772, + "TCGCCA": 1773, + "TCGCCT": 1774, + "TCGCCC": 1775, + "TCGCCG": 1776, + "TCGCGA": 1777, + "TCGCGT": 1778, + "TCGCGC": 1779, + "TCGCGG": 1780, + "TCGGAA": 1781, + "TCGGAT": 1782, + "TCGGAC": 1783, + "TCGGAG": 1784, + "TCGGTA": 1785, + "TCGGTT": 1786, + "TCGGTC": 1787, + "TCGGTG": 1788, + "TCGGCA": 1789, + "TCGGCT": 1790, + "TCGGCC": 1791, + "TCGGCG": 1792, + "TCGGGA": 1793, + "TCGGGT": 1794, + "TCGGGC": 1795, + "TCGGGG": 1796, + "TGAAAA": 1797, + "TGAAAT": 1798, + "TGAAAC": 1799, + "TGAAAG": 1800, + "TGAATA": 1801, + "TGAATT": 1802, + "TGAATC": 1803, + "TGAATG": 1804, + "TGAACA": 1805, + "TGAACT": 1806, + "TGAACC": 1807, + "TGAACG": 1808, + "TGAAGA": 1809, + "TGAAGT": 1810, + "TGAAGC": 1811, + "TGAAGG": 1812, + "TGATAA": 1813, + "TGATAT": 1814, + "TGATAC": 1815, + "TGATAG": 1816, + "TGATTA": 1817, + "TGATTT": 1818, + "TGATTC": 1819, + "TGATTG": 1820, + "TGATCA": 1821, + "TGATCT": 1822, + "TGATCC": 1823, + "TGATCG": 1824, + "TGATGA": 1825, + "TGATGT": 1826, + "TGATGC": 1827, + "TGATGG": 1828, + "TGACAA": 1829, + "TGACAT": 1830, + "TGACAC": 1831, + "TGACAG": 1832, + "TGACTA": 1833, + "TGACTT": 1834, + "TGACTC": 1835, + "TGACTG": 1836, + "TGACCA": 1837, + "TGACCT": 1838, + "TGACCC": 1839, + "TGACCG": 1840, + "TGACGA": 1841, + "TGACGT": 1842, + "TGACGC": 1843, + "TGACGG": 1844, + "TGAGAA": 1845, + "TGAGAT": 1846, + "TGAGAC": 1847, + "TGAGAG": 1848, + "TGAGTA": 1849, + "TGAGTT": 1850, + "TGAGTC": 1851, + "TGAGTG": 1852, + "TGAGCA": 1853, + "TGAGCT": 1854, + "TGAGCC": 1855, + "TGAGCG": 1856, + "TGAGGA": 1857, + "TGAGGT": 1858, + "TGAGGC": 1859, + "TGAGGG": 1860, + "TGTAAA": 1861, + "TGTAAT": 1862, + "TGTAAC": 1863, + "TGTAAG": 1864, + "TGTATA": 1865, + "TGTATT": 1866, + "TGTATC": 1867, + "TGTATG": 1868, + "TGTACA": 1869, + "TGTACT": 1870, + "TGTACC": 1871, + "TGTACG": 1872, + "TGTAGA": 1873, + "TGTAGT": 1874, + "TGTAGC": 1875, + "TGTAGG": 1876, + "TGTTAA": 1877, + "TGTTAT": 1878, + "TGTTAC": 1879, + "TGTTAG": 1880, + "TGTTTA": 1881, + "TGTTTT": 1882, + "TGTTTC": 1883, + "TGTTTG": 1884, + "TGTTCA": 1885, + "TGTTCT": 1886, + "TGTTCC": 1887, + "TGTTCG": 1888, + "TGTTGA": 1889, + "TGTTGT": 1890, + "TGTTGC": 1891, + "TGTTGG": 1892, + "TGTCAA": 1893, + "TGTCAT": 1894, + "TGTCAC": 1895, + "TGTCAG": 1896, + "TGTCTA": 1897, + "TGTCTT": 1898, + "TGTCTC": 1899, + "TGTCTG": 1900, + "TGTCCA": 1901, + "TGTCCT": 1902, + "TGTCCC": 1903, + "TGTCCG": 1904, + "TGTCGA": 1905, + "TGTCGT": 1906, + "TGTCGC": 1907, + "TGTCGG": 1908, + "TGTGAA": 1909, + "TGTGAT": 1910, + "TGTGAC": 1911, + "TGTGAG": 1912, + "TGTGTA": 1913, + "TGTGTT": 1914, + "TGTGTC": 1915, + "TGTGTG": 1916, + "TGTGCA": 1917, + "TGTGCT": 1918, + "TGTGCC": 1919, + "TGTGCG": 1920, + "TGTGGA": 1921, + "TGTGGT": 1922, + "TGTGGC": 1923, + "TGTGGG": 1924, + "TGCAAA": 1925, + "TGCAAT": 1926, + "TGCAAC": 1927, + "TGCAAG": 1928, + "TGCATA": 1929, + "TGCATT": 1930, + "TGCATC": 1931, + "TGCATG": 1932, + "TGCACA": 1933, + "TGCACT": 1934, + "TGCACC": 1935, + "TGCACG": 1936, + "TGCAGA": 1937, + "TGCAGT": 1938, + "TGCAGC": 1939, + "TGCAGG": 1940, + "TGCTAA": 1941, + "TGCTAT": 1942, + "TGCTAC": 1943, + "TGCTAG": 1944, + "TGCTTA": 1945, + "TGCTTT": 1946, + "TGCTTC": 1947, + "TGCTTG": 1948, + "TGCTCA": 1949, + "TGCTCT": 1950, + "TGCTCC": 1951, + "TGCTCG": 1952, + "TGCTGA": 1953, + "TGCTGT": 1954, + "TGCTGC": 1955, + "TGCTGG": 1956, + "TGCCAA": 1957, + "TGCCAT": 1958, + "TGCCAC": 1959, + "TGCCAG": 1960, + "TGCCTA": 1961, + "TGCCTT": 1962, + "TGCCTC": 1963, + "TGCCTG": 1964, + "TGCCCA": 1965, + "TGCCCT": 1966, + "TGCCCC": 1967, + "TGCCCG": 1968, + "TGCCGA": 1969, + "TGCCGT": 1970, + "TGCCGC": 1971, + "TGCCGG": 1972, + "TGCGAA": 1973, + "TGCGAT": 1974, + "TGCGAC": 1975, + "TGCGAG": 1976, + "TGCGTA": 1977, + "TGCGTT": 1978, + "TGCGTC": 1979, + "TGCGTG": 1980, + "TGCGCA": 1981, + "TGCGCT": 1982, + "TGCGCC": 1983, + "TGCGCG": 1984, + "TGCGGA": 1985, + "TGCGGT": 1986, + "TGCGGC": 1987, + "TGCGGG": 1988, + "TGGAAA": 1989, + "TGGAAT": 1990, + "TGGAAC": 1991, + "TGGAAG": 1992, + "TGGATA": 1993, + "TGGATT": 1994, + "TGGATC": 1995, + "TGGATG": 1996, + "TGGACA": 1997, + "TGGACT": 1998, + "TGGACC": 1999, + "TGGACG": 2000, + "TGGAGA": 2001, + "TGGAGT": 2002, + "TGGAGC": 2003, + "TGGAGG": 2004, + "TGGTAA": 2005, + "TGGTAT": 2006, + "TGGTAC": 2007, + "TGGTAG": 2008, + "TGGTTA": 2009, + "TGGTTT": 2010, + "TGGTTC": 2011, + "TGGTTG": 2012, + "TGGTCA": 2013, + "TGGTCT": 2014, + "TGGTCC": 2015, + "TGGTCG": 2016, + "TGGTGA": 2017, + "TGGTGT": 2018, + "TGGTGC": 2019, + "TGGTGG": 2020, + "TGGCAA": 2021, + "TGGCAT": 2022, + "TGGCAC": 2023, + "TGGCAG": 2024, + "TGGCTA": 2025, + "TGGCTT": 2026, + "TGGCTC": 2027, + "TGGCTG": 2028, + "TGGCCA": 2029, + "TGGCCT": 2030, + "TGGCCC": 2031, + "TGGCCG": 2032, + "TGGCGA": 2033, + "TGGCGT": 2034, + "TGGCGC": 2035, + "TGGCGG": 2036, + "TGGGAA": 2037, + "TGGGAT": 2038, + "TGGGAC": 2039, + "TGGGAG": 2040, + "TGGGTA": 2041, + "TGGGTT": 2042, + "TGGGTC": 2043, + "TGGGTG": 2044, + "TGGGCA": 2045, + "TGGGCT": 2046, + "TGGGCC": 2047, + "TGGGCG": 2048, + "TGGGGA": 2049, + "TGGGGT": 2050, + "TGGGGC": 2051, + "TGGGGG": 2052, + "CAAAAA": 2053, + "CAAAAT": 2054, + "CAAAAC": 2055, + "CAAAAG": 2056, + "CAAATA": 2057, + "CAAATT": 2058, + "CAAATC": 2059, + "CAAATG": 2060, + "CAAACA": 2061, + "CAAACT": 2062, + "CAAACC": 2063, + "CAAACG": 2064, + "CAAAGA": 2065, + "CAAAGT": 2066, + "CAAAGC": 2067, + "CAAAGG": 2068, + "CAATAA": 2069, + "CAATAT": 2070, + "CAATAC": 2071, + "CAATAG": 2072, + "CAATTA": 2073, + "CAATTT": 2074, + "CAATTC": 2075, + "CAATTG": 2076, + "CAATCA": 2077, + "CAATCT": 2078, + "CAATCC": 2079, + "CAATCG": 2080, + "CAATGA": 2081, + "CAATGT": 2082, + "CAATGC": 2083, + "CAATGG": 2084, + "CAACAA": 2085, + "CAACAT": 2086, + "CAACAC": 2087, + "CAACAG": 2088, + "CAACTA": 2089, + "CAACTT": 2090, + "CAACTC": 2091, + "CAACTG": 2092, + "CAACCA": 2093, + "CAACCT": 2094, + "CAACCC": 2095, + "CAACCG": 2096, + "CAACGA": 2097, + "CAACGT": 2098, + "CAACGC": 2099, + "CAACGG": 2100, + "CAAGAA": 2101, + "CAAGAT": 2102, + "CAAGAC": 2103, + "CAAGAG": 2104, + "CAAGTA": 2105, + "CAAGTT": 2106, + "CAAGTC": 2107, + "CAAGTG": 2108, + "CAAGCA": 2109, + "CAAGCT": 2110, + "CAAGCC": 2111, + "CAAGCG": 2112, + "CAAGGA": 2113, + "CAAGGT": 2114, + "CAAGGC": 2115, + "CAAGGG": 2116, + "CATAAA": 2117, + "CATAAT": 2118, + "CATAAC": 2119, + "CATAAG": 2120, + "CATATA": 2121, + "CATATT": 2122, + "CATATC": 2123, + "CATATG": 2124, + "CATACA": 2125, + "CATACT": 2126, + "CATACC": 2127, + "CATACG": 2128, + "CATAGA": 2129, + "CATAGT": 2130, + "CATAGC": 2131, + "CATAGG": 2132, + "CATTAA": 2133, + "CATTAT": 2134, + "CATTAC": 2135, + "CATTAG": 2136, + "CATTTA": 2137, + "CATTTT": 2138, + "CATTTC": 2139, + "CATTTG": 2140, + "CATTCA": 2141, + "CATTCT": 2142, + "CATTCC": 2143, + "CATTCG": 2144, + "CATTGA": 2145, + "CATTGT": 2146, + "CATTGC": 2147, + "CATTGG": 2148, + "CATCAA": 2149, + "CATCAT": 2150, + "CATCAC": 2151, + "CATCAG": 2152, + "CATCTA": 2153, + "CATCTT": 2154, + "CATCTC": 2155, + "CATCTG": 2156, + "CATCCA": 2157, + "CATCCT": 2158, + "CATCCC": 2159, + "CATCCG": 2160, + "CATCGA": 2161, + "CATCGT": 2162, + "CATCGC": 2163, + "CATCGG": 2164, + "CATGAA": 2165, + "CATGAT": 2166, + "CATGAC": 2167, + "CATGAG": 2168, + "CATGTA": 2169, + "CATGTT": 2170, + "CATGTC": 2171, + "CATGTG": 2172, + "CATGCA": 2173, + "CATGCT": 2174, + "CATGCC": 2175, + "CATGCG": 2176, + "CATGGA": 2177, + "CATGGT": 2178, + "CATGGC": 2179, + "CATGGG": 2180, + "CACAAA": 2181, + "CACAAT": 2182, + "CACAAC": 2183, + "CACAAG": 2184, + "CACATA": 2185, + "CACATT": 2186, + "CACATC": 2187, + "CACATG": 2188, + "CACACA": 2189, + "CACACT": 2190, + "CACACC": 2191, + "CACACG": 2192, + "CACAGA": 2193, + "CACAGT": 2194, + "CACAGC": 2195, + "CACAGG": 2196, + "CACTAA": 2197, + "CACTAT": 2198, + "CACTAC": 2199, + "CACTAG": 2200, + "CACTTA": 2201, + "CACTTT": 2202, + "CACTTC": 2203, + "CACTTG": 2204, + "CACTCA": 2205, + "CACTCT": 2206, + "CACTCC": 2207, + "CACTCG": 2208, + "CACTGA": 2209, + "CACTGT": 2210, + "CACTGC": 2211, + "CACTGG": 2212, + "CACCAA": 2213, + "CACCAT": 2214, + "CACCAC": 2215, + "CACCAG": 2216, + "CACCTA": 2217, + "CACCTT": 2218, + "CACCTC": 2219, + "CACCTG": 2220, + "CACCCA": 2221, + "CACCCT": 2222, + "CACCCC": 2223, + "CACCCG": 2224, + "CACCGA": 2225, + "CACCGT": 2226, + "CACCGC": 2227, + "CACCGG": 2228, + "CACGAA": 2229, + "CACGAT": 2230, + "CACGAC": 2231, + "CACGAG": 2232, + "CACGTA": 2233, + "CACGTT": 2234, + "CACGTC": 2235, + "CACGTG": 2236, + "CACGCA": 2237, + "CACGCT": 2238, + "CACGCC": 2239, + "CACGCG": 2240, + "CACGGA": 2241, + "CACGGT": 2242, + "CACGGC": 2243, + "CACGGG": 2244, + "CAGAAA": 2245, + "CAGAAT": 2246, + "CAGAAC": 2247, + "CAGAAG": 2248, + "CAGATA": 2249, + "CAGATT": 2250, + "CAGATC": 2251, + "CAGATG": 2252, + "CAGACA": 2253, + "CAGACT": 2254, + "CAGACC": 2255, + "CAGACG": 2256, + "CAGAGA": 2257, + "CAGAGT": 2258, + "CAGAGC": 2259, + "CAGAGG": 2260, + "CAGTAA": 2261, + "CAGTAT": 2262, + "CAGTAC": 2263, + "CAGTAG": 2264, + "CAGTTA": 2265, + "CAGTTT": 2266, + "CAGTTC": 2267, + "CAGTTG": 2268, + "CAGTCA": 2269, + "CAGTCT": 2270, + "CAGTCC": 2271, + "CAGTCG": 2272, + "CAGTGA": 2273, + "CAGTGT": 2274, + "CAGTGC": 2275, + "CAGTGG": 2276, + "CAGCAA": 2277, + "CAGCAT": 2278, + "CAGCAC": 2279, + "CAGCAG": 2280, + "CAGCTA": 2281, + "CAGCTT": 2282, + "CAGCTC": 2283, + "CAGCTG": 2284, + "CAGCCA": 2285, + "CAGCCT": 2286, + "CAGCCC": 2287, + "CAGCCG": 2288, + "CAGCGA": 2289, + "CAGCGT": 2290, + "CAGCGC": 2291, + "CAGCGG": 2292, + "CAGGAA": 2293, + "CAGGAT": 2294, + "CAGGAC": 2295, + "CAGGAG": 2296, + "CAGGTA": 2297, + "CAGGTT": 2298, + "CAGGTC": 2299, + "CAGGTG": 2300, + "CAGGCA": 2301, + "CAGGCT": 2302, + "CAGGCC": 2303, + "CAGGCG": 2304, + "CAGGGA": 2305, + "CAGGGT": 2306, + "CAGGGC": 2307, + "CAGGGG": 2308, + "CTAAAA": 2309, + "CTAAAT": 2310, + "CTAAAC": 2311, + "CTAAAG": 2312, + "CTAATA": 2313, + "CTAATT": 2314, + "CTAATC": 2315, + "CTAATG": 2316, + "CTAACA": 2317, + "CTAACT": 2318, + "CTAACC": 2319, + "CTAACG": 2320, + "CTAAGA": 2321, + "CTAAGT": 2322, + "CTAAGC": 2323, + "CTAAGG": 2324, + "CTATAA": 2325, + "CTATAT": 2326, + "CTATAC": 2327, + "CTATAG": 2328, + "CTATTA": 2329, + "CTATTT": 2330, + "CTATTC": 2331, + "CTATTG": 2332, + "CTATCA": 2333, + "CTATCT": 2334, + "CTATCC": 2335, + "CTATCG": 2336, + "CTATGA": 2337, + "CTATGT": 2338, + "CTATGC": 2339, + "CTATGG": 2340, + "CTACAA": 2341, + "CTACAT": 2342, + "CTACAC": 2343, + "CTACAG": 2344, + "CTACTA": 2345, + "CTACTT": 2346, + "CTACTC": 2347, + "CTACTG": 2348, + "CTACCA": 2349, + "CTACCT": 2350, + "CTACCC": 2351, + "CTACCG": 2352, + "CTACGA": 2353, + "CTACGT": 2354, + "CTACGC": 2355, + "CTACGG": 2356, + "CTAGAA": 2357, + "CTAGAT": 2358, + "CTAGAC": 2359, + "CTAGAG": 2360, + "CTAGTA": 2361, + "CTAGTT": 2362, + "CTAGTC": 2363, + "CTAGTG": 2364, + "CTAGCA": 2365, + "CTAGCT": 2366, + "CTAGCC": 2367, + "CTAGCG": 2368, + "CTAGGA": 2369, + "CTAGGT": 2370, + "CTAGGC": 2371, + "CTAGGG": 2372, + "CTTAAA": 2373, + "CTTAAT": 2374, + "CTTAAC": 2375, + "CTTAAG": 2376, + "CTTATA": 2377, + "CTTATT": 2378, + "CTTATC": 2379, + "CTTATG": 2380, + "CTTACA": 2381, + "CTTACT": 2382, + "CTTACC": 2383, + "CTTACG": 2384, + "CTTAGA": 2385, + "CTTAGT": 2386, + "CTTAGC": 2387, + "CTTAGG": 2388, + "CTTTAA": 2389, + "CTTTAT": 2390, + "CTTTAC": 2391, + "CTTTAG": 2392, + "CTTTTA": 2393, + "CTTTTT": 2394, + "CTTTTC": 2395, + "CTTTTG": 2396, + "CTTTCA": 2397, + "CTTTCT": 2398, + "CTTTCC": 2399, + "CTTTCG": 2400, + "CTTTGA": 2401, + "CTTTGT": 2402, + "CTTTGC": 2403, + "CTTTGG": 2404, + "CTTCAA": 2405, + "CTTCAT": 2406, + "CTTCAC": 2407, + "CTTCAG": 2408, + "CTTCTA": 2409, + "CTTCTT": 2410, + "CTTCTC": 2411, + "CTTCTG": 2412, + "CTTCCA": 2413, + "CTTCCT": 2414, + "CTTCCC": 2415, + "CTTCCG": 2416, + "CTTCGA": 2417, + "CTTCGT": 2418, + "CTTCGC": 2419, + "CTTCGG": 2420, + "CTTGAA": 2421, + "CTTGAT": 2422, + "CTTGAC": 2423, + "CTTGAG": 2424, + "CTTGTA": 2425, + "CTTGTT": 2426, + "CTTGTC": 2427, + "CTTGTG": 2428, + "CTTGCA": 2429, + "CTTGCT": 2430, + "CTTGCC": 2431, + "CTTGCG": 2432, + "CTTGGA": 2433, + "CTTGGT": 2434, + "CTTGGC": 2435, + "CTTGGG": 2436, + "CTCAAA": 2437, + "CTCAAT": 2438, + "CTCAAC": 2439, + "CTCAAG": 2440, + "CTCATA": 2441, + "CTCATT": 2442, + "CTCATC": 2443, + "CTCATG": 2444, + "CTCACA": 2445, + "CTCACT": 2446, + "CTCACC": 2447, + "CTCACG": 2448, + "CTCAGA": 2449, + "CTCAGT": 2450, + "CTCAGC": 2451, + "CTCAGG": 2452, + "CTCTAA": 2453, + "CTCTAT": 2454, + "CTCTAC": 2455, + "CTCTAG": 2456, + "CTCTTA": 2457, + "CTCTTT": 2458, + "CTCTTC": 2459, + "CTCTTG": 2460, + "CTCTCA": 2461, + "CTCTCT": 2462, + "CTCTCC": 2463, + "CTCTCG": 2464, + "CTCTGA": 2465, + "CTCTGT": 2466, + "CTCTGC": 2467, + "CTCTGG": 2468, + "CTCCAA": 2469, + "CTCCAT": 2470, + "CTCCAC": 2471, + "CTCCAG": 2472, + "CTCCTA": 2473, + "CTCCTT": 2474, + "CTCCTC": 2475, + "CTCCTG": 2476, + "CTCCCA": 2477, + "CTCCCT": 2478, + "CTCCCC": 2479, + "CTCCCG": 2480, + "CTCCGA": 2481, + "CTCCGT": 2482, + "CTCCGC": 2483, + "CTCCGG": 2484, + "CTCGAA": 2485, + "CTCGAT": 2486, + "CTCGAC": 2487, + "CTCGAG": 2488, + "CTCGTA": 2489, + "CTCGTT": 2490, + "CTCGTC": 2491, + "CTCGTG": 2492, + "CTCGCA": 2493, + "CTCGCT": 2494, + "CTCGCC": 2495, + "CTCGCG": 2496, + "CTCGGA": 2497, + "CTCGGT": 2498, + "CTCGGC": 2499, + "CTCGGG": 2500, + "CTGAAA": 2501, + "CTGAAT": 2502, + "CTGAAC": 2503, + "CTGAAG": 2504, + "CTGATA": 2505, + "CTGATT": 2506, + "CTGATC": 2507, + "CTGATG": 2508, + "CTGACA": 2509, + "CTGACT": 2510, + "CTGACC": 2511, + "CTGACG": 2512, + "CTGAGA": 2513, + "CTGAGT": 2514, + "CTGAGC": 2515, + "CTGAGG": 2516, + "CTGTAA": 2517, + "CTGTAT": 2518, + "CTGTAC": 2519, + "CTGTAG": 2520, + "CTGTTA": 2521, + "CTGTTT": 2522, + "CTGTTC": 2523, + "CTGTTG": 2524, + "CTGTCA": 2525, + "CTGTCT": 2526, + "CTGTCC": 2527, + "CTGTCG": 2528, + "CTGTGA": 2529, + "CTGTGT": 2530, + "CTGTGC": 2531, + "CTGTGG": 2532, + "CTGCAA": 2533, + "CTGCAT": 2534, + "CTGCAC": 2535, + "CTGCAG": 2536, + "CTGCTA": 2537, + "CTGCTT": 2538, + "CTGCTC": 2539, + "CTGCTG": 2540, + "CTGCCA": 2541, + "CTGCCT": 2542, + "CTGCCC": 2543, + "CTGCCG": 2544, + "CTGCGA": 2545, + "CTGCGT": 2546, + "CTGCGC": 2547, + "CTGCGG": 2548, + "CTGGAA": 2549, + "CTGGAT": 2550, + "CTGGAC": 2551, + "CTGGAG": 2552, + "CTGGTA": 2553, + "CTGGTT": 2554, + "CTGGTC": 2555, + "CTGGTG": 2556, + "CTGGCA": 2557, + "CTGGCT": 2558, + "CTGGCC": 2559, + "CTGGCG": 2560, + "CTGGGA": 2561, + "CTGGGT": 2562, + "CTGGGC": 2563, + "CTGGGG": 2564, + "CCAAAA": 2565, + "CCAAAT": 2566, + "CCAAAC": 2567, + "CCAAAG": 2568, + "CCAATA": 2569, + "CCAATT": 2570, + "CCAATC": 2571, + "CCAATG": 2572, + "CCAACA": 2573, + "CCAACT": 2574, + "CCAACC": 2575, + "CCAACG": 2576, + "CCAAGA": 2577, + "CCAAGT": 2578, + "CCAAGC": 2579, + "CCAAGG": 2580, + "CCATAA": 2581, + "CCATAT": 2582, + "CCATAC": 2583, + "CCATAG": 2584, + "CCATTA": 2585, + "CCATTT": 2586, + "CCATTC": 2587, + "CCATTG": 2588, + "CCATCA": 2589, + "CCATCT": 2590, + "CCATCC": 2591, + "CCATCG": 2592, + "CCATGA": 2593, + "CCATGT": 2594, + "CCATGC": 2595, + "CCATGG": 2596, + "CCACAA": 2597, + "CCACAT": 2598, + "CCACAC": 2599, + "CCACAG": 2600, + "CCACTA": 2601, + "CCACTT": 2602, + "CCACTC": 2603, + "CCACTG": 2604, + "CCACCA": 2605, + "CCACCT": 2606, + "CCACCC": 2607, + "CCACCG": 2608, + "CCACGA": 2609, + "CCACGT": 2610, + "CCACGC": 2611, + "CCACGG": 2612, + "CCAGAA": 2613, + "CCAGAT": 2614, + "CCAGAC": 2615, + "CCAGAG": 2616, + "CCAGTA": 2617, + "CCAGTT": 2618, + "CCAGTC": 2619, + "CCAGTG": 2620, + "CCAGCA": 2621, + "CCAGCT": 2622, + "CCAGCC": 2623, + "CCAGCG": 2624, + "CCAGGA": 2625, + "CCAGGT": 2626, + "CCAGGC": 2627, + "CCAGGG": 2628, + "CCTAAA": 2629, + "CCTAAT": 2630, + "CCTAAC": 2631, + "CCTAAG": 2632, + "CCTATA": 2633, + "CCTATT": 2634, + "CCTATC": 2635, + "CCTATG": 2636, + "CCTACA": 2637, + "CCTACT": 2638, + "CCTACC": 2639, + "CCTACG": 2640, + "CCTAGA": 2641, + "CCTAGT": 2642, + "CCTAGC": 2643, + "CCTAGG": 2644, + "CCTTAA": 2645, + "CCTTAT": 2646, + "CCTTAC": 2647, + "CCTTAG": 2648, + "CCTTTA": 2649, + "CCTTTT": 2650, + "CCTTTC": 2651, + "CCTTTG": 2652, + "CCTTCA": 2653, + "CCTTCT": 2654, + "CCTTCC": 2655, + "CCTTCG": 2656, + "CCTTGA": 2657, + "CCTTGT": 2658, + "CCTTGC": 2659, + "CCTTGG": 2660, + "CCTCAA": 2661, + "CCTCAT": 2662, + "CCTCAC": 2663, + "CCTCAG": 2664, + "CCTCTA": 2665, + "CCTCTT": 2666, + "CCTCTC": 2667, + "CCTCTG": 2668, + "CCTCCA": 2669, + "CCTCCT": 2670, + "CCTCCC": 2671, + "CCTCCG": 2672, + "CCTCGA": 2673, + "CCTCGT": 2674, + "CCTCGC": 2675, + "CCTCGG": 2676, + "CCTGAA": 2677, + "CCTGAT": 2678, + "CCTGAC": 2679, + "CCTGAG": 2680, + "CCTGTA": 2681, + "CCTGTT": 2682, + "CCTGTC": 2683, + "CCTGTG": 2684, + "CCTGCA": 2685, + "CCTGCT": 2686, + "CCTGCC": 2687, + "CCTGCG": 2688, + "CCTGGA": 2689, + "CCTGGT": 2690, + "CCTGGC": 2691, + "CCTGGG": 2692, + "CCCAAA": 2693, + "CCCAAT": 2694, + "CCCAAC": 2695, + "CCCAAG": 2696, + "CCCATA": 2697, + "CCCATT": 2698, + "CCCATC": 2699, + "CCCATG": 2700, + "CCCACA": 2701, + "CCCACT": 2702, + "CCCACC": 2703, + "CCCACG": 2704, + "CCCAGA": 2705, + "CCCAGT": 2706, + "CCCAGC": 2707, + "CCCAGG": 2708, + "CCCTAA": 2709, + "CCCTAT": 2710, + "CCCTAC": 2711, + "CCCTAG": 2712, + "CCCTTA": 2713, + "CCCTTT": 2714, + "CCCTTC": 2715, + "CCCTTG": 2716, + "CCCTCA": 2717, + "CCCTCT": 2718, + "CCCTCC": 2719, + "CCCTCG": 2720, + "CCCTGA": 2721, + "CCCTGT": 2722, + "CCCTGC": 2723, + "CCCTGG": 2724, + "CCCCAA": 2725, + "CCCCAT": 2726, + "CCCCAC": 2727, + "CCCCAG": 2728, + "CCCCTA": 2729, + "CCCCTT": 2730, + "CCCCTC": 2731, + "CCCCTG": 2732, + "CCCCCA": 2733, + "CCCCCT": 2734, + "CCCCCC": 2735, + "CCCCCG": 2736, + "CCCCGA": 2737, + "CCCCGT": 2738, + "CCCCGC": 2739, + "CCCCGG": 2740, + "CCCGAA": 2741, + "CCCGAT": 2742, + "CCCGAC": 2743, + "CCCGAG": 2744, + "CCCGTA": 2745, + "CCCGTT": 2746, + "CCCGTC": 2747, + "CCCGTG": 2748, + "CCCGCA": 2749, + "CCCGCT": 2750, + "CCCGCC": 2751, + "CCCGCG": 2752, + "CCCGGA": 2753, + "CCCGGT": 2754, + "CCCGGC": 2755, + "CCCGGG": 2756, + "CCGAAA": 2757, + "CCGAAT": 2758, + "CCGAAC": 2759, + "CCGAAG": 2760, + "CCGATA": 2761, + "CCGATT": 2762, + "CCGATC": 2763, + "CCGATG": 2764, + "CCGACA": 2765, + "CCGACT": 2766, + "CCGACC": 2767, + "CCGACG": 2768, + "CCGAGA": 2769, + "CCGAGT": 2770, + "CCGAGC": 2771, + "CCGAGG": 2772, + "CCGTAA": 2773, + "CCGTAT": 2774, + "CCGTAC": 2775, + "CCGTAG": 2776, + "CCGTTA": 2777, + "CCGTTT": 2778, + "CCGTTC": 2779, + "CCGTTG": 2780, + "CCGTCA": 2781, + "CCGTCT": 2782, + "CCGTCC": 2783, + "CCGTCG": 2784, + "CCGTGA": 2785, + "CCGTGT": 2786, + "CCGTGC": 2787, + "CCGTGG": 2788, + "CCGCAA": 2789, + "CCGCAT": 2790, + "CCGCAC": 2791, + "CCGCAG": 2792, + "CCGCTA": 2793, + "CCGCTT": 2794, + "CCGCTC": 2795, + "CCGCTG": 2796, + "CCGCCA": 2797, + "CCGCCT": 2798, + "CCGCCC": 2799, + "CCGCCG": 2800, + "CCGCGA": 2801, + "CCGCGT": 2802, + "CCGCGC": 2803, + "CCGCGG": 2804, + "CCGGAA": 2805, + "CCGGAT": 2806, + "CCGGAC": 2807, + "CCGGAG": 2808, + "CCGGTA": 2809, + "CCGGTT": 2810, + "CCGGTC": 2811, + "CCGGTG": 2812, + "CCGGCA": 2813, + "CCGGCT": 2814, + "CCGGCC": 2815, + "CCGGCG": 2816, + "CCGGGA": 2817, + "CCGGGT": 2818, + "CCGGGC": 2819, + "CCGGGG": 2820, + "CGAAAA": 2821, + "CGAAAT": 2822, + "CGAAAC": 2823, + "CGAAAG": 2824, + "CGAATA": 2825, + "CGAATT": 2826, + "CGAATC": 2827, + "CGAATG": 2828, + "CGAACA": 2829, + "CGAACT": 2830, + "CGAACC": 2831, + "CGAACG": 2832, + "CGAAGA": 2833, + "CGAAGT": 2834, + "CGAAGC": 2835, + "CGAAGG": 2836, + "CGATAA": 2837, + "CGATAT": 2838, + "CGATAC": 2839, + "CGATAG": 2840, + "CGATTA": 2841, + "CGATTT": 2842, + "CGATTC": 2843, + "CGATTG": 2844, + "CGATCA": 2845, + "CGATCT": 2846, + "CGATCC": 2847, + "CGATCG": 2848, + "CGATGA": 2849, + "CGATGT": 2850, + "CGATGC": 2851, + "CGATGG": 2852, + "CGACAA": 2853, + "CGACAT": 2854, + "CGACAC": 2855, + "CGACAG": 2856, + "CGACTA": 2857, + "CGACTT": 2858, + "CGACTC": 2859, + "CGACTG": 2860, + "CGACCA": 2861, + "CGACCT": 2862, + "CGACCC": 2863, + "CGACCG": 2864, + "CGACGA": 2865, + "CGACGT": 2866, + "CGACGC": 2867, + "CGACGG": 2868, + "CGAGAA": 2869, + "CGAGAT": 2870, + "CGAGAC": 2871, + "CGAGAG": 2872, + "CGAGTA": 2873, + "CGAGTT": 2874, + "CGAGTC": 2875, + "CGAGTG": 2876, + "CGAGCA": 2877, + "CGAGCT": 2878, + "CGAGCC": 2879, + "CGAGCG": 2880, + "CGAGGA": 2881, + "CGAGGT": 2882, + "CGAGGC": 2883, + "CGAGGG": 2884, + "CGTAAA": 2885, + "CGTAAT": 2886, + "CGTAAC": 2887, + "CGTAAG": 2888, + "CGTATA": 2889, + "CGTATT": 2890, + "CGTATC": 2891, + "CGTATG": 2892, + "CGTACA": 2893, + "CGTACT": 2894, + "CGTACC": 2895, + "CGTACG": 2896, + "CGTAGA": 2897, + "CGTAGT": 2898, + "CGTAGC": 2899, + "CGTAGG": 2900, + "CGTTAA": 2901, + "CGTTAT": 2902, + "CGTTAC": 2903, + "CGTTAG": 2904, + "CGTTTA": 2905, + "CGTTTT": 2906, + "CGTTTC": 2907, + "CGTTTG": 2908, + "CGTTCA": 2909, + "CGTTCT": 2910, + "CGTTCC": 2911, + "CGTTCG": 2912, + "CGTTGA": 2913, + "CGTTGT": 2914, + "CGTTGC": 2915, + "CGTTGG": 2916, + "CGTCAA": 2917, + "CGTCAT": 2918, + "CGTCAC": 2919, + "CGTCAG": 2920, + "CGTCTA": 2921, + "CGTCTT": 2922, + "CGTCTC": 2923, + "CGTCTG": 2924, + "CGTCCA": 2925, + "CGTCCT": 2926, + "CGTCCC": 2927, + "CGTCCG": 2928, + "CGTCGA": 2929, + "CGTCGT": 2930, + "CGTCGC": 2931, + "CGTCGG": 2932, + "CGTGAA": 2933, + "CGTGAT": 2934, + "CGTGAC": 2935, + "CGTGAG": 2936, + "CGTGTA": 2937, + "CGTGTT": 2938, + "CGTGTC": 2939, + "CGTGTG": 2940, + "CGTGCA": 2941, + "CGTGCT": 2942, + "CGTGCC": 2943, + "CGTGCG": 2944, + "CGTGGA": 2945, + "CGTGGT": 2946, + "CGTGGC": 2947, + "CGTGGG": 2948, + "CGCAAA": 2949, + "CGCAAT": 2950, + "CGCAAC": 2951, + "CGCAAG": 2952, + "CGCATA": 2953, + "CGCATT": 2954, + "CGCATC": 2955, + "CGCATG": 2956, + "CGCACA": 2957, + "CGCACT": 2958, + "CGCACC": 2959, + "CGCACG": 2960, + "CGCAGA": 2961, + "CGCAGT": 2962, + "CGCAGC": 2963, + "CGCAGG": 2964, + "CGCTAA": 2965, + "CGCTAT": 2966, + "CGCTAC": 2967, + "CGCTAG": 2968, + "CGCTTA": 2969, + "CGCTTT": 2970, + "CGCTTC": 2971, + "CGCTTG": 2972, + "CGCTCA": 2973, + "CGCTCT": 2974, + "CGCTCC": 2975, + "CGCTCG": 2976, + "CGCTGA": 2977, + "CGCTGT": 2978, + "CGCTGC": 2979, + "CGCTGG": 2980, + "CGCCAA": 2981, + "CGCCAT": 2982, + "CGCCAC": 2983, + "CGCCAG": 2984, + "CGCCTA": 2985, + "CGCCTT": 2986, + "CGCCTC": 2987, + "CGCCTG": 2988, + "CGCCCA": 2989, + "CGCCCT": 2990, + "CGCCCC": 2991, + "CGCCCG": 2992, + "CGCCGA": 2993, + "CGCCGT": 2994, + "CGCCGC": 2995, + "CGCCGG": 2996, + "CGCGAA": 2997, + "CGCGAT": 2998, + "CGCGAC": 2999, + "CGCGAG": 3000, + "CGCGTA": 3001, + "CGCGTT": 3002, + "CGCGTC": 3003, + "CGCGTG": 3004, + "CGCGCA": 3005, + "CGCGCT": 3006, + "CGCGCC": 3007, + "CGCGCG": 3008, + "CGCGGA": 3009, + "CGCGGT": 3010, + "CGCGGC": 3011, + "CGCGGG": 3012, + "CGGAAA": 3013, + "CGGAAT": 3014, + "CGGAAC": 3015, + "CGGAAG": 3016, + "CGGATA": 3017, + "CGGATT": 3018, + "CGGATC": 3019, + "CGGATG": 3020, + "CGGACA": 3021, + "CGGACT": 3022, + "CGGACC": 3023, + "CGGACG": 3024, + "CGGAGA": 3025, + "CGGAGT": 3026, + "CGGAGC": 3027, + "CGGAGG": 3028, + "CGGTAA": 3029, + "CGGTAT": 3030, + "CGGTAC": 3031, + "CGGTAG": 3032, + "CGGTTA": 3033, + "CGGTTT": 3034, + "CGGTTC": 3035, + "CGGTTG": 3036, + "CGGTCA": 3037, + "CGGTCT": 3038, + "CGGTCC": 3039, + "CGGTCG": 3040, + "CGGTGA": 3041, + "CGGTGT": 3042, + "CGGTGC": 3043, + "CGGTGG": 3044, + "CGGCAA": 3045, + "CGGCAT": 3046, + "CGGCAC": 3047, + "CGGCAG": 3048, + "CGGCTA": 3049, + "CGGCTT": 3050, + "CGGCTC": 3051, + "CGGCTG": 3052, + "CGGCCA": 3053, + "CGGCCT": 3054, + "CGGCCC": 3055, + "CGGCCG": 3056, + "CGGCGA": 3057, + "CGGCGT": 3058, + "CGGCGC": 3059, + "CGGCGG": 3060, + "CGGGAA": 3061, + "CGGGAT": 3062, + "CGGGAC": 3063, + "CGGGAG": 3064, + "CGGGTA": 3065, + "CGGGTT": 3066, + "CGGGTC": 3067, + "CGGGTG": 3068, + "CGGGCA": 3069, + "CGGGCT": 3070, + "CGGGCC": 3071, + "CGGGCG": 3072, + "CGGGGA": 3073, + "CGGGGT": 3074, + "CGGGGC": 3075, + "CGGGGG": 3076, + "GAAAAA": 3077, + "GAAAAT": 3078, + "GAAAAC": 3079, + "GAAAAG": 3080, + "GAAATA": 3081, + "GAAATT": 3082, + "GAAATC": 3083, + "GAAATG": 3084, + "GAAACA": 3085, + "GAAACT": 3086, + "GAAACC": 3087, + "GAAACG": 3088, + "GAAAGA": 3089, + "GAAAGT": 3090, + "GAAAGC": 3091, + "GAAAGG": 3092, + "GAATAA": 3093, + "GAATAT": 3094, + "GAATAC": 3095, + "GAATAG": 3096, + "GAATTA": 3097, + "GAATTT": 3098, + "GAATTC": 3099, + "GAATTG": 3100, + "GAATCA": 3101, + "GAATCT": 3102, + "GAATCC": 3103, + "GAATCG": 3104, + "GAATGA": 3105, + "GAATGT": 3106, + "GAATGC": 3107, + "GAATGG": 3108, + "GAACAA": 3109, + "GAACAT": 3110, + "GAACAC": 3111, + "GAACAG": 3112, + "GAACTA": 3113, + "GAACTT": 3114, + "GAACTC": 3115, + "GAACTG": 3116, + "GAACCA": 3117, + "GAACCT": 3118, + "GAACCC": 3119, + "GAACCG": 3120, + "GAACGA": 3121, + "GAACGT": 3122, + "GAACGC": 3123, + "GAACGG": 3124, + "GAAGAA": 3125, + "GAAGAT": 3126, + "GAAGAC": 3127, + "GAAGAG": 3128, + "GAAGTA": 3129, + "GAAGTT": 3130, + "GAAGTC": 3131, + "GAAGTG": 3132, + "GAAGCA": 3133, + "GAAGCT": 3134, + "GAAGCC": 3135, + "GAAGCG": 3136, + "GAAGGA": 3137, + "GAAGGT": 3138, + "GAAGGC": 3139, + "GAAGGG": 3140, + "GATAAA": 3141, + "GATAAT": 3142, + "GATAAC": 3143, + "GATAAG": 3144, + "GATATA": 3145, + "GATATT": 3146, + "GATATC": 3147, + "GATATG": 3148, + "GATACA": 3149, + "GATACT": 3150, + "GATACC": 3151, + "GATACG": 3152, + "GATAGA": 3153, + "GATAGT": 3154, + "GATAGC": 3155, + "GATAGG": 3156, + "GATTAA": 3157, + "GATTAT": 3158, + "GATTAC": 3159, + "GATTAG": 3160, + "GATTTA": 3161, + "GATTTT": 3162, + "GATTTC": 3163, + "GATTTG": 3164, + "GATTCA": 3165, + "GATTCT": 3166, + "GATTCC": 3167, + "GATTCG": 3168, + "GATTGA": 3169, + "GATTGT": 3170, + "GATTGC": 3171, + "GATTGG": 3172, + "GATCAA": 3173, + "GATCAT": 3174, + "GATCAC": 3175, + "GATCAG": 3176, + "GATCTA": 3177, + "GATCTT": 3178, + "GATCTC": 3179, + "GATCTG": 3180, + "GATCCA": 3181, + "GATCCT": 3182, + "GATCCC": 3183, + "GATCCG": 3184, + "GATCGA": 3185, + "GATCGT": 3186, + "GATCGC": 3187, + "GATCGG": 3188, + "GATGAA": 3189, + "GATGAT": 3190, + "GATGAC": 3191, + "GATGAG": 3192, + "GATGTA": 3193, + "GATGTT": 3194, + "GATGTC": 3195, + "GATGTG": 3196, + "GATGCA": 3197, + "GATGCT": 3198, + "GATGCC": 3199, + "GATGCG": 3200, + "GATGGA": 3201, + "GATGGT": 3202, + "GATGGC": 3203, + "GATGGG": 3204, + "GACAAA": 3205, + "GACAAT": 3206, + "GACAAC": 3207, + "GACAAG": 3208, + "GACATA": 3209, + "GACATT": 3210, + "GACATC": 3211, + "GACATG": 3212, + "GACACA": 3213, + "GACACT": 3214, + "GACACC": 3215, + "GACACG": 3216, + "GACAGA": 3217, + "GACAGT": 3218, + "GACAGC": 3219, + "GACAGG": 3220, + "GACTAA": 3221, + "GACTAT": 3222, + "GACTAC": 3223, + "GACTAG": 3224, + "GACTTA": 3225, + "GACTTT": 3226, + "GACTTC": 3227, + "GACTTG": 3228, + "GACTCA": 3229, + "GACTCT": 3230, + "GACTCC": 3231, + "GACTCG": 3232, + "GACTGA": 3233, + "GACTGT": 3234, + "GACTGC": 3235, + "GACTGG": 3236, + "GACCAA": 3237, + "GACCAT": 3238, + "GACCAC": 3239, + "GACCAG": 3240, + "GACCTA": 3241, + "GACCTT": 3242, + "GACCTC": 3243, + "GACCTG": 3244, + "GACCCA": 3245, + "GACCCT": 3246, + "GACCCC": 3247, + "GACCCG": 3248, + "GACCGA": 3249, + "GACCGT": 3250, + "GACCGC": 3251, + "GACCGG": 3252, + "GACGAA": 3253, + "GACGAT": 3254, + "GACGAC": 3255, + "GACGAG": 3256, + "GACGTA": 3257, + "GACGTT": 3258, + "GACGTC": 3259, + "GACGTG": 3260, + "GACGCA": 3261, + "GACGCT": 3262, + "GACGCC": 3263, + "GACGCG": 3264, + "GACGGA": 3265, + "GACGGT": 3266, + "GACGGC": 3267, + "GACGGG": 3268, + "GAGAAA": 3269, + "GAGAAT": 3270, + "GAGAAC": 3271, + "GAGAAG": 3272, + "GAGATA": 3273, + "GAGATT": 3274, + "GAGATC": 3275, + "GAGATG": 3276, + "GAGACA": 3277, + "GAGACT": 3278, + "GAGACC": 3279, + "GAGACG": 3280, + "GAGAGA": 3281, + "GAGAGT": 3282, + "GAGAGC": 3283, + "GAGAGG": 3284, + "GAGTAA": 3285, + "GAGTAT": 3286, + "GAGTAC": 3287, + "GAGTAG": 3288, + "GAGTTA": 3289, + "GAGTTT": 3290, + "GAGTTC": 3291, + "GAGTTG": 3292, + "GAGTCA": 3293, + "GAGTCT": 3294, + "GAGTCC": 3295, + "GAGTCG": 3296, + "GAGTGA": 3297, + "GAGTGT": 3298, + "GAGTGC": 3299, + "GAGTGG": 3300, + "GAGCAA": 3301, + "GAGCAT": 3302, + "GAGCAC": 3303, + "GAGCAG": 3304, + "GAGCTA": 3305, + "GAGCTT": 3306, + "GAGCTC": 3307, + "GAGCTG": 3308, + "GAGCCA": 3309, + "GAGCCT": 3310, + "GAGCCC": 3311, + "GAGCCG": 3312, + "GAGCGA": 3313, + "GAGCGT": 3314, + "GAGCGC": 3315, + "GAGCGG": 3316, + "GAGGAA": 3317, + "GAGGAT": 3318, + "GAGGAC": 3319, + "GAGGAG": 3320, + "GAGGTA": 3321, + "GAGGTT": 3322, + "GAGGTC": 3323, + "GAGGTG": 3324, + "GAGGCA": 3325, + "GAGGCT": 3326, + "GAGGCC": 3327, + "GAGGCG": 3328, + "GAGGGA": 3329, + "GAGGGT": 3330, + "GAGGGC": 3331, + "GAGGGG": 3332, + "GTAAAA": 3333, + "GTAAAT": 3334, + "GTAAAC": 3335, + "GTAAAG": 3336, + "GTAATA": 3337, + "GTAATT": 3338, + "GTAATC": 3339, + "GTAATG": 3340, + "GTAACA": 3341, + "GTAACT": 3342, + "GTAACC": 3343, + "GTAACG": 3344, + "GTAAGA": 3345, + "GTAAGT": 3346, + "GTAAGC": 3347, + "GTAAGG": 3348, + "GTATAA": 3349, + "GTATAT": 3350, + "GTATAC": 3351, + "GTATAG": 3352, + "GTATTA": 3353, + "GTATTT": 3354, + "GTATTC": 3355, + "GTATTG": 3356, + "GTATCA": 3357, + "GTATCT": 3358, + "GTATCC": 3359, + "GTATCG": 3360, + "GTATGA": 3361, + "GTATGT": 3362, + "GTATGC": 3363, + "GTATGG": 3364, + "GTACAA": 3365, + "GTACAT": 3366, + "GTACAC": 3367, + "GTACAG": 3368, + "GTACTA": 3369, + "GTACTT": 3370, + "GTACTC": 3371, + "GTACTG": 3372, + "GTACCA": 3373, + "GTACCT": 3374, + "GTACCC": 3375, + "GTACCG": 3376, + "GTACGA": 3377, + "GTACGT": 3378, + "GTACGC": 3379, + "GTACGG": 3380, + "GTAGAA": 3381, + "GTAGAT": 3382, + "GTAGAC": 3383, + "GTAGAG": 3384, + "GTAGTA": 3385, + "GTAGTT": 3386, + "GTAGTC": 3387, + "GTAGTG": 3388, + "GTAGCA": 3389, + "GTAGCT": 3390, + "GTAGCC": 3391, + "GTAGCG": 3392, + "GTAGGA": 3393, + "GTAGGT": 3394, + "GTAGGC": 3395, + "GTAGGG": 3396, + "GTTAAA": 3397, + "GTTAAT": 3398, + "GTTAAC": 3399, + "GTTAAG": 3400, + "GTTATA": 3401, + "GTTATT": 3402, + "GTTATC": 3403, + "GTTATG": 3404, + "GTTACA": 3405, + "GTTACT": 3406, + "GTTACC": 3407, + "GTTACG": 3408, + "GTTAGA": 3409, + "GTTAGT": 3410, + "GTTAGC": 3411, + "GTTAGG": 3412, + "GTTTAA": 3413, + "GTTTAT": 3414, + "GTTTAC": 3415, + "GTTTAG": 3416, + "GTTTTA": 3417, + "GTTTTT": 3418, + "GTTTTC": 3419, + "GTTTTG": 3420, + "GTTTCA": 3421, + "GTTTCT": 3422, + "GTTTCC": 3423, + "GTTTCG": 3424, + "GTTTGA": 3425, + "GTTTGT": 3426, + "GTTTGC": 3427, + "GTTTGG": 3428, + "GTTCAA": 3429, + "GTTCAT": 3430, + "GTTCAC": 3431, + "GTTCAG": 3432, + "GTTCTA": 3433, + "GTTCTT": 3434, + "GTTCTC": 3435, + "GTTCTG": 3436, + "GTTCCA": 3437, + "GTTCCT": 3438, + "GTTCCC": 3439, + "GTTCCG": 3440, + "GTTCGA": 3441, + "GTTCGT": 3442, + "GTTCGC": 3443, + "GTTCGG": 3444, + "GTTGAA": 3445, + "GTTGAT": 3446, + "GTTGAC": 3447, + "GTTGAG": 3448, + "GTTGTA": 3449, + "GTTGTT": 3450, + "GTTGTC": 3451, + "GTTGTG": 3452, + "GTTGCA": 3453, + "GTTGCT": 3454, + "GTTGCC": 3455, + "GTTGCG": 3456, + "GTTGGA": 3457, + "GTTGGT": 3458, + "GTTGGC": 3459, + "GTTGGG": 3460, + "GTCAAA": 3461, + "GTCAAT": 3462, + "GTCAAC": 3463, + "GTCAAG": 3464, + "GTCATA": 3465, + "GTCATT": 3466, + "GTCATC": 3467, + "GTCATG": 3468, + "GTCACA": 3469, + "GTCACT": 3470, + "GTCACC": 3471, + "GTCACG": 3472, + "GTCAGA": 3473, + "GTCAGT": 3474, + "GTCAGC": 3475, + "GTCAGG": 3476, + "GTCTAA": 3477, + "GTCTAT": 3478, + "GTCTAC": 3479, + "GTCTAG": 3480, + "GTCTTA": 3481, + "GTCTTT": 3482, + "GTCTTC": 3483, + "GTCTTG": 3484, + "GTCTCA": 3485, + "GTCTCT": 3486, + "GTCTCC": 3487, + "GTCTCG": 3488, + "GTCTGA": 3489, + "GTCTGT": 3490, + "GTCTGC": 3491, + "GTCTGG": 3492, + "GTCCAA": 3493, + "GTCCAT": 3494, + "GTCCAC": 3495, + "GTCCAG": 3496, + "GTCCTA": 3497, + "GTCCTT": 3498, + "GTCCTC": 3499, + "GTCCTG": 3500, + "GTCCCA": 3501, + "GTCCCT": 3502, + "GTCCCC": 3503, + "GTCCCG": 3504, + "GTCCGA": 3505, + "GTCCGT": 3506, + "GTCCGC": 3507, + "GTCCGG": 3508, + "GTCGAA": 3509, + "GTCGAT": 3510, + "GTCGAC": 3511, + "GTCGAG": 3512, + "GTCGTA": 3513, + "GTCGTT": 3514, + "GTCGTC": 3515, + "GTCGTG": 3516, + "GTCGCA": 3517, + "GTCGCT": 3518, + "GTCGCC": 3519, + "GTCGCG": 3520, + "GTCGGA": 3521, + "GTCGGT": 3522, + "GTCGGC": 3523, + "GTCGGG": 3524, + "GTGAAA": 3525, + "GTGAAT": 3526, + "GTGAAC": 3527, + "GTGAAG": 3528, + "GTGATA": 3529, + "GTGATT": 3530, + "GTGATC": 3531, + "GTGATG": 3532, + "GTGACA": 3533, + "GTGACT": 3534, + "GTGACC": 3535, + "GTGACG": 3536, + "GTGAGA": 3537, + "GTGAGT": 3538, + "GTGAGC": 3539, + "GTGAGG": 3540, + "GTGTAA": 3541, + "GTGTAT": 3542, + "GTGTAC": 3543, + "GTGTAG": 3544, + "GTGTTA": 3545, + "GTGTTT": 3546, + "GTGTTC": 3547, + "GTGTTG": 3548, + "GTGTCA": 3549, + "GTGTCT": 3550, + "GTGTCC": 3551, + "GTGTCG": 3552, + "GTGTGA": 3553, + "GTGTGT": 3554, + "GTGTGC": 3555, + "GTGTGG": 3556, + "GTGCAA": 3557, + "GTGCAT": 3558, + "GTGCAC": 3559, + "GTGCAG": 3560, + "GTGCTA": 3561, + "GTGCTT": 3562, + "GTGCTC": 3563, + "GTGCTG": 3564, + "GTGCCA": 3565, + "GTGCCT": 3566, + "GTGCCC": 3567, + "GTGCCG": 3568, + "GTGCGA": 3569, + "GTGCGT": 3570, + "GTGCGC": 3571, + "GTGCGG": 3572, + "GTGGAA": 3573, + "GTGGAT": 3574, + "GTGGAC": 3575, + "GTGGAG": 3576, + "GTGGTA": 3577, + "GTGGTT": 3578, + "GTGGTC": 3579, + "GTGGTG": 3580, + "GTGGCA": 3581, + "GTGGCT": 3582, + "GTGGCC": 3583, + "GTGGCG": 3584, + "GTGGGA": 3585, + "GTGGGT": 3586, + "GTGGGC": 3587, + "GTGGGG": 3588, + "GCAAAA": 3589, + "GCAAAT": 3590, + "GCAAAC": 3591, + "GCAAAG": 3592, + "GCAATA": 3593, + "GCAATT": 3594, + "GCAATC": 3595, + "GCAATG": 3596, + "GCAACA": 3597, + "GCAACT": 3598, + "GCAACC": 3599, + "GCAACG": 3600, + "GCAAGA": 3601, + "GCAAGT": 3602, + "GCAAGC": 3603, + "GCAAGG": 3604, + "GCATAA": 3605, + "GCATAT": 3606, + "GCATAC": 3607, + "GCATAG": 3608, + "GCATTA": 3609, + "GCATTT": 3610, + "GCATTC": 3611, + "GCATTG": 3612, + "GCATCA": 3613, + "GCATCT": 3614, + "GCATCC": 3615, + "GCATCG": 3616, + "GCATGA": 3617, + "GCATGT": 3618, + "GCATGC": 3619, + "GCATGG": 3620, + "GCACAA": 3621, + "GCACAT": 3622, + "GCACAC": 3623, + "GCACAG": 3624, + "GCACTA": 3625, + "GCACTT": 3626, + "GCACTC": 3627, + "GCACTG": 3628, + "GCACCA": 3629, + "GCACCT": 3630, + "GCACCC": 3631, + "GCACCG": 3632, + "GCACGA": 3633, + "GCACGT": 3634, + "GCACGC": 3635, + "GCACGG": 3636, + "GCAGAA": 3637, + "GCAGAT": 3638, + "GCAGAC": 3639, + "GCAGAG": 3640, + "GCAGTA": 3641, + "GCAGTT": 3642, + "GCAGTC": 3643, + "GCAGTG": 3644, + "GCAGCA": 3645, + "GCAGCT": 3646, + "GCAGCC": 3647, + "GCAGCG": 3648, + "GCAGGA": 3649, + "GCAGGT": 3650, + "GCAGGC": 3651, + "GCAGGG": 3652, + "GCTAAA": 3653, + "GCTAAT": 3654, + "GCTAAC": 3655, + "GCTAAG": 3656, + "GCTATA": 3657, + "GCTATT": 3658, + "GCTATC": 3659, + "GCTATG": 3660, + "GCTACA": 3661, + "GCTACT": 3662, + "GCTACC": 3663, + "GCTACG": 3664, + "GCTAGA": 3665, + "GCTAGT": 3666, + "GCTAGC": 3667, + "GCTAGG": 3668, + "GCTTAA": 3669, + "GCTTAT": 3670, + "GCTTAC": 3671, + "GCTTAG": 3672, + "GCTTTA": 3673, + "GCTTTT": 3674, + "GCTTTC": 3675, + "GCTTTG": 3676, + "GCTTCA": 3677, + "GCTTCT": 3678, + "GCTTCC": 3679, + "GCTTCG": 3680, + "GCTTGA": 3681, + "GCTTGT": 3682, + "GCTTGC": 3683, + "GCTTGG": 3684, + "GCTCAA": 3685, + "GCTCAT": 3686, + "GCTCAC": 3687, + "GCTCAG": 3688, + "GCTCTA": 3689, + "GCTCTT": 3690, + "GCTCTC": 3691, + "GCTCTG": 3692, + "GCTCCA": 3693, + "GCTCCT": 3694, + "GCTCCC": 3695, + "GCTCCG": 3696, + "GCTCGA": 3697, + "GCTCGT": 3698, + "GCTCGC": 3699, + "GCTCGG": 3700, + "GCTGAA": 3701, + "GCTGAT": 3702, + "GCTGAC": 3703, + "GCTGAG": 3704, + "GCTGTA": 3705, + "GCTGTT": 3706, + "GCTGTC": 3707, + "GCTGTG": 3708, + "GCTGCA": 3709, + "GCTGCT": 3710, + "GCTGCC": 3711, + "GCTGCG": 3712, + "GCTGGA": 3713, + "GCTGGT": 3714, + "GCTGGC": 3715, + "GCTGGG": 3716, + "GCCAAA": 3717, + "GCCAAT": 3718, + "GCCAAC": 3719, + "GCCAAG": 3720, + "GCCATA": 3721, + "GCCATT": 3722, + "GCCATC": 3723, + "GCCATG": 3724, + "GCCACA": 3725, + "GCCACT": 3726, + "GCCACC": 3727, + "GCCACG": 3728, + "GCCAGA": 3729, + "GCCAGT": 3730, + "GCCAGC": 3731, + "GCCAGG": 3732, + "GCCTAA": 3733, + "GCCTAT": 3734, + "GCCTAC": 3735, + "GCCTAG": 3736, + "GCCTTA": 3737, + "GCCTTT": 3738, + "GCCTTC": 3739, + "GCCTTG": 3740, + "GCCTCA": 3741, + "GCCTCT": 3742, + "GCCTCC": 3743, + "GCCTCG": 3744, + "GCCTGA": 3745, + "GCCTGT": 3746, + "GCCTGC": 3747, + "GCCTGG": 3748, + "GCCCAA": 3749, + "GCCCAT": 3750, + "GCCCAC": 3751, + "GCCCAG": 3752, + "GCCCTA": 3753, + "GCCCTT": 3754, + "GCCCTC": 3755, + "GCCCTG": 3756, + "GCCCCA": 3757, + "GCCCCT": 3758, + "GCCCCC": 3759, + "GCCCCG": 3760, + "GCCCGA": 3761, + "GCCCGT": 3762, + "GCCCGC": 3763, + "GCCCGG": 3764, + "GCCGAA": 3765, + "GCCGAT": 3766, + "GCCGAC": 3767, + "GCCGAG": 3768, + "GCCGTA": 3769, + "GCCGTT": 3770, + "GCCGTC": 3771, + "GCCGTG": 3772, + "GCCGCA": 3773, + "GCCGCT": 3774, + "GCCGCC": 3775, + "GCCGCG": 3776, + "GCCGGA": 3777, + "GCCGGT": 3778, + "GCCGGC": 3779, + "GCCGGG": 3780, + "GCGAAA": 3781, + "GCGAAT": 3782, + "GCGAAC": 3783, + "GCGAAG": 3784, + "GCGATA": 3785, + "GCGATT": 3786, + "GCGATC": 3787, + "GCGATG": 3788, + "GCGACA": 3789, + "GCGACT": 3790, + "GCGACC": 3791, + "GCGACG": 3792, + "GCGAGA": 3793, + "GCGAGT": 3794, + "GCGAGC": 3795, + "GCGAGG": 3796, + "GCGTAA": 3797, + "GCGTAT": 3798, + "GCGTAC": 3799, + "GCGTAG": 3800, + "GCGTTA": 3801, + "GCGTTT": 3802, + "GCGTTC": 3803, + "GCGTTG": 3804, + "GCGTCA": 3805, + "GCGTCT": 3806, + "GCGTCC": 3807, + "GCGTCG": 3808, + "GCGTGA": 3809, + "GCGTGT": 3810, + "GCGTGC": 3811, + "GCGTGG": 3812, + "GCGCAA": 3813, + "GCGCAT": 3814, + "GCGCAC": 3815, + "GCGCAG": 3816, + "GCGCTA": 3817, + "GCGCTT": 3818, + "GCGCTC": 3819, + "GCGCTG": 3820, + "GCGCCA": 3821, + "GCGCCT": 3822, + "GCGCCC": 3823, + "GCGCCG": 3824, + "GCGCGA": 3825, + "GCGCGT": 3826, + "GCGCGC": 3827, + "GCGCGG": 3828, + "GCGGAA": 3829, + "GCGGAT": 3830, + "GCGGAC": 3831, + "GCGGAG": 3832, + "GCGGTA": 3833, + "GCGGTT": 3834, + "GCGGTC": 3835, + "GCGGTG": 3836, + "GCGGCA": 3837, + "GCGGCT": 3838, + "GCGGCC": 3839, + "GCGGCG": 3840, + "GCGGGA": 3841, + "GCGGGT": 3842, + "GCGGGC": 3843, + "GCGGGG": 3844, + "GGAAAA": 3845, + "GGAAAT": 3846, + "GGAAAC": 3847, + "GGAAAG": 3848, + "GGAATA": 3849, + "GGAATT": 3850, + "GGAATC": 3851, + "GGAATG": 3852, + "GGAACA": 3853, + "GGAACT": 3854, + "GGAACC": 3855, + "GGAACG": 3856, + "GGAAGA": 3857, + "GGAAGT": 3858, + "GGAAGC": 3859, + "GGAAGG": 3860, + "GGATAA": 3861, + "GGATAT": 3862, + "GGATAC": 3863, + "GGATAG": 3864, + "GGATTA": 3865, + "GGATTT": 3866, + "GGATTC": 3867, + "GGATTG": 3868, + "GGATCA": 3869, + "GGATCT": 3870, + "GGATCC": 3871, + "GGATCG": 3872, + "GGATGA": 3873, + "GGATGT": 3874, + "GGATGC": 3875, + "GGATGG": 3876, + "GGACAA": 3877, + "GGACAT": 3878, + "GGACAC": 3879, + "GGACAG": 3880, + "GGACTA": 3881, + "GGACTT": 3882, + "GGACTC": 3883, + "GGACTG": 3884, + "GGACCA": 3885, + "GGACCT": 3886, + "GGACCC": 3887, + "GGACCG": 3888, + "GGACGA": 3889, + "GGACGT": 3890, + "GGACGC": 3891, + "GGACGG": 3892, + "GGAGAA": 3893, + "GGAGAT": 3894, + "GGAGAC": 3895, + "GGAGAG": 3896, + "GGAGTA": 3897, + "GGAGTT": 3898, + "GGAGTC": 3899, + "GGAGTG": 3900, + "GGAGCA": 3901, + "GGAGCT": 3902, + "GGAGCC": 3903, + "GGAGCG": 3904, + "GGAGGA": 3905, + "GGAGGT": 3906, + "GGAGGC": 3907, + "GGAGGG": 3908, + "GGTAAA": 3909, + "GGTAAT": 3910, + "GGTAAC": 3911, + "GGTAAG": 3912, + "GGTATA": 3913, + "GGTATT": 3914, + "GGTATC": 3915, + "GGTATG": 3916, + "GGTACA": 3917, + "GGTACT": 3918, + "GGTACC": 3919, + "GGTACG": 3920, + "GGTAGA": 3921, + "GGTAGT": 3922, + "GGTAGC": 3923, + "GGTAGG": 3924, + "GGTTAA": 3925, + "GGTTAT": 3926, + "GGTTAC": 3927, + "GGTTAG": 3928, + "GGTTTA": 3929, + "GGTTTT": 3930, + "GGTTTC": 3931, + "GGTTTG": 3932, + "GGTTCA": 3933, + "GGTTCT": 3934, + "GGTTCC": 3935, + "GGTTCG": 3936, + "GGTTGA": 3937, + "GGTTGT": 3938, + "GGTTGC": 3939, + "GGTTGG": 3940, + "GGTCAA": 3941, + "GGTCAT": 3942, + "GGTCAC": 3943, + "GGTCAG": 3944, + "GGTCTA": 3945, + "GGTCTT": 3946, + "GGTCTC": 3947, + "GGTCTG": 3948, + "GGTCCA": 3949, + "GGTCCT": 3950, + "GGTCCC": 3951, + "GGTCCG": 3952, + "GGTCGA": 3953, + "GGTCGT": 3954, + "GGTCGC": 3955, + "GGTCGG": 3956, + "GGTGAA": 3957, + "GGTGAT": 3958, + "GGTGAC": 3959, + "GGTGAG": 3960, + "GGTGTA": 3961, + "GGTGTT": 3962, + "GGTGTC": 3963, + "GGTGTG": 3964, + "GGTGCA": 3965, + "GGTGCT": 3966, + "GGTGCC": 3967, + "GGTGCG": 3968, + "GGTGGA": 3969, + "GGTGGT": 3970, + "GGTGGC": 3971, + "GGTGGG": 3972, + "GGCAAA": 3973, + "GGCAAT": 3974, + "GGCAAC": 3975, + "GGCAAG": 3976, + "GGCATA": 3977, + "GGCATT": 3978, + "GGCATC": 3979, + "GGCATG": 3980, + "GGCACA": 3981, + "GGCACT": 3982, + "GGCACC": 3983, + "GGCACG": 3984, + "GGCAGA": 3985, + "GGCAGT": 3986, + "GGCAGC": 3987, + "GGCAGG": 3988, + "GGCTAA": 3989, + "GGCTAT": 3990, + "GGCTAC": 3991, + "GGCTAG": 3992, + "GGCTTA": 3993, + "GGCTTT": 3994, + "GGCTTC": 3995, + "GGCTTG": 3996, + "GGCTCA": 3997, + "GGCTCT": 3998, + "GGCTCC": 3999, + "GGCTCG": 4000, + "GGCTGA": 4001, + "GGCTGT": 4002, + "GGCTGC": 4003, + "GGCTGG": 4004, + "GGCCAA": 4005, + "GGCCAT": 4006, + "GGCCAC": 4007, + "GGCCAG": 4008, + "GGCCTA": 4009, + "GGCCTT": 4010, + "GGCCTC": 4011, + "GGCCTG": 4012, + "GGCCCA": 4013, + "GGCCCT": 4014, + "GGCCCC": 4015, + "GGCCCG": 4016, + "GGCCGA": 4017, + "GGCCGT": 4018, + "GGCCGC": 4019, + "GGCCGG": 4020, + "GGCGAA": 4021, + "GGCGAT": 4022, + "GGCGAC": 4023, + "GGCGAG": 4024, + "GGCGTA": 4025, + "GGCGTT": 4026, + "GGCGTC": 4027, + "GGCGTG": 4028, + "GGCGCA": 4029, + "GGCGCT": 4030, + "GGCGCC": 4031, + "GGCGCG": 4032, + "GGCGGA": 4033, + "GGCGGT": 4034, + "GGCGGC": 4035, + "GGCGGG": 4036, + "GGGAAA": 4037, + "GGGAAT": 4038, + "GGGAAC": 4039, + "GGGAAG": 4040, + "GGGATA": 4041, + "GGGATT": 4042, + "GGGATC": 4043, + "GGGATG": 4044, + "GGGACA": 4045, + "GGGACT": 4046, + "GGGACC": 4047, + "GGGACG": 4048, + "GGGAGA": 4049, + "GGGAGT": 4050, + "GGGAGC": 4051, + "GGGAGG": 4052, + "GGGTAA": 4053, + "GGGTAT": 4054, + "GGGTAC": 4055, + "GGGTAG": 4056, + "GGGTTA": 4057, + "GGGTTT": 4058, + "GGGTTC": 4059, + "GGGTTG": 4060, + "GGGTCA": 4061, + "GGGTCT": 4062, + "GGGTCC": 4063, + "GGGTCG": 4064, + "GGGTGA": 4065, + "GGGTGT": 4066, + "GGGTGC": 4067, + "GGGTGG": 4068, + "GGGCAA": 4069, + "GGGCAT": 4070, + "GGGCAC": 4071, + "GGGCAG": 4072, + "GGGCTA": 4073, + "GGGCTT": 4074, + "GGGCTC": 4075, + "GGGCTG": 4076, + "GGGCCA": 4077, + "GGGCCT": 4078, + "GGGCCC": 4079, + "GGGCCG": 4080, + "GGGCGA": 4081, + "GGGCGT": 4082, + "GGGCGC": 4083, + "GGGCGG": 4084, + "GGGGAA": 4085, + "GGGGAT": 4086, + "GGGGAC": 4087, + "GGGGAG": 4088, + "GGGGTA": 4089, + "GGGGTT": 4090, + "GGGGTC": 4091, + "GGGGTG": 4092, + "GGGGCA": 4093, + "GGGGCT": 4094, + "GGGGCC": 4095, + "GGGGCG": 4096, + "GGGGGA": 4097, + "GGGGGT": 4098, + "GGGGGC": 4099, + "GGGGGG": 4100 + } + } +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-5000/tokenizer_config.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-5000/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..604e2b1b29af1dcb3920d2828b6a90276e5e3d02 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-5000/tokenizer_config.json @@ -0,0 +1,58 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "[UNK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "[CLS]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "[SEP]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "4": { + "content": "[MASK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "clean_up_tokenization_spaces": true, + "cls_token": "[CLS]", + "do_basic_tokenize": true, + "do_lower_case": false, + "mask_token": "[MASK]", + "max_len": 512, + "model_max_length": 512, + "never_split": null, + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "strip_accents": null, + "tokenize_chinese_chars": true, + "tokenizer_class": "BertTokenizer", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-5000/trainer_state.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-5000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..992555f5bc7b26925e27e15a5f3d88955e91e11f --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-5000/trainer_state.json @@ -0,0 +1,1433 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 20.0, + "eval_steps": 500, + "global_step": 5000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.1, + "grad_norm": 0.6038870215415955, + "learning_rate": 5.319148936170213e-05, + "loss": 0.6945, + "step": 25 + }, + { + "epoch": 0.2, + "grad_norm": 0.6668091416358948, + "learning_rate": 0.00010638297872340425, + "loss": 0.6929, + "step": 50 + }, + { + "epoch": 0.3, + "grad_norm": 0.8282511234283447, + "learning_rate": 0.00015957446808510637, + "loss": 0.6925, + "step": 75 + }, + { + "epoch": 0.4, + "grad_norm": 1.3149235248565674, + "learning_rate": 0.0002127659574468085, + "loss": 0.6919, + "step": 100 + }, + { + "epoch": 0.5, + "grad_norm": 1.8841172456741333, + "learning_rate": 0.00026595744680851064, + "loss": 0.6788, + "step": 125 + }, + { + "epoch": 0.6, + "grad_norm": 4.435206413269043, + "learning_rate": 0.00031914893617021275, + "loss": 0.6285, + "step": 150 + }, + { + "epoch": 0.7, + "grad_norm": 8.858148574829102, + "learning_rate": 0.0003723404255319149, + "loss": 0.5386, + "step": 175 + }, + { + "epoch": 0.8, + "grad_norm": 5.627648830413818, + "learning_rate": 0.0003992081821181128, + "loss": 0.5537, + "step": 200 + }, + { + "epoch": 0.9, + "grad_norm": 5.086580753326416, + "learning_rate": 0.0003975585615308479, + "loss": 0.5595, + "step": 225 + }, + { + "epoch": 1.0, + "grad_norm": 5.627719402313232, + "learning_rate": 0.00039590894094358297, + "loss": 0.4824, + "step": 250 + }, + { + "epoch": 1.1, + "grad_norm": 5.888038158416748, + "learning_rate": 0.0003942593203563181, + "loss": 0.4404, + "step": 275 + }, + { + "epoch": 1.2, + "grad_norm": 5.646231651306152, + "learning_rate": 0.0003926096997690532, + "loss": 0.4354, + "step": 300 + }, + { + "epoch": 1.3, + "grad_norm": 5.191990852355957, + "learning_rate": 0.00039096007918178817, + "loss": 0.4469, + "step": 325 + }, + { + "epoch": 1.4, + "grad_norm": 13.90101432800293, + "learning_rate": 0.0003893104585945233, + "loss": 0.3858, + "step": 350 + }, + { + "epoch": 1.5, + "grad_norm": 17.75942039489746, + "learning_rate": 0.0003876608380072583, + "loss": 0.4131, + "step": 375 + }, + { + "epoch": 1.6, + "grad_norm": 7.398252487182617, + "learning_rate": 0.00038601121741999343, + "loss": 0.3842, + "step": 400 + }, + { + "epoch": 1.7, + "grad_norm": 11.010066986083984, + "learning_rate": 0.0003843615968327285, + "loss": 0.3959, + "step": 425 + }, + { + "epoch": 1.8, + "grad_norm": 9.277341842651367, + "learning_rate": 0.0003827119762454636, + "loss": 0.3552, + "step": 450 + }, + { + "epoch": 1.9, + "grad_norm": 26.011449813842773, + "learning_rate": 0.00038106235565819863, + "loss": 0.3961, + "step": 475 + }, + { + "epoch": 2.0, + "grad_norm": 11.123054504394531, + "learning_rate": 0.0003794127350709337, + "loss": 0.3548, + "step": 500 + }, + { + "epoch": 2.1, + "grad_norm": 20.128774642944336, + "learning_rate": 0.0003777631144836688, + "loss": 0.302, + "step": 525 + }, + { + "epoch": 2.2, + "grad_norm": 4.64390230178833, + "learning_rate": 0.00037611349389640383, + "loss": 0.2853, + "step": 550 + }, + { + "epoch": 2.3, + "grad_norm": 4.199934482574463, + "learning_rate": 0.00037446387330913894, + "loss": 0.2724, + "step": 575 + }, + { + "epoch": 2.4, + "grad_norm": 6.909411430358887, + "learning_rate": 0.000372814252721874, + "loss": 0.2976, + "step": 600 + }, + { + "epoch": 2.5, + "grad_norm": 9.234820365905762, + "learning_rate": 0.00037116463213460903, + "loss": 0.2798, + "step": 625 + }, + { + "epoch": 2.6, + "grad_norm": 10.472943305969238, + "learning_rate": 0.00036951501154734414, + "loss": 0.3007, + "step": 650 + }, + { + "epoch": 2.7, + "grad_norm": 19.46732521057129, + "learning_rate": 0.0003678653909600792, + "loss": 0.2701, + "step": 675 + }, + { + "epoch": 2.8, + "grad_norm": 7.501255512237549, + "learning_rate": 0.0003662157703728143, + "loss": 0.2896, + "step": 700 + }, + { + "epoch": 2.9, + "grad_norm": 10.645585060119629, + "learning_rate": 0.00036456614978554934, + "loss": 0.2748, + "step": 725 + }, + { + "epoch": 3.0, + "grad_norm": 30.26616859436035, + "learning_rate": 0.00036291652919828444, + "loss": 0.268, + "step": 750 + }, + { + "epoch": 3.1, + "grad_norm": 14.13516616821289, + "learning_rate": 0.00036126690861101944, + "loss": 0.2159, + "step": 775 + }, + { + "epoch": 3.2, + "grad_norm": 10.091023445129395, + "learning_rate": 0.00035961728802375454, + "loss": 0.2071, + "step": 800 + }, + { + "epoch": 3.3, + "grad_norm": 22.201452255249023, + "learning_rate": 0.0003579676674364896, + "loss": 0.2146, + "step": 825 + }, + { + "epoch": 3.4, + "grad_norm": 10.044412612915039, + "learning_rate": 0.0003563180468492247, + "loss": 0.1827, + "step": 850 + }, + { + "epoch": 3.5, + "grad_norm": 21.04941177368164, + "learning_rate": 0.0003546684262619598, + "loss": 0.202, + "step": 875 + }, + { + "epoch": 3.6, + "grad_norm": 11.754891395568848, + "learning_rate": 0.00035301880567469485, + "loss": 0.2021, + "step": 900 + }, + { + "epoch": 3.7, + "grad_norm": 9.113747596740723, + "learning_rate": 0.0003513691850874299, + "loss": 0.2067, + "step": 925 + }, + { + "epoch": 3.8, + "grad_norm": 9.075196266174316, + "learning_rate": 0.00034971956450016495, + "loss": 0.3285, + "step": 950 + }, + { + "epoch": 3.9, + "grad_norm": 8.568796157836914, + "learning_rate": 0.00034806994391290005, + "loss": 0.2254, + "step": 975 + }, + { + "epoch": 4.0, + "grad_norm": 3.9587976932525635, + "learning_rate": 0.0003464203233256351, + "loss": 0.2245, + "step": 1000 + }, + { + "epoch": 4.1, + "grad_norm": 11.143651008605957, + "learning_rate": 0.0003447707027383702, + "loss": 0.1755, + "step": 1025 + }, + { + "epoch": 4.2, + "grad_norm": 12.53118896484375, + "learning_rate": 0.00034312108215110525, + "loss": 0.1808, + "step": 1050 + }, + { + "epoch": 4.3, + "grad_norm": 9.880820274353027, + "learning_rate": 0.0003414714615638403, + "loss": 0.1387, + "step": 1075 + }, + { + "epoch": 4.4, + "grad_norm": 5.729593753814697, + "learning_rate": 0.0003398218409765754, + "loss": 0.1958, + "step": 1100 + }, + { + "epoch": 4.5, + "grad_norm": 8.633047103881836, + "learning_rate": 0.00033817222038931045, + "loss": 0.1515, + "step": 1125 + }, + { + "epoch": 4.6, + "grad_norm": 19.237892150878906, + "learning_rate": 0.00033652259980204556, + "loss": 0.1452, + "step": 1150 + }, + { + "epoch": 4.7, + "grad_norm": 6.129818439483643, + "learning_rate": 0.0003348729792147806, + "loss": 0.1617, + "step": 1175 + }, + { + "epoch": 4.8, + "grad_norm": 5.073274612426758, + "learning_rate": 0.0003332233586275157, + "loss": 0.1967, + "step": 1200 + }, + { + "epoch": 4.9, + "grad_norm": 5.729922771453857, + "learning_rate": 0.00033157373804025076, + "loss": 0.1482, + "step": 1225 + }, + { + "epoch": 5.0, + "grad_norm": 12.314518928527832, + "learning_rate": 0.0003299241174529858, + "loss": 0.1588, + "step": 1250 + }, + { + "epoch": 5.1, + "grad_norm": 3.2636606693267822, + "learning_rate": 0.0003282744968657209, + "loss": 0.1588, + "step": 1275 + }, + { + "epoch": 5.2, + "grad_norm": 5.914486408233643, + "learning_rate": 0.00032662487627845596, + "loss": 0.1288, + "step": 1300 + }, + { + "epoch": 5.3, + "grad_norm": 13.849296569824219, + "learning_rate": 0.00032497525569119106, + "loss": 0.0961, + "step": 1325 + }, + { + "epoch": 5.4, + "grad_norm": 7.7317328453063965, + "learning_rate": 0.0003233256351039261, + "loss": 0.1148, + "step": 1350 + }, + { + "epoch": 5.5, + "grad_norm": 1.768608808517456, + "learning_rate": 0.00032167601451666116, + "loss": 0.121, + "step": 1375 + }, + { + "epoch": 5.6, + "grad_norm": 10.802785873413086, + "learning_rate": 0.00032002639392939627, + "loss": 0.1786, + "step": 1400 + }, + { + "epoch": 5.7, + "grad_norm": 4.158301830291748, + "learning_rate": 0.0003183767733421313, + "loss": 0.1078, + "step": 1425 + }, + { + "epoch": 5.8, + "grad_norm": 10.65804672241211, + "learning_rate": 0.0003167271527548664, + "loss": 0.1597, + "step": 1450 + }, + { + "epoch": 5.9, + "grad_norm": 10.880438804626465, + "learning_rate": 0.00031507753216760147, + "loss": 0.1225, + "step": 1475 + }, + { + "epoch": 6.0, + "grad_norm": 9.95138168334961, + "learning_rate": 0.00031342791158033657, + "loss": 0.1254, + "step": 1500 + }, + { + "epoch": 6.1, + "grad_norm": 1.5018162727355957, + "learning_rate": 0.00031177829099307157, + "loss": 0.0902, + "step": 1525 + }, + { + "epoch": 6.2, + "grad_norm": 14.281820297241211, + "learning_rate": 0.00031012867040580667, + "loss": 0.1026, + "step": 1550 + }, + { + "epoch": 6.3, + "grad_norm": 4.7540435791015625, + "learning_rate": 0.0003084790498185417, + "loss": 0.0876, + "step": 1575 + }, + { + "epoch": 6.4, + "grad_norm": 4.304183006286621, + "learning_rate": 0.0003068294292312768, + "loss": 0.1355, + "step": 1600 + }, + { + "epoch": 6.5, + "grad_norm": 14.653755187988281, + "learning_rate": 0.0003051798086440119, + "loss": 0.0878, + "step": 1625 + }, + { + "epoch": 6.6, + "grad_norm": 11.13028335571289, + "learning_rate": 0.000303530188056747, + "loss": 0.1203, + "step": 1650 + }, + { + "epoch": 6.7, + "grad_norm": 9.729938507080078, + "learning_rate": 0.000301880567469482, + "loss": 0.0775, + "step": 1675 + }, + { + "epoch": 6.8, + "grad_norm": 10.12938117980957, + "learning_rate": 0.0003002309468822171, + "loss": 0.1015, + "step": 1700 + }, + { + "epoch": 6.9, + "grad_norm": 8.506614685058594, + "learning_rate": 0.0002985813262949522, + "loss": 0.1362, + "step": 1725 + }, + { + "epoch": 7.0, + "grad_norm": 25.618762969970703, + "learning_rate": 0.00029693170570768723, + "loss": 0.1269, + "step": 1750 + }, + { + "epoch": 7.1, + "grad_norm": 27.155860900878906, + "learning_rate": 0.00029528208512042233, + "loss": 0.0555, + "step": 1775 + }, + { + "epoch": 7.2, + "grad_norm": 9.412527084350586, + "learning_rate": 0.0002936324645331574, + "loss": 0.0695, + "step": 1800 + }, + { + "epoch": 7.3, + "grad_norm": 6.4328293800354, + "learning_rate": 0.00029198284394589243, + "loss": 0.0808, + "step": 1825 + }, + { + "epoch": 7.4, + "grad_norm": 15.267293930053711, + "learning_rate": 0.00029033322335862753, + "loss": 0.0871, + "step": 1850 + }, + { + "epoch": 7.5, + "grad_norm": 6.774560928344727, + "learning_rate": 0.0002886836027713626, + "loss": 0.0819, + "step": 1875 + }, + { + "epoch": 7.6, + "grad_norm": 18.67180633544922, + "learning_rate": 0.0002870339821840977, + "loss": 0.0565, + "step": 1900 + }, + { + "epoch": 7.7, + "grad_norm": 16.39004898071289, + "learning_rate": 0.00028538436159683273, + "loss": 0.0907, + "step": 1925 + }, + { + "epoch": 7.8, + "grad_norm": 6.351583003997803, + "learning_rate": 0.00028373474100956784, + "loss": 0.1289, + "step": 1950 + }, + { + "epoch": 7.9, + "grad_norm": 11.396740913391113, + "learning_rate": 0.0002820851204223029, + "loss": 0.1149, + "step": 1975 + }, + { + "epoch": 8.0, + "grad_norm": 6.491570949554443, + "learning_rate": 0.00028043549983503794, + "loss": 0.0891, + "step": 2000 + }, + { + "epoch": 8.1, + "grad_norm": 0.22946956753730774, + "learning_rate": 0.00027878587924777304, + "loss": 0.0413, + "step": 2025 + }, + { + "epoch": 8.2, + "grad_norm": 42.31606674194336, + "learning_rate": 0.0002771362586605081, + "loss": 0.0881, + "step": 2050 + }, + { + "epoch": 8.3, + "grad_norm": 18.490406036376953, + "learning_rate": 0.0002754866380732432, + "loss": 0.1238, + "step": 2075 + }, + { + "epoch": 8.4, + "grad_norm": 17.950319290161133, + "learning_rate": 0.00027383701748597824, + "loss": 0.0642, + "step": 2100 + }, + { + "epoch": 8.5, + "grad_norm": 37.16359329223633, + "learning_rate": 0.0002721873968987133, + "loss": 0.0916, + "step": 2125 + }, + { + "epoch": 8.6, + "grad_norm": 39.75199890136719, + "learning_rate": 0.00027053777631144834, + "loss": 0.1087, + "step": 2150 + }, + { + "epoch": 8.7, + "grad_norm": 0.08794238418340683, + "learning_rate": 0.00026888815572418344, + "loss": 0.0825, + "step": 2175 + }, + { + "epoch": 8.8, + "grad_norm": 14.681429862976074, + "learning_rate": 0.00026723853513691855, + "loss": 0.0925, + "step": 2200 + }, + { + "epoch": 8.9, + "grad_norm": 17.90182876586914, + "learning_rate": 0.0002655889145496536, + "loss": 0.0933, + "step": 2225 + }, + { + "epoch": 9.0, + "grad_norm": 0.9108687043190002, + "learning_rate": 0.0002639392939623887, + "loss": 0.0602, + "step": 2250 + }, + { + "epoch": 9.1, + "grad_norm": 9.671142578125, + "learning_rate": 0.0002622896733751237, + "loss": 0.0438, + "step": 2275 + }, + { + "epoch": 9.2, + "grad_norm": 12.797035217285156, + "learning_rate": 0.0002606400527878588, + "loss": 0.0495, + "step": 2300 + }, + { + "epoch": 9.3, + "grad_norm": 0.4585617184638977, + "learning_rate": 0.00025899043220059385, + "loss": 0.032, + "step": 2325 + }, + { + "epoch": 9.4, + "grad_norm": 0.05073042958974838, + "learning_rate": 0.00025734081161332895, + "loss": 0.067, + "step": 2350 + }, + { + "epoch": 9.5, + "grad_norm": 0.013653877191245556, + "learning_rate": 0.000255691191026064, + "loss": 0.0898, + "step": 2375 + }, + { + "epoch": 9.6, + "grad_norm": 18.198965072631836, + "learning_rate": 0.0002540415704387991, + "loss": 0.0804, + "step": 2400 + }, + { + "epoch": 9.7, + "grad_norm": 0.1363959163427353, + "learning_rate": 0.00025239194985153415, + "loss": 0.0584, + "step": 2425 + }, + { + "epoch": 9.8, + "grad_norm": 17.61290740966797, + "learning_rate": 0.0002507423292642692, + "loss": 0.0778, + "step": 2450 + }, + { + "epoch": 9.9, + "grad_norm": 0.10548887401819229, + "learning_rate": 0.0002490927086770043, + "loss": 0.0667, + "step": 2475 + }, + { + "epoch": 10.0, + "grad_norm": 0.1262211948633194, + "learning_rate": 0.00024744308808973936, + "loss": 0.0625, + "step": 2500 + }, + { + "epoch": 10.1, + "grad_norm": 0.019293297082185745, + "learning_rate": 0.00024579346750247446, + "loss": 0.0762, + "step": 2525 + }, + { + "epoch": 10.2, + "grad_norm": 0.9273566603660583, + "learning_rate": 0.0002441438469152095, + "loss": 0.0357, + "step": 2550 + }, + { + "epoch": 10.3, + "grad_norm": 10.720550537109375, + "learning_rate": 0.00024249422632794456, + "loss": 0.046, + "step": 2575 + }, + { + "epoch": 10.4, + "grad_norm": 1.8441611528396606, + "learning_rate": 0.00024084460574067963, + "loss": 0.0369, + "step": 2600 + }, + { + "epoch": 10.5, + "grad_norm": 0.8760197758674622, + "learning_rate": 0.0002391949851534147, + "loss": 0.046, + "step": 2625 + }, + { + "epoch": 10.6, + "grad_norm": 8.899499893188477, + "learning_rate": 0.0002375453645661498, + "loss": 0.08, + "step": 2650 + }, + { + "epoch": 10.7, + "grad_norm": 0.2664288580417633, + "learning_rate": 0.00023589574397888486, + "loss": 0.0616, + "step": 2675 + }, + { + "epoch": 10.8, + "grad_norm": 0.06609692424535751, + "learning_rate": 0.00023424612339161997, + "loss": 0.05, + "step": 2700 + }, + { + "epoch": 10.9, + "grad_norm": 32.44829177856445, + "learning_rate": 0.000232596502804355, + "loss": 0.0487, + "step": 2725 + }, + { + "epoch": 11.0, + "grad_norm": 7.25365686416626, + "learning_rate": 0.00023094688221709007, + "loss": 0.0456, + "step": 2750 + }, + { + "epoch": 11.1, + "grad_norm": 0.10101272910833359, + "learning_rate": 0.00022929726162982514, + "loss": 0.0328, + "step": 2775 + }, + { + "epoch": 11.2, + "grad_norm": 1.3087996244430542, + "learning_rate": 0.00022764764104256022, + "loss": 0.0606, + "step": 2800 + }, + { + "epoch": 11.3, + "grad_norm": 1.247656226158142, + "learning_rate": 0.0002259980204552953, + "loss": 0.0838, + "step": 2825 + }, + { + "epoch": 11.4, + "grad_norm": 1.7708797454833984, + "learning_rate": 0.00022434839986803037, + "loss": 0.0416, + "step": 2850 + }, + { + "epoch": 11.5, + "grad_norm": 11.131721496582031, + "learning_rate": 0.00022269877928076542, + "loss": 0.0528, + "step": 2875 + }, + { + "epoch": 11.6, + "grad_norm": 3.824920892715454, + "learning_rate": 0.0002210491586935005, + "loss": 0.0526, + "step": 2900 + }, + { + "epoch": 11.7, + "grad_norm": 0.8507682085037231, + "learning_rate": 0.00021939953810623557, + "loss": 0.0965, + "step": 2925 + }, + { + "epoch": 11.8, + "grad_norm": 0.011997719295322895, + "learning_rate": 0.00021774991751897065, + "loss": 0.0446, + "step": 2950 + }, + { + "epoch": 11.9, + "grad_norm": 0.0792650505900383, + "learning_rate": 0.00021610029693170573, + "loss": 0.0556, + "step": 2975 + }, + { + "epoch": 12.0, + "grad_norm": 0.007047992665320635, + "learning_rate": 0.0002144506763444408, + "loss": 0.0759, + "step": 3000 + }, + { + "epoch": 12.1, + "grad_norm": 0.7784481644630432, + "learning_rate": 0.00021280105575717582, + "loss": 0.0505, + "step": 3025 + }, + { + "epoch": 12.2, + "grad_norm": 9.635031700134277, + "learning_rate": 0.00021115143516991093, + "loss": 0.0284, + "step": 3050 + }, + { + "epoch": 12.3, + "grad_norm": 12.473057746887207, + "learning_rate": 0.000209501814582646, + "loss": 0.0531, + "step": 3075 + }, + { + "epoch": 12.4, + "grad_norm": 10.904062271118164, + "learning_rate": 0.00020785219399538108, + "loss": 0.0212, + "step": 3100 + }, + { + "epoch": 12.5, + "grad_norm": 9.258865356445312, + "learning_rate": 0.00020620257340811616, + "loss": 0.0367, + "step": 3125 + }, + { + "epoch": 12.6, + "grad_norm": 2.6826488971710205, + "learning_rate": 0.00020455295282085123, + "loss": 0.0457, + "step": 3150 + }, + { + "epoch": 12.7, + "grad_norm": 8.800697326660156, + "learning_rate": 0.00020290333223358626, + "loss": 0.0504, + "step": 3175 + }, + { + "epoch": 12.8, + "grad_norm": 0.023783667013049126, + "learning_rate": 0.00020125371164632133, + "loss": 0.0255, + "step": 3200 + }, + { + "epoch": 12.9, + "grad_norm": 5.948949337005615, + "learning_rate": 0.0001996040910590564, + "loss": 0.0413, + "step": 3225 + }, + { + "epoch": 13.0, + "grad_norm": 0.026645230129361153, + "learning_rate": 0.00019795447047179148, + "loss": 0.0249, + "step": 3250 + }, + { + "epoch": 13.1, + "grad_norm": 60.826637268066406, + "learning_rate": 0.0001963048498845266, + "loss": 0.0293, + "step": 3275 + }, + { + "epoch": 13.2, + "grad_norm": 26.056686401367188, + "learning_rate": 0.00019465522929726164, + "loss": 0.0638, + "step": 3300 + }, + { + "epoch": 13.3, + "grad_norm": 18.143638610839844, + "learning_rate": 0.00019300560870999671, + "loss": 0.0206, + "step": 3325 + }, + { + "epoch": 13.4, + "grad_norm": 27.899967193603516, + "learning_rate": 0.0001913559881227318, + "loss": 0.0423, + "step": 3350 + }, + { + "epoch": 13.5, + "grad_norm": 51.9879264831543, + "learning_rate": 0.00018970636753546684, + "loss": 0.0274, + "step": 3375 + }, + { + "epoch": 13.6, + "grad_norm": 0.028627146035432816, + "learning_rate": 0.00018805674694820192, + "loss": 0.0513, + "step": 3400 + }, + { + "epoch": 13.7, + "grad_norm": 33.211944580078125, + "learning_rate": 0.000186407126360937, + "loss": 0.0446, + "step": 3425 + }, + { + "epoch": 13.8, + "grad_norm": 32.37785339355469, + "learning_rate": 0.00018475750577367207, + "loss": 0.0538, + "step": 3450 + }, + { + "epoch": 13.9, + "grad_norm": 30.558673858642578, + "learning_rate": 0.00018310788518640715, + "loss": 0.0339, + "step": 3475 + }, + { + "epoch": 14.0, + "grad_norm": 19.316938400268555, + "learning_rate": 0.00018145826459914222, + "loss": 0.071, + "step": 3500 + }, + { + "epoch": 14.1, + "grad_norm": 0.1271055042743683, + "learning_rate": 0.00017980864401187727, + "loss": 0.0159, + "step": 3525 + }, + { + "epoch": 14.2, + "grad_norm": 51.1268196105957, + "learning_rate": 0.00017815902342461235, + "loss": 0.028, + "step": 3550 + }, + { + "epoch": 14.3, + "grad_norm": 0.003214583033695817, + "learning_rate": 0.00017650940283734742, + "loss": 0.0342, + "step": 3575 + }, + { + "epoch": 14.4, + "grad_norm": 94.75691223144531, + "learning_rate": 0.00017485978225008247, + "loss": 0.0243, + "step": 3600 + }, + { + "epoch": 14.5, + "grad_norm": 0.019844651222229004, + "learning_rate": 0.00017321016166281755, + "loss": 0.0205, + "step": 3625 + }, + { + "epoch": 14.6, + "grad_norm": 0.10069809854030609, + "learning_rate": 0.00017156054107555263, + "loss": 0.0402, + "step": 3650 + }, + { + "epoch": 14.7, + "grad_norm": 0.0023937481455504894, + "learning_rate": 0.0001699109204882877, + "loss": 0.0173, + "step": 3675 + }, + { + "epoch": 14.8, + "grad_norm": 0.2573629319667816, + "learning_rate": 0.00016826129990102278, + "loss": 0.0312, + "step": 3700 + }, + { + "epoch": 14.9, + "grad_norm": 0.0067031364887952805, + "learning_rate": 0.00016661167931375785, + "loss": 0.0418, + "step": 3725 + }, + { + "epoch": 15.0, + "grad_norm": 0.08329538255929947, + "learning_rate": 0.0001649620587264929, + "loss": 0.0164, + "step": 3750 + }, + { + "epoch": 15.1, + "grad_norm": 0.04175519943237305, + "learning_rate": 0.00016331243813922798, + "loss": 0.044, + "step": 3775 + }, + { + "epoch": 15.2, + "grad_norm": 0.005420651752501726, + "learning_rate": 0.00016166281755196306, + "loss": 0.0599, + "step": 3800 + }, + { + "epoch": 15.3, + "grad_norm": 20.68973731994629, + "learning_rate": 0.00016001319696469813, + "loss": 0.047, + "step": 3825 + }, + { + "epoch": 15.4, + "grad_norm": 1.1241929531097412, + "learning_rate": 0.0001583635763774332, + "loss": 0.0184, + "step": 3850 + }, + { + "epoch": 15.5, + "grad_norm": 0.00928444042801857, + "learning_rate": 0.00015671395579016829, + "loss": 0.0146, + "step": 3875 + }, + { + "epoch": 15.6, + "grad_norm": 3.2419674396514893, + "learning_rate": 0.00015506433520290334, + "loss": 0.0047, + "step": 3900 + }, + { + "epoch": 15.7, + "grad_norm": 0.0019528602715581656, + "learning_rate": 0.0001534147146156384, + "loss": 0.0177, + "step": 3925 + }, + { + "epoch": 15.8, + "grad_norm": 0.003142274683341384, + "learning_rate": 0.0001517650940283735, + "loss": 0.0319, + "step": 3950 + }, + { + "epoch": 15.9, + "grad_norm": 45.58081817626953, + "learning_rate": 0.00015011547344110854, + "loss": 0.0353, + "step": 3975 + }, + { + "epoch": 16.0, + "grad_norm": 0.003515947377309203, + "learning_rate": 0.00014846585285384361, + "loss": 0.0208, + "step": 4000 + }, + { + "epoch": 16.1, + "grad_norm": 0.01574423350393772, + "learning_rate": 0.0001468162322665787, + "loss": 0.0198, + "step": 4025 + }, + { + "epoch": 16.2, + "grad_norm": 22.461671829223633, + "learning_rate": 0.00014516661167931377, + "loss": 0.0262, + "step": 4050 + }, + { + "epoch": 16.3, + "grad_norm": 17.334550857543945, + "learning_rate": 0.00014351699109204884, + "loss": 0.0567, + "step": 4075 + }, + { + "epoch": 16.4, + "grad_norm": 0.00328136351890862, + "learning_rate": 0.00014186737050478392, + "loss": 0.0182, + "step": 4100 + }, + { + "epoch": 16.5, + "grad_norm": 1.2645615339279175, + "learning_rate": 0.00014021774991751897, + "loss": 0.0116, + "step": 4125 + }, + { + "epoch": 16.6, + "grad_norm": 0.0031980539206415415, + "learning_rate": 0.00013856812933025404, + "loss": 0.0079, + "step": 4150 + }, + { + "epoch": 16.7, + "grad_norm": 0.01752198301255703, + "learning_rate": 0.00013691850874298912, + "loss": 0.0576, + "step": 4175 + }, + { + "epoch": 16.8, + "grad_norm": 48.27894592285156, + "learning_rate": 0.00013526888815572417, + "loss": 0.0264, + "step": 4200 + }, + { + "epoch": 16.9, + "grad_norm": 0.008802427910268307, + "learning_rate": 0.00013361926756845927, + "loss": 0.0051, + "step": 4225 + }, + { + "epoch": 17.0, + "grad_norm": 0.0350751169025898, + "learning_rate": 0.00013196964698119435, + "loss": 0.0403, + "step": 4250 + }, + { + "epoch": 17.1, + "grad_norm": 0.0020171860232949257, + "learning_rate": 0.0001303200263939294, + "loss": 0.0012, + "step": 4275 + }, + { + "epoch": 17.2, + "grad_norm": 0.014922365546226501, + "learning_rate": 0.00012867040580666448, + "loss": 0.036, + "step": 4300 + }, + { + "epoch": 17.3, + "grad_norm": 0.01983438804745674, + "learning_rate": 0.00012702078521939955, + "loss": 0.0094, + "step": 4325 + }, + { + "epoch": 17.4, + "grad_norm": 0.0031381379812955856, + "learning_rate": 0.0001253711646321346, + "loss": 0.011, + "step": 4350 + }, + { + "epoch": 17.5, + "grad_norm": 0.656819224357605, + "learning_rate": 0.00012372154404486968, + "loss": 0.0094, + "step": 4375 + }, + { + "epoch": 17.6, + "grad_norm": 0.008435660973191261, + "learning_rate": 0.00012207192345760475, + "loss": 0.0445, + "step": 4400 + }, + { + "epoch": 17.7, + "grad_norm": 23.819961547851562, + "learning_rate": 0.00012042230287033982, + "loss": 0.0394, + "step": 4425 + }, + { + "epoch": 17.8, + "grad_norm": 0.0055985236540436745, + "learning_rate": 0.0001187726822830749, + "loss": 0.0106, + "step": 4450 + }, + { + "epoch": 17.9, + "grad_norm": 11.063871383666992, + "learning_rate": 0.00011712306169580998, + "loss": 0.0128, + "step": 4475 + }, + { + "epoch": 18.0, + "grad_norm": 0.003767346730455756, + "learning_rate": 0.00011547344110854503, + "loss": 0.0096, + "step": 4500 + }, + { + "epoch": 18.1, + "grad_norm": 0.018198227509856224, + "learning_rate": 0.00011382382052128011, + "loss": 0.0395, + "step": 4525 + }, + { + "epoch": 18.2, + "grad_norm": 0.004393396433442831, + "learning_rate": 0.00011217419993401519, + "loss": 0.0378, + "step": 4550 + }, + { + "epoch": 18.3, + "grad_norm": 0.01423373818397522, + "learning_rate": 0.00011052457934675025, + "loss": 0.013, + "step": 4575 + }, + { + "epoch": 18.4, + "grad_norm": 25.492542266845703, + "learning_rate": 0.00010887495875948532, + "loss": 0.0094, + "step": 4600 + }, + { + "epoch": 18.5, + "grad_norm": 31.223493576049805, + "learning_rate": 0.0001072253381722204, + "loss": 0.0161, + "step": 4625 + }, + { + "epoch": 18.6, + "grad_norm": 0.003598716575652361, + "learning_rate": 0.00010557571758495546, + "loss": 0.0106, + "step": 4650 + }, + { + "epoch": 18.7, + "grad_norm": 50.249603271484375, + "learning_rate": 0.00010392609699769054, + "loss": 0.024, + "step": 4675 + }, + { + "epoch": 18.8, + "grad_norm": 63.50950622558594, + "learning_rate": 0.00010227647641042562, + "loss": 0.018, + "step": 4700 + }, + { + "epoch": 18.9, + "grad_norm": 0.002336911391466856, + "learning_rate": 0.00010062685582316067, + "loss": 0.0091, + "step": 4725 + }, + { + "epoch": 19.0, + "grad_norm": 0.005011886823922396, + "learning_rate": 9.897723523589574e-05, + "loss": 0.0061, + "step": 4750 + }, + { + "epoch": 19.1, + "grad_norm": 0.027857890352606773, + "learning_rate": 9.732761464863082e-05, + "loss": 0.0221, + "step": 4775 + }, + { + "epoch": 19.2, + "grad_norm": 0.0016227838350459933, + "learning_rate": 9.56779940613659e-05, + "loss": 0.0195, + "step": 4800 + }, + { + "epoch": 19.3, + "grad_norm": 0.005687279161065817, + "learning_rate": 9.402837347410096e-05, + "loss": 0.0203, + "step": 4825 + }, + { + "epoch": 19.4, + "grad_norm": 0.009174875915050507, + "learning_rate": 9.237875288683603e-05, + "loss": 0.0086, + "step": 4850 + }, + { + "epoch": 19.5, + "grad_norm": 0.0032144756987690926, + "learning_rate": 9.072913229957111e-05, + "loss": 0.0289, + "step": 4875 + }, + { + "epoch": 19.6, + "grad_norm": 0.003061591414734721, + "learning_rate": 8.907951171230617e-05, + "loss": 0.0015, + "step": 4900 + }, + { + "epoch": 19.7, + "grad_norm": 0.006960512604564428, + "learning_rate": 8.742989112504124e-05, + "loss": 0.0176, + "step": 4925 + }, + { + "epoch": 19.8, + "grad_norm": 0.0027057609986513853, + "learning_rate": 8.578027053777631e-05, + "loss": 0.0047, + "step": 4950 + }, + { + "epoch": 19.9, + "grad_norm": 0.22216364741325378, + "learning_rate": 8.413064995051139e-05, + "loss": 0.0099, + "step": 4975 + }, + { + "epoch": 20.0, + "grad_norm": 0.0022015271242707968, + "learning_rate": 8.248102936324645e-05, + "loss": 0.0219, + "step": 5000 + } + ], + "logging_steps": 25, + "max_steps": 6250, + "num_input_tokens_seen": 0, + "num_train_epochs": 25, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 254097408000000.0, + "train_batch_size": 20, + "trial_name": null, + "trial_params": null +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-5000/training_args.bin b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-5000/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..77692b6666e83a688b38e799c964e4694a57eec9 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-5000/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f73b41315068a430d2439f5707c62152a34210df62454b9549f9642446dd90b2 +size 5304 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-5000/vocab.txt b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-5000/vocab.txt new file mode 100644 index 0000000000000000000000000000000000000000..cc9ad53732e21ea9180f18e1fb67a38cc5702d4c --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-5000/vocab.txt @@ -0,0 +1,4101 @@ +[PAD] +[UNK] +[CLS] +[SEP] +[MASK] +AAAAAA +AAAAAT +AAAAAC +AAAAAG +AAAATA +AAAATT +AAAATC +AAAATG +AAAACA +AAAACT +AAAACC +AAAACG +AAAAGA +AAAAGT +AAAAGC +AAAAGG +AAATAA +AAATAT +AAATAC +AAATAG +AAATTA +AAATTT +AAATTC +AAATTG +AAATCA +AAATCT +AAATCC +AAATCG +AAATGA +AAATGT +AAATGC +AAATGG +AAACAA +AAACAT +AAACAC +AAACAG +AAACTA +AAACTT +AAACTC +AAACTG +AAACCA +AAACCT +AAACCC +AAACCG +AAACGA +AAACGT +AAACGC +AAACGG +AAAGAA +AAAGAT +AAAGAC +AAAGAG +AAAGTA +AAAGTT +AAAGTC +AAAGTG +AAAGCA +AAAGCT +AAAGCC +AAAGCG +AAAGGA +AAAGGT +AAAGGC +AAAGGG +AATAAA +AATAAT +AATAAC +AATAAG +AATATA +AATATT +AATATC +AATATG +AATACA +AATACT +AATACC +AATACG +AATAGA +AATAGT +AATAGC +AATAGG +AATTAA +AATTAT +AATTAC +AATTAG +AATTTA +AATTTT +AATTTC +AATTTG +AATTCA +AATTCT +AATTCC +AATTCG +AATTGA +AATTGT +AATTGC +AATTGG +AATCAA +AATCAT +AATCAC +AATCAG +AATCTA +AATCTT +AATCTC +AATCTG +AATCCA +AATCCT +AATCCC +AATCCG +AATCGA +AATCGT +AATCGC +AATCGG +AATGAA +AATGAT +AATGAC +AATGAG +AATGTA +AATGTT +AATGTC +AATGTG +AATGCA +AATGCT +AATGCC +AATGCG +AATGGA +AATGGT +AATGGC +AATGGG +AACAAA +AACAAT +AACAAC +AACAAG +AACATA +AACATT +AACATC +AACATG +AACACA +AACACT +AACACC +AACACG +AACAGA +AACAGT +AACAGC +AACAGG +AACTAA +AACTAT +AACTAC +AACTAG +AACTTA +AACTTT +AACTTC +AACTTG +AACTCA +AACTCT +AACTCC +AACTCG +AACTGA +AACTGT +AACTGC +AACTGG +AACCAA +AACCAT +AACCAC +AACCAG +AACCTA +AACCTT +AACCTC +AACCTG +AACCCA +AACCCT +AACCCC +AACCCG +AACCGA +AACCGT +AACCGC +AACCGG +AACGAA +AACGAT +AACGAC +AACGAG +AACGTA +AACGTT +AACGTC +AACGTG +AACGCA +AACGCT +AACGCC +AACGCG +AACGGA +AACGGT +AACGGC +AACGGG +AAGAAA +AAGAAT +AAGAAC +AAGAAG +AAGATA +AAGATT +AAGATC +AAGATG +AAGACA +AAGACT +AAGACC +AAGACG +AAGAGA +AAGAGT +AAGAGC +AAGAGG +AAGTAA +AAGTAT +AAGTAC +AAGTAG +AAGTTA +AAGTTT +AAGTTC +AAGTTG +AAGTCA +AAGTCT +AAGTCC +AAGTCG +AAGTGA +AAGTGT +AAGTGC +AAGTGG +AAGCAA +AAGCAT +AAGCAC +AAGCAG +AAGCTA +AAGCTT +AAGCTC +AAGCTG +AAGCCA +AAGCCT +AAGCCC +AAGCCG +AAGCGA +AAGCGT +AAGCGC +AAGCGG +AAGGAA +AAGGAT +AAGGAC +AAGGAG +AAGGTA +AAGGTT +AAGGTC +AAGGTG +AAGGCA +AAGGCT +AAGGCC +AAGGCG +AAGGGA +AAGGGT +AAGGGC +AAGGGG +ATAAAA +ATAAAT +ATAAAC +ATAAAG +ATAATA +ATAATT +ATAATC +ATAATG +ATAACA +ATAACT +ATAACC +ATAACG +ATAAGA +ATAAGT +ATAAGC +ATAAGG +ATATAA +ATATAT +ATATAC +ATATAG +ATATTA +ATATTT +ATATTC +ATATTG +ATATCA +ATATCT +ATATCC +ATATCG +ATATGA +ATATGT +ATATGC +ATATGG +ATACAA +ATACAT +ATACAC +ATACAG +ATACTA +ATACTT +ATACTC +ATACTG +ATACCA +ATACCT +ATACCC +ATACCG +ATACGA +ATACGT +ATACGC +ATACGG +ATAGAA +ATAGAT +ATAGAC +ATAGAG +ATAGTA +ATAGTT +ATAGTC +ATAGTG +ATAGCA +ATAGCT +ATAGCC +ATAGCG +ATAGGA +ATAGGT +ATAGGC +ATAGGG +ATTAAA +ATTAAT +ATTAAC +ATTAAG +ATTATA +ATTATT +ATTATC +ATTATG +ATTACA +ATTACT +ATTACC +ATTACG +ATTAGA +ATTAGT +ATTAGC +ATTAGG +ATTTAA +ATTTAT +ATTTAC +ATTTAG +ATTTTA +ATTTTT +ATTTTC +ATTTTG +ATTTCA +ATTTCT +ATTTCC +ATTTCG +ATTTGA +ATTTGT +ATTTGC +ATTTGG +ATTCAA +ATTCAT +ATTCAC +ATTCAG +ATTCTA +ATTCTT +ATTCTC +ATTCTG +ATTCCA +ATTCCT +ATTCCC +ATTCCG +ATTCGA +ATTCGT +ATTCGC +ATTCGG +ATTGAA +ATTGAT +ATTGAC +ATTGAG +ATTGTA +ATTGTT +ATTGTC +ATTGTG +ATTGCA +ATTGCT +ATTGCC +ATTGCG +ATTGGA +ATTGGT +ATTGGC +ATTGGG +ATCAAA +ATCAAT +ATCAAC +ATCAAG +ATCATA +ATCATT +ATCATC +ATCATG +ATCACA +ATCACT +ATCACC +ATCACG +ATCAGA +ATCAGT +ATCAGC +ATCAGG +ATCTAA +ATCTAT +ATCTAC +ATCTAG +ATCTTA +ATCTTT +ATCTTC +ATCTTG +ATCTCA +ATCTCT +ATCTCC +ATCTCG +ATCTGA +ATCTGT +ATCTGC +ATCTGG +ATCCAA +ATCCAT +ATCCAC +ATCCAG +ATCCTA +ATCCTT +ATCCTC +ATCCTG +ATCCCA +ATCCCT +ATCCCC +ATCCCG +ATCCGA +ATCCGT +ATCCGC +ATCCGG +ATCGAA +ATCGAT +ATCGAC +ATCGAG +ATCGTA +ATCGTT +ATCGTC +ATCGTG +ATCGCA +ATCGCT +ATCGCC +ATCGCG +ATCGGA +ATCGGT +ATCGGC +ATCGGG +ATGAAA +ATGAAT +ATGAAC +ATGAAG +ATGATA +ATGATT +ATGATC +ATGATG +ATGACA +ATGACT +ATGACC +ATGACG +ATGAGA +ATGAGT +ATGAGC +ATGAGG +ATGTAA +ATGTAT +ATGTAC +ATGTAG +ATGTTA +ATGTTT +ATGTTC +ATGTTG +ATGTCA +ATGTCT +ATGTCC +ATGTCG +ATGTGA +ATGTGT +ATGTGC +ATGTGG +ATGCAA +ATGCAT +ATGCAC +ATGCAG +ATGCTA +ATGCTT +ATGCTC +ATGCTG +ATGCCA +ATGCCT +ATGCCC +ATGCCG +ATGCGA +ATGCGT +ATGCGC +ATGCGG +ATGGAA +ATGGAT +ATGGAC +ATGGAG +ATGGTA +ATGGTT +ATGGTC +ATGGTG +ATGGCA +ATGGCT +ATGGCC +ATGGCG +ATGGGA +ATGGGT +ATGGGC +ATGGGG +ACAAAA +ACAAAT +ACAAAC +ACAAAG +ACAATA +ACAATT +ACAATC +ACAATG +ACAACA +ACAACT +ACAACC +ACAACG +ACAAGA +ACAAGT +ACAAGC +ACAAGG +ACATAA +ACATAT +ACATAC +ACATAG +ACATTA +ACATTT +ACATTC +ACATTG +ACATCA +ACATCT +ACATCC +ACATCG +ACATGA +ACATGT +ACATGC +ACATGG +ACACAA +ACACAT +ACACAC +ACACAG +ACACTA +ACACTT +ACACTC +ACACTG +ACACCA +ACACCT +ACACCC +ACACCG +ACACGA +ACACGT +ACACGC +ACACGG +ACAGAA +ACAGAT +ACAGAC +ACAGAG +ACAGTA +ACAGTT +ACAGTC +ACAGTG +ACAGCA +ACAGCT +ACAGCC +ACAGCG +ACAGGA +ACAGGT +ACAGGC +ACAGGG +ACTAAA +ACTAAT +ACTAAC +ACTAAG +ACTATA +ACTATT +ACTATC +ACTATG +ACTACA +ACTACT +ACTACC +ACTACG +ACTAGA +ACTAGT +ACTAGC +ACTAGG +ACTTAA +ACTTAT +ACTTAC +ACTTAG +ACTTTA +ACTTTT +ACTTTC +ACTTTG +ACTTCA +ACTTCT +ACTTCC +ACTTCG +ACTTGA +ACTTGT +ACTTGC +ACTTGG +ACTCAA +ACTCAT +ACTCAC +ACTCAG +ACTCTA +ACTCTT +ACTCTC +ACTCTG +ACTCCA +ACTCCT +ACTCCC +ACTCCG +ACTCGA +ACTCGT +ACTCGC +ACTCGG +ACTGAA +ACTGAT +ACTGAC +ACTGAG +ACTGTA +ACTGTT +ACTGTC +ACTGTG +ACTGCA +ACTGCT +ACTGCC +ACTGCG +ACTGGA +ACTGGT +ACTGGC +ACTGGG +ACCAAA +ACCAAT +ACCAAC +ACCAAG +ACCATA +ACCATT +ACCATC +ACCATG +ACCACA +ACCACT +ACCACC +ACCACG +ACCAGA +ACCAGT +ACCAGC +ACCAGG +ACCTAA +ACCTAT +ACCTAC +ACCTAG +ACCTTA +ACCTTT +ACCTTC +ACCTTG +ACCTCA +ACCTCT +ACCTCC +ACCTCG +ACCTGA +ACCTGT +ACCTGC +ACCTGG +ACCCAA +ACCCAT +ACCCAC +ACCCAG +ACCCTA +ACCCTT +ACCCTC +ACCCTG +ACCCCA +ACCCCT +ACCCCC +ACCCCG +ACCCGA +ACCCGT +ACCCGC +ACCCGG +ACCGAA +ACCGAT +ACCGAC +ACCGAG +ACCGTA +ACCGTT +ACCGTC +ACCGTG +ACCGCA +ACCGCT +ACCGCC +ACCGCG +ACCGGA +ACCGGT +ACCGGC +ACCGGG +ACGAAA +ACGAAT +ACGAAC +ACGAAG +ACGATA +ACGATT +ACGATC +ACGATG +ACGACA +ACGACT +ACGACC +ACGACG +ACGAGA +ACGAGT +ACGAGC +ACGAGG +ACGTAA +ACGTAT +ACGTAC +ACGTAG +ACGTTA +ACGTTT +ACGTTC +ACGTTG +ACGTCA +ACGTCT +ACGTCC +ACGTCG +ACGTGA +ACGTGT +ACGTGC +ACGTGG +ACGCAA +ACGCAT +ACGCAC +ACGCAG +ACGCTA +ACGCTT +ACGCTC +ACGCTG +ACGCCA +ACGCCT +ACGCCC +ACGCCG +ACGCGA +ACGCGT +ACGCGC +ACGCGG +ACGGAA +ACGGAT +ACGGAC +ACGGAG +ACGGTA +ACGGTT +ACGGTC +ACGGTG +ACGGCA +ACGGCT +ACGGCC +ACGGCG +ACGGGA +ACGGGT +ACGGGC +ACGGGG +AGAAAA +AGAAAT +AGAAAC +AGAAAG +AGAATA +AGAATT +AGAATC +AGAATG +AGAACA +AGAACT +AGAACC +AGAACG +AGAAGA +AGAAGT +AGAAGC +AGAAGG +AGATAA +AGATAT +AGATAC +AGATAG +AGATTA +AGATTT +AGATTC +AGATTG +AGATCA +AGATCT +AGATCC +AGATCG +AGATGA +AGATGT +AGATGC +AGATGG +AGACAA +AGACAT +AGACAC +AGACAG +AGACTA +AGACTT +AGACTC +AGACTG +AGACCA +AGACCT +AGACCC +AGACCG +AGACGA +AGACGT +AGACGC +AGACGG +AGAGAA +AGAGAT +AGAGAC +AGAGAG +AGAGTA +AGAGTT +AGAGTC +AGAGTG +AGAGCA +AGAGCT +AGAGCC +AGAGCG +AGAGGA +AGAGGT +AGAGGC +AGAGGG +AGTAAA +AGTAAT +AGTAAC +AGTAAG +AGTATA +AGTATT +AGTATC +AGTATG +AGTACA +AGTACT +AGTACC +AGTACG +AGTAGA +AGTAGT +AGTAGC +AGTAGG +AGTTAA +AGTTAT +AGTTAC +AGTTAG +AGTTTA +AGTTTT +AGTTTC +AGTTTG +AGTTCA +AGTTCT +AGTTCC +AGTTCG +AGTTGA +AGTTGT +AGTTGC +AGTTGG +AGTCAA +AGTCAT +AGTCAC +AGTCAG +AGTCTA +AGTCTT +AGTCTC +AGTCTG +AGTCCA +AGTCCT +AGTCCC +AGTCCG +AGTCGA +AGTCGT +AGTCGC +AGTCGG +AGTGAA +AGTGAT +AGTGAC +AGTGAG +AGTGTA +AGTGTT +AGTGTC +AGTGTG +AGTGCA +AGTGCT +AGTGCC +AGTGCG +AGTGGA +AGTGGT +AGTGGC +AGTGGG +AGCAAA +AGCAAT +AGCAAC +AGCAAG +AGCATA +AGCATT +AGCATC +AGCATG +AGCACA +AGCACT +AGCACC +AGCACG +AGCAGA +AGCAGT +AGCAGC +AGCAGG +AGCTAA +AGCTAT +AGCTAC +AGCTAG +AGCTTA +AGCTTT +AGCTTC +AGCTTG +AGCTCA +AGCTCT +AGCTCC +AGCTCG +AGCTGA +AGCTGT +AGCTGC +AGCTGG +AGCCAA +AGCCAT +AGCCAC +AGCCAG +AGCCTA +AGCCTT +AGCCTC +AGCCTG +AGCCCA +AGCCCT +AGCCCC +AGCCCG +AGCCGA +AGCCGT +AGCCGC +AGCCGG +AGCGAA +AGCGAT +AGCGAC +AGCGAG +AGCGTA +AGCGTT +AGCGTC +AGCGTG +AGCGCA +AGCGCT +AGCGCC +AGCGCG +AGCGGA +AGCGGT +AGCGGC +AGCGGG +AGGAAA +AGGAAT +AGGAAC +AGGAAG +AGGATA +AGGATT +AGGATC +AGGATG +AGGACA +AGGACT +AGGACC +AGGACG +AGGAGA +AGGAGT +AGGAGC +AGGAGG +AGGTAA +AGGTAT +AGGTAC +AGGTAG +AGGTTA +AGGTTT +AGGTTC +AGGTTG +AGGTCA +AGGTCT +AGGTCC +AGGTCG +AGGTGA +AGGTGT +AGGTGC +AGGTGG +AGGCAA +AGGCAT +AGGCAC +AGGCAG +AGGCTA +AGGCTT +AGGCTC +AGGCTG +AGGCCA +AGGCCT +AGGCCC +AGGCCG +AGGCGA +AGGCGT +AGGCGC +AGGCGG +AGGGAA +AGGGAT +AGGGAC +AGGGAG +AGGGTA +AGGGTT +AGGGTC +AGGGTG +AGGGCA +AGGGCT +AGGGCC +AGGGCG +AGGGGA +AGGGGT +AGGGGC +AGGGGG +TAAAAA +TAAAAT +TAAAAC +TAAAAG +TAAATA +TAAATT +TAAATC +TAAATG +TAAACA +TAAACT +TAAACC +TAAACG +TAAAGA +TAAAGT +TAAAGC +TAAAGG +TAATAA +TAATAT +TAATAC +TAATAG +TAATTA +TAATTT +TAATTC +TAATTG +TAATCA +TAATCT +TAATCC +TAATCG +TAATGA +TAATGT +TAATGC +TAATGG +TAACAA +TAACAT +TAACAC +TAACAG +TAACTA +TAACTT +TAACTC +TAACTG +TAACCA +TAACCT +TAACCC +TAACCG +TAACGA +TAACGT +TAACGC +TAACGG +TAAGAA +TAAGAT +TAAGAC +TAAGAG +TAAGTA +TAAGTT +TAAGTC +TAAGTG +TAAGCA +TAAGCT +TAAGCC +TAAGCG +TAAGGA +TAAGGT +TAAGGC +TAAGGG +TATAAA +TATAAT +TATAAC +TATAAG +TATATA +TATATT +TATATC +TATATG +TATACA +TATACT +TATACC +TATACG +TATAGA +TATAGT +TATAGC +TATAGG +TATTAA +TATTAT +TATTAC +TATTAG +TATTTA +TATTTT +TATTTC +TATTTG +TATTCA +TATTCT +TATTCC +TATTCG +TATTGA +TATTGT +TATTGC +TATTGG +TATCAA +TATCAT +TATCAC +TATCAG +TATCTA +TATCTT +TATCTC +TATCTG +TATCCA +TATCCT +TATCCC +TATCCG +TATCGA +TATCGT +TATCGC +TATCGG +TATGAA +TATGAT +TATGAC +TATGAG +TATGTA +TATGTT +TATGTC +TATGTG +TATGCA +TATGCT +TATGCC +TATGCG +TATGGA +TATGGT +TATGGC +TATGGG +TACAAA +TACAAT +TACAAC +TACAAG +TACATA +TACATT +TACATC +TACATG +TACACA +TACACT +TACACC +TACACG +TACAGA +TACAGT +TACAGC +TACAGG +TACTAA +TACTAT +TACTAC +TACTAG +TACTTA +TACTTT +TACTTC +TACTTG +TACTCA +TACTCT +TACTCC +TACTCG +TACTGA +TACTGT +TACTGC +TACTGG +TACCAA +TACCAT +TACCAC +TACCAG +TACCTA +TACCTT +TACCTC +TACCTG +TACCCA +TACCCT +TACCCC +TACCCG +TACCGA +TACCGT +TACCGC +TACCGG +TACGAA +TACGAT +TACGAC +TACGAG +TACGTA +TACGTT +TACGTC +TACGTG +TACGCA +TACGCT +TACGCC +TACGCG +TACGGA +TACGGT +TACGGC +TACGGG +TAGAAA +TAGAAT +TAGAAC +TAGAAG +TAGATA +TAGATT +TAGATC +TAGATG +TAGACA +TAGACT +TAGACC +TAGACG +TAGAGA +TAGAGT +TAGAGC +TAGAGG +TAGTAA +TAGTAT +TAGTAC +TAGTAG +TAGTTA +TAGTTT +TAGTTC +TAGTTG +TAGTCA +TAGTCT +TAGTCC +TAGTCG +TAGTGA +TAGTGT +TAGTGC +TAGTGG +TAGCAA +TAGCAT +TAGCAC +TAGCAG +TAGCTA +TAGCTT +TAGCTC +TAGCTG +TAGCCA +TAGCCT +TAGCCC +TAGCCG +TAGCGA +TAGCGT +TAGCGC +TAGCGG +TAGGAA +TAGGAT +TAGGAC +TAGGAG +TAGGTA +TAGGTT +TAGGTC +TAGGTG +TAGGCA +TAGGCT +TAGGCC +TAGGCG +TAGGGA +TAGGGT +TAGGGC +TAGGGG +TTAAAA +TTAAAT +TTAAAC +TTAAAG +TTAATA +TTAATT +TTAATC +TTAATG +TTAACA +TTAACT +TTAACC +TTAACG +TTAAGA +TTAAGT +TTAAGC +TTAAGG +TTATAA +TTATAT +TTATAC +TTATAG +TTATTA +TTATTT +TTATTC +TTATTG +TTATCA +TTATCT +TTATCC +TTATCG +TTATGA +TTATGT +TTATGC +TTATGG +TTACAA +TTACAT +TTACAC +TTACAG +TTACTA +TTACTT +TTACTC +TTACTG +TTACCA +TTACCT +TTACCC +TTACCG +TTACGA +TTACGT +TTACGC +TTACGG +TTAGAA +TTAGAT +TTAGAC +TTAGAG +TTAGTA +TTAGTT +TTAGTC +TTAGTG +TTAGCA +TTAGCT +TTAGCC +TTAGCG +TTAGGA +TTAGGT +TTAGGC +TTAGGG +TTTAAA +TTTAAT +TTTAAC +TTTAAG +TTTATA +TTTATT +TTTATC +TTTATG +TTTACA +TTTACT +TTTACC +TTTACG +TTTAGA +TTTAGT +TTTAGC +TTTAGG +TTTTAA +TTTTAT +TTTTAC +TTTTAG +TTTTTA +TTTTTT +TTTTTC +TTTTTG +TTTTCA +TTTTCT +TTTTCC +TTTTCG +TTTTGA +TTTTGT +TTTTGC +TTTTGG +TTTCAA +TTTCAT +TTTCAC +TTTCAG +TTTCTA +TTTCTT +TTTCTC +TTTCTG +TTTCCA +TTTCCT +TTTCCC +TTTCCG +TTTCGA +TTTCGT +TTTCGC +TTTCGG +TTTGAA +TTTGAT +TTTGAC +TTTGAG +TTTGTA +TTTGTT +TTTGTC +TTTGTG +TTTGCA +TTTGCT +TTTGCC +TTTGCG +TTTGGA +TTTGGT +TTTGGC +TTTGGG +TTCAAA +TTCAAT +TTCAAC +TTCAAG +TTCATA +TTCATT +TTCATC +TTCATG +TTCACA +TTCACT +TTCACC +TTCACG +TTCAGA +TTCAGT +TTCAGC +TTCAGG +TTCTAA +TTCTAT +TTCTAC +TTCTAG +TTCTTA +TTCTTT +TTCTTC +TTCTTG +TTCTCA +TTCTCT +TTCTCC +TTCTCG +TTCTGA +TTCTGT +TTCTGC +TTCTGG +TTCCAA +TTCCAT +TTCCAC +TTCCAG +TTCCTA +TTCCTT +TTCCTC +TTCCTG +TTCCCA +TTCCCT +TTCCCC +TTCCCG +TTCCGA +TTCCGT +TTCCGC +TTCCGG +TTCGAA +TTCGAT +TTCGAC +TTCGAG +TTCGTA +TTCGTT +TTCGTC +TTCGTG +TTCGCA +TTCGCT +TTCGCC +TTCGCG +TTCGGA +TTCGGT +TTCGGC +TTCGGG +TTGAAA +TTGAAT +TTGAAC +TTGAAG +TTGATA +TTGATT +TTGATC +TTGATG +TTGACA +TTGACT +TTGACC +TTGACG +TTGAGA +TTGAGT +TTGAGC +TTGAGG +TTGTAA +TTGTAT +TTGTAC +TTGTAG +TTGTTA +TTGTTT +TTGTTC +TTGTTG +TTGTCA +TTGTCT +TTGTCC +TTGTCG +TTGTGA +TTGTGT +TTGTGC +TTGTGG +TTGCAA +TTGCAT +TTGCAC +TTGCAG +TTGCTA +TTGCTT +TTGCTC +TTGCTG +TTGCCA +TTGCCT +TTGCCC +TTGCCG +TTGCGA +TTGCGT +TTGCGC +TTGCGG +TTGGAA +TTGGAT +TTGGAC +TTGGAG +TTGGTA +TTGGTT +TTGGTC +TTGGTG +TTGGCA +TTGGCT +TTGGCC +TTGGCG +TTGGGA +TTGGGT +TTGGGC +TTGGGG +TCAAAA +TCAAAT +TCAAAC +TCAAAG +TCAATA +TCAATT +TCAATC +TCAATG +TCAACA +TCAACT +TCAACC +TCAACG +TCAAGA +TCAAGT +TCAAGC +TCAAGG +TCATAA +TCATAT +TCATAC +TCATAG +TCATTA +TCATTT +TCATTC +TCATTG +TCATCA +TCATCT +TCATCC +TCATCG +TCATGA +TCATGT +TCATGC +TCATGG +TCACAA +TCACAT +TCACAC +TCACAG +TCACTA +TCACTT +TCACTC +TCACTG +TCACCA +TCACCT +TCACCC +TCACCG +TCACGA +TCACGT +TCACGC +TCACGG +TCAGAA +TCAGAT +TCAGAC +TCAGAG +TCAGTA +TCAGTT +TCAGTC +TCAGTG +TCAGCA +TCAGCT +TCAGCC +TCAGCG +TCAGGA +TCAGGT +TCAGGC +TCAGGG +TCTAAA +TCTAAT +TCTAAC +TCTAAG +TCTATA +TCTATT +TCTATC +TCTATG +TCTACA +TCTACT +TCTACC +TCTACG +TCTAGA +TCTAGT +TCTAGC +TCTAGG +TCTTAA +TCTTAT +TCTTAC +TCTTAG +TCTTTA +TCTTTT +TCTTTC +TCTTTG +TCTTCA +TCTTCT +TCTTCC +TCTTCG +TCTTGA +TCTTGT +TCTTGC +TCTTGG +TCTCAA +TCTCAT +TCTCAC +TCTCAG +TCTCTA +TCTCTT +TCTCTC +TCTCTG +TCTCCA +TCTCCT +TCTCCC +TCTCCG +TCTCGA +TCTCGT +TCTCGC +TCTCGG +TCTGAA +TCTGAT +TCTGAC +TCTGAG +TCTGTA +TCTGTT +TCTGTC +TCTGTG +TCTGCA +TCTGCT +TCTGCC +TCTGCG +TCTGGA +TCTGGT +TCTGGC +TCTGGG +TCCAAA +TCCAAT +TCCAAC +TCCAAG +TCCATA +TCCATT +TCCATC +TCCATG +TCCACA +TCCACT +TCCACC +TCCACG +TCCAGA +TCCAGT +TCCAGC +TCCAGG +TCCTAA +TCCTAT +TCCTAC +TCCTAG +TCCTTA +TCCTTT +TCCTTC +TCCTTG +TCCTCA +TCCTCT +TCCTCC +TCCTCG +TCCTGA +TCCTGT +TCCTGC +TCCTGG +TCCCAA +TCCCAT +TCCCAC +TCCCAG +TCCCTA +TCCCTT +TCCCTC +TCCCTG +TCCCCA +TCCCCT +TCCCCC +TCCCCG +TCCCGA +TCCCGT +TCCCGC +TCCCGG +TCCGAA +TCCGAT +TCCGAC +TCCGAG +TCCGTA +TCCGTT +TCCGTC +TCCGTG +TCCGCA +TCCGCT +TCCGCC +TCCGCG +TCCGGA +TCCGGT +TCCGGC +TCCGGG +TCGAAA +TCGAAT +TCGAAC +TCGAAG +TCGATA +TCGATT +TCGATC +TCGATG +TCGACA +TCGACT +TCGACC +TCGACG +TCGAGA +TCGAGT +TCGAGC +TCGAGG +TCGTAA +TCGTAT +TCGTAC +TCGTAG +TCGTTA +TCGTTT +TCGTTC +TCGTTG +TCGTCA +TCGTCT +TCGTCC +TCGTCG +TCGTGA +TCGTGT +TCGTGC +TCGTGG +TCGCAA +TCGCAT +TCGCAC +TCGCAG +TCGCTA +TCGCTT +TCGCTC +TCGCTG +TCGCCA +TCGCCT +TCGCCC +TCGCCG +TCGCGA +TCGCGT +TCGCGC +TCGCGG +TCGGAA +TCGGAT +TCGGAC +TCGGAG +TCGGTA +TCGGTT +TCGGTC +TCGGTG +TCGGCA +TCGGCT +TCGGCC +TCGGCG +TCGGGA +TCGGGT +TCGGGC +TCGGGG +TGAAAA +TGAAAT +TGAAAC +TGAAAG +TGAATA +TGAATT +TGAATC +TGAATG +TGAACA +TGAACT +TGAACC +TGAACG +TGAAGA +TGAAGT +TGAAGC +TGAAGG +TGATAA +TGATAT +TGATAC +TGATAG +TGATTA +TGATTT +TGATTC +TGATTG +TGATCA +TGATCT +TGATCC +TGATCG +TGATGA +TGATGT +TGATGC +TGATGG +TGACAA +TGACAT +TGACAC +TGACAG +TGACTA +TGACTT +TGACTC +TGACTG +TGACCA +TGACCT +TGACCC +TGACCG +TGACGA +TGACGT +TGACGC +TGACGG +TGAGAA +TGAGAT +TGAGAC +TGAGAG +TGAGTA +TGAGTT +TGAGTC +TGAGTG +TGAGCA +TGAGCT +TGAGCC +TGAGCG +TGAGGA +TGAGGT +TGAGGC +TGAGGG +TGTAAA +TGTAAT +TGTAAC +TGTAAG +TGTATA +TGTATT +TGTATC +TGTATG +TGTACA +TGTACT +TGTACC +TGTACG +TGTAGA +TGTAGT +TGTAGC +TGTAGG +TGTTAA +TGTTAT +TGTTAC +TGTTAG +TGTTTA +TGTTTT +TGTTTC +TGTTTG +TGTTCA +TGTTCT +TGTTCC +TGTTCG +TGTTGA +TGTTGT +TGTTGC +TGTTGG +TGTCAA +TGTCAT +TGTCAC +TGTCAG +TGTCTA +TGTCTT +TGTCTC +TGTCTG +TGTCCA +TGTCCT +TGTCCC +TGTCCG +TGTCGA +TGTCGT +TGTCGC +TGTCGG +TGTGAA +TGTGAT +TGTGAC +TGTGAG +TGTGTA +TGTGTT +TGTGTC +TGTGTG +TGTGCA +TGTGCT +TGTGCC +TGTGCG +TGTGGA +TGTGGT +TGTGGC +TGTGGG +TGCAAA +TGCAAT +TGCAAC +TGCAAG +TGCATA +TGCATT +TGCATC +TGCATG +TGCACA +TGCACT +TGCACC +TGCACG +TGCAGA +TGCAGT +TGCAGC +TGCAGG +TGCTAA +TGCTAT +TGCTAC +TGCTAG +TGCTTA +TGCTTT +TGCTTC +TGCTTG +TGCTCA +TGCTCT +TGCTCC +TGCTCG +TGCTGA +TGCTGT +TGCTGC +TGCTGG +TGCCAA +TGCCAT +TGCCAC +TGCCAG +TGCCTA +TGCCTT +TGCCTC +TGCCTG +TGCCCA +TGCCCT +TGCCCC +TGCCCG +TGCCGA +TGCCGT +TGCCGC +TGCCGG +TGCGAA +TGCGAT +TGCGAC +TGCGAG +TGCGTA +TGCGTT +TGCGTC +TGCGTG +TGCGCA +TGCGCT +TGCGCC +TGCGCG +TGCGGA +TGCGGT +TGCGGC +TGCGGG +TGGAAA +TGGAAT +TGGAAC +TGGAAG +TGGATA +TGGATT +TGGATC +TGGATG +TGGACA +TGGACT +TGGACC +TGGACG +TGGAGA +TGGAGT +TGGAGC +TGGAGG +TGGTAA +TGGTAT +TGGTAC +TGGTAG +TGGTTA +TGGTTT +TGGTTC +TGGTTG +TGGTCA +TGGTCT +TGGTCC +TGGTCG +TGGTGA +TGGTGT +TGGTGC +TGGTGG +TGGCAA +TGGCAT +TGGCAC +TGGCAG +TGGCTA +TGGCTT +TGGCTC +TGGCTG +TGGCCA +TGGCCT +TGGCCC +TGGCCG +TGGCGA +TGGCGT +TGGCGC +TGGCGG +TGGGAA +TGGGAT +TGGGAC +TGGGAG +TGGGTA +TGGGTT +TGGGTC +TGGGTG +TGGGCA +TGGGCT +TGGGCC +TGGGCG +TGGGGA +TGGGGT +TGGGGC +TGGGGG +CAAAAA +CAAAAT +CAAAAC +CAAAAG +CAAATA +CAAATT +CAAATC +CAAATG +CAAACA +CAAACT +CAAACC +CAAACG +CAAAGA +CAAAGT +CAAAGC +CAAAGG +CAATAA +CAATAT +CAATAC +CAATAG +CAATTA +CAATTT +CAATTC +CAATTG +CAATCA +CAATCT +CAATCC +CAATCG +CAATGA +CAATGT +CAATGC +CAATGG +CAACAA +CAACAT +CAACAC +CAACAG +CAACTA +CAACTT +CAACTC +CAACTG +CAACCA +CAACCT +CAACCC +CAACCG +CAACGA +CAACGT +CAACGC +CAACGG +CAAGAA +CAAGAT +CAAGAC +CAAGAG +CAAGTA +CAAGTT +CAAGTC +CAAGTG +CAAGCA +CAAGCT +CAAGCC +CAAGCG +CAAGGA +CAAGGT +CAAGGC +CAAGGG +CATAAA +CATAAT +CATAAC +CATAAG +CATATA +CATATT +CATATC +CATATG +CATACA +CATACT +CATACC +CATACG +CATAGA +CATAGT +CATAGC +CATAGG +CATTAA +CATTAT +CATTAC +CATTAG +CATTTA +CATTTT +CATTTC +CATTTG +CATTCA +CATTCT +CATTCC +CATTCG +CATTGA +CATTGT +CATTGC +CATTGG +CATCAA +CATCAT +CATCAC +CATCAG +CATCTA +CATCTT +CATCTC +CATCTG +CATCCA +CATCCT +CATCCC +CATCCG +CATCGA +CATCGT +CATCGC +CATCGG +CATGAA +CATGAT +CATGAC +CATGAG +CATGTA +CATGTT +CATGTC +CATGTG +CATGCA +CATGCT +CATGCC +CATGCG +CATGGA +CATGGT +CATGGC +CATGGG +CACAAA +CACAAT +CACAAC +CACAAG +CACATA +CACATT +CACATC +CACATG +CACACA +CACACT +CACACC +CACACG +CACAGA +CACAGT +CACAGC +CACAGG +CACTAA +CACTAT +CACTAC +CACTAG +CACTTA +CACTTT +CACTTC +CACTTG +CACTCA +CACTCT +CACTCC +CACTCG +CACTGA +CACTGT +CACTGC +CACTGG +CACCAA +CACCAT +CACCAC +CACCAG +CACCTA +CACCTT +CACCTC +CACCTG +CACCCA +CACCCT +CACCCC +CACCCG +CACCGA +CACCGT +CACCGC +CACCGG +CACGAA +CACGAT +CACGAC +CACGAG +CACGTA +CACGTT +CACGTC +CACGTG +CACGCA +CACGCT +CACGCC +CACGCG +CACGGA +CACGGT +CACGGC +CACGGG +CAGAAA +CAGAAT +CAGAAC +CAGAAG +CAGATA +CAGATT +CAGATC +CAGATG +CAGACA +CAGACT +CAGACC +CAGACG +CAGAGA +CAGAGT +CAGAGC +CAGAGG +CAGTAA +CAGTAT +CAGTAC +CAGTAG +CAGTTA +CAGTTT +CAGTTC +CAGTTG +CAGTCA +CAGTCT +CAGTCC +CAGTCG +CAGTGA +CAGTGT +CAGTGC +CAGTGG +CAGCAA +CAGCAT +CAGCAC +CAGCAG +CAGCTA +CAGCTT +CAGCTC +CAGCTG +CAGCCA +CAGCCT +CAGCCC +CAGCCG +CAGCGA +CAGCGT +CAGCGC +CAGCGG +CAGGAA +CAGGAT +CAGGAC +CAGGAG +CAGGTA +CAGGTT +CAGGTC +CAGGTG +CAGGCA +CAGGCT +CAGGCC +CAGGCG +CAGGGA +CAGGGT +CAGGGC +CAGGGG +CTAAAA +CTAAAT +CTAAAC +CTAAAG +CTAATA +CTAATT +CTAATC +CTAATG +CTAACA +CTAACT +CTAACC +CTAACG +CTAAGA +CTAAGT +CTAAGC +CTAAGG +CTATAA +CTATAT +CTATAC +CTATAG +CTATTA +CTATTT +CTATTC +CTATTG +CTATCA +CTATCT +CTATCC +CTATCG +CTATGA +CTATGT +CTATGC +CTATGG +CTACAA +CTACAT +CTACAC +CTACAG +CTACTA +CTACTT +CTACTC +CTACTG +CTACCA +CTACCT +CTACCC +CTACCG +CTACGA +CTACGT +CTACGC +CTACGG +CTAGAA +CTAGAT +CTAGAC +CTAGAG +CTAGTA +CTAGTT +CTAGTC +CTAGTG +CTAGCA +CTAGCT +CTAGCC +CTAGCG +CTAGGA +CTAGGT +CTAGGC +CTAGGG +CTTAAA +CTTAAT +CTTAAC +CTTAAG +CTTATA +CTTATT +CTTATC +CTTATG +CTTACA +CTTACT +CTTACC +CTTACG +CTTAGA +CTTAGT +CTTAGC +CTTAGG +CTTTAA +CTTTAT +CTTTAC +CTTTAG +CTTTTA +CTTTTT +CTTTTC +CTTTTG +CTTTCA +CTTTCT +CTTTCC +CTTTCG +CTTTGA +CTTTGT +CTTTGC +CTTTGG +CTTCAA +CTTCAT +CTTCAC +CTTCAG +CTTCTA +CTTCTT +CTTCTC +CTTCTG +CTTCCA +CTTCCT +CTTCCC +CTTCCG +CTTCGA +CTTCGT +CTTCGC +CTTCGG +CTTGAA +CTTGAT +CTTGAC +CTTGAG +CTTGTA +CTTGTT +CTTGTC +CTTGTG +CTTGCA +CTTGCT +CTTGCC +CTTGCG +CTTGGA +CTTGGT +CTTGGC +CTTGGG +CTCAAA +CTCAAT +CTCAAC +CTCAAG +CTCATA +CTCATT +CTCATC +CTCATG +CTCACA +CTCACT +CTCACC +CTCACG +CTCAGA +CTCAGT +CTCAGC +CTCAGG +CTCTAA +CTCTAT +CTCTAC +CTCTAG +CTCTTA +CTCTTT +CTCTTC +CTCTTG +CTCTCA +CTCTCT +CTCTCC +CTCTCG +CTCTGA +CTCTGT +CTCTGC +CTCTGG +CTCCAA +CTCCAT +CTCCAC +CTCCAG +CTCCTA +CTCCTT +CTCCTC +CTCCTG +CTCCCA +CTCCCT +CTCCCC +CTCCCG +CTCCGA +CTCCGT +CTCCGC +CTCCGG +CTCGAA +CTCGAT +CTCGAC +CTCGAG +CTCGTA +CTCGTT +CTCGTC +CTCGTG +CTCGCA +CTCGCT +CTCGCC +CTCGCG +CTCGGA +CTCGGT +CTCGGC +CTCGGG +CTGAAA +CTGAAT +CTGAAC +CTGAAG +CTGATA +CTGATT +CTGATC +CTGATG +CTGACA +CTGACT +CTGACC +CTGACG +CTGAGA +CTGAGT +CTGAGC +CTGAGG +CTGTAA +CTGTAT +CTGTAC +CTGTAG +CTGTTA +CTGTTT +CTGTTC +CTGTTG +CTGTCA +CTGTCT +CTGTCC +CTGTCG +CTGTGA +CTGTGT +CTGTGC +CTGTGG +CTGCAA +CTGCAT +CTGCAC +CTGCAG +CTGCTA +CTGCTT +CTGCTC +CTGCTG +CTGCCA +CTGCCT +CTGCCC +CTGCCG +CTGCGA +CTGCGT +CTGCGC +CTGCGG +CTGGAA +CTGGAT +CTGGAC +CTGGAG +CTGGTA +CTGGTT +CTGGTC +CTGGTG +CTGGCA +CTGGCT +CTGGCC +CTGGCG +CTGGGA +CTGGGT +CTGGGC +CTGGGG +CCAAAA +CCAAAT +CCAAAC +CCAAAG +CCAATA +CCAATT +CCAATC +CCAATG +CCAACA +CCAACT +CCAACC +CCAACG +CCAAGA +CCAAGT +CCAAGC +CCAAGG +CCATAA +CCATAT +CCATAC +CCATAG +CCATTA +CCATTT +CCATTC +CCATTG +CCATCA +CCATCT +CCATCC +CCATCG +CCATGA +CCATGT +CCATGC +CCATGG +CCACAA +CCACAT +CCACAC +CCACAG +CCACTA +CCACTT +CCACTC +CCACTG +CCACCA +CCACCT +CCACCC +CCACCG +CCACGA +CCACGT +CCACGC +CCACGG +CCAGAA +CCAGAT +CCAGAC +CCAGAG +CCAGTA +CCAGTT +CCAGTC +CCAGTG +CCAGCA +CCAGCT +CCAGCC +CCAGCG +CCAGGA +CCAGGT +CCAGGC +CCAGGG +CCTAAA +CCTAAT +CCTAAC +CCTAAG +CCTATA +CCTATT +CCTATC +CCTATG +CCTACA +CCTACT +CCTACC +CCTACG +CCTAGA +CCTAGT +CCTAGC +CCTAGG +CCTTAA +CCTTAT +CCTTAC +CCTTAG +CCTTTA +CCTTTT +CCTTTC +CCTTTG +CCTTCA +CCTTCT +CCTTCC +CCTTCG +CCTTGA +CCTTGT +CCTTGC +CCTTGG +CCTCAA +CCTCAT +CCTCAC +CCTCAG +CCTCTA +CCTCTT +CCTCTC +CCTCTG +CCTCCA +CCTCCT +CCTCCC +CCTCCG +CCTCGA +CCTCGT +CCTCGC +CCTCGG +CCTGAA +CCTGAT +CCTGAC +CCTGAG +CCTGTA +CCTGTT +CCTGTC +CCTGTG +CCTGCA +CCTGCT +CCTGCC +CCTGCG +CCTGGA +CCTGGT +CCTGGC +CCTGGG +CCCAAA +CCCAAT +CCCAAC +CCCAAG +CCCATA +CCCATT +CCCATC +CCCATG +CCCACA +CCCACT +CCCACC +CCCACG +CCCAGA +CCCAGT +CCCAGC +CCCAGG +CCCTAA +CCCTAT +CCCTAC +CCCTAG +CCCTTA +CCCTTT +CCCTTC +CCCTTG +CCCTCA +CCCTCT +CCCTCC +CCCTCG +CCCTGA +CCCTGT +CCCTGC +CCCTGG +CCCCAA +CCCCAT +CCCCAC +CCCCAG +CCCCTA +CCCCTT +CCCCTC +CCCCTG +CCCCCA +CCCCCT +CCCCCC +CCCCCG +CCCCGA +CCCCGT +CCCCGC +CCCCGG +CCCGAA +CCCGAT +CCCGAC +CCCGAG +CCCGTA +CCCGTT +CCCGTC +CCCGTG +CCCGCA +CCCGCT +CCCGCC +CCCGCG +CCCGGA +CCCGGT +CCCGGC +CCCGGG +CCGAAA +CCGAAT +CCGAAC +CCGAAG +CCGATA +CCGATT +CCGATC +CCGATG +CCGACA +CCGACT +CCGACC +CCGACG +CCGAGA +CCGAGT +CCGAGC +CCGAGG +CCGTAA +CCGTAT +CCGTAC +CCGTAG +CCGTTA +CCGTTT +CCGTTC +CCGTTG +CCGTCA +CCGTCT +CCGTCC +CCGTCG +CCGTGA +CCGTGT +CCGTGC +CCGTGG +CCGCAA +CCGCAT +CCGCAC +CCGCAG +CCGCTA +CCGCTT +CCGCTC +CCGCTG +CCGCCA +CCGCCT +CCGCCC +CCGCCG +CCGCGA +CCGCGT +CCGCGC +CCGCGG +CCGGAA +CCGGAT +CCGGAC +CCGGAG +CCGGTA +CCGGTT +CCGGTC +CCGGTG +CCGGCA +CCGGCT +CCGGCC +CCGGCG +CCGGGA +CCGGGT +CCGGGC +CCGGGG +CGAAAA +CGAAAT +CGAAAC +CGAAAG +CGAATA +CGAATT +CGAATC +CGAATG +CGAACA +CGAACT +CGAACC +CGAACG +CGAAGA +CGAAGT +CGAAGC +CGAAGG +CGATAA +CGATAT +CGATAC +CGATAG +CGATTA +CGATTT +CGATTC +CGATTG +CGATCA +CGATCT +CGATCC +CGATCG +CGATGA +CGATGT +CGATGC +CGATGG +CGACAA +CGACAT +CGACAC +CGACAG +CGACTA +CGACTT +CGACTC +CGACTG +CGACCA +CGACCT +CGACCC +CGACCG +CGACGA +CGACGT +CGACGC +CGACGG +CGAGAA +CGAGAT +CGAGAC +CGAGAG +CGAGTA +CGAGTT +CGAGTC +CGAGTG +CGAGCA +CGAGCT +CGAGCC +CGAGCG +CGAGGA +CGAGGT +CGAGGC +CGAGGG +CGTAAA +CGTAAT +CGTAAC +CGTAAG +CGTATA +CGTATT +CGTATC +CGTATG +CGTACA +CGTACT +CGTACC +CGTACG +CGTAGA +CGTAGT +CGTAGC +CGTAGG +CGTTAA +CGTTAT +CGTTAC +CGTTAG +CGTTTA +CGTTTT +CGTTTC +CGTTTG +CGTTCA +CGTTCT +CGTTCC +CGTTCG +CGTTGA +CGTTGT +CGTTGC +CGTTGG +CGTCAA +CGTCAT +CGTCAC +CGTCAG +CGTCTA +CGTCTT +CGTCTC +CGTCTG +CGTCCA +CGTCCT +CGTCCC +CGTCCG +CGTCGA +CGTCGT +CGTCGC +CGTCGG +CGTGAA +CGTGAT +CGTGAC +CGTGAG +CGTGTA +CGTGTT +CGTGTC +CGTGTG +CGTGCA +CGTGCT +CGTGCC +CGTGCG +CGTGGA +CGTGGT +CGTGGC +CGTGGG +CGCAAA +CGCAAT +CGCAAC +CGCAAG +CGCATA +CGCATT +CGCATC +CGCATG +CGCACA +CGCACT +CGCACC +CGCACG +CGCAGA +CGCAGT +CGCAGC +CGCAGG +CGCTAA +CGCTAT +CGCTAC +CGCTAG +CGCTTA +CGCTTT +CGCTTC +CGCTTG +CGCTCA +CGCTCT +CGCTCC +CGCTCG +CGCTGA +CGCTGT +CGCTGC +CGCTGG +CGCCAA +CGCCAT +CGCCAC +CGCCAG +CGCCTA +CGCCTT +CGCCTC +CGCCTG +CGCCCA +CGCCCT +CGCCCC +CGCCCG +CGCCGA +CGCCGT +CGCCGC +CGCCGG +CGCGAA +CGCGAT +CGCGAC +CGCGAG +CGCGTA +CGCGTT +CGCGTC +CGCGTG +CGCGCA +CGCGCT +CGCGCC +CGCGCG +CGCGGA +CGCGGT +CGCGGC +CGCGGG +CGGAAA +CGGAAT +CGGAAC +CGGAAG +CGGATA +CGGATT +CGGATC +CGGATG +CGGACA +CGGACT +CGGACC +CGGACG +CGGAGA +CGGAGT +CGGAGC +CGGAGG +CGGTAA +CGGTAT +CGGTAC +CGGTAG +CGGTTA +CGGTTT +CGGTTC +CGGTTG +CGGTCA +CGGTCT +CGGTCC +CGGTCG +CGGTGA +CGGTGT +CGGTGC +CGGTGG +CGGCAA +CGGCAT +CGGCAC +CGGCAG +CGGCTA +CGGCTT +CGGCTC +CGGCTG +CGGCCA +CGGCCT +CGGCCC +CGGCCG +CGGCGA +CGGCGT +CGGCGC +CGGCGG +CGGGAA +CGGGAT +CGGGAC +CGGGAG +CGGGTA +CGGGTT +CGGGTC +CGGGTG +CGGGCA +CGGGCT +CGGGCC +CGGGCG +CGGGGA +CGGGGT +CGGGGC +CGGGGG +GAAAAA +GAAAAT +GAAAAC +GAAAAG +GAAATA +GAAATT +GAAATC +GAAATG +GAAACA +GAAACT +GAAACC +GAAACG +GAAAGA +GAAAGT +GAAAGC +GAAAGG +GAATAA +GAATAT +GAATAC +GAATAG +GAATTA +GAATTT +GAATTC +GAATTG +GAATCA +GAATCT +GAATCC +GAATCG +GAATGA +GAATGT +GAATGC +GAATGG +GAACAA +GAACAT +GAACAC +GAACAG +GAACTA +GAACTT +GAACTC +GAACTG +GAACCA +GAACCT +GAACCC +GAACCG +GAACGA +GAACGT +GAACGC +GAACGG +GAAGAA +GAAGAT +GAAGAC +GAAGAG +GAAGTA +GAAGTT +GAAGTC +GAAGTG +GAAGCA +GAAGCT +GAAGCC +GAAGCG +GAAGGA +GAAGGT +GAAGGC +GAAGGG +GATAAA +GATAAT +GATAAC +GATAAG +GATATA +GATATT +GATATC +GATATG +GATACA +GATACT +GATACC +GATACG +GATAGA +GATAGT +GATAGC +GATAGG +GATTAA +GATTAT +GATTAC +GATTAG +GATTTA +GATTTT +GATTTC +GATTTG +GATTCA +GATTCT +GATTCC +GATTCG +GATTGA +GATTGT +GATTGC +GATTGG +GATCAA +GATCAT +GATCAC +GATCAG +GATCTA +GATCTT +GATCTC +GATCTG +GATCCA +GATCCT +GATCCC +GATCCG +GATCGA +GATCGT +GATCGC +GATCGG +GATGAA +GATGAT +GATGAC +GATGAG +GATGTA +GATGTT +GATGTC +GATGTG +GATGCA +GATGCT +GATGCC +GATGCG +GATGGA +GATGGT +GATGGC +GATGGG +GACAAA +GACAAT +GACAAC +GACAAG +GACATA +GACATT +GACATC +GACATG +GACACA +GACACT +GACACC +GACACG +GACAGA +GACAGT +GACAGC +GACAGG +GACTAA +GACTAT +GACTAC +GACTAG +GACTTA +GACTTT +GACTTC +GACTTG +GACTCA +GACTCT +GACTCC +GACTCG +GACTGA +GACTGT +GACTGC +GACTGG +GACCAA +GACCAT +GACCAC +GACCAG +GACCTA +GACCTT +GACCTC +GACCTG +GACCCA +GACCCT +GACCCC +GACCCG +GACCGA +GACCGT +GACCGC +GACCGG +GACGAA +GACGAT +GACGAC +GACGAG +GACGTA +GACGTT +GACGTC +GACGTG +GACGCA +GACGCT +GACGCC +GACGCG +GACGGA +GACGGT +GACGGC +GACGGG +GAGAAA +GAGAAT +GAGAAC +GAGAAG +GAGATA +GAGATT +GAGATC +GAGATG +GAGACA +GAGACT +GAGACC +GAGACG +GAGAGA +GAGAGT +GAGAGC +GAGAGG +GAGTAA +GAGTAT +GAGTAC +GAGTAG +GAGTTA +GAGTTT +GAGTTC +GAGTTG +GAGTCA +GAGTCT +GAGTCC +GAGTCG +GAGTGA +GAGTGT +GAGTGC +GAGTGG +GAGCAA +GAGCAT +GAGCAC +GAGCAG +GAGCTA +GAGCTT +GAGCTC +GAGCTG +GAGCCA +GAGCCT +GAGCCC +GAGCCG +GAGCGA +GAGCGT +GAGCGC +GAGCGG +GAGGAA +GAGGAT +GAGGAC +GAGGAG +GAGGTA +GAGGTT +GAGGTC +GAGGTG +GAGGCA +GAGGCT +GAGGCC +GAGGCG +GAGGGA +GAGGGT +GAGGGC +GAGGGG +GTAAAA +GTAAAT +GTAAAC +GTAAAG +GTAATA +GTAATT +GTAATC +GTAATG +GTAACA +GTAACT +GTAACC +GTAACG +GTAAGA +GTAAGT +GTAAGC +GTAAGG +GTATAA +GTATAT +GTATAC +GTATAG +GTATTA +GTATTT +GTATTC +GTATTG +GTATCA +GTATCT +GTATCC +GTATCG +GTATGA +GTATGT +GTATGC +GTATGG +GTACAA +GTACAT +GTACAC +GTACAG +GTACTA +GTACTT +GTACTC +GTACTG +GTACCA +GTACCT +GTACCC +GTACCG +GTACGA +GTACGT +GTACGC +GTACGG +GTAGAA +GTAGAT +GTAGAC +GTAGAG +GTAGTA +GTAGTT +GTAGTC +GTAGTG +GTAGCA +GTAGCT +GTAGCC +GTAGCG +GTAGGA +GTAGGT +GTAGGC +GTAGGG +GTTAAA +GTTAAT +GTTAAC +GTTAAG +GTTATA +GTTATT +GTTATC +GTTATG +GTTACA +GTTACT +GTTACC +GTTACG +GTTAGA +GTTAGT +GTTAGC +GTTAGG +GTTTAA +GTTTAT +GTTTAC +GTTTAG +GTTTTA +GTTTTT +GTTTTC +GTTTTG +GTTTCA +GTTTCT +GTTTCC +GTTTCG +GTTTGA +GTTTGT +GTTTGC +GTTTGG +GTTCAA +GTTCAT +GTTCAC +GTTCAG +GTTCTA +GTTCTT +GTTCTC +GTTCTG +GTTCCA +GTTCCT +GTTCCC +GTTCCG +GTTCGA +GTTCGT +GTTCGC +GTTCGG +GTTGAA +GTTGAT +GTTGAC +GTTGAG +GTTGTA +GTTGTT +GTTGTC +GTTGTG +GTTGCA +GTTGCT +GTTGCC +GTTGCG +GTTGGA +GTTGGT +GTTGGC +GTTGGG +GTCAAA +GTCAAT +GTCAAC +GTCAAG +GTCATA +GTCATT +GTCATC +GTCATG +GTCACA +GTCACT +GTCACC +GTCACG +GTCAGA +GTCAGT +GTCAGC +GTCAGG +GTCTAA +GTCTAT +GTCTAC +GTCTAG +GTCTTA +GTCTTT +GTCTTC +GTCTTG +GTCTCA +GTCTCT +GTCTCC +GTCTCG +GTCTGA +GTCTGT +GTCTGC +GTCTGG +GTCCAA +GTCCAT +GTCCAC +GTCCAG +GTCCTA +GTCCTT +GTCCTC +GTCCTG +GTCCCA +GTCCCT +GTCCCC +GTCCCG +GTCCGA +GTCCGT +GTCCGC +GTCCGG +GTCGAA +GTCGAT +GTCGAC +GTCGAG +GTCGTA +GTCGTT +GTCGTC +GTCGTG +GTCGCA +GTCGCT +GTCGCC +GTCGCG +GTCGGA +GTCGGT +GTCGGC +GTCGGG +GTGAAA +GTGAAT +GTGAAC +GTGAAG +GTGATA +GTGATT +GTGATC +GTGATG +GTGACA +GTGACT +GTGACC +GTGACG +GTGAGA +GTGAGT +GTGAGC +GTGAGG +GTGTAA +GTGTAT +GTGTAC +GTGTAG +GTGTTA +GTGTTT +GTGTTC +GTGTTG +GTGTCA +GTGTCT +GTGTCC +GTGTCG +GTGTGA +GTGTGT +GTGTGC +GTGTGG +GTGCAA +GTGCAT +GTGCAC +GTGCAG +GTGCTA +GTGCTT +GTGCTC +GTGCTG +GTGCCA +GTGCCT +GTGCCC +GTGCCG +GTGCGA +GTGCGT +GTGCGC +GTGCGG +GTGGAA +GTGGAT +GTGGAC +GTGGAG +GTGGTA +GTGGTT +GTGGTC +GTGGTG +GTGGCA +GTGGCT +GTGGCC +GTGGCG +GTGGGA +GTGGGT +GTGGGC +GTGGGG +GCAAAA +GCAAAT +GCAAAC +GCAAAG +GCAATA +GCAATT +GCAATC +GCAATG +GCAACA +GCAACT +GCAACC +GCAACG +GCAAGA +GCAAGT +GCAAGC +GCAAGG +GCATAA +GCATAT +GCATAC +GCATAG +GCATTA +GCATTT +GCATTC +GCATTG +GCATCA +GCATCT +GCATCC +GCATCG +GCATGA +GCATGT +GCATGC +GCATGG +GCACAA +GCACAT +GCACAC +GCACAG +GCACTA +GCACTT +GCACTC +GCACTG +GCACCA +GCACCT +GCACCC +GCACCG +GCACGA +GCACGT +GCACGC +GCACGG +GCAGAA +GCAGAT +GCAGAC +GCAGAG +GCAGTA +GCAGTT +GCAGTC +GCAGTG +GCAGCA +GCAGCT +GCAGCC +GCAGCG +GCAGGA +GCAGGT +GCAGGC +GCAGGG +GCTAAA +GCTAAT +GCTAAC +GCTAAG +GCTATA +GCTATT +GCTATC +GCTATG +GCTACA +GCTACT +GCTACC +GCTACG +GCTAGA +GCTAGT +GCTAGC +GCTAGG +GCTTAA +GCTTAT +GCTTAC +GCTTAG +GCTTTA +GCTTTT +GCTTTC +GCTTTG +GCTTCA +GCTTCT +GCTTCC +GCTTCG +GCTTGA +GCTTGT +GCTTGC +GCTTGG +GCTCAA +GCTCAT +GCTCAC +GCTCAG +GCTCTA +GCTCTT +GCTCTC +GCTCTG +GCTCCA +GCTCCT +GCTCCC +GCTCCG +GCTCGA +GCTCGT +GCTCGC +GCTCGG +GCTGAA +GCTGAT +GCTGAC +GCTGAG +GCTGTA +GCTGTT +GCTGTC +GCTGTG +GCTGCA +GCTGCT +GCTGCC +GCTGCG +GCTGGA +GCTGGT +GCTGGC +GCTGGG +GCCAAA +GCCAAT +GCCAAC +GCCAAG +GCCATA +GCCATT +GCCATC +GCCATG +GCCACA +GCCACT +GCCACC +GCCACG +GCCAGA +GCCAGT +GCCAGC +GCCAGG +GCCTAA +GCCTAT +GCCTAC +GCCTAG +GCCTTA +GCCTTT +GCCTTC +GCCTTG +GCCTCA +GCCTCT +GCCTCC +GCCTCG +GCCTGA +GCCTGT +GCCTGC +GCCTGG +GCCCAA +GCCCAT +GCCCAC +GCCCAG +GCCCTA +GCCCTT +GCCCTC +GCCCTG +GCCCCA +GCCCCT +GCCCCC +GCCCCG +GCCCGA +GCCCGT +GCCCGC +GCCCGG +GCCGAA +GCCGAT +GCCGAC +GCCGAG +GCCGTA +GCCGTT +GCCGTC +GCCGTG +GCCGCA +GCCGCT +GCCGCC +GCCGCG +GCCGGA +GCCGGT +GCCGGC +GCCGGG +GCGAAA +GCGAAT +GCGAAC +GCGAAG +GCGATA +GCGATT +GCGATC +GCGATG +GCGACA +GCGACT +GCGACC +GCGACG +GCGAGA +GCGAGT +GCGAGC +GCGAGG +GCGTAA +GCGTAT +GCGTAC +GCGTAG +GCGTTA +GCGTTT +GCGTTC +GCGTTG +GCGTCA +GCGTCT +GCGTCC +GCGTCG +GCGTGA +GCGTGT +GCGTGC +GCGTGG +GCGCAA +GCGCAT +GCGCAC +GCGCAG +GCGCTA +GCGCTT +GCGCTC +GCGCTG +GCGCCA +GCGCCT +GCGCCC +GCGCCG +GCGCGA +GCGCGT +GCGCGC +GCGCGG +GCGGAA +GCGGAT +GCGGAC +GCGGAG +GCGGTA +GCGGTT +GCGGTC +GCGGTG +GCGGCA +GCGGCT +GCGGCC +GCGGCG +GCGGGA +GCGGGT +GCGGGC +GCGGGG +GGAAAA +GGAAAT +GGAAAC +GGAAAG +GGAATA +GGAATT +GGAATC +GGAATG +GGAACA +GGAACT +GGAACC +GGAACG +GGAAGA +GGAAGT +GGAAGC +GGAAGG +GGATAA +GGATAT +GGATAC +GGATAG +GGATTA +GGATTT +GGATTC +GGATTG +GGATCA +GGATCT +GGATCC +GGATCG +GGATGA +GGATGT +GGATGC +GGATGG +GGACAA +GGACAT +GGACAC +GGACAG +GGACTA +GGACTT +GGACTC +GGACTG +GGACCA +GGACCT +GGACCC +GGACCG +GGACGA +GGACGT +GGACGC +GGACGG +GGAGAA +GGAGAT +GGAGAC +GGAGAG +GGAGTA +GGAGTT +GGAGTC +GGAGTG +GGAGCA +GGAGCT +GGAGCC +GGAGCG +GGAGGA +GGAGGT +GGAGGC +GGAGGG +GGTAAA +GGTAAT +GGTAAC +GGTAAG +GGTATA +GGTATT +GGTATC +GGTATG +GGTACA +GGTACT +GGTACC +GGTACG +GGTAGA +GGTAGT +GGTAGC +GGTAGG +GGTTAA +GGTTAT +GGTTAC +GGTTAG +GGTTTA +GGTTTT +GGTTTC +GGTTTG +GGTTCA +GGTTCT +GGTTCC +GGTTCG +GGTTGA +GGTTGT +GGTTGC +GGTTGG +GGTCAA +GGTCAT +GGTCAC +GGTCAG +GGTCTA +GGTCTT +GGTCTC +GGTCTG +GGTCCA +GGTCCT +GGTCCC +GGTCCG +GGTCGA +GGTCGT +GGTCGC +GGTCGG +GGTGAA +GGTGAT +GGTGAC +GGTGAG +GGTGTA +GGTGTT +GGTGTC +GGTGTG +GGTGCA +GGTGCT +GGTGCC +GGTGCG +GGTGGA +GGTGGT +GGTGGC +GGTGGG +GGCAAA +GGCAAT +GGCAAC +GGCAAG +GGCATA +GGCATT +GGCATC +GGCATG +GGCACA +GGCACT +GGCACC +GGCACG +GGCAGA +GGCAGT +GGCAGC +GGCAGG +GGCTAA +GGCTAT +GGCTAC +GGCTAG +GGCTTA +GGCTTT +GGCTTC +GGCTTG +GGCTCA +GGCTCT +GGCTCC +GGCTCG +GGCTGA +GGCTGT +GGCTGC +GGCTGG +GGCCAA +GGCCAT +GGCCAC +GGCCAG +GGCCTA +GGCCTT +GGCCTC +GGCCTG +GGCCCA +GGCCCT +GGCCCC +GGCCCG +GGCCGA +GGCCGT +GGCCGC +GGCCGG +GGCGAA +GGCGAT +GGCGAC +GGCGAG +GGCGTA +GGCGTT +GGCGTC +GGCGTG +GGCGCA +GGCGCT +GGCGCC +GGCGCG +GGCGGA +GGCGGT +GGCGGC +GGCGGG +GGGAAA +GGGAAT +GGGAAC +GGGAAG +GGGATA +GGGATT +GGGATC +GGGATG +GGGACA +GGGACT +GGGACC +GGGACG +GGGAGA +GGGAGT +GGGAGC +GGGAGG +GGGTAA +GGGTAT +GGGTAC +GGGTAG +GGGTTA +GGGTTT +GGGTTC +GGGTTG +GGGTCA +GGGTCT +GGGTCC +GGGTCG +GGGTGA +GGGTGT +GGGTGC +GGGTGG +GGGCAA +GGGCAT +GGGCAC +GGGCAG +GGGCTA +GGGCTT +GGGCTC +GGGCTG +GGGCCA +GGGCCT +GGGCCC +GGGCCG +GGGCGA +GGGCGT +GGGCGC +GGGCGG +GGGGAA +GGGGAT +GGGGAC +GGGGAG +GGGGTA +GGGGTT +GGGGTC +GGGGTG +GGGGCA +GGGGCT +GGGGCC +GGGGCG +GGGGGA +GGGGGT +GGGGGC +GGGGGG diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6000/config.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..9b14e61d327bf586fcff6e817cf9d365bae18ad2 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6000/config.json @@ -0,0 +1,26 @@ +{ + "_name_or_path": "fabihamakhdoomi/TinyDNABERT", + "architectures": [ + "BertForSequenceClassification" + ], + "attention_probs_dropout_prob": 0.1, + "classifier_dropout": null, + "hidden_act": "gelu", + "hidden_dropout_prob": 0.1, + "hidden_size": 128, + "initializer_range": 0.02, + "intermediate_size": 512, + "layer_norm_eps": 1e-12, + "max_position_embeddings": 512, + "model_type": "bert", + "num_attention_heads": 2, + "num_hidden_layers": 2, + "pad_token_id": 0, + "position_embedding_type": "absolute", + "problem_type": "single_label_classification", + "torch_dtype": "float32", + "transformers_version": "4.44.0", + "type_vocab_size": 2, + "use_cache": true, + "vocab_size": 4101 +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6000/model.safetensors b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6000/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..2079df53198c970227a8fc057b4b2b6b536aba14 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6000/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2ecc88dae0ef072761a5cc1798cc026e954d4a785e3d3ecb062876253857ccf3 +size 4021688 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6000/optimizer.pt b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6000/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..4152d536efe751d87694a833b3a8fcac89527d93 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6000/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:27884c7f92f21aa74b1849da39eb0621170e86780e41d3f791454af49ed40d4e +size 8058700 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6000/rng_state.pth b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6000/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..c78d09ac076b437860c4e8c07df963c522b10249 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6000/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d44e448e2a0dd138d067ef3ddf4b7abb608f898bed2f0d0880205ed4bbbe2136 +size 14244 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6000/scheduler.pt b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6000/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..f3ebc22eebe7fac9f20742118527e91760861e6d --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6000/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7e153e480933b4bdcc7d98226c2fdfde72db931809685729c8748251ae3ecae9 +size 1064 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6000/special_tokens_map.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6000/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..a8b3208c2884c4efb86e49300fdd3dc877220cdf --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6000/special_tokens_map.json @@ -0,0 +1,7 @@ +{ + "cls_token": "[CLS]", + "mask_token": "[MASK]", + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6000/tokenizer.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6000/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..ebd782de6963b9032ce9374cfe30f5275be1a30b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6000/tokenizer.json @@ -0,0 +1,4265 @@ +{ + "version": "1.0", + "truncation": { + "direction": "Right", + "max_length": 512, + "strategy": "LongestFirst", + "stride": 0 + }, + "padding": { + "strategy": { + "Fixed": 512 + }, + "direction": "Right", + "pad_to_multiple_of": null, + "pad_id": 0, + "pad_type_id": 0, + "pad_token": "[PAD]" + }, + "added_tokens": [ + { + "id": 0, + "content": "[PAD]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 1, + "content": "[UNK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 2, + "content": "[CLS]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 3, + "content": "[SEP]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 4, + "content": "[MASK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + } + ], + "normalizer": { + "type": "BertNormalizer", + "clean_text": true, + "handle_chinese_chars": true, + "strip_accents": null, + "lowercase": false + }, + "pre_tokenizer": { + "type": "BertPreTokenizer" + }, + "post_processor": { + "type": "TemplateProcessing", + "single": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + } + ], + "pair": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "B", + "type_id": 1 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 1 + } + } + ], + "special_tokens": { + "[CLS]": { + "id": "[CLS]", + "ids": [ + 2 + ], + "tokens": [ + "[CLS]" + ] + }, + "[SEP]": { + "id": "[SEP]", + "ids": [ + 3 + ], + "tokens": [ + "[SEP]" + ] + } + } + }, + "decoder": { + "type": "WordPiece", + "prefix": "##", + "cleanup": true + }, + "model": { + "type": "WordPiece", + "unk_token": "[UNK]", + "continuing_subword_prefix": "##", + "max_input_chars_per_word": 100, + "vocab": { + "[PAD]": 0, + "[UNK]": 1, + "[CLS]": 2, + "[SEP]": 3, + "[MASK]": 4, + "AAAAAA": 5, + "AAAAAT": 6, + "AAAAAC": 7, + "AAAAAG": 8, + "AAAATA": 9, + "AAAATT": 10, + "AAAATC": 11, + "AAAATG": 12, + "AAAACA": 13, + "AAAACT": 14, + "AAAACC": 15, + "AAAACG": 16, + "AAAAGA": 17, + "AAAAGT": 18, + "AAAAGC": 19, + "AAAAGG": 20, + "AAATAA": 21, + "AAATAT": 22, + "AAATAC": 23, + "AAATAG": 24, + "AAATTA": 25, + "AAATTT": 26, + "AAATTC": 27, + "AAATTG": 28, + "AAATCA": 29, + "AAATCT": 30, + "AAATCC": 31, + "AAATCG": 32, + "AAATGA": 33, + "AAATGT": 34, + "AAATGC": 35, + "AAATGG": 36, + "AAACAA": 37, + "AAACAT": 38, + "AAACAC": 39, + "AAACAG": 40, + "AAACTA": 41, + "AAACTT": 42, + "AAACTC": 43, + "AAACTG": 44, + "AAACCA": 45, + "AAACCT": 46, + "AAACCC": 47, + "AAACCG": 48, + "AAACGA": 49, + "AAACGT": 50, + "AAACGC": 51, + "AAACGG": 52, + "AAAGAA": 53, + "AAAGAT": 54, + "AAAGAC": 55, + "AAAGAG": 56, + "AAAGTA": 57, + "AAAGTT": 58, + "AAAGTC": 59, + "AAAGTG": 60, + "AAAGCA": 61, + "AAAGCT": 62, + "AAAGCC": 63, + "AAAGCG": 64, + "AAAGGA": 65, + "AAAGGT": 66, + "AAAGGC": 67, + "AAAGGG": 68, + "AATAAA": 69, + "AATAAT": 70, + "AATAAC": 71, + "AATAAG": 72, + "AATATA": 73, + "AATATT": 74, + "AATATC": 75, + "AATATG": 76, + "AATACA": 77, + "AATACT": 78, + "AATACC": 79, + "AATACG": 80, + "AATAGA": 81, + "AATAGT": 82, + "AATAGC": 83, + "AATAGG": 84, + "AATTAA": 85, + "AATTAT": 86, + "AATTAC": 87, + "AATTAG": 88, + "AATTTA": 89, + "AATTTT": 90, + "AATTTC": 91, + "AATTTG": 92, + "AATTCA": 93, + "AATTCT": 94, + "AATTCC": 95, + "AATTCG": 96, + "AATTGA": 97, + "AATTGT": 98, + "AATTGC": 99, + "AATTGG": 100, + "AATCAA": 101, + "AATCAT": 102, + "AATCAC": 103, + "AATCAG": 104, + "AATCTA": 105, + "AATCTT": 106, + "AATCTC": 107, + "AATCTG": 108, + "AATCCA": 109, + "AATCCT": 110, + "AATCCC": 111, + "AATCCG": 112, + "AATCGA": 113, + "AATCGT": 114, + "AATCGC": 115, + "AATCGG": 116, + "AATGAA": 117, + "AATGAT": 118, + "AATGAC": 119, + "AATGAG": 120, + "AATGTA": 121, + "AATGTT": 122, + "AATGTC": 123, + "AATGTG": 124, + "AATGCA": 125, + "AATGCT": 126, + "AATGCC": 127, + "AATGCG": 128, + "AATGGA": 129, + "AATGGT": 130, + "AATGGC": 131, + "AATGGG": 132, + "AACAAA": 133, + "AACAAT": 134, + "AACAAC": 135, + "AACAAG": 136, + "AACATA": 137, + "AACATT": 138, + "AACATC": 139, + "AACATG": 140, + "AACACA": 141, + "AACACT": 142, + "AACACC": 143, + "AACACG": 144, + "AACAGA": 145, + "AACAGT": 146, + "AACAGC": 147, + "AACAGG": 148, + "AACTAA": 149, + "AACTAT": 150, + "AACTAC": 151, + "AACTAG": 152, + "AACTTA": 153, + "AACTTT": 154, + "AACTTC": 155, + "AACTTG": 156, + "AACTCA": 157, + "AACTCT": 158, + "AACTCC": 159, + "AACTCG": 160, + "AACTGA": 161, + "AACTGT": 162, + "AACTGC": 163, + "AACTGG": 164, + "AACCAA": 165, + "AACCAT": 166, + "AACCAC": 167, + "AACCAG": 168, + "AACCTA": 169, + "AACCTT": 170, + "AACCTC": 171, + "AACCTG": 172, + "AACCCA": 173, + "AACCCT": 174, + "AACCCC": 175, + "AACCCG": 176, + "AACCGA": 177, + "AACCGT": 178, + "AACCGC": 179, + "AACCGG": 180, + "AACGAA": 181, + "AACGAT": 182, + "AACGAC": 183, + "AACGAG": 184, + "AACGTA": 185, + "AACGTT": 186, + "AACGTC": 187, + "AACGTG": 188, + "AACGCA": 189, + "AACGCT": 190, + "AACGCC": 191, + "AACGCG": 192, + "AACGGA": 193, + "AACGGT": 194, + "AACGGC": 195, + "AACGGG": 196, + "AAGAAA": 197, + "AAGAAT": 198, + "AAGAAC": 199, + "AAGAAG": 200, + "AAGATA": 201, + "AAGATT": 202, + "AAGATC": 203, + "AAGATG": 204, + "AAGACA": 205, + "AAGACT": 206, + "AAGACC": 207, + "AAGACG": 208, + "AAGAGA": 209, + "AAGAGT": 210, + "AAGAGC": 211, + "AAGAGG": 212, + "AAGTAA": 213, + "AAGTAT": 214, + "AAGTAC": 215, + "AAGTAG": 216, + "AAGTTA": 217, + "AAGTTT": 218, + "AAGTTC": 219, + "AAGTTG": 220, + "AAGTCA": 221, + "AAGTCT": 222, + "AAGTCC": 223, + "AAGTCG": 224, + "AAGTGA": 225, + "AAGTGT": 226, + "AAGTGC": 227, + "AAGTGG": 228, + "AAGCAA": 229, + "AAGCAT": 230, + "AAGCAC": 231, + "AAGCAG": 232, + "AAGCTA": 233, + "AAGCTT": 234, + "AAGCTC": 235, + "AAGCTG": 236, + "AAGCCA": 237, + "AAGCCT": 238, + "AAGCCC": 239, + "AAGCCG": 240, + "AAGCGA": 241, + "AAGCGT": 242, + "AAGCGC": 243, + "AAGCGG": 244, + "AAGGAA": 245, + "AAGGAT": 246, + "AAGGAC": 247, + "AAGGAG": 248, + "AAGGTA": 249, + "AAGGTT": 250, + "AAGGTC": 251, + "AAGGTG": 252, + "AAGGCA": 253, + "AAGGCT": 254, + "AAGGCC": 255, + "AAGGCG": 256, + "AAGGGA": 257, + "AAGGGT": 258, + "AAGGGC": 259, + "AAGGGG": 260, + "ATAAAA": 261, + "ATAAAT": 262, + "ATAAAC": 263, + "ATAAAG": 264, + "ATAATA": 265, + "ATAATT": 266, + "ATAATC": 267, + "ATAATG": 268, + "ATAACA": 269, + "ATAACT": 270, + "ATAACC": 271, + "ATAACG": 272, + "ATAAGA": 273, + "ATAAGT": 274, + "ATAAGC": 275, + "ATAAGG": 276, + "ATATAA": 277, + "ATATAT": 278, + "ATATAC": 279, + "ATATAG": 280, + "ATATTA": 281, + "ATATTT": 282, + "ATATTC": 283, + "ATATTG": 284, + "ATATCA": 285, + "ATATCT": 286, + "ATATCC": 287, + "ATATCG": 288, + "ATATGA": 289, + "ATATGT": 290, + "ATATGC": 291, + "ATATGG": 292, + "ATACAA": 293, + "ATACAT": 294, + "ATACAC": 295, + "ATACAG": 296, + "ATACTA": 297, + "ATACTT": 298, + "ATACTC": 299, + "ATACTG": 300, + "ATACCA": 301, + "ATACCT": 302, + "ATACCC": 303, + "ATACCG": 304, + "ATACGA": 305, + "ATACGT": 306, + "ATACGC": 307, + "ATACGG": 308, + "ATAGAA": 309, + "ATAGAT": 310, + "ATAGAC": 311, + "ATAGAG": 312, + "ATAGTA": 313, + "ATAGTT": 314, + "ATAGTC": 315, + "ATAGTG": 316, + "ATAGCA": 317, + "ATAGCT": 318, + "ATAGCC": 319, + "ATAGCG": 320, + "ATAGGA": 321, + "ATAGGT": 322, + "ATAGGC": 323, + "ATAGGG": 324, + "ATTAAA": 325, + "ATTAAT": 326, + "ATTAAC": 327, + "ATTAAG": 328, + "ATTATA": 329, + "ATTATT": 330, + "ATTATC": 331, + "ATTATG": 332, + "ATTACA": 333, + "ATTACT": 334, + "ATTACC": 335, + "ATTACG": 336, + "ATTAGA": 337, + "ATTAGT": 338, + "ATTAGC": 339, + "ATTAGG": 340, + "ATTTAA": 341, + "ATTTAT": 342, + "ATTTAC": 343, + "ATTTAG": 344, + "ATTTTA": 345, + "ATTTTT": 346, + "ATTTTC": 347, + "ATTTTG": 348, + "ATTTCA": 349, + "ATTTCT": 350, + "ATTTCC": 351, + "ATTTCG": 352, + "ATTTGA": 353, + "ATTTGT": 354, + "ATTTGC": 355, + "ATTTGG": 356, + "ATTCAA": 357, + "ATTCAT": 358, + "ATTCAC": 359, + "ATTCAG": 360, + "ATTCTA": 361, + "ATTCTT": 362, + "ATTCTC": 363, + "ATTCTG": 364, + "ATTCCA": 365, + "ATTCCT": 366, + "ATTCCC": 367, + "ATTCCG": 368, + "ATTCGA": 369, + "ATTCGT": 370, + "ATTCGC": 371, + "ATTCGG": 372, + "ATTGAA": 373, + "ATTGAT": 374, + "ATTGAC": 375, + "ATTGAG": 376, + "ATTGTA": 377, + "ATTGTT": 378, + "ATTGTC": 379, + "ATTGTG": 380, + "ATTGCA": 381, + "ATTGCT": 382, + "ATTGCC": 383, + "ATTGCG": 384, + "ATTGGA": 385, + "ATTGGT": 386, + "ATTGGC": 387, + "ATTGGG": 388, + "ATCAAA": 389, + "ATCAAT": 390, + "ATCAAC": 391, + "ATCAAG": 392, + "ATCATA": 393, + "ATCATT": 394, + "ATCATC": 395, + "ATCATG": 396, + "ATCACA": 397, + "ATCACT": 398, + "ATCACC": 399, + "ATCACG": 400, + "ATCAGA": 401, + "ATCAGT": 402, + "ATCAGC": 403, + "ATCAGG": 404, + "ATCTAA": 405, + "ATCTAT": 406, + "ATCTAC": 407, + "ATCTAG": 408, + "ATCTTA": 409, + "ATCTTT": 410, + "ATCTTC": 411, + "ATCTTG": 412, + "ATCTCA": 413, + "ATCTCT": 414, + "ATCTCC": 415, + "ATCTCG": 416, + "ATCTGA": 417, + "ATCTGT": 418, + "ATCTGC": 419, + "ATCTGG": 420, + "ATCCAA": 421, + "ATCCAT": 422, + "ATCCAC": 423, + "ATCCAG": 424, + "ATCCTA": 425, + "ATCCTT": 426, + "ATCCTC": 427, + "ATCCTG": 428, + "ATCCCA": 429, + "ATCCCT": 430, + "ATCCCC": 431, + "ATCCCG": 432, + "ATCCGA": 433, + "ATCCGT": 434, + "ATCCGC": 435, + "ATCCGG": 436, + "ATCGAA": 437, + "ATCGAT": 438, + "ATCGAC": 439, + "ATCGAG": 440, + "ATCGTA": 441, + "ATCGTT": 442, + "ATCGTC": 443, + "ATCGTG": 444, + "ATCGCA": 445, + "ATCGCT": 446, + "ATCGCC": 447, + "ATCGCG": 448, + "ATCGGA": 449, + "ATCGGT": 450, + "ATCGGC": 451, + "ATCGGG": 452, + "ATGAAA": 453, + "ATGAAT": 454, + "ATGAAC": 455, + "ATGAAG": 456, + "ATGATA": 457, + "ATGATT": 458, + "ATGATC": 459, + "ATGATG": 460, + "ATGACA": 461, + "ATGACT": 462, + "ATGACC": 463, + "ATGACG": 464, + "ATGAGA": 465, + "ATGAGT": 466, + "ATGAGC": 467, + "ATGAGG": 468, + "ATGTAA": 469, + "ATGTAT": 470, + "ATGTAC": 471, + "ATGTAG": 472, + "ATGTTA": 473, + "ATGTTT": 474, + "ATGTTC": 475, + "ATGTTG": 476, + "ATGTCA": 477, + "ATGTCT": 478, + "ATGTCC": 479, + "ATGTCG": 480, + "ATGTGA": 481, + "ATGTGT": 482, + "ATGTGC": 483, + "ATGTGG": 484, + "ATGCAA": 485, + "ATGCAT": 486, + "ATGCAC": 487, + "ATGCAG": 488, + "ATGCTA": 489, + "ATGCTT": 490, + "ATGCTC": 491, + "ATGCTG": 492, + "ATGCCA": 493, + "ATGCCT": 494, + "ATGCCC": 495, + "ATGCCG": 496, + "ATGCGA": 497, + "ATGCGT": 498, + "ATGCGC": 499, + "ATGCGG": 500, + "ATGGAA": 501, + "ATGGAT": 502, + "ATGGAC": 503, + "ATGGAG": 504, + "ATGGTA": 505, + "ATGGTT": 506, + "ATGGTC": 507, + "ATGGTG": 508, + "ATGGCA": 509, + "ATGGCT": 510, + "ATGGCC": 511, + "ATGGCG": 512, + "ATGGGA": 513, + "ATGGGT": 514, + "ATGGGC": 515, + "ATGGGG": 516, + "ACAAAA": 517, + "ACAAAT": 518, + "ACAAAC": 519, + "ACAAAG": 520, + "ACAATA": 521, + "ACAATT": 522, + "ACAATC": 523, + "ACAATG": 524, + "ACAACA": 525, + "ACAACT": 526, + "ACAACC": 527, + "ACAACG": 528, + "ACAAGA": 529, + "ACAAGT": 530, + "ACAAGC": 531, + "ACAAGG": 532, + "ACATAA": 533, + "ACATAT": 534, + "ACATAC": 535, + "ACATAG": 536, + "ACATTA": 537, + "ACATTT": 538, + "ACATTC": 539, + "ACATTG": 540, + "ACATCA": 541, + "ACATCT": 542, + "ACATCC": 543, + "ACATCG": 544, + "ACATGA": 545, + "ACATGT": 546, + "ACATGC": 547, + "ACATGG": 548, + "ACACAA": 549, + "ACACAT": 550, + "ACACAC": 551, + "ACACAG": 552, + "ACACTA": 553, + "ACACTT": 554, + "ACACTC": 555, + "ACACTG": 556, + "ACACCA": 557, + "ACACCT": 558, + "ACACCC": 559, + "ACACCG": 560, + "ACACGA": 561, + "ACACGT": 562, + "ACACGC": 563, + "ACACGG": 564, + "ACAGAA": 565, + "ACAGAT": 566, + "ACAGAC": 567, + "ACAGAG": 568, + "ACAGTA": 569, + "ACAGTT": 570, + "ACAGTC": 571, + "ACAGTG": 572, + "ACAGCA": 573, + "ACAGCT": 574, + "ACAGCC": 575, + "ACAGCG": 576, + "ACAGGA": 577, + "ACAGGT": 578, + "ACAGGC": 579, + "ACAGGG": 580, + "ACTAAA": 581, + "ACTAAT": 582, + "ACTAAC": 583, + "ACTAAG": 584, + "ACTATA": 585, + "ACTATT": 586, + "ACTATC": 587, + "ACTATG": 588, + "ACTACA": 589, + "ACTACT": 590, + "ACTACC": 591, + "ACTACG": 592, + "ACTAGA": 593, + "ACTAGT": 594, + "ACTAGC": 595, + "ACTAGG": 596, + "ACTTAA": 597, + "ACTTAT": 598, + "ACTTAC": 599, + "ACTTAG": 600, + "ACTTTA": 601, + "ACTTTT": 602, + "ACTTTC": 603, + "ACTTTG": 604, + "ACTTCA": 605, + "ACTTCT": 606, + "ACTTCC": 607, + "ACTTCG": 608, + "ACTTGA": 609, + "ACTTGT": 610, + "ACTTGC": 611, + "ACTTGG": 612, + "ACTCAA": 613, + "ACTCAT": 614, + "ACTCAC": 615, + "ACTCAG": 616, + "ACTCTA": 617, + "ACTCTT": 618, + "ACTCTC": 619, + "ACTCTG": 620, + "ACTCCA": 621, + "ACTCCT": 622, + "ACTCCC": 623, + "ACTCCG": 624, + "ACTCGA": 625, + "ACTCGT": 626, + "ACTCGC": 627, + "ACTCGG": 628, + "ACTGAA": 629, + "ACTGAT": 630, + "ACTGAC": 631, + "ACTGAG": 632, + "ACTGTA": 633, + "ACTGTT": 634, + "ACTGTC": 635, + "ACTGTG": 636, + "ACTGCA": 637, + "ACTGCT": 638, + "ACTGCC": 639, + "ACTGCG": 640, + "ACTGGA": 641, + "ACTGGT": 642, + "ACTGGC": 643, + "ACTGGG": 644, + "ACCAAA": 645, + "ACCAAT": 646, + "ACCAAC": 647, + "ACCAAG": 648, + "ACCATA": 649, + "ACCATT": 650, + "ACCATC": 651, + "ACCATG": 652, + "ACCACA": 653, + "ACCACT": 654, + "ACCACC": 655, + "ACCACG": 656, + "ACCAGA": 657, + "ACCAGT": 658, + "ACCAGC": 659, + "ACCAGG": 660, + "ACCTAA": 661, + "ACCTAT": 662, + "ACCTAC": 663, + "ACCTAG": 664, + "ACCTTA": 665, + "ACCTTT": 666, + "ACCTTC": 667, + "ACCTTG": 668, + "ACCTCA": 669, + "ACCTCT": 670, + "ACCTCC": 671, + "ACCTCG": 672, + "ACCTGA": 673, + "ACCTGT": 674, + "ACCTGC": 675, + "ACCTGG": 676, + "ACCCAA": 677, + "ACCCAT": 678, + "ACCCAC": 679, + "ACCCAG": 680, + "ACCCTA": 681, + "ACCCTT": 682, + "ACCCTC": 683, + "ACCCTG": 684, + "ACCCCA": 685, + "ACCCCT": 686, + "ACCCCC": 687, + "ACCCCG": 688, + "ACCCGA": 689, + "ACCCGT": 690, + "ACCCGC": 691, + "ACCCGG": 692, + "ACCGAA": 693, + "ACCGAT": 694, + "ACCGAC": 695, + "ACCGAG": 696, + "ACCGTA": 697, + "ACCGTT": 698, + "ACCGTC": 699, + "ACCGTG": 700, + "ACCGCA": 701, + "ACCGCT": 702, + "ACCGCC": 703, + "ACCGCG": 704, + "ACCGGA": 705, + "ACCGGT": 706, + "ACCGGC": 707, + "ACCGGG": 708, + "ACGAAA": 709, + "ACGAAT": 710, + "ACGAAC": 711, + "ACGAAG": 712, + "ACGATA": 713, + "ACGATT": 714, + "ACGATC": 715, + "ACGATG": 716, + "ACGACA": 717, + "ACGACT": 718, + "ACGACC": 719, + "ACGACG": 720, + "ACGAGA": 721, + "ACGAGT": 722, + "ACGAGC": 723, + "ACGAGG": 724, + "ACGTAA": 725, + "ACGTAT": 726, + "ACGTAC": 727, + "ACGTAG": 728, + "ACGTTA": 729, + "ACGTTT": 730, + "ACGTTC": 731, + "ACGTTG": 732, + "ACGTCA": 733, + "ACGTCT": 734, + "ACGTCC": 735, + "ACGTCG": 736, + "ACGTGA": 737, + "ACGTGT": 738, + "ACGTGC": 739, + "ACGTGG": 740, + "ACGCAA": 741, + "ACGCAT": 742, + "ACGCAC": 743, + "ACGCAG": 744, + "ACGCTA": 745, + "ACGCTT": 746, + "ACGCTC": 747, + "ACGCTG": 748, + "ACGCCA": 749, + "ACGCCT": 750, + "ACGCCC": 751, + "ACGCCG": 752, + "ACGCGA": 753, + "ACGCGT": 754, + "ACGCGC": 755, + "ACGCGG": 756, + "ACGGAA": 757, + "ACGGAT": 758, + "ACGGAC": 759, + "ACGGAG": 760, + "ACGGTA": 761, + "ACGGTT": 762, + "ACGGTC": 763, + "ACGGTG": 764, + "ACGGCA": 765, + "ACGGCT": 766, + "ACGGCC": 767, + "ACGGCG": 768, + "ACGGGA": 769, + "ACGGGT": 770, + "ACGGGC": 771, + "ACGGGG": 772, + "AGAAAA": 773, + "AGAAAT": 774, + "AGAAAC": 775, + "AGAAAG": 776, + "AGAATA": 777, + "AGAATT": 778, + "AGAATC": 779, + "AGAATG": 780, + "AGAACA": 781, + "AGAACT": 782, + "AGAACC": 783, + "AGAACG": 784, + "AGAAGA": 785, + "AGAAGT": 786, + "AGAAGC": 787, + "AGAAGG": 788, + "AGATAA": 789, + "AGATAT": 790, + "AGATAC": 791, + "AGATAG": 792, + "AGATTA": 793, + "AGATTT": 794, + "AGATTC": 795, + "AGATTG": 796, + "AGATCA": 797, + "AGATCT": 798, + "AGATCC": 799, + "AGATCG": 800, + "AGATGA": 801, + "AGATGT": 802, + "AGATGC": 803, + "AGATGG": 804, + "AGACAA": 805, + "AGACAT": 806, + "AGACAC": 807, + "AGACAG": 808, + "AGACTA": 809, + "AGACTT": 810, + "AGACTC": 811, + "AGACTG": 812, + "AGACCA": 813, + "AGACCT": 814, + "AGACCC": 815, + "AGACCG": 816, + "AGACGA": 817, + "AGACGT": 818, + "AGACGC": 819, + "AGACGG": 820, + "AGAGAA": 821, + "AGAGAT": 822, + "AGAGAC": 823, + "AGAGAG": 824, + "AGAGTA": 825, + "AGAGTT": 826, + "AGAGTC": 827, + "AGAGTG": 828, + "AGAGCA": 829, + "AGAGCT": 830, + "AGAGCC": 831, + "AGAGCG": 832, + "AGAGGA": 833, + "AGAGGT": 834, + "AGAGGC": 835, + "AGAGGG": 836, + "AGTAAA": 837, + "AGTAAT": 838, + "AGTAAC": 839, + "AGTAAG": 840, + "AGTATA": 841, + "AGTATT": 842, + "AGTATC": 843, + "AGTATG": 844, + "AGTACA": 845, + "AGTACT": 846, + "AGTACC": 847, + "AGTACG": 848, + "AGTAGA": 849, + "AGTAGT": 850, + "AGTAGC": 851, + "AGTAGG": 852, + "AGTTAA": 853, + "AGTTAT": 854, + "AGTTAC": 855, + "AGTTAG": 856, + "AGTTTA": 857, + "AGTTTT": 858, + "AGTTTC": 859, + "AGTTTG": 860, + "AGTTCA": 861, + "AGTTCT": 862, + "AGTTCC": 863, + "AGTTCG": 864, + "AGTTGA": 865, + "AGTTGT": 866, + "AGTTGC": 867, + "AGTTGG": 868, + "AGTCAA": 869, + "AGTCAT": 870, + "AGTCAC": 871, + "AGTCAG": 872, + "AGTCTA": 873, + "AGTCTT": 874, + "AGTCTC": 875, + "AGTCTG": 876, + "AGTCCA": 877, + "AGTCCT": 878, + "AGTCCC": 879, + "AGTCCG": 880, + "AGTCGA": 881, + "AGTCGT": 882, + "AGTCGC": 883, + "AGTCGG": 884, + "AGTGAA": 885, + "AGTGAT": 886, + "AGTGAC": 887, + "AGTGAG": 888, + "AGTGTA": 889, + "AGTGTT": 890, + "AGTGTC": 891, + "AGTGTG": 892, + "AGTGCA": 893, + "AGTGCT": 894, + "AGTGCC": 895, + "AGTGCG": 896, + "AGTGGA": 897, + "AGTGGT": 898, + "AGTGGC": 899, + "AGTGGG": 900, + "AGCAAA": 901, + "AGCAAT": 902, + "AGCAAC": 903, + "AGCAAG": 904, + "AGCATA": 905, + "AGCATT": 906, + "AGCATC": 907, + "AGCATG": 908, + "AGCACA": 909, + "AGCACT": 910, + "AGCACC": 911, + "AGCACG": 912, + "AGCAGA": 913, + "AGCAGT": 914, + "AGCAGC": 915, + "AGCAGG": 916, + "AGCTAA": 917, + "AGCTAT": 918, + "AGCTAC": 919, + "AGCTAG": 920, + "AGCTTA": 921, + "AGCTTT": 922, + "AGCTTC": 923, + "AGCTTG": 924, + "AGCTCA": 925, + "AGCTCT": 926, + "AGCTCC": 927, + "AGCTCG": 928, + "AGCTGA": 929, + "AGCTGT": 930, + "AGCTGC": 931, + "AGCTGG": 932, + "AGCCAA": 933, + "AGCCAT": 934, + "AGCCAC": 935, + "AGCCAG": 936, + "AGCCTA": 937, + "AGCCTT": 938, + "AGCCTC": 939, + "AGCCTG": 940, + "AGCCCA": 941, + "AGCCCT": 942, + "AGCCCC": 943, + "AGCCCG": 944, + "AGCCGA": 945, + "AGCCGT": 946, + "AGCCGC": 947, + "AGCCGG": 948, + "AGCGAA": 949, + "AGCGAT": 950, + "AGCGAC": 951, + "AGCGAG": 952, + "AGCGTA": 953, + "AGCGTT": 954, + "AGCGTC": 955, + "AGCGTG": 956, + "AGCGCA": 957, + "AGCGCT": 958, + "AGCGCC": 959, + "AGCGCG": 960, + "AGCGGA": 961, + "AGCGGT": 962, + "AGCGGC": 963, + "AGCGGG": 964, + "AGGAAA": 965, + "AGGAAT": 966, + "AGGAAC": 967, + "AGGAAG": 968, + "AGGATA": 969, + "AGGATT": 970, + "AGGATC": 971, + "AGGATG": 972, + "AGGACA": 973, + "AGGACT": 974, + "AGGACC": 975, + "AGGACG": 976, + "AGGAGA": 977, + "AGGAGT": 978, + "AGGAGC": 979, + "AGGAGG": 980, + "AGGTAA": 981, + "AGGTAT": 982, + "AGGTAC": 983, + "AGGTAG": 984, + "AGGTTA": 985, + "AGGTTT": 986, + "AGGTTC": 987, + "AGGTTG": 988, + "AGGTCA": 989, + "AGGTCT": 990, + "AGGTCC": 991, + "AGGTCG": 992, + "AGGTGA": 993, + "AGGTGT": 994, + "AGGTGC": 995, + "AGGTGG": 996, + "AGGCAA": 997, + "AGGCAT": 998, + "AGGCAC": 999, + "AGGCAG": 1000, + "AGGCTA": 1001, + "AGGCTT": 1002, + "AGGCTC": 1003, + "AGGCTG": 1004, + "AGGCCA": 1005, + "AGGCCT": 1006, + "AGGCCC": 1007, + "AGGCCG": 1008, + "AGGCGA": 1009, + "AGGCGT": 1010, + "AGGCGC": 1011, + "AGGCGG": 1012, + "AGGGAA": 1013, + "AGGGAT": 1014, + "AGGGAC": 1015, + "AGGGAG": 1016, + "AGGGTA": 1017, + "AGGGTT": 1018, + "AGGGTC": 1019, + "AGGGTG": 1020, + "AGGGCA": 1021, + "AGGGCT": 1022, + "AGGGCC": 1023, + "AGGGCG": 1024, + "AGGGGA": 1025, + "AGGGGT": 1026, + "AGGGGC": 1027, + "AGGGGG": 1028, + "TAAAAA": 1029, + "TAAAAT": 1030, + "TAAAAC": 1031, + "TAAAAG": 1032, + "TAAATA": 1033, + "TAAATT": 1034, + "TAAATC": 1035, + "TAAATG": 1036, + "TAAACA": 1037, + "TAAACT": 1038, + "TAAACC": 1039, + "TAAACG": 1040, + "TAAAGA": 1041, + "TAAAGT": 1042, + "TAAAGC": 1043, + "TAAAGG": 1044, + "TAATAA": 1045, + "TAATAT": 1046, + "TAATAC": 1047, + "TAATAG": 1048, + "TAATTA": 1049, + "TAATTT": 1050, + "TAATTC": 1051, + "TAATTG": 1052, + "TAATCA": 1053, + "TAATCT": 1054, + "TAATCC": 1055, + "TAATCG": 1056, + "TAATGA": 1057, + "TAATGT": 1058, + "TAATGC": 1059, + "TAATGG": 1060, + "TAACAA": 1061, + "TAACAT": 1062, + "TAACAC": 1063, + "TAACAG": 1064, + "TAACTA": 1065, + "TAACTT": 1066, + "TAACTC": 1067, + "TAACTG": 1068, + "TAACCA": 1069, + "TAACCT": 1070, + "TAACCC": 1071, + "TAACCG": 1072, + "TAACGA": 1073, + "TAACGT": 1074, + "TAACGC": 1075, + "TAACGG": 1076, + "TAAGAA": 1077, + "TAAGAT": 1078, + "TAAGAC": 1079, + "TAAGAG": 1080, + "TAAGTA": 1081, + "TAAGTT": 1082, + "TAAGTC": 1083, + "TAAGTG": 1084, + "TAAGCA": 1085, + "TAAGCT": 1086, + "TAAGCC": 1087, + "TAAGCG": 1088, + "TAAGGA": 1089, + "TAAGGT": 1090, + "TAAGGC": 1091, + "TAAGGG": 1092, + "TATAAA": 1093, + "TATAAT": 1094, + "TATAAC": 1095, + "TATAAG": 1096, + "TATATA": 1097, + "TATATT": 1098, + "TATATC": 1099, + "TATATG": 1100, + "TATACA": 1101, + "TATACT": 1102, + "TATACC": 1103, + "TATACG": 1104, + "TATAGA": 1105, + "TATAGT": 1106, + "TATAGC": 1107, + "TATAGG": 1108, + "TATTAA": 1109, + "TATTAT": 1110, + "TATTAC": 1111, + "TATTAG": 1112, + "TATTTA": 1113, + "TATTTT": 1114, + "TATTTC": 1115, + "TATTTG": 1116, + "TATTCA": 1117, + "TATTCT": 1118, + "TATTCC": 1119, + "TATTCG": 1120, + "TATTGA": 1121, + "TATTGT": 1122, + "TATTGC": 1123, + "TATTGG": 1124, + "TATCAA": 1125, + "TATCAT": 1126, + "TATCAC": 1127, + "TATCAG": 1128, + "TATCTA": 1129, + "TATCTT": 1130, + "TATCTC": 1131, + "TATCTG": 1132, + "TATCCA": 1133, + "TATCCT": 1134, + "TATCCC": 1135, + "TATCCG": 1136, + "TATCGA": 1137, + "TATCGT": 1138, + "TATCGC": 1139, + "TATCGG": 1140, + "TATGAA": 1141, + "TATGAT": 1142, + "TATGAC": 1143, + "TATGAG": 1144, + "TATGTA": 1145, + "TATGTT": 1146, + "TATGTC": 1147, + "TATGTG": 1148, + "TATGCA": 1149, + "TATGCT": 1150, + "TATGCC": 1151, + "TATGCG": 1152, + "TATGGA": 1153, + "TATGGT": 1154, + "TATGGC": 1155, + "TATGGG": 1156, + "TACAAA": 1157, + "TACAAT": 1158, + "TACAAC": 1159, + "TACAAG": 1160, + "TACATA": 1161, + "TACATT": 1162, + "TACATC": 1163, + "TACATG": 1164, + "TACACA": 1165, + "TACACT": 1166, + "TACACC": 1167, + "TACACG": 1168, + "TACAGA": 1169, + "TACAGT": 1170, + "TACAGC": 1171, + "TACAGG": 1172, + "TACTAA": 1173, + "TACTAT": 1174, + "TACTAC": 1175, + "TACTAG": 1176, + "TACTTA": 1177, + "TACTTT": 1178, + "TACTTC": 1179, + "TACTTG": 1180, + "TACTCA": 1181, + "TACTCT": 1182, + "TACTCC": 1183, + "TACTCG": 1184, + "TACTGA": 1185, + "TACTGT": 1186, + "TACTGC": 1187, + "TACTGG": 1188, + "TACCAA": 1189, + "TACCAT": 1190, + "TACCAC": 1191, + "TACCAG": 1192, + "TACCTA": 1193, + "TACCTT": 1194, + "TACCTC": 1195, + "TACCTG": 1196, + "TACCCA": 1197, + "TACCCT": 1198, + "TACCCC": 1199, + "TACCCG": 1200, + "TACCGA": 1201, + "TACCGT": 1202, + "TACCGC": 1203, + "TACCGG": 1204, + "TACGAA": 1205, + "TACGAT": 1206, + "TACGAC": 1207, + "TACGAG": 1208, + "TACGTA": 1209, + "TACGTT": 1210, + "TACGTC": 1211, + "TACGTG": 1212, + "TACGCA": 1213, + "TACGCT": 1214, + "TACGCC": 1215, + "TACGCG": 1216, + "TACGGA": 1217, + "TACGGT": 1218, + "TACGGC": 1219, + "TACGGG": 1220, + "TAGAAA": 1221, + "TAGAAT": 1222, + "TAGAAC": 1223, + "TAGAAG": 1224, + "TAGATA": 1225, + "TAGATT": 1226, + "TAGATC": 1227, + "TAGATG": 1228, + "TAGACA": 1229, + "TAGACT": 1230, + "TAGACC": 1231, + "TAGACG": 1232, + "TAGAGA": 1233, + "TAGAGT": 1234, + "TAGAGC": 1235, + "TAGAGG": 1236, + "TAGTAA": 1237, + "TAGTAT": 1238, + "TAGTAC": 1239, + "TAGTAG": 1240, + "TAGTTA": 1241, + "TAGTTT": 1242, + "TAGTTC": 1243, + "TAGTTG": 1244, + "TAGTCA": 1245, + "TAGTCT": 1246, + "TAGTCC": 1247, + "TAGTCG": 1248, + "TAGTGA": 1249, + "TAGTGT": 1250, + "TAGTGC": 1251, + "TAGTGG": 1252, + "TAGCAA": 1253, + "TAGCAT": 1254, + "TAGCAC": 1255, + "TAGCAG": 1256, + "TAGCTA": 1257, + "TAGCTT": 1258, + "TAGCTC": 1259, + "TAGCTG": 1260, + "TAGCCA": 1261, + "TAGCCT": 1262, + "TAGCCC": 1263, + "TAGCCG": 1264, + "TAGCGA": 1265, + "TAGCGT": 1266, + "TAGCGC": 1267, + "TAGCGG": 1268, + "TAGGAA": 1269, + "TAGGAT": 1270, + "TAGGAC": 1271, + "TAGGAG": 1272, + "TAGGTA": 1273, + "TAGGTT": 1274, + "TAGGTC": 1275, + "TAGGTG": 1276, + "TAGGCA": 1277, + "TAGGCT": 1278, + "TAGGCC": 1279, + "TAGGCG": 1280, + "TAGGGA": 1281, + "TAGGGT": 1282, + "TAGGGC": 1283, + "TAGGGG": 1284, + "TTAAAA": 1285, + "TTAAAT": 1286, + "TTAAAC": 1287, + "TTAAAG": 1288, + "TTAATA": 1289, + "TTAATT": 1290, + "TTAATC": 1291, + "TTAATG": 1292, + "TTAACA": 1293, + "TTAACT": 1294, + "TTAACC": 1295, + "TTAACG": 1296, + "TTAAGA": 1297, + "TTAAGT": 1298, + "TTAAGC": 1299, + "TTAAGG": 1300, + "TTATAA": 1301, + "TTATAT": 1302, + "TTATAC": 1303, + "TTATAG": 1304, + "TTATTA": 1305, + "TTATTT": 1306, + "TTATTC": 1307, + "TTATTG": 1308, + "TTATCA": 1309, + "TTATCT": 1310, + "TTATCC": 1311, + "TTATCG": 1312, + "TTATGA": 1313, + "TTATGT": 1314, + "TTATGC": 1315, + "TTATGG": 1316, + "TTACAA": 1317, + "TTACAT": 1318, + "TTACAC": 1319, + "TTACAG": 1320, + "TTACTA": 1321, + "TTACTT": 1322, + "TTACTC": 1323, + "TTACTG": 1324, + "TTACCA": 1325, + "TTACCT": 1326, + "TTACCC": 1327, + "TTACCG": 1328, + "TTACGA": 1329, + "TTACGT": 1330, + "TTACGC": 1331, + "TTACGG": 1332, + "TTAGAA": 1333, + "TTAGAT": 1334, + "TTAGAC": 1335, + "TTAGAG": 1336, + "TTAGTA": 1337, + "TTAGTT": 1338, + "TTAGTC": 1339, + "TTAGTG": 1340, + "TTAGCA": 1341, + "TTAGCT": 1342, + "TTAGCC": 1343, + "TTAGCG": 1344, + "TTAGGA": 1345, + "TTAGGT": 1346, + "TTAGGC": 1347, + "TTAGGG": 1348, + "TTTAAA": 1349, + "TTTAAT": 1350, + "TTTAAC": 1351, + "TTTAAG": 1352, + "TTTATA": 1353, + "TTTATT": 1354, + "TTTATC": 1355, + "TTTATG": 1356, + "TTTACA": 1357, + "TTTACT": 1358, + "TTTACC": 1359, + "TTTACG": 1360, + "TTTAGA": 1361, + "TTTAGT": 1362, + "TTTAGC": 1363, + "TTTAGG": 1364, + "TTTTAA": 1365, + "TTTTAT": 1366, + "TTTTAC": 1367, + "TTTTAG": 1368, + "TTTTTA": 1369, + "TTTTTT": 1370, + "TTTTTC": 1371, + "TTTTTG": 1372, + "TTTTCA": 1373, + "TTTTCT": 1374, + "TTTTCC": 1375, + "TTTTCG": 1376, + "TTTTGA": 1377, + "TTTTGT": 1378, + "TTTTGC": 1379, + "TTTTGG": 1380, + "TTTCAA": 1381, + "TTTCAT": 1382, + "TTTCAC": 1383, + "TTTCAG": 1384, + "TTTCTA": 1385, + "TTTCTT": 1386, + "TTTCTC": 1387, + "TTTCTG": 1388, + "TTTCCA": 1389, + "TTTCCT": 1390, + "TTTCCC": 1391, + "TTTCCG": 1392, + "TTTCGA": 1393, + "TTTCGT": 1394, + "TTTCGC": 1395, + "TTTCGG": 1396, + "TTTGAA": 1397, + "TTTGAT": 1398, + "TTTGAC": 1399, + "TTTGAG": 1400, + "TTTGTA": 1401, + "TTTGTT": 1402, + "TTTGTC": 1403, + "TTTGTG": 1404, + "TTTGCA": 1405, + "TTTGCT": 1406, + "TTTGCC": 1407, + "TTTGCG": 1408, + "TTTGGA": 1409, + "TTTGGT": 1410, + "TTTGGC": 1411, + "TTTGGG": 1412, + "TTCAAA": 1413, + "TTCAAT": 1414, + "TTCAAC": 1415, + "TTCAAG": 1416, + "TTCATA": 1417, + "TTCATT": 1418, + "TTCATC": 1419, + "TTCATG": 1420, + "TTCACA": 1421, + "TTCACT": 1422, + "TTCACC": 1423, + "TTCACG": 1424, + "TTCAGA": 1425, + "TTCAGT": 1426, + "TTCAGC": 1427, + "TTCAGG": 1428, + "TTCTAA": 1429, + "TTCTAT": 1430, + "TTCTAC": 1431, + "TTCTAG": 1432, + "TTCTTA": 1433, + "TTCTTT": 1434, + "TTCTTC": 1435, + "TTCTTG": 1436, + "TTCTCA": 1437, + "TTCTCT": 1438, + "TTCTCC": 1439, + "TTCTCG": 1440, + "TTCTGA": 1441, + "TTCTGT": 1442, + "TTCTGC": 1443, + "TTCTGG": 1444, + "TTCCAA": 1445, + "TTCCAT": 1446, + "TTCCAC": 1447, + "TTCCAG": 1448, + "TTCCTA": 1449, + "TTCCTT": 1450, + "TTCCTC": 1451, + "TTCCTG": 1452, + "TTCCCA": 1453, + "TTCCCT": 1454, + "TTCCCC": 1455, + "TTCCCG": 1456, + "TTCCGA": 1457, + "TTCCGT": 1458, + "TTCCGC": 1459, + "TTCCGG": 1460, + "TTCGAA": 1461, + "TTCGAT": 1462, + "TTCGAC": 1463, + "TTCGAG": 1464, + "TTCGTA": 1465, + "TTCGTT": 1466, + "TTCGTC": 1467, + "TTCGTG": 1468, + "TTCGCA": 1469, + "TTCGCT": 1470, + "TTCGCC": 1471, + "TTCGCG": 1472, + "TTCGGA": 1473, + "TTCGGT": 1474, + "TTCGGC": 1475, + "TTCGGG": 1476, + "TTGAAA": 1477, + "TTGAAT": 1478, + "TTGAAC": 1479, + "TTGAAG": 1480, + "TTGATA": 1481, + "TTGATT": 1482, + "TTGATC": 1483, + "TTGATG": 1484, + "TTGACA": 1485, + "TTGACT": 1486, + "TTGACC": 1487, + "TTGACG": 1488, + "TTGAGA": 1489, + "TTGAGT": 1490, + "TTGAGC": 1491, + "TTGAGG": 1492, + "TTGTAA": 1493, + "TTGTAT": 1494, + "TTGTAC": 1495, + "TTGTAG": 1496, + "TTGTTA": 1497, + "TTGTTT": 1498, + "TTGTTC": 1499, + "TTGTTG": 1500, + "TTGTCA": 1501, + "TTGTCT": 1502, + "TTGTCC": 1503, + "TTGTCG": 1504, + "TTGTGA": 1505, + "TTGTGT": 1506, + "TTGTGC": 1507, + "TTGTGG": 1508, + "TTGCAA": 1509, + "TTGCAT": 1510, + "TTGCAC": 1511, + "TTGCAG": 1512, + "TTGCTA": 1513, + "TTGCTT": 1514, + "TTGCTC": 1515, + "TTGCTG": 1516, + "TTGCCA": 1517, + "TTGCCT": 1518, + "TTGCCC": 1519, + "TTGCCG": 1520, + "TTGCGA": 1521, + "TTGCGT": 1522, + "TTGCGC": 1523, + "TTGCGG": 1524, + "TTGGAA": 1525, + "TTGGAT": 1526, + "TTGGAC": 1527, + "TTGGAG": 1528, + "TTGGTA": 1529, + "TTGGTT": 1530, + "TTGGTC": 1531, + "TTGGTG": 1532, + "TTGGCA": 1533, + "TTGGCT": 1534, + "TTGGCC": 1535, + "TTGGCG": 1536, + "TTGGGA": 1537, + "TTGGGT": 1538, + "TTGGGC": 1539, + "TTGGGG": 1540, + "TCAAAA": 1541, + "TCAAAT": 1542, + "TCAAAC": 1543, + "TCAAAG": 1544, + "TCAATA": 1545, + "TCAATT": 1546, + "TCAATC": 1547, + "TCAATG": 1548, + "TCAACA": 1549, + "TCAACT": 1550, + "TCAACC": 1551, + "TCAACG": 1552, + "TCAAGA": 1553, + "TCAAGT": 1554, + "TCAAGC": 1555, + "TCAAGG": 1556, + "TCATAA": 1557, + "TCATAT": 1558, + "TCATAC": 1559, + "TCATAG": 1560, + "TCATTA": 1561, + "TCATTT": 1562, + "TCATTC": 1563, + "TCATTG": 1564, + "TCATCA": 1565, + "TCATCT": 1566, + "TCATCC": 1567, + "TCATCG": 1568, + "TCATGA": 1569, + "TCATGT": 1570, + "TCATGC": 1571, + "TCATGG": 1572, + "TCACAA": 1573, + "TCACAT": 1574, + "TCACAC": 1575, + "TCACAG": 1576, + "TCACTA": 1577, + "TCACTT": 1578, + "TCACTC": 1579, + "TCACTG": 1580, + "TCACCA": 1581, + "TCACCT": 1582, + "TCACCC": 1583, + "TCACCG": 1584, + "TCACGA": 1585, + "TCACGT": 1586, + "TCACGC": 1587, + "TCACGG": 1588, + "TCAGAA": 1589, + "TCAGAT": 1590, + "TCAGAC": 1591, + "TCAGAG": 1592, + "TCAGTA": 1593, + "TCAGTT": 1594, + "TCAGTC": 1595, + "TCAGTG": 1596, + "TCAGCA": 1597, + "TCAGCT": 1598, + "TCAGCC": 1599, + "TCAGCG": 1600, + "TCAGGA": 1601, + "TCAGGT": 1602, + "TCAGGC": 1603, + "TCAGGG": 1604, + "TCTAAA": 1605, + "TCTAAT": 1606, + "TCTAAC": 1607, + "TCTAAG": 1608, + "TCTATA": 1609, + "TCTATT": 1610, + "TCTATC": 1611, + "TCTATG": 1612, + "TCTACA": 1613, + "TCTACT": 1614, + "TCTACC": 1615, + "TCTACG": 1616, + "TCTAGA": 1617, + "TCTAGT": 1618, + "TCTAGC": 1619, + "TCTAGG": 1620, + "TCTTAA": 1621, + "TCTTAT": 1622, + "TCTTAC": 1623, + "TCTTAG": 1624, + "TCTTTA": 1625, + "TCTTTT": 1626, + "TCTTTC": 1627, + "TCTTTG": 1628, + "TCTTCA": 1629, + "TCTTCT": 1630, + "TCTTCC": 1631, + "TCTTCG": 1632, + "TCTTGA": 1633, + "TCTTGT": 1634, + "TCTTGC": 1635, + "TCTTGG": 1636, + "TCTCAA": 1637, + "TCTCAT": 1638, + "TCTCAC": 1639, + "TCTCAG": 1640, + "TCTCTA": 1641, + "TCTCTT": 1642, + "TCTCTC": 1643, + "TCTCTG": 1644, + "TCTCCA": 1645, + "TCTCCT": 1646, + "TCTCCC": 1647, + "TCTCCG": 1648, + "TCTCGA": 1649, + "TCTCGT": 1650, + "TCTCGC": 1651, + "TCTCGG": 1652, + "TCTGAA": 1653, + "TCTGAT": 1654, + "TCTGAC": 1655, + "TCTGAG": 1656, + "TCTGTA": 1657, + "TCTGTT": 1658, + "TCTGTC": 1659, + "TCTGTG": 1660, + "TCTGCA": 1661, + "TCTGCT": 1662, + "TCTGCC": 1663, + "TCTGCG": 1664, + "TCTGGA": 1665, + "TCTGGT": 1666, + "TCTGGC": 1667, + "TCTGGG": 1668, + "TCCAAA": 1669, + "TCCAAT": 1670, + "TCCAAC": 1671, + "TCCAAG": 1672, + "TCCATA": 1673, + "TCCATT": 1674, + "TCCATC": 1675, + "TCCATG": 1676, + "TCCACA": 1677, + "TCCACT": 1678, + "TCCACC": 1679, + "TCCACG": 1680, + "TCCAGA": 1681, + "TCCAGT": 1682, + "TCCAGC": 1683, + "TCCAGG": 1684, + "TCCTAA": 1685, + "TCCTAT": 1686, + "TCCTAC": 1687, + "TCCTAG": 1688, + "TCCTTA": 1689, + "TCCTTT": 1690, + "TCCTTC": 1691, + "TCCTTG": 1692, + "TCCTCA": 1693, + "TCCTCT": 1694, + "TCCTCC": 1695, + "TCCTCG": 1696, + "TCCTGA": 1697, + "TCCTGT": 1698, + "TCCTGC": 1699, + "TCCTGG": 1700, + "TCCCAA": 1701, + "TCCCAT": 1702, + "TCCCAC": 1703, + "TCCCAG": 1704, + "TCCCTA": 1705, + "TCCCTT": 1706, + "TCCCTC": 1707, + "TCCCTG": 1708, + "TCCCCA": 1709, + "TCCCCT": 1710, + "TCCCCC": 1711, + "TCCCCG": 1712, + "TCCCGA": 1713, + "TCCCGT": 1714, + "TCCCGC": 1715, + "TCCCGG": 1716, + "TCCGAA": 1717, + "TCCGAT": 1718, + "TCCGAC": 1719, + "TCCGAG": 1720, + "TCCGTA": 1721, + "TCCGTT": 1722, + "TCCGTC": 1723, + "TCCGTG": 1724, + "TCCGCA": 1725, + "TCCGCT": 1726, + "TCCGCC": 1727, + "TCCGCG": 1728, + "TCCGGA": 1729, + "TCCGGT": 1730, + "TCCGGC": 1731, + "TCCGGG": 1732, + "TCGAAA": 1733, + "TCGAAT": 1734, + "TCGAAC": 1735, + "TCGAAG": 1736, + "TCGATA": 1737, + "TCGATT": 1738, + "TCGATC": 1739, + "TCGATG": 1740, + "TCGACA": 1741, + "TCGACT": 1742, + "TCGACC": 1743, + "TCGACG": 1744, + "TCGAGA": 1745, + "TCGAGT": 1746, + "TCGAGC": 1747, + "TCGAGG": 1748, + "TCGTAA": 1749, + "TCGTAT": 1750, + "TCGTAC": 1751, + "TCGTAG": 1752, + "TCGTTA": 1753, + "TCGTTT": 1754, + "TCGTTC": 1755, + "TCGTTG": 1756, + "TCGTCA": 1757, + "TCGTCT": 1758, + "TCGTCC": 1759, + "TCGTCG": 1760, + "TCGTGA": 1761, + "TCGTGT": 1762, + "TCGTGC": 1763, + "TCGTGG": 1764, + "TCGCAA": 1765, + "TCGCAT": 1766, + "TCGCAC": 1767, + "TCGCAG": 1768, + "TCGCTA": 1769, + "TCGCTT": 1770, + "TCGCTC": 1771, + "TCGCTG": 1772, + "TCGCCA": 1773, + "TCGCCT": 1774, + "TCGCCC": 1775, + "TCGCCG": 1776, + "TCGCGA": 1777, + "TCGCGT": 1778, + "TCGCGC": 1779, + "TCGCGG": 1780, + "TCGGAA": 1781, + "TCGGAT": 1782, + "TCGGAC": 1783, + "TCGGAG": 1784, + "TCGGTA": 1785, + "TCGGTT": 1786, + "TCGGTC": 1787, + "TCGGTG": 1788, + "TCGGCA": 1789, + "TCGGCT": 1790, + "TCGGCC": 1791, + "TCGGCG": 1792, + "TCGGGA": 1793, + "TCGGGT": 1794, + "TCGGGC": 1795, + "TCGGGG": 1796, + "TGAAAA": 1797, + "TGAAAT": 1798, + "TGAAAC": 1799, + "TGAAAG": 1800, + "TGAATA": 1801, + "TGAATT": 1802, + "TGAATC": 1803, + "TGAATG": 1804, + "TGAACA": 1805, + "TGAACT": 1806, + "TGAACC": 1807, + "TGAACG": 1808, + "TGAAGA": 1809, + "TGAAGT": 1810, + "TGAAGC": 1811, + "TGAAGG": 1812, + "TGATAA": 1813, + "TGATAT": 1814, + "TGATAC": 1815, + "TGATAG": 1816, + "TGATTA": 1817, + "TGATTT": 1818, + "TGATTC": 1819, + "TGATTG": 1820, + "TGATCA": 1821, + "TGATCT": 1822, + "TGATCC": 1823, + "TGATCG": 1824, + "TGATGA": 1825, + "TGATGT": 1826, + "TGATGC": 1827, + "TGATGG": 1828, + "TGACAA": 1829, + "TGACAT": 1830, + "TGACAC": 1831, + "TGACAG": 1832, + "TGACTA": 1833, + "TGACTT": 1834, + "TGACTC": 1835, + "TGACTG": 1836, + "TGACCA": 1837, + "TGACCT": 1838, + "TGACCC": 1839, + "TGACCG": 1840, + "TGACGA": 1841, + "TGACGT": 1842, + "TGACGC": 1843, + "TGACGG": 1844, + "TGAGAA": 1845, + "TGAGAT": 1846, + "TGAGAC": 1847, + "TGAGAG": 1848, + "TGAGTA": 1849, + "TGAGTT": 1850, + "TGAGTC": 1851, + "TGAGTG": 1852, + "TGAGCA": 1853, + "TGAGCT": 1854, + "TGAGCC": 1855, + "TGAGCG": 1856, + "TGAGGA": 1857, + "TGAGGT": 1858, + "TGAGGC": 1859, + "TGAGGG": 1860, + "TGTAAA": 1861, + "TGTAAT": 1862, + "TGTAAC": 1863, + "TGTAAG": 1864, + "TGTATA": 1865, + "TGTATT": 1866, + "TGTATC": 1867, + "TGTATG": 1868, + "TGTACA": 1869, + "TGTACT": 1870, + "TGTACC": 1871, + "TGTACG": 1872, + "TGTAGA": 1873, + "TGTAGT": 1874, + "TGTAGC": 1875, + "TGTAGG": 1876, + "TGTTAA": 1877, + "TGTTAT": 1878, + "TGTTAC": 1879, + "TGTTAG": 1880, + "TGTTTA": 1881, + "TGTTTT": 1882, + "TGTTTC": 1883, + "TGTTTG": 1884, + "TGTTCA": 1885, + "TGTTCT": 1886, + "TGTTCC": 1887, + "TGTTCG": 1888, + "TGTTGA": 1889, + "TGTTGT": 1890, + "TGTTGC": 1891, + "TGTTGG": 1892, + "TGTCAA": 1893, + "TGTCAT": 1894, + "TGTCAC": 1895, + "TGTCAG": 1896, + "TGTCTA": 1897, + "TGTCTT": 1898, + "TGTCTC": 1899, + "TGTCTG": 1900, + "TGTCCA": 1901, + "TGTCCT": 1902, + "TGTCCC": 1903, + "TGTCCG": 1904, + "TGTCGA": 1905, + "TGTCGT": 1906, + "TGTCGC": 1907, + "TGTCGG": 1908, + "TGTGAA": 1909, + "TGTGAT": 1910, + "TGTGAC": 1911, + "TGTGAG": 1912, + "TGTGTA": 1913, + "TGTGTT": 1914, + "TGTGTC": 1915, + "TGTGTG": 1916, + "TGTGCA": 1917, + "TGTGCT": 1918, + "TGTGCC": 1919, + "TGTGCG": 1920, + "TGTGGA": 1921, + "TGTGGT": 1922, + "TGTGGC": 1923, + "TGTGGG": 1924, + "TGCAAA": 1925, + "TGCAAT": 1926, + "TGCAAC": 1927, + "TGCAAG": 1928, + "TGCATA": 1929, + "TGCATT": 1930, + "TGCATC": 1931, + "TGCATG": 1932, + "TGCACA": 1933, + "TGCACT": 1934, + "TGCACC": 1935, + "TGCACG": 1936, + "TGCAGA": 1937, + "TGCAGT": 1938, + "TGCAGC": 1939, + "TGCAGG": 1940, + "TGCTAA": 1941, + "TGCTAT": 1942, + "TGCTAC": 1943, + "TGCTAG": 1944, + "TGCTTA": 1945, + "TGCTTT": 1946, + "TGCTTC": 1947, + "TGCTTG": 1948, + "TGCTCA": 1949, + "TGCTCT": 1950, + "TGCTCC": 1951, + "TGCTCG": 1952, + "TGCTGA": 1953, + "TGCTGT": 1954, + "TGCTGC": 1955, + "TGCTGG": 1956, + "TGCCAA": 1957, + "TGCCAT": 1958, + "TGCCAC": 1959, + "TGCCAG": 1960, + "TGCCTA": 1961, + "TGCCTT": 1962, + "TGCCTC": 1963, + "TGCCTG": 1964, + "TGCCCA": 1965, + "TGCCCT": 1966, + "TGCCCC": 1967, + "TGCCCG": 1968, + "TGCCGA": 1969, + "TGCCGT": 1970, + "TGCCGC": 1971, + "TGCCGG": 1972, + "TGCGAA": 1973, + "TGCGAT": 1974, + "TGCGAC": 1975, + "TGCGAG": 1976, + "TGCGTA": 1977, + "TGCGTT": 1978, + "TGCGTC": 1979, + "TGCGTG": 1980, + "TGCGCA": 1981, + "TGCGCT": 1982, + "TGCGCC": 1983, + "TGCGCG": 1984, + "TGCGGA": 1985, + "TGCGGT": 1986, + "TGCGGC": 1987, + "TGCGGG": 1988, + "TGGAAA": 1989, + "TGGAAT": 1990, + "TGGAAC": 1991, + "TGGAAG": 1992, + "TGGATA": 1993, + "TGGATT": 1994, + "TGGATC": 1995, + "TGGATG": 1996, + "TGGACA": 1997, + "TGGACT": 1998, + "TGGACC": 1999, + "TGGACG": 2000, + "TGGAGA": 2001, + "TGGAGT": 2002, + "TGGAGC": 2003, + "TGGAGG": 2004, + "TGGTAA": 2005, + "TGGTAT": 2006, + "TGGTAC": 2007, + "TGGTAG": 2008, + "TGGTTA": 2009, + "TGGTTT": 2010, + "TGGTTC": 2011, + "TGGTTG": 2012, + "TGGTCA": 2013, + "TGGTCT": 2014, + "TGGTCC": 2015, + "TGGTCG": 2016, + "TGGTGA": 2017, + "TGGTGT": 2018, + "TGGTGC": 2019, + "TGGTGG": 2020, + "TGGCAA": 2021, + "TGGCAT": 2022, + "TGGCAC": 2023, + "TGGCAG": 2024, + "TGGCTA": 2025, + "TGGCTT": 2026, + "TGGCTC": 2027, + "TGGCTG": 2028, + "TGGCCA": 2029, + "TGGCCT": 2030, + "TGGCCC": 2031, + "TGGCCG": 2032, + "TGGCGA": 2033, + "TGGCGT": 2034, + "TGGCGC": 2035, + "TGGCGG": 2036, + "TGGGAA": 2037, + "TGGGAT": 2038, + "TGGGAC": 2039, + "TGGGAG": 2040, + "TGGGTA": 2041, + "TGGGTT": 2042, + "TGGGTC": 2043, + "TGGGTG": 2044, + "TGGGCA": 2045, + "TGGGCT": 2046, + "TGGGCC": 2047, + "TGGGCG": 2048, + "TGGGGA": 2049, + "TGGGGT": 2050, + "TGGGGC": 2051, + "TGGGGG": 2052, + "CAAAAA": 2053, + "CAAAAT": 2054, + "CAAAAC": 2055, + "CAAAAG": 2056, + "CAAATA": 2057, + "CAAATT": 2058, + "CAAATC": 2059, + "CAAATG": 2060, + "CAAACA": 2061, + "CAAACT": 2062, + "CAAACC": 2063, + "CAAACG": 2064, + "CAAAGA": 2065, + "CAAAGT": 2066, + "CAAAGC": 2067, + "CAAAGG": 2068, + "CAATAA": 2069, + "CAATAT": 2070, + "CAATAC": 2071, + "CAATAG": 2072, + "CAATTA": 2073, + "CAATTT": 2074, + "CAATTC": 2075, + "CAATTG": 2076, + "CAATCA": 2077, + "CAATCT": 2078, + "CAATCC": 2079, + "CAATCG": 2080, + "CAATGA": 2081, + "CAATGT": 2082, + "CAATGC": 2083, + "CAATGG": 2084, + "CAACAA": 2085, + "CAACAT": 2086, + "CAACAC": 2087, + "CAACAG": 2088, + "CAACTA": 2089, + "CAACTT": 2090, + "CAACTC": 2091, + "CAACTG": 2092, + "CAACCA": 2093, + "CAACCT": 2094, + "CAACCC": 2095, + "CAACCG": 2096, + "CAACGA": 2097, + "CAACGT": 2098, + "CAACGC": 2099, + "CAACGG": 2100, + "CAAGAA": 2101, + "CAAGAT": 2102, + "CAAGAC": 2103, + "CAAGAG": 2104, + "CAAGTA": 2105, + "CAAGTT": 2106, + "CAAGTC": 2107, + "CAAGTG": 2108, + "CAAGCA": 2109, + "CAAGCT": 2110, + "CAAGCC": 2111, + "CAAGCG": 2112, + "CAAGGA": 2113, + "CAAGGT": 2114, + "CAAGGC": 2115, + "CAAGGG": 2116, + "CATAAA": 2117, + "CATAAT": 2118, + "CATAAC": 2119, + "CATAAG": 2120, + "CATATA": 2121, + "CATATT": 2122, + "CATATC": 2123, + "CATATG": 2124, + "CATACA": 2125, + "CATACT": 2126, + "CATACC": 2127, + "CATACG": 2128, + "CATAGA": 2129, + "CATAGT": 2130, + "CATAGC": 2131, + "CATAGG": 2132, + "CATTAA": 2133, + "CATTAT": 2134, + "CATTAC": 2135, + "CATTAG": 2136, + "CATTTA": 2137, + "CATTTT": 2138, + "CATTTC": 2139, + "CATTTG": 2140, + "CATTCA": 2141, + "CATTCT": 2142, + "CATTCC": 2143, + "CATTCG": 2144, + "CATTGA": 2145, + "CATTGT": 2146, + "CATTGC": 2147, + "CATTGG": 2148, + "CATCAA": 2149, + "CATCAT": 2150, + "CATCAC": 2151, + "CATCAG": 2152, + "CATCTA": 2153, + "CATCTT": 2154, + "CATCTC": 2155, + "CATCTG": 2156, + "CATCCA": 2157, + "CATCCT": 2158, + "CATCCC": 2159, + "CATCCG": 2160, + "CATCGA": 2161, + "CATCGT": 2162, + "CATCGC": 2163, + "CATCGG": 2164, + "CATGAA": 2165, + "CATGAT": 2166, + "CATGAC": 2167, + "CATGAG": 2168, + "CATGTA": 2169, + "CATGTT": 2170, + "CATGTC": 2171, + "CATGTG": 2172, + "CATGCA": 2173, + "CATGCT": 2174, + "CATGCC": 2175, + "CATGCG": 2176, + "CATGGA": 2177, + "CATGGT": 2178, + "CATGGC": 2179, + "CATGGG": 2180, + "CACAAA": 2181, + "CACAAT": 2182, + "CACAAC": 2183, + "CACAAG": 2184, + "CACATA": 2185, + "CACATT": 2186, + "CACATC": 2187, + "CACATG": 2188, + "CACACA": 2189, + "CACACT": 2190, + "CACACC": 2191, + "CACACG": 2192, + "CACAGA": 2193, + "CACAGT": 2194, + "CACAGC": 2195, + "CACAGG": 2196, + "CACTAA": 2197, + "CACTAT": 2198, + "CACTAC": 2199, + "CACTAG": 2200, + "CACTTA": 2201, + "CACTTT": 2202, + "CACTTC": 2203, + "CACTTG": 2204, + "CACTCA": 2205, + "CACTCT": 2206, + "CACTCC": 2207, + "CACTCG": 2208, + "CACTGA": 2209, + "CACTGT": 2210, + "CACTGC": 2211, + "CACTGG": 2212, + "CACCAA": 2213, + "CACCAT": 2214, + "CACCAC": 2215, + "CACCAG": 2216, + "CACCTA": 2217, + "CACCTT": 2218, + "CACCTC": 2219, + "CACCTG": 2220, + "CACCCA": 2221, + "CACCCT": 2222, + "CACCCC": 2223, + "CACCCG": 2224, + "CACCGA": 2225, + "CACCGT": 2226, + "CACCGC": 2227, + "CACCGG": 2228, + "CACGAA": 2229, + "CACGAT": 2230, + "CACGAC": 2231, + "CACGAG": 2232, + "CACGTA": 2233, + "CACGTT": 2234, + "CACGTC": 2235, + "CACGTG": 2236, + "CACGCA": 2237, + "CACGCT": 2238, + "CACGCC": 2239, + "CACGCG": 2240, + "CACGGA": 2241, + "CACGGT": 2242, + "CACGGC": 2243, + "CACGGG": 2244, + "CAGAAA": 2245, + "CAGAAT": 2246, + "CAGAAC": 2247, + "CAGAAG": 2248, + "CAGATA": 2249, + "CAGATT": 2250, + "CAGATC": 2251, + "CAGATG": 2252, + "CAGACA": 2253, + "CAGACT": 2254, + "CAGACC": 2255, + "CAGACG": 2256, + "CAGAGA": 2257, + "CAGAGT": 2258, + "CAGAGC": 2259, + "CAGAGG": 2260, + "CAGTAA": 2261, + "CAGTAT": 2262, + "CAGTAC": 2263, + "CAGTAG": 2264, + "CAGTTA": 2265, + "CAGTTT": 2266, + "CAGTTC": 2267, + "CAGTTG": 2268, + "CAGTCA": 2269, + "CAGTCT": 2270, + "CAGTCC": 2271, + "CAGTCG": 2272, + "CAGTGA": 2273, + "CAGTGT": 2274, + "CAGTGC": 2275, + "CAGTGG": 2276, + "CAGCAA": 2277, + "CAGCAT": 2278, + "CAGCAC": 2279, + "CAGCAG": 2280, + "CAGCTA": 2281, + "CAGCTT": 2282, + "CAGCTC": 2283, + "CAGCTG": 2284, + "CAGCCA": 2285, + "CAGCCT": 2286, + "CAGCCC": 2287, + "CAGCCG": 2288, + "CAGCGA": 2289, + "CAGCGT": 2290, + "CAGCGC": 2291, + "CAGCGG": 2292, + "CAGGAA": 2293, + "CAGGAT": 2294, + "CAGGAC": 2295, + "CAGGAG": 2296, + "CAGGTA": 2297, + "CAGGTT": 2298, + "CAGGTC": 2299, + "CAGGTG": 2300, + "CAGGCA": 2301, + "CAGGCT": 2302, + "CAGGCC": 2303, + "CAGGCG": 2304, + "CAGGGA": 2305, + "CAGGGT": 2306, + "CAGGGC": 2307, + "CAGGGG": 2308, + "CTAAAA": 2309, + "CTAAAT": 2310, + "CTAAAC": 2311, + "CTAAAG": 2312, + "CTAATA": 2313, + "CTAATT": 2314, + "CTAATC": 2315, + "CTAATG": 2316, + "CTAACA": 2317, + "CTAACT": 2318, + "CTAACC": 2319, + "CTAACG": 2320, + "CTAAGA": 2321, + "CTAAGT": 2322, + "CTAAGC": 2323, + "CTAAGG": 2324, + "CTATAA": 2325, + "CTATAT": 2326, + "CTATAC": 2327, + "CTATAG": 2328, + "CTATTA": 2329, + "CTATTT": 2330, + "CTATTC": 2331, + "CTATTG": 2332, + "CTATCA": 2333, + "CTATCT": 2334, + "CTATCC": 2335, + "CTATCG": 2336, + "CTATGA": 2337, + "CTATGT": 2338, + "CTATGC": 2339, + "CTATGG": 2340, + "CTACAA": 2341, + "CTACAT": 2342, + "CTACAC": 2343, + "CTACAG": 2344, + "CTACTA": 2345, + "CTACTT": 2346, + "CTACTC": 2347, + "CTACTG": 2348, + "CTACCA": 2349, + "CTACCT": 2350, + "CTACCC": 2351, + "CTACCG": 2352, + "CTACGA": 2353, + "CTACGT": 2354, + "CTACGC": 2355, + "CTACGG": 2356, + "CTAGAA": 2357, + "CTAGAT": 2358, + "CTAGAC": 2359, + "CTAGAG": 2360, + "CTAGTA": 2361, + "CTAGTT": 2362, + "CTAGTC": 2363, + "CTAGTG": 2364, + "CTAGCA": 2365, + "CTAGCT": 2366, + "CTAGCC": 2367, + "CTAGCG": 2368, + "CTAGGA": 2369, + "CTAGGT": 2370, + "CTAGGC": 2371, + "CTAGGG": 2372, + "CTTAAA": 2373, + "CTTAAT": 2374, + "CTTAAC": 2375, + "CTTAAG": 2376, + "CTTATA": 2377, + "CTTATT": 2378, + "CTTATC": 2379, + "CTTATG": 2380, + "CTTACA": 2381, + "CTTACT": 2382, + "CTTACC": 2383, + "CTTACG": 2384, + "CTTAGA": 2385, + "CTTAGT": 2386, + "CTTAGC": 2387, + "CTTAGG": 2388, + "CTTTAA": 2389, + "CTTTAT": 2390, + "CTTTAC": 2391, + "CTTTAG": 2392, + "CTTTTA": 2393, + "CTTTTT": 2394, + "CTTTTC": 2395, + "CTTTTG": 2396, + "CTTTCA": 2397, + "CTTTCT": 2398, + "CTTTCC": 2399, + "CTTTCG": 2400, + "CTTTGA": 2401, + "CTTTGT": 2402, + "CTTTGC": 2403, + "CTTTGG": 2404, + "CTTCAA": 2405, + "CTTCAT": 2406, + "CTTCAC": 2407, + "CTTCAG": 2408, + "CTTCTA": 2409, + "CTTCTT": 2410, + "CTTCTC": 2411, + "CTTCTG": 2412, + "CTTCCA": 2413, + "CTTCCT": 2414, + "CTTCCC": 2415, + "CTTCCG": 2416, + "CTTCGA": 2417, + "CTTCGT": 2418, + "CTTCGC": 2419, + "CTTCGG": 2420, + "CTTGAA": 2421, + "CTTGAT": 2422, + "CTTGAC": 2423, + "CTTGAG": 2424, + "CTTGTA": 2425, + "CTTGTT": 2426, + "CTTGTC": 2427, + "CTTGTG": 2428, + "CTTGCA": 2429, + "CTTGCT": 2430, + "CTTGCC": 2431, + "CTTGCG": 2432, + "CTTGGA": 2433, + "CTTGGT": 2434, + "CTTGGC": 2435, + "CTTGGG": 2436, + "CTCAAA": 2437, + "CTCAAT": 2438, + "CTCAAC": 2439, + "CTCAAG": 2440, + "CTCATA": 2441, + "CTCATT": 2442, + "CTCATC": 2443, + "CTCATG": 2444, + "CTCACA": 2445, + "CTCACT": 2446, + "CTCACC": 2447, + "CTCACG": 2448, + "CTCAGA": 2449, + "CTCAGT": 2450, + "CTCAGC": 2451, + "CTCAGG": 2452, + "CTCTAA": 2453, + "CTCTAT": 2454, + "CTCTAC": 2455, + "CTCTAG": 2456, + "CTCTTA": 2457, + "CTCTTT": 2458, + "CTCTTC": 2459, + "CTCTTG": 2460, + "CTCTCA": 2461, + "CTCTCT": 2462, + "CTCTCC": 2463, + "CTCTCG": 2464, + "CTCTGA": 2465, + "CTCTGT": 2466, + "CTCTGC": 2467, + "CTCTGG": 2468, + "CTCCAA": 2469, + "CTCCAT": 2470, + "CTCCAC": 2471, + "CTCCAG": 2472, + "CTCCTA": 2473, + "CTCCTT": 2474, + "CTCCTC": 2475, + "CTCCTG": 2476, + "CTCCCA": 2477, + "CTCCCT": 2478, + "CTCCCC": 2479, + "CTCCCG": 2480, + "CTCCGA": 2481, + "CTCCGT": 2482, + "CTCCGC": 2483, + "CTCCGG": 2484, + "CTCGAA": 2485, + "CTCGAT": 2486, + "CTCGAC": 2487, + "CTCGAG": 2488, + "CTCGTA": 2489, + "CTCGTT": 2490, + "CTCGTC": 2491, + "CTCGTG": 2492, + "CTCGCA": 2493, + "CTCGCT": 2494, + "CTCGCC": 2495, + "CTCGCG": 2496, + "CTCGGA": 2497, + "CTCGGT": 2498, + "CTCGGC": 2499, + "CTCGGG": 2500, + "CTGAAA": 2501, + "CTGAAT": 2502, + "CTGAAC": 2503, + "CTGAAG": 2504, + "CTGATA": 2505, + "CTGATT": 2506, + "CTGATC": 2507, + "CTGATG": 2508, + "CTGACA": 2509, + "CTGACT": 2510, + "CTGACC": 2511, + "CTGACG": 2512, + "CTGAGA": 2513, + "CTGAGT": 2514, + "CTGAGC": 2515, + "CTGAGG": 2516, + "CTGTAA": 2517, + "CTGTAT": 2518, + "CTGTAC": 2519, + "CTGTAG": 2520, + "CTGTTA": 2521, + "CTGTTT": 2522, + "CTGTTC": 2523, + "CTGTTG": 2524, + "CTGTCA": 2525, + "CTGTCT": 2526, + "CTGTCC": 2527, + "CTGTCG": 2528, + "CTGTGA": 2529, + "CTGTGT": 2530, + "CTGTGC": 2531, + "CTGTGG": 2532, + "CTGCAA": 2533, + "CTGCAT": 2534, + "CTGCAC": 2535, + "CTGCAG": 2536, + "CTGCTA": 2537, + "CTGCTT": 2538, + "CTGCTC": 2539, + "CTGCTG": 2540, + "CTGCCA": 2541, + "CTGCCT": 2542, + "CTGCCC": 2543, + "CTGCCG": 2544, + "CTGCGA": 2545, + "CTGCGT": 2546, + "CTGCGC": 2547, + "CTGCGG": 2548, + "CTGGAA": 2549, + "CTGGAT": 2550, + "CTGGAC": 2551, + "CTGGAG": 2552, + "CTGGTA": 2553, + "CTGGTT": 2554, + "CTGGTC": 2555, + "CTGGTG": 2556, + "CTGGCA": 2557, + "CTGGCT": 2558, + "CTGGCC": 2559, + "CTGGCG": 2560, + "CTGGGA": 2561, + "CTGGGT": 2562, + "CTGGGC": 2563, + "CTGGGG": 2564, + "CCAAAA": 2565, + "CCAAAT": 2566, + "CCAAAC": 2567, + "CCAAAG": 2568, + "CCAATA": 2569, + "CCAATT": 2570, + "CCAATC": 2571, + "CCAATG": 2572, + "CCAACA": 2573, + "CCAACT": 2574, + "CCAACC": 2575, + "CCAACG": 2576, + "CCAAGA": 2577, + "CCAAGT": 2578, + "CCAAGC": 2579, + "CCAAGG": 2580, + "CCATAA": 2581, + "CCATAT": 2582, + "CCATAC": 2583, + "CCATAG": 2584, + "CCATTA": 2585, + "CCATTT": 2586, + "CCATTC": 2587, + "CCATTG": 2588, + "CCATCA": 2589, + "CCATCT": 2590, + "CCATCC": 2591, + "CCATCG": 2592, + "CCATGA": 2593, + "CCATGT": 2594, + "CCATGC": 2595, + "CCATGG": 2596, + "CCACAA": 2597, + "CCACAT": 2598, + "CCACAC": 2599, + "CCACAG": 2600, + "CCACTA": 2601, + "CCACTT": 2602, + "CCACTC": 2603, + "CCACTG": 2604, + "CCACCA": 2605, + "CCACCT": 2606, + "CCACCC": 2607, + "CCACCG": 2608, + "CCACGA": 2609, + "CCACGT": 2610, + "CCACGC": 2611, + "CCACGG": 2612, + "CCAGAA": 2613, + "CCAGAT": 2614, + "CCAGAC": 2615, + "CCAGAG": 2616, + "CCAGTA": 2617, + "CCAGTT": 2618, + "CCAGTC": 2619, + "CCAGTG": 2620, + "CCAGCA": 2621, + "CCAGCT": 2622, + "CCAGCC": 2623, + "CCAGCG": 2624, + "CCAGGA": 2625, + "CCAGGT": 2626, + "CCAGGC": 2627, + "CCAGGG": 2628, + "CCTAAA": 2629, + "CCTAAT": 2630, + "CCTAAC": 2631, + "CCTAAG": 2632, + "CCTATA": 2633, + "CCTATT": 2634, + "CCTATC": 2635, + "CCTATG": 2636, + "CCTACA": 2637, + "CCTACT": 2638, + "CCTACC": 2639, + "CCTACG": 2640, + "CCTAGA": 2641, + "CCTAGT": 2642, + "CCTAGC": 2643, + "CCTAGG": 2644, + "CCTTAA": 2645, + "CCTTAT": 2646, + "CCTTAC": 2647, + "CCTTAG": 2648, + "CCTTTA": 2649, + "CCTTTT": 2650, + "CCTTTC": 2651, + "CCTTTG": 2652, + "CCTTCA": 2653, + "CCTTCT": 2654, + "CCTTCC": 2655, + "CCTTCG": 2656, + "CCTTGA": 2657, + "CCTTGT": 2658, + "CCTTGC": 2659, + "CCTTGG": 2660, + "CCTCAA": 2661, + "CCTCAT": 2662, + "CCTCAC": 2663, + "CCTCAG": 2664, + "CCTCTA": 2665, + "CCTCTT": 2666, + "CCTCTC": 2667, + "CCTCTG": 2668, + "CCTCCA": 2669, + "CCTCCT": 2670, + "CCTCCC": 2671, + "CCTCCG": 2672, + "CCTCGA": 2673, + "CCTCGT": 2674, + "CCTCGC": 2675, + "CCTCGG": 2676, + "CCTGAA": 2677, + "CCTGAT": 2678, + "CCTGAC": 2679, + "CCTGAG": 2680, + "CCTGTA": 2681, + "CCTGTT": 2682, + "CCTGTC": 2683, + "CCTGTG": 2684, + "CCTGCA": 2685, + "CCTGCT": 2686, + "CCTGCC": 2687, + "CCTGCG": 2688, + "CCTGGA": 2689, + "CCTGGT": 2690, + "CCTGGC": 2691, + "CCTGGG": 2692, + "CCCAAA": 2693, + "CCCAAT": 2694, + "CCCAAC": 2695, + "CCCAAG": 2696, + "CCCATA": 2697, + "CCCATT": 2698, + "CCCATC": 2699, + "CCCATG": 2700, + "CCCACA": 2701, + "CCCACT": 2702, + "CCCACC": 2703, + "CCCACG": 2704, + "CCCAGA": 2705, + "CCCAGT": 2706, + "CCCAGC": 2707, + "CCCAGG": 2708, + "CCCTAA": 2709, + "CCCTAT": 2710, + "CCCTAC": 2711, + "CCCTAG": 2712, + "CCCTTA": 2713, + "CCCTTT": 2714, + "CCCTTC": 2715, + "CCCTTG": 2716, + "CCCTCA": 2717, + "CCCTCT": 2718, + "CCCTCC": 2719, + "CCCTCG": 2720, + "CCCTGA": 2721, + "CCCTGT": 2722, + "CCCTGC": 2723, + "CCCTGG": 2724, + "CCCCAA": 2725, + "CCCCAT": 2726, + "CCCCAC": 2727, + "CCCCAG": 2728, + "CCCCTA": 2729, + "CCCCTT": 2730, + "CCCCTC": 2731, + "CCCCTG": 2732, + "CCCCCA": 2733, + "CCCCCT": 2734, + "CCCCCC": 2735, + "CCCCCG": 2736, + "CCCCGA": 2737, + "CCCCGT": 2738, + "CCCCGC": 2739, + "CCCCGG": 2740, + "CCCGAA": 2741, + "CCCGAT": 2742, + "CCCGAC": 2743, + "CCCGAG": 2744, + "CCCGTA": 2745, + "CCCGTT": 2746, + "CCCGTC": 2747, + "CCCGTG": 2748, + "CCCGCA": 2749, + "CCCGCT": 2750, + "CCCGCC": 2751, + "CCCGCG": 2752, + "CCCGGA": 2753, + "CCCGGT": 2754, + "CCCGGC": 2755, + "CCCGGG": 2756, + "CCGAAA": 2757, + "CCGAAT": 2758, + "CCGAAC": 2759, + "CCGAAG": 2760, + "CCGATA": 2761, + "CCGATT": 2762, + "CCGATC": 2763, + "CCGATG": 2764, + "CCGACA": 2765, + "CCGACT": 2766, + "CCGACC": 2767, + "CCGACG": 2768, + "CCGAGA": 2769, + "CCGAGT": 2770, + "CCGAGC": 2771, + "CCGAGG": 2772, + "CCGTAA": 2773, + "CCGTAT": 2774, + "CCGTAC": 2775, + "CCGTAG": 2776, + "CCGTTA": 2777, + "CCGTTT": 2778, + "CCGTTC": 2779, + "CCGTTG": 2780, + "CCGTCA": 2781, + "CCGTCT": 2782, + "CCGTCC": 2783, + "CCGTCG": 2784, + "CCGTGA": 2785, + "CCGTGT": 2786, + "CCGTGC": 2787, + "CCGTGG": 2788, + "CCGCAA": 2789, + "CCGCAT": 2790, + "CCGCAC": 2791, + "CCGCAG": 2792, + "CCGCTA": 2793, + "CCGCTT": 2794, + "CCGCTC": 2795, + "CCGCTG": 2796, + "CCGCCA": 2797, + "CCGCCT": 2798, + "CCGCCC": 2799, + "CCGCCG": 2800, + "CCGCGA": 2801, + "CCGCGT": 2802, + "CCGCGC": 2803, + "CCGCGG": 2804, + "CCGGAA": 2805, + "CCGGAT": 2806, + "CCGGAC": 2807, + "CCGGAG": 2808, + "CCGGTA": 2809, + "CCGGTT": 2810, + "CCGGTC": 2811, + "CCGGTG": 2812, + "CCGGCA": 2813, + "CCGGCT": 2814, + "CCGGCC": 2815, + "CCGGCG": 2816, + "CCGGGA": 2817, + "CCGGGT": 2818, + "CCGGGC": 2819, + "CCGGGG": 2820, + "CGAAAA": 2821, + "CGAAAT": 2822, + "CGAAAC": 2823, + "CGAAAG": 2824, + "CGAATA": 2825, + "CGAATT": 2826, + "CGAATC": 2827, + "CGAATG": 2828, + "CGAACA": 2829, + "CGAACT": 2830, + "CGAACC": 2831, + "CGAACG": 2832, + "CGAAGA": 2833, + "CGAAGT": 2834, + "CGAAGC": 2835, + "CGAAGG": 2836, + "CGATAA": 2837, + "CGATAT": 2838, + "CGATAC": 2839, + "CGATAG": 2840, + "CGATTA": 2841, + "CGATTT": 2842, + "CGATTC": 2843, + "CGATTG": 2844, + "CGATCA": 2845, + "CGATCT": 2846, + "CGATCC": 2847, + "CGATCG": 2848, + "CGATGA": 2849, + "CGATGT": 2850, + "CGATGC": 2851, + "CGATGG": 2852, + "CGACAA": 2853, + "CGACAT": 2854, + "CGACAC": 2855, + "CGACAG": 2856, + "CGACTA": 2857, + "CGACTT": 2858, + "CGACTC": 2859, + "CGACTG": 2860, + "CGACCA": 2861, + "CGACCT": 2862, + "CGACCC": 2863, + "CGACCG": 2864, + "CGACGA": 2865, + "CGACGT": 2866, + "CGACGC": 2867, + "CGACGG": 2868, + "CGAGAA": 2869, + "CGAGAT": 2870, + "CGAGAC": 2871, + "CGAGAG": 2872, + "CGAGTA": 2873, + "CGAGTT": 2874, + "CGAGTC": 2875, + "CGAGTG": 2876, + "CGAGCA": 2877, + "CGAGCT": 2878, + "CGAGCC": 2879, + "CGAGCG": 2880, + "CGAGGA": 2881, + "CGAGGT": 2882, + "CGAGGC": 2883, + "CGAGGG": 2884, + "CGTAAA": 2885, + "CGTAAT": 2886, + "CGTAAC": 2887, + "CGTAAG": 2888, + "CGTATA": 2889, + "CGTATT": 2890, + "CGTATC": 2891, + "CGTATG": 2892, + "CGTACA": 2893, + "CGTACT": 2894, + "CGTACC": 2895, + "CGTACG": 2896, + "CGTAGA": 2897, + "CGTAGT": 2898, + "CGTAGC": 2899, + "CGTAGG": 2900, + "CGTTAA": 2901, + "CGTTAT": 2902, + "CGTTAC": 2903, + "CGTTAG": 2904, + "CGTTTA": 2905, + "CGTTTT": 2906, + "CGTTTC": 2907, + "CGTTTG": 2908, + "CGTTCA": 2909, + "CGTTCT": 2910, + "CGTTCC": 2911, + "CGTTCG": 2912, + "CGTTGA": 2913, + "CGTTGT": 2914, + "CGTTGC": 2915, + "CGTTGG": 2916, + "CGTCAA": 2917, + "CGTCAT": 2918, + "CGTCAC": 2919, + "CGTCAG": 2920, + "CGTCTA": 2921, + "CGTCTT": 2922, + "CGTCTC": 2923, + "CGTCTG": 2924, + "CGTCCA": 2925, + "CGTCCT": 2926, + "CGTCCC": 2927, + "CGTCCG": 2928, + "CGTCGA": 2929, + "CGTCGT": 2930, + "CGTCGC": 2931, + "CGTCGG": 2932, + "CGTGAA": 2933, + "CGTGAT": 2934, + "CGTGAC": 2935, + "CGTGAG": 2936, + "CGTGTA": 2937, + "CGTGTT": 2938, + "CGTGTC": 2939, + "CGTGTG": 2940, + "CGTGCA": 2941, + "CGTGCT": 2942, + "CGTGCC": 2943, + "CGTGCG": 2944, + "CGTGGA": 2945, + "CGTGGT": 2946, + "CGTGGC": 2947, + "CGTGGG": 2948, + "CGCAAA": 2949, + "CGCAAT": 2950, + "CGCAAC": 2951, + "CGCAAG": 2952, + "CGCATA": 2953, + "CGCATT": 2954, + "CGCATC": 2955, + "CGCATG": 2956, + "CGCACA": 2957, + "CGCACT": 2958, + "CGCACC": 2959, + "CGCACG": 2960, + "CGCAGA": 2961, + "CGCAGT": 2962, + "CGCAGC": 2963, + "CGCAGG": 2964, + "CGCTAA": 2965, + "CGCTAT": 2966, + "CGCTAC": 2967, + "CGCTAG": 2968, + "CGCTTA": 2969, + "CGCTTT": 2970, + "CGCTTC": 2971, + "CGCTTG": 2972, + "CGCTCA": 2973, + "CGCTCT": 2974, + "CGCTCC": 2975, + "CGCTCG": 2976, + "CGCTGA": 2977, + "CGCTGT": 2978, + "CGCTGC": 2979, + "CGCTGG": 2980, + "CGCCAA": 2981, + "CGCCAT": 2982, + "CGCCAC": 2983, + "CGCCAG": 2984, + "CGCCTA": 2985, + "CGCCTT": 2986, + "CGCCTC": 2987, + "CGCCTG": 2988, + "CGCCCA": 2989, + "CGCCCT": 2990, + "CGCCCC": 2991, + "CGCCCG": 2992, + "CGCCGA": 2993, + "CGCCGT": 2994, + "CGCCGC": 2995, + "CGCCGG": 2996, + "CGCGAA": 2997, + "CGCGAT": 2998, + "CGCGAC": 2999, + "CGCGAG": 3000, + "CGCGTA": 3001, + "CGCGTT": 3002, + "CGCGTC": 3003, + "CGCGTG": 3004, + "CGCGCA": 3005, + "CGCGCT": 3006, + "CGCGCC": 3007, + "CGCGCG": 3008, + "CGCGGA": 3009, + "CGCGGT": 3010, + "CGCGGC": 3011, + "CGCGGG": 3012, + "CGGAAA": 3013, + "CGGAAT": 3014, + "CGGAAC": 3015, + "CGGAAG": 3016, + "CGGATA": 3017, + "CGGATT": 3018, + "CGGATC": 3019, + "CGGATG": 3020, + "CGGACA": 3021, + "CGGACT": 3022, + "CGGACC": 3023, + "CGGACG": 3024, + "CGGAGA": 3025, + "CGGAGT": 3026, + "CGGAGC": 3027, + "CGGAGG": 3028, + "CGGTAA": 3029, + "CGGTAT": 3030, + "CGGTAC": 3031, + "CGGTAG": 3032, + "CGGTTA": 3033, + "CGGTTT": 3034, + "CGGTTC": 3035, + "CGGTTG": 3036, + "CGGTCA": 3037, + "CGGTCT": 3038, + "CGGTCC": 3039, + "CGGTCG": 3040, + "CGGTGA": 3041, + "CGGTGT": 3042, + "CGGTGC": 3043, + "CGGTGG": 3044, + "CGGCAA": 3045, + "CGGCAT": 3046, + "CGGCAC": 3047, + "CGGCAG": 3048, + "CGGCTA": 3049, + "CGGCTT": 3050, + "CGGCTC": 3051, + "CGGCTG": 3052, + "CGGCCA": 3053, + "CGGCCT": 3054, + "CGGCCC": 3055, + "CGGCCG": 3056, + "CGGCGA": 3057, + "CGGCGT": 3058, + "CGGCGC": 3059, + "CGGCGG": 3060, + "CGGGAA": 3061, + "CGGGAT": 3062, + "CGGGAC": 3063, + "CGGGAG": 3064, + "CGGGTA": 3065, + "CGGGTT": 3066, + "CGGGTC": 3067, + "CGGGTG": 3068, + "CGGGCA": 3069, + "CGGGCT": 3070, + "CGGGCC": 3071, + "CGGGCG": 3072, + "CGGGGA": 3073, + "CGGGGT": 3074, + "CGGGGC": 3075, + "CGGGGG": 3076, + "GAAAAA": 3077, + "GAAAAT": 3078, + "GAAAAC": 3079, + "GAAAAG": 3080, + "GAAATA": 3081, + "GAAATT": 3082, + "GAAATC": 3083, + "GAAATG": 3084, + "GAAACA": 3085, + "GAAACT": 3086, + "GAAACC": 3087, + "GAAACG": 3088, + "GAAAGA": 3089, + "GAAAGT": 3090, + "GAAAGC": 3091, + "GAAAGG": 3092, + "GAATAA": 3093, + "GAATAT": 3094, + "GAATAC": 3095, + "GAATAG": 3096, + "GAATTA": 3097, + "GAATTT": 3098, + "GAATTC": 3099, + "GAATTG": 3100, + "GAATCA": 3101, + "GAATCT": 3102, + "GAATCC": 3103, + "GAATCG": 3104, + "GAATGA": 3105, + "GAATGT": 3106, + "GAATGC": 3107, + "GAATGG": 3108, + "GAACAA": 3109, + "GAACAT": 3110, + "GAACAC": 3111, + "GAACAG": 3112, + "GAACTA": 3113, + "GAACTT": 3114, + "GAACTC": 3115, + "GAACTG": 3116, + "GAACCA": 3117, + "GAACCT": 3118, + "GAACCC": 3119, + "GAACCG": 3120, + "GAACGA": 3121, + "GAACGT": 3122, + "GAACGC": 3123, + "GAACGG": 3124, + "GAAGAA": 3125, + "GAAGAT": 3126, + "GAAGAC": 3127, + "GAAGAG": 3128, + "GAAGTA": 3129, + "GAAGTT": 3130, + "GAAGTC": 3131, + "GAAGTG": 3132, + "GAAGCA": 3133, + "GAAGCT": 3134, + "GAAGCC": 3135, + "GAAGCG": 3136, + "GAAGGA": 3137, + "GAAGGT": 3138, + "GAAGGC": 3139, + "GAAGGG": 3140, + "GATAAA": 3141, + "GATAAT": 3142, + "GATAAC": 3143, + "GATAAG": 3144, + "GATATA": 3145, + "GATATT": 3146, + "GATATC": 3147, + "GATATG": 3148, + "GATACA": 3149, + "GATACT": 3150, + "GATACC": 3151, + "GATACG": 3152, + "GATAGA": 3153, + "GATAGT": 3154, + "GATAGC": 3155, + "GATAGG": 3156, + "GATTAA": 3157, + "GATTAT": 3158, + "GATTAC": 3159, + "GATTAG": 3160, + "GATTTA": 3161, + "GATTTT": 3162, + "GATTTC": 3163, + "GATTTG": 3164, + "GATTCA": 3165, + "GATTCT": 3166, + "GATTCC": 3167, + "GATTCG": 3168, + "GATTGA": 3169, + "GATTGT": 3170, + "GATTGC": 3171, + "GATTGG": 3172, + "GATCAA": 3173, + "GATCAT": 3174, + "GATCAC": 3175, + "GATCAG": 3176, + "GATCTA": 3177, + "GATCTT": 3178, + "GATCTC": 3179, + "GATCTG": 3180, + "GATCCA": 3181, + "GATCCT": 3182, + "GATCCC": 3183, + "GATCCG": 3184, + "GATCGA": 3185, + "GATCGT": 3186, + "GATCGC": 3187, + "GATCGG": 3188, + "GATGAA": 3189, + "GATGAT": 3190, + "GATGAC": 3191, + "GATGAG": 3192, + "GATGTA": 3193, + "GATGTT": 3194, + "GATGTC": 3195, + "GATGTG": 3196, + "GATGCA": 3197, + "GATGCT": 3198, + "GATGCC": 3199, + "GATGCG": 3200, + "GATGGA": 3201, + "GATGGT": 3202, + "GATGGC": 3203, + "GATGGG": 3204, + "GACAAA": 3205, + "GACAAT": 3206, + "GACAAC": 3207, + "GACAAG": 3208, + "GACATA": 3209, + "GACATT": 3210, + "GACATC": 3211, + "GACATG": 3212, + "GACACA": 3213, + "GACACT": 3214, + "GACACC": 3215, + "GACACG": 3216, + "GACAGA": 3217, + "GACAGT": 3218, + "GACAGC": 3219, + "GACAGG": 3220, + "GACTAA": 3221, + "GACTAT": 3222, + "GACTAC": 3223, + "GACTAG": 3224, + "GACTTA": 3225, + "GACTTT": 3226, + "GACTTC": 3227, + "GACTTG": 3228, + "GACTCA": 3229, + "GACTCT": 3230, + "GACTCC": 3231, + "GACTCG": 3232, + "GACTGA": 3233, + "GACTGT": 3234, + "GACTGC": 3235, + "GACTGG": 3236, + "GACCAA": 3237, + "GACCAT": 3238, + "GACCAC": 3239, + "GACCAG": 3240, + "GACCTA": 3241, + "GACCTT": 3242, + "GACCTC": 3243, + "GACCTG": 3244, + "GACCCA": 3245, + "GACCCT": 3246, + "GACCCC": 3247, + "GACCCG": 3248, + "GACCGA": 3249, + "GACCGT": 3250, + "GACCGC": 3251, + "GACCGG": 3252, + "GACGAA": 3253, + "GACGAT": 3254, + "GACGAC": 3255, + "GACGAG": 3256, + "GACGTA": 3257, + "GACGTT": 3258, + "GACGTC": 3259, + "GACGTG": 3260, + "GACGCA": 3261, + "GACGCT": 3262, + "GACGCC": 3263, + "GACGCG": 3264, + "GACGGA": 3265, + "GACGGT": 3266, + "GACGGC": 3267, + "GACGGG": 3268, + "GAGAAA": 3269, + "GAGAAT": 3270, + "GAGAAC": 3271, + "GAGAAG": 3272, + "GAGATA": 3273, + "GAGATT": 3274, + "GAGATC": 3275, + "GAGATG": 3276, + "GAGACA": 3277, + "GAGACT": 3278, + "GAGACC": 3279, + "GAGACG": 3280, + "GAGAGA": 3281, + "GAGAGT": 3282, + "GAGAGC": 3283, + "GAGAGG": 3284, + "GAGTAA": 3285, + "GAGTAT": 3286, + "GAGTAC": 3287, + "GAGTAG": 3288, + "GAGTTA": 3289, + "GAGTTT": 3290, + "GAGTTC": 3291, + "GAGTTG": 3292, + "GAGTCA": 3293, + "GAGTCT": 3294, + "GAGTCC": 3295, + "GAGTCG": 3296, + "GAGTGA": 3297, + "GAGTGT": 3298, + "GAGTGC": 3299, + "GAGTGG": 3300, + "GAGCAA": 3301, + "GAGCAT": 3302, + "GAGCAC": 3303, + "GAGCAG": 3304, + "GAGCTA": 3305, + "GAGCTT": 3306, + "GAGCTC": 3307, + "GAGCTG": 3308, + "GAGCCA": 3309, + "GAGCCT": 3310, + "GAGCCC": 3311, + "GAGCCG": 3312, + "GAGCGA": 3313, + "GAGCGT": 3314, + "GAGCGC": 3315, + "GAGCGG": 3316, + "GAGGAA": 3317, + "GAGGAT": 3318, + "GAGGAC": 3319, + "GAGGAG": 3320, + "GAGGTA": 3321, + "GAGGTT": 3322, + "GAGGTC": 3323, + "GAGGTG": 3324, + "GAGGCA": 3325, + "GAGGCT": 3326, + "GAGGCC": 3327, + "GAGGCG": 3328, + "GAGGGA": 3329, + "GAGGGT": 3330, + "GAGGGC": 3331, + "GAGGGG": 3332, + "GTAAAA": 3333, + "GTAAAT": 3334, + "GTAAAC": 3335, + "GTAAAG": 3336, + "GTAATA": 3337, + "GTAATT": 3338, + "GTAATC": 3339, + "GTAATG": 3340, + "GTAACA": 3341, + "GTAACT": 3342, + "GTAACC": 3343, + "GTAACG": 3344, + "GTAAGA": 3345, + "GTAAGT": 3346, + "GTAAGC": 3347, + "GTAAGG": 3348, + "GTATAA": 3349, + "GTATAT": 3350, + "GTATAC": 3351, + "GTATAG": 3352, + "GTATTA": 3353, + "GTATTT": 3354, + "GTATTC": 3355, + "GTATTG": 3356, + "GTATCA": 3357, + "GTATCT": 3358, + "GTATCC": 3359, + "GTATCG": 3360, + "GTATGA": 3361, + "GTATGT": 3362, + "GTATGC": 3363, + "GTATGG": 3364, + "GTACAA": 3365, + "GTACAT": 3366, + "GTACAC": 3367, + "GTACAG": 3368, + "GTACTA": 3369, + "GTACTT": 3370, + "GTACTC": 3371, + "GTACTG": 3372, + "GTACCA": 3373, + "GTACCT": 3374, + "GTACCC": 3375, + "GTACCG": 3376, + "GTACGA": 3377, + "GTACGT": 3378, + "GTACGC": 3379, + "GTACGG": 3380, + "GTAGAA": 3381, + "GTAGAT": 3382, + "GTAGAC": 3383, + "GTAGAG": 3384, + "GTAGTA": 3385, + "GTAGTT": 3386, + "GTAGTC": 3387, + "GTAGTG": 3388, + "GTAGCA": 3389, + "GTAGCT": 3390, + "GTAGCC": 3391, + "GTAGCG": 3392, + "GTAGGA": 3393, + "GTAGGT": 3394, + "GTAGGC": 3395, + "GTAGGG": 3396, + "GTTAAA": 3397, + "GTTAAT": 3398, + "GTTAAC": 3399, + "GTTAAG": 3400, + "GTTATA": 3401, + "GTTATT": 3402, + "GTTATC": 3403, + "GTTATG": 3404, + "GTTACA": 3405, + "GTTACT": 3406, + "GTTACC": 3407, + "GTTACG": 3408, + "GTTAGA": 3409, + "GTTAGT": 3410, + "GTTAGC": 3411, + "GTTAGG": 3412, + "GTTTAA": 3413, + "GTTTAT": 3414, + "GTTTAC": 3415, + "GTTTAG": 3416, + "GTTTTA": 3417, + "GTTTTT": 3418, + "GTTTTC": 3419, + "GTTTTG": 3420, + "GTTTCA": 3421, + "GTTTCT": 3422, + "GTTTCC": 3423, + "GTTTCG": 3424, + "GTTTGA": 3425, + "GTTTGT": 3426, + "GTTTGC": 3427, + "GTTTGG": 3428, + "GTTCAA": 3429, + "GTTCAT": 3430, + "GTTCAC": 3431, + "GTTCAG": 3432, + "GTTCTA": 3433, + "GTTCTT": 3434, + "GTTCTC": 3435, + "GTTCTG": 3436, + "GTTCCA": 3437, + "GTTCCT": 3438, + "GTTCCC": 3439, + "GTTCCG": 3440, + "GTTCGA": 3441, + "GTTCGT": 3442, + "GTTCGC": 3443, + "GTTCGG": 3444, + "GTTGAA": 3445, + "GTTGAT": 3446, + "GTTGAC": 3447, + "GTTGAG": 3448, + "GTTGTA": 3449, + "GTTGTT": 3450, + "GTTGTC": 3451, + "GTTGTG": 3452, + "GTTGCA": 3453, + "GTTGCT": 3454, + "GTTGCC": 3455, + "GTTGCG": 3456, + "GTTGGA": 3457, + "GTTGGT": 3458, + "GTTGGC": 3459, + "GTTGGG": 3460, + "GTCAAA": 3461, + "GTCAAT": 3462, + "GTCAAC": 3463, + "GTCAAG": 3464, + "GTCATA": 3465, + "GTCATT": 3466, + "GTCATC": 3467, + "GTCATG": 3468, + "GTCACA": 3469, + "GTCACT": 3470, + "GTCACC": 3471, + "GTCACG": 3472, + "GTCAGA": 3473, + "GTCAGT": 3474, + "GTCAGC": 3475, + "GTCAGG": 3476, + "GTCTAA": 3477, + "GTCTAT": 3478, + "GTCTAC": 3479, + "GTCTAG": 3480, + "GTCTTA": 3481, + "GTCTTT": 3482, + "GTCTTC": 3483, + "GTCTTG": 3484, + "GTCTCA": 3485, + "GTCTCT": 3486, + "GTCTCC": 3487, + "GTCTCG": 3488, + "GTCTGA": 3489, + "GTCTGT": 3490, + "GTCTGC": 3491, + "GTCTGG": 3492, + "GTCCAA": 3493, + "GTCCAT": 3494, + "GTCCAC": 3495, + "GTCCAG": 3496, + "GTCCTA": 3497, + "GTCCTT": 3498, + "GTCCTC": 3499, + "GTCCTG": 3500, + "GTCCCA": 3501, + "GTCCCT": 3502, + "GTCCCC": 3503, + "GTCCCG": 3504, + "GTCCGA": 3505, + "GTCCGT": 3506, + "GTCCGC": 3507, + "GTCCGG": 3508, + "GTCGAA": 3509, + "GTCGAT": 3510, + "GTCGAC": 3511, + "GTCGAG": 3512, + "GTCGTA": 3513, + "GTCGTT": 3514, + "GTCGTC": 3515, + "GTCGTG": 3516, + "GTCGCA": 3517, + "GTCGCT": 3518, + "GTCGCC": 3519, + "GTCGCG": 3520, + "GTCGGA": 3521, + "GTCGGT": 3522, + "GTCGGC": 3523, + "GTCGGG": 3524, + "GTGAAA": 3525, + "GTGAAT": 3526, + "GTGAAC": 3527, + "GTGAAG": 3528, + "GTGATA": 3529, + "GTGATT": 3530, + "GTGATC": 3531, + "GTGATG": 3532, + "GTGACA": 3533, + "GTGACT": 3534, + "GTGACC": 3535, + "GTGACG": 3536, + "GTGAGA": 3537, + "GTGAGT": 3538, + "GTGAGC": 3539, + "GTGAGG": 3540, + "GTGTAA": 3541, + "GTGTAT": 3542, + "GTGTAC": 3543, + "GTGTAG": 3544, + "GTGTTA": 3545, + "GTGTTT": 3546, + "GTGTTC": 3547, + "GTGTTG": 3548, + "GTGTCA": 3549, + "GTGTCT": 3550, + "GTGTCC": 3551, + "GTGTCG": 3552, + "GTGTGA": 3553, + "GTGTGT": 3554, + "GTGTGC": 3555, + "GTGTGG": 3556, + "GTGCAA": 3557, + "GTGCAT": 3558, + "GTGCAC": 3559, + "GTGCAG": 3560, + "GTGCTA": 3561, + "GTGCTT": 3562, + "GTGCTC": 3563, + "GTGCTG": 3564, + "GTGCCA": 3565, + "GTGCCT": 3566, + "GTGCCC": 3567, + "GTGCCG": 3568, + "GTGCGA": 3569, + "GTGCGT": 3570, + "GTGCGC": 3571, + "GTGCGG": 3572, + "GTGGAA": 3573, + "GTGGAT": 3574, + "GTGGAC": 3575, + "GTGGAG": 3576, + "GTGGTA": 3577, + "GTGGTT": 3578, + "GTGGTC": 3579, + "GTGGTG": 3580, + "GTGGCA": 3581, + "GTGGCT": 3582, + "GTGGCC": 3583, + "GTGGCG": 3584, + "GTGGGA": 3585, + "GTGGGT": 3586, + "GTGGGC": 3587, + "GTGGGG": 3588, + "GCAAAA": 3589, + "GCAAAT": 3590, + "GCAAAC": 3591, + "GCAAAG": 3592, + "GCAATA": 3593, + "GCAATT": 3594, + "GCAATC": 3595, + "GCAATG": 3596, + "GCAACA": 3597, + "GCAACT": 3598, + "GCAACC": 3599, + "GCAACG": 3600, + "GCAAGA": 3601, + "GCAAGT": 3602, + "GCAAGC": 3603, + "GCAAGG": 3604, + "GCATAA": 3605, + "GCATAT": 3606, + "GCATAC": 3607, + "GCATAG": 3608, + "GCATTA": 3609, + "GCATTT": 3610, + "GCATTC": 3611, + "GCATTG": 3612, + "GCATCA": 3613, + "GCATCT": 3614, + "GCATCC": 3615, + "GCATCG": 3616, + "GCATGA": 3617, + "GCATGT": 3618, + "GCATGC": 3619, + "GCATGG": 3620, + "GCACAA": 3621, + "GCACAT": 3622, + "GCACAC": 3623, + "GCACAG": 3624, + "GCACTA": 3625, + "GCACTT": 3626, + "GCACTC": 3627, + "GCACTG": 3628, + "GCACCA": 3629, + "GCACCT": 3630, + "GCACCC": 3631, + "GCACCG": 3632, + "GCACGA": 3633, + "GCACGT": 3634, + "GCACGC": 3635, + "GCACGG": 3636, + "GCAGAA": 3637, + "GCAGAT": 3638, + "GCAGAC": 3639, + "GCAGAG": 3640, + "GCAGTA": 3641, + "GCAGTT": 3642, + "GCAGTC": 3643, + "GCAGTG": 3644, + "GCAGCA": 3645, + "GCAGCT": 3646, + "GCAGCC": 3647, + "GCAGCG": 3648, + "GCAGGA": 3649, + "GCAGGT": 3650, + "GCAGGC": 3651, + "GCAGGG": 3652, + "GCTAAA": 3653, + "GCTAAT": 3654, + "GCTAAC": 3655, + "GCTAAG": 3656, + "GCTATA": 3657, + "GCTATT": 3658, + "GCTATC": 3659, + "GCTATG": 3660, + "GCTACA": 3661, + "GCTACT": 3662, + "GCTACC": 3663, + "GCTACG": 3664, + "GCTAGA": 3665, + "GCTAGT": 3666, + "GCTAGC": 3667, + "GCTAGG": 3668, + "GCTTAA": 3669, + "GCTTAT": 3670, + "GCTTAC": 3671, + "GCTTAG": 3672, + "GCTTTA": 3673, + "GCTTTT": 3674, + "GCTTTC": 3675, + "GCTTTG": 3676, + "GCTTCA": 3677, + "GCTTCT": 3678, + "GCTTCC": 3679, + "GCTTCG": 3680, + "GCTTGA": 3681, + "GCTTGT": 3682, + "GCTTGC": 3683, + "GCTTGG": 3684, + "GCTCAA": 3685, + "GCTCAT": 3686, + "GCTCAC": 3687, + "GCTCAG": 3688, + "GCTCTA": 3689, + "GCTCTT": 3690, + "GCTCTC": 3691, + "GCTCTG": 3692, + "GCTCCA": 3693, + "GCTCCT": 3694, + "GCTCCC": 3695, + "GCTCCG": 3696, + "GCTCGA": 3697, + "GCTCGT": 3698, + "GCTCGC": 3699, + "GCTCGG": 3700, + "GCTGAA": 3701, + "GCTGAT": 3702, + "GCTGAC": 3703, + "GCTGAG": 3704, + "GCTGTA": 3705, + "GCTGTT": 3706, + "GCTGTC": 3707, + "GCTGTG": 3708, + "GCTGCA": 3709, + "GCTGCT": 3710, + "GCTGCC": 3711, + "GCTGCG": 3712, + "GCTGGA": 3713, + "GCTGGT": 3714, + "GCTGGC": 3715, + "GCTGGG": 3716, + "GCCAAA": 3717, + "GCCAAT": 3718, + "GCCAAC": 3719, + "GCCAAG": 3720, + "GCCATA": 3721, + "GCCATT": 3722, + "GCCATC": 3723, + "GCCATG": 3724, + "GCCACA": 3725, + "GCCACT": 3726, + "GCCACC": 3727, + "GCCACG": 3728, + "GCCAGA": 3729, + "GCCAGT": 3730, + "GCCAGC": 3731, + "GCCAGG": 3732, + "GCCTAA": 3733, + "GCCTAT": 3734, + "GCCTAC": 3735, + "GCCTAG": 3736, + "GCCTTA": 3737, + "GCCTTT": 3738, + "GCCTTC": 3739, + "GCCTTG": 3740, + "GCCTCA": 3741, + "GCCTCT": 3742, + "GCCTCC": 3743, + "GCCTCG": 3744, + "GCCTGA": 3745, + "GCCTGT": 3746, + "GCCTGC": 3747, + "GCCTGG": 3748, + "GCCCAA": 3749, + "GCCCAT": 3750, + "GCCCAC": 3751, + "GCCCAG": 3752, + "GCCCTA": 3753, + "GCCCTT": 3754, + "GCCCTC": 3755, + "GCCCTG": 3756, + "GCCCCA": 3757, + "GCCCCT": 3758, + "GCCCCC": 3759, + "GCCCCG": 3760, + "GCCCGA": 3761, + "GCCCGT": 3762, + "GCCCGC": 3763, + "GCCCGG": 3764, + "GCCGAA": 3765, + "GCCGAT": 3766, + "GCCGAC": 3767, + "GCCGAG": 3768, + "GCCGTA": 3769, + "GCCGTT": 3770, + "GCCGTC": 3771, + "GCCGTG": 3772, + "GCCGCA": 3773, + "GCCGCT": 3774, + "GCCGCC": 3775, + "GCCGCG": 3776, + "GCCGGA": 3777, + "GCCGGT": 3778, + "GCCGGC": 3779, + "GCCGGG": 3780, + "GCGAAA": 3781, + "GCGAAT": 3782, + "GCGAAC": 3783, + "GCGAAG": 3784, + "GCGATA": 3785, + "GCGATT": 3786, + "GCGATC": 3787, + "GCGATG": 3788, + "GCGACA": 3789, + "GCGACT": 3790, + "GCGACC": 3791, + "GCGACG": 3792, + "GCGAGA": 3793, + "GCGAGT": 3794, + "GCGAGC": 3795, + "GCGAGG": 3796, + "GCGTAA": 3797, + "GCGTAT": 3798, + "GCGTAC": 3799, + "GCGTAG": 3800, + "GCGTTA": 3801, + "GCGTTT": 3802, + "GCGTTC": 3803, + "GCGTTG": 3804, + "GCGTCA": 3805, + "GCGTCT": 3806, + "GCGTCC": 3807, + "GCGTCG": 3808, + "GCGTGA": 3809, + "GCGTGT": 3810, + "GCGTGC": 3811, + "GCGTGG": 3812, + "GCGCAA": 3813, + "GCGCAT": 3814, + "GCGCAC": 3815, + "GCGCAG": 3816, + "GCGCTA": 3817, + "GCGCTT": 3818, + "GCGCTC": 3819, + "GCGCTG": 3820, + "GCGCCA": 3821, + "GCGCCT": 3822, + "GCGCCC": 3823, + "GCGCCG": 3824, + "GCGCGA": 3825, + "GCGCGT": 3826, + "GCGCGC": 3827, + "GCGCGG": 3828, + "GCGGAA": 3829, + "GCGGAT": 3830, + "GCGGAC": 3831, + "GCGGAG": 3832, + "GCGGTA": 3833, + "GCGGTT": 3834, + "GCGGTC": 3835, + "GCGGTG": 3836, + "GCGGCA": 3837, + "GCGGCT": 3838, + "GCGGCC": 3839, + "GCGGCG": 3840, + "GCGGGA": 3841, + "GCGGGT": 3842, + "GCGGGC": 3843, + "GCGGGG": 3844, + "GGAAAA": 3845, + "GGAAAT": 3846, + "GGAAAC": 3847, + "GGAAAG": 3848, + "GGAATA": 3849, + "GGAATT": 3850, + "GGAATC": 3851, + "GGAATG": 3852, + "GGAACA": 3853, + "GGAACT": 3854, + "GGAACC": 3855, + "GGAACG": 3856, + "GGAAGA": 3857, + "GGAAGT": 3858, + "GGAAGC": 3859, + "GGAAGG": 3860, + "GGATAA": 3861, + "GGATAT": 3862, + "GGATAC": 3863, + "GGATAG": 3864, + "GGATTA": 3865, + "GGATTT": 3866, + "GGATTC": 3867, + "GGATTG": 3868, + "GGATCA": 3869, + "GGATCT": 3870, + "GGATCC": 3871, + "GGATCG": 3872, + "GGATGA": 3873, + "GGATGT": 3874, + "GGATGC": 3875, + "GGATGG": 3876, + "GGACAA": 3877, + "GGACAT": 3878, + "GGACAC": 3879, + "GGACAG": 3880, + "GGACTA": 3881, + "GGACTT": 3882, + "GGACTC": 3883, + "GGACTG": 3884, + "GGACCA": 3885, + "GGACCT": 3886, + "GGACCC": 3887, + "GGACCG": 3888, + "GGACGA": 3889, + "GGACGT": 3890, + "GGACGC": 3891, + "GGACGG": 3892, + "GGAGAA": 3893, + "GGAGAT": 3894, + "GGAGAC": 3895, + "GGAGAG": 3896, + "GGAGTA": 3897, + "GGAGTT": 3898, + "GGAGTC": 3899, + "GGAGTG": 3900, + "GGAGCA": 3901, + "GGAGCT": 3902, + "GGAGCC": 3903, + "GGAGCG": 3904, + "GGAGGA": 3905, + "GGAGGT": 3906, + "GGAGGC": 3907, + "GGAGGG": 3908, + "GGTAAA": 3909, + "GGTAAT": 3910, + "GGTAAC": 3911, + "GGTAAG": 3912, + "GGTATA": 3913, + "GGTATT": 3914, + "GGTATC": 3915, + "GGTATG": 3916, + "GGTACA": 3917, + "GGTACT": 3918, + "GGTACC": 3919, + "GGTACG": 3920, + "GGTAGA": 3921, + "GGTAGT": 3922, + "GGTAGC": 3923, + "GGTAGG": 3924, + "GGTTAA": 3925, + "GGTTAT": 3926, + "GGTTAC": 3927, + "GGTTAG": 3928, + "GGTTTA": 3929, + "GGTTTT": 3930, + "GGTTTC": 3931, + "GGTTTG": 3932, + "GGTTCA": 3933, + "GGTTCT": 3934, + "GGTTCC": 3935, + "GGTTCG": 3936, + "GGTTGA": 3937, + "GGTTGT": 3938, + "GGTTGC": 3939, + "GGTTGG": 3940, + "GGTCAA": 3941, + "GGTCAT": 3942, + "GGTCAC": 3943, + "GGTCAG": 3944, + "GGTCTA": 3945, + "GGTCTT": 3946, + "GGTCTC": 3947, + "GGTCTG": 3948, + "GGTCCA": 3949, + "GGTCCT": 3950, + "GGTCCC": 3951, + "GGTCCG": 3952, + "GGTCGA": 3953, + "GGTCGT": 3954, + "GGTCGC": 3955, + "GGTCGG": 3956, + "GGTGAA": 3957, + "GGTGAT": 3958, + "GGTGAC": 3959, + "GGTGAG": 3960, + "GGTGTA": 3961, + "GGTGTT": 3962, + "GGTGTC": 3963, + "GGTGTG": 3964, + "GGTGCA": 3965, + "GGTGCT": 3966, + "GGTGCC": 3967, + "GGTGCG": 3968, + "GGTGGA": 3969, + "GGTGGT": 3970, + "GGTGGC": 3971, + "GGTGGG": 3972, + "GGCAAA": 3973, + "GGCAAT": 3974, + "GGCAAC": 3975, + "GGCAAG": 3976, + "GGCATA": 3977, + "GGCATT": 3978, + "GGCATC": 3979, + "GGCATG": 3980, + "GGCACA": 3981, + "GGCACT": 3982, + "GGCACC": 3983, + "GGCACG": 3984, + "GGCAGA": 3985, + "GGCAGT": 3986, + "GGCAGC": 3987, + "GGCAGG": 3988, + "GGCTAA": 3989, + "GGCTAT": 3990, + "GGCTAC": 3991, + "GGCTAG": 3992, + "GGCTTA": 3993, + "GGCTTT": 3994, + "GGCTTC": 3995, + "GGCTTG": 3996, + "GGCTCA": 3997, + "GGCTCT": 3998, + "GGCTCC": 3999, + "GGCTCG": 4000, + "GGCTGA": 4001, + "GGCTGT": 4002, + "GGCTGC": 4003, + "GGCTGG": 4004, + "GGCCAA": 4005, + "GGCCAT": 4006, + "GGCCAC": 4007, + "GGCCAG": 4008, + "GGCCTA": 4009, + "GGCCTT": 4010, + "GGCCTC": 4011, + "GGCCTG": 4012, + "GGCCCA": 4013, + "GGCCCT": 4014, + "GGCCCC": 4015, + "GGCCCG": 4016, + "GGCCGA": 4017, + "GGCCGT": 4018, + "GGCCGC": 4019, + "GGCCGG": 4020, + "GGCGAA": 4021, + "GGCGAT": 4022, + "GGCGAC": 4023, + "GGCGAG": 4024, + "GGCGTA": 4025, + "GGCGTT": 4026, + "GGCGTC": 4027, + "GGCGTG": 4028, + "GGCGCA": 4029, + "GGCGCT": 4030, + "GGCGCC": 4031, + "GGCGCG": 4032, + "GGCGGA": 4033, + "GGCGGT": 4034, + "GGCGGC": 4035, + "GGCGGG": 4036, + "GGGAAA": 4037, + "GGGAAT": 4038, + "GGGAAC": 4039, + "GGGAAG": 4040, + "GGGATA": 4041, + "GGGATT": 4042, + "GGGATC": 4043, + "GGGATG": 4044, + "GGGACA": 4045, + "GGGACT": 4046, + "GGGACC": 4047, + "GGGACG": 4048, + "GGGAGA": 4049, + "GGGAGT": 4050, + "GGGAGC": 4051, + "GGGAGG": 4052, + "GGGTAA": 4053, + "GGGTAT": 4054, + "GGGTAC": 4055, + "GGGTAG": 4056, + "GGGTTA": 4057, + "GGGTTT": 4058, + "GGGTTC": 4059, + "GGGTTG": 4060, + "GGGTCA": 4061, + "GGGTCT": 4062, + "GGGTCC": 4063, + "GGGTCG": 4064, + "GGGTGA": 4065, + "GGGTGT": 4066, + "GGGTGC": 4067, + "GGGTGG": 4068, + "GGGCAA": 4069, + "GGGCAT": 4070, + "GGGCAC": 4071, + "GGGCAG": 4072, + "GGGCTA": 4073, + "GGGCTT": 4074, + "GGGCTC": 4075, + "GGGCTG": 4076, + "GGGCCA": 4077, + "GGGCCT": 4078, + "GGGCCC": 4079, + "GGGCCG": 4080, + "GGGCGA": 4081, + "GGGCGT": 4082, + "GGGCGC": 4083, + "GGGCGG": 4084, + "GGGGAA": 4085, + "GGGGAT": 4086, + "GGGGAC": 4087, + "GGGGAG": 4088, + "GGGGTA": 4089, + "GGGGTT": 4090, + "GGGGTC": 4091, + "GGGGTG": 4092, + "GGGGCA": 4093, + "GGGGCT": 4094, + "GGGGCC": 4095, + "GGGGCG": 4096, + "GGGGGA": 4097, + "GGGGGT": 4098, + "GGGGGC": 4099, + "GGGGGG": 4100 + } + } +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6000/tokenizer_config.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6000/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..604e2b1b29af1dcb3920d2828b6a90276e5e3d02 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6000/tokenizer_config.json @@ -0,0 +1,58 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "[UNK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "[CLS]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "[SEP]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "4": { + "content": "[MASK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "clean_up_tokenization_spaces": true, + "cls_token": "[CLS]", + "do_basic_tokenize": true, + "do_lower_case": false, + "mask_token": "[MASK]", + "max_len": 512, + "model_max_length": 512, + "never_split": null, + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "strip_accents": null, + "tokenize_chinese_chars": true, + "tokenizer_class": "BertTokenizer", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6000/trainer_state.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..ccabf0b67bd1e1b8a5013d107183d629db6bd405 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6000/trainer_state.json @@ -0,0 +1,1713 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 24.0, + "eval_steps": 500, + "global_step": 6000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.1, + "grad_norm": 0.6038870215415955, + "learning_rate": 5.319148936170213e-05, + "loss": 0.6945, + "step": 25 + }, + { + "epoch": 0.2, + "grad_norm": 0.6668091416358948, + "learning_rate": 0.00010638297872340425, + "loss": 0.6929, + "step": 50 + }, + { + "epoch": 0.3, + "grad_norm": 0.8282511234283447, + "learning_rate": 0.00015957446808510637, + "loss": 0.6925, + "step": 75 + }, + { + "epoch": 0.4, + "grad_norm": 1.3149235248565674, + "learning_rate": 0.0002127659574468085, + "loss": 0.6919, + "step": 100 + }, + { + "epoch": 0.5, + "grad_norm": 1.8841172456741333, + "learning_rate": 0.00026595744680851064, + "loss": 0.6788, + "step": 125 + }, + { + "epoch": 0.6, + "grad_norm": 4.435206413269043, + "learning_rate": 0.00031914893617021275, + "loss": 0.6285, + "step": 150 + }, + { + "epoch": 0.7, + "grad_norm": 8.858148574829102, + "learning_rate": 0.0003723404255319149, + "loss": 0.5386, + "step": 175 + }, + { + "epoch": 0.8, + "grad_norm": 5.627648830413818, + "learning_rate": 0.0003992081821181128, + "loss": 0.5537, + "step": 200 + }, + { + "epoch": 0.9, + "grad_norm": 5.086580753326416, + "learning_rate": 0.0003975585615308479, + "loss": 0.5595, + "step": 225 + }, + { + "epoch": 1.0, + "grad_norm": 5.627719402313232, + "learning_rate": 0.00039590894094358297, + "loss": 0.4824, + "step": 250 + }, + { + "epoch": 1.1, + "grad_norm": 5.888038158416748, + "learning_rate": 0.0003942593203563181, + "loss": 0.4404, + "step": 275 + }, + { + "epoch": 1.2, + "grad_norm": 5.646231651306152, + "learning_rate": 0.0003926096997690532, + "loss": 0.4354, + "step": 300 + }, + { + "epoch": 1.3, + "grad_norm": 5.191990852355957, + "learning_rate": 0.00039096007918178817, + "loss": 0.4469, + "step": 325 + }, + { + "epoch": 1.4, + "grad_norm": 13.90101432800293, + "learning_rate": 0.0003893104585945233, + "loss": 0.3858, + "step": 350 + }, + { + "epoch": 1.5, + "grad_norm": 17.75942039489746, + "learning_rate": 0.0003876608380072583, + "loss": 0.4131, + "step": 375 + }, + { + "epoch": 1.6, + "grad_norm": 7.398252487182617, + "learning_rate": 0.00038601121741999343, + "loss": 0.3842, + "step": 400 + }, + { + "epoch": 1.7, + "grad_norm": 11.010066986083984, + "learning_rate": 0.0003843615968327285, + "loss": 0.3959, + "step": 425 + }, + { + "epoch": 1.8, + "grad_norm": 9.277341842651367, + "learning_rate": 0.0003827119762454636, + "loss": 0.3552, + "step": 450 + }, + { + "epoch": 1.9, + "grad_norm": 26.011449813842773, + "learning_rate": 0.00038106235565819863, + "loss": 0.3961, + "step": 475 + }, + { + "epoch": 2.0, + "grad_norm": 11.123054504394531, + "learning_rate": 0.0003794127350709337, + "loss": 0.3548, + "step": 500 + }, + { + "epoch": 2.1, + "grad_norm": 20.128774642944336, + "learning_rate": 0.0003777631144836688, + "loss": 0.302, + "step": 525 + }, + { + "epoch": 2.2, + "grad_norm": 4.64390230178833, + "learning_rate": 0.00037611349389640383, + "loss": 0.2853, + "step": 550 + }, + { + "epoch": 2.3, + "grad_norm": 4.199934482574463, + "learning_rate": 0.00037446387330913894, + "loss": 0.2724, + "step": 575 + }, + { + "epoch": 2.4, + "grad_norm": 6.909411430358887, + "learning_rate": 0.000372814252721874, + "loss": 0.2976, + "step": 600 + }, + { + "epoch": 2.5, + "grad_norm": 9.234820365905762, + "learning_rate": 0.00037116463213460903, + "loss": 0.2798, + "step": 625 + }, + { + "epoch": 2.6, + "grad_norm": 10.472943305969238, + "learning_rate": 0.00036951501154734414, + "loss": 0.3007, + "step": 650 + }, + { + "epoch": 2.7, + "grad_norm": 19.46732521057129, + "learning_rate": 0.0003678653909600792, + "loss": 0.2701, + "step": 675 + }, + { + "epoch": 2.8, + "grad_norm": 7.501255512237549, + "learning_rate": 0.0003662157703728143, + "loss": 0.2896, + "step": 700 + }, + { + "epoch": 2.9, + "grad_norm": 10.645585060119629, + "learning_rate": 0.00036456614978554934, + "loss": 0.2748, + "step": 725 + }, + { + "epoch": 3.0, + "grad_norm": 30.26616859436035, + "learning_rate": 0.00036291652919828444, + "loss": 0.268, + "step": 750 + }, + { + "epoch": 3.1, + "grad_norm": 14.13516616821289, + "learning_rate": 0.00036126690861101944, + "loss": 0.2159, + "step": 775 + }, + { + "epoch": 3.2, + "grad_norm": 10.091023445129395, + "learning_rate": 0.00035961728802375454, + "loss": 0.2071, + "step": 800 + }, + { + "epoch": 3.3, + "grad_norm": 22.201452255249023, + "learning_rate": 0.0003579676674364896, + "loss": 0.2146, + "step": 825 + }, + { + "epoch": 3.4, + "grad_norm": 10.044412612915039, + "learning_rate": 0.0003563180468492247, + "loss": 0.1827, + "step": 850 + }, + { + "epoch": 3.5, + "grad_norm": 21.04941177368164, + "learning_rate": 0.0003546684262619598, + "loss": 0.202, + "step": 875 + }, + { + "epoch": 3.6, + "grad_norm": 11.754891395568848, + "learning_rate": 0.00035301880567469485, + "loss": 0.2021, + "step": 900 + }, + { + "epoch": 3.7, + "grad_norm": 9.113747596740723, + "learning_rate": 0.0003513691850874299, + "loss": 0.2067, + "step": 925 + }, + { + "epoch": 3.8, + "grad_norm": 9.075196266174316, + "learning_rate": 0.00034971956450016495, + "loss": 0.3285, + "step": 950 + }, + { + "epoch": 3.9, + "grad_norm": 8.568796157836914, + "learning_rate": 0.00034806994391290005, + "loss": 0.2254, + "step": 975 + }, + { + "epoch": 4.0, + "grad_norm": 3.9587976932525635, + "learning_rate": 0.0003464203233256351, + "loss": 0.2245, + "step": 1000 + }, + { + "epoch": 4.1, + "grad_norm": 11.143651008605957, + "learning_rate": 0.0003447707027383702, + "loss": 0.1755, + "step": 1025 + }, + { + "epoch": 4.2, + "grad_norm": 12.53118896484375, + "learning_rate": 0.00034312108215110525, + "loss": 0.1808, + "step": 1050 + }, + { + "epoch": 4.3, + "grad_norm": 9.880820274353027, + "learning_rate": 0.0003414714615638403, + "loss": 0.1387, + "step": 1075 + }, + { + "epoch": 4.4, + "grad_norm": 5.729593753814697, + "learning_rate": 0.0003398218409765754, + "loss": 0.1958, + "step": 1100 + }, + { + "epoch": 4.5, + "grad_norm": 8.633047103881836, + "learning_rate": 0.00033817222038931045, + "loss": 0.1515, + "step": 1125 + }, + { + "epoch": 4.6, + "grad_norm": 19.237892150878906, + "learning_rate": 0.00033652259980204556, + "loss": 0.1452, + "step": 1150 + }, + { + "epoch": 4.7, + "grad_norm": 6.129818439483643, + "learning_rate": 0.0003348729792147806, + "loss": 0.1617, + "step": 1175 + }, + { + "epoch": 4.8, + "grad_norm": 5.073274612426758, + "learning_rate": 0.0003332233586275157, + "loss": 0.1967, + "step": 1200 + }, + { + "epoch": 4.9, + "grad_norm": 5.729922771453857, + "learning_rate": 0.00033157373804025076, + "loss": 0.1482, + "step": 1225 + }, + { + "epoch": 5.0, + "grad_norm": 12.314518928527832, + "learning_rate": 0.0003299241174529858, + "loss": 0.1588, + "step": 1250 + }, + { + "epoch": 5.1, + "grad_norm": 3.2636606693267822, + "learning_rate": 0.0003282744968657209, + "loss": 0.1588, + "step": 1275 + }, + { + "epoch": 5.2, + "grad_norm": 5.914486408233643, + "learning_rate": 0.00032662487627845596, + "loss": 0.1288, + "step": 1300 + }, + { + "epoch": 5.3, + "grad_norm": 13.849296569824219, + "learning_rate": 0.00032497525569119106, + "loss": 0.0961, + "step": 1325 + }, + { + "epoch": 5.4, + "grad_norm": 7.7317328453063965, + "learning_rate": 0.0003233256351039261, + "loss": 0.1148, + "step": 1350 + }, + { + "epoch": 5.5, + "grad_norm": 1.768608808517456, + "learning_rate": 0.00032167601451666116, + "loss": 0.121, + "step": 1375 + }, + { + "epoch": 5.6, + "grad_norm": 10.802785873413086, + "learning_rate": 0.00032002639392939627, + "loss": 0.1786, + "step": 1400 + }, + { + "epoch": 5.7, + "grad_norm": 4.158301830291748, + "learning_rate": 0.0003183767733421313, + "loss": 0.1078, + "step": 1425 + }, + { + "epoch": 5.8, + "grad_norm": 10.65804672241211, + "learning_rate": 0.0003167271527548664, + "loss": 0.1597, + "step": 1450 + }, + { + "epoch": 5.9, + "grad_norm": 10.880438804626465, + "learning_rate": 0.00031507753216760147, + "loss": 0.1225, + "step": 1475 + }, + { + "epoch": 6.0, + "grad_norm": 9.95138168334961, + "learning_rate": 0.00031342791158033657, + "loss": 0.1254, + "step": 1500 + }, + { + "epoch": 6.1, + "grad_norm": 1.5018162727355957, + "learning_rate": 0.00031177829099307157, + "loss": 0.0902, + "step": 1525 + }, + { + "epoch": 6.2, + "grad_norm": 14.281820297241211, + "learning_rate": 0.00031012867040580667, + "loss": 0.1026, + "step": 1550 + }, + { + "epoch": 6.3, + "grad_norm": 4.7540435791015625, + "learning_rate": 0.0003084790498185417, + "loss": 0.0876, + "step": 1575 + }, + { + "epoch": 6.4, + "grad_norm": 4.304183006286621, + "learning_rate": 0.0003068294292312768, + "loss": 0.1355, + "step": 1600 + }, + { + "epoch": 6.5, + "grad_norm": 14.653755187988281, + "learning_rate": 0.0003051798086440119, + "loss": 0.0878, + "step": 1625 + }, + { + "epoch": 6.6, + "grad_norm": 11.13028335571289, + "learning_rate": 0.000303530188056747, + "loss": 0.1203, + "step": 1650 + }, + { + "epoch": 6.7, + "grad_norm": 9.729938507080078, + "learning_rate": 0.000301880567469482, + "loss": 0.0775, + "step": 1675 + }, + { + "epoch": 6.8, + "grad_norm": 10.12938117980957, + "learning_rate": 0.0003002309468822171, + "loss": 0.1015, + "step": 1700 + }, + { + "epoch": 6.9, + "grad_norm": 8.506614685058594, + "learning_rate": 0.0002985813262949522, + "loss": 0.1362, + "step": 1725 + }, + { + "epoch": 7.0, + "grad_norm": 25.618762969970703, + "learning_rate": 0.00029693170570768723, + "loss": 0.1269, + "step": 1750 + }, + { + "epoch": 7.1, + "grad_norm": 27.155860900878906, + "learning_rate": 0.00029528208512042233, + "loss": 0.0555, + "step": 1775 + }, + { + "epoch": 7.2, + "grad_norm": 9.412527084350586, + "learning_rate": 0.0002936324645331574, + "loss": 0.0695, + "step": 1800 + }, + { + "epoch": 7.3, + "grad_norm": 6.4328293800354, + "learning_rate": 0.00029198284394589243, + "loss": 0.0808, + "step": 1825 + }, + { + "epoch": 7.4, + "grad_norm": 15.267293930053711, + "learning_rate": 0.00029033322335862753, + "loss": 0.0871, + "step": 1850 + }, + { + "epoch": 7.5, + "grad_norm": 6.774560928344727, + "learning_rate": 0.0002886836027713626, + "loss": 0.0819, + "step": 1875 + }, + { + "epoch": 7.6, + "grad_norm": 18.67180633544922, + "learning_rate": 0.0002870339821840977, + "loss": 0.0565, + "step": 1900 + }, + { + "epoch": 7.7, + "grad_norm": 16.39004898071289, + "learning_rate": 0.00028538436159683273, + "loss": 0.0907, + "step": 1925 + }, + { + "epoch": 7.8, + "grad_norm": 6.351583003997803, + "learning_rate": 0.00028373474100956784, + "loss": 0.1289, + "step": 1950 + }, + { + "epoch": 7.9, + "grad_norm": 11.396740913391113, + "learning_rate": 0.0002820851204223029, + "loss": 0.1149, + "step": 1975 + }, + { + "epoch": 8.0, + "grad_norm": 6.491570949554443, + "learning_rate": 0.00028043549983503794, + "loss": 0.0891, + "step": 2000 + }, + { + "epoch": 8.1, + "grad_norm": 0.22946956753730774, + "learning_rate": 0.00027878587924777304, + "loss": 0.0413, + "step": 2025 + }, + { + "epoch": 8.2, + "grad_norm": 42.31606674194336, + "learning_rate": 0.0002771362586605081, + "loss": 0.0881, + "step": 2050 + }, + { + "epoch": 8.3, + "grad_norm": 18.490406036376953, + "learning_rate": 0.0002754866380732432, + "loss": 0.1238, + "step": 2075 + }, + { + "epoch": 8.4, + "grad_norm": 17.950319290161133, + "learning_rate": 0.00027383701748597824, + "loss": 0.0642, + "step": 2100 + }, + { + "epoch": 8.5, + "grad_norm": 37.16359329223633, + "learning_rate": 0.0002721873968987133, + "loss": 0.0916, + "step": 2125 + }, + { + "epoch": 8.6, + "grad_norm": 39.75199890136719, + "learning_rate": 0.00027053777631144834, + "loss": 0.1087, + "step": 2150 + }, + { + "epoch": 8.7, + "grad_norm": 0.08794238418340683, + "learning_rate": 0.00026888815572418344, + "loss": 0.0825, + "step": 2175 + }, + { + "epoch": 8.8, + "grad_norm": 14.681429862976074, + "learning_rate": 0.00026723853513691855, + "loss": 0.0925, + "step": 2200 + }, + { + "epoch": 8.9, + "grad_norm": 17.90182876586914, + "learning_rate": 0.0002655889145496536, + "loss": 0.0933, + "step": 2225 + }, + { + "epoch": 9.0, + "grad_norm": 0.9108687043190002, + "learning_rate": 0.0002639392939623887, + "loss": 0.0602, + "step": 2250 + }, + { + "epoch": 9.1, + "grad_norm": 9.671142578125, + "learning_rate": 0.0002622896733751237, + "loss": 0.0438, + "step": 2275 + }, + { + "epoch": 9.2, + "grad_norm": 12.797035217285156, + "learning_rate": 0.0002606400527878588, + "loss": 0.0495, + "step": 2300 + }, + { + "epoch": 9.3, + "grad_norm": 0.4585617184638977, + "learning_rate": 0.00025899043220059385, + "loss": 0.032, + "step": 2325 + }, + { + "epoch": 9.4, + "grad_norm": 0.05073042958974838, + "learning_rate": 0.00025734081161332895, + "loss": 0.067, + "step": 2350 + }, + { + "epoch": 9.5, + "grad_norm": 0.013653877191245556, + "learning_rate": 0.000255691191026064, + "loss": 0.0898, + "step": 2375 + }, + { + "epoch": 9.6, + "grad_norm": 18.198965072631836, + "learning_rate": 0.0002540415704387991, + "loss": 0.0804, + "step": 2400 + }, + { + "epoch": 9.7, + "grad_norm": 0.1363959163427353, + "learning_rate": 0.00025239194985153415, + "loss": 0.0584, + "step": 2425 + }, + { + "epoch": 9.8, + "grad_norm": 17.61290740966797, + "learning_rate": 0.0002507423292642692, + "loss": 0.0778, + "step": 2450 + }, + { + "epoch": 9.9, + "grad_norm": 0.10548887401819229, + "learning_rate": 0.0002490927086770043, + "loss": 0.0667, + "step": 2475 + }, + { + "epoch": 10.0, + "grad_norm": 0.1262211948633194, + "learning_rate": 0.00024744308808973936, + "loss": 0.0625, + "step": 2500 + }, + { + "epoch": 10.1, + "grad_norm": 0.019293297082185745, + "learning_rate": 0.00024579346750247446, + "loss": 0.0762, + "step": 2525 + }, + { + "epoch": 10.2, + "grad_norm": 0.9273566603660583, + "learning_rate": 0.0002441438469152095, + "loss": 0.0357, + "step": 2550 + }, + { + "epoch": 10.3, + "grad_norm": 10.720550537109375, + "learning_rate": 0.00024249422632794456, + "loss": 0.046, + "step": 2575 + }, + { + "epoch": 10.4, + "grad_norm": 1.8441611528396606, + "learning_rate": 0.00024084460574067963, + "loss": 0.0369, + "step": 2600 + }, + { + "epoch": 10.5, + "grad_norm": 0.8760197758674622, + "learning_rate": 0.0002391949851534147, + "loss": 0.046, + "step": 2625 + }, + { + "epoch": 10.6, + "grad_norm": 8.899499893188477, + "learning_rate": 0.0002375453645661498, + "loss": 0.08, + "step": 2650 + }, + { + "epoch": 10.7, + "grad_norm": 0.2664288580417633, + "learning_rate": 0.00023589574397888486, + "loss": 0.0616, + "step": 2675 + }, + { + "epoch": 10.8, + "grad_norm": 0.06609692424535751, + "learning_rate": 0.00023424612339161997, + "loss": 0.05, + "step": 2700 + }, + { + "epoch": 10.9, + "grad_norm": 32.44829177856445, + "learning_rate": 0.000232596502804355, + "loss": 0.0487, + "step": 2725 + }, + { + "epoch": 11.0, + "grad_norm": 7.25365686416626, + "learning_rate": 0.00023094688221709007, + "loss": 0.0456, + "step": 2750 + }, + { + "epoch": 11.1, + "grad_norm": 0.10101272910833359, + "learning_rate": 0.00022929726162982514, + "loss": 0.0328, + "step": 2775 + }, + { + "epoch": 11.2, + "grad_norm": 1.3087996244430542, + "learning_rate": 0.00022764764104256022, + "loss": 0.0606, + "step": 2800 + }, + { + "epoch": 11.3, + "grad_norm": 1.247656226158142, + "learning_rate": 0.0002259980204552953, + "loss": 0.0838, + "step": 2825 + }, + { + "epoch": 11.4, + "grad_norm": 1.7708797454833984, + "learning_rate": 0.00022434839986803037, + "loss": 0.0416, + "step": 2850 + }, + { + "epoch": 11.5, + "grad_norm": 11.131721496582031, + "learning_rate": 0.00022269877928076542, + "loss": 0.0528, + "step": 2875 + }, + { + "epoch": 11.6, + "grad_norm": 3.824920892715454, + "learning_rate": 0.0002210491586935005, + "loss": 0.0526, + "step": 2900 + }, + { + "epoch": 11.7, + "grad_norm": 0.8507682085037231, + "learning_rate": 0.00021939953810623557, + "loss": 0.0965, + "step": 2925 + }, + { + "epoch": 11.8, + "grad_norm": 0.011997719295322895, + "learning_rate": 0.00021774991751897065, + "loss": 0.0446, + "step": 2950 + }, + { + "epoch": 11.9, + "grad_norm": 0.0792650505900383, + "learning_rate": 0.00021610029693170573, + "loss": 0.0556, + "step": 2975 + }, + { + "epoch": 12.0, + "grad_norm": 0.007047992665320635, + "learning_rate": 0.0002144506763444408, + "loss": 0.0759, + "step": 3000 + }, + { + "epoch": 12.1, + "grad_norm": 0.7784481644630432, + "learning_rate": 0.00021280105575717582, + "loss": 0.0505, + "step": 3025 + }, + { + "epoch": 12.2, + "grad_norm": 9.635031700134277, + "learning_rate": 0.00021115143516991093, + "loss": 0.0284, + "step": 3050 + }, + { + "epoch": 12.3, + "grad_norm": 12.473057746887207, + "learning_rate": 0.000209501814582646, + "loss": 0.0531, + "step": 3075 + }, + { + "epoch": 12.4, + "grad_norm": 10.904062271118164, + "learning_rate": 0.00020785219399538108, + "loss": 0.0212, + "step": 3100 + }, + { + "epoch": 12.5, + "grad_norm": 9.258865356445312, + "learning_rate": 0.00020620257340811616, + "loss": 0.0367, + "step": 3125 + }, + { + "epoch": 12.6, + "grad_norm": 2.6826488971710205, + "learning_rate": 0.00020455295282085123, + "loss": 0.0457, + "step": 3150 + }, + { + "epoch": 12.7, + "grad_norm": 8.800697326660156, + "learning_rate": 0.00020290333223358626, + "loss": 0.0504, + "step": 3175 + }, + { + "epoch": 12.8, + "grad_norm": 0.023783667013049126, + "learning_rate": 0.00020125371164632133, + "loss": 0.0255, + "step": 3200 + }, + { + "epoch": 12.9, + "grad_norm": 5.948949337005615, + "learning_rate": 0.0001996040910590564, + "loss": 0.0413, + "step": 3225 + }, + { + "epoch": 13.0, + "grad_norm": 0.026645230129361153, + "learning_rate": 0.00019795447047179148, + "loss": 0.0249, + "step": 3250 + }, + { + "epoch": 13.1, + "grad_norm": 60.826637268066406, + "learning_rate": 0.0001963048498845266, + "loss": 0.0293, + "step": 3275 + }, + { + "epoch": 13.2, + "grad_norm": 26.056686401367188, + "learning_rate": 0.00019465522929726164, + "loss": 0.0638, + "step": 3300 + }, + { + "epoch": 13.3, + "grad_norm": 18.143638610839844, + "learning_rate": 0.00019300560870999671, + "loss": 0.0206, + "step": 3325 + }, + { + "epoch": 13.4, + "grad_norm": 27.899967193603516, + "learning_rate": 0.0001913559881227318, + "loss": 0.0423, + "step": 3350 + }, + { + "epoch": 13.5, + "grad_norm": 51.9879264831543, + "learning_rate": 0.00018970636753546684, + "loss": 0.0274, + "step": 3375 + }, + { + "epoch": 13.6, + "grad_norm": 0.028627146035432816, + "learning_rate": 0.00018805674694820192, + "loss": 0.0513, + "step": 3400 + }, + { + "epoch": 13.7, + "grad_norm": 33.211944580078125, + "learning_rate": 0.000186407126360937, + "loss": 0.0446, + "step": 3425 + }, + { + "epoch": 13.8, + "grad_norm": 32.37785339355469, + "learning_rate": 0.00018475750577367207, + "loss": 0.0538, + "step": 3450 + }, + { + "epoch": 13.9, + "grad_norm": 30.558673858642578, + "learning_rate": 0.00018310788518640715, + "loss": 0.0339, + "step": 3475 + }, + { + "epoch": 14.0, + "grad_norm": 19.316938400268555, + "learning_rate": 0.00018145826459914222, + "loss": 0.071, + "step": 3500 + }, + { + "epoch": 14.1, + "grad_norm": 0.1271055042743683, + "learning_rate": 0.00017980864401187727, + "loss": 0.0159, + "step": 3525 + }, + { + "epoch": 14.2, + "grad_norm": 51.1268196105957, + "learning_rate": 0.00017815902342461235, + "loss": 0.028, + "step": 3550 + }, + { + "epoch": 14.3, + "grad_norm": 0.003214583033695817, + "learning_rate": 0.00017650940283734742, + "loss": 0.0342, + "step": 3575 + }, + { + "epoch": 14.4, + "grad_norm": 94.75691223144531, + "learning_rate": 0.00017485978225008247, + "loss": 0.0243, + "step": 3600 + }, + { + "epoch": 14.5, + "grad_norm": 0.019844651222229004, + "learning_rate": 0.00017321016166281755, + "loss": 0.0205, + "step": 3625 + }, + { + "epoch": 14.6, + "grad_norm": 0.10069809854030609, + "learning_rate": 0.00017156054107555263, + "loss": 0.0402, + "step": 3650 + }, + { + "epoch": 14.7, + "grad_norm": 0.0023937481455504894, + "learning_rate": 0.0001699109204882877, + "loss": 0.0173, + "step": 3675 + }, + { + "epoch": 14.8, + "grad_norm": 0.2573629319667816, + "learning_rate": 0.00016826129990102278, + "loss": 0.0312, + "step": 3700 + }, + { + "epoch": 14.9, + "grad_norm": 0.0067031364887952805, + "learning_rate": 0.00016661167931375785, + "loss": 0.0418, + "step": 3725 + }, + { + "epoch": 15.0, + "grad_norm": 0.08329538255929947, + "learning_rate": 0.0001649620587264929, + "loss": 0.0164, + "step": 3750 + }, + { + "epoch": 15.1, + "grad_norm": 0.04175519943237305, + "learning_rate": 0.00016331243813922798, + "loss": 0.044, + "step": 3775 + }, + { + "epoch": 15.2, + "grad_norm": 0.005420651752501726, + "learning_rate": 0.00016166281755196306, + "loss": 0.0599, + "step": 3800 + }, + { + "epoch": 15.3, + "grad_norm": 20.68973731994629, + "learning_rate": 0.00016001319696469813, + "loss": 0.047, + "step": 3825 + }, + { + "epoch": 15.4, + "grad_norm": 1.1241929531097412, + "learning_rate": 0.0001583635763774332, + "loss": 0.0184, + "step": 3850 + }, + { + "epoch": 15.5, + "grad_norm": 0.00928444042801857, + "learning_rate": 0.00015671395579016829, + "loss": 0.0146, + "step": 3875 + }, + { + "epoch": 15.6, + "grad_norm": 3.2419674396514893, + "learning_rate": 0.00015506433520290334, + "loss": 0.0047, + "step": 3900 + }, + { + "epoch": 15.7, + "grad_norm": 0.0019528602715581656, + "learning_rate": 0.0001534147146156384, + "loss": 0.0177, + "step": 3925 + }, + { + "epoch": 15.8, + "grad_norm": 0.003142274683341384, + "learning_rate": 0.0001517650940283735, + "loss": 0.0319, + "step": 3950 + }, + { + "epoch": 15.9, + "grad_norm": 45.58081817626953, + "learning_rate": 0.00015011547344110854, + "loss": 0.0353, + "step": 3975 + }, + { + "epoch": 16.0, + "grad_norm": 0.003515947377309203, + "learning_rate": 0.00014846585285384361, + "loss": 0.0208, + "step": 4000 + }, + { + "epoch": 16.1, + "grad_norm": 0.01574423350393772, + "learning_rate": 0.0001468162322665787, + "loss": 0.0198, + "step": 4025 + }, + { + "epoch": 16.2, + "grad_norm": 22.461671829223633, + "learning_rate": 0.00014516661167931377, + "loss": 0.0262, + "step": 4050 + }, + { + "epoch": 16.3, + "grad_norm": 17.334550857543945, + "learning_rate": 0.00014351699109204884, + "loss": 0.0567, + "step": 4075 + }, + { + "epoch": 16.4, + "grad_norm": 0.00328136351890862, + "learning_rate": 0.00014186737050478392, + "loss": 0.0182, + "step": 4100 + }, + { + "epoch": 16.5, + "grad_norm": 1.2645615339279175, + "learning_rate": 0.00014021774991751897, + "loss": 0.0116, + "step": 4125 + }, + { + "epoch": 16.6, + "grad_norm": 0.0031980539206415415, + "learning_rate": 0.00013856812933025404, + "loss": 0.0079, + "step": 4150 + }, + { + "epoch": 16.7, + "grad_norm": 0.01752198301255703, + "learning_rate": 0.00013691850874298912, + "loss": 0.0576, + "step": 4175 + }, + { + "epoch": 16.8, + "grad_norm": 48.27894592285156, + "learning_rate": 0.00013526888815572417, + "loss": 0.0264, + "step": 4200 + }, + { + "epoch": 16.9, + "grad_norm": 0.008802427910268307, + "learning_rate": 0.00013361926756845927, + "loss": 0.0051, + "step": 4225 + }, + { + "epoch": 17.0, + "grad_norm": 0.0350751169025898, + "learning_rate": 0.00013196964698119435, + "loss": 0.0403, + "step": 4250 + }, + { + "epoch": 17.1, + "grad_norm": 0.0020171860232949257, + "learning_rate": 0.0001303200263939294, + "loss": 0.0012, + "step": 4275 + }, + { + "epoch": 17.2, + "grad_norm": 0.014922365546226501, + "learning_rate": 0.00012867040580666448, + "loss": 0.036, + "step": 4300 + }, + { + "epoch": 17.3, + "grad_norm": 0.01983438804745674, + "learning_rate": 0.00012702078521939955, + "loss": 0.0094, + "step": 4325 + }, + { + "epoch": 17.4, + "grad_norm": 0.0031381379812955856, + "learning_rate": 0.0001253711646321346, + "loss": 0.011, + "step": 4350 + }, + { + "epoch": 17.5, + "grad_norm": 0.656819224357605, + "learning_rate": 0.00012372154404486968, + "loss": 0.0094, + "step": 4375 + }, + { + "epoch": 17.6, + "grad_norm": 0.008435660973191261, + "learning_rate": 0.00012207192345760475, + "loss": 0.0445, + "step": 4400 + }, + { + "epoch": 17.7, + "grad_norm": 23.819961547851562, + "learning_rate": 0.00012042230287033982, + "loss": 0.0394, + "step": 4425 + }, + { + "epoch": 17.8, + "grad_norm": 0.0055985236540436745, + "learning_rate": 0.0001187726822830749, + "loss": 0.0106, + "step": 4450 + }, + { + "epoch": 17.9, + "grad_norm": 11.063871383666992, + "learning_rate": 0.00011712306169580998, + "loss": 0.0128, + "step": 4475 + }, + { + "epoch": 18.0, + "grad_norm": 0.003767346730455756, + "learning_rate": 0.00011547344110854503, + "loss": 0.0096, + "step": 4500 + }, + { + "epoch": 18.1, + "grad_norm": 0.018198227509856224, + "learning_rate": 0.00011382382052128011, + "loss": 0.0395, + "step": 4525 + }, + { + "epoch": 18.2, + "grad_norm": 0.004393396433442831, + "learning_rate": 0.00011217419993401519, + "loss": 0.0378, + "step": 4550 + }, + { + "epoch": 18.3, + "grad_norm": 0.01423373818397522, + "learning_rate": 0.00011052457934675025, + "loss": 0.013, + "step": 4575 + }, + { + "epoch": 18.4, + "grad_norm": 25.492542266845703, + "learning_rate": 0.00010887495875948532, + "loss": 0.0094, + "step": 4600 + }, + { + "epoch": 18.5, + "grad_norm": 31.223493576049805, + "learning_rate": 0.0001072253381722204, + "loss": 0.0161, + "step": 4625 + }, + { + "epoch": 18.6, + "grad_norm": 0.003598716575652361, + "learning_rate": 0.00010557571758495546, + "loss": 0.0106, + "step": 4650 + }, + { + "epoch": 18.7, + "grad_norm": 50.249603271484375, + "learning_rate": 0.00010392609699769054, + "loss": 0.024, + "step": 4675 + }, + { + "epoch": 18.8, + "grad_norm": 63.50950622558594, + "learning_rate": 0.00010227647641042562, + "loss": 0.018, + "step": 4700 + }, + { + "epoch": 18.9, + "grad_norm": 0.002336911391466856, + "learning_rate": 0.00010062685582316067, + "loss": 0.0091, + "step": 4725 + }, + { + "epoch": 19.0, + "grad_norm": 0.005011886823922396, + "learning_rate": 9.897723523589574e-05, + "loss": 0.0061, + "step": 4750 + }, + { + "epoch": 19.1, + "grad_norm": 0.027857890352606773, + "learning_rate": 9.732761464863082e-05, + "loss": 0.0221, + "step": 4775 + }, + { + "epoch": 19.2, + "grad_norm": 0.0016227838350459933, + "learning_rate": 9.56779940613659e-05, + "loss": 0.0195, + "step": 4800 + }, + { + "epoch": 19.3, + "grad_norm": 0.005687279161065817, + "learning_rate": 9.402837347410096e-05, + "loss": 0.0203, + "step": 4825 + }, + { + "epoch": 19.4, + "grad_norm": 0.009174875915050507, + "learning_rate": 9.237875288683603e-05, + "loss": 0.0086, + "step": 4850 + }, + { + "epoch": 19.5, + "grad_norm": 0.0032144756987690926, + "learning_rate": 9.072913229957111e-05, + "loss": 0.0289, + "step": 4875 + }, + { + "epoch": 19.6, + "grad_norm": 0.003061591414734721, + "learning_rate": 8.907951171230617e-05, + "loss": 0.0015, + "step": 4900 + }, + { + "epoch": 19.7, + "grad_norm": 0.006960512604564428, + "learning_rate": 8.742989112504124e-05, + "loss": 0.0176, + "step": 4925 + }, + { + "epoch": 19.8, + "grad_norm": 0.0027057609986513853, + "learning_rate": 8.578027053777631e-05, + "loss": 0.0047, + "step": 4950 + }, + { + "epoch": 19.9, + "grad_norm": 0.22216364741325378, + "learning_rate": 8.413064995051139e-05, + "loss": 0.0099, + "step": 4975 + }, + { + "epoch": 20.0, + "grad_norm": 0.0022015271242707968, + "learning_rate": 8.248102936324645e-05, + "loss": 0.0219, + "step": 5000 + }, + { + "epoch": 20.1, + "grad_norm": 0.019280511885881424, + "learning_rate": 8.083140877598153e-05, + "loss": 0.0129, + "step": 5025 + }, + { + "epoch": 20.2, + "grad_norm": 0.005342870019376278, + "learning_rate": 7.91817881887166e-05, + "loss": 0.016, + "step": 5050 + }, + { + "epoch": 20.3, + "grad_norm": 0.013655520975589752, + "learning_rate": 7.753216760145167e-05, + "loss": 0.0174, + "step": 5075 + }, + { + "epoch": 20.4, + "grad_norm": 0.0037927369121462107, + "learning_rate": 7.588254701418674e-05, + "loss": 0.021, + "step": 5100 + }, + { + "epoch": 20.5, + "grad_norm": 0.0036746144760400057, + "learning_rate": 7.423292642692181e-05, + "loss": 0.004, + "step": 5125 + }, + { + "epoch": 20.6, + "grad_norm": 0.014823048375546932, + "learning_rate": 7.258330583965688e-05, + "loss": 0.0161, + "step": 5150 + }, + { + "epoch": 20.7, + "grad_norm": 0.0033355404157191515, + "learning_rate": 7.093368525239196e-05, + "loss": 0.0035, + "step": 5175 + }, + { + "epoch": 20.8, + "grad_norm": 0.0018589649116620421, + "learning_rate": 6.928406466512702e-05, + "loss": 0.0105, + "step": 5200 + }, + { + "epoch": 20.9, + "grad_norm": 0.013408798724412918, + "learning_rate": 6.763444407786209e-05, + "loss": 0.0001, + "step": 5225 + }, + { + "epoch": 21.0, + "grad_norm": 70.37215423583984, + "learning_rate": 6.598482349059718e-05, + "loss": 0.0419, + "step": 5250 + }, + { + "epoch": 21.1, + "grad_norm": 0.09737736731767654, + "learning_rate": 6.433520290333224e-05, + "loss": 0.0146, + "step": 5275 + }, + { + "epoch": 21.2, + "grad_norm": 0.012592921033501625, + "learning_rate": 6.26855823160673e-05, + "loss": 0.0333, + "step": 5300 + }, + { + "epoch": 21.3, + "grad_norm": 0.0022445865906774998, + "learning_rate": 6.103596172880238e-05, + "loss": 0.0241, + "step": 5325 + }, + { + "epoch": 21.4, + "grad_norm": 0.01678585074841976, + "learning_rate": 5.938634114153745e-05, + "loss": 0.0047, + "step": 5350 + }, + { + "epoch": 21.5, + "grad_norm": 0.07437444478273392, + "learning_rate": 5.7736720554272516e-05, + "loss": 0.0213, + "step": 5375 + }, + { + "epoch": 21.6, + "grad_norm": 0.002057570032775402, + "learning_rate": 5.608709996700759e-05, + "loss": 0.0001, + "step": 5400 + }, + { + "epoch": 21.7, + "grad_norm": 0.4772427976131439, + "learning_rate": 5.443747937974266e-05, + "loss": 0.0038, + "step": 5425 + }, + { + "epoch": 21.8, + "grad_norm": 0.0016257546376436949, + "learning_rate": 5.278785879247773e-05, + "loss": 0.0076, + "step": 5450 + }, + { + "epoch": 21.9, + "grad_norm": 0.02936253882944584, + "learning_rate": 5.113823820521281e-05, + "loss": 0.0004, + "step": 5475 + }, + { + "epoch": 22.0, + "grad_norm": 0.12796524167060852, + "learning_rate": 4.948861761794787e-05, + "loss": 0.011, + "step": 5500 + }, + { + "epoch": 22.1, + "grad_norm": 0.0015867372276261449, + "learning_rate": 4.783899703068295e-05, + "loss": 0.0066, + "step": 5525 + }, + { + "epoch": 22.2, + "grad_norm": 54.19424819946289, + "learning_rate": 4.618937644341802e-05, + "loss": 0.0102, + "step": 5550 + }, + { + "epoch": 22.3, + "grad_norm": 0.07260770350694656, + "learning_rate": 4.453975585615309e-05, + "loss": 0.0016, + "step": 5575 + }, + { + "epoch": 22.4, + "grad_norm": 0.004677828401327133, + "learning_rate": 4.2890135268888156e-05, + "loss": 0.0005, + "step": 5600 + }, + { + "epoch": 22.5, + "grad_norm": 0.0017690203385427594, + "learning_rate": 4.1240514681623226e-05, + "loss": 0.0084, + "step": 5625 + }, + { + "epoch": 22.6, + "grad_norm": 0.0025316874962300062, + "learning_rate": 3.95908940943583e-05, + "loss": 0.006, + "step": 5650 + }, + { + "epoch": 22.7, + "grad_norm": 0.5509907007217407, + "learning_rate": 3.794127350709337e-05, + "loss": 0.001, + "step": 5675 + }, + { + "epoch": 22.8, + "grad_norm": 0.002628073561936617, + "learning_rate": 3.629165291982844e-05, + "loss": 0.0129, + "step": 5700 + }, + { + "epoch": 22.9, + "grad_norm": 0.004700347315520048, + "learning_rate": 3.464203233256351e-05, + "loss": 0.0002, + "step": 5725 + }, + { + "epoch": 23.0, + "grad_norm": 0.0015652852598577738, + "learning_rate": 3.299241174529859e-05, + "loss": 0.0001, + "step": 5750 + }, + { + "epoch": 23.1, + "grad_norm": 0.0014234065311029553, + "learning_rate": 3.134279115803365e-05, + "loss": 0.0002, + "step": 5775 + }, + { + "epoch": 23.2, + "grad_norm": 0.001536093419417739, + "learning_rate": 2.9693170570768723e-05, + "loss": 0.0007, + "step": 5800 + }, + { + "epoch": 23.3, + "grad_norm": 0.002034024801105261, + "learning_rate": 2.8043549983503796e-05, + "loss": 0.0069, + "step": 5825 + }, + { + "epoch": 23.4, + "grad_norm": 0.001170129980891943, + "learning_rate": 2.6393929396238866e-05, + "loss": 0.0182, + "step": 5850 + }, + { + "epoch": 23.5, + "grad_norm": 1.254554271697998, + "learning_rate": 2.4744308808973936e-05, + "loss": 0.0002, + "step": 5875 + }, + { + "epoch": 23.6, + "grad_norm": 0.0014751444105058908, + "learning_rate": 2.309468822170901e-05, + "loss": 0.0001, + "step": 5900 + }, + { + "epoch": 23.7, + "grad_norm": 1.3905378580093384, + "learning_rate": 2.1445067634444078e-05, + "loss": 0.0018, + "step": 5925 + }, + { + "epoch": 23.8, + "grad_norm": 0.001244895625859499, + "learning_rate": 1.979544704717915e-05, + "loss": 0.0076, + "step": 5950 + }, + { + "epoch": 23.9, + "grad_norm": 0.0016068242257460952, + "learning_rate": 1.814582645991422e-05, + "loss": 0.0017, + "step": 5975 + }, + { + "epoch": 24.0, + "grad_norm": 0.0011312123388051987, + "learning_rate": 1.6496205872649294e-05, + "loss": 0.0007, + "step": 6000 + } + ], + "logging_steps": 25, + "max_steps": 6250, + "num_input_tokens_seen": 0, + "num_train_epochs": 25, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 304916889600000.0, + "train_batch_size": 20, + "trial_name": null, + "trial_params": null +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6000/training_args.bin b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6000/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..77692b6666e83a688b38e799c964e4694a57eec9 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6000/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f73b41315068a430d2439f5707c62152a34210df62454b9549f9642446dd90b2 +size 5304 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6000/vocab.txt b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6000/vocab.txt new file mode 100644 index 0000000000000000000000000000000000000000..cc9ad53732e21ea9180f18e1fb67a38cc5702d4c --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6000/vocab.txt @@ -0,0 +1,4101 @@ +[PAD] +[UNK] +[CLS] +[SEP] +[MASK] +AAAAAA +AAAAAT +AAAAAC +AAAAAG +AAAATA +AAAATT +AAAATC +AAAATG +AAAACA +AAAACT +AAAACC +AAAACG +AAAAGA +AAAAGT +AAAAGC +AAAAGG +AAATAA +AAATAT +AAATAC +AAATAG +AAATTA +AAATTT +AAATTC +AAATTG +AAATCA +AAATCT +AAATCC +AAATCG +AAATGA +AAATGT +AAATGC +AAATGG +AAACAA +AAACAT +AAACAC +AAACAG +AAACTA +AAACTT +AAACTC +AAACTG +AAACCA +AAACCT +AAACCC +AAACCG +AAACGA +AAACGT +AAACGC +AAACGG +AAAGAA +AAAGAT +AAAGAC +AAAGAG +AAAGTA +AAAGTT +AAAGTC +AAAGTG +AAAGCA +AAAGCT +AAAGCC +AAAGCG +AAAGGA +AAAGGT +AAAGGC +AAAGGG +AATAAA +AATAAT +AATAAC +AATAAG +AATATA +AATATT +AATATC +AATATG +AATACA +AATACT +AATACC +AATACG +AATAGA +AATAGT +AATAGC +AATAGG +AATTAA +AATTAT +AATTAC +AATTAG +AATTTA +AATTTT +AATTTC +AATTTG +AATTCA +AATTCT +AATTCC +AATTCG +AATTGA +AATTGT +AATTGC +AATTGG +AATCAA +AATCAT +AATCAC +AATCAG +AATCTA +AATCTT +AATCTC +AATCTG +AATCCA +AATCCT +AATCCC +AATCCG +AATCGA +AATCGT +AATCGC +AATCGG +AATGAA +AATGAT +AATGAC +AATGAG +AATGTA +AATGTT +AATGTC +AATGTG +AATGCA +AATGCT +AATGCC +AATGCG +AATGGA +AATGGT +AATGGC +AATGGG +AACAAA +AACAAT +AACAAC +AACAAG +AACATA +AACATT +AACATC +AACATG +AACACA +AACACT +AACACC +AACACG +AACAGA +AACAGT +AACAGC +AACAGG +AACTAA +AACTAT +AACTAC +AACTAG +AACTTA +AACTTT +AACTTC +AACTTG +AACTCA +AACTCT +AACTCC +AACTCG +AACTGA +AACTGT +AACTGC +AACTGG +AACCAA +AACCAT +AACCAC +AACCAG +AACCTA +AACCTT +AACCTC +AACCTG +AACCCA +AACCCT +AACCCC +AACCCG +AACCGA +AACCGT +AACCGC +AACCGG +AACGAA +AACGAT +AACGAC +AACGAG +AACGTA +AACGTT +AACGTC +AACGTG +AACGCA +AACGCT +AACGCC +AACGCG +AACGGA +AACGGT +AACGGC +AACGGG +AAGAAA +AAGAAT +AAGAAC +AAGAAG +AAGATA +AAGATT +AAGATC +AAGATG +AAGACA +AAGACT +AAGACC +AAGACG +AAGAGA +AAGAGT +AAGAGC +AAGAGG +AAGTAA +AAGTAT +AAGTAC +AAGTAG +AAGTTA +AAGTTT +AAGTTC +AAGTTG +AAGTCA +AAGTCT +AAGTCC +AAGTCG +AAGTGA +AAGTGT +AAGTGC +AAGTGG +AAGCAA +AAGCAT +AAGCAC +AAGCAG +AAGCTA +AAGCTT +AAGCTC +AAGCTG +AAGCCA +AAGCCT +AAGCCC +AAGCCG +AAGCGA +AAGCGT +AAGCGC +AAGCGG +AAGGAA +AAGGAT +AAGGAC +AAGGAG +AAGGTA +AAGGTT +AAGGTC +AAGGTG +AAGGCA +AAGGCT +AAGGCC +AAGGCG +AAGGGA +AAGGGT +AAGGGC +AAGGGG +ATAAAA +ATAAAT +ATAAAC +ATAAAG +ATAATA +ATAATT +ATAATC +ATAATG +ATAACA +ATAACT +ATAACC +ATAACG +ATAAGA +ATAAGT +ATAAGC +ATAAGG +ATATAA +ATATAT +ATATAC +ATATAG +ATATTA +ATATTT +ATATTC +ATATTG +ATATCA +ATATCT +ATATCC +ATATCG +ATATGA +ATATGT +ATATGC +ATATGG +ATACAA +ATACAT +ATACAC +ATACAG +ATACTA +ATACTT +ATACTC +ATACTG +ATACCA +ATACCT +ATACCC +ATACCG +ATACGA +ATACGT +ATACGC +ATACGG +ATAGAA +ATAGAT +ATAGAC +ATAGAG +ATAGTA +ATAGTT +ATAGTC +ATAGTG +ATAGCA +ATAGCT +ATAGCC +ATAGCG +ATAGGA +ATAGGT +ATAGGC +ATAGGG +ATTAAA +ATTAAT +ATTAAC +ATTAAG +ATTATA +ATTATT +ATTATC +ATTATG +ATTACA +ATTACT +ATTACC +ATTACG +ATTAGA +ATTAGT +ATTAGC +ATTAGG +ATTTAA +ATTTAT +ATTTAC +ATTTAG +ATTTTA +ATTTTT +ATTTTC +ATTTTG +ATTTCA +ATTTCT +ATTTCC +ATTTCG +ATTTGA +ATTTGT +ATTTGC +ATTTGG +ATTCAA +ATTCAT +ATTCAC +ATTCAG +ATTCTA +ATTCTT +ATTCTC +ATTCTG +ATTCCA +ATTCCT +ATTCCC +ATTCCG +ATTCGA +ATTCGT +ATTCGC +ATTCGG +ATTGAA +ATTGAT +ATTGAC +ATTGAG +ATTGTA +ATTGTT +ATTGTC +ATTGTG +ATTGCA +ATTGCT +ATTGCC +ATTGCG +ATTGGA +ATTGGT +ATTGGC +ATTGGG +ATCAAA +ATCAAT +ATCAAC +ATCAAG +ATCATA +ATCATT +ATCATC +ATCATG +ATCACA +ATCACT +ATCACC +ATCACG +ATCAGA +ATCAGT +ATCAGC +ATCAGG +ATCTAA +ATCTAT +ATCTAC +ATCTAG +ATCTTA +ATCTTT +ATCTTC +ATCTTG +ATCTCA +ATCTCT +ATCTCC +ATCTCG +ATCTGA +ATCTGT +ATCTGC +ATCTGG +ATCCAA +ATCCAT +ATCCAC +ATCCAG +ATCCTA +ATCCTT +ATCCTC +ATCCTG +ATCCCA +ATCCCT +ATCCCC +ATCCCG +ATCCGA +ATCCGT +ATCCGC +ATCCGG +ATCGAA +ATCGAT +ATCGAC +ATCGAG +ATCGTA +ATCGTT +ATCGTC +ATCGTG +ATCGCA +ATCGCT +ATCGCC +ATCGCG +ATCGGA +ATCGGT +ATCGGC +ATCGGG +ATGAAA +ATGAAT +ATGAAC +ATGAAG +ATGATA +ATGATT +ATGATC +ATGATG +ATGACA +ATGACT +ATGACC +ATGACG +ATGAGA +ATGAGT +ATGAGC +ATGAGG +ATGTAA +ATGTAT +ATGTAC +ATGTAG +ATGTTA +ATGTTT +ATGTTC +ATGTTG +ATGTCA +ATGTCT +ATGTCC +ATGTCG +ATGTGA +ATGTGT +ATGTGC +ATGTGG +ATGCAA +ATGCAT +ATGCAC +ATGCAG +ATGCTA +ATGCTT +ATGCTC +ATGCTG +ATGCCA +ATGCCT +ATGCCC +ATGCCG +ATGCGA +ATGCGT +ATGCGC +ATGCGG +ATGGAA +ATGGAT +ATGGAC +ATGGAG +ATGGTA +ATGGTT +ATGGTC +ATGGTG +ATGGCA +ATGGCT +ATGGCC +ATGGCG +ATGGGA +ATGGGT +ATGGGC +ATGGGG +ACAAAA +ACAAAT +ACAAAC +ACAAAG +ACAATA +ACAATT +ACAATC +ACAATG +ACAACA +ACAACT +ACAACC +ACAACG +ACAAGA +ACAAGT +ACAAGC +ACAAGG +ACATAA +ACATAT +ACATAC +ACATAG +ACATTA +ACATTT +ACATTC +ACATTG +ACATCA +ACATCT +ACATCC +ACATCG +ACATGA +ACATGT +ACATGC +ACATGG +ACACAA +ACACAT +ACACAC +ACACAG +ACACTA +ACACTT +ACACTC +ACACTG +ACACCA +ACACCT +ACACCC +ACACCG +ACACGA +ACACGT +ACACGC +ACACGG +ACAGAA +ACAGAT +ACAGAC +ACAGAG +ACAGTA +ACAGTT +ACAGTC +ACAGTG +ACAGCA +ACAGCT +ACAGCC +ACAGCG +ACAGGA +ACAGGT +ACAGGC +ACAGGG +ACTAAA +ACTAAT +ACTAAC +ACTAAG +ACTATA +ACTATT +ACTATC +ACTATG +ACTACA +ACTACT +ACTACC +ACTACG +ACTAGA +ACTAGT +ACTAGC +ACTAGG +ACTTAA +ACTTAT +ACTTAC +ACTTAG +ACTTTA +ACTTTT +ACTTTC +ACTTTG +ACTTCA +ACTTCT +ACTTCC +ACTTCG +ACTTGA +ACTTGT +ACTTGC +ACTTGG +ACTCAA +ACTCAT +ACTCAC +ACTCAG +ACTCTA +ACTCTT +ACTCTC +ACTCTG +ACTCCA +ACTCCT +ACTCCC +ACTCCG +ACTCGA +ACTCGT +ACTCGC +ACTCGG +ACTGAA +ACTGAT +ACTGAC +ACTGAG +ACTGTA +ACTGTT +ACTGTC +ACTGTG +ACTGCA +ACTGCT +ACTGCC +ACTGCG +ACTGGA +ACTGGT +ACTGGC +ACTGGG +ACCAAA +ACCAAT +ACCAAC +ACCAAG +ACCATA +ACCATT +ACCATC +ACCATG +ACCACA +ACCACT +ACCACC +ACCACG +ACCAGA +ACCAGT +ACCAGC +ACCAGG +ACCTAA +ACCTAT +ACCTAC +ACCTAG +ACCTTA +ACCTTT +ACCTTC +ACCTTG +ACCTCA +ACCTCT +ACCTCC +ACCTCG +ACCTGA +ACCTGT +ACCTGC +ACCTGG +ACCCAA +ACCCAT +ACCCAC +ACCCAG +ACCCTA +ACCCTT +ACCCTC +ACCCTG +ACCCCA +ACCCCT +ACCCCC +ACCCCG +ACCCGA +ACCCGT +ACCCGC +ACCCGG +ACCGAA +ACCGAT +ACCGAC +ACCGAG +ACCGTA +ACCGTT +ACCGTC +ACCGTG +ACCGCA +ACCGCT +ACCGCC +ACCGCG +ACCGGA +ACCGGT +ACCGGC +ACCGGG +ACGAAA +ACGAAT +ACGAAC +ACGAAG +ACGATA +ACGATT +ACGATC +ACGATG +ACGACA +ACGACT +ACGACC +ACGACG +ACGAGA +ACGAGT +ACGAGC +ACGAGG +ACGTAA +ACGTAT +ACGTAC +ACGTAG +ACGTTA +ACGTTT +ACGTTC +ACGTTG +ACGTCA +ACGTCT +ACGTCC +ACGTCG +ACGTGA +ACGTGT +ACGTGC +ACGTGG +ACGCAA +ACGCAT +ACGCAC +ACGCAG +ACGCTA +ACGCTT +ACGCTC +ACGCTG +ACGCCA +ACGCCT +ACGCCC +ACGCCG +ACGCGA +ACGCGT +ACGCGC +ACGCGG +ACGGAA +ACGGAT +ACGGAC +ACGGAG +ACGGTA +ACGGTT +ACGGTC +ACGGTG +ACGGCA +ACGGCT +ACGGCC +ACGGCG +ACGGGA +ACGGGT +ACGGGC +ACGGGG +AGAAAA +AGAAAT +AGAAAC +AGAAAG +AGAATA +AGAATT +AGAATC +AGAATG +AGAACA +AGAACT +AGAACC +AGAACG +AGAAGA +AGAAGT +AGAAGC +AGAAGG +AGATAA +AGATAT +AGATAC +AGATAG +AGATTA +AGATTT +AGATTC +AGATTG +AGATCA +AGATCT +AGATCC +AGATCG +AGATGA +AGATGT +AGATGC +AGATGG +AGACAA +AGACAT +AGACAC +AGACAG +AGACTA +AGACTT +AGACTC +AGACTG +AGACCA +AGACCT +AGACCC +AGACCG +AGACGA +AGACGT +AGACGC +AGACGG +AGAGAA +AGAGAT +AGAGAC +AGAGAG +AGAGTA +AGAGTT +AGAGTC +AGAGTG +AGAGCA +AGAGCT +AGAGCC +AGAGCG +AGAGGA +AGAGGT +AGAGGC +AGAGGG +AGTAAA +AGTAAT +AGTAAC +AGTAAG +AGTATA +AGTATT +AGTATC +AGTATG +AGTACA +AGTACT +AGTACC +AGTACG +AGTAGA +AGTAGT +AGTAGC +AGTAGG +AGTTAA +AGTTAT +AGTTAC +AGTTAG +AGTTTA +AGTTTT +AGTTTC +AGTTTG +AGTTCA +AGTTCT +AGTTCC +AGTTCG +AGTTGA +AGTTGT +AGTTGC +AGTTGG +AGTCAA +AGTCAT +AGTCAC +AGTCAG +AGTCTA +AGTCTT +AGTCTC +AGTCTG +AGTCCA +AGTCCT +AGTCCC +AGTCCG +AGTCGA +AGTCGT +AGTCGC +AGTCGG +AGTGAA +AGTGAT +AGTGAC +AGTGAG +AGTGTA +AGTGTT +AGTGTC +AGTGTG +AGTGCA +AGTGCT +AGTGCC +AGTGCG +AGTGGA +AGTGGT +AGTGGC +AGTGGG +AGCAAA +AGCAAT +AGCAAC +AGCAAG +AGCATA +AGCATT +AGCATC +AGCATG +AGCACA +AGCACT +AGCACC +AGCACG +AGCAGA +AGCAGT +AGCAGC +AGCAGG +AGCTAA +AGCTAT +AGCTAC +AGCTAG +AGCTTA +AGCTTT +AGCTTC +AGCTTG +AGCTCA +AGCTCT +AGCTCC +AGCTCG +AGCTGA +AGCTGT +AGCTGC +AGCTGG +AGCCAA +AGCCAT +AGCCAC +AGCCAG +AGCCTA +AGCCTT +AGCCTC +AGCCTG +AGCCCA +AGCCCT +AGCCCC +AGCCCG +AGCCGA +AGCCGT +AGCCGC +AGCCGG +AGCGAA +AGCGAT +AGCGAC +AGCGAG +AGCGTA +AGCGTT +AGCGTC +AGCGTG +AGCGCA +AGCGCT +AGCGCC +AGCGCG +AGCGGA +AGCGGT +AGCGGC +AGCGGG +AGGAAA +AGGAAT +AGGAAC +AGGAAG +AGGATA +AGGATT +AGGATC +AGGATG +AGGACA +AGGACT +AGGACC +AGGACG +AGGAGA +AGGAGT +AGGAGC +AGGAGG +AGGTAA +AGGTAT +AGGTAC +AGGTAG +AGGTTA +AGGTTT +AGGTTC +AGGTTG +AGGTCA +AGGTCT +AGGTCC +AGGTCG +AGGTGA +AGGTGT +AGGTGC +AGGTGG +AGGCAA +AGGCAT +AGGCAC +AGGCAG +AGGCTA +AGGCTT +AGGCTC +AGGCTG +AGGCCA +AGGCCT +AGGCCC +AGGCCG +AGGCGA +AGGCGT +AGGCGC +AGGCGG +AGGGAA +AGGGAT +AGGGAC +AGGGAG +AGGGTA +AGGGTT +AGGGTC +AGGGTG +AGGGCA +AGGGCT +AGGGCC +AGGGCG +AGGGGA +AGGGGT +AGGGGC +AGGGGG +TAAAAA +TAAAAT +TAAAAC +TAAAAG +TAAATA +TAAATT +TAAATC +TAAATG +TAAACA +TAAACT +TAAACC +TAAACG +TAAAGA +TAAAGT +TAAAGC +TAAAGG +TAATAA +TAATAT +TAATAC +TAATAG +TAATTA +TAATTT +TAATTC +TAATTG +TAATCA +TAATCT +TAATCC +TAATCG +TAATGA +TAATGT +TAATGC +TAATGG +TAACAA +TAACAT +TAACAC +TAACAG +TAACTA +TAACTT +TAACTC +TAACTG +TAACCA +TAACCT +TAACCC +TAACCG +TAACGA +TAACGT +TAACGC +TAACGG +TAAGAA +TAAGAT +TAAGAC +TAAGAG +TAAGTA +TAAGTT +TAAGTC +TAAGTG +TAAGCA +TAAGCT +TAAGCC +TAAGCG +TAAGGA +TAAGGT +TAAGGC +TAAGGG +TATAAA +TATAAT +TATAAC +TATAAG +TATATA +TATATT +TATATC +TATATG +TATACA +TATACT +TATACC +TATACG +TATAGA +TATAGT +TATAGC +TATAGG +TATTAA +TATTAT +TATTAC +TATTAG +TATTTA +TATTTT +TATTTC +TATTTG +TATTCA +TATTCT +TATTCC +TATTCG +TATTGA +TATTGT +TATTGC +TATTGG +TATCAA +TATCAT +TATCAC +TATCAG +TATCTA +TATCTT +TATCTC +TATCTG +TATCCA +TATCCT +TATCCC +TATCCG +TATCGA +TATCGT +TATCGC +TATCGG +TATGAA +TATGAT +TATGAC +TATGAG +TATGTA +TATGTT +TATGTC +TATGTG +TATGCA +TATGCT +TATGCC +TATGCG +TATGGA +TATGGT +TATGGC +TATGGG +TACAAA +TACAAT +TACAAC +TACAAG +TACATA +TACATT +TACATC +TACATG +TACACA +TACACT +TACACC +TACACG +TACAGA +TACAGT +TACAGC +TACAGG +TACTAA +TACTAT +TACTAC +TACTAG +TACTTA +TACTTT +TACTTC +TACTTG +TACTCA +TACTCT +TACTCC +TACTCG +TACTGA +TACTGT +TACTGC +TACTGG +TACCAA +TACCAT +TACCAC +TACCAG +TACCTA +TACCTT +TACCTC +TACCTG +TACCCA +TACCCT +TACCCC +TACCCG +TACCGA +TACCGT +TACCGC +TACCGG +TACGAA +TACGAT +TACGAC +TACGAG +TACGTA +TACGTT +TACGTC +TACGTG +TACGCA +TACGCT +TACGCC +TACGCG +TACGGA +TACGGT +TACGGC +TACGGG +TAGAAA +TAGAAT +TAGAAC +TAGAAG +TAGATA +TAGATT +TAGATC +TAGATG +TAGACA +TAGACT +TAGACC +TAGACG +TAGAGA +TAGAGT +TAGAGC +TAGAGG +TAGTAA +TAGTAT +TAGTAC +TAGTAG +TAGTTA +TAGTTT +TAGTTC +TAGTTG +TAGTCA +TAGTCT +TAGTCC +TAGTCG +TAGTGA +TAGTGT +TAGTGC +TAGTGG +TAGCAA +TAGCAT +TAGCAC +TAGCAG +TAGCTA +TAGCTT +TAGCTC +TAGCTG +TAGCCA +TAGCCT +TAGCCC +TAGCCG +TAGCGA +TAGCGT +TAGCGC +TAGCGG +TAGGAA +TAGGAT +TAGGAC +TAGGAG +TAGGTA +TAGGTT +TAGGTC +TAGGTG +TAGGCA +TAGGCT +TAGGCC +TAGGCG +TAGGGA +TAGGGT +TAGGGC +TAGGGG +TTAAAA +TTAAAT +TTAAAC +TTAAAG +TTAATA +TTAATT +TTAATC +TTAATG +TTAACA +TTAACT +TTAACC +TTAACG +TTAAGA +TTAAGT +TTAAGC +TTAAGG +TTATAA +TTATAT +TTATAC +TTATAG +TTATTA +TTATTT +TTATTC +TTATTG +TTATCA +TTATCT +TTATCC +TTATCG +TTATGA +TTATGT +TTATGC +TTATGG +TTACAA +TTACAT +TTACAC +TTACAG +TTACTA +TTACTT +TTACTC +TTACTG +TTACCA +TTACCT +TTACCC +TTACCG +TTACGA +TTACGT +TTACGC +TTACGG +TTAGAA +TTAGAT +TTAGAC +TTAGAG +TTAGTA +TTAGTT +TTAGTC +TTAGTG +TTAGCA +TTAGCT +TTAGCC +TTAGCG +TTAGGA +TTAGGT +TTAGGC +TTAGGG +TTTAAA +TTTAAT +TTTAAC +TTTAAG +TTTATA +TTTATT +TTTATC +TTTATG +TTTACA +TTTACT +TTTACC +TTTACG +TTTAGA +TTTAGT +TTTAGC +TTTAGG +TTTTAA +TTTTAT +TTTTAC +TTTTAG +TTTTTA +TTTTTT +TTTTTC +TTTTTG +TTTTCA +TTTTCT +TTTTCC +TTTTCG +TTTTGA +TTTTGT +TTTTGC +TTTTGG +TTTCAA +TTTCAT +TTTCAC +TTTCAG +TTTCTA +TTTCTT +TTTCTC +TTTCTG +TTTCCA +TTTCCT +TTTCCC +TTTCCG +TTTCGA +TTTCGT +TTTCGC +TTTCGG +TTTGAA +TTTGAT +TTTGAC +TTTGAG +TTTGTA +TTTGTT +TTTGTC +TTTGTG +TTTGCA +TTTGCT +TTTGCC +TTTGCG +TTTGGA +TTTGGT +TTTGGC +TTTGGG +TTCAAA +TTCAAT +TTCAAC +TTCAAG +TTCATA +TTCATT +TTCATC +TTCATG +TTCACA +TTCACT +TTCACC +TTCACG +TTCAGA +TTCAGT +TTCAGC +TTCAGG +TTCTAA +TTCTAT +TTCTAC +TTCTAG +TTCTTA +TTCTTT +TTCTTC +TTCTTG +TTCTCA +TTCTCT +TTCTCC +TTCTCG +TTCTGA +TTCTGT +TTCTGC +TTCTGG +TTCCAA +TTCCAT +TTCCAC +TTCCAG +TTCCTA +TTCCTT +TTCCTC +TTCCTG +TTCCCA +TTCCCT +TTCCCC +TTCCCG +TTCCGA +TTCCGT +TTCCGC +TTCCGG +TTCGAA +TTCGAT +TTCGAC +TTCGAG +TTCGTA +TTCGTT +TTCGTC +TTCGTG +TTCGCA +TTCGCT +TTCGCC +TTCGCG +TTCGGA +TTCGGT +TTCGGC +TTCGGG +TTGAAA +TTGAAT +TTGAAC +TTGAAG +TTGATA +TTGATT +TTGATC +TTGATG +TTGACA +TTGACT +TTGACC +TTGACG +TTGAGA +TTGAGT +TTGAGC +TTGAGG +TTGTAA +TTGTAT +TTGTAC +TTGTAG +TTGTTA +TTGTTT +TTGTTC +TTGTTG +TTGTCA +TTGTCT +TTGTCC +TTGTCG +TTGTGA +TTGTGT +TTGTGC +TTGTGG +TTGCAA +TTGCAT +TTGCAC +TTGCAG +TTGCTA +TTGCTT +TTGCTC +TTGCTG +TTGCCA +TTGCCT +TTGCCC +TTGCCG +TTGCGA +TTGCGT +TTGCGC +TTGCGG +TTGGAA +TTGGAT +TTGGAC +TTGGAG +TTGGTA +TTGGTT +TTGGTC +TTGGTG +TTGGCA +TTGGCT +TTGGCC +TTGGCG +TTGGGA +TTGGGT +TTGGGC +TTGGGG +TCAAAA +TCAAAT +TCAAAC +TCAAAG +TCAATA +TCAATT +TCAATC +TCAATG +TCAACA +TCAACT +TCAACC +TCAACG +TCAAGA +TCAAGT +TCAAGC +TCAAGG +TCATAA +TCATAT +TCATAC +TCATAG +TCATTA +TCATTT +TCATTC +TCATTG +TCATCA +TCATCT +TCATCC +TCATCG +TCATGA +TCATGT +TCATGC +TCATGG +TCACAA +TCACAT +TCACAC +TCACAG +TCACTA +TCACTT +TCACTC +TCACTG +TCACCA +TCACCT +TCACCC +TCACCG +TCACGA +TCACGT +TCACGC +TCACGG +TCAGAA +TCAGAT +TCAGAC +TCAGAG +TCAGTA +TCAGTT +TCAGTC +TCAGTG +TCAGCA +TCAGCT +TCAGCC +TCAGCG +TCAGGA +TCAGGT +TCAGGC +TCAGGG +TCTAAA +TCTAAT +TCTAAC +TCTAAG +TCTATA +TCTATT +TCTATC +TCTATG +TCTACA +TCTACT +TCTACC +TCTACG +TCTAGA +TCTAGT +TCTAGC +TCTAGG +TCTTAA +TCTTAT +TCTTAC +TCTTAG +TCTTTA +TCTTTT +TCTTTC +TCTTTG +TCTTCA +TCTTCT +TCTTCC +TCTTCG +TCTTGA +TCTTGT +TCTTGC +TCTTGG +TCTCAA +TCTCAT +TCTCAC +TCTCAG +TCTCTA +TCTCTT +TCTCTC +TCTCTG +TCTCCA +TCTCCT +TCTCCC +TCTCCG +TCTCGA +TCTCGT +TCTCGC +TCTCGG +TCTGAA +TCTGAT +TCTGAC +TCTGAG +TCTGTA +TCTGTT +TCTGTC +TCTGTG +TCTGCA +TCTGCT +TCTGCC +TCTGCG +TCTGGA +TCTGGT +TCTGGC +TCTGGG +TCCAAA +TCCAAT +TCCAAC +TCCAAG +TCCATA +TCCATT +TCCATC +TCCATG +TCCACA +TCCACT +TCCACC +TCCACG +TCCAGA +TCCAGT +TCCAGC +TCCAGG +TCCTAA +TCCTAT +TCCTAC +TCCTAG +TCCTTA +TCCTTT +TCCTTC +TCCTTG +TCCTCA +TCCTCT +TCCTCC +TCCTCG +TCCTGA +TCCTGT +TCCTGC +TCCTGG +TCCCAA +TCCCAT +TCCCAC +TCCCAG +TCCCTA +TCCCTT +TCCCTC +TCCCTG +TCCCCA +TCCCCT +TCCCCC +TCCCCG +TCCCGA +TCCCGT +TCCCGC +TCCCGG +TCCGAA +TCCGAT +TCCGAC +TCCGAG +TCCGTA +TCCGTT +TCCGTC +TCCGTG +TCCGCA +TCCGCT +TCCGCC +TCCGCG +TCCGGA +TCCGGT +TCCGGC +TCCGGG +TCGAAA +TCGAAT +TCGAAC +TCGAAG +TCGATA +TCGATT +TCGATC +TCGATG +TCGACA +TCGACT +TCGACC +TCGACG +TCGAGA +TCGAGT +TCGAGC +TCGAGG +TCGTAA +TCGTAT +TCGTAC +TCGTAG +TCGTTA +TCGTTT +TCGTTC +TCGTTG +TCGTCA +TCGTCT +TCGTCC +TCGTCG +TCGTGA +TCGTGT +TCGTGC +TCGTGG +TCGCAA +TCGCAT +TCGCAC +TCGCAG +TCGCTA +TCGCTT +TCGCTC +TCGCTG +TCGCCA +TCGCCT +TCGCCC +TCGCCG +TCGCGA +TCGCGT +TCGCGC +TCGCGG +TCGGAA +TCGGAT +TCGGAC +TCGGAG +TCGGTA +TCGGTT +TCGGTC +TCGGTG +TCGGCA +TCGGCT +TCGGCC +TCGGCG +TCGGGA +TCGGGT +TCGGGC +TCGGGG +TGAAAA +TGAAAT +TGAAAC +TGAAAG +TGAATA +TGAATT +TGAATC +TGAATG +TGAACA +TGAACT +TGAACC +TGAACG +TGAAGA +TGAAGT +TGAAGC +TGAAGG +TGATAA +TGATAT +TGATAC +TGATAG +TGATTA +TGATTT +TGATTC +TGATTG +TGATCA +TGATCT +TGATCC +TGATCG +TGATGA +TGATGT +TGATGC +TGATGG +TGACAA +TGACAT +TGACAC +TGACAG +TGACTA +TGACTT +TGACTC +TGACTG +TGACCA +TGACCT +TGACCC +TGACCG +TGACGA +TGACGT +TGACGC +TGACGG +TGAGAA +TGAGAT +TGAGAC +TGAGAG +TGAGTA +TGAGTT +TGAGTC +TGAGTG +TGAGCA +TGAGCT +TGAGCC +TGAGCG +TGAGGA +TGAGGT +TGAGGC +TGAGGG +TGTAAA +TGTAAT +TGTAAC +TGTAAG +TGTATA +TGTATT +TGTATC +TGTATG +TGTACA +TGTACT +TGTACC +TGTACG +TGTAGA +TGTAGT +TGTAGC +TGTAGG +TGTTAA +TGTTAT +TGTTAC +TGTTAG +TGTTTA +TGTTTT +TGTTTC +TGTTTG +TGTTCA +TGTTCT +TGTTCC +TGTTCG +TGTTGA +TGTTGT +TGTTGC +TGTTGG +TGTCAA +TGTCAT +TGTCAC +TGTCAG +TGTCTA +TGTCTT +TGTCTC +TGTCTG +TGTCCA +TGTCCT +TGTCCC +TGTCCG +TGTCGA +TGTCGT +TGTCGC +TGTCGG +TGTGAA +TGTGAT +TGTGAC +TGTGAG +TGTGTA +TGTGTT +TGTGTC +TGTGTG +TGTGCA +TGTGCT +TGTGCC +TGTGCG +TGTGGA +TGTGGT +TGTGGC +TGTGGG +TGCAAA +TGCAAT +TGCAAC +TGCAAG +TGCATA +TGCATT +TGCATC +TGCATG +TGCACA +TGCACT +TGCACC +TGCACG +TGCAGA +TGCAGT +TGCAGC +TGCAGG +TGCTAA +TGCTAT +TGCTAC +TGCTAG +TGCTTA +TGCTTT +TGCTTC +TGCTTG +TGCTCA +TGCTCT +TGCTCC +TGCTCG +TGCTGA +TGCTGT +TGCTGC +TGCTGG +TGCCAA +TGCCAT +TGCCAC +TGCCAG +TGCCTA +TGCCTT +TGCCTC +TGCCTG +TGCCCA +TGCCCT +TGCCCC +TGCCCG +TGCCGA +TGCCGT +TGCCGC +TGCCGG +TGCGAA +TGCGAT +TGCGAC +TGCGAG +TGCGTA +TGCGTT +TGCGTC +TGCGTG +TGCGCA +TGCGCT +TGCGCC +TGCGCG +TGCGGA +TGCGGT +TGCGGC +TGCGGG +TGGAAA +TGGAAT +TGGAAC +TGGAAG +TGGATA +TGGATT +TGGATC +TGGATG +TGGACA +TGGACT +TGGACC +TGGACG +TGGAGA +TGGAGT +TGGAGC +TGGAGG +TGGTAA +TGGTAT +TGGTAC +TGGTAG +TGGTTA +TGGTTT +TGGTTC +TGGTTG +TGGTCA +TGGTCT +TGGTCC +TGGTCG +TGGTGA +TGGTGT +TGGTGC +TGGTGG +TGGCAA +TGGCAT +TGGCAC +TGGCAG +TGGCTA +TGGCTT +TGGCTC +TGGCTG +TGGCCA +TGGCCT +TGGCCC +TGGCCG +TGGCGA +TGGCGT +TGGCGC +TGGCGG +TGGGAA +TGGGAT +TGGGAC +TGGGAG +TGGGTA +TGGGTT +TGGGTC +TGGGTG +TGGGCA +TGGGCT +TGGGCC +TGGGCG +TGGGGA +TGGGGT +TGGGGC +TGGGGG +CAAAAA +CAAAAT +CAAAAC +CAAAAG +CAAATA +CAAATT +CAAATC +CAAATG +CAAACA +CAAACT +CAAACC +CAAACG +CAAAGA +CAAAGT +CAAAGC +CAAAGG +CAATAA +CAATAT +CAATAC +CAATAG +CAATTA +CAATTT +CAATTC +CAATTG +CAATCA +CAATCT +CAATCC +CAATCG +CAATGA +CAATGT +CAATGC +CAATGG +CAACAA +CAACAT +CAACAC +CAACAG +CAACTA +CAACTT +CAACTC +CAACTG +CAACCA +CAACCT +CAACCC +CAACCG +CAACGA +CAACGT +CAACGC +CAACGG +CAAGAA +CAAGAT +CAAGAC +CAAGAG +CAAGTA +CAAGTT +CAAGTC +CAAGTG +CAAGCA +CAAGCT +CAAGCC +CAAGCG +CAAGGA +CAAGGT +CAAGGC +CAAGGG +CATAAA +CATAAT +CATAAC +CATAAG +CATATA +CATATT +CATATC +CATATG +CATACA +CATACT +CATACC +CATACG +CATAGA +CATAGT +CATAGC +CATAGG +CATTAA +CATTAT +CATTAC +CATTAG +CATTTA +CATTTT +CATTTC +CATTTG +CATTCA +CATTCT +CATTCC +CATTCG +CATTGA +CATTGT +CATTGC +CATTGG +CATCAA +CATCAT +CATCAC +CATCAG +CATCTA +CATCTT +CATCTC +CATCTG +CATCCA +CATCCT +CATCCC +CATCCG +CATCGA +CATCGT +CATCGC +CATCGG +CATGAA +CATGAT +CATGAC +CATGAG +CATGTA +CATGTT +CATGTC +CATGTG +CATGCA +CATGCT +CATGCC +CATGCG +CATGGA +CATGGT +CATGGC +CATGGG +CACAAA +CACAAT +CACAAC +CACAAG +CACATA +CACATT +CACATC +CACATG +CACACA +CACACT +CACACC +CACACG +CACAGA +CACAGT +CACAGC +CACAGG +CACTAA +CACTAT +CACTAC +CACTAG +CACTTA +CACTTT +CACTTC +CACTTG +CACTCA +CACTCT +CACTCC +CACTCG +CACTGA +CACTGT +CACTGC +CACTGG +CACCAA +CACCAT +CACCAC +CACCAG +CACCTA +CACCTT +CACCTC +CACCTG +CACCCA +CACCCT +CACCCC +CACCCG +CACCGA +CACCGT +CACCGC +CACCGG +CACGAA +CACGAT +CACGAC +CACGAG +CACGTA +CACGTT +CACGTC +CACGTG +CACGCA +CACGCT +CACGCC +CACGCG +CACGGA +CACGGT +CACGGC +CACGGG +CAGAAA +CAGAAT +CAGAAC +CAGAAG +CAGATA +CAGATT +CAGATC +CAGATG +CAGACA +CAGACT +CAGACC +CAGACG +CAGAGA +CAGAGT +CAGAGC +CAGAGG +CAGTAA +CAGTAT +CAGTAC +CAGTAG +CAGTTA +CAGTTT +CAGTTC +CAGTTG +CAGTCA +CAGTCT +CAGTCC +CAGTCG +CAGTGA +CAGTGT +CAGTGC +CAGTGG +CAGCAA +CAGCAT +CAGCAC +CAGCAG +CAGCTA +CAGCTT +CAGCTC +CAGCTG +CAGCCA +CAGCCT +CAGCCC +CAGCCG +CAGCGA +CAGCGT +CAGCGC +CAGCGG +CAGGAA +CAGGAT +CAGGAC +CAGGAG +CAGGTA +CAGGTT +CAGGTC +CAGGTG +CAGGCA +CAGGCT +CAGGCC +CAGGCG +CAGGGA +CAGGGT +CAGGGC +CAGGGG +CTAAAA +CTAAAT +CTAAAC +CTAAAG +CTAATA +CTAATT +CTAATC +CTAATG +CTAACA +CTAACT +CTAACC +CTAACG +CTAAGA +CTAAGT +CTAAGC +CTAAGG +CTATAA +CTATAT +CTATAC +CTATAG +CTATTA +CTATTT +CTATTC +CTATTG +CTATCA +CTATCT +CTATCC +CTATCG +CTATGA +CTATGT +CTATGC +CTATGG +CTACAA +CTACAT +CTACAC +CTACAG +CTACTA +CTACTT +CTACTC +CTACTG +CTACCA +CTACCT +CTACCC +CTACCG +CTACGA +CTACGT +CTACGC +CTACGG +CTAGAA +CTAGAT +CTAGAC +CTAGAG +CTAGTA +CTAGTT +CTAGTC +CTAGTG +CTAGCA +CTAGCT +CTAGCC +CTAGCG +CTAGGA +CTAGGT +CTAGGC +CTAGGG +CTTAAA +CTTAAT +CTTAAC +CTTAAG +CTTATA +CTTATT +CTTATC +CTTATG +CTTACA +CTTACT +CTTACC +CTTACG +CTTAGA +CTTAGT +CTTAGC +CTTAGG +CTTTAA +CTTTAT +CTTTAC +CTTTAG +CTTTTA +CTTTTT +CTTTTC +CTTTTG +CTTTCA +CTTTCT +CTTTCC +CTTTCG +CTTTGA +CTTTGT +CTTTGC +CTTTGG +CTTCAA +CTTCAT +CTTCAC +CTTCAG +CTTCTA +CTTCTT +CTTCTC +CTTCTG +CTTCCA +CTTCCT +CTTCCC +CTTCCG +CTTCGA +CTTCGT +CTTCGC +CTTCGG +CTTGAA +CTTGAT +CTTGAC +CTTGAG +CTTGTA +CTTGTT +CTTGTC +CTTGTG +CTTGCA +CTTGCT +CTTGCC +CTTGCG +CTTGGA +CTTGGT +CTTGGC +CTTGGG +CTCAAA +CTCAAT +CTCAAC +CTCAAG +CTCATA +CTCATT +CTCATC +CTCATG +CTCACA +CTCACT +CTCACC +CTCACG +CTCAGA +CTCAGT +CTCAGC +CTCAGG +CTCTAA +CTCTAT +CTCTAC +CTCTAG +CTCTTA +CTCTTT +CTCTTC +CTCTTG +CTCTCA +CTCTCT +CTCTCC +CTCTCG +CTCTGA +CTCTGT +CTCTGC +CTCTGG +CTCCAA +CTCCAT +CTCCAC +CTCCAG +CTCCTA +CTCCTT +CTCCTC +CTCCTG +CTCCCA +CTCCCT +CTCCCC +CTCCCG +CTCCGA +CTCCGT +CTCCGC +CTCCGG +CTCGAA +CTCGAT +CTCGAC +CTCGAG +CTCGTA +CTCGTT +CTCGTC +CTCGTG +CTCGCA +CTCGCT +CTCGCC +CTCGCG +CTCGGA +CTCGGT +CTCGGC +CTCGGG +CTGAAA +CTGAAT +CTGAAC +CTGAAG +CTGATA +CTGATT +CTGATC +CTGATG +CTGACA +CTGACT +CTGACC +CTGACG +CTGAGA +CTGAGT +CTGAGC +CTGAGG +CTGTAA +CTGTAT +CTGTAC +CTGTAG +CTGTTA +CTGTTT +CTGTTC +CTGTTG +CTGTCA +CTGTCT +CTGTCC +CTGTCG +CTGTGA +CTGTGT +CTGTGC +CTGTGG +CTGCAA +CTGCAT +CTGCAC +CTGCAG +CTGCTA +CTGCTT +CTGCTC +CTGCTG +CTGCCA +CTGCCT +CTGCCC +CTGCCG +CTGCGA +CTGCGT +CTGCGC +CTGCGG +CTGGAA +CTGGAT +CTGGAC +CTGGAG +CTGGTA +CTGGTT +CTGGTC +CTGGTG +CTGGCA +CTGGCT +CTGGCC +CTGGCG +CTGGGA +CTGGGT +CTGGGC +CTGGGG +CCAAAA +CCAAAT +CCAAAC +CCAAAG +CCAATA +CCAATT +CCAATC +CCAATG +CCAACA +CCAACT +CCAACC +CCAACG +CCAAGA +CCAAGT +CCAAGC +CCAAGG +CCATAA +CCATAT +CCATAC +CCATAG +CCATTA +CCATTT +CCATTC +CCATTG +CCATCA +CCATCT +CCATCC +CCATCG +CCATGA +CCATGT +CCATGC +CCATGG +CCACAA +CCACAT +CCACAC +CCACAG +CCACTA +CCACTT +CCACTC +CCACTG +CCACCA +CCACCT +CCACCC +CCACCG +CCACGA +CCACGT +CCACGC +CCACGG +CCAGAA +CCAGAT +CCAGAC +CCAGAG +CCAGTA +CCAGTT +CCAGTC +CCAGTG +CCAGCA +CCAGCT +CCAGCC +CCAGCG +CCAGGA +CCAGGT +CCAGGC +CCAGGG +CCTAAA +CCTAAT +CCTAAC +CCTAAG +CCTATA +CCTATT +CCTATC +CCTATG +CCTACA +CCTACT +CCTACC +CCTACG +CCTAGA +CCTAGT +CCTAGC +CCTAGG +CCTTAA +CCTTAT +CCTTAC +CCTTAG +CCTTTA +CCTTTT +CCTTTC +CCTTTG +CCTTCA +CCTTCT +CCTTCC +CCTTCG +CCTTGA +CCTTGT +CCTTGC +CCTTGG +CCTCAA +CCTCAT +CCTCAC +CCTCAG +CCTCTA +CCTCTT +CCTCTC +CCTCTG +CCTCCA +CCTCCT +CCTCCC +CCTCCG +CCTCGA +CCTCGT +CCTCGC +CCTCGG +CCTGAA +CCTGAT +CCTGAC +CCTGAG +CCTGTA +CCTGTT +CCTGTC +CCTGTG +CCTGCA +CCTGCT +CCTGCC +CCTGCG +CCTGGA +CCTGGT +CCTGGC +CCTGGG +CCCAAA +CCCAAT +CCCAAC +CCCAAG +CCCATA +CCCATT +CCCATC +CCCATG +CCCACA +CCCACT +CCCACC +CCCACG +CCCAGA +CCCAGT +CCCAGC +CCCAGG +CCCTAA +CCCTAT +CCCTAC +CCCTAG +CCCTTA +CCCTTT +CCCTTC +CCCTTG +CCCTCA +CCCTCT +CCCTCC +CCCTCG +CCCTGA +CCCTGT +CCCTGC +CCCTGG +CCCCAA +CCCCAT +CCCCAC +CCCCAG +CCCCTA +CCCCTT +CCCCTC +CCCCTG +CCCCCA +CCCCCT +CCCCCC +CCCCCG +CCCCGA +CCCCGT +CCCCGC +CCCCGG +CCCGAA +CCCGAT +CCCGAC +CCCGAG +CCCGTA +CCCGTT +CCCGTC +CCCGTG +CCCGCA +CCCGCT +CCCGCC +CCCGCG +CCCGGA +CCCGGT +CCCGGC +CCCGGG +CCGAAA +CCGAAT +CCGAAC +CCGAAG +CCGATA +CCGATT +CCGATC +CCGATG +CCGACA +CCGACT +CCGACC +CCGACG +CCGAGA +CCGAGT +CCGAGC +CCGAGG +CCGTAA +CCGTAT +CCGTAC +CCGTAG +CCGTTA +CCGTTT +CCGTTC +CCGTTG +CCGTCA +CCGTCT +CCGTCC +CCGTCG +CCGTGA +CCGTGT +CCGTGC +CCGTGG +CCGCAA +CCGCAT +CCGCAC +CCGCAG +CCGCTA +CCGCTT +CCGCTC +CCGCTG +CCGCCA +CCGCCT +CCGCCC +CCGCCG +CCGCGA +CCGCGT +CCGCGC +CCGCGG +CCGGAA +CCGGAT +CCGGAC +CCGGAG +CCGGTA +CCGGTT +CCGGTC +CCGGTG +CCGGCA +CCGGCT +CCGGCC +CCGGCG +CCGGGA +CCGGGT +CCGGGC +CCGGGG +CGAAAA +CGAAAT +CGAAAC +CGAAAG +CGAATA +CGAATT +CGAATC +CGAATG +CGAACA +CGAACT +CGAACC +CGAACG +CGAAGA +CGAAGT +CGAAGC +CGAAGG +CGATAA +CGATAT +CGATAC +CGATAG +CGATTA +CGATTT +CGATTC +CGATTG +CGATCA +CGATCT +CGATCC +CGATCG +CGATGA +CGATGT +CGATGC +CGATGG +CGACAA +CGACAT +CGACAC +CGACAG +CGACTA +CGACTT +CGACTC +CGACTG +CGACCA +CGACCT +CGACCC +CGACCG +CGACGA +CGACGT +CGACGC +CGACGG +CGAGAA +CGAGAT +CGAGAC +CGAGAG +CGAGTA +CGAGTT +CGAGTC +CGAGTG +CGAGCA +CGAGCT +CGAGCC +CGAGCG +CGAGGA +CGAGGT +CGAGGC +CGAGGG +CGTAAA +CGTAAT +CGTAAC +CGTAAG +CGTATA +CGTATT +CGTATC +CGTATG +CGTACA +CGTACT +CGTACC +CGTACG +CGTAGA +CGTAGT +CGTAGC +CGTAGG +CGTTAA +CGTTAT +CGTTAC +CGTTAG +CGTTTA +CGTTTT +CGTTTC +CGTTTG +CGTTCA +CGTTCT +CGTTCC +CGTTCG +CGTTGA +CGTTGT +CGTTGC +CGTTGG +CGTCAA +CGTCAT +CGTCAC +CGTCAG +CGTCTA +CGTCTT +CGTCTC +CGTCTG +CGTCCA +CGTCCT +CGTCCC +CGTCCG +CGTCGA +CGTCGT +CGTCGC +CGTCGG +CGTGAA +CGTGAT +CGTGAC +CGTGAG +CGTGTA +CGTGTT +CGTGTC +CGTGTG +CGTGCA +CGTGCT +CGTGCC +CGTGCG +CGTGGA +CGTGGT +CGTGGC +CGTGGG +CGCAAA +CGCAAT +CGCAAC +CGCAAG +CGCATA +CGCATT +CGCATC +CGCATG +CGCACA +CGCACT +CGCACC +CGCACG +CGCAGA +CGCAGT +CGCAGC +CGCAGG +CGCTAA +CGCTAT +CGCTAC +CGCTAG +CGCTTA +CGCTTT +CGCTTC +CGCTTG +CGCTCA +CGCTCT +CGCTCC +CGCTCG +CGCTGA +CGCTGT +CGCTGC +CGCTGG +CGCCAA +CGCCAT +CGCCAC +CGCCAG +CGCCTA +CGCCTT +CGCCTC +CGCCTG +CGCCCA +CGCCCT +CGCCCC +CGCCCG +CGCCGA +CGCCGT +CGCCGC +CGCCGG +CGCGAA +CGCGAT +CGCGAC +CGCGAG +CGCGTA +CGCGTT +CGCGTC +CGCGTG +CGCGCA +CGCGCT +CGCGCC +CGCGCG +CGCGGA +CGCGGT +CGCGGC +CGCGGG +CGGAAA +CGGAAT +CGGAAC +CGGAAG +CGGATA +CGGATT +CGGATC +CGGATG +CGGACA +CGGACT +CGGACC +CGGACG +CGGAGA +CGGAGT +CGGAGC +CGGAGG +CGGTAA +CGGTAT +CGGTAC +CGGTAG +CGGTTA +CGGTTT +CGGTTC +CGGTTG +CGGTCA +CGGTCT +CGGTCC +CGGTCG +CGGTGA +CGGTGT +CGGTGC +CGGTGG +CGGCAA +CGGCAT +CGGCAC +CGGCAG +CGGCTA +CGGCTT +CGGCTC +CGGCTG +CGGCCA +CGGCCT +CGGCCC +CGGCCG +CGGCGA +CGGCGT +CGGCGC +CGGCGG +CGGGAA +CGGGAT +CGGGAC +CGGGAG +CGGGTA +CGGGTT +CGGGTC +CGGGTG +CGGGCA +CGGGCT +CGGGCC +CGGGCG +CGGGGA +CGGGGT +CGGGGC +CGGGGG +GAAAAA +GAAAAT +GAAAAC +GAAAAG +GAAATA +GAAATT +GAAATC +GAAATG +GAAACA +GAAACT +GAAACC +GAAACG +GAAAGA +GAAAGT +GAAAGC +GAAAGG +GAATAA +GAATAT +GAATAC +GAATAG +GAATTA +GAATTT +GAATTC +GAATTG +GAATCA +GAATCT +GAATCC +GAATCG +GAATGA +GAATGT +GAATGC +GAATGG +GAACAA +GAACAT +GAACAC +GAACAG +GAACTA +GAACTT +GAACTC +GAACTG +GAACCA +GAACCT +GAACCC +GAACCG +GAACGA +GAACGT +GAACGC +GAACGG +GAAGAA +GAAGAT +GAAGAC +GAAGAG +GAAGTA +GAAGTT +GAAGTC +GAAGTG +GAAGCA +GAAGCT +GAAGCC +GAAGCG +GAAGGA +GAAGGT +GAAGGC +GAAGGG +GATAAA +GATAAT +GATAAC +GATAAG +GATATA +GATATT +GATATC +GATATG +GATACA +GATACT +GATACC +GATACG +GATAGA +GATAGT +GATAGC +GATAGG +GATTAA +GATTAT +GATTAC +GATTAG +GATTTA +GATTTT +GATTTC +GATTTG +GATTCA +GATTCT +GATTCC +GATTCG +GATTGA +GATTGT +GATTGC +GATTGG +GATCAA +GATCAT +GATCAC +GATCAG +GATCTA +GATCTT +GATCTC +GATCTG +GATCCA +GATCCT +GATCCC +GATCCG +GATCGA +GATCGT +GATCGC +GATCGG +GATGAA +GATGAT +GATGAC +GATGAG +GATGTA +GATGTT +GATGTC +GATGTG +GATGCA +GATGCT +GATGCC +GATGCG +GATGGA +GATGGT +GATGGC +GATGGG +GACAAA +GACAAT +GACAAC +GACAAG +GACATA +GACATT +GACATC +GACATG +GACACA +GACACT +GACACC +GACACG +GACAGA +GACAGT +GACAGC +GACAGG +GACTAA +GACTAT +GACTAC +GACTAG +GACTTA +GACTTT +GACTTC +GACTTG +GACTCA +GACTCT +GACTCC +GACTCG +GACTGA +GACTGT +GACTGC +GACTGG +GACCAA +GACCAT +GACCAC +GACCAG +GACCTA +GACCTT +GACCTC +GACCTG +GACCCA +GACCCT +GACCCC +GACCCG +GACCGA +GACCGT +GACCGC +GACCGG +GACGAA +GACGAT +GACGAC +GACGAG +GACGTA +GACGTT +GACGTC +GACGTG +GACGCA +GACGCT +GACGCC +GACGCG +GACGGA +GACGGT +GACGGC +GACGGG +GAGAAA +GAGAAT +GAGAAC +GAGAAG +GAGATA +GAGATT +GAGATC +GAGATG +GAGACA +GAGACT +GAGACC +GAGACG +GAGAGA +GAGAGT +GAGAGC +GAGAGG +GAGTAA +GAGTAT +GAGTAC +GAGTAG +GAGTTA +GAGTTT +GAGTTC +GAGTTG +GAGTCA +GAGTCT +GAGTCC +GAGTCG +GAGTGA +GAGTGT +GAGTGC +GAGTGG +GAGCAA +GAGCAT +GAGCAC +GAGCAG +GAGCTA +GAGCTT +GAGCTC +GAGCTG +GAGCCA +GAGCCT +GAGCCC +GAGCCG +GAGCGA +GAGCGT +GAGCGC +GAGCGG +GAGGAA +GAGGAT +GAGGAC +GAGGAG +GAGGTA +GAGGTT +GAGGTC +GAGGTG +GAGGCA +GAGGCT +GAGGCC +GAGGCG +GAGGGA +GAGGGT +GAGGGC +GAGGGG +GTAAAA +GTAAAT +GTAAAC +GTAAAG +GTAATA +GTAATT +GTAATC +GTAATG +GTAACA +GTAACT +GTAACC +GTAACG +GTAAGA +GTAAGT +GTAAGC +GTAAGG +GTATAA +GTATAT +GTATAC +GTATAG +GTATTA +GTATTT +GTATTC +GTATTG +GTATCA +GTATCT +GTATCC +GTATCG +GTATGA +GTATGT +GTATGC +GTATGG +GTACAA +GTACAT +GTACAC +GTACAG +GTACTA +GTACTT +GTACTC +GTACTG +GTACCA +GTACCT +GTACCC +GTACCG +GTACGA +GTACGT +GTACGC +GTACGG +GTAGAA +GTAGAT +GTAGAC +GTAGAG +GTAGTA +GTAGTT +GTAGTC +GTAGTG +GTAGCA +GTAGCT +GTAGCC +GTAGCG +GTAGGA +GTAGGT +GTAGGC +GTAGGG +GTTAAA +GTTAAT +GTTAAC +GTTAAG +GTTATA +GTTATT +GTTATC +GTTATG +GTTACA +GTTACT +GTTACC +GTTACG +GTTAGA +GTTAGT +GTTAGC +GTTAGG +GTTTAA +GTTTAT +GTTTAC +GTTTAG +GTTTTA +GTTTTT +GTTTTC +GTTTTG +GTTTCA +GTTTCT +GTTTCC +GTTTCG +GTTTGA +GTTTGT +GTTTGC +GTTTGG +GTTCAA +GTTCAT +GTTCAC +GTTCAG +GTTCTA +GTTCTT +GTTCTC +GTTCTG +GTTCCA +GTTCCT +GTTCCC +GTTCCG +GTTCGA +GTTCGT +GTTCGC +GTTCGG +GTTGAA +GTTGAT +GTTGAC +GTTGAG +GTTGTA +GTTGTT +GTTGTC +GTTGTG +GTTGCA +GTTGCT +GTTGCC +GTTGCG +GTTGGA +GTTGGT +GTTGGC +GTTGGG +GTCAAA +GTCAAT +GTCAAC +GTCAAG +GTCATA +GTCATT +GTCATC +GTCATG +GTCACA +GTCACT +GTCACC +GTCACG +GTCAGA +GTCAGT +GTCAGC +GTCAGG +GTCTAA +GTCTAT +GTCTAC +GTCTAG +GTCTTA +GTCTTT +GTCTTC +GTCTTG +GTCTCA +GTCTCT +GTCTCC +GTCTCG +GTCTGA +GTCTGT +GTCTGC +GTCTGG +GTCCAA +GTCCAT +GTCCAC +GTCCAG +GTCCTA +GTCCTT +GTCCTC +GTCCTG +GTCCCA +GTCCCT +GTCCCC +GTCCCG +GTCCGA +GTCCGT +GTCCGC +GTCCGG +GTCGAA +GTCGAT +GTCGAC +GTCGAG +GTCGTA +GTCGTT +GTCGTC +GTCGTG +GTCGCA +GTCGCT +GTCGCC +GTCGCG +GTCGGA +GTCGGT +GTCGGC +GTCGGG +GTGAAA +GTGAAT +GTGAAC +GTGAAG +GTGATA +GTGATT +GTGATC +GTGATG +GTGACA +GTGACT +GTGACC +GTGACG +GTGAGA +GTGAGT +GTGAGC +GTGAGG +GTGTAA +GTGTAT +GTGTAC +GTGTAG +GTGTTA +GTGTTT +GTGTTC +GTGTTG +GTGTCA +GTGTCT +GTGTCC +GTGTCG +GTGTGA +GTGTGT +GTGTGC +GTGTGG +GTGCAA +GTGCAT +GTGCAC +GTGCAG +GTGCTA +GTGCTT +GTGCTC +GTGCTG +GTGCCA +GTGCCT +GTGCCC +GTGCCG +GTGCGA +GTGCGT +GTGCGC +GTGCGG +GTGGAA +GTGGAT +GTGGAC +GTGGAG +GTGGTA +GTGGTT +GTGGTC +GTGGTG +GTGGCA +GTGGCT +GTGGCC +GTGGCG +GTGGGA +GTGGGT +GTGGGC +GTGGGG +GCAAAA +GCAAAT +GCAAAC +GCAAAG +GCAATA +GCAATT +GCAATC +GCAATG +GCAACA +GCAACT +GCAACC +GCAACG +GCAAGA +GCAAGT +GCAAGC +GCAAGG +GCATAA +GCATAT +GCATAC +GCATAG +GCATTA +GCATTT +GCATTC +GCATTG +GCATCA +GCATCT +GCATCC +GCATCG +GCATGA +GCATGT +GCATGC +GCATGG +GCACAA +GCACAT +GCACAC +GCACAG +GCACTA +GCACTT +GCACTC +GCACTG +GCACCA +GCACCT +GCACCC +GCACCG +GCACGA +GCACGT +GCACGC +GCACGG +GCAGAA +GCAGAT +GCAGAC +GCAGAG +GCAGTA +GCAGTT +GCAGTC +GCAGTG +GCAGCA +GCAGCT +GCAGCC +GCAGCG +GCAGGA +GCAGGT +GCAGGC +GCAGGG +GCTAAA +GCTAAT +GCTAAC +GCTAAG +GCTATA +GCTATT +GCTATC +GCTATG +GCTACA +GCTACT +GCTACC +GCTACG +GCTAGA +GCTAGT +GCTAGC +GCTAGG +GCTTAA +GCTTAT +GCTTAC +GCTTAG +GCTTTA +GCTTTT +GCTTTC +GCTTTG +GCTTCA +GCTTCT +GCTTCC +GCTTCG +GCTTGA +GCTTGT +GCTTGC +GCTTGG +GCTCAA +GCTCAT +GCTCAC +GCTCAG +GCTCTA +GCTCTT +GCTCTC +GCTCTG +GCTCCA +GCTCCT +GCTCCC +GCTCCG +GCTCGA +GCTCGT +GCTCGC +GCTCGG +GCTGAA +GCTGAT +GCTGAC +GCTGAG +GCTGTA +GCTGTT +GCTGTC +GCTGTG +GCTGCA +GCTGCT +GCTGCC +GCTGCG +GCTGGA +GCTGGT +GCTGGC +GCTGGG +GCCAAA +GCCAAT +GCCAAC +GCCAAG +GCCATA +GCCATT +GCCATC +GCCATG +GCCACA +GCCACT +GCCACC +GCCACG +GCCAGA +GCCAGT +GCCAGC +GCCAGG +GCCTAA +GCCTAT +GCCTAC +GCCTAG +GCCTTA +GCCTTT +GCCTTC +GCCTTG +GCCTCA +GCCTCT +GCCTCC +GCCTCG +GCCTGA +GCCTGT +GCCTGC +GCCTGG +GCCCAA +GCCCAT +GCCCAC +GCCCAG +GCCCTA +GCCCTT +GCCCTC +GCCCTG +GCCCCA +GCCCCT +GCCCCC +GCCCCG +GCCCGA +GCCCGT +GCCCGC +GCCCGG +GCCGAA +GCCGAT +GCCGAC +GCCGAG +GCCGTA +GCCGTT +GCCGTC +GCCGTG +GCCGCA +GCCGCT +GCCGCC +GCCGCG +GCCGGA +GCCGGT +GCCGGC +GCCGGG +GCGAAA +GCGAAT +GCGAAC +GCGAAG +GCGATA +GCGATT +GCGATC +GCGATG +GCGACA +GCGACT +GCGACC +GCGACG +GCGAGA +GCGAGT +GCGAGC +GCGAGG +GCGTAA +GCGTAT +GCGTAC +GCGTAG +GCGTTA +GCGTTT +GCGTTC +GCGTTG +GCGTCA +GCGTCT +GCGTCC +GCGTCG +GCGTGA +GCGTGT +GCGTGC +GCGTGG +GCGCAA +GCGCAT +GCGCAC +GCGCAG +GCGCTA +GCGCTT +GCGCTC +GCGCTG +GCGCCA +GCGCCT +GCGCCC +GCGCCG +GCGCGA +GCGCGT +GCGCGC +GCGCGG +GCGGAA +GCGGAT +GCGGAC +GCGGAG +GCGGTA +GCGGTT +GCGGTC +GCGGTG +GCGGCA +GCGGCT +GCGGCC +GCGGCG +GCGGGA +GCGGGT +GCGGGC +GCGGGG +GGAAAA +GGAAAT +GGAAAC +GGAAAG +GGAATA +GGAATT +GGAATC +GGAATG +GGAACA +GGAACT +GGAACC +GGAACG +GGAAGA +GGAAGT +GGAAGC +GGAAGG +GGATAA +GGATAT +GGATAC +GGATAG +GGATTA +GGATTT +GGATTC +GGATTG +GGATCA +GGATCT +GGATCC +GGATCG +GGATGA +GGATGT +GGATGC +GGATGG +GGACAA +GGACAT +GGACAC +GGACAG +GGACTA +GGACTT +GGACTC +GGACTG +GGACCA +GGACCT +GGACCC +GGACCG +GGACGA +GGACGT +GGACGC +GGACGG +GGAGAA +GGAGAT +GGAGAC +GGAGAG +GGAGTA +GGAGTT +GGAGTC +GGAGTG +GGAGCA +GGAGCT +GGAGCC +GGAGCG +GGAGGA +GGAGGT +GGAGGC +GGAGGG +GGTAAA +GGTAAT +GGTAAC +GGTAAG +GGTATA +GGTATT +GGTATC +GGTATG +GGTACA +GGTACT +GGTACC +GGTACG +GGTAGA +GGTAGT +GGTAGC +GGTAGG +GGTTAA +GGTTAT +GGTTAC +GGTTAG +GGTTTA +GGTTTT +GGTTTC +GGTTTG +GGTTCA +GGTTCT +GGTTCC +GGTTCG +GGTTGA +GGTTGT +GGTTGC +GGTTGG +GGTCAA +GGTCAT +GGTCAC +GGTCAG +GGTCTA +GGTCTT +GGTCTC +GGTCTG +GGTCCA +GGTCCT +GGTCCC +GGTCCG +GGTCGA +GGTCGT +GGTCGC +GGTCGG +GGTGAA +GGTGAT +GGTGAC +GGTGAG +GGTGTA +GGTGTT +GGTGTC +GGTGTG +GGTGCA +GGTGCT +GGTGCC +GGTGCG +GGTGGA +GGTGGT +GGTGGC +GGTGGG +GGCAAA +GGCAAT +GGCAAC +GGCAAG +GGCATA +GGCATT +GGCATC +GGCATG +GGCACA +GGCACT +GGCACC +GGCACG +GGCAGA +GGCAGT +GGCAGC +GGCAGG +GGCTAA +GGCTAT +GGCTAC +GGCTAG +GGCTTA +GGCTTT +GGCTTC +GGCTTG +GGCTCA +GGCTCT +GGCTCC +GGCTCG +GGCTGA +GGCTGT +GGCTGC +GGCTGG +GGCCAA +GGCCAT +GGCCAC +GGCCAG +GGCCTA +GGCCTT +GGCCTC +GGCCTG +GGCCCA +GGCCCT +GGCCCC +GGCCCG +GGCCGA +GGCCGT +GGCCGC +GGCCGG +GGCGAA +GGCGAT +GGCGAC +GGCGAG +GGCGTA +GGCGTT +GGCGTC +GGCGTG +GGCGCA +GGCGCT +GGCGCC +GGCGCG +GGCGGA +GGCGGT +GGCGGC +GGCGGG +GGGAAA +GGGAAT +GGGAAC +GGGAAG +GGGATA +GGGATT +GGGATC +GGGATG +GGGACA +GGGACT +GGGACC +GGGACG +GGGAGA +GGGAGT +GGGAGC +GGGAGG +GGGTAA +GGGTAT +GGGTAC +GGGTAG +GGGTTA +GGGTTT +GGGTTC +GGGTTG +GGGTCA +GGGTCT +GGGTCC +GGGTCG +GGGTGA +GGGTGT +GGGTGC +GGGTGG +GGGCAA +GGGCAT +GGGCAC +GGGCAG +GGGCTA +GGGCTT +GGGCTC +GGGCTG +GGGCCA +GGGCCT +GGGCCC +GGGCCG +GGGCGA +GGGCGT +GGGCGC +GGGCGG +GGGGAA +GGGGAT +GGGGAC +GGGGAG +GGGGTA +GGGGTT +GGGGTC +GGGGTG +GGGGCA +GGGGCT +GGGGCC +GGGGCG +GGGGGA +GGGGGT +GGGGGC +GGGGGG diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6250/config.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6250/config.json new file mode 100644 index 0000000000000000000000000000000000000000..9b14e61d327bf586fcff6e817cf9d365bae18ad2 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6250/config.json @@ -0,0 +1,26 @@ +{ + "_name_or_path": "fabihamakhdoomi/TinyDNABERT", + "architectures": [ + "BertForSequenceClassification" + ], + "attention_probs_dropout_prob": 0.1, + "classifier_dropout": null, + "hidden_act": "gelu", + "hidden_dropout_prob": 0.1, + "hidden_size": 128, + "initializer_range": 0.02, + "intermediate_size": 512, + "layer_norm_eps": 1e-12, + "max_position_embeddings": 512, + "model_type": "bert", + "num_attention_heads": 2, + "num_hidden_layers": 2, + "pad_token_id": 0, + "position_embedding_type": "absolute", + "problem_type": "single_label_classification", + "torch_dtype": "float32", + "transformers_version": "4.44.0", + "type_vocab_size": 2, + "use_cache": true, + "vocab_size": 4101 +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6250/model.safetensors b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6250/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..2e7dad0d964e750562229653a966906313acd014 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6250/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b9c1dbd7e2fc2948c959cb33a33c36087f7ca8ecbc39af3eee0d8ff559b587a2 +size 4021688 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6250/optimizer.pt b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6250/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..54058ffbaac1a17d4c0699085283ae40f4a721ad --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6250/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f984a85346cd8eb29376410d07fd344291853bb91b01203a899fefd9ae530134 +size 8058700 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6250/rng_state.pth b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6250/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..e666d96a1e7d953ad715b4f92cb4f8713cf02642 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6250/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:539fe12b01f7e37cbd1c254718c50ce0226ac2f3489b5b349456040c507b4a44 +size 14244 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6250/scheduler.pt b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6250/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..4c34d9dcafc0eb1507a84da1af7c2c5228be082e --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6250/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:20611d193eaf6658e6192005c679e5aaa561ac33254f734047a2032edc3e8f9b +size 1064 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6250/special_tokens_map.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6250/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..a8b3208c2884c4efb86e49300fdd3dc877220cdf --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6250/special_tokens_map.json @@ -0,0 +1,7 @@ +{ + "cls_token": "[CLS]", + "mask_token": "[MASK]", + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6250/tokenizer.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6250/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..ebd782de6963b9032ce9374cfe30f5275be1a30b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6250/tokenizer.json @@ -0,0 +1,4265 @@ +{ + "version": "1.0", + "truncation": { + "direction": "Right", + "max_length": 512, + "strategy": "LongestFirst", + "stride": 0 + }, + "padding": { + "strategy": { + "Fixed": 512 + }, + "direction": "Right", + "pad_to_multiple_of": null, + "pad_id": 0, + "pad_type_id": 0, + "pad_token": "[PAD]" + }, + "added_tokens": [ + { + "id": 0, + "content": "[PAD]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 1, + "content": "[UNK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 2, + "content": "[CLS]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 3, + "content": "[SEP]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 4, + "content": "[MASK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + } + ], + "normalizer": { + "type": "BertNormalizer", + "clean_text": true, + "handle_chinese_chars": true, + "strip_accents": null, + "lowercase": false + }, + "pre_tokenizer": { + "type": "BertPreTokenizer" + }, + "post_processor": { + "type": "TemplateProcessing", + "single": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + } + ], + "pair": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "B", + "type_id": 1 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 1 + } + } + ], + "special_tokens": { + "[CLS]": { + "id": "[CLS]", + "ids": [ + 2 + ], + "tokens": [ + "[CLS]" + ] + }, + "[SEP]": { + "id": "[SEP]", + "ids": [ + 3 + ], + "tokens": [ + "[SEP]" + ] + } + } + }, + "decoder": { + "type": "WordPiece", + "prefix": "##", + "cleanup": true + }, + "model": { + "type": "WordPiece", + "unk_token": "[UNK]", + "continuing_subword_prefix": "##", + "max_input_chars_per_word": 100, + "vocab": { + "[PAD]": 0, + "[UNK]": 1, + "[CLS]": 2, + "[SEP]": 3, + "[MASK]": 4, + "AAAAAA": 5, + "AAAAAT": 6, + "AAAAAC": 7, + "AAAAAG": 8, + "AAAATA": 9, + "AAAATT": 10, + "AAAATC": 11, + "AAAATG": 12, + "AAAACA": 13, + "AAAACT": 14, + "AAAACC": 15, + "AAAACG": 16, + "AAAAGA": 17, + "AAAAGT": 18, + "AAAAGC": 19, + "AAAAGG": 20, + "AAATAA": 21, + "AAATAT": 22, + "AAATAC": 23, + "AAATAG": 24, + "AAATTA": 25, + "AAATTT": 26, + "AAATTC": 27, + "AAATTG": 28, + "AAATCA": 29, + "AAATCT": 30, + "AAATCC": 31, + "AAATCG": 32, + "AAATGA": 33, + "AAATGT": 34, + "AAATGC": 35, + "AAATGG": 36, + "AAACAA": 37, + "AAACAT": 38, + "AAACAC": 39, + "AAACAG": 40, + "AAACTA": 41, + "AAACTT": 42, + "AAACTC": 43, + "AAACTG": 44, + "AAACCA": 45, + "AAACCT": 46, + "AAACCC": 47, + "AAACCG": 48, + "AAACGA": 49, + "AAACGT": 50, + "AAACGC": 51, + "AAACGG": 52, + "AAAGAA": 53, + "AAAGAT": 54, + "AAAGAC": 55, + "AAAGAG": 56, + "AAAGTA": 57, + "AAAGTT": 58, + "AAAGTC": 59, + "AAAGTG": 60, + "AAAGCA": 61, + "AAAGCT": 62, + "AAAGCC": 63, + "AAAGCG": 64, + "AAAGGA": 65, + "AAAGGT": 66, + "AAAGGC": 67, + "AAAGGG": 68, + "AATAAA": 69, + "AATAAT": 70, + "AATAAC": 71, + "AATAAG": 72, + "AATATA": 73, + "AATATT": 74, + "AATATC": 75, + "AATATG": 76, + "AATACA": 77, + "AATACT": 78, + "AATACC": 79, + "AATACG": 80, + "AATAGA": 81, + "AATAGT": 82, + "AATAGC": 83, + "AATAGG": 84, + "AATTAA": 85, + "AATTAT": 86, + "AATTAC": 87, + "AATTAG": 88, + "AATTTA": 89, + "AATTTT": 90, + "AATTTC": 91, + "AATTTG": 92, + "AATTCA": 93, + "AATTCT": 94, + "AATTCC": 95, + "AATTCG": 96, + "AATTGA": 97, + "AATTGT": 98, + "AATTGC": 99, + "AATTGG": 100, + "AATCAA": 101, + "AATCAT": 102, + "AATCAC": 103, + "AATCAG": 104, + "AATCTA": 105, + "AATCTT": 106, + "AATCTC": 107, + "AATCTG": 108, + "AATCCA": 109, + "AATCCT": 110, + "AATCCC": 111, + "AATCCG": 112, + "AATCGA": 113, + "AATCGT": 114, + "AATCGC": 115, + "AATCGG": 116, + "AATGAA": 117, + "AATGAT": 118, + "AATGAC": 119, + "AATGAG": 120, + "AATGTA": 121, + "AATGTT": 122, + "AATGTC": 123, + "AATGTG": 124, + "AATGCA": 125, + "AATGCT": 126, + "AATGCC": 127, + "AATGCG": 128, + "AATGGA": 129, + "AATGGT": 130, + "AATGGC": 131, + "AATGGG": 132, + "AACAAA": 133, + "AACAAT": 134, + "AACAAC": 135, + "AACAAG": 136, + "AACATA": 137, + "AACATT": 138, + "AACATC": 139, + "AACATG": 140, + "AACACA": 141, + "AACACT": 142, + "AACACC": 143, + "AACACG": 144, + "AACAGA": 145, + "AACAGT": 146, + "AACAGC": 147, + "AACAGG": 148, + "AACTAA": 149, + "AACTAT": 150, + "AACTAC": 151, + "AACTAG": 152, + "AACTTA": 153, + "AACTTT": 154, + "AACTTC": 155, + "AACTTG": 156, + "AACTCA": 157, + "AACTCT": 158, + "AACTCC": 159, + "AACTCG": 160, + "AACTGA": 161, + "AACTGT": 162, + "AACTGC": 163, + "AACTGG": 164, + "AACCAA": 165, + "AACCAT": 166, + "AACCAC": 167, + "AACCAG": 168, + "AACCTA": 169, + "AACCTT": 170, + "AACCTC": 171, + "AACCTG": 172, + "AACCCA": 173, + "AACCCT": 174, + "AACCCC": 175, + "AACCCG": 176, + "AACCGA": 177, + "AACCGT": 178, + "AACCGC": 179, + "AACCGG": 180, + "AACGAA": 181, + "AACGAT": 182, + "AACGAC": 183, + "AACGAG": 184, + "AACGTA": 185, + "AACGTT": 186, + "AACGTC": 187, + "AACGTG": 188, + "AACGCA": 189, + "AACGCT": 190, + "AACGCC": 191, + "AACGCG": 192, + "AACGGA": 193, + "AACGGT": 194, + "AACGGC": 195, + "AACGGG": 196, + "AAGAAA": 197, + "AAGAAT": 198, + "AAGAAC": 199, + "AAGAAG": 200, + "AAGATA": 201, + "AAGATT": 202, + "AAGATC": 203, + "AAGATG": 204, + "AAGACA": 205, + "AAGACT": 206, + "AAGACC": 207, + "AAGACG": 208, + "AAGAGA": 209, + "AAGAGT": 210, + "AAGAGC": 211, + "AAGAGG": 212, + "AAGTAA": 213, + "AAGTAT": 214, + "AAGTAC": 215, + "AAGTAG": 216, + "AAGTTA": 217, + "AAGTTT": 218, + "AAGTTC": 219, + "AAGTTG": 220, + "AAGTCA": 221, + "AAGTCT": 222, + "AAGTCC": 223, + "AAGTCG": 224, + "AAGTGA": 225, + "AAGTGT": 226, + "AAGTGC": 227, + "AAGTGG": 228, + "AAGCAA": 229, + "AAGCAT": 230, + "AAGCAC": 231, + "AAGCAG": 232, + "AAGCTA": 233, + "AAGCTT": 234, + "AAGCTC": 235, + "AAGCTG": 236, + "AAGCCA": 237, + "AAGCCT": 238, + "AAGCCC": 239, + "AAGCCG": 240, + "AAGCGA": 241, + "AAGCGT": 242, + "AAGCGC": 243, + "AAGCGG": 244, + "AAGGAA": 245, + "AAGGAT": 246, + "AAGGAC": 247, + "AAGGAG": 248, + "AAGGTA": 249, + "AAGGTT": 250, + "AAGGTC": 251, + "AAGGTG": 252, + "AAGGCA": 253, + "AAGGCT": 254, + "AAGGCC": 255, + "AAGGCG": 256, + "AAGGGA": 257, + "AAGGGT": 258, + "AAGGGC": 259, + "AAGGGG": 260, + "ATAAAA": 261, + "ATAAAT": 262, + "ATAAAC": 263, + "ATAAAG": 264, + "ATAATA": 265, + "ATAATT": 266, + "ATAATC": 267, + "ATAATG": 268, + "ATAACA": 269, + "ATAACT": 270, + "ATAACC": 271, + "ATAACG": 272, + "ATAAGA": 273, + "ATAAGT": 274, + "ATAAGC": 275, + "ATAAGG": 276, + "ATATAA": 277, + "ATATAT": 278, + "ATATAC": 279, + "ATATAG": 280, + "ATATTA": 281, + "ATATTT": 282, + "ATATTC": 283, + "ATATTG": 284, + "ATATCA": 285, + "ATATCT": 286, + "ATATCC": 287, + "ATATCG": 288, + "ATATGA": 289, + "ATATGT": 290, + "ATATGC": 291, + "ATATGG": 292, + "ATACAA": 293, + "ATACAT": 294, + "ATACAC": 295, + "ATACAG": 296, + "ATACTA": 297, + "ATACTT": 298, + "ATACTC": 299, + "ATACTG": 300, + "ATACCA": 301, + "ATACCT": 302, + "ATACCC": 303, + "ATACCG": 304, + "ATACGA": 305, + "ATACGT": 306, + "ATACGC": 307, + "ATACGG": 308, + "ATAGAA": 309, + "ATAGAT": 310, + "ATAGAC": 311, + "ATAGAG": 312, + "ATAGTA": 313, + "ATAGTT": 314, + "ATAGTC": 315, + "ATAGTG": 316, + "ATAGCA": 317, + "ATAGCT": 318, + "ATAGCC": 319, + "ATAGCG": 320, + "ATAGGA": 321, + "ATAGGT": 322, + "ATAGGC": 323, + "ATAGGG": 324, + "ATTAAA": 325, + "ATTAAT": 326, + "ATTAAC": 327, + "ATTAAG": 328, + "ATTATA": 329, + "ATTATT": 330, + "ATTATC": 331, + "ATTATG": 332, + "ATTACA": 333, + "ATTACT": 334, + "ATTACC": 335, + "ATTACG": 336, + "ATTAGA": 337, + "ATTAGT": 338, + "ATTAGC": 339, + "ATTAGG": 340, + "ATTTAA": 341, + "ATTTAT": 342, + "ATTTAC": 343, + "ATTTAG": 344, + "ATTTTA": 345, + "ATTTTT": 346, + "ATTTTC": 347, + "ATTTTG": 348, + "ATTTCA": 349, + "ATTTCT": 350, + "ATTTCC": 351, + "ATTTCG": 352, + "ATTTGA": 353, + "ATTTGT": 354, + "ATTTGC": 355, + "ATTTGG": 356, + "ATTCAA": 357, + "ATTCAT": 358, + "ATTCAC": 359, + "ATTCAG": 360, + "ATTCTA": 361, + "ATTCTT": 362, + "ATTCTC": 363, + "ATTCTG": 364, + "ATTCCA": 365, + "ATTCCT": 366, + "ATTCCC": 367, + "ATTCCG": 368, + "ATTCGA": 369, + "ATTCGT": 370, + "ATTCGC": 371, + "ATTCGG": 372, + "ATTGAA": 373, + "ATTGAT": 374, + "ATTGAC": 375, + "ATTGAG": 376, + "ATTGTA": 377, + "ATTGTT": 378, + "ATTGTC": 379, + "ATTGTG": 380, + "ATTGCA": 381, + "ATTGCT": 382, + "ATTGCC": 383, + "ATTGCG": 384, + "ATTGGA": 385, + "ATTGGT": 386, + "ATTGGC": 387, + "ATTGGG": 388, + "ATCAAA": 389, + "ATCAAT": 390, + "ATCAAC": 391, + "ATCAAG": 392, + "ATCATA": 393, + "ATCATT": 394, + "ATCATC": 395, + "ATCATG": 396, + "ATCACA": 397, + "ATCACT": 398, + "ATCACC": 399, + "ATCACG": 400, + "ATCAGA": 401, + "ATCAGT": 402, + "ATCAGC": 403, + "ATCAGG": 404, + "ATCTAA": 405, + "ATCTAT": 406, + "ATCTAC": 407, + "ATCTAG": 408, + "ATCTTA": 409, + "ATCTTT": 410, + "ATCTTC": 411, + "ATCTTG": 412, + "ATCTCA": 413, + "ATCTCT": 414, + "ATCTCC": 415, + "ATCTCG": 416, + "ATCTGA": 417, + "ATCTGT": 418, + "ATCTGC": 419, + "ATCTGG": 420, + "ATCCAA": 421, + "ATCCAT": 422, + "ATCCAC": 423, + "ATCCAG": 424, + "ATCCTA": 425, + "ATCCTT": 426, + "ATCCTC": 427, + "ATCCTG": 428, + "ATCCCA": 429, + "ATCCCT": 430, + "ATCCCC": 431, + "ATCCCG": 432, + "ATCCGA": 433, + "ATCCGT": 434, + "ATCCGC": 435, + "ATCCGG": 436, + "ATCGAA": 437, + "ATCGAT": 438, + "ATCGAC": 439, + "ATCGAG": 440, + "ATCGTA": 441, + "ATCGTT": 442, + "ATCGTC": 443, + "ATCGTG": 444, + "ATCGCA": 445, + "ATCGCT": 446, + "ATCGCC": 447, + "ATCGCG": 448, + "ATCGGA": 449, + "ATCGGT": 450, + "ATCGGC": 451, + "ATCGGG": 452, + "ATGAAA": 453, + "ATGAAT": 454, + "ATGAAC": 455, + "ATGAAG": 456, + "ATGATA": 457, + "ATGATT": 458, + "ATGATC": 459, + "ATGATG": 460, + "ATGACA": 461, + "ATGACT": 462, + "ATGACC": 463, + "ATGACG": 464, + "ATGAGA": 465, + "ATGAGT": 466, + "ATGAGC": 467, + "ATGAGG": 468, + "ATGTAA": 469, + "ATGTAT": 470, + "ATGTAC": 471, + "ATGTAG": 472, + "ATGTTA": 473, + "ATGTTT": 474, + "ATGTTC": 475, + "ATGTTG": 476, + "ATGTCA": 477, + "ATGTCT": 478, + "ATGTCC": 479, + "ATGTCG": 480, + "ATGTGA": 481, + "ATGTGT": 482, + "ATGTGC": 483, + "ATGTGG": 484, + "ATGCAA": 485, + "ATGCAT": 486, + "ATGCAC": 487, + "ATGCAG": 488, + "ATGCTA": 489, + "ATGCTT": 490, + "ATGCTC": 491, + "ATGCTG": 492, + "ATGCCA": 493, + "ATGCCT": 494, + "ATGCCC": 495, + "ATGCCG": 496, + "ATGCGA": 497, + "ATGCGT": 498, + "ATGCGC": 499, + "ATGCGG": 500, + "ATGGAA": 501, + "ATGGAT": 502, + "ATGGAC": 503, + "ATGGAG": 504, + "ATGGTA": 505, + "ATGGTT": 506, + "ATGGTC": 507, + "ATGGTG": 508, + "ATGGCA": 509, + "ATGGCT": 510, + "ATGGCC": 511, + "ATGGCG": 512, + "ATGGGA": 513, + "ATGGGT": 514, + "ATGGGC": 515, + "ATGGGG": 516, + "ACAAAA": 517, + "ACAAAT": 518, + "ACAAAC": 519, + "ACAAAG": 520, + "ACAATA": 521, + "ACAATT": 522, + "ACAATC": 523, + "ACAATG": 524, + "ACAACA": 525, + "ACAACT": 526, + "ACAACC": 527, + "ACAACG": 528, + "ACAAGA": 529, + "ACAAGT": 530, + "ACAAGC": 531, + "ACAAGG": 532, + "ACATAA": 533, + "ACATAT": 534, + "ACATAC": 535, + "ACATAG": 536, + "ACATTA": 537, + "ACATTT": 538, + "ACATTC": 539, + "ACATTG": 540, + "ACATCA": 541, + "ACATCT": 542, + "ACATCC": 543, + "ACATCG": 544, + "ACATGA": 545, + "ACATGT": 546, + "ACATGC": 547, + "ACATGG": 548, + "ACACAA": 549, + "ACACAT": 550, + "ACACAC": 551, + "ACACAG": 552, + "ACACTA": 553, + "ACACTT": 554, + "ACACTC": 555, + "ACACTG": 556, + "ACACCA": 557, + "ACACCT": 558, + "ACACCC": 559, + "ACACCG": 560, + "ACACGA": 561, + "ACACGT": 562, + "ACACGC": 563, + "ACACGG": 564, + "ACAGAA": 565, + "ACAGAT": 566, + "ACAGAC": 567, + "ACAGAG": 568, + "ACAGTA": 569, + "ACAGTT": 570, + "ACAGTC": 571, + "ACAGTG": 572, + "ACAGCA": 573, + "ACAGCT": 574, + "ACAGCC": 575, + "ACAGCG": 576, + "ACAGGA": 577, + "ACAGGT": 578, + "ACAGGC": 579, + "ACAGGG": 580, + "ACTAAA": 581, + "ACTAAT": 582, + "ACTAAC": 583, + "ACTAAG": 584, + "ACTATA": 585, + "ACTATT": 586, + "ACTATC": 587, + "ACTATG": 588, + "ACTACA": 589, + "ACTACT": 590, + "ACTACC": 591, + "ACTACG": 592, + "ACTAGA": 593, + "ACTAGT": 594, + "ACTAGC": 595, + "ACTAGG": 596, + "ACTTAA": 597, + "ACTTAT": 598, + "ACTTAC": 599, + "ACTTAG": 600, + "ACTTTA": 601, + "ACTTTT": 602, + "ACTTTC": 603, + "ACTTTG": 604, + "ACTTCA": 605, + "ACTTCT": 606, + "ACTTCC": 607, + "ACTTCG": 608, + "ACTTGA": 609, + "ACTTGT": 610, + "ACTTGC": 611, + "ACTTGG": 612, + "ACTCAA": 613, + "ACTCAT": 614, + "ACTCAC": 615, + "ACTCAG": 616, + "ACTCTA": 617, + "ACTCTT": 618, + "ACTCTC": 619, + "ACTCTG": 620, + "ACTCCA": 621, + "ACTCCT": 622, + "ACTCCC": 623, + "ACTCCG": 624, + "ACTCGA": 625, + "ACTCGT": 626, + "ACTCGC": 627, + "ACTCGG": 628, + "ACTGAA": 629, + "ACTGAT": 630, + "ACTGAC": 631, + "ACTGAG": 632, + "ACTGTA": 633, + "ACTGTT": 634, + "ACTGTC": 635, + "ACTGTG": 636, + "ACTGCA": 637, + "ACTGCT": 638, + "ACTGCC": 639, + "ACTGCG": 640, + "ACTGGA": 641, + "ACTGGT": 642, + "ACTGGC": 643, + "ACTGGG": 644, + "ACCAAA": 645, + "ACCAAT": 646, + "ACCAAC": 647, + "ACCAAG": 648, + "ACCATA": 649, + "ACCATT": 650, + "ACCATC": 651, + "ACCATG": 652, + "ACCACA": 653, + "ACCACT": 654, + "ACCACC": 655, + "ACCACG": 656, + "ACCAGA": 657, + "ACCAGT": 658, + "ACCAGC": 659, + "ACCAGG": 660, + "ACCTAA": 661, + "ACCTAT": 662, + "ACCTAC": 663, + "ACCTAG": 664, + "ACCTTA": 665, + "ACCTTT": 666, + "ACCTTC": 667, + "ACCTTG": 668, + "ACCTCA": 669, + "ACCTCT": 670, + "ACCTCC": 671, + "ACCTCG": 672, + "ACCTGA": 673, + "ACCTGT": 674, + "ACCTGC": 675, + "ACCTGG": 676, + "ACCCAA": 677, + "ACCCAT": 678, + "ACCCAC": 679, + "ACCCAG": 680, + "ACCCTA": 681, + "ACCCTT": 682, + "ACCCTC": 683, + "ACCCTG": 684, + "ACCCCA": 685, + "ACCCCT": 686, + "ACCCCC": 687, + "ACCCCG": 688, + "ACCCGA": 689, + "ACCCGT": 690, + "ACCCGC": 691, + "ACCCGG": 692, + "ACCGAA": 693, + "ACCGAT": 694, + "ACCGAC": 695, + "ACCGAG": 696, + "ACCGTA": 697, + "ACCGTT": 698, + "ACCGTC": 699, + "ACCGTG": 700, + "ACCGCA": 701, + "ACCGCT": 702, + "ACCGCC": 703, + "ACCGCG": 704, + "ACCGGA": 705, + "ACCGGT": 706, + "ACCGGC": 707, + "ACCGGG": 708, + "ACGAAA": 709, + "ACGAAT": 710, + "ACGAAC": 711, + "ACGAAG": 712, + "ACGATA": 713, + "ACGATT": 714, + "ACGATC": 715, + "ACGATG": 716, + "ACGACA": 717, + "ACGACT": 718, + "ACGACC": 719, + "ACGACG": 720, + "ACGAGA": 721, + "ACGAGT": 722, + "ACGAGC": 723, + "ACGAGG": 724, + "ACGTAA": 725, + "ACGTAT": 726, + "ACGTAC": 727, + "ACGTAG": 728, + "ACGTTA": 729, + "ACGTTT": 730, + "ACGTTC": 731, + "ACGTTG": 732, + "ACGTCA": 733, + "ACGTCT": 734, + "ACGTCC": 735, + "ACGTCG": 736, + "ACGTGA": 737, + "ACGTGT": 738, + "ACGTGC": 739, + "ACGTGG": 740, + "ACGCAA": 741, + "ACGCAT": 742, + "ACGCAC": 743, + "ACGCAG": 744, + "ACGCTA": 745, + "ACGCTT": 746, + "ACGCTC": 747, + "ACGCTG": 748, + "ACGCCA": 749, + "ACGCCT": 750, + "ACGCCC": 751, + "ACGCCG": 752, + "ACGCGA": 753, + "ACGCGT": 754, + "ACGCGC": 755, + "ACGCGG": 756, + "ACGGAA": 757, + "ACGGAT": 758, + "ACGGAC": 759, + "ACGGAG": 760, + "ACGGTA": 761, + "ACGGTT": 762, + "ACGGTC": 763, + "ACGGTG": 764, + "ACGGCA": 765, + "ACGGCT": 766, + "ACGGCC": 767, + "ACGGCG": 768, + "ACGGGA": 769, + "ACGGGT": 770, + "ACGGGC": 771, + "ACGGGG": 772, + "AGAAAA": 773, + "AGAAAT": 774, + "AGAAAC": 775, + "AGAAAG": 776, + "AGAATA": 777, + "AGAATT": 778, + "AGAATC": 779, + "AGAATG": 780, + "AGAACA": 781, + "AGAACT": 782, + "AGAACC": 783, + "AGAACG": 784, + "AGAAGA": 785, + "AGAAGT": 786, + "AGAAGC": 787, + "AGAAGG": 788, + "AGATAA": 789, + "AGATAT": 790, + "AGATAC": 791, + "AGATAG": 792, + "AGATTA": 793, + "AGATTT": 794, + "AGATTC": 795, + "AGATTG": 796, + "AGATCA": 797, + "AGATCT": 798, + "AGATCC": 799, + "AGATCG": 800, + "AGATGA": 801, + "AGATGT": 802, + "AGATGC": 803, + "AGATGG": 804, + "AGACAA": 805, + "AGACAT": 806, + "AGACAC": 807, + "AGACAG": 808, + "AGACTA": 809, + "AGACTT": 810, + "AGACTC": 811, + "AGACTG": 812, + "AGACCA": 813, + "AGACCT": 814, + "AGACCC": 815, + "AGACCG": 816, + "AGACGA": 817, + "AGACGT": 818, + "AGACGC": 819, + "AGACGG": 820, + "AGAGAA": 821, + "AGAGAT": 822, + "AGAGAC": 823, + "AGAGAG": 824, + "AGAGTA": 825, + "AGAGTT": 826, + "AGAGTC": 827, + "AGAGTG": 828, + "AGAGCA": 829, + "AGAGCT": 830, + "AGAGCC": 831, + "AGAGCG": 832, + "AGAGGA": 833, + "AGAGGT": 834, + "AGAGGC": 835, + "AGAGGG": 836, + "AGTAAA": 837, + "AGTAAT": 838, + "AGTAAC": 839, + "AGTAAG": 840, + "AGTATA": 841, + "AGTATT": 842, + "AGTATC": 843, + "AGTATG": 844, + "AGTACA": 845, + "AGTACT": 846, + "AGTACC": 847, + "AGTACG": 848, + "AGTAGA": 849, + "AGTAGT": 850, + "AGTAGC": 851, + "AGTAGG": 852, + "AGTTAA": 853, + "AGTTAT": 854, + "AGTTAC": 855, + "AGTTAG": 856, + "AGTTTA": 857, + "AGTTTT": 858, + "AGTTTC": 859, + "AGTTTG": 860, + "AGTTCA": 861, + "AGTTCT": 862, + "AGTTCC": 863, + "AGTTCG": 864, + "AGTTGA": 865, + "AGTTGT": 866, + "AGTTGC": 867, + "AGTTGG": 868, + "AGTCAA": 869, + "AGTCAT": 870, + "AGTCAC": 871, + "AGTCAG": 872, + "AGTCTA": 873, + "AGTCTT": 874, + "AGTCTC": 875, + "AGTCTG": 876, + "AGTCCA": 877, + "AGTCCT": 878, + "AGTCCC": 879, + "AGTCCG": 880, + "AGTCGA": 881, + "AGTCGT": 882, + "AGTCGC": 883, + "AGTCGG": 884, + "AGTGAA": 885, + "AGTGAT": 886, + "AGTGAC": 887, + "AGTGAG": 888, + "AGTGTA": 889, + "AGTGTT": 890, + "AGTGTC": 891, + "AGTGTG": 892, + "AGTGCA": 893, + "AGTGCT": 894, + "AGTGCC": 895, + "AGTGCG": 896, + "AGTGGA": 897, + "AGTGGT": 898, + "AGTGGC": 899, + "AGTGGG": 900, + "AGCAAA": 901, + "AGCAAT": 902, + "AGCAAC": 903, + "AGCAAG": 904, + "AGCATA": 905, + "AGCATT": 906, + "AGCATC": 907, + "AGCATG": 908, + "AGCACA": 909, + "AGCACT": 910, + "AGCACC": 911, + "AGCACG": 912, + "AGCAGA": 913, + "AGCAGT": 914, + "AGCAGC": 915, + "AGCAGG": 916, + "AGCTAA": 917, + "AGCTAT": 918, + "AGCTAC": 919, + "AGCTAG": 920, + "AGCTTA": 921, + "AGCTTT": 922, + "AGCTTC": 923, + "AGCTTG": 924, + "AGCTCA": 925, + "AGCTCT": 926, + "AGCTCC": 927, + "AGCTCG": 928, + "AGCTGA": 929, + "AGCTGT": 930, + "AGCTGC": 931, + "AGCTGG": 932, + "AGCCAA": 933, + "AGCCAT": 934, + "AGCCAC": 935, + "AGCCAG": 936, + "AGCCTA": 937, + "AGCCTT": 938, + "AGCCTC": 939, + "AGCCTG": 940, + "AGCCCA": 941, + "AGCCCT": 942, + "AGCCCC": 943, + "AGCCCG": 944, + "AGCCGA": 945, + "AGCCGT": 946, + "AGCCGC": 947, + "AGCCGG": 948, + "AGCGAA": 949, + "AGCGAT": 950, + "AGCGAC": 951, + "AGCGAG": 952, + "AGCGTA": 953, + "AGCGTT": 954, + "AGCGTC": 955, + "AGCGTG": 956, + "AGCGCA": 957, + "AGCGCT": 958, + "AGCGCC": 959, + "AGCGCG": 960, + "AGCGGA": 961, + "AGCGGT": 962, + "AGCGGC": 963, + "AGCGGG": 964, + "AGGAAA": 965, + "AGGAAT": 966, + "AGGAAC": 967, + "AGGAAG": 968, + "AGGATA": 969, + "AGGATT": 970, + "AGGATC": 971, + "AGGATG": 972, + "AGGACA": 973, + "AGGACT": 974, + "AGGACC": 975, + "AGGACG": 976, + "AGGAGA": 977, + "AGGAGT": 978, + "AGGAGC": 979, + "AGGAGG": 980, + "AGGTAA": 981, + "AGGTAT": 982, + "AGGTAC": 983, + "AGGTAG": 984, + "AGGTTA": 985, + "AGGTTT": 986, + "AGGTTC": 987, + "AGGTTG": 988, + "AGGTCA": 989, + "AGGTCT": 990, + "AGGTCC": 991, + "AGGTCG": 992, + "AGGTGA": 993, + "AGGTGT": 994, + "AGGTGC": 995, + "AGGTGG": 996, + "AGGCAA": 997, + "AGGCAT": 998, + "AGGCAC": 999, + "AGGCAG": 1000, + "AGGCTA": 1001, + "AGGCTT": 1002, + "AGGCTC": 1003, + "AGGCTG": 1004, + "AGGCCA": 1005, + "AGGCCT": 1006, + "AGGCCC": 1007, + "AGGCCG": 1008, + "AGGCGA": 1009, + "AGGCGT": 1010, + "AGGCGC": 1011, + "AGGCGG": 1012, + "AGGGAA": 1013, + "AGGGAT": 1014, + "AGGGAC": 1015, + "AGGGAG": 1016, + "AGGGTA": 1017, + "AGGGTT": 1018, + "AGGGTC": 1019, + "AGGGTG": 1020, + "AGGGCA": 1021, + "AGGGCT": 1022, + "AGGGCC": 1023, + "AGGGCG": 1024, + "AGGGGA": 1025, + "AGGGGT": 1026, + "AGGGGC": 1027, + "AGGGGG": 1028, + "TAAAAA": 1029, + "TAAAAT": 1030, + "TAAAAC": 1031, + "TAAAAG": 1032, + "TAAATA": 1033, + "TAAATT": 1034, + "TAAATC": 1035, + "TAAATG": 1036, + "TAAACA": 1037, + "TAAACT": 1038, + "TAAACC": 1039, + "TAAACG": 1040, + "TAAAGA": 1041, + "TAAAGT": 1042, + "TAAAGC": 1043, + "TAAAGG": 1044, + "TAATAA": 1045, + "TAATAT": 1046, + "TAATAC": 1047, + "TAATAG": 1048, + "TAATTA": 1049, + "TAATTT": 1050, + "TAATTC": 1051, + "TAATTG": 1052, + "TAATCA": 1053, + "TAATCT": 1054, + "TAATCC": 1055, + "TAATCG": 1056, + "TAATGA": 1057, + "TAATGT": 1058, + "TAATGC": 1059, + "TAATGG": 1060, + "TAACAA": 1061, + "TAACAT": 1062, + "TAACAC": 1063, + "TAACAG": 1064, + "TAACTA": 1065, + "TAACTT": 1066, + "TAACTC": 1067, + "TAACTG": 1068, + "TAACCA": 1069, + "TAACCT": 1070, + "TAACCC": 1071, + "TAACCG": 1072, + "TAACGA": 1073, + "TAACGT": 1074, + "TAACGC": 1075, + "TAACGG": 1076, + "TAAGAA": 1077, + "TAAGAT": 1078, + "TAAGAC": 1079, + "TAAGAG": 1080, + "TAAGTA": 1081, + "TAAGTT": 1082, + "TAAGTC": 1083, + "TAAGTG": 1084, + "TAAGCA": 1085, + "TAAGCT": 1086, + "TAAGCC": 1087, + "TAAGCG": 1088, + "TAAGGA": 1089, + "TAAGGT": 1090, + "TAAGGC": 1091, + "TAAGGG": 1092, + "TATAAA": 1093, + "TATAAT": 1094, + "TATAAC": 1095, + "TATAAG": 1096, + "TATATA": 1097, + "TATATT": 1098, + "TATATC": 1099, + "TATATG": 1100, + "TATACA": 1101, + "TATACT": 1102, + "TATACC": 1103, + "TATACG": 1104, + "TATAGA": 1105, + "TATAGT": 1106, + "TATAGC": 1107, + "TATAGG": 1108, + "TATTAA": 1109, + "TATTAT": 1110, + "TATTAC": 1111, + "TATTAG": 1112, + "TATTTA": 1113, + "TATTTT": 1114, + "TATTTC": 1115, + "TATTTG": 1116, + "TATTCA": 1117, + "TATTCT": 1118, + "TATTCC": 1119, + "TATTCG": 1120, + "TATTGA": 1121, + "TATTGT": 1122, + "TATTGC": 1123, + "TATTGG": 1124, + "TATCAA": 1125, + "TATCAT": 1126, + "TATCAC": 1127, + "TATCAG": 1128, + "TATCTA": 1129, + "TATCTT": 1130, + "TATCTC": 1131, + "TATCTG": 1132, + "TATCCA": 1133, + "TATCCT": 1134, + "TATCCC": 1135, + "TATCCG": 1136, + "TATCGA": 1137, + "TATCGT": 1138, + "TATCGC": 1139, + "TATCGG": 1140, + "TATGAA": 1141, + "TATGAT": 1142, + "TATGAC": 1143, + "TATGAG": 1144, + "TATGTA": 1145, + "TATGTT": 1146, + "TATGTC": 1147, + "TATGTG": 1148, + "TATGCA": 1149, + "TATGCT": 1150, + "TATGCC": 1151, + "TATGCG": 1152, + "TATGGA": 1153, + "TATGGT": 1154, + "TATGGC": 1155, + "TATGGG": 1156, + "TACAAA": 1157, + "TACAAT": 1158, + "TACAAC": 1159, + "TACAAG": 1160, + "TACATA": 1161, + "TACATT": 1162, + "TACATC": 1163, + "TACATG": 1164, + "TACACA": 1165, + "TACACT": 1166, + "TACACC": 1167, + "TACACG": 1168, + "TACAGA": 1169, + "TACAGT": 1170, + "TACAGC": 1171, + "TACAGG": 1172, + "TACTAA": 1173, + "TACTAT": 1174, + "TACTAC": 1175, + "TACTAG": 1176, + "TACTTA": 1177, + "TACTTT": 1178, + "TACTTC": 1179, + "TACTTG": 1180, + "TACTCA": 1181, + "TACTCT": 1182, + "TACTCC": 1183, + "TACTCG": 1184, + "TACTGA": 1185, + "TACTGT": 1186, + "TACTGC": 1187, + "TACTGG": 1188, + "TACCAA": 1189, + "TACCAT": 1190, + "TACCAC": 1191, + "TACCAG": 1192, + "TACCTA": 1193, + "TACCTT": 1194, + "TACCTC": 1195, + "TACCTG": 1196, + "TACCCA": 1197, + "TACCCT": 1198, + "TACCCC": 1199, + "TACCCG": 1200, + "TACCGA": 1201, + "TACCGT": 1202, + "TACCGC": 1203, + "TACCGG": 1204, + "TACGAA": 1205, + "TACGAT": 1206, + "TACGAC": 1207, + "TACGAG": 1208, + "TACGTA": 1209, + "TACGTT": 1210, + "TACGTC": 1211, + "TACGTG": 1212, + "TACGCA": 1213, + "TACGCT": 1214, + "TACGCC": 1215, + "TACGCG": 1216, + "TACGGA": 1217, + "TACGGT": 1218, + "TACGGC": 1219, + "TACGGG": 1220, + "TAGAAA": 1221, + "TAGAAT": 1222, + "TAGAAC": 1223, + "TAGAAG": 1224, + "TAGATA": 1225, + "TAGATT": 1226, + "TAGATC": 1227, + "TAGATG": 1228, + "TAGACA": 1229, + "TAGACT": 1230, + "TAGACC": 1231, + "TAGACG": 1232, + "TAGAGA": 1233, + "TAGAGT": 1234, + "TAGAGC": 1235, + "TAGAGG": 1236, + "TAGTAA": 1237, + "TAGTAT": 1238, + "TAGTAC": 1239, + "TAGTAG": 1240, + "TAGTTA": 1241, + "TAGTTT": 1242, + "TAGTTC": 1243, + "TAGTTG": 1244, + "TAGTCA": 1245, + "TAGTCT": 1246, + "TAGTCC": 1247, + "TAGTCG": 1248, + "TAGTGA": 1249, + "TAGTGT": 1250, + "TAGTGC": 1251, + "TAGTGG": 1252, + "TAGCAA": 1253, + "TAGCAT": 1254, + "TAGCAC": 1255, + "TAGCAG": 1256, + "TAGCTA": 1257, + "TAGCTT": 1258, + "TAGCTC": 1259, + "TAGCTG": 1260, + "TAGCCA": 1261, + "TAGCCT": 1262, + "TAGCCC": 1263, + "TAGCCG": 1264, + "TAGCGA": 1265, + "TAGCGT": 1266, + "TAGCGC": 1267, + "TAGCGG": 1268, + "TAGGAA": 1269, + "TAGGAT": 1270, + "TAGGAC": 1271, + "TAGGAG": 1272, + "TAGGTA": 1273, + "TAGGTT": 1274, + "TAGGTC": 1275, + "TAGGTG": 1276, + "TAGGCA": 1277, + "TAGGCT": 1278, + "TAGGCC": 1279, + "TAGGCG": 1280, + "TAGGGA": 1281, + "TAGGGT": 1282, + "TAGGGC": 1283, + "TAGGGG": 1284, + "TTAAAA": 1285, + "TTAAAT": 1286, + "TTAAAC": 1287, + "TTAAAG": 1288, + "TTAATA": 1289, + "TTAATT": 1290, + "TTAATC": 1291, + "TTAATG": 1292, + "TTAACA": 1293, + "TTAACT": 1294, + "TTAACC": 1295, + "TTAACG": 1296, + "TTAAGA": 1297, + "TTAAGT": 1298, + "TTAAGC": 1299, + "TTAAGG": 1300, + "TTATAA": 1301, + "TTATAT": 1302, + "TTATAC": 1303, + "TTATAG": 1304, + "TTATTA": 1305, + "TTATTT": 1306, + "TTATTC": 1307, + "TTATTG": 1308, + "TTATCA": 1309, + "TTATCT": 1310, + "TTATCC": 1311, + "TTATCG": 1312, + "TTATGA": 1313, + "TTATGT": 1314, + "TTATGC": 1315, + "TTATGG": 1316, + "TTACAA": 1317, + "TTACAT": 1318, + "TTACAC": 1319, + "TTACAG": 1320, + "TTACTA": 1321, + "TTACTT": 1322, + "TTACTC": 1323, + "TTACTG": 1324, + "TTACCA": 1325, + "TTACCT": 1326, + "TTACCC": 1327, + "TTACCG": 1328, + "TTACGA": 1329, + "TTACGT": 1330, + "TTACGC": 1331, + "TTACGG": 1332, + "TTAGAA": 1333, + "TTAGAT": 1334, + "TTAGAC": 1335, + "TTAGAG": 1336, + "TTAGTA": 1337, + "TTAGTT": 1338, + "TTAGTC": 1339, + "TTAGTG": 1340, + "TTAGCA": 1341, + "TTAGCT": 1342, + "TTAGCC": 1343, + "TTAGCG": 1344, + "TTAGGA": 1345, + "TTAGGT": 1346, + "TTAGGC": 1347, + "TTAGGG": 1348, + "TTTAAA": 1349, + "TTTAAT": 1350, + "TTTAAC": 1351, + "TTTAAG": 1352, + "TTTATA": 1353, + "TTTATT": 1354, + "TTTATC": 1355, + "TTTATG": 1356, + "TTTACA": 1357, + "TTTACT": 1358, + "TTTACC": 1359, + "TTTACG": 1360, + "TTTAGA": 1361, + "TTTAGT": 1362, + "TTTAGC": 1363, + "TTTAGG": 1364, + "TTTTAA": 1365, + "TTTTAT": 1366, + "TTTTAC": 1367, + "TTTTAG": 1368, + "TTTTTA": 1369, + "TTTTTT": 1370, + "TTTTTC": 1371, + "TTTTTG": 1372, + "TTTTCA": 1373, + "TTTTCT": 1374, + "TTTTCC": 1375, + "TTTTCG": 1376, + "TTTTGA": 1377, + "TTTTGT": 1378, + "TTTTGC": 1379, + "TTTTGG": 1380, + "TTTCAA": 1381, + "TTTCAT": 1382, + "TTTCAC": 1383, + "TTTCAG": 1384, + "TTTCTA": 1385, + "TTTCTT": 1386, + "TTTCTC": 1387, + "TTTCTG": 1388, + "TTTCCA": 1389, + "TTTCCT": 1390, + "TTTCCC": 1391, + "TTTCCG": 1392, + "TTTCGA": 1393, + "TTTCGT": 1394, + "TTTCGC": 1395, + "TTTCGG": 1396, + "TTTGAA": 1397, + "TTTGAT": 1398, + "TTTGAC": 1399, + "TTTGAG": 1400, + "TTTGTA": 1401, + "TTTGTT": 1402, + "TTTGTC": 1403, + "TTTGTG": 1404, + "TTTGCA": 1405, + "TTTGCT": 1406, + "TTTGCC": 1407, + "TTTGCG": 1408, + "TTTGGA": 1409, + "TTTGGT": 1410, + "TTTGGC": 1411, + "TTTGGG": 1412, + "TTCAAA": 1413, + "TTCAAT": 1414, + "TTCAAC": 1415, + "TTCAAG": 1416, + "TTCATA": 1417, + "TTCATT": 1418, + "TTCATC": 1419, + "TTCATG": 1420, + "TTCACA": 1421, + "TTCACT": 1422, + "TTCACC": 1423, + "TTCACG": 1424, + "TTCAGA": 1425, + "TTCAGT": 1426, + "TTCAGC": 1427, + "TTCAGG": 1428, + "TTCTAA": 1429, + "TTCTAT": 1430, + "TTCTAC": 1431, + "TTCTAG": 1432, + "TTCTTA": 1433, + "TTCTTT": 1434, + "TTCTTC": 1435, + "TTCTTG": 1436, + "TTCTCA": 1437, + "TTCTCT": 1438, + "TTCTCC": 1439, + "TTCTCG": 1440, + "TTCTGA": 1441, + "TTCTGT": 1442, + "TTCTGC": 1443, + "TTCTGG": 1444, + "TTCCAA": 1445, + "TTCCAT": 1446, + "TTCCAC": 1447, + "TTCCAG": 1448, + "TTCCTA": 1449, + "TTCCTT": 1450, + "TTCCTC": 1451, + "TTCCTG": 1452, + "TTCCCA": 1453, + "TTCCCT": 1454, + "TTCCCC": 1455, + "TTCCCG": 1456, + "TTCCGA": 1457, + "TTCCGT": 1458, + "TTCCGC": 1459, + "TTCCGG": 1460, + "TTCGAA": 1461, + "TTCGAT": 1462, + "TTCGAC": 1463, + "TTCGAG": 1464, + "TTCGTA": 1465, + "TTCGTT": 1466, + "TTCGTC": 1467, + "TTCGTG": 1468, + "TTCGCA": 1469, + "TTCGCT": 1470, + "TTCGCC": 1471, + "TTCGCG": 1472, + "TTCGGA": 1473, + "TTCGGT": 1474, + "TTCGGC": 1475, + "TTCGGG": 1476, + "TTGAAA": 1477, + "TTGAAT": 1478, + "TTGAAC": 1479, + "TTGAAG": 1480, + "TTGATA": 1481, + "TTGATT": 1482, + "TTGATC": 1483, + "TTGATG": 1484, + "TTGACA": 1485, + "TTGACT": 1486, + "TTGACC": 1487, + "TTGACG": 1488, + "TTGAGA": 1489, + "TTGAGT": 1490, + "TTGAGC": 1491, + "TTGAGG": 1492, + "TTGTAA": 1493, + "TTGTAT": 1494, + "TTGTAC": 1495, + "TTGTAG": 1496, + "TTGTTA": 1497, + "TTGTTT": 1498, + "TTGTTC": 1499, + "TTGTTG": 1500, + "TTGTCA": 1501, + "TTGTCT": 1502, + "TTGTCC": 1503, + "TTGTCG": 1504, + "TTGTGA": 1505, + "TTGTGT": 1506, + "TTGTGC": 1507, + "TTGTGG": 1508, + "TTGCAA": 1509, + "TTGCAT": 1510, + "TTGCAC": 1511, + "TTGCAG": 1512, + "TTGCTA": 1513, + "TTGCTT": 1514, + "TTGCTC": 1515, + "TTGCTG": 1516, + "TTGCCA": 1517, + "TTGCCT": 1518, + "TTGCCC": 1519, + "TTGCCG": 1520, + "TTGCGA": 1521, + "TTGCGT": 1522, + "TTGCGC": 1523, + "TTGCGG": 1524, + "TTGGAA": 1525, + "TTGGAT": 1526, + "TTGGAC": 1527, + "TTGGAG": 1528, + "TTGGTA": 1529, + "TTGGTT": 1530, + "TTGGTC": 1531, + "TTGGTG": 1532, + "TTGGCA": 1533, + "TTGGCT": 1534, + "TTGGCC": 1535, + "TTGGCG": 1536, + "TTGGGA": 1537, + "TTGGGT": 1538, + "TTGGGC": 1539, + "TTGGGG": 1540, + "TCAAAA": 1541, + "TCAAAT": 1542, + "TCAAAC": 1543, + "TCAAAG": 1544, + "TCAATA": 1545, + "TCAATT": 1546, + "TCAATC": 1547, + "TCAATG": 1548, + "TCAACA": 1549, + "TCAACT": 1550, + "TCAACC": 1551, + "TCAACG": 1552, + "TCAAGA": 1553, + "TCAAGT": 1554, + "TCAAGC": 1555, + "TCAAGG": 1556, + "TCATAA": 1557, + "TCATAT": 1558, + "TCATAC": 1559, + "TCATAG": 1560, + "TCATTA": 1561, + "TCATTT": 1562, + "TCATTC": 1563, + "TCATTG": 1564, + "TCATCA": 1565, + "TCATCT": 1566, + "TCATCC": 1567, + "TCATCG": 1568, + "TCATGA": 1569, + "TCATGT": 1570, + "TCATGC": 1571, + "TCATGG": 1572, + "TCACAA": 1573, + "TCACAT": 1574, + "TCACAC": 1575, + "TCACAG": 1576, + "TCACTA": 1577, + "TCACTT": 1578, + "TCACTC": 1579, + "TCACTG": 1580, + "TCACCA": 1581, + "TCACCT": 1582, + "TCACCC": 1583, + "TCACCG": 1584, + "TCACGA": 1585, + "TCACGT": 1586, + "TCACGC": 1587, + "TCACGG": 1588, + "TCAGAA": 1589, + "TCAGAT": 1590, + "TCAGAC": 1591, + "TCAGAG": 1592, + "TCAGTA": 1593, + "TCAGTT": 1594, + "TCAGTC": 1595, + "TCAGTG": 1596, + "TCAGCA": 1597, + "TCAGCT": 1598, + "TCAGCC": 1599, + "TCAGCG": 1600, + "TCAGGA": 1601, + "TCAGGT": 1602, + "TCAGGC": 1603, + "TCAGGG": 1604, + "TCTAAA": 1605, + "TCTAAT": 1606, + "TCTAAC": 1607, + "TCTAAG": 1608, + "TCTATA": 1609, + "TCTATT": 1610, + "TCTATC": 1611, + "TCTATG": 1612, + "TCTACA": 1613, + "TCTACT": 1614, + "TCTACC": 1615, + "TCTACG": 1616, + "TCTAGA": 1617, + "TCTAGT": 1618, + "TCTAGC": 1619, + "TCTAGG": 1620, + "TCTTAA": 1621, + "TCTTAT": 1622, + "TCTTAC": 1623, + "TCTTAG": 1624, + "TCTTTA": 1625, + "TCTTTT": 1626, + "TCTTTC": 1627, + "TCTTTG": 1628, + "TCTTCA": 1629, + "TCTTCT": 1630, + "TCTTCC": 1631, + "TCTTCG": 1632, + "TCTTGA": 1633, + "TCTTGT": 1634, + "TCTTGC": 1635, + "TCTTGG": 1636, + "TCTCAA": 1637, + "TCTCAT": 1638, + "TCTCAC": 1639, + "TCTCAG": 1640, + "TCTCTA": 1641, + "TCTCTT": 1642, + "TCTCTC": 1643, + "TCTCTG": 1644, + "TCTCCA": 1645, + "TCTCCT": 1646, + "TCTCCC": 1647, + "TCTCCG": 1648, + "TCTCGA": 1649, + "TCTCGT": 1650, + "TCTCGC": 1651, + "TCTCGG": 1652, + "TCTGAA": 1653, + "TCTGAT": 1654, + "TCTGAC": 1655, + "TCTGAG": 1656, + "TCTGTA": 1657, + "TCTGTT": 1658, + "TCTGTC": 1659, + "TCTGTG": 1660, + "TCTGCA": 1661, + "TCTGCT": 1662, + "TCTGCC": 1663, + "TCTGCG": 1664, + "TCTGGA": 1665, + "TCTGGT": 1666, + "TCTGGC": 1667, + "TCTGGG": 1668, + "TCCAAA": 1669, + "TCCAAT": 1670, + "TCCAAC": 1671, + "TCCAAG": 1672, + "TCCATA": 1673, + "TCCATT": 1674, + "TCCATC": 1675, + "TCCATG": 1676, + "TCCACA": 1677, + "TCCACT": 1678, + "TCCACC": 1679, + "TCCACG": 1680, + "TCCAGA": 1681, + "TCCAGT": 1682, + "TCCAGC": 1683, + "TCCAGG": 1684, + "TCCTAA": 1685, + "TCCTAT": 1686, + "TCCTAC": 1687, + "TCCTAG": 1688, + "TCCTTA": 1689, + "TCCTTT": 1690, + "TCCTTC": 1691, + "TCCTTG": 1692, + "TCCTCA": 1693, + "TCCTCT": 1694, + "TCCTCC": 1695, + "TCCTCG": 1696, + "TCCTGA": 1697, + "TCCTGT": 1698, + "TCCTGC": 1699, + "TCCTGG": 1700, + "TCCCAA": 1701, + "TCCCAT": 1702, + "TCCCAC": 1703, + "TCCCAG": 1704, + "TCCCTA": 1705, + "TCCCTT": 1706, + "TCCCTC": 1707, + "TCCCTG": 1708, + "TCCCCA": 1709, + "TCCCCT": 1710, + "TCCCCC": 1711, + "TCCCCG": 1712, + "TCCCGA": 1713, + "TCCCGT": 1714, + "TCCCGC": 1715, + "TCCCGG": 1716, + "TCCGAA": 1717, + "TCCGAT": 1718, + "TCCGAC": 1719, + "TCCGAG": 1720, + "TCCGTA": 1721, + "TCCGTT": 1722, + "TCCGTC": 1723, + "TCCGTG": 1724, + "TCCGCA": 1725, + "TCCGCT": 1726, + "TCCGCC": 1727, + "TCCGCG": 1728, + "TCCGGA": 1729, + "TCCGGT": 1730, + "TCCGGC": 1731, + "TCCGGG": 1732, + "TCGAAA": 1733, + "TCGAAT": 1734, + "TCGAAC": 1735, + "TCGAAG": 1736, + "TCGATA": 1737, + "TCGATT": 1738, + "TCGATC": 1739, + "TCGATG": 1740, + "TCGACA": 1741, + "TCGACT": 1742, + "TCGACC": 1743, + "TCGACG": 1744, + "TCGAGA": 1745, + "TCGAGT": 1746, + "TCGAGC": 1747, + "TCGAGG": 1748, + "TCGTAA": 1749, + "TCGTAT": 1750, + "TCGTAC": 1751, + "TCGTAG": 1752, + "TCGTTA": 1753, + "TCGTTT": 1754, + "TCGTTC": 1755, + "TCGTTG": 1756, + "TCGTCA": 1757, + "TCGTCT": 1758, + "TCGTCC": 1759, + "TCGTCG": 1760, + "TCGTGA": 1761, + "TCGTGT": 1762, + "TCGTGC": 1763, + "TCGTGG": 1764, + "TCGCAA": 1765, + "TCGCAT": 1766, + "TCGCAC": 1767, + "TCGCAG": 1768, + "TCGCTA": 1769, + "TCGCTT": 1770, + "TCGCTC": 1771, + "TCGCTG": 1772, + "TCGCCA": 1773, + "TCGCCT": 1774, + "TCGCCC": 1775, + "TCGCCG": 1776, + "TCGCGA": 1777, + "TCGCGT": 1778, + "TCGCGC": 1779, + "TCGCGG": 1780, + "TCGGAA": 1781, + "TCGGAT": 1782, + "TCGGAC": 1783, + "TCGGAG": 1784, + "TCGGTA": 1785, + "TCGGTT": 1786, + "TCGGTC": 1787, + "TCGGTG": 1788, + "TCGGCA": 1789, + "TCGGCT": 1790, + "TCGGCC": 1791, + "TCGGCG": 1792, + "TCGGGA": 1793, + "TCGGGT": 1794, + "TCGGGC": 1795, + "TCGGGG": 1796, + "TGAAAA": 1797, + "TGAAAT": 1798, + "TGAAAC": 1799, + "TGAAAG": 1800, + "TGAATA": 1801, + "TGAATT": 1802, + "TGAATC": 1803, + "TGAATG": 1804, + "TGAACA": 1805, + "TGAACT": 1806, + "TGAACC": 1807, + "TGAACG": 1808, + "TGAAGA": 1809, + "TGAAGT": 1810, + "TGAAGC": 1811, + "TGAAGG": 1812, + "TGATAA": 1813, + "TGATAT": 1814, + "TGATAC": 1815, + "TGATAG": 1816, + "TGATTA": 1817, + "TGATTT": 1818, + "TGATTC": 1819, + "TGATTG": 1820, + "TGATCA": 1821, + "TGATCT": 1822, + "TGATCC": 1823, + "TGATCG": 1824, + "TGATGA": 1825, + "TGATGT": 1826, + "TGATGC": 1827, + "TGATGG": 1828, + "TGACAA": 1829, + "TGACAT": 1830, + "TGACAC": 1831, + "TGACAG": 1832, + "TGACTA": 1833, + "TGACTT": 1834, + "TGACTC": 1835, + "TGACTG": 1836, + "TGACCA": 1837, + "TGACCT": 1838, + "TGACCC": 1839, + "TGACCG": 1840, + "TGACGA": 1841, + "TGACGT": 1842, + "TGACGC": 1843, + "TGACGG": 1844, + "TGAGAA": 1845, + "TGAGAT": 1846, + "TGAGAC": 1847, + "TGAGAG": 1848, + "TGAGTA": 1849, + "TGAGTT": 1850, + "TGAGTC": 1851, + "TGAGTG": 1852, + "TGAGCA": 1853, + "TGAGCT": 1854, + "TGAGCC": 1855, + "TGAGCG": 1856, + "TGAGGA": 1857, + "TGAGGT": 1858, + "TGAGGC": 1859, + "TGAGGG": 1860, + "TGTAAA": 1861, + "TGTAAT": 1862, + "TGTAAC": 1863, + "TGTAAG": 1864, + "TGTATA": 1865, + "TGTATT": 1866, + "TGTATC": 1867, + "TGTATG": 1868, + "TGTACA": 1869, + "TGTACT": 1870, + "TGTACC": 1871, + "TGTACG": 1872, + "TGTAGA": 1873, + "TGTAGT": 1874, + "TGTAGC": 1875, + "TGTAGG": 1876, + "TGTTAA": 1877, + "TGTTAT": 1878, + "TGTTAC": 1879, + "TGTTAG": 1880, + "TGTTTA": 1881, + "TGTTTT": 1882, + "TGTTTC": 1883, + "TGTTTG": 1884, + "TGTTCA": 1885, + "TGTTCT": 1886, + "TGTTCC": 1887, + "TGTTCG": 1888, + "TGTTGA": 1889, + "TGTTGT": 1890, + "TGTTGC": 1891, + "TGTTGG": 1892, + "TGTCAA": 1893, + "TGTCAT": 1894, + "TGTCAC": 1895, + "TGTCAG": 1896, + "TGTCTA": 1897, + "TGTCTT": 1898, + "TGTCTC": 1899, + "TGTCTG": 1900, + "TGTCCA": 1901, + "TGTCCT": 1902, + "TGTCCC": 1903, + "TGTCCG": 1904, + "TGTCGA": 1905, + "TGTCGT": 1906, + "TGTCGC": 1907, + "TGTCGG": 1908, + "TGTGAA": 1909, + "TGTGAT": 1910, + "TGTGAC": 1911, + "TGTGAG": 1912, + "TGTGTA": 1913, + "TGTGTT": 1914, + "TGTGTC": 1915, + "TGTGTG": 1916, + "TGTGCA": 1917, + "TGTGCT": 1918, + "TGTGCC": 1919, + "TGTGCG": 1920, + "TGTGGA": 1921, + "TGTGGT": 1922, + "TGTGGC": 1923, + "TGTGGG": 1924, + "TGCAAA": 1925, + "TGCAAT": 1926, + "TGCAAC": 1927, + "TGCAAG": 1928, + "TGCATA": 1929, + "TGCATT": 1930, + "TGCATC": 1931, + "TGCATG": 1932, + "TGCACA": 1933, + "TGCACT": 1934, + "TGCACC": 1935, + "TGCACG": 1936, + "TGCAGA": 1937, + "TGCAGT": 1938, + "TGCAGC": 1939, + "TGCAGG": 1940, + "TGCTAA": 1941, + "TGCTAT": 1942, + "TGCTAC": 1943, + "TGCTAG": 1944, + "TGCTTA": 1945, + "TGCTTT": 1946, + "TGCTTC": 1947, + "TGCTTG": 1948, + "TGCTCA": 1949, + "TGCTCT": 1950, + "TGCTCC": 1951, + "TGCTCG": 1952, + "TGCTGA": 1953, + "TGCTGT": 1954, + "TGCTGC": 1955, + "TGCTGG": 1956, + "TGCCAA": 1957, + "TGCCAT": 1958, + "TGCCAC": 1959, + "TGCCAG": 1960, + "TGCCTA": 1961, + "TGCCTT": 1962, + "TGCCTC": 1963, + "TGCCTG": 1964, + "TGCCCA": 1965, + "TGCCCT": 1966, + "TGCCCC": 1967, + "TGCCCG": 1968, + "TGCCGA": 1969, + "TGCCGT": 1970, + "TGCCGC": 1971, + "TGCCGG": 1972, + "TGCGAA": 1973, + "TGCGAT": 1974, + "TGCGAC": 1975, + "TGCGAG": 1976, + "TGCGTA": 1977, + "TGCGTT": 1978, + "TGCGTC": 1979, + "TGCGTG": 1980, + "TGCGCA": 1981, + "TGCGCT": 1982, + "TGCGCC": 1983, + "TGCGCG": 1984, + "TGCGGA": 1985, + "TGCGGT": 1986, + "TGCGGC": 1987, + "TGCGGG": 1988, + "TGGAAA": 1989, + "TGGAAT": 1990, + "TGGAAC": 1991, + "TGGAAG": 1992, + "TGGATA": 1993, + "TGGATT": 1994, + "TGGATC": 1995, + "TGGATG": 1996, + "TGGACA": 1997, + "TGGACT": 1998, + "TGGACC": 1999, + "TGGACG": 2000, + "TGGAGA": 2001, + "TGGAGT": 2002, + "TGGAGC": 2003, + "TGGAGG": 2004, + "TGGTAA": 2005, + "TGGTAT": 2006, + "TGGTAC": 2007, + "TGGTAG": 2008, + "TGGTTA": 2009, + "TGGTTT": 2010, + "TGGTTC": 2011, + "TGGTTG": 2012, + "TGGTCA": 2013, + "TGGTCT": 2014, + "TGGTCC": 2015, + "TGGTCG": 2016, + "TGGTGA": 2017, + "TGGTGT": 2018, + "TGGTGC": 2019, + "TGGTGG": 2020, + "TGGCAA": 2021, + "TGGCAT": 2022, + "TGGCAC": 2023, + "TGGCAG": 2024, + "TGGCTA": 2025, + "TGGCTT": 2026, + "TGGCTC": 2027, + "TGGCTG": 2028, + "TGGCCA": 2029, + "TGGCCT": 2030, + "TGGCCC": 2031, + "TGGCCG": 2032, + "TGGCGA": 2033, + "TGGCGT": 2034, + "TGGCGC": 2035, + "TGGCGG": 2036, + "TGGGAA": 2037, + "TGGGAT": 2038, + "TGGGAC": 2039, + "TGGGAG": 2040, + "TGGGTA": 2041, + "TGGGTT": 2042, + "TGGGTC": 2043, + "TGGGTG": 2044, + "TGGGCA": 2045, + "TGGGCT": 2046, + "TGGGCC": 2047, + "TGGGCG": 2048, + "TGGGGA": 2049, + "TGGGGT": 2050, + "TGGGGC": 2051, + "TGGGGG": 2052, + "CAAAAA": 2053, + "CAAAAT": 2054, + "CAAAAC": 2055, + "CAAAAG": 2056, + "CAAATA": 2057, + "CAAATT": 2058, + "CAAATC": 2059, + "CAAATG": 2060, + "CAAACA": 2061, + "CAAACT": 2062, + "CAAACC": 2063, + "CAAACG": 2064, + "CAAAGA": 2065, + "CAAAGT": 2066, + "CAAAGC": 2067, + "CAAAGG": 2068, + "CAATAA": 2069, + "CAATAT": 2070, + "CAATAC": 2071, + "CAATAG": 2072, + "CAATTA": 2073, + "CAATTT": 2074, + "CAATTC": 2075, + "CAATTG": 2076, + "CAATCA": 2077, + "CAATCT": 2078, + "CAATCC": 2079, + "CAATCG": 2080, + "CAATGA": 2081, + "CAATGT": 2082, + "CAATGC": 2083, + "CAATGG": 2084, + "CAACAA": 2085, + "CAACAT": 2086, + "CAACAC": 2087, + "CAACAG": 2088, + "CAACTA": 2089, + "CAACTT": 2090, + "CAACTC": 2091, + "CAACTG": 2092, + "CAACCA": 2093, + "CAACCT": 2094, + "CAACCC": 2095, + "CAACCG": 2096, + "CAACGA": 2097, + "CAACGT": 2098, + "CAACGC": 2099, + "CAACGG": 2100, + "CAAGAA": 2101, + "CAAGAT": 2102, + "CAAGAC": 2103, + "CAAGAG": 2104, + "CAAGTA": 2105, + "CAAGTT": 2106, + "CAAGTC": 2107, + "CAAGTG": 2108, + "CAAGCA": 2109, + "CAAGCT": 2110, + "CAAGCC": 2111, + "CAAGCG": 2112, + "CAAGGA": 2113, + "CAAGGT": 2114, + "CAAGGC": 2115, + "CAAGGG": 2116, + "CATAAA": 2117, + "CATAAT": 2118, + "CATAAC": 2119, + "CATAAG": 2120, + "CATATA": 2121, + "CATATT": 2122, + "CATATC": 2123, + "CATATG": 2124, + "CATACA": 2125, + "CATACT": 2126, + "CATACC": 2127, + "CATACG": 2128, + "CATAGA": 2129, + "CATAGT": 2130, + "CATAGC": 2131, + "CATAGG": 2132, + "CATTAA": 2133, + "CATTAT": 2134, + "CATTAC": 2135, + "CATTAG": 2136, + "CATTTA": 2137, + "CATTTT": 2138, + "CATTTC": 2139, + "CATTTG": 2140, + "CATTCA": 2141, + "CATTCT": 2142, + "CATTCC": 2143, + "CATTCG": 2144, + "CATTGA": 2145, + "CATTGT": 2146, + "CATTGC": 2147, + "CATTGG": 2148, + "CATCAA": 2149, + "CATCAT": 2150, + "CATCAC": 2151, + "CATCAG": 2152, + "CATCTA": 2153, + "CATCTT": 2154, + "CATCTC": 2155, + "CATCTG": 2156, + "CATCCA": 2157, + "CATCCT": 2158, + "CATCCC": 2159, + "CATCCG": 2160, + "CATCGA": 2161, + "CATCGT": 2162, + "CATCGC": 2163, + "CATCGG": 2164, + "CATGAA": 2165, + "CATGAT": 2166, + "CATGAC": 2167, + "CATGAG": 2168, + "CATGTA": 2169, + "CATGTT": 2170, + "CATGTC": 2171, + "CATGTG": 2172, + "CATGCA": 2173, + "CATGCT": 2174, + "CATGCC": 2175, + "CATGCG": 2176, + "CATGGA": 2177, + "CATGGT": 2178, + "CATGGC": 2179, + "CATGGG": 2180, + "CACAAA": 2181, + "CACAAT": 2182, + "CACAAC": 2183, + "CACAAG": 2184, + "CACATA": 2185, + "CACATT": 2186, + "CACATC": 2187, + "CACATG": 2188, + "CACACA": 2189, + "CACACT": 2190, + "CACACC": 2191, + "CACACG": 2192, + "CACAGA": 2193, + "CACAGT": 2194, + "CACAGC": 2195, + "CACAGG": 2196, + "CACTAA": 2197, + "CACTAT": 2198, + "CACTAC": 2199, + "CACTAG": 2200, + "CACTTA": 2201, + "CACTTT": 2202, + "CACTTC": 2203, + "CACTTG": 2204, + "CACTCA": 2205, + "CACTCT": 2206, + "CACTCC": 2207, + "CACTCG": 2208, + "CACTGA": 2209, + "CACTGT": 2210, + "CACTGC": 2211, + "CACTGG": 2212, + "CACCAA": 2213, + "CACCAT": 2214, + "CACCAC": 2215, + "CACCAG": 2216, + "CACCTA": 2217, + "CACCTT": 2218, + "CACCTC": 2219, + "CACCTG": 2220, + "CACCCA": 2221, + "CACCCT": 2222, + "CACCCC": 2223, + "CACCCG": 2224, + "CACCGA": 2225, + "CACCGT": 2226, + "CACCGC": 2227, + "CACCGG": 2228, + "CACGAA": 2229, + "CACGAT": 2230, + "CACGAC": 2231, + "CACGAG": 2232, + "CACGTA": 2233, + "CACGTT": 2234, + "CACGTC": 2235, + "CACGTG": 2236, + "CACGCA": 2237, + "CACGCT": 2238, + "CACGCC": 2239, + "CACGCG": 2240, + "CACGGA": 2241, + "CACGGT": 2242, + "CACGGC": 2243, + "CACGGG": 2244, + "CAGAAA": 2245, + "CAGAAT": 2246, + "CAGAAC": 2247, + "CAGAAG": 2248, + "CAGATA": 2249, + "CAGATT": 2250, + "CAGATC": 2251, + "CAGATG": 2252, + "CAGACA": 2253, + "CAGACT": 2254, + "CAGACC": 2255, + "CAGACG": 2256, + "CAGAGA": 2257, + "CAGAGT": 2258, + "CAGAGC": 2259, + "CAGAGG": 2260, + "CAGTAA": 2261, + "CAGTAT": 2262, + "CAGTAC": 2263, + "CAGTAG": 2264, + "CAGTTA": 2265, + "CAGTTT": 2266, + "CAGTTC": 2267, + "CAGTTG": 2268, + "CAGTCA": 2269, + "CAGTCT": 2270, + "CAGTCC": 2271, + "CAGTCG": 2272, + "CAGTGA": 2273, + "CAGTGT": 2274, + "CAGTGC": 2275, + "CAGTGG": 2276, + "CAGCAA": 2277, + "CAGCAT": 2278, + "CAGCAC": 2279, + "CAGCAG": 2280, + "CAGCTA": 2281, + "CAGCTT": 2282, + "CAGCTC": 2283, + "CAGCTG": 2284, + "CAGCCA": 2285, + "CAGCCT": 2286, + "CAGCCC": 2287, + "CAGCCG": 2288, + "CAGCGA": 2289, + "CAGCGT": 2290, + "CAGCGC": 2291, + "CAGCGG": 2292, + "CAGGAA": 2293, + "CAGGAT": 2294, + "CAGGAC": 2295, + "CAGGAG": 2296, + "CAGGTA": 2297, + "CAGGTT": 2298, + "CAGGTC": 2299, + "CAGGTG": 2300, + "CAGGCA": 2301, + "CAGGCT": 2302, + "CAGGCC": 2303, + "CAGGCG": 2304, + "CAGGGA": 2305, + "CAGGGT": 2306, + "CAGGGC": 2307, + "CAGGGG": 2308, + "CTAAAA": 2309, + "CTAAAT": 2310, + "CTAAAC": 2311, + "CTAAAG": 2312, + "CTAATA": 2313, + "CTAATT": 2314, + "CTAATC": 2315, + "CTAATG": 2316, + "CTAACA": 2317, + "CTAACT": 2318, + "CTAACC": 2319, + "CTAACG": 2320, + "CTAAGA": 2321, + "CTAAGT": 2322, + "CTAAGC": 2323, + "CTAAGG": 2324, + "CTATAA": 2325, + "CTATAT": 2326, + "CTATAC": 2327, + "CTATAG": 2328, + "CTATTA": 2329, + "CTATTT": 2330, + "CTATTC": 2331, + "CTATTG": 2332, + "CTATCA": 2333, + "CTATCT": 2334, + "CTATCC": 2335, + "CTATCG": 2336, + "CTATGA": 2337, + "CTATGT": 2338, + "CTATGC": 2339, + "CTATGG": 2340, + "CTACAA": 2341, + "CTACAT": 2342, + "CTACAC": 2343, + "CTACAG": 2344, + "CTACTA": 2345, + "CTACTT": 2346, + "CTACTC": 2347, + "CTACTG": 2348, + "CTACCA": 2349, + "CTACCT": 2350, + "CTACCC": 2351, + "CTACCG": 2352, + "CTACGA": 2353, + "CTACGT": 2354, + "CTACGC": 2355, + "CTACGG": 2356, + "CTAGAA": 2357, + "CTAGAT": 2358, + "CTAGAC": 2359, + "CTAGAG": 2360, + "CTAGTA": 2361, + "CTAGTT": 2362, + "CTAGTC": 2363, + "CTAGTG": 2364, + "CTAGCA": 2365, + "CTAGCT": 2366, + "CTAGCC": 2367, + "CTAGCG": 2368, + "CTAGGA": 2369, + "CTAGGT": 2370, + "CTAGGC": 2371, + "CTAGGG": 2372, + "CTTAAA": 2373, + "CTTAAT": 2374, + "CTTAAC": 2375, + "CTTAAG": 2376, + "CTTATA": 2377, + "CTTATT": 2378, + "CTTATC": 2379, + "CTTATG": 2380, + "CTTACA": 2381, + "CTTACT": 2382, + "CTTACC": 2383, + "CTTACG": 2384, + "CTTAGA": 2385, + "CTTAGT": 2386, + "CTTAGC": 2387, + "CTTAGG": 2388, + "CTTTAA": 2389, + "CTTTAT": 2390, + "CTTTAC": 2391, + "CTTTAG": 2392, + "CTTTTA": 2393, + "CTTTTT": 2394, + "CTTTTC": 2395, + "CTTTTG": 2396, + "CTTTCA": 2397, + "CTTTCT": 2398, + "CTTTCC": 2399, + "CTTTCG": 2400, + "CTTTGA": 2401, + "CTTTGT": 2402, + "CTTTGC": 2403, + "CTTTGG": 2404, + "CTTCAA": 2405, + "CTTCAT": 2406, + "CTTCAC": 2407, + "CTTCAG": 2408, + "CTTCTA": 2409, + "CTTCTT": 2410, + "CTTCTC": 2411, + "CTTCTG": 2412, + "CTTCCA": 2413, + "CTTCCT": 2414, + "CTTCCC": 2415, + "CTTCCG": 2416, + "CTTCGA": 2417, + "CTTCGT": 2418, + "CTTCGC": 2419, + "CTTCGG": 2420, + "CTTGAA": 2421, + "CTTGAT": 2422, + "CTTGAC": 2423, + "CTTGAG": 2424, + "CTTGTA": 2425, + "CTTGTT": 2426, + "CTTGTC": 2427, + "CTTGTG": 2428, + "CTTGCA": 2429, + "CTTGCT": 2430, + "CTTGCC": 2431, + "CTTGCG": 2432, + "CTTGGA": 2433, + "CTTGGT": 2434, + "CTTGGC": 2435, + "CTTGGG": 2436, + "CTCAAA": 2437, + "CTCAAT": 2438, + "CTCAAC": 2439, + "CTCAAG": 2440, + "CTCATA": 2441, + "CTCATT": 2442, + "CTCATC": 2443, + "CTCATG": 2444, + "CTCACA": 2445, + "CTCACT": 2446, + "CTCACC": 2447, + "CTCACG": 2448, + "CTCAGA": 2449, + "CTCAGT": 2450, + "CTCAGC": 2451, + "CTCAGG": 2452, + "CTCTAA": 2453, + "CTCTAT": 2454, + "CTCTAC": 2455, + "CTCTAG": 2456, + "CTCTTA": 2457, + "CTCTTT": 2458, + "CTCTTC": 2459, + "CTCTTG": 2460, + "CTCTCA": 2461, + "CTCTCT": 2462, + "CTCTCC": 2463, + "CTCTCG": 2464, + "CTCTGA": 2465, + "CTCTGT": 2466, + "CTCTGC": 2467, + "CTCTGG": 2468, + "CTCCAA": 2469, + "CTCCAT": 2470, + "CTCCAC": 2471, + "CTCCAG": 2472, + "CTCCTA": 2473, + "CTCCTT": 2474, + "CTCCTC": 2475, + "CTCCTG": 2476, + "CTCCCA": 2477, + "CTCCCT": 2478, + "CTCCCC": 2479, + "CTCCCG": 2480, + "CTCCGA": 2481, + "CTCCGT": 2482, + "CTCCGC": 2483, + "CTCCGG": 2484, + "CTCGAA": 2485, + "CTCGAT": 2486, + "CTCGAC": 2487, + "CTCGAG": 2488, + "CTCGTA": 2489, + "CTCGTT": 2490, + "CTCGTC": 2491, + "CTCGTG": 2492, + "CTCGCA": 2493, + "CTCGCT": 2494, + "CTCGCC": 2495, + "CTCGCG": 2496, + "CTCGGA": 2497, + "CTCGGT": 2498, + "CTCGGC": 2499, + "CTCGGG": 2500, + "CTGAAA": 2501, + "CTGAAT": 2502, + "CTGAAC": 2503, + "CTGAAG": 2504, + "CTGATA": 2505, + "CTGATT": 2506, + "CTGATC": 2507, + "CTGATG": 2508, + "CTGACA": 2509, + "CTGACT": 2510, + "CTGACC": 2511, + "CTGACG": 2512, + "CTGAGA": 2513, + "CTGAGT": 2514, + "CTGAGC": 2515, + "CTGAGG": 2516, + "CTGTAA": 2517, + "CTGTAT": 2518, + "CTGTAC": 2519, + "CTGTAG": 2520, + "CTGTTA": 2521, + "CTGTTT": 2522, + "CTGTTC": 2523, + "CTGTTG": 2524, + "CTGTCA": 2525, + "CTGTCT": 2526, + "CTGTCC": 2527, + "CTGTCG": 2528, + "CTGTGA": 2529, + "CTGTGT": 2530, + "CTGTGC": 2531, + "CTGTGG": 2532, + "CTGCAA": 2533, + "CTGCAT": 2534, + "CTGCAC": 2535, + "CTGCAG": 2536, + "CTGCTA": 2537, + "CTGCTT": 2538, + "CTGCTC": 2539, + "CTGCTG": 2540, + "CTGCCA": 2541, + "CTGCCT": 2542, + "CTGCCC": 2543, + "CTGCCG": 2544, + "CTGCGA": 2545, + "CTGCGT": 2546, + "CTGCGC": 2547, + "CTGCGG": 2548, + "CTGGAA": 2549, + "CTGGAT": 2550, + "CTGGAC": 2551, + "CTGGAG": 2552, + "CTGGTA": 2553, + "CTGGTT": 2554, + "CTGGTC": 2555, + "CTGGTG": 2556, + "CTGGCA": 2557, + "CTGGCT": 2558, + "CTGGCC": 2559, + "CTGGCG": 2560, + "CTGGGA": 2561, + "CTGGGT": 2562, + "CTGGGC": 2563, + "CTGGGG": 2564, + "CCAAAA": 2565, + "CCAAAT": 2566, + "CCAAAC": 2567, + "CCAAAG": 2568, + "CCAATA": 2569, + "CCAATT": 2570, + "CCAATC": 2571, + "CCAATG": 2572, + "CCAACA": 2573, + "CCAACT": 2574, + "CCAACC": 2575, + "CCAACG": 2576, + "CCAAGA": 2577, + "CCAAGT": 2578, + "CCAAGC": 2579, + "CCAAGG": 2580, + "CCATAA": 2581, + "CCATAT": 2582, + "CCATAC": 2583, + "CCATAG": 2584, + "CCATTA": 2585, + "CCATTT": 2586, + "CCATTC": 2587, + "CCATTG": 2588, + "CCATCA": 2589, + "CCATCT": 2590, + "CCATCC": 2591, + "CCATCG": 2592, + "CCATGA": 2593, + "CCATGT": 2594, + "CCATGC": 2595, + "CCATGG": 2596, + "CCACAA": 2597, + "CCACAT": 2598, + "CCACAC": 2599, + "CCACAG": 2600, + "CCACTA": 2601, + "CCACTT": 2602, + "CCACTC": 2603, + "CCACTG": 2604, + "CCACCA": 2605, + "CCACCT": 2606, + "CCACCC": 2607, + "CCACCG": 2608, + "CCACGA": 2609, + "CCACGT": 2610, + "CCACGC": 2611, + "CCACGG": 2612, + "CCAGAA": 2613, + "CCAGAT": 2614, + "CCAGAC": 2615, + "CCAGAG": 2616, + "CCAGTA": 2617, + "CCAGTT": 2618, + "CCAGTC": 2619, + "CCAGTG": 2620, + "CCAGCA": 2621, + "CCAGCT": 2622, + "CCAGCC": 2623, + "CCAGCG": 2624, + "CCAGGA": 2625, + "CCAGGT": 2626, + "CCAGGC": 2627, + "CCAGGG": 2628, + "CCTAAA": 2629, + "CCTAAT": 2630, + "CCTAAC": 2631, + "CCTAAG": 2632, + "CCTATA": 2633, + "CCTATT": 2634, + "CCTATC": 2635, + "CCTATG": 2636, + "CCTACA": 2637, + "CCTACT": 2638, + "CCTACC": 2639, + "CCTACG": 2640, + "CCTAGA": 2641, + "CCTAGT": 2642, + "CCTAGC": 2643, + "CCTAGG": 2644, + "CCTTAA": 2645, + "CCTTAT": 2646, + "CCTTAC": 2647, + "CCTTAG": 2648, + "CCTTTA": 2649, + "CCTTTT": 2650, + "CCTTTC": 2651, + "CCTTTG": 2652, + "CCTTCA": 2653, + "CCTTCT": 2654, + "CCTTCC": 2655, + "CCTTCG": 2656, + "CCTTGA": 2657, + "CCTTGT": 2658, + "CCTTGC": 2659, + "CCTTGG": 2660, + "CCTCAA": 2661, + "CCTCAT": 2662, + "CCTCAC": 2663, + "CCTCAG": 2664, + "CCTCTA": 2665, + "CCTCTT": 2666, + "CCTCTC": 2667, + "CCTCTG": 2668, + "CCTCCA": 2669, + "CCTCCT": 2670, + "CCTCCC": 2671, + "CCTCCG": 2672, + "CCTCGA": 2673, + "CCTCGT": 2674, + "CCTCGC": 2675, + "CCTCGG": 2676, + "CCTGAA": 2677, + "CCTGAT": 2678, + "CCTGAC": 2679, + "CCTGAG": 2680, + "CCTGTA": 2681, + "CCTGTT": 2682, + "CCTGTC": 2683, + "CCTGTG": 2684, + "CCTGCA": 2685, + "CCTGCT": 2686, + "CCTGCC": 2687, + "CCTGCG": 2688, + "CCTGGA": 2689, + "CCTGGT": 2690, + "CCTGGC": 2691, + "CCTGGG": 2692, + "CCCAAA": 2693, + "CCCAAT": 2694, + "CCCAAC": 2695, + "CCCAAG": 2696, + "CCCATA": 2697, + "CCCATT": 2698, + "CCCATC": 2699, + "CCCATG": 2700, + "CCCACA": 2701, + "CCCACT": 2702, + "CCCACC": 2703, + "CCCACG": 2704, + "CCCAGA": 2705, + "CCCAGT": 2706, + "CCCAGC": 2707, + "CCCAGG": 2708, + "CCCTAA": 2709, + "CCCTAT": 2710, + "CCCTAC": 2711, + "CCCTAG": 2712, + "CCCTTA": 2713, + "CCCTTT": 2714, + "CCCTTC": 2715, + "CCCTTG": 2716, + "CCCTCA": 2717, + "CCCTCT": 2718, + "CCCTCC": 2719, + "CCCTCG": 2720, + "CCCTGA": 2721, + "CCCTGT": 2722, + "CCCTGC": 2723, + "CCCTGG": 2724, + "CCCCAA": 2725, + "CCCCAT": 2726, + "CCCCAC": 2727, + "CCCCAG": 2728, + "CCCCTA": 2729, + "CCCCTT": 2730, + "CCCCTC": 2731, + "CCCCTG": 2732, + "CCCCCA": 2733, + "CCCCCT": 2734, + "CCCCCC": 2735, + "CCCCCG": 2736, + "CCCCGA": 2737, + "CCCCGT": 2738, + "CCCCGC": 2739, + "CCCCGG": 2740, + "CCCGAA": 2741, + "CCCGAT": 2742, + "CCCGAC": 2743, + "CCCGAG": 2744, + "CCCGTA": 2745, + "CCCGTT": 2746, + "CCCGTC": 2747, + "CCCGTG": 2748, + "CCCGCA": 2749, + "CCCGCT": 2750, + "CCCGCC": 2751, + "CCCGCG": 2752, + "CCCGGA": 2753, + "CCCGGT": 2754, + "CCCGGC": 2755, + "CCCGGG": 2756, + "CCGAAA": 2757, + "CCGAAT": 2758, + "CCGAAC": 2759, + "CCGAAG": 2760, + "CCGATA": 2761, + "CCGATT": 2762, + "CCGATC": 2763, + "CCGATG": 2764, + "CCGACA": 2765, + "CCGACT": 2766, + "CCGACC": 2767, + "CCGACG": 2768, + "CCGAGA": 2769, + "CCGAGT": 2770, + "CCGAGC": 2771, + "CCGAGG": 2772, + "CCGTAA": 2773, + "CCGTAT": 2774, + "CCGTAC": 2775, + "CCGTAG": 2776, + "CCGTTA": 2777, + "CCGTTT": 2778, + "CCGTTC": 2779, + "CCGTTG": 2780, + "CCGTCA": 2781, + "CCGTCT": 2782, + "CCGTCC": 2783, + "CCGTCG": 2784, + "CCGTGA": 2785, + "CCGTGT": 2786, + "CCGTGC": 2787, + "CCGTGG": 2788, + "CCGCAA": 2789, + "CCGCAT": 2790, + "CCGCAC": 2791, + "CCGCAG": 2792, + "CCGCTA": 2793, + "CCGCTT": 2794, + "CCGCTC": 2795, + "CCGCTG": 2796, + "CCGCCA": 2797, + "CCGCCT": 2798, + "CCGCCC": 2799, + "CCGCCG": 2800, + "CCGCGA": 2801, + "CCGCGT": 2802, + "CCGCGC": 2803, + "CCGCGG": 2804, + "CCGGAA": 2805, + "CCGGAT": 2806, + "CCGGAC": 2807, + "CCGGAG": 2808, + "CCGGTA": 2809, + "CCGGTT": 2810, + "CCGGTC": 2811, + "CCGGTG": 2812, + "CCGGCA": 2813, + "CCGGCT": 2814, + "CCGGCC": 2815, + "CCGGCG": 2816, + "CCGGGA": 2817, + "CCGGGT": 2818, + "CCGGGC": 2819, + "CCGGGG": 2820, + "CGAAAA": 2821, + "CGAAAT": 2822, + "CGAAAC": 2823, + "CGAAAG": 2824, + "CGAATA": 2825, + "CGAATT": 2826, + "CGAATC": 2827, + "CGAATG": 2828, + "CGAACA": 2829, + "CGAACT": 2830, + "CGAACC": 2831, + "CGAACG": 2832, + "CGAAGA": 2833, + "CGAAGT": 2834, + "CGAAGC": 2835, + "CGAAGG": 2836, + "CGATAA": 2837, + "CGATAT": 2838, + "CGATAC": 2839, + "CGATAG": 2840, + "CGATTA": 2841, + "CGATTT": 2842, + "CGATTC": 2843, + "CGATTG": 2844, + "CGATCA": 2845, + "CGATCT": 2846, + "CGATCC": 2847, + "CGATCG": 2848, + "CGATGA": 2849, + "CGATGT": 2850, + "CGATGC": 2851, + "CGATGG": 2852, + "CGACAA": 2853, + "CGACAT": 2854, + "CGACAC": 2855, + "CGACAG": 2856, + "CGACTA": 2857, + "CGACTT": 2858, + "CGACTC": 2859, + "CGACTG": 2860, + "CGACCA": 2861, + "CGACCT": 2862, + "CGACCC": 2863, + "CGACCG": 2864, + "CGACGA": 2865, + "CGACGT": 2866, + "CGACGC": 2867, + "CGACGG": 2868, + "CGAGAA": 2869, + "CGAGAT": 2870, + "CGAGAC": 2871, + "CGAGAG": 2872, + "CGAGTA": 2873, + "CGAGTT": 2874, + "CGAGTC": 2875, + "CGAGTG": 2876, + "CGAGCA": 2877, + "CGAGCT": 2878, + "CGAGCC": 2879, + "CGAGCG": 2880, + "CGAGGA": 2881, + "CGAGGT": 2882, + "CGAGGC": 2883, + "CGAGGG": 2884, + "CGTAAA": 2885, + "CGTAAT": 2886, + "CGTAAC": 2887, + "CGTAAG": 2888, + "CGTATA": 2889, + "CGTATT": 2890, + "CGTATC": 2891, + "CGTATG": 2892, + "CGTACA": 2893, + "CGTACT": 2894, + "CGTACC": 2895, + "CGTACG": 2896, + "CGTAGA": 2897, + "CGTAGT": 2898, + "CGTAGC": 2899, + "CGTAGG": 2900, + "CGTTAA": 2901, + "CGTTAT": 2902, + "CGTTAC": 2903, + "CGTTAG": 2904, + "CGTTTA": 2905, + "CGTTTT": 2906, + "CGTTTC": 2907, + "CGTTTG": 2908, + "CGTTCA": 2909, + "CGTTCT": 2910, + "CGTTCC": 2911, + "CGTTCG": 2912, + "CGTTGA": 2913, + "CGTTGT": 2914, + "CGTTGC": 2915, + "CGTTGG": 2916, + "CGTCAA": 2917, + "CGTCAT": 2918, + "CGTCAC": 2919, + "CGTCAG": 2920, + "CGTCTA": 2921, + "CGTCTT": 2922, + "CGTCTC": 2923, + "CGTCTG": 2924, + "CGTCCA": 2925, + "CGTCCT": 2926, + "CGTCCC": 2927, + "CGTCCG": 2928, + "CGTCGA": 2929, + "CGTCGT": 2930, + "CGTCGC": 2931, + "CGTCGG": 2932, + "CGTGAA": 2933, + "CGTGAT": 2934, + "CGTGAC": 2935, + "CGTGAG": 2936, + "CGTGTA": 2937, + "CGTGTT": 2938, + "CGTGTC": 2939, + "CGTGTG": 2940, + "CGTGCA": 2941, + "CGTGCT": 2942, + "CGTGCC": 2943, + "CGTGCG": 2944, + "CGTGGA": 2945, + "CGTGGT": 2946, + "CGTGGC": 2947, + "CGTGGG": 2948, + "CGCAAA": 2949, + "CGCAAT": 2950, + "CGCAAC": 2951, + "CGCAAG": 2952, + "CGCATA": 2953, + "CGCATT": 2954, + "CGCATC": 2955, + "CGCATG": 2956, + "CGCACA": 2957, + "CGCACT": 2958, + "CGCACC": 2959, + "CGCACG": 2960, + "CGCAGA": 2961, + "CGCAGT": 2962, + "CGCAGC": 2963, + "CGCAGG": 2964, + "CGCTAA": 2965, + "CGCTAT": 2966, + "CGCTAC": 2967, + "CGCTAG": 2968, + "CGCTTA": 2969, + "CGCTTT": 2970, + "CGCTTC": 2971, + "CGCTTG": 2972, + "CGCTCA": 2973, + "CGCTCT": 2974, + "CGCTCC": 2975, + "CGCTCG": 2976, + "CGCTGA": 2977, + "CGCTGT": 2978, + "CGCTGC": 2979, + "CGCTGG": 2980, + "CGCCAA": 2981, + "CGCCAT": 2982, + "CGCCAC": 2983, + "CGCCAG": 2984, + "CGCCTA": 2985, + "CGCCTT": 2986, + "CGCCTC": 2987, + "CGCCTG": 2988, + "CGCCCA": 2989, + "CGCCCT": 2990, + "CGCCCC": 2991, + "CGCCCG": 2992, + "CGCCGA": 2993, + "CGCCGT": 2994, + "CGCCGC": 2995, + "CGCCGG": 2996, + "CGCGAA": 2997, + "CGCGAT": 2998, + "CGCGAC": 2999, + "CGCGAG": 3000, + "CGCGTA": 3001, + "CGCGTT": 3002, + "CGCGTC": 3003, + "CGCGTG": 3004, + "CGCGCA": 3005, + "CGCGCT": 3006, + "CGCGCC": 3007, + "CGCGCG": 3008, + "CGCGGA": 3009, + "CGCGGT": 3010, + "CGCGGC": 3011, + "CGCGGG": 3012, + "CGGAAA": 3013, + "CGGAAT": 3014, + "CGGAAC": 3015, + "CGGAAG": 3016, + "CGGATA": 3017, + "CGGATT": 3018, + "CGGATC": 3019, + "CGGATG": 3020, + "CGGACA": 3021, + "CGGACT": 3022, + "CGGACC": 3023, + "CGGACG": 3024, + "CGGAGA": 3025, + "CGGAGT": 3026, + "CGGAGC": 3027, + "CGGAGG": 3028, + "CGGTAA": 3029, + "CGGTAT": 3030, + "CGGTAC": 3031, + "CGGTAG": 3032, + "CGGTTA": 3033, + "CGGTTT": 3034, + "CGGTTC": 3035, + "CGGTTG": 3036, + "CGGTCA": 3037, + "CGGTCT": 3038, + "CGGTCC": 3039, + "CGGTCG": 3040, + "CGGTGA": 3041, + "CGGTGT": 3042, + "CGGTGC": 3043, + "CGGTGG": 3044, + "CGGCAA": 3045, + "CGGCAT": 3046, + "CGGCAC": 3047, + "CGGCAG": 3048, + "CGGCTA": 3049, + "CGGCTT": 3050, + "CGGCTC": 3051, + "CGGCTG": 3052, + "CGGCCA": 3053, + "CGGCCT": 3054, + "CGGCCC": 3055, + "CGGCCG": 3056, + "CGGCGA": 3057, + "CGGCGT": 3058, + "CGGCGC": 3059, + "CGGCGG": 3060, + "CGGGAA": 3061, + "CGGGAT": 3062, + "CGGGAC": 3063, + "CGGGAG": 3064, + "CGGGTA": 3065, + "CGGGTT": 3066, + "CGGGTC": 3067, + "CGGGTG": 3068, + "CGGGCA": 3069, + "CGGGCT": 3070, + "CGGGCC": 3071, + "CGGGCG": 3072, + "CGGGGA": 3073, + "CGGGGT": 3074, + "CGGGGC": 3075, + "CGGGGG": 3076, + "GAAAAA": 3077, + "GAAAAT": 3078, + "GAAAAC": 3079, + "GAAAAG": 3080, + "GAAATA": 3081, + "GAAATT": 3082, + "GAAATC": 3083, + "GAAATG": 3084, + "GAAACA": 3085, + "GAAACT": 3086, + "GAAACC": 3087, + "GAAACG": 3088, + "GAAAGA": 3089, + "GAAAGT": 3090, + "GAAAGC": 3091, + "GAAAGG": 3092, + "GAATAA": 3093, + "GAATAT": 3094, + "GAATAC": 3095, + "GAATAG": 3096, + "GAATTA": 3097, + "GAATTT": 3098, + "GAATTC": 3099, + "GAATTG": 3100, + "GAATCA": 3101, + "GAATCT": 3102, + "GAATCC": 3103, + "GAATCG": 3104, + "GAATGA": 3105, + "GAATGT": 3106, + "GAATGC": 3107, + "GAATGG": 3108, + "GAACAA": 3109, + "GAACAT": 3110, + "GAACAC": 3111, + "GAACAG": 3112, + "GAACTA": 3113, + "GAACTT": 3114, + "GAACTC": 3115, + "GAACTG": 3116, + "GAACCA": 3117, + "GAACCT": 3118, + "GAACCC": 3119, + "GAACCG": 3120, + "GAACGA": 3121, + "GAACGT": 3122, + "GAACGC": 3123, + "GAACGG": 3124, + "GAAGAA": 3125, + "GAAGAT": 3126, + "GAAGAC": 3127, + "GAAGAG": 3128, + "GAAGTA": 3129, + "GAAGTT": 3130, + "GAAGTC": 3131, + "GAAGTG": 3132, + "GAAGCA": 3133, + "GAAGCT": 3134, + "GAAGCC": 3135, + "GAAGCG": 3136, + "GAAGGA": 3137, + "GAAGGT": 3138, + "GAAGGC": 3139, + "GAAGGG": 3140, + "GATAAA": 3141, + "GATAAT": 3142, + "GATAAC": 3143, + "GATAAG": 3144, + "GATATA": 3145, + "GATATT": 3146, + "GATATC": 3147, + "GATATG": 3148, + "GATACA": 3149, + "GATACT": 3150, + "GATACC": 3151, + "GATACG": 3152, + "GATAGA": 3153, + "GATAGT": 3154, + "GATAGC": 3155, + "GATAGG": 3156, + "GATTAA": 3157, + "GATTAT": 3158, + "GATTAC": 3159, + "GATTAG": 3160, + "GATTTA": 3161, + "GATTTT": 3162, + "GATTTC": 3163, + "GATTTG": 3164, + "GATTCA": 3165, + "GATTCT": 3166, + "GATTCC": 3167, + "GATTCG": 3168, + "GATTGA": 3169, + "GATTGT": 3170, + "GATTGC": 3171, + "GATTGG": 3172, + "GATCAA": 3173, + "GATCAT": 3174, + "GATCAC": 3175, + "GATCAG": 3176, + "GATCTA": 3177, + "GATCTT": 3178, + "GATCTC": 3179, + "GATCTG": 3180, + "GATCCA": 3181, + "GATCCT": 3182, + "GATCCC": 3183, + "GATCCG": 3184, + "GATCGA": 3185, + "GATCGT": 3186, + "GATCGC": 3187, + "GATCGG": 3188, + "GATGAA": 3189, + "GATGAT": 3190, + "GATGAC": 3191, + "GATGAG": 3192, + "GATGTA": 3193, + "GATGTT": 3194, + "GATGTC": 3195, + "GATGTG": 3196, + "GATGCA": 3197, + "GATGCT": 3198, + "GATGCC": 3199, + "GATGCG": 3200, + "GATGGA": 3201, + "GATGGT": 3202, + "GATGGC": 3203, + "GATGGG": 3204, + "GACAAA": 3205, + "GACAAT": 3206, + "GACAAC": 3207, + "GACAAG": 3208, + "GACATA": 3209, + "GACATT": 3210, + "GACATC": 3211, + "GACATG": 3212, + "GACACA": 3213, + "GACACT": 3214, + "GACACC": 3215, + "GACACG": 3216, + "GACAGA": 3217, + "GACAGT": 3218, + "GACAGC": 3219, + "GACAGG": 3220, + "GACTAA": 3221, + "GACTAT": 3222, + "GACTAC": 3223, + "GACTAG": 3224, + "GACTTA": 3225, + "GACTTT": 3226, + "GACTTC": 3227, + "GACTTG": 3228, + "GACTCA": 3229, + "GACTCT": 3230, + "GACTCC": 3231, + "GACTCG": 3232, + "GACTGA": 3233, + "GACTGT": 3234, + "GACTGC": 3235, + "GACTGG": 3236, + "GACCAA": 3237, + "GACCAT": 3238, + "GACCAC": 3239, + "GACCAG": 3240, + "GACCTA": 3241, + "GACCTT": 3242, + "GACCTC": 3243, + "GACCTG": 3244, + "GACCCA": 3245, + "GACCCT": 3246, + "GACCCC": 3247, + "GACCCG": 3248, + "GACCGA": 3249, + "GACCGT": 3250, + "GACCGC": 3251, + "GACCGG": 3252, + "GACGAA": 3253, + "GACGAT": 3254, + "GACGAC": 3255, + "GACGAG": 3256, + "GACGTA": 3257, + "GACGTT": 3258, + "GACGTC": 3259, + "GACGTG": 3260, + "GACGCA": 3261, + "GACGCT": 3262, + "GACGCC": 3263, + "GACGCG": 3264, + "GACGGA": 3265, + "GACGGT": 3266, + "GACGGC": 3267, + "GACGGG": 3268, + "GAGAAA": 3269, + "GAGAAT": 3270, + "GAGAAC": 3271, + "GAGAAG": 3272, + "GAGATA": 3273, + "GAGATT": 3274, + "GAGATC": 3275, + "GAGATG": 3276, + "GAGACA": 3277, + "GAGACT": 3278, + "GAGACC": 3279, + "GAGACG": 3280, + "GAGAGA": 3281, + "GAGAGT": 3282, + "GAGAGC": 3283, + "GAGAGG": 3284, + "GAGTAA": 3285, + "GAGTAT": 3286, + "GAGTAC": 3287, + "GAGTAG": 3288, + "GAGTTA": 3289, + "GAGTTT": 3290, + "GAGTTC": 3291, + "GAGTTG": 3292, + "GAGTCA": 3293, + "GAGTCT": 3294, + "GAGTCC": 3295, + "GAGTCG": 3296, + "GAGTGA": 3297, + "GAGTGT": 3298, + "GAGTGC": 3299, + "GAGTGG": 3300, + "GAGCAA": 3301, + "GAGCAT": 3302, + "GAGCAC": 3303, + "GAGCAG": 3304, + "GAGCTA": 3305, + "GAGCTT": 3306, + "GAGCTC": 3307, + "GAGCTG": 3308, + "GAGCCA": 3309, + "GAGCCT": 3310, + "GAGCCC": 3311, + "GAGCCG": 3312, + "GAGCGA": 3313, + "GAGCGT": 3314, + "GAGCGC": 3315, + "GAGCGG": 3316, + "GAGGAA": 3317, + "GAGGAT": 3318, + "GAGGAC": 3319, + "GAGGAG": 3320, + "GAGGTA": 3321, + "GAGGTT": 3322, + "GAGGTC": 3323, + "GAGGTG": 3324, + "GAGGCA": 3325, + "GAGGCT": 3326, + "GAGGCC": 3327, + "GAGGCG": 3328, + "GAGGGA": 3329, + "GAGGGT": 3330, + "GAGGGC": 3331, + "GAGGGG": 3332, + "GTAAAA": 3333, + "GTAAAT": 3334, + "GTAAAC": 3335, + "GTAAAG": 3336, + "GTAATA": 3337, + "GTAATT": 3338, + "GTAATC": 3339, + "GTAATG": 3340, + "GTAACA": 3341, + "GTAACT": 3342, + "GTAACC": 3343, + "GTAACG": 3344, + "GTAAGA": 3345, + "GTAAGT": 3346, + "GTAAGC": 3347, + "GTAAGG": 3348, + "GTATAA": 3349, + "GTATAT": 3350, + "GTATAC": 3351, + "GTATAG": 3352, + "GTATTA": 3353, + "GTATTT": 3354, + "GTATTC": 3355, + "GTATTG": 3356, + "GTATCA": 3357, + "GTATCT": 3358, + "GTATCC": 3359, + "GTATCG": 3360, + "GTATGA": 3361, + "GTATGT": 3362, + "GTATGC": 3363, + "GTATGG": 3364, + "GTACAA": 3365, + "GTACAT": 3366, + "GTACAC": 3367, + "GTACAG": 3368, + "GTACTA": 3369, + "GTACTT": 3370, + "GTACTC": 3371, + "GTACTG": 3372, + "GTACCA": 3373, + "GTACCT": 3374, + "GTACCC": 3375, + "GTACCG": 3376, + "GTACGA": 3377, + "GTACGT": 3378, + "GTACGC": 3379, + "GTACGG": 3380, + "GTAGAA": 3381, + "GTAGAT": 3382, + "GTAGAC": 3383, + "GTAGAG": 3384, + "GTAGTA": 3385, + "GTAGTT": 3386, + "GTAGTC": 3387, + "GTAGTG": 3388, + "GTAGCA": 3389, + "GTAGCT": 3390, + "GTAGCC": 3391, + "GTAGCG": 3392, + "GTAGGA": 3393, + "GTAGGT": 3394, + "GTAGGC": 3395, + "GTAGGG": 3396, + "GTTAAA": 3397, + "GTTAAT": 3398, + "GTTAAC": 3399, + "GTTAAG": 3400, + "GTTATA": 3401, + "GTTATT": 3402, + "GTTATC": 3403, + "GTTATG": 3404, + "GTTACA": 3405, + "GTTACT": 3406, + "GTTACC": 3407, + "GTTACG": 3408, + "GTTAGA": 3409, + "GTTAGT": 3410, + "GTTAGC": 3411, + "GTTAGG": 3412, + "GTTTAA": 3413, + "GTTTAT": 3414, + "GTTTAC": 3415, + "GTTTAG": 3416, + "GTTTTA": 3417, + "GTTTTT": 3418, + "GTTTTC": 3419, + "GTTTTG": 3420, + "GTTTCA": 3421, + "GTTTCT": 3422, + "GTTTCC": 3423, + "GTTTCG": 3424, + "GTTTGA": 3425, + "GTTTGT": 3426, + "GTTTGC": 3427, + "GTTTGG": 3428, + "GTTCAA": 3429, + "GTTCAT": 3430, + "GTTCAC": 3431, + "GTTCAG": 3432, + "GTTCTA": 3433, + "GTTCTT": 3434, + "GTTCTC": 3435, + "GTTCTG": 3436, + "GTTCCA": 3437, + "GTTCCT": 3438, + "GTTCCC": 3439, + "GTTCCG": 3440, + "GTTCGA": 3441, + "GTTCGT": 3442, + "GTTCGC": 3443, + "GTTCGG": 3444, + "GTTGAA": 3445, + "GTTGAT": 3446, + "GTTGAC": 3447, + "GTTGAG": 3448, + "GTTGTA": 3449, + "GTTGTT": 3450, + "GTTGTC": 3451, + "GTTGTG": 3452, + "GTTGCA": 3453, + "GTTGCT": 3454, + "GTTGCC": 3455, + "GTTGCG": 3456, + "GTTGGA": 3457, + "GTTGGT": 3458, + "GTTGGC": 3459, + "GTTGGG": 3460, + "GTCAAA": 3461, + "GTCAAT": 3462, + "GTCAAC": 3463, + "GTCAAG": 3464, + "GTCATA": 3465, + "GTCATT": 3466, + "GTCATC": 3467, + "GTCATG": 3468, + "GTCACA": 3469, + "GTCACT": 3470, + "GTCACC": 3471, + "GTCACG": 3472, + "GTCAGA": 3473, + "GTCAGT": 3474, + "GTCAGC": 3475, + "GTCAGG": 3476, + "GTCTAA": 3477, + "GTCTAT": 3478, + "GTCTAC": 3479, + "GTCTAG": 3480, + "GTCTTA": 3481, + "GTCTTT": 3482, + "GTCTTC": 3483, + "GTCTTG": 3484, + "GTCTCA": 3485, + "GTCTCT": 3486, + "GTCTCC": 3487, + "GTCTCG": 3488, + "GTCTGA": 3489, + "GTCTGT": 3490, + "GTCTGC": 3491, + "GTCTGG": 3492, + "GTCCAA": 3493, + "GTCCAT": 3494, + "GTCCAC": 3495, + "GTCCAG": 3496, + "GTCCTA": 3497, + "GTCCTT": 3498, + "GTCCTC": 3499, + "GTCCTG": 3500, + "GTCCCA": 3501, + "GTCCCT": 3502, + "GTCCCC": 3503, + "GTCCCG": 3504, + "GTCCGA": 3505, + "GTCCGT": 3506, + "GTCCGC": 3507, + "GTCCGG": 3508, + "GTCGAA": 3509, + "GTCGAT": 3510, + "GTCGAC": 3511, + "GTCGAG": 3512, + "GTCGTA": 3513, + "GTCGTT": 3514, + "GTCGTC": 3515, + "GTCGTG": 3516, + "GTCGCA": 3517, + "GTCGCT": 3518, + "GTCGCC": 3519, + "GTCGCG": 3520, + "GTCGGA": 3521, + "GTCGGT": 3522, + "GTCGGC": 3523, + "GTCGGG": 3524, + "GTGAAA": 3525, + "GTGAAT": 3526, + "GTGAAC": 3527, + "GTGAAG": 3528, + "GTGATA": 3529, + "GTGATT": 3530, + "GTGATC": 3531, + "GTGATG": 3532, + "GTGACA": 3533, + "GTGACT": 3534, + "GTGACC": 3535, + "GTGACG": 3536, + "GTGAGA": 3537, + "GTGAGT": 3538, + "GTGAGC": 3539, + "GTGAGG": 3540, + "GTGTAA": 3541, + "GTGTAT": 3542, + "GTGTAC": 3543, + "GTGTAG": 3544, + "GTGTTA": 3545, + "GTGTTT": 3546, + "GTGTTC": 3547, + "GTGTTG": 3548, + "GTGTCA": 3549, + "GTGTCT": 3550, + "GTGTCC": 3551, + "GTGTCG": 3552, + "GTGTGA": 3553, + "GTGTGT": 3554, + "GTGTGC": 3555, + "GTGTGG": 3556, + "GTGCAA": 3557, + "GTGCAT": 3558, + "GTGCAC": 3559, + "GTGCAG": 3560, + "GTGCTA": 3561, + "GTGCTT": 3562, + "GTGCTC": 3563, + "GTGCTG": 3564, + "GTGCCA": 3565, + "GTGCCT": 3566, + "GTGCCC": 3567, + "GTGCCG": 3568, + "GTGCGA": 3569, + "GTGCGT": 3570, + "GTGCGC": 3571, + "GTGCGG": 3572, + "GTGGAA": 3573, + "GTGGAT": 3574, + "GTGGAC": 3575, + "GTGGAG": 3576, + "GTGGTA": 3577, + "GTGGTT": 3578, + "GTGGTC": 3579, + "GTGGTG": 3580, + "GTGGCA": 3581, + "GTGGCT": 3582, + "GTGGCC": 3583, + "GTGGCG": 3584, + "GTGGGA": 3585, + "GTGGGT": 3586, + "GTGGGC": 3587, + "GTGGGG": 3588, + "GCAAAA": 3589, + "GCAAAT": 3590, + "GCAAAC": 3591, + "GCAAAG": 3592, + "GCAATA": 3593, + "GCAATT": 3594, + "GCAATC": 3595, + "GCAATG": 3596, + "GCAACA": 3597, + "GCAACT": 3598, + "GCAACC": 3599, + "GCAACG": 3600, + "GCAAGA": 3601, + "GCAAGT": 3602, + "GCAAGC": 3603, + "GCAAGG": 3604, + "GCATAA": 3605, + "GCATAT": 3606, + "GCATAC": 3607, + "GCATAG": 3608, + "GCATTA": 3609, + "GCATTT": 3610, + "GCATTC": 3611, + "GCATTG": 3612, + "GCATCA": 3613, + "GCATCT": 3614, + "GCATCC": 3615, + "GCATCG": 3616, + "GCATGA": 3617, + "GCATGT": 3618, + "GCATGC": 3619, + "GCATGG": 3620, + "GCACAA": 3621, + "GCACAT": 3622, + "GCACAC": 3623, + "GCACAG": 3624, + "GCACTA": 3625, + "GCACTT": 3626, + "GCACTC": 3627, + "GCACTG": 3628, + "GCACCA": 3629, + "GCACCT": 3630, + "GCACCC": 3631, + "GCACCG": 3632, + "GCACGA": 3633, + "GCACGT": 3634, + "GCACGC": 3635, + "GCACGG": 3636, + "GCAGAA": 3637, + "GCAGAT": 3638, + "GCAGAC": 3639, + "GCAGAG": 3640, + "GCAGTA": 3641, + "GCAGTT": 3642, + "GCAGTC": 3643, + "GCAGTG": 3644, + "GCAGCA": 3645, + "GCAGCT": 3646, + "GCAGCC": 3647, + "GCAGCG": 3648, + "GCAGGA": 3649, + "GCAGGT": 3650, + "GCAGGC": 3651, + "GCAGGG": 3652, + "GCTAAA": 3653, + "GCTAAT": 3654, + "GCTAAC": 3655, + "GCTAAG": 3656, + "GCTATA": 3657, + "GCTATT": 3658, + "GCTATC": 3659, + "GCTATG": 3660, + "GCTACA": 3661, + "GCTACT": 3662, + "GCTACC": 3663, + "GCTACG": 3664, + "GCTAGA": 3665, + "GCTAGT": 3666, + "GCTAGC": 3667, + "GCTAGG": 3668, + "GCTTAA": 3669, + "GCTTAT": 3670, + "GCTTAC": 3671, + "GCTTAG": 3672, + "GCTTTA": 3673, + "GCTTTT": 3674, + "GCTTTC": 3675, + "GCTTTG": 3676, + "GCTTCA": 3677, + "GCTTCT": 3678, + "GCTTCC": 3679, + "GCTTCG": 3680, + "GCTTGA": 3681, + "GCTTGT": 3682, + "GCTTGC": 3683, + "GCTTGG": 3684, + "GCTCAA": 3685, + "GCTCAT": 3686, + "GCTCAC": 3687, + "GCTCAG": 3688, + "GCTCTA": 3689, + "GCTCTT": 3690, + "GCTCTC": 3691, + "GCTCTG": 3692, + "GCTCCA": 3693, + "GCTCCT": 3694, + "GCTCCC": 3695, + "GCTCCG": 3696, + "GCTCGA": 3697, + "GCTCGT": 3698, + "GCTCGC": 3699, + "GCTCGG": 3700, + "GCTGAA": 3701, + "GCTGAT": 3702, + "GCTGAC": 3703, + "GCTGAG": 3704, + "GCTGTA": 3705, + "GCTGTT": 3706, + "GCTGTC": 3707, + "GCTGTG": 3708, + "GCTGCA": 3709, + "GCTGCT": 3710, + "GCTGCC": 3711, + "GCTGCG": 3712, + "GCTGGA": 3713, + "GCTGGT": 3714, + "GCTGGC": 3715, + "GCTGGG": 3716, + "GCCAAA": 3717, + "GCCAAT": 3718, + "GCCAAC": 3719, + "GCCAAG": 3720, + "GCCATA": 3721, + "GCCATT": 3722, + "GCCATC": 3723, + "GCCATG": 3724, + "GCCACA": 3725, + "GCCACT": 3726, + "GCCACC": 3727, + "GCCACG": 3728, + "GCCAGA": 3729, + "GCCAGT": 3730, + "GCCAGC": 3731, + "GCCAGG": 3732, + "GCCTAA": 3733, + "GCCTAT": 3734, + "GCCTAC": 3735, + "GCCTAG": 3736, + "GCCTTA": 3737, + "GCCTTT": 3738, + "GCCTTC": 3739, + "GCCTTG": 3740, + "GCCTCA": 3741, + "GCCTCT": 3742, + "GCCTCC": 3743, + "GCCTCG": 3744, + "GCCTGA": 3745, + "GCCTGT": 3746, + "GCCTGC": 3747, + "GCCTGG": 3748, + "GCCCAA": 3749, + "GCCCAT": 3750, + "GCCCAC": 3751, + "GCCCAG": 3752, + "GCCCTA": 3753, + "GCCCTT": 3754, + "GCCCTC": 3755, + "GCCCTG": 3756, + "GCCCCA": 3757, + "GCCCCT": 3758, + "GCCCCC": 3759, + "GCCCCG": 3760, + "GCCCGA": 3761, + "GCCCGT": 3762, + "GCCCGC": 3763, + "GCCCGG": 3764, + "GCCGAA": 3765, + "GCCGAT": 3766, + "GCCGAC": 3767, + "GCCGAG": 3768, + "GCCGTA": 3769, + "GCCGTT": 3770, + "GCCGTC": 3771, + "GCCGTG": 3772, + "GCCGCA": 3773, + "GCCGCT": 3774, + "GCCGCC": 3775, + "GCCGCG": 3776, + "GCCGGA": 3777, + "GCCGGT": 3778, + "GCCGGC": 3779, + "GCCGGG": 3780, + "GCGAAA": 3781, + "GCGAAT": 3782, + "GCGAAC": 3783, + "GCGAAG": 3784, + "GCGATA": 3785, + "GCGATT": 3786, + "GCGATC": 3787, + "GCGATG": 3788, + "GCGACA": 3789, + "GCGACT": 3790, + "GCGACC": 3791, + "GCGACG": 3792, + "GCGAGA": 3793, + "GCGAGT": 3794, + "GCGAGC": 3795, + "GCGAGG": 3796, + "GCGTAA": 3797, + "GCGTAT": 3798, + "GCGTAC": 3799, + "GCGTAG": 3800, + "GCGTTA": 3801, + "GCGTTT": 3802, + "GCGTTC": 3803, + "GCGTTG": 3804, + "GCGTCA": 3805, + "GCGTCT": 3806, + "GCGTCC": 3807, + "GCGTCG": 3808, + "GCGTGA": 3809, + "GCGTGT": 3810, + "GCGTGC": 3811, + "GCGTGG": 3812, + "GCGCAA": 3813, + "GCGCAT": 3814, + "GCGCAC": 3815, + "GCGCAG": 3816, + "GCGCTA": 3817, + "GCGCTT": 3818, + "GCGCTC": 3819, + "GCGCTG": 3820, + "GCGCCA": 3821, + "GCGCCT": 3822, + "GCGCCC": 3823, + "GCGCCG": 3824, + "GCGCGA": 3825, + "GCGCGT": 3826, + "GCGCGC": 3827, + "GCGCGG": 3828, + "GCGGAA": 3829, + "GCGGAT": 3830, + "GCGGAC": 3831, + "GCGGAG": 3832, + "GCGGTA": 3833, + "GCGGTT": 3834, + "GCGGTC": 3835, + "GCGGTG": 3836, + "GCGGCA": 3837, + "GCGGCT": 3838, + "GCGGCC": 3839, + "GCGGCG": 3840, + "GCGGGA": 3841, + "GCGGGT": 3842, + "GCGGGC": 3843, + "GCGGGG": 3844, + "GGAAAA": 3845, + "GGAAAT": 3846, + "GGAAAC": 3847, + "GGAAAG": 3848, + "GGAATA": 3849, + "GGAATT": 3850, + "GGAATC": 3851, + "GGAATG": 3852, + "GGAACA": 3853, + "GGAACT": 3854, + "GGAACC": 3855, + "GGAACG": 3856, + "GGAAGA": 3857, + "GGAAGT": 3858, + "GGAAGC": 3859, + "GGAAGG": 3860, + "GGATAA": 3861, + "GGATAT": 3862, + "GGATAC": 3863, + "GGATAG": 3864, + "GGATTA": 3865, + "GGATTT": 3866, + "GGATTC": 3867, + "GGATTG": 3868, + "GGATCA": 3869, + "GGATCT": 3870, + "GGATCC": 3871, + "GGATCG": 3872, + "GGATGA": 3873, + "GGATGT": 3874, + "GGATGC": 3875, + "GGATGG": 3876, + "GGACAA": 3877, + "GGACAT": 3878, + "GGACAC": 3879, + "GGACAG": 3880, + "GGACTA": 3881, + "GGACTT": 3882, + "GGACTC": 3883, + "GGACTG": 3884, + "GGACCA": 3885, + "GGACCT": 3886, + "GGACCC": 3887, + "GGACCG": 3888, + "GGACGA": 3889, + "GGACGT": 3890, + "GGACGC": 3891, + "GGACGG": 3892, + "GGAGAA": 3893, + "GGAGAT": 3894, + "GGAGAC": 3895, + "GGAGAG": 3896, + "GGAGTA": 3897, + "GGAGTT": 3898, + "GGAGTC": 3899, + "GGAGTG": 3900, + "GGAGCA": 3901, + "GGAGCT": 3902, + "GGAGCC": 3903, + "GGAGCG": 3904, + "GGAGGA": 3905, + "GGAGGT": 3906, + "GGAGGC": 3907, + "GGAGGG": 3908, + "GGTAAA": 3909, + "GGTAAT": 3910, + "GGTAAC": 3911, + "GGTAAG": 3912, + "GGTATA": 3913, + "GGTATT": 3914, + "GGTATC": 3915, + "GGTATG": 3916, + "GGTACA": 3917, + "GGTACT": 3918, + "GGTACC": 3919, + "GGTACG": 3920, + "GGTAGA": 3921, + "GGTAGT": 3922, + "GGTAGC": 3923, + "GGTAGG": 3924, + "GGTTAA": 3925, + "GGTTAT": 3926, + "GGTTAC": 3927, + "GGTTAG": 3928, + "GGTTTA": 3929, + "GGTTTT": 3930, + "GGTTTC": 3931, + "GGTTTG": 3932, + "GGTTCA": 3933, + "GGTTCT": 3934, + "GGTTCC": 3935, + "GGTTCG": 3936, + "GGTTGA": 3937, + "GGTTGT": 3938, + "GGTTGC": 3939, + "GGTTGG": 3940, + "GGTCAA": 3941, + "GGTCAT": 3942, + "GGTCAC": 3943, + "GGTCAG": 3944, + "GGTCTA": 3945, + "GGTCTT": 3946, + "GGTCTC": 3947, + "GGTCTG": 3948, + "GGTCCA": 3949, + "GGTCCT": 3950, + "GGTCCC": 3951, + "GGTCCG": 3952, + "GGTCGA": 3953, + "GGTCGT": 3954, + "GGTCGC": 3955, + "GGTCGG": 3956, + "GGTGAA": 3957, + "GGTGAT": 3958, + "GGTGAC": 3959, + "GGTGAG": 3960, + "GGTGTA": 3961, + "GGTGTT": 3962, + "GGTGTC": 3963, + "GGTGTG": 3964, + "GGTGCA": 3965, + "GGTGCT": 3966, + "GGTGCC": 3967, + "GGTGCG": 3968, + "GGTGGA": 3969, + "GGTGGT": 3970, + "GGTGGC": 3971, + "GGTGGG": 3972, + "GGCAAA": 3973, + "GGCAAT": 3974, + "GGCAAC": 3975, + "GGCAAG": 3976, + "GGCATA": 3977, + "GGCATT": 3978, + "GGCATC": 3979, + "GGCATG": 3980, + "GGCACA": 3981, + "GGCACT": 3982, + "GGCACC": 3983, + "GGCACG": 3984, + "GGCAGA": 3985, + "GGCAGT": 3986, + "GGCAGC": 3987, + "GGCAGG": 3988, + "GGCTAA": 3989, + "GGCTAT": 3990, + "GGCTAC": 3991, + "GGCTAG": 3992, + "GGCTTA": 3993, + "GGCTTT": 3994, + "GGCTTC": 3995, + "GGCTTG": 3996, + "GGCTCA": 3997, + "GGCTCT": 3998, + "GGCTCC": 3999, + "GGCTCG": 4000, + "GGCTGA": 4001, + "GGCTGT": 4002, + "GGCTGC": 4003, + "GGCTGG": 4004, + "GGCCAA": 4005, + "GGCCAT": 4006, + "GGCCAC": 4007, + "GGCCAG": 4008, + "GGCCTA": 4009, + "GGCCTT": 4010, + "GGCCTC": 4011, + "GGCCTG": 4012, + "GGCCCA": 4013, + "GGCCCT": 4014, + "GGCCCC": 4015, + "GGCCCG": 4016, + "GGCCGA": 4017, + "GGCCGT": 4018, + "GGCCGC": 4019, + "GGCCGG": 4020, + "GGCGAA": 4021, + "GGCGAT": 4022, + "GGCGAC": 4023, + "GGCGAG": 4024, + "GGCGTA": 4025, + "GGCGTT": 4026, + "GGCGTC": 4027, + "GGCGTG": 4028, + "GGCGCA": 4029, + "GGCGCT": 4030, + "GGCGCC": 4031, + "GGCGCG": 4032, + "GGCGGA": 4033, + "GGCGGT": 4034, + "GGCGGC": 4035, + "GGCGGG": 4036, + "GGGAAA": 4037, + "GGGAAT": 4038, + "GGGAAC": 4039, + "GGGAAG": 4040, + "GGGATA": 4041, + "GGGATT": 4042, + "GGGATC": 4043, + "GGGATG": 4044, + "GGGACA": 4045, + "GGGACT": 4046, + "GGGACC": 4047, + "GGGACG": 4048, + "GGGAGA": 4049, + "GGGAGT": 4050, + "GGGAGC": 4051, + "GGGAGG": 4052, + "GGGTAA": 4053, + "GGGTAT": 4054, + "GGGTAC": 4055, + "GGGTAG": 4056, + "GGGTTA": 4057, + "GGGTTT": 4058, + "GGGTTC": 4059, + "GGGTTG": 4060, + "GGGTCA": 4061, + "GGGTCT": 4062, + "GGGTCC": 4063, + "GGGTCG": 4064, + "GGGTGA": 4065, + "GGGTGT": 4066, + "GGGTGC": 4067, + "GGGTGG": 4068, + "GGGCAA": 4069, + "GGGCAT": 4070, + "GGGCAC": 4071, + "GGGCAG": 4072, + "GGGCTA": 4073, + "GGGCTT": 4074, + "GGGCTC": 4075, + "GGGCTG": 4076, + "GGGCCA": 4077, + "GGGCCT": 4078, + "GGGCCC": 4079, + "GGGCCG": 4080, + "GGGCGA": 4081, + "GGGCGT": 4082, + "GGGCGC": 4083, + "GGGCGG": 4084, + "GGGGAA": 4085, + "GGGGAT": 4086, + "GGGGAC": 4087, + "GGGGAG": 4088, + "GGGGTA": 4089, + "GGGGTT": 4090, + "GGGGTC": 4091, + "GGGGTG": 4092, + "GGGGCA": 4093, + "GGGGCT": 4094, + "GGGGCC": 4095, + "GGGGCG": 4096, + "GGGGGA": 4097, + "GGGGGT": 4098, + "GGGGGC": 4099, + "GGGGGG": 4100 + } + } +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6250/tokenizer_config.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6250/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..604e2b1b29af1dcb3920d2828b6a90276e5e3d02 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6250/tokenizer_config.json @@ -0,0 +1,58 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "[UNK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "[CLS]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "[SEP]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "4": { + "content": "[MASK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "clean_up_tokenization_spaces": true, + "cls_token": "[CLS]", + "do_basic_tokenize": true, + "do_lower_case": false, + "mask_token": "[MASK]", + "max_len": 512, + "model_max_length": 512, + "never_split": null, + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "strip_accents": null, + "tokenize_chinese_chars": true, + "tokenizer_class": "BertTokenizer", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6250/trainer_state.json b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6250/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..3903b9a82a92b40fb403739b65c2becc05a9d6e0 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6250/trainer_state.json @@ -0,0 +1,1783 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 25.0, + "eval_steps": 500, + "global_step": 6250, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.1, + "grad_norm": 0.6038870215415955, + "learning_rate": 5.319148936170213e-05, + "loss": 0.6945, + "step": 25 + }, + { + "epoch": 0.2, + "grad_norm": 0.6668091416358948, + "learning_rate": 0.00010638297872340425, + "loss": 0.6929, + "step": 50 + }, + { + "epoch": 0.3, + "grad_norm": 0.8282511234283447, + "learning_rate": 0.00015957446808510637, + "loss": 0.6925, + "step": 75 + }, + { + "epoch": 0.4, + "grad_norm": 1.3149235248565674, + "learning_rate": 0.0002127659574468085, + "loss": 0.6919, + "step": 100 + }, + { + "epoch": 0.5, + "grad_norm": 1.8841172456741333, + "learning_rate": 0.00026595744680851064, + "loss": 0.6788, + "step": 125 + }, + { + "epoch": 0.6, + "grad_norm": 4.435206413269043, + "learning_rate": 0.00031914893617021275, + "loss": 0.6285, + "step": 150 + }, + { + "epoch": 0.7, + "grad_norm": 8.858148574829102, + "learning_rate": 0.0003723404255319149, + "loss": 0.5386, + "step": 175 + }, + { + "epoch": 0.8, + "grad_norm": 5.627648830413818, + "learning_rate": 0.0003992081821181128, + "loss": 0.5537, + "step": 200 + }, + { + "epoch": 0.9, + "grad_norm": 5.086580753326416, + "learning_rate": 0.0003975585615308479, + "loss": 0.5595, + "step": 225 + }, + { + "epoch": 1.0, + "grad_norm": 5.627719402313232, + "learning_rate": 0.00039590894094358297, + "loss": 0.4824, + "step": 250 + }, + { + "epoch": 1.1, + "grad_norm": 5.888038158416748, + "learning_rate": 0.0003942593203563181, + "loss": 0.4404, + "step": 275 + }, + { + "epoch": 1.2, + "grad_norm": 5.646231651306152, + "learning_rate": 0.0003926096997690532, + "loss": 0.4354, + "step": 300 + }, + { + "epoch": 1.3, + "grad_norm": 5.191990852355957, + "learning_rate": 0.00039096007918178817, + "loss": 0.4469, + "step": 325 + }, + { + "epoch": 1.4, + "grad_norm": 13.90101432800293, + "learning_rate": 0.0003893104585945233, + "loss": 0.3858, + "step": 350 + }, + { + "epoch": 1.5, + "grad_norm": 17.75942039489746, + "learning_rate": 0.0003876608380072583, + "loss": 0.4131, + "step": 375 + }, + { + "epoch": 1.6, + "grad_norm": 7.398252487182617, + "learning_rate": 0.00038601121741999343, + "loss": 0.3842, + "step": 400 + }, + { + "epoch": 1.7, + "grad_norm": 11.010066986083984, + "learning_rate": 0.0003843615968327285, + "loss": 0.3959, + "step": 425 + }, + { + "epoch": 1.8, + "grad_norm": 9.277341842651367, + "learning_rate": 0.0003827119762454636, + "loss": 0.3552, + "step": 450 + }, + { + "epoch": 1.9, + "grad_norm": 26.011449813842773, + "learning_rate": 0.00038106235565819863, + "loss": 0.3961, + "step": 475 + }, + { + "epoch": 2.0, + "grad_norm": 11.123054504394531, + "learning_rate": 0.0003794127350709337, + "loss": 0.3548, + "step": 500 + }, + { + "epoch": 2.1, + "grad_norm": 20.128774642944336, + "learning_rate": 0.0003777631144836688, + "loss": 0.302, + "step": 525 + }, + { + "epoch": 2.2, + "grad_norm": 4.64390230178833, + "learning_rate": 0.00037611349389640383, + "loss": 0.2853, + "step": 550 + }, + { + "epoch": 2.3, + "grad_norm": 4.199934482574463, + "learning_rate": 0.00037446387330913894, + "loss": 0.2724, + "step": 575 + }, + { + "epoch": 2.4, + "grad_norm": 6.909411430358887, + "learning_rate": 0.000372814252721874, + "loss": 0.2976, + "step": 600 + }, + { + "epoch": 2.5, + "grad_norm": 9.234820365905762, + "learning_rate": 0.00037116463213460903, + "loss": 0.2798, + "step": 625 + }, + { + "epoch": 2.6, + "grad_norm": 10.472943305969238, + "learning_rate": 0.00036951501154734414, + "loss": 0.3007, + "step": 650 + }, + { + "epoch": 2.7, + "grad_norm": 19.46732521057129, + "learning_rate": 0.0003678653909600792, + "loss": 0.2701, + "step": 675 + }, + { + "epoch": 2.8, + "grad_norm": 7.501255512237549, + "learning_rate": 0.0003662157703728143, + "loss": 0.2896, + "step": 700 + }, + { + "epoch": 2.9, + "grad_norm": 10.645585060119629, + "learning_rate": 0.00036456614978554934, + "loss": 0.2748, + "step": 725 + }, + { + "epoch": 3.0, + "grad_norm": 30.26616859436035, + "learning_rate": 0.00036291652919828444, + "loss": 0.268, + "step": 750 + }, + { + "epoch": 3.1, + "grad_norm": 14.13516616821289, + "learning_rate": 0.00036126690861101944, + "loss": 0.2159, + "step": 775 + }, + { + "epoch": 3.2, + "grad_norm": 10.091023445129395, + "learning_rate": 0.00035961728802375454, + "loss": 0.2071, + "step": 800 + }, + { + "epoch": 3.3, + "grad_norm": 22.201452255249023, + "learning_rate": 0.0003579676674364896, + "loss": 0.2146, + "step": 825 + }, + { + "epoch": 3.4, + "grad_norm": 10.044412612915039, + "learning_rate": 0.0003563180468492247, + "loss": 0.1827, + "step": 850 + }, + { + "epoch": 3.5, + "grad_norm": 21.04941177368164, + "learning_rate": 0.0003546684262619598, + "loss": 0.202, + "step": 875 + }, + { + "epoch": 3.6, + "grad_norm": 11.754891395568848, + "learning_rate": 0.00035301880567469485, + "loss": 0.2021, + "step": 900 + }, + { + "epoch": 3.7, + "grad_norm": 9.113747596740723, + "learning_rate": 0.0003513691850874299, + "loss": 0.2067, + "step": 925 + }, + { + "epoch": 3.8, + "grad_norm": 9.075196266174316, + "learning_rate": 0.00034971956450016495, + "loss": 0.3285, + "step": 950 + }, + { + "epoch": 3.9, + "grad_norm": 8.568796157836914, + "learning_rate": 0.00034806994391290005, + "loss": 0.2254, + "step": 975 + }, + { + "epoch": 4.0, + "grad_norm": 3.9587976932525635, + "learning_rate": 0.0003464203233256351, + "loss": 0.2245, + "step": 1000 + }, + { + "epoch": 4.1, + "grad_norm": 11.143651008605957, + "learning_rate": 0.0003447707027383702, + "loss": 0.1755, + "step": 1025 + }, + { + "epoch": 4.2, + "grad_norm": 12.53118896484375, + "learning_rate": 0.00034312108215110525, + "loss": 0.1808, + "step": 1050 + }, + { + "epoch": 4.3, + "grad_norm": 9.880820274353027, + "learning_rate": 0.0003414714615638403, + "loss": 0.1387, + "step": 1075 + }, + { + "epoch": 4.4, + "grad_norm": 5.729593753814697, + "learning_rate": 0.0003398218409765754, + "loss": 0.1958, + "step": 1100 + }, + { + "epoch": 4.5, + "grad_norm": 8.633047103881836, + "learning_rate": 0.00033817222038931045, + "loss": 0.1515, + "step": 1125 + }, + { + "epoch": 4.6, + "grad_norm": 19.237892150878906, + "learning_rate": 0.00033652259980204556, + "loss": 0.1452, + "step": 1150 + }, + { + "epoch": 4.7, + "grad_norm": 6.129818439483643, + "learning_rate": 0.0003348729792147806, + "loss": 0.1617, + "step": 1175 + }, + { + "epoch": 4.8, + "grad_norm": 5.073274612426758, + "learning_rate": 0.0003332233586275157, + "loss": 0.1967, + "step": 1200 + }, + { + "epoch": 4.9, + "grad_norm": 5.729922771453857, + "learning_rate": 0.00033157373804025076, + "loss": 0.1482, + "step": 1225 + }, + { + "epoch": 5.0, + "grad_norm": 12.314518928527832, + "learning_rate": 0.0003299241174529858, + "loss": 0.1588, + "step": 1250 + }, + { + "epoch": 5.1, + "grad_norm": 3.2636606693267822, + "learning_rate": 0.0003282744968657209, + "loss": 0.1588, + "step": 1275 + }, + { + "epoch": 5.2, + "grad_norm": 5.914486408233643, + "learning_rate": 0.00032662487627845596, + "loss": 0.1288, + "step": 1300 + }, + { + "epoch": 5.3, + "grad_norm": 13.849296569824219, + "learning_rate": 0.00032497525569119106, + "loss": 0.0961, + "step": 1325 + }, + { + "epoch": 5.4, + "grad_norm": 7.7317328453063965, + "learning_rate": 0.0003233256351039261, + "loss": 0.1148, + "step": 1350 + }, + { + "epoch": 5.5, + "grad_norm": 1.768608808517456, + "learning_rate": 0.00032167601451666116, + "loss": 0.121, + "step": 1375 + }, + { + "epoch": 5.6, + "grad_norm": 10.802785873413086, + "learning_rate": 0.00032002639392939627, + "loss": 0.1786, + "step": 1400 + }, + { + "epoch": 5.7, + "grad_norm": 4.158301830291748, + "learning_rate": 0.0003183767733421313, + "loss": 0.1078, + "step": 1425 + }, + { + "epoch": 5.8, + "grad_norm": 10.65804672241211, + "learning_rate": 0.0003167271527548664, + "loss": 0.1597, + "step": 1450 + }, + { + "epoch": 5.9, + "grad_norm": 10.880438804626465, + "learning_rate": 0.00031507753216760147, + "loss": 0.1225, + "step": 1475 + }, + { + "epoch": 6.0, + "grad_norm": 9.95138168334961, + "learning_rate": 0.00031342791158033657, + "loss": 0.1254, + "step": 1500 + }, + { + "epoch": 6.1, + "grad_norm": 1.5018162727355957, + "learning_rate": 0.00031177829099307157, + "loss": 0.0902, + "step": 1525 + }, + { + "epoch": 6.2, + "grad_norm": 14.281820297241211, + "learning_rate": 0.00031012867040580667, + "loss": 0.1026, + "step": 1550 + }, + { + "epoch": 6.3, + "grad_norm": 4.7540435791015625, + "learning_rate": 0.0003084790498185417, + "loss": 0.0876, + "step": 1575 + }, + { + "epoch": 6.4, + "grad_norm": 4.304183006286621, + "learning_rate": 0.0003068294292312768, + "loss": 0.1355, + "step": 1600 + }, + { + "epoch": 6.5, + "grad_norm": 14.653755187988281, + "learning_rate": 0.0003051798086440119, + "loss": 0.0878, + "step": 1625 + }, + { + "epoch": 6.6, + "grad_norm": 11.13028335571289, + "learning_rate": 0.000303530188056747, + "loss": 0.1203, + "step": 1650 + }, + { + "epoch": 6.7, + "grad_norm": 9.729938507080078, + "learning_rate": 0.000301880567469482, + "loss": 0.0775, + "step": 1675 + }, + { + "epoch": 6.8, + "grad_norm": 10.12938117980957, + "learning_rate": 0.0003002309468822171, + "loss": 0.1015, + "step": 1700 + }, + { + "epoch": 6.9, + "grad_norm": 8.506614685058594, + "learning_rate": 0.0002985813262949522, + "loss": 0.1362, + "step": 1725 + }, + { + "epoch": 7.0, + "grad_norm": 25.618762969970703, + "learning_rate": 0.00029693170570768723, + "loss": 0.1269, + "step": 1750 + }, + { + "epoch": 7.1, + "grad_norm": 27.155860900878906, + "learning_rate": 0.00029528208512042233, + "loss": 0.0555, + "step": 1775 + }, + { + "epoch": 7.2, + "grad_norm": 9.412527084350586, + "learning_rate": 0.0002936324645331574, + "loss": 0.0695, + "step": 1800 + }, + { + "epoch": 7.3, + "grad_norm": 6.4328293800354, + "learning_rate": 0.00029198284394589243, + "loss": 0.0808, + "step": 1825 + }, + { + "epoch": 7.4, + "grad_norm": 15.267293930053711, + "learning_rate": 0.00029033322335862753, + "loss": 0.0871, + "step": 1850 + }, + { + "epoch": 7.5, + "grad_norm": 6.774560928344727, + "learning_rate": 0.0002886836027713626, + "loss": 0.0819, + "step": 1875 + }, + { + "epoch": 7.6, + "grad_norm": 18.67180633544922, + "learning_rate": 0.0002870339821840977, + "loss": 0.0565, + "step": 1900 + }, + { + "epoch": 7.7, + "grad_norm": 16.39004898071289, + "learning_rate": 0.00028538436159683273, + "loss": 0.0907, + "step": 1925 + }, + { + "epoch": 7.8, + "grad_norm": 6.351583003997803, + "learning_rate": 0.00028373474100956784, + "loss": 0.1289, + "step": 1950 + }, + { + "epoch": 7.9, + "grad_norm": 11.396740913391113, + "learning_rate": 0.0002820851204223029, + "loss": 0.1149, + "step": 1975 + }, + { + "epoch": 8.0, + "grad_norm": 6.491570949554443, + "learning_rate": 0.00028043549983503794, + "loss": 0.0891, + "step": 2000 + }, + { + "epoch": 8.1, + "grad_norm": 0.22946956753730774, + "learning_rate": 0.00027878587924777304, + "loss": 0.0413, + "step": 2025 + }, + { + "epoch": 8.2, + "grad_norm": 42.31606674194336, + "learning_rate": 0.0002771362586605081, + "loss": 0.0881, + "step": 2050 + }, + { + "epoch": 8.3, + "grad_norm": 18.490406036376953, + "learning_rate": 0.0002754866380732432, + "loss": 0.1238, + "step": 2075 + }, + { + "epoch": 8.4, + "grad_norm": 17.950319290161133, + "learning_rate": 0.00027383701748597824, + "loss": 0.0642, + "step": 2100 + }, + { + "epoch": 8.5, + "grad_norm": 37.16359329223633, + "learning_rate": 0.0002721873968987133, + "loss": 0.0916, + "step": 2125 + }, + { + "epoch": 8.6, + "grad_norm": 39.75199890136719, + "learning_rate": 0.00027053777631144834, + "loss": 0.1087, + "step": 2150 + }, + { + "epoch": 8.7, + "grad_norm": 0.08794238418340683, + "learning_rate": 0.00026888815572418344, + "loss": 0.0825, + "step": 2175 + }, + { + "epoch": 8.8, + "grad_norm": 14.681429862976074, + "learning_rate": 0.00026723853513691855, + "loss": 0.0925, + "step": 2200 + }, + { + "epoch": 8.9, + "grad_norm": 17.90182876586914, + "learning_rate": 0.0002655889145496536, + "loss": 0.0933, + "step": 2225 + }, + { + "epoch": 9.0, + "grad_norm": 0.9108687043190002, + "learning_rate": 0.0002639392939623887, + "loss": 0.0602, + "step": 2250 + }, + { + "epoch": 9.1, + "grad_norm": 9.671142578125, + "learning_rate": 0.0002622896733751237, + "loss": 0.0438, + "step": 2275 + }, + { + "epoch": 9.2, + "grad_norm": 12.797035217285156, + "learning_rate": 0.0002606400527878588, + "loss": 0.0495, + "step": 2300 + }, + { + "epoch": 9.3, + "grad_norm": 0.4585617184638977, + "learning_rate": 0.00025899043220059385, + "loss": 0.032, + "step": 2325 + }, + { + "epoch": 9.4, + "grad_norm": 0.05073042958974838, + "learning_rate": 0.00025734081161332895, + "loss": 0.067, + "step": 2350 + }, + { + "epoch": 9.5, + "grad_norm": 0.013653877191245556, + "learning_rate": 0.000255691191026064, + "loss": 0.0898, + "step": 2375 + }, + { + "epoch": 9.6, + "grad_norm": 18.198965072631836, + "learning_rate": 0.0002540415704387991, + "loss": 0.0804, + "step": 2400 + }, + { + "epoch": 9.7, + "grad_norm": 0.1363959163427353, + "learning_rate": 0.00025239194985153415, + "loss": 0.0584, + "step": 2425 + }, + { + "epoch": 9.8, + "grad_norm": 17.61290740966797, + "learning_rate": 0.0002507423292642692, + "loss": 0.0778, + "step": 2450 + }, + { + "epoch": 9.9, + "grad_norm": 0.10548887401819229, + "learning_rate": 0.0002490927086770043, + "loss": 0.0667, + "step": 2475 + }, + { + "epoch": 10.0, + "grad_norm": 0.1262211948633194, + "learning_rate": 0.00024744308808973936, + "loss": 0.0625, + "step": 2500 + }, + { + "epoch": 10.1, + "grad_norm": 0.019293297082185745, + "learning_rate": 0.00024579346750247446, + "loss": 0.0762, + "step": 2525 + }, + { + "epoch": 10.2, + "grad_norm": 0.9273566603660583, + "learning_rate": 0.0002441438469152095, + "loss": 0.0357, + "step": 2550 + }, + { + "epoch": 10.3, + "grad_norm": 10.720550537109375, + "learning_rate": 0.00024249422632794456, + "loss": 0.046, + "step": 2575 + }, + { + "epoch": 10.4, + "grad_norm": 1.8441611528396606, + "learning_rate": 0.00024084460574067963, + "loss": 0.0369, + "step": 2600 + }, + { + "epoch": 10.5, + "grad_norm": 0.8760197758674622, + "learning_rate": 0.0002391949851534147, + "loss": 0.046, + "step": 2625 + }, + { + "epoch": 10.6, + "grad_norm": 8.899499893188477, + "learning_rate": 0.0002375453645661498, + "loss": 0.08, + "step": 2650 + }, + { + "epoch": 10.7, + "grad_norm": 0.2664288580417633, + "learning_rate": 0.00023589574397888486, + "loss": 0.0616, + "step": 2675 + }, + { + "epoch": 10.8, + "grad_norm": 0.06609692424535751, + "learning_rate": 0.00023424612339161997, + "loss": 0.05, + "step": 2700 + }, + { + "epoch": 10.9, + "grad_norm": 32.44829177856445, + "learning_rate": 0.000232596502804355, + "loss": 0.0487, + "step": 2725 + }, + { + "epoch": 11.0, + "grad_norm": 7.25365686416626, + "learning_rate": 0.00023094688221709007, + "loss": 0.0456, + "step": 2750 + }, + { + "epoch": 11.1, + "grad_norm": 0.10101272910833359, + "learning_rate": 0.00022929726162982514, + "loss": 0.0328, + "step": 2775 + }, + { + "epoch": 11.2, + "grad_norm": 1.3087996244430542, + "learning_rate": 0.00022764764104256022, + "loss": 0.0606, + "step": 2800 + }, + { + "epoch": 11.3, + "grad_norm": 1.247656226158142, + "learning_rate": 0.0002259980204552953, + "loss": 0.0838, + "step": 2825 + }, + { + "epoch": 11.4, + "grad_norm": 1.7708797454833984, + "learning_rate": 0.00022434839986803037, + "loss": 0.0416, + "step": 2850 + }, + { + "epoch": 11.5, + "grad_norm": 11.131721496582031, + "learning_rate": 0.00022269877928076542, + "loss": 0.0528, + "step": 2875 + }, + { + "epoch": 11.6, + "grad_norm": 3.824920892715454, + "learning_rate": 0.0002210491586935005, + "loss": 0.0526, + "step": 2900 + }, + { + "epoch": 11.7, + "grad_norm": 0.8507682085037231, + "learning_rate": 0.00021939953810623557, + "loss": 0.0965, + "step": 2925 + }, + { + "epoch": 11.8, + "grad_norm": 0.011997719295322895, + "learning_rate": 0.00021774991751897065, + "loss": 0.0446, + "step": 2950 + }, + { + "epoch": 11.9, + "grad_norm": 0.0792650505900383, + "learning_rate": 0.00021610029693170573, + "loss": 0.0556, + "step": 2975 + }, + { + "epoch": 12.0, + "grad_norm": 0.007047992665320635, + "learning_rate": 0.0002144506763444408, + "loss": 0.0759, + "step": 3000 + }, + { + "epoch": 12.1, + "grad_norm": 0.7784481644630432, + "learning_rate": 0.00021280105575717582, + "loss": 0.0505, + "step": 3025 + }, + { + "epoch": 12.2, + "grad_norm": 9.635031700134277, + "learning_rate": 0.00021115143516991093, + "loss": 0.0284, + "step": 3050 + }, + { + "epoch": 12.3, + "grad_norm": 12.473057746887207, + "learning_rate": 0.000209501814582646, + "loss": 0.0531, + "step": 3075 + }, + { + "epoch": 12.4, + "grad_norm": 10.904062271118164, + "learning_rate": 0.00020785219399538108, + "loss": 0.0212, + "step": 3100 + }, + { + "epoch": 12.5, + "grad_norm": 9.258865356445312, + "learning_rate": 0.00020620257340811616, + "loss": 0.0367, + "step": 3125 + }, + { + "epoch": 12.6, + "grad_norm": 2.6826488971710205, + "learning_rate": 0.00020455295282085123, + "loss": 0.0457, + "step": 3150 + }, + { + "epoch": 12.7, + "grad_norm": 8.800697326660156, + "learning_rate": 0.00020290333223358626, + "loss": 0.0504, + "step": 3175 + }, + { + "epoch": 12.8, + "grad_norm": 0.023783667013049126, + "learning_rate": 0.00020125371164632133, + "loss": 0.0255, + "step": 3200 + }, + { + "epoch": 12.9, + "grad_norm": 5.948949337005615, + "learning_rate": 0.0001996040910590564, + "loss": 0.0413, + "step": 3225 + }, + { + "epoch": 13.0, + "grad_norm": 0.026645230129361153, + "learning_rate": 0.00019795447047179148, + "loss": 0.0249, + "step": 3250 + }, + { + "epoch": 13.1, + "grad_norm": 60.826637268066406, + "learning_rate": 0.0001963048498845266, + "loss": 0.0293, + "step": 3275 + }, + { + "epoch": 13.2, + "grad_norm": 26.056686401367188, + "learning_rate": 0.00019465522929726164, + "loss": 0.0638, + "step": 3300 + }, + { + "epoch": 13.3, + "grad_norm": 18.143638610839844, + "learning_rate": 0.00019300560870999671, + "loss": 0.0206, + "step": 3325 + }, + { + "epoch": 13.4, + "grad_norm": 27.899967193603516, + "learning_rate": 0.0001913559881227318, + "loss": 0.0423, + "step": 3350 + }, + { + "epoch": 13.5, + "grad_norm": 51.9879264831543, + "learning_rate": 0.00018970636753546684, + "loss": 0.0274, + "step": 3375 + }, + { + "epoch": 13.6, + "grad_norm": 0.028627146035432816, + "learning_rate": 0.00018805674694820192, + "loss": 0.0513, + "step": 3400 + }, + { + "epoch": 13.7, + "grad_norm": 33.211944580078125, + "learning_rate": 0.000186407126360937, + "loss": 0.0446, + "step": 3425 + }, + { + "epoch": 13.8, + "grad_norm": 32.37785339355469, + "learning_rate": 0.00018475750577367207, + "loss": 0.0538, + "step": 3450 + }, + { + "epoch": 13.9, + "grad_norm": 30.558673858642578, + "learning_rate": 0.00018310788518640715, + "loss": 0.0339, + "step": 3475 + }, + { + "epoch": 14.0, + "grad_norm": 19.316938400268555, + "learning_rate": 0.00018145826459914222, + "loss": 0.071, + "step": 3500 + }, + { + "epoch": 14.1, + "grad_norm": 0.1271055042743683, + "learning_rate": 0.00017980864401187727, + "loss": 0.0159, + "step": 3525 + }, + { + "epoch": 14.2, + "grad_norm": 51.1268196105957, + "learning_rate": 0.00017815902342461235, + "loss": 0.028, + "step": 3550 + }, + { + "epoch": 14.3, + "grad_norm": 0.003214583033695817, + "learning_rate": 0.00017650940283734742, + "loss": 0.0342, + "step": 3575 + }, + { + "epoch": 14.4, + "grad_norm": 94.75691223144531, + "learning_rate": 0.00017485978225008247, + "loss": 0.0243, + "step": 3600 + }, + { + "epoch": 14.5, + "grad_norm": 0.019844651222229004, + "learning_rate": 0.00017321016166281755, + "loss": 0.0205, + "step": 3625 + }, + { + "epoch": 14.6, + "grad_norm": 0.10069809854030609, + "learning_rate": 0.00017156054107555263, + "loss": 0.0402, + "step": 3650 + }, + { + "epoch": 14.7, + "grad_norm": 0.0023937481455504894, + "learning_rate": 0.0001699109204882877, + "loss": 0.0173, + "step": 3675 + }, + { + "epoch": 14.8, + "grad_norm": 0.2573629319667816, + "learning_rate": 0.00016826129990102278, + "loss": 0.0312, + "step": 3700 + }, + { + "epoch": 14.9, + "grad_norm": 0.0067031364887952805, + "learning_rate": 0.00016661167931375785, + "loss": 0.0418, + "step": 3725 + }, + { + "epoch": 15.0, + "grad_norm": 0.08329538255929947, + "learning_rate": 0.0001649620587264929, + "loss": 0.0164, + "step": 3750 + }, + { + "epoch": 15.1, + "grad_norm": 0.04175519943237305, + "learning_rate": 0.00016331243813922798, + "loss": 0.044, + "step": 3775 + }, + { + "epoch": 15.2, + "grad_norm": 0.005420651752501726, + "learning_rate": 0.00016166281755196306, + "loss": 0.0599, + "step": 3800 + }, + { + "epoch": 15.3, + "grad_norm": 20.68973731994629, + "learning_rate": 0.00016001319696469813, + "loss": 0.047, + "step": 3825 + }, + { + "epoch": 15.4, + "grad_norm": 1.1241929531097412, + "learning_rate": 0.0001583635763774332, + "loss": 0.0184, + "step": 3850 + }, + { + "epoch": 15.5, + "grad_norm": 0.00928444042801857, + "learning_rate": 0.00015671395579016829, + "loss": 0.0146, + "step": 3875 + }, + { + "epoch": 15.6, + "grad_norm": 3.2419674396514893, + "learning_rate": 0.00015506433520290334, + "loss": 0.0047, + "step": 3900 + }, + { + "epoch": 15.7, + "grad_norm": 0.0019528602715581656, + "learning_rate": 0.0001534147146156384, + "loss": 0.0177, + "step": 3925 + }, + { + "epoch": 15.8, + "grad_norm": 0.003142274683341384, + "learning_rate": 0.0001517650940283735, + "loss": 0.0319, + "step": 3950 + }, + { + "epoch": 15.9, + "grad_norm": 45.58081817626953, + "learning_rate": 0.00015011547344110854, + "loss": 0.0353, + "step": 3975 + }, + { + "epoch": 16.0, + "grad_norm": 0.003515947377309203, + "learning_rate": 0.00014846585285384361, + "loss": 0.0208, + "step": 4000 + }, + { + "epoch": 16.1, + "grad_norm": 0.01574423350393772, + "learning_rate": 0.0001468162322665787, + "loss": 0.0198, + "step": 4025 + }, + { + "epoch": 16.2, + "grad_norm": 22.461671829223633, + "learning_rate": 0.00014516661167931377, + "loss": 0.0262, + "step": 4050 + }, + { + "epoch": 16.3, + "grad_norm": 17.334550857543945, + "learning_rate": 0.00014351699109204884, + "loss": 0.0567, + "step": 4075 + }, + { + "epoch": 16.4, + "grad_norm": 0.00328136351890862, + "learning_rate": 0.00014186737050478392, + "loss": 0.0182, + "step": 4100 + }, + { + "epoch": 16.5, + "grad_norm": 1.2645615339279175, + "learning_rate": 0.00014021774991751897, + "loss": 0.0116, + "step": 4125 + }, + { + "epoch": 16.6, + "grad_norm": 0.0031980539206415415, + "learning_rate": 0.00013856812933025404, + "loss": 0.0079, + "step": 4150 + }, + { + "epoch": 16.7, + "grad_norm": 0.01752198301255703, + "learning_rate": 0.00013691850874298912, + "loss": 0.0576, + "step": 4175 + }, + { + "epoch": 16.8, + "grad_norm": 48.27894592285156, + "learning_rate": 0.00013526888815572417, + "loss": 0.0264, + "step": 4200 + }, + { + "epoch": 16.9, + "grad_norm": 0.008802427910268307, + "learning_rate": 0.00013361926756845927, + "loss": 0.0051, + "step": 4225 + }, + { + "epoch": 17.0, + "grad_norm": 0.0350751169025898, + "learning_rate": 0.00013196964698119435, + "loss": 0.0403, + "step": 4250 + }, + { + "epoch": 17.1, + "grad_norm": 0.0020171860232949257, + "learning_rate": 0.0001303200263939294, + "loss": 0.0012, + "step": 4275 + }, + { + "epoch": 17.2, + "grad_norm": 0.014922365546226501, + "learning_rate": 0.00012867040580666448, + "loss": 0.036, + "step": 4300 + }, + { + "epoch": 17.3, + "grad_norm": 0.01983438804745674, + "learning_rate": 0.00012702078521939955, + "loss": 0.0094, + "step": 4325 + }, + { + "epoch": 17.4, + "grad_norm": 0.0031381379812955856, + "learning_rate": 0.0001253711646321346, + "loss": 0.011, + "step": 4350 + }, + { + "epoch": 17.5, + "grad_norm": 0.656819224357605, + "learning_rate": 0.00012372154404486968, + "loss": 0.0094, + "step": 4375 + }, + { + "epoch": 17.6, + "grad_norm": 0.008435660973191261, + "learning_rate": 0.00012207192345760475, + "loss": 0.0445, + "step": 4400 + }, + { + "epoch": 17.7, + "grad_norm": 23.819961547851562, + "learning_rate": 0.00012042230287033982, + "loss": 0.0394, + "step": 4425 + }, + { + "epoch": 17.8, + "grad_norm": 0.0055985236540436745, + "learning_rate": 0.0001187726822830749, + "loss": 0.0106, + "step": 4450 + }, + { + "epoch": 17.9, + "grad_norm": 11.063871383666992, + "learning_rate": 0.00011712306169580998, + "loss": 0.0128, + "step": 4475 + }, + { + "epoch": 18.0, + "grad_norm": 0.003767346730455756, + "learning_rate": 0.00011547344110854503, + "loss": 0.0096, + "step": 4500 + }, + { + "epoch": 18.1, + "grad_norm": 0.018198227509856224, + "learning_rate": 0.00011382382052128011, + "loss": 0.0395, + "step": 4525 + }, + { + "epoch": 18.2, + "grad_norm": 0.004393396433442831, + "learning_rate": 0.00011217419993401519, + "loss": 0.0378, + "step": 4550 + }, + { + "epoch": 18.3, + "grad_norm": 0.01423373818397522, + "learning_rate": 0.00011052457934675025, + "loss": 0.013, + "step": 4575 + }, + { + "epoch": 18.4, + "grad_norm": 25.492542266845703, + "learning_rate": 0.00010887495875948532, + "loss": 0.0094, + "step": 4600 + }, + { + "epoch": 18.5, + "grad_norm": 31.223493576049805, + "learning_rate": 0.0001072253381722204, + "loss": 0.0161, + "step": 4625 + }, + { + "epoch": 18.6, + "grad_norm": 0.003598716575652361, + "learning_rate": 0.00010557571758495546, + "loss": 0.0106, + "step": 4650 + }, + { + "epoch": 18.7, + "grad_norm": 50.249603271484375, + "learning_rate": 0.00010392609699769054, + "loss": 0.024, + "step": 4675 + }, + { + "epoch": 18.8, + "grad_norm": 63.50950622558594, + "learning_rate": 0.00010227647641042562, + "loss": 0.018, + "step": 4700 + }, + { + "epoch": 18.9, + "grad_norm": 0.002336911391466856, + "learning_rate": 0.00010062685582316067, + "loss": 0.0091, + "step": 4725 + }, + { + "epoch": 19.0, + "grad_norm": 0.005011886823922396, + "learning_rate": 9.897723523589574e-05, + "loss": 0.0061, + "step": 4750 + }, + { + "epoch": 19.1, + "grad_norm": 0.027857890352606773, + "learning_rate": 9.732761464863082e-05, + "loss": 0.0221, + "step": 4775 + }, + { + "epoch": 19.2, + "grad_norm": 0.0016227838350459933, + "learning_rate": 9.56779940613659e-05, + "loss": 0.0195, + "step": 4800 + }, + { + "epoch": 19.3, + "grad_norm": 0.005687279161065817, + "learning_rate": 9.402837347410096e-05, + "loss": 0.0203, + "step": 4825 + }, + { + "epoch": 19.4, + "grad_norm": 0.009174875915050507, + "learning_rate": 9.237875288683603e-05, + "loss": 0.0086, + "step": 4850 + }, + { + "epoch": 19.5, + "grad_norm": 0.0032144756987690926, + "learning_rate": 9.072913229957111e-05, + "loss": 0.0289, + "step": 4875 + }, + { + "epoch": 19.6, + "grad_norm": 0.003061591414734721, + "learning_rate": 8.907951171230617e-05, + "loss": 0.0015, + "step": 4900 + }, + { + "epoch": 19.7, + "grad_norm": 0.006960512604564428, + "learning_rate": 8.742989112504124e-05, + "loss": 0.0176, + "step": 4925 + }, + { + "epoch": 19.8, + "grad_norm": 0.0027057609986513853, + "learning_rate": 8.578027053777631e-05, + "loss": 0.0047, + "step": 4950 + }, + { + "epoch": 19.9, + "grad_norm": 0.22216364741325378, + "learning_rate": 8.413064995051139e-05, + "loss": 0.0099, + "step": 4975 + }, + { + "epoch": 20.0, + "grad_norm": 0.0022015271242707968, + "learning_rate": 8.248102936324645e-05, + "loss": 0.0219, + "step": 5000 + }, + { + "epoch": 20.1, + "grad_norm": 0.019280511885881424, + "learning_rate": 8.083140877598153e-05, + "loss": 0.0129, + "step": 5025 + }, + { + "epoch": 20.2, + "grad_norm": 0.005342870019376278, + "learning_rate": 7.91817881887166e-05, + "loss": 0.016, + "step": 5050 + }, + { + "epoch": 20.3, + "grad_norm": 0.013655520975589752, + "learning_rate": 7.753216760145167e-05, + "loss": 0.0174, + "step": 5075 + }, + { + "epoch": 20.4, + "grad_norm": 0.0037927369121462107, + "learning_rate": 7.588254701418674e-05, + "loss": 0.021, + "step": 5100 + }, + { + "epoch": 20.5, + "grad_norm": 0.0036746144760400057, + "learning_rate": 7.423292642692181e-05, + "loss": 0.004, + "step": 5125 + }, + { + "epoch": 20.6, + "grad_norm": 0.014823048375546932, + "learning_rate": 7.258330583965688e-05, + "loss": 0.0161, + "step": 5150 + }, + { + "epoch": 20.7, + "grad_norm": 0.0033355404157191515, + "learning_rate": 7.093368525239196e-05, + "loss": 0.0035, + "step": 5175 + }, + { + "epoch": 20.8, + "grad_norm": 0.0018589649116620421, + "learning_rate": 6.928406466512702e-05, + "loss": 0.0105, + "step": 5200 + }, + { + "epoch": 20.9, + "grad_norm": 0.013408798724412918, + "learning_rate": 6.763444407786209e-05, + "loss": 0.0001, + "step": 5225 + }, + { + "epoch": 21.0, + "grad_norm": 70.37215423583984, + "learning_rate": 6.598482349059718e-05, + "loss": 0.0419, + "step": 5250 + }, + { + "epoch": 21.1, + "grad_norm": 0.09737736731767654, + "learning_rate": 6.433520290333224e-05, + "loss": 0.0146, + "step": 5275 + }, + { + "epoch": 21.2, + "grad_norm": 0.012592921033501625, + "learning_rate": 6.26855823160673e-05, + "loss": 0.0333, + "step": 5300 + }, + { + "epoch": 21.3, + "grad_norm": 0.0022445865906774998, + "learning_rate": 6.103596172880238e-05, + "loss": 0.0241, + "step": 5325 + }, + { + "epoch": 21.4, + "grad_norm": 0.01678585074841976, + "learning_rate": 5.938634114153745e-05, + "loss": 0.0047, + "step": 5350 + }, + { + "epoch": 21.5, + "grad_norm": 0.07437444478273392, + "learning_rate": 5.7736720554272516e-05, + "loss": 0.0213, + "step": 5375 + }, + { + "epoch": 21.6, + "grad_norm": 0.002057570032775402, + "learning_rate": 5.608709996700759e-05, + "loss": 0.0001, + "step": 5400 + }, + { + "epoch": 21.7, + "grad_norm": 0.4772427976131439, + "learning_rate": 5.443747937974266e-05, + "loss": 0.0038, + "step": 5425 + }, + { + "epoch": 21.8, + "grad_norm": 0.0016257546376436949, + "learning_rate": 5.278785879247773e-05, + "loss": 0.0076, + "step": 5450 + }, + { + "epoch": 21.9, + "grad_norm": 0.02936253882944584, + "learning_rate": 5.113823820521281e-05, + "loss": 0.0004, + "step": 5475 + }, + { + "epoch": 22.0, + "grad_norm": 0.12796524167060852, + "learning_rate": 4.948861761794787e-05, + "loss": 0.011, + "step": 5500 + }, + { + "epoch": 22.1, + "grad_norm": 0.0015867372276261449, + "learning_rate": 4.783899703068295e-05, + "loss": 0.0066, + "step": 5525 + }, + { + "epoch": 22.2, + "grad_norm": 54.19424819946289, + "learning_rate": 4.618937644341802e-05, + "loss": 0.0102, + "step": 5550 + }, + { + "epoch": 22.3, + "grad_norm": 0.07260770350694656, + "learning_rate": 4.453975585615309e-05, + "loss": 0.0016, + "step": 5575 + }, + { + "epoch": 22.4, + "grad_norm": 0.004677828401327133, + "learning_rate": 4.2890135268888156e-05, + "loss": 0.0005, + "step": 5600 + }, + { + "epoch": 22.5, + "grad_norm": 0.0017690203385427594, + "learning_rate": 4.1240514681623226e-05, + "loss": 0.0084, + "step": 5625 + }, + { + "epoch": 22.6, + "grad_norm": 0.0025316874962300062, + "learning_rate": 3.95908940943583e-05, + "loss": 0.006, + "step": 5650 + }, + { + "epoch": 22.7, + "grad_norm": 0.5509907007217407, + "learning_rate": 3.794127350709337e-05, + "loss": 0.001, + "step": 5675 + }, + { + "epoch": 22.8, + "grad_norm": 0.002628073561936617, + "learning_rate": 3.629165291982844e-05, + "loss": 0.0129, + "step": 5700 + }, + { + "epoch": 22.9, + "grad_norm": 0.004700347315520048, + "learning_rate": 3.464203233256351e-05, + "loss": 0.0002, + "step": 5725 + }, + { + "epoch": 23.0, + "grad_norm": 0.0015652852598577738, + "learning_rate": 3.299241174529859e-05, + "loss": 0.0001, + "step": 5750 + }, + { + "epoch": 23.1, + "grad_norm": 0.0014234065311029553, + "learning_rate": 3.134279115803365e-05, + "loss": 0.0002, + "step": 5775 + }, + { + "epoch": 23.2, + "grad_norm": 0.001536093419417739, + "learning_rate": 2.9693170570768723e-05, + "loss": 0.0007, + "step": 5800 + }, + { + "epoch": 23.3, + "grad_norm": 0.002034024801105261, + "learning_rate": 2.8043549983503796e-05, + "loss": 0.0069, + "step": 5825 + }, + { + "epoch": 23.4, + "grad_norm": 0.001170129980891943, + "learning_rate": 2.6393929396238866e-05, + "loss": 0.0182, + "step": 5850 + }, + { + "epoch": 23.5, + "grad_norm": 1.254554271697998, + "learning_rate": 2.4744308808973936e-05, + "loss": 0.0002, + "step": 5875 + }, + { + "epoch": 23.6, + "grad_norm": 0.0014751444105058908, + "learning_rate": 2.309468822170901e-05, + "loss": 0.0001, + "step": 5900 + }, + { + "epoch": 23.7, + "grad_norm": 1.3905378580093384, + "learning_rate": 2.1445067634444078e-05, + "loss": 0.0018, + "step": 5925 + }, + { + "epoch": 23.8, + "grad_norm": 0.001244895625859499, + "learning_rate": 1.979544704717915e-05, + "loss": 0.0076, + "step": 5950 + }, + { + "epoch": 23.9, + "grad_norm": 0.0016068242257460952, + "learning_rate": 1.814582645991422e-05, + "loss": 0.0017, + "step": 5975 + }, + { + "epoch": 24.0, + "grad_norm": 0.0011312123388051987, + "learning_rate": 1.6496205872649294e-05, + "loss": 0.0007, + "step": 6000 + }, + { + "epoch": 24.1, + "grad_norm": 0.0016296221874654293, + "learning_rate": 1.4846585285384362e-05, + "loss": 0.0011, + "step": 6025 + }, + { + "epoch": 24.2, + "grad_norm": 0.0028211891185492277, + "learning_rate": 1.3196964698119433e-05, + "loss": 0.0074, + "step": 6050 + }, + { + "epoch": 24.3, + "grad_norm": 41.444637298583984, + "learning_rate": 1.1547344110854504e-05, + "loss": 0.0068, + "step": 6075 + }, + { + "epoch": 24.4, + "grad_norm": 0.0014097857056185603, + "learning_rate": 9.897723523589576e-06, + "loss": 0.0026, + "step": 6100 + }, + { + "epoch": 24.5, + "grad_norm": 0.002990308217704296, + "learning_rate": 8.248102936324647e-06, + "loss": 0.0004, + "step": 6125 + }, + { + "epoch": 24.6, + "grad_norm": 0.0019173379987478256, + "learning_rate": 6.5984823490597165e-06, + "loss": 0.0141, + "step": 6150 + }, + { + "epoch": 24.7, + "grad_norm": 0.0025832897517830133, + "learning_rate": 4.948861761794788e-06, + "loss": 0.0001, + "step": 6175 + }, + { + "epoch": 24.8, + "grad_norm": 0.0010854025604203343, + "learning_rate": 3.2992411745298582e-06, + "loss": 0.0016, + "step": 6200 + }, + { + "epoch": 24.9, + "grad_norm": 0.0013974973699077964, + "learning_rate": 1.6496205872649291e-06, + "loss": 0.0001, + "step": 6225 + }, + { + "epoch": 25.0, + "grad_norm": 0.001211145892739296, + "learning_rate": 0.0, + "loss": 0.0004, + "step": 6250 + } + ], + "logging_steps": 25, + "max_steps": 6250, + "num_input_tokens_seen": 0, + "num_train_epochs": 25, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 317621760000000.0, + "train_batch_size": 20, + "trial_name": null, + "trial_params": null +} diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6250/training_args.bin b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6250/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..77692b6666e83a688b38e799c964e4694a57eec9 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6250/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f73b41315068a430d2439f5707c62152a34210df62454b9549f9642446dd90b2 +size 5304 diff --git a/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6250/vocab.txt b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6250/vocab.txt new file mode 100644 index 0000000000000000000000000000000000000000..cc9ad53732e21ea9180f18e1fb67a38cc5702d4c --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnaprom/TinyDNABERT_base_model/checkpoint-6250/vocab.txt @@ -0,0 +1,4101 @@ +[PAD] +[UNK] +[CLS] +[SEP] +[MASK] +AAAAAA +AAAAAT +AAAAAC +AAAAAG +AAAATA +AAAATT +AAAATC +AAAATG +AAAACA +AAAACT +AAAACC +AAAACG +AAAAGA +AAAAGT +AAAAGC +AAAAGG +AAATAA +AAATAT +AAATAC +AAATAG +AAATTA +AAATTT +AAATTC +AAATTG +AAATCA +AAATCT +AAATCC +AAATCG +AAATGA +AAATGT +AAATGC +AAATGG +AAACAA +AAACAT +AAACAC +AAACAG +AAACTA +AAACTT +AAACTC +AAACTG +AAACCA +AAACCT +AAACCC +AAACCG +AAACGA +AAACGT +AAACGC +AAACGG +AAAGAA +AAAGAT +AAAGAC +AAAGAG +AAAGTA +AAAGTT +AAAGTC +AAAGTG +AAAGCA +AAAGCT +AAAGCC +AAAGCG +AAAGGA +AAAGGT +AAAGGC +AAAGGG +AATAAA +AATAAT +AATAAC +AATAAG +AATATA +AATATT +AATATC +AATATG +AATACA +AATACT +AATACC +AATACG +AATAGA +AATAGT +AATAGC +AATAGG +AATTAA +AATTAT +AATTAC +AATTAG +AATTTA +AATTTT +AATTTC +AATTTG +AATTCA +AATTCT +AATTCC +AATTCG +AATTGA +AATTGT +AATTGC +AATTGG +AATCAA +AATCAT +AATCAC +AATCAG +AATCTA +AATCTT +AATCTC +AATCTG +AATCCA +AATCCT +AATCCC +AATCCG +AATCGA +AATCGT +AATCGC +AATCGG +AATGAA +AATGAT +AATGAC +AATGAG +AATGTA +AATGTT +AATGTC +AATGTG +AATGCA +AATGCT +AATGCC +AATGCG +AATGGA +AATGGT +AATGGC +AATGGG +AACAAA +AACAAT +AACAAC +AACAAG +AACATA +AACATT +AACATC +AACATG +AACACA +AACACT +AACACC +AACACG +AACAGA +AACAGT +AACAGC +AACAGG +AACTAA +AACTAT +AACTAC +AACTAG +AACTTA +AACTTT +AACTTC +AACTTG +AACTCA +AACTCT +AACTCC +AACTCG +AACTGA +AACTGT +AACTGC +AACTGG +AACCAA +AACCAT +AACCAC +AACCAG +AACCTA +AACCTT +AACCTC +AACCTG +AACCCA +AACCCT +AACCCC +AACCCG +AACCGA +AACCGT +AACCGC +AACCGG +AACGAA +AACGAT +AACGAC +AACGAG +AACGTA +AACGTT +AACGTC +AACGTG +AACGCA +AACGCT +AACGCC +AACGCG +AACGGA +AACGGT +AACGGC +AACGGG +AAGAAA +AAGAAT +AAGAAC +AAGAAG +AAGATA +AAGATT +AAGATC +AAGATG +AAGACA +AAGACT +AAGACC +AAGACG +AAGAGA +AAGAGT +AAGAGC +AAGAGG +AAGTAA +AAGTAT +AAGTAC +AAGTAG +AAGTTA +AAGTTT +AAGTTC +AAGTTG +AAGTCA +AAGTCT +AAGTCC +AAGTCG +AAGTGA +AAGTGT +AAGTGC +AAGTGG +AAGCAA +AAGCAT +AAGCAC +AAGCAG +AAGCTA +AAGCTT +AAGCTC +AAGCTG +AAGCCA +AAGCCT +AAGCCC +AAGCCG +AAGCGA +AAGCGT +AAGCGC +AAGCGG +AAGGAA +AAGGAT +AAGGAC +AAGGAG +AAGGTA +AAGGTT +AAGGTC +AAGGTG +AAGGCA +AAGGCT +AAGGCC +AAGGCG +AAGGGA +AAGGGT +AAGGGC +AAGGGG +ATAAAA +ATAAAT +ATAAAC +ATAAAG +ATAATA +ATAATT +ATAATC +ATAATG +ATAACA +ATAACT +ATAACC +ATAACG +ATAAGA +ATAAGT +ATAAGC +ATAAGG +ATATAA +ATATAT +ATATAC +ATATAG +ATATTA +ATATTT +ATATTC +ATATTG +ATATCA +ATATCT +ATATCC +ATATCG +ATATGA +ATATGT +ATATGC +ATATGG +ATACAA +ATACAT +ATACAC +ATACAG +ATACTA +ATACTT +ATACTC +ATACTG +ATACCA +ATACCT +ATACCC +ATACCG +ATACGA +ATACGT +ATACGC +ATACGG +ATAGAA +ATAGAT +ATAGAC +ATAGAG +ATAGTA +ATAGTT +ATAGTC +ATAGTG +ATAGCA +ATAGCT +ATAGCC +ATAGCG +ATAGGA +ATAGGT +ATAGGC +ATAGGG +ATTAAA +ATTAAT +ATTAAC +ATTAAG +ATTATA +ATTATT +ATTATC +ATTATG +ATTACA +ATTACT +ATTACC +ATTACG +ATTAGA +ATTAGT +ATTAGC +ATTAGG +ATTTAA +ATTTAT +ATTTAC +ATTTAG +ATTTTA +ATTTTT +ATTTTC +ATTTTG +ATTTCA +ATTTCT +ATTTCC +ATTTCG +ATTTGA +ATTTGT +ATTTGC +ATTTGG +ATTCAA +ATTCAT +ATTCAC +ATTCAG +ATTCTA +ATTCTT +ATTCTC +ATTCTG +ATTCCA +ATTCCT +ATTCCC +ATTCCG +ATTCGA +ATTCGT +ATTCGC +ATTCGG +ATTGAA +ATTGAT +ATTGAC +ATTGAG +ATTGTA +ATTGTT +ATTGTC +ATTGTG +ATTGCA +ATTGCT +ATTGCC +ATTGCG +ATTGGA +ATTGGT +ATTGGC +ATTGGG +ATCAAA +ATCAAT +ATCAAC +ATCAAG +ATCATA +ATCATT +ATCATC +ATCATG +ATCACA +ATCACT +ATCACC +ATCACG +ATCAGA +ATCAGT +ATCAGC +ATCAGG +ATCTAA +ATCTAT +ATCTAC +ATCTAG +ATCTTA +ATCTTT +ATCTTC +ATCTTG +ATCTCA +ATCTCT +ATCTCC +ATCTCG +ATCTGA +ATCTGT +ATCTGC +ATCTGG +ATCCAA +ATCCAT +ATCCAC +ATCCAG +ATCCTA +ATCCTT +ATCCTC +ATCCTG +ATCCCA +ATCCCT +ATCCCC +ATCCCG +ATCCGA +ATCCGT +ATCCGC +ATCCGG +ATCGAA +ATCGAT +ATCGAC +ATCGAG +ATCGTA +ATCGTT +ATCGTC +ATCGTG +ATCGCA +ATCGCT +ATCGCC +ATCGCG +ATCGGA +ATCGGT +ATCGGC +ATCGGG +ATGAAA +ATGAAT +ATGAAC +ATGAAG +ATGATA +ATGATT +ATGATC +ATGATG +ATGACA +ATGACT +ATGACC +ATGACG +ATGAGA +ATGAGT +ATGAGC +ATGAGG +ATGTAA +ATGTAT +ATGTAC +ATGTAG +ATGTTA +ATGTTT +ATGTTC +ATGTTG +ATGTCA +ATGTCT +ATGTCC +ATGTCG +ATGTGA +ATGTGT +ATGTGC +ATGTGG +ATGCAA +ATGCAT +ATGCAC +ATGCAG +ATGCTA +ATGCTT +ATGCTC +ATGCTG +ATGCCA +ATGCCT +ATGCCC +ATGCCG +ATGCGA +ATGCGT +ATGCGC +ATGCGG +ATGGAA +ATGGAT +ATGGAC +ATGGAG +ATGGTA +ATGGTT +ATGGTC +ATGGTG +ATGGCA +ATGGCT +ATGGCC +ATGGCG +ATGGGA +ATGGGT +ATGGGC +ATGGGG +ACAAAA +ACAAAT +ACAAAC +ACAAAG +ACAATA +ACAATT +ACAATC +ACAATG +ACAACA +ACAACT +ACAACC +ACAACG +ACAAGA +ACAAGT +ACAAGC +ACAAGG +ACATAA +ACATAT +ACATAC +ACATAG +ACATTA +ACATTT +ACATTC +ACATTG +ACATCA +ACATCT +ACATCC +ACATCG +ACATGA +ACATGT +ACATGC +ACATGG +ACACAA +ACACAT +ACACAC +ACACAG +ACACTA +ACACTT +ACACTC +ACACTG +ACACCA +ACACCT +ACACCC +ACACCG +ACACGA +ACACGT +ACACGC +ACACGG +ACAGAA +ACAGAT +ACAGAC +ACAGAG +ACAGTA +ACAGTT +ACAGTC +ACAGTG +ACAGCA +ACAGCT +ACAGCC +ACAGCG +ACAGGA +ACAGGT +ACAGGC +ACAGGG +ACTAAA +ACTAAT +ACTAAC +ACTAAG +ACTATA +ACTATT +ACTATC +ACTATG +ACTACA +ACTACT +ACTACC +ACTACG +ACTAGA +ACTAGT +ACTAGC +ACTAGG +ACTTAA +ACTTAT +ACTTAC +ACTTAG +ACTTTA +ACTTTT +ACTTTC +ACTTTG +ACTTCA +ACTTCT +ACTTCC +ACTTCG +ACTTGA +ACTTGT +ACTTGC +ACTTGG +ACTCAA +ACTCAT +ACTCAC +ACTCAG +ACTCTA +ACTCTT +ACTCTC +ACTCTG +ACTCCA +ACTCCT +ACTCCC +ACTCCG +ACTCGA +ACTCGT +ACTCGC +ACTCGG +ACTGAA +ACTGAT +ACTGAC +ACTGAG +ACTGTA +ACTGTT +ACTGTC +ACTGTG +ACTGCA +ACTGCT +ACTGCC +ACTGCG +ACTGGA +ACTGGT +ACTGGC +ACTGGG +ACCAAA +ACCAAT +ACCAAC +ACCAAG +ACCATA +ACCATT +ACCATC +ACCATG +ACCACA +ACCACT +ACCACC +ACCACG +ACCAGA +ACCAGT +ACCAGC +ACCAGG +ACCTAA +ACCTAT +ACCTAC +ACCTAG +ACCTTA +ACCTTT +ACCTTC +ACCTTG +ACCTCA +ACCTCT +ACCTCC +ACCTCG +ACCTGA +ACCTGT +ACCTGC +ACCTGG +ACCCAA +ACCCAT +ACCCAC +ACCCAG +ACCCTA +ACCCTT +ACCCTC +ACCCTG +ACCCCA +ACCCCT +ACCCCC +ACCCCG +ACCCGA +ACCCGT +ACCCGC +ACCCGG +ACCGAA +ACCGAT +ACCGAC +ACCGAG +ACCGTA +ACCGTT +ACCGTC +ACCGTG +ACCGCA +ACCGCT +ACCGCC +ACCGCG +ACCGGA +ACCGGT +ACCGGC +ACCGGG +ACGAAA +ACGAAT +ACGAAC +ACGAAG +ACGATA +ACGATT +ACGATC +ACGATG +ACGACA +ACGACT +ACGACC +ACGACG +ACGAGA +ACGAGT +ACGAGC +ACGAGG +ACGTAA +ACGTAT +ACGTAC +ACGTAG +ACGTTA +ACGTTT +ACGTTC +ACGTTG +ACGTCA +ACGTCT +ACGTCC +ACGTCG +ACGTGA +ACGTGT +ACGTGC +ACGTGG +ACGCAA +ACGCAT +ACGCAC +ACGCAG +ACGCTA +ACGCTT +ACGCTC +ACGCTG +ACGCCA +ACGCCT +ACGCCC +ACGCCG +ACGCGA +ACGCGT +ACGCGC +ACGCGG +ACGGAA +ACGGAT +ACGGAC +ACGGAG +ACGGTA +ACGGTT +ACGGTC +ACGGTG +ACGGCA +ACGGCT +ACGGCC +ACGGCG +ACGGGA +ACGGGT +ACGGGC +ACGGGG +AGAAAA +AGAAAT +AGAAAC +AGAAAG +AGAATA +AGAATT +AGAATC +AGAATG +AGAACA +AGAACT +AGAACC +AGAACG +AGAAGA +AGAAGT +AGAAGC +AGAAGG +AGATAA +AGATAT +AGATAC +AGATAG +AGATTA +AGATTT +AGATTC +AGATTG +AGATCA +AGATCT +AGATCC +AGATCG +AGATGA +AGATGT +AGATGC +AGATGG +AGACAA +AGACAT +AGACAC +AGACAG +AGACTA +AGACTT +AGACTC +AGACTG +AGACCA +AGACCT +AGACCC +AGACCG +AGACGA +AGACGT +AGACGC +AGACGG +AGAGAA +AGAGAT +AGAGAC +AGAGAG +AGAGTA +AGAGTT +AGAGTC +AGAGTG +AGAGCA +AGAGCT +AGAGCC +AGAGCG +AGAGGA +AGAGGT +AGAGGC +AGAGGG +AGTAAA +AGTAAT +AGTAAC +AGTAAG +AGTATA +AGTATT +AGTATC +AGTATG +AGTACA +AGTACT +AGTACC +AGTACG +AGTAGA +AGTAGT +AGTAGC +AGTAGG +AGTTAA +AGTTAT +AGTTAC +AGTTAG +AGTTTA +AGTTTT +AGTTTC +AGTTTG +AGTTCA +AGTTCT +AGTTCC +AGTTCG +AGTTGA +AGTTGT +AGTTGC +AGTTGG +AGTCAA +AGTCAT +AGTCAC +AGTCAG +AGTCTA +AGTCTT +AGTCTC +AGTCTG +AGTCCA +AGTCCT +AGTCCC +AGTCCG +AGTCGA +AGTCGT +AGTCGC +AGTCGG +AGTGAA +AGTGAT +AGTGAC +AGTGAG +AGTGTA +AGTGTT +AGTGTC +AGTGTG +AGTGCA +AGTGCT +AGTGCC +AGTGCG +AGTGGA +AGTGGT +AGTGGC +AGTGGG +AGCAAA +AGCAAT +AGCAAC +AGCAAG +AGCATA +AGCATT +AGCATC +AGCATG +AGCACA +AGCACT +AGCACC +AGCACG +AGCAGA +AGCAGT +AGCAGC +AGCAGG +AGCTAA +AGCTAT +AGCTAC +AGCTAG +AGCTTA +AGCTTT +AGCTTC +AGCTTG +AGCTCA +AGCTCT +AGCTCC +AGCTCG +AGCTGA +AGCTGT +AGCTGC +AGCTGG +AGCCAA +AGCCAT +AGCCAC +AGCCAG +AGCCTA +AGCCTT +AGCCTC +AGCCTG +AGCCCA +AGCCCT +AGCCCC +AGCCCG +AGCCGA +AGCCGT +AGCCGC +AGCCGG +AGCGAA +AGCGAT +AGCGAC +AGCGAG +AGCGTA +AGCGTT +AGCGTC +AGCGTG +AGCGCA +AGCGCT +AGCGCC +AGCGCG +AGCGGA +AGCGGT +AGCGGC +AGCGGG +AGGAAA +AGGAAT +AGGAAC +AGGAAG +AGGATA +AGGATT +AGGATC +AGGATG +AGGACA +AGGACT +AGGACC +AGGACG +AGGAGA +AGGAGT +AGGAGC +AGGAGG +AGGTAA +AGGTAT +AGGTAC +AGGTAG +AGGTTA +AGGTTT +AGGTTC +AGGTTG +AGGTCA +AGGTCT +AGGTCC +AGGTCG +AGGTGA +AGGTGT +AGGTGC +AGGTGG +AGGCAA +AGGCAT +AGGCAC +AGGCAG +AGGCTA +AGGCTT +AGGCTC +AGGCTG +AGGCCA +AGGCCT +AGGCCC +AGGCCG +AGGCGA +AGGCGT +AGGCGC +AGGCGG +AGGGAA +AGGGAT +AGGGAC +AGGGAG +AGGGTA +AGGGTT +AGGGTC +AGGGTG +AGGGCA +AGGGCT +AGGGCC +AGGGCG +AGGGGA +AGGGGT +AGGGGC +AGGGGG +TAAAAA +TAAAAT +TAAAAC +TAAAAG +TAAATA +TAAATT +TAAATC +TAAATG +TAAACA +TAAACT +TAAACC +TAAACG +TAAAGA +TAAAGT +TAAAGC +TAAAGG +TAATAA +TAATAT +TAATAC +TAATAG +TAATTA +TAATTT +TAATTC +TAATTG +TAATCA +TAATCT +TAATCC +TAATCG +TAATGA +TAATGT +TAATGC +TAATGG +TAACAA +TAACAT +TAACAC +TAACAG +TAACTA +TAACTT +TAACTC +TAACTG +TAACCA +TAACCT +TAACCC +TAACCG +TAACGA +TAACGT +TAACGC +TAACGG +TAAGAA +TAAGAT +TAAGAC +TAAGAG +TAAGTA +TAAGTT +TAAGTC +TAAGTG +TAAGCA +TAAGCT +TAAGCC +TAAGCG +TAAGGA +TAAGGT +TAAGGC +TAAGGG +TATAAA +TATAAT +TATAAC +TATAAG +TATATA +TATATT +TATATC +TATATG +TATACA +TATACT +TATACC +TATACG +TATAGA +TATAGT +TATAGC +TATAGG +TATTAA +TATTAT +TATTAC +TATTAG +TATTTA +TATTTT +TATTTC +TATTTG +TATTCA +TATTCT +TATTCC +TATTCG +TATTGA +TATTGT +TATTGC +TATTGG +TATCAA +TATCAT +TATCAC +TATCAG +TATCTA +TATCTT +TATCTC +TATCTG +TATCCA +TATCCT +TATCCC +TATCCG +TATCGA +TATCGT +TATCGC +TATCGG +TATGAA +TATGAT +TATGAC +TATGAG +TATGTA +TATGTT +TATGTC +TATGTG +TATGCA +TATGCT +TATGCC +TATGCG +TATGGA +TATGGT +TATGGC +TATGGG +TACAAA +TACAAT +TACAAC +TACAAG +TACATA +TACATT +TACATC +TACATG +TACACA +TACACT +TACACC +TACACG +TACAGA +TACAGT +TACAGC +TACAGG +TACTAA +TACTAT +TACTAC +TACTAG +TACTTA +TACTTT +TACTTC +TACTTG +TACTCA +TACTCT +TACTCC +TACTCG +TACTGA +TACTGT +TACTGC +TACTGG +TACCAA +TACCAT +TACCAC +TACCAG +TACCTA +TACCTT +TACCTC +TACCTG +TACCCA +TACCCT +TACCCC +TACCCG +TACCGA +TACCGT +TACCGC +TACCGG +TACGAA +TACGAT +TACGAC +TACGAG +TACGTA +TACGTT +TACGTC +TACGTG +TACGCA +TACGCT +TACGCC +TACGCG +TACGGA +TACGGT +TACGGC +TACGGG +TAGAAA +TAGAAT +TAGAAC +TAGAAG +TAGATA +TAGATT +TAGATC +TAGATG +TAGACA +TAGACT +TAGACC +TAGACG +TAGAGA +TAGAGT +TAGAGC +TAGAGG +TAGTAA +TAGTAT +TAGTAC +TAGTAG +TAGTTA +TAGTTT +TAGTTC +TAGTTG +TAGTCA +TAGTCT +TAGTCC +TAGTCG +TAGTGA +TAGTGT +TAGTGC +TAGTGG +TAGCAA +TAGCAT +TAGCAC +TAGCAG +TAGCTA +TAGCTT +TAGCTC +TAGCTG +TAGCCA +TAGCCT +TAGCCC +TAGCCG +TAGCGA +TAGCGT +TAGCGC +TAGCGG +TAGGAA +TAGGAT +TAGGAC +TAGGAG +TAGGTA +TAGGTT +TAGGTC +TAGGTG +TAGGCA +TAGGCT +TAGGCC +TAGGCG +TAGGGA +TAGGGT +TAGGGC +TAGGGG +TTAAAA +TTAAAT +TTAAAC +TTAAAG +TTAATA +TTAATT +TTAATC +TTAATG +TTAACA +TTAACT +TTAACC +TTAACG +TTAAGA +TTAAGT +TTAAGC +TTAAGG +TTATAA +TTATAT +TTATAC +TTATAG +TTATTA +TTATTT +TTATTC +TTATTG +TTATCA +TTATCT +TTATCC +TTATCG +TTATGA +TTATGT +TTATGC +TTATGG +TTACAA +TTACAT +TTACAC +TTACAG +TTACTA +TTACTT +TTACTC +TTACTG +TTACCA +TTACCT +TTACCC +TTACCG +TTACGA +TTACGT +TTACGC +TTACGG +TTAGAA +TTAGAT +TTAGAC +TTAGAG +TTAGTA +TTAGTT +TTAGTC +TTAGTG +TTAGCA +TTAGCT +TTAGCC +TTAGCG +TTAGGA +TTAGGT +TTAGGC +TTAGGG +TTTAAA +TTTAAT +TTTAAC +TTTAAG +TTTATA +TTTATT +TTTATC +TTTATG +TTTACA +TTTACT +TTTACC +TTTACG +TTTAGA +TTTAGT +TTTAGC +TTTAGG +TTTTAA +TTTTAT +TTTTAC +TTTTAG +TTTTTA +TTTTTT +TTTTTC +TTTTTG +TTTTCA +TTTTCT +TTTTCC +TTTTCG +TTTTGA +TTTTGT +TTTTGC +TTTTGG +TTTCAA +TTTCAT +TTTCAC +TTTCAG +TTTCTA +TTTCTT +TTTCTC +TTTCTG +TTTCCA +TTTCCT +TTTCCC +TTTCCG +TTTCGA +TTTCGT +TTTCGC +TTTCGG +TTTGAA +TTTGAT +TTTGAC +TTTGAG +TTTGTA +TTTGTT +TTTGTC +TTTGTG +TTTGCA +TTTGCT +TTTGCC +TTTGCG +TTTGGA +TTTGGT +TTTGGC +TTTGGG +TTCAAA +TTCAAT +TTCAAC +TTCAAG +TTCATA +TTCATT +TTCATC +TTCATG +TTCACA +TTCACT +TTCACC +TTCACG +TTCAGA +TTCAGT +TTCAGC +TTCAGG +TTCTAA +TTCTAT +TTCTAC +TTCTAG +TTCTTA +TTCTTT +TTCTTC +TTCTTG +TTCTCA +TTCTCT +TTCTCC +TTCTCG +TTCTGA +TTCTGT +TTCTGC +TTCTGG +TTCCAA +TTCCAT +TTCCAC +TTCCAG +TTCCTA +TTCCTT +TTCCTC +TTCCTG +TTCCCA +TTCCCT +TTCCCC +TTCCCG +TTCCGA +TTCCGT +TTCCGC +TTCCGG +TTCGAA +TTCGAT +TTCGAC +TTCGAG +TTCGTA +TTCGTT +TTCGTC +TTCGTG +TTCGCA +TTCGCT +TTCGCC +TTCGCG +TTCGGA +TTCGGT +TTCGGC +TTCGGG +TTGAAA +TTGAAT +TTGAAC +TTGAAG +TTGATA +TTGATT +TTGATC +TTGATG +TTGACA +TTGACT +TTGACC +TTGACG +TTGAGA +TTGAGT +TTGAGC +TTGAGG +TTGTAA +TTGTAT +TTGTAC +TTGTAG +TTGTTA +TTGTTT +TTGTTC +TTGTTG +TTGTCA +TTGTCT +TTGTCC +TTGTCG +TTGTGA +TTGTGT +TTGTGC +TTGTGG +TTGCAA +TTGCAT +TTGCAC +TTGCAG +TTGCTA +TTGCTT +TTGCTC +TTGCTG +TTGCCA +TTGCCT +TTGCCC +TTGCCG +TTGCGA +TTGCGT +TTGCGC +TTGCGG +TTGGAA +TTGGAT +TTGGAC +TTGGAG +TTGGTA +TTGGTT +TTGGTC +TTGGTG +TTGGCA +TTGGCT +TTGGCC +TTGGCG +TTGGGA +TTGGGT +TTGGGC +TTGGGG +TCAAAA +TCAAAT +TCAAAC +TCAAAG +TCAATA +TCAATT +TCAATC +TCAATG +TCAACA +TCAACT +TCAACC +TCAACG +TCAAGA +TCAAGT +TCAAGC +TCAAGG +TCATAA +TCATAT +TCATAC +TCATAG +TCATTA +TCATTT +TCATTC +TCATTG +TCATCA +TCATCT +TCATCC +TCATCG +TCATGA +TCATGT +TCATGC +TCATGG +TCACAA +TCACAT +TCACAC +TCACAG +TCACTA +TCACTT +TCACTC +TCACTG +TCACCA +TCACCT +TCACCC +TCACCG +TCACGA +TCACGT +TCACGC +TCACGG +TCAGAA +TCAGAT +TCAGAC +TCAGAG +TCAGTA +TCAGTT +TCAGTC +TCAGTG +TCAGCA +TCAGCT +TCAGCC +TCAGCG +TCAGGA +TCAGGT +TCAGGC +TCAGGG +TCTAAA +TCTAAT +TCTAAC +TCTAAG +TCTATA +TCTATT +TCTATC +TCTATG +TCTACA +TCTACT +TCTACC +TCTACG +TCTAGA +TCTAGT +TCTAGC +TCTAGG +TCTTAA +TCTTAT +TCTTAC +TCTTAG +TCTTTA +TCTTTT +TCTTTC +TCTTTG +TCTTCA +TCTTCT +TCTTCC +TCTTCG +TCTTGA +TCTTGT +TCTTGC +TCTTGG +TCTCAA +TCTCAT +TCTCAC +TCTCAG +TCTCTA +TCTCTT +TCTCTC +TCTCTG +TCTCCA +TCTCCT +TCTCCC +TCTCCG +TCTCGA +TCTCGT +TCTCGC +TCTCGG +TCTGAA +TCTGAT +TCTGAC +TCTGAG +TCTGTA +TCTGTT +TCTGTC +TCTGTG +TCTGCA +TCTGCT +TCTGCC +TCTGCG +TCTGGA +TCTGGT +TCTGGC +TCTGGG +TCCAAA +TCCAAT +TCCAAC +TCCAAG +TCCATA +TCCATT +TCCATC +TCCATG +TCCACA +TCCACT +TCCACC +TCCACG +TCCAGA +TCCAGT +TCCAGC +TCCAGG +TCCTAA +TCCTAT +TCCTAC +TCCTAG +TCCTTA +TCCTTT +TCCTTC +TCCTTG +TCCTCA +TCCTCT +TCCTCC +TCCTCG +TCCTGA +TCCTGT +TCCTGC +TCCTGG +TCCCAA +TCCCAT +TCCCAC +TCCCAG +TCCCTA +TCCCTT +TCCCTC +TCCCTG +TCCCCA +TCCCCT +TCCCCC +TCCCCG +TCCCGA +TCCCGT +TCCCGC +TCCCGG +TCCGAA +TCCGAT +TCCGAC +TCCGAG +TCCGTA +TCCGTT +TCCGTC +TCCGTG +TCCGCA +TCCGCT +TCCGCC +TCCGCG +TCCGGA +TCCGGT +TCCGGC +TCCGGG +TCGAAA +TCGAAT +TCGAAC +TCGAAG +TCGATA +TCGATT +TCGATC +TCGATG +TCGACA +TCGACT +TCGACC +TCGACG +TCGAGA +TCGAGT +TCGAGC +TCGAGG +TCGTAA +TCGTAT +TCGTAC +TCGTAG +TCGTTA +TCGTTT +TCGTTC +TCGTTG +TCGTCA +TCGTCT +TCGTCC +TCGTCG +TCGTGA +TCGTGT +TCGTGC +TCGTGG +TCGCAA +TCGCAT +TCGCAC +TCGCAG +TCGCTA +TCGCTT +TCGCTC +TCGCTG +TCGCCA +TCGCCT +TCGCCC +TCGCCG +TCGCGA +TCGCGT +TCGCGC +TCGCGG +TCGGAA +TCGGAT +TCGGAC +TCGGAG +TCGGTA +TCGGTT +TCGGTC +TCGGTG +TCGGCA +TCGGCT +TCGGCC +TCGGCG +TCGGGA +TCGGGT +TCGGGC +TCGGGG +TGAAAA +TGAAAT +TGAAAC +TGAAAG +TGAATA +TGAATT +TGAATC +TGAATG +TGAACA +TGAACT +TGAACC +TGAACG +TGAAGA +TGAAGT +TGAAGC +TGAAGG +TGATAA +TGATAT +TGATAC +TGATAG +TGATTA +TGATTT +TGATTC +TGATTG +TGATCA +TGATCT +TGATCC +TGATCG +TGATGA +TGATGT +TGATGC +TGATGG +TGACAA +TGACAT +TGACAC +TGACAG +TGACTA +TGACTT +TGACTC +TGACTG +TGACCA +TGACCT +TGACCC +TGACCG +TGACGA +TGACGT +TGACGC +TGACGG +TGAGAA +TGAGAT +TGAGAC +TGAGAG +TGAGTA +TGAGTT +TGAGTC +TGAGTG +TGAGCA +TGAGCT +TGAGCC +TGAGCG +TGAGGA +TGAGGT +TGAGGC +TGAGGG +TGTAAA +TGTAAT +TGTAAC +TGTAAG +TGTATA +TGTATT +TGTATC +TGTATG +TGTACA +TGTACT +TGTACC +TGTACG +TGTAGA +TGTAGT +TGTAGC +TGTAGG +TGTTAA +TGTTAT +TGTTAC +TGTTAG +TGTTTA +TGTTTT +TGTTTC +TGTTTG +TGTTCA +TGTTCT +TGTTCC +TGTTCG +TGTTGA +TGTTGT +TGTTGC +TGTTGG +TGTCAA +TGTCAT +TGTCAC +TGTCAG +TGTCTA +TGTCTT +TGTCTC +TGTCTG +TGTCCA +TGTCCT +TGTCCC +TGTCCG +TGTCGA +TGTCGT +TGTCGC +TGTCGG +TGTGAA +TGTGAT +TGTGAC +TGTGAG +TGTGTA +TGTGTT +TGTGTC +TGTGTG +TGTGCA +TGTGCT +TGTGCC +TGTGCG +TGTGGA +TGTGGT +TGTGGC +TGTGGG +TGCAAA +TGCAAT +TGCAAC +TGCAAG +TGCATA +TGCATT +TGCATC +TGCATG +TGCACA +TGCACT +TGCACC +TGCACG +TGCAGA +TGCAGT +TGCAGC +TGCAGG +TGCTAA +TGCTAT +TGCTAC +TGCTAG +TGCTTA +TGCTTT +TGCTTC +TGCTTG +TGCTCA +TGCTCT +TGCTCC +TGCTCG +TGCTGA +TGCTGT +TGCTGC +TGCTGG +TGCCAA +TGCCAT +TGCCAC +TGCCAG +TGCCTA +TGCCTT +TGCCTC +TGCCTG +TGCCCA +TGCCCT +TGCCCC +TGCCCG +TGCCGA +TGCCGT +TGCCGC +TGCCGG +TGCGAA +TGCGAT +TGCGAC +TGCGAG +TGCGTA +TGCGTT +TGCGTC +TGCGTG +TGCGCA +TGCGCT +TGCGCC +TGCGCG +TGCGGA +TGCGGT +TGCGGC +TGCGGG +TGGAAA +TGGAAT +TGGAAC +TGGAAG +TGGATA +TGGATT +TGGATC +TGGATG +TGGACA +TGGACT +TGGACC +TGGACG +TGGAGA +TGGAGT +TGGAGC +TGGAGG +TGGTAA +TGGTAT +TGGTAC +TGGTAG +TGGTTA +TGGTTT +TGGTTC +TGGTTG +TGGTCA +TGGTCT +TGGTCC +TGGTCG +TGGTGA +TGGTGT +TGGTGC +TGGTGG +TGGCAA +TGGCAT +TGGCAC +TGGCAG +TGGCTA +TGGCTT +TGGCTC +TGGCTG +TGGCCA +TGGCCT +TGGCCC +TGGCCG +TGGCGA +TGGCGT +TGGCGC +TGGCGG +TGGGAA +TGGGAT +TGGGAC +TGGGAG +TGGGTA +TGGGTT +TGGGTC +TGGGTG +TGGGCA +TGGGCT +TGGGCC +TGGGCG +TGGGGA +TGGGGT +TGGGGC +TGGGGG +CAAAAA +CAAAAT +CAAAAC +CAAAAG +CAAATA +CAAATT +CAAATC +CAAATG +CAAACA +CAAACT +CAAACC +CAAACG +CAAAGA +CAAAGT +CAAAGC +CAAAGG +CAATAA +CAATAT +CAATAC +CAATAG +CAATTA +CAATTT +CAATTC +CAATTG +CAATCA +CAATCT +CAATCC +CAATCG +CAATGA +CAATGT +CAATGC +CAATGG +CAACAA +CAACAT +CAACAC +CAACAG +CAACTA +CAACTT +CAACTC +CAACTG +CAACCA +CAACCT +CAACCC +CAACCG +CAACGA +CAACGT +CAACGC +CAACGG +CAAGAA +CAAGAT +CAAGAC +CAAGAG +CAAGTA +CAAGTT +CAAGTC +CAAGTG +CAAGCA +CAAGCT +CAAGCC +CAAGCG +CAAGGA +CAAGGT +CAAGGC +CAAGGG +CATAAA +CATAAT +CATAAC +CATAAG +CATATA +CATATT +CATATC +CATATG +CATACA +CATACT +CATACC +CATACG +CATAGA +CATAGT +CATAGC +CATAGG +CATTAA +CATTAT +CATTAC +CATTAG +CATTTA +CATTTT +CATTTC +CATTTG +CATTCA +CATTCT +CATTCC +CATTCG +CATTGA +CATTGT +CATTGC +CATTGG +CATCAA +CATCAT +CATCAC +CATCAG +CATCTA +CATCTT +CATCTC +CATCTG +CATCCA +CATCCT +CATCCC +CATCCG +CATCGA +CATCGT +CATCGC +CATCGG +CATGAA +CATGAT +CATGAC +CATGAG +CATGTA +CATGTT +CATGTC +CATGTG +CATGCA +CATGCT +CATGCC +CATGCG +CATGGA +CATGGT +CATGGC +CATGGG +CACAAA +CACAAT +CACAAC +CACAAG +CACATA +CACATT +CACATC +CACATG +CACACA +CACACT +CACACC +CACACG +CACAGA +CACAGT +CACAGC +CACAGG +CACTAA +CACTAT +CACTAC +CACTAG +CACTTA +CACTTT +CACTTC +CACTTG +CACTCA +CACTCT +CACTCC +CACTCG +CACTGA +CACTGT +CACTGC +CACTGG +CACCAA +CACCAT +CACCAC +CACCAG +CACCTA +CACCTT +CACCTC +CACCTG +CACCCA +CACCCT +CACCCC +CACCCG +CACCGA +CACCGT +CACCGC +CACCGG +CACGAA +CACGAT +CACGAC +CACGAG +CACGTA +CACGTT +CACGTC +CACGTG +CACGCA +CACGCT +CACGCC +CACGCG +CACGGA +CACGGT +CACGGC +CACGGG +CAGAAA +CAGAAT +CAGAAC +CAGAAG +CAGATA +CAGATT +CAGATC +CAGATG +CAGACA +CAGACT +CAGACC +CAGACG +CAGAGA +CAGAGT +CAGAGC +CAGAGG +CAGTAA +CAGTAT +CAGTAC +CAGTAG +CAGTTA +CAGTTT +CAGTTC +CAGTTG +CAGTCA +CAGTCT +CAGTCC +CAGTCG +CAGTGA +CAGTGT +CAGTGC +CAGTGG +CAGCAA +CAGCAT +CAGCAC +CAGCAG +CAGCTA +CAGCTT +CAGCTC +CAGCTG +CAGCCA +CAGCCT +CAGCCC +CAGCCG +CAGCGA +CAGCGT +CAGCGC +CAGCGG +CAGGAA +CAGGAT +CAGGAC +CAGGAG +CAGGTA +CAGGTT +CAGGTC +CAGGTG +CAGGCA +CAGGCT +CAGGCC +CAGGCG +CAGGGA +CAGGGT +CAGGGC +CAGGGG +CTAAAA +CTAAAT +CTAAAC +CTAAAG +CTAATA +CTAATT +CTAATC +CTAATG +CTAACA +CTAACT +CTAACC +CTAACG +CTAAGA +CTAAGT +CTAAGC +CTAAGG +CTATAA +CTATAT +CTATAC +CTATAG +CTATTA +CTATTT +CTATTC +CTATTG +CTATCA +CTATCT +CTATCC +CTATCG +CTATGA +CTATGT +CTATGC +CTATGG +CTACAA +CTACAT +CTACAC +CTACAG +CTACTA +CTACTT +CTACTC +CTACTG +CTACCA +CTACCT +CTACCC +CTACCG +CTACGA +CTACGT +CTACGC +CTACGG +CTAGAA +CTAGAT +CTAGAC +CTAGAG +CTAGTA +CTAGTT +CTAGTC +CTAGTG +CTAGCA +CTAGCT +CTAGCC +CTAGCG +CTAGGA +CTAGGT +CTAGGC +CTAGGG +CTTAAA +CTTAAT +CTTAAC +CTTAAG +CTTATA +CTTATT +CTTATC +CTTATG +CTTACA +CTTACT +CTTACC +CTTACG +CTTAGA +CTTAGT +CTTAGC +CTTAGG +CTTTAA +CTTTAT +CTTTAC +CTTTAG +CTTTTA +CTTTTT +CTTTTC +CTTTTG +CTTTCA +CTTTCT +CTTTCC +CTTTCG +CTTTGA +CTTTGT +CTTTGC +CTTTGG +CTTCAA +CTTCAT +CTTCAC +CTTCAG +CTTCTA +CTTCTT +CTTCTC +CTTCTG +CTTCCA +CTTCCT +CTTCCC +CTTCCG +CTTCGA +CTTCGT +CTTCGC +CTTCGG +CTTGAA +CTTGAT +CTTGAC +CTTGAG +CTTGTA +CTTGTT +CTTGTC +CTTGTG +CTTGCA +CTTGCT +CTTGCC +CTTGCG +CTTGGA +CTTGGT +CTTGGC +CTTGGG +CTCAAA +CTCAAT +CTCAAC +CTCAAG +CTCATA +CTCATT +CTCATC +CTCATG +CTCACA +CTCACT +CTCACC +CTCACG +CTCAGA +CTCAGT +CTCAGC +CTCAGG +CTCTAA +CTCTAT +CTCTAC +CTCTAG +CTCTTA +CTCTTT +CTCTTC +CTCTTG +CTCTCA +CTCTCT +CTCTCC +CTCTCG +CTCTGA +CTCTGT +CTCTGC +CTCTGG +CTCCAA +CTCCAT +CTCCAC +CTCCAG +CTCCTA +CTCCTT +CTCCTC +CTCCTG +CTCCCA +CTCCCT +CTCCCC +CTCCCG +CTCCGA +CTCCGT +CTCCGC +CTCCGG +CTCGAA +CTCGAT +CTCGAC +CTCGAG +CTCGTA +CTCGTT +CTCGTC +CTCGTG +CTCGCA +CTCGCT +CTCGCC +CTCGCG +CTCGGA +CTCGGT +CTCGGC +CTCGGG +CTGAAA +CTGAAT +CTGAAC +CTGAAG +CTGATA +CTGATT +CTGATC +CTGATG +CTGACA +CTGACT +CTGACC +CTGACG +CTGAGA +CTGAGT +CTGAGC +CTGAGG +CTGTAA +CTGTAT +CTGTAC +CTGTAG +CTGTTA +CTGTTT +CTGTTC +CTGTTG +CTGTCA +CTGTCT +CTGTCC +CTGTCG +CTGTGA +CTGTGT +CTGTGC +CTGTGG +CTGCAA +CTGCAT +CTGCAC +CTGCAG +CTGCTA +CTGCTT +CTGCTC +CTGCTG +CTGCCA +CTGCCT +CTGCCC +CTGCCG +CTGCGA +CTGCGT +CTGCGC +CTGCGG +CTGGAA +CTGGAT +CTGGAC +CTGGAG +CTGGTA +CTGGTT +CTGGTC +CTGGTG +CTGGCA +CTGGCT +CTGGCC +CTGGCG +CTGGGA +CTGGGT +CTGGGC +CTGGGG +CCAAAA +CCAAAT +CCAAAC +CCAAAG +CCAATA +CCAATT +CCAATC +CCAATG +CCAACA +CCAACT +CCAACC +CCAACG +CCAAGA +CCAAGT +CCAAGC +CCAAGG +CCATAA +CCATAT +CCATAC +CCATAG +CCATTA +CCATTT +CCATTC +CCATTG +CCATCA +CCATCT +CCATCC +CCATCG +CCATGA +CCATGT +CCATGC +CCATGG +CCACAA +CCACAT +CCACAC +CCACAG +CCACTA +CCACTT +CCACTC +CCACTG +CCACCA +CCACCT +CCACCC +CCACCG +CCACGA +CCACGT +CCACGC +CCACGG +CCAGAA +CCAGAT +CCAGAC +CCAGAG +CCAGTA +CCAGTT +CCAGTC +CCAGTG +CCAGCA +CCAGCT +CCAGCC +CCAGCG +CCAGGA +CCAGGT +CCAGGC +CCAGGG +CCTAAA +CCTAAT +CCTAAC +CCTAAG +CCTATA +CCTATT +CCTATC +CCTATG +CCTACA +CCTACT +CCTACC +CCTACG +CCTAGA +CCTAGT +CCTAGC +CCTAGG +CCTTAA +CCTTAT +CCTTAC +CCTTAG +CCTTTA +CCTTTT +CCTTTC +CCTTTG +CCTTCA +CCTTCT +CCTTCC +CCTTCG +CCTTGA +CCTTGT +CCTTGC +CCTTGG +CCTCAA +CCTCAT +CCTCAC +CCTCAG +CCTCTA +CCTCTT +CCTCTC +CCTCTG +CCTCCA +CCTCCT +CCTCCC +CCTCCG +CCTCGA +CCTCGT +CCTCGC +CCTCGG +CCTGAA +CCTGAT +CCTGAC +CCTGAG +CCTGTA +CCTGTT +CCTGTC +CCTGTG +CCTGCA +CCTGCT +CCTGCC +CCTGCG +CCTGGA +CCTGGT +CCTGGC +CCTGGG +CCCAAA +CCCAAT +CCCAAC +CCCAAG +CCCATA +CCCATT +CCCATC +CCCATG +CCCACA +CCCACT +CCCACC +CCCACG +CCCAGA +CCCAGT +CCCAGC +CCCAGG +CCCTAA +CCCTAT +CCCTAC +CCCTAG +CCCTTA +CCCTTT +CCCTTC +CCCTTG +CCCTCA +CCCTCT +CCCTCC +CCCTCG +CCCTGA +CCCTGT +CCCTGC +CCCTGG +CCCCAA +CCCCAT +CCCCAC +CCCCAG +CCCCTA +CCCCTT +CCCCTC +CCCCTG +CCCCCA +CCCCCT +CCCCCC +CCCCCG +CCCCGA +CCCCGT +CCCCGC +CCCCGG +CCCGAA +CCCGAT +CCCGAC +CCCGAG +CCCGTA +CCCGTT +CCCGTC +CCCGTG +CCCGCA +CCCGCT +CCCGCC +CCCGCG +CCCGGA +CCCGGT +CCCGGC +CCCGGG +CCGAAA +CCGAAT +CCGAAC +CCGAAG +CCGATA +CCGATT +CCGATC +CCGATG +CCGACA +CCGACT +CCGACC +CCGACG +CCGAGA +CCGAGT +CCGAGC +CCGAGG +CCGTAA +CCGTAT +CCGTAC +CCGTAG +CCGTTA +CCGTTT +CCGTTC +CCGTTG +CCGTCA +CCGTCT +CCGTCC +CCGTCG +CCGTGA +CCGTGT +CCGTGC +CCGTGG +CCGCAA +CCGCAT +CCGCAC +CCGCAG +CCGCTA +CCGCTT +CCGCTC +CCGCTG +CCGCCA +CCGCCT +CCGCCC +CCGCCG +CCGCGA +CCGCGT +CCGCGC +CCGCGG +CCGGAA +CCGGAT +CCGGAC +CCGGAG +CCGGTA +CCGGTT +CCGGTC +CCGGTG +CCGGCA +CCGGCT +CCGGCC +CCGGCG +CCGGGA +CCGGGT +CCGGGC +CCGGGG +CGAAAA +CGAAAT +CGAAAC +CGAAAG +CGAATA +CGAATT +CGAATC +CGAATG +CGAACA +CGAACT +CGAACC +CGAACG +CGAAGA +CGAAGT +CGAAGC +CGAAGG +CGATAA +CGATAT +CGATAC +CGATAG +CGATTA +CGATTT +CGATTC +CGATTG +CGATCA +CGATCT +CGATCC +CGATCG +CGATGA +CGATGT +CGATGC +CGATGG +CGACAA +CGACAT +CGACAC +CGACAG +CGACTA +CGACTT +CGACTC +CGACTG +CGACCA +CGACCT +CGACCC +CGACCG +CGACGA +CGACGT +CGACGC +CGACGG +CGAGAA +CGAGAT +CGAGAC +CGAGAG +CGAGTA +CGAGTT +CGAGTC +CGAGTG +CGAGCA +CGAGCT +CGAGCC +CGAGCG +CGAGGA +CGAGGT +CGAGGC +CGAGGG +CGTAAA +CGTAAT +CGTAAC +CGTAAG +CGTATA +CGTATT +CGTATC +CGTATG +CGTACA +CGTACT +CGTACC +CGTACG +CGTAGA +CGTAGT +CGTAGC +CGTAGG +CGTTAA +CGTTAT +CGTTAC +CGTTAG +CGTTTA +CGTTTT +CGTTTC +CGTTTG +CGTTCA +CGTTCT +CGTTCC +CGTTCG +CGTTGA +CGTTGT +CGTTGC +CGTTGG +CGTCAA +CGTCAT +CGTCAC +CGTCAG +CGTCTA +CGTCTT +CGTCTC +CGTCTG +CGTCCA +CGTCCT +CGTCCC +CGTCCG +CGTCGA +CGTCGT +CGTCGC +CGTCGG +CGTGAA +CGTGAT +CGTGAC +CGTGAG +CGTGTA +CGTGTT +CGTGTC +CGTGTG +CGTGCA +CGTGCT +CGTGCC +CGTGCG +CGTGGA +CGTGGT +CGTGGC +CGTGGG +CGCAAA +CGCAAT +CGCAAC +CGCAAG +CGCATA +CGCATT +CGCATC +CGCATG +CGCACA +CGCACT +CGCACC +CGCACG +CGCAGA +CGCAGT +CGCAGC +CGCAGG +CGCTAA +CGCTAT +CGCTAC +CGCTAG +CGCTTA +CGCTTT +CGCTTC +CGCTTG +CGCTCA +CGCTCT +CGCTCC +CGCTCG +CGCTGA +CGCTGT +CGCTGC +CGCTGG +CGCCAA +CGCCAT +CGCCAC +CGCCAG +CGCCTA +CGCCTT +CGCCTC +CGCCTG +CGCCCA +CGCCCT +CGCCCC +CGCCCG +CGCCGA +CGCCGT +CGCCGC +CGCCGG +CGCGAA +CGCGAT +CGCGAC +CGCGAG +CGCGTA +CGCGTT +CGCGTC +CGCGTG +CGCGCA +CGCGCT +CGCGCC +CGCGCG +CGCGGA +CGCGGT +CGCGGC +CGCGGG +CGGAAA +CGGAAT +CGGAAC +CGGAAG +CGGATA +CGGATT +CGGATC +CGGATG +CGGACA +CGGACT +CGGACC +CGGACG +CGGAGA +CGGAGT +CGGAGC +CGGAGG +CGGTAA +CGGTAT +CGGTAC +CGGTAG +CGGTTA +CGGTTT +CGGTTC +CGGTTG +CGGTCA +CGGTCT +CGGTCC +CGGTCG +CGGTGA +CGGTGT +CGGTGC +CGGTGG +CGGCAA +CGGCAT +CGGCAC +CGGCAG +CGGCTA +CGGCTT +CGGCTC +CGGCTG +CGGCCA +CGGCCT +CGGCCC +CGGCCG +CGGCGA +CGGCGT +CGGCGC +CGGCGG +CGGGAA +CGGGAT +CGGGAC +CGGGAG +CGGGTA +CGGGTT +CGGGTC +CGGGTG +CGGGCA +CGGGCT +CGGGCC +CGGGCG +CGGGGA +CGGGGT +CGGGGC +CGGGGG +GAAAAA +GAAAAT +GAAAAC +GAAAAG +GAAATA +GAAATT +GAAATC +GAAATG +GAAACA +GAAACT +GAAACC +GAAACG +GAAAGA +GAAAGT +GAAAGC +GAAAGG +GAATAA +GAATAT +GAATAC +GAATAG +GAATTA +GAATTT +GAATTC +GAATTG +GAATCA +GAATCT +GAATCC +GAATCG +GAATGA +GAATGT +GAATGC +GAATGG +GAACAA +GAACAT +GAACAC +GAACAG +GAACTA +GAACTT +GAACTC +GAACTG +GAACCA +GAACCT +GAACCC +GAACCG +GAACGA +GAACGT +GAACGC +GAACGG +GAAGAA +GAAGAT +GAAGAC +GAAGAG +GAAGTA +GAAGTT +GAAGTC +GAAGTG +GAAGCA +GAAGCT +GAAGCC +GAAGCG +GAAGGA +GAAGGT +GAAGGC +GAAGGG +GATAAA +GATAAT +GATAAC +GATAAG +GATATA +GATATT +GATATC +GATATG +GATACA +GATACT +GATACC +GATACG +GATAGA +GATAGT +GATAGC +GATAGG +GATTAA +GATTAT +GATTAC +GATTAG +GATTTA +GATTTT +GATTTC +GATTTG +GATTCA +GATTCT +GATTCC +GATTCG +GATTGA +GATTGT +GATTGC +GATTGG +GATCAA +GATCAT +GATCAC +GATCAG +GATCTA +GATCTT +GATCTC +GATCTG +GATCCA +GATCCT +GATCCC +GATCCG +GATCGA +GATCGT +GATCGC +GATCGG +GATGAA +GATGAT +GATGAC +GATGAG +GATGTA +GATGTT +GATGTC +GATGTG +GATGCA +GATGCT +GATGCC +GATGCG +GATGGA +GATGGT +GATGGC +GATGGG +GACAAA +GACAAT +GACAAC +GACAAG +GACATA +GACATT +GACATC +GACATG +GACACA +GACACT +GACACC +GACACG +GACAGA +GACAGT +GACAGC +GACAGG +GACTAA +GACTAT +GACTAC +GACTAG +GACTTA +GACTTT +GACTTC +GACTTG +GACTCA +GACTCT +GACTCC +GACTCG +GACTGA +GACTGT +GACTGC +GACTGG +GACCAA +GACCAT +GACCAC +GACCAG +GACCTA +GACCTT +GACCTC +GACCTG +GACCCA +GACCCT +GACCCC +GACCCG +GACCGA +GACCGT +GACCGC +GACCGG +GACGAA +GACGAT +GACGAC +GACGAG +GACGTA +GACGTT +GACGTC +GACGTG +GACGCA +GACGCT +GACGCC +GACGCG +GACGGA +GACGGT +GACGGC +GACGGG +GAGAAA +GAGAAT +GAGAAC +GAGAAG +GAGATA +GAGATT +GAGATC +GAGATG +GAGACA +GAGACT +GAGACC +GAGACG +GAGAGA +GAGAGT +GAGAGC +GAGAGG +GAGTAA +GAGTAT +GAGTAC +GAGTAG +GAGTTA +GAGTTT +GAGTTC +GAGTTG +GAGTCA +GAGTCT +GAGTCC +GAGTCG +GAGTGA +GAGTGT +GAGTGC +GAGTGG +GAGCAA +GAGCAT +GAGCAC +GAGCAG +GAGCTA +GAGCTT +GAGCTC +GAGCTG +GAGCCA +GAGCCT +GAGCCC +GAGCCG +GAGCGA +GAGCGT +GAGCGC +GAGCGG +GAGGAA +GAGGAT +GAGGAC +GAGGAG +GAGGTA +GAGGTT +GAGGTC +GAGGTG +GAGGCA +GAGGCT +GAGGCC +GAGGCG +GAGGGA +GAGGGT +GAGGGC +GAGGGG +GTAAAA +GTAAAT +GTAAAC +GTAAAG +GTAATA +GTAATT +GTAATC +GTAATG +GTAACA +GTAACT +GTAACC +GTAACG +GTAAGA +GTAAGT +GTAAGC +GTAAGG +GTATAA +GTATAT +GTATAC +GTATAG +GTATTA +GTATTT +GTATTC +GTATTG +GTATCA +GTATCT +GTATCC +GTATCG +GTATGA +GTATGT +GTATGC +GTATGG +GTACAA +GTACAT +GTACAC +GTACAG +GTACTA +GTACTT +GTACTC +GTACTG +GTACCA +GTACCT +GTACCC +GTACCG +GTACGA +GTACGT +GTACGC +GTACGG +GTAGAA +GTAGAT +GTAGAC +GTAGAG +GTAGTA +GTAGTT +GTAGTC +GTAGTG +GTAGCA +GTAGCT +GTAGCC +GTAGCG +GTAGGA +GTAGGT +GTAGGC +GTAGGG +GTTAAA +GTTAAT +GTTAAC +GTTAAG +GTTATA +GTTATT +GTTATC +GTTATG +GTTACA +GTTACT +GTTACC +GTTACG +GTTAGA +GTTAGT +GTTAGC +GTTAGG +GTTTAA +GTTTAT +GTTTAC +GTTTAG +GTTTTA +GTTTTT +GTTTTC +GTTTTG +GTTTCA +GTTTCT +GTTTCC +GTTTCG +GTTTGA +GTTTGT +GTTTGC +GTTTGG +GTTCAA +GTTCAT +GTTCAC +GTTCAG +GTTCTA +GTTCTT +GTTCTC +GTTCTG +GTTCCA +GTTCCT +GTTCCC +GTTCCG +GTTCGA +GTTCGT +GTTCGC +GTTCGG +GTTGAA +GTTGAT +GTTGAC +GTTGAG +GTTGTA +GTTGTT +GTTGTC +GTTGTG +GTTGCA +GTTGCT +GTTGCC +GTTGCG +GTTGGA +GTTGGT +GTTGGC +GTTGGG +GTCAAA +GTCAAT +GTCAAC +GTCAAG +GTCATA +GTCATT +GTCATC +GTCATG +GTCACA +GTCACT +GTCACC +GTCACG +GTCAGA +GTCAGT +GTCAGC +GTCAGG +GTCTAA +GTCTAT +GTCTAC +GTCTAG +GTCTTA +GTCTTT +GTCTTC +GTCTTG +GTCTCA +GTCTCT +GTCTCC +GTCTCG +GTCTGA +GTCTGT +GTCTGC +GTCTGG +GTCCAA +GTCCAT +GTCCAC +GTCCAG +GTCCTA +GTCCTT +GTCCTC +GTCCTG +GTCCCA +GTCCCT +GTCCCC +GTCCCG +GTCCGA +GTCCGT +GTCCGC +GTCCGG +GTCGAA +GTCGAT +GTCGAC +GTCGAG +GTCGTA +GTCGTT +GTCGTC +GTCGTG +GTCGCA +GTCGCT +GTCGCC +GTCGCG +GTCGGA +GTCGGT +GTCGGC +GTCGGG +GTGAAA +GTGAAT +GTGAAC +GTGAAG +GTGATA +GTGATT +GTGATC +GTGATG +GTGACA +GTGACT +GTGACC +GTGACG +GTGAGA +GTGAGT +GTGAGC +GTGAGG +GTGTAA +GTGTAT +GTGTAC +GTGTAG +GTGTTA +GTGTTT +GTGTTC +GTGTTG +GTGTCA +GTGTCT +GTGTCC +GTGTCG +GTGTGA +GTGTGT +GTGTGC +GTGTGG +GTGCAA +GTGCAT +GTGCAC +GTGCAG +GTGCTA +GTGCTT +GTGCTC +GTGCTG +GTGCCA +GTGCCT +GTGCCC +GTGCCG +GTGCGA +GTGCGT +GTGCGC +GTGCGG +GTGGAA +GTGGAT +GTGGAC +GTGGAG +GTGGTA +GTGGTT +GTGGTC +GTGGTG +GTGGCA +GTGGCT +GTGGCC +GTGGCG +GTGGGA +GTGGGT +GTGGGC +GTGGGG +GCAAAA +GCAAAT +GCAAAC +GCAAAG +GCAATA +GCAATT +GCAATC +GCAATG +GCAACA +GCAACT +GCAACC +GCAACG +GCAAGA +GCAAGT +GCAAGC +GCAAGG +GCATAA +GCATAT +GCATAC +GCATAG +GCATTA +GCATTT +GCATTC +GCATTG +GCATCA +GCATCT +GCATCC +GCATCG +GCATGA +GCATGT +GCATGC +GCATGG +GCACAA +GCACAT +GCACAC +GCACAG +GCACTA +GCACTT +GCACTC +GCACTG +GCACCA +GCACCT +GCACCC +GCACCG +GCACGA +GCACGT +GCACGC +GCACGG +GCAGAA +GCAGAT +GCAGAC +GCAGAG +GCAGTA +GCAGTT +GCAGTC +GCAGTG +GCAGCA +GCAGCT +GCAGCC +GCAGCG +GCAGGA +GCAGGT +GCAGGC +GCAGGG +GCTAAA +GCTAAT +GCTAAC +GCTAAG +GCTATA +GCTATT +GCTATC +GCTATG +GCTACA +GCTACT +GCTACC +GCTACG +GCTAGA +GCTAGT +GCTAGC +GCTAGG +GCTTAA +GCTTAT +GCTTAC +GCTTAG +GCTTTA +GCTTTT +GCTTTC +GCTTTG +GCTTCA +GCTTCT +GCTTCC +GCTTCG +GCTTGA +GCTTGT +GCTTGC +GCTTGG +GCTCAA +GCTCAT +GCTCAC +GCTCAG +GCTCTA +GCTCTT +GCTCTC +GCTCTG +GCTCCA +GCTCCT +GCTCCC +GCTCCG +GCTCGA +GCTCGT +GCTCGC +GCTCGG +GCTGAA +GCTGAT +GCTGAC +GCTGAG +GCTGTA +GCTGTT +GCTGTC +GCTGTG +GCTGCA +GCTGCT +GCTGCC +GCTGCG +GCTGGA +GCTGGT +GCTGGC +GCTGGG +GCCAAA +GCCAAT +GCCAAC +GCCAAG +GCCATA +GCCATT +GCCATC +GCCATG +GCCACA +GCCACT +GCCACC +GCCACG +GCCAGA +GCCAGT +GCCAGC +GCCAGG +GCCTAA +GCCTAT +GCCTAC +GCCTAG +GCCTTA +GCCTTT +GCCTTC +GCCTTG +GCCTCA +GCCTCT +GCCTCC +GCCTCG +GCCTGA +GCCTGT +GCCTGC +GCCTGG +GCCCAA +GCCCAT +GCCCAC +GCCCAG +GCCCTA +GCCCTT +GCCCTC +GCCCTG +GCCCCA +GCCCCT +GCCCCC +GCCCCG +GCCCGA +GCCCGT +GCCCGC +GCCCGG +GCCGAA +GCCGAT +GCCGAC +GCCGAG +GCCGTA +GCCGTT +GCCGTC +GCCGTG +GCCGCA +GCCGCT +GCCGCC +GCCGCG +GCCGGA +GCCGGT +GCCGGC +GCCGGG +GCGAAA +GCGAAT +GCGAAC +GCGAAG +GCGATA +GCGATT +GCGATC +GCGATG +GCGACA +GCGACT +GCGACC +GCGACG +GCGAGA +GCGAGT +GCGAGC +GCGAGG +GCGTAA +GCGTAT +GCGTAC +GCGTAG +GCGTTA +GCGTTT +GCGTTC +GCGTTG +GCGTCA +GCGTCT +GCGTCC +GCGTCG +GCGTGA +GCGTGT +GCGTGC +GCGTGG +GCGCAA +GCGCAT +GCGCAC +GCGCAG +GCGCTA +GCGCTT +GCGCTC +GCGCTG +GCGCCA +GCGCCT +GCGCCC +GCGCCG +GCGCGA +GCGCGT +GCGCGC +GCGCGG +GCGGAA +GCGGAT +GCGGAC +GCGGAG +GCGGTA +GCGGTT +GCGGTC +GCGGTG +GCGGCA +GCGGCT +GCGGCC +GCGGCG +GCGGGA +GCGGGT +GCGGGC +GCGGGG +GGAAAA +GGAAAT +GGAAAC +GGAAAG +GGAATA +GGAATT +GGAATC +GGAATG +GGAACA +GGAACT +GGAACC +GGAACG +GGAAGA +GGAAGT +GGAAGC +GGAAGG +GGATAA +GGATAT +GGATAC +GGATAG +GGATTA +GGATTT +GGATTC +GGATTG +GGATCA +GGATCT +GGATCC +GGATCG +GGATGA +GGATGT +GGATGC +GGATGG +GGACAA +GGACAT +GGACAC +GGACAG +GGACTA +GGACTT +GGACTC +GGACTG +GGACCA +GGACCT +GGACCC +GGACCG +GGACGA +GGACGT +GGACGC +GGACGG +GGAGAA +GGAGAT +GGAGAC +GGAGAG +GGAGTA +GGAGTT +GGAGTC +GGAGTG +GGAGCA +GGAGCT +GGAGCC +GGAGCG +GGAGGA +GGAGGT +GGAGGC +GGAGGG +GGTAAA +GGTAAT +GGTAAC +GGTAAG +GGTATA +GGTATT +GGTATC +GGTATG +GGTACA +GGTACT +GGTACC +GGTACG +GGTAGA +GGTAGT +GGTAGC +GGTAGG +GGTTAA +GGTTAT +GGTTAC +GGTTAG +GGTTTA +GGTTTT +GGTTTC +GGTTTG +GGTTCA +GGTTCT +GGTTCC +GGTTCG +GGTTGA +GGTTGT +GGTTGC +GGTTGG +GGTCAA +GGTCAT +GGTCAC +GGTCAG +GGTCTA +GGTCTT +GGTCTC +GGTCTG +GGTCCA +GGTCCT +GGTCCC +GGTCCG +GGTCGA +GGTCGT +GGTCGC +GGTCGG +GGTGAA +GGTGAT +GGTGAC +GGTGAG +GGTGTA +GGTGTT +GGTGTC +GGTGTG +GGTGCA +GGTGCT +GGTGCC +GGTGCG +GGTGGA +GGTGGT +GGTGGC +GGTGGG +GGCAAA +GGCAAT +GGCAAC +GGCAAG +GGCATA +GGCATT +GGCATC +GGCATG +GGCACA +GGCACT +GGCACC +GGCACG +GGCAGA +GGCAGT +GGCAGC +GGCAGG +GGCTAA +GGCTAT +GGCTAC +GGCTAG +GGCTTA +GGCTTT +GGCTTC +GGCTTG +GGCTCA +GGCTCT +GGCTCC +GGCTCG +GGCTGA +GGCTGT +GGCTGC +GGCTGG +GGCCAA +GGCCAT +GGCCAC +GGCCAG +GGCCTA +GGCCTT +GGCCTC +GGCCTG +GGCCCA +GGCCCT +GGCCCC +GGCCCG +GGCCGA +GGCCGT +GGCCGC +GGCCGG +GGCGAA +GGCGAT +GGCGAC +GGCGAG +GGCGTA +GGCGTT +GGCGTC +GGCGTG +GGCGCA +GGCGCT +GGCGCC +GGCGCG +GGCGGA +GGCGGT +GGCGGC +GGCGGG +GGGAAA +GGGAAT +GGGAAC +GGGAAG +GGGATA +GGGATT +GGGATC +GGGATG +GGGACA +GGGACT +GGGACC +GGGACG +GGGAGA +GGGAGT +GGGAGC +GGGAGG +GGGTAA +GGGTAT +GGGTAC +GGGTAG +GGGTTA +GGGTTT +GGGTTC +GGGTTG +GGGTCA +GGGTCT +GGGTCC +GGGTCG +GGGTGA +GGGTGT +GGGTGC +GGGTGG +GGGCAA +GGGCAT +GGGCAC +GGGCAG +GGGCTA +GGGCTT +GGGCTC +GGGCTG +GGGCCA +GGGCCT +GGGCCC +GGGCCG +GGGCGA +GGGCGT +GGGCGC +GGGCGG +GGGGAA +GGGGAT +GGGGAC +GGGGAG +GGGGTA +GGGGTT +GGGGTC +GGGGTG +GGGGCA +GGGGCT +GGGGCC +GGGGCG +GGGGGA +GGGGGT +GGGGGC +GGGGGG diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/README.md b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..500fd976d92b217455f181bc98c77e612d9a18d9 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/README.md @@ -0,0 +1,202 @@ +--- +base_model: fabihamakhdoomi/TinyDNABERT +library_name: peft +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.12.0 \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/adapter_config.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..28c871c93775eedfaf266f00da8c02f96bc02f29 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/adapter_config.json @@ -0,0 +1,42 @@ +{ + "alpha_pattern": {}, + "auto_mapping": null, + "base_model_name_or_path": "fabihamakhdoomi/TinyDNABERT", + "bias": "none", + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 16, + "lora_dropout": 0.2, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": [ + "classifier", + "score" + ], + "peft_type": "LORA", + "r": 4, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "bert.encoder.layer.1.attention.self.query", + "bert.encoder.layer.0.output.dense", + "bert.encoder.layer.1.output.dense", + "bert.encoder.layer.0.intermediate.dense", + "bert.encoder.layer.1.intermediate.dense", + "bert.encoder.layer.0.attention.output.dense", + "bert.encoder.layer.1.attention.output.dense", + "bert.encoder.layer.0.attention.self.query", + "bert.encoder.layer.0.attention.self.value", + "bert.encoder.layer.1.attention.self.key", + "bert.encoder.layer.1.attention.self.value", + "bert.encoder.layer.0.attention.self.key" + ], + "task_type": "SEQ_CLS", + "use_dora": false, + "use_rslora": false +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/adapter_model.safetensors b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/adapter_model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..f3b81ded5e9f4e5d38c5d6a5f85f346589965425 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/adapter_model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:283e92c174df9a767ba574c268e1b44ebf2228ebfb311742573eea98b43a2304 +size 78208 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/optimizer.pt b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..443635bc6ec8427e69dcc3f15faed2eca494b919 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b5ff20c372bbf1021bb193c2d5a06dbe5f9092fa5bb345af59f76df1dc63999c +size 165226 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/rng_state.pth b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..d769758d72ed276350b7475a3c6ab31d03d6d464 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:af365fb722ff3640b9c3cbda0c33c89309e74956fafe530249e5f878f489cae3 +size 14244 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/scheduler.pt b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..e21fdaa14c288938c4b2c1277194110b759d8724 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7480edaf7ecb5003556fedbf146cf26ec75a9ee19931ce031bbb3b509f9927ce +size 1064 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/special_tokens_map.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..a8b3208c2884c4efb86e49300fdd3dc877220cdf --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/special_tokens_map.json @@ -0,0 +1,7 @@ +{ + "cls_token": "[CLS]", + "mask_token": "[MASK]", + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/tokenizer.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..ebd782de6963b9032ce9374cfe30f5275be1a30b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/tokenizer.json @@ -0,0 +1,4265 @@ +{ + "version": "1.0", + "truncation": { + "direction": "Right", + "max_length": 512, + "strategy": "LongestFirst", + "stride": 0 + }, + "padding": { + "strategy": { + "Fixed": 512 + }, + "direction": "Right", + "pad_to_multiple_of": null, + "pad_id": 0, + "pad_type_id": 0, + "pad_token": "[PAD]" + }, + "added_tokens": [ + { + "id": 0, + "content": "[PAD]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 1, + "content": "[UNK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 2, + "content": "[CLS]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 3, + "content": "[SEP]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 4, + "content": "[MASK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + } + ], + "normalizer": { + "type": "BertNormalizer", + "clean_text": true, + "handle_chinese_chars": true, + "strip_accents": null, + "lowercase": false + }, + "pre_tokenizer": { + "type": "BertPreTokenizer" + }, + "post_processor": { + "type": "TemplateProcessing", + "single": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + } + ], + "pair": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "B", + "type_id": 1 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 1 + } + } + ], + "special_tokens": { + "[CLS]": { + "id": "[CLS]", + "ids": [ + 2 + ], + "tokens": [ + "[CLS]" + ] + }, + "[SEP]": { + "id": "[SEP]", + "ids": [ + 3 + ], + "tokens": [ + "[SEP]" + ] + } + } + }, + "decoder": { + "type": "WordPiece", + "prefix": "##", + "cleanup": true + }, + "model": { + "type": "WordPiece", + "unk_token": "[UNK]", + "continuing_subword_prefix": "##", + "max_input_chars_per_word": 100, + "vocab": { + "[PAD]": 0, + "[UNK]": 1, + "[CLS]": 2, + "[SEP]": 3, + "[MASK]": 4, + "AAAAAA": 5, + "AAAAAT": 6, + "AAAAAC": 7, + "AAAAAG": 8, + "AAAATA": 9, + "AAAATT": 10, + "AAAATC": 11, + "AAAATG": 12, + "AAAACA": 13, + "AAAACT": 14, + "AAAACC": 15, + "AAAACG": 16, + "AAAAGA": 17, + "AAAAGT": 18, + "AAAAGC": 19, + "AAAAGG": 20, + "AAATAA": 21, + "AAATAT": 22, + "AAATAC": 23, + "AAATAG": 24, + "AAATTA": 25, + "AAATTT": 26, + "AAATTC": 27, + "AAATTG": 28, + "AAATCA": 29, + "AAATCT": 30, + "AAATCC": 31, + "AAATCG": 32, + "AAATGA": 33, + "AAATGT": 34, + "AAATGC": 35, + "AAATGG": 36, + "AAACAA": 37, + "AAACAT": 38, + "AAACAC": 39, + "AAACAG": 40, + "AAACTA": 41, + "AAACTT": 42, + "AAACTC": 43, + "AAACTG": 44, + "AAACCA": 45, + "AAACCT": 46, + "AAACCC": 47, + "AAACCG": 48, + "AAACGA": 49, + "AAACGT": 50, + "AAACGC": 51, + "AAACGG": 52, + "AAAGAA": 53, + "AAAGAT": 54, + "AAAGAC": 55, + "AAAGAG": 56, + "AAAGTA": 57, + "AAAGTT": 58, + "AAAGTC": 59, + "AAAGTG": 60, + "AAAGCA": 61, + "AAAGCT": 62, + "AAAGCC": 63, + "AAAGCG": 64, + "AAAGGA": 65, + "AAAGGT": 66, + "AAAGGC": 67, + "AAAGGG": 68, + "AATAAA": 69, + "AATAAT": 70, + "AATAAC": 71, + "AATAAG": 72, + "AATATA": 73, + "AATATT": 74, + "AATATC": 75, + "AATATG": 76, + "AATACA": 77, + "AATACT": 78, + "AATACC": 79, + "AATACG": 80, + "AATAGA": 81, + "AATAGT": 82, + "AATAGC": 83, + "AATAGG": 84, + "AATTAA": 85, + "AATTAT": 86, + "AATTAC": 87, + "AATTAG": 88, + "AATTTA": 89, + "AATTTT": 90, + "AATTTC": 91, + "AATTTG": 92, + "AATTCA": 93, + "AATTCT": 94, + "AATTCC": 95, + "AATTCG": 96, + "AATTGA": 97, + "AATTGT": 98, + "AATTGC": 99, + "AATTGG": 100, + "AATCAA": 101, + "AATCAT": 102, + "AATCAC": 103, + "AATCAG": 104, + "AATCTA": 105, + "AATCTT": 106, + "AATCTC": 107, + "AATCTG": 108, + "AATCCA": 109, + "AATCCT": 110, + "AATCCC": 111, + "AATCCG": 112, + "AATCGA": 113, + "AATCGT": 114, + "AATCGC": 115, + "AATCGG": 116, + "AATGAA": 117, + "AATGAT": 118, + "AATGAC": 119, + "AATGAG": 120, + "AATGTA": 121, + "AATGTT": 122, + "AATGTC": 123, + "AATGTG": 124, + "AATGCA": 125, + "AATGCT": 126, + "AATGCC": 127, + "AATGCG": 128, + "AATGGA": 129, + "AATGGT": 130, + "AATGGC": 131, + "AATGGG": 132, + "AACAAA": 133, + "AACAAT": 134, + "AACAAC": 135, + "AACAAG": 136, + "AACATA": 137, + "AACATT": 138, + "AACATC": 139, + "AACATG": 140, + "AACACA": 141, + "AACACT": 142, + "AACACC": 143, + "AACACG": 144, + "AACAGA": 145, + "AACAGT": 146, + "AACAGC": 147, + "AACAGG": 148, + "AACTAA": 149, + "AACTAT": 150, + "AACTAC": 151, + "AACTAG": 152, + "AACTTA": 153, + "AACTTT": 154, + "AACTTC": 155, + "AACTTG": 156, + "AACTCA": 157, + "AACTCT": 158, + "AACTCC": 159, + "AACTCG": 160, + "AACTGA": 161, + "AACTGT": 162, + "AACTGC": 163, + "AACTGG": 164, + "AACCAA": 165, + "AACCAT": 166, + "AACCAC": 167, + "AACCAG": 168, + "AACCTA": 169, + "AACCTT": 170, + "AACCTC": 171, + "AACCTG": 172, + "AACCCA": 173, + "AACCCT": 174, + "AACCCC": 175, + "AACCCG": 176, + "AACCGA": 177, + "AACCGT": 178, + "AACCGC": 179, + "AACCGG": 180, + "AACGAA": 181, + "AACGAT": 182, + "AACGAC": 183, + "AACGAG": 184, + "AACGTA": 185, + "AACGTT": 186, + "AACGTC": 187, + "AACGTG": 188, + "AACGCA": 189, + "AACGCT": 190, + "AACGCC": 191, + "AACGCG": 192, + "AACGGA": 193, + "AACGGT": 194, + "AACGGC": 195, + "AACGGG": 196, + "AAGAAA": 197, + "AAGAAT": 198, + "AAGAAC": 199, + "AAGAAG": 200, + "AAGATA": 201, + "AAGATT": 202, + "AAGATC": 203, + "AAGATG": 204, + "AAGACA": 205, + "AAGACT": 206, + "AAGACC": 207, + "AAGACG": 208, + "AAGAGA": 209, + "AAGAGT": 210, + "AAGAGC": 211, + "AAGAGG": 212, + "AAGTAA": 213, + "AAGTAT": 214, + "AAGTAC": 215, + "AAGTAG": 216, + "AAGTTA": 217, + "AAGTTT": 218, + "AAGTTC": 219, + "AAGTTG": 220, + "AAGTCA": 221, + "AAGTCT": 222, + "AAGTCC": 223, + "AAGTCG": 224, + "AAGTGA": 225, + "AAGTGT": 226, + "AAGTGC": 227, + "AAGTGG": 228, + "AAGCAA": 229, + "AAGCAT": 230, + "AAGCAC": 231, + "AAGCAG": 232, + "AAGCTA": 233, + "AAGCTT": 234, + "AAGCTC": 235, + "AAGCTG": 236, + "AAGCCA": 237, + "AAGCCT": 238, + "AAGCCC": 239, + "AAGCCG": 240, + "AAGCGA": 241, + "AAGCGT": 242, + "AAGCGC": 243, + "AAGCGG": 244, + "AAGGAA": 245, + "AAGGAT": 246, + "AAGGAC": 247, + "AAGGAG": 248, + "AAGGTA": 249, + "AAGGTT": 250, + "AAGGTC": 251, + "AAGGTG": 252, + "AAGGCA": 253, + "AAGGCT": 254, + "AAGGCC": 255, + "AAGGCG": 256, + "AAGGGA": 257, + "AAGGGT": 258, + "AAGGGC": 259, + "AAGGGG": 260, + "ATAAAA": 261, + "ATAAAT": 262, + "ATAAAC": 263, + "ATAAAG": 264, + "ATAATA": 265, + "ATAATT": 266, + "ATAATC": 267, + "ATAATG": 268, + "ATAACA": 269, + "ATAACT": 270, + "ATAACC": 271, + "ATAACG": 272, + "ATAAGA": 273, + "ATAAGT": 274, + "ATAAGC": 275, + "ATAAGG": 276, + "ATATAA": 277, + "ATATAT": 278, + "ATATAC": 279, + "ATATAG": 280, + "ATATTA": 281, + "ATATTT": 282, + "ATATTC": 283, + "ATATTG": 284, + "ATATCA": 285, + "ATATCT": 286, + "ATATCC": 287, + "ATATCG": 288, + "ATATGA": 289, + "ATATGT": 290, + "ATATGC": 291, + "ATATGG": 292, + "ATACAA": 293, + "ATACAT": 294, + "ATACAC": 295, + "ATACAG": 296, + "ATACTA": 297, + "ATACTT": 298, + "ATACTC": 299, + "ATACTG": 300, + "ATACCA": 301, + "ATACCT": 302, + "ATACCC": 303, + "ATACCG": 304, + "ATACGA": 305, + "ATACGT": 306, + "ATACGC": 307, + "ATACGG": 308, + "ATAGAA": 309, + "ATAGAT": 310, + "ATAGAC": 311, + "ATAGAG": 312, + "ATAGTA": 313, + "ATAGTT": 314, + "ATAGTC": 315, + "ATAGTG": 316, + "ATAGCA": 317, + "ATAGCT": 318, + "ATAGCC": 319, + "ATAGCG": 320, + "ATAGGA": 321, + "ATAGGT": 322, + "ATAGGC": 323, + "ATAGGG": 324, + "ATTAAA": 325, + "ATTAAT": 326, + "ATTAAC": 327, + "ATTAAG": 328, + "ATTATA": 329, + "ATTATT": 330, + "ATTATC": 331, + "ATTATG": 332, + "ATTACA": 333, + "ATTACT": 334, + "ATTACC": 335, + "ATTACG": 336, + "ATTAGA": 337, + "ATTAGT": 338, + "ATTAGC": 339, + "ATTAGG": 340, + "ATTTAA": 341, + "ATTTAT": 342, + "ATTTAC": 343, + "ATTTAG": 344, + "ATTTTA": 345, + "ATTTTT": 346, + "ATTTTC": 347, + "ATTTTG": 348, + "ATTTCA": 349, + "ATTTCT": 350, + "ATTTCC": 351, + "ATTTCG": 352, + "ATTTGA": 353, + "ATTTGT": 354, + "ATTTGC": 355, + "ATTTGG": 356, + "ATTCAA": 357, + "ATTCAT": 358, + "ATTCAC": 359, + "ATTCAG": 360, + "ATTCTA": 361, + "ATTCTT": 362, + "ATTCTC": 363, + "ATTCTG": 364, + "ATTCCA": 365, + "ATTCCT": 366, + "ATTCCC": 367, + "ATTCCG": 368, + "ATTCGA": 369, + "ATTCGT": 370, + "ATTCGC": 371, + "ATTCGG": 372, + "ATTGAA": 373, + "ATTGAT": 374, + "ATTGAC": 375, + "ATTGAG": 376, + "ATTGTA": 377, + "ATTGTT": 378, + "ATTGTC": 379, + "ATTGTG": 380, + "ATTGCA": 381, + "ATTGCT": 382, + "ATTGCC": 383, + "ATTGCG": 384, + "ATTGGA": 385, + "ATTGGT": 386, + "ATTGGC": 387, + "ATTGGG": 388, + "ATCAAA": 389, + "ATCAAT": 390, + "ATCAAC": 391, + "ATCAAG": 392, + "ATCATA": 393, + "ATCATT": 394, + "ATCATC": 395, + "ATCATG": 396, + "ATCACA": 397, + "ATCACT": 398, + "ATCACC": 399, + "ATCACG": 400, + "ATCAGA": 401, + "ATCAGT": 402, + "ATCAGC": 403, + "ATCAGG": 404, + "ATCTAA": 405, + "ATCTAT": 406, + "ATCTAC": 407, + "ATCTAG": 408, + "ATCTTA": 409, + "ATCTTT": 410, + "ATCTTC": 411, + "ATCTTG": 412, + "ATCTCA": 413, + "ATCTCT": 414, + "ATCTCC": 415, + "ATCTCG": 416, + "ATCTGA": 417, + "ATCTGT": 418, + "ATCTGC": 419, + "ATCTGG": 420, + "ATCCAA": 421, + "ATCCAT": 422, + "ATCCAC": 423, + "ATCCAG": 424, + "ATCCTA": 425, + "ATCCTT": 426, + "ATCCTC": 427, + "ATCCTG": 428, + "ATCCCA": 429, + "ATCCCT": 430, + "ATCCCC": 431, + "ATCCCG": 432, + "ATCCGA": 433, + "ATCCGT": 434, + "ATCCGC": 435, + "ATCCGG": 436, + "ATCGAA": 437, + "ATCGAT": 438, + "ATCGAC": 439, + "ATCGAG": 440, + "ATCGTA": 441, + "ATCGTT": 442, + "ATCGTC": 443, + "ATCGTG": 444, + "ATCGCA": 445, + "ATCGCT": 446, + "ATCGCC": 447, + "ATCGCG": 448, + "ATCGGA": 449, + "ATCGGT": 450, + "ATCGGC": 451, + "ATCGGG": 452, + "ATGAAA": 453, + "ATGAAT": 454, + "ATGAAC": 455, + "ATGAAG": 456, + "ATGATA": 457, + "ATGATT": 458, + "ATGATC": 459, + "ATGATG": 460, + "ATGACA": 461, + "ATGACT": 462, + "ATGACC": 463, + "ATGACG": 464, + "ATGAGA": 465, + "ATGAGT": 466, + "ATGAGC": 467, + "ATGAGG": 468, + "ATGTAA": 469, + "ATGTAT": 470, + "ATGTAC": 471, + "ATGTAG": 472, + "ATGTTA": 473, + "ATGTTT": 474, + "ATGTTC": 475, + "ATGTTG": 476, + "ATGTCA": 477, + "ATGTCT": 478, + "ATGTCC": 479, + "ATGTCG": 480, + "ATGTGA": 481, + "ATGTGT": 482, + "ATGTGC": 483, + "ATGTGG": 484, + "ATGCAA": 485, + "ATGCAT": 486, + "ATGCAC": 487, + "ATGCAG": 488, + "ATGCTA": 489, + "ATGCTT": 490, + "ATGCTC": 491, + "ATGCTG": 492, + "ATGCCA": 493, + "ATGCCT": 494, + "ATGCCC": 495, + "ATGCCG": 496, + "ATGCGA": 497, + "ATGCGT": 498, + "ATGCGC": 499, + "ATGCGG": 500, + "ATGGAA": 501, + "ATGGAT": 502, + "ATGGAC": 503, + "ATGGAG": 504, + "ATGGTA": 505, + "ATGGTT": 506, + "ATGGTC": 507, + "ATGGTG": 508, + "ATGGCA": 509, + "ATGGCT": 510, + "ATGGCC": 511, + "ATGGCG": 512, + "ATGGGA": 513, + "ATGGGT": 514, + "ATGGGC": 515, + "ATGGGG": 516, + "ACAAAA": 517, + "ACAAAT": 518, + "ACAAAC": 519, + "ACAAAG": 520, + "ACAATA": 521, + "ACAATT": 522, + "ACAATC": 523, + "ACAATG": 524, + "ACAACA": 525, + "ACAACT": 526, + "ACAACC": 527, + "ACAACG": 528, + "ACAAGA": 529, + "ACAAGT": 530, + "ACAAGC": 531, + "ACAAGG": 532, + "ACATAA": 533, + "ACATAT": 534, + "ACATAC": 535, + "ACATAG": 536, + "ACATTA": 537, + "ACATTT": 538, + "ACATTC": 539, + "ACATTG": 540, + "ACATCA": 541, + "ACATCT": 542, + "ACATCC": 543, + "ACATCG": 544, + "ACATGA": 545, + "ACATGT": 546, + "ACATGC": 547, + "ACATGG": 548, + "ACACAA": 549, + "ACACAT": 550, + "ACACAC": 551, + "ACACAG": 552, + "ACACTA": 553, + "ACACTT": 554, + "ACACTC": 555, + "ACACTG": 556, + "ACACCA": 557, + "ACACCT": 558, + "ACACCC": 559, + "ACACCG": 560, + "ACACGA": 561, + "ACACGT": 562, + "ACACGC": 563, + "ACACGG": 564, + "ACAGAA": 565, + "ACAGAT": 566, + "ACAGAC": 567, + "ACAGAG": 568, + "ACAGTA": 569, + "ACAGTT": 570, + "ACAGTC": 571, + "ACAGTG": 572, + "ACAGCA": 573, + "ACAGCT": 574, + "ACAGCC": 575, + "ACAGCG": 576, + "ACAGGA": 577, + "ACAGGT": 578, + "ACAGGC": 579, + "ACAGGG": 580, + "ACTAAA": 581, + "ACTAAT": 582, + "ACTAAC": 583, + "ACTAAG": 584, + "ACTATA": 585, + "ACTATT": 586, + "ACTATC": 587, + "ACTATG": 588, + "ACTACA": 589, + "ACTACT": 590, + "ACTACC": 591, + "ACTACG": 592, + "ACTAGA": 593, + "ACTAGT": 594, + "ACTAGC": 595, + "ACTAGG": 596, + "ACTTAA": 597, + "ACTTAT": 598, + "ACTTAC": 599, + "ACTTAG": 600, + "ACTTTA": 601, + "ACTTTT": 602, + "ACTTTC": 603, + "ACTTTG": 604, + "ACTTCA": 605, + "ACTTCT": 606, + "ACTTCC": 607, + "ACTTCG": 608, + "ACTTGA": 609, + "ACTTGT": 610, + "ACTTGC": 611, + "ACTTGG": 612, + "ACTCAA": 613, + "ACTCAT": 614, + "ACTCAC": 615, + "ACTCAG": 616, + "ACTCTA": 617, + "ACTCTT": 618, + "ACTCTC": 619, + "ACTCTG": 620, + "ACTCCA": 621, + "ACTCCT": 622, + "ACTCCC": 623, + "ACTCCG": 624, + "ACTCGA": 625, + "ACTCGT": 626, + "ACTCGC": 627, + "ACTCGG": 628, + "ACTGAA": 629, + "ACTGAT": 630, + "ACTGAC": 631, + "ACTGAG": 632, + "ACTGTA": 633, + "ACTGTT": 634, + "ACTGTC": 635, + "ACTGTG": 636, + "ACTGCA": 637, + "ACTGCT": 638, + "ACTGCC": 639, + "ACTGCG": 640, + "ACTGGA": 641, + "ACTGGT": 642, + "ACTGGC": 643, + "ACTGGG": 644, + "ACCAAA": 645, + "ACCAAT": 646, + "ACCAAC": 647, + "ACCAAG": 648, + "ACCATA": 649, + "ACCATT": 650, + "ACCATC": 651, + "ACCATG": 652, + "ACCACA": 653, + "ACCACT": 654, + "ACCACC": 655, + "ACCACG": 656, + "ACCAGA": 657, + "ACCAGT": 658, + "ACCAGC": 659, + "ACCAGG": 660, + "ACCTAA": 661, + "ACCTAT": 662, + "ACCTAC": 663, + "ACCTAG": 664, + "ACCTTA": 665, + "ACCTTT": 666, + "ACCTTC": 667, + "ACCTTG": 668, + "ACCTCA": 669, + "ACCTCT": 670, + "ACCTCC": 671, + "ACCTCG": 672, + "ACCTGA": 673, + "ACCTGT": 674, + "ACCTGC": 675, + "ACCTGG": 676, + "ACCCAA": 677, + "ACCCAT": 678, + "ACCCAC": 679, + "ACCCAG": 680, + "ACCCTA": 681, + "ACCCTT": 682, + "ACCCTC": 683, + "ACCCTG": 684, + "ACCCCA": 685, + "ACCCCT": 686, + "ACCCCC": 687, + "ACCCCG": 688, + "ACCCGA": 689, + "ACCCGT": 690, + "ACCCGC": 691, + "ACCCGG": 692, + "ACCGAA": 693, + "ACCGAT": 694, + "ACCGAC": 695, + "ACCGAG": 696, + "ACCGTA": 697, + "ACCGTT": 698, + "ACCGTC": 699, + "ACCGTG": 700, + "ACCGCA": 701, + "ACCGCT": 702, + "ACCGCC": 703, + "ACCGCG": 704, + "ACCGGA": 705, + "ACCGGT": 706, + "ACCGGC": 707, + "ACCGGG": 708, + "ACGAAA": 709, + "ACGAAT": 710, + "ACGAAC": 711, + "ACGAAG": 712, + "ACGATA": 713, + "ACGATT": 714, + "ACGATC": 715, + "ACGATG": 716, + "ACGACA": 717, + "ACGACT": 718, + "ACGACC": 719, + "ACGACG": 720, + "ACGAGA": 721, + "ACGAGT": 722, + "ACGAGC": 723, + "ACGAGG": 724, + "ACGTAA": 725, + "ACGTAT": 726, + "ACGTAC": 727, + "ACGTAG": 728, + "ACGTTA": 729, + "ACGTTT": 730, + "ACGTTC": 731, + "ACGTTG": 732, + "ACGTCA": 733, + "ACGTCT": 734, + "ACGTCC": 735, + "ACGTCG": 736, + "ACGTGA": 737, + "ACGTGT": 738, + "ACGTGC": 739, + "ACGTGG": 740, + "ACGCAA": 741, + "ACGCAT": 742, + "ACGCAC": 743, + "ACGCAG": 744, + "ACGCTA": 745, + "ACGCTT": 746, + "ACGCTC": 747, + "ACGCTG": 748, + "ACGCCA": 749, + "ACGCCT": 750, + "ACGCCC": 751, + "ACGCCG": 752, + "ACGCGA": 753, + "ACGCGT": 754, + "ACGCGC": 755, + "ACGCGG": 756, + "ACGGAA": 757, + "ACGGAT": 758, + "ACGGAC": 759, + "ACGGAG": 760, + "ACGGTA": 761, + "ACGGTT": 762, + "ACGGTC": 763, + "ACGGTG": 764, + "ACGGCA": 765, + "ACGGCT": 766, + "ACGGCC": 767, + "ACGGCG": 768, + "ACGGGA": 769, + "ACGGGT": 770, + "ACGGGC": 771, + "ACGGGG": 772, + "AGAAAA": 773, + "AGAAAT": 774, + "AGAAAC": 775, + "AGAAAG": 776, + "AGAATA": 777, + "AGAATT": 778, + "AGAATC": 779, + "AGAATG": 780, + "AGAACA": 781, + "AGAACT": 782, + "AGAACC": 783, + "AGAACG": 784, + "AGAAGA": 785, + "AGAAGT": 786, + "AGAAGC": 787, + "AGAAGG": 788, + "AGATAA": 789, + "AGATAT": 790, + "AGATAC": 791, + "AGATAG": 792, + "AGATTA": 793, + "AGATTT": 794, + "AGATTC": 795, + "AGATTG": 796, + "AGATCA": 797, + "AGATCT": 798, + "AGATCC": 799, + "AGATCG": 800, + "AGATGA": 801, + "AGATGT": 802, + "AGATGC": 803, + "AGATGG": 804, + "AGACAA": 805, + "AGACAT": 806, + "AGACAC": 807, + "AGACAG": 808, + "AGACTA": 809, + "AGACTT": 810, + "AGACTC": 811, + "AGACTG": 812, + "AGACCA": 813, + "AGACCT": 814, + "AGACCC": 815, + "AGACCG": 816, + "AGACGA": 817, + "AGACGT": 818, + "AGACGC": 819, + "AGACGG": 820, + "AGAGAA": 821, + "AGAGAT": 822, + "AGAGAC": 823, + "AGAGAG": 824, + "AGAGTA": 825, + "AGAGTT": 826, + "AGAGTC": 827, + "AGAGTG": 828, + "AGAGCA": 829, + "AGAGCT": 830, + "AGAGCC": 831, + "AGAGCG": 832, + "AGAGGA": 833, + "AGAGGT": 834, + "AGAGGC": 835, + "AGAGGG": 836, + "AGTAAA": 837, + "AGTAAT": 838, + "AGTAAC": 839, + "AGTAAG": 840, + "AGTATA": 841, + "AGTATT": 842, + "AGTATC": 843, + "AGTATG": 844, + "AGTACA": 845, + "AGTACT": 846, + "AGTACC": 847, + "AGTACG": 848, + "AGTAGA": 849, + "AGTAGT": 850, + "AGTAGC": 851, + "AGTAGG": 852, + "AGTTAA": 853, + "AGTTAT": 854, + "AGTTAC": 855, + "AGTTAG": 856, + "AGTTTA": 857, + "AGTTTT": 858, + "AGTTTC": 859, + "AGTTTG": 860, + "AGTTCA": 861, + "AGTTCT": 862, + "AGTTCC": 863, + "AGTTCG": 864, + "AGTTGA": 865, + "AGTTGT": 866, + "AGTTGC": 867, + "AGTTGG": 868, + "AGTCAA": 869, + "AGTCAT": 870, + "AGTCAC": 871, + "AGTCAG": 872, + "AGTCTA": 873, + "AGTCTT": 874, + "AGTCTC": 875, + "AGTCTG": 876, + "AGTCCA": 877, + "AGTCCT": 878, + "AGTCCC": 879, + "AGTCCG": 880, + "AGTCGA": 881, + "AGTCGT": 882, + "AGTCGC": 883, + "AGTCGG": 884, + "AGTGAA": 885, + "AGTGAT": 886, + "AGTGAC": 887, + "AGTGAG": 888, + "AGTGTA": 889, + "AGTGTT": 890, + "AGTGTC": 891, + "AGTGTG": 892, + "AGTGCA": 893, + "AGTGCT": 894, + "AGTGCC": 895, + "AGTGCG": 896, + "AGTGGA": 897, + "AGTGGT": 898, + "AGTGGC": 899, + "AGTGGG": 900, + "AGCAAA": 901, + "AGCAAT": 902, + "AGCAAC": 903, + "AGCAAG": 904, + "AGCATA": 905, + "AGCATT": 906, + "AGCATC": 907, + "AGCATG": 908, + "AGCACA": 909, + "AGCACT": 910, + "AGCACC": 911, + "AGCACG": 912, + "AGCAGA": 913, + "AGCAGT": 914, + "AGCAGC": 915, + "AGCAGG": 916, + "AGCTAA": 917, + "AGCTAT": 918, + "AGCTAC": 919, + "AGCTAG": 920, + "AGCTTA": 921, + "AGCTTT": 922, + "AGCTTC": 923, + "AGCTTG": 924, + "AGCTCA": 925, + "AGCTCT": 926, + "AGCTCC": 927, + "AGCTCG": 928, + "AGCTGA": 929, + "AGCTGT": 930, + "AGCTGC": 931, + "AGCTGG": 932, + "AGCCAA": 933, + "AGCCAT": 934, + "AGCCAC": 935, + "AGCCAG": 936, + "AGCCTA": 937, + "AGCCTT": 938, + "AGCCTC": 939, + "AGCCTG": 940, + "AGCCCA": 941, + "AGCCCT": 942, + "AGCCCC": 943, + "AGCCCG": 944, + "AGCCGA": 945, + "AGCCGT": 946, + "AGCCGC": 947, + "AGCCGG": 948, + "AGCGAA": 949, + "AGCGAT": 950, + "AGCGAC": 951, + "AGCGAG": 952, + "AGCGTA": 953, + "AGCGTT": 954, + "AGCGTC": 955, + "AGCGTG": 956, + "AGCGCA": 957, + "AGCGCT": 958, + "AGCGCC": 959, + "AGCGCG": 960, + "AGCGGA": 961, + "AGCGGT": 962, + "AGCGGC": 963, + "AGCGGG": 964, + "AGGAAA": 965, + "AGGAAT": 966, + "AGGAAC": 967, + "AGGAAG": 968, + "AGGATA": 969, + "AGGATT": 970, + "AGGATC": 971, + "AGGATG": 972, + "AGGACA": 973, + "AGGACT": 974, + "AGGACC": 975, + "AGGACG": 976, + "AGGAGA": 977, + "AGGAGT": 978, + "AGGAGC": 979, + "AGGAGG": 980, + "AGGTAA": 981, + "AGGTAT": 982, + "AGGTAC": 983, + "AGGTAG": 984, + "AGGTTA": 985, + "AGGTTT": 986, + "AGGTTC": 987, + "AGGTTG": 988, + "AGGTCA": 989, + "AGGTCT": 990, + "AGGTCC": 991, + "AGGTCG": 992, + "AGGTGA": 993, + "AGGTGT": 994, + "AGGTGC": 995, + "AGGTGG": 996, + "AGGCAA": 997, + "AGGCAT": 998, + "AGGCAC": 999, + "AGGCAG": 1000, + "AGGCTA": 1001, + "AGGCTT": 1002, + "AGGCTC": 1003, + "AGGCTG": 1004, + "AGGCCA": 1005, + "AGGCCT": 1006, + "AGGCCC": 1007, + "AGGCCG": 1008, + "AGGCGA": 1009, + "AGGCGT": 1010, + "AGGCGC": 1011, + "AGGCGG": 1012, + "AGGGAA": 1013, + "AGGGAT": 1014, + "AGGGAC": 1015, + "AGGGAG": 1016, + "AGGGTA": 1017, + "AGGGTT": 1018, + "AGGGTC": 1019, + "AGGGTG": 1020, + "AGGGCA": 1021, + "AGGGCT": 1022, + "AGGGCC": 1023, + "AGGGCG": 1024, + "AGGGGA": 1025, + "AGGGGT": 1026, + "AGGGGC": 1027, + "AGGGGG": 1028, + "TAAAAA": 1029, + "TAAAAT": 1030, + "TAAAAC": 1031, + "TAAAAG": 1032, + "TAAATA": 1033, + "TAAATT": 1034, + "TAAATC": 1035, + "TAAATG": 1036, + "TAAACA": 1037, + "TAAACT": 1038, + "TAAACC": 1039, + "TAAACG": 1040, + "TAAAGA": 1041, + "TAAAGT": 1042, + "TAAAGC": 1043, + "TAAAGG": 1044, + "TAATAA": 1045, + "TAATAT": 1046, + "TAATAC": 1047, + "TAATAG": 1048, + "TAATTA": 1049, + "TAATTT": 1050, + "TAATTC": 1051, + "TAATTG": 1052, + "TAATCA": 1053, + "TAATCT": 1054, + "TAATCC": 1055, + "TAATCG": 1056, + "TAATGA": 1057, + "TAATGT": 1058, + "TAATGC": 1059, + "TAATGG": 1060, + "TAACAA": 1061, + "TAACAT": 1062, + "TAACAC": 1063, + "TAACAG": 1064, + "TAACTA": 1065, + "TAACTT": 1066, + "TAACTC": 1067, + "TAACTG": 1068, + "TAACCA": 1069, + "TAACCT": 1070, + "TAACCC": 1071, + "TAACCG": 1072, + "TAACGA": 1073, + "TAACGT": 1074, + "TAACGC": 1075, + "TAACGG": 1076, + "TAAGAA": 1077, + "TAAGAT": 1078, + "TAAGAC": 1079, + "TAAGAG": 1080, + "TAAGTA": 1081, + "TAAGTT": 1082, + "TAAGTC": 1083, + "TAAGTG": 1084, + "TAAGCA": 1085, + "TAAGCT": 1086, + "TAAGCC": 1087, + "TAAGCG": 1088, + "TAAGGA": 1089, + "TAAGGT": 1090, + "TAAGGC": 1091, + "TAAGGG": 1092, + "TATAAA": 1093, + "TATAAT": 1094, + "TATAAC": 1095, + "TATAAG": 1096, + "TATATA": 1097, + "TATATT": 1098, + "TATATC": 1099, + "TATATG": 1100, + "TATACA": 1101, + "TATACT": 1102, + "TATACC": 1103, + "TATACG": 1104, + "TATAGA": 1105, + "TATAGT": 1106, + "TATAGC": 1107, + "TATAGG": 1108, + "TATTAA": 1109, + "TATTAT": 1110, + "TATTAC": 1111, + "TATTAG": 1112, + "TATTTA": 1113, + "TATTTT": 1114, + "TATTTC": 1115, + "TATTTG": 1116, + "TATTCA": 1117, + "TATTCT": 1118, + "TATTCC": 1119, + "TATTCG": 1120, + "TATTGA": 1121, + "TATTGT": 1122, + "TATTGC": 1123, + "TATTGG": 1124, + "TATCAA": 1125, + "TATCAT": 1126, + "TATCAC": 1127, + "TATCAG": 1128, + "TATCTA": 1129, + "TATCTT": 1130, + "TATCTC": 1131, + "TATCTG": 1132, + "TATCCA": 1133, + "TATCCT": 1134, + "TATCCC": 1135, + "TATCCG": 1136, + "TATCGA": 1137, + "TATCGT": 1138, + "TATCGC": 1139, + "TATCGG": 1140, + "TATGAA": 1141, + "TATGAT": 1142, + "TATGAC": 1143, + "TATGAG": 1144, + "TATGTA": 1145, + "TATGTT": 1146, + "TATGTC": 1147, + "TATGTG": 1148, + "TATGCA": 1149, + "TATGCT": 1150, + "TATGCC": 1151, + "TATGCG": 1152, + "TATGGA": 1153, + "TATGGT": 1154, + "TATGGC": 1155, + "TATGGG": 1156, + "TACAAA": 1157, + "TACAAT": 1158, + "TACAAC": 1159, + "TACAAG": 1160, + "TACATA": 1161, + "TACATT": 1162, + "TACATC": 1163, + "TACATG": 1164, + "TACACA": 1165, + "TACACT": 1166, + "TACACC": 1167, + "TACACG": 1168, + "TACAGA": 1169, + "TACAGT": 1170, + "TACAGC": 1171, + "TACAGG": 1172, + "TACTAA": 1173, + "TACTAT": 1174, + "TACTAC": 1175, + "TACTAG": 1176, + "TACTTA": 1177, + "TACTTT": 1178, + "TACTTC": 1179, + "TACTTG": 1180, + "TACTCA": 1181, + "TACTCT": 1182, + "TACTCC": 1183, + "TACTCG": 1184, + "TACTGA": 1185, + "TACTGT": 1186, + "TACTGC": 1187, + "TACTGG": 1188, + "TACCAA": 1189, + "TACCAT": 1190, + "TACCAC": 1191, + "TACCAG": 1192, + "TACCTA": 1193, + "TACCTT": 1194, + "TACCTC": 1195, + "TACCTG": 1196, + "TACCCA": 1197, + "TACCCT": 1198, + "TACCCC": 1199, + "TACCCG": 1200, + "TACCGA": 1201, + "TACCGT": 1202, + "TACCGC": 1203, + "TACCGG": 1204, + "TACGAA": 1205, + "TACGAT": 1206, + "TACGAC": 1207, + "TACGAG": 1208, + "TACGTA": 1209, + "TACGTT": 1210, + "TACGTC": 1211, + "TACGTG": 1212, + "TACGCA": 1213, + "TACGCT": 1214, + "TACGCC": 1215, + "TACGCG": 1216, + "TACGGA": 1217, + "TACGGT": 1218, + "TACGGC": 1219, + "TACGGG": 1220, + "TAGAAA": 1221, + "TAGAAT": 1222, + "TAGAAC": 1223, + "TAGAAG": 1224, + "TAGATA": 1225, + "TAGATT": 1226, + "TAGATC": 1227, + "TAGATG": 1228, + "TAGACA": 1229, + "TAGACT": 1230, + "TAGACC": 1231, + "TAGACG": 1232, + "TAGAGA": 1233, + "TAGAGT": 1234, + "TAGAGC": 1235, + "TAGAGG": 1236, + "TAGTAA": 1237, + "TAGTAT": 1238, + "TAGTAC": 1239, + "TAGTAG": 1240, + "TAGTTA": 1241, + "TAGTTT": 1242, + "TAGTTC": 1243, + "TAGTTG": 1244, + "TAGTCA": 1245, + "TAGTCT": 1246, + "TAGTCC": 1247, + "TAGTCG": 1248, + "TAGTGA": 1249, + "TAGTGT": 1250, + "TAGTGC": 1251, + "TAGTGG": 1252, + "TAGCAA": 1253, + "TAGCAT": 1254, + "TAGCAC": 1255, + "TAGCAG": 1256, + "TAGCTA": 1257, + "TAGCTT": 1258, + "TAGCTC": 1259, + "TAGCTG": 1260, + "TAGCCA": 1261, + "TAGCCT": 1262, + "TAGCCC": 1263, + "TAGCCG": 1264, + "TAGCGA": 1265, + "TAGCGT": 1266, + "TAGCGC": 1267, + "TAGCGG": 1268, + "TAGGAA": 1269, + "TAGGAT": 1270, + "TAGGAC": 1271, + "TAGGAG": 1272, + "TAGGTA": 1273, + "TAGGTT": 1274, + "TAGGTC": 1275, + "TAGGTG": 1276, + "TAGGCA": 1277, + "TAGGCT": 1278, + "TAGGCC": 1279, + "TAGGCG": 1280, + "TAGGGA": 1281, + "TAGGGT": 1282, + "TAGGGC": 1283, + "TAGGGG": 1284, + "TTAAAA": 1285, + "TTAAAT": 1286, + "TTAAAC": 1287, + "TTAAAG": 1288, + "TTAATA": 1289, + "TTAATT": 1290, + "TTAATC": 1291, + "TTAATG": 1292, + "TTAACA": 1293, + "TTAACT": 1294, + "TTAACC": 1295, + "TTAACG": 1296, + "TTAAGA": 1297, + "TTAAGT": 1298, + "TTAAGC": 1299, + "TTAAGG": 1300, + "TTATAA": 1301, + "TTATAT": 1302, + "TTATAC": 1303, + "TTATAG": 1304, + "TTATTA": 1305, + "TTATTT": 1306, + "TTATTC": 1307, + "TTATTG": 1308, + "TTATCA": 1309, + "TTATCT": 1310, + "TTATCC": 1311, + "TTATCG": 1312, + "TTATGA": 1313, + "TTATGT": 1314, + "TTATGC": 1315, + "TTATGG": 1316, + "TTACAA": 1317, + "TTACAT": 1318, + "TTACAC": 1319, + "TTACAG": 1320, + "TTACTA": 1321, + "TTACTT": 1322, + "TTACTC": 1323, + "TTACTG": 1324, + "TTACCA": 1325, + "TTACCT": 1326, + "TTACCC": 1327, + "TTACCG": 1328, + "TTACGA": 1329, + "TTACGT": 1330, + "TTACGC": 1331, + "TTACGG": 1332, + "TTAGAA": 1333, + "TTAGAT": 1334, + "TTAGAC": 1335, + "TTAGAG": 1336, + "TTAGTA": 1337, + "TTAGTT": 1338, + "TTAGTC": 1339, + "TTAGTG": 1340, + "TTAGCA": 1341, + "TTAGCT": 1342, + "TTAGCC": 1343, + "TTAGCG": 1344, + "TTAGGA": 1345, + "TTAGGT": 1346, + "TTAGGC": 1347, + "TTAGGG": 1348, + "TTTAAA": 1349, + "TTTAAT": 1350, + "TTTAAC": 1351, + "TTTAAG": 1352, + "TTTATA": 1353, + "TTTATT": 1354, + "TTTATC": 1355, + "TTTATG": 1356, + "TTTACA": 1357, + "TTTACT": 1358, + "TTTACC": 1359, + "TTTACG": 1360, + "TTTAGA": 1361, + "TTTAGT": 1362, + "TTTAGC": 1363, + "TTTAGG": 1364, + "TTTTAA": 1365, + "TTTTAT": 1366, + "TTTTAC": 1367, + "TTTTAG": 1368, + "TTTTTA": 1369, + "TTTTTT": 1370, + "TTTTTC": 1371, + "TTTTTG": 1372, + "TTTTCA": 1373, + "TTTTCT": 1374, + "TTTTCC": 1375, + "TTTTCG": 1376, + "TTTTGA": 1377, + "TTTTGT": 1378, + "TTTTGC": 1379, + "TTTTGG": 1380, + "TTTCAA": 1381, + "TTTCAT": 1382, + "TTTCAC": 1383, + "TTTCAG": 1384, + "TTTCTA": 1385, + "TTTCTT": 1386, + "TTTCTC": 1387, + "TTTCTG": 1388, + "TTTCCA": 1389, + "TTTCCT": 1390, + "TTTCCC": 1391, + "TTTCCG": 1392, + "TTTCGA": 1393, + "TTTCGT": 1394, + "TTTCGC": 1395, + "TTTCGG": 1396, + "TTTGAA": 1397, + "TTTGAT": 1398, + "TTTGAC": 1399, + "TTTGAG": 1400, + "TTTGTA": 1401, + "TTTGTT": 1402, + "TTTGTC": 1403, + "TTTGTG": 1404, + "TTTGCA": 1405, + "TTTGCT": 1406, + "TTTGCC": 1407, + "TTTGCG": 1408, + "TTTGGA": 1409, + "TTTGGT": 1410, + "TTTGGC": 1411, + "TTTGGG": 1412, + "TTCAAA": 1413, + "TTCAAT": 1414, + "TTCAAC": 1415, + "TTCAAG": 1416, + "TTCATA": 1417, + "TTCATT": 1418, + "TTCATC": 1419, + "TTCATG": 1420, + "TTCACA": 1421, + "TTCACT": 1422, + "TTCACC": 1423, + "TTCACG": 1424, + "TTCAGA": 1425, + "TTCAGT": 1426, + "TTCAGC": 1427, + "TTCAGG": 1428, + "TTCTAA": 1429, + "TTCTAT": 1430, + "TTCTAC": 1431, + "TTCTAG": 1432, + "TTCTTA": 1433, + "TTCTTT": 1434, + "TTCTTC": 1435, + "TTCTTG": 1436, + "TTCTCA": 1437, + "TTCTCT": 1438, + "TTCTCC": 1439, + "TTCTCG": 1440, + "TTCTGA": 1441, + "TTCTGT": 1442, + "TTCTGC": 1443, + "TTCTGG": 1444, + "TTCCAA": 1445, + "TTCCAT": 1446, + "TTCCAC": 1447, + "TTCCAG": 1448, + "TTCCTA": 1449, + "TTCCTT": 1450, + "TTCCTC": 1451, + "TTCCTG": 1452, + "TTCCCA": 1453, + "TTCCCT": 1454, + "TTCCCC": 1455, + "TTCCCG": 1456, + "TTCCGA": 1457, + "TTCCGT": 1458, + "TTCCGC": 1459, + "TTCCGG": 1460, + "TTCGAA": 1461, + "TTCGAT": 1462, + "TTCGAC": 1463, + "TTCGAG": 1464, + "TTCGTA": 1465, + "TTCGTT": 1466, + "TTCGTC": 1467, + "TTCGTG": 1468, + "TTCGCA": 1469, + "TTCGCT": 1470, + "TTCGCC": 1471, + "TTCGCG": 1472, + "TTCGGA": 1473, + "TTCGGT": 1474, + "TTCGGC": 1475, + "TTCGGG": 1476, + "TTGAAA": 1477, + "TTGAAT": 1478, + "TTGAAC": 1479, + "TTGAAG": 1480, + "TTGATA": 1481, + "TTGATT": 1482, + "TTGATC": 1483, + "TTGATG": 1484, + "TTGACA": 1485, + "TTGACT": 1486, + "TTGACC": 1487, + "TTGACG": 1488, + "TTGAGA": 1489, + "TTGAGT": 1490, + "TTGAGC": 1491, + "TTGAGG": 1492, + "TTGTAA": 1493, + "TTGTAT": 1494, + "TTGTAC": 1495, + "TTGTAG": 1496, + "TTGTTA": 1497, + "TTGTTT": 1498, + "TTGTTC": 1499, + "TTGTTG": 1500, + "TTGTCA": 1501, + "TTGTCT": 1502, + "TTGTCC": 1503, + "TTGTCG": 1504, + "TTGTGA": 1505, + "TTGTGT": 1506, + "TTGTGC": 1507, + "TTGTGG": 1508, + "TTGCAA": 1509, + "TTGCAT": 1510, + "TTGCAC": 1511, + "TTGCAG": 1512, + "TTGCTA": 1513, + "TTGCTT": 1514, + "TTGCTC": 1515, + "TTGCTG": 1516, + "TTGCCA": 1517, + "TTGCCT": 1518, + "TTGCCC": 1519, + "TTGCCG": 1520, + "TTGCGA": 1521, + "TTGCGT": 1522, + "TTGCGC": 1523, + "TTGCGG": 1524, + "TTGGAA": 1525, + "TTGGAT": 1526, + "TTGGAC": 1527, + "TTGGAG": 1528, + "TTGGTA": 1529, + "TTGGTT": 1530, + "TTGGTC": 1531, + "TTGGTG": 1532, + "TTGGCA": 1533, + "TTGGCT": 1534, + "TTGGCC": 1535, + "TTGGCG": 1536, + "TTGGGA": 1537, + "TTGGGT": 1538, + "TTGGGC": 1539, + "TTGGGG": 1540, + "TCAAAA": 1541, + "TCAAAT": 1542, + "TCAAAC": 1543, + "TCAAAG": 1544, + "TCAATA": 1545, + "TCAATT": 1546, + "TCAATC": 1547, + "TCAATG": 1548, + "TCAACA": 1549, + "TCAACT": 1550, + "TCAACC": 1551, + "TCAACG": 1552, + "TCAAGA": 1553, + "TCAAGT": 1554, + "TCAAGC": 1555, + "TCAAGG": 1556, + "TCATAA": 1557, + "TCATAT": 1558, + "TCATAC": 1559, + "TCATAG": 1560, + "TCATTA": 1561, + "TCATTT": 1562, + "TCATTC": 1563, + "TCATTG": 1564, + "TCATCA": 1565, + "TCATCT": 1566, + "TCATCC": 1567, + "TCATCG": 1568, + "TCATGA": 1569, + "TCATGT": 1570, + "TCATGC": 1571, + "TCATGG": 1572, + "TCACAA": 1573, + "TCACAT": 1574, + "TCACAC": 1575, + "TCACAG": 1576, + "TCACTA": 1577, + "TCACTT": 1578, + "TCACTC": 1579, + "TCACTG": 1580, + "TCACCA": 1581, + "TCACCT": 1582, + "TCACCC": 1583, + "TCACCG": 1584, + "TCACGA": 1585, + "TCACGT": 1586, + "TCACGC": 1587, + "TCACGG": 1588, + "TCAGAA": 1589, + "TCAGAT": 1590, + "TCAGAC": 1591, + "TCAGAG": 1592, + "TCAGTA": 1593, + "TCAGTT": 1594, + "TCAGTC": 1595, + "TCAGTG": 1596, + "TCAGCA": 1597, + "TCAGCT": 1598, + "TCAGCC": 1599, + "TCAGCG": 1600, + "TCAGGA": 1601, + "TCAGGT": 1602, + "TCAGGC": 1603, + "TCAGGG": 1604, + "TCTAAA": 1605, + "TCTAAT": 1606, + "TCTAAC": 1607, + "TCTAAG": 1608, + "TCTATA": 1609, + "TCTATT": 1610, + "TCTATC": 1611, + "TCTATG": 1612, + "TCTACA": 1613, + "TCTACT": 1614, + "TCTACC": 1615, + "TCTACG": 1616, + "TCTAGA": 1617, + "TCTAGT": 1618, + "TCTAGC": 1619, + "TCTAGG": 1620, + "TCTTAA": 1621, + "TCTTAT": 1622, + "TCTTAC": 1623, + "TCTTAG": 1624, + "TCTTTA": 1625, + "TCTTTT": 1626, + "TCTTTC": 1627, + "TCTTTG": 1628, + "TCTTCA": 1629, + "TCTTCT": 1630, + "TCTTCC": 1631, + "TCTTCG": 1632, + "TCTTGA": 1633, + "TCTTGT": 1634, + "TCTTGC": 1635, + "TCTTGG": 1636, + "TCTCAA": 1637, + "TCTCAT": 1638, + "TCTCAC": 1639, + "TCTCAG": 1640, + "TCTCTA": 1641, + "TCTCTT": 1642, + "TCTCTC": 1643, + "TCTCTG": 1644, + "TCTCCA": 1645, + "TCTCCT": 1646, + "TCTCCC": 1647, + "TCTCCG": 1648, + "TCTCGA": 1649, + "TCTCGT": 1650, + "TCTCGC": 1651, + "TCTCGG": 1652, + "TCTGAA": 1653, + "TCTGAT": 1654, + "TCTGAC": 1655, + "TCTGAG": 1656, + "TCTGTA": 1657, + "TCTGTT": 1658, + "TCTGTC": 1659, + "TCTGTG": 1660, + "TCTGCA": 1661, + "TCTGCT": 1662, + "TCTGCC": 1663, + "TCTGCG": 1664, + "TCTGGA": 1665, + "TCTGGT": 1666, + "TCTGGC": 1667, + "TCTGGG": 1668, + "TCCAAA": 1669, + "TCCAAT": 1670, + "TCCAAC": 1671, + "TCCAAG": 1672, + "TCCATA": 1673, + "TCCATT": 1674, + "TCCATC": 1675, + "TCCATG": 1676, + "TCCACA": 1677, + "TCCACT": 1678, + "TCCACC": 1679, + "TCCACG": 1680, + "TCCAGA": 1681, + "TCCAGT": 1682, + "TCCAGC": 1683, + "TCCAGG": 1684, + "TCCTAA": 1685, + "TCCTAT": 1686, + "TCCTAC": 1687, + "TCCTAG": 1688, + "TCCTTA": 1689, + "TCCTTT": 1690, + "TCCTTC": 1691, + "TCCTTG": 1692, + "TCCTCA": 1693, + "TCCTCT": 1694, + "TCCTCC": 1695, + "TCCTCG": 1696, + "TCCTGA": 1697, + "TCCTGT": 1698, + "TCCTGC": 1699, + "TCCTGG": 1700, + "TCCCAA": 1701, + "TCCCAT": 1702, + "TCCCAC": 1703, + "TCCCAG": 1704, + "TCCCTA": 1705, + "TCCCTT": 1706, + "TCCCTC": 1707, + "TCCCTG": 1708, + "TCCCCA": 1709, + "TCCCCT": 1710, + "TCCCCC": 1711, + "TCCCCG": 1712, + "TCCCGA": 1713, + "TCCCGT": 1714, + "TCCCGC": 1715, + "TCCCGG": 1716, + "TCCGAA": 1717, + "TCCGAT": 1718, + "TCCGAC": 1719, + "TCCGAG": 1720, + "TCCGTA": 1721, + "TCCGTT": 1722, + "TCCGTC": 1723, + "TCCGTG": 1724, + "TCCGCA": 1725, + "TCCGCT": 1726, + "TCCGCC": 1727, + "TCCGCG": 1728, + "TCCGGA": 1729, + "TCCGGT": 1730, + "TCCGGC": 1731, + "TCCGGG": 1732, + "TCGAAA": 1733, + "TCGAAT": 1734, + "TCGAAC": 1735, + "TCGAAG": 1736, + "TCGATA": 1737, + "TCGATT": 1738, + "TCGATC": 1739, + "TCGATG": 1740, + "TCGACA": 1741, + "TCGACT": 1742, + "TCGACC": 1743, + "TCGACG": 1744, + "TCGAGA": 1745, + "TCGAGT": 1746, + "TCGAGC": 1747, + "TCGAGG": 1748, + "TCGTAA": 1749, + "TCGTAT": 1750, + "TCGTAC": 1751, + "TCGTAG": 1752, + "TCGTTA": 1753, + "TCGTTT": 1754, + "TCGTTC": 1755, + "TCGTTG": 1756, + "TCGTCA": 1757, + "TCGTCT": 1758, + "TCGTCC": 1759, + "TCGTCG": 1760, + "TCGTGA": 1761, + "TCGTGT": 1762, + "TCGTGC": 1763, + "TCGTGG": 1764, + "TCGCAA": 1765, + "TCGCAT": 1766, + "TCGCAC": 1767, + "TCGCAG": 1768, + "TCGCTA": 1769, + "TCGCTT": 1770, + "TCGCTC": 1771, + "TCGCTG": 1772, + "TCGCCA": 1773, + "TCGCCT": 1774, + "TCGCCC": 1775, + "TCGCCG": 1776, + "TCGCGA": 1777, + "TCGCGT": 1778, + "TCGCGC": 1779, + "TCGCGG": 1780, + "TCGGAA": 1781, + "TCGGAT": 1782, + "TCGGAC": 1783, + "TCGGAG": 1784, + "TCGGTA": 1785, + "TCGGTT": 1786, + "TCGGTC": 1787, + "TCGGTG": 1788, + "TCGGCA": 1789, + "TCGGCT": 1790, + "TCGGCC": 1791, + "TCGGCG": 1792, + "TCGGGA": 1793, + "TCGGGT": 1794, + "TCGGGC": 1795, + "TCGGGG": 1796, + "TGAAAA": 1797, + "TGAAAT": 1798, + "TGAAAC": 1799, + "TGAAAG": 1800, + "TGAATA": 1801, + "TGAATT": 1802, + "TGAATC": 1803, + "TGAATG": 1804, + "TGAACA": 1805, + "TGAACT": 1806, + "TGAACC": 1807, + "TGAACG": 1808, + "TGAAGA": 1809, + "TGAAGT": 1810, + "TGAAGC": 1811, + "TGAAGG": 1812, + "TGATAA": 1813, + "TGATAT": 1814, + "TGATAC": 1815, + "TGATAG": 1816, + "TGATTA": 1817, + "TGATTT": 1818, + "TGATTC": 1819, + "TGATTG": 1820, + "TGATCA": 1821, + "TGATCT": 1822, + "TGATCC": 1823, + "TGATCG": 1824, + "TGATGA": 1825, + "TGATGT": 1826, + "TGATGC": 1827, + "TGATGG": 1828, + "TGACAA": 1829, + "TGACAT": 1830, + "TGACAC": 1831, + "TGACAG": 1832, + "TGACTA": 1833, + "TGACTT": 1834, + "TGACTC": 1835, + "TGACTG": 1836, + "TGACCA": 1837, + "TGACCT": 1838, + "TGACCC": 1839, + "TGACCG": 1840, + "TGACGA": 1841, + "TGACGT": 1842, + "TGACGC": 1843, + "TGACGG": 1844, + "TGAGAA": 1845, + "TGAGAT": 1846, + "TGAGAC": 1847, + "TGAGAG": 1848, + "TGAGTA": 1849, + "TGAGTT": 1850, + "TGAGTC": 1851, + "TGAGTG": 1852, + "TGAGCA": 1853, + "TGAGCT": 1854, + "TGAGCC": 1855, + "TGAGCG": 1856, + "TGAGGA": 1857, + "TGAGGT": 1858, + "TGAGGC": 1859, + "TGAGGG": 1860, + "TGTAAA": 1861, + "TGTAAT": 1862, + "TGTAAC": 1863, + "TGTAAG": 1864, + "TGTATA": 1865, + "TGTATT": 1866, + "TGTATC": 1867, + "TGTATG": 1868, + "TGTACA": 1869, + "TGTACT": 1870, + "TGTACC": 1871, + "TGTACG": 1872, + "TGTAGA": 1873, + "TGTAGT": 1874, + "TGTAGC": 1875, + "TGTAGG": 1876, + "TGTTAA": 1877, + "TGTTAT": 1878, + "TGTTAC": 1879, + "TGTTAG": 1880, + "TGTTTA": 1881, + "TGTTTT": 1882, + "TGTTTC": 1883, + "TGTTTG": 1884, + "TGTTCA": 1885, + "TGTTCT": 1886, + "TGTTCC": 1887, + "TGTTCG": 1888, + "TGTTGA": 1889, + "TGTTGT": 1890, + "TGTTGC": 1891, + "TGTTGG": 1892, + "TGTCAA": 1893, + "TGTCAT": 1894, + "TGTCAC": 1895, + "TGTCAG": 1896, + "TGTCTA": 1897, + "TGTCTT": 1898, + "TGTCTC": 1899, + "TGTCTG": 1900, + "TGTCCA": 1901, + "TGTCCT": 1902, + "TGTCCC": 1903, + "TGTCCG": 1904, + "TGTCGA": 1905, + "TGTCGT": 1906, + "TGTCGC": 1907, + "TGTCGG": 1908, + "TGTGAA": 1909, + "TGTGAT": 1910, + "TGTGAC": 1911, + "TGTGAG": 1912, + "TGTGTA": 1913, + "TGTGTT": 1914, + "TGTGTC": 1915, + "TGTGTG": 1916, + "TGTGCA": 1917, + "TGTGCT": 1918, + "TGTGCC": 1919, + "TGTGCG": 1920, + "TGTGGA": 1921, + "TGTGGT": 1922, + "TGTGGC": 1923, + "TGTGGG": 1924, + "TGCAAA": 1925, + "TGCAAT": 1926, + "TGCAAC": 1927, + "TGCAAG": 1928, + "TGCATA": 1929, + "TGCATT": 1930, + "TGCATC": 1931, + "TGCATG": 1932, + "TGCACA": 1933, + "TGCACT": 1934, + "TGCACC": 1935, + "TGCACG": 1936, + "TGCAGA": 1937, + "TGCAGT": 1938, + "TGCAGC": 1939, + "TGCAGG": 1940, + "TGCTAA": 1941, + "TGCTAT": 1942, + "TGCTAC": 1943, + "TGCTAG": 1944, + "TGCTTA": 1945, + "TGCTTT": 1946, + "TGCTTC": 1947, + "TGCTTG": 1948, + "TGCTCA": 1949, + "TGCTCT": 1950, + "TGCTCC": 1951, + "TGCTCG": 1952, + "TGCTGA": 1953, + "TGCTGT": 1954, + "TGCTGC": 1955, + "TGCTGG": 1956, + "TGCCAA": 1957, + "TGCCAT": 1958, + "TGCCAC": 1959, + "TGCCAG": 1960, + "TGCCTA": 1961, + "TGCCTT": 1962, + "TGCCTC": 1963, + "TGCCTG": 1964, + "TGCCCA": 1965, + "TGCCCT": 1966, + "TGCCCC": 1967, + "TGCCCG": 1968, + "TGCCGA": 1969, + "TGCCGT": 1970, + "TGCCGC": 1971, + "TGCCGG": 1972, + "TGCGAA": 1973, + "TGCGAT": 1974, + "TGCGAC": 1975, + "TGCGAG": 1976, + "TGCGTA": 1977, + "TGCGTT": 1978, + "TGCGTC": 1979, + "TGCGTG": 1980, + "TGCGCA": 1981, + "TGCGCT": 1982, + "TGCGCC": 1983, + "TGCGCG": 1984, + "TGCGGA": 1985, + "TGCGGT": 1986, + "TGCGGC": 1987, + "TGCGGG": 1988, + "TGGAAA": 1989, + "TGGAAT": 1990, + "TGGAAC": 1991, + "TGGAAG": 1992, + "TGGATA": 1993, + "TGGATT": 1994, + "TGGATC": 1995, + "TGGATG": 1996, + "TGGACA": 1997, + "TGGACT": 1998, + "TGGACC": 1999, + "TGGACG": 2000, + "TGGAGA": 2001, + "TGGAGT": 2002, + "TGGAGC": 2003, + "TGGAGG": 2004, + "TGGTAA": 2005, + "TGGTAT": 2006, + "TGGTAC": 2007, + "TGGTAG": 2008, + "TGGTTA": 2009, + "TGGTTT": 2010, + "TGGTTC": 2011, + "TGGTTG": 2012, + "TGGTCA": 2013, + "TGGTCT": 2014, + "TGGTCC": 2015, + "TGGTCG": 2016, + "TGGTGA": 2017, + "TGGTGT": 2018, + "TGGTGC": 2019, + "TGGTGG": 2020, + "TGGCAA": 2021, + "TGGCAT": 2022, + "TGGCAC": 2023, + "TGGCAG": 2024, + "TGGCTA": 2025, + "TGGCTT": 2026, + "TGGCTC": 2027, + "TGGCTG": 2028, + "TGGCCA": 2029, + "TGGCCT": 2030, + "TGGCCC": 2031, + "TGGCCG": 2032, + "TGGCGA": 2033, + "TGGCGT": 2034, + "TGGCGC": 2035, + "TGGCGG": 2036, + "TGGGAA": 2037, + "TGGGAT": 2038, + "TGGGAC": 2039, + "TGGGAG": 2040, + "TGGGTA": 2041, + "TGGGTT": 2042, + "TGGGTC": 2043, + "TGGGTG": 2044, + "TGGGCA": 2045, + "TGGGCT": 2046, + "TGGGCC": 2047, + "TGGGCG": 2048, + "TGGGGA": 2049, + "TGGGGT": 2050, + "TGGGGC": 2051, + "TGGGGG": 2052, + "CAAAAA": 2053, + "CAAAAT": 2054, + "CAAAAC": 2055, + "CAAAAG": 2056, + "CAAATA": 2057, + "CAAATT": 2058, + "CAAATC": 2059, + "CAAATG": 2060, + "CAAACA": 2061, + "CAAACT": 2062, + "CAAACC": 2063, + "CAAACG": 2064, + "CAAAGA": 2065, + "CAAAGT": 2066, + "CAAAGC": 2067, + "CAAAGG": 2068, + "CAATAA": 2069, + "CAATAT": 2070, + "CAATAC": 2071, + "CAATAG": 2072, + "CAATTA": 2073, + "CAATTT": 2074, + "CAATTC": 2075, + "CAATTG": 2076, + "CAATCA": 2077, + "CAATCT": 2078, + "CAATCC": 2079, + "CAATCG": 2080, + "CAATGA": 2081, + "CAATGT": 2082, + "CAATGC": 2083, + "CAATGG": 2084, + "CAACAA": 2085, + "CAACAT": 2086, + "CAACAC": 2087, + "CAACAG": 2088, + "CAACTA": 2089, + "CAACTT": 2090, + "CAACTC": 2091, + "CAACTG": 2092, + "CAACCA": 2093, + "CAACCT": 2094, + "CAACCC": 2095, + "CAACCG": 2096, + "CAACGA": 2097, + "CAACGT": 2098, + "CAACGC": 2099, + "CAACGG": 2100, + "CAAGAA": 2101, + "CAAGAT": 2102, + "CAAGAC": 2103, + "CAAGAG": 2104, + "CAAGTA": 2105, + "CAAGTT": 2106, + "CAAGTC": 2107, + "CAAGTG": 2108, + "CAAGCA": 2109, + "CAAGCT": 2110, + "CAAGCC": 2111, + "CAAGCG": 2112, + "CAAGGA": 2113, + "CAAGGT": 2114, + "CAAGGC": 2115, + "CAAGGG": 2116, + "CATAAA": 2117, + "CATAAT": 2118, + "CATAAC": 2119, + "CATAAG": 2120, + "CATATA": 2121, + "CATATT": 2122, + "CATATC": 2123, + "CATATG": 2124, + "CATACA": 2125, + "CATACT": 2126, + "CATACC": 2127, + "CATACG": 2128, + "CATAGA": 2129, + "CATAGT": 2130, + "CATAGC": 2131, + "CATAGG": 2132, + "CATTAA": 2133, + "CATTAT": 2134, + "CATTAC": 2135, + "CATTAG": 2136, + "CATTTA": 2137, + "CATTTT": 2138, + "CATTTC": 2139, + "CATTTG": 2140, + "CATTCA": 2141, + "CATTCT": 2142, + "CATTCC": 2143, + "CATTCG": 2144, + "CATTGA": 2145, + "CATTGT": 2146, + "CATTGC": 2147, + "CATTGG": 2148, + "CATCAA": 2149, + "CATCAT": 2150, + "CATCAC": 2151, + "CATCAG": 2152, + "CATCTA": 2153, + "CATCTT": 2154, + "CATCTC": 2155, + "CATCTG": 2156, + "CATCCA": 2157, + "CATCCT": 2158, + "CATCCC": 2159, + "CATCCG": 2160, + "CATCGA": 2161, + "CATCGT": 2162, + "CATCGC": 2163, + "CATCGG": 2164, + "CATGAA": 2165, + "CATGAT": 2166, + "CATGAC": 2167, + "CATGAG": 2168, + "CATGTA": 2169, + "CATGTT": 2170, + "CATGTC": 2171, + "CATGTG": 2172, + "CATGCA": 2173, + "CATGCT": 2174, + "CATGCC": 2175, + "CATGCG": 2176, + "CATGGA": 2177, + "CATGGT": 2178, + "CATGGC": 2179, + "CATGGG": 2180, + "CACAAA": 2181, + "CACAAT": 2182, + "CACAAC": 2183, + "CACAAG": 2184, + "CACATA": 2185, + "CACATT": 2186, + "CACATC": 2187, + "CACATG": 2188, + "CACACA": 2189, + "CACACT": 2190, + "CACACC": 2191, + "CACACG": 2192, + "CACAGA": 2193, + "CACAGT": 2194, + "CACAGC": 2195, + "CACAGG": 2196, + "CACTAA": 2197, + "CACTAT": 2198, + "CACTAC": 2199, + "CACTAG": 2200, + "CACTTA": 2201, + "CACTTT": 2202, + "CACTTC": 2203, + "CACTTG": 2204, + "CACTCA": 2205, + "CACTCT": 2206, + "CACTCC": 2207, + "CACTCG": 2208, + "CACTGA": 2209, + "CACTGT": 2210, + "CACTGC": 2211, + "CACTGG": 2212, + "CACCAA": 2213, + "CACCAT": 2214, + "CACCAC": 2215, + "CACCAG": 2216, + "CACCTA": 2217, + "CACCTT": 2218, + "CACCTC": 2219, + "CACCTG": 2220, + "CACCCA": 2221, + "CACCCT": 2222, + "CACCCC": 2223, + "CACCCG": 2224, + "CACCGA": 2225, + "CACCGT": 2226, + "CACCGC": 2227, + "CACCGG": 2228, + "CACGAA": 2229, + "CACGAT": 2230, + "CACGAC": 2231, + "CACGAG": 2232, + "CACGTA": 2233, + "CACGTT": 2234, + "CACGTC": 2235, + "CACGTG": 2236, + "CACGCA": 2237, + "CACGCT": 2238, + "CACGCC": 2239, + "CACGCG": 2240, + "CACGGA": 2241, + "CACGGT": 2242, + "CACGGC": 2243, + "CACGGG": 2244, + "CAGAAA": 2245, + "CAGAAT": 2246, + "CAGAAC": 2247, + "CAGAAG": 2248, + "CAGATA": 2249, + "CAGATT": 2250, + "CAGATC": 2251, + "CAGATG": 2252, + "CAGACA": 2253, + "CAGACT": 2254, + "CAGACC": 2255, + "CAGACG": 2256, + "CAGAGA": 2257, + "CAGAGT": 2258, + "CAGAGC": 2259, + "CAGAGG": 2260, + "CAGTAA": 2261, + "CAGTAT": 2262, + "CAGTAC": 2263, + "CAGTAG": 2264, + "CAGTTA": 2265, + "CAGTTT": 2266, + "CAGTTC": 2267, + "CAGTTG": 2268, + "CAGTCA": 2269, + "CAGTCT": 2270, + "CAGTCC": 2271, + "CAGTCG": 2272, + "CAGTGA": 2273, + "CAGTGT": 2274, + "CAGTGC": 2275, + "CAGTGG": 2276, + "CAGCAA": 2277, + "CAGCAT": 2278, + "CAGCAC": 2279, + "CAGCAG": 2280, + "CAGCTA": 2281, + "CAGCTT": 2282, + "CAGCTC": 2283, + "CAGCTG": 2284, + "CAGCCA": 2285, + "CAGCCT": 2286, + "CAGCCC": 2287, + "CAGCCG": 2288, + "CAGCGA": 2289, + "CAGCGT": 2290, + "CAGCGC": 2291, + "CAGCGG": 2292, + "CAGGAA": 2293, + "CAGGAT": 2294, + "CAGGAC": 2295, + "CAGGAG": 2296, + "CAGGTA": 2297, + "CAGGTT": 2298, + "CAGGTC": 2299, + "CAGGTG": 2300, + "CAGGCA": 2301, + "CAGGCT": 2302, + "CAGGCC": 2303, + "CAGGCG": 2304, + "CAGGGA": 2305, + "CAGGGT": 2306, + "CAGGGC": 2307, + "CAGGGG": 2308, + "CTAAAA": 2309, + "CTAAAT": 2310, + "CTAAAC": 2311, + "CTAAAG": 2312, + "CTAATA": 2313, + "CTAATT": 2314, + "CTAATC": 2315, + "CTAATG": 2316, + "CTAACA": 2317, + "CTAACT": 2318, + "CTAACC": 2319, + "CTAACG": 2320, + "CTAAGA": 2321, + "CTAAGT": 2322, + "CTAAGC": 2323, + "CTAAGG": 2324, + "CTATAA": 2325, + "CTATAT": 2326, + "CTATAC": 2327, + "CTATAG": 2328, + "CTATTA": 2329, + "CTATTT": 2330, + "CTATTC": 2331, + "CTATTG": 2332, + "CTATCA": 2333, + "CTATCT": 2334, + "CTATCC": 2335, + "CTATCG": 2336, + "CTATGA": 2337, + "CTATGT": 2338, + "CTATGC": 2339, + "CTATGG": 2340, + "CTACAA": 2341, + "CTACAT": 2342, + "CTACAC": 2343, + "CTACAG": 2344, + "CTACTA": 2345, + "CTACTT": 2346, + "CTACTC": 2347, + "CTACTG": 2348, + "CTACCA": 2349, + "CTACCT": 2350, + "CTACCC": 2351, + "CTACCG": 2352, + "CTACGA": 2353, + "CTACGT": 2354, + "CTACGC": 2355, + "CTACGG": 2356, + "CTAGAA": 2357, + "CTAGAT": 2358, + "CTAGAC": 2359, + "CTAGAG": 2360, + "CTAGTA": 2361, + "CTAGTT": 2362, + "CTAGTC": 2363, + "CTAGTG": 2364, + "CTAGCA": 2365, + "CTAGCT": 2366, + "CTAGCC": 2367, + "CTAGCG": 2368, + "CTAGGA": 2369, + "CTAGGT": 2370, + "CTAGGC": 2371, + "CTAGGG": 2372, + "CTTAAA": 2373, + "CTTAAT": 2374, + "CTTAAC": 2375, + "CTTAAG": 2376, + "CTTATA": 2377, + "CTTATT": 2378, + "CTTATC": 2379, + "CTTATG": 2380, + "CTTACA": 2381, + "CTTACT": 2382, + "CTTACC": 2383, + "CTTACG": 2384, + "CTTAGA": 2385, + "CTTAGT": 2386, + "CTTAGC": 2387, + "CTTAGG": 2388, + "CTTTAA": 2389, + "CTTTAT": 2390, + "CTTTAC": 2391, + "CTTTAG": 2392, + "CTTTTA": 2393, + "CTTTTT": 2394, + "CTTTTC": 2395, + "CTTTTG": 2396, + "CTTTCA": 2397, + "CTTTCT": 2398, + "CTTTCC": 2399, + "CTTTCG": 2400, + "CTTTGA": 2401, + "CTTTGT": 2402, + "CTTTGC": 2403, + "CTTTGG": 2404, + "CTTCAA": 2405, + "CTTCAT": 2406, + "CTTCAC": 2407, + "CTTCAG": 2408, + "CTTCTA": 2409, + "CTTCTT": 2410, + "CTTCTC": 2411, + "CTTCTG": 2412, + "CTTCCA": 2413, + "CTTCCT": 2414, + "CTTCCC": 2415, + "CTTCCG": 2416, + "CTTCGA": 2417, + "CTTCGT": 2418, + "CTTCGC": 2419, + "CTTCGG": 2420, + "CTTGAA": 2421, + "CTTGAT": 2422, + "CTTGAC": 2423, + "CTTGAG": 2424, + "CTTGTA": 2425, + "CTTGTT": 2426, + "CTTGTC": 2427, + "CTTGTG": 2428, + "CTTGCA": 2429, + "CTTGCT": 2430, + "CTTGCC": 2431, + "CTTGCG": 2432, + "CTTGGA": 2433, + "CTTGGT": 2434, + "CTTGGC": 2435, + "CTTGGG": 2436, + "CTCAAA": 2437, + "CTCAAT": 2438, + "CTCAAC": 2439, + "CTCAAG": 2440, + "CTCATA": 2441, + "CTCATT": 2442, + "CTCATC": 2443, + "CTCATG": 2444, + "CTCACA": 2445, + "CTCACT": 2446, + "CTCACC": 2447, + "CTCACG": 2448, + "CTCAGA": 2449, + "CTCAGT": 2450, + "CTCAGC": 2451, + "CTCAGG": 2452, + "CTCTAA": 2453, + "CTCTAT": 2454, + "CTCTAC": 2455, + "CTCTAG": 2456, + "CTCTTA": 2457, + "CTCTTT": 2458, + "CTCTTC": 2459, + "CTCTTG": 2460, + "CTCTCA": 2461, + "CTCTCT": 2462, + "CTCTCC": 2463, + "CTCTCG": 2464, + "CTCTGA": 2465, + "CTCTGT": 2466, + "CTCTGC": 2467, + "CTCTGG": 2468, + "CTCCAA": 2469, + "CTCCAT": 2470, + "CTCCAC": 2471, + "CTCCAG": 2472, + "CTCCTA": 2473, + "CTCCTT": 2474, + "CTCCTC": 2475, + "CTCCTG": 2476, + "CTCCCA": 2477, + "CTCCCT": 2478, + "CTCCCC": 2479, + "CTCCCG": 2480, + "CTCCGA": 2481, + "CTCCGT": 2482, + "CTCCGC": 2483, + "CTCCGG": 2484, + "CTCGAA": 2485, + "CTCGAT": 2486, + "CTCGAC": 2487, + "CTCGAG": 2488, + "CTCGTA": 2489, + "CTCGTT": 2490, + "CTCGTC": 2491, + "CTCGTG": 2492, + "CTCGCA": 2493, + "CTCGCT": 2494, + "CTCGCC": 2495, + "CTCGCG": 2496, + "CTCGGA": 2497, + "CTCGGT": 2498, + "CTCGGC": 2499, + "CTCGGG": 2500, + "CTGAAA": 2501, + "CTGAAT": 2502, + "CTGAAC": 2503, + "CTGAAG": 2504, + "CTGATA": 2505, + "CTGATT": 2506, + "CTGATC": 2507, + "CTGATG": 2508, + "CTGACA": 2509, + "CTGACT": 2510, + "CTGACC": 2511, + "CTGACG": 2512, + "CTGAGA": 2513, + "CTGAGT": 2514, + "CTGAGC": 2515, + "CTGAGG": 2516, + "CTGTAA": 2517, + "CTGTAT": 2518, + "CTGTAC": 2519, + "CTGTAG": 2520, + "CTGTTA": 2521, + "CTGTTT": 2522, + "CTGTTC": 2523, + "CTGTTG": 2524, + "CTGTCA": 2525, + "CTGTCT": 2526, + "CTGTCC": 2527, + "CTGTCG": 2528, + "CTGTGA": 2529, + "CTGTGT": 2530, + "CTGTGC": 2531, + "CTGTGG": 2532, + "CTGCAA": 2533, + "CTGCAT": 2534, + "CTGCAC": 2535, + "CTGCAG": 2536, + "CTGCTA": 2537, + "CTGCTT": 2538, + "CTGCTC": 2539, + "CTGCTG": 2540, + "CTGCCA": 2541, + "CTGCCT": 2542, + "CTGCCC": 2543, + "CTGCCG": 2544, + "CTGCGA": 2545, + "CTGCGT": 2546, + "CTGCGC": 2547, + "CTGCGG": 2548, + "CTGGAA": 2549, + "CTGGAT": 2550, + "CTGGAC": 2551, + "CTGGAG": 2552, + "CTGGTA": 2553, + "CTGGTT": 2554, + "CTGGTC": 2555, + "CTGGTG": 2556, + "CTGGCA": 2557, + "CTGGCT": 2558, + "CTGGCC": 2559, + "CTGGCG": 2560, + "CTGGGA": 2561, + "CTGGGT": 2562, + "CTGGGC": 2563, + "CTGGGG": 2564, + "CCAAAA": 2565, + "CCAAAT": 2566, + "CCAAAC": 2567, + "CCAAAG": 2568, + "CCAATA": 2569, + "CCAATT": 2570, + "CCAATC": 2571, + "CCAATG": 2572, + "CCAACA": 2573, + "CCAACT": 2574, + "CCAACC": 2575, + "CCAACG": 2576, + "CCAAGA": 2577, + "CCAAGT": 2578, + "CCAAGC": 2579, + "CCAAGG": 2580, + "CCATAA": 2581, + "CCATAT": 2582, + "CCATAC": 2583, + "CCATAG": 2584, + "CCATTA": 2585, + "CCATTT": 2586, + "CCATTC": 2587, + "CCATTG": 2588, + "CCATCA": 2589, + "CCATCT": 2590, + "CCATCC": 2591, + "CCATCG": 2592, + "CCATGA": 2593, + "CCATGT": 2594, + "CCATGC": 2595, + "CCATGG": 2596, + "CCACAA": 2597, + "CCACAT": 2598, + "CCACAC": 2599, + "CCACAG": 2600, + "CCACTA": 2601, + "CCACTT": 2602, + "CCACTC": 2603, + "CCACTG": 2604, + "CCACCA": 2605, + "CCACCT": 2606, + "CCACCC": 2607, + "CCACCG": 2608, + "CCACGA": 2609, + "CCACGT": 2610, + "CCACGC": 2611, + "CCACGG": 2612, + "CCAGAA": 2613, + "CCAGAT": 2614, + "CCAGAC": 2615, + "CCAGAG": 2616, + "CCAGTA": 2617, + "CCAGTT": 2618, + "CCAGTC": 2619, + "CCAGTG": 2620, + "CCAGCA": 2621, + "CCAGCT": 2622, + "CCAGCC": 2623, + "CCAGCG": 2624, + "CCAGGA": 2625, + "CCAGGT": 2626, + "CCAGGC": 2627, + "CCAGGG": 2628, + "CCTAAA": 2629, + "CCTAAT": 2630, + "CCTAAC": 2631, + "CCTAAG": 2632, + "CCTATA": 2633, + "CCTATT": 2634, + "CCTATC": 2635, + "CCTATG": 2636, + "CCTACA": 2637, + "CCTACT": 2638, + "CCTACC": 2639, + "CCTACG": 2640, + "CCTAGA": 2641, + "CCTAGT": 2642, + "CCTAGC": 2643, + "CCTAGG": 2644, + "CCTTAA": 2645, + "CCTTAT": 2646, + "CCTTAC": 2647, + "CCTTAG": 2648, + "CCTTTA": 2649, + "CCTTTT": 2650, + "CCTTTC": 2651, + "CCTTTG": 2652, + "CCTTCA": 2653, + "CCTTCT": 2654, + "CCTTCC": 2655, + "CCTTCG": 2656, + "CCTTGA": 2657, + "CCTTGT": 2658, + "CCTTGC": 2659, + "CCTTGG": 2660, + "CCTCAA": 2661, + "CCTCAT": 2662, + "CCTCAC": 2663, + "CCTCAG": 2664, + "CCTCTA": 2665, + "CCTCTT": 2666, + "CCTCTC": 2667, + "CCTCTG": 2668, + "CCTCCA": 2669, + "CCTCCT": 2670, + "CCTCCC": 2671, + "CCTCCG": 2672, + "CCTCGA": 2673, + "CCTCGT": 2674, + "CCTCGC": 2675, + "CCTCGG": 2676, + "CCTGAA": 2677, + "CCTGAT": 2678, + "CCTGAC": 2679, + "CCTGAG": 2680, + "CCTGTA": 2681, + "CCTGTT": 2682, + "CCTGTC": 2683, + "CCTGTG": 2684, + "CCTGCA": 2685, + "CCTGCT": 2686, + "CCTGCC": 2687, + "CCTGCG": 2688, + "CCTGGA": 2689, + "CCTGGT": 2690, + "CCTGGC": 2691, + "CCTGGG": 2692, + "CCCAAA": 2693, + "CCCAAT": 2694, + "CCCAAC": 2695, + "CCCAAG": 2696, + "CCCATA": 2697, + "CCCATT": 2698, + "CCCATC": 2699, + "CCCATG": 2700, + "CCCACA": 2701, + "CCCACT": 2702, + "CCCACC": 2703, + "CCCACG": 2704, + "CCCAGA": 2705, + "CCCAGT": 2706, + "CCCAGC": 2707, + "CCCAGG": 2708, + "CCCTAA": 2709, + "CCCTAT": 2710, + "CCCTAC": 2711, + "CCCTAG": 2712, + "CCCTTA": 2713, + "CCCTTT": 2714, + "CCCTTC": 2715, + "CCCTTG": 2716, + "CCCTCA": 2717, + "CCCTCT": 2718, + "CCCTCC": 2719, + "CCCTCG": 2720, + "CCCTGA": 2721, + "CCCTGT": 2722, + "CCCTGC": 2723, + "CCCTGG": 2724, + "CCCCAA": 2725, + "CCCCAT": 2726, + "CCCCAC": 2727, + "CCCCAG": 2728, + "CCCCTA": 2729, + "CCCCTT": 2730, + "CCCCTC": 2731, + "CCCCTG": 2732, + "CCCCCA": 2733, + "CCCCCT": 2734, + "CCCCCC": 2735, + "CCCCCG": 2736, + "CCCCGA": 2737, + "CCCCGT": 2738, + "CCCCGC": 2739, + "CCCCGG": 2740, + "CCCGAA": 2741, + "CCCGAT": 2742, + "CCCGAC": 2743, + "CCCGAG": 2744, + "CCCGTA": 2745, + "CCCGTT": 2746, + "CCCGTC": 2747, + "CCCGTG": 2748, + "CCCGCA": 2749, + "CCCGCT": 2750, + "CCCGCC": 2751, + "CCCGCG": 2752, + "CCCGGA": 2753, + "CCCGGT": 2754, + "CCCGGC": 2755, + "CCCGGG": 2756, + "CCGAAA": 2757, + "CCGAAT": 2758, + "CCGAAC": 2759, + "CCGAAG": 2760, + "CCGATA": 2761, + "CCGATT": 2762, + "CCGATC": 2763, + "CCGATG": 2764, + "CCGACA": 2765, + "CCGACT": 2766, + "CCGACC": 2767, + "CCGACG": 2768, + "CCGAGA": 2769, + "CCGAGT": 2770, + "CCGAGC": 2771, + "CCGAGG": 2772, + "CCGTAA": 2773, + "CCGTAT": 2774, + "CCGTAC": 2775, + "CCGTAG": 2776, + "CCGTTA": 2777, + "CCGTTT": 2778, + "CCGTTC": 2779, + "CCGTTG": 2780, + "CCGTCA": 2781, + "CCGTCT": 2782, + "CCGTCC": 2783, + "CCGTCG": 2784, + "CCGTGA": 2785, + "CCGTGT": 2786, + "CCGTGC": 2787, + "CCGTGG": 2788, + "CCGCAA": 2789, + "CCGCAT": 2790, + "CCGCAC": 2791, + "CCGCAG": 2792, + "CCGCTA": 2793, + "CCGCTT": 2794, + "CCGCTC": 2795, + "CCGCTG": 2796, + "CCGCCA": 2797, + "CCGCCT": 2798, + "CCGCCC": 2799, + "CCGCCG": 2800, + "CCGCGA": 2801, + "CCGCGT": 2802, + "CCGCGC": 2803, + "CCGCGG": 2804, + "CCGGAA": 2805, + "CCGGAT": 2806, + "CCGGAC": 2807, + "CCGGAG": 2808, + "CCGGTA": 2809, + "CCGGTT": 2810, + "CCGGTC": 2811, + "CCGGTG": 2812, + "CCGGCA": 2813, + "CCGGCT": 2814, + "CCGGCC": 2815, + "CCGGCG": 2816, + "CCGGGA": 2817, + "CCGGGT": 2818, + "CCGGGC": 2819, + "CCGGGG": 2820, + "CGAAAA": 2821, + "CGAAAT": 2822, + "CGAAAC": 2823, + "CGAAAG": 2824, + "CGAATA": 2825, + "CGAATT": 2826, + "CGAATC": 2827, + "CGAATG": 2828, + "CGAACA": 2829, + "CGAACT": 2830, + "CGAACC": 2831, + "CGAACG": 2832, + "CGAAGA": 2833, + "CGAAGT": 2834, + "CGAAGC": 2835, + "CGAAGG": 2836, + "CGATAA": 2837, + "CGATAT": 2838, + "CGATAC": 2839, + "CGATAG": 2840, + "CGATTA": 2841, + "CGATTT": 2842, + "CGATTC": 2843, + "CGATTG": 2844, + "CGATCA": 2845, + "CGATCT": 2846, + "CGATCC": 2847, + "CGATCG": 2848, + "CGATGA": 2849, + "CGATGT": 2850, + "CGATGC": 2851, + "CGATGG": 2852, + "CGACAA": 2853, + "CGACAT": 2854, + "CGACAC": 2855, + "CGACAG": 2856, + "CGACTA": 2857, + "CGACTT": 2858, + "CGACTC": 2859, + "CGACTG": 2860, + "CGACCA": 2861, + "CGACCT": 2862, + "CGACCC": 2863, + "CGACCG": 2864, + "CGACGA": 2865, + "CGACGT": 2866, + "CGACGC": 2867, + "CGACGG": 2868, + "CGAGAA": 2869, + "CGAGAT": 2870, + "CGAGAC": 2871, + "CGAGAG": 2872, + "CGAGTA": 2873, + "CGAGTT": 2874, + "CGAGTC": 2875, + "CGAGTG": 2876, + "CGAGCA": 2877, + "CGAGCT": 2878, + "CGAGCC": 2879, + "CGAGCG": 2880, + "CGAGGA": 2881, + "CGAGGT": 2882, + "CGAGGC": 2883, + "CGAGGG": 2884, + "CGTAAA": 2885, + "CGTAAT": 2886, + "CGTAAC": 2887, + "CGTAAG": 2888, + "CGTATA": 2889, + "CGTATT": 2890, + "CGTATC": 2891, + "CGTATG": 2892, + "CGTACA": 2893, + "CGTACT": 2894, + "CGTACC": 2895, + "CGTACG": 2896, + "CGTAGA": 2897, + "CGTAGT": 2898, + "CGTAGC": 2899, + "CGTAGG": 2900, + "CGTTAA": 2901, + "CGTTAT": 2902, + "CGTTAC": 2903, + "CGTTAG": 2904, + "CGTTTA": 2905, + "CGTTTT": 2906, + "CGTTTC": 2907, + "CGTTTG": 2908, + "CGTTCA": 2909, + "CGTTCT": 2910, + "CGTTCC": 2911, + "CGTTCG": 2912, + "CGTTGA": 2913, + "CGTTGT": 2914, + "CGTTGC": 2915, + "CGTTGG": 2916, + "CGTCAA": 2917, + "CGTCAT": 2918, + "CGTCAC": 2919, + "CGTCAG": 2920, + "CGTCTA": 2921, + "CGTCTT": 2922, + "CGTCTC": 2923, + "CGTCTG": 2924, + "CGTCCA": 2925, + "CGTCCT": 2926, + "CGTCCC": 2927, + "CGTCCG": 2928, + "CGTCGA": 2929, + "CGTCGT": 2930, + "CGTCGC": 2931, + "CGTCGG": 2932, + "CGTGAA": 2933, + "CGTGAT": 2934, + "CGTGAC": 2935, + "CGTGAG": 2936, + "CGTGTA": 2937, + "CGTGTT": 2938, + "CGTGTC": 2939, + "CGTGTG": 2940, + "CGTGCA": 2941, + "CGTGCT": 2942, + "CGTGCC": 2943, + "CGTGCG": 2944, + "CGTGGA": 2945, + "CGTGGT": 2946, + "CGTGGC": 2947, + "CGTGGG": 2948, + "CGCAAA": 2949, + "CGCAAT": 2950, + "CGCAAC": 2951, + "CGCAAG": 2952, + "CGCATA": 2953, + "CGCATT": 2954, + "CGCATC": 2955, + "CGCATG": 2956, + "CGCACA": 2957, + "CGCACT": 2958, + "CGCACC": 2959, + "CGCACG": 2960, + "CGCAGA": 2961, + "CGCAGT": 2962, + "CGCAGC": 2963, + "CGCAGG": 2964, + "CGCTAA": 2965, + "CGCTAT": 2966, + "CGCTAC": 2967, + "CGCTAG": 2968, + "CGCTTA": 2969, + "CGCTTT": 2970, + "CGCTTC": 2971, + "CGCTTG": 2972, + "CGCTCA": 2973, + "CGCTCT": 2974, + "CGCTCC": 2975, + "CGCTCG": 2976, + "CGCTGA": 2977, + "CGCTGT": 2978, + "CGCTGC": 2979, + "CGCTGG": 2980, + "CGCCAA": 2981, + "CGCCAT": 2982, + "CGCCAC": 2983, + "CGCCAG": 2984, + "CGCCTA": 2985, + "CGCCTT": 2986, + "CGCCTC": 2987, + "CGCCTG": 2988, + "CGCCCA": 2989, + "CGCCCT": 2990, + "CGCCCC": 2991, + "CGCCCG": 2992, + "CGCCGA": 2993, + "CGCCGT": 2994, + "CGCCGC": 2995, + "CGCCGG": 2996, + "CGCGAA": 2997, + "CGCGAT": 2998, + "CGCGAC": 2999, + "CGCGAG": 3000, + "CGCGTA": 3001, + "CGCGTT": 3002, + "CGCGTC": 3003, + "CGCGTG": 3004, + "CGCGCA": 3005, + "CGCGCT": 3006, + "CGCGCC": 3007, + "CGCGCG": 3008, + "CGCGGA": 3009, + "CGCGGT": 3010, + "CGCGGC": 3011, + "CGCGGG": 3012, + "CGGAAA": 3013, + "CGGAAT": 3014, + "CGGAAC": 3015, + "CGGAAG": 3016, + "CGGATA": 3017, + "CGGATT": 3018, + "CGGATC": 3019, + "CGGATG": 3020, + "CGGACA": 3021, + "CGGACT": 3022, + "CGGACC": 3023, + "CGGACG": 3024, + "CGGAGA": 3025, + "CGGAGT": 3026, + "CGGAGC": 3027, + "CGGAGG": 3028, + "CGGTAA": 3029, + "CGGTAT": 3030, + "CGGTAC": 3031, + "CGGTAG": 3032, + "CGGTTA": 3033, + "CGGTTT": 3034, + "CGGTTC": 3035, + "CGGTTG": 3036, + "CGGTCA": 3037, + "CGGTCT": 3038, + "CGGTCC": 3039, + "CGGTCG": 3040, + "CGGTGA": 3041, + "CGGTGT": 3042, + "CGGTGC": 3043, + "CGGTGG": 3044, + "CGGCAA": 3045, + "CGGCAT": 3046, + "CGGCAC": 3047, + "CGGCAG": 3048, + "CGGCTA": 3049, + "CGGCTT": 3050, + "CGGCTC": 3051, + "CGGCTG": 3052, + "CGGCCA": 3053, + "CGGCCT": 3054, + "CGGCCC": 3055, + "CGGCCG": 3056, + "CGGCGA": 3057, + "CGGCGT": 3058, + "CGGCGC": 3059, + "CGGCGG": 3060, + "CGGGAA": 3061, + "CGGGAT": 3062, + "CGGGAC": 3063, + "CGGGAG": 3064, + "CGGGTA": 3065, + "CGGGTT": 3066, + "CGGGTC": 3067, + "CGGGTG": 3068, + "CGGGCA": 3069, + "CGGGCT": 3070, + "CGGGCC": 3071, + "CGGGCG": 3072, + "CGGGGA": 3073, + "CGGGGT": 3074, + "CGGGGC": 3075, + "CGGGGG": 3076, + "GAAAAA": 3077, + "GAAAAT": 3078, + "GAAAAC": 3079, + "GAAAAG": 3080, + "GAAATA": 3081, + "GAAATT": 3082, + "GAAATC": 3083, + "GAAATG": 3084, + "GAAACA": 3085, + "GAAACT": 3086, + "GAAACC": 3087, + "GAAACG": 3088, + "GAAAGA": 3089, + "GAAAGT": 3090, + "GAAAGC": 3091, + "GAAAGG": 3092, + "GAATAA": 3093, + "GAATAT": 3094, + "GAATAC": 3095, + "GAATAG": 3096, + "GAATTA": 3097, + "GAATTT": 3098, + "GAATTC": 3099, + "GAATTG": 3100, + "GAATCA": 3101, + "GAATCT": 3102, + "GAATCC": 3103, + "GAATCG": 3104, + "GAATGA": 3105, + "GAATGT": 3106, + "GAATGC": 3107, + "GAATGG": 3108, + "GAACAA": 3109, + "GAACAT": 3110, + "GAACAC": 3111, + "GAACAG": 3112, + "GAACTA": 3113, + "GAACTT": 3114, + "GAACTC": 3115, + "GAACTG": 3116, + "GAACCA": 3117, + "GAACCT": 3118, + "GAACCC": 3119, + "GAACCG": 3120, + "GAACGA": 3121, + "GAACGT": 3122, + "GAACGC": 3123, + "GAACGG": 3124, + "GAAGAA": 3125, + "GAAGAT": 3126, + "GAAGAC": 3127, + "GAAGAG": 3128, + "GAAGTA": 3129, + "GAAGTT": 3130, + "GAAGTC": 3131, + "GAAGTG": 3132, + "GAAGCA": 3133, + "GAAGCT": 3134, + "GAAGCC": 3135, + "GAAGCG": 3136, + "GAAGGA": 3137, + "GAAGGT": 3138, + "GAAGGC": 3139, + "GAAGGG": 3140, + "GATAAA": 3141, + "GATAAT": 3142, + "GATAAC": 3143, + "GATAAG": 3144, + "GATATA": 3145, + "GATATT": 3146, + "GATATC": 3147, + "GATATG": 3148, + "GATACA": 3149, + "GATACT": 3150, + "GATACC": 3151, + "GATACG": 3152, + "GATAGA": 3153, + "GATAGT": 3154, + "GATAGC": 3155, + "GATAGG": 3156, + "GATTAA": 3157, + "GATTAT": 3158, + "GATTAC": 3159, + "GATTAG": 3160, + "GATTTA": 3161, + "GATTTT": 3162, + "GATTTC": 3163, + "GATTTG": 3164, + "GATTCA": 3165, + "GATTCT": 3166, + "GATTCC": 3167, + "GATTCG": 3168, + "GATTGA": 3169, + "GATTGT": 3170, + "GATTGC": 3171, + "GATTGG": 3172, + "GATCAA": 3173, + "GATCAT": 3174, + "GATCAC": 3175, + "GATCAG": 3176, + "GATCTA": 3177, + "GATCTT": 3178, + "GATCTC": 3179, + "GATCTG": 3180, + "GATCCA": 3181, + "GATCCT": 3182, + "GATCCC": 3183, + "GATCCG": 3184, + "GATCGA": 3185, + "GATCGT": 3186, + "GATCGC": 3187, + "GATCGG": 3188, + "GATGAA": 3189, + "GATGAT": 3190, + "GATGAC": 3191, + "GATGAG": 3192, + "GATGTA": 3193, + "GATGTT": 3194, + "GATGTC": 3195, + "GATGTG": 3196, + "GATGCA": 3197, + "GATGCT": 3198, + "GATGCC": 3199, + "GATGCG": 3200, + "GATGGA": 3201, + "GATGGT": 3202, + "GATGGC": 3203, + "GATGGG": 3204, + "GACAAA": 3205, + "GACAAT": 3206, + "GACAAC": 3207, + "GACAAG": 3208, + "GACATA": 3209, + "GACATT": 3210, + "GACATC": 3211, + "GACATG": 3212, + "GACACA": 3213, + "GACACT": 3214, + "GACACC": 3215, + "GACACG": 3216, + "GACAGA": 3217, + "GACAGT": 3218, + "GACAGC": 3219, + "GACAGG": 3220, + "GACTAA": 3221, + "GACTAT": 3222, + "GACTAC": 3223, + "GACTAG": 3224, + "GACTTA": 3225, + "GACTTT": 3226, + "GACTTC": 3227, + "GACTTG": 3228, + "GACTCA": 3229, + "GACTCT": 3230, + "GACTCC": 3231, + "GACTCG": 3232, + "GACTGA": 3233, + "GACTGT": 3234, + "GACTGC": 3235, + "GACTGG": 3236, + "GACCAA": 3237, + "GACCAT": 3238, + "GACCAC": 3239, + "GACCAG": 3240, + "GACCTA": 3241, + "GACCTT": 3242, + "GACCTC": 3243, + "GACCTG": 3244, + "GACCCA": 3245, + "GACCCT": 3246, + "GACCCC": 3247, + "GACCCG": 3248, + "GACCGA": 3249, + "GACCGT": 3250, + "GACCGC": 3251, + "GACCGG": 3252, + "GACGAA": 3253, + "GACGAT": 3254, + "GACGAC": 3255, + "GACGAG": 3256, + "GACGTA": 3257, + "GACGTT": 3258, + "GACGTC": 3259, + "GACGTG": 3260, + "GACGCA": 3261, + "GACGCT": 3262, + "GACGCC": 3263, + "GACGCG": 3264, + "GACGGA": 3265, + "GACGGT": 3266, + "GACGGC": 3267, + "GACGGG": 3268, + "GAGAAA": 3269, + "GAGAAT": 3270, + "GAGAAC": 3271, + "GAGAAG": 3272, + "GAGATA": 3273, + "GAGATT": 3274, + "GAGATC": 3275, + "GAGATG": 3276, + "GAGACA": 3277, + "GAGACT": 3278, + "GAGACC": 3279, + "GAGACG": 3280, + "GAGAGA": 3281, + "GAGAGT": 3282, + "GAGAGC": 3283, + "GAGAGG": 3284, + "GAGTAA": 3285, + "GAGTAT": 3286, + "GAGTAC": 3287, + "GAGTAG": 3288, + "GAGTTA": 3289, + "GAGTTT": 3290, + "GAGTTC": 3291, + "GAGTTG": 3292, + "GAGTCA": 3293, + "GAGTCT": 3294, + "GAGTCC": 3295, + "GAGTCG": 3296, + "GAGTGA": 3297, + "GAGTGT": 3298, + "GAGTGC": 3299, + "GAGTGG": 3300, + "GAGCAA": 3301, + "GAGCAT": 3302, + "GAGCAC": 3303, + "GAGCAG": 3304, + "GAGCTA": 3305, + "GAGCTT": 3306, + "GAGCTC": 3307, + "GAGCTG": 3308, + "GAGCCA": 3309, + "GAGCCT": 3310, + "GAGCCC": 3311, + "GAGCCG": 3312, + "GAGCGA": 3313, + "GAGCGT": 3314, + "GAGCGC": 3315, + "GAGCGG": 3316, + "GAGGAA": 3317, + "GAGGAT": 3318, + "GAGGAC": 3319, + "GAGGAG": 3320, + "GAGGTA": 3321, + "GAGGTT": 3322, + "GAGGTC": 3323, + "GAGGTG": 3324, + "GAGGCA": 3325, + "GAGGCT": 3326, + "GAGGCC": 3327, + "GAGGCG": 3328, + "GAGGGA": 3329, + "GAGGGT": 3330, + "GAGGGC": 3331, + "GAGGGG": 3332, + "GTAAAA": 3333, + "GTAAAT": 3334, + "GTAAAC": 3335, + "GTAAAG": 3336, + "GTAATA": 3337, + "GTAATT": 3338, + "GTAATC": 3339, + "GTAATG": 3340, + "GTAACA": 3341, + "GTAACT": 3342, + "GTAACC": 3343, + "GTAACG": 3344, + "GTAAGA": 3345, + "GTAAGT": 3346, + "GTAAGC": 3347, + "GTAAGG": 3348, + "GTATAA": 3349, + "GTATAT": 3350, + "GTATAC": 3351, + "GTATAG": 3352, + "GTATTA": 3353, + "GTATTT": 3354, + "GTATTC": 3355, + "GTATTG": 3356, + "GTATCA": 3357, + "GTATCT": 3358, + "GTATCC": 3359, + "GTATCG": 3360, + "GTATGA": 3361, + "GTATGT": 3362, + "GTATGC": 3363, + "GTATGG": 3364, + "GTACAA": 3365, + "GTACAT": 3366, + "GTACAC": 3367, + "GTACAG": 3368, + "GTACTA": 3369, + "GTACTT": 3370, + "GTACTC": 3371, + "GTACTG": 3372, + "GTACCA": 3373, + "GTACCT": 3374, + "GTACCC": 3375, + "GTACCG": 3376, + "GTACGA": 3377, + "GTACGT": 3378, + "GTACGC": 3379, + "GTACGG": 3380, + "GTAGAA": 3381, + "GTAGAT": 3382, + "GTAGAC": 3383, + "GTAGAG": 3384, + "GTAGTA": 3385, + "GTAGTT": 3386, + "GTAGTC": 3387, + "GTAGTG": 3388, + "GTAGCA": 3389, + "GTAGCT": 3390, + "GTAGCC": 3391, + "GTAGCG": 3392, + "GTAGGA": 3393, + "GTAGGT": 3394, + "GTAGGC": 3395, + "GTAGGG": 3396, + "GTTAAA": 3397, + "GTTAAT": 3398, + "GTTAAC": 3399, + "GTTAAG": 3400, + "GTTATA": 3401, + "GTTATT": 3402, + "GTTATC": 3403, + "GTTATG": 3404, + "GTTACA": 3405, + "GTTACT": 3406, + "GTTACC": 3407, + "GTTACG": 3408, + "GTTAGA": 3409, + "GTTAGT": 3410, + "GTTAGC": 3411, + "GTTAGG": 3412, + "GTTTAA": 3413, + "GTTTAT": 3414, + "GTTTAC": 3415, + "GTTTAG": 3416, + "GTTTTA": 3417, + "GTTTTT": 3418, + "GTTTTC": 3419, + "GTTTTG": 3420, + "GTTTCA": 3421, + "GTTTCT": 3422, + "GTTTCC": 3423, + "GTTTCG": 3424, + "GTTTGA": 3425, + "GTTTGT": 3426, + "GTTTGC": 3427, + "GTTTGG": 3428, + "GTTCAA": 3429, + "GTTCAT": 3430, + "GTTCAC": 3431, + "GTTCAG": 3432, + "GTTCTA": 3433, + "GTTCTT": 3434, + "GTTCTC": 3435, + "GTTCTG": 3436, + "GTTCCA": 3437, + "GTTCCT": 3438, + "GTTCCC": 3439, + "GTTCCG": 3440, + "GTTCGA": 3441, + "GTTCGT": 3442, + "GTTCGC": 3443, + "GTTCGG": 3444, + "GTTGAA": 3445, + "GTTGAT": 3446, + "GTTGAC": 3447, + "GTTGAG": 3448, + "GTTGTA": 3449, + "GTTGTT": 3450, + "GTTGTC": 3451, + "GTTGTG": 3452, + "GTTGCA": 3453, + "GTTGCT": 3454, + "GTTGCC": 3455, + "GTTGCG": 3456, + "GTTGGA": 3457, + "GTTGGT": 3458, + "GTTGGC": 3459, + "GTTGGG": 3460, + "GTCAAA": 3461, + "GTCAAT": 3462, + "GTCAAC": 3463, + "GTCAAG": 3464, + "GTCATA": 3465, + "GTCATT": 3466, + "GTCATC": 3467, + "GTCATG": 3468, + "GTCACA": 3469, + "GTCACT": 3470, + "GTCACC": 3471, + "GTCACG": 3472, + "GTCAGA": 3473, + "GTCAGT": 3474, + "GTCAGC": 3475, + "GTCAGG": 3476, + "GTCTAA": 3477, + "GTCTAT": 3478, + "GTCTAC": 3479, + "GTCTAG": 3480, + "GTCTTA": 3481, + "GTCTTT": 3482, + "GTCTTC": 3483, + "GTCTTG": 3484, + "GTCTCA": 3485, + "GTCTCT": 3486, + "GTCTCC": 3487, + "GTCTCG": 3488, + "GTCTGA": 3489, + "GTCTGT": 3490, + "GTCTGC": 3491, + "GTCTGG": 3492, + "GTCCAA": 3493, + "GTCCAT": 3494, + "GTCCAC": 3495, + "GTCCAG": 3496, + "GTCCTA": 3497, + "GTCCTT": 3498, + "GTCCTC": 3499, + "GTCCTG": 3500, + "GTCCCA": 3501, + "GTCCCT": 3502, + "GTCCCC": 3503, + "GTCCCG": 3504, + "GTCCGA": 3505, + "GTCCGT": 3506, + "GTCCGC": 3507, + "GTCCGG": 3508, + "GTCGAA": 3509, + "GTCGAT": 3510, + "GTCGAC": 3511, + "GTCGAG": 3512, + "GTCGTA": 3513, + "GTCGTT": 3514, + "GTCGTC": 3515, + "GTCGTG": 3516, + "GTCGCA": 3517, + "GTCGCT": 3518, + "GTCGCC": 3519, + "GTCGCG": 3520, + "GTCGGA": 3521, + "GTCGGT": 3522, + "GTCGGC": 3523, + "GTCGGG": 3524, + "GTGAAA": 3525, + "GTGAAT": 3526, + "GTGAAC": 3527, + "GTGAAG": 3528, + "GTGATA": 3529, + "GTGATT": 3530, + "GTGATC": 3531, + "GTGATG": 3532, + "GTGACA": 3533, + "GTGACT": 3534, + "GTGACC": 3535, + "GTGACG": 3536, + "GTGAGA": 3537, + "GTGAGT": 3538, + "GTGAGC": 3539, + "GTGAGG": 3540, + "GTGTAA": 3541, + "GTGTAT": 3542, + "GTGTAC": 3543, + "GTGTAG": 3544, + "GTGTTA": 3545, + "GTGTTT": 3546, + "GTGTTC": 3547, + "GTGTTG": 3548, + "GTGTCA": 3549, + "GTGTCT": 3550, + "GTGTCC": 3551, + "GTGTCG": 3552, + "GTGTGA": 3553, + "GTGTGT": 3554, + "GTGTGC": 3555, + "GTGTGG": 3556, + "GTGCAA": 3557, + "GTGCAT": 3558, + "GTGCAC": 3559, + "GTGCAG": 3560, + "GTGCTA": 3561, + "GTGCTT": 3562, + "GTGCTC": 3563, + "GTGCTG": 3564, + "GTGCCA": 3565, + "GTGCCT": 3566, + "GTGCCC": 3567, + "GTGCCG": 3568, + "GTGCGA": 3569, + "GTGCGT": 3570, + "GTGCGC": 3571, + "GTGCGG": 3572, + "GTGGAA": 3573, + "GTGGAT": 3574, + "GTGGAC": 3575, + "GTGGAG": 3576, + "GTGGTA": 3577, + "GTGGTT": 3578, + "GTGGTC": 3579, + "GTGGTG": 3580, + "GTGGCA": 3581, + "GTGGCT": 3582, + "GTGGCC": 3583, + "GTGGCG": 3584, + "GTGGGA": 3585, + "GTGGGT": 3586, + "GTGGGC": 3587, + "GTGGGG": 3588, + "GCAAAA": 3589, + "GCAAAT": 3590, + "GCAAAC": 3591, + "GCAAAG": 3592, + "GCAATA": 3593, + "GCAATT": 3594, + "GCAATC": 3595, + "GCAATG": 3596, + "GCAACA": 3597, + "GCAACT": 3598, + "GCAACC": 3599, + "GCAACG": 3600, + "GCAAGA": 3601, + "GCAAGT": 3602, + "GCAAGC": 3603, + "GCAAGG": 3604, + "GCATAA": 3605, + "GCATAT": 3606, + "GCATAC": 3607, + "GCATAG": 3608, + "GCATTA": 3609, + "GCATTT": 3610, + "GCATTC": 3611, + "GCATTG": 3612, + "GCATCA": 3613, + "GCATCT": 3614, + "GCATCC": 3615, + "GCATCG": 3616, + "GCATGA": 3617, + "GCATGT": 3618, + "GCATGC": 3619, + "GCATGG": 3620, + "GCACAA": 3621, + "GCACAT": 3622, + "GCACAC": 3623, + "GCACAG": 3624, + "GCACTA": 3625, + "GCACTT": 3626, + "GCACTC": 3627, + "GCACTG": 3628, + "GCACCA": 3629, + "GCACCT": 3630, + "GCACCC": 3631, + "GCACCG": 3632, + "GCACGA": 3633, + "GCACGT": 3634, + "GCACGC": 3635, + "GCACGG": 3636, + "GCAGAA": 3637, + "GCAGAT": 3638, + "GCAGAC": 3639, + "GCAGAG": 3640, + "GCAGTA": 3641, + "GCAGTT": 3642, + "GCAGTC": 3643, + "GCAGTG": 3644, + "GCAGCA": 3645, + "GCAGCT": 3646, + "GCAGCC": 3647, + "GCAGCG": 3648, + "GCAGGA": 3649, + "GCAGGT": 3650, + "GCAGGC": 3651, + "GCAGGG": 3652, + "GCTAAA": 3653, + "GCTAAT": 3654, + "GCTAAC": 3655, + "GCTAAG": 3656, + "GCTATA": 3657, + "GCTATT": 3658, + "GCTATC": 3659, + "GCTATG": 3660, + "GCTACA": 3661, + "GCTACT": 3662, + "GCTACC": 3663, + "GCTACG": 3664, + "GCTAGA": 3665, + "GCTAGT": 3666, + "GCTAGC": 3667, + "GCTAGG": 3668, + "GCTTAA": 3669, + "GCTTAT": 3670, + "GCTTAC": 3671, + "GCTTAG": 3672, + "GCTTTA": 3673, + "GCTTTT": 3674, + "GCTTTC": 3675, + "GCTTTG": 3676, + "GCTTCA": 3677, + "GCTTCT": 3678, + "GCTTCC": 3679, + "GCTTCG": 3680, + "GCTTGA": 3681, + "GCTTGT": 3682, + "GCTTGC": 3683, + "GCTTGG": 3684, + "GCTCAA": 3685, + "GCTCAT": 3686, + "GCTCAC": 3687, + "GCTCAG": 3688, + "GCTCTA": 3689, + "GCTCTT": 3690, + "GCTCTC": 3691, + "GCTCTG": 3692, + "GCTCCA": 3693, + "GCTCCT": 3694, + "GCTCCC": 3695, + "GCTCCG": 3696, + "GCTCGA": 3697, + "GCTCGT": 3698, + "GCTCGC": 3699, + "GCTCGG": 3700, + "GCTGAA": 3701, + "GCTGAT": 3702, + "GCTGAC": 3703, + "GCTGAG": 3704, + "GCTGTA": 3705, + "GCTGTT": 3706, + "GCTGTC": 3707, + "GCTGTG": 3708, + "GCTGCA": 3709, + "GCTGCT": 3710, + "GCTGCC": 3711, + "GCTGCG": 3712, + "GCTGGA": 3713, + "GCTGGT": 3714, + "GCTGGC": 3715, + "GCTGGG": 3716, + "GCCAAA": 3717, + "GCCAAT": 3718, + "GCCAAC": 3719, + "GCCAAG": 3720, + "GCCATA": 3721, + "GCCATT": 3722, + "GCCATC": 3723, + "GCCATG": 3724, + "GCCACA": 3725, + "GCCACT": 3726, + "GCCACC": 3727, + "GCCACG": 3728, + "GCCAGA": 3729, + "GCCAGT": 3730, + "GCCAGC": 3731, + "GCCAGG": 3732, + "GCCTAA": 3733, + "GCCTAT": 3734, + "GCCTAC": 3735, + "GCCTAG": 3736, + "GCCTTA": 3737, + "GCCTTT": 3738, + "GCCTTC": 3739, + "GCCTTG": 3740, + "GCCTCA": 3741, + "GCCTCT": 3742, + "GCCTCC": 3743, + "GCCTCG": 3744, + "GCCTGA": 3745, + "GCCTGT": 3746, + "GCCTGC": 3747, + "GCCTGG": 3748, + "GCCCAA": 3749, + "GCCCAT": 3750, + "GCCCAC": 3751, + "GCCCAG": 3752, + "GCCCTA": 3753, + "GCCCTT": 3754, + "GCCCTC": 3755, + "GCCCTG": 3756, + "GCCCCA": 3757, + "GCCCCT": 3758, + "GCCCCC": 3759, + "GCCCCG": 3760, + "GCCCGA": 3761, + "GCCCGT": 3762, + "GCCCGC": 3763, + "GCCCGG": 3764, + "GCCGAA": 3765, + "GCCGAT": 3766, + "GCCGAC": 3767, + "GCCGAG": 3768, + "GCCGTA": 3769, + "GCCGTT": 3770, + "GCCGTC": 3771, + "GCCGTG": 3772, + "GCCGCA": 3773, + "GCCGCT": 3774, + "GCCGCC": 3775, + "GCCGCG": 3776, + "GCCGGA": 3777, + "GCCGGT": 3778, + "GCCGGC": 3779, + "GCCGGG": 3780, + "GCGAAA": 3781, + "GCGAAT": 3782, + "GCGAAC": 3783, + "GCGAAG": 3784, + "GCGATA": 3785, + "GCGATT": 3786, + "GCGATC": 3787, + "GCGATG": 3788, + "GCGACA": 3789, + "GCGACT": 3790, + "GCGACC": 3791, + "GCGACG": 3792, + "GCGAGA": 3793, + "GCGAGT": 3794, + "GCGAGC": 3795, + "GCGAGG": 3796, + "GCGTAA": 3797, + "GCGTAT": 3798, + "GCGTAC": 3799, + "GCGTAG": 3800, + "GCGTTA": 3801, + "GCGTTT": 3802, + "GCGTTC": 3803, + "GCGTTG": 3804, + "GCGTCA": 3805, + "GCGTCT": 3806, + "GCGTCC": 3807, + "GCGTCG": 3808, + "GCGTGA": 3809, + "GCGTGT": 3810, + "GCGTGC": 3811, + "GCGTGG": 3812, + "GCGCAA": 3813, + "GCGCAT": 3814, + "GCGCAC": 3815, + "GCGCAG": 3816, + "GCGCTA": 3817, + "GCGCTT": 3818, + "GCGCTC": 3819, + "GCGCTG": 3820, + "GCGCCA": 3821, + "GCGCCT": 3822, + "GCGCCC": 3823, + "GCGCCG": 3824, + "GCGCGA": 3825, + "GCGCGT": 3826, + "GCGCGC": 3827, + "GCGCGG": 3828, + "GCGGAA": 3829, + "GCGGAT": 3830, + "GCGGAC": 3831, + "GCGGAG": 3832, + "GCGGTA": 3833, + "GCGGTT": 3834, + "GCGGTC": 3835, + "GCGGTG": 3836, + "GCGGCA": 3837, + "GCGGCT": 3838, + "GCGGCC": 3839, + "GCGGCG": 3840, + "GCGGGA": 3841, + "GCGGGT": 3842, + "GCGGGC": 3843, + "GCGGGG": 3844, + "GGAAAA": 3845, + "GGAAAT": 3846, + "GGAAAC": 3847, + "GGAAAG": 3848, + "GGAATA": 3849, + "GGAATT": 3850, + "GGAATC": 3851, + "GGAATG": 3852, + "GGAACA": 3853, + "GGAACT": 3854, + "GGAACC": 3855, + "GGAACG": 3856, + "GGAAGA": 3857, + "GGAAGT": 3858, + "GGAAGC": 3859, + "GGAAGG": 3860, + "GGATAA": 3861, + "GGATAT": 3862, + "GGATAC": 3863, + "GGATAG": 3864, + "GGATTA": 3865, + "GGATTT": 3866, + "GGATTC": 3867, + "GGATTG": 3868, + "GGATCA": 3869, + "GGATCT": 3870, + "GGATCC": 3871, + "GGATCG": 3872, + "GGATGA": 3873, + "GGATGT": 3874, + "GGATGC": 3875, + "GGATGG": 3876, + "GGACAA": 3877, + "GGACAT": 3878, + "GGACAC": 3879, + "GGACAG": 3880, + "GGACTA": 3881, + "GGACTT": 3882, + "GGACTC": 3883, + "GGACTG": 3884, + "GGACCA": 3885, + "GGACCT": 3886, + "GGACCC": 3887, + "GGACCG": 3888, + "GGACGA": 3889, + "GGACGT": 3890, + "GGACGC": 3891, + "GGACGG": 3892, + "GGAGAA": 3893, + "GGAGAT": 3894, + "GGAGAC": 3895, + "GGAGAG": 3896, + "GGAGTA": 3897, + "GGAGTT": 3898, + "GGAGTC": 3899, + "GGAGTG": 3900, + "GGAGCA": 3901, + "GGAGCT": 3902, + "GGAGCC": 3903, + "GGAGCG": 3904, + "GGAGGA": 3905, + "GGAGGT": 3906, + "GGAGGC": 3907, + "GGAGGG": 3908, + "GGTAAA": 3909, + "GGTAAT": 3910, + "GGTAAC": 3911, + "GGTAAG": 3912, + "GGTATA": 3913, + "GGTATT": 3914, + "GGTATC": 3915, + "GGTATG": 3916, + "GGTACA": 3917, + "GGTACT": 3918, + "GGTACC": 3919, + "GGTACG": 3920, + "GGTAGA": 3921, + "GGTAGT": 3922, + "GGTAGC": 3923, + "GGTAGG": 3924, + "GGTTAA": 3925, + "GGTTAT": 3926, + "GGTTAC": 3927, + "GGTTAG": 3928, + "GGTTTA": 3929, + "GGTTTT": 3930, + "GGTTTC": 3931, + "GGTTTG": 3932, + "GGTTCA": 3933, + "GGTTCT": 3934, + "GGTTCC": 3935, + "GGTTCG": 3936, + "GGTTGA": 3937, + "GGTTGT": 3938, + "GGTTGC": 3939, + "GGTTGG": 3940, + "GGTCAA": 3941, + "GGTCAT": 3942, + "GGTCAC": 3943, + "GGTCAG": 3944, + "GGTCTA": 3945, + "GGTCTT": 3946, + "GGTCTC": 3947, + "GGTCTG": 3948, + "GGTCCA": 3949, + "GGTCCT": 3950, + "GGTCCC": 3951, + "GGTCCG": 3952, + "GGTCGA": 3953, + "GGTCGT": 3954, + "GGTCGC": 3955, + "GGTCGG": 3956, + "GGTGAA": 3957, + "GGTGAT": 3958, + "GGTGAC": 3959, + "GGTGAG": 3960, + "GGTGTA": 3961, + "GGTGTT": 3962, + "GGTGTC": 3963, + "GGTGTG": 3964, + "GGTGCA": 3965, + "GGTGCT": 3966, + "GGTGCC": 3967, + "GGTGCG": 3968, + "GGTGGA": 3969, + "GGTGGT": 3970, + "GGTGGC": 3971, + "GGTGGG": 3972, + "GGCAAA": 3973, + "GGCAAT": 3974, + "GGCAAC": 3975, + "GGCAAG": 3976, + "GGCATA": 3977, + "GGCATT": 3978, + "GGCATC": 3979, + "GGCATG": 3980, + "GGCACA": 3981, + "GGCACT": 3982, + "GGCACC": 3983, + "GGCACG": 3984, + "GGCAGA": 3985, + "GGCAGT": 3986, + "GGCAGC": 3987, + "GGCAGG": 3988, + "GGCTAA": 3989, + "GGCTAT": 3990, + "GGCTAC": 3991, + "GGCTAG": 3992, + "GGCTTA": 3993, + "GGCTTT": 3994, + "GGCTTC": 3995, + "GGCTTG": 3996, + "GGCTCA": 3997, + "GGCTCT": 3998, + "GGCTCC": 3999, + "GGCTCG": 4000, + "GGCTGA": 4001, + "GGCTGT": 4002, + "GGCTGC": 4003, + "GGCTGG": 4004, + "GGCCAA": 4005, + "GGCCAT": 4006, + "GGCCAC": 4007, + "GGCCAG": 4008, + "GGCCTA": 4009, + "GGCCTT": 4010, + "GGCCTC": 4011, + "GGCCTG": 4012, + "GGCCCA": 4013, + "GGCCCT": 4014, + "GGCCCC": 4015, + "GGCCCG": 4016, + "GGCCGA": 4017, + "GGCCGT": 4018, + "GGCCGC": 4019, + "GGCCGG": 4020, + "GGCGAA": 4021, + "GGCGAT": 4022, + "GGCGAC": 4023, + "GGCGAG": 4024, + "GGCGTA": 4025, + "GGCGTT": 4026, + "GGCGTC": 4027, + "GGCGTG": 4028, + "GGCGCA": 4029, + "GGCGCT": 4030, + "GGCGCC": 4031, + "GGCGCG": 4032, + "GGCGGA": 4033, + "GGCGGT": 4034, + "GGCGGC": 4035, + "GGCGGG": 4036, + "GGGAAA": 4037, + "GGGAAT": 4038, + "GGGAAC": 4039, + "GGGAAG": 4040, + "GGGATA": 4041, + "GGGATT": 4042, + "GGGATC": 4043, + "GGGATG": 4044, + "GGGACA": 4045, + "GGGACT": 4046, + "GGGACC": 4047, + "GGGACG": 4048, + "GGGAGA": 4049, + "GGGAGT": 4050, + "GGGAGC": 4051, + "GGGAGG": 4052, + "GGGTAA": 4053, + "GGGTAT": 4054, + "GGGTAC": 4055, + "GGGTAG": 4056, + "GGGTTA": 4057, + "GGGTTT": 4058, + "GGGTTC": 4059, + "GGGTTG": 4060, + "GGGTCA": 4061, + "GGGTCT": 4062, + "GGGTCC": 4063, + "GGGTCG": 4064, + "GGGTGA": 4065, + "GGGTGT": 4066, + "GGGTGC": 4067, + "GGGTGG": 4068, + "GGGCAA": 4069, + "GGGCAT": 4070, + "GGGCAC": 4071, + "GGGCAG": 4072, + "GGGCTA": 4073, + "GGGCTT": 4074, + "GGGCTC": 4075, + "GGGCTG": 4076, + "GGGCCA": 4077, + "GGGCCT": 4078, + "GGGCCC": 4079, + "GGGCCG": 4080, + "GGGCGA": 4081, + "GGGCGT": 4082, + "GGGCGC": 4083, + "GGGCGG": 4084, + "GGGGAA": 4085, + "GGGGAT": 4086, + "GGGGAC": 4087, + "GGGGAG": 4088, + "GGGGTA": 4089, + "GGGGTT": 4090, + "GGGGTC": 4091, + "GGGGTG": 4092, + "GGGGCA": 4093, + "GGGGCT": 4094, + "GGGGCC": 4095, + "GGGGCG": 4096, + "GGGGGA": 4097, + "GGGGGT": 4098, + "GGGGGC": 4099, + "GGGGGG": 4100 + } + } +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/tokenizer_config.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..604e2b1b29af1dcb3920d2828b6a90276e5e3d02 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/tokenizer_config.json @@ -0,0 +1,58 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "[UNK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "[CLS]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "[SEP]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "4": { + "content": "[MASK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "clean_up_tokenization_spaces": true, + "cls_token": "[CLS]", + "do_basic_tokenize": true, + "do_lower_case": false, + "mask_token": "[MASK]", + "max_len": 512, + "model_max_length": 512, + "never_split": null, + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "strip_accents": null, + "tokenize_chinese_chars": true, + "tokenizer_class": "BertTokenizer", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/trainer_state.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..db112a679464c2584ba557a53d265b1b791a9c4f --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/trainer_state.json @@ -0,0 +1,313 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 4.0, + "eval_steps": 500, + "global_step": 1000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.1, + "grad_norm": 0.9416134357452393, + "learning_rate": 5.319148936170213e-05, + "loss": 0.6912, + "step": 25 + }, + { + "epoch": 0.2, + "grad_norm": 0.9283172488212585, + "learning_rate": 0.00010638297872340425, + "loss": 0.6691, + "step": 50 + }, + { + "epoch": 0.3, + "grad_norm": 0.5870445370674133, + "learning_rate": 0.00015957446808510637, + "loss": 0.6164, + "step": 75 + }, + { + "epoch": 0.4, + "grad_norm": 0.33945509791374207, + "learning_rate": 0.0002127659574468085, + "loss": 0.5453, + "step": 100 + }, + { + "epoch": 0.5, + "grad_norm": 0.21052460372447968, + "learning_rate": 0.00026595744680851064, + "loss": 0.5025, + "step": 125 + }, + { + "epoch": 0.6, + "grad_norm": 0.25948837399482727, + "learning_rate": 0.00031914893617021275, + "loss": 0.4881, + "step": 150 + }, + { + "epoch": 0.7, + "grad_norm": 0.6165992021560669, + "learning_rate": 0.0003723404255319149, + "loss": 0.4852, + "step": 175 + }, + { + "epoch": 0.8, + "grad_norm": 0.33296748995780945, + "learning_rate": 0.0003992081821181128, + "loss": 0.4867, + "step": 200 + }, + { + "epoch": 0.9, + "grad_norm": 0.40745800733566284, + "learning_rate": 0.0003975585615308479, + "loss": 0.4767, + "step": 225 + }, + { + "epoch": 1.0, + "grad_norm": 0.8070425391197205, + "learning_rate": 0.00039590894094358297, + "loss": 0.4784, + "step": 250 + }, + { + "epoch": 1.1, + "grad_norm": 1.2249579429626465, + "learning_rate": 0.0003942593203563181, + "loss": 0.4543, + "step": 275 + }, + { + "epoch": 1.2, + "grad_norm": 1.0143898725509644, + "learning_rate": 0.0003926096997690532, + "loss": 0.4319, + "step": 300 + }, + { + "epoch": 1.3, + "grad_norm": 2.3284952640533447, + "learning_rate": 0.00039096007918178817, + "loss": 0.4455, + "step": 325 + }, + { + "epoch": 1.4, + "grad_norm": 1.189352035522461, + "learning_rate": 0.0003893104585945233, + "loss": 0.4246, + "step": 350 + }, + { + "epoch": 1.5, + "grad_norm": 0.947853684425354, + "learning_rate": 0.0003876608380072583, + "loss": 0.4326, + "step": 375 + }, + { + "epoch": 1.6, + "grad_norm": 0.7130059003829956, + "learning_rate": 0.00038601121741999343, + "loss": 0.4034, + "step": 400 + }, + { + "epoch": 1.7, + "grad_norm": 3.517078161239624, + "learning_rate": 0.0003843615968327285, + "loss": 0.4114, + "step": 425 + }, + { + "epoch": 1.8, + "grad_norm": 0.3982729911804199, + "learning_rate": 0.0003827119762454636, + "loss": 0.4345, + "step": 450 + }, + { + "epoch": 1.9, + "grad_norm": 0.6208440065383911, + "learning_rate": 0.00038106235565819863, + "loss": 0.4135, + "step": 475 + }, + { + "epoch": 2.0, + "grad_norm": 0.9085566401481628, + "learning_rate": 0.0003794127350709337, + "loss": 0.3983, + "step": 500 + }, + { + "epoch": 2.1, + "grad_norm": 1.2225189208984375, + "learning_rate": 0.0003777631144836688, + "loss": 0.3915, + "step": 525 + }, + { + "epoch": 2.2, + "grad_norm": 1.1379530429840088, + "learning_rate": 0.00037611349389640383, + "loss": 0.4335, + "step": 550 + }, + { + "epoch": 2.3, + "grad_norm": 0.8535892367362976, + "learning_rate": 0.00037446387330913894, + "loss": 0.4063, + "step": 575 + }, + { + "epoch": 2.4, + "grad_norm": 1.3587419986724854, + "learning_rate": 0.000372814252721874, + "loss": 0.4065, + "step": 600 + }, + { + "epoch": 2.5, + "grad_norm": 0.9311383962631226, + "learning_rate": 0.00037116463213460903, + "loss": 0.4007, + "step": 625 + }, + { + "epoch": 2.6, + "grad_norm": 0.9912465810775757, + "learning_rate": 0.00036951501154734414, + "loss": 0.3992, + "step": 650 + }, + { + "epoch": 2.7, + "grad_norm": 1.2762119770050049, + "learning_rate": 0.0003678653909600792, + "loss": 0.3827, + "step": 675 + }, + { + "epoch": 2.8, + "grad_norm": 2.3506247997283936, + "learning_rate": 0.0003662157703728143, + "loss": 0.3734, + "step": 700 + }, + { + "epoch": 2.9, + "grad_norm": 1.5375425815582275, + "learning_rate": 0.00036456614978554934, + "loss": 0.4005, + "step": 725 + }, + { + "epoch": 3.0, + "grad_norm": 1.3143396377563477, + "learning_rate": 0.00036291652919828444, + "loss": 0.3548, + "step": 750 + }, + { + "epoch": 3.1, + "grad_norm": 0.662200391292572, + "learning_rate": 0.00036126690861101944, + "loss": 0.4129, + "step": 775 + }, + { + "epoch": 3.2, + "grad_norm": 1.3569878339767456, + "learning_rate": 0.00035961728802375454, + "loss": 0.38, + "step": 800 + }, + { + "epoch": 3.3, + "grad_norm": 0.9888588190078735, + "learning_rate": 0.0003579676674364896, + "loss": 0.3778, + "step": 825 + }, + { + "epoch": 3.4, + "grad_norm": 0.679969310760498, + "learning_rate": 0.0003563180468492247, + "loss": 0.3934, + "step": 850 + }, + { + "epoch": 3.5, + "grad_norm": 1.608539342880249, + "learning_rate": 0.0003546684262619598, + "loss": 0.418, + "step": 875 + }, + { + "epoch": 3.6, + "grad_norm": 1.7037910223007202, + "learning_rate": 0.00035301880567469485, + "loss": 0.3934, + "step": 900 + }, + { + "epoch": 3.7, + "grad_norm": 1.4417132139205933, + "learning_rate": 0.0003513691850874299, + "loss": 0.3947, + "step": 925 + }, + { + "epoch": 3.8, + "grad_norm": 1.1482582092285156, + "learning_rate": 0.00034971956450016495, + "loss": 0.3621, + "step": 950 + }, + { + "epoch": 3.9, + "grad_norm": 2.046616792678833, + "learning_rate": 0.00034806994391290005, + "loss": 0.3774, + "step": 975 + }, + { + "epoch": 4.0, + "grad_norm": 1.0832113027572632, + "learning_rate": 0.0003464203233256351, + "loss": 0.3542, + "step": 1000 + } + ], + "logging_steps": 25, + "max_steps": 6250, + "num_input_tokens_seen": 0, + "num_train_epochs": 25, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 53116108800000.0, + "train_batch_size": 20, + "trial_name": null, + "trial_params": null +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/training_args.bin b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..d61f3500eb70763e1f33496c1787bdf3f07a651b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b6e0f7aa622786274a1ea4195b72f4da224050767f7b89529b85721865c5075a +size 5304 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/vocab.txt b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/vocab.txt new file mode 100644 index 0000000000000000000000000000000000000000..cc9ad53732e21ea9180f18e1fb67a38cc5702d4c --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/vocab.txt @@ -0,0 +1,4101 @@ +[PAD] +[UNK] +[CLS] +[SEP] +[MASK] +AAAAAA +AAAAAT +AAAAAC +AAAAAG +AAAATA +AAAATT +AAAATC +AAAATG +AAAACA +AAAACT +AAAACC +AAAACG +AAAAGA +AAAAGT +AAAAGC +AAAAGG +AAATAA +AAATAT +AAATAC +AAATAG +AAATTA +AAATTT +AAATTC +AAATTG +AAATCA +AAATCT +AAATCC +AAATCG +AAATGA +AAATGT +AAATGC +AAATGG +AAACAA +AAACAT +AAACAC +AAACAG +AAACTA +AAACTT +AAACTC +AAACTG +AAACCA +AAACCT +AAACCC +AAACCG +AAACGA +AAACGT +AAACGC +AAACGG +AAAGAA +AAAGAT +AAAGAC +AAAGAG +AAAGTA +AAAGTT +AAAGTC +AAAGTG +AAAGCA +AAAGCT +AAAGCC +AAAGCG +AAAGGA +AAAGGT +AAAGGC +AAAGGG +AATAAA +AATAAT +AATAAC +AATAAG +AATATA +AATATT +AATATC +AATATG +AATACA +AATACT +AATACC +AATACG +AATAGA +AATAGT +AATAGC +AATAGG +AATTAA +AATTAT +AATTAC +AATTAG +AATTTA +AATTTT +AATTTC +AATTTG +AATTCA +AATTCT +AATTCC +AATTCG +AATTGA +AATTGT +AATTGC +AATTGG +AATCAA +AATCAT +AATCAC +AATCAG +AATCTA +AATCTT +AATCTC +AATCTG +AATCCA +AATCCT +AATCCC +AATCCG +AATCGA +AATCGT +AATCGC +AATCGG +AATGAA +AATGAT +AATGAC +AATGAG +AATGTA +AATGTT +AATGTC +AATGTG +AATGCA +AATGCT +AATGCC +AATGCG +AATGGA +AATGGT +AATGGC +AATGGG +AACAAA +AACAAT +AACAAC +AACAAG +AACATA +AACATT +AACATC +AACATG +AACACA +AACACT +AACACC +AACACG +AACAGA +AACAGT +AACAGC +AACAGG +AACTAA +AACTAT +AACTAC +AACTAG +AACTTA +AACTTT +AACTTC +AACTTG +AACTCA +AACTCT +AACTCC +AACTCG +AACTGA +AACTGT +AACTGC +AACTGG +AACCAA +AACCAT +AACCAC +AACCAG +AACCTA +AACCTT +AACCTC +AACCTG +AACCCA +AACCCT +AACCCC +AACCCG +AACCGA +AACCGT +AACCGC +AACCGG +AACGAA +AACGAT +AACGAC +AACGAG +AACGTA +AACGTT +AACGTC +AACGTG +AACGCA +AACGCT +AACGCC +AACGCG +AACGGA +AACGGT +AACGGC +AACGGG +AAGAAA +AAGAAT +AAGAAC +AAGAAG +AAGATA +AAGATT +AAGATC +AAGATG +AAGACA +AAGACT +AAGACC +AAGACG +AAGAGA +AAGAGT +AAGAGC +AAGAGG +AAGTAA +AAGTAT +AAGTAC +AAGTAG +AAGTTA +AAGTTT +AAGTTC +AAGTTG +AAGTCA +AAGTCT +AAGTCC +AAGTCG +AAGTGA +AAGTGT +AAGTGC +AAGTGG +AAGCAA +AAGCAT +AAGCAC +AAGCAG +AAGCTA +AAGCTT +AAGCTC +AAGCTG +AAGCCA +AAGCCT +AAGCCC +AAGCCG +AAGCGA +AAGCGT +AAGCGC +AAGCGG +AAGGAA +AAGGAT +AAGGAC +AAGGAG +AAGGTA +AAGGTT +AAGGTC +AAGGTG +AAGGCA +AAGGCT +AAGGCC +AAGGCG +AAGGGA +AAGGGT +AAGGGC +AAGGGG +ATAAAA +ATAAAT +ATAAAC +ATAAAG +ATAATA +ATAATT +ATAATC +ATAATG +ATAACA +ATAACT +ATAACC +ATAACG +ATAAGA +ATAAGT +ATAAGC +ATAAGG +ATATAA +ATATAT +ATATAC +ATATAG +ATATTA +ATATTT +ATATTC +ATATTG +ATATCA +ATATCT +ATATCC +ATATCG +ATATGA +ATATGT +ATATGC +ATATGG +ATACAA +ATACAT +ATACAC +ATACAG +ATACTA +ATACTT +ATACTC +ATACTG +ATACCA +ATACCT +ATACCC +ATACCG +ATACGA +ATACGT +ATACGC +ATACGG +ATAGAA +ATAGAT +ATAGAC +ATAGAG +ATAGTA +ATAGTT +ATAGTC +ATAGTG +ATAGCA +ATAGCT +ATAGCC +ATAGCG +ATAGGA +ATAGGT +ATAGGC +ATAGGG +ATTAAA +ATTAAT +ATTAAC +ATTAAG +ATTATA +ATTATT +ATTATC +ATTATG +ATTACA +ATTACT +ATTACC +ATTACG +ATTAGA +ATTAGT +ATTAGC +ATTAGG +ATTTAA +ATTTAT +ATTTAC +ATTTAG +ATTTTA +ATTTTT +ATTTTC +ATTTTG +ATTTCA +ATTTCT +ATTTCC +ATTTCG +ATTTGA +ATTTGT +ATTTGC +ATTTGG +ATTCAA +ATTCAT +ATTCAC +ATTCAG +ATTCTA +ATTCTT +ATTCTC +ATTCTG +ATTCCA +ATTCCT +ATTCCC +ATTCCG +ATTCGA +ATTCGT +ATTCGC +ATTCGG +ATTGAA +ATTGAT +ATTGAC +ATTGAG +ATTGTA +ATTGTT +ATTGTC +ATTGTG +ATTGCA +ATTGCT +ATTGCC +ATTGCG +ATTGGA +ATTGGT +ATTGGC +ATTGGG +ATCAAA +ATCAAT +ATCAAC +ATCAAG +ATCATA +ATCATT +ATCATC +ATCATG +ATCACA +ATCACT +ATCACC +ATCACG +ATCAGA +ATCAGT +ATCAGC +ATCAGG +ATCTAA +ATCTAT +ATCTAC +ATCTAG +ATCTTA +ATCTTT +ATCTTC +ATCTTG +ATCTCA +ATCTCT +ATCTCC +ATCTCG +ATCTGA +ATCTGT +ATCTGC +ATCTGG +ATCCAA +ATCCAT +ATCCAC +ATCCAG +ATCCTA +ATCCTT +ATCCTC +ATCCTG +ATCCCA +ATCCCT +ATCCCC +ATCCCG +ATCCGA +ATCCGT +ATCCGC +ATCCGG +ATCGAA +ATCGAT +ATCGAC +ATCGAG +ATCGTA +ATCGTT +ATCGTC +ATCGTG +ATCGCA +ATCGCT +ATCGCC +ATCGCG +ATCGGA +ATCGGT +ATCGGC +ATCGGG +ATGAAA +ATGAAT +ATGAAC +ATGAAG +ATGATA +ATGATT +ATGATC +ATGATG +ATGACA +ATGACT +ATGACC +ATGACG +ATGAGA +ATGAGT +ATGAGC +ATGAGG +ATGTAA +ATGTAT +ATGTAC +ATGTAG +ATGTTA +ATGTTT +ATGTTC +ATGTTG +ATGTCA +ATGTCT +ATGTCC +ATGTCG +ATGTGA +ATGTGT +ATGTGC +ATGTGG +ATGCAA +ATGCAT +ATGCAC +ATGCAG +ATGCTA +ATGCTT +ATGCTC +ATGCTG +ATGCCA +ATGCCT +ATGCCC +ATGCCG +ATGCGA +ATGCGT +ATGCGC +ATGCGG +ATGGAA +ATGGAT +ATGGAC +ATGGAG +ATGGTA +ATGGTT +ATGGTC +ATGGTG +ATGGCA +ATGGCT +ATGGCC +ATGGCG +ATGGGA +ATGGGT +ATGGGC +ATGGGG +ACAAAA +ACAAAT +ACAAAC +ACAAAG +ACAATA +ACAATT +ACAATC +ACAATG +ACAACA +ACAACT +ACAACC +ACAACG +ACAAGA +ACAAGT +ACAAGC +ACAAGG +ACATAA +ACATAT +ACATAC +ACATAG +ACATTA +ACATTT +ACATTC +ACATTG +ACATCA +ACATCT +ACATCC +ACATCG +ACATGA +ACATGT +ACATGC +ACATGG +ACACAA +ACACAT +ACACAC +ACACAG +ACACTA +ACACTT +ACACTC +ACACTG +ACACCA +ACACCT +ACACCC +ACACCG +ACACGA +ACACGT +ACACGC +ACACGG +ACAGAA +ACAGAT +ACAGAC +ACAGAG +ACAGTA +ACAGTT +ACAGTC +ACAGTG +ACAGCA +ACAGCT +ACAGCC +ACAGCG +ACAGGA +ACAGGT +ACAGGC +ACAGGG +ACTAAA +ACTAAT +ACTAAC +ACTAAG +ACTATA +ACTATT +ACTATC +ACTATG +ACTACA +ACTACT +ACTACC +ACTACG +ACTAGA +ACTAGT +ACTAGC +ACTAGG +ACTTAA +ACTTAT +ACTTAC +ACTTAG +ACTTTA +ACTTTT +ACTTTC +ACTTTG +ACTTCA +ACTTCT +ACTTCC +ACTTCG +ACTTGA +ACTTGT +ACTTGC +ACTTGG +ACTCAA +ACTCAT +ACTCAC +ACTCAG +ACTCTA +ACTCTT +ACTCTC +ACTCTG +ACTCCA +ACTCCT +ACTCCC +ACTCCG +ACTCGA +ACTCGT +ACTCGC +ACTCGG +ACTGAA +ACTGAT +ACTGAC +ACTGAG +ACTGTA +ACTGTT +ACTGTC +ACTGTG +ACTGCA +ACTGCT +ACTGCC +ACTGCG +ACTGGA +ACTGGT +ACTGGC +ACTGGG +ACCAAA +ACCAAT +ACCAAC +ACCAAG +ACCATA +ACCATT +ACCATC +ACCATG +ACCACA +ACCACT +ACCACC +ACCACG +ACCAGA +ACCAGT +ACCAGC +ACCAGG +ACCTAA +ACCTAT +ACCTAC +ACCTAG +ACCTTA +ACCTTT +ACCTTC +ACCTTG +ACCTCA +ACCTCT +ACCTCC +ACCTCG +ACCTGA +ACCTGT +ACCTGC +ACCTGG +ACCCAA +ACCCAT +ACCCAC +ACCCAG +ACCCTA +ACCCTT +ACCCTC +ACCCTG +ACCCCA +ACCCCT +ACCCCC +ACCCCG +ACCCGA +ACCCGT +ACCCGC +ACCCGG +ACCGAA +ACCGAT +ACCGAC +ACCGAG +ACCGTA +ACCGTT +ACCGTC +ACCGTG +ACCGCA +ACCGCT +ACCGCC +ACCGCG +ACCGGA +ACCGGT +ACCGGC +ACCGGG +ACGAAA +ACGAAT +ACGAAC +ACGAAG +ACGATA +ACGATT +ACGATC +ACGATG +ACGACA +ACGACT +ACGACC +ACGACG +ACGAGA +ACGAGT +ACGAGC +ACGAGG +ACGTAA +ACGTAT +ACGTAC +ACGTAG +ACGTTA +ACGTTT +ACGTTC +ACGTTG +ACGTCA +ACGTCT +ACGTCC +ACGTCG +ACGTGA +ACGTGT +ACGTGC +ACGTGG +ACGCAA +ACGCAT +ACGCAC +ACGCAG +ACGCTA +ACGCTT +ACGCTC +ACGCTG +ACGCCA +ACGCCT +ACGCCC +ACGCCG +ACGCGA +ACGCGT +ACGCGC +ACGCGG +ACGGAA +ACGGAT +ACGGAC +ACGGAG +ACGGTA +ACGGTT +ACGGTC +ACGGTG +ACGGCA +ACGGCT +ACGGCC +ACGGCG +ACGGGA +ACGGGT +ACGGGC +ACGGGG +AGAAAA +AGAAAT +AGAAAC +AGAAAG +AGAATA +AGAATT +AGAATC +AGAATG +AGAACA +AGAACT +AGAACC +AGAACG +AGAAGA +AGAAGT +AGAAGC +AGAAGG +AGATAA +AGATAT +AGATAC +AGATAG +AGATTA +AGATTT +AGATTC +AGATTG +AGATCA +AGATCT +AGATCC +AGATCG +AGATGA +AGATGT +AGATGC +AGATGG +AGACAA +AGACAT +AGACAC +AGACAG +AGACTA +AGACTT +AGACTC +AGACTG +AGACCA +AGACCT +AGACCC +AGACCG +AGACGA +AGACGT +AGACGC +AGACGG +AGAGAA +AGAGAT +AGAGAC +AGAGAG +AGAGTA +AGAGTT +AGAGTC +AGAGTG +AGAGCA +AGAGCT +AGAGCC +AGAGCG +AGAGGA +AGAGGT +AGAGGC +AGAGGG +AGTAAA +AGTAAT +AGTAAC +AGTAAG +AGTATA +AGTATT +AGTATC +AGTATG +AGTACA +AGTACT +AGTACC +AGTACG +AGTAGA +AGTAGT +AGTAGC +AGTAGG +AGTTAA +AGTTAT +AGTTAC +AGTTAG +AGTTTA +AGTTTT +AGTTTC +AGTTTG +AGTTCA +AGTTCT +AGTTCC +AGTTCG +AGTTGA +AGTTGT +AGTTGC +AGTTGG +AGTCAA +AGTCAT +AGTCAC +AGTCAG +AGTCTA +AGTCTT +AGTCTC +AGTCTG +AGTCCA +AGTCCT +AGTCCC +AGTCCG +AGTCGA +AGTCGT +AGTCGC +AGTCGG +AGTGAA +AGTGAT +AGTGAC +AGTGAG +AGTGTA +AGTGTT +AGTGTC +AGTGTG +AGTGCA +AGTGCT +AGTGCC +AGTGCG +AGTGGA +AGTGGT +AGTGGC +AGTGGG +AGCAAA +AGCAAT +AGCAAC +AGCAAG +AGCATA +AGCATT +AGCATC +AGCATG +AGCACA +AGCACT +AGCACC +AGCACG +AGCAGA +AGCAGT +AGCAGC +AGCAGG +AGCTAA +AGCTAT +AGCTAC +AGCTAG +AGCTTA +AGCTTT +AGCTTC +AGCTTG +AGCTCA +AGCTCT +AGCTCC +AGCTCG +AGCTGA +AGCTGT +AGCTGC +AGCTGG +AGCCAA +AGCCAT +AGCCAC +AGCCAG +AGCCTA +AGCCTT +AGCCTC +AGCCTG +AGCCCA +AGCCCT +AGCCCC +AGCCCG +AGCCGA +AGCCGT +AGCCGC +AGCCGG +AGCGAA +AGCGAT +AGCGAC +AGCGAG +AGCGTA +AGCGTT +AGCGTC +AGCGTG +AGCGCA +AGCGCT +AGCGCC +AGCGCG +AGCGGA +AGCGGT +AGCGGC +AGCGGG +AGGAAA +AGGAAT +AGGAAC +AGGAAG +AGGATA +AGGATT +AGGATC +AGGATG +AGGACA +AGGACT +AGGACC +AGGACG +AGGAGA +AGGAGT +AGGAGC +AGGAGG +AGGTAA +AGGTAT +AGGTAC +AGGTAG +AGGTTA +AGGTTT +AGGTTC +AGGTTG +AGGTCA +AGGTCT +AGGTCC +AGGTCG +AGGTGA +AGGTGT +AGGTGC +AGGTGG +AGGCAA +AGGCAT +AGGCAC +AGGCAG +AGGCTA +AGGCTT +AGGCTC +AGGCTG +AGGCCA +AGGCCT +AGGCCC +AGGCCG +AGGCGA +AGGCGT +AGGCGC +AGGCGG +AGGGAA +AGGGAT +AGGGAC +AGGGAG +AGGGTA +AGGGTT +AGGGTC +AGGGTG +AGGGCA +AGGGCT +AGGGCC +AGGGCG +AGGGGA +AGGGGT +AGGGGC +AGGGGG +TAAAAA +TAAAAT +TAAAAC +TAAAAG +TAAATA +TAAATT +TAAATC +TAAATG +TAAACA +TAAACT +TAAACC +TAAACG +TAAAGA +TAAAGT +TAAAGC +TAAAGG +TAATAA +TAATAT +TAATAC +TAATAG +TAATTA +TAATTT +TAATTC +TAATTG +TAATCA +TAATCT +TAATCC +TAATCG +TAATGA +TAATGT +TAATGC +TAATGG +TAACAA +TAACAT +TAACAC +TAACAG +TAACTA +TAACTT +TAACTC +TAACTG +TAACCA +TAACCT +TAACCC +TAACCG +TAACGA +TAACGT +TAACGC +TAACGG +TAAGAA +TAAGAT +TAAGAC +TAAGAG +TAAGTA +TAAGTT +TAAGTC +TAAGTG +TAAGCA +TAAGCT +TAAGCC +TAAGCG +TAAGGA +TAAGGT +TAAGGC +TAAGGG +TATAAA +TATAAT +TATAAC +TATAAG +TATATA +TATATT +TATATC +TATATG +TATACA +TATACT +TATACC +TATACG +TATAGA +TATAGT +TATAGC +TATAGG +TATTAA +TATTAT +TATTAC +TATTAG +TATTTA +TATTTT +TATTTC +TATTTG +TATTCA +TATTCT +TATTCC +TATTCG +TATTGA +TATTGT +TATTGC +TATTGG +TATCAA +TATCAT +TATCAC +TATCAG +TATCTA +TATCTT +TATCTC +TATCTG +TATCCA +TATCCT +TATCCC +TATCCG +TATCGA +TATCGT +TATCGC +TATCGG +TATGAA +TATGAT +TATGAC +TATGAG +TATGTA +TATGTT +TATGTC +TATGTG +TATGCA +TATGCT +TATGCC +TATGCG +TATGGA +TATGGT +TATGGC +TATGGG +TACAAA +TACAAT +TACAAC +TACAAG +TACATA +TACATT +TACATC +TACATG +TACACA +TACACT +TACACC +TACACG +TACAGA +TACAGT +TACAGC +TACAGG +TACTAA +TACTAT +TACTAC +TACTAG +TACTTA +TACTTT +TACTTC +TACTTG +TACTCA +TACTCT +TACTCC +TACTCG +TACTGA +TACTGT +TACTGC +TACTGG +TACCAA +TACCAT +TACCAC +TACCAG +TACCTA +TACCTT +TACCTC +TACCTG +TACCCA +TACCCT +TACCCC +TACCCG +TACCGA +TACCGT +TACCGC +TACCGG +TACGAA +TACGAT +TACGAC +TACGAG +TACGTA +TACGTT +TACGTC +TACGTG +TACGCA +TACGCT +TACGCC +TACGCG +TACGGA +TACGGT +TACGGC +TACGGG +TAGAAA +TAGAAT +TAGAAC +TAGAAG +TAGATA +TAGATT +TAGATC +TAGATG +TAGACA +TAGACT +TAGACC +TAGACG +TAGAGA +TAGAGT +TAGAGC +TAGAGG +TAGTAA +TAGTAT +TAGTAC +TAGTAG +TAGTTA +TAGTTT +TAGTTC +TAGTTG +TAGTCA +TAGTCT +TAGTCC +TAGTCG +TAGTGA +TAGTGT +TAGTGC +TAGTGG +TAGCAA +TAGCAT +TAGCAC +TAGCAG +TAGCTA +TAGCTT +TAGCTC +TAGCTG +TAGCCA +TAGCCT +TAGCCC +TAGCCG +TAGCGA +TAGCGT +TAGCGC +TAGCGG +TAGGAA +TAGGAT +TAGGAC +TAGGAG +TAGGTA +TAGGTT +TAGGTC +TAGGTG +TAGGCA +TAGGCT +TAGGCC +TAGGCG +TAGGGA +TAGGGT +TAGGGC +TAGGGG +TTAAAA +TTAAAT +TTAAAC +TTAAAG +TTAATA +TTAATT +TTAATC +TTAATG +TTAACA +TTAACT +TTAACC +TTAACG +TTAAGA +TTAAGT +TTAAGC +TTAAGG +TTATAA +TTATAT +TTATAC +TTATAG +TTATTA +TTATTT +TTATTC +TTATTG +TTATCA +TTATCT +TTATCC +TTATCG +TTATGA +TTATGT +TTATGC +TTATGG +TTACAA +TTACAT +TTACAC +TTACAG +TTACTA +TTACTT +TTACTC +TTACTG +TTACCA +TTACCT +TTACCC +TTACCG +TTACGA +TTACGT +TTACGC +TTACGG +TTAGAA +TTAGAT +TTAGAC +TTAGAG +TTAGTA +TTAGTT +TTAGTC +TTAGTG +TTAGCA +TTAGCT +TTAGCC +TTAGCG +TTAGGA +TTAGGT +TTAGGC +TTAGGG +TTTAAA +TTTAAT +TTTAAC +TTTAAG +TTTATA +TTTATT +TTTATC +TTTATG +TTTACA +TTTACT +TTTACC +TTTACG +TTTAGA +TTTAGT +TTTAGC +TTTAGG +TTTTAA +TTTTAT +TTTTAC +TTTTAG +TTTTTA +TTTTTT +TTTTTC +TTTTTG +TTTTCA +TTTTCT +TTTTCC +TTTTCG +TTTTGA +TTTTGT +TTTTGC +TTTTGG +TTTCAA +TTTCAT +TTTCAC +TTTCAG +TTTCTA +TTTCTT +TTTCTC +TTTCTG +TTTCCA +TTTCCT +TTTCCC +TTTCCG +TTTCGA +TTTCGT +TTTCGC +TTTCGG +TTTGAA +TTTGAT +TTTGAC +TTTGAG +TTTGTA +TTTGTT +TTTGTC +TTTGTG +TTTGCA +TTTGCT +TTTGCC +TTTGCG +TTTGGA +TTTGGT +TTTGGC +TTTGGG +TTCAAA +TTCAAT +TTCAAC +TTCAAG +TTCATA +TTCATT +TTCATC +TTCATG +TTCACA +TTCACT +TTCACC +TTCACG +TTCAGA +TTCAGT +TTCAGC +TTCAGG +TTCTAA +TTCTAT +TTCTAC +TTCTAG +TTCTTA +TTCTTT +TTCTTC +TTCTTG +TTCTCA +TTCTCT +TTCTCC +TTCTCG +TTCTGA +TTCTGT +TTCTGC +TTCTGG +TTCCAA +TTCCAT +TTCCAC +TTCCAG +TTCCTA +TTCCTT +TTCCTC +TTCCTG +TTCCCA +TTCCCT +TTCCCC +TTCCCG +TTCCGA +TTCCGT +TTCCGC +TTCCGG +TTCGAA +TTCGAT +TTCGAC +TTCGAG +TTCGTA +TTCGTT +TTCGTC +TTCGTG +TTCGCA +TTCGCT +TTCGCC +TTCGCG +TTCGGA +TTCGGT +TTCGGC +TTCGGG +TTGAAA +TTGAAT +TTGAAC +TTGAAG +TTGATA +TTGATT +TTGATC +TTGATG +TTGACA +TTGACT +TTGACC +TTGACG +TTGAGA +TTGAGT +TTGAGC +TTGAGG +TTGTAA +TTGTAT +TTGTAC +TTGTAG +TTGTTA +TTGTTT +TTGTTC +TTGTTG +TTGTCA +TTGTCT +TTGTCC +TTGTCG +TTGTGA +TTGTGT +TTGTGC +TTGTGG +TTGCAA +TTGCAT +TTGCAC +TTGCAG +TTGCTA +TTGCTT +TTGCTC +TTGCTG +TTGCCA +TTGCCT +TTGCCC +TTGCCG +TTGCGA +TTGCGT +TTGCGC +TTGCGG +TTGGAA +TTGGAT +TTGGAC +TTGGAG +TTGGTA +TTGGTT +TTGGTC +TTGGTG +TTGGCA +TTGGCT +TTGGCC +TTGGCG +TTGGGA +TTGGGT +TTGGGC +TTGGGG +TCAAAA +TCAAAT +TCAAAC +TCAAAG +TCAATA +TCAATT +TCAATC +TCAATG +TCAACA +TCAACT +TCAACC +TCAACG +TCAAGA +TCAAGT +TCAAGC +TCAAGG +TCATAA +TCATAT +TCATAC +TCATAG +TCATTA +TCATTT +TCATTC +TCATTG +TCATCA +TCATCT +TCATCC +TCATCG +TCATGA +TCATGT +TCATGC +TCATGG +TCACAA +TCACAT +TCACAC +TCACAG +TCACTA +TCACTT +TCACTC +TCACTG +TCACCA +TCACCT +TCACCC +TCACCG +TCACGA +TCACGT +TCACGC +TCACGG +TCAGAA +TCAGAT +TCAGAC +TCAGAG +TCAGTA +TCAGTT +TCAGTC +TCAGTG +TCAGCA +TCAGCT +TCAGCC +TCAGCG +TCAGGA +TCAGGT +TCAGGC +TCAGGG +TCTAAA +TCTAAT +TCTAAC +TCTAAG +TCTATA +TCTATT +TCTATC +TCTATG +TCTACA +TCTACT +TCTACC +TCTACG +TCTAGA +TCTAGT +TCTAGC +TCTAGG +TCTTAA +TCTTAT +TCTTAC +TCTTAG +TCTTTA +TCTTTT +TCTTTC +TCTTTG +TCTTCA +TCTTCT +TCTTCC +TCTTCG +TCTTGA +TCTTGT +TCTTGC +TCTTGG +TCTCAA +TCTCAT +TCTCAC +TCTCAG +TCTCTA +TCTCTT +TCTCTC +TCTCTG +TCTCCA +TCTCCT +TCTCCC +TCTCCG +TCTCGA +TCTCGT +TCTCGC +TCTCGG +TCTGAA +TCTGAT +TCTGAC +TCTGAG +TCTGTA +TCTGTT +TCTGTC +TCTGTG +TCTGCA +TCTGCT +TCTGCC +TCTGCG +TCTGGA +TCTGGT +TCTGGC +TCTGGG +TCCAAA +TCCAAT +TCCAAC +TCCAAG +TCCATA +TCCATT +TCCATC +TCCATG +TCCACA +TCCACT +TCCACC +TCCACG +TCCAGA +TCCAGT +TCCAGC +TCCAGG +TCCTAA +TCCTAT +TCCTAC +TCCTAG +TCCTTA +TCCTTT +TCCTTC +TCCTTG +TCCTCA +TCCTCT +TCCTCC +TCCTCG +TCCTGA +TCCTGT +TCCTGC +TCCTGG +TCCCAA +TCCCAT +TCCCAC +TCCCAG +TCCCTA +TCCCTT +TCCCTC +TCCCTG +TCCCCA +TCCCCT +TCCCCC +TCCCCG +TCCCGA +TCCCGT +TCCCGC +TCCCGG +TCCGAA +TCCGAT +TCCGAC +TCCGAG +TCCGTA +TCCGTT +TCCGTC +TCCGTG +TCCGCA +TCCGCT +TCCGCC +TCCGCG +TCCGGA +TCCGGT +TCCGGC +TCCGGG +TCGAAA +TCGAAT +TCGAAC +TCGAAG +TCGATA +TCGATT +TCGATC +TCGATG +TCGACA +TCGACT +TCGACC +TCGACG +TCGAGA +TCGAGT +TCGAGC +TCGAGG +TCGTAA +TCGTAT +TCGTAC +TCGTAG +TCGTTA +TCGTTT +TCGTTC +TCGTTG +TCGTCA +TCGTCT +TCGTCC +TCGTCG +TCGTGA +TCGTGT +TCGTGC +TCGTGG +TCGCAA +TCGCAT +TCGCAC +TCGCAG +TCGCTA +TCGCTT +TCGCTC +TCGCTG +TCGCCA +TCGCCT +TCGCCC +TCGCCG +TCGCGA +TCGCGT +TCGCGC +TCGCGG +TCGGAA +TCGGAT +TCGGAC +TCGGAG +TCGGTA +TCGGTT +TCGGTC +TCGGTG +TCGGCA +TCGGCT +TCGGCC +TCGGCG +TCGGGA +TCGGGT +TCGGGC +TCGGGG +TGAAAA +TGAAAT +TGAAAC +TGAAAG +TGAATA +TGAATT +TGAATC +TGAATG +TGAACA +TGAACT +TGAACC +TGAACG +TGAAGA +TGAAGT +TGAAGC +TGAAGG +TGATAA +TGATAT +TGATAC +TGATAG +TGATTA +TGATTT +TGATTC +TGATTG +TGATCA +TGATCT +TGATCC +TGATCG +TGATGA +TGATGT +TGATGC +TGATGG +TGACAA +TGACAT +TGACAC +TGACAG +TGACTA +TGACTT +TGACTC +TGACTG +TGACCA +TGACCT +TGACCC +TGACCG +TGACGA +TGACGT +TGACGC +TGACGG +TGAGAA +TGAGAT +TGAGAC +TGAGAG +TGAGTA +TGAGTT +TGAGTC +TGAGTG +TGAGCA +TGAGCT +TGAGCC +TGAGCG +TGAGGA +TGAGGT +TGAGGC +TGAGGG +TGTAAA +TGTAAT +TGTAAC +TGTAAG +TGTATA +TGTATT +TGTATC +TGTATG +TGTACA +TGTACT +TGTACC +TGTACG +TGTAGA +TGTAGT +TGTAGC +TGTAGG +TGTTAA +TGTTAT +TGTTAC +TGTTAG +TGTTTA +TGTTTT +TGTTTC +TGTTTG +TGTTCA +TGTTCT +TGTTCC +TGTTCG +TGTTGA +TGTTGT +TGTTGC +TGTTGG +TGTCAA +TGTCAT +TGTCAC +TGTCAG +TGTCTA +TGTCTT +TGTCTC +TGTCTG +TGTCCA +TGTCCT +TGTCCC +TGTCCG +TGTCGA +TGTCGT +TGTCGC +TGTCGG +TGTGAA +TGTGAT +TGTGAC +TGTGAG +TGTGTA +TGTGTT +TGTGTC +TGTGTG +TGTGCA +TGTGCT +TGTGCC +TGTGCG +TGTGGA +TGTGGT +TGTGGC +TGTGGG +TGCAAA +TGCAAT +TGCAAC +TGCAAG +TGCATA +TGCATT +TGCATC +TGCATG +TGCACA +TGCACT +TGCACC +TGCACG +TGCAGA +TGCAGT +TGCAGC +TGCAGG +TGCTAA +TGCTAT +TGCTAC +TGCTAG +TGCTTA +TGCTTT +TGCTTC +TGCTTG +TGCTCA +TGCTCT +TGCTCC +TGCTCG +TGCTGA +TGCTGT +TGCTGC +TGCTGG +TGCCAA +TGCCAT +TGCCAC +TGCCAG +TGCCTA +TGCCTT +TGCCTC +TGCCTG +TGCCCA +TGCCCT +TGCCCC +TGCCCG +TGCCGA +TGCCGT +TGCCGC +TGCCGG +TGCGAA +TGCGAT +TGCGAC +TGCGAG +TGCGTA +TGCGTT +TGCGTC +TGCGTG +TGCGCA +TGCGCT +TGCGCC +TGCGCG +TGCGGA +TGCGGT +TGCGGC +TGCGGG +TGGAAA +TGGAAT +TGGAAC +TGGAAG +TGGATA +TGGATT +TGGATC +TGGATG +TGGACA +TGGACT +TGGACC +TGGACG +TGGAGA +TGGAGT +TGGAGC +TGGAGG +TGGTAA +TGGTAT +TGGTAC +TGGTAG +TGGTTA +TGGTTT +TGGTTC +TGGTTG +TGGTCA +TGGTCT +TGGTCC +TGGTCG +TGGTGA +TGGTGT +TGGTGC +TGGTGG +TGGCAA +TGGCAT +TGGCAC +TGGCAG +TGGCTA +TGGCTT +TGGCTC +TGGCTG +TGGCCA +TGGCCT +TGGCCC +TGGCCG +TGGCGA +TGGCGT +TGGCGC +TGGCGG +TGGGAA +TGGGAT +TGGGAC +TGGGAG +TGGGTA +TGGGTT +TGGGTC +TGGGTG +TGGGCA +TGGGCT +TGGGCC +TGGGCG +TGGGGA +TGGGGT +TGGGGC +TGGGGG +CAAAAA +CAAAAT +CAAAAC +CAAAAG +CAAATA +CAAATT +CAAATC +CAAATG +CAAACA +CAAACT +CAAACC +CAAACG +CAAAGA +CAAAGT +CAAAGC +CAAAGG +CAATAA +CAATAT +CAATAC +CAATAG +CAATTA +CAATTT +CAATTC +CAATTG +CAATCA +CAATCT +CAATCC +CAATCG +CAATGA +CAATGT +CAATGC +CAATGG +CAACAA +CAACAT +CAACAC +CAACAG +CAACTA +CAACTT +CAACTC +CAACTG +CAACCA +CAACCT +CAACCC +CAACCG +CAACGA +CAACGT +CAACGC +CAACGG +CAAGAA +CAAGAT +CAAGAC +CAAGAG +CAAGTA +CAAGTT +CAAGTC +CAAGTG +CAAGCA +CAAGCT +CAAGCC +CAAGCG +CAAGGA +CAAGGT +CAAGGC +CAAGGG +CATAAA +CATAAT +CATAAC +CATAAG +CATATA +CATATT +CATATC +CATATG +CATACA +CATACT +CATACC +CATACG +CATAGA +CATAGT +CATAGC +CATAGG +CATTAA +CATTAT +CATTAC +CATTAG +CATTTA +CATTTT +CATTTC +CATTTG +CATTCA +CATTCT +CATTCC +CATTCG +CATTGA +CATTGT +CATTGC +CATTGG +CATCAA +CATCAT +CATCAC +CATCAG +CATCTA +CATCTT +CATCTC +CATCTG +CATCCA +CATCCT +CATCCC +CATCCG +CATCGA +CATCGT +CATCGC +CATCGG +CATGAA +CATGAT +CATGAC +CATGAG +CATGTA +CATGTT +CATGTC +CATGTG +CATGCA +CATGCT +CATGCC +CATGCG +CATGGA +CATGGT +CATGGC +CATGGG +CACAAA +CACAAT +CACAAC +CACAAG +CACATA +CACATT +CACATC +CACATG +CACACA +CACACT +CACACC +CACACG +CACAGA +CACAGT +CACAGC +CACAGG +CACTAA +CACTAT +CACTAC +CACTAG +CACTTA +CACTTT +CACTTC +CACTTG +CACTCA +CACTCT +CACTCC +CACTCG +CACTGA +CACTGT +CACTGC +CACTGG +CACCAA +CACCAT +CACCAC +CACCAG +CACCTA +CACCTT +CACCTC +CACCTG +CACCCA +CACCCT +CACCCC +CACCCG +CACCGA +CACCGT +CACCGC +CACCGG +CACGAA +CACGAT +CACGAC +CACGAG +CACGTA +CACGTT +CACGTC +CACGTG +CACGCA +CACGCT +CACGCC +CACGCG +CACGGA +CACGGT +CACGGC +CACGGG +CAGAAA +CAGAAT +CAGAAC +CAGAAG +CAGATA +CAGATT +CAGATC +CAGATG +CAGACA +CAGACT +CAGACC +CAGACG +CAGAGA +CAGAGT +CAGAGC +CAGAGG +CAGTAA +CAGTAT +CAGTAC +CAGTAG +CAGTTA +CAGTTT +CAGTTC +CAGTTG +CAGTCA +CAGTCT +CAGTCC +CAGTCG +CAGTGA +CAGTGT +CAGTGC +CAGTGG +CAGCAA +CAGCAT +CAGCAC +CAGCAG +CAGCTA +CAGCTT +CAGCTC +CAGCTG +CAGCCA +CAGCCT +CAGCCC +CAGCCG +CAGCGA +CAGCGT +CAGCGC +CAGCGG +CAGGAA +CAGGAT +CAGGAC +CAGGAG +CAGGTA +CAGGTT +CAGGTC +CAGGTG +CAGGCA +CAGGCT +CAGGCC +CAGGCG +CAGGGA +CAGGGT +CAGGGC +CAGGGG +CTAAAA +CTAAAT +CTAAAC +CTAAAG +CTAATA +CTAATT +CTAATC +CTAATG +CTAACA +CTAACT +CTAACC +CTAACG +CTAAGA +CTAAGT +CTAAGC +CTAAGG +CTATAA +CTATAT +CTATAC +CTATAG +CTATTA +CTATTT +CTATTC +CTATTG +CTATCA +CTATCT +CTATCC +CTATCG +CTATGA +CTATGT +CTATGC +CTATGG +CTACAA +CTACAT +CTACAC +CTACAG +CTACTA +CTACTT +CTACTC +CTACTG +CTACCA +CTACCT +CTACCC +CTACCG +CTACGA +CTACGT +CTACGC +CTACGG +CTAGAA +CTAGAT +CTAGAC +CTAGAG +CTAGTA +CTAGTT +CTAGTC +CTAGTG +CTAGCA +CTAGCT +CTAGCC +CTAGCG +CTAGGA +CTAGGT +CTAGGC +CTAGGG +CTTAAA +CTTAAT +CTTAAC +CTTAAG +CTTATA +CTTATT +CTTATC +CTTATG +CTTACA +CTTACT +CTTACC +CTTACG +CTTAGA +CTTAGT +CTTAGC +CTTAGG +CTTTAA +CTTTAT +CTTTAC +CTTTAG +CTTTTA +CTTTTT +CTTTTC +CTTTTG +CTTTCA +CTTTCT +CTTTCC +CTTTCG +CTTTGA +CTTTGT +CTTTGC +CTTTGG +CTTCAA +CTTCAT +CTTCAC +CTTCAG +CTTCTA +CTTCTT +CTTCTC +CTTCTG +CTTCCA +CTTCCT +CTTCCC +CTTCCG +CTTCGA +CTTCGT +CTTCGC +CTTCGG +CTTGAA +CTTGAT +CTTGAC +CTTGAG +CTTGTA +CTTGTT +CTTGTC +CTTGTG +CTTGCA +CTTGCT +CTTGCC +CTTGCG +CTTGGA +CTTGGT +CTTGGC +CTTGGG +CTCAAA +CTCAAT +CTCAAC +CTCAAG +CTCATA +CTCATT +CTCATC +CTCATG +CTCACA +CTCACT +CTCACC +CTCACG +CTCAGA +CTCAGT +CTCAGC +CTCAGG +CTCTAA +CTCTAT +CTCTAC +CTCTAG +CTCTTA +CTCTTT +CTCTTC +CTCTTG +CTCTCA +CTCTCT +CTCTCC +CTCTCG +CTCTGA +CTCTGT +CTCTGC +CTCTGG +CTCCAA +CTCCAT +CTCCAC +CTCCAG +CTCCTA +CTCCTT +CTCCTC +CTCCTG +CTCCCA +CTCCCT +CTCCCC +CTCCCG +CTCCGA +CTCCGT +CTCCGC +CTCCGG +CTCGAA +CTCGAT +CTCGAC +CTCGAG +CTCGTA +CTCGTT +CTCGTC +CTCGTG +CTCGCA +CTCGCT +CTCGCC +CTCGCG +CTCGGA +CTCGGT +CTCGGC +CTCGGG +CTGAAA +CTGAAT +CTGAAC +CTGAAG +CTGATA +CTGATT +CTGATC +CTGATG +CTGACA +CTGACT +CTGACC +CTGACG +CTGAGA +CTGAGT +CTGAGC +CTGAGG +CTGTAA +CTGTAT +CTGTAC +CTGTAG +CTGTTA +CTGTTT +CTGTTC +CTGTTG +CTGTCA +CTGTCT +CTGTCC +CTGTCG +CTGTGA +CTGTGT +CTGTGC +CTGTGG +CTGCAA +CTGCAT +CTGCAC +CTGCAG +CTGCTA +CTGCTT +CTGCTC +CTGCTG +CTGCCA +CTGCCT +CTGCCC +CTGCCG +CTGCGA +CTGCGT +CTGCGC +CTGCGG +CTGGAA +CTGGAT +CTGGAC +CTGGAG +CTGGTA +CTGGTT +CTGGTC +CTGGTG +CTGGCA +CTGGCT +CTGGCC +CTGGCG +CTGGGA +CTGGGT +CTGGGC +CTGGGG +CCAAAA +CCAAAT +CCAAAC +CCAAAG +CCAATA +CCAATT +CCAATC +CCAATG +CCAACA +CCAACT +CCAACC +CCAACG +CCAAGA +CCAAGT +CCAAGC +CCAAGG +CCATAA +CCATAT +CCATAC +CCATAG +CCATTA +CCATTT +CCATTC +CCATTG +CCATCA +CCATCT +CCATCC +CCATCG +CCATGA +CCATGT +CCATGC +CCATGG +CCACAA +CCACAT +CCACAC +CCACAG +CCACTA +CCACTT +CCACTC +CCACTG +CCACCA +CCACCT +CCACCC +CCACCG +CCACGA +CCACGT +CCACGC +CCACGG +CCAGAA +CCAGAT +CCAGAC +CCAGAG +CCAGTA +CCAGTT +CCAGTC +CCAGTG +CCAGCA +CCAGCT +CCAGCC +CCAGCG +CCAGGA +CCAGGT +CCAGGC +CCAGGG +CCTAAA +CCTAAT +CCTAAC +CCTAAG +CCTATA +CCTATT +CCTATC +CCTATG +CCTACA +CCTACT +CCTACC +CCTACG +CCTAGA +CCTAGT +CCTAGC +CCTAGG +CCTTAA +CCTTAT +CCTTAC +CCTTAG +CCTTTA +CCTTTT +CCTTTC +CCTTTG +CCTTCA +CCTTCT +CCTTCC +CCTTCG +CCTTGA +CCTTGT +CCTTGC +CCTTGG +CCTCAA +CCTCAT +CCTCAC +CCTCAG +CCTCTA +CCTCTT +CCTCTC +CCTCTG +CCTCCA +CCTCCT +CCTCCC +CCTCCG +CCTCGA +CCTCGT +CCTCGC +CCTCGG +CCTGAA +CCTGAT +CCTGAC +CCTGAG +CCTGTA +CCTGTT +CCTGTC +CCTGTG +CCTGCA +CCTGCT +CCTGCC +CCTGCG +CCTGGA +CCTGGT +CCTGGC +CCTGGG +CCCAAA +CCCAAT +CCCAAC +CCCAAG +CCCATA +CCCATT +CCCATC +CCCATG +CCCACA +CCCACT +CCCACC +CCCACG +CCCAGA +CCCAGT +CCCAGC +CCCAGG +CCCTAA +CCCTAT +CCCTAC +CCCTAG +CCCTTA +CCCTTT +CCCTTC +CCCTTG +CCCTCA +CCCTCT +CCCTCC +CCCTCG +CCCTGA +CCCTGT +CCCTGC +CCCTGG +CCCCAA +CCCCAT +CCCCAC +CCCCAG +CCCCTA +CCCCTT +CCCCTC +CCCCTG +CCCCCA +CCCCCT +CCCCCC +CCCCCG +CCCCGA +CCCCGT +CCCCGC +CCCCGG +CCCGAA +CCCGAT +CCCGAC +CCCGAG +CCCGTA +CCCGTT +CCCGTC +CCCGTG +CCCGCA +CCCGCT +CCCGCC +CCCGCG +CCCGGA +CCCGGT +CCCGGC +CCCGGG +CCGAAA +CCGAAT +CCGAAC +CCGAAG +CCGATA +CCGATT +CCGATC +CCGATG +CCGACA +CCGACT +CCGACC +CCGACG +CCGAGA +CCGAGT +CCGAGC +CCGAGG +CCGTAA +CCGTAT +CCGTAC +CCGTAG +CCGTTA +CCGTTT +CCGTTC +CCGTTG +CCGTCA +CCGTCT +CCGTCC +CCGTCG +CCGTGA +CCGTGT +CCGTGC +CCGTGG +CCGCAA +CCGCAT +CCGCAC +CCGCAG +CCGCTA +CCGCTT +CCGCTC +CCGCTG +CCGCCA +CCGCCT +CCGCCC +CCGCCG +CCGCGA +CCGCGT +CCGCGC +CCGCGG +CCGGAA +CCGGAT +CCGGAC +CCGGAG +CCGGTA +CCGGTT +CCGGTC +CCGGTG +CCGGCA +CCGGCT +CCGGCC +CCGGCG +CCGGGA +CCGGGT +CCGGGC +CCGGGG +CGAAAA +CGAAAT +CGAAAC +CGAAAG +CGAATA +CGAATT +CGAATC +CGAATG +CGAACA +CGAACT +CGAACC +CGAACG +CGAAGA +CGAAGT +CGAAGC +CGAAGG +CGATAA +CGATAT +CGATAC +CGATAG +CGATTA +CGATTT +CGATTC +CGATTG +CGATCA +CGATCT +CGATCC +CGATCG +CGATGA +CGATGT +CGATGC +CGATGG +CGACAA +CGACAT +CGACAC +CGACAG +CGACTA +CGACTT +CGACTC +CGACTG +CGACCA +CGACCT +CGACCC +CGACCG +CGACGA +CGACGT +CGACGC +CGACGG +CGAGAA +CGAGAT +CGAGAC +CGAGAG +CGAGTA +CGAGTT +CGAGTC +CGAGTG +CGAGCA +CGAGCT +CGAGCC +CGAGCG +CGAGGA +CGAGGT +CGAGGC +CGAGGG +CGTAAA +CGTAAT +CGTAAC +CGTAAG +CGTATA +CGTATT +CGTATC +CGTATG +CGTACA +CGTACT +CGTACC +CGTACG +CGTAGA +CGTAGT +CGTAGC +CGTAGG +CGTTAA +CGTTAT +CGTTAC +CGTTAG +CGTTTA +CGTTTT +CGTTTC +CGTTTG +CGTTCA +CGTTCT +CGTTCC +CGTTCG +CGTTGA +CGTTGT +CGTTGC +CGTTGG +CGTCAA +CGTCAT +CGTCAC +CGTCAG +CGTCTA +CGTCTT +CGTCTC +CGTCTG +CGTCCA +CGTCCT +CGTCCC +CGTCCG +CGTCGA +CGTCGT +CGTCGC +CGTCGG +CGTGAA +CGTGAT +CGTGAC +CGTGAG +CGTGTA +CGTGTT +CGTGTC +CGTGTG +CGTGCA +CGTGCT +CGTGCC +CGTGCG +CGTGGA +CGTGGT +CGTGGC +CGTGGG +CGCAAA +CGCAAT +CGCAAC +CGCAAG +CGCATA +CGCATT +CGCATC +CGCATG +CGCACA +CGCACT +CGCACC +CGCACG +CGCAGA +CGCAGT +CGCAGC +CGCAGG +CGCTAA +CGCTAT +CGCTAC +CGCTAG +CGCTTA +CGCTTT +CGCTTC +CGCTTG +CGCTCA +CGCTCT +CGCTCC +CGCTCG +CGCTGA +CGCTGT +CGCTGC +CGCTGG +CGCCAA +CGCCAT +CGCCAC +CGCCAG +CGCCTA +CGCCTT +CGCCTC +CGCCTG +CGCCCA +CGCCCT +CGCCCC +CGCCCG +CGCCGA +CGCCGT +CGCCGC +CGCCGG +CGCGAA +CGCGAT +CGCGAC +CGCGAG +CGCGTA +CGCGTT +CGCGTC +CGCGTG +CGCGCA +CGCGCT +CGCGCC +CGCGCG +CGCGGA +CGCGGT +CGCGGC +CGCGGG +CGGAAA +CGGAAT +CGGAAC +CGGAAG +CGGATA +CGGATT +CGGATC +CGGATG +CGGACA +CGGACT +CGGACC +CGGACG +CGGAGA +CGGAGT +CGGAGC +CGGAGG +CGGTAA +CGGTAT +CGGTAC +CGGTAG +CGGTTA +CGGTTT +CGGTTC +CGGTTG +CGGTCA +CGGTCT +CGGTCC +CGGTCG +CGGTGA +CGGTGT +CGGTGC +CGGTGG +CGGCAA +CGGCAT +CGGCAC +CGGCAG +CGGCTA +CGGCTT +CGGCTC +CGGCTG +CGGCCA +CGGCCT +CGGCCC +CGGCCG +CGGCGA +CGGCGT +CGGCGC +CGGCGG +CGGGAA +CGGGAT +CGGGAC +CGGGAG +CGGGTA +CGGGTT +CGGGTC +CGGGTG +CGGGCA +CGGGCT +CGGGCC +CGGGCG +CGGGGA +CGGGGT +CGGGGC +CGGGGG +GAAAAA +GAAAAT +GAAAAC +GAAAAG +GAAATA +GAAATT +GAAATC +GAAATG +GAAACA +GAAACT +GAAACC +GAAACG +GAAAGA +GAAAGT +GAAAGC +GAAAGG +GAATAA +GAATAT +GAATAC +GAATAG +GAATTA +GAATTT +GAATTC +GAATTG +GAATCA +GAATCT +GAATCC +GAATCG +GAATGA +GAATGT +GAATGC +GAATGG +GAACAA +GAACAT +GAACAC +GAACAG +GAACTA +GAACTT +GAACTC +GAACTG +GAACCA +GAACCT +GAACCC +GAACCG +GAACGA +GAACGT +GAACGC +GAACGG +GAAGAA +GAAGAT +GAAGAC +GAAGAG +GAAGTA +GAAGTT +GAAGTC +GAAGTG +GAAGCA +GAAGCT +GAAGCC +GAAGCG +GAAGGA +GAAGGT +GAAGGC +GAAGGG +GATAAA +GATAAT +GATAAC +GATAAG +GATATA +GATATT +GATATC +GATATG +GATACA +GATACT +GATACC +GATACG +GATAGA +GATAGT +GATAGC +GATAGG +GATTAA +GATTAT +GATTAC +GATTAG +GATTTA +GATTTT +GATTTC +GATTTG +GATTCA +GATTCT +GATTCC +GATTCG +GATTGA +GATTGT +GATTGC +GATTGG +GATCAA +GATCAT +GATCAC +GATCAG +GATCTA +GATCTT +GATCTC +GATCTG +GATCCA +GATCCT +GATCCC +GATCCG +GATCGA +GATCGT +GATCGC +GATCGG +GATGAA +GATGAT +GATGAC +GATGAG +GATGTA +GATGTT +GATGTC +GATGTG +GATGCA +GATGCT +GATGCC +GATGCG +GATGGA +GATGGT +GATGGC +GATGGG +GACAAA +GACAAT +GACAAC +GACAAG +GACATA +GACATT +GACATC +GACATG +GACACA +GACACT +GACACC +GACACG +GACAGA +GACAGT +GACAGC +GACAGG +GACTAA +GACTAT +GACTAC +GACTAG +GACTTA +GACTTT +GACTTC +GACTTG +GACTCA +GACTCT +GACTCC +GACTCG +GACTGA +GACTGT +GACTGC +GACTGG +GACCAA +GACCAT +GACCAC +GACCAG +GACCTA +GACCTT +GACCTC +GACCTG +GACCCA +GACCCT +GACCCC +GACCCG +GACCGA +GACCGT +GACCGC +GACCGG +GACGAA +GACGAT +GACGAC +GACGAG +GACGTA +GACGTT +GACGTC +GACGTG +GACGCA +GACGCT +GACGCC +GACGCG +GACGGA +GACGGT +GACGGC +GACGGG +GAGAAA +GAGAAT +GAGAAC +GAGAAG +GAGATA +GAGATT +GAGATC +GAGATG +GAGACA +GAGACT +GAGACC +GAGACG +GAGAGA +GAGAGT +GAGAGC +GAGAGG +GAGTAA +GAGTAT +GAGTAC +GAGTAG +GAGTTA +GAGTTT +GAGTTC +GAGTTG +GAGTCA +GAGTCT +GAGTCC +GAGTCG +GAGTGA +GAGTGT +GAGTGC +GAGTGG +GAGCAA +GAGCAT +GAGCAC +GAGCAG +GAGCTA +GAGCTT +GAGCTC +GAGCTG +GAGCCA +GAGCCT +GAGCCC +GAGCCG +GAGCGA +GAGCGT +GAGCGC +GAGCGG +GAGGAA +GAGGAT +GAGGAC +GAGGAG +GAGGTA +GAGGTT +GAGGTC +GAGGTG +GAGGCA +GAGGCT +GAGGCC +GAGGCG +GAGGGA +GAGGGT +GAGGGC +GAGGGG +GTAAAA +GTAAAT +GTAAAC +GTAAAG +GTAATA +GTAATT +GTAATC +GTAATG +GTAACA +GTAACT +GTAACC +GTAACG +GTAAGA +GTAAGT +GTAAGC +GTAAGG +GTATAA +GTATAT +GTATAC +GTATAG +GTATTA +GTATTT +GTATTC +GTATTG +GTATCA +GTATCT +GTATCC +GTATCG +GTATGA +GTATGT +GTATGC +GTATGG +GTACAA +GTACAT +GTACAC +GTACAG +GTACTA +GTACTT +GTACTC +GTACTG +GTACCA +GTACCT +GTACCC +GTACCG +GTACGA +GTACGT +GTACGC +GTACGG +GTAGAA +GTAGAT +GTAGAC +GTAGAG +GTAGTA +GTAGTT +GTAGTC +GTAGTG +GTAGCA +GTAGCT +GTAGCC +GTAGCG +GTAGGA +GTAGGT +GTAGGC +GTAGGG +GTTAAA +GTTAAT +GTTAAC +GTTAAG +GTTATA +GTTATT +GTTATC +GTTATG +GTTACA +GTTACT +GTTACC +GTTACG +GTTAGA +GTTAGT +GTTAGC +GTTAGG +GTTTAA +GTTTAT +GTTTAC +GTTTAG +GTTTTA +GTTTTT +GTTTTC +GTTTTG +GTTTCA +GTTTCT +GTTTCC +GTTTCG +GTTTGA +GTTTGT +GTTTGC +GTTTGG +GTTCAA +GTTCAT +GTTCAC +GTTCAG +GTTCTA +GTTCTT +GTTCTC +GTTCTG +GTTCCA +GTTCCT +GTTCCC +GTTCCG +GTTCGA +GTTCGT +GTTCGC +GTTCGG +GTTGAA +GTTGAT +GTTGAC +GTTGAG +GTTGTA +GTTGTT +GTTGTC +GTTGTG +GTTGCA +GTTGCT +GTTGCC +GTTGCG +GTTGGA +GTTGGT +GTTGGC +GTTGGG +GTCAAA +GTCAAT +GTCAAC +GTCAAG +GTCATA +GTCATT +GTCATC +GTCATG +GTCACA +GTCACT +GTCACC +GTCACG +GTCAGA +GTCAGT +GTCAGC +GTCAGG +GTCTAA +GTCTAT +GTCTAC +GTCTAG +GTCTTA +GTCTTT +GTCTTC +GTCTTG +GTCTCA +GTCTCT +GTCTCC +GTCTCG +GTCTGA +GTCTGT +GTCTGC +GTCTGG +GTCCAA +GTCCAT +GTCCAC +GTCCAG +GTCCTA +GTCCTT +GTCCTC +GTCCTG +GTCCCA +GTCCCT +GTCCCC +GTCCCG +GTCCGA +GTCCGT +GTCCGC +GTCCGG +GTCGAA +GTCGAT +GTCGAC +GTCGAG +GTCGTA +GTCGTT +GTCGTC +GTCGTG +GTCGCA +GTCGCT +GTCGCC +GTCGCG +GTCGGA +GTCGGT +GTCGGC +GTCGGG +GTGAAA +GTGAAT +GTGAAC +GTGAAG +GTGATA +GTGATT +GTGATC +GTGATG +GTGACA +GTGACT +GTGACC +GTGACG +GTGAGA +GTGAGT +GTGAGC +GTGAGG +GTGTAA +GTGTAT +GTGTAC +GTGTAG +GTGTTA +GTGTTT +GTGTTC +GTGTTG +GTGTCA +GTGTCT +GTGTCC +GTGTCG +GTGTGA +GTGTGT +GTGTGC +GTGTGG +GTGCAA +GTGCAT +GTGCAC +GTGCAG +GTGCTA +GTGCTT +GTGCTC +GTGCTG +GTGCCA +GTGCCT +GTGCCC +GTGCCG +GTGCGA +GTGCGT +GTGCGC +GTGCGG +GTGGAA +GTGGAT +GTGGAC +GTGGAG +GTGGTA +GTGGTT +GTGGTC +GTGGTG +GTGGCA +GTGGCT +GTGGCC +GTGGCG +GTGGGA +GTGGGT +GTGGGC +GTGGGG +GCAAAA +GCAAAT +GCAAAC +GCAAAG +GCAATA +GCAATT +GCAATC +GCAATG +GCAACA +GCAACT +GCAACC +GCAACG +GCAAGA +GCAAGT +GCAAGC +GCAAGG +GCATAA +GCATAT +GCATAC +GCATAG +GCATTA +GCATTT +GCATTC +GCATTG +GCATCA +GCATCT +GCATCC +GCATCG +GCATGA +GCATGT +GCATGC +GCATGG +GCACAA +GCACAT +GCACAC +GCACAG +GCACTA +GCACTT +GCACTC +GCACTG +GCACCA +GCACCT +GCACCC +GCACCG +GCACGA +GCACGT +GCACGC +GCACGG +GCAGAA +GCAGAT +GCAGAC +GCAGAG +GCAGTA +GCAGTT +GCAGTC +GCAGTG +GCAGCA +GCAGCT +GCAGCC +GCAGCG +GCAGGA +GCAGGT +GCAGGC +GCAGGG +GCTAAA +GCTAAT +GCTAAC +GCTAAG +GCTATA +GCTATT +GCTATC +GCTATG +GCTACA +GCTACT +GCTACC +GCTACG +GCTAGA +GCTAGT +GCTAGC +GCTAGG +GCTTAA +GCTTAT +GCTTAC +GCTTAG +GCTTTA +GCTTTT +GCTTTC +GCTTTG +GCTTCA +GCTTCT +GCTTCC +GCTTCG +GCTTGA +GCTTGT +GCTTGC +GCTTGG +GCTCAA +GCTCAT +GCTCAC +GCTCAG +GCTCTA +GCTCTT +GCTCTC +GCTCTG +GCTCCA +GCTCCT +GCTCCC +GCTCCG +GCTCGA +GCTCGT +GCTCGC +GCTCGG +GCTGAA +GCTGAT +GCTGAC +GCTGAG +GCTGTA +GCTGTT +GCTGTC +GCTGTG +GCTGCA +GCTGCT +GCTGCC +GCTGCG +GCTGGA +GCTGGT +GCTGGC +GCTGGG +GCCAAA +GCCAAT +GCCAAC +GCCAAG +GCCATA +GCCATT +GCCATC +GCCATG +GCCACA +GCCACT +GCCACC +GCCACG +GCCAGA +GCCAGT +GCCAGC +GCCAGG +GCCTAA +GCCTAT +GCCTAC +GCCTAG +GCCTTA +GCCTTT +GCCTTC +GCCTTG +GCCTCA +GCCTCT +GCCTCC +GCCTCG +GCCTGA +GCCTGT +GCCTGC +GCCTGG +GCCCAA +GCCCAT +GCCCAC +GCCCAG +GCCCTA +GCCCTT +GCCCTC +GCCCTG +GCCCCA +GCCCCT +GCCCCC +GCCCCG +GCCCGA +GCCCGT +GCCCGC +GCCCGG +GCCGAA +GCCGAT +GCCGAC +GCCGAG +GCCGTA +GCCGTT +GCCGTC +GCCGTG +GCCGCA +GCCGCT +GCCGCC +GCCGCG +GCCGGA +GCCGGT +GCCGGC +GCCGGG +GCGAAA +GCGAAT +GCGAAC +GCGAAG +GCGATA +GCGATT +GCGATC +GCGATG +GCGACA +GCGACT +GCGACC +GCGACG +GCGAGA +GCGAGT +GCGAGC +GCGAGG +GCGTAA +GCGTAT +GCGTAC +GCGTAG +GCGTTA +GCGTTT +GCGTTC +GCGTTG +GCGTCA +GCGTCT +GCGTCC +GCGTCG +GCGTGA +GCGTGT +GCGTGC +GCGTGG +GCGCAA +GCGCAT +GCGCAC +GCGCAG +GCGCTA +GCGCTT +GCGCTC +GCGCTG +GCGCCA +GCGCCT +GCGCCC +GCGCCG +GCGCGA +GCGCGT +GCGCGC +GCGCGG +GCGGAA +GCGGAT +GCGGAC +GCGGAG +GCGGTA +GCGGTT +GCGGTC +GCGGTG +GCGGCA +GCGGCT +GCGGCC +GCGGCG +GCGGGA +GCGGGT +GCGGGC +GCGGGG +GGAAAA +GGAAAT +GGAAAC +GGAAAG +GGAATA +GGAATT +GGAATC +GGAATG +GGAACA +GGAACT +GGAACC +GGAACG +GGAAGA +GGAAGT +GGAAGC +GGAAGG +GGATAA +GGATAT +GGATAC +GGATAG +GGATTA +GGATTT +GGATTC +GGATTG +GGATCA +GGATCT +GGATCC +GGATCG +GGATGA +GGATGT +GGATGC +GGATGG +GGACAA +GGACAT +GGACAC +GGACAG +GGACTA +GGACTT +GGACTC +GGACTG +GGACCA +GGACCT +GGACCC +GGACCG +GGACGA +GGACGT +GGACGC +GGACGG +GGAGAA +GGAGAT +GGAGAC +GGAGAG +GGAGTA +GGAGTT +GGAGTC +GGAGTG +GGAGCA +GGAGCT +GGAGCC +GGAGCG +GGAGGA +GGAGGT +GGAGGC +GGAGGG +GGTAAA +GGTAAT +GGTAAC +GGTAAG +GGTATA +GGTATT +GGTATC +GGTATG +GGTACA +GGTACT +GGTACC +GGTACG +GGTAGA +GGTAGT +GGTAGC +GGTAGG +GGTTAA +GGTTAT +GGTTAC +GGTTAG +GGTTTA +GGTTTT +GGTTTC +GGTTTG +GGTTCA +GGTTCT +GGTTCC +GGTTCG +GGTTGA +GGTTGT +GGTTGC +GGTTGG +GGTCAA +GGTCAT +GGTCAC +GGTCAG +GGTCTA +GGTCTT +GGTCTC +GGTCTG +GGTCCA +GGTCCT +GGTCCC +GGTCCG +GGTCGA +GGTCGT +GGTCGC +GGTCGG +GGTGAA +GGTGAT +GGTGAC +GGTGAG +GGTGTA +GGTGTT +GGTGTC +GGTGTG +GGTGCA +GGTGCT +GGTGCC +GGTGCG +GGTGGA +GGTGGT +GGTGGC +GGTGGG +GGCAAA +GGCAAT +GGCAAC +GGCAAG +GGCATA +GGCATT +GGCATC +GGCATG +GGCACA +GGCACT +GGCACC +GGCACG +GGCAGA +GGCAGT +GGCAGC +GGCAGG +GGCTAA +GGCTAT +GGCTAC +GGCTAG +GGCTTA +GGCTTT +GGCTTC +GGCTTG +GGCTCA +GGCTCT +GGCTCC +GGCTCG +GGCTGA +GGCTGT +GGCTGC +GGCTGG +GGCCAA +GGCCAT +GGCCAC +GGCCAG +GGCCTA +GGCCTT +GGCCTC +GGCCTG +GGCCCA +GGCCCT +GGCCCC +GGCCCG +GGCCGA +GGCCGT +GGCCGC +GGCCGG +GGCGAA +GGCGAT +GGCGAC +GGCGAG +GGCGTA +GGCGTT +GGCGTC +GGCGTG +GGCGCA +GGCGCT +GGCGCC +GGCGCG +GGCGGA +GGCGGT +GGCGGC +GGCGGG +GGGAAA +GGGAAT +GGGAAC +GGGAAG +GGGATA +GGGATT +GGGATC +GGGATG +GGGACA +GGGACT +GGGACC +GGGACG +GGGAGA +GGGAGT +GGGAGC +GGGAGG +GGGTAA +GGGTAT +GGGTAC +GGGTAG +GGGTTA +GGGTTT +GGGTTC +GGGTTG +GGGTCA +GGGTCT +GGGTCC +GGGTCG +GGGTGA +GGGTGT +GGGTGC +GGGTGG +GGGCAA +GGGCAT +GGGCAC +GGGCAG +GGGCTA +GGGCTT +GGGCTC +GGGCTG +GGGCCA +GGGCCT +GGGCCC +GGGCCG +GGGCGA +GGGCGT +GGGCGC +GGGCGG +GGGGAA +GGGGAT +GGGGAC +GGGGAG +GGGGTA +GGGGTT +GGGGTC +GGGGTG +GGGGCA +GGGGCT +GGGGCC +GGGGCG +GGGGGA +GGGGGT +GGGGGC +GGGGGG diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/README.md b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..500fd976d92b217455f181bc98c77e612d9a18d9 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/README.md @@ -0,0 +1,202 @@ +--- +base_model: fabihamakhdoomi/TinyDNABERT +library_name: peft +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.12.0 \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/adapter_config.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..28c871c93775eedfaf266f00da8c02f96bc02f29 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/adapter_config.json @@ -0,0 +1,42 @@ +{ + "alpha_pattern": {}, + "auto_mapping": null, + "base_model_name_or_path": "fabihamakhdoomi/TinyDNABERT", + "bias": "none", + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 16, + "lora_dropout": 0.2, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": [ + "classifier", + "score" + ], + "peft_type": "LORA", + "r": 4, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "bert.encoder.layer.1.attention.self.query", + "bert.encoder.layer.0.output.dense", + "bert.encoder.layer.1.output.dense", + "bert.encoder.layer.0.intermediate.dense", + "bert.encoder.layer.1.intermediate.dense", + "bert.encoder.layer.0.attention.output.dense", + "bert.encoder.layer.1.attention.output.dense", + "bert.encoder.layer.0.attention.self.query", + "bert.encoder.layer.0.attention.self.value", + "bert.encoder.layer.1.attention.self.key", + "bert.encoder.layer.1.attention.self.value", + "bert.encoder.layer.0.attention.self.key" + ], + "task_type": "SEQ_CLS", + "use_dora": false, + "use_rslora": false +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/adapter_model.safetensors b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/adapter_model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..4a291c70f54d0fcdc188a9a5590318ea5dac1191 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/adapter_model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:121067fe052db1fa34b1d9552edbb81cc33ea802ae5f226a6405670afe9b3583 +size 78208 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/optimizer.pt b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..dbf643f98dcdd60858260e287d450686ed574335 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:622d89bfe7b4ff60acf1c31c998ab6372f178fa0522a26c1c6832c86345dac38 +size 165226 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/rng_state.pth b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..f70e226d95e8a928fcaafa095fe05a0982a8afb3 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fed19b5a5643714b6623a6032bf908ff668a150f2ea673a45992e0ae4072b338 +size 14244 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/scheduler.pt b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..2287c4e5e02bffea6969638a53c3d5402c6dc2b0 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:da119a7607c7d2d96a92ebe57fbaf5fb22ebfce3399b1702b2209f460c0bfe35 +size 1064 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/special_tokens_map.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..a8b3208c2884c4efb86e49300fdd3dc877220cdf --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/special_tokens_map.json @@ -0,0 +1,7 @@ +{ + "cls_token": "[CLS]", + "mask_token": "[MASK]", + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/tokenizer.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..ebd782de6963b9032ce9374cfe30f5275be1a30b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/tokenizer.json @@ -0,0 +1,4265 @@ +{ + "version": "1.0", + "truncation": { + "direction": "Right", + "max_length": 512, + "strategy": "LongestFirst", + "stride": 0 + }, + "padding": { + "strategy": { + "Fixed": 512 + }, + "direction": "Right", + "pad_to_multiple_of": null, + "pad_id": 0, + "pad_type_id": 0, + "pad_token": "[PAD]" + }, + "added_tokens": [ + { + "id": 0, + "content": "[PAD]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 1, + "content": "[UNK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 2, + "content": "[CLS]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 3, + "content": "[SEP]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 4, + "content": "[MASK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + } + ], + "normalizer": { + "type": "BertNormalizer", + "clean_text": true, + "handle_chinese_chars": true, + "strip_accents": null, + "lowercase": false + }, + "pre_tokenizer": { + "type": "BertPreTokenizer" + }, + "post_processor": { + "type": "TemplateProcessing", + "single": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + } + ], + "pair": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "B", + "type_id": 1 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 1 + } + } + ], + "special_tokens": { + "[CLS]": { + "id": "[CLS]", + "ids": [ + 2 + ], + "tokens": [ + "[CLS]" + ] + }, + "[SEP]": { + "id": "[SEP]", + "ids": [ + 3 + ], + "tokens": [ + "[SEP]" + ] + } + } + }, + "decoder": { + "type": "WordPiece", + "prefix": "##", + "cleanup": true + }, + "model": { + "type": "WordPiece", + "unk_token": "[UNK]", + "continuing_subword_prefix": "##", + "max_input_chars_per_word": 100, + "vocab": { + "[PAD]": 0, + "[UNK]": 1, + "[CLS]": 2, + "[SEP]": 3, + "[MASK]": 4, + "AAAAAA": 5, + "AAAAAT": 6, + "AAAAAC": 7, + "AAAAAG": 8, + "AAAATA": 9, + "AAAATT": 10, + "AAAATC": 11, + "AAAATG": 12, + "AAAACA": 13, + "AAAACT": 14, + "AAAACC": 15, + "AAAACG": 16, + "AAAAGA": 17, + "AAAAGT": 18, + "AAAAGC": 19, + "AAAAGG": 20, + "AAATAA": 21, + "AAATAT": 22, + "AAATAC": 23, + "AAATAG": 24, + "AAATTA": 25, + "AAATTT": 26, + "AAATTC": 27, + "AAATTG": 28, + "AAATCA": 29, + "AAATCT": 30, + "AAATCC": 31, + "AAATCG": 32, + "AAATGA": 33, + "AAATGT": 34, + "AAATGC": 35, + "AAATGG": 36, + "AAACAA": 37, + "AAACAT": 38, + "AAACAC": 39, + "AAACAG": 40, + "AAACTA": 41, + "AAACTT": 42, + "AAACTC": 43, + "AAACTG": 44, + "AAACCA": 45, + "AAACCT": 46, + "AAACCC": 47, + "AAACCG": 48, + "AAACGA": 49, + "AAACGT": 50, + "AAACGC": 51, + "AAACGG": 52, + "AAAGAA": 53, + "AAAGAT": 54, + "AAAGAC": 55, + "AAAGAG": 56, + "AAAGTA": 57, + "AAAGTT": 58, + "AAAGTC": 59, + "AAAGTG": 60, + "AAAGCA": 61, + "AAAGCT": 62, + "AAAGCC": 63, + "AAAGCG": 64, + "AAAGGA": 65, + "AAAGGT": 66, + "AAAGGC": 67, + "AAAGGG": 68, + "AATAAA": 69, + "AATAAT": 70, + "AATAAC": 71, + "AATAAG": 72, + "AATATA": 73, + "AATATT": 74, + "AATATC": 75, + "AATATG": 76, + "AATACA": 77, + "AATACT": 78, + "AATACC": 79, + "AATACG": 80, + "AATAGA": 81, + "AATAGT": 82, + "AATAGC": 83, + "AATAGG": 84, + "AATTAA": 85, + "AATTAT": 86, + "AATTAC": 87, + "AATTAG": 88, + "AATTTA": 89, + "AATTTT": 90, + "AATTTC": 91, + "AATTTG": 92, + "AATTCA": 93, + "AATTCT": 94, + "AATTCC": 95, + "AATTCG": 96, + "AATTGA": 97, + "AATTGT": 98, + "AATTGC": 99, + "AATTGG": 100, + "AATCAA": 101, + "AATCAT": 102, + "AATCAC": 103, + "AATCAG": 104, + "AATCTA": 105, + "AATCTT": 106, + "AATCTC": 107, + "AATCTG": 108, + "AATCCA": 109, + "AATCCT": 110, + "AATCCC": 111, + "AATCCG": 112, + "AATCGA": 113, + "AATCGT": 114, + "AATCGC": 115, + "AATCGG": 116, + "AATGAA": 117, + "AATGAT": 118, + "AATGAC": 119, + "AATGAG": 120, + "AATGTA": 121, + "AATGTT": 122, + "AATGTC": 123, + "AATGTG": 124, + "AATGCA": 125, + "AATGCT": 126, + "AATGCC": 127, + "AATGCG": 128, + "AATGGA": 129, + "AATGGT": 130, + "AATGGC": 131, + "AATGGG": 132, + "AACAAA": 133, + "AACAAT": 134, + "AACAAC": 135, + "AACAAG": 136, + "AACATA": 137, + "AACATT": 138, + "AACATC": 139, + "AACATG": 140, + "AACACA": 141, + "AACACT": 142, + "AACACC": 143, + "AACACG": 144, + "AACAGA": 145, + "AACAGT": 146, + "AACAGC": 147, + "AACAGG": 148, + "AACTAA": 149, + "AACTAT": 150, + "AACTAC": 151, + "AACTAG": 152, + "AACTTA": 153, + "AACTTT": 154, + "AACTTC": 155, + "AACTTG": 156, + "AACTCA": 157, + "AACTCT": 158, + "AACTCC": 159, + "AACTCG": 160, + "AACTGA": 161, + "AACTGT": 162, + "AACTGC": 163, + "AACTGG": 164, + "AACCAA": 165, + "AACCAT": 166, + "AACCAC": 167, + "AACCAG": 168, + "AACCTA": 169, + "AACCTT": 170, + "AACCTC": 171, + "AACCTG": 172, + "AACCCA": 173, + "AACCCT": 174, + "AACCCC": 175, + "AACCCG": 176, + "AACCGA": 177, + "AACCGT": 178, + "AACCGC": 179, + "AACCGG": 180, + "AACGAA": 181, + "AACGAT": 182, + "AACGAC": 183, + "AACGAG": 184, + "AACGTA": 185, + "AACGTT": 186, + "AACGTC": 187, + "AACGTG": 188, + "AACGCA": 189, + "AACGCT": 190, + "AACGCC": 191, + "AACGCG": 192, + "AACGGA": 193, + "AACGGT": 194, + "AACGGC": 195, + "AACGGG": 196, + "AAGAAA": 197, + "AAGAAT": 198, + "AAGAAC": 199, + "AAGAAG": 200, + "AAGATA": 201, + "AAGATT": 202, + "AAGATC": 203, + "AAGATG": 204, + "AAGACA": 205, + "AAGACT": 206, + "AAGACC": 207, + "AAGACG": 208, + "AAGAGA": 209, + "AAGAGT": 210, + "AAGAGC": 211, + "AAGAGG": 212, + "AAGTAA": 213, + "AAGTAT": 214, + "AAGTAC": 215, + "AAGTAG": 216, + "AAGTTA": 217, + "AAGTTT": 218, + "AAGTTC": 219, + "AAGTTG": 220, + "AAGTCA": 221, + "AAGTCT": 222, + "AAGTCC": 223, + "AAGTCG": 224, + "AAGTGA": 225, + "AAGTGT": 226, + "AAGTGC": 227, + "AAGTGG": 228, + "AAGCAA": 229, + "AAGCAT": 230, + "AAGCAC": 231, + "AAGCAG": 232, + "AAGCTA": 233, + "AAGCTT": 234, + "AAGCTC": 235, + "AAGCTG": 236, + "AAGCCA": 237, + "AAGCCT": 238, + "AAGCCC": 239, + "AAGCCG": 240, + "AAGCGA": 241, + "AAGCGT": 242, + "AAGCGC": 243, + "AAGCGG": 244, + "AAGGAA": 245, + "AAGGAT": 246, + "AAGGAC": 247, + "AAGGAG": 248, + "AAGGTA": 249, + "AAGGTT": 250, + "AAGGTC": 251, + "AAGGTG": 252, + "AAGGCA": 253, + "AAGGCT": 254, + "AAGGCC": 255, + "AAGGCG": 256, + "AAGGGA": 257, + "AAGGGT": 258, + "AAGGGC": 259, + "AAGGGG": 260, + "ATAAAA": 261, + "ATAAAT": 262, + "ATAAAC": 263, + "ATAAAG": 264, + "ATAATA": 265, + "ATAATT": 266, + "ATAATC": 267, + "ATAATG": 268, + "ATAACA": 269, + "ATAACT": 270, + "ATAACC": 271, + "ATAACG": 272, + "ATAAGA": 273, + "ATAAGT": 274, + "ATAAGC": 275, + "ATAAGG": 276, + "ATATAA": 277, + "ATATAT": 278, + "ATATAC": 279, + "ATATAG": 280, + "ATATTA": 281, + "ATATTT": 282, + "ATATTC": 283, + "ATATTG": 284, + "ATATCA": 285, + "ATATCT": 286, + "ATATCC": 287, + "ATATCG": 288, + "ATATGA": 289, + "ATATGT": 290, + "ATATGC": 291, + "ATATGG": 292, + "ATACAA": 293, + "ATACAT": 294, + "ATACAC": 295, + "ATACAG": 296, + "ATACTA": 297, + "ATACTT": 298, + "ATACTC": 299, + "ATACTG": 300, + "ATACCA": 301, + "ATACCT": 302, + "ATACCC": 303, + "ATACCG": 304, + "ATACGA": 305, + "ATACGT": 306, + "ATACGC": 307, + "ATACGG": 308, + "ATAGAA": 309, + "ATAGAT": 310, + "ATAGAC": 311, + "ATAGAG": 312, + "ATAGTA": 313, + "ATAGTT": 314, + "ATAGTC": 315, + "ATAGTG": 316, + "ATAGCA": 317, + "ATAGCT": 318, + "ATAGCC": 319, + "ATAGCG": 320, + "ATAGGA": 321, + "ATAGGT": 322, + "ATAGGC": 323, + "ATAGGG": 324, + "ATTAAA": 325, + "ATTAAT": 326, + "ATTAAC": 327, + "ATTAAG": 328, + "ATTATA": 329, + "ATTATT": 330, + "ATTATC": 331, + "ATTATG": 332, + "ATTACA": 333, + "ATTACT": 334, + "ATTACC": 335, + "ATTACG": 336, + "ATTAGA": 337, + "ATTAGT": 338, + "ATTAGC": 339, + "ATTAGG": 340, + "ATTTAA": 341, + "ATTTAT": 342, + "ATTTAC": 343, + "ATTTAG": 344, + "ATTTTA": 345, + "ATTTTT": 346, + "ATTTTC": 347, + "ATTTTG": 348, + "ATTTCA": 349, + "ATTTCT": 350, + "ATTTCC": 351, + "ATTTCG": 352, + "ATTTGA": 353, + "ATTTGT": 354, + "ATTTGC": 355, + "ATTTGG": 356, + "ATTCAA": 357, + "ATTCAT": 358, + "ATTCAC": 359, + "ATTCAG": 360, + "ATTCTA": 361, + "ATTCTT": 362, + "ATTCTC": 363, + "ATTCTG": 364, + "ATTCCA": 365, + "ATTCCT": 366, + "ATTCCC": 367, + "ATTCCG": 368, + "ATTCGA": 369, + "ATTCGT": 370, + "ATTCGC": 371, + "ATTCGG": 372, + "ATTGAA": 373, + "ATTGAT": 374, + "ATTGAC": 375, + "ATTGAG": 376, + "ATTGTA": 377, + "ATTGTT": 378, + "ATTGTC": 379, + "ATTGTG": 380, + "ATTGCA": 381, + "ATTGCT": 382, + "ATTGCC": 383, + "ATTGCG": 384, + "ATTGGA": 385, + "ATTGGT": 386, + "ATTGGC": 387, + "ATTGGG": 388, + "ATCAAA": 389, + "ATCAAT": 390, + "ATCAAC": 391, + "ATCAAG": 392, + "ATCATA": 393, + "ATCATT": 394, + "ATCATC": 395, + "ATCATG": 396, + "ATCACA": 397, + "ATCACT": 398, + "ATCACC": 399, + "ATCACG": 400, + "ATCAGA": 401, + "ATCAGT": 402, + "ATCAGC": 403, + "ATCAGG": 404, + "ATCTAA": 405, + "ATCTAT": 406, + "ATCTAC": 407, + "ATCTAG": 408, + "ATCTTA": 409, + "ATCTTT": 410, + "ATCTTC": 411, + "ATCTTG": 412, + "ATCTCA": 413, + "ATCTCT": 414, + "ATCTCC": 415, + "ATCTCG": 416, + "ATCTGA": 417, + "ATCTGT": 418, + "ATCTGC": 419, + "ATCTGG": 420, + "ATCCAA": 421, + "ATCCAT": 422, + "ATCCAC": 423, + "ATCCAG": 424, + "ATCCTA": 425, + "ATCCTT": 426, + "ATCCTC": 427, + "ATCCTG": 428, + "ATCCCA": 429, + "ATCCCT": 430, + "ATCCCC": 431, + "ATCCCG": 432, + "ATCCGA": 433, + "ATCCGT": 434, + "ATCCGC": 435, + "ATCCGG": 436, + "ATCGAA": 437, + "ATCGAT": 438, + "ATCGAC": 439, + "ATCGAG": 440, + "ATCGTA": 441, + "ATCGTT": 442, + "ATCGTC": 443, + "ATCGTG": 444, + "ATCGCA": 445, + "ATCGCT": 446, + "ATCGCC": 447, + "ATCGCG": 448, + "ATCGGA": 449, + "ATCGGT": 450, + "ATCGGC": 451, + "ATCGGG": 452, + "ATGAAA": 453, + "ATGAAT": 454, + "ATGAAC": 455, + "ATGAAG": 456, + "ATGATA": 457, + "ATGATT": 458, + "ATGATC": 459, + "ATGATG": 460, + "ATGACA": 461, + "ATGACT": 462, + "ATGACC": 463, + "ATGACG": 464, + "ATGAGA": 465, + "ATGAGT": 466, + "ATGAGC": 467, + "ATGAGG": 468, + "ATGTAA": 469, + "ATGTAT": 470, + "ATGTAC": 471, + "ATGTAG": 472, + "ATGTTA": 473, + "ATGTTT": 474, + "ATGTTC": 475, + "ATGTTG": 476, + "ATGTCA": 477, + "ATGTCT": 478, + "ATGTCC": 479, + "ATGTCG": 480, + "ATGTGA": 481, + "ATGTGT": 482, + "ATGTGC": 483, + "ATGTGG": 484, + "ATGCAA": 485, + "ATGCAT": 486, + "ATGCAC": 487, + "ATGCAG": 488, + "ATGCTA": 489, + "ATGCTT": 490, + "ATGCTC": 491, + "ATGCTG": 492, + "ATGCCA": 493, + "ATGCCT": 494, + "ATGCCC": 495, + "ATGCCG": 496, + "ATGCGA": 497, + "ATGCGT": 498, + "ATGCGC": 499, + "ATGCGG": 500, + "ATGGAA": 501, + "ATGGAT": 502, + "ATGGAC": 503, + "ATGGAG": 504, + "ATGGTA": 505, + "ATGGTT": 506, + "ATGGTC": 507, + "ATGGTG": 508, + "ATGGCA": 509, + "ATGGCT": 510, + "ATGGCC": 511, + "ATGGCG": 512, + "ATGGGA": 513, + "ATGGGT": 514, + "ATGGGC": 515, + "ATGGGG": 516, + "ACAAAA": 517, + "ACAAAT": 518, + "ACAAAC": 519, + "ACAAAG": 520, + "ACAATA": 521, + "ACAATT": 522, + "ACAATC": 523, + "ACAATG": 524, + "ACAACA": 525, + "ACAACT": 526, + "ACAACC": 527, + "ACAACG": 528, + "ACAAGA": 529, + "ACAAGT": 530, + "ACAAGC": 531, + "ACAAGG": 532, + "ACATAA": 533, + "ACATAT": 534, + "ACATAC": 535, + "ACATAG": 536, + "ACATTA": 537, + "ACATTT": 538, + "ACATTC": 539, + "ACATTG": 540, + "ACATCA": 541, + "ACATCT": 542, + "ACATCC": 543, + "ACATCG": 544, + "ACATGA": 545, + "ACATGT": 546, + "ACATGC": 547, + "ACATGG": 548, + "ACACAA": 549, + "ACACAT": 550, + "ACACAC": 551, + "ACACAG": 552, + "ACACTA": 553, + "ACACTT": 554, + "ACACTC": 555, + "ACACTG": 556, + "ACACCA": 557, + "ACACCT": 558, + "ACACCC": 559, + "ACACCG": 560, + "ACACGA": 561, + "ACACGT": 562, + "ACACGC": 563, + "ACACGG": 564, + "ACAGAA": 565, + "ACAGAT": 566, + "ACAGAC": 567, + "ACAGAG": 568, + "ACAGTA": 569, + "ACAGTT": 570, + "ACAGTC": 571, + "ACAGTG": 572, + "ACAGCA": 573, + "ACAGCT": 574, + "ACAGCC": 575, + "ACAGCG": 576, + "ACAGGA": 577, + "ACAGGT": 578, + "ACAGGC": 579, + "ACAGGG": 580, + "ACTAAA": 581, + "ACTAAT": 582, + "ACTAAC": 583, + "ACTAAG": 584, + "ACTATA": 585, + "ACTATT": 586, + "ACTATC": 587, + "ACTATG": 588, + "ACTACA": 589, + "ACTACT": 590, + "ACTACC": 591, + "ACTACG": 592, + "ACTAGA": 593, + "ACTAGT": 594, + "ACTAGC": 595, + "ACTAGG": 596, + "ACTTAA": 597, + "ACTTAT": 598, + "ACTTAC": 599, + "ACTTAG": 600, + "ACTTTA": 601, + "ACTTTT": 602, + "ACTTTC": 603, + "ACTTTG": 604, + "ACTTCA": 605, + "ACTTCT": 606, + "ACTTCC": 607, + "ACTTCG": 608, + "ACTTGA": 609, + "ACTTGT": 610, + "ACTTGC": 611, + "ACTTGG": 612, + "ACTCAA": 613, + "ACTCAT": 614, + "ACTCAC": 615, + "ACTCAG": 616, + "ACTCTA": 617, + "ACTCTT": 618, + "ACTCTC": 619, + "ACTCTG": 620, + "ACTCCA": 621, + "ACTCCT": 622, + "ACTCCC": 623, + "ACTCCG": 624, + "ACTCGA": 625, + "ACTCGT": 626, + "ACTCGC": 627, + "ACTCGG": 628, + "ACTGAA": 629, + "ACTGAT": 630, + "ACTGAC": 631, + "ACTGAG": 632, + "ACTGTA": 633, + "ACTGTT": 634, + "ACTGTC": 635, + "ACTGTG": 636, + "ACTGCA": 637, + "ACTGCT": 638, + "ACTGCC": 639, + "ACTGCG": 640, + "ACTGGA": 641, + "ACTGGT": 642, + "ACTGGC": 643, + "ACTGGG": 644, + "ACCAAA": 645, + "ACCAAT": 646, + "ACCAAC": 647, + "ACCAAG": 648, + "ACCATA": 649, + "ACCATT": 650, + "ACCATC": 651, + "ACCATG": 652, + "ACCACA": 653, + "ACCACT": 654, + "ACCACC": 655, + "ACCACG": 656, + "ACCAGA": 657, + "ACCAGT": 658, + "ACCAGC": 659, + "ACCAGG": 660, + "ACCTAA": 661, + "ACCTAT": 662, + "ACCTAC": 663, + "ACCTAG": 664, + "ACCTTA": 665, + "ACCTTT": 666, + "ACCTTC": 667, + "ACCTTG": 668, + "ACCTCA": 669, + "ACCTCT": 670, + "ACCTCC": 671, + "ACCTCG": 672, + "ACCTGA": 673, + "ACCTGT": 674, + "ACCTGC": 675, + "ACCTGG": 676, + "ACCCAA": 677, + "ACCCAT": 678, + "ACCCAC": 679, + "ACCCAG": 680, + "ACCCTA": 681, + "ACCCTT": 682, + "ACCCTC": 683, + "ACCCTG": 684, + "ACCCCA": 685, + "ACCCCT": 686, + "ACCCCC": 687, + "ACCCCG": 688, + "ACCCGA": 689, + "ACCCGT": 690, + "ACCCGC": 691, + "ACCCGG": 692, + "ACCGAA": 693, + "ACCGAT": 694, + "ACCGAC": 695, + "ACCGAG": 696, + "ACCGTA": 697, + "ACCGTT": 698, + "ACCGTC": 699, + "ACCGTG": 700, + "ACCGCA": 701, + "ACCGCT": 702, + "ACCGCC": 703, + "ACCGCG": 704, + "ACCGGA": 705, + "ACCGGT": 706, + "ACCGGC": 707, + "ACCGGG": 708, + "ACGAAA": 709, + "ACGAAT": 710, + "ACGAAC": 711, + "ACGAAG": 712, + "ACGATA": 713, + "ACGATT": 714, + "ACGATC": 715, + "ACGATG": 716, + "ACGACA": 717, + "ACGACT": 718, + "ACGACC": 719, + "ACGACG": 720, + "ACGAGA": 721, + "ACGAGT": 722, + "ACGAGC": 723, + "ACGAGG": 724, + "ACGTAA": 725, + "ACGTAT": 726, + "ACGTAC": 727, + "ACGTAG": 728, + "ACGTTA": 729, + "ACGTTT": 730, + "ACGTTC": 731, + "ACGTTG": 732, + "ACGTCA": 733, + "ACGTCT": 734, + "ACGTCC": 735, + "ACGTCG": 736, + "ACGTGA": 737, + "ACGTGT": 738, + "ACGTGC": 739, + "ACGTGG": 740, + "ACGCAA": 741, + "ACGCAT": 742, + "ACGCAC": 743, + "ACGCAG": 744, + "ACGCTA": 745, + "ACGCTT": 746, + "ACGCTC": 747, + "ACGCTG": 748, + "ACGCCA": 749, + "ACGCCT": 750, + "ACGCCC": 751, + "ACGCCG": 752, + "ACGCGA": 753, + "ACGCGT": 754, + "ACGCGC": 755, + "ACGCGG": 756, + "ACGGAA": 757, + "ACGGAT": 758, + "ACGGAC": 759, + "ACGGAG": 760, + "ACGGTA": 761, + "ACGGTT": 762, + "ACGGTC": 763, + "ACGGTG": 764, + "ACGGCA": 765, + "ACGGCT": 766, + "ACGGCC": 767, + "ACGGCG": 768, + "ACGGGA": 769, + "ACGGGT": 770, + "ACGGGC": 771, + "ACGGGG": 772, + "AGAAAA": 773, + "AGAAAT": 774, + "AGAAAC": 775, + "AGAAAG": 776, + "AGAATA": 777, + "AGAATT": 778, + "AGAATC": 779, + "AGAATG": 780, + "AGAACA": 781, + "AGAACT": 782, + "AGAACC": 783, + "AGAACG": 784, + "AGAAGA": 785, + "AGAAGT": 786, + "AGAAGC": 787, + "AGAAGG": 788, + "AGATAA": 789, + "AGATAT": 790, + "AGATAC": 791, + "AGATAG": 792, + "AGATTA": 793, + "AGATTT": 794, + "AGATTC": 795, + "AGATTG": 796, + "AGATCA": 797, + "AGATCT": 798, + "AGATCC": 799, + "AGATCG": 800, + "AGATGA": 801, + "AGATGT": 802, + "AGATGC": 803, + "AGATGG": 804, + "AGACAA": 805, + "AGACAT": 806, + "AGACAC": 807, + "AGACAG": 808, + "AGACTA": 809, + "AGACTT": 810, + "AGACTC": 811, + "AGACTG": 812, + "AGACCA": 813, + "AGACCT": 814, + "AGACCC": 815, + "AGACCG": 816, + "AGACGA": 817, + "AGACGT": 818, + "AGACGC": 819, + "AGACGG": 820, + "AGAGAA": 821, + "AGAGAT": 822, + "AGAGAC": 823, + "AGAGAG": 824, + "AGAGTA": 825, + "AGAGTT": 826, + "AGAGTC": 827, + "AGAGTG": 828, + "AGAGCA": 829, + "AGAGCT": 830, + "AGAGCC": 831, + "AGAGCG": 832, + "AGAGGA": 833, + "AGAGGT": 834, + "AGAGGC": 835, + "AGAGGG": 836, + "AGTAAA": 837, + "AGTAAT": 838, + "AGTAAC": 839, + "AGTAAG": 840, + "AGTATA": 841, + "AGTATT": 842, + "AGTATC": 843, + "AGTATG": 844, + "AGTACA": 845, + "AGTACT": 846, + "AGTACC": 847, + "AGTACG": 848, + "AGTAGA": 849, + "AGTAGT": 850, + "AGTAGC": 851, + "AGTAGG": 852, + "AGTTAA": 853, + "AGTTAT": 854, + "AGTTAC": 855, + "AGTTAG": 856, + "AGTTTA": 857, + "AGTTTT": 858, + "AGTTTC": 859, + "AGTTTG": 860, + "AGTTCA": 861, + "AGTTCT": 862, + "AGTTCC": 863, + "AGTTCG": 864, + "AGTTGA": 865, + "AGTTGT": 866, + "AGTTGC": 867, + "AGTTGG": 868, + "AGTCAA": 869, + "AGTCAT": 870, + "AGTCAC": 871, + "AGTCAG": 872, + "AGTCTA": 873, + "AGTCTT": 874, + "AGTCTC": 875, + "AGTCTG": 876, + "AGTCCA": 877, + "AGTCCT": 878, + "AGTCCC": 879, + "AGTCCG": 880, + "AGTCGA": 881, + "AGTCGT": 882, + "AGTCGC": 883, + "AGTCGG": 884, + "AGTGAA": 885, + "AGTGAT": 886, + "AGTGAC": 887, + "AGTGAG": 888, + "AGTGTA": 889, + "AGTGTT": 890, + "AGTGTC": 891, + "AGTGTG": 892, + "AGTGCA": 893, + "AGTGCT": 894, + "AGTGCC": 895, + "AGTGCG": 896, + "AGTGGA": 897, + "AGTGGT": 898, + "AGTGGC": 899, + "AGTGGG": 900, + "AGCAAA": 901, + "AGCAAT": 902, + "AGCAAC": 903, + "AGCAAG": 904, + "AGCATA": 905, + "AGCATT": 906, + "AGCATC": 907, + "AGCATG": 908, + "AGCACA": 909, + "AGCACT": 910, + "AGCACC": 911, + "AGCACG": 912, + "AGCAGA": 913, + "AGCAGT": 914, + "AGCAGC": 915, + "AGCAGG": 916, + "AGCTAA": 917, + "AGCTAT": 918, + "AGCTAC": 919, + "AGCTAG": 920, + "AGCTTA": 921, + "AGCTTT": 922, + "AGCTTC": 923, + "AGCTTG": 924, + "AGCTCA": 925, + "AGCTCT": 926, + "AGCTCC": 927, + "AGCTCG": 928, + "AGCTGA": 929, + "AGCTGT": 930, + "AGCTGC": 931, + "AGCTGG": 932, + "AGCCAA": 933, + "AGCCAT": 934, + "AGCCAC": 935, + "AGCCAG": 936, + "AGCCTA": 937, + "AGCCTT": 938, + "AGCCTC": 939, + "AGCCTG": 940, + "AGCCCA": 941, + "AGCCCT": 942, + "AGCCCC": 943, + "AGCCCG": 944, + "AGCCGA": 945, + "AGCCGT": 946, + "AGCCGC": 947, + "AGCCGG": 948, + "AGCGAA": 949, + "AGCGAT": 950, + "AGCGAC": 951, + "AGCGAG": 952, + "AGCGTA": 953, + "AGCGTT": 954, + "AGCGTC": 955, + "AGCGTG": 956, + "AGCGCA": 957, + "AGCGCT": 958, + "AGCGCC": 959, + "AGCGCG": 960, + "AGCGGA": 961, + "AGCGGT": 962, + "AGCGGC": 963, + "AGCGGG": 964, + "AGGAAA": 965, + "AGGAAT": 966, + "AGGAAC": 967, + "AGGAAG": 968, + "AGGATA": 969, + "AGGATT": 970, + "AGGATC": 971, + "AGGATG": 972, + "AGGACA": 973, + "AGGACT": 974, + "AGGACC": 975, + "AGGACG": 976, + "AGGAGA": 977, + "AGGAGT": 978, + "AGGAGC": 979, + "AGGAGG": 980, + "AGGTAA": 981, + "AGGTAT": 982, + "AGGTAC": 983, + "AGGTAG": 984, + "AGGTTA": 985, + "AGGTTT": 986, + "AGGTTC": 987, + "AGGTTG": 988, + "AGGTCA": 989, + "AGGTCT": 990, + "AGGTCC": 991, + "AGGTCG": 992, + "AGGTGA": 993, + "AGGTGT": 994, + "AGGTGC": 995, + "AGGTGG": 996, + "AGGCAA": 997, + "AGGCAT": 998, + "AGGCAC": 999, + "AGGCAG": 1000, + "AGGCTA": 1001, + "AGGCTT": 1002, + "AGGCTC": 1003, + "AGGCTG": 1004, + "AGGCCA": 1005, + "AGGCCT": 1006, + "AGGCCC": 1007, + "AGGCCG": 1008, + "AGGCGA": 1009, + "AGGCGT": 1010, + "AGGCGC": 1011, + "AGGCGG": 1012, + "AGGGAA": 1013, + "AGGGAT": 1014, + "AGGGAC": 1015, + "AGGGAG": 1016, + "AGGGTA": 1017, + "AGGGTT": 1018, + "AGGGTC": 1019, + "AGGGTG": 1020, + "AGGGCA": 1021, + "AGGGCT": 1022, + "AGGGCC": 1023, + "AGGGCG": 1024, + "AGGGGA": 1025, + "AGGGGT": 1026, + "AGGGGC": 1027, + "AGGGGG": 1028, + "TAAAAA": 1029, + "TAAAAT": 1030, + "TAAAAC": 1031, + "TAAAAG": 1032, + "TAAATA": 1033, + "TAAATT": 1034, + "TAAATC": 1035, + "TAAATG": 1036, + "TAAACA": 1037, + "TAAACT": 1038, + "TAAACC": 1039, + "TAAACG": 1040, + "TAAAGA": 1041, + "TAAAGT": 1042, + "TAAAGC": 1043, + "TAAAGG": 1044, + "TAATAA": 1045, + "TAATAT": 1046, + "TAATAC": 1047, + "TAATAG": 1048, + "TAATTA": 1049, + "TAATTT": 1050, + "TAATTC": 1051, + "TAATTG": 1052, + "TAATCA": 1053, + "TAATCT": 1054, + "TAATCC": 1055, + "TAATCG": 1056, + "TAATGA": 1057, + "TAATGT": 1058, + "TAATGC": 1059, + "TAATGG": 1060, + "TAACAA": 1061, + "TAACAT": 1062, + "TAACAC": 1063, + "TAACAG": 1064, + "TAACTA": 1065, + "TAACTT": 1066, + "TAACTC": 1067, + "TAACTG": 1068, + "TAACCA": 1069, + "TAACCT": 1070, + "TAACCC": 1071, + "TAACCG": 1072, + "TAACGA": 1073, + "TAACGT": 1074, + "TAACGC": 1075, + "TAACGG": 1076, + "TAAGAA": 1077, + "TAAGAT": 1078, + "TAAGAC": 1079, + "TAAGAG": 1080, + "TAAGTA": 1081, + "TAAGTT": 1082, + "TAAGTC": 1083, + "TAAGTG": 1084, + "TAAGCA": 1085, + "TAAGCT": 1086, + "TAAGCC": 1087, + "TAAGCG": 1088, + "TAAGGA": 1089, + "TAAGGT": 1090, + "TAAGGC": 1091, + "TAAGGG": 1092, + "TATAAA": 1093, + "TATAAT": 1094, + "TATAAC": 1095, + "TATAAG": 1096, + "TATATA": 1097, + "TATATT": 1098, + "TATATC": 1099, + "TATATG": 1100, + "TATACA": 1101, + "TATACT": 1102, + "TATACC": 1103, + "TATACG": 1104, + "TATAGA": 1105, + "TATAGT": 1106, + "TATAGC": 1107, + "TATAGG": 1108, + "TATTAA": 1109, + "TATTAT": 1110, + "TATTAC": 1111, + "TATTAG": 1112, + "TATTTA": 1113, + "TATTTT": 1114, + "TATTTC": 1115, + "TATTTG": 1116, + "TATTCA": 1117, + "TATTCT": 1118, + "TATTCC": 1119, + "TATTCG": 1120, + "TATTGA": 1121, + "TATTGT": 1122, + "TATTGC": 1123, + "TATTGG": 1124, + "TATCAA": 1125, + "TATCAT": 1126, + "TATCAC": 1127, + "TATCAG": 1128, + "TATCTA": 1129, + "TATCTT": 1130, + "TATCTC": 1131, + "TATCTG": 1132, + "TATCCA": 1133, + "TATCCT": 1134, + "TATCCC": 1135, + "TATCCG": 1136, + "TATCGA": 1137, + "TATCGT": 1138, + "TATCGC": 1139, + "TATCGG": 1140, + "TATGAA": 1141, + "TATGAT": 1142, + "TATGAC": 1143, + "TATGAG": 1144, + "TATGTA": 1145, + "TATGTT": 1146, + "TATGTC": 1147, + "TATGTG": 1148, + "TATGCA": 1149, + "TATGCT": 1150, + "TATGCC": 1151, + "TATGCG": 1152, + "TATGGA": 1153, + "TATGGT": 1154, + "TATGGC": 1155, + "TATGGG": 1156, + "TACAAA": 1157, + "TACAAT": 1158, + "TACAAC": 1159, + "TACAAG": 1160, + "TACATA": 1161, + "TACATT": 1162, + "TACATC": 1163, + "TACATG": 1164, + "TACACA": 1165, + "TACACT": 1166, + "TACACC": 1167, + "TACACG": 1168, + "TACAGA": 1169, + "TACAGT": 1170, + "TACAGC": 1171, + "TACAGG": 1172, + "TACTAA": 1173, + "TACTAT": 1174, + "TACTAC": 1175, + "TACTAG": 1176, + "TACTTA": 1177, + "TACTTT": 1178, + "TACTTC": 1179, + "TACTTG": 1180, + "TACTCA": 1181, + "TACTCT": 1182, + "TACTCC": 1183, + "TACTCG": 1184, + "TACTGA": 1185, + "TACTGT": 1186, + "TACTGC": 1187, + "TACTGG": 1188, + "TACCAA": 1189, + "TACCAT": 1190, + "TACCAC": 1191, + "TACCAG": 1192, + "TACCTA": 1193, + "TACCTT": 1194, + "TACCTC": 1195, + "TACCTG": 1196, + "TACCCA": 1197, + "TACCCT": 1198, + "TACCCC": 1199, + "TACCCG": 1200, + "TACCGA": 1201, + "TACCGT": 1202, + "TACCGC": 1203, + "TACCGG": 1204, + "TACGAA": 1205, + "TACGAT": 1206, + "TACGAC": 1207, + "TACGAG": 1208, + "TACGTA": 1209, + "TACGTT": 1210, + "TACGTC": 1211, + "TACGTG": 1212, + "TACGCA": 1213, + "TACGCT": 1214, + "TACGCC": 1215, + "TACGCG": 1216, + "TACGGA": 1217, + "TACGGT": 1218, + "TACGGC": 1219, + "TACGGG": 1220, + "TAGAAA": 1221, + "TAGAAT": 1222, + "TAGAAC": 1223, + "TAGAAG": 1224, + "TAGATA": 1225, + "TAGATT": 1226, + "TAGATC": 1227, + "TAGATG": 1228, + "TAGACA": 1229, + "TAGACT": 1230, + "TAGACC": 1231, + "TAGACG": 1232, + "TAGAGA": 1233, + "TAGAGT": 1234, + "TAGAGC": 1235, + "TAGAGG": 1236, + "TAGTAA": 1237, + "TAGTAT": 1238, + "TAGTAC": 1239, + "TAGTAG": 1240, + "TAGTTA": 1241, + "TAGTTT": 1242, + "TAGTTC": 1243, + "TAGTTG": 1244, + "TAGTCA": 1245, + "TAGTCT": 1246, + "TAGTCC": 1247, + "TAGTCG": 1248, + "TAGTGA": 1249, + "TAGTGT": 1250, + "TAGTGC": 1251, + "TAGTGG": 1252, + "TAGCAA": 1253, + "TAGCAT": 1254, + "TAGCAC": 1255, + "TAGCAG": 1256, + "TAGCTA": 1257, + "TAGCTT": 1258, + "TAGCTC": 1259, + "TAGCTG": 1260, + "TAGCCA": 1261, + "TAGCCT": 1262, + "TAGCCC": 1263, + "TAGCCG": 1264, + "TAGCGA": 1265, + "TAGCGT": 1266, + "TAGCGC": 1267, + "TAGCGG": 1268, + "TAGGAA": 1269, + "TAGGAT": 1270, + "TAGGAC": 1271, + "TAGGAG": 1272, + "TAGGTA": 1273, + "TAGGTT": 1274, + "TAGGTC": 1275, + "TAGGTG": 1276, + "TAGGCA": 1277, + "TAGGCT": 1278, + "TAGGCC": 1279, + "TAGGCG": 1280, + "TAGGGA": 1281, + "TAGGGT": 1282, + "TAGGGC": 1283, + "TAGGGG": 1284, + "TTAAAA": 1285, + "TTAAAT": 1286, + "TTAAAC": 1287, + "TTAAAG": 1288, + "TTAATA": 1289, + "TTAATT": 1290, + "TTAATC": 1291, + "TTAATG": 1292, + "TTAACA": 1293, + "TTAACT": 1294, + "TTAACC": 1295, + "TTAACG": 1296, + "TTAAGA": 1297, + "TTAAGT": 1298, + "TTAAGC": 1299, + "TTAAGG": 1300, + "TTATAA": 1301, + "TTATAT": 1302, + "TTATAC": 1303, + "TTATAG": 1304, + "TTATTA": 1305, + "TTATTT": 1306, + "TTATTC": 1307, + "TTATTG": 1308, + "TTATCA": 1309, + "TTATCT": 1310, + "TTATCC": 1311, + "TTATCG": 1312, + "TTATGA": 1313, + "TTATGT": 1314, + "TTATGC": 1315, + "TTATGG": 1316, + "TTACAA": 1317, + "TTACAT": 1318, + "TTACAC": 1319, + "TTACAG": 1320, + "TTACTA": 1321, + "TTACTT": 1322, + "TTACTC": 1323, + "TTACTG": 1324, + "TTACCA": 1325, + "TTACCT": 1326, + "TTACCC": 1327, + "TTACCG": 1328, + "TTACGA": 1329, + "TTACGT": 1330, + "TTACGC": 1331, + "TTACGG": 1332, + "TTAGAA": 1333, + "TTAGAT": 1334, + "TTAGAC": 1335, + "TTAGAG": 1336, + "TTAGTA": 1337, + "TTAGTT": 1338, + "TTAGTC": 1339, + "TTAGTG": 1340, + "TTAGCA": 1341, + "TTAGCT": 1342, + "TTAGCC": 1343, + "TTAGCG": 1344, + "TTAGGA": 1345, + "TTAGGT": 1346, + "TTAGGC": 1347, + "TTAGGG": 1348, + "TTTAAA": 1349, + "TTTAAT": 1350, + "TTTAAC": 1351, + "TTTAAG": 1352, + "TTTATA": 1353, + "TTTATT": 1354, + "TTTATC": 1355, + "TTTATG": 1356, + "TTTACA": 1357, + "TTTACT": 1358, + "TTTACC": 1359, + "TTTACG": 1360, + "TTTAGA": 1361, + "TTTAGT": 1362, + "TTTAGC": 1363, + "TTTAGG": 1364, + "TTTTAA": 1365, + "TTTTAT": 1366, + "TTTTAC": 1367, + "TTTTAG": 1368, + "TTTTTA": 1369, + "TTTTTT": 1370, + "TTTTTC": 1371, + "TTTTTG": 1372, + "TTTTCA": 1373, + "TTTTCT": 1374, + "TTTTCC": 1375, + "TTTTCG": 1376, + "TTTTGA": 1377, + "TTTTGT": 1378, + "TTTTGC": 1379, + "TTTTGG": 1380, + "TTTCAA": 1381, + "TTTCAT": 1382, + "TTTCAC": 1383, + "TTTCAG": 1384, + "TTTCTA": 1385, + "TTTCTT": 1386, + "TTTCTC": 1387, + "TTTCTG": 1388, + "TTTCCA": 1389, + "TTTCCT": 1390, + "TTTCCC": 1391, + "TTTCCG": 1392, + "TTTCGA": 1393, + "TTTCGT": 1394, + "TTTCGC": 1395, + "TTTCGG": 1396, + "TTTGAA": 1397, + "TTTGAT": 1398, + "TTTGAC": 1399, + "TTTGAG": 1400, + "TTTGTA": 1401, + "TTTGTT": 1402, + "TTTGTC": 1403, + "TTTGTG": 1404, + "TTTGCA": 1405, + "TTTGCT": 1406, + "TTTGCC": 1407, + "TTTGCG": 1408, + "TTTGGA": 1409, + "TTTGGT": 1410, + "TTTGGC": 1411, + "TTTGGG": 1412, + "TTCAAA": 1413, + "TTCAAT": 1414, + "TTCAAC": 1415, + "TTCAAG": 1416, + "TTCATA": 1417, + "TTCATT": 1418, + "TTCATC": 1419, + "TTCATG": 1420, + "TTCACA": 1421, + "TTCACT": 1422, + "TTCACC": 1423, + "TTCACG": 1424, + "TTCAGA": 1425, + "TTCAGT": 1426, + "TTCAGC": 1427, + "TTCAGG": 1428, + "TTCTAA": 1429, + "TTCTAT": 1430, + "TTCTAC": 1431, + "TTCTAG": 1432, + "TTCTTA": 1433, + "TTCTTT": 1434, + "TTCTTC": 1435, + "TTCTTG": 1436, + "TTCTCA": 1437, + "TTCTCT": 1438, + "TTCTCC": 1439, + "TTCTCG": 1440, + "TTCTGA": 1441, + "TTCTGT": 1442, + "TTCTGC": 1443, + "TTCTGG": 1444, + "TTCCAA": 1445, + "TTCCAT": 1446, + "TTCCAC": 1447, + "TTCCAG": 1448, + "TTCCTA": 1449, + "TTCCTT": 1450, + "TTCCTC": 1451, + "TTCCTG": 1452, + "TTCCCA": 1453, + "TTCCCT": 1454, + "TTCCCC": 1455, + "TTCCCG": 1456, + "TTCCGA": 1457, + "TTCCGT": 1458, + "TTCCGC": 1459, + "TTCCGG": 1460, + "TTCGAA": 1461, + "TTCGAT": 1462, + "TTCGAC": 1463, + "TTCGAG": 1464, + "TTCGTA": 1465, + "TTCGTT": 1466, + "TTCGTC": 1467, + "TTCGTG": 1468, + "TTCGCA": 1469, + "TTCGCT": 1470, + "TTCGCC": 1471, + "TTCGCG": 1472, + "TTCGGA": 1473, + "TTCGGT": 1474, + "TTCGGC": 1475, + "TTCGGG": 1476, + "TTGAAA": 1477, + "TTGAAT": 1478, + "TTGAAC": 1479, + "TTGAAG": 1480, + "TTGATA": 1481, + "TTGATT": 1482, + "TTGATC": 1483, + "TTGATG": 1484, + "TTGACA": 1485, + "TTGACT": 1486, + "TTGACC": 1487, + "TTGACG": 1488, + "TTGAGA": 1489, + "TTGAGT": 1490, + "TTGAGC": 1491, + "TTGAGG": 1492, + "TTGTAA": 1493, + "TTGTAT": 1494, + "TTGTAC": 1495, + "TTGTAG": 1496, + "TTGTTA": 1497, + "TTGTTT": 1498, + "TTGTTC": 1499, + "TTGTTG": 1500, + "TTGTCA": 1501, + "TTGTCT": 1502, + "TTGTCC": 1503, + "TTGTCG": 1504, + "TTGTGA": 1505, + "TTGTGT": 1506, + "TTGTGC": 1507, + "TTGTGG": 1508, + "TTGCAA": 1509, + "TTGCAT": 1510, + "TTGCAC": 1511, + "TTGCAG": 1512, + "TTGCTA": 1513, + "TTGCTT": 1514, + "TTGCTC": 1515, + "TTGCTG": 1516, + "TTGCCA": 1517, + "TTGCCT": 1518, + "TTGCCC": 1519, + "TTGCCG": 1520, + "TTGCGA": 1521, + "TTGCGT": 1522, + "TTGCGC": 1523, + "TTGCGG": 1524, + "TTGGAA": 1525, + "TTGGAT": 1526, + "TTGGAC": 1527, + "TTGGAG": 1528, + "TTGGTA": 1529, + "TTGGTT": 1530, + "TTGGTC": 1531, + "TTGGTG": 1532, + "TTGGCA": 1533, + "TTGGCT": 1534, + "TTGGCC": 1535, + "TTGGCG": 1536, + "TTGGGA": 1537, + "TTGGGT": 1538, + "TTGGGC": 1539, + "TTGGGG": 1540, + "TCAAAA": 1541, + "TCAAAT": 1542, + "TCAAAC": 1543, + "TCAAAG": 1544, + "TCAATA": 1545, + "TCAATT": 1546, + "TCAATC": 1547, + "TCAATG": 1548, + "TCAACA": 1549, + "TCAACT": 1550, + "TCAACC": 1551, + "TCAACG": 1552, + "TCAAGA": 1553, + "TCAAGT": 1554, + "TCAAGC": 1555, + "TCAAGG": 1556, + "TCATAA": 1557, + "TCATAT": 1558, + "TCATAC": 1559, + "TCATAG": 1560, + "TCATTA": 1561, + "TCATTT": 1562, + "TCATTC": 1563, + "TCATTG": 1564, + "TCATCA": 1565, + "TCATCT": 1566, + "TCATCC": 1567, + "TCATCG": 1568, + "TCATGA": 1569, + "TCATGT": 1570, + "TCATGC": 1571, + "TCATGG": 1572, + "TCACAA": 1573, + "TCACAT": 1574, + "TCACAC": 1575, + "TCACAG": 1576, + "TCACTA": 1577, + "TCACTT": 1578, + "TCACTC": 1579, + "TCACTG": 1580, + "TCACCA": 1581, + "TCACCT": 1582, + "TCACCC": 1583, + "TCACCG": 1584, + "TCACGA": 1585, + "TCACGT": 1586, + "TCACGC": 1587, + "TCACGG": 1588, + "TCAGAA": 1589, + "TCAGAT": 1590, + "TCAGAC": 1591, + "TCAGAG": 1592, + "TCAGTA": 1593, + "TCAGTT": 1594, + "TCAGTC": 1595, + "TCAGTG": 1596, + "TCAGCA": 1597, + "TCAGCT": 1598, + "TCAGCC": 1599, + "TCAGCG": 1600, + "TCAGGA": 1601, + "TCAGGT": 1602, + "TCAGGC": 1603, + "TCAGGG": 1604, + "TCTAAA": 1605, + "TCTAAT": 1606, + "TCTAAC": 1607, + "TCTAAG": 1608, + "TCTATA": 1609, + "TCTATT": 1610, + "TCTATC": 1611, + "TCTATG": 1612, + "TCTACA": 1613, + "TCTACT": 1614, + "TCTACC": 1615, + "TCTACG": 1616, + "TCTAGA": 1617, + "TCTAGT": 1618, + "TCTAGC": 1619, + "TCTAGG": 1620, + "TCTTAA": 1621, + "TCTTAT": 1622, + "TCTTAC": 1623, + "TCTTAG": 1624, + "TCTTTA": 1625, + "TCTTTT": 1626, + "TCTTTC": 1627, + "TCTTTG": 1628, + "TCTTCA": 1629, + "TCTTCT": 1630, + "TCTTCC": 1631, + "TCTTCG": 1632, + "TCTTGA": 1633, + "TCTTGT": 1634, + "TCTTGC": 1635, + "TCTTGG": 1636, + "TCTCAA": 1637, + "TCTCAT": 1638, + "TCTCAC": 1639, + "TCTCAG": 1640, + "TCTCTA": 1641, + "TCTCTT": 1642, + "TCTCTC": 1643, + "TCTCTG": 1644, + "TCTCCA": 1645, + "TCTCCT": 1646, + "TCTCCC": 1647, + "TCTCCG": 1648, + "TCTCGA": 1649, + "TCTCGT": 1650, + "TCTCGC": 1651, + "TCTCGG": 1652, + "TCTGAA": 1653, + "TCTGAT": 1654, + "TCTGAC": 1655, + "TCTGAG": 1656, + "TCTGTA": 1657, + "TCTGTT": 1658, + "TCTGTC": 1659, + "TCTGTG": 1660, + "TCTGCA": 1661, + "TCTGCT": 1662, + "TCTGCC": 1663, + "TCTGCG": 1664, + "TCTGGA": 1665, + "TCTGGT": 1666, + "TCTGGC": 1667, + "TCTGGG": 1668, + "TCCAAA": 1669, + "TCCAAT": 1670, + "TCCAAC": 1671, + "TCCAAG": 1672, + "TCCATA": 1673, + "TCCATT": 1674, + "TCCATC": 1675, + "TCCATG": 1676, + "TCCACA": 1677, + "TCCACT": 1678, + "TCCACC": 1679, + "TCCACG": 1680, + "TCCAGA": 1681, + "TCCAGT": 1682, + "TCCAGC": 1683, + "TCCAGG": 1684, + "TCCTAA": 1685, + "TCCTAT": 1686, + "TCCTAC": 1687, + "TCCTAG": 1688, + "TCCTTA": 1689, + "TCCTTT": 1690, + "TCCTTC": 1691, + "TCCTTG": 1692, + "TCCTCA": 1693, + "TCCTCT": 1694, + "TCCTCC": 1695, + "TCCTCG": 1696, + "TCCTGA": 1697, + "TCCTGT": 1698, + "TCCTGC": 1699, + "TCCTGG": 1700, + "TCCCAA": 1701, + "TCCCAT": 1702, + "TCCCAC": 1703, + "TCCCAG": 1704, + "TCCCTA": 1705, + "TCCCTT": 1706, + "TCCCTC": 1707, + "TCCCTG": 1708, + "TCCCCA": 1709, + "TCCCCT": 1710, + "TCCCCC": 1711, + "TCCCCG": 1712, + "TCCCGA": 1713, + "TCCCGT": 1714, + "TCCCGC": 1715, + "TCCCGG": 1716, + "TCCGAA": 1717, + "TCCGAT": 1718, + "TCCGAC": 1719, + "TCCGAG": 1720, + "TCCGTA": 1721, + "TCCGTT": 1722, + "TCCGTC": 1723, + "TCCGTG": 1724, + "TCCGCA": 1725, + "TCCGCT": 1726, + "TCCGCC": 1727, + "TCCGCG": 1728, + "TCCGGA": 1729, + "TCCGGT": 1730, + "TCCGGC": 1731, + "TCCGGG": 1732, + "TCGAAA": 1733, + "TCGAAT": 1734, + "TCGAAC": 1735, + "TCGAAG": 1736, + "TCGATA": 1737, + "TCGATT": 1738, + "TCGATC": 1739, + "TCGATG": 1740, + "TCGACA": 1741, + "TCGACT": 1742, + "TCGACC": 1743, + "TCGACG": 1744, + "TCGAGA": 1745, + "TCGAGT": 1746, + "TCGAGC": 1747, + "TCGAGG": 1748, + "TCGTAA": 1749, + "TCGTAT": 1750, + "TCGTAC": 1751, + "TCGTAG": 1752, + "TCGTTA": 1753, + "TCGTTT": 1754, + "TCGTTC": 1755, + "TCGTTG": 1756, + "TCGTCA": 1757, + "TCGTCT": 1758, + "TCGTCC": 1759, + "TCGTCG": 1760, + "TCGTGA": 1761, + "TCGTGT": 1762, + "TCGTGC": 1763, + "TCGTGG": 1764, + "TCGCAA": 1765, + "TCGCAT": 1766, + "TCGCAC": 1767, + "TCGCAG": 1768, + "TCGCTA": 1769, + "TCGCTT": 1770, + "TCGCTC": 1771, + "TCGCTG": 1772, + "TCGCCA": 1773, + "TCGCCT": 1774, + "TCGCCC": 1775, + "TCGCCG": 1776, + "TCGCGA": 1777, + "TCGCGT": 1778, + "TCGCGC": 1779, + "TCGCGG": 1780, + "TCGGAA": 1781, + "TCGGAT": 1782, + "TCGGAC": 1783, + "TCGGAG": 1784, + "TCGGTA": 1785, + "TCGGTT": 1786, + "TCGGTC": 1787, + "TCGGTG": 1788, + "TCGGCA": 1789, + "TCGGCT": 1790, + "TCGGCC": 1791, + "TCGGCG": 1792, + "TCGGGA": 1793, + "TCGGGT": 1794, + "TCGGGC": 1795, + "TCGGGG": 1796, + "TGAAAA": 1797, + "TGAAAT": 1798, + "TGAAAC": 1799, + "TGAAAG": 1800, + "TGAATA": 1801, + "TGAATT": 1802, + "TGAATC": 1803, + "TGAATG": 1804, + "TGAACA": 1805, + "TGAACT": 1806, + "TGAACC": 1807, + "TGAACG": 1808, + "TGAAGA": 1809, + "TGAAGT": 1810, + "TGAAGC": 1811, + "TGAAGG": 1812, + "TGATAA": 1813, + "TGATAT": 1814, + "TGATAC": 1815, + "TGATAG": 1816, + "TGATTA": 1817, + "TGATTT": 1818, + "TGATTC": 1819, + "TGATTG": 1820, + "TGATCA": 1821, + "TGATCT": 1822, + "TGATCC": 1823, + "TGATCG": 1824, + "TGATGA": 1825, + "TGATGT": 1826, + "TGATGC": 1827, + "TGATGG": 1828, + "TGACAA": 1829, + "TGACAT": 1830, + "TGACAC": 1831, + "TGACAG": 1832, + "TGACTA": 1833, + "TGACTT": 1834, + "TGACTC": 1835, + "TGACTG": 1836, + "TGACCA": 1837, + "TGACCT": 1838, + "TGACCC": 1839, + "TGACCG": 1840, + "TGACGA": 1841, + "TGACGT": 1842, + "TGACGC": 1843, + "TGACGG": 1844, + "TGAGAA": 1845, + "TGAGAT": 1846, + "TGAGAC": 1847, + "TGAGAG": 1848, + "TGAGTA": 1849, + "TGAGTT": 1850, + "TGAGTC": 1851, + "TGAGTG": 1852, + "TGAGCA": 1853, + "TGAGCT": 1854, + "TGAGCC": 1855, + "TGAGCG": 1856, + "TGAGGA": 1857, + "TGAGGT": 1858, + "TGAGGC": 1859, + "TGAGGG": 1860, + "TGTAAA": 1861, + "TGTAAT": 1862, + "TGTAAC": 1863, + "TGTAAG": 1864, + "TGTATA": 1865, + "TGTATT": 1866, + "TGTATC": 1867, + "TGTATG": 1868, + "TGTACA": 1869, + "TGTACT": 1870, + "TGTACC": 1871, + "TGTACG": 1872, + "TGTAGA": 1873, + "TGTAGT": 1874, + "TGTAGC": 1875, + "TGTAGG": 1876, + "TGTTAA": 1877, + "TGTTAT": 1878, + "TGTTAC": 1879, + "TGTTAG": 1880, + "TGTTTA": 1881, + "TGTTTT": 1882, + "TGTTTC": 1883, + "TGTTTG": 1884, + "TGTTCA": 1885, + "TGTTCT": 1886, + "TGTTCC": 1887, + "TGTTCG": 1888, + "TGTTGA": 1889, + "TGTTGT": 1890, + "TGTTGC": 1891, + "TGTTGG": 1892, + "TGTCAA": 1893, + "TGTCAT": 1894, + "TGTCAC": 1895, + "TGTCAG": 1896, + "TGTCTA": 1897, + "TGTCTT": 1898, + "TGTCTC": 1899, + "TGTCTG": 1900, + "TGTCCA": 1901, + "TGTCCT": 1902, + "TGTCCC": 1903, + "TGTCCG": 1904, + "TGTCGA": 1905, + "TGTCGT": 1906, + "TGTCGC": 1907, + "TGTCGG": 1908, + "TGTGAA": 1909, + "TGTGAT": 1910, + "TGTGAC": 1911, + "TGTGAG": 1912, + "TGTGTA": 1913, + "TGTGTT": 1914, + "TGTGTC": 1915, + "TGTGTG": 1916, + "TGTGCA": 1917, + "TGTGCT": 1918, + "TGTGCC": 1919, + "TGTGCG": 1920, + "TGTGGA": 1921, + "TGTGGT": 1922, + "TGTGGC": 1923, + "TGTGGG": 1924, + "TGCAAA": 1925, + "TGCAAT": 1926, + "TGCAAC": 1927, + "TGCAAG": 1928, + "TGCATA": 1929, + "TGCATT": 1930, + "TGCATC": 1931, + "TGCATG": 1932, + "TGCACA": 1933, + "TGCACT": 1934, + "TGCACC": 1935, + "TGCACG": 1936, + "TGCAGA": 1937, + "TGCAGT": 1938, + "TGCAGC": 1939, + "TGCAGG": 1940, + "TGCTAA": 1941, + "TGCTAT": 1942, + "TGCTAC": 1943, + "TGCTAG": 1944, + "TGCTTA": 1945, + "TGCTTT": 1946, + "TGCTTC": 1947, + "TGCTTG": 1948, + "TGCTCA": 1949, + "TGCTCT": 1950, + "TGCTCC": 1951, + "TGCTCG": 1952, + "TGCTGA": 1953, + "TGCTGT": 1954, + "TGCTGC": 1955, + "TGCTGG": 1956, + "TGCCAA": 1957, + "TGCCAT": 1958, + "TGCCAC": 1959, + "TGCCAG": 1960, + "TGCCTA": 1961, + "TGCCTT": 1962, + "TGCCTC": 1963, + "TGCCTG": 1964, + "TGCCCA": 1965, + "TGCCCT": 1966, + "TGCCCC": 1967, + "TGCCCG": 1968, + "TGCCGA": 1969, + "TGCCGT": 1970, + "TGCCGC": 1971, + "TGCCGG": 1972, + "TGCGAA": 1973, + "TGCGAT": 1974, + "TGCGAC": 1975, + "TGCGAG": 1976, + "TGCGTA": 1977, + "TGCGTT": 1978, + "TGCGTC": 1979, + "TGCGTG": 1980, + "TGCGCA": 1981, + "TGCGCT": 1982, + "TGCGCC": 1983, + "TGCGCG": 1984, + "TGCGGA": 1985, + "TGCGGT": 1986, + "TGCGGC": 1987, + "TGCGGG": 1988, + "TGGAAA": 1989, + "TGGAAT": 1990, + "TGGAAC": 1991, + "TGGAAG": 1992, + "TGGATA": 1993, + "TGGATT": 1994, + "TGGATC": 1995, + "TGGATG": 1996, + "TGGACA": 1997, + "TGGACT": 1998, + "TGGACC": 1999, + "TGGACG": 2000, + "TGGAGA": 2001, + "TGGAGT": 2002, + "TGGAGC": 2003, + "TGGAGG": 2004, + "TGGTAA": 2005, + "TGGTAT": 2006, + "TGGTAC": 2007, + "TGGTAG": 2008, + "TGGTTA": 2009, + "TGGTTT": 2010, + "TGGTTC": 2011, + "TGGTTG": 2012, + "TGGTCA": 2013, + "TGGTCT": 2014, + "TGGTCC": 2015, + "TGGTCG": 2016, + "TGGTGA": 2017, + "TGGTGT": 2018, + "TGGTGC": 2019, + "TGGTGG": 2020, + "TGGCAA": 2021, + "TGGCAT": 2022, + "TGGCAC": 2023, + "TGGCAG": 2024, + "TGGCTA": 2025, + "TGGCTT": 2026, + "TGGCTC": 2027, + "TGGCTG": 2028, + "TGGCCA": 2029, + "TGGCCT": 2030, + "TGGCCC": 2031, + "TGGCCG": 2032, + "TGGCGA": 2033, + "TGGCGT": 2034, + "TGGCGC": 2035, + "TGGCGG": 2036, + "TGGGAA": 2037, + "TGGGAT": 2038, + "TGGGAC": 2039, + "TGGGAG": 2040, + "TGGGTA": 2041, + "TGGGTT": 2042, + "TGGGTC": 2043, + "TGGGTG": 2044, + "TGGGCA": 2045, + "TGGGCT": 2046, + "TGGGCC": 2047, + "TGGGCG": 2048, + "TGGGGA": 2049, + "TGGGGT": 2050, + "TGGGGC": 2051, + "TGGGGG": 2052, + "CAAAAA": 2053, + "CAAAAT": 2054, + "CAAAAC": 2055, + "CAAAAG": 2056, + "CAAATA": 2057, + "CAAATT": 2058, + "CAAATC": 2059, + "CAAATG": 2060, + "CAAACA": 2061, + "CAAACT": 2062, + "CAAACC": 2063, + "CAAACG": 2064, + "CAAAGA": 2065, + "CAAAGT": 2066, + "CAAAGC": 2067, + "CAAAGG": 2068, + "CAATAA": 2069, + "CAATAT": 2070, + "CAATAC": 2071, + "CAATAG": 2072, + "CAATTA": 2073, + "CAATTT": 2074, + "CAATTC": 2075, + "CAATTG": 2076, + "CAATCA": 2077, + "CAATCT": 2078, + "CAATCC": 2079, + "CAATCG": 2080, + "CAATGA": 2081, + "CAATGT": 2082, + "CAATGC": 2083, + "CAATGG": 2084, + "CAACAA": 2085, + "CAACAT": 2086, + "CAACAC": 2087, + "CAACAG": 2088, + "CAACTA": 2089, + "CAACTT": 2090, + "CAACTC": 2091, + "CAACTG": 2092, + "CAACCA": 2093, + "CAACCT": 2094, + "CAACCC": 2095, + "CAACCG": 2096, + "CAACGA": 2097, + "CAACGT": 2098, + "CAACGC": 2099, + "CAACGG": 2100, + "CAAGAA": 2101, + "CAAGAT": 2102, + "CAAGAC": 2103, + "CAAGAG": 2104, + "CAAGTA": 2105, + "CAAGTT": 2106, + "CAAGTC": 2107, + "CAAGTG": 2108, + "CAAGCA": 2109, + "CAAGCT": 2110, + "CAAGCC": 2111, + "CAAGCG": 2112, + "CAAGGA": 2113, + "CAAGGT": 2114, + "CAAGGC": 2115, + "CAAGGG": 2116, + "CATAAA": 2117, + "CATAAT": 2118, + "CATAAC": 2119, + "CATAAG": 2120, + "CATATA": 2121, + "CATATT": 2122, + "CATATC": 2123, + "CATATG": 2124, + "CATACA": 2125, + "CATACT": 2126, + "CATACC": 2127, + "CATACG": 2128, + "CATAGA": 2129, + "CATAGT": 2130, + "CATAGC": 2131, + "CATAGG": 2132, + "CATTAA": 2133, + "CATTAT": 2134, + "CATTAC": 2135, + "CATTAG": 2136, + "CATTTA": 2137, + "CATTTT": 2138, + "CATTTC": 2139, + "CATTTG": 2140, + "CATTCA": 2141, + "CATTCT": 2142, + "CATTCC": 2143, + "CATTCG": 2144, + "CATTGA": 2145, + "CATTGT": 2146, + "CATTGC": 2147, + "CATTGG": 2148, + "CATCAA": 2149, + "CATCAT": 2150, + "CATCAC": 2151, + "CATCAG": 2152, + "CATCTA": 2153, + "CATCTT": 2154, + "CATCTC": 2155, + "CATCTG": 2156, + "CATCCA": 2157, + "CATCCT": 2158, + "CATCCC": 2159, + "CATCCG": 2160, + "CATCGA": 2161, + "CATCGT": 2162, + "CATCGC": 2163, + "CATCGG": 2164, + "CATGAA": 2165, + "CATGAT": 2166, + "CATGAC": 2167, + "CATGAG": 2168, + "CATGTA": 2169, + "CATGTT": 2170, + "CATGTC": 2171, + "CATGTG": 2172, + "CATGCA": 2173, + "CATGCT": 2174, + "CATGCC": 2175, + "CATGCG": 2176, + "CATGGA": 2177, + "CATGGT": 2178, + "CATGGC": 2179, + "CATGGG": 2180, + "CACAAA": 2181, + "CACAAT": 2182, + "CACAAC": 2183, + "CACAAG": 2184, + "CACATA": 2185, + "CACATT": 2186, + "CACATC": 2187, + "CACATG": 2188, + "CACACA": 2189, + "CACACT": 2190, + "CACACC": 2191, + "CACACG": 2192, + "CACAGA": 2193, + "CACAGT": 2194, + "CACAGC": 2195, + "CACAGG": 2196, + "CACTAA": 2197, + "CACTAT": 2198, + "CACTAC": 2199, + "CACTAG": 2200, + "CACTTA": 2201, + "CACTTT": 2202, + "CACTTC": 2203, + "CACTTG": 2204, + "CACTCA": 2205, + "CACTCT": 2206, + "CACTCC": 2207, + "CACTCG": 2208, + "CACTGA": 2209, + "CACTGT": 2210, + "CACTGC": 2211, + "CACTGG": 2212, + "CACCAA": 2213, + "CACCAT": 2214, + "CACCAC": 2215, + "CACCAG": 2216, + "CACCTA": 2217, + "CACCTT": 2218, + "CACCTC": 2219, + "CACCTG": 2220, + "CACCCA": 2221, + "CACCCT": 2222, + "CACCCC": 2223, + "CACCCG": 2224, + "CACCGA": 2225, + "CACCGT": 2226, + "CACCGC": 2227, + "CACCGG": 2228, + "CACGAA": 2229, + "CACGAT": 2230, + "CACGAC": 2231, + "CACGAG": 2232, + "CACGTA": 2233, + "CACGTT": 2234, + "CACGTC": 2235, + "CACGTG": 2236, + "CACGCA": 2237, + "CACGCT": 2238, + "CACGCC": 2239, + "CACGCG": 2240, + "CACGGA": 2241, + "CACGGT": 2242, + "CACGGC": 2243, + "CACGGG": 2244, + "CAGAAA": 2245, + "CAGAAT": 2246, + "CAGAAC": 2247, + "CAGAAG": 2248, + "CAGATA": 2249, + "CAGATT": 2250, + "CAGATC": 2251, + "CAGATG": 2252, + "CAGACA": 2253, + "CAGACT": 2254, + "CAGACC": 2255, + "CAGACG": 2256, + "CAGAGA": 2257, + "CAGAGT": 2258, + "CAGAGC": 2259, + "CAGAGG": 2260, + "CAGTAA": 2261, + "CAGTAT": 2262, + "CAGTAC": 2263, + "CAGTAG": 2264, + "CAGTTA": 2265, + "CAGTTT": 2266, + "CAGTTC": 2267, + "CAGTTG": 2268, + "CAGTCA": 2269, + "CAGTCT": 2270, + "CAGTCC": 2271, + "CAGTCG": 2272, + "CAGTGA": 2273, + "CAGTGT": 2274, + "CAGTGC": 2275, + "CAGTGG": 2276, + "CAGCAA": 2277, + "CAGCAT": 2278, + "CAGCAC": 2279, + "CAGCAG": 2280, + "CAGCTA": 2281, + "CAGCTT": 2282, + "CAGCTC": 2283, + "CAGCTG": 2284, + "CAGCCA": 2285, + "CAGCCT": 2286, + "CAGCCC": 2287, + "CAGCCG": 2288, + "CAGCGA": 2289, + "CAGCGT": 2290, + "CAGCGC": 2291, + "CAGCGG": 2292, + "CAGGAA": 2293, + "CAGGAT": 2294, + "CAGGAC": 2295, + "CAGGAG": 2296, + "CAGGTA": 2297, + "CAGGTT": 2298, + "CAGGTC": 2299, + "CAGGTG": 2300, + "CAGGCA": 2301, + "CAGGCT": 2302, + "CAGGCC": 2303, + "CAGGCG": 2304, + "CAGGGA": 2305, + "CAGGGT": 2306, + "CAGGGC": 2307, + "CAGGGG": 2308, + "CTAAAA": 2309, + "CTAAAT": 2310, + "CTAAAC": 2311, + "CTAAAG": 2312, + "CTAATA": 2313, + "CTAATT": 2314, + "CTAATC": 2315, + "CTAATG": 2316, + "CTAACA": 2317, + "CTAACT": 2318, + "CTAACC": 2319, + "CTAACG": 2320, + "CTAAGA": 2321, + "CTAAGT": 2322, + "CTAAGC": 2323, + "CTAAGG": 2324, + "CTATAA": 2325, + "CTATAT": 2326, + "CTATAC": 2327, + "CTATAG": 2328, + "CTATTA": 2329, + "CTATTT": 2330, + "CTATTC": 2331, + "CTATTG": 2332, + "CTATCA": 2333, + "CTATCT": 2334, + "CTATCC": 2335, + "CTATCG": 2336, + "CTATGA": 2337, + "CTATGT": 2338, + "CTATGC": 2339, + "CTATGG": 2340, + "CTACAA": 2341, + "CTACAT": 2342, + "CTACAC": 2343, + "CTACAG": 2344, + "CTACTA": 2345, + "CTACTT": 2346, + "CTACTC": 2347, + "CTACTG": 2348, + "CTACCA": 2349, + "CTACCT": 2350, + "CTACCC": 2351, + "CTACCG": 2352, + "CTACGA": 2353, + "CTACGT": 2354, + "CTACGC": 2355, + "CTACGG": 2356, + "CTAGAA": 2357, + "CTAGAT": 2358, + "CTAGAC": 2359, + "CTAGAG": 2360, + "CTAGTA": 2361, + "CTAGTT": 2362, + "CTAGTC": 2363, + "CTAGTG": 2364, + "CTAGCA": 2365, + "CTAGCT": 2366, + "CTAGCC": 2367, + "CTAGCG": 2368, + "CTAGGA": 2369, + "CTAGGT": 2370, + "CTAGGC": 2371, + "CTAGGG": 2372, + "CTTAAA": 2373, + "CTTAAT": 2374, + "CTTAAC": 2375, + "CTTAAG": 2376, + "CTTATA": 2377, + "CTTATT": 2378, + "CTTATC": 2379, + "CTTATG": 2380, + "CTTACA": 2381, + "CTTACT": 2382, + "CTTACC": 2383, + "CTTACG": 2384, + "CTTAGA": 2385, + "CTTAGT": 2386, + "CTTAGC": 2387, + "CTTAGG": 2388, + "CTTTAA": 2389, + "CTTTAT": 2390, + "CTTTAC": 2391, + "CTTTAG": 2392, + "CTTTTA": 2393, + "CTTTTT": 2394, + "CTTTTC": 2395, + "CTTTTG": 2396, + "CTTTCA": 2397, + "CTTTCT": 2398, + "CTTTCC": 2399, + "CTTTCG": 2400, + "CTTTGA": 2401, + "CTTTGT": 2402, + "CTTTGC": 2403, + "CTTTGG": 2404, + "CTTCAA": 2405, + "CTTCAT": 2406, + "CTTCAC": 2407, + "CTTCAG": 2408, + "CTTCTA": 2409, + "CTTCTT": 2410, + "CTTCTC": 2411, + "CTTCTG": 2412, + "CTTCCA": 2413, + "CTTCCT": 2414, + "CTTCCC": 2415, + "CTTCCG": 2416, + "CTTCGA": 2417, + "CTTCGT": 2418, + "CTTCGC": 2419, + "CTTCGG": 2420, + "CTTGAA": 2421, + "CTTGAT": 2422, + "CTTGAC": 2423, + "CTTGAG": 2424, + "CTTGTA": 2425, + "CTTGTT": 2426, + "CTTGTC": 2427, + "CTTGTG": 2428, + "CTTGCA": 2429, + "CTTGCT": 2430, + "CTTGCC": 2431, + "CTTGCG": 2432, + "CTTGGA": 2433, + "CTTGGT": 2434, + "CTTGGC": 2435, + "CTTGGG": 2436, + "CTCAAA": 2437, + "CTCAAT": 2438, + "CTCAAC": 2439, + "CTCAAG": 2440, + "CTCATA": 2441, + "CTCATT": 2442, + "CTCATC": 2443, + "CTCATG": 2444, + "CTCACA": 2445, + "CTCACT": 2446, + "CTCACC": 2447, + "CTCACG": 2448, + "CTCAGA": 2449, + "CTCAGT": 2450, + "CTCAGC": 2451, + "CTCAGG": 2452, + "CTCTAA": 2453, + "CTCTAT": 2454, + "CTCTAC": 2455, + "CTCTAG": 2456, + "CTCTTA": 2457, + "CTCTTT": 2458, + "CTCTTC": 2459, + "CTCTTG": 2460, + "CTCTCA": 2461, + "CTCTCT": 2462, + "CTCTCC": 2463, + "CTCTCG": 2464, + "CTCTGA": 2465, + "CTCTGT": 2466, + "CTCTGC": 2467, + "CTCTGG": 2468, + "CTCCAA": 2469, + "CTCCAT": 2470, + "CTCCAC": 2471, + "CTCCAG": 2472, + "CTCCTA": 2473, + "CTCCTT": 2474, + "CTCCTC": 2475, + "CTCCTG": 2476, + "CTCCCA": 2477, + "CTCCCT": 2478, + "CTCCCC": 2479, + "CTCCCG": 2480, + "CTCCGA": 2481, + "CTCCGT": 2482, + "CTCCGC": 2483, + "CTCCGG": 2484, + "CTCGAA": 2485, + "CTCGAT": 2486, + "CTCGAC": 2487, + "CTCGAG": 2488, + "CTCGTA": 2489, + "CTCGTT": 2490, + "CTCGTC": 2491, + "CTCGTG": 2492, + "CTCGCA": 2493, + "CTCGCT": 2494, + "CTCGCC": 2495, + "CTCGCG": 2496, + "CTCGGA": 2497, + "CTCGGT": 2498, + "CTCGGC": 2499, + "CTCGGG": 2500, + "CTGAAA": 2501, + "CTGAAT": 2502, + "CTGAAC": 2503, + "CTGAAG": 2504, + "CTGATA": 2505, + "CTGATT": 2506, + "CTGATC": 2507, + "CTGATG": 2508, + "CTGACA": 2509, + "CTGACT": 2510, + "CTGACC": 2511, + "CTGACG": 2512, + "CTGAGA": 2513, + "CTGAGT": 2514, + "CTGAGC": 2515, + "CTGAGG": 2516, + "CTGTAA": 2517, + "CTGTAT": 2518, + "CTGTAC": 2519, + "CTGTAG": 2520, + "CTGTTA": 2521, + "CTGTTT": 2522, + "CTGTTC": 2523, + "CTGTTG": 2524, + "CTGTCA": 2525, + "CTGTCT": 2526, + "CTGTCC": 2527, + "CTGTCG": 2528, + "CTGTGA": 2529, + "CTGTGT": 2530, + "CTGTGC": 2531, + "CTGTGG": 2532, + "CTGCAA": 2533, + "CTGCAT": 2534, + "CTGCAC": 2535, + "CTGCAG": 2536, + "CTGCTA": 2537, + "CTGCTT": 2538, + "CTGCTC": 2539, + "CTGCTG": 2540, + "CTGCCA": 2541, + "CTGCCT": 2542, + "CTGCCC": 2543, + "CTGCCG": 2544, + "CTGCGA": 2545, + "CTGCGT": 2546, + "CTGCGC": 2547, + "CTGCGG": 2548, + "CTGGAA": 2549, + "CTGGAT": 2550, + "CTGGAC": 2551, + "CTGGAG": 2552, + "CTGGTA": 2553, + "CTGGTT": 2554, + "CTGGTC": 2555, + "CTGGTG": 2556, + "CTGGCA": 2557, + "CTGGCT": 2558, + "CTGGCC": 2559, + "CTGGCG": 2560, + "CTGGGA": 2561, + "CTGGGT": 2562, + "CTGGGC": 2563, + "CTGGGG": 2564, + "CCAAAA": 2565, + "CCAAAT": 2566, + "CCAAAC": 2567, + "CCAAAG": 2568, + "CCAATA": 2569, + "CCAATT": 2570, + "CCAATC": 2571, + "CCAATG": 2572, + "CCAACA": 2573, + "CCAACT": 2574, + "CCAACC": 2575, + "CCAACG": 2576, + "CCAAGA": 2577, + "CCAAGT": 2578, + "CCAAGC": 2579, + "CCAAGG": 2580, + "CCATAA": 2581, + "CCATAT": 2582, + "CCATAC": 2583, + "CCATAG": 2584, + "CCATTA": 2585, + "CCATTT": 2586, + "CCATTC": 2587, + "CCATTG": 2588, + "CCATCA": 2589, + "CCATCT": 2590, + "CCATCC": 2591, + "CCATCG": 2592, + "CCATGA": 2593, + "CCATGT": 2594, + "CCATGC": 2595, + "CCATGG": 2596, + "CCACAA": 2597, + "CCACAT": 2598, + "CCACAC": 2599, + "CCACAG": 2600, + "CCACTA": 2601, + "CCACTT": 2602, + "CCACTC": 2603, + "CCACTG": 2604, + "CCACCA": 2605, + "CCACCT": 2606, + "CCACCC": 2607, + "CCACCG": 2608, + "CCACGA": 2609, + "CCACGT": 2610, + "CCACGC": 2611, + "CCACGG": 2612, + "CCAGAA": 2613, + "CCAGAT": 2614, + "CCAGAC": 2615, + "CCAGAG": 2616, + "CCAGTA": 2617, + "CCAGTT": 2618, + "CCAGTC": 2619, + "CCAGTG": 2620, + "CCAGCA": 2621, + "CCAGCT": 2622, + "CCAGCC": 2623, + "CCAGCG": 2624, + "CCAGGA": 2625, + "CCAGGT": 2626, + "CCAGGC": 2627, + "CCAGGG": 2628, + "CCTAAA": 2629, + "CCTAAT": 2630, + "CCTAAC": 2631, + "CCTAAG": 2632, + "CCTATA": 2633, + "CCTATT": 2634, + "CCTATC": 2635, + "CCTATG": 2636, + "CCTACA": 2637, + "CCTACT": 2638, + "CCTACC": 2639, + "CCTACG": 2640, + "CCTAGA": 2641, + "CCTAGT": 2642, + "CCTAGC": 2643, + "CCTAGG": 2644, + "CCTTAA": 2645, + "CCTTAT": 2646, + "CCTTAC": 2647, + "CCTTAG": 2648, + "CCTTTA": 2649, + "CCTTTT": 2650, + "CCTTTC": 2651, + "CCTTTG": 2652, + "CCTTCA": 2653, + "CCTTCT": 2654, + "CCTTCC": 2655, + "CCTTCG": 2656, + "CCTTGA": 2657, + "CCTTGT": 2658, + "CCTTGC": 2659, + "CCTTGG": 2660, + "CCTCAA": 2661, + "CCTCAT": 2662, + "CCTCAC": 2663, + "CCTCAG": 2664, + "CCTCTA": 2665, + "CCTCTT": 2666, + "CCTCTC": 2667, + "CCTCTG": 2668, + "CCTCCA": 2669, + "CCTCCT": 2670, + "CCTCCC": 2671, + "CCTCCG": 2672, + "CCTCGA": 2673, + "CCTCGT": 2674, + "CCTCGC": 2675, + "CCTCGG": 2676, + "CCTGAA": 2677, + "CCTGAT": 2678, + "CCTGAC": 2679, + "CCTGAG": 2680, + "CCTGTA": 2681, + "CCTGTT": 2682, + "CCTGTC": 2683, + "CCTGTG": 2684, + "CCTGCA": 2685, + "CCTGCT": 2686, + "CCTGCC": 2687, + "CCTGCG": 2688, + "CCTGGA": 2689, + "CCTGGT": 2690, + "CCTGGC": 2691, + "CCTGGG": 2692, + "CCCAAA": 2693, + "CCCAAT": 2694, + "CCCAAC": 2695, + "CCCAAG": 2696, + "CCCATA": 2697, + "CCCATT": 2698, + "CCCATC": 2699, + "CCCATG": 2700, + "CCCACA": 2701, + "CCCACT": 2702, + "CCCACC": 2703, + "CCCACG": 2704, + "CCCAGA": 2705, + "CCCAGT": 2706, + "CCCAGC": 2707, + "CCCAGG": 2708, + "CCCTAA": 2709, + "CCCTAT": 2710, + "CCCTAC": 2711, + "CCCTAG": 2712, + "CCCTTA": 2713, + "CCCTTT": 2714, + "CCCTTC": 2715, + "CCCTTG": 2716, + "CCCTCA": 2717, + "CCCTCT": 2718, + "CCCTCC": 2719, + "CCCTCG": 2720, + "CCCTGA": 2721, + "CCCTGT": 2722, + "CCCTGC": 2723, + "CCCTGG": 2724, + "CCCCAA": 2725, + "CCCCAT": 2726, + "CCCCAC": 2727, + "CCCCAG": 2728, + "CCCCTA": 2729, + "CCCCTT": 2730, + "CCCCTC": 2731, + "CCCCTG": 2732, + "CCCCCA": 2733, + "CCCCCT": 2734, + "CCCCCC": 2735, + "CCCCCG": 2736, + "CCCCGA": 2737, + "CCCCGT": 2738, + "CCCCGC": 2739, + "CCCCGG": 2740, + "CCCGAA": 2741, + "CCCGAT": 2742, + "CCCGAC": 2743, + "CCCGAG": 2744, + "CCCGTA": 2745, + "CCCGTT": 2746, + "CCCGTC": 2747, + "CCCGTG": 2748, + "CCCGCA": 2749, + "CCCGCT": 2750, + "CCCGCC": 2751, + "CCCGCG": 2752, + "CCCGGA": 2753, + "CCCGGT": 2754, + "CCCGGC": 2755, + "CCCGGG": 2756, + "CCGAAA": 2757, + "CCGAAT": 2758, + "CCGAAC": 2759, + "CCGAAG": 2760, + "CCGATA": 2761, + "CCGATT": 2762, + "CCGATC": 2763, + "CCGATG": 2764, + "CCGACA": 2765, + "CCGACT": 2766, + "CCGACC": 2767, + "CCGACG": 2768, + "CCGAGA": 2769, + "CCGAGT": 2770, + "CCGAGC": 2771, + "CCGAGG": 2772, + "CCGTAA": 2773, + "CCGTAT": 2774, + "CCGTAC": 2775, + "CCGTAG": 2776, + "CCGTTA": 2777, + "CCGTTT": 2778, + "CCGTTC": 2779, + "CCGTTG": 2780, + "CCGTCA": 2781, + "CCGTCT": 2782, + "CCGTCC": 2783, + "CCGTCG": 2784, + "CCGTGA": 2785, + "CCGTGT": 2786, + "CCGTGC": 2787, + "CCGTGG": 2788, + "CCGCAA": 2789, + "CCGCAT": 2790, + "CCGCAC": 2791, + "CCGCAG": 2792, + "CCGCTA": 2793, + "CCGCTT": 2794, + "CCGCTC": 2795, + "CCGCTG": 2796, + "CCGCCA": 2797, + "CCGCCT": 2798, + "CCGCCC": 2799, + "CCGCCG": 2800, + "CCGCGA": 2801, + "CCGCGT": 2802, + "CCGCGC": 2803, + "CCGCGG": 2804, + "CCGGAA": 2805, + "CCGGAT": 2806, + "CCGGAC": 2807, + "CCGGAG": 2808, + "CCGGTA": 2809, + "CCGGTT": 2810, + "CCGGTC": 2811, + "CCGGTG": 2812, + "CCGGCA": 2813, + "CCGGCT": 2814, + "CCGGCC": 2815, + "CCGGCG": 2816, + "CCGGGA": 2817, + "CCGGGT": 2818, + "CCGGGC": 2819, + "CCGGGG": 2820, + "CGAAAA": 2821, + "CGAAAT": 2822, + "CGAAAC": 2823, + "CGAAAG": 2824, + "CGAATA": 2825, + "CGAATT": 2826, + "CGAATC": 2827, + "CGAATG": 2828, + "CGAACA": 2829, + "CGAACT": 2830, + "CGAACC": 2831, + "CGAACG": 2832, + "CGAAGA": 2833, + "CGAAGT": 2834, + "CGAAGC": 2835, + "CGAAGG": 2836, + "CGATAA": 2837, + "CGATAT": 2838, + "CGATAC": 2839, + "CGATAG": 2840, + "CGATTA": 2841, + "CGATTT": 2842, + "CGATTC": 2843, + "CGATTG": 2844, + "CGATCA": 2845, + "CGATCT": 2846, + "CGATCC": 2847, + "CGATCG": 2848, + "CGATGA": 2849, + "CGATGT": 2850, + "CGATGC": 2851, + "CGATGG": 2852, + "CGACAA": 2853, + "CGACAT": 2854, + "CGACAC": 2855, + "CGACAG": 2856, + "CGACTA": 2857, + "CGACTT": 2858, + "CGACTC": 2859, + "CGACTG": 2860, + "CGACCA": 2861, + "CGACCT": 2862, + "CGACCC": 2863, + "CGACCG": 2864, + "CGACGA": 2865, + "CGACGT": 2866, + "CGACGC": 2867, + "CGACGG": 2868, + "CGAGAA": 2869, + "CGAGAT": 2870, + "CGAGAC": 2871, + "CGAGAG": 2872, + "CGAGTA": 2873, + "CGAGTT": 2874, + "CGAGTC": 2875, + "CGAGTG": 2876, + "CGAGCA": 2877, + "CGAGCT": 2878, + "CGAGCC": 2879, + "CGAGCG": 2880, + "CGAGGA": 2881, + "CGAGGT": 2882, + "CGAGGC": 2883, + "CGAGGG": 2884, + "CGTAAA": 2885, + "CGTAAT": 2886, + "CGTAAC": 2887, + "CGTAAG": 2888, + "CGTATA": 2889, + "CGTATT": 2890, + "CGTATC": 2891, + "CGTATG": 2892, + "CGTACA": 2893, + "CGTACT": 2894, + "CGTACC": 2895, + "CGTACG": 2896, + "CGTAGA": 2897, + "CGTAGT": 2898, + "CGTAGC": 2899, + "CGTAGG": 2900, + "CGTTAA": 2901, + "CGTTAT": 2902, + "CGTTAC": 2903, + "CGTTAG": 2904, + "CGTTTA": 2905, + "CGTTTT": 2906, + "CGTTTC": 2907, + "CGTTTG": 2908, + "CGTTCA": 2909, + "CGTTCT": 2910, + "CGTTCC": 2911, + "CGTTCG": 2912, + "CGTTGA": 2913, + "CGTTGT": 2914, + "CGTTGC": 2915, + "CGTTGG": 2916, + "CGTCAA": 2917, + "CGTCAT": 2918, + "CGTCAC": 2919, + "CGTCAG": 2920, + "CGTCTA": 2921, + "CGTCTT": 2922, + "CGTCTC": 2923, + "CGTCTG": 2924, + "CGTCCA": 2925, + "CGTCCT": 2926, + "CGTCCC": 2927, + "CGTCCG": 2928, + "CGTCGA": 2929, + "CGTCGT": 2930, + "CGTCGC": 2931, + "CGTCGG": 2932, + "CGTGAA": 2933, + "CGTGAT": 2934, + "CGTGAC": 2935, + "CGTGAG": 2936, + "CGTGTA": 2937, + "CGTGTT": 2938, + "CGTGTC": 2939, + "CGTGTG": 2940, + "CGTGCA": 2941, + "CGTGCT": 2942, + "CGTGCC": 2943, + "CGTGCG": 2944, + "CGTGGA": 2945, + "CGTGGT": 2946, + "CGTGGC": 2947, + "CGTGGG": 2948, + "CGCAAA": 2949, + "CGCAAT": 2950, + "CGCAAC": 2951, + "CGCAAG": 2952, + "CGCATA": 2953, + "CGCATT": 2954, + "CGCATC": 2955, + "CGCATG": 2956, + "CGCACA": 2957, + "CGCACT": 2958, + "CGCACC": 2959, + "CGCACG": 2960, + "CGCAGA": 2961, + "CGCAGT": 2962, + "CGCAGC": 2963, + "CGCAGG": 2964, + "CGCTAA": 2965, + "CGCTAT": 2966, + "CGCTAC": 2967, + "CGCTAG": 2968, + "CGCTTA": 2969, + "CGCTTT": 2970, + "CGCTTC": 2971, + "CGCTTG": 2972, + "CGCTCA": 2973, + "CGCTCT": 2974, + "CGCTCC": 2975, + "CGCTCG": 2976, + "CGCTGA": 2977, + "CGCTGT": 2978, + "CGCTGC": 2979, + "CGCTGG": 2980, + "CGCCAA": 2981, + "CGCCAT": 2982, + "CGCCAC": 2983, + "CGCCAG": 2984, + "CGCCTA": 2985, + "CGCCTT": 2986, + "CGCCTC": 2987, + "CGCCTG": 2988, + "CGCCCA": 2989, + "CGCCCT": 2990, + "CGCCCC": 2991, + "CGCCCG": 2992, + "CGCCGA": 2993, + "CGCCGT": 2994, + "CGCCGC": 2995, + "CGCCGG": 2996, + "CGCGAA": 2997, + "CGCGAT": 2998, + "CGCGAC": 2999, + "CGCGAG": 3000, + "CGCGTA": 3001, + "CGCGTT": 3002, + "CGCGTC": 3003, + "CGCGTG": 3004, + "CGCGCA": 3005, + "CGCGCT": 3006, + "CGCGCC": 3007, + "CGCGCG": 3008, + "CGCGGA": 3009, + "CGCGGT": 3010, + "CGCGGC": 3011, + "CGCGGG": 3012, + "CGGAAA": 3013, + "CGGAAT": 3014, + "CGGAAC": 3015, + "CGGAAG": 3016, + "CGGATA": 3017, + "CGGATT": 3018, + "CGGATC": 3019, + "CGGATG": 3020, + "CGGACA": 3021, + "CGGACT": 3022, + "CGGACC": 3023, + "CGGACG": 3024, + "CGGAGA": 3025, + "CGGAGT": 3026, + "CGGAGC": 3027, + "CGGAGG": 3028, + "CGGTAA": 3029, + "CGGTAT": 3030, + "CGGTAC": 3031, + "CGGTAG": 3032, + "CGGTTA": 3033, + "CGGTTT": 3034, + "CGGTTC": 3035, + "CGGTTG": 3036, + "CGGTCA": 3037, + "CGGTCT": 3038, + "CGGTCC": 3039, + "CGGTCG": 3040, + "CGGTGA": 3041, + "CGGTGT": 3042, + "CGGTGC": 3043, + "CGGTGG": 3044, + "CGGCAA": 3045, + "CGGCAT": 3046, + "CGGCAC": 3047, + "CGGCAG": 3048, + "CGGCTA": 3049, + "CGGCTT": 3050, + "CGGCTC": 3051, + "CGGCTG": 3052, + "CGGCCA": 3053, + "CGGCCT": 3054, + "CGGCCC": 3055, + "CGGCCG": 3056, + "CGGCGA": 3057, + "CGGCGT": 3058, + "CGGCGC": 3059, + "CGGCGG": 3060, + "CGGGAA": 3061, + "CGGGAT": 3062, + "CGGGAC": 3063, + "CGGGAG": 3064, + "CGGGTA": 3065, + "CGGGTT": 3066, + "CGGGTC": 3067, + "CGGGTG": 3068, + "CGGGCA": 3069, + "CGGGCT": 3070, + "CGGGCC": 3071, + "CGGGCG": 3072, + "CGGGGA": 3073, + "CGGGGT": 3074, + "CGGGGC": 3075, + "CGGGGG": 3076, + "GAAAAA": 3077, + "GAAAAT": 3078, + "GAAAAC": 3079, + "GAAAAG": 3080, + "GAAATA": 3081, + "GAAATT": 3082, + "GAAATC": 3083, + "GAAATG": 3084, + "GAAACA": 3085, + "GAAACT": 3086, + "GAAACC": 3087, + "GAAACG": 3088, + "GAAAGA": 3089, + "GAAAGT": 3090, + "GAAAGC": 3091, + "GAAAGG": 3092, + "GAATAA": 3093, + "GAATAT": 3094, + "GAATAC": 3095, + "GAATAG": 3096, + "GAATTA": 3097, + "GAATTT": 3098, + "GAATTC": 3099, + "GAATTG": 3100, + "GAATCA": 3101, + "GAATCT": 3102, + "GAATCC": 3103, + "GAATCG": 3104, + "GAATGA": 3105, + "GAATGT": 3106, + "GAATGC": 3107, + "GAATGG": 3108, + "GAACAA": 3109, + "GAACAT": 3110, + "GAACAC": 3111, + "GAACAG": 3112, + "GAACTA": 3113, + "GAACTT": 3114, + "GAACTC": 3115, + "GAACTG": 3116, + "GAACCA": 3117, + "GAACCT": 3118, + "GAACCC": 3119, + "GAACCG": 3120, + "GAACGA": 3121, + "GAACGT": 3122, + "GAACGC": 3123, + "GAACGG": 3124, + "GAAGAA": 3125, + "GAAGAT": 3126, + "GAAGAC": 3127, + "GAAGAG": 3128, + "GAAGTA": 3129, + "GAAGTT": 3130, + "GAAGTC": 3131, + "GAAGTG": 3132, + "GAAGCA": 3133, + "GAAGCT": 3134, + "GAAGCC": 3135, + "GAAGCG": 3136, + "GAAGGA": 3137, + "GAAGGT": 3138, + "GAAGGC": 3139, + "GAAGGG": 3140, + "GATAAA": 3141, + "GATAAT": 3142, + "GATAAC": 3143, + "GATAAG": 3144, + "GATATA": 3145, + "GATATT": 3146, + "GATATC": 3147, + "GATATG": 3148, + "GATACA": 3149, + "GATACT": 3150, + "GATACC": 3151, + "GATACG": 3152, + "GATAGA": 3153, + "GATAGT": 3154, + "GATAGC": 3155, + "GATAGG": 3156, + "GATTAA": 3157, + "GATTAT": 3158, + "GATTAC": 3159, + "GATTAG": 3160, + "GATTTA": 3161, + "GATTTT": 3162, + "GATTTC": 3163, + "GATTTG": 3164, + "GATTCA": 3165, + "GATTCT": 3166, + "GATTCC": 3167, + "GATTCG": 3168, + "GATTGA": 3169, + "GATTGT": 3170, + "GATTGC": 3171, + "GATTGG": 3172, + "GATCAA": 3173, + "GATCAT": 3174, + "GATCAC": 3175, + "GATCAG": 3176, + "GATCTA": 3177, + "GATCTT": 3178, + "GATCTC": 3179, + "GATCTG": 3180, + "GATCCA": 3181, + "GATCCT": 3182, + "GATCCC": 3183, + "GATCCG": 3184, + "GATCGA": 3185, + "GATCGT": 3186, + "GATCGC": 3187, + "GATCGG": 3188, + "GATGAA": 3189, + "GATGAT": 3190, + "GATGAC": 3191, + "GATGAG": 3192, + "GATGTA": 3193, + "GATGTT": 3194, + "GATGTC": 3195, + "GATGTG": 3196, + "GATGCA": 3197, + "GATGCT": 3198, + "GATGCC": 3199, + "GATGCG": 3200, + "GATGGA": 3201, + "GATGGT": 3202, + "GATGGC": 3203, + "GATGGG": 3204, + "GACAAA": 3205, + "GACAAT": 3206, + "GACAAC": 3207, + "GACAAG": 3208, + "GACATA": 3209, + "GACATT": 3210, + "GACATC": 3211, + "GACATG": 3212, + "GACACA": 3213, + "GACACT": 3214, + "GACACC": 3215, + "GACACG": 3216, + "GACAGA": 3217, + "GACAGT": 3218, + "GACAGC": 3219, + "GACAGG": 3220, + "GACTAA": 3221, + "GACTAT": 3222, + "GACTAC": 3223, + "GACTAG": 3224, + "GACTTA": 3225, + "GACTTT": 3226, + "GACTTC": 3227, + "GACTTG": 3228, + "GACTCA": 3229, + "GACTCT": 3230, + "GACTCC": 3231, + "GACTCG": 3232, + "GACTGA": 3233, + "GACTGT": 3234, + "GACTGC": 3235, + "GACTGG": 3236, + "GACCAA": 3237, + "GACCAT": 3238, + "GACCAC": 3239, + "GACCAG": 3240, + "GACCTA": 3241, + "GACCTT": 3242, + "GACCTC": 3243, + "GACCTG": 3244, + "GACCCA": 3245, + "GACCCT": 3246, + "GACCCC": 3247, + "GACCCG": 3248, + "GACCGA": 3249, + "GACCGT": 3250, + "GACCGC": 3251, + "GACCGG": 3252, + "GACGAA": 3253, + "GACGAT": 3254, + "GACGAC": 3255, + "GACGAG": 3256, + "GACGTA": 3257, + "GACGTT": 3258, + "GACGTC": 3259, + "GACGTG": 3260, + "GACGCA": 3261, + "GACGCT": 3262, + "GACGCC": 3263, + "GACGCG": 3264, + "GACGGA": 3265, + "GACGGT": 3266, + "GACGGC": 3267, + "GACGGG": 3268, + "GAGAAA": 3269, + "GAGAAT": 3270, + "GAGAAC": 3271, + "GAGAAG": 3272, + "GAGATA": 3273, + "GAGATT": 3274, + "GAGATC": 3275, + "GAGATG": 3276, + "GAGACA": 3277, + "GAGACT": 3278, + "GAGACC": 3279, + "GAGACG": 3280, + "GAGAGA": 3281, + "GAGAGT": 3282, + "GAGAGC": 3283, + "GAGAGG": 3284, + "GAGTAA": 3285, + "GAGTAT": 3286, + "GAGTAC": 3287, + "GAGTAG": 3288, + "GAGTTA": 3289, + "GAGTTT": 3290, + "GAGTTC": 3291, + "GAGTTG": 3292, + "GAGTCA": 3293, + "GAGTCT": 3294, + "GAGTCC": 3295, + "GAGTCG": 3296, + "GAGTGA": 3297, + "GAGTGT": 3298, + "GAGTGC": 3299, + "GAGTGG": 3300, + "GAGCAA": 3301, + "GAGCAT": 3302, + "GAGCAC": 3303, + "GAGCAG": 3304, + "GAGCTA": 3305, + "GAGCTT": 3306, + "GAGCTC": 3307, + "GAGCTG": 3308, + "GAGCCA": 3309, + "GAGCCT": 3310, + "GAGCCC": 3311, + "GAGCCG": 3312, + "GAGCGA": 3313, + "GAGCGT": 3314, + "GAGCGC": 3315, + "GAGCGG": 3316, + "GAGGAA": 3317, + "GAGGAT": 3318, + "GAGGAC": 3319, + "GAGGAG": 3320, + "GAGGTA": 3321, + "GAGGTT": 3322, + "GAGGTC": 3323, + "GAGGTG": 3324, + "GAGGCA": 3325, + "GAGGCT": 3326, + "GAGGCC": 3327, + "GAGGCG": 3328, + "GAGGGA": 3329, + "GAGGGT": 3330, + "GAGGGC": 3331, + "GAGGGG": 3332, + "GTAAAA": 3333, + "GTAAAT": 3334, + "GTAAAC": 3335, + "GTAAAG": 3336, + "GTAATA": 3337, + "GTAATT": 3338, + "GTAATC": 3339, + "GTAATG": 3340, + "GTAACA": 3341, + "GTAACT": 3342, + "GTAACC": 3343, + "GTAACG": 3344, + "GTAAGA": 3345, + "GTAAGT": 3346, + "GTAAGC": 3347, + "GTAAGG": 3348, + "GTATAA": 3349, + "GTATAT": 3350, + "GTATAC": 3351, + "GTATAG": 3352, + "GTATTA": 3353, + "GTATTT": 3354, + "GTATTC": 3355, + "GTATTG": 3356, + "GTATCA": 3357, + "GTATCT": 3358, + "GTATCC": 3359, + "GTATCG": 3360, + "GTATGA": 3361, + "GTATGT": 3362, + "GTATGC": 3363, + "GTATGG": 3364, + "GTACAA": 3365, + "GTACAT": 3366, + "GTACAC": 3367, + "GTACAG": 3368, + "GTACTA": 3369, + "GTACTT": 3370, + "GTACTC": 3371, + "GTACTG": 3372, + "GTACCA": 3373, + "GTACCT": 3374, + "GTACCC": 3375, + "GTACCG": 3376, + "GTACGA": 3377, + "GTACGT": 3378, + "GTACGC": 3379, + "GTACGG": 3380, + "GTAGAA": 3381, + "GTAGAT": 3382, + "GTAGAC": 3383, + "GTAGAG": 3384, + "GTAGTA": 3385, + "GTAGTT": 3386, + "GTAGTC": 3387, + "GTAGTG": 3388, + "GTAGCA": 3389, + "GTAGCT": 3390, + "GTAGCC": 3391, + "GTAGCG": 3392, + "GTAGGA": 3393, + "GTAGGT": 3394, + "GTAGGC": 3395, + "GTAGGG": 3396, + "GTTAAA": 3397, + "GTTAAT": 3398, + "GTTAAC": 3399, + "GTTAAG": 3400, + "GTTATA": 3401, + "GTTATT": 3402, + "GTTATC": 3403, + "GTTATG": 3404, + "GTTACA": 3405, + "GTTACT": 3406, + "GTTACC": 3407, + "GTTACG": 3408, + "GTTAGA": 3409, + "GTTAGT": 3410, + "GTTAGC": 3411, + "GTTAGG": 3412, + "GTTTAA": 3413, + "GTTTAT": 3414, + "GTTTAC": 3415, + "GTTTAG": 3416, + "GTTTTA": 3417, + "GTTTTT": 3418, + "GTTTTC": 3419, + "GTTTTG": 3420, + "GTTTCA": 3421, + "GTTTCT": 3422, + "GTTTCC": 3423, + "GTTTCG": 3424, + "GTTTGA": 3425, + "GTTTGT": 3426, + "GTTTGC": 3427, + "GTTTGG": 3428, + "GTTCAA": 3429, + "GTTCAT": 3430, + "GTTCAC": 3431, + "GTTCAG": 3432, + "GTTCTA": 3433, + "GTTCTT": 3434, + "GTTCTC": 3435, + "GTTCTG": 3436, + "GTTCCA": 3437, + "GTTCCT": 3438, + "GTTCCC": 3439, + "GTTCCG": 3440, + "GTTCGA": 3441, + "GTTCGT": 3442, + "GTTCGC": 3443, + "GTTCGG": 3444, + "GTTGAA": 3445, + "GTTGAT": 3446, + "GTTGAC": 3447, + "GTTGAG": 3448, + "GTTGTA": 3449, + "GTTGTT": 3450, + "GTTGTC": 3451, + "GTTGTG": 3452, + "GTTGCA": 3453, + "GTTGCT": 3454, + "GTTGCC": 3455, + "GTTGCG": 3456, + "GTTGGA": 3457, + "GTTGGT": 3458, + "GTTGGC": 3459, + "GTTGGG": 3460, + "GTCAAA": 3461, + "GTCAAT": 3462, + "GTCAAC": 3463, + "GTCAAG": 3464, + "GTCATA": 3465, + "GTCATT": 3466, + "GTCATC": 3467, + "GTCATG": 3468, + "GTCACA": 3469, + "GTCACT": 3470, + "GTCACC": 3471, + "GTCACG": 3472, + "GTCAGA": 3473, + "GTCAGT": 3474, + "GTCAGC": 3475, + "GTCAGG": 3476, + "GTCTAA": 3477, + "GTCTAT": 3478, + "GTCTAC": 3479, + "GTCTAG": 3480, + "GTCTTA": 3481, + "GTCTTT": 3482, + "GTCTTC": 3483, + "GTCTTG": 3484, + "GTCTCA": 3485, + "GTCTCT": 3486, + "GTCTCC": 3487, + "GTCTCG": 3488, + "GTCTGA": 3489, + "GTCTGT": 3490, + "GTCTGC": 3491, + "GTCTGG": 3492, + "GTCCAA": 3493, + "GTCCAT": 3494, + "GTCCAC": 3495, + "GTCCAG": 3496, + "GTCCTA": 3497, + "GTCCTT": 3498, + "GTCCTC": 3499, + "GTCCTG": 3500, + "GTCCCA": 3501, + "GTCCCT": 3502, + "GTCCCC": 3503, + "GTCCCG": 3504, + "GTCCGA": 3505, + "GTCCGT": 3506, + "GTCCGC": 3507, + "GTCCGG": 3508, + "GTCGAA": 3509, + "GTCGAT": 3510, + "GTCGAC": 3511, + "GTCGAG": 3512, + "GTCGTA": 3513, + "GTCGTT": 3514, + "GTCGTC": 3515, + "GTCGTG": 3516, + "GTCGCA": 3517, + "GTCGCT": 3518, + "GTCGCC": 3519, + "GTCGCG": 3520, + "GTCGGA": 3521, + "GTCGGT": 3522, + "GTCGGC": 3523, + "GTCGGG": 3524, + "GTGAAA": 3525, + "GTGAAT": 3526, + "GTGAAC": 3527, + "GTGAAG": 3528, + "GTGATA": 3529, + "GTGATT": 3530, + "GTGATC": 3531, + "GTGATG": 3532, + "GTGACA": 3533, + "GTGACT": 3534, + "GTGACC": 3535, + "GTGACG": 3536, + "GTGAGA": 3537, + "GTGAGT": 3538, + "GTGAGC": 3539, + "GTGAGG": 3540, + "GTGTAA": 3541, + "GTGTAT": 3542, + "GTGTAC": 3543, + "GTGTAG": 3544, + "GTGTTA": 3545, + "GTGTTT": 3546, + "GTGTTC": 3547, + "GTGTTG": 3548, + "GTGTCA": 3549, + "GTGTCT": 3550, + "GTGTCC": 3551, + "GTGTCG": 3552, + "GTGTGA": 3553, + "GTGTGT": 3554, + "GTGTGC": 3555, + "GTGTGG": 3556, + "GTGCAA": 3557, + "GTGCAT": 3558, + "GTGCAC": 3559, + "GTGCAG": 3560, + "GTGCTA": 3561, + "GTGCTT": 3562, + "GTGCTC": 3563, + "GTGCTG": 3564, + "GTGCCA": 3565, + "GTGCCT": 3566, + "GTGCCC": 3567, + "GTGCCG": 3568, + "GTGCGA": 3569, + "GTGCGT": 3570, + "GTGCGC": 3571, + "GTGCGG": 3572, + "GTGGAA": 3573, + "GTGGAT": 3574, + "GTGGAC": 3575, + "GTGGAG": 3576, + "GTGGTA": 3577, + "GTGGTT": 3578, + "GTGGTC": 3579, + "GTGGTG": 3580, + "GTGGCA": 3581, + "GTGGCT": 3582, + "GTGGCC": 3583, + "GTGGCG": 3584, + "GTGGGA": 3585, + "GTGGGT": 3586, + "GTGGGC": 3587, + "GTGGGG": 3588, + "GCAAAA": 3589, + "GCAAAT": 3590, + "GCAAAC": 3591, + "GCAAAG": 3592, + "GCAATA": 3593, + "GCAATT": 3594, + "GCAATC": 3595, + "GCAATG": 3596, + "GCAACA": 3597, + "GCAACT": 3598, + "GCAACC": 3599, + "GCAACG": 3600, + "GCAAGA": 3601, + "GCAAGT": 3602, + "GCAAGC": 3603, + "GCAAGG": 3604, + "GCATAA": 3605, + "GCATAT": 3606, + "GCATAC": 3607, + "GCATAG": 3608, + "GCATTA": 3609, + "GCATTT": 3610, + "GCATTC": 3611, + "GCATTG": 3612, + "GCATCA": 3613, + "GCATCT": 3614, + "GCATCC": 3615, + "GCATCG": 3616, + "GCATGA": 3617, + "GCATGT": 3618, + "GCATGC": 3619, + "GCATGG": 3620, + "GCACAA": 3621, + "GCACAT": 3622, + "GCACAC": 3623, + "GCACAG": 3624, + "GCACTA": 3625, + "GCACTT": 3626, + "GCACTC": 3627, + "GCACTG": 3628, + "GCACCA": 3629, + "GCACCT": 3630, + "GCACCC": 3631, + "GCACCG": 3632, + "GCACGA": 3633, + "GCACGT": 3634, + "GCACGC": 3635, + "GCACGG": 3636, + "GCAGAA": 3637, + "GCAGAT": 3638, + "GCAGAC": 3639, + "GCAGAG": 3640, + "GCAGTA": 3641, + "GCAGTT": 3642, + "GCAGTC": 3643, + "GCAGTG": 3644, + "GCAGCA": 3645, + "GCAGCT": 3646, + "GCAGCC": 3647, + "GCAGCG": 3648, + "GCAGGA": 3649, + "GCAGGT": 3650, + "GCAGGC": 3651, + "GCAGGG": 3652, + "GCTAAA": 3653, + "GCTAAT": 3654, + "GCTAAC": 3655, + "GCTAAG": 3656, + "GCTATA": 3657, + "GCTATT": 3658, + "GCTATC": 3659, + "GCTATG": 3660, + "GCTACA": 3661, + "GCTACT": 3662, + "GCTACC": 3663, + "GCTACG": 3664, + "GCTAGA": 3665, + "GCTAGT": 3666, + "GCTAGC": 3667, + "GCTAGG": 3668, + "GCTTAA": 3669, + "GCTTAT": 3670, + "GCTTAC": 3671, + "GCTTAG": 3672, + "GCTTTA": 3673, + "GCTTTT": 3674, + "GCTTTC": 3675, + "GCTTTG": 3676, + "GCTTCA": 3677, + "GCTTCT": 3678, + "GCTTCC": 3679, + "GCTTCG": 3680, + "GCTTGA": 3681, + "GCTTGT": 3682, + "GCTTGC": 3683, + "GCTTGG": 3684, + "GCTCAA": 3685, + "GCTCAT": 3686, + "GCTCAC": 3687, + "GCTCAG": 3688, + "GCTCTA": 3689, + "GCTCTT": 3690, + "GCTCTC": 3691, + "GCTCTG": 3692, + "GCTCCA": 3693, + "GCTCCT": 3694, + "GCTCCC": 3695, + "GCTCCG": 3696, + "GCTCGA": 3697, + "GCTCGT": 3698, + "GCTCGC": 3699, + "GCTCGG": 3700, + "GCTGAA": 3701, + "GCTGAT": 3702, + "GCTGAC": 3703, + "GCTGAG": 3704, + "GCTGTA": 3705, + "GCTGTT": 3706, + "GCTGTC": 3707, + "GCTGTG": 3708, + "GCTGCA": 3709, + "GCTGCT": 3710, + "GCTGCC": 3711, + "GCTGCG": 3712, + "GCTGGA": 3713, + "GCTGGT": 3714, + "GCTGGC": 3715, + "GCTGGG": 3716, + "GCCAAA": 3717, + "GCCAAT": 3718, + "GCCAAC": 3719, + "GCCAAG": 3720, + "GCCATA": 3721, + "GCCATT": 3722, + "GCCATC": 3723, + "GCCATG": 3724, + "GCCACA": 3725, + "GCCACT": 3726, + "GCCACC": 3727, + "GCCACG": 3728, + "GCCAGA": 3729, + "GCCAGT": 3730, + "GCCAGC": 3731, + "GCCAGG": 3732, + "GCCTAA": 3733, + "GCCTAT": 3734, + "GCCTAC": 3735, + "GCCTAG": 3736, + "GCCTTA": 3737, + "GCCTTT": 3738, + "GCCTTC": 3739, + "GCCTTG": 3740, + "GCCTCA": 3741, + "GCCTCT": 3742, + "GCCTCC": 3743, + "GCCTCG": 3744, + "GCCTGA": 3745, + "GCCTGT": 3746, + "GCCTGC": 3747, + "GCCTGG": 3748, + "GCCCAA": 3749, + "GCCCAT": 3750, + "GCCCAC": 3751, + "GCCCAG": 3752, + "GCCCTA": 3753, + "GCCCTT": 3754, + "GCCCTC": 3755, + "GCCCTG": 3756, + "GCCCCA": 3757, + "GCCCCT": 3758, + "GCCCCC": 3759, + "GCCCCG": 3760, + "GCCCGA": 3761, + "GCCCGT": 3762, + "GCCCGC": 3763, + "GCCCGG": 3764, + "GCCGAA": 3765, + "GCCGAT": 3766, + "GCCGAC": 3767, + "GCCGAG": 3768, + "GCCGTA": 3769, + "GCCGTT": 3770, + "GCCGTC": 3771, + "GCCGTG": 3772, + "GCCGCA": 3773, + "GCCGCT": 3774, + "GCCGCC": 3775, + "GCCGCG": 3776, + "GCCGGA": 3777, + "GCCGGT": 3778, + "GCCGGC": 3779, + "GCCGGG": 3780, + "GCGAAA": 3781, + "GCGAAT": 3782, + "GCGAAC": 3783, + "GCGAAG": 3784, + "GCGATA": 3785, + "GCGATT": 3786, + "GCGATC": 3787, + "GCGATG": 3788, + "GCGACA": 3789, + "GCGACT": 3790, + "GCGACC": 3791, + "GCGACG": 3792, + "GCGAGA": 3793, + "GCGAGT": 3794, + "GCGAGC": 3795, + "GCGAGG": 3796, + "GCGTAA": 3797, + "GCGTAT": 3798, + "GCGTAC": 3799, + "GCGTAG": 3800, + "GCGTTA": 3801, + "GCGTTT": 3802, + "GCGTTC": 3803, + "GCGTTG": 3804, + "GCGTCA": 3805, + "GCGTCT": 3806, + "GCGTCC": 3807, + "GCGTCG": 3808, + "GCGTGA": 3809, + "GCGTGT": 3810, + "GCGTGC": 3811, + "GCGTGG": 3812, + "GCGCAA": 3813, + "GCGCAT": 3814, + "GCGCAC": 3815, + "GCGCAG": 3816, + "GCGCTA": 3817, + "GCGCTT": 3818, + "GCGCTC": 3819, + "GCGCTG": 3820, + "GCGCCA": 3821, + "GCGCCT": 3822, + "GCGCCC": 3823, + "GCGCCG": 3824, + "GCGCGA": 3825, + "GCGCGT": 3826, + "GCGCGC": 3827, + "GCGCGG": 3828, + "GCGGAA": 3829, + "GCGGAT": 3830, + "GCGGAC": 3831, + "GCGGAG": 3832, + "GCGGTA": 3833, + "GCGGTT": 3834, + "GCGGTC": 3835, + "GCGGTG": 3836, + "GCGGCA": 3837, + "GCGGCT": 3838, + "GCGGCC": 3839, + "GCGGCG": 3840, + "GCGGGA": 3841, + "GCGGGT": 3842, + "GCGGGC": 3843, + "GCGGGG": 3844, + "GGAAAA": 3845, + "GGAAAT": 3846, + "GGAAAC": 3847, + "GGAAAG": 3848, + "GGAATA": 3849, + "GGAATT": 3850, + "GGAATC": 3851, + "GGAATG": 3852, + "GGAACA": 3853, + "GGAACT": 3854, + "GGAACC": 3855, + "GGAACG": 3856, + "GGAAGA": 3857, + "GGAAGT": 3858, + "GGAAGC": 3859, + "GGAAGG": 3860, + "GGATAA": 3861, + "GGATAT": 3862, + "GGATAC": 3863, + "GGATAG": 3864, + "GGATTA": 3865, + "GGATTT": 3866, + "GGATTC": 3867, + "GGATTG": 3868, + "GGATCA": 3869, + "GGATCT": 3870, + "GGATCC": 3871, + "GGATCG": 3872, + "GGATGA": 3873, + "GGATGT": 3874, + "GGATGC": 3875, + "GGATGG": 3876, + "GGACAA": 3877, + "GGACAT": 3878, + "GGACAC": 3879, + "GGACAG": 3880, + "GGACTA": 3881, + "GGACTT": 3882, + "GGACTC": 3883, + "GGACTG": 3884, + "GGACCA": 3885, + "GGACCT": 3886, + "GGACCC": 3887, + "GGACCG": 3888, + "GGACGA": 3889, + "GGACGT": 3890, + "GGACGC": 3891, + "GGACGG": 3892, + "GGAGAA": 3893, + "GGAGAT": 3894, + "GGAGAC": 3895, + "GGAGAG": 3896, + "GGAGTA": 3897, + "GGAGTT": 3898, + "GGAGTC": 3899, + "GGAGTG": 3900, + "GGAGCA": 3901, + "GGAGCT": 3902, + "GGAGCC": 3903, + "GGAGCG": 3904, + "GGAGGA": 3905, + "GGAGGT": 3906, + "GGAGGC": 3907, + "GGAGGG": 3908, + "GGTAAA": 3909, + "GGTAAT": 3910, + "GGTAAC": 3911, + "GGTAAG": 3912, + "GGTATA": 3913, + "GGTATT": 3914, + "GGTATC": 3915, + "GGTATG": 3916, + "GGTACA": 3917, + "GGTACT": 3918, + "GGTACC": 3919, + "GGTACG": 3920, + "GGTAGA": 3921, + "GGTAGT": 3922, + "GGTAGC": 3923, + "GGTAGG": 3924, + "GGTTAA": 3925, + "GGTTAT": 3926, + "GGTTAC": 3927, + "GGTTAG": 3928, + "GGTTTA": 3929, + "GGTTTT": 3930, + "GGTTTC": 3931, + "GGTTTG": 3932, + "GGTTCA": 3933, + "GGTTCT": 3934, + "GGTTCC": 3935, + "GGTTCG": 3936, + "GGTTGA": 3937, + "GGTTGT": 3938, + "GGTTGC": 3939, + "GGTTGG": 3940, + "GGTCAA": 3941, + "GGTCAT": 3942, + "GGTCAC": 3943, + "GGTCAG": 3944, + "GGTCTA": 3945, + "GGTCTT": 3946, + "GGTCTC": 3947, + "GGTCTG": 3948, + "GGTCCA": 3949, + "GGTCCT": 3950, + "GGTCCC": 3951, + "GGTCCG": 3952, + "GGTCGA": 3953, + "GGTCGT": 3954, + "GGTCGC": 3955, + "GGTCGG": 3956, + "GGTGAA": 3957, + "GGTGAT": 3958, + "GGTGAC": 3959, + "GGTGAG": 3960, + "GGTGTA": 3961, + "GGTGTT": 3962, + "GGTGTC": 3963, + "GGTGTG": 3964, + "GGTGCA": 3965, + "GGTGCT": 3966, + "GGTGCC": 3967, + "GGTGCG": 3968, + "GGTGGA": 3969, + "GGTGGT": 3970, + "GGTGGC": 3971, + "GGTGGG": 3972, + "GGCAAA": 3973, + "GGCAAT": 3974, + "GGCAAC": 3975, + "GGCAAG": 3976, + "GGCATA": 3977, + "GGCATT": 3978, + "GGCATC": 3979, + "GGCATG": 3980, + "GGCACA": 3981, + "GGCACT": 3982, + "GGCACC": 3983, + "GGCACG": 3984, + "GGCAGA": 3985, + "GGCAGT": 3986, + "GGCAGC": 3987, + "GGCAGG": 3988, + "GGCTAA": 3989, + "GGCTAT": 3990, + "GGCTAC": 3991, + "GGCTAG": 3992, + "GGCTTA": 3993, + "GGCTTT": 3994, + "GGCTTC": 3995, + "GGCTTG": 3996, + "GGCTCA": 3997, + "GGCTCT": 3998, + "GGCTCC": 3999, + "GGCTCG": 4000, + "GGCTGA": 4001, + "GGCTGT": 4002, + "GGCTGC": 4003, + "GGCTGG": 4004, + "GGCCAA": 4005, + "GGCCAT": 4006, + "GGCCAC": 4007, + "GGCCAG": 4008, + "GGCCTA": 4009, + "GGCCTT": 4010, + "GGCCTC": 4011, + "GGCCTG": 4012, + "GGCCCA": 4013, + "GGCCCT": 4014, + "GGCCCC": 4015, + "GGCCCG": 4016, + "GGCCGA": 4017, + "GGCCGT": 4018, + "GGCCGC": 4019, + "GGCCGG": 4020, + "GGCGAA": 4021, + "GGCGAT": 4022, + "GGCGAC": 4023, + "GGCGAG": 4024, + "GGCGTA": 4025, + "GGCGTT": 4026, + "GGCGTC": 4027, + "GGCGTG": 4028, + "GGCGCA": 4029, + "GGCGCT": 4030, + "GGCGCC": 4031, + "GGCGCG": 4032, + "GGCGGA": 4033, + "GGCGGT": 4034, + "GGCGGC": 4035, + "GGCGGG": 4036, + "GGGAAA": 4037, + "GGGAAT": 4038, + "GGGAAC": 4039, + "GGGAAG": 4040, + "GGGATA": 4041, + "GGGATT": 4042, + "GGGATC": 4043, + "GGGATG": 4044, + "GGGACA": 4045, + "GGGACT": 4046, + "GGGACC": 4047, + "GGGACG": 4048, + "GGGAGA": 4049, + "GGGAGT": 4050, + "GGGAGC": 4051, + "GGGAGG": 4052, + "GGGTAA": 4053, + "GGGTAT": 4054, + "GGGTAC": 4055, + "GGGTAG": 4056, + "GGGTTA": 4057, + "GGGTTT": 4058, + "GGGTTC": 4059, + "GGGTTG": 4060, + "GGGTCA": 4061, + "GGGTCT": 4062, + "GGGTCC": 4063, + "GGGTCG": 4064, + "GGGTGA": 4065, + "GGGTGT": 4066, + "GGGTGC": 4067, + "GGGTGG": 4068, + "GGGCAA": 4069, + "GGGCAT": 4070, + "GGGCAC": 4071, + "GGGCAG": 4072, + "GGGCTA": 4073, + "GGGCTT": 4074, + "GGGCTC": 4075, + "GGGCTG": 4076, + "GGGCCA": 4077, + "GGGCCT": 4078, + "GGGCCC": 4079, + "GGGCCG": 4080, + "GGGCGA": 4081, + "GGGCGT": 4082, + "GGGCGC": 4083, + "GGGCGG": 4084, + "GGGGAA": 4085, + "GGGGAT": 4086, + "GGGGAC": 4087, + "GGGGAG": 4088, + "GGGGTA": 4089, + "GGGGTT": 4090, + "GGGGTC": 4091, + "GGGGTG": 4092, + "GGGGCA": 4093, + "GGGGCT": 4094, + "GGGGCC": 4095, + "GGGGCG": 4096, + "GGGGGA": 4097, + "GGGGGT": 4098, + "GGGGGC": 4099, + "GGGGGG": 4100 + } + } +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/tokenizer_config.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..604e2b1b29af1dcb3920d2828b6a90276e5e3d02 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/tokenizer_config.json @@ -0,0 +1,58 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "[UNK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "[CLS]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "[SEP]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "4": { + "content": "[MASK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "clean_up_tokenization_spaces": true, + "cls_token": "[CLS]", + "do_basic_tokenize": true, + "do_lower_case": false, + "mask_token": "[MASK]", + "max_len": 512, + "model_max_length": 512, + "never_split": null, + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "strip_accents": null, + "tokenize_chinese_chars": true, + "tokenizer_class": "BertTokenizer", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/trainer_state.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..f0abceb1ba72d951cf072928eebed90588cb00b6 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/trainer_state.json @@ -0,0 +1,593 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 8.0, + "eval_steps": 500, + "global_step": 2000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.1, + "grad_norm": 0.9416134357452393, + "learning_rate": 5.319148936170213e-05, + "loss": 0.6912, + "step": 25 + }, + { + "epoch": 0.2, + "grad_norm": 0.9283172488212585, + "learning_rate": 0.00010638297872340425, + "loss": 0.6691, + "step": 50 + }, + { + "epoch": 0.3, + "grad_norm": 0.5870445370674133, + "learning_rate": 0.00015957446808510637, + "loss": 0.6164, + "step": 75 + }, + { + "epoch": 0.4, + "grad_norm": 0.33945509791374207, + "learning_rate": 0.0002127659574468085, + "loss": 0.5453, + "step": 100 + }, + { + "epoch": 0.5, + "grad_norm": 0.21052460372447968, + "learning_rate": 0.00026595744680851064, + "loss": 0.5025, + "step": 125 + }, + { + "epoch": 0.6, + "grad_norm": 0.25948837399482727, + "learning_rate": 0.00031914893617021275, + "loss": 0.4881, + "step": 150 + }, + { + "epoch": 0.7, + "grad_norm": 0.6165992021560669, + "learning_rate": 0.0003723404255319149, + "loss": 0.4852, + "step": 175 + }, + { + "epoch": 0.8, + "grad_norm": 0.33296748995780945, + "learning_rate": 0.0003992081821181128, + "loss": 0.4867, + "step": 200 + }, + { + "epoch": 0.9, + "grad_norm": 0.40745800733566284, + "learning_rate": 0.0003975585615308479, + "loss": 0.4767, + "step": 225 + }, + { + "epoch": 1.0, + "grad_norm": 0.8070425391197205, + "learning_rate": 0.00039590894094358297, + "loss": 0.4784, + "step": 250 + }, + { + "epoch": 1.1, + "grad_norm": 1.2249579429626465, + "learning_rate": 0.0003942593203563181, + "loss": 0.4543, + "step": 275 + }, + { + "epoch": 1.2, + "grad_norm": 1.0143898725509644, + "learning_rate": 0.0003926096997690532, + "loss": 0.4319, + "step": 300 + }, + { + "epoch": 1.3, + "grad_norm": 2.3284952640533447, + "learning_rate": 0.00039096007918178817, + "loss": 0.4455, + "step": 325 + }, + { + "epoch": 1.4, + "grad_norm": 1.189352035522461, + "learning_rate": 0.0003893104585945233, + "loss": 0.4246, + "step": 350 + }, + { + "epoch": 1.5, + "grad_norm": 0.947853684425354, + "learning_rate": 0.0003876608380072583, + "loss": 0.4326, + "step": 375 + }, + { + "epoch": 1.6, + "grad_norm": 0.7130059003829956, + "learning_rate": 0.00038601121741999343, + "loss": 0.4034, + "step": 400 + }, + { + "epoch": 1.7, + "grad_norm": 3.517078161239624, + "learning_rate": 0.0003843615968327285, + "loss": 0.4114, + "step": 425 + }, + { + "epoch": 1.8, + "grad_norm": 0.3982729911804199, + "learning_rate": 0.0003827119762454636, + "loss": 0.4345, + "step": 450 + }, + { + "epoch": 1.9, + "grad_norm": 0.6208440065383911, + "learning_rate": 0.00038106235565819863, + "loss": 0.4135, + "step": 475 + }, + { + "epoch": 2.0, + "grad_norm": 0.9085566401481628, + "learning_rate": 0.0003794127350709337, + "loss": 0.3983, + "step": 500 + }, + { + "epoch": 2.1, + "grad_norm": 1.2225189208984375, + "learning_rate": 0.0003777631144836688, + "loss": 0.3915, + "step": 525 + }, + { + "epoch": 2.2, + "grad_norm": 1.1379530429840088, + "learning_rate": 0.00037611349389640383, + "loss": 0.4335, + "step": 550 + }, + { + "epoch": 2.3, + "grad_norm": 0.8535892367362976, + "learning_rate": 0.00037446387330913894, + "loss": 0.4063, + "step": 575 + }, + { + "epoch": 2.4, + "grad_norm": 1.3587419986724854, + "learning_rate": 0.000372814252721874, + "loss": 0.4065, + "step": 600 + }, + { + "epoch": 2.5, + "grad_norm": 0.9311383962631226, + "learning_rate": 0.00037116463213460903, + "loss": 0.4007, + "step": 625 + }, + { + "epoch": 2.6, + "grad_norm": 0.9912465810775757, + "learning_rate": 0.00036951501154734414, + "loss": 0.3992, + "step": 650 + }, + { + "epoch": 2.7, + "grad_norm": 1.2762119770050049, + "learning_rate": 0.0003678653909600792, + "loss": 0.3827, + "step": 675 + }, + { + "epoch": 2.8, + "grad_norm": 2.3506247997283936, + "learning_rate": 0.0003662157703728143, + "loss": 0.3734, + "step": 700 + }, + { + "epoch": 2.9, + "grad_norm": 1.5375425815582275, + "learning_rate": 0.00036456614978554934, + "loss": 0.4005, + "step": 725 + }, + { + "epoch": 3.0, + "grad_norm": 1.3143396377563477, + "learning_rate": 0.00036291652919828444, + "loss": 0.3548, + "step": 750 + }, + { + "epoch": 3.1, + "grad_norm": 0.662200391292572, + "learning_rate": 0.00036126690861101944, + "loss": 0.4129, + "step": 775 + }, + { + "epoch": 3.2, + "grad_norm": 1.3569878339767456, + "learning_rate": 0.00035961728802375454, + "loss": 0.38, + "step": 800 + }, + { + "epoch": 3.3, + "grad_norm": 0.9888588190078735, + "learning_rate": 0.0003579676674364896, + "loss": 0.3778, + "step": 825 + }, + { + "epoch": 3.4, + "grad_norm": 0.679969310760498, + "learning_rate": 0.0003563180468492247, + "loss": 0.3934, + "step": 850 + }, + { + "epoch": 3.5, + "grad_norm": 1.608539342880249, + "learning_rate": 0.0003546684262619598, + "loss": 0.418, + "step": 875 + }, + { + "epoch": 3.6, + "grad_norm": 1.7037910223007202, + "learning_rate": 0.00035301880567469485, + "loss": 0.3934, + "step": 900 + }, + { + "epoch": 3.7, + "grad_norm": 1.4417132139205933, + "learning_rate": 0.0003513691850874299, + "loss": 0.3947, + "step": 925 + }, + { + "epoch": 3.8, + "grad_norm": 1.1482582092285156, + "learning_rate": 0.00034971956450016495, + "loss": 0.3621, + "step": 950 + }, + { + "epoch": 3.9, + "grad_norm": 2.046616792678833, + "learning_rate": 0.00034806994391290005, + "loss": 0.3774, + "step": 975 + }, + { + "epoch": 4.0, + "grad_norm": 1.0832113027572632, + "learning_rate": 0.0003464203233256351, + "loss": 0.3542, + "step": 1000 + }, + { + "epoch": 4.1, + "grad_norm": 0.8189190626144409, + "learning_rate": 0.0003447707027383702, + "loss": 0.381, + "step": 1025 + }, + { + "epoch": 4.2, + "grad_norm": 1.6974774599075317, + "learning_rate": 0.00034312108215110525, + "loss": 0.3775, + "step": 1050 + }, + { + "epoch": 4.3, + "grad_norm": 1.2468841075897217, + "learning_rate": 0.0003414714615638403, + "loss": 0.3527, + "step": 1075 + }, + { + "epoch": 4.4, + "grad_norm": 0.9022793173789978, + "learning_rate": 0.0003398218409765754, + "loss": 0.3637, + "step": 1100 + }, + { + "epoch": 4.5, + "grad_norm": 1.121774673461914, + "learning_rate": 0.00033817222038931045, + "loss": 0.3935, + "step": 1125 + }, + { + "epoch": 4.6, + "grad_norm": 0.810937762260437, + "learning_rate": 0.00033652259980204556, + "loss": 0.4103, + "step": 1150 + }, + { + "epoch": 4.7, + "grad_norm": 0.7376565337181091, + "learning_rate": 0.0003348729792147806, + "loss": 0.362, + "step": 1175 + }, + { + "epoch": 4.8, + "grad_norm": 1.7271156311035156, + "learning_rate": 0.0003332233586275157, + "loss": 0.3597, + "step": 1200 + }, + { + "epoch": 4.9, + "grad_norm": 1.2273073196411133, + "learning_rate": 0.00033157373804025076, + "loss": 0.3853, + "step": 1225 + }, + { + "epoch": 5.0, + "grad_norm": 0.9998250603675842, + "learning_rate": 0.0003299241174529858, + "loss": 0.3461, + "step": 1250 + }, + { + "epoch": 5.1, + "grad_norm": 0.9112564325332642, + "learning_rate": 0.0003282744968657209, + "loss": 0.3659, + "step": 1275 + }, + { + "epoch": 5.2, + "grad_norm": 3.246051073074341, + "learning_rate": 0.00032662487627845596, + "loss": 0.3488, + "step": 1300 + }, + { + "epoch": 5.3, + "grad_norm": 1.0283201932907104, + "learning_rate": 0.00032497525569119106, + "loss": 0.3528, + "step": 1325 + }, + { + "epoch": 5.4, + "grad_norm": 1.112660527229309, + "learning_rate": 0.0003233256351039261, + "loss": 0.3723, + "step": 1350 + }, + { + "epoch": 5.5, + "grad_norm": 0.9069312810897827, + "learning_rate": 0.00032167601451666116, + "loss": 0.3416, + "step": 1375 + }, + { + "epoch": 5.6, + "grad_norm": 0.710412323474884, + "learning_rate": 0.00032002639392939627, + "loss": 0.3607, + "step": 1400 + }, + { + "epoch": 5.7, + "grad_norm": 1.185612678527832, + "learning_rate": 0.0003183767733421313, + "loss": 0.3777, + "step": 1425 + }, + { + "epoch": 5.8, + "grad_norm": 1.340174913406372, + "learning_rate": 0.0003167271527548664, + "loss": 0.3563, + "step": 1450 + }, + { + "epoch": 5.9, + "grad_norm": 0.9387012124061584, + "learning_rate": 0.00031507753216760147, + "loss": 0.3647, + "step": 1475 + }, + { + "epoch": 6.0, + "grad_norm": 1.1005347967147827, + "learning_rate": 0.00031342791158033657, + "loss": 0.3484, + "step": 1500 + }, + { + "epoch": 6.1, + "grad_norm": 2.64058780670166, + "learning_rate": 0.00031177829099307157, + "loss": 0.3542, + "step": 1525 + }, + { + "epoch": 6.2, + "grad_norm": 1.3007783889770508, + "learning_rate": 0.00031012867040580667, + "loss": 0.3379, + "step": 1550 + }, + { + "epoch": 6.3, + "grad_norm": 3.7190418243408203, + "learning_rate": 0.0003084790498185417, + "loss": 0.3514, + "step": 1575 + }, + { + "epoch": 6.4, + "grad_norm": 1.7263462543487549, + "learning_rate": 0.0003068294292312768, + "loss": 0.3434, + "step": 1600 + }, + { + "epoch": 6.5, + "grad_norm": 3.3360602855682373, + "learning_rate": 0.0003051798086440119, + "loss": 0.3507, + "step": 1625 + }, + { + "epoch": 6.6, + "grad_norm": 1.9620813131332397, + "learning_rate": 0.000303530188056747, + "loss": 0.3455, + "step": 1650 + }, + { + "epoch": 6.7, + "grad_norm": 1.3080660104751587, + "learning_rate": 0.000301880567469482, + "loss": 0.3678, + "step": 1675 + }, + { + "epoch": 6.8, + "grad_norm": 0.6183618307113647, + "learning_rate": 0.0003002309468822171, + "loss": 0.3704, + "step": 1700 + }, + { + "epoch": 6.9, + "grad_norm": 1.0411962270736694, + "learning_rate": 0.0002985813262949522, + "loss": 0.3505, + "step": 1725 + }, + { + "epoch": 7.0, + "grad_norm": 2.040569543838501, + "learning_rate": 0.00029693170570768723, + "loss": 0.3666, + "step": 1750 + }, + { + "epoch": 7.1, + "grad_norm": 0.713647723197937, + "learning_rate": 0.00029528208512042233, + "loss": 0.3529, + "step": 1775 + }, + { + "epoch": 7.2, + "grad_norm": 3.0373058319091797, + "learning_rate": 0.0002936324645331574, + "loss": 0.361, + "step": 1800 + }, + { + "epoch": 7.3, + "grad_norm": 1.819886326789856, + "learning_rate": 0.00029198284394589243, + "loss": 0.3482, + "step": 1825 + }, + { + "epoch": 7.4, + "grad_norm": 0.7930757403373718, + "learning_rate": 0.00029033322335862753, + "loss": 0.3695, + "step": 1850 + }, + { + "epoch": 7.5, + "grad_norm": 0.7889612913131714, + "learning_rate": 0.0002886836027713626, + "loss": 0.3527, + "step": 1875 + }, + { + "epoch": 7.6, + "grad_norm": 2.4007863998413086, + "learning_rate": 0.0002870339821840977, + "loss": 0.3275, + "step": 1900 + }, + { + "epoch": 7.7, + "grad_norm": 2.3483331203460693, + "learning_rate": 0.00028538436159683273, + "loss": 0.3571, + "step": 1925 + }, + { + "epoch": 7.8, + "grad_norm": 0.8911663293838501, + "learning_rate": 0.00028373474100956784, + "loss": 0.3079, + "step": 1950 + }, + { + "epoch": 7.9, + "grad_norm": 1.0521246194839478, + "learning_rate": 0.0002820851204223029, + "loss": 0.3578, + "step": 1975 + }, + { + "epoch": 8.0, + "grad_norm": 1.3923832178115845, + "learning_rate": 0.00028043549983503794, + "loss": 0.3588, + "step": 2000 + } + ], + "logging_steps": 25, + "max_steps": 6250, + "num_input_tokens_seen": 0, + "num_train_epochs": 25, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 106232217600000.0, + "train_batch_size": 20, + "trial_name": null, + "trial_params": null +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/training_args.bin b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..d61f3500eb70763e1f33496c1787bdf3f07a651b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b6e0f7aa622786274a1ea4195b72f4da224050767f7b89529b85721865c5075a +size 5304 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/vocab.txt b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/vocab.txt new file mode 100644 index 0000000000000000000000000000000000000000..cc9ad53732e21ea9180f18e1fb67a38cc5702d4c --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/vocab.txt @@ -0,0 +1,4101 @@ +[PAD] +[UNK] +[CLS] +[SEP] +[MASK] +AAAAAA +AAAAAT +AAAAAC +AAAAAG +AAAATA +AAAATT +AAAATC +AAAATG +AAAACA +AAAACT +AAAACC +AAAACG +AAAAGA +AAAAGT +AAAAGC +AAAAGG +AAATAA +AAATAT +AAATAC +AAATAG +AAATTA +AAATTT +AAATTC +AAATTG +AAATCA +AAATCT +AAATCC +AAATCG +AAATGA +AAATGT +AAATGC +AAATGG +AAACAA +AAACAT +AAACAC +AAACAG +AAACTA +AAACTT +AAACTC +AAACTG +AAACCA +AAACCT +AAACCC +AAACCG +AAACGA +AAACGT +AAACGC +AAACGG +AAAGAA +AAAGAT +AAAGAC +AAAGAG +AAAGTA +AAAGTT +AAAGTC +AAAGTG +AAAGCA +AAAGCT +AAAGCC +AAAGCG +AAAGGA +AAAGGT +AAAGGC +AAAGGG +AATAAA +AATAAT +AATAAC +AATAAG +AATATA +AATATT +AATATC +AATATG +AATACA +AATACT +AATACC +AATACG +AATAGA +AATAGT +AATAGC +AATAGG +AATTAA +AATTAT +AATTAC +AATTAG +AATTTA +AATTTT +AATTTC +AATTTG +AATTCA +AATTCT +AATTCC +AATTCG +AATTGA +AATTGT +AATTGC +AATTGG +AATCAA +AATCAT +AATCAC +AATCAG +AATCTA +AATCTT +AATCTC +AATCTG +AATCCA +AATCCT +AATCCC +AATCCG +AATCGA +AATCGT +AATCGC +AATCGG +AATGAA +AATGAT +AATGAC +AATGAG +AATGTA +AATGTT +AATGTC +AATGTG +AATGCA +AATGCT +AATGCC +AATGCG +AATGGA +AATGGT +AATGGC +AATGGG +AACAAA +AACAAT +AACAAC +AACAAG +AACATA +AACATT +AACATC +AACATG +AACACA +AACACT +AACACC +AACACG +AACAGA +AACAGT +AACAGC +AACAGG +AACTAA +AACTAT +AACTAC +AACTAG +AACTTA +AACTTT +AACTTC +AACTTG +AACTCA +AACTCT +AACTCC +AACTCG +AACTGA +AACTGT +AACTGC +AACTGG +AACCAA +AACCAT +AACCAC +AACCAG +AACCTA +AACCTT +AACCTC +AACCTG +AACCCA +AACCCT +AACCCC +AACCCG +AACCGA +AACCGT +AACCGC +AACCGG +AACGAA +AACGAT +AACGAC +AACGAG +AACGTA +AACGTT +AACGTC +AACGTG +AACGCA +AACGCT +AACGCC +AACGCG +AACGGA +AACGGT +AACGGC +AACGGG +AAGAAA +AAGAAT +AAGAAC +AAGAAG +AAGATA +AAGATT +AAGATC +AAGATG +AAGACA +AAGACT +AAGACC +AAGACG +AAGAGA +AAGAGT +AAGAGC +AAGAGG +AAGTAA +AAGTAT +AAGTAC +AAGTAG +AAGTTA +AAGTTT +AAGTTC +AAGTTG +AAGTCA +AAGTCT +AAGTCC +AAGTCG +AAGTGA +AAGTGT +AAGTGC +AAGTGG +AAGCAA +AAGCAT +AAGCAC +AAGCAG +AAGCTA +AAGCTT +AAGCTC +AAGCTG +AAGCCA +AAGCCT +AAGCCC +AAGCCG +AAGCGA +AAGCGT +AAGCGC +AAGCGG +AAGGAA +AAGGAT +AAGGAC +AAGGAG +AAGGTA +AAGGTT +AAGGTC +AAGGTG +AAGGCA +AAGGCT +AAGGCC +AAGGCG +AAGGGA +AAGGGT +AAGGGC +AAGGGG +ATAAAA +ATAAAT +ATAAAC +ATAAAG +ATAATA +ATAATT +ATAATC +ATAATG +ATAACA +ATAACT +ATAACC +ATAACG +ATAAGA +ATAAGT +ATAAGC +ATAAGG +ATATAA +ATATAT +ATATAC +ATATAG +ATATTA +ATATTT +ATATTC +ATATTG +ATATCA +ATATCT +ATATCC +ATATCG +ATATGA +ATATGT +ATATGC +ATATGG +ATACAA +ATACAT +ATACAC +ATACAG +ATACTA +ATACTT +ATACTC +ATACTG +ATACCA +ATACCT +ATACCC +ATACCG +ATACGA +ATACGT +ATACGC +ATACGG +ATAGAA +ATAGAT +ATAGAC +ATAGAG +ATAGTA +ATAGTT +ATAGTC +ATAGTG +ATAGCA +ATAGCT +ATAGCC +ATAGCG +ATAGGA +ATAGGT +ATAGGC +ATAGGG +ATTAAA +ATTAAT +ATTAAC +ATTAAG +ATTATA +ATTATT +ATTATC +ATTATG +ATTACA +ATTACT +ATTACC +ATTACG +ATTAGA +ATTAGT +ATTAGC +ATTAGG +ATTTAA +ATTTAT +ATTTAC +ATTTAG +ATTTTA +ATTTTT +ATTTTC +ATTTTG +ATTTCA +ATTTCT +ATTTCC +ATTTCG +ATTTGA +ATTTGT +ATTTGC +ATTTGG +ATTCAA +ATTCAT +ATTCAC +ATTCAG +ATTCTA +ATTCTT +ATTCTC +ATTCTG +ATTCCA +ATTCCT +ATTCCC +ATTCCG +ATTCGA +ATTCGT +ATTCGC +ATTCGG +ATTGAA +ATTGAT +ATTGAC +ATTGAG +ATTGTA +ATTGTT +ATTGTC +ATTGTG +ATTGCA +ATTGCT +ATTGCC +ATTGCG +ATTGGA +ATTGGT +ATTGGC +ATTGGG +ATCAAA +ATCAAT +ATCAAC +ATCAAG +ATCATA +ATCATT +ATCATC +ATCATG +ATCACA +ATCACT +ATCACC +ATCACG +ATCAGA +ATCAGT +ATCAGC +ATCAGG +ATCTAA +ATCTAT +ATCTAC +ATCTAG +ATCTTA +ATCTTT +ATCTTC +ATCTTG +ATCTCA +ATCTCT +ATCTCC +ATCTCG +ATCTGA +ATCTGT +ATCTGC +ATCTGG +ATCCAA +ATCCAT +ATCCAC +ATCCAG +ATCCTA +ATCCTT +ATCCTC +ATCCTG +ATCCCA +ATCCCT +ATCCCC +ATCCCG +ATCCGA +ATCCGT +ATCCGC +ATCCGG +ATCGAA +ATCGAT +ATCGAC +ATCGAG +ATCGTA +ATCGTT +ATCGTC +ATCGTG +ATCGCA +ATCGCT +ATCGCC +ATCGCG +ATCGGA +ATCGGT +ATCGGC +ATCGGG +ATGAAA +ATGAAT +ATGAAC +ATGAAG +ATGATA +ATGATT +ATGATC +ATGATG +ATGACA +ATGACT +ATGACC +ATGACG +ATGAGA +ATGAGT +ATGAGC +ATGAGG +ATGTAA +ATGTAT +ATGTAC +ATGTAG +ATGTTA +ATGTTT +ATGTTC +ATGTTG +ATGTCA +ATGTCT +ATGTCC +ATGTCG +ATGTGA +ATGTGT +ATGTGC +ATGTGG +ATGCAA +ATGCAT +ATGCAC +ATGCAG +ATGCTA +ATGCTT +ATGCTC +ATGCTG +ATGCCA +ATGCCT +ATGCCC +ATGCCG +ATGCGA +ATGCGT +ATGCGC +ATGCGG +ATGGAA +ATGGAT +ATGGAC +ATGGAG +ATGGTA +ATGGTT +ATGGTC +ATGGTG +ATGGCA +ATGGCT +ATGGCC +ATGGCG +ATGGGA +ATGGGT +ATGGGC +ATGGGG +ACAAAA +ACAAAT +ACAAAC +ACAAAG +ACAATA +ACAATT +ACAATC +ACAATG +ACAACA +ACAACT +ACAACC +ACAACG +ACAAGA +ACAAGT +ACAAGC +ACAAGG +ACATAA +ACATAT +ACATAC +ACATAG +ACATTA +ACATTT +ACATTC +ACATTG +ACATCA +ACATCT +ACATCC +ACATCG +ACATGA +ACATGT +ACATGC +ACATGG +ACACAA +ACACAT +ACACAC +ACACAG +ACACTA +ACACTT +ACACTC +ACACTG +ACACCA +ACACCT +ACACCC +ACACCG +ACACGA +ACACGT +ACACGC +ACACGG +ACAGAA +ACAGAT +ACAGAC +ACAGAG +ACAGTA +ACAGTT +ACAGTC +ACAGTG +ACAGCA +ACAGCT +ACAGCC +ACAGCG +ACAGGA +ACAGGT +ACAGGC +ACAGGG +ACTAAA +ACTAAT +ACTAAC +ACTAAG +ACTATA +ACTATT +ACTATC +ACTATG +ACTACA +ACTACT +ACTACC +ACTACG +ACTAGA +ACTAGT +ACTAGC +ACTAGG +ACTTAA +ACTTAT +ACTTAC +ACTTAG +ACTTTA +ACTTTT +ACTTTC +ACTTTG +ACTTCA +ACTTCT +ACTTCC +ACTTCG +ACTTGA +ACTTGT +ACTTGC +ACTTGG +ACTCAA +ACTCAT +ACTCAC +ACTCAG +ACTCTA +ACTCTT +ACTCTC +ACTCTG +ACTCCA +ACTCCT +ACTCCC +ACTCCG +ACTCGA +ACTCGT +ACTCGC +ACTCGG +ACTGAA +ACTGAT +ACTGAC +ACTGAG +ACTGTA +ACTGTT +ACTGTC +ACTGTG +ACTGCA +ACTGCT +ACTGCC +ACTGCG +ACTGGA +ACTGGT +ACTGGC +ACTGGG +ACCAAA +ACCAAT +ACCAAC +ACCAAG +ACCATA +ACCATT +ACCATC +ACCATG +ACCACA +ACCACT +ACCACC +ACCACG +ACCAGA +ACCAGT +ACCAGC +ACCAGG +ACCTAA +ACCTAT +ACCTAC +ACCTAG +ACCTTA +ACCTTT +ACCTTC +ACCTTG +ACCTCA +ACCTCT +ACCTCC +ACCTCG +ACCTGA +ACCTGT +ACCTGC +ACCTGG +ACCCAA +ACCCAT +ACCCAC +ACCCAG +ACCCTA +ACCCTT +ACCCTC +ACCCTG +ACCCCA +ACCCCT +ACCCCC +ACCCCG +ACCCGA +ACCCGT +ACCCGC +ACCCGG +ACCGAA +ACCGAT +ACCGAC +ACCGAG +ACCGTA +ACCGTT +ACCGTC +ACCGTG +ACCGCA +ACCGCT +ACCGCC +ACCGCG +ACCGGA +ACCGGT +ACCGGC +ACCGGG +ACGAAA +ACGAAT +ACGAAC +ACGAAG +ACGATA +ACGATT +ACGATC +ACGATG +ACGACA +ACGACT +ACGACC +ACGACG +ACGAGA +ACGAGT +ACGAGC +ACGAGG +ACGTAA +ACGTAT +ACGTAC +ACGTAG +ACGTTA +ACGTTT +ACGTTC +ACGTTG +ACGTCA +ACGTCT +ACGTCC +ACGTCG +ACGTGA +ACGTGT +ACGTGC +ACGTGG +ACGCAA +ACGCAT +ACGCAC +ACGCAG +ACGCTA +ACGCTT +ACGCTC +ACGCTG +ACGCCA +ACGCCT +ACGCCC +ACGCCG +ACGCGA +ACGCGT +ACGCGC +ACGCGG +ACGGAA +ACGGAT +ACGGAC +ACGGAG +ACGGTA +ACGGTT +ACGGTC +ACGGTG +ACGGCA +ACGGCT +ACGGCC +ACGGCG +ACGGGA +ACGGGT +ACGGGC +ACGGGG +AGAAAA +AGAAAT +AGAAAC +AGAAAG +AGAATA +AGAATT +AGAATC +AGAATG +AGAACA +AGAACT +AGAACC +AGAACG +AGAAGA +AGAAGT +AGAAGC +AGAAGG +AGATAA +AGATAT +AGATAC +AGATAG +AGATTA +AGATTT +AGATTC +AGATTG +AGATCA +AGATCT +AGATCC +AGATCG +AGATGA +AGATGT +AGATGC +AGATGG +AGACAA +AGACAT +AGACAC +AGACAG +AGACTA +AGACTT +AGACTC +AGACTG +AGACCA +AGACCT +AGACCC +AGACCG +AGACGA +AGACGT +AGACGC +AGACGG +AGAGAA +AGAGAT +AGAGAC +AGAGAG +AGAGTA +AGAGTT +AGAGTC +AGAGTG +AGAGCA +AGAGCT +AGAGCC +AGAGCG +AGAGGA +AGAGGT +AGAGGC +AGAGGG +AGTAAA +AGTAAT +AGTAAC +AGTAAG +AGTATA +AGTATT +AGTATC +AGTATG +AGTACA +AGTACT +AGTACC +AGTACG +AGTAGA +AGTAGT +AGTAGC +AGTAGG +AGTTAA +AGTTAT +AGTTAC +AGTTAG +AGTTTA +AGTTTT +AGTTTC +AGTTTG +AGTTCA +AGTTCT +AGTTCC +AGTTCG +AGTTGA +AGTTGT +AGTTGC +AGTTGG +AGTCAA +AGTCAT +AGTCAC +AGTCAG +AGTCTA +AGTCTT +AGTCTC +AGTCTG +AGTCCA +AGTCCT +AGTCCC +AGTCCG +AGTCGA +AGTCGT +AGTCGC +AGTCGG +AGTGAA +AGTGAT +AGTGAC +AGTGAG +AGTGTA +AGTGTT +AGTGTC +AGTGTG +AGTGCA +AGTGCT +AGTGCC +AGTGCG +AGTGGA +AGTGGT +AGTGGC +AGTGGG +AGCAAA +AGCAAT +AGCAAC +AGCAAG +AGCATA +AGCATT +AGCATC +AGCATG +AGCACA +AGCACT +AGCACC +AGCACG +AGCAGA +AGCAGT +AGCAGC +AGCAGG +AGCTAA +AGCTAT +AGCTAC +AGCTAG +AGCTTA +AGCTTT +AGCTTC +AGCTTG +AGCTCA +AGCTCT +AGCTCC +AGCTCG +AGCTGA +AGCTGT +AGCTGC +AGCTGG +AGCCAA +AGCCAT +AGCCAC +AGCCAG +AGCCTA +AGCCTT +AGCCTC +AGCCTG +AGCCCA +AGCCCT +AGCCCC +AGCCCG +AGCCGA +AGCCGT +AGCCGC +AGCCGG +AGCGAA +AGCGAT +AGCGAC +AGCGAG +AGCGTA +AGCGTT +AGCGTC +AGCGTG +AGCGCA +AGCGCT +AGCGCC +AGCGCG +AGCGGA +AGCGGT +AGCGGC +AGCGGG +AGGAAA +AGGAAT +AGGAAC +AGGAAG +AGGATA +AGGATT +AGGATC +AGGATG +AGGACA +AGGACT +AGGACC +AGGACG +AGGAGA +AGGAGT +AGGAGC +AGGAGG +AGGTAA +AGGTAT +AGGTAC +AGGTAG +AGGTTA +AGGTTT +AGGTTC +AGGTTG +AGGTCA +AGGTCT +AGGTCC +AGGTCG +AGGTGA +AGGTGT +AGGTGC +AGGTGG +AGGCAA +AGGCAT +AGGCAC +AGGCAG +AGGCTA +AGGCTT +AGGCTC +AGGCTG +AGGCCA +AGGCCT +AGGCCC +AGGCCG +AGGCGA +AGGCGT +AGGCGC +AGGCGG +AGGGAA +AGGGAT +AGGGAC +AGGGAG +AGGGTA +AGGGTT +AGGGTC +AGGGTG +AGGGCA +AGGGCT +AGGGCC +AGGGCG +AGGGGA +AGGGGT +AGGGGC +AGGGGG +TAAAAA +TAAAAT +TAAAAC +TAAAAG +TAAATA +TAAATT +TAAATC +TAAATG +TAAACA +TAAACT +TAAACC +TAAACG +TAAAGA +TAAAGT +TAAAGC +TAAAGG +TAATAA +TAATAT +TAATAC +TAATAG +TAATTA +TAATTT +TAATTC +TAATTG +TAATCA +TAATCT +TAATCC +TAATCG +TAATGA +TAATGT +TAATGC +TAATGG +TAACAA +TAACAT +TAACAC +TAACAG +TAACTA +TAACTT +TAACTC +TAACTG +TAACCA +TAACCT +TAACCC +TAACCG +TAACGA +TAACGT +TAACGC +TAACGG +TAAGAA +TAAGAT +TAAGAC +TAAGAG +TAAGTA +TAAGTT +TAAGTC +TAAGTG +TAAGCA +TAAGCT +TAAGCC +TAAGCG +TAAGGA +TAAGGT +TAAGGC +TAAGGG +TATAAA +TATAAT +TATAAC +TATAAG +TATATA +TATATT +TATATC +TATATG +TATACA +TATACT +TATACC +TATACG +TATAGA +TATAGT +TATAGC +TATAGG +TATTAA +TATTAT +TATTAC +TATTAG +TATTTA +TATTTT +TATTTC +TATTTG +TATTCA +TATTCT +TATTCC +TATTCG +TATTGA +TATTGT +TATTGC +TATTGG +TATCAA +TATCAT +TATCAC +TATCAG +TATCTA +TATCTT +TATCTC +TATCTG +TATCCA +TATCCT +TATCCC +TATCCG +TATCGA +TATCGT +TATCGC +TATCGG +TATGAA +TATGAT +TATGAC +TATGAG +TATGTA +TATGTT +TATGTC +TATGTG +TATGCA +TATGCT +TATGCC +TATGCG +TATGGA +TATGGT +TATGGC +TATGGG +TACAAA +TACAAT +TACAAC +TACAAG +TACATA +TACATT +TACATC +TACATG +TACACA +TACACT +TACACC +TACACG +TACAGA +TACAGT +TACAGC +TACAGG +TACTAA +TACTAT +TACTAC +TACTAG +TACTTA +TACTTT +TACTTC +TACTTG +TACTCA +TACTCT +TACTCC +TACTCG +TACTGA +TACTGT +TACTGC +TACTGG +TACCAA +TACCAT +TACCAC +TACCAG +TACCTA +TACCTT +TACCTC +TACCTG +TACCCA +TACCCT +TACCCC +TACCCG +TACCGA +TACCGT +TACCGC +TACCGG +TACGAA +TACGAT +TACGAC +TACGAG +TACGTA +TACGTT +TACGTC +TACGTG +TACGCA +TACGCT +TACGCC +TACGCG +TACGGA +TACGGT +TACGGC +TACGGG +TAGAAA +TAGAAT +TAGAAC +TAGAAG +TAGATA +TAGATT +TAGATC +TAGATG +TAGACA +TAGACT +TAGACC +TAGACG +TAGAGA +TAGAGT +TAGAGC +TAGAGG +TAGTAA +TAGTAT +TAGTAC +TAGTAG +TAGTTA +TAGTTT +TAGTTC +TAGTTG +TAGTCA +TAGTCT +TAGTCC +TAGTCG +TAGTGA +TAGTGT +TAGTGC +TAGTGG +TAGCAA +TAGCAT +TAGCAC +TAGCAG +TAGCTA +TAGCTT +TAGCTC +TAGCTG +TAGCCA +TAGCCT +TAGCCC +TAGCCG +TAGCGA +TAGCGT +TAGCGC +TAGCGG +TAGGAA +TAGGAT +TAGGAC +TAGGAG +TAGGTA +TAGGTT +TAGGTC +TAGGTG +TAGGCA +TAGGCT +TAGGCC +TAGGCG +TAGGGA +TAGGGT +TAGGGC +TAGGGG +TTAAAA +TTAAAT +TTAAAC +TTAAAG +TTAATA +TTAATT +TTAATC +TTAATG +TTAACA +TTAACT +TTAACC +TTAACG +TTAAGA +TTAAGT +TTAAGC +TTAAGG +TTATAA +TTATAT +TTATAC +TTATAG +TTATTA +TTATTT +TTATTC +TTATTG +TTATCA +TTATCT +TTATCC +TTATCG +TTATGA +TTATGT +TTATGC +TTATGG +TTACAA +TTACAT +TTACAC +TTACAG +TTACTA +TTACTT +TTACTC +TTACTG +TTACCA +TTACCT +TTACCC +TTACCG +TTACGA +TTACGT +TTACGC +TTACGG +TTAGAA +TTAGAT +TTAGAC +TTAGAG +TTAGTA +TTAGTT +TTAGTC +TTAGTG +TTAGCA +TTAGCT +TTAGCC +TTAGCG +TTAGGA +TTAGGT +TTAGGC +TTAGGG +TTTAAA +TTTAAT +TTTAAC +TTTAAG +TTTATA +TTTATT +TTTATC +TTTATG +TTTACA +TTTACT +TTTACC +TTTACG +TTTAGA +TTTAGT +TTTAGC +TTTAGG +TTTTAA +TTTTAT +TTTTAC +TTTTAG +TTTTTA +TTTTTT +TTTTTC +TTTTTG +TTTTCA +TTTTCT +TTTTCC +TTTTCG +TTTTGA +TTTTGT +TTTTGC +TTTTGG +TTTCAA +TTTCAT +TTTCAC +TTTCAG +TTTCTA +TTTCTT +TTTCTC +TTTCTG +TTTCCA +TTTCCT +TTTCCC +TTTCCG +TTTCGA +TTTCGT +TTTCGC +TTTCGG +TTTGAA +TTTGAT +TTTGAC +TTTGAG +TTTGTA +TTTGTT +TTTGTC +TTTGTG +TTTGCA +TTTGCT +TTTGCC +TTTGCG +TTTGGA +TTTGGT +TTTGGC +TTTGGG +TTCAAA +TTCAAT +TTCAAC +TTCAAG +TTCATA +TTCATT +TTCATC +TTCATG +TTCACA +TTCACT +TTCACC +TTCACG +TTCAGA +TTCAGT +TTCAGC +TTCAGG +TTCTAA +TTCTAT +TTCTAC +TTCTAG +TTCTTA +TTCTTT +TTCTTC +TTCTTG +TTCTCA +TTCTCT +TTCTCC +TTCTCG +TTCTGA +TTCTGT +TTCTGC +TTCTGG +TTCCAA +TTCCAT +TTCCAC +TTCCAG +TTCCTA +TTCCTT +TTCCTC +TTCCTG +TTCCCA +TTCCCT +TTCCCC +TTCCCG +TTCCGA +TTCCGT +TTCCGC +TTCCGG +TTCGAA +TTCGAT +TTCGAC +TTCGAG +TTCGTA +TTCGTT +TTCGTC +TTCGTG +TTCGCA +TTCGCT +TTCGCC +TTCGCG +TTCGGA +TTCGGT +TTCGGC +TTCGGG +TTGAAA +TTGAAT +TTGAAC +TTGAAG +TTGATA +TTGATT +TTGATC +TTGATG +TTGACA +TTGACT +TTGACC +TTGACG +TTGAGA +TTGAGT +TTGAGC +TTGAGG +TTGTAA +TTGTAT +TTGTAC +TTGTAG +TTGTTA +TTGTTT +TTGTTC +TTGTTG +TTGTCA +TTGTCT +TTGTCC +TTGTCG +TTGTGA +TTGTGT +TTGTGC +TTGTGG +TTGCAA +TTGCAT +TTGCAC +TTGCAG +TTGCTA +TTGCTT +TTGCTC +TTGCTG +TTGCCA +TTGCCT +TTGCCC +TTGCCG +TTGCGA +TTGCGT +TTGCGC +TTGCGG +TTGGAA +TTGGAT +TTGGAC +TTGGAG +TTGGTA +TTGGTT +TTGGTC +TTGGTG +TTGGCA +TTGGCT +TTGGCC +TTGGCG +TTGGGA +TTGGGT +TTGGGC +TTGGGG +TCAAAA +TCAAAT +TCAAAC +TCAAAG +TCAATA +TCAATT +TCAATC +TCAATG +TCAACA +TCAACT +TCAACC +TCAACG +TCAAGA +TCAAGT +TCAAGC +TCAAGG +TCATAA +TCATAT +TCATAC +TCATAG +TCATTA +TCATTT +TCATTC +TCATTG +TCATCA +TCATCT +TCATCC +TCATCG +TCATGA +TCATGT +TCATGC +TCATGG +TCACAA +TCACAT +TCACAC +TCACAG +TCACTA +TCACTT +TCACTC +TCACTG +TCACCA +TCACCT +TCACCC +TCACCG +TCACGA +TCACGT +TCACGC +TCACGG +TCAGAA +TCAGAT +TCAGAC +TCAGAG +TCAGTA +TCAGTT +TCAGTC +TCAGTG +TCAGCA +TCAGCT +TCAGCC +TCAGCG +TCAGGA +TCAGGT +TCAGGC +TCAGGG +TCTAAA +TCTAAT +TCTAAC +TCTAAG +TCTATA +TCTATT +TCTATC +TCTATG +TCTACA +TCTACT +TCTACC +TCTACG +TCTAGA +TCTAGT +TCTAGC +TCTAGG +TCTTAA +TCTTAT +TCTTAC +TCTTAG +TCTTTA +TCTTTT +TCTTTC +TCTTTG +TCTTCA +TCTTCT +TCTTCC +TCTTCG +TCTTGA +TCTTGT +TCTTGC +TCTTGG +TCTCAA +TCTCAT +TCTCAC +TCTCAG +TCTCTA +TCTCTT +TCTCTC +TCTCTG +TCTCCA +TCTCCT +TCTCCC +TCTCCG +TCTCGA +TCTCGT +TCTCGC +TCTCGG +TCTGAA +TCTGAT +TCTGAC +TCTGAG +TCTGTA +TCTGTT +TCTGTC +TCTGTG +TCTGCA +TCTGCT +TCTGCC +TCTGCG +TCTGGA +TCTGGT +TCTGGC +TCTGGG +TCCAAA +TCCAAT +TCCAAC +TCCAAG +TCCATA +TCCATT +TCCATC +TCCATG +TCCACA +TCCACT +TCCACC +TCCACG +TCCAGA +TCCAGT +TCCAGC +TCCAGG +TCCTAA +TCCTAT +TCCTAC +TCCTAG +TCCTTA +TCCTTT +TCCTTC +TCCTTG +TCCTCA +TCCTCT +TCCTCC +TCCTCG +TCCTGA +TCCTGT +TCCTGC +TCCTGG +TCCCAA +TCCCAT +TCCCAC +TCCCAG +TCCCTA +TCCCTT +TCCCTC +TCCCTG +TCCCCA +TCCCCT +TCCCCC +TCCCCG +TCCCGA +TCCCGT +TCCCGC +TCCCGG +TCCGAA +TCCGAT +TCCGAC +TCCGAG +TCCGTA +TCCGTT +TCCGTC +TCCGTG +TCCGCA +TCCGCT +TCCGCC +TCCGCG +TCCGGA +TCCGGT +TCCGGC +TCCGGG +TCGAAA +TCGAAT +TCGAAC +TCGAAG +TCGATA +TCGATT +TCGATC +TCGATG +TCGACA +TCGACT +TCGACC +TCGACG +TCGAGA +TCGAGT +TCGAGC +TCGAGG +TCGTAA +TCGTAT +TCGTAC +TCGTAG +TCGTTA +TCGTTT +TCGTTC +TCGTTG +TCGTCA +TCGTCT +TCGTCC +TCGTCG +TCGTGA +TCGTGT +TCGTGC +TCGTGG +TCGCAA +TCGCAT +TCGCAC +TCGCAG +TCGCTA +TCGCTT +TCGCTC +TCGCTG +TCGCCA +TCGCCT +TCGCCC +TCGCCG +TCGCGA +TCGCGT +TCGCGC +TCGCGG +TCGGAA +TCGGAT +TCGGAC +TCGGAG +TCGGTA +TCGGTT +TCGGTC +TCGGTG +TCGGCA +TCGGCT +TCGGCC +TCGGCG +TCGGGA +TCGGGT +TCGGGC +TCGGGG +TGAAAA +TGAAAT +TGAAAC +TGAAAG +TGAATA +TGAATT +TGAATC +TGAATG +TGAACA +TGAACT +TGAACC +TGAACG +TGAAGA +TGAAGT +TGAAGC +TGAAGG +TGATAA +TGATAT +TGATAC +TGATAG +TGATTA +TGATTT +TGATTC +TGATTG +TGATCA +TGATCT +TGATCC +TGATCG +TGATGA +TGATGT +TGATGC +TGATGG +TGACAA +TGACAT +TGACAC +TGACAG +TGACTA +TGACTT +TGACTC +TGACTG +TGACCA +TGACCT +TGACCC +TGACCG +TGACGA +TGACGT +TGACGC +TGACGG +TGAGAA +TGAGAT +TGAGAC +TGAGAG +TGAGTA +TGAGTT +TGAGTC +TGAGTG +TGAGCA +TGAGCT +TGAGCC +TGAGCG +TGAGGA +TGAGGT +TGAGGC +TGAGGG +TGTAAA +TGTAAT +TGTAAC +TGTAAG +TGTATA +TGTATT +TGTATC +TGTATG +TGTACA +TGTACT +TGTACC +TGTACG +TGTAGA +TGTAGT +TGTAGC +TGTAGG +TGTTAA +TGTTAT +TGTTAC +TGTTAG +TGTTTA +TGTTTT +TGTTTC +TGTTTG +TGTTCA +TGTTCT +TGTTCC +TGTTCG +TGTTGA +TGTTGT +TGTTGC +TGTTGG +TGTCAA +TGTCAT +TGTCAC +TGTCAG +TGTCTA +TGTCTT +TGTCTC +TGTCTG +TGTCCA +TGTCCT +TGTCCC +TGTCCG +TGTCGA +TGTCGT +TGTCGC +TGTCGG +TGTGAA +TGTGAT +TGTGAC +TGTGAG +TGTGTA +TGTGTT +TGTGTC +TGTGTG +TGTGCA +TGTGCT +TGTGCC +TGTGCG +TGTGGA +TGTGGT +TGTGGC +TGTGGG +TGCAAA +TGCAAT +TGCAAC +TGCAAG +TGCATA +TGCATT +TGCATC +TGCATG +TGCACA +TGCACT +TGCACC +TGCACG +TGCAGA +TGCAGT +TGCAGC +TGCAGG +TGCTAA +TGCTAT +TGCTAC +TGCTAG +TGCTTA +TGCTTT +TGCTTC +TGCTTG +TGCTCA +TGCTCT +TGCTCC +TGCTCG +TGCTGA +TGCTGT +TGCTGC +TGCTGG +TGCCAA +TGCCAT +TGCCAC +TGCCAG +TGCCTA +TGCCTT +TGCCTC +TGCCTG +TGCCCA +TGCCCT +TGCCCC +TGCCCG +TGCCGA +TGCCGT +TGCCGC +TGCCGG +TGCGAA +TGCGAT +TGCGAC +TGCGAG +TGCGTA +TGCGTT +TGCGTC +TGCGTG +TGCGCA +TGCGCT +TGCGCC +TGCGCG +TGCGGA +TGCGGT +TGCGGC +TGCGGG +TGGAAA +TGGAAT +TGGAAC +TGGAAG +TGGATA +TGGATT +TGGATC +TGGATG +TGGACA +TGGACT +TGGACC +TGGACG +TGGAGA +TGGAGT +TGGAGC +TGGAGG +TGGTAA +TGGTAT +TGGTAC +TGGTAG +TGGTTA +TGGTTT +TGGTTC +TGGTTG +TGGTCA +TGGTCT +TGGTCC +TGGTCG +TGGTGA +TGGTGT +TGGTGC +TGGTGG +TGGCAA +TGGCAT +TGGCAC +TGGCAG +TGGCTA +TGGCTT +TGGCTC +TGGCTG +TGGCCA +TGGCCT +TGGCCC +TGGCCG +TGGCGA +TGGCGT +TGGCGC +TGGCGG +TGGGAA +TGGGAT +TGGGAC +TGGGAG +TGGGTA +TGGGTT +TGGGTC +TGGGTG +TGGGCA +TGGGCT +TGGGCC +TGGGCG +TGGGGA +TGGGGT +TGGGGC +TGGGGG +CAAAAA +CAAAAT +CAAAAC +CAAAAG +CAAATA +CAAATT +CAAATC +CAAATG +CAAACA +CAAACT +CAAACC +CAAACG +CAAAGA +CAAAGT +CAAAGC +CAAAGG +CAATAA +CAATAT +CAATAC +CAATAG +CAATTA +CAATTT +CAATTC +CAATTG +CAATCA +CAATCT +CAATCC +CAATCG +CAATGA +CAATGT +CAATGC +CAATGG +CAACAA +CAACAT +CAACAC +CAACAG +CAACTA +CAACTT +CAACTC +CAACTG +CAACCA +CAACCT +CAACCC +CAACCG +CAACGA +CAACGT +CAACGC +CAACGG +CAAGAA +CAAGAT +CAAGAC +CAAGAG +CAAGTA +CAAGTT +CAAGTC +CAAGTG +CAAGCA +CAAGCT +CAAGCC +CAAGCG +CAAGGA +CAAGGT +CAAGGC +CAAGGG +CATAAA +CATAAT +CATAAC +CATAAG +CATATA +CATATT +CATATC +CATATG +CATACA +CATACT +CATACC +CATACG +CATAGA +CATAGT +CATAGC +CATAGG +CATTAA +CATTAT +CATTAC +CATTAG +CATTTA +CATTTT +CATTTC +CATTTG +CATTCA +CATTCT +CATTCC +CATTCG +CATTGA +CATTGT +CATTGC +CATTGG +CATCAA +CATCAT +CATCAC +CATCAG +CATCTA +CATCTT +CATCTC +CATCTG +CATCCA +CATCCT +CATCCC +CATCCG +CATCGA +CATCGT +CATCGC +CATCGG +CATGAA +CATGAT +CATGAC +CATGAG +CATGTA +CATGTT +CATGTC +CATGTG +CATGCA +CATGCT +CATGCC +CATGCG +CATGGA +CATGGT +CATGGC +CATGGG +CACAAA +CACAAT +CACAAC +CACAAG +CACATA +CACATT +CACATC +CACATG +CACACA +CACACT +CACACC +CACACG +CACAGA +CACAGT +CACAGC +CACAGG +CACTAA +CACTAT +CACTAC +CACTAG +CACTTA +CACTTT +CACTTC +CACTTG +CACTCA +CACTCT +CACTCC +CACTCG +CACTGA +CACTGT +CACTGC +CACTGG +CACCAA +CACCAT +CACCAC +CACCAG +CACCTA +CACCTT +CACCTC +CACCTG +CACCCA +CACCCT +CACCCC +CACCCG +CACCGA +CACCGT +CACCGC +CACCGG +CACGAA +CACGAT +CACGAC +CACGAG +CACGTA +CACGTT +CACGTC +CACGTG +CACGCA +CACGCT +CACGCC +CACGCG +CACGGA +CACGGT +CACGGC +CACGGG +CAGAAA +CAGAAT +CAGAAC +CAGAAG +CAGATA +CAGATT +CAGATC +CAGATG +CAGACA +CAGACT +CAGACC +CAGACG +CAGAGA +CAGAGT +CAGAGC +CAGAGG +CAGTAA +CAGTAT +CAGTAC +CAGTAG +CAGTTA +CAGTTT +CAGTTC +CAGTTG +CAGTCA +CAGTCT +CAGTCC +CAGTCG +CAGTGA +CAGTGT +CAGTGC +CAGTGG +CAGCAA +CAGCAT +CAGCAC +CAGCAG +CAGCTA +CAGCTT +CAGCTC +CAGCTG +CAGCCA +CAGCCT +CAGCCC +CAGCCG +CAGCGA +CAGCGT +CAGCGC +CAGCGG +CAGGAA +CAGGAT +CAGGAC +CAGGAG +CAGGTA +CAGGTT +CAGGTC +CAGGTG +CAGGCA +CAGGCT +CAGGCC +CAGGCG +CAGGGA +CAGGGT +CAGGGC +CAGGGG +CTAAAA +CTAAAT +CTAAAC +CTAAAG +CTAATA +CTAATT +CTAATC +CTAATG +CTAACA +CTAACT +CTAACC +CTAACG +CTAAGA +CTAAGT +CTAAGC +CTAAGG +CTATAA +CTATAT +CTATAC +CTATAG +CTATTA +CTATTT +CTATTC +CTATTG +CTATCA +CTATCT +CTATCC +CTATCG +CTATGA +CTATGT +CTATGC +CTATGG +CTACAA +CTACAT +CTACAC +CTACAG +CTACTA +CTACTT +CTACTC +CTACTG +CTACCA +CTACCT +CTACCC +CTACCG +CTACGA +CTACGT +CTACGC +CTACGG +CTAGAA +CTAGAT +CTAGAC +CTAGAG +CTAGTA +CTAGTT +CTAGTC +CTAGTG +CTAGCA +CTAGCT +CTAGCC +CTAGCG +CTAGGA +CTAGGT +CTAGGC +CTAGGG +CTTAAA +CTTAAT +CTTAAC +CTTAAG +CTTATA +CTTATT +CTTATC +CTTATG +CTTACA +CTTACT +CTTACC +CTTACG +CTTAGA +CTTAGT +CTTAGC +CTTAGG +CTTTAA +CTTTAT +CTTTAC +CTTTAG +CTTTTA +CTTTTT +CTTTTC +CTTTTG +CTTTCA +CTTTCT +CTTTCC +CTTTCG +CTTTGA +CTTTGT +CTTTGC +CTTTGG +CTTCAA +CTTCAT +CTTCAC +CTTCAG +CTTCTA +CTTCTT +CTTCTC +CTTCTG +CTTCCA +CTTCCT +CTTCCC +CTTCCG +CTTCGA +CTTCGT +CTTCGC +CTTCGG +CTTGAA +CTTGAT +CTTGAC +CTTGAG +CTTGTA +CTTGTT +CTTGTC +CTTGTG +CTTGCA +CTTGCT +CTTGCC +CTTGCG +CTTGGA +CTTGGT +CTTGGC +CTTGGG +CTCAAA +CTCAAT +CTCAAC +CTCAAG +CTCATA +CTCATT +CTCATC +CTCATG +CTCACA +CTCACT +CTCACC +CTCACG +CTCAGA +CTCAGT +CTCAGC +CTCAGG +CTCTAA +CTCTAT +CTCTAC +CTCTAG +CTCTTA +CTCTTT +CTCTTC +CTCTTG +CTCTCA +CTCTCT +CTCTCC +CTCTCG +CTCTGA +CTCTGT +CTCTGC +CTCTGG +CTCCAA +CTCCAT +CTCCAC +CTCCAG +CTCCTA +CTCCTT +CTCCTC +CTCCTG +CTCCCA +CTCCCT +CTCCCC +CTCCCG +CTCCGA +CTCCGT +CTCCGC +CTCCGG +CTCGAA +CTCGAT +CTCGAC +CTCGAG +CTCGTA +CTCGTT +CTCGTC +CTCGTG +CTCGCA +CTCGCT +CTCGCC +CTCGCG +CTCGGA +CTCGGT +CTCGGC +CTCGGG +CTGAAA +CTGAAT +CTGAAC +CTGAAG +CTGATA +CTGATT +CTGATC +CTGATG +CTGACA +CTGACT +CTGACC +CTGACG +CTGAGA +CTGAGT +CTGAGC +CTGAGG +CTGTAA +CTGTAT +CTGTAC +CTGTAG +CTGTTA +CTGTTT +CTGTTC +CTGTTG +CTGTCA +CTGTCT +CTGTCC +CTGTCG +CTGTGA +CTGTGT +CTGTGC +CTGTGG +CTGCAA +CTGCAT +CTGCAC +CTGCAG +CTGCTA +CTGCTT +CTGCTC +CTGCTG +CTGCCA +CTGCCT +CTGCCC +CTGCCG +CTGCGA +CTGCGT +CTGCGC +CTGCGG +CTGGAA +CTGGAT +CTGGAC +CTGGAG +CTGGTA +CTGGTT +CTGGTC +CTGGTG +CTGGCA +CTGGCT +CTGGCC +CTGGCG +CTGGGA +CTGGGT +CTGGGC +CTGGGG +CCAAAA +CCAAAT +CCAAAC +CCAAAG +CCAATA +CCAATT +CCAATC +CCAATG +CCAACA +CCAACT +CCAACC +CCAACG +CCAAGA +CCAAGT +CCAAGC +CCAAGG +CCATAA +CCATAT +CCATAC +CCATAG +CCATTA +CCATTT +CCATTC +CCATTG +CCATCA +CCATCT +CCATCC +CCATCG +CCATGA +CCATGT +CCATGC +CCATGG +CCACAA +CCACAT +CCACAC +CCACAG +CCACTA +CCACTT +CCACTC +CCACTG +CCACCA +CCACCT +CCACCC +CCACCG +CCACGA +CCACGT +CCACGC +CCACGG +CCAGAA +CCAGAT +CCAGAC +CCAGAG +CCAGTA +CCAGTT +CCAGTC +CCAGTG +CCAGCA +CCAGCT +CCAGCC +CCAGCG +CCAGGA +CCAGGT +CCAGGC +CCAGGG +CCTAAA +CCTAAT +CCTAAC +CCTAAG +CCTATA +CCTATT +CCTATC +CCTATG +CCTACA +CCTACT +CCTACC +CCTACG +CCTAGA +CCTAGT +CCTAGC +CCTAGG +CCTTAA +CCTTAT +CCTTAC +CCTTAG +CCTTTA +CCTTTT +CCTTTC +CCTTTG +CCTTCA +CCTTCT +CCTTCC +CCTTCG +CCTTGA +CCTTGT +CCTTGC +CCTTGG +CCTCAA +CCTCAT +CCTCAC +CCTCAG +CCTCTA +CCTCTT +CCTCTC +CCTCTG +CCTCCA +CCTCCT +CCTCCC +CCTCCG +CCTCGA +CCTCGT +CCTCGC +CCTCGG +CCTGAA +CCTGAT +CCTGAC +CCTGAG +CCTGTA +CCTGTT +CCTGTC +CCTGTG +CCTGCA +CCTGCT +CCTGCC +CCTGCG +CCTGGA +CCTGGT +CCTGGC +CCTGGG +CCCAAA +CCCAAT +CCCAAC +CCCAAG +CCCATA +CCCATT +CCCATC +CCCATG +CCCACA +CCCACT +CCCACC +CCCACG +CCCAGA +CCCAGT +CCCAGC +CCCAGG +CCCTAA +CCCTAT +CCCTAC +CCCTAG +CCCTTA +CCCTTT +CCCTTC +CCCTTG +CCCTCA +CCCTCT +CCCTCC +CCCTCG +CCCTGA +CCCTGT +CCCTGC +CCCTGG +CCCCAA +CCCCAT +CCCCAC +CCCCAG +CCCCTA +CCCCTT +CCCCTC +CCCCTG +CCCCCA +CCCCCT +CCCCCC +CCCCCG +CCCCGA +CCCCGT +CCCCGC +CCCCGG +CCCGAA +CCCGAT +CCCGAC +CCCGAG +CCCGTA +CCCGTT +CCCGTC +CCCGTG +CCCGCA +CCCGCT +CCCGCC +CCCGCG +CCCGGA +CCCGGT +CCCGGC +CCCGGG +CCGAAA +CCGAAT +CCGAAC +CCGAAG +CCGATA +CCGATT +CCGATC +CCGATG +CCGACA +CCGACT +CCGACC +CCGACG +CCGAGA +CCGAGT +CCGAGC +CCGAGG +CCGTAA +CCGTAT +CCGTAC +CCGTAG +CCGTTA +CCGTTT +CCGTTC +CCGTTG +CCGTCA +CCGTCT +CCGTCC +CCGTCG +CCGTGA +CCGTGT +CCGTGC +CCGTGG +CCGCAA +CCGCAT +CCGCAC +CCGCAG +CCGCTA +CCGCTT +CCGCTC +CCGCTG +CCGCCA +CCGCCT +CCGCCC +CCGCCG +CCGCGA +CCGCGT +CCGCGC +CCGCGG +CCGGAA +CCGGAT +CCGGAC +CCGGAG +CCGGTA +CCGGTT +CCGGTC +CCGGTG +CCGGCA +CCGGCT +CCGGCC +CCGGCG +CCGGGA +CCGGGT +CCGGGC +CCGGGG +CGAAAA +CGAAAT +CGAAAC +CGAAAG +CGAATA +CGAATT +CGAATC +CGAATG +CGAACA +CGAACT +CGAACC +CGAACG +CGAAGA +CGAAGT +CGAAGC +CGAAGG +CGATAA +CGATAT +CGATAC +CGATAG +CGATTA +CGATTT +CGATTC +CGATTG +CGATCA +CGATCT +CGATCC +CGATCG +CGATGA +CGATGT +CGATGC +CGATGG +CGACAA +CGACAT +CGACAC +CGACAG +CGACTA +CGACTT +CGACTC +CGACTG +CGACCA +CGACCT +CGACCC +CGACCG +CGACGA +CGACGT +CGACGC +CGACGG +CGAGAA +CGAGAT +CGAGAC +CGAGAG +CGAGTA +CGAGTT +CGAGTC +CGAGTG +CGAGCA +CGAGCT +CGAGCC +CGAGCG +CGAGGA +CGAGGT +CGAGGC +CGAGGG +CGTAAA +CGTAAT +CGTAAC +CGTAAG +CGTATA +CGTATT +CGTATC +CGTATG +CGTACA +CGTACT +CGTACC +CGTACG +CGTAGA +CGTAGT +CGTAGC +CGTAGG +CGTTAA +CGTTAT +CGTTAC +CGTTAG +CGTTTA +CGTTTT +CGTTTC +CGTTTG +CGTTCA +CGTTCT +CGTTCC +CGTTCG +CGTTGA +CGTTGT +CGTTGC +CGTTGG +CGTCAA +CGTCAT +CGTCAC +CGTCAG +CGTCTA +CGTCTT +CGTCTC +CGTCTG +CGTCCA +CGTCCT +CGTCCC +CGTCCG +CGTCGA +CGTCGT +CGTCGC +CGTCGG +CGTGAA +CGTGAT +CGTGAC +CGTGAG +CGTGTA +CGTGTT +CGTGTC +CGTGTG +CGTGCA +CGTGCT +CGTGCC +CGTGCG +CGTGGA +CGTGGT +CGTGGC +CGTGGG +CGCAAA +CGCAAT +CGCAAC +CGCAAG +CGCATA +CGCATT +CGCATC +CGCATG +CGCACA +CGCACT +CGCACC +CGCACG +CGCAGA +CGCAGT +CGCAGC +CGCAGG +CGCTAA +CGCTAT +CGCTAC +CGCTAG +CGCTTA +CGCTTT +CGCTTC +CGCTTG +CGCTCA +CGCTCT +CGCTCC +CGCTCG +CGCTGA +CGCTGT +CGCTGC +CGCTGG +CGCCAA +CGCCAT +CGCCAC +CGCCAG +CGCCTA +CGCCTT +CGCCTC +CGCCTG +CGCCCA +CGCCCT +CGCCCC +CGCCCG +CGCCGA +CGCCGT +CGCCGC +CGCCGG +CGCGAA +CGCGAT +CGCGAC +CGCGAG +CGCGTA +CGCGTT +CGCGTC +CGCGTG +CGCGCA +CGCGCT +CGCGCC +CGCGCG +CGCGGA +CGCGGT +CGCGGC +CGCGGG +CGGAAA +CGGAAT +CGGAAC +CGGAAG +CGGATA +CGGATT +CGGATC +CGGATG +CGGACA +CGGACT +CGGACC +CGGACG +CGGAGA +CGGAGT +CGGAGC +CGGAGG +CGGTAA +CGGTAT +CGGTAC +CGGTAG +CGGTTA +CGGTTT +CGGTTC +CGGTTG +CGGTCA +CGGTCT +CGGTCC +CGGTCG +CGGTGA +CGGTGT +CGGTGC +CGGTGG +CGGCAA +CGGCAT +CGGCAC +CGGCAG +CGGCTA +CGGCTT +CGGCTC +CGGCTG +CGGCCA +CGGCCT +CGGCCC +CGGCCG +CGGCGA +CGGCGT +CGGCGC +CGGCGG +CGGGAA +CGGGAT +CGGGAC +CGGGAG +CGGGTA +CGGGTT +CGGGTC +CGGGTG +CGGGCA +CGGGCT +CGGGCC +CGGGCG +CGGGGA +CGGGGT +CGGGGC +CGGGGG +GAAAAA +GAAAAT +GAAAAC +GAAAAG +GAAATA +GAAATT +GAAATC +GAAATG +GAAACA +GAAACT +GAAACC +GAAACG +GAAAGA +GAAAGT +GAAAGC +GAAAGG +GAATAA +GAATAT +GAATAC +GAATAG +GAATTA +GAATTT +GAATTC +GAATTG +GAATCA +GAATCT +GAATCC +GAATCG +GAATGA +GAATGT +GAATGC +GAATGG +GAACAA +GAACAT +GAACAC +GAACAG +GAACTA +GAACTT +GAACTC +GAACTG +GAACCA +GAACCT +GAACCC +GAACCG +GAACGA +GAACGT +GAACGC +GAACGG +GAAGAA +GAAGAT +GAAGAC +GAAGAG +GAAGTA +GAAGTT +GAAGTC +GAAGTG +GAAGCA +GAAGCT +GAAGCC +GAAGCG +GAAGGA +GAAGGT +GAAGGC +GAAGGG +GATAAA +GATAAT +GATAAC +GATAAG +GATATA +GATATT +GATATC +GATATG +GATACA +GATACT +GATACC +GATACG +GATAGA +GATAGT +GATAGC +GATAGG +GATTAA +GATTAT +GATTAC +GATTAG +GATTTA +GATTTT +GATTTC +GATTTG +GATTCA +GATTCT +GATTCC +GATTCG +GATTGA +GATTGT +GATTGC +GATTGG +GATCAA +GATCAT +GATCAC +GATCAG +GATCTA +GATCTT +GATCTC +GATCTG +GATCCA +GATCCT +GATCCC +GATCCG +GATCGA +GATCGT +GATCGC +GATCGG +GATGAA +GATGAT +GATGAC +GATGAG +GATGTA +GATGTT +GATGTC +GATGTG +GATGCA +GATGCT +GATGCC +GATGCG +GATGGA +GATGGT +GATGGC +GATGGG +GACAAA +GACAAT +GACAAC +GACAAG +GACATA +GACATT +GACATC +GACATG +GACACA +GACACT +GACACC +GACACG +GACAGA +GACAGT +GACAGC +GACAGG +GACTAA +GACTAT +GACTAC +GACTAG +GACTTA +GACTTT +GACTTC +GACTTG +GACTCA +GACTCT +GACTCC +GACTCG +GACTGA +GACTGT +GACTGC +GACTGG +GACCAA +GACCAT +GACCAC +GACCAG +GACCTA +GACCTT +GACCTC +GACCTG +GACCCA +GACCCT +GACCCC +GACCCG +GACCGA +GACCGT +GACCGC +GACCGG +GACGAA +GACGAT +GACGAC +GACGAG +GACGTA +GACGTT +GACGTC +GACGTG +GACGCA +GACGCT +GACGCC +GACGCG +GACGGA +GACGGT +GACGGC +GACGGG +GAGAAA +GAGAAT +GAGAAC +GAGAAG +GAGATA +GAGATT +GAGATC +GAGATG +GAGACA +GAGACT +GAGACC +GAGACG +GAGAGA +GAGAGT +GAGAGC +GAGAGG +GAGTAA +GAGTAT +GAGTAC +GAGTAG +GAGTTA +GAGTTT +GAGTTC +GAGTTG +GAGTCA +GAGTCT +GAGTCC +GAGTCG +GAGTGA +GAGTGT +GAGTGC +GAGTGG +GAGCAA +GAGCAT +GAGCAC +GAGCAG +GAGCTA +GAGCTT +GAGCTC +GAGCTG +GAGCCA +GAGCCT +GAGCCC +GAGCCG +GAGCGA +GAGCGT +GAGCGC +GAGCGG +GAGGAA +GAGGAT +GAGGAC +GAGGAG +GAGGTA +GAGGTT +GAGGTC +GAGGTG +GAGGCA +GAGGCT +GAGGCC +GAGGCG +GAGGGA +GAGGGT +GAGGGC +GAGGGG +GTAAAA +GTAAAT +GTAAAC +GTAAAG +GTAATA +GTAATT +GTAATC +GTAATG +GTAACA +GTAACT +GTAACC +GTAACG +GTAAGA +GTAAGT +GTAAGC +GTAAGG +GTATAA +GTATAT +GTATAC +GTATAG +GTATTA +GTATTT +GTATTC +GTATTG +GTATCA +GTATCT +GTATCC +GTATCG +GTATGA +GTATGT +GTATGC +GTATGG +GTACAA +GTACAT +GTACAC +GTACAG +GTACTA +GTACTT +GTACTC +GTACTG +GTACCA +GTACCT +GTACCC +GTACCG +GTACGA +GTACGT +GTACGC +GTACGG +GTAGAA +GTAGAT +GTAGAC +GTAGAG +GTAGTA +GTAGTT +GTAGTC +GTAGTG +GTAGCA +GTAGCT +GTAGCC +GTAGCG +GTAGGA +GTAGGT +GTAGGC +GTAGGG +GTTAAA +GTTAAT +GTTAAC +GTTAAG +GTTATA +GTTATT +GTTATC +GTTATG +GTTACA +GTTACT +GTTACC +GTTACG +GTTAGA +GTTAGT +GTTAGC +GTTAGG +GTTTAA +GTTTAT +GTTTAC +GTTTAG +GTTTTA +GTTTTT +GTTTTC +GTTTTG +GTTTCA +GTTTCT +GTTTCC +GTTTCG +GTTTGA +GTTTGT +GTTTGC +GTTTGG +GTTCAA +GTTCAT +GTTCAC +GTTCAG +GTTCTA +GTTCTT +GTTCTC +GTTCTG +GTTCCA +GTTCCT +GTTCCC +GTTCCG +GTTCGA +GTTCGT +GTTCGC +GTTCGG +GTTGAA +GTTGAT +GTTGAC +GTTGAG +GTTGTA +GTTGTT +GTTGTC +GTTGTG +GTTGCA +GTTGCT +GTTGCC +GTTGCG +GTTGGA +GTTGGT +GTTGGC +GTTGGG +GTCAAA +GTCAAT +GTCAAC +GTCAAG +GTCATA +GTCATT +GTCATC +GTCATG +GTCACA +GTCACT +GTCACC +GTCACG +GTCAGA +GTCAGT +GTCAGC +GTCAGG +GTCTAA +GTCTAT +GTCTAC +GTCTAG +GTCTTA +GTCTTT +GTCTTC +GTCTTG +GTCTCA +GTCTCT +GTCTCC +GTCTCG +GTCTGA +GTCTGT +GTCTGC +GTCTGG +GTCCAA +GTCCAT +GTCCAC +GTCCAG +GTCCTA +GTCCTT +GTCCTC +GTCCTG +GTCCCA +GTCCCT +GTCCCC +GTCCCG +GTCCGA +GTCCGT +GTCCGC +GTCCGG +GTCGAA +GTCGAT +GTCGAC +GTCGAG +GTCGTA +GTCGTT +GTCGTC +GTCGTG +GTCGCA +GTCGCT +GTCGCC +GTCGCG +GTCGGA +GTCGGT +GTCGGC +GTCGGG +GTGAAA +GTGAAT +GTGAAC +GTGAAG +GTGATA +GTGATT +GTGATC +GTGATG +GTGACA +GTGACT +GTGACC +GTGACG +GTGAGA +GTGAGT +GTGAGC +GTGAGG +GTGTAA +GTGTAT +GTGTAC +GTGTAG +GTGTTA +GTGTTT +GTGTTC +GTGTTG +GTGTCA +GTGTCT +GTGTCC +GTGTCG +GTGTGA +GTGTGT +GTGTGC +GTGTGG +GTGCAA +GTGCAT +GTGCAC +GTGCAG +GTGCTA +GTGCTT +GTGCTC +GTGCTG +GTGCCA +GTGCCT +GTGCCC +GTGCCG +GTGCGA +GTGCGT +GTGCGC +GTGCGG +GTGGAA +GTGGAT +GTGGAC +GTGGAG +GTGGTA +GTGGTT +GTGGTC +GTGGTG +GTGGCA +GTGGCT +GTGGCC +GTGGCG +GTGGGA +GTGGGT +GTGGGC +GTGGGG +GCAAAA +GCAAAT +GCAAAC +GCAAAG +GCAATA +GCAATT +GCAATC +GCAATG +GCAACA +GCAACT +GCAACC +GCAACG +GCAAGA +GCAAGT +GCAAGC +GCAAGG +GCATAA +GCATAT +GCATAC +GCATAG +GCATTA +GCATTT +GCATTC +GCATTG +GCATCA +GCATCT +GCATCC +GCATCG +GCATGA +GCATGT +GCATGC +GCATGG +GCACAA +GCACAT +GCACAC +GCACAG +GCACTA +GCACTT +GCACTC +GCACTG +GCACCA +GCACCT +GCACCC +GCACCG +GCACGA +GCACGT +GCACGC +GCACGG +GCAGAA +GCAGAT +GCAGAC +GCAGAG +GCAGTA +GCAGTT +GCAGTC +GCAGTG +GCAGCA +GCAGCT +GCAGCC +GCAGCG +GCAGGA +GCAGGT +GCAGGC +GCAGGG +GCTAAA +GCTAAT +GCTAAC +GCTAAG +GCTATA +GCTATT +GCTATC +GCTATG +GCTACA +GCTACT +GCTACC +GCTACG +GCTAGA +GCTAGT +GCTAGC +GCTAGG +GCTTAA +GCTTAT +GCTTAC +GCTTAG +GCTTTA +GCTTTT +GCTTTC +GCTTTG +GCTTCA +GCTTCT +GCTTCC +GCTTCG +GCTTGA +GCTTGT +GCTTGC +GCTTGG +GCTCAA +GCTCAT +GCTCAC +GCTCAG +GCTCTA +GCTCTT +GCTCTC +GCTCTG +GCTCCA +GCTCCT +GCTCCC +GCTCCG +GCTCGA +GCTCGT +GCTCGC +GCTCGG +GCTGAA +GCTGAT +GCTGAC +GCTGAG +GCTGTA +GCTGTT +GCTGTC +GCTGTG +GCTGCA +GCTGCT +GCTGCC +GCTGCG +GCTGGA +GCTGGT +GCTGGC +GCTGGG +GCCAAA +GCCAAT +GCCAAC +GCCAAG +GCCATA +GCCATT +GCCATC +GCCATG +GCCACA +GCCACT +GCCACC +GCCACG +GCCAGA +GCCAGT +GCCAGC +GCCAGG +GCCTAA +GCCTAT +GCCTAC +GCCTAG +GCCTTA +GCCTTT +GCCTTC +GCCTTG +GCCTCA +GCCTCT +GCCTCC +GCCTCG +GCCTGA +GCCTGT +GCCTGC +GCCTGG +GCCCAA +GCCCAT +GCCCAC +GCCCAG +GCCCTA +GCCCTT +GCCCTC +GCCCTG +GCCCCA +GCCCCT +GCCCCC +GCCCCG +GCCCGA +GCCCGT +GCCCGC +GCCCGG +GCCGAA +GCCGAT +GCCGAC +GCCGAG +GCCGTA +GCCGTT +GCCGTC +GCCGTG +GCCGCA +GCCGCT +GCCGCC +GCCGCG +GCCGGA +GCCGGT +GCCGGC +GCCGGG +GCGAAA +GCGAAT +GCGAAC +GCGAAG +GCGATA +GCGATT +GCGATC +GCGATG +GCGACA +GCGACT +GCGACC +GCGACG +GCGAGA +GCGAGT +GCGAGC +GCGAGG +GCGTAA +GCGTAT +GCGTAC +GCGTAG +GCGTTA +GCGTTT +GCGTTC +GCGTTG +GCGTCA +GCGTCT +GCGTCC +GCGTCG +GCGTGA +GCGTGT +GCGTGC +GCGTGG +GCGCAA +GCGCAT +GCGCAC +GCGCAG +GCGCTA +GCGCTT +GCGCTC +GCGCTG +GCGCCA +GCGCCT +GCGCCC +GCGCCG +GCGCGA +GCGCGT +GCGCGC +GCGCGG +GCGGAA +GCGGAT +GCGGAC +GCGGAG +GCGGTA +GCGGTT +GCGGTC +GCGGTG +GCGGCA +GCGGCT +GCGGCC +GCGGCG +GCGGGA +GCGGGT +GCGGGC +GCGGGG +GGAAAA +GGAAAT +GGAAAC +GGAAAG +GGAATA +GGAATT +GGAATC +GGAATG +GGAACA +GGAACT +GGAACC +GGAACG +GGAAGA +GGAAGT +GGAAGC +GGAAGG +GGATAA +GGATAT +GGATAC +GGATAG +GGATTA +GGATTT +GGATTC +GGATTG +GGATCA +GGATCT +GGATCC +GGATCG +GGATGA +GGATGT +GGATGC +GGATGG +GGACAA +GGACAT +GGACAC +GGACAG +GGACTA +GGACTT +GGACTC +GGACTG +GGACCA +GGACCT +GGACCC +GGACCG +GGACGA +GGACGT +GGACGC +GGACGG +GGAGAA +GGAGAT +GGAGAC +GGAGAG +GGAGTA +GGAGTT +GGAGTC +GGAGTG +GGAGCA +GGAGCT +GGAGCC +GGAGCG +GGAGGA +GGAGGT +GGAGGC +GGAGGG +GGTAAA +GGTAAT +GGTAAC +GGTAAG +GGTATA +GGTATT +GGTATC +GGTATG +GGTACA +GGTACT +GGTACC +GGTACG +GGTAGA +GGTAGT +GGTAGC +GGTAGG +GGTTAA +GGTTAT +GGTTAC +GGTTAG +GGTTTA +GGTTTT +GGTTTC +GGTTTG +GGTTCA +GGTTCT +GGTTCC +GGTTCG +GGTTGA +GGTTGT +GGTTGC +GGTTGG +GGTCAA +GGTCAT +GGTCAC +GGTCAG +GGTCTA +GGTCTT +GGTCTC +GGTCTG +GGTCCA +GGTCCT +GGTCCC +GGTCCG +GGTCGA +GGTCGT +GGTCGC +GGTCGG +GGTGAA +GGTGAT +GGTGAC +GGTGAG +GGTGTA +GGTGTT +GGTGTC +GGTGTG +GGTGCA +GGTGCT +GGTGCC +GGTGCG +GGTGGA +GGTGGT +GGTGGC +GGTGGG +GGCAAA +GGCAAT +GGCAAC +GGCAAG +GGCATA +GGCATT +GGCATC +GGCATG +GGCACA +GGCACT +GGCACC +GGCACG +GGCAGA +GGCAGT +GGCAGC +GGCAGG +GGCTAA +GGCTAT +GGCTAC +GGCTAG +GGCTTA +GGCTTT +GGCTTC +GGCTTG +GGCTCA +GGCTCT +GGCTCC +GGCTCG +GGCTGA +GGCTGT +GGCTGC +GGCTGG +GGCCAA +GGCCAT +GGCCAC +GGCCAG +GGCCTA +GGCCTT +GGCCTC +GGCCTG +GGCCCA +GGCCCT +GGCCCC +GGCCCG +GGCCGA +GGCCGT +GGCCGC +GGCCGG +GGCGAA +GGCGAT +GGCGAC +GGCGAG +GGCGTA +GGCGTT +GGCGTC +GGCGTG +GGCGCA +GGCGCT +GGCGCC +GGCGCG +GGCGGA +GGCGGT +GGCGGC +GGCGGG +GGGAAA +GGGAAT +GGGAAC +GGGAAG +GGGATA +GGGATT +GGGATC +GGGATG +GGGACA +GGGACT +GGGACC +GGGACG +GGGAGA +GGGAGT +GGGAGC +GGGAGG +GGGTAA +GGGTAT +GGGTAC +GGGTAG +GGGTTA +GGGTTT +GGGTTC +GGGTTG +GGGTCA +GGGTCT +GGGTCC +GGGTCG +GGGTGA +GGGTGT +GGGTGC +GGGTGG +GGGCAA +GGGCAT +GGGCAC +GGGCAG +GGGCTA +GGGCTT +GGGCTC +GGGCTG +GGGCCA +GGGCCT +GGGCCC +GGGCCG +GGGCGA +GGGCGT +GGGCGC +GGGCGG +GGGGAA +GGGGAT +GGGGAC +GGGGAG +GGGGTA +GGGGTT +GGGGTC +GGGGTG +GGGGCA +GGGGCT +GGGGCC +GGGGCG +GGGGGA +GGGGGT +GGGGGC +GGGGGG diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/README.md b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..500fd976d92b217455f181bc98c77e612d9a18d9 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/README.md @@ -0,0 +1,202 @@ +--- +base_model: fabihamakhdoomi/TinyDNABERT +library_name: peft +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.12.0 \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/adapter_config.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..28c871c93775eedfaf266f00da8c02f96bc02f29 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/adapter_config.json @@ -0,0 +1,42 @@ +{ + "alpha_pattern": {}, + "auto_mapping": null, + "base_model_name_or_path": "fabihamakhdoomi/TinyDNABERT", + "bias": "none", + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 16, + "lora_dropout": 0.2, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": [ + "classifier", + "score" + ], + "peft_type": "LORA", + "r": 4, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "bert.encoder.layer.1.attention.self.query", + "bert.encoder.layer.0.output.dense", + "bert.encoder.layer.1.output.dense", + "bert.encoder.layer.0.intermediate.dense", + "bert.encoder.layer.1.intermediate.dense", + "bert.encoder.layer.0.attention.output.dense", + "bert.encoder.layer.1.attention.output.dense", + "bert.encoder.layer.0.attention.self.query", + "bert.encoder.layer.0.attention.self.value", + "bert.encoder.layer.1.attention.self.key", + "bert.encoder.layer.1.attention.self.value", + "bert.encoder.layer.0.attention.self.key" + ], + "task_type": "SEQ_CLS", + "use_dora": false, + "use_rslora": false +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/adapter_model.safetensors b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/adapter_model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..674d607fe07cacded20d22453e9d7fa7c58598f5 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/adapter_model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5be68580cc3e64f622d60aaf0ab837aaec2252f73e47941733a14bd395c2468c +size 78208 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/optimizer.pt b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..f12eac588bc7a8b84de4e16b0b66745985bf6c47 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:91912c38a3a7a0df66cba257ddd4651bb82bc3464516bed01b75d543a24cc17f +size 165226 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/rng_state.pth b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..b07d245718b57ab2fe28aa8fcd87fd566f5ca37a --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a7b08a9b6cab3ce1ea404a1560c73adefdab4cd5f0d8619baa241f06062472ab +size 14244 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/scheduler.pt b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..6c655a9f3dd5d4d532f6c0a84b19b9ae5b766437 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dbc2253b3cc1ddd11640ea56d3925f925017f9aa6e1577411dac42d629457af6 +size 1064 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/special_tokens_map.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..a8b3208c2884c4efb86e49300fdd3dc877220cdf --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/special_tokens_map.json @@ -0,0 +1,7 @@ +{ + "cls_token": "[CLS]", + "mask_token": "[MASK]", + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/tokenizer.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..ebd782de6963b9032ce9374cfe30f5275be1a30b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/tokenizer.json @@ -0,0 +1,4265 @@ +{ + "version": "1.0", + "truncation": { + "direction": "Right", + "max_length": 512, + "strategy": "LongestFirst", + "stride": 0 + }, + "padding": { + "strategy": { + "Fixed": 512 + }, + "direction": "Right", + "pad_to_multiple_of": null, + "pad_id": 0, + "pad_type_id": 0, + "pad_token": "[PAD]" + }, + "added_tokens": [ + { + "id": 0, + "content": "[PAD]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 1, + "content": "[UNK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 2, + "content": "[CLS]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 3, + "content": "[SEP]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 4, + "content": "[MASK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + } + ], + "normalizer": { + "type": "BertNormalizer", + "clean_text": true, + "handle_chinese_chars": true, + "strip_accents": null, + "lowercase": false + }, + "pre_tokenizer": { + "type": "BertPreTokenizer" + }, + "post_processor": { + "type": "TemplateProcessing", + "single": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + } + ], + "pair": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "B", + "type_id": 1 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 1 + } + } + ], + "special_tokens": { + "[CLS]": { + "id": "[CLS]", + "ids": [ + 2 + ], + "tokens": [ + "[CLS]" + ] + }, + "[SEP]": { + "id": "[SEP]", + "ids": [ + 3 + ], + "tokens": [ + "[SEP]" + ] + } + } + }, + "decoder": { + "type": "WordPiece", + "prefix": "##", + "cleanup": true + }, + "model": { + "type": "WordPiece", + "unk_token": "[UNK]", + "continuing_subword_prefix": "##", + "max_input_chars_per_word": 100, + "vocab": { + "[PAD]": 0, + "[UNK]": 1, + "[CLS]": 2, + "[SEP]": 3, + "[MASK]": 4, + "AAAAAA": 5, + "AAAAAT": 6, + "AAAAAC": 7, + "AAAAAG": 8, + "AAAATA": 9, + "AAAATT": 10, + "AAAATC": 11, + "AAAATG": 12, + "AAAACA": 13, + "AAAACT": 14, + "AAAACC": 15, + "AAAACG": 16, + "AAAAGA": 17, + "AAAAGT": 18, + "AAAAGC": 19, + "AAAAGG": 20, + "AAATAA": 21, + "AAATAT": 22, + "AAATAC": 23, + "AAATAG": 24, + "AAATTA": 25, + "AAATTT": 26, + "AAATTC": 27, + "AAATTG": 28, + "AAATCA": 29, + "AAATCT": 30, + "AAATCC": 31, + "AAATCG": 32, + "AAATGA": 33, + "AAATGT": 34, + "AAATGC": 35, + "AAATGG": 36, + "AAACAA": 37, + "AAACAT": 38, + "AAACAC": 39, + "AAACAG": 40, + "AAACTA": 41, + "AAACTT": 42, + "AAACTC": 43, + "AAACTG": 44, + "AAACCA": 45, + "AAACCT": 46, + "AAACCC": 47, + "AAACCG": 48, + "AAACGA": 49, + "AAACGT": 50, + "AAACGC": 51, + "AAACGG": 52, + "AAAGAA": 53, + "AAAGAT": 54, + "AAAGAC": 55, + "AAAGAG": 56, + "AAAGTA": 57, + "AAAGTT": 58, + "AAAGTC": 59, + "AAAGTG": 60, + "AAAGCA": 61, + "AAAGCT": 62, + "AAAGCC": 63, + "AAAGCG": 64, + "AAAGGA": 65, + "AAAGGT": 66, + "AAAGGC": 67, + "AAAGGG": 68, + "AATAAA": 69, + "AATAAT": 70, + "AATAAC": 71, + "AATAAG": 72, + "AATATA": 73, + "AATATT": 74, + "AATATC": 75, + "AATATG": 76, + "AATACA": 77, + "AATACT": 78, + "AATACC": 79, + "AATACG": 80, + "AATAGA": 81, + "AATAGT": 82, + "AATAGC": 83, + "AATAGG": 84, + "AATTAA": 85, + "AATTAT": 86, + "AATTAC": 87, + "AATTAG": 88, + "AATTTA": 89, + "AATTTT": 90, + "AATTTC": 91, + "AATTTG": 92, + "AATTCA": 93, + "AATTCT": 94, + "AATTCC": 95, + "AATTCG": 96, + "AATTGA": 97, + "AATTGT": 98, + "AATTGC": 99, + "AATTGG": 100, + "AATCAA": 101, + "AATCAT": 102, + "AATCAC": 103, + "AATCAG": 104, + "AATCTA": 105, + "AATCTT": 106, + "AATCTC": 107, + "AATCTG": 108, + "AATCCA": 109, + "AATCCT": 110, + "AATCCC": 111, + "AATCCG": 112, + "AATCGA": 113, + "AATCGT": 114, + "AATCGC": 115, + "AATCGG": 116, + "AATGAA": 117, + "AATGAT": 118, + "AATGAC": 119, + "AATGAG": 120, + "AATGTA": 121, + "AATGTT": 122, + "AATGTC": 123, + "AATGTG": 124, + "AATGCA": 125, + "AATGCT": 126, + "AATGCC": 127, + "AATGCG": 128, + "AATGGA": 129, + "AATGGT": 130, + "AATGGC": 131, + "AATGGG": 132, + "AACAAA": 133, + "AACAAT": 134, + "AACAAC": 135, + "AACAAG": 136, + "AACATA": 137, + "AACATT": 138, + "AACATC": 139, + "AACATG": 140, + "AACACA": 141, + "AACACT": 142, + "AACACC": 143, + "AACACG": 144, + "AACAGA": 145, + "AACAGT": 146, + "AACAGC": 147, + "AACAGG": 148, + "AACTAA": 149, + "AACTAT": 150, + "AACTAC": 151, + "AACTAG": 152, + "AACTTA": 153, + "AACTTT": 154, + "AACTTC": 155, + "AACTTG": 156, + "AACTCA": 157, + "AACTCT": 158, + "AACTCC": 159, + "AACTCG": 160, + "AACTGA": 161, + "AACTGT": 162, + "AACTGC": 163, + "AACTGG": 164, + "AACCAA": 165, + "AACCAT": 166, + "AACCAC": 167, + "AACCAG": 168, + "AACCTA": 169, + "AACCTT": 170, + "AACCTC": 171, + "AACCTG": 172, + "AACCCA": 173, + "AACCCT": 174, + "AACCCC": 175, + "AACCCG": 176, + "AACCGA": 177, + "AACCGT": 178, + "AACCGC": 179, + "AACCGG": 180, + "AACGAA": 181, + "AACGAT": 182, + "AACGAC": 183, + "AACGAG": 184, + "AACGTA": 185, + "AACGTT": 186, + "AACGTC": 187, + "AACGTG": 188, + "AACGCA": 189, + "AACGCT": 190, + "AACGCC": 191, + "AACGCG": 192, + "AACGGA": 193, + "AACGGT": 194, + "AACGGC": 195, + "AACGGG": 196, + "AAGAAA": 197, + "AAGAAT": 198, + "AAGAAC": 199, + "AAGAAG": 200, + "AAGATA": 201, + "AAGATT": 202, + "AAGATC": 203, + "AAGATG": 204, + "AAGACA": 205, + "AAGACT": 206, + "AAGACC": 207, + "AAGACG": 208, + "AAGAGA": 209, + "AAGAGT": 210, + "AAGAGC": 211, + "AAGAGG": 212, + "AAGTAA": 213, + "AAGTAT": 214, + "AAGTAC": 215, + "AAGTAG": 216, + "AAGTTA": 217, + "AAGTTT": 218, + "AAGTTC": 219, + "AAGTTG": 220, + "AAGTCA": 221, + "AAGTCT": 222, + "AAGTCC": 223, + "AAGTCG": 224, + "AAGTGA": 225, + "AAGTGT": 226, + "AAGTGC": 227, + "AAGTGG": 228, + "AAGCAA": 229, + "AAGCAT": 230, + "AAGCAC": 231, + "AAGCAG": 232, + "AAGCTA": 233, + "AAGCTT": 234, + "AAGCTC": 235, + "AAGCTG": 236, + "AAGCCA": 237, + "AAGCCT": 238, + "AAGCCC": 239, + "AAGCCG": 240, + "AAGCGA": 241, + "AAGCGT": 242, + "AAGCGC": 243, + "AAGCGG": 244, + "AAGGAA": 245, + "AAGGAT": 246, + "AAGGAC": 247, + "AAGGAG": 248, + "AAGGTA": 249, + "AAGGTT": 250, + "AAGGTC": 251, + "AAGGTG": 252, + "AAGGCA": 253, + "AAGGCT": 254, + "AAGGCC": 255, + "AAGGCG": 256, + "AAGGGA": 257, + "AAGGGT": 258, + "AAGGGC": 259, + "AAGGGG": 260, + "ATAAAA": 261, + "ATAAAT": 262, + "ATAAAC": 263, + "ATAAAG": 264, + "ATAATA": 265, + "ATAATT": 266, + "ATAATC": 267, + "ATAATG": 268, + "ATAACA": 269, + "ATAACT": 270, + "ATAACC": 271, + "ATAACG": 272, + "ATAAGA": 273, + "ATAAGT": 274, + "ATAAGC": 275, + "ATAAGG": 276, + "ATATAA": 277, + "ATATAT": 278, + "ATATAC": 279, + "ATATAG": 280, + "ATATTA": 281, + "ATATTT": 282, + "ATATTC": 283, + "ATATTG": 284, + "ATATCA": 285, + "ATATCT": 286, + "ATATCC": 287, + "ATATCG": 288, + "ATATGA": 289, + "ATATGT": 290, + "ATATGC": 291, + "ATATGG": 292, + "ATACAA": 293, + "ATACAT": 294, + "ATACAC": 295, + "ATACAG": 296, + "ATACTA": 297, + "ATACTT": 298, + "ATACTC": 299, + "ATACTG": 300, + "ATACCA": 301, + "ATACCT": 302, + "ATACCC": 303, + "ATACCG": 304, + "ATACGA": 305, + "ATACGT": 306, + "ATACGC": 307, + "ATACGG": 308, + "ATAGAA": 309, + "ATAGAT": 310, + "ATAGAC": 311, + "ATAGAG": 312, + "ATAGTA": 313, + "ATAGTT": 314, + "ATAGTC": 315, + "ATAGTG": 316, + "ATAGCA": 317, + "ATAGCT": 318, + "ATAGCC": 319, + "ATAGCG": 320, + "ATAGGA": 321, + "ATAGGT": 322, + "ATAGGC": 323, + "ATAGGG": 324, + "ATTAAA": 325, + "ATTAAT": 326, + "ATTAAC": 327, + "ATTAAG": 328, + "ATTATA": 329, + "ATTATT": 330, + "ATTATC": 331, + "ATTATG": 332, + "ATTACA": 333, + "ATTACT": 334, + "ATTACC": 335, + "ATTACG": 336, + "ATTAGA": 337, + "ATTAGT": 338, + "ATTAGC": 339, + "ATTAGG": 340, + "ATTTAA": 341, + "ATTTAT": 342, + "ATTTAC": 343, + "ATTTAG": 344, + "ATTTTA": 345, + "ATTTTT": 346, + "ATTTTC": 347, + "ATTTTG": 348, + "ATTTCA": 349, + "ATTTCT": 350, + "ATTTCC": 351, + "ATTTCG": 352, + "ATTTGA": 353, + "ATTTGT": 354, + "ATTTGC": 355, + "ATTTGG": 356, + "ATTCAA": 357, + "ATTCAT": 358, + "ATTCAC": 359, + "ATTCAG": 360, + "ATTCTA": 361, + "ATTCTT": 362, + "ATTCTC": 363, + "ATTCTG": 364, + "ATTCCA": 365, + "ATTCCT": 366, + "ATTCCC": 367, + "ATTCCG": 368, + "ATTCGA": 369, + "ATTCGT": 370, + "ATTCGC": 371, + "ATTCGG": 372, + "ATTGAA": 373, + "ATTGAT": 374, + "ATTGAC": 375, + "ATTGAG": 376, + "ATTGTA": 377, + "ATTGTT": 378, + "ATTGTC": 379, + "ATTGTG": 380, + "ATTGCA": 381, + "ATTGCT": 382, + "ATTGCC": 383, + "ATTGCG": 384, + "ATTGGA": 385, + "ATTGGT": 386, + "ATTGGC": 387, + "ATTGGG": 388, + "ATCAAA": 389, + "ATCAAT": 390, + "ATCAAC": 391, + "ATCAAG": 392, + "ATCATA": 393, + "ATCATT": 394, + "ATCATC": 395, + "ATCATG": 396, + "ATCACA": 397, + "ATCACT": 398, + "ATCACC": 399, + "ATCACG": 400, + "ATCAGA": 401, + "ATCAGT": 402, + "ATCAGC": 403, + "ATCAGG": 404, + "ATCTAA": 405, + "ATCTAT": 406, + "ATCTAC": 407, + "ATCTAG": 408, + "ATCTTA": 409, + "ATCTTT": 410, + "ATCTTC": 411, + "ATCTTG": 412, + "ATCTCA": 413, + "ATCTCT": 414, + "ATCTCC": 415, + "ATCTCG": 416, + "ATCTGA": 417, + "ATCTGT": 418, + "ATCTGC": 419, + "ATCTGG": 420, + "ATCCAA": 421, + "ATCCAT": 422, + "ATCCAC": 423, + "ATCCAG": 424, + "ATCCTA": 425, + "ATCCTT": 426, + "ATCCTC": 427, + "ATCCTG": 428, + "ATCCCA": 429, + "ATCCCT": 430, + "ATCCCC": 431, + "ATCCCG": 432, + "ATCCGA": 433, + "ATCCGT": 434, + "ATCCGC": 435, + "ATCCGG": 436, + "ATCGAA": 437, + "ATCGAT": 438, + "ATCGAC": 439, + "ATCGAG": 440, + "ATCGTA": 441, + "ATCGTT": 442, + "ATCGTC": 443, + "ATCGTG": 444, + "ATCGCA": 445, + "ATCGCT": 446, + "ATCGCC": 447, + "ATCGCG": 448, + "ATCGGA": 449, + "ATCGGT": 450, + "ATCGGC": 451, + "ATCGGG": 452, + "ATGAAA": 453, + "ATGAAT": 454, + "ATGAAC": 455, + "ATGAAG": 456, + "ATGATA": 457, + "ATGATT": 458, + "ATGATC": 459, + "ATGATG": 460, + "ATGACA": 461, + "ATGACT": 462, + "ATGACC": 463, + "ATGACG": 464, + "ATGAGA": 465, + "ATGAGT": 466, + "ATGAGC": 467, + "ATGAGG": 468, + "ATGTAA": 469, + "ATGTAT": 470, + "ATGTAC": 471, + "ATGTAG": 472, + "ATGTTA": 473, + "ATGTTT": 474, + "ATGTTC": 475, + "ATGTTG": 476, + "ATGTCA": 477, + "ATGTCT": 478, + "ATGTCC": 479, + "ATGTCG": 480, + "ATGTGA": 481, + "ATGTGT": 482, + "ATGTGC": 483, + "ATGTGG": 484, + "ATGCAA": 485, + "ATGCAT": 486, + "ATGCAC": 487, + "ATGCAG": 488, + "ATGCTA": 489, + "ATGCTT": 490, + "ATGCTC": 491, + "ATGCTG": 492, + "ATGCCA": 493, + "ATGCCT": 494, + "ATGCCC": 495, + "ATGCCG": 496, + "ATGCGA": 497, + "ATGCGT": 498, + "ATGCGC": 499, + "ATGCGG": 500, + "ATGGAA": 501, + "ATGGAT": 502, + "ATGGAC": 503, + "ATGGAG": 504, + "ATGGTA": 505, + "ATGGTT": 506, + "ATGGTC": 507, + "ATGGTG": 508, + "ATGGCA": 509, + "ATGGCT": 510, + "ATGGCC": 511, + "ATGGCG": 512, + "ATGGGA": 513, + "ATGGGT": 514, + "ATGGGC": 515, + "ATGGGG": 516, + "ACAAAA": 517, + "ACAAAT": 518, + "ACAAAC": 519, + "ACAAAG": 520, + "ACAATA": 521, + "ACAATT": 522, + "ACAATC": 523, + "ACAATG": 524, + "ACAACA": 525, + "ACAACT": 526, + "ACAACC": 527, + "ACAACG": 528, + "ACAAGA": 529, + "ACAAGT": 530, + "ACAAGC": 531, + "ACAAGG": 532, + "ACATAA": 533, + "ACATAT": 534, + "ACATAC": 535, + "ACATAG": 536, + "ACATTA": 537, + "ACATTT": 538, + "ACATTC": 539, + "ACATTG": 540, + "ACATCA": 541, + "ACATCT": 542, + "ACATCC": 543, + "ACATCG": 544, + "ACATGA": 545, + "ACATGT": 546, + "ACATGC": 547, + "ACATGG": 548, + "ACACAA": 549, + "ACACAT": 550, + "ACACAC": 551, + "ACACAG": 552, + "ACACTA": 553, + "ACACTT": 554, + "ACACTC": 555, + "ACACTG": 556, + "ACACCA": 557, + "ACACCT": 558, + "ACACCC": 559, + "ACACCG": 560, + "ACACGA": 561, + "ACACGT": 562, + "ACACGC": 563, + "ACACGG": 564, + "ACAGAA": 565, + "ACAGAT": 566, + "ACAGAC": 567, + "ACAGAG": 568, + "ACAGTA": 569, + "ACAGTT": 570, + "ACAGTC": 571, + "ACAGTG": 572, + "ACAGCA": 573, + "ACAGCT": 574, + "ACAGCC": 575, + "ACAGCG": 576, + "ACAGGA": 577, + "ACAGGT": 578, + "ACAGGC": 579, + "ACAGGG": 580, + "ACTAAA": 581, + "ACTAAT": 582, + "ACTAAC": 583, + "ACTAAG": 584, + "ACTATA": 585, + "ACTATT": 586, + "ACTATC": 587, + "ACTATG": 588, + "ACTACA": 589, + "ACTACT": 590, + "ACTACC": 591, + "ACTACG": 592, + "ACTAGA": 593, + "ACTAGT": 594, + "ACTAGC": 595, + "ACTAGG": 596, + "ACTTAA": 597, + "ACTTAT": 598, + "ACTTAC": 599, + "ACTTAG": 600, + "ACTTTA": 601, + "ACTTTT": 602, + "ACTTTC": 603, + "ACTTTG": 604, + "ACTTCA": 605, + "ACTTCT": 606, + "ACTTCC": 607, + "ACTTCG": 608, + "ACTTGA": 609, + "ACTTGT": 610, + "ACTTGC": 611, + "ACTTGG": 612, + "ACTCAA": 613, + "ACTCAT": 614, + "ACTCAC": 615, + "ACTCAG": 616, + "ACTCTA": 617, + "ACTCTT": 618, + "ACTCTC": 619, + "ACTCTG": 620, + "ACTCCA": 621, + "ACTCCT": 622, + "ACTCCC": 623, + "ACTCCG": 624, + "ACTCGA": 625, + "ACTCGT": 626, + "ACTCGC": 627, + "ACTCGG": 628, + "ACTGAA": 629, + "ACTGAT": 630, + "ACTGAC": 631, + "ACTGAG": 632, + "ACTGTA": 633, + "ACTGTT": 634, + "ACTGTC": 635, + "ACTGTG": 636, + "ACTGCA": 637, + "ACTGCT": 638, + "ACTGCC": 639, + "ACTGCG": 640, + "ACTGGA": 641, + "ACTGGT": 642, + "ACTGGC": 643, + "ACTGGG": 644, + "ACCAAA": 645, + "ACCAAT": 646, + "ACCAAC": 647, + "ACCAAG": 648, + "ACCATA": 649, + "ACCATT": 650, + "ACCATC": 651, + "ACCATG": 652, + "ACCACA": 653, + "ACCACT": 654, + "ACCACC": 655, + "ACCACG": 656, + "ACCAGA": 657, + "ACCAGT": 658, + "ACCAGC": 659, + "ACCAGG": 660, + "ACCTAA": 661, + "ACCTAT": 662, + "ACCTAC": 663, + "ACCTAG": 664, + "ACCTTA": 665, + "ACCTTT": 666, + "ACCTTC": 667, + "ACCTTG": 668, + "ACCTCA": 669, + "ACCTCT": 670, + "ACCTCC": 671, + "ACCTCG": 672, + "ACCTGA": 673, + "ACCTGT": 674, + "ACCTGC": 675, + "ACCTGG": 676, + "ACCCAA": 677, + "ACCCAT": 678, + "ACCCAC": 679, + "ACCCAG": 680, + "ACCCTA": 681, + "ACCCTT": 682, + "ACCCTC": 683, + "ACCCTG": 684, + "ACCCCA": 685, + "ACCCCT": 686, + "ACCCCC": 687, + "ACCCCG": 688, + "ACCCGA": 689, + "ACCCGT": 690, + "ACCCGC": 691, + "ACCCGG": 692, + "ACCGAA": 693, + "ACCGAT": 694, + "ACCGAC": 695, + "ACCGAG": 696, + "ACCGTA": 697, + "ACCGTT": 698, + "ACCGTC": 699, + "ACCGTG": 700, + "ACCGCA": 701, + "ACCGCT": 702, + "ACCGCC": 703, + "ACCGCG": 704, + "ACCGGA": 705, + "ACCGGT": 706, + "ACCGGC": 707, + "ACCGGG": 708, + "ACGAAA": 709, + "ACGAAT": 710, + "ACGAAC": 711, + "ACGAAG": 712, + "ACGATA": 713, + "ACGATT": 714, + "ACGATC": 715, + "ACGATG": 716, + "ACGACA": 717, + "ACGACT": 718, + "ACGACC": 719, + "ACGACG": 720, + "ACGAGA": 721, + "ACGAGT": 722, + "ACGAGC": 723, + "ACGAGG": 724, + "ACGTAA": 725, + "ACGTAT": 726, + "ACGTAC": 727, + "ACGTAG": 728, + "ACGTTA": 729, + "ACGTTT": 730, + "ACGTTC": 731, + "ACGTTG": 732, + "ACGTCA": 733, + "ACGTCT": 734, + "ACGTCC": 735, + "ACGTCG": 736, + "ACGTGA": 737, + "ACGTGT": 738, + "ACGTGC": 739, + "ACGTGG": 740, + "ACGCAA": 741, + "ACGCAT": 742, + "ACGCAC": 743, + "ACGCAG": 744, + "ACGCTA": 745, + "ACGCTT": 746, + "ACGCTC": 747, + "ACGCTG": 748, + "ACGCCA": 749, + "ACGCCT": 750, + "ACGCCC": 751, + "ACGCCG": 752, + "ACGCGA": 753, + "ACGCGT": 754, + "ACGCGC": 755, + "ACGCGG": 756, + "ACGGAA": 757, + "ACGGAT": 758, + "ACGGAC": 759, + "ACGGAG": 760, + "ACGGTA": 761, + "ACGGTT": 762, + "ACGGTC": 763, + "ACGGTG": 764, + "ACGGCA": 765, + "ACGGCT": 766, + "ACGGCC": 767, + "ACGGCG": 768, + "ACGGGA": 769, + "ACGGGT": 770, + "ACGGGC": 771, + "ACGGGG": 772, + "AGAAAA": 773, + "AGAAAT": 774, + "AGAAAC": 775, + "AGAAAG": 776, + "AGAATA": 777, + "AGAATT": 778, + "AGAATC": 779, + "AGAATG": 780, + "AGAACA": 781, + "AGAACT": 782, + "AGAACC": 783, + "AGAACG": 784, + "AGAAGA": 785, + "AGAAGT": 786, + "AGAAGC": 787, + "AGAAGG": 788, + "AGATAA": 789, + "AGATAT": 790, + "AGATAC": 791, + "AGATAG": 792, + "AGATTA": 793, + "AGATTT": 794, + "AGATTC": 795, + "AGATTG": 796, + "AGATCA": 797, + "AGATCT": 798, + "AGATCC": 799, + "AGATCG": 800, + "AGATGA": 801, + "AGATGT": 802, + "AGATGC": 803, + "AGATGG": 804, + "AGACAA": 805, + "AGACAT": 806, + "AGACAC": 807, + "AGACAG": 808, + "AGACTA": 809, + "AGACTT": 810, + "AGACTC": 811, + "AGACTG": 812, + "AGACCA": 813, + "AGACCT": 814, + "AGACCC": 815, + "AGACCG": 816, + "AGACGA": 817, + "AGACGT": 818, + "AGACGC": 819, + "AGACGG": 820, + "AGAGAA": 821, + "AGAGAT": 822, + "AGAGAC": 823, + "AGAGAG": 824, + "AGAGTA": 825, + "AGAGTT": 826, + "AGAGTC": 827, + "AGAGTG": 828, + "AGAGCA": 829, + "AGAGCT": 830, + "AGAGCC": 831, + "AGAGCG": 832, + "AGAGGA": 833, + "AGAGGT": 834, + "AGAGGC": 835, + "AGAGGG": 836, + "AGTAAA": 837, + "AGTAAT": 838, + "AGTAAC": 839, + "AGTAAG": 840, + "AGTATA": 841, + "AGTATT": 842, + "AGTATC": 843, + "AGTATG": 844, + "AGTACA": 845, + "AGTACT": 846, + "AGTACC": 847, + "AGTACG": 848, + "AGTAGA": 849, + "AGTAGT": 850, + "AGTAGC": 851, + "AGTAGG": 852, + "AGTTAA": 853, + "AGTTAT": 854, + "AGTTAC": 855, + "AGTTAG": 856, + "AGTTTA": 857, + "AGTTTT": 858, + "AGTTTC": 859, + "AGTTTG": 860, + "AGTTCA": 861, + "AGTTCT": 862, + "AGTTCC": 863, + "AGTTCG": 864, + "AGTTGA": 865, + "AGTTGT": 866, + "AGTTGC": 867, + "AGTTGG": 868, + "AGTCAA": 869, + "AGTCAT": 870, + "AGTCAC": 871, + "AGTCAG": 872, + "AGTCTA": 873, + "AGTCTT": 874, + "AGTCTC": 875, + "AGTCTG": 876, + "AGTCCA": 877, + "AGTCCT": 878, + "AGTCCC": 879, + "AGTCCG": 880, + "AGTCGA": 881, + "AGTCGT": 882, + "AGTCGC": 883, + "AGTCGG": 884, + "AGTGAA": 885, + "AGTGAT": 886, + "AGTGAC": 887, + "AGTGAG": 888, + "AGTGTA": 889, + "AGTGTT": 890, + "AGTGTC": 891, + "AGTGTG": 892, + "AGTGCA": 893, + "AGTGCT": 894, + "AGTGCC": 895, + "AGTGCG": 896, + "AGTGGA": 897, + "AGTGGT": 898, + "AGTGGC": 899, + "AGTGGG": 900, + "AGCAAA": 901, + "AGCAAT": 902, + "AGCAAC": 903, + "AGCAAG": 904, + "AGCATA": 905, + "AGCATT": 906, + "AGCATC": 907, + "AGCATG": 908, + "AGCACA": 909, + "AGCACT": 910, + "AGCACC": 911, + "AGCACG": 912, + "AGCAGA": 913, + "AGCAGT": 914, + "AGCAGC": 915, + "AGCAGG": 916, + "AGCTAA": 917, + "AGCTAT": 918, + "AGCTAC": 919, + "AGCTAG": 920, + "AGCTTA": 921, + "AGCTTT": 922, + "AGCTTC": 923, + "AGCTTG": 924, + "AGCTCA": 925, + "AGCTCT": 926, + "AGCTCC": 927, + "AGCTCG": 928, + "AGCTGA": 929, + "AGCTGT": 930, + "AGCTGC": 931, + "AGCTGG": 932, + "AGCCAA": 933, + "AGCCAT": 934, + "AGCCAC": 935, + "AGCCAG": 936, + "AGCCTA": 937, + "AGCCTT": 938, + "AGCCTC": 939, + "AGCCTG": 940, + "AGCCCA": 941, + "AGCCCT": 942, + "AGCCCC": 943, + "AGCCCG": 944, + "AGCCGA": 945, + "AGCCGT": 946, + "AGCCGC": 947, + "AGCCGG": 948, + "AGCGAA": 949, + "AGCGAT": 950, + "AGCGAC": 951, + "AGCGAG": 952, + "AGCGTA": 953, + "AGCGTT": 954, + "AGCGTC": 955, + "AGCGTG": 956, + "AGCGCA": 957, + "AGCGCT": 958, + "AGCGCC": 959, + "AGCGCG": 960, + "AGCGGA": 961, + "AGCGGT": 962, + "AGCGGC": 963, + "AGCGGG": 964, + "AGGAAA": 965, + "AGGAAT": 966, + "AGGAAC": 967, + "AGGAAG": 968, + "AGGATA": 969, + "AGGATT": 970, + "AGGATC": 971, + "AGGATG": 972, + "AGGACA": 973, + "AGGACT": 974, + "AGGACC": 975, + "AGGACG": 976, + "AGGAGA": 977, + "AGGAGT": 978, + "AGGAGC": 979, + "AGGAGG": 980, + "AGGTAA": 981, + "AGGTAT": 982, + "AGGTAC": 983, + "AGGTAG": 984, + "AGGTTA": 985, + "AGGTTT": 986, + "AGGTTC": 987, + "AGGTTG": 988, + "AGGTCA": 989, + "AGGTCT": 990, + "AGGTCC": 991, + "AGGTCG": 992, + "AGGTGA": 993, + "AGGTGT": 994, + "AGGTGC": 995, + "AGGTGG": 996, + "AGGCAA": 997, + "AGGCAT": 998, + "AGGCAC": 999, + "AGGCAG": 1000, + "AGGCTA": 1001, + "AGGCTT": 1002, + "AGGCTC": 1003, + "AGGCTG": 1004, + "AGGCCA": 1005, + "AGGCCT": 1006, + "AGGCCC": 1007, + "AGGCCG": 1008, + "AGGCGA": 1009, + "AGGCGT": 1010, + "AGGCGC": 1011, + "AGGCGG": 1012, + "AGGGAA": 1013, + "AGGGAT": 1014, + "AGGGAC": 1015, + "AGGGAG": 1016, + "AGGGTA": 1017, + "AGGGTT": 1018, + "AGGGTC": 1019, + "AGGGTG": 1020, + "AGGGCA": 1021, + "AGGGCT": 1022, + "AGGGCC": 1023, + "AGGGCG": 1024, + "AGGGGA": 1025, + "AGGGGT": 1026, + "AGGGGC": 1027, + "AGGGGG": 1028, + "TAAAAA": 1029, + "TAAAAT": 1030, + "TAAAAC": 1031, + "TAAAAG": 1032, + "TAAATA": 1033, + "TAAATT": 1034, + "TAAATC": 1035, + "TAAATG": 1036, + "TAAACA": 1037, + "TAAACT": 1038, + "TAAACC": 1039, + "TAAACG": 1040, + "TAAAGA": 1041, + "TAAAGT": 1042, + "TAAAGC": 1043, + "TAAAGG": 1044, + "TAATAA": 1045, + "TAATAT": 1046, + "TAATAC": 1047, + "TAATAG": 1048, + "TAATTA": 1049, + "TAATTT": 1050, + "TAATTC": 1051, + "TAATTG": 1052, + "TAATCA": 1053, + "TAATCT": 1054, + "TAATCC": 1055, + "TAATCG": 1056, + "TAATGA": 1057, + "TAATGT": 1058, + "TAATGC": 1059, + "TAATGG": 1060, + "TAACAA": 1061, + "TAACAT": 1062, + "TAACAC": 1063, + "TAACAG": 1064, + "TAACTA": 1065, + "TAACTT": 1066, + "TAACTC": 1067, + "TAACTG": 1068, + "TAACCA": 1069, + "TAACCT": 1070, + "TAACCC": 1071, + "TAACCG": 1072, + "TAACGA": 1073, + "TAACGT": 1074, + "TAACGC": 1075, + "TAACGG": 1076, + "TAAGAA": 1077, + "TAAGAT": 1078, + "TAAGAC": 1079, + "TAAGAG": 1080, + "TAAGTA": 1081, + "TAAGTT": 1082, + "TAAGTC": 1083, + "TAAGTG": 1084, + "TAAGCA": 1085, + "TAAGCT": 1086, + "TAAGCC": 1087, + "TAAGCG": 1088, + "TAAGGA": 1089, + "TAAGGT": 1090, + "TAAGGC": 1091, + "TAAGGG": 1092, + "TATAAA": 1093, + "TATAAT": 1094, + "TATAAC": 1095, + "TATAAG": 1096, + "TATATA": 1097, + "TATATT": 1098, + "TATATC": 1099, + "TATATG": 1100, + "TATACA": 1101, + "TATACT": 1102, + "TATACC": 1103, + "TATACG": 1104, + "TATAGA": 1105, + "TATAGT": 1106, + "TATAGC": 1107, + "TATAGG": 1108, + "TATTAA": 1109, + "TATTAT": 1110, + "TATTAC": 1111, + "TATTAG": 1112, + "TATTTA": 1113, + "TATTTT": 1114, + "TATTTC": 1115, + "TATTTG": 1116, + "TATTCA": 1117, + "TATTCT": 1118, + "TATTCC": 1119, + "TATTCG": 1120, + "TATTGA": 1121, + "TATTGT": 1122, + "TATTGC": 1123, + "TATTGG": 1124, + "TATCAA": 1125, + "TATCAT": 1126, + "TATCAC": 1127, + "TATCAG": 1128, + "TATCTA": 1129, + "TATCTT": 1130, + "TATCTC": 1131, + "TATCTG": 1132, + "TATCCA": 1133, + "TATCCT": 1134, + "TATCCC": 1135, + "TATCCG": 1136, + "TATCGA": 1137, + "TATCGT": 1138, + "TATCGC": 1139, + "TATCGG": 1140, + "TATGAA": 1141, + "TATGAT": 1142, + "TATGAC": 1143, + "TATGAG": 1144, + "TATGTA": 1145, + "TATGTT": 1146, + "TATGTC": 1147, + "TATGTG": 1148, + "TATGCA": 1149, + "TATGCT": 1150, + "TATGCC": 1151, + "TATGCG": 1152, + "TATGGA": 1153, + "TATGGT": 1154, + "TATGGC": 1155, + "TATGGG": 1156, + "TACAAA": 1157, + "TACAAT": 1158, + "TACAAC": 1159, + "TACAAG": 1160, + "TACATA": 1161, + "TACATT": 1162, + "TACATC": 1163, + "TACATG": 1164, + "TACACA": 1165, + "TACACT": 1166, + "TACACC": 1167, + "TACACG": 1168, + "TACAGA": 1169, + "TACAGT": 1170, + "TACAGC": 1171, + "TACAGG": 1172, + "TACTAA": 1173, + "TACTAT": 1174, + "TACTAC": 1175, + "TACTAG": 1176, + "TACTTA": 1177, + "TACTTT": 1178, + "TACTTC": 1179, + "TACTTG": 1180, + "TACTCA": 1181, + "TACTCT": 1182, + "TACTCC": 1183, + "TACTCG": 1184, + "TACTGA": 1185, + "TACTGT": 1186, + "TACTGC": 1187, + "TACTGG": 1188, + "TACCAA": 1189, + "TACCAT": 1190, + "TACCAC": 1191, + "TACCAG": 1192, + "TACCTA": 1193, + "TACCTT": 1194, + "TACCTC": 1195, + "TACCTG": 1196, + "TACCCA": 1197, + "TACCCT": 1198, + "TACCCC": 1199, + "TACCCG": 1200, + "TACCGA": 1201, + "TACCGT": 1202, + "TACCGC": 1203, + "TACCGG": 1204, + "TACGAA": 1205, + "TACGAT": 1206, + "TACGAC": 1207, + "TACGAG": 1208, + "TACGTA": 1209, + "TACGTT": 1210, + "TACGTC": 1211, + "TACGTG": 1212, + "TACGCA": 1213, + "TACGCT": 1214, + "TACGCC": 1215, + "TACGCG": 1216, + "TACGGA": 1217, + "TACGGT": 1218, + "TACGGC": 1219, + "TACGGG": 1220, + "TAGAAA": 1221, + "TAGAAT": 1222, + "TAGAAC": 1223, + "TAGAAG": 1224, + "TAGATA": 1225, + "TAGATT": 1226, + "TAGATC": 1227, + "TAGATG": 1228, + "TAGACA": 1229, + "TAGACT": 1230, + "TAGACC": 1231, + "TAGACG": 1232, + "TAGAGA": 1233, + "TAGAGT": 1234, + "TAGAGC": 1235, + "TAGAGG": 1236, + "TAGTAA": 1237, + "TAGTAT": 1238, + "TAGTAC": 1239, + "TAGTAG": 1240, + "TAGTTA": 1241, + "TAGTTT": 1242, + "TAGTTC": 1243, + "TAGTTG": 1244, + "TAGTCA": 1245, + "TAGTCT": 1246, + "TAGTCC": 1247, + "TAGTCG": 1248, + "TAGTGA": 1249, + "TAGTGT": 1250, + "TAGTGC": 1251, + "TAGTGG": 1252, + "TAGCAA": 1253, + "TAGCAT": 1254, + "TAGCAC": 1255, + "TAGCAG": 1256, + "TAGCTA": 1257, + "TAGCTT": 1258, + "TAGCTC": 1259, + "TAGCTG": 1260, + "TAGCCA": 1261, + "TAGCCT": 1262, + "TAGCCC": 1263, + "TAGCCG": 1264, + "TAGCGA": 1265, + "TAGCGT": 1266, + "TAGCGC": 1267, + "TAGCGG": 1268, + "TAGGAA": 1269, + "TAGGAT": 1270, + "TAGGAC": 1271, + "TAGGAG": 1272, + "TAGGTA": 1273, + "TAGGTT": 1274, + "TAGGTC": 1275, + "TAGGTG": 1276, + "TAGGCA": 1277, + "TAGGCT": 1278, + "TAGGCC": 1279, + "TAGGCG": 1280, + "TAGGGA": 1281, + "TAGGGT": 1282, + "TAGGGC": 1283, + "TAGGGG": 1284, + "TTAAAA": 1285, + "TTAAAT": 1286, + "TTAAAC": 1287, + "TTAAAG": 1288, + "TTAATA": 1289, + "TTAATT": 1290, + "TTAATC": 1291, + "TTAATG": 1292, + "TTAACA": 1293, + "TTAACT": 1294, + "TTAACC": 1295, + "TTAACG": 1296, + "TTAAGA": 1297, + "TTAAGT": 1298, + "TTAAGC": 1299, + "TTAAGG": 1300, + "TTATAA": 1301, + "TTATAT": 1302, + "TTATAC": 1303, + "TTATAG": 1304, + "TTATTA": 1305, + "TTATTT": 1306, + "TTATTC": 1307, + "TTATTG": 1308, + "TTATCA": 1309, + "TTATCT": 1310, + "TTATCC": 1311, + "TTATCG": 1312, + "TTATGA": 1313, + "TTATGT": 1314, + "TTATGC": 1315, + "TTATGG": 1316, + "TTACAA": 1317, + "TTACAT": 1318, + "TTACAC": 1319, + "TTACAG": 1320, + "TTACTA": 1321, + "TTACTT": 1322, + "TTACTC": 1323, + "TTACTG": 1324, + "TTACCA": 1325, + "TTACCT": 1326, + "TTACCC": 1327, + "TTACCG": 1328, + "TTACGA": 1329, + "TTACGT": 1330, + "TTACGC": 1331, + "TTACGG": 1332, + "TTAGAA": 1333, + "TTAGAT": 1334, + "TTAGAC": 1335, + "TTAGAG": 1336, + "TTAGTA": 1337, + "TTAGTT": 1338, + "TTAGTC": 1339, + "TTAGTG": 1340, + "TTAGCA": 1341, + "TTAGCT": 1342, + "TTAGCC": 1343, + "TTAGCG": 1344, + "TTAGGA": 1345, + "TTAGGT": 1346, + "TTAGGC": 1347, + "TTAGGG": 1348, + "TTTAAA": 1349, + "TTTAAT": 1350, + "TTTAAC": 1351, + "TTTAAG": 1352, + "TTTATA": 1353, + "TTTATT": 1354, + "TTTATC": 1355, + "TTTATG": 1356, + "TTTACA": 1357, + "TTTACT": 1358, + "TTTACC": 1359, + "TTTACG": 1360, + "TTTAGA": 1361, + "TTTAGT": 1362, + "TTTAGC": 1363, + "TTTAGG": 1364, + "TTTTAA": 1365, + "TTTTAT": 1366, + "TTTTAC": 1367, + "TTTTAG": 1368, + "TTTTTA": 1369, + "TTTTTT": 1370, + "TTTTTC": 1371, + "TTTTTG": 1372, + "TTTTCA": 1373, + "TTTTCT": 1374, + "TTTTCC": 1375, + "TTTTCG": 1376, + "TTTTGA": 1377, + "TTTTGT": 1378, + "TTTTGC": 1379, + "TTTTGG": 1380, + "TTTCAA": 1381, + "TTTCAT": 1382, + "TTTCAC": 1383, + "TTTCAG": 1384, + "TTTCTA": 1385, + "TTTCTT": 1386, + "TTTCTC": 1387, + "TTTCTG": 1388, + "TTTCCA": 1389, + "TTTCCT": 1390, + "TTTCCC": 1391, + "TTTCCG": 1392, + "TTTCGA": 1393, + "TTTCGT": 1394, + "TTTCGC": 1395, + "TTTCGG": 1396, + "TTTGAA": 1397, + "TTTGAT": 1398, + "TTTGAC": 1399, + "TTTGAG": 1400, + "TTTGTA": 1401, + "TTTGTT": 1402, + "TTTGTC": 1403, + "TTTGTG": 1404, + "TTTGCA": 1405, + "TTTGCT": 1406, + "TTTGCC": 1407, + "TTTGCG": 1408, + "TTTGGA": 1409, + "TTTGGT": 1410, + "TTTGGC": 1411, + "TTTGGG": 1412, + "TTCAAA": 1413, + "TTCAAT": 1414, + "TTCAAC": 1415, + "TTCAAG": 1416, + "TTCATA": 1417, + "TTCATT": 1418, + "TTCATC": 1419, + "TTCATG": 1420, + "TTCACA": 1421, + "TTCACT": 1422, + "TTCACC": 1423, + "TTCACG": 1424, + "TTCAGA": 1425, + "TTCAGT": 1426, + "TTCAGC": 1427, + "TTCAGG": 1428, + "TTCTAA": 1429, + "TTCTAT": 1430, + "TTCTAC": 1431, + "TTCTAG": 1432, + "TTCTTA": 1433, + "TTCTTT": 1434, + "TTCTTC": 1435, + "TTCTTG": 1436, + "TTCTCA": 1437, + "TTCTCT": 1438, + "TTCTCC": 1439, + "TTCTCG": 1440, + "TTCTGA": 1441, + "TTCTGT": 1442, + "TTCTGC": 1443, + "TTCTGG": 1444, + "TTCCAA": 1445, + "TTCCAT": 1446, + "TTCCAC": 1447, + "TTCCAG": 1448, + "TTCCTA": 1449, + "TTCCTT": 1450, + "TTCCTC": 1451, + "TTCCTG": 1452, + "TTCCCA": 1453, + "TTCCCT": 1454, + "TTCCCC": 1455, + "TTCCCG": 1456, + "TTCCGA": 1457, + "TTCCGT": 1458, + "TTCCGC": 1459, + "TTCCGG": 1460, + "TTCGAA": 1461, + "TTCGAT": 1462, + "TTCGAC": 1463, + "TTCGAG": 1464, + "TTCGTA": 1465, + "TTCGTT": 1466, + "TTCGTC": 1467, + "TTCGTG": 1468, + "TTCGCA": 1469, + "TTCGCT": 1470, + "TTCGCC": 1471, + "TTCGCG": 1472, + "TTCGGA": 1473, + "TTCGGT": 1474, + "TTCGGC": 1475, + "TTCGGG": 1476, + "TTGAAA": 1477, + "TTGAAT": 1478, + "TTGAAC": 1479, + "TTGAAG": 1480, + "TTGATA": 1481, + "TTGATT": 1482, + "TTGATC": 1483, + "TTGATG": 1484, + "TTGACA": 1485, + "TTGACT": 1486, + "TTGACC": 1487, + "TTGACG": 1488, + "TTGAGA": 1489, + "TTGAGT": 1490, + "TTGAGC": 1491, + "TTGAGG": 1492, + "TTGTAA": 1493, + "TTGTAT": 1494, + "TTGTAC": 1495, + "TTGTAG": 1496, + "TTGTTA": 1497, + "TTGTTT": 1498, + "TTGTTC": 1499, + "TTGTTG": 1500, + "TTGTCA": 1501, + "TTGTCT": 1502, + "TTGTCC": 1503, + "TTGTCG": 1504, + "TTGTGA": 1505, + "TTGTGT": 1506, + "TTGTGC": 1507, + "TTGTGG": 1508, + "TTGCAA": 1509, + "TTGCAT": 1510, + "TTGCAC": 1511, + "TTGCAG": 1512, + "TTGCTA": 1513, + "TTGCTT": 1514, + "TTGCTC": 1515, + "TTGCTG": 1516, + "TTGCCA": 1517, + "TTGCCT": 1518, + "TTGCCC": 1519, + "TTGCCG": 1520, + "TTGCGA": 1521, + "TTGCGT": 1522, + "TTGCGC": 1523, + "TTGCGG": 1524, + "TTGGAA": 1525, + "TTGGAT": 1526, + "TTGGAC": 1527, + "TTGGAG": 1528, + "TTGGTA": 1529, + "TTGGTT": 1530, + "TTGGTC": 1531, + "TTGGTG": 1532, + "TTGGCA": 1533, + "TTGGCT": 1534, + "TTGGCC": 1535, + "TTGGCG": 1536, + "TTGGGA": 1537, + "TTGGGT": 1538, + "TTGGGC": 1539, + "TTGGGG": 1540, + "TCAAAA": 1541, + "TCAAAT": 1542, + "TCAAAC": 1543, + "TCAAAG": 1544, + "TCAATA": 1545, + "TCAATT": 1546, + "TCAATC": 1547, + "TCAATG": 1548, + "TCAACA": 1549, + "TCAACT": 1550, + "TCAACC": 1551, + "TCAACG": 1552, + "TCAAGA": 1553, + "TCAAGT": 1554, + "TCAAGC": 1555, + "TCAAGG": 1556, + "TCATAA": 1557, + "TCATAT": 1558, + "TCATAC": 1559, + "TCATAG": 1560, + "TCATTA": 1561, + "TCATTT": 1562, + "TCATTC": 1563, + "TCATTG": 1564, + "TCATCA": 1565, + "TCATCT": 1566, + "TCATCC": 1567, + "TCATCG": 1568, + "TCATGA": 1569, + "TCATGT": 1570, + "TCATGC": 1571, + "TCATGG": 1572, + "TCACAA": 1573, + "TCACAT": 1574, + "TCACAC": 1575, + "TCACAG": 1576, + "TCACTA": 1577, + "TCACTT": 1578, + "TCACTC": 1579, + "TCACTG": 1580, + "TCACCA": 1581, + "TCACCT": 1582, + "TCACCC": 1583, + "TCACCG": 1584, + "TCACGA": 1585, + "TCACGT": 1586, + "TCACGC": 1587, + "TCACGG": 1588, + "TCAGAA": 1589, + "TCAGAT": 1590, + "TCAGAC": 1591, + "TCAGAG": 1592, + "TCAGTA": 1593, + "TCAGTT": 1594, + "TCAGTC": 1595, + "TCAGTG": 1596, + "TCAGCA": 1597, + "TCAGCT": 1598, + "TCAGCC": 1599, + "TCAGCG": 1600, + "TCAGGA": 1601, + "TCAGGT": 1602, + "TCAGGC": 1603, + "TCAGGG": 1604, + "TCTAAA": 1605, + "TCTAAT": 1606, + "TCTAAC": 1607, + "TCTAAG": 1608, + "TCTATA": 1609, + "TCTATT": 1610, + "TCTATC": 1611, + "TCTATG": 1612, + "TCTACA": 1613, + "TCTACT": 1614, + "TCTACC": 1615, + "TCTACG": 1616, + "TCTAGA": 1617, + "TCTAGT": 1618, + "TCTAGC": 1619, + "TCTAGG": 1620, + "TCTTAA": 1621, + "TCTTAT": 1622, + "TCTTAC": 1623, + "TCTTAG": 1624, + "TCTTTA": 1625, + "TCTTTT": 1626, + "TCTTTC": 1627, + "TCTTTG": 1628, + "TCTTCA": 1629, + "TCTTCT": 1630, + "TCTTCC": 1631, + "TCTTCG": 1632, + "TCTTGA": 1633, + "TCTTGT": 1634, + "TCTTGC": 1635, + "TCTTGG": 1636, + "TCTCAA": 1637, + "TCTCAT": 1638, + "TCTCAC": 1639, + "TCTCAG": 1640, + "TCTCTA": 1641, + "TCTCTT": 1642, + "TCTCTC": 1643, + "TCTCTG": 1644, + "TCTCCA": 1645, + "TCTCCT": 1646, + "TCTCCC": 1647, + "TCTCCG": 1648, + "TCTCGA": 1649, + "TCTCGT": 1650, + "TCTCGC": 1651, + "TCTCGG": 1652, + "TCTGAA": 1653, + "TCTGAT": 1654, + "TCTGAC": 1655, + "TCTGAG": 1656, + "TCTGTA": 1657, + "TCTGTT": 1658, + "TCTGTC": 1659, + "TCTGTG": 1660, + "TCTGCA": 1661, + "TCTGCT": 1662, + "TCTGCC": 1663, + "TCTGCG": 1664, + "TCTGGA": 1665, + "TCTGGT": 1666, + "TCTGGC": 1667, + "TCTGGG": 1668, + "TCCAAA": 1669, + "TCCAAT": 1670, + "TCCAAC": 1671, + "TCCAAG": 1672, + "TCCATA": 1673, + "TCCATT": 1674, + "TCCATC": 1675, + "TCCATG": 1676, + "TCCACA": 1677, + "TCCACT": 1678, + "TCCACC": 1679, + "TCCACG": 1680, + "TCCAGA": 1681, + "TCCAGT": 1682, + "TCCAGC": 1683, + "TCCAGG": 1684, + "TCCTAA": 1685, + "TCCTAT": 1686, + "TCCTAC": 1687, + "TCCTAG": 1688, + "TCCTTA": 1689, + "TCCTTT": 1690, + "TCCTTC": 1691, + "TCCTTG": 1692, + "TCCTCA": 1693, + "TCCTCT": 1694, + "TCCTCC": 1695, + "TCCTCG": 1696, + "TCCTGA": 1697, + "TCCTGT": 1698, + "TCCTGC": 1699, + "TCCTGG": 1700, + "TCCCAA": 1701, + "TCCCAT": 1702, + "TCCCAC": 1703, + "TCCCAG": 1704, + "TCCCTA": 1705, + "TCCCTT": 1706, + "TCCCTC": 1707, + "TCCCTG": 1708, + "TCCCCA": 1709, + "TCCCCT": 1710, + "TCCCCC": 1711, + "TCCCCG": 1712, + "TCCCGA": 1713, + "TCCCGT": 1714, + "TCCCGC": 1715, + "TCCCGG": 1716, + "TCCGAA": 1717, + "TCCGAT": 1718, + "TCCGAC": 1719, + "TCCGAG": 1720, + "TCCGTA": 1721, + "TCCGTT": 1722, + "TCCGTC": 1723, + "TCCGTG": 1724, + "TCCGCA": 1725, + "TCCGCT": 1726, + "TCCGCC": 1727, + "TCCGCG": 1728, + "TCCGGA": 1729, + "TCCGGT": 1730, + "TCCGGC": 1731, + "TCCGGG": 1732, + "TCGAAA": 1733, + "TCGAAT": 1734, + "TCGAAC": 1735, + "TCGAAG": 1736, + "TCGATA": 1737, + "TCGATT": 1738, + "TCGATC": 1739, + "TCGATG": 1740, + "TCGACA": 1741, + "TCGACT": 1742, + "TCGACC": 1743, + "TCGACG": 1744, + "TCGAGA": 1745, + "TCGAGT": 1746, + "TCGAGC": 1747, + "TCGAGG": 1748, + "TCGTAA": 1749, + "TCGTAT": 1750, + "TCGTAC": 1751, + "TCGTAG": 1752, + "TCGTTA": 1753, + "TCGTTT": 1754, + "TCGTTC": 1755, + "TCGTTG": 1756, + "TCGTCA": 1757, + "TCGTCT": 1758, + "TCGTCC": 1759, + "TCGTCG": 1760, + "TCGTGA": 1761, + "TCGTGT": 1762, + "TCGTGC": 1763, + "TCGTGG": 1764, + "TCGCAA": 1765, + "TCGCAT": 1766, + "TCGCAC": 1767, + "TCGCAG": 1768, + "TCGCTA": 1769, + "TCGCTT": 1770, + "TCGCTC": 1771, + "TCGCTG": 1772, + "TCGCCA": 1773, + "TCGCCT": 1774, + "TCGCCC": 1775, + "TCGCCG": 1776, + "TCGCGA": 1777, + "TCGCGT": 1778, + "TCGCGC": 1779, + "TCGCGG": 1780, + "TCGGAA": 1781, + "TCGGAT": 1782, + "TCGGAC": 1783, + "TCGGAG": 1784, + "TCGGTA": 1785, + "TCGGTT": 1786, + "TCGGTC": 1787, + "TCGGTG": 1788, + "TCGGCA": 1789, + "TCGGCT": 1790, + "TCGGCC": 1791, + "TCGGCG": 1792, + "TCGGGA": 1793, + "TCGGGT": 1794, + "TCGGGC": 1795, + "TCGGGG": 1796, + "TGAAAA": 1797, + "TGAAAT": 1798, + "TGAAAC": 1799, + "TGAAAG": 1800, + "TGAATA": 1801, + "TGAATT": 1802, + "TGAATC": 1803, + "TGAATG": 1804, + "TGAACA": 1805, + "TGAACT": 1806, + "TGAACC": 1807, + "TGAACG": 1808, + "TGAAGA": 1809, + "TGAAGT": 1810, + "TGAAGC": 1811, + "TGAAGG": 1812, + "TGATAA": 1813, + "TGATAT": 1814, + "TGATAC": 1815, + "TGATAG": 1816, + "TGATTA": 1817, + "TGATTT": 1818, + "TGATTC": 1819, + "TGATTG": 1820, + "TGATCA": 1821, + "TGATCT": 1822, + "TGATCC": 1823, + "TGATCG": 1824, + "TGATGA": 1825, + "TGATGT": 1826, + "TGATGC": 1827, + "TGATGG": 1828, + "TGACAA": 1829, + "TGACAT": 1830, + "TGACAC": 1831, + "TGACAG": 1832, + "TGACTA": 1833, + "TGACTT": 1834, + "TGACTC": 1835, + "TGACTG": 1836, + "TGACCA": 1837, + "TGACCT": 1838, + "TGACCC": 1839, + "TGACCG": 1840, + "TGACGA": 1841, + "TGACGT": 1842, + "TGACGC": 1843, + "TGACGG": 1844, + "TGAGAA": 1845, + "TGAGAT": 1846, + "TGAGAC": 1847, + "TGAGAG": 1848, + "TGAGTA": 1849, + "TGAGTT": 1850, + "TGAGTC": 1851, + "TGAGTG": 1852, + "TGAGCA": 1853, + "TGAGCT": 1854, + "TGAGCC": 1855, + "TGAGCG": 1856, + "TGAGGA": 1857, + "TGAGGT": 1858, + "TGAGGC": 1859, + "TGAGGG": 1860, + "TGTAAA": 1861, + "TGTAAT": 1862, + "TGTAAC": 1863, + "TGTAAG": 1864, + "TGTATA": 1865, + "TGTATT": 1866, + "TGTATC": 1867, + "TGTATG": 1868, + "TGTACA": 1869, + "TGTACT": 1870, + "TGTACC": 1871, + "TGTACG": 1872, + "TGTAGA": 1873, + "TGTAGT": 1874, + "TGTAGC": 1875, + "TGTAGG": 1876, + "TGTTAA": 1877, + "TGTTAT": 1878, + "TGTTAC": 1879, + "TGTTAG": 1880, + "TGTTTA": 1881, + "TGTTTT": 1882, + "TGTTTC": 1883, + "TGTTTG": 1884, + "TGTTCA": 1885, + "TGTTCT": 1886, + "TGTTCC": 1887, + "TGTTCG": 1888, + "TGTTGA": 1889, + "TGTTGT": 1890, + "TGTTGC": 1891, + "TGTTGG": 1892, + "TGTCAA": 1893, + "TGTCAT": 1894, + "TGTCAC": 1895, + "TGTCAG": 1896, + "TGTCTA": 1897, + "TGTCTT": 1898, + "TGTCTC": 1899, + "TGTCTG": 1900, + "TGTCCA": 1901, + "TGTCCT": 1902, + "TGTCCC": 1903, + "TGTCCG": 1904, + "TGTCGA": 1905, + "TGTCGT": 1906, + "TGTCGC": 1907, + "TGTCGG": 1908, + "TGTGAA": 1909, + "TGTGAT": 1910, + "TGTGAC": 1911, + "TGTGAG": 1912, + "TGTGTA": 1913, + "TGTGTT": 1914, + "TGTGTC": 1915, + "TGTGTG": 1916, + "TGTGCA": 1917, + "TGTGCT": 1918, + "TGTGCC": 1919, + "TGTGCG": 1920, + "TGTGGA": 1921, + "TGTGGT": 1922, + "TGTGGC": 1923, + "TGTGGG": 1924, + "TGCAAA": 1925, + "TGCAAT": 1926, + "TGCAAC": 1927, + "TGCAAG": 1928, + "TGCATA": 1929, + "TGCATT": 1930, + "TGCATC": 1931, + "TGCATG": 1932, + "TGCACA": 1933, + "TGCACT": 1934, + "TGCACC": 1935, + "TGCACG": 1936, + "TGCAGA": 1937, + "TGCAGT": 1938, + "TGCAGC": 1939, + "TGCAGG": 1940, + "TGCTAA": 1941, + "TGCTAT": 1942, + "TGCTAC": 1943, + "TGCTAG": 1944, + "TGCTTA": 1945, + "TGCTTT": 1946, + "TGCTTC": 1947, + "TGCTTG": 1948, + "TGCTCA": 1949, + "TGCTCT": 1950, + "TGCTCC": 1951, + "TGCTCG": 1952, + "TGCTGA": 1953, + "TGCTGT": 1954, + "TGCTGC": 1955, + "TGCTGG": 1956, + "TGCCAA": 1957, + "TGCCAT": 1958, + "TGCCAC": 1959, + "TGCCAG": 1960, + "TGCCTA": 1961, + "TGCCTT": 1962, + "TGCCTC": 1963, + "TGCCTG": 1964, + "TGCCCA": 1965, + "TGCCCT": 1966, + "TGCCCC": 1967, + "TGCCCG": 1968, + "TGCCGA": 1969, + "TGCCGT": 1970, + "TGCCGC": 1971, + "TGCCGG": 1972, + "TGCGAA": 1973, + "TGCGAT": 1974, + "TGCGAC": 1975, + "TGCGAG": 1976, + "TGCGTA": 1977, + "TGCGTT": 1978, + "TGCGTC": 1979, + "TGCGTG": 1980, + "TGCGCA": 1981, + "TGCGCT": 1982, + "TGCGCC": 1983, + "TGCGCG": 1984, + "TGCGGA": 1985, + "TGCGGT": 1986, + "TGCGGC": 1987, + "TGCGGG": 1988, + "TGGAAA": 1989, + "TGGAAT": 1990, + "TGGAAC": 1991, + "TGGAAG": 1992, + "TGGATA": 1993, + "TGGATT": 1994, + "TGGATC": 1995, + "TGGATG": 1996, + "TGGACA": 1997, + "TGGACT": 1998, + "TGGACC": 1999, + "TGGACG": 2000, + "TGGAGA": 2001, + "TGGAGT": 2002, + "TGGAGC": 2003, + "TGGAGG": 2004, + "TGGTAA": 2005, + "TGGTAT": 2006, + "TGGTAC": 2007, + "TGGTAG": 2008, + "TGGTTA": 2009, + "TGGTTT": 2010, + "TGGTTC": 2011, + "TGGTTG": 2012, + "TGGTCA": 2013, + "TGGTCT": 2014, + "TGGTCC": 2015, + "TGGTCG": 2016, + "TGGTGA": 2017, + "TGGTGT": 2018, + "TGGTGC": 2019, + "TGGTGG": 2020, + "TGGCAA": 2021, + "TGGCAT": 2022, + "TGGCAC": 2023, + "TGGCAG": 2024, + "TGGCTA": 2025, + "TGGCTT": 2026, + "TGGCTC": 2027, + "TGGCTG": 2028, + "TGGCCA": 2029, + "TGGCCT": 2030, + "TGGCCC": 2031, + "TGGCCG": 2032, + "TGGCGA": 2033, + "TGGCGT": 2034, + "TGGCGC": 2035, + "TGGCGG": 2036, + "TGGGAA": 2037, + "TGGGAT": 2038, + "TGGGAC": 2039, + "TGGGAG": 2040, + "TGGGTA": 2041, + "TGGGTT": 2042, + "TGGGTC": 2043, + "TGGGTG": 2044, + "TGGGCA": 2045, + "TGGGCT": 2046, + "TGGGCC": 2047, + "TGGGCG": 2048, + "TGGGGA": 2049, + "TGGGGT": 2050, + "TGGGGC": 2051, + "TGGGGG": 2052, + "CAAAAA": 2053, + "CAAAAT": 2054, + "CAAAAC": 2055, + "CAAAAG": 2056, + "CAAATA": 2057, + "CAAATT": 2058, + "CAAATC": 2059, + "CAAATG": 2060, + "CAAACA": 2061, + "CAAACT": 2062, + "CAAACC": 2063, + "CAAACG": 2064, + "CAAAGA": 2065, + "CAAAGT": 2066, + "CAAAGC": 2067, + "CAAAGG": 2068, + "CAATAA": 2069, + "CAATAT": 2070, + "CAATAC": 2071, + "CAATAG": 2072, + "CAATTA": 2073, + "CAATTT": 2074, + "CAATTC": 2075, + "CAATTG": 2076, + "CAATCA": 2077, + "CAATCT": 2078, + "CAATCC": 2079, + "CAATCG": 2080, + "CAATGA": 2081, + "CAATGT": 2082, + "CAATGC": 2083, + "CAATGG": 2084, + "CAACAA": 2085, + "CAACAT": 2086, + "CAACAC": 2087, + "CAACAG": 2088, + "CAACTA": 2089, + "CAACTT": 2090, + "CAACTC": 2091, + "CAACTG": 2092, + "CAACCA": 2093, + "CAACCT": 2094, + "CAACCC": 2095, + "CAACCG": 2096, + "CAACGA": 2097, + "CAACGT": 2098, + "CAACGC": 2099, + "CAACGG": 2100, + "CAAGAA": 2101, + "CAAGAT": 2102, + "CAAGAC": 2103, + "CAAGAG": 2104, + "CAAGTA": 2105, + "CAAGTT": 2106, + "CAAGTC": 2107, + "CAAGTG": 2108, + "CAAGCA": 2109, + "CAAGCT": 2110, + "CAAGCC": 2111, + "CAAGCG": 2112, + "CAAGGA": 2113, + "CAAGGT": 2114, + "CAAGGC": 2115, + "CAAGGG": 2116, + "CATAAA": 2117, + "CATAAT": 2118, + "CATAAC": 2119, + "CATAAG": 2120, + "CATATA": 2121, + "CATATT": 2122, + "CATATC": 2123, + "CATATG": 2124, + "CATACA": 2125, + "CATACT": 2126, + "CATACC": 2127, + "CATACG": 2128, + "CATAGA": 2129, + "CATAGT": 2130, + "CATAGC": 2131, + "CATAGG": 2132, + "CATTAA": 2133, + "CATTAT": 2134, + "CATTAC": 2135, + "CATTAG": 2136, + "CATTTA": 2137, + "CATTTT": 2138, + "CATTTC": 2139, + "CATTTG": 2140, + "CATTCA": 2141, + "CATTCT": 2142, + "CATTCC": 2143, + "CATTCG": 2144, + "CATTGA": 2145, + "CATTGT": 2146, + "CATTGC": 2147, + "CATTGG": 2148, + "CATCAA": 2149, + "CATCAT": 2150, + "CATCAC": 2151, + "CATCAG": 2152, + "CATCTA": 2153, + "CATCTT": 2154, + "CATCTC": 2155, + "CATCTG": 2156, + "CATCCA": 2157, + "CATCCT": 2158, + "CATCCC": 2159, + "CATCCG": 2160, + "CATCGA": 2161, + "CATCGT": 2162, + "CATCGC": 2163, + "CATCGG": 2164, + "CATGAA": 2165, + "CATGAT": 2166, + "CATGAC": 2167, + "CATGAG": 2168, + "CATGTA": 2169, + "CATGTT": 2170, + "CATGTC": 2171, + "CATGTG": 2172, + "CATGCA": 2173, + "CATGCT": 2174, + "CATGCC": 2175, + "CATGCG": 2176, + "CATGGA": 2177, + "CATGGT": 2178, + "CATGGC": 2179, + "CATGGG": 2180, + "CACAAA": 2181, + "CACAAT": 2182, + "CACAAC": 2183, + "CACAAG": 2184, + "CACATA": 2185, + "CACATT": 2186, + "CACATC": 2187, + "CACATG": 2188, + "CACACA": 2189, + "CACACT": 2190, + "CACACC": 2191, + "CACACG": 2192, + "CACAGA": 2193, + "CACAGT": 2194, + "CACAGC": 2195, + "CACAGG": 2196, + "CACTAA": 2197, + "CACTAT": 2198, + "CACTAC": 2199, + "CACTAG": 2200, + "CACTTA": 2201, + "CACTTT": 2202, + "CACTTC": 2203, + "CACTTG": 2204, + "CACTCA": 2205, + "CACTCT": 2206, + "CACTCC": 2207, + "CACTCG": 2208, + "CACTGA": 2209, + "CACTGT": 2210, + "CACTGC": 2211, + "CACTGG": 2212, + "CACCAA": 2213, + "CACCAT": 2214, + "CACCAC": 2215, + "CACCAG": 2216, + "CACCTA": 2217, + "CACCTT": 2218, + "CACCTC": 2219, + "CACCTG": 2220, + "CACCCA": 2221, + "CACCCT": 2222, + "CACCCC": 2223, + "CACCCG": 2224, + "CACCGA": 2225, + "CACCGT": 2226, + "CACCGC": 2227, + "CACCGG": 2228, + "CACGAA": 2229, + "CACGAT": 2230, + "CACGAC": 2231, + "CACGAG": 2232, + "CACGTA": 2233, + "CACGTT": 2234, + "CACGTC": 2235, + "CACGTG": 2236, + "CACGCA": 2237, + "CACGCT": 2238, + "CACGCC": 2239, + "CACGCG": 2240, + "CACGGA": 2241, + "CACGGT": 2242, + "CACGGC": 2243, + "CACGGG": 2244, + "CAGAAA": 2245, + "CAGAAT": 2246, + "CAGAAC": 2247, + "CAGAAG": 2248, + "CAGATA": 2249, + "CAGATT": 2250, + "CAGATC": 2251, + "CAGATG": 2252, + "CAGACA": 2253, + "CAGACT": 2254, + "CAGACC": 2255, + "CAGACG": 2256, + "CAGAGA": 2257, + "CAGAGT": 2258, + "CAGAGC": 2259, + "CAGAGG": 2260, + "CAGTAA": 2261, + "CAGTAT": 2262, + "CAGTAC": 2263, + "CAGTAG": 2264, + "CAGTTA": 2265, + "CAGTTT": 2266, + "CAGTTC": 2267, + "CAGTTG": 2268, + "CAGTCA": 2269, + "CAGTCT": 2270, + "CAGTCC": 2271, + "CAGTCG": 2272, + "CAGTGA": 2273, + "CAGTGT": 2274, + "CAGTGC": 2275, + "CAGTGG": 2276, + "CAGCAA": 2277, + "CAGCAT": 2278, + "CAGCAC": 2279, + "CAGCAG": 2280, + "CAGCTA": 2281, + "CAGCTT": 2282, + "CAGCTC": 2283, + "CAGCTG": 2284, + "CAGCCA": 2285, + "CAGCCT": 2286, + "CAGCCC": 2287, + "CAGCCG": 2288, + "CAGCGA": 2289, + "CAGCGT": 2290, + "CAGCGC": 2291, + "CAGCGG": 2292, + "CAGGAA": 2293, + "CAGGAT": 2294, + "CAGGAC": 2295, + "CAGGAG": 2296, + "CAGGTA": 2297, + "CAGGTT": 2298, + "CAGGTC": 2299, + "CAGGTG": 2300, + "CAGGCA": 2301, + "CAGGCT": 2302, + "CAGGCC": 2303, + "CAGGCG": 2304, + "CAGGGA": 2305, + "CAGGGT": 2306, + "CAGGGC": 2307, + "CAGGGG": 2308, + "CTAAAA": 2309, + "CTAAAT": 2310, + "CTAAAC": 2311, + "CTAAAG": 2312, + "CTAATA": 2313, + "CTAATT": 2314, + "CTAATC": 2315, + "CTAATG": 2316, + "CTAACA": 2317, + "CTAACT": 2318, + "CTAACC": 2319, + "CTAACG": 2320, + "CTAAGA": 2321, + "CTAAGT": 2322, + "CTAAGC": 2323, + "CTAAGG": 2324, + "CTATAA": 2325, + "CTATAT": 2326, + "CTATAC": 2327, + "CTATAG": 2328, + "CTATTA": 2329, + "CTATTT": 2330, + "CTATTC": 2331, + "CTATTG": 2332, + "CTATCA": 2333, + "CTATCT": 2334, + "CTATCC": 2335, + "CTATCG": 2336, + "CTATGA": 2337, + "CTATGT": 2338, + "CTATGC": 2339, + "CTATGG": 2340, + "CTACAA": 2341, + "CTACAT": 2342, + "CTACAC": 2343, + "CTACAG": 2344, + "CTACTA": 2345, + "CTACTT": 2346, + "CTACTC": 2347, + "CTACTG": 2348, + "CTACCA": 2349, + "CTACCT": 2350, + "CTACCC": 2351, + "CTACCG": 2352, + "CTACGA": 2353, + "CTACGT": 2354, + "CTACGC": 2355, + "CTACGG": 2356, + "CTAGAA": 2357, + "CTAGAT": 2358, + "CTAGAC": 2359, + "CTAGAG": 2360, + "CTAGTA": 2361, + "CTAGTT": 2362, + "CTAGTC": 2363, + "CTAGTG": 2364, + "CTAGCA": 2365, + "CTAGCT": 2366, + "CTAGCC": 2367, + "CTAGCG": 2368, + "CTAGGA": 2369, + "CTAGGT": 2370, + "CTAGGC": 2371, + "CTAGGG": 2372, + "CTTAAA": 2373, + "CTTAAT": 2374, + "CTTAAC": 2375, + "CTTAAG": 2376, + "CTTATA": 2377, + "CTTATT": 2378, + "CTTATC": 2379, + "CTTATG": 2380, + "CTTACA": 2381, + "CTTACT": 2382, + "CTTACC": 2383, + "CTTACG": 2384, + "CTTAGA": 2385, + "CTTAGT": 2386, + "CTTAGC": 2387, + "CTTAGG": 2388, + "CTTTAA": 2389, + "CTTTAT": 2390, + "CTTTAC": 2391, + "CTTTAG": 2392, + "CTTTTA": 2393, + "CTTTTT": 2394, + "CTTTTC": 2395, + "CTTTTG": 2396, + "CTTTCA": 2397, + "CTTTCT": 2398, + "CTTTCC": 2399, + "CTTTCG": 2400, + "CTTTGA": 2401, + "CTTTGT": 2402, + "CTTTGC": 2403, + "CTTTGG": 2404, + "CTTCAA": 2405, + "CTTCAT": 2406, + "CTTCAC": 2407, + "CTTCAG": 2408, + "CTTCTA": 2409, + "CTTCTT": 2410, + "CTTCTC": 2411, + "CTTCTG": 2412, + "CTTCCA": 2413, + "CTTCCT": 2414, + "CTTCCC": 2415, + "CTTCCG": 2416, + "CTTCGA": 2417, + "CTTCGT": 2418, + "CTTCGC": 2419, + "CTTCGG": 2420, + "CTTGAA": 2421, + "CTTGAT": 2422, + "CTTGAC": 2423, + "CTTGAG": 2424, + "CTTGTA": 2425, + "CTTGTT": 2426, + "CTTGTC": 2427, + "CTTGTG": 2428, + "CTTGCA": 2429, + "CTTGCT": 2430, + "CTTGCC": 2431, + "CTTGCG": 2432, + "CTTGGA": 2433, + "CTTGGT": 2434, + "CTTGGC": 2435, + "CTTGGG": 2436, + "CTCAAA": 2437, + "CTCAAT": 2438, + "CTCAAC": 2439, + "CTCAAG": 2440, + "CTCATA": 2441, + "CTCATT": 2442, + "CTCATC": 2443, + "CTCATG": 2444, + "CTCACA": 2445, + "CTCACT": 2446, + "CTCACC": 2447, + "CTCACG": 2448, + "CTCAGA": 2449, + "CTCAGT": 2450, + "CTCAGC": 2451, + "CTCAGG": 2452, + "CTCTAA": 2453, + "CTCTAT": 2454, + "CTCTAC": 2455, + "CTCTAG": 2456, + "CTCTTA": 2457, + "CTCTTT": 2458, + "CTCTTC": 2459, + "CTCTTG": 2460, + "CTCTCA": 2461, + "CTCTCT": 2462, + "CTCTCC": 2463, + "CTCTCG": 2464, + "CTCTGA": 2465, + "CTCTGT": 2466, + "CTCTGC": 2467, + "CTCTGG": 2468, + "CTCCAA": 2469, + "CTCCAT": 2470, + "CTCCAC": 2471, + "CTCCAG": 2472, + "CTCCTA": 2473, + "CTCCTT": 2474, + "CTCCTC": 2475, + "CTCCTG": 2476, + "CTCCCA": 2477, + "CTCCCT": 2478, + "CTCCCC": 2479, + "CTCCCG": 2480, + "CTCCGA": 2481, + "CTCCGT": 2482, + "CTCCGC": 2483, + "CTCCGG": 2484, + "CTCGAA": 2485, + "CTCGAT": 2486, + "CTCGAC": 2487, + "CTCGAG": 2488, + "CTCGTA": 2489, + "CTCGTT": 2490, + "CTCGTC": 2491, + "CTCGTG": 2492, + "CTCGCA": 2493, + "CTCGCT": 2494, + "CTCGCC": 2495, + "CTCGCG": 2496, + "CTCGGA": 2497, + "CTCGGT": 2498, + "CTCGGC": 2499, + "CTCGGG": 2500, + "CTGAAA": 2501, + "CTGAAT": 2502, + "CTGAAC": 2503, + "CTGAAG": 2504, + "CTGATA": 2505, + "CTGATT": 2506, + "CTGATC": 2507, + "CTGATG": 2508, + "CTGACA": 2509, + "CTGACT": 2510, + "CTGACC": 2511, + "CTGACG": 2512, + "CTGAGA": 2513, + "CTGAGT": 2514, + "CTGAGC": 2515, + "CTGAGG": 2516, + "CTGTAA": 2517, + "CTGTAT": 2518, + "CTGTAC": 2519, + "CTGTAG": 2520, + "CTGTTA": 2521, + "CTGTTT": 2522, + "CTGTTC": 2523, + "CTGTTG": 2524, + "CTGTCA": 2525, + "CTGTCT": 2526, + "CTGTCC": 2527, + "CTGTCG": 2528, + "CTGTGA": 2529, + "CTGTGT": 2530, + "CTGTGC": 2531, + "CTGTGG": 2532, + "CTGCAA": 2533, + "CTGCAT": 2534, + "CTGCAC": 2535, + "CTGCAG": 2536, + "CTGCTA": 2537, + "CTGCTT": 2538, + "CTGCTC": 2539, + "CTGCTG": 2540, + "CTGCCA": 2541, + "CTGCCT": 2542, + "CTGCCC": 2543, + "CTGCCG": 2544, + "CTGCGA": 2545, + "CTGCGT": 2546, + "CTGCGC": 2547, + "CTGCGG": 2548, + "CTGGAA": 2549, + "CTGGAT": 2550, + "CTGGAC": 2551, + "CTGGAG": 2552, + "CTGGTA": 2553, + "CTGGTT": 2554, + "CTGGTC": 2555, + "CTGGTG": 2556, + "CTGGCA": 2557, + "CTGGCT": 2558, + "CTGGCC": 2559, + "CTGGCG": 2560, + "CTGGGA": 2561, + "CTGGGT": 2562, + "CTGGGC": 2563, + "CTGGGG": 2564, + "CCAAAA": 2565, + "CCAAAT": 2566, + "CCAAAC": 2567, + "CCAAAG": 2568, + "CCAATA": 2569, + "CCAATT": 2570, + "CCAATC": 2571, + "CCAATG": 2572, + "CCAACA": 2573, + "CCAACT": 2574, + "CCAACC": 2575, + "CCAACG": 2576, + "CCAAGA": 2577, + "CCAAGT": 2578, + "CCAAGC": 2579, + "CCAAGG": 2580, + "CCATAA": 2581, + "CCATAT": 2582, + "CCATAC": 2583, + "CCATAG": 2584, + "CCATTA": 2585, + "CCATTT": 2586, + "CCATTC": 2587, + "CCATTG": 2588, + "CCATCA": 2589, + "CCATCT": 2590, + "CCATCC": 2591, + "CCATCG": 2592, + "CCATGA": 2593, + "CCATGT": 2594, + "CCATGC": 2595, + "CCATGG": 2596, + "CCACAA": 2597, + "CCACAT": 2598, + "CCACAC": 2599, + "CCACAG": 2600, + "CCACTA": 2601, + "CCACTT": 2602, + "CCACTC": 2603, + "CCACTG": 2604, + "CCACCA": 2605, + "CCACCT": 2606, + "CCACCC": 2607, + "CCACCG": 2608, + "CCACGA": 2609, + "CCACGT": 2610, + "CCACGC": 2611, + "CCACGG": 2612, + "CCAGAA": 2613, + "CCAGAT": 2614, + "CCAGAC": 2615, + "CCAGAG": 2616, + "CCAGTA": 2617, + "CCAGTT": 2618, + "CCAGTC": 2619, + "CCAGTG": 2620, + "CCAGCA": 2621, + "CCAGCT": 2622, + "CCAGCC": 2623, + "CCAGCG": 2624, + "CCAGGA": 2625, + "CCAGGT": 2626, + "CCAGGC": 2627, + "CCAGGG": 2628, + "CCTAAA": 2629, + "CCTAAT": 2630, + "CCTAAC": 2631, + "CCTAAG": 2632, + "CCTATA": 2633, + "CCTATT": 2634, + "CCTATC": 2635, + "CCTATG": 2636, + "CCTACA": 2637, + "CCTACT": 2638, + "CCTACC": 2639, + "CCTACG": 2640, + "CCTAGA": 2641, + "CCTAGT": 2642, + "CCTAGC": 2643, + "CCTAGG": 2644, + "CCTTAA": 2645, + "CCTTAT": 2646, + "CCTTAC": 2647, + "CCTTAG": 2648, + "CCTTTA": 2649, + "CCTTTT": 2650, + "CCTTTC": 2651, + "CCTTTG": 2652, + "CCTTCA": 2653, + "CCTTCT": 2654, + "CCTTCC": 2655, + "CCTTCG": 2656, + "CCTTGA": 2657, + "CCTTGT": 2658, + "CCTTGC": 2659, + "CCTTGG": 2660, + "CCTCAA": 2661, + "CCTCAT": 2662, + "CCTCAC": 2663, + "CCTCAG": 2664, + "CCTCTA": 2665, + "CCTCTT": 2666, + "CCTCTC": 2667, + "CCTCTG": 2668, + "CCTCCA": 2669, + "CCTCCT": 2670, + "CCTCCC": 2671, + "CCTCCG": 2672, + "CCTCGA": 2673, + "CCTCGT": 2674, + "CCTCGC": 2675, + "CCTCGG": 2676, + "CCTGAA": 2677, + "CCTGAT": 2678, + "CCTGAC": 2679, + "CCTGAG": 2680, + "CCTGTA": 2681, + "CCTGTT": 2682, + "CCTGTC": 2683, + "CCTGTG": 2684, + "CCTGCA": 2685, + "CCTGCT": 2686, + "CCTGCC": 2687, + "CCTGCG": 2688, + "CCTGGA": 2689, + "CCTGGT": 2690, + "CCTGGC": 2691, + "CCTGGG": 2692, + "CCCAAA": 2693, + "CCCAAT": 2694, + "CCCAAC": 2695, + "CCCAAG": 2696, + "CCCATA": 2697, + "CCCATT": 2698, + "CCCATC": 2699, + "CCCATG": 2700, + "CCCACA": 2701, + "CCCACT": 2702, + "CCCACC": 2703, + "CCCACG": 2704, + "CCCAGA": 2705, + "CCCAGT": 2706, + "CCCAGC": 2707, + "CCCAGG": 2708, + "CCCTAA": 2709, + "CCCTAT": 2710, + "CCCTAC": 2711, + "CCCTAG": 2712, + "CCCTTA": 2713, + "CCCTTT": 2714, + "CCCTTC": 2715, + "CCCTTG": 2716, + "CCCTCA": 2717, + "CCCTCT": 2718, + "CCCTCC": 2719, + "CCCTCG": 2720, + "CCCTGA": 2721, + "CCCTGT": 2722, + "CCCTGC": 2723, + "CCCTGG": 2724, + "CCCCAA": 2725, + "CCCCAT": 2726, + "CCCCAC": 2727, + "CCCCAG": 2728, + "CCCCTA": 2729, + "CCCCTT": 2730, + "CCCCTC": 2731, + "CCCCTG": 2732, + "CCCCCA": 2733, + "CCCCCT": 2734, + "CCCCCC": 2735, + "CCCCCG": 2736, + "CCCCGA": 2737, + "CCCCGT": 2738, + "CCCCGC": 2739, + "CCCCGG": 2740, + "CCCGAA": 2741, + "CCCGAT": 2742, + "CCCGAC": 2743, + "CCCGAG": 2744, + "CCCGTA": 2745, + "CCCGTT": 2746, + "CCCGTC": 2747, + "CCCGTG": 2748, + "CCCGCA": 2749, + "CCCGCT": 2750, + "CCCGCC": 2751, + "CCCGCG": 2752, + "CCCGGA": 2753, + "CCCGGT": 2754, + "CCCGGC": 2755, + "CCCGGG": 2756, + "CCGAAA": 2757, + "CCGAAT": 2758, + "CCGAAC": 2759, + "CCGAAG": 2760, + "CCGATA": 2761, + "CCGATT": 2762, + "CCGATC": 2763, + "CCGATG": 2764, + "CCGACA": 2765, + "CCGACT": 2766, + "CCGACC": 2767, + "CCGACG": 2768, + "CCGAGA": 2769, + "CCGAGT": 2770, + "CCGAGC": 2771, + "CCGAGG": 2772, + "CCGTAA": 2773, + "CCGTAT": 2774, + "CCGTAC": 2775, + "CCGTAG": 2776, + "CCGTTA": 2777, + "CCGTTT": 2778, + "CCGTTC": 2779, + "CCGTTG": 2780, + "CCGTCA": 2781, + "CCGTCT": 2782, + "CCGTCC": 2783, + "CCGTCG": 2784, + "CCGTGA": 2785, + "CCGTGT": 2786, + "CCGTGC": 2787, + "CCGTGG": 2788, + "CCGCAA": 2789, + "CCGCAT": 2790, + "CCGCAC": 2791, + "CCGCAG": 2792, + "CCGCTA": 2793, + "CCGCTT": 2794, + "CCGCTC": 2795, + "CCGCTG": 2796, + "CCGCCA": 2797, + "CCGCCT": 2798, + "CCGCCC": 2799, + "CCGCCG": 2800, + "CCGCGA": 2801, + "CCGCGT": 2802, + "CCGCGC": 2803, + "CCGCGG": 2804, + "CCGGAA": 2805, + "CCGGAT": 2806, + "CCGGAC": 2807, + "CCGGAG": 2808, + "CCGGTA": 2809, + "CCGGTT": 2810, + "CCGGTC": 2811, + "CCGGTG": 2812, + "CCGGCA": 2813, + "CCGGCT": 2814, + "CCGGCC": 2815, + "CCGGCG": 2816, + "CCGGGA": 2817, + "CCGGGT": 2818, + "CCGGGC": 2819, + "CCGGGG": 2820, + "CGAAAA": 2821, + "CGAAAT": 2822, + "CGAAAC": 2823, + "CGAAAG": 2824, + "CGAATA": 2825, + "CGAATT": 2826, + "CGAATC": 2827, + "CGAATG": 2828, + "CGAACA": 2829, + "CGAACT": 2830, + "CGAACC": 2831, + "CGAACG": 2832, + "CGAAGA": 2833, + "CGAAGT": 2834, + "CGAAGC": 2835, + "CGAAGG": 2836, + "CGATAA": 2837, + "CGATAT": 2838, + "CGATAC": 2839, + "CGATAG": 2840, + "CGATTA": 2841, + "CGATTT": 2842, + "CGATTC": 2843, + "CGATTG": 2844, + "CGATCA": 2845, + "CGATCT": 2846, + "CGATCC": 2847, + "CGATCG": 2848, + "CGATGA": 2849, + "CGATGT": 2850, + "CGATGC": 2851, + "CGATGG": 2852, + "CGACAA": 2853, + "CGACAT": 2854, + "CGACAC": 2855, + "CGACAG": 2856, + "CGACTA": 2857, + "CGACTT": 2858, + "CGACTC": 2859, + "CGACTG": 2860, + "CGACCA": 2861, + "CGACCT": 2862, + "CGACCC": 2863, + "CGACCG": 2864, + "CGACGA": 2865, + "CGACGT": 2866, + "CGACGC": 2867, + "CGACGG": 2868, + "CGAGAA": 2869, + "CGAGAT": 2870, + "CGAGAC": 2871, + "CGAGAG": 2872, + "CGAGTA": 2873, + "CGAGTT": 2874, + "CGAGTC": 2875, + "CGAGTG": 2876, + "CGAGCA": 2877, + "CGAGCT": 2878, + "CGAGCC": 2879, + "CGAGCG": 2880, + "CGAGGA": 2881, + "CGAGGT": 2882, + "CGAGGC": 2883, + "CGAGGG": 2884, + "CGTAAA": 2885, + "CGTAAT": 2886, + "CGTAAC": 2887, + "CGTAAG": 2888, + "CGTATA": 2889, + "CGTATT": 2890, + "CGTATC": 2891, + "CGTATG": 2892, + "CGTACA": 2893, + "CGTACT": 2894, + "CGTACC": 2895, + "CGTACG": 2896, + "CGTAGA": 2897, + "CGTAGT": 2898, + "CGTAGC": 2899, + "CGTAGG": 2900, + "CGTTAA": 2901, + "CGTTAT": 2902, + "CGTTAC": 2903, + "CGTTAG": 2904, + "CGTTTA": 2905, + "CGTTTT": 2906, + "CGTTTC": 2907, + "CGTTTG": 2908, + "CGTTCA": 2909, + "CGTTCT": 2910, + "CGTTCC": 2911, + "CGTTCG": 2912, + "CGTTGA": 2913, + "CGTTGT": 2914, + "CGTTGC": 2915, + "CGTTGG": 2916, + "CGTCAA": 2917, + "CGTCAT": 2918, + "CGTCAC": 2919, + "CGTCAG": 2920, + "CGTCTA": 2921, + "CGTCTT": 2922, + "CGTCTC": 2923, + "CGTCTG": 2924, + "CGTCCA": 2925, + "CGTCCT": 2926, + "CGTCCC": 2927, + "CGTCCG": 2928, + "CGTCGA": 2929, + "CGTCGT": 2930, + "CGTCGC": 2931, + "CGTCGG": 2932, + "CGTGAA": 2933, + "CGTGAT": 2934, + "CGTGAC": 2935, + "CGTGAG": 2936, + "CGTGTA": 2937, + "CGTGTT": 2938, + "CGTGTC": 2939, + "CGTGTG": 2940, + "CGTGCA": 2941, + "CGTGCT": 2942, + "CGTGCC": 2943, + "CGTGCG": 2944, + "CGTGGA": 2945, + "CGTGGT": 2946, + "CGTGGC": 2947, + "CGTGGG": 2948, + "CGCAAA": 2949, + "CGCAAT": 2950, + "CGCAAC": 2951, + "CGCAAG": 2952, + "CGCATA": 2953, + "CGCATT": 2954, + "CGCATC": 2955, + "CGCATG": 2956, + "CGCACA": 2957, + "CGCACT": 2958, + "CGCACC": 2959, + "CGCACG": 2960, + "CGCAGA": 2961, + "CGCAGT": 2962, + "CGCAGC": 2963, + "CGCAGG": 2964, + "CGCTAA": 2965, + "CGCTAT": 2966, + "CGCTAC": 2967, + "CGCTAG": 2968, + "CGCTTA": 2969, + "CGCTTT": 2970, + "CGCTTC": 2971, + "CGCTTG": 2972, + "CGCTCA": 2973, + "CGCTCT": 2974, + "CGCTCC": 2975, + "CGCTCG": 2976, + "CGCTGA": 2977, + "CGCTGT": 2978, + "CGCTGC": 2979, + "CGCTGG": 2980, + "CGCCAA": 2981, + "CGCCAT": 2982, + "CGCCAC": 2983, + "CGCCAG": 2984, + "CGCCTA": 2985, + "CGCCTT": 2986, + "CGCCTC": 2987, + "CGCCTG": 2988, + "CGCCCA": 2989, + "CGCCCT": 2990, + "CGCCCC": 2991, + "CGCCCG": 2992, + "CGCCGA": 2993, + "CGCCGT": 2994, + "CGCCGC": 2995, + "CGCCGG": 2996, + "CGCGAA": 2997, + "CGCGAT": 2998, + "CGCGAC": 2999, + "CGCGAG": 3000, + "CGCGTA": 3001, + "CGCGTT": 3002, + "CGCGTC": 3003, + "CGCGTG": 3004, + "CGCGCA": 3005, + "CGCGCT": 3006, + "CGCGCC": 3007, + "CGCGCG": 3008, + "CGCGGA": 3009, + "CGCGGT": 3010, + "CGCGGC": 3011, + "CGCGGG": 3012, + "CGGAAA": 3013, + "CGGAAT": 3014, + "CGGAAC": 3015, + "CGGAAG": 3016, + "CGGATA": 3017, + "CGGATT": 3018, + "CGGATC": 3019, + "CGGATG": 3020, + "CGGACA": 3021, + "CGGACT": 3022, + "CGGACC": 3023, + "CGGACG": 3024, + "CGGAGA": 3025, + "CGGAGT": 3026, + "CGGAGC": 3027, + "CGGAGG": 3028, + "CGGTAA": 3029, + "CGGTAT": 3030, + "CGGTAC": 3031, + "CGGTAG": 3032, + "CGGTTA": 3033, + "CGGTTT": 3034, + "CGGTTC": 3035, + "CGGTTG": 3036, + "CGGTCA": 3037, + "CGGTCT": 3038, + "CGGTCC": 3039, + "CGGTCG": 3040, + "CGGTGA": 3041, + "CGGTGT": 3042, + "CGGTGC": 3043, + "CGGTGG": 3044, + "CGGCAA": 3045, + "CGGCAT": 3046, + "CGGCAC": 3047, + "CGGCAG": 3048, + "CGGCTA": 3049, + "CGGCTT": 3050, + "CGGCTC": 3051, + "CGGCTG": 3052, + "CGGCCA": 3053, + "CGGCCT": 3054, + "CGGCCC": 3055, + "CGGCCG": 3056, + "CGGCGA": 3057, + "CGGCGT": 3058, + "CGGCGC": 3059, + "CGGCGG": 3060, + "CGGGAA": 3061, + "CGGGAT": 3062, + "CGGGAC": 3063, + "CGGGAG": 3064, + "CGGGTA": 3065, + "CGGGTT": 3066, + "CGGGTC": 3067, + "CGGGTG": 3068, + "CGGGCA": 3069, + "CGGGCT": 3070, + "CGGGCC": 3071, + "CGGGCG": 3072, + "CGGGGA": 3073, + "CGGGGT": 3074, + "CGGGGC": 3075, + "CGGGGG": 3076, + "GAAAAA": 3077, + "GAAAAT": 3078, + "GAAAAC": 3079, + "GAAAAG": 3080, + "GAAATA": 3081, + "GAAATT": 3082, + "GAAATC": 3083, + "GAAATG": 3084, + "GAAACA": 3085, + "GAAACT": 3086, + "GAAACC": 3087, + "GAAACG": 3088, + "GAAAGA": 3089, + "GAAAGT": 3090, + "GAAAGC": 3091, + "GAAAGG": 3092, + "GAATAA": 3093, + "GAATAT": 3094, + "GAATAC": 3095, + "GAATAG": 3096, + "GAATTA": 3097, + "GAATTT": 3098, + "GAATTC": 3099, + "GAATTG": 3100, + "GAATCA": 3101, + "GAATCT": 3102, + "GAATCC": 3103, + "GAATCG": 3104, + "GAATGA": 3105, + "GAATGT": 3106, + "GAATGC": 3107, + "GAATGG": 3108, + "GAACAA": 3109, + "GAACAT": 3110, + "GAACAC": 3111, + "GAACAG": 3112, + "GAACTA": 3113, + "GAACTT": 3114, + "GAACTC": 3115, + "GAACTG": 3116, + "GAACCA": 3117, + "GAACCT": 3118, + "GAACCC": 3119, + "GAACCG": 3120, + "GAACGA": 3121, + "GAACGT": 3122, + "GAACGC": 3123, + "GAACGG": 3124, + "GAAGAA": 3125, + "GAAGAT": 3126, + "GAAGAC": 3127, + "GAAGAG": 3128, + "GAAGTA": 3129, + "GAAGTT": 3130, + "GAAGTC": 3131, + "GAAGTG": 3132, + "GAAGCA": 3133, + "GAAGCT": 3134, + "GAAGCC": 3135, + "GAAGCG": 3136, + "GAAGGA": 3137, + "GAAGGT": 3138, + "GAAGGC": 3139, + "GAAGGG": 3140, + "GATAAA": 3141, + "GATAAT": 3142, + "GATAAC": 3143, + "GATAAG": 3144, + "GATATA": 3145, + "GATATT": 3146, + "GATATC": 3147, + "GATATG": 3148, + "GATACA": 3149, + "GATACT": 3150, + "GATACC": 3151, + "GATACG": 3152, + "GATAGA": 3153, + "GATAGT": 3154, + "GATAGC": 3155, + "GATAGG": 3156, + "GATTAA": 3157, + "GATTAT": 3158, + "GATTAC": 3159, + "GATTAG": 3160, + "GATTTA": 3161, + "GATTTT": 3162, + "GATTTC": 3163, + "GATTTG": 3164, + "GATTCA": 3165, + "GATTCT": 3166, + "GATTCC": 3167, + "GATTCG": 3168, + "GATTGA": 3169, + "GATTGT": 3170, + "GATTGC": 3171, + "GATTGG": 3172, + "GATCAA": 3173, + "GATCAT": 3174, + "GATCAC": 3175, + "GATCAG": 3176, + "GATCTA": 3177, + "GATCTT": 3178, + "GATCTC": 3179, + "GATCTG": 3180, + "GATCCA": 3181, + "GATCCT": 3182, + "GATCCC": 3183, + "GATCCG": 3184, + "GATCGA": 3185, + "GATCGT": 3186, + "GATCGC": 3187, + "GATCGG": 3188, + "GATGAA": 3189, + "GATGAT": 3190, + "GATGAC": 3191, + "GATGAG": 3192, + "GATGTA": 3193, + "GATGTT": 3194, + "GATGTC": 3195, + "GATGTG": 3196, + "GATGCA": 3197, + "GATGCT": 3198, + "GATGCC": 3199, + "GATGCG": 3200, + "GATGGA": 3201, + "GATGGT": 3202, + "GATGGC": 3203, + "GATGGG": 3204, + "GACAAA": 3205, + "GACAAT": 3206, + "GACAAC": 3207, + "GACAAG": 3208, + "GACATA": 3209, + "GACATT": 3210, + "GACATC": 3211, + "GACATG": 3212, + "GACACA": 3213, + "GACACT": 3214, + "GACACC": 3215, + "GACACG": 3216, + "GACAGA": 3217, + "GACAGT": 3218, + "GACAGC": 3219, + "GACAGG": 3220, + "GACTAA": 3221, + "GACTAT": 3222, + "GACTAC": 3223, + "GACTAG": 3224, + "GACTTA": 3225, + "GACTTT": 3226, + "GACTTC": 3227, + "GACTTG": 3228, + "GACTCA": 3229, + "GACTCT": 3230, + "GACTCC": 3231, + "GACTCG": 3232, + "GACTGA": 3233, + "GACTGT": 3234, + "GACTGC": 3235, + "GACTGG": 3236, + "GACCAA": 3237, + "GACCAT": 3238, + "GACCAC": 3239, + "GACCAG": 3240, + "GACCTA": 3241, + "GACCTT": 3242, + "GACCTC": 3243, + "GACCTG": 3244, + "GACCCA": 3245, + "GACCCT": 3246, + "GACCCC": 3247, + "GACCCG": 3248, + "GACCGA": 3249, + "GACCGT": 3250, + "GACCGC": 3251, + "GACCGG": 3252, + "GACGAA": 3253, + "GACGAT": 3254, + "GACGAC": 3255, + "GACGAG": 3256, + "GACGTA": 3257, + "GACGTT": 3258, + "GACGTC": 3259, + "GACGTG": 3260, + "GACGCA": 3261, + "GACGCT": 3262, + "GACGCC": 3263, + "GACGCG": 3264, + "GACGGA": 3265, + "GACGGT": 3266, + "GACGGC": 3267, + "GACGGG": 3268, + "GAGAAA": 3269, + "GAGAAT": 3270, + "GAGAAC": 3271, + "GAGAAG": 3272, + "GAGATA": 3273, + "GAGATT": 3274, + "GAGATC": 3275, + "GAGATG": 3276, + "GAGACA": 3277, + "GAGACT": 3278, + "GAGACC": 3279, + "GAGACG": 3280, + "GAGAGA": 3281, + "GAGAGT": 3282, + "GAGAGC": 3283, + "GAGAGG": 3284, + "GAGTAA": 3285, + "GAGTAT": 3286, + "GAGTAC": 3287, + "GAGTAG": 3288, + "GAGTTA": 3289, + "GAGTTT": 3290, + "GAGTTC": 3291, + "GAGTTG": 3292, + "GAGTCA": 3293, + "GAGTCT": 3294, + "GAGTCC": 3295, + "GAGTCG": 3296, + "GAGTGA": 3297, + "GAGTGT": 3298, + "GAGTGC": 3299, + "GAGTGG": 3300, + "GAGCAA": 3301, + "GAGCAT": 3302, + "GAGCAC": 3303, + "GAGCAG": 3304, + "GAGCTA": 3305, + "GAGCTT": 3306, + "GAGCTC": 3307, + "GAGCTG": 3308, + "GAGCCA": 3309, + "GAGCCT": 3310, + "GAGCCC": 3311, + "GAGCCG": 3312, + "GAGCGA": 3313, + "GAGCGT": 3314, + "GAGCGC": 3315, + "GAGCGG": 3316, + "GAGGAA": 3317, + "GAGGAT": 3318, + "GAGGAC": 3319, + "GAGGAG": 3320, + "GAGGTA": 3321, + "GAGGTT": 3322, + "GAGGTC": 3323, + "GAGGTG": 3324, + "GAGGCA": 3325, + "GAGGCT": 3326, + "GAGGCC": 3327, + "GAGGCG": 3328, + "GAGGGA": 3329, + "GAGGGT": 3330, + "GAGGGC": 3331, + "GAGGGG": 3332, + "GTAAAA": 3333, + "GTAAAT": 3334, + "GTAAAC": 3335, + "GTAAAG": 3336, + "GTAATA": 3337, + "GTAATT": 3338, + "GTAATC": 3339, + "GTAATG": 3340, + "GTAACA": 3341, + "GTAACT": 3342, + "GTAACC": 3343, + "GTAACG": 3344, + "GTAAGA": 3345, + "GTAAGT": 3346, + "GTAAGC": 3347, + "GTAAGG": 3348, + "GTATAA": 3349, + "GTATAT": 3350, + "GTATAC": 3351, + "GTATAG": 3352, + "GTATTA": 3353, + "GTATTT": 3354, + "GTATTC": 3355, + "GTATTG": 3356, + "GTATCA": 3357, + "GTATCT": 3358, + "GTATCC": 3359, + "GTATCG": 3360, + "GTATGA": 3361, + "GTATGT": 3362, + "GTATGC": 3363, + "GTATGG": 3364, + "GTACAA": 3365, + "GTACAT": 3366, + "GTACAC": 3367, + "GTACAG": 3368, + "GTACTA": 3369, + "GTACTT": 3370, + "GTACTC": 3371, + "GTACTG": 3372, + "GTACCA": 3373, + "GTACCT": 3374, + "GTACCC": 3375, + "GTACCG": 3376, + "GTACGA": 3377, + "GTACGT": 3378, + "GTACGC": 3379, + "GTACGG": 3380, + "GTAGAA": 3381, + "GTAGAT": 3382, + "GTAGAC": 3383, + "GTAGAG": 3384, + "GTAGTA": 3385, + "GTAGTT": 3386, + "GTAGTC": 3387, + "GTAGTG": 3388, + "GTAGCA": 3389, + "GTAGCT": 3390, + "GTAGCC": 3391, + "GTAGCG": 3392, + "GTAGGA": 3393, + "GTAGGT": 3394, + "GTAGGC": 3395, + "GTAGGG": 3396, + "GTTAAA": 3397, + "GTTAAT": 3398, + "GTTAAC": 3399, + "GTTAAG": 3400, + "GTTATA": 3401, + "GTTATT": 3402, + "GTTATC": 3403, + "GTTATG": 3404, + "GTTACA": 3405, + "GTTACT": 3406, + "GTTACC": 3407, + "GTTACG": 3408, + "GTTAGA": 3409, + "GTTAGT": 3410, + "GTTAGC": 3411, + "GTTAGG": 3412, + "GTTTAA": 3413, + "GTTTAT": 3414, + "GTTTAC": 3415, + "GTTTAG": 3416, + "GTTTTA": 3417, + "GTTTTT": 3418, + "GTTTTC": 3419, + "GTTTTG": 3420, + "GTTTCA": 3421, + "GTTTCT": 3422, + "GTTTCC": 3423, + "GTTTCG": 3424, + "GTTTGA": 3425, + "GTTTGT": 3426, + "GTTTGC": 3427, + "GTTTGG": 3428, + "GTTCAA": 3429, + "GTTCAT": 3430, + "GTTCAC": 3431, + "GTTCAG": 3432, + "GTTCTA": 3433, + "GTTCTT": 3434, + "GTTCTC": 3435, + "GTTCTG": 3436, + "GTTCCA": 3437, + "GTTCCT": 3438, + "GTTCCC": 3439, + "GTTCCG": 3440, + "GTTCGA": 3441, + "GTTCGT": 3442, + "GTTCGC": 3443, + "GTTCGG": 3444, + "GTTGAA": 3445, + "GTTGAT": 3446, + "GTTGAC": 3447, + "GTTGAG": 3448, + "GTTGTA": 3449, + "GTTGTT": 3450, + "GTTGTC": 3451, + "GTTGTG": 3452, + "GTTGCA": 3453, + "GTTGCT": 3454, + "GTTGCC": 3455, + "GTTGCG": 3456, + "GTTGGA": 3457, + "GTTGGT": 3458, + "GTTGGC": 3459, + "GTTGGG": 3460, + "GTCAAA": 3461, + "GTCAAT": 3462, + "GTCAAC": 3463, + "GTCAAG": 3464, + "GTCATA": 3465, + "GTCATT": 3466, + "GTCATC": 3467, + "GTCATG": 3468, + "GTCACA": 3469, + "GTCACT": 3470, + "GTCACC": 3471, + "GTCACG": 3472, + "GTCAGA": 3473, + "GTCAGT": 3474, + "GTCAGC": 3475, + "GTCAGG": 3476, + "GTCTAA": 3477, + "GTCTAT": 3478, + "GTCTAC": 3479, + "GTCTAG": 3480, + "GTCTTA": 3481, + "GTCTTT": 3482, + "GTCTTC": 3483, + "GTCTTG": 3484, + "GTCTCA": 3485, + "GTCTCT": 3486, + "GTCTCC": 3487, + "GTCTCG": 3488, + "GTCTGA": 3489, + "GTCTGT": 3490, + "GTCTGC": 3491, + "GTCTGG": 3492, + "GTCCAA": 3493, + "GTCCAT": 3494, + "GTCCAC": 3495, + "GTCCAG": 3496, + "GTCCTA": 3497, + "GTCCTT": 3498, + "GTCCTC": 3499, + "GTCCTG": 3500, + "GTCCCA": 3501, + "GTCCCT": 3502, + "GTCCCC": 3503, + "GTCCCG": 3504, + "GTCCGA": 3505, + "GTCCGT": 3506, + "GTCCGC": 3507, + "GTCCGG": 3508, + "GTCGAA": 3509, + "GTCGAT": 3510, + "GTCGAC": 3511, + "GTCGAG": 3512, + "GTCGTA": 3513, + "GTCGTT": 3514, + "GTCGTC": 3515, + "GTCGTG": 3516, + "GTCGCA": 3517, + "GTCGCT": 3518, + "GTCGCC": 3519, + "GTCGCG": 3520, + "GTCGGA": 3521, + "GTCGGT": 3522, + "GTCGGC": 3523, + "GTCGGG": 3524, + "GTGAAA": 3525, + "GTGAAT": 3526, + "GTGAAC": 3527, + "GTGAAG": 3528, + "GTGATA": 3529, + "GTGATT": 3530, + "GTGATC": 3531, + "GTGATG": 3532, + "GTGACA": 3533, + "GTGACT": 3534, + "GTGACC": 3535, + "GTGACG": 3536, + "GTGAGA": 3537, + "GTGAGT": 3538, + "GTGAGC": 3539, + "GTGAGG": 3540, + "GTGTAA": 3541, + "GTGTAT": 3542, + "GTGTAC": 3543, + "GTGTAG": 3544, + "GTGTTA": 3545, + "GTGTTT": 3546, + "GTGTTC": 3547, + "GTGTTG": 3548, + "GTGTCA": 3549, + "GTGTCT": 3550, + "GTGTCC": 3551, + "GTGTCG": 3552, + "GTGTGA": 3553, + "GTGTGT": 3554, + "GTGTGC": 3555, + "GTGTGG": 3556, + "GTGCAA": 3557, + "GTGCAT": 3558, + "GTGCAC": 3559, + "GTGCAG": 3560, + "GTGCTA": 3561, + "GTGCTT": 3562, + "GTGCTC": 3563, + "GTGCTG": 3564, + "GTGCCA": 3565, + "GTGCCT": 3566, + "GTGCCC": 3567, + "GTGCCG": 3568, + "GTGCGA": 3569, + "GTGCGT": 3570, + "GTGCGC": 3571, + "GTGCGG": 3572, + "GTGGAA": 3573, + "GTGGAT": 3574, + "GTGGAC": 3575, + "GTGGAG": 3576, + "GTGGTA": 3577, + "GTGGTT": 3578, + "GTGGTC": 3579, + "GTGGTG": 3580, + "GTGGCA": 3581, + "GTGGCT": 3582, + "GTGGCC": 3583, + "GTGGCG": 3584, + "GTGGGA": 3585, + "GTGGGT": 3586, + "GTGGGC": 3587, + "GTGGGG": 3588, + "GCAAAA": 3589, + "GCAAAT": 3590, + "GCAAAC": 3591, + "GCAAAG": 3592, + "GCAATA": 3593, + "GCAATT": 3594, + "GCAATC": 3595, + "GCAATG": 3596, + "GCAACA": 3597, + "GCAACT": 3598, + "GCAACC": 3599, + "GCAACG": 3600, + "GCAAGA": 3601, + "GCAAGT": 3602, + "GCAAGC": 3603, + "GCAAGG": 3604, + "GCATAA": 3605, + "GCATAT": 3606, + "GCATAC": 3607, + "GCATAG": 3608, + "GCATTA": 3609, + "GCATTT": 3610, + "GCATTC": 3611, + "GCATTG": 3612, + "GCATCA": 3613, + "GCATCT": 3614, + "GCATCC": 3615, + "GCATCG": 3616, + "GCATGA": 3617, + "GCATGT": 3618, + "GCATGC": 3619, + "GCATGG": 3620, + "GCACAA": 3621, + "GCACAT": 3622, + "GCACAC": 3623, + "GCACAG": 3624, + "GCACTA": 3625, + "GCACTT": 3626, + "GCACTC": 3627, + "GCACTG": 3628, + "GCACCA": 3629, + "GCACCT": 3630, + "GCACCC": 3631, + "GCACCG": 3632, + "GCACGA": 3633, + "GCACGT": 3634, + "GCACGC": 3635, + "GCACGG": 3636, + "GCAGAA": 3637, + "GCAGAT": 3638, + "GCAGAC": 3639, + "GCAGAG": 3640, + "GCAGTA": 3641, + "GCAGTT": 3642, + "GCAGTC": 3643, + "GCAGTG": 3644, + "GCAGCA": 3645, + "GCAGCT": 3646, + "GCAGCC": 3647, + "GCAGCG": 3648, + "GCAGGA": 3649, + "GCAGGT": 3650, + "GCAGGC": 3651, + "GCAGGG": 3652, + "GCTAAA": 3653, + "GCTAAT": 3654, + "GCTAAC": 3655, + "GCTAAG": 3656, + "GCTATA": 3657, + "GCTATT": 3658, + "GCTATC": 3659, + "GCTATG": 3660, + "GCTACA": 3661, + "GCTACT": 3662, + "GCTACC": 3663, + "GCTACG": 3664, + "GCTAGA": 3665, + "GCTAGT": 3666, + "GCTAGC": 3667, + "GCTAGG": 3668, + "GCTTAA": 3669, + "GCTTAT": 3670, + "GCTTAC": 3671, + "GCTTAG": 3672, + "GCTTTA": 3673, + "GCTTTT": 3674, + "GCTTTC": 3675, + "GCTTTG": 3676, + "GCTTCA": 3677, + "GCTTCT": 3678, + "GCTTCC": 3679, + "GCTTCG": 3680, + "GCTTGA": 3681, + "GCTTGT": 3682, + "GCTTGC": 3683, + "GCTTGG": 3684, + "GCTCAA": 3685, + "GCTCAT": 3686, + "GCTCAC": 3687, + "GCTCAG": 3688, + "GCTCTA": 3689, + "GCTCTT": 3690, + "GCTCTC": 3691, + "GCTCTG": 3692, + "GCTCCA": 3693, + "GCTCCT": 3694, + "GCTCCC": 3695, + "GCTCCG": 3696, + "GCTCGA": 3697, + "GCTCGT": 3698, + "GCTCGC": 3699, + "GCTCGG": 3700, + "GCTGAA": 3701, + "GCTGAT": 3702, + "GCTGAC": 3703, + "GCTGAG": 3704, + "GCTGTA": 3705, + "GCTGTT": 3706, + "GCTGTC": 3707, + "GCTGTG": 3708, + "GCTGCA": 3709, + "GCTGCT": 3710, + "GCTGCC": 3711, + "GCTGCG": 3712, + "GCTGGA": 3713, + "GCTGGT": 3714, + "GCTGGC": 3715, + "GCTGGG": 3716, + "GCCAAA": 3717, + "GCCAAT": 3718, + "GCCAAC": 3719, + "GCCAAG": 3720, + "GCCATA": 3721, + "GCCATT": 3722, + "GCCATC": 3723, + "GCCATG": 3724, + "GCCACA": 3725, + "GCCACT": 3726, + "GCCACC": 3727, + "GCCACG": 3728, + "GCCAGA": 3729, + "GCCAGT": 3730, + "GCCAGC": 3731, + "GCCAGG": 3732, + "GCCTAA": 3733, + "GCCTAT": 3734, + "GCCTAC": 3735, + "GCCTAG": 3736, + "GCCTTA": 3737, + "GCCTTT": 3738, + "GCCTTC": 3739, + "GCCTTG": 3740, + "GCCTCA": 3741, + "GCCTCT": 3742, + "GCCTCC": 3743, + "GCCTCG": 3744, + "GCCTGA": 3745, + "GCCTGT": 3746, + "GCCTGC": 3747, + "GCCTGG": 3748, + "GCCCAA": 3749, + "GCCCAT": 3750, + "GCCCAC": 3751, + "GCCCAG": 3752, + "GCCCTA": 3753, + "GCCCTT": 3754, + "GCCCTC": 3755, + "GCCCTG": 3756, + "GCCCCA": 3757, + "GCCCCT": 3758, + "GCCCCC": 3759, + "GCCCCG": 3760, + "GCCCGA": 3761, + "GCCCGT": 3762, + "GCCCGC": 3763, + "GCCCGG": 3764, + "GCCGAA": 3765, + "GCCGAT": 3766, + "GCCGAC": 3767, + "GCCGAG": 3768, + "GCCGTA": 3769, + "GCCGTT": 3770, + "GCCGTC": 3771, + "GCCGTG": 3772, + "GCCGCA": 3773, + "GCCGCT": 3774, + "GCCGCC": 3775, + "GCCGCG": 3776, + "GCCGGA": 3777, + "GCCGGT": 3778, + "GCCGGC": 3779, + "GCCGGG": 3780, + "GCGAAA": 3781, + "GCGAAT": 3782, + "GCGAAC": 3783, + "GCGAAG": 3784, + "GCGATA": 3785, + "GCGATT": 3786, + "GCGATC": 3787, + "GCGATG": 3788, + "GCGACA": 3789, + "GCGACT": 3790, + "GCGACC": 3791, + "GCGACG": 3792, + "GCGAGA": 3793, + "GCGAGT": 3794, + "GCGAGC": 3795, + "GCGAGG": 3796, + "GCGTAA": 3797, + "GCGTAT": 3798, + "GCGTAC": 3799, + "GCGTAG": 3800, + "GCGTTA": 3801, + "GCGTTT": 3802, + "GCGTTC": 3803, + "GCGTTG": 3804, + "GCGTCA": 3805, + "GCGTCT": 3806, + "GCGTCC": 3807, + "GCGTCG": 3808, + "GCGTGA": 3809, + "GCGTGT": 3810, + "GCGTGC": 3811, + "GCGTGG": 3812, + "GCGCAA": 3813, + "GCGCAT": 3814, + "GCGCAC": 3815, + "GCGCAG": 3816, + "GCGCTA": 3817, + "GCGCTT": 3818, + "GCGCTC": 3819, + "GCGCTG": 3820, + "GCGCCA": 3821, + "GCGCCT": 3822, + "GCGCCC": 3823, + "GCGCCG": 3824, + "GCGCGA": 3825, + "GCGCGT": 3826, + "GCGCGC": 3827, + "GCGCGG": 3828, + "GCGGAA": 3829, + "GCGGAT": 3830, + "GCGGAC": 3831, + "GCGGAG": 3832, + "GCGGTA": 3833, + "GCGGTT": 3834, + "GCGGTC": 3835, + "GCGGTG": 3836, + "GCGGCA": 3837, + "GCGGCT": 3838, + "GCGGCC": 3839, + "GCGGCG": 3840, + "GCGGGA": 3841, + "GCGGGT": 3842, + "GCGGGC": 3843, + "GCGGGG": 3844, + "GGAAAA": 3845, + "GGAAAT": 3846, + "GGAAAC": 3847, + "GGAAAG": 3848, + "GGAATA": 3849, + "GGAATT": 3850, + "GGAATC": 3851, + "GGAATG": 3852, + "GGAACA": 3853, + "GGAACT": 3854, + "GGAACC": 3855, + "GGAACG": 3856, + "GGAAGA": 3857, + "GGAAGT": 3858, + "GGAAGC": 3859, + "GGAAGG": 3860, + "GGATAA": 3861, + "GGATAT": 3862, + "GGATAC": 3863, + "GGATAG": 3864, + "GGATTA": 3865, + "GGATTT": 3866, + "GGATTC": 3867, + "GGATTG": 3868, + "GGATCA": 3869, + "GGATCT": 3870, + "GGATCC": 3871, + "GGATCG": 3872, + "GGATGA": 3873, + "GGATGT": 3874, + "GGATGC": 3875, + "GGATGG": 3876, + "GGACAA": 3877, + "GGACAT": 3878, + "GGACAC": 3879, + "GGACAG": 3880, + "GGACTA": 3881, + "GGACTT": 3882, + "GGACTC": 3883, + "GGACTG": 3884, + "GGACCA": 3885, + "GGACCT": 3886, + "GGACCC": 3887, + "GGACCG": 3888, + "GGACGA": 3889, + "GGACGT": 3890, + "GGACGC": 3891, + "GGACGG": 3892, + "GGAGAA": 3893, + "GGAGAT": 3894, + "GGAGAC": 3895, + "GGAGAG": 3896, + "GGAGTA": 3897, + "GGAGTT": 3898, + "GGAGTC": 3899, + "GGAGTG": 3900, + "GGAGCA": 3901, + "GGAGCT": 3902, + "GGAGCC": 3903, + "GGAGCG": 3904, + "GGAGGA": 3905, + "GGAGGT": 3906, + "GGAGGC": 3907, + "GGAGGG": 3908, + "GGTAAA": 3909, + "GGTAAT": 3910, + "GGTAAC": 3911, + "GGTAAG": 3912, + "GGTATA": 3913, + "GGTATT": 3914, + "GGTATC": 3915, + "GGTATG": 3916, + "GGTACA": 3917, + "GGTACT": 3918, + "GGTACC": 3919, + "GGTACG": 3920, + "GGTAGA": 3921, + "GGTAGT": 3922, + "GGTAGC": 3923, + "GGTAGG": 3924, + "GGTTAA": 3925, + "GGTTAT": 3926, + "GGTTAC": 3927, + "GGTTAG": 3928, + "GGTTTA": 3929, + "GGTTTT": 3930, + "GGTTTC": 3931, + "GGTTTG": 3932, + "GGTTCA": 3933, + "GGTTCT": 3934, + "GGTTCC": 3935, + "GGTTCG": 3936, + "GGTTGA": 3937, + "GGTTGT": 3938, + "GGTTGC": 3939, + "GGTTGG": 3940, + "GGTCAA": 3941, + "GGTCAT": 3942, + "GGTCAC": 3943, + "GGTCAG": 3944, + "GGTCTA": 3945, + "GGTCTT": 3946, + "GGTCTC": 3947, + "GGTCTG": 3948, + "GGTCCA": 3949, + "GGTCCT": 3950, + "GGTCCC": 3951, + "GGTCCG": 3952, + "GGTCGA": 3953, + "GGTCGT": 3954, + "GGTCGC": 3955, + "GGTCGG": 3956, + "GGTGAA": 3957, + "GGTGAT": 3958, + "GGTGAC": 3959, + "GGTGAG": 3960, + "GGTGTA": 3961, + "GGTGTT": 3962, + "GGTGTC": 3963, + "GGTGTG": 3964, + "GGTGCA": 3965, + "GGTGCT": 3966, + "GGTGCC": 3967, + "GGTGCG": 3968, + "GGTGGA": 3969, + "GGTGGT": 3970, + "GGTGGC": 3971, + "GGTGGG": 3972, + "GGCAAA": 3973, + "GGCAAT": 3974, + "GGCAAC": 3975, + "GGCAAG": 3976, + "GGCATA": 3977, + "GGCATT": 3978, + "GGCATC": 3979, + "GGCATG": 3980, + "GGCACA": 3981, + "GGCACT": 3982, + "GGCACC": 3983, + "GGCACG": 3984, + "GGCAGA": 3985, + "GGCAGT": 3986, + "GGCAGC": 3987, + "GGCAGG": 3988, + "GGCTAA": 3989, + "GGCTAT": 3990, + "GGCTAC": 3991, + "GGCTAG": 3992, + "GGCTTA": 3993, + "GGCTTT": 3994, + "GGCTTC": 3995, + "GGCTTG": 3996, + "GGCTCA": 3997, + "GGCTCT": 3998, + "GGCTCC": 3999, + "GGCTCG": 4000, + "GGCTGA": 4001, + "GGCTGT": 4002, + "GGCTGC": 4003, + "GGCTGG": 4004, + "GGCCAA": 4005, + "GGCCAT": 4006, + "GGCCAC": 4007, + "GGCCAG": 4008, + "GGCCTA": 4009, + "GGCCTT": 4010, + "GGCCTC": 4011, + "GGCCTG": 4012, + "GGCCCA": 4013, + "GGCCCT": 4014, + "GGCCCC": 4015, + "GGCCCG": 4016, + "GGCCGA": 4017, + "GGCCGT": 4018, + "GGCCGC": 4019, + "GGCCGG": 4020, + "GGCGAA": 4021, + "GGCGAT": 4022, + "GGCGAC": 4023, + "GGCGAG": 4024, + "GGCGTA": 4025, + "GGCGTT": 4026, + "GGCGTC": 4027, + "GGCGTG": 4028, + "GGCGCA": 4029, + "GGCGCT": 4030, + "GGCGCC": 4031, + "GGCGCG": 4032, + "GGCGGA": 4033, + "GGCGGT": 4034, + "GGCGGC": 4035, + "GGCGGG": 4036, + "GGGAAA": 4037, + "GGGAAT": 4038, + "GGGAAC": 4039, + "GGGAAG": 4040, + "GGGATA": 4041, + "GGGATT": 4042, + "GGGATC": 4043, + "GGGATG": 4044, + "GGGACA": 4045, + "GGGACT": 4046, + "GGGACC": 4047, + "GGGACG": 4048, + "GGGAGA": 4049, + "GGGAGT": 4050, + "GGGAGC": 4051, + "GGGAGG": 4052, + "GGGTAA": 4053, + "GGGTAT": 4054, + "GGGTAC": 4055, + "GGGTAG": 4056, + "GGGTTA": 4057, + "GGGTTT": 4058, + "GGGTTC": 4059, + "GGGTTG": 4060, + "GGGTCA": 4061, + "GGGTCT": 4062, + "GGGTCC": 4063, + "GGGTCG": 4064, + "GGGTGA": 4065, + "GGGTGT": 4066, + "GGGTGC": 4067, + "GGGTGG": 4068, + "GGGCAA": 4069, + "GGGCAT": 4070, + "GGGCAC": 4071, + "GGGCAG": 4072, + "GGGCTA": 4073, + "GGGCTT": 4074, + "GGGCTC": 4075, + "GGGCTG": 4076, + "GGGCCA": 4077, + "GGGCCT": 4078, + "GGGCCC": 4079, + "GGGCCG": 4080, + "GGGCGA": 4081, + "GGGCGT": 4082, + "GGGCGC": 4083, + "GGGCGG": 4084, + "GGGGAA": 4085, + "GGGGAT": 4086, + "GGGGAC": 4087, + "GGGGAG": 4088, + "GGGGTA": 4089, + "GGGGTT": 4090, + "GGGGTC": 4091, + "GGGGTG": 4092, + "GGGGCA": 4093, + "GGGGCT": 4094, + "GGGGCC": 4095, + "GGGGCG": 4096, + "GGGGGA": 4097, + "GGGGGT": 4098, + "GGGGGC": 4099, + "GGGGGG": 4100 + } + } +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/tokenizer_config.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..604e2b1b29af1dcb3920d2828b6a90276e5e3d02 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/tokenizer_config.json @@ -0,0 +1,58 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "[UNK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "[CLS]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "[SEP]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "4": { + "content": "[MASK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "clean_up_tokenization_spaces": true, + "cls_token": "[CLS]", + "do_basic_tokenize": true, + "do_lower_case": false, + "mask_token": "[MASK]", + "max_len": 512, + "model_max_length": 512, + "never_split": null, + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "strip_accents": null, + "tokenize_chinese_chars": true, + "tokenizer_class": "BertTokenizer", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/trainer_state.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..12c63bd1e33b409e22fb9380055f5039c462aeca --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/trainer_state.json @@ -0,0 +1,873 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 12.0, + "eval_steps": 500, + "global_step": 3000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.1, + "grad_norm": 0.9416134357452393, + "learning_rate": 5.319148936170213e-05, + "loss": 0.6912, + "step": 25 + }, + { + "epoch": 0.2, + "grad_norm": 0.9283172488212585, + "learning_rate": 0.00010638297872340425, + "loss": 0.6691, + "step": 50 + }, + { + "epoch": 0.3, + "grad_norm": 0.5870445370674133, + "learning_rate": 0.00015957446808510637, + "loss": 0.6164, + "step": 75 + }, + { + "epoch": 0.4, + "grad_norm": 0.33945509791374207, + "learning_rate": 0.0002127659574468085, + "loss": 0.5453, + "step": 100 + }, + { + "epoch": 0.5, + "grad_norm": 0.21052460372447968, + "learning_rate": 0.00026595744680851064, + "loss": 0.5025, + "step": 125 + }, + { + "epoch": 0.6, + "grad_norm": 0.25948837399482727, + "learning_rate": 0.00031914893617021275, + "loss": 0.4881, + "step": 150 + }, + { + "epoch": 0.7, + "grad_norm": 0.6165992021560669, + "learning_rate": 0.0003723404255319149, + "loss": 0.4852, + "step": 175 + }, + { + "epoch": 0.8, + "grad_norm": 0.33296748995780945, + "learning_rate": 0.0003992081821181128, + "loss": 0.4867, + "step": 200 + }, + { + "epoch": 0.9, + "grad_norm": 0.40745800733566284, + "learning_rate": 0.0003975585615308479, + "loss": 0.4767, + "step": 225 + }, + { + "epoch": 1.0, + "grad_norm": 0.8070425391197205, + "learning_rate": 0.00039590894094358297, + "loss": 0.4784, + "step": 250 + }, + { + "epoch": 1.1, + "grad_norm": 1.2249579429626465, + "learning_rate": 0.0003942593203563181, + "loss": 0.4543, + "step": 275 + }, + { + "epoch": 1.2, + "grad_norm": 1.0143898725509644, + "learning_rate": 0.0003926096997690532, + "loss": 0.4319, + "step": 300 + }, + { + "epoch": 1.3, + "grad_norm": 2.3284952640533447, + "learning_rate": 0.00039096007918178817, + "loss": 0.4455, + "step": 325 + }, + { + "epoch": 1.4, + "grad_norm": 1.189352035522461, + "learning_rate": 0.0003893104585945233, + "loss": 0.4246, + "step": 350 + }, + { + "epoch": 1.5, + "grad_norm": 0.947853684425354, + "learning_rate": 0.0003876608380072583, + "loss": 0.4326, + "step": 375 + }, + { + "epoch": 1.6, + "grad_norm": 0.7130059003829956, + "learning_rate": 0.00038601121741999343, + "loss": 0.4034, + "step": 400 + }, + { + "epoch": 1.7, + "grad_norm": 3.517078161239624, + "learning_rate": 0.0003843615968327285, + "loss": 0.4114, + "step": 425 + }, + { + "epoch": 1.8, + "grad_norm": 0.3982729911804199, + "learning_rate": 0.0003827119762454636, + "loss": 0.4345, + "step": 450 + }, + { + "epoch": 1.9, + "grad_norm": 0.6208440065383911, + "learning_rate": 0.00038106235565819863, + "loss": 0.4135, + "step": 475 + }, + { + "epoch": 2.0, + "grad_norm": 0.9085566401481628, + "learning_rate": 0.0003794127350709337, + "loss": 0.3983, + "step": 500 + }, + { + "epoch": 2.1, + "grad_norm": 1.2225189208984375, + "learning_rate": 0.0003777631144836688, + "loss": 0.3915, + "step": 525 + }, + { + "epoch": 2.2, + "grad_norm": 1.1379530429840088, + "learning_rate": 0.00037611349389640383, + "loss": 0.4335, + "step": 550 + }, + { + "epoch": 2.3, + "grad_norm": 0.8535892367362976, + "learning_rate": 0.00037446387330913894, + "loss": 0.4063, + "step": 575 + }, + { + "epoch": 2.4, + "grad_norm": 1.3587419986724854, + "learning_rate": 0.000372814252721874, + "loss": 0.4065, + "step": 600 + }, + { + "epoch": 2.5, + "grad_norm": 0.9311383962631226, + "learning_rate": 0.00037116463213460903, + "loss": 0.4007, + "step": 625 + }, + { + "epoch": 2.6, + "grad_norm": 0.9912465810775757, + "learning_rate": 0.00036951501154734414, + "loss": 0.3992, + "step": 650 + }, + { + "epoch": 2.7, + "grad_norm": 1.2762119770050049, + "learning_rate": 0.0003678653909600792, + "loss": 0.3827, + "step": 675 + }, + { + "epoch": 2.8, + "grad_norm": 2.3506247997283936, + "learning_rate": 0.0003662157703728143, + "loss": 0.3734, + "step": 700 + }, + { + "epoch": 2.9, + "grad_norm": 1.5375425815582275, + "learning_rate": 0.00036456614978554934, + "loss": 0.4005, + "step": 725 + }, + { + "epoch": 3.0, + "grad_norm": 1.3143396377563477, + "learning_rate": 0.00036291652919828444, + "loss": 0.3548, + "step": 750 + }, + { + "epoch": 3.1, + "grad_norm": 0.662200391292572, + "learning_rate": 0.00036126690861101944, + "loss": 0.4129, + "step": 775 + }, + { + "epoch": 3.2, + "grad_norm": 1.3569878339767456, + "learning_rate": 0.00035961728802375454, + "loss": 0.38, + "step": 800 + }, + { + "epoch": 3.3, + "grad_norm": 0.9888588190078735, + "learning_rate": 0.0003579676674364896, + "loss": 0.3778, + "step": 825 + }, + { + "epoch": 3.4, + "grad_norm": 0.679969310760498, + "learning_rate": 0.0003563180468492247, + "loss": 0.3934, + "step": 850 + }, + { + "epoch": 3.5, + "grad_norm": 1.608539342880249, + "learning_rate": 0.0003546684262619598, + "loss": 0.418, + "step": 875 + }, + { + "epoch": 3.6, + "grad_norm": 1.7037910223007202, + "learning_rate": 0.00035301880567469485, + "loss": 0.3934, + "step": 900 + }, + { + "epoch": 3.7, + "grad_norm": 1.4417132139205933, + "learning_rate": 0.0003513691850874299, + "loss": 0.3947, + "step": 925 + }, + { + "epoch": 3.8, + "grad_norm": 1.1482582092285156, + "learning_rate": 0.00034971956450016495, + "loss": 0.3621, + "step": 950 + }, + { + "epoch": 3.9, + "grad_norm": 2.046616792678833, + "learning_rate": 0.00034806994391290005, + "loss": 0.3774, + "step": 975 + }, + { + "epoch": 4.0, + "grad_norm": 1.0832113027572632, + "learning_rate": 0.0003464203233256351, + "loss": 0.3542, + "step": 1000 + }, + { + "epoch": 4.1, + "grad_norm": 0.8189190626144409, + "learning_rate": 0.0003447707027383702, + "loss": 0.381, + "step": 1025 + }, + { + "epoch": 4.2, + "grad_norm": 1.6974774599075317, + "learning_rate": 0.00034312108215110525, + "loss": 0.3775, + "step": 1050 + }, + { + "epoch": 4.3, + "grad_norm": 1.2468841075897217, + "learning_rate": 0.0003414714615638403, + "loss": 0.3527, + "step": 1075 + }, + { + "epoch": 4.4, + "grad_norm": 0.9022793173789978, + "learning_rate": 0.0003398218409765754, + "loss": 0.3637, + "step": 1100 + }, + { + "epoch": 4.5, + "grad_norm": 1.121774673461914, + "learning_rate": 0.00033817222038931045, + "loss": 0.3935, + "step": 1125 + }, + { + "epoch": 4.6, + "grad_norm": 0.810937762260437, + "learning_rate": 0.00033652259980204556, + "loss": 0.4103, + "step": 1150 + }, + { + "epoch": 4.7, + "grad_norm": 0.7376565337181091, + "learning_rate": 0.0003348729792147806, + "loss": 0.362, + "step": 1175 + }, + { + "epoch": 4.8, + "grad_norm": 1.7271156311035156, + "learning_rate": 0.0003332233586275157, + "loss": 0.3597, + "step": 1200 + }, + { + "epoch": 4.9, + "grad_norm": 1.2273073196411133, + "learning_rate": 0.00033157373804025076, + "loss": 0.3853, + "step": 1225 + }, + { + "epoch": 5.0, + "grad_norm": 0.9998250603675842, + "learning_rate": 0.0003299241174529858, + "loss": 0.3461, + "step": 1250 + }, + { + "epoch": 5.1, + "grad_norm": 0.9112564325332642, + "learning_rate": 0.0003282744968657209, + "loss": 0.3659, + "step": 1275 + }, + { + "epoch": 5.2, + "grad_norm": 3.246051073074341, + "learning_rate": 0.00032662487627845596, + "loss": 0.3488, + "step": 1300 + }, + { + "epoch": 5.3, + "grad_norm": 1.0283201932907104, + "learning_rate": 0.00032497525569119106, + "loss": 0.3528, + "step": 1325 + }, + { + "epoch": 5.4, + "grad_norm": 1.112660527229309, + "learning_rate": 0.0003233256351039261, + "loss": 0.3723, + "step": 1350 + }, + { + "epoch": 5.5, + "grad_norm": 0.9069312810897827, + "learning_rate": 0.00032167601451666116, + "loss": 0.3416, + "step": 1375 + }, + { + "epoch": 5.6, + "grad_norm": 0.710412323474884, + "learning_rate": 0.00032002639392939627, + "loss": 0.3607, + "step": 1400 + }, + { + "epoch": 5.7, + "grad_norm": 1.185612678527832, + "learning_rate": 0.0003183767733421313, + "loss": 0.3777, + "step": 1425 + }, + { + "epoch": 5.8, + "grad_norm": 1.340174913406372, + "learning_rate": 0.0003167271527548664, + "loss": 0.3563, + "step": 1450 + }, + { + "epoch": 5.9, + "grad_norm": 0.9387012124061584, + "learning_rate": 0.00031507753216760147, + "loss": 0.3647, + "step": 1475 + }, + { + "epoch": 6.0, + "grad_norm": 1.1005347967147827, + "learning_rate": 0.00031342791158033657, + "loss": 0.3484, + "step": 1500 + }, + { + "epoch": 6.1, + "grad_norm": 2.64058780670166, + "learning_rate": 0.00031177829099307157, + "loss": 0.3542, + "step": 1525 + }, + { + "epoch": 6.2, + "grad_norm": 1.3007783889770508, + "learning_rate": 0.00031012867040580667, + "loss": 0.3379, + "step": 1550 + }, + { + "epoch": 6.3, + "grad_norm": 3.7190418243408203, + "learning_rate": 0.0003084790498185417, + "loss": 0.3514, + "step": 1575 + }, + { + "epoch": 6.4, + "grad_norm": 1.7263462543487549, + "learning_rate": 0.0003068294292312768, + "loss": 0.3434, + "step": 1600 + }, + { + "epoch": 6.5, + "grad_norm": 3.3360602855682373, + "learning_rate": 0.0003051798086440119, + "loss": 0.3507, + "step": 1625 + }, + { + "epoch": 6.6, + "grad_norm": 1.9620813131332397, + "learning_rate": 0.000303530188056747, + "loss": 0.3455, + "step": 1650 + }, + { + "epoch": 6.7, + "grad_norm": 1.3080660104751587, + "learning_rate": 0.000301880567469482, + "loss": 0.3678, + "step": 1675 + }, + { + "epoch": 6.8, + "grad_norm": 0.6183618307113647, + "learning_rate": 0.0003002309468822171, + "loss": 0.3704, + "step": 1700 + }, + { + "epoch": 6.9, + "grad_norm": 1.0411962270736694, + "learning_rate": 0.0002985813262949522, + "loss": 0.3505, + "step": 1725 + }, + { + "epoch": 7.0, + "grad_norm": 2.040569543838501, + "learning_rate": 0.00029693170570768723, + "loss": 0.3666, + "step": 1750 + }, + { + "epoch": 7.1, + "grad_norm": 0.713647723197937, + "learning_rate": 0.00029528208512042233, + "loss": 0.3529, + "step": 1775 + }, + { + "epoch": 7.2, + "grad_norm": 3.0373058319091797, + "learning_rate": 0.0002936324645331574, + "loss": 0.361, + "step": 1800 + }, + { + "epoch": 7.3, + "grad_norm": 1.819886326789856, + "learning_rate": 0.00029198284394589243, + "loss": 0.3482, + "step": 1825 + }, + { + "epoch": 7.4, + "grad_norm": 0.7930757403373718, + "learning_rate": 0.00029033322335862753, + "loss": 0.3695, + "step": 1850 + }, + { + "epoch": 7.5, + "grad_norm": 0.7889612913131714, + "learning_rate": 0.0002886836027713626, + "loss": 0.3527, + "step": 1875 + }, + { + "epoch": 7.6, + "grad_norm": 2.4007863998413086, + "learning_rate": 0.0002870339821840977, + "loss": 0.3275, + "step": 1900 + }, + { + "epoch": 7.7, + "grad_norm": 2.3483331203460693, + "learning_rate": 0.00028538436159683273, + "loss": 0.3571, + "step": 1925 + }, + { + "epoch": 7.8, + "grad_norm": 0.8911663293838501, + "learning_rate": 0.00028373474100956784, + "loss": 0.3079, + "step": 1950 + }, + { + "epoch": 7.9, + "grad_norm": 1.0521246194839478, + "learning_rate": 0.0002820851204223029, + "loss": 0.3578, + "step": 1975 + }, + { + "epoch": 8.0, + "grad_norm": 1.3923832178115845, + "learning_rate": 0.00028043549983503794, + "loss": 0.3588, + "step": 2000 + }, + { + "epoch": 8.1, + "grad_norm": 1.4007149934768677, + "learning_rate": 0.00027878587924777304, + "loss": 0.3458, + "step": 2025 + }, + { + "epoch": 8.2, + "grad_norm": 1.103198766708374, + "learning_rate": 0.0002771362586605081, + "loss": 0.3736, + "step": 2050 + }, + { + "epoch": 8.3, + "grad_norm": 2.936150550842285, + "learning_rate": 0.0002754866380732432, + "loss": 0.3223, + "step": 2075 + }, + { + "epoch": 8.4, + "grad_norm": 1.907762885093689, + "learning_rate": 0.00027383701748597824, + "loss": 0.3348, + "step": 2100 + }, + { + "epoch": 8.5, + "grad_norm": 1.4422886371612549, + "learning_rate": 0.0002721873968987133, + "loss": 0.3222, + "step": 2125 + }, + { + "epoch": 8.6, + "grad_norm": 1.8203483819961548, + "learning_rate": 0.00027053777631144834, + "loss": 0.3968, + "step": 2150 + }, + { + "epoch": 8.7, + "grad_norm": 0.8046119213104248, + "learning_rate": 0.00026888815572418344, + "loss": 0.3422, + "step": 2175 + }, + { + "epoch": 8.8, + "grad_norm": 1.2612605094909668, + "learning_rate": 0.00026723853513691855, + "loss": 0.3387, + "step": 2200 + }, + { + "epoch": 8.9, + "grad_norm": 1.6534866094589233, + "learning_rate": 0.0002655889145496536, + "loss": 0.3548, + "step": 2225 + }, + { + "epoch": 9.0, + "grad_norm": 1.889664649963379, + "learning_rate": 0.0002639392939623887, + "loss": 0.3367, + "step": 2250 + }, + { + "epoch": 9.1, + "grad_norm": 2.6691551208496094, + "learning_rate": 0.0002622896733751237, + "loss": 0.3628, + "step": 2275 + }, + { + "epoch": 9.2, + "grad_norm": 2.0130255222320557, + "learning_rate": 0.0002606400527878588, + "loss": 0.3366, + "step": 2300 + }, + { + "epoch": 9.3, + "grad_norm": 1.404814600944519, + "learning_rate": 0.00025899043220059385, + "loss": 0.3327, + "step": 2325 + }, + { + "epoch": 9.4, + "grad_norm": 0.9150362610816956, + "learning_rate": 0.00025734081161332895, + "loss": 0.308, + "step": 2350 + }, + { + "epoch": 9.5, + "grad_norm": 1.099320411682129, + "learning_rate": 0.000255691191026064, + "loss": 0.3499, + "step": 2375 + }, + { + "epoch": 9.6, + "grad_norm": 1.502451777458191, + "learning_rate": 0.0002540415704387991, + "loss": 0.3728, + "step": 2400 + }, + { + "epoch": 9.7, + "grad_norm": 1.1115409135818481, + "learning_rate": 0.00025239194985153415, + "loss": 0.348, + "step": 2425 + }, + { + "epoch": 9.8, + "grad_norm": 0.8831422328948975, + "learning_rate": 0.0002507423292642692, + "loss": 0.3392, + "step": 2450 + }, + { + "epoch": 9.9, + "grad_norm": 2.104572057723999, + "learning_rate": 0.0002490927086770043, + "loss": 0.3511, + "step": 2475 + }, + { + "epoch": 10.0, + "grad_norm": 2.4081919193267822, + "learning_rate": 0.00024744308808973936, + "loss": 0.3284, + "step": 2500 + }, + { + "epoch": 10.1, + "grad_norm": 1.9861253499984741, + "learning_rate": 0.00024579346750247446, + "loss": 0.3621, + "step": 2525 + }, + { + "epoch": 10.2, + "grad_norm": 1.9042344093322754, + "learning_rate": 0.0002441438469152095, + "loss": 0.3341, + "step": 2550 + }, + { + "epoch": 10.3, + "grad_norm": 0.5676223635673523, + "learning_rate": 0.00024249422632794456, + "loss": 0.3203, + "step": 2575 + }, + { + "epoch": 10.4, + "grad_norm": 1.5105270147323608, + "learning_rate": 0.00024084460574067963, + "loss": 0.3711, + "step": 2600 + }, + { + "epoch": 10.5, + "grad_norm": 0.9969163537025452, + "learning_rate": 0.0002391949851534147, + "loss": 0.3296, + "step": 2625 + }, + { + "epoch": 10.6, + "grad_norm": 1.297096848487854, + "learning_rate": 0.0002375453645661498, + "loss": 0.349, + "step": 2650 + }, + { + "epoch": 10.7, + "grad_norm": 1.0513092279434204, + "learning_rate": 0.00023589574397888486, + "loss": 0.3308, + "step": 2675 + }, + { + "epoch": 10.8, + "grad_norm": 1.5543969869613647, + "learning_rate": 0.00023424612339161997, + "loss": 0.3289, + "step": 2700 + }, + { + "epoch": 10.9, + "grad_norm": 0.9587257504463196, + "learning_rate": 0.000232596502804355, + "loss": 0.3128, + "step": 2725 + }, + { + "epoch": 11.0, + "grad_norm": 2.1693837642669678, + "learning_rate": 0.00023094688221709007, + "loss": 0.349, + "step": 2750 + }, + { + "epoch": 11.1, + "grad_norm": 1.5624831914901733, + "learning_rate": 0.00022929726162982514, + "loss": 0.3391, + "step": 2775 + }, + { + "epoch": 11.2, + "grad_norm": 1.4600939750671387, + "learning_rate": 0.00022764764104256022, + "loss": 0.3371, + "step": 2800 + }, + { + "epoch": 11.3, + "grad_norm": 1.6139200925827026, + "learning_rate": 0.0002259980204552953, + "loss": 0.3447, + "step": 2825 + }, + { + "epoch": 11.4, + "grad_norm": 2.2703235149383545, + "learning_rate": 0.00022434839986803037, + "loss": 0.3157, + "step": 2850 + }, + { + "epoch": 11.5, + "grad_norm": 2.0282037258148193, + "learning_rate": 0.00022269877928076542, + "loss": 0.3339, + "step": 2875 + }, + { + "epoch": 11.6, + "grad_norm": 1.389129638671875, + "learning_rate": 0.0002210491586935005, + "loss": 0.3195, + "step": 2900 + }, + { + "epoch": 11.7, + "grad_norm": 1.0041213035583496, + "learning_rate": 0.00021939953810623557, + "loss": 0.3382, + "step": 2925 + }, + { + "epoch": 11.8, + "grad_norm": 1.7116957902908325, + "learning_rate": 0.00021774991751897065, + "loss": 0.3261, + "step": 2950 + }, + { + "epoch": 11.9, + "grad_norm": 1.080194354057312, + "learning_rate": 0.00021610029693170573, + "loss": 0.3433, + "step": 2975 + }, + { + "epoch": 12.0, + "grad_norm": 0.6561412811279297, + "learning_rate": 0.0002144506763444408, + "loss": 0.3465, + "step": 3000 + } + ], + "logging_steps": 25, + "max_steps": 6250, + "num_input_tokens_seen": 0, + "num_train_epochs": 25, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 159348326400000.0, + "train_batch_size": 20, + "trial_name": null, + "trial_params": null +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/training_args.bin b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..d61f3500eb70763e1f33496c1787bdf3f07a651b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b6e0f7aa622786274a1ea4195b72f4da224050767f7b89529b85721865c5075a +size 5304 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/vocab.txt b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/vocab.txt new file mode 100644 index 0000000000000000000000000000000000000000..cc9ad53732e21ea9180f18e1fb67a38cc5702d4c --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/vocab.txt @@ -0,0 +1,4101 @@ +[PAD] +[UNK] +[CLS] +[SEP] +[MASK] +AAAAAA +AAAAAT +AAAAAC +AAAAAG +AAAATA +AAAATT +AAAATC +AAAATG +AAAACA +AAAACT +AAAACC +AAAACG +AAAAGA +AAAAGT +AAAAGC +AAAAGG +AAATAA +AAATAT +AAATAC +AAATAG +AAATTA +AAATTT +AAATTC +AAATTG +AAATCA +AAATCT +AAATCC +AAATCG +AAATGA +AAATGT +AAATGC +AAATGG +AAACAA +AAACAT +AAACAC +AAACAG +AAACTA +AAACTT +AAACTC +AAACTG +AAACCA +AAACCT +AAACCC +AAACCG +AAACGA +AAACGT +AAACGC +AAACGG +AAAGAA +AAAGAT +AAAGAC +AAAGAG +AAAGTA +AAAGTT +AAAGTC +AAAGTG +AAAGCA +AAAGCT +AAAGCC +AAAGCG +AAAGGA +AAAGGT +AAAGGC +AAAGGG +AATAAA +AATAAT +AATAAC +AATAAG +AATATA +AATATT +AATATC +AATATG +AATACA +AATACT +AATACC +AATACG +AATAGA +AATAGT +AATAGC +AATAGG +AATTAA +AATTAT +AATTAC +AATTAG +AATTTA +AATTTT +AATTTC +AATTTG +AATTCA +AATTCT +AATTCC +AATTCG +AATTGA +AATTGT +AATTGC +AATTGG +AATCAA +AATCAT +AATCAC +AATCAG +AATCTA +AATCTT +AATCTC +AATCTG +AATCCA +AATCCT +AATCCC +AATCCG +AATCGA +AATCGT +AATCGC +AATCGG +AATGAA +AATGAT +AATGAC +AATGAG +AATGTA +AATGTT +AATGTC +AATGTG +AATGCA +AATGCT +AATGCC +AATGCG +AATGGA +AATGGT +AATGGC +AATGGG +AACAAA +AACAAT +AACAAC +AACAAG +AACATA +AACATT +AACATC +AACATG +AACACA +AACACT +AACACC +AACACG +AACAGA +AACAGT +AACAGC +AACAGG +AACTAA +AACTAT +AACTAC +AACTAG +AACTTA +AACTTT +AACTTC +AACTTG +AACTCA +AACTCT +AACTCC +AACTCG +AACTGA +AACTGT +AACTGC +AACTGG +AACCAA +AACCAT +AACCAC +AACCAG +AACCTA +AACCTT +AACCTC +AACCTG +AACCCA +AACCCT +AACCCC +AACCCG +AACCGA +AACCGT +AACCGC +AACCGG +AACGAA +AACGAT +AACGAC +AACGAG +AACGTA +AACGTT +AACGTC +AACGTG +AACGCA +AACGCT +AACGCC +AACGCG +AACGGA +AACGGT +AACGGC +AACGGG +AAGAAA +AAGAAT +AAGAAC +AAGAAG +AAGATA +AAGATT +AAGATC +AAGATG +AAGACA +AAGACT +AAGACC +AAGACG +AAGAGA +AAGAGT +AAGAGC +AAGAGG +AAGTAA +AAGTAT +AAGTAC +AAGTAG +AAGTTA +AAGTTT +AAGTTC +AAGTTG +AAGTCA +AAGTCT +AAGTCC +AAGTCG +AAGTGA +AAGTGT +AAGTGC +AAGTGG +AAGCAA +AAGCAT +AAGCAC +AAGCAG +AAGCTA +AAGCTT +AAGCTC +AAGCTG +AAGCCA +AAGCCT +AAGCCC +AAGCCG +AAGCGA +AAGCGT +AAGCGC +AAGCGG +AAGGAA +AAGGAT +AAGGAC +AAGGAG +AAGGTA +AAGGTT +AAGGTC +AAGGTG +AAGGCA +AAGGCT +AAGGCC +AAGGCG +AAGGGA +AAGGGT +AAGGGC +AAGGGG +ATAAAA +ATAAAT +ATAAAC +ATAAAG +ATAATA +ATAATT +ATAATC +ATAATG +ATAACA +ATAACT +ATAACC +ATAACG +ATAAGA +ATAAGT +ATAAGC +ATAAGG +ATATAA +ATATAT +ATATAC +ATATAG +ATATTA +ATATTT +ATATTC +ATATTG +ATATCA +ATATCT +ATATCC +ATATCG +ATATGA +ATATGT +ATATGC +ATATGG +ATACAA +ATACAT +ATACAC +ATACAG +ATACTA +ATACTT +ATACTC +ATACTG +ATACCA +ATACCT +ATACCC +ATACCG +ATACGA +ATACGT +ATACGC +ATACGG +ATAGAA +ATAGAT +ATAGAC +ATAGAG +ATAGTA +ATAGTT +ATAGTC +ATAGTG +ATAGCA +ATAGCT +ATAGCC +ATAGCG +ATAGGA +ATAGGT +ATAGGC +ATAGGG +ATTAAA +ATTAAT +ATTAAC +ATTAAG +ATTATA +ATTATT +ATTATC +ATTATG +ATTACA +ATTACT +ATTACC +ATTACG +ATTAGA +ATTAGT +ATTAGC +ATTAGG +ATTTAA +ATTTAT +ATTTAC +ATTTAG +ATTTTA +ATTTTT +ATTTTC +ATTTTG +ATTTCA +ATTTCT +ATTTCC +ATTTCG +ATTTGA +ATTTGT +ATTTGC +ATTTGG +ATTCAA +ATTCAT +ATTCAC +ATTCAG +ATTCTA +ATTCTT +ATTCTC +ATTCTG +ATTCCA +ATTCCT +ATTCCC +ATTCCG +ATTCGA +ATTCGT +ATTCGC +ATTCGG +ATTGAA +ATTGAT +ATTGAC +ATTGAG +ATTGTA +ATTGTT +ATTGTC +ATTGTG +ATTGCA +ATTGCT +ATTGCC +ATTGCG +ATTGGA +ATTGGT +ATTGGC +ATTGGG +ATCAAA +ATCAAT +ATCAAC +ATCAAG +ATCATA +ATCATT +ATCATC +ATCATG +ATCACA +ATCACT +ATCACC +ATCACG +ATCAGA +ATCAGT +ATCAGC +ATCAGG +ATCTAA +ATCTAT +ATCTAC +ATCTAG +ATCTTA +ATCTTT +ATCTTC +ATCTTG +ATCTCA +ATCTCT +ATCTCC +ATCTCG +ATCTGA +ATCTGT +ATCTGC +ATCTGG +ATCCAA +ATCCAT +ATCCAC +ATCCAG +ATCCTA +ATCCTT +ATCCTC +ATCCTG +ATCCCA +ATCCCT +ATCCCC +ATCCCG +ATCCGA +ATCCGT +ATCCGC +ATCCGG +ATCGAA +ATCGAT +ATCGAC +ATCGAG +ATCGTA +ATCGTT +ATCGTC +ATCGTG +ATCGCA +ATCGCT +ATCGCC +ATCGCG +ATCGGA +ATCGGT +ATCGGC +ATCGGG +ATGAAA +ATGAAT +ATGAAC +ATGAAG +ATGATA +ATGATT +ATGATC +ATGATG +ATGACA +ATGACT +ATGACC +ATGACG +ATGAGA +ATGAGT +ATGAGC +ATGAGG +ATGTAA +ATGTAT +ATGTAC +ATGTAG +ATGTTA +ATGTTT +ATGTTC +ATGTTG +ATGTCA +ATGTCT +ATGTCC +ATGTCG +ATGTGA +ATGTGT +ATGTGC +ATGTGG +ATGCAA +ATGCAT +ATGCAC +ATGCAG +ATGCTA +ATGCTT +ATGCTC +ATGCTG +ATGCCA +ATGCCT +ATGCCC +ATGCCG +ATGCGA +ATGCGT +ATGCGC +ATGCGG +ATGGAA +ATGGAT +ATGGAC +ATGGAG +ATGGTA +ATGGTT +ATGGTC +ATGGTG +ATGGCA +ATGGCT +ATGGCC +ATGGCG +ATGGGA +ATGGGT +ATGGGC +ATGGGG +ACAAAA +ACAAAT +ACAAAC +ACAAAG +ACAATA +ACAATT +ACAATC +ACAATG +ACAACA +ACAACT +ACAACC +ACAACG +ACAAGA +ACAAGT +ACAAGC +ACAAGG +ACATAA +ACATAT +ACATAC +ACATAG +ACATTA +ACATTT +ACATTC +ACATTG +ACATCA +ACATCT +ACATCC +ACATCG +ACATGA +ACATGT +ACATGC +ACATGG +ACACAA +ACACAT +ACACAC +ACACAG +ACACTA +ACACTT +ACACTC +ACACTG +ACACCA +ACACCT +ACACCC +ACACCG +ACACGA +ACACGT +ACACGC +ACACGG +ACAGAA +ACAGAT +ACAGAC +ACAGAG +ACAGTA +ACAGTT +ACAGTC +ACAGTG +ACAGCA +ACAGCT +ACAGCC +ACAGCG +ACAGGA +ACAGGT +ACAGGC +ACAGGG +ACTAAA +ACTAAT +ACTAAC +ACTAAG +ACTATA +ACTATT +ACTATC +ACTATG +ACTACA +ACTACT +ACTACC +ACTACG +ACTAGA +ACTAGT +ACTAGC +ACTAGG +ACTTAA +ACTTAT +ACTTAC +ACTTAG +ACTTTA +ACTTTT +ACTTTC +ACTTTG +ACTTCA +ACTTCT +ACTTCC +ACTTCG +ACTTGA +ACTTGT +ACTTGC +ACTTGG +ACTCAA +ACTCAT +ACTCAC +ACTCAG +ACTCTA +ACTCTT +ACTCTC +ACTCTG +ACTCCA +ACTCCT +ACTCCC +ACTCCG +ACTCGA +ACTCGT +ACTCGC +ACTCGG +ACTGAA +ACTGAT +ACTGAC +ACTGAG +ACTGTA +ACTGTT +ACTGTC +ACTGTG +ACTGCA +ACTGCT +ACTGCC +ACTGCG +ACTGGA +ACTGGT +ACTGGC +ACTGGG +ACCAAA +ACCAAT +ACCAAC +ACCAAG +ACCATA +ACCATT +ACCATC +ACCATG +ACCACA +ACCACT +ACCACC +ACCACG +ACCAGA +ACCAGT +ACCAGC +ACCAGG +ACCTAA +ACCTAT +ACCTAC +ACCTAG +ACCTTA +ACCTTT +ACCTTC +ACCTTG +ACCTCA +ACCTCT +ACCTCC +ACCTCG +ACCTGA +ACCTGT +ACCTGC +ACCTGG +ACCCAA +ACCCAT +ACCCAC +ACCCAG +ACCCTA +ACCCTT +ACCCTC +ACCCTG +ACCCCA +ACCCCT +ACCCCC +ACCCCG +ACCCGA +ACCCGT +ACCCGC +ACCCGG +ACCGAA +ACCGAT +ACCGAC +ACCGAG +ACCGTA +ACCGTT +ACCGTC +ACCGTG +ACCGCA +ACCGCT +ACCGCC +ACCGCG +ACCGGA +ACCGGT +ACCGGC +ACCGGG +ACGAAA +ACGAAT +ACGAAC +ACGAAG +ACGATA +ACGATT +ACGATC +ACGATG +ACGACA +ACGACT +ACGACC +ACGACG +ACGAGA +ACGAGT +ACGAGC +ACGAGG +ACGTAA +ACGTAT +ACGTAC +ACGTAG +ACGTTA +ACGTTT +ACGTTC +ACGTTG +ACGTCA +ACGTCT +ACGTCC +ACGTCG +ACGTGA +ACGTGT +ACGTGC +ACGTGG +ACGCAA +ACGCAT +ACGCAC +ACGCAG +ACGCTA +ACGCTT +ACGCTC +ACGCTG +ACGCCA +ACGCCT +ACGCCC +ACGCCG +ACGCGA +ACGCGT +ACGCGC +ACGCGG +ACGGAA +ACGGAT +ACGGAC +ACGGAG +ACGGTA +ACGGTT +ACGGTC +ACGGTG +ACGGCA +ACGGCT +ACGGCC +ACGGCG +ACGGGA +ACGGGT +ACGGGC +ACGGGG +AGAAAA +AGAAAT +AGAAAC +AGAAAG +AGAATA +AGAATT +AGAATC +AGAATG +AGAACA +AGAACT +AGAACC +AGAACG +AGAAGA +AGAAGT +AGAAGC +AGAAGG +AGATAA +AGATAT +AGATAC +AGATAG +AGATTA +AGATTT +AGATTC +AGATTG +AGATCA +AGATCT +AGATCC +AGATCG +AGATGA +AGATGT +AGATGC +AGATGG +AGACAA +AGACAT +AGACAC +AGACAG +AGACTA +AGACTT +AGACTC +AGACTG +AGACCA +AGACCT +AGACCC +AGACCG +AGACGA +AGACGT +AGACGC +AGACGG +AGAGAA +AGAGAT +AGAGAC +AGAGAG +AGAGTA +AGAGTT +AGAGTC +AGAGTG +AGAGCA +AGAGCT +AGAGCC +AGAGCG +AGAGGA +AGAGGT +AGAGGC +AGAGGG +AGTAAA +AGTAAT +AGTAAC +AGTAAG +AGTATA +AGTATT +AGTATC +AGTATG +AGTACA +AGTACT +AGTACC +AGTACG +AGTAGA +AGTAGT +AGTAGC +AGTAGG +AGTTAA +AGTTAT +AGTTAC +AGTTAG +AGTTTA +AGTTTT +AGTTTC +AGTTTG +AGTTCA +AGTTCT +AGTTCC +AGTTCG +AGTTGA +AGTTGT +AGTTGC +AGTTGG +AGTCAA +AGTCAT +AGTCAC +AGTCAG +AGTCTA +AGTCTT +AGTCTC +AGTCTG +AGTCCA +AGTCCT +AGTCCC +AGTCCG +AGTCGA +AGTCGT +AGTCGC +AGTCGG +AGTGAA +AGTGAT +AGTGAC +AGTGAG +AGTGTA +AGTGTT +AGTGTC +AGTGTG +AGTGCA +AGTGCT +AGTGCC +AGTGCG +AGTGGA +AGTGGT +AGTGGC +AGTGGG +AGCAAA +AGCAAT +AGCAAC +AGCAAG +AGCATA +AGCATT +AGCATC +AGCATG +AGCACA +AGCACT +AGCACC +AGCACG +AGCAGA +AGCAGT +AGCAGC +AGCAGG +AGCTAA +AGCTAT +AGCTAC +AGCTAG +AGCTTA +AGCTTT +AGCTTC +AGCTTG +AGCTCA +AGCTCT +AGCTCC +AGCTCG +AGCTGA +AGCTGT +AGCTGC +AGCTGG +AGCCAA +AGCCAT +AGCCAC +AGCCAG +AGCCTA +AGCCTT +AGCCTC +AGCCTG +AGCCCA +AGCCCT +AGCCCC +AGCCCG +AGCCGA +AGCCGT +AGCCGC +AGCCGG +AGCGAA +AGCGAT +AGCGAC +AGCGAG +AGCGTA +AGCGTT +AGCGTC +AGCGTG +AGCGCA +AGCGCT +AGCGCC +AGCGCG +AGCGGA +AGCGGT +AGCGGC +AGCGGG +AGGAAA +AGGAAT +AGGAAC +AGGAAG +AGGATA +AGGATT +AGGATC +AGGATG +AGGACA +AGGACT +AGGACC +AGGACG +AGGAGA +AGGAGT +AGGAGC +AGGAGG +AGGTAA +AGGTAT +AGGTAC +AGGTAG +AGGTTA +AGGTTT +AGGTTC +AGGTTG +AGGTCA +AGGTCT +AGGTCC +AGGTCG +AGGTGA +AGGTGT +AGGTGC +AGGTGG +AGGCAA +AGGCAT +AGGCAC +AGGCAG +AGGCTA +AGGCTT +AGGCTC +AGGCTG +AGGCCA +AGGCCT +AGGCCC +AGGCCG +AGGCGA +AGGCGT +AGGCGC +AGGCGG +AGGGAA +AGGGAT +AGGGAC +AGGGAG +AGGGTA +AGGGTT +AGGGTC +AGGGTG +AGGGCA +AGGGCT +AGGGCC +AGGGCG +AGGGGA +AGGGGT +AGGGGC +AGGGGG +TAAAAA +TAAAAT +TAAAAC +TAAAAG +TAAATA +TAAATT +TAAATC +TAAATG +TAAACA +TAAACT +TAAACC +TAAACG +TAAAGA +TAAAGT +TAAAGC +TAAAGG +TAATAA +TAATAT +TAATAC +TAATAG +TAATTA +TAATTT +TAATTC +TAATTG +TAATCA +TAATCT +TAATCC +TAATCG +TAATGA +TAATGT +TAATGC +TAATGG +TAACAA +TAACAT +TAACAC +TAACAG +TAACTA +TAACTT +TAACTC +TAACTG +TAACCA +TAACCT +TAACCC +TAACCG +TAACGA +TAACGT +TAACGC +TAACGG +TAAGAA +TAAGAT +TAAGAC +TAAGAG +TAAGTA +TAAGTT +TAAGTC +TAAGTG +TAAGCA +TAAGCT +TAAGCC +TAAGCG +TAAGGA +TAAGGT +TAAGGC +TAAGGG +TATAAA +TATAAT +TATAAC +TATAAG +TATATA +TATATT +TATATC +TATATG +TATACA +TATACT +TATACC +TATACG +TATAGA +TATAGT +TATAGC +TATAGG +TATTAA +TATTAT +TATTAC +TATTAG +TATTTA +TATTTT +TATTTC +TATTTG +TATTCA +TATTCT +TATTCC +TATTCG +TATTGA +TATTGT +TATTGC +TATTGG +TATCAA +TATCAT +TATCAC +TATCAG +TATCTA +TATCTT +TATCTC +TATCTG +TATCCA +TATCCT +TATCCC +TATCCG +TATCGA +TATCGT +TATCGC +TATCGG +TATGAA +TATGAT +TATGAC +TATGAG +TATGTA +TATGTT +TATGTC +TATGTG +TATGCA +TATGCT +TATGCC +TATGCG +TATGGA +TATGGT +TATGGC +TATGGG +TACAAA +TACAAT +TACAAC +TACAAG +TACATA +TACATT +TACATC +TACATG +TACACA +TACACT +TACACC +TACACG +TACAGA +TACAGT +TACAGC +TACAGG +TACTAA +TACTAT +TACTAC +TACTAG +TACTTA +TACTTT +TACTTC +TACTTG +TACTCA +TACTCT +TACTCC +TACTCG +TACTGA +TACTGT +TACTGC +TACTGG +TACCAA +TACCAT +TACCAC +TACCAG +TACCTA +TACCTT +TACCTC +TACCTG +TACCCA +TACCCT +TACCCC +TACCCG +TACCGA +TACCGT +TACCGC +TACCGG +TACGAA +TACGAT +TACGAC +TACGAG +TACGTA +TACGTT +TACGTC +TACGTG +TACGCA +TACGCT +TACGCC +TACGCG +TACGGA +TACGGT +TACGGC +TACGGG +TAGAAA +TAGAAT +TAGAAC +TAGAAG +TAGATA +TAGATT +TAGATC +TAGATG +TAGACA +TAGACT +TAGACC +TAGACG +TAGAGA +TAGAGT +TAGAGC +TAGAGG +TAGTAA +TAGTAT +TAGTAC +TAGTAG +TAGTTA +TAGTTT +TAGTTC +TAGTTG +TAGTCA +TAGTCT +TAGTCC +TAGTCG +TAGTGA +TAGTGT +TAGTGC +TAGTGG +TAGCAA +TAGCAT +TAGCAC +TAGCAG +TAGCTA +TAGCTT +TAGCTC +TAGCTG +TAGCCA +TAGCCT +TAGCCC +TAGCCG +TAGCGA +TAGCGT +TAGCGC +TAGCGG +TAGGAA +TAGGAT +TAGGAC +TAGGAG +TAGGTA +TAGGTT +TAGGTC +TAGGTG +TAGGCA +TAGGCT +TAGGCC +TAGGCG +TAGGGA +TAGGGT +TAGGGC +TAGGGG +TTAAAA +TTAAAT +TTAAAC +TTAAAG +TTAATA +TTAATT +TTAATC +TTAATG +TTAACA +TTAACT +TTAACC +TTAACG +TTAAGA +TTAAGT +TTAAGC +TTAAGG +TTATAA +TTATAT +TTATAC +TTATAG +TTATTA +TTATTT +TTATTC +TTATTG +TTATCA +TTATCT +TTATCC +TTATCG +TTATGA +TTATGT +TTATGC +TTATGG +TTACAA +TTACAT +TTACAC +TTACAG +TTACTA +TTACTT +TTACTC +TTACTG +TTACCA +TTACCT +TTACCC +TTACCG +TTACGA +TTACGT +TTACGC +TTACGG +TTAGAA +TTAGAT +TTAGAC +TTAGAG +TTAGTA +TTAGTT +TTAGTC +TTAGTG +TTAGCA +TTAGCT +TTAGCC +TTAGCG +TTAGGA +TTAGGT +TTAGGC +TTAGGG +TTTAAA +TTTAAT +TTTAAC +TTTAAG +TTTATA +TTTATT +TTTATC +TTTATG +TTTACA +TTTACT +TTTACC +TTTACG +TTTAGA +TTTAGT +TTTAGC +TTTAGG +TTTTAA +TTTTAT +TTTTAC +TTTTAG +TTTTTA +TTTTTT +TTTTTC +TTTTTG +TTTTCA +TTTTCT +TTTTCC +TTTTCG +TTTTGA +TTTTGT +TTTTGC +TTTTGG +TTTCAA +TTTCAT +TTTCAC +TTTCAG +TTTCTA +TTTCTT +TTTCTC +TTTCTG +TTTCCA +TTTCCT +TTTCCC +TTTCCG +TTTCGA +TTTCGT +TTTCGC +TTTCGG +TTTGAA +TTTGAT +TTTGAC +TTTGAG +TTTGTA +TTTGTT +TTTGTC +TTTGTG +TTTGCA +TTTGCT +TTTGCC +TTTGCG +TTTGGA +TTTGGT +TTTGGC +TTTGGG +TTCAAA +TTCAAT +TTCAAC +TTCAAG +TTCATA +TTCATT +TTCATC +TTCATG +TTCACA +TTCACT +TTCACC +TTCACG +TTCAGA +TTCAGT +TTCAGC +TTCAGG +TTCTAA +TTCTAT +TTCTAC +TTCTAG +TTCTTA +TTCTTT +TTCTTC +TTCTTG +TTCTCA +TTCTCT +TTCTCC +TTCTCG +TTCTGA +TTCTGT +TTCTGC +TTCTGG +TTCCAA +TTCCAT +TTCCAC +TTCCAG +TTCCTA +TTCCTT +TTCCTC +TTCCTG +TTCCCA +TTCCCT +TTCCCC +TTCCCG +TTCCGA +TTCCGT +TTCCGC +TTCCGG +TTCGAA +TTCGAT +TTCGAC +TTCGAG +TTCGTA +TTCGTT +TTCGTC +TTCGTG +TTCGCA +TTCGCT +TTCGCC +TTCGCG +TTCGGA +TTCGGT +TTCGGC +TTCGGG +TTGAAA +TTGAAT +TTGAAC +TTGAAG +TTGATA +TTGATT +TTGATC +TTGATG +TTGACA +TTGACT +TTGACC +TTGACG +TTGAGA +TTGAGT +TTGAGC +TTGAGG +TTGTAA +TTGTAT +TTGTAC +TTGTAG +TTGTTA +TTGTTT +TTGTTC +TTGTTG +TTGTCA +TTGTCT +TTGTCC +TTGTCG +TTGTGA +TTGTGT +TTGTGC +TTGTGG +TTGCAA +TTGCAT +TTGCAC +TTGCAG +TTGCTA +TTGCTT +TTGCTC +TTGCTG +TTGCCA +TTGCCT +TTGCCC +TTGCCG +TTGCGA +TTGCGT +TTGCGC +TTGCGG +TTGGAA +TTGGAT +TTGGAC +TTGGAG +TTGGTA +TTGGTT +TTGGTC +TTGGTG +TTGGCA +TTGGCT +TTGGCC +TTGGCG +TTGGGA +TTGGGT +TTGGGC +TTGGGG +TCAAAA +TCAAAT +TCAAAC +TCAAAG +TCAATA +TCAATT +TCAATC +TCAATG +TCAACA +TCAACT +TCAACC +TCAACG +TCAAGA +TCAAGT +TCAAGC +TCAAGG +TCATAA +TCATAT +TCATAC +TCATAG +TCATTA +TCATTT +TCATTC +TCATTG +TCATCA +TCATCT +TCATCC +TCATCG +TCATGA +TCATGT +TCATGC +TCATGG +TCACAA +TCACAT +TCACAC +TCACAG +TCACTA +TCACTT +TCACTC +TCACTG +TCACCA +TCACCT +TCACCC +TCACCG +TCACGA +TCACGT +TCACGC +TCACGG +TCAGAA +TCAGAT +TCAGAC +TCAGAG +TCAGTA +TCAGTT +TCAGTC +TCAGTG +TCAGCA +TCAGCT +TCAGCC +TCAGCG +TCAGGA +TCAGGT +TCAGGC +TCAGGG +TCTAAA +TCTAAT +TCTAAC +TCTAAG +TCTATA +TCTATT +TCTATC +TCTATG +TCTACA +TCTACT +TCTACC +TCTACG +TCTAGA +TCTAGT +TCTAGC +TCTAGG +TCTTAA +TCTTAT +TCTTAC +TCTTAG +TCTTTA +TCTTTT +TCTTTC +TCTTTG +TCTTCA +TCTTCT +TCTTCC +TCTTCG +TCTTGA +TCTTGT +TCTTGC +TCTTGG +TCTCAA +TCTCAT +TCTCAC +TCTCAG +TCTCTA +TCTCTT +TCTCTC +TCTCTG +TCTCCA +TCTCCT +TCTCCC +TCTCCG +TCTCGA +TCTCGT +TCTCGC +TCTCGG +TCTGAA +TCTGAT +TCTGAC +TCTGAG +TCTGTA +TCTGTT +TCTGTC +TCTGTG +TCTGCA +TCTGCT +TCTGCC +TCTGCG +TCTGGA +TCTGGT +TCTGGC +TCTGGG +TCCAAA +TCCAAT +TCCAAC +TCCAAG +TCCATA +TCCATT +TCCATC +TCCATG +TCCACA +TCCACT +TCCACC +TCCACG +TCCAGA +TCCAGT +TCCAGC +TCCAGG +TCCTAA +TCCTAT +TCCTAC +TCCTAG +TCCTTA +TCCTTT +TCCTTC +TCCTTG +TCCTCA +TCCTCT +TCCTCC +TCCTCG +TCCTGA +TCCTGT +TCCTGC +TCCTGG +TCCCAA +TCCCAT +TCCCAC +TCCCAG +TCCCTA +TCCCTT +TCCCTC +TCCCTG +TCCCCA +TCCCCT +TCCCCC +TCCCCG +TCCCGA +TCCCGT +TCCCGC +TCCCGG +TCCGAA +TCCGAT +TCCGAC +TCCGAG +TCCGTA +TCCGTT +TCCGTC +TCCGTG +TCCGCA +TCCGCT +TCCGCC +TCCGCG +TCCGGA +TCCGGT +TCCGGC +TCCGGG +TCGAAA +TCGAAT +TCGAAC +TCGAAG +TCGATA +TCGATT +TCGATC +TCGATG +TCGACA +TCGACT +TCGACC +TCGACG +TCGAGA +TCGAGT +TCGAGC +TCGAGG +TCGTAA +TCGTAT +TCGTAC +TCGTAG +TCGTTA +TCGTTT +TCGTTC +TCGTTG +TCGTCA +TCGTCT +TCGTCC +TCGTCG +TCGTGA +TCGTGT +TCGTGC +TCGTGG +TCGCAA +TCGCAT +TCGCAC +TCGCAG +TCGCTA +TCGCTT +TCGCTC +TCGCTG +TCGCCA +TCGCCT +TCGCCC +TCGCCG +TCGCGA +TCGCGT +TCGCGC +TCGCGG +TCGGAA +TCGGAT +TCGGAC +TCGGAG +TCGGTA +TCGGTT +TCGGTC +TCGGTG +TCGGCA +TCGGCT +TCGGCC +TCGGCG +TCGGGA +TCGGGT +TCGGGC +TCGGGG +TGAAAA +TGAAAT +TGAAAC +TGAAAG +TGAATA +TGAATT +TGAATC +TGAATG +TGAACA +TGAACT +TGAACC +TGAACG +TGAAGA +TGAAGT +TGAAGC +TGAAGG +TGATAA +TGATAT +TGATAC +TGATAG +TGATTA +TGATTT +TGATTC +TGATTG +TGATCA +TGATCT +TGATCC +TGATCG +TGATGA +TGATGT +TGATGC +TGATGG +TGACAA +TGACAT +TGACAC +TGACAG +TGACTA +TGACTT +TGACTC +TGACTG +TGACCA +TGACCT +TGACCC +TGACCG +TGACGA +TGACGT +TGACGC +TGACGG +TGAGAA +TGAGAT +TGAGAC +TGAGAG +TGAGTA +TGAGTT +TGAGTC +TGAGTG +TGAGCA +TGAGCT +TGAGCC +TGAGCG +TGAGGA +TGAGGT +TGAGGC +TGAGGG +TGTAAA +TGTAAT +TGTAAC +TGTAAG +TGTATA +TGTATT +TGTATC +TGTATG +TGTACA +TGTACT +TGTACC +TGTACG +TGTAGA +TGTAGT +TGTAGC +TGTAGG +TGTTAA +TGTTAT +TGTTAC +TGTTAG +TGTTTA +TGTTTT +TGTTTC +TGTTTG +TGTTCA +TGTTCT +TGTTCC +TGTTCG +TGTTGA +TGTTGT +TGTTGC +TGTTGG +TGTCAA +TGTCAT +TGTCAC +TGTCAG +TGTCTA +TGTCTT +TGTCTC +TGTCTG +TGTCCA +TGTCCT +TGTCCC +TGTCCG +TGTCGA +TGTCGT +TGTCGC +TGTCGG +TGTGAA +TGTGAT +TGTGAC +TGTGAG +TGTGTA +TGTGTT +TGTGTC +TGTGTG +TGTGCA +TGTGCT +TGTGCC +TGTGCG +TGTGGA +TGTGGT +TGTGGC +TGTGGG +TGCAAA +TGCAAT +TGCAAC +TGCAAG +TGCATA +TGCATT +TGCATC +TGCATG +TGCACA +TGCACT +TGCACC +TGCACG +TGCAGA +TGCAGT +TGCAGC +TGCAGG +TGCTAA +TGCTAT +TGCTAC +TGCTAG +TGCTTA +TGCTTT +TGCTTC +TGCTTG +TGCTCA +TGCTCT +TGCTCC +TGCTCG +TGCTGA +TGCTGT +TGCTGC +TGCTGG +TGCCAA +TGCCAT +TGCCAC +TGCCAG +TGCCTA +TGCCTT +TGCCTC +TGCCTG +TGCCCA +TGCCCT +TGCCCC +TGCCCG +TGCCGA +TGCCGT +TGCCGC +TGCCGG +TGCGAA +TGCGAT +TGCGAC +TGCGAG +TGCGTA +TGCGTT +TGCGTC +TGCGTG +TGCGCA +TGCGCT +TGCGCC +TGCGCG +TGCGGA +TGCGGT +TGCGGC +TGCGGG +TGGAAA +TGGAAT +TGGAAC +TGGAAG +TGGATA +TGGATT +TGGATC +TGGATG +TGGACA +TGGACT +TGGACC +TGGACG +TGGAGA +TGGAGT +TGGAGC +TGGAGG +TGGTAA +TGGTAT +TGGTAC +TGGTAG +TGGTTA +TGGTTT +TGGTTC +TGGTTG +TGGTCA +TGGTCT +TGGTCC +TGGTCG +TGGTGA +TGGTGT +TGGTGC +TGGTGG +TGGCAA +TGGCAT +TGGCAC +TGGCAG +TGGCTA +TGGCTT +TGGCTC +TGGCTG +TGGCCA +TGGCCT +TGGCCC +TGGCCG +TGGCGA +TGGCGT +TGGCGC +TGGCGG +TGGGAA +TGGGAT +TGGGAC +TGGGAG +TGGGTA +TGGGTT +TGGGTC +TGGGTG +TGGGCA +TGGGCT +TGGGCC +TGGGCG +TGGGGA +TGGGGT +TGGGGC +TGGGGG +CAAAAA +CAAAAT +CAAAAC +CAAAAG +CAAATA +CAAATT +CAAATC +CAAATG +CAAACA +CAAACT +CAAACC +CAAACG +CAAAGA +CAAAGT +CAAAGC +CAAAGG +CAATAA +CAATAT +CAATAC +CAATAG +CAATTA +CAATTT +CAATTC +CAATTG +CAATCA +CAATCT +CAATCC +CAATCG +CAATGA +CAATGT +CAATGC +CAATGG +CAACAA +CAACAT +CAACAC +CAACAG +CAACTA +CAACTT +CAACTC +CAACTG +CAACCA +CAACCT +CAACCC +CAACCG +CAACGA +CAACGT +CAACGC +CAACGG +CAAGAA +CAAGAT +CAAGAC +CAAGAG +CAAGTA +CAAGTT +CAAGTC +CAAGTG +CAAGCA +CAAGCT +CAAGCC +CAAGCG +CAAGGA +CAAGGT +CAAGGC +CAAGGG +CATAAA +CATAAT +CATAAC +CATAAG +CATATA +CATATT +CATATC +CATATG +CATACA +CATACT +CATACC +CATACG +CATAGA +CATAGT +CATAGC +CATAGG +CATTAA +CATTAT +CATTAC +CATTAG +CATTTA +CATTTT +CATTTC +CATTTG +CATTCA +CATTCT +CATTCC +CATTCG +CATTGA +CATTGT +CATTGC +CATTGG +CATCAA +CATCAT +CATCAC +CATCAG +CATCTA +CATCTT +CATCTC +CATCTG +CATCCA +CATCCT +CATCCC +CATCCG +CATCGA +CATCGT +CATCGC +CATCGG +CATGAA +CATGAT +CATGAC +CATGAG +CATGTA +CATGTT +CATGTC +CATGTG +CATGCA +CATGCT +CATGCC +CATGCG +CATGGA +CATGGT +CATGGC +CATGGG +CACAAA +CACAAT +CACAAC +CACAAG +CACATA +CACATT +CACATC +CACATG +CACACA +CACACT +CACACC +CACACG +CACAGA +CACAGT +CACAGC +CACAGG +CACTAA +CACTAT +CACTAC +CACTAG +CACTTA +CACTTT +CACTTC +CACTTG +CACTCA +CACTCT +CACTCC +CACTCG +CACTGA +CACTGT +CACTGC +CACTGG +CACCAA +CACCAT +CACCAC +CACCAG +CACCTA +CACCTT +CACCTC +CACCTG +CACCCA +CACCCT +CACCCC +CACCCG +CACCGA +CACCGT +CACCGC +CACCGG +CACGAA +CACGAT +CACGAC +CACGAG +CACGTA +CACGTT +CACGTC +CACGTG +CACGCA +CACGCT +CACGCC +CACGCG +CACGGA +CACGGT +CACGGC +CACGGG +CAGAAA +CAGAAT +CAGAAC +CAGAAG +CAGATA +CAGATT +CAGATC +CAGATG +CAGACA +CAGACT +CAGACC +CAGACG +CAGAGA +CAGAGT +CAGAGC +CAGAGG +CAGTAA +CAGTAT +CAGTAC +CAGTAG +CAGTTA +CAGTTT +CAGTTC +CAGTTG +CAGTCA +CAGTCT +CAGTCC +CAGTCG +CAGTGA +CAGTGT +CAGTGC +CAGTGG +CAGCAA +CAGCAT +CAGCAC +CAGCAG +CAGCTA +CAGCTT +CAGCTC +CAGCTG +CAGCCA +CAGCCT +CAGCCC +CAGCCG +CAGCGA +CAGCGT +CAGCGC +CAGCGG +CAGGAA +CAGGAT +CAGGAC +CAGGAG +CAGGTA +CAGGTT +CAGGTC +CAGGTG +CAGGCA +CAGGCT +CAGGCC +CAGGCG +CAGGGA +CAGGGT +CAGGGC +CAGGGG +CTAAAA +CTAAAT +CTAAAC +CTAAAG +CTAATA +CTAATT +CTAATC +CTAATG +CTAACA +CTAACT +CTAACC +CTAACG +CTAAGA +CTAAGT +CTAAGC +CTAAGG +CTATAA +CTATAT +CTATAC +CTATAG +CTATTA +CTATTT +CTATTC +CTATTG +CTATCA +CTATCT +CTATCC +CTATCG +CTATGA +CTATGT +CTATGC +CTATGG +CTACAA +CTACAT +CTACAC +CTACAG +CTACTA +CTACTT +CTACTC +CTACTG +CTACCA +CTACCT +CTACCC +CTACCG +CTACGA +CTACGT +CTACGC +CTACGG +CTAGAA +CTAGAT +CTAGAC +CTAGAG +CTAGTA +CTAGTT +CTAGTC +CTAGTG +CTAGCA +CTAGCT +CTAGCC +CTAGCG +CTAGGA +CTAGGT +CTAGGC +CTAGGG +CTTAAA +CTTAAT +CTTAAC +CTTAAG +CTTATA +CTTATT +CTTATC +CTTATG +CTTACA +CTTACT +CTTACC +CTTACG +CTTAGA +CTTAGT +CTTAGC +CTTAGG +CTTTAA +CTTTAT +CTTTAC +CTTTAG +CTTTTA +CTTTTT +CTTTTC +CTTTTG +CTTTCA +CTTTCT +CTTTCC +CTTTCG +CTTTGA +CTTTGT +CTTTGC +CTTTGG +CTTCAA +CTTCAT +CTTCAC +CTTCAG +CTTCTA +CTTCTT +CTTCTC +CTTCTG +CTTCCA +CTTCCT +CTTCCC +CTTCCG +CTTCGA +CTTCGT +CTTCGC +CTTCGG +CTTGAA +CTTGAT +CTTGAC +CTTGAG +CTTGTA +CTTGTT +CTTGTC +CTTGTG +CTTGCA +CTTGCT +CTTGCC +CTTGCG +CTTGGA +CTTGGT +CTTGGC +CTTGGG +CTCAAA +CTCAAT +CTCAAC +CTCAAG +CTCATA +CTCATT +CTCATC +CTCATG +CTCACA +CTCACT +CTCACC +CTCACG +CTCAGA +CTCAGT +CTCAGC +CTCAGG +CTCTAA +CTCTAT +CTCTAC +CTCTAG +CTCTTA +CTCTTT +CTCTTC +CTCTTG +CTCTCA +CTCTCT +CTCTCC +CTCTCG +CTCTGA +CTCTGT +CTCTGC +CTCTGG +CTCCAA +CTCCAT +CTCCAC +CTCCAG +CTCCTA +CTCCTT +CTCCTC +CTCCTG +CTCCCA +CTCCCT +CTCCCC +CTCCCG +CTCCGA +CTCCGT +CTCCGC +CTCCGG +CTCGAA +CTCGAT +CTCGAC +CTCGAG +CTCGTA +CTCGTT +CTCGTC +CTCGTG +CTCGCA +CTCGCT +CTCGCC +CTCGCG +CTCGGA +CTCGGT +CTCGGC +CTCGGG +CTGAAA +CTGAAT +CTGAAC +CTGAAG +CTGATA +CTGATT +CTGATC +CTGATG +CTGACA +CTGACT +CTGACC +CTGACG +CTGAGA +CTGAGT +CTGAGC +CTGAGG +CTGTAA +CTGTAT +CTGTAC +CTGTAG +CTGTTA +CTGTTT +CTGTTC +CTGTTG +CTGTCA +CTGTCT +CTGTCC +CTGTCG +CTGTGA +CTGTGT +CTGTGC +CTGTGG +CTGCAA +CTGCAT +CTGCAC +CTGCAG +CTGCTA +CTGCTT +CTGCTC +CTGCTG +CTGCCA +CTGCCT +CTGCCC +CTGCCG +CTGCGA +CTGCGT +CTGCGC +CTGCGG +CTGGAA +CTGGAT +CTGGAC +CTGGAG +CTGGTA +CTGGTT +CTGGTC +CTGGTG +CTGGCA +CTGGCT +CTGGCC +CTGGCG +CTGGGA +CTGGGT +CTGGGC +CTGGGG +CCAAAA +CCAAAT +CCAAAC +CCAAAG +CCAATA +CCAATT +CCAATC +CCAATG +CCAACA +CCAACT +CCAACC +CCAACG +CCAAGA +CCAAGT +CCAAGC +CCAAGG +CCATAA +CCATAT +CCATAC +CCATAG +CCATTA +CCATTT +CCATTC +CCATTG +CCATCA +CCATCT +CCATCC +CCATCG +CCATGA +CCATGT +CCATGC +CCATGG +CCACAA +CCACAT +CCACAC +CCACAG +CCACTA +CCACTT +CCACTC +CCACTG +CCACCA +CCACCT +CCACCC +CCACCG +CCACGA +CCACGT +CCACGC +CCACGG +CCAGAA +CCAGAT +CCAGAC +CCAGAG +CCAGTA +CCAGTT +CCAGTC +CCAGTG +CCAGCA +CCAGCT +CCAGCC +CCAGCG +CCAGGA +CCAGGT +CCAGGC +CCAGGG +CCTAAA +CCTAAT +CCTAAC +CCTAAG +CCTATA +CCTATT +CCTATC +CCTATG +CCTACA +CCTACT +CCTACC +CCTACG +CCTAGA +CCTAGT +CCTAGC +CCTAGG +CCTTAA +CCTTAT +CCTTAC +CCTTAG +CCTTTA +CCTTTT +CCTTTC +CCTTTG +CCTTCA +CCTTCT +CCTTCC +CCTTCG +CCTTGA +CCTTGT +CCTTGC +CCTTGG +CCTCAA +CCTCAT +CCTCAC +CCTCAG +CCTCTA +CCTCTT +CCTCTC +CCTCTG +CCTCCA +CCTCCT +CCTCCC +CCTCCG +CCTCGA +CCTCGT +CCTCGC +CCTCGG +CCTGAA +CCTGAT +CCTGAC +CCTGAG +CCTGTA +CCTGTT +CCTGTC +CCTGTG +CCTGCA +CCTGCT +CCTGCC +CCTGCG +CCTGGA +CCTGGT +CCTGGC +CCTGGG +CCCAAA +CCCAAT +CCCAAC +CCCAAG +CCCATA +CCCATT +CCCATC +CCCATG +CCCACA +CCCACT +CCCACC +CCCACG +CCCAGA +CCCAGT +CCCAGC +CCCAGG +CCCTAA +CCCTAT +CCCTAC +CCCTAG +CCCTTA +CCCTTT +CCCTTC +CCCTTG +CCCTCA +CCCTCT +CCCTCC +CCCTCG +CCCTGA +CCCTGT +CCCTGC +CCCTGG +CCCCAA +CCCCAT +CCCCAC +CCCCAG +CCCCTA +CCCCTT +CCCCTC +CCCCTG +CCCCCA +CCCCCT +CCCCCC +CCCCCG +CCCCGA +CCCCGT +CCCCGC +CCCCGG +CCCGAA +CCCGAT +CCCGAC +CCCGAG +CCCGTA +CCCGTT +CCCGTC +CCCGTG +CCCGCA +CCCGCT +CCCGCC +CCCGCG +CCCGGA +CCCGGT +CCCGGC +CCCGGG +CCGAAA +CCGAAT +CCGAAC +CCGAAG +CCGATA +CCGATT +CCGATC +CCGATG +CCGACA +CCGACT +CCGACC +CCGACG +CCGAGA +CCGAGT +CCGAGC +CCGAGG +CCGTAA +CCGTAT +CCGTAC +CCGTAG +CCGTTA +CCGTTT +CCGTTC +CCGTTG +CCGTCA +CCGTCT +CCGTCC +CCGTCG +CCGTGA +CCGTGT +CCGTGC +CCGTGG +CCGCAA +CCGCAT +CCGCAC +CCGCAG +CCGCTA +CCGCTT +CCGCTC +CCGCTG +CCGCCA +CCGCCT +CCGCCC +CCGCCG +CCGCGA +CCGCGT +CCGCGC +CCGCGG +CCGGAA +CCGGAT +CCGGAC +CCGGAG +CCGGTA +CCGGTT +CCGGTC +CCGGTG +CCGGCA +CCGGCT +CCGGCC +CCGGCG +CCGGGA +CCGGGT +CCGGGC +CCGGGG +CGAAAA +CGAAAT +CGAAAC +CGAAAG +CGAATA +CGAATT +CGAATC +CGAATG +CGAACA +CGAACT +CGAACC +CGAACG +CGAAGA +CGAAGT +CGAAGC +CGAAGG +CGATAA +CGATAT +CGATAC +CGATAG +CGATTA +CGATTT +CGATTC +CGATTG +CGATCA +CGATCT +CGATCC +CGATCG +CGATGA +CGATGT +CGATGC +CGATGG +CGACAA +CGACAT +CGACAC +CGACAG +CGACTA +CGACTT +CGACTC +CGACTG +CGACCA +CGACCT +CGACCC +CGACCG +CGACGA +CGACGT +CGACGC +CGACGG +CGAGAA +CGAGAT +CGAGAC +CGAGAG +CGAGTA +CGAGTT +CGAGTC +CGAGTG +CGAGCA +CGAGCT +CGAGCC +CGAGCG +CGAGGA +CGAGGT +CGAGGC +CGAGGG +CGTAAA +CGTAAT +CGTAAC +CGTAAG +CGTATA +CGTATT +CGTATC +CGTATG +CGTACA +CGTACT +CGTACC +CGTACG +CGTAGA +CGTAGT +CGTAGC +CGTAGG +CGTTAA +CGTTAT +CGTTAC +CGTTAG +CGTTTA +CGTTTT +CGTTTC +CGTTTG +CGTTCA +CGTTCT +CGTTCC +CGTTCG +CGTTGA +CGTTGT +CGTTGC +CGTTGG +CGTCAA +CGTCAT +CGTCAC +CGTCAG +CGTCTA +CGTCTT +CGTCTC +CGTCTG +CGTCCA +CGTCCT +CGTCCC +CGTCCG +CGTCGA +CGTCGT +CGTCGC +CGTCGG +CGTGAA +CGTGAT +CGTGAC +CGTGAG +CGTGTA +CGTGTT +CGTGTC +CGTGTG +CGTGCA +CGTGCT +CGTGCC +CGTGCG +CGTGGA +CGTGGT +CGTGGC +CGTGGG +CGCAAA +CGCAAT +CGCAAC +CGCAAG +CGCATA +CGCATT +CGCATC +CGCATG +CGCACA +CGCACT +CGCACC +CGCACG +CGCAGA +CGCAGT +CGCAGC +CGCAGG +CGCTAA +CGCTAT +CGCTAC +CGCTAG +CGCTTA +CGCTTT +CGCTTC +CGCTTG +CGCTCA +CGCTCT +CGCTCC +CGCTCG +CGCTGA +CGCTGT +CGCTGC +CGCTGG +CGCCAA +CGCCAT +CGCCAC +CGCCAG +CGCCTA +CGCCTT +CGCCTC +CGCCTG +CGCCCA +CGCCCT +CGCCCC +CGCCCG +CGCCGA +CGCCGT +CGCCGC +CGCCGG +CGCGAA +CGCGAT +CGCGAC +CGCGAG +CGCGTA +CGCGTT +CGCGTC +CGCGTG +CGCGCA +CGCGCT +CGCGCC +CGCGCG +CGCGGA +CGCGGT +CGCGGC +CGCGGG +CGGAAA +CGGAAT +CGGAAC +CGGAAG +CGGATA +CGGATT +CGGATC +CGGATG +CGGACA +CGGACT +CGGACC +CGGACG +CGGAGA +CGGAGT +CGGAGC +CGGAGG +CGGTAA +CGGTAT +CGGTAC +CGGTAG +CGGTTA +CGGTTT +CGGTTC +CGGTTG +CGGTCA +CGGTCT +CGGTCC +CGGTCG +CGGTGA +CGGTGT +CGGTGC +CGGTGG +CGGCAA +CGGCAT +CGGCAC +CGGCAG +CGGCTA +CGGCTT +CGGCTC +CGGCTG +CGGCCA +CGGCCT +CGGCCC +CGGCCG +CGGCGA +CGGCGT +CGGCGC +CGGCGG +CGGGAA +CGGGAT +CGGGAC +CGGGAG +CGGGTA +CGGGTT +CGGGTC +CGGGTG +CGGGCA +CGGGCT +CGGGCC +CGGGCG +CGGGGA +CGGGGT +CGGGGC +CGGGGG +GAAAAA +GAAAAT +GAAAAC +GAAAAG +GAAATA +GAAATT +GAAATC +GAAATG +GAAACA +GAAACT +GAAACC +GAAACG +GAAAGA +GAAAGT +GAAAGC +GAAAGG +GAATAA +GAATAT +GAATAC +GAATAG +GAATTA +GAATTT +GAATTC +GAATTG +GAATCA +GAATCT +GAATCC +GAATCG +GAATGA +GAATGT +GAATGC +GAATGG +GAACAA +GAACAT +GAACAC +GAACAG +GAACTA +GAACTT +GAACTC +GAACTG +GAACCA +GAACCT +GAACCC +GAACCG +GAACGA +GAACGT +GAACGC +GAACGG +GAAGAA +GAAGAT +GAAGAC +GAAGAG +GAAGTA +GAAGTT +GAAGTC +GAAGTG +GAAGCA +GAAGCT +GAAGCC +GAAGCG +GAAGGA +GAAGGT +GAAGGC +GAAGGG +GATAAA +GATAAT +GATAAC +GATAAG +GATATA +GATATT +GATATC +GATATG +GATACA +GATACT +GATACC +GATACG +GATAGA +GATAGT +GATAGC +GATAGG +GATTAA +GATTAT +GATTAC +GATTAG +GATTTA +GATTTT +GATTTC +GATTTG +GATTCA +GATTCT +GATTCC +GATTCG +GATTGA +GATTGT +GATTGC +GATTGG +GATCAA +GATCAT +GATCAC +GATCAG +GATCTA +GATCTT +GATCTC +GATCTG +GATCCA +GATCCT +GATCCC +GATCCG +GATCGA +GATCGT +GATCGC +GATCGG +GATGAA +GATGAT +GATGAC +GATGAG +GATGTA +GATGTT +GATGTC +GATGTG +GATGCA +GATGCT +GATGCC +GATGCG +GATGGA +GATGGT +GATGGC +GATGGG +GACAAA +GACAAT +GACAAC +GACAAG +GACATA +GACATT +GACATC +GACATG +GACACA +GACACT +GACACC +GACACG +GACAGA +GACAGT +GACAGC +GACAGG +GACTAA +GACTAT +GACTAC +GACTAG +GACTTA +GACTTT +GACTTC +GACTTG +GACTCA +GACTCT +GACTCC +GACTCG +GACTGA +GACTGT +GACTGC +GACTGG +GACCAA +GACCAT +GACCAC +GACCAG +GACCTA +GACCTT +GACCTC +GACCTG +GACCCA +GACCCT +GACCCC +GACCCG +GACCGA +GACCGT +GACCGC +GACCGG +GACGAA +GACGAT +GACGAC +GACGAG +GACGTA +GACGTT +GACGTC +GACGTG +GACGCA +GACGCT +GACGCC +GACGCG +GACGGA +GACGGT +GACGGC +GACGGG +GAGAAA +GAGAAT +GAGAAC +GAGAAG +GAGATA +GAGATT +GAGATC +GAGATG +GAGACA +GAGACT +GAGACC +GAGACG +GAGAGA +GAGAGT +GAGAGC +GAGAGG +GAGTAA +GAGTAT +GAGTAC +GAGTAG +GAGTTA +GAGTTT +GAGTTC +GAGTTG +GAGTCA +GAGTCT +GAGTCC +GAGTCG +GAGTGA +GAGTGT +GAGTGC +GAGTGG +GAGCAA +GAGCAT +GAGCAC +GAGCAG +GAGCTA +GAGCTT +GAGCTC +GAGCTG +GAGCCA +GAGCCT +GAGCCC +GAGCCG +GAGCGA +GAGCGT +GAGCGC +GAGCGG +GAGGAA +GAGGAT +GAGGAC +GAGGAG +GAGGTA +GAGGTT +GAGGTC +GAGGTG +GAGGCA +GAGGCT +GAGGCC +GAGGCG +GAGGGA +GAGGGT +GAGGGC +GAGGGG +GTAAAA +GTAAAT +GTAAAC +GTAAAG +GTAATA +GTAATT +GTAATC +GTAATG +GTAACA +GTAACT +GTAACC +GTAACG +GTAAGA +GTAAGT +GTAAGC +GTAAGG +GTATAA +GTATAT +GTATAC +GTATAG +GTATTA +GTATTT +GTATTC +GTATTG +GTATCA +GTATCT +GTATCC +GTATCG +GTATGA +GTATGT +GTATGC +GTATGG +GTACAA +GTACAT +GTACAC +GTACAG +GTACTA +GTACTT +GTACTC +GTACTG +GTACCA +GTACCT +GTACCC +GTACCG +GTACGA +GTACGT +GTACGC +GTACGG +GTAGAA +GTAGAT +GTAGAC +GTAGAG +GTAGTA +GTAGTT +GTAGTC +GTAGTG +GTAGCA +GTAGCT +GTAGCC +GTAGCG +GTAGGA +GTAGGT +GTAGGC +GTAGGG +GTTAAA +GTTAAT +GTTAAC +GTTAAG +GTTATA +GTTATT +GTTATC +GTTATG +GTTACA +GTTACT +GTTACC +GTTACG +GTTAGA +GTTAGT +GTTAGC +GTTAGG +GTTTAA +GTTTAT +GTTTAC +GTTTAG +GTTTTA +GTTTTT +GTTTTC +GTTTTG +GTTTCA +GTTTCT +GTTTCC +GTTTCG +GTTTGA +GTTTGT +GTTTGC +GTTTGG +GTTCAA +GTTCAT +GTTCAC +GTTCAG +GTTCTA +GTTCTT +GTTCTC +GTTCTG +GTTCCA +GTTCCT +GTTCCC +GTTCCG +GTTCGA +GTTCGT +GTTCGC +GTTCGG +GTTGAA +GTTGAT +GTTGAC +GTTGAG +GTTGTA +GTTGTT +GTTGTC +GTTGTG +GTTGCA +GTTGCT +GTTGCC +GTTGCG +GTTGGA +GTTGGT +GTTGGC +GTTGGG +GTCAAA +GTCAAT +GTCAAC +GTCAAG +GTCATA +GTCATT +GTCATC +GTCATG +GTCACA +GTCACT +GTCACC +GTCACG +GTCAGA +GTCAGT +GTCAGC +GTCAGG +GTCTAA +GTCTAT +GTCTAC +GTCTAG +GTCTTA +GTCTTT +GTCTTC +GTCTTG +GTCTCA +GTCTCT +GTCTCC +GTCTCG +GTCTGA +GTCTGT +GTCTGC +GTCTGG +GTCCAA +GTCCAT +GTCCAC +GTCCAG +GTCCTA +GTCCTT +GTCCTC +GTCCTG +GTCCCA +GTCCCT +GTCCCC +GTCCCG +GTCCGA +GTCCGT +GTCCGC +GTCCGG +GTCGAA +GTCGAT +GTCGAC +GTCGAG +GTCGTA +GTCGTT +GTCGTC +GTCGTG +GTCGCA +GTCGCT +GTCGCC +GTCGCG +GTCGGA +GTCGGT +GTCGGC +GTCGGG +GTGAAA +GTGAAT +GTGAAC +GTGAAG +GTGATA +GTGATT +GTGATC +GTGATG +GTGACA +GTGACT +GTGACC +GTGACG +GTGAGA +GTGAGT +GTGAGC +GTGAGG +GTGTAA +GTGTAT +GTGTAC +GTGTAG +GTGTTA +GTGTTT +GTGTTC +GTGTTG +GTGTCA +GTGTCT +GTGTCC +GTGTCG +GTGTGA +GTGTGT +GTGTGC +GTGTGG +GTGCAA +GTGCAT +GTGCAC +GTGCAG +GTGCTA +GTGCTT +GTGCTC +GTGCTG +GTGCCA +GTGCCT +GTGCCC +GTGCCG +GTGCGA +GTGCGT +GTGCGC +GTGCGG +GTGGAA +GTGGAT +GTGGAC +GTGGAG +GTGGTA +GTGGTT +GTGGTC +GTGGTG +GTGGCA +GTGGCT +GTGGCC +GTGGCG +GTGGGA +GTGGGT +GTGGGC +GTGGGG +GCAAAA +GCAAAT +GCAAAC +GCAAAG +GCAATA +GCAATT +GCAATC +GCAATG +GCAACA +GCAACT +GCAACC +GCAACG +GCAAGA +GCAAGT +GCAAGC +GCAAGG +GCATAA +GCATAT +GCATAC +GCATAG +GCATTA +GCATTT +GCATTC +GCATTG +GCATCA +GCATCT +GCATCC +GCATCG +GCATGA +GCATGT +GCATGC +GCATGG +GCACAA +GCACAT +GCACAC +GCACAG +GCACTA +GCACTT +GCACTC +GCACTG +GCACCA +GCACCT +GCACCC +GCACCG +GCACGA +GCACGT +GCACGC +GCACGG +GCAGAA +GCAGAT +GCAGAC +GCAGAG +GCAGTA +GCAGTT +GCAGTC +GCAGTG +GCAGCA +GCAGCT +GCAGCC +GCAGCG +GCAGGA +GCAGGT +GCAGGC +GCAGGG +GCTAAA +GCTAAT +GCTAAC +GCTAAG +GCTATA +GCTATT +GCTATC +GCTATG +GCTACA +GCTACT +GCTACC +GCTACG +GCTAGA +GCTAGT +GCTAGC +GCTAGG +GCTTAA +GCTTAT +GCTTAC +GCTTAG +GCTTTA +GCTTTT +GCTTTC +GCTTTG +GCTTCA +GCTTCT +GCTTCC +GCTTCG +GCTTGA +GCTTGT +GCTTGC +GCTTGG +GCTCAA +GCTCAT +GCTCAC +GCTCAG +GCTCTA +GCTCTT +GCTCTC +GCTCTG +GCTCCA +GCTCCT +GCTCCC +GCTCCG +GCTCGA +GCTCGT +GCTCGC +GCTCGG +GCTGAA +GCTGAT +GCTGAC +GCTGAG +GCTGTA +GCTGTT +GCTGTC +GCTGTG +GCTGCA +GCTGCT +GCTGCC +GCTGCG +GCTGGA +GCTGGT +GCTGGC +GCTGGG +GCCAAA +GCCAAT +GCCAAC +GCCAAG +GCCATA +GCCATT +GCCATC +GCCATG +GCCACA +GCCACT +GCCACC +GCCACG +GCCAGA +GCCAGT +GCCAGC +GCCAGG +GCCTAA +GCCTAT +GCCTAC +GCCTAG +GCCTTA +GCCTTT +GCCTTC +GCCTTG +GCCTCA +GCCTCT +GCCTCC +GCCTCG +GCCTGA +GCCTGT +GCCTGC +GCCTGG +GCCCAA +GCCCAT +GCCCAC +GCCCAG +GCCCTA +GCCCTT +GCCCTC +GCCCTG +GCCCCA +GCCCCT +GCCCCC +GCCCCG +GCCCGA +GCCCGT +GCCCGC +GCCCGG +GCCGAA +GCCGAT +GCCGAC +GCCGAG +GCCGTA +GCCGTT +GCCGTC +GCCGTG +GCCGCA +GCCGCT +GCCGCC +GCCGCG +GCCGGA +GCCGGT +GCCGGC +GCCGGG +GCGAAA +GCGAAT +GCGAAC +GCGAAG +GCGATA +GCGATT +GCGATC +GCGATG +GCGACA +GCGACT +GCGACC +GCGACG +GCGAGA +GCGAGT +GCGAGC +GCGAGG +GCGTAA +GCGTAT +GCGTAC +GCGTAG +GCGTTA +GCGTTT +GCGTTC +GCGTTG +GCGTCA +GCGTCT +GCGTCC +GCGTCG +GCGTGA +GCGTGT +GCGTGC +GCGTGG +GCGCAA +GCGCAT +GCGCAC +GCGCAG +GCGCTA +GCGCTT +GCGCTC +GCGCTG +GCGCCA +GCGCCT +GCGCCC +GCGCCG +GCGCGA +GCGCGT +GCGCGC +GCGCGG +GCGGAA +GCGGAT +GCGGAC +GCGGAG +GCGGTA +GCGGTT +GCGGTC +GCGGTG +GCGGCA +GCGGCT +GCGGCC +GCGGCG +GCGGGA +GCGGGT +GCGGGC +GCGGGG +GGAAAA +GGAAAT +GGAAAC +GGAAAG +GGAATA +GGAATT +GGAATC +GGAATG +GGAACA +GGAACT +GGAACC +GGAACG +GGAAGA +GGAAGT +GGAAGC +GGAAGG +GGATAA +GGATAT +GGATAC +GGATAG +GGATTA +GGATTT +GGATTC +GGATTG +GGATCA +GGATCT +GGATCC +GGATCG +GGATGA +GGATGT +GGATGC +GGATGG +GGACAA +GGACAT +GGACAC +GGACAG +GGACTA +GGACTT +GGACTC +GGACTG +GGACCA +GGACCT +GGACCC +GGACCG +GGACGA +GGACGT +GGACGC +GGACGG +GGAGAA +GGAGAT +GGAGAC +GGAGAG +GGAGTA +GGAGTT +GGAGTC +GGAGTG +GGAGCA +GGAGCT +GGAGCC +GGAGCG +GGAGGA +GGAGGT +GGAGGC +GGAGGG +GGTAAA +GGTAAT +GGTAAC +GGTAAG +GGTATA +GGTATT +GGTATC +GGTATG +GGTACA +GGTACT +GGTACC +GGTACG +GGTAGA +GGTAGT +GGTAGC +GGTAGG +GGTTAA +GGTTAT +GGTTAC +GGTTAG +GGTTTA +GGTTTT +GGTTTC +GGTTTG +GGTTCA +GGTTCT +GGTTCC +GGTTCG +GGTTGA +GGTTGT +GGTTGC +GGTTGG +GGTCAA +GGTCAT +GGTCAC +GGTCAG +GGTCTA +GGTCTT +GGTCTC +GGTCTG +GGTCCA +GGTCCT +GGTCCC +GGTCCG +GGTCGA +GGTCGT +GGTCGC +GGTCGG +GGTGAA +GGTGAT +GGTGAC +GGTGAG +GGTGTA +GGTGTT +GGTGTC +GGTGTG +GGTGCA +GGTGCT +GGTGCC +GGTGCG +GGTGGA +GGTGGT +GGTGGC +GGTGGG +GGCAAA +GGCAAT +GGCAAC +GGCAAG +GGCATA +GGCATT +GGCATC +GGCATG +GGCACA +GGCACT +GGCACC +GGCACG +GGCAGA +GGCAGT +GGCAGC +GGCAGG +GGCTAA +GGCTAT +GGCTAC +GGCTAG +GGCTTA +GGCTTT +GGCTTC +GGCTTG +GGCTCA +GGCTCT +GGCTCC +GGCTCG +GGCTGA +GGCTGT +GGCTGC +GGCTGG +GGCCAA +GGCCAT +GGCCAC +GGCCAG +GGCCTA +GGCCTT +GGCCTC +GGCCTG +GGCCCA +GGCCCT +GGCCCC +GGCCCG +GGCCGA +GGCCGT +GGCCGC +GGCCGG +GGCGAA +GGCGAT +GGCGAC +GGCGAG +GGCGTA +GGCGTT +GGCGTC +GGCGTG +GGCGCA +GGCGCT +GGCGCC +GGCGCG +GGCGGA +GGCGGT +GGCGGC +GGCGGG +GGGAAA +GGGAAT +GGGAAC +GGGAAG +GGGATA +GGGATT +GGGATC +GGGATG +GGGACA +GGGACT +GGGACC +GGGACG +GGGAGA +GGGAGT +GGGAGC +GGGAGG +GGGTAA +GGGTAT +GGGTAC +GGGTAG +GGGTTA +GGGTTT +GGGTTC +GGGTTG +GGGTCA +GGGTCT +GGGTCC +GGGTCG +GGGTGA +GGGTGT +GGGTGC +GGGTGG +GGGCAA +GGGCAT +GGGCAC +GGGCAG +GGGCTA +GGGCTT +GGGCTC +GGGCTG +GGGCCA +GGGCCT +GGGCCC +GGGCCG +GGGCGA +GGGCGT +GGGCGC +GGGCGG +GGGGAA +GGGGAT +GGGGAC +GGGGAG +GGGGTA +GGGGTT +GGGGTC +GGGGTG +GGGGCA +GGGGCT +GGGGCC +GGGGCG +GGGGGA +GGGGGT +GGGGGC +GGGGGG diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/README.md b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..500fd976d92b217455f181bc98c77e612d9a18d9 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/README.md @@ -0,0 +1,202 @@ +--- +base_model: fabihamakhdoomi/TinyDNABERT +library_name: peft +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.12.0 \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/adapter_config.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..28c871c93775eedfaf266f00da8c02f96bc02f29 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/adapter_config.json @@ -0,0 +1,42 @@ +{ + "alpha_pattern": {}, + "auto_mapping": null, + "base_model_name_or_path": "fabihamakhdoomi/TinyDNABERT", + "bias": "none", + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 16, + "lora_dropout": 0.2, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": [ + "classifier", + "score" + ], + "peft_type": "LORA", + "r": 4, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "bert.encoder.layer.1.attention.self.query", + "bert.encoder.layer.0.output.dense", + "bert.encoder.layer.1.output.dense", + "bert.encoder.layer.0.intermediate.dense", + "bert.encoder.layer.1.intermediate.dense", + "bert.encoder.layer.0.attention.output.dense", + "bert.encoder.layer.1.attention.output.dense", + "bert.encoder.layer.0.attention.self.query", + "bert.encoder.layer.0.attention.self.value", + "bert.encoder.layer.1.attention.self.key", + "bert.encoder.layer.1.attention.self.value", + "bert.encoder.layer.0.attention.self.key" + ], + "task_type": "SEQ_CLS", + "use_dora": false, + "use_rslora": false +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/adapter_model.safetensors b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/adapter_model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..383d6c7ef863462d253ee0f0564a88abf90eeaae --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/adapter_model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b962a3253b3e5c29f041e8987191e8928def01d00568ebd7e5b47940de712393 +size 78208 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/optimizer.pt b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..5e3b00de61c6a13b986ba2a6a9209a11ede2b10f --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a0af68c7172219feef9cc292e3bec410d40287f3eb0343e62ca5e247daf012c0 +size 165226 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/rng_state.pth b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..ab07d6382b5d2f619b3927fc83d6571e1bcccc54 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f0f0d7ddc5f130ac1babf491a193bd5c163e7a1b572c7a4eac20de52cd6b2bd7 +size 14244 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/scheduler.pt b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..8c9a6adac55ac4d1afd6e1b61889589acfe31330 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f3e454de3c0f52ddb8e5884c9783308658a6f61ede2ed6e377ca9a68d47c4515 +size 1064 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/special_tokens_map.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..a8b3208c2884c4efb86e49300fdd3dc877220cdf --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/special_tokens_map.json @@ -0,0 +1,7 @@ +{ + "cls_token": "[CLS]", + "mask_token": "[MASK]", + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/tokenizer.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..ebd782de6963b9032ce9374cfe30f5275be1a30b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/tokenizer.json @@ -0,0 +1,4265 @@ +{ + "version": "1.0", + "truncation": { + "direction": "Right", + "max_length": 512, + "strategy": "LongestFirst", + "stride": 0 + }, + "padding": { + "strategy": { + "Fixed": 512 + }, + "direction": "Right", + "pad_to_multiple_of": null, + "pad_id": 0, + "pad_type_id": 0, + "pad_token": "[PAD]" + }, + "added_tokens": [ + { + "id": 0, + "content": "[PAD]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 1, + "content": "[UNK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 2, + "content": "[CLS]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 3, + "content": "[SEP]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 4, + "content": "[MASK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + } + ], + "normalizer": { + "type": "BertNormalizer", + "clean_text": true, + "handle_chinese_chars": true, + "strip_accents": null, + "lowercase": false + }, + "pre_tokenizer": { + "type": "BertPreTokenizer" + }, + "post_processor": { + "type": "TemplateProcessing", + "single": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + } + ], + "pair": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "B", + "type_id": 1 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 1 + } + } + ], + "special_tokens": { + "[CLS]": { + "id": "[CLS]", + "ids": [ + 2 + ], + "tokens": [ + "[CLS]" + ] + }, + "[SEP]": { + "id": "[SEP]", + "ids": [ + 3 + ], + "tokens": [ + "[SEP]" + ] + } + } + }, + "decoder": { + "type": "WordPiece", + "prefix": "##", + "cleanup": true + }, + "model": { + "type": "WordPiece", + "unk_token": "[UNK]", + "continuing_subword_prefix": "##", + "max_input_chars_per_word": 100, + "vocab": { + "[PAD]": 0, + "[UNK]": 1, + "[CLS]": 2, + "[SEP]": 3, + "[MASK]": 4, + "AAAAAA": 5, + "AAAAAT": 6, + "AAAAAC": 7, + "AAAAAG": 8, + "AAAATA": 9, + "AAAATT": 10, + "AAAATC": 11, + "AAAATG": 12, + "AAAACA": 13, + "AAAACT": 14, + "AAAACC": 15, + "AAAACG": 16, + "AAAAGA": 17, + "AAAAGT": 18, + "AAAAGC": 19, + "AAAAGG": 20, + "AAATAA": 21, + "AAATAT": 22, + "AAATAC": 23, + "AAATAG": 24, + "AAATTA": 25, + "AAATTT": 26, + "AAATTC": 27, + "AAATTG": 28, + "AAATCA": 29, + "AAATCT": 30, + "AAATCC": 31, + "AAATCG": 32, + "AAATGA": 33, + "AAATGT": 34, + "AAATGC": 35, + "AAATGG": 36, + "AAACAA": 37, + "AAACAT": 38, + "AAACAC": 39, + "AAACAG": 40, + "AAACTA": 41, + "AAACTT": 42, + "AAACTC": 43, + "AAACTG": 44, + "AAACCA": 45, + "AAACCT": 46, + "AAACCC": 47, + "AAACCG": 48, + "AAACGA": 49, + "AAACGT": 50, + "AAACGC": 51, + "AAACGG": 52, + "AAAGAA": 53, + "AAAGAT": 54, + "AAAGAC": 55, + "AAAGAG": 56, + "AAAGTA": 57, + "AAAGTT": 58, + "AAAGTC": 59, + "AAAGTG": 60, + "AAAGCA": 61, + "AAAGCT": 62, + "AAAGCC": 63, + "AAAGCG": 64, + "AAAGGA": 65, + "AAAGGT": 66, + "AAAGGC": 67, + "AAAGGG": 68, + "AATAAA": 69, + "AATAAT": 70, + "AATAAC": 71, + "AATAAG": 72, + "AATATA": 73, + "AATATT": 74, + "AATATC": 75, + "AATATG": 76, + "AATACA": 77, + "AATACT": 78, + "AATACC": 79, + "AATACG": 80, + "AATAGA": 81, + "AATAGT": 82, + "AATAGC": 83, + "AATAGG": 84, + "AATTAA": 85, + "AATTAT": 86, + "AATTAC": 87, + "AATTAG": 88, + "AATTTA": 89, + "AATTTT": 90, + "AATTTC": 91, + "AATTTG": 92, + "AATTCA": 93, + "AATTCT": 94, + "AATTCC": 95, + "AATTCG": 96, + "AATTGA": 97, + "AATTGT": 98, + "AATTGC": 99, + "AATTGG": 100, + "AATCAA": 101, + "AATCAT": 102, + "AATCAC": 103, + "AATCAG": 104, + "AATCTA": 105, + "AATCTT": 106, + "AATCTC": 107, + "AATCTG": 108, + "AATCCA": 109, + "AATCCT": 110, + "AATCCC": 111, + "AATCCG": 112, + "AATCGA": 113, + "AATCGT": 114, + "AATCGC": 115, + "AATCGG": 116, + "AATGAA": 117, + "AATGAT": 118, + "AATGAC": 119, + "AATGAG": 120, + "AATGTA": 121, + "AATGTT": 122, + "AATGTC": 123, + "AATGTG": 124, + "AATGCA": 125, + "AATGCT": 126, + "AATGCC": 127, + "AATGCG": 128, + "AATGGA": 129, + "AATGGT": 130, + "AATGGC": 131, + "AATGGG": 132, + "AACAAA": 133, + "AACAAT": 134, + "AACAAC": 135, + "AACAAG": 136, + "AACATA": 137, + "AACATT": 138, + "AACATC": 139, + "AACATG": 140, + "AACACA": 141, + "AACACT": 142, + "AACACC": 143, + "AACACG": 144, + "AACAGA": 145, + "AACAGT": 146, + "AACAGC": 147, + "AACAGG": 148, + "AACTAA": 149, + "AACTAT": 150, + "AACTAC": 151, + "AACTAG": 152, + "AACTTA": 153, + "AACTTT": 154, + "AACTTC": 155, + "AACTTG": 156, + "AACTCA": 157, + "AACTCT": 158, + "AACTCC": 159, + "AACTCG": 160, + "AACTGA": 161, + "AACTGT": 162, + "AACTGC": 163, + "AACTGG": 164, + "AACCAA": 165, + "AACCAT": 166, + "AACCAC": 167, + "AACCAG": 168, + "AACCTA": 169, + "AACCTT": 170, + "AACCTC": 171, + "AACCTG": 172, + "AACCCA": 173, + "AACCCT": 174, + "AACCCC": 175, + "AACCCG": 176, + "AACCGA": 177, + "AACCGT": 178, + "AACCGC": 179, + "AACCGG": 180, + "AACGAA": 181, + "AACGAT": 182, + "AACGAC": 183, + "AACGAG": 184, + "AACGTA": 185, + "AACGTT": 186, + "AACGTC": 187, + "AACGTG": 188, + "AACGCA": 189, + "AACGCT": 190, + "AACGCC": 191, + "AACGCG": 192, + "AACGGA": 193, + "AACGGT": 194, + "AACGGC": 195, + "AACGGG": 196, + "AAGAAA": 197, + "AAGAAT": 198, + "AAGAAC": 199, + "AAGAAG": 200, + "AAGATA": 201, + "AAGATT": 202, + "AAGATC": 203, + "AAGATG": 204, + "AAGACA": 205, + "AAGACT": 206, + "AAGACC": 207, + "AAGACG": 208, + "AAGAGA": 209, + "AAGAGT": 210, + "AAGAGC": 211, + "AAGAGG": 212, + "AAGTAA": 213, + "AAGTAT": 214, + "AAGTAC": 215, + "AAGTAG": 216, + "AAGTTA": 217, + "AAGTTT": 218, + "AAGTTC": 219, + "AAGTTG": 220, + "AAGTCA": 221, + "AAGTCT": 222, + "AAGTCC": 223, + "AAGTCG": 224, + "AAGTGA": 225, + "AAGTGT": 226, + "AAGTGC": 227, + "AAGTGG": 228, + "AAGCAA": 229, + "AAGCAT": 230, + "AAGCAC": 231, + "AAGCAG": 232, + "AAGCTA": 233, + "AAGCTT": 234, + "AAGCTC": 235, + "AAGCTG": 236, + "AAGCCA": 237, + "AAGCCT": 238, + "AAGCCC": 239, + "AAGCCG": 240, + "AAGCGA": 241, + "AAGCGT": 242, + "AAGCGC": 243, + "AAGCGG": 244, + "AAGGAA": 245, + "AAGGAT": 246, + "AAGGAC": 247, + "AAGGAG": 248, + "AAGGTA": 249, + "AAGGTT": 250, + "AAGGTC": 251, + "AAGGTG": 252, + "AAGGCA": 253, + "AAGGCT": 254, + "AAGGCC": 255, + "AAGGCG": 256, + "AAGGGA": 257, + "AAGGGT": 258, + "AAGGGC": 259, + "AAGGGG": 260, + "ATAAAA": 261, + "ATAAAT": 262, + "ATAAAC": 263, + "ATAAAG": 264, + "ATAATA": 265, + "ATAATT": 266, + "ATAATC": 267, + "ATAATG": 268, + "ATAACA": 269, + "ATAACT": 270, + "ATAACC": 271, + "ATAACG": 272, + "ATAAGA": 273, + "ATAAGT": 274, + "ATAAGC": 275, + "ATAAGG": 276, + "ATATAA": 277, + "ATATAT": 278, + "ATATAC": 279, + "ATATAG": 280, + "ATATTA": 281, + "ATATTT": 282, + "ATATTC": 283, + "ATATTG": 284, + "ATATCA": 285, + "ATATCT": 286, + "ATATCC": 287, + "ATATCG": 288, + "ATATGA": 289, + "ATATGT": 290, + "ATATGC": 291, + "ATATGG": 292, + "ATACAA": 293, + "ATACAT": 294, + "ATACAC": 295, + "ATACAG": 296, + "ATACTA": 297, + "ATACTT": 298, + "ATACTC": 299, + "ATACTG": 300, + "ATACCA": 301, + "ATACCT": 302, + "ATACCC": 303, + "ATACCG": 304, + "ATACGA": 305, + "ATACGT": 306, + "ATACGC": 307, + "ATACGG": 308, + "ATAGAA": 309, + "ATAGAT": 310, + "ATAGAC": 311, + "ATAGAG": 312, + "ATAGTA": 313, + "ATAGTT": 314, + "ATAGTC": 315, + "ATAGTG": 316, + "ATAGCA": 317, + "ATAGCT": 318, + "ATAGCC": 319, + "ATAGCG": 320, + "ATAGGA": 321, + "ATAGGT": 322, + "ATAGGC": 323, + "ATAGGG": 324, + "ATTAAA": 325, + "ATTAAT": 326, + "ATTAAC": 327, + "ATTAAG": 328, + "ATTATA": 329, + "ATTATT": 330, + "ATTATC": 331, + "ATTATG": 332, + "ATTACA": 333, + "ATTACT": 334, + "ATTACC": 335, + "ATTACG": 336, + "ATTAGA": 337, + "ATTAGT": 338, + "ATTAGC": 339, + "ATTAGG": 340, + "ATTTAA": 341, + "ATTTAT": 342, + "ATTTAC": 343, + "ATTTAG": 344, + "ATTTTA": 345, + "ATTTTT": 346, + "ATTTTC": 347, + "ATTTTG": 348, + "ATTTCA": 349, + "ATTTCT": 350, + "ATTTCC": 351, + "ATTTCG": 352, + "ATTTGA": 353, + "ATTTGT": 354, + "ATTTGC": 355, + "ATTTGG": 356, + "ATTCAA": 357, + "ATTCAT": 358, + "ATTCAC": 359, + "ATTCAG": 360, + "ATTCTA": 361, + "ATTCTT": 362, + "ATTCTC": 363, + "ATTCTG": 364, + "ATTCCA": 365, + "ATTCCT": 366, + "ATTCCC": 367, + "ATTCCG": 368, + "ATTCGA": 369, + "ATTCGT": 370, + "ATTCGC": 371, + "ATTCGG": 372, + "ATTGAA": 373, + "ATTGAT": 374, + "ATTGAC": 375, + "ATTGAG": 376, + "ATTGTA": 377, + "ATTGTT": 378, + "ATTGTC": 379, + "ATTGTG": 380, + "ATTGCA": 381, + "ATTGCT": 382, + "ATTGCC": 383, + "ATTGCG": 384, + "ATTGGA": 385, + "ATTGGT": 386, + "ATTGGC": 387, + "ATTGGG": 388, + "ATCAAA": 389, + "ATCAAT": 390, + "ATCAAC": 391, + "ATCAAG": 392, + "ATCATA": 393, + "ATCATT": 394, + "ATCATC": 395, + "ATCATG": 396, + "ATCACA": 397, + "ATCACT": 398, + "ATCACC": 399, + "ATCACG": 400, + "ATCAGA": 401, + "ATCAGT": 402, + "ATCAGC": 403, + "ATCAGG": 404, + "ATCTAA": 405, + "ATCTAT": 406, + "ATCTAC": 407, + "ATCTAG": 408, + "ATCTTA": 409, + "ATCTTT": 410, + "ATCTTC": 411, + "ATCTTG": 412, + "ATCTCA": 413, + "ATCTCT": 414, + "ATCTCC": 415, + "ATCTCG": 416, + "ATCTGA": 417, + "ATCTGT": 418, + "ATCTGC": 419, + "ATCTGG": 420, + "ATCCAA": 421, + "ATCCAT": 422, + "ATCCAC": 423, + "ATCCAG": 424, + "ATCCTA": 425, + "ATCCTT": 426, + "ATCCTC": 427, + "ATCCTG": 428, + "ATCCCA": 429, + "ATCCCT": 430, + "ATCCCC": 431, + "ATCCCG": 432, + "ATCCGA": 433, + "ATCCGT": 434, + "ATCCGC": 435, + "ATCCGG": 436, + "ATCGAA": 437, + "ATCGAT": 438, + "ATCGAC": 439, + "ATCGAG": 440, + "ATCGTA": 441, + "ATCGTT": 442, + "ATCGTC": 443, + "ATCGTG": 444, + "ATCGCA": 445, + "ATCGCT": 446, + "ATCGCC": 447, + "ATCGCG": 448, + "ATCGGA": 449, + "ATCGGT": 450, + "ATCGGC": 451, + "ATCGGG": 452, + "ATGAAA": 453, + "ATGAAT": 454, + "ATGAAC": 455, + "ATGAAG": 456, + "ATGATA": 457, + "ATGATT": 458, + "ATGATC": 459, + "ATGATG": 460, + "ATGACA": 461, + "ATGACT": 462, + "ATGACC": 463, + "ATGACG": 464, + "ATGAGA": 465, + "ATGAGT": 466, + "ATGAGC": 467, + "ATGAGG": 468, + "ATGTAA": 469, + "ATGTAT": 470, + "ATGTAC": 471, + "ATGTAG": 472, + "ATGTTA": 473, + "ATGTTT": 474, + "ATGTTC": 475, + "ATGTTG": 476, + "ATGTCA": 477, + "ATGTCT": 478, + "ATGTCC": 479, + "ATGTCG": 480, + "ATGTGA": 481, + "ATGTGT": 482, + "ATGTGC": 483, + "ATGTGG": 484, + "ATGCAA": 485, + "ATGCAT": 486, + "ATGCAC": 487, + "ATGCAG": 488, + "ATGCTA": 489, + "ATGCTT": 490, + "ATGCTC": 491, + "ATGCTG": 492, + "ATGCCA": 493, + "ATGCCT": 494, + "ATGCCC": 495, + "ATGCCG": 496, + "ATGCGA": 497, + "ATGCGT": 498, + "ATGCGC": 499, + "ATGCGG": 500, + "ATGGAA": 501, + "ATGGAT": 502, + "ATGGAC": 503, + "ATGGAG": 504, + "ATGGTA": 505, + "ATGGTT": 506, + "ATGGTC": 507, + "ATGGTG": 508, + "ATGGCA": 509, + "ATGGCT": 510, + "ATGGCC": 511, + "ATGGCG": 512, + "ATGGGA": 513, + "ATGGGT": 514, + "ATGGGC": 515, + "ATGGGG": 516, + "ACAAAA": 517, + "ACAAAT": 518, + "ACAAAC": 519, + "ACAAAG": 520, + "ACAATA": 521, + "ACAATT": 522, + "ACAATC": 523, + "ACAATG": 524, + "ACAACA": 525, + "ACAACT": 526, + "ACAACC": 527, + "ACAACG": 528, + "ACAAGA": 529, + "ACAAGT": 530, + "ACAAGC": 531, + "ACAAGG": 532, + "ACATAA": 533, + "ACATAT": 534, + "ACATAC": 535, + "ACATAG": 536, + "ACATTA": 537, + "ACATTT": 538, + "ACATTC": 539, + "ACATTG": 540, + "ACATCA": 541, + "ACATCT": 542, + "ACATCC": 543, + "ACATCG": 544, + "ACATGA": 545, + "ACATGT": 546, + "ACATGC": 547, + "ACATGG": 548, + "ACACAA": 549, + "ACACAT": 550, + "ACACAC": 551, + "ACACAG": 552, + "ACACTA": 553, + "ACACTT": 554, + "ACACTC": 555, + "ACACTG": 556, + "ACACCA": 557, + "ACACCT": 558, + "ACACCC": 559, + "ACACCG": 560, + "ACACGA": 561, + "ACACGT": 562, + "ACACGC": 563, + "ACACGG": 564, + "ACAGAA": 565, + "ACAGAT": 566, + "ACAGAC": 567, + "ACAGAG": 568, + "ACAGTA": 569, + "ACAGTT": 570, + "ACAGTC": 571, + "ACAGTG": 572, + "ACAGCA": 573, + "ACAGCT": 574, + "ACAGCC": 575, + "ACAGCG": 576, + "ACAGGA": 577, + "ACAGGT": 578, + "ACAGGC": 579, + "ACAGGG": 580, + "ACTAAA": 581, + "ACTAAT": 582, + "ACTAAC": 583, + "ACTAAG": 584, + "ACTATA": 585, + "ACTATT": 586, + "ACTATC": 587, + "ACTATG": 588, + "ACTACA": 589, + "ACTACT": 590, + "ACTACC": 591, + "ACTACG": 592, + "ACTAGA": 593, + "ACTAGT": 594, + "ACTAGC": 595, + "ACTAGG": 596, + "ACTTAA": 597, + "ACTTAT": 598, + "ACTTAC": 599, + "ACTTAG": 600, + "ACTTTA": 601, + "ACTTTT": 602, + "ACTTTC": 603, + "ACTTTG": 604, + "ACTTCA": 605, + "ACTTCT": 606, + "ACTTCC": 607, + "ACTTCG": 608, + "ACTTGA": 609, + "ACTTGT": 610, + "ACTTGC": 611, + "ACTTGG": 612, + "ACTCAA": 613, + "ACTCAT": 614, + "ACTCAC": 615, + "ACTCAG": 616, + "ACTCTA": 617, + "ACTCTT": 618, + "ACTCTC": 619, + "ACTCTG": 620, + "ACTCCA": 621, + "ACTCCT": 622, + "ACTCCC": 623, + "ACTCCG": 624, + "ACTCGA": 625, + "ACTCGT": 626, + "ACTCGC": 627, + "ACTCGG": 628, + "ACTGAA": 629, + "ACTGAT": 630, + "ACTGAC": 631, + "ACTGAG": 632, + "ACTGTA": 633, + "ACTGTT": 634, + "ACTGTC": 635, + "ACTGTG": 636, + "ACTGCA": 637, + "ACTGCT": 638, + "ACTGCC": 639, + "ACTGCG": 640, + "ACTGGA": 641, + "ACTGGT": 642, + "ACTGGC": 643, + "ACTGGG": 644, + "ACCAAA": 645, + "ACCAAT": 646, + "ACCAAC": 647, + "ACCAAG": 648, + "ACCATA": 649, + "ACCATT": 650, + "ACCATC": 651, + "ACCATG": 652, + "ACCACA": 653, + "ACCACT": 654, + "ACCACC": 655, + "ACCACG": 656, + "ACCAGA": 657, + "ACCAGT": 658, + "ACCAGC": 659, + "ACCAGG": 660, + "ACCTAA": 661, + "ACCTAT": 662, + "ACCTAC": 663, + "ACCTAG": 664, + "ACCTTA": 665, + "ACCTTT": 666, + "ACCTTC": 667, + "ACCTTG": 668, + "ACCTCA": 669, + "ACCTCT": 670, + "ACCTCC": 671, + "ACCTCG": 672, + "ACCTGA": 673, + "ACCTGT": 674, + "ACCTGC": 675, + "ACCTGG": 676, + "ACCCAA": 677, + "ACCCAT": 678, + "ACCCAC": 679, + "ACCCAG": 680, + "ACCCTA": 681, + "ACCCTT": 682, + "ACCCTC": 683, + "ACCCTG": 684, + "ACCCCA": 685, + "ACCCCT": 686, + "ACCCCC": 687, + "ACCCCG": 688, + "ACCCGA": 689, + "ACCCGT": 690, + "ACCCGC": 691, + "ACCCGG": 692, + "ACCGAA": 693, + "ACCGAT": 694, + "ACCGAC": 695, + "ACCGAG": 696, + "ACCGTA": 697, + "ACCGTT": 698, + "ACCGTC": 699, + "ACCGTG": 700, + "ACCGCA": 701, + "ACCGCT": 702, + "ACCGCC": 703, + "ACCGCG": 704, + "ACCGGA": 705, + "ACCGGT": 706, + "ACCGGC": 707, + "ACCGGG": 708, + "ACGAAA": 709, + "ACGAAT": 710, + "ACGAAC": 711, + "ACGAAG": 712, + "ACGATA": 713, + "ACGATT": 714, + "ACGATC": 715, + "ACGATG": 716, + "ACGACA": 717, + "ACGACT": 718, + "ACGACC": 719, + "ACGACG": 720, + "ACGAGA": 721, + "ACGAGT": 722, + "ACGAGC": 723, + "ACGAGG": 724, + "ACGTAA": 725, + "ACGTAT": 726, + "ACGTAC": 727, + "ACGTAG": 728, + "ACGTTA": 729, + "ACGTTT": 730, + "ACGTTC": 731, + "ACGTTG": 732, + "ACGTCA": 733, + "ACGTCT": 734, + "ACGTCC": 735, + "ACGTCG": 736, + "ACGTGA": 737, + "ACGTGT": 738, + "ACGTGC": 739, + "ACGTGG": 740, + "ACGCAA": 741, + "ACGCAT": 742, + "ACGCAC": 743, + "ACGCAG": 744, + "ACGCTA": 745, + "ACGCTT": 746, + "ACGCTC": 747, + "ACGCTG": 748, + "ACGCCA": 749, + "ACGCCT": 750, + "ACGCCC": 751, + "ACGCCG": 752, + "ACGCGA": 753, + "ACGCGT": 754, + "ACGCGC": 755, + "ACGCGG": 756, + "ACGGAA": 757, + "ACGGAT": 758, + "ACGGAC": 759, + "ACGGAG": 760, + "ACGGTA": 761, + "ACGGTT": 762, + "ACGGTC": 763, + "ACGGTG": 764, + "ACGGCA": 765, + "ACGGCT": 766, + "ACGGCC": 767, + "ACGGCG": 768, + "ACGGGA": 769, + "ACGGGT": 770, + "ACGGGC": 771, + "ACGGGG": 772, + "AGAAAA": 773, + "AGAAAT": 774, + "AGAAAC": 775, + "AGAAAG": 776, + "AGAATA": 777, + "AGAATT": 778, + "AGAATC": 779, + "AGAATG": 780, + "AGAACA": 781, + "AGAACT": 782, + "AGAACC": 783, + "AGAACG": 784, + "AGAAGA": 785, + "AGAAGT": 786, + "AGAAGC": 787, + "AGAAGG": 788, + "AGATAA": 789, + "AGATAT": 790, + "AGATAC": 791, + "AGATAG": 792, + "AGATTA": 793, + "AGATTT": 794, + "AGATTC": 795, + "AGATTG": 796, + "AGATCA": 797, + "AGATCT": 798, + "AGATCC": 799, + "AGATCG": 800, + "AGATGA": 801, + "AGATGT": 802, + "AGATGC": 803, + "AGATGG": 804, + "AGACAA": 805, + "AGACAT": 806, + "AGACAC": 807, + "AGACAG": 808, + "AGACTA": 809, + "AGACTT": 810, + "AGACTC": 811, + "AGACTG": 812, + "AGACCA": 813, + "AGACCT": 814, + "AGACCC": 815, + "AGACCG": 816, + "AGACGA": 817, + "AGACGT": 818, + "AGACGC": 819, + "AGACGG": 820, + "AGAGAA": 821, + "AGAGAT": 822, + "AGAGAC": 823, + "AGAGAG": 824, + "AGAGTA": 825, + "AGAGTT": 826, + "AGAGTC": 827, + "AGAGTG": 828, + "AGAGCA": 829, + "AGAGCT": 830, + "AGAGCC": 831, + "AGAGCG": 832, + "AGAGGA": 833, + "AGAGGT": 834, + "AGAGGC": 835, + "AGAGGG": 836, + "AGTAAA": 837, + "AGTAAT": 838, + "AGTAAC": 839, + "AGTAAG": 840, + "AGTATA": 841, + "AGTATT": 842, + "AGTATC": 843, + "AGTATG": 844, + "AGTACA": 845, + "AGTACT": 846, + "AGTACC": 847, + "AGTACG": 848, + "AGTAGA": 849, + "AGTAGT": 850, + "AGTAGC": 851, + "AGTAGG": 852, + "AGTTAA": 853, + "AGTTAT": 854, + "AGTTAC": 855, + "AGTTAG": 856, + "AGTTTA": 857, + "AGTTTT": 858, + "AGTTTC": 859, + "AGTTTG": 860, + "AGTTCA": 861, + "AGTTCT": 862, + "AGTTCC": 863, + "AGTTCG": 864, + "AGTTGA": 865, + "AGTTGT": 866, + "AGTTGC": 867, + "AGTTGG": 868, + "AGTCAA": 869, + "AGTCAT": 870, + "AGTCAC": 871, + "AGTCAG": 872, + "AGTCTA": 873, + "AGTCTT": 874, + "AGTCTC": 875, + "AGTCTG": 876, + "AGTCCA": 877, + "AGTCCT": 878, + "AGTCCC": 879, + "AGTCCG": 880, + "AGTCGA": 881, + "AGTCGT": 882, + "AGTCGC": 883, + "AGTCGG": 884, + "AGTGAA": 885, + "AGTGAT": 886, + "AGTGAC": 887, + "AGTGAG": 888, + "AGTGTA": 889, + "AGTGTT": 890, + "AGTGTC": 891, + "AGTGTG": 892, + "AGTGCA": 893, + "AGTGCT": 894, + "AGTGCC": 895, + "AGTGCG": 896, + "AGTGGA": 897, + "AGTGGT": 898, + "AGTGGC": 899, + "AGTGGG": 900, + "AGCAAA": 901, + "AGCAAT": 902, + "AGCAAC": 903, + "AGCAAG": 904, + "AGCATA": 905, + "AGCATT": 906, + "AGCATC": 907, + "AGCATG": 908, + "AGCACA": 909, + "AGCACT": 910, + "AGCACC": 911, + "AGCACG": 912, + "AGCAGA": 913, + "AGCAGT": 914, + "AGCAGC": 915, + "AGCAGG": 916, + "AGCTAA": 917, + "AGCTAT": 918, + "AGCTAC": 919, + "AGCTAG": 920, + "AGCTTA": 921, + "AGCTTT": 922, + "AGCTTC": 923, + "AGCTTG": 924, + "AGCTCA": 925, + "AGCTCT": 926, + "AGCTCC": 927, + "AGCTCG": 928, + "AGCTGA": 929, + "AGCTGT": 930, + "AGCTGC": 931, + "AGCTGG": 932, + "AGCCAA": 933, + "AGCCAT": 934, + "AGCCAC": 935, + "AGCCAG": 936, + "AGCCTA": 937, + "AGCCTT": 938, + "AGCCTC": 939, + "AGCCTG": 940, + "AGCCCA": 941, + "AGCCCT": 942, + "AGCCCC": 943, + "AGCCCG": 944, + "AGCCGA": 945, + "AGCCGT": 946, + "AGCCGC": 947, + "AGCCGG": 948, + "AGCGAA": 949, + "AGCGAT": 950, + "AGCGAC": 951, + "AGCGAG": 952, + "AGCGTA": 953, + "AGCGTT": 954, + "AGCGTC": 955, + "AGCGTG": 956, + "AGCGCA": 957, + "AGCGCT": 958, + "AGCGCC": 959, + "AGCGCG": 960, + "AGCGGA": 961, + "AGCGGT": 962, + "AGCGGC": 963, + "AGCGGG": 964, + "AGGAAA": 965, + "AGGAAT": 966, + "AGGAAC": 967, + "AGGAAG": 968, + "AGGATA": 969, + "AGGATT": 970, + "AGGATC": 971, + "AGGATG": 972, + "AGGACA": 973, + "AGGACT": 974, + "AGGACC": 975, + "AGGACG": 976, + "AGGAGA": 977, + "AGGAGT": 978, + "AGGAGC": 979, + "AGGAGG": 980, + "AGGTAA": 981, + "AGGTAT": 982, + "AGGTAC": 983, + "AGGTAG": 984, + "AGGTTA": 985, + "AGGTTT": 986, + "AGGTTC": 987, + "AGGTTG": 988, + "AGGTCA": 989, + "AGGTCT": 990, + "AGGTCC": 991, + "AGGTCG": 992, + "AGGTGA": 993, + "AGGTGT": 994, + "AGGTGC": 995, + "AGGTGG": 996, + "AGGCAA": 997, + "AGGCAT": 998, + "AGGCAC": 999, + "AGGCAG": 1000, + "AGGCTA": 1001, + "AGGCTT": 1002, + "AGGCTC": 1003, + "AGGCTG": 1004, + "AGGCCA": 1005, + "AGGCCT": 1006, + "AGGCCC": 1007, + "AGGCCG": 1008, + "AGGCGA": 1009, + "AGGCGT": 1010, + "AGGCGC": 1011, + "AGGCGG": 1012, + "AGGGAA": 1013, + "AGGGAT": 1014, + "AGGGAC": 1015, + "AGGGAG": 1016, + "AGGGTA": 1017, + "AGGGTT": 1018, + "AGGGTC": 1019, + "AGGGTG": 1020, + "AGGGCA": 1021, + "AGGGCT": 1022, + "AGGGCC": 1023, + "AGGGCG": 1024, + "AGGGGA": 1025, + "AGGGGT": 1026, + "AGGGGC": 1027, + "AGGGGG": 1028, + "TAAAAA": 1029, + "TAAAAT": 1030, + "TAAAAC": 1031, + "TAAAAG": 1032, + "TAAATA": 1033, + "TAAATT": 1034, + "TAAATC": 1035, + "TAAATG": 1036, + "TAAACA": 1037, + "TAAACT": 1038, + "TAAACC": 1039, + "TAAACG": 1040, + "TAAAGA": 1041, + "TAAAGT": 1042, + "TAAAGC": 1043, + "TAAAGG": 1044, + "TAATAA": 1045, + "TAATAT": 1046, + "TAATAC": 1047, + "TAATAG": 1048, + "TAATTA": 1049, + "TAATTT": 1050, + "TAATTC": 1051, + "TAATTG": 1052, + "TAATCA": 1053, + "TAATCT": 1054, + "TAATCC": 1055, + "TAATCG": 1056, + "TAATGA": 1057, + "TAATGT": 1058, + "TAATGC": 1059, + "TAATGG": 1060, + "TAACAA": 1061, + "TAACAT": 1062, + "TAACAC": 1063, + "TAACAG": 1064, + "TAACTA": 1065, + "TAACTT": 1066, + "TAACTC": 1067, + "TAACTG": 1068, + "TAACCA": 1069, + "TAACCT": 1070, + "TAACCC": 1071, + "TAACCG": 1072, + "TAACGA": 1073, + "TAACGT": 1074, + "TAACGC": 1075, + "TAACGG": 1076, + "TAAGAA": 1077, + "TAAGAT": 1078, + "TAAGAC": 1079, + "TAAGAG": 1080, + "TAAGTA": 1081, + "TAAGTT": 1082, + "TAAGTC": 1083, + "TAAGTG": 1084, + "TAAGCA": 1085, + "TAAGCT": 1086, + "TAAGCC": 1087, + "TAAGCG": 1088, + "TAAGGA": 1089, + "TAAGGT": 1090, + "TAAGGC": 1091, + "TAAGGG": 1092, + "TATAAA": 1093, + "TATAAT": 1094, + "TATAAC": 1095, + "TATAAG": 1096, + "TATATA": 1097, + "TATATT": 1098, + "TATATC": 1099, + "TATATG": 1100, + "TATACA": 1101, + "TATACT": 1102, + "TATACC": 1103, + "TATACG": 1104, + "TATAGA": 1105, + "TATAGT": 1106, + "TATAGC": 1107, + "TATAGG": 1108, + "TATTAA": 1109, + "TATTAT": 1110, + "TATTAC": 1111, + "TATTAG": 1112, + "TATTTA": 1113, + "TATTTT": 1114, + "TATTTC": 1115, + "TATTTG": 1116, + "TATTCA": 1117, + "TATTCT": 1118, + "TATTCC": 1119, + "TATTCG": 1120, + "TATTGA": 1121, + "TATTGT": 1122, + "TATTGC": 1123, + "TATTGG": 1124, + "TATCAA": 1125, + "TATCAT": 1126, + "TATCAC": 1127, + "TATCAG": 1128, + "TATCTA": 1129, + "TATCTT": 1130, + "TATCTC": 1131, + "TATCTG": 1132, + "TATCCA": 1133, + "TATCCT": 1134, + "TATCCC": 1135, + "TATCCG": 1136, + "TATCGA": 1137, + "TATCGT": 1138, + "TATCGC": 1139, + "TATCGG": 1140, + "TATGAA": 1141, + "TATGAT": 1142, + "TATGAC": 1143, + "TATGAG": 1144, + "TATGTA": 1145, + "TATGTT": 1146, + "TATGTC": 1147, + "TATGTG": 1148, + "TATGCA": 1149, + "TATGCT": 1150, + "TATGCC": 1151, + "TATGCG": 1152, + "TATGGA": 1153, + "TATGGT": 1154, + "TATGGC": 1155, + "TATGGG": 1156, + "TACAAA": 1157, + "TACAAT": 1158, + "TACAAC": 1159, + "TACAAG": 1160, + "TACATA": 1161, + "TACATT": 1162, + "TACATC": 1163, + "TACATG": 1164, + "TACACA": 1165, + "TACACT": 1166, + "TACACC": 1167, + "TACACG": 1168, + "TACAGA": 1169, + "TACAGT": 1170, + "TACAGC": 1171, + "TACAGG": 1172, + "TACTAA": 1173, + "TACTAT": 1174, + "TACTAC": 1175, + "TACTAG": 1176, + "TACTTA": 1177, + "TACTTT": 1178, + "TACTTC": 1179, + "TACTTG": 1180, + "TACTCA": 1181, + "TACTCT": 1182, + "TACTCC": 1183, + "TACTCG": 1184, + "TACTGA": 1185, + "TACTGT": 1186, + "TACTGC": 1187, + "TACTGG": 1188, + "TACCAA": 1189, + "TACCAT": 1190, + "TACCAC": 1191, + "TACCAG": 1192, + "TACCTA": 1193, + "TACCTT": 1194, + "TACCTC": 1195, + "TACCTG": 1196, + "TACCCA": 1197, + "TACCCT": 1198, + "TACCCC": 1199, + "TACCCG": 1200, + "TACCGA": 1201, + "TACCGT": 1202, + "TACCGC": 1203, + "TACCGG": 1204, + "TACGAA": 1205, + "TACGAT": 1206, + "TACGAC": 1207, + "TACGAG": 1208, + "TACGTA": 1209, + "TACGTT": 1210, + "TACGTC": 1211, + "TACGTG": 1212, + "TACGCA": 1213, + "TACGCT": 1214, + "TACGCC": 1215, + "TACGCG": 1216, + "TACGGA": 1217, + "TACGGT": 1218, + "TACGGC": 1219, + "TACGGG": 1220, + "TAGAAA": 1221, + "TAGAAT": 1222, + "TAGAAC": 1223, + "TAGAAG": 1224, + "TAGATA": 1225, + "TAGATT": 1226, + "TAGATC": 1227, + "TAGATG": 1228, + "TAGACA": 1229, + "TAGACT": 1230, + "TAGACC": 1231, + "TAGACG": 1232, + "TAGAGA": 1233, + "TAGAGT": 1234, + "TAGAGC": 1235, + "TAGAGG": 1236, + "TAGTAA": 1237, + "TAGTAT": 1238, + "TAGTAC": 1239, + "TAGTAG": 1240, + "TAGTTA": 1241, + "TAGTTT": 1242, + "TAGTTC": 1243, + "TAGTTG": 1244, + "TAGTCA": 1245, + "TAGTCT": 1246, + "TAGTCC": 1247, + "TAGTCG": 1248, + "TAGTGA": 1249, + "TAGTGT": 1250, + "TAGTGC": 1251, + "TAGTGG": 1252, + "TAGCAA": 1253, + "TAGCAT": 1254, + "TAGCAC": 1255, + "TAGCAG": 1256, + "TAGCTA": 1257, + "TAGCTT": 1258, + "TAGCTC": 1259, + "TAGCTG": 1260, + "TAGCCA": 1261, + "TAGCCT": 1262, + "TAGCCC": 1263, + "TAGCCG": 1264, + "TAGCGA": 1265, + "TAGCGT": 1266, + "TAGCGC": 1267, + "TAGCGG": 1268, + "TAGGAA": 1269, + "TAGGAT": 1270, + "TAGGAC": 1271, + "TAGGAG": 1272, + "TAGGTA": 1273, + "TAGGTT": 1274, + "TAGGTC": 1275, + "TAGGTG": 1276, + "TAGGCA": 1277, + "TAGGCT": 1278, + "TAGGCC": 1279, + "TAGGCG": 1280, + "TAGGGA": 1281, + "TAGGGT": 1282, + "TAGGGC": 1283, + "TAGGGG": 1284, + "TTAAAA": 1285, + "TTAAAT": 1286, + "TTAAAC": 1287, + "TTAAAG": 1288, + "TTAATA": 1289, + "TTAATT": 1290, + "TTAATC": 1291, + "TTAATG": 1292, + "TTAACA": 1293, + "TTAACT": 1294, + "TTAACC": 1295, + "TTAACG": 1296, + "TTAAGA": 1297, + "TTAAGT": 1298, + "TTAAGC": 1299, + "TTAAGG": 1300, + "TTATAA": 1301, + "TTATAT": 1302, + "TTATAC": 1303, + "TTATAG": 1304, + "TTATTA": 1305, + "TTATTT": 1306, + "TTATTC": 1307, + "TTATTG": 1308, + "TTATCA": 1309, + "TTATCT": 1310, + "TTATCC": 1311, + "TTATCG": 1312, + "TTATGA": 1313, + "TTATGT": 1314, + "TTATGC": 1315, + "TTATGG": 1316, + "TTACAA": 1317, + "TTACAT": 1318, + "TTACAC": 1319, + "TTACAG": 1320, + "TTACTA": 1321, + "TTACTT": 1322, + "TTACTC": 1323, + "TTACTG": 1324, + "TTACCA": 1325, + "TTACCT": 1326, + "TTACCC": 1327, + "TTACCG": 1328, + "TTACGA": 1329, + "TTACGT": 1330, + "TTACGC": 1331, + "TTACGG": 1332, + "TTAGAA": 1333, + "TTAGAT": 1334, + "TTAGAC": 1335, + "TTAGAG": 1336, + "TTAGTA": 1337, + "TTAGTT": 1338, + "TTAGTC": 1339, + "TTAGTG": 1340, + "TTAGCA": 1341, + "TTAGCT": 1342, + "TTAGCC": 1343, + "TTAGCG": 1344, + "TTAGGA": 1345, + "TTAGGT": 1346, + "TTAGGC": 1347, + "TTAGGG": 1348, + "TTTAAA": 1349, + "TTTAAT": 1350, + "TTTAAC": 1351, + "TTTAAG": 1352, + "TTTATA": 1353, + "TTTATT": 1354, + "TTTATC": 1355, + "TTTATG": 1356, + "TTTACA": 1357, + "TTTACT": 1358, + "TTTACC": 1359, + "TTTACG": 1360, + "TTTAGA": 1361, + "TTTAGT": 1362, + "TTTAGC": 1363, + "TTTAGG": 1364, + "TTTTAA": 1365, + "TTTTAT": 1366, + "TTTTAC": 1367, + "TTTTAG": 1368, + "TTTTTA": 1369, + "TTTTTT": 1370, + "TTTTTC": 1371, + "TTTTTG": 1372, + "TTTTCA": 1373, + "TTTTCT": 1374, + "TTTTCC": 1375, + "TTTTCG": 1376, + "TTTTGA": 1377, + "TTTTGT": 1378, + "TTTTGC": 1379, + "TTTTGG": 1380, + "TTTCAA": 1381, + "TTTCAT": 1382, + "TTTCAC": 1383, + "TTTCAG": 1384, + "TTTCTA": 1385, + "TTTCTT": 1386, + "TTTCTC": 1387, + "TTTCTG": 1388, + "TTTCCA": 1389, + "TTTCCT": 1390, + "TTTCCC": 1391, + "TTTCCG": 1392, + "TTTCGA": 1393, + "TTTCGT": 1394, + "TTTCGC": 1395, + "TTTCGG": 1396, + "TTTGAA": 1397, + "TTTGAT": 1398, + "TTTGAC": 1399, + "TTTGAG": 1400, + "TTTGTA": 1401, + "TTTGTT": 1402, + "TTTGTC": 1403, + "TTTGTG": 1404, + "TTTGCA": 1405, + "TTTGCT": 1406, + "TTTGCC": 1407, + "TTTGCG": 1408, + "TTTGGA": 1409, + "TTTGGT": 1410, + "TTTGGC": 1411, + "TTTGGG": 1412, + "TTCAAA": 1413, + "TTCAAT": 1414, + "TTCAAC": 1415, + "TTCAAG": 1416, + "TTCATA": 1417, + "TTCATT": 1418, + "TTCATC": 1419, + "TTCATG": 1420, + "TTCACA": 1421, + "TTCACT": 1422, + "TTCACC": 1423, + "TTCACG": 1424, + "TTCAGA": 1425, + "TTCAGT": 1426, + "TTCAGC": 1427, + "TTCAGG": 1428, + "TTCTAA": 1429, + "TTCTAT": 1430, + "TTCTAC": 1431, + "TTCTAG": 1432, + "TTCTTA": 1433, + "TTCTTT": 1434, + "TTCTTC": 1435, + "TTCTTG": 1436, + "TTCTCA": 1437, + "TTCTCT": 1438, + "TTCTCC": 1439, + "TTCTCG": 1440, + "TTCTGA": 1441, + "TTCTGT": 1442, + "TTCTGC": 1443, + "TTCTGG": 1444, + "TTCCAA": 1445, + "TTCCAT": 1446, + "TTCCAC": 1447, + "TTCCAG": 1448, + "TTCCTA": 1449, + "TTCCTT": 1450, + "TTCCTC": 1451, + "TTCCTG": 1452, + "TTCCCA": 1453, + "TTCCCT": 1454, + "TTCCCC": 1455, + "TTCCCG": 1456, + "TTCCGA": 1457, + "TTCCGT": 1458, + "TTCCGC": 1459, + "TTCCGG": 1460, + "TTCGAA": 1461, + "TTCGAT": 1462, + "TTCGAC": 1463, + "TTCGAG": 1464, + "TTCGTA": 1465, + "TTCGTT": 1466, + "TTCGTC": 1467, + "TTCGTG": 1468, + "TTCGCA": 1469, + "TTCGCT": 1470, + "TTCGCC": 1471, + "TTCGCG": 1472, + "TTCGGA": 1473, + "TTCGGT": 1474, + "TTCGGC": 1475, + "TTCGGG": 1476, + "TTGAAA": 1477, + "TTGAAT": 1478, + "TTGAAC": 1479, + "TTGAAG": 1480, + "TTGATA": 1481, + "TTGATT": 1482, + "TTGATC": 1483, + "TTGATG": 1484, + "TTGACA": 1485, + "TTGACT": 1486, + "TTGACC": 1487, + "TTGACG": 1488, + "TTGAGA": 1489, + "TTGAGT": 1490, + "TTGAGC": 1491, + "TTGAGG": 1492, + "TTGTAA": 1493, + "TTGTAT": 1494, + "TTGTAC": 1495, + "TTGTAG": 1496, + "TTGTTA": 1497, + "TTGTTT": 1498, + "TTGTTC": 1499, + "TTGTTG": 1500, + "TTGTCA": 1501, + "TTGTCT": 1502, + "TTGTCC": 1503, + "TTGTCG": 1504, + "TTGTGA": 1505, + "TTGTGT": 1506, + "TTGTGC": 1507, + "TTGTGG": 1508, + "TTGCAA": 1509, + "TTGCAT": 1510, + "TTGCAC": 1511, + "TTGCAG": 1512, + "TTGCTA": 1513, + "TTGCTT": 1514, + "TTGCTC": 1515, + "TTGCTG": 1516, + "TTGCCA": 1517, + "TTGCCT": 1518, + "TTGCCC": 1519, + "TTGCCG": 1520, + "TTGCGA": 1521, + "TTGCGT": 1522, + "TTGCGC": 1523, + "TTGCGG": 1524, + "TTGGAA": 1525, + "TTGGAT": 1526, + "TTGGAC": 1527, + "TTGGAG": 1528, + "TTGGTA": 1529, + "TTGGTT": 1530, + "TTGGTC": 1531, + "TTGGTG": 1532, + "TTGGCA": 1533, + "TTGGCT": 1534, + "TTGGCC": 1535, + "TTGGCG": 1536, + "TTGGGA": 1537, + "TTGGGT": 1538, + "TTGGGC": 1539, + "TTGGGG": 1540, + "TCAAAA": 1541, + "TCAAAT": 1542, + "TCAAAC": 1543, + "TCAAAG": 1544, + "TCAATA": 1545, + "TCAATT": 1546, + "TCAATC": 1547, + "TCAATG": 1548, + "TCAACA": 1549, + "TCAACT": 1550, + "TCAACC": 1551, + "TCAACG": 1552, + "TCAAGA": 1553, + "TCAAGT": 1554, + "TCAAGC": 1555, + "TCAAGG": 1556, + "TCATAA": 1557, + "TCATAT": 1558, + "TCATAC": 1559, + "TCATAG": 1560, + "TCATTA": 1561, + "TCATTT": 1562, + "TCATTC": 1563, + "TCATTG": 1564, + "TCATCA": 1565, + "TCATCT": 1566, + "TCATCC": 1567, + "TCATCG": 1568, + "TCATGA": 1569, + "TCATGT": 1570, + "TCATGC": 1571, + "TCATGG": 1572, + "TCACAA": 1573, + "TCACAT": 1574, + "TCACAC": 1575, + "TCACAG": 1576, + "TCACTA": 1577, + "TCACTT": 1578, + "TCACTC": 1579, + "TCACTG": 1580, + "TCACCA": 1581, + "TCACCT": 1582, + "TCACCC": 1583, + "TCACCG": 1584, + "TCACGA": 1585, + "TCACGT": 1586, + "TCACGC": 1587, + "TCACGG": 1588, + "TCAGAA": 1589, + "TCAGAT": 1590, + "TCAGAC": 1591, + "TCAGAG": 1592, + "TCAGTA": 1593, + "TCAGTT": 1594, + "TCAGTC": 1595, + "TCAGTG": 1596, + "TCAGCA": 1597, + "TCAGCT": 1598, + "TCAGCC": 1599, + "TCAGCG": 1600, + "TCAGGA": 1601, + "TCAGGT": 1602, + "TCAGGC": 1603, + "TCAGGG": 1604, + "TCTAAA": 1605, + "TCTAAT": 1606, + "TCTAAC": 1607, + "TCTAAG": 1608, + "TCTATA": 1609, + "TCTATT": 1610, + "TCTATC": 1611, + "TCTATG": 1612, + "TCTACA": 1613, + "TCTACT": 1614, + "TCTACC": 1615, + "TCTACG": 1616, + "TCTAGA": 1617, + "TCTAGT": 1618, + "TCTAGC": 1619, + "TCTAGG": 1620, + "TCTTAA": 1621, + "TCTTAT": 1622, + "TCTTAC": 1623, + "TCTTAG": 1624, + "TCTTTA": 1625, + "TCTTTT": 1626, + "TCTTTC": 1627, + "TCTTTG": 1628, + "TCTTCA": 1629, + "TCTTCT": 1630, + "TCTTCC": 1631, + "TCTTCG": 1632, + "TCTTGA": 1633, + "TCTTGT": 1634, + "TCTTGC": 1635, + "TCTTGG": 1636, + "TCTCAA": 1637, + "TCTCAT": 1638, + "TCTCAC": 1639, + "TCTCAG": 1640, + "TCTCTA": 1641, + "TCTCTT": 1642, + "TCTCTC": 1643, + "TCTCTG": 1644, + "TCTCCA": 1645, + "TCTCCT": 1646, + "TCTCCC": 1647, + "TCTCCG": 1648, + "TCTCGA": 1649, + "TCTCGT": 1650, + "TCTCGC": 1651, + "TCTCGG": 1652, + "TCTGAA": 1653, + "TCTGAT": 1654, + "TCTGAC": 1655, + "TCTGAG": 1656, + "TCTGTA": 1657, + "TCTGTT": 1658, + "TCTGTC": 1659, + "TCTGTG": 1660, + "TCTGCA": 1661, + "TCTGCT": 1662, + "TCTGCC": 1663, + "TCTGCG": 1664, + "TCTGGA": 1665, + "TCTGGT": 1666, + "TCTGGC": 1667, + "TCTGGG": 1668, + "TCCAAA": 1669, + "TCCAAT": 1670, + "TCCAAC": 1671, + "TCCAAG": 1672, + "TCCATA": 1673, + "TCCATT": 1674, + "TCCATC": 1675, + "TCCATG": 1676, + "TCCACA": 1677, + "TCCACT": 1678, + "TCCACC": 1679, + "TCCACG": 1680, + "TCCAGA": 1681, + "TCCAGT": 1682, + "TCCAGC": 1683, + "TCCAGG": 1684, + "TCCTAA": 1685, + "TCCTAT": 1686, + "TCCTAC": 1687, + "TCCTAG": 1688, + "TCCTTA": 1689, + "TCCTTT": 1690, + "TCCTTC": 1691, + "TCCTTG": 1692, + "TCCTCA": 1693, + "TCCTCT": 1694, + "TCCTCC": 1695, + "TCCTCG": 1696, + "TCCTGA": 1697, + "TCCTGT": 1698, + "TCCTGC": 1699, + "TCCTGG": 1700, + "TCCCAA": 1701, + "TCCCAT": 1702, + "TCCCAC": 1703, + "TCCCAG": 1704, + "TCCCTA": 1705, + "TCCCTT": 1706, + "TCCCTC": 1707, + "TCCCTG": 1708, + "TCCCCA": 1709, + "TCCCCT": 1710, + "TCCCCC": 1711, + "TCCCCG": 1712, + "TCCCGA": 1713, + "TCCCGT": 1714, + "TCCCGC": 1715, + "TCCCGG": 1716, + "TCCGAA": 1717, + "TCCGAT": 1718, + "TCCGAC": 1719, + "TCCGAG": 1720, + "TCCGTA": 1721, + "TCCGTT": 1722, + "TCCGTC": 1723, + "TCCGTG": 1724, + "TCCGCA": 1725, + "TCCGCT": 1726, + "TCCGCC": 1727, + "TCCGCG": 1728, + "TCCGGA": 1729, + "TCCGGT": 1730, + "TCCGGC": 1731, + "TCCGGG": 1732, + "TCGAAA": 1733, + "TCGAAT": 1734, + "TCGAAC": 1735, + "TCGAAG": 1736, + "TCGATA": 1737, + "TCGATT": 1738, + "TCGATC": 1739, + "TCGATG": 1740, + "TCGACA": 1741, + "TCGACT": 1742, + "TCGACC": 1743, + "TCGACG": 1744, + "TCGAGA": 1745, + "TCGAGT": 1746, + "TCGAGC": 1747, + "TCGAGG": 1748, + "TCGTAA": 1749, + "TCGTAT": 1750, + "TCGTAC": 1751, + "TCGTAG": 1752, + "TCGTTA": 1753, + "TCGTTT": 1754, + "TCGTTC": 1755, + "TCGTTG": 1756, + "TCGTCA": 1757, + "TCGTCT": 1758, + "TCGTCC": 1759, + "TCGTCG": 1760, + "TCGTGA": 1761, + "TCGTGT": 1762, + "TCGTGC": 1763, + "TCGTGG": 1764, + "TCGCAA": 1765, + "TCGCAT": 1766, + "TCGCAC": 1767, + "TCGCAG": 1768, + "TCGCTA": 1769, + "TCGCTT": 1770, + "TCGCTC": 1771, + "TCGCTG": 1772, + "TCGCCA": 1773, + "TCGCCT": 1774, + "TCGCCC": 1775, + "TCGCCG": 1776, + "TCGCGA": 1777, + "TCGCGT": 1778, + "TCGCGC": 1779, + "TCGCGG": 1780, + "TCGGAA": 1781, + "TCGGAT": 1782, + "TCGGAC": 1783, + "TCGGAG": 1784, + "TCGGTA": 1785, + "TCGGTT": 1786, + "TCGGTC": 1787, + "TCGGTG": 1788, + "TCGGCA": 1789, + "TCGGCT": 1790, + "TCGGCC": 1791, + "TCGGCG": 1792, + "TCGGGA": 1793, + "TCGGGT": 1794, + "TCGGGC": 1795, + "TCGGGG": 1796, + "TGAAAA": 1797, + "TGAAAT": 1798, + "TGAAAC": 1799, + "TGAAAG": 1800, + "TGAATA": 1801, + "TGAATT": 1802, + "TGAATC": 1803, + "TGAATG": 1804, + "TGAACA": 1805, + "TGAACT": 1806, + "TGAACC": 1807, + "TGAACG": 1808, + "TGAAGA": 1809, + "TGAAGT": 1810, + "TGAAGC": 1811, + "TGAAGG": 1812, + "TGATAA": 1813, + "TGATAT": 1814, + "TGATAC": 1815, + "TGATAG": 1816, + "TGATTA": 1817, + "TGATTT": 1818, + "TGATTC": 1819, + "TGATTG": 1820, + "TGATCA": 1821, + "TGATCT": 1822, + "TGATCC": 1823, + "TGATCG": 1824, + "TGATGA": 1825, + "TGATGT": 1826, + "TGATGC": 1827, + "TGATGG": 1828, + "TGACAA": 1829, + "TGACAT": 1830, + "TGACAC": 1831, + "TGACAG": 1832, + "TGACTA": 1833, + "TGACTT": 1834, + "TGACTC": 1835, + "TGACTG": 1836, + "TGACCA": 1837, + "TGACCT": 1838, + "TGACCC": 1839, + "TGACCG": 1840, + "TGACGA": 1841, + "TGACGT": 1842, + "TGACGC": 1843, + "TGACGG": 1844, + "TGAGAA": 1845, + "TGAGAT": 1846, + "TGAGAC": 1847, + "TGAGAG": 1848, + "TGAGTA": 1849, + "TGAGTT": 1850, + "TGAGTC": 1851, + "TGAGTG": 1852, + "TGAGCA": 1853, + "TGAGCT": 1854, + "TGAGCC": 1855, + "TGAGCG": 1856, + "TGAGGA": 1857, + "TGAGGT": 1858, + "TGAGGC": 1859, + "TGAGGG": 1860, + "TGTAAA": 1861, + "TGTAAT": 1862, + "TGTAAC": 1863, + "TGTAAG": 1864, + "TGTATA": 1865, + "TGTATT": 1866, + "TGTATC": 1867, + "TGTATG": 1868, + "TGTACA": 1869, + "TGTACT": 1870, + "TGTACC": 1871, + "TGTACG": 1872, + "TGTAGA": 1873, + "TGTAGT": 1874, + "TGTAGC": 1875, + "TGTAGG": 1876, + "TGTTAA": 1877, + "TGTTAT": 1878, + "TGTTAC": 1879, + "TGTTAG": 1880, + "TGTTTA": 1881, + "TGTTTT": 1882, + "TGTTTC": 1883, + "TGTTTG": 1884, + "TGTTCA": 1885, + "TGTTCT": 1886, + "TGTTCC": 1887, + "TGTTCG": 1888, + "TGTTGA": 1889, + "TGTTGT": 1890, + "TGTTGC": 1891, + "TGTTGG": 1892, + "TGTCAA": 1893, + "TGTCAT": 1894, + "TGTCAC": 1895, + "TGTCAG": 1896, + "TGTCTA": 1897, + "TGTCTT": 1898, + "TGTCTC": 1899, + "TGTCTG": 1900, + "TGTCCA": 1901, + "TGTCCT": 1902, + "TGTCCC": 1903, + "TGTCCG": 1904, + "TGTCGA": 1905, + "TGTCGT": 1906, + "TGTCGC": 1907, + "TGTCGG": 1908, + "TGTGAA": 1909, + "TGTGAT": 1910, + "TGTGAC": 1911, + "TGTGAG": 1912, + "TGTGTA": 1913, + "TGTGTT": 1914, + "TGTGTC": 1915, + "TGTGTG": 1916, + "TGTGCA": 1917, + "TGTGCT": 1918, + "TGTGCC": 1919, + "TGTGCG": 1920, + "TGTGGA": 1921, + "TGTGGT": 1922, + "TGTGGC": 1923, + "TGTGGG": 1924, + "TGCAAA": 1925, + "TGCAAT": 1926, + "TGCAAC": 1927, + "TGCAAG": 1928, + "TGCATA": 1929, + "TGCATT": 1930, + "TGCATC": 1931, + "TGCATG": 1932, + "TGCACA": 1933, + "TGCACT": 1934, + "TGCACC": 1935, + "TGCACG": 1936, + "TGCAGA": 1937, + "TGCAGT": 1938, + "TGCAGC": 1939, + "TGCAGG": 1940, + "TGCTAA": 1941, + "TGCTAT": 1942, + "TGCTAC": 1943, + "TGCTAG": 1944, + "TGCTTA": 1945, + "TGCTTT": 1946, + "TGCTTC": 1947, + "TGCTTG": 1948, + "TGCTCA": 1949, + "TGCTCT": 1950, + "TGCTCC": 1951, + "TGCTCG": 1952, + "TGCTGA": 1953, + "TGCTGT": 1954, + "TGCTGC": 1955, + "TGCTGG": 1956, + "TGCCAA": 1957, + "TGCCAT": 1958, + "TGCCAC": 1959, + "TGCCAG": 1960, + "TGCCTA": 1961, + "TGCCTT": 1962, + "TGCCTC": 1963, + "TGCCTG": 1964, + "TGCCCA": 1965, + "TGCCCT": 1966, + "TGCCCC": 1967, + "TGCCCG": 1968, + "TGCCGA": 1969, + "TGCCGT": 1970, + "TGCCGC": 1971, + "TGCCGG": 1972, + "TGCGAA": 1973, + "TGCGAT": 1974, + "TGCGAC": 1975, + "TGCGAG": 1976, + "TGCGTA": 1977, + "TGCGTT": 1978, + "TGCGTC": 1979, + "TGCGTG": 1980, + "TGCGCA": 1981, + "TGCGCT": 1982, + "TGCGCC": 1983, + "TGCGCG": 1984, + "TGCGGA": 1985, + "TGCGGT": 1986, + "TGCGGC": 1987, + "TGCGGG": 1988, + "TGGAAA": 1989, + "TGGAAT": 1990, + "TGGAAC": 1991, + "TGGAAG": 1992, + "TGGATA": 1993, + "TGGATT": 1994, + "TGGATC": 1995, + "TGGATG": 1996, + "TGGACA": 1997, + "TGGACT": 1998, + "TGGACC": 1999, + "TGGACG": 2000, + "TGGAGA": 2001, + "TGGAGT": 2002, + "TGGAGC": 2003, + "TGGAGG": 2004, + "TGGTAA": 2005, + "TGGTAT": 2006, + "TGGTAC": 2007, + "TGGTAG": 2008, + "TGGTTA": 2009, + "TGGTTT": 2010, + "TGGTTC": 2011, + "TGGTTG": 2012, + "TGGTCA": 2013, + "TGGTCT": 2014, + "TGGTCC": 2015, + "TGGTCG": 2016, + "TGGTGA": 2017, + "TGGTGT": 2018, + "TGGTGC": 2019, + "TGGTGG": 2020, + "TGGCAA": 2021, + "TGGCAT": 2022, + "TGGCAC": 2023, + "TGGCAG": 2024, + "TGGCTA": 2025, + "TGGCTT": 2026, + "TGGCTC": 2027, + "TGGCTG": 2028, + "TGGCCA": 2029, + "TGGCCT": 2030, + "TGGCCC": 2031, + "TGGCCG": 2032, + "TGGCGA": 2033, + "TGGCGT": 2034, + "TGGCGC": 2035, + "TGGCGG": 2036, + "TGGGAA": 2037, + "TGGGAT": 2038, + "TGGGAC": 2039, + "TGGGAG": 2040, + "TGGGTA": 2041, + "TGGGTT": 2042, + "TGGGTC": 2043, + "TGGGTG": 2044, + "TGGGCA": 2045, + "TGGGCT": 2046, + "TGGGCC": 2047, + "TGGGCG": 2048, + "TGGGGA": 2049, + "TGGGGT": 2050, + "TGGGGC": 2051, + "TGGGGG": 2052, + "CAAAAA": 2053, + "CAAAAT": 2054, + "CAAAAC": 2055, + "CAAAAG": 2056, + "CAAATA": 2057, + "CAAATT": 2058, + "CAAATC": 2059, + "CAAATG": 2060, + "CAAACA": 2061, + "CAAACT": 2062, + "CAAACC": 2063, + "CAAACG": 2064, + "CAAAGA": 2065, + "CAAAGT": 2066, + "CAAAGC": 2067, + "CAAAGG": 2068, + "CAATAA": 2069, + "CAATAT": 2070, + "CAATAC": 2071, + "CAATAG": 2072, + "CAATTA": 2073, + "CAATTT": 2074, + "CAATTC": 2075, + "CAATTG": 2076, + "CAATCA": 2077, + "CAATCT": 2078, + "CAATCC": 2079, + "CAATCG": 2080, + "CAATGA": 2081, + "CAATGT": 2082, + "CAATGC": 2083, + "CAATGG": 2084, + "CAACAA": 2085, + "CAACAT": 2086, + "CAACAC": 2087, + "CAACAG": 2088, + "CAACTA": 2089, + "CAACTT": 2090, + "CAACTC": 2091, + "CAACTG": 2092, + "CAACCA": 2093, + "CAACCT": 2094, + "CAACCC": 2095, + "CAACCG": 2096, + "CAACGA": 2097, + "CAACGT": 2098, + "CAACGC": 2099, + "CAACGG": 2100, + "CAAGAA": 2101, + "CAAGAT": 2102, + "CAAGAC": 2103, + "CAAGAG": 2104, + "CAAGTA": 2105, + "CAAGTT": 2106, + "CAAGTC": 2107, + "CAAGTG": 2108, + "CAAGCA": 2109, + "CAAGCT": 2110, + "CAAGCC": 2111, + "CAAGCG": 2112, + "CAAGGA": 2113, + "CAAGGT": 2114, + "CAAGGC": 2115, + "CAAGGG": 2116, + "CATAAA": 2117, + "CATAAT": 2118, + "CATAAC": 2119, + "CATAAG": 2120, + "CATATA": 2121, + "CATATT": 2122, + "CATATC": 2123, + "CATATG": 2124, + "CATACA": 2125, + "CATACT": 2126, + "CATACC": 2127, + "CATACG": 2128, + "CATAGA": 2129, + "CATAGT": 2130, + "CATAGC": 2131, + "CATAGG": 2132, + "CATTAA": 2133, + "CATTAT": 2134, + "CATTAC": 2135, + "CATTAG": 2136, + "CATTTA": 2137, + "CATTTT": 2138, + "CATTTC": 2139, + "CATTTG": 2140, + "CATTCA": 2141, + "CATTCT": 2142, + "CATTCC": 2143, + "CATTCG": 2144, + "CATTGA": 2145, + "CATTGT": 2146, + "CATTGC": 2147, + "CATTGG": 2148, + "CATCAA": 2149, + "CATCAT": 2150, + "CATCAC": 2151, + "CATCAG": 2152, + "CATCTA": 2153, + "CATCTT": 2154, + "CATCTC": 2155, + "CATCTG": 2156, + "CATCCA": 2157, + "CATCCT": 2158, + "CATCCC": 2159, + "CATCCG": 2160, + "CATCGA": 2161, + "CATCGT": 2162, + "CATCGC": 2163, + "CATCGG": 2164, + "CATGAA": 2165, + "CATGAT": 2166, + "CATGAC": 2167, + "CATGAG": 2168, + "CATGTA": 2169, + "CATGTT": 2170, + "CATGTC": 2171, + "CATGTG": 2172, + "CATGCA": 2173, + "CATGCT": 2174, + "CATGCC": 2175, + "CATGCG": 2176, + "CATGGA": 2177, + "CATGGT": 2178, + "CATGGC": 2179, + "CATGGG": 2180, + "CACAAA": 2181, + "CACAAT": 2182, + "CACAAC": 2183, + "CACAAG": 2184, + "CACATA": 2185, + "CACATT": 2186, + "CACATC": 2187, + "CACATG": 2188, + "CACACA": 2189, + "CACACT": 2190, + "CACACC": 2191, + "CACACG": 2192, + "CACAGA": 2193, + "CACAGT": 2194, + "CACAGC": 2195, + "CACAGG": 2196, + "CACTAA": 2197, + "CACTAT": 2198, + "CACTAC": 2199, + "CACTAG": 2200, + "CACTTA": 2201, + "CACTTT": 2202, + "CACTTC": 2203, + "CACTTG": 2204, + "CACTCA": 2205, + "CACTCT": 2206, + "CACTCC": 2207, + "CACTCG": 2208, + "CACTGA": 2209, + "CACTGT": 2210, + "CACTGC": 2211, + "CACTGG": 2212, + "CACCAA": 2213, + "CACCAT": 2214, + "CACCAC": 2215, + "CACCAG": 2216, + "CACCTA": 2217, + "CACCTT": 2218, + "CACCTC": 2219, + "CACCTG": 2220, + "CACCCA": 2221, + "CACCCT": 2222, + "CACCCC": 2223, + "CACCCG": 2224, + "CACCGA": 2225, + "CACCGT": 2226, + "CACCGC": 2227, + "CACCGG": 2228, + "CACGAA": 2229, + "CACGAT": 2230, + "CACGAC": 2231, + "CACGAG": 2232, + "CACGTA": 2233, + "CACGTT": 2234, + "CACGTC": 2235, + "CACGTG": 2236, + "CACGCA": 2237, + "CACGCT": 2238, + "CACGCC": 2239, + "CACGCG": 2240, + "CACGGA": 2241, + "CACGGT": 2242, + "CACGGC": 2243, + "CACGGG": 2244, + "CAGAAA": 2245, + "CAGAAT": 2246, + "CAGAAC": 2247, + "CAGAAG": 2248, + "CAGATA": 2249, + "CAGATT": 2250, + "CAGATC": 2251, + "CAGATG": 2252, + "CAGACA": 2253, + "CAGACT": 2254, + "CAGACC": 2255, + "CAGACG": 2256, + "CAGAGA": 2257, + "CAGAGT": 2258, + "CAGAGC": 2259, + "CAGAGG": 2260, + "CAGTAA": 2261, + "CAGTAT": 2262, + "CAGTAC": 2263, + "CAGTAG": 2264, + "CAGTTA": 2265, + "CAGTTT": 2266, + "CAGTTC": 2267, + "CAGTTG": 2268, + "CAGTCA": 2269, + "CAGTCT": 2270, + "CAGTCC": 2271, + "CAGTCG": 2272, + "CAGTGA": 2273, + "CAGTGT": 2274, + "CAGTGC": 2275, + "CAGTGG": 2276, + "CAGCAA": 2277, + "CAGCAT": 2278, + "CAGCAC": 2279, + "CAGCAG": 2280, + "CAGCTA": 2281, + "CAGCTT": 2282, + "CAGCTC": 2283, + "CAGCTG": 2284, + "CAGCCA": 2285, + "CAGCCT": 2286, + "CAGCCC": 2287, + "CAGCCG": 2288, + "CAGCGA": 2289, + "CAGCGT": 2290, + "CAGCGC": 2291, + "CAGCGG": 2292, + "CAGGAA": 2293, + "CAGGAT": 2294, + "CAGGAC": 2295, + "CAGGAG": 2296, + "CAGGTA": 2297, + "CAGGTT": 2298, + "CAGGTC": 2299, + "CAGGTG": 2300, + "CAGGCA": 2301, + "CAGGCT": 2302, + "CAGGCC": 2303, + "CAGGCG": 2304, + "CAGGGA": 2305, + "CAGGGT": 2306, + "CAGGGC": 2307, + "CAGGGG": 2308, + "CTAAAA": 2309, + "CTAAAT": 2310, + "CTAAAC": 2311, + "CTAAAG": 2312, + "CTAATA": 2313, + "CTAATT": 2314, + "CTAATC": 2315, + "CTAATG": 2316, + "CTAACA": 2317, + "CTAACT": 2318, + "CTAACC": 2319, + "CTAACG": 2320, + "CTAAGA": 2321, + "CTAAGT": 2322, + "CTAAGC": 2323, + "CTAAGG": 2324, + "CTATAA": 2325, + "CTATAT": 2326, + "CTATAC": 2327, + "CTATAG": 2328, + "CTATTA": 2329, + "CTATTT": 2330, + "CTATTC": 2331, + "CTATTG": 2332, + "CTATCA": 2333, + "CTATCT": 2334, + "CTATCC": 2335, + "CTATCG": 2336, + "CTATGA": 2337, + "CTATGT": 2338, + "CTATGC": 2339, + "CTATGG": 2340, + "CTACAA": 2341, + "CTACAT": 2342, + "CTACAC": 2343, + "CTACAG": 2344, + "CTACTA": 2345, + "CTACTT": 2346, + "CTACTC": 2347, + "CTACTG": 2348, + "CTACCA": 2349, + "CTACCT": 2350, + "CTACCC": 2351, + "CTACCG": 2352, + "CTACGA": 2353, + "CTACGT": 2354, + "CTACGC": 2355, + "CTACGG": 2356, + "CTAGAA": 2357, + "CTAGAT": 2358, + "CTAGAC": 2359, + "CTAGAG": 2360, + "CTAGTA": 2361, + "CTAGTT": 2362, + "CTAGTC": 2363, + "CTAGTG": 2364, + "CTAGCA": 2365, + "CTAGCT": 2366, + "CTAGCC": 2367, + "CTAGCG": 2368, + "CTAGGA": 2369, + "CTAGGT": 2370, + "CTAGGC": 2371, + "CTAGGG": 2372, + "CTTAAA": 2373, + "CTTAAT": 2374, + "CTTAAC": 2375, + "CTTAAG": 2376, + "CTTATA": 2377, + "CTTATT": 2378, + "CTTATC": 2379, + "CTTATG": 2380, + "CTTACA": 2381, + "CTTACT": 2382, + "CTTACC": 2383, + "CTTACG": 2384, + "CTTAGA": 2385, + "CTTAGT": 2386, + "CTTAGC": 2387, + "CTTAGG": 2388, + "CTTTAA": 2389, + "CTTTAT": 2390, + "CTTTAC": 2391, + "CTTTAG": 2392, + "CTTTTA": 2393, + "CTTTTT": 2394, + "CTTTTC": 2395, + "CTTTTG": 2396, + "CTTTCA": 2397, + "CTTTCT": 2398, + "CTTTCC": 2399, + "CTTTCG": 2400, + "CTTTGA": 2401, + "CTTTGT": 2402, + "CTTTGC": 2403, + "CTTTGG": 2404, + "CTTCAA": 2405, + "CTTCAT": 2406, + "CTTCAC": 2407, + "CTTCAG": 2408, + "CTTCTA": 2409, + "CTTCTT": 2410, + "CTTCTC": 2411, + "CTTCTG": 2412, + "CTTCCA": 2413, + "CTTCCT": 2414, + "CTTCCC": 2415, + "CTTCCG": 2416, + "CTTCGA": 2417, + "CTTCGT": 2418, + "CTTCGC": 2419, + "CTTCGG": 2420, + "CTTGAA": 2421, + "CTTGAT": 2422, + "CTTGAC": 2423, + "CTTGAG": 2424, + "CTTGTA": 2425, + "CTTGTT": 2426, + "CTTGTC": 2427, + "CTTGTG": 2428, + "CTTGCA": 2429, + "CTTGCT": 2430, + "CTTGCC": 2431, + "CTTGCG": 2432, + "CTTGGA": 2433, + "CTTGGT": 2434, + "CTTGGC": 2435, + "CTTGGG": 2436, + "CTCAAA": 2437, + "CTCAAT": 2438, + "CTCAAC": 2439, + "CTCAAG": 2440, + "CTCATA": 2441, + "CTCATT": 2442, + "CTCATC": 2443, + "CTCATG": 2444, + "CTCACA": 2445, + "CTCACT": 2446, + "CTCACC": 2447, + "CTCACG": 2448, + "CTCAGA": 2449, + "CTCAGT": 2450, + "CTCAGC": 2451, + "CTCAGG": 2452, + "CTCTAA": 2453, + "CTCTAT": 2454, + "CTCTAC": 2455, + "CTCTAG": 2456, + "CTCTTA": 2457, + "CTCTTT": 2458, + "CTCTTC": 2459, + "CTCTTG": 2460, + "CTCTCA": 2461, + "CTCTCT": 2462, + "CTCTCC": 2463, + "CTCTCG": 2464, + "CTCTGA": 2465, + "CTCTGT": 2466, + "CTCTGC": 2467, + "CTCTGG": 2468, + "CTCCAA": 2469, + "CTCCAT": 2470, + "CTCCAC": 2471, + "CTCCAG": 2472, + "CTCCTA": 2473, + "CTCCTT": 2474, + "CTCCTC": 2475, + "CTCCTG": 2476, + "CTCCCA": 2477, + "CTCCCT": 2478, + "CTCCCC": 2479, + "CTCCCG": 2480, + "CTCCGA": 2481, + "CTCCGT": 2482, + "CTCCGC": 2483, + "CTCCGG": 2484, + "CTCGAA": 2485, + "CTCGAT": 2486, + "CTCGAC": 2487, + "CTCGAG": 2488, + "CTCGTA": 2489, + "CTCGTT": 2490, + "CTCGTC": 2491, + "CTCGTG": 2492, + "CTCGCA": 2493, + "CTCGCT": 2494, + "CTCGCC": 2495, + "CTCGCG": 2496, + "CTCGGA": 2497, + "CTCGGT": 2498, + "CTCGGC": 2499, + "CTCGGG": 2500, + "CTGAAA": 2501, + "CTGAAT": 2502, + "CTGAAC": 2503, + "CTGAAG": 2504, + "CTGATA": 2505, + "CTGATT": 2506, + "CTGATC": 2507, + "CTGATG": 2508, + "CTGACA": 2509, + "CTGACT": 2510, + "CTGACC": 2511, + "CTGACG": 2512, + "CTGAGA": 2513, + "CTGAGT": 2514, + "CTGAGC": 2515, + "CTGAGG": 2516, + "CTGTAA": 2517, + "CTGTAT": 2518, + "CTGTAC": 2519, + "CTGTAG": 2520, + "CTGTTA": 2521, + "CTGTTT": 2522, + "CTGTTC": 2523, + "CTGTTG": 2524, + "CTGTCA": 2525, + "CTGTCT": 2526, + "CTGTCC": 2527, + "CTGTCG": 2528, + "CTGTGA": 2529, + "CTGTGT": 2530, + "CTGTGC": 2531, + "CTGTGG": 2532, + "CTGCAA": 2533, + "CTGCAT": 2534, + "CTGCAC": 2535, + "CTGCAG": 2536, + "CTGCTA": 2537, + "CTGCTT": 2538, + "CTGCTC": 2539, + "CTGCTG": 2540, + "CTGCCA": 2541, + "CTGCCT": 2542, + "CTGCCC": 2543, + "CTGCCG": 2544, + "CTGCGA": 2545, + "CTGCGT": 2546, + "CTGCGC": 2547, + "CTGCGG": 2548, + "CTGGAA": 2549, + "CTGGAT": 2550, + "CTGGAC": 2551, + "CTGGAG": 2552, + "CTGGTA": 2553, + "CTGGTT": 2554, + "CTGGTC": 2555, + "CTGGTG": 2556, + "CTGGCA": 2557, + "CTGGCT": 2558, + "CTGGCC": 2559, + "CTGGCG": 2560, + "CTGGGA": 2561, + "CTGGGT": 2562, + "CTGGGC": 2563, + "CTGGGG": 2564, + "CCAAAA": 2565, + "CCAAAT": 2566, + "CCAAAC": 2567, + "CCAAAG": 2568, + "CCAATA": 2569, + "CCAATT": 2570, + "CCAATC": 2571, + "CCAATG": 2572, + "CCAACA": 2573, + "CCAACT": 2574, + "CCAACC": 2575, + "CCAACG": 2576, + "CCAAGA": 2577, + "CCAAGT": 2578, + "CCAAGC": 2579, + "CCAAGG": 2580, + "CCATAA": 2581, + "CCATAT": 2582, + "CCATAC": 2583, + "CCATAG": 2584, + "CCATTA": 2585, + "CCATTT": 2586, + "CCATTC": 2587, + "CCATTG": 2588, + "CCATCA": 2589, + "CCATCT": 2590, + "CCATCC": 2591, + "CCATCG": 2592, + "CCATGA": 2593, + "CCATGT": 2594, + "CCATGC": 2595, + "CCATGG": 2596, + "CCACAA": 2597, + "CCACAT": 2598, + "CCACAC": 2599, + "CCACAG": 2600, + "CCACTA": 2601, + "CCACTT": 2602, + "CCACTC": 2603, + "CCACTG": 2604, + "CCACCA": 2605, + "CCACCT": 2606, + "CCACCC": 2607, + "CCACCG": 2608, + "CCACGA": 2609, + "CCACGT": 2610, + "CCACGC": 2611, + "CCACGG": 2612, + "CCAGAA": 2613, + "CCAGAT": 2614, + "CCAGAC": 2615, + "CCAGAG": 2616, + "CCAGTA": 2617, + "CCAGTT": 2618, + "CCAGTC": 2619, + "CCAGTG": 2620, + "CCAGCA": 2621, + "CCAGCT": 2622, + "CCAGCC": 2623, + "CCAGCG": 2624, + "CCAGGA": 2625, + "CCAGGT": 2626, + "CCAGGC": 2627, + "CCAGGG": 2628, + "CCTAAA": 2629, + "CCTAAT": 2630, + "CCTAAC": 2631, + "CCTAAG": 2632, + "CCTATA": 2633, + "CCTATT": 2634, + "CCTATC": 2635, + "CCTATG": 2636, + "CCTACA": 2637, + "CCTACT": 2638, + "CCTACC": 2639, + "CCTACG": 2640, + "CCTAGA": 2641, + "CCTAGT": 2642, + "CCTAGC": 2643, + "CCTAGG": 2644, + "CCTTAA": 2645, + "CCTTAT": 2646, + "CCTTAC": 2647, + "CCTTAG": 2648, + "CCTTTA": 2649, + "CCTTTT": 2650, + "CCTTTC": 2651, + "CCTTTG": 2652, + "CCTTCA": 2653, + "CCTTCT": 2654, + "CCTTCC": 2655, + "CCTTCG": 2656, + "CCTTGA": 2657, + "CCTTGT": 2658, + "CCTTGC": 2659, + "CCTTGG": 2660, + "CCTCAA": 2661, + "CCTCAT": 2662, + "CCTCAC": 2663, + "CCTCAG": 2664, + "CCTCTA": 2665, + "CCTCTT": 2666, + "CCTCTC": 2667, + "CCTCTG": 2668, + "CCTCCA": 2669, + "CCTCCT": 2670, + "CCTCCC": 2671, + "CCTCCG": 2672, + "CCTCGA": 2673, + "CCTCGT": 2674, + "CCTCGC": 2675, + "CCTCGG": 2676, + "CCTGAA": 2677, + "CCTGAT": 2678, + "CCTGAC": 2679, + "CCTGAG": 2680, + "CCTGTA": 2681, + "CCTGTT": 2682, + "CCTGTC": 2683, + "CCTGTG": 2684, + "CCTGCA": 2685, + "CCTGCT": 2686, + "CCTGCC": 2687, + "CCTGCG": 2688, + "CCTGGA": 2689, + "CCTGGT": 2690, + "CCTGGC": 2691, + "CCTGGG": 2692, + "CCCAAA": 2693, + "CCCAAT": 2694, + "CCCAAC": 2695, + "CCCAAG": 2696, + "CCCATA": 2697, + "CCCATT": 2698, + "CCCATC": 2699, + "CCCATG": 2700, + "CCCACA": 2701, + "CCCACT": 2702, + "CCCACC": 2703, + "CCCACG": 2704, + "CCCAGA": 2705, + "CCCAGT": 2706, + "CCCAGC": 2707, + "CCCAGG": 2708, + "CCCTAA": 2709, + "CCCTAT": 2710, + "CCCTAC": 2711, + "CCCTAG": 2712, + "CCCTTA": 2713, + "CCCTTT": 2714, + "CCCTTC": 2715, + "CCCTTG": 2716, + "CCCTCA": 2717, + "CCCTCT": 2718, + "CCCTCC": 2719, + "CCCTCG": 2720, + "CCCTGA": 2721, + "CCCTGT": 2722, + "CCCTGC": 2723, + "CCCTGG": 2724, + "CCCCAA": 2725, + "CCCCAT": 2726, + "CCCCAC": 2727, + "CCCCAG": 2728, + "CCCCTA": 2729, + "CCCCTT": 2730, + "CCCCTC": 2731, + "CCCCTG": 2732, + "CCCCCA": 2733, + "CCCCCT": 2734, + "CCCCCC": 2735, + "CCCCCG": 2736, + "CCCCGA": 2737, + "CCCCGT": 2738, + "CCCCGC": 2739, + "CCCCGG": 2740, + "CCCGAA": 2741, + "CCCGAT": 2742, + "CCCGAC": 2743, + "CCCGAG": 2744, + "CCCGTA": 2745, + "CCCGTT": 2746, + "CCCGTC": 2747, + "CCCGTG": 2748, + "CCCGCA": 2749, + "CCCGCT": 2750, + "CCCGCC": 2751, + "CCCGCG": 2752, + "CCCGGA": 2753, + "CCCGGT": 2754, + "CCCGGC": 2755, + "CCCGGG": 2756, + "CCGAAA": 2757, + "CCGAAT": 2758, + "CCGAAC": 2759, + "CCGAAG": 2760, + "CCGATA": 2761, + "CCGATT": 2762, + "CCGATC": 2763, + "CCGATG": 2764, + "CCGACA": 2765, + "CCGACT": 2766, + "CCGACC": 2767, + "CCGACG": 2768, + "CCGAGA": 2769, + "CCGAGT": 2770, + "CCGAGC": 2771, + "CCGAGG": 2772, + "CCGTAA": 2773, + "CCGTAT": 2774, + "CCGTAC": 2775, + "CCGTAG": 2776, + "CCGTTA": 2777, + "CCGTTT": 2778, + "CCGTTC": 2779, + "CCGTTG": 2780, + "CCGTCA": 2781, + "CCGTCT": 2782, + "CCGTCC": 2783, + "CCGTCG": 2784, + "CCGTGA": 2785, + "CCGTGT": 2786, + "CCGTGC": 2787, + "CCGTGG": 2788, + "CCGCAA": 2789, + "CCGCAT": 2790, + "CCGCAC": 2791, + "CCGCAG": 2792, + "CCGCTA": 2793, + "CCGCTT": 2794, + "CCGCTC": 2795, + "CCGCTG": 2796, + "CCGCCA": 2797, + "CCGCCT": 2798, + "CCGCCC": 2799, + "CCGCCG": 2800, + "CCGCGA": 2801, + "CCGCGT": 2802, + "CCGCGC": 2803, + "CCGCGG": 2804, + "CCGGAA": 2805, + "CCGGAT": 2806, + "CCGGAC": 2807, + "CCGGAG": 2808, + "CCGGTA": 2809, + "CCGGTT": 2810, + "CCGGTC": 2811, + "CCGGTG": 2812, + "CCGGCA": 2813, + "CCGGCT": 2814, + "CCGGCC": 2815, + "CCGGCG": 2816, + "CCGGGA": 2817, + "CCGGGT": 2818, + "CCGGGC": 2819, + "CCGGGG": 2820, + "CGAAAA": 2821, + "CGAAAT": 2822, + "CGAAAC": 2823, + "CGAAAG": 2824, + "CGAATA": 2825, + "CGAATT": 2826, + "CGAATC": 2827, + "CGAATG": 2828, + "CGAACA": 2829, + "CGAACT": 2830, + "CGAACC": 2831, + "CGAACG": 2832, + "CGAAGA": 2833, + "CGAAGT": 2834, + "CGAAGC": 2835, + "CGAAGG": 2836, + "CGATAA": 2837, + "CGATAT": 2838, + "CGATAC": 2839, + "CGATAG": 2840, + "CGATTA": 2841, + "CGATTT": 2842, + "CGATTC": 2843, + "CGATTG": 2844, + "CGATCA": 2845, + "CGATCT": 2846, + "CGATCC": 2847, + "CGATCG": 2848, + "CGATGA": 2849, + "CGATGT": 2850, + "CGATGC": 2851, + "CGATGG": 2852, + "CGACAA": 2853, + "CGACAT": 2854, + "CGACAC": 2855, + "CGACAG": 2856, + "CGACTA": 2857, + "CGACTT": 2858, + "CGACTC": 2859, + "CGACTG": 2860, + "CGACCA": 2861, + "CGACCT": 2862, + "CGACCC": 2863, + "CGACCG": 2864, + "CGACGA": 2865, + "CGACGT": 2866, + "CGACGC": 2867, + "CGACGG": 2868, + "CGAGAA": 2869, + "CGAGAT": 2870, + "CGAGAC": 2871, + "CGAGAG": 2872, + "CGAGTA": 2873, + "CGAGTT": 2874, + "CGAGTC": 2875, + "CGAGTG": 2876, + "CGAGCA": 2877, + "CGAGCT": 2878, + "CGAGCC": 2879, + "CGAGCG": 2880, + "CGAGGA": 2881, + "CGAGGT": 2882, + "CGAGGC": 2883, + "CGAGGG": 2884, + "CGTAAA": 2885, + "CGTAAT": 2886, + "CGTAAC": 2887, + "CGTAAG": 2888, + "CGTATA": 2889, + "CGTATT": 2890, + "CGTATC": 2891, + "CGTATG": 2892, + "CGTACA": 2893, + "CGTACT": 2894, + "CGTACC": 2895, + "CGTACG": 2896, + "CGTAGA": 2897, + "CGTAGT": 2898, + "CGTAGC": 2899, + "CGTAGG": 2900, + "CGTTAA": 2901, + "CGTTAT": 2902, + "CGTTAC": 2903, + "CGTTAG": 2904, + "CGTTTA": 2905, + "CGTTTT": 2906, + "CGTTTC": 2907, + "CGTTTG": 2908, + "CGTTCA": 2909, + "CGTTCT": 2910, + "CGTTCC": 2911, + "CGTTCG": 2912, + "CGTTGA": 2913, + "CGTTGT": 2914, + "CGTTGC": 2915, + "CGTTGG": 2916, + "CGTCAA": 2917, + "CGTCAT": 2918, + "CGTCAC": 2919, + "CGTCAG": 2920, + "CGTCTA": 2921, + "CGTCTT": 2922, + "CGTCTC": 2923, + "CGTCTG": 2924, + "CGTCCA": 2925, + "CGTCCT": 2926, + "CGTCCC": 2927, + "CGTCCG": 2928, + "CGTCGA": 2929, + "CGTCGT": 2930, + "CGTCGC": 2931, + "CGTCGG": 2932, + "CGTGAA": 2933, + "CGTGAT": 2934, + "CGTGAC": 2935, + "CGTGAG": 2936, + "CGTGTA": 2937, + "CGTGTT": 2938, + "CGTGTC": 2939, + "CGTGTG": 2940, + "CGTGCA": 2941, + "CGTGCT": 2942, + "CGTGCC": 2943, + "CGTGCG": 2944, + "CGTGGA": 2945, + "CGTGGT": 2946, + "CGTGGC": 2947, + "CGTGGG": 2948, + "CGCAAA": 2949, + "CGCAAT": 2950, + "CGCAAC": 2951, + "CGCAAG": 2952, + "CGCATA": 2953, + "CGCATT": 2954, + "CGCATC": 2955, + "CGCATG": 2956, + "CGCACA": 2957, + "CGCACT": 2958, + "CGCACC": 2959, + "CGCACG": 2960, + "CGCAGA": 2961, + "CGCAGT": 2962, + "CGCAGC": 2963, + "CGCAGG": 2964, + "CGCTAA": 2965, + "CGCTAT": 2966, + "CGCTAC": 2967, + "CGCTAG": 2968, + "CGCTTA": 2969, + "CGCTTT": 2970, + "CGCTTC": 2971, + "CGCTTG": 2972, + "CGCTCA": 2973, + "CGCTCT": 2974, + "CGCTCC": 2975, + "CGCTCG": 2976, + "CGCTGA": 2977, + "CGCTGT": 2978, + "CGCTGC": 2979, + "CGCTGG": 2980, + "CGCCAA": 2981, + "CGCCAT": 2982, + "CGCCAC": 2983, + "CGCCAG": 2984, + "CGCCTA": 2985, + "CGCCTT": 2986, + "CGCCTC": 2987, + "CGCCTG": 2988, + "CGCCCA": 2989, + "CGCCCT": 2990, + "CGCCCC": 2991, + "CGCCCG": 2992, + "CGCCGA": 2993, + "CGCCGT": 2994, + "CGCCGC": 2995, + "CGCCGG": 2996, + "CGCGAA": 2997, + "CGCGAT": 2998, + "CGCGAC": 2999, + "CGCGAG": 3000, + "CGCGTA": 3001, + "CGCGTT": 3002, + "CGCGTC": 3003, + "CGCGTG": 3004, + "CGCGCA": 3005, + "CGCGCT": 3006, + "CGCGCC": 3007, + "CGCGCG": 3008, + "CGCGGA": 3009, + "CGCGGT": 3010, + "CGCGGC": 3011, + "CGCGGG": 3012, + "CGGAAA": 3013, + "CGGAAT": 3014, + "CGGAAC": 3015, + "CGGAAG": 3016, + "CGGATA": 3017, + "CGGATT": 3018, + "CGGATC": 3019, + "CGGATG": 3020, + "CGGACA": 3021, + "CGGACT": 3022, + "CGGACC": 3023, + "CGGACG": 3024, + "CGGAGA": 3025, + "CGGAGT": 3026, + "CGGAGC": 3027, + "CGGAGG": 3028, + "CGGTAA": 3029, + "CGGTAT": 3030, + "CGGTAC": 3031, + "CGGTAG": 3032, + "CGGTTA": 3033, + "CGGTTT": 3034, + "CGGTTC": 3035, + "CGGTTG": 3036, + "CGGTCA": 3037, + "CGGTCT": 3038, + "CGGTCC": 3039, + "CGGTCG": 3040, + "CGGTGA": 3041, + "CGGTGT": 3042, + "CGGTGC": 3043, + "CGGTGG": 3044, + "CGGCAA": 3045, + "CGGCAT": 3046, + "CGGCAC": 3047, + "CGGCAG": 3048, + "CGGCTA": 3049, + "CGGCTT": 3050, + "CGGCTC": 3051, + "CGGCTG": 3052, + "CGGCCA": 3053, + "CGGCCT": 3054, + "CGGCCC": 3055, + "CGGCCG": 3056, + "CGGCGA": 3057, + "CGGCGT": 3058, + "CGGCGC": 3059, + "CGGCGG": 3060, + "CGGGAA": 3061, + "CGGGAT": 3062, + "CGGGAC": 3063, + "CGGGAG": 3064, + "CGGGTA": 3065, + "CGGGTT": 3066, + "CGGGTC": 3067, + "CGGGTG": 3068, + "CGGGCA": 3069, + "CGGGCT": 3070, + "CGGGCC": 3071, + "CGGGCG": 3072, + "CGGGGA": 3073, + "CGGGGT": 3074, + "CGGGGC": 3075, + "CGGGGG": 3076, + "GAAAAA": 3077, + "GAAAAT": 3078, + "GAAAAC": 3079, + "GAAAAG": 3080, + "GAAATA": 3081, + "GAAATT": 3082, + "GAAATC": 3083, + "GAAATG": 3084, + "GAAACA": 3085, + "GAAACT": 3086, + "GAAACC": 3087, + "GAAACG": 3088, + "GAAAGA": 3089, + "GAAAGT": 3090, + "GAAAGC": 3091, + "GAAAGG": 3092, + "GAATAA": 3093, + "GAATAT": 3094, + "GAATAC": 3095, + "GAATAG": 3096, + "GAATTA": 3097, + "GAATTT": 3098, + "GAATTC": 3099, + "GAATTG": 3100, + "GAATCA": 3101, + "GAATCT": 3102, + "GAATCC": 3103, + "GAATCG": 3104, + "GAATGA": 3105, + "GAATGT": 3106, + "GAATGC": 3107, + "GAATGG": 3108, + "GAACAA": 3109, + "GAACAT": 3110, + "GAACAC": 3111, + "GAACAG": 3112, + "GAACTA": 3113, + "GAACTT": 3114, + "GAACTC": 3115, + "GAACTG": 3116, + "GAACCA": 3117, + "GAACCT": 3118, + "GAACCC": 3119, + "GAACCG": 3120, + "GAACGA": 3121, + "GAACGT": 3122, + "GAACGC": 3123, + "GAACGG": 3124, + "GAAGAA": 3125, + "GAAGAT": 3126, + "GAAGAC": 3127, + "GAAGAG": 3128, + "GAAGTA": 3129, + "GAAGTT": 3130, + "GAAGTC": 3131, + "GAAGTG": 3132, + "GAAGCA": 3133, + "GAAGCT": 3134, + "GAAGCC": 3135, + "GAAGCG": 3136, + "GAAGGA": 3137, + "GAAGGT": 3138, + "GAAGGC": 3139, + "GAAGGG": 3140, + "GATAAA": 3141, + "GATAAT": 3142, + "GATAAC": 3143, + "GATAAG": 3144, + "GATATA": 3145, + "GATATT": 3146, + "GATATC": 3147, + "GATATG": 3148, + "GATACA": 3149, + "GATACT": 3150, + "GATACC": 3151, + "GATACG": 3152, + "GATAGA": 3153, + "GATAGT": 3154, + "GATAGC": 3155, + "GATAGG": 3156, + "GATTAA": 3157, + "GATTAT": 3158, + "GATTAC": 3159, + "GATTAG": 3160, + "GATTTA": 3161, + "GATTTT": 3162, + "GATTTC": 3163, + "GATTTG": 3164, + "GATTCA": 3165, + "GATTCT": 3166, + "GATTCC": 3167, + "GATTCG": 3168, + "GATTGA": 3169, + "GATTGT": 3170, + "GATTGC": 3171, + "GATTGG": 3172, + "GATCAA": 3173, + "GATCAT": 3174, + "GATCAC": 3175, + "GATCAG": 3176, + "GATCTA": 3177, + "GATCTT": 3178, + "GATCTC": 3179, + "GATCTG": 3180, + "GATCCA": 3181, + "GATCCT": 3182, + "GATCCC": 3183, + "GATCCG": 3184, + "GATCGA": 3185, + "GATCGT": 3186, + "GATCGC": 3187, + "GATCGG": 3188, + "GATGAA": 3189, + "GATGAT": 3190, + "GATGAC": 3191, + "GATGAG": 3192, + "GATGTA": 3193, + "GATGTT": 3194, + "GATGTC": 3195, + "GATGTG": 3196, + "GATGCA": 3197, + "GATGCT": 3198, + "GATGCC": 3199, + "GATGCG": 3200, + "GATGGA": 3201, + "GATGGT": 3202, + "GATGGC": 3203, + "GATGGG": 3204, + "GACAAA": 3205, + "GACAAT": 3206, + "GACAAC": 3207, + "GACAAG": 3208, + "GACATA": 3209, + "GACATT": 3210, + "GACATC": 3211, + "GACATG": 3212, + "GACACA": 3213, + "GACACT": 3214, + "GACACC": 3215, + "GACACG": 3216, + "GACAGA": 3217, + "GACAGT": 3218, + "GACAGC": 3219, + "GACAGG": 3220, + "GACTAA": 3221, + "GACTAT": 3222, + "GACTAC": 3223, + "GACTAG": 3224, + "GACTTA": 3225, + "GACTTT": 3226, + "GACTTC": 3227, + "GACTTG": 3228, + "GACTCA": 3229, + "GACTCT": 3230, + "GACTCC": 3231, + "GACTCG": 3232, + "GACTGA": 3233, + "GACTGT": 3234, + "GACTGC": 3235, + "GACTGG": 3236, + "GACCAA": 3237, + "GACCAT": 3238, + "GACCAC": 3239, + "GACCAG": 3240, + "GACCTA": 3241, + "GACCTT": 3242, + "GACCTC": 3243, + "GACCTG": 3244, + "GACCCA": 3245, + "GACCCT": 3246, + "GACCCC": 3247, + "GACCCG": 3248, + "GACCGA": 3249, + "GACCGT": 3250, + "GACCGC": 3251, + "GACCGG": 3252, + "GACGAA": 3253, + "GACGAT": 3254, + "GACGAC": 3255, + "GACGAG": 3256, + "GACGTA": 3257, + "GACGTT": 3258, + "GACGTC": 3259, + "GACGTG": 3260, + "GACGCA": 3261, + "GACGCT": 3262, + "GACGCC": 3263, + "GACGCG": 3264, + "GACGGA": 3265, + "GACGGT": 3266, + "GACGGC": 3267, + "GACGGG": 3268, + "GAGAAA": 3269, + "GAGAAT": 3270, + "GAGAAC": 3271, + "GAGAAG": 3272, + "GAGATA": 3273, + "GAGATT": 3274, + "GAGATC": 3275, + "GAGATG": 3276, + "GAGACA": 3277, + "GAGACT": 3278, + "GAGACC": 3279, + "GAGACG": 3280, + "GAGAGA": 3281, + "GAGAGT": 3282, + "GAGAGC": 3283, + "GAGAGG": 3284, + "GAGTAA": 3285, + "GAGTAT": 3286, + "GAGTAC": 3287, + "GAGTAG": 3288, + "GAGTTA": 3289, + "GAGTTT": 3290, + "GAGTTC": 3291, + "GAGTTG": 3292, + "GAGTCA": 3293, + "GAGTCT": 3294, + "GAGTCC": 3295, + "GAGTCG": 3296, + "GAGTGA": 3297, + "GAGTGT": 3298, + "GAGTGC": 3299, + "GAGTGG": 3300, + "GAGCAA": 3301, + "GAGCAT": 3302, + "GAGCAC": 3303, + "GAGCAG": 3304, + "GAGCTA": 3305, + "GAGCTT": 3306, + "GAGCTC": 3307, + "GAGCTG": 3308, + "GAGCCA": 3309, + "GAGCCT": 3310, + "GAGCCC": 3311, + "GAGCCG": 3312, + "GAGCGA": 3313, + "GAGCGT": 3314, + "GAGCGC": 3315, + "GAGCGG": 3316, + "GAGGAA": 3317, + "GAGGAT": 3318, + "GAGGAC": 3319, + "GAGGAG": 3320, + "GAGGTA": 3321, + "GAGGTT": 3322, + "GAGGTC": 3323, + "GAGGTG": 3324, + "GAGGCA": 3325, + "GAGGCT": 3326, + "GAGGCC": 3327, + "GAGGCG": 3328, + "GAGGGA": 3329, + "GAGGGT": 3330, + "GAGGGC": 3331, + "GAGGGG": 3332, + "GTAAAA": 3333, + "GTAAAT": 3334, + "GTAAAC": 3335, + "GTAAAG": 3336, + "GTAATA": 3337, + "GTAATT": 3338, + "GTAATC": 3339, + "GTAATG": 3340, + "GTAACA": 3341, + "GTAACT": 3342, + "GTAACC": 3343, + "GTAACG": 3344, + "GTAAGA": 3345, + "GTAAGT": 3346, + "GTAAGC": 3347, + "GTAAGG": 3348, + "GTATAA": 3349, + "GTATAT": 3350, + "GTATAC": 3351, + "GTATAG": 3352, + "GTATTA": 3353, + "GTATTT": 3354, + "GTATTC": 3355, + "GTATTG": 3356, + "GTATCA": 3357, + "GTATCT": 3358, + "GTATCC": 3359, + "GTATCG": 3360, + "GTATGA": 3361, + "GTATGT": 3362, + "GTATGC": 3363, + "GTATGG": 3364, + "GTACAA": 3365, + "GTACAT": 3366, + "GTACAC": 3367, + "GTACAG": 3368, + "GTACTA": 3369, + "GTACTT": 3370, + "GTACTC": 3371, + "GTACTG": 3372, + "GTACCA": 3373, + "GTACCT": 3374, + "GTACCC": 3375, + "GTACCG": 3376, + "GTACGA": 3377, + "GTACGT": 3378, + "GTACGC": 3379, + "GTACGG": 3380, + "GTAGAA": 3381, + "GTAGAT": 3382, + "GTAGAC": 3383, + "GTAGAG": 3384, + "GTAGTA": 3385, + "GTAGTT": 3386, + "GTAGTC": 3387, + "GTAGTG": 3388, + "GTAGCA": 3389, + "GTAGCT": 3390, + "GTAGCC": 3391, + "GTAGCG": 3392, + "GTAGGA": 3393, + "GTAGGT": 3394, + "GTAGGC": 3395, + "GTAGGG": 3396, + "GTTAAA": 3397, + "GTTAAT": 3398, + "GTTAAC": 3399, + "GTTAAG": 3400, + "GTTATA": 3401, + "GTTATT": 3402, + "GTTATC": 3403, + "GTTATG": 3404, + "GTTACA": 3405, + "GTTACT": 3406, + "GTTACC": 3407, + "GTTACG": 3408, + "GTTAGA": 3409, + "GTTAGT": 3410, + "GTTAGC": 3411, + "GTTAGG": 3412, + "GTTTAA": 3413, + "GTTTAT": 3414, + "GTTTAC": 3415, + "GTTTAG": 3416, + "GTTTTA": 3417, + "GTTTTT": 3418, + "GTTTTC": 3419, + "GTTTTG": 3420, + "GTTTCA": 3421, + "GTTTCT": 3422, + "GTTTCC": 3423, + "GTTTCG": 3424, + "GTTTGA": 3425, + "GTTTGT": 3426, + "GTTTGC": 3427, + "GTTTGG": 3428, + "GTTCAA": 3429, + "GTTCAT": 3430, + "GTTCAC": 3431, + "GTTCAG": 3432, + "GTTCTA": 3433, + "GTTCTT": 3434, + "GTTCTC": 3435, + "GTTCTG": 3436, + "GTTCCA": 3437, + "GTTCCT": 3438, + "GTTCCC": 3439, + "GTTCCG": 3440, + "GTTCGA": 3441, + "GTTCGT": 3442, + "GTTCGC": 3443, + "GTTCGG": 3444, + "GTTGAA": 3445, + "GTTGAT": 3446, + "GTTGAC": 3447, + "GTTGAG": 3448, + "GTTGTA": 3449, + "GTTGTT": 3450, + "GTTGTC": 3451, + "GTTGTG": 3452, + "GTTGCA": 3453, + "GTTGCT": 3454, + "GTTGCC": 3455, + "GTTGCG": 3456, + "GTTGGA": 3457, + "GTTGGT": 3458, + "GTTGGC": 3459, + "GTTGGG": 3460, + "GTCAAA": 3461, + "GTCAAT": 3462, + "GTCAAC": 3463, + "GTCAAG": 3464, + "GTCATA": 3465, + "GTCATT": 3466, + "GTCATC": 3467, + "GTCATG": 3468, + "GTCACA": 3469, + "GTCACT": 3470, + "GTCACC": 3471, + "GTCACG": 3472, + "GTCAGA": 3473, + "GTCAGT": 3474, + "GTCAGC": 3475, + "GTCAGG": 3476, + "GTCTAA": 3477, + "GTCTAT": 3478, + "GTCTAC": 3479, + "GTCTAG": 3480, + "GTCTTA": 3481, + "GTCTTT": 3482, + "GTCTTC": 3483, + "GTCTTG": 3484, + "GTCTCA": 3485, + "GTCTCT": 3486, + "GTCTCC": 3487, + "GTCTCG": 3488, + "GTCTGA": 3489, + "GTCTGT": 3490, + "GTCTGC": 3491, + "GTCTGG": 3492, + "GTCCAA": 3493, + "GTCCAT": 3494, + "GTCCAC": 3495, + "GTCCAG": 3496, + "GTCCTA": 3497, + "GTCCTT": 3498, + "GTCCTC": 3499, + "GTCCTG": 3500, + "GTCCCA": 3501, + "GTCCCT": 3502, + "GTCCCC": 3503, + "GTCCCG": 3504, + "GTCCGA": 3505, + "GTCCGT": 3506, + "GTCCGC": 3507, + "GTCCGG": 3508, + "GTCGAA": 3509, + "GTCGAT": 3510, + "GTCGAC": 3511, + "GTCGAG": 3512, + "GTCGTA": 3513, + "GTCGTT": 3514, + "GTCGTC": 3515, + "GTCGTG": 3516, + "GTCGCA": 3517, + "GTCGCT": 3518, + "GTCGCC": 3519, + "GTCGCG": 3520, + "GTCGGA": 3521, + "GTCGGT": 3522, + "GTCGGC": 3523, + "GTCGGG": 3524, + "GTGAAA": 3525, + "GTGAAT": 3526, + "GTGAAC": 3527, + "GTGAAG": 3528, + "GTGATA": 3529, + "GTGATT": 3530, + "GTGATC": 3531, + "GTGATG": 3532, + "GTGACA": 3533, + "GTGACT": 3534, + "GTGACC": 3535, + "GTGACG": 3536, + "GTGAGA": 3537, + "GTGAGT": 3538, + "GTGAGC": 3539, + "GTGAGG": 3540, + "GTGTAA": 3541, + "GTGTAT": 3542, + "GTGTAC": 3543, + "GTGTAG": 3544, + "GTGTTA": 3545, + "GTGTTT": 3546, + "GTGTTC": 3547, + "GTGTTG": 3548, + "GTGTCA": 3549, + "GTGTCT": 3550, + "GTGTCC": 3551, + "GTGTCG": 3552, + "GTGTGA": 3553, + "GTGTGT": 3554, + "GTGTGC": 3555, + "GTGTGG": 3556, + "GTGCAA": 3557, + "GTGCAT": 3558, + "GTGCAC": 3559, + "GTGCAG": 3560, + "GTGCTA": 3561, + "GTGCTT": 3562, + "GTGCTC": 3563, + "GTGCTG": 3564, + "GTGCCA": 3565, + "GTGCCT": 3566, + "GTGCCC": 3567, + "GTGCCG": 3568, + "GTGCGA": 3569, + "GTGCGT": 3570, + "GTGCGC": 3571, + "GTGCGG": 3572, + "GTGGAA": 3573, + "GTGGAT": 3574, + "GTGGAC": 3575, + "GTGGAG": 3576, + "GTGGTA": 3577, + "GTGGTT": 3578, + "GTGGTC": 3579, + "GTGGTG": 3580, + "GTGGCA": 3581, + "GTGGCT": 3582, + "GTGGCC": 3583, + "GTGGCG": 3584, + "GTGGGA": 3585, + "GTGGGT": 3586, + "GTGGGC": 3587, + "GTGGGG": 3588, + "GCAAAA": 3589, + "GCAAAT": 3590, + "GCAAAC": 3591, + "GCAAAG": 3592, + "GCAATA": 3593, + "GCAATT": 3594, + "GCAATC": 3595, + "GCAATG": 3596, + "GCAACA": 3597, + "GCAACT": 3598, + "GCAACC": 3599, + "GCAACG": 3600, + "GCAAGA": 3601, + "GCAAGT": 3602, + "GCAAGC": 3603, + "GCAAGG": 3604, + "GCATAA": 3605, + "GCATAT": 3606, + "GCATAC": 3607, + "GCATAG": 3608, + "GCATTA": 3609, + "GCATTT": 3610, + "GCATTC": 3611, + "GCATTG": 3612, + "GCATCA": 3613, + "GCATCT": 3614, + "GCATCC": 3615, + "GCATCG": 3616, + "GCATGA": 3617, + "GCATGT": 3618, + "GCATGC": 3619, + "GCATGG": 3620, + "GCACAA": 3621, + "GCACAT": 3622, + "GCACAC": 3623, + "GCACAG": 3624, + "GCACTA": 3625, + "GCACTT": 3626, + "GCACTC": 3627, + "GCACTG": 3628, + "GCACCA": 3629, + "GCACCT": 3630, + "GCACCC": 3631, + "GCACCG": 3632, + "GCACGA": 3633, + "GCACGT": 3634, + "GCACGC": 3635, + "GCACGG": 3636, + "GCAGAA": 3637, + "GCAGAT": 3638, + "GCAGAC": 3639, + "GCAGAG": 3640, + "GCAGTA": 3641, + "GCAGTT": 3642, + "GCAGTC": 3643, + "GCAGTG": 3644, + "GCAGCA": 3645, + "GCAGCT": 3646, + "GCAGCC": 3647, + "GCAGCG": 3648, + "GCAGGA": 3649, + "GCAGGT": 3650, + "GCAGGC": 3651, + "GCAGGG": 3652, + "GCTAAA": 3653, + "GCTAAT": 3654, + "GCTAAC": 3655, + "GCTAAG": 3656, + "GCTATA": 3657, + "GCTATT": 3658, + "GCTATC": 3659, + "GCTATG": 3660, + "GCTACA": 3661, + "GCTACT": 3662, + "GCTACC": 3663, + "GCTACG": 3664, + "GCTAGA": 3665, + "GCTAGT": 3666, + "GCTAGC": 3667, + "GCTAGG": 3668, + "GCTTAA": 3669, + "GCTTAT": 3670, + "GCTTAC": 3671, + "GCTTAG": 3672, + "GCTTTA": 3673, + "GCTTTT": 3674, + "GCTTTC": 3675, + "GCTTTG": 3676, + "GCTTCA": 3677, + "GCTTCT": 3678, + "GCTTCC": 3679, + "GCTTCG": 3680, + "GCTTGA": 3681, + "GCTTGT": 3682, + "GCTTGC": 3683, + "GCTTGG": 3684, + "GCTCAA": 3685, + "GCTCAT": 3686, + "GCTCAC": 3687, + "GCTCAG": 3688, + "GCTCTA": 3689, + "GCTCTT": 3690, + "GCTCTC": 3691, + "GCTCTG": 3692, + "GCTCCA": 3693, + "GCTCCT": 3694, + "GCTCCC": 3695, + "GCTCCG": 3696, + "GCTCGA": 3697, + "GCTCGT": 3698, + "GCTCGC": 3699, + "GCTCGG": 3700, + "GCTGAA": 3701, + "GCTGAT": 3702, + "GCTGAC": 3703, + "GCTGAG": 3704, + "GCTGTA": 3705, + "GCTGTT": 3706, + "GCTGTC": 3707, + "GCTGTG": 3708, + "GCTGCA": 3709, + "GCTGCT": 3710, + "GCTGCC": 3711, + "GCTGCG": 3712, + "GCTGGA": 3713, + "GCTGGT": 3714, + "GCTGGC": 3715, + "GCTGGG": 3716, + "GCCAAA": 3717, + "GCCAAT": 3718, + "GCCAAC": 3719, + "GCCAAG": 3720, + "GCCATA": 3721, + "GCCATT": 3722, + "GCCATC": 3723, + "GCCATG": 3724, + "GCCACA": 3725, + "GCCACT": 3726, + "GCCACC": 3727, + "GCCACG": 3728, + "GCCAGA": 3729, + "GCCAGT": 3730, + "GCCAGC": 3731, + "GCCAGG": 3732, + "GCCTAA": 3733, + "GCCTAT": 3734, + "GCCTAC": 3735, + "GCCTAG": 3736, + "GCCTTA": 3737, + "GCCTTT": 3738, + "GCCTTC": 3739, + "GCCTTG": 3740, + "GCCTCA": 3741, + "GCCTCT": 3742, + "GCCTCC": 3743, + "GCCTCG": 3744, + "GCCTGA": 3745, + "GCCTGT": 3746, + "GCCTGC": 3747, + "GCCTGG": 3748, + "GCCCAA": 3749, + "GCCCAT": 3750, + "GCCCAC": 3751, + "GCCCAG": 3752, + "GCCCTA": 3753, + "GCCCTT": 3754, + "GCCCTC": 3755, + "GCCCTG": 3756, + "GCCCCA": 3757, + "GCCCCT": 3758, + "GCCCCC": 3759, + "GCCCCG": 3760, + "GCCCGA": 3761, + "GCCCGT": 3762, + "GCCCGC": 3763, + "GCCCGG": 3764, + "GCCGAA": 3765, + "GCCGAT": 3766, + "GCCGAC": 3767, + "GCCGAG": 3768, + "GCCGTA": 3769, + "GCCGTT": 3770, + "GCCGTC": 3771, + "GCCGTG": 3772, + "GCCGCA": 3773, + "GCCGCT": 3774, + "GCCGCC": 3775, + "GCCGCG": 3776, + "GCCGGA": 3777, + "GCCGGT": 3778, + "GCCGGC": 3779, + "GCCGGG": 3780, + "GCGAAA": 3781, + "GCGAAT": 3782, + "GCGAAC": 3783, + "GCGAAG": 3784, + "GCGATA": 3785, + "GCGATT": 3786, + "GCGATC": 3787, + "GCGATG": 3788, + "GCGACA": 3789, + "GCGACT": 3790, + "GCGACC": 3791, + "GCGACG": 3792, + "GCGAGA": 3793, + "GCGAGT": 3794, + "GCGAGC": 3795, + "GCGAGG": 3796, + "GCGTAA": 3797, + "GCGTAT": 3798, + "GCGTAC": 3799, + "GCGTAG": 3800, + "GCGTTA": 3801, + "GCGTTT": 3802, + "GCGTTC": 3803, + "GCGTTG": 3804, + "GCGTCA": 3805, + "GCGTCT": 3806, + "GCGTCC": 3807, + "GCGTCG": 3808, + "GCGTGA": 3809, + "GCGTGT": 3810, + "GCGTGC": 3811, + "GCGTGG": 3812, + "GCGCAA": 3813, + "GCGCAT": 3814, + "GCGCAC": 3815, + "GCGCAG": 3816, + "GCGCTA": 3817, + "GCGCTT": 3818, + "GCGCTC": 3819, + "GCGCTG": 3820, + "GCGCCA": 3821, + "GCGCCT": 3822, + "GCGCCC": 3823, + "GCGCCG": 3824, + "GCGCGA": 3825, + "GCGCGT": 3826, + "GCGCGC": 3827, + "GCGCGG": 3828, + "GCGGAA": 3829, + "GCGGAT": 3830, + "GCGGAC": 3831, + "GCGGAG": 3832, + "GCGGTA": 3833, + "GCGGTT": 3834, + "GCGGTC": 3835, + "GCGGTG": 3836, + "GCGGCA": 3837, + "GCGGCT": 3838, + "GCGGCC": 3839, + "GCGGCG": 3840, + "GCGGGA": 3841, + "GCGGGT": 3842, + "GCGGGC": 3843, + "GCGGGG": 3844, + "GGAAAA": 3845, + "GGAAAT": 3846, + "GGAAAC": 3847, + "GGAAAG": 3848, + "GGAATA": 3849, + "GGAATT": 3850, + "GGAATC": 3851, + "GGAATG": 3852, + "GGAACA": 3853, + "GGAACT": 3854, + "GGAACC": 3855, + "GGAACG": 3856, + "GGAAGA": 3857, + "GGAAGT": 3858, + "GGAAGC": 3859, + "GGAAGG": 3860, + "GGATAA": 3861, + "GGATAT": 3862, + "GGATAC": 3863, + "GGATAG": 3864, + "GGATTA": 3865, + "GGATTT": 3866, + "GGATTC": 3867, + "GGATTG": 3868, + "GGATCA": 3869, + "GGATCT": 3870, + "GGATCC": 3871, + "GGATCG": 3872, + "GGATGA": 3873, + "GGATGT": 3874, + "GGATGC": 3875, + "GGATGG": 3876, + "GGACAA": 3877, + "GGACAT": 3878, + "GGACAC": 3879, + "GGACAG": 3880, + "GGACTA": 3881, + "GGACTT": 3882, + "GGACTC": 3883, + "GGACTG": 3884, + "GGACCA": 3885, + "GGACCT": 3886, + "GGACCC": 3887, + "GGACCG": 3888, + "GGACGA": 3889, + "GGACGT": 3890, + "GGACGC": 3891, + "GGACGG": 3892, + "GGAGAA": 3893, + "GGAGAT": 3894, + "GGAGAC": 3895, + "GGAGAG": 3896, + "GGAGTA": 3897, + "GGAGTT": 3898, + "GGAGTC": 3899, + "GGAGTG": 3900, + "GGAGCA": 3901, + "GGAGCT": 3902, + "GGAGCC": 3903, + "GGAGCG": 3904, + "GGAGGA": 3905, + "GGAGGT": 3906, + "GGAGGC": 3907, + "GGAGGG": 3908, + "GGTAAA": 3909, + "GGTAAT": 3910, + "GGTAAC": 3911, + "GGTAAG": 3912, + "GGTATA": 3913, + "GGTATT": 3914, + "GGTATC": 3915, + "GGTATG": 3916, + "GGTACA": 3917, + "GGTACT": 3918, + "GGTACC": 3919, + "GGTACG": 3920, + "GGTAGA": 3921, + "GGTAGT": 3922, + "GGTAGC": 3923, + "GGTAGG": 3924, + "GGTTAA": 3925, + "GGTTAT": 3926, + "GGTTAC": 3927, + "GGTTAG": 3928, + "GGTTTA": 3929, + "GGTTTT": 3930, + "GGTTTC": 3931, + "GGTTTG": 3932, + "GGTTCA": 3933, + "GGTTCT": 3934, + "GGTTCC": 3935, + "GGTTCG": 3936, + "GGTTGA": 3937, + "GGTTGT": 3938, + "GGTTGC": 3939, + "GGTTGG": 3940, + "GGTCAA": 3941, + "GGTCAT": 3942, + "GGTCAC": 3943, + "GGTCAG": 3944, + "GGTCTA": 3945, + "GGTCTT": 3946, + "GGTCTC": 3947, + "GGTCTG": 3948, + "GGTCCA": 3949, + "GGTCCT": 3950, + "GGTCCC": 3951, + "GGTCCG": 3952, + "GGTCGA": 3953, + "GGTCGT": 3954, + "GGTCGC": 3955, + "GGTCGG": 3956, + "GGTGAA": 3957, + "GGTGAT": 3958, + "GGTGAC": 3959, + "GGTGAG": 3960, + "GGTGTA": 3961, + "GGTGTT": 3962, + "GGTGTC": 3963, + "GGTGTG": 3964, + "GGTGCA": 3965, + "GGTGCT": 3966, + "GGTGCC": 3967, + "GGTGCG": 3968, + "GGTGGA": 3969, + "GGTGGT": 3970, + "GGTGGC": 3971, + "GGTGGG": 3972, + "GGCAAA": 3973, + "GGCAAT": 3974, + "GGCAAC": 3975, + "GGCAAG": 3976, + "GGCATA": 3977, + "GGCATT": 3978, + "GGCATC": 3979, + "GGCATG": 3980, + "GGCACA": 3981, + "GGCACT": 3982, + "GGCACC": 3983, + "GGCACG": 3984, + "GGCAGA": 3985, + "GGCAGT": 3986, + "GGCAGC": 3987, + "GGCAGG": 3988, + "GGCTAA": 3989, + "GGCTAT": 3990, + "GGCTAC": 3991, + "GGCTAG": 3992, + "GGCTTA": 3993, + "GGCTTT": 3994, + "GGCTTC": 3995, + "GGCTTG": 3996, + "GGCTCA": 3997, + "GGCTCT": 3998, + "GGCTCC": 3999, + "GGCTCG": 4000, + "GGCTGA": 4001, + "GGCTGT": 4002, + "GGCTGC": 4003, + "GGCTGG": 4004, + "GGCCAA": 4005, + "GGCCAT": 4006, + "GGCCAC": 4007, + "GGCCAG": 4008, + "GGCCTA": 4009, + "GGCCTT": 4010, + "GGCCTC": 4011, + "GGCCTG": 4012, + "GGCCCA": 4013, + "GGCCCT": 4014, + "GGCCCC": 4015, + "GGCCCG": 4016, + "GGCCGA": 4017, + "GGCCGT": 4018, + "GGCCGC": 4019, + "GGCCGG": 4020, + "GGCGAA": 4021, + "GGCGAT": 4022, + "GGCGAC": 4023, + "GGCGAG": 4024, + "GGCGTA": 4025, + "GGCGTT": 4026, + "GGCGTC": 4027, + "GGCGTG": 4028, + "GGCGCA": 4029, + "GGCGCT": 4030, + "GGCGCC": 4031, + "GGCGCG": 4032, + "GGCGGA": 4033, + "GGCGGT": 4034, + "GGCGGC": 4035, + "GGCGGG": 4036, + "GGGAAA": 4037, + "GGGAAT": 4038, + "GGGAAC": 4039, + "GGGAAG": 4040, + "GGGATA": 4041, + "GGGATT": 4042, + "GGGATC": 4043, + "GGGATG": 4044, + "GGGACA": 4045, + "GGGACT": 4046, + "GGGACC": 4047, + "GGGACG": 4048, + "GGGAGA": 4049, + "GGGAGT": 4050, + "GGGAGC": 4051, + "GGGAGG": 4052, + "GGGTAA": 4053, + "GGGTAT": 4054, + "GGGTAC": 4055, + "GGGTAG": 4056, + "GGGTTA": 4057, + "GGGTTT": 4058, + "GGGTTC": 4059, + "GGGTTG": 4060, + "GGGTCA": 4061, + "GGGTCT": 4062, + "GGGTCC": 4063, + "GGGTCG": 4064, + "GGGTGA": 4065, + "GGGTGT": 4066, + "GGGTGC": 4067, + "GGGTGG": 4068, + "GGGCAA": 4069, + "GGGCAT": 4070, + "GGGCAC": 4071, + "GGGCAG": 4072, + "GGGCTA": 4073, + "GGGCTT": 4074, + "GGGCTC": 4075, + "GGGCTG": 4076, + "GGGCCA": 4077, + "GGGCCT": 4078, + "GGGCCC": 4079, + "GGGCCG": 4080, + "GGGCGA": 4081, + "GGGCGT": 4082, + "GGGCGC": 4083, + "GGGCGG": 4084, + "GGGGAA": 4085, + "GGGGAT": 4086, + "GGGGAC": 4087, + "GGGGAG": 4088, + "GGGGTA": 4089, + "GGGGTT": 4090, + "GGGGTC": 4091, + "GGGGTG": 4092, + "GGGGCA": 4093, + "GGGGCT": 4094, + "GGGGCC": 4095, + "GGGGCG": 4096, + "GGGGGA": 4097, + "GGGGGT": 4098, + "GGGGGC": 4099, + "GGGGGG": 4100 + } + } +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/tokenizer_config.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..604e2b1b29af1dcb3920d2828b6a90276e5e3d02 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/tokenizer_config.json @@ -0,0 +1,58 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "[UNK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "[CLS]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "[SEP]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "4": { + "content": "[MASK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "clean_up_tokenization_spaces": true, + "cls_token": "[CLS]", + "do_basic_tokenize": true, + "do_lower_case": false, + "mask_token": "[MASK]", + "max_len": 512, + "model_max_length": 512, + "never_split": null, + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "strip_accents": null, + "tokenize_chinese_chars": true, + "tokenizer_class": "BertTokenizer", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/trainer_state.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..317c7143d8d138dedf91137cf85e06c4bd3e4a6d --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/trainer_state.json @@ -0,0 +1,1153 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 16.0, + "eval_steps": 500, + "global_step": 4000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.1, + "grad_norm": 0.9416134357452393, + "learning_rate": 5.319148936170213e-05, + "loss": 0.6912, + "step": 25 + }, + { + "epoch": 0.2, + "grad_norm": 0.9283172488212585, + "learning_rate": 0.00010638297872340425, + "loss": 0.6691, + "step": 50 + }, + { + "epoch": 0.3, + "grad_norm": 0.5870445370674133, + "learning_rate": 0.00015957446808510637, + "loss": 0.6164, + "step": 75 + }, + { + "epoch": 0.4, + "grad_norm": 0.33945509791374207, + "learning_rate": 0.0002127659574468085, + "loss": 0.5453, + "step": 100 + }, + { + "epoch": 0.5, + "grad_norm": 0.21052460372447968, + "learning_rate": 0.00026595744680851064, + "loss": 0.5025, + "step": 125 + }, + { + "epoch": 0.6, + "grad_norm": 0.25948837399482727, + "learning_rate": 0.00031914893617021275, + "loss": 0.4881, + "step": 150 + }, + { + "epoch": 0.7, + "grad_norm": 0.6165992021560669, + "learning_rate": 0.0003723404255319149, + "loss": 0.4852, + "step": 175 + }, + { + "epoch": 0.8, + "grad_norm": 0.33296748995780945, + "learning_rate": 0.0003992081821181128, + "loss": 0.4867, + "step": 200 + }, + { + "epoch": 0.9, + "grad_norm": 0.40745800733566284, + "learning_rate": 0.0003975585615308479, + "loss": 0.4767, + "step": 225 + }, + { + "epoch": 1.0, + "grad_norm": 0.8070425391197205, + "learning_rate": 0.00039590894094358297, + "loss": 0.4784, + "step": 250 + }, + { + "epoch": 1.1, + "grad_norm": 1.2249579429626465, + "learning_rate": 0.0003942593203563181, + "loss": 0.4543, + "step": 275 + }, + { + "epoch": 1.2, + "grad_norm": 1.0143898725509644, + "learning_rate": 0.0003926096997690532, + "loss": 0.4319, + "step": 300 + }, + { + "epoch": 1.3, + "grad_norm": 2.3284952640533447, + "learning_rate": 0.00039096007918178817, + "loss": 0.4455, + "step": 325 + }, + { + "epoch": 1.4, + "grad_norm": 1.189352035522461, + "learning_rate": 0.0003893104585945233, + "loss": 0.4246, + "step": 350 + }, + { + "epoch": 1.5, + "grad_norm": 0.947853684425354, + "learning_rate": 0.0003876608380072583, + "loss": 0.4326, + "step": 375 + }, + { + "epoch": 1.6, + "grad_norm": 0.7130059003829956, + "learning_rate": 0.00038601121741999343, + "loss": 0.4034, + "step": 400 + }, + { + "epoch": 1.7, + "grad_norm": 3.517078161239624, + "learning_rate": 0.0003843615968327285, + "loss": 0.4114, + "step": 425 + }, + { + "epoch": 1.8, + "grad_norm": 0.3982729911804199, + "learning_rate": 0.0003827119762454636, + "loss": 0.4345, + "step": 450 + }, + { + "epoch": 1.9, + "grad_norm": 0.6208440065383911, + "learning_rate": 0.00038106235565819863, + "loss": 0.4135, + "step": 475 + }, + { + "epoch": 2.0, + "grad_norm": 0.9085566401481628, + "learning_rate": 0.0003794127350709337, + "loss": 0.3983, + "step": 500 + }, + { + "epoch": 2.1, + "grad_norm": 1.2225189208984375, + "learning_rate": 0.0003777631144836688, + "loss": 0.3915, + "step": 525 + }, + { + "epoch": 2.2, + "grad_norm": 1.1379530429840088, + "learning_rate": 0.00037611349389640383, + "loss": 0.4335, + "step": 550 + }, + { + "epoch": 2.3, + "grad_norm": 0.8535892367362976, + "learning_rate": 0.00037446387330913894, + "loss": 0.4063, + "step": 575 + }, + { + "epoch": 2.4, + "grad_norm": 1.3587419986724854, + "learning_rate": 0.000372814252721874, + "loss": 0.4065, + "step": 600 + }, + { + "epoch": 2.5, + "grad_norm": 0.9311383962631226, + "learning_rate": 0.00037116463213460903, + "loss": 0.4007, + "step": 625 + }, + { + "epoch": 2.6, + "grad_norm": 0.9912465810775757, + "learning_rate": 0.00036951501154734414, + "loss": 0.3992, + "step": 650 + }, + { + "epoch": 2.7, + "grad_norm": 1.2762119770050049, + "learning_rate": 0.0003678653909600792, + "loss": 0.3827, + "step": 675 + }, + { + "epoch": 2.8, + "grad_norm": 2.3506247997283936, + "learning_rate": 0.0003662157703728143, + "loss": 0.3734, + "step": 700 + }, + { + "epoch": 2.9, + "grad_norm": 1.5375425815582275, + "learning_rate": 0.00036456614978554934, + "loss": 0.4005, + "step": 725 + }, + { + "epoch": 3.0, + "grad_norm": 1.3143396377563477, + "learning_rate": 0.00036291652919828444, + "loss": 0.3548, + "step": 750 + }, + { + "epoch": 3.1, + "grad_norm": 0.662200391292572, + "learning_rate": 0.00036126690861101944, + "loss": 0.4129, + "step": 775 + }, + { + "epoch": 3.2, + "grad_norm": 1.3569878339767456, + "learning_rate": 0.00035961728802375454, + "loss": 0.38, + "step": 800 + }, + { + "epoch": 3.3, + "grad_norm": 0.9888588190078735, + "learning_rate": 0.0003579676674364896, + "loss": 0.3778, + "step": 825 + }, + { + "epoch": 3.4, + "grad_norm": 0.679969310760498, + "learning_rate": 0.0003563180468492247, + "loss": 0.3934, + "step": 850 + }, + { + "epoch": 3.5, + "grad_norm": 1.608539342880249, + "learning_rate": 0.0003546684262619598, + "loss": 0.418, + "step": 875 + }, + { + "epoch": 3.6, + "grad_norm": 1.7037910223007202, + "learning_rate": 0.00035301880567469485, + "loss": 0.3934, + "step": 900 + }, + { + "epoch": 3.7, + "grad_norm": 1.4417132139205933, + "learning_rate": 0.0003513691850874299, + "loss": 0.3947, + "step": 925 + }, + { + "epoch": 3.8, + "grad_norm": 1.1482582092285156, + "learning_rate": 0.00034971956450016495, + "loss": 0.3621, + "step": 950 + }, + { + "epoch": 3.9, + "grad_norm": 2.046616792678833, + "learning_rate": 0.00034806994391290005, + "loss": 0.3774, + "step": 975 + }, + { + "epoch": 4.0, + "grad_norm": 1.0832113027572632, + "learning_rate": 0.0003464203233256351, + "loss": 0.3542, + "step": 1000 + }, + { + "epoch": 4.1, + "grad_norm": 0.8189190626144409, + "learning_rate": 0.0003447707027383702, + "loss": 0.381, + "step": 1025 + }, + { + "epoch": 4.2, + "grad_norm": 1.6974774599075317, + "learning_rate": 0.00034312108215110525, + "loss": 0.3775, + "step": 1050 + }, + { + "epoch": 4.3, + "grad_norm": 1.2468841075897217, + "learning_rate": 0.0003414714615638403, + "loss": 0.3527, + "step": 1075 + }, + { + "epoch": 4.4, + "grad_norm": 0.9022793173789978, + "learning_rate": 0.0003398218409765754, + "loss": 0.3637, + "step": 1100 + }, + { + "epoch": 4.5, + "grad_norm": 1.121774673461914, + "learning_rate": 0.00033817222038931045, + "loss": 0.3935, + "step": 1125 + }, + { + "epoch": 4.6, + "grad_norm": 0.810937762260437, + "learning_rate": 0.00033652259980204556, + "loss": 0.4103, + "step": 1150 + }, + { + "epoch": 4.7, + "grad_norm": 0.7376565337181091, + "learning_rate": 0.0003348729792147806, + "loss": 0.362, + "step": 1175 + }, + { + "epoch": 4.8, + "grad_norm": 1.7271156311035156, + "learning_rate": 0.0003332233586275157, + "loss": 0.3597, + "step": 1200 + }, + { + "epoch": 4.9, + "grad_norm": 1.2273073196411133, + "learning_rate": 0.00033157373804025076, + "loss": 0.3853, + "step": 1225 + }, + { + "epoch": 5.0, + "grad_norm": 0.9998250603675842, + "learning_rate": 0.0003299241174529858, + "loss": 0.3461, + "step": 1250 + }, + { + "epoch": 5.1, + "grad_norm": 0.9112564325332642, + "learning_rate": 0.0003282744968657209, + "loss": 0.3659, + "step": 1275 + }, + { + "epoch": 5.2, + "grad_norm": 3.246051073074341, + "learning_rate": 0.00032662487627845596, + "loss": 0.3488, + "step": 1300 + }, + { + "epoch": 5.3, + "grad_norm": 1.0283201932907104, + "learning_rate": 0.00032497525569119106, + "loss": 0.3528, + "step": 1325 + }, + { + "epoch": 5.4, + "grad_norm": 1.112660527229309, + "learning_rate": 0.0003233256351039261, + "loss": 0.3723, + "step": 1350 + }, + { + "epoch": 5.5, + "grad_norm": 0.9069312810897827, + "learning_rate": 0.00032167601451666116, + "loss": 0.3416, + "step": 1375 + }, + { + "epoch": 5.6, + "grad_norm": 0.710412323474884, + "learning_rate": 0.00032002639392939627, + "loss": 0.3607, + "step": 1400 + }, + { + "epoch": 5.7, + "grad_norm": 1.185612678527832, + "learning_rate": 0.0003183767733421313, + "loss": 0.3777, + "step": 1425 + }, + { + "epoch": 5.8, + "grad_norm": 1.340174913406372, + "learning_rate": 0.0003167271527548664, + "loss": 0.3563, + "step": 1450 + }, + { + "epoch": 5.9, + "grad_norm": 0.9387012124061584, + "learning_rate": 0.00031507753216760147, + "loss": 0.3647, + "step": 1475 + }, + { + "epoch": 6.0, + "grad_norm": 1.1005347967147827, + "learning_rate": 0.00031342791158033657, + "loss": 0.3484, + "step": 1500 + }, + { + "epoch": 6.1, + "grad_norm": 2.64058780670166, + "learning_rate": 0.00031177829099307157, + "loss": 0.3542, + "step": 1525 + }, + { + "epoch": 6.2, + "grad_norm": 1.3007783889770508, + "learning_rate": 0.00031012867040580667, + "loss": 0.3379, + "step": 1550 + }, + { + "epoch": 6.3, + "grad_norm": 3.7190418243408203, + "learning_rate": 0.0003084790498185417, + "loss": 0.3514, + "step": 1575 + }, + { + "epoch": 6.4, + "grad_norm": 1.7263462543487549, + "learning_rate": 0.0003068294292312768, + "loss": 0.3434, + "step": 1600 + }, + { + "epoch": 6.5, + "grad_norm": 3.3360602855682373, + "learning_rate": 0.0003051798086440119, + "loss": 0.3507, + "step": 1625 + }, + { + "epoch": 6.6, + "grad_norm": 1.9620813131332397, + "learning_rate": 0.000303530188056747, + "loss": 0.3455, + "step": 1650 + }, + { + "epoch": 6.7, + "grad_norm": 1.3080660104751587, + "learning_rate": 0.000301880567469482, + "loss": 0.3678, + "step": 1675 + }, + { + "epoch": 6.8, + "grad_norm": 0.6183618307113647, + "learning_rate": 0.0003002309468822171, + "loss": 0.3704, + "step": 1700 + }, + { + "epoch": 6.9, + "grad_norm": 1.0411962270736694, + "learning_rate": 0.0002985813262949522, + "loss": 0.3505, + "step": 1725 + }, + { + "epoch": 7.0, + "grad_norm": 2.040569543838501, + "learning_rate": 0.00029693170570768723, + "loss": 0.3666, + "step": 1750 + }, + { + "epoch": 7.1, + "grad_norm": 0.713647723197937, + "learning_rate": 0.00029528208512042233, + "loss": 0.3529, + "step": 1775 + }, + { + "epoch": 7.2, + "grad_norm": 3.0373058319091797, + "learning_rate": 0.0002936324645331574, + "loss": 0.361, + "step": 1800 + }, + { + "epoch": 7.3, + "grad_norm": 1.819886326789856, + "learning_rate": 0.00029198284394589243, + "loss": 0.3482, + "step": 1825 + }, + { + "epoch": 7.4, + "grad_norm": 0.7930757403373718, + "learning_rate": 0.00029033322335862753, + "loss": 0.3695, + "step": 1850 + }, + { + "epoch": 7.5, + "grad_norm": 0.7889612913131714, + "learning_rate": 0.0002886836027713626, + "loss": 0.3527, + "step": 1875 + }, + { + "epoch": 7.6, + "grad_norm": 2.4007863998413086, + "learning_rate": 0.0002870339821840977, + "loss": 0.3275, + "step": 1900 + }, + { + "epoch": 7.7, + "grad_norm": 2.3483331203460693, + "learning_rate": 0.00028538436159683273, + "loss": 0.3571, + "step": 1925 + }, + { + "epoch": 7.8, + "grad_norm": 0.8911663293838501, + "learning_rate": 0.00028373474100956784, + "loss": 0.3079, + "step": 1950 + }, + { + "epoch": 7.9, + "grad_norm": 1.0521246194839478, + "learning_rate": 0.0002820851204223029, + "loss": 0.3578, + "step": 1975 + }, + { + "epoch": 8.0, + "grad_norm": 1.3923832178115845, + "learning_rate": 0.00028043549983503794, + "loss": 0.3588, + "step": 2000 + }, + { + "epoch": 8.1, + "grad_norm": 1.4007149934768677, + "learning_rate": 0.00027878587924777304, + "loss": 0.3458, + "step": 2025 + }, + { + "epoch": 8.2, + "grad_norm": 1.103198766708374, + "learning_rate": 0.0002771362586605081, + "loss": 0.3736, + "step": 2050 + }, + { + "epoch": 8.3, + "grad_norm": 2.936150550842285, + "learning_rate": 0.0002754866380732432, + "loss": 0.3223, + "step": 2075 + }, + { + "epoch": 8.4, + "grad_norm": 1.907762885093689, + "learning_rate": 0.00027383701748597824, + "loss": 0.3348, + "step": 2100 + }, + { + "epoch": 8.5, + "grad_norm": 1.4422886371612549, + "learning_rate": 0.0002721873968987133, + "loss": 0.3222, + "step": 2125 + }, + { + "epoch": 8.6, + "grad_norm": 1.8203483819961548, + "learning_rate": 0.00027053777631144834, + "loss": 0.3968, + "step": 2150 + }, + { + "epoch": 8.7, + "grad_norm": 0.8046119213104248, + "learning_rate": 0.00026888815572418344, + "loss": 0.3422, + "step": 2175 + }, + { + "epoch": 8.8, + "grad_norm": 1.2612605094909668, + "learning_rate": 0.00026723853513691855, + "loss": 0.3387, + "step": 2200 + }, + { + "epoch": 8.9, + "grad_norm": 1.6534866094589233, + "learning_rate": 0.0002655889145496536, + "loss": 0.3548, + "step": 2225 + }, + { + "epoch": 9.0, + "grad_norm": 1.889664649963379, + "learning_rate": 0.0002639392939623887, + "loss": 0.3367, + "step": 2250 + }, + { + "epoch": 9.1, + "grad_norm": 2.6691551208496094, + "learning_rate": 0.0002622896733751237, + "loss": 0.3628, + "step": 2275 + }, + { + "epoch": 9.2, + "grad_norm": 2.0130255222320557, + "learning_rate": 0.0002606400527878588, + "loss": 0.3366, + "step": 2300 + }, + { + "epoch": 9.3, + "grad_norm": 1.404814600944519, + "learning_rate": 0.00025899043220059385, + "loss": 0.3327, + "step": 2325 + }, + { + "epoch": 9.4, + "grad_norm": 0.9150362610816956, + "learning_rate": 0.00025734081161332895, + "loss": 0.308, + "step": 2350 + }, + { + "epoch": 9.5, + "grad_norm": 1.099320411682129, + "learning_rate": 0.000255691191026064, + "loss": 0.3499, + "step": 2375 + }, + { + "epoch": 9.6, + "grad_norm": 1.502451777458191, + "learning_rate": 0.0002540415704387991, + "loss": 0.3728, + "step": 2400 + }, + { + "epoch": 9.7, + "grad_norm": 1.1115409135818481, + "learning_rate": 0.00025239194985153415, + "loss": 0.348, + "step": 2425 + }, + { + "epoch": 9.8, + "grad_norm": 0.8831422328948975, + "learning_rate": 0.0002507423292642692, + "loss": 0.3392, + "step": 2450 + }, + { + "epoch": 9.9, + "grad_norm": 2.104572057723999, + "learning_rate": 0.0002490927086770043, + "loss": 0.3511, + "step": 2475 + }, + { + "epoch": 10.0, + "grad_norm": 2.4081919193267822, + "learning_rate": 0.00024744308808973936, + "loss": 0.3284, + "step": 2500 + }, + { + "epoch": 10.1, + "grad_norm": 1.9861253499984741, + "learning_rate": 0.00024579346750247446, + "loss": 0.3621, + "step": 2525 + }, + { + "epoch": 10.2, + "grad_norm": 1.9042344093322754, + "learning_rate": 0.0002441438469152095, + "loss": 0.3341, + "step": 2550 + }, + { + "epoch": 10.3, + "grad_norm": 0.5676223635673523, + "learning_rate": 0.00024249422632794456, + "loss": 0.3203, + "step": 2575 + }, + { + "epoch": 10.4, + "grad_norm": 1.5105270147323608, + "learning_rate": 0.00024084460574067963, + "loss": 0.3711, + "step": 2600 + }, + { + "epoch": 10.5, + "grad_norm": 0.9969163537025452, + "learning_rate": 0.0002391949851534147, + "loss": 0.3296, + "step": 2625 + }, + { + "epoch": 10.6, + "grad_norm": 1.297096848487854, + "learning_rate": 0.0002375453645661498, + "loss": 0.349, + "step": 2650 + }, + { + "epoch": 10.7, + "grad_norm": 1.0513092279434204, + "learning_rate": 0.00023589574397888486, + "loss": 0.3308, + "step": 2675 + }, + { + "epoch": 10.8, + "grad_norm": 1.5543969869613647, + "learning_rate": 0.00023424612339161997, + "loss": 0.3289, + "step": 2700 + }, + { + "epoch": 10.9, + "grad_norm": 0.9587257504463196, + "learning_rate": 0.000232596502804355, + "loss": 0.3128, + "step": 2725 + }, + { + "epoch": 11.0, + "grad_norm": 2.1693837642669678, + "learning_rate": 0.00023094688221709007, + "loss": 0.349, + "step": 2750 + }, + { + "epoch": 11.1, + "grad_norm": 1.5624831914901733, + "learning_rate": 0.00022929726162982514, + "loss": 0.3391, + "step": 2775 + }, + { + "epoch": 11.2, + "grad_norm": 1.4600939750671387, + "learning_rate": 0.00022764764104256022, + "loss": 0.3371, + "step": 2800 + }, + { + "epoch": 11.3, + "grad_norm": 1.6139200925827026, + "learning_rate": 0.0002259980204552953, + "loss": 0.3447, + "step": 2825 + }, + { + "epoch": 11.4, + "grad_norm": 2.2703235149383545, + "learning_rate": 0.00022434839986803037, + "loss": 0.3157, + "step": 2850 + }, + { + "epoch": 11.5, + "grad_norm": 2.0282037258148193, + "learning_rate": 0.00022269877928076542, + "loss": 0.3339, + "step": 2875 + }, + { + "epoch": 11.6, + "grad_norm": 1.389129638671875, + "learning_rate": 0.0002210491586935005, + "loss": 0.3195, + "step": 2900 + }, + { + "epoch": 11.7, + "grad_norm": 1.0041213035583496, + "learning_rate": 0.00021939953810623557, + "loss": 0.3382, + "step": 2925 + }, + { + "epoch": 11.8, + "grad_norm": 1.7116957902908325, + "learning_rate": 0.00021774991751897065, + "loss": 0.3261, + "step": 2950 + }, + { + "epoch": 11.9, + "grad_norm": 1.080194354057312, + "learning_rate": 0.00021610029693170573, + "loss": 0.3433, + "step": 2975 + }, + { + "epoch": 12.0, + "grad_norm": 0.6561412811279297, + "learning_rate": 0.0002144506763444408, + "loss": 0.3465, + "step": 3000 + }, + { + "epoch": 12.1, + "grad_norm": 1.5294175148010254, + "learning_rate": 0.00021280105575717582, + "loss": 0.3195, + "step": 3025 + }, + { + "epoch": 12.2, + "grad_norm": 0.6313191652297974, + "learning_rate": 0.00021115143516991093, + "loss": 0.3214, + "step": 3050 + }, + { + "epoch": 12.3, + "grad_norm": 1.4611085653305054, + "learning_rate": 0.000209501814582646, + "loss": 0.3091, + "step": 3075 + }, + { + "epoch": 12.4, + "grad_norm": 1.6590704917907715, + "learning_rate": 0.00020785219399538108, + "loss": 0.3633, + "step": 3100 + }, + { + "epoch": 12.5, + "grad_norm": 0.982976496219635, + "learning_rate": 0.00020620257340811616, + "loss": 0.3688, + "step": 3125 + }, + { + "epoch": 12.6, + "grad_norm": 0.9717169404029846, + "learning_rate": 0.00020455295282085123, + "loss": 0.3532, + "step": 3150 + }, + { + "epoch": 12.7, + "grad_norm": 1.5903713703155518, + "learning_rate": 0.00020290333223358626, + "loss": 0.2906, + "step": 3175 + }, + { + "epoch": 12.8, + "grad_norm": 1.7043166160583496, + "learning_rate": 0.00020125371164632133, + "loss": 0.322, + "step": 3200 + }, + { + "epoch": 12.9, + "grad_norm": 0.7691648006439209, + "learning_rate": 0.0001996040910590564, + "loss": 0.3477, + "step": 3225 + }, + { + "epoch": 13.0, + "grad_norm": 1.1726899147033691, + "learning_rate": 0.00019795447047179148, + "loss": 0.3551, + "step": 3250 + }, + { + "epoch": 13.1, + "grad_norm": 1.0323086977005005, + "learning_rate": 0.0001963048498845266, + "loss": 0.3248, + "step": 3275 + }, + { + "epoch": 13.2, + "grad_norm": 2.920222520828247, + "learning_rate": 0.00019465522929726164, + "loss": 0.3298, + "step": 3300 + }, + { + "epoch": 13.3, + "grad_norm": 1.101644515991211, + "learning_rate": 0.00019300560870999671, + "loss": 0.3011, + "step": 3325 + }, + { + "epoch": 13.4, + "grad_norm": 1.111193299293518, + "learning_rate": 0.0001913559881227318, + "loss": 0.325, + "step": 3350 + }, + { + "epoch": 13.5, + "grad_norm": 1.5074782371520996, + "learning_rate": 0.00018970636753546684, + "loss": 0.3435, + "step": 3375 + }, + { + "epoch": 13.6, + "grad_norm": 1.0705279111862183, + "learning_rate": 0.00018805674694820192, + "loss": 0.3378, + "step": 3400 + }, + { + "epoch": 13.7, + "grad_norm": 1.5590386390686035, + "learning_rate": 0.000186407126360937, + "loss": 0.3349, + "step": 3425 + }, + { + "epoch": 13.8, + "grad_norm": 2.2993671894073486, + "learning_rate": 0.00018475750577367207, + "loss": 0.3479, + "step": 3450 + }, + { + "epoch": 13.9, + "grad_norm": 1.8973578214645386, + "learning_rate": 0.00018310788518640715, + "loss": 0.3358, + "step": 3475 + }, + { + "epoch": 14.0, + "grad_norm": 1.5887267589569092, + "learning_rate": 0.00018145826459914222, + "loss": 0.3095, + "step": 3500 + }, + { + "epoch": 14.1, + "grad_norm": 2.1970202922821045, + "learning_rate": 0.00017980864401187727, + "loss": 0.3435, + "step": 3525 + }, + { + "epoch": 14.2, + "grad_norm": 0.9183377027511597, + "learning_rate": 0.00017815902342461235, + "loss": 0.3454, + "step": 3550 + }, + { + "epoch": 14.3, + "grad_norm": 1.253682017326355, + "learning_rate": 0.00017650940283734742, + "loss": 0.3128, + "step": 3575 + }, + { + "epoch": 14.4, + "grad_norm": 2.4595584869384766, + "learning_rate": 0.00017485978225008247, + "loss": 0.3181, + "step": 3600 + }, + { + "epoch": 14.5, + "grad_norm": 2.0421340465545654, + "learning_rate": 0.00017321016166281755, + "loss": 0.3555, + "step": 3625 + }, + { + "epoch": 14.6, + "grad_norm": 2.0127763748168945, + "learning_rate": 0.00017156054107555263, + "loss": 0.2952, + "step": 3650 + }, + { + "epoch": 14.7, + "grad_norm": 0.9481540322303772, + "learning_rate": 0.0001699109204882877, + "loss": 0.3154, + "step": 3675 + }, + { + "epoch": 14.8, + "grad_norm": 1.5080980062484741, + "learning_rate": 0.00016826129990102278, + "loss": 0.336, + "step": 3700 + }, + { + "epoch": 14.9, + "grad_norm": 1.504712700843811, + "learning_rate": 0.00016661167931375785, + "loss": 0.3511, + "step": 3725 + }, + { + "epoch": 15.0, + "grad_norm": 1.4532283544540405, + "learning_rate": 0.0001649620587264929, + "loss": 0.3216, + "step": 3750 + }, + { + "epoch": 15.1, + "grad_norm": 1.215970754623413, + "learning_rate": 0.00016331243813922798, + "loss": 0.3597, + "step": 3775 + }, + { + "epoch": 15.2, + "grad_norm": 2.054934024810791, + "learning_rate": 0.00016166281755196306, + "loss": 0.3095, + "step": 3800 + }, + { + "epoch": 15.3, + "grad_norm": 1.6162620782852173, + "learning_rate": 0.00016001319696469813, + "loss": 0.3172, + "step": 3825 + }, + { + "epoch": 15.4, + "grad_norm": 2.974353790283203, + "learning_rate": 0.0001583635763774332, + "loss": 0.3186, + "step": 3850 + }, + { + "epoch": 15.5, + "grad_norm": 1.1828879117965698, + "learning_rate": 0.00015671395579016829, + "loss": 0.3141, + "step": 3875 + }, + { + "epoch": 15.6, + "grad_norm": 2.170670509338379, + "learning_rate": 0.00015506433520290334, + "loss": 0.3054, + "step": 3900 + }, + { + "epoch": 15.7, + "grad_norm": 2.5053343772888184, + "learning_rate": 0.0001534147146156384, + "loss": 0.3348, + "step": 3925 + }, + { + "epoch": 15.8, + "grad_norm": 2.761240005493164, + "learning_rate": 0.0001517650940283735, + "loss": 0.3513, + "step": 3950 + }, + { + "epoch": 15.9, + "grad_norm": 2.6656720638275146, + "learning_rate": 0.00015011547344110854, + "loss": 0.3477, + "step": 3975 + }, + { + "epoch": 16.0, + "grad_norm": 1.2667946815490723, + "learning_rate": 0.00014846585285384361, + "loss": 0.3107, + "step": 4000 + } + ], + "logging_steps": 25, + "max_steps": 6250, + "num_input_tokens_seen": 0, + "num_train_epochs": 25, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 212464435200000.0, + "train_batch_size": 20, + "trial_name": null, + "trial_params": null +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/training_args.bin b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..d61f3500eb70763e1f33496c1787bdf3f07a651b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b6e0f7aa622786274a1ea4195b72f4da224050767f7b89529b85721865c5075a +size 5304 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/vocab.txt b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/vocab.txt new file mode 100644 index 0000000000000000000000000000000000000000..cc9ad53732e21ea9180f18e1fb67a38cc5702d4c --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/vocab.txt @@ -0,0 +1,4101 @@ +[PAD] +[UNK] +[CLS] +[SEP] +[MASK] +AAAAAA +AAAAAT +AAAAAC +AAAAAG +AAAATA +AAAATT +AAAATC +AAAATG +AAAACA +AAAACT +AAAACC +AAAACG +AAAAGA +AAAAGT +AAAAGC +AAAAGG +AAATAA +AAATAT +AAATAC +AAATAG +AAATTA +AAATTT +AAATTC +AAATTG +AAATCA +AAATCT +AAATCC +AAATCG +AAATGA +AAATGT +AAATGC +AAATGG +AAACAA +AAACAT +AAACAC +AAACAG +AAACTA +AAACTT +AAACTC +AAACTG +AAACCA +AAACCT +AAACCC +AAACCG +AAACGA +AAACGT +AAACGC +AAACGG +AAAGAA +AAAGAT +AAAGAC +AAAGAG +AAAGTA +AAAGTT +AAAGTC +AAAGTG +AAAGCA +AAAGCT +AAAGCC +AAAGCG +AAAGGA +AAAGGT +AAAGGC +AAAGGG +AATAAA +AATAAT +AATAAC +AATAAG +AATATA +AATATT +AATATC +AATATG +AATACA +AATACT +AATACC +AATACG +AATAGA +AATAGT +AATAGC +AATAGG +AATTAA +AATTAT +AATTAC +AATTAG +AATTTA +AATTTT +AATTTC +AATTTG +AATTCA +AATTCT +AATTCC +AATTCG +AATTGA +AATTGT +AATTGC +AATTGG +AATCAA +AATCAT +AATCAC +AATCAG +AATCTA +AATCTT +AATCTC +AATCTG +AATCCA +AATCCT +AATCCC +AATCCG +AATCGA +AATCGT +AATCGC +AATCGG +AATGAA +AATGAT +AATGAC +AATGAG +AATGTA +AATGTT +AATGTC +AATGTG +AATGCA +AATGCT +AATGCC +AATGCG +AATGGA +AATGGT +AATGGC +AATGGG +AACAAA +AACAAT +AACAAC +AACAAG +AACATA +AACATT +AACATC +AACATG +AACACA +AACACT +AACACC +AACACG +AACAGA +AACAGT +AACAGC +AACAGG +AACTAA +AACTAT +AACTAC +AACTAG +AACTTA +AACTTT +AACTTC +AACTTG +AACTCA +AACTCT +AACTCC +AACTCG +AACTGA +AACTGT +AACTGC +AACTGG +AACCAA +AACCAT +AACCAC +AACCAG +AACCTA +AACCTT +AACCTC +AACCTG +AACCCA +AACCCT +AACCCC +AACCCG +AACCGA +AACCGT +AACCGC +AACCGG +AACGAA +AACGAT +AACGAC +AACGAG +AACGTA +AACGTT +AACGTC +AACGTG +AACGCA +AACGCT +AACGCC +AACGCG +AACGGA +AACGGT +AACGGC +AACGGG +AAGAAA +AAGAAT +AAGAAC +AAGAAG +AAGATA +AAGATT +AAGATC +AAGATG +AAGACA +AAGACT +AAGACC +AAGACG +AAGAGA +AAGAGT +AAGAGC +AAGAGG +AAGTAA +AAGTAT +AAGTAC +AAGTAG +AAGTTA +AAGTTT +AAGTTC +AAGTTG +AAGTCA +AAGTCT +AAGTCC +AAGTCG +AAGTGA +AAGTGT +AAGTGC +AAGTGG +AAGCAA +AAGCAT +AAGCAC +AAGCAG +AAGCTA +AAGCTT +AAGCTC +AAGCTG +AAGCCA +AAGCCT +AAGCCC +AAGCCG +AAGCGA +AAGCGT +AAGCGC +AAGCGG +AAGGAA +AAGGAT +AAGGAC +AAGGAG +AAGGTA +AAGGTT +AAGGTC +AAGGTG +AAGGCA +AAGGCT +AAGGCC +AAGGCG +AAGGGA +AAGGGT +AAGGGC +AAGGGG +ATAAAA +ATAAAT +ATAAAC +ATAAAG +ATAATA +ATAATT +ATAATC +ATAATG +ATAACA +ATAACT +ATAACC +ATAACG +ATAAGA +ATAAGT +ATAAGC +ATAAGG +ATATAA +ATATAT +ATATAC +ATATAG +ATATTA +ATATTT +ATATTC +ATATTG +ATATCA +ATATCT +ATATCC +ATATCG +ATATGA +ATATGT +ATATGC +ATATGG +ATACAA +ATACAT +ATACAC +ATACAG +ATACTA +ATACTT +ATACTC +ATACTG +ATACCA +ATACCT +ATACCC +ATACCG +ATACGA +ATACGT +ATACGC +ATACGG +ATAGAA +ATAGAT +ATAGAC +ATAGAG +ATAGTA +ATAGTT +ATAGTC +ATAGTG +ATAGCA +ATAGCT +ATAGCC +ATAGCG +ATAGGA +ATAGGT +ATAGGC +ATAGGG +ATTAAA +ATTAAT +ATTAAC +ATTAAG +ATTATA +ATTATT +ATTATC +ATTATG +ATTACA +ATTACT +ATTACC +ATTACG +ATTAGA +ATTAGT +ATTAGC +ATTAGG +ATTTAA +ATTTAT +ATTTAC +ATTTAG +ATTTTA +ATTTTT +ATTTTC +ATTTTG +ATTTCA +ATTTCT +ATTTCC +ATTTCG +ATTTGA +ATTTGT +ATTTGC +ATTTGG +ATTCAA +ATTCAT +ATTCAC +ATTCAG +ATTCTA +ATTCTT +ATTCTC +ATTCTG +ATTCCA +ATTCCT +ATTCCC +ATTCCG +ATTCGA +ATTCGT +ATTCGC +ATTCGG +ATTGAA +ATTGAT +ATTGAC +ATTGAG +ATTGTA +ATTGTT +ATTGTC +ATTGTG +ATTGCA +ATTGCT +ATTGCC +ATTGCG +ATTGGA +ATTGGT +ATTGGC +ATTGGG +ATCAAA +ATCAAT +ATCAAC +ATCAAG +ATCATA +ATCATT +ATCATC +ATCATG +ATCACA +ATCACT +ATCACC +ATCACG +ATCAGA +ATCAGT +ATCAGC +ATCAGG +ATCTAA +ATCTAT +ATCTAC +ATCTAG +ATCTTA +ATCTTT +ATCTTC +ATCTTG +ATCTCA +ATCTCT +ATCTCC +ATCTCG +ATCTGA +ATCTGT +ATCTGC +ATCTGG +ATCCAA +ATCCAT +ATCCAC +ATCCAG +ATCCTA +ATCCTT +ATCCTC +ATCCTG +ATCCCA +ATCCCT +ATCCCC +ATCCCG +ATCCGA +ATCCGT +ATCCGC +ATCCGG +ATCGAA +ATCGAT +ATCGAC +ATCGAG +ATCGTA +ATCGTT +ATCGTC +ATCGTG +ATCGCA +ATCGCT +ATCGCC +ATCGCG +ATCGGA +ATCGGT +ATCGGC +ATCGGG +ATGAAA +ATGAAT +ATGAAC +ATGAAG +ATGATA +ATGATT +ATGATC +ATGATG +ATGACA +ATGACT +ATGACC +ATGACG +ATGAGA +ATGAGT +ATGAGC +ATGAGG +ATGTAA +ATGTAT +ATGTAC +ATGTAG +ATGTTA +ATGTTT +ATGTTC +ATGTTG +ATGTCA +ATGTCT +ATGTCC +ATGTCG +ATGTGA +ATGTGT +ATGTGC +ATGTGG +ATGCAA +ATGCAT +ATGCAC +ATGCAG +ATGCTA +ATGCTT +ATGCTC +ATGCTG +ATGCCA +ATGCCT +ATGCCC +ATGCCG +ATGCGA +ATGCGT +ATGCGC +ATGCGG +ATGGAA +ATGGAT +ATGGAC +ATGGAG +ATGGTA +ATGGTT +ATGGTC +ATGGTG +ATGGCA +ATGGCT +ATGGCC +ATGGCG +ATGGGA +ATGGGT +ATGGGC +ATGGGG +ACAAAA +ACAAAT +ACAAAC +ACAAAG +ACAATA +ACAATT +ACAATC +ACAATG +ACAACA +ACAACT +ACAACC +ACAACG +ACAAGA +ACAAGT +ACAAGC +ACAAGG +ACATAA +ACATAT +ACATAC +ACATAG +ACATTA +ACATTT +ACATTC +ACATTG +ACATCA +ACATCT +ACATCC +ACATCG +ACATGA +ACATGT +ACATGC +ACATGG +ACACAA +ACACAT +ACACAC +ACACAG +ACACTA +ACACTT +ACACTC +ACACTG +ACACCA +ACACCT +ACACCC +ACACCG +ACACGA +ACACGT +ACACGC +ACACGG +ACAGAA +ACAGAT +ACAGAC +ACAGAG +ACAGTA +ACAGTT +ACAGTC +ACAGTG +ACAGCA +ACAGCT +ACAGCC +ACAGCG +ACAGGA +ACAGGT +ACAGGC +ACAGGG +ACTAAA +ACTAAT +ACTAAC +ACTAAG +ACTATA +ACTATT +ACTATC +ACTATG +ACTACA +ACTACT +ACTACC +ACTACG +ACTAGA +ACTAGT +ACTAGC +ACTAGG +ACTTAA +ACTTAT +ACTTAC +ACTTAG +ACTTTA +ACTTTT +ACTTTC +ACTTTG +ACTTCA +ACTTCT +ACTTCC +ACTTCG +ACTTGA +ACTTGT +ACTTGC +ACTTGG +ACTCAA +ACTCAT +ACTCAC +ACTCAG +ACTCTA +ACTCTT +ACTCTC +ACTCTG +ACTCCA +ACTCCT +ACTCCC +ACTCCG +ACTCGA +ACTCGT +ACTCGC +ACTCGG +ACTGAA +ACTGAT +ACTGAC +ACTGAG +ACTGTA +ACTGTT +ACTGTC +ACTGTG +ACTGCA +ACTGCT +ACTGCC +ACTGCG +ACTGGA +ACTGGT +ACTGGC +ACTGGG +ACCAAA +ACCAAT +ACCAAC +ACCAAG +ACCATA +ACCATT +ACCATC +ACCATG +ACCACA +ACCACT +ACCACC +ACCACG +ACCAGA +ACCAGT +ACCAGC +ACCAGG +ACCTAA +ACCTAT +ACCTAC +ACCTAG +ACCTTA +ACCTTT +ACCTTC +ACCTTG +ACCTCA +ACCTCT +ACCTCC +ACCTCG +ACCTGA +ACCTGT +ACCTGC +ACCTGG +ACCCAA +ACCCAT +ACCCAC +ACCCAG +ACCCTA +ACCCTT +ACCCTC +ACCCTG +ACCCCA +ACCCCT +ACCCCC +ACCCCG +ACCCGA +ACCCGT +ACCCGC +ACCCGG +ACCGAA +ACCGAT +ACCGAC +ACCGAG +ACCGTA +ACCGTT +ACCGTC +ACCGTG +ACCGCA +ACCGCT +ACCGCC +ACCGCG +ACCGGA +ACCGGT +ACCGGC +ACCGGG +ACGAAA +ACGAAT +ACGAAC +ACGAAG +ACGATA +ACGATT +ACGATC +ACGATG +ACGACA +ACGACT +ACGACC +ACGACG +ACGAGA +ACGAGT +ACGAGC +ACGAGG +ACGTAA +ACGTAT +ACGTAC +ACGTAG +ACGTTA +ACGTTT +ACGTTC +ACGTTG +ACGTCA +ACGTCT +ACGTCC +ACGTCG +ACGTGA +ACGTGT +ACGTGC +ACGTGG +ACGCAA +ACGCAT +ACGCAC +ACGCAG +ACGCTA +ACGCTT +ACGCTC +ACGCTG +ACGCCA +ACGCCT +ACGCCC +ACGCCG +ACGCGA +ACGCGT +ACGCGC +ACGCGG +ACGGAA +ACGGAT +ACGGAC +ACGGAG +ACGGTA +ACGGTT +ACGGTC +ACGGTG +ACGGCA +ACGGCT +ACGGCC +ACGGCG +ACGGGA +ACGGGT +ACGGGC +ACGGGG +AGAAAA +AGAAAT +AGAAAC +AGAAAG +AGAATA +AGAATT +AGAATC +AGAATG +AGAACA +AGAACT +AGAACC +AGAACG +AGAAGA +AGAAGT +AGAAGC +AGAAGG +AGATAA +AGATAT +AGATAC +AGATAG +AGATTA +AGATTT +AGATTC +AGATTG +AGATCA +AGATCT +AGATCC +AGATCG +AGATGA +AGATGT +AGATGC +AGATGG +AGACAA +AGACAT +AGACAC +AGACAG +AGACTA +AGACTT +AGACTC +AGACTG +AGACCA +AGACCT +AGACCC +AGACCG +AGACGA +AGACGT +AGACGC +AGACGG +AGAGAA +AGAGAT +AGAGAC +AGAGAG +AGAGTA +AGAGTT +AGAGTC +AGAGTG +AGAGCA +AGAGCT +AGAGCC +AGAGCG +AGAGGA +AGAGGT +AGAGGC +AGAGGG +AGTAAA +AGTAAT +AGTAAC +AGTAAG +AGTATA +AGTATT +AGTATC +AGTATG +AGTACA +AGTACT +AGTACC +AGTACG +AGTAGA +AGTAGT +AGTAGC +AGTAGG +AGTTAA +AGTTAT +AGTTAC +AGTTAG +AGTTTA +AGTTTT +AGTTTC +AGTTTG +AGTTCA +AGTTCT +AGTTCC +AGTTCG +AGTTGA +AGTTGT +AGTTGC +AGTTGG +AGTCAA +AGTCAT +AGTCAC +AGTCAG +AGTCTA +AGTCTT +AGTCTC +AGTCTG +AGTCCA +AGTCCT +AGTCCC +AGTCCG +AGTCGA +AGTCGT +AGTCGC +AGTCGG +AGTGAA +AGTGAT +AGTGAC +AGTGAG +AGTGTA +AGTGTT +AGTGTC +AGTGTG +AGTGCA +AGTGCT +AGTGCC +AGTGCG +AGTGGA +AGTGGT +AGTGGC +AGTGGG +AGCAAA +AGCAAT +AGCAAC +AGCAAG +AGCATA +AGCATT +AGCATC +AGCATG +AGCACA +AGCACT +AGCACC +AGCACG +AGCAGA +AGCAGT +AGCAGC +AGCAGG +AGCTAA +AGCTAT +AGCTAC +AGCTAG +AGCTTA +AGCTTT +AGCTTC +AGCTTG +AGCTCA +AGCTCT +AGCTCC +AGCTCG +AGCTGA +AGCTGT +AGCTGC +AGCTGG +AGCCAA +AGCCAT +AGCCAC +AGCCAG +AGCCTA +AGCCTT +AGCCTC +AGCCTG +AGCCCA +AGCCCT +AGCCCC +AGCCCG +AGCCGA +AGCCGT +AGCCGC +AGCCGG +AGCGAA +AGCGAT +AGCGAC +AGCGAG +AGCGTA +AGCGTT +AGCGTC +AGCGTG +AGCGCA +AGCGCT +AGCGCC +AGCGCG +AGCGGA +AGCGGT +AGCGGC +AGCGGG +AGGAAA +AGGAAT +AGGAAC +AGGAAG +AGGATA +AGGATT +AGGATC +AGGATG +AGGACA +AGGACT +AGGACC +AGGACG +AGGAGA +AGGAGT +AGGAGC +AGGAGG +AGGTAA +AGGTAT +AGGTAC +AGGTAG +AGGTTA +AGGTTT +AGGTTC +AGGTTG +AGGTCA +AGGTCT +AGGTCC +AGGTCG +AGGTGA +AGGTGT +AGGTGC +AGGTGG +AGGCAA +AGGCAT +AGGCAC +AGGCAG +AGGCTA +AGGCTT +AGGCTC +AGGCTG +AGGCCA +AGGCCT +AGGCCC +AGGCCG +AGGCGA +AGGCGT +AGGCGC +AGGCGG +AGGGAA +AGGGAT +AGGGAC +AGGGAG +AGGGTA +AGGGTT +AGGGTC +AGGGTG +AGGGCA +AGGGCT +AGGGCC +AGGGCG +AGGGGA +AGGGGT +AGGGGC +AGGGGG +TAAAAA +TAAAAT +TAAAAC +TAAAAG +TAAATA +TAAATT +TAAATC +TAAATG +TAAACA +TAAACT +TAAACC +TAAACG +TAAAGA +TAAAGT +TAAAGC +TAAAGG +TAATAA +TAATAT +TAATAC +TAATAG +TAATTA +TAATTT +TAATTC +TAATTG +TAATCA +TAATCT +TAATCC +TAATCG +TAATGA +TAATGT +TAATGC +TAATGG +TAACAA +TAACAT +TAACAC +TAACAG +TAACTA +TAACTT +TAACTC +TAACTG +TAACCA +TAACCT +TAACCC +TAACCG +TAACGA +TAACGT +TAACGC +TAACGG +TAAGAA +TAAGAT +TAAGAC +TAAGAG +TAAGTA +TAAGTT +TAAGTC +TAAGTG +TAAGCA +TAAGCT +TAAGCC +TAAGCG +TAAGGA +TAAGGT +TAAGGC +TAAGGG +TATAAA +TATAAT +TATAAC +TATAAG +TATATA +TATATT +TATATC +TATATG +TATACA +TATACT +TATACC +TATACG +TATAGA +TATAGT +TATAGC +TATAGG +TATTAA +TATTAT +TATTAC +TATTAG +TATTTA +TATTTT +TATTTC +TATTTG +TATTCA +TATTCT +TATTCC +TATTCG +TATTGA +TATTGT +TATTGC +TATTGG +TATCAA +TATCAT +TATCAC +TATCAG +TATCTA +TATCTT +TATCTC +TATCTG +TATCCA +TATCCT +TATCCC +TATCCG +TATCGA +TATCGT +TATCGC +TATCGG +TATGAA +TATGAT +TATGAC +TATGAG +TATGTA +TATGTT +TATGTC +TATGTG +TATGCA +TATGCT +TATGCC +TATGCG +TATGGA +TATGGT +TATGGC +TATGGG +TACAAA +TACAAT +TACAAC +TACAAG +TACATA +TACATT +TACATC +TACATG +TACACA +TACACT +TACACC +TACACG +TACAGA +TACAGT +TACAGC +TACAGG +TACTAA +TACTAT +TACTAC +TACTAG +TACTTA +TACTTT +TACTTC +TACTTG +TACTCA +TACTCT +TACTCC +TACTCG +TACTGA +TACTGT +TACTGC +TACTGG +TACCAA +TACCAT +TACCAC +TACCAG +TACCTA +TACCTT +TACCTC +TACCTG +TACCCA +TACCCT +TACCCC +TACCCG +TACCGA +TACCGT +TACCGC +TACCGG +TACGAA +TACGAT +TACGAC +TACGAG +TACGTA +TACGTT +TACGTC +TACGTG +TACGCA +TACGCT +TACGCC +TACGCG +TACGGA +TACGGT +TACGGC +TACGGG +TAGAAA +TAGAAT +TAGAAC +TAGAAG +TAGATA +TAGATT +TAGATC +TAGATG +TAGACA +TAGACT +TAGACC +TAGACG +TAGAGA +TAGAGT +TAGAGC +TAGAGG +TAGTAA +TAGTAT +TAGTAC +TAGTAG +TAGTTA +TAGTTT +TAGTTC +TAGTTG +TAGTCA +TAGTCT +TAGTCC +TAGTCG +TAGTGA +TAGTGT +TAGTGC +TAGTGG +TAGCAA +TAGCAT +TAGCAC +TAGCAG +TAGCTA +TAGCTT +TAGCTC +TAGCTG +TAGCCA +TAGCCT +TAGCCC +TAGCCG +TAGCGA +TAGCGT +TAGCGC +TAGCGG +TAGGAA +TAGGAT +TAGGAC +TAGGAG +TAGGTA +TAGGTT +TAGGTC +TAGGTG +TAGGCA +TAGGCT +TAGGCC +TAGGCG +TAGGGA +TAGGGT +TAGGGC +TAGGGG +TTAAAA +TTAAAT +TTAAAC +TTAAAG +TTAATA +TTAATT +TTAATC +TTAATG +TTAACA +TTAACT +TTAACC +TTAACG +TTAAGA +TTAAGT +TTAAGC +TTAAGG +TTATAA +TTATAT +TTATAC +TTATAG +TTATTA +TTATTT +TTATTC +TTATTG +TTATCA +TTATCT +TTATCC +TTATCG +TTATGA +TTATGT +TTATGC +TTATGG +TTACAA +TTACAT +TTACAC +TTACAG +TTACTA +TTACTT +TTACTC +TTACTG +TTACCA +TTACCT +TTACCC +TTACCG +TTACGA +TTACGT +TTACGC +TTACGG +TTAGAA +TTAGAT +TTAGAC +TTAGAG +TTAGTA +TTAGTT +TTAGTC +TTAGTG +TTAGCA +TTAGCT +TTAGCC +TTAGCG +TTAGGA +TTAGGT +TTAGGC +TTAGGG +TTTAAA +TTTAAT +TTTAAC +TTTAAG +TTTATA +TTTATT +TTTATC +TTTATG +TTTACA +TTTACT +TTTACC +TTTACG +TTTAGA +TTTAGT +TTTAGC +TTTAGG +TTTTAA +TTTTAT +TTTTAC +TTTTAG +TTTTTA +TTTTTT +TTTTTC +TTTTTG +TTTTCA +TTTTCT +TTTTCC +TTTTCG +TTTTGA +TTTTGT +TTTTGC +TTTTGG +TTTCAA +TTTCAT +TTTCAC +TTTCAG +TTTCTA +TTTCTT +TTTCTC +TTTCTG +TTTCCA +TTTCCT +TTTCCC +TTTCCG +TTTCGA +TTTCGT +TTTCGC +TTTCGG +TTTGAA +TTTGAT +TTTGAC +TTTGAG +TTTGTA +TTTGTT +TTTGTC +TTTGTG +TTTGCA +TTTGCT +TTTGCC +TTTGCG +TTTGGA +TTTGGT +TTTGGC +TTTGGG +TTCAAA +TTCAAT +TTCAAC +TTCAAG +TTCATA +TTCATT +TTCATC +TTCATG +TTCACA +TTCACT +TTCACC +TTCACG +TTCAGA +TTCAGT +TTCAGC +TTCAGG +TTCTAA +TTCTAT +TTCTAC +TTCTAG +TTCTTA +TTCTTT +TTCTTC +TTCTTG +TTCTCA +TTCTCT +TTCTCC +TTCTCG +TTCTGA +TTCTGT +TTCTGC +TTCTGG +TTCCAA +TTCCAT +TTCCAC +TTCCAG +TTCCTA +TTCCTT +TTCCTC +TTCCTG +TTCCCA +TTCCCT +TTCCCC +TTCCCG +TTCCGA +TTCCGT +TTCCGC +TTCCGG +TTCGAA +TTCGAT +TTCGAC +TTCGAG +TTCGTA +TTCGTT +TTCGTC +TTCGTG +TTCGCA +TTCGCT +TTCGCC +TTCGCG +TTCGGA +TTCGGT +TTCGGC +TTCGGG +TTGAAA +TTGAAT +TTGAAC +TTGAAG +TTGATA +TTGATT +TTGATC +TTGATG +TTGACA +TTGACT +TTGACC +TTGACG +TTGAGA +TTGAGT +TTGAGC +TTGAGG +TTGTAA +TTGTAT +TTGTAC +TTGTAG +TTGTTA +TTGTTT +TTGTTC +TTGTTG +TTGTCA +TTGTCT +TTGTCC +TTGTCG +TTGTGA +TTGTGT +TTGTGC +TTGTGG +TTGCAA +TTGCAT +TTGCAC +TTGCAG +TTGCTA +TTGCTT +TTGCTC +TTGCTG +TTGCCA +TTGCCT +TTGCCC +TTGCCG +TTGCGA +TTGCGT +TTGCGC +TTGCGG +TTGGAA +TTGGAT +TTGGAC +TTGGAG +TTGGTA +TTGGTT +TTGGTC +TTGGTG +TTGGCA +TTGGCT +TTGGCC +TTGGCG +TTGGGA +TTGGGT +TTGGGC +TTGGGG +TCAAAA +TCAAAT +TCAAAC +TCAAAG +TCAATA +TCAATT +TCAATC +TCAATG +TCAACA +TCAACT +TCAACC +TCAACG +TCAAGA +TCAAGT +TCAAGC +TCAAGG +TCATAA +TCATAT +TCATAC +TCATAG +TCATTA +TCATTT +TCATTC +TCATTG +TCATCA +TCATCT +TCATCC +TCATCG +TCATGA +TCATGT +TCATGC +TCATGG +TCACAA +TCACAT +TCACAC +TCACAG +TCACTA +TCACTT +TCACTC +TCACTG +TCACCA +TCACCT +TCACCC +TCACCG +TCACGA +TCACGT +TCACGC +TCACGG +TCAGAA +TCAGAT +TCAGAC +TCAGAG +TCAGTA +TCAGTT +TCAGTC +TCAGTG +TCAGCA +TCAGCT +TCAGCC +TCAGCG +TCAGGA +TCAGGT +TCAGGC +TCAGGG +TCTAAA +TCTAAT +TCTAAC +TCTAAG +TCTATA +TCTATT +TCTATC +TCTATG +TCTACA +TCTACT +TCTACC +TCTACG +TCTAGA +TCTAGT +TCTAGC +TCTAGG +TCTTAA +TCTTAT +TCTTAC +TCTTAG +TCTTTA +TCTTTT +TCTTTC +TCTTTG +TCTTCA +TCTTCT +TCTTCC +TCTTCG +TCTTGA +TCTTGT +TCTTGC +TCTTGG +TCTCAA +TCTCAT +TCTCAC +TCTCAG +TCTCTA +TCTCTT +TCTCTC +TCTCTG +TCTCCA +TCTCCT +TCTCCC +TCTCCG +TCTCGA +TCTCGT +TCTCGC +TCTCGG +TCTGAA +TCTGAT +TCTGAC +TCTGAG +TCTGTA +TCTGTT +TCTGTC +TCTGTG +TCTGCA +TCTGCT +TCTGCC +TCTGCG +TCTGGA +TCTGGT +TCTGGC +TCTGGG +TCCAAA +TCCAAT +TCCAAC +TCCAAG +TCCATA +TCCATT +TCCATC +TCCATG +TCCACA +TCCACT +TCCACC +TCCACG +TCCAGA +TCCAGT +TCCAGC +TCCAGG +TCCTAA +TCCTAT +TCCTAC +TCCTAG +TCCTTA +TCCTTT +TCCTTC +TCCTTG +TCCTCA +TCCTCT +TCCTCC +TCCTCG +TCCTGA +TCCTGT +TCCTGC +TCCTGG +TCCCAA +TCCCAT +TCCCAC +TCCCAG +TCCCTA +TCCCTT +TCCCTC +TCCCTG +TCCCCA +TCCCCT +TCCCCC +TCCCCG +TCCCGA +TCCCGT +TCCCGC +TCCCGG +TCCGAA +TCCGAT +TCCGAC +TCCGAG +TCCGTA +TCCGTT +TCCGTC +TCCGTG +TCCGCA +TCCGCT +TCCGCC +TCCGCG +TCCGGA +TCCGGT +TCCGGC +TCCGGG +TCGAAA +TCGAAT +TCGAAC +TCGAAG +TCGATA +TCGATT +TCGATC +TCGATG +TCGACA +TCGACT +TCGACC +TCGACG +TCGAGA +TCGAGT +TCGAGC +TCGAGG +TCGTAA +TCGTAT +TCGTAC +TCGTAG +TCGTTA +TCGTTT +TCGTTC +TCGTTG +TCGTCA +TCGTCT +TCGTCC +TCGTCG +TCGTGA +TCGTGT +TCGTGC +TCGTGG +TCGCAA +TCGCAT +TCGCAC +TCGCAG +TCGCTA +TCGCTT +TCGCTC +TCGCTG +TCGCCA +TCGCCT +TCGCCC +TCGCCG +TCGCGA +TCGCGT +TCGCGC +TCGCGG +TCGGAA +TCGGAT +TCGGAC +TCGGAG +TCGGTA +TCGGTT +TCGGTC +TCGGTG +TCGGCA +TCGGCT +TCGGCC +TCGGCG +TCGGGA +TCGGGT +TCGGGC +TCGGGG +TGAAAA +TGAAAT +TGAAAC +TGAAAG +TGAATA +TGAATT +TGAATC +TGAATG +TGAACA +TGAACT +TGAACC +TGAACG +TGAAGA +TGAAGT +TGAAGC +TGAAGG +TGATAA +TGATAT +TGATAC +TGATAG +TGATTA +TGATTT +TGATTC +TGATTG +TGATCA +TGATCT +TGATCC +TGATCG +TGATGA +TGATGT +TGATGC +TGATGG +TGACAA +TGACAT +TGACAC +TGACAG +TGACTA +TGACTT +TGACTC +TGACTG +TGACCA +TGACCT +TGACCC +TGACCG +TGACGA +TGACGT +TGACGC +TGACGG +TGAGAA +TGAGAT +TGAGAC +TGAGAG +TGAGTA +TGAGTT +TGAGTC +TGAGTG +TGAGCA +TGAGCT +TGAGCC +TGAGCG +TGAGGA +TGAGGT +TGAGGC +TGAGGG +TGTAAA +TGTAAT +TGTAAC +TGTAAG +TGTATA +TGTATT +TGTATC +TGTATG +TGTACA +TGTACT +TGTACC +TGTACG +TGTAGA +TGTAGT +TGTAGC +TGTAGG +TGTTAA +TGTTAT +TGTTAC +TGTTAG +TGTTTA +TGTTTT +TGTTTC +TGTTTG +TGTTCA +TGTTCT +TGTTCC +TGTTCG +TGTTGA +TGTTGT +TGTTGC +TGTTGG +TGTCAA +TGTCAT +TGTCAC +TGTCAG +TGTCTA +TGTCTT +TGTCTC +TGTCTG +TGTCCA +TGTCCT +TGTCCC +TGTCCG +TGTCGA +TGTCGT +TGTCGC +TGTCGG +TGTGAA +TGTGAT +TGTGAC +TGTGAG +TGTGTA +TGTGTT +TGTGTC +TGTGTG +TGTGCA +TGTGCT +TGTGCC +TGTGCG +TGTGGA +TGTGGT +TGTGGC +TGTGGG +TGCAAA +TGCAAT +TGCAAC +TGCAAG +TGCATA +TGCATT +TGCATC +TGCATG +TGCACA +TGCACT +TGCACC +TGCACG +TGCAGA +TGCAGT +TGCAGC +TGCAGG +TGCTAA +TGCTAT +TGCTAC +TGCTAG +TGCTTA +TGCTTT +TGCTTC +TGCTTG +TGCTCA +TGCTCT +TGCTCC +TGCTCG +TGCTGA +TGCTGT +TGCTGC +TGCTGG +TGCCAA +TGCCAT +TGCCAC +TGCCAG +TGCCTA +TGCCTT +TGCCTC +TGCCTG +TGCCCA +TGCCCT +TGCCCC +TGCCCG +TGCCGA +TGCCGT +TGCCGC +TGCCGG +TGCGAA +TGCGAT +TGCGAC +TGCGAG +TGCGTA +TGCGTT +TGCGTC +TGCGTG +TGCGCA +TGCGCT +TGCGCC +TGCGCG +TGCGGA +TGCGGT +TGCGGC +TGCGGG +TGGAAA +TGGAAT +TGGAAC +TGGAAG +TGGATA +TGGATT +TGGATC +TGGATG +TGGACA +TGGACT +TGGACC +TGGACG +TGGAGA +TGGAGT +TGGAGC +TGGAGG +TGGTAA +TGGTAT +TGGTAC +TGGTAG +TGGTTA +TGGTTT +TGGTTC +TGGTTG +TGGTCA +TGGTCT +TGGTCC +TGGTCG +TGGTGA +TGGTGT +TGGTGC +TGGTGG +TGGCAA +TGGCAT +TGGCAC +TGGCAG +TGGCTA +TGGCTT +TGGCTC +TGGCTG +TGGCCA +TGGCCT +TGGCCC +TGGCCG +TGGCGA +TGGCGT +TGGCGC +TGGCGG +TGGGAA +TGGGAT +TGGGAC +TGGGAG +TGGGTA +TGGGTT +TGGGTC +TGGGTG +TGGGCA +TGGGCT +TGGGCC +TGGGCG +TGGGGA +TGGGGT +TGGGGC +TGGGGG +CAAAAA +CAAAAT +CAAAAC +CAAAAG +CAAATA +CAAATT +CAAATC +CAAATG +CAAACA +CAAACT +CAAACC +CAAACG +CAAAGA +CAAAGT +CAAAGC +CAAAGG +CAATAA +CAATAT +CAATAC +CAATAG +CAATTA +CAATTT +CAATTC +CAATTG +CAATCA +CAATCT +CAATCC +CAATCG +CAATGA +CAATGT +CAATGC +CAATGG +CAACAA +CAACAT +CAACAC +CAACAG +CAACTA +CAACTT +CAACTC +CAACTG +CAACCA +CAACCT +CAACCC +CAACCG +CAACGA +CAACGT +CAACGC +CAACGG +CAAGAA +CAAGAT +CAAGAC +CAAGAG +CAAGTA +CAAGTT +CAAGTC +CAAGTG +CAAGCA +CAAGCT +CAAGCC +CAAGCG +CAAGGA +CAAGGT +CAAGGC +CAAGGG +CATAAA +CATAAT +CATAAC +CATAAG +CATATA +CATATT +CATATC +CATATG +CATACA +CATACT +CATACC +CATACG +CATAGA +CATAGT +CATAGC +CATAGG +CATTAA +CATTAT +CATTAC +CATTAG +CATTTA +CATTTT +CATTTC +CATTTG +CATTCA +CATTCT +CATTCC +CATTCG +CATTGA +CATTGT +CATTGC +CATTGG +CATCAA +CATCAT +CATCAC +CATCAG +CATCTA +CATCTT +CATCTC +CATCTG +CATCCA +CATCCT +CATCCC +CATCCG +CATCGA +CATCGT +CATCGC +CATCGG +CATGAA +CATGAT +CATGAC +CATGAG +CATGTA +CATGTT +CATGTC +CATGTG +CATGCA +CATGCT +CATGCC +CATGCG +CATGGA +CATGGT +CATGGC +CATGGG +CACAAA +CACAAT +CACAAC +CACAAG +CACATA +CACATT +CACATC +CACATG +CACACA +CACACT +CACACC +CACACG +CACAGA +CACAGT +CACAGC +CACAGG +CACTAA +CACTAT +CACTAC +CACTAG +CACTTA +CACTTT +CACTTC +CACTTG +CACTCA +CACTCT +CACTCC +CACTCG +CACTGA +CACTGT +CACTGC +CACTGG +CACCAA +CACCAT +CACCAC +CACCAG +CACCTA +CACCTT +CACCTC +CACCTG +CACCCA +CACCCT +CACCCC +CACCCG +CACCGA +CACCGT +CACCGC +CACCGG +CACGAA +CACGAT +CACGAC +CACGAG +CACGTA +CACGTT +CACGTC +CACGTG +CACGCA +CACGCT +CACGCC +CACGCG +CACGGA +CACGGT +CACGGC +CACGGG +CAGAAA +CAGAAT +CAGAAC +CAGAAG +CAGATA +CAGATT +CAGATC +CAGATG +CAGACA +CAGACT +CAGACC +CAGACG +CAGAGA +CAGAGT +CAGAGC +CAGAGG +CAGTAA +CAGTAT +CAGTAC +CAGTAG +CAGTTA +CAGTTT +CAGTTC +CAGTTG +CAGTCA +CAGTCT +CAGTCC +CAGTCG +CAGTGA +CAGTGT +CAGTGC +CAGTGG +CAGCAA +CAGCAT +CAGCAC +CAGCAG +CAGCTA +CAGCTT +CAGCTC +CAGCTG +CAGCCA +CAGCCT +CAGCCC +CAGCCG +CAGCGA +CAGCGT +CAGCGC +CAGCGG +CAGGAA +CAGGAT +CAGGAC +CAGGAG +CAGGTA +CAGGTT +CAGGTC +CAGGTG +CAGGCA +CAGGCT +CAGGCC +CAGGCG +CAGGGA +CAGGGT +CAGGGC +CAGGGG +CTAAAA +CTAAAT +CTAAAC +CTAAAG +CTAATA +CTAATT +CTAATC +CTAATG +CTAACA +CTAACT +CTAACC +CTAACG +CTAAGA +CTAAGT +CTAAGC +CTAAGG +CTATAA +CTATAT +CTATAC +CTATAG +CTATTA +CTATTT +CTATTC +CTATTG +CTATCA +CTATCT +CTATCC +CTATCG +CTATGA +CTATGT +CTATGC +CTATGG +CTACAA +CTACAT +CTACAC +CTACAG +CTACTA +CTACTT +CTACTC +CTACTG +CTACCA +CTACCT +CTACCC +CTACCG +CTACGA +CTACGT +CTACGC +CTACGG +CTAGAA +CTAGAT +CTAGAC +CTAGAG +CTAGTA +CTAGTT +CTAGTC +CTAGTG +CTAGCA +CTAGCT +CTAGCC +CTAGCG +CTAGGA +CTAGGT +CTAGGC +CTAGGG +CTTAAA +CTTAAT +CTTAAC +CTTAAG +CTTATA +CTTATT +CTTATC +CTTATG +CTTACA +CTTACT +CTTACC +CTTACG +CTTAGA +CTTAGT +CTTAGC +CTTAGG +CTTTAA +CTTTAT +CTTTAC +CTTTAG +CTTTTA +CTTTTT +CTTTTC +CTTTTG +CTTTCA +CTTTCT +CTTTCC +CTTTCG +CTTTGA +CTTTGT +CTTTGC +CTTTGG +CTTCAA +CTTCAT +CTTCAC +CTTCAG +CTTCTA +CTTCTT +CTTCTC +CTTCTG +CTTCCA +CTTCCT +CTTCCC +CTTCCG +CTTCGA +CTTCGT +CTTCGC +CTTCGG +CTTGAA +CTTGAT +CTTGAC +CTTGAG +CTTGTA +CTTGTT +CTTGTC +CTTGTG +CTTGCA +CTTGCT +CTTGCC +CTTGCG +CTTGGA +CTTGGT +CTTGGC +CTTGGG +CTCAAA +CTCAAT +CTCAAC +CTCAAG +CTCATA +CTCATT +CTCATC +CTCATG +CTCACA +CTCACT +CTCACC +CTCACG +CTCAGA +CTCAGT +CTCAGC +CTCAGG +CTCTAA +CTCTAT +CTCTAC +CTCTAG +CTCTTA +CTCTTT +CTCTTC +CTCTTG +CTCTCA +CTCTCT +CTCTCC +CTCTCG +CTCTGA +CTCTGT +CTCTGC +CTCTGG +CTCCAA +CTCCAT +CTCCAC +CTCCAG +CTCCTA +CTCCTT +CTCCTC +CTCCTG +CTCCCA +CTCCCT +CTCCCC +CTCCCG +CTCCGA +CTCCGT +CTCCGC +CTCCGG +CTCGAA +CTCGAT +CTCGAC +CTCGAG +CTCGTA +CTCGTT +CTCGTC +CTCGTG +CTCGCA +CTCGCT +CTCGCC +CTCGCG +CTCGGA +CTCGGT +CTCGGC +CTCGGG +CTGAAA +CTGAAT +CTGAAC +CTGAAG +CTGATA +CTGATT +CTGATC +CTGATG +CTGACA +CTGACT +CTGACC +CTGACG +CTGAGA +CTGAGT +CTGAGC +CTGAGG +CTGTAA +CTGTAT +CTGTAC +CTGTAG +CTGTTA +CTGTTT +CTGTTC +CTGTTG +CTGTCA +CTGTCT +CTGTCC +CTGTCG +CTGTGA +CTGTGT +CTGTGC +CTGTGG +CTGCAA +CTGCAT +CTGCAC +CTGCAG +CTGCTA +CTGCTT +CTGCTC +CTGCTG +CTGCCA +CTGCCT +CTGCCC +CTGCCG +CTGCGA +CTGCGT +CTGCGC +CTGCGG +CTGGAA +CTGGAT +CTGGAC +CTGGAG +CTGGTA +CTGGTT +CTGGTC +CTGGTG +CTGGCA +CTGGCT +CTGGCC +CTGGCG +CTGGGA +CTGGGT +CTGGGC +CTGGGG +CCAAAA +CCAAAT +CCAAAC +CCAAAG +CCAATA +CCAATT +CCAATC +CCAATG +CCAACA +CCAACT +CCAACC +CCAACG +CCAAGA +CCAAGT +CCAAGC +CCAAGG +CCATAA +CCATAT +CCATAC +CCATAG +CCATTA +CCATTT +CCATTC +CCATTG +CCATCA +CCATCT +CCATCC +CCATCG +CCATGA +CCATGT +CCATGC +CCATGG +CCACAA +CCACAT +CCACAC +CCACAG +CCACTA +CCACTT +CCACTC +CCACTG +CCACCA +CCACCT +CCACCC +CCACCG +CCACGA +CCACGT +CCACGC +CCACGG +CCAGAA +CCAGAT +CCAGAC +CCAGAG +CCAGTA +CCAGTT +CCAGTC +CCAGTG +CCAGCA +CCAGCT +CCAGCC +CCAGCG +CCAGGA +CCAGGT +CCAGGC +CCAGGG +CCTAAA +CCTAAT +CCTAAC +CCTAAG +CCTATA +CCTATT +CCTATC +CCTATG +CCTACA +CCTACT +CCTACC +CCTACG +CCTAGA +CCTAGT +CCTAGC +CCTAGG +CCTTAA +CCTTAT +CCTTAC +CCTTAG +CCTTTA +CCTTTT +CCTTTC +CCTTTG +CCTTCA +CCTTCT +CCTTCC +CCTTCG +CCTTGA +CCTTGT +CCTTGC +CCTTGG +CCTCAA +CCTCAT +CCTCAC +CCTCAG +CCTCTA +CCTCTT +CCTCTC +CCTCTG +CCTCCA +CCTCCT +CCTCCC +CCTCCG +CCTCGA +CCTCGT +CCTCGC +CCTCGG +CCTGAA +CCTGAT +CCTGAC +CCTGAG +CCTGTA +CCTGTT +CCTGTC +CCTGTG +CCTGCA +CCTGCT +CCTGCC +CCTGCG +CCTGGA +CCTGGT +CCTGGC +CCTGGG +CCCAAA +CCCAAT +CCCAAC +CCCAAG +CCCATA +CCCATT +CCCATC +CCCATG +CCCACA +CCCACT +CCCACC +CCCACG +CCCAGA +CCCAGT +CCCAGC +CCCAGG +CCCTAA +CCCTAT +CCCTAC +CCCTAG +CCCTTA +CCCTTT +CCCTTC +CCCTTG +CCCTCA +CCCTCT +CCCTCC +CCCTCG +CCCTGA +CCCTGT +CCCTGC +CCCTGG +CCCCAA +CCCCAT +CCCCAC +CCCCAG +CCCCTA +CCCCTT +CCCCTC +CCCCTG +CCCCCA +CCCCCT +CCCCCC +CCCCCG +CCCCGA +CCCCGT +CCCCGC +CCCCGG +CCCGAA +CCCGAT +CCCGAC +CCCGAG +CCCGTA +CCCGTT +CCCGTC +CCCGTG +CCCGCA +CCCGCT +CCCGCC +CCCGCG +CCCGGA +CCCGGT +CCCGGC +CCCGGG +CCGAAA +CCGAAT +CCGAAC +CCGAAG +CCGATA +CCGATT +CCGATC +CCGATG +CCGACA +CCGACT +CCGACC +CCGACG +CCGAGA +CCGAGT +CCGAGC +CCGAGG +CCGTAA +CCGTAT +CCGTAC +CCGTAG +CCGTTA +CCGTTT +CCGTTC +CCGTTG +CCGTCA +CCGTCT +CCGTCC +CCGTCG +CCGTGA +CCGTGT +CCGTGC +CCGTGG +CCGCAA +CCGCAT +CCGCAC +CCGCAG +CCGCTA +CCGCTT +CCGCTC +CCGCTG +CCGCCA +CCGCCT +CCGCCC +CCGCCG +CCGCGA +CCGCGT +CCGCGC +CCGCGG +CCGGAA +CCGGAT +CCGGAC +CCGGAG +CCGGTA +CCGGTT +CCGGTC +CCGGTG +CCGGCA +CCGGCT +CCGGCC +CCGGCG +CCGGGA +CCGGGT +CCGGGC +CCGGGG +CGAAAA +CGAAAT +CGAAAC +CGAAAG +CGAATA +CGAATT +CGAATC +CGAATG +CGAACA +CGAACT +CGAACC +CGAACG +CGAAGA +CGAAGT +CGAAGC +CGAAGG +CGATAA +CGATAT +CGATAC +CGATAG +CGATTA +CGATTT +CGATTC +CGATTG +CGATCA +CGATCT +CGATCC +CGATCG +CGATGA +CGATGT +CGATGC +CGATGG +CGACAA +CGACAT +CGACAC +CGACAG +CGACTA +CGACTT +CGACTC +CGACTG +CGACCA +CGACCT +CGACCC +CGACCG +CGACGA +CGACGT +CGACGC +CGACGG +CGAGAA +CGAGAT +CGAGAC +CGAGAG +CGAGTA +CGAGTT +CGAGTC +CGAGTG +CGAGCA +CGAGCT +CGAGCC +CGAGCG +CGAGGA +CGAGGT +CGAGGC +CGAGGG +CGTAAA +CGTAAT +CGTAAC +CGTAAG +CGTATA +CGTATT +CGTATC +CGTATG +CGTACA +CGTACT +CGTACC +CGTACG +CGTAGA +CGTAGT +CGTAGC +CGTAGG +CGTTAA +CGTTAT +CGTTAC +CGTTAG +CGTTTA +CGTTTT +CGTTTC +CGTTTG +CGTTCA +CGTTCT +CGTTCC +CGTTCG +CGTTGA +CGTTGT +CGTTGC +CGTTGG +CGTCAA +CGTCAT +CGTCAC +CGTCAG +CGTCTA +CGTCTT +CGTCTC +CGTCTG +CGTCCA +CGTCCT +CGTCCC +CGTCCG +CGTCGA +CGTCGT +CGTCGC +CGTCGG +CGTGAA +CGTGAT +CGTGAC +CGTGAG +CGTGTA +CGTGTT +CGTGTC +CGTGTG +CGTGCA +CGTGCT +CGTGCC +CGTGCG +CGTGGA +CGTGGT +CGTGGC +CGTGGG +CGCAAA +CGCAAT +CGCAAC +CGCAAG +CGCATA +CGCATT +CGCATC +CGCATG +CGCACA +CGCACT +CGCACC +CGCACG +CGCAGA +CGCAGT +CGCAGC +CGCAGG +CGCTAA +CGCTAT +CGCTAC +CGCTAG +CGCTTA +CGCTTT +CGCTTC +CGCTTG +CGCTCA +CGCTCT +CGCTCC +CGCTCG +CGCTGA +CGCTGT +CGCTGC +CGCTGG +CGCCAA +CGCCAT +CGCCAC +CGCCAG +CGCCTA +CGCCTT +CGCCTC +CGCCTG +CGCCCA +CGCCCT +CGCCCC +CGCCCG +CGCCGA +CGCCGT +CGCCGC +CGCCGG +CGCGAA +CGCGAT +CGCGAC +CGCGAG +CGCGTA +CGCGTT +CGCGTC +CGCGTG +CGCGCA +CGCGCT +CGCGCC +CGCGCG +CGCGGA +CGCGGT +CGCGGC +CGCGGG +CGGAAA +CGGAAT +CGGAAC +CGGAAG +CGGATA +CGGATT +CGGATC +CGGATG +CGGACA +CGGACT +CGGACC +CGGACG +CGGAGA +CGGAGT +CGGAGC +CGGAGG +CGGTAA +CGGTAT +CGGTAC +CGGTAG +CGGTTA +CGGTTT +CGGTTC +CGGTTG +CGGTCA +CGGTCT +CGGTCC +CGGTCG +CGGTGA +CGGTGT +CGGTGC +CGGTGG +CGGCAA +CGGCAT +CGGCAC +CGGCAG +CGGCTA +CGGCTT +CGGCTC +CGGCTG +CGGCCA +CGGCCT +CGGCCC +CGGCCG +CGGCGA +CGGCGT +CGGCGC +CGGCGG +CGGGAA +CGGGAT +CGGGAC +CGGGAG +CGGGTA +CGGGTT +CGGGTC +CGGGTG +CGGGCA +CGGGCT +CGGGCC +CGGGCG +CGGGGA +CGGGGT +CGGGGC +CGGGGG +GAAAAA +GAAAAT +GAAAAC +GAAAAG +GAAATA +GAAATT +GAAATC +GAAATG +GAAACA +GAAACT +GAAACC +GAAACG +GAAAGA +GAAAGT +GAAAGC +GAAAGG +GAATAA +GAATAT +GAATAC +GAATAG +GAATTA +GAATTT +GAATTC +GAATTG +GAATCA +GAATCT +GAATCC +GAATCG +GAATGA +GAATGT +GAATGC +GAATGG +GAACAA +GAACAT +GAACAC +GAACAG +GAACTA +GAACTT +GAACTC +GAACTG +GAACCA +GAACCT +GAACCC +GAACCG +GAACGA +GAACGT +GAACGC +GAACGG +GAAGAA +GAAGAT +GAAGAC +GAAGAG +GAAGTA +GAAGTT +GAAGTC +GAAGTG +GAAGCA +GAAGCT +GAAGCC +GAAGCG +GAAGGA +GAAGGT +GAAGGC +GAAGGG +GATAAA +GATAAT +GATAAC +GATAAG +GATATA +GATATT +GATATC +GATATG +GATACA +GATACT +GATACC +GATACG +GATAGA +GATAGT +GATAGC +GATAGG +GATTAA +GATTAT +GATTAC +GATTAG +GATTTA +GATTTT +GATTTC +GATTTG +GATTCA +GATTCT +GATTCC +GATTCG +GATTGA +GATTGT +GATTGC +GATTGG +GATCAA +GATCAT +GATCAC +GATCAG +GATCTA +GATCTT +GATCTC +GATCTG +GATCCA +GATCCT +GATCCC +GATCCG +GATCGA +GATCGT +GATCGC +GATCGG +GATGAA +GATGAT +GATGAC +GATGAG +GATGTA +GATGTT +GATGTC +GATGTG +GATGCA +GATGCT +GATGCC +GATGCG +GATGGA +GATGGT +GATGGC +GATGGG +GACAAA +GACAAT +GACAAC +GACAAG +GACATA +GACATT +GACATC +GACATG +GACACA +GACACT +GACACC +GACACG +GACAGA +GACAGT +GACAGC +GACAGG +GACTAA +GACTAT +GACTAC +GACTAG +GACTTA +GACTTT +GACTTC +GACTTG +GACTCA +GACTCT +GACTCC +GACTCG +GACTGA +GACTGT +GACTGC +GACTGG +GACCAA +GACCAT +GACCAC +GACCAG +GACCTA +GACCTT +GACCTC +GACCTG +GACCCA +GACCCT +GACCCC +GACCCG +GACCGA +GACCGT +GACCGC +GACCGG +GACGAA +GACGAT +GACGAC +GACGAG +GACGTA +GACGTT +GACGTC +GACGTG +GACGCA +GACGCT +GACGCC +GACGCG +GACGGA +GACGGT +GACGGC +GACGGG +GAGAAA +GAGAAT +GAGAAC +GAGAAG +GAGATA +GAGATT +GAGATC +GAGATG +GAGACA +GAGACT +GAGACC +GAGACG +GAGAGA +GAGAGT +GAGAGC +GAGAGG +GAGTAA +GAGTAT +GAGTAC +GAGTAG +GAGTTA +GAGTTT +GAGTTC +GAGTTG +GAGTCA +GAGTCT +GAGTCC +GAGTCG +GAGTGA +GAGTGT +GAGTGC +GAGTGG +GAGCAA +GAGCAT +GAGCAC +GAGCAG +GAGCTA +GAGCTT +GAGCTC +GAGCTG +GAGCCA +GAGCCT +GAGCCC +GAGCCG +GAGCGA +GAGCGT +GAGCGC +GAGCGG +GAGGAA +GAGGAT +GAGGAC +GAGGAG +GAGGTA +GAGGTT +GAGGTC +GAGGTG +GAGGCA +GAGGCT +GAGGCC +GAGGCG +GAGGGA +GAGGGT +GAGGGC +GAGGGG +GTAAAA +GTAAAT +GTAAAC +GTAAAG +GTAATA +GTAATT +GTAATC +GTAATG +GTAACA +GTAACT +GTAACC +GTAACG +GTAAGA +GTAAGT +GTAAGC +GTAAGG +GTATAA +GTATAT +GTATAC +GTATAG +GTATTA +GTATTT +GTATTC +GTATTG +GTATCA +GTATCT +GTATCC +GTATCG +GTATGA +GTATGT +GTATGC +GTATGG +GTACAA +GTACAT +GTACAC +GTACAG +GTACTA +GTACTT +GTACTC +GTACTG +GTACCA +GTACCT +GTACCC +GTACCG +GTACGA +GTACGT +GTACGC +GTACGG +GTAGAA +GTAGAT +GTAGAC +GTAGAG +GTAGTA +GTAGTT +GTAGTC +GTAGTG +GTAGCA +GTAGCT +GTAGCC +GTAGCG +GTAGGA +GTAGGT +GTAGGC +GTAGGG +GTTAAA +GTTAAT +GTTAAC +GTTAAG +GTTATA +GTTATT +GTTATC +GTTATG +GTTACA +GTTACT +GTTACC +GTTACG +GTTAGA +GTTAGT +GTTAGC +GTTAGG +GTTTAA +GTTTAT +GTTTAC +GTTTAG +GTTTTA +GTTTTT +GTTTTC +GTTTTG +GTTTCA +GTTTCT +GTTTCC +GTTTCG +GTTTGA +GTTTGT +GTTTGC +GTTTGG +GTTCAA +GTTCAT +GTTCAC +GTTCAG +GTTCTA +GTTCTT +GTTCTC +GTTCTG +GTTCCA +GTTCCT +GTTCCC +GTTCCG +GTTCGA +GTTCGT +GTTCGC +GTTCGG +GTTGAA +GTTGAT +GTTGAC +GTTGAG +GTTGTA +GTTGTT +GTTGTC +GTTGTG +GTTGCA +GTTGCT +GTTGCC +GTTGCG +GTTGGA +GTTGGT +GTTGGC +GTTGGG +GTCAAA +GTCAAT +GTCAAC +GTCAAG +GTCATA +GTCATT +GTCATC +GTCATG +GTCACA +GTCACT +GTCACC +GTCACG +GTCAGA +GTCAGT +GTCAGC +GTCAGG +GTCTAA +GTCTAT +GTCTAC +GTCTAG +GTCTTA +GTCTTT +GTCTTC +GTCTTG +GTCTCA +GTCTCT +GTCTCC +GTCTCG +GTCTGA +GTCTGT +GTCTGC +GTCTGG +GTCCAA +GTCCAT +GTCCAC +GTCCAG +GTCCTA +GTCCTT +GTCCTC +GTCCTG +GTCCCA +GTCCCT +GTCCCC +GTCCCG +GTCCGA +GTCCGT +GTCCGC +GTCCGG +GTCGAA +GTCGAT +GTCGAC +GTCGAG +GTCGTA +GTCGTT +GTCGTC +GTCGTG +GTCGCA +GTCGCT +GTCGCC +GTCGCG +GTCGGA +GTCGGT +GTCGGC +GTCGGG +GTGAAA +GTGAAT +GTGAAC +GTGAAG +GTGATA +GTGATT +GTGATC +GTGATG +GTGACA +GTGACT +GTGACC +GTGACG +GTGAGA +GTGAGT +GTGAGC +GTGAGG +GTGTAA +GTGTAT +GTGTAC +GTGTAG +GTGTTA +GTGTTT +GTGTTC +GTGTTG +GTGTCA +GTGTCT +GTGTCC +GTGTCG +GTGTGA +GTGTGT +GTGTGC +GTGTGG +GTGCAA +GTGCAT +GTGCAC +GTGCAG +GTGCTA +GTGCTT +GTGCTC +GTGCTG +GTGCCA +GTGCCT +GTGCCC +GTGCCG +GTGCGA +GTGCGT +GTGCGC +GTGCGG +GTGGAA +GTGGAT +GTGGAC +GTGGAG +GTGGTA +GTGGTT +GTGGTC +GTGGTG +GTGGCA +GTGGCT +GTGGCC +GTGGCG +GTGGGA +GTGGGT +GTGGGC +GTGGGG +GCAAAA +GCAAAT +GCAAAC +GCAAAG +GCAATA +GCAATT +GCAATC +GCAATG +GCAACA +GCAACT +GCAACC +GCAACG +GCAAGA +GCAAGT +GCAAGC +GCAAGG +GCATAA +GCATAT +GCATAC +GCATAG +GCATTA +GCATTT +GCATTC +GCATTG +GCATCA +GCATCT +GCATCC +GCATCG +GCATGA +GCATGT +GCATGC +GCATGG +GCACAA +GCACAT +GCACAC +GCACAG +GCACTA +GCACTT +GCACTC +GCACTG +GCACCA +GCACCT +GCACCC +GCACCG +GCACGA +GCACGT +GCACGC +GCACGG +GCAGAA +GCAGAT +GCAGAC +GCAGAG +GCAGTA +GCAGTT +GCAGTC +GCAGTG +GCAGCA +GCAGCT +GCAGCC +GCAGCG +GCAGGA +GCAGGT +GCAGGC +GCAGGG +GCTAAA +GCTAAT +GCTAAC +GCTAAG +GCTATA +GCTATT +GCTATC +GCTATG +GCTACA +GCTACT +GCTACC +GCTACG +GCTAGA +GCTAGT +GCTAGC +GCTAGG +GCTTAA +GCTTAT +GCTTAC +GCTTAG +GCTTTA +GCTTTT +GCTTTC +GCTTTG +GCTTCA +GCTTCT +GCTTCC +GCTTCG +GCTTGA +GCTTGT +GCTTGC +GCTTGG +GCTCAA +GCTCAT +GCTCAC +GCTCAG +GCTCTA +GCTCTT +GCTCTC +GCTCTG +GCTCCA +GCTCCT +GCTCCC +GCTCCG +GCTCGA +GCTCGT +GCTCGC +GCTCGG +GCTGAA +GCTGAT +GCTGAC +GCTGAG +GCTGTA +GCTGTT +GCTGTC +GCTGTG +GCTGCA +GCTGCT +GCTGCC +GCTGCG +GCTGGA +GCTGGT +GCTGGC +GCTGGG +GCCAAA +GCCAAT +GCCAAC +GCCAAG +GCCATA +GCCATT +GCCATC +GCCATG +GCCACA +GCCACT +GCCACC +GCCACG +GCCAGA +GCCAGT +GCCAGC +GCCAGG +GCCTAA +GCCTAT +GCCTAC +GCCTAG +GCCTTA +GCCTTT +GCCTTC +GCCTTG +GCCTCA +GCCTCT +GCCTCC +GCCTCG +GCCTGA +GCCTGT +GCCTGC +GCCTGG +GCCCAA +GCCCAT +GCCCAC +GCCCAG +GCCCTA +GCCCTT +GCCCTC +GCCCTG +GCCCCA +GCCCCT +GCCCCC +GCCCCG +GCCCGA +GCCCGT +GCCCGC +GCCCGG +GCCGAA +GCCGAT +GCCGAC +GCCGAG +GCCGTA +GCCGTT +GCCGTC +GCCGTG +GCCGCA +GCCGCT +GCCGCC +GCCGCG +GCCGGA +GCCGGT +GCCGGC +GCCGGG +GCGAAA +GCGAAT +GCGAAC +GCGAAG +GCGATA +GCGATT +GCGATC +GCGATG +GCGACA +GCGACT +GCGACC +GCGACG +GCGAGA +GCGAGT +GCGAGC +GCGAGG +GCGTAA +GCGTAT +GCGTAC +GCGTAG +GCGTTA +GCGTTT +GCGTTC +GCGTTG +GCGTCA +GCGTCT +GCGTCC +GCGTCG +GCGTGA +GCGTGT +GCGTGC +GCGTGG +GCGCAA +GCGCAT +GCGCAC +GCGCAG +GCGCTA +GCGCTT +GCGCTC +GCGCTG +GCGCCA +GCGCCT +GCGCCC +GCGCCG +GCGCGA +GCGCGT +GCGCGC +GCGCGG +GCGGAA +GCGGAT +GCGGAC +GCGGAG +GCGGTA +GCGGTT +GCGGTC +GCGGTG +GCGGCA +GCGGCT +GCGGCC +GCGGCG +GCGGGA +GCGGGT +GCGGGC +GCGGGG +GGAAAA +GGAAAT +GGAAAC +GGAAAG +GGAATA +GGAATT +GGAATC +GGAATG +GGAACA +GGAACT +GGAACC +GGAACG +GGAAGA +GGAAGT +GGAAGC +GGAAGG +GGATAA +GGATAT +GGATAC +GGATAG +GGATTA +GGATTT +GGATTC +GGATTG +GGATCA +GGATCT +GGATCC +GGATCG +GGATGA +GGATGT +GGATGC +GGATGG +GGACAA +GGACAT +GGACAC +GGACAG +GGACTA +GGACTT +GGACTC +GGACTG +GGACCA +GGACCT +GGACCC +GGACCG +GGACGA +GGACGT +GGACGC +GGACGG +GGAGAA +GGAGAT +GGAGAC +GGAGAG +GGAGTA +GGAGTT +GGAGTC +GGAGTG +GGAGCA +GGAGCT +GGAGCC +GGAGCG +GGAGGA +GGAGGT +GGAGGC +GGAGGG +GGTAAA +GGTAAT +GGTAAC +GGTAAG +GGTATA +GGTATT +GGTATC +GGTATG +GGTACA +GGTACT +GGTACC +GGTACG +GGTAGA +GGTAGT +GGTAGC +GGTAGG +GGTTAA +GGTTAT +GGTTAC +GGTTAG +GGTTTA +GGTTTT +GGTTTC +GGTTTG +GGTTCA +GGTTCT +GGTTCC +GGTTCG +GGTTGA +GGTTGT +GGTTGC +GGTTGG +GGTCAA +GGTCAT +GGTCAC +GGTCAG +GGTCTA +GGTCTT +GGTCTC +GGTCTG +GGTCCA +GGTCCT +GGTCCC +GGTCCG +GGTCGA +GGTCGT +GGTCGC +GGTCGG +GGTGAA +GGTGAT +GGTGAC +GGTGAG +GGTGTA +GGTGTT +GGTGTC +GGTGTG +GGTGCA +GGTGCT +GGTGCC +GGTGCG +GGTGGA +GGTGGT +GGTGGC +GGTGGG +GGCAAA +GGCAAT +GGCAAC +GGCAAG +GGCATA +GGCATT +GGCATC +GGCATG +GGCACA +GGCACT +GGCACC +GGCACG +GGCAGA +GGCAGT +GGCAGC +GGCAGG +GGCTAA +GGCTAT +GGCTAC +GGCTAG +GGCTTA +GGCTTT +GGCTTC +GGCTTG +GGCTCA +GGCTCT +GGCTCC +GGCTCG +GGCTGA +GGCTGT +GGCTGC +GGCTGG +GGCCAA +GGCCAT +GGCCAC +GGCCAG +GGCCTA +GGCCTT +GGCCTC +GGCCTG +GGCCCA +GGCCCT +GGCCCC +GGCCCG +GGCCGA +GGCCGT +GGCCGC +GGCCGG +GGCGAA +GGCGAT +GGCGAC +GGCGAG +GGCGTA +GGCGTT +GGCGTC +GGCGTG +GGCGCA +GGCGCT +GGCGCC +GGCGCG +GGCGGA +GGCGGT +GGCGGC +GGCGGG +GGGAAA +GGGAAT +GGGAAC +GGGAAG +GGGATA +GGGATT +GGGATC +GGGATG +GGGACA +GGGACT +GGGACC +GGGACG +GGGAGA +GGGAGT +GGGAGC +GGGAGG +GGGTAA +GGGTAT +GGGTAC +GGGTAG +GGGTTA +GGGTTT +GGGTTC +GGGTTG +GGGTCA +GGGTCT +GGGTCC +GGGTCG +GGGTGA +GGGTGT +GGGTGC +GGGTGG +GGGCAA +GGGCAT +GGGCAC +GGGCAG +GGGCTA +GGGCTT +GGGCTC +GGGCTG +GGGCCA +GGGCCT +GGGCCC +GGGCCG +GGGCGA +GGGCGT +GGGCGC +GGGCGG +GGGGAA +GGGGAT +GGGGAC +GGGGAG +GGGGTA +GGGGTT +GGGGTC +GGGGTG +GGGGCA +GGGGCT +GGGGCC +GGGGCG +GGGGGA +GGGGGT +GGGGGC +GGGGGG diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/README.md b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..500fd976d92b217455f181bc98c77e612d9a18d9 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/README.md @@ -0,0 +1,202 @@ +--- +base_model: fabihamakhdoomi/TinyDNABERT +library_name: peft +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.12.0 \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/adapter_config.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..28c871c93775eedfaf266f00da8c02f96bc02f29 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/adapter_config.json @@ -0,0 +1,42 @@ +{ + "alpha_pattern": {}, + "auto_mapping": null, + "base_model_name_or_path": "fabihamakhdoomi/TinyDNABERT", + "bias": "none", + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 16, + "lora_dropout": 0.2, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": [ + "classifier", + "score" + ], + "peft_type": "LORA", + "r": 4, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "bert.encoder.layer.1.attention.self.query", + "bert.encoder.layer.0.output.dense", + "bert.encoder.layer.1.output.dense", + "bert.encoder.layer.0.intermediate.dense", + "bert.encoder.layer.1.intermediate.dense", + "bert.encoder.layer.0.attention.output.dense", + "bert.encoder.layer.1.attention.output.dense", + "bert.encoder.layer.0.attention.self.query", + "bert.encoder.layer.0.attention.self.value", + "bert.encoder.layer.1.attention.self.key", + "bert.encoder.layer.1.attention.self.value", + "bert.encoder.layer.0.attention.self.key" + ], + "task_type": "SEQ_CLS", + "use_dora": false, + "use_rslora": false +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/adapter_model.safetensors b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/adapter_model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..6f0a44a434556bed016185e0b9fbdc44029372cc --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/adapter_model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:837d40d9b9e8859c93f2f588fe7525b9d3a54c35070a2aa719020e553d19fd35 +size 78208 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/optimizer.pt b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..a7d5ec13dc73e05e03dee315a70dcc96cba912cd --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4c2dc29b538aaaf3b507977f2e63aa473fa3e9e143977a73112b355662e922ad +size 165226 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/rng_state.pth b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..bd3d1cad75d68826b1d075775d8e5141764e5725 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:855054099544841bdc3a4d2a62131897c2be69ce849e4324026c75100489f631 +size 14244 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/scheduler.pt b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..14aa799e6f5c93344d1b233b91d0bda9d8eb59e7 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:eca4da5723e14faf16c3be96d23919390da5288d0d9db40021d332ee1df25b93 +size 1064 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/special_tokens_map.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..a8b3208c2884c4efb86e49300fdd3dc877220cdf --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/special_tokens_map.json @@ -0,0 +1,7 @@ +{ + "cls_token": "[CLS]", + "mask_token": "[MASK]", + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/tokenizer.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..ebd782de6963b9032ce9374cfe30f5275be1a30b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/tokenizer.json @@ -0,0 +1,4265 @@ +{ + "version": "1.0", + "truncation": { + "direction": "Right", + "max_length": 512, + "strategy": "LongestFirst", + "stride": 0 + }, + "padding": { + "strategy": { + "Fixed": 512 + }, + "direction": "Right", + "pad_to_multiple_of": null, + "pad_id": 0, + "pad_type_id": 0, + "pad_token": "[PAD]" + }, + "added_tokens": [ + { + "id": 0, + "content": "[PAD]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 1, + "content": "[UNK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 2, + "content": "[CLS]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 3, + "content": "[SEP]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 4, + "content": "[MASK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + } + ], + "normalizer": { + "type": "BertNormalizer", + "clean_text": true, + "handle_chinese_chars": true, + "strip_accents": null, + "lowercase": false + }, + "pre_tokenizer": { + "type": "BertPreTokenizer" + }, + "post_processor": { + "type": "TemplateProcessing", + "single": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + } + ], + "pair": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "B", + "type_id": 1 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 1 + } + } + ], + "special_tokens": { + "[CLS]": { + "id": "[CLS]", + "ids": [ + 2 + ], + "tokens": [ + "[CLS]" + ] + }, + "[SEP]": { + "id": "[SEP]", + "ids": [ + 3 + ], + "tokens": [ + "[SEP]" + ] + } + } + }, + "decoder": { + "type": "WordPiece", + "prefix": "##", + "cleanup": true + }, + "model": { + "type": "WordPiece", + "unk_token": "[UNK]", + "continuing_subword_prefix": "##", + "max_input_chars_per_word": 100, + "vocab": { + "[PAD]": 0, + "[UNK]": 1, + "[CLS]": 2, + "[SEP]": 3, + "[MASK]": 4, + "AAAAAA": 5, + "AAAAAT": 6, + "AAAAAC": 7, + "AAAAAG": 8, + "AAAATA": 9, + "AAAATT": 10, + "AAAATC": 11, + "AAAATG": 12, + "AAAACA": 13, + "AAAACT": 14, + "AAAACC": 15, + "AAAACG": 16, + "AAAAGA": 17, + "AAAAGT": 18, + "AAAAGC": 19, + "AAAAGG": 20, + "AAATAA": 21, + "AAATAT": 22, + "AAATAC": 23, + "AAATAG": 24, + "AAATTA": 25, + "AAATTT": 26, + "AAATTC": 27, + "AAATTG": 28, + "AAATCA": 29, + "AAATCT": 30, + "AAATCC": 31, + "AAATCG": 32, + "AAATGA": 33, + "AAATGT": 34, + "AAATGC": 35, + "AAATGG": 36, + "AAACAA": 37, + "AAACAT": 38, + "AAACAC": 39, + "AAACAG": 40, + "AAACTA": 41, + "AAACTT": 42, + "AAACTC": 43, + "AAACTG": 44, + "AAACCA": 45, + "AAACCT": 46, + "AAACCC": 47, + "AAACCG": 48, + "AAACGA": 49, + "AAACGT": 50, + "AAACGC": 51, + "AAACGG": 52, + "AAAGAA": 53, + "AAAGAT": 54, + "AAAGAC": 55, + "AAAGAG": 56, + "AAAGTA": 57, + "AAAGTT": 58, + "AAAGTC": 59, + "AAAGTG": 60, + "AAAGCA": 61, + "AAAGCT": 62, + "AAAGCC": 63, + "AAAGCG": 64, + "AAAGGA": 65, + "AAAGGT": 66, + "AAAGGC": 67, + "AAAGGG": 68, + "AATAAA": 69, + "AATAAT": 70, + "AATAAC": 71, + "AATAAG": 72, + "AATATA": 73, + "AATATT": 74, + "AATATC": 75, + "AATATG": 76, + "AATACA": 77, + "AATACT": 78, + "AATACC": 79, + "AATACG": 80, + "AATAGA": 81, + "AATAGT": 82, + "AATAGC": 83, + "AATAGG": 84, + "AATTAA": 85, + "AATTAT": 86, + "AATTAC": 87, + "AATTAG": 88, + "AATTTA": 89, + "AATTTT": 90, + "AATTTC": 91, + "AATTTG": 92, + "AATTCA": 93, + "AATTCT": 94, + "AATTCC": 95, + "AATTCG": 96, + "AATTGA": 97, + "AATTGT": 98, + "AATTGC": 99, + "AATTGG": 100, + "AATCAA": 101, + "AATCAT": 102, + "AATCAC": 103, + "AATCAG": 104, + "AATCTA": 105, + "AATCTT": 106, + "AATCTC": 107, + "AATCTG": 108, + "AATCCA": 109, + "AATCCT": 110, + "AATCCC": 111, + "AATCCG": 112, + "AATCGA": 113, + "AATCGT": 114, + "AATCGC": 115, + "AATCGG": 116, + "AATGAA": 117, + "AATGAT": 118, + "AATGAC": 119, + "AATGAG": 120, + "AATGTA": 121, + "AATGTT": 122, + "AATGTC": 123, + "AATGTG": 124, + "AATGCA": 125, + "AATGCT": 126, + "AATGCC": 127, + "AATGCG": 128, + "AATGGA": 129, + "AATGGT": 130, + "AATGGC": 131, + "AATGGG": 132, + "AACAAA": 133, + "AACAAT": 134, + "AACAAC": 135, + "AACAAG": 136, + "AACATA": 137, + "AACATT": 138, + "AACATC": 139, + "AACATG": 140, + "AACACA": 141, + "AACACT": 142, + "AACACC": 143, + "AACACG": 144, + "AACAGA": 145, + "AACAGT": 146, + "AACAGC": 147, + "AACAGG": 148, + "AACTAA": 149, + "AACTAT": 150, + "AACTAC": 151, + "AACTAG": 152, + "AACTTA": 153, + "AACTTT": 154, + "AACTTC": 155, + "AACTTG": 156, + "AACTCA": 157, + "AACTCT": 158, + "AACTCC": 159, + "AACTCG": 160, + "AACTGA": 161, + "AACTGT": 162, + "AACTGC": 163, + "AACTGG": 164, + "AACCAA": 165, + "AACCAT": 166, + "AACCAC": 167, + "AACCAG": 168, + "AACCTA": 169, + "AACCTT": 170, + "AACCTC": 171, + "AACCTG": 172, + "AACCCA": 173, + "AACCCT": 174, + "AACCCC": 175, + "AACCCG": 176, + "AACCGA": 177, + "AACCGT": 178, + "AACCGC": 179, + "AACCGG": 180, + "AACGAA": 181, + "AACGAT": 182, + "AACGAC": 183, + "AACGAG": 184, + "AACGTA": 185, + "AACGTT": 186, + "AACGTC": 187, + "AACGTG": 188, + "AACGCA": 189, + "AACGCT": 190, + "AACGCC": 191, + "AACGCG": 192, + "AACGGA": 193, + "AACGGT": 194, + "AACGGC": 195, + "AACGGG": 196, + "AAGAAA": 197, + "AAGAAT": 198, + "AAGAAC": 199, + "AAGAAG": 200, + "AAGATA": 201, + "AAGATT": 202, + "AAGATC": 203, + "AAGATG": 204, + "AAGACA": 205, + "AAGACT": 206, + "AAGACC": 207, + "AAGACG": 208, + "AAGAGA": 209, + "AAGAGT": 210, + "AAGAGC": 211, + "AAGAGG": 212, + "AAGTAA": 213, + "AAGTAT": 214, + "AAGTAC": 215, + "AAGTAG": 216, + "AAGTTA": 217, + "AAGTTT": 218, + "AAGTTC": 219, + "AAGTTG": 220, + "AAGTCA": 221, + "AAGTCT": 222, + "AAGTCC": 223, + "AAGTCG": 224, + "AAGTGA": 225, + "AAGTGT": 226, + "AAGTGC": 227, + "AAGTGG": 228, + "AAGCAA": 229, + "AAGCAT": 230, + "AAGCAC": 231, + "AAGCAG": 232, + "AAGCTA": 233, + "AAGCTT": 234, + "AAGCTC": 235, + "AAGCTG": 236, + "AAGCCA": 237, + "AAGCCT": 238, + "AAGCCC": 239, + "AAGCCG": 240, + "AAGCGA": 241, + "AAGCGT": 242, + "AAGCGC": 243, + "AAGCGG": 244, + "AAGGAA": 245, + "AAGGAT": 246, + "AAGGAC": 247, + "AAGGAG": 248, + "AAGGTA": 249, + "AAGGTT": 250, + "AAGGTC": 251, + "AAGGTG": 252, + "AAGGCA": 253, + "AAGGCT": 254, + "AAGGCC": 255, + "AAGGCG": 256, + "AAGGGA": 257, + "AAGGGT": 258, + "AAGGGC": 259, + "AAGGGG": 260, + "ATAAAA": 261, + "ATAAAT": 262, + "ATAAAC": 263, + "ATAAAG": 264, + "ATAATA": 265, + "ATAATT": 266, + "ATAATC": 267, + "ATAATG": 268, + "ATAACA": 269, + "ATAACT": 270, + "ATAACC": 271, + "ATAACG": 272, + "ATAAGA": 273, + "ATAAGT": 274, + "ATAAGC": 275, + "ATAAGG": 276, + "ATATAA": 277, + "ATATAT": 278, + "ATATAC": 279, + "ATATAG": 280, + "ATATTA": 281, + "ATATTT": 282, + "ATATTC": 283, + "ATATTG": 284, + "ATATCA": 285, + "ATATCT": 286, + "ATATCC": 287, + "ATATCG": 288, + "ATATGA": 289, + "ATATGT": 290, + "ATATGC": 291, + "ATATGG": 292, + "ATACAA": 293, + "ATACAT": 294, + "ATACAC": 295, + "ATACAG": 296, + "ATACTA": 297, + "ATACTT": 298, + "ATACTC": 299, + "ATACTG": 300, + "ATACCA": 301, + "ATACCT": 302, + "ATACCC": 303, + "ATACCG": 304, + "ATACGA": 305, + "ATACGT": 306, + "ATACGC": 307, + "ATACGG": 308, + "ATAGAA": 309, + "ATAGAT": 310, + "ATAGAC": 311, + "ATAGAG": 312, + "ATAGTA": 313, + "ATAGTT": 314, + "ATAGTC": 315, + "ATAGTG": 316, + "ATAGCA": 317, + "ATAGCT": 318, + "ATAGCC": 319, + "ATAGCG": 320, + "ATAGGA": 321, + "ATAGGT": 322, + "ATAGGC": 323, + "ATAGGG": 324, + "ATTAAA": 325, + "ATTAAT": 326, + "ATTAAC": 327, + "ATTAAG": 328, + "ATTATA": 329, + "ATTATT": 330, + "ATTATC": 331, + "ATTATG": 332, + "ATTACA": 333, + "ATTACT": 334, + "ATTACC": 335, + "ATTACG": 336, + "ATTAGA": 337, + "ATTAGT": 338, + "ATTAGC": 339, + "ATTAGG": 340, + "ATTTAA": 341, + "ATTTAT": 342, + "ATTTAC": 343, + "ATTTAG": 344, + "ATTTTA": 345, + "ATTTTT": 346, + "ATTTTC": 347, + "ATTTTG": 348, + "ATTTCA": 349, + "ATTTCT": 350, + "ATTTCC": 351, + "ATTTCG": 352, + "ATTTGA": 353, + "ATTTGT": 354, + "ATTTGC": 355, + "ATTTGG": 356, + "ATTCAA": 357, + "ATTCAT": 358, + "ATTCAC": 359, + "ATTCAG": 360, + "ATTCTA": 361, + "ATTCTT": 362, + "ATTCTC": 363, + "ATTCTG": 364, + "ATTCCA": 365, + "ATTCCT": 366, + "ATTCCC": 367, + "ATTCCG": 368, + "ATTCGA": 369, + "ATTCGT": 370, + "ATTCGC": 371, + "ATTCGG": 372, + "ATTGAA": 373, + "ATTGAT": 374, + "ATTGAC": 375, + "ATTGAG": 376, + "ATTGTA": 377, + "ATTGTT": 378, + "ATTGTC": 379, + "ATTGTG": 380, + "ATTGCA": 381, + "ATTGCT": 382, + "ATTGCC": 383, + "ATTGCG": 384, + "ATTGGA": 385, + "ATTGGT": 386, + "ATTGGC": 387, + "ATTGGG": 388, + "ATCAAA": 389, + "ATCAAT": 390, + "ATCAAC": 391, + "ATCAAG": 392, + "ATCATA": 393, + "ATCATT": 394, + "ATCATC": 395, + "ATCATG": 396, + "ATCACA": 397, + "ATCACT": 398, + "ATCACC": 399, + "ATCACG": 400, + "ATCAGA": 401, + "ATCAGT": 402, + "ATCAGC": 403, + "ATCAGG": 404, + "ATCTAA": 405, + "ATCTAT": 406, + "ATCTAC": 407, + "ATCTAG": 408, + "ATCTTA": 409, + "ATCTTT": 410, + "ATCTTC": 411, + "ATCTTG": 412, + "ATCTCA": 413, + "ATCTCT": 414, + "ATCTCC": 415, + "ATCTCG": 416, + "ATCTGA": 417, + "ATCTGT": 418, + "ATCTGC": 419, + "ATCTGG": 420, + "ATCCAA": 421, + "ATCCAT": 422, + "ATCCAC": 423, + "ATCCAG": 424, + "ATCCTA": 425, + "ATCCTT": 426, + "ATCCTC": 427, + "ATCCTG": 428, + "ATCCCA": 429, + "ATCCCT": 430, + "ATCCCC": 431, + "ATCCCG": 432, + "ATCCGA": 433, + "ATCCGT": 434, + "ATCCGC": 435, + "ATCCGG": 436, + "ATCGAA": 437, + "ATCGAT": 438, + "ATCGAC": 439, + "ATCGAG": 440, + "ATCGTA": 441, + "ATCGTT": 442, + "ATCGTC": 443, + "ATCGTG": 444, + "ATCGCA": 445, + "ATCGCT": 446, + "ATCGCC": 447, + "ATCGCG": 448, + "ATCGGA": 449, + "ATCGGT": 450, + "ATCGGC": 451, + "ATCGGG": 452, + "ATGAAA": 453, + "ATGAAT": 454, + "ATGAAC": 455, + "ATGAAG": 456, + "ATGATA": 457, + "ATGATT": 458, + "ATGATC": 459, + "ATGATG": 460, + "ATGACA": 461, + "ATGACT": 462, + "ATGACC": 463, + "ATGACG": 464, + "ATGAGA": 465, + "ATGAGT": 466, + "ATGAGC": 467, + "ATGAGG": 468, + "ATGTAA": 469, + "ATGTAT": 470, + "ATGTAC": 471, + "ATGTAG": 472, + "ATGTTA": 473, + "ATGTTT": 474, + "ATGTTC": 475, + "ATGTTG": 476, + "ATGTCA": 477, + "ATGTCT": 478, + "ATGTCC": 479, + "ATGTCG": 480, + "ATGTGA": 481, + "ATGTGT": 482, + "ATGTGC": 483, + "ATGTGG": 484, + "ATGCAA": 485, + "ATGCAT": 486, + "ATGCAC": 487, + "ATGCAG": 488, + "ATGCTA": 489, + "ATGCTT": 490, + "ATGCTC": 491, + "ATGCTG": 492, + "ATGCCA": 493, + "ATGCCT": 494, + "ATGCCC": 495, + "ATGCCG": 496, + "ATGCGA": 497, + "ATGCGT": 498, + "ATGCGC": 499, + "ATGCGG": 500, + "ATGGAA": 501, + "ATGGAT": 502, + "ATGGAC": 503, + "ATGGAG": 504, + "ATGGTA": 505, + "ATGGTT": 506, + "ATGGTC": 507, + "ATGGTG": 508, + "ATGGCA": 509, + "ATGGCT": 510, + "ATGGCC": 511, + "ATGGCG": 512, + "ATGGGA": 513, + "ATGGGT": 514, + "ATGGGC": 515, + "ATGGGG": 516, + "ACAAAA": 517, + "ACAAAT": 518, + "ACAAAC": 519, + "ACAAAG": 520, + "ACAATA": 521, + "ACAATT": 522, + "ACAATC": 523, + "ACAATG": 524, + "ACAACA": 525, + "ACAACT": 526, + "ACAACC": 527, + "ACAACG": 528, + "ACAAGA": 529, + "ACAAGT": 530, + "ACAAGC": 531, + "ACAAGG": 532, + "ACATAA": 533, + "ACATAT": 534, + "ACATAC": 535, + "ACATAG": 536, + "ACATTA": 537, + "ACATTT": 538, + "ACATTC": 539, + "ACATTG": 540, + "ACATCA": 541, + "ACATCT": 542, + "ACATCC": 543, + "ACATCG": 544, + "ACATGA": 545, + "ACATGT": 546, + "ACATGC": 547, + "ACATGG": 548, + "ACACAA": 549, + "ACACAT": 550, + "ACACAC": 551, + "ACACAG": 552, + "ACACTA": 553, + "ACACTT": 554, + "ACACTC": 555, + "ACACTG": 556, + "ACACCA": 557, + "ACACCT": 558, + "ACACCC": 559, + "ACACCG": 560, + "ACACGA": 561, + "ACACGT": 562, + "ACACGC": 563, + "ACACGG": 564, + "ACAGAA": 565, + "ACAGAT": 566, + "ACAGAC": 567, + "ACAGAG": 568, + "ACAGTA": 569, + "ACAGTT": 570, + "ACAGTC": 571, + "ACAGTG": 572, + "ACAGCA": 573, + "ACAGCT": 574, + "ACAGCC": 575, + "ACAGCG": 576, + "ACAGGA": 577, + "ACAGGT": 578, + "ACAGGC": 579, + "ACAGGG": 580, + "ACTAAA": 581, + "ACTAAT": 582, + "ACTAAC": 583, + "ACTAAG": 584, + "ACTATA": 585, + "ACTATT": 586, + "ACTATC": 587, + "ACTATG": 588, + "ACTACA": 589, + "ACTACT": 590, + "ACTACC": 591, + "ACTACG": 592, + "ACTAGA": 593, + "ACTAGT": 594, + "ACTAGC": 595, + "ACTAGG": 596, + "ACTTAA": 597, + "ACTTAT": 598, + "ACTTAC": 599, + "ACTTAG": 600, + "ACTTTA": 601, + "ACTTTT": 602, + "ACTTTC": 603, + "ACTTTG": 604, + "ACTTCA": 605, + "ACTTCT": 606, + "ACTTCC": 607, + "ACTTCG": 608, + "ACTTGA": 609, + "ACTTGT": 610, + "ACTTGC": 611, + "ACTTGG": 612, + "ACTCAA": 613, + "ACTCAT": 614, + "ACTCAC": 615, + "ACTCAG": 616, + "ACTCTA": 617, + "ACTCTT": 618, + "ACTCTC": 619, + "ACTCTG": 620, + "ACTCCA": 621, + "ACTCCT": 622, + "ACTCCC": 623, + "ACTCCG": 624, + "ACTCGA": 625, + "ACTCGT": 626, + "ACTCGC": 627, + "ACTCGG": 628, + "ACTGAA": 629, + "ACTGAT": 630, + "ACTGAC": 631, + "ACTGAG": 632, + "ACTGTA": 633, + "ACTGTT": 634, + "ACTGTC": 635, + "ACTGTG": 636, + "ACTGCA": 637, + "ACTGCT": 638, + "ACTGCC": 639, + "ACTGCG": 640, + "ACTGGA": 641, + "ACTGGT": 642, + "ACTGGC": 643, + "ACTGGG": 644, + "ACCAAA": 645, + "ACCAAT": 646, + "ACCAAC": 647, + "ACCAAG": 648, + "ACCATA": 649, + "ACCATT": 650, + "ACCATC": 651, + "ACCATG": 652, + "ACCACA": 653, + "ACCACT": 654, + "ACCACC": 655, + "ACCACG": 656, + "ACCAGA": 657, + "ACCAGT": 658, + "ACCAGC": 659, + "ACCAGG": 660, + "ACCTAA": 661, + "ACCTAT": 662, + "ACCTAC": 663, + "ACCTAG": 664, + "ACCTTA": 665, + "ACCTTT": 666, + "ACCTTC": 667, + "ACCTTG": 668, + "ACCTCA": 669, + "ACCTCT": 670, + "ACCTCC": 671, + "ACCTCG": 672, + "ACCTGA": 673, + "ACCTGT": 674, + "ACCTGC": 675, + "ACCTGG": 676, + "ACCCAA": 677, + "ACCCAT": 678, + "ACCCAC": 679, + "ACCCAG": 680, + "ACCCTA": 681, + "ACCCTT": 682, + "ACCCTC": 683, + "ACCCTG": 684, + "ACCCCA": 685, + "ACCCCT": 686, + "ACCCCC": 687, + "ACCCCG": 688, + "ACCCGA": 689, + "ACCCGT": 690, + "ACCCGC": 691, + "ACCCGG": 692, + "ACCGAA": 693, + "ACCGAT": 694, + "ACCGAC": 695, + "ACCGAG": 696, + "ACCGTA": 697, + "ACCGTT": 698, + "ACCGTC": 699, + "ACCGTG": 700, + "ACCGCA": 701, + "ACCGCT": 702, + "ACCGCC": 703, + "ACCGCG": 704, + "ACCGGA": 705, + "ACCGGT": 706, + "ACCGGC": 707, + "ACCGGG": 708, + "ACGAAA": 709, + "ACGAAT": 710, + "ACGAAC": 711, + "ACGAAG": 712, + "ACGATA": 713, + "ACGATT": 714, + "ACGATC": 715, + "ACGATG": 716, + "ACGACA": 717, + "ACGACT": 718, + "ACGACC": 719, + "ACGACG": 720, + "ACGAGA": 721, + "ACGAGT": 722, + "ACGAGC": 723, + "ACGAGG": 724, + "ACGTAA": 725, + "ACGTAT": 726, + "ACGTAC": 727, + "ACGTAG": 728, + "ACGTTA": 729, + "ACGTTT": 730, + "ACGTTC": 731, + "ACGTTG": 732, + "ACGTCA": 733, + "ACGTCT": 734, + "ACGTCC": 735, + "ACGTCG": 736, + "ACGTGA": 737, + "ACGTGT": 738, + "ACGTGC": 739, + "ACGTGG": 740, + "ACGCAA": 741, + "ACGCAT": 742, + "ACGCAC": 743, + "ACGCAG": 744, + "ACGCTA": 745, + "ACGCTT": 746, + "ACGCTC": 747, + "ACGCTG": 748, + "ACGCCA": 749, + "ACGCCT": 750, + "ACGCCC": 751, + "ACGCCG": 752, + "ACGCGA": 753, + "ACGCGT": 754, + "ACGCGC": 755, + "ACGCGG": 756, + "ACGGAA": 757, + "ACGGAT": 758, + "ACGGAC": 759, + "ACGGAG": 760, + "ACGGTA": 761, + "ACGGTT": 762, + "ACGGTC": 763, + "ACGGTG": 764, + "ACGGCA": 765, + "ACGGCT": 766, + "ACGGCC": 767, + "ACGGCG": 768, + "ACGGGA": 769, + "ACGGGT": 770, + "ACGGGC": 771, + "ACGGGG": 772, + "AGAAAA": 773, + "AGAAAT": 774, + "AGAAAC": 775, + "AGAAAG": 776, + "AGAATA": 777, + "AGAATT": 778, + "AGAATC": 779, + "AGAATG": 780, + "AGAACA": 781, + "AGAACT": 782, + "AGAACC": 783, + "AGAACG": 784, + "AGAAGA": 785, + "AGAAGT": 786, + "AGAAGC": 787, + "AGAAGG": 788, + "AGATAA": 789, + "AGATAT": 790, + "AGATAC": 791, + "AGATAG": 792, + "AGATTA": 793, + "AGATTT": 794, + "AGATTC": 795, + "AGATTG": 796, + "AGATCA": 797, + "AGATCT": 798, + "AGATCC": 799, + "AGATCG": 800, + "AGATGA": 801, + "AGATGT": 802, + "AGATGC": 803, + "AGATGG": 804, + "AGACAA": 805, + "AGACAT": 806, + "AGACAC": 807, + "AGACAG": 808, + "AGACTA": 809, + "AGACTT": 810, + "AGACTC": 811, + "AGACTG": 812, + "AGACCA": 813, + "AGACCT": 814, + "AGACCC": 815, + "AGACCG": 816, + "AGACGA": 817, + "AGACGT": 818, + "AGACGC": 819, + "AGACGG": 820, + "AGAGAA": 821, + "AGAGAT": 822, + "AGAGAC": 823, + "AGAGAG": 824, + "AGAGTA": 825, + "AGAGTT": 826, + "AGAGTC": 827, + "AGAGTG": 828, + "AGAGCA": 829, + "AGAGCT": 830, + "AGAGCC": 831, + "AGAGCG": 832, + "AGAGGA": 833, + "AGAGGT": 834, + "AGAGGC": 835, + "AGAGGG": 836, + "AGTAAA": 837, + "AGTAAT": 838, + "AGTAAC": 839, + "AGTAAG": 840, + "AGTATA": 841, + "AGTATT": 842, + "AGTATC": 843, + "AGTATG": 844, + "AGTACA": 845, + "AGTACT": 846, + "AGTACC": 847, + "AGTACG": 848, + "AGTAGA": 849, + "AGTAGT": 850, + "AGTAGC": 851, + "AGTAGG": 852, + "AGTTAA": 853, + "AGTTAT": 854, + "AGTTAC": 855, + "AGTTAG": 856, + "AGTTTA": 857, + "AGTTTT": 858, + "AGTTTC": 859, + "AGTTTG": 860, + "AGTTCA": 861, + "AGTTCT": 862, + "AGTTCC": 863, + "AGTTCG": 864, + "AGTTGA": 865, + "AGTTGT": 866, + "AGTTGC": 867, + "AGTTGG": 868, + "AGTCAA": 869, + "AGTCAT": 870, + "AGTCAC": 871, + "AGTCAG": 872, + "AGTCTA": 873, + "AGTCTT": 874, + "AGTCTC": 875, + "AGTCTG": 876, + "AGTCCA": 877, + "AGTCCT": 878, + "AGTCCC": 879, + "AGTCCG": 880, + "AGTCGA": 881, + "AGTCGT": 882, + "AGTCGC": 883, + "AGTCGG": 884, + "AGTGAA": 885, + "AGTGAT": 886, + "AGTGAC": 887, + "AGTGAG": 888, + "AGTGTA": 889, + "AGTGTT": 890, + "AGTGTC": 891, + "AGTGTG": 892, + "AGTGCA": 893, + "AGTGCT": 894, + "AGTGCC": 895, + "AGTGCG": 896, + "AGTGGA": 897, + "AGTGGT": 898, + "AGTGGC": 899, + "AGTGGG": 900, + "AGCAAA": 901, + "AGCAAT": 902, + "AGCAAC": 903, + "AGCAAG": 904, + "AGCATA": 905, + "AGCATT": 906, + "AGCATC": 907, + "AGCATG": 908, + "AGCACA": 909, + "AGCACT": 910, + "AGCACC": 911, + "AGCACG": 912, + "AGCAGA": 913, + "AGCAGT": 914, + "AGCAGC": 915, + "AGCAGG": 916, + "AGCTAA": 917, + "AGCTAT": 918, + "AGCTAC": 919, + "AGCTAG": 920, + "AGCTTA": 921, + "AGCTTT": 922, + "AGCTTC": 923, + "AGCTTG": 924, + "AGCTCA": 925, + "AGCTCT": 926, + "AGCTCC": 927, + "AGCTCG": 928, + "AGCTGA": 929, + "AGCTGT": 930, + "AGCTGC": 931, + "AGCTGG": 932, + "AGCCAA": 933, + "AGCCAT": 934, + "AGCCAC": 935, + "AGCCAG": 936, + "AGCCTA": 937, + "AGCCTT": 938, + "AGCCTC": 939, + "AGCCTG": 940, + "AGCCCA": 941, + "AGCCCT": 942, + "AGCCCC": 943, + "AGCCCG": 944, + "AGCCGA": 945, + "AGCCGT": 946, + "AGCCGC": 947, + "AGCCGG": 948, + "AGCGAA": 949, + "AGCGAT": 950, + "AGCGAC": 951, + "AGCGAG": 952, + "AGCGTA": 953, + "AGCGTT": 954, + "AGCGTC": 955, + "AGCGTG": 956, + "AGCGCA": 957, + "AGCGCT": 958, + "AGCGCC": 959, + "AGCGCG": 960, + "AGCGGA": 961, + "AGCGGT": 962, + "AGCGGC": 963, + "AGCGGG": 964, + "AGGAAA": 965, + "AGGAAT": 966, + "AGGAAC": 967, + "AGGAAG": 968, + "AGGATA": 969, + "AGGATT": 970, + "AGGATC": 971, + "AGGATG": 972, + "AGGACA": 973, + "AGGACT": 974, + "AGGACC": 975, + "AGGACG": 976, + "AGGAGA": 977, + "AGGAGT": 978, + "AGGAGC": 979, + "AGGAGG": 980, + "AGGTAA": 981, + "AGGTAT": 982, + "AGGTAC": 983, + "AGGTAG": 984, + "AGGTTA": 985, + "AGGTTT": 986, + "AGGTTC": 987, + "AGGTTG": 988, + "AGGTCA": 989, + "AGGTCT": 990, + "AGGTCC": 991, + "AGGTCG": 992, + "AGGTGA": 993, + "AGGTGT": 994, + "AGGTGC": 995, + "AGGTGG": 996, + "AGGCAA": 997, + "AGGCAT": 998, + "AGGCAC": 999, + "AGGCAG": 1000, + "AGGCTA": 1001, + "AGGCTT": 1002, + "AGGCTC": 1003, + "AGGCTG": 1004, + "AGGCCA": 1005, + "AGGCCT": 1006, + "AGGCCC": 1007, + "AGGCCG": 1008, + "AGGCGA": 1009, + "AGGCGT": 1010, + "AGGCGC": 1011, + "AGGCGG": 1012, + "AGGGAA": 1013, + "AGGGAT": 1014, + "AGGGAC": 1015, + "AGGGAG": 1016, + "AGGGTA": 1017, + "AGGGTT": 1018, + "AGGGTC": 1019, + "AGGGTG": 1020, + "AGGGCA": 1021, + "AGGGCT": 1022, + "AGGGCC": 1023, + "AGGGCG": 1024, + "AGGGGA": 1025, + "AGGGGT": 1026, + "AGGGGC": 1027, + "AGGGGG": 1028, + "TAAAAA": 1029, + "TAAAAT": 1030, + "TAAAAC": 1031, + "TAAAAG": 1032, + "TAAATA": 1033, + "TAAATT": 1034, + "TAAATC": 1035, + "TAAATG": 1036, + "TAAACA": 1037, + "TAAACT": 1038, + "TAAACC": 1039, + "TAAACG": 1040, + "TAAAGA": 1041, + "TAAAGT": 1042, + "TAAAGC": 1043, + "TAAAGG": 1044, + "TAATAA": 1045, + "TAATAT": 1046, + "TAATAC": 1047, + "TAATAG": 1048, + "TAATTA": 1049, + "TAATTT": 1050, + "TAATTC": 1051, + "TAATTG": 1052, + "TAATCA": 1053, + "TAATCT": 1054, + "TAATCC": 1055, + "TAATCG": 1056, + "TAATGA": 1057, + "TAATGT": 1058, + "TAATGC": 1059, + "TAATGG": 1060, + "TAACAA": 1061, + "TAACAT": 1062, + "TAACAC": 1063, + "TAACAG": 1064, + "TAACTA": 1065, + "TAACTT": 1066, + "TAACTC": 1067, + "TAACTG": 1068, + "TAACCA": 1069, + "TAACCT": 1070, + "TAACCC": 1071, + "TAACCG": 1072, + "TAACGA": 1073, + "TAACGT": 1074, + "TAACGC": 1075, + "TAACGG": 1076, + "TAAGAA": 1077, + "TAAGAT": 1078, + "TAAGAC": 1079, + "TAAGAG": 1080, + "TAAGTA": 1081, + "TAAGTT": 1082, + "TAAGTC": 1083, + "TAAGTG": 1084, + "TAAGCA": 1085, + "TAAGCT": 1086, + "TAAGCC": 1087, + "TAAGCG": 1088, + "TAAGGA": 1089, + "TAAGGT": 1090, + "TAAGGC": 1091, + "TAAGGG": 1092, + "TATAAA": 1093, + "TATAAT": 1094, + "TATAAC": 1095, + "TATAAG": 1096, + "TATATA": 1097, + "TATATT": 1098, + "TATATC": 1099, + "TATATG": 1100, + "TATACA": 1101, + "TATACT": 1102, + "TATACC": 1103, + "TATACG": 1104, + "TATAGA": 1105, + "TATAGT": 1106, + "TATAGC": 1107, + "TATAGG": 1108, + "TATTAA": 1109, + "TATTAT": 1110, + "TATTAC": 1111, + "TATTAG": 1112, + "TATTTA": 1113, + "TATTTT": 1114, + "TATTTC": 1115, + "TATTTG": 1116, + "TATTCA": 1117, + "TATTCT": 1118, + "TATTCC": 1119, + "TATTCG": 1120, + "TATTGA": 1121, + "TATTGT": 1122, + "TATTGC": 1123, + "TATTGG": 1124, + "TATCAA": 1125, + "TATCAT": 1126, + "TATCAC": 1127, + "TATCAG": 1128, + "TATCTA": 1129, + "TATCTT": 1130, + "TATCTC": 1131, + "TATCTG": 1132, + "TATCCA": 1133, + "TATCCT": 1134, + "TATCCC": 1135, + "TATCCG": 1136, + "TATCGA": 1137, + "TATCGT": 1138, + "TATCGC": 1139, + "TATCGG": 1140, + "TATGAA": 1141, + "TATGAT": 1142, + "TATGAC": 1143, + "TATGAG": 1144, + "TATGTA": 1145, + "TATGTT": 1146, + "TATGTC": 1147, + "TATGTG": 1148, + "TATGCA": 1149, + "TATGCT": 1150, + "TATGCC": 1151, + "TATGCG": 1152, + "TATGGA": 1153, + "TATGGT": 1154, + "TATGGC": 1155, + "TATGGG": 1156, + "TACAAA": 1157, + "TACAAT": 1158, + "TACAAC": 1159, + "TACAAG": 1160, + "TACATA": 1161, + "TACATT": 1162, + "TACATC": 1163, + "TACATG": 1164, + "TACACA": 1165, + "TACACT": 1166, + "TACACC": 1167, + "TACACG": 1168, + "TACAGA": 1169, + "TACAGT": 1170, + "TACAGC": 1171, + "TACAGG": 1172, + "TACTAA": 1173, + "TACTAT": 1174, + "TACTAC": 1175, + "TACTAG": 1176, + "TACTTA": 1177, + "TACTTT": 1178, + "TACTTC": 1179, + "TACTTG": 1180, + "TACTCA": 1181, + "TACTCT": 1182, + "TACTCC": 1183, + "TACTCG": 1184, + "TACTGA": 1185, + "TACTGT": 1186, + "TACTGC": 1187, + "TACTGG": 1188, + "TACCAA": 1189, + "TACCAT": 1190, + "TACCAC": 1191, + "TACCAG": 1192, + "TACCTA": 1193, + "TACCTT": 1194, + "TACCTC": 1195, + "TACCTG": 1196, + "TACCCA": 1197, + "TACCCT": 1198, + "TACCCC": 1199, + "TACCCG": 1200, + "TACCGA": 1201, + "TACCGT": 1202, + "TACCGC": 1203, + "TACCGG": 1204, + "TACGAA": 1205, + "TACGAT": 1206, + "TACGAC": 1207, + "TACGAG": 1208, + "TACGTA": 1209, + "TACGTT": 1210, + "TACGTC": 1211, + "TACGTG": 1212, + "TACGCA": 1213, + "TACGCT": 1214, + "TACGCC": 1215, + "TACGCG": 1216, + "TACGGA": 1217, + "TACGGT": 1218, + "TACGGC": 1219, + "TACGGG": 1220, + "TAGAAA": 1221, + "TAGAAT": 1222, + "TAGAAC": 1223, + "TAGAAG": 1224, + "TAGATA": 1225, + "TAGATT": 1226, + "TAGATC": 1227, + "TAGATG": 1228, + "TAGACA": 1229, + "TAGACT": 1230, + "TAGACC": 1231, + "TAGACG": 1232, + "TAGAGA": 1233, + "TAGAGT": 1234, + "TAGAGC": 1235, + "TAGAGG": 1236, + "TAGTAA": 1237, + "TAGTAT": 1238, + "TAGTAC": 1239, + "TAGTAG": 1240, + "TAGTTA": 1241, + "TAGTTT": 1242, + "TAGTTC": 1243, + "TAGTTG": 1244, + "TAGTCA": 1245, + "TAGTCT": 1246, + "TAGTCC": 1247, + "TAGTCG": 1248, + "TAGTGA": 1249, + "TAGTGT": 1250, + "TAGTGC": 1251, + "TAGTGG": 1252, + "TAGCAA": 1253, + "TAGCAT": 1254, + "TAGCAC": 1255, + "TAGCAG": 1256, + "TAGCTA": 1257, + "TAGCTT": 1258, + "TAGCTC": 1259, + "TAGCTG": 1260, + "TAGCCA": 1261, + "TAGCCT": 1262, + "TAGCCC": 1263, + "TAGCCG": 1264, + "TAGCGA": 1265, + "TAGCGT": 1266, + "TAGCGC": 1267, + "TAGCGG": 1268, + "TAGGAA": 1269, + "TAGGAT": 1270, + "TAGGAC": 1271, + "TAGGAG": 1272, + "TAGGTA": 1273, + "TAGGTT": 1274, + "TAGGTC": 1275, + "TAGGTG": 1276, + "TAGGCA": 1277, + "TAGGCT": 1278, + "TAGGCC": 1279, + "TAGGCG": 1280, + "TAGGGA": 1281, + "TAGGGT": 1282, + "TAGGGC": 1283, + "TAGGGG": 1284, + "TTAAAA": 1285, + "TTAAAT": 1286, + "TTAAAC": 1287, + "TTAAAG": 1288, + "TTAATA": 1289, + "TTAATT": 1290, + "TTAATC": 1291, + "TTAATG": 1292, + "TTAACA": 1293, + "TTAACT": 1294, + "TTAACC": 1295, + "TTAACG": 1296, + "TTAAGA": 1297, + "TTAAGT": 1298, + "TTAAGC": 1299, + "TTAAGG": 1300, + "TTATAA": 1301, + "TTATAT": 1302, + "TTATAC": 1303, + "TTATAG": 1304, + "TTATTA": 1305, + "TTATTT": 1306, + "TTATTC": 1307, + "TTATTG": 1308, + "TTATCA": 1309, + "TTATCT": 1310, + "TTATCC": 1311, + "TTATCG": 1312, + "TTATGA": 1313, + "TTATGT": 1314, + "TTATGC": 1315, + "TTATGG": 1316, + "TTACAA": 1317, + "TTACAT": 1318, + "TTACAC": 1319, + "TTACAG": 1320, + "TTACTA": 1321, + "TTACTT": 1322, + "TTACTC": 1323, + "TTACTG": 1324, + "TTACCA": 1325, + "TTACCT": 1326, + "TTACCC": 1327, + "TTACCG": 1328, + "TTACGA": 1329, + "TTACGT": 1330, + "TTACGC": 1331, + "TTACGG": 1332, + "TTAGAA": 1333, + "TTAGAT": 1334, + "TTAGAC": 1335, + "TTAGAG": 1336, + "TTAGTA": 1337, + "TTAGTT": 1338, + "TTAGTC": 1339, + "TTAGTG": 1340, + "TTAGCA": 1341, + "TTAGCT": 1342, + "TTAGCC": 1343, + "TTAGCG": 1344, + "TTAGGA": 1345, + "TTAGGT": 1346, + "TTAGGC": 1347, + "TTAGGG": 1348, + "TTTAAA": 1349, + "TTTAAT": 1350, + "TTTAAC": 1351, + "TTTAAG": 1352, + "TTTATA": 1353, + "TTTATT": 1354, + "TTTATC": 1355, + "TTTATG": 1356, + "TTTACA": 1357, + "TTTACT": 1358, + "TTTACC": 1359, + "TTTACG": 1360, + "TTTAGA": 1361, + "TTTAGT": 1362, + "TTTAGC": 1363, + "TTTAGG": 1364, + "TTTTAA": 1365, + "TTTTAT": 1366, + "TTTTAC": 1367, + "TTTTAG": 1368, + "TTTTTA": 1369, + "TTTTTT": 1370, + "TTTTTC": 1371, + "TTTTTG": 1372, + "TTTTCA": 1373, + "TTTTCT": 1374, + "TTTTCC": 1375, + "TTTTCG": 1376, + "TTTTGA": 1377, + "TTTTGT": 1378, + "TTTTGC": 1379, + "TTTTGG": 1380, + "TTTCAA": 1381, + "TTTCAT": 1382, + "TTTCAC": 1383, + "TTTCAG": 1384, + "TTTCTA": 1385, + "TTTCTT": 1386, + "TTTCTC": 1387, + "TTTCTG": 1388, + "TTTCCA": 1389, + "TTTCCT": 1390, + "TTTCCC": 1391, + "TTTCCG": 1392, + "TTTCGA": 1393, + "TTTCGT": 1394, + "TTTCGC": 1395, + "TTTCGG": 1396, + "TTTGAA": 1397, + "TTTGAT": 1398, + "TTTGAC": 1399, + "TTTGAG": 1400, + "TTTGTA": 1401, + "TTTGTT": 1402, + "TTTGTC": 1403, + "TTTGTG": 1404, + "TTTGCA": 1405, + "TTTGCT": 1406, + "TTTGCC": 1407, + "TTTGCG": 1408, + "TTTGGA": 1409, + "TTTGGT": 1410, + "TTTGGC": 1411, + "TTTGGG": 1412, + "TTCAAA": 1413, + "TTCAAT": 1414, + "TTCAAC": 1415, + "TTCAAG": 1416, + "TTCATA": 1417, + "TTCATT": 1418, + "TTCATC": 1419, + "TTCATG": 1420, + "TTCACA": 1421, + "TTCACT": 1422, + "TTCACC": 1423, + "TTCACG": 1424, + "TTCAGA": 1425, + "TTCAGT": 1426, + "TTCAGC": 1427, + "TTCAGG": 1428, + "TTCTAA": 1429, + "TTCTAT": 1430, + "TTCTAC": 1431, + "TTCTAG": 1432, + "TTCTTA": 1433, + "TTCTTT": 1434, + "TTCTTC": 1435, + "TTCTTG": 1436, + "TTCTCA": 1437, + "TTCTCT": 1438, + "TTCTCC": 1439, + "TTCTCG": 1440, + "TTCTGA": 1441, + "TTCTGT": 1442, + "TTCTGC": 1443, + "TTCTGG": 1444, + "TTCCAA": 1445, + "TTCCAT": 1446, + "TTCCAC": 1447, + "TTCCAG": 1448, + "TTCCTA": 1449, + "TTCCTT": 1450, + "TTCCTC": 1451, + "TTCCTG": 1452, + "TTCCCA": 1453, + "TTCCCT": 1454, + "TTCCCC": 1455, + "TTCCCG": 1456, + "TTCCGA": 1457, + "TTCCGT": 1458, + "TTCCGC": 1459, + "TTCCGG": 1460, + "TTCGAA": 1461, + "TTCGAT": 1462, + "TTCGAC": 1463, + "TTCGAG": 1464, + "TTCGTA": 1465, + "TTCGTT": 1466, + "TTCGTC": 1467, + "TTCGTG": 1468, + "TTCGCA": 1469, + "TTCGCT": 1470, + "TTCGCC": 1471, + "TTCGCG": 1472, + "TTCGGA": 1473, + "TTCGGT": 1474, + "TTCGGC": 1475, + "TTCGGG": 1476, + "TTGAAA": 1477, + "TTGAAT": 1478, + "TTGAAC": 1479, + "TTGAAG": 1480, + "TTGATA": 1481, + "TTGATT": 1482, + "TTGATC": 1483, + "TTGATG": 1484, + "TTGACA": 1485, + "TTGACT": 1486, + "TTGACC": 1487, + "TTGACG": 1488, + "TTGAGA": 1489, + "TTGAGT": 1490, + "TTGAGC": 1491, + "TTGAGG": 1492, + "TTGTAA": 1493, + "TTGTAT": 1494, + "TTGTAC": 1495, + "TTGTAG": 1496, + "TTGTTA": 1497, + "TTGTTT": 1498, + "TTGTTC": 1499, + "TTGTTG": 1500, + "TTGTCA": 1501, + "TTGTCT": 1502, + "TTGTCC": 1503, + "TTGTCG": 1504, + "TTGTGA": 1505, + "TTGTGT": 1506, + "TTGTGC": 1507, + "TTGTGG": 1508, + "TTGCAA": 1509, + "TTGCAT": 1510, + "TTGCAC": 1511, + "TTGCAG": 1512, + "TTGCTA": 1513, + "TTGCTT": 1514, + "TTGCTC": 1515, + "TTGCTG": 1516, + "TTGCCA": 1517, + "TTGCCT": 1518, + "TTGCCC": 1519, + "TTGCCG": 1520, + "TTGCGA": 1521, + "TTGCGT": 1522, + "TTGCGC": 1523, + "TTGCGG": 1524, + "TTGGAA": 1525, + "TTGGAT": 1526, + "TTGGAC": 1527, + "TTGGAG": 1528, + "TTGGTA": 1529, + "TTGGTT": 1530, + "TTGGTC": 1531, + "TTGGTG": 1532, + "TTGGCA": 1533, + "TTGGCT": 1534, + "TTGGCC": 1535, + "TTGGCG": 1536, + "TTGGGA": 1537, + "TTGGGT": 1538, + "TTGGGC": 1539, + "TTGGGG": 1540, + "TCAAAA": 1541, + "TCAAAT": 1542, + "TCAAAC": 1543, + "TCAAAG": 1544, + "TCAATA": 1545, + "TCAATT": 1546, + "TCAATC": 1547, + "TCAATG": 1548, + "TCAACA": 1549, + "TCAACT": 1550, + "TCAACC": 1551, + "TCAACG": 1552, + "TCAAGA": 1553, + "TCAAGT": 1554, + "TCAAGC": 1555, + "TCAAGG": 1556, + "TCATAA": 1557, + "TCATAT": 1558, + "TCATAC": 1559, + "TCATAG": 1560, + "TCATTA": 1561, + "TCATTT": 1562, + "TCATTC": 1563, + "TCATTG": 1564, + "TCATCA": 1565, + "TCATCT": 1566, + "TCATCC": 1567, + "TCATCG": 1568, + "TCATGA": 1569, + "TCATGT": 1570, + "TCATGC": 1571, + "TCATGG": 1572, + "TCACAA": 1573, + "TCACAT": 1574, + "TCACAC": 1575, + "TCACAG": 1576, + "TCACTA": 1577, + "TCACTT": 1578, + "TCACTC": 1579, + "TCACTG": 1580, + "TCACCA": 1581, + "TCACCT": 1582, + "TCACCC": 1583, + "TCACCG": 1584, + "TCACGA": 1585, + "TCACGT": 1586, + "TCACGC": 1587, + "TCACGG": 1588, + "TCAGAA": 1589, + "TCAGAT": 1590, + "TCAGAC": 1591, + "TCAGAG": 1592, + "TCAGTA": 1593, + "TCAGTT": 1594, + "TCAGTC": 1595, + "TCAGTG": 1596, + "TCAGCA": 1597, + "TCAGCT": 1598, + "TCAGCC": 1599, + "TCAGCG": 1600, + "TCAGGA": 1601, + "TCAGGT": 1602, + "TCAGGC": 1603, + "TCAGGG": 1604, + "TCTAAA": 1605, + "TCTAAT": 1606, + "TCTAAC": 1607, + "TCTAAG": 1608, + "TCTATA": 1609, + "TCTATT": 1610, + "TCTATC": 1611, + "TCTATG": 1612, + "TCTACA": 1613, + "TCTACT": 1614, + "TCTACC": 1615, + "TCTACG": 1616, + "TCTAGA": 1617, + "TCTAGT": 1618, + "TCTAGC": 1619, + "TCTAGG": 1620, + "TCTTAA": 1621, + "TCTTAT": 1622, + "TCTTAC": 1623, + "TCTTAG": 1624, + "TCTTTA": 1625, + "TCTTTT": 1626, + "TCTTTC": 1627, + "TCTTTG": 1628, + "TCTTCA": 1629, + "TCTTCT": 1630, + "TCTTCC": 1631, + "TCTTCG": 1632, + "TCTTGA": 1633, + "TCTTGT": 1634, + "TCTTGC": 1635, + "TCTTGG": 1636, + "TCTCAA": 1637, + "TCTCAT": 1638, + "TCTCAC": 1639, + "TCTCAG": 1640, + "TCTCTA": 1641, + "TCTCTT": 1642, + "TCTCTC": 1643, + "TCTCTG": 1644, + "TCTCCA": 1645, + "TCTCCT": 1646, + "TCTCCC": 1647, + "TCTCCG": 1648, + "TCTCGA": 1649, + "TCTCGT": 1650, + "TCTCGC": 1651, + "TCTCGG": 1652, + "TCTGAA": 1653, + "TCTGAT": 1654, + "TCTGAC": 1655, + "TCTGAG": 1656, + "TCTGTA": 1657, + "TCTGTT": 1658, + "TCTGTC": 1659, + "TCTGTG": 1660, + "TCTGCA": 1661, + "TCTGCT": 1662, + "TCTGCC": 1663, + "TCTGCG": 1664, + "TCTGGA": 1665, + "TCTGGT": 1666, + "TCTGGC": 1667, + "TCTGGG": 1668, + "TCCAAA": 1669, + "TCCAAT": 1670, + "TCCAAC": 1671, + "TCCAAG": 1672, + "TCCATA": 1673, + "TCCATT": 1674, + "TCCATC": 1675, + "TCCATG": 1676, + "TCCACA": 1677, + "TCCACT": 1678, + "TCCACC": 1679, + "TCCACG": 1680, + "TCCAGA": 1681, + "TCCAGT": 1682, + "TCCAGC": 1683, + "TCCAGG": 1684, + "TCCTAA": 1685, + "TCCTAT": 1686, + "TCCTAC": 1687, + "TCCTAG": 1688, + "TCCTTA": 1689, + "TCCTTT": 1690, + "TCCTTC": 1691, + "TCCTTG": 1692, + "TCCTCA": 1693, + "TCCTCT": 1694, + "TCCTCC": 1695, + "TCCTCG": 1696, + "TCCTGA": 1697, + "TCCTGT": 1698, + "TCCTGC": 1699, + "TCCTGG": 1700, + "TCCCAA": 1701, + "TCCCAT": 1702, + "TCCCAC": 1703, + "TCCCAG": 1704, + "TCCCTA": 1705, + "TCCCTT": 1706, + "TCCCTC": 1707, + "TCCCTG": 1708, + "TCCCCA": 1709, + "TCCCCT": 1710, + "TCCCCC": 1711, + "TCCCCG": 1712, + "TCCCGA": 1713, + "TCCCGT": 1714, + "TCCCGC": 1715, + "TCCCGG": 1716, + "TCCGAA": 1717, + "TCCGAT": 1718, + "TCCGAC": 1719, + "TCCGAG": 1720, + "TCCGTA": 1721, + "TCCGTT": 1722, + "TCCGTC": 1723, + "TCCGTG": 1724, + "TCCGCA": 1725, + "TCCGCT": 1726, + "TCCGCC": 1727, + "TCCGCG": 1728, + "TCCGGA": 1729, + "TCCGGT": 1730, + "TCCGGC": 1731, + "TCCGGG": 1732, + "TCGAAA": 1733, + "TCGAAT": 1734, + "TCGAAC": 1735, + "TCGAAG": 1736, + "TCGATA": 1737, + "TCGATT": 1738, + "TCGATC": 1739, + "TCGATG": 1740, + "TCGACA": 1741, + "TCGACT": 1742, + "TCGACC": 1743, + "TCGACG": 1744, + "TCGAGA": 1745, + "TCGAGT": 1746, + "TCGAGC": 1747, + "TCGAGG": 1748, + "TCGTAA": 1749, + "TCGTAT": 1750, + "TCGTAC": 1751, + "TCGTAG": 1752, + "TCGTTA": 1753, + "TCGTTT": 1754, + "TCGTTC": 1755, + "TCGTTG": 1756, + "TCGTCA": 1757, + "TCGTCT": 1758, + "TCGTCC": 1759, + "TCGTCG": 1760, + "TCGTGA": 1761, + "TCGTGT": 1762, + "TCGTGC": 1763, + "TCGTGG": 1764, + "TCGCAA": 1765, + "TCGCAT": 1766, + "TCGCAC": 1767, + "TCGCAG": 1768, + "TCGCTA": 1769, + "TCGCTT": 1770, + "TCGCTC": 1771, + "TCGCTG": 1772, + "TCGCCA": 1773, + "TCGCCT": 1774, + "TCGCCC": 1775, + "TCGCCG": 1776, + "TCGCGA": 1777, + "TCGCGT": 1778, + "TCGCGC": 1779, + "TCGCGG": 1780, + "TCGGAA": 1781, + "TCGGAT": 1782, + "TCGGAC": 1783, + "TCGGAG": 1784, + "TCGGTA": 1785, + "TCGGTT": 1786, + "TCGGTC": 1787, + "TCGGTG": 1788, + "TCGGCA": 1789, + "TCGGCT": 1790, + "TCGGCC": 1791, + "TCGGCG": 1792, + "TCGGGA": 1793, + "TCGGGT": 1794, + "TCGGGC": 1795, + "TCGGGG": 1796, + "TGAAAA": 1797, + "TGAAAT": 1798, + "TGAAAC": 1799, + "TGAAAG": 1800, + "TGAATA": 1801, + "TGAATT": 1802, + "TGAATC": 1803, + "TGAATG": 1804, + "TGAACA": 1805, + "TGAACT": 1806, + "TGAACC": 1807, + "TGAACG": 1808, + "TGAAGA": 1809, + "TGAAGT": 1810, + "TGAAGC": 1811, + "TGAAGG": 1812, + "TGATAA": 1813, + "TGATAT": 1814, + "TGATAC": 1815, + "TGATAG": 1816, + "TGATTA": 1817, + "TGATTT": 1818, + "TGATTC": 1819, + "TGATTG": 1820, + "TGATCA": 1821, + "TGATCT": 1822, + "TGATCC": 1823, + "TGATCG": 1824, + "TGATGA": 1825, + "TGATGT": 1826, + "TGATGC": 1827, + "TGATGG": 1828, + "TGACAA": 1829, + "TGACAT": 1830, + "TGACAC": 1831, + "TGACAG": 1832, + "TGACTA": 1833, + "TGACTT": 1834, + "TGACTC": 1835, + "TGACTG": 1836, + "TGACCA": 1837, + "TGACCT": 1838, + "TGACCC": 1839, + "TGACCG": 1840, + "TGACGA": 1841, + "TGACGT": 1842, + "TGACGC": 1843, + "TGACGG": 1844, + "TGAGAA": 1845, + "TGAGAT": 1846, + "TGAGAC": 1847, + "TGAGAG": 1848, + "TGAGTA": 1849, + "TGAGTT": 1850, + "TGAGTC": 1851, + "TGAGTG": 1852, + "TGAGCA": 1853, + "TGAGCT": 1854, + "TGAGCC": 1855, + "TGAGCG": 1856, + "TGAGGA": 1857, + "TGAGGT": 1858, + "TGAGGC": 1859, + "TGAGGG": 1860, + "TGTAAA": 1861, + "TGTAAT": 1862, + "TGTAAC": 1863, + "TGTAAG": 1864, + "TGTATA": 1865, + "TGTATT": 1866, + "TGTATC": 1867, + "TGTATG": 1868, + "TGTACA": 1869, + "TGTACT": 1870, + "TGTACC": 1871, + "TGTACG": 1872, + "TGTAGA": 1873, + "TGTAGT": 1874, + "TGTAGC": 1875, + "TGTAGG": 1876, + "TGTTAA": 1877, + "TGTTAT": 1878, + "TGTTAC": 1879, + "TGTTAG": 1880, + "TGTTTA": 1881, + "TGTTTT": 1882, + "TGTTTC": 1883, + "TGTTTG": 1884, + "TGTTCA": 1885, + "TGTTCT": 1886, + "TGTTCC": 1887, + "TGTTCG": 1888, + "TGTTGA": 1889, + "TGTTGT": 1890, + "TGTTGC": 1891, + "TGTTGG": 1892, + "TGTCAA": 1893, + "TGTCAT": 1894, + "TGTCAC": 1895, + "TGTCAG": 1896, + "TGTCTA": 1897, + "TGTCTT": 1898, + "TGTCTC": 1899, + "TGTCTG": 1900, + "TGTCCA": 1901, + "TGTCCT": 1902, + "TGTCCC": 1903, + "TGTCCG": 1904, + "TGTCGA": 1905, + "TGTCGT": 1906, + "TGTCGC": 1907, + "TGTCGG": 1908, + "TGTGAA": 1909, + "TGTGAT": 1910, + "TGTGAC": 1911, + "TGTGAG": 1912, + "TGTGTA": 1913, + "TGTGTT": 1914, + "TGTGTC": 1915, + "TGTGTG": 1916, + "TGTGCA": 1917, + "TGTGCT": 1918, + "TGTGCC": 1919, + "TGTGCG": 1920, + "TGTGGA": 1921, + "TGTGGT": 1922, + "TGTGGC": 1923, + "TGTGGG": 1924, + "TGCAAA": 1925, + "TGCAAT": 1926, + "TGCAAC": 1927, + "TGCAAG": 1928, + "TGCATA": 1929, + "TGCATT": 1930, + "TGCATC": 1931, + "TGCATG": 1932, + "TGCACA": 1933, + "TGCACT": 1934, + "TGCACC": 1935, + "TGCACG": 1936, + "TGCAGA": 1937, + "TGCAGT": 1938, + "TGCAGC": 1939, + "TGCAGG": 1940, + "TGCTAA": 1941, + "TGCTAT": 1942, + "TGCTAC": 1943, + "TGCTAG": 1944, + "TGCTTA": 1945, + "TGCTTT": 1946, + "TGCTTC": 1947, + "TGCTTG": 1948, + "TGCTCA": 1949, + "TGCTCT": 1950, + "TGCTCC": 1951, + "TGCTCG": 1952, + "TGCTGA": 1953, + "TGCTGT": 1954, + "TGCTGC": 1955, + "TGCTGG": 1956, + "TGCCAA": 1957, + "TGCCAT": 1958, + "TGCCAC": 1959, + "TGCCAG": 1960, + "TGCCTA": 1961, + "TGCCTT": 1962, + "TGCCTC": 1963, + "TGCCTG": 1964, + "TGCCCA": 1965, + "TGCCCT": 1966, + "TGCCCC": 1967, + "TGCCCG": 1968, + "TGCCGA": 1969, + "TGCCGT": 1970, + "TGCCGC": 1971, + "TGCCGG": 1972, + "TGCGAA": 1973, + "TGCGAT": 1974, + "TGCGAC": 1975, + "TGCGAG": 1976, + "TGCGTA": 1977, + "TGCGTT": 1978, + "TGCGTC": 1979, + "TGCGTG": 1980, + "TGCGCA": 1981, + "TGCGCT": 1982, + "TGCGCC": 1983, + "TGCGCG": 1984, + "TGCGGA": 1985, + "TGCGGT": 1986, + "TGCGGC": 1987, + "TGCGGG": 1988, + "TGGAAA": 1989, + "TGGAAT": 1990, + "TGGAAC": 1991, + "TGGAAG": 1992, + "TGGATA": 1993, + "TGGATT": 1994, + "TGGATC": 1995, + "TGGATG": 1996, + "TGGACA": 1997, + "TGGACT": 1998, + "TGGACC": 1999, + "TGGACG": 2000, + "TGGAGA": 2001, + "TGGAGT": 2002, + "TGGAGC": 2003, + "TGGAGG": 2004, + "TGGTAA": 2005, + "TGGTAT": 2006, + "TGGTAC": 2007, + "TGGTAG": 2008, + "TGGTTA": 2009, + "TGGTTT": 2010, + "TGGTTC": 2011, + "TGGTTG": 2012, + "TGGTCA": 2013, + "TGGTCT": 2014, + "TGGTCC": 2015, + "TGGTCG": 2016, + "TGGTGA": 2017, + "TGGTGT": 2018, + "TGGTGC": 2019, + "TGGTGG": 2020, + "TGGCAA": 2021, + "TGGCAT": 2022, + "TGGCAC": 2023, + "TGGCAG": 2024, + "TGGCTA": 2025, + "TGGCTT": 2026, + "TGGCTC": 2027, + "TGGCTG": 2028, + "TGGCCA": 2029, + "TGGCCT": 2030, + "TGGCCC": 2031, + "TGGCCG": 2032, + "TGGCGA": 2033, + "TGGCGT": 2034, + "TGGCGC": 2035, + "TGGCGG": 2036, + "TGGGAA": 2037, + "TGGGAT": 2038, + "TGGGAC": 2039, + "TGGGAG": 2040, + "TGGGTA": 2041, + "TGGGTT": 2042, + "TGGGTC": 2043, + "TGGGTG": 2044, + "TGGGCA": 2045, + "TGGGCT": 2046, + "TGGGCC": 2047, + "TGGGCG": 2048, + "TGGGGA": 2049, + "TGGGGT": 2050, + "TGGGGC": 2051, + "TGGGGG": 2052, + "CAAAAA": 2053, + "CAAAAT": 2054, + "CAAAAC": 2055, + "CAAAAG": 2056, + "CAAATA": 2057, + "CAAATT": 2058, + "CAAATC": 2059, + "CAAATG": 2060, + "CAAACA": 2061, + "CAAACT": 2062, + "CAAACC": 2063, + "CAAACG": 2064, + "CAAAGA": 2065, + "CAAAGT": 2066, + "CAAAGC": 2067, + "CAAAGG": 2068, + "CAATAA": 2069, + "CAATAT": 2070, + "CAATAC": 2071, + "CAATAG": 2072, + "CAATTA": 2073, + "CAATTT": 2074, + "CAATTC": 2075, + "CAATTG": 2076, + "CAATCA": 2077, + "CAATCT": 2078, + "CAATCC": 2079, + "CAATCG": 2080, + "CAATGA": 2081, + "CAATGT": 2082, + "CAATGC": 2083, + "CAATGG": 2084, + "CAACAA": 2085, + "CAACAT": 2086, + "CAACAC": 2087, + "CAACAG": 2088, + "CAACTA": 2089, + "CAACTT": 2090, + "CAACTC": 2091, + "CAACTG": 2092, + "CAACCA": 2093, + "CAACCT": 2094, + "CAACCC": 2095, + "CAACCG": 2096, + "CAACGA": 2097, + "CAACGT": 2098, + "CAACGC": 2099, + "CAACGG": 2100, + "CAAGAA": 2101, + "CAAGAT": 2102, + "CAAGAC": 2103, + "CAAGAG": 2104, + "CAAGTA": 2105, + "CAAGTT": 2106, + "CAAGTC": 2107, + "CAAGTG": 2108, + "CAAGCA": 2109, + "CAAGCT": 2110, + "CAAGCC": 2111, + "CAAGCG": 2112, + "CAAGGA": 2113, + "CAAGGT": 2114, + "CAAGGC": 2115, + "CAAGGG": 2116, + "CATAAA": 2117, + "CATAAT": 2118, + "CATAAC": 2119, + "CATAAG": 2120, + "CATATA": 2121, + "CATATT": 2122, + "CATATC": 2123, + "CATATG": 2124, + "CATACA": 2125, + "CATACT": 2126, + "CATACC": 2127, + "CATACG": 2128, + "CATAGA": 2129, + "CATAGT": 2130, + "CATAGC": 2131, + "CATAGG": 2132, + "CATTAA": 2133, + "CATTAT": 2134, + "CATTAC": 2135, + "CATTAG": 2136, + "CATTTA": 2137, + "CATTTT": 2138, + "CATTTC": 2139, + "CATTTG": 2140, + "CATTCA": 2141, + "CATTCT": 2142, + "CATTCC": 2143, + "CATTCG": 2144, + "CATTGA": 2145, + "CATTGT": 2146, + "CATTGC": 2147, + "CATTGG": 2148, + "CATCAA": 2149, + "CATCAT": 2150, + "CATCAC": 2151, + "CATCAG": 2152, + "CATCTA": 2153, + "CATCTT": 2154, + "CATCTC": 2155, + "CATCTG": 2156, + "CATCCA": 2157, + "CATCCT": 2158, + "CATCCC": 2159, + "CATCCG": 2160, + "CATCGA": 2161, + "CATCGT": 2162, + "CATCGC": 2163, + "CATCGG": 2164, + "CATGAA": 2165, + "CATGAT": 2166, + "CATGAC": 2167, + "CATGAG": 2168, + "CATGTA": 2169, + "CATGTT": 2170, + "CATGTC": 2171, + "CATGTG": 2172, + "CATGCA": 2173, + "CATGCT": 2174, + "CATGCC": 2175, + "CATGCG": 2176, + "CATGGA": 2177, + "CATGGT": 2178, + "CATGGC": 2179, + "CATGGG": 2180, + "CACAAA": 2181, + "CACAAT": 2182, + "CACAAC": 2183, + "CACAAG": 2184, + "CACATA": 2185, + "CACATT": 2186, + "CACATC": 2187, + "CACATG": 2188, + "CACACA": 2189, + "CACACT": 2190, + "CACACC": 2191, + "CACACG": 2192, + "CACAGA": 2193, + "CACAGT": 2194, + "CACAGC": 2195, + "CACAGG": 2196, + "CACTAA": 2197, + "CACTAT": 2198, + "CACTAC": 2199, + "CACTAG": 2200, + "CACTTA": 2201, + "CACTTT": 2202, + "CACTTC": 2203, + "CACTTG": 2204, + "CACTCA": 2205, + "CACTCT": 2206, + "CACTCC": 2207, + "CACTCG": 2208, + "CACTGA": 2209, + "CACTGT": 2210, + "CACTGC": 2211, + "CACTGG": 2212, + "CACCAA": 2213, + "CACCAT": 2214, + "CACCAC": 2215, + "CACCAG": 2216, + "CACCTA": 2217, + "CACCTT": 2218, + "CACCTC": 2219, + "CACCTG": 2220, + "CACCCA": 2221, + "CACCCT": 2222, + "CACCCC": 2223, + "CACCCG": 2224, + "CACCGA": 2225, + "CACCGT": 2226, + "CACCGC": 2227, + "CACCGG": 2228, + "CACGAA": 2229, + "CACGAT": 2230, + "CACGAC": 2231, + "CACGAG": 2232, + "CACGTA": 2233, + "CACGTT": 2234, + "CACGTC": 2235, + "CACGTG": 2236, + "CACGCA": 2237, + "CACGCT": 2238, + "CACGCC": 2239, + "CACGCG": 2240, + "CACGGA": 2241, + "CACGGT": 2242, + "CACGGC": 2243, + "CACGGG": 2244, + "CAGAAA": 2245, + "CAGAAT": 2246, + "CAGAAC": 2247, + "CAGAAG": 2248, + "CAGATA": 2249, + "CAGATT": 2250, + "CAGATC": 2251, + "CAGATG": 2252, + "CAGACA": 2253, + "CAGACT": 2254, + "CAGACC": 2255, + "CAGACG": 2256, + "CAGAGA": 2257, + "CAGAGT": 2258, + "CAGAGC": 2259, + "CAGAGG": 2260, + "CAGTAA": 2261, + "CAGTAT": 2262, + "CAGTAC": 2263, + "CAGTAG": 2264, + "CAGTTA": 2265, + "CAGTTT": 2266, + "CAGTTC": 2267, + "CAGTTG": 2268, + "CAGTCA": 2269, + "CAGTCT": 2270, + "CAGTCC": 2271, + "CAGTCG": 2272, + "CAGTGA": 2273, + "CAGTGT": 2274, + "CAGTGC": 2275, + "CAGTGG": 2276, + "CAGCAA": 2277, + "CAGCAT": 2278, + "CAGCAC": 2279, + "CAGCAG": 2280, + "CAGCTA": 2281, + "CAGCTT": 2282, + "CAGCTC": 2283, + "CAGCTG": 2284, + "CAGCCA": 2285, + "CAGCCT": 2286, + "CAGCCC": 2287, + "CAGCCG": 2288, + "CAGCGA": 2289, + "CAGCGT": 2290, + "CAGCGC": 2291, + "CAGCGG": 2292, + "CAGGAA": 2293, + "CAGGAT": 2294, + "CAGGAC": 2295, + "CAGGAG": 2296, + "CAGGTA": 2297, + "CAGGTT": 2298, + "CAGGTC": 2299, + "CAGGTG": 2300, + "CAGGCA": 2301, + "CAGGCT": 2302, + "CAGGCC": 2303, + "CAGGCG": 2304, + "CAGGGA": 2305, + "CAGGGT": 2306, + "CAGGGC": 2307, + "CAGGGG": 2308, + "CTAAAA": 2309, + "CTAAAT": 2310, + "CTAAAC": 2311, + "CTAAAG": 2312, + "CTAATA": 2313, + "CTAATT": 2314, + "CTAATC": 2315, + "CTAATG": 2316, + "CTAACA": 2317, + "CTAACT": 2318, + "CTAACC": 2319, + "CTAACG": 2320, + "CTAAGA": 2321, + "CTAAGT": 2322, + "CTAAGC": 2323, + "CTAAGG": 2324, + "CTATAA": 2325, + "CTATAT": 2326, + "CTATAC": 2327, + "CTATAG": 2328, + "CTATTA": 2329, + "CTATTT": 2330, + "CTATTC": 2331, + "CTATTG": 2332, + "CTATCA": 2333, + "CTATCT": 2334, + "CTATCC": 2335, + "CTATCG": 2336, + "CTATGA": 2337, + "CTATGT": 2338, + "CTATGC": 2339, + "CTATGG": 2340, + "CTACAA": 2341, + "CTACAT": 2342, + "CTACAC": 2343, + "CTACAG": 2344, + "CTACTA": 2345, + "CTACTT": 2346, + "CTACTC": 2347, + "CTACTG": 2348, + "CTACCA": 2349, + "CTACCT": 2350, + "CTACCC": 2351, + "CTACCG": 2352, + "CTACGA": 2353, + "CTACGT": 2354, + "CTACGC": 2355, + "CTACGG": 2356, + "CTAGAA": 2357, + "CTAGAT": 2358, + "CTAGAC": 2359, + "CTAGAG": 2360, + "CTAGTA": 2361, + "CTAGTT": 2362, + "CTAGTC": 2363, + "CTAGTG": 2364, + "CTAGCA": 2365, + "CTAGCT": 2366, + "CTAGCC": 2367, + "CTAGCG": 2368, + "CTAGGA": 2369, + "CTAGGT": 2370, + "CTAGGC": 2371, + "CTAGGG": 2372, + "CTTAAA": 2373, + "CTTAAT": 2374, + "CTTAAC": 2375, + "CTTAAG": 2376, + "CTTATA": 2377, + "CTTATT": 2378, + "CTTATC": 2379, + "CTTATG": 2380, + "CTTACA": 2381, + "CTTACT": 2382, + "CTTACC": 2383, + "CTTACG": 2384, + "CTTAGA": 2385, + "CTTAGT": 2386, + "CTTAGC": 2387, + "CTTAGG": 2388, + "CTTTAA": 2389, + "CTTTAT": 2390, + "CTTTAC": 2391, + "CTTTAG": 2392, + "CTTTTA": 2393, + "CTTTTT": 2394, + "CTTTTC": 2395, + "CTTTTG": 2396, + "CTTTCA": 2397, + "CTTTCT": 2398, + "CTTTCC": 2399, + "CTTTCG": 2400, + "CTTTGA": 2401, + "CTTTGT": 2402, + "CTTTGC": 2403, + "CTTTGG": 2404, + "CTTCAA": 2405, + "CTTCAT": 2406, + "CTTCAC": 2407, + "CTTCAG": 2408, + "CTTCTA": 2409, + "CTTCTT": 2410, + "CTTCTC": 2411, + "CTTCTG": 2412, + "CTTCCA": 2413, + "CTTCCT": 2414, + "CTTCCC": 2415, + "CTTCCG": 2416, + "CTTCGA": 2417, + "CTTCGT": 2418, + "CTTCGC": 2419, + "CTTCGG": 2420, + "CTTGAA": 2421, + "CTTGAT": 2422, + "CTTGAC": 2423, + "CTTGAG": 2424, + "CTTGTA": 2425, + "CTTGTT": 2426, + "CTTGTC": 2427, + "CTTGTG": 2428, + "CTTGCA": 2429, + "CTTGCT": 2430, + "CTTGCC": 2431, + "CTTGCG": 2432, + "CTTGGA": 2433, + "CTTGGT": 2434, + "CTTGGC": 2435, + "CTTGGG": 2436, + "CTCAAA": 2437, + "CTCAAT": 2438, + "CTCAAC": 2439, + "CTCAAG": 2440, + "CTCATA": 2441, + "CTCATT": 2442, + "CTCATC": 2443, + "CTCATG": 2444, + "CTCACA": 2445, + "CTCACT": 2446, + "CTCACC": 2447, + "CTCACG": 2448, + "CTCAGA": 2449, + "CTCAGT": 2450, + "CTCAGC": 2451, + "CTCAGG": 2452, + "CTCTAA": 2453, + "CTCTAT": 2454, + "CTCTAC": 2455, + "CTCTAG": 2456, + "CTCTTA": 2457, + "CTCTTT": 2458, + "CTCTTC": 2459, + "CTCTTG": 2460, + "CTCTCA": 2461, + "CTCTCT": 2462, + "CTCTCC": 2463, + "CTCTCG": 2464, + "CTCTGA": 2465, + "CTCTGT": 2466, + "CTCTGC": 2467, + "CTCTGG": 2468, + "CTCCAA": 2469, + "CTCCAT": 2470, + "CTCCAC": 2471, + "CTCCAG": 2472, + "CTCCTA": 2473, + "CTCCTT": 2474, + "CTCCTC": 2475, + "CTCCTG": 2476, + "CTCCCA": 2477, + "CTCCCT": 2478, + "CTCCCC": 2479, + "CTCCCG": 2480, + "CTCCGA": 2481, + "CTCCGT": 2482, + "CTCCGC": 2483, + "CTCCGG": 2484, + "CTCGAA": 2485, + "CTCGAT": 2486, + "CTCGAC": 2487, + "CTCGAG": 2488, + "CTCGTA": 2489, + "CTCGTT": 2490, + "CTCGTC": 2491, + "CTCGTG": 2492, + "CTCGCA": 2493, + "CTCGCT": 2494, + "CTCGCC": 2495, + "CTCGCG": 2496, + "CTCGGA": 2497, + "CTCGGT": 2498, + "CTCGGC": 2499, + "CTCGGG": 2500, + "CTGAAA": 2501, + "CTGAAT": 2502, + "CTGAAC": 2503, + "CTGAAG": 2504, + "CTGATA": 2505, + "CTGATT": 2506, + "CTGATC": 2507, + "CTGATG": 2508, + "CTGACA": 2509, + "CTGACT": 2510, + "CTGACC": 2511, + "CTGACG": 2512, + "CTGAGA": 2513, + "CTGAGT": 2514, + "CTGAGC": 2515, + "CTGAGG": 2516, + "CTGTAA": 2517, + "CTGTAT": 2518, + "CTGTAC": 2519, + "CTGTAG": 2520, + "CTGTTA": 2521, + "CTGTTT": 2522, + "CTGTTC": 2523, + "CTGTTG": 2524, + "CTGTCA": 2525, + "CTGTCT": 2526, + "CTGTCC": 2527, + "CTGTCG": 2528, + "CTGTGA": 2529, + "CTGTGT": 2530, + "CTGTGC": 2531, + "CTGTGG": 2532, + "CTGCAA": 2533, + "CTGCAT": 2534, + "CTGCAC": 2535, + "CTGCAG": 2536, + "CTGCTA": 2537, + "CTGCTT": 2538, + "CTGCTC": 2539, + "CTGCTG": 2540, + "CTGCCA": 2541, + "CTGCCT": 2542, + "CTGCCC": 2543, + "CTGCCG": 2544, + "CTGCGA": 2545, + "CTGCGT": 2546, + "CTGCGC": 2547, + "CTGCGG": 2548, + "CTGGAA": 2549, + "CTGGAT": 2550, + "CTGGAC": 2551, + "CTGGAG": 2552, + "CTGGTA": 2553, + "CTGGTT": 2554, + "CTGGTC": 2555, + "CTGGTG": 2556, + "CTGGCA": 2557, + "CTGGCT": 2558, + "CTGGCC": 2559, + "CTGGCG": 2560, + "CTGGGA": 2561, + "CTGGGT": 2562, + "CTGGGC": 2563, + "CTGGGG": 2564, + "CCAAAA": 2565, + "CCAAAT": 2566, + "CCAAAC": 2567, + "CCAAAG": 2568, + "CCAATA": 2569, + "CCAATT": 2570, + "CCAATC": 2571, + "CCAATG": 2572, + "CCAACA": 2573, + "CCAACT": 2574, + "CCAACC": 2575, + "CCAACG": 2576, + "CCAAGA": 2577, + "CCAAGT": 2578, + "CCAAGC": 2579, + "CCAAGG": 2580, + "CCATAA": 2581, + "CCATAT": 2582, + "CCATAC": 2583, + "CCATAG": 2584, + "CCATTA": 2585, + "CCATTT": 2586, + "CCATTC": 2587, + "CCATTG": 2588, + "CCATCA": 2589, + "CCATCT": 2590, + "CCATCC": 2591, + "CCATCG": 2592, + "CCATGA": 2593, + "CCATGT": 2594, + "CCATGC": 2595, + "CCATGG": 2596, + "CCACAA": 2597, + "CCACAT": 2598, + "CCACAC": 2599, + "CCACAG": 2600, + "CCACTA": 2601, + "CCACTT": 2602, + "CCACTC": 2603, + "CCACTG": 2604, + "CCACCA": 2605, + "CCACCT": 2606, + "CCACCC": 2607, + "CCACCG": 2608, + "CCACGA": 2609, + "CCACGT": 2610, + "CCACGC": 2611, + "CCACGG": 2612, + "CCAGAA": 2613, + "CCAGAT": 2614, + "CCAGAC": 2615, + "CCAGAG": 2616, + "CCAGTA": 2617, + "CCAGTT": 2618, + "CCAGTC": 2619, + "CCAGTG": 2620, + "CCAGCA": 2621, + "CCAGCT": 2622, + "CCAGCC": 2623, + "CCAGCG": 2624, + "CCAGGA": 2625, + "CCAGGT": 2626, + "CCAGGC": 2627, + "CCAGGG": 2628, + "CCTAAA": 2629, + "CCTAAT": 2630, + "CCTAAC": 2631, + "CCTAAG": 2632, + "CCTATA": 2633, + "CCTATT": 2634, + "CCTATC": 2635, + "CCTATG": 2636, + "CCTACA": 2637, + "CCTACT": 2638, + "CCTACC": 2639, + "CCTACG": 2640, + "CCTAGA": 2641, + "CCTAGT": 2642, + "CCTAGC": 2643, + "CCTAGG": 2644, + "CCTTAA": 2645, + "CCTTAT": 2646, + "CCTTAC": 2647, + "CCTTAG": 2648, + "CCTTTA": 2649, + "CCTTTT": 2650, + "CCTTTC": 2651, + "CCTTTG": 2652, + "CCTTCA": 2653, + "CCTTCT": 2654, + "CCTTCC": 2655, + "CCTTCG": 2656, + "CCTTGA": 2657, + "CCTTGT": 2658, + "CCTTGC": 2659, + "CCTTGG": 2660, + "CCTCAA": 2661, + "CCTCAT": 2662, + "CCTCAC": 2663, + "CCTCAG": 2664, + "CCTCTA": 2665, + "CCTCTT": 2666, + "CCTCTC": 2667, + "CCTCTG": 2668, + "CCTCCA": 2669, + "CCTCCT": 2670, + "CCTCCC": 2671, + "CCTCCG": 2672, + "CCTCGA": 2673, + "CCTCGT": 2674, + "CCTCGC": 2675, + "CCTCGG": 2676, + "CCTGAA": 2677, + "CCTGAT": 2678, + "CCTGAC": 2679, + "CCTGAG": 2680, + "CCTGTA": 2681, + "CCTGTT": 2682, + "CCTGTC": 2683, + "CCTGTG": 2684, + "CCTGCA": 2685, + "CCTGCT": 2686, + "CCTGCC": 2687, + "CCTGCG": 2688, + "CCTGGA": 2689, + "CCTGGT": 2690, + "CCTGGC": 2691, + "CCTGGG": 2692, + "CCCAAA": 2693, + "CCCAAT": 2694, + "CCCAAC": 2695, + "CCCAAG": 2696, + "CCCATA": 2697, + "CCCATT": 2698, + "CCCATC": 2699, + "CCCATG": 2700, + "CCCACA": 2701, + "CCCACT": 2702, + "CCCACC": 2703, + "CCCACG": 2704, + "CCCAGA": 2705, + "CCCAGT": 2706, + "CCCAGC": 2707, + "CCCAGG": 2708, + "CCCTAA": 2709, + "CCCTAT": 2710, + "CCCTAC": 2711, + "CCCTAG": 2712, + "CCCTTA": 2713, + "CCCTTT": 2714, + "CCCTTC": 2715, + "CCCTTG": 2716, + "CCCTCA": 2717, + "CCCTCT": 2718, + "CCCTCC": 2719, + "CCCTCG": 2720, + "CCCTGA": 2721, + "CCCTGT": 2722, + "CCCTGC": 2723, + "CCCTGG": 2724, + "CCCCAA": 2725, + "CCCCAT": 2726, + "CCCCAC": 2727, + "CCCCAG": 2728, + "CCCCTA": 2729, + "CCCCTT": 2730, + "CCCCTC": 2731, + "CCCCTG": 2732, + "CCCCCA": 2733, + "CCCCCT": 2734, + "CCCCCC": 2735, + "CCCCCG": 2736, + "CCCCGA": 2737, + "CCCCGT": 2738, + "CCCCGC": 2739, + "CCCCGG": 2740, + "CCCGAA": 2741, + "CCCGAT": 2742, + "CCCGAC": 2743, + "CCCGAG": 2744, + "CCCGTA": 2745, + "CCCGTT": 2746, + "CCCGTC": 2747, + "CCCGTG": 2748, + "CCCGCA": 2749, + "CCCGCT": 2750, + "CCCGCC": 2751, + "CCCGCG": 2752, + "CCCGGA": 2753, + "CCCGGT": 2754, + "CCCGGC": 2755, + "CCCGGG": 2756, + "CCGAAA": 2757, + "CCGAAT": 2758, + "CCGAAC": 2759, + "CCGAAG": 2760, + "CCGATA": 2761, + "CCGATT": 2762, + "CCGATC": 2763, + "CCGATG": 2764, + "CCGACA": 2765, + "CCGACT": 2766, + "CCGACC": 2767, + "CCGACG": 2768, + "CCGAGA": 2769, + "CCGAGT": 2770, + "CCGAGC": 2771, + "CCGAGG": 2772, + "CCGTAA": 2773, + "CCGTAT": 2774, + "CCGTAC": 2775, + "CCGTAG": 2776, + "CCGTTA": 2777, + "CCGTTT": 2778, + "CCGTTC": 2779, + "CCGTTG": 2780, + "CCGTCA": 2781, + "CCGTCT": 2782, + "CCGTCC": 2783, + "CCGTCG": 2784, + "CCGTGA": 2785, + "CCGTGT": 2786, + "CCGTGC": 2787, + "CCGTGG": 2788, + "CCGCAA": 2789, + "CCGCAT": 2790, + "CCGCAC": 2791, + "CCGCAG": 2792, + "CCGCTA": 2793, + "CCGCTT": 2794, + "CCGCTC": 2795, + "CCGCTG": 2796, + "CCGCCA": 2797, + "CCGCCT": 2798, + "CCGCCC": 2799, + "CCGCCG": 2800, + "CCGCGA": 2801, + "CCGCGT": 2802, + "CCGCGC": 2803, + "CCGCGG": 2804, + "CCGGAA": 2805, + "CCGGAT": 2806, + "CCGGAC": 2807, + "CCGGAG": 2808, + "CCGGTA": 2809, + "CCGGTT": 2810, + "CCGGTC": 2811, + "CCGGTG": 2812, + "CCGGCA": 2813, + "CCGGCT": 2814, + "CCGGCC": 2815, + "CCGGCG": 2816, + "CCGGGA": 2817, + "CCGGGT": 2818, + "CCGGGC": 2819, + "CCGGGG": 2820, + "CGAAAA": 2821, + "CGAAAT": 2822, + "CGAAAC": 2823, + "CGAAAG": 2824, + "CGAATA": 2825, + "CGAATT": 2826, + "CGAATC": 2827, + "CGAATG": 2828, + "CGAACA": 2829, + "CGAACT": 2830, + "CGAACC": 2831, + "CGAACG": 2832, + "CGAAGA": 2833, + "CGAAGT": 2834, + "CGAAGC": 2835, + "CGAAGG": 2836, + "CGATAA": 2837, + "CGATAT": 2838, + "CGATAC": 2839, + "CGATAG": 2840, + "CGATTA": 2841, + "CGATTT": 2842, + "CGATTC": 2843, + "CGATTG": 2844, + "CGATCA": 2845, + "CGATCT": 2846, + "CGATCC": 2847, + "CGATCG": 2848, + "CGATGA": 2849, + "CGATGT": 2850, + "CGATGC": 2851, + "CGATGG": 2852, + "CGACAA": 2853, + "CGACAT": 2854, + "CGACAC": 2855, + "CGACAG": 2856, + "CGACTA": 2857, + "CGACTT": 2858, + "CGACTC": 2859, + "CGACTG": 2860, + "CGACCA": 2861, + "CGACCT": 2862, + "CGACCC": 2863, + "CGACCG": 2864, + "CGACGA": 2865, + "CGACGT": 2866, + "CGACGC": 2867, + "CGACGG": 2868, + "CGAGAA": 2869, + "CGAGAT": 2870, + "CGAGAC": 2871, + "CGAGAG": 2872, + "CGAGTA": 2873, + "CGAGTT": 2874, + "CGAGTC": 2875, + "CGAGTG": 2876, + "CGAGCA": 2877, + "CGAGCT": 2878, + "CGAGCC": 2879, + "CGAGCG": 2880, + "CGAGGA": 2881, + "CGAGGT": 2882, + "CGAGGC": 2883, + "CGAGGG": 2884, + "CGTAAA": 2885, + "CGTAAT": 2886, + "CGTAAC": 2887, + "CGTAAG": 2888, + "CGTATA": 2889, + "CGTATT": 2890, + "CGTATC": 2891, + "CGTATG": 2892, + "CGTACA": 2893, + "CGTACT": 2894, + "CGTACC": 2895, + "CGTACG": 2896, + "CGTAGA": 2897, + "CGTAGT": 2898, + "CGTAGC": 2899, + "CGTAGG": 2900, + "CGTTAA": 2901, + "CGTTAT": 2902, + "CGTTAC": 2903, + "CGTTAG": 2904, + "CGTTTA": 2905, + "CGTTTT": 2906, + "CGTTTC": 2907, + "CGTTTG": 2908, + "CGTTCA": 2909, + "CGTTCT": 2910, + "CGTTCC": 2911, + "CGTTCG": 2912, + "CGTTGA": 2913, + "CGTTGT": 2914, + "CGTTGC": 2915, + "CGTTGG": 2916, + "CGTCAA": 2917, + "CGTCAT": 2918, + "CGTCAC": 2919, + "CGTCAG": 2920, + "CGTCTA": 2921, + "CGTCTT": 2922, + "CGTCTC": 2923, + "CGTCTG": 2924, + "CGTCCA": 2925, + "CGTCCT": 2926, + "CGTCCC": 2927, + "CGTCCG": 2928, + "CGTCGA": 2929, + "CGTCGT": 2930, + "CGTCGC": 2931, + "CGTCGG": 2932, + "CGTGAA": 2933, + "CGTGAT": 2934, + "CGTGAC": 2935, + "CGTGAG": 2936, + "CGTGTA": 2937, + "CGTGTT": 2938, + "CGTGTC": 2939, + "CGTGTG": 2940, + "CGTGCA": 2941, + "CGTGCT": 2942, + "CGTGCC": 2943, + "CGTGCG": 2944, + "CGTGGA": 2945, + "CGTGGT": 2946, + "CGTGGC": 2947, + "CGTGGG": 2948, + "CGCAAA": 2949, + "CGCAAT": 2950, + "CGCAAC": 2951, + "CGCAAG": 2952, + "CGCATA": 2953, + "CGCATT": 2954, + "CGCATC": 2955, + "CGCATG": 2956, + "CGCACA": 2957, + "CGCACT": 2958, + "CGCACC": 2959, + "CGCACG": 2960, + "CGCAGA": 2961, + "CGCAGT": 2962, + "CGCAGC": 2963, + "CGCAGG": 2964, + "CGCTAA": 2965, + "CGCTAT": 2966, + "CGCTAC": 2967, + "CGCTAG": 2968, + "CGCTTA": 2969, + "CGCTTT": 2970, + "CGCTTC": 2971, + "CGCTTG": 2972, + "CGCTCA": 2973, + "CGCTCT": 2974, + "CGCTCC": 2975, + "CGCTCG": 2976, + "CGCTGA": 2977, + "CGCTGT": 2978, + "CGCTGC": 2979, + "CGCTGG": 2980, + "CGCCAA": 2981, + "CGCCAT": 2982, + "CGCCAC": 2983, + "CGCCAG": 2984, + "CGCCTA": 2985, + "CGCCTT": 2986, + "CGCCTC": 2987, + "CGCCTG": 2988, + "CGCCCA": 2989, + "CGCCCT": 2990, + "CGCCCC": 2991, + "CGCCCG": 2992, + "CGCCGA": 2993, + "CGCCGT": 2994, + "CGCCGC": 2995, + "CGCCGG": 2996, + "CGCGAA": 2997, + "CGCGAT": 2998, + "CGCGAC": 2999, + "CGCGAG": 3000, + "CGCGTA": 3001, + "CGCGTT": 3002, + "CGCGTC": 3003, + "CGCGTG": 3004, + "CGCGCA": 3005, + "CGCGCT": 3006, + "CGCGCC": 3007, + "CGCGCG": 3008, + "CGCGGA": 3009, + "CGCGGT": 3010, + "CGCGGC": 3011, + "CGCGGG": 3012, + "CGGAAA": 3013, + "CGGAAT": 3014, + "CGGAAC": 3015, + "CGGAAG": 3016, + "CGGATA": 3017, + "CGGATT": 3018, + "CGGATC": 3019, + "CGGATG": 3020, + "CGGACA": 3021, + "CGGACT": 3022, + "CGGACC": 3023, + "CGGACG": 3024, + "CGGAGA": 3025, + "CGGAGT": 3026, + "CGGAGC": 3027, + "CGGAGG": 3028, + "CGGTAA": 3029, + "CGGTAT": 3030, + "CGGTAC": 3031, + "CGGTAG": 3032, + "CGGTTA": 3033, + "CGGTTT": 3034, + "CGGTTC": 3035, + "CGGTTG": 3036, + "CGGTCA": 3037, + "CGGTCT": 3038, + "CGGTCC": 3039, + "CGGTCG": 3040, + "CGGTGA": 3041, + "CGGTGT": 3042, + "CGGTGC": 3043, + "CGGTGG": 3044, + "CGGCAA": 3045, + "CGGCAT": 3046, + "CGGCAC": 3047, + "CGGCAG": 3048, + "CGGCTA": 3049, + "CGGCTT": 3050, + "CGGCTC": 3051, + "CGGCTG": 3052, + "CGGCCA": 3053, + "CGGCCT": 3054, + "CGGCCC": 3055, + "CGGCCG": 3056, + "CGGCGA": 3057, + "CGGCGT": 3058, + "CGGCGC": 3059, + "CGGCGG": 3060, + "CGGGAA": 3061, + "CGGGAT": 3062, + "CGGGAC": 3063, + "CGGGAG": 3064, + "CGGGTA": 3065, + "CGGGTT": 3066, + "CGGGTC": 3067, + "CGGGTG": 3068, + "CGGGCA": 3069, + "CGGGCT": 3070, + "CGGGCC": 3071, + "CGGGCG": 3072, + "CGGGGA": 3073, + "CGGGGT": 3074, + "CGGGGC": 3075, + "CGGGGG": 3076, + "GAAAAA": 3077, + "GAAAAT": 3078, + "GAAAAC": 3079, + "GAAAAG": 3080, + "GAAATA": 3081, + "GAAATT": 3082, + "GAAATC": 3083, + "GAAATG": 3084, + "GAAACA": 3085, + "GAAACT": 3086, + "GAAACC": 3087, + "GAAACG": 3088, + "GAAAGA": 3089, + "GAAAGT": 3090, + "GAAAGC": 3091, + "GAAAGG": 3092, + "GAATAA": 3093, + "GAATAT": 3094, + "GAATAC": 3095, + "GAATAG": 3096, + "GAATTA": 3097, + "GAATTT": 3098, + "GAATTC": 3099, + "GAATTG": 3100, + "GAATCA": 3101, + "GAATCT": 3102, + "GAATCC": 3103, + "GAATCG": 3104, + "GAATGA": 3105, + "GAATGT": 3106, + "GAATGC": 3107, + "GAATGG": 3108, + "GAACAA": 3109, + "GAACAT": 3110, + "GAACAC": 3111, + "GAACAG": 3112, + "GAACTA": 3113, + "GAACTT": 3114, + "GAACTC": 3115, + "GAACTG": 3116, + "GAACCA": 3117, + "GAACCT": 3118, + "GAACCC": 3119, + "GAACCG": 3120, + "GAACGA": 3121, + "GAACGT": 3122, + "GAACGC": 3123, + "GAACGG": 3124, + "GAAGAA": 3125, + "GAAGAT": 3126, + "GAAGAC": 3127, + "GAAGAG": 3128, + "GAAGTA": 3129, + "GAAGTT": 3130, + "GAAGTC": 3131, + "GAAGTG": 3132, + "GAAGCA": 3133, + "GAAGCT": 3134, + "GAAGCC": 3135, + "GAAGCG": 3136, + "GAAGGA": 3137, + "GAAGGT": 3138, + "GAAGGC": 3139, + "GAAGGG": 3140, + "GATAAA": 3141, + "GATAAT": 3142, + "GATAAC": 3143, + "GATAAG": 3144, + "GATATA": 3145, + "GATATT": 3146, + "GATATC": 3147, + "GATATG": 3148, + "GATACA": 3149, + "GATACT": 3150, + "GATACC": 3151, + "GATACG": 3152, + "GATAGA": 3153, + "GATAGT": 3154, + "GATAGC": 3155, + "GATAGG": 3156, + "GATTAA": 3157, + "GATTAT": 3158, + "GATTAC": 3159, + "GATTAG": 3160, + "GATTTA": 3161, + "GATTTT": 3162, + "GATTTC": 3163, + "GATTTG": 3164, + "GATTCA": 3165, + "GATTCT": 3166, + "GATTCC": 3167, + "GATTCG": 3168, + "GATTGA": 3169, + "GATTGT": 3170, + "GATTGC": 3171, + "GATTGG": 3172, + "GATCAA": 3173, + "GATCAT": 3174, + "GATCAC": 3175, + "GATCAG": 3176, + "GATCTA": 3177, + "GATCTT": 3178, + "GATCTC": 3179, + "GATCTG": 3180, + "GATCCA": 3181, + "GATCCT": 3182, + "GATCCC": 3183, + "GATCCG": 3184, + "GATCGA": 3185, + "GATCGT": 3186, + "GATCGC": 3187, + "GATCGG": 3188, + "GATGAA": 3189, + "GATGAT": 3190, + "GATGAC": 3191, + "GATGAG": 3192, + "GATGTA": 3193, + "GATGTT": 3194, + "GATGTC": 3195, + "GATGTG": 3196, + "GATGCA": 3197, + "GATGCT": 3198, + "GATGCC": 3199, + "GATGCG": 3200, + "GATGGA": 3201, + "GATGGT": 3202, + "GATGGC": 3203, + "GATGGG": 3204, + "GACAAA": 3205, + "GACAAT": 3206, + "GACAAC": 3207, + "GACAAG": 3208, + "GACATA": 3209, + "GACATT": 3210, + "GACATC": 3211, + "GACATG": 3212, + "GACACA": 3213, + "GACACT": 3214, + "GACACC": 3215, + "GACACG": 3216, + "GACAGA": 3217, + "GACAGT": 3218, + "GACAGC": 3219, + "GACAGG": 3220, + "GACTAA": 3221, + "GACTAT": 3222, + "GACTAC": 3223, + "GACTAG": 3224, + "GACTTA": 3225, + "GACTTT": 3226, + "GACTTC": 3227, + "GACTTG": 3228, + "GACTCA": 3229, + "GACTCT": 3230, + "GACTCC": 3231, + "GACTCG": 3232, + "GACTGA": 3233, + "GACTGT": 3234, + "GACTGC": 3235, + "GACTGG": 3236, + "GACCAA": 3237, + "GACCAT": 3238, + "GACCAC": 3239, + "GACCAG": 3240, + "GACCTA": 3241, + "GACCTT": 3242, + "GACCTC": 3243, + "GACCTG": 3244, + "GACCCA": 3245, + "GACCCT": 3246, + "GACCCC": 3247, + "GACCCG": 3248, + "GACCGA": 3249, + "GACCGT": 3250, + "GACCGC": 3251, + "GACCGG": 3252, + "GACGAA": 3253, + "GACGAT": 3254, + "GACGAC": 3255, + "GACGAG": 3256, + "GACGTA": 3257, + "GACGTT": 3258, + "GACGTC": 3259, + "GACGTG": 3260, + "GACGCA": 3261, + "GACGCT": 3262, + "GACGCC": 3263, + "GACGCG": 3264, + "GACGGA": 3265, + "GACGGT": 3266, + "GACGGC": 3267, + "GACGGG": 3268, + "GAGAAA": 3269, + "GAGAAT": 3270, + "GAGAAC": 3271, + "GAGAAG": 3272, + "GAGATA": 3273, + "GAGATT": 3274, + "GAGATC": 3275, + "GAGATG": 3276, + "GAGACA": 3277, + "GAGACT": 3278, + "GAGACC": 3279, + "GAGACG": 3280, + "GAGAGA": 3281, + "GAGAGT": 3282, + "GAGAGC": 3283, + "GAGAGG": 3284, + "GAGTAA": 3285, + "GAGTAT": 3286, + "GAGTAC": 3287, + "GAGTAG": 3288, + "GAGTTA": 3289, + "GAGTTT": 3290, + "GAGTTC": 3291, + "GAGTTG": 3292, + "GAGTCA": 3293, + "GAGTCT": 3294, + "GAGTCC": 3295, + "GAGTCG": 3296, + "GAGTGA": 3297, + "GAGTGT": 3298, + "GAGTGC": 3299, + "GAGTGG": 3300, + "GAGCAA": 3301, + "GAGCAT": 3302, + "GAGCAC": 3303, + "GAGCAG": 3304, + "GAGCTA": 3305, + "GAGCTT": 3306, + "GAGCTC": 3307, + "GAGCTG": 3308, + "GAGCCA": 3309, + "GAGCCT": 3310, + "GAGCCC": 3311, + "GAGCCG": 3312, + "GAGCGA": 3313, + "GAGCGT": 3314, + "GAGCGC": 3315, + "GAGCGG": 3316, + "GAGGAA": 3317, + "GAGGAT": 3318, + "GAGGAC": 3319, + "GAGGAG": 3320, + "GAGGTA": 3321, + "GAGGTT": 3322, + "GAGGTC": 3323, + "GAGGTG": 3324, + "GAGGCA": 3325, + "GAGGCT": 3326, + "GAGGCC": 3327, + "GAGGCG": 3328, + "GAGGGA": 3329, + "GAGGGT": 3330, + "GAGGGC": 3331, + "GAGGGG": 3332, + "GTAAAA": 3333, + "GTAAAT": 3334, + "GTAAAC": 3335, + "GTAAAG": 3336, + "GTAATA": 3337, + "GTAATT": 3338, + "GTAATC": 3339, + "GTAATG": 3340, + "GTAACA": 3341, + "GTAACT": 3342, + "GTAACC": 3343, + "GTAACG": 3344, + "GTAAGA": 3345, + "GTAAGT": 3346, + "GTAAGC": 3347, + "GTAAGG": 3348, + "GTATAA": 3349, + "GTATAT": 3350, + "GTATAC": 3351, + "GTATAG": 3352, + "GTATTA": 3353, + "GTATTT": 3354, + "GTATTC": 3355, + "GTATTG": 3356, + "GTATCA": 3357, + "GTATCT": 3358, + "GTATCC": 3359, + "GTATCG": 3360, + "GTATGA": 3361, + "GTATGT": 3362, + "GTATGC": 3363, + "GTATGG": 3364, + "GTACAA": 3365, + "GTACAT": 3366, + "GTACAC": 3367, + "GTACAG": 3368, + "GTACTA": 3369, + "GTACTT": 3370, + "GTACTC": 3371, + "GTACTG": 3372, + "GTACCA": 3373, + "GTACCT": 3374, + "GTACCC": 3375, + "GTACCG": 3376, + "GTACGA": 3377, + "GTACGT": 3378, + "GTACGC": 3379, + "GTACGG": 3380, + "GTAGAA": 3381, + "GTAGAT": 3382, + "GTAGAC": 3383, + "GTAGAG": 3384, + "GTAGTA": 3385, + "GTAGTT": 3386, + "GTAGTC": 3387, + "GTAGTG": 3388, + "GTAGCA": 3389, + "GTAGCT": 3390, + "GTAGCC": 3391, + "GTAGCG": 3392, + "GTAGGA": 3393, + "GTAGGT": 3394, + "GTAGGC": 3395, + "GTAGGG": 3396, + "GTTAAA": 3397, + "GTTAAT": 3398, + "GTTAAC": 3399, + "GTTAAG": 3400, + "GTTATA": 3401, + "GTTATT": 3402, + "GTTATC": 3403, + "GTTATG": 3404, + "GTTACA": 3405, + "GTTACT": 3406, + "GTTACC": 3407, + "GTTACG": 3408, + "GTTAGA": 3409, + "GTTAGT": 3410, + "GTTAGC": 3411, + "GTTAGG": 3412, + "GTTTAA": 3413, + "GTTTAT": 3414, + "GTTTAC": 3415, + "GTTTAG": 3416, + "GTTTTA": 3417, + "GTTTTT": 3418, + "GTTTTC": 3419, + "GTTTTG": 3420, + "GTTTCA": 3421, + "GTTTCT": 3422, + "GTTTCC": 3423, + "GTTTCG": 3424, + "GTTTGA": 3425, + "GTTTGT": 3426, + "GTTTGC": 3427, + "GTTTGG": 3428, + "GTTCAA": 3429, + "GTTCAT": 3430, + "GTTCAC": 3431, + "GTTCAG": 3432, + "GTTCTA": 3433, + "GTTCTT": 3434, + "GTTCTC": 3435, + "GTTCTG": 3436, + "GTTCCA": 3437, + "GTTCCT": 3438, + "GTTCCC": 3439, + "GTTCCG": 3440, + "GTTCGA": 3441, + "GTTCGT": 3442, + "GTTCGC": 3443, + "GTTCGG": 3444, + "GTTGAA": 3445, + "GTTGAT": 3446, + "GTTGAC": 3447, + "GTTGAG": 3448, + "GTTGTA": 3449, + "GTTGTT": 3450, + "GTTGTC": 3451, + "GTTGTG": 3452, + "GTTGCA": 3453, + "GTTGCT": 3454, + "GTTGCC": 3455, + "GTTGCG": 3456, + "GTTGGA": 3457, + "GTTGGT": 3458, + "GTTGGC": 3459, + "GTTGGG": 3460, + "GTCAAA": 3461, + "GTCAAT": 3462, + "GTCAAC": 3463, + "GTCAAG": 3464, + "GTCATA": 3465, + "GTCATT": 3466, + "GTCATC": 3467, + "GTCATG": 3468, + "GTCACA": 3469, + "GTCACT": 3470, + "GTCACC": 3471, + "GTCACG": 3472, + "GTCAGA": 3473, + "GTCAGT": 3474, + "GTCAGC": 3475, + "GTCAGG": 3476, + "GTCTAA": 3477, + "GTCTAT": 3478, + "GTCTAC": 3479, + "GTCTAG": 3480, + "GTCTTA": 3481, + "GTCTTT": 3482, + "GTCTTC": 3483, + "GTCTTG": 3484, + "GTCTCA": 3485, + "GTCTCT": 3486, + "GTCTCC": 3487, + "GTCTCG": 3488, + "GTCTGA": 3489, + "GTCTGT": 3490, + "GTCTGC": 3491, + "GTCTGG": 3492, + "GTCCAA": 3493, + "GTCCAT": 3494, + "GTCCAC": 3495, + "GTCCAG": 3496, + "GTCCTA": 3497, + "GTCCTT": 3498, + "GTCCTC": 3499, + "GTCCTG": 3500, + "GTCCCA": 3501, + "GTCCCT": 3502, + "GTCCCC": 3503, + "GTCCCG": 3504, + "GTCCGA": 3505, + "GTCCGT": 3506, + "GTCCGC": 3507, + "GTCCGG": 3508, + "GTCGAA": 3509, + "GTCGAT": 3510, + "GTCGAC": 3511, + "GTCGAG": 3512, + "GTCGTA": 3513, + "GTCGTT": 3514, + "GTCGTC": 3515, + "GTCGTG": 3516, + "GTCGCA": 3517, + "GTCGCT": 3518, + "GTCGCC": 3519, + "GTCGCG": 3520, + "GTCGGA": 3521, + "GTCGGT": 3522, + "GTCGGC": 3523, + "GTCGGG": 3524, + "GTGAAA": 3525, + "GTGAAT": 3526, + "GTGAAC": 3527, + "GTGAAG": 3528, + "GTGATA": 3529, + "GTGATT": 3530, + "GTGATC": 3531, + "GTGATG": 3532, + "GTGACA": 3533, + "GTGACT": 3534, + "GTGACC": 3535, + "GTGACG": 3536, + "GTGAGA": 3537, + "GTGAGT": 3538, + "GTGAGC": 3539, + "GTGAGG": 3540, + "GTGTAA": 3541, + "GTGTAT": 3542, + "GTGTAC": 3543, + "GTGTAG": 3544, + "GTGTTA": 3545, + "GTGTTT": 3546, + "GTGTTC": 3547, + "GTGTTG": 3548, + "GTGTCA": 3549, + "GTGTCT": 3550, + "GTGTCC": 3551, + "GTGTCG": 3552, + "GTGTGA": 3553, + "GTGTGT": 3554, + "GTGTGC": 3555, + "GTGTGG": 3556, + "GTGCAA": 3557, + "GTGCAT": 3558, + "GTGCAC": 3559, + "GTGCAG": 3560, + "GTGCTA": 3561, + "GTGCTT": 3562, + "GTGCTC": 3563, + "GTGCTG": 3564, + "GTGCCA": 3565, + "GTGCCT": 3566, + "GTGCCC": 3567, + "GTGCCG": 3568, + "GTGCGA": 3569, + "GTGCGT": 3570, + "GTGCGC": 3571, + "GTGCGG": 3572, + "GTGGAA": 3573, + "GTGGAT": 3574, + "GTGGAC": 3575, + "GTGGAG": 3576, + "GTGGTA": 3577, + "GTGGTT": 3578, + "GTGGTC": 3579, + "GTGGTG": 3580, + "GTGGCA": 3581, + "GTGGCT": 3582, + "GTGGCC": 3583, + "GTGGCG": 3584, + "GTGGGA": 3585, + "GTGGGT": 3586, + "GTGGGC": 3587, + "GTGGGG": 3588, + "GCAAAA": 3589, + "GCAAAT": 3590, + "GCAAAC": 3591, + "GCAAAG": 3592, + "GCAATA": 3593, + "GCAATT": 3594, + "GCAATC": 3595, + "GCAATG": 3596, + "GCAACA": 3597, + "GCAACT": 3598, + "GCAACC": 3599, + "GCAACG": 3600, + "GCAAGA": 3601, + "GCAAGT": 3602, + "GCAAGC": 3603, + "GCAAGG": 3604, + "GCATAA": 3605, + "GCATAT": 3606, + "GCATAC": 3607, + "GCATAG": 3608, + "GCATTA": 3609, + "GCATTT": 3610, + "GCATTC": 3611, + "GCATTG": 3612, + "GCATCA": 3613, + "GCATCT": 3614, + "GCATCC": 3615, + "GCATCG": 3616, + "GCATGA": 3617, + "GCATGT": 3618, + "GCATGC": 3619, + "GCATGG": 3620, + "GCACAA": 3621, + "GCACAT": 3622, + "GCACAC": 3623, + "GCACAG": 3624, + "GCACTA": 3625, + "GCACTT": 3626, + "GCACTC": 3627, + "GCACTG": 3628, + "GCACCA": 3629, + "GCACCT": 3630, + "GCACCC": 3631, + "GCACCG": 3632, + "GCACGA": 3633, + "GCACGT": 3634, + "GCACGC": 3635, + "GCACGG": 3636, + "GCAGAA": 3637, + "GCAGAT": 3638, + "GCAGAC": 3639, + "GCAGAG": 3640, + "GCAGTA": 3641, + "GCAGTT": 3642, + "GCAGTC": 3643, + "GCAGTG": 3644, + "GCAGCA": 3645, + "GCAGCT": 3646, + "GCAGCC": 3647, + "GCAGCG": 3648, + "GCAGGA": 3649, + "GCAGGT": 3650, + "GCAGGC": 3651, + "GCAGGG": 3652, + "GCTAAA": 3653, + "GCTAAT": 3654, + "GCTAAC": 3655, + "GCTAAG": 3656, + "GCTATA": 3657, + "GCTATT": 3658, + "GCTATC": 3659, + "GCTATG": 3660, + "GCTACA": 3661, + "GCTACT": 3662, + "GCTACC": 3663, + "GCTACG": 3664, + "GCTAGA": 3665, + "GCTAGT": 3666, + "GCTAGC": 3667, + "GCTAGG": 3668, + "GCTTAA": 3669, + "GCTTAT": 3670, + "GCTTAC": 3671, + "GCTTAG": 3672, + "GCTTTA": 3673, + "GCTTTT": 3674, + "GCTTTC": 3675, + "GCTTTG": 3676, + "GCTTCA": 3677, + "GCTTCT": 3678, + "GCTTCC": 3679, + "GCTTCG": 3680, + "GCTTGA": 3681, + "GCTTGT": 3682, + "GCTTGC": 3683, + "GCTTGG": 3684, + "GCTCAA": 3685, + "GCTCAT": 3686, + "GCTCAC": 3687, + "GCTCAG": 3688, + "GCTCTA": 3689, + "GCTCTT": 3690, + "GCTCTC": 3691, + "GCTCTG": 3692, + "GCTCCA": 3693, + "GCTCCT": 3694, + "GCTCCC": 3695, + "GCTCCG": 3696, + "GCTCGA": 3697, + "GCTCGT": 3698, + "GCTCGC": 3699, + "GCTCGG": 3700, + "GCTGAA": 3701, + "GCTGAT": 3702, + "GCTGAC": 3703, + "GCTGAG": 3704, + "GCTGTA": 3705, + "GCTGTT": 3706, + "GCTGTC": 3707, + "GCTGTG": 3708, + "GCTGCA": 3709, + "GCTGCT": 3710, + "GCTGCC": 3711, + "GCTGCG": 3712, + "GCTGGA": 3713, + "GCTGGT": 3714, + "GCTGGC": 3715, + "GCTGGG": 3716, + "GCCAAA": 3717, + "GCCAAT": 3718, + "GCCAAC": 3719, + "GCCAAG": 3720, + "GCCATA": 3721, + "GCCATT": 3722, + "GCCATC": 3723, + "GCCATG": 3724, + "GCCACA": 3725, + "GCCACT": 3726, + "GCCACC": 3727, + "GCCACG": 3728, + "GCCAGA": 3729, + "GCCAGT": 3730, + "GCCAGC": 3731, + "GCCAGG": 3732, + "GCCTAA": 3733, + "GCCTAT": 3734, + "GCCTAC": 3735, + "GCCTAG": 3736, + "GCCTTA": 3737, + "GCCTTT": 3738, + "GCCTTC": 3739, + "GCCTTG": 3740, + "GCCTCA": 3741, + "GCCTCT": 3742, + "GCCTCC": 3743, + "GCCTCG": 3744, + "GCCTGA": 3745, + "GCCTGT": 3746, + "GCCTGC": 3747, + "GCCTGG": 3748, + "GCCCAA": 3749, + "GCCCAT": 3750, + "GCCCAC": 3751, + "GCCCAG": 3752, + "GCCCTA": 3753, + "GCCCTT": 3754, + "GCCCTC": 3755, + "GCCCTG": 3756, + "GCCCCA": 3757, + "GCCCCT": 3758, + "GCCCCC": 3759, + "GCCCCG": 3760, + "GCCCGA": 3761, + "GCCCGT": 3762, + "GCCCGC": 3763, + "GCCCGG": 3764, + "GCCGAA": 3765, + "GCCGAT": 3766, + "GCCGAC": 3767, + "GCCGAG": 3768, + "GCCGTA": 3769, + "GCCGTT": 3770, + "GCCGTC": 3771, + "GCCGTG": 3772, + "GCCGCA": 3773, + "GCCGCT": 3774, + "GCCGCC": 3775, + "GCCGCG": 3776, + "GCCGGA": 3777, + "GCCGGT": 3778, + "GCCGGC": 3779, + "GCCGGG": 3780, + "GCGAAA": 3781, + "GCGAAT": 3782, + "GCGAAC": 3783, + "GCGAAG": 3784, + "GCGATA": 3785, + "GCGATT": 3786, + "GCGATC": 3787, + "GCGATG": 3788, + "GCGACA": 3789, + "GCGACT": 3790, + "GCGACC": 3791, + "GCGACG": 3792, + "GCGAGA": 3793, + "GCGAGT": 3794, + "GCGAGC": 3795, + "GCGAGG": 3796, + "GCGTAA": 3797, + "GCGTAT": 3798, + "GCGTAC": 3799, + "GCGTAG": 3800, + "GCGTTA": 3801, + "GCGTTT": 3802, + "GCGTTC": 3803, + "GCGTTG": 3804, + "GCGTCA": 3805, + "GCGTCT": 3806, + "GCGTCC": 3807, + "GCGTCG": 3808, + "GCGTGA": 3809, + "GCGTGT": 3810, + "GCGTGC": 3811, + "GCGTGG": 3812, + "GCGCAA": 3813, + "GCGCAT": 3814, + "GCGCAC": 3815, + "GCGCAG": 3816, + "GCGCTA": 3817, + "GCGCTT": 3818, + "GCGCTC": 3819, + "GCGCTG": 3820, + "GCGCCA": 3821, + "GCGCCT": 3822, + "GCGCCC": 3823, + "GCGCCG": 3824, + "GCGCGA": 3825, + "GCGCGT": 3826, + "GCGCGC": 3827, + "GCGCGG": 3828, + "GCGGAA": 3829, + "GCGGAT": 3830, + "GCGGAC": 3831, + "GCGGAG": 3832, + "GCGGTA": 3833, + "GCGGTT": 3834, + "GCGGTC": 3835, + "GCGGTG": 3836, + "GCGGCA": 3837, + "GCGGCT": 3838, + "GCGGCC": 3839, + "GCGGCG": 3840, + "GCGGGA": 3841, + "GCGGGT": 3842, + "GCGGGC": 3843, + "GCGGGG": 3844, + "GGAAAA": 3845, + "GGAAAT": 3846, + "GGAAAC": 3847, + "GGAAAG": 3848, + "GGAATA": 3849, + "GGAATT": 3850, + "GGAATC": 3851, + "GGAATG": 3852, + "GGAACA": 3853, + "GGAACT": 3854, + "GGAACC": 3855, + "GGAACG": 3856, + "GGAAGA": 3857, + "GGAAGT": 3858, + "GGAAGC": 3859, + "GGAAGG": 3860, + "GGATAA": 3861, + "GGATAT": 3862, + "GGATAC": 3863, + "GGATAG": 3864, + "GGATTA": 3865, + "GGATTT": 3866, + "GGATTC": 3867, + "GGATTG": 3868, + "GGATCA": 3869, + "GGATCT": 3870, + "GGATCC": 3871, + "GGATCG": 3872, + "GGATGA": 3873, + "GGATGT": 3874, + "GGATGC": 3875, + "GGATGG": 3876, + "GGACAA": 3877, + "GGACAT": 3878, + "GGACAC": 3879, + "GGACAG": 3880, + "GGACTA": 3881, + "GGACTT": 3882, + "GGACTC": 3883, + "GGACTG": 3884, + "GGACCA": 3885, + "GGACCT": 3886, + "GGACCC": 3887, + "GGACCG": 3888, + "GGACGA": 3889, + "GGACGT": 3890, + "GGACGC": 3891, + "GGACGG": 3892, + "GGAGAA": 3893, + "GGAGAT": 3894, + "GGAGAC": 3895, + "GGAGAG": 3896, + "GGAGTA": 3897, + "GGAGTT": 3898, + "GGAGTC": 3899, + "GGAGTG": 3900, + "GGAGCA": 3901, + "GGAGCT": 3902, + "GGAGCC": 3903, + "GGAGCG": 3904, + "GGAGGA": 3905, + "GGAGGT": 3906, + "GGAGGC": 3907, + "GGAGGG": 3908, + "GGTAAA": 3909, + "GGTAAT": 3910, + "GGTAAC": 3911, + "GGTAAG": 3912, + "GGTATA": 3913, + "GGTATT": 3914, + "GGTATC": 3915, + "GGTATG": 3916, + "GGTACA": 3917, + "GGTACT": 3918, + "GGTACC": 3919, + "GGTACG": 3920, + "GGTAGA": 3921, + "GGTAGT": 3922, + "GGTAGC": 3923, + "GGTAGG": 3924, + "GGTTAA": 3925, + "GGTTAT": 3926, + "GGTTAC": 3927, + "GGTTAG": 3928, + "GGTTTA": 3929, + "GGTTTT": 3930, + "GGTTTC": 3931, + "GGTTTG": 3932, + "GGTTCA": 3933, + "GGTTCT": 3934, + "GGTTCC": 3935, + "GGTTCG": 3936, + "GGTTGA": 3937, + "GGTTGT": 3938, + "GGTTGC": 3939, + "GGTTGG": 3940, + "GGTCAA": 3941, + "GGTCAT": 3942, + "GGTCAC": 3943, + "GGTCAG": 3944, + "GGTCTA": 3945, + "GGTCTT": 3946, + "GGTCTC": 3947, + "GGTCTG": 3948, + "GGTCCA": 3949, + "GGTCCT": 3950, + "GGTCCC": 3951, + "GGTCCG": 3952, + "GGTCGA": 3953, + "GGTCGT": 3954, + "GGTCGC": 3955, + "GGTCGG": 3956, + "GGTGAA": 3957, + "GGTGAT": 3958, + "GGTGAC": 3959, + "GGTGAG": 3960, + "GGTGTA": 3961, + "GGTGTT": 3962, + "GGTGTC": 3963, + "GGTGTG": 3964, + "GGTGCA": 3965, + "GGTGCT": 3966, + "GGTGCC": 3967, + "GGTGCG": 3968, + "GGTGGA": 3969, + "GGTGGT": 3970, + "GGTGGC": 3971, + "GGTGGG": 3972, + "GGCAAA": 3973, + "GGCAAT": 3974, + "GGCAAC": 3975, + "GGCAAG": 3976, + "GGCATA": 3977, + "GGCATT": 3978, + "GGCATC": 3979, + "GGCATG": 3980, + "GGCACA": 3981, + "GGCACT": 3982, + "GGCACC": 3983, + "GGCACG": 3984, + "GGCAGA": 3985, + "GGCAGT": 3986, + "GGCAGC": 3987, + "GGCAGG": 3988, + "GGCTAA": 3989, + "GGCTAT": 3990, + "GGCTAC": 3991, + "GGCTAG": 3992, + "GGCTTA": 3993, + "GGCTTT": 3994, + "GGCTTC": 3995, + "GGCTTG": 3996, + "GGCTCA": 3997, + "GGCTCT": 3998, + "GGCTCC": 3999, + "GGCTCG": 4000, + "GGCTGA": 4001, + "GGCTGT": 4002, + "GGCTGC": 4003, + "GGCTGG": 4004, + "GGCCAA": 4005, + "GGCCAT": 4006, + "GGCCAC": 4007, + "GGCCAG": 4008, + "GGCCTA": 4009, + "GGCCTT": 4010, + "GGCCTC": 4011, + "GGCCTG": 4012, + "GGCCCA": 4013, + "GGCCCT": 4014, + "GGCCCC": 4015, + "GGCCCG": 4016, + "GGCCGA": 4017, + "GGCCGT": 4018, + "GGCCGC": 4019, + "GGCCGG": 4020, + "GGCGAA": 4021, + "GGCGAT": 4022, + "GGCGAC": 4023, + "GGCGAG": 4024, + "GGCGTA": 4025, + "GGCGTT": 4026, + "GGCGTC": 4027, + "GGCGTG": 4028, + "GGCGCA": 4029, + "GGCGCT": 4030, + "GGCGCC": 4031, + "GGCGCG": 4032, + "GGCGGA": 4033, + "GGCGGT": 4034, + "GGCGGC": 4035, + "GGCGGG": 4036, + "GGGAAA": 4037, + "GGGAAT": 4038, + "GGGAAC": 4039, + "GGGAAG": 4040, + "GGGATA": 4041, + "GGGATT": 4042, + "GGGATC": 4043, + "GGGATG": 4044, + "GGGACA": 4045, + "GGGACT": 4046, + "GGGACC": 4047, + "GGGACG": 4048, + "GGGAGA": 4049, + "GGGAGT": 4050, + "GGGAGC": 4051, + "GGGAGG": 4052, + "GGGTAA": 4053, + "GGGTAT": 4054, + "GGGTAC": 4055, + "GGGTAG": 4056, + "GGGTTA": 4057, + "GGGTTT": 4058, + "GGGTTC": 4059, + "GGGTTG": 4060, + "GGGTCA": 4061, + "GGGTCT": 4062, + "GGGTCC": 4063, + "GGGTCG": 4064, + "GGGTGA": 4065, + "GGGTGT": 4066, + "GGGTGC": 4067, + "GGGTGG": 4068, + "GGGCAA": 4069, + "GGGCAT": 4070, + "GGGCAC": 4071, + "GGGCAG": 4072, + "GGGCTA": 4073, + "GGGCTT": 4074, + "GGGCTC": 4075, + "GGGCTG": 4076, + "GGGCCA": 4077, + "GGGCCT": 4078, + "GGGCCC": 4079, + "GGGCCG": 4080, + "GGGCGA": 4081, + "GGGCGT": 4082, + "GGGCGC": 4083, + "GGGCGG": 4084, + "GGGGAA": 4085, + "GGGGAT": 4086, + "GGGGAC": 4087, + "GGGGAG": 4088, + "GGGGTA": 4089, + "GGGGTT": 4090, + "GGGGTC": 4091, + "GGGGTG": 4092, + "GGGGCA": 4093, + "GGGGCT": 4094, + "GGGGCC": 4095, + "GGGGCG": 4096, + "GGGGGA": 4097, + "GGGGGT": 4098, + "GGGGGC": 4099, + "GGGGGG": 4100 + } + } +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/tokenizer_config.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..604e2b1b29af1dcb3920d2828b6a90276e5e3d02 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/tokenizer_config.json @@ -0,0 +1,58 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "[UNK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "[CLS]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "[SEP]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "4": { + "content": "[MASK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "clean_up_tokenization_spaces": true, + "cls_token": "[CLS]", + "do_basic_tokenize": true, + "do_lower_case": false, + "mask_token": "[MASK]", + "max_len": 512, + "model_max_length": 512, + "never_split": null, + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "strip_accents": null, + "tokenize_chinese_chars": true, + "tokenizer_class": "BertTokenizer", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/trainer_state.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..74e87ed148e8f7d985d6818db1cbe66de84b07fd --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/trainer_state.json @@ -0,0 +1,1433 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 20.0, + "eval_steps": 500, + "global_step": 5000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.1, + "grad_norm": 0.9416134357452393, + "learning_rate": 5.319148936170213e-05, + "loss": 0.6912, + "step": 25 + }, + { + "epoch": 0.2, + "grad_norm": 0.9283172488212585, + "learning_rate": 0.00010638297872340425, + "loss": 0.6691, + "step": 50 + }, + { + "epoch": 0.3, + "grad_norm": 0.5870445370674133, + "learning_rate": 0.00015957446808510637, + "loss": 0.6164, + "step": 75 + }, + { + "epoch": 0.4, + "grad_norm": 0.33945509791374207, + "learning_rate": 0.0002127659574468085, + "loss": 0.5453, + "step": 100 + }, + { + "epoch": 0.5, + "grad_norm": 0.21052460372447968, + "learning_rate": 0.00026595744680851064, + "loss": 0.5025, + "step": 125 + }, + { + "epoch": 0.6, + "grad_norm": 0.25948837399482727, + "learning_rate": 0.00031914893617021275, + "loss": 0.4881, + "step": 150 + }, + { + "epoch": 0.7, + "grad_norm": 0.6165992021560669, + "learning_rate": 0.0003723404255319149, + "loss": 0.4852, + "step": 175 + }, + { + "epoch": 0.8, + "grad_norm": 0.33296748995780945, + "learning_rate": 0.0003992081821181128, + "loss": 0.4867, + "step": 200 + }, + { + "epoch": 0.9, + "grad_norm": 0.40745800733566284, + "learning_rate": 0.0003975585615308479, + "loss": 0.4767, + "step": 225 + }, + { + "epoch": 1.0, + "grad_norm": 0.8070425391197205, + "learning_rate": 0.00039590894094358297, + "loss": 0.4784, + "step": 250 + }, + { + "epoch": 1.1, + "grad_norm": 1.2249579429626465, + "learning_rate": 0.0003942593203563181, + "loss": 0.4543, + "step": 275 + }, + { + "epoch": 1.2, + "grad_norm": 1.0143898725509644, + "learning_rate": 0.0003926096997690532, + "loss": 0.4319, + "step": 300 + }, + { + "epoch": 1.3, + "grad_norm": 2.3284952640533447, + "learning_rate": 0.00039096007918178817, + "loss": 0.4455, + "step": 325 + }, + { + "epoch": 1.4, + "grad_norm": 1.189352035522461, + "learning_rate": 0.0003893104585945233, + "loss": 0.4246, + "step": 350 + }, + { + "epoch": 1.5, + "grad_norm": 0.947853684425354, + "learning_rate": 0.0003876608380072583, + "loss": 0.4326, + "step": 375 + }, + { + "epoch": 1.6, + "grad_norm": 0.7130059003829956, + "learning_rate": 0.00038601121741999343, + "loss": 0.4034, + "step": 400 + }, + { + "epoch": 1.7, + "grad_norm": 3.517078161239624, + "learning_rate": 0.0003843615968327285, + "loss": 0.4114, + "step": 425 + }, + { + "epoch": 1.8, + "grad_norm": 0.3982729911804199, + "learning_rate": 0.0003827119762454636, + "loss": 0.4345, + "step": 450 + }, + { + "epoch": 1.9, + "grad_norm": 0.6208440065383911, + "learning_rate": 0.00038106235565819863, + "loss": 0.4135, + "step": 475 + }, + { + "epoch": 2.0, + "grad_norm": 0.9085566401481628, + "learning_rate": 0.0003794127350709337, + "loss": 0.3983, + "step": 500 + }, + { + "epoch": 2.1, + "grad_norm": 1.2225189208984375, + "learning_rate": 0.0003777631144836688, + "loss": 0.3915, + "step": 525 + }, + { + "epoch": 2.2, + "grad_norm": 1.1379530429840088, + "learning_rate": 0.00037611349389640383, + "loss": 0.4335, + "step": 550 + }, + { + "epoch": 2.3, + "grad_norm": 0.8535892367362976, + "learning_rate": 0.00037446387330913894, + "loss": 0.4063, + "step": 575 + }, + { + "epoch": 2.4, + "grad_norm": 1.3587419986724854, + "learning_rate": 0.000372814252721874, + "loss": 0.4065, + "step": 600 + }, + { + "epoch": 2.5, + "grad_norm": 0.9311383962631226, + "learning_rate": 0.00037116463213460903, + "loss": 0.4007, + "step": 625 + }, + { + "epoch": 2.6, + "grad_norm": 0.9912465810775757, + "learning_rate": 0.00036951501154734414, + "loss": 0.3992, + "step": 650 + }, + { + "epoch": 2.7, + "grad_norm": 1.2762119770050049, + "learning_rate": 0.0003678653909600792, + "loss": 0.3827, + "step": 675 + }, + { + "epoch": 2.8, + "grad_norm": 2.3506247997283936, + "learning_rate": 0.0003662157703728143, + "loss": 0.3734, + "step": 700 + }, + { + "epoch": 2.9, + "grad_norm": 1.5375425815582275, + "learning_rate": 0.00036456614978554934, + "loss": 0.4005, + "step": 725 + }, + { + "epoch": 3.0, + "grad_norm": 1.3143396377563477, + "learning_rate": 0.00036291652919828444, + "loss": 0.3548, + "step": 750 + }, + { + "epoch": 3.1, + "grad_norm": 0.662200391292572, + "learning_rate": 0.00036126690861101944, + "loss": 0.4129, + "step": 775 + }, + { + "epoch": 3.2, + "grad_norm": 1.3569878339767456, + "learning_rate": 0.00035961728802375454, + "loss": 0.38, + "step": 800 + }, + { + "epoch": 3.3, + "grad_norm": 0.9888588190078735, + "learning_rate": 0.0003579676674364896, + "loss": 0.3778, + "step": 825 + }, + { + "epoch": 3.4, + "grad_norm": 0.679969310760498, + "learning_rate": 0.0003563180468492247, + "loss": 0.3934, + "step": 850 + }, + { + "epoch": 3.5, + "grad_norm": 1.608539342880249, + "learning_rate": 0.0003546684262619598, + "loss": 0.418, + "step": 875 + }, + { + "epoch": 3.6, + "grad_norm": 1.7037910223007202, + "learning_rate": 0.00035301880567469485, + "loss": 0.3934, + "step": 900 + }, + { + "epoch": 3.7, + "grad_norm": 1.4417132139205933, + "learning_rate": 0.0003513691850874299, + "loss": 0.3947, + "step": 925 + }, + { + "epoch": 3.8, + "grad_norm": 1.1482582092285156, + "learning_rate": 0.00034971956450016495, + "loss": 0.3621, + "step": 950 + }, + { + "epoch": 3.9, + "grad_norm": 2.046616792678833, + "learning_rate": 0.00034806994391290005, + "loss": 0.3774, + "step": 975 + }, + { + "epoch": 4.0, + "grad_norm": 1.0832113027572632, + "learning_rate": 0.0003464203233256351, + "loss": 0.3542, + "step": 1000 + }, + { + "epoch": 4.1, + "grad_norm": 0.8189190626144409, + "learning_rate": 0.0003447707027383702, + "loss": 0.381, + "step": 1025 + }, + { + "epoch": 4.2, + "grad_norm": 1.6974774599075317, + "learning_rate": 0.00034312108215110525, + "loss": 0.3775, + "step": 1050 + }, + { + "epoch": 4.3, + "grad_norm": 1.2468841075897217, + "learning_rate": 0.0003414714615638403, + "loss": 0.3527, + "step": 1075 + }, + { + "epoch": 4.4, + "grad_norm": 0.9022793173789978, + "learning_rate": 0.0003398218409765754, + "loss": 0.3637, + "step": 1100 + }, + { + "epoch": 4.5, + "grad_norm": 1.121774673461914, + "learning_rate": 0.00033817222038931045, + "loss": 0.3935, + "step": 1125 + }, + { + "epoch": 4.6, + "grad_norm": 0.810937762260437, + "learning_rate": 0.00033652259980204556, + "loss": 0.4103, + "step": 1150 + }, + { + "epoch": 4.7, + "grad_norm": 0.7376565337181091, + "learning_rate": 0.0003348729792147806, + "loss": 0.362, + "step": 1175 + }, + { + "epoch": 4.8, + "grad_norm": 1.7271156311035156, + "learning_rate": 0.0003332233586275157, + "loss": 0.3597, + "step": 1200 + }, + { + "epoch": 4.9, + "grad_norm": 1.2273073196411133, + "learning_rate": 0.00033157373804025076, + "loss": 0.3853, + "step": 1225 + }, + { + "epoch": 5.0, + "grad_norm": 0.9998250603675842, + "learning_rate": 0.0003299241174529858, + "loss": 0.3461, + "step": 1250 + }, + { + "epoch": 5.1, + "grad_norm": 0.9112564325332642, + "learning_rate": 0.0003282744968657209, + "loss": 0.3659, + "step": 1275 + }, + { + "epoch": 5.2, + "grad_norm": 3.246051073074341, + "learning_rate": 0.00032662487627845596, + "loss": 0.3488, + "step": 1300 + }, + { + "epoch": 5.3, + "grad_norm": 1.0283201932907104, + "learning_rate": 0.00032497525569119106, + "loss": 0.3528, + "step": 1325 + }, + { + "epoch": 5.4, + "grad_norm": 1.112660527229309, + "learning_rate": 0.0003233256351039261, + "loss": 0.3723, + "step": 1350 + }, + { + "epoch": 5.5, + "grad_norm": 0.9069312810897827, + "learning_rate": 0.00032167601451666116, + "loss": 0.3416, + "step": 1375 + }, + { + "epoch": 5.6, + "grad_norm": 0.710412323474884, + "learning_rate": 0.00032002639392939627, + "loss": 0.3607, + "step": 1400 + }, + { + "epoch": 5.7, + "grad_norm": 1.185612678527832, + "learning_rate": 0.0003183767733421313, + "loss": 0.3777, + "step": 1425 + }, + { + "epoch": 5.8, + "grad_norm": 1.340174913406372, + "learning_rate": 0.0003167271527548664, + "loss": 0.3563, + "step": 1450 + }, + { + "epoch": 5.9, + "grad_norm": 0.9387012124061584, + "learning_rate": 0.00031507753216760147, + "loss": 0.3647, + "step": 1475 + }, + { + "epoch": 6.0, + "grad_norm": 1.1005347967147827, + "learning_rate": 0.00031342791158033657, + "loss": 0.3484, + "step": 1500 + }, + { + "epoch": 6.1, + "grad_norm": 2.64058780670166, + "learning_rate": 0.00031177829099307157, + "loss": 0.3542, + "step": 1525 + }, + { + "epoch": 6.2, + "grad_norm": 1.3007783889770508, + "learning_rate": 0.00031012867040580667, + "loss": 0.3379, + "step": 1550 + }, + { + "epoch": 6.3, + "grad_norm": 3.7190418243408203, + "learning_rate": 0.0003084790498185417, + "loss": 0.3514, + "step": 1575 + }, + { + "epoch": 6.4, + "grad_norm": 1.7263462543487549, + "learning_rate": 0.0003068294292312768, + "loss": 0.3434, + "step": 1600 + }, + { + "epoch": 6.5, + "grad_norm": 3.3360602855682373, + "learning_rate": 0.0003051798086440119, + "loss": 0.3507, + "step": 1625 + }, + { + "epoch": 6.6, + "grad_norm": 1.9620813131332397, + "learning_rate": 0.000303530188056747, + "loss": 0.3455, + "step": 1650 + }, + { + "epoch": 6.7, + "grad_norm": 1.3080660104751587, + "learning_rate": 0.000301880567469482, + "loss": 0.3678, + "step": 1675 + }, + { + "epoch": 6.8, + "grad_norm": 0.6183618307113647, + "learning_rate": 0.0003002309468822171, + "loss": 0.3704, + "step": 1700 + }, + { + "epoch": 6.9, + "grad_norm": 1.0411962270736694, + "learning_rate": 0.0002985813262949522, + "loss": 0.3505, + "step": 1725 + }, + { + "epoch": 7.0, + "grad_norm": 2.040569543838501, + "learning_rate": 0.00029693170570768723, + "loss": 0.3666, + "step": 1750 + }, + { + "epoch": 7.1, + "grad_norm": 0.713647723197937, + "learning_rate": 0.00029528208512042233, + "loss": 0.3529, + "step": 1775 + }, + { + "epoch": 7.2, + "grad_norm": 3.0373058319091797, + "learning_rate": 0.0002936324645331574, + "loss": 0.361, + "step": 1800 + }, + { + "epoch": 7.3, + "grad_norm": 1.819886326789856, + "learning_rate": 0.00029198284394589243, + "loss": 0.3482, + "step": 1825 + }, + { + "epoch": 7.4, + "grad_norm": 0.7930757403373718, + "learning_rate": 0.00029033322335862753, + "loss": 0.3695, + "step": 1850 + }, + { + "epoch": 7.5, + "grad_norm": 0.7889612913131714, + "learning_rate": 0.0002886836027713626, + "loss": 0.3527, + "step": 1875 + }, + { + "epoch": 7.6, + "grad_norm": 2.4007863998413086, + "learning_rate": 0.0002870339821840977, + "loss": 0.3275, + "step": 1900 + }, + { + "epoch": 7.7, + "grad_norm": 2.3483331203460693, + "learning_rate": 0.00028538436159683273, + "loss": 0.3571, + "step": 1925 + }, + { + "epoch": 7.8, + "grad_norm": 0.8911663293838501, + "learning_rate": 0.00028373474100956784, + "loss": 0.3079, + "step": 1950 + }, + { + "epoch": 7.9, + "grad_norm": 1.0521246194839478, + "learning_rate": 0.0002820851204223029, + "loss": 0.3578, + "step": 1975 + }, + { + "epoch": 8.0, + "grad_norm": 1.3923832178115845, + "learning_rate": 0.00028043549983503794, + "loss": 0.3588, + "step": 2000 + }, + { + "epoch": 8.1, + "grad_norm": 1.4007149934768677, + "learning_rate": 0.00027878587924777304, + "loss": 0.3458, + "step": 2025 + }, + { + "epoch": 8.2, + "grad_norm": 1.103198766708374, + "learning_rate": 0.0002771362586605081, + "loss": 0.3736, + "step": 2050 + }, + { + "epoch": 8.3, + "grad_norm": 2.936150550842285, + "learning_rate": 0.0002754866380732432, + "loss": 0.3223, + "step": 2075 + }, + { + "epoch": 8.4, + "grad_norm": 1.907762885093689, + "learning_rate": 0.00027383701748597824, + "loss": 0.3348, + "step": 2100 + }, + { + "epoch": 8.5, + "grad_norm": 1.4422886371612549, + "learning_rate": 0.0002721873968987133, + "loss": 0.3222, + "step": 2125 + }, + { + "epoch": 8.6, + "grad_norm": 1.8203483819961548, + "learning_rate": 0.00027053777631144834, + "loss": 0.3968, + "step": 2150 + }, + { + "epoch": 8.7, + "grad_norm": 0.8046119213104248, + "learning_rate": 0.00026888815572418344, + "loss": 0.3422, + "step": 2175 + }, + { + "epoch": 8.8, + "grad_norm": 1.2612605094909668, + "learning_rate": 0.00026723853513691855, + "loss": 0.3387, + "step": 2200 + }, + { + "epoch": 8.9, + "grad_norm": 1.6534866094589233, + "learning_rate": 0.0002655889145496536, + "loss": 0.3548, + "step": 2225 + }, + { + "epoch": 9.0, + "grad_norm": 1.889664649963379, + "learning_rate": 0.0002639392939623887, + "loss": 0.3367, + "step": 2250 + }, + { + "epoch": 9.1, + "grad_norm": 2.6691551208496094, + "learning_rate": 0.0002622896733751237, + "loss": 0.3628, + "step": 2275 + }, + { + "epoch": 9.2, + "grad_norm": 2.0130255222320557, + "learning_rate": 0.0002606400527878588, + "loss": 0.3366, + "step": 2300 + }, + { + "epoch": 9.3, + "grad_norm": 1.404814600944519, + "learning_rate": 0.00025899043220059385, + "loss": 0.3327, + "step": 2325 + }, + { + "epoch": 9.4, + "grad_norm": 0.9150362610816956, + "learning_rate": 0.00025734081161332895, + "loss": 0.308, + "step": 2350 + }, + { + "epoch": 9.5, + "grad_norm": 1.099320411682129, + "learning_rate": 0.000255691191026064, + "loss": 0.3499, + "step": 2375 + }, + { + "epoch": 9.6, + "grad_norm": 1.502451777458191, + "learning_rate": 0.0002540415704387991, + "loss": 0.3728, + "step": 2400 + }, + { + "epoch": 9.7, + "grad_norm": 1.1115409135818481, + "learning_rate": 0.00025239194985153415, + "loss": 0.348, + "step": 2425 + }, + { + "epoch": 9.8, + "grad_norm": 0.8831422328948975, + "learning_rate": 0.0002507423292642692, + "loss": 0.3392, + "step": 2450 + }, + { + "epoch": 9.9, + "grad_norm": 2.104572057723999, + "learning_rate": 0.0002490927086770043, + "loss": 0.3511, + "step": 2475 + }, + { + "epoch": 10.0, + "grad_norm": 2.4081919193267822, + "learning_rate": 0.00024744308808973936, + "loss": 0.3284, + "step": 2500 + }, + { + "epoch": 10.1, + "grad_norm": 1.9861253499984741, + "learning_rate": 0.00024579346750247446, + "loss": 0.3621, + "step": 2525 + }, + { + "epoch": 10.2, + "grad_norm": 1.9042344093322754, + "learning_rate": 0.0002441438469152095, + "loss": 0.3341, + "step": 2550 + }, + { + "epoch": 10.3, + "grad_norm": 0.5676223635673523, + "learning_rate": 0.00024249422632794456, + "loss": 0.3203, + "step": 2575 + }, + { + "epoch": 10.4, + "grad_norm": 1.5105270147323608, + "learning_rate": 0.00024084460574067963, + "loss": 0.3711, + "step": 2600 + }, + { + "epoch": 10.5, + "grad_norm": 0.9969163537025452, + "learning_rate": 0.0002391949851534147, + "loss": 0.3296, + "step": 2625 + }, + { + "epoch": 10.6, + "grad_norm": 1.297096848487854, + "learning_rate": 0.0002375453645661498, + "loss": 0.349, + "step": 2650 + }, + { + "epoch": 10.7, + "grad_norm": 1.0513092279434204, + "learning_rate": 0.00023589574397888486, + "loss": 0.3308, + "step": 2675 + }, + { + "epoch": 10.8, + "grad_norm": 1.5543969869613647, + "learning_rate": 0.00023424612339161997, + "loss": 0.3289, + "step": 2700 + }, + { + "epoch": 10.9, + "grad_norm": 0.9587257504463196, + "learning_rate": 0.000232596502804355, + "loss": 0.3128, + "step": 2725 + }, + { + "epoch": 11.0, + "grad_norm": 2.1693837642669678, + "learning_rate": 0.00023094688221709007, + "loss": 0.349, + "step": 2750 + }, + { + "epoch": 11.1, + "grad_norm": 1.5624831914901733, + "learning_rate": 0.00022929726162982514, + "loss": 0.3391, + "step": 2775 + }, + { + "epoch": 11.2, + "grad_norm": 1.4600939750671387, + "learning_rate": 0.00022764764104256022, + "loss": 0.3371, + "step": 2800 + }, + { + "epoch": 11.3, + "grad_norm": 1.6139200925827026, + "learning_rate": 0.0002259980204552953, + "loss": 0.3447, + "step": 2825 + }, + { + "epoch": 11.4, + "grad_norm": 2.2703235149383545, + "learning_rate": 0.00022434839986803037, + "loss": 0.3157, + "step": 2850 + }, + { + "epoch": 11.5, + "grad_norm": 2.0282037258148193, + "learning_rate": 0.00022269877928076542, + "loss": 0.3339, + "step": 2875 + }, + { + "epoch": 11.6, + "grad_norm": 1.389129638671875, + "learning_rate": 0.0002210491586935005, + "loss": 0.3195, + "step": 2900 + }, + { + "epoch": 11.7, + "grad_norm": 1.0041213035583496, + "learning_rate": 0.00021939953810623557, + "loss": 0.3382, + "step": 2925 + }, + { + "epoch": 11.8, + "grad_norm": 1.7116957902908325, + "learning_rate": 0.00021774991751897065, + "loss": 0.3261, + "step": 2950 + }, + { + "epoch": 11.9, + "grad_norm": 1.080194354057312, + "learning_rate": 0.00021610029693170573, + "loss": 0.3433, + "step": 2975 + }, + { + "epoch": 12.0, + "grad_norm": 0.6561412811279297, + "learning_rate": 0.0002144506763444408, + "loss": 0.3465, + "step": 3000 + }, + { + "epoch": 12.1, + "grad_norm": 1.5294175148010254, + "learning_rate": 0.00021280105575717582, + "loss": 0.3195, + "step": 3025 + }, + { + "epoch": 12.2, + "grad_norm": 0.6313191652297974, + "learning_rate": 0.00021115143516991093, + "loss": 0.3214, + "step": 3050 + }, + { + "epoch": 12.3, + "grad_norm": 1.4611085653305054, + "learning_rate": 0.000209501814582646, + "loss": 0.3091, + "step": 3075 + }, + { + "epoch": 12.4, + "grad_norm": 1.6590704917907715, + "learning_rate": 0.00020785219399538108, + "loss": 0.3633, + "step": 3100 + }, + { + "epoch": 12.5, + "grad_norm": 0.982976496219635, + "learning_rate": 0.00020620257340811616, + "loss": 0.3688, + "step": 3125 + }, + { + "epoch": 12.6, + "grad_norm": 0.9717169404029846, + "learning_rate": 0.00020455295282085123, + "loss": 0.3532, + "step": 3150 + }, + { + "epoch": 12.7, + "grad_norm": 1.5903713703155518, + "learning_rate": 0.00020290333223358626, + "loss": 0.2906, + "step": 3175 + }, + { + "epoch": 12.8, + "grad_norm": 1.7043166160583496, + "learning_rate": 0.00020125371164632133, + "loss": 0.322, + "step": 3200 + }, + { + "epoch": 12.9, + "grad_norm": 0.7691648006439209, + "learning_rate": 0.0001996040910590564, + "loss": 0.3477, + "step": 3225 + }, + { + "epoch": 13.0, + "grad_norm": 1.1726899147033691, + "learning_rate": 0.00019795447047179148, + "loss": 0.3551, + "step": 3250 + }, + { + "epoch": 13.1, + "grad_norm": 1.0323086977005005, + "learning_rate": 0.0001963048498845266, + "loss": 0.3248, + "step": 3275 + }, + { + "epoch": 13.2, + "grad_norm": 2.920222520828247, + "learning_rate": 0.00019465522929726164, + "loss": 0.3298, + "step": 3300 + }, + { + "epoch": 13.3, + "grad_norm": 1.101644515991211, + "learning_rate": 0.00019300560870999671, + "loss": 0.3011, + "step": 3325 + }, + { + "epoch": 13.4, + "grad_norm": 1.111193299293518, + "learning_rate": 0.0001913559881227318, + "loss": 0.325, + "step": 3350 + }, + { + "epoch": 13.5, + "grad_norm": 1.5074782371520996, + "learning_rate": 0.00018970636753546684, + "loss": 0.3435, + "step": 3375 + }, + { + "epoch": 13.6, + "grad_norm": 1.0705279111862183, + "learning_rate": 0.00018805674694820192, + "loss": 0.3378, + "step": 3400 + }, + { + "epoch": 13.7, + "grad_norm": 1.5590386390686035, + "learning_rate": 0.000186407126360937, + "loss": 0.3349, + "step": 3425 + }, + { + "epoch": 13.8, + "grad_norm": 2.2993671894073486, + "learning_rate": 0.00018475750577367207, + "loss": 0.3479, + "step": 3450 + }, + { + "epoch": 13.9, + "grad_norm": 1.8973578214645386, + "learning_rate": 0.00018310788518640715, + "loss": 0.3358, + "step": 3475 + }, + { + "epoch": 14.0, + "grad_norm": 1.5887267589569092, + "learning_rate": 0.00018145826459914222, + "loss": 0.3095, + "step": 3500 + }, + { + "epoch": 14.1, + "grad_norm": 2.1970202922821045, + "learning_rate": 0.00017980864401187727, + "loss": 0.3435, + "step": 3525 + }, + { + "epoch": 14.2, + "grad_norm": 0.9183377027511597, + "learning_rate": 0.00017815902342461235, + "loss": 0.3454, + "step": 3550 + }, + { + "epoch": 14.3, + "grad_norm": 1.253682017326355, + "learning_rate": 0.00017650940283734742, + "loss": 0.3128, + "step": 3575 + }, + { + "epoch": 14.4, + "grad_norm": 2.4595584869384766, + "learning_rate": 0.00017485978225008247, + "loss": 0.3181, + "step": 3600 + }, + { + "epoch": 14.5, + "grad_norm": 2.0421340465545654, + "learning_rate": 0.00017321016166281755, + "loss": 0.3555, + "step": 3625 + }, + { + "epoch": 14.6, + "grad_norm": 2.0127763748168945, + "learning_rate": 0.00017156054107555263, + "loss": 0.2952, + "step": 3650 + }, + { + "epoch": 14.7, + "grad_norm": 0.9481540322303772, + "learning_rate": 0.0001699109204882877, + "loss": 0.3154, + "step": 3675 + }, + { + "epoch": 14.8, + "grad_norm": 1.5080980062484741, + "learning_rate": 0.00016826129990102278, + "loss": 0.336, + "step": 3700 + }, + { + "epoch": 14.9, + "grad_norm": 1.504712700843811, + "learning_rate": 0.00016661167931375785, + "loss": 0.3511, + "step": 3725 + }, + { + "epoch": 15.0, + "grad_norm": 1.4532283544540405, + "learning_rate": 0.0001649620587264929, + "loss": 0.3216, + "step": 3750 + }, + { + "epoch": 15.1, + "grad_norm": 1.215970754623413, + "learning_rate": 0.00016331243813922798, + "loss": 0.3597, + "step": 3775 + }, + { + "epoch": 15.2, + "grad_norm": 2.054934024810791, + "learning_rate": 0.00016166281755196306, + "loss": 0.3095, + "step": 3800 + }, + { + "epoch": 15.3, + "grad_norm": 1.6162620782852173, + "learning_rate": 0.00016001319696469813, + "loss": 0.3172, + "step": 3825 + }, + { + "epoch": 15.4, + "grad_norm": 2.974353790283203, + "learning_rate": 0.0001583635763774332, + "loss": 0.3186, + "step": 3850 + }, + { + "epoch": 15.5, + "grad_norm": 1.1828879117965698, + "learning_rate": 0.00015671395579016829, + "loss": 0.3141, + "step": 3875 + }, + { + "epoch": 15.6, + "grad_norm": 2.170670509338379, + "learning_rate": 0.00015506433520290334, + "loss": 0.3054, + "step": 3900 + }, + { + "epoch": 15.7, + "grad_norm": 2.5053343772888184, + "learning_rate": 0.0001534147146156384, + "loss": 0.3348, + "step": 3925 + }, + { + "epoch": 15.8, + "grad_norm": 2.761240005493164, + "learning_rate": 0.0001517650940283735, + "loss": 0.3513, + "step": 3950 + }, + { + "epoch": 15.9, + "grad_norm": 2.6656720638275146, + "learning_rate": 0.00015011547344110854, + "loss": 0.3477, + "step": 3975 + }, + { + "epoch": 16.0, + "grad_norm": 1.2667946815490723, + "learning_rate": 0.00014846585285384361, + "loss": 0.3107, + "step": 4000 + }, + { + "epoch": 16.1, + "grad_norm": 1.4039124250411987, + "learning_rate": 0.0001468162322665787, + "loss": 0.3242, + "step": 4025 + }, + { + "epoch": 16.2, + "grad_norm": 1.3851566314697266, + "learning_rate": 0.00014516661167931377, + "loss": 0.3144, + "step": 4050 + }, + { + "epoch": 16.3, + "grad_norm": 1.2302920818328857, + "learning_rate": 0.00014351699109204884, + "loss": 0.3376, + "step": 4075 + }, + { + "epoch": 16.4, + "grad_norm": 1.8713226318359375, + "learning_rate": 0.00014186737050478392, + "loss": 0.3266, + "step": 4100 + }, + { + "epoch": 16.5, + "grad_norm": 1.3579822778701782, + "learning_rate": 0.00014021774991751897, + "loss": 0.33, + "step": 4125 + }, + { + "epoch": 16.6, + "grad_norm": 1.4338854551315308, + "learning_rate": 0.00013856812933025404, + "loss": 0.3218, + "step": 4150 + }, + { + "epoch": 16.7, + "grad_norm": 1.9794092178344727, + "learning_rate": 0.00013691850874298912, + "loss": 0.3375, + "step": 4175 + }, + { + "epoch": 16.8, + "grad_norm": 1.4431006908416748, + "learning_rate": 0.00013526888815572417, + "loss": 0.3177, + "step": 4200 + }, + { + "epoch": 16.9, + "grad_norm": 1.1928157806396484, + "learning_rate": 0.00013361926756845927, + "loss": 0.3227, + "step": 4225 + }, + { + "epoch": 17.0, + "grad_norm": 2.0860953330993652, + "learning_rate": 0.00013196964698119435, + "loss": 0.3345, + "step": 4250 + }, + { + "epoch": 17.1, + "grad_norm": 1.0021557807922363, + "learning_rate": 0.0001303200263939294, + "loss": 0.3299, + "step": 4275 + }, + { + "epoch": 17.2, + "grad_norm": 1.6104000806808472, + "learning_rate": 0.00012867040580666448, + "loss": 0.3066, + "step": 4300 + }, + { + "epoch": 17.3, + "grad_norm": 2.3414576053619385, + "learning_rate": 0.00012702078521939955, + "loss": 0.3474, + "step": 4325 + }, + { + "epoch": 17.4, + "grad_norm": 1.8101857900619507, + "learning_rate": 0.0001253711646321346, + "loss": 0.3022, + "step": 4350 + }, + { + "epoch": 17.5, + "grad_norm": 1.1479285955429077, + "learning_rate": 0.00012372154404486968, + "loss": 0.3133, + "step": 4375 + }, + { + "epoch": 17.6, + "grad_norm": 2.7935454845428467, + "learning_rate": 0.00012207192345760475, + "loss": 0.3027, + "step": 4400 + }, + { + "epoch": 17.7, + "grad_norm": 2.0491445064544678, + "learning_rate": 0.00012042230287033982, + "loss": 0.3451, + "step": 4425 + }, + { + "epoch": 17.8, + "grad_norm": 0.9633632302284241, + "learning_rate": 0.0001187726822830749, + "loss": 0.3171, + "step": 4450 + }, + { + "epoch": 17.9, + "grad_norm": 1.5458674430847168, + "learning_rate": 0.00011712306169580998, + "loss": 0.3011, + "step": 4475 + }, + { + "epoch": 18.0, + "grad_norm": 1.4501872062683105, + "learning_rate": 0.00011547344110854503, + "loss": 0.3185, + "step": 4500 + }, + { + "epoch": 18.1, + "grad_norm": 2.6352083683013916, + "learning_rate": 0.00011382382052128011, + "loss": 0.3191, + "step": 4525 + }, + { + "epoch": 18.2, + "grad_norm": 2.366875410079956, + "learning_rate": 0.00011217419993401519, + "loss": 0.311, + "step": 4550 + }, + { + "epoch": 18.3, + "grad_norm": 1.9184437990188599, + "learning_rate": 0.00011052457934675025, + "loss": 0.2812, + "step": 4575 + }, + { + "epoch": 18.4, + "grad_norm": 1.466262936592102, + "learning_rate": 0.00010887495875948532, + "loss": 0.3075, + "step": 4600 + }, + { + "epoch": 18.5, + "grad_norm": 1.94712233543396, + "learning_rate": 0.0001072253381722204, + "loss": 0.3163, + "step": 4625 + }, + { + "epoch": 18.6, + "grad_norm": 2.5286598205566406, + "learning_rate": 0.00010557571758495546, + "loss": 0.3346, + "step": 4650 + }, + { + "epoch": 18.7, + "grad_norm": 1.604040265083313, + "learning_rate": 0.00010392609699769054, + "loss": 0.3138, + "step": 4675 + }, + { + "epoch": 18.8, + "grad_norm": 1.0606187582015991, + "learning_rate": 0.00010227647641042562, + "loss": 0.3524, + "step": 4700 + }, + { + "epoch": 18.9, + "grad_norm": 1.275098443031311, + "learning_rate": 0.00010062685582316067, + "loss": 0.3057, + "step": 4725 + }, + { + "epoch": 19.0, + "grad_norm": 1.971979022026062, + "learning_rate": 9.897723523589574e-05, + "loss": 0.3089, + "step": 4750 + }, + { + "epoch": 19.1, + "grad_norm": 4.085920810699463, + "learning_rate": 9.732761464863082e-05, + "loss": 0.3251, + "step": 4775 + }, + { + "epoch": 19.2, + "grad_norm": 1.5097062587738037, + "learning_rate": 9.56779940613659e-05, + "loss": 0.2869, + "step": 4800 + }, + { + "epoch": 19.3, + "grad_norm": 2.4897541999816895, + "learning_rate": 9.402837347410096e-05, + "loss": 0.28, + "step": 4825 + }, + { + "epoch": 19.4, + "grad_norm": 2.232405662536621, + "learning_rate": 9.237875288683603e-05, + "loss": 0.3585, + "step": 4850 + }, + { + "epoch": 19.5, + "grad_norm": 2.3231348991394043, + "learning_rate": 9.072913229957111e-05, + "loss": 0.3578, + "step": 4875 + }, + { + "epoch": 19.6, + "grad_norm": 1.1693369150161743, + "learning_rate": 8.907951171230617e-05, + "loss": 0.3201, + "step": 4900 + }, + { + "epoch": 19.7, + "grad_norm": 1.5570422410964966, + "learning_rate": 8.742989112504124e-05, + "loss": 0.3142, + "step": 4925 + }, + { + "epoch": 19.8, + "grad_norm": 1.122075080871582, + "learning_rate": 8.578027053777631e-05, + "loss": 0.3274, + "step": 4950 + }, + { + "epoch": 19.9, + "grad_norm": 1.4783751964569092, + "learning_rate": 8.413064995051139e-05, + "loss": 0.3246, + "step": 4975 + }, + { + "epoch": 20.0, + "grad_norm": 1.6368613243103027, + "learning_rate": 8.248102936324645e-05, + "loss": 0.2984, + "step": 5000 + } + ], + "logging_steps": 25, + "max_steps": 6250, + "num_input_tokens_seen": 0, + "num_train_epochs": 25, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 265580544000000.0, + "train_batch_size": 20, + "trial_name": null, + "trial_params": null +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/training_args.bin b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..d61f3500eb70763e1f33496c1787bdf3f07a651b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b6e0f7aa622786274a1ea4195b72f4da224050767f7b89529b85721865c5075a +size 5304 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/vocab.txt b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/vocab.txt new file mode 100644 index 0000000000000000000000000000000000000000..cc9ad53732e21ea9180f18e1fb67a38cc5702d4c --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/vocab.txt @@ -0,0 +1,4101 @@ +[PAD] +[UNK] +[CLS] +[SEP] +[MASK] +AAAAAA +AAAAAT +AAAAAC +AAAAAG +AAAATA +AAAATT +AAAATC +AAAATG +AAAACA +AAAACT +AAAACC +AAAACG +AAAAGA +AAAAGT +AAAAGC +AAAAGG +AAATAA +AAATAT +AAATAC +AAATAG +AAATTA +AAATTT +AAATTC +AAATTG +AAATCA +AAATCT +AAATCC +AAATCG +AAATGA +AAATGT +AAATGC +AAATGG +AAACAA +AAACAT +AAACAC +AAACAG +AAACTA +AAACTT +AAACTC +AAACTG +AAACCA +AAACCT +AAACCC +AAACCG +AAACGA +AAACGT +AAACGC +AAACGG +AAAGAA +AAAGAT +AAAGAC +AAAGAG +AAAGTA +AAAGTT +AAAGTC +AAAGTG +AAAGCA +AAAGCT +AAAGCC +AAAGCG +AAAGGA +AAAGGT +AAAGGC +AAAGGG +AATAAA +AATAAT +AATAAC +AATAAG +AATATA +AATATT +AATATC +AATATG +AATACA +AATACT +AATACC +AATACG +AATAGA +AATAGT +AATAGC +AATAGG +AATTAA +AATTAT +AATTAC +AATTAG +AATTTA +AATTTT +AATTTC +AATTTG +AATTCA +AATTCT +AATTCC +AATTCG +AATTGA +AATTGT +AATTGC +AATTGG +AATCAA +AATCAT +AATCAC +AATCAG +AATCTA +AATCTT +AATCTC +AATCTG +AATCCA +AATCCT +AATCCC +AATCCG +AATCGA +AATCGT +AATCGC +AATCGG +AATGAA +AATGAT +AATGAC +AATGAG +AATGTA +AATGTT +AATGTC +AATGTG +AATGCA +AATGCT +AATGCC +AATGCG +AATGGA +AATGGT +AATGGC +AATGGG +AACAAA +AACAAT +AACAAC +AACAAG +AACATA +AACATT +AACATC +AACATG +AACACA +AACACT +AACACC +AACACG +AACAGA +AACAGT +AACAGC +AACAGG +AACTAA +AACTAT +AACTAC +AACTAG +AACTTA +AACTTT +AACTTC +AACTTG +AACTCA +AACTCT +AACTCC +AACTCG +AACTGA +AACTGT +AACTGC +AACTGG +AACCAA +AACCAT +AACCAC +AACCAG +AACCTA +AACCTT +AACCTC +AACCTG +AACCCA +AACCCT +AACCCC +AACCCG +AACCGA +AACCGT +AACCGC +AACCGG +AACGAA +AACGAT +AACGAC +AACGAG +AACGTA +AACGTT +AACGTC +AACGTG +AACGCA +AACGCT +AACGCC +AACGCG +AACGGA +AACGGT +AACGGC +AACGGG +AAGAAA +AAGAAT +AAGAAC +AAGAAG +AAGATA +AAGATT +AAGATC +AAGATG +AAGACA +AAGACT +AAGACC +AAGACG +AAGAGA +AAGAGT +AAGAGC +AAGAGG +AAGTAA +AAGTAT +AAGTAC +AAGTAG +AAGTTA +AAGTTT +AAGTTC +AAGTTG +AAGTCA +AAGTCT +AAGTCC +AAGTCG +AAGTGA +AAGTGT +AAGTGC +AAGTGG +AAGCAA +AAGCAT +AAGCAC +AAGCAG +AAGCTA +AAGCTT +AAGCTC +AAGCTG +AAGCCA +AAGCCT +AAGCCC +AAGCCG +AAGCGA +AAGCGT +AAGCGC +AAGCGG +AAGGAA +AAGGAT +AAGGAC +AAGGAG +AAGGTA +AAGGTT +AAGGTC +AAGGTG +AAGGCA +AAGGCT +AAGGCC +AAGGCG +AAGGGA +AAGGGT +AAGGGC +AAGGGG +ATAAAA +ATAAAT +ATAAAC +ATAAAG +ATAATA +ATAATT +ATAATC +ATAATG +ATAACA +ATAACT +ATAACC +ATAACG +ATAAGA +ATAAGT +ATAAGC +ATAAGG +ATATAA +ATATAT +ATATAC +ATATAG +ATATTA +ATATTT +ATATTC +ATATTG +ATATCA +ATATCT +ATATCC +ATATCG +ATATGA +ATATGT +ATATGC +ATATGG +ATACAA +ATACAT +ATACAC +ATACAG +ATACTA +ATACTT +ATACTC +ATACTG +ATACCA +ATACCT +ATACCC +ATACCG +ATACGA +ATACGT +ATACGC +ATACGG +ATAGAA +ATAGAT +ATAGAC +ATAGAG +ATAGTA +ATAGTT +ATAGTC +ATAGTG +ATAGCA +ATAGCT +ATAGCC +ATAGCG +ATAGGA +ATAGGT +ATAGGC +ATAGGG +ATTAAA +ATTAAT +ATTAAC +ATTAAG +ATTATA +ATTATT +ATTATC +ATTATG +ATTACA +ATTACT +ATTACC +ATTACG +ATTAGA +ATTAGT +ATTAGC +ATTAGG +ATTTAA +ATTTAT +ATTTAC +ATTTAG +ATTTTA +ATTTTT +ATTTTC +ATTTTG +ATTTCA +ATTTCT +ATTTCC +ATTTCG +ATTTGA +ATTTGT +ATTTGC +ATTTGG +ATTCAA +ATTCAT +ATTCAC +ATTCAG +ATTCTA +ATTCTT +ATTCTC +ATTCTG +ATTCCA +ATTCCT +ATTCCC +ATTCCG +ATTCGA +ATTCGT +ATTCGC +ATTCGG +ATTGAA +ATTGAT +ATTGAC +ATTGAG +ATTGTA +ATTGTT +ATTGTC +ATTGTG +ATTGCA +ATTGCT +ATTGCC +ATTGCG +ATTGGA +ATTGGT +ATTGGC +ATTGGG +ATCAAA +ATCAAT +ATCAAC +ATCAAG +ATCATA +ATCATT +ATCATC +ATCATG +ATCACA +ATCACT +ATCACC +ATCACG +ATCAGA +ATCAGT +ATCAGC +ATCAGG +ATCTAA +ATCTAT +ATCTAC +ATCTAG +ATCTTA +ATCTTT +ATCTTC +ATCTTG +ATCTCA +ATCTCT +ATCTCC +ATCTCG +ATCTGA +ATCTGT +ATCTGC +ATCTGG +ATCCAA +ATCCAT +ATCCAC +ATCCAG +ATCCTA +ATCCTT +ATCCTC +ATCCTG +ATCCCA +ATCCCT +ATCCCC +ATCCCG +ATCCGA +ATCCGT +ATCCGC +ATCCGG +ATCGAA +ATCGAT +ATCGAC +ATCGAG +ATCGTA +ATCGTT +ATCGTC +ATCGTG +ATCGCA +ATCGCT +ATCGCC +ATCGCG +ATCGGA +ATCGGT +ATCGGC +ATCGGG +ATGAAA +ATGAAT +ATGAAC +ATGAAG +ATGATA +ATGATT +ATGATC +ATGATG +ATGACA +ATGACT +ATGACC +ATGACG +ATGAGA +ATGAGT +ATGAGC +ATGAGG +ATGTAA +ATGTAT +ATGTAC +ATGTAG +ATGTTA +ATGTTT +ATGTTC +ATGTTG +ATGTCA +ATGTCT +ATGTCC +ATGTCG +ATGTGA +ATGTGT +ATGTGC +ATGTGG +ATGCAA +ATGCAT +ATGCAC +ATGCAG +ATGCTA +ATGCTT +ATGCTC +ATGCTG +ATGCCA +ATGCCT +ATGCCC +ATGCCG +ATGCGA +ATGCGT +ATGCGC +ATGCGG +ATGGAA +ATGGAT +ATGGAC +ATGGAG +ATGGTA +ATGGTT +ATGGTC +ATGGTG +ATGGCA +ATGGCT +ATGGCC +ATGGCG +ATGGGA +ATGGGT +ATGGGC +ATGGGG +ACAAAA +ACAAAT +ACAAAC +ACAAAG +ACAATA +ACAATT +ACAATC +ACAATG +ACAACA +ACAACT +ACAACC +ACAACG +ACAAGA +ACAAGT +ACAAGC +ACAAGG +ACATAA +ACATAT +ACATAC +ACATAG +ACATTA +ACATTT +ACATTC +ACATTG +ACATCA +ACATCT +ACATCC +ACATCG +ACATGA +ACATGT +ACATGC +ACATGG +ACACAA +ACACAT +ACACAC +ACACAG +ACACTA +ACACTT +ACACTC +ACACTG +ACACCA +ACACCT +ACACCC +ACACCG +ACACGA +ACACGT +ACACGC +ACACGG +ACAGAA +ACAGAT +ACAGAC +ACAGAG +ACAGTA +ACAGTT +ACAGTC +ACAGTG +ACAGCA +ACAGCT +ACAGCC +ACAGCG +ACAGGA +ACAGGT +ACAGGC +ACAGGG +ACTAAA +ACTAAT +ACTAAC +ACTAAG +ACTATA +ACTATT +ACTATC +ACTATG +ACTACA +ACTACT +ACTACC +ACTACG +ACTAGA +ACTAGT +ACTAGC +ACTAGG +ACTTAA +ACTTAT +ACTTAC +ACTTAG +ACTTTA +ACTTTT +ACTTTC +ACTTTG +ACTTCA +ACTTCT +ACTTCC +ACTTCG +ACTTGA +ACTTGT +ACTTGC +ACTTGG +ACTCAA +ACTCAT +ACTCAC +ACTCAG +ACTCTA +ACTCTT +ACTCTC +ACTCTG +ACTCCA +ACTCCT +ACTCCC +ACTCCG +ACTCGA +ACTCGT +ACTCGC +ACTCGG +ACTGAA +ACTGAT +ACTGAC +ACTGAG +ACTGTA +ACTGTT +ACTGTC +ACTGTG +ACTGCA +ACTGCT +ACTGCC +ACTGCG +ACTGGA +ACTGGT +ACTGGC +ACTGGG +ACCAAA +ACCAAT +ACCAAC +ACCAAG +ACCATA +ACCATT +ACCATC +ACCATG +ACCACA +ACCACT +ACCACC +ACCACG +ACCAGA +ACCAGT +ACCAGC +ACCAGG +ACCTAA +ACCTAT +ACCTAC +ACCTAG +ACCTTA +ACCTTT +ACCTTC +ACCTTG +ACCTCA +ACCTCT +ACCTCC +ACCTCG +ACCTGA +ACCTGT +ACCTGC +ACCTGG +ACCCAA +ACCCAT +ACCCAC +ACCCAG +ACCCTA +ACCCTT +ACCCTC +ACCCTG +ACCCCA +ACCCCT +ACCCCC +ACCCCG +ACCCGA +ACCCGT +ACCCGC +ACCCGG +ACCGAA +ACCGAT +ACCGAC +ACCGAG +ACCGTA +ACCGTT +ACCGTC +ACCGTG +ACCGCA +ACCGCT +ACCGCC +ACCGCG +ACCGGA +ACCGGT +ACCGGC +ACCGGG +ACGAAA +ACGAAT +ACGAAC +ACGAAG +ACGATA +ACGATT +ACGATC +ACGATG +ACGACA +ACGACT +ACGACC +ACGACG +ACGAGA +ACGAGT +ACGAGC +ACGAGG +ACGTAA +ACGTAT +ACGTAC +ACGTAG +ACGTTA +ACGTTT +ACGTTC +ACGTTG +ACGTCA +ACGTCT +ACGTCC +ACGTCG +ACGTGA +ACGTGT +ACGTGC +ACGTGG +ACGCAA +ACGCAT +ACGCAC +ACGCAG +ACGCTA +ACGCTT +ACGCTC +ACGCTG +ACGCCA +ACGCCT +ACGCCC +ACGCCG +ACGCGA +ACGCGT +ACGCGC +ACGCGG +ACGGAA +ACGGAT +ACGGAC +ACGGAG +ACGGTA +ACGGTT +ACGGTC +ACGGTG +ACGGCA +ACGGCT +ACGGCC +ACGGCG +ACGGGA +ACGGGT +ACGGGC +ACGGGG +AGAAAA +AGAAAT +AGAAAC +AGAAAG +AGAATA +AGAATT +AGAATC +AGAATG +AGAACA +AGAACT +AGAACC +AGAACG +AGAAGA +AGAAGT +AGAAGC +AGAAGG +AGATAA +AGATAT +AGATAC +AGATAG +AGATTA +AGATTT +AGATTC +AGATTG +AGATCA +AGATCT +AGATCC +AGATCG +AGATGA +AGATGT +AGATGC +AGATGG +AGACAA +AGACAT +AGACAC +AGACAG +AGACTA +AGACTT +AGACTC +AGACTG +AGACCA +AGACCT +AGACCC +AGACCG +AGACGA +AGACGT +AGACGC +AGACGG +AGAGAA +AGAGAT +AGAGAC +AGAGAG +AGAGTA +AGAGTT +AGAGTC +AGAGTG +AGAGCA +AGAGCT +AGAGCC +AGAGCG +AGAGGA +AGAGGT +AGAGGC +AGAGGG +AGTAAA +AGTAAT +AGTAAC +AGTAAG +AGTATA +AGTATT +AGTATC +AGTATG +AGTACA +AGTACT +AGTACC +AGTACG +AGTAGA +AGTAGT +AGTAGC +AGTAGG +AGTTAA +AGTTAT +AGTTAC +AGTTAG +AGTTTA +AGTTTT +AGTTTC +AGTTTG +AGTTCA +AGTTCT +AGTTCC +AGTTCG +AGTTGA +AGTTGT +AGTTGC +AGTTGG +AGTCAA +AGTCAT +AGTCAC +AGTCAG +AGTCTA +AGTCTT +AGTCTC +AGTCTG +AGTCCA +AGTCCT +AGTCCC +AGTCCG +AGTCGA +AGTCGT +AGTCGC +AGTCGG +AGTGAA +AGTGAT +AGTGAC +AGTGAG +AGTGTA +AGTGTT +AGTGTC +AGTGTG +AGTGCA +AGTGCT +AGTGCC +AGTGCG +AGTGGA +AGTGGT +AGTGGC +AGTGGG +AGCAAA +AGCAAT +AGCAAC +AGCAAG +AGCATA +AGCATT +AGCATC +AGCATG +AGCACA +AGCACT +AGCACC +AGCACG +AGCAGA +AGCAGT +AGCAGC +AGCAGG +AGCTAA +AGCTAT +AGCTAC +AGCTAG +AGCTTA +AGCTTT +AGCTTC +AGCTTG +AGCTCA +AGCTCT +AGCTCC +AGCTCG +AGCTGA +AGCTGT +AGCTGC +AGCTGG +AGCCAA +AGCCAT +AGCCAC +AGCCAG +AGCCTA +AGCCTT +AGCCTC +AGCCTG +AGCCCA +AGCCCT +AGCCCC +AGCCCG +AGCCGA +AGCCGT +AGCCGC +AGCCGG +AGCGAA +AGCGAT +AGCGAC +AGCGAG +AGCGTA +AGCGTT +AGCGTC +AGCGTG +AGCGCA +AGCGCT +AGCGCC +AGCGCG +AGCGGA +AGCGGT +AGCGGC +AGCGGG +AGGAAA +AGGAAT +AGGAAC +AGGAAG +AGGATA +AGGATT +AGGATC +AGGATG +AGGACA +AGGACT +AGGACC +AGGACG +AGGAGA +AGGAGT +AGGAGC +AGGAGG +AGGTAA +AGGTAT +AGGTAC +AGGTAG +AGGTTA +AGGTTT +AGGTTC +AGGTTG +AGGTCA +AGGTCT +AGGTCC +AGGTCG +AGGTGA +AGGTGT +AGGTGC +AGGTGG +AGGCAA +AGGCAT +AGGCAC +AGGCAG +AGGCTA +AGGCTT +AGGCTC +AGGCTG +AGGCCA +AGGCCT +AGGCCC +AGGCCG +AGGCGA +AGGCGT +AGGCGC +AGGCGG +AGGGAA +AGGGAT +AGGGAC +AGGGAG +AGGGTA +AGGGTT +AGGGTC +AGGGTG +AGGGCA +AGGGCT +AGGGCC +AGGGCG +AGGGGA +AGGGGT +AGGGGC +AGGGGG +TAAAAA +TAAAAT +TAAAAC +TAAAAG +TAAATA +TAAATT +TAAATC +TAAATG +TAAACA +TAAACT +TAAACC +TAAACG +TAAAGA +TAAAGT +TAAAGC +TAAAGG +TAATAA +TAATAT +TAATAC +TAATAG +TAATTA +TAATTT +TAATTC +TAATTG +TAATCA +TAATCT +TAATCC +TAATCG +TAATGA +TAATGT +TAATGC +TAATGG +TAACAA +TAACAT +TAACAC +TAACAG +TAACTA +TAACTT +TAACTC +TAACTG +TAACCA +TAACCT +TAACCC +TAACCG +TAACGA +TAACGT +TAACGC +TAACGG +TAAGAA +TAAGAT +TAAGAC +TAAGAG +TAAGTA +TAAGTT +TAAGTC +TAAGTG +TAAGCA +TAAGCT +TAAGCC +TAAGCG +TAAGGA +TAAGGT +TAAGGC +TAAGGG +TATAAA +TATAAT +TATAAC +TATAAG +TATATA +TATATT +TATATC +TATATG +TATACA +TATACT +TATACC +TATACG +TATAGA +TATAGT +TATAGC +TATAGG +TATTAA +TATTAT +TATTAC +TATTAG +TATTTA +TATTTT +TATTTC +TATTTG +TATTCA +TATTCT +TATTCC +TATTCG +TATTGA +TATTGT +TATTGC +TATTGG +TATCAA +TATCAT +TATCAC +TATCAG +TATCTA +TATCTT +TATCTC +TATCTG +TATCCA +TATCCT +TATCCC +TATCCG +TATCGA +TATCGT +TATCGC +TATCGG +TATGAA +TATGAT +TATGAC +TATGAG +TATGTA +TATGTT +TATGTC +TATGTG +TATGCA +TATGCT +TATGCC +TATGCG +TATGGA +TATGGT +TATGGC +TATGGG +TACAAA +TACAAT +TACAAC +TACAAG +TACATA +TACATT +TACATC +TACATG +TACACA +TACACT +TACACC +TACACG +TACAGA +TACAGT +TACAGC +TACAGG +TACTAA +TACTAT +TACTAC +TACTAG +TACTTA +TACTTT +TACTTC +TACTTG +TACTCA +TACTCT +TACTCC +TACTCG +TACTGA +TACTGT +TACTGC +TACTGG +TACCAA +TACCAT +TACCAC +TACCAG +TACCTA +TACCTT +TACCTC +TACCTG +TACCCA +TACCCT +TACCCC +TACCCG +TACCGA +TACCGT +TACCGC +TACCGG +TACGAA +TACGAT +TACGAC +TACGAG +TACGTA +TACGTT +TACGTC +TACGTG +TACGCA +TACGCT +TACGCC +TACGCG +TACGGA +TACGGT +TACGGC +TACGGG +TAGAAA +TAGAAT +TAGAAC +TAGAAG +TAGATA +TAGATT +TAGATC +TAGATG +TAGACA +TAGACT +TAGACC +TAGACG +TAGAGA +TAGAGT +TAGAGC +TAGAGG +TAGTAA +TAGTAT +TAGTAC +TAGTAG +TAGTTA +TAGTTT +TAGTTC +TAGTTG +TAGTCA +TAGTCT +TAGTCC +TAGTCG +TAGTGA +TAGTGT +TAGTGC +TAGTGG +TAGCAA +TAGCAT +TAGCAC +TAGCAG +TAGCTA +TAGCTT +TAGCTC +TAGCTG +TAGCCA +TAGCCT +TAGCCC +TAGCCG +TAGCGA +TAGCGT +TAGCGC +TAGCGG +TAGGAA +TAGGAT +TAGGAC +TAGGAG +TAGGTA +TAGGTT +TAGGTC +TAGGTG +TAGGCA +TAGGCT +TAGGCC +TAGGCG +TAGGGA +TAGGGT +TAGGGC +TAGGGG +TTAAAA +TTAAAT +TTAAAC +TTAAAG +TTAATA +TTAATT +TTAATC +TTAATG +TTAACA +TTAACT +TTAACC +TTAACG +TTAAGA +TTAAGT +TTAAGC +TTAAGG +TTATAA +TTATAT +TTATAC +TTATAG +TTATTA +TTATTT +TTATTC +TTATTG +TTATCA +TTATCT +TTATCC +TTATCG +TTATGA +TTATGT +TTATGC +TTATGG +TTACAA +TTACAT +TTACAC +TTACAG +TTACTA +TTACTT +TTACTC +TTACTG +TTACCA +TTACCT +TTACCC +TTACCG +TTACGA +TTACGT +TTACGC +TTACGG +TTAGAA +TTAGAT +TTAGAC +TTAGAG +TTAGTA +TTAGTT +TTAGTC +TTAGTG +TTAGCA +TTAGCT +TTAGCC +TTAGCG +TTAGGA +TTAGGT +TTAGGC +TTAGGG +TTTAAA +TTTAAT +TTTAAC +TTTAAG +TTTATA +TTTATT +TTTATC +TTTATG +TTTACA +TTTACT +TTTACC +TTTACG +TTTAGA +TTTAGT +TTTAGC +TTTAGG +TTTTAA +TTTTAT +TTTTAC +TTTTAG +TTTTTA +TTTTTT +TTTTTC +TTTTTG +TTTTCA +TTTTCT +TTTTCC +TTTTCG +TTTTGA +TTTTGT +TTTTGC +TTTTGG +TTTCAA +TTTCAT +TTTCAC +TTTCAG +TTTCTA +TTTCTT +TTTCTC +TTTCTG +TTTCCA +TTTCCT +TTTCCC +TTTCCG +TTTCGA +TTTCGT +TTTCGC +TTTCGG +TTTGAA +TTTGAT +TTTGAC +TTTGAG +TTTGTA +TTTGTT +TTTGTC +TTTGTG +TTTGCA +TTTGCT +TTTGCC +TTTGCG +TTTGGA +TTTGGT +TTTGGC +TTTGGG +TTCAAA +TTCAAT +TTCAAC +TTCAAG +TTCATA +TTCATT +TTCATC +TTCATG +TTCACA +TTCACT +TTCACC +TTCACG +TTCAGA +TTCAGT +TTCAGC +TTCAGG +TTCTAA +TTCTAT +TTCTAC +TTCTAG +TTCTTA +TTCTTT +TTCTTC +TTCTTG +TTCTCA +TTCTCT +TTCTCC +TTCTCG +TTCTGA +TTCTGT +TTCTGC +TTCTGG +TTCCAA +TTCCAT +TTCCAC +TTCCAG +TTCCTA +TTCCTT +TTCCTC +TTCCTG +TTCCCA +TTCCCT +TTCCCC +TTCCCG +TTCCGA +TTCCGT +TTCCGC +TTCCGG +TTCGAA +TTCGAT +TTCGAC +TTCGAG +TTCGTA +TTCGTT +TTCGTC +TTCGTG +TTCGCA +TTCGCT +TTCGCC +TTCGCG +TTCGGA +TTCGGT +TTCGGC +TTCGGG +TTGAAA +TTGAAT +TTGAAC +TTGAAG +TTGATA +TTGATT +TTGATC +TTGATG +TTGACA +TTGACT +TTGACC +TTGACG +TTGAGA +TTGAGT +TTGAGC +TTGAGG +TTGTAA +TTGTAT +TTGTAC +TTGTAG +TTGTTA +TTGTTT +TTGTTC +TTGTTG +TTGTCA +TTGTCT +TTGTCC +TTGTCG +TTGTGA +TTGTGT +TTGTGC +TTGTGG +TTGCAA +TTGCAT +TTGCAC +TTGCAG +TTGCTA +TTGCTT +TTGCTC +TTGCTG +TTGCCA +TTGCCT +TTGCCC +TTGCCG +TTGCGA +TTGCGT +TTGCGC +TTGCGG +TTGGAA +TTGGAT +TTGGAC +TTGGAG +TTGGTA +TTGGTT +TTGGTC +TTGGTG +TTGGCA +TTGGCT +TTGGCC +TTGGCG +TTGGGA +TTGGGT +TTGGGC +TTGGGG +TCAAAA +TCAAAT +TCAAAC +TCAAAG +TCAATA +TCAATT +TCAATC +TCAATG +TCAACA +TCAACT +TCAACC +TCAACG +TCAAGA +TCAAGT +TCAAGC +TCAAGG +TCATAA +TCATAT +TCATAC +TCATAG +TCATTA +TCATTT +TCATTC +TCATTG +TCATCA +TCATCT +TCATCC +TCATCG +TCATGA +TCATGT +TCATGC +TCATGG +TCACAA +TCACAT +TCACAC +TCACAG +TCACTA +TCACTT +TCACTC +TCACTG +TCACCA +TCACCT +TCACCC +TCACCG +TCACGA +TCACGT +TCACGC +TCACGG +TCAGAA +TCAGAT +TCAGAC +TCAGAG +TCAGTA +TCAGTT +TCAGTC +TCAGTG +TCAGCA +TCAGCT +TCAGCC +TCAGCG +TCAGGA +TCAGGT +TCAGGC +TCAGGG +TCTAAA +TCTAAT +TCTAAC +TCTAAG +TCTATA +TCTATT +TCTATC +TCTATG +TCTACA +TCTACT +TCTACC +TCTACG +TCTAGA +TCTAGT +TCTAGC +TCTAGG +TCTTAA +TCTTAT +TCTTAC +TCTTAG +TCTTTA +TCTTTT +TCTTTC +TCTTTG +TCTTCA +TCTTCT +TCTTCC +TCTTCG +TCTTGA +TCTTGT +TCTTGC +TCTTGG +TCTCAA +TCTCAT +TCTCAC +TCTCAG +TCTCTA +TCTCTT +TCTCTC +TCTCTG +TCTCCA +TCTCCT +TCTCCC +TCTCCG +TCTCGA +TCTCGT +TCTCGC +TCTCGG +TCTGAA +TCTGAT +TCTGAC +TCTGAG +TCTGTA +TCTGTT +TCTGTC +TCTGTG +TCTGCA +TCTGCT +TCTGCC +TCTGCG +TCTGGA +TCTGGT +TCTGGC +TCTGGG +TCCAAA +TCCAAT +TCCAAC +TCCAAG +TCCATA +TCCATT +TCCATC +TCCATG +TCCACA +TCCACT +TCCACC +TCCACG +TCCAGA +TCCAGT +TCCAGC +TCCAGG +TCCTAA +TCCTAT +TCCTAC +TCCTAG +TCCTTA +TCCTTT +TCCTTC +TCCTTG +TCCTCA +TCCTCT +TCCTCC +TCCTCG +TCCTGA +TCCTGT +TCCTGC +TCCTGG +TCCCAA +TCCCAT +TCCCAC +TCCCAG +TCCCTA +TCCCTT +TCCCTC +TCCCTG +TCCCCA +TCCCCT +TCCCCC +TCCCCG +TCCCGA +TCCCGT +TCCCGC +TCCCGG +TCCGAA +TCCGAT +TCCGAC +TCCGAG +TCCGTA +TCCGTT +TCCGTC +TCCGTG +TCCGCA +TCCGCT +TCCGCC +TCCGCG +TCCGGA +TCCGGT +TCCGGC +TCCGGG +TCGAAA +TCGAAT +TCGAAC +TCGAAG +TCGATA +TCGATT +TCGATC +TCGATG +TCGACA +TCGACT +TCGACC +TCGACG +TCGAGA +TCGAGT +TCGAGC +TCGAGG +TCGTAA +TCGTAT +TCGTAC +TCGTAG +TCGTTA +TCGTTT +TCGTTC +TCGTTG +TCGTCA +TCGTCT +TCGTCC +TCGTCG +TCGTGA +TCGTGT +TCGTGC +TCGTGG +TCGCAA +TCGCAT +TCGCAC +TCGCAG +TCGCTA +TCGCTT +TCGCTC +TCGCTG +TCGCCA +TCGCCT +TCGCCC +TCGCCG +TCGCGA +TCGCGT +TCGCGC +TCGCGG +TCGGAA +TCGGAT +TCGGAC +TCGGAG +TCGGTA +TCGGTT +TCGGTC +TCGGTG +TCGGCA +TCGGCT +TCGGCC +TCGGCG +TCGGGA +TCGGGT +TCGGGC +TCGGGG +TGAAAA +TGAAAT +TGAAAC +TGAAAG +TGAATA +TGAATT +TGAATC +TGAATG +TGAACA +TGAACT +TGAACC +TGAACG +TGAAGA +TGAAGT +TGAAGC +TGAAGG +TGATAA +TGATAT +TGATAC +TGATAG +TGATTA +TGATTT +TGATTC +TGATTG +TGATCA +TGATCT +TGATCC +TGATCG +TGATGA +TGATGT +TGATGC +TGATGG +TGACAA +TGACAT +TGACAC +TGACAG +TGACTA +TGACTT +TGACTC +TGACTG +TGACCA +TGACCT +TGACCC +TGACCG +TGACGA +TGACGT +TGACGC +TGACGG +TGAGAA +TGAGAT +TGAGAC +TGAGAG +TGAGTA +TGAGTT +TGAGTC +TGAGTG +TGAGCA +TGAGCT +TGAGCC +TGAGCG +TGAGGA +TGAGGT +TGAGGC +TGAGGG +TGTAAA +TGTAAT +TGTAAC +TGTAAG +TGTATA +TGTATT +TGTATC +TGTATG +TGTACA +TGTACT +TGTACC +TGTACG +TGTAGA +TGTAGT +TGTAGC +TGTAGG +TGTTAA +TGTTAT +TGTTAC +TGTTAG +TGTTTA +TGTTTT +TGTTTC +TGTTTG +TGTTCA +TGTTCT +TGTTCC +TGTTCG +TGTTGA +TGTTGT +TGTTGC +TGTTGG +TGTCAA +TGTCAT +TGTCAC +TGTCAG +TGTCTA +TGTCTT +TGTCTC +TGTCTG +TGTCCA +TGTCCT +TGTCCC +TGTCCG +TGTCGA +TGTCGT +TGTCGC +TGTCGG +TGTGAA +TGTGAT +TGTGAC +TGTGAG +TGTGTA +TGTGTT +TGTGTC +TGTGTG +TGTGCA +TGTGCT +TGTGCC +TGTGCG +TGTGGA +TGTGGT +TGTGGC +TGTGGG +TGCAAA +TGCAAT +TGCAAC +TGCAAG +TGCATA +TGCATT +TGCATC +TGCATG +TGCACA +TGCACT +TGCACC +TGCACG +TGCAGA +TGCAGT +TGCAGC +TGCAGG +TGCTAA +TGCTAT +TGCTAC +TGCTAG +TGCTTA +TGCTTT +TGCTTC +TGCTTG +TGCTCA +TGCTCT +TGCTCC +TGCTCG +TGCTGA +TGCTGT +TGCTGC +TGCTGG +TGCCAA +TGCCAT +TGCCAC +TGCCAG +TGCCTA +TGCCTT +TGCCTC +TGCCTG +TGCCCA +TGCCCT +TGCCCC +TGCCCG +TGCCGA +TGCCGT +TGCCGC +TGCCGG +TGCGAA +TGCGAT +TGCGAC +TGCGAG +TGCGTA +TGCGTT +TGCGTC +TGCGTG +TGCGCA +TGCGCT +TGCGCC +TGCGCG +TGCGGA +TGCGGT +TGCGGC +TGCGGG +TGGAAA +TGGAAT +TGGAAC +TGGAAG +TGGATA +TGGATT +TGGATC +TGGATG +TGGACA +TGGACT +TGGACC +TGGACG +TGGAGA +TGGAGT +TGGAGC +TGGAGG +TGGTAA +TGGTAT +TGGTAC +TGGTAG +TGGTTA +TGGTTT +TGGTTC +TGGTTG +TGGTCA +TGGTCT +TGGTCC +TGGTCG +TGGTGA +TGGTGT +TGGTGC +TGGTGG +TGGCAA +TGGCAT +TGGCAC +TGGCAG +TGGCTA +TGGCTT +TGGCTC +TGGCTG +TGGCCA +TGGCCT +TGGCCC +TGGCCG +TGGCGA +TGGCGT +TGGCGC +TGGCGG +TGGGAA +TGGGAT +TGGGAC +TGGGAG +TGGGTA +TGGGTT +TGGGTC +TGGGTG +TGGGCA +TGGGCT +TGGGCC +TGGGCG +TGGGGA +TGGGGT +TGGGGC +TGGGGG +CAAAAA +CAAAAT +CAAAAC +CAAAAG +CAAATA +CAAATT +CAAATC +CAAATG +CAAACA +CAAACT +CAAACC +CAAACG +CAAAGA +CAAAGT +CAAAGC +CAAAGG +CAATAA +CAATAT +CAATAC +CAATAG +CAATTA +CAATTT +CAATTC +CAATTG +CAATCA +CAATCT +CAATCC +CAATCG +CAATGA +CAATGT +CAATGC +CAATGG +CAACAA +CAACAT +CAACAC +CAACAG +CAACTA +CAACTT +CAACTC +CAACTG +CAACCA +CAACCT +CAACCC +CAACCG +CAACGA +CAACGT +CAACGC +CAACGG +CAAGAA +CAAGAT +CAAGAC +CAAGAG +CAAGTA +CAAGTT +CAAGTC +CAAGTG +CAAGCA +CAAGCT +CAAGCC +CAAGCG +CAAGGA +CAAGGT +CAAGGC +CAAGGG +CATAAA +CATAAT +CATAAC +CATAAG +CATATA +CATATT +CATATC +CATATG +CATACA +CATACT +CATACC +CATACG +CATAGA +CATAGT +CATAGC +CATAGG +CATTAA +CATTAT +CATTAC +CATTAG +CATTTA +CATTTT +CATTTC +CATTTG +CATTCA +CATTCT +CATTCC +CATTCG +CATTGA +CATTGT +CATTGC +CATTGG +CATCAA +CATCAT +CATCAC +CATCAG +CATCTA +CATCTT +CATCTC +CATCTG +CATCCA +CATCCT +CATCCC +CATCCG +CATCGA +CATCGT +CATCGC +CATCGG +CATGAA +CATGAT +CATGAC +CATGAG +CATGTA +CATGTT +CATGTC +CATGTG +CATGCA +CATGCT +CATGCC +CATGCG +CATGGA +CATGGT +CATGGC +CATGGG +CACAAA +CACAAT +CACAAC +CACAAG +CACATA +CACATT +CACATC +CACATG +CACACA +CACACT +CACACC +CACACG +CACAGA +CACAGT +CACAGC +CACAGG +CACTAA +CACTAT +CACTAC +CACTAG +CACTTA +CACTTT +CACTTC +CACTTG +CACTCA +CACTCT +CACTCC +CACTCG +CACTGA +CACTGT +CACTGC +CACTGG +CACCAA +CACCAT +CACCAC +CACCAG +CACCTA +CACCTT +CACCTC +CACCTG +CACCCA +CACCCT +CACCCC +CACCCG +CACCGA +CACCGT +CACCGC +CACCGG +CACGAA +CACGAT +CACGAC +CACGAG +CACGTA +CACGTT +CACGTC +CACGTG +CACGCA +CACGCT +CACGCC +CACGCG +CACGGA +CACGGT +CACGGC +CACGGG +CAGAAA +CAGAAT +CAGAAC +CAGAAG +CAGATA +CAGATT +CAGATC +CAGATG +CAGACA +CAGACT +CAGACC +CAGACG +CAGAGA +CAGAGT +CAGAGC +CAGAGG +CAGTAA +CAGTAT +CAGTAC +CAGTAG +CAGTTA +CAGTTT +CAGTTC +CAGTTG +CAGTCA +CAGTCT +CAGTCC +CAGTCG +CAGTGA +CAGTGT +CAGTGC +CAGTGG +CAGCAA +CAGCAT +CAGCAC +CAGCAG +CAGCTA +CAGCTT +CAGCTC +CAGCTG +CAGCCA +CAGCCT +CAGCCC +CAGCCG +CAGCGA +CAGCGT +CAGCGC +CAGCGG +CAGGAA +CAGGAT +CAGGAC +CAGGAG +CAGGTA +CAGGTT +CAGGTC +CAGGTG +CAGGCA +CAGGCT +CAGGCC +CAGGCG +CAGGGA +CAGGGT +CAGGGC +CAGGGG +CTAAAA +CTAAAT +CTAAAC +CTAAAG +CTAATA +CTAATT +CTAATC +CTAATG +CTAACA +CTAACT +CTAACC +CTAACG +CTAAGA +CTAAGT +CTAAGC +CTAAGG +CTATAA +CTATAT +CTATAC +CTATAG +CTATTA +CTATTT +CTATTC +CTATTG +CTATCA +CTATCT +CTATCC +CTATCG +CTATGA +CTATGT +CTATGC +CTATGG +CTACAA +CTACAT +CTACAC +CTACAG +CTACTA +CTACTT +CTACTC +CTACTG +CTACCA +CTACCT +CTACCC +CTACCG +CTACGA +CTACGT +CTACGC +CTACGG +CTAGAA +CTAGAT +CTAGAC +CTAGAG +CTAGTA +CTAGTT +CTAGTC +CTAGTG +CTAGCA +CTAGCT +CTAGCC +CTAGCG +CTAGGA +CTAGGT +CTAGGC +CTAGGG +CTTAAA +CTTAAT +CTTAAC +CTTAAG +CTTATA +CTTATT +CTTATC +CTTATG +CTTACA +CTTACT +CTTACC +CTTACG +CTTAGA +CTTAGT +CTTAGC +CTTAGG +CTTTAA +CTTTAT +CTTTAC +CTTTAG +CTTTTA +CTTTTT +CTTTTC +CTTTTG +CTTTCA +CTTTCT +CTTTCC +CTTTCG +CTTTGA +CTTTGT +CTTTGC +CTTTGG +CTTCAA +CTTCAT +CTTCAC +CTTCAG +CTTCTA +CTTCTT +CTTCTC +CTTCTG +CTTCCA +CTTCCT +CTTCCC +CTTCCG +CTTCGA +CTTCGT +CTTCGC +CTTCGG +CTTGAA +CTTGAT +CTTGAC +CTTGAG +CTTGTA +CTTGTT +CTTGTC +CTTGTG +CTTGCA +CTTGCT +CTTGCC +CTTGCG +CTTGGA +CTTGGT +CTTGGC +CTTGGG +CTCAAA +CTCAAT +CTCAAC +CTCAAG +CTCATA +CTCATT +CTCATC +CTCATG +CTCACA +CTCACT +CTCACC +CTCACG +CTCAGA +CTCAGT +CTCAGC +CTCAGG +CTCTAA +CTCTAT +CTCTAC +CTCTAG +CTCTTA +CTCTTT +CTCTTC +CTCTTG +CTCTCA +CTCTCT +CTCTCC +CTCTCG +CTCTGA +CTCTGT +CTCTGC +CTCTGG +CTCCAA +CTCCAT +CTCCAC +CTCCAG +CTCCTA +CTCCTT +CTCCTC +CTCCTG +CTCCCA +CTCCCT +CTCCCC +CTCCCG +CTCCGA +CTCCGT +CTCCGC +CTCCGG +CTCGAA +CTCGAT +CTCGAC +CTCGAG +CTCGTA +CTCGTT +CTCGTC +CTCGTG +CTCGCA +CTCGCT +CTCGCC +CTCGCG +CTCGGA +CTCGGT +CTCGGC +CTCGGG +CTGAAA +CTGAAT +CTGAAC +CTGAAG +CTGATA +CTGATT +CTGATC +CTGATG +CTGACA +CTGACT +CTGACC +CTGACG +CTGAGA +CTGAGT +CTGAGC +CTGAGG +CTGTAA +CTGTAT +CTGTAC +CTGTAG +CTGTTA +CTGTTT +CTGTTC +CTGTTG +CTGTCA +CTGTCT +CTGTCC +CTGTCG +CTGTGA +CTGTGT +CTGTGC +CTGTGG +CTGCAA +CTGCAT +CTGCAC +CTGCAG +CTGCTA +CTGCTT +CTGCTC +CTGCTG +CTGCCA +CTGCCT +CTGCCC +CTGCCG +CTGCGA +CTGCGT +CTGCGC +CTGCGG +CTGGAA +CTGGAT +CTGGAC +CTGGAG +CTGGTA +CTGGTT +CTGGTC +CTGGTG +CTGGCA +CTGGCT +CTGGCC +CTGGCG +CTGGGA +CTGGGT +CTGGGC +CTGGGG +CCAAAA +CCAAAT +CCAAAC +CCAAAG +CCAATA +CCAATT +CCAATC +CCAATG +CCAACA +CCAACT +CCAACC +CCAACG +CCAAGA +CCAAGT +CCAAGC +CCAAGG +CCATAA +CCATAT +CCATAC +CCATAG +CCATTA +CCATTT +CCATTC +CCATTG +CCATCA +CCATCT +CCATCC +CCATCG +CCATGA +CCATGT +CCATGC +CCATGG +CCACAA +CCACAT +CCACAC +CCACAG +CCACTA +CCACTT +CCACTC +CCACTG +CCACCA +CCACCT +CCACCC +CCACCG +CCACGA +CCACGT +CCACGC +CCACGG +CCAGAA +CCAGAT +CCAGAC +CCAGAG +CCAGTA +CCAGTT +CCAGTC +CCAGTG +CCAGCA +CCAGCT +CCAGCC +CCAGCG +CCAGGA +CCAGGT +CCAGGC +CCAGGG +CCTAAA +CCTAAT +CCTAAC +CCTAAG +CCTATA +CCTATT +CCTATC +CCTATG +CCTACA +CCTACT +CCTACC +CCTACG +CCTAGA +CCTAGT +CCTAGC +CCTAGG +CCTTAA +CCTTAT +CCTTAC +CCTTAG +CCTTTA +CCTTTT +CCTTTC +CCTTTG +CCTTCA +CCTTCT +CCTTCC +CCTTCG +CCTTGA +CCTTGT +CCTTGC +CCTTGG +CCTCAA +CCTCAT +CCTCAC +CCTCAG +CCTCTA +CCTCTT +CCTCTC +CCTCTG +CCTCCA +CCTCCT +CCTCCC +CCTCCG +CCTCGA +CCTCGT +CCTCGC +CCTCGG +CCTGAA +CCTGAT +CCTGAC +CCTGAG +CCTGTA +CCTGTT +CCTGTC +CCTGTG +CCTGCA +CCTGCT +CCTGCC +CCTGCG +CCTGGA +CCTGGT +CCTGGC +CCTGGG +CCCAAA +CCCAAT +CCCAAC +CCCAAG +CCCATA +CCCATT +CCCATC +CCCATG +CCCACA +CCCACT +CCCACC +CCCACG +CCCAGA +CCCAGT +CCCAGC +CCCAGG +CCCTAA +CCCTAT +CCCTAC +CCCTAG +CCCTTA +CCCTTT +CCCTTC +CCCTTG +CCCTCA +CCCTCT +CCCTCC +CCCTCG +CCCTGA +CCCTGT +CCCTGC +CCCTGG +CCCCAA +CCCCAT +CCCCAC +CCCCAG +CCCCTA +CCCCTT +CCCCTC +CCCCTG +CCCCCA +CCCCCT +CCCCCC +CCCCCG +CCCCGA +CCCCGT +CCCCGC +CCCCGG +CCCGAA +CCCGAT +CCCGAC +CCCGAG +CCCGTA +CCCGTT +CCCGTC +CCCGTG +CCCGCA +CCCGCT +CCCGCC +CCCGCG +CCCGGA +CCCGGT +CCCGGC +CCCGGG +CCGAAA +CCGAAT +CCGAAC +CCGAAG +CCGATA +CCGATT +CCGATC +CCGATG +CCGACA +CCGACT +CCGACC +CCGACG +CCGAGA +CCGAGT +CCGAGC +CCGAGG +CCGTAA +CCGTAT +CCGTAC +CCGTAG +CCGTTA +CCGTTT +CCGTTC +CCGTTG +CCGTCA +CCGTCT +CCGTCC +CCGTCG +CCGTGA +CCGTGT +CCGTGC +CCGTGG +CCGCAA +CCGCAT +CCGCAC +CCGCAG +CCGCTA +CCGCTT +CCGCTC +CCGCTG +CCGCCA +CCGCCT +CCGCCC +CCGCCG +CCGCGA +CCGCGT +CCGCGC +CCGCGG +CCGGAA +CCGGAT +CCGGAC +CCGGAG +CCGGTA +CCGGTT +CCGGTC +CCGGTG +CCGGCA +CCGGCT +CCGGCC +CCGGCG +CCGGGA +CCGGGT +CCGGGC +CCGGGG +CGAAAA +CGAAAT +CGAAAC +CGAAAG +CGAATA +CGAATT +CGAATC +CGAATG +CGAACA +CGAACT +CGAACC +CGAACG +CGAAGA +CGAAGT +CGAAGC +CGAAGG +CGATAA +CGATAT +CGATAC +CGATAG +CGATTA +CGATTT +CGATTC +CGATTG +CGATCA +CGATCT +CGATCC +CGATCG +CGATGA +CGATGT +CGATGC +CGATGG +CGACAA +CGACAT +CGACAC +CGACAG +CGACTA +CGACTT +CGACTC +CGACTG +CGACCA +CGACCT +CGACCC +CGACCG +CGACGA +CGACGT +CGACGC +CGACGG +CGAGAA +CGAGAT +CGAGAC +CGAGAG +CGAGTA +CGAGTT +CGAGTC +CGAGTG +CGAGCA +CGAGCT +CGAGCC +CGAGCG +CGAGGA +CGAGGT +CGAGGC +CGAGGG +CGTAAA +CGTAAT +CGTAAC +CGTAAG +CGTATA +CGTATT +CGTATC +CGTATG +CGTACA +CGTACT +CGTACC +CGTACG +CGTAGA +CGTAGT +CGTAGC +CGTAGG +CGTTAA +CGTTAT +CGTTAC +CGTTAG +CGTTTA +CGTTTT +CGTTTC +CGTTTG +CGTTCA +CGTTCT +CGTTCC +CGTTCG +CGTTGA +CGTTGT +CGTTGC +CGTTGG +CGTCAA +CGTCAT +CGTCAC +CGTCAG +CGTCTA +CGTCTT +CGTCTC +CGTCTG +CGTCCA +CGTCCT +CGTCCC +CGTCCG +CGTCGA +CGTCGT +CGTCGC +CGTCGG +CGTGAA +CGTGAT +CGTGAC +CGTGAG +CGTGTA +CGTGTT +CGTGTC +CGTGTG +CGTGCA +CGTGCT +CGTGCC +CGTGCG +CGTGGA +CGTGGT +CGTGGC +CGTGGG +CGCAAA +CGCAAT +CGCAAC +CGCAAG +CGCATA +CGCATT +CGCATC +CGCATG +CGCACA +CGCACT +CGCACC +CGCACG +CGCAGA +CGCAGT +CGCAGC +CGCAGG +CGCTAA +CGCTAT +CGCTAC +CGCTAG +CGCTTA +CGCTTT +CGCTTC +CGCTTG +CGCTCA +CGCTCT +CGCTCC +CGCTCG +CGCTGA +CGCTGT +CGCTGC +CGCTGG +CGCCAA +CGCCAT +CGCCAC +CGCCAG +CGCCTA +CGCCTT +CGCCTC +CGCCTG +CGCCCA +CGCCCT +CGCCCC +CGCCCG +CGCCGA +CGCCGT +CGCCGC +CGCCGG +CGCGAA +CGCGAT +CGCGAC +CGCGAG +CGCGTA +CGCGTT +CGCGTC +CGCGTG +CGCGCA +CGCGCT +CGCGCC +CGCGCG +CGCGGA +CGCGGT +CGCGGC +CGCGGG +CGGAAA +CGGAAT +CGGAAC +CGGAAG +CGGATA +CGGATT +CGGATC +CGGATG +CGGACA +CGGACT +CGGACC +CGGACG +CGGAGA +CGGAGT +CGGAGC +CGGAGG +CGGTAA +CGGTAT +CGGTAC +CGGTAG +CGGTTA +CGGTTT +CGGTTC +CGGTTG +CGGTCA +CGGTCT +CGGTCC +CGGTCG +CGGTGA +CGGTGT +CGGTGC +CGGTGG +CGGCAA +CGGCAT +CGGCAC +CGGCAG +CGGCTA +CGGCTT +CGGCTC +CGGCTG +CGGCCA +CGGCCT +CGGCCC +CGGCCG +CGGCGA +CGGCGT +CGGCGC +CGGCGG +CGGGAA +CGGGAT +CGGGAC +CGGGAG +CGGGTA +CGGGTT +CGGGTC +CGGGTG +CGGGCA +CGGGCT +CGGGCC +CGGGCG +CGGGGA +CGGGGT +CGGGGC +CGGGGG +GAAAAA +GAAAAT +GAAAAC +GAAAAG +GAAATA +GAAATT +GAAATC +GAAATG +GAAACA +GAAACT +GAAACC +GAAACG +GAAAGA +GAAAGT +GAAAGC +GAAAGG +GAATAA +GAATAT +GAATAC +GAATAG +GAATTA +GAATTT +GAATTC +GAATTG +GAATCA +GAATCT +GAATCC +GAATCG +GAATGA +GAATGT +GAATGC +GAATGG +GAACAA +GAACAT +GAACAC +GAACAG +GAACTA +GAACTT +GAACTC +GAACTG +GAACCA +GAACCT +GAACCC +GAACCG +GAACGA +GAACGT +GAACGC +GAACGG +GAAGAA +GAAGAT +GAAGAC +GAAGAG +GAAGTA +GAAGTT +GAAGTC +GAAGTG +GAAGCA +GAAGCT +GAAGCC +GAAGCG +GAAGGA +GAAGGT +GAAGGC +GAAGGG +GATAAA +GATAAT +GATAAC +GATAAG +GATATA +GATATT +GATATC +GATATG +GATACA +GATACT +GATACC +GATACG +GATAGA +GATAGT +GATAGC +GATAGG +GATTAA +GATTAT +GATTAC +GATTAG +GATTTA +GATTTT +GATTTC +GATTTG +GATTCA +GATTCT +GATTCC +GATTCG +GATTGA +GATTGT +GATTGC +GATTGG +GATCAA +GATCAT +GATCAC +GATCAG +GATCTA +GATCTT +GATCTC +GATCTG +GATCCA +GATCCT +GATCCC +GATCCG +GATCGA +GATCGT +GATCGC +GATCGG +GATGAA +GATGAT +GATGAC +GATGAG +GATGTA +GATGTT +GATGTC +GATGTG +GATGCA +GATGCT +GATGCC +GATGCG +GATGGA +GATGGT +GATGGC +GATGGG +GACAAA +GACAAT +GACAAC +GACAAG +GACATA +GACATT +GACATC +GACATG +GACACA +GACACT +GACACC +GACACG +GACAGA +GACAGT +GACAGC +GACAGG +GACTAA +GACTAT +GACTAC +GACTAG +GACTTA +GACTTT +GACTTC +GACTTG +GACTCA +GACTCT +GACTCC +GACTCG +GACTGA +GACTGT +GACTGC +GACTGG +GACCAA +GACCAT +GACCAC +GACCAG +GACCTA +GACCTT +GACCTC +GACCTG +GACCCA +GACCCT +GACCCC +GACCCG +GACCGA +GACCGT +GACCGC +GACCGG +GACGAA +GACGAT +GACGAC +GACGAG +GACGTA +GACGTT +GACGTC +GACGTG +GACGCA +GACGCT +GACGCC +GACGCG +GACGGA +GACGGT +GACGGC +GACGGG +GAGAAA +GAGAAT +GAGAAC +GAGAAG +GAGATA +GAGATT +GAGATC +GAGATG +GAGACA +GAGACT +GAGACC +GAGACG +GAGAGA +GAGAGT +GAGAGC +GAGAGG +GAGTAA +GAGTAT +GAGTAC +GAGTAG +GAGTTA +GAGTTT +GAGTTC +GAGTTG +GAGTCA +GAGTCT +GAGTCC +GAGTCG +GAGTGA +GAGTGT +GAGTGC +GAGTGG +GAGCAA +GAGCAT +GAGCAC +GAGCAG +GAGCTA +GAGCTT +GAGCTC +GAGCTG +GAGCCA +GAGCCT +GAGCCC +GAGCCG +GAGCGA +GAGCGT +GAGCGC +GAGCGG +GAGGAA +GAGGAT +GAGGAC +GAGGAG +GAGGTA +GAGGTT +GAGGTC +GAGGTG +GAGGCA +GAGGCT +GAGGCC +GAGGCG +GAGGGA +GAGGGT +GAGGGC +GAGGGG +GTAAAA +GTAAAT +GTAAAC +GTAAAG +GTAATA +GTAATT +GTAATC +GTAATG +GTAACA +GTAACT +GTAACC +GTAACG +GTAAGA +GTAAGT +GTAAGC +GTAAGG +GTATAA +GTATAT +GTATAC +GTATAG +GTATTA +GTATTT +GTATTC +GTATTG +GTATCA +GTATCT +GTATCC +GTATCG +GTATGA +GTATGT +GTATGC +GTATGG +GTACAA +GTACAT +GTACAC +GTACAG +GTACTA +GTACTT +GTACTC +GTACTG +GTACCA +GTACCT +GTACCC +GTACCG +GTACGA +GTACGT +GTACGC +GTACGG +GTAGAA +GTAGAT +GTAGAC +GTAGAG +GTAGTA +GTAGTT +GTAGTC +GTAGTG +GTAGCA +GTAGCT +GTAGCC +GTAGCG +GTAGGA +GTAGGT +GTAGGC +GTAGGG +GTTAAA +GTTAAT +GTTAAC +GTTAAG +GTTATA +GTTATT +GTTATC +GTTATG +GTTACA +GTTACT +GTTACC +GTTACG +GTTAGA +GTTAGT +GTTAGC +GTTAGG +GTTTAA +GTTTAT +GTTTAC +GTTTAG +GTTTTA +GTTTTT +GTTTTC +GTTTTG +GTTTCA +GTTTCT +GTTTCC +GTTTCG +GTTTGA +GTTTGT +GTTTGC +GTTTGG +GTTCAA +GTTCAT +GTTCAC +GTTCAG +GTTCTA +GTTCTT +GTTCTC +GTTCTG +GTTCCA +GTTCCT +GTTCCC +GTTCCG +GTTCGA +GTTCGT +GTTCGC +GTTCGG +GTTGAA +GTTGAT +GTTGAC +GTTGAG +GTTGTA +GTTGTT +GTTGTC +GTTGTG +GTTGCA +GTTGCT +GTTGCC +GTTGCG +GTTGGA +GTTGGT +GTTGGC +GTTGGG +GTCAAA +GTCAAT +GTCAAC +GTCAAG +GTCATA +GTCATT +GTCATC +GTCATG +GTCACA +GTCACT +GTCACC +GTCACG +GTCAGA +GTCAGT +GTCAGC +GTCAGG +GTCTAA +GTCTAT +GTCTAC +GTCTAG +GTCTTA +GTCTTT +GTCTTC +GTCTTG +GTCTCA +GTCTCT +GTCTCC +GTCTCG +GTCTGA +GTCTGT +GTCTGC +GTCTGG +GTCCAA +GTCCAT +GTCCAC +GTCCAG +GTCCTA +GTCCTT +GTCCTC +GTCCTG +GTCCCA +GTCCCT +GTCCCC +GTCCCG +GTCCGA +GTCCGT +GTCCGC +GTCCGG +GTCGAA +GTCGAT +GTCGAC +GTCGAG +GTCGTA +GTCGTT +GTCGTC +GTCGTG +GTCGCA +GTCGCT +GTCGCC +GTCGCG +GTCGGA +GTCGGT +GTCGGC +GTCGGG +GTGAAA +GTGAAT +GTGAAC +GTGAAG +GTGATA +GTGATT +GTGATC +GTGATG +GTGACA +GTGACT +GTGACC +GTGACG +GTGAGA +GTGAGT +GTGAGC +GTGAGG +GTGTAA +GTGTAT +GTGTAC +GTGTAG +GTGTTA +GTGTTT +GTGTTC +GTGTTG +GTGTCA +GTGTCT +GTGTCC +GTGTCG +GTGTGA +GTGTGT +GTGTGC +GTGTGG +GTGCAA +GTGCAT +GTGCAC +GTGCAG +GTGCTA +GTGCTT +GTGCTC +GTGCTG +GTGCCA +GTGCCT +GTGCCC +GTGCCG +GTGCGA +GTGCGT +GTGCGC +GTGCGG +GTGGAA +GTGGAT +GTGGAC +GTGGAG +GTGGTA +GTGGTT +GTGGTC +GTGGTG +GTGGCA +GTGGCT +GTGGCC +GTGGCG +GTGGGA +GTGGGT +GTGGGC +GTGGGG +GCAAAA +GCAAAT +GCAAAC +GCAAAG +GCAATA +GCAATT +GCAATC +GCAATG +GCAACA +GCAACT +GCAACC +GCAACG +GCAAGA +GCAAGT +GCAAGC +GCAAGG +GCATAA +GCATAT +GCATAC +GCATAG +GCATTA +GCATTT +GCATTC +GCATTG +GCATCA +GCATCT +GCATCC +GCATCG +GCATGA +GCATGT +GCATGC +GCATGG +GCACAA +GCACAT +GCACAC +GCACAG +GCACTA +GCACTT +GCACTC +GCACTG +GCACCA +GCACCT +GCACCC +GCACCG +GCACGA +GCACGT +GCACGC +GCACGG +GCAGAA +GCAGAT +GCAGAC +GCAGAG +GCAGTA +GCAGTT +GCAGTC +GCAGTG +GCAGCA +GCAGCT +GCAGCC +GCAGCG +GCAGGA +GCAGGT +GCAGGC +GCAGGG +GCTAAA +GCTAAT +GCTAAC +GCTAAG +GCTATA +GCTATT +GCTATC +GCTATG +GCTACA +GCTACT +GCTACC +GCTACG +GCTAGA +GCTAGT +GCTAGC +GCTAGG +GCTTAA +GCTTAT +GCTTAC +GCTTAG +GCTTTA +GCTTTT +GCTTTC +GCTTTG +GCTTCA +GCTTCT +GCTTCC +GCTTCG +GCTTGA +GCTTGT +GCTTGC +GCTTGG +GCTCAA +GCTCAT +GCTCAC +GCTCAG +GCTCTA +GCTCTT +GCTCTC +GCTCTG +GCTCCA +GCTCCT +GCTCCC +GCTCCG +GCTCGA +GCTCGT +GCTCGC +GCTCGG +GCTGAA +GCTGAT +GCTGAC +GCTGAG +GCTGTA +GCTGTT +GCTGTC +GCTGTG +GCTGCA +GCTGCT +GCTGCC +GCTGCG +GCTGGA +GCTGGT +GCTGGC +GCTGGG +GCCAAA +GCCAAT +GCCAAC +GCCAAG +GCCATA +GCCATT +GCCATC +GCCATG +GCCACA +GCCACT +GCCACC +GCCACG +GCCAGA +GCCAGT +GCCAGC +GCCAGG +GCCTAA +GCCTAT +GCCTAC +GCCTAG +GCCTTA +GCCTTT +GCCTTC +GCCTTG +GCCTCA +GCCTCT +GCCTCC +GCCTCG +GCCTGA +GCCTGT +GCCTGC +GCCTGG +GCCCAA +GCCCAT +GCCCAC +GCCCAG +GCCCTA +GCCCTT +GCCCTC +GCCCTG +GCCCCA +GCCCCT +GCCCCC +GCCCCG +GCCCGA +GCCCGT +GCCCGC +GCCCGG +GCCGAA +GCCGAT +GCCGAC +GCCGAG +GCCGTA +GCCGTT +GCCGTC +GCCGTG +GCCGCA +GCCGCT +GCCGCC +GCCGCG +GCCGGA +GCCGGT +GCCGGC +GCCGGG +GCGAAA +GCGAAT +GCGAAC +GCGAAG +GCGATA +GCGATT +GCGATC +GCGATG +GCGACA +GCGACT +GCGACC +GCGACG +GCGAGA +GCGAGT +GCGAGC +GCGAGG +GCGTAA +GCGTAT +GCGTAC +GCGTAG +GCGTTA +GCGTTT +GCGTTC +GCGTTG +GCGTCA +GCGTCT +GCGTCC +GCGTCG +GCGTGA +GCGTGT +GCGTGC +GCGTGG +GCGCAA +GCGCAT +GCGCAC +GCGCAG +GCGCTA +GCGCTT +GCGCTC +GCGCTG +GCGCCA +GCGCCT +GCGCCC +GCGCCG +GCGCGA +GCGCGT +GCGCGC +GCGCGG +GCGGAA +GCGGAT +GCGGAC +GCGGAG +GCGGTA +GCGGTT +GCGGTC +GCGGTG +GCGGCA +GCGGCT +GCGGCC +GCGGCG +GCGGGA +GCGGGT +GCGGGC +GCGGGG +GGAAAA +GGAAAT +GGAAAC +GGAAAG +GGAATA +GGAATT +GGAATC +GGAATG +GGAACA +GGAACT +GGAACC +GGAACG +GGAAGA +GGAAGT +GGAAGC +GGAAGG +GGATAA +GGATAT +GGATAC +GGATAG +GGATTA +GGATTT +GGATTC +GGATTG +GGATCA +GGATCT +GGATCC +GGATCG +GGATGA +GGATGT +GGATGC +GGATGG +GGACAA +GGACAT +GGACAC +GGACAG +GGACTA +GGACTT +GGACTC +GGACTG +GGACCA +GGACCT +GGACCC +GGACCG +GGACGA +GGACGT +GGACGC +GGACGG +GGAGAA +GGAGAT +GGAGAC +GGAGAG +GGAGTA +GGAGTT +GGAGTC +GGAGTG +GGAGCA +GGAGCT +GGAGCC +GGAGCG +GGAGGA +GGAGGT +GGAGGC +GGAGGG +GGTAAA +GGTAAT +GGTAAC +GGTAAG +GGTATA +GGTATT +GGTATC +GGTATG +GGTACA +GGTACT +GGTACC +GGTACG +GGTAGA +GGTAGT +GGTAGC +GGTAGG +GGTTAA +GGTTAT +GGTTAC +GGTTAG +GGTTTA +GGTTTT +GGTTTC +GGTTTG +GGTTCA +GGTTCT +GGTTCC +GGTTCG +GGTTGA +GGTTGT +GGTTGC +GGTTGG +GGTCAA +GGTCAT +GGTCAC +GGTCAG +GGTCTA +GGTCTT +GGTCTC +GGTCTG +GGTCCA +GGTCCT +GGTCCC +GGTCCG +GGTCGA +GGTCGT +GGTCGC +GGTCGG +GGTGAA +GGTGAT +GGTGAC +GGTGAG +GGTGTA +GGTGTT +GGTGTC +GGTGTG +GGTGCA +GGTGCT +GGTGCC +GGTGCG +GGTGGA +GGTGGT +GGTGGC +GGTGGG +GGCAAA +GGCAAT +GGCAAC +GGCAAG +GGCATA +GGCATT +GGCATC +GGCATG +GGCACA +GGCACT +GGCACC +GGCACG +GGCAGA +GGCAGT +GGCAGC +GGCAGG +GGCTAA +GGCTAT +GGCTAC +GGCTAG +GGCTTA +GGCTTT +GGCTTC +GGCTTG +GGCTCA +GGCTCT +GGCTCC +GGCTCG +GGCTGA +GGCTGT +GGCTGC +GGCTGG +GGCCAA +GGCCAT +GGCCAC +GGCCAG +GGCCTA +GGCCTT +GGCCTC +GGCCTG +GGCCCA +GGCCCT +GGCCCC +GGCCCG +GGCCGA +GGCCGT +GGCCGC +GGCCGG +GGCGAA +GGCGAT +GGCGAC +GGCGAG +GGCGTA +GGCGTT +GGCGTC +GGCGTG +GGCGCA +GGCGCT +GGCGCC +GGCGCG +GGCGGA +GGCGGT +GGCGGC +GGCGGG +GGGAAA +GGGAAT +GGGAAC +GGGAAG +GGGATA +GGGATT +GGGATC +GGGATG +GGGACA +GGGACT +GGGACC +GGGACG +GGGAGA +GGGAGT +GGGAGC +GGGAGG +GGGTAA +GGGTAT +GGGTAC +GGGTAG +GGGTTA +GGGTTT +GGGTTC +GGGTTG +GGGTCA +GGGTCT +GGGTCC +GGGTCG +GGGTGA +GGGTGT +GGGTGC +GGGTGG +GGGCAA +GGGCAT +GGGCAC +GGGCAG +GGGCTA +GGGCTT +GGGCTC +GGGCTG +GGGCCA +GGGCCT +GGGCCC +GGGCCG +GGGCGA +GGGCGT +GGGCGC +GGGCGG +GGGGAA +GGGGAT +GGGGAC +GGGGAG +GGGGTA +GGGGTT +GGGGTC +GGGGTG +GGGGCA +GGGGCT +GGGGCC +GGGGCG +GGGGGA +GGGGGT +GGGGGC +GGGGGG diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/README.md b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..500fd976d92b217455f181bc98c77e612d9a18d9 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/README.md @@ -0,0 +1,202 @@ +--- +base_model: fabihamakhdoomi/TinyDNABERT +library_name: peft +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.12.0 \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/adapter_config.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..28c871c93775eedfaf266f00da8c02f96bc02f29 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/adapter_config.json @@ -0,0 +1,42 @@ +{ + "alpha_pattern": {}, + "auto_mapping": null, + "base_model_name_or_path": "fabihamakhdoomi/TinyDNABERT", + "bias": "none", + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 16, + "lora_dropout": 0.2, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": [ + "classifier", + "score" + ], + "peft_type": "LORA", + "r": 4, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "bert.encoder.layer.1.attention.self.query", + "bert.encoder.layer.0.output.dense", + "bert.encoder.layer.1.output.dense", + "bert.encoder.layer.0.intermediate.dense", + "bert.encoder.layer.1.intermediate.dense", + "bert.encoder.layer.0.attention.output.dense", + "bert.encoder.layer.1.attention.output.dense", + "bert.encoder.layer.0.attention.self.query", + "bert.encoder.layer.0.attention.self.value", + "bert.encoder.layer.1.attention.self.key", + "bert.encoder.layer.1.attention.self.value", + "bert.encoder.layer.0.attention.self.key" + ], + "task_type": "SEQ_CLS", + "use_dora": false, + "use_rslora": false +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/adapter_model.safetensors b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/adapter_model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..657967c0bf0628f3e21be90952a43ec415c4def8 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/adapter_model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:059d8930d00d0b5c382f801d32bcf31044c0693be948b80c3d8d7c5276eea728 +size 78208 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/optimizer.pt b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..1912a25c21bb2ea8b9791755cbdbfdb60f53992d --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7af16bd81d8b81b1968267f705623c61a387921ee4729bbb580bbbe4278eb032 +size 165226 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/rng_state.pth b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..7d59ecb6c9b7d6ec50b060f3843e47cd8ba9b667 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:84106ec8b7c96df034e207cd31f8dd32759188fe95dd21501975d40030684bf4 +size 14244 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/scheduler.pt b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..f3ebc22eebe7fac9f20742118527e91760861e6d --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7e153e480933b4bdcc7d98226c2fdfde72db931809685729c8748251ae3ecae9 +size 1064 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/special_tokens_map.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..a8b3208c2884c4efb86e49300fdd3dc877220cdf --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/special_tokens_map.json @@ -0,0 +1,7 @@ +{ + "cls_token": "[CLS]", + "mask_token": "[MASK]", + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/tokenizer.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..ebd782de6963b9032ce9374cfe30f5275be1a30b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/tokenizer.json @@ -0,0 +1,4265 @@ +{ + "version": "1.0", + "truncation": { + "direction": "Right", + "max_length": 512, + "strategy": "LongestFirst", + "stride": 0 + }, + "padding": { + "strategy": { + "Fixed": 512 + }, + "direction": "Right", + "pad_to_multiple_of": null, + "pad_id": 0, + "pad_type_id": 0, + "pad_token": "[PAD]" + }, + "added_tokens": [ + { + "id": 0, + "content": "[PAD]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 1, + "content": "[UNK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 2, + "content": "[CLS]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 3, + "content": "[SEP]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 4, + "content": "[MASK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + } + ], + "normalizer": { + "type": "BertNormalizer", + "clean_text": true, + "handle_chinese_chars": true, + "strip_accents": null, + "lowercase": false + }, + "pre_tokenizer": { + "type": "BertPreTokenizer" + }, + "post_processor": { + "type": "TemplateProcessing", + "single": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + } + ], + "pair": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "B", + "type_id": 1 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 1 + } + } + ], + "special_tokens": { + "[CLS]": { + "id": "[CLS]", + "ids": [ + 2 + ], + "tokens": [ + "[CLS]" + ] + }, + "[SEP]": { + "id": "[SEP]", + "ids": [ + 3 + ], + "tokens": [ + "[SEP]" + ] + } + } + }, + "decoder": { + "type": "WordPiece", + "prefix": "##", + "cleanup": true + }, + "model": { + "type": "WordPiece", + "unk_token": "[UNK]", + "continuing_subword_prefix": "##", + "max_input_chars_per_word": 100, + "vocab": { + "[PAD]": 0, + "[UNK]": 1, + "[CLS]": 2, + "[SEP]": 3, + "[MASK]": 4, + "AAAAAA": 5, + "AAAAAT": 6, + "AAAAAC": 7, + "AAAAAG": 8, + "AAAATA": 9, + "AAAATT": 10, + "AAAATC": 11, + "AAAATG": 12, + "AAAACA": 13, + "AAAACT": 14, + "AAAACC": 15, + "AAAACG": 16, + "AAAAGA": 17, + "AAAAGT": 18, + "AAAAGC": 19, + "AAAAGG": 20, + "AAATAA": 21, + "AAATAT": 22, + "AAATAC": 23, + "AAATAG": 24, + "AAATTA": 25, + "AAATTT": 26, + "AAATTC": 27, + "AAATTG": 28, + "AAATCA": 29, + "AAATCT": 30, + "AAATCC": 31, + "AAATCG": 32, + "AAATGA": 33, + "AAATGT": 34, + "AAATGC": 35, + "AAATGG": 36, + "AAACAA": 37, + "AAACAT": 38, + "AAACAC": 39, + "AAACAG": 40, + "AAACTA": 41, + "AAACTT": 42, + "AAACTC": 43, + "AAACTG": 44, + "AAACCA": 45, + "AAACCT": 46, + "AAACCC": 47, + "AAACCG": 48, + "AAACGA": 49, + "AAACGT": 50, + "AAACGC": 51, + "AAACGG": 52, + "AAAGAA": 53, + "AAAGAT": 54, + "AAAGAC": 55, + "AAAGAG": 56, + "AAAGTA": 57, + "AAAGTT": 58, + "AAAGTC": 59, + "AAAGTG": 60, + "AAAGCA": 61, + "AAAGCT": 62, + "AAAGCC": 63, + "AAAGCG": 64, + "AAAGGA": 65, + "AAAGGT": 66, + "AAAGGC": 67, + "AAAGGG": 68, + "AATAAA": 69, + "AATAAT": 70, + "AATAAC": 71, + "AATAAG": 72, + "AATATA": 73, + "AATATT": 74, + "AATATC": 75, + "AATATG": 76, + "AATACA": 77, + "AATACT": 78, + "AATACC": 79, + "AATACG": 80, + "AATAGA": 81, + "AATAGT": 82, + "AATAGC": 83, + "AATAGG": 84, + "AATTAA": 85, + "AATTAT": 86, + "AATTAC": 87, + "AATTAG": 88, + "AATTTA": 89, + "AATTTT": 90, + "AATTTC": 91, + "AATTTG": 92, + "AATTCA": 93, + "AATTCT": 94, + "AATTCC": 95, + "AATTCG": 96, + "AATTGA": 97, + "AATTGT": 98, + "AATTGC": 99, + "AATTGG": 100, + "AATCAA": 101, + "AATCAT": 102, + "AATCAC": 103, + "AATCAG": 104, + "AATCTA": 105, + "AATCTT": 106, + "AATCTC": 107, + "AATCTG": 108, + "AATCCA": 109, + "AATCCT": 110, + "AATCCC": 111, + "AATCCG": 112, + "AATCGA": 113, + "AATCGT": 114, + "AATCGC": 115, + "AATCGG": 116, + "AATGAA": 117, + "AATGAT": 118, + "AATGAC": 119, + "AATGAG": 120, + "AATGTA": 121, + "AATGTT": 122, + "AATGTC": 123, + "AATGTG": 124, + "AATGCA": 125, + "AATGCT": 126, + "AATGCC": 127, + "AATGCG": 128, + "AATGGA": 129, + "AATGGT": 130, + "AATGGC": 131, + "AATGGG": 132, + "AACAAA": 133, + "AACAAT": 134, + "AACAAC": 135, + "AACAAG": 136, + "AACATA": 137, + "AACATT": 138, + "AACATC": 139, + "AACATG": 140, + "AACACA": 141, + "AACACT": 142, + "AACACC": 143, + "AACACG": 144, + "AACAGA": 145, + "AACAGT": 146, + "AACAGC": 147, + "AACAGG": 148, + "AACTAA": 149, + "AACTAT": 150, + "AACTAC": 151, + "AACTAG": 152, + "AACTTA": 153, + "AACTTT": 154, + "AACTTC": 155, + "AACTTG": 156, + "AACTCA": 157, + "AACTCT": 158, + "AACTCC": 159, + "AACTCG": 160, + "AACTGA": 161, + "AACTGT": 162, + "AACTGC": 163, + "AACTGG": 164, + "AACCAA": 165, + "AACCAT": 166, + "AACCAC": 167, + "AACCAG": 168, + "AACCTA": 169, + "AACCTT": 170, + "AACCTC": 171, + "AACCTG": 172, + "AACCCA": 173, + "AACCCT": 174, + "AACCCC": 175, + "AACCCG": 176, + "AACCGA": 177, + "AACCGT": 178, + "AACCGC": 179, + "AACCGG": 180, + "AACGAA": 181, + "AACGAT": 182, + "AACGAC": 183, + "AACGAG": 184, + "AACGTA": 185, + "AACGTT": 186, + "AACGTC": 187, + "AACGTG": 188, + "AACGCA": 189, + "AACGCT": 190, + "AACGCC": 191, + "AACGCG": 192, + "AACGGA": 193, + "AACGGT": 194, + "AACGGC": 195, + "AACGGG": 196, + "AAGAAA": 197, + "AAGAAT": 198, + "AAGAAC": 199, + "AAGAAG": 200, + "AAGATA": 201, + "AAGATT": 202, + "AAGATC": 203, + "AAGATG": 204, + "AAGACA": 205, + "AAGACT": 206, + "AAGACC": 207, + "AAGACG": 208, + "AAGAGA": 209, + "AAGAGT": 210, + "AAGAGC": 211, + "AAGAGG": 212, + "AAGTAA": 213, + "AAGTAT": 214, + "AAGTAC": 215, + "AAGTAG": 216, + "AAGTTA": 217, + "AAGTTT": 218, + "AAGTTC": 219, + "AAGTTG": 220, + "AAGTCA": 221, + "AAGTCT": 222, + "AAGTCC": 223, + "AAGTCG": 224, + "AAGTGA": 225, + "AAGTGT": 226, + "AAGTGC": 227, + "AAGTGG": 228, + "AAGCAA": 229, + "AAGCAT": 230, + "AAGCAC": 231, + "AAGCAG": 232, + "AAGCTA": 233, + "AAGCTT": 234, + "AAGCTC": 235, + "AAGCTG": 236, + "AAGCCA": 237, + "AAGCCT": 238, + "AAGCCC": 239, + "AAGCCG": 240, + "AAGCGA": 241, + "AAGCGT": 242, + "AAGCGC": 243, + "AAGCGG": 244, + "AAGGAA": 245, + "AAGGAT": 246, + "AAGGAC": 247, + "AAGGAG": 248, + "AAGGTA": 249, + "AAGGTT": 250, + "AAGGTC": 251, + "AAGGTG": 252, + "AAGGCA": 253, + "AAGGCT": 254, + "AAGGCC": 255, + "AAGGCG": 256, + "AAGGGA": 257, + "AAGGGT": 258, + "AAGGGC": 259, + "AAGGGG": 260, + "ATAAAA": 261, + "ATAAAT": 262, + "ATAAAC": 263, + "ATAAAG": 264, + "ATAATA": 265, + "ATAATT": 266, + "ATAATC": 267, + "ATAATG": 268, + "ATAACA": 269, + "ATAACT": 270, + "ATAACC": 271, + "ATAACG": 272, + "ATAAGA": 273, + "ATAAGT": 274, + "ATAAGC": 275, + "ATAAGG": 276, + "ATATAA": 277, + "ATATAT": 278, + "ATATAC": 279, + "ATATAG": 280, + "ATATTA": 281, + "ATATTT": 282, + "ATATTC": 283, + "ATATTG": 284, + "ATATCA": 285, + "ATATCT": 286, + "ATATCC": 287, + "ATATCG": 288, + "ATATGA": 289, + "ATATGT": 290, + "ATATGC": 291, + "ATATGG": 292, + "ATACAA": 293, + "ATACAT": 294, + "ATACAC": 295, + "ATACAG": 296, + "ATACTA": 297, + "ATACTT": 298, + "ATACTC": 299, + "ATACTG": 300, + "ATACCA": 301, + "ATACCT": 302, + "ATACCC": 303, + "ATACCG": 304, + "ATACGA": 305, + "ATACGT": 306, + "ATACGC": 307, + "ATACGG": 308, + "ATAGAA": 309, + "ATAGAT": 310, + "ATAGAC": 311, + "ATAGAG": 312, + "ATAGTA": 313, + "ATAGTT": 314, + "ATAGTC": 315, + "ATAGTG": 316, + "ATAGCA": 317, + "ATAGCT": 318, + "ATAGCC": 319, + "ATAGCG": 320, + "ATAGGA": 321, + "ATAGGT": 322, + "ATAGGC": 323, + "ATAGGG": 324, + "ATTAAA": 325, + "ATTAAT": 326, + "ATTAAC": 327, + "ATTAAG": 328, + "ATTATA": 329, + "ATTATT": 330, + "ATTATC": 331, + "ATTATG": 332, + "ATTACA": 333, + "ATTACT": 334, + "ATTACC": 335, + "ATTACG": 336, + "ATTAGA": 337, + "ATTAGT": 338, + "ATTAGC": 339, + "ATTAGG": 340, + "ATTTAA": 341, + "ATTTAT": 342, + "ATTTAC": 343, + "ATTTAG": 344, + "ATTTTA": 345, + "ATTTTT": 346, + "ATTTTC": 347, + "ATTTTG": 348, + "ATTTCA": 349, + "ATTTCT": 350, + "ATTTCC": 351, + "ATTTCG": 352, + "ATTTGA": 353, + "ATTTGT": 354, + "ATTTGC": 355, + "ATTTGG": 356, + "ATTCAA": 357, + "ATTCAT": 358, + "ATTCAC": 359, + "ATTCAG": 360, + "ATTCTA": 361, + "ATTCTT": 362, + "ATTCTC": 363, + "ATTCTG": 364, + "ATTCCA": 365, + "ATTCCT": 366, + "ATTCCC": 367, + "ATTCCG": 368, + "ATTCGA": 369, + "ATTCGT": 370, + "ATTCGC": 371, + "ATTCGG": 372, + "ATTGAA": 373, + "ATTGAT": 374, + "ATTGAC": 375, + "ATTGAG": 376, + "ATTGTA": 377, + "ATTGTT": 378, + "ATTGTC": 379, + "ATTGTG": 380, + "ATTGCA": 381, + "ATTGCT": 382, + "ATTGCC": 383, + "ATTGCG": 384, + "ATTGGA": 385, + "ATTGGT": 386, + "ATTGGC": 387, + "ATTGGG": 388, + "ATCAAA": 389, + "ATCAAT": 390, + "ATCAAC": 391, + "ATCAAG": 392, + "ATCATA": 393, + "ATCATT": 394, + "ATCATC": 395, + "ATCATG": 396, + "ATCACA": 397, + "ATCACT": 398, + "ATCACC": 399, + "ATCACG": 400, + "ATCAGA": 401, + "ATCAGT": 402, + "ATCAGC": 403, + "ATCAGG": 404, + "ATCTAA": 405, + "ATCTAT": 406, + "ATCTAC": 407, + "ATCTAG": 408, + "ATCTTA": 409, + "ATCTTT": 410, + "ATCTTC": 411, + "ATCTTG": 412, + "ATCTCA": 413, + "ATCTCT": 414, + "ATCTCC": 415, + "ATCTCG": 416, + "ATCTGA": 417, + "ATCTGT": 418, + "ATCTGC": 419, + "ATCTGG": 420, + "ATCCAA": 421, + "ATCCAT": 422, + "ATCCAC": 423, + "ATCCAG": 424, + "ATCCTA": 425, + "ATCCTT": 426, + "ATCCTC": 427, + "ATCCTG": 428, + "ATCCCA": 429, + "ATCCCT": 430, + "ATCCCC": 431, + "ATCCCG": 432, + "ATCCGA": 433, + "ATCCGT": 434, + "ATCCGC": 435, + "ATCCGG": 436, + "ATCGAA": 437, + "ATCGAT": 438, + "ATCGAC": 439, + "ATCGAG": 440, + "ATCGTA": 441, + "ATCGTT": 442, + "ATCGTC": 443, + "ATCGTG": 444, + "ATCGCA": 445, + "ATCGCT": 446, + "ATCGCC": 447, + "ATCGCG": 448, + "ATCGGA": 449, + "ATCGGT": 450, + "ATCGGC": 451, + "ATCGGG": 452, + "ATGAAA": 453, + "ATGAAT": 454, + "ATGAAC": 455, + "ATGAAG": 456, + "ATGATA": 457, + "ATGATT": 458, + "ATGATC": 459, + "ATGATG": 460, + "ATGACA": 461, + "ATGACT": 462, + "ATGACC": 463, + "ATGACG": 464, + "ATGAGA": 465, + "ATGAGT": 466, + "ATGAGC": 467, + "ATGAGG": 468, + "ATGTAA": 469, + "ATGTAT": 470, + "ATGTAC": 471, + "ATGTAG": 472, + "ATGTTA": 473, + "ATGTTT": 474, + "ATGTTC": 475, + "ATGTTG": 476, + "ATGTCA": 477, + "ATGTCT": 478, + "ATGTCC": 479, + "ATGTCG": 480, + "ATGTGA": 481, + "ATGTGT": 482, + "ATGTGC": 483, + "ATGTGG": 484, + "ATGCAA": 485, + "ATGCAT": 486, + "ATGCAC": 487, + "ATGCAG": 488, + "ATGCTA": 489, + "ATGCTT": 490, + "ATGCTC": 491, + "ATGCTG": 492, + "ATGCCA": 493, + "ATGCCT": 494, + "ATGCCC": 495, + "ATGCCG": 496, + "ATGCGA": 497, + "ATGCGT": 498, + "ATGCGC": 499, + "ATGCGG": 500, + "ATGGAA": 501, + "ATGGAT": 502, + "ATGGAC": 503, + "ATGGAG": 504, + "ATGGTA": 505, + "ATGGTT": 506, + "ATGGTC": 507, + "ATGGTG": 508, + "ATGGCA": 509, + "ATGGCT": 510, + "ATGGCC": 511, + "ATGGCG": 512, + "ATGGGA": 513, + "ATGGGT": 514, + "ATGGGC": 515, + "ATGGGG": 516, + "ACAAAA": 517, + "ACAAAT": 518, + "ACAAAC": 519, + "ACAAAG": 520, + "ACAATA": 521, + "ACAATT": 522, + "ACAATC": 523, + "ACAATG": 524, + "ACAACA": 525, + "ACAACT": 526, + "ACAACC": 527, + "ACAACG": 528, + "ACAAGA": 529, + "ACAAGT": 530, + "ACAAGC": 531, + "ACAAGG": 532, + "ACATAA": 533, + "ACATAT": 534, + "ACATAC": 535, + "ACATAG": 536, + "ACATTA": 537, + "ACATTT": 538, + "ACATTC": 539, + "ACATTG": 540, + "ACATCA": 541, + "ACATCT": 542, + "ACATCC": 543, + "ACATCG": 544, + "ACATGA": 545, + "ACATGT": 546, + "ACATGC": 547, + "ACATGG": 548, + "ACACAA": 549, + "ACACAT": 550, + "ACACAC": 551, + "ACACAG": 552, + "ACACTA": 553, + "ACACTT": 554, + "ACACTC": 555, + "ACACTG": 556, + "ACACCA": 557, + "ACACCT": 558, + "ACACCC": 559, + "ACACCG": 560, + "ACACGA": 561, + "ACACGT": 562, + "ACACGC": 563, + "ACACGG": 564, + "ACAGAA": 565, + "ACAGAT": 566, + "ACAGAC": 567, + "ACAGAG": 568, + "ACAGTA": 569, + "ACAGTT": 570, + "ACAGTC": 571, + "ACAGTG": 572, + "ACAGCA": 573, + "ACAGCT": 574, + "ACAGCC": 575, + "ACAGCG": 576, + "ACAGGA": 577, + "ACAGGT": 578, + "ACAGGC": 579, + "ACAGGG": 580, + "ACTAAA": 581, + "ACTAAT": 582, + "ACTAAC": 583, + "ACTAAG": 584, + "ACTATA": 585, + "ACTATT": 586, + "ACTATC": 587, + "ACTATG": 588, + "ACTACA": 589, + "ACTACT": 590, + "ACTACC": 591, + "ACTACG": 592, + "ACTAGA": 593, + "ACTAGT": 594, + "ACTAGC": 595, + "ACTAGG": 596, + "ACTTAA": 597, + "ACTTAT": 598, + "ACTTAC": 599, + "ACTTAG": 600, + "ACTTTA": 601, + "ACTTTT": 602, + "ACTTTC": 603, + "ACTTTG": 604, + "ACTTCA": 605, + "ACTTCT": 606, + "ACTTCC": 607, + "ACTTCG": 608, + "ACTTGA": 609, + "ACTTGT": 610, + "ACTTGC": 611, + "ACTTGG": 612, + "ACTCAA": 613, + "ACTCAT": 614, + "ACTCAC": 615, + "ACTCAG": 616, + "ACTCTA": 617, + "ACTCTT": 618, + "ACTCTC": 619, + "ACTCTG": 620, + "ACTCCA": 621, + "ACTCCT": 622, + "ACTCCC": 623, + "ACTCCG": 624, + "ACTCGA": 625, + "ACTCGT": 626, + "ACTCGC": 627, + "ACTCGG": 628, + "ACTGAA": 629, + "ACTGAT": 630, + "ACTGAC": 631, + "ACTGAG": 632, + "ACTGTA": 633, + "ACTGTT": 634, + "ACTGTC": 635, + "ACTGTG": 636, + "ACTGCA": 637, + "ACTGCT": 638, + "ACTGCC": 639, + "ACTGCG": 640, + "ACTGGA": 641, + "ACTGGT": 642, + "ACTGGC": 643, + "ACTGGG": 644, + "ACCAAA": 645, + "ACCAAT": 646, + "ACCAAC": 647, + "ACCAAG": 648, + "ACCATA": 649, + "ACCATT": 650, + "ACCATC": 651, + "ACCATG": 652, + "ACCACA": 653, + "ACCACT": 654, + "ACCACC": 655, + "ACCACG": 656, + "ACCAGA": 657, + "ACCAGT": 658, + "ACCAGC": 659, + "ACCAGG": 660, + "ACCTAA": 661, + "ACCTAT": 662, + "ACCTAC": 663, + "ACCTAG": 664, + "ACCTTA": 665, + "ACCTTT": 666, + "ACCTTC": 667, + "ACCTTG": 668, + "ACCTCA": 669, + "ACCTCT": 670, + "ACCTCC": 671, + "ACCTCG": 672, + "ACCTGA": 673, + "ACCTGT": 674, + "ACCTGC": 675, + "ACCTGG": 676, + "ACCCAA": 677, + "ACCCAT": 678, + "ACCCAC": 679, + "ACCCAG": 680, + "ACCCTA": 681, + "ACCCTT": 682, + "ACCCTC": 683, + "ACCCTG": 684, + "ACCCCA": 685, + "ACCCCT": 686, + "ACCCCC": 687, + "ACCCCG": 688, + "ACCCGA": 689, + "ACCCGT": 690, + "ACCCGC": 691, + "ACCCGG": 692, + "ACCGAA": 693, + "ACCGAT": 694, + "ACCGAC": 695, + "ACCGAG": 696, + "ACCGTA": 697, + "ACCGTT": 698, + "ACCGTC": 699, + "ACCGTG": 700, + "ACCGCA": 701, + "ACCGCT": 702, + "ACCGCC": 703, + "ACCGCG": 704, + "ACCGGA": 705, + "ACCGGT": 706, + "ACCGGC": 707, + "ACCGGG": 708, + "ACGAAA": 709, + "ACGAAT": 710, + "ACGAAC": 711, + "ACGAAG": 712, + "ACGATA": 713, + "ACGATT": 714, + "ACGATC": 715, + "ACGATG": 716, + "ACGACA": 717, + "ACGACT": 718, + "ACGACC": 719, + "ACGACG": 720, + "ACGAGA": 721, + "ACGAGT": 722, + "ACGAGC": 723, + "ACGAGG": 724, + "ACGTAA": 725, + "ACGTAT": 726, + "ACGTAC": 727, + "ACGTAG": 728, + "ACGTTA": 729, + "ACGTTT": 730, + "ACGTTC": 731, + "ACGTTG": 732, + "ACGTCA": 733, + "ACGTCT": 734, + "ACGTCC": 735, + "ACGTCG": 736, + "ACGTGA": 737, + "ACGTGT": 738, + "ACGTGC": 739, + "ACGTGG": 740, + "ACGCAA": 741, + "ACGCAT": 742, + "ACGCAC": 743, + "ACGCAG": 744, + "ACGCTA": 745, + "ACGCTT": 746, + "ACGCTC": 747, + "ACGCTG": 748, + "ACGCCA": 749, + "ACGCCT": 750, + "ACGCCC": 751, + "ACGCCG": 752, + "ACGCGA": 753, + "ACGCGT": 754, + "ACGCGC": 755, + "ACGCGG": 756, + "ACGGAA": 757, + "ACGGAT": 758, + "ACGGAC": 759, + "ACGGAG": 760, + "ACGGTA": 761, + "ACGGTT": 762, + "ACGGTC": 763, + "ACGGTG": 764, + "ACGGCA": 765, + "ACGGCT": 766, + "ACGGCC": 767, + "ACGGCG": 768, + "ACGGGA": 769, + "ACGGGT": 770, + "ACGGGC": 771, + "ACGGGG": 772, + "AGAAAA": 773, + "AGAAAT": 774, + "AGAAAC": 775, + "AGAAAG": 776, + "AGAATA": 777, + "AGAATT": 778, + "AGAATC": 779, + "AGAATG": 780, + "AGAACA": 781, + "AGAACT": 782, + "AGAACC": 783, + "AGAACG": 784, + "AGAAGA": 785, + "AGAAGT": 786, + "AGAAGC": 787, + "AGAAGG": 788, + "AGATAA": 789, + "AGATAT": 790, + "AGATAC": 791, + "AGATAG": 792, + "AGATTA": 793, + "AGATTT": 794, + "AGATTC": 795, + "AGATTG": 796, + "AGATCA": 797, + "AGATCT": 798, + "AGATCC": 799, + "AGATCG": 800, + "AGATGA": 801, + "AGATGT": 802, + "AGATGC": 803, + "AGATGG": 804, + "AGACAA": 805, + "AGACAT": 806, + "AGACAC": 807, + "AGACAG": 808, + "AGACTA": 809, + "AGACTT": 810, + "AGACTC": 811, + "AGACTG": 812, + "AGACCA": 813, + "AGACCT": 814, + "AGACCC": 815, + "AGACCG": 816, + "AGACGA": 817, + "AGACGT": 818, + "AGACGC": 819, + "AGACGG": 820, + "AGAGAA": 821, + "AGAGAT": 822, + "AGAGAC": 823, + "AGAGAG": 824, + "AGAGTA": 825, + "AGAGTT": 826, + "AGAGTC": 827, + "AGAGTG": 828, + "AGAGCA": 829, + "AGAGCT": 830, + "AGAGCC": 831, + "AGAGCG": 832, + "AGAGGA": 833, + "AGAGGT": 834, + "AGAGGC": 835, + "AGAGGG": 836, + "AGTAAA": 837, + "AGTAAT": 838, + "AGTAAC": 839, + "AGTAAG": 840, + "AGTATA": 841, + "AGTATT": 842, + "AGTATC": 843, + "AGTATG": 844, + "AGTACA": 845, + "AGTACT": 846, + "AGTACC": 847, + "AGTACG": 848, + "AGTAGA": 849, + "AGTAGT": 850, + "AGTAGC": 851, + "AGTAGG": 852, + "AGTTAA": 853, + "AGTTAT": 854, + "AGTTAC": 855, + "AGTTAG": 856, + "AGTTTA": 857, + "AGTTTT": 858, + "AGTTTC": 859, + "AGTTTG": 860, + "AGTTCA": 861, + "AGTTCT": 862, + "AGTTCC": 863, + "AGTTCG": 864, + "AGTTGA": 865, + "AGTTGT": 866, + "AGTTGC": 867, + "AGTTGG": 868, + "AGTCAA": 869, + "AGTCAT": 870, + "AGTCAC": 871, + "AGTCAG": 872, + "AGTCTA": 873, + "AGTCTT": 874, + "AGTCTC": 875, + "AGTCTG": 876, + "AGTCCA": 877, + "AGTCCT": 878, + "AGTCCC": 879, + "AGTCCG": 880, + "AGTCGA": 881, + "AGTCGT": 882, + "AGTCGC": 883, + "AGTCGG": 884, + "AGTGAA": 885, + "AGTGAT": 886, + "AGTGAC": 887, + "AGTGAG": 888, + "AGTGTA": 889, + "AGTGTT": 890, + "AGTGTC": 891, + "AGTGTG": 892, + "AGTGCA": 893, + "AGTGCT": 894, + "AGTGCC": 895, + "AGTGCG": 896, + "AGTGGA": 897, + "AGTGGT": 898, + "AGTGGC": 899, + "AGTGGG": 900, + "AGCAAA": 901, + "AGCAAT": 902, + "AGCAAC": 903, + "AGCAAG": 904, + "AGCATA": 905, + "AGCATT": 906, + "AGCATC": 907, + "AGCATG": 908, + "AGCACA": 909, + "AGCACT": 910, + "AGCACC": 911, + "AGCACG": 912, + "AGCAGA": 913, + "AGCAGT": 914, + "AGCAGC": 915, + "AGCAGG": 916, + "AGCTAA": 917, + "AGCTAT": 918, + "AGCTAC": 919, + "AGCTAG": 920, + "AGCTTA": 921, + "AGCTTT": 922, + "AGCTTC": 923, + "AGCTTG": 924, + "AGCTCA": 925, + "AGCTCT": 926, + "AGCTCC": 927, + "AGCTCG": 928, + "AGCTGA": 929, + "AGCTGT": 930, + "AGCTGC": 931, + "AGCTGG": 932, + "AGCCAA": 933, + "AGCCAT": 934, + "AGCCAC": 935, + "AGCCAG": 936, + "AGCCTA": 937, + "AGCCTT": 938, + "AGCCTC": 939, + "AGCCTG": 940, + "AGCCCA": 941, + "AGCCCT": 942, + "AGCCCC": 943, + "AGCCCG": 944, + "AGCCGA": 945, + "AGCCGT": 946, + "AGCCGC": 947, + "AGCCGG": 948, + "AGCGAA": 949, + "AGCGAT": 950, + "AGCGAC": 951, + "AGCGAG": 952, + "AGCGTA": 953, + "AGCGTT": 954, + "AGCGTC": 955, + "AGCGTG": 956, + "AGCGCA": 957, + "AGCGCT": 958, + "AGCGCC": 959, + "AGCGCG": 960, + "AGCGGA": 961, + "AGCGGT": 962, + "AGCGGC": 963, + "AGCGGG": 964, + "AGGAAA": 965, + "AGGAAT": 966, + "AGGAAC": 967, + "AGGAAG": 968, + "AGGATA": 969, + "AGGATT": 970, + "AGGATC": 971, + "AGGATG": 972, + "AGGACA": 973, + "AGGACT": 974, + "AGGACC": 975, + "AGGACG": 976, + "AGGAGA": 977, + "AGGAGT": 978, + "AGGAGC": 979, + "AGGAGG": 980, + "AGGTAA": 981, + "AGGTAT": 982, + "AGGTAC": 983, + "AGGTAG": 984, + "AGGTTA": 985, + "AGGTTT": 986, + "AGGTTC": 987, + "AGGTTG": 988, + "AGGTCA": 989, + "AGGTCT": 990, + "AGGTCC": 991, + "AGGTCG": 992, + "AGGTGA": 993, + "AGGTGT": 994, + "AGGTGC": 995, + "AGGTGG": 996, + "AGGCAA": 997, + "AGGCAT": 998, + "AGGCAC": 999, + "AGGCAG": 1000, + "AGGCTA": 1001, + "AGGCTT": 1002, + "AGGCTC": 1003, + "AGGCTG": 1004, + "AGGCCA": 1005, + "AGGCCT": 1006, + "AGGCCC": 1007, + "AGGCCG": 1008, + "AGGCGA": 1009, + "AGGCGT": 1010, + "AGGCGC": 1011, + "AGGCGG": 1012, + "AGGGAA": 1013, + "AGGGAT": 1014, + "AGGGAC": 1015, + "AGGGAG": 1016, + "AGGGTA": 1017, + "AGGGTT": 1018, + "AGGGTC": 1019, + "AGGGTG": 1020, + "AGGGCA": 1021, + "AGGGCT": 1022, + "AGGGCC": 1023, + "AGGGCG": 1024, + "AGGGGA": 1025, + "AGGGGT": 1026, + "AGGGGC": 1027, + "AGGGGG": 1028, + "TAAAAA": 1029, + "TAAAAT": 1030, + "TAAAAC": 1031, + "TAAAAG": 1032, + "TAAATA": 1033, + "TAAATT": 1034, + "TAAATC": 1035, + "TAAATG": 1036, + "TAAACA": 1037, + "TAAACT": 1038, + "TAAACC": 1039, + "TAAACG": 1040, + "TAAAGA": 1041, + "TAAAGT": 1042, + "TAAAGC": 1043, + "TAAAGG": 1044, + "TAATAA": 1045, + "TAATAT": 1046, + "TAATAC": 1047, + "TAATAG": 1048, + "TAATTA": 1049, + "TAATTT": 1050, + "TAATTC": 1051, + "TAATTG": 1052, + "TAATCA": 1053, + "TAATCT": 1054, + "TAATCC": 1055, + "TAATCG": 1056, + "TAATGA": 1057, + "TAATGT": 1058, + "TAATGC": 1059, + "TAATGG": 1060, + "TAACAA": 1061, + "TAACAT": 1062, + "TAACAC": 1063, + "TAACAG": 1064, + "TAACTA": 1065, + "TAACTT": 1066, + "TAACTC": 1067, + "TAACTG": 1068, + "TAACCA": 1069, + "TAACCT": 1070, + "TAACCC": 1071, + "TAACCG": 1072, + "TAACGA": 1073, + "TAACGT": 1074, + "TAACGC": 1075, + "TAACGG": 1076, + "TAAGAA": 1077, + "TAAGAT": 1078, + "TAAGAC": 1079, + "TAAGAG": 1080, + "TAAGTA": 1081, + "TAAGTT": 1082, + "TAAGTC": 1083, + "TAAGTG": 1084, + "TAAGCA": 1085, + "TAAGCT": 1086, + "TAAGCC": 1087, + "TAAGCG": 1088, + "TAAGGA": 1089, + "TAAGGT": 1090, + "TAAGGC": 1091, + "TAAGGG": 1092, + "TATAAA": 1093, + "TATAAT": 1094, + "TATAAC": 1095, + "TATAAG": 1096, + "TATATA": 1097, + "TATATT": 1098, + "TATATC": 1099, + "TATATG": 1100, + "TATACA": 1101, + "TATACT": 1102, + "TATACC": 1103, + "TATACG": 1104, + "TATAGA": 1105, + "TATAGT": 1106, + "TATAGC": 1107, + "TATAGG": 1108, + "TATTAA": 1109, + "TATTAT": 1110, + "TATTAC": 1111, + "TATTAG": 1112, + "TATTTA": 1113, + "TATTTT": 1114, + "TATTTC": 1115, + "TATTTG": 1116, + "TATTCA": 1117, + "TATTCT": 1118, + "TATTCC": 1119, + "TATTCG": 1120, + "TATTGA": 1121, + "TATTGT": 1122, + "TATTGC": 1123, + "TATTGG": 1124, + "TATCAA": 1125, + "TATCAT": 1126, + "TATCAC": 1127, + "TATCAG": 1128, + "TATCTA": 1129, + "TATCTT": 1130, + "TATCTC": 1131, + "TATCTG": 1132, + "TATCCA": 1133, + "TATCCT": 1134, + "TATCCC": 1135, + "TATCCG": 1136, + "TATCGA": 1137, + "TATCGT": 1138, + "TATCGC": 1139, + "TATCGG": 1140, + "TATGAA": 1141, + "TATGAT": 1142, + "TATGAC": 1143, + "TATGAG": 1144, + "TATGTA": 1145, + "TATGTT": 1146, + "TATGTC": 1147, + "TATGTG": 1148, + "TATGCA": 1149, + "TATGCT": 1150, + "TATGCC": 1151, + "TATGCG": 1152, + "TATGGA": 1153, + "TATGGT": 1154, + "TATGGC": 1155, + "TATGGG": 1156, + "TACAAA": 1157, + "TACAAT": 1158, + "TACAAC": 1159, + "TACAAG": 1160, + "TACATA": 1161, + "TACATT": 1162, + "TACATC": 1163, + "TACATG": 1164, + "TACACA": 1165, + "TACACT": 1166, + "TACACC": 1167, + "TACACG": 1168, + "TACAGA": 1169, + "TACAGT": 1170, + "TACAGC": 1171, + "TACAGG": 1172, + "TACTAA": 1173, + "TACTAT": 1174, + "TACTAC": 1175, + "TACTAG": 1176, + "TACTTA": 1177, + "TACTTT": 1178, + "TACTTC": 1179, + "TACTTG": 1180, + "TACTCA": 1181, + "TACTCT": 1182, + "TACTCC": 1183, + "TACTCG": 1184, + "TACTGA": 1185, + "TACTGT": 1186, + "TACTGC": 1187, + "TACTGG": 1188, + "TACCAA": 1189, + "TACCAT": 1190, + "TACCAC": 1191, + "TACCAG": 1192, + "TACCTA": 1193, + "TACCTT": 1194, + "TACCTC": 1195, + "TACCTG": 1196, + "TACCCA": 1197, + "TACCCT": 1198, + "TACCCC": 1199, + "TACCCG": 1200, + "TACCGA": 1201, + "TACCGT": 1202, + "TACCGC": 1203, + "TACCGG": 1204, + "TACGAA": 1205, + "TACGAT": 1206, + "TACGAC": 1207, + "TACGAG": 1208, + "TACGTA": 1209, + "TACGTT": 1210, + "TACGTC": 1211, + "TACGTG": 1212, + "TACGCA": 1213, + "TACGCT": 1214, + "TACGCC": 1215, + "TACGCG": 1216, + "TACGGA": 1217, + "TACGGT": 1218, + "TACGGC": 1219, + "TACGGG": 1220, + "TAGAAA": 1221, + "TAGAAT": 1222, + "TAGAAC": 1223, + "TAGAAG": 1224, + "TAGATA": 1225, + "TAGATT": 1226, + "TAGATC": 1227, + "TAGATG": 1228, + "TAGACA": 1229, + "TAGACT": 1230, + "TAGACC": 1231, + "TAGACG": 1232, + "TAGAGA": 1233, + "TAGAGT": 1234, + "TAGAGC": 1235, + "TAGAGG": 1236, + "TAGTAA": 1237, + "TAGTAT": 1238, + "TAGTAC": 1239, + "TAGTAG": 1240, + "TAGTTA": 1241, + "TAGTTT": 1242, + "TAGTTC": 1243, + "TAGTTG": 1244, + "TAGTCA": 1245, + "TAGTCT": 1246, + "TAGTCC": 1247, + "TAGTCG": 1248, + "TAGTGA": 1249, + "TAGTGT": 1250, + "TAGTGC": 1251, + "TAGTGG": 1252, + "TAGCAA": 1253, + "TAGCAT": 1254, + "TAGCAC": 1255, + "TAGCAG": 1256, + "TAGCTA": 1257, + "TAGCTT": 1258, + "TAGCTC": 1259, + "TAGCTG": 1260, + "TAGCCA": 1261, + "TAGCCT": 1262, + "TAGCCC": 1263, + "TAGCCG": 1264, + "TAGCGA": 1265, + "TAGCGT": 1266, + "TAGCGC": 1267, + "TAGCGG": 1268, + "TAGGAA": 1269, + "TAGGAT": 1270, + "TAGGAC": 1271, + "TAGGAG": 1272, + "TAGGTA": 1273, + "TAGGTT": 1274, + "TAGGTC": 1275, + "TAGGTG": 1276, + "TAGGCA": 1277, + "TAGGCT": 1278, + "TAGGCC": 1279, + "TAGGCG": 1280, + "TAGGGA": 1281, + "TAGGGT": 1282, + "TAGGGC": 1283, + "TAGGGG": 1284, + "TTAAAA": 1285, + "TTAAAT": 1286, + "TTAAAC": 1287, + "TTAAAG": 1288, + "TTAATA": 1289, + "TTAATT": 1290, + "TTAATC": 1291, + "TTAATG": 1292, + "TTAACA": 1293, + "TTAACT": 1294, + "TTAACC": 1295, + "TTAACG": 1296, + "TTAAGA": 1297, + "TTAAGT": 1298, + "TTAAGC": 1299, + "TTAAGG": 1300, + "TTATAA": 1301, + "TTATAT": 1302, + "TTATAC": 1303, + "TTATAG": 1304, + "TTATTA": 1305, + "TTATTT": 1306, + "TTATTC": 1307, + "TTATTG": 1308, + "TTATCA": 1309, + "TTATCT": 1310, + "TTATCC": 1311, + "TTATCG": 1312, + "TTATGA": 1313, + "TTATGT": 1314, + "TTATGC": 1315, + "TTATGG": 1316, + "TTACAA": 1317, + "TTACAT": 1318, + "TTACAC": 1319, + "TTACAG": 1320, + "TTACTA": 1321, + "TTACTT": 1322, + "TTACTC": 1323, + "TTACTG": 1324, + "TTACCA": 1325, + "TTACCT": 1326, + "TTACCC": 1327, + "TTACCG": 1328, + "TTACGA": 1329, + "TTACGT": 1330, + "TTACGC": 1331, + "TTACGG": 1332, + "TTAGAA": 1333, + "TTAGAT": 1334, + "TTAGAC": 1335, + "TTAGAG": 1336, + "TTAGTA": 1337, + "TTAGTT": 1338, + "TTAGTC": 1339, + "TTAGTG": 1340, + "TTAGCA": 1341, + "TTAGCT": 1342, + "TTAGCC": 1343, + "TTAGCG": 1344, + "TTAGGA": 1345, + "TTAGGT": 1346, + "TTAGGC": 1347, + "TTAGGG": 1348, + "TTTAAA": 1349, + "TTTAAT": 1350, + "TTTAAC": 1351, + "TTTAAG": 1352, + "TTTATA": 1353, + "TTTATT": 1354, + "TTTATC": 1355, + "TTTATG": 1356, + "TTTACA": 1357, + "TTTACT": 1358, + "TTTACC": 1359, + "TTTACG": 1360, + "TTTAGA": 1361, + "TTTAGT": 1362, + "TTTAGC": 1363, + "TTTAGG": 1364, + "TTTTAA": 1365, + "TTTTAT": 1366, + "TTTTAC": 1367, + "TTTTAG": 1368, + "TTTTTA": 1369, + "TTTTTT": 1370, + "TTTTTC": 1371, + "TTTTTG": 1372, + "TTTTCA": 1373, + "TTTTCT": 1374, + "TTTTCC": 1375, + "TTTTCG": 1376, + "TTTTGA": 1377, + "TTTTGT": 1378, + "TTTTGC": 1379, + "TTTTGG": 1380, + "TTTCAA": 1381, + "TTTCAT": 1382, + "TTTCAC": 1383, + "TTTCAG": 1384, + "TTTCTA": 1385, + "TTTCTT": 1386, + "TTTCTC": 1387, + "TTTCTG": 1388, + "TTTCCA": 1389, + "TTTCCT": 1390, + "TTTCCC": 1391, + "TTTCCG": 1392, + "TTTCGA": 1393, + "TTTCGT": 1394, + "TTTCGC": 1395, + "TTTCGG": 1396, + "TTTGAA": 1397, + "TTTGAT": 1398, + "TTTGAC": 1399, + "TTTGAG": 1400, + "TTTGTA": 1401, + "TTTGTT": 1402, + "TTTGTC": 1403, + "TTTGTG": 1404, + "TTTGCA": 1405, + "TTTGCT": 1406, + "TTTGCC": 1407, + "TTTGCG": 1408, + "TTTGGA": 1409, + "TTTGGT": 1410, + "TTTGGC": 1411, + "TTTGGG": 1412, + "TTCAAA": 1413, + "TTCAAT": 1414, + "TTCAAC": 1415, + "TTCAAG": 1416, + "TTCATA": 1417, + "TTCATT": 1418, + "TTCATC": 1419, + "TTCATG": 1420, + "TTCACA": 1421, + "TTCACT": 1422, + "TTCACC": 1423, + "TTCACG": 1424, + "TTCAGA": 1425, + "TTCAGT": 1426, + "TTCAGC": 1427, + "TTCAGG": 1428, + "TTCTAA": 1429, + "TTCTAT": 1430, + "TTCTAC": 1431, + "TTCTAG": 1432, + "TTCTTA": 1433, + "TTCTTT": 1434, + "TTCTTC": 1435, + "TTCTTG": 1436, + "TTCTCA": 1437, + "TTCTCT": 1438, + "TTCTCC": 1439, + "TTCTCG": 1440, + "TTCTGA": 1441, + "TTCTGT": 1442, + "TTCTGC": 1443, + "TTCTGG": 1444, + "TTCCAA": 1445, + "TTCCAT": 1446, + "TTCCAC": 1447, + "TTCCAG": 1448, + "TTCCTA": 1449, + "TTCCTT": 1450, + "TTCCTC": 1451, + "TTCCTG": 1452, + "TTCCCA": 1453, + "TTCCCT": 1454, + "TTCCCC": 1455, + "TTCCCG": 1456, + "TTCCGA": 1457, + "TTCCGT": 1458, + "TTCCGC": 1459, + "TTCCGG": 1460, + "TTCGAA": 1461, + "TTCGAT": 1462, + "TTCGAC": 1463, + "TTCGAG": 1464, + "TTCGTA": 1465, + "TTCGTT": 1466, + "TTCGTC": 1467, + "TTCGTG": 1468, + "TTCGCA": 1469, + "TTCGCT": 1470, + "TTCGCC": 1471, + "TTCGCG": 1472, + "TTCGGA": 1473, + "TTCGGT": 1474, + "TTCGGC": 1475, + "TTCGGG": 1476, + "TTGAAA": 1477, + "TTGAAT": 1478, + "TTGAAC": 1479, + "TTGAAG": 1480, + "TTGATA": 1481, + "TTGATT": 1482, + "TTGATC": 1483, + "TTGATG": 1484, + "TTGACA": 1485, + "TTGACT": 1486, + "TTGACC": 1487, + "TTGACG": 1488, + "TTGAGA": 1489, + "TTGAGT": 1490, + "TTGAGC": 1491, + "TTGAGG": 1492, + "TTGTAA": 1493, + "TTGTAT": 1494, + "TTGTAC": 1495, + "TTGTAG": 1496, + "TTGTTA": 1497, + "TTGTTT": 1498, + "TTGTTC": 1499, + "TTGTTG": 1500, + "TTGTCA": 1501, + "TTGTCT": 1502, + "TTGTCC": 1503, + "TTGTCG": 1504, + "TTGTGA": 1505, + "TTGTGT": 1506, + "TTGTGC": 1507, + "TTGTGG": 1508, + "TTGCAA": 1509, + "TTGCAT": 1510, + "TTGCAC": 1511, + "TTGCAG": 1512, + "TTGCTA": 1513, + "TTGCTT": 1514, + "TTGCTC": 1515, + "TTGCTG": 1516, + "TTGCCA": 1517, + "TTGCCT": 1518, + "TTGCCC": 1519, + "TTGCCG": 1520, + "TTGCGA": 1521, + "TTGCGT": 1522, + "TTGCGC": 1523, + "TTGCGG": 1524, + "TTGGAA": 1525, + "TTGGAT": 1526, + "TTGGAC": 1527, + "TTGGAG": 1528, + "TTGGTA": 1529, + "TTGGTT": 1530, + "TTGGTC": 1531, + "TTGGTG": 1532, + "TTGGCA": 1533, + "TTGGCT": 1534, + "TTGGCC": 1535, + "TTGGCG": 1536, + "TTGGGA": 1537, + "TTGGGT": 1538, + "TTGGGC": 1539, + "TTGGGG": 1540, + "TCAAAA": 1541, + "TCAAAT": 1542, + "TCAAAC": 1543, + "TCAAAG": 1544, + "TCAATA": 1545, + "TCAATT": 1546, + "TCAATC": 1547, + "TCAATG": 1548, + "TCAACA": 1549, + "TCAACT": 1550, + "TCAACC": 1551, + "TCAACG": 1552, + "TCAAGA": 1553, + "TCAAGT": 1554, + "TCAAGC": 1555, + "TCAAGG": 1556, + "TCATAA": 1557, + "TCATAT": 1558, + "TCATAC": 1559, + "TCATAG": 1560, + "TCATTA": 1561, + "TCATTT": 1562, + "TCATTC": 1563, + "TCATTG": 1564, + "TCATCA": 1565, + "TCATCT": 1566, + "TCATCC": 1567, + "TCATCG": 1568, + "TCATGA": 1569, + "TCATGT": 1570, + "TCATGC": 1571, + "TCATGG": 1572, + "TCACAA": 1573, + "TCACAT": 1574, + "TCACAC": 1575, + "TCACAG": 1576, + "TCACTA": 1577, + "TCACTT": 1578, + "TCACTC": 1579, + "TCACTG": 1580, + "TCACCA": 1581, + "TCACCT": 1582, + "TCACCC": 1583, + "TCACCG": 1584, + "TCACGA": 1585, + "TCACGT": 1586, + "TCACGC": 1587, + "TCACGG": 1588, + "TCAGAA": 1589, + "TCAGAT": 1590, + "TCAGAC": 1591, + "TCAGAG": 1592, + "TCAGTA": 1593, + "TCAGTT": 1594, + "TCAGTC": 1595, + "TCAGTG": 1596, + "TCAGCA": 1597, + "TCAGCT": 1598, + "TCAGCC": 1599, + "TCAGCG": 1600, + "TCAGGA": 1601, + "TCAGGT": 1602, + "TCAGGC": 1603, + "TCAGGG": 1604, + "TCTAAA": 1605, + "TCTAAT": 1606, + "TCTAAC": 1607, + "TCTAAG": 1608, + "TCTATA": 1609, + "TCTATT": 1610, + "TCTATC": 1611, + "TCTATG": 1612, + "TCTACA": 1613, + "TCTACT": 1614, + "TCTACC": 1615, + "TCTACG": 1616, + "TCTAGA": 1617, + "TCTAGT": 1618, + "TCTAGC": 1619, + "TCTAGG": 1620, + "TCTTAA": 1621, + "TCTTAT": 1622, + "TCTTAC": 1623, + "TCTTAG": 1624, + "TCTTTA": 1625, + "TCTTTT": 1626, + "TCTTTC": 1627, + "TCTTTG": 1628, + "TCTTCA": 1629, + "TCTTCT": 1630, + "TCTTCC": 1631, + "TCTTCG": 1632, + "TCTTGA": 1633, + "TCTTGT": 1634, + "TCTTGC": 1635, + "TCTTGG": 1636, + "TCTCAA": 1637, + "TCTCAT": 1638, + "TCTCAC": 1639, + "TCTCAG": 1640, + "TCTCTA": 1641, + "TCTCTT": 1642, + "TCTCTC": 1643, + "TCTCTG": 1644, + "TCTCCA": 1645, + "TCTCCT": 1646, + "TCTCCC": 1647, + "TCTCCG": 1648, + "TCTCGA": 1649, + "TCTCGT": 1650, + "TCTCGC": 1651, + "TCTCGG": 1652, + "TCTGAA": 1653, + "TCTGAT": 1654, + "TCTGAC": 1655, + "TCTGAG": 1656, + "TCTGTA": 1657, + "TCTGTT": 1658, + "TCTGTC": 1659, + "TCTGTG": 1660, + "TCTGCA": 1661, + "TCTGCT": 1662, + "TCTGCC": 1663, + "TCTGCG": 1664, + "TCTGGA": 1665, + "TCTGGT": 1666, + "TCTGGC": 1667, + "TCTGGG": 1668, + "TCCAAA": 1669, + "TCCAAT": 1670, + "TCCAAC": 1671, + "TCCAAG": 1672, + "TCCATA": 1673, + "TCCATT": 1674, + "TCCATC": 1675, + "TCCATG": 1676, + "TCCACA": 1677, + "TCCACT": 1678, + "TCCACC": 1679, + "TCCACG": 1680, + "TCCAGA": 1681, + "TCCAGT": 1682, + "TCCAGC": 1683, + "TCCAGG": 1684, + "TCCTAA": 1685, + "TCCTAT": 1686, + "TCCTAC": 1687, + "TCCTAG": 1688, + "TCCTTA": 1689, + "TCCTTT": 1690, + "TCCTTC": 1691, + "TCCTTG": 1692, + "TCCTCA": 1693, + "TCCTCT": 1694, + "TCCTCC": 1695, + "TCCTCG": 1696, + "TCCTGA": 1697, + "TCCTGT": 1698, + "TCCTGC": 1699, + "TCCTGG": 1700, + "TCCCAA": 1701, + "TCCCAT": 1702, + "TCCCAC": 1703, + "TCCCAG": 1704, + "TCCCTA": 1705, + "TCCCTT": 1706, + "TCCCTC": 1707, + "TCCCTG": 1708, + "TCCCCA": 1709, + "TCCCCT": 1710, + "TCCCCC": 1711, + "TCCCCG": 1712, + "TCCCGA": 1713, + "TCCCGT": 1714, + "TCCCGC": 1715, + "TCCCGG": 1716, + "TCCGAA": 1717, + "TCCGAT": 1718, + "TCCGAC": 1719, + "TCCGAG": 1720, + "TCCGTA": 1721, + "TCCGTT": 1722, + "TCCGTC": 1723, + "TCCGTG": 1724, + "TCCGCA": 1725, + "TCCGCT": 1726, + "TCCGCC": 1727, + "TCCGCG": 1728, + "TCCGGA": 1729, + "TCCGGT": 1730, + "TCCGGC": 1731, + "TCCGGG": 1732, + "TCGAAA": 1733, + "TCGAAT": 1734, + "TCGAAC": 1735, + "TCGAAG": 1736, + "TCGATA": 1737, + "TCGATT": 1738, + "TCGATC": 1739, + "TCGATG": 1740, + "TCGACA": 1741, + "TCGACT": 1742, + "TCGACC": 1743, + "TCGACG": 1744, + "TCGAGA": 1745, + "TCGAGT": 1746, + "TCGAGC": 1747, + "TCGAGG": 1748, + "TCGTAA": 1749, + "TCGTAT": 1750, + "TCGTAC": 1751, + "TCGTAG": 1752, + "TCGTTA": 1753, + "TCGTTT": 1754, + "TCGTTC": 1755, + "TCGTTG": 1756, + "TCGTCA": 1757, + "TCGTCT": 1758, + "TCGTCC": 1759, + "TCGTCG": 1760, + "TCGTGA": 1761, + "TCGTGT": 1762, + "TCGTGC": 1763, + "TCGTGG": 1764, + "TCGCAA": 1765, + "TCGCAT": 1766, + "TCGCAC": 1767, + "TCGCAG": 1768, + "TCGCTA": 1769, + "TCGCTT": 1770, + "TCGCTC": 1771, + "TCGCTG": 1772, + "TCGCCA": 1773, + "TCGCCT": 1774, + "TCGCCC": 1775, + "TCGCCG": 1776, + "TCGCGA": 1777, + "TCGCGT": 1778, + "TCGCGC": 1779, + "TCGCGG": 1780, + "TCGGAA": 1781, + "TCGGAT": 1782, + "TCGGAC": 1783, + "TCGGAG": 1784, + "TCGGTA": 1785, + "TCGGTT": 1786, + "TCGGTC": 1787, + "TCGGTG": 1788, + "TCGGCA": 1789, + "TCGGCT": 1790, + "TCGGCC": 1791, + "TCGGCG": 1792, + "TCGGGA": 1793, + "TCGGGT": 1794, + "TCGGGC": 1795, + "TCGGGG": 1796, + "TGAAAA": 1797, + "TGAAAT": 1798, + "TGAAAC": 1799, + "TGAAAG": 1800, + "TGAATA": 1801, + "TGAATT": 1802, + "TGAATC": 1803, + "TGAATG": 1804, + "TGAACA": 1805, + "TGAACT": 1806, + "TGAACC": 1807, + "TGAACG": 1808, + "TGAAGA": 1809, + "TGAAGT": 1810, + "TGAAGC": 1811, + "TGAAGG": 1812, + "TGATAA": 1813, + "TGATAT": 1814, + "TGATAC": 1815, + "TGATAG": 1816, + "TGATTA": 1817, + "TGATTT": 1818, + "TGATTC": 1819, + "TGATTG": 1820, + "TGATCA": 1821, + "TGATCT": 1822, + "TGATCC": 1823, + "TGATCG": 1824, + "TGATGA": 1825, + "TGATGT": 1826, + "TGATGC": 1827, + "TGATGG": 1828, + "TGACAA": 1829, + "TGACAT": 1830, + "TGACAC": 1831, + "TGACAG": 1832, + "TGACTA": 1833, + "TGACTT": 1834, + "TGACTC": 1835, + "TGACTG": 1836, + "TGACCA": 1837, + "TGACCT": 1838, + "TGACCC": 1839, + "TGACCG": 1840, + "TGACGA": 1841, + "TGACGT": 1842, + "TGACGC": 1843, + "TGACGG": 1844, + "TGAGAA": 1845, + "TGAGAT": 1846, + "TGAGAC": 1847, + "TGAGAG": 1848, + "TGAGTA": 1849, + "TGAGTT": 1850, + "TGAGTC": 1851, + "TGAGTG": 1852, + "TGAGCA": 1853, + "TGAGCT": 1854, + "TGAGCC": 1855, + "TGAGCG": 1856, + "TGAGGA": 1857, + "TGAGGT": 1858, + "TGAGGC": 1859, + "TGAGGG": 1860, + "TGTAAA": 1861, + "TGTAAT": 1862, + "TGTAAC": 1863, + "TGTAAG": 1864, + "TGTATA": 1865, + "TGTATT": 1866, + "TGTATC": 1867, + "TGTATG": 1868, + "TGTACA": 1869, + "TGTACT": 1870, + "TGTACC": 1871, + "TGTACG": 1872, + "TGTAGA": 1873, + "TGTAGT": 1874, + "TGTAGC": 1875, + "TGTAGG": 1876, + "TGTTAA": 1877, + "TGTTAT": 1878, + "TGTTAC": 1879, + "TGTTAG": 1880, + "TGTTTA": 1881, + "TGTTTT": 1882, + "TGTTTC": 1883, + "TGTTTG": 1884, + "TGTTCA": 1885, + "TGTTCT": 1886, + "TGTTCC": 1887, + "TGTTCG": 1888, + "TGTTGA": 1889, + "TGTTGT": 1890, + "TGTTGC": 1891, + "TGTTGG": 1892, + "TGTCAA": 1893, + "TGTCAT": 1894, + "TGTCAC": 1895, + "TGTCAG": 1896, + "TGTCTA": 1897, + "TGTCTT": 1898, + "TGTCTC": 1899, + "TGTCTG": 1900, + "TGTCCA": 1901, + "TGTCCT": 1902, + "TGTCCC": 1903, + "TGTCCG": 1904, + "TGTCGA": 1905, + "TGTCGT": 1906, + "TGTCGC": 1907, + "TGTCGG": 1908, + "TGTGAA": 1909, + "TGTGAT": 1910, + "TGTGAC": 1911, + "TGTGAG": 1912, + "TGTGTA": 1913, + "TGTGTT": 1914, + "TGTGTC": 1915, + "TGTGTG": 1916, + "TGTGCA": 1917, + "TGTGCT": 1918, + "TGTGCC": 1919, + "TGTGCG": 1920, + "TGTGGA": 1921, + "TGTGGT": 1922, + "TGTGGC": 1923, + "TGTGGG": 1924, + "TGCAAA": 1925, + "TGCAAT": 1926, + "TGCAAC": 1927, + "TGCAAG": 1928, + "TGCATA": 1929, + "TGCATT": 1930, + "TGCATC": 1931, + "TGCATG": 1932, + "TGCACA": 1933, + "TGCACT": 1934, + "TGCACC": 1935, + "TGCACG": 1936, + "TGCAGA": 1937, + "TGCAGT": 1938, + "TGCAGC": 1939, + "TGCAGG": 1940, + "TGCTAA": 1941, + "TGCTAT": 1942, + "TGCTAC": 1943, + "TGCTAG": 1944, + "TGCTTA": 1945, + "TGCTTT": 1946, + "TGCTTC": 1947, + "TGCTTG": 1948, + "TGCTCA": 1949, + "TGCTCT": 1950, + "TGCTCC": 1951, + "TGCTCG": 1952, + "TGCTGA": 1953, + "TGCTGT": 1954, + "TGCTGC": 1955, + "TGCTGG": 1956, + "TGCCAA": 1957, + "TGCCAT": 1958, + "TGCCAC": 1959, + "TGCCAG": 1960, + "TGCCTA": 1961, + "TGCCTT": 1962, + "TGCCTC": 1963, + "TGCCTG": 1964, + "TGCCCA": 1965, + "TGCCCT": 1966, + "TGCCCC": 1967, + "TGCCCG": 1968, + "TGCCGA": 1969, + "TGCCGT": 1970, + "TGCCGC": 1971, + "TGCCGG": 1972, + "TGCGAA": 1973, + "TGCGAT": 1974, + "TGCGAC": 1975, + "TGCGAG": 1976, + "TGCGTA": 1977, + "TGCGTT": 1978, + "TGCGTC": 1979, + "TGCGTG": 1980, + "TGCGCA": 1981, + "TGCGCT": 1982, + "TGCGCC": 1983, + "TGCGCG": 1984, + "TGCGGA": 1985, + "TGCGGT": 1986, + "TGCGGC": 1987, + "TGCGGG": 1988, + "TGGAAA": 1989, + "TGGAAT": 1990, + "TGGAAC": 1991, + "TGGAAG": 1992, + "TGGATA": 1993, + "TGGATT": 1994, + "TGGATC": 1995, + "TGGATG": 1996, + "TGGACA": 1997, + "TGGACT": 1998, + "TGGACC": 1999, + "TGGACG": 2000, + "TGGAGA": 2001, + "TGGAGT": 2002, + "TGGAGC": 2003, + "TGGAGG": 2004, + "TGGTAA": 2005, + "TGGTAT": 2006, + "TGGTAC": 2007, + "TGGTAG": 2008, + "TGGTTA": 2009, + "TGGTTT": 2010, + "TGGTTC": 2011, + "TGGTTG": 2012, + "TGGTCA": 2013, + "TGGTCT": 2014, + "TGGTCC": 2015, + "TGGTCG": 2016, + "TGGTGA": 2017, + "TGGTGT": 2018, + "TGGTGC": 2019, + "TGGTGG": 2020, + "TGGCAA": 2021, + "TGGCAT": 2022, + "TGGCAC": 2023, + "TGGCAG": 2024, + "TGGCTA": 2025, + "TGGCTT": 2026, + "TGGCTC": 2027, + "TGGCTG": 2028, + "TGGCCA": 2029, + "TGGCCT": 2030, + "TGGCCC": 2031, + "TGGCCG": 2032, + "TGGCGA": 2033, + "TGGCGT": 2034, + "TGGCGC": 2035, + "TGGCGG": 2036, + "TGGGAA": 2037, + "TGGGAT": 2038, + "TGGGAC": 2039, + "TGGGAG": 2040, + "TGGGTA": 2041, + "TGGGTT": 2042, + "TGGGTC": 2043, + "TGGGTG": 2044, + "TGGGCA": 2045, + "TGGGCT": 2046, + "TGGGCC": 2047, + "TGGGCG": 2048, + "TGGGGA": 2049, + "TGGGGT": 2050, + "TGGGGC": 2051, + "TGGGGG": 2052, + "CAAAAA": 2053, + "CAAAAT": 2054, + "CAAAAC": 2055, + "CAAAAG": 2056, + "CAAATA": 2057, + "CAAATT": 2058, + "CAAATC": 2059, + "CAAATG": 2060, + "CAAACA": 2061, + "CAAACT": 2062, + "CAAACC": 2063, + "CAAACG": 2064, + "CAAAGA": 2065, + "CAAAGT": 2066, + "CAAAGC": 2067, + "CAAAGG": 2068, + "CAATAA": 2069, + "CAATAT": 2070, + "CAATAC": 2071, + "CAATAG": 2072, + "CAATTA": 2073, + "CAATTT": 2074, + "CAATTC": 2075, + "CAATTG": 2076, + "CAATCA": 2077, + "CAATCT": 2078, + "CAATCC": 2079, + "CAATCG": 2080, + "CAATGA": 2081, + "CAATGT": 2082, + "CAATGC": 2083, + "CAATGG": 2084, + "CAACAA": 2085, + "CAACAT": 2086, + "CAACAC": 2087, + "CAACAG": 2088, + "CAACTA": 2089, + "CAACTT": 2090, + "CAACTC": 2091, + "CAACTG": 2092, + "CAACCA": 2093, + "CAACCT": 2094, + "CAACCC": 2095, + "CAACCG": 2096, + "CAACGA": 2097, + "CAACGT": 2098, + "CAACGC": 2099, + "CAACGG": 2100, + "CAAGAA": 2101, + "CAAGAT": 2102, + "CAAGAC": 2103, + "CAAGAG": 2104, + "CAAGTA": 2105, + "CAAGTT": 2106, + "CAAGTC": 2107, + "CAAGTG": 2108, + "CAAGCA": 2109, + "CAAGCT": 2110, + "CAAGCC": 2111, + "CAAGCG": 2112, + "CAAGGA": 2113, + "CAAGGT": 2114, + "CAAGGC": 2115, + "CAAGGG": 2116, + "CATAAA": 2117, + "CATAAT": 2118, + "CATAAC": 2119, + "CATAAG": 2120, + "CATATA": 2121, + "CATATT": 2122, + "CATATC": 2123, + "CATATG": 2124, + "CATACA": 2125, + "CATACT": 2126, + "CATACC": 2127, + "CATACG": 2128, + "CATAGA": 2129, + "CATAGT": 2130, + "CATAGC": 2131, + "CATAGG": 2132, + "CATTAA": 2133, + "CATTAT": 2134, + "CATTAC": 2135, + "CATTAG": 2136, + "CATTTA": 2137, + "CATTTT": 2138, + "CATTTC": 2139, + "CATTTG": 2140, + "CATTCA": 2141, + "CATTCT": 2142, + "CATTCC": 2143, + "CATTCG": 2144, + "CATTGA": 2145, + "CATTGT": 2146, + "CATTGC": 2147, + "CATTGG": 2148, + "CATCAA": 2149, + "CATCAT": 2150, + "CATCAC": 2151, + "CATCAG": 2152, + "CATCTA": 2153, + "CATCTT": 2154, + "CATCTC": 2155, + "CATCTG": 2156, + "CATCCA": 2157, + "CATCCT": 2158, + "CATCCC": 2159, + "CATCCG": 2160, + "CATCGA": 2161, + "CATCGT": 2162, + "CATCGC": 2163, + "CATCGG": 2164, + "CATGAA": 2165, + "CATGAT": 2166, + "CATGAC": 2167, + "CATGAG": 2168, + "CATGTA": 2169, + "CATGTT": 2170, + "CATGTC": 2171, + "CATGTG": 2172, + "CATGCA": 2173, + "CATGCT": 2174, + "CATGCC": 2175, + "CATGCG": 2176, + "CATGGA": 2177, + "CATGGT": 2178, + "CATGGC": 2179, + "CATGGG": 2180, + "CACAAA": 2181, + "CACAAT": 2182, + "CACAAC": 2183, + "CACAAG": 2184, + "CACATA": 2185, + "CACATT": 2186, + "CACATC": 2187, + "CACATG": 2188, + "CACACA": 2189, + "CACACT": 2190, + "CACACC": 2191, + "CACACG": 2192, + "CACAGA": 2193, + "CACAGT": 2194, + "CACAGC": 2195, + "CACAGG": 2196, + "CACTAA": 2197, + "CACTAT": 2198, + "CACTAC": 2199, + "CACTAG": 2200, + "CACTTA": 2201, + "CACTTT": 2202, + "CACTTC": 2203, + "CACTTG": 2204, + "CACTCA": 2205, + "CACTCT": 2206, + "CACTCC": 2207, + "CACTCG": 2208, + "CACTGA": 2209, + "CACTGT": 2210, + "CACTGC": 2211, + "CACTGG": 2212, + "CACCAA": 2213, + "CACCAT": 2214, + "CACCAC": 2215, + "CACCAG": 2216, + "CACCTA": 2217, + "CACCTT": 2218, + "CACCTC": 2219, + "CACCTG": 2220, + "CACCCA": 2221, + "CACCCT": 2222, + "CACCCC": 2223, + "CACCCG": 2224, + "CACCGA": 2225, + "CACCGT": 2226, + "CACCGC": 2227, + "CACCGG": 2228, + "CACGAA": 2229, + "CACGAT": 2230, + "CACGAC": 2231, + "CACGAG": 2232, + "CACGTA": 2233, + "CACGTT": 2234, + "CACGTC": 2235, + "CACGTG": 2236, + "CACGCA": 2237, + "CACGCT": 2238, + "CACGCC": 2239, + "CACGCG": 2240, + "CACGGA": 2241, + "CACGGT": 2242, + "CACGGC": 2243, + "CACGGG": 2244, + "CAGAAA": 2245, + "CAGAAT": 2246, + "CAGAAC": 2247, + "CAGAAG": 2248, + "CAGATA": 2249, + "CAGATT": 2250, + "CAGATC": 2251, + "CAGATG": 2252, + "CAGACA": 2253, + "CAGACT": 2254, + "CAGACC": 2255, + "CAGACG": 2256, + "CAGAGA": 2257, + "CAGAGT": 2258, + "CAGAGC": 2259, + "CAGAGG": 2260, + "CAGTAA": 2261, + "CAGTAT": 2262, + "CAGTAC": 2263, + "CAGTAG": 2264, + "CAGTTA": 2265, + "CAGTTT": 2266, + "CAGTTC": 2267, + "CAGTTG": 2268, + "CAGTCA": 2269, + "CAGTCT": 2270, + "CAGTCC": 2271, + "CAGTCG": 2272, + "CAGTGA": 2273, + "CAGTGT": 2274, + "CAGTGC": 2275, + "CAGTGG": 2276, + "CAGCAA": 2277, + "CAGCAT": 2278, + "CAGCAC": 2279, + "CAGCAG": 2280, + "CAGCTA": 2281, + "CAGCTT": 2282, + "CAGCTC": 2283, + "CAGCTG": 2284, + "CAGCCA": 2285, + "CAGCCT": 2286, + "CAGCCC": 2287, + "CAGCCG": 2288, + "CAGCGA": 2289, + "CAGCGT": 2290, + "CAGCGC": 2291, + "CAGCGG": 2292, + "CAGGAA": 2293, + "CAGGAT": 2294, + "CAGGAC": 2295, + "CAGGAG": 2296, + "CAGGTA": 2297, + "CAGGTT": 2298, + "CAGGTC": 2299, + "CAGGTG": 2300, + "CAGGCA": 2301, + "CAGGCT": 2302, + "CAGGCC": 2303, + "CAGGCG": 2304, + "CAGGGA": 2305, + "CAGGGT": 2306, + "CAGGGC": 2307, + "CAGGGG": 2308, + "CTAAAA": 2309, + "CTAAAT": 2310, + "CTAAAC": 2311, + "CTAAAG": 2312, + "CTAATA": 2313, + "CTAATT": 2314, + "CTAATC": 2315, + "CTAATG": 2316, + "CTAACA": 2317, + "CTAACT": 2318, + "CTAACC": 2319, + "CTAACG": 2320, + "CTAAGA": 2321, + "CTAAGT": 2322, + "CTAAGC": 2323, + "CTAAGG": 2324, + "CTATAA": 2325, + "CTATAT": 2326, + "CTATAC": 2327, + "CTATAG": 2328, + "CTATTA": 2329, + "CTATTT": 2330, + "CTATTC": 2331, + "CTATTG": 2332, + "CTATCA": 2333, + "CTATCT": 2334, + "CTATCC": 2335, + "CTATCG": 2336, + "CTATGA": 2337, + "CTATGT": 2338, + "CTATGC": 2339, + "CTATGG": 2340, + "CTACAA": 2341, + "CTACAT": 2342, + "CTACAC": 2343, + "CTACAG": 2344, + "CTACTA": 2345, + "CTACTT": 2346, + "CTACTC": 2347, + "CTACTG": 2348, + "CTACCA": 2349, + "CTACCT": 2350, + "CTACCC": 2351, + "CTACCG": 2352, + "CTACGA": 2353, + "CTACGT": 2354, + "CTACGC": 2355, + "CTACGG": 2356, + "CTAGAA": 2357, + "CTAGAT": 2358, + "CTAGAC": 2359, + "CTAGAG": 2360, + "CTAGTA": 2361, + "CTAGTT": 2362, + "CTAGTC": 2363, + "CTAGTG": 2364, + "CTAGCA": 2365, + "CTAGCT": 2366, + "CTAGCC": 2367, + "CTAGCG": 2368, + "CTAGGA": 2369, + "CTAGGT": 2370, + "CTAGGC": 2371, + "CTAGGG": 2372, + "CTTAAA": 2373, + "CTTAAT": 2374, + "CTTAAC": 2375, + "CTTAAG": 2376, + "CTTATA": 2377, + "CTTATT": 2378, + "CTTATC": 2379, + "CTTATG": 2380, + "CTTACA": 2381, + "CTTACT": 2382, + "CTTACC": 2383, + "CTTACG": 2384, + "CTTAGA": 2385, + "CTTAGT": 2386, + "CTTAGC": 2387, + "CTTAGG": 2388, + "CTTTAA": 2389, + "CTTTAT": 2390, + "CTTTAC": 2391, + "CTTTAG": 2392, + "CTTTTA": 2393, + "CTTTTT": 2394, + "CTTTTC": 2395, + "CTTTTG": 2396, + "CTTTCA": 2397, + "CTTTCT": 2398, + "CTTTCC": 2399, + "CTTTCG": 2400, + "CTTTGA": 2401, + "CTTTGT": 2402, + "CTTTGC": 2403, + "CTTTGG": 2404, + "CTTCAA": 2405, + "CTTCAT": 2406, + "CTTCAC": 2407, + "CTTCAG": 2408, + "CTTCTA": 2409, + "CTTCTT": 2410, + "CTTCTC": 2411, + "CTTCTG": 2412, + "CTTCCA": 2413, + "CTTCCT": 2414, + "CTTCCC": 2415, + "CTTCCG": 2416, + "CTTCGA": 2417, + "CTTCGT": 2418, + "CTTCGC": 2419, + "CTTCGG": 2420, + "CTTGAA": 2421, + "CTTGAT": 2422, + "CTTGAC": 2423, + "CTTGAG": 2424, + "CTTGTA": 2425, + "CTTGTT": 2426, + "CTTGTC": 2427, + "CTTGTG": 2428, + "CTTGCA": 2429, + "CTTGCT": 2430, + "CTTGCC": 2431, + "CTTGCG": 2432, + "CTTGGA": 2433, + "CTTGGT": 2434, + "CTTGGC": 2435, + "CTTGGG": 2436, + "CTCAAA": 2437, + "CTCAAT": 2438, + "CTCAAC": 2439, + "CTCAAG": 2440, + "CTCATA": 2441, + "CTCATT": 2442, + "CTCATC": 2443, + "CTCATG": 2444, + "CTCACA": 2445, + "CTCACT": 2446, + "CTCACC": 2447, + "CTCACG": 2448, + "CTCAGA": 2449, + "CTCAGT": 2450, + "CTCAGC": 2451, + "CTCAGG": 2452, + "CTCTAA": 2453, + "CTCTAT": 2454, + "CTCTAC": 2455, + "CTCTAG": 2456, + "CTCTTA": 2457, + "CTCTTT": 2458, + "CTCTTC": 2459, + "CTCTTG": 2460, + "CTCTCA": 2461, + "CTCTCT": 2462, + "CTCTCC": 2463, + "CTCTCG": 2464, + "CTCTGA": 2465, + "CTCTGT": 2466, + "CTCTGC": 2467, + "CTCTGG": 2468, + "CTCCAA": 2469, + "CTCCAT": 2470, + "CTCCAC": 2471, + "CTCCAG": 2472, + "CTCCTA": 2473, + "CTCCTT": 2474, + "CTCCTC": 2475, + "CTCCTG": 2476, + "CTCCCA": 2477, + "CTCCCT": 2478, + "CTCCCC": 2479, + "CTCCCG": 2480, + "CTCCGA": 2481, + "CTCCGT": 2482, + "CTCCGC": 2483, + "CTCCGG": 2484, + "CTCGAA": 2485, + "CTCGAT": 2486, + "CTCGAC": 2487, + "CTCGAG": 2488, + "CTCGTA": 2489, + "CTCGTT": 2490, + "CTCGTC": 2491, + "CTCGTG": 2492, + "CTCGCA": 2493, + "CTCGCT": 2494, + "CTCGCC": 2495, + "CTCGCG": 2496, + "CTCGGA": 2497, + "CTCGGT": 2498, + "CTCGGC": 2499, + "CTCGGG": 2500, + "CTGAAA": 2501, + "CTGAAT": 2502, + "CTGAAC": 2503, + "CTGAAG": 2504, + "CTGATA": 2505, + "CTGATT": 2506, + "CTGATC": 2507, + "CTGATG": 2508, + "CTGACA": 2509, + "CTGACT": 2510, + "CTGACC": 2511, + "CTGACG": 2512, + "CTGAGA": 2513, + "CTGAGT": 2514, + "CTGAGC": 2515, + "CTGAGG": 2516, + "CTGTAA": 2517, + "CTGTAT": 2518, + "CTGTAC": 2519, + "CTGTAG": 2520, + "CTGTTA": 2521, + "CTGTTT": 2522, + "CTGTTC": 2523, + "CTGTTG": 2524, + "CTGTCA": 2525, + "CTGTCT": 2526, + "CTGTCC": 2527, + "CTGTCG": 2528, + "CTGTGA": 2529, + "CTGTGT": 2530, + "CTGTGC": 2531, + "CTGTGG": 2532, + "CTGCAA": 2533, + "CTGCAT": 2534, + "CTGCAC": 2535, + "CTGCAG": 2536, + "CTGCTA": 2537, + "CTGCTT": 2538, + "CTGCTC": 2539, + "CTGCTG": 2540, + "CTGCCA": 2541, + "CTGCCT": 2542, + "CTGCCC": 2543, + "CTGCCG": 2544, + "CTGCGA": 2545, + "CTGCGT": 2546, + "CTGCGC": 2547, + "CTGCGG": 2548, + "CTGGAA": 2549, + "CTGGAT": 2550, + "CTGGAC": 2551, + "CTGGAG": 2552, + "CTGGTA": 2553, + "CTGGTT": 2554, + "CTGGTC": 2555, + "CTGGTG": 2556, + "CTGGCA": 2557, + "CTGGCT": 2558, + "CTGGCC": 2559, + "CTGGCG": 2560, + "CTGGGA": 2561, + "CTGGGT": 2562, + "CTGGGC": 2563, + "CTGGGG": 2564, + "CCAAAA": 2565, + "CCAAAT": 2566, + "CCAAAC": 2567, + "CCAAAG": 2568, + "CCAATA": 2569, + "CCAATT": 2570, + "CCAATC": 2571, + "CCAATG": 2572, + "CCAACA": 2573, + "CCAACT": 2574, + "CCAACC": 2575, + "CCAACG": 2576, + "CCAAGA": 2577, + "CCAAGT": 2578, + "CCAAGC": 2579, + "CCAAGG": 2580, + "CCATAA": 2581, + "CCATAT": 2582, + "CCATAC": 2583, + "CCATAG": 2584, + "CCATTA": 2585, + "CCATTT": 2586, + "CCATTC": 2587, + "CCATTG": 2588, + "CCATCA": 2589, + "CCATCT": 2590, + "CCATCC": 2591, + "CCATCG": 2592, + "CCATGA": 2593, + "CCATGT": 2594, + "CCATGC": 2595, + "CCATGG": 2596, + "CCACAA": 2597, + "CCACAT": 2598, + "CCACAC": 2599, + "CCACAG": 2600, + "CCACTA": 2601, + "CCACTT": 2602, + "CCACTC": 2603, + "CCACTG": 2604, + "CCACCA": 2605, + "CCACCT": 2606, + "CCACCC": 2607, + "CCACCG": 2608, + "CCACGA": 2609, + "CCACGT": 2610, + "CCACGC": 2611, + "CCACGG": 2612, + "CCAGAA": 2613, + "CCAGAT": 2614, + "CCAGAC": 2615, + "CCAGAG": 2616, + "CCAGTA": 2617, + "CCAGTT": 2618, + "CCAGTC": 2619, + "CCAGTG": 2620, + "CCAGCA": 2621, + "CCAGCT": 2622, + "CCAGCC": 2623, + "CCAGCG": 2624, + "CCAGGA": 2625, + "CCAGGT": 2626, + "CCAGGC": 2627, + "CCAGGG": 2628, + "CCTAAA": 2629, + "CCTAAT": 2630, + "CCTAAC": 2631, + "CCTAAG": 2632, + "CCTATA": 2633, + "CCTATT": 2634, + "CCTATC": 2635, + "CCTATG": 2636, + "CCTACA": 2637, + "CCTACT": 2638, + "CCTACC": 2639, + "CCTACG": 2640, + "CCTAGA": 2641, + "CCTAGT": 2642, + "CCTAGC": 2643, + "CCTAGG": 2644, + "CCTTAA": 2645, + "CCTTAT": 2646, + "CCTTAC": 2647, + "CCTTAG": 2648, + "CCTTTA": 2649, + "CCTTTT": 2650, + "CCTTTC": 2651, + "CCTTTG": 2652, + "CCTTCA": 2653, + "CCTTCT": 2654, + "CCTTCC": 2655, + "CCTTCG": 2656, + "CCTTGA": 2657, + "CCTTGT": 2658, + "CCTTGC": 2659, + "CCTTGG": 2660, + "CCTCAA": 2661, + "CCTCAT": 2662, + "CCTCAC": 2663, + "CCTCAG": 2664, + "CCTCTA": 2665, + "CCTCTT": 2666, + "CCTCTC": 2667, + "CCTCTG": 2668, + "CCTCCA": 2669, + "CCTCCT": 2670, + "CCTCCC": 2671, + "CCTCCG": 2672, + "CCTCGA": 2673, + "CCTCGT": 2674, + "CCTCGC": 2675, + "CCTCGG": 2676, + "CCTGAA": 2677, + "CCTGAT": 2678, + "CCTGAC": 2679, + "CCTGAG": 2680, + "CCTGTA": 2681, + "CCTGTT": 2682, + "CCTGTC": 2683, + "CCTGTG": 2684, + "CCTGCA": 2685, + "CCTGCT": 2686, + "CCTGCC": 2687, + "CCTGCG": 2688, + "CCTGGA": 2689, + "CCTGGT": 2690, + "CCTGGC": 2691, + "CCTGGG": 2692, + "CCCAAA": 2693, + "CCCAAT": 2694, + "CCCAAC": 2695, + "CCCAAG": 2696, + "CCCATA": 2697, + "CCCATT": 2698, + "CCCATC": 2699, + "CCCATG": 2700, + "CCCACA": 2701, + "CCCACT": 2702, + "CCCACC": 2703, + "CCCACG": 2704, + "CCCAGA": 2705, + "CCCAGT": 2706, + "CCCAGC": 2707, + "CCCAGG": 2708, + "CCCTAA": 2709, + "CCCTAT": 2710, + "CCCTAC": 2711, + "CCCTAG": 2712, + "CCCTTA": 2713, + "CCCTTT": 2714, + "CCCTTC": 2715, + "CCCTTG": 2716, + "CCCTCA": 2717, + "CCCTCT": 2718, + "CCCTCC": 2719, + "CCCTCG": 2720, + "CCCTGA": 2721, + "CCCTGT": 2722, + "CCCTGC": 2723, + "CCCTGG": 2724, + "CCCCAA": 2725, + "CCCCAT": 2726, + "CCCCAC": 2727, + "CCCCAG": 2728, + "CCCCTA": 2729, + "CCCCTT": 2730, + "CCCCTC": 2731, + "CCCCTG": 2732, + "CCCCCA": 2733, + "CCCCCT": 2734, + "CCCCCC": 2735, + "CCCCCG": 2736, + "CCCCGA": 2737, + "CCCCGT": 2738, + "CCCCGC": 2739, + "CCCCGG": 2740, + "CCCGAA": 2741, + "CCCGAT": 2742, + "CCCGAC": 2743, + "CCCGAG": 2744, + "CCCGTA": 2745, + "CCCGTT": 2746, + "CCCGTC": 2747, + "CCCGTG": 2748, + "CCCGCA": 2749, + "CCCGCT": 2750, + "CCCGCC": 2751, + "CCCGCG": 2752, + "CCCGGA": 2753, + "CCCGGT": 2754, + "CCCGGC": 2755, + "CCCGGG": 2756, + "CCGAAA": 2757, + "CCGAAT": 2758, + "CCGAAC": 2759, + "CCGAAG": 2760, + "CCGATA": 2761, + "CCGATT": 2762, + "CCGATC": 2763, + "CCGATG": 2764, + "CCGACA": 2765, + "CCGACT": 2766, + "CCGACC": 2767, + "CCGACG": 2768, + "CCGAGA": 2769, + "CCGAGT": 2770, + "CCGAGC": 2771, + "CCGAGG": 2772, + "CCGTAA": 2773, + "CCGTAT": 2774, + "CCGTAC": 2775, + "CCGTAG": 2776, + "CCGTTA": 2777, + "CCGTTT": 2778, + "CCGTTC": 2779, + "CCGTTG": 2780, + "CCGTCA": 2781, + "CCGTCT": 2782, + "CCGTCC": 2783, + "CCGTCG": 2784, + "CCGTGA": 2785, + "CCGTGT": 2786, + "CCGTGC": 2787, + "CCGTGG": 2788, + "CCGCAA": 2789, + "CCGCAT": 2790, + "CCGCAC": 2791, + "CCGCAG": 2792, + "CCGCTA": 2793, + "CCGCTT": 2794, + "CCGCTC": 2795, + "CCGCTG": 2796, + "CCGCCA": 2797, + "CCGCCT": 2798, + "CCGCCC": 2799, + "CCGCCG": 2800, + "CCGCGA": 2801, + "CCGCGT": 2802, + "CCGCGC": 2803, + "CCGCGG": 2804, + "CCGGAA": 2805, + "CCGGAT": 2806, + "CCGGAC": 2807, + "CCGGAG": 2808, + "CCGGTA": 2809, + "CCGGTT": 2810, + "CCGGTC": 2811, + "CCGGTG": 2812, + "CCGGCA": 2813, + "CCGGCT": 2814, + "CCGGCC": 2815, + "CCGGCG": 2816, + "CCGGGA": 2817, + "CCGGGT": 2818, + "CCGGGC": 2819, + "CCGGGG": 2820, + "CGAAAA": 2821, + "CGAAAT": 2822, + "CGAAAC": 2823, + "CGAAAG": 2824, + "CGAATA": 2825, + "CGAATT": 2826, + "CGAATC": 2827, + "CGAATG": 2828, + "CGAACA": 2829, + "CGAACT": 2830, + "CGAACC": 2831, + "CGAACG": 2832, + "CGAAGA": 2833, + "CGAAGT": 2834, + "CGAAGC": 2835, + "CGAAGG": 2836, + "CGATAA": 2837, + "CGATAT": 2838, + "CGATAC": 2839, + "CGATAG": 2840, + "CGATTA": 2841, + "CGATTT": 2842, + "CGATTC": 2843, + "CGATTG": 2844, + "CGATCA": 2845, + "CGATCT": 2846, + "CGATCC": 2847, + "CGATCG": 2848, + "CGATGA": 2849, + "CGATGT": 2850, + "CGATGC": 2851, + "CGATGG": 2852, + "CGACAA": 2853, + "CGACAT": 2854, + "CGACAC": 2855, + "CGACAG": 2856, + "CGACTA": 2857, + "CGACTT": 2858, + "CGACTC": 2859, + "CGACTG": 2860, + "CGACCA": 2861, + "CGACCT": 2862, + "CGACCC": 2863, + "CGACCG": 2864, + "CGACGA": 2865, + "CGACGT": 2866, + "CGACGC": 2867, + "CGACGG": 2868, + "CGAGAA": 2869, + "CGAGAT": 2870, + "CGAGAC": 2871, + "CGAGAG": 2872, + "CGAGTA": 2873, + "CGAGTT": 2874, + "CGAGTC": 2875, + "CGAGTG": 2876, + "CGAGCA": 2877, + "CGAGCT": 2878, + "CGAGCC": 2879, + "CGAGCG": 2880, + "CGAGGA": 2881, + "CGAGGT": 2882, + "CGAGGC": 2883, + "CGAGGG": 2884, + "CGTAAA": 2885, + "CGTAAT": 2886, + "CGTAAC": 2887, + "CGTAAG": 2888, + "CGTATA": 2889, + "CGTATT": 2890, + "CGTATC": 2891, + "CGTATG": 2892, + "CGTACA": 2893, + "CGTACT": 2894, + "CGTACC": 2895, + "CGTACG": 2896, + "CGTAGA": 2897, + "CGTAGT": 2898, + "CGTAGC": 2899, + "CGTAGG": 2900, + "CGTTAA": 2901, + "CGTTAT": 2902, + "CGTTAC": 2903, + "CGTTAG": 2904, + "CGTTTA": 2905, + "CGTTTT": 2906, + "CGTTTC": 2907, + "CGTTTG": 2908, + "CGTTCA": 2909, + "CGTTCT": 2910, + "CGTTCC": 2911, + "CGTTCG": 2912, + "CGTTGA": 2913, + "CGTTGT": 2914, + "CGTTGC": 2915, + "CGTTGG": 2916, + "CGTCAA": 2917, + "CGTCAT": 2918, + "CGTCAC": 2919, + "CGTCAG": 2920, + "CGTCTA": 2921, + "CGTCTT": 2922, + "CGTCTC": 2923, + "CGTCTG": 2924, + "CGTCCA": 2925, + "CGTCCT": 2926, + "CGTCCC": 2927, + "CGTCCG": 2928, + "CGTCGA": 2929, + "CGTCGT": 2930, + "CGTCGC": 2931, + "CGTCGG": 2932, + "CGTGAA": 2933, + "CGTGAT": 2934, + "CGTGAC": 2935, + "CGTGAG": 2936, + "CGTGTA": 2937, + "CGTGTT": 2938, + "CGTGTC": 2939, + "CGTGTG": 2940, + "CGTGCA": 2941, + "CGTGCT": 2942, + "CGTGCC": 2943, + "CGTGCG": 2944, + "CGTGGA": 2945, + "CGTGGT": 2946, + "CGTGGC": 2947, + "CGTGGG": 2948, + "CGCAAA": 2949, + "CGCAAT": 2950, + "CGCAAC": 2951, + "CGCAAG": 2952, + "CGCATA": 2953, + "CGCATT": 2954, + "CGCATC": 2955, + "CGCATG": 2956, + "CGCACA": 2957, + "CGCACT": 2958, + "CGCACC": 2959, + "CGCACG": 2960, + "CGCAGA": 2961, + "CGCAGT": 2962, + "CGCAGC": 2963, + "CGCAGG": 2964, + "CGCTAA": 2965, + "CGCTAT": 2966, + "CGCTAC": 2967, + "CGCTAG": 2968, + "CGCTTA": 2969, + "CGCTTT": 2970, + "CGCTTC": 2971, + "CGCTTG": 2972, + "CGCTCA": 2973, + "CGCTCT": 2974, + "CGCTCC": 2975, + "CGCTCG": 2976, + "CGCTGA": 2977, + "CGCTGT": 2978, + "CGCTGC": 2979, + "CGCTGG": 2980, + "CGCCAA": 2981, + "CGCCAT": 2982, + "CGCCAC": 2983, + "CGCCAG": 2984, + "CGCCTA": 2985, + "CGCCTT": 2986, + "CGCCTC": 2987, + "CGCCTG": 2988, + "CGCCCA": 2989, + "CGCCCT": 2990, + "CGCCCC": 2991, + "CGCCCG": 2992, + "CGCCGA": 2993, + "CGCCGT": 2994, + "CGCCGC": 2995, + "CGCCGG": 2996, + "CGCGAA": 2997, + "CGCGAT": 2998, + "CGCGAC": 2999, + "CGCGAG": 3000, + "CGCGTA": 3001, + "CGCGTT": 3002, + "CGCGTC": 3003, + "CGCGTG": 3004, + "CGCGCA": 3005, + "CGCGCT": 3006, + "CGCGCC": 3007, + "CGCGCG": 3008, + "CGCGGA": 3009, + "CGCGGT": 3010, + "CGCGGC": 3011, + "CGCGGG": 3012, + "CGGAAA": 3013, + "CGGAAT": 3014, + "CGGAAC": 3015, + "CGGAAG": 3016, + "CGGATA": 3017, + "CGGATT": 3018, + "CGGATC": 3019, + "CGGATG": 3020, + "CGGACA": 3021, + "CGGACT": 3022, + "CGGACC": 3023, + "CGGACG": 3024, + "CGGAGA": 3025, + "CGGAGT": 3026, + "CGGAGC": 3027, + "CGGAGG": 3028, + "CGGTAA": 3029, + "CGGTAT": 3030, + "CGGTAC": 3031, + "CGGTAG": 3032, + "CGGTTA": 3033, + "CGGTTT": 3034, + "CGGTTC": 3035, + "CGGTTG": 3036, + "CGGTCA": 3037, + "CGGTCT": 3038, + "CGGTCC": 3039, + "CGGTCG": 3040, + "CGGTGA": 3041, + "CGGTGT": 3042, + "CGGTGC": 3043, + "CGGTGG": 3044, + "CGGCAA": 3045, + "CGGCAT": 3046, + "CGGCAC": 3047, + "CGGCAG": 3048, + "CGGCTA": 3049, + "CGGCTT": 3050, + "CGGCTC": 3051, + "CGGCTG": 3052, + "CGGCCA": 3053, + "CGGCCT": 3054, + "CGGCCC": 3055, + "CGGCCG": 3056, + "CGGCGA": 3057, + "CGGCGT": 3058, + "CGGCGC": 3059, + "CGGCGG": 3060, + "CGGGAA": 3061, + "CGGGAT": 3062, + "CGGGAC": 3063, + "CGGGAG": 3064, + "CGGGTA": 3065, + "CGGGTT": 3066, + "CGGGTC": 3067, + "CGGGTG": 3068, + "CGGGCA": 3069, + "CGGGCT": 3070, + "CGGGCC": 3071, + "CGGGCG": 3072, + "CGGGGA": 3073, + "CGGGGT": 3074, + "CGGGGC": 3075, + "CGGGGG": 3076, + "GAAAAA": 3077, + "GAAAAT": 3078, + "GAAAAC": 3079, + "GAAAAG": 3080, + "GAAATA": 3081, + "GAAATT": 3082, + "GAAATC": 3083, + "GAAATG": 3084, + "GAAACA": 3085, + "GAAACT": 3086, + "GAAACC": 3087, + "GAAACG": 3088, + "GAAAGA": 3089, + "GAAAGT": 3090, + "GAAAGC": 3091, + "GAAAGG": 3092, + "GAATAA": 3093, + "GAATAT": 3094, + "GAATAC": 3095, + "GAATAG": 3096, + "GAATTA": 3097, + "GAATTT": 3098, + "GAATTC": 3099, + "GAATTG": 3100, + "GAATCA": 3101, + "GAATCT": 3102, + "GAATCC": 3103, + "GAATCG": 3104, + "GAATGA": 3105, + "GAATGT": 3106, + "GAATGC": 3107, + "GAATGG": 3108, + "GAACAA": 3109, + "GAACAT": 3110, + "GAACAC": 3111, + "GAACAG": 3112, + "GAACTA": 3113, + "GAACTT": 3114, + "GAACTC": 3115, + "GAACTG": 3116, + "GAACCA": 3117, + "GAACCT": 3118, + "GAACCC": 3119, + "GAACCG": 3120, + "GAACGA": 3121, + "GAACGT": 3122, + "GAACGC": 3123, + "GAACGG": 3124, + "GAAGAA": 3125, + "GAAGAT": 3126, + "GAAGAC": 3127, + "GAAGAG": 3128, + "GAAGTA": 3129, + "GAAGTT": 3130, + "GAAGTC": 3131, + "GAAGTG": 3132, + "GAAGCA": 3133, + "GAAGCT": 3134, + "GAAGCC": 3135, + "GAAGCG": 3136, + "GAAGGA": 3137, + "GAAGGT": 3138, + "GAAGGC": 3139, + "GAAGGG": 3140, + "GATAAA": 3141, + "GATAAT": 3142, + "GATAAC": 3143, + "GATAAG": 3144, + "GATATA": 3145, + "GATATT": 3146, + "GATATC": 3147, + "GATATG": 3148, + "GATACA": 3149, + "GATACT": 3150, + "GATACC": 3151, + "GATACG": 3152, + "GATAGA": 3153, + "GATAGT": 3154, + "GATAGC": 3155, + "GATAGG": 3156, + "GATTAA": 3157, + "GATTAT": 3158, + "GATTAC": 3159, + "GATTAG": 3160, + "GATTTA": 3161, + "GATTTT": 3162, + "GATTTC": 3163, + "GATTTG": 3164, + "GATTCA": 3165, + "GATTCT": 3166, + "GATTCC": 3167, + "GATTCG": 3168, + "GATTGA": 3169, + "GATTGT": 3170, + "GATTGC": 3171, + "GATTGG": 3172, + "GATCAA": 3173, + "GATCAT": 3174, + "GATCAC": 3175, + "GATCAG": 3176, + "GATCTA": 3177, + "GATCTT": 3178, + "GATCTC": 3179, + "GATCTG": 3180, + "GATCCA": 3181, + "GATCCT": 3182, + "GATCCC": 3183, + "GATCCG": 3184, + "GATCGA": 3185, + "GATCGT": 3186, + "GATCGC": 3187, + "GATCGG": 3188, + "GATGAA": 3189, + "GATGAT": 3190, + "GATGAC": 3191, + "GATGAG": 3192, + "GATGTA": 3193, + "GATGTT": 3194, + "GATGTC": 3195, + "GATGTG": 3196, + "GATGCA": 3197, + "GATGCT": 3198, + "GATGCC": 3199, + "GATGCG": 3200, + "GATGGA": 3201, + "GATGGT": 3202, + "GATGGC": 3203, + "GATGGG": 3204, + "GACAAA": 3205, + "GACAAT": 3206, + "GACAAC": 3207, + "GACAAG": 3208, + "GACATA": 3209, + "GACATT": 3210, + "GACATC": 3211, + "GACATG": 3212, + "GACACA": 3213, + "GACACT": 3214, + "GACACC": 3215, + "GACACG": 3216, + "GACAGA": 3217, + "GACAGT": 3218, + "GACAGC": 3219, + "GACAGG": 3220, + "GACTAA": 3221, + "GACTAT": 3222, + "GACTAC": 3223, + "GACTAG": 3224, + "GACTTA": 3225, + "GACTTT": 3226, + "GACTTC": 3227, + "GACTTG": 3228, + "GACTCA": 3229, + "GACTCT": 3230, + "GACTCC": 3231, + "GACTCG": 3232, + "GACTGA": 3233, + "GACTGT": 3234, + "GACTGC": 3235, + "GACTGG": 3236, + "GACCAA": 3237, + "GACCAT": 3238, + "GACCAC": 3239, + "GACCAG": 3240, + "GACCTA": 3241, + "GACCTT": 3242, + "GACCTC": 3243, + "GACCTG": 3244, + "GACCCA": 3245, + "GACCCT": 3246, + "GACCCC": 3247, + "GACCCG": 3248, + "GACCGA": 3249, + "GACCGT": 3250, + "GACCGC": 3251, + "GACCGG": 3252, + "GACGAA": 3253, + "GACGAT": 3254, + "GACGAC": 3255, + "GACGAG": 3256, + "GACGTA": 3257, + "GACGTT": 3258, + "GACGTC": 3259, + "GACGTG": 3260, + "GACGCA": 3261, + "GACGCT": 3262, + "GACGCC": 3263, + "GACGCG": 3264, + "GACGGA": 3265, + "GACGGT": 3266, + "GACGGC": 3267, + "GACGGG": 3268, + "GAGAAA": 3269, + "GAGAAT": 3270, + "GAGAAC": 3271, + "GAGAAG": 3272, + "GAGATA": 3273, + "GAGATT": 3274, + "GAGATC": 3275, + "GAGATG": 3276, + "GAGACA": 3277, + "GAGACT": 3278, + "GAGACC": 3279, + "GAGACG": 3280, + "GAGAGA": 3281, + "GAGAGT": 3282, + "GAGAGC": 3283, + "GAGAGG": 3284, + "GAGTAA": 3285, + "GAGTAT": 3286, + "GAGTAC": 3287, + "GAGTAG": 3288, + "GAGTTA": 3289, + "GAGTTT": 3290, + "GAGTTC": 3291, + "GAGTTG": 3292, + "GAGTCA": 3293, + "GAGTCT": 3294, + "GAGTCC": 3295, + "GAGTCG": 3296, + "GAGTGA": 3297, + "GAGTGT": 3298, + "GAGTGC": 3299, + "GAGTGG": 3300, + "GAGCAA": 3301, + "GAGCAT": 3302, + "GAGCAC": 3303, + "GAGCAG": 3304, + "GAGCTA": 3305, + "GAGCTT": 3306, + "GAGCTC": 3307, + "GAGCTG": 3308, + "GAGCCA": 3309, + "GAGCCT": 3310, + "GAGCCC": 3311, + "GAGCCG": 3312, + "GAGCGA": 3313, + "GAGCGT": 3314, + "GAGCGC": 3315, + "GAGCGG": 3316, + "GAGGAA": 3317, + "GAGGAT": 3318, + "GAGGAC": 3319, + "GAGGAG": 3320, + "GAGGTA": 3321, + "GAGGTT": 3322, + "GAGGTC": 3323, + "GAGGTG": 3324, + "GAGGCA": 3325, + "GAGGCT": 3326, + "GAGGCC": 3327, + "GAGGCG": 3328, + "GAGGGA": 3329, + "GAGGGT": 3330, + "GAGGGC": 3331, + "GAGGGG": 3332, + "GTAAAA": 3333, + "GTAAAT": 3334, + "GTAAAC": 3335, + "GTAAAG": 3336, + "GTAATA": 3337, + "GTAATT": 3338, + "GTAATC": 3339, + "GTAATG": 3340, + "GTAACA": 3341, + "GTAACT": 3342, + "GTAACC": 3343, + "GTAACG": 3344, + "GTAAGA": 3345, + "GTAAGT": 3346, + "GTAAGC": 3347, + "GTAAGG": 3348, + "GTATAA": 3349, + "GTATAT": 3350, + "GTATAC": 3351, + "GTATAG": 3352, + "GTATTA": 3353, + "GTATTT": 3354, + "GTATTC": 3355, + "GTATTG": 3356, + "GTATCA": 3357, + "GTATCT": 3358, + "GTATCC": 3359, + "GTATCG": 3360, + "GTATGA": 3361, + "GTATGT": 3362, + "GTATGC": 3363, + "GTATGG": 3364, + "GTACAA": 3365, + "GTACAT": 3366, + "GTACAC": 3367, + "GTACAG": 3368, + "GTACTA": 3369, + "GTACTT": 3370, + "GTACTC": 3371, + "GTACTG": 3372, + "GTACCA": 3373, + "GTACCT": 3374, + "GTACCC": 3375, + "GTACCG": 3376, + "GTACGA": 3377, + "GTACGT": 3378, + "GTACGC": 3379, + "GTACGG": 3380, + "GTAGAA": 3381, + "GTAGAT": 3382, + "GTAGAC": 3383, + "GTAGAG": 3384, + "GTAGTA": 3385, + "GTAGTT": 3386, + "GTAGTC": 3387, + "GTAGTG": 3388, + "GTAGCA": 3389, + "GTAGCT": 3390, + "GTAGCC": 3391, + "GTAGCG": 3392, + "GTAGGA": 3393, + "GTAGGT": 3394, + "GTAGGC": 3395, + "GTAGGG": 3396, + "GTTAAA": 3397, + "GTTAAT": 3398, + "GTTAAC": 3399, + "GTTAAG": 3400, + "GTTATA": 3401, + "GTTATT": 3402, + "GTTATC": 3403, + "GTTATG": 3404, + "GTTACA": 3405, + "GTTACT": 3406, + "GTTACC": 3407, + "GTTACG": 3408, + "GTTAGA": 3409, + "GTTAGT": 3410, + "GTTAGC": 3411, + "GTTAGG": 3412, + "GTTTAA": 3413, + "GTTTAT": 3414, + "GTTTAC": 3415, + "GTTTAG": 3416, + "GTTTTA": 3417, + "GTTTTT": 3418, + "GTTTTC": 3419, + "GTTTTG": 3420, + "GTTTCA": 3421, + "GTTTCT": 3422, + "GTTTCC": 3423, + "GTTTCG": 3424, + "GTTTGA": 3425, + "GTTTGT": 3426, + "GTTTGC": 3427, + "GTTTGG": 3428, + "GTTCAA": 3429, + "GTTCAT": 3430, + "GTTCAC": 3431, + "GTTCAG": 3432, + "GTTCTA": 3433, + "GTTCTT": 3434, + "GTTCTC": 3435, + "GTTCTG": 3436, + "GTTCCA": 3437, + "GTTCCT": 3438, + "GTTCCC": 3439, + "GTTCCG": 3440, + "GTTCGA": 3441, + "GTTCGT": 3442, + "GTTCGC": 3443, + "GTTCGG": 3444, + "GTTGAA": 3445, + "GTTGAT": 3446, + "GTTGAC": 3447, + "GTTGAG": 3448, + "GTTGTA": 3449, + "GTTGTT": 3450, + "GTTGTC": 3451, + "GTTGTG": 3452, + "GTTGCA": 3453, + "GTTGCT": 3454, + "GTTGCC": 3455, + "GTTGCG": 3456, + "GTTGGA": 3457, + "GTTGGT": 3458, + "GTTGGC": 3459, + "GTTGGG": 3460, + "GTCAAA": 3461, + "GTCAAT": 3462, + "GTCAAC": 3463, + "GTCAAG": 3464, + "GTCATA": 3465, + "GTCATT": 3466, + "GTCATC": 3467, + "GTCATG": 3468, + "GTCACA": 3469, + "GTCACT": 3470, + "GTCACC": 3471, + "GTCACG": 3472, + "GTCAGA": 3473, + "GTCAGT": 3474, + "GTCAGC": 3475, + "GTCAGG": 3476, + "GTCTAA": 3477, + "GTCTAT": 3478, + "GTCTAC": 3479, + "GTCTAG": 3480, + "GTCTTA": 3481, + "GTCTTT": 3482, + "GTCTTC": 3483, + "GTCTTG": 3484, + "GTCTCA": 3485, + "GTCTCT": 3486, + "GTCTCC": 3487, + "GTCTCG": 3488, + "GTCTGA": 3489, + "GTCTGT": 3490, + "GTCTGC": 3491, + "GTCTGG": 3492, + "GTCCAA": 3493, + "GTCCAT": 3494, + "GTCCAC": 3495, + "GTCCAG": 3496, + "GTCCTA": 3497, + "GTCCTT": 3498, + "GTCCTC": 3499, + "GTCCTG": 3500, + "GTCCCA": 3501, + "GTCCCT": 3502, + "GTCCCC": 3503, + "GTCCCG": 3504, + "GTCCGA": 3505, + "GTCCGT": 3506, + "GTCCGC": 3507, + "GTCCGG": 3508, + "GTCGAA": 3509, + "GTCGAT": 3510, + "GTCGAC": 3511, + "GTCGAG": 3512, + "GTCGTA": 3513, + "GTCGTT": 3514, + "GTCGTC": 3515, + "GTCGTG": 3516, + "GTCGCA": 3517, + "GTCGCT": 3518, + "GTCGCC": 3519, + "GTCGCG": 3520, + "GTCGGA": 3521, + "GTCGGT": 3522, + "GTCGGC": 3523, + "GTCGGG": 3524, + "GTGAAA": 3525, + "GTGAAT": 3526, + "GTGAAC": 3527, + "GTGAAG": 3528, + "GTGATA": 3529, + "GTGATT": 3530, + "GTGATC": 3531, + "GTGATG": 3532, + "GTGACA": 3533, + "GTGACT": 3534, + "GTGACC": 3535, + "GTGACG": 3536, + "GTGAGA": 3537, + "GTGAGT": 3538, + "GTGAGC": 3539, + "GTGAGG": 3540, + "GTGTAA": 3541, + "GTGTAT": 3542, + "GTGTAC": 3543, + "GTGTAG": 3544, + "GTGTTA": 3545, + "GTGTTT": 3546, + "GTGTTC": 3547, + "GTGTTG": 3548, + "GTGTCA": 3549, + "GTGTCT": 3550, + "GTGTCC": 3551, + "GTGTCG": 3552, + "GTGTGA": 3553, + "GTGTGT": 3554, + "GTGTGC": 3555, + "GTGTGG": 3556, + "GTGCAA": 3557, + "GTGCAT": 3558, + "GTGCAC": 3559, + "GTGCAG": 3560, + "GTGCTA": 3561, + "GTGCTT": 3562, + "GTGCTC": 3563, + "GTGCTG": 3564, + "GTGCCA": 3565, + "GTGCCT": 3566, + "GTGCCC": 3567, + "GTGCCG": 3568, + "GTGCGA": 3569, + "GTGCGT": 3570, + "GTGCGC": 3571, + "GTGCGG": 3572, + "GTGGAA": 3573, + "GTGGAT": 3574, + "GTGGAC": 3575, + "GTGGAG": 3576, + "GTGGTA": 3577, + "GTGGTT": 3578, + "GTGGTC": 3579, + "GTGGTG": 3580, + "GTGGCA": 3581, + "GTGGCT": 3582, + "GTGGCC": 3583, + "GTGGCG": 3584, + "GTGGGA": 3585, + "GTGGGT": 3586, + "GTGGGC": 3587, + "GTGGGG": 3588, + "GCAAAA": 3589, + "GCAAAT": 3590, + "GCAAAC": 3591, + "GCAAAG": 3592, + "GCAATA": 3593, + "GCAATT": 3594, + "GCAATC": 3595, + "GCAATG": 3596, + "GCAACA": 3597, + "GCAACT": 3598, + "GCAACC": 3599, + "GCAACG": 3600, + "GCAAGA": 3601, + "GCAAGT": 3602, + "GCAAGC": 3603, + "GCAAGG": 3604, + "GCATAA": 3605, + "GCATAT": 3606, + "GCATAC": 3607, + "GCATAG": 3608, + "GCATTA": 3609, + "GCATTT": 3610, + "GCATTC": 3611, + "GCATTG": 3612, + "GCATCA": 3613, + "GCATCT": 3614, + "GCATCC": 3615, + "GCATCG": 3616, + "GCATGA": 3617, + "GCATGT": 3618, + "GCATGC": 3619, + "GCATGG": 3620, + "GCACAA": 3621, + "GCACAT": 3622, + "GCACAC": 3623, + "GCACAG": 3624, + "GCACTA": 3625, + "GCACTT": 3626, + "GCACTC": 3627, + "GCACTG": 3628, + "GCACCA": 3629, + "GCACCT": 3630, + "GCACCC": 3631, + "GCACCG": 3632, + "GCACGA": 3633, + "GCACGT": 3634, + "GCACGC": 3635, + "GCACGG": 3636, + "GCAGAA": 3637, + "GCAGAT": 3638, + "GCAGAC": 3639, + "GCAGAG": 3640, + "GCAGTA": 3641, + "GCAGTT": 3642, + "GCAGTC": 3643, + "GCAGTG": 3644, + "GCAGCA": 3645, + "GCAGCT": 3646, + "GCAGCC": 3647, + "GCAGCG": 3648, + "GCAGGA": 3649, + "GCAGGT": 3650, + "GCAGGC": 3651, + "GCAGGG": 3652, + "GCTAAA": 3653, + "GCTAAT": 3654, + "GCTAAC": 3655, + "GCTAAG": 3656, + "GCTATA": 3657, + "GCTATT": 3658, + "GCTATC": 3659, + "GCTATG": 3660, + "GCTACA": 3661, + "GCTACT": 3662, + "GCTACC": 3663, + "GCTACG": 3664, + "GCTAGA": 3665, + "GCTAGT": 3666, + "GCTAGC": 3667, + "GCTAGG": 3668, + "GCTTAA": 3669, + "GCTTAT": 3670, + "GCTTAC": 3671, + "GCTTAG": 3672, + "GCTTTA": 3673, + "GCTTTT": 3674, + "GCTTTC": 3675, + "GCTTTG": 3676, + "GCTTCA": 3677, + "GCTTCT": 3678, + "GCTTCC": 3679, + "GCTTCG": 3680, + "GCTTGA": 3681, + "GCTTGT": 3682, + "GCTTGC": 3683, + "GCTTGG": 3684, + "GCTCAA": 3685, + "GCTCAT": 3686, + "GCTCAC": 3687, + "GCTCAG": 3688, + "GCTCTA": 3689, + "GCTCTT": 3690, + "GCTCTC": 3691, + "GCTCTG": 3692, + "GCTCCA": 3693, + "GCTCCT": 3694, + "GCTCCC": 3695, + "GCTCCG": 3696, + "GCTCGA": 3697, + "GCTCGT": 3698, + "GCTCGC": 3699, + "GCTCGG": 3700, + "GCTGAA": 3701, + "GCTGAT": 3702, + "GCTGAC": 3703, + "GCTGAG": 3704, + "GCTGTA": 3705, + "GCTGTT": 3706, + "GCTGTC": 3707, + "GCTGTG": 3708, + "GCTGCA": 3709, + "GCTGCT": 3710, + "GCTGCC": 3711, + "GCTGCG": 3712, + "GCTGGA": 3713, + "GCTGGT": 3714, + "GCTGGC": 3715, + "GCTGGG": 3716, + "GCCAAA": 3717, + "GCCAAT": 3718, + "GCCAAC": 3719, + "GCCAAG": 3720, + "GCCATA": 3721, + "GCCATT": 3722, + "GCCATC": 3723, + "GCCATG": 3724, + "GCCACA": 3725, + "GCCACT": 3726, + "GCCACC": 3727, + "GCCACG": 3728, + "GCCAGA": 3729, + "GCCAGT": 3730, + "GCCAGC": 3731, + "GCCAGG": 3732, + "GCCTAA": 3733, + "GCCTAT": 3734, + "GCCTAC": 3735, + "GCCTAG": 3736, + "GCCTTA": 3737, + "GCCTTT": 3738, + "GCCTTC": 3739, + "GCCTTG": 3740, + "GCCTCA": 3741, + "GCCTCT": 3742, + "GCCTCC": 3743, + "GCCTCG": 3744, + "GCCTGA": 3745, + "GCCTGT": 3746, + "GCCTGC": 3747, + "GCCTGG": 3748, + "GCCCAA": 3749, + "GCCCAT": 3750, + "GCCCAC": 3751, + "GCCCAG": 3752, + "GCCCTA": 3753, + "GCCCTT": 3754, + "GCCCTC": 3755, + "GCCCTG": 3756, + "GCCCCA": 3757, + "GCCCCT": 3758, + "GCCCCC": 3759, + "GCCCCG": 3760, + "GCCCGA": 3761, + "GCCCGT": 3762, + "GCCCGC": 3763, + "GCCCGG": 3764, + "GCCGAA": 3765, + "GCCGAT": 3766, + "GCCGAC": 3767, + "GCCGAG": 3768, + "GCCGTA": 3769, + "GCCGTT": 3770, + "GCCGTC": 3771, + "GCCGTG": 3772, + "GCCGCA": 3773, + "GCCGCT": 3774, + "GCCGCC": 3775, + "GCCGCG": 3776, + "GCCGGA": 3777, + "GCCGGT": 3778, + "GCCGGC": 3779, + "GCCGGG": 3780, + "GCGAAA": 3781, + "GCGAAT": 3782, + "GCGAAC": 3783, + "GCGAAG": 3784, + "GCGATA": 3785, + "GCGATT": 3786, + "GCGATC": 3787, + "GCGATG": 3788, + "GCGACA": 3789, + "GCGACT": 3790, + "GCGACC": 3791, + "GCGACG": 3792, + "GCGAGA": 3793, + "GCGAGT": 3794, + "GCGAGC": 3795, + "GCGAGG": 3796, + "GCGTAA": 3797, + "GCGTAT": 3798, + "GCGTAC": 3799, + "GCGTAG": 3800, + "GCGTTA": 3801, + "GCGTTT": 3802, + "GCGTTC": 3803, + "GCGTTG": 3804, + "GCGTCA": 3805, + "GCGTCT": 3806, + "GCGTCC": 3807, + "GCGTCG": 3808, + "GCGTGA": 3809, + "GCGTGT": 3810, + "GCGTGC": 3811, + "GCGTGG": 3812, + "GCGCAA": 3813, + "GCGCAT": 3814, + "GCGCAC": 3815, + "GCGCAG": 3816, + "GCGCTA": 3817, + "GCGCTT": 3818, + "GCGCTC": 3819, + "GCGCTG": 3820, + "GCGCCA": 3821, + "GCGCCT": 3822, + "GCGCCC": 3823, + "GCGCCG": 3824, + "GCGCGA": 3825, + "GCGCGT": 3826, + "GCGCGC": 3827, + "GCGCGG": 3828, + "GCGGAA": 3829, + "GCGGAT": 3830, + "GCGGAC": 3831, + "GCGGAG": 3832, + "GCGGTA": 3833, + "GCGGTT": 3834, + "GCGGTC": 3835, + "GCGGTG": 3836, + "GCGGCA": 3837, + "GCGGCT": 3838, + "GCGGCC": 3839, + "GCGGCG": 3840, + "GCGGGA": 3841, + "GCGGGT": 3842, + "GCGGGC": 3843, + "GCGGGG": 3844, + "GGAAAA": 3845, + "GGAAAT": 3846, + "GGAAAC": 3847, + "GGAAAG": 3848, + "GGAATA": 3849, + "GGAATT": 3850, + "GGAATC": 3851, + "GGAATG": 3852, + "GGAACA": 3853, + "GGAACT": 3854, + "GGAACC": 3855, + "GGAACG": 3856, + "GGAAGA": 3857, + "GGAAGT": 3858, + "GGAAGC": 3859, + "GGAAGG": 3860, + "GGATAA": 3861, + "GGATAT": 3862, + "GGATAC": 3863, + "GGATAG": 3864, + "GGATTA": 3865, + "GGATTT": 3866, + "GGATTC": 3867, + "GGATTG": 3868, + "GGATCA": 3869, + "GGATCT": 3870, + "GGATCC": 3871, + "GGATCG": 3872, + "GGATGA": 3873, + "GGATGT": 3874, + "GGATGC": 3875, + "GGATGG": 3876, + "GGACAA": 3877, + "GGACAT": 3878, + "GGACAC": 3879, + "GGACAG": 3880, + "GGACTA": 3881, + "GGACTT": 3882, + "GGACTC": 3883, + "GGACTG": 3884, + "GGACCA": 3885, + "GGACCT": 3886, + "GGACCC": 3887, + "GGACCG": 3888, + "GGACGA": 3889, + "GGACGT": 3890, + "GGACGC": 3891, + "GGACGG": 3892, + "GGAGAA": 3893, + "GGAGAT": 3894, + "GGAGAC": 3895, + "GGAGAG": 3896, + "GGAGTA": 3897, + "GGAGTT": 3898, + "GGAGTC": 3899, + "GGAGTG": 3900, + "GGAGCA": 3901, + "GGAGCT": 3902, + "GGAGCC": 3903, + "GGAGCG": 3904, + "GGAGGA": 3905, + "GGAGGT": 3906, + "GGAGGC": 3907, + "GGAGGG": 3908, + "GGTAAA": 3909, + "GGTAAT": 3910, + "GGTAAC": 3911, + "GGTAAG": 3912, + "GGTATA": 3913, + "GGTATT": 3914, + "GGTATC": 3915, + "GGTATG": 3916, + "GGTACA": 3917, + "GGTACT": 3918, + "GGTACC": 3919, + "GGTACG": 3920, + "GGTAGA": 3921, + "GGTAGT": 3922, + "GGTAGC": 3923, + "GGTAGG": 3924, + "GGTTAA": 3925, + "GGTTAT": 3926, + "GGTTAC": 3927, + "GGTTAG": 3928, + "GGTTTA": 3929, + "GGTTTT": 3930, + "GGTTTC": 3931, + "GGTTTG": 3932, + "GGTTCA": 3933, + "GGTTCT": 3934, + "GGTTCC": 3935, + "GGTTCG": 3936, + "GGTTGA": 3937, + "GGTTGT": 3938, + "GGTTGC": 3939, + "GGTTGG": 3940, + "GGTCAA": 3941, + "GGTCAT": 3942, + "GGTCAC": 3943, + "GGTCAG": 3944, + "GGTCTA": 3945, + "GGTCTT": 3946, + "GGTCTC": 3947, + "GGTCTG": 3948, + "GGTCCA": 3949, + "GGTCCT": 3950, + "GGTCCC": 3951, + "GGTCCG": 3952, + "GGTCGA": 3953, + "GGTCGT": 3954, + "GGTCGC": 3955, + "GGTCGG": 3956, + "GGTGAA": 3957, + "GGTGAT": 3958, + "GGTGAC": 3959, + "GGTGAG": 3960, + "GGTGTA": 3961, + "GGTGTT": 3962, + "GGTGTC": 3963, + "GGTGTG": 3964, + "GGTGCA": 3965, + "GGTGCT": 3966, + "GGTGCC": 3967, + "GGTGCG": 3968, + "GGTGGA": 3969, + "GGTGGT": 3970, + "GGTGGC": 3971, + "GGTGGG": 3972, + "GGCAAA": 3973, + "GGCAAT": 3974, + "GGCAAC": 3975, + "GGCAAG": 3976, + "GGCATA": 3977, + "GGCATT": 3978, + "GGCATC": 3979, + "GGCATG": 3980, + "GGCACA": 3981, + "GGCACT": 3982, + "GGCACC": 3983, + "GGCACG": 3984, + "GGCAGA": 3985, + "GGCAGT": 3986, + "GGCAGC": 3987, + "GGCAGG": 3988, + "GGCTAA": 3989, + "GGCTAT": 3990, + "GGCTAC": 3991, + "GGCTAG": 3992, + "GGCTTA": 3993, + "GGCTTT": 3994, + "GGCTTC": 3995, + "GGCTTG": 3996, + "GGCTCA": 3997, + "GGCTCT": 3998, + "GGCTCC": 3999, + "GGCTCG": 4000, + "GGCTGA": 4001, + "GGCTGT": 4002, + "GGCTGC": 4003, + "GGCTGG": 4004, + "GGCCAA": 4005, + "GGCCAT": 4006, + "GGCCAC": 4007, + "GGCCAG": 4008, + "GGCCTA": 4009, + "GGCCTT": 4010, + "GGCCTC": 4011, + "GGCCTG": 4012, + "GGCCCA": 4013, + "GGCCCT": 4014, + "GGCCCC": 4015, + "GGCCCG": 4016, + "GGCCGA": 4017, + "GGCCGT": 4018, + "GGCCGC": 4019, + "GGCCGG": 4020, + "GGCGAA": 4021, + "GGCGAT": 4022, + "GGCGAC": 4023, + "GGCGAG": 4024, + "GGCGTA": 4025, + "GGCGTT": 4026, + "GGCGTC": 4027, + "GGCGTG": 4028, + "GGCGCA": 4029, + "GGCGCT": 4030, + "GGCGCC": 4031, + "GGCGCG": 4032, + "GGCGGA": 4033, + "GGCGGT": 4034, + "GGCGGC": 4035, + "GGCGGG": 4036, + "GGGAAA": 4037, + "GGGAAT": 4038, + "GGGAAC": 4039, + "GGGAAG": 4040, + "GGGATA": 4041, + "GGGATT": 4042, + "GGGATC": 4043, + "GGGATG": 4044, + "GGGACA": 4045, + "GGGACT": 4046, + "GGGACC": 4047, + "GGGACG": 4048, + "GGGAGA": 4049, + "GGGAGT": 4050, + "GGGAGC": 4051, + "GGGAGG": 4052, + "GGGTAA": 4053, + "GGGTAT": 4054, + "GGGTAC": 4055, + "GGGTAG": 4056, + "GGGTTA": 4057, + "GGGTTT": 4058, + "GGGTTC": 4059, + "GGGTTG": 4060, + "GGGTCA": 4061, + "GGGTCT": 4062, + "GGGTCC": 4063, + "GGGTCG": 4064, + "GGGTGA": 4065, + "GGGTGT": 4066, + "GGGTGC": 4067, + "GGGTGG": 4068, + "GGGCAA": 4069, + "GGGCAT": 4070, + "GGGCAC": 4071, + "GGGCAG": 4072, + "GGGCTA": 4073, + "GGGCTT": 4074, + "GGGCTC": 4075, + "GGGCTG": 4076, + "GGGCCA": 4077, + "GGGCCT": 4078, + "GGGCCC": 4079, + "GGGCCG": 4080, + "GGGCGA": 4081, + "GGGCGT": 4082, + "GGGCGC": 4083, + "GGGCGG": 4084, + "GGGGAA": 4085, + "GGGGAT": 4086, + "GGGGAC": 4087, + "GGGGAG": 4088, + "GGGGTA": 4089, + "GGGGTT": 4090, + "GGGGTC": 4091, + "GGGGTG": 4092, + "GGGGCA": 4093, + "GGGGCT": 4094, + "GGGGCC": 4095, + "GGGGCG": 4096, + "GGGGGA": 4097, + "GGGGGT": 4098, + "GGGGGC": 4099, + "GGGGGG": 4100 + } + } +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/tokenizer_config.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..604e2b1b29af1dcb3920d2828b6a90276e5e3d02 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/tokenizer_config.json @@ -0,0 +1,58 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "[UNK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "[CLS]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "[SEP]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "4": { + "content": "[MASK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "clean_up_tokenization_spaces": true, + "cls_token": "[CLS]", + "do_basic_tokenize": true, + "do_lower_case": false, + "mask_token": "[MASK]", + "max_len": 512, + "model_max_length": 512, + "never_split": null, + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "strip_accents": null, + "tokenize_chinese_chars": true, + "tokenizer_class": "BertTokenizer", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/trainer_state.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..e39d5589bcff5b5370b275aa58d76c9c2b3ae12c --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/trainer_state.json @@ -0,0 +1,1713 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 24.0, + "eval_steps": 500, + "global_step": 6000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.1, + "grad_norm": 0.9416134357452393, + "learning_rate": 5.319148936170213e-05, + "loss": 0.6912, + "step": 25 + }, + { + "epoch": 0.2, + "grad_norm": 0.9283172488212585, + "learning_rate": 0.00010638297872340425, + "loss": 0.6691, + "step": 50 + }, + { + "epoch": 0.3, + "grad_norm": 0.5870445370674133, + "learning_rate": 0.00015957446808510637, + "loss": 0.6164, + "step": 75 + }, + { + "epoch": 0.4, + "grad_norm": 0.33945509791374207, + "learning_rate": 0.0002127659574468085, + "loss": 0.5453, + "step": 100 + }, + { + "epoch": 0.5, + "grad_norm": 0.21052460372447968, + "learning_rate": 0.00026595744680851064, + "loss": 0.5025, + "step": 125 + }, + { + "epoch": 0.6, + "grad_norm": 0.25948837399482727, + "learning_rate": 0.00031914893617021275, + "loss": 0.4881, + "step": 150 + }, + { + "epoch": 0.7, + "grad_norm": 0.6165992021560669, + "learning_rate": 0.0003723404255319149, + "loss": 0.4852, + "step": 175 + }, + { + "epoch": 0.8, + "grad_norm": 0.33296748995780945, + "learning_rate": 0.0003992081821181128, + "loss": 0.4867, + "step": 200 + }, + { + "epoch": 0.9, + "grad_norm": 0.40745800733566284, + "learning_rate": 0.0003975585615308479, + "loss": 0.4767, + "step": 225 + }, + { + "epoch": 1.0, + "grad_norm": 0.8070425391197205, + "learning_rate": 0.00039590894094358297, + "loss": 0.4784, + "step": 250 + }, + { + "epoch": 1.1, + "grad_norm": 1.2249579429626465, + "learning_rate": 0.0003942593203563181, + "loss": 0.4543, + "step": 275 + }, + { + "epoch": 1.2, + "grad_norm": 1.0143898725509644, + "learning_rate": 0.0003926096997690532, + "loss": 0.4319, + "step": 300 + }, + { + "epoch": 1.3, + "grad_norm": 2.3284952640533447, + "learning_rate": 0.00039096007918178817, + "loss": 0.4455, + "step": 325 + }, + { + "epoch": 1.4, + "grad_norm": 1.189352035522461, + "learning_rate": 0.0003893104585945233, + "loss": 0.4246, + "step": 350 + }, + { + "epoch": 1.5, + "grad_norm": 0.947853684425354, + "learning_rate": 0.0003876608380072583, + "loss": 0.4326, + "step": 375 + }, + { + "epoch": 1.6, + "grad_norm": 0.7130059003829956, + "learning_rate": 0.00038601121741999343, + "loss": 0.4034, + "step": 400 + }, + { + "epoch": 1.7, + "grad_norm": 3.517078161239624, + "learning_rate": 0.0003843615968327285, + "loss": 0.4114, + "step": 425 + }, + { + "epoch": 1.8, + "grad_norm": 0.3982729911804199, + "learning_rate": 0.0003827119762454636, + "loss": 0.4345, + "step": 450 + }, + { + "epoch": 1.9, + "grad_norm": 0.6208440065383911, + "learning_rate": 0.00038106235565819863, + "loss": 0.4135, + "step": 475 + }, + { + "epoch": 2.0, + "grad_norm": 0.9085566401481628, + "learning_rate": 0.0003794127350709337, + "loss": 0.3983, + "step": 500 + }, + { + "epoch": 2.1, + "grad_norm": 1.2225189208984375, + "learning_rate": 0.0003777631144836688, + "loss": 0.3915, + "step": 525 + }, + { + "epoch": 2.2, + "grad_norm": 1.1379530429840088, + "learning_rate": 0.00037611349389640383, + "loss": 0.4335, + "step": 550 + }, + { + "epoch": 2.3, + "grad_norm": 0.8535892367362976, + "learning_rate": 0.00037446387330913894, + "loss": 0.4063, + "step": 575 + }, + { + "epoch": 2.4, + "grad_norm": 1.3587419986724854, + "learning_rate": 0.000372814252721874, + "loss": 0.4065, + "step": 600 + }, + { + "epoch": 2.5, + "grad_norm": 0.9311383962631226, + "learning_rate": 0.00037116463213460903, + "loss": 0.4007, + "step": 625 + }, + { + "epoch": 2.6, + "grad_norm": 0.9912465810775757, + "learning_rate": 0.00036951501154734414, + "loss": 0.3992, + "step": 650 + }, + { + "epoch": 2.7, + "grad_norm": 1.2762119770050049, + "learning_rate": 0.0003678653909600792, + "loss": 0.3827, + "step": 675 + }, + { + "epoch": 2.8, + "grad_norm": 2.3506247997283936, + "learning_rate": 0.0003662157703728143, + "loss": 0.3734, + "step": 700 + }, + { + "epoch": 2.9, + "grad_norm": 1.5375425815582275, + "learning_rate": 0.00036456614978554934, + "loss": 0.4005, + "step": 725 + }, + { + "epoch": 3.0, + "grad_norm": 1.3143396377563477, + "learning_rate": 0.00036291652919828444, + "loss": 0.3548, + "step": 750 + }, + { + "epoch": 3.1, + "grad_norm": 0.662200391292572, + "learning_rate": 0.00036126690861101944, + "loss": 0.4129, + "step": 775 + }, + { + "epoch": 3.2, + "grad_norm": 1.3569878339767456, + "learning_rate": 0.00035961728802375454, + "loss": 0.38, + "step": 800 + }, + { + "epoch": 3.3, + "grad_norm": 0.9888588190078735, + "learning_rate": 0.0003579676674364896, + "loss": 0.3778, + "step": 825 + }, + { + "epoch": 3.4, + "grad_norm": 0.679969310760498, + "learning_rate": 0.0003563180468492247, + "loss": 0.3934, + "step": 850 + }, + { + "epoch": 3.5, + "grad_norm": 1.608539342880249, + "learning_rate": 0.0003546684262619598, + "loss": 0.418, + "step": 875 + }, + { + "epoch": 3.6, + "grad_norm": 1.7037910223007202, + "learning_rate": 0.00035301880567469485, + "loss": 0.3934, + "step": 900 + }, + { + "epoch": 3.7, + "grad_norm": 1.4417132139205933, + "learning_rate": 0.0003513691850874299, + "loss": 0.3947, + "step": 925 + }, + { + "epoch": 3.8, + "grad_norm": 1.1482582092285156, + "learning_rate": 0.00034971956450016495, + "loss": 0.3621, + "step": 950 + }, + { + "epoch": 3.9, + "grad_norm": 2.046616792678833, + "learning_rate": 0.00034806994391290005, + "loss": 0.3774, + "step": 975 + }, + { + "epoch": 4.0, + "grad_norm": 1.0832113027572632, + "learning_rate": 0.0003464203233256351, + "loss": 0.3542, + "step": 1000 + }, + { + "epoch": 4.1, + "grad_norm": 0.8189190626144409, + "learning_rate": 0.0003447707027383702, + "loss": 0.381, + "step": 1025 + }, + { + "epoch": 4.2, + "grad_norm": 1.6974774599075317, + "learning_rate": 0.00034312108215110525, + "loss": 0.3775, + "step": 1050 + }, + { + "epoch": 4.3, + "grad_norm": 1.2468841075897217, + "learning_rate": 0.0003414714615638403, + "loss": 0.3527, + "step": 1075 + }, + { + "epoch": 4.4, + "grad_norm": 0.9022793173789978, + "learning_rate": 0.0003398218409765754, + "loss": 0.3637, + "step": 1100 + }, + { + "epoch": 4.5, + "grad_norm": 1.121774673461914, + "learning_rate": 0.00033817222038931045, + "loss": 0.3935, + "step": 1125 + }, + { + "epoch": 4.6, + "grad_norm": 0.810937762260437, + "learning_rate": 0.00033652259980204556, + "loss": 0.4103, + "step": 1150 + }, + { + "epoch": 4.7, + "grad_norm": 0.7376565337181091, + "learning_rate": 0.0003348729792147806, + "loss": 0.362, + "step": 1175 + }, + { + "epoch": 4.8, + "grad_norm": 1.7271156311035156, + "learning_rate": 0.0003332233586275157, + "loss": 0.3597, + "step": 1200 + }, + { + "epoch": 4.9, + "grad_norm": 1.2273073196411133, + "learning_rate": 0.00033157373804025076, + "loss": 0.3853, + "step": 1225 + }, + { + "epoch": 5.0, + "grad_norm": 0.9998250603675842, + "learning_rate": 0.0003299241174529858, + "loss": 0.3461, + "step": 1250 + }, + { + "epoch": 5.1, + "grad_norm": 0.9112564325332642, + "learning_rate": 0.0003282744968657209, + "loss": 0.3659, + "step": 1275 + }, + { + "epoch": 5.2, + "grad_norm": 3.246051073074341, + "learning_rate": 0.00032662487627845596, + "loss": 0.3488, + "step": 1300 + }, + { + "epoch": 5.3, + "grad_norm": 1.0283201932907104, + "learning_rate": 0.00032497525569119106, + "loss": 0.3528, + "step": 1325 + }, + { + "epoch": 5.4, + "grad_norm": 1.112660527229309, + "learning_rate": 0.0003233256351039261, + "loss": 0.3723, + "step": 1350 + }, + { + "epoch": 5.5, + "grad_norm": 0.9069312810897827, + "learning_rate": 0.00032167601451666116, + "loss": 0.3416, + "step": 1375 + }, + { + "epoch": 5.6, + "grad_norm": 0.710412323474884, + "learning_rate": 0.00032002639392939627, + "loss": 0.3607, + "step": 1400 + }, + { + "epoch": 5.7, + "grad_norm": 1.185612678527832, + "learning_rate": 0.0003183767733421313, + "loss": 0.3777, + "step": 1425 + }, + { + "epoch": 5.8, + "grad_norm": 1.340174913406372, + "learning_rate": 0.0003167271527548664, + "loss": 0.3563, + "step": 1450 + }, + { + "epoch": 5.9, + "grad_norm": 0.9387012124061584, + "learning_rate": 0.00031507753216760147, + "loss": 0.3647, + "step": 1475 + }, + { + "epoch": 6.0, + "grad_norm": 1.1005347967147827, + "learning_rate": 0.00031342791158033657, + "loss": 0.3484, + "step": 1500 + }, + { + "epoch": 6.1, + "grad_norm": 2.64058780670166, + "learning_rate": 0.00031177829099307157, + "loss": 0.3542, + "step": 1525 + }, + { + "epoch": 6.2, + "grad_norm": 1.3007783889770508, + "learning_rate": 0.00031012867040580667, + "loss": 0.3379, + "step": 1550 + }, + { + "epoch": 6.3, + "grad_norm": 3.7190418243408203, + "learning_rate": 0.0003084790498185417, + "loss": 0.3514, + "step": 1575 + }, + { + "epoch": 6.4, + "grad_norm": 1.7263462543487549, + "learning_rate": 0.0003068294292312768, + "loss": 0.3434, + "step": 1600 + }, + { + "epoch": 6.5, + "grad_norm": 3.3360602855682373, + "learning_rate": 0.0003051798086440119, + "loss": 0.3507, + "step": 1625 + }, + { + "epoch": 6.6, + "grad_norm": 1.9620813131332397, + "learning_rate": 0.000303530188056747, + "loss": 0.3455, + "step": 1650 + }, + { + "epoch": 6.7, + "grad_norm": 1.3080660104751587, + "learning_rate": 0.000301880567469482, + "loss": 0.3678, + "step": 1675 + }, + { + "epoch": 6.8, + "grad_norm": 0.6183618307113647, + "learning_rate": 0.0003002309468822171, + "loss": 0.3704, + "step": 1700 + }, + { + "epoch": 6.9, + "grad_norm": 1.0411962270736694, + "learning_rate": 0.0002985813262949522, + "loss": 0.3505, + "step": 1725 + }, + { + "epoch": 7.0, + "grad_norm": 2.040569543838501, + "learning_rate": 0.00029693170570768723, + "loss": 0.3666, + "step": 1750 + }, + { + "epoch": 7.1, + "grad_norm": 0.713647723197937, + "learning_rate": 0.00029528208512042233, + "loss": 0.3529, + "step": 1775 + }, + { + "epoch": 7.2, + "grad_norm": 3.0373058319091797, + "learning_rate": 0.0002936324645331574, + "loss": 0.361, + "step": 1800 + }, + { + "epoch": 7.3, + "grad_norm": 1.819886326789856, + "learning_rate": 0.00029198284394589243, + "loss": 0.3482, + "step": 1825 + }, + { + "epoch": 7.4, + "grad_norm": 0.7930757403373718, + "learning_rate": 0.00029033322335862753, + "loss": 0.3695, + "step": 1850 + }, + { + "epoch": 7.5, + "grad_norm": 0.7889612913131714, + "learning_rate": 0.0002886836027713626, + "loss": 0.3527, + "step": 1875 + }, + { + "epoch": 7.6, + "grad_norm": 2.4007863998413086, + "learning_rate": 0.0002870339821840977, + "loss": 0.3275, + "step": 1900 + }, + { + "epoch": 7.7, + "grad_norm": 2.3483331203460693, + "learning_rate": 0.00028538436159683273, + "loss": 0.3571, + "step": 1925 + }, + { + "epoch": 7.8, + "grad_norm": 0.8911663293838501, + "learning_rate": 0.00028373474100956784, + "loss": 0.3079, + "step": 1950 + }, + { + "epoch": 7.9, + "grad_norm": 1.0521246194839478, + "learning_rate": 0.0002820851204223029, + "loss": 0.3578, + "step": 1975 + }, + { + "epoch": 8.0, + "grad_norm": 1.3923832178115845, + "learning_rate": 0.00028043549983503794, + "loss": 0.3588, + "step": 2000 + }, + { + "epoch": 8.1, + "grad_norm": 1.4007149934768677, + "learning_rate": 0.00027878587924777304, + "loss": 0.3458, + "step": 2025 + }, + { + "epoch": 8.2, + "grad_norm": 1.103198766708374, + "learning_rate": 0.0002771362586605081, + "loss": 0.3736, + "step": 2050 + }, + { + "epoch": 8.3, + "grad_norm": 2.936150550842285, + "learning_rate": 0.0002754866380732432, + "loss": 0.3223, + "step": 2075 + }, + { + "epoch": 8.4, + "grad_norm": 1.907762885093689, + "learning_rate": 0.00027383701748597824, + "loss": 0.3348, + "step": 2100 + }, + { + "epoch": 8.5, + "grad_norm": 1.4422886371612549, + "learning_rate": 0.0002721873968987133, + "loss": 0.3222, + "step": 2125 + }, + { + "epoch": 8.6, + "grad_norm": 1.8203483819961548, + "learning_rate": 0.00027053777631144834, + "loss": 0.3968, + "step": 2150 + }, + { + "epoch": 8.7, + "grad_norm": 0.8046119213104248, + "learning_rate": 0.00026888815572418344, + "loss": 0.3422, + "step": 2175 + }, + { + "epoch": 8.8, + "grad_norm": 1.2612605094909668, + "learning_rate": 0.00026723853513691855, + "loss": 0.3387, + "step": 2200 + }, + { + "epoch": 8.9, + "grad_norm": 1.6534866094589233, + "learning_rate": 0.0002655889145496536, + "loss": 0.3548, + "step": 2225 + }, + { + "epoch": 9.0, + "grad_norm": 1.889664649963379, + "learning_rate": 0.0002639392939623887, + "loss": 0.3367, + "step": 2250 + }, + { + "epoch": 9.1, + "grad_norm": 2.6691551208496094, + "learning_rate": 0.0002622896733751237, + "loss": 0.3628, + "step": 2275 + }, + { + "epoch": 9.2, + "grad_norm": 2.0130255222320557, + "learning_rate": 0.0002606400527878588, + "loss": 0.3366, + "step": 2300 + }, + { + "epoch": 9.3, + "grad_norm": 1.404814600944519, + "learning_rate": 0.00025899043220059385, + "loss": 0.3327, + "step": 2325 + }, + { + "epoch": 9.4, + "grad_norm": 0.9150362610816956, + "learning_rate": 0.00025734081161332895, + "loss": 0.308, + "step": 2350 + }, + { + "epoch": 9.5, + "grad_norm": 1.099320411682129, + "learning_rate": 0.000255691191026064, + "loss": 0.3499, + "step": 2375 + }, + { + "epoch": 9.6, + "grad_norm": 1.502451777458191, + "learning_rate": 0.0002540415704387991, + "loss": 0.3728, + "step": 2400 + }, + { + "epoch": 9.7, + "grad_norm": 1.1115409135818481, + "learning_rate": 0.00025239194985153415, + "loss": 0.348, + "step": 2425 + }, + { + "epoch": 9.8, + "grad_norm": 0.8831422328948975, + "learning_rate": 0.0002507423292642692, + "loss": 0.3392, + "step": 2450 + }, + { + "epoch": 9.9, + "grad_norm": 2.104572057723999, + "learning_rate": 0.0002490927086770043, + "loss": 0.3511, + "step": 2475 + }, + { + "epoch": 10.0, + "grad_norm": 2.4081919193267822, + "learning_rate": 0.00024744308808973936, + "loss": 0.3284, + "step": 2500 + }, + { + "epoch": 10.1, + "grad_norm": 1.9861253499984741, + "learning_rate": 0.00024579346750247446, + "loss": 0.3621, + "step": 2525 + }, + { + "epoch": 10.2, + "grad_norm": 1.9042344093322754, + "learning_rate": 0.0002441438469152095, + "loss": 0.3341, + "step": 2550 + }, + { + "epoch": 10.3, + "grad_norm": 0.5676223635673523, + "learning_rate": 0.00024249422632794456, + "loss": 0.3203, + "step": 2575 + }, + { + "epoch": 10.4, + "grad_norm": 1.5105270147323608, + "learning_rate": 0.00024084460574067963, + "loss": 0.3711, + "step": 2600 + }, + { + "epoch": 10.5, + "grad_norm": 0.9969163537025452, + "learning_rate": 0.0002391949851534147, + "loss": 0.3296, + "step": 2625 + }, + { + "epoch": 10.6, + "grad_norm": 1.297096848487854, + "learning_rate": 0.0002375453645661498, + "loss": 0.349, + "step": 2650 + }, + { + "epoch": 10.7, + "grad_norm": 1.0513092279434204, + "learning_rate": 0.00023589574397888486, + "loss": 0.3308, + "step": 2675 + }, + { + "epoch": 10.8, + "grad_norm": 1.5543969869613647, + "learning_rate": 0.00023424612339161997, + "loss": 0.3289, + "step": 2700 + }, + { + "epoch": 10.9, + "grad_norm": 0.9587257504463196, + "learning_rate": 0.000232596502804355, + "loss": 0.3128, + "step": 2725 + }, + { + "epoch": 11.0, + "grad_norm": 2.1693837642669678, + "learning_rate": 0.00023094688221709007, + "loss": 0.349, + "step": 2750 + }, + { + "epoch": 11.1, + "grad_norm": 1.5624831914901733, + "learning_rate": 0.00022929726162982514, + "loss": 0.3391, + "step": 2775 + }, + { + "epoch": 11.2, + "grad_norm": 1.4600939750671387, + "learning_rate": 0.00022764764104256022, + "loss": 0.3371, + "step": 2800 + }, + { + "epoch": 11.3, + "grad_norm": 1.6139200925827026, + "learning_rate": 0.0002259980204552953, + "loss": 0.3447, + "step": 2825 + }, + { + "epoch": 11.4, + "grad_norm": 2.2703235149383545, + "learning_rate": 0.00022434839986803037, + "loss": 0.3157, + "step": 2850 + }, + { + "epoch": 11.5, + "grad_norm": 2.0282037258148193, + "learning_rate": 0.00022269877928076542, + "loss": 0.3339, + "step": 2875 + }, + { + "epoch": 11.6, + "grad_norm": 1.389129638671875, + "learning_rate": 0.0002210491586935005, + "loss": 0.3195, + "step": 2900 + }, + { + "epoch": 11.7, + "grad_norm": 1.0041213035583496, + "learning_rate": 0.00021939953810623557, + "loss": 0.3382, + "step": 2925 + }, + { + "epoch": 11.8, + "grad_norm": 1.7116957902908325, + "learning_rate": 0.00021774991751897065, + "loss": 0.3261, + "step": 2950 + }, + { + "epoch": 11.9, + "grad_norm": 1.080194354057312, + "learning_rate": 0.00021610029693170573, + "loss": 0.3433, + "step": 2975 + }, + { + "epoch": 12.0, + "grad_norm": 0.6561412811279297, + "learning_rate": 0.0002144506763444408, + "loss": 0.3465, + "step": 3000 + }, + { + "epoch": 12.1, + "grad_norm": 1.5294175148010254, + "learning_rate": 0.00021280105575717582, + "loss": 0.3195, + "step": 3025 + }, + { + "epoch": 12.2, + "grad_norm": 0.6313191652297974, + "learning_rate": 0.00021115143516991093, + "loss": 0.3214, + "step": 3050 + }, + { + "epoch": 12.3, + "grad_norm": 1.4611085653305054, + "learning_rate": 0.000209501814582646, + "loss": 0.3091, + "step": 3075 + }, + { + "epoch": 12.4, + "grad_norm": 1.6590704917907715, + "learning_rate": 0.00020785219399538108, + "loss": 0.3633, + "step": 3100 + }, + { + "epoch": 12.5, + "grad_norm": 0.982976496219635, + "learning_rate": 0.00020620257340811616, + "loss": 0.3688, + "step": 3125 + }, + { + "epoch": 12.6, + "grad_norm": 0.9717169404029846, + "learning_rate": 0.00020455295282085123, + "loss": 0.3532, + "step": 3150 + }, + { + "epoch": 12.7, + "grad_norm": 1.5903713703155518, + "learning_rate": 0.00020290333223358626, + "loss": 0.2906, + "step": 3175 + }, + { + "epoch": 12.8, + "grad_norm": 1.7043166160583496, + "learning_rate": 0.00020125371164632133, + "loss": 0.322, + "step": 3200 + }, + { + "epoch": 12.9, + "grad_norm": 0.7691648006439209, + "learning_rate": 0.0001996040910590564, + "loss": 0.3477, + "step": 3225 + }, + { + "epoch": 13.0, + "grad_norm": 1.1726899147033691, + "learning_rate": 0.00019795447047179148, + "loss": 0.3551, + "step": 3250 + }, + { + "epoch": 13.1, + "grad_norm": 1.0323086977005005, + "learning_rate": 0.0001963048498845266, + "loss": 0.3248, + "step": 3275 + }, + { + "epoch": 13.2, + "grad_norm": 2.920222520828247, + "learning_rate": 0.00019465522929726164, + "loss": 0.3298, + "step": 3300 + }, + { + "epoch": 13.3, + "grad_norm": 1.101644515991211, + "learning_rate": 0.00019300560870999671, + "loss": 0.3011, + "step": 3325 + }, + { + "epoch": 13.4, + "grad_norm": 1.111193299293518, + "learning_rate": 0.0001913559881227318, + "loss": 0.325, + "step": 3350 + }, + { + "epoch": 13.5, + "grad_norm": 1.5074782371520996, + "learning_rate": 0.00018970636753546684, + "loss": 0.3435, + "step": 3375 + }, + { + "epoch": 13.6, + "grad_norm": 1.0705279111862183, + "learning_rate": 0.00018805674694820192, + "loss": 0.3378, + "step": 3400 + }, + { + "epoch": 13.7, + "grad_norm": 1.5590386390686035, + "learning_rate": 0.000186407126360937, + "loss": 0.3349, + "step": 3425 + }, + { + "epoch": 13.8, + "grad_norm": 2.2993671894073486, + "learning_rate": 0.00018475750577367207, + "loss": 0.3479, + "step": 3450 + }, + { + "epoch": 13.9, + "grad_norm": 1.8973578214645386, + "learning_rate": 0.00018310788518640715, + "loss": 0.3358, + "step": 3475 + }, + { + "epoch": 14.0, + "grad_norm": 1.5887267589569092, + "learning_rate": 0.00018145826459914222, + "loss": 0.3095, + "step": 3500 + }, + { + "epoch": 14.1, + "grad_norm": 2.1970202922821045, + "learning_rate": 0.00017980864401187727, + "loss": 0.3435, + "step": 3525 + }, + { + "epoch": 14.2, + "grad_norm": 0.9183377027511597, + "learning_rate": 0.00017815902342461235, + "loss": 0.3454, + "step": 3550 + }, + { + "epoch": 14.3, + "grad_norm": 1.253682017326355, + "learning_rate": 0.00017650940283734742, + "loss": 0.3128, + "step": 3575 + }, + { + "epoch": 14.4, + "grad_norm": 2.4595584869384766, + "learning_rate": 0.00017485978225008247, + "loss": 0.3181, + "step": 3600 + }, + { + "epoch": 14.5, + "grad_norm": 2.0421340465545654, + "learning_rate": 0.00017321016166281755, + "loss": 0.3555, + "step": 3625 + }, + { + "epoch": 14.6, + "grad_norm": 2.0127763748168945, + "learning_rate": 0.00017156054107555263, + "loss": 0.2952, + "step": 3650 + }, + { + "epoch": 14.7, + "grad_norm": 0.9481540322303772, + "learning_rate": 0.0001699109204882877, + "loss": 0.3154, + "step": 3675 + }, + { + "epoch": 14.8, + "grad_norm": 1.5080980062484741, + "learning_rate": 0.00016826129990102278, + "loss": 0.336, + "step": 3700 + }, + { + "epoch": 14.9, + "grad_norm": 1.504712700843811, + "learning_rate": 0.00016661167931375785, + "loss": 0.3511, + "step": 3725 + }, + { + "epoch": 15.0, + "grad_norm": 1.4532283544540405, + "learning_rate": 0.0001649620587264929, + "loss": 0.3216, + "step": 3750 + }, + { + "epoch": 15.1, + "grad_norm": 1.215970754623413, + "learning_rate": 0.00016331243813922798, + "loss": 0.3597, + "step": 3775 + }, + { + "epoch": 15.2, + "grad_norm": 2.054934024810791, + "learning_rate": 0.00016166281755196306, + "loss": 0.3095, + "step": 3800 + }, + { + "epoch": 15.3, + "grad_norm": 1.6162620782852173, + "learning_rate": 0.00016001319696469813, + "loss": 0.3172, + "step": 3825 + }, + { + "epoch": 15.4, + "grad_norm": 2.974353790283203, + "learning_rate": 0.0001583635763774332, + "loss": 0.3186, + "step": 3850 + }, + { + "epoch": 15.5, + "grad_norm": 1.1828879117965698, + "learning_rate": 0.00015671395579016829, + "loss": 0.3141, + "step": 3875 + }, + { + "epoch": 15.6, + "grad_norm": 2.170670509338379, + "learning_rate": 0.00015506433520290334, + "loss": 0.3054, + "step": 3900 + }, + { + "epoch": 15.7, + "grad_norm": 2.5053343772888184, + "learning_rate": 0.0001534147146156384, + "loss": 0.3348, + "step": 3925 + }, + { + "epoch": 15.8, + "grad_norm": 2.761240005493164, + "learning_rate": 0.0001517650940283735, + "loss": 0.3513, + "step": 3950 + }, + { + "epoch": 15.9, + "grad_norm": 2.6656720638275146, + "learning_rate": 0.00015011547344110854, + "loss": 0.3477, + "step": 3975 + }, + { + "epoch": 16.0, + "grad_norm": 1.2667946815490723, + "learning_rate": 0.00014846585285384361, + "loss": 0.3107, + "step": 4000 + }, + { + "epoch": 16.1, + "grad_norm": 1.4039124250411987, + "learning_rate": 0.0001468162322665787, + "loss": 0.3242, + "step": 4025 + }, + { + "epoch": 16.2, + "grad_norm": 1.3851566314697266, + "learning_rate": 0.00014516661167931377, + "loss": 0.3144, + "step": 4050 + }, + { + "epoch": 16.3, + "grad_norm": 1.2302920818328857, + "learning_rate": 0.00014351699109204884, + "loss": 0.3376, + "step": 4075 + }, + { + "epoch": 16.4, + "grad_norm": 1.8713226318359375, + "learning_rate": 0.00014186737050478392, + "loss": 0.3266, + "step": 4100 + }, + { + "epoch": 16.5, + "grad_norm": 1.3579822778701782, + "learning_rate": 0.00014021774991751897, + "loss": 0.33, + "step": 4125 + }, + { + "epoch": 16.6, + "grad_norm": 1.4338854551315308, + "learning_rate": 0.00013856812933025404, + "loss": 0.3218, + "step": 4150 + }, + { + "epoch": 16.7, + "grad_norm": 1.9794092178344727, + "learning_rate": 0.00013691850874298912, + "loss": 0.3375, + "step": 4175 + }, + { + "epoch": 16.8, + "grad_norm": 1.4431006908416748, + "learning_rate": 0.00013526888815572417, + "loss": 0.3177, + "step": 4200 + }, + { + "epoch": 16.9, + "grad_norm": 1.1928157806396484, + "learning_rate": 0.00013361926756845927, + "loss": 0.3227, + "step": 4225 + }, + { + "epoch": 17.0, + "grad_norm": 2.0860953330993652, + "learning_rate": 0.00013196964698119435, + "loss": 0.3345, + "step": 4250 + }, + { + "epoch": 17.1, + "grad_norm": 1.0021557807922363, + "learning_rate": 0.0001303200263939294, + "loss": 0.3299, + "step": 4275 + }, + { + "epoch": 17.2, + "grad_norm": 1.6104000806808472, + "learning_rate": 0.00012867040580666448, + "loss": 0.3066, + "step": 4300 + }, + { + "epoch": 17.3, + "grad_norm": 2.3414576053619385, + "learning_rate": 0.00012702078521939955, + "loss": 0.3474, + "step": 4325 + }, + { + "epoch": 17.4, + "grad_norm": 1.8101857900619507, + "learning_rate": 0.0001253711646321346, + "loss": 0.3022, + "step": 4350 + }, + { + "epoch": 17.5, + "grad_norm": 1.1479285955429077, + "learning_rate": 0.00012372154404486968, + "loss": 0.3133, + "step": 4375 + }, + { + "epoch": 17.6, + "grad_norm": 2.7935454845428467, + "learning_rate": 0.00012207192345760475, + "loss": 0.3027, + "step": 4400 + }, + { + "epoch": 17.7, + "grad_norm": 2.0491445064544678, + "learning_rate": 0.00012042230287033982, + "loss": 0.3451, + "step": 4425 + }, + { + "epoch": 17.8, + "grad_norm": 0.9633632302284241, + "learning_rate": 0.0001187726822830749, + "loss": 0.3171, + "step": 4450 + }, + { + "epoch": 17.9, + "grad_norm": 1.5458674430847168, + "learning_rate": 0.00011712306169580998, + "loss": 0.3011, + "step": 4475 + }, + { + "epoch": 18.0, + "grad_norm": 1.4501872062683105, + "learning_rate": 0.00011547344110854503, + "loss": 0.3185, + "step": 4500 + }, + { + "epoch": 18.1, + "grad_norm": 2.6352083683013916, + "learning_rate": 0.00011382382052128011, + "loss": 0.3191, + "step": 4525 + }, + { + "epoch": 18.2, + "grad_norm": 2.366875410079956, + "learning_rate": 0.00011217419993401519, + "loss": 0.311, + "step": 4550 + }, + { + "epoch": 18.3, + "grad_norm": 1.9184437990188599, + "learning_rate": 0.00011052457934675025, + "loss": 0.2812, + "step": 4575 + }, + { + "epoch": 18.4, + "grad_norm": 1.466262936592102, + "learning_rate": 0.00010887495875948532, + "loss": 0.3075, + "step": 4600 + }, + { + "epoch": 18.5, + "grad_norm": 1.94712233543396, + "learning_rate": 0.0001072253381722204, + "loss": 0.3163, + "step": 4625 + }, + { + "epoch": 18.6, + "grad_norm": 2.5286598205566406, + "learning_rate": 0.00010557571758495546, + "loss": 0.3346, + "step": 4650 + }, + { + "epoch": 18.7, + "grad_norm": 1.604040265083313, + "learning_rate": 0.00010392609699769054, + "loss": 0.3138, + "step": 4675 + }, + { + "epoch": 18.8, + "grad_norm": 1.0606187582015991, + "learning_rate": 0.00010227647641042562, + "loss": 0.3524, + "step": 4700 + }, + { + "epoch": 18.9, + "grad_norm": 1.275098443031311, + "learning_rate": 0.00010062685582316067, + "loss": 0.3057, + "step": 4725 + }, + { + "epoch": 19.0, + "grad_norm": 1.971979022026062, + "learning_rate": 9.897723523589574e-05, + "loss": 0.3089, + "step": 4750 + }, + { + "epoch": 19.1, + "grad_norm": 4.085920810699463, + "learning_rate": 9.732761464863082e-05, + "loss": 0.3251, + "step": 4775 + }, + { + "epoch": 19.2, + "grad_norm": 1.5097062587738037, + "learning_rate": 9.56779940613659e-05, + "loss": 0.2869, + "step": 4800 + }, + { + "epoch": 19.3, + "grad_norm": 2.4897541999816895, + "learning_rate": 9.402837347410096e-05, + "loss": 0.28, + "step": 4825 + }, + { + "epoch": 19.4, + "grad_norm": 2.232405662536621, + "learning_rate": 9.237875288683603e-05, + "loss": 0.3585, + "step": 4850 + }, + { + "epoch": 19.5, + "grad_norm": 2.3231348991394043, + "learning_rate": 9.072913229957111e-05, + "loss": 0.3578, + "step": 4875 + }, + { + "epoch": 19.6, + "grad_norm": 1.1693369150161743, + "learning_rate": 8.907951171230617e-05, + "loss": 0.3201, + "step": 4900 + }, + { + "epoch": 19.7, + "grad_norm": 1.5570422410964966, + "learning_rate": 8.742989112504124e-05, + "loss": 0.3142, + "step": 4925 + }, + { + "epoch": 19.8, + "grad_norm": 1.122075080871582, + "learning_rate": 8.578027053777631e-05, + "loss": 0.3274, + "step": 4950 + }, + { + "epoch": 19.9, + "grad_norm": 1.4783751964569092, + "learning_rate": 8.413064995051139e-05, + "loss": 0.3246, + "step": 4975 + }, + { + "epoch": 20.0, + "grad_norm": 1.6368613243103027, + "learning_rate": 8.248102936324645e-05, + "loss": 0.2984, + "step": 5000 + }, + { + "epoch": 20.1, + "grad_norm": 1.4082392454147339, + "learning_rate": 8.083140877598153e-05, + "loss": 0.3359, + "step": 5025 + }, + { + "epoch": 20.2, + "grad_norm": 1.4183349609375, + "learning_rate": 7.91817881887166e-05, + "loss": 0.327, + "step": 5050 + }, + { + "epoch": 20.3, + "grad_norm": 1.4731121063232422, + "learning_rate": 7.753216760145167e-05, + "loss": 0.3184, + "step": 5075 + }, + { + "epoch": 20.4, + "grad_norm": 2.123331069946289, + "learning_rate": 7.588254701418674e-05, + "loss": 0.3366, + "step": 5100 + }, + { + "epoch": 20.5, + "grad_norm": 1.2433639764785767, + "learning_rate": 7.423292642692181e-05, + "loss": 0.2926, + "step": 5125 + }, + { + "epoch": 20.6, + "grad_norm": 1.4048691987991333, + "learning_rate": 7.258330583965688e-05, + "loss": 0.3067, + "step": 5150 + }, + { + "epoch": 20.7, + "grad_norm": 1.4444036483764648, + "learning_rate": 7.093368525239196e-05, + "loss": 0.2933, + "step": 5175 + }, + { + "epoch": 20.8, + "grad_norm": 1.5258699655532837, + "learning_rate": 6.928406466512702e-05, + "loss": 0.3075, + "step": 5200 + }, + { + "epoch": 20.9, + "grad_norm": 1.6690775156021118, + "learning_rate": 6.763444407786209e-05, + "loss": 0.3002, + "step": 5225 + }, + { + "epoch": 21.0, + "grad_norm": 1.2771008014678955, + "learning_rate": 6.598482349059718e-05, + "loss": 0.3457, + "step": 5250 + }, + { + "epoch": 21.1, + "grad_norm": 0.9566493630409241, + "learning_rate": 6.433520290333224e-05, + "loss": 0.3132, + "step": 5275 + }, + { + "epoch": 21.2, + "grad_norm": 1.8674430847167969, + "learning_rate": 6.26855823160673e-05, + "loss": 0.3158, + "step": 5300 + }, + { + "epoch": 21.3, + "grad_norm": 1.9155611991882324, + "learning_rate": 6.103596172880238e-05, + "loss": 0.3322, + "step": 5325 + }, + { + "epoch": 21.4, + "grad_norm": 1.1892235279083252, + "learning_rate": 5.938634114153745e-05, + "loss": 0.3221, + "step": 5350 + }, + { + "epoch": 21.5, + "grad_norm": 1.3518539667129517, + "learning_rate": 5.7736720554272516e-05, + "loss": 0.2993, + "step": 5375 + }, + { + "epoch": 21.6, + "grad_norm": 1.9076350927352905, + "learning_rate": 5.608709996700759e-05, + "loss": 0.2872, + "step": 5400 + }, + { + "epoch": 21.7, + "grad_norm": 1.7502110004425049, + "learning_rate": 5.443747937974266e-05, + "loss": 0.3233, + "step": 5425 + }, + { + "epoch": 21.8, + "grad_norm": 2.491464853286743, + "learning_rate": 5.278785879247773e-05, + "loss": 0.3307, + "step": 5450 + }, + { + "epoch": 21.9, + "grad_norm": 2.3630316257476807, + "learning_rate": 5.113823820521281e-05, + "loss": 0.3266, + "step": 5475 + }, + { + "epoch": 22.0, + "grad_norm": 1.6543325185775757, + "learning_rate": 4.948861761794787e-05, + "loss": 0.3026, + "step": 5500 + }, + { + "epoch": 22.1, + "grad_norm": 1.8995773792266846, + "learning_rate": 4.783899703068295e-05, + "loss": 0.3158, + "step": 5525 + }, + { + "epoch": 22.2, + "grad_norm": 2.4362375736236572, + "learning_rate": 4.618937644341802e-05, + "loss": 0.3068, + "step": 5550 + }, + { + "epoch": 22.3, + "grad_norm": 2.885927200317383, + "learning_rate": 4.453975585615309e-05, + "loss": 0.3263, + "step": 5575 + }, + { + "epoch": 22.4, + "grad_norm": 3.3690805435180664, + "learning_rate": 4.2890135268888156e-05, + "loss": 0.3186, + "step": 5600 + }, + { + "epoch": 22.5, + "grad_norm": 1.418340802192688, + "learning_rate": 4.1240514681623226e-05, + "loss": 0.2831, + "step": 5625 + }, + { + "epoch": 22.6, + "grad_norm": 2.099191904067993, + "learning_rate": 3.95908940943583e-05, + "loss": 0.3079, + "step": 5650 + }, + { + "epoch": 22.7, + "grad_norm": 3.213982582092285, + "learning_rate": 3.794127350709337e-05, + "loss": 0.2997, + "step": 5675 + }, + { + "epoch": 22.8, + "grad_norm": 1.909794807434082, + "learning_rate": 3.629165291982844e-05, + "loss": 0.3414, + "step": 5700 + }, + { + "epoch": 22.9, + "grad_norm": 1.5416369438171387, + "learning_rate": 3.464203233256351e-05, + "loss": 0.3103, + "step": 5725 + }, + { + "epoch": 23.0, + "grad_norm": 2.4521079063415527, + "learning_rate": 3.299241174529859e-05, + "loss": 0.315, + "step": 5750 + }, + { + "epoch": 23.1, + "grad_norm": 1.9134442806243896, + "learning_rate": 3.134279115803365e-05, + "loss": 0.3141, + "step": 5775 + }, + { + "epoch": 23.2, + "grad_norm": 1.115973711013794, + "learning_rate": 2.9693170570768723e-05, + "loss": 0.319, + "step": 5800 + }, + { + "epoch": 23.3, + "grad_norm": 3.0362348556518555, + "learning_rate": 2.8043549983503796e-05, + "loss": 0.299, + "step": 5825 + }, + { + "epoch": 23.4, + "grad_norm": 1.9920865297317505, + "learning_rate": 2.6393929396238866e-05, + "loss": 0.2653, + "step": 5850 + }, + { + "epoch": 23.5, + "grad_norm": 3.130983591079712, + "learning_rate": 2.4744308808973936e-05, + "loss": 0.3371, + "step": 5875 + }, + { + "epoch": 23.6, + "grad_norm": 2.5702381134033203, + "learning_rate": 2.309468822170901e-05, + "loss": 0.3061, + "step": 5900 + }, + { + "epoch": 23.7, + "grad_norm": 1.6236166954040527, + "learning_rate": 2.1445067634444078e-05, + "loss": 0.3187, + "step": 5925 + }, + { + "epoch": 23.8, + "grad_norm": 1.791316032409668, + "learning_rate": 1.979544704717915e-05, + "loss": 0.3168, + "step": 5950 + }, + { + "epoch": 23.9, + "grad_norm": 2.133009433746338, + "learning_rate": 1.814582645991422e-05, + "loss": 0.3183, + "step": 5975 + }, + { + "epoch": 24.0, + "grad_norm": 1.2213393449783325, + "learning_rate": 1.6496205872649294e-05, + "loss": 0.3336, + "step": 6000 + } + ], + "logging_steps": 25, + "max_steps": 6250, + "num_input_tokens_seen": 0, + "num_train_epochs": 25, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 318696652800000.0, + "train_batch_size": 20, + "trial_name": null, + "trial_params": null +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/training_args.bin b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..d61f3500eb70763e1f33496c1787bdf3f07a651b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b6e0f7aa622786274a1ea4195b72f4da224050767f7b89529b85721865c5075a +size 5304 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/vocab.txt b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/vocab.txt new file mode 100644 index 0000000000000000000000000000000000000000..cc9ad53732e21ea9180f18e1fb67a38cc5702d4c --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/vocab.txt @@ -0,0 +1,4101 @@ +[PAD] +[UNK] +[CLS] +[SEP] +[MASK] +AAAAAA +AAAAAT +AAAAAC +AAAAAG +AAAATA +AAAATT +AAAATC +AAAATG +AAAACA +AAAACT +AAAACC +AAAACG +AAAAGA +AAAAGT +AAAAGC +AAAAGG +AAATAA +AAATAT +AAATAC +AAATAG +AAATTA +AAATTT +AAATTC +AAATTG +AAATCA +AAATCT +AAATCC +AAATCG +AAATGA +AAATGT +AAATGC +AAATGG +AAACAA +AAACAT +AAACAC +AAACAG +AAACTA +AAACTT +AAACTC +AAACTG +AAACCA +AAACCT +AAACCC +AAACCG +AAACGA +AAACGT +AAACGC +AAACGG +AAAGAA +AAAGAT +AAAGAC +AAAGAG +AAAGTA +AAAGTT +AAAGTC +AAAGTG +AAAGCA +AAAGCT +AAAGCC +AAAGCG +AAAGGA +AAAGGT +AAAGGC +AAAGGG +AATAAA +AATAAT +AATAAC +AATAAG +AATATA +AATATT +AATATC +AATATG +AATACA +AATACT +AATACC +AATACG +AATAGA +AATAGT +AATAGC +AATAGG +AATTAA +AATTAT +AATTAC +AATTAG +AATTTA +AATTTT +AATTTC +AATTTG +AATTCA +AATTCT +AATTCC +AATTCG +AATTGA +AATTGT +AATTGC +AATTGG +AATCAA +AATCAT +AATCAC +AATCAG +AATCTA +AATCTT +AATCTC +AATCTG +AATCCA +AATCCT +AATCCC +AATCCG +AATCGA +AATCGT +AATCGC +AATCGG +AATGAA +AATGAT +AATGAC +AATGAG +AATGTA +AATGTT +AATGTC +AATGTG +AATGCA +AATGCT +AATGCC +AATGCG +AATGGA +AATGGT +AATGGC +AATGGG +AACAAA +AACAAT +AACAAC +AACAAG +AACATA +AACATT +AACATC +AACATG +AACACA +AACACT +AACACC +AACACG +AACAGA +AACAGT +AACAGC +AACAGG +AACTAA +AACTAT +AACTAC +AACTAG +AACTTA +AACTTT +AACTTC +AACTTG +AACTCA +AACTCT +AACTCC +AACTCG +AACTGA +AACTGT +AACTGC +AACTGG +AACCAA +AACCAT +AACCAC +AACCAG +AACCTA +AACCTT +AACCTC +AACCTG +AACCCA +AACCCT +AACCCC +AACCCG +AACCGA +AACCGT +AACCGC +AACCGG +AACGAA +AACGAT +AACGAC +AACGAG +AACGTA +AACGTT +AACGTC +AACGTG +AACGCA +AACGCT +AACGCC +AACGCG +AACGGA +AACGGT +AACGGC +AACGGG +AAGAAA +AAGAAT +AAGAAC +AAGAAG +AAGATA +AAGATT +AAGATC +AAGATG +AAGACA +AAGACT +AAGACC +AAGACG +AAGAGA +AAGAGT +AAGAGC +AAGAGG +AAGTAA +AAGTAT +AAGTAC +AAGTAG +AAGTTA +AAGTTT +AAGTTC +AAGTTG +AAGTCA +AAGTCT +AAGTCC +AAGTCG +AAGTGA +AAGTGT +AAGTGC +AAGTGG +AAGCAA +AAGCAT +AAGCAC +AAGCAG +AAGCTA +AAGCTT +AAGCTC +AAGCTG +AAGCCA +AAGCCT +AAGCCC +AAGCCG +AAGCGA +AAGCGT +AAGCGC +AAGCGG +AAGGAA +AAGGAT +AAGGAC +AAGGAG +AAGGTA +AAGGTT +AAGGTC +AAGGTG +AAGGCA +AAGGCT +AAGGCC +AAGGCG +AAGGGA +AAGGGT +AAGGGC +AAGGGG +ATAAAA +ATAAAT +ATAAAC +ATAAAG +ATAATA +ATAATT +ATAATC +ATAATG +ATAACA +ATAACT +ATAACC +ATAACG +ATAAGA +ATAAGT +ATAAGC +ATAAGG +ATATAA +ATATAT +ATATAC +ATATAG +ATATTA +ATATTT +ATATTC +ATATTG +ATATCA +ATATCT +ATATCC +ATATCG +ATATGA +ATATGT +ATATGC +ATATGG +ATACAA +ATACAT +ATACAC +ATACAG +ATACTA +ATACTT +ATACTC +ATACTG +ATACCA +ATACCT +ATACCC +ATACCG +ATACGA +ATACGT +ATACGC +ATACGG +ATAGAA +ATAGAT +ATAGAC +ATAGAG +ATAGTA +ATAGTT +ATAGTC +ATAGTG +ATAGCA +ATAGCT +ATAGCC +ATAGCG +ATAGGA +ATAGGT +ATAGGC +ATAGGG +ATTAAA +ATTAAT +ATTAAC +ATTAAG +ATTATA +ATTATT +ATTATC +ATTATG +ATTACA +ATTACT +ATTACC +ATTACG +ATTAGA +ATTAGT +ATTAGC +ATTAGG +ATTTAA +ATTTAT +ATTTAC +ATTTAG +ATTTTA +ATTTTT +ATTTTC +ATTTTG +ATTTCA +ATTTCT +ATTTCC +ATTTCG +ATTTGA +ATTTGT +ATTTGC +ATTTGG +ATTCAA +ATTCAT +ATTCAC +ATTCAG +ATTCTA +ATTCTT +ATTCTC +ATTCTG +ATTCCA +ATTCCT +ATTCCC +ATTCCG +ATTCGA +ATTCGT +ATTCGC +ATTCGG +ATTGAA +ATTGAT +ATTGAC +ATTGAG +ATTGTA +ATTGTT +ATTGTC +ATTGTG +ATTGCA +ATTGCT +ATTGCC +ATTGCG +ATTGGA +ATTGGT +ATTGGC +ATTGGG +ATCAAA +ATCAAT +ATCAAC +ATCAAG +ATCATA +ATCATT +ATCATC +ATCATG +ATCACA +ATCACT +ATCACC +ATCACG +ATCAGA +ATCAGT +ATCAGC +ATCAGG +ATCTAA +ATCTAT +ATCTAC +ATCTAG +ATCTTA +ATCTTT +ATCTTC +ATCTTG +ATCTCA +ATCTCT +ATCTCC +ATCTCG +ATCTGA +ATCTGT +ATCTGC +ATCTGG +ATCCAA +ATCCAT +ATCCAC +ATCCAG +ATCCTA +ATCCTT +ATCCTC +ATCCTG +ATCCCA +ATCCCT +ATCCCC +ATCCCG +ATCCGA +ATCCGT +ATCCGC +ATCCGG +ATCGAA +ATCGAT +ATCGAC +ATCGAG +ATCGTA +ATCGTT +ATCGTC +ATCGTG +ATCGCA +ATCGCT +ATCGCC +ATCGCG +ATCGGA +ATCGGT +ATCGGC +ATCGGG +ATGAAA +ATGAAT +ATGAAC +ATGAAG +ATGATA +ATGATT +ATGATC +ATGATG +ATGACA +ATGACT +ATGACC +ATGACG +ATGAGA +ATGAGT +ATGAGC +ATGAGG +ATGTAA +ATGTAT +ATGTAC +ATGTAG +ATGTTA +ATGTTT +ATGTTC +ATGTTG +ATGTCA +ATGTCT +ATGTCC +ATGTCG +ATGTGA +ATGTGT +ATGTGC +ATGTGG +ATGCAA +ATGCAT +ATGCAC +ATGCAG +ATGCTA +ATGCTT +ATGCTC +ATGCTG +ATGCCA +ATGCCT +ATGCCC +ATGCCG +ATGCGA +ATGCGT +ATGCGC +ATGCGG +ATGGAA +ATGGAT +ATGGAC +ATGGAG +ATGGTA +ATGGTT +ATGGTC +ATGGTG +ATGGCA +ATGGCT +ATGGCC +ATGGCG +ATGGGA +ATGGGT +ATGGGC +ATGGGG +ACAAAA +ACAAAT +ACAAAC +ACAAAG +ACAATA +ACAATT +ACAATC +ACAATG +ACAACA +ACAACT +ACAACC +ACAACG +ACAAGA +ACAAGT +ACAAGC +ACAAGG +ACATAA +ACATAT +ACATAC +ACATAG +ACATTA +ACATTT +ACATTC +ACATTG +ACATCA +ACATCT +ACATCC +ACATCG +ACATGA +ACATGT +ACATGC +ACATGG +ACACAA +ACACAT +ACACAC +ACACAG +ACACTA +ACACTT +ACACTC +ACACTG +ACACCA +ACACCT +ACACCC +ACACCG +ACACGA +ACACGT +ACACGC +ACACGG +ACAGAA +ACAGAT +ACAGAC +ACAGAG +ACAGTA +ACAGTT +ACAGTC +ACAGTG +ACAGCA +ACAGCT +ACAGCC +ACAGCG +ACAGGA +ACAGGT +ACAGGC +ACAGGG +ACTAAA +ACTAAT +ACTAAC +ACTAAG +ACTATA +ACTATT +ACTATC +ACTATG +ACTACA +ACTACT +ACTACC +ACTACG +ACTAGA +ACTAGT +ACTAGC +ACTAGG +ACTTAA +ACTTAT +ACTTAC +ACTTAG +ACTTTA +ACTTTT +ACTTTC +ACTTTG +ACTTCA +ACTTCT +ACTTCC +ACTTCG +ACTTGA +ACTTGT +ACTTGC +ACTTGG +ACTCAA +ACTCAT +ACTCAC +ACTCAG +ACTCTA +ACTCTT +ACTCTC +ACTCTG +ACTCCA +ACTCCT +ACTCCC +ACTCCG +ACTCGA +ACTCGT +ACTCGC +ACTCGG +ACTGAA +ACTGAT +ACTGAC +ACTGAG +ACTGTA +ACTGTT +ACTGTC +ACTGTG +ACTGCA +ACTGCT +ACTGCC +ACTGCG +ACTGGA +ACTGGT +ACTGGC +ACTGGG +ACCAAA +ACCAAT +ACCAAC +ACCAAG +ACCATA +ACCATT +ACCATC +ACCATG +ACCACA +ACCACT +ACCACC +ACCACG +ACCAGA +ACCAGT +ACCAGC +ACCAGG +ACCTAA +ACCTAT +ACCTAC +ACCTAG +ACCTTA +ACCTTT +ACCTTC +ACCTTG +ACCTCA +ACCTCT +ACCTCC +ACCTCG +ACCTGA +ACCTGT +ACCTGC +ACCTGG +ACCCAA +ACCCAT +ACCCAC +ACCCAG +ACCCTA +ACCCTT +ACCCTC +ACCCTG +ACCCCA +ACCCCT +ACCCCC +ACCCCG +ACCCGA +ACCCGT +ACCCGC +ACCCGG +ACCGAA +ACCGAT +ACCGAC +ACCGAG +ACCGTA +ACCGTT +ACCGTC +ACCGTG +ACCGCA +ACCGCT +ACCGCC +ACCGCG +ACCGGA +ACCGGT +ACCGGC +ACCGGG +ACGAAA +ACGAAT +ACGAAC +ACGAAG +ACGATA +ACGATT +ACGATC +ACGATG +ACGACA +ACGACT +ACGACC +ACGACG +ACGAGA +ACGAGT +ACGAGC +ACGAGG +ACGTAA +ACGTAT +ACGTAC +ACGTAG +ACGTTA +ACGTTT +ACGTTC +ACGTTG +ACGTCA +ACGTCT +ACGTCC +ACGTCG +ACGTGA +ACGTGT +ACGTGC +ACGTGG +ACGCAA +ACGCAT +ACGCAC +ACGCAG +ACGCTA +ACGCTT +ACGCTC +ACGCTG +ACGCCA +ACGCCT +ACGCCC +ACGCCG +ACGCGA +ACGCGT +ACGCGC +ACGCGG +ACGGAA +ACGGAT +ACGGAC +ACGGAG +ACGGTA +ACGGTT +ACGGTC +ACGGTG +ACGGCA +ACGGCT +ACGGCC +ACGGCG +ACGGGA +ACGGGT +ACGGGC +ACGGGG +AGAAAA +AGAAAT +AGAAAC +AGAAAG +AGAATA +AGAATT +AGAATC +AGAATG +AGAACA +AGAACT +AGAACC +AGAACG +AGAAGA +AGAAGT +AGAAGC +AGAAGG +AGATAA +AGATAT +AGATAC +AGATAG +AGATTA +AGATTT +AGATTC +AGATTG +AGATCA +AGATCT +AGATCC +AGATCG +AGATGA +AGATGT +AGATGC +AGATGG +AGACAA +AGACAT +AGACAC +AGACAG +AGACTA +AGACTT +AGACTC +AGACTG +AGACCA +AGACCT +AGACCC +AGACCG +AGACGA +AGACGT +AGACGC +AGACGG +AGAGAA +AGAGAT +AGAGAC +AGAGAG +AGAGTA +AGAGTT +AGAGTC +AGAGTG +AGAGCA +AGAGCT +AGAGCC +AGAGCG +AGAGGA +AGAGGT +AGAGGC +AGAGGG +AGTAAA +AGTAAT +AGTAAC +AGTAAG +AGTATA +AGTATT +AGTATC +AGTATG +AGTACA +AGTACT +AGTACC +AGTACG +AGTAGA +AGTAGT +AGTAGC +AGTAGG +AGTTAA +AGTTAT +AGTTAC +AGTTAG +AGTTTA +AGTTTT +AGTTTC +AGTTTG +AGTTCA +AGTTCT +AGTTCC +AGTTCG +AGTTGA +AGTTGT +AGTTGC +AGTTGG +AGTCAA +AGTCAT +AGTCAC +AGTCAG +AGTCTA +AGTCTT +AGTCTC +AGTCTG +AGTCCA +AGTCCT +AGTCCC +AGTCCG +AGTCGA +AGTCGT +AGTCGC +AGTCGG +AGTGAA +AGTGAT +AGTGAC +AGTGAG +AGTGTA +AGTGTT +AGTGTC +AGTGTG +AGTGCA +AGTGCT +AGTGCC +AGTGCG +AGTGGA +AGTGGT +AGTGGC +AGTGGG +AGCAAA +AGCAAT +AGCAAC +AGCAAG +AGCATA +AGCATT +AGCATC +AGCATG +AGCACA +AGCACT +AGCACC +AGCACG +AGCAGA +AGCAGT +AGCAGC +AGCAGG +AGCTAA +AGCTAT +AGCTAC +AGCTAG +AGCTTA +AGCTTT +AGCTTC +AGCTTG +AGCTCA +AGCTCT +AGCTCC +AGCTCG +AGCTGA +AGCTGT +AGCTGC +AGCTGG +AGCCAA +AGCCAT +AGCCAC +AGCCAG +AGCCTA +AGCCTT +AGCCTC +AGCCTG +AGCCCA +AGCCCT +AGCCCC +AGCCCG +AGCCGA +AGCCGT +AGCCGC +AGCCGG +AGCGAA +AGCGAT +AGCGAC +AGCGAG +AGCGTA +AGCGTT +AGCGTC +AGCGTG +AGCGCA +AGCGCT +AGCGCC +AGCGCG +AGCGGA +AGCGGT +AGCGGC +AGCGGG +AGGAAA +AGGAAT +AGGAAC +AGGAAG +AGGATA +AGGATT +AGGATC +AGGATG +AGGACA +AGGACT +AGGACC +AGGACG +AGGAGA +AGGAGT +AGGAGC +AGGAGG +AGGTAA +AGGTAT +AGGTAC +AGGTAG +AGGTTA +AGGTTT +AGGTTC +AGGTTG +AGGTCA +AGGTCT +AGGTCC +AGGTCG +AGGTGA +AGGTGT +AGGTGC +AGGTGG +AGGCAA +AGGCAT +AGGCAC +AGGCAG +AGGCTA +AGGCTT +AGGCTC +AGGCTG +AGGCCA +AGGCCT +AGGCCC +AGGCCG +AGGCGA +AGGCGT +AGGCGC +AGGCGG +AGGGAA +AGGGAT +AGGGAC +AGGGAG +AGGGTA +AGGGTT +AGGGTC +AGGGTG +AGGGCA +AGGGCT +AGGGCC +AGGGCG +AGGGGA +AGGGGT +AGGGGC +AGGGGG +TAAAAA +TAAAAT +TAAAAC +TAAAAG +TAAATA +TAAATT +TAAATC +TAAATG +TAAACA +TAAACT +TAAACC +TAAACG +TAAAGA +TAAAGT +TAAAGC +TAAAGG +TAATAA +TAATAT +TAATAC +TAATAG +TAATTA +TAATTT +TAATTC +TAATTG +TAATCA +TAATCT +TAATCC +TAATCG +TAATGA +TAATGT +TAATGC +TAATGG +TAACAA +TAACAT +TAACAC +TAACAG +TAACTA +TAACTT +TAACTC +TAACTG +TAACCA +TAACCT +TAACCC +TAACCG +TAACGA +TAACGT +TAACGC +TAACGG +TAAGAA +TAAGAT +TAAGAC +TAAGAG +TAAGTA +TAAGTT +TAAGTC +TAAGTG +TAAGCA +TAAGCT +TAAGCC +TAAGCG +TAAGGA +TAAGGT +TAAGGC +TAAGGG +TATAAA +TATAAT +TATAAC +TATAAG +TATATA +TATATT +TATATC +TATATG +TATACA +TATACT +TATACC +TATACG +TATAGA +TATAGT +TATAGC +TATAGG +TATTAA +TATTAT +TATTAC +TATTAG +TATTTA +TATTTT +TATTTC +TATTTG +TATTCA +TATTCT +TATTCC +TATTCG +TATTGA +TATTGT +TATTGC +TATTGG +TATCAA +TATCAT +TATCAC +TATCAG +TATCTA +TATCTT +TATCTC +TATCTG +TATCCA +TATCCT +TATCCC +TATCCG +TATCGA +TATCGT +TATCGC +TATCGG +TATGAA +TATGAT +TATGAC +TATGAG +TATGTA +TATGTT +TATGTC +TATGTG +TATGCA +TATGCT +TATGCC +TATGCG +TATGGA +TATGGT +TATGGC +TATGGG +TACAAA +TACAAT +TACAAC +TACAAG +TACATA +TACATT +TACATC +TACATG +TACACA +TACACT +TACACC +TACACG +TACAGA +TACAGT +TACAGC +TACAGG +TACTAA +TACTAT +TACTAC +TACTAG +TACTTA +TACTTT +TACTTC +TACTTG +TACTCA +TACTCT +TACTCC +TACTCG +TACTGA +TACTGT +TACTGC +TACTGG +TACCAA +TACCAT +TACCAC +TACCAG +TACCTA +TACCTT +TACCTC +TACCTG +TACCCA +TACCCT +TACCCC +TACCCG +TACCGA +TACCGT +TACCGC +TACCGG +TACGAA +TACGAT +TACGAC +TACGAG +TACGTA +TACGTT +TACGTC +TACGTG +TACGCA +TACGCT +TACGCC +TACGCG +TACGGA +TACGGT +TACGGC +TACGGG +TAGAAA +TAGAAT +TAGAAC +TAGAAG +TAGATA +TAGATT +TAGATC +TAGATG +TAGACA +TAGACT +TAGACC +TAGACG +TAGAGA +TAGAGT +TAGAGC +TAGAGG +TAGTAA +TAGTAT +TAGTAC +TAGTAG +TAGTTA +TAGTTT +TAGTTC +TAGTTG +TAGTCA +TAGTCT +TAGTCC +TAGTCG +TAGTGA +TAGTGT +TAGTGC +TAGTGG +TAGCAA +TAGCAT +TAGCAC +TAGCAG +TAGCTA +TAGCTT +TAGCTC +TAGCTG +TAGCCA +TAGCCT +TAGCCC +TAGCCG +TAGCGA +TAGCGT +TAGCGC +TAGCGG +TAGGAA +TAGGAT +TAGGAC +TAGGAG +TAGGTA +TAGGTT +TAGGTC +TAGGTG +TAGGCA +TAGGCT +TAGGCC +TAGGCG +TAGGGA +TAGGGT +TAGGGC +TAGGGG +TTAAAA +TTAAAT +TTAAAC +TTAAAG +TTAATA +TTAATT +TTAATC +TTAATG +TTAACA +TTAACT +TTAACC +TTAACG +TTAAGA +TTAAGT +TTAAGC +TTAAGG +TTATAA +TTATAT +TTATAC +TTATAG +TTATTA +TTATTT +TTATTC +TTATTG +TTATCA +TTATCT +TTATCC +TTATCG +TTATGA +TTATGT +TTATGC +TTATGG +TTACAA +TTACAT +TTACAC +TTACAG +TTACTA +TTACTT +TTACTC +TTACTG +TTACCA +TTACCT +TTACCC +TTACCG +TTACGA +TTACGT +TTACGC +TTACGG +TTAGAA +TTAGAT +TTAGAC +TTAGAG +TTAGTA +TTAGTT +TTAGTC +TTAGTG +TTAGCA +TTAGCT +TTAGCC +TTAGCG +TTAGGA +TTAGGT +TTAGGC +TTAGGG +TTTAAA +TTTAAT +TTTAAC +TTTAAG +TTTATA +TTTATT +TTTATC +TTTATG +TTTACA +TTTACT +TTTACC +TTTACG +TTTAGA +TTTAGT +TTTAGC +TTTAGG +TTTTAA +TTTTAT +TTTTAC +TTTTAG +TTTTTA +TTTTTT +TTTTTC +TTTTTG +TTTTCA +TTTTCT +TTTTCC +TTTTCG +TTTTGA +TTTTGT +TTTTGC +TTTTGG +TTTCAA +TTTCAT +TTTCAC +TTTCAG +TTTCTA +TTTCTT +TTTCTC +TTTCTG +TTTCCA +TTTCCT +TTTCCC +TTTCCG +TTTCGA +TTTCGT +TTTCGC +TTTCGG +TTTGAA +TTTGAT +TTTGAC +TTTGAG +TTTGTA +TTTGTT +TTTGTC +TTTGTG +TTTGCA +TTTGCT +TTTGCC +TTTGCG +TTTGGA +TTTGGT +TTTGGC +TTTGGG +TTCAAA +TTCAAT +TTCAAC +TTCAAG +TTCATA +TTCATT +TTCATC +TTCATG +TTCACA +TTCACT +TTCACC +TTCACG +TTCAGA +TTCAGT +TTCAGC +TTCAGG +TTCTAA +TTCTAT +TTCTAC +TTCTAG +TTCTTA +TTCTTT +TTCTTC +TTCTTG +TTCTCA +TTCTCT +TTCTCC +TTCTCG +TTCTGA +TTCTGT +TTCTGC +TTCTGG +TTCCAA +TTCCAT +TTCCAC +TTCCAG +TTCCTA +TTCCTT +TTCCTC +TTCCTG +TTCCCA +TTCCCT +TTCCCC +TTCCCG +TTCCGA +TTCCGT +TTCCGC +TTCCGG +TTCGAA +TTCGAT +TTCGAC +TTCGAG +TTCGTA +TTCGTT +TTCGTC +TTCGTG +TTCGCA +TTCGCT +TTCGCC +TTCGCG +TTCGGA +TTCGGT +TTCGGC +TTCGGG +TTGAAA +TTGAAT +TTGAAC +TTGAAG +TTGATA +TTGATT +TTGATC +TTGATG +TTGACA +TTGACT +TTGACC +TTGACG +TTGAGA +TTGAGT +TTGAGC +TTGAGG +TTGTAA +TTGTAT +TTGTAC +TTGTAG +TTGTTA +TTGTTT +TTGTTC +TTGTTG +TTGTCA +TTGTCT +TTGTCC +TTGTCG +TTGTGA +TTGTGT +TTGTGC +TTGTGG +TTGCAA +TTGCAT +TTGCAC +TTGCAG +TTGCTA +TTGCTT +TTGCTC +TTGCTG +TTGCCA +TTGCCT +TTGCCC +TTGCCG +TTGCGA +TTGCGT +TTGCGC +TTGCGG +TTGGAA +TTGGAT +TTGGAC +TTGGAG +TTGGTA +TTGGTT +TTGGTC +TTGGTG +TTGGCA +TTGGCT +TTGGCC +TTGGCG +TTGGGA +TTGGGT +TTGGGC +TTGGGG +TCAAAA +TCAAAT +TCAAAC +TCAAAG +TCAATA +TCAATT +TCAATC +TCAATG +TCAACA +TCAACT +TCAACC +TCAACG +TCAAGA +TCAAGT +TCAAGC +TCAAGG +TCATAA +TCATAT +TCATAC +TCATAG +TCATTA +TCATTT +TCATTC +TCATTG +TCATCA +TCATCT +TCATCC +TCATCG +TCATGA +TCATGT +TCATGC +TCATGG +TCACAA +TCACAT +TCACAC +TCACAG +TCACTA +TCACTT +TCACTC +TCACTG +TCACCA +TCACCT +TCACCC +TCACCG +TCACGA +TCACGT +TCACGC +TCACGG +TCAGAA +TCAGAT +TCAGAC +TCAGAG +TCAGTA +TCAGTT +TCAGTC +TCAGTG +TCAGCA +TCAGCT +TCAGCC +TCAGCG +TCAGGA +TCAGGT +TCAGGC +TCAGGG +TCTAAA +TCTAAT +TCTAAC +TCTAAG +TCTATA +TCTATT +TCTATC +TCTATG +TCTACA +TCTACT +TCTACC +TCTACG +TCTAGA +TCTAGT +TCTAGC +TCTAGG +TCTTAA +TCTTAT +TCTTAC +TCTTAG +TCTTTA +TCTTTT +TCTTTC +TCTTTG +TCTTCA +TCTTCT +TCTTCC +TCTTCG +TCTTGA +TCTTGT +TCTTGC +TCTTGG +TCTCAA +TCTCAT +TCTCAC +TCTCAG +TCTCTA +TCTCTT +TCTCTC +TCTCTG +TCTCCA +TCTCCT +TCTCCC +TCTCCG +TCTCGA +TCTCGT +TCTCGC +TCTCGG +TCTGAA +TCTGAT +TCTGAC +TCTGAG +TCTGTA +TCTGTT +TCTGTC +TCTGTG +TCTGCA +TCTGCT +TCTGCC +TCTGCG +TCTGGA +TCTGGT +TCTGGC +TCTGGG +TCCAAA +TCCAAT +TCCAAC +TCCAAG +TCCATA +TCCATT +TCCATC +TCCATG +TCCACA +TCCACT +TCCACC +TCCACG +TCCAGA +TCCAGT +TCCAGC +TCCAGG +TCCTAA +TCCTAT +TCCTAC +TCCTAG +TCCTTA +TCCTTT +TCCTTC +TCCTTG +TCCTCA +TCCTCT +TCCTCC +TCCTCG +TCCTGA +TCCTGT +TCCTGC +TCCTGG +TCCCAA +TCCCAT +TCCCAC +TCCCAG +TCCCTA +TCCCTT +TCCCTC +TCCCTG +TCCCCA +TCCCCT +TCCCCC +TCCCCG +TCCCGA +TCCCGT +TCCCGC +TCCCGG +TCCGAA +TCCGAT +TCCGAC +TCCGAG +TCCGTA +TCCGTT +TCCGTC +TCCGTG +TCCGCA +TCCGCT +TCCGCC +TCCGCG +TCCGGA +TCCGGT +TCCGGC +TCCGGG +TCGAAA +TCGAAT +TCGAAC +TCGAAG +TCGATA +TCGATT +TCGATC +TCGATG +TCGACA +TCGACT +TCGACC +TCGACG +TCGAGA +TCGAGT +TCGAGC +TCGAGG +TCGTAA +TCGTAT +TCGTAC +TCGTAG +TCGTTA +TCGTTT +TCGTTC +TCGTTG +TCGTCA +TCGTCT +TCGTCC +TCGTCG +TCGTGA +TCGTGT +TCGTGC +TCGTGG +TCGCAA +TCGCAT +TCGCAC +TCGCAG +TCGCTA +TCGCTT +TCGCTC +TCGCTG +TCGCCA +TCGCCT +TCGCCC +TCGCCG +TCGCGA +TCGCGT +TCGCGC +TCGCGG +TCGGAA +TCGGAT +TCGGAC +TCGGAG +TCGGTA +TCGGTT +TCGGTC +TCGGTG +TCGGCA +TCGGCT +TCGGCC +TCGGCG +TCGGGA +TCGGGT +TCGGGC +TCGGGG +TGAAAA +TGAAAT +TGAAAC +TGAAAG +TGAATA +TGAATT +TGAATC +TGAATG +TGAACA +TGAACT +TGAACC +TGAACG +TGAAGA +TGAAGT +TGAAGC +TGAAGG +TGATAA +TGATAT +TGATAC +TGATAG +TGATTA +TGATTT +TGATTC +TGATTG +TGATCA +TGATCT +TGATCC +TGATCG +TGATGA +TGATGT +TGATGC +TGATGG +TGACAA +TGACAT +TGACAC +TGACAG +TGACTA +TGACTT +TGACTC +TGACTG +TGACCA +TGACCT +TGACCC +TGACCG +TGACGA +TGACGT +TGACGC +TGACGG +TGAGAA +TGAGAT +TGAGAC +TGAGAG +TGAGTA +TGAGTT +TGAGTC +TGAGTG +TGAGCA +TGAGCT +TGAGCC +TGAGCG +TGAGGA +TGAGGT +TGAGGC +TGAGGG +TGTAAA +TGTAAT +TGTAAC +TGTAAG +TGTATA +TGTATT +TGTATC +TGTATG +TGTACA +TGTACT +TGTACC +TGTACG +TGTAGA +TGTAGT +TGTAGC +TGTAGG +TGTTAA +TGTTAT +TGTTAC +TGTTAG +TGTTTA +TGTTTT +TGTTTC +TGTTTG +TGTTCA +TGTTCT +TGTTCC +TGTTCG +TGTTGA +TGTTGT +TGTTGC +TGTTGG +TGTCAA +TGTCAT +TGTCAC +TGTCAG +TGTCTA +TGTCTT +TGTCTC +TGTCTG +TGTCCA +TGTCCT +TGTCCC +TGTCCG +TGTCGA +TGTCGT +TGTCGC +TGTCGG +TGTGAA +TGTGAT +TGTGAC +TGTGAG +TGTGTA +TGTGTT +TGTGTC +TGTGTG +TGTGCA +TGTGCT +TGTGCC +TGTGCG +TGTGGA +TGTGGT +TGTGGC +TGTGGG +TGCAAA +TGCAAT +TGCAAC +TGCAAG +TGCATA +TGCATT +TGCATC +TGCATG +TGCACA +TGCACT +TGCACC +TGCACG +TGCAGA +TGCAGT +TGCAGC +TGCAGG +TGCTAA +TGCTAT +TGCTAC +TGCTAG +TGCTTA +TGCTTT +TGCTTC +TGCTTG +TGCTCA +TGCTCT +TGCTCC +TGCTCG +TGCTGA +TGCTGT +TGCTGC +TGCTGG +TGCCAA +TGCCAT +TGCCAC +TGCCAG +TGCCTA +TGCCTT +TGCCTC +TGCCTG +TGCCCA +TGCCCT +TGCCCC +TGCCCG +TGCCGA +TGCCGT +TGCCGC +TGCCGG +TGCGAA +TGCGAT +TGCGAC +TGCGAG +TGCGTA +TGCGTT +TGCGTC +TGCGTG +TGCGCA +TGCGCT +TGCGCC +TGCGCG +TGCGGA +TGCGGT +TGCGGC +TGCGGG +TGGAAA +TGGAAT +TGGAAC +TGGAAG +TGGATA +TGGATT +TGGATC +TGGATG +TGGACA +TGGACT +TGGACC +TGGACG +TGGAGA +TGGAGT +TGGAGC +TGGAGG +TGGTAA +TGGTAT +TGGTAC +TGGTAG +TGGTTA +TGGTTT +TGGTTC +TGGTTG +TGGTCA +TGGTCT +TGGTCC +TGGTCG +TGGTGA +TGGTGT +TGGTGC +TGGTGG +TGGCAA +TGGCAT +TGGCAC +TGGCAG +TGGCTA +TGGCTT +TGGCTC +TGGCTG +TGGCCA +TGGCCT +TGGCCC +TGGCCG +TGGCGA +TGGCGT +TGGCGC +TGGCGG +TGGGAA +TGGGAT +TGGGAC +TGGGAG +TGGGTA +TGGGTT +TGGGTC +TGGGTG +TGGGCA +TGGGCT +TGGGCC +TGGGCG +TGGGGA +TGGGGT +TGGGGC +TGGGGG +CAAAAA +CAAAAT +CAAAAC +CAAAAG +CAAATA +CAAATT +CAAATC +CAAATG +CAAACA +CAAACT +CAAACC +CAAACG +CAAAGA +CAAAGT +CAAAGC +CAAAGG +CAATAA +CAATAT +CAATAC +CAATAG +CAATTA +CAATTT +CAATTC +CAATTG +CAATCA +CAATCT +CAATCC +CAATCG +CAATGA +CAATGT +CAATGC +CAATGG +CAACAA +CAACAT +CAACAC +CAACAG +CAACTA +CAACTT +CAACTC +CAACTG +CAACCA +CAACCT +CAACCC +CAACCG +CAACGA +CAACGT +CAACGC +CAACGG +CAAGAA +CAAGAT +CAAGAC +CAAGAG +CAAGTA +CAAGTT +CAAGTC +CAAGTG +CAAGCA +CAAGCT +CAAGCC +CAAGCG +CAAGGA +CAAGGT +CAAGGC +CAAGGG +CATAAA +CATAAT +CATAAC +CATAAG +CATATA +CATATT +CATATC +CATATG +CATACA +CATACT +CATACC +CATACG +CATAGA +CATAGT +CATAGC +CATAGG +CATTAA +CATTAT +CATTAC +CATTAG +CATTTA +CATTTT +CATTTC +CATTTG +CATTCA +CATTCT +CATTCC +CATTCG +CATTGA +CATTGT +CATTGC +CATTGG +CATCAA +CATCAT +CATCAC +CATCAG +CATCTA +CATCTT +CATCTC +CATCTG +CATCCA +CATCCT +CATCCC +CATCCG +CATCGA +CATCGT +CATCGC +CATCGG +CATGAA +CATGAT +CATGAC +CATGAG +CATGTA +CATGTT +CATGTC +CATGTG +CATGCA +CATGCT +CATGCC +CATGCG +CATGGA +CATGGT +CATGGC +CATGGG +CACAAA +CACAAT +CACAAC +CACAAG +CACATA +CACATT +CACATC +CACATG +CACACA +CACACT +CACACC +CACACG +CACAGA +CACAGT +CACAGC +CACAGG +CACTAA +CACTAT +CACTAC +CACTAG +CACTTA +CACTTT +CACTTC +CACTTG +CACTCA +CACTCT +CACTCC +CACTCG +CACTGA +CACTGT +CACTGC +CACTGG +CACCAA +CACCAT +CACCAC +CACCAG +CACCTA +CACCTT +CACCTC +CACCTG +CACCCA +CACCCT +CACCCC +CACCCG +CACCGA +CACCGT +CACCGC +CACCGG +CACGAA +CACGAT +CACGAC +CACGAG +CACGTA +CACGTT +CACGTC +CACGTG +CACGCA +CACGCT +CACGCC +CACGCG +CACGGA +CACGGT +CACGGC +CACGGG +CAGAAA +CAGAAT +CAGAAC +CAGAAG +CAGATA +CAGATT +CAGATC +CAGATG +CAGACA +CAGACT +CAGACC +CAGACG +CAGAGA +CAGAGT +CAGAGC +CAGAGG +CAGTAA +CAGTAT +CAGTAC +CAGTAG +CAGTTA +CAGTTT +CAGTTC +CAGTTG +CAGTCA +CAGTCT +CAGTCC +CAGTCG +CAGTGA +CAGTGT +CAGTGC +CAGTGG +CAGCAA +CAGCAT +CAGCAC +CAGCAG +CAGCTA +CAGCTT +CAGCTC +CAGCTG +CAGCCA +CAGCCT +CAGCCC +CAGCCG +CAGCGA +CAGCGT +CAGCGC +CAGCGG +CAGGAA +CAGGAT +CAGGAC +CAGGAG +CAGGTA +CAGGTT +CAGGTC +CAGGTG +CAGGCA +CAGGCT +CAGGCC +CAGGCG +CAGGGA +CAGGGT +CAGGGC +CAGGGG +CTAAAA +CTAAAT +CTAAAC +CTAAAG +CTAATA +CTAATT +CTAATC +CTAATG +CTAACA +CTAACT +CTAACC +CTAACG +CTAAGA +CTAAGT +CTAAGC +CTAAGG +CTATAA +CTATAT +CTATAC +CTATAG +CTATTA +CTATTT +CTATTC +CTATTG +CTATCA +CTATCT +CTATCC +CTATCG +CTATGA +CTATGT +CTATGC +CTATGG +CTACAA +CTACAT +CTACAC +CTACAG +CTACTA +CTACTT +CTACTC +CTACTG +CTACCA +CTACCT +CTACCC +CTACCG +CTACGA +CTACGT +CTACGC +CTACGG +CTAGAA +CTAGAT +CTAGAC +CTAGAG +CTAGTA +CTAGTT +CTAGTC +CTAGTG +CTAGCA +CTAGCT +CTAGCC +CTAGCG +CTAGGA +CTAGGT +CTAGGC +CTAGGG +CTTAAA +CTTAAT +CTTAAC +CTTAAG +CTTATA +CTTATT +CTTATC +CTTATG +CTTACA +CTTACT +CTTACC +CTTACG +CTTAGA +CTTAGT +CTTAGC +CTTAGG +CTTTAA +CTTTAT +CTTTAC +CTTTAG +CTTTTA +CTTTTT +CTTTTC +CTTTTG +CTTTCA +CTTTCT +CTTTCC +CTTTCG +CTTTGA +CTTTGT +CTTTGC +CTTTGG +CTTCAA +CTTCAT +CTTCAC +CTTCAG +CTTCTA +CTTCTT +CTTCTC +CTTCTG +CTTCCA +CTTCCT +CTTCCC +CTTCCG +CTTCGA +CTTCGT +CTTCGC +CTTCGG +CTTGAA +CTTGAT +CTTGAC +CTTGAG +CTTGTA +CTTGTT +CTTGTC +CTTGTG +CTTGCA +CTTGCT +CTTGCC +CTTGCG +CTTGGA +CTTGGT +CTTGGC +CTTGGG +CTCAAA +CTCAAT +CTCAAC +CTCAAG +CTCATA +CTCATT +CTCATC +CTCATG +CTCACA +CTCACT +CTCACC +CTCACG +CTCAGA +CTCAGT +CTCAGC +CTCAGG +CTCTAA +CTCTAT +CTCTAC +CTCTAG +CTCTTA +CTCTTT +CTCTTC +CTCTTG +CTCTCA +CTCTCT +CTCTCC +CTCTCG +CTCTGA +CTCTGT +CTCTGC +CTCTGG +CTCCAA +CTCCAT +CTCCAC +CTCCAG +CTCCTA +CTCCTT +CTCCTC +CTCCTG +CTCCCA +CTCCCT +CTCCCC +CTCCCG +CTCCGA +CTCCGT +CTCCGC +CTCCGG +CTCGAA +CTCGAT +CTCGAC +CTCGAG +CTCGTA +CTCGTT +CTCGTC +CTCGTG +CTCGCA +CTCGCT +CTCGCC +CTCGCG +CTCGGA +CTCGGT +CTCGGC +CTCGGG +CTGAAA +CTGAAT +CTGAAC +CTGAAG +CTGATA +CTGATT +CTGATC +CTGATG +CTGACA +CTGACT +CTGACC +CTGACG +CTGAGA +CTGAGT +CTGAGC +CTGAGG +CTGTAA +CTGTAT +CTGTAC +CTGTAG +CTGTTA +CTGTTT +CTGTTC +CTGTTG +CTGTCA +CTGTCT +CTGTCC +CTGTCG +CTGTGA +CTGTGT +CTGTGC +CTGTGG +CTGCAA +CTGCAT +CTGCAC +CTGCAG +CTGCTA +CTGCTT +CTGCTC +CTGCTG +CTGCCA +CTGCCT +CTGCCC +CTGCCG +CTGCGA +CTGCGT +CTGCGC +CTGCGG +CTGGAA +CTGGAT +CTGGAC +CTGGAG +CTGGTA +CTGGTT +CTGGTC +CTGGTG +CTGGCA +CTGGCT +CTGGCC +CTGGCG +CTGGGA +CTGGGT +CTGGGC +CTGGGG +CCAAAA +CCAAAT +CCAAAC +CCAAAG +CCAATA +CCAATT +CCAATC +CCAATG +CCAACA +CCAACT +CCAACC +CCAACG +CCAAGA +CCAAGT +CCAAGC +CCAAGG +CCATAA +CCATAT +CCATAC +CCATAG +CCATTA +CCATTT +CCATTC +CCATTG +CCATCA +CCATCT +CCATCC +CCATCG +CCATGA +CCATGT +CCATGC +CCATGG +CCACAA +CCACAT +CCACAC +CCACAG +CCACTA +CCACTT +CCACTC +CCACTG +CCACCA +CCACCT +CCACCC +CCACCG +CCACGA +CCACGT +CCACGC +CCACGG +CCAGAA +CCAGAT +CCAGAC +CCAGAG +CCAGTA +CCAGTT +CCAGTC +CCAGTG +CCAGCA +CCAGCT +CCAGCC +CCAGCG +CCAGGA +CCAGGT +CCAGGC +CCAGGG +CCTAAA +CCTAAT +CCTAAC +CCTAAG +CCTATA +CCTATT +CCTATC +CCTATG +CCTACA +CCTACT +CCTACC +CCTACG +CCTAGA +CCTAGT +CCTAGC +CCTAGG +CCTTAA +CCTTAT +CCTTAC +CCTTAG +CCTTTA +CCTTTT +CCTTTC +CCTTTG +CCTTCA +CCTTCT +CCTTCC +CCTTCG +CCTTGA +CCTTGT +CCTTGC +CCTTGG +CCTCAA +CCTCAT +CCTCAC +CCTCAG +CCTCTA +CCTCTT +CCTCTC +CCTCTG +CCTCCA +CCTCCT +CCTCCC +CCTCCG +CCTCGA +CCTCGT +CCTCGC +CCTCGG +CCTGAA +CCTGAT +CCTGAC +CCTGAG +CCTGTA +CCTGTT +CCTGTC +CCTGTG +CCTGCA +CCTGCT +CCTGCC +CCTGCG +CCTGGA +CCTGGT +CCTGGC +CCTGGG +CCCAAA +CCCAAT +CCCAAC +CCCAAG +CCCATA +CCCATT +CCCATC +CCCATG +CCCACA +CCCACT +CCCACC +CCCACG +CCCAGA +CCCAGT +CCCAGC +CCCAGG +CCCTAA +CCCTAT +CCCTAC +CCCTAG +CCCTTA +CCCTTT +CCCTTC +CCCTTG +CCCTCA +CCCTCT +CCCTCC +CCCTCG +CCCTGA +CCCTGT +CCCTGC +CCCTGG +CCCCAA +CCCCAT +CCCCAC +CCCCAG +CCCCTA +CCCCTT +CCCCTC +CCCCTG +CCCCCA +CCCCCT +CCCCCC +CCCCCG +CCCCGA +CCCCGT +CCCCGC +CCCCGG +CCCGAA +CCCGAT +CCCGAC +CCCGAG +CCCGTA +CCCGTT +CCCGTC +CCCGTG +CCCGCA +CCCGCT +CCCGCC +CCCGCG +CCCGGA +CCCGGT +CCCGGC +CCCGGG +CCGAAA +CCGAAT +CCGAAC +CCGAAG +CCGATA +CCGATT +CCGATC +CCGATG +CCGACA +CCGACT +CCGACC +CCGACG +CCGAGA +CCGAGT +CCGAGC +CCGAGG +CCGTAA +CCGTAT +CCGTAC +CCGTAG +CCGTTA +CCGTTT +CCGTTC +CCGTTG +CCGTCA +CCGTCT +CCGTCC +CCGTCG +CCGTGA +CCGTGT +CCGTGC +CCGTGG +CCGCAA +CCGCAT +CCGCAC +CCGCAG +CCGCTA +CCGCTT +CCGCTC +CCGCTG +CCGCCA +CCGCCT +CCGCCC +CCGCCG +CCGCGA +CCGCGT +CCGCGC +CCGCGG +CCGGAA +CCGGAT +CCGGAC +CCGGAG +CCGGTA +CCGGTT +CCGGTC +CCGGTG +CCGGCA +CCGGCT +CCGGCC +CCGGCG +CCGGGA +CCGGGT +CCGGGC +CCGGGG +CGAAAA +CGAAAT +CGAAAC +CGAAAG +CGAATA +CGAATT +CGAATC +CGAATG +CGAACA +CGAACT +CGAACC +CGAACG +CGAAGA +CGAAGT +CGAAGC +CGAAGG +CGATAA +CGATAT +CGATAC +CGATAG +CGATTA +CGATTT +CGATTC +CGATTG +CGATCA +CGATCT +CGATCC +CGATCG +CGATGA +CGATGT +CGATGC +CGATGG +CGACAA +CGACAT +CGACAC +CGACAG +CGACTA +CGACTT +CGACTC +CGACTG +CGACCA +CGACCT +CGACCC +CGACCG +CGACGA +CGACGT +CGACGC +CGACGG +CGAGAA +CGAGAT +CGAGAC +CGAGAG +CGAGTA +CGAGTT +CGAGTC +CGAGTG +CGAGCA +CGAGCT +CGAGCC +CGAGCG +CGAGGA +CGAGGT +CGAGGC +CGAGGG +CGTAAA +CGTAAT +CGTAAC +CGTAAG +CGTATA +CGTATT +CGTATC +CGTATG +CGTACA +CGTACT +CGTACC +CGTACG +CGTAGA +CGTAGT +CGTAGC +CGTAGG +CGTTAA +CGTTAT +CGTTAC +CGTTAG +CGTTTA +CGTTTT +CGTTTC +CGTTTG +CGTTCA +CGTTCT +CGTTCC +CGTTCG +CGTTGA +CGTTGT +CGTTGC +CGTTGG +CGTCAA +CGTCAT +CGTCAC +CGTCAG +CGTCTA +CGTCTT +CGTCTC +CGTCTG +CGTCCA +CGTCCT +CGTCCC +CGTCCG +CGTCGA +CGTCGT +CGTCGC +CGTCGG +CGTGAA +CGTGAT +CGTGAC +CGTGAG +CGTGTA +CGTGTT +CGTGTC +CGTGTG +CGTGCA +CGTGCT +CGTGCC +CGTGCG +CGTGGA +CGTGGT +CGTGGC +CGTGGG +CGCAAA +CGCAAT +CGCAAC +CGCAAG +CGCATA +CGCATT +CGCATC +CGCATG +CGCACA +CGCACT +CGCACC +CGCACG +CGCAGA +CGCAGT +CGCAGC +CGCAGG +CGCTAA +CGCTAT +CGCTAC +CGCTAG +CGCTTA +CGCTTT +CGCTTC +CGCTTG +CGCTCA +CGCTCT +CGCTCC +CGCTCG +CGCTGA +CGCTGT +CGCTGC +CGCTGG +CGCCAA +CGCCAT +CGCCAC +CGCCAG +CGCCTA +CGCCTT +CGCCTC +CGCCTG +CGCCCA +CGCCCT +CGCCCC +CGCCCG +CGCCGA +CGCCGT +CGCCGC +CGCCGG +CGCGAA +CGCGAT +CGCGAC +CGCGAG +CGCGTA +CGCGTT +CGCGTC +CGCGTG +CGCGCA +CGCGCT +CGCGCC +CGCGCG +CGCGGA +CGCGGT +CGCGGC +CGCGGG +CGGAAA +CGGAAT +CGGAAC +CGGAAG +CGGATA +CGGATT +CGGATC +CGGATG +CGGACA +CGGACT +CGGACC +CGGACG +CGGAGA +CGGAGT +CGGAGC +CGGAGG +CGGTAA +CGGTAT +CGGTAC +CGGTAG +CGGTTA +CGGTTT +CGGTTC +CGGTTG +CGGTCA +CGGTCT +CGGTCC +CGGTCG +CGGTGA +CGGTGT +CGGTGC +CGGTGG +CGGCAA +CGGCAT +CGGCAC +CGGCAG +CGGCTA +CGGCTT +CGGCTC +CGGCTG +CGGCCA +CGGCCT +CGGCCC +CGGCCG +CGGCGA +CGGCGT +CGGCGC +CGGCGG +CGGGAA +CGGGAT +CGGGAC +CGGGAG +CGGGTA +CGGGTT +CGGGTC +CGGGTG +CGGGCA +CGGGCT +CGGGCC +CGGGCG +CGGGGA +CGGGGT +CGGGGC +CGGGGG +GAAAAA +GAAAAT +GAAAAC +GAAAAG +GAAATA +GAAATT +GAAATC +GAAATG +GAAACA +GAAACT +GAAACC +GAAACG +GAAAGA +GAAAGT +GAAAGC +GAAAGG +GAATAA +GAATAT +GAATAC +GAATAG +GAATTA +GAATTT +GAATTC +GAATTG +GAATCA +GAATCT +GAATCC +GAATCG +GAATGA +GAATGT +GAATGC +GAATGG +GAACAA +GAACAT +GAACAC +GAACAG +GAACTA +GAACTT +GAACTC +GAACTG +GAACCA +GAACCT +GAACCC +GAACCG +GAACGA +GAACGT +GAACGC +GAACGG +GAAGAA +GAAGAT +GAAGAC +GAAGAG +GAAGTA +GAAGTT +GAAGTC +GAAGTG +GAAGCA +GAAGCT +GAAGCC +GAAGCG +GAAGGA +GAAGGT +GAAGGC +GAAGGG +GATAAA +GATAAT +GATAAC +GATAAG +GATATA +GATATT +GATATC +GATATG +GATACA +GATACT +GATACC +GATACG +GATAGA +GATAGT +GATAGC +GATAGG +GATTAA +GATTAT +GATTAC +GATTAG +GATTTA +GATTTT +GATTTC +GATTTG +GATTCA +GATTCT +GATTCC +GATTCG +GATTGA +GATTGT +GATTGC +GATTGG +GATCAA +GATCAT +GATCAC +GATCAG +GATCTA +GATCTT +GATCTC +GATCTG +GATCCA +GATCCT +GATCCC +GATCCG +GATCGA +GATCGT +GATCGC +GATCGG +GATGAA +GATGAT +GATGAC +GATGAG +GATGTA +GATGTT +GATGTC +GATGTG +GATGCA +GATGCT +GATGCC +GATGCG +GATGGA +GATGGT +GATGGC +GATGGG +GACAAA +GACAAT +GACAAC +GACAAG +GACATA +GACATT +GACATC +GACATG +GACACA +GACACT +GACACC +GACACG +GACAGA +GACAGT +GACAGC +GACAGG +GACTAA +GACTAT +GACTAC +GACTAG +GACTTA +GACTTT +GACTTC +GACTTG +GACTCA +GACTCT +GACTCC +GACTCG +GACTGA +GACTGT +GACTGC +GACTGG +GACCAA +GACCAT +GACCAC +GACCAG +GACCTA +GACCTT +GACCTC +GACCTG +GACCCA +GACCCT +GACCCC +GACCCG +GACCGA +GACCGT +GACCGC +GACCGG +GACGAA +GACGAT +GACGAC +GACGAG +GACGTA +GACGTT +GACGTC +GACGTG +GACGCA +GACGCT +GACGCC +GACGCG +GACGGA +GACGGT +GACGGC +GACGGG +GAGAAA +GAGAAT +GAGAAC +GAGAAG +GAGATA +GAGATT +GAGATC +GAGATG +GAGACA +GAGACT +GAGACC +GAGACG +GAGAGA +GAGAGT +GAGAGC +GAGAGG +GAGTAA +GAGTAT +GAGTAC +GAGTAG +GAGTTA +GAGTTT +GAGTTC +GAGTTG +GAGTCA +GAGTCT +GAGTCC +GAGTCG +GAGTGA +GAGTGT +GAGTGC +GAGTGG +GAGCAA +GAGCAT +GAGCAC +GAGCAG +GAGCTA +GAGCTT +GAGCTC +GAGCTG +GAGCCA +GAGCCT +GAGCCC +GAGCCG +GAGCGA +GAGCGT +GAGCGC +GAGCGG +GAGGAA +GAGGAT +GAGGAC +GAGGAG +GAGGTA +GAGGTT +GAGGTC +GAGGTG +GAGGCA +GAGGCT +GAGGCC +GAGGCG +GAGGGA +GAGGGT +GAGGGC +GAGGGG +GTAAAA +GTAAAT +GTAAAC +GTAAAG +GTAATA +GTAATT +GTAATC +GTAATG +GTAACA +GTAACT +GTAACC +GTAACG +GTAAGA +GTAAGT +GTAAGC +GTAAGG +GTATAA +GTATAT +GTATAC +GTATAG +GTATTA +GTATTT +GTATTC +GTATTG +GTATCA +GTATCT +GTATCC +GTATCG +GTATGA +GTATGT +GTATGC +GTATGG +GTACAA +GTACAT +GTACAC +GTACAG +GTACTA +GTACTT +GTACTC +GTACTG +GTACCA +GTACCT +GTACCC +GTACCG +GTACGA +GTACGT +GTACGC +GTACGG +GTAGAA +GTAGAT +GTAGAC +GTAGAG +GTAGTA +GTAGTT +GTAGTC +GTAGTG +GTAGCA +GTAGCT +GTAGCC +GTAGCG +GTAGGA +GTAGGT +GTAGGC +GTAGGG +GTTAAA +GTTAAT +GTTAAC +GTTAAG +GTTATA +GTTATT +GTTATC +GTTATG +GTTACA +GTTACT +GTTACC +GTTACG +GTTAGA +GTTAGT +GTTAGC +GTTAGG +GTTTAA +GTTTAT +GTTTAC +GTTTAG +GTTTTA +GTTTTT +GTTTTC +GTTTTG +GTTTCA +GTTTCT +GTTTCC +GTTTCG +GTTTGA +GTTTGT +GTTTGC +GTTTGG +GTTCAA +GTTCAT +GTTCAC +GTTCAG +GTTCTA +GTTCTT +GTTCTC +GTTCTG +GTTCCA +GTTCCT +GTTCCC +GTTCCG +GTTCGA +GTTCGT +GTTCGC +GTTCGG +GTTGAA +GTTGAT +GTTGAC +GTTGAG +GTTGTA +GTTGTT +GTTGTC +GTTGTG +GTTGCA +GTTGCT +GTTGCC +GTTGCG +GTTGGA +GTTGGT +GTTGGC +GTTGGG +GTCAAA +GTCAAT +GTCAAC +GTCAAG +GTCATA +GTCATT +GTCATC +GTCATG +GTCACA +GTCACT +GTCACC +GTCACG +GTCAGA +GTCAGT +GTCAGC +GTCAGG +GTCTAA +GTCTAT +GTCTAC +GTCTAG +GTCTTA +GTCTTT +GTCTTC +GTCTTG +GTCTCA +GTCTCT +GTCTCC +GTCTCG +GTCTGA +GTCTGT +GTCTGC +GTCTGG +GTCCAA +GTCCAT +GTCCAC +GTCCAG +GTCCTA +GTCCTT +GTCCTC +GTCCTG +GTCCCA +GTCCCT +GTCCCC +GTCCCG +GTCCGA +GTCCGT +GTCCGC +GTCCGG +GTCGAA +GTCGAT +GTCGAC +GTCGAG +GTCGTA +GTCGTT +GTCGTC +GTCGTG +GTCGCA +GTCGCT +GTCGCC +GTCGCG +GTCGGA +GTCGGT +GTCGGC +GTCGGG +GTGAAA +GTGAAT +GTGAAC +GTGAAG +GTGATA +GTGATT +GTGATC +GTGATG +GTGACA +GTGACT +GTGACC +GTGACG +GTGAGA +GTGAGT +GTGAGC +GTGAGG +GTGTAA +GTGTAT +GTGTAC +GTGTAG +GTGTTA +GTGTTT +GTGTTC +GTGTTG +GTGTCA +GTGTCT +GTGTCC +GTGTCG +GTGTGA +GTGTGT +GTGTGC +GTGTGG +GTGCAA +GTGCAT +GTGCAC +GTGCAG +GTGCTA +GTGCTT +GTGCTC +GTGCTG +GTGCCA +GTGCCT +GTGCCC +GTGCCG +GTGCGA +GTGCGT +GTGCGC +GTGCGG +GTGGAA +GTGGAT +GTGGAC +GTGGAG +GTGGTA +GTGGTT +GTGGTC +GTGGTG +GTGGCA +GTGGCT +GTGGCC +GTGGCG +GTGGGA +GTGGGT +GTGGGC +GTGGGG +GCAAAA +GCAAAT +GCAAAC +GCAAAG +GCAATA +GCAATT +GCAATC +GCAATG +GCAACA +GCAACT +GCAACC +GCAACG +GCAAGA +GCAAGT +GCAAGC +GCAAGG +GCATAA +GCATAT +GCATAC +GCATAG +GCATTA +GCATTT +GCATTC +GCATTG +GCATCA +GCATCT +GCATCC +GCATCG +GCATGA +GCATGT +GCATGC +GCATGG +GCACAA +GCACAT +GCACAC +GCACAG +GCACTA +GCACTT +GCACTC +GCACTG +GCACCA +GCACCT +GCACCC +GCACCG +GCACGA +GCACGT +GCACGC +GCACGG +GCAGAA +GCAGAT +GCAGAC +GCAGAG +GCAGTA +GCAGTT +GCAGTC +GCAGTG +GCAGCA +GCAGCT +GCAGCC +GCAGCG +GCAGGA +GCAGGT +GCAGGC +GCAGGG +GCTAAA +GCTAAT +GCTAAC +GCTAAG +GCTATA +GCTATT +GCTATC +GCTATG +GCTACA +GCTACT +GCTACC +GCTACG +GCTAGA +GCTAGT +GCTAGC +GCTAGG +GCTTAA +GCTTAT +GCTTAC +GCTTAG +GCTTTA +GCTTTT +GCTTTC +GCTTTG +GCTTCA +GCTTCT +GCTTCC +GCTTCG +GCTTGA +GCTTGT +GCTTGC +GCTTGG +GCTCAA +GCTCAT +GCTCAC +GCTCAG +GCTCTA +GCTCTT +GCTCTC +GCTCTG +GCTCCA +GCTCCT +GCTCCC +GCTCCG +GCTCGA +GCTCGT +GCTCGC +GCTCGG +GCTGAA +GCTGAT +GCTGAC +GCTGAG +GCTGTA +GCTGTT +GCTGTC +GCTGTG +GCTGCA +GCTGCT +GCTGCC +GCTGCG +GCTGGA +GCTGGT +GCTGGC +GCTGGG +GCCAAA +GCCAAT +GCCAAC +GCCAAG +GCCATA +GCCATT +GCCATC +GCCATG +GCCACA +GCCACT +GCCACC +GCCACG +GCCAGA +GCCAGT +GCCAGC +GCCAGG +GCCTAA +GCCTAT +GCCTAC +GCCTAG +GCCTTA +GCCTTT +GCCTTC +GCCTTG +GCCTCA +GCCTCT +GCCTCC +GCCTCG +GCCTGA +GCCTGT +GCCTGC +GCCTGG +GCCCAA +GCCCAT +GCCCAC +GCCCAG +GCCCTA +GCCCTT +GCCCTC +GCCCTG +GCCCCA +GCCCCT +GCCCCC +GCCCCG +GCCCGA +GCCCGT +GCCCGC +GCCCGG +GCCGAA +GCCGAT +GCCGAC +GCCGAG +GCCGTA +GCCGTT +GCCGTC +GCCGTG +GCCGCA +GCCGCT +GCCGCC +GCCGCG +GCCGGA +GCCGGT +GCCGGC +GCCGGG +GCGAAA +GCGAAT +GCGAAC +GCGAAG +GCGATA +GCGATT +GCGATC +GCGATG +GCGACA +GCGACT +GCGACC +GCGACG +GCGAGA +GCGAGT +GCGAGC +GCGAGG +GCGTAA +GCGTAT +GCGTAC +GCGTAG +GCGTTA +GCGTTT +GCGTTC +GCGTTG +GCGTCA +GCGTCT +GCGTCC +GCGTCG +GCGTGA +GCGTGT +GCGTGC +GCGTGG +GCGCAA +GCGCAT +GCGCAC +GCGCAG +GCGCTA +GCGCTT +GCGCTC +GCGCTG +GCGCCA +GCGCCT +GCGCCC +GCGCCG +GCGCGA +GCGCGT +GCGCGC +GCGCGG +GCGGAA +GCGGAT +GCGGAC +GCGGAG +GCGGTA +GCGGTT +GCGGTC +GCGGTG +GCGGCA +GCGGCT +GCGGCC +GCGGCG +GCGGGA +GCGGGT +GCGGGC +GCGGGG +GGAAAA +GGAAAT +GGAAAC +GGAAAG +GGAATA +GGAATT +GGAATC +GGAATG +GGAACA +GGAACT +GGAACC +GGAACG +GGAAGA +GGAAGT +GGAAGC +GGAAGG +GGATAA +GGATAT +GGATAC +GGATAG +GGATTA +GGATTT +GGATTC +GGATTG +GGATCA +GGATCT +GGATCC +GGATCG +GGATGA +GGATGT +GGATGC +GGATGG +GGACAA +GGACAT +GGACAC +GGACAG +GGACTA +GGACTT +GGACTC +GGACTG +GGACCA +GGACCT +GGACCC +GGACCG +GGACGA +GGACGT +GGACGC +GGACGG +GGAGAA +GGAGAT +GGAGAC +GGAGAG +GGAGTA +GGAGTT +GGAGTC +GGAGTG +GGAGCA +GGAGCT +GGAGCC +GGAGCG +GGAGGA +GGAGGT +GGAGGC +GGAGGG +GGTAAA +GGTAAT +GGTAAC +GGTAAG +GGTATA +GGTATT +GGTATC +GGTATG +GGTACA +GGTACT +GGTACC +GGTACG +GGTAGA +GGTAGT +GGTAGC +GGTAGG +GGTTAA +GGTTAT +GGTTAC +GGTTAG +GGTTTA +GGTTTT +GGTTTC +GGTTTG +GGTTCA +GGTTCT +GGTTCC +GGTTCG +GGTTGA +GGTTGT +GGTTGC +GGTTGG +GGTCAA +GGTCAT +GGTCAC +GGTCAG +GGTCTA +GGTCTT +GGTCTC +GGTCTG +GGTCCA +GGTCCT +GGTCCC +GGTCCG +GGTCGA +GGTCGT +GGTCGC +GGTCGG +GGTGAA +GGTGAT +GGTGAC +GGTGAG +GGTGTA +GGTGTT +GGTGTC +GGTGTG +GGTGCA +GGTGCT +GGTGCC +GGTGCG +GGTGGA +GGTGGT +GGTGGC +GGTGGG +GGCAAA +GGCAAT +GGCAAC +GGCAAG +GGCATA +GGCATT +GGCATC +GGCATG +GGCACA +GGCACT +GGCACC +GGCACG +GGCAGA +GGCAGT +GGCAGC +GGCAGG +GGCTAA +GGCTAT +GGCTAC +GGCTAG +GGCTTA +GGCTTT +GGCTTC +GGCTTG +GGCTCA +GGCTCT +GGCTCC +GGCTCG +GGCTGA +GGCTGT +GGCTGC +GGCTGG +GGCCAA +GGCCAT +GGCCAC +GGCCAG +GGCCTA +GGCCTT +GGCCTC +GGCCTG +GGCCCA +GGCCCT +GGCCCC +GGCCCG +GGCCGA +GGCCGT +GGCCGC +GGCCGG +GGCGAA +GGCGAT +GGCGAC +GGCGAG +GGCGTA +GGCGTT +GGCGTC +GGCGTG +GGCGCA +GGCGCT +GGCGCC +GGCGCG +GGCGGA +GGCGGT +GGCGGC +GGCGGG +GGGAAA +GGGAAT +GGGAAC +GGGAAG +GGGATA +GGGATT +GGGATC +GGGATG +GGGACA +GGGACT +GGGACC +GGGACG +GGGAGA +GGGAGT +GGGAGC +GGGAGG +GGGTAA +GGGTAT +GGGTAC +GGGTAG +GGGTTA +GGGTTT +GGGTTC +GGGTTG +GGGTCA +GGGTCT +GGGTCC +GGGTCG +GGGTGA +GGGTGT +GGGTGC +GGGTGG +GGGCAA +GGGCAT +GGGCAC +GGGCAG +GGGCTA +GGGCTT +GGGCTC +GGGCTG +GGGCCA +GGGCCT +GGGCCC +GGGCCG +GGGCGA +GGGCGT +GGGCGC +GGGCGG +GGGGAA +GGGGAT +GGGGAC +GGGGAG +GGGGTA +GGGGTT +GGGGTC +GGGGTG +GGGGCA +GGGGCT +GGGGCC +GGGGCG +GGGGGA +GGGGGT +GGGGGC +GGGGGG diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/README.md b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/README.md new file mode 100644 index 0000000000000000000000000000000000000000..500fd976d92b217455f181bc98c77e612d9a18d9 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/README.md @@ -0,0 +1,202 @@ +--- +base_model: fabihamakhdoomi/TinyDNABERT +library_name: peft +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.12.0 \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/adapter_config.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..28c871c93775eedfaf266f00da8c02f96bc02f29 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/adapter_config.json @@ -0,0 +1,42 @@ +{ + "alpha_pattern": {}, + "auto_mapping": null, + "base_model_name_or_path": "fabihamakhdoomi/TinyDNABERT", + "bias": "none", + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 16, + "lora_dropout": 0.2, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": [ + "classifier", + "score" + ], + "peft_type": "LORA", + "r": 4, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "bert.encoder.layer.1.attention.self.query", + "bert.encoder.layer.0.output.dense", + "bert.encoder.layer.1.output.dense", + "bert.encoder.layer.0.intermediate.dense", + "bert.encoder.layer.1.intermediate.dense", + "bert.encoder.layer.0.attention.output.dense", + "bert.encoder.layer.1.attention.output.dense", + "bert.encoder.layer.0.attention.self.query", + "bert.encoder.layer.0.attention.self.value", + "bert.encoder.layer.1.attention.self.key", + "bert.encoder.layer.1.attention.self.value", + "bert.encoder.layer.0.attention.self.key" + ], + "task_type": "SEQ_CLS", + "use_dora": false, + "use_rslora": false +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/adapter_model.safetensors b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/adapter_model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..63257bc32400fc7dddb0ba4de9ae099ff2ba1525 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/adapter_model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:117914ee9ad91aee1b4e9239d554d9be32e8633c9d60c430763a6681afbaf9f1 +size 78208 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/optimizer.pt b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..4367f15bb400bf89aceb9db33681f23e2068c6a2 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c5fd1a86bee5e05f13c0f7fe96cd3596d5927e9355f14bfc96afce351cb812d8 +size 165226 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/rng_state.pth b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..506b93da132609b3f461ebc7645f04db448a3371 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:77b9489769cad6c0f8fc553b5b129c47588998775afee5582baf3b7ef2be1cfa +size 14244 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/scheduler.pt b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..4c34d9dcafc0eb1507a84da1af7c2c5228be082e --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:20611d193eaf6658e6192005c679e5aaa561ac33254f734047a2032edc3e8f9b +size 1064 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/special_tokens_map.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..a8b3208c2884c4efb86e49300fdd3dc877220cdf --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/special_tokens_map.json @@ -0,0 +1,7 @@ +{ + "cls_token": "[CLS]", + "mask_token": "[MASK]", + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/tokenizer.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..ebd782de6963b9032ce9374cfe30f5275be1a30b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/tokenizer.json @@ -0,0 +1,4265 @@ +{ + "version": "1.0", + "truncation": { + "direction": "Right", + "max_length": 512, + "strategy": "LongestFirst", + "stride": 0 + }, + "padding": { + "strategy": { + "Fixed": 512 + }, + "direction": "Right", + "pad_to_multiple_of": null, + "pad_id": 0, + "pad_type_id": 0, + "pad_token": "[PAD]" + }, + "added_tokens": [ + { + "id": 0, + "content": "[PAD]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 1, + "content": "[UNK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 2, + "content": "[CLS]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 3, + "content": "[SEP]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 4, + "content": "[MASK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + } + ], + "normalizer": { + "type": "BertNormalizer", + "clean_text": true, + "handle_chinese_chars": true, + "strip_accents": null, + "lowercase": false + }, + "pre_tokenizer": { + "type": "BertPreTokenizer" + }, + "post_processor": { + "type": "TemplateProcessing", + "single": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + } + ], + "pair": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "B", + "type_id": 1 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 1 + } + } + ], + "special_tokens": { + "[CLS]": { + "id": "[CLS]", + "ids": [ + 2 + ], + "tokens": [ + "[CLS]" + ] + }, + "[SEP]": { + "id": "[SEP]", + "ids": [ + 3 + ], + "tokens": [ + "[SEP]" + ] + } + } + }, + "decoder": { + "type": "WordPiece", + "prefix": "##", + "cleanup": true + }, + "model": { + "type": "WordPiece", + "unk_token": "[UNK]", + "continuing_subword_prefix": "##", + "max_input_chars_per_word": 100, + "vocab": { + "[PAD]": 0, + "[UNK]": 1, + "[CLS]": 2, + "[SEP]": 3, + "[MASK]": 4, + "AAAAAA": 5, + "AAAAAT": 6, + "AAAAAC": 7, + "AAAAAG": 8, + "AAAATA": 9, + "AAAATT": 10, + "AAAATC": 11, + "AAAATG": 12, + "AAAACA": 13, + "AAAACT": 14, + "AAAACC": 15, + "AAAACG": 16, + "AAAAGA": 17, + "AAAAGT": 18, + "AAAAGC": 19, + "AAAAGG": 20, + "AAATAA": 21, + "AAATAT": 22, + "AAATAC": 23, + "AAATAG": 24, + "AAATTA": 25, + "AAATTT": 26, + "AAATTC": 27, + "AAATTG": 28, + "AAATCA": 29, + "AAATCT": 30, + "AAATCC": 31, + "AAATCG": 32, + "AAATGA": 33, + "AAATGT": 34, + "AAATGC": 35, + "AAATGG": 36, + "AAACAA": 37, + "AAACAT": 38, + "AAACAC": 39, + "AAACAG": 40, + "AAACTA": 41, + "AAACTT": 42, + "AAACTC": 43, + "AAACTG": 44, + "AAACCA": 45, + "AAACCT": 46, + "AAACCC": 47, + "AAACCG": 48, + "AAACGA": 49, + "AAACGT": 50, + "AAACGC": 51, + "AAACGG": 52, + "AAAGAA": 53, + "AAAGAT": 54, + "AAAGAC": 55, + "AAAGAG": 56, + "AAAGTA": 57, + "AAAGTT": 58, + "AAAGTC": 59, + "AAAGTG": 60, + "AAAGCA": 61, + "AAAGCT": 62, + "AAAGCC": 63, + "AAAGCG": 64, + "AAAGGA": 65, + "AAAGGT": 66, + "AAAGGC": 67, + "AAAGGG": 68, + "AATAAA": 69, + "AATAAT": 70, + "AATAAC": 71, + "AATAAG": 72, + "AATATA": 73, + "AATATT": 74, + "AATATC": 75, + "AATATG": 76, + "AATACA": 77, + "AATACT": 78, + "AATACC": 79, + "AATACG": 80, + "AATAGA": 81, + "AATAGT": 82, + "AATAGC": 83, + "AATAGG": 84, + "AATTAA": 85, + "AATTAT": 86, + "AATTAC": 87, + "AATTAG": 88, + "AATTTA": 89, + "AATTTT": 90, + "AATTTC": 91, + "AATTTG": 92, + "AATTCA": 93, + "AATTCT": 94, + "AATTCC": 95, + "AATTCG": 96, + "AATTGA": 97, + "AATTGT": 98, + "AATTGC": 99, + "AATTGG": 100, + "AATCAA": 101, + "AATCAT": 102, + "AATCAC": 103, + "AATCAG": 104, + "AATCTA": 105, + "AATCTT": 106, + "AATCTC": 107, + "AATCTG": 108, + "AATCCA": 109, + "AATCCT": 110, + "AATCCC": 111, + "AATCCG": 112, + "AATCGA": 113, + "AATCGT": 114, + "AATCGC": 115, + "AATCGG": 116, + "AATGAA": 117, + "AATGAT": 118, + "AATGAC": 119, + "AATGAG": 120, + "AATGTA": 121, + "AATGTT": 122, + "AATGTC": 123, + "AATGTG": 124, + "AATGCA": 125, + "AATGCT": 126, + "AATGCC": 127, + "AATGCG": 128, + "AATGGA": 129, + "AATGGT": 130, + "AATGGC": 131, + "AATGGG": 132, + "AACAAA": 133, + "AACAAT": 134, + "AACAAC": 135, + "AACAAG": 136, + "AACATA": 137, + "AACATT": 138, + "AACATC": 139, + "AACATG": 140, + "AACACA": 141, + "AACACT": 142, + "AACACC": 143, + "AACACG": 144, + "AACAGA": 145, + "AACAGT": 146, + "AACAGC": 147, + "AACAGG": 148, + "AACTAA": 149, + "AACTAT": 150, + "AACTAC": 151, + "AACTAG": 152, + "AACTTA": 153, + "AACTTT": 154, + "AACTTC": 155, + "AACTTG": 156, + "AACTCA": 157, + "AACTCT": 158, + "AACTCC": 159, + "AACTCG": 160, + "AACTGA": 161, + "AACTGT": 162, + "AACTGC": 163, + "AACTGG": 164, + "AACCAA": 165, + "AACCAT": 166, + "AACCAC": 167, + "AACCAG": 168, + "AACCTA": 169, + "AACCTT": 170, + "AACCTC": 171, + "AACCTG": 172, + "AACCCA": 173, + "AACCCT": 174, + "AACCCC": 175, + "AACCCG": 176, + "AACCGA": 177, + "AACCGT": 178, + "AACCGC": 179, + "AACCGG": 180, + "AACGAA": 181, + "AACGAT": 182, + "AACGAC": 183, + "AACGAG": 184, + "AACGTA": 185, + "AACGTT": 186, + "AACGTC": 187, + "AACGTG": 188, + "AACGCA": 189, + "AACGCT": 190, + "AACGCC": 191, + "AACGCG": 192, + "AACGGA": 193, + "AACGGT": 194, + "AACGGC": 195, + "AACGGG": 196, + "AAGAAA": 197, + "AAGAAT": 198, + "AAGAAC": 199, + "AAGAAG": 200, + "AAGATA": 201, + "AAGATT": 202, + "AAGATC": 203, + "AAGATG": 204, + "AAGACA": 205, + "AAGACT": 206, + "AAGACC": 207, + "AAGACG": 208, + "AAGAGA": 209, + "AAGAGT": 210, + "AAGAGC": 211, + "AAGAGG": 212, + "AAGTAA": 213, + "AAGTAT": 214, + "AAGTAC": 215, + "AAGTAG": 216, + "AAGTTA": 217, + "AAGTTT": 218, + "AAGTTC": 219, + "AAGTTG": 220, + "AAGTCA": 221, + "AAGTCT": 222, + "AAGTCC": 223, + "AAGTCG": 224, + "AAGTGA": 225, + "AAGTGT": 226, + "AAGTGC": 227, + "AAGTGG": 228, + "AAGCAA": 229, + "AAGCAT": 230, + "AAGCAC": 231, + "AAGCAG": 232, + "AAGCTA": 233, + "AAGCTT": 234, + "AAGCTC": 235, + "AAGCTG": 236, + "AAGCCA": 237, + "AAGCCT": 238, + "AAGCCC": 239, + "AAGCCG": 240, + "AAGCGA": 241, + "AAGCGT": 242, + "AAGCGC": 243, + "AAGCGG": 244, + "AAGGAA": 245, + "AAGGAT": 246, + "AAGGAC": 247, + "AAGGAG": 248, + "AAGGTA": 249, + "AAGGTT": 250, + "AAGGTC": 251, + "AAGGTG": 252, + "AAGGCA": 253, + "AAGGCT": 254, + "AAGGCC": 255, + "AAGGCG": 256, + "AAGGGA": 257, + "AAGGGT": 258, + "AAGGGC": 259, + "AAGGGG": 260, + "ATAAAA": 261, + "ATAAAT": 262, + "ATAAAC": 263, + "ATAAAG": 264, + "ATAATA": 265, + "ATAATT": 266, + "ATAATC": 267, + "ATAATG": 268, + "ATAACA": 269, + "ATAACT": 270, + "ATAACC": 271, + "ATAACG": 272, + "ATAAGA": 273, + "ATAAGT": 274, + "ATAAGC": 275, + "ATAAGG": 276, + "ATATAA": 277, + "ATATAT": 278, + "ATATAC": 279, + "ATATAG": 280, + "ATATTA": 281, + "ATATTT": 282, + "ATATTC": 283, + "ATATTG": 284, + "ATATCA": 285, + "ATATCT": 286, + "ATATCC": 287, + "ATATCG": 288, + "ATATGA": 289, + "ATATGT": 290, + "ATATGC": 291, + "ATATGG": 292, + "ATACAA": 293, + "ATACAT": 294, + "ATACAC": 295, + "ATACAG": 296, + "ATACTA": 297, + "ATACTT": 298, + "ATACTC": 299, + "ATACTG": 300, + "ATACCA": 301, + "ATACCT": 302, + "ATACCC": 303, + "ATACCG": 304, + "ATACGA": 305, + "ATACGT": 306, + "ATACGC": 307, + "ATACGG": 308, + "ATAGAA": 309, + "ATAGAT": 310, + "ATAGAC": 311, + "ATAGAG": 312, + "ATAGTA": 313, + "ATAGTT": 314, + "ATAGTC": 315, + "ATAGTG": 316, + "ATAGCA": 317, + "ATAGCT": 318, + "ATAGCC": 319, + "ATAGCG": 320, + "ATAGGA": 321, + "ATAGGT": 322, + "ATAGGC": 323, + "ATAGGG": 324, + "ATTAAA": 325, + "ATTAAT": 326, + "ATTAAC": 327, + "ATTAAG": 328, + "ATTATA": 329, + "ATTATT": 330, + "ATTATC": 331, + "ATTATG": 332, + "ATTACA": 333, + "ATTACT": 334, + "ATTACC": 335, + "ATTACG": 336, + "ATTAGA": 337, + "ATTAGT": 338, + "ATTAGC": 339, + "ATTAGG": 340, + "ATTTAA": 341, + "ATTTAT": 342, + "ATTTAC": 343, + "ATTTAG": 344, + "ATTTTA": 345, + "ATTTTT": 346, + "ATTTTC": 347, + "ATTTTG": 348, + "ATTTCA": 349, + "ATTTCT": 350, + "ATTTCC": 351, + "ATTTCG": 352, + "ATTTGA": 353, + "ATTTGT": 354, + "ATTTGC": 355, + "ATTTGG": 356, + "ATTCAA": 357, + "ATTCAT": 358, + "ATTCAC": 359, + "ATTCAG": 360, + "ATTCTA": 361, + "ATTCTT": 362, + "ATTCTC": 363, + "ATTCTG": 364, + "ATTCCA": 365, + "ATTCCT": 366, + "ATTCCC": 367, + "ATTCCG": 368, + "ATTCGA": 369, + "ATTCGT": 370, + "ATTCGC": 371, + "ATTCGG": 372, + "ATTGAA": 373, + "ATTGAT": 374, + "ATTGAC": 375, + "ATTGAG": 376, + "ATTGTA": 377, + "ATTGTT": 378, + "ATTGTC": 379, + "ATTGTG": 380, + "ATTGCA": 381, + "ATTGCT": 382, + "ATTGCC": 383, + "ATTGCG": 384, + "ATTGGA": 385, + "ATTGGT": 386, + "ATTGGC": 387, + "ATTGGG": 388, + "ATCAAA": 389, + "ATCAAT": 390, + "ATCAAC": 391, + "ATCAAG": 392, + "ATCATA": 393, + "ATCATT": 394, + "ATCATC": 395, + "ATCATG": 396, + "ATCACA": 397, + "ATCACT": 398, + "ATCACC": 399, + "ATCACG": 400, + "ATCAGA": 401, + "ATCAGT": 402, + "ATCAGC": 403, + "ATCAGG": 404, + "ATCTAA": 405, + "ATCTAT": 406, + "ATCTAC": 407, + "ATCTAG": 408, + "ATCTTA": 409, + "ATCTTT": 410, + "ATCTTC": 411, + "ATCTTG": 412, + "ATCTCA": 413, + "ATCTCT": 414, + "ATCTCC": 415, + "ATCTCG": 416, + "ATCTGA": 417, + "ATCTGT": 418, + "ATCTGC": 419, + "ATCTGG": 420, + "ATCCAA": 421, + "ATCCAT": 422, + "ATCCAC": 423, + "ATCCAG": 424, + "ATCCTA": 425, + "ATCCTT": 426, + "ATCCTC": 427, + "ATCCTG": 428, + "ATCCCA": 429, + "ATCCCT": 430, + "ATCCCC": 431, + "ATCCCG": 432, + "ATCCGA": 433, + "ATCCGT": 434, + "ATCCGC": 435, + "ATCCGG": 436, + "ATCGAA": 437, + "ATCGAT": 438, + "ATCGAC": 439, + "ATCGAG": 440, + "ATCGTA": 441, + "ATCGTT": 442, + "ATCGTC": 443, + "ATCGTG": 444, + "ATCGCA": 445, + "ATCGCT": 446, + "ATCGCC": 447, + "ATCGCG": 448, + "ATCGGA": 449, + "ATCGGT": 450, + "ATCGGC": 451, + "ATCGGG": 452, + "ATGAAA": 453, + "ATGAAT": 454, + "ATGAAC": 455, + "ATGAAG": 456, + "ATGATA": 457, + "ATGATT": 458, + "ATGATC": 459, + "ATGATG": 460, + "ATGACA": 461, + "ATGACT": 462, + "ATGACC": 463, + "ATGACG": 464, + "ATGAGA": 465, + "ATGAGT": 466, + "ATGAGC": 467, + "ATGAGG": 468, + "ATGTAA": 469, + "ATGTAT": 470, + "ATGTAC": 471, + "ATGTAG": 472, + "ATGTTA": 473, + "ATGTTT": 474, + "ATGTTC": 475, + "ATGTTG": 476, + "ATGTCA": 477, + "ATGTCT": 478, + "ATGTCC": 479, + "ATGTCG": 480, + "ATGTGA": 481, + "ATGTGT": 482, + "ATGTGC": 483, + "ATGTGG": 484, + "ATGCAA": 485, + "ATGCAT": 486, + "ATGCAC": 487, + "ATGCAG": 488, + "ATGCTA": 489, + "ATGCTT": 490, + "ATGCTC": 491, + "ATGCTG": 492, + "ATGCCA": 493, + "ATGCCT": 494, + "ATGCCC": 495, + "ATGCCG": 496, + "ATGCGA": 497, + "ATGCGT": 498, + "ATGCGC": 499, + "ATGCGG": 500, + "ATGGAA": 501, + "ATGGAT": 502, + "ATGGAC": 503, + "ATGGAG": 504, + "ATGGTA": 505, + "ATGGTT": 506, + "ATGGTC": 507, + "ATGGTG": 508, + "ATGGCA": 509, + "ATGGCT": 510, + "ATGGCC": 511, + "ATGGCG": 512, + "ATGGGA": 513, + "ATGGGT": 514, + "ATGGGC": 515, + "ATGGGG": 516, + "ACAAAA": 517, + "ACAAAT": 518, + "ACAAAC": 519, + "ACAAAG": 520, + "ACAATA": 521, + "ACAATT": 522, + "ACAATC": 523, + "ACAATG": 524, + "ACAACA": 525, + "ACAACT": 526, + "ACAACC": 527, + "ACAACG": 528, + "ACAAGA": 529, + "ACAAGT": 530, + "ACAAGC": 531, + "ACAAGG": 532, + "ACATAA": 533, + "ACATAT": 534, + "ACATAC": 535, + "ACATAG": 536, + "ACATTA": 537, + "ACATTT": 538, + "ACATTC": 539, + "ACATTG": 540, + "ACATCA": 541, + "ACATCT": 542, + "ACATCC": 543, + "ACATCG": 544, + "ACATGA": 545, + "ACATGT": 546, + "ACATGC": 547, + "ACATGG": 548, + "ACACAA": 549, + "ACACAT": 550, + "ACACAC": 551, + "ACACAG": 552, + "ACACTA": 553, + "ACACTT": 554, + "ACACTC": 555, + "ACACTG": 556, + "ACACCA": 557, + "ACACCT": 558, + "ACACCC": 559, + "ACACCG": 560, + "ACACGA": 561, + "ACACGT": 562, + "ACACGC": 563, + "ACACGG": 564, + "ACAGAA": 565, + "ACAGAT": 566, + "ACAGAC": 567, + "ACAGAG": 568, + "ACAGTA": 569, + "ACAGTT": 570, + "ACAGTC": 571, + "ACAGTG": 572, + "ACAGCA": 573, + "ACAGCT": 574, + "ACAGCC": 575, + "ACAGCG": 576, + "ACAGGA": 577, + "ACAGGT": 578, + "ACAGGC": 579, + "ACAGGG": 580, + "ACTAAA": 581, + "ACTAAT": 582, + "ACTAAC": 583, + "ACTAAG": 584, + "ACTATA": 585, + "ACTATT": 586, + "ACTATC": 587, + "ACTATG": 588, + "ACTACA": 589, + "ACTACT": 590, + "ACTACC": 591, + "ACTACG": 592, + "ACTAGA": 593, + "ACTAGT": 594, + "ACTAGC": 595, + "ACTAGG": 596, + "ACTTAA": 597, + "ACTTAT": 598, + "ACTTAC": 599, + "ACTTAG": 600, + "ACTTTA": 601, + "ACTTTT": 602, + "ACTTTC": 603, + "ACTTTG": 604, + "ACTTCA": 605, + "ACTTCT": 606, + "ACTTCC": 607, + "ACTTCG": 608, + "ACTTGA": 609, + "ACTTGT": 610, + "ACTTGC": 611, + "ACTTGG": 612, + "ACTCAA": 613, + "ACTCAT": 614, + "ACTCAC": 615, + "ACTCAG": 616, + "ACTCTA": 617, + "ACTCTT": 618, + "ACTCTC": 619, + "ACTCTG": 620, + "ACTCCA": 621, + "ACTCCT": 622, + "ACTCCC": 623, + "ACTCCG": 624, + "ACTCGA": 625, + "ACTCGT": 626, + "ACTCGC": 627, + "ACTCGG": 628, + "ACTGAA": 629, + "ACTGAT": 630, + "ACTGAC": 631, + "ACTGAG": 632, + "ACTGTA": 633, + "ACTGTT": 634, + "ACTGTC": 635, + "ACTGTG": 636, + "ACTGCA": 637, + "ACTGCT": 638, + "ACTGCC": 639, + "ACTGCG": 640, + "ACTGGA": 641, + "ACTGGT": 642, + "ACTGGC": 643, + "ACTGGG": 644, + "ACCAAA": 645, + "ACCAAT": 646, + "ACCAAC": 647, + "ACCAAG": 648, + "ACCATA": 649, + "ACCATT": 650, + "ACCATC": 651, + "ACCATG": 652, + "ACCACA": 653, + "ACCACT": 654, + "ACCACC": 655, + "ACCACG": 656, + "ACCAGA": 657, + "ACCAGT": 658, + "ACCAGC": 659, + "ACCAGG": 660, + "ACCTAA": 661, + "ACCTAT": 662, + "ACCTAC": 663, + "ACCTAG": 664, + "ACCTTA": 665, + "ACCTTT": 666, + "ACCTTC": 667, + "ACCTTG": 668, + "ACCTCA": 669, + "ACCTCT": 670, + "ACCTCC": 671, + "ACCTCG": 672, + "ACCTGA": 673, + "ACCTGT": 674, + "ACCTGC": 675, + "ACCTGG": 676, + "ACCCAA": 677, + "ACCCAT": 678, + "ACCCAC": 679, + "ACCCAG": 680, + "ACCCTA": 681, + "ACCCTT": 682, + "ACCCTC": 683, + "ACCCTG": 684, + "ACCCCA": 685, + "ACCCCT": 686, + "ACCCCC": 687, + "ACCCCG": 688, + "ACCCGA": 689, + "ACCCGT": 690, + "ACCCGC": 691, + "ACCCGG": 692, + "ACCGAA": 693, + "ACCGAT": 694, + "ACCGAC": 695, + "ACCGAG": 696, + "ACCGTA": 697, + "ACCGTT": 698, + "ACCGTC": 699, + "ACCGTG": 700, + "ACCGCA": 701, + "ACCGCT": 702, + "ACCGCC": 703, + "ACCGCG": 704, + "ACCGGA": 705, + "ACCGGT": 706, + "ACCGGC": 707, + "ACCGGG": 708, + "ACGAAA": 709, + "ACGAAT": 710, + "ACGAAC": 711, + "ACGAAG": 712, + "ACGATA": 713, + "ACGATT": 714, + "ACGATC": 715, + "ACGATG": 716, + "ACGACA": 717, + "ACGACT": 718, + "ACGACC": 719, + "ACGACG": 720, + "ACGAGA": 721, + "ACGAGT": 722, + "ACGAGC": 723, + "ACGAGG": 724, + "ACGTAA": 725, + "ACGTAT": 726, + "ACGTAC": 727, + "ACGTAG": 728, + "ACGTTA": 729, + "ACGTTT": 730, + "ACGTTC": 731, + "ACGTTG": 732, + "ACGTCA": 733, + "ACGTCT": 734, + "ACGTCC": 735, + "ACGTCG": 736, + "ACGTGA": 737, + "ACGTGT": 738, + "ACGTGC": 739, + "ACGTGG": 740, + "ACGCAA": 741, + "ACGCAT": 742, + "ACGCAC": 743, + "ACGCAG": 744, + "ACGCTA": 745, + "ACGCTT": 746, + "ACGCTC": 747, + "ACGCTG": 748, + "ACGCCA": 749, + "ACGCCT": 750, + "ACGCCC": 751, + "ACGCCG": 752, + "ACGCGA": 753, + "ACGCGT": 754, + "ACGCGC": 755, + "ACGCGG": 756, + "ACGGAA": 757, + "ACGGAT": 758, + "ACGGAC": 759, + "ACGGAG": 760, + "ACGGTA": 761, + "ACGGTT": 762, + "ACGGTC": 763, + "ACGGTG": 764, + "ACGGCA": 765, + "ACGGCT": 766, + "ACGGCC": 767, + "ACGGCG": 768, + "ACGGGA": 769, + "ACGGGT": 770, + "ACGGGC": 771, + "ACGGGG": 772, + "AGAAAA": 773, + "AGAAAT": 774, + "AGAAAC": 775, + "AGAAAG": 776, + "AGAATA": 777, + "AGAATT": 778, + "AGAATC": 779, + "AGAATG": 780, + "AGAACA": 781, + "AGAACT": 782, + "AGAACC": 783, + "AGAACG": 784, + "AGAAGA": 785, + "AGAAGT": 786, + "AGAAGC": 787, + "AGAAGG": 788, + "AGATAA": 789, + "AGATAT": 790, + "AGATAC": 791, + "AGATAG": 792, + "AGATTA": 793, + "AGATTT": 794, + "AGATTC": 795, + "AGATTG": 796, + "AGATCA": 797, + "AGATCT": 798, + "AGATCC": 799, + "AGATCG": 800, + "AGATGA": 801, + "AGATGT": 802, + "AGATGC": 803, + "AGATGG": 804, + "AGACAA": 805, + "AGACAT": 806, + "AGACAC": 807, + "AGACAG": 808, + "AGACTA": 809, + "AGACTT": 810, + "AGACTC": 811, + "AGACTG": 812, + "AGACCA": 813, + "AGACCT": 814, + "AGACCC": 815, + "AGACCG": 816, + "AGACGA": 817, + "AGACGT": 818, + "AGACGC": 819, + "AGACGG": 820, + "AGAGAA": 821, + "AGAGAT": 822, + "AGAGAC": 823, + "AGAGAG": 824, + "AGAGTA": 825, + "AGAGTT": 826, + "AGAGTC": 827, + "AGAGTG": 828, + "AGAGCA": 829, + "AGAGCT": 830, + "AGAGCC": 831, + "AGAGCG": 832, + "AGAGGA": 833, + "AGAGGT": 834, + "AGAGGC": 835, + "AGAGGG": 836, + "AGTAAA": 837, + "AGTAAT": 838, + "AGTAAC": 839, + "AGTAAG": 840, + "AGTATA": 841, + "AGTATT": 842, + "AGTATC": 843, + "AGTATG": 844, + "AGTACA": 845, + "AGTACT": 846, + "AGTACC": 847, + "AGTACG": 848, + "AGTAGA": 849, + "AGTAGT": 850, + "AGTAGC": 851, + "AGTAGG": 852, + "AGTTAA": 853, + "AGTTAT": 854, + "AGTTAC": 855, + "AGTTAG": 856, + "AGTTTA": 857, + "AGTTTT": 858, + "AGTTTC": 859, + "AGTTTG": 860, + "AGTTCA": 861, + "AGTTCT": 862, + "AGTTCC": 863, + "AGTTCG": 864, + "AGTTGA": 865, + "AGTTGT": 866, + "AGTTGC": 867, + "AGTTGG": 868, + "AGTCAA": 869, + "AGTCAT": 870, + "AGTCAC": 871, + "AGTCAG": 872, + "AGTCTA": 873, + "AGTCTT": 874, + "AGTCTC": 875, + "AGTCTG": 876, + "AGTCCA": 877, + "AGTCCT": 878, + "AGTCCC": 879, + "AGTCCG": 880, + "AGTCGA": 881, + "AGTCGT": 882, + "AGTCGC": 883, + "AGTCGG": 884, + "AGTGAA": 885, + "AGTGAT": 886, + "AGTGAC": 887, + "AGTGAG": 888, + "AGTGTA": 889, + "AGTGTT": 890, + "AGTGTC": 891, + "AGTGTG": 892, + "AGTGCA": 893, + "AGTGCT": 894, + "AGTGCC": 895, + "AGTGCG": 896, + "AGTGGA": 897, + "AGTGGT": 898, + "AGTGGC": 899, + "AGTGGG": 900, + "AGCAAA": 901, + "AGCAAT": 902, + "AGCAAC": 903, + "AGCAAG": 904, + "AGCATA": 905, + "AGCATT": 906, + "AGCATC": 907, + "AGCATG": 908, + "AGCACA": 909, + "AGCACT": 910, + "AGCACC": 911, + "AGCACG": 912, + "AGCAGA": 913, + "AGCAGT": 914, + "AGCAGC": 915, + "AGCAGG": 916, + "AGCTAA": 917, + "AGCTAT": 918, + "AGCTAC": 919, + "AGCTAG": 920, + "AGCTTA": 921, + "AGCTTT": 922, + "AGCTTC": 923, + "AGCTTG": 924, + "AGCTCA": 925, + "AGCTCT": 926, + "AGCTCC": 927, + "AGCTCG": 928, + "AGCTGA": 929, + "AGCTGT": 930, + "AGCTGC": 931, + "AGCTGG": 932, + "AGCCAA": 933, + "AGCCAT": 934, + "AGCCAC": 935, + "AGCCAG": 936, + "AGCCTA": 937, + "AGCCTT": 938, + "AGCCTC": 939, + "AGCCTG": 940, + "AGCCCA": 941, + "AGCCCT": 942, + "AGCCCC": 943, + "AGCCCG": 944, + "AGCCGA": 945, + "AGCCGT": 946, + "AGCCGC": 947, + "AGCCGG": 948, + "AGCGAA": 949, + "AGCGAT": 950, + "AGCGAC": 951, + "AGCGAG": 952, + "AGCGTA": 953, + "AGCGTT": 954, + "AGCGTC": 955, + "AGCGTG": 956, + "AGCGCA": 957, + "AGCGCT": 958, + "AGCGCC": 959, + "AGCGCG": 960, + "AGCGGA": 961, + "AGCGGT": 962, + "AGCGGC": 963, + "AGCGGG": 964, + "AGGAAA": 965, + "AGGAAT": 966, + "AGGAAC": 967, + "AGGAAG": 968, + "AGGATA": 969, + "AGGATT": 970, + "AGGATC": 971, + "AGGATG": 972, + "AGGACA": 973, + "AGGACT": 974, + "AGGACC": 975, + "AGGACG": 976, + "AGGAGA": 977, + "AGGAGT": 978, + "AGGAGC": 979, + "AGGAGG": 980, + "AGGTAA": 981, + "AGGTAT": 982, + "AGGTAC": 983, + "AGGTAG": 984, + "AGGTTA": 985, + "AGGTTT": 986, + "AGGTTC": 987, + "AGGTTG": 988, + "AGGTCA": 989, + "AGGTCT": 990, + "AGGTCC": 991, + "AGGTCG": 992, + "AGGTGA": 993, + "AGGTGT": 994, + "AGGTGC": 995, + "AGGTGG": 996, + "AGGCAA": 997, + "AGGCAT": 998, + "AGGCAC": 999, + "AGGCAG": 1000, + "AGGCTA": 1001, + "AGGCTT": 1002, + "AGGCTC": 1003, + "AGGCTG": 1004, + "AGGCCA": 1005, + "AGGCCT": 1006, + "AGGCCC": 1007, + "AGGCCG": 1008, + "AGGCGA": 1009, + "AGGCGT": 1010, + "AGGCGC": 1011, + "AGGCGG": 1012, + "AGGGAA": 1013, + "AGGGAT": 1014, + "AGGGAC": 1015, + "AGGGAG": 1016, + "AGGGTA": 1017, + "AGGGTT": 1018, + "AGGGTC": 1019, + "AGGGTG": 1020, + "AGGGCA": 1021, + "AGGGCT": 1022, + "AGGGCC": 1023, + "AGGGCG": 1024, + "AGGGGA": 1025, + "AGGGGT": 1026, + "AGGGGC": 1027, + "AGGGGG": 1028, + "TAAAAA": 1029, + "TAAAAT": 1030, + "TAAAAC": 1031, + "TAAAAG": 1032, + "TAAATA": 1033, + "TAAATT": 1034, + "TAAATC": 1035, + "TAAATG": 1036, + "TAAACA": 1037, + "TAAACT": 1038, + "TAAACC": 1039, + "TAAACG": 1040, + "TAAAGA": 1041, + "TAAAGT": 1042, + "TAAAGC": 1043, + "TAAAGG": 1044, + "TAATAA": 1045, + "TAATAT": 1046, + "TAATAC": 1047, + "TAATAG": 1048, + "TAATTA": 1049, + "TAATTT": 1050, + "TAATTC": 1051, + "TAATTG": 1052, + "TAATCA": 1053, + "TAATCT": 1054, + "TAATCC": 1055, + "TAATCG": 1056, + "TAATGA": 1057, + "TAATGT": 1058, + "TAATGC": 1059, + "TAATGG": 1060, + "TAACAA": 1061, + "TAACAT": 1062, + "TAACAC": 1063, + "TAACAG": 1064, + "TAACTA": 1065, + "TAACTT": 1066, + "TAACTC": 1067, + "TAACTG": 1068, + "TAACCA": 1069, + "TAACCT": 1070, + "TAACCC": 1071, + "TAACCG": 1072, + "TAACGA": 1073, + "TAACGT": 1074, + "TAACGC": 1075, + "TAACGG": 1076, + "TAAGAA": 1077, + "TAAGAT": 1078, + "TAAGAC": 1079, + "TAAGAG": 1080, + "TAAGTA": 1081, + "TAAGTT": 1082, + "TAAGTC": 1083, + "TAAGTG": 1084, + "TAAGCA": 1085, + "TAAGCT": 1086, + "TAAGCC": 1087, + "TAAGCG": 1088, + "TAAGGA": 1089, + "TAAGGT": 1090, + "TAAGGC": 1091, + "TAAGGG": 1092, + "TATAAA": 1093, + "TATAAT": 1094, + "TATAAC": 1095, + "TATAAG": 1096, + "TATATA": 1097, + "TATATT": 1098, + "TATATC": 1099, + "TATATG": 1100, + "TATACA": 1101, + "TATACT": 1102, + "TATACC": 1103, + "TATACG": 1104, + "TATAGA": 1105, + "TATAGT": 1106, + "TATAGC": 1107, + "TATAGG": 1108, + "TATTAA": 1109, + "TATTAT": 1110, + "TATTAC": 1111, + "TATTAG": 1112, + "TATTTA": 1113, + "TATTTT": 1114, + "TATTTC": 1115, + "TATTTG": 1116, + "TATTCA": 1117, + "TATTCT": 1118, + "TATTCC": 1119, + "TATTCG": 1120, + "TATTGA": 1121, + "TATTGT": 1122, + "TATTGC": 1123, + "TATTGG": 1124, + "TATCAA": 1125, + "TATCAT": 1126, + "TATCAC": 1127, + "TATCAG": 1128, + "TATCTA": 1129, + "TATCTT": 1130, + "TATCTC": 1131, + "TATCTG": 1132, + "TATCCA": 1133, + "TATCCT": 1134, + "TATCCC": 1135, + "TATCCG": 1136, + "TATCGA": 1137, + "TATCGT": 1138, + "TATCGC": 1139, + "TATCGG": 1140, + "TATGAA": 1141, + "TATGAT": 1142, + "TATGAC": 1143, + "TATGAG": 1144, + "TATGTA": 1145, + "TATGTT": 1146, + "TATGTC": 1147, + "TATGTG": 1148, + "TATGCA": 1149, + "TATGCT": 1150, + "TATGCC": 1151, + "TATGCG": 1152, + "TATGGA": 1153, + "TATGGT": 1154, + "TATGGC": 1155, + "TATGGG": 1156, + "TACAAA": 1157, + "TACAAT": 1158, + "TACAAC": 1159, + "TACAAG": 1160, + "TACATA": 1161, + "TACATT": 1162, + "TACATC": 1163, + "TACATG": 1164, + "TACACA": 1165, + "TACACT": 1166, + "TACACC": 1167, + "TACACG": 1168, + "TACAGA": 1169, + "TACAGT": 1170, + "TACAGC": 1171, + "TACAGG": 1172, + "TACTAA": 1173, + "TACTAT": 1174, + "TACTAC": 1175, + "TACTAG": 1176, + "TACTTA": 1177, + "TACTTT": 1178, + "TACTTC": 1179, + "TACTTG": 1180, + "TACTCA": 1181, + "TACTCT": 1182, + "TACTCC": 1183, + "TACTCG": 1184, + "TACTGA": 1185, + "TACTGT": 1186, + "TACTGC": 1187, + "TACTGG": 1188, + "TACCAA": 1189, + "TACCAT": 1190, + "TACCAC": 1191, + "TACCAG": 1192, + "TACCTA": 1193, + "TACCTT": 1194, + "TACCTC": 1195, + "TACCTG": 1196, + "TACCCA": 1197, + "TACCCT": 1198, + "TACCCC": 1199, + "TACCCG": 1200, + "TACCGA": 1201, + "TACCGT": 1202, + "TACCGC": 1203, + "TACCGG": 1204, + "TACGAA": 1205, + "TACGAT": 1206, + "TACGAC": 1207, + "TACGAG": 1208, + "TACGTA": 1209, + "TACGTT": 1210, + "TACGTC": 1211, + "TACGTG": 1212, + "TACGCA": 1213, + "TACGCT": 1214, + "TACGCC": 1215, + "TACGCG": 1216, + "TACGGA": 1217, + "TACGGT": 1218, + "TACGGC": 1219, + "TACGGG": 1220, + "TAGAAA": 1221, + "TAGAAT": 1222, + "TAGAAC": 1223, + "TAGAAG": 1224, + "TAGATA": 1225, + "TAGATT": 1226, + "TAGATC": 1227, + "TAGATG": 1228, + "TAGACA": 1229, + "TAGACT": 1230, + "TAGACC": 1231, + "TAGACG": 1232, + "TAGAGA": 1233, + "TAGAGT": 1234, + "TAGAGC": 1235, + "TAGAGG": 1236, + "TAGTAA": 1237, + "TAGTAT": 1238, + "TAGTAC": 1239, + "TAGTAG": 1240, + "TAGTTA": 1241, + "TAGTTT": 1242, + "TAGTTC": 1243, + "TAGTTG": 1244, + "TAGTCA": 1245, + "TAGTCT": 1246, + "TAGTCC": 1247, + "TAGTCG": 1248, + "TAGTGA": 1249, + "TAGTGT": 1250, + "TAGTGC": 1251, + "TAGTGG": 1252, + "TAGCAA": 1253, + "TAGCAT": 1254, + "TAGCAC": 1255, + "TAGCAG": 1256, + "TAGCTA": 1257, + "TAGCTT": 1258, + "TAGCTC": 1259, + "TAGCTG": 1260, + "TAGCCA": 1261, + "TAGCCT": 1262, + "TAGCCC": 1263, + "TAGCCG": 1264, + "TAGCGA": 1265, + "TAGCGT": 1266, + "TAGCGC": 1267, + "TAGCGG": 1268, + "TAGGAA": 1269, + "TAGGAT": 1270, + "TAGGAC": 1271, + "TAGGAG": 1272, + "TAGGTA": 1273, + "TAGGTT": 1274, + "TAGGTC": 1275, + "TAGGTG": 1276, + "TAGGCA": 1277, + "TAGGCT": 1278, + "TAGGCC": 1279, + "TAGGCG": 1280, + "TAGGGA": 1281, + "TAGGGT": 1282, + "TAGGGC": 1283, + "TAGGGG": 1284, + "TTAAAA": 1285, + "TTAAAT": 1286, + "TTAAAC": 1287, + "TTAAAG": 1288, + "TTAATA": 1289, + "TTAATT": 1290, + "TTAATC": 1291, + "TTAATG": 1292, + "TTAACA": 1293, + "TTAACT": 1294, + "TTAACC": 1295, + "TTAACG": 1296, + "TTAAGA": 1297, + "TTAAGT": 1298, + "TTAAGC": 1299, + "TTAAGG": 1300, + "TTATAA": 1301, + "TTATAT": 1302, + "TTATAC": 1303, + "TTATAG": 1304, + "TTATTA": 1305, + "TTATTT": 1306, + "TTATTC": 1307, + "TTATTG": 1308, + "TTATCA": 1309, + "TTATCT": 1310, + "TTATCC": 1311, + "TTATCG": 1312, + "TTATGA": 1313, + "TTATGT": 1314, + "TTATGC": 1315, + "TTATGG": 1316, + "TTACAA": 1317, + "TTACAT": 1318, + "TTACAC": 1319, + "TTACAG": 1320, + "TTACTA": 1321, + "TTACTT": 1322, + "TTACTC": 1323, + "TTACTG": 1324, + "TTACCA": 1325, + "TTACCT": 1326, + "TTACCC": 1327, + "TTACCG": 1328, + "TTACGA": 1329, + "TTACGT": 1330, + "TTACGC": 1331, + "TTACGG": 1332, + "TTAGAA": 1333, + "TTAGAT": 1334, + "TTAGAC": 1335, + "TTAGAG": 1336, + "TTAGTA": 1337, + "TTAGTT": 1338, + "TTAGTC": 1339, + "TTAGTG": 1340, + "TTAGCA": 1341, + "TTAGCT": 1342, + "TTAGCC": 1343, + "TTAGCG": 1344, + "TTAGGA": 1345, + "TTAGGT": 1346, + "TTAGGC": 1347, + "TTAGGG": 1348, + "TTTAAA": 1349, + "TTTAAT": 1350, + "TTTAAC": 1351, + "TTTAAG": 1352, + "TTTATA": 1353, + "TTTATT": 1354, + "TTTATC": 1355, + "TTTATG": 1356, + "TTTACA": 1357, + "TTTACT": 1358, + "TTTACC": 1359, + "TTTACG": 1360, + "TTTAGA": 1361, + "TTTAGT": 1362, + "TTTAGC": 1363, + "TTTAGG": 1364, + "TTTTAA": 1365, + "TTTTAT": 1366, + "TTTTAC": 1367, + "TTTTAG": 1368, + "TTTTTA": 1369, + "TTTTTT": 1370, + "TTTTTC": 1371, + "TTTTTG": 1372, + "TTTTCA": 1373, + "TTTTCT": 1374, + "TTTTCC": 1375, + "TTTTCG": 1376, + "TTTTGA": 1377, + "TTTTGT": 1378, + "TTTTGC": 1379, + "TTTTGG": 1380, + "TTTCAA": 1381, + "TTTCAT": 1382, + "TTTCAC": 1383, + "TTTCAG": 1384, + "TTTCTA": 1385, + "TTTCTT": 1386, + "TTTCTC": 1387, + "TTTCTG": 1388, + "TTTCCA": 1389, + "TTTCCT": 1390, + "TTTCCC": 1391, + "TTTCCG": 1392, + "TTTCGA": 1393, + "TTTCGT": 1394, + "TTTCGC": 1395, + "TTTCGG": 1396, + "TTTGAA": 1397, + "TTTGAT": 1398, + "TTTGAC": 1399, + "TTTGAG": 1400, + "TTTGTA": 1401, + "TTTGTT": 1402, + "TTTGTC": 1403, + "TTTGTG": 1404, + "TTTGCA": 1405, + "TTTGCT": 1406, + "TTTGCC": 1407, + "TTTGCG": 1408, + "TTTGGA": 1409, + "TTTGGT": 1410, + "TTTGGC": 1411, + "TTTGGG": 1412, + "TTCAAA": 1413, + "TTCAAT": 1414, + "TTCAAC": 1415, + "TTCAAG": 1416, + "TTCATA": 1417, + "TTCATT": 1418, + "TTCATC": 1419, + "TTCATG": 1420, + "TTCACA": 1421, + "TTCACT": 1422, + "TTCACC": 1423, + "TTCACG": 1424, + "TTCAGA": 1425, + "TTCAGT": 1426, + "TTCAGC": 1427, + "TTCAGG": 1428, + "TTCTAA": 1429, + "TTCTAT": 1430, + "TTCTAC": 1431, + "TTCTAG": 1432, + "TTCTTA": 1433, + "TTCTTT": 1434, + "TTCTTC": 1435, + "TTCTTG": 1436, + "TTCTCA": 1437, + "TTCTCT": 1438, + "TTCTCC": 1439, + "TTCTCG": 1440, + "TTCTGA": 1441, + "TTCTGT": 1442, + "TTCTGC": 1443, + "TTCTGG": 1444, + "TTCCAA": 1445, + "TTCCAT": 1446, + "TTCCAC": 1447, + "TTCCAG": 1448, + "TTCCTA": 1449, + "TTCCTT": 1450, + "TTCCTC": 1451, + "TTCCTG": 1452, + "TTCCCA": 1453, + "TTCCCT": 1454, + "TTCCCC": 1455, + "TTCCCG": 1456, + "TTCCGA": 1457, + "TTCCGT": 1458, + "TTCCGC": 1459, + "TTCCGG": 1460, + "TTCGAA": 1461, + "TTCGAT": 1462, + "TTCGAC": 1463, + "TTCGAG": 1464, + "TTCGTA": 1465, + "TTCGTT": 1466, + "TTCGTC": 1467, + "TTCGTG": 1468, + "TTCGCA": 1469, + "TTCGCT": 1470, + "TTCGCC": 1471, + "TTCGCG": 1472, + "TTCGGA": 1473, + "TTCGGT": 1474, + "TTCGGC": 1475, + "TTCGGG": 1476, + "TTGAAA": 1477, + "TTGAAT": 1478, + "TTGAAC": 1479, + "TTGAAG": 1480, + "TTGATA": 1481, + "TTGATT": 1482, + "TTGATC": 1483, + "TTGATG": 1484, + "TTGACA": 1485, + "TTGACT": 1486, + "TTGACC": 1487, + "TTGACG": 1488, + "TTGAGA": 1489, + "TTGAGT": 1490, + "TTGAGC": 1491, + "TTGAGG": 1492, + "TTGTAA": 1493, + "TTGTAT": 1494, + "TTGTAC": 1495, + "TTGTAG": 1496, + "TTGTTA": 1497, + "TTGTTT": 1498, + "TTGTTC": 1499, + "TTGTTG": 1500, + "TTGTCA": 1501, + "TTGTCT": 1502, + "TTGTCC": 1503, + "TTGTCG": 1504, + "TTGTGA": 1505, + "TTGTGT": 1506, + "TTGTGC": 1507, + "TTGTGG": 1508, + "TTGCAA": 1509, + "TTGCAT": 1510, + "TTGCAC": 1511, + "TTGCAG": 1512, + "TTGCTA": 1513, + "TTGCTT": 1514, + "TTGCTC": 1515, + "TTGCTG": 1516, + "TTGCCA": 1517, + "TTGCCT": 1518, + "TTGCCC": 1519, + "TTGCCG": 1520, + "TTGCGA": 1521, + "TTGCGT": 1522, + "TTGCGC": 1523, + "TTGCGG": 1524, + "TTGGAA": 1525, + "TTGGAT": 1526, + "TTGGAC": 1527, + "TTGGAG": 1528, + "TTGGTA": 1529, + "TTGGTT": 1530, + "TTGGTC": 1531, + "TTGGTG": 1532, + "TTGGCA": 1533, + "TTGGCT": 1534, + "TTGGCC": 1535, + "TTGGCG": 1536, + "TTGGGA": 1537, + "TTGGGT": 1538, + "TTGGGC": 1539, + "TTGGGG": 1540, + "TCAAAA": 1541, + "TCAAAT": 1542, + "TCAAAC": 1543, + "TCAAAG": 1544, + "TCAATA": 1545, + "TCAATT": 1546, + "TCAATC": 1547, + "TCAATG": 1548, + "TCAACA": 1549, + "TCAACT": 1550, + "TCAACC": 1551, + "TCAACG": 1552, + "TCAAGA": 1553, + "TCAAGT": 1554, + "TCAAGC": 1555, + "TCAAGG": 1556, + "TCATAA": 1557, + "TCATAT": 1558, + "TCATAC": 1559, + "TCATAG": 1560, + "TCATTA": 1561, + "TCATTT": 1562, + "TCATTC": 1563, + "TCATTG": 1564, + "TCATCA": 1565, + "TCATCT": 1566, + "TCATCC": 1567, + "TCATCG": 1568, + "TCATGA": 1569, + "TCATGT": 1570, + "TCATGC": 1571, + "TCATGG": 1572, + "TCACAA": 1573, + "TCACAT": 1574, + "TCACAC": 1575, + "TCACAG": 1576, + "TCACTA": 1577, + "TCACTT": 1578, + "TCACTC": 1579, + "TCACTG": 1580, + "TCACCA": 1581, + "TCACCT": 1582, + "TCACCC": 1583, + "TCACCG": 1584, + "TCACGA": 1585, + "TCACGT": 1586, + "TCACGC": 1587, + "TCACGG": 1588, + "TCAGAA": 1589, + "TCAGAT": 1590, + "TCAGAC": 1591, + "TCAGAG": 1592, + "TCAGTA": 1593, + "TCAGTT": 1594, + "TCAGTC": 1595, + "TCAGTG": 1596, + "TCAGCA": 1597, + "TCAGCT": 1598, + "TCAGCC": 1599, + "TCAGCG": 1600, + "TCAGGA": 1601, + "TCAGGT": 1602, + "TCAGGC": 1603, + "TCAGGG": 1604, + "TCTAAA": 1605, + "TCTAAT": 1606, + "TCTAAC": 1607, + "TCTAAG": 1608, + "TCTATA": 1609, + "TCTATT": 1610, + "TCTATC": 1611, + "TCTATG": 1612, + "TCTACA": 1613, + "TCTACT": 1614, + "TCTACC": 1615, + "TCTACG": 1616, + "TCTAGA": 1617, + "TCTAGT": 1618, + "TCTAGC": 1619, + "TCTAGG": 1620, + "TCTTAA": 1621, + "TCTTAT": 1622, + "TCTTAC": 1623, + "TCTTAG": 1624, + "TCTTTA": 1625, + "TCTTTT": 1626, + "TCTTTC": 1627, + "TCTTTG": 1628, + "TCTTCA": 1629, + "TCTTCT": 1630, + "TCTTCC": 1631, + "TCTTCG": 1632, + "TCTTGA": 1633, + "TCTTGT": 1634, + "TCTTGC": 1635, + "TCTTGG": 1636, + "TCTCAA": 1637, + "TCTCAT": 1638, + "TCTCAC": 1639, + "TCTCAG": 1640, + "TCTCTA": 1641, + "TCTCTT": 1642, + "TCTCTC": 1643, + "TCTCTG": 1644, + "TCTCCA": 1645, + "TCTCCT": 1646, + "TCTCCC": 1647, + "TCTCCG": 1648, + "TCTCGA": 1649, + "TCTCGT": 1650, + "TCTCGC": 1651, + "TCTCGG": 1652, + "TCTGAA": 1653, + "TCTGAT": 1654, + "TCTGAC": 1655, + "TCTGAG": 1656, + "TCTGTA": 1657, + "TCTGTT": 1658, + "TCTGTC": 1659, + "TCTGTG": 1660, + "TCTGCA": 1661, + "TCTGCT": 1662, + "TCTGCC": 1663, + "TCTGCG": 1664, + "TCTGGA": 1665, + "TCTGGT": 1666, + "TCTGGC": 1667, + "TCTGGG": 1668, + "TCCAAA": 1669, + "TCCAAT": 1670, + "TCCAAC": 1671, + "TCCAAG": 1672, + "TCCATA": 1673, + "TCCATT": 1674, + "TCCATC": 1675, + "TCCATG": 1676, + "TCCACA": 1677, + "TCCACT": 1678, + "TCCACC": 1679, + "TCCACG": 1680, + "TCCAGA": 1681, + "TCCAGT": 1682, + "TCCAGC": 1683, + "TCCAGG": 1684, + "TCCTAA": 1685, + "TCCTAT": 1686, + "TCCTAC": 1687, + "TCCTAG": 1688, + "TCCTTA": 1689, + "TCCTTT": 1690, + "TCCTTC": 1691, + "TCCTTG": 1692, + "TCCTCA": 1693, + "TCCTCT": 1694, + "TCCTCC": 1695, + "TCCTCG": 1696, + "TCCTGA": 1697, + "TCCTGT": 1698, + "TCCTGC": 1699, + "TCCTGG": 1700, + "TCCCAA": 1701, + "TCCCAT": 1702, + "TCCCAC": 1703, + "TCCCAG": 1704, + "TCCCTA": 1705, + "TCCCTT": 1706, + "TCCCTC": 1707, + "TCCCTG": 1708, + "TCCCCA": 1709, + "TCCCCT": 1710, + "TCCCCC": 1711, + "TCCCCG": 1712, + "TCCCGA": 1713, + "TCCCGT": 1714, + "TCCCGC": 1715, + "TCCCGG": 1716, + "TCCGAA": 1717, + "TCCGAT": 1718, + "TCCGAC": 1719, + "TCCGAG": 1720, + "TCCGTA": 1721, + "TCCGTT": 1722, + "TCCGTC": 1723, + "TCCGTG": 1724, + "TCCGCA": 1725, + "TCCGCT": 1726, + "TCCGCC": 1727, + "TCCGCG": 1728, + "TCCGGA": 1729, + "TCCGGT": 1730, + "TCCGGC": 1731, + "TCCGGG": 1732, + "TCGAAA": 1733, + "TCGAAT": 1734, + "TCGAAC": 1735, + "TCGAAG": 1736, + "TCGATA": 1737, + "TCGATT": 1738, + "TCGATC": 1739, + "TCGATG": 1740, + "TCGACA": 1741, + "TCGACT": 1742, + "TCGACC": 1743, + "TCGACG": 1744, + "TCGAGA": 1745, + "TCGAGT": 1746, + "TCGAGC": 1747, + "TCGAGG": 1748, + "TCGTAA": 1749, + "TCGTAT": 1750, + "TCGTAC": 1751, + "TCGTAG": 1752, + "TCGTTA": 1753, + "TCGTTT": 1754, + "TCGTTC": 1755, + "TCGTTG": 1756, + "TCGTCA": 1757, + "TCGTCT": 1758, + "TCGTCC": 1759, + "TCGTCG": 1760, + "TCGTGA": 1761, + "TCGTGT": 1762, + "TCGTGC": 1763, + "TCGTGG": 1764, + "TCGCAA": 1765, + "TCGCAT": 1766, + "TCGCAC": 1767, + "TCGCAG": 1768, + "TCGCTA": 1769, + "TCGCTT": 1770, + "TCGCTC": 1771, + "TCGCTG": 1772, + "TCGCCA": 1773, + "TCGCCT": 1774, + "TCGCCC": 1775, + "TCGCCG": 1776, + "TCGCGA": 1777, + "TCGCGT": 1778, + "TCGCGC": 1779, + "TCGCGG": 1780, + "TCGGAA": 1781, + "TCGGAT": 1782, + "TCGGAC": 1783, + "TCGGAG": 1784, + "TCGGTA": 1785, + "TCGGTT": 1786, + "TCGGTC": 1787, + "TCGGTG": 1788, + "TCGGCA": 1789, + "TCGGCT": 1790, + "TCGGCC": 1791, + "TCGGCG": 1792, + "TCGGGA": 1793, + "TCGGGT": 1794, + "TCGGGC": 1795, + "TCGGGG": 1796, + "TGAAAA": 1797, + "TGAAAT": 1798, + "TGAAAC": 1799, + "TGAAAG": 1800, + "TGAATA": 1801, + "TGAATT": 1802, + "TGAATC": 1803, + "TGAATG": 1804, + "TGAACA": 1805, + "TGAACT": 1806, + "TGAACC": 1807, + "TGAACG": 1808, + "TGAAGA": 1809, + "TGAAGT": 1810, + "TGAAGC": 1811, + "TGAAGG": 1812, + "TGATAA": 1813, + "TGATAT": 1814, + "TGATAC": 1815, + "TGATAG": 1816, + "TGATTA": 1817, + "TGATTT": 1818, + "TGATTC": 1819, + "TGATTG": 1820, + "TGATCA": 1821, + "TGATCT": 1822, + "TGATCC": 1823, + "TGATCG": 1824, + "TGATGA": 1825, + "TGATGT": 1826, + "TGATGC": 1827, + "TGATGG": 1828, + "TGACAA": 1829, + "TGACAT": 1830, + "TGACAC": 1831, + "TGACAG": 1832, + "TGACTA": 1833, + "TGACTT": 1834, + "TGACTC": 1835, + "TGACTG": 1836, + "TGACCA": 1837, + "TGACCT": 1838, + "TGACCC": 1839, + "TGACCG": 1840, + "TGACGA": 1841, + "TGACGT": 1842, + "TGACGC": 1843, + "TGACGG": 1844, + "TGAGAA": 1845, + "TGAGAT": 1846, + "TGAGAC": 1847, + "TGAGAG": 1848, + "TGAGTA": 1849, + "TGAGTT": 1850, + "TGAGTC": 1851, + "TGAGTG": 1852, + "TGAGCA": 1853, + "TGAGCT": 1854, + "TGAGCC": 1855, + "TGAGCG": 1856, + "TGAGGA": 1857, + "TGAGGT": 1858, + "TGAGGC": 1859, + "TGAGGG": 1860, + "TGTAAA": 1861, + "TGTAAT": 1862, + "TGTAAC": 1863, + "TGTAAG": 1864, + "TGTATA": 1865, + "TGTATT": 1866, + "TGTATC": 1867, + "TGTATG": 1868, + "TGTACA": 1869, + "TGTACT": 1870, + "TGTACC": 1871, + "TGTACG": 1872, + "TGTAGA": 1873, + "TGTAGT": 1874, + "TGTAGC": 1875, + "TGTAGG": 1876, + "TGTTAA": 1877, + "TGTTAT": 1878, + "TGTTAC": 1879, + "TGTTAG": 1880, + "TGTTTA": 1881, + "TGTTTT": 1882, + "TGTTTC": 1883, + "TGTTTG": 1884, + "TGTTCA": 1885, + "TGTTCT": 1886, + "TGTTCC": 1887, + "TGTTCG": 1888, + "TGTTGA": 1889, + "TGTTGT": 1890, + "TGTTGC": 1891, + "TGTTGG": 1892, + "TGTCAA": 1893, + "TGTCAT": 1894, + "TGTCAC": 1895, + "TGTCAG": 1896, + "TGTCTA": 1897, + "TGTCTT": 1898, + "TGTCTC": 1899, + "TGTCTG": 1900, + "TGTCCA": 1901, + "TGTCCT": 1902, + "TGTCCC": 1903, + "TGTCCG": 1904, + "TGTCGA": 1905, + "TGTCGT": 1906, + "TGTCGC": 1907, + "TGTCGG": 1908, + "TGTGAA": 1909, + "TGTGAT": 1910, + "TGTGAC": 1911, + "TGTGAG": 1912, + "TGTGTA": 1913, + "TGTGTT": 1914, + "TGTGTC": 1915, + "TGTGTG": 1916, + "TGTGCA": 1917, + "TGTGCT": 1918, + "TGTGCC": 1919, + "TGTGCG": 1920, + "TGTGGA": 1921, + "TGTGGT": 1922, + "TGTGGC": 1923, + "TGTGGG": 1924, + "TGCAAA": 1925, + "TGCAAT": 1926, + "TGCAAC": 1927, + "TGCAAG": 1928, + "TGCATA": 1929, + "TGCATT": 1930, + "TGCATC": 1931, + "TGCATG": 1932, + "TGCACA": 1933, + "TGCACT": 1934, + "TGCACC": 1935, + "TGCACG": 1936, + "TGCAGA": 1937, + "TGCAGT": 1938, + "TGCAGC": 1939, + "TGCAGG": 1940, + "TGCTAA": 1941, + "TGCTAT": 1942, + "TGCTAC": 1943, + "TGCTAG": 1944, + "TGCTTA": 1945, + "TGCTTT": 1946, + "TGCTTC": 1947, + "TGCTTG": 1948, + "TGCTCA": 1949, + "TGCTCT": 1950, + "TGCTCC": 1951, + "TGCTCG": 1952, + "TGCTGA": 1953, + "TGCTGT": 1954, + "TGCTGC": 1955, + "TGCTGG": 1956, + "TGCCAA": 1957, + "TGCCAT": 1958, + "TGCCAC": 1959, + "TGCCAG": 1960, + "TGCCTA": 1961, + "TGCCTT": 1962, + "TGCCTC": 1963, + "TGCCTG": 1964, + "TGCCCA": 1965, + "TGCCCT": 1966, + "TGCCCC": 1967, + "TGCCCG": 1968, + "TGCCGA": 1969, + "TGCCGT": 1970, + "TGCCGC": 1971, + "TGCCGG": 1972, + "TGCGAA": 1973, + "TGCGAT": 1974, + "TGCGAC": 1975, + "TGCGAG": 1976, + "TGCGTA": 1977, + "TGCGTT": 1978, + "TGCGTC": 1979, + "TGCGTG": 1980, + "TGCGCA": 1981, + "TGCGCT": 1982, + "TGCGCC": 1983, + "TGCGCG": 1984, + "TGCGGA": 1985, + "TGCGGT": 1986, + "TGCGGC": 1987, + "TGCGGG": 1988, + "TGGAAA": 1989, + "TGGAAT": 1990, + "TGGAAC": 1991, + "TGGAAG": 1992, + "TGGATA": 1993, + "TGGATT": 1994, + "TGGATC": 1995, + "TGGATG": 1996, + "TGGACA": 1997, + "TGGACT": 1998, + "TGGACC": 1999, + "TGGACG": 2000, + "TGGAGA": 2001, + "TGGAGT": 2002, + "TGGAGC": 2003, + "TGGAGG": 2004, + "TGGTAA": 2005, + "TGGTAT": 2006, + "TGGTAC": 2007, + "TGGTAG": 2008, + "TGGTTA": 2009, + "TGGTTT": 2010, + "TGGTTC": 2011, + "TGGTTG": 2012, + "TGGTCA": 2013, + "TGGTCT": 2014, + "TGGTCC": 2015, + "TGGTCG": 2016, + "TGGTGA": 2017, + "TGGTGT": 2018, + "TGGTGC": 2019, + "TGGTGG": 2020, + "TGGCAA": 2021, + "TGGCAT": 2022, + "TGGCAC": 2023, + "TGGCAG": 2024, + "TGGCTA": 2025, + "TGGCTT": 2026, + "TGGCTC": 2027, + "TGGCTG": 2028, + "TGGCCA": 2029, + "TGGCCT": 2030, + "TGGCCC": 2031, + "TGGCCG": 2032, + "TGGCGA": 2033, + "TGGCGT": 2034, + "TGGCGC": 2035, + "TGGCGG": 2036, + "TGGGAA": 2037, + "TGGGAT": 2038, + "TGGGAC": 2039, + "TGGGAG": 2040, + "TGGGTA": 2041, + "TGGGTT": 2042, + "TGGGTC": 2043, + "TGGGTG": 2044, + "TGGGCA": 2045, + "TGGGCT": 2046, + "TGGGCC": 2047, + "TGGGCG": 2048, + "TGGGGA": 2049, + "TGGGGT": 2050, + "TGGGGC": 2051, + "TGGGGG": 2052, + "CAAAAA": 2053, + "CAAAAT": 2054, + "CAAAAC": 2055, + "CAAAAG": 2056, + "CAAATA": 2057, + "CAAATT": 2058, + "CAAATC": 2059, + "CAAATG": 2060, + "CAAACA": 2061, + "CAAACT": 2062, + "CAAACC": 2063, + "CAAACG": 2064, + "CAAAGA": 2065, + "CAAAGT": 2066, + "CAAAGC": 2067, + "CAAAGG": 2068, + "CAATAA": 2069, + "CAATAT": 2070, + "CAATAC": 2071, + "CAATAG": 2072, + "CAATTA": 2073, + "CAATTT": 2074, + "CAATTC": 2075, + "CAATTG": 2076, + "CAATCA": 2077, + "CAATCT": 2078, + "CAATCC": 2079, + "CAATCG": 2080, + "CAATGA": 2081, + "CAATGT": 2082, + "CAATGC": 2083, + "CAATGG": 2084, + "CAACAA": 2085, + "CAACAT": 2086, + "CAACAC": 2087, + "CAACAG": 2088, + "CAACTA": 2089, + "CAACTT": 2090, + "CAACTC": 2091, + "CAACTG": 2092, + "CAACCA": 2093, + "CAACCT": 2094, + "CAACCC": 2095, + "CAACCG": 2096, + "CAACGA": 2097, + "CAACGT": 2098, + "CAACGC": 2099, + "CAACGG": 2100, + "CAAGAA": 2101, + "CAAGAT": 2102, + "CAAGAC": 2103, + "CAAGAG": 2104, + "CAAGTA": 2105, + "CAAGTT": 2106, + "CAAGTC": 2107, + "CAAGTG": 2108, + "CAAGCA": 2109, + "CAAGCT": 2110, + "CAAGCC": 2111, + "CAAGCG": 2112, + "CAAGGA": 2113, + "CAAGGT": 2114, + "CAAGGC": 2115, + "CAAGGG": 2116, + "CATAAA": 2117, + "CATAAT": 2118, + "CATAAC": 2119, + "CATAAG": 2120, + "CATATA": 2121, + "CATATT": 2122, + "CATATC": 2123, + "CATATG": 2124, + "CATACA": 2125, + "CATACT": 2126, + "CATACC": 2127, + "CATACG": 2128, + "CATAGA": 2129, + "CATAGT": 2130, + "CATAGC": 2131, + "CATAGG": 2132, + "CATTAA": 2133, + "CATTAT": 2134, + "CATTAC": 2135, + "CATTAG": 2136, + "CATTTA": 2137, + "CATTTT": 2138, + "CATTTC": 2139, + "CATTTG": 2140, + "CATTCA": 2141, + "CATTCT": 2142, + "CATTCC": 2143, + "CATTCG": 2144, + "CATTGA": 2145, + "CATTGT": 2146, + "CATTGC": 2147, + "CATTGG": 2148, + "CATCAA": 2149, + "CATCAT": 2150, + "CATCAC": 2151, + "CATCAG": 2152, + "CATCTA": 2153, + "CATCTT": 2154, + "CATCTC": 2155, + "CATCTG": 2156, + "CATCCA": 2157, + "CATCCT": 2158, + "CATCCC": 2159, + "CATCCG": 2160, + "CATCGA": 2161, + "CATCGT": 2162, + "CATCGC": 2163, + "CATCGG": 2164, + "CATGAA": 2165, + "CATGAT": 2166, + "CATGAC": 2167, + "CATGAG": 2168, + "CATGTA": 2169, + "CATGTT": 2170, + "CATGTC": 2171, + "CATGTG": 2172, + "CATGCA": 2173, + "CATGCT": 2174, + "CATGCC": 2175, + "CATGCG": 2176, + "CATGGA": 2177, + "CATGGT": 2178, + "CATGGC": 2179, + "CATGGG": 2180, + "CACAAA": 2181, + "CACAAT": 2182, + "CACAAC": 2183, + "CACAAG": 2184, + "CACATA": 2185, + "CACATT": 2186, + "CACATC": 2187, + "CACATG": 2188, + "CACACA": 2189, + "CACACT": 2190, + "CACACC": 2191, + "CACACG": 2192, + "CACAGA": 2193, + "CACAGT": 2194, + "CACAGC": 2195, + "CACAGG": 2196, + "CACTAA": 2197, + "CACTAT": 2198, + "CACTAC": 2199, + "CACTAG": 2200, + "CACTTA": 2201, + "CACTTT": 2202, + "CACTTC": 2203, + "CACTTG": 2204, + "CACTCA": 2205, + "CACTCT": 2206, + "CACTCC": 2207, + "CACTCG": 2208, + "CACTGA": 2209, + "CACTGT": 2210, + "CACTGC": 2211, + "CACTGG": 2212, + "CACCAA": 2213, + "CACCAT": 2214, + "CACCAC": 2215, + "CACCAG": 2216, + "CACCTA": 2217, + "CACCTT": 2218, + "CACCTC": 2219, + "CACCTG": 2220, + "CACCCA": 2221, + "CACCCT": 2222, + "CACCCC": 2223, + "CACCCG": 2224, + "CACCGA": 2225, + "CACCGT": 2226, + "CACCGC": 2227, + "CACCGG": 2228, + "CACGAA": 2229, + "CACGAT": 2230, + "CACGAC": 2231, + "CACGAG": 2232, + "CACGTA": 2233, + "CACGTT": 2234, + "CACGTC": 2235, + "CACGTG": 2236, + "CACGCA": 2237, + "CACGCT": 2238, + "CACGCC": 2239, + "CACGCG": 2240, + "CACGGA": 2241, + "CACGGT": 2242, + "CACGGC": 2243, + "CACGGG": 2244, + "CAGAAA": 2245, + "CAGAAT": 2246, + "CAGAAC": 2247, + "CAGAAG": 2248, + "CAGATA": 2249, + "CAGATT": 2250, + "CAGATC": 2251, + "CAGATG": 2252, + "CAGACA": 2253, + "CAGACT": 2254, + "CAGACC": 2255, + "CAGACG": 2256, + "CAGAGA": 2257, + "CAGAGT": 2258, + "CAGAGC": 2259, + "CAGAGG": 2260, + "CAGTAA": 2261, + "CAGTAT": 2262, + "CAGTAC": 2263, + "CAGTAG": 2264, + "CAGTTA": 2265, + "CAGTTT": 2266, + "CAGTTC": 2267, + "CAGTTG": 2268, + "CAGTCA": 2269, + "CAGTCT": 2270, + "CAGTCC": 2271, + "CAGTCG": 2272, + "CAGTGA": 2273, + "CAGTGT": 2274, + "CAGTGC": 2275, + "CAGTGG": 2276, + "CAGCAA": 2277, + "CAGCAT": 2278, + "CAGCAC": 2279, + "CAGCAG": 2280, + "CAGCTA": 2281, + "CAGCTT": 2282, + "CAGCTC": 2283, + "CAGCTG": 2284, + "CAGCCA": 2285, + "CAGCCT": 2286, + "CAGCCC": 2287, + "CAGCCG": 2288, + "CAGCGA": 2289, + "CAGCGT": 2290, + "CAGCGC": 2291, + "CAGCGG": 2292, + "CAGGAA": 2293, + "CAGGAT": 2294, + "CAGGAC": 2295, + "CAGGAG": 2296, + "CAGGTA": 2297, + "CAGGTT": 2298, + "CAGGTC": 2299, + "CAGGTG": 2300, + "CAGGCA": 2301, + "CAGGCT": 2302, + "CAGGCC": 2303, + "CAGGCG": 2304, + "CAGGGA": 2305, + "CAGGGT": 2306, + "CAGGGC": 2307, + "CAGGGG": 2308, + "CTAAAA": 2309, + "CTAAAT": 2310, + "CTAAAC": 2311, + "CTAAAG": 2312, + "CTAATA": 2313, + "CTAATT": 2314, + "CTAATC": 2315, + "CTAATG": 2316, + "CTAACA": 2317, + "CTAACT": 2318, + "CTAACC": 2319, + "CTAACG": 2320, + "CTAAGA": 2321, + "CTAAGT": 2322, + "CTAAGC": 2323, + "CTAAGG": 2324, + "CTATAA": 2325, + "CTATAT": 2326, + "CTATAC": 2327, + "CTATAG": 2328, + "CTATTA": 2329, + "CTATTT": 2330, + "CTATTC": 2331, + "CTATTG": 2332, + "CTATCA": 2333, + "CTATCT": 2334, + "CTATCC": 2335, + "CTATCG": 2336, + "CTATGA": 2337, + "CTATGT": 2338, + "CTATGC": 2339, + "CTATGG": 2340, + "CTACAA": 2341, + "CTACAT": 2342, + "CTACAC": 2343, + "CTACAG": 2344, + "CTACTA": 2345, + "CTACTT": 2346, + "CTACTC": 2347, + "CTACTG": 2348, + "CTACCA": 2349, + "CTACCT": 2350, + "CTACCC": 2351, + "CTACCG": 2352, + "CTACGA": 2353, + "CTACGT": 2354, + "CTACGC": 2355, + "CTACGG": 2356, + "CTAGAA": 2357, + "CTAGAT": 2358, + "CTAGAC": 2359, + "CTAGAG": 2360, + "CTAGTA": 2361, + "CTAGTT": 2362, + "CTAGTC": 2363, + "CTAGTG": 2364, + "CTAGCA": 2365, + "CTAGCT": 2366, + "CTAGCC": 2367, + "CTAGCG": 2368, + "CTAGGA": 2369, + "CTAGGT": 2370, + "CTAGGC": 2371, + "CTAGGG": 2372, + "CTTAAA": 2373, + "CTTAAT": 2374, + "CTTAAC": 2375, + "CTTAAG": 2376, + "CTTATA": 2377, + "CTTATT": 2378, + "CTTATC": 2379, + "CTTATG": 2380, + "CTTACA": 2381, + "CTTACT": 2382, + "CTTACC": 2383, + "CTTACG": 2384, + "CTTAGA": 2385, + "CTTAGT": 2386, + "CTTAGC": 2387, + "CTTAGG": 2388, + "CTTTAA": 2389, + "CTTTAT": 2390, + "CTTTAC": 2391, + "CTTTAG": 2392, + "CTTTTA": 2393, + "CTTTTT": 2394, + "CTTTTC": 2395, + "CTTTTG": 2396, + "CTTTCA": 2397, + "CTTTCT": 2398, + "CTTTCC": 2399, + "CTTTCG": 2400, + "CTTTGA": 2401, + "CTTTGT": 2402, + "CTTTGC": 2403, + "CTTTGG": 2404, + "CTTCAA": 2405, + "CTTCAT": 2406, + "CTTCAC": 2407, + "CTTCAG": 2408, + "CTTCTA": 2409, + "CTTCTT": 2410, + "CTTCTC": 2411, + "CTTCTG": 2412, + "CTTCCA": 2413, + "CTTCCT": 2414, + "CTTCCC": 2415, + "CTTCCG": 2416, + "CTTCGA": 2417, + "CTTCGT": 2418, + "CTTCGC": 2419, + "CTTCGG": 2420, + "CTTGAA": 2421, + "CTTGAT": 2422, + "CTTGAC": 2423, + "CTTGAG": 2424, + "CTTGTA": 2425, + "CTTGTT": 2426, + "CTTGTC": 2427, + "CTTGTG": 2428, + "CTTGCA": 2429, + "CTTGCT": 2430, + "CTTGCC": 2431, + "CTTGCG": 2432, + "CTTGGA": 2433, + "CTTGGT": 2434, + "CTTGGC": 2435, + "CTTGGG": 2436, + "CTCAAA": 2437, + "CTCAAT": 2438, + "CTCAAC": 2439, + "CTCAAG": 2440, + "CTCATA": 2441, + "CTCATT": 2442, + "CTCATC": 2443, + "CTCATG": 2444, + "CTCACA": 2445, + "CTCACT": 2446, + "CTCACC": 2447, + "CTCACG": 2448, + "CTCAGA": 2449, + "CTCAGT": 2450, + "CTCAGC": 2451, + "CTCAGG": 2452, + "CTCTAA": 2453, + "CTCTAT": 2454, + "CTCTAC": 2455, + "CTCTAG": 2456, + "CTCTTA": 2457, + "CTCTTT": 2458, + "CTCTTC": 2459, + "CTCTTG": 2460, + "CTCTCA": 2461, + "CTCTCT": 2462, + "CTCTCC": 2463, + "CTCTCG": 2464, + "CTCTGA": 2465, + "CTCTGT": 2466, + "CTCTGC": 2467, + "CTCTGG": 2468, + "CTCCAA": 2469, + "CTCCAT": 2470, + "CTCCAC": 2471, + "CTCCAG": 2472, + "CTCCTA": 2473, + "CTCCTT": 2474, + "CTCCTC": 2475, + "CTCCTG": 2476, + "CTCCCA": 2477, + "CTCCCT": 2478, + "CTCCCC": 2479, + "CTCCCG": 2480, + "CTCCGA": 2481, + "CTCCGT": 2482, + "CTCCGC": 2483, + "CTCCGG": 2484, + "CTCGAA": 2485, + "CTCGAT": 2486, + "CTCGAC": 2487, + "CTCGAG": 2488, + "CTCGTA": 2489, + "CTCGTT": 2490, + "CTCGTC": 2491, + "CTCGTG": 2492, + "CTCGCA": 2493, + "CTCGCT": 2494, + "CTCGCC": 2495, + "CTCGCG": 2496, + "CTCGGA": 2497, + "CTCGGT": 2498, + "CTCGGC": 2499, + "CTCGGG": 2500, + "CTGAAA": 2501, + "CTGAAT": 2502, + "CTGAAC": 2503, + "CTGAAG": 2504, + "CTGATA": 2505, + "CTGATT": 2506, + "CTGATC": 2507, + "CTGATG": 2508, + "CTGACA": 2509, + "CTGACT": 2510, + "CTGACC": 2511, + "CTGACG": 2512, + "CTGAGA": 2513, + "CTGAGT": 2514, + "CTGAGC": 2515, + "CTGAGG": 2516, + "CTGTAA": 2517, + "CTGTAT": 2518, + "CTGTAC": 2519, + "CTGTAG": 2520, + "CTGTTA": 2521, + "CTGTTT": 2522, + "CTGTTC": 2523, + "CTGTTG": 2524, + "CTGTCA": 2525, + "CTGTCT": 2526, + "CTGTCC": 2527, + "CTGTCG": 2528, + "CTGTGA": 2529, + "CTGTGT": 2530, + "CTGTGC": 2531, + "CTGTGG": 2532, + "CTGCAA": 2533, + "CTGCAT": 2534, + "CTGCAC": 2535, + "CTGCAG": 2536, + "CTGCTA": 2537, + "CTGCTT": 2538, + "CTGCTC": 2539, + "CTGCTG": 2540, + "CTGCCA": 2541, + "CTGCCT": 2542, + "CTGCCC": 2543, + "CTGCCG": 2544, + "CTGCGA": 2545, + "CTGCGT": 2546, + "CTGCGC": 2547, + "CTGCGG": 2548, + "CTGGAA": 2549, + "CTGGAT": 2550, + "CTGGAC": 2551, + "CTGGAG": 2552, + "CTGGTA": 2553, + "CTGGTT": 2554, + "CTGGTC": 2555, + "CTGGTG": 2556, + "CTGGCA": 2557, + "CTGGCT": 2558, + "CTGGCC": 2559, + "CTGGCG": 2560, + "CTGGGA": 2561, + "CTGGGT": 2562, + "CTGGGC": 2563, + "CTGGGG": 2564, + "CCAAAA": 2565, + "CCAAAT": 2566, + "CCAAAC": 2567, + "CCAAAG": 2568, + "CCAATA": 2569, + "CCAATT": 2570, + "CCAATC": 2571, + "CCAATG": 2572, + "CCAACA": 2573, + "CCAACT": 2574, + "CCAACC": 2575, + "CCAACG": 2576, + "CCAAGA": 2577, + "CCAAGT": 2578, + "CCAAGC": 2579, + "CCAAGG": 2580, + "CCATAA": 2581, + "CCATAT": 2582, + "CCATAC": 2583, + "CCATAG": 2584, + "CCATTA": 2585, + "CCATTT": 2586, + "CCATTC": 2587, + "CCATTG": 2588, + "CCATCA": 2589, + "CCATCT": 2590, + "CCATCC": 2591, + "CCATCG": 2592, + "CCATGA": 2593, + "CCATGT": 2594, + "CCATGC": 2595, + "CCATGG": 2596, + "CCACAA": 2597, + "CCACAT": 2598, + "CCACAC": 2599, + "CCACAG": 2600, + "CCACTA": 2601, + "CCACTT": 2602, + "CCACTC": 2603, + "CCACTG": 2604, + "CCACCA": 2605, + "CCACCT": 2606, + "CCACCC": 2607, + "CCACCG": 2608, + "CCACGA": 2609, + "CCACGT": 2610, + "CCACGC": 2611, + "CCACGG": 2612, + "CCAGAA": 2613, + "CCAGAT": 2614, + "CCAGAC": 2615, + "CCAGAG": 2616, + "CCAGTA": 2617, + "CCAGTT": 2618, + "CCAGTC": 2619, + "CCAGTG": 2620, + "CCAGCA": 2621, + "CCAGCT": 2622, + "CCAGCC": 2623, + "CCAGCG": 2624, + "CCAGGA": 2625, + "CCAGGT": 2626, + "CCAGGC": 2627, + "CCAGGG": 2628, + "CCTAAA": 2629, + "CCTAAT": 2630, + "CCTAAC": 2631, + "CCTAAG": 2632, + "CCTATA": 2633, + "CCTATT": 2634, + "CCTATC": 2635, + "CCTATG": 2636, + "CCTACA": 2637, + "CCTACT": 2638, + "CCTACC": 2639, + "CCTACG": 2640, + "CCTAGA": 2641, + "CCTAGT": 2642, + "CCTAGC": 2643, + "CCTAGG": 2644, + "CCTTAA": 2645, + "CCTTAT": 2646, + "CCTTAC": 2647, + "CCTTAG": 2648, + "CCTTTA": 2649, + "CCTTTT": 2650, + "CCTTTC": 2651, + "CCTTTG": 2652, + "CCTTCA": 2653, + "CCTTCT": 2654, + "CCTTCC": 2655, + "CCTTCG": 2656, + "CCTTGA": 2657, + "CCTTGT": 2658, + "CCTTGC": 2659, + "CCTTGG": 2660, + "CCTCAA": 2661, + "CCTCAT": 2662, + "CCTCAC": 2663, + "CCTCAG": 2664, + "CCTCTA": 2665, + "CCTCTT": 2666, + "CCTCTC": 2667, + "CCTCTG": 2668, + "CCTCCA": 2669, + "CCTCCT": 2670, + "CCTCCC": 2671, + "CCTCCG": 2672, + "CCTCGA": 2673, + "CCTCGT": 2674, + "CCTCGC": 2675, + "CCTCGG": 2676, + "CCTGAA": 2677, + "CCTGAT": 2678, + "CCTGAC": 2679, + "CCTGAG": 2680, + "CCTGTA": 2681, + "CCTGTT": 2682, + "CCTGTC": 2683, + "CCTGTG": 2684, + "CCTGCA": 2685, + "CCTGCT": 2686, + "CCTGCC": 2687, + "CCTGCG": 2688, + "CCTGGA": 2689, + "CCTGGT": 2690, + "CCTGGC": 2691, + "CCTGGG": 2692, + "CCCAAA": 2693, + "CCCAAT": 2694, + "CCCAAC": 2695, + "CCCAAG": 2696, + "CCCATA": 2697, + "CCCATT": 2698, + "CCCATC": 2699, + "CCCATG": 2700, + "CCCACA": 2701, + "CCCACT": 2702, + "CCCACC": 2703, + "CCCACG": 2704, + "CCCAGA": 2705, + "CCCAGT": 2706, + "CCCAGC": 2707, + "CCCAGG": 2708, + "CCCTAA": 2709, + "CCCTAT": 2710, + "CCCTAC": 2711, + "CCCTAG": 2712, + "CCCTTA": 2713, + "CCCTTT": 2714, + "CCCTTC": 2715, + "CCCTTG": 2716, + "CCCTCA": 2717, + "CCCTCT": 2718, + "CCCTCC": 2719, + "CCCTCG": 2720, + "CCCTGA": 2721, + "CCCTGT": 2722, + "CCCTGC": 2723, + "CCCTGG": 2724, + "CCCCAA": 2725, + "CCCCAT": 2726, + "CCCCAC": 2727, + "CCCCAG": 2728, + "CCCCTA": 2729, + "CCCCTT": 2730, + "CCCCTC": 2731, + "CCCCTG": 2732, + "CCCCCA": 2733, + "CCCCCT": 2734, + "CCCCCC": 2735, + "CCCCCG": 2736, + "CCCCGA": 2737, + "CCCCGT": 2738, + "CCCCGC": 2739, + "CCCCGG": 2740, + "CCCGAA": 2741, + "CCCGAT": 2742, + "CCCGAC": 2743, + "CCCGAG": 2744, + "CCCGTA": 2745, + "CCCGTT": 2746, + "CCCGTC": 2747, + "CCCGTG": 2748, + "CCCGCA": 2749, + "CCCGCT": 2750, + "CCCGCC": 2751, + "CCCGCG": 2752, + "CCCGGA": 2753, + "CCCGGT": 2754, + "CCCGGC": 2755, + "CCCGGG": 2756, + "CCGAAA": 2757, + "CCGAAT": 2758, + "CCGAAC": 2759, + "CCGAAG": 2760, + "CCGATA": 2761, + "CCGATT": 2762, + "CCGATC": 2763, + "CCGATG": 2764, + "CCGACA": 2765, + "CCGACT": 2766, + "CCGACC": 2767, + "CCGACG": 2768, + "CCGAGA": 2769, + "CCGAGT": 2770, + "CCGAGC": 2771, + "CCGAGG": 2772, + "CCGTAA": 2773, + "CCGTAT": 2774, + "CCGTAC": 2775, + "CCGTAG": 2776, + "CCGTTA": 2777, + "CCGTTT": 2778, + "CCGTTC": 2779, + "CCGTTG": 2780, + "CCGTCA": 2781, + "CCGTCT": 2782, + "CCGTCC": 2783, + "CCGTCG": 2784, + "CCGTGA": 2785, + "CCGTGT": 2786, + "CCGTGC": 2787, + "CCGTGG": 2788, + "CCGCAA": 2789, + "CCGCAT": 2790, + "CCGCAC": 2791, + "CCGCAG": 2792, + "CCGCTA": 2793, + "CCGCTT": 2794, + "CCGCTC": 2795, + "CCGCTG": 2796, + "CCGCCA": 2797, + "CCGCCT": 2798, + "CCGCCC": 2799, + "CCGCCG": 2800, + "CCGCGA": 2801, + "CCGCGT": 2802, + "CCGCGC": 2803, + "CCGCGG": 2804, + "CCGGAA": 2805, + "CCGGAT": 2806, + "CCGGAC": 2807, + "CCGGAG": 2808, + "CCGGTA": 2809, + "CCGGTT": 2810, + "CCGGTC": 2811, + "CCGGTG": 2812, + "CCGGCA": 2813, + "CCGGCT": 2814, + "CCGGCC": 2815, + "CCGGCG": 2816, + "CCGGGA": 2817, + "CCGGGT": 2818, + "CCGGGC": 2819, + "CCGGGG": 2820, + "CGAAAA": 2821, + "CGAAAT": 2822, + "CGAAAC": 2823, + "CGAAAG": 2824, + "CGAATA": 2825, + "CGAATT": 2826, + "CGAATC": 2827, + "CGAATG": 2828, + "CGAACA": 2829, + "CGAACT": 2830, + "CGAACC": 2831, + "CGAACG": 2832, + "CGAAGA": 2833, + "CGAAGT": 2834, + "CGAAGC": 2835, + "CGAAGG": 2836, + "CGATAA": 2837, + "CGATAT": 2838, + "CGATAC": 2839, + "CGATAG": 2840, + "CGATTA": 2841, + "CGATTT": 2842, + "CGATTC": 2843, + "CGATTG": 2844, + "CGATCA": 2845, + "CGATCT": 2846, + "CGATCC": 2847, + "CGATCG": 2848, + "CGATGA": 2849, + "CGATGT": 2850, + "CGATGC": 2851, + "CGATGG": 2852, + "CGACAA": 2853, + "CGACAT": 2854, + "CGACAC": 2855, + "CGACAG": 2856, + "CGACTA": 2857, + "CGACTT": 2858, + "CGACTC": 2859, + "CGACTG": 2860, + "CGACCA": 2861, + "CGACCT": 2862, + "CGACCC": 2863, + "CGACCG": 2864, + "CGACGA": 2865, + "CGACGT": 2866, + "CGACGC": 2867, + "CGACGG": 2868, + "CGAGAA": 2869, + "CGAGAT": 2870, + "CGAGAC": 2871, + "CGAGAG": 2872, + "CGAGTA": 2873, + "CGAGTT": 2874, + "CGAGTC": 2875, + "CGAGTG": 2876, + "CGAGCA": 2877, + "CGAGCT": 2878, + "CGAGCC": 2879, + "CGAGCG": 2880, + "CGAGGA": 2881, + "CGAGGT": 2882, + "CGAGGC": 2883, + "CGAGGG": 2884, + "CGTAAA": 2885, + "CGTAAT": 2886, + "CGTAAC": 2887, + "CGTAAG": 2888, + "CGTATA": 2889, + "CGTATT": 2890, + "CGTATC": 2891, + "CGTATG": 2892, + "CGTACA": 2893, + "CGTACT": 2894, + "CGTACC": 2895, + "CGTACG": 2896, + "CGTAGA": 2897, + "CGTAGT": 2898, + "CGTAGC": 2899, + "CGTAGG": 2900, + "CGTTAA": 2901, + "CGTTAT": 2902, + "CGTTAC": 2903, + "CGTTAG": 2904, + "CGTTTA": 2905, + "CGTTTT": 2906, + "CGTTTC": 2907, + "CGTTTG": 2908, + "CGTTCA": 2909, + "CGTTCT": 2910, + "CGTTCC": 2911, + "CGTTCG": 2912, + "CGTTGA": 2913, + "CGTTGT": 2914, + "CGTTGC": 2915, + "CGTTGG": 2916, + "CGTCAA": 2917, + "CGTCAT": 2918, + "CGTCAC": 2919, + "CGTCAG": 2920, + "CGTCTA": 2921, + "CGTCTT": 2922, + "CGTCTC": 2923, + "CGTCTG": 2924, + "CGTCCA": 2925, + "CGTCCT": 2926, + "CGTCCC": 2927, + "CGTCCG": 2928, + "CGTCGA": 2929, + "CGTCGT": 2930, + "CGTCGC": 2931, + "CGTCGG": 2932, + "CGTGAA": 2933, + "CGTGAT": 2934, + "CGTGAC": 2935, + "CGTGAG": 2936, + "CGTGTA": 2937, + "CGTGTT": 2938, + "CGTGTC": 2939, + "CGTGTG": 2940, + "CGTGCA": 2941, + "CGTGCT": 2942, + "CGTGCC": 2943, + "CGTGCG": 2944, + "CGTGGA": 2945, + "CGTGGT": 2946, + "CGTGGC": 2947, + "CGTGGG": 2948, + "CGCAAA": 2949, + "CGCAAT": 2950, + "CGCAAC": 2951, + "CGCAAG": 2952, + "CGCATA": 2953, + "CGCATT": 2954, + "CGCATC": 2955, + "CGCATG": 2956, + "CGCACA": 2957, + "CGCACT": 2958, + "CGCACC": 2959, + "CGCACG": 2960, + "CGCAGA": 2961, + "CGCAGT": 2962, + "CGCAGC": 2963, + "CGCAGG": 2964, + "CGCTAA": 2965, + "CGCTAT": 2966, + "CGCTAC": 2967, + "CGCTAG": 2968, + "CGCTTA": 2969, + "CGCTTT": 2970, + "CGCTTC": 2971, + "CGCTTG": 2972, + "CGCTCA": 2973, + "CGCTCT": 2974, + "CGCTCC": 2975, + "CGCTCG": 2976, + "CGCTGA": 2977, + "CGCTGT": 2978, + "CGCTGC": 2979, + "CGCTGG": 2980, + "CGCCAA": 2981, + "CGCCAT": 2982, + "CGCCAC": 2983, + "CGCCAG": 2984, + "CGCCTA": 2985, + "CGCCTT": 2986, + "CGCCTC": 2987, + "CGCCTG": 2988, + "CGCCCA": 2989, + "CGCCCT": 2990, + "CGCCCC": 2991, + "CGCCCG": 2992, + "CGCCGA": 2993, + "CGCCGT": 2994, + "CGCCGC": 2995, + "CGCCGG": 2996, + "CGCGAA": 2997, + "CGCGAT": 2998, + "CGCGAC": 2999, + "CGCGAG": 3000, + "CGCGTA": 3001, + "CGCGTT": 3002, + "CGCGTC": 3003, + "CGCGTG": 3004, + "CGCGCA": 3005, + "CGCGCT": 3006, + "CGCGCC": 3007, + "CGCGCG": 3008, + "CGCGGA": 3009, + "CGCGGT": 3010, + "CGCGGC": 3011, + "CGCGGG": 3012, + "CGGAAA": 3013, + "CGGAAT": 3014, + "CGGAAC": 3015, + "CGGAAG": 3016, + "CGGATA": 3017, + "CGGATT": 3018, + "CGGATC": 3019, + "CGGATG": 3020, + "CGGACA": 3021, + "CGGACT": 3022, + "CGGACC": 3023, + "CGGACG": 3024, + "CGGAGA": 3025, + "CGGAGT": 3026, + "CGGAGC": 3027, + "CGGAGG": 3028, + "CGGTAA": 3029, + "CGGTAT": 3030, + "CGGTAC": 3031, + "CGGTAG": 3032, + "CGGTTA": 3033, + "CGGTTT": 3034, + "CGGTTC": 3035, + "CGGTTG": 3036, + "CGGTCA": 3037, + "CGGTCT": 3038, + "CGGTCC": 3039, + "CGGTCG": 3040, + "CGGTGA": 3041, + "CGGTGT": 3042, + "CGGTGC": 3043, + "CGGTGG": 3044, + "CGGCAA": 3045, + "CGGCAT": 3046, + "CGGCAC": 3047, + "CGGCAG": 3048, + "CGGCTA": 3049, + "CGGCTT": 3050, + "CGGCTC": 3051, + "CGGCTG": 3052, + "CGGCCA": 3053, + "CGGCCT": 3054, + "CGGCCC": 3055, + "CGGCCG": 3056, + "CGGCGA": 3057, + "CGGCGT": 3058, + "CGGCGC": 3059, + "CGGCGG": 3060, + "CGGGAA": 3061, + "CGGGAT": 3062, + "CGGGAC": 3063, + "CGGGAG": 3064, + "CGGGTA": 3065, + "CGGGTT": 3066, + "CGGGTC": 3067, + "CGGGTG": 3068, + "CGGGCA": 3069, + "CGGGCT": 3070, + "CGGGCC": 3071, + "CGGGCG": 3072, + "CGGGGA": 3073, + "CGGGGT": 3074, + "CGGGGC": 3075, + "CGGGGG": 3076, + "GAAAAA": 3077, + "GAAAAT": 3078, + "GAAAAC": 3079, + "GAAAAG": 3080, + "GAAATA": 3081, + "GAAATT": 3082, + "GAAATC": 3083, + "GAAATG": 3084, + "GAAACA": 3085, + "GAAACT": 3086, + "GAAACC": 3087, + "GAAACG": 3088, + "GAAAGA": 3089, + "GAAAGT": 3090, + "GAAAGC": 3091, + "GAAAGG": 3092, + "GAATAA": 3093, + "GAATAT": 3094, + "GAATAC": 3095, + "GAATAG": 3096, + "GAATTA": 3097, + "GAATTT": 3098, + "GAATTC": 3099, + "GAATTG": 3100, + "GAATCA": 3101, + "GAATCT": 3102, + "GAATCC": 3103, + "GAATCG": 3104, + "GAATGA": 3105, + "GAATGT": 3106, + "GAATGC": 3107, + "GAATGG": 3108, + "GAACAA": 3109, + "GAACAT": 3110, + "GAACAC": 3111, + "GAACAG": 3112, + "GAACTA": 3113, + "GAACTT": 3114, + "GAACTC": 3115, + "GAACTG": 3116, + "GAACCA": 3117, + "GAACCT": 3118, + "GAACCC": 3119, + "GAACCG": 3120, + "GAACGA": 3121, + "GAACGT": 3122, + "GAACGC": 3123, + "GAACGG": 3124, + "GAAGAA": 3125, + "GAAGAT": 3126, + "GAAGAC": 3127, + "GAAGAG": 3128, + "GAAGTA": 3129, + "GAAGTT": 3130, + "GAAGTC": 3131, + "GAAGTG": 3132, + "GAAGCA": 3133, + "GAAGCT": 3134, + "GAAGCC": 3135, + "GAAGCG": 3136, + "GAAGGA": 3137, + "GAAGGT": 3138, + "GAAGGC": 3139, + "GAAGGG": 3140, + "GATAAA": 3141, + "GATAAT": 3142, + "GATAAC": 3143, + "GATAAG": 3144, + "GATATA": 3145, + "GATATT": 3146, + "GATATC": 3147, + "GATATG": 3148, + "GATACA": 3149, + "GATACT": 3150, + "GATACC": 3151, + "GATACG": 3152, + "GATAGA": 3153, + "GATAGT": 3154, + "GATAGC": 3155, + "GATAGG": 3156, + "GATTAA": 3157, + "GATTAT": 3158, + "GATTAC": 3159, + "GATTAG": 3160, + "GATTTA": 3161, + "GATTTT": 3162, + "GATTTC": 3163, + "GATTTG": 3164, + "GATTCA": 3165, + "GATTCT": 3166, + "GATTCC": 3167, + "GATTCG": 3168, + "GATTGA": 3169, + "GATTGT": 3170, + "GATTGC": 3171, + "GATTGG": 3172, + "GATCAA": 3173, + "GATCAT": 3174, + "GATCAC": 3175, + "GATCAG": 3176, + "GATCTA": 3177, + "GATCTT": 3178, + "GATCTC": 3179, + "GATCTG": 3180, + "GATCCA": 3181, + "GATCCT": 3182, + "GATCCC": 3183, + "GATCCG": 3184, + "GATCGA": 3185, + "GATCGT": 3186, + "GATCGC": 3187, + "GATCGG": 3188, + "GATGAA": 3189, + "GATGAT": 3190, + "GATGAC": 3191, + "GATGAG": 3192, + "GATGTA": 3193, + "GATGTT": 3194, + "GATGTC": 3195, + "GATGTG": 3196, + "GATGCA": 3197, + "GATGCT": 3198, + "GATGCC": 3199, + "GATGCG": 3200, + "GATGGA": 3201, + "GATGGT": 3202, + "GATGGC": 3203, + "GATGGG": 3204, + "GACAAA": 3205, + "GACAAT": 3206, + "GACAAC": 3207, + "GACAAG": 3208, + "GACATA": 3209, + "GACATT": 3210, + "GACATC": 3211, + "GACATG": 3212, + "GACACA": 3213, + "GACACT": 3214, + "GACACC": 3215, + "GACACG": 3216, + "GACAGA": 3217, + "GACAGT": 3218, + "GACAGC": 3219, + "GACAGG": 3220, + "GACTAA": 3221, + "GACTAT": 3222, + "GACTAC": 3223, + "GACTAG": 3224, + "GACTTA": 3225, + "GACTTT": 3226, + "GACTTC": 3227, + "GACTTG": 3228, + "GACTCA": 3229, + "GACTCT": 3230, + "GACTCC": 3231, + "GACTCG": 3232, + "GACTGA": 3233, + "GACTGT": 3234, + "GACTGC": 3235, + "GACTGG": 3236, + "GACCAA": 3237, + "GACCAT": 3238, + "GACCAC": 3239, + "GACCAG": 3240, + "GACCTA": 3241, + "GACCTT": 3242, + "GACCTC": 3243, + "GACCTG": 3244, + "GACCCA": 3245, + "GACCCT": 3246, + "GACCCC": 3247, + "GACCCG": 3248, + "GACCGA": 3249, + "GACCGT": 3250, + "GACCGC": 3251, + "GACCGG": 3252, + "GACGAA": 3253, + "GACGAT": 3254, + "GACGAC": 3255, + "GACGAG": 3256, + "GACGTA": 3257, + "GACGTT": 3258, + "GACGTC": 3259, + "GACGTG": 3260, + "GACGCA": 3261, + "GACGCT": 3262, + "GACGCC": 3263, + "GACGCG": 3264, + "GACGGA": 3265, + "GACGGT": 3266, + "GACGGC": 3267, + "GACGGG": 3268, + "GAGAAA": 3269, + "GAGAAT": 3270, + "GAGAAC": 3271, + "GAGAAG": 3272, + "GAGATA": 3273, + "GAGATT": 3274, + "GAGATC": 3275, + "GAGATG": 3276, + "GAGACA": 3277, + "GAGACT": 3278, + "GAGACC": 3279, + "GAGACG": 3280, + "GAGAGA": 3281, + "GAGAGT": 3282, + "GAGAGC": 3283, + "GAGAGG": 3284, + "GAGTAA": 3285, + "GAGTAT": 3286, + "GAGTAC": 3287, + "GAGTAG": 3288, + "GAGTTA": 3289, + "GAGTTT": 3290, + "GAGTTC": 3291, + "GAGTTG": 3292, + "GAGTCA": 3293, + "GAGTCT": 3294, + "GAGTCC": 3295, + "GAGTCG": 3296, + "GAGTGA": 3297, + "GAGTGT": 3298, + "GAGTGC": 3299, + "GAGTGG": 3300, + "GAGCAA": 3301, + "GAGCAT": 3302, + "GAGCAC": 3303, + "GAGCAG": 3304, + "GAGCTA": 3305, + "GAGCTT": 3306, + "GAGCTC": 3307, + "GAGCTG": 3308, + "GAGCCA": 3309, + "GAGCCT": 3310, + "GAGCCC": 3311, + "GAGCCG": 3312, + "GAGCGA": 3313, + "GAGCGT": 3314, + "GAGCGC": 3315, + "GAGCGG": 3316, + "GAGGAA": 3317, + "GAGGAT": 3318, + "GAGGAC": 3319, + "GAGGAG": 3320, + "GAGGTA": 3321, + "GAGGTT": 3322, + "GAGGTC": 3323, + "GAGGTG": 3324, + "GAGGCA": 3325, + "GAGGCT": 3326, + "GAGGCC": 3327, + "GAGGCG": 3328, + "GAGGGA": 3329, + "GAGGGT": 3330, + "GAGGGC": 3331, + "GAGGGG": 3332, + "GTAAAA": 3333, + "GTAAAT": 3334, + "GTAAAC": 3335, + "GTAAAG": 3336, + "GTAATA": 3337, + "GTAATT": 3338, + "GTAATC": 3339, + "GTAATG": 3340, + "GTAACA": 3341, + "GTAACT": 3342, + "GTAACC": 3343, + "GTAACG": 3344, + "GTAAGA": 3345, + "GTAAGT": 3346, + "GTAAGC": 3347, + "GTAAGG": 3348, + "GTATAA": 3349, + "GTATAT": 3350, + "GTATAC": 3351, + "GTATAG": 3352, + "GTATTA": 3353, + "GTATTT": 3354, + "GTATTC": 3355, + "GTATTG": 3356, + "GTATCA": 3357, + "GTATCT": 3358, + "GTATCC": 3359, + "GTATCG": 3360, + "GTATGA": 3361, + "GTATGT": 3362, + "GTATGC": 3363, + "GTATGG": 3364, + "GTACAA": 3365, + "GTACAT": 3366, + "GTACAC": 3367, + "GTACAG": 3368, + "GTACTA": 3369, + "GTACTT": 3370, + "GTACTC": 3371, + "GTACTG": 3372, + "GTACCA": 3373, + "GTACCT": 3374, + "GTACCC": 3375, + "GTACCG": 3376, + "GTACGA": 3377, + "GTACGT": 3378, + "GTACGC": 3379, + "GTACGG": 3380, + "GTAGAA": 3381, + "GTAGAT": 3382, + "GTAGAC": 3383, + "GTAGAG": 3384, + "GTAGTA": 3385, + "GTAGTT": 3386, + "GTAGTC": 3387, + "GTAGTG": 3388, + "GTAGCA": 3389, + "GTAGCT": 3390, + "GTAGCC": 3391, + "GTAGCG": 3392, + "GTAGGA": 3393, + "GTAGGT": 3394, + "GTAGGC": 3395, + "GTAGGG": 3396, + "GTTAAA": 3397, + "GTTAAT": 3398, + "GTTAAC": 3399, + "GTTAAG": 3400, + "GTTATA": 3401, + "GTTATT": 3402, + "GTTATC": 3403, + "GTTATG": 3404, + "GTTACA": 3405, + "GTTACT": 3406, + "GTTACC": 3407, + "GTTACG": 3408, + "GTTAGA": 3409, + "GTTAGT": 3410, + "GTTAGC": 3411, + "GTTAGG": 3412, + "GTTTAA": 3413, + "GTTTAT": 3414, + "GTTTAC": 3415, + "GTTTAG": 3416, + "GTTTTA": 3417, + "GTTTTT": 3418, + "GTTTTC": 3419, + "GTTTTG": 3420, + "GTTTCA": 3421, + "GTTTCT": 3422, + "GTTTCC": 3423, + "GTTTCG": 3424, + "GTTTGA": 3425, + "GTTTGT": 3426, + "GTTTGC": 3427, + "GTTTGG": 3428, + "GTTCAA": 3429, + "GTTCAT": 3430, + "GTTCAC": 3431, + "GTTCAG": 3432, + "GTTCTA": 3433, + "GTTCTT": 3434, + "GTTCTC": 3435, + "GTTCTG": 3436, + "GTTCCA": 3437, + "GTTCCT": 3438, + "GTTCCC": 3439, + "GTTCCG": 3440, + "GTTCGA": 3441, + "GTTCGT": 3442, + "GTTCGC": 3443, + "GTTCGG": 3444, + "GTTGAA": 3445, + "GTTGAT": 3446, + "GTTGAC": 3447, + "GTTGAG": 3448, + "GTTGTA": 3449, + "GTTGTT": 3450, + "GTTGTC": 3451, + "GTTGTG": 3452, + "GTTGCA": 3453, + "GTTGCT": 3454, + "GTTGCC": 3455, + "GTTGCG": 3456, + "GTTGGA": 3457, + "GTTGGT": 3458, + "GTTGGC": 3459, + "GTTGGG": 3460, + "GTCAAA": 3461, + "GTCAAT": 3462, + "GTCAAC": 3463, + "GTCAAG": 3464, + "GTCATA": 3465, + "GTCATT": 3466, + "GTCATC": 3467, + "GTCATG": 3468, + "GTCACA": 3469, + "GTCACT": 3470, + "GTCACC": 3471, + "GTCACG": 3472, + "GTCAGA": 3473, + "GTCAGT": 3474, + "GTCAGC": 3475, + "GTCAGG": 3476, + "GTCTAA": 3477, + "GTCTAT": 3478, + "GTCTAC": 3479, + "GTCTAG": 3480, + "GTCTTA": 3481, + "GTCTTT": 3482, + "GTCTTC": 3483, + "GTCTTG": 3484, + "GTCTCA": 3485, + "GTCTCT": 3486, + "GTCTCC": 3487, + "GTCTCG": 3488, + "GTCTGA": 3489, + "GTCTGT": 3490, + "GTCTGC": 3491, + "GTCTGG": 3492, + "GTCCAA": 3493, + "GTCCAT": 3494, + "GTCCAC": 3495, + "GTCCAG": 3496, + "GTCCTA": 3497, + "GTCCTT": 3498, + "GTCCTC": 3499, + "GTCCTG": 3500, + "GTCCCA": 3501, + "GTCCCT": 3502, + "GTCCCC": 3503, + "GTCCCG": 3504, + "GTCCGA": 3505, + "GTCCGT": 3506, + "GTCCGC": 3507, + "GTCCGG": 3508, + "GTCGAA": 3509, + "GTCGAT": 3510, + "GTCGAC": 3511, + "GTCGAG": 3512, + "GTCGTA": 3513, + "GTCGTT": 3514, + "GTCGTC": 3515, + "GTCGTG": 3516, + "GTCGCA": 3517, + "GTCGCT": 3518, + "GTCGCC": 3519, + "GTCGCG": 3520, + "GTCGGA": 3521, + "GTCGGT": 3522, + "GTCGGC": 3523, + "GTCGGG": 3524, + "GTGAAA": 3525, + "GTGAAT": 3526, + "GTGAAC": 3527, + "GTGAAG": 3528, + "GTGATA": 3529, + "GTGATT": 3530, + "GTGATC": 3531, + "GTGATG": 3532, + "GTGACA": 3533, + "GTGACT": 3534, + "GTGACC": 3535, + "GTGACG": 3536, + "GTGAGA": 3537, + "GTGAGT": 3538, + "GTGAGC": 3539, + "GTGAGG": 3540, + "GTGTAA": 3541, + "GTGTAT": 3542, + "GTGTAC": 3543, + "GTGTAG": 3544, + "GTGTTA": 3545, + "GTGTTT": 3546, + "GTGTTC": 3547, + "GTGTTG": 3548, + "GTGTCA": 3549, + "GTGTCT": 3550, + "GTGTCC": 3551, + "GTGTCG": 3552, + "GTGTGA": 3553, + "GTGTGT": 3554, + "GTGTGC": 3555, + "GTGTGG": 3556, + "GTGCAA": 3557, + "GTGCAT": 3558, + "GTGCAC": 3559, + "GTGCAG": 3560, + "GTGCTA": 3561, + "GTGCTT": 3562, + "GTGCTC": 3563, + "GTGCTG": 3564, + "GTGCCA": 3565, + "GTGCCT": 3566, + "GTGCCC": 3567, + "GTGCCG": 3568, + "GTGCGA": 3569, + "GTGCGT": 3570, + "GTGCGC": 3571, + "GTGCGG": 3572, + "GTGGAA": 3573, + "GTGGAT": 3574, + "GTGGAC": 3575, + "GTGGAG": 3576, + "GTGGTA": 3577, + "GTGGTT": 3578, + "GTGGTC": 3579, + "GTGGTG": 3580, + "GTGGCA": 3581, + "GTGGCT": 3582, + "GTGGCC": 3583, + "GTGGCG": 3584, + "GTGGGA": 3585, + "GTGGGT": 3586, + "GTGGGC": 3587, + "GTGGGG": 3588, + "GCAAAA": 3589, + "GCAAAT": 3590, + "GCAAAC": 3591, + "GCAAAG": 3592, + "GCAATA": 3593, + "GCAATT": 3594, + "GCAATC": 3595, + "GCAATG": 3596, + "GCAACA": 3597, + "GCAACT": 3598, + "GCAACC": 3599, + "GCAACG": 3600, + "GCAAGA": 3601, + "GCAAGT": 3602, + "GCAAGC": 3603, + "GCAAGG": 3604, + "GCATAA": 3605, + "GCATAT": 3606, + "GCATAC": 3607, + "GCATAG": 3608, + "GCATTA": 3609, + "GCATTT": 3610, + "GCATTC": 3611, + "GCATTG": 3612, + "GCATCA": 3613, + "GCATCT": 3614, + "GCATCC": 3615, + "GCATCG": 3616, + "GCATGA": 3617, + "GCATGT": 3618, + "GCATGC": 3619, + "GCATGG": 3620, + "GCACAA": 3621, + "GCACAT": 3622, + "GCACAC": 3623, + "GCACAG": 3624, + "GCACTA": 3625, + "GCACTT": 3626, + "GCACTC": 3627, + "GCACTG": 3628, + "GCACCA": 3629, + "GCACCT": 3630, + "GCACCC": 3631, + "GCACCG": 3632, + "GCACGA": 3633, + "GCACGT": 3634, + "GCACGC": 3635, + "GCACGG": 3636, + "GCAGAA": 3637, + "GCAGAT": 3638, + "GCAGAC": 3639, + "GCAGAG": 3640, + "GCAGTA": 3641, + "GCAGTT": 3642, + "GCAGTC": 3643, + "GCAGTG": 3644, + "GCAGCA": 3645, + "GCAGCT": 3646, + "GCAGCC": 3647, + "GCAGCG": 3648, + "GCAGGA": 3649, + "GCAGGT": 3650, + "GCAGGC": 3651, + "GCAGGG": 3652, + "GCTAAA": 3653, + "GCTAAT": 3654, + "GCTAAC": 3655, + "GCTAAG": 3656, + "GCTATA": 3657, + "GCTATT": 3658, + "GCTATC": 3659, + "GCTATG": 3660, + "GCTACA": 3661, + "GCTACT": 3662, + "GCTACC": 3663, + "GCTACG": 3664, + "GCTAGA": 3665, + "GCTAGT": 3666, + "GCTAGC": 3667, + "GCTAGG": 3668, + "GCTTAA": 3669, + "GCTTAT": 3670, + "GCTTAC": 3671, + "GCTTAG": 3672, + "GCTTTA": 3673, + "GCTTTT": 3674, + "GCTTTC": 3675, + "GCTTTG": 3676, + "GCTTCA": 3677, + "GCTTCT": 3678, + "GCTTCC": 3679, + "GCTTCG": 3680, + "GCTTGA": 3681, + "GCTTGT": 3682, + "GCTTGC": 3683, + "GCTTGG": 3684, + "GCTCAA": 3685, + "GCTCAT": 3686, + "GCTCAC": 3687, + "GCTCAG": 3688, + "GCTCTA": 3689, + "GCTCTT": 3690, + "GCTCTC": 3691, + "GCTCTG": 3692, + "GCTCCA": 3693, + "GCTCCT": 3694, + "GCTCCC": 3695, + "GCTCCG": 3696, + "GCTCGA": 3697, + "GCTCGT": 3698, + "GCTCGC": 3699, + "GCTCGG": 3700, + "GCTGAA": 3701, + "GCTGAT": 3702, + "GCTGAC": 3703, + "GCTGAG": 3704, + "GCTGTA": 3705, + "GCTGTT": 3706, + "GCTGTC": 3707, + "GCTGTG": 3708, + "GCTGCA": 3709, + "GCTGCT": 3710, + "GCTGCC": 3711, + "GCTGCG": 3712, + "GCTGGA": 3713, + "GCTGGT": 3714, + "GCTGGC": 3715, + "GCTGGG": 3716, + "GCCAAA": 3717, + "GCCAAT": 3718, + "GCCAAC": 3719, + "GCCAAG": 3720, + "GCCATA": 3721, + "GCCATT": 3722, + "GCCATC": 3723, + "GCCATG": 3724, + "GCCACA": 3725, + "GCCACT": 3726, + "GCCACC": 3727, + "GCCACG": 3728, + "GCCAGA": 3729, + "GCCAGT": 3730, + "GCCAGC": 3731, + "GCCAGG": 3732, + "GCCTAA": 3733, + "GCCTAT": 3734, + "GCCTAC": 3735, + "GCCTAG": 3736, + "GCCTTA": 3737, + "GCCTTT": 3738, + "GCCTTC": 3739, + "GCCTTG": 3740, + "GCCTCA": 3741, + "GCCTCT": 3742, + "GCCTCC": 3743, + "GCCTCG": 3744, + "GCCTGA": 3745, + "GCCTGT": 3746, + "GCCTGC": 3747, + "GCCTGG": 3748, + "GCCCAA": 3749, + "GCCCAT": 3750, + "GCCCAC": 3751, + "GCCCAG": 3752, + "GCCCTA": 3753, + "GCCCTT": 3754, + "GCCCTC": 3755, + "GCCCTG": 3756, + "GCCCCA": 3757, + "GCCCCT": 3758, + "GCCCCC": 3759, + "GCCCCG": 3760, + "GCCCGA": 3761, + "GCCCGT": 3762, + "GCCCGC": 3763, + "GCCCGG": 3764, + "GCCGAA": 3765, + "GCCGAT": 3766, + "GCCGAC": 3767, + "GCCGAG": 3768, + "GCCGTA": 3769, + "GCCGTT": 3770, + "GCCGTC": 3771, + "GCCGTG": 3772, + "GCCGCA": 3773, + "GCCGCT": 3774, + "GCCGCC": 3775, + "GCCGCG": 3776, + "GCCGGA": 3777, + "GCCGGT": 3778, + "GCCGGC": 3779, + "GCCGGG": 3780, + "GCGAAA": 3781, + "GCGAAT": 3782, + "GCGAAC": 3783, + "GCGAAG": 3784, + "GCGATA": 3785, + "GCGATT": 3786, + "GCGATC": 3787, + "GCGATG": 3788, + "GCGACA": 3789, + "GCGACT": 3790, + "GCGACC": 3791, + "GCGACG": 3792, + "GCGAGA": 3793, + "GCGAGT": 3794, + "GCGAGC": 3795, + "GCGAGG": 3796, + "GCGTAA": 3797, + "GCGTAT": 3798, + "GCGTAC": 3799, + "GCGTAG": 3800, + "GCGTTA": 3801, + "GCGTTT": 3802, + "GCGTTC": 3803, + "GCGTTG": 3804, + "GCGTCA": 3805, + "GCGTCT": 3806, + "GCGTCC": 3807, + "GCGTCG": 3808, + "GCGTGA": 3809, + "GCGTGT": 3810, + "GCGTGC": 3811, + "GCGTGG": 3812, + "GCGCAA": 3813, + "GCGCAT": 3814, + "GCGCAC": 3815, + "GCGCAG": 3816, + "GCGCTA": 3817, + "GCGCTT": 3818, + "GCGCTC": 3819, + "GCGCTG": 3820, + "GCGCCA": 3821, + "GCGCCT": 3822, + "GCGCCC": 3823, + "GCGCCG": 3824, + "GCGCGA": 3825, + "GCGCGT": 3826, + "GCGCGC": 3827, + "GCGCGG": 3828, + "GCGGAA": 3829, + "GCGGAT": 3830, + "GCGGAC": 3831, + "GCGGAG": 3832, + "GCGGTA": 3833, + "GCGGTT": 3834, + "GCGGTC": 3835, + "GCGGTG": 3836, + "GCGGCA": 3837, + "GCGGCT": 3838, + "GCGGCC": 3839, + "GCGGCG": 3840, + "GCGGGA": 3841, + "GCGGGT": 3842, + "GCGGGC": 3843, + "GCGGGG": 3844, + "GGAAAA": 3845, + "GGAAAT": 3846, + "GGAAAC": 3847, + "GGAAAG": 3848, + "GGAATA": 3849, + "GGAATT": 3850, + "GGAATC": 3851, + "GGAATG": 3852, + "GGAACA": 3853, + "GGAACT": 3854, + "GGAACC": 3855, + "GGAACG": 3856, + "GGAAGA": 3857, + "GGAAGT": 3858, + "GGAAGC": 3859, + "GGAAGG": 3860, + "GGATAA": 3861, + "GGATAT": 3862, + "GGATAC": 3863, + "GGATAG": 3864, + "GGATTA": 3865, + "GGATTT": 3866, + "GGATTC": 3867, + "GGATTG": 3868, + "GGATCA": 3869, + "GGATCT": 3870, + "GGATCC": 3871, + "GGATCG": 3872, + "GGATGA": 3873, + "GGATGT": 3874, + "GGATGC": 3875, + "GGATGG": 3876, + "GGACAA": 3877, + "GGACAT": 3878, + "GGACAC": 3879, + "GGACAG": 3880, + "GGACTA": 3881, + "GGACTT": 3882, + "GGACTC": 3883, + "GGACTG": 3884, + "GGACCA": 3885, + "GGACCT": 3886, + "GGACCC": 3887, + "GGACCG": 3888, + "GGACGA": 3889, + "GGACGT": 3890, + "GGACGC": 3891, + "GGACGG": 3892, + "GGAGAA": 3893, + "GGAGAT": 3894, + "GGAGAC": 3895, + "GGAGAG": 3896, + "GGAGTA": 3897, + "GGAGTT": 3898, + "GGAGTC": 3899, + "GGAGTG": 3900, + "GGAGCA": 3901, + "GGAGCT": 3902, + "GGAGCC": 3903, + "GGAGCG": 3904, + "GGAGGA": 3905, + "GGAGGT": 3906, + "GGAGGC": 3907, + "GGAGGG": 3908, + "GGTAAA": 3909, + "GGTAAT": 3910, + "GGTAAC": 3911, + "GGTAAG": 3912, + "GGTATA": 3913, + "GGTATT": 3914, + "GGTATC": 3915, + "GGTATG": 3916, + "GGTACA": 3917, + "GGTACT": 3918, + "GGTACC": 3919, + "GGTACG": 3920, + "GGTAGA": 3921, + "GGTAGT": 3922, + "GGTAGC": 3923, + "GGTAGG": 3924, + "GGTTAA": 3925, + "GGTTAT": 3926, + "GGTTAC": 3927, + "GGTTAG": 3928, + "GGTTTA": 3929, + "GGTTTT": 3930, + "GGTTTC": 3931, + "GGTTTG": 3932, + "GGTTCA": 3933, + "GGTTCT": 3934, + "GGTTCC": 3935, + "GGTTCG": 3936, + "GGTTGA": 3937, + "GGTTGT": 3938, + "GGTTGC": 3939, + "GGTTGG": 3940, + "GGTCAA": 3941, + "GGTCAT": 3942, + "GGTCAC": 3943, + "GGTCAG": 3944, + "GGTCTA": 3945, + "GGTCTT": 3946, + "GGTCTC": 3947, + "GGTCTG": 3948, + "GGTCCA": 3949, + "GGTCCT": 3950, + "GGTCCC": 3951, + "GGTCCG": 3952, + "GGTCGA": 3953, + "GGTCGT": 3954, + "GGTCGC": 3955, + "GGTCGG": 3956, + "GGTGAA": 3957, + "GGTGAT": 3958, + "GGTGAC": 3959, + "GGTGAG": 3960, + "GGTGTA": 3961, + "GGTGTT": 3962, + "GGTGTC": 3963, + "GGTGTG": 3964, + "GGTGCA": 3965, + "GGTGCT": 3966, + "GGTGCC": 3967, + "GGTGCG": 3968, + "GGTGGA": 3969, + "GGTGGT": 3970, + "GGTGGC": 3971, + "GGTGGG": 3972, + "GGCAAA": 3973, + "GGCAAT": 3974, + "GGCAAC": 3975, + "GGCAAG": 3976, + "GGCATA": 3977, + "GGCATT": 3978, + "GGCATC": 3979, + "GGCATG": 3980, + "GGCACA": 3981, + "GGCACT": 3982, + "GGCACC": 3983, + "GGCACG": 3984, + "GGCAGA": 3985, + "GGCAGT": 3986, + "GGCAGC": 3987, + "GGCAGG": 3988, + "GGCTAA": 3989, + "GGCTAT": 3990, + "GGCTAC": 3991, + "GGCTAG": 3992, + "GGCTTA": 3993, + "GGCTTT": 3994, + "GGCTTC": 3995, + "GGCTTG": 3996, + "GGCTCA": 3997, + "GGCTCT": 3998, + "GGCTCC": 3999, + "GGCTCG": 4000, + "GGCTGA": 4001, + "GGCTGT": 4002, + "GGCTGC": 4003, + "GGCTGG": 4004, + "GGCCAA": 4005, + "GGCCAT": 4006, + "GGCCAC": 4007, + "GGCCAG": 4008, + "GGCCTA": 4009, + "GGCCTT": 4010, + "GGCCTC": 4011, + "GGCCTG": 4012, + "GGCCCA": 4013, + "GGCCCT": 4014, + "GGCCCC": 4015, + "GGCCCG": 4016, + "GGCCGA": 4017, + "GGCCGT": 4018, + "GGCCGC": 4019, + "GGCCGG": 4020, + "GGCGAA": 4021, + "GGCGAT": 4022, + "GGCGAC": 4023, + "GGCGAG": 4024, + "GGCGTA": 4025, + "GGCGTT": 4026, + "GGCGTC": 4027, + "GGCGTG": 4028, + "GGCGCA": 4029, + "GGCGCT": 4030, + "GGCGCC": 4031, + "GGCGCG": 4032, + "GGCGGA": 4033, + "GGCGGT": 4034, + "GGCGGC": 4035, + "GGCGGG": 4036, + "GGGAAA": 4037, + "GGGAAT": 4038, + "GGGAAC": 4039, + "GGGAAG": 4040, + "GGGATA": 4041, + "GGGATT": 4042, + "GGGATC": 4043, + "GGGATG": 4044, + "GGGACA": 4045, + "GGGACT": 4046, + "GGGACC": 4047, + "GGGACG": 4048, + "GGGAGA": 4049, + "GGGAGT": 4050, + "GGGAGC": 4051, + "GGGAGG": 4052, + "GGGTAA": 4053, + "GGGTAT": 4054, + "GGGTAC": 4055, + "GGGTAG": 4056, + "GGGTTA": 4057, + "GGGTTT": 4058, + "GGGTTC": 4059, + "GGGTTG": 4060, + "GGGTCA": 4061, + "GGGTCT": 4062, + "GGGTCC": 4063, + "GGGTCG": 4064, + "GGGTGA": 4065, + "GGGTGT": 4066, + "GGGTGC": 4067, + "GGGTGG": 4068, + "GGGCAA": 4069, + "GGGCAT": 4070, + "GGGCAC": 4071, + "GGGCAG": 4072, + "GGGCTA": 4073, + "GGGCTT": 4074, + "GGGCTC": 4075, + "GGGCTG": 4076, + "GGGCCA": 4077, + "GGGCCT": 4078, + "GGGCCC": 4079, + "GGGCCG": 4080, + "GGGCGA": 4081, + "GGGCGT": 4082, + "GGGCGC": 4083, + "GGGCGG": 4084, + "GGGGAA": 4085, + "GGGGAT": 4086, + "GGGGAC": 4087, + "GGGGAG": 4088, + "GGGGTA": 4089, + "GGGGTT": 4090, + "GGGGTC": 4091, + "GGGGTG": 4092, + "GGGGCA": 4093, + "GGGGCT": 4094, + "GGGGCC": 4095, + "GGGGCG": 4096, + "GGGGGA": 4097, + "GGGGGT": 4098, + "GGGGGC": 4099, + "GGGGGG": 4100 + } + } +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/tokenizer_config.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..604e2b1b29af1dcb3920d2828b6a90276e5e3d02 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/tokenizer_config.json @@ -0,0 +1,58 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "[UNK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "[CLS]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "[SEP]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "4": { + "content": "[MASK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "clean_up_tokenization_spaces": true, + "cls_token": "[CLS]", + "do_basic_tokenize": true, + "do_lower_case": false, + "mask_token": "[MASK]", + "max_len": 512, + "model_max_length": 512, + "never_split": null, + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "strip_accents": null, + "tokenize_chinese_chars": true, + "tokenizer_class": "BertTokenizer", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/trainer_state.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..c4e697afb3e3f7fcd762c3096fec180b92217c74 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/trainer_state.json @@ -0,0 +1,1783 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 25.0, + "eval_steps": 500, + "global_step": 6250, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.1, + "grad_norm": 0.9416134357452393, + "learning_rate": 5.319148936170213e-05, + "loss": 0.6912, + "step": 25 + }, + { + "epoch": 0.2, + "grad_norm": 0.9283172488212585, + "learning_rate": 0.00010638297872340425, + "loss": 0.6691, + "step": 50 + }, + { + "epoch": 0.3, + "grad_norm": 0.5870445370674133, + "learning_rate": 0.00015957446808510637, + "loss": 0.6164, + "step": 75 + }, + { + "epoch": 0.4, + "grad_norm": 0.33945509791374207, + "learning_rate": 0.0002127659574468085, + "loss": 0.5453, + "step": 100 + }, + { + "epoch": 0.5, + "grad_norm": 0.21052460372447968, + "learning_rate": 0.00026595744680851064, + "loss": 0.5025, + "step": 125 + }, + { + "epoch": 0.6, + "grad_norm": 0.25948837399482727, + "learning_rate": 0.00031914893617021275, + "loss": 0.4881, + "step": 150 + }, + { + "epoch": 0.7, + "grad_norm": 0.6165992021560669, + "learning_rate": 0.0003723404255319149, + "loss": 0.4852, + "step": 175 + }, + { + "epoch": 0.8, + "grad_norm": 0.33296748995780945, + "learning_rate": 0.0003992081821181128, + "loss": 0.4867, + "step": 200 + }, + { + "epoch": 0.9, + "grad_norm": 0.40745800733566284, + "learning_rate": 0.0003975585615308479, + "loss": 0.4767, + "step": 225 + }, + { + "epoch": 1.0, + "grad_norm": 0.8070425391197205, + "learning_rate": 0.00039590894094358297, + "loss": 0.4784, + "step": 250 + }, + { + "epoch": 1.1, + "grad_norm": 1.2249579429626465, + "learning_rate": 0.0003942593203563181, + "loss": 0.4543, + "step": 275 + }, + { + "epoch": 1.2, + "grad_norm": 1.0143898725509644, + "learning_rate": 0.0003926096997690532, + "loss": 0.4319, + "step": 300 + }, + { + "epoch": 1.3, + "grad_norm": 2.3284952640533447, + "learning_rate": 0.00039096007918178817, + "loss": 0.4455, + "step": 325 + }, + { + "epoch": 1.4, + "grad_norm": 1.189352035522461, + "learning_rate": 0.0003893104585945233, + "loss": 0.4246, + "step": 350 + }, + { + "epoch": 1.5, + "grad_norm": 0.947853684425354, + "learning_rate": 0.0003876608380072583, + "loss": 0.4326, + "step": 375 + }, + { + "epoch": 1.6, + "grad_norm": 0.7130059003829956, + "learning_rate": 0.00038601121741999343, + "loss": 0.4034, + "step": 400 + }, + { + "epoch": 1.7, + "grad_norm": 3.517078161239624, + "learning_rate": 0.0003843615968327285, + "loss": 0.4114, + "step": 425 + }, + { + "epoch": 1.8, + "grad_norm": 0.3982729911804199, + "learning_rate": 0.0003827119762454636, + "loss": 0.4345, + "step": 450 + }, + { + "epoch": 1.9, + "grad_norm": 0.6208440065383911, + "learning_rate": 0.00038106235565819863, + "loss": 0.4135, + "step": 475 + }, + { + "epoch": 2.0, + "grad_norm": 0.9085566401481628, + "learning_rate": 0.0003794127350709337, + "loss": 0.3983, + "step": 500 + }, + { + "epoch": 2.1, + "grad_norm": 1.2225189208984375, + "learning_rate": 0.0003777631144836688, + "loss": 0.3915, + "step": 525 + }, + { + "epoch": 2.2, + "grad_norm": 1.1379530429840088, + "learning_rate": 0.00037611349389640383, + "loss": 0.4335, + "step": 550 + }, + { + "epoch": 2.3, + "grad_norm": 0.8535892367362976, + "learning_rate": 0.00037446387330913894, + "loss": 0.4063, + "step": 575 + }, + { + "epoch": 2.4, + "grad_norm": 1.3587419986724854, + "learning_rate": 0.000372814252721874, + "loss": 0.4065, + "step": 600 + }, + { + "epoch": 2.5, + "grad_norm": 0.9311383962631226, + "learning_rate": 0.00037116463213460903, + "loss": 0.4007, + "step": 625 + }, + { + "epoch": 2.6, + "grad_norm": 0.9912465810775757, + "learning_rate": 0.00036951501154734414, + "loss": 0.3992, + "step": 650 + }, + { + "epoch": 2.7, + "grad_norm": 1.2762119770050049, + "learning_rate": 0.0003678653909600792, + "loss": 0.3827, + "step": 675 + }, + { + "epoch": 2.8, + "grad_norm": 2.3506247997283936, + "learning_rate": 0.0003662157703728143, + "loss": 0.3734, + "step": 700 + }, + { + "epoch": 2.9, + "grad_norm": 1.5375425815582275, + "learning_rate": 0.00036456614978554934, + "loss": 0.4005, + "step": 725 + }, + { + "epoch": 3.0, + "grad_norm": 1.3143396377563477, + "learning_rate": 0.00036291652919828444, + "loss": 0.3548, + "step": 750 + }, + { + "epoch": 3.1, + "grad_norm": 0.662200391292572, + "learning_rate": 0.00036126690861101944, + "loss": 0.4129, + "step": 775 + }, + { + "epoch": 3.2, + "grad_norm": 1.3569878339767456, + "learning_rate": 0.00035961728802375454, + "loss": 0.38, + "step": 800 + }, + { + "epoch": 3.3, + "grad_norm": 0.9888588190078735, + "learning_rate": 0.0003579676674364896, + "loss": 0.3778, + "step": 825 + }, + { + "epoch": 3.4, + "grad_norm": 0.679969310760498, + "learning_rate": 0.0003563180468492247, + "loss": 0.3934, + "step": 850 + }, + { + "epoch": 3.5, + "grad_norm": 1.608539342880249, + "learning_rate": 0.0003546684262619598, + "loss": 0.418, + "step": 875 + }, + { + "epoch": 3.6, + "grad_norm": 1.7037910223007202, + "learning_rate": 0.00035301880567469485, + "loss": 0.3934, + "step": 900 + }, + { + "epoch": 3.7, + "grad_norm": 1.4417132139205933, + "learning_rate": 0.0003513691850874299, + "loss": 0.3947, + "step": 925 + }, + { + "epoch": 3.8, + "grad_norm": 1.1482582092285156, + "learning_rate": 0.00034971956450016495, + "loss": 0.3621, + "step": 950 + }, + { + "epoch": 3.9, + "grad_norm": 2.046616792678833, + "learning_rate": 0.00034806994391290005, + "loss": 0.3774, + "step": 975 + }, + { + "epoch": 4.0, + "grad_norm": 1.0832113027572632, + "learning_rate": 0.0003464203233256351, + "loss": 0.3542, + "step": 1000 + }, + { + "epoch": 4.1, + "grad_norm": 0.8189190626144409, + "learning_rate": 0.0003447707027383702, + "loss": 0.381, + "step": 1025 + }, + { + "epoch": 4.2, + "grad_norm": 1.6974774599075317, + "learning_rate": 0.00034312108215110525, + "loss": 0.3775, + "step": 1050 + }, + { + "epoch": 4.3, + "grad_norm": 1.2468841075897217, + "learning_rate": 0.0003414714615638403, + "loss": 0.3527, + "step": 1075 + }, + { + "epoch": 4.4, + "grad_norm": 0.9022793173789978, + "learning_rate": 0.0003398218409765754, + "loss": 0.3637, + "step": 1100 + }, + { + "epoch": 4.5, + "grad_norm": 1.121774673461914, + "learning_rate": 0.00033817222038931045, + "loss": 0.3935, + "step": 1125 + }, + { + "epoch": 4.6, + "grad_norm": 0.810937762260437, + "learning_rate": 0.00033652259980204556, + "loss": 0.4103, + "step": 1150 + }, + { + "epoch": 4.7, + "grad_norm": 0.7376565337181091, + "learning_rate": 0.0003348729792147806, + "loss": 0.362, + "step": 1175 + }, + { + "epoch": 4.8, + "grad_norm": 1.7271156311035156, + "learning_rate": 0.0003332233586275157, + "loss": 0.3597, + "step": 1200 + }, + { + "epoch": 4.9, + "grad_norm": 1.2273073196411133, + "learning_rate": 0.00033157373804025076, + "loss": 0.3853, + "step": 1225 + }, + { + "epoch": 5.0, + "grad_norm": 0.9998250603675842, + "learning_rate": 0.0003299241174529858, + "loss": 0.3461, + "step": 1250 + }, + { + "epoch": 5.1, + "grad_norm": 0.9112564325332642, + "learning_rate": 0.0003282744968657209, + "loss": 0.3659, + "step": 1275 + }, + { + "epoch": 5.2, + "grad_norm": 3.246051073074341, + "learning_rate": 0.00032662487627845596, + "loss": 0.3488, + "step": 1300 + }, + { + "epoch": 5.3, + "grad_norm": 1.0283201932907104, + "learning_rate": 0.00032497525569119106, + "loss": 0.3528, + "step": 1325 + }, + { + "epoch": 5.4, + "grad_norm": 1.112660527229309, + "learning_rate": 0.0003233256351039261, + "loss": 0.3723, + "step": 1350 + }, + { + "epoch": 5.5, + "grad_norm": 0.9069312810897827, + "learning_rate": 0.00032167601451666116, + "loss": 0.3416, + "step": 1375 + }, + { + "epoch": 5.6, + "grad_norm": 0.710412323474884, + "learning_rate": 0.00032002639392939627, + "loss": 0.3607, + "step": 1400 + }, + { + "epoch": 5.7, + "grad_norm": 1.185612678527832, + "learning_rate": 0.0003183767733421313, + "loss": 0.3777, + "step": 1425 + }, + { + "epoch": 5.8, + "grad_norm": 1.340174913406372, + "learning_rate": 0.0003167271527548664, + "loss": 0.3563, + "step": 1450 + }, + { + "epoch": 5.9, + "grad_norm": 0.9387012124061584, + "learning_rate": 0.00031507753216760147, + "loss": 0.3647, + "step": 1475 + }, + { + "epoch": 6.0, + "grad_norm": 1.1005347967147827, + "learning_rate": 0.00031342791158033657, + "loss": 0.3484, + "step": 1500 + }, + { + "epoch": 6.1, + "grad_norm": 2.64058780670166, + "learning_rate": 0.00031177829099307157, + "loss": 0.3542, + "step": 1525 + }, + { + "epoch": 6.2, + "grad_norm": 1.3007783889770508, + "learning_rate": 0.00031012867040580667, + "loss": 0.3379, + "step": 1550 + }, + { + "epoch": 6.3, + "grad_norm": 3.7190418243408203, + "learning_rate": 0.0003084790498185417, + "loss": 0.3514, + "step": 1575 + }, + { + "epoch": 6.4, + "grad_norm": 1.7263462543487549, + "learning_rate": 0.0003068294292312768, + "loss": 0.3434, + "step": 1600 + }, + { + "epoch": 6.5, + "grad_norm": 3.3360602855682373, + "learning_rate": 0.0003051798086440119, + "loss": 0.3507, + "step": 1625 + }, + { + "epoch": 6.6, + "grad_norm": 1.9620813131332397, + "learning_rate": 0.000303530188056747, + "loss": 0.3455, + "step": 1650 + }, + { + "epoch": 6.7, + "grad_norm": 1.3080660104751587, + "learning_rate": 0.000301880567469482, + "loss": 0.3678, + "step": 1675 + }, + { + "epoch": 6.8, + "grad_norm": 0.6183618307113647, + "learning_rate": 0.0003002309468822171, + "loss": 0.3704, + "step": 1700 + }, + { + "epoch": 6.9, + "grad_norm": 1.0411962270736694, + "learning_rate": 0.0002985813262949522, + "loss": 0.3505, + "step": 1725 + }, + { + "epoch": 7.0, + "grad_norm": 2.040569543838501, + "learning_rate": 0.00029693170570768723, + "loss": 0.3666, + "step": 1750 + }, + { + "epoch": 7.1, + "grad_norm": 0.713647723197937, + "learning_rate": 0.00029528208512042233, + "loss": 0.3529, + "step": 1775 + }, + { + "epoch": 7.2, + "grad_norm": 3.0373058319091797, + "learning_rate": 0.0002936324645331574, + "loss": 0.361, + "step": 1800 + }, + { + "epoch": 7.3, + "grad_norm": 1.819886326789856, + "learning_rate": 0.00029198284394589243, + "loss": 0.3482, + "step": 1825 + }, + { + "epoch": 7.4, + "grad_norm": 0.7930757403373718, + "learning_rate": 0.00029033322335862753, + "loss": 0.3695, + "step": 1850 + }, + { + "epoch": 7.5, + "grad_norm": 0.7889612913131714, + "learning_rate": 0.0002886836027713626, + "loss": 0.3527, + "step": 1875 + }, + { + "epoch": 7.6, + "grad_norm": 2.4007863998413086, + "learning_rate": 0.0002870339821840977, + "loss": 0.3275, + "step": 1900 + }, + { + "epoch": 7.7, + "grad_norm": 2.3483331203460693, + "learning_rate": 0.00028538436159683273, + "loss": 0.3571, + "step": 1925 + }, + { + "epoch": 7.8, + "grad_norm": 0.8911663293838501, + "learning_rate": 0.00028373474100956784, + "loss": 0.3079, + "step": 1950 + }, + { + "epoch": 7.9, + "grad_norm": 1.0521246194839478, + "learning_rate": 0.0002820851204223029, + "loss": 0.3578, + "step": 1975 + }, + { + "epoch": 8.0, + "grad_norm": 1.3923832178115845, + "learning_rate": 0.00028043549983503794, + "loss": 0.3588, + "step": 2000 + }, + { + "epoch": 8.1, + "grad_norm": 1.4007149934768677, + "learning_rate": 0.00027878587924777304, + "loss": 0.3458, + "step": 2025 + }, + { + "epoch": 8.2, + "grad_norm": 1.103198766708374, + "learning_rate": 0.0002771362586605081, + "loss": 0.3736, + "step": 2050 + }, + { + "epoch": 8.3, + "grad_norm": 2.936150550842285, + "learning_rate": 0.0002754866380732432, + "loss": 0.3223, + "step": 2075 + }, + { + "epoch": 8.4, + "grad_norm": 1.907762885093689, + "learning_rate": 0.00027383701748597824, + "loss": 0.3348, + "step": 2100 + }, + { + "epoch": 8.5, + "grad_norm": 1.4422886371612549, + "learning_rate": 0.0002721873968987133, + "loss": 0.3222, + "step": 2125 + }, + { + "epoch": 8.6, + "grad_norm": 1.8203483819961548, + "learning_rate": 0.00027053777631144834, + "loss": 0.3968, + "step": 2150 + }, + { + "epoch": 8.7, + "grad_norm": 0.8046119213104248, + "learning_rate": 0.00026888815572418344, + "loss": 0.3422, + "step": 2175 + }, + { + "epoch": 8.8, + "grad_norm": 1.2612605094909668, + "learning_rate": 0.00026723853513691855, + "loss": 0.3387, + "step": 2200 + }, + { + "epoch": 8.9, + "grad_norm": 1.6534866094589233, + "learning_rate": 0.0002655889145496536, + "loss": 0.3548, + "step": 2225 + }, + { + "epoch": 9.0, + "grad_norm": 1.889664649963379, + "learning_rate": 0.0002639392939623887, + "loss": 0.3367, + "step": 2250 + }, + { + "epoch": 9.1, + "grad_norm": 2.6691551208496094, + "learning_rate": 0.0002622896733751237, + "loss": 0.3628, + "step": 2275 + }, + { + "epoch": 9.2, + "grad_norm": 2.0130255222320557, + "learning_rate": 0.0002606400527878588, + "loss": 0.3366, + "step": 2300 + }, + { + "epoch": 9.3, + "grad_norm": 1.404814600944519, + "learning_rate": 0.00025899043220059385, + "loss": 0.3327, + "step": 2325 + }, + { + "epoch": 9.4, + "grad_norm": 0.9150362610816956, + "learning_rate": 0.00025734081161332895, + "loss": 0.308, + "step": 2350 + }, + { + "epoch": 9.5, + "grad_norm": 1.099320411682129, + "learning_rate": 0.000255691191026064, + "loss": 0.3499, + "step": 2375 + }, + { + "epoch": 9.6, + "grad_norm": 1.502451777458191, + "learning_rate": 0.0002540415704387991, + "loss": 0.3728, + "step": 2400 + }, + { + "epoch": 9.7, + "grad_norm": 1.1115409135818481, + "learning_rate": 0.00025239194985153415, + "loss": 0.348, + "step": 2425 + }, + { + "epoch": 9.8, + "grad_norm": 0.8831422328948975, + "learning_rate": 0.0002507423292642692, + "loss": 0.3392, + "step": 2450 + }, + { + "epoch": 9.9, + "grad_norm": 2.104572057723999, + "learning_rate": 0.0002490927086770043, + "loss": 0.3511, + "step": 2475 + }, + { + "epoch": 10.0, + "grad_norm": 2.4081919193267822, + "learning_rate": 0.00024744308808973936, + "loss": 0.3284, + "step": 2500 + }, + { + "epoch": 10.1, + "grad_norm": 1.9861253499984741, + "learning_rate": 0.00024579346750247446, + "loss": 0.3621, + "step": 2525 + }, + { + "epoch": 10.2, + "grad_norm": 1.9042344093322754, + "learning_rate": 0.0002441438469152095, + "loss": 0.3341, + "step": 2550 + }, + { + "epoch": 10.3, + "grad_norm": 0.5676223635673523, + "learning_rate": 0.00024249422632794456, + "loss": 0.3203, + "step": 2575 + }, + { + "epoch": 10.4, + "grad_norm": 1.5105270147323608, + "learning_rate": 0.00024084460574067963, + "loss": 0.3711, + "step": 2600 + }, + { + "epoch": 10.5, + "grad_norm": 0.9969163537025452, + "learning_rate": 0.0002391949851534147, + "loss": 0.3296, + "step": 2625 + }, + { + "epoch": 10.6, + "grad_norm": 1.297096848487854, + "learning_rate": 0.0002375453645661498, + "loss": 0.349, + "step": 2650 + }, + { + "epoch": 10.7, + "grad_norm": 1.0513092279434204, + "learning_rate": 0.00023589574397888486, + "loss": 0.3308, + "step": 2675 + }, + { + "epoch": 10.8, + "grad_norm": 1.5543969869613647, + "learning_rate": 0.00023424612339161997, + "loss": 0.3289, + "step": 2700 + }, + { + "epoch": 10.9, + "grad_norm": 0.9587257504463196, + "learning_rate": 0.000232596502804355, + "loss": 0.3128, + "step": 2725 + }, + { + "epoch": 11.0, + "grad_norm": 2.1693837642669678, + "learning_rate": 0.00023094688221709007, + "loss": 0.349, + "step": 2750 + }, + { + "epoch": 11.1, + "grad_norm": 1.5624831914901733, + "learning_rate": 0.00022929726162982514, + "loss": 0.3391, + "step": 2775 + }, + { + "epoch": 11.2, + "grad_norm": 1.4600939750671387, + "learning_rate": 0.00022764764104256022, + "loss": 0.3371, + "step": 2800 + }, + { + "epoch": 11.3, + "grad_norm": 1.6139200925827026, + "learning_rate": 0.0002259980204552953, + "loss": 0.3447, + "step": 2825 + }, + { + "epoch": 11.4, + "grad_norm": 2.2703235149383545, + "learning_rate": 0.00022434839986803037, + "loss": 0.3157, + "step": 2850 + }, + { + "epoch": 11.5, + "grad_norm": 2.0282037258148193, + "learning_rate": 0.00022269877928076542, + "loss": 0.3339, + "step": 2875 + }, + { + "epoch": 11.6, + "grad_norm": 1.389129638671875, + "learning_rate": 0.0002210491586935005, + "loss": 0.3195, + "step": 2900 + }, + { + "epoch": 11.7, + "grad_norm": 1.0041213035583496, + "learning_rate": 0.00021939953810623557, + "loss": 0.3382, + "step": 2925 + }, + { + "epoch": 11.8, + "grad_norm": 1.7116957902908325, + "learning_rate": 0.00021774991751897065, + "loss": 0.3261, + "step": 2950 + }, + { + "epoch": 11.9, + "grad_norm": 1.080194354057312, + "learning_rate": 0.00021610029693170573, + "loss": 0.3433, + "step": 2975 + }, + { + "epoch": 12.0, + "grad_norm": 0.6561412811279297, + "learning_rate": 0.0002144506763444408, + "loss": 0.3465, + "step": 3000 + }, + { + "epoch": 12.1, + "grad_norm": 1.5294175148010254, + "learning_rate": 0.00021280105575717582, + "loss": 0.3195, + "step": 3025 + }, + { + "epoch": 12.2, + "grad_norm": 0.6313191652297974, + "learning_rate": 0.00021115143516991093, + "loss": 0.3214, + "step": 3050 + }, + { + "epoch": 12.3, + "grad_norm": 1.4611085653305054, + "learning_rate": 0.000209501814582646, + "loss": 0.3091, + "step": 3075 + }, + { + "epoch": 12.4, + "grad_norm": 1.6590704917907715, + "learning_rate": 0.00020785219399538108, + "loss": 0.3633, + "step": 3100 + }, + { + "epoch": 12.5, + "grad_norm": 0.982976496219635, + "learning_rate": 0.00020620257340811616, + "loss": 0.3688, + "step": 3125 + }, + { + "epoch": 12.6, + "grad_norm": 0.9717169404029846, + "learning_rate": 0.00020455295282085123, + "loss": 0.3532, + "step": 3150 + }, + { + "epoch": 12.7, + "grad_norm": 1.5903713703155518, + "learning_rate": 0.00020290333223358626, + "loss": 0.2906, + "step": 3175 + }, + { + "epoch": 12.8, + "grad_norm": 1.7043166160583496, + "learning_rate": 0.00020125371164632133, + "loss": 0.322, + "step": 3200 + }, + { + "epoch": 12.9, + "grad_norm": 0.7691648006439209, + "learning_rate": 0.0001996040910590564, + "loss": 0.3477, + "step": 3225 + }, + { + "epoch": 13.0, + "grad_norm": 1.1726899147033691, + "learning_rate": 0.00019795447047179148, + "loss": 0.3551, + "step": 3250 + }, + { + "epoch": 13.1, + "grad_norm": 1.0323086977005005, + "learning_rate": 0.0001963048498845266, + "loss": 0.3248, + "step": 3275 + }, + { + "epoch": 13.2, + "grad_norm": 2.920222520828247, + "learning_rate": 0.00019465522929726164, + "loss": 0.3298, + "step": 3300 + }, + { + "epoch": 13.3, + "grad_norm": 1.101644515991211, + "learning_rate": 0.00019300560870999671, + "loss": 0.3011, + "step": 3325 + }, + { + "epoch": 13.4, + "grad_norm": 1.111193299293518, + "learning_rate": 0.0001913559881227318, + "loss": 0.325, + "step": 3350 + }, + { + "epoch": 13.5, + "grad_norm": 1.5074782371520996, + "learning_rate": 0.00018970636753546684, + "loss": 0.3435, + "step": 3375 + }, + { + "epoch": 13.6, + "grad_norm": 1.0705279111862183, + "learning_rate": 0.00018805674694820192, + "loss": 0.3378, + "step": 3400 + }, + { + "epoch": 13.7, + "grad_norm": 1.5590386390686035, + "learning_rate": 0.000186407126360937, + "loss": 0.3349, + "step": 3425 + }, + { + "epoch": 13.8, + "grad_norm": 2.2993671894073486, + "learning_rate": 0.00018475750577367207, + "loss": 0.3479, + "step": 3450 + }, + { + "epoch": 13.9, + "grad_norm": 1.8973578214645386, + "learning_rate": 0.00018310788518640715, + "loss": 0.3358, + "step": 3475 + }, + { + "epoch": 14.0, + "grad_norm": 1.5887267589569092, + "learning_rate": 0.00018145826459914222, + "loss": 0.3095, + "step": 3500 + }, + { + "epoch": 14.1, + "grad_norm": 2.1970202922821045, + "learning_rate": 0.00017980864401187727, + "loss": 0.3435, + "step": 3525 + }, + { + "epoch": 14.2, + "grad_norm": 0.9183377027511597, + "learning_rate": 0.00017815902342461235, + "loss": 0.3454, + "step": 3550 + }, + { + "epoch": 14.3, + "grad_norm": 1.253682017326355, + "learning_rate": 0.00017650940283734742, + "loss": 0.3128, + "step": 3575 + }, + { + "epoch": 14.4, + "grad_norm": 2.4595584869384766, + "learning_rate": 0.00017485978225008247, + "loss": 0.3181, + "step": 3600 + }, + { + "epoch": 14.5, + "grad_norm": 2.0421340465545654, + "learning_rate": 0.00017321016166281755, + "loss": 0.3555, + "step": 3625 + }, + { + "epoch": 14.6, + "grad_norm": 2.0127763748168945, + "learning_rate": 0.00017156054107555263, + "loss": 0.2952, + "step": 3650 + }, + { + "epoch": 14.7, + "grad_norm": 0.9481540322303772, + "learning_rate": 0.0001699109204882877, + "loss": 0.3154, + "step": 3675 + }, + { + "epoch": 14.8, + "grad_norm": 1.5080980062484741, + "learning_rate": 0.00016826129990102278, + "loss": 0.336, + "step": 3700 + }, + { + "epoch": 14.9, + "grad_norm": 1.504712700843811, + "learning_rate": 0.00016661167931375785, + "loss": 0.3511, + "step": 3725 + }, + { + "epoch": 15.0, + "grad_norm": 1.4532283544540405, + "learning_rate": 0.0001649620587264929, + "loss": 0.3216, + "step": 3750 + }, + { + "epoch": 15.1, + "grad_norm": 1.215970754623413, + "learning_rate": 0.00016331243813922798, + "loss": 0.3597, + "step": 3775 + }, + { + "epoch": 15.2, + "grad_norm": 2.054934024810791, + "learning_rate": 0.00016166281755196306, + "loss": 0.3095, + "step": 3800 + }, + { + "epoch": 15.3, + "grad_norm": 1.6162620782852173, + "learning_rate": 0.00016001319696469813, + "loss": 0.3172, + "step": 3825 + }, + { + "epoch": 15.4, + "grad_norm": 2.974353790283203, + "learning_rate": 0.0001583635763774332, + "loss": 0.3186, + "step": 3850 + }, + { + "epoch": 15.5, + "grad_norm": 1.1828879117965698, + "learning_rate": 0.00015671395579016829, + "loss": 0.3141, + "step": 3875 + }, + { + "epoch": 15.6, + "grad_norm": 2.170670509338379, + "learning_rate": 0.00015506433520290334, + "loss": 0.3054, + "step": 3900 + }, + { + "epoch": 15.7, + "grad_norm": 2.5053343772888184, + "learning_rate": 0.0001534147146156384, + "loss": 0.3348, + "step": 3925 + }, + { + "epoch": 15.8, + "grad_norm": 2.761240005493164, + "learning_rate": 0.0001517650940283735, + "loss": 0.3513, + "step": 3950 + }, + { + "epoch": 15.9, + "grad_norm": 2.6656720638275146, + "learning_rate": 0.00015011547344110854, + "loss": 0.3477, + "step": 3975 + }, + { + "epoch": 16.0, + "grad_norm": 1.2667946815490723, + "learning_rate": 0.00014846585285384361, + "loss": 0.3107, + "step": 4000 + }, + { + "epoch": 16.1, + "grad_norm": 1.4039124250411987, + "learning_rate": 0.0001468162322665787, + "loss": 0.3242, + "step": 4025 + }, + { + "epoch": 16.2, + "grad_norm": 1.3851566314697266, + "learning_rate": 0.00014516661167931377, + "loss": 0.3144, + "step": 4050 + }, + { + "epoch": 16.3, + "grad_norm": 1.2302920818328857, + "learning_rate": 0.00014351699109204884, + "loss": 0.3376, + "step": 4075 + }, + { + "epoch": 16.4, + "grad_norm": 1.8713226318359375, + "learning_rate": 0.00014186737050478392, + "loss": 0.3266, + "step": 4100 + }, + { + "epoch": 16.5, + "grad_norm": 1.3579822778701782, + "learning_rate": 0.00014021774991751897, + "loss": 0.33, + "step": 4125 + }, + { + "epoch": 16.6, + "grad_norm": 1.4338854551315308, + "learning_rate": 0.00013856812933025404, + "loss": 0.3218, + "step": 4150 + }, + { + "epoch": 16.7, + "grad_norm": 1.9794092178344727, + "learning_rate": 0.00013691850874298912, + "loss": 0.3375, + "step": 4175 + }, + { + "epoch": 16.8, + "grad_norm": 1.4431006908416748, + "learning_rate": 0.00013526888815572417, + "loss": 0.3177, + "step": 4200 + }, + { + "epoch": 16.9, + "grad_norm": 1.1928157806396484, + "learning_rate": 0.00013361926756845927, + "loss": 0.3227, + "step": 4225 + }, + { + "epoch": 17.0, + "grad_norm": 2.0860953330993652, + "learning_rate": 0.00013196964698119435, + "loss": 0.3345, + "step": 4250 + }, + { + "epoch": 17.1, + "grad_norm": 1.0021557807922363, + "learning_rate": 0.0001303200263939294, + "loss": 0.3299, + "step": 4275 + }, + { + "epoch": 17.2, + "grad_norm": 1.6104000806808472, + "learning_rate": 0.00012867040580666448, + "loss": 0.3066, + "step": 4300 + }, + { + "epoch": 17.3, + "grad_norm": 2.3414576053619385, + "learning_rate": 0.00012702078521939955, + "loss": 0.3474, + "step": 4325 + }, + { + "epoch": 17.4, + "grad_norm": 1.8101857900619507, + "learning_rate": 0.0001253711646321346, + "loss": 0.3022, + "step": 4350 + }, + { + "epoch": 17.5, + "grad_norm": 1.1479285955429077, + "learning_rate": 0.00012372154404486968, + "loss": 0.3133, + "step": 4375 + }, + { + "epoch": 17.6, + "grad_norm": 2.7935454845428467, + "learning_rate": 0.00012207192345760475, + "loss": 0.3027, + "step": 4400 + }, + { + "epoch": 17.7, + "grad_norm": 2.0491445064544678, + "learning_rate": 0.00012042230287033982, + "loss": 0.3451, + "step": 4425 + }, + { + "epoch": 17.8, + "grad_norm": 0.9633632302284241, + "learning_rate": 0.0001187726822830749, + "loss": 0.3171, + "step": 4450 + }, + { + "epoch": 17.9, + "grad_norm": 1.5458674430847168, + "learning_rate": 0.00011712306169580998, + "loss": 0.3011, + "step": 4475 + }, + { + "epoch": 18.0, + "grad_norm": 1.4501872062683105, + "learning_rate": 0.00011547344110854503, + "loss": 0.3185, + "step": 4500 + }, + { + "epoch": 18.1, + "grad_norm": 2.6352083683013916, + "learning_rate": 0.00011382382052128011, + "loss": 0.3191, + "step": 4525 + }, + { + "epoch": 18.2, + "grad_norm": 2.366875410079956, + "learning_rate": 0.00011217419993401519, + "loss": 0.311, + "step": 4550 + }, + { + "epoch": 18.3, + "grad_norm": 1.9184437990188599, + "learning_rate": 0.00011052457934675025, + "loss": 0.2812, + "step": 4575 + }, + { + "epoch": 18.4, + "grad_norm": 1.466262936592102, + "learning_rate": 0.00010887495875948532, + "loss": 0.3075, + "step": 4600 + }, + { + "epoch": 18.5, + "grad_norm": 1.94712233543396, + "learning_rate": 0.0001072253381722204, + "loss": 0.3163, + "step": 4625 + }, + { + "epoch": 18.6, + "grad_norm": 2.5286598205566406, + "learning_rate": 0.00010557571758495546, + "loss": 0.3346, + "step": 4650 + }, + { + "epoch": 18.7, + "grad_norm": 1.604040265083313, + "learning_rate": 0.00010392609699769054, + "loss": 0.3138, + "step": 4675 + }, + { + "epoch": 18.8, + "grad_norm": 1.0606187582015991, + "learning_rate": 0.00010227647641042562, + "loss": 0.3524, + "step": 4700 + }, + { + "epoch": 18.9, + "grad_norm": 1.275098443031311, + "learning_rate": 0.00010062685582316067, + "loss": 0.3057, + "step": 4725 + }, + { + "epoch": 19.0, + "grad_norm": 1.971979022026062, + "learning_rate": 9.897723523589574e-05, + "loss": 0.3089, + "step": 4750 + }, + { + "epoch": 19.1, + "grad_norm": 4.085920810699463, + "learning_rate": 9.732761464863082e-05, + "loss": 0.3251, + "step": 4775 + }, + { + "epoch": 19.2, + "grad_norm": 1.5097062587738037, + "learning_rate": 9.56779940613659e-05, + "loss": 0.2869, + "step": 4800 + }, + { + "epoch": 19.3, + "grad_norm": 2.4897541999816895, + "learning_rate": 9.402837347410096e-05, + "loss": 0.28, + "step": 4825 + }, + { + "epoch": 19.4, + "grad_norm": 2.232405662536621, + "learning_rate": 9.237875288683603e-05, + "loss": 0.3585, + "step": 4850 + }, + { + "epoch": 19.5, + "grad_norm": 2.3231348991394043, + "learning_rate": 9.072913229957111e-05, + "loss": 0.3578, + "step": 4875 + }, + { + "epoch": 19.6, + "grad_norm": 1.1693369150161743, + "learning_rate": 8.907951171230617e-05, + "loss": 0.3201, + "step": 4900 + }, + { + "epoch": 19.7, + "grad_norm": 1.5570422410964966, + "learning_rate": 8.742989112504124e-05, + "loss": 0.3142, + "step": 4925 + }, + { + "epoch": 19.8, + "grad_norm": 1.122075080871582, + "learning_rate": 8.578027053777631e-05, + "loss": 0.3274, + "step": 4950 + }, + { + "epoch": 19.9, + "grad_norm": 1.4783751964569092, + "learning_rate": 8.413064995051139e-05, + "loss": 0.3246, + "step": 4975 + }, + { + "epoch": 20.0, + "grad_norm": 1.6368613243103027, + "learning_rate": 8.248102936324645e-05, + "loss": 0.2984, + "step": 5000 + }, + { + "epoch": 20.1, + "grad_norm": 1.4082392454147339, + "learning_rate": 8.083140877598153e-05, + "loss": 0.3359, + "step": 5025 + }, + { + "epoch": 20.2, + "grad_norm": 1.4183349609375, + "learning_rate": 7.91817881887166e-05, + "loss": 0.327, + "step": 5050 + }, + { + "epoch": 20.3, + "grad_norm": 1.4731121063232422, + "learning_rate": 7.753216760145167e-05, + "loss": 0.3184, + "step": 5075 + }, + { + "epoch": 20.4, + "grad_norm": 2.123331069946289, + "learning_rate": 7.588254701418674e-05, + "loss": 0.3366, + "step": 5100 + }, + { + "epoch": 20.5, + "grad_norm": 1.2433639764785767, + "learning_rate": 7.423292642692181e-05, + "loss": 0.2926, + "step": 5125 + }, + { + "epoch": 20.6, + "grad_norm": 1.4048691987991333, + "learning_rate": 7.258330583965688e-05, + "loss": 0.3067, + "step": 5150 + }, + { + "epoch": 20.7, + "grad_norm": 1.4444036483764648, + "learning_rate": 7.093368525239196e-05, + "loss": 0.2933, + "step": 5175 + }, + { + "epoch": 20.8, + "grad_norm": 1.5258699655532837, + "learning_rate": 6.928406466512702e-05, + "loss": 0.3075, + "step": 5200 + }, + { + "epoch": 20.9, + "grad_norm": 1.6690775156021118, + "learning_rate": 6.763444407786209e-05, + "loss": 0.3002, + "step": 5225 + }, + { + "epoch": 21.0, + "grad_norm": 1.2771008014678955, + "learning_rate": 6.598482349059718e-05, + "loss": 0.3457, + "step": 5250 + }, + { + "epoch": 21.1, + "grad_norm": 0.9566493630409241, + "learning_rate": 6.433520290333224e-05, + "loss": 0.3132, + "step": 5275 + }, + { + "epoch": 21.2, + "grad_norm": 1.8674430847167969, + "learning_rate": 6.26855823160673e-05, + "loss": 0.3158, + "step": 5300 + }, + { + "epoch": 21.3, + "grad_norm": 1.9155611991882324, + "learning_rate": 6.103596172880238e-05, + "loss": 0.3322, + "step": 5325 + }, + { + "epoch": 21.4, + "grad_norm": 1.1892235279083252, + "learning_rate": 5.938634114153745e-05, + "loss": 0.3221, + "step": 5350 + }, + { + "epoch": 21.5, + "grad_norm": 1.3518539667129517, + "learning_rate": 5.7736720554272516e-05, + "loss": 0.2993, + "step": 5375 + }, + { + "epoch": 21.6, + "grad_norm": 1.9076350927352905, + "learning_rate": 5.608709996700759e-05, + "loss": 0.2872, + "step": 5400 + }, + { + "epoch": 21.7, + "grad_norm": 1.7502110004425049, + "learning_rate": 5.443747937974266e-05, + "loss": 0.3233, + "step": 5425 + }, + { + "epoch": 21.8, + "grad_norm": 2.491464853286743, + "learning_rate": 5.278785879247773e-05, + "loss": 0.3307, + "step": 5450 + }, + { + "epoch": 21.9, + "grad_norm": 2.3630316257476807, + "learning_rate": 5.113823820521281e-05, + "loss": 0.3266, + "step": 5475 + }, + { + "epoch": 22.0, + "grad_norm": 1.6543325185775757, + "learning_rate": 4.948861761794787e-05, + "loss": 0.3026, + "step": 5500 + }, + { + "epoch": 22.1, + "grad_norm": 1.8995773792266846, + "learning_rate": 4.783899703068295e-05, + "loss": 0.3158, + "step": 5525 + }, + { + "epoch": 22.2, + "grad_norm": 2.4362375736236572, + "learning_rate": 4.618937644341802e-05, + "loss": 0.3068, + "step": 5550 + }, + { + "epoch": 22.3, + "grad_norm": 2.885927200317383, + "learning_rate": 4.453975585615309e-05, + "loss": 0.3263, + "step": 5575 + }, + { + "epoch": 22.4, + "grad_norm": 3.3690805435180664, + "learning_rate": 4.2890135268888156e-05, + "loss": 0.3186, + "step": 5600 + }, + { + "epoch": 22.5, + "grad_norm": 1.418340802192688, + "learning_rate": 4.1240514681623226e-05, + "loss": 0.2831, + "step": 5625 + }, + { + "epoch": 22.6, + "grad_norm": 2.099191904067993, + "learning_rate": 3.95908940943583e-05, + "loss": 0.3079, + "step": 5650 + }, + { + "epoch": 22.7, + "grad_norm": 3.213982582092285, + "learning_rate": 3.794127350709337e-05, + "loss": 0.2997, + "step": 5675 + }, + { + "epoch": 22.8, + "grad_norm": 1.909794807434082, + "learning_rate": 3.629165291982844e-05, + "loss": 0.3414, + "step": 5700 + }, + { + "epoch": 22.9, + "grad_norm": 1.5416369438171387, + "learning_rate": 3.464203233256351e-05, + "loss": 0.3103, + "step": 5725 + }, + { + "epoch": 23.0, + "grad_norm": 2.4521079063415527, + "learning_rate": 3.299241174529859e-05, + "loss": 0.315, + "step": 5750 + }, + { + "epoch": 23.1, + "grad_norm": 1.9134442806243896, + "learning_rate": 3.134279115803365e-05, + "loss": 0.3141, + "step": 5775 + }, + { + "epoch": 23.2, + "grad_norm": 1.115973711013794, + "learning_rate": 2.9693170570768723e-05, + "loss": 0.319, + "step": 5800 + }, + { + "epoch": 23.3, + "grad_norm": 3.0362348556518555, + "learning_rate": 2.8043549983503796e-05, + "loss": 0.299, + "step": 5825 + }, + { + "epoch": 23.4, + "grad_norm": 1.9920865297317505, + "learning_rate": 2.6393929396238866e-05, + "loss": 0.2653, + "step": 5850 + }, + { + "epoch": 23.5, + "grad_norm": 3.130983591079712, + "learning_rate": 2.4744308808973936e-05, + "loss": 0.3371, + "step": 5875 + }, + { + "epoch": 23.6, + "grad_norm": 2.5702381134033203, + "learning_rate": 2.309468822170901e-05, + "loss": 0.3061, + "step": 5900 + }, + { + "epoch": 23.7, + "grad_norm": 1.6236166954040527, + "learning_rate": 2.1445067634444078e-05, + "loss": 0.3187, + "step": 5925 + }, + { + "epoch": 23.8, + "grad_norm": 1.791316032409668, + "learning_rate": 1.979544704717915e-05, + "loss": 0.3168, + "step": 5950 + }, + { + "epoch": 23.9, + "grad_norm": 2.133009433746338, + "learning_rate": 1.814582645991422e-05, + "loss": 0.3183, + "step": 5975 + }, + { + "epoch": 24.0, + "grad_norm": 1.2213393449783325, + "learning_rate": 1.6496205872649294e-05, + "loss": 0.3336, + "step": 6000 + }, + { + "epoch": 24.1, + "grad_norm": 2.972104072570801, + "learning_rate": 1.4846585285384362e-05, + "loss": 0.3104, + "step": 6025 + }, + { + "epoch": 24.2, + "grad_norm": 2.2249817848205566, + "learning_rate": 1.3196964698119433e-05, + "loss": 0.3187, + "step": 6050 + }, + { + "epoch": 24.3, + "grad_norm": 1.7277551889419556, + "learning_rate": 1.1547344110854504e-05, + "loss": 0.3161, + "step": 6075 + }, + { + "epoch": 24.4, + "grad_norm": 2.0481982231140137, + "learning_rate": 9.897723523589576e-06, + "loss": 0.2824, + "step": 6100 + }, + { + "epoch": 24.5, + "grad_norm": 1.879083514213562, + "learning_rate": 8.248102936324647e-06, + "loss": 0.3276, + "step": 6125 + }, + { + "epoch": 24.6, + "grad_norm": 3.247365713119507, + "learning_rate": 6.5984823490597165e-06, + "loss": 0.32, + "step": 6150 + }, + { + "epoch": 24.7, + "grad_norm": 1.5685606002807617, + "learning_rate": 4.948861761794788e-06, + "loss": 0.3091, + "step": 6175 + }, + { + "epoch": 24.8, + "grad_norm": 1.6192843914031982, + "learning_rate": 3.2992411745298582e-06, + "loss": 0.2968, + "step": 6200 + }, + { + "epoch": 24.9, + "grad_norm": 1.8681565523147583, + "learning_rate": 1.6496205872649291e-06, + "loss": 0.3142, + "step": 6225 + }, + { + "epoch": 25.0, + "grad_norm": 1.358121633529663, + "learning_rate": 0.0, + "loss": 0.3263, + "step": 6250 + } + ], + "logging_steps": 25, + "max_steps": 6250, + "num_input_tokens_seen": 0, + "num_train_epochs": 25, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 331975680000000.0, + "train_batch_size": 20, + "trial_name": null, + "trial_params": null +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/training_args.bin b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..d61f3500eb70763e1f33496c1787bdf3f07a651b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b6e0f7aa622786274a1ea4195b72f4da224050767f7b89529b85721865c5075a +size 5304 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/vocab.txt b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/vocab.txt new file mode 100644 index 0000000000000000000000000000000000000000..cc9ad53732e21ea9180f18e1fb67a38cc5702d4c --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/vocab.txt @@ -0,0 +1,4101 @@ +[PAD] +[UNK] +[CLS] +[SEP] +[MASK] +AAAAAA +AAAAAT +AAAAAC +AAAAAG +AAAATA +AAAATT +AAAATC +AAAATG +AAAACA +AAAACT +AAAACC +AAAACG +AAAAGA +AAAAGT +AAAAGC +AAAAGG +AAATAA +AAATAT +AAATAC +AAATAG +AAATTA +AAATTT +AAATTC +AAATTG +AAATCA +AAATCT +AAATCC +AAATCG +AAATGA +AAATGT +AAATGC +AAATGG +AAACAA +AAACAT +AAACAC +AAACAG +AAACTA +AAACTT +AAACTC +AAACTG +AAACCA +AAACCT +AAACCC +AAACCG +AAACGA +AAACGT +AAACGC +AAACGG +AAAGAA +AAAGAT +AAAGAC +AAAGAG +AAAGTA +AAAGTT +AAAGTC +AAAGTG +AAAGCA +AAAGCT +AAAGCC +AAAGCG +AAAGGA +AAAGGT +AAAGGC +AAAGGG +AATAAA +AATAAT +AATAAC +AATAAG +AATATA +AATATT +AATATC +AATATG +AATACA +AATACT +AATACC +AATACG +AATAGA +AATAGT +AATAGC +AATAGG +AATTAA +AATTAT +AATTAC +AATTAG +AATTTA +AATTTT +AATTTC +AATTTG +AATTCA +AATTCT +AATTCC +AATTCG +AATTGA +AATTGT +AATTGC +AATTGG +AATCAA +AATCAT +AATCAC +AATCAG +AATCTA +AATCTT +AATCTC +AATCTG +AATCCA +AATCCT +AATCCC +AATCCG +AATCGA +AATCGT +AATCGC +AATCGG +AATGAA +AATGAT +AATGAC +AATGAG +AATGTA +AATGTT +AATGTC +AATGTG +AATGCA +AATGCT +AATGCC +AATGCG +AATGGA +AATGGT +AATGGC +AATGGG +AACAAA +AACAAT +AACAAC +AACAAG +AACATA +AACATT +AACATC +AACATG +AACACA +AACACT +AACACC +AACACG +AACAGA +AACAGT +AACAGC +AACAGG +AACTAA +AACTAT +AACTAC +AACTAG +AACTTA +AACTTT +AACTTC +AACTTG +AACTCA +AACTCT +AACTCC +AACTCG +AACTGA +AACTGT +AACTGC +AACTGG +AACCAA +AACCAT +AACCAC +AACCAG +AACCTA +AACCTT +AACCTC +AACCTG +AACCCA +AACCCT +AACCCC +AACCCG +AACCGA +AACCGT +AACCGC +AACCGG +AACGAA +AACGAT +AACGAC +AACGAG +AACGTA +AACGTT +AACGTC +AACGTG +AACGCA +AACGCT +AACGCC +AACGCG +AACGGA +AACGGT +AACGGC +AACGGG +AAGAAA +AAGAAT +AAGAAC +AAGAAG +AAGATA +AAGATT +AAGATC +AAGATG +AAGACA +AAGACT +AAGACC +AAGACG +AAGAGA +AAGAGT +AAGAGC +AAGAGG +AAGTAA +AAGTAT +AAGTAC +AAGTAG +AAGTTA +AAGTTT +AAGTTC +AAGTTG +AAGTCA +AAGTCT +AAGTCC +AAGTCG +AAGTGA +AAGTGT +AAGTGC +AAGTGG +AAGCAA +AAGCAT +AAGCAC +AAGCAG +AAGCTA +AAGCTT +AAGCTC +AAGCTG +AAGCCA +AAGCCT +AAGCCC +AAGCCG +AAGCGA +AAGCGT +AAGCGC +AAGCGG +AAGGAA +AAGGAT +AAGGAC +AAGGAG +AAGGTA +AAGGTT +AAGGTC +AAGGTG +AAGGCA +AAGGCT +AAGGCC +AAGGCG +AAGGGA +AAGGGT +AAGGGC +AAGGGG +ATAAAA +ATAAAT +ATAAAC +ATAAAG +ATAATA +ATAATT +ATAATC +ATAATG +ATAACA +ATAACT +ATAACC +ATAACG +ATAAGA +ATAAGT +ATAAGC +ATAAGG +ATATAA +ATATAT +ATATAC +ATATAG +ATATTA +ATATTT +ATATTC +ATATTG +ATATCA +ATATCT +ATATCC +ATATCG +ATATGA +ATATGT +ATATGC +ATATGG +ATACAA +ATACAT +ATACAC +ATACAG +ATACTA +ATACTT +ATACTC +ATACTG +ATACCA +ATACCT +ATACCC +ATACCG +ATACGA +ATACGT +ATACGC +ATACGG +ATAGAA +ATAGAT +ATAGAC +ATAGAG +ATAGTA +ATAGTT +ATAGTC +ATAGTG +ATAGCA +ATAGCT +ATAGCC +ATAGCG +ATAGGA +ATAGGT +ATAGGC +ATAGGG +ATTAAA +ATTAAT +ATTAAC +ATTAAG +ATTATA +ATTATT +ATTATC +ATTATG +ATTACA +ATTACT +ATTACC +ATTACG +ATTAGA +ATTAGT +ATTAGC +ATTAGG +ATTTAA +ATTTAT +ATTTAC +ATTTAG +ATTTTA +ATTTTT +ATTTTC +ATTTTG +ATTTCA +ATTTCT +ATTTCC +ATTTCG +ATTTGA +ATTTGT +ATTTGC +ATTTGG +ATTCAA +ATTCAT +ATTCAC +ATTCAG +ATTCTA +ATTCTT +ATTCTC +ATTCTG +ATTCCA +ATTCCT +ATTCCC +ATTCCG +ATTCGA +ATTCGT +ATTCGC +ATTCGG +ATTGAA +ATTGAT +ATTGAC +ATTGAG +ATTGTA +ATTGTT +ATTGTC +ATTGTG +ATTGCA +ATTGCT +ATTGCC +ATTGCG +ATTGGA +ATTGGT +ATTGGC +ATTGGG +ATCAAA +ATCAAT +ATCAAC +ATCAAG +ATCATA +ATCATT +ATCATC +ATCATG +ATCACA +ATCACT +ATCACC +ATCACG +ATCAGA +ATCAGT +ATCAGC +ATCAGG +ATCTAA +ATCTAT +ATCTAC +ATCTAG +ATCTTA +ATCTTT +ATCTTC +ATCTTG +ATCTCA +ATCTCT +ATCTCC +ATCTCG +ATCTGA +ATCTGT +ATCTGC +ATCTGG +ATCCAA +ATCCAT +ATCCAC +ATCCAG +ATCCTA +ATCCTT +ATCCTC +ATCCTG +ATCCCA +ATCCCT +ATCCCC +ATCCCG +ATCCGA +ATCCGT +ATCCGC +ATCCGG +ATCGAA +ATCGAT +ATCGAC +ATCGAG +ATCGTA +ATCGTT +ATCGTC +ATCGTG +ATCGCA +ATCGCT +ATCGCC +ATCGCG +ATCGGA +ATCGGT +ATCGGC +ATCGGG +ATGAAA +ATGAAT +ATGAAC +ATGAAG +ATGATA +ATGATT +ATGATC +ATGATG +ATGACA +ATGACT +ATGACC +ATGACG +ATGAGA +ATGAGT +ATGAGC +ATGAGG +ATGTAA +ATGTAT +ATGTAC +ATGTAG +ATGTTA +ATGTTT +ATGTTC +ATGTTG +ATGTCA +ATGTCT +ATGTCC +ATGTCG +ATGTGA +ATGTGT +ATGTGC +ATGTGG +ATGCAA +ATGCAT +ATGCAC +ATGCAG +ATGCTA +ATGCTT +ATGCTC +ATGCTG +ATGCCA +ATGCCT +ATGCCC +ATGCCG +ATGCGA +ATGCGT +ATGCGC +ATGCGG +ATGGAA +ATGGAT +ATGGAC +ATGGAG +ATGGTA +ATGGTT +ATGGTC +ATGGTG +ATGGCA +ATGGCT +ATGGCC +ATGGCG +ATGGGA +ATGGGT +ATGGGC +ATGGGG +ACAAAA +ACAAAT +ACAAAC +ACAAAG +ACAATA +ACAATT +ACAATC +ACAATG +ACAACA +ACAACT +ACAACC +ACAACG +ACAAGA +ACAAGT +ACAAGC +ACAAGG +ACATAA +ACATAT +ACATAC +ACATAG +ACATTA +ACATTT +ACATTC +ACATTG +ACATCA +ACATCT +ACATCC +ACATCG +ACATGA +ACATGT +ACATGC +ACATGG +ACACAA +ACACAT +ACACAC +ACACAG +ACACTA +ACACTT +ACACTC +ACACTG +ACACCA +ACACCT +ACACCC +ACACCG +ACACGA +ACACGT +ACACGC +ACACGG +ACAGAA +ACAGAT +ACAGAC +ACAGAG +ACAGTA +ACAGTT +ACAGTC +ACAGTG +ACAGCA +ACAGCT +ACAGCC +ACAGCG +ACAGGA +ACAGGT +ACAGGC +ACAGGG +ACTAAA +ACTAAT +ACTAAC +ACTAAG +ACTATA +ACTATT +ACTATC +ACTATG +ACTACA +ACTACT +ACTACC +ACTACG +ACTAGA +ACTAGT +ACTAGC +ACTAGG +ACTTAA +ACTTAT +ACTTAC +ACTTAG +ACTTTA +ACTTTT +ACTTTC +ACTTTG +ACTTCA +ACTTCT +ACTTCC +ACTTCG +ACTTGA +ACTTGT +ACTTGC +ACTTGG +ACTCAA +ACTCAT +ACTCAC +ACTCAG +ACTCTA +ACTCTT +ACTCTC +ACTCTG +ACTCCA +ACTCCT +ACTCCC +ACTCCG +ACTCGA +ACTCGT +ACTCGC +ACTCGG +ACTGAA +ACTGAT +ACTGAC +ACTGAG +ACTGTA +ACTGTT +ACTGTC +ACTGTG +ACTGCA +ACTGCT +ACTGCC +ACTGCG +ACTGGA +ACTGGT +ACTGGC +ACTGGG +ACCAAA +ACCAAT +ACCAAC +ACCAAG +ACCATA +ACCATT +ACCATC +ACCATG +ACCACA +ACCACT +ACCACC +ACCACG +ACCAGA +ACCAGT +ACCAGC +ACCAGG +ACCTAA +ACCTAT +ACCTAC +ACCTAG +ACCTTA +ACCTTT +ACCTTC +ACCTTG +ACCTCA +ACCTCT +ACCTCC +ACCTCG +ACCTGA +ACCTGT +ACCTGC +ACCTGG +ACCCAA +ACCCAT +ACCCAC +ACCCAG +ACCCTA +ACCCTT +ACCCTC +ACCCTG +ACCCCA +ACCCCT +ACCCCC +ACCCCG +ACCCGA +ACCCGT +ACCCGC +ACCCGG +ACCGAA +ACCGAT +ACCGAC +ACCGAG +ACCGTA +ACCGTT +ACCGTC +ACCGTG +ACCGCA +ACCGCT +ACCGCC +ACCGCG +ACCGGA +ACCGGT +ACCGGC +ACCGGG +ACGAAA +ACGAAT +ACGAAC +ACGAAG +ACGATA +ACGATT +ACGATC +ACGATG +ACGACA +ACGACT +ACGACC +ACGACG +ACGAGA +ACGAGT +ACGAGC +ACGAGG +ACGTAA +ACGTAT +ACGTAC +ACGTAG +ACGTTA +ACGTTT +ACGTTC +ACGTTG +ACGTCA +ACGTCT +ACGTCC +ACGTCG +ACGTGA +ACGTGT +ACGTGC +ACGTGG +ACGCAA +ACGCAT +ACGCAC +ACGCAG +ACGCTA +ACGCTT +ACGCTC +ACGCTG +ACGCCA +ACGCCT +ACGCCC +ACGCCG +ACGCGA +ACGCGT +ACGCGC +ACGCGG +ACGGAA +ACGGAT +ACGGAC +ACGGAG +ACGGTA +ACGGTT +ACGGTC +ACGGTG +ACGGCA +ACGGCT +ACGGCC +ACGGCG +ACGGGA +ACGGGT +ACGGGC +ACGGGG +AGAAAA +AGAAAT +AGAAAC +AGAAAG +AGAATA +AGAATT +AGAATC +AGAATG +AGAACA +AGAACT +AGAACC +AGAACG +AGAAGA +AGAAGT +AGAAGC +AGAAGG +AGATAA +AGATAT +AGATAC +AGATAG +AGATTA +AGATTT +AGATTC +AGATTG +AGATCA +AGATCT +AGATCC +AGATCG +AGATGA +AGATGT +AGATGC +AGATGG +AGACAA +AGACAT +AGACAC +AGACAG +AGACTA +AGACTT +AGACTC +AGACTG +AGACCA +AGACCT +AGACCC +AGACCG +AGACGA +AGACGT +AGACGC +AGACGG +AGAGAA +AGAGAT +AGAGAC +AGAGAG +AGAGTA +AGAGTT +AGAGTC +AGAGTG +AGAGCA +AGAGCT +AGAGCC +AGAGCG +AGAGGA +AGAGGT +AGAGGC +AGAGGG +AGTAAA +AGTAAT +AGTAAC +AGTAAG +AGTATA +AGTATT +AGTATC +AGTATG +AGTACA +AGTACT +AGTACC +AGTACG +AGTAGA +AGTAGT +AGTAGC +AGTAGG +AGTTAA +AGTTAT +AGTTAC +AGTTAG +AGTTTA +AGTTTT +AGTTTC +AGTTTG +AGTTCA +AGTTCT +AGTTCC +AGTTCG +AGTTGA +AGTTGT +AGTTGC +AGTTGG +AGTCAA +AGTCAT +AGTCAC +AGTCAG +AGTCTA +AGTCTT +AGTCTC +AGTCTG +AGTCCA +AGTCCT +AGTCCC +AGTCCG +AGTCGA +AGTCGT +AGTCGC +AGTCGG +AGTGAA +AGTGAT +AGTGAC +AGTGAG +AGTGTA +AGTGTT +AGTGTC +AGTGTG +AGTGCA +AGTGCT +AGTGCC +AGTGCG +AGTGGA +AGTGGT +AGTGGC +AGTGGG +AGCAAA +AGCAAT +AGCAAC +AGCAAG +AGCATA +AGCATT +AGCATC +AGCATG +AGCACA +AGCACT +AGCACC +AGCACG +AGCAGA +AGCAGT +AGCAGC +AGCAGG +AGCTAA +AGCTAT +AGCTAC +AGCTAG +AGCTTA +AGCTTT +AGCTTC +AGCTTG +AGCTCA +AGCTCT +AGCTCC +AGCTCG +AGCTGA +AGCTGT +AGCTGC +AGCTGG +AGCCAA +AGCCAT +AGCCAC +AGCCAG +AGCCTA +AGCCTT +AGCCTC +AGCCTG +AGCCCA +AGCCCT +AGCCCC +AGCCCG +AGCCGA +AGCCGT +AGCCGC +AGCCGG +AGCGAA +AGCGAT +AGCGAC +AGCGAG +AGCGTA +AGCGTT +AGCGTC +AGCGTG +AGCGCA +AGCGCT +AGCGCC +AGCGCG +AGCGGA +AGCGGT +AGCGGC +AGCGGG +AGGAAA +AGGAAT +AGGAAC +AGGAAG +AGGATA +AGGATT +AGGATC +AGGATG +AGGACA +AGGACT +AGGACC +AGGACG +AGGAGA +AGGAGT +AGGAGC +AGGAGG +AGGTAA +AGGTAT +AGGTAC +AGGTAG +AGGTTA +AGGTTT +AGGTTC +AGGTTG +AGGTCA +AGGTCT +AGGTCC +AGGTCG +AGGTGA +AGGTGT +AGGTGC +AGGTGG +AGGCAA +AGGCAT +AGGCAC +AGGCAG +AGGCTA +AGGCTT +AGGCTC +AGGCTG +AGGCCA +AGGCCT +AGGCCC +AGGCCG +AGGCGA +AGGCGT +AGGCGC +AGGCGG +AGGGAA +AGGGAT +AGGGAC +AGGGAG +AGGGTA +AGGGTT +AGGGTC +AGGGTG +AGGGCA +AGGGCT +AGGGCC +AGGGCG +AGGGGA +AGGGGT +AGGGGC +AGGGGG +TAAAAA +TAAAAT +TAAAAC +TAAAAG +TAAATA +TAAATT +TAAATC +TAAATG +TAAACA +TAAACT +TAAACC +TAAACG +TAAAGA +TAAAGT +TAAAGC +TAAAGG +TAATAA +TAATAT +TAATAC +TAATAG +TAATTA +TAATTT +TAATTC +TAATTG +TAATCA +TAATCT +TAATCC +TAATCG +TAATGA +TAATGT +TAATGC +TAATGG +TAACAA +TAACAT +TAACAC +TAACAG +TAACTA +TAACTT +TAACTC +TAACTG +TAACCA +TAACCT +TAACCC +TAACCG +TAACGA +TAACGT +TAACGC +TAACGG +TAAGAA +TAAGAT +TAAGAC +TAAGAG +TAAGTA +TAAGTT +TAAGTC +TAAGTG +TAAGCA +TAAGCT +TAAGCC +TAAGCG +TAAGGA +TAAGGT +TAAGGC +TAAGGG +TATAAA +TATAAT +TATAAC +TATAAG +TATATA +TATATT +TATATC +TATATG +TATACA +TATACT +TATACC +TATACG +TATAGA +TATAGT +TATAGC +TATAGG +TATTAA +TATTAT +TATTAC +TATTAG +TATTTA +TATTTT +TATTTC +TATTTG +TATTCA +TATTCT +TATTCC +TATTCG +TATTGA +TATTGT +TATTGC +TATTGG +TATCAA +TATCAT +TATCAC +TATCAG +TATCTA +TATCTT +TATCTC +TATCTG +TATCCA +TATCCT +TATCCC +TATCCG +TATCGA +TATCGT +TATCGC +TATCGG +TATGAA +TATGAT +TATGAC +TATGAG +TATGTA +TATGTT +TATGTC +TATGTG +TATGCA +TATGCT +TATGCC +TATGCG +TATGGA +TATGGT +TATGGC +TATGGG +TACAAA +TACAAT +TACAAC +TACAAG +TACATA +TACATT +TACATC +TACATG +TACACA +TACACT +TACACC +TACACG +TACAGA +TACAGT +TACAGC +TACAGG +TACTAA +TACTAT +TACTAC +TACTAG +TACTTA +TACTTT +TACTTC +TACTTG +TACTCA +TACTCT +TACTCC +TACTCG +TACTGA +TACTGT +TACTGC +TACTGG +TACCAA +TACCAT +TACCAC +TACCAG +TACCTA +TACCTT +TACCTC +TACCTG +TACCCA +TACCCT +TACCCC +TACCCG +TACCGA +TACCGT +TACCGC +TACCGG +TACGAA +TACGAT +TACGAC +TACGAG +TACGTA +TACGTT +TACGTC +TACGTG +TACGCA +TACGCT +TACGCC +TACGCG +TACGGA +TACGGT +TACGGC +TACGGG +TAGAAA +TAGAAT +TAGAAC +TAGAAG +TAGATA +TAGATT +TAGATC +TAGATG +TAGACA +TAGACT +TAGACC +TAGACG +TAGAGA +TAGAGT +TAGAGC +TAGAGG +TAGTAA +TAGTAT +TAGTAC +TAGTAG +TAGTTA +TAGTTT +TAGTTC +TAGTTG +TAGTCA +TAGTCT +TAGTCC +TAGTCG +TAGTGA +TAGTGT +TAGTGC +TAGTGG +TAGCAA +TAGCAT +TAGCAC +TAGCAG +TAGCTA +TAGCTT +TAGCTC +TAGCTG +TAGCCA +TAGCCT +TAGCCC +TAGCCG +TAGCGA +TAGCGT +TAGCGC +TAGCGG +TAGGAA +TAGGAT +TAGGAC +TAGGAG +TAGGTA +TAGGTT +TAGGTC +TAGGTG +TAGGCA +TAGGCT +TAGGCC +TAGGCG +TAGGGA +TAGGGT +TAGGGC +TAGGGG +TTAAAA +TTAAAT +TTAAAC +TTAAAG +TTAATA +TTAATT +TTAATC +TTAATG +TTAACA +TTAACT +TTAACC +TTAACG +TTAAGA +TTAAGT +TTAAGC +TTAAGG +TTATAA +TTATAT +TTATAC +TTATAG +TTATTA +TTATTT +TTATTC +TTATTG +TTATCA +TTATCT +TTATCC +TTATCG +TTATGA +TTATGT +TTATGC +TTATGG +TTACAA +TTACAT +TTACAC +TTACAG +TTACTA +TTACTT +TTACTC +TTACTG +TTACCA +TTACCT +TTACCC +TTACCG +TTACGA +TTACGT +TTACGC +TTACGG +TTAGAA +TTAGAT +TTAGAC +TTAGAG +TTAGTA +TTAGTT +TTAGTC +TTAGTG +TTAGCA +TTAGCT +TTAGCC +TTAGCG +TTAGGA +TTAGGT +TTAGGC +TTAGGG +TTTAAA +TTTAAT +TTTAAC +TTTAAG +TTTATA +TTTATT +TTTATC +TTTATG +TTTACA +TTTACT +TTTACC +TTTACG +TTTAGA +TTTAGT +TTTAGC +TTTAGG +TTTTAA +TTTTAT +TTTTAC +TTTTAG +TTTTTA +TTTTTT +TTTTTC +TTTTTG +TTTTCA +TTTTCT +TTTTCC +TTTTCG +TTTTGA +TTTTGT +TTTTGC +TTTTGG +TTTCAA +TTTCAT +TTTCAC +TTTCAG +TTTCTA +TTTCTT +TTTCTC +TTTCTG +TTTCCA +TTTCCT +TTTCCC +TTTCCG +TTTCGA +TTTCGT +TTTCGC +TTTCGG +TTTGAA +TTTGAT +TTTGAC +TTTGAG +TTTGTA +TTTGTT +TTTGTC +TTTGTG +TTTGCA +TTTGCT +TTTGCC +TTTGCG +TTTGGA +TTTGGT +TTTGGC +TTTGGG +TTCAAA +TTCAAT +TTCAAC +TTCAAG +TTCATA +TTCATT +TTCATC +TTCATG +TTCACA +TTCACT +TTCACC +TTCACG +TTCAGA +TTCAGT +TTCAGC +TTCAGG +TTCTAA +TTCTAT +TTCTAC +TTCTAG +TTCTTA +TTCTTT +TTCTTC +TTCTTG +TTCTCA +TTCTCT +TTCTCC +TTCTCG +TTCTGA +TTCTGT +TTCTGC +TTCTGG +TTCCAA +TTCCAT +TTCCAC +TTCCAG +TTCCTA +TTCCTT +TTCCTC +TTCCTG +TTCCCA +TTCCCT +TTCCCC +TTCCCG +TTCCGA +TTCCGT +TTCCGC +TTCCGG +TTCGAA +TTCGAT +TTCGAC +TTCGAG +TTCGTA +TTCGTT +TTCGTC +TTCGTG +TTCGCA +TTCGCT +TTCGCC +TTCGCG +TTCGGA +TTCGGT +TTCGGC +TTCGGG +TTGAAA +TTGAAT +TTGAAC +TTGAAG +TTGATA +TTGATT +TTGATC +TTGATG +TTGACA +TTGACT +TTGACC +TTGACG +TTGAGA +TTGAGT +TTGAGC +TTGAGG +TTGTAA +TTGTAT +TTGTAC +TTGTAG +TTGTTA +TTGTTT +TTGTTC +TTGTTG +TTGTCA +TTGTCT +TTGTCC +TTGTCG +TTGTGA +TTGTGT +TTGTGC +TTGTGG +TTGCAA +TTGCAT +TTGCAC +TTGCAG +TTGCTA +TTGCTT +TTGCTC +TTGCTG +TTGCCA +TTGCCT +TTGCCC +TTGCCG +TTGCGA +TTGCGT +TTGCGC +TTGCGG +TTGGAA +TTGGAT +TTGGAC +TTGGAG +TTGGTA +TTGGTT +TTGGTC +TTGGTG +TTGGCA +TTGGCT +TTGGCC +TTGGCG +TTGGGA +TTGGGT +TTGGGC +TTGGGG +TCAAAA +TCAAAT +TCAAAC +TCAAAG +TCAATA +TCAATT +TCAATC +TCAATG +TCAACA +TCAACT +TCAACC +TCAACG +TCAAGA +TCAAGT +TCAAGC +TCAAGG +TCATAA +TCATAT +TCATAC +TCATAG +TCATTA +TCATTT +TCATTC +TCATTG +TCATCA +TCATCT +TCATCC +TCATCG +TCATGA +TCATGT +TCATGC +TCATGG +TCACAA +TCACAT +TCACAC +TCACAG +TCACTA +TCACTT +TCACTC +TCACTG +TCACCA +TCACCT +TCACCC +TCACCG +TCACGA +TCACGT +TCACGC +TCACGG +TCAGAA +TCAGAT +TCAGAC +TCAGAG +TCAGTA +TCAGTT +TCAGTC +TCAGTG +TCAGCA +TCAGCT +TCAGCC +TCAGCG +TCAGGA +TCAGGT +TCAGGC +TCAGGG +TCTAAA +TCTAAT +TCTAAC +TCTAAG +TCTATA +TCTATT +TCTATC +TCTATG +TCTACA +TCTACT +TCTACC +TCTACG +TCTAGA +TCTAGT +TCTAGC +TCTAGG +TCTTAA +TCTTAT +TCTTAC +TCTTAG +TCTTTA +TCTTTT +TCTTTC +TCTTTG +TCTTCA +TCTTCT +TCTTCC +TCTTCG +TCTTGA +TCTTGT +TCTTGC +TCTTGG +TCTCAA +TCTCAT +TCTCAC +TCTCAG +TCTCTA +TCTCTT +TCTCTC +TCTCTG +TCTCCA +TCTCCT +TCTCCC +TCTCCG +TCTCGA +TCTCGT +TCTCGC +TCTCGG +TCTGAA +TCTGAT +TCTGAC +TCTGAG +TCTGTA +TCTGTT +TCTGTC +TCTGTG +TCTGCA +TCTGCT +TCTGCC +TCTGCG +TCTGGA +TCTGGT +TCTGGC +TCTGGG +TCCAAA +TCCAAT +TCCAAC +TCCAAG +TCCATA +TCCATT +TCCATC +TCCATG +TCCACA +TCCACT +TCCACC +TCCACG +TCCAGA +TCCAGT +TCCAGC +TCCAGG +TCCTAA +TCCTAT +TCCTAC +TCCTAG +TCCTTA +TCCTTT +TCCTTC +TCCTTG +TCCTCA +TCCTCT +TCCTCC +TCCTCG +TCCTGA +TCCTGT +TCCTGC +TCCTGG +TCCCAA +TCCCAT +TCCCAC +TCCCAG +TCCCTA +TCCCTT +TCCCTC +TCCCTG +TCCCCA +TCCCCT +TCCCCC +TCCCCG +TCCCGA +TCCCGT +TCCCGC +TCCCGG +TCCGAA +TCCGAT +TCCGAC +TCCGAG +TCCGTA +TCCGTT +TCCGTC +TCCGTG +TCCGCA +TCCGCT +TCCGCC +TCCGCG +TCCGGA +TCCGGT +TCCGGC +TCCGGG +TCGAAA +TCGAAT +TCGAAC +TCGAAG +TCGATA +TCGATT +TCGATC +TCGATG +TCGACA +TCGACT +TCGACC +TCGACG +TCGAGA +TCGAGT +TCGAGC +TCGAGG +TCGTAA +TCGTAT +TCGTAC +TCGTAG +TCGTTA +TCGTTT +TCGTTC +TCGTTG +TCGTCA +TCGTCT +TCGTCC +TCGTCG +TCGTGA +TCGTGT +TCGTGC +TCGTGG +TCGCAA +TCGCAT +TCGCAC +TCGCAG +TCGCTA +TCGCTT +TCGCTC +TCGCTG +TCGCCA +TCGCCT +TCGCCC +TCGCCG +TCGCGA +TCGCGT +TCGCGC +TCGCGG +TCGGAA +TCGGAT +TCGGAC +TCGGAG +TCGGTA +TCGGTT +TCGGTC +TCGGTG +TCGGCA +TCGGCT +TCGGCC +TCGGCG +TCGGGA +TCGGGT +TCGGGC +TCGGGG +TGAAAA +TGAAAT +TGAAAC +TGAAAG +TGAATA +TGAATT +TGAATC +TGAATG +TGAACA +TGAACT +TGAACC +TGAACG +TGAAGA +TGAAGT +TGAAGC +TGAAGG +TGATAA +TGATAT +TGATAC +TGATAG +TGATTA +TGATTT +TGATTC +TGATTG +TGATCA +TGATCT +TGATCC +TGATCG +TGATGA +TGATGT +TGATGC +TGATGG +TGACAA +TGACAT +TGACAC +TGACAG +TGACTA +TGACTT +TGACTC +TGACTG +TGACCA +TGACCT +TGACCC +TGACCG +TGACGA +TGACGT +TGACGC +TGACGG +TGAGAA +TGAGAT +TGAGAC +TGAGAG +TGAGTA +TGAGTT +TGAGTC +TGAGTG +TGAGCA +TGAGCT +TGAGCC +TGAGCG +TGAGGA +TGAGGT +TGAGGC +TGAGGG +TGTAAA +TGTAAT +TGTAAC +TGTAAG +TGTATA +TGTATT +TGTATC +TGTATG +TGTACA +TGTACT +TGTACC +TGTACG +TGTAGA +TGTAGT +TGTAGC +TGTAGG +TGTTAA +TGTTAT +TGTTAC +TGTTAG +TGTTTA +TGTTTT +TGTTTC +TGTTTG +TGTTCA +TGTTCT +TGTTCC +TGTTCG +TGTTGA +TGTTGT +TGTTGC +TGTTGG +TGTCAA +TGTCAT +TGTCAC +TGTCAG +TGTCTA +TGTCTT +TGTCTC +TGTCTG +TGTCCA +TGTCCT +TGTCCC +TGTCCG +TGTCGA +TGTCGT +TGTCGC +TGTCGG +TGTGAA +TGTGAT +TGTGAC +TGTGAG +TGTGTA +TGTGTT +TGTGTC +TGTGTG +TGTGCA +TGTGCT +TGTGCC +TGTGCG +TGTGGA +TGTGGT +TGTGGC +TGTGGG +TGCAAA +TGCAAT +TGCAAC +TGCAAG +TGCATA +TGCATT +TGCATC +TGCATG +TGCACA +TGCACT +TGCACC +TGCACG +TGCAGA +TGCAGT +TGCAGC +TGCAGG +TGCTAA +TGCTAT +TGCTAC +TGCTAG +TGCTTA +TGCTTT +TGCTTC +TGCTTG +TGCTCA +TGCTCT +TGCTCC +TGCTCG +TGCTGA +TGCTGT +TGCTGC +TGCTGG +TGCCAA +TGCCAT +TGCCAC +TGCCAG +TGCCTA +TGCCTT +TGCCTC +TGCCTG +TGCCCA +TGCCCT +TGCCCC +TGCCCG +TGCCGA +TGCCGT +TGCCGC +TGCCGG +TGCGAA +TGCGAT +TGCGAC +TGCGAG +TGCGTA +TGCGTT +TGCGTC +TGCGTG +TGCGCA +TGCGCT +TGCGCC +TGCGCG +TGCGGA +TGCGGT +TGCGGC +TGCGGG +TGGAAA +TGGAAT +TGGAAC +TGGAAG +TGGATA +TGGATT +TGGATC +TGGATG +TGGACA +TGGACT +TGGACC +TGGACG +TGGAGA +TGGAGT +TGGAGC +TGGAGG +TGGTAA +TGGTAT +TGGTAC +TGGTAG +TGGTTA +TGGTTT +TGGTTC +TGGTTG +TGGTCA +TGGTCT +TGGTCC +TGGTCG +TGGTGA +TGGTGT +TGGTGC +TGGTGG +TGGCAA +TGGCAT +TGGCAC +TGGCAG +TGGCTA +TGGCTT +TGGCTC +TGGCTG +TGGCCA +TGGCCT +TGGCCC +TGGCCG +TGGCGA +TGGCGT +TGGCGC +TGGCGG +TGGGAA +TGGGAT +TGGGAC +TGGGAG +TGGGTA +TGGGTT +TGGGTC +TGGGTG +TGGGCA +TGGGCT +TGGGCC +TGGGCG +TGGGGA +TGGGGT +TGGGGC +TGGGGG +CAAAAA +CAAAAT +CAAAAC +CAAAAG +CAAATA +CAAATT +CAAATC +CAAATG +CAAACA +CAAACT +CAAACC +CAAACG +CAAAGA +CAAAGT +CAAAGC +CAAAGG +CAATAA +CAATAT +CAATAC +CAATAG +CAATTA +CAATTT +CAATTC +CAATTG +CAATCA +CAATCT +CAATCC +CAATCG +CAATGA +CAATGT +CAATGC +CAATGG +CAACAA +CAACAT +CAACAC +CAACAG +CAACTA +CAACTT +CAACTC +CAACTG +CAACCA +CAACCT +CAACCC +CAACCG +CAACGA +CAACGT +CAACGC +CAACGG +CAAGAA +CAAGAT +CAAGAC +CAAGAG +CAAGTA +CAAGTT +CAAGTC +CAAGTG +CAAGCA +CAAGCT +CAAGCC +CAAGCG +CAAGGA +CAAGGT +CAAGGC +CAAGGG +CATAAA +CATAAT +CATAAC +CATAAG +CATATA +CATATT +CATATC +CATATG +CATACA +CATACT +CATACC +CATACG +CATAGA +CATAGT +CATAGC +CATAGG +CATTAA +CATTAT +CATTAC +CATTAG +CATTTA +CATTTT +CATTTC +CATTTG +CATTCA +CATTCT +CATTCC +CATTCG +CATTGA +CATTGT +CATTGC +CATTGG +CATCAA +CATCAT +CATCAC +CATCAG +CATCTA +CATCTT +CATCTC +CATCTG +CATCCA +CATCCT +CATCCC +CATCCG +CATCGA +CATCGT +CATCGC +CATCGG +CATGAA +CATGAT +CATGAC +CATGAG +CATGTA +CATGTT +CATGTC +CATGTG +CATGCA +CATGCT +CATGCC +CATGCG +CATGGA +CATGGT +CATGGC +CATGGG +CACAAA +CACAAT +CACAAC +CACAAG +CACATA +CACATT +CACATC +CACATG +CACACA +CACACT +CACACC +CACACG +CACAGA +CACAGT +CACAGC +CACAGG +CACTAA +CACTAT +CACTAC +CACTAG +CACTTA +CACTTT +CACTTC +CACTTG +CACTCA +CACTCT +CACTCC +CACTCG +CACTGA +CACTGT +CACTGC +CACTGG +CACCAA +CACCAT +CACCAC +CACCAG +CACCTA +CACCTT +CACCTC +CACCTG +CACCCA +CACCCT +CACCCC +CACCCG +CACCGA +CACCGT +CACCGC +CACCGG +CACGAA +CACGAT +CACGAC +CACGAG +CACGTA +CACGTT +CACGTC +CACGTG +CACGCA +CACGCT +CACGCC +CACGCG +CACGGA +CACGGT +CACGGC +CACGGG +CAGAAA +CAGAAT +CAGAAC +CAGAAG +CAGATA +CAGATT +CAGATC +CAGATG +CAGACA +CAGACT +CAGACC +CAGACG +CAGAGA +CAGAGT +CAGAGC +CAGAGG +CAGTAA +CAGTAT +CAGTAC +CAGTAG +CAGTTA +CAGTTT +CAGTTC +CAGTTG +CAGTCA +CAGTCT +CAGTCC +CAGTCG +CAGTGA +CAGTGT +CAGTGC +CAGTGG +CAGCAA +CAGCAT +CAGCAC +CAGCAG +CAGCTA +CAGCTT +CAGCTC +CAGCTG +CAGCCA +CAGCCT +CAGCCC +CAGCCG +CAGCGA +CAGCGT +CAGCGC +CAGCGG +CAGGAA +CAGGAT +CAGGAC +CAGGAG +CAGGTA +CAGGTT +CAGGTC +CAGGTG +CAGGCA +CAGGCT +CAGGCC +CAGGCG +CAGGGA +CAGGGT +CAGGGC +CAGGGG +CTAAAA +CTAAAT +CTAAAC +CTAAAG +CTAATA +CTAATT +CTAATC +CTAATG +CTAACA +CTAACT +CTAACC +CTAACG +CTAAGA +CTAAGT +CTAAGC +CTAAGG +CTATAA +CTATAT +CTATAC +CTATAG +CTATTA +CTATTT +CTATTC +CTATTG +CTATCA +CTATCT +CTATCC +CTATCG +CTATGA +CTATGT +CTATGC +CTATGG +CTACAA +CTACAT +CTACAC +CTACAG +CTACTA +CTACTT +CTACTC +CTACTG +CTACCA +CTACCT +CTACCC +CTACCG +CTACGA +CTACGT +CTACGC +CTACGG +CTAGAA +CTAGAT +CTAGAC +CTAGAG +CTAGTA +CTAGTT +CTAGTC +CTAGTG +CTAGCA +CTAGCT +CTAGCC +CTAGCG +CTAGGA +CTAGGT +CTAGGC +CTAGGG +CTTAAA +CTTAAT +CTTAAC +CTTAAG +CTTATA +CTTATT +CTTATC +CTTATG +CTTACA +CTTACT +CTTACC +CTTACG +CTTAGA +CTTAGT +CTTAGC +CTTAGG +CTTTAA +CTTTAT +CTTTAC +CTTTAG +CTTTTA +CTTTTT +CTTTTC +CTTTTG +CTTTCA +CTTTCT +CTTTCC +CTTTCG +CTTTGA +CTTTGT +CTTTGC +CTTTGG +CTTCAA +CTTCAT +CTTCAC +CTTCAG +CTTCTA +CTTCTT +CTTCTC +CTTCTG +CTTCCA +CTTCCT +CTTCCC +CTTCCG +CTTCGA +CTTCGT +CTTCGC +CTTCGG +CTTGAA +CTTGAT +CTTGAC +CTTGAG +CTTGTA +CTTGTT +CTTGTC +CTTGTG +CTTGCA +CTTGCT +CTTGCC +CTTGCG +CTTGGA +CTTGGT +CTTGGC +CTTGGG +CTCAAA +CTCAAT +CTCAAC +CTCAAG +CTCATA +CTCATT +CTCATC +CTCATG +CTCACA +CTCACT +CTCACC +CTCACG +CTCAGA +CTCAGT +CTCAGC +CTCAGG +CTCTAA +CTCTAT +CTCTAC +CTCTAG +CTCTTA +CTCTTT +CTCTTC +CTCTTG +CTCTCA +CTCTCT +CTCTCC +CTCTCG +CTCTGA +CTCTGT +CTCTGC +CTCTGG +CTCCAA +CTCCAT +CTCCAC +CTCCAG +CTCCTA +CTCCTT +CTCCTC +CTCCTG +CTCCCA +CTCCCT +CTCCCC +CTCCCG +CTCCGA +CTCCGT +CTCCGC +CTCCGG +CTCGAA +CTCGAT +CTCGAC +CTCGAG +CTCGTA +CTCGTT +CTCGTC +CTCGTG +CTCGCA +CTCGCT +CTCGCC +CTCGCG +CTCGGA +CTCGGT +CTCGGC +CTCGGG +CTGAAA +CTGAAT +CTGAAC +CTGAAG +CTGATA +CTGATT +CTGATC +CTGATG +CTGACA +CTGACT +CTGACC +CTGACG +CTGAGA +CTGAGT +CTGAGC +CTGAGG +CTGTAA +CTGTAT +CTGTAC +CTGTAG +CTGTTA +CTGTTT +CTGTTC +CTGTTG +CTGTCA +CTGTCT +CTGTCC +CTGTCG +CTGTGA +CTGTGT +CTGTGC +CTGTGG +CTGCAA +CTGCAT +CTGCAC +CTGCAG +CTGCTA +CTGCTT +CTGCTC +CTGCTG +CTGCCA +CTGCCT +CTGCCC +CTGCCG +CTGCGA +CTGCGT +CTGCGC +CTGCGG +CTGGAA +CTGGAT +CTGGAC +CTGGAG +CTGGTA +CTGGTT +CTGGTC +CTGGTG +CTGGCA +CTGGCT +CTGGCC +CTGGCG +CTGGGA +CTGGGT +CTGGGC +CTGGGG +CCAAAA +CCAAAT +CCAAAC +CCAAAG +CCAATA +CCAATT +CCAATC +CCAATG +CCAACA +CCAACT +CCAACC +CCAACG +CCAAGA +CCAAGT +CCAAGC +CCAAGG +CCATAA +CCATAT +CCATAC +CCATAG +CCATTA +CCATTT +CCATTC +CCATTG +CCATCA +CCATCT +CCATCC +CCATCG +CCATGA +CCATGT +CCATGC +CCATGG +CCACAA +CCACAT +CCACAC +CCACAG +CCACTA +CCACTT +CCACTC +CCACTG +CCACCA +CCACCT +CCACCC +CCACCG +CCACGA +CCACGT +CCACGC +CCACGG +CCAGAA +CCAGAT +CCAGAC +CCAGAG +CCAGTA +CCAGTT +CCAGTC +CCAGTG +CCAGCA +CCAGCT +CCAGCC +CCAGCG +CCAGGA +CCAGGT +CCAGGC +CCAGGG +CCTAAA +CCTAAT +CCTAAC +CCTAAG +CCTATA +CCTATT +CCTATC +CCTATG +CCTACA +CCTACT +CCTACC +CCTACG +CCTAGA +CCTAGT +CCTAGC +CCTAGG +CCTTAA +CCTTAT +CCTTAC +CCTTAG +CCTTTA +CCTTTT +CCTTTC +CCTTTG +CCTTCA +CCTTCT +CCTTCC +CCTTCG +CCTTGA +CCTTGT +CCTTGC +CCTTGG +CCTCAA +CCTCAT +CCTCAC +CCTCAG +CCTCTA +CCTCTT +CCTCTC +CCTCTG +CCTCCA +CCTCCT +CCTCCC +CCTCCG +CCTCGA +CCTCGT +CCTCGC +CCTCGG +CCTGAA +CCTGAT +CCTGAC +CCTGAG +CCTGTA +CCTGTT +CCTGTC +CCTGTG +CCTGCA +CCTGCT +CCTGCC +CCTGCG +CCTGGA +CCTGGT +CCTGGC +CCTGGG +CCCAAA +CCCAAT +CCCAAC +CCCAAG +CCCATA +CCCATT +CCCATC +CCCATG +CCCACA +CCCACT +CCCACC +CCCACG +CCCAGA +CCCAGT +CCCAGC +CCCAGG +CCCTAA +CCCTAT +CCCTAC +CCCTAG +CCCTTA +CCCTTT +CCCTTC +CCCTTG +CCCTCA +CCCTCT +CCCTCC +CCCTCG +CCCTGA +CCCTGT +CCCTGC +CCCTGG +CCCCAA +CCCCAT +CCCCAC +CCCCAG +CCCCTA +CCCCTT +CCCCTC +CCCCTG +CCCCCA +CCCCCT +CCCCCC +CCCCCG +CCCCGA +CCCCGT +CCCCGC +CCCCGG +CCCGAA +CCCGAT +CCCGAC +CCCGAG +CCCGTA +CCCGTT +CCCGTC +CCCGTG +CCCGCA +CCCGCT +CCCGCC +CCCGCG +CCCGGA +CCCGGT +CCCGGC +CCCGGG +CCGAAA +CCGAAT +CCGAAC +CCGAAG +CCGATA +CCGATT +CCGATC +CCGATG +CCGACA +CCGACT +CCGACC +CCGACG +CCGAGA +CCGAGT +CCGAGC +CCGAGG +CCGTAA +CCGTAT +CCGTAC +CCGTAG +CCGTTA +CCGTTT +CCGTTC +CCGTTG +CCGTCA +CCGTCT +CCGTCC +CCGTCG +CCGTGA +CCGTGT +CCGTGC +CCGTGG +CCGCAA +CCGCAT +CCGCAC +CCGCAG +CCGCTA +CCGCTT +CCGCTC +CCGCTG +CCGCCA +CCGCCT +CCGCCC +CCGCCG +CCGCGA +CCGCGT +CCGCGC +CCGCGG +CCGGAA +CCGGAT +CCGGAC +CCGGAG +CCGGTA +CCGGTT +CCGGTC +CCGGTG +CCGGCA +CCGGCT +CCGGCC +CCGGCG +CCGGGA +CCGGGT +CCGGGC +CCGGGG +CGAAAA +CGAAAT +CGAAAC +CGAAAG +CGAATA +CGAATT +CGAATC +CGAATG +CGAACA +CGAACT +CGAACC +CGAACG +CGAAGA +CGAAGT +CGAAGC +CGAAGG +CGATAA +CGATAT +CGATAC +CGATAG +CGATTA +CGATTT +CGATTC +CGATTG +CGATCA +CGATCT +CGATCC +CGATCG +CGATGA +CGATGT +CGATGC +CGATGG +CGACAA +CGACAT +CGACAC +CGACAG +CGACTA +CGACTT +CGACTC +CGACTG +CGACCA +CGACCT +CGACCC +CGACCG +CGACGA +CGACGT +CGACGC +CGACGG +CGAGAA +CGAGAT +CGAGAC +CGAGAG +CGAGTA +CGAGTT +CGAGTC +CGAGTG +CGAGCA +CGAGCT +CGAGCC +CGAGCG +CGAGGA +CGAGGT +CGAGGC +CGAGGG +CGTAAA +CGTAAT +CGTAAC +CGTAAG +CGTATA +CGTATT +CGTATC +CGTATG +CGTACA +CGTACT +CGTACC +CGTACG +CGTAGA +CGTAGT +CGTAGC +CGTAGG +CGTTAA +CGTTAT +CGTTAC +CGTTAG +CGTTTA +CGTTTT +CGTTTC +CGTTTG +CGTTCA +CGTTCT +CGTTCC +CGTTCG +CGTTGA +CGTTGT +CGTTGC +CGTTGG +CGTCAA +CGTCAT +CGTCAC +CGTCAG +CGTCTA +CGTCTT +CGTCTC +CGTCTG +CGTCCA +CGTCCT +CGTCCC +CGTCCG +CGTCGA +CGTCGT +CGTCGC +CGTCGG +CGTGAA +CGTGAT +CGTGAC +CGTGAG +CGTGTA +CGTGTT +CGTGTC +CGTGTG +CGTGCA +CGTGCT +CGTGCC +CGTGCG +CGTGGA +CGTGGT +CGTGGC +CGTGGG +CGCAAA +CGCAAT +CGCAAC +CGCAAG +CGCATA +CGCATT +CGCATC +CGCATG +CGCACA +CGCACT +CGCACC +CGCACG +CGCAGA +CGCAGT +CGCAGC +CGCAGG +CGCTAA +CGCTAT +CGCTAC +CGCTAG +CGCTTA +CGCTTT +CGCTTC +CGCTTG +CGCTCA +CGCTCT +CGCTCC +CGCTCG +CGCTGA +CGCTGT +CGCTGC +CGCTGG +CGCCAA +CGCCAT +CGCCAC +CGCCAG +CGCCTA +CGCCTT +CGCCTC +CGCCTG +CGCCCA +CGCCCT +CGCCCC +CGCCCG +CGCCGA +CGCCGT +CGCCGC +CGCCGG +CGCGAA +CGCGAT +CGCGAC +CGCGAG +CGCGTA +CGCGTT +CGCGTC +CGCGTG +CGCGCA +CGCGCT +CGCGCC +CGCGCG +CGCGGA +CGCGGT +CGCGGC +CGCGGG +CGGAAA +CGGAAT +CGGAAC +CGGAAG +CGGATA +CGGATT +CGGATC +CGGATG +CGGACA +CGGACT +CGGACC +CGGACG +CGGAGA +CGGAGT +CGGAGC +CGGAGG +CGGTAA +CGGTAT +CGGTAC +CGGTAG +CGGTTA +CGGTTT +CGGTTC +CGGTTG +CGGTCA +CGGTCT +CGGTCC +CGGTCG +CGGTGA +CGGTGT +CGGTGC +CGGTGG +CGGCAA +CGGCAT +CGGCAC +CGGCAG +CGGCTA +CGGCTT +CGGCTC +CGGCTG +CGGCCA +CGGCCT +CGGCCC +CGGCCG +CGGCGA +CGGCGT +CGGCGC +CGGCGG +CGGGAA +CGGGAT +CGGGAC +CGGGAG +CGGGTA +CGGGTT +CGGGTC +CGGGTG +CGGGCA +CGGGCT +CGGGCC +CGGGCG +CGGGGA +CGGGGT +CGGGGC +CGGGGG +GAAAAA +GAAAAT +GAAAAC +GAAAAG +GAAATA +GAAATT +GAAATC +GAAATG +GAAACA +GAAACT +GAAACC +GAAACG +GAAAGA +GAAAGT +GAAAGC +GAAAGG +GAATAA +GAATAT +GAATAC +GAATAG +GAATTA +GAATTT +GAATTC +GAATTG +GAATCA +GAATCT +GAATCC +GAATCG +GAATGA +GAATGT +GAATGC +GAATGG +GAACAA +GAACAT +GAACAC +GAACAG +GAACTA +GAACTT +GAACTC +GAACTG +GAACCA +GAACCT +GAACCC +GAACCG +GAACGA +GAACGT +GAACGC +GAACGG +GAAGAA +GAAGAT +GAAGAC +GAAGAG +GAAGTA +GAAGTT +GAAGTC +GAAGTG +GAAGCA +GAAGCT +GAAGCC +GAAGCG +GAAGGA +GAAGGT +GAAGGC +GAAGGG +GATAAA +GATAAT +GATAAC +GATAAG +GATATA +GATATT +GATATC +GATATG +GATACA +GATACT +GATACC +GATACG +GATAGA +GATAGT +GATAGC +GATAGG +GATTAA +GATTAT +GATTAC +GATTAG +GATTTA +GATTTT +GATTTC +GATTTG +GATTCA +GATTCT +GATTCC +GATTCG +GATTGA +GATTGT +GATTGC +GATTGG +GATCAA +GATCAT +GATCAC +GATCAG +GATCTA +GATCTT +GATCTC +GATCTG +GATCCA +GATCCT +GATCCC +GATCCG +GATCGA +GATCGT +GATCGC +GATCGG +GATGAA +GATGAT +GATGAC +GATGAG +GATGTA +GATGTT +GATGTC +GATGTG +GATGCA +GATGCT +GATGCC +GATGCG +GATGGA +GATGGT +GATGGC +GATGGG +GACAAA +GACAAT +GACAAC +GACAAG +GACATA +GACATT +GACATC +GACATG +GACACA +GACACT +GACACC +GACACG +GACAGA +GACAGT +GACAGC +GACAGG +GACTAA +GACTAT +GACTAC +GACTAG +GACTTA +GACTTT +GACTTC +GACTTG +GACTCA +GACTCT +GACTCC +GACTCG +GACTGA +GACTGT +GACTGC +GACTGG +GACCAA +GACCAT +GACCAC +GACCAG +GACCTA +GACCTT +GACCTC +GACCTG +GACCCA +GACCCT +GACCCC +GACCCG +GACCGA +GACCGT +GACCGC +GACCGG +GACGAA +GACGAT +GACGAC +GACGAG +GACGTA +GACGTT +GACGTC +GACGTG +GACGCA +GACGCT +GACGCC +GACGCG +GACGGA +GACGGT +GACGGC +GACGGG +GAGAAA +GAGAAT +GAGAAC +GAGAAG +GAGATA +GAGATT +GAGATC +GAGATG +GAGACA +GAGACT +GAGACC +GAGACG +GAGAGA +GAGAGT +GAGAGC +GAGAGG +GAGTAA +GAGTAT +GAGTAC +GAGTAG +GAGTTA +GAGTTT +GAGTTC +GAGTTG +GAGTCA +GAGTCT +GAGTCC +GAGTCG +GAGTGA +GAGTGT +GAGTGC +GAGTGG +GAGCAA +GAGCAT +GAGCAC +GAGCAG +GAGCTA +GAGCTT +GAGCTC +GAGCTG +GAGCCA +GAGCCT +GAGCCC +GAGCCG +GAGCGA +GAGCGT +GAGCGC +GAGCGG +GAGGAA +GAGGAT +GAGGAC +GAGGAG +GAGGTA +GAGGTT +GAGGTC +GAGGTG +GAGGCA +GAGGCT +GAGGCC +GAGGCG +GAGGGA +GAGGGT +GAGGGC +GAGGGG +GTAAAA +GTAAAT +GTAAAC +GTAAAG +GTAATA +GTAATT +GTAATC +GTAATG +GTAACA +GTAACT +GTAACC +GTAACG +GTAAGA +GTAAGT +GTAAGC +GTAAGG +GTATAA +GTATAT +GTATAC +GTATAG +GTATTA +GTATTT +GTATTC +GTATTG +GTATCA +GTATCT +GTATCC +GTATCG +GTATGA +GTATGT +GTATGC +GTATGG +GTACAA +GTACAT +GTACAC +GTACAG +GTACTA +GTACTT +GTACTC +GTACTG +GTACCA +GTACCT +GTACCC +GTACCG +GTACGA +GTACGT +GTACGC +GTACGG +GTAGAA +GTAGAT +GTAGAC +GTAGAG +GTAGTA +GTAGTT +GTAGTC +GTAGTG +GTAGCA +GTAGCT +GTAGCC +GTAGCG +GTAGGA +GTAGGT +GTAGGC +GTAGGG +GTTAAA +GTTAAT +GTTAAC +GTTAAG +GTTATA +GTTATT +GTTATC +GTTATG +GTTACA +GTTACT +GTTACC +GTTACG +GTTAGA +GTTAGT +GTTAGC +GTTAGG +GTTTAA +GTTTAT +GTTTAC +GTTTAG +GTTTTA +GTTTTT +GTTTTC +GTTTTG +GTTTCA +GTTTCT +GTTTCC +GTTTCG +GTTTGA +GTTTGT +GTTTGC +GTTTGG +GTTCAA +GTTCAT +GTTCAC +GTTCAG +GTTCTA +GTTCTT +GTTCTC +GTTCTG +GTTCCA +GTTCCT +GTTCCC +GTTCCG +GTTCGA +GTTCGT +GTTCGC +GTTCGG +GTTGAA +GTTGAT +GTTGAC +GTTGAG +GTTGTA +GTTGTT +GTTGTC +GTTGTG +GTTGCA +GTTGCT +GTTGCC +GTTGCG +GTTGGA +GTTGGT +GTTGGC +GTTGGG +GTCAAA +GTCAAT +GTCAAC +GTCAAG +GTCATA +GTCATT +GTCATC +GTCATG +GTCACA +GTCACT +GTCACC +GTCACG +GTCAGA +GTCAGT +GTCAGC +GTCAGG +GTCTAA +GTCTAT +GTCTAC +GTCTAG +GTCTTA +GTCTTT +GTCTTC +GTCTTG +GTCTCA +GTCTCT +GTCTCC +GTCTCG +GTCTGA +GTCTGT +GTCTGC +GTCTGG +GTCCAA +GTCCAT +GTCCAC +GTCCAG +GTCCTA +GTCCTT +GTCCTC +GTCCTG +GTCCCA +GTCCCT +GTCCCC +GTCCCG +GTCCGA +GTCCGT +GTCCGC +GTCCGG +GTCGAA +GTCGAT +GTCGAC +GTCGAG +GTCGTA +GTCGTT +GTCGTC +GTCGTG +GTCGCA +GTCGCT +GTCGCC +GTCGCG +GTCGGA +GTCGGT +GTCGGC +GTCGGG +GTGAAA +GTGAAT +GTGAAC +GTGAAG +GTGATA +GTGATT +GTGATC +GTGATG +GTGACA +GTGACT +GTGACC +GTGACG +GTGAGA +GTGAGT +GTGAGC +GTGAGG +GTGTAA +GTGTAT +GTGTAC +GTGTAG +GTGTTA +GTGTTT +GTGTTC +GTGTTG +GTGTCA +GTGTCT +GTGTCC +GTGTCG +GTGTGA +GTGTGT +GTGTGC +GTGTGG +GTGCAA +GTGCAT +GTGCAC +GTGCAG +GTGCTA +GTGCTT +GTGCTC +GTGCTG +GTGCCA +GTGCCT +GTGCCC +GTGCCG +GTGCGA +GTGCGT +GTGCGC +GTGCGG +GTGGAA +GTGGAT +GTGGAC +GTGGAG +GTGGTA +GTGGTT +GTGGTC +GTGGTG +GTGGCA +GTGGCT +GTGGCC +GTGGCG +GTGGGA +GTGGGT +GTGGGC +GTGGGG +GCAAAA +GCAAAT +GCAAAC +GCAAAG +GCAATA +GCAATT +GCAATC +GCAATG +GCAACA +GCAACT +GCAACC +GCAACG +GCAAGA +GCAAGT +GCAAGC +GCAAGG +GCATAA +GCATAT +GCATAC +GCATAG +GCATTA +GCATTT +GCATTC +GCATTG +GCATCA +GCATCT +GCATCC +GCATCG +GCATGA +GCATGT +GCATGC +GCATGG +GCACAA +GCACAT +GCACAC +GCACAG +GCACTA +GCACTT +GCACTC +GCACTG +GCACCA +GCACCT +GCACCC +GCACCG +GCACGA +GCACGT +GCACGC +GCACGG +GCAGAA +GCAGAT +GCAGAC +GCAGAG +GCAGTA +GCAGTT +GCAGTC +GCAGTG +GCAGCA +GCAGCT +GCAGCC +GCAGCG +GCAGGA +GCAGGT +GCAGGC +GCAGGG +GCTAAA +GCTAAT +GCTAAC +GCTAAG +GCTATA +GCTATT +GCTATC +GCTATG +GCTACA +GCTACT +GCTACC +GCTACG +GCTAGA +GCTAGT +GCTAGC +GCTAGG +GCTTAA +GCTTAT +GCTTAC +GCTTAG +GCTTTA +GCTTTT +GCTTTC +GCTTTG +GCTTCA +GCTTCT +GCTTCC +GCTTCG +GCTTGA +GCTTGT +GCTTGC +GCTTGG +GCTCAA +GCTCAT +GCTCAC +GCTCAG +GCTCTA +GCTCTT +GCTCTC +GCTCTG +GCTCCA +GCTCCT +GCTCCC +GCTCCG +GCTCGA +GCTCGT +GCTCGC +GCTCGG +GCTGAA +GCTGAT +GCTGAC +GCTGAG +GCTGTA +GCTGTT +GCTGTC +GCTGTG +GCTGCA +GCTGCT +GCTGCC +GCTGCG +GCTGGA +GCTGGT +GCTGGC +GCTGGG +GCCAAA +GCCAAT +GCCAAC +GCCAAG +GCCATA +GCCATT +GCCATC +GCCATG +GCCACA +GCCACT +GCCACC +GCCACG +GCCAGA +GCCAGT +GCCAGC +GCCAGG +GCCTAA +GCCTAT +GCCTAC +GCCTAG +GCCTTA +GCCTTT +GCCTTC +GCCTTG +GCCTCA +GCCTCT +GCCTCC +GCCTCG +GCCTGA +GCCTGT +GCCTGC +GCCTGG +GCCCAA +GCCCAT +GCCCAC +GCCCAG +GCCCTA +GCCCTT +GCCCTC +GCCCTG +GCCCCA +GCCCCT +GCCCCC +GCCCCG +GCCCGA +GCCCGT +GCCCGC +GCCCGG +GCCGAA +GCCGAT +GCCGAC +GCCGAG +GCCGTA +GCCGTT +GCCGTC +GCCGTG +GCCGCA +GCCGCT +GCCGCC +GCCGCG +GCCGGA +GCCGGT +GCCGGC +GCCGGG +GCGAAA +GCGAAT +GCGAAC +GCGAAG +GCGATA +GCGATT +GCGATC +GCGATG +GCGACA +GCGACT +GCGACC +GCGACG +GCGAGA +GCGAGT +GCGAGC +GCGAGG +GCGTAA +GCGTAT +GCGTAC +GCGTAG +GCGTTA +GCGTTT +GCGTTC +GCGTTG +GCGTCA +GCGTCT +GCGTCC +GCGTCG +GCGTGA +GCGTGT +GCGTGC +GCGTGG +GCGCAA +GCGCAT +GCGCAC +GCGCAG +GCGCTA +GCGCTT +GCGCTC +GCGCTG +GCGCCA +GCGCCT +GCGCCC +GCGCCG +GCGCGA +GCGCGT +GCGCGC +GCGCGG +GCGGAA +GCGGAT +GCGGAC +GCGGAG +GCGGTA +GCGGTT +GCGGTC +GCGGTG +GCGGCA +GCGGCT +GCGGCC +GCGGCG +GCGGGA +GCGGGT +GCGGGC +GCGGGG +GGAAAA +GGAAAT +GGAAAC +GGAAAG +GGAATA +GGAATT +GGAATC +GGAATG +GGAACA +GGAACT +GGAACC +GGAACG +GGAAGA +GGAAGT +GGAAGC +GGAAGG +GGATAA +GGATAT +GGATAC +GGATAG +GGATTA +GGATTT +GGATTC +GGATTG +GGATCA +GGATCT +GGATCC +GGATCG +GGATGA +GGATGT +GGATGC +GGATGG +GGACAA +GGACAT +GGACAC +GGACAG +GGACTA +GGACTT +GGACTC +GGACTG +GGACCA +GGACCT +GGACCC +GGACCG +GGACGA +GGACGT +GGACGC +GGACGG +GGAGAA +GGAGAT +GGAGAC +GGAGAG +GGAGTA +GGAGTT +GGAGTC +GGAGTG +GGAGCA +GGAGCT +GGAGCC +GGAGCG +GGAGGA +GGAGGT +GGAGGC +GGAGGG +GGTAAA +GGTAAT +GGTAAC +GGTAAG +GGTATA +GGTATT +GGTATC +GGTATG +GGTACA +GGTACT +GGTACC +GGTACG +GGTAGA +GGTAGT +GGTAGC +GGTAGG +GGTTAA +GGTTAT +GGTTAC +GGTTAG +GGTTTA +GGTTTT +GGTTTC +GGTTTG +GGTTCA +GGTTCT +GGTTCC +GGTTCG +GGTTGA +GGTTGT +GGTTGC +GGTTGG +GGTCAA +GGTCAT +GGTCAC +GGTCAG +GGTCTA +GGTCTT +GGTCTC +GGTCTG +GGTCCA +GGTCCT +GGTCCC +GGTCCG +GGTCGA +GGTCGT +GGTCGC +GGTCGG +GGTGAA +GGTGAT +GGTGAC +GGTGAG +GGTGTA +GGTGTT +GGTGTC +GGTGTG +GGTGCA +GGTGCT +GGTGCC +GGTGCG +GGTGGA +GGTGGT +GGTGGC +GGTGGG +GGCAAA +GGCAAT +GGCAAC +GGCAAG +GGCATA +GGCATT +GGCATC +GGCATG +GGCACA +GGCACT +GGCACC +GGCACG +GGCAGA +GGCAGT +GGCAGC +GGCAGG +GGCTAA +GGCTAT +GGCTAC +GGCTAG +GGCTTA +GGCTTT +GGCTTC +GGCTTG +GGCTCA +GGCTCT +GGCTCC +GGCTCG +GGCTGA +GGCTGT +GGCTGC +GGCTGG +GGCCAA +GGCCAT +GGCCAC +GGCCAG +GGCCTA +GGCCTT +GGCCTC +GGCCTG +GGCCCA +GGCCCT +GGCCCC +GGCCCG +GGCCGA +GGCCGT +GGCCGC +GGCCGG +GGCGAA +GGCGAT +GGCGAC +GGCGAG +GGCGTA +GGCGTT +GGCGTC +GGCGTG +GGCGCA +GGCGCT +GGCGCC +GGCGCG +GGCGGA +GGCGGT +GGCGGC +GGCGGG +GGGAAA +GGGAAT +GGGAAC +GGGAAG +GGGATA +GGGATT +GGGATC +GGGATG +GGGACA +GGGACT +GGGACC +GGGACG +GGGAGA +GGGAGT +GGGAGC +GGGAGG +GGGTAA +GGGTAT +GGGTAC +GGGTAG +GGGTTA +GGGTTT +GGGTTC +GGGTTG +GGGTCA +GGGTCT +GGGTCC +GGGTCG +GGGTGA +GGGTGT +GGGTGC +GGGTGG +GGGCAA +GGGCAT +GGGCAC +GGGCAG +GGGCTA +GGGCTT +GGGCTC +GGGCTG +GGGCCA +GGGCCT +GGGCCC +GGGCCG +GGGCGA +GGGCGT +GGGCGC +GGGCGG +GGGGAA +GGGGAT +GGGGAC +GGGGAG +GGGGTA +GGGGTT +GGGGTC +GGGGTG +GGGGCA +GGGGCT +GGGGCC +GGGGCG +GGGGGA +GGGGGT +GGGGGC +GGGGGG diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-1000/config.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..9b14e61d327bf586fcff6e817cf9d365bae18ad2 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-1000/config.json @@ -0,0 +1,26 @@ +{ + "_name_or_path": "fabihamakhdoomi/TinyDNABERT", + "architectures": [ + "BertForSequenceClassification" + ], + "attention_probs_dropout_prob": 0.1, + "classifier_dropout": null, + "hidden_act": "gelu", + "hidden_dropout_prob": 0.1, + "hidden_size": 128, + "initializer_range": 0.02, + "intermediate_size": 512, + "layer_norm_eps": 1e-12, + "max_position_embeddings": 512, + "model_type": "bert", + "num_attention_heads": 2, + "num_hidden_layers": 2, + "pad_token_id": 0, + "position_embedding_type": "absolute", + "problem_type": "single_label_classification", + "torch_dtype": "float32", + "transformers_version": "4.44.0", + "type_vocab_size": 2, + "use_cache": true, + "vocab_size": 4101 +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-1000/model.safetensors b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-1000/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..f17be0203cd91be066913fd2267394445723260d --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-1000/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a4cbd05a53c9d7135c0521b5cdc2fe8cadb99ed586771b58bfdad68ee655b7df +size 4021688 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-1000/optimizer.pt b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-1000/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..34776bea776706c0d2df0451da49cee2c5cc2aaf --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-1000/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e13079234deaa2961bd228dc45e3c3dab60a397c92fc468ca385a6f4067af89f +size 8058700 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-1000/rng_state.pth b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-1000/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..031c7b1895cd8940e942badcb833c16894800a59 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-1000/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e570743f570544924d9d30f7a79905faa2dba99076a5c64ec39e4f6a500cf60e +size 14244 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-1000/scheduler.pt b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-1000/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..e21fdaa14c288938c4b2c1277194110b759d8724 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-1000/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7480edaf7ecb5003556fedbf146cf26ec75a9ee19931ce031bbb3b509f9927ce +size 1064 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-1000/special_tokens_map.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-1000/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..a8b3208c2884c4efb86e49300fdd3dc877220cdf --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-1000/special_tokens_map.json @@ -0,0 +1,7 @@ +{ + "cls_token": "[CLS]", + "mask_token": "[MASK]", + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-1000/tokenizer.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-1000/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..ebd782de6963b9032ce9374cfe30f5275be1a30b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-1000/tokenizer.json @@ -0,0 +1,4265 @@ +{ + "version": "1.0", + "truncation": { + "direction": "Right", + "max_length": 512, + "strategy": "LongestFirst", + "stride": 0 + }, + "padding": { + "strategy": { + "Fixed": 512 + }, + "direction": "Right", + "pad_to_multiple_of": null, + "pad_id": 0, + "pad_type_id": 0, + "pad_token": "[PAD]" + }, + "added_tokens": [ + { + "id": 0, + "content": "[PAD]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 1, + "content": "[UNK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 2, + "content": "[CLS]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 3, + "content": "[SEP]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 4, + "content": "[MASK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + } + ], + "normalizer": { + "type": "BertNormalizer", + "clean_text": true, + "handle_chinese_chars": true, + "strip_accents": null, + "lowercase": false + }, + "pre_tokenizer": { + "type": "BertPreTokenizer" + }, + "post_processor": { + "type": "TemplateProcessing", + "single": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + } + ], + "pair": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "B", + "type_id": 1 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 1 + } + } + ], + "special_tokens": { + "[CLS]": { + "id": "[CLS]", + "ids": [ + 2 + ], + "tokens": [ + "[CLS]" + ] + }, + "[SEP]": { + "id": "[SEP]", + "ids": [ + 3 + ], + "tokens": [ + "[SEP]" + ] + } + } + }, + "decoder": { + "type": "WordPiece", + "prefix": "##", + "cleanup": true + }, + "model": { + "type": "WordPiece", + "unk_token": "[UNK]", + "continuing_subword_prefix": "##", + "max_input_chars_per_word": 100, + "vocab": { + "[PAD]": 0, + "[UNK]": 1, + "[CLS]": 2, + "[SEP]": 3, + "[MASK]": 4, + "AAAAAA": 5, + "AAAAAT": 6, + "AAAAAC": 7, + "AAAAAG": 8, + "AAAATA": 9, + "AAAATT": 10, + "AAAATC": 11, + "AAAATG": 12, + "AAAACA": 13, + "AAAACT": 14, + "AAAACC": 15, + "AAAACG": 16, + "AAAAGA": 17, + "AAAAGT": 18, + "AAAAGC": 19, + "AAAAGG": 20, + "AAATAA": 21, + "AAATAT": 22, + "AAATAC": 23, + "AAATAG": 24, + "AAATTA": 25, + "AAATTT": 26, + "AAATTC": 27, + "AAATTG": 28, + "AAATCA": 29, + "AAATCT": 30, + "AAATCC": 31, + "AAATCG": 32, + "AAATGA": 33, + "AAATGT": 34, + "AAATGC": 35, + "AAATGG": 36, + "AAACAA": 37, + "AAACAT": 38, + "AAACAC": 39, + "AAACAG": 40, + "AAACTA": 41, + "AAACTT": 42, + "AAACTC": 43, + "AAACTG": 44, + "AAACCA": 45, + "AAACCT": 46, + "AAACCC": 47, + "AAACCG": 48, + "AAACGA": 49, + "AAACGT": 50, + "AAACGC": 51, + "AAACGG": 52, + "AAAGAA": 53, + "AAAGAT": 54, + "AAAGAC": 55, + "AAAGAG": 56, + "AAAGTA": 57, + "AAAGTT": 58, + "AAAGTC": 59, + "AAAGTG": 60, + "AAAGCA": 61, + "AAAGCT": 62, + "AAAGCC": 63, + "AAAGCG": 64, + "AAAGGA": 65, + "AAAGGT": 66, + "AAAGGC": 67, + "AAAGGG": 68, + "AATAAA": 69, + "AATAAT": 70, + "AATAAC": 71, + "AATAAG": 72, + "AATATA": 73, + "AATATT": 74, + "AATATC": 75, + "AATATG": 76, + "AATACA": 77, + "AATACT": 78, + "AATACC": 79, + "AATACG": 80, + "AATAGA": 81, + "AATAGT": 82, + "AATAGC": 83, + "AATAGG": 84, + "AATTAA": 85, + "AATTAT": 86, + "AATTAC": 87, + "AATTAG": 88, + "AATTTA": 89, + "AATTTT": 90, + "AATTTC": 91, + "AATTTG": 92, + "AATTCA": 93, + "AATTCT": 94, + "AATTCC": 95, + "AATTCG": 96, + "AATTGA": 97, + "AATTGT": 98, + "AATTGC": 99, + "AATTGG": 100, + "AATCAA": 101, + "AATCAT": 102, + "AATCAC": 103, + "AATCAG": 104, + "AATCTA": 105, + "AATCTT": 106, + "AATCTC": 107, + "AATCTG": 108, + "AATCCA": 109, + "AATCCT": 110, + "AATCCC": 111, + "AATCCG": 112, + "AATCGA": 113, + "AATCGT": 114, + "AATCGC": 115, + "AATCGG": 116, + "AATGAA": 117, + "AATGAT": 118, + "AATGAC": 119, + "AATGAG": 120, + "AATGTA": 121, + "AATGTT": 122, + "AATGTC": 123, + "AATGTG": 124, + "AATGCA": 125, + "AATGCT": 126, + "AATGCC": 127, + "AATGCG": 128, + "AATGGA": 129, + "AATGGT": 130, + "AATGGC": 131, + "AATGGG": 132, + "AACAAA": 133, + "AACAAT": 134, + "AACAAC": 135, + "AACAAG": 136, + "AACATA": 137, + "AACATT": 138, + "AACATC": 139, + "AACATG": 140, + "AACACA": 141, + "AACACT": 142, + "AACACC": 143, + "AACACG": 144, + "AACAGA": 145, + "AACAGT": 146, + "AACAGC": 147, + "AACAGG": 148, + "AACTAA": 149, + "AACTAT": 150, + "AACTAC": 151, + "AACTAG": 152, + "AACTTA": 153, + "AACTTT": 154, + "AACTTC": 155, + "AACTTG": 156, + "AACTCA": 157, + "AACTCT": 158, + "AACTCC": 159, + "AACTCG": 160, + "AACTGA": 161, + "AACTGT": 162, + "AACTGC": 163, + "AACTGG": 164, + "AACCAA": 165, + "AACCAT": 166, + "AACCAC": 167, + "AACCAG": 168, + "AACCTA": 169, + "AACCTT": 170, + "AACCTC": 171, + "AACCTG": 172, + "AACCCA": 173, + "AACCCT": 174, + "AACCCC": 175, + "AACCCG": 176, + "AACCGA": 177, + "AACCGT": 178, + "AACCGC": 179, + "AACCGG": 180, + "AACGAA": 181, + "AACGAT": 182, + "AACGAC": 183, + "AACGAG": 184, + "AACGTA": 185, + "AACGTT": 186, + "AACGTC": 187, + "AACGTG": 188, + "AACGCA": 189, + "AACGCT": 190, + "AACGCC": 191, + "AACGCG": 192, + "AACGGA": 193, + "AACGGT": 194, + "AACGGC": 195, + "AACGGG": 196, + "AAGAAA": 197, + "AAGAAT": 198, + "AAGAAC": 199, + "AAGAAG": 200, + "AAGATA": 201, + "AAGATT": 202, + "AAGATC": 203, + "AAGATG": 204, + "AAGACA": 205, + "AAGACT": 206, + "AAGACC": 207, + "AAGACG": 208, + "AAGAGA": 209, + "AAGAGT": 210, + "AAGAGC": 211, + "AAGAGG": 212, + "AAGTAA": 213, + "AAGTAT": 214, + "AAGTAC": 215, + "AAGTAG": 216, + "AAGTTA": 217, + "AAGTTT": 218, + "AAGTTC": 219, + "AAGTTG": 220, + "AAGTCA": 221, + "AAGTCT": 222, + "AAGTCC": 223, + "AAGTCG": 224, + "AAGTGA": 225, + "AAGTGT": 226, + "AAGTGC": 227, + "AAGTGG": 228, + "AAGCAA": 229, + "AAGCAT": 230, + "AAGCAC": 231, + "AAGCAG": 232, + "AAGCTA": 233, + "AAGCTT": 234, + "AAGCTC": 235, + "AAGCTG": 236, + "AAGCCA": 237, + "AAGCCT": 238, + "AAGCCC": 239, + "AAGCCG": 240, + "AAGCGA": 241, + "AAGCGT": 242, + "AAGCGC": 243, + "AAGCGG": 244, + "AAGGAA": 245, + "AAGGAT": 246, + "AAGGAC": 247, + "AAGGAG": 248, + "AAGGTA": 249, + "AAGGTT": 250, + "AAGGTC": 251, + "AAGGTG": 252, + "AAGGCA": 253, + "AAGGCT": 254, + "AAGGCC": 255, + "AAGGCG": 256, + "AAGGGA": 257, + "AAGGGT": 258, + "AAGGGC": 259, + "AAGGGG": 260, + "ATAAAA": 261, + "ATAAAT": 262, + "ATAAAC": 263, + "ATAAAG": 264, + "ATAATA": 265, + "ATAATT": 266, + "ATAATC": 267, + "ATAATG": 268, + "ATAACA": 269, + "ATAACT": 270, + "ATAACC": 271, + "ATAACG": 272, + "ATAAGA": 273, + "ATAAGT": 274, + "ATAAGC": 275, + "ATAAGG": 276, + "ATATAA": 277, + "ATATAT": 278, + "ATATAC": 279, + "ATATAG": 280, + "ATATTA": 281, + "ATATTT": 282, + "ATATTC": 283, + "ATATTG": 284, + "ATATCA": 285, + "ATATCT": 286, + "ATATCC": 287, + "ATATCG": 288, + "ATATGA": 289, + "ATATGT": 290, + "ATATGC": 291, + "ATATGG": 292, + "ATACAA": 293, + "ATACAT": 294, + "ATACAC": 295, + "ATACAG": 296, + "ATACTA": 297, + "ATACTT": 298, + "ATACTC": 299, + "ATACTG": 300, + "ATACCA": 301, + "ATACCT": 302, + "ATACCC": 303, + "ATACCG": 304, + "ATACGA": 305, + "ATACGT": 306, + "ATACGC": 307, + "ATACGG": 308, + "ATAGAA": 309, + "ATAGAT": 310, + "ATAGAC": 311, + "ATAGAG": 312, + "ATAGTA": 313, + "ATAGTT": 314, + "ATAGTC": 315, + "ATAGTG": 316, + "ATAGCA": 317, + "ATAGCT": 318, + "ATAGCC": 319, + "ATAGCG": 320, + "ATAGGA": 321, + "ATAGGT": 322, + "ATAGGC": 323, + "ATAGGG": 324, + "ATTAAA": 325, + "ATTAAT": 326, + "ATTAAC": 327, + "ATTAAG": 328, + "ATTATA": 329, + "ATTATT": 330, + "ATTATC": 331, + "ATTATG": 332, + "ATTACA": 333, + "ATTACT": 334, + "ATTACC": 335, + "ATTACG": 336, + "ATTAGA": 337, + "ATTAGT": 338, + "ATTAGC": 339, + "ATTAGG": 340, + "ATTTAA": 341, + "ATTTAT": 342, + "ATTTAC": 343, + "ATTTAG": 344, + "ATTTTA": 345, + "ATTTTT": 346, + "ATTTTC": 347, + "ATTTTG": 348, + "ATTTCA": 349, + "ATTTCT": 350, + "ATTTCC": 351, + "ATTTCG": 352, + "ATTTGA": 353, + "ATTTGT": 354, + "ATTTGC": 355, + "ATTTGG": 356, + "ATTCAA": 357, + "ATTCAT": 358, + "ATTCAC": 359, + "ATTCAG": 360, + "ATTCTA": 361, + "ATTCTT": 362, + "ATTCTC": 363, + "ATTCTG": 364, + "ATTCCA": 365, + "ATTCCT": 366, + "ATTCCC": 367, + "ATTCCG": 368, + "ATTCGA": 369, + "ATTCGT": 370, + "ATTCGC": 371, + "ATTCGG": 372, + "ATTGAA": 373, + "ATTGAT": 374, + "ATTGAC": 375, + "ATTGAG": 376, + "ATTGTA": 377, + "ATTGTT": 378, + "ATTGTC": 379, + "ATTGTG": 380, + "ATTGCA": 381, + "ATTGCT": 382, + "ATTGCC": 383, + "ATTGCG": 384, + "ATTGGA": 385, + "ATTGGT": 386, + "ATTGGC": 387, + "ATTGGG": 388, + "ATCAAA": 389, + "ATCAAT": 390, + "ATCAAC": 391, + "ATCAAG": 392, + "ATCATA": 393, + "ATCATT": 394, + "ATCATC": 395, + "ATCATG": 396, + "ATCACA": 397, + "ATCACT": 398, + "ATCACC": 399, + "ATCACG": 400, + "ATCAGA": 401, + "ATCAGT": 402, + "ATCAGC": 403, + "ATCAGG": 404, + "ATCTAA": 405, + "ATCTAT": 406, + "ATCTAC": 407, + "ATCTAG": 408, + "ATCTTA": 409, + "ATCTTT": 410, + "ATCTTC": 411, + "ATCTTG": 412, + "ATCTCA": 413, + "ATCTCT": 414, + "ATCTCC": 415, + "ATCTCG": 416, + "ATCTGA": 417, + "ATCTGT": 418, + "ATCTGC": 419, + "ATCTGG": 420, + "ATCCAA": 421, + "ATCCAT": 422, + "ATCCAC": 423, + "ATCCAG": 424, + "ATCCTA": 425, + "ATCCTT": 426, + "ATCCTC": 427, + "ATCCTG": 428, + "ATCCCA": 429, + "ATCCCT": 430, + "ATCCCC": 431, + "ATCCCG": 432, + "ATCCGA": 433, + "ATCCGT": 434, + "ATCCGC": 435, + "ATCCGG": 436, + "ATCGAA": 437, + "ATCGAT": 438, + "ATCGAC": 439, + "ATCGAG": 440, + "ATCGTA": 441, + "ATCGTT": 442, + "ATCGTC": 443, + "ATCGTG": 444, + "ATCGCA": 445, + "ATCGCT": 446, + "ATCGCC": 447, + "ATCGCG": 448, + "ATCGGA": 449, + "ATCGGT": 450, + "ATCGGC": 451, + "ATCGGG": 452, + "ATGAAA": 453, + "ATGAAT": 454, + "ATGAAC": 455, + "ATGAAG": 456, + "ATGATA": 457, + "ATGATT": 458, + "ATGATC": 459, + "ATGATG": 460, + "ATGACA": 461, + "ATGACT": 462, + "ATGACC": 463, + "ATGACG": 464, + "ATGAGA": 465, + "ATGAGT": 466, + "ATGAGC": 467, + "ATGAGG": 468, + "ATGTAA": 469, + "ATGTAT": 470, + "ATGTAC": 471, + "ATGTAG": 472, + "ATGTTA": 473, + "ATGTTT": 474, + "ATGTTC": 475, + "ATGTTG": 476, + "ATGTCA": 477, + "ATGTCT": 478, + "ATGTCC": 479, + "ATGTCG": 480, + "ATGTGA": 481, + "ATGTGT": 482, + "ATGTGC": 483, + "ATGTGG": 484, + "ATGCAA": 485, + "ATGCAT": 486, + "ATGCAC": 487, + "ATGCAG": 488, + "ATGCTA": 489, + "ATGCTT": 490, + "ATGCTC": 491, + "ATGCTG": 492, + "ATGCCA": 493, + "ATGCCT": 494, + "ATGCCC": 495, + "ATGCCG": 496, + "ATGCGA": 497, + "ATGCGT": 498, + "ATGCGC": 499, + "ATGCGG": 500, + "ATGGAA": 501, + "ATGGAT": 502, + "ATGGAC": 503, + "ATGGAG": 504, + "ATGGTA": 505, + "ATGGTT": 506, + "ATGGTC": 507, + "ATGGTG": 508, + "ATGGCA": 509, + "ATGGCT": 510, + "ATGGCC": 511, + "ATGGCG": 512, + "ATGGGA": 513, + "ATGGGT": 514, + "ATGGGC": 515, + "ATGGGG": 516, + "ACAAAA": 517, + "ACAAAT": 518, + "ACAAAC": 519, + "ACAAAG": 520, + "ACAATA": 521, + "ACAATT": 522, + "ACAATC": 523, + "ACAATG": 524, + "ACAACA": 525, + "ACAACT": 526, + "ACAACC": 527, + "ACAACG": 528, + "ACAAGA": 529, + "ACAAGT": 530, + "ACAAGC": 531, + "ACAAGG": 532, + "ACATAA": 533, + "ACATAT": 534, + "ACATAC": 535, + "ACATAG": 536, + "ACATTA": 537, + "ACATTT": 538, + "ACATTC": 539, + "ACATTG": 540, + "ACATCA": 541, + "ACATCT": 542, + "ACATCC": 543, + "ACATCG": 544, + "ACATGA": 545, + "ACATGT": 546, + "ACATGC": 547, + "ACATGG": 548, + "ACACAA": 549, + "ACACAT": 550, + "ACACAC": 551, + "ACACAG": 552, + "ACACTA": 553, + "ACACTT": 554, + "ACACTC": 555, + "ACACTG": 556, + "ACACCA": 557, + "ACACCT": 558, + "ACACCC": 559, + "ACACCG": 560, + "ACACGA": 561, + "ACACGT": 562, + "ACACGC": 563, + "ACACGG": 564, + "ACAGAA": 565, + "ACAGAT": 566, + "ACAGAC": 567, + "ACAGAG": 568, + "ACAGTA": 569, + "ACAGTT": 570, + "ACAGTC": 571, + "ACAGTG": 572, + "ACAGCA": 573, + "ACAGCT": 574, + "ACAGCC": 575, + "ACAGCG": 576, + "ACAGGA": 577, + "ACAGGT": 578, + "ACAGGC": 579, + "ACAGGG": 580, + "ACTAAA": 581, + "ACTAAT": 582, + "ACTAAC": 583, + "ACTAAG": 584, + "ACTATA": 585, + "ACTATT": 586, + "ACTATC": 587, + "ACTATG": 588, + "ACTACA": 589, + "ACTACT": 590, + "ACTACC": 591, + "ACTACG": 592, + "ACTAGA": 593, + "ACTAGT": 594, + "ACTAGC": 595, + "ACTAGG": 596, + "ACTTAA": 597, + "ACTTAT": 598, + "ACTTAC": 599, + "ACTTAG": 600, + "ACTTTA": 601, + "ACTTTT": 602, + "ACTTTC": 603, + "ACTTTG": 604, + "ACTTCA": 605, + "ACTTCT": 606, + "ACTTCC": 607, + "ACTTCG": 608, + "ACTTGA": 609, + "ACTTGT": 610, + "ACTTGC": 611, + "ACTTGG": 612, + "ACTCAA": 613, + "ACTCAT": 614, + "ACTCAC": 615, + "ACTCAG": 616, + "ACTCTA": 617, + "ACTCTT": 618, + "ACTCTC": 619, + "ACTCTG": 620, + "ACTCCA": 621, + "ACTCCT": 622, + "ACTCCC": 623, + "ACTCCG": 624, + "ACTCGA": 625, + "ACTCGT": 626, + "ACTCGC": 627, + "ACTCGG": 628, + "ACTGAA": 629, + "ACTGAT": 630, + "ACTGAC": 631, + "ACTGAG": 632, + "ACTGTA": 633, + "ACTGTT": 634, + "ACTGTC": 635, + "ACTGTG": 636, + "ACTGCA": 637, + "ACTGCT": 638, + "ACTGCC": 639, + "ACTGCG": 640, + "ACTGGA": 641, + "ACTGGT": 642, + "ACTGGC": 643, + "ACTGGG": 644, + "ACCAAA": 645, + "ACCAAT": 646, + "ACCAAC": 647, + "ACCAAG": 648, + "ACCATA": 649, + "ACCATT": 650, + "ACCATC": 651, + "ACCATG": 652, + "ACCACA": 653, + "ACCACT": 654, + "ACCACC": 655, + "ACCACG": 656, + "ACCAGA": 657, + "ACCAGT": 658, + "ACCAGC": 659, + "ACCAGG": 660, + "ACCTAA": 661, + "ACCTAT": 662, + "ACCTAC": 663, + "ACCTAG": 664, + "ACCTTA": 665, + "ACCTTT": 666, + "ACCTTC": 667, + "ACCTTG": 668, + "ACCTCA": 669, + "ACCTCT": 670, + "ACCTCC": 671, + "ACCTCG": 672, + "ACCTGA": 673, + "ACCTGT": 674, + "ACCTGC": 675, + "ACCTGG": 676, + "ACCCAA": 677, + "ACCCAT": 678, + "ACCCAC": 679, + "ACCCAG": 680, + "ACCCTA": 681, + "ACCCTT": 682, + "ACCCTC": 683, + "ACCCTG": 684, + "ACCCCA": 685, + "ACCCCT": 686, + "ACCCCC": 687, + "ACCCCG": 688, + "ACCCGA": 689, + "ACCCGT": 690, + "ACCCGC": 691, + "ACCCGG": 692, + "ACCGAA": 693, + "ACCGAT": 694, + "ACCGAC": 695, + "ACCGAG": 696, + "ACCGTA": 697, + "ACCGTT": 698, + "ACCGTC": 699, + "ACCGTG": 700, + "ACCGCA": 701, + "ACCGCT": 702, + "ACCGCC": 703, + "ACCGCG": 704, + "ACCGGA": 705, + "ACCGGT": 706, + "ACCGGC": 707, + "ACCGGG": 708, + "ACGAAA": 709, + "ACGAAT": 710, + "ACGAAC": 711, + "ACGAAG": 712, + "ACGATA": 713, + "ACGATT": 714, + "ACGATC": 715, + "ACGATG": 716, + "ACGACA": 717, + "ACGACT": 718, + "ACGACC": 719, + "ACGACG": 720, + "ACGAGA": 721, + "ACGAGT": 722, + "ACGAGC": 723, + "ACGAGG": 724, + "ACGTAA": 725, + "ACGTAT": 726, + "ACGTAC": 727, + "ACGTAG": 728, + "ACGTTA": 729, + "ACGTTT": 730, + "ACGTTC": 731, + "ACGTTG": 732, + "ACGTCA": 733, + "ACGTCT": 734, + "ACGTCC": 735, + "ACGTCG": 736, + "ACGTGA": 737, + "ACGTGT": 738, + "ACGTGC": 739, + "ACGTGG": 740, + "ACGCAA": 741, + "ACGCAT": 742, + "ACGCAC": 743, + "ACGCAG": 744, + "ACGCTA": 745, + "ACGCTT": 746, + "ACGCTC": 747, + "ACGCTG": 748, + "ACGCCA": 749, + "ACGCCT": 750, + "ACGCCC": 751, + "ACGCCG": 752, + "ACGCGA": 753, + "ACGCGT": 754, + "ACGCGC": 755, + "ACGCGG": 756, + "ACGGAA": 757, + "ACGGAT": 758, + "ACGGAC": 759, + "ACGGAG": 760, + "ACGGTA": 761, + "ACGGTT": 762, + "ACGGTC": 763, + "ACGGTG": 764, + "ACGGCA": 765, + "ACGGCT": 766, + "ACGGCC": 767, + "ACGGCG": 768, + "ACGGGA": 769, + "ACGGGT": 770, + "ACGGGC": 771, + "ACGGGG": 772, + "AGAAAA": 773, + "AGAAAT": 774, + "AGAAAC": 775, + "AGAAAG": 776, + "AGAATA": 777, + "AGAATT": 778, + "AGAATC": 779, + "AGAATG": 780, + "AGAACA": 781, + "AGAACT": 782, + "AGAACC": 783, + "AGAACG": 784, + "AGAAGA": 785, + "AGAAGT": 786, + "AGAAGC": 787, + "AGAAGG": 788, + "AGATAA": 789, + "AGATAT": 790, + "AGATAC": 791, + "AGATAG": 792, + "AGATTA": 793, + "AGATTT": 794, + "AGATTC": 795, + "AGATTG": 796, + "AGATCA": 797, + "AGATCT": 798, + "AGATCC": 799, + "AGATCG": 800, + "AGATGA": 801, + "AGATGT": 802, + "AGATGC": 803, + "AGATGG": 804, + "AGACAA": 805, + "AGACAT": 806, + "AGACAC": 807, + "AGACAG": 808, + "AGACTA": 809, + "AGACTT": 810, + "AGACTC": 811, + "AGACTG": 812, + "AGACCA": 813, + "AGACCT": 814, + "AGACCC": 815, + "AGACCG": 816, + "AGACGA": 817, + "AGACGT": 818, + "AGACGC": 819, + "AGACGG": 820, + "AGAGAA": 821, + "AGAGAT": 822, + "AGAGAC": 823, + "AGAGAG": 824, + "AGAGTA": 825, + "AGAGTT": 826, + "AGAGTC": 827, + "AGAGTG": 828, + "AGAGCA": 829, + "AGAGCT": 830, + "AGAGCC": 831, + "AGAGCG": 832, + "AGAGGA": 833, + "AGAGGT": 834, + "AGAGGC": 835, + "AGAGGG": 836, + "AGTAAA": 837, + "AGTAAT": 838, + "AGTAAC": 839, + "AGTAAG": 840, + "AGTATA": 841, + "AGTATT": 842, + "AGTATC": 843, + "AGTATG": 844, + "AGTACA": 845, + "AGTACT": 846, + "AGTACC": 847, + "AGTACG": 848, + "AGTAGA": 849, + "AGTAGT": 850, + "AGTAGC": 851, + "AGTAGG": 852, + "AGTTAA": 853, + "AGTTAT": 854, + "AGTTAC": 855, + "AGTTAG": 856, + "AGTTTA": 857, + "AGTTTT": 858, + "AGTTTC": 859, + "AGTTTG": 860, + "AGTTCA": 861, + "AGTTCT": 862, + "AGTTCC": 863, + "AGTTCG": 864, + "AGTTGA": 865, + "AGTTGT": 866, + "AGTTGC": 867, + "AGTTGG": 868, + "AGTCAA": 869, + "AGTCAT": 870, + "AGTCAC": 871, + "AGTCAG": 872, + "AGTCTA": 873, + "AGTCTT": 874, + "AGTCTC": 875, + "AGTCTG": 876, + "AGTCCA": 877, + "AGTCCT": 878, + "AGTCCC": 879, + "AGTCCG": 880, + "AGTCGA": 881, + "AGTCGT": 882, + "AGTCGC": 883, + "AGTCGG": 884, + "AGTGAA": 885, + "AGTGAT": 886, + "AGTGAC": 887, + "AGTGAG": 888, + "AGTGTA": 889, + "AGTGTT": 890, + "AGTGTC": 891, + "AGTGTG": 892, + "AGTGCA": 893, + "AGTGCT": 894, + "AGTGCC": 895, + "AGTGCG": 896, + "AGTGGA": 897, + "AGTGGT": 898, + "AGTGGC": 899, + "AGTGGG": 900, + "AGCAAA": 901, + "AGCAAT": 902, + "AGCAAC": 903, + "AGCAAG": 904, + "AGCATA": 905, + "AGCATT": 906, + "AGCATC": 907, + "AGCATG": 908, + "AGCACA": 909, + "AGCACT": 910, + "AGCACC": 911, + "AGCACG": 912, + "AGCAGA": 913, + "AGCAGT": 914, + "AGCAGC": 915, + "AGCAGG": 916, + "AGCTAA": 917, + "AGCTAT": 918, + "AGCTAC": 919, + "AGCTAG": 920, + "AGCTTA": 921, + "AGCTTT": 922, + "AGCTTC": 923, + "AGCTTG": 924, + "AGCTCA": 925, + "AGCTCT": 926, + "AGCTCC": 927, + "AGCTCG": 928, + "AGCTGA": 929, + "AGCTGT": 930, + "AGCTGC": 931, + "AGCTGG": 932, + "AGCCAA": 933, + "AGCCAT": 934, + "AGCCAC": 935, + "AGCCAG": 936, + "AGCCTA": 937, + "AGCCTT": 938, + "AGCCTC": 939, + "AGCCTG": 940, + "AGCCCA": 941, + "AGCCCT": 942, + "AGCCCC": 943, + "AGCCCG": 944, + "AGCCGA": 945, + "AGCCGT": 946, + "AGCCGC": 947, + "AGCCGG": 948, + "AGCGAA": 949, + "AGCGAT": 950, + "AGCGAC": 951, + "AGCGAG": 952, + "AGCGTA": 953, + "AGCGTT": 954, + "AGCGTC": 955, + "AGCGTG": 956, + "AGCGCA": 957, + "AGCGCT": 958, + "AGCGCC": 959, + "AGCGCG": 960, + "AGCGGA": 961, + "AGCGGT": 962, + "AGCGGC": 963, + "AGCGGG": 964, + "AGGAAA": 965, + "AGGAAT": 966, + "AGGAAC": 967, + "AGGAAG": 968, + "AGGATA": 969, + "AGGATT": 970, + "AGGATC": 971, + "AGGATG": 972, + "AGGACA": 973, + "AGGACT": 974, + "AGGACC": 975, + "AGGACG": 976, + "AGGAGA": 977, + "AGGAGT": 978, + "AGGAGC": 979, + "AGGAGG": 980, + "AGGTAA": 981, + "AGGTAT": 982, + "AGGTAC": 983, + "AGGTAG": 984, + "AGGTTA": 985, + "AGGTTT": 986, + "AGGTTC": 987, + "AGGTTG": 988, + "AGGTCA": 989, + "AGGTCT": 990, + "AGGTCC": 991, + "AGGTCG": 992, + "AGGTGA": 993, + "AGGTGT": 994, + "AGGTGC": 995, + "AGGTGG": 996, + "AGGCAA": 997, + "AGGCAT": 998, + "AGGCAC": 999, + "AGGCAG": 1000, + "AGGCTA": 1001, + "AGGCTT": 1002, + "AGGCTC": 1003, + "AGGCTG": 1004, + "AGGCCA": 1005, + "AGGCCT": 1006, + "AGGCCC": 1007, + "AGGCCG": 1008, + "AGGCGA": 1009, + "AGGCGT": 1010, + "AGGCGC": 1011, + "AGGCGG": 1012, + "AGGGAA": 1013, + "AGGGAT": 1014, + "AGGGAC": 1015, + "AGGGAG": 1016, + "AGGGTA": 1017, + "AGGGTT": 1018, + "AGGGTC": 1019, + "AGGGTG": 1020, + "AGGGCA": 1021, + "AGGGCT": 1022, + "AGGGCC": 1023, + "AGGGCG": 1024, + "AGGGGA": 1025, + "AGGGGT": 1026, + "AGGGGC": 1027, + "AGGGGG": 1028, + "TAAAAA": 1029, + "TAAAAT": 1030, + "TAAAAC": 1031, + "TAAAAG": 1032, + "TAAATA": 1033, + "TAAATT": 1034, + "TAAATC": 1035, + "TAAATG": 1036, + "TAAACA": 1037, + "TAAACT": 1038, + "TAAACC": 1039, + "TAAACG": 1040, + "TAAAGA": 1041, + "TAAAGT": 1042, + "TAAAGC": 1043, + "TAAAGG": 1044, + "TAATAA": 1045, + "TAATAT": 1046, + "TAATAC": 1047, + "TAATAG": 1048, + "TAATTA": 1049, + "TAATTT": 1050, + "TAATTC": 1051, + "TAATTG": 1052, + "TAATCA": 1053, + "TAATCT": 1054, + "TAATCC": 1055, + "TAATCG": 1056, + "TAATGA": 1057, + "TAATGT": 1058, + "TAATGC": 1059, + "TAATGG": 1060, + "TAACAA": 1061, + "TAACAT": 1062, + "TAACAC": 1063, + "TAACAG": 1064, + "TAACTA": 1065, + "TAACTT": 1066, + "TAACTC": 1067, + "TAACTG": 1068, + "TAACCA": 1069, + "TAACCT": 1070, + "TAACCC": 1071, + "TAACCG": 1072, + "TAACGA": 1073, + "TAACGT": 1074, + "TAACGC": 1075, + "TAACGG": 1076, + "TAAGAA": 1077, + "TAAGAT": 1078, + "TAAGAC": 1079, + "TAAGAG": 1080, + "TAAGTA": 1081, + "TAAGTT": 1082, + "TAAGTC": 1083, + "TAAGTG": 1084, + "TAAGCA": 1085, + "TAAGCT": 1086, + "TAAGCC": 1087, + "TAAGCG": 1088, + "TAAGGA": 1089, + "TAAGGT": 1090, + "TAAGGC": 1091, + "TAAGGG": 1092, + "TATAAA": 1093, + "TATAAT": 1094, + "TATAAC": 1095, + "TATAAG": 1096, + "TATATA": 1097, + "TATATT": 1098, + "TATATC": 1099, + "TATATG": 1100, + "TATACA": 1101, + "TATACT": 1102, + "TATACC": 1103, + "TATACG": 1104, + "TATAGA": 1105, + "TATAGT": 1106, + "TATAGC": 1107, + "TATAGG": 1108, + "TATTAA": 1109, + "TATTAT": 1110, + "TATTAC": 1111, + "TATTAG": 1112, + "TATTTA": 1113, + "TATTTT": 1114, + "TATTTC": 1115, + "TATTTG": 1116, + "TATTCA": 1117, + "TATTCT": 1118, + "TATTCC": 1119, + "TATTCG": 1120, + "TATTGA": 1121, + "TATTGT": 1122, + "TATTGC": 1123, + "TATTGG": 1124, + "TATCAA": 1125, + "TATCAT": 1126, + "TATCAC": 1127, + "TATCAG": 1128, + "TATCTA": 1129, + "TATCTT": 1130, + "TATCTC": 1131, + "TATCTG": 1132, + "TATCCA": 1133, + "TATCCT": 1134, + "TATCCC": 1135, + "TATCCG": 1136, + "TATCGA": 1137, + "TATCGT": 1138, + "TATCGC": 1139, + "TATCGG": 1140, + "TATGAA": 1141, + "TATGAT": 1142, + "TATGAC": 1143, + "TATGAG": 1144, + "TATGTA": 1145, + "TATGTT": 1146, + "TATGTC": 1147, + "TATGTG": 1148, + "TATGCA": 1149, + "TATGCT": 1150, + "TATGCC": 1151, + "TATGCG": 1152, + "TATGGA": 1153, + "TATGGT": 1154, + "TATGGC": 1155, + "TATGGG": 1156, + "TACAAA": 1157, + "TACAAT": 1158, + "TACAAC": 1159, + "TACAAG": 1160, + "TACATA": 1161, + "TACATT": 1162, + "TACATC": 1163, + "TACATG": 1164, + "TACACA": 1165, + "TACACT": 1166, + "TACACC": 1167, + "TACACG": 1168, + "TACAGA": 1169, + "TACAGT": 1170, + "TACAGC": 1171, + "TACAGG": 1172, + "TACTAA": 1173, + "TACTAT": 1174, + "TACTAC": 1175, + "TACTAG": 1176, + "TACTTA": 1177, + "TACTTT": 1178, + "TACTTC": 1179, + "TACTTG": 1180, + "TACTCA": 1181, + "TACTCT": 1182, + "TACTCC": 1183, + "TACTCG": 1184, + "TACTGA": 1185, + "TACTGT": 1186, + "TACTGC": 1187, + "TACTGG": 1188, + "TACCAA": 1189, + "TACCAT": 1190, + "TACCAC": 1191, + "TACCAG": 1192, + "TACCTA": 1193, + "TACCTT": 1194, + "TACCTC": 1195, + "TACCTG": 1196, + "TACCCA": 1197, + "TACCCT": 1198, + "TACCCC": 1199, + "TACCCG": 1200, + "TACCGA": 1201, + "TACCGT": 1202, + "TACCGC": 1203, + "TACCGG": 1204, + "TACGAA": 1205, + "TACGAT": 1206, + "TACGAC": 1207, + "TACGAG": 1208, + "TACGTA": 1209, + "TACGTT": 1210, + "TACGTC": 1211, + "TACGTG": 1212, + "TACGCA": 1213, + "TACGCT": 1214, + "TACGCC": 1215, + "TACGCG": 1216, + "TACGGA": 1217, + "TACGGT": 1218, + "TACGGC": 1219, + "TACGGG": 1220, + "TAGAAA": 1221, + "TAGAAT": 1222, + "TAGAAC": 1223, + "TAGAAG": 1224, + "TAGATA": 1225, + "TAGATT": 1226, + "TAGATC": 1227, + "TAGATG": 1228, + "TAGACA": 1229, + "TAGACT": 1230, + "TAGACC": 1231, + "TAGACG": 1232, + "TAGAGA": 1233, + "TAGAGT": 1234, + "TAGAGC": 1235, + "TAGAGG": 1236, + "TAGTAA": 1237, + "TAGTAT": 1238, + "TAGTAC": 1239, + "TAGTAG": 1240, + "TAGTTA": 1241, + "TAGTTT": 1242, + "TAGTTC": 1243, + "TAGTTG": 1244, + "TAGTCA": 1245, + "TAGTCT": 1246, + "TAGTCC": 1247, + "TAGTCG": 1248, + "TAGTGA": 1249, + "TAGTGT": 1250, + "TAGTGC": 1251, + "TAGTGG": 1252, + "TAGCAA": 1253, + "TAGCAT": 1254, + "TAGCAC": 1255, + "TAGCAG": 1256, + "TAGCTA": 1257, + "TAGCTT": 1258, + "TAGCTC": 1259, + "TAGCTG": 1260, + "TAGCCA": 1261, + "TAGCCT": 1262, + "TAGCCC": 1263, + "TAGCCG": 1264, + "TAGCGA": 1265, + "TAGCGT": 1266, + "TAGCGC": 1267, + "TAGCGG": 1268, + "TAGGAA": 1269, + "TAGGAT": 1270, + "TAGGAC": 1271, + "TAGGAG": 1272, + "TAGGTA": 1273, + "TAGGTT": 1274, + "TAGGTC": 1275, + "TAGGTG": 1276, + "TAGGCA": 1277, + "TAGGCT": 1278, + "TAGGCC": 1279, + "TAGGCG": 1280, + "TAGGGA": 1281, + "TAGGGT": 1282, + "TAGGGC": 1283, + "TAGGGG": 1284, + "TTAAAA": 1285, + "TTAAAT": 1286, + "TTAAAC": 1287, + "TTAAAG": 1288, + "TTAATA": 1289, + "TTAATT": 1290, + "TTAATC": 1291, + "TTAATG": 1292, + "TTAACA": 1293, + "TTAACT": 1294, + "TTAACC": 1295, + "TTAACG": 1296, + "TTAAGA": 1297, + "TTAAGT": 1298, + "TTAAGC": 1299, + "TTAAGG": 1300, + "TTATAA": 1301, + "TTATAT": 1302, + "TTATAC": 1303, + "TTATAG": 1304, + "TTATTA": 1305, + "TTATTT": 1306, + "TTATTC": 1307, + "TTATTG": 1308, + "TTATCA": 1309, + "TTATCT": 1310, + "TTATCC": 1311, + "TTATCG": 1312, + "TTATGA": 1313, + "TTATGT": 1314, + "TTATGC": 1315, + "TTATGG": 1316, + "TTACAA": 1317, + "TTACAT": 1318, + "TTACAC": 1319, + "TTACAG": 1320, + "TTACTA": 1321, + "TTACTT": 1322, + "TTACTC": 1323, + "TTACTG": 1324, + "TTACCA": 1325, + "TTACCT": 1326, + "TTACCC": 1327, + "TTACCG": 1328, + "TTACGA": 1329, + "TTACGT": 1330, + "TTACGC": 1331, + "TTACGG": 1332, + "TTAGAA": 1333, + "TTAGAT": 1334, + "TTAGAC": 1335, + "TTAGAG": 1336, + "TTAGTA": 1337, + "TTAGTT": 1338, + "TTAGTC": 1339, + "TTAGTG": 1340, + "TTAGCA": 1341, + "TTAGCT": 1342, + "TTAGCC": 1343, + "TTAGCG": 1344, + "TTAGGA": 1345, + "TTAGGT": 1346, + "TTAGGC": 1347, + "TTAGGG": 1348, + "TTTAAA": 1349, + "TTTAAT": 1350, + "TTTAAC": 1351, + "TTTAAG": 1352, + "TTTATA": 1353, + "TTTATT": 1354, + "TTTATC": 1355, + "TTTATG": 1356, + "TTTACA": 1357, + "TTTACT": 1358, + "TTTACC": 1359, + "TTTACG": 1360, + "TTTAGA": 1361, + "TTTAGT": 1362, + "TTTAGC": 1363, + "TTTAGG": 1364, + "TTTTAA": 1365, + "TTTTAT": 1366, + "TTTTAC": 1367, + "TTTTAG": 1368, + "TTTTTA": 1369, + "TTTTTT": 1370, + "TTTTTC": 1371, + "TTTTTG": 1372, + "TTTTCA": 1373, + "TTTTCT": 1374, + "TTTTCC": 1375, + "TTTTCG": 1376, + "TTTTGA": 1377, + "TTTTGT": 1378, + "TTTTGC": 1379, + "TTTTGG": 1380, + "TTTCAA": 1381, + "TTTCAT": 1382, + "TTTCAC": 1383, + "TTTCAG": 1384, + "TTTCTA": 1385, + "TTTCTT": 1386, + "TTTCTC": 1387, + "TTTCTG": 1388, + "TTTCCA": 1389, + "TTTCCT": 1390, + "TTTCCC": 1391, + "TTTCCG": 1392, + "TTTCGA": 1393, + "TTTCGT": 1394, + "TTTCGC": 1395, + "TTTCGG": 1396, + "TTTGAA": 1397, + "TTTGAT": 1398, + "TTTGAC": 1399, + "TTTGAG": 1400, + "TTTGTA": 1401, + "TTTGTT": 1402, + "TTTGTC": 1403, + "TTTGTG": 1404, + "TTTGCA": 1405, + "TTTGCT": 1406, + "TTTGCC": 1407, + "TTTGCG": 1408, + "TTTGGA": 1409, + "TTTGGT": 1410, + "TTTGGC": 1411, + "TTTGGG": 1412, + "TTCAAA": 1413, + "TTCAAT": 1414, + "TTCAAC": 1415, + "TTCAAG": 1416, + "TTCATA": 1417, + "TTCATT": 1418, + "TTCATC": 1419, + "TTCATG": 1420, + "TTCACA": 1421, + "TTCACT": 1422, + "TTCACC": 1423, + "TTCACG": 1424, + "TTCAGA": 1425, + "TTCAGT": 1426, + "TTCAGC": 1427, + "TTCAGG": 1428, + "TTCTAA": 1429, + "TTCTAT": 1430, + "TTCTAC": 1431, + "TTCTAG": 1432, + "TTCTTA": 1433, + "TTCTTT": 1434, + "TTCTTC": 1435, + "TTCTTG": 1436, + "TTCTCA": 1437, + "TTCTCT": 1438, + "TTCTCC": 1439, + "TTCTCG": 1440, + "TTCTGA": 1441, + "TTCTGT": 1442, + "TTCTGC": 1443, + "TTCTGG": 1444, + "TTCCAA": 1445, + "TTCCAT": 1446, + "TTCCAC": 1447, + "TTCCAG": 1448, + "TTCCTA": 1449, + "TTCCTT": 1450, + "TTCCTC": 1451, + "TTCCTG": 1452, + "TTCCCA": 1453, + "TTCCCT": 1454, + "TTCCCC": 1455, + "TTCCCG": 1456, + "TTCCGA": 1457, + "TTCCGT": 1458, + "TTCCGC": 1459, + "TTCCGG": 1460, + "TTCGAA": 1461, + "TTCGAT": 1462, + "TTCGAC": 1463, + "TTCGAG": 1464, + "TTCGTA": 1465, + "TTCGTT": 1466, + "TTCGTC": 1467, + "TTCGTG": 1468, + "TTCGCA": 1469, + "TTCGCT": 1470, + "TTCGCC": 1471, + "TTCGCG": 1472, + "TTCGGA": 1473, + "TTCGGT": 1474, + "TTCGGC": 1475, + "TTCGGG": 1476, + "TTGAAA": 1477, + "TTGAAT": 1478, + "TTGAAC": 1479, + "TTGAAG": 1480, + "TTGATA": 1481, + "TTGATT": 1482, + "TTGATC": 1483, + "TTGATG": 1484, + "TTGACA": 1485, + "TTGACT": 1486, + "TTGACC": 1487, + "TTGACG": 1488, + "TTGAGA": 1489, + "TTGAGT": 1490, + "TTGAGC": 1491, + "TTGAGG": 1492, + "TTGTAA": 1493, + "TTGTAT": 1494, + "TTGTAC": 1495, + "TTGTAG": 1496, + "TTGTTA": 1497, + "TTGTTT": 1498, + "TTGTTC": 1499, + "TTGTTG": 1500, + "TTGTCA": 1501, + "TTGTCT": 1502, + "TTGTCC": 1503, + "TTGTCG": 1504, + "TTGTGA": 1505, + "TTGTGT": 1506, + "TTGTGC": 1507, + "TTGTGG": 1508, + "TTGCAA": 1509, + "TTGCAT": 1510, + "TTGCAC": 1511, + "TTGCAG": 1512, + "TTGCTA": 1513, + "TTGCTT": 1514, + "TTGCTC": 1515, + "TTGCTG": 1516, + "TTGCCA": 1517, + "TTGCCT": 1518, + "TTGCCC": 1519, + "TTGCCG": 1520, + "TTGCGA": 1521, + "TTGCGT": 1522, + "TTGCGC": 1523, + "TTGCGG": 1524, + "TTGGAA": 1525, + "TTGGAT": 1526, + "TTGGAC": 1527, + "TTGGAG": 1528, + "TTGGTA": 1529, + "TTGGTT": 1530, + "TTGGTC": 1531, + "TTGGTG": 1532, + "TTGGCA": 1533, + "TTGGCT": 1534, + "TTGGCC": 1535, + "TTGGCG": 1536, + "TTGGGA": 1537, + "TTGGGT": 1538, + "TTGGGC": 1539, + "TTGGGG": 1540, + "TCAAAA": 1541, + "TCAAAT": 1542, + "TCAAAC": 1543, + "TCAAAG": 1544, + "TCAATA": 1545, + "TCAATT": 1546, + "TCAATC": 1547, + "TCAATG": 1548, + "TCAACA": 1549, + "TCAACT": 1550, + "TCAACC": 1551, + "TCAACG": 1552, + "TCAAGA": 1553, + "TCAAGT": 1554, + "TCAAGC": 1555, + "TCAAGG": 1556, + "TCATAA": 1557, + "TCATAT": 1558, + "TCATAC": 1559, + "TCATAG": 1560, + "TCATTA": 1561, + "TCATTT": 1562, + "TCATTC": 1563, + "TCATTG": 1564, + "TCATCA": 1565, + "TCATCT": 1566, + "TCATCC": 1567, + "TCATCG": 1568, + "TCATGA": 1569, + "TCATGT": 1570, + "TCATGC": 1571, + "TCATGG": 1572, + "TCACAA": 1573, + "TCACAT": 1574, + "TCACAC": 1575, + "TCACAG": 1576, + "TCACTA": 1577, + "TCACTT": 1578, + "TCACTC": 1579, + "TCACTG": 1580, + "TCACCA": 1581, + "TCACCT": 1582, + "TCACCC": 1583, + "TCACCG": 1584, + "TCACGA": 1585, + "TCACGT": 1586, + "TCACGC": 1587, + "TCACGG": 1588, + "TCAGAA": 1589, + "TCAGAT": 1590, + "TCAGAC": 1591, + "TCAGAG": 1592, + "TCAGTA": 1593, + "TCAGTT": 1594, + "TCAGTC": 1595, + "TCAGTG": 1596, + "TCAGCA": 1597, + "TCAGCT": 1598, + "TCAGCC": 1599, + "TCAGCG": 1600, + "TCAGGA": 1601, + "TCAGGT": 1602, + "TCAGGC": 1603, + "TCAGGG": 1604, + "TCTAAA": 1605, + "TCTAAT": 1606, + "TCTAAC": 1607, + "TCTAAG": 1608, + "TCTATA": 1609, + "TCTATT": 1610, + "TCTATC": 1611, + "TCTATG": 1612, + "TCTACA": 1613, + "TCTACT": 1614, + "TCTACC": 1615, + "TCTACG": 1616, + "TCTAGA": 1617, + "TCTAGT": 1618, + "TCTAGC": 1619, + "TCTAGG": 1620, + "TCTTAA": 1621, + "TCTTAT": 1622, + "TCTTAC": 1623, + "TCTTAG": 1624, + "TCTTTA": 1625, + "TCTTTT": 1626, + "TCTTTC": 1627, + "TCTTTG": 1628, + "TCTTCA": 1629, + "TCTTCT": 1630, + "TCTTCC": 1631, + "TCTTCG": 1632, + "TCTTGA": 1633, + "TCTTGT": 1634, + "TCTTGC": 1635, + "TCTTGG": 1636, + "TCTCAA": 1637, + "TCTCAT": 1638, + "TCTCAC": 1639, + "TCTCAG": 1640, + "TCTCTA": 1641, + "TCTCTT": 1642, + "TCTCTC": 1643, + "TCTCTG": 1644, + "TCTCCA": 1645, + "TCTCCT": 1646, + "TCTCCC": 1647, + "TCTCCG": 1648, + "TCTCGA": 1649, + "TCTCGT": 1650, + "TCTCGC": 1651, + "TCTCGG": 1652, + "TCTGAA": 1653, + "TCTGAT": 1654, + "TCTGAC": 1655, + "TCTGAG": 1656, + "TCTGTA": 1657, + "TCTGTT": 1658, + "TCTGTC": 1659, + "TCTGTG": 1660, + "TCTGCA": 1661, + "TCTGCT": 1662, + "TCTGCC": 1663, + "TCTGCG": 1664, + "TCTGGA": 1665, + "TCTGGT": 1666, + "TCTGGC": 1667, + "TCTGGG": 1668, + "TCCAAA": 1669, + "TCCAAT": 1670, + "TCCAAC": 1671, + "TCCAAG": 1672, + "TCCATA": 1673, + "TCCATT": 1674, + "TCCATC": 1675, + "TCCATG": 1676, + "TCCACA": 1677, + "TCCACT": 1678, + "TCCACC": 1679, + "TCCACG": 1680, + "TCCAGA": 1681, + "TCCAGT": 1682, + "TCCAGC": 1683, + "TCCAGG": 1684, + "TCCTAA": 1685, + "TCCTAT": 1686, + "TCCTAC": 1687, + "TCCTAG": 1688, + "TCCTTA": 1689, + "TCCTTT": 1690, + "TCCTTC": 1691, + "TCCTTG": 1692, + "TCCTCA": 1693, + "TCCTCT": 1694, + "TCCTCC": 1695, + "TCCTCG": 1696, + "TCCTGA": 1697, + "TCCTGT": 1698, + "TCCTGC": 1699, + "TCCTGG": 1700, + "TCCCAA": 1701, + "TCCCAT": 1702, + "TCCCAC": 1703, + "TCCCAG": 1704, + "TCCCTA": 1705, + "TCCCTT": 1706, + "TCCCTC": 1707, + "TCCCTG": 1708, + "TCCCCA": 1709, + "TCCCCT": 1710, + "TCCCCC": 1711, + "TCCCCG": 1712, + "TCCCGA": 1713, + "TCCCGT": 1714, + "TCCCGC": 1715, + "TCCCGG": 1716, + "TCCGAA": 1717, + "TCCGAT": 1718, + "TCCGAC": 1719, + "TCCGAG": 1720, + "TCCGTA": 1721, + "TCCGTT": 1722, + "TCCGTC": 1723, + "TCCGTG": 1724, + "TCCGCA": 1725, + "TCCGCT": 1726, + "TCCGCC": 1727, + "TCCGCG": 1728, + "TCCGGA": 1729, + "TCCGGT": 1730, + "TCCGGC": 1731, + "TCCGGG": 1732, + "TCGAAA": 1733, + "TCGAAT": 1734, + "TCGAAC": 1735, + "TCGAAG": 1736, + "TCGATA": 1737, + "TCGATT": 1738, + "TCGATC": 1739, + "TCGATG": 1740, + "TCGACA": 1741, + "TCGACT": 1742, + "TCGACC": 1743, + "TCGACG": 1744, + "TCGAGA": 1745, + "TCGAGT": 1746, + "TCGAGC": 1747, + "TCGAGG": 1748, + "TCGTAA": 1749, + "TCGTAT": 1750, + "TCGTAC": 1751, + "TCGTAG": 1752, + "TCGTTA": 1753, + "TCGTTT": 1754, + "TCGTTC": 1755, + "TCGTTG": 1756, + "TCGTCA": 1757, + "TCGTCT": 1758, + "TCGTCC": 1759, + "TCGTCG": 1760, + "TCGTGA": 1761, + "TCGTGT": 1762, + "TCGTGC": 1763, + "TCGTGG": 1764, + "TCGCAA": 1765, + "TCGCAT": 1766, + "TCGCAC": 1767, + "TCGCAG": 1768, + "TCGCTA": 1769, + "TCGCTT": 1770, + "TCGCTC": 1771, + "TCGCTG": 1772, + "TCGCCA": 1773, + "TCGCCT": 1774, + "TCGCCC": 1775, + "TCGCCG": 1776, + "TCGCGA": 1777, + "TCGCGT": 1778, + "TCGCGC": 1779, + "TCGCGG": 1780, + "TCGGAA": 1781, + "TCGGAT": 1782, + "TCGGAC": 1783, + "TCGGAG": 1784, + "TCGGTA": 1785, + "TCGGTT": 1786, + "TCGGTC": 1787, + "TCGGTG": 1788, + "TCGGCA": 1789, + "TCGGCT": 1790, + "TCGGCC": 1791, + "TCGGCG": 1792, + "TCGGGA": 1793, + "TCGGGT": 1794, + "TCGGGC": 1795, + "TCGGGG": 1796, + "TGAAAA": 1797, + "TGAAAT": 1798, + "TGAAAC": 1799, + "TGAAAG": 1800, + "TGAATA": 1801, + "TGAATT": 1802, + "TGAATC": 1803, + "TGAATG": 1804, + "TGAACA": 1805, + "TGAACT": 1806, + "TGAACC": 1807, + "TGAACG": 1808, + "TGAAGA": 1809, + "TGAAGT": 1810, + "TGAAGC": 1811, + "TGAAGG": 1812, + "TGATAA": 1813, + "TGATAT": 1814, + "TGATAC": 1815, + "TGATAG": 1816, + "TGATTA": 1817, + "TGATTT": 1818, + "TGATTC": 1819, + "TGATTG": 1820, + "TGATCA": 1821, + "TGATCT": 1822, + "TGATCC": 1823, + "TGATCG": 1824, + "TGATGA": 1825, + "TGATGT": 1826, + "TGATGC": 1827, + "TGATGG": 1828, + "TGACAA": 1829, + "TGACAT": 1830, + "TGACAC": 1831, + "TGACAG": 1832, + "TGACTA": 1833, + "TGACTT": 1834, + "TGACTC": 1835, + "TGACTG": 1836, + "TGACCA": 1837, + "TGACCT": 1838, + "TGACCC": 1839, + "TGACCG": 1840, + "TGACGA": 1841, + "TGACGT": 1842, + "TGACGC": 1843, + "TGACGG": 1844, + "TGAGAA": 1845, + "TGAGAT": 1846, + "TGAGAC": 1847, + "TGAGAG": 1848, + "TGAGTA": 1849, + "TGAGTT": 1850, + "TGAGTC": 1851, + "TGAGTG": 1852, + "TGAGCA": 1853, + "TGAGCT": 1854, + "TGAGCC": 1855, + "TGAGCG": 1856, + "TGAGGA": 1857, + "TGAGGT": 1858, + "TGAGGC": 1859, + "TGAGGG": 1860, + "TGTAAA": 1861, + "TGTAAT": 1862, + "TGTAAC": 1863, + "TGTAAG": 1864, + "TGTATA": 1865, + "TGTATT": 1866, + "TGTATC": 1867, + "TGTATG": 1868, + "TGTACA": 1869, + "TGTACT": 1870, + "TGTACC": 1871, + "TGTACG": 1872, + "TGTAGA": 1873, + "TGTAGT": 1874, + "TGTAGC": 1875, + "TGTAGG": 1876, + "TGTTAA": 1877, + "TGTTAT": 1878, + "TGTTAC": 1879, + "TGTTAG": 1880, + "TGTTTA": 1881, + "TGTTTT": 1882, + "TGTTTC": 1883, + "TGTTTG": 1884, + "TGTTCA": 1885, + "TGTTCT": 1886, + "TGTTCC": 1887, + "TGTTCG": 1888, + "TGTTGA": 1889, + "TGTTGT": 1890, + "TGTTGC": 1891, + "TGTTGG": 1892, + "TGTCAA": 1893, + "TGTCAT": 1894, + "TGTCAC": 1895, + "TGTCAG": 1896, + "TGTCTA": 1897, + "TGTCTT": 1898, + "TGTCTC": 1899, + "TGTCTG": 1900, + "TGTCCA": 1901, + "TGTCCT": 1902, + "TGTCCC": 1903, + "TGTCCG": 1904, + "TGTCGA": 1905, + "TGTCGT": 1906, + "TGTCGC": 1907, + "TGTCGG": 1908, + "TGTGAA": 1909, + "TGTGAT": 1910, + "TGTGAC": 1911, + "TGTGAG": 1912, + "TGTGTA": 1913, + "TGTGTT": 1914, + "TGTGTC": 1915, + "TGTGTG": 1916, + "TGTGCA": 1917, + "TGTGCT": 1918, + "TGTGCC": 1919, + "TGTGCG": 1920, + "TGTGGA": 1921, + "TGTGGT": 1922, + "TGTGGC": 1923, + "TGTGGG": 1924, + "TGCAAA": 1925, + "TGCAAT": 1926, + "TGCAAC": 1927, + "TGCAAG": 1928, + "TGCATA": 1929, + "TGCATT": 1930, + "TGCATC": 1931, + "TGCATG": 1932, + "TGCACA": 1933, + "TGCACT": 1934, + "TGCACC": 1935, + "TGCACG": 1936, + "TGCAGA": 1937, + "TGCAGT": 1938, + "TGCAGC": 1939, + "TGCAGG": 1940, + "TGCTAA": 1941, + "TGCTAT": 1942, + "TGCTAC": 1943, + "TGCTAG": 1944, + "TGCTTA": 1945, + "TGCTTT": 1946, + "TGCTTC": 1947, + "TGCTTG": 1948, + "TGCTCA": 1949, + "TGCTCT": 1950, + "TGCTCC": 1951, + "TGCTCG": 1952, + "TGCTGA": 1953, + "TGCTGT": 1954, + "TGCTGC": 1955, + "TGCTGG": 1956, + "TGCCAA": 1957, + "TGCCAT": 1958, + "TGCCAC": 1959, + "TGCCAG": 1960, + "TGCCTA": 1961, + "TGCCTT": 1962, + "TGCCTC": 1963, + "TGCCTG": 1964, + "TGCCCA": 1965, + "TGCCCT": 1966, + "TGCCCC": 1967, + "TGCCCG": 1968, + "TGCCGA": 1969, + "TGCCGT": 1970, + "TGCCGC": 1971, + "TGCCGG": 1972, + "TGCGAA": 1973, + "TGCGAT": 1974, + "TGCGAC": 1975, + "TGCGAG": 1976, + "TGCGTA": 1977, + "TGCGTT": 1978, + "TGCGTC": 1979, + "TGCGTG": 1980, + "TGCGCA": 1981, + "TGCGCT": 1982, + "TGCGCC": 1983, + "TGCGCG": 1984, + "TGCGGA": 1985, + "TGCGGT": 1986, + "TGCGGC": 1987, + "TGCGGG": 1988, + "TGGAAA": 1989, + "TGGAAT": 1990, + "TGGAAC": 1991, + "TGGAAG": 1992, + "TGGATA": 1993, + "TGGATT": 1994, + "TGGATC": 1995, + "TGGATG": 1996, + "TGGACA": 1997, + "TGGACT": 1998, + "TGGACC": 1999, + "TGGACG": 2000, + "TGGAGA": 2001, + "TGGAGT": 2002, + "TGGAGC": 2003, + "TGGAGG": 2004, + "TGGTAA": 2005, + "TGGTAT": 2006, + "TGGTAC": 2007, + "TGGTAG": 2008, + "TGGTTA": 2009, + "TGGTTT": 2010, + "TGGTTC": 2011, + "TGGTTG": 2012, + "TGGTCA": 2013, + "TGGTCT": 2014, + "TGGTCC": 2015, + "TGGTCG": 2016, + "TGGTGA": 2017, + "TGGTGT": 2018, + "TGGTGC": 2019, + "TGGTGG": 2020, + "TGGCAA": 2021, + "TGGCAT": 2022, + "TGGCAC": 2023, + "TGGCAG": 2024, + "TGGCTA": 2025, + "TGGCTT": 2026, + "TGGCTC": 2027, + "TGGCTG": 2028, + "TGGCCA": 2029, + "TGGCCT": 2030, + "TGGCCC": 2031, + "TGGCCG": 2032, + "TGGCGA": 2033, + "TGGCGT": 2034, + "TGGCGC": 2035, + "TGGCGG": 2036, + "TGGGAA": 2037, + "TGGGAT": 2038, + "TGGGAC": 2039, + "TGGGAG": 2040, + "TGGGTA": 2041, + "TGGGTT": 2042, + "TGGGTC": 2043, + "TGGGTG": 2044, + "TGGGCA": 2045, + "TGGGCT": 2046, + "TGGGCC": 2047, + "TGGGCG": 2048, + "TGGGGA": 2049, + "TGGGGT": 2050, + "TGGGGC": 2051, + "TGGGGG": 2052, + "CAAAAA": 2053, + "CAAAAT": 2054, + "CAAAAC": 2055, + "CAAAAG": 2056, + "CAAATA": 2057, + "CAAATT": 2058, + "CAAATC": 2059, + "CAAATG": 2060, + "CAAACA": 2061, + "CAAACT": 2062, + "CAAACC": 2063, + "CAAACG": 2064, + "CAAAGA": 2065, + "CAAAGT": 2066, + "CAAAGC": 2067, + "CAAAGG": 2068, + "CAATAA": 2069, + "CAATAT": 2070, + "CAATAC": 2071, + "CAATAG": 2072, + "CAATTA": 2073, + "CAATTT": 2074, + "CAATTC": 2075, + "CAATTG": 2076, + "CAATCA": 2077, + "CAATCT": 2078, + "CAATCC": 2079, + "CAATCG": 2080, + "CAATGA": 2081, + "CAATGT": 2082, + "CAATGC": 2083, + "CAATGG": 2084, + "CAACAA": 2085, + "CAACAT": 2086, + "CAACAC": 2087, + "CAACAG": 2088, + "CAACTA": 2089, + "CAACTT": 2090, + "CAACTC": 2091, + "CAACTG": 2092, + "CAACCA": 2093, + "CAACCT": 2094, + "CAACCC": 2095, + "CAACCG": 2096, + "CAACGA": 2097, + "CAACGT": 2098, + "CAACGC": 2099, + "CAACGG": 2100, + "CAAGAA": 2101, + "CAAGAT": 2102, + "CAAGAC": 2103, + "CAAGAG": 2104, + "CAAGTA": 2105, + "CAAGTT": 2106, + "CAAGTC": 2107, + "CAAGTG": 2108, + "CAAGCA": 2109, + "CAAGCT": 2110, + "CAAGCC": 2111, + "CAAGCG": 2112, + "CAAGGA": 2113, + "CAAGGT": 2114, + "CAAGGC": 2115, + "CAAGGG": 2116, + "CATAAA": 2117, + "CATAAT": 2118, + "CATAAC": 2119, + "CATAAG": 2120, + "CATATA": 2121, + "CATATT": 2122, + "CATATC": 2123, + "CATATG": 2124, + "CATACA": 2125, + "CATACT": 2126, + "CATACC": 2127, + "CATACG": 2128, + "CATAGA": 2129, + "CATAGT": 2130, + "CATAGC": 2131, + "CATAGG": 2132, + "CATTAA": 2133, + "CATTAT": 2134, + "CATTAC": 2135, + "CATTAG": 2136, + "CATTTA": 2137, + "CATTTT": 2138, + "CATTTC": 2139, + "CATTTG": 2140, + "CATTCA": 2141, + "CATTCT": 2142, + "CATTCC": 2143, + "CATTCG": 2144, + "CATTGA": 2145, + "CATTGT": 2146, + "CATTGC": 2147, + "CATTGG": 2148, + "CATCAA": 2149, + "CATCAT": 2150, + "CATCAC": 2151, + "CATCAG": 2152, + "CATCTA": 2153, + "CATCTT": 2154, + "CATCTC": 2155, + "CATCTG": 2156, + "CATCCA": 2157, + "CATCCT": 2158, + "CATCCC": 2159, + "CATCCG": 2160, + "CATCGA": 2161, + "CATCGT": 2162, + "CATCGC": 2163, + "CATCGG": 2164, + "CATGAA": 2165, + "CATGAT": 2166, + "CATGAC": 2167, + "CATGAG": 2168, + "CATGTA": 2169, + "CATGTT": 2170, + "CATGTC": 2171, + "CATGTG": 2172, + "CATGCA": 2173, + "CATGCT": 2174, + "CATGCC": 2175, + "CATGCG": 2176, + "CATGGA": 2177, + "CATGGT": 2178, + "CATGGC": 2179, + "CATGGG": 2180, + "CACAAA": 2181, + "CACAAT": 2182, + "CACAAC": 2183, + "CACAAG": 2184, + "CACATA": 2185, + "CACATT": 2186, + "CACATC": 2187, + "CACATG": 2188, + "CACACA": 2189, + "CACACT": 2190, + "CACACC": 2191, + "CACACG": 2192, + "CACAGA": 2193, + "CACAGT": 2194, + "CACAGC": 2195, + "CACAGG": 2196, + "CACTAA": 2197, + "CACTAT": 2198, + "CACTAC": 2199, + "CACTAG": 2200, + "CACTTA": 2201, + "CACTTT": 2202, + "CACTTC": 2203, + "CACTTG": 2204, + "CACTCA": 2205, + "CACTCT": 2206, + "CACTCC": 2207, + "CACTCG": 2208, + "CACTGA": 2209, + "CACTGT": 2210, + "CACTGC": 2211, + "CACTGG": 2212, + "CACCAA": 2213, + "CACCAT": 2214, + "CACCAC": 2215, + "CACCAG": 2216, + "CACCTA": 2217, + "CACCTT": 2218, + "CACCTC": 2219, + "CACCTG": 2220, + "CACCCA": 2221, + "CACCCT": 2222, + "CACCCC": 2223, + "CACCCG": 2224, + "CACCGA": 2225, + "CACCGT": 2226, + "CACCGC": 2227, + "CACCGG": 2228, + "CACGAA": 2229, + "CACGAT": 2230, + "CACGAC": 2231, + "CACGAG": 2232, + "CACGTA": 2233, + "CACGTT": 2234, + "CACGTC": 2235, + "CACGTG": 2236, + "CACGCA": 2237, + "CACGCT": 2238, + "CACGCC": 2239, + "CACGCG": 2240, + "CACGGA": 2241, + "CACGGT": 2242, + "CACGGC": 2243, + "CACGGG": 2244, + "CAGAAA": 2245, + "CAGAAT": 2246, + "CAGAAC": 2247, + "CAGAAG": 2248, + "CAGATA": 2249, + "CAGATT": 2250, + "CAGATC": 2251, + "CAGATG": 2252, + "CAGACA": 2253, + "CAGACT": 2254, + "CAGACC": 2255, + "CAGACG": 2256, + "CAGAGA": 2257, + "CAGAGT": 2258, + "CAGAGC": 2259, + "CAGAGG": 2260, + "CAGTAA": 2261, + "CAGTAT": 2262, + "CAGTAC": 2263, + "CAGTAG": 2264, + "CAGTTA": 2265, + "CAGTTT": 2266, + "CAGTTC": 2267, + "CAGTTG": 2268, + "CAGTCA": 2269, + "CAGTCT": 2270, + "CAGTCC": 2271, + "CAGTCG": 2272, + "CAGTGA": 2273, + "CAGTGT": 2274, + "CAGTGC": 2275, + "CAGTGG": 2276, + "CAGCAA": 2277, + "CAGCAT": 2278, + "CAGCAC": 2279, + "CAGCAG": 2280, + "CAGCTA": 2281, + "CAGCTT": 2282, + "CAGCTC": 2283, + "CAGCTG": 2284, + "CAGCCA": 2285, + "CAGCCT": 2286, + "CAGCCC": 2287, + "CAGCCG": 2288, + "CAGCGA": 2289, + "CAGCGT": 2290, + "CAGCGC": 2291, + "CAGCGG": 2292, + "CAGGAA": 2293, + "CAGGAT": 2294, + "CAGGAC": 2295, + "CAGGAG": 2296, + "CAGGTA": 2297, + "CAGGTT": 2298, + "CAGGTC": 2299, + "CAGGTG": 2300, + "CAGGCA": 2301, + "CAGGCT": 2302, + "CAGGCC": 2303, + "CAGGCG": 2304, + "CAGGGA": 2305, + "CAGGGT": 2306, + "CAGGGC": 2307, + "CAGGGG": 2308, + "CTAAAA": 2309, + "CTAAAT": 2310, + "CTAAAC": 2311, + "CTAAAG": 2312, + "CTAATA": 2313, + "CTAATT": 2314, + "CTAATC": 2315, + "CTAATG": 2316, + "CTAACA": 2317, + "CTAACT": 2318, + "CTAACC": 2319, + "CTAACG": 2320, + "CTAAGA": 2321, + "CTAAGT": 2322, + "CTAAGC": 2323, + "CTAAGG": 2324, + "CTATAA": 2325, + "CTATAT": 2326, + "CTATAC": 2327, + "CTATAG": 2328, + "CTATTA": 2329, + "CTATTT": 2330, + "CTATTC": 2331, + "CTATTG": 2332, + "CTATCA": 2333, + "CTATCT": 2334, + "CTATCC": 2335, + "CTATCG": 2336, + "CTATGA": 2337, + "CTATGT": 2338, + "CTATGC": 2339, + "CTATGG": 2340, + "CTACAA": 2341, + "CTACAT": 2342, + "CTACAC": 2343, + "CTACAG": 2344, + "CTACTA": 2345, + "CTACTT": 2346, + "CTACTC": 2347, + "CTACTG": 2348, + "CTACCA": 2349, + "CTACCT": 2350, + "CTACCC": 2351, + "CTACCG": 2352, + "CTACGA": 2353, + "CTACGT": 2354, + "CTACGC": 2355, + "CTACGG": 2356, + "CTAGAA": 2357, + "CTAGAT": 2358, + "CTAGAC": 2359, + "CTAGAG": 2360, + "CTAGTA": 2361, + "CTAGTT": 2362, + "CTAGTC": 2363, + "CTAGTG": 2364, + "CTAGCA": 2365, + "CTAGCT": 2366, + "CTAGCC": 2367, + "CTAGCG": 2368, + "CTAGGA": 2369, + "CTAGGT": 2370, + "CTAGGC": 2371, + "CTAGGG": 2372, + "CTTAAA": 2373, + "CTTAAT": 2374, + "CTTAAC": 2375, + "CTTAAG": 2376, + "CTTATA": 2377, + "CTTATT": 2378, + "CTTATC": 2379, + "CTTATG": 2380, + "CTTACA": 2381, + "CTTACT": 2382, + "CTTACC": 2383, + "CTTACG": 2384, + "CTTAGA": 2385, + "CTTAGT": 2386, + "CTTAGC": 2387, + "CTTAGG": 2388, + "CTTTAA": 2389, + "CTTTAT": 2390, + "CTTTAC": 2391, + "CTTTAG": 2392, + "CTTTTA": 2393, + "CTTTTT": 2394, + "CTTTTC": 2395, + "CTTTTG": 2396, + "CTTTCA": 2397, + "CTTTCT": 2398, + "CTTTCC": 2399, + "CTTTCG": 2400, + "CTTTGA": 2401, + "CTTTGT": 2402, + "CTTTGC": 2403, + "CTTTGG": 2404, + "CTTCAA": 2405, + "CTTCAT": 2406, + "CTTCAC": 2407, + "CTTCAG": 2408, + "CTTCTA": 2409, + "CTTCTT": 2410, + "CTTCTC": 2411, + "CTTCTG": 2412, + "CTTCCA": 2413, + "CTTCCT": 2414, + "CTTCCC": 2415, + "CTTCCG": 2416, + "CTTCGA": 2417, + "CTTCGT": 2418, + "CTTCGC": 2419, + "CTTCGG": 2420, + "CTTGAA": 2421, + "CTTGAT": 2422, + "CTTGAC": 2423, + "CTTGAG": 2424, + "CTTGTA": 2425, + "CTTGTT": 2426, + "CTTGTC": 2427, + "CTTGTG": 2428, + "CTTGCA": 2429, + "CTTGCT": 2430, + "CTTGCC": 2431, + "CTTGCG": 2432, + "CTTGGA": 2433, + "CTTGGT": 2434, + "CTTGGC": 2435, + "CTTGGG": 2436, + "CTCAAA": 2437, + "CTCAAT": 2438, + "CTCAAC": 2439, + "CTCAAG": 2440, + "CTCATA": 2441, + "CTCATT": 2442, + "CTCATC": 2443, + "CTCATG": 2444, + "CTCACA": 2445, + "CTCACT": 2446, + "CTCACC": 2447, + "CTCACG": 2448, + "CTCAGA": 2449, + "CTCAGT": 2450, + "CTCAGC": 2451, + "CTCAGG": 2452, + "CTCTAA": 2453, + "CTCTAT": 2454, + "CTCTAC": 2455, + "CTCTAG": 2456, + "CTCTTA": 2457, + "CTCTTT": 2458, + "CTCTTC": 2459, + "CTCTTG": 2460, + "CTCTCA": 2461, + "CTCTCT": 2462, + "CTCTCC": 2463, + "CTCTCG": 2464, + "CTCTGA": 2465, + "CTCTGT": 2466, + "CTCTGC": 2467, + "CTCTGG": 2468, + "CTCCAA": 2469, + "CTCCAT": 2470, + "CTCCAC": 2471, + "CTCCAG": 2472, + "CTCCTA": 2473, + "CTCCTT": 2474, + "CTCCTC": 2475, + "CTCCTG": 2476, + "CTCCCA": 2477, + "CTCCCT": 2478, + "CTCCCC": 2479, + "CTCCCG": 2480, + "CTCCGA": 2481, + "CTCCGT": 2482, + "CTCCGC": 2483, + "CTCCGG": 2484, + "CTCGAA": 2485, + "CTCGAT": 2486, + "CTCGAC": 2487, + "CTCGAG": 2488, + "CTCGTA": 2489, + "CTCGTT": 2490, + "CTCGTC": 2491, + "CTCGTG": 2492, + "CTCGCA": 2493, + "CTCGCT": 2494, + "CTCGCC": 2495, + "CTCGCG": 2496, + "CTCGGA": 2497, + "CTCGGT": 2498, + "CTCGGC": 2499, + "CTCGGG": 2500, + "CTGAAA": 2501, + "CTGAAT": 2502, + "CTGAAC": 2503, + "CTGAAG": 2504, + "CTGATA": 2505, + "CTGATT": 2506, + "CTGATC": 2507, + "CTGATG": 2508, + "CTGACA": 2509, + "CTGACT": 2510, + "CTGACC": 2511, + "CTGACG": 2512, + "CTGAGA": 2513, + "CTGAGT": 2514, + "CTGAGC": 2515, + "CTGAGG": 2516, + "CTGTAA": 2517, + "CTGTAT": 2518, + "CTGTAC": 2519, + "CTGTAG": 2520, + "CTGTTA": 2521, + "CTGTTT": 2522, + "CTGTTC": 2523, + "CTGTTG": 2524, + "CTGTCA": 2525, + "CTGTCT": 2526, + "CTGTCC": 2527, + "CTGTCG": 2528, + "CTGTGA": 2529, + "CTGTGT": 2530, + "CTGTGC": 2531, + "CTGTGG": 2532, + "CTGCAA": 2533, + "CTGCAT": 2534, + "CTGCAC": 2535, + "CTGCAG": 2536, + "CTGCTA": 2537, + "CTGCTT": 2538, + "CTGCTC": 2539, + "CTGCTG": 2540, + "CTGCCA": 2541, + "CTGCCT": 2542, + "CTGCCC": 2543, + "CTGCCG": 2544, + "CTGCGA": 2545, + "CTGCGT": 2546, + "CTGCGC": 2547, + "CTGCGG": 2548, + "CTGGAA": 2549, + "CTGGAT": 2550, + "CTGGAC": 2551, + "CTGGAG": 2552, + "CTGGTA": 2553, + "CTGGTT": 2554, + "CTGGTC": 2555, + "CTGGTG": 2556, + "CTGGCA": 2557, + "CTGGCT": 2558, + "CTGGCC": 2559, + "CTGGCG": 2560, + "CTGGGA": 2561, + "CTGGGT": 2562, + "CTGGGC": 2563, + "CTGGGG": 2564, + "CCAAAA": 2565, + "CCAAAT": 2566, + "CCAAAC": 2567, + "CCAAAG": 2568, + "CCAATA": 2569, + "CCAATT": 2570, + "CCAATC": 2571, + "CCAATG": 2572, + "CCAACA": 2573, + "CCAACT": 2574, + "CCAACC": 2575, + "CCAACG": 2576, + "CCAAGA": 2577, + "CCAAGT": 2578, + "CCAAGC": 2579, + "CCAAGG": 2580, + "CCATAA": 2581, + "CCATAT": 2582, + "CCATAC": 2583, + "CCATAG": 2584, + "CCATTA": 2585, + "CCATTT": 2586, + "CCATTC": 2587, + "CCATTG": 2588, + "CCATCA": 2589, + "CCATCT": 2590, + "CCATCC": 2591, + "CCATCG": 2592, + "CCATGA": 2593, + "CCATGT": 2594, + "CCATGC": 2595, + "CCATGG": 2596, + "CCACAA": 2597, + "CCACAT": 2598, + "CCACAC": 2599, + "CCACAG": 2600, + "CCACTA": 2601, + "CCACTT": 2602, + "CCACTC": 2603, + "CCACTG": 2604, + "CCACCA": 2605, + "CCACCT": 2606, + "CCACCC": 2607, + "CCACCG": 2608, + "CCACGA": 2609, + "CCACGT": 2610, + "CCACGC": 2611, + "CCACGG": 2612, + "CCAGAA": 2613, + "CCAGAT": 2614, + "CCAGAC": 2615, + "CCAGAG": 2616, + "CCAGTA": 2617, + "CCAGTT": 2618, + "CCAGTC": 2619, + "CCAGTG": 2620, + "CCAGCA": 2621, + "CCAGCT": 2622, + "CCAGCC": 2623, + "CCAGCG": 2624, + "CCAGGA": 2625, + "CCAGGT": 2626, + "CCAGGC": 2627, + "CCAGGG": 2628, + "CCTAAA": 2629, + "CCTAAT": 2630, + "CCTAAC": 2631, + "CCTAAG": 2632, + "CCTATA": 2633, + "CCTATT": 2634, + "CCTATC": 2635, + "CCTATG": 2636, + "CCTACA": 2637, + "CCTACT": 2638, + "CCTACC": 2639, + "CCTACG": 2640, + "CCTAGA": 2641, + "CCTAGT": 2642, + "CCTAGC": 2643, + "CCTAGG": 2644, + "CCTTAA": 2645, + "CCTTAT": 2646, + "CCTTAC": 2647, + "CCTTAG": 2648, + "CCTTTA": 2649, + "CCTTTT": 2650, + "CCTTTC": 2651, + "CCTTTG": 2652, + "CCTTCA": 2653, + "CCTTCT": 2654, + "CCTTCC": 2655, + "CCTTCG": 2656, + "CCTTGA": 2657, + "CCTTGT": 2658, + "CCTTGC": 2659, + "CCTTGG": 2660, + "CCTCAA": 2661, + "CCTCAT": 2662, + "CCTCAC": 2663, + "CCTCAG": 2664, + "CCTCTA": 2665, + "CCTCTT": 2666, + "CCTCTC": 2667, + "CCTCTG": 2668, + "CCTCCA": 2669, + "CCTCCT": 2670, + "CCTCCC": 2671, + "CCTCCG": 2672, + "CCTCGA": 2673, + "CCTCGT": 2674, + "CCTCGC": 2675, + "CCTCGG": 2676, + "CCTGAA": 2677, + "CCTGAT": 2678, + "CCTGAC": 2679, + "CCTGAG": 2680, + "CCTGTA": 2681, + "CCTGTT": 2682, + "CCTGTC": 2683, + "CCTGTG": 2684, + "CCTGCA": 2685, + "CCTGCT": 2686, + "CCTGCC": 2687, + "CCTGCG": 2688, + "CCTGGA": 2689, + "CCTGGT": 2690, + "CCTGGC": 2691, + "CCTGGG": 2692, + "CCCAAA": 2693, + "CCCAAT": 2694, + "CCCAAC": 2695, + "CCCAAG": 2696, + "CCCATA": 2697, + "CCCATT": 2698, + "CCCATC": 2699, + "CCCATG": 2700, + "CCCACA": 2701, + "CCCACT": 2702, + "CCCACC": 2703, + "CCCACG": 2704, + "CCCAGA": 2705, + "CCCAGT": 2706, + "CCCAGC": 2707, + "CCCAGG": 2708, + "CCCTAA": 2709, + "CCCTAT": 2710, + "CCCTAC": 2711, + "CCCTAG": 2712, + "CCCTTA": 2713, + "CCCTTT": 2714, + "CCCTTC": 2715, + "CCCTTG": 2716, + "CCCTCA": 2717, + "CCCTCT": 2718, + "CCCTCC": 2719, + "CCCTCG": 2720, + "CCCTGA": 2721, + "CCCTGT": 2722, + "CCCTGC": 2723, + "CCCTGG": 2724, + "CCCCAA": 2725, + "CCCCAT": 2726, + "CCCCAC": 2727, + "CCCCAG": 2728, + "CCCCTA": 2729, + "CCCCTT": 2730, + "CCCCTC": 2731, + "CCCCTG": 2732, + "CCCCCA": 2733, + "CCCCCT": 2734, + "CCCCCC": 2735, + "CCCCCG": 2736, + "CCCCGA": 2737, + "CCCCGT": 2738, + "CCCCGC": 2739, + "CCCCGG": 2740, + "CCCGAA": 2741, + "CCCGAT": 2742, + "CCCGAC": 2743, + "CCCGAG": 2744, + "CCCGTA": 2745, + "CCCGTT": 2746, + "CCCGTC": 2747, + "CCCGTG": 2748, + "CCCGCA": 2749, + "CCCGCT": 2750, + "CCCGCC": 2751, + "CCCGCG": 2752, + "CCCGGA": 2753, + "CCCGGT": 2754, + "CCCGGC": 2755, + "CCCGGG": 2756, + "CCGAAA": 2757, + "CCGAAT": 2758, + "CCGAAC": 2759, + "CCGAAG": 2760, + "CCGATA": 2761, + "CCGATT": 2762, + "CCGATC": 2763, + "CCGATG": 2764, + "CCGACA": 2765, + "CCGACT": 2766, + "CCGACC": 2767, + "CCGACG": 2768, + "CCGAGA": 2769, + "CCGAGT": 2770, + "CCGAGC": 2771, + "CCGAGG": 2772, + "CCGTAA": 2773, + "CCGTAT": 2774, + "CCGTAC": 2775, + "CCGTAG": 2776, + "CCGTTA": 2777, + "CCGTTT": 2778, + "CCGTTC": 2779, + "CCGTTG": 2780, + "CCGTCA": 2781, + "CCGTCT": 2782, + "CCGTCC": 2783, + "CCGTCG": 2784, + "CCGTGA": 2785, + "CCGTGT": 2786, + "CCGTGC": 2787, + "CCGTGG": 2788, + "CCGCAA": 2789, + "CCGCAT": 2790, + "CCGCAC": 2791, + "CCGCAG": 2792, + "CCGCTA": 2793, + "CCGCTT": 2794, + "CCGCTC": 2795, + "CCGCTG": 2796, + "CCGCCA": 2797, + "CCGCCT": 2798, + "CCGCCC": 2799, + "CCGCCG": 2800, + "CCGCGA": 2801, + "CCGCGT": 2802, + "CCGCGC": 2803, + "CCGCGG": 2804, + "CCGGAA": 2805, + "CCGGAT": 2806, + "CCGGAC": 2807, + "CCGGAG": 2808, + "CCGGTA": 2809, + "CCGGTT": 2810, + "CCGGTC": 2811, + "CCGGTG": 2812, + "CCGGCA": 2813, + "CCGGCT": 2814, + "CCGGCC": 2815, + "CCGGCG": 2816, + "CCGGGA": 2817, + "CCGGGT": 2818, + "CCGGGC": 2819, + "CCGGGG": 2820, + "CGAAAA": 2821, + "CGAAAT": 2822, + "CGAAAC": 2823, + "CGAAAG": 2824, + "CGAATA": 2825, + "CGAATT": 2826, + "CGAATC": 2827, + "CGAATG": 2828, + "CGAACA": 2829, + "CGAACT": 2830, + "CGAACC": 2831, + "CGAACG": 2832, + "CGAAGA": 2833, + "CGAAGT": 2834, + "CGAAGC": 2835, + "CGAAGG": 2836, + "CGATAA": 2837, + "CGATAT": 2838, + "CGATAC": 2839, + "CGATAG": 2840, + "CGATTA": 2841, + "CGATTT": 2842, + "CGATTC": 2843, + "CGATTG": 2844, + "CGATCA": 2845, + "CGATCT": 2846, + "CGATCC": 2847, + "CGATCG": 2848, + "CGATGA": 2849, + "CGATGT": 2850, + "CGATGC": 2851, + "CGATGG": 2852, + "CGACAA": 2853, + "CGACAT": 2854, + "CGACAC": 2855, + "CGACAG": 2856, + "CGACTA": 2857, + "CGACTT": 2858, + "CGACTC": 2859, + "CGACTG": 2860, + "CGACCA": 2861, + "CGACCT": 2862, + "CGACCC": 2863, + "CGACCG": 2864, + "CGACGA": 2865, + "CGACGT": 2866, + "CGACGC": 2867, + "CGACGG": 2868, + "CGAGAA": 2869, + "CGAGAT": 2870, + "CGAGAC": 2871, + "CGAGAG": 2872, + "CGAGTA": 2873, + "CGAGTT": 2874, + "CGAGTC": 2875, + "CGAGTG": 2876, + "CGAGCA": 2877, + "CGAGCT": 2878, + "CGAGCC": 2879, + "CGAGCG": 2880, + "CGAGGA": 2881, + "CGAGGT": 2882, + "CGAGGC": 2883, + "CGAGGG": 2884, + "CGTAAA": 2885, + "CGTAAT": 2886, + "CGTAAC": 2887, + "CGTAAG": 2888, + "CGTATA": 2889, + "CGTATT": 2890, + "CGTATC": 2891, + "CGTATG": 2892, + "CGTACA": 2893, + "CGTACT": 2894, + "CGTACC": 2895, + "CGTACG": 2896, + "CGTAGA": 2897, + "CGTAGT": 2898, + "CGTAGC": 2899, + "CGTAGG": 2900, + "CGTTAA": 2901, + "CGTTAT": 2902, + "CGTTAC": 2903, + "CGTTAG": 2904, + "CGTTTA": 2905, + "CGTTTT": 2906, + "CGTTTC": 2907, + "CGTTTG": 2908, + "CGTTCA": 2909, + "CGTTCT": 2910, + "CGTTCC": 2911, + "CGTTCG": 2912, + "CGTTGA": 2913, + "CGTTGT": 2914, + "CGTTGC": 2915, + "CGTTGG": 2916, + "CGTCAA": 2917, + "CGTCAT": 2918, + "CGTCAC": 2919, + "CGTCAG": 2920, + "CGTCTA": 2921, + "CGTCTT": 2922, + "CGTCTC": 2923, + "CGTCTG": 2924, + "CGTCCA": 2925, + "CGTCCT": 2926, + "CGTCCC": 2927, + "CGTCCG": 2928, + "CGTCGA": 2929, + "CGTCGT": 2930, + "CGTCGC": 2931, + "CGTCGG": 2932, + "CGTGAA": 2933, + "CGTGAT": 2934, + "CGTGAC": 2935, + "CGTGAG": 2936, + "CGTGTA": 2937, + "CGTGTT": 2938, + "CGTGTC": 2939, + "CGTGTG": 2940, + "CGTGCA": 2941, + "CGTGCT": 2942, + "CGTGCC": 2943, + "CGTGCG": 2944, + "CGTGGA": 2945, + "CGTGGT": 2946, + "CGTGGC": 2947, + "CGTGGG": 2948, + "CGCAAA": 2949, + "CGCAAT": 2950, + "CGCAAC": 2951, + "CGCAAG": 2952, + "CGCATA": 2953, + "CGCATT": 2954, + "CGCATC": 2955, + "CGCATG": 2956, + "CGCACA": 2957, + "CGCACT": 2958, + "CGCACC": 2959, + "CGCACG": 2960, + "CGCAGA": 2961, + "CGCAGT": 2962, + "CGCAGC": 2963, + "CGCAGG": 2964, + "CGCTAA": 2965, + "CGCTAT": 2966, + "CGCTAC": 2967, + "CGCTAG": 2968, + "CGCTTA": 2969, + "CGCTTT": 2970, + "CGCTTC": 2971, + "CGCTTG": 2972, + "CGCTCA": 2973, + "CGCTCT": 2974, + "CGCTCC": 2975, + "CGCTCG": 2976, + "CGCTGA": 2977, + "CGCTGT": 2978, + "CGCTGC": 2979, + "CGCTGG": 2980, + "CGCCAA": 2981, + "CGCCAT": 2982, + "CGCCAC": 2983, + "CGCCAG": 2984, + "CGCCTA": 2985, + "CGCCTT": 2986, + "CGCCTC": 2987, + "CGCCTG": 2988, + "CGCCCA": 2989, + "CGCCCT": 2990, + "CGCCCC": 2991, + "CGCCCG": 2992, + "CGCCGA": 2993, + "CGCCGT": 2994, + "CGCCGC": 2995, + "CGCCGG": 2996, + "CGCGAA": 2997, + "CGCGAT": 2998, + "CGCGAC": 2999, + "CGCGAG": 3000, + "CGCGTA": 3001, + "CGCGTT": 3002, + "CGCGTC": 3003, + "CGCGTG": 3004, + "CGCGCA": 3005, + "CGCGCT": 3006, + "CGCGCC": 3007, + "CGCGCG": 3008, + "CGCGGA": 3009, + "CGCGGT": 3010, + "CGCGGC": 3011, + "CGCGGG": 3012, + "CGGAAA": 3013, + "CGGAAT": 3014, + "CGGAAC": 3015, + "CGGAAG": 3016, + "CGGATA": 3017, + "CGGATT": 3018, + "CGGATC": 3019, + "CGGATG": 3020, + "CGGACA": 3021, + "CGGACT": 3022, + "CGGACC": 3023, + "CGGACG": 3024, + "CGGAGA": 3025, + "CGGAGT": 3026, + "CGGAGC": 3027, + "CGGAGG": 3028, + "CGGTAA": 3029, + "CGGTAT": 3030, + "CGGTAC": 3031, + "CGGTAG": 3032, + "CGGTTA": 3033, + "CGGTTT": 3034, + "CGGTTC": 3035, + "CGGTTG": 3036, + "CGGTCA": 3037, + "CGGTCT": 3038, + "CGGTCC": 3039, + "CGGTCG": 3040, + "CGGTGA": 3041, + "CGGTGT": 3042, + "CGGTGC": 3043, + "CGGTGG": 3044, + "CGGCAA": 3045, + "CGGCAT": 3046, + "CGGCAC": 3047, + "CGGCAG": 3048, + "CGGCTA": 3049, + "CGGCTT": 3050, + "CGGCTC": 3051, + "CGGCTG": 3052, + "CGGCCA": 3053, + "CGGCCT": 3054, + "CGGCCC": 3055, + "CGGCCG": 3056, + "CGGCGA": 3057, + "CGGCGT": 3058, + "CGGCGC": 3059, + "CGGCGG": 3060, + "CGGGAA": 3061, + "CGGGAT": 3062, + "CGGGAC": 3063, + "CGGGAG": 3064, + "CGGGTA": 3065, + "CGGGTT": 3066, + "CGGGTC": 3067, + "CGGGTG": 3068, + "CGGGCA": 3069, + "CGGGCT": 3070, + "CGGGCC": 3071, + "CGGGCG": 3072, + "CGGGGA": 3073, + "CGGGGT": 3074, + "CGGGGC": 3075, + "CGGGGG": 3076, + "GAAAAA": 3077, + "GAAAAT": 3078, + "GAAAAC": 3079, + "GAAAAG": 3080, + "GAAATA": 3081, + "GAAATT": 3082, + "GAAATC": 3083, + "GAAATG": 3084, + "GAAACA": 3085, + "GAAACT": 3086, + "GAAACC": 3087, + "GAAACG": 3088, + "GAAAGA": 3089, + "GAAAGT": 3090, + "GAAAGC": 3091, + "GAAAGG": 3092, + "GAATAA": 3093, + "GAATAT": 3094, + "GAATAC": 3095, + "GAATAG": 3096, + "GAATTA": 3097, + "GAATTT": 3098, + "GAATTC": 3099, + "GAATTG": 3100, + "GAATCA": 3101, + "GAATCT": 3102, + "GAATCC": 3103, + "GAATCG": 3104, + "GAATGA": 3105, + "GAATGT": 3106, + "GAATGC": 3107, + "GAATGG": 3108, + "GAACAA": 3109, + "GAACAT": 3110, + "GAACAC": 3111, + "GAACAG": 3112, + "GAACTA": 3113, + "GAACTT": 3114, + "GAACTC": 3115, + "GAACTG": 3116, + "GAACCA": 3117, + "GAACCT": 3118, + "GAACCC": 3119, + "GAACCG": 3120, + "GAACGA": 3121, + "GAACGT": 3122, + "GAACGC": 3123, + "GAACGG": 3124, + "GAAGAA": 3125, + "GAAGAT": 3126, + "GAAGAC": 3127, + "GAAGAG": 3128, + "GAAGTA": 3129, + "GAAGTT": 3130, + "GAAGTC": 3131, + "GAAGTG": 3132, + "GAAGCA": 3133, + "GAAGCT": 3134, + "GAAGCC": 3135, + "GAAGCG": 3136, + "GAAGGA": 3137, + "GAAGGT": 3138, + "GAAGGC": 3139, + "GAAGGG": 3140, + "GATAAA": 3141, + "GATAAT": 3142, + "GATAAC": 3143, + "GATAAG": 3144, + "GATATA": 3145, + "GATATT": 3146, + "GATATC": 3147, + "GATATG": 3148, + "GATACA": 3149, + "GATACT": 3150, + "GATACC": 3151, + "GATACG": 3152, + "GATAGA": 3153, + "GATAGT": 3154, + "GATAGC": 3155, + "GATAGG": 3156, + "GATTAA": 3157, + "GATTAT": 3158, + "GATTAC": 3159, + "GATTAG": 3160, + "GATTTA": 3161, + "GATTTT": 3162, + "GATTTC": 3163, + "GATTTG": 3164, + "GATTCA": 3165, + "GATTCT": 3166, + "GATTCC": 3167, + "GATTCG": 3168, + "GATTGA": 3169, + "GATTGT": 3170, + "GATTGC": 3171, + "GATTGG": 3172, + "GATCAA": 3173, + "GATCAT": 3174, + "GATCAC": 3175, + "GATCAG": 3176, + "GATCTA": 3177, + "GATCTT": 3178, + "GATCTC": 3179, + "GATCTG": 3180, + "GATCCA": 3181, + "GATCCT": 3182, + "GATCCC": 3183, + "GATCCG": 3184, + "GATCGA": 3185, + "GATCGT": 3186, + "GATCGC": 3187, + "GATCGG": 3188, + "GATGAA": 3189, + "GATGAT": 3190, + "GATGAC": 3191, + "GATGAG": 3192, + "GATGTA": 3193, + "GATGTT": 3194, + "GATGTC": 3195, + "GATGTG": 3196, + "GATGCA": 3197, + "GATGCT": 3198, + "GATGCC": 3199, + "GATGCG": 3200, + "GATGGA": 3201, + "GATGGT": 3202, + "GATGGC": 3203, + "GATGGG": 3204, + "GACAAA": 3205, + "GACAAT": 3206, + "GACAAC": 3207, + "GACAAG": 3208, + "GACATA": 3209, + "GACATT": 3210, + "GACATC": 3211, + "GACATG": 3212, + "GACACA": 3213, + "GACACT": 3214, + "GACACC": 3215, + "GACACG": 3216, + "GACAGA": 3217, + "GACAGT": 3218, + "GACAGC": 3219, + "GACAGG": 3220, + "GACTAA": 3221, + "GACTAT": 3222, + "GACTAC": 3223, + "GACTAG": 3224, + "GACTTA": 3225, + "GACTTT": 3226, + "GACTTC": 3227, + "GACTTG": 3228, + "GACTCA": 3229, + "GACTCT": 3230, + "GACTCC": 3231, + "GACTCG": 3232, + "GACTGA": 3233, + "GACTGT": 3234, + "GACTGC": 3235, + "GACTGG": 3236, + "GACCAA": 3237, + "GACCAT": 3238, + "GACCAC": 3239, + "GACCAG": 3240, + "GACCTA": 3241, + "GACCTT": 3242, + "GACCTC": 3243, + "GACCTG": 3244, + "GACCCA": 3245, + "GACCCT": 3246, + "GACCCC": 3247, + "GACCCG": 3248, + "GACCGA": 3249, + "GACCGT": 3250, + "GACCGC": 3251, + "GACCGG": 3252, + "GACGAA": 3253, + "GACGAT": 3254, + "GACGAC": 3255, + "GACGAG": 3256, + "GACGTA": 3257, + "GACGTT": 3258, + "GACGTC": 3259, + "GACGTG": 3260, + "GACGCA": 3261, + "GACGCT": 3262, + "GACGCC": 3263, + "GACGCG": 3264, + "GACGGA": 3265, + "GACGGT": 3266, + "GACGGC": 3267, + "GACGGG": 3268, + "GAGAAA": 3269, + "GAGAAT": 3270, + "GAGAAC": 3271, + "GAGAAG": 3272, + "GAGATA": 3273, + "GAGATT": 3274, + "GAGATC": 3275, + "GAGATG": 3276, + "GAGACA": 3277, + "GAGACT": 3278, + "GAGACC": 3279, + "GAGACG": 3280, + "GAGAGA": 3281, + "GAGAGT": 3282, + "GAGAGC": 3283, + "GAGAGG": 3284, + "GAGTAA": 3285, + "GAGTAT": 3286, + "GAGTAC": 3287, + "GAGTAG": 3288, + "GAGTTA": 3289, + "GAGTTT": 3290, + "GAGTTC": 3291, + "GAGTTG": 3292, + "GAGTCA": 3293, + "GAGTCT": 3294, + "GAGTCC": 3295, + "GAGTCG": 3296, + "GAGTGA": 3297, + "GAGTGT": 3298, + "GAGTGC": 3299, + "GAGTGG": 3300, + "GAGCAA": 3301, + "GAGCAT": 3302, + "GAGCAC": 3303, + "GAGCAG": 3304, + "GAGCTA": 3305, + "GAGCTT": 3306, + "GAGCTC": 3307, + "GAGCTG": 3308, + "GAGCCA": 3309, + "GAGCCT": 3310, + "GAGCCC": 3311, + "GAGCCG": 3312, + "GAGCGA": 3313, + "GAGCGT": 3314, + "GAGCGC": 3315, + "GAGCGG": 3316, + "GAGGAA": 3317, + "GAGGAT": 3318, + "GAGGAC": 3319, + "GAGGAG": 3320, + "GAGGTA": 3321, + "GAGGTT": 3322, + "GAGGTC": 3323, + "GAGGTG": 3324, + "GAGGCA": 3325, + "GAGGCT": 3326, + "GAGGCC": 3327, + "GAGGCG": 3328, + "GAGGGA": 3329, + "GAGGGT": 3330, + "GAGGGC": 3331, + "GAGGGG": 3332, + "GTAAAA": 3333, + "GTAAAT": 3334, + "GTAAAC": 3335, + "GTAAAG": 3336, + "GTAATA": 3337, + "GTAATT": 3338, + "GTAATC": 3339, + "GTAATG": 3340, + "GTAACA": 3341, + "GTAACT": 3342, + "GTAACC": 3343, + "GTAACG": 3344, + "GTAAGA": 3345, + "GTAAGT": 3346, + "GTAAGC": 3347, + "GTAAGG": 3348, + "GTATAA": 3349, + "GTATAT": 3350, + "GTATAC": 3351, + "GTATAG": 3352, + "GTATTA": 3353, + "GTATTT": 3354, + "GTATTC": 3355, + "GTATTG": 3356, + "GTATCA": 3357, + "GTATCT": 3358, + "GTATCC": 3359, + "GTATCG": 3360, + "GTATGA": 3361, + "GTATGT": 3362, + "GTATGC": 3363, + "GTATGG": 3364, + "GTACAA": 3365, + "GTACAT": 3366, + "GTACAC": 3367, + "GTACAG": 3368, + "GTACTA": 3369, + "GTACTT": 3370, + "GTACTC": 3371, + "GTACTG": 3372, + "GTACCA": 3373, + "GTACCT": 3374, + "GTACCC": 3375, + "GTACCG": 3376, + "GTACGA": 3377, + "GTACGT": 3378, + "GTACGC": 3379, + "GTACGG": 3380, + "GTAGAA": 3381, + "GTAGAT": 3382, + "GTAGAC": 3383, + "GTAGAG": 3384, + "GTAGTA": 3385, + "GTAGTT": 3386, + "GTAGTC": 3387, + "GTAGTG": 3388, + "GTAGCA": 3389, + "GTAGCT": 3390, + "GTAGCC": 3391, + "GTAGCG": 3392, + "GTAGGA": 3393, + "GTAGGT": 3394, + "GTAGGC": 3395, + "GTAGGG": 3396, + "GTTAAA": 3397, + "GTTAAT": 3398, + "GTTAAC": 3399, + "GTTAAG": 3400, + "GTTATA": 3401, + "GTTATT": 3402, + "GTTATC": 3403, + "GTTATG": 3404, + "GTTACA": 3405, + "GTTACT": 3406, + "GTTACC": 3407, + "GTTACG": 3408, + "GTTAGA": 3409, + "GTTAGT": 3410, + "GTTAGC": 3411, + "GTTAGG": 3412, + "GTTTAA": 3413, + "GTTTAT": 3414, + "GTTTAC": 3415, + "GTTTAG": 3416, + "GTTTTA": 3417, + "GTTTTT": 3418, + "GTTTTC": 3419, + "GTTTTG": 3420, + "GTTTCA": 3421, + "GTTTCT": 3422, + "GTTTCC": 3423, + "GTTTCG": 3424, + "GTTTGA": 3425, + "GTTTGT": 3426, + "GTTTGC": 3427, + "GTTTGG": 3428, + "GTTCAA": 3429, + "GTTCAT": 3430, + "GTTCAC": 3431, + "GTTCAG": 3432, + "GTTCTA": 3433, + "GTTCTT": 3434, + "GTTCTC": 3435, + "GTTCTG": 3436, + "GTTCCA": 3437, + "GTTCCT": 3438, + "GTTCCC": 3439, + "GTTCCG": 3440, + "GTTCGA": 3441, + "GTTCGT": 3442, + "GTTCGC": 3443, + "GTTCGG": 3444, + "GTTGAA": 3445, + "GTTGAT": 3446, + "GTTGAC": 3447, + "GTTGAG": 3448, + "GTTGTA": 3449, + "GTTGTT": 3450, + "GTTGTC": 3451, + "GTTGTG": 3452, + "GTTGCA": 3453, + "GTTGCT": 3454, + "GTTGCC": 3455, + "GTTGCG": 3456, + "GTTGGA": 3457, + "GTTGGT": 3458, + "GTTGGC": 3459, + "GTTGGG": 3460, + "GTCAAA": 3461, + "GTCAAT": 3462, + "GTCAAC": 3463, + "GTCAAG": 3464, + "GTCATA": 3465, + "GTCATT": 3466, + "GTCATC": 3467, + "GTCATG": 3468, + "GTCACA": 3469, + "GTCACT": 3470, + "GTCACC": 3471, + "GTCACG": 3472, + "GTCAGA": 3473, + "GTCAGT": 3474, + "GTCAGC": 3475, + "GTCAGG": 3476, + "GTCTAA": 3477, + "GTCTAT": 3478, + "GTCTAC": 3479, + "GTCTAG": 3480, + "GTCTTA": 3481, + "GTCTTT": 3482, + "GTCTTC": 3483, + "GTCTTG": 3484, + "GTCTCA": 3485, + "GTCTCT": 3486, + "GTCTCC": 3487, + "GTCTCG": 3488, + "GTCTGA": 3489, + "GTCTGT": 3490, + "GTCTGC": 3491, + "GTCTGG": 3492, + "GTCCAA": 3493, + "GTCCAT": 3494, + "GTCCAC": 3495, + "GTCCAG": 3496, + "GTCCTA": 3497, + "GTCCTT": 3498, + "GTCCTC": 3499, + "GTCCTG": 3500, + "GTCCCA": 3501, + "GTCCCT": 3502, + "GTCCCC": 3503, + "GTCCCG": 3504, + "GTCCGA": 3505, + "GTCCGT": 3506, + "GTCCGC": 3507, + "GTCCGG": 3508, + "GTCGAA": 3509, + "GTCGAT": 3510, + "GTCGAC": 3511, + "GTCGAG": 3512, + "GTCGTA": 3513, + "GTCGTT": 3514, + "GTCGTC": 3515, + "GTCGTG": 3516, + "GTCGCA": 3517, + "GTCGCT": 3518, + "GTCGCC": 3519, + "GTCGCG": 3520, + "GTCGGA": 3521, + "GTCGGT": 3522, + "GTCGGC": 3523, + "GTCGGG": 3524, + "GTGAAA": 3525, + "GTGAAT": 3526, + "GTGAAC": 3527, + "GTGAAG": 3528, + "GTGATA": 3529, + "GTGATT": 3530, + "GTGATC": 3531, + "GTGATG": 3532, + "GTGACA": 3533, + "GTGACT": 3534, + "GTGACC": 3535, + "GTGACG": 3536, + "GTGAGA": 3537, + "GTGAGT": 3538, + "GTGAGC": 3539, + "GTGAGG": 3540, + "GTGTAA": 3541, + "GTGTAT": 3542, + "GTGTAC": 3543, + "GTGTAG": 3544, + "GTGTTA": 3545, + "GTGTTT": 3546, + "GTGTTC": 3547, + "GTGTTG": 3548, + "GTGTCA": 3549, + "GTGTCT": 3550, + "GTGTCC": 3551, + "GTGTCG": 3552, + "GTGTGA": 3553, + "GTGTGT": 3554, + "GTGTGC": 3555, + "GTGTGG": 3556, + "GTGCAA": 3557, + "GTGCAT": 3558, + "GTGCAC": 3559, + "GTGCAG": 3560, + "GTGCTA": 3561, + "GTGCTT": 3562, + "GTGCTC": 3563, + "GTGCTG": 3564, + "GTGCCA": 3565, + "GTGCCT": 3566, + "GTGCCC": 3567, + "GTGCCG": 3568, + "GTGCGA": 3569, + "GTGCGT": 3570, + "GTGCGC": 3571, + "GTGCGG": 3572, + "GTGGAA": 3573, + "GTGGAT": 3574, + "GTGGAC": 3575, + "GTGGAG": 3576, + "GTGGTA": 3577, + "GTGGTT": 3578, + "GTGGTC": 3579, + "GTGGTG": 3580, + "GTGGCA": 3581, + "GTGGCT": 3582, + "GTGGCC": 3583, + "GTGGCG": 3584, + "GTGGGA": 3585, + "GTGGGT": 3586, + "GTGGGC": 3587, + "GTGGGG": 3588, + "GCAAAA": 3589, + "GCAAAT": 3590, + "GCAAAC": 3591, + "GCAAAG": 3592, + "GCAATA": 3593, + "GCAATT": 3594, + "GCAATC": 3595, + "GCAATG": 3596, + "GCAACA": 3597, + "GCAACT": 3598, + "GCAACC": 3599, + "GCAACG": 3600, + "GCAAGA": 3601, + "GCAAGT": 3602, + "GCAAGC": 3603, + "GCAAGG": 3604, + "GCATAA": 3605, + "GCATAT": 3606, + "GCATAC": 3607, + "GCATAG": 3608, + "GCATTA": 3609, + "GCATTT": 3610, + "GCATTC": 3611, + "GCATTG": 3612, + "GCATCA": 3613, + "GCATCT": 3614, + "GCATCC": 3615, + "GCATCG": 3616, + "GCATGA": 3617, + "GCATGT": 3618, + "GCATGC": 3619, + "GCATGG": 3620, + "GCACAA": 3621, + "GCACAT": 3622, + "GCACAC": 3623, + "GCACAG": 3624, + "GCACTA": 3625, + "GCACTT": 3626, + "GCACTC": 3627, + "GCACTG": 3628, + "GCACCA": 3629, + "GCACCT": 3630, + "GCACCC": 3631, + "GCACCG": 3632, + "GCACGA": 3633, + "GCACGT": 3634, + "GCACGC": 3635, + "GCACGG": 3636, + "GCAGAA": 3637, + "GCAGAT": 3638, + "GCAGAC": 3639, + "GCAGAG": 3640, + "GCAGTA": 3641, + "GCAGTT": 3642, + "GCAGTC": 3643, + "GCAGTG": 3644, + "GCAGCA": 3645, + "GCAGCT": 3646, + "GCAGCC": 3647, + "GCAGCG": 3648, + "GCAGGA": 3649, + "GCAGGT": 3650, + "GCAGGC": 3651, + "GCAGGG": 3652, + "GCTAAA": 3653, + "GCTAAT": 3654, + "GCTAAC": 3655, + "GCTAAG": 3656, + "GCTATA": 3657, + "GCTATT": 3658, + "GCTATC": 3659, + "GCTATG": 3660, + "GCTACA": 3661, + "GCTACT": 3662, + "GCTACC": 3663, + "GCTACG": 3664, + "GCTAGA": 3665, + "GCTAGT": 3666, + "GCTAGC": 3667, + "GCTAGG": 3668, + "GCTTAA": 3669, + "GCTTAT": 3670, + "GCTTAC": 3671, + "GCTTAG": 3672, + "GCTTTA": 3673, + "GCTTTT": 3674, + "GCTTTC": 3675, + "GCTTTG": 3676, + "GCTTCA": 3677, + "GCTTCT": 3678, + "GCTTCC": 3679, + "GCTTCG": 3680, + "GCTTGA": 3681, + "GCTTGT": 3682, + "GCTTGC": 3683, + "GCTTGG": 3684, + "GCTCAA": 3685, + "GCTCAT": 3686, + "GCTCAC": 3687, + "GCTCAG": 3688, + "GCTCTA": 3689, + "GCTCTT": 3690, + "GCTCTC": 3691, + "GCTCTG": 3692, + "GCTCCA": 3693, + "GCTCCT": 3694, + "GCTCCC": 3695, + "GCTCCG": 3696, + "GCTCGA": 3697, + "GCTCGT": 3698, + "GCTCGC": 3699, + "GCTCGG": 3700, + "GCTGAA": 3701, + "GCTGAT": 3702, + "GCTGAC": 3703, + "GCTGAG": 3704, + "GCTGTA": 3705, + "GCTGTT": 3706, + "GCTGTC": 3707, + "GCTGTG": 3708, + "GCTGCA": 3709, + "GCTGCT": 3710, + "GCTGCC": 3711, + "GCTGCG": 3712, + "GCTGGA": 3713, + "GCTGGT": 3714, + "GCTGGC": 3715, + "GCTGGG": 3716, + "GCCAAA": 3717, + "GCCAAT": 3718, + "GCCAAC": 3719, + "GCCAAG": 3720, + "GCCATA": 3721, + "GCCATT": 3722, + "GCCATC": 3723, + "GCCATG": 3724, + "GCCACA": 3725, + "GCCACT": 3726, + "GCCACC": 3727, + "GCCACG": 3728, + "GCCAGA": 3729, + "GCCAGT": 3730, + "GCCAGC": 3731, + "GCCAGG": 3732, + "GCCTAA": 3733, + "GCCTAT": 3734, + "GCCTAC": 3735, + "GCCTAG": 3736, + "GCCTTA": 3737, + "GCCTTT": 3738, + "GCCTTC": 3739, + "GCCTTG": 3740, + "GCCTCA": 3741, + "GCCTCT": 3742, + "GCCTCC": 3743, + "GCCTCG": 3744, + "GCCTGA": 3745, + "GCCTGT": 3746, + "GCCTGC": 3747, + "GCCTGG": 3748, + "GCCCAA": 3749, + "GCCCAT": 3750, + "GCCCAC": 3751, + "GCCCAG": 3752, + "GCCCTA": 3753, + "GCCCTT": 3754, + "GCCCTC": 3755, + "GCCCTG": 3756, + "GCCCCA": 3757, + "GCCCCT": 3758, + "GCCCCC": 3759, + "GCCCCG": 3760, + "GCCCGA": 3761, + "GCCCGT": 3762, + "GCCCGC": 3763, + "GCCCGG": 3764, + "GCCGAA": 3765, + "GCCGAT": 3766, + "GCCGAC": 3767, + "GCCGAG": 3768, + "GCCGTA": 3769, + "GCCGTT": 3770, + "GCCGTC": 3771, + "GCCGTG": 3772, + "GCCGCA": 3773, + "GCCGCT": 3774, + "GCCGCC": 3775, + "GCCGCG": 3776, + "GCCGGA": 3777, + "GCCGGT": 3778, + "GCCGGC": 3779, + "GCCGGG": 3780, + "GCGAAA": 3781, + "GCGAAT": 3782, + "GCGAAC": 3783, + "GCGAAG": 3784, + "GCGATA": 3785, + "GCGATT": 3786, + "GCGATC": 3787, + "GCGATG": 3788, + "GCGACA": 3789, + "GCGACT": 3790, + "GCGACC": 3791, + "GCGACG": 3792, + "GCGAGA": 3793, + "GCGAGT": 3794, + "GCGAGC": 3795, + "GCGAGG": 3796, + "GCGTAA": 3797, + "GCGTAT": 3798, + "GCGTAC": 3799, + "GCGTAG": 3800, + "GCGTTA": 3801, + "GCGTTT": 3802, + "GCGTTC": 3803, + "GCGTTG": 3804, + "GCGTCA": 3805, + "GCGTCT": 3806, + "GCGTCC": 3807, + "GCGTCG": 3808, + "GCGTGA": 3809, + "GCGTGT": 3810, + "GCGTGC": 3811, + "GCGTGG": 3812, + "GCGCAA": 3813, + "GCGCAT": 3814, + "GCGCAC": 3815, + "GCGCAG": 3816, + "GCGCTA": 3817, + "GCGCTT": 3818, + "GCGCTC": 3819, + "GCGCTG": 3820, + "GCGCCA": 3821, + "GCGCCT": 3822, + "GCGCCC": 3823, + "GCGCCG": 3824, + "GCGCGA": 3825, + "GCGCGT": 3826, + "GCGCGC": 3827, + "GCGCGG": 3828, + "GCGGAA": 3829, + "GCGGAT": 3830, + "GCGGAC": 3831, + "GCGGAG": 3832, + "GCGGTA": 3833, + "GCGGTT": 3834, + "GCGGTC": 3835, + "GCGGTG": 3836, + "GCGGCA": 3837, + "GCGGCT": 3838, + "GCGGCC": 3839, + "GCGGCG": 3840, + "GCGGGA": 3841, + "GCGGGT": 3842, + "GCGGGC": 3843, + "GCGGGG": 3844, + "GGAAAA": 3845, + "GGAAAT": 3846, + "GGAAAC": 3847, + "GGAAAG": 3848, + "GGAATA": 3849, + "GGAATT": 3850, + "GGAATC": 3851, + "GGAATG": 3852, + "GGAACA": 3853, + "GGAACT": 3854, + "GGAACC": 3855, + "GGAACG": 3856, + "GGAAGA": 3857, + "GGAAGT": 3858, + "GGAAGC": 3859, + "GGAAGG": 3860, + "GGATAA": 3861, + "GGATAT": 3862, + "GGATAC": 3863, + "GGATAG": 3864, + "GGATTA": 3865, + "GGATTT": 3866, + "GGATTC": 3867, + "GGATTG": 3868, + "GGATCA": 3869, + "GGATCT": 3870, + "GGATCC": 3871, + "GGATCG": 3872, + "GGATGA": 3873, + "GGATGT": 3874, + "GGATGC": 3875, + "GGATGG": 3876, + "GGACAA": 3877, + "GGACAT": 3878, + "GGACAC": 3879, + "GGACAG": 3880, + "GGACTA": 3881, + "GGACTT": 3882, + "GGACTC": 3883, + "GGACTG": 3884, + "GGACCA": 3885, + "GGACCT": 3886, + "GGACCC": 3887, + "GGACCG": 3888, + "GGACGA": 3889, + "GGACGT": 3890, + "GGACGC": 3891, + "GGACGG": 3892, + "GGAGAA": 3893, + "GGAGAT": 3894, + "GGAGAC": 3895, + "GGAGAG": 3896, + "GGAGTA": 3897, + "GGAGTT": 3898, + "GGAGTC": 3899, + "GGAGTG": 3900, + "GGAGCA": 3901, + "GGAGCT": 3902, + "GGAGCC": 3903, + "GGAGCG": 3904, + "GGAGGA": 3905, + "GGAGGT": 3906, + "GGAGGC": 3907, + "GGAGGG": 3908, + "GGTAAA": 3909, + "GGTAAT": 3910, + "GGTAAC": 3911, + "GGTAAG": 3912, + "GGTATA": 3913, + "GGTATT": 3914, + "GGTATC": 3915, + "GGTATG": 3916, + "GGTACA": 3917, + "GGTACT": 3918, + "GGTACC": 3919, + "GGTACG": 3920, + "GGTAGA": 3921, + "GGTAGT": 3922, + "GGTAGC": 3923, + "GGTAGG": 3924, + "GGTTAA": 3925, + "GGTTAT": 3926, + "GGTTAC": 3927, + "GGTTAG": 3928, + "GGTTTA": 3929, + "GGTTTT": 3930, + "GGTTTC": 3931, + "GGTTTG": 3932, + "GGTTCA": 3933, + "GGTTCT": 3934, + "GGTTCC": 3935, + "GGTTCG": 3936, + "GGTTGA": 3937, + "GGTTGT": 3938, + "GGTTGC": 3939, + "GGTTGG": 3940, + "GGTCAA": 3941, + "GGTCAT": 3942, + "GGTCAC": 3943, + "GGTCAG": 3944, + "GGTCTA": 3945, + "GGTCTT": 3946, + "GGTCTC": 3947, + "GGTCTG": 3948, + "GGTCCA": 3949, + "GGTCCT": 3950, + "GGTCCC": 3951, + "GGTCCG": 3952, + "GGTCGA": 3953, + "GGTCGT": 3954, + "GGTCGC": 3955, + "GGTCGG": 3956, + "GGTGAA": 3957, + "GGTGAT": 3958, + "GGTGAC": 3959, + "GGTGAG": 3960, + "GGTGTA": 3961, + "GGTGTT": 3962, + "GGTGTC": 3963, + "GGTGTG": 3964, + "GGTGCA": 3965, + "GGTGCT": 3966, + "GGTGCC": 3967, + "GGTGCG": 3968, + "GGTGGA": 3969, + "GGTGGT": 3970, + "GGTGGC": 3971, + "GGTGGG": 3972, + "GGCAAA": 3973, + "GGCAAT": 3974, + "GGCAAC": 3975, + "GGCAAG": 3976, + "GGCATA": 3977, + "GGCATT": 3978, + "GGCATC": 3979, + "GGCATG": 3980, + "GGCACA": 3981, + "GGCACT": 3982, + "GGCACC": 3983, + "GGCACG": 3984, + "GGCAGA": 3985, + "GGCAGT": 3986, + "GGCAGC": 3987, + "GGCAGG": 3988, + "GGCTAA": 3989, + "GGCTAT": 3990, + "GGCTAC": 3991, + "GGCTAG": 3992, + "GGCTTA": 3993, + "GGCTTT": 3994, + "GGCTTC": 3995, + "GGCTTG": 3996, + "GGCTCA": 3997, + "GGCTCT": 3998, + "GGCTCC": 3999, + "GGCTCG": 4000, + "GGCTGA": 4001, + "GGCTGT": 4002, + "GGCTGC": 4003, + "GGCTGG": 4004, + "GGCCAA": 4005, + "GGCCAT": 4006, + "GGCCAC": 4007, + "GGCCAG": 4008, + "GGCCTA": 4009, + "GGCCTT": 4010, + "GGCCTC": 4011, + "GGCCTG": 4012, + "GGCCCA": 4013, + "GGCCCT": 4014, + "GGCCCC": 4015, + "GGCCCG": 4016, + "GGCCGA": 4017, + "GGCCGT": 4018, + "GGCCGC": 4019, + "GGCCGG": 4020, + "GGCGAA": 4021, + "GGCGAT": 4022, + "GGCGAC": 4023, + "GGCGAG": 4024, + "GGCGTA": 4025, + "GGCGTT": 4026, + "GGCGTC": 4027, + "GGCGTG": 4028, + "GGCGCA": 4029, + "GGCGCT": 4030, + "GGCGCC": 4031, + "GGCGCG": 4032, + "GGCGGA": 4033, + "GGCGGT": 4034, + "GGCGGC": 4035, + "GGCGGG": 4036, + "GGGAAA": 4037, + "GGGAAT": 4038, + "GGGAAC": 4039, + "GGGAAG": 4040, + "GGGATA": 4041, + "GGGATT": 4042, + "GGGATC": 4043, + "GGGATG": 4044, + "GGGACA": 4045, + "GGGACT": 4046, + "GGGACC": 4047, + "GGGACG": 4048, + "GGGAGA": 4049, + "GGGAGT": 4050, + "GGGAGC": 4051, + "GGGAGG": 4052, + "GGGTAA": 4053, + "GGGTAT": 4054, + "GGGTAC": 4055, + "GGGTAG": 4056, + "GGGTTA": 4057, + "GGGTTT": 4058, + "GGGTTC": 4059, + "GGGTTG": 4060, + "GGGTCA": 4061, + "GGGTCT": 4062, + "GGGTCC": 4063, + "GGGTCG": 4064, + "GGGTGA": 4065, + "GGGTGT": 4066, + "GGGTGC": 4067, + "GGGTGG": 4068, + "GGGCAA": 4069, + "GGGCAT": 4070, + "GGGCAC": 4071, + "GGGCAG": 4072, + "GGGCTA": 4073, + "GGGCTT": 4074, + "GGGCTC": 4075, + "GGGCTG": 4076, + "GGGCCA": 4077, + "GGGCCT": 4078, + "GGGCCC": 4079, + "GGGCCG": 4080, + "GGGCGA": 4081, + "GGGCGT": 4082, + "GGGCGC": 4083, + "GGGCGG": 4084, + "GGGGAA": 4085, + "GGGGAT": 4086, + "GGGGAC": 4087, + "GGGGAG": 4088, + "GGGGTA": 4089, + "GGGGTT": 4090, + "GGGGTC": 4091, + "GGGGTG": 4092, + "GGGGCA": 4093, + "GGGGCT": 4094, + "GGGGCC": 4095, + "GGGGCG": 4096, + "GGGGGA": 4097, + "GGGGGT": 4098, + "GGGGGC": 4099, + "GGGGGG": 4100 + } + } +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-1000/tokenizer_config.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-1000/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..604e2b1b29af1dcb3920d2828b6a90276e5e3d02 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-1000/tokenizer_config.json @@ -0,0 +1,58 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "[UNK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "[CLS]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "[SEP]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "4": { + "content": "[MASK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "clean_up_tokenization_spaces": true, + "cls_token": "[CLS]", + "do_basic_tokenize": true, + "do_lower_case": false, + "mask_token": "[MASK]", + "max_len": 512, + "model_max_length": 512, + "never_split": null, + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "strip_accents": null, + "tokenize_chinese_chars": true, + "tokenizer_class": "BertTokenizer", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-1000/trainer_state.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-1000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..0a4426df9efc68545f9c7c8338a5b969bda99e61 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-1000/trainer_state.json @@ -0,0 +1,313 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 4.0, + "eval_steps": 500, + "global_step": 1000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.1, + "grad_norm": 1.1046091318130493, + "learning_rate": 5.319148936170213e-05, + "loss": 0.6562, + "step": 25 + }, + { + "epoch": 0.2, + "grad_norm": 0.4066573977470398, + "learning_rate": 0.00010638297872340425, + "loss": 0.5373, + "step": 50 + }, + { + "epoch": 0.3, + "grad_norm": 0.7340325713157654, + "learning_rate": 0.00015957446808510637, + "loss": 0.4995, + "step": 75 + }, + { + "epoch": 0.4, + "grad_norm": 2.134040117263794, + "learning_rate": 0.0002127659574468085, + "loss": 0.4871, + "step": 100 + }, + { + "epoch": 0.5, + "grad_norm": 2.911975860595703, + "learning_rate": 0.00026595744680851064, + "loss": 0.4507, + "step": 125 + }, + { + "epoch": 0.6, + "grad_norm": 2.040365219116211, + "learning_rate": 0.00031914893617021275, + "loss": 0.3944, + "step": 150 + }, + { + "epoch": 0.7, + "grad_norm": 5.643661022186279, + "learning_rate": 0.0003723404255319149, + "loss": 0.3863, + "step": 175 + }, + { + "epoch": 0.8, + "grad_norm": 3.882492780685425, + "learning_rate": 0.0003992081821181128, + "loss": 0.3875, + "step": 200 + }, + { + "epoch": 0.9, + "grad_norm": 5.87437105178833, + "learning_rate": 0.0003975585615308479, + "loss": 0.3856, + "step": 225 + }, + { + "epoch": 1.0, + "grad_norm": 4.611206531524658, + "learning_rate": 0.00039590894094358297, + "loss": 0.3866, + "step": 250 + }, + { + "epoch": 1.1, + "grad_norm": 3.6404929161071777, + "learning_rate": 0.0003942593203563181, + "loss": 0.3437, + "step": 275 + }, + { + "epoch": 1.2, + "grad_norm": 6.469444751739502, + "learning_rate": 0.0003926096997690532, + "loss": 0.3192, + "step": 300 + }, + { + "epoch": 1.3, + "grad_norm": 3.6251447200775146, + "learning_rate": 0.00039096007918178817, + "loss": 0.349, + "step": 325 + }, + { + "epoch": 1.4, + "grad_norm": 2.854794979095459, + "learning_rate": 0.0003893104585945233, + "loss": 0.3274, + "step": 350 + }, + { + "epoch": 1.5, + "grad_norm": 4.805097579956055, + "learning_rate": 0.0003876608380072583, + "loss": 0.3279, + "step": 375 + }, + { + "epoch": 1.6, + "grad_norm": 2.9891581535339355, + "learning_rate": 0.00038601121741999343, + "loss": 0.3166, + "step": 400 + }, + { + "epoch": 1.7, + "grad_norm": 5.333752155303955, + "learning_rate": 0.0003843615968327285, + "loss": 0.2856, + "step": 425 + }, + { + "epoch": 1.8, + "grad_norm": 2.615621566772461, + "learning_rate": 0.0003827119762454636, + "loss": 0.3286, + "step": 450 + }, + { + "epoch": 1.9, + "grad_norm": 2.4228415489196777, + "learning_rate": 0.00038106235565819863, + "loss": 0.2812, + "step": 475 + }, + { + "epoch": 2.0, + "grad_norm": 3.6015517711639404, + "learning_rate": 0.0003794127350709337, + "loss": 0.2884, + "step": 500 + }, + { + "epoch": 2.1, + "grad_norm": 2.7867860794067383, + "learning_rate": 0.0003777631144836688, + "loss": 0.2582, + "step": 525 + }, + { + "epoch": 2.2, + "grad_norm": 10.079645156860352, + "learning_rate": 0.00037611349389640383, + "loss": 0.245, + "step": 550 + }, + { + "epoch": 2.3, + "grad_norm": 2.475917100906372, + "learning_rate": 0.00037446387330913894, + "loss": 0.2615, + "step": 575 + }, + { + "epoch": 2.4, + "grad_norm": 8.098592758178711, + "learning_rate": 0.000372814252721874, + "loss": 0.2484, + "step": 600 + }, + { + "epoch": 2.5, + "grad_norm": 2.318723201751709, + "learning_rate": 0.00037116463213460903, + "loss": 0.2056, + "step": 625 + }, + { + "epoch": 2.6, + "grad_norm": 3.616283893585205, + "learning_rate": 0.00036951501154734414, + "loss": 0.2358, + "step": 650 + }, + { + "epoch": 2.7, + "grad_norm": 6.419492721557617, + "learning_rate": 0.0003678653909600792, + "loss": 0.2204, + "step": 675 + }, + { + "epoch": 2.8, + "grad_norm": 2.9333388805389404, + "learning_rate": 0.0003662157703728143, + "loss": 0.1857, + "step": 700 + }, + { + "epoch": 2.9, + "grad_norm": 7.691112995147705, + "learning_rate": 0.00036456614978554934, + "loss": 0.2054, + "step": 725 + }, + { + "epoch": 3.0, + "grad_norm": 4.908492565155029, + "learning_rate": 0.00036291652919828444, + "loss": 0.1956, + "step": 750 + }, + { + "epoch": 3.1, + "grad_norm": 1.8923287391662598, + "learning_rate": 0.00036126690861101944, + "loss": 0.1708, + "step": 775 + }, + { + "epoch": 3.2, + "grad_norm": 4.403504848480225, + "learning_rate": 0.00035961728802375454, + "loss": 0.1441, + "step": 800 + }, + { + "epoch": 3.3, + "grad_norm": 4.117386817932129, + "learning_rate": 0.0003579676674364896, + "loss": 0.1586, + "step": 825 + }, + { + "epoch": 3.4, + "grad_norm": 7.119816303253174, + "learning_rate": 0.0003563180468492247, + "loss": 0.1472, + "step": 850 + }, + { + "epoch": 3.5, + "grad_norm": 3.5796430110931396, + "learning_rate": 0.0003546684262619598, + "loss": 0.1802, + "step": 875 + }, + { + "epoch": 3.6, + "grad_norm": 2.97688889503479, + "learning_rate": 0.00035301880567469485, + "loss": 0.1578, + "step": 900 + }, + { + "epoch": 3.7, + "grad_norm": 3.716148614883423, + "learning_rate": 0.0003513691850874299, + "loss": 0.1662, + "step": 925 + }, + { + "epoch": 3.8, + "grad_norm": 6.1249566078186035, + "learning_rate": 0.00034971956450016495, + "loss": 0.159, + "step": 950 + }, + { + "epoch": 3.9, + "grad_norm": 3.0592427253723145, + "learning_rate": 0.00034806994391290005, + "loss": 0.148, + "step": 975 + }, + { + "epoch": 4.0, + "grad_norm": 4.467265605926514, + "learning_rate": 0.0003464203233256351, + "loss": 0.1442, + "step": 1000 + } + ], + "logging_steps": 25, + "max_steps": 6250, + "num_input_tokens_seen": 0, + "num_train_epochs": 25, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 50819481600000.0, + "train_batch_size": 20, + "trial_name": null, + "trial_params": null +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-1000/training_args.bin b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-1000/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..ef3fc22e099df49861bca49d15da2b959cfcb6df --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-1000/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ec1d8ed36c45af65d67a23de6c5216a486932bfb09b720e1a4417b82eaaa215d +size 5304 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-1000/vocab.txt b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-1000/vocab.txt new file mode 100644 index 0000000000000000000000000000000000000000..cc9ad53732e21ea9180f18e1fb67a38cc5702d4c --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-1000/vocab.txt @@ -0,0 +1,4101 @@ +[PAD] +[UNK] +[CLS] +[SEP] +[MASK] +AAAAAA +AAAAAT +AAAAAC +AAAAAG +AAAATA +AAAATT +AAAATC +AAAATG +AAAACA +AAAACT +AAAACC +AAAACG +AAAAGA +AAAAGT +AAAAGC +AAAAGG +AAATAA +AAATAT +AAATAC +AAATAG +AAATTA +AAATTT +AAATTC +AAATTG +AAATCA +AAATCT +AAATCC +AAATCG +AAATGA +AAATGT +AAATGC +AAATGG +AAACAA +AAACAT +AAACAC +AAACAG +AAACTA +AAACTT +AAACTC +AAACTG +AAACCA +AAACCT +AAACCC +AAACCG +AAACGA +AAACGT +AAACGC +AAACGG +AAAGAA +AAAGAT +AAAGAC +AAAGAG +AAAGTA +AAAGTT +AAAGTC +AAAGTG +AAAGCA +AAAGCT +AAAGCC +AAAGCG +AAAGGA +AAAGGT +AAAGGC +AAAGGG +AATAAA +AATAAT +AATAAC +AATAAG +AATATA +AATATT +AATATC +AATATG +AATACA +AATACT +AATACC +AATACG +AATAGA +AATAGT +AATAGC +AATAGG +AATTAA +AATTAT +AATTAC +AATTAG +AATTTA +AATTTT +AATTTC +AATTTG +AATTCA +AATTCT +AATTCC +AATTCG +AATTGA +AATTGT +AATTGC +AATTGG +AATCAA +AATCAT +AATCAC +AATCAG +AATCTA +AATCTT +AATCTC +AATCTG +AATCCA +AATCCT +AATCCC +AATCCG +AATCGA +AATCGT +AATCGC +AATCGG +AATGAA +AATGAT +AATGAC +AATGAG +AATGTA +AATGTT +AATGTC +AATGTG +AATGCA +AATGCT +AATGCC +AATGCG +AATGGA +AATGGT +AATGGC +AATGGG +AACAAA +AACAAT +AACAAC +AACAAG +AACATA +AACATT +AACATC +AACATG +AACACA +AACACT +AACACC +AACACG +AACAGA +AACAGT +AACAGC +AACAGG +AACTAA +AACTAT +AACTAC +AACTAG +AACTTA +AACTTT +AACTTC +AACTTG +AACTCA +AACTCT +AACTCC +AACTCG +AACTGA +AACTGT +AACTGC +AACTGG +AACCAA +AACCAT +AACCAC +AACCAG +AACCTA +AACCTT +AACCTC +AACCTG +AACCCA +AACCCT +AACCCC +AACCCG +AACCGA +AACCGT +AACCGC +AACCGG +AACGAA +AACGAT +AACGAC +AACGAG +AACGTA +AACGTT +AACGTC +AACGTG +AACGCA +AACGCT +AACGCC +AACGCG +AACGGA +AACGGT +AACGGC +AACGGG +AAGAAA +AAGAAT +AAGAAC +AAGAAG +AAGATA +AAGATT +AAGATC +AAGATG +AAGACA +AAGACT +AAGACC +AAGACG +AAGAGA +AAGAGT +AAGAGC +AAGAGG +AAGTAA +AAGTAT +AAGTAC +AAGTAG +AAGTTA +AAGTTT +AAGTTC +AAGTTG +AAGTCA +AAGTCT +AAGTCC +AAGTCG +AAGTGA +AAGTGT +AAGTGC +AAGTGG +AAGCAA +AAGCAT +AAGCAC +AAGCAG +AAGCTA +AAGCTT +AAGCTC +AAGCTG +AAGCCA +AAGCCT +AAGCCC +AAGCCG +AAGCGA +AAGCGT +AAGCGC +AAGCGG +AAGGAA +AAGGAT +AAGGAC +AAGGAG +AAGGTA +AAGGTT +AAGGTC +AAGGTG +AAGGCA +AAGGCT +AAGGCC +AAGGCG +AAGGGA +AAGGGT +AAGGGC +AAGGGG +ATAAAA +ATAAAT +ATAAAC +ATAAAG +ATAATA +ATAATT +ATAATC +ATAATG +ATAACA +ATAACT +ATAACC +ATAACG +ATAAGA +ATAAGT +ATAAGC +ATAAGG +ATATAA +ATATAT +ATATAC +ATATAG +ATATTA +ATATTT +ATATTC +ATATTG +ATATCA +ATATCT +ATATCC +ATATCG +ATATGA +ATATGT +ATATGC +ATATGG +ATACAA +ATACAT +ATACAC +ATACAG +ATACTA +ATACTT +ATACTC +ATACTG +ATACCA +ATACCT +ATACCC +ATACCG +ATACGA +ATACGT +ATACGC +ATACGG +ATAGAA +ATAGAT +ATAGAC +ATAGAG +ATAGTA +ATAGTT +ATAGTC +ATAGTG +ATAGCA +ATAGCT +ATAGCC +ATAGCG +ATAGGA +ATAGGT +ATAGGC +ATAGGG +ATTAAA +ATTAAT +ATTAAC +ATTAAG +ATTATA +ATTATT +ATTATC +ATTATG +ATTACA +ATTACT +ATTACC +ATTACG +ATTAGA +ATTAGT +ATTAGC +ATTAGG +ATTTAA +ATTTAT +ATTTAC +ATTTAG +ATTTTA +ATTTTT +ATTTTC +ATTTTG +ATTTCA +ATTTCT +ATTTCC +ATTTCG +ATTTGA +ATTTGT +ATTTGC +ATTTGG +ATTCAA +ATTCAT +ATTCAC +ATTCAG +ATTCTA +ATTCTT +ATTCTC +ATTCTG +ATTCCA +ATTCCT +ATTCCC +ATTCCG +ATTCGA +ATTCGT +ATTCGC +ATTCGG +ATTGAA +ATTGAT +ATTGAC +ATTGAG +ATTGTA +ATTGTT +ATTGTC +ATTGTG +ATTGCA +ATTGCT +ATTGCC +ATTGCG +ATTGGA +ATTGGT +ATTGGC +ATTGGG +ATCAAA +ATCAAT +ATCAAC +ATCAAG +ATCATA +ATCATT +ATCATC +ATCATG +ATCACA +ATCACT +ATCACC +ATCACG +ATCAGA +ATCAGT +ATCAGC +ATCAGG +ATCTAA +ATCTAT +ATCTAC +ATCTAG +ATCTTA +ATCTTT +ATCTTC +ATCTTG +ATCTCA +ATCTCT +ATCTCC +ATCTCG +ATCTGA +ATCTGT +ATCTGC +ATCTGG +ATCCAA +ATCCAT +ATCCAC +ATCCAG +ATCCTA +ATCCTT +ATCCTC +ATCCTG +ATCCCA +ATCCCT +ATCCCC +ATCCCG +ATCCGA +ATCCGT +ATCCGC +ATCCGG +ATCGAA +ATCGAT +ATCGAC +ATCGAG +ATCGTA +ATCGTT +ATCGTC +ATCGTG +ATCGCA +ATCGCT +ATCGCC +ATCGCG +ATCGGA +ATCGGT +ATCGGC +ATCGGG +ATGAAA +ATGAAT +ATGAAC +ATGAAG +ATGATA +ATGATT +ATGATC +ATGATG +ATGACA +ATGACT +ATGACC +ATGACG +ATGAGA +ATGAGT +ATGAGC +ATGAGG +ATGTAA +ATGTAT +ATGTAC +ATGTAG +ATGTTA +ATGTTT +ATGTTC +ATGTTG +ATGTCA +ATGTCT +ATGTCC +ATGTCG +ATGTGA +ATGTGT +ATGTGC +ATGTGG +ATGCAA +ATGCAT +ATGCAC +ATGCAG +ATGCTA +ATGCTT +ATGCTC +ATGCTG +ATGCCA +ATGCCT +ATGCCC +ATGCCG +ATGCGA +ATGCGT +ATGCGC +ATGCGG +ATGGAA +ATGGAT +ATGGAC +ATGGAG +ATGGTA +ATGGTT +ATGGTC +ATGGTG +ATGGCA +ATGGCT +ATGGCC +ATGGCG +ATGGGA +ATGGGT +ATGGGC +ATGGGG +ACAAAA +ACAAAT +ACAAAC +ACAAAG +ACAATA +ACAATT +ACAATC +ACAATG +ACAACA +ACAACT +ACAACC +ACAACG +ACAAGA +ACAAGT +ACAAGC +ACAAGG +ACATAA +ACATAT +ACATAC +ACATAG +ACATTA +ACATTT +ACATTC +ACATTG +ACATCA +ACATCT +ACATCC +ACATCG +ACATGA +ACATGT +ACATGC +ACATGG +ACACAA +ACACAT +ACACAC +ACACAG +ACACTA +ACACTT +ACACTC +ACACTG +ACACCA +ACACCT +ACACCC +ACACCG +ACACGA +ACACGT +ACACGC +ACACGG +ACAGAA +ACAGAT +ACAGAC +ACAGAG +ACAGTA +ACAGTT +ACAGTC +ACAGTG +ACAGCA +ACAGCT +ACAGCC +ACAGCG +ACAGGA +ACAGGT +ACAGGC +ACAGGG +ACTAAA +ACTAAT +ACTAAC +ACTAAG +ACTATA +ACTATT +ACTATC +ACTATG +ACTACA +ACTACT +ACTACC +ACTACG +ACTAGA +ACTAGT +ACTAGC +ACTAGG +ACTTAA +ACTTAT +ACTTAC +ACTTAG +ACTTTA +ACTTTT +ACTTTC +ACTTTG +ACTTCA +ACTTCT +ACTTCC +ACTTCG +ACTTGA +ACTTGT +ACTTGC +ACTTGG +ACTCAA +ACTCAT +ACTCAC +ACTCAG +ACTCTA +ACTCTT +ACTCTC +ACTCTG +ACTCCA +ACTCCT +ACTCCC +ACTCCG +ACTCGA +ACTCGT +ACTCGC +ACTCGG +ACTGAA +ACTGAT +ACTGAC +ACTGAG +ACTGTA +ACTGTT +ACTGTC +ACTGTG +ACTGCA +ACTGCT +ACTGCC +ACTGCG +ACTGGA +ACTGGT +ACTGGC +ACTGGG +ACCAAA +ACCAAT +ACCAAC +ACCAAG +ACCATA +ACCATT +ACCATC +ACCATG +ACCACA +ACCACT +ACCACC +ACCACG +ACCAGA +ACCAGT +ACCAGC +ACCAGG +ACCTAA +ACCTAT +ACCTAC +ACCTAG +ACCTTA +ACCTTT +ACCTTC +ACCTTG +ACCTCA +ACCTCT +ACCTCC +ACCTCG +ACCTGA +ACCTGT +ACCTGC +ACCTGG +ACCCAA +ACCCAT +ACCCAC +ACCCAG +ACCCTA +ACCCTT +ACCCTC +ACCCTG +ACCCCA +ACCCCT +ACCCCC +ACCCCG +ACCCGA +ACCCGT +ACCCGC +ACCCGG +ACCGAA +ACCGAT +ACCGAC +ACCGAG +ACCGTA +ACCGTT +ACCGTC +ACCGTG +ACCGCA +ACCGCT +ACCGCC +ACCGCG +ACCGGA +ACCGGT +ACCGGC +ACCGGG +ACGAAA +ACGAAT +ACGAAC +ACGAAG +ACGATA +ACGATT +ACGATC +ACGATG +ACGACA +ACGACT +ACGACC +ACGACG +ACGAGA +ACGAGT +ACGAGC +ACGAGG +ACGTAA +ACGTAT +ACGTAC +ACGTAG +ACGTTA +ACGTTT +ACGTTC +ACGTTG +ACGTCA +ACGTCT +ACGTCC +ACGTCG +ACGTGA +ACGTGT +ACGTGC +ACGTGG +ACGCAA +ACGCAT +ACGCAC +ACGCAG +ACGCTA +ACGCTT +ACGCTC +ACGCTG +ACGCCA +ACGCCT +ACGCCC +ACGCCG +ACGCGA +ACGCGT +ACGCGC +ACGCGG +ACGGAA +ACGGAT +ACGGAC +ACGGAG +ACGGTA +ACGGTT +ACGGTC +ACGGTG +ACGGCA +ACGGCT +ACGGCC +ACGGCG +ACGGGA +ACGGGT +ACGGGC +ACGGGG +AGAAAA +AGAAAT +AGAAAC +AGAAAG +AGAATA +AGAATT +AGAATC +AGAATG +AGAACA +AGAACT +AGAACC +AGAACG +AGAAGA +AGAAGT +AGAAGC +AGAAGG +AGATAA +AGATAT +AGATAC +AGATAG +AGATTA +AGATTT +AGATTC +AGATTG +AGATCA +AGATCT +AGATCC +AGATCG +AGATGA +AGATGT +AGATGC +AGATGG +AGACAA +AGACAT +AGACAC +AGACAG +AGACTA +AGACTT +AGACTC +AGACTG +AGACCA +AGACCT +AGACCC +AGACCG +AGACGA +AGACGT +AGACGC +AGACGG +AGAGAA +AGAGAT +AGAGAC +AGAGAG +AGAGTA +AGAGTT +AGAGTC +AGAGTG +AGAGCA +AGAGCT +AGAGCC +AGAGCG +AGAGGA +AGAGGT +AGAGGC +AGAGGG +AGTAAA +AGTAAT +AGTAAC +AGTAAG +AGTATA +AGTATT +AGTATC +AGTATG +AGTACA +AGTACT +AGTACC +AGTACG +AGTAGA +AGTAGT +AGTAGC +AGTAGG +AGTTAA +AGTTAT +AGTTAC +AGTTAG +AGTTTA +AGTTTT +AGTTTC +AGTTTG +AGTTCA +AGTTCT +AGTTCC +AGTTCG +AGTTGA +AGTTGT +AGTTGC +AGTTGG +AGTCAA +AGTCAT +AGTCAC +AGTCAG +AGTCTA +AGTCTT +AGTCTC +AGTCTG +AGTCCA +AGTCCT +AGTCCC +AGTCCG +AGTCGA +AGTCGT +AGTCGC +AGTCGG +AGTGAA +AGTGAT +AGTGAC +AGTGAG +AGTGTA +AGTGTT +AGTGTC +AGTGTG +AGTGCA +AGTGCT +AGTGCC +AGTGCG +AGTGGA +AGTGGT +AGTGGC +AGTGGG +AGCAAA +AGCAAT +AGCAAC +AGCAAG +AGCATA +AGCATT +AGCATC +AGCATG +AGCACA +AGCACT +AGCACC +AGCACG +AGCAGA +AGCAGT +AGCAGC +AGCAGG +AGCTAA +AGCTAT +AGCTAC +AGCTAG +AGCTTA +AGCTTT +AGCTTC +AGCTTG +AGCTCA +AGCTCT +AGCTCC +AGCTCG +AGCTGA +AGCTGT +AGCTGC +AGCTGG +AGCCAA +AGCCAT +AGCCAC +AGCCAG +AGCCTA +AGCCTT +AGCCTC +AGCCTG +AGCCCA +AGCCCT +AGCCCC +AGCCCG +AGCCGA +AGCCGT +AGCCGC +AGCCGG +AGCGAA +AGCGAT +AGCGAC +AGCGAG +AGCGTA +AGCGTT +AGCGTC +AGCGTG +AGCGCA +AGCGCT +AGCGCC +AGCGCG +AGCGGA +AGCGGT +AGCGGC +AGCGGG +AGGAAA +AGGAAT +AGGAAC +AGGAAG +AGGATA +AGGATT +AGGATC +AGGATG +AGGACA +AGGACT +AGGACC +AGGACG +AGGAGA +AGGAGT +AGGAGC +AGGAGG +AGGTAA +AGGTAT +AGGTAC +AGGTAG +AGGTTA +AGGTTT +AGGTTC +AGGTTG +AGGTCA +AGGTCT +AGGTCC +AGGTCG +AGGTGA +AGGTGT +AGGTGC +AGGTGG +AGGCAA +AGGCAT +AGGCAC +AGGCAG +AGGCTA +AGGCTT +AGGCTC +AGGCTG +AGGCCA +AGGCCT +AGGCCC +AGGCCG +AGGCGA +AGGCGT +AGGCGC +AGGCGG +AGGGAA +AGGGAT +AGGGAC +AGGGAG +AGGGTA +AGGGTT +AGGGTC +AGGGTG +AGGGCA +AGGGCT +AGGGCC +AGGGCG +AGGGGA +AGGGGT +AGGGGC +AGGGGG +TAAAAA +TAAAAT +TAAAAC +TAAAAG +TAAATA +TAAATT +TAAATC +TAAATG +TAAACA +TAAACT +TAAACC +TAAACG +TAAAGA +TAAAGT +TAAAGC +TAAAGG +TAATAA +TAATAT +TAATAC +TAATAG +TAATTA +TAATTT +TAATTC +TAATTG +TAATCA +TAATCT +TAATCC +TAATCG +TAATGA +TAATGT +TAATGC +TAATGG +TAACAA +TAACAT +TAACAC +TAACAG +TAACTA +TAACTT +TAACTC +TAACTG +TAACCA +TAACCT +TAACCC +TAACCG +TAACGA +TAACGT +TAACGC +TAACGG +TAAGAA +TAAGAT +TAAGAC +TAAGAG +TAAGTA +TAAGTT +TAAGTC +TAAGTG +TAAGCA +TAAGCT +TAAGCC +TAAGCG +TAAGGA +TAAGGT +TAAGGC +TAAGGG +TATAAA +TATAAT +TATAAC +TATAAG +TATATA +TATATT +TATATC +TATATG +TATACA +TATACT +TATACC +TATACG +TATAGA +TATAGT +TATAGC +TATAGG +TATTAA +TATTAT +TATTAC +TATTAG +TATTTA +TATTTT +TATTTC +TATTTG +TATTCA +TATTCT +TATTCC +TATTCG +TATTGA +TATTGT +TATTGC +TATTGG +TATCAA +TATCAT +TATCAC +TATCAG +TATCTA +TATCTT +TATCTC +TATCTG +TATCCA +TATCCT +TATCCC +TATCCG +TATCGA +TATCGT +TATCGC +TATCGG +TATGAA +TATGAT +TATGAC +TATGAG +TATGTA +TATGTT +TATGTC +TATGTG +TATGCA +TATGCT +TATGCC +TATGCG +TATGGA +TATGGT +TATGGC +TATGGG +TACAAA +TACAAT +TACAAC +TACAAG +TACATA +TACATT +TACATC +TACATG +TACACA +TACACT +TACACC +TACACG +TACAGA +TACAGT +TACAGC +TACAGG +TACTAA +TACTAT +TACTAC +TACTAG +TACTTA +TACTTT +TACTTC +TACTTG +TACTCA +TACTCT +TACTCC +TACTCG +TACTGA +TACTGT +TACTGC +TACTGG +TACCAA +TACCAT +TACCAC +TACCAG +TACCTA +TACCTT +TACCTC +TACCTG +TACCCA +TACCCT +TACCCC +TACCCG +TACCGA +TACCGT +TACCGC +TACCGG +TACGAA +TACGAT +TACGAC +TACGAG +TACGTA +TACGTT +TACGTC +TACGTG +TACGCA +TACGCT +TACGCC +TACGCG +TACGGA +TACGGT +TACGGC +TACGGG +TAGAAA +TAGAAT +TAGAAC +TAGAAG +TAGATA +TAGATT +TAGATC +TAGATG +TAGACA +TAGACT +TAGACC +TAGACG +TAGAGA +TAGAGT +TAGAGC +TAGAGG +TAGTAA +TAGTAT +TAGTAC +TAGTAG +TAGTTA +TAGTTT +TAGTTC +TAGTTG +TAGTCA +TAGTCT +TAGTCC +TAGTCG +TAGTGA +TAGTGT +TAGTGC +TAGTGG +TAGCAA +TAGCAT +TAGCAC +TAGCAG +TAGCTA +TAGCTT +TAGCTC +TAGCTG +TAGCCA +TAGCCT +TAGCCC +TAGCCG +TAGCGA +TAGCGT +TAGCGC +TAGCGG +TAGGAA +TAGGAT +TAGGAC +TAGGAG +TAGGTA +TAGGTT +TAGGTC +TAGGTG +TAGGCA +TAGGCT +TAGGCC +TAGGCG +TAGGGA +TAGGGT +TAGGGC +TAGGGG +TTAAAA +TTAAAT +TTAAAC +TTAAAG +TTAATA +TTAATT +TTAATC +TTAATG +TTAACA +TTAACT +TTAACC +TTAACG +TTAAGA +TTAAGT +TTAAGC +TTAAGG +TTATAA +TTATAT +TTATAC +TTATAG +TTATTA +TTATTT +TTATTC +TTATTG +TTATCA +TTATCT +TTATCC +TTATCG +TTATGA +TTATGT +TTATGC +TTATGG +TTACAA +TTACAT +TTACAC +TTACAG +TTACTA +TTACTT +TTACTC +TTACTG +TTACCA +TTACCT +TTACCC +TTACCG +TTACGA +TTACGT +TTACGC +TTACGG +TTAGAA +TTAGAT +TTAGAC +TTAGAG +TTAGTA +TTAGTT +TTAGTC +TTAGTG +TTAGCA +TTAGCT +TTAGCC +TTAGCG +TTAGGA +TTAGGT +TTAGGC +TTAGGG +TTTAAA +TTTAAT +TTTAAC +TTTAAG +TTTATA +TTTATT +TTTATC +TTTATG +TTTACA +TTTACT +TTTACC +TTTACG +TTTAGA +TTTAGT +TTTAGC +TTTAGG +TTTTAA +TTTTAT +TTTTAC +TTTTAG +TTTTTA +TTTTTT +TTTTTC +TTTTTG +TTTTCA +TTTTCT +TTTTCC +TTTTCG +TTTTGA +TTTTGT +TTTTGC +TTTTGG +TTTCAA +TTTCAT +TTTCAC +TTTCAG +TTTCTA +TTTCTT +TTTCTC +TTTCTG +TTTCCA +TTTCCT +TTTCCC +TTTCCG +TTTCGA +TTTCGT +TTTCGC +TTTCGG +TTTGAA +TTTGAT +TTTGAC +TTTGAG +TTTGTA +TTTGTT +TTTGTC +TTTGTG +TTTGCA +TTTGCT +TTTGCC +TTTGCG +TTTGGA +TTTGGT +TTTGGC +TTTGGG +TTCAAA +TTCAAT +TTCAAC +TTCAAG +TTCATA +TTCATT +TTCATC +TTCATG +TTCACA +TTCACT +TTCACC +TTCACG +TTCAGA +TTCAGT +TTCAGC +TTCAGG +TTCTAA +TTCTAT +TTCTAC +TTCTAG +TTCTTA +TTCTTT +TTCTTC +TTCTTG +TTCTCA +TTCTCT +TTCTCC +TTCTCG +TTCTGA +TTCTGT +TTCTGC +TTCTGG +TTCCAA +TTCCAT +TTCCAC +TTCCAG +TTCCTA +TTCCTT +TTCCTC +TTCCTG +TTCCCA +TTCCCT +TTCCCC +TTCCCG +TTCCGA +TTCCGT +TTCCGC +TTCCGG +TTCGAA +TTCGAT +TTCGAC +TTCGAG +TTCGTA +TTCGTT +TTCGTC +TTCGTG +TTCGCA +TTCGCT +TTCGCC +TTCGCG +TTCGGA +TTCGGT +TTCGGC +TTCGGG +TTGAAA +TTGAAT +TTGAAC +TTGAAG +TTGATA +TTGATT +TTGATC +TTGATG +TTGACA +TTGACT +TTGACC +TTGACG +TTGAGA +TTGAGT +TTGAGC +TTGAGG +TTGTAA +TTGTAT +TTGTAC +TTGTAG +TTGTTA +TTGTTT +TTGTTC +TTGTTG +TTGTCA +TTGTCT +TTGTCC +TTGTCG +TTGTGA +TTGTGT +TTGTGC +TTGTGG +TTGCAA +TTGCAT +TTGCAC +TTGCAG +TTGCTA +TTGCTT +TTGCTC +TTGCTG +TTGCCA +TTGCCT +TTGCCC +TTGCCG +TTGCGA +TTGCGT +TTGCGC +TTGCGG +TTGGAA +TTGGAT +TTGGAC +TTGGAG +TTGGTA +TTGGTT +TTGGTC +TTGGTG +TTGGCA +TTGGCT +TTGGCC +TTGGCG +TTGGGA +TTGGGT +TTGGGC +TTGGGG +TCAAAA +TCAAAT +TCAAAC +TCAAAG +TCAATA +TCAATT +TCAATC +TCAATG +TCAACA +TCAACT +TCAACC +TCAACG +TCAAGA +TCAAGT +TCAAGC +TCAAGG +TCATAA +TCATAT +TCATAC +TCATAG +TCATTA +TCATTT +TCATTC +TCATTG +TCATCA +TCATCT +TCATCC +TCATCG +TCATGA +TCATGT +TCATGC +TCATGG +TCACAA +TCACAT +TCACAC +TCACAG +TCACTA +TCACTT +TCACTC +TCACTG +TCACCA +TCACCT +TCACCC +TCACCG +TCACGA +TCACGT +TCACGC +TCACGG +TCAGAA +TCAGAT +TCAGAC +TCAGAG +TCAGTA +TCAGTT +TCAGTC +TCAGTG +TCAGCA +TCAGCT +TCAGCC +TCAGCG +TCAGGA +TCAGGT +TCAGGC +TCAGGG +TCTAAA +TCTAAT +TCTAAC +TCTAAG +TCTATA +TCTATT +TCTATC +TCTATG +TCTACA +TCTACT +TCTACC +TCTACG +TCTAGA +TCTAGT +TCTAGC +TCTAGG +TCTTAA +TCTTAT +TCTTAC +TCTTAG +TCTTTA +TCTTTT +TCTTTC +TCTTTG +TCTTCA +TCTTCT +TCTTCC +TCTTCG +TCTTGA +TCTTGT +TCTTGC +TCTTGG +TCTCAA +TCTCAT +TCTCAC +TCTCAG +TCTCTA +TCTCTT +TCTCTC +TCTCTG +TCTCCA +TCTCCT +TCTCCC +TCTCCG +TCTCGA +TCTCGT +TCTCGC +TCTCGG +TCTGAA +TCTGAT +TCTGAC +TCTGAG +TCTGTA +TCTGTT +TCTGTC +TCTGTG +TCTGCA +TCTGCT +TCTGCC +TCTGCG +TCTGGA +TCTGGT +TCTGGC +TCTGGG +TCCAAA +TCCAAT +TCCAAC +TCCAAG +TCCATA +TCCATT +TCCATC +TCCATG +TCCACA +TCCACT +TCCACC +TCCACG +TCCAGA +TCCAGT +TCCAGC +TCCAGG +TCCTAA +TCCTAT +TCCTAC +TCCTAG +TCCTTA +TCCTTT +TCCTTC +TCCTTG +TCCTCA +TCCTCT +TCCTCC +TCCTCG +TCCTGA +TCCTGT +TCCTGC +TCCTGG +TCCCAA +TCCCAT +TCCCAC +TCCCAG +TCCCTA +TCCCTT +TCCCTC +TCCCTG +TCCCCA +TCCCCT +TCCCCC +TCCCCG +TCCCGA +TCCCGT +TCCCGC +TCCCGG +TCCGAA +TCCGAT +TCCGAC +TCCGAG +TCCGTA +TCCGTT +TCCGTC +TCCGTG +TCCGCA +TCCGCT +TCCGCC +TCCGCG +TCCGGA +TCCGGT +TCCGGC +TCCGGG +TCGAAA +TCGAAT +TCGAAC +TCGAAG +TCGATA +TCGATT +TCGATC +TCGATG +TCGACA +TCGACT +TCGACC +TCGACG +TCGAGA +TCGAGT +TCGAGC +TCGAGG +TCGTAA +TCGTAT +TCGTAC +TCGTAG +TCGTTA +TCGTTT +TCGTTC +TCGTTG +TCGTCA +TCGTCT +TCGTCC +TCGTCG +TCGTGA +TCGTGT +TCGTGC +TCGTGG +TCGCAA +TCGCAT +TCGCAC +TCGCAG +TCGCTA +TCGCTT +TCGCTC +TCGCTG +TCGCCA +TCGCCT +TCGCCC +TCGCCG +TCGCGA +TCGCGT +TCGCGC +TCGCGG +TCGGAA +TCGGAT +TCGGAC +TCGGAG +TCGGTA +TCGGTT +TCGGTC +TCGGTG +TCGGCA +TCGGCT +TCGGCC +TCGGCG +TCGGGA +TCGGGT +TCGGGC +TCGGGG +TGAAAA +TGAAAT +TGAAAC +TGAAAG +TGAATA +TGAATT +TGAATC +TGAATG +TGAACA +TGAACT +TGAACC +TGAACG +TGAAGA +TGAAGT +TGAAGC +TGAAGG +TGATAA +TGATAT +TGATAC +TGATAG +TGATTA +TGATTT +TGATTC +TGATTG +TGATCA +TGATCT +TGATCC +TGATCG +TGATGA +TGATGT +TGATGC +TGATGG +TGACAA +TGACAT +TGACAC +TGACAG +TGACTA +TGACTT +TGACTC +TGACTG +TGACCA +TGACCT +TGACCC +TGACCG +TGACGA +TGACGT +TGACGC +TGACGG +TGAGAA +TGAGAT +TGAGAC +TGAGAG +TGAGTA +TGAGTT +TGAGTC +TGAGTG +TGAGCA +TGAGCT +TGAGCC +TGAGCG +TGAGGA +TGAGGT +TGAGGC +TGAGGG +TGTAAA +TGTAAT +TGTAAC +TGTAAG +TGTATA +TGTATT +TGTATC +TGTATG +TGTACA +TGTACT +TGTACC +TGTACG +TGTAGA +TGTAGT +TGTAGC +TGTAGG +TGTTAA +TGTTAT +TGTTAC +TGTTAG +TGTTTA +TGTTTT +TGTTTC +TGTTTG +TGTTCA +TGTTCT +TGTTCC +TGTTCG +TGTTGA +TGTTGT +TGTTGC +TGTTGG +TGTCAA +TGTCAT +TGTCAC +TGTCAG +TGTCTA +TGTCTT +TGTCTC +TGTCTG +TGTCCA +TGTCCT +TGTCCC +TGTCCG +TGTCGA +TGTCGT +TGTCGC +TGTCGG +TGTGAA +TGTGAT +TGTGAC +TGTGAG +TGTGTA +TGTGTT +TGTGTC +TGTGTG +TGTGCA +TGTGCT +TGTGCC +TGTGCG +TGTGGA +TGTGGT +TGTGGC +TGTGGG +TGCAAA +TGCAAT +TGCAAC +TGCAAG +TGCATA +TGCATT +TGCATC +TGCATG +TGCACA +TGCACT +TGCACC +TGCACG +TGCAGA +TGCAGT +TGCAGC +TGCAGG +TGCTAA +TGCTAT +TGCTAC +TGCTAG +TGCTTA +TGCTTT +TGCTTC +TGCTTG +TGCTCA +TGCTCT +TGCTCC +TGCTCG +TGCTGA +TGCTGT +TGCTGC +TGCTGG +TGCCAA +TGCCAT +TGCCAC +TGCCAG +TGCCTA +TGCCTT +TGCCTC +TGCCTG +TGCCCA +TGCCCT +TGCCCC +TGCCCG +TGCCGA +TGCCGT +TGCCGC +TGCCGG +TGCGAA +TGCGAT +TGCGAC +TGCGAG +TGCGTA +TGCGTT +TGCGTC +TGCGTG +TGCGCA +TGCGCT +TGCGCC +TGCGCG +TGCGGA +TGCGGT +TGCGGC +TGCGGG +TGGAAA +TGGAAT +TGGAAC +TGGAAG +TGGATA +TGGATT +TGGATC +TGGATG +TGGACA +TGGACT +TGGACC +TGGACG +TGGAGA +TGGAGT +TGGAGC +TGGAGG +TGGTAA +TGGTAT +TGGTAC +TGGTAG +TGGTTA +TGGTTT +TGGTTC +TGGTTG +TGGTCA +TGGTCT +TGGTCC +TGGTCG +TGGTGA +TGGTGT +TGGTGC +TGGTGG +TGGCAA +TGGCAT +TGGCAC +TGGCAG +TGGCTA +TGGCTT +TGGCTC +TGGCTG +TGGCCA +TGGCCT +TGGCCC +TGGCCG +TGGCGA +TGGCGT +TGGCGC +TGGCGG +TGGGAA +TGGGAT +TGGGAC +TGGGAG +TGGGTA +TGGGTT +TGGGTC +TGGGTG +TGGGCA +TGGGCT +TGGGCC +TGGGCG +TGGGGA +TGGGGT +TGGGGC +TGGGGG +CAAAAA +CAAAAT +CAAAAC +CAAAAG +CAAATA +CAAATT +CAAATC +CAAATG +CAAACA +CAAACT +CAAACC +CAAACG +CAAAGA +CAAAGT +CAAAGC +CAAAGG +CAATAA +CAATAT +CAATAC +CAATAG +CAATTA +CAATTT +CAATTC +CAATTG +CAATCA +CAATCT +CAATCC +CAATCG +CAATGA +CAATGT +CAATGC +CAATGG +CAACAA +CAACAT +CAACAC +CAACAG +CAACTA +CAACTT +CAACTC +CAACTG +CAACCA +CAACCT +CAACCC +CAACCG +CAACGA +CAACGT +CAACGC +CAACGG +CAAGAA +CAAGAT +CAAGAC +CAAGAG +CAAGTA +CAAGTT +CAAGTC +CAAGTG +CAAGCA +CAAGCT +CAAGCC +CAAGCG +CAAGGA +CAAGGT +CAAGGC +CAAGGG +CATAAA +CATAAT +CATAAC +CATAAG +CATATA +CATATT +CATATC +CATATG +CATACA +CATACT +CATACC +CATACG +CATAGA +CATAGT +CATAGC +CATAGG +CATTAA +CATTAT +CATTAC +CATTAG +CATTTA +CATTTT +CATTTC +CATTTG +CATTCA +CATTCT +CATTCC +CATTCG +CATTGA +CATTGT +CATTGC +CATTGG +CATCAA +CATCAT +CATCAC +CATCAG +CATCTA +CATCTT +CATCTC +CATCTG +CATCCA +CATCCT +CATCCC +CATCCG +CATCGA +CATCGT +CATCGC +CATCGG +CATGAA +CATGAT +CATGAC +CATGAG +CATGTA +CATGTT +CATGTC +CATGTG +CATGCA +CATGCT +CATGCC +CATGCG +CATGGA +CATGGT +CATGGC +CATGGG +CACAAA +CACAAT +CACAAC +CACAAG +CACATA +CACATT +CACATC +CACATG +CACACA +CACACT +CACACC +CACACG +CACAGA +CACAGT +CACAGC +CACAGG +CACTAA +CACTAT +CACTAC +CACTAG +CACTTA +CACTTT +CACTTC +CACTTG +CACTCA +CACTCT +CACTCC +CACTCG +CACTGA +CACTGT +CACTGC +CACTGG +CACCAA +CACCAT +CACCAC +CACCAG +CACCTA +CACCTT +CACCTC +CACCTG +CACCCA +CACCCT +CACCCC +CACCCG +CACCGA +CACCGT +CACCGC +CACCGG +CACGAA +CACGAT +CACGAC +CACGAG +CACGTA +CACGTT +CACGTC +CACGTG +CACGCA +CACGCT +CACGCC +CACGCG +CACGGA +CACGGT +CACGGC +CACGGG +CAGAAA +CAGAAT +CAGAAC +CAGAAG +CAGATA +CAGATT +CAGATC +CAGATG +CAGACA +CAGACT +CAGACC +CAGACG +CAGAGA +CAGAGT +CAGAGC +CAGAGG +CAGTAA +CAGTAT +CAGTAC +CAGTAG +CAGTTA +CAGTTT +CAGTTC +CAGTTG +CAGTCA +CAGTCT +CAGTCC +CAGTCG +CAGTGA +CAGTGT +CAGTGC +CAGTGG +CAGCAA +CAGCAT +CAGCAC +CAGCAG +CAGCTA +CAGCTT +CAGCTC +CAGCTG +CAGCCA +CAGCCT +CAGCCC +CAGCCG +CAGCGA +CAGCGT +CAGCGC +CAGCGG +CAGGAA +CAGGAT +CAGGAC +CAGGAG +CAGGTA +CAGGTT +CAGGTC +CAGGTG +CAGGCA +CAGGCT +CAGGCC +CAGGCG +CAGGGA +CAGGGT +CAGGGC +CAGGGG +CTAAAA +CTAAAT +CTAAAC +CTAAAG +CTAATA +CTAATT +CTAATC +CTAATG +CTAACA +CTAACT +CTAACC +CTAACG +CTAAGA +CTAAGT +CTAAGC +CTAAGG +CTATAA +CTATAT +CTATAC +CTATAG +CTATTA +CTATTT +CTATTC +CTATTG +CTATCA +CTATCT +CTATCC +CTATCG +CTATGA +CTATGT +CTATGC +CTATGG +CTACAA +CTACAT +CTACAC +CTACAG +CTACTA +CTACTT +CTACTC +CTACTG +CTACCA +CTACCT +CTACCC +CTACCG +CTACGA +CTACGT +CTACGC +CTACGG +CTAGAA +CTAGAT +CTAGAC +CTAGAG +CTAGTA +CTAGTT +CTAGTC +CTAGTG +CTAGCA +CTAGCT +CTAGCC +CTAGCG +CTAGGA +CTAGGT +CTAGGC +CTAGGG +CTTAAA +CTTAAT +CTTAAC +CTTAAG +CTTATA +CTTATT +CTTATC +CTTATG +CTTACA +CTTACT +CTTACC +CTTACG +CTTAGA +CTTAGT +CTTAGC +CTTAGG +CTTTAA +CTTTAT +CTTTAC +CTTTAG +CTTTTA +CTTTTT +CTTTTC +CTTTTG +CTTTCA +CTTTCT +CTTTCC +CTTTCG +CTTTGA +CTTTGT +CTTTGC +CTTTGG +CTTCAA +CTTCAT +CTTCAC +CTTCAG +CTTCTA +CTTCTT +CTTCTC +CTTCTG +CTTCCA +CTTCCT +CTTCCC +CTTCCG +CTTCGA +CTTCGT +CTTCGC +CTTCGG +CTTGAA +CTTGAT +CTTGAC +CTTGAG +CTTGTA +CTTGTT +CTTGTC +CTTGTG +CTTGCA +CTTGCT +CTTGCC +CTTGCG +CTTGGA +CTTGGT +CTTGGC +CTTGGG +CTCAAA +CTCAAT +CTCAAC +CTCAAG +CTCATA +CTCATT +CTCATC +CTCATG +CTCACA +CTCACT +CTCACC +CTCACG +CTCAGA +CTCAGT +CTCAGC +CTCAGG +CTCTAA +CTCTAT +CTCTAC +CTCTAG +CTCTTA +CTCTTT +CTCTTC +CTCTTG +CTCTCA +CTCTCT +CTCTCC +CTCTCG +CTCTGA +CTCTGT +CTCTGC +CTCTGG +CTCCAA +CTCCAT +CTCCAC +CTCCAG +CTCCTA +CTCCTT +CTCCTC +CTCCTG +CTCCCA +CTCCCT +CTCCCC +CTCCCG +CTCCGA +CTCCGT +CTCCGC +CTCCGG +CTCGAA +CTCGAT +CTCGAC +CTCGAG +CTCGTA +CTCGTT +CTCGTC +CTCGTG +CTCGCA +CTCGCT +CTCGCC +CTCGCG +CTCGGA +CTCGGT +CTCGGC +CTCGGG +CTGAAA +CTGAAT +CTGAAC +CTGAAG +CTGATA +CTGATT +CTGATC +CTGATG +CTGACA +CTGACT +CTGACC +CTGACG +CTGAGA +CTGAGT +CTGAGC +CTGAGG +CTGTAA +CTGTAT +CTGTAC +CTGTAG +CTGTTA +CTGTTT +CTGTTC +CTGTTG +CTGTCA +CTGTCT +CTGTCC +CTGTCG +CTGTGA +CTGTGT +CTGTGC +CTGTGG +CTGCAA +CTGCAT +CTGCAC +CTGCAG +CTGCTA +CTGCTT +CTGCTC +CTGCTG +CTGCCA +CTGCCT +CTGCCC +CTGCCG +CTGCGA +CTGCGT +CTGCGC +CTGCGG +CTGGAA +CTGGAT +CTGGAC +CTGGAG +CTGGTA +CTGGTT +CTGGTC +CTGGTG +CTGGCA +CTGGCT +CTGGCC +CTGGCG +CTGGGA +CTGGGT +CTGGGC +CTGGGG +CCAAAA +CCAAAT +CCAAAC +CCAAAG +CCAATA +CCAATT +CCAATC +CCAATG +CCAACA +CCAACT +CCAACC +CCAACG +CCAAGA +CCAAGT +CCAAGC +CCAAGG +CCATAA +CCATAT +CCATAC +CCATAG +CCATTA +CCATTT +CCATTC +CCATTG +CCATCA +CCATCT +CCATCC +CCATCG +CCATGA +CCATGT +CCATGC +CCATGG +CCACAA +CCACAT +CCACAC +CCACAG +CCACTA +CCACTT +CCACTC +CCACTG +CCACCA +CCACCT +CCACCC +CCACCG +CCACGA +CCACGT +CCACGC +CCACGG +CCAGAA +CCAGAT +CCAGAC +CCAGAG +CCAGTA +CCAGTT +CCAGTC +CCAGTG +CCAGCA +CCAGCT +CCAGCC +CCAGCG +CCAGGA +CCAGGT +CCAGGC +CCAGGG +CCTAAA +CCTAAT +CCTAAC +CCTAAG +CCTATA +CCTATT +CCTATC +CCTATG +CCTACA +CCTACT +CCTACC +CCTACG +CCTAGA +CCTAGT +CCTAGC +CCTAGG +CCTTAA +CCTTAT +CCTTAC +CCTTAG +CCTTTA +CCTTTT +CCTTTC +CCTTTG +CCTTCA +CCTTCT +CCTTCC +CCTTCG +CCTTGA +CCTTGT +CCTTGC +CCTTGG +CCTCAA +CCTCAT +CCTCAC +CCTCAG +CCTCTA +CCTCTT +CCTCTC +CCTCTG +CCTCCA +CCTCCT +CCTCCC +CCTCCG +CCTCGA +CCTCGT +CCTCGC +CCTCGG +CCTGAA +CCTGAT +CCTGAC +CCTGAG +CCTGTA +CCTGTT +CCTGTC +CCTGTG +CCTGCA +CCTGCT +CCTGCC +CCTGCG +CCTGGA +CCTGGT +CCTGGC +CCTGGG +CCCAAA +CCCAAT +CCCAAC +CCCAAG +CCCATA +CCCATT +CCCATC +CCCATG +CCCACA +CCCACT +CCCACC +CCCACG +CCCAGA +CCCAGT +CCCAGC +CCCAGG +CCCTAA +CCCTAT +CCCTAC +CCCTAG +CCCTTA +CCCTTT +CCCTTC +CCCTTG +CCCTCA +CCCTCT +CCCTCC +CCCTCG +CCCTGA +CCCTGT +CCCTGC +CCCTGG +CCCCAA +CCCCAT +CCCCAC +CCCCAG +CCCCTA +CCCCTT +CCCCTC +CCCCTG +CCCCCA +CCCCCT +CCCCCC +CCCCCG +CCCCGA +CCCCGT +CCCCGC +CCCCGG +CCCGAA +CCCGAT +CCCGAC +CCCGAG +CCCGTA +CCCGTT +CCCGTC +CCCGTG +CCCGCA +CCCGCT +CCCGCC +CCCGCG +CCCGGA +CCCGGT +CCCGGC +CCCGGG +CCGAAA +CCGAAT +CCGAAC +CCGAAG +CCGATA +CCGATT +CCGATC +CCGATG +CCGACA +CCGACT +CCGACC +CCGACG +CCGAGA +CCGAGT +CCGAGC +CCGAGG +CCGTAA +CCGTAT +CCGTAC +CCGTAG +CCGTTA +CCGTTT +CCGTTC +CCGTTG +CCGTCA +CCGTCT +CCGTCC +CCGTCG +CCGTGA +CCGTGT +CCGTGC +CCGTGG +CCGCAA +CCGCAT +CCGCAC +CCGCAG +CCGCTA +CCGCTT +CCGCTC +CCGCTG +CCGCCA +CCGCCT +CCGCCC +CCGCCG +CCGCGA +CCGCGT +CCGCGC +CCGCGG +CCGGAA +CCGGAT +CCGGAC +CCGGAG +CCGGTA +CCGGTT +CCGGTC +CCGGTG +CCGGCA +CCGGCT +CCGGCC +CCGGCG +CCGGGA +CCGGGT +CCGGGC +CCGGGG +CGAAAA +CGAAAT +CGAAAC +CGAAAG +CGAATA +CGAATT +CGAATC +CGAATG +CGAACA +CGAACT +CGAACC +CGAACG +CGAAGA +CGAAGT +CGAAGC +CGAAGG +CGATAA +CGATAT +CGATAC +CGATAG +CGATTA +CGATTT +CGATTC +CGATTG +CGATCA +CGATCT +CGATCC +CGATCG +CGATGA +CGATGT +CGATGC +CGATGG +CGACAA +CGACAT +CGACAC +CGACAG +CGACTA +CGACTT +CGACTC +CGACTG +CGACCA +CGACCT +CGACCC +CGACCG +CGACGA +CGACGT +CGACGC +CGACGG +CGAGAA +CGAGAT +CGAGAC +CGAGAG +CGAGTA +CGAGTT +CGAGTC +CGAGTG +CGAGCA +CGAGCT +CGAGCC +CGAGCG +CGAGGA +CGAGGT +CGAGGC +CGAGGG +CGTAAA +CGTAAT +CGTAAC +CGTAAG +CGTATA +CGTATT +CGTATC +CGTATG +CGTACA +CGTACT +CGTACC +CGTACG +CGTAGA +CGTAGT +CGTAGC +CGTAGG +CGTTAA +CGTTAT +CGTTAC +CGTTAG +CGTTTA +CGTTTT +CGTTTC +CGTTTG +CGTTCA +CGTTCT +CGTTCC +CGTTCG +CGTTGA +CGTTGT +CGTTGC +CGTTGG +CGTCAA +CGTCAT +CGTCAC +CGTCAG +CGTCTA +CGTCTT +CGTCTC +CGTCTG +CGTCCA +CGTCCT +CGTCCC +CGTCCG +CGTCGA +CGTCGT +CGTCGC +CGTCGG +CGTGAA +CGTGAT +CGTGAC +CGTGAG +CGTGTA +CGTGTT +CGTGTC +CGTGTG +CGTGCA +CGTGCT +CGTGCC +CGTGCG +CGTGGA +CGTGGT +CGTGGC +CGTGGG +CGCAAA +CGCAAT +CGCAAC +CGCAAG +CGCATA +CGCATT +CGCATC +CGCATG +CGCACA +CGCACT +CGCACC +CGCACG +CGCAGA +CGCAGT +CGCAGC +CGCAGG +CGCTAA +CGCTAT +CGCTAC +CGCTAG +CGCTTA +CGCTTT +CGCTTC +CGCTTG +CGCTCA +CGCTCT +CGCTCC +CGCTCG +CGCTGA +CGCTGT +CGCTGC +CGCTGG +CGCCAA +CGCCAT +CGCCAC +CGCCAG +CGCCTA +CGCCTT +CGCCTC +CGCCTG +CGCCCA +CGCCCT +CGCCCC +CGCCCG +CGCCGA +CGCCGT +CGCCGC +CGCCGG +CGCGAA +CGCGAT +CGCGAC +CGCGAG +CGCGTA +CGCGTT +CGCGTC +CGCGTG +CGCGCA +CGCGCT +CGCGCC +CGCGCG +CGCGGA +CGCGGT +CGCGGC +CGCGGG +CGGAAA +CGGAAT +CGGAAC +CGGAAG +CGGATA +CGGATT +CGGATC +CGGATG +CGGACA +CGGACT +CGGACC +CGGACG +CGGAGA +CGGAGT +CGGAGC +CGGAGG +CGGTAA +CGGTAT +CGGTAC +CGGTAG +CGGTTA +CGGTTT +CGGTTC +CGGTTG +CGGTCA +CGGTCT +CGGTCC +CGGTCG +CGGTGA +CGGTGT +CGGTGC +CGGTGG +CGGCAA +CGGCAT +CGGCAC +CGGCAG +CGGCTA +CGGCTT +CGGCTC +CGGCTG +CGGCCA +CGGCCT +CGGCCC +CGGCCG +CGGCGA +CGGCGT +CGGCGC +CGGCGG +CGGGAA +CGGGAT +CGGGAC +CGGGAG +CGGGTA +CGGGTT +CGGGTC +CGGGTG +CGGGCA +CGGGCT +CGGGCC +CGGGCG +CGGGGA +CGGGGT +CGGGGC +CGGGGG +GAAAAA +GAAAAT +GAAAAC +GAAAAG +GAAATA +GAAATT +GAAATC +GAAATG +GAAACA +GAAACT +GAAACC +GAAACG +GAAAGA +GAAAGT +GAAAGC +GAAAGG +GAATAA +GAATAT +GAATAC +GAATAG +GAATTA +GAATTT +GAATTC +GAATTG +GAATCA +GAATCT +GAATCC +GAATCG +GAATGA +GAATGT +GAATGC +GAATGG +GAACAA +GAACAT +GAACAC +GAACAG +GAACTA +GAACTT +GAACTC +GAACTG +GAACCA +GAACCT +GAACCC +GAACCG +GAACGA +GAACGT +GAACGC +GAACGG +GAAGAA +GAAGAT +GAAGAC +GAAGAG +GAAGTA +GAAGTT +GAAGTC +GAAGTG +GAAGCA +GAAGCT +GAAGCC +GAAGCG +GAAGGA +GAAGGT +GAAGGC +GAAGGG +GATAAA +GATAAT +GATAAC +GATAAG +GATATA +GATATT +GATATC +GATATG +GATACA +GATACT +GATACC +GATACG +GATAGA +GATAGT +GATAGC +GATAGG +GATTAA +GATTAT +GATTAC +GATTAG +GATTTA +GATTTT +GATTTC +GATTTG +GATTCA +GATTCT +GATTCC +GATTCG +GATTGA +GATTGT +GATTGC +GATTGG +GATCAA +GATCAT +GATCAC +GATCAG +GATCTA +GATCTT +GATCTC +GATCTG +GATCCA +GATCCT +GATCCC +GATCCG +GATCGA +GATCGT +GATCGC +GATCGG +GATGAA +GATGAT +GATGAC +GATGAG +GATGTA +GATGTT +GATGTC +GATGTG +GATGCA +GATGCT +GATGCC +GATGCG +GATGGA +GATGGT +GATGGC +GATGGG +GACAAA +GACAAT +GACAAC +GACAAG +GACATA +GACATT +GACATC +GACATG +GACACA +GACACT +GACACC +GACACG +GACAGA +GACAGT +GACAGC +GACAGG +GACTAA +GACTAT +GACTAC +GACTAG +GACTTA +GACTTT +GACTTC +GACTTG +GACTCA +GACTCT +GACTCC +GACTCG +GACTGA +GACTGT +GACTGC +GACTGG +GACCAA +GACCAT +GACCAC +GACCAG +GACCTA +GACCTT +GACCTC +GACCTG +GACCCA +GACCCT +GACCCC +GACCCG +GACCGA +GACCGT +GACCGC +GACCGG +GACGAA +GACGAT +GACGAC +GACGAG +GACGTA +GACGTT +GACGTC +GACGTG +GACGCA +GACGCT +GACGCC +GACGCG +GACGGA +GACGGT +GACGGC +GACGGG +GAGAAA +GAGAAT +GAGAAC +GAGAAG +GAGATA +GAGATT +GAGATC +GAGATG +GAGACA +GAGACT +GAGACC +GAGACG +GAGAGA +GAGAGT +GAGAGC +GAGAGG +GAGTAA +GAGTAT +GAGTAC +GAGTAG +GAGTTA +GAGTTT +GAGTTC +GAGTTG +GAGTCA +GAGTCT +GAGTCC +GAGTCG +GAGTGA +GAGTGT +GAGTGC +GAGTGG +GAGCAA +GAGCAT +GAGCAC +GAGCAG +GAGCTA +GAGCTT +GAGCTC +GAGCTG +GAGCCA +GAGCCT +GAGCCC +GAGCCG +GAGCGA +GAGCGT +GAGCGC +GAGCGG +GAGGAA +GAGGAT +GAGGAC +GAGGAG +GAGGTA +GAGGTT +GAGGTC +GAGGTG +GAGGCA +GAGGCT +GAGGCC +GAGGCG +GAGGGA +GAGGGT +GAGGGC +GAGGGG +GTAAAA +GTAAAT +GTAAAC +GTAAAG +GTAATA +GTAATT +GTAATC +GTAATG +GTAACA +GTAACT +GTAACC +GTAACG +GTAAGA +GTAAGT +GTAAGC +GTAAGG +GTATAA +GTATAT +GTATAC +GTATAG +GTATTA +GTATTT +GTATTC +GTATTG +GTATCA +GTATCT +GTATCC +GTATCG +GTATGA +GTATGT +GTATGC +GTATGG +GTACAA +GTACAT +GTACAC +GTACAG +GTACTA +GTACTT +GTACTC +GTACTG +GTACCA +GTACCT +GTACCC +GTACCG +GTACGA +GTACGT +GTACGC +GTACGG +GTAGAA +GTAGAT +GTAGAC +GTAGAG +GTAGTA +GTAGTT +GTAGTC +GTAGTG +GTAGCA +GTAGCT +GTAGCC +GTAGCG +GTAGGA +GTAGGT +GTAGGC +GTAGGG +GTTAAA +GTTAAT +GTTAAC +GTTAAG +GTTATA +GTTATT +GTTATC +GTTATG +GTTACA +GTTACT +GTTACC +GTTACG +GTTAGA +GTTAGT +GTTAGC +GTTAGG +GTTTAA +GTTTAT +GTTTAC +GTTTAG +GTTTTA +GTTTTT +GTTTTC +GTTTTG +GTTTCA +GTTTCT +GTTTCC +GTTTCG +GTTTGA +GTTTGT +GTTTGC +GTTTGG +GTTCAA +GTTCAT +GTTCAC +GTTCAG +GTTCTA +GTTCTT +GTTCTC +GTTCTG +GTTCCA +GTTCCT +GTTCCC +GTTCCG +GTTCGA +GTTCGT +GTTCGC +GTTCGG +GTTGAA +GTTGAT +GTTGAC +GTTGAG +GTTGTA +GTTGTT +GTTGTC +GTTGTG +GTTGCA +GTTGCT +GTTGCC +GTTGCG +GTTGGA +GTTGGT +GTTGGC +GTTGGG +GTCAAA +GTCAAT +GTCAAC +GTCAAG +GTCATA +GTCATT +GTCATC +GTCATG +GTCACA +GTCACT +GTCACC +GTCACG +GTCAGA +GTCAGT +GTCAGC +GTCAGG +GTCTAA +GTCTAT +GTCTAC +GTCTAG +GTCTTA +GTCTTT +GTCTTC +GTCTTG +GTCTCA +GTCTCT +GTCTCC +GTCTCG +GTCTGA +GTCTGT +GTCTGC +GTCTGG +GTCCAA +GTCCAT +GTCCAC +GTCCAG +GTCCTA +GTCCTT +GTCCTC +GTCCTG +GTCCCA +GTCCCT +GTCCCC +GTCCCG +GTCCGA +GTCCGT +GTCCGC +GTCCGG +GTCGAA +GTCGAT +GTCGAC +GTCGAG +GTCGTA +GTCGTT +GTCGTC +GTCGTG +GTCGCA +GTCGCT +GTCGCC +GTCGCG +GTCGGA +GTCGGT +GTCGGC +GTCGGG +GTGAAA +GTGAAT +GTGAAC +GTGAAG +GTGATA +GTGATT +GTGATC +GTGATG +GTGACA +GTGACT +GTGACC +GTGACG +GTGAGA +GTGAGT +GTGAGC +GTGAGG +GTGTAA +GTGTAT +GTGTAC +GTGTAG +GTGTTA +GTGTTT +GTGTTC +GTGTTG +GTGTCA +GTGTCT +GTGTCC +GTGTCG +GTGTGA +GTGTGT +GTGTGC +GTGTGG +GTGCAA +GTGCAT +GTGCAC +GTGCAG +GTGCTA +GTGCTT +GTGCTC +GTGCTG +GTGCCA +GTGCCT +GTGCCC +GTGCCG +GTGCGA +GTGCGT +GTGCGC +GTGCGG +GTGGAA +GTGGAT +GTGGAC +GTGGAG +GTGGTA +GTGGTT +GTGGTC +GTGGTG +GTGGCA +GTGGCT +GTGGCC +GTGGCG +GTGGGA +GTGGGT +GTGGGC +GTGGGG +GCAAAA +GCAAAT +GCAAAC +GCAAAG +GCAATA +GCAATT +GCAATC +GCAATG +GCAACA +GCAACT +GCAACC +GCAACG +GCAAGA +GCAAGT +GCAAGC +GCAAGG +GCATAA +GCATAT +GCATAC +GCATAG +GCATTA +GCATTT +GCATTC +GCATTG +GCATCA +GCATCT +GCATCC +GCATCG +GCATGA +GCATGT +GCATGC +GCATGG +GCACAA +GCACAT +GCACAC +GCACAG +GCACTA +GCACTT +GCACTC +GCACTG +GCACCA +GCACCT +GCACCC +GCACCG +GCACGA +GCACGT +GCACGC +GCACGG +GCAGAA +GCAGAT +GCAGAC +GCAGAG +GCAGTA +GCAGTT +GCAGTC +GCAGTG +GCAGCA +GCAGCT +GCAGCC +GCAGCG +GCAGGA +GCAGGT +GCAGGC +GCAGGG +GCTAAA +GCTAAT +GCTAAC +GCTAAG +GCTATA +GCTATT +GCTATC +GCTATG +GCTACA +GCTACT +GCTACC +GCTACG +GCTAGA +GCTAGT +GCTAGC +GCTAGG +GCTTAA +GCTTAT +GCTTAC +GCTTAG +GCTTTA +GCTTTT +GCTTTC +GCTTTG +GCTTCA +GCTTCT +GCTTCC +GCTTCG +GCTTGA +GCTTGT +GCTTGC +GCTTGG +GCTCAA +GCTCAT +GCTCAC +GCTCAG +GCTCTA +GCTCTT +GCTCTC +GCTCTG +GCTCCA +GCTCCT +GCTCCC +GCTCCG +GCTCGA +GCTCGT +GCTCGC +GCTCGG +GCTGAA +GCTGAT +GCTGAC +GCTGAG +GCTGTA +GCTGTT +GCTGTC +GCTGTG +GCTGCA +GCTGCT +GCTGCC +GCTGCG +GCTGGA +GCTGGT +GCTGGC +GCTGGG +GCCAAA +GCCAAT +GCCAAC +GCCAAG +GCCATA +GCCATT +GCCATC +GCCATG +GCCACA +GCCACT +GCCACC +GCCACG +GCCAGA +GCCAGT +GCCAGC +GCCAGG +GCCTAA +GCCTAT +GCCTAC +GCCTAG +GCCTTA +GCCTTT +GCCTTC +GCCTTG +GCCTCA +GCCTCT +GCCTCC +GCCTCG +GCCTGA +GCCTGT +GCCTGC +GCCTGG +GCCCAA +GCCCAT +GCCCAC +GCCCAG +GCCCTA +GCCCTT +GCCCTC +GCCCTG +GCCCCA +GCCCCT +GCCCCC +GCCCCG +GCCCGA +GCCCGT +GCCCGC +GCCCGG +GCCGAA +GCCGAT +GCCGAC +GCCGAG +GCCGTA +GCCGTT +GCCGTC +GCCGTG +GCCGCA +GCCGCT +GCCGCC +GCCGCG +GCCGGA +GCCGGT +GCCGGC +GCCGGG +GCGAAA +GCGAAT +GCGAAC +GCGAAG +GCGATA +GCGATT +GCGATC +GCGATG +GCGACA +GCGACT +GCGACC +GCGACG +GCGAGA +GCGAGT +GCGAGC +GCGAGG +GCGTAA +GCGTAT +GCGTAC +GCGTAG +GCGTTA +GCGTTT +GCGTTC +GCGTTG +GCGTCA +GCGTCT +GCGTCC +GCGTCG +GCGTGA +GCGTGT +GCGTGC +GCGTGG +GCGCAA +GCGCAT +GCGCAC +GCGCAG +GCGCTA +GCGCTT +GCGCTC +GCGCTG +GCGCCA +GCGCCT +GCGCCC +GCGCCG +GCGCGA +GCGCGT +GCGCGC +GCGCGG +GCGGAA +GCGGAT +GCGGAC +GCGGAG +GCGGTA +GCGGTT +GCGGTC +GCGGTG +GCGGCA +GCGGCT +GCGGCC +GCGGCG +GCGGGA +GCGGGT +GCGGGC +GCGGGG +GGAAAA +GGAAAT +GGAAAC +GGAAAG +GGAATA +GGAATT +GGAATC +GGAATG +GGAACA +GGAACT +GGAACC +GGAACG +GGAAGA +GGAAGT +GGAAGC +GGAAGG +GGATAA +GGATAT +GGATAC +GGATAG +GGATTA +GGATTT +GGATTC +GGATTG +GGATCA +GGATCT +GGATCC +GGATCG +GGATGA +GGATGT +GGATGC +GGATGG +GGACAA +GGACAT +GGACAC +GGACAG +GGACTA +GGACTT +GGACTC +GGACTG +GGACCA +GGACCT +GGACCC +GGACCG +GGACGA +GGACGT +GGACGC +GGACGG +GGAGAA +GGAGAT +GGAGAC +GGAGAG +GGAGTA +GGAGTT +GGAGTC +GGAGTG +GGAGCA +GGAGCT +GGAGCC +GGAGCG +GGAGGA +GGAGGT +GGAGGC +GGAGGG +GGTAAA +GGTAAT +GGTAAC +GGTAAG +GGTATA +GGTATT +GGTATC +GGTATG +GGTACA +GGTACT +GGTACC +GGTACG +GGTAGA +GGTAGT +GGTAGC +GGTAGG +GGTTAA +GGTTAT +GGTTAC +GGTTAG +GGTTTA +GGTTTT +GGTTTC +GGTTTG +GGTTCA +GGTTCT +GGTTCC +GGTTCG +GGTTGA +GGTTGT +GGTTGC +GGTTGG +GGTCAA +GGTCAT +GGTCAC +GGTCAG +GGTCTA +GGTCTT +GGTCTC +GGTCTG +GGTCCA +GGTCCT +GGTCCC +GGTCCG +GGTCGA +GGTCGT +GGTCGC +GGTCGG +GGTGAA +GGTGAT +GGTGAC +GGTGAG +GGTGTA +GGTGTT +GGTGTC +GGTGTG +GGTGCA +GGTGCT +GGTGCC +GGTGCG +GGTGGA +GGTGGT +GGTGGC +GGTGGG +GGCAAA +GGCAAT +GGCAAC +GGCAAG +GGCATA +GGCATT +GGCATC +GGCATG +GGCACA +GGCACT +GGCACC +GGCACG +GGCAGA +GGCAGT +GGCAGC +GGCAGG +GGCTAA +GGCTAT +GGCTAC +GGCTAG +GGCTTA +GGCTTT +GGCTTC +GGCTTG +GGCTCA +GGCTCT +GGCTCC +GGCTCG +GGCTGA +GGCTGT +GGCTGC +GGCTGG +GGCCAA +GGCCAT +GGCCAC +GGCCAG +GGCCTA +GGCCTT +GGCCTC +GGCCTG +GGCCCA +GGCCCT +GGCCCC +GGCCCG +GGCCGA +GGCCGT +GGCCGC +GGCCGG +GGCGAA +GGCGAT +GGCGAC +GGCGAG +GGCGTA +GGCGTT +GGCGTC +GGCGTG +GGCGCA +GGCGCT +GGCGCC +GGCGCG +GGCGGA +GGCGGT +GGCGGC +GGCGGG +GGGAAA +GGGAAT +GGGAAC +GGGAAG +GGGATA +GGGATT +GGGATC +GGGATG +GGGACA +GGGACT +GGGACC +GGGACG +GGGAGA +GGGAGT +GGGAGC +GGGAGG +GGGTAA +GGGTAT +GGGTAC +GGGTAG +GGGTTA +GGGTTT +GGGTTC +GGGTTG +GGGTCA +GGGTCT +GGGTCC +GGGTCG +GGGTGA +GGGTGT +GGGTGC +GGGTGG +GGGCAA +GGGCAT +GGGCAC +GGGCAG +GGGCTA +GGGCTT +GGGCTC +GGGCTG +GGGCCA +GGGCCT +GGGCCC +GGGCCG +GGGCGA +GGGCGT +GGGCGC +GGGCGG +GGGGAA +GGGGAT +GGGGAC +GGGGAG +GGGGTA +GGGGTT +GGGGTC +GGGGTG +GGGGCA +GGGGCT +GGGGCC +GGGGCG +GGGGGA +GGGGGT +GGGGGC +GGGGGG diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-2000/config.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..9b14e61d327bf586fcff6e817cf9d365bae18ad2 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-2000/config.json @@ -0,0 +1,26 @@ +{ + "_name_or_path": "fabihamakhdoomi/TinyDNABERT", + "architectures": [ + "BertForSequenceClassification" + ], + "attention_probs_dropout_prob": 0.1, + "classifier_dropout": null, + "hidden_act": "gelu", + "hidden_dropout_prob": 0.1, + "hidden_size": 128, + "initializer_range": 0.02, + "intermediate_size": 512, + "layer_norm_eps": 1e-12, + "max_position_embeddings": 512, + "model_type": "bert", + "num_attention_heads": 2, + "num_hidden_layers": 2, + "pad_token_id": 0, + "position_embedding_type": "absolute", + "problem_type": "single_label_classification", + "torch_dtype": "float32", + "transformers_version": "4.44.0", + "type_vocab_size": 2, + "use_cache": true, + "vocab_size": 4101 +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-2000/model.safetensors b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-2000/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..148bf903d3e49085062cce14e25318bb96cfdc6c --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-2000/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c9d6a798369cceb122f17267595079cb4fa5146fd81f08366a4a99626828facb +size 4021688 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-2000/optimizer.pt b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-2000/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..3b87bd99b288be888629a380bc82c4cb7fb757ed --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-2000/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:45c74c5ab11f6657eddd53ff8077d4492ee89094872ccc177a39616f16727c51 +size 8058700 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-2000/rng_state.pth b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-2000/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..dc62b940907551fbd6412f1b948afb4ecb3c061f --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-2000/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b1753d15a4935b9bbb100d316f8edf3879525b0e6b5b0c8edeffb1787132e3cd +size 14244 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-2000/scheduler.pt b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-2000/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..2287c4e5e02bffea6969638a53c3d5402c6dc2b0 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-2000/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:da119a7607c7d2d96a92ebe57fbaf5fb22ebfce3399b1702b2209f460c0bfe35 +size 1064 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-2000/special_tokens_map.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-2000/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..a8b3208c2884c4efb86e49300fdd3dc877220cdf --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-2000/special_tokens_map.json @@ -0,0 +1,7 @@ +{ + "cls_token": "[CLS]", + "mask_token": "[MASK]", + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-2000/tokenizer.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-2000/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..ebd782de6963b9032ce9374cfe30f5275be1a30b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-2000/tokenizer.json @@ -0,0 +1,4265 @@ +{ + "version": "1.0", + "truncation": { + "direction": "Right", + "max_length": 512, + "strategy": "LongestFirst", + "stride": 0 + }, + "padding": { + "strategy": { + "Fixed": 512 + }, + "direction": "Right", + "pad_to_multiple_of": null, + "pad_id": 0, + "pad_type_id": 0, + "pad_token": "[PAD]" + }, + "added_tokens": [ + { + "id": 0, + "content": "[PAD]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 1, + "content": "[UNK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 2, + "content": "[CLS]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 3, + "content": "[SEP]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 4, + "content": "[MASK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + } + ], + "normalizer": { + "type": "BertNormalizer", + "clean_text": true, + "handle_chinese_chars": true, + "strip_accents": null, + "lowercase": false + }, + "pre_tokenizer": { + "type": "BertPreTokenizer" + }, + "post_processor": { + "type": "TemplateProcessing", + "single": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + } + ], + "pair": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "B", + "type_id": 1 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 1 + } + } + ], + "special_tokens": { + "[CLS]": { + "id": "[CLS]", + "ids": [ + 2 + ], + "tokens": [ + "[CLS]" + ] + }, + "[SEP]": { + "id": "[SEP]", + "ids": [ + 3 + ], + "tokens": [ + "[SEP]" + ] + } + } + }, + "decoder": { + "type": "WordPiece", + "prefix": "##", + "cleanup": true + }, + "model": { + "type": "WordPiece", + "unk_token": "[UNK]", + "continuing_subword_prefix": "##", + "max_input_chars_per_word": 100, + "vocab": { + "[PAD]": 0, + "[UNK]": 1, + "[CLS]": 2, + "[SEP]": 3, + "[MASK]": 4, + "AAAAAA": 5, + "AAAAAT": 6, + "AAAAAC": 7, + "AAAAAG": 8, + "AAAATA": 9, + "AAAATT": 10, + "AAAATC": 11, + "AAAATG": 12, + "AAAACA": 13, + "AAAACT": 14, + "AAAACC": 15, + "AAAACG": 16, + "AAAAGA": 17, + "AAAAGT": 18, + "AAAAGC": 19, + "AAAAGG": 20, + "AAATAA": 21, + "AAATAT": 22, + "AAATAC": 23, + "AAATAG": 24, + "AAATTA": 25, + "AAATTT": 26, + "AAATTC": 27, + "AAATTG": 28, + "AAATCA": 29, + "AAATCT": 30, + "AAATCC": 31, + "AAATCG": 32, + "AAATGA": 33, + "AAATGT": 34, + "AAATGC": 35, + "AAATGG": 36, + "AAACAA": 37, + "AAACAT": 38, + "AAACAC": 39, + "AAACAG": 40, + "AAACTA": 41, + "AAACTT": 42, + "AAACTC": 43, + "AAACTG": 44, + "AAACCA": 45, + "AAACCT": 46, + "AAACCC": 47, + "AAACCG": 48, + "AAACGA": 49, + "AAACGT": 50, + "AAACGC": 51, + "AAACGG": 52, + "AAAGAA": 53, + "AAAGAT": 54, + "AAAGAC": 55, + "AAAGAG": 56, + "AAAGTA": 57, + "AAAGTT": 58, + "AAAGTC": 59, + "AAAGTG": 60, + "AAAGCA": 61, + "AAAGCT": 62, + "AAAGCC": 63, + "AAAGCG": 64, + "AAAGGA": 65, + "AAAGGT": 66, + "AAAGGC": 67, + "AAAGGG": 68, + "AATAAA": 69, + "AATAAT": 70, + "AATAAC": 71, + "AATAAG": 72, + "AATATA": 73, + "AATATT": 74, + "AATATC": 75, + "AATATG": 76, + "AATACA": 77, + "AATACT": 78, + "AATACC": 79, + "AATACG": 80, + "AATAGA": 81, + "AATAGT": 82, + "AATAGC": 83, + "AATAGG": 84, + "AATTAA": 85, + "AATTAT": 86, + "AATTAC": 87, + "AATTAG": 88, + "AATTTA": 89, + "AATTTT": 90, + "AATTTC": 91, + "AATTTG": 92, + "AATTCA": 93, + "AATTCT": 94, + "AATTCC": 95, + "AATTCG": 96, + "AATTGA": 97, + "AATTGT": 98, + "AATTGC": 99, + "AATTGG": 100, + "AATCAA": 101, + "AATCAT": 102, + "AATCAC": 103, + "AATCAG": 104, + "AATCTA": 105, + "AATCTT": 106, + "AATCTC": 107, + "AATCTG": 108, + "AATCCA": 109, + "AATCCT": 110, + "AATCCC": 111, + "AATCCG": 112, + "AATCGA": 113, + "AATCGT": 114, + "AATCGC": 115, + "AATCGG": 116, + "AATGAA": 117, + "AATGAT": 118, + "AATGAC": 119, + "AATGAG": 120, + "AATGTA": 121, + "AATGTT": 122, + "AATGTC": 123, + "AATGTG": 124, + "AATGCA": 125, + "AATGCT": 126, + "AATGCC": 127, + "AATGCG": 128, + "AATGGA": 129, + "AATGGT": 130, + "AATGGC": 131, + "AATGGG": 132, + "AACAAA": 133, + "AACAAT": 134, + "AACAAC": 135, + "AACAAG": 136, + "AACATA": 137, + "AACATT": 138, + "AACATC": 139, + "AACATG": 140, + "AACACA": 141, + "AACACT": 142, + "AACACC": 143, + "AACACG": 144, + "AACAGA": 145, + "AACAGT": 146, + "AACAGC": 147, + "AACAGG": 148, + "AACTAA": 149, + "AACTAT": 150, + "AACTAC": 151, + "AACTAG": 152, + "AACTTA": 153, + "AACTTT": 154, + "AACTTC": 155, + "AACTTG": 156, + "AACTCA": 157, + "AACTCT": 158, + "AACTCC": 159, + "AACTCG": 160, + "AACTGA": 161, + "AACTGT": 162, + "AACTGC": 163, + "AACTGG": 164, + "AACCAA": 165, + "AACCAT": 166, + "AACCAC": 167, + "AACCAG": 168, + "AACCTA": 169, + "AACCTT": 170, + "AACCTC": 171, + "AACCTG": 172, + "AACCCA": 173, + "AACCCT": 174, + "AACCCC": 175, + "AACCCG": 176, + "AACCGA": 177, + "AACCGT": 178, + "AACCGC": 179, + "AACCGG": 180, + "AACGAA": 181, + "AACGAT": 182, + "AACGAC": 183, + "AACGAG": 184, + "AACGTA": 185, + "AACGTT": 186, + "AACGTC": 187, + "AACGTG": 188, + "AACGCA": 189, + "AACGCT": 190, + "AACGCC": 191, + "AACGCG": 192, + "AACGGA": 193, + "AACGGT": 194, + "AACGGC": 195, + "AACGGG": 196, + "AAGAAA": 197, + "AAGAAT": 198, + "AAGAAC": 199, + "AAGAAG": 200, + "AAGATA": 201, + "AAGATT": 202, + "AAGATC": 203, + "AAGATG": 204, + "AAGACA": 205, + "AAGACT": 206, + "AAGACC": 207, + "AAGACG": 208, + "AAGAGA": 209, + "AAGAGT": 210, + "AAGAGC": 211, + "AAGAGG": 212, + "AAGTAA": 213, + "AAGTAT": 214, + "AAGTAC": 215, + "AAGTAG": 216, + "AAGTTA": 217, + "AAGTTT": 218, + "AAGTTC": 219, + "AAGTTG": 220, + "AAGTCA": 221, + "AAGTCT": 222, + "AAGTCC": 223, + "AAGTCG": 224, + "AAGTGA": 225, + "AAGTGT": 226, + "AAGTGC": 227, + "AAGTGG": 228, + "AAGCAA": 229, + "AAGCAT": 230, + "AAGCAC": 231, + "AAGCAG": 232, + "AAGCTA": 233, + "AAGCTT": 234, + "AAGCTC": 235, + "AAGCTG": 236, + "AAGCCA": 237, + "AAGCCT": 238, + "AAGCCC": 239, + "AAGCCG": 240, + "AAGCGA": 241, + "AAGCGT": 242, + "AAGCGC": 243, + "AAGCGG": 244, + "AAGGAA": 245, + "AAGGAT": 246, + "AAGGAC": 247, + "AAGGAG": 248, + "AAGGTA": 249, + "AAGGTT": 250, + "AAGGTC": 251, + "AAGGTG": 252, + "AAGGCA": 253, + "AAGGCT": 254, + "AAGGCC": 255, + "AAGGCG": 256, + "AAGGGA": 257, + "AAGGGT": 258, + "AAGGGC": 259, + "AAGGGG": 260, + "ATAAAA": 261, + "ATAAAT": 262, + "ATAAAC": 263, + "ATAAAG": 264, + "ATAATA": 265, + "ATAATT": 266, + "ATAATC": 267, + "ATAATG": 268, + "ATAACA": 269, + "ATAACT": 270, + "ATAACC": 271, + "ATAACG": 272, + "ATAAGA": 273, + "ATAAGT": 274, + "ATAAGC": 275, + "ATAAGG": 276, + "ATATAA": 277, + "ATATAT": 278, + "ATATAC": 279, + "ATATAG": 280, + "ATATTA": 281, + "ATATTT": 282, + "ATATTC": 283, + "ATATTG": 284, + "ATATCA": 285, + "ATATCT": 286, + "ATATCC": 287, + "ATATCG": 288, + "ATATGA": 289, + "ATATGT": 290, + "ATATGC": 291, + "ATATGG": 292, + "ATACAA": 293, + "ATACAT": 294, + "ATACAC": 295, + "ATACAG": 296, + "ATACTA": 297, + "ATACTT": 298, + "ATACTC": 299, + "ATACTG": 300, + "ATACCA": 301, + "ATACCT": 302, + "ATACCC": 303, + "ATACCG": 304, + "ATACGA": 305, + "ATACGT": 306, + "ATACGC": 307, + "ATACGG": 308, + "ATAGAA": 309, + "ATAGAT": 310, + "ATAGAC": 311, + "ATAGAG": 312, + "ATAGTA": 313, + "ATAGTT": 314, + "ATAGTC": 315, + "ATAGTG": 316, + "ATAGCA": 317, + "ATAGCT": 318, + "ATAGCC": 319, + "ATAGCG": 320, + "ATAGGA": 321, + "ATAGGT": 322, + "ATAGGC": 323, + "ATAGGG": 324, + "ATTAAA": 325, + "ATTAAT": 326, + "ATTAAC": 327, + "ATTAAG": 328, + "ATTATA": 329, + "ATTATT": 330, + "ATTATC": 331, + "ATTATG": 332, + "ATTACA": 333, + "ATTACT": 334, + "ATTACC": 335, + "ATTACG": 336, + "ATTAGA": 337, + "ATTAGT": 338, + "ATTAGC": 339, + "ATTAGG": 340, + "ATTTAA": 341, + "ATTTAT": 342, + "ATTTAC": 343, + "ATTTAG": 344, + "ATTTTA": 345, + "ATTTTT": 346, + "ATTTTC": 347, + "ATTTTG": 348, + "ATTTCA": 349, + "ATTTCT": 350, + "ATTTCC": 351, + "ATTTCG": 352, + "ATTTGA": 353, + "ATTTGT": 354, + "ATTTGC": 355, + "ATTTGG": 356, + "ATTCAA": 357, + "ATTCAT": 358, + "ATTCAC": 359, + "ATTCAG": 360, + "ATTCTA": 361, + "ATTCTT": 362, + "ATTCTC": 363, + "ATTCTG": 364, + "ATTCCA": 365, + "ATTCCT": 366, + "ATTCCC": 367, + "ATTCCG": 368, + "ATTCGA": 369, + "ATTCGT": 370, + "ATTCGC": 371, + "ATTCGG": 372, + "ATTGAA": 373, + "ATTGAT": 374, + "ATTGAC": 375, + "ATTGAG": 376, + "ATTGTA": 377, + "ATTGTT": 378, + "ATTGTC": 379, + "ATTGTG": 380, + "ATTGCA": 381, + "ATTGCT": 382, + "ATTGCC": 383, + "ATTGCG": 384, + "ATTGGA": 385, + "ATTGGT": 386, + "ATTGGC": 387, + "ATTGGG": 388, + "ATCAAA": 389, + "ATCAAT": 390, + "ATCAAC": 391, + "ATCAAG": 392, + "ATCATA": 393, + "ATCATT": 394, + "ATCATC": 395, + "ATCATG": 396, + "ATCACA": 397, + "ATCACT": 398, + "ATCACC": 399, + "ATCACG": 400, + "ATCAGA": 401, + "ATCAGT": 402, + "ATCAGC": 403, + "ATCAGG": 404, + "ATCTAA": 405, + "ATCTAT": 406, + "ATCTAC": 407, + "ATCTAG": 408, + "ATCTTA": 409, + "ATCTTT": 410, + "ATCTTC": 411, + "ATCTTG": 412, + "ATCTCA": 413, + "ATCTCT": 414, + "ATCTCC": 415, + "ATCTCG": 416, + "ATCTGA": 417, + "ATCTGT": 418, + "ATCTGC": 419, + "ATCTGG": 420, + "ATCCAA": 421, + "ATCCAT": 422, + "ATCCAC": 423, + "ATCCAG": 424, + "ATCCTA": 425, + "ATCCTT": 426, + "ATCCTC": 427, + "ATCCTG": 428, + "ATCCCA": 429, + "ATCCCT": 430, + "ATCCCC": 431, + "ATCCCG": 432, + "ATCCGA": 433, + "ATCCGT": 434, + "ATCCGC": 435, + "ATCCGG": 436, + "ATCGAA": 437, + "ATCGAT": 438, + "ATCGAC": 439, + "ATCGAG": 440, + "ATCGTA": 441, + "ATCGTT": 442, + "ATCGTC": 443, + "ATCGTG": 444, + "ATCGCA": 445, + "ATCGCT": 446, + "ATCGCC": 447, + "ATCGCG": 448, + "ATCGGA": 449, + "ATCGGT": 450, + "ATCGGC": 451, + "ATCGGG": 452, + "ATGAAA": 453, + "ATGAAT": 454, + "ATGAAC": 455, + "ATGAAG": 456, + "ATGATA": 457, + "ATGATT": 458, + "ATGATC": 459, + "ATGATG": 460, + "ATGACA": 461, + "ATGACT": 462, + "ATGACC": 463, + "ATGACG": 464, + "ATGAGA": 465, + "ATGAGT": 466, + "ATGAGC": 467, + "ATGAGG": 468, + "ATGTAA": 469, + "ATGTAT": 470, + "ATGTAC": 471, + "ATGTAG": 472, + "ATGTTA": 473, + "ATGTTT": 474, + "ATGTTC": 475, + "ATGTTG": 476, + "ATGTCA": 477, + "ATGTCT": 478, + "ATGTCC": 479, + "ATGTCG": 480, + "ATGTGA": 481, + "ATGTGT": 482, + "ATGTGC": 483, + "ATGTGG": 484, + "ATGCAA": 485, + "ATGCAT": 486, + "ATGCAC": 487, + "ATGCAG": 488, + "ATGCTA": 489, + "ATGCTT": 490, + "ATGCTC": 491, + "ATGCTG": 492, + "ATGCCA": 493, + "ATGCCT": 494, + "ATGCCC": 495, + "ATGCCG": 496, + "ATGCGA": 497, + "ATGCGT": 498, + "ATGCGC": 499, + "ATGCGG": 500, + "ATGGAA": 501, + "ATGGAT": 502, + "ATGGAC": 503, + "ATGGAG": 504, + "ATGGTA": 505, + "ATGGTT": 506, + "ATGGTC": 507, + "ATGGTG": 508, + "ATGGCA": 509, + "ATGGCT": 510, + "ATGGCC": 511, + "ATGGCG": 512, + "ATGGGA": 513, + "ATGGGT": 514, + "ATGGGC": 515, + "ATGGGG": 516, + "ACAAAA": 517, + "ACAAAT": 518, + "ACAAAC": 519, + "ACAAAG": 520, + "ACAATA": 521, + "ACAATT": 522, + "ACAATC": 523, + "ACAATG": 524, + "ACAACA": 525, + "ACAACT": 526, + "ACAACC": 527, + "ACAACG": 528, + "ACAAGA": 529, + "ACAAGT": 530, + "ACAAGC": 531, + "ACAAGG": 532, + "ACATAA": 533, + "ACATAT": 534, + "ACATAC": 535, + "ACATAG": 536, + "ACATTA": 537, + "ACATTT": 538, + "ACATTC": 539, + "ACATTG": 540, + "ACATCA": 541, + "ACATCT": 542, + "ACATCC": 543, + "ACATCG": 544, + "ACATGA": 545, + "ACATGT": 546, + "ACATGC": 547, + "ACATGG": 548, + "ACACAA": 549, + "ACACAT": 550, + "ACACAC": 551, + "ACACAG": 552, + "ACACTA": 553, + "ACACTT": 554, + "ACACTC": 555, + "ACACTG": 556, + "ACACCA": 557, + "ACACCT": 558, + "ACACCC": 559, + "ACACCG": 560, + "ACACGA": 561, + "ACACGT": 562, + "ACACGC": 563, + "ACACGG": 564, + "ACAGAA": 565, + "ACAGAT": 566, + "ACAGAC": 567, + "ACAGAG": 568, + "ACAGTA": 569, + "ACAGTT": 570, + "ACAGTC": 571, + "ACAGTG": 572, + "ACAGCA": 573, + "ACAGCT": 574, + "ACAGCC": 575, + "ACAGCG": 576, + "ACAGGA": 577, + "ACAGGT": 578, + "ACAGGC": 579, + "ACAGGG": 580, + "ACTAAA": 581, + "ACTAAT": 582, + "ACTAAC": 583, + "ACTAAG": 584, + "ACTATA": 585, + "ACTATT": 586, + "ACTATC": 587, + "ACTATG": 588, + "ACTACA": 589, + "ACTACT": 590, + "ACTACC": 591, + "ACTACG": 592, + "ACTAGA": 593, + "ACTAGT": 594, + "ACTAGC": 595, + "ACTAGG": 596, + "ACTTAA": 597, + "ACTTAT": 598, + "ACTTAC": 599, + "ACTTAG": 600, + "ACTTTA": 601, + "ACTTTT": 602, + "ACTTTC": 603, + "ACTTTG": 604, + "ACTTCA": 605, + "ACTTCT": 606, + "ACTTCC": 607, + "ACTTCG": 608, + "ACTTGA": 609, + "ACTTGT": 610, + "ACTTGC": 611, + "ACTTGG": 612, + "ACTCAA": 613, + "ACTCAT": 614, + "ACTCAC": 615, + "ACTCAG": 616, + "ACTCTA": 617, + "ACTCTT": 618, + "ACTCTC": 619, + "ACTCTG": 620, + "ACTCCA": 621, + "ACTCCT": 622, + "ACTCCC": 623, + "ACTCCG": 624, + "ACTCGA": 625, + "ACTCGT": 626, + "ACTCGC": 627, + "ACTCGG": 628, + "ACTGAA": 629, + "ACTGAT": 630, + "ACTGAC": 631, + "ACTGAG": 632, + "ACTGTA": 633, + "ACTGTT": 634, + "ACTGTC": 635, + "ACTGTG": 636, + "ACTGCA": 637, + "ACTGCT": 638, + "ACTGCC": 639, + "ACTGCG": 640, + "ACTGGA": 641, + "ACTGGT": 642, + "ACTGGC": 643, + "ACTGGG": 644, + "ACCAAA": 645, + "ACCAAT": 646, + "ACCAAC": 647, + "ACCAAG": 648, + "ACCATA": 649, + "ACCATT": 650, + "ACCATC": 651, + "ACCATG": 652, + "ACCACA": 653, + "ACCACT": 654, + "ACCACC": 655, + "ACCACG": 656, + "ACCAGA": 657, + "ACCAGT": 658, + "ACCAGC": 659, + "ACCAGG": 660, + "ACCTAA": 661, + "ACCTAT": 662, + "ACCTAC": 663, + "ACCTAG": 664, + "ACCTTA": 665, + "ACCTTT": 666, + "ACCTTC": 667, + "ACCTTG": 668, + "ACCTCA": 669, + "ACCTCT": 670, + "ACCTCC": 671, + "ACCTCG": 672, + "ACCTGA": 673, + "ACCTGT": 674, + "ACCTGC": 675, + "ACCTGG": 676, + "ACCCAA": 677, + "ACCCAT": 678, + "ACCCAC": 679, + "ACCCAG": 680, + "ACCCTA": 681, + "ACCCTT": 682, + "ACCCTC": 683, + "ACCCTG": 684, + "ACCCCA": 685, + "ACCCCT": 686, + "ACCCCC": 687, + "ACCCCG": 688, + "ACCCGA": 689, + "ACCCGT": 690, + "ACCCGC": 691, + "ACCCGG": 692, + "ACCGAA": 693, + "ACCGAT": 694, + "ACCGAC": 695, + "ACCGAG": 696, + "ACCGTA": 697, + "ACCGTT": 698, + "ACCGTC": 699, + "ACCGTG": 700, + "ACCGCA": 701, + "ACCGCT": 702, + "ACCGCC": 703, + "ACCGCG": 704, + "ACCGGA": 705, + "ACCGGT": 706, + "ACCGGC": 707, + "ACCGGG": 708, + "ACGAAA": 709, + "ACGAAT": 710, + "ACGAAC": 711, + "ACGAAG": 712, + "ACGATA": 713, + "ACGATT": 714, + "ACGATC": 715, + "ACGATG": 716, + "ACGACA": 717, + "ACGACT": 718, + "ACGACC": 719, + "ACGACG": 720, + "ACGAGA": 721, + "ACGAGT": 722, + "ACGAGC": 723, + "ACGAGG": 724, + "ACGTAA": 725, + "ACGTAT": 726, + "ACGTAC": 727, + "ACGTAG": 728, + "ACGTTA": 729, + "ACGTTT": 730, + "ACGTTC": 731, + "ACGTTG": 732, + "ACGTCA": 733, + "ACGTCT": 734, + "ACGTCC": 735, + "ACGTCG": 736, + "ACGTGA": 737, + "ACGTGT": 738, + "ACGTGC": 739, + "ACGTGG": 740, + "ACGCAA": 741, + "ACGCAT": 742, + "ACGCAC": 743, + "ACGCAG": 744, + "ACGCTA": 745, + "ACGCTT": 746, + "ACGCTC": 747, + "ACGCTG": 748, + "ACGCCA": 749, + "ACGCCT": 750, + "ACGCCC": 751, + "ACGCCG": 752, + "ACGCGA": 753, + "ACGCGT": 754, + "ACGCGC": 755, + "ACGCGG": 756, + "ACGGAA": 757, + "ACGGAT": 758, + "ACGGAC": 759, + "ACGGAG": 760, + "ACGGTA": 761, + "ACGGTT": 762, + "ACGGTC": 763, + "ACGGTG": 764, + "ACGGCA": 765, + "ACGGCT": 766, + "ACGGCC": 767, + "ACGGCG": 768, + "ACGGGA": 769, + "ACGGGT": 770, + "ACGGGC": 771, + "ACGGGG": 772, + "AGAAAA": 773, + "AGAAAT": 774, + "AGAAAC": 775, + "AGAAAG": 776, + "AGAATA": 777, + "AGAATT": 778, + "AGAATC": 779, + "AGAATG": 780, + "AGAACA": 781, + "AGAACT": 782, + "AGAACC": 783, + "AGAACG": 784, + "AGAAGA": 785, + "AGAAGT": 786, + "AGAAGC": 787, + "AGAAGG": 788, + "AGATAA": 789, + "AGATAT": 790, + "AGATAC": 791, + "AGATAG": 792, + "AGATTA": 793, + "AGATTT": 794, + "AGATTC": 795, + "AGATTG": 796, + "AGATCA": 797, + "AGATCT": 798, + "AGATCC": 799, + "AGATCG": 800, + "AGATGA": 801, + "AGATGT": 802, + "AGATGC": 803, + "AGATGG": 804, + "AGACAA": 805, + "AGACAT": 806, + "AGACAC": 807, + "AGACAG": 808, + "AGACTA": 809, + "AGACTT": 810, + "AGACTC": 811, + "AGACTG": 812, + "AGACCA": 813, + "AGACCT": 814, + "AGACCC": 815, + "AGACCG": 816, + "AGACGA": 817, + "AGACGT": 818, + "AGACGC": 819, + "AGACGG": 820, + "AGAGAA": 821, + "AGAGAT": 822, + "AGAGAC": 823, + "AGAGAG": 824, + "AGAGTA": 825, + "AGAGTT": 826, + "AGAGTC": 827, + "AGAGTG": 828, + "AGAGCA": 829, + "AGAGCT": 830, + "AGAGCC": 831, + "AGAGCG": 832, + "AGAGGA": 833, + "AGAGGT": 834, + "AGAGGC": 835, + "AGAGGG": 836, + "AGTAAA": 837, + "AGTAAT": 838, + "AGTAAC": 839, + "AGTAAG": 840, + "AGTATA": 841, + "AGTATT": 842, + "AGTATC": 843, + "AGTATG": 844, + "AGTACA": 845, + "AGTACT": 846, + "AGTACC": 847, + "AGTACG": 848, + "AGTAGA": 849, + "AGTAGT": 850, + "AGTAGC": 851, + "AGTAGG": 852, + "AGTTAA": 853, + "AGTTAT": 854, + "AGTTAC": 855, + "AGTTAG": 856, + "AGTTTA": 857, + "AGTTTT": 858, + "AGTTTC": 859, + "AGTTTG": 860, + "AGTTCA": 861, + "AGTTCT": 862, + "AGTTCC": 863, + "AGTTCG": 864, + "AGTTGA": 865, + "AGTTGT": 866, + "AGTTGC": 867, + "AGTTGG": 868, + "AGTCAA": 869, + "AGTCAT": 870, + "AGTCAC": 871, + "AGTCAG": 872, + "AGTCTA": 873, + "AGTCTT": 874, + "AGTCTC": 875, + "AGTCTG": 876, + "AGTCCA": 877, + "AGTCCT": 878, + "AGTCCC": 879, + "AGTCCG": 880, + "AGTCGA": 881, + "AGTCGT": 882, + "AGTCGC": 883, + "AGTCGG": 884, + "AGTGAA": 885, + "AGTGAT": 886, + "AGTGAC": 887, + "AGTGAG": 888, + "AGTGTA": 889, + "AGTGTT": 890, + "AGTGTC": 891, + "AGTGTG": 892, + "AGTGCA": 893, + "AGTGCT": 894, + "AGTGCC": 895, + "AGTGCG": 896, + "AGTGGA": 897, + "AGTGGT": 898, + "AGTGGC": 899, + "AGTGGG": 900, + "AGCAAA": 901, + "AGCAAT": 902, + "AGCAAC": 903, + "AGCAAG": 904, + "AGCATA": 905, + "AGCATT": 906, + "AGCATC": 907, + "AGCATG": 908, + "AGCACA": 909, + "AGCACT": 910, + "AGCACC": 911, + "AGCACG": 912, + "AGCAGA": 913, + "AGCAGT": 914, + "AGCAGC": 915, + "AGCAGG": 916, + "AGCTAA": 917, + "AGCTAT": 918, + "AGCTAC": 919, + "AGCTAG": 920, + "AGCTTA": 921, + "AGCTTT": 922, + "AGCTTC": 923, + "AGCTTG": 924, + "AGCTCA": 925, + "AGCTCT": 926, + "AGCTCC": 927, + "AGCTCG": 928, + "AGCTGA": 929, + "AGCTGT": 930, + "AGCTGC": 931, + "AGCTGG": 932, + "AGCCAA": 933, + "AGCCAT": 934, + "AGCCAC": 935, + "AGCCAG": 936, + "AGCCTA": 937, + "AGCCTT": 938, + "AGCCTC": 939, + "AGCCTG": 940, + "AGCCCA": 941, + "AGCCCT": 942, + "AGCCCC": 943, + "AGCCCG": 944, + "AGCCGA": 945, + "AGCCGT": 946, + "AGCCGC": 947, + "AGCCGG": 948, + "AGCGAA": 949, + "AGCGAT": 950, + "AGCGAC": 951, + "AGCGAG": 952, + "AGCGTA": 953, + "AGCGTT": 954, + "AGCGTC": 955, + "AGCGTG": 956, + "AGCGCA": 957, + "AGCGCT": 958, + "AGCGCC": 959, + "AGCGCG": 960, + "AGCGGA": 961, + "AGCGGT": 962, + "AGCGGC": 963, + "AGCGGG": 964, + "AGGAAA": 965, + "AGGAAT": 966, + "AGGAAC": 967, + "AGGAAG": 968, + "AGGATA": 969, + "AGGATT": 970, + "AGGATC": 971, + "AGGATG": 972, + "AGGACA": 973, + "AGGACT": 974, + "AGGACC": 975, + "AGGACG": 976, + "AGGAGA": 977, + "AGGAGT": 978, + "AGGAGC": 979, + "AGGAGG": 980, + "AGGTAA": 981, + "AGGTAT": 982, + "AGGTAC": 983, + "AGGTAG": 984, + "AGGTTA": 985, + "AGGTTT": 986, + "AGGTTC": 987, + "AGGTTG": 988, + "AGGTCA": 989, + "AGGTCT": 990, + "AGGTCC": 991, + "AGGTCG": 992, + "AGGTGA": 993, + "AGGTGT": 994, + "AGGTGC": 995, + "AGGTGG": 996, + "AGGCAA": 997, + "AGGCAT": 998, + "AGGCAC": 999, + "AGGCAG": 1000, + "AGGCTA": 1001, + "AGGCTT": 1002, + "AGGCTC": 1003, + "AGGCTG": 1004, + "AGGCCA": 1005, + "AGGCCT": 1006, + "AGGCCC": 1007, + "AGGCCG": 1008, + "AGGCGA": 1009, + "AGGCGT": 1010, + "AGGCGC": 1011, + "AGGCGG": 1012, + "AGGGAA": 1013, + "AGGGAT": 1014, + "AGGGAC": 1015, + "AGGGAG": 1016, + "AGGGTA": 1017, + "AGGGTT": 1018, + "AGGGTC": 1019, + "AGGGTG": 1020, + "AGGGCA": 1021, + "AGGGCT": 1022, + "AGGGCC": 1023, + "AGGGCG": 1024, + "AGGGGA": 1025, + "AGGGGT": 1026, + "AGGGGC": 1027, + "AGGGGG": 1028, + "TAAAAA": 1029, + "TAAAAT": 1030, + "TAAAAC": 1031, + "TAAAAG": 1032, + "TAAATA": 1033, + "TAAATT": 1034, + "TAAATC": 1035, + "TAAATG": 1036, + "TAAACA": 1037, + "TAAACT": 1038, + "TAAACC": 1039, + "TAAACG": 1040, + "TAAAGA": 1041, + "TAAAGT": 1042, + "TAAAGC": 1043, + "TAAAGG": 1044, + "TAATAA": 1045, + "TAATAT": 1046, + "TAATAC": 1047, + "TAATAG": 1048, + "TAATTA": 1049, + "TAATTT": 1050, + "TAATTC": 1051, + "TAATTG": 1052, + "TAATCA": 1053, + "TAATCT": 1054, + "TAATCC": 1055, + "TAATCG": 1056, + "TAATGA": 1057, + "TAATGT": 1058, + "TAATGC": 1059, + "TAATGG": 1060, + "TAACAA": 1061, + "TAACAT": 1062, + "TAACAC": 1063, + "TAACAG": 1064, + "TAACTA": 1065, + "TAACTT": 1066, + "TAACTC": 1067, + "TAACTG": 1068, + "TAACCA": 1069, + "TAACCT": 1070, + "TAACCC": 1071, + "TAACCG": 1072, + "TAACGA": 1073, + "TAACGT": 1074, + "TAACGC": 1075, + "TAACGG": 1076, + "TAAGAA": 1077, + "TAAGAT": 1078, + "TAAGAC": 1079, + "TAAGAG": 1080, + "TAAGTA": 1081, + "TAAGTT": 1082, + "TAAGTC": 1083, + "TAAGTG": 1084, + "TAAGCA": 1085, + "TAAGCT": 1086, + "TAAGCC": 1087, + "TAAGCG": 1088, + "TAAGGA": 1089, + "TAAGGT": 1090, + "TAAGGC": 1091, + "TAAGGG": 1092, + "TATAAA": 1093, + "TATAAT": 1094, + "TATAAC": 1095, + "TATAAG": 1096, + "TATATA": 1097, + "TATATT": 1098, + "TATATC": 1099, + "TATATG": 1100, + "TATACA": 1101, + "TATACT": 1102, + "TATACC": 1103, + "TATACG": 1104, + "TATAGA": 1105, + "TATAGT": 1106, + "TATAGC": 1107, + "TATAGG": 1108, + "TATTAA": 1109, + "TATTAT": 1110, + "TATTAC": 1111, + "TATTAG": 1112, + "TATTTA": 1113, + "TATTTT": 1114, + "TATTTC": 1115, + "TATTTG": 1116, + "TATTCA": 1117, + "TATTCT": 1118, + "TATTCC": 1119, + "TATTCG": 1120, + "TATTGA": 1121, + "TATTGT": 1122, + "TATTGC": 1123, + "TATTGG": 1124, + "TATCAA": 1125, + "TATCAT": 1126, + "TATCAC": 1127, + "TATCAG": 1128, + "TATCTA": 1129, + "TATCTT": 1130, + "TATCTC": 1131, + "TATCTG": 1132, + "TATCCA": 1133, + "TATCCT": 1134, + "TATCCC": 1135, + "TATCCG": 1136, + "TATCGA": 1137, + "TATCGT": 1138, + "TATCGC": 1139, + "TATCGG": 1140, + "TATGAA": 1141, + "TATGAT": 1142, + "TATGAC": 1143, + "TATGAG": 1144, + "TATGTA": 1145, + "TATGTT": 1146, + "TATGTC": 1147, + "TATGTG": 1148, + "TATGCA": 1149, + "TATGCT": 1150, + "TATGCC": 1151, + "TATGCG": 1152, + "TATGGA": 1153, + "TATGGT": 1154, + "TATGGC": 1155, + "TATGGG": 1156, + "TACAAA": 1157, + "TACAAT": 1158, + "TACAAC": 1159, + "TACAAG": 1160, + "TACATA": 1161, + "TACATT": 1162, + "TACATC": 1163, + "TACATG": 1164, + "TACACA": 1165, + "TACACT": 1166, + "TACACC": 1167, + "TACACG": 1168, + "TACAGA": 1169, + "TACAGT": 1170, + "TACAGC": 1171, + "TACAGG": 1172, + "TACTAA": 1173, + "TACTAT": 1174, + "TACTAC": 1175, + "TACTAG": 1176, + "TACTTA": 1177, + "TACTTT": 1178, + "TACTTC": 1179, + "TACTTG": 1180, + "TACTCA": 1181, + "TACTCT": 1182, + "TACTCC": 1183, + "TACTCG": 1184, + "TACTGA": 1185, + "TACTGT": 1186, + "TACTGC": 1187, + "TACTGG": 1188, + "TACCAA": 1189, + "TACCAT": 1190, + "TACCAC": 1191, + "TACCAG": 1192, + "TACCTA": 1193, + "TACCTT": 1194, + "TACCTC": 1195, + "TACCTG": 1196, + "TACCCA": 1197, + "TACCCT": 1198, + "TACCCC": 1199, + "TACCCG": 1200, + "TACCGA": 1201, + "TACCGT": 1202, + "TACCGC": 1203, + "TACCGG": 1204, + "TACGAA": 1205, + "TACGAT": 1206, + "TACGAC": 1207, + "TACGAG": 1208, + "TACGTA": 1209, + "TACGTT": 1210, + "TACGTC": 1211, + "TACGTG": 1212, + "TACGCA": 1213, + "TACGCT": 1214, + "TACGCC": 1215, + "TACGCG": 1216, + "TACGGA": 1217, + "TACGGT": 1218, + "TACGGC": 1219, + "TACGGG": 1220, + "TAGAAA": 1221, + "TAGAAT": 1222, + "TAGAAC": 1223, + "TAGAAG": 1224, + "TAGATA": 1225, + "TAGATT": 1226, + "TAGATC": 1227, + "TAGATG": 1228, + "TAGACA": 1229, + "TAGACT": 1230, + "TAGACC": 1231, + "TAGACG": 1232, + "TAGAGA": 1233, + "TAGAGT": 1234, + "TAGAGC": 1235, + "TAGAGG": 1236, + "TAGTAA": 1237, + "TAGTAT": 1238, + "TAGTAC": 1239, + "TAGTAG": 1240, + "TAGTTA": 1241, + "TAGTTT": 1242, + "TAGTTC": 1243, + "TAGTTG": 1244, + "TAGTCA": 1245, + "TAGTCT": 1246, + "TAGTCC": 1247, + "TAGTCG": 1248, + "TAGTGA": 1249, + "TAGTGT": 1250, + "TAGTGC": 1251, + "TAGTGG": 1252, + "TAGCAA": 1253, + "TAGCAT": 1254, + "TAGCAC": 1255, + "TAGCAG": 1256, + "TAGCTA": 1257, + "TAGCTT": 1258, + "TAGCTC": 1259, + "TAGCTG": 1260, + "TAGCCA": 1261, + "TAGCCT": 1262, + "TAGCCC": 1263, + "TAGCCG": 1264, + "TAGCGA": 1265, + "TAGCGT": 1266, + "TAGCGC": 1267, + "TAGCGG": 1268, + "TAGGAA": 1269, + "TAGGAT": 1270, + "TAGGAC": 1271, + "TAGGAG": 1272, + "TAGGTA": 1273, + "TAGGTT": 1274, + "TAGGTC": 1275, + "TAGGTG": 1276, + "TAGGCA": 1277, + "TAGGCT": 1278, + "TAGGCC": 1279, + "TAGGCG": 1280, + "TAGGGA": 1281, + "TAGGGT": 1282, + "TAGGGC": 1283, + "TAGGGG": 1284, + "TTAAAA": 1285, + "TTAAAT": 1286, + "TTAAAC": 1287, + "TTAAAG": 1288, + "TTAATA": 1289, + "TTAATT": 1290, + "TTAATC": 1291, + "TTAATG": 1292, + "TTAACA": 1293, + "TTAACT": 1294, + "TTAACC": 1295, + "TTAACG": 1296, + "TTAAGA": 1297, + "TTAAGT": 1298, + "TTAAGC": 1299, + "TTAAGG": 1300, + "TTATAA": 1301, + "TTATAT": 1302, + "TTATAC": 1303, + "TTATAG": 1304, + "TTATTA": 1305, + "TTATTT": 1306, + "TTATTC": 1307, + "TTATTG": 1308, + "TTATCA": 1309, + "TTATCT": 1310, + "TTATCC": 1311, + "TTATCG": 1312, + "TTATGA": 1313, + "TTATGT": 1314, + "TTATGC": 1315, + "TTATGG": 1316, + "TTACAA": 1317, + "TTACAT": 1318, + "TTACAC": 1319, + "TTACAG": 1320, + "TTACTA": 1321, + "TTACTT": 1322, + "TTACTC": 1323, + "TTACTG": 1324, + "TTACCA": 1325, + "TTACCT": 1326, + "TTACCC": 1327, + "TTACCG": 1328, + "TTACGA": 1329, + "TTACGT": 1330, + "TTACGC": 1331, + "TTACGG": 1332, + "TTAGAA": 1333, + "TTAGAT": 1334, + "TTAGAC": 1335, + "TTAGAG": 1336, + "TTAGTA": 1337, + "TTAGTT": 1338, + "TTAGTC": 1339, + "TTAGTG": 1340, + "TTAGCA": 1341, + "TTAGCT": 1342, + "TTAGCC": 1343, + "TTAGCG": 1344, + "TTAGGA": 1345, + "TTAGGT": 1346, + "TTAGGC": 1347, + "TTAGGG": 1348, + "TTTAAA": 1349, + "TTTAAT": 1350, + "TTTAAC": 1351, + "TTTAAG": 1352, + "TTTATA": 1353, + "TTTATT": 1354, + "TTTATC": 1355, + "TTTATG": 1356, + "TTTACA": 1357, + "TTTACT": 1358, + "TTTACC": 1359, + "TTTACG": 1360, + "TTTAGA": 1361, + "TTTAGT": 1362, + "TTTAGC": 1363, + "TTTAGG": 1364, + "TTTTAA": 1365, + "TTTTAT": 1366, + "TTTTAC": 1367, + "TTTTAG": 1368, + "TTTTTA": 1369, + "TTTTTT": 1370, + "TTTTTC": 1371, + "TTTTTG": 1372, + "TTTTCA": 1373, + "TTTTCT": 1374, + "TTTTCC": 1375, + "TTTTCG": 1376, + "TTTTGA": 1377, + "TTTTGT": 1378, + "TTTTGC": 1379, + "TTTTGG": 1380, + "TTTCAA": 1381, + "TTTCAT": 1382, + "TTTCAC": 1383, + "TTTCAG": 1384, + "TTTCTA": 1385, + "TTTCTT": 1386, + "TTTCTC": 1387, + "TTTCTG": 1388, + "TTTCCA": 1389, + "TTTCCT": 1390, + "TTTCCC": 1391, + "TTTCCG": 1392, + "TTTCGA": 1393, + "TTTCGT": 1394, + "TTTCGC": 1395, + "TTTCGG": 1396, + "TTTGAA": 1397, + "TTTGAT": 1398, + "TTTGAC": 1399, + "TTTGAG": 1400, + "TTTGTA": 1401, + "TTTGTT": 1402, + "TTTGTC": 1403, + "TTTGTG": 1404, + "TTTGCA": 1405, + "TTTGCT": 1406, + "TTTGCC": 1407, + "TTTGCG": 1408, + "TTTGGA": 1409, + "TTTGGT": 1410, + "TTTGGC": 1411, + "TTTGGG": 1412, + "TTCAAA": 1413, + "TTCAAT": 1414, + "TTCAAC": 1415, + "TTCAAG": 1416, + "TTCATA": 1417, + "TTCATT": 1418, + "TTCATC": 1419, + "TTCATG": 1420, + "TTCACA": 1421, + "TTCACT": 1422, + "TTCACC": 1423, + "TTCACG": 1424, + "TTCAGA": 1425, + "TTCAGT": 1426, + "TTCAGC": 1427, + "TTCAGG": 1428, + "TTCTAA": 1429, + "TTCTAT": 1430, + "TTCTAC": 1431, + "TTCTAG": 1432, + "TTCTTA": 1433, + "TTCTTT": 1434, + "TTCTTC": 1435, + "TTCTTG": 1436, + "TTCTCA": 1437, + "TTCTCT": 1438, + "TTCTCC": 1439, + "TTCTCG": 1440, + "TTCTGA": 1441, + "TTCTGT": 1442, + "TTCTGC": 1443, + "TTCTGG": 1444, + "TTCCAA": 1445, + "TTCCAT": 1446, + "TTCCAC": 1447, + "TTCCAG": 1448, + "TTCCTA": 1449, + "TTCCTT": 1450, + "TTCCTC": 1451, + "TTCCTG": 1452, + "TTCCCA": 1453, + "TTCCCT": 1454, + "TTCCCC": 1455, + "TTCCCG": 1456, + "TTCCGA": 1457, + "TTCCGT": 1458, + "TTCCGC": 1459, + "TTCCGG": 1460, + "TTCGAA": 1461, + "TTCGAT": 1462, + "TTCGAC": 1463, + "TTCGAG": 1464, + "TTCGTA": 1465, + "TTCGTT": 1466, + "TTCGTC": 1467, + "TTCGTG": 1468, + "TTCGCA": 1469, + "TTCGCT": 1470, + "TTCGCC": 1471, + "TTCGCG": 1472, + "TTCGGA": 1473, + "TTCGGT": 1474, + "TTCGGC": 1475, + "TTCGGG": 1476, + "TTGAAA": 1477, + "TTGAAT": 1478, + "TTGAAC": 1479, + "TTGAAG": 1480, + "TTGATA": 1481, + "TTGATT": 1482, + "TTGATC": 1483, + "TTGATG": 1484, + "TTGACA": 1485, + "TTGACT": 1486, + "TTGACC": 1487, + "TTGACG": 1488, + "TTGAGA": 1489, + "TTGAGT": 1490, + "TTGAGC": 1491, + "TTGAGG": 1492, + "TTGTAA": 1493, + "TTGTAT": 1494, + "TTGTAC": 1495, + "TTGTAG": 1496, + "TTGTTA": 1497, + "TTGTTT": 1498, + "TTGTTC": 1499, + "TTGTTG": 1500, + "TTGTCA": 1501, + "TTGTCT": 1502, + "TTGTCC": 1503, + "TTGTCG": 1504, + "TTGTGA": 1505, + "TTGTGT": 1506, + "TTGTGC": 1507, + "TTGTGG": 1508, + "TTGCAA": 1509, + "TTGCAT": 1510, + "TTGCAC": 1511, + "TTGCAG": 1512, + "TTGCTA": 1513, + "TTGCTT": 1514, + "TTGCTC": 1515, + "TTGCTG": 1516, + "TTGCCA": 1517, + "TTGCCT": 1518, + "TTGCCC": 1519, + "TTGCCG": 1520, + "TTGCGA": 1521, + "TTGCGT": 1522, + "TTGCGC": 1523, + "TTGCGG": 1524, + "TTGGAA": 1525, + "TTGGAT": 1526, + "TTGGAC": 1527, + "TTGGAG": 1528, + "TTGGTA": 1529, + "TTGGTT": 1530, + "TTGGTC": 1531, + "TTGGTG": 1532, + "TTGGCA": 1533, + "TTGGCT": 1534, + "TTGGCC": 1535, + "TTGGCG": 1536, + "TTGGGA": 1537, + "TTGGGT": 1538, + "TTGGGC": 1539, + "TTGGGG": 1540, + "TCAAAA": 1541, + "TCAAAT": 1542, + "TCAAAC": 1543, + "TCAAAG": 1544, + "TCAATA": 1545, + "TCAATT": 1546, + "TCAATC": 1547, + "TCAATG": 1548, + "TCAACA": 1549, + "TCAACT": 1550, + "TCAACC": 1551, + "TCAACG": 1552, + "TCAAGA": 1553, + "TCAAGT": 1554, + "TCAAGC": 1555, + "TCAAGG": 1556, + "TCATAA": 1557, + "TCATAT": 1558, + "TCATAC": 1559, + "TCATAG": 1560, + "TCATTA": 1561, + "TCATTT": 1562, + "TCATTC": 1563, + "TCATTG": 1564, + "TCATCA": 1565, + "TCATCT": 1566, + "TCATCC": 1567, + "TCATCG": 1568, + "TCATGA": 1569, + "TCATGT": 1570, + "TCATGC": 1571, + "TCATGG": 1572, + "TCACAA": 1573, + "TCACAT": 1574, + "TCACAC": 1575, + "TCACAG": 1576, + "TCACTA": 1577, + "TCACTT": 1578, + "TCACTC": 1579, + "TCACTG": 1580, + "TCACCA": 1581, + "TCACCT": 1582, + "TCACCC": 1583, + "TCACCG": 1584, + "TCACGA": 1585, + "TCACGT": 1586, + "TCACGC": 1587, + "TCACGG": 1588, + "TCAGAA": 1589, + "TCAGAT": 1590, + "TCAGAC": 1591, + "TCAGAG": 1592, + "TCAGTA": 1593, + "TCAGTT": 1594, + "TCAGTC": 1595, + "TCAGTG": 1596, + "TCAGCA": 1597, + "TCAGCT": 1598, + "TCAGCC": 1599, + "TCAGCG": 1600, + "TCAGGA": 1601, + "TCAGGT": 1602, + "TCAGGC": 1603, + "TCAGGG": 1604, + "TCTAAA": 1605, + "TCTAAT": 1606, + "TCTAAC": 1607, + "TCTAAG": 1608, + "TCTATA": 1609, + "TCTATT": 1610, + "TCTATC": 1611, + "TCTATG": 1612, + "TCTACA": 1613, + "TCTACT": 1614, + "TCTACC": 1615, + "TCTACG": 1616, + "TCTAGA": 1617, + "TCTAGT": 1618, + "TCTAGC": 1619, + "TCTAGG": 1620, + "TCTTAA": 1621, + "TCTTAT": 1622, + "TCTTAC": 1623, + "TCTTAG": 1624, + "TCTTTA": 1625, + "TCTTTT": 1626, + "TCTTTC": 1627, + "TCTTTG": 1628, + "TCTTCA": 1629, + "TCTTCT": 1630, + "TCTTCC": 1631, + "TCTTCG": 1632, + "TCTTGA": 1633, + "TCTTGT": 1634, + "TCTTGC": 1635, + "TCTTGG": 1636, + "TCTCAA": 1637, + "TCTCAT": 1638, + "TCTCAC": 1639, + "TCTCAG": 1640, + "TCTCTA": 1641, + "TCTCTT": 1642, + "TCTCTC": 1643, + "TCTCTG": 1644, + "TCTCCA": 1645, + "TCTCCT": 1646, + "TCTCCC": 1647, + "TCTCCG": 1648, + "TCTCGA": 1649, + "TCTCGT": 1650, + "TCTCGC": 1651, + "TCTCGG": 1652, + "TCTGAA": 1653, + "TCTGAT": 1654, + "TCTGAC": 1655, + "TCTGAG": 1656, + "TCTGTA": 1657, + "TCTGTT": 1658, + "TCTGTC": 1659, + "TCTGTG": 1660, + "TCTGCA": 1661, + "TCTGCT": 1662, + "TCTGCC": 1663, + "TCTGCG": 1664, + "TCTGGA": 1665, + "TCTGGT": 1666, + "TCTGGC": 1667, + "TCTGGG": 1668, + "TCCAAA": 1669, + "TCCAAT": 1670, + "TCCAAC": 1671, + "TCCAAG": 1672, + "TCCATA": 1673, + "TCCATT": 1674, + "TCCATC": 1675, + "TCCATG": 1676, + "TCCACA": 1677, + "TCCACT": 1678, + "TCCACC": 1679, + "TCCACG": 1680, + "TCCAGA": 1681, + "TCCAGT": 1682, + "TCCAGC": 1683, + "TCCAGG": 1684, + "TCCTAA": 1685, + "TCCTAT": 1686, + "TCCTAC": 1687, + "TCCTAG": 1688, + "TCCTTA": 1689, + "TCCTTT": 1690, + "TCCTTC": 1691, + "TCCTTG": 1692, + "TCCTCA": 1693, + "TCCTCT": 1694, + "TCCTCC": 1695, + "TCCTCG": 1696, + "TCCTGA": 1697, + "TCCTGT": 1698, + "TCCTGC": 1699, + "TCCTGG": 1700, + "TCCCAA": 1701, + "TCCCAT": 1702, + "TCCCAC": 1703, + "TCCCAG": 1704, + "TCCCTA": 1705, + "TCCCTT": 1706, + "TCCCTC": 1707, + "TCCCTG": 1708, + "TCCCCA": 1709, + "TCCCCT": 1710, + "TCCCCC": 1711, + "TCCCCG": 1712, + "TCCCGA": 1713, + "TCCCGT": 1714, + "TCCCGC": 1715, + "TCCCGG": 1716, + "TCCGAA": 1717, + "TCCGAT": 1718, + "TCCGAC": 1719, + "TCCGAG": 1720, + "TCCGTA": 1721, + "TCCGTT": 1722, + "TCCGTC": 1723, + "TCCGTG": 1724, + "TCCGCA": 1725, + "TCCGCT": 1726, + "TCCGCC": 1727, + "TCCGCG": 1728, + "TCCGGA": 1729, + "TCCGGT": 1730, + "TCCGGC": 1731, + "TCCGGG": 1732, + "TCGAAA": 1733, + "TCGAAT": 1734, + "TCGAAC": 1735, + "TCGAAG": 1736, + "TCGATA": 1737, + "TCGATT": 1738, + "TCGATC": 1739, + "TCGATG": 1740, + "TCGACA": 1741, + "TCGACT": 1742, + "TCGACC": 1743, + "TCGACG": 1744, + "TCGAGA": 1745, + "TCGAGT": 1746, + "TCGAGC": 1747, + "TCGAGG": 1748, + "TCGTAA": 1749, + "TCGTAT": 1750, + "TCGTAC": 1751, + "TCGTAG": 1752, + "TCGTTA": 1753, + "TCGTTT": 1754, + "TCGTTC": 1755, + "TCGTTG": 1756, + "TCGTCA": 1757, + "TCGTCT": 1758, + "TCGTCC": 1759, + "TCGTCG": 1760, + "TCGTGA": 1761, + "TCGTGT": 1762, + "TCGTGC": 1763, + "TCGTGG": 1764, + "TCGCAA": 1765, + "TCGCAT": 1766, + "TCGCAC": 1767, + "TCGCAG": 1768, + "TCGCTA": 1769, + "TCGCTT": 1770, + "TCGCTC": 1771, + "TCGCTG": 1772, + "TCGCCA": 1773, + "TCGCCT": 1774, + "TCGCCC": 1775, + "TCGCCG": 1776, + "TCGCGA": 1777, + "TCGCGT": 1778, + "TCGCGC": 1779, + "TCGCGG": 1780, + "TCGGAA": 1781, + "TCGGAT": 1782, + "TCGGAC": 1783, + "TCGGAG": 1784, + "TCGGTA": 1785, + "TCGGTT": 1786, + "TCGGTC": 1787, + "TCGGTG": 1788, + "TCGGCA": 1789, + "TCGGCT": 1790, + "TCGGCC": 1791, + "TCGGCG": 1792, + "TCGGGA": 1793, + "TCGGGT": 1794, + "TCGGGC": 1795, + "TCGGGG": 1796, + "TGAAAA": 1797, + "TGAAAT": 1798, + "TGAAAC": 1799, + "TGAAAG": 1800, + "TGAATA": 1801, + "TGAATT": 1802, + "TGAATC": 1803, + "TGAATG": 1804, + "TGAACA": 1805, + "TGAACT": 1806, + "TGAACC": 1807, + "TGAACG": 1808, + "TGAAGA": 1809, + "TGAAGT": 1810, + "TGAAGC": 1811, + "TGAAGG": 1812, + "TGATAA": 1813, + "TGATAT": 1814, + "TGATAC": 1815, + "TGATAG": 1816, + "TGATTA": 1817, + "TGATTT": 1818, + "TGATTC": 1819, + "TGATTG": 1820, + "TGATCA": 1821, + "TGATCT": 1822, + "TGATCC": 1823, + "TGATCG": 1824, + "TGATGA": 1825, + "TGATGT": 1826, + "TGATGC": 1827, + "TGATGG": 1828, + "TGACAA": 1829, + "TGACAT": 1830, + "TGACAC": 1831, + "TGACAG": 1832, + "TGACTA": 1833, + "TGACTT": 1834, + "TGACTC": 1835, + "TGACTG": 1836, + "TGACCA": 1837, + "TGACCT": 1838, + "TGACCC": 1839, + "TGACCG": 1840, + "TGACGA": 1841, + "TGACGT": 1842, + "TGACGC": 1843, + "TGACGG": 1844, + "TGAGAA": 1845, + "TGAGAT": 1846, + "TGAGAC": 1847, + "TGAGAG": 1848, + "TGAGTA": 1849, + "TGAGTT": 1850, + "TGAGTC": 1851, + "TGAGTG": 1852, + "TGAGCA": 1853, + "TGAGCT": 1854, + "TGAGCC": 1855, + "TGAGCG": 1856, + "TGAGGA": 1857, + "TGAGGT": 1858, + "TGAGGC": 1859, + "TGAGGG": 1860, + "TGTAAA": 1861, + "TGTAAT": 1862, + "TGTAAC": 1863, + "TGTAAG": 1864, + "TGTATA": 1865, + "TGTATT": 1866, + "TGTATC": 1867, + "TGTATG": 1868, + "TGTACA": 1869, + "TGTACT": 1870, + "TGTACC": 1871, + "TGTACG": 1872, + "TGTAGA": 1873, + "TGTAGT": 1874, + "TGTAGC": 1875, + "TGTAGG": 1876, + "TGTTAA": 1877, + "TGTTAT": 1878, + "TGTTAC": 1879, + "TGTTAG": 1880, + "TGTTTA": 1881, + "TGTTTT": 1882, + "TGTTTC": 1883, + "TGTTTG": 1884, + "TGTTCA": 1885, + "TGTTCT": 1886, + "TGTTCC": 1887, + "TGTTCG": 1888, + "TGTTGA": 1889, + "TGTTGT": 1890, + "TGTTGC": 1891, + "TGTTGG": 1892, + "TGTCAA": 1893, + "TGTCAT": 1894, + "TGTCAC": 1895, + "TGTCAG": 1896, + "TGTCTA": 1897, + "TGTCTT": 1898, + "TGTCTC": 1899, + "TGTCTG": 1900, + "TGTCCA": 1901, + "TGTCCT": 1902, + "TGTCCC": 1903, + "TGTCCG": 1904, + "TGTCGA": 1905, + "TGTCGT": 1906, + "TGTCGC": 1907, + "TGTCGG": 1908, + "TGTGAA": 1909, + "TGTGAT": 1910, + "TGTGAC": 1911, + "TGTGAG": 1912, + "TGTGTA": 1913, + "TGTGTT": 1914, + "TGTGTC": 1915, + "TGTGTG": 1916, + "TGTGCA": 1917, + "TGTGCT": 1918, + "TGTGCC": 1919, + "TGTGCG": 1920, + "TGTGGA": 1921, + "TGTGGT": 1922, + "TGTGGC": 1923, + "TGTGGG": 1924, + "TGCAAA": 1925, + "TGCAAT": 1926, + "TGCAAC": 1927, + "TGCAAG": 1928, + "TGCATA": 1929, + "TGCATT": 1930, + "TGCATC": 1931, + "TGCATG": 1932, + "TGCACA": 1933, + "TGCACT": 1934, + "TGCACC": 1935, + "TGCACG": 1936, + "TGCAGA": 1937, + "TGCAGT": 1938, + "TGCAGC": 1939, + "TGCAGG": 1940, + "TGCTAA": 1941, + "TGCTAT": 1942, + "TGCTAC": 1943, + "TGCTAG": 1944, + "TGCTTA": 1945, + "TGCTTT": 1946, + "TGCTTC": 1947, + "TGCTTG": 1948, + "TGCTCA": 1949, + "TGCTCT": 1950, + "TGCTCC": 1951, + "TGCTCG": 1952, + "TGCTGA": 1953, + "TGCTGT": 1954, + "TGCTGC": 1955, + "TGCTGG": 1956, + "TGCCAA": 1957, + "TGCCAT": 1958, + "TGCCAC": 1959, + "TGCCAG": 1960, + "TGCCTA": 1961, + "TGCCTT": 1962, + "TGCCTC": 1963, + "TGCCTG": 1964, + "TGCCCA": 1965, + "TGCCCT": 1966, + "TGCCCC": 1967, + "TGCCCG": 1968, + "TGCCGA": 1969, + "TGCCGT": 1970, + "TGCCGC": 1971, + "TGCCGG": 1972, + "TGCGAA": 1973, + "TGCGAT": 1974, + "TGCGAC": 1975, + "TGCGAG": 1976, + "TGCGTA": 1977, + "TGCGTT": 1978, + "TGCGTC": 1979, + "TGCGTG": 1980, + "TGCGCA": 1981, + "TGCGCT": 1982, + "TGCGCC": 1983, + "TGCGCG": 1984, + "TGCGGA": 1985, + "TGCGGT": 1986, + "TGCGGC": 1987, + "TGCGGG": 1988, + "TGGAAA": 1989, + "TGGAAT": 1990, + "TGGAAC": 1991, + "TGGAAG": 1992, + "TGGATA": 1993, + "TGGATT": 1994, + "TGGATC": 1995, + "TGGATG": 1996, + "TGGACA": 1997, + "TGGACT": 1998, + "TGGACC": 1999, + "TGGACG": 2000, + "TGGAGA": 2001, + "TGGAGT": 2002, + "TGGAGC": 2003, + "TGGAGG": 2004, + "TGGTAA": 2005, + "TGGTAT": 2006, + "TGGTAC": 2007, + "TGGTAG": 2008, + "TGGTTA": 2009, + "TGGTTT": 2010, + "TGGTTC": 2011, + "TGGTTG": 2012, + "TGGTCA": 2013, + "TGGTCT": 2014, + "TGGTCC": 2015, + "TGGTCG": 2016, + "TGGTGA": 2017, + "TGGTGT": 2018, + "TGGTGC": 2019, + "TGGTGG": 2020, + "TGGCAA": 2021, + "TGGCAT": 2022, + "TGGCAC": 2023, + "TGGCAG": 2024, + "TGGCTA": 2025, + "TGGCTT": 2026, + "TGGCTC": 2027, + "TGGCTG": 2028, + "TGGCCA": 2029, + "TGGCCT": 2030, + "TGGCCC": 2031, + "TGGCCG": 2032, + "TGGCGA": 2033, + "TGGCGT": 2034, + "TGGCGC": 2035, + "TGGCGG": 2036, + "TGGGAA": 2037, + "TGGGAT": 2038, + "TGGGAC": 2039, + "TGGGAG": 2040, + "TGGGTA": 2041, + "TGGGTT": 2042, + "TGGGTC": 2043, + "TGGGTG": 2044, + "TGGGCA": 2045, + "TGGGCT": 2046, + "TGGGCC": 2047, + "TGGGCG": 2048, + "TGGGGA": 2049, + "TGGGGT": 2050, + "TGGGGC": 2051, + "TGGGGG": 2052, + "CAAAAA": 2053, + "CAAAAT": 2054, + "CAAAAC": 2055, + "CAAAAG": 2056, + "CAAATA": 2057, + "CAAATT": 2058, + "CAAATC": 2059, + "CAAATG": 2060, + "CAAACA": 2061, + "CAAACT": 2062, + "CAAACC": 2063, + "CAAACG": 2064, + "CAAAGA": 2065, + "CAAAGT": 2066, + "CAAAGC": 2067, + "CAAAGG": 2068, + "CAATAA": 2069, + "CAATAT": 2070, + "CAATAC": 2071, + "CAATAG": 2072, + "CAATTA": 2073, + "CAATTT": 2074, + "CAATTC": 2075, + "CAATTG": 2076, + "CAATCA": 2077, + "CAATCT": 2078, + "CAATCC": 2079, + "CAATCG": 2080, + "CAATGA": 2081, + "CAATGT": 2082, + "CAATGC": 2083, + "CAATGG": 2084, + "CAACAA": 2085, + "CAACAT": 2086, + "CAACAC": 2087, + "CAACAG": 2088, + "CAACTA": 2089, + "CAACTT": 2090, + "CAACTC": 2091, + "CAACTG": 2092, + "CAACCA": 2093, + "CAACCT": 2094, + "CAACCC": 2095, + "CAACCG": 2096, + "CAACGA": 2097, + "CAACGT": 2098, + "CAACGC": 2099, + "CAACGG": 2100, + "CAAGAA": 2101, + "CAAGAT": 2102, + "CAAGAC": 2103, + "CAAGAG": 2104, + "CAAGTA": 2105, + "CAAGTT": 2106, + "CAAGTC": 2107, + "CAAGTG": 2108, + "CAAGCA": 2109, + "CAAGCT": 2110, + "CAAGCC": 2111, + "CAAGCG": 2112, + "CAAGGA": 2113, + "CAAGGT": 2114, + "CAAGGC": 2115, + "CAAGGG": 2116, + "CATAAA": 2117, + "CATAAT": 2118, + "CATAAC": 2119, + "CATAAG": 2120, + "CATATA": 2121, + "CATATT": 2122, + "CATATC": 2123, + "CATATG": 2124, + "CATACA": 2125, + "CATACT": 2126, + "CATACC": 2127, + "CATACG": 2128, + "CATAGA": 2129, + "CATAGT": 2130, + "CATAGC": 2131, + "CATAGG": 2132, + "CATTAA": 2133, + "CATTAT": 2134, + "CATTAC": 2135, + "CATTAG": 2136, + "CATTTA": 2137, + "CATTTT": 2138, + "CATTTC": 2139, + "CATTTG": 2140, + "CATTCA": 2141, + "CATTCT": 2142, + "CATTCC": 2143, + "CATTCG": 2144, + "CATTGA": 2145, + "CATTGT": 2146, + "CATTGC": 2147, + "CATTGG": 2148, + "CATCAA": 2149, + "CATCAT": 2150, + "CATCAC": 2151, + "CATCAG": 2152, + "CATCTA": 2153, + "CATCTT": 2154, + "CATCTC": 2155, + "CATCTG": 2156, + "CATCCA": 2157, + "CATCCT": 2158, + "CATCCC": 2159, + "CATCCG": 2160, + "CATCGA": 2161, + "CATCGT": 2162, + "CATCGC": 2163, + "CATCGG": 2164, + "CATGAA": 2165, + "CATGAT": 2166, + "CATGAC": 2167, + "CATGAG": 2168, + "CATGTA": 2169, + "CATGTT": 2170, + "CATGTC": 2171, + "CATGTG": 2172, + "CATGCA": 2173, + "CATGCT": 2174, + "CATGCC": 2175, + "CATGCG": 2176, + "CATGGA": 2177, + "CATGGT": 2178, + "CATGGC": 2179, + "CATGGG": 2180, + "CACAAA": 2181, + "CACAAT": 2182, + "CACAAC": 2183, + "CACAAG": 2184, + "CACATA": 2185, + "CACATT": 2186, + "CACATC": 2187, + "CACATG": 2188, + "CACACA": 2189, + "CACACT": 2190, + "CACACC": 2191, + "CACACG": 2192, + "CACAGA": 2193, + "CACAGT": 2194, + "CACAGC": 2195, + "CACAGG": 2196, + "CACTAA": 2197, + "CACTAT": 2198, + "CACTAC": 2199, + "CACTAG": 2200, + "CACTTA": 2201, + "CACTTT": 2202, + "CACTTC": 2203, + "CACTTG": 2204, + "CACTCA": 2205, + "CACTCT": 2206, + "CACTCC": 2207, + "CACTCG": 2208, + "CACTGA": 2209, + "CACTGT": 2210, + "CACTGC": 2211, + "CACTGG": 2212, + "CACCAA": 2213, + "CACCAT": 2214, + "CACCAC": 2215, + "CACCAG": 2216, + "CACCTA": 2217, + "CACCTT": 2218, + "CACCTC": 2219, + "CACCTG": 2220, + "CACCCA": 2221, + "CACCCT": 2222, + "CACCCC": 2223, + "CACCCG": 2224, + "CACCGA": 2225, + "CACCGT": 2226, + "CACCGC": 2227, + "CACCGG": 2228, + "CACGAA": 2229, + "CACGAT": 2230, + "CACGAC": 2231, + "CACGAG": 2232, + "CACGTA": 2233, + "CACGTT": 2234, + "CACGTC": 2235, + "CACGTG": 2236, + "CACGCA": 2237, + "CACGCT": 2238, + "CACGCC": 2239, + "CACGCG": 2240, + "CACGGA": 2241, + "CACGGT": 2242, + "CACGGC": 2243, + "CACGGG": 2244, + "CAGAAA": 2245, + "CAGAAT": 2246, + "CAGAAC": 2247, + "CAGAAG": 2248, + "CAGATA": 2249, + "CAGATT": 2250, + "CAGATC": 2251, + "CAGATG": 2252, + "CAGACA": 2253, + "CAGACT": 2254, + "CAGACC": 2255, + "CAGACG": 2256, + "CAGAGA": 2257, + "CAGAGT": 2258, + "CAGAGC": 2259, + "CAGAGG": 2260, + "CAGTAA": 2261, + "CAGTAT": 2262, + "CAGTAC": 2263, + "CAGTAG": 2264, + "CAGTTA": 2265, + "CAGTTT": 2266, + "CAGTTC": 2267, + "CAGTTG": 2268, + "CAGTCA": 2269, + "CAGTCT": 2270, + "CAGTCC": 2271, + "CAGTCG": 2272, + "CAGTGA": 2273, + "CAGTGT": 2274, + "CAGTGC": 2275, + "CAGTGG": 2276, + "CAGCAA": 2277, + "CAGCAT": 2278, + "CAGCAC": 2279, + "CAGCAG": 2280, + "CAGCTA": 2281, + "CAGCTT": 2282, + "CAGCTC": 2283, + "CAGCTG": 2284, + "CAGCCA": 2285, + "CAGCCT": 2286, + "CAGCCC": 2287, + "CAGCCG": 2288, + "CAGCGA": 2289, + "CAGCGT": 2290, + "CAGCGC": 2291, + "CAGCGG": 2292, + "CAGGAA": 2293, + "CAGGAT": 2294, + "CAGGAC": 2295, + "CAGGAG": 2296, + "CAGGTA": 2297, + "CAGGTT": 2298, + "CAGGTC": 2299, + "CAGGTG": 2300, + "CAGGCA": 2301, + "CAGGCT": 2302, + "CAGGCC": 2303, + "CAGGCG": 2304, + "CAGGGA": 2305, + "CAGGGT": 2306, + "CAGGGC": 2307, + "CAGGGG": 2308, + "CTAAAA": 2309, + "CTAAAT": 2310, + "CTAAAC": 2311, + "CTAAAG": 2312, + "CTAATA": 2313, + "CTAATT": 2314, + "CTAATC": 2315, + "CTAATG": 2316, + "CTAACA": 2317, + "CTAACT": 2318, + "CTAACC": 2319, + "CTAACG": 2320, + "CTAAGA": 2321, + "CTAAGT": 2322, + "CTAAGC": 2323, + "CTAAGG": 2324, + "CTATAA": 2325, + "CTATAT": 2326, + "CTATAC": 2327, + "CTATAG": 2328, + "CTATTA": 2329, + "CTATTT": 2330, + "CTATTC": 2331, + "CTATTG": 2332, + "CTATCA": 2333, + "CTATCT": 2334, + "CTATCC": 2335, + "CTATCG": 2336, + "CTATGA": 2337, + "CTATGT": 2338, + "CTATGC": 2339, + "CTATGG": 2340, + "CTACAA": 2341, + "CTACAT": 2342, + "CTACAC": 2343, + "CTACAG": 2344, + "CTACTA": 2345, + "CTACTT": 2346, + "CTACTC": 2347, + "CTACTG": 2348, + "CTACCA": 2349, + "CTACCT": 2350, + "CTACCC": 2351, + "CTACCG": 2352, + "CTACGA": 2353, + "CTACGT": 2354, + "CTACGC": 2355, + "CTACGG": 2356, + "CTAGAA": 2357, + "CTAGAT": 2358, + "CTAGAC": 2359, + "CTAGAG": 2360, + "CTAGTA": 2361, + "CTAGTT": 2362, + "CTAGTC": 2363, + "CTAGTG": 2364, + "CTAGCA": 2365, + "CTAGCT": 2366, + "CTAGCC": 2367, + "CTAGCG": 2368, + "CTAGGA": 2369, + "CTAGGT": 2370, + "CTAGGC": 2371, + "CTAGGG": 2372, + "CTTAAA": 2373, + "CTTAAT": 2374, + "CTTAAC": 2375, + "CTTAAG": 2376, + "CTTATA": 2377, + "CTTATT": 2378, + "CTTATC": 2379, + "CTTATG": 2380, + "CTTACA": 2381, + "CTTACT": 2382, + "CTTACC": 2383, + "CTTACG": 2384, + "CTTAGA": 2385, + "CTTAGT": 2386, + "CTTAGC": 2387, + "CTTAGG": 2388, + "CTTTAA": 2389, + "CTTTAT": 2390, + "CTTTAC": 2391, + "CTTTAG": 2392, + "CTTTTA": 2393, + "CTTTTT": 2394, + "CTTTTC": 2395, + "CTTTTG": 2396, + "CTTTCA": 2397, + "CTTTCT": 2398, + "CTTTCC": 2399, + "CTTTCG": 2400, + "CTTTGA": 2401, + "CTTTGT": 2402, + "CTTTGC": 2403, + "CTTTGG": 2404, + "CTTCAA": 2405, + "CTTCAT": 2406, + "CTTCAC": 2407, + "CTTCAG": 2408, + "CTTCTA": 2409, + "CTTCTT": 2410, + "CTTCTC": 2411, + "CTTCTG": 2412, + "CTTCCA": 2413, + "CTTCCT": 2414, + "CTTCCC": 2415, + "CTTCCG": 2416, + "CTTCGA": 2417, + "CTTCGT": 2418, + "CTTCGC": 2419, + "CTTCGG": 2420, + "CTTGAA": 2421, + "CTTGAT": 2422, + "CTTGAC": 2423, + "CTTGAG": 2424, + "CTTGTA": 2425, + "CTTGTT": 2426, + "CTTGTC": 2427, + "CTTGTG": 2428, + "CTTGCA": 2429, + "CTTGCT": 2430, + "CTTGCC": 2431, + "CTTGCG": 2432, + "CTTGGA": 2433, + "CTTGGT": 2434, + "CTTGGC": 2435, + "CTTGGG": 2436, + "CTCAAA": 2437, + "CTCAAT": 2438, + "CTCAAC": 2439, + "CTCAAG": 2440, + "CTCATA": 2441, + "CTCATT": 2442, + "CTCATC": 2443, + "CTCATG": 2444, + "CTCACA": 2445, + "CTCACT": 2446, + "CTCACC": 2447, + "CTCACG": 2448, + "CTCAGA": 2449, + "CTCAGT": 2450, + "CTCAGC": 2451, + "CTCAGG": 2452, + "CTCTAA": 2453, + "CTCTAT": 2454, + "CTCTAC": 2455, + "CTCTAG": 2456, + "CTCTTA": 2457, + "CTCTTT": 2458, + "CTCTTC": 2459, + "CTCTTG": 2460, + "CTCTCA": 2461, + "CTCTCT": 2462, + "CTCTCC": 2463, + "CTCTCG": 2464, + "CTCTGA": 2465, + "CTCTGT": 2466, + "CTCTGC": 2467, + "CTCTGG": 2468, + "CTCCAA": 2469, + "CTCCAT": 2470, + "CTCCAC": 2471, + "CTCCAG": 2472, + "CTCCTA": 2473, + "CTCCTT": 2474, + "CTCCTC": 2475, + "CTCCTG": 2476, + "CTCCCA": 2477, + "CTCCCT": 2478, + "CTCCCC": 2479, + "CTCCCG": 2480, + "CTCCGA": 2481, + "CTCCGT": 2482, + "CTCCGC": 2483, + "CTCCGG": 2484, + "CTCGAA": 2485, + "CTCGAT": 2486, + "CTCGAC": 2487, + "CTCGAG": 2488, + "CTCGTA": 2489, + "CTCGTT": 2490, + "CTCGTC": 2491, + "CTCGTG": 2492, + "CTCGCA": 2493, + "CTCGCT": 2494, + "CTCGCC": 2495, + "CTCGCG": 2496, + "CTCGGA": 2497, + "CTCGGT": 2498, + "CTCGGC": 2499, + "CTCGGG": 2500, + "CTGAAA": 2501, + "CTGAAT": 2502, + "CTGAAC": 2503, + "CTGAAG": 2504, + "CTGATA": 2505, + "CTGATT": 2506, + "CTGATC": 2507, + "CTGATG": 2508, + "CTGACA": 2509, + "CTGACT": 2510, + "CTGACC": 2511, + "CTGACG": 2512, + "CTGAGA": 2513, + "CTGAGT": 2514, + "CTGAGC": 2515, + "CTGAGG": 2516, + "CTGTAA": 2517, + "CTGTAT": 2518, + "CTGTAC": 2519, + "CTGTAG": 2520, + "CTGTTA": 2521, + "CTGTTT": 2522, + "CTGTTC": 2523, + "CTGTTG": 2524, + "CTGTCA": 2525, + "CTGTCT": 2526, + "CTGTCC": 2527, + "CTGTCG": 2528, + "CTGTGA": 2529, + "CTGTGT": 2530, + "CTGTGC": 2531, + "CTGTGG": 2532, + "CTGCAA": 2533, + "CTGCAT": 2534, + "CTGCAC": 2535, + "CTGCAG": 2536, + "CTGCTA": 2537, + "CTGCTT": 2538, + "CTGCTC": 2539, + "CTGCTG": 2540, + "CTGCCA": 2541, + "CTGCCT": 2542, + "CTGCCC": 2543, + "CTGCCG": 2544, + "CTGCGA": 2545, + "CTGCGT": 2546, + "CTGCGC": 2547, + "CTGCGG": 2548, + "CTGGAA": 2549, + "CTGGAT": 2550, + "CTGGAC": 2551, + "CTGGAG": 2552, + "CTGGTA": 2553, + "CTGGTT": 2554, + "CTGGTC": 2555, + "CTGGTG": 2556, + "CTGGCA": 2557, + "CTGGCT": 2558, + "CTGGCC": 2559, + "CTGGCG": 2560, + "CTGGGA": 2561, + "CTGGGT": 2562, + "CTGGGC": 2563, + "CTGGGG": 2564, + "CCAAAA": 2565, + "CCAAAT": 2566, + "CCAAAC": 2567, + "CCAAAG": 2568, + "CCAATA": 2569, + "CCAATT": 2570, + "CCAATC": 2571, + "CCAATG": 2572, + "CCAACA": 2573, + "CCAACT": 2574, + "CCAACC": 2575, + "CCAACG": 2576, + "CCAAGA": 2577, + "CCAAGT": 2578, + "CCAAGC": 2579, + "CCAAGG": 2580, + "CCATAA": 2581, + "CCATAT": 2582, + "CCATAC": 2583, + "CCATAG": 2584, + "CCATTA": 2585, + "CCATTT": 2586, + "CCATTC": 2587, + "CCATTG": 2588, + "CCATCA": 2589, + "CCATCT": 2590, + "CCATCC": 2591, + "CCATCG": 2592, + "CCATGA": 2593, + "CCATGT": 2594, + "CCATGC": 2595, + "CCATGG": 2596, + "CCACAA": 2597, + "CCACAT": 2598, + "CCACAC": 2599, + "CCACAG": 2600, + "CCACTA": 2601, + "CCACTT": 2602, + "CCACTC": 2603, + "CCACTG": 2604, + "CCACCA": 2605, + "CCACCT": 2606, + "CCACCC": 2607, + "CCACCG": 2608, + "CCACGA": 2609, + "CCACGT": 2610, + "CCACGC": 2611, + "CCACGG": 2612, + "CCAGAA": 2613, + "CCAGAT": 2614, + "CCAGAC": 2615, + "CCAGAG": 2616, + "CCAGTA": 2617, + "CCAGTT": 2618, + "CCAGTC": 2619, + "CCAGTG": 2620, + "CCAGCA": 2621, + "CCAGCT": 2622, + "CCAGCC": 2623, + "CCAGCG": 2624, + "CCAGGA": 2625, + "CCAGGT": 2626, + "CCAGGC": 2627, + "CCAGGG": 2628, + "CCTAAA": 2629, + "CCTAAT": 2630, + "CCTAAC": 2631, + "CCTAAG": 2632, + "CCTATA": 2633, + "CCTATT": 2634, + "CCTATC": 2635, + "CCTATG": 2636, + "CCTACA": 2637, + "CCTACT": 2638, + "CCTACC": 2639, + "CCTACG": 2640, + "CCTAGA": 2641, + "CCTAGT": 2642, + "CCTAGC": 2643, + "CCTAGG": 2644, + "CCTTAA": 2645, + "CCTTAT": 2646, + "CCTTAC": 2647, + "CCTTAG": 2648, + "CCTTTA": 2649, + "CCTTTT": 2650, + "CCTTTC": 2651, + "CCTTTG": 2652, + "CCTTCA": 2653, + "CCTTCT": 2654, + "CCTTCC": 2655, + "CCTTCG": 2656, + "CCTTGA": 2657, + "CCTTGT": 2658, + "CCTTGC": 2659, + "CCTTGG": 2660, + "CCTCAA": 2661, + "CCTCAT": 2662, + "CCTCAC": 2663, + "CCTCAG": 2664, + "CCTCTA": 2665, + "CCTCTT": 2666, + "CCTCTC": 2667, + "CCTCTG": 2668, + "CCTCCA": 2669, + "CCTCCT": 2670, + "CCTCCC": 2671, + "CCTCCG": 2672, + "CCTCGA": 2673, + "CCTCGT": 2674, + "CCTCGC": 2675, + "CCTCGG": 2676, + "CCTGAA": 2677, + "CCTGAT": 2678, + "CCTGAC": 2679, + "CCTGAG": 2680, + "CCTGTA": 2681, + "CCTGTT": 2682, + "CCTGTC": 2683, + "CCTGTG": 2684, + "CCTGCA": 2685, + "CCTGCT": 2686, + "CCTGCC": 2687, + "CCTGCG": 2688, + "CCTGGA": 2689, + "CCTGGT": 2690, + "CCTGGC": 2691, + "CCTGGG": 2692, + "CCCAAA": 2693, + "CCCAAT": 2694, + "CCCAAC": 2695, + "CCCAAG": 2696, + "CCCATA": 2697, + "CCCATT": 2698, + "CCCATC": 2699, + "CCCATG": 2700, + "CCCACA": 2701, + "CCCACT": 2702, + "CCCACC": 2703, + "CCCACG": 2704, + "CCCAGA": 2705, + "CCCAGT": 2706, + "CCCAGC": 2707, + "CCCAGG": 2708, + "CCCTAA": 2709, + "CCCTAT": 2710, + "CCCTAC": 2711, + "CCCTAG": 2712, + "CCCTTA": 2713, + "CCCTTT": 2714, + "CCCTTC": 2715, + "CCCTTG": 2716, + "CCCTCA": 2717, + "CCCTCT": 2718, + "CCCTCC": 2719, + "CCCTCG": 2720, + "CCCTGA": 2721, + "CCCTGT": 2722, + "CCCTGC": 2723, + "CCCTGG": 2724, + "CCCCAA": 2725, + "CCCCAT": 2726, + "CCCCAC": 2727, + "CCCCAG": 2728, + "CCCCTA": 2729, + "CCCCTT": 2730, + "CCCCTC": 2731, + "CCCCTG": 2732, + "CCCCCA": 2733, + "CCCCCT": 2734, + "CCCCCC": 2735, + "CCCCCG": 2736, + "CCCCGA": 2737, + "CCCCGT": 2738, + "CCCCGC": 2739, + "CCCCGG": 2740, + "CCCGAA": 2741, + "CCCGAT": 2742, + "CCCGAC": 2743, + "CCCGAG": 2744, + "CCCGTA": 2745, + "CCCGTT": 2746, + "CCCGTC": 2747, + "CCCGTG": 2748, + "CCCGCA": 2749, + "CCCGCT": 2750, + "CCCGCC": 2751, + "CCCGCG": 2752, + "CCCGGA": 2753, + "CCCGGT": 2754, + "CCCGGC": 2755, + "CCCGGG": 2756, + "CCGAAA": 2757, + "CCGAAT": 2758, + "CCGAAC": 2759, + "CCGAAG": 2760, + "CCGATA": 2761, + "CCGATT": 2762, + "CCGATC": 2763, + "CCGATG": 2764, + "CCGACA": 2765, + "CCGACT": 2766, + "CCGACC": 2767, + "CCGACG": 2768, + "CCGAGA": 2769, + "CCGAGT": 2770, + "CCGAGC": 2771, + "CCGAGG": 2772, + "CCGTAA": 2773, + "CCGTAT": 2774, + "CCGTAC": 2775, + "CCGTAG": 2776, + "CCGTTA": 2777, + "CCGTTT": 2778, + "CCGTTC": 2779, + "CCGTTG": 2780, + "CCGTCA": 2781, + "CCGTCT": 2782, + "CCGTCC": 2783, + "CCGTCG": 2784, + "CCGTGA": 2785, + "CCGTGT": 2786, + "CCGTGC": 2787, + "CCGTGG": 2788, + "CCGCAA": 2789, + "CCGCAT": 2790, + "CCGCAC": 2791, + "CCGCAG": 2792, + "CCGCTA": 2793, + "CCGCTT": 2794, + "CCGCTC": 2795, + "CCGCTG": 2796, + "CCGCCA": 2797, + "CCGCCT": 2798, + "CCGCCC": 2799, + "CCGCCG": 2800, + "CCGCGA": 2801, + "CCGCGT": 2802, + "CCGCGC": 2803, + "CCGCGG": 2804, + "CCGGAA": 2805, + "CCGGAT": 2806, + "CCGGAC": 2807, + "CCGGAG": 2808, + "CCGGTA": 2809, + "CCGGTT": 2810, + "CCGGTC": 2811, + "CCGGTG": 2812, + "CCGGCA": 2813, + "CCGGCT": 2814, + "CCGGCC": 2815, + "CCGGCG": 2816, + "CCGGGA": 2817, + "CCGGGT": 2818, + "CCGGGC": 2819, + "CCGGGG": 2820, + "CGAAAA": 2821, + "CGAAAT": 2822, + "CGAAAC": 2823, + "CGAAAG": 2824, + "CGAATA": 2825, + "CGAATT": 2826, + "CGAATC": 2827, + "CGAATG": 2828, + "CGAACA": 2829, + "CGAACT": 2830, + "CGAACC": 2831, + "CGAACG": 2832, + "CGAAGA": 2833, + "CGAAGT": 2834, + "CGAAGC": 2835, + "CGAAGG": 2836, + "CGATAA": 2837, + "CGATAT": 2838, + "CGATAC": 2839, + "CGATAG": 2840, + "CGATTA": 2841, + "CGATTT": 2842, + "CGATTC": 2843, + "CGATTG": 2844, + "CGATCA": 2845, + "CGATCT": 2846, + "CGATCC": 2847, + "CGATCG": 2848, + "CGATGA": 2849, + "CGATGT": 2850, + "CGATGC": 2851, + "CGATGG": 2852, + "CGACAA": 2853, + "CGACAT": 2854, + "CGACAC": 2855, + "CGACAG": 2856, + "CGACTA": 2857, + "CGACTT": 2858, + "CGACTC": 2859, + "CGACTG": 2860, + "CGACCA": 2861, + "CGACCT": 2862, + "CGACCC": 2863, + "CGACCG": 2864, + "CGACGA": 2865, + "CGACGT": 2866, + "CGACGC": 2867, + "CGACGG": 2868, + "CGAGAA": 2869, + "CGAGAT": 2870, + "CGAGAC": 2871, + "CGAGAG": 2872, + "CGAGTA": 2873, + "CGAGTT": 2874, + "CGAGTC": 2875, + "CGAGTG": 2876, + "CGAGCA": 2877, + "CGAGCT": 2878, + "CGAGCC": 2879, + "CGAGCG": 2880, + "CGAGGA": 2881, + "CGAGGT": 2882, + "CGAGGC": 2883, + "CGAGGG": 2884, + "CGTAAA": 2885, + "CGTAAT": 2886, + "CGTAAC": 2887, + "CGTAAG": 2888, + "CGTATA": 2889, + "CGTATT": 2890, + "CGTATC": 2891, + "CGTATG": 2892, + "CGTACA": 2893, + "CGTACT": 2894, + "CGTACC": 2895, + "CGTACG": 2896, + "CGTAGA": 2897, + "CGTAGT": 2898, + "CGTAGC": 2899, + "CGTAGG": 2900, + "CGTTAA": 2901, + "CGTTAT": 2902, + "CGTTAC": 2903, + "CGTTAG": 2904, + "CGTTTA": 2905, + "CGTTTT": 2906, + "CGTTTC": 2907, + "CGTTTG": 2908, + "CGTTCA": 2909, + "CGTTCT": 2910, + "CGTTCC": 2911, + "CGTTCG": 2912, + "CGTTGA": 2913, + "CGTTGT": 2914, + "CGTTGC": 2915, + "CGTTGG": 2916, + "CGTCAA": 2917, + "CGTCAT": 2918, + "CGTCAC": 2919, + "CGTCAG": 2920, + "CGTCTA": 2921, + "CGTCTT": 2922, + "CGTCTC": 2923, + "CGTCTG": 2924, + "CGTCCA": 2925, + "CGTCCT": 2926, + "CGTCCC": 2927, + "CGTCCG": 2928, + "CGTCGA": 2929, + "CGTCGT": 2930, + "CGTCGC": 2931, + "CGTCGG": 2932, + "CGTGAA": 2933, + "CGTGAT": 2934, + "CGTGAC": 2935, + "CGTGAG": 2936, + "CGTGTA": 2937, + "CGTGTT": 2938, + "CGTGTC": 2939, + "CGTGTG": 2940, + "CGTGCA": 2941, + "CGTGCT": 2942, + "CGTGCC": 2943, + "CGTGCG": 2944, + "CGTGGA": 2945, + "CGTGGT": 2946, + "CGTGGC": 2947, + "CGTGGG": 2948, + "CGCAAA": 2949, + "CGCAAT": 2950, + "CGCAAC": 2951, + "CGCAAG": 2952, + "CGCATA": 2953, + "CGCATT": 2954, + "CGCATC": 2955, + "CGCATG": 2956, + "CGCACA": 2957, + "CGCACT": 2958, + "CGCACC": 2959, + "CGCACG": 2960, + "CGCAGA": 2961, + "CGCAGT": 2962, + "CGCAGC": 2963, + "CGCAGG": 2964, + "CGCTAA": 2965, + "CGCTAT": 2966, + "CGCTAC": 2967, + "CGCTAG": 2968, + "CGCTTA": 2969, + "CGCTTT": 2970, + "CGCTTC": 2971, + "CGCTTG": 2972, + "CGCTCA": 2973, + "CGCTCT": 2974, + "CGCTCC": 2975, + "CGCTCG": 2976, + "CGCTGA": 2977, + "CGCTGT": 2978, + "CGCTGC": 2979, + "CGCTGG": 2980, + "CGCCAA": 2981, + "CGCCAT": 2982, + "CGCCAC": 2983, + "CGCCAG": 2984, + "CGCCTA": 2985, + "CGCCTT": 2986, + "CGCCTC": 2987, + "CGCCTG": 2988, + "CGCCCA": 2989, + "CGCCCT": 2990, + "CGCCCC": 2991, + "CGCCCG": 2992, + "CGCCGA": 2993, + "CGCCGT": 2994, + "CGCCGC": 2995, + "CGCCGG": 2996, + "CGCGAA": 2997, + "CGCGAT": 2998, + "CGCGAC": 2999, + "CGCGAG": 3000, + "CGCGTA": 3001, + "CGCGTT": 3002, + "CGCGTC": 3003, + "CGCGTG": 3004, + "CGCGCA": 3005, + "CGCGCT": 3006, + "CGCGCC": 3007, + "CGCGCG": 3008, + "CGCGGA": 3009, + "CGCGGT": 3010, + "CGCGGC": 3011, + "CGCGGG": 3012, + "CGGAAA": 3013, + "CGGAAT": 3014, + "CGGAAC": 3015, + "CGGAAG": 3016, + "CGGATA": 3017, + "CGGATT": 3018, + "CGGATC": 3019, + "CGGATG": 3020, + "CGGACA": 3021, + "CGGACT": 3022, + "CGGACC": 3023, + "CGGACG": 3024, + "CGGAGA": 3025, + "CGGAGT": 3026, + "CGGAGC": 3027, + "CGGAGG": 3028, + "CGGTAA": 3029, + "CGGTAT": 3030, + "CGGTAC": 3031, + "CGGTAG": 3032, + "CGGTTA": 3033, + "CGGTTT": 3034, + "CGGTTC": 3035, + "CGGTTG": 3036, + "CGGTCA": 3037, + "CGGTCT": 3038, + "CGGTCC": 3039, + "CGGTCG": 3040, + "CGGTGA": 3041, + "CGGTGT": 3042, + "CGGTGC": 3043, + "CGGTGG": 3044, + "CGGCAA": 3045, + "CGGCAT": 3046, + "CGGCAC": 3047, + "CGGCAG": 3048, + "CGGCTA": 3049, + "CGGCTT": 3050, + "CGGCTC": 3051, + "CGGCTG": 3052, + "CGGCCA": 3053, + "CGGCCT": 3054, + "CGGCCC": 3055, + "CGGCCG": 3056, + "CGGCGA": 3057, + "CGGCGT": 3058, + "CGGCGC": 3059, + "CGGCGG": 3060, + "CGGGAA": 3061, + "CGGGAT": 3062, + "CGGGAC": 3063, + "CGGGAG": 3064, + "CGGGTA": 3065, + "CGGGTT": 3066, + "CGGGTC": 3067, + "CGGGTG": 3068, + "CGGGCA": 3069, + "CGGGCT": 3070, + "CGGGCC": 3071, + "CGGGCG": 3072, + "CGGGGA": 3073, + "CGGGGT": 3074, + "CGGGGC": 3075, + "CGGGGG": 3076, + "GAAAAA": 3077, + "GAAAAT": 3078, + "GAAAAC": 3079, + "GAAAAG": 3080, + "GAAATA": 3081, + "GAAATT": 3082, + "GAAATC": 3083, + "GAAATG": 3084, + "GAAACA": 3085, + "GAAACT": 3086, + "GAAACC": 3087, + "GAAACG": 3088, + "GAAAGA": 3089, + "GAAAGT": 3090, + "GAAAGC": 3091, + "GAAAGG": 3092, + "GAATAA": 3093, + "GAATAT": 3094, + "GAATAC": 3095, + "GAATAG": 3096, + "GAATTA": 3097, + "GAATTT": 3098, + "GAATTC": 3099, + "GAATTG": 3100, + "GAATCA": 3101, + "GAATCT": 3102, + "GAATCC": 3103, + "GAATCG": 3104, + "GAATGA": 3105, + "GAATGT": 3106, + "GAATGC": 3107, + "GAATGG": 3108, + "GAACAA": 3109, + "GAACAT": 3110, + "GAACAC": 3111, + "GAACAG": 3112, + "GAACTA": 3113, + "GAACTT": 3114, + "GAACTC": 3115, + "GAACTG": 3116, + "GAACCA": 3117, + "GAACCT": 3118, + "GAACCC": 3119, + "GAACCG": 3120, + "GAACGA": 3121, + "GAACGT": 3122, + "GAACGC": 3123, + "GAACGG": 3124, + "GAAGAA": 3125, + "GAAGAT": 3126, + "GAAGAC": 3127, + "GAAGAG": 3128, + "GAAGTA": 3129, + "GAAGTT": 3130, + "GAAGTC": 3131, + "GAAGTG": 3132, + "GAAGCA": 3133, + "GAAGCT": 3134, + "GAAGCC": 3135, + "GAAGCG": 3136, + "GAAGGA": 3137, + "GAAGGT": 3138, + "GAAGGC": 3139, + "GAAGGG": 3140, + "GATAAA": 3141, + "GATAAT": 3142, + "GATAAC": 3143, + "GATAAG": 3144, + "GATATA": 3145, + "GATATT": 3146, + "GATATC": 3147, + "GATATG": 3148, + "GATACA": 3149, + "GATACT": 3150, + "GATACC": 3151, + "GATACG": 3152, + "GATAGA": 3153, + "GATAGT": 3154, + "GATAGC": 3155, + "GATAGG": 3156, + "GATTAA": 3157, + "GATTAT": 3158, + "GATTAC": 3159, + "GATTAG": 3160, + "GATTTA": 3161, + "GATTTT": 3162, + "GATTTC": 3163, + "GATTTG": 3164, + "GATTCA": 3165, + "GATTCT": 3166, + "GATTCC": 3167, + "GATTCG": 3168, + "GATTGA": 3169, + "GATTGT": 3170, + "GATTGC": 3171, + "GATTGG": 3172, + "GATCAA": 3173, + "GATCAT": 3174, + "GATCAC": 3175, + "GATCAG": 3176, + "GATCTA": 3177, + "GATCTT": 3178, + "GATCTC": 3179, + "GATCTG": 3180, + "GATCCA": 3181, + "GATCCT": 3182, + "GATCCC": 3183, + "GATCCG": 3184, + "GATCGA": 3185, + "GATCGT": 3186, + "GATCGC": 3187, + "GATCGG": 3188, + "GATGAA": 3189, + "GATGAT": 3190, + "GATGAC": 3191, + "GATGAG": 3192, + "GATGTA": 3193, + "GATGTT": 3194, + "GATGTC": 3195, + "GATGTG": 3196, + "GATGCA": 3197, + "GATGCT": 3198, + "GATGCC": 3199, + "GATGCG": 3200, + "GATGGA": 3201, + "GATGGT": 3202, + "GATGGC": 3203, + "GATGGG": 3204, + "GACAAA": 3205, + "GACAAT": 3206, + "GACAAC": 3207, + "GACAAG": 3208, + "GACATA": 3209, + "GACATT": 3210, + "GACATC": 3211, + "GACATG": 3212, + "GACACA": 3213, + "GACACT": 3214, + "GACACC": 3215, + "GACACG": 3216, + "GACAGA": 3217, + "GACAGT": 3218, + "GACAGC": 3219, + "GACAGG": 3220, + "GACTAA": 3221, + "GACTAT": 3222, + "GACTAC": 3223, + "GACTAG": 3224, + "GACTTA": 3225, + "GACTTT": 3226, + "GACTTC": 3227, + "GACTTG": 3228, + "GACTCA": 3229, + "GACTCT": 3230, + "GACTCC": 3231, + "GACTCG": 3232, + "GACTGA": 3233, + "GACTGT": 3234, + "GACTGC": 3235, + "GACTGG": 3236, + "GACCAA": 3237, + "GACCAT": 3238, + "GACCAC": 3239, + "GACCAG": 3240, + "GACCTA": 3241, + "GACCTT": 3242, + "GACCTC": 3243, + "GACCTG": 3244, + "GACCCA": 3245, + "GACCCT": 3246, + "GACCCC": 3247, + "GACCCG": 3248, + "GACCGA": 3249, + "GACCGT": 3250, + "GACCGC": 3251, + "GACCGG": 3252, + "GACGAA": 3253, + "GACGAT": 3254, + "GACGAC": 3255, + "GACGAG": 3256, + "GACGTA": 3257, + "GACGTT": 3258, + "GACGTC": 3259, + "GACGTG": 3260, + "GACGCA": 3261, + "GACGCT": 3262, + "GACGCC": 3263, + "GACGCG": 3264, + "GACGGA": 3265, + "GACGGT": 3266, + "GACGGC": 3267, + "GACGGG": 3268, + "GAGAAA": 3269, + "GAGAAT": 3270, + "GAGAAC": 3271, + "GAGAAG": 3272, + "GAGATA": 3273, + "GAGATT": 3274, + "GAGATC": 3275, + "GAGATG": 3276, + "GAGACA": 3277, + "GAGACT": 3278, + "GAGACC": 3279, + "GAGACG": 3280, + "GAGAGA": 3281, + "GAGAGT": 3282, + "GAGAGC": 3283, + "GAGAGG": 3284, + "GAGTAA": 3285, + "GAGTAT": 3286, + "GAGTAC": 3287, + "GAGTAG": 3288, + "GAGTTA": 3289, + "GAGTTT": 3290, + "GAGTTC": 3291, + "GAGTTG": 3292, + "GAGTCA": 3293, + "GAGTCT": 3294, + "GAGTCC": 3295, + "GAGTCG": 3296, + "GAGTGA": 3297, + "GAGTGT": 3298, + "GAGTGC": 3299, + "GAGTGG": 3300, + "GAGCAA": 3301, + "GAGCAT": 3302, + "GAGCAC": 3303, + "GAGCAG": 3304, + "GAGCTA": 3305, + "GAGCTT": 3306, + "GAGCTC": 3307, + "GAGCTG": 3308, + "GAGCCA": 3309, + "GAGCCT": 3310, + "GAGCCC": 3311, + "GAGCCG": 3312, + "GAGCGA": 3313, + "GAGCGT": 3314, + "GAGCGC": 3315, + "GAGCGG": 3316, + "GAGGAA": 3317, + "GAGGAT": 3318, + "GAGGAC": 3319, + "GAGGAG": 3320, + "GAGGTA": 3321, + "GAGGTT": 3322, + "GAGGTC": 3323, + "GAGGTG": 3324, + "GAGGCA": 3325, + "GAGGCT": 3326, + "GAGGCC": 3327, + "GAGGCG": 3328, + "GAGGGA": 3329, + "GAGGGT": 3330, + "GAGGGC": 3331, + "GAGGGG": 3332, + "GTAAAA": 3333, + "GTAAAT": 3334, + "GTAAAC": 3335, + "GTAAAG": 3336, + "GTAATA": 3337, + "GTAATT": 3338, + "GTAATC": 3339, + "GTAATG": 3340, + "GTAACA": 3341, + "GTAACT": 3342, + "GTAACC": 3343, + "GTAACG": 3344, + "GTAAGA": 3345, + "GTAAGT": 3346, + "GTAAGC": 3347, + "GTAAGG": 3348, + "GTATAA": 3349, + "GTATAT": 3350, + "GTATAC": 3351, + "GTATAG": 3352, + "GTATTA": 3353, + "GTATTT": 3354, + "GTATTC": 3355, + "GTATTG": 3356, + "GTATCA": 3357, + "GTATCT": 3358, + "GTATCC": 3359, + "GTATCG": 3360, + "GTATGA": 3361, + "GTATGT": 3362, + "GTATGC": 3363, + "GTATGG": 3364, + "GTACAA": 3365, + "GTACAT": 3366, + "GTACAC": 3367, + "GTACAG": 3368, + "GTACTA": 3369, + "GTACTT": 3370, + "GTACTC": 3371, + "GTACTG": 3372, + "GTACCA": 3373, + "GTACCT": 3374, + "GTACCC": 3375, + "GTACCG": 3376, + "GTACGA": 3377, + "GTACGT": 3378, + "GTACGC": 3379, + "GTACGG": 3380, + "GTAGAA": 3381, + "GTAGAT": 3382, + "GTAGAC": 3383, + "GTAGAG": 3384, + "GTAGTA": 3385, + "GTAGTT": 3386, + "GTAGTC": 3387, + "GTAGTG": 3388, + "GTAGCA": 3389, + "GTAGCT": 3390, + "GTAGCC": 3391, + "GTAGCG": 3392, + "GTAGGA": 3393, + "GTAGGT": 3394, + "GTAGGC": 3395, + "GTAGGG": 3396, + "GTTAAA": 3397, + "GTTAAT": 3398, + "GTTAAC": 3399, + "GTTAAG": 3400, + "GTTATA": 3401, + "GTTATT": 3402, + "GTTATC": 3403, + "GTTATG": 3404, + "GTTACA": 3405, + "GTTACT": 3406, + "GTTACC": 3407, + "GTTACG": 3408, + "GTTAGA": 3409, + "GTTAGT": 3410, + "GTTAGC": 3411, + "GTTAGG": 3412, + "GTTTAA": 3413, + "GTTTAT": 3414, + "GTTTAC": 3415, + "GTTTAG": 3416, + "GTTTTA": 3417, + "GTTTTT": 3418, + "GTTTTC": 3419, + "GTTTTG": 3420, + "GTTTCA": 3421, + "GTTTCT": 3422, + "GTTTCC": 3423, + "GTTTCG": 3424, + "GTTTGA": 3425, + "GTTTGT": 3426, + "GTTTGC": 3427, + "GTTTGG": 3428, + "GTTCAA": 3429, + "GTTCAT": 3430, + "GTTCAC": 3431, + "GTTCAG": 3432, + "GTTCTA": 3433, + "GTTCTT": 3434, + "GTTCTC": 3435, + "GTTCTG": 3436, + "GTTCCA": 3437, + "GTTCCT": 3438, + "GTTCCC": 3439, + "GTTCCG": 3440, + "GTTCGA": 3441, + "GTTCGT": 3442, + "GTTCGC": 3443, + "GTTCGG": 3444, + "GTTGAA": 3445, + "GTTGAT": 3446, + "GTTGAC": 3447, + "GTTGAG": 3448, + "GTTGTA": 3449, + "GTTGTT": 3450, + "GTTGTC": 3451, + "GTTGTG": 3452, + "GTTGCA": 3453, + "GTTGCT": 3454, + "GTTGCC": 3455, + "GTTGCG": 3456, + "GTTGGA": 3457, + "GTTGGT": 3458, + "GTTGGC": 3459, + "GTTGGG": 3460, + "GTCAAA": 3461, + "GTCAAT": 3462, + "GTCAAC": 3463, + "GTCAAG": 3464, + "GTCATA": 3465, + "GTCATT": 3466, + "GTCATC": 3467, + "GTCATG": 3468, + "GTCACA": 3469, + "GTCACT": 3470, + "GTCACC": 3471, + "GTCACG": 3472, + "GTCAGA": 3473, + "GTCAGT": 3474, + "GTCAGC": 3475, + "GTCAGG": 3476, + "GTCTAA": 3477, + "GTCTAT": 3478, + "GTCTAC": 3479, + "GTCTAG": 3480, + "GTCTTA": 3481, + "GTCTTT": 3482, + "GTCTTC": 3483, + "GTCTTG": 3484, + "GTCTCA": 3485, + "GTCTCT": 3486, + "GTCTCC": 3487, + "GTCTCG": 3488, + "GTCTGA": 3489, + "GTCTGT": 3490, + "GTCTGC": 3491, + "GTCTGG": 3492, + "GTCCAA": 3493, + "GTCCAT": 3494, + "GTCCAC": 3495, + "GTCCAG": 3496, + "GTCCTA": 3497, + "GTCCTT": 3498, + "GTCCTC": 3499, + "GTCCTG": 3500, + "GTCCCA": 3501, + "GTCCCT": 3502, + "GTCCCC": 3503, + "GTCCCG": 3504, + "GTCCGA": 3505, + "GTCCGT": 3506, + "GTCCGC": 3507, + "GTCCGG": 3508, + "GTCGAA": 3509, + "GTCGAT": 3510, + "GTCGAC": 3511, + "GTCGAG": 3512, + "GTCGTA": 3513, + "GTCGTT": 3514, + "GTCGTC": 3515, + "GTCGTG": 3516, + "GTCGCA": 3517, + "GTCGCT": 3518, + "GTCGCC": 3519, + "GTCGCG": 3520, + "GTCGGA": 3521, + "GTCGGT": 3522, + "GTCGGC": 3523, + "GTCGGG": 3524, + "GTGAAA": 3525, + "GTGAAT": 3526, + "GTGAAC": 3527, + "GTGAAG": 3528, + "GTGATA": 3529, + "GTGATT": 3530, + "GTGATC": 3531, + "GTGATG": 3532, + "GTGACA": 3533, + "GTGACT": 3534, + "GTGACC": 3535, + "GTGACG": 3536, + "GTGAGA": 3537, + "GTGAGT": 3538, + "GTGAGC": 3539, + "GTGAGG": 3540, + "GTGTAA": 3541, + "GTGTAT": 3542, + "GTGTAC": 3543, + "GTGTAG": 3544, + "GTGTTA": 3545, + "GTGTTT": 3546, + "GTGTTC": 3547, + "GTGTTG": 3548, + "GTGTCA": 3549, + "GTGTCT": 3550, + "GTGTCC": 3551, + "GTGTCG": 3552, + "GTGTGA": 3553, + "GTGTGT": 3554, + "GTGTGC": 3555, + "GTGTGG": 3556, + "GTGCAA": 3557, + "GTGCAT": 3558, + "GTGCAC": 3559, + "GTGCAG": 3560, + "GTGCTA": 3561, + "GTGCTT": 3562, + "GTGCTC": 3563, + "GTGCTG": 3564, + "GTGCCA": 3565, + "GTGCCT": 3566, + "GTGCCC": 3567, + "GTGCCG": 3568, + "GTGCGA": 3569, + "GTGCGT": 3570, + "GTGCGC": 3571, + "GTGCGG": 3572, + "GTGGAA": 3573, + "GTGGAT": 3574, + "GTGGAC": 3575, + "GTGGAG": 3576, + "GTGGTA": 3577, + "GTGGTT": 3578, + "GTGGTC": 3579, + "GTGGTG": 3580, + "GTGGCA": 3581, + "GTGGCT": 3582, + "GTGGCC": 3583, + "GTGGCG": 3584, + "GTGGGA": 3585, + "GTGGGT": 3586, + "GTGGGC": 3587, + "GTGGGG": 3588, + "GCAAAA": 3589, + "GCAAAT": 3590, + "GCAAAC": 3591, + "GCAAAG": 3592, + "GCAATA": 3593, + "GCAATT": 3594, + "GCAATC": 3595, + "GCAATG": 3596, + "GCAACA": 3597, + "GCAACT": 3598, + "GCAACC": 3599, + "GCAACG": 3600, + "GCAAGA": 3601, + "GCAAGT": 3602, + "GCAAGC": 3603, + "GCAAGG": 3604, + "GCATAA": 3605, + "GCATAT": 3606, + "GCATAC": 3607, + "GCATAG": 3608, + "GCATTA": 3609, + "GCATTT": 3610, + "GCATTC": 3611, + "GCATTG": 3612, + "GCATCA": 3613, + "GCATCT": 3614, + "GCATCC": 3615, + "GCATCG": 3616, + "GCATGA": 3617, + "GCATGT": 3618, + "GCATGC": 3619, + "GCATGG": 3620, + "GCACAA": 3621, + "GCACAT": 3622, + "GCACAC": 3623, + "GCACAG": 3624, + "GCACTA": 3625, + "GCACTT": 3626, + "GCACTC": 3627, + "GCACTG": 3628, + "GCACCA": 3629, + "GCACCT": 3630, + "GCACCC": 3631, + "GCACCG": 3632, + "GCACGA": 3633, + "GCACGT": 3634, + "GCACGC": 3635, + "GCACGG": 3636, + "GCAGAA": 3637, + "GCAGAT": 3638, + "GCAGAC": 3639, + "GCAGAG": 3640, + "GCAGTA": 3641, + "GCAGTT": 3642, + "GCAGTC": 3643, + "GCAGTG": 3644, + "GCAGCA": 3645, + "GCAGCT": 3646, + "GCAGCC": 3647, + "GCAGCG": 3648, + "GCAGGA": 3649, + "GCAGGT": 3650, + "GCAGGC": 3651, + "GCAGGG": 3652, + "GCTAAA": 3653, + "GCTAAT": 3654, + "GCTAAC": 3655, + "GCTAAG": 3656, + "GCTATA": 3657, + "GCTATT": 3658, + "GCTATC": 3659, + "GCTATG": 3660, + "GCTACA": 3661, + "GCTACT": 3662, + "GCTACC": 3663, + "GCTACG": 3664, + "GCTAGA": 3665, + "GCTAGT": 3666, + "GCTAGC": 3667, + "GCTAGG": 3668, + "GCTTAA": 3669, + "GCTTAT": 3670, + "GCTTAC": 3671, + "GCTTAG": 3672, + "GCTTTA": 3673, + "GCTTTT": 3674, + "GCTTTC": 3675, + "GCTTTG": 3676, + "GCTTCA": 3677, + "GCTTCT": 3678, + "GCTTCC": 3679, + "GCTTCG": 3680, + "GCTTGA": 3681, + "GCTTGT": 3682, + "GCTTGC": 3683, + "GCTTGG": 3684, + "GCTCAA": 3685, + "GCTCAT": 3686, + "GCTCAC": 3687, + "GCTCAG": 3688, + "GCTCTA": 3689, + "GCTCTT": 3690, + "GCTCTC": 3691, + "GCTCTG": 3692, + "GCTCCA": 3693, + "GCTCCT": 3694, + "GCTCCC": 3695, + "GCTCCG": 3696, + "GCTCGA": 3697, + "GCTCGT": 3698, + "GCTCGC": 3699, + "GCTCGG": 3700, + "GCTGAA": 3701, + "GCTGAT": 3702, + "GCTGAC": 3703, + "GCTGAG": 3704, + "GCTGTA": 3705, + "GCTGTT": 3706, + "GCTGTC": 3707, + "GCTGTG": 3708, + "GCTGCA": 3709, + "GCTGCT": 3710, + "GCTGCC": 3711, + "GCTGCG": 3712, + "GCTGGA": 3713, + "GCTGGT": 3714, + "GCTGGC": 3715, + "GCTGGG": 3716, + "GCCAAA": 3717, + "GCCAAT": 3718, + "GCCAAC": 3719, + "GCCAAG": 3720, + "GCCATA": 3721, + "GCCATT": 3722, + "GCCATC": 3723, + "GCCATG": 3724, + "GCCACA": 3725, + "GCCACT": 3726, + "GCCACC": 3727, + "GCCACG": 3728, + "GCCAGA": 3729, + "GCCAGT": 3730, + "GCCAGC": 3731, + "GCCAGG": 3732, + "GCCTAA": 3733, + "GCCTAT": 3734, + "GCCTAC": 3735, + "GCCTAG": 3736, + "GCCTTA": 3737, + "GCCTTT": 3738, + "GCCTTC": 3739, + "GCCTTG": 3740, + "GCCTCA": 3741, + "GCCTCT": 3742, + "GCCTCC": 3743, + "GCCTCG": 3744, + "GCCTGA": 3745, + "GCCTGT": 3746, + "GCCTGC": 3747, + "GCCTGG": 3748, + "GCCCAA": 3749, + "GCCCAT": 3750, + "GCCCAC": 3751, + "GCCCAG": 3752, + "GCCCTA": 3753, + "GCCCTT": 3754, + "GCCCTC": 3755, + "GCCCTG": 3756, + "GCCCCA": 3757, + "GCCCCT": 3758, + "GCCCCC": 3759, + "GCCCCG": 3760, + "GCCCGA": 3761, + "GCCCGT": 3762, + "GCCCGC": 3763, + "GCCCGG": 3764, + "GCCGAA": 3765, + "GCCGAT": 3766, + "GCCGAC": 3767, + "GCCGAG": 3768, + "GCCGTA": 3769, + "GCCGTT": 3770, + "GCCGTC": 3771, + "GCCGTG": 3772, + "GCCGCA": 3773, + "GCCGCT": 3774, + "GCCGCC": 3775, + "GCCGCG": 3776, + "GCCGGA": 3777, + "GCCGGT": 3778, + "GCCGGC": 3779, + "GCCGGG": 3780, + "GCGAAA": 3781, + "GCGAAT": 3782, + "GCGAAC": 3783, + "GCGAAG": 3784, + "GCGATA": 3785, + "GCGATT": 3786, + "GCGATC": 3787, + "GCGATG": 3788, + "GCGACA": 3789, + "GCGACT": 3790, + "GCGACC": 3791, + "GCGACG": 3792, + "GCGAGA": 3793, + "GCGAGT": 3794, + "GCGAGC": 3795, + "GCGAGG": 3796, + "GCGTAA": 3797, + "GCGTAT": 3798, + "GCGTAC": 3799, + "GCGTAG": 3800, + "GCGTTA": 3801, + "GCGTTT": 3802, + "GCGTTC": 3803, + "GCGTTG": 3804, + "GCGTCA": 3805, + "GCGTCT": 3806, + "GCGTCC": 3807, + "GCGTCG": 3808, + "GCGTGA": 3809, + "GCGTGT": 3810, + "GCGTGC": 3811, + "GCGTGG": 3812, + "GCGCAA": 3813, + "GCGCAT": 3814, + "GCGCAC": 3815, + "GCGCAG": 3816, + "GCGCTA": 3817, + "GCGCTT": 3818, + "GCGCTC": 3819, + "GCGCTG": 3820, + "GCGCCA": 3821, + "GCGCCT": 3822, + "GCGCCC": 3823, + "GCGCCG": 3824, + "GCGCGA": 3825, + "GCGCGT": 3826, + "GCGCGC": 3827, + "GCGCGG": 3828, + "GCGGAA": 3829, + "GCGGAT": 3830, + "GCGGAC": 3831, + "GCGGAG": 3832, + "GCGGTA": 3833, + "GCGGTT": 3834, + "GCGGTC": 3835, + "GCGGTG": 3836, + "GCGGCA": 3837, + "GCGGCT": 3838, + "GCGGCC": 3839, + "GCGGCG": 3840, + "GCGGGA": 3841, + "GCGGGT": 3842, + "GCGGGC": 3843, + "GCGGGG": 3844, + "GGAAAA": 3845, + "GGAAAT": 3846, + "GGAAAC": 3847, + "GGAAAG": 3848, + "GGAATA": 3849, + "GGAATT": 3850, + "GGAATC": 3851, + "GGAATG": 3852, + "GGAACA": 3853, + "GGAACT": 3854, + "GGAACC": 3855, + "GGAACG": 3856, + "GGAAGA": 3857, + "GGAAGT": 3858, + "GGAAGC": 3859, + "GGAAGG": 3860, + "GGATAA": 3861, + "GGATAT": 3862, + "GGATAC": 3863, + "GGATAG": 3864, + "GGATTA": 3865, + "GGATTT": 3866, + "GGATTC": 3867, + "GGATTG": 3868, + "GGATCA": 3869, + "GGATCT": 3870, + "GGATCC": 3871, + "GGATCG": 3872, + "GGATGA": 3873, + "GGATGT": 3874, + "GGATGC": 3875, + "GGATGG": 3876, + "GGACAA": 3877, + "GGACAT": 3878, + "GGACAC": 3879, + "GGACAG": 3880, + "GGACTA": 3881, + "GGACTT": 3882, + "GGACTC": 3883, + "GGACTG": 3884, + "GGACCA": 3885, + "GGACCT": 3886, + "GGACCC": 3887, + "GGACCG": 3888, + "GGACGA": 3889, + "GGACGT": 3890, + "GGACGC": 3891, + "GGACGG": 3892, + "GGAGAA": 3893, + "GGAGAT": 3894, + "GGAGAC": 3895, + "GGAGAG": 3896, + "GGAGTA": 3897, + "GGAGTT": 3898, + "GGAGTC": 3899, + "GGAGTG": 3900, + "GGAGCA": 3901, + "GGAGCT": 3902, + "GGAGCC": 3903, + "GGAGCG": 3904, + "GGAGGA": 3905, + "GGAGGT": 3906, + "GGAGGC": 3907, + "GGAGGG": 3908, + "GGTAAA": 3909, + "GGTAAT": 3910, + "GGTAAC": 3911, + "GGTAAG": 3912, + "GGTATA": 3913, + "GGTATT": 3914, + "GGTATC": 3915, + "GGTATG": 3916, + "GGTACA": 3917, + "GGTACT": 3918, + "GGTACC": 3919, + "GGTACG": 3920, + "GGTAGA": 3921, + "GGTAGT": 3922, + "GGTAGC": 3923, + "GGTAGG": 3924, + "GGTTAA": 3925, + "GGTTAT": 3926, + "GGTTAC": 3927, + "GGTTAG": 3928, + "GGTTTA": 3929, + "GGTTTT": 3930, + "GGTTTC": 3931, + "GGTTTG": 3932, + "GGTTCA": 3933, + "GGTTCT": 3934, + "GGTTCC": 3935, + "GGTTCG": 3936, + "GGTTGA": 3937, + "GGTTGT": 3938, + "GGTTGC": 3939, + "GGTTGG": 3940, + "GGTCAA": 3941, + "GGTCAT": 3942, + "GGTCAC": 3943, + "GGTCAG": 3944, + "GGTCTA": 3945, + "GGTCTT": 3946, + "GGTCTC": 3947, + "GGTCTG": 3948, + "GGTCCA": 3949, + "GGTCCT": 3950, + "GGTCCC": 3951, + "GGTCCG": 3952, + "GGTCGA": 3953, + "GGTCGT": 3954, + "GGTCGC": 3955, + "GGTCGG": 3956, + "GGTGAA": 3957, + "GGTGAT": 3958, + "GGTGAC": 3959, + "GGTGAG": 3960, + "GGTGTA": 3961, + "GGTGTT": 3962, + "GGTGTC": 3963, + "GGTGTG": 3964, + "GGTGCA": 3965, + "GGTGCT": 3966, + "GGTGCC": 3967, + "GGTGCG": 3968, + "GGTGGA": 3969, + "GGTGGT": 3970, + "GGTGGC": 3971, + "GGTGGG": 3972, + "GGCAAA": 3973, + "GGCAAT": 3974, + "GGCAAC": 3975, + "GGCAAG": 3976, + "GGCATA": 3977, + "GGCATT": 3978, + "GGCATC": 3979, + "GGCATG": 3980, + "GGCACA": 3981, + "GGCACT": 3982, + "GGCACC": 3983, + "GGCACG": 3984, + "GGCAGA": 3985, + "GGCAGT": 3986, + "GGCAGC": 3987, + "GGCAGG": 3988, + "GGCTAA": 3989, + "GGCTAT": 3990, + "GGCTAC": 3991, + "GGCTAG": 3992, + "GGCTTA": 3993, + "GGCTTT": 3994, + "GGCTTC": 3995, + "GGCTTG": 3996, + "GGCTCA": 3997, + "GGCTCT": 3998, + "GGCTCC": 3999, + "GGCTCG": 4000, + "GGCTGA": 4001, + "GGCTGT": 4002, + "GGCTGC": 4003, + "GGCTGG": 4004, + "GGCCAA": 4005, + "GGCCAT": 4006, + "GGCCAC": 4007, + "GGCCAG": 4008, + "GGCCTA": 4009, + "GGCCTT": 4010, + "GGCCTC": 4011, + "GGCCTG": 4012, + "GGCCCA": 4013, + "GGCCCT": 4014, + "GGCCCC": 4015, + "GGCCCG": 4016, + "GGCCGA": 4017, + "GGCCGT": 4018, + "GGCCGC": 4019, + "GGCCGG": 4020, + "GGCGAA": 4021, + "GGCGAT": 4022, + "GGCGAC": 4023, + "GGCGAG": 4024, + "GGCGTA": 4025, + "GGCGTT": 4026, + "GGCGTC": 4027, + "GGCGTG": 4028, + "GGCGCA": 4029, + "GGCGCT": 4030, + "GGCGCC": 4031, + "GGCGCG": 4032, + "GGCGGA": 4033, + "GGCGGT": 4034, + "GGCGGC": 4035, + "GGCGGG": 4036, + "GGGAAA": 4037, + "GGGAAT": 4038, + "GGGAAC": 4039, + "GGGAAG": 4040, + "GGGATA": 4041, + "GGGATT": 4042, + "GGGATC": 4043, + "GGGATG": 4044, + "GGGACA": 4045, + "GGGACT": 4046, + "GGGACC": 4047, + "GGGACG": 4048, + "GGGAGA": 4049, + "GGGAGT": 4050, + "GGGAGC": 4051, + "GGGAGG": 4052, + "GGGTAA": 4053, + "GGGTAT": 4054, + "GGGTAC": 4055, + "GGGTAG": 4056, + "GGGTTA": 4057, + "GGGTTT": 4058, + "GGGTTC": 4059, + "GGGTTG": 4060, + "GGGTCA": 4061, + "GGGTCT": 4062, + "GGGTCC": 4063, + "GGGTCG": 4064, + "GGGTGA": 4065, + "GGGTGT": 4066, + "GGGTGC": 4067, + "GGGTGG": 4068, + "GGGCAA": 4069, + "GGGCAT": 4070, + "GGGCAC": 4071, + "GGGCAG": 4072, + "GGGCTA": 4073, + "GGGCTT": 4074, + "GGGCTC": 4075, + "GGGCTG": 4076, + "GGGCCA": 4077, + "GGGCCT": 4078, + "GGGCCC": 4079, + "GGGCCG": 4080, + "GGGCGA": 4081, + "GGGCGT": 4082, + "GGGCGC": 4083, + "GGGCGG": 4084, + "GGGGAA": 4085, + "GGGGAT": 4086, + "GGGGAC": 4087, + "GGGGAG": 4088, + "GGGGTA": 4089, + "GGGGTT": 4090, + "GGGGTC": 4091, + "GGGGTG": 4092, + "GGGGCA": 4093, + "GGGGCT": 4094, + "GGGGCC": 4095, + "GGGGCG": 4096, + "GGGGGA": 4097, + "GGGGGT": 4098, + "GGGGGC": 4099, + "GGGGGG": 4100 + } + } +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-2000/tokenizer_config.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-2000/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..604e2b1b29af1dcb3920d2828b6a90276e5e3d02 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-2000/tokenizer_config.json @@ -0,0 +1,58 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "[UNK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "[CLS]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "[SEP]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "4": { + "content": "[MASK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "clean_up_tokenization_spaces": true, + "cls_token": "[CLS]", + "do_basic_tokenize": true, + "do_lower_case": false, + "mask_token": "[MASK]", + "max_len": 512, + "model_max_length": 512, + "never_split": null, + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "strip_accents": null, + "tokenize_chinese_chars": true, + "tokenizer_class": "BertTokenizer", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-2000/trainer_state.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-2000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..462cf97866825498e8a91934e1b8d00aeffe8339 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-2000/trainer_state.json @@ -0,0 +1,593 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 8.0, + "eval_steps": 500, + "global_step": 2000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.1, + "grad_norm": 1.1046091318130493, + "learning_rate": 5.319148936170213e-05, + "loss": 0.6562, + "step": 25 + }, + { + "epoch": 0.2, + "grad_norm": 0.4066573977470398, + "learning_rate": 0.00010638297872340425, + "loss": 0.5373, + "step": 50 + }, + { + "epoch": 0.3, + "grad_norm": 0.7340325713157654, + "learning_rate": 0.00015957446808510637, + "loss": 0.4995, + "step": 75 + }, + { + "epoch": 0.4, + "grad_norm": 2.134040117263794, + "learning_rate": 0.0002127659574468085, + "loss": 0.4871, + "step": 100 + }, + { + "epoch": 0.5, + "grad_norm": 2.911975860595703, + "learning_rate": 0.00026595744680851064, + "loss": 0.4507, + "step": 125 + }, + { + "epoch": 0.6, + "grad_norm": 2.040365219116211, + "learning_rate": 0.00031914893617021275, + "loss": 0.3944, + "step": 150 + }, + { + "epoch": 0.7, + "grad_norm": 5.643661022186279, + "learning_rate": 0.0003723404255319149, + "loss": 0.3863, + "step": 175 + }, + { + "epoch": 0.8, + "grad_norm": 3.882492780685425, + "learning_rate": 0.0003992081821181128, + "loss": 0.3875, + "step": 200 + }, + { + "epoch": 0.9, + "grad_norm": 5.87437105178833, + "learning_rate": 0.0003975585615308479, + "loss": 0.3856, + "step": 225 + }, + { + "epoch": 1.0, + "grad_norm": 4.611206531524658, + "learning_rate": 0.00039590894094358297, + "loss": 0.3866, + "step": 250 + }, + { + "epoch": 1.1, + "grad_norm": 3.6404929161071777, + "learning_rate": 0.0003942593203563181, + "loss": 0.3437, + "step": 275 + }, + { + "epoch": 1.2, + "grad_norm": 6.469444751739502, + "learning_rate": 0.0003926096997690532, + "loss": 0.3192, + "step": 300 + }, + { + "epoch": 1.3, + "grad_norm": 3.6251447200775146, + "learning_rate": 0.00039096007918178817, + "loss": 0.349, + "step": 325 + }, + { + "epoch": 1.4, + "grad_norm": 2.854794979095459, + "learning_rate": 0.0003893104585945233, + "loss": 0.3274, + "step": 350 + }, + { + "epoch": 1.5, + "grad_norm": 4.805097579956055, + "learning_rate": 0.0003876608380072583, + "loss": 0.3279, + "step": 375 + }, + { + "epoch": 1.6, + "grad_norm": 2.9891581535339355, + "learning_rate": 0.00038601121741999343, + "loss": 0.3166, + "step": 400 + }, + { + "epoch": 1.7, + "grad_norm": 5.333752155303955, + "learning_rate": 0.0003843615968327285, + "loss": 0.2856, + "step": 425 + }, + { + "epoch": 1.8, + "grad_norm": 2.615621566772461, + "learning_rate": 0.0003827119762454636, + "loss": 0.3286, + "step": 450 + }, + { + "epoch": 1.9, + "grad_norm": 2.4228415489196777, + "learning_rate": 0.00038106235565819863, + "loss": 0.2812, + "step": 475 + }, + { + "epoch": 2.0, + "grad_norm": 3.6015517711639404, + "learning_rate": 0.0003794127350709337, + "loss": 0.2884, + "step": 500 + }, + { + "epoch": 2.1, + "grad_norm": 2.7867860794067383, + "learning_rate": 0.0003777631144836688, + "loss": 0.2582, + "step": 525 + }, + { + "epoch": 2.2, + "grad_norm": 10.079645156860352, + "learning_rate": 0.00037611349389640383, + "loss": 0.245, + "step": 550 + }, + { + "epoch": 2.3, + "grad_norm": 2.475917100906372, + "learning_rate": 0.00037446387330913894, + "loss": 0.2615, + "step": 575 + }, + { + "epoch": 2.4, + "grad_norm": 8.098592758178711, + "learning_rate": 0.000372814252721874, + "loss": 0.2484, + "step": 600 + }, + { + "epoch": 2.5, + "grad_norm": 2.318723201751709, + "learning_rate": 0.00037116463213460903, + "loss": 0.2056, + "step": 625 + }, + { + "epoch": 2.6, + "grad_norm": 3.616283893585205, + "learning_rate": 0.00036951501154734414, + "loss": 0.2358, + "step": 650 + }, + { + "epoch": 2.7, + "grad_norm": 6.419492721557617, + "learning_rate": 0.0003678653909600792, + "loss": 0.2204, + "step": 675 + }, + { + "epoch": 2.8, + "grad_norm": 2.9333388805389404, + "learning_rate": 0.0003662157703728143, + "loss": 0.1857, + "step": 700 + }, + { + "epoch": 2.9, + "grad_norm": 7.691112995147705, + "learning_rate": 0.00036456614978554934, + "loss": 0.2054, + "step": 725 + }, + { + "epoch": 3.0, + "grad_norm": 4.908492565155029, + "learning_rate": 0.00036291652919828444, + "loss": 0.1956, + "step": 750 + }, + { + "epoch": 3.1, + "grad_norm": 1.8923287391662598, + "learning_rate": 0.00036126690861101944, + "loss": 0.1708, + "step": 775 + }, + { + "epoch": 3.2, + "grad_norm": 4.403504848480225, + "learning_rate": 0.00035961728802375454, + "loss": 0.1441, + "step": 800 + }, + { + "epoch": 3.3, + "grad_norm": 4.117386817932129, + "learning_rate": 0.0003579676674364896, + "loss": 0.1586, + "step": 825 + }, + { + "epoch": 3.4, + "grad_norm": 7.119816303253174, + "learning_rate": 0.0003563180468492247, + "loss": 0.1472, + "step": 850 + }, + { + "epoch": 3.5, + "grad_norm": 3.5796430110931396, + "learning_rate": 0.0003546684262619598, + "loss": 0.1802, + "step": 875 + }, + { + "epoch": 3.6, + "grad_norm": 2.97688889503479, + "learning_rate": 0.00035301880567469485, + "loss": 0.1578, + "step": 900 + }, + { + "epoch": 3.7, + "grad_norm": 3.716148614883423, + "learning_rate": 0.0003513691850874299, + "loss": 0.1662, + "step": 925 + }, + { + "epoch": 3.8, + "grad_norm": 6.1249566078186035, + "learning_rate": 0.00034971956450016495, + "loss": 0.159, + "step": 950 + }, + { + "epoch": 3.9, + "grad_norm": 3.0592427253723145, + "learning_rate": 0.00034806994391290005, + "loss": 0.148, + "step": 975 + }, + { + "epoch": 4.0, + "grad_norm": 4.467265605926514, + "learning_rate": 0.0003464203233256351, + "loss": 0.1442, + "step": 1000 + }, + { + "epoch": 4.1, + "grad_norm": 2.7223546504974365, + "learning_rate": 0.0003447707027383702, + "loss": 0.1339, + "step": 1025 + }, + { + "epoch": 4.2, + "grad_norm": 3.698854923248291, + "learning_rate": 0.00034312108215110525, + "loss": 0.1381, + "step": 1050 + }, + { + "epoch": 4.3, + "grad_norm": 7.418384552001953, + "learning_rate": 0.0003414714615638403, + "loss": 0.1158, + "step": 1075 + }, + { + "epoch": 4.4, + "grad_norm": 1.2887814044952393, + "learning_rate": 0.0003398218409765754, + "loss": 0.0855, + "step": 1100 + }, + { + "epoch": 4.5, + "grad_norm": 0.838731586933136, + "learning_rate": 0.00033817222038931045, + "loss": 0.1272, + "step": 1125 + }, + { + "epoch": 4.6, + "grad_norm": 5.912592887878418, + "learning_rate": 0.00033652259980204556, + "loss": 0.1156, + "step": 1150 + }, + { + "epoch": 4.7, + "grad_norm": 8.49521255493164, + "learning_rate": 0.0003348729792147806, + "loss": 0.1182, + "step": 1175 + }, + { + "epoch": 4.8, + "grad_norm": 10.278315544128418, + "learning_rate": 0.0003332233586275157, + "loss": 0.1071, + "step": 1200 + }, + { + "epoch": 4.9, + "grad_norm": 0.724703311920166, + "learning_rate": 0.00033157373804025076, + "loss": 0.11, + "step": 1225 + }, + { + "epoch": 5.0, + "grad_norm": 1.6808199882507324, + "learning_rate": 0.0003299241174529858, + "loss": 0.1101, + "step": 1250 + }, + { + "epoch": 5.1, + "grad_norm": 0.1038585901260376, + "learning_rate": 0.0003282744968657209, + "loss": 0.0791, + "step": 1275 + }, + { + "epoch": 5.2, + "grad_norm": 0.15866787731647491, + "learning_rate": 0.00032662487627845596, + "loss": 0.0638, + "step": 1300 + }, + { + "epoch": 5.3, + "grad_norm": 6.748287200927734, + "learning_rate": 0.00032497525569119106, + "loss": 0.0961, + "step": 1325 + }, + { + "epoch": 5.4, + "grad_norm": 1.5960336923599243, + "learning_rate": 0.0003233256351039261, + "loss": 0.1057, + "step": 1350 + }, + { + "epoch": 5.5, + "grad_norm": 3.3678526878356934, + "learning_rate": 0.00032167601451666116, + "loss": 0.1092, + "step": 1375 + }, + { + "epoch": 5.6, + "grad_norm": 4.743562698364258, + "learning_rate": 0.00032002639392939627, + "loss": 0.1054, + "step": 1400 + }, + { + "epoch": 5.7, + "grad_norm": 12.860156059265137, + "learning_rate": 0.0003183767733421313, + "loss": 0.1009, + "step": 1425 + }, + { + "epoch": 5.8, + "grad_norm": 0.8820792436599731, + "learning_rate": 0.0003167271527548664, + "loss": 0.1219, + "step": 1450 + }, + { + "epoch": 5.9, + "grad_norm": 2.732572317123413, + "learning_rate": 0.00031507753216760147, + "loss": 0.0682, + "step": 1475 + }, + { + "epoch": 6.0, + "grad_norm": 9.810711860656738, + "learning_rate": 0.00031342791158033657, + "loss": 0.1134, + "step": 1500 + }, + { + "epoch": 6.1, + "grad_norm": 6.699489593505859, + "learning_rate": 0.00031177829099307157, + "loss": 0.0557, + "step": 1525 + }, + { + "epoch": 6.2, + "grad_norm": 14.691123008728027, + "learning_rate": 0.00031012867040580667, + "loss": 0.0599, + "step": 1550 + }, + { + "epoch": 6.3, + "grad_norm": 24.44135093688965, + "learning_rate": 0.0003084790498185417, + "loss": 0.0818, + "step": 1575 + }, + { + "epoch": 6.4, + "grad_norm": 0.033995576202869415, + "learning_rate": 0.0003068294292312768, + "loss": 0.0692, + "step": 1600 + }, + { + "epoch": 6.5, + "grad_norm": 6.722524642944336, + "learning_rate": 0.0003051798086440119, + "loss": 0.0702, + "step": 1625 + }, + { + "epoch": 6.6, + "grad_norm": 2.9519436359405518, + "learning_rate": 0.000303530188056747, + "loss": 0.073, + "step": 1650 + }, + { + "epoch": 6.7, + "grad_norm": 1.5118328332901, + "learning_rate": 0.000301880567469482, + "loss": 0.0926, + "step": 1675 + }, + { + "epoch": 6.8, + "grad_norm": 2.1923201084136963, + "learning_rate": 0.0003002309468822171, + "loss": 0.0821, + "step": 1700 + }, + { + "epoch": 6.9, + "grad_norm": 3.34309458732605, + "learning_rate": 0.0002985813262949522, + "loss": 0.0827, + "step": 1725 + }, + { + "epoch": 7.0, + "grad_norm": 10.289947509765625, + "learning_rate": 0.00029693170570768723, + "loss": 0.1162, + "step": 1750 + }, + { + "epoch": 7.1, + "grad_norm": 0.051936667412519455, + "learning_rate": 0.00029528208512042233, + "loss": 0.0626, + "step": 1775 + }, + { + "epoch": 7.2, + "grad_norm": 1.8913339376449585, + "learning_rate": 0.0002936324645331574, + "loss": 0.0657, + "step": 1800 + }, + { + "epoch": 7.3, + "grad_norm": 15.073695182800293, + "learning_rate": 0.00029198284394589243, + "loss": 0.0539, + "step": 1825 + }, + { + "epoch": 7.4, + "grad_norm": 4.514949798583984, + "learning_rate": 0.00029033322335862753, + "loss": 0.056, + "step": 1850 + }, + { + "epoch": 7.5, + "grad_norm": 0.46094000339508057, + "learning_rate": 0.0002886836027713626, + "loss": 0.0667, + "step": 1875 + }, + { + "epoch": 7.6, + "grad_norm": 4.682873725891113, + "learning_rate": 0.0002870339821840977, + "loss": 0.0673, + "step": 1900 + }, + { + "epoch": 7.7, + "grad_norm": 3.581599235534668, + "learning_rate": 0.00028538436159683273, + "loss": 0.0699, + "step": 1925 + }, + { + "epoch": 7.8, + "grad_norm": 3.2518703937530518, + "learning_rate": 0.00028373474100956784, + "loss": 0.0628, + "step": 1950 + }, + { + "epoch": 7.9, + "grad_norm": 0.24071219563484192, + "learning_rate": 0.0002820851204223029, + "loss": 0.0747, + "step": 1975 + }, + { + "epoch": 8.0, + "grad_norm": 2.8545596599578857, + "learning_rate": 0.00028043549983503794, + "loss": 0.0688, + "step": 2000 + } + ], + "logging_steps": 25, + "max_steps": 6250, + "num_input_tokens_seen": 0, + "num_train_epochs": 25, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 101638963200000.0, + "train_batch_size": 20, + "trial_name": null, + "trial_params": null +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-2000/training_args.bin b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-2000/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..ef3fc22e099df49861bca49d15da2b959cfcb6df --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-2000/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ec1d8ed36c45af65d67a23de6c5216a486932bfb09b720e1a4417b82eaaa215d +size 5304 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-2000/vocab.txt b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-2000/vocab.txt new file mode 100644 index 0000000000000000000000000000000000000000..cc9ad53732e21ea9180f18e1fb67a38cc5702d4c --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-2000/vocab.txt @@ -0,0 +1,4101 @@ +[PAD] +[UNK] +[CLS] +[SEP] +[MASK] +AAAAAA +AAAAAT +AAAAAC +AAAAAG +AAAATA +AAAATT +AAAATC +AAAATG +AAAACA +AAAACT +AAAACC +AAAACG +AAAAGA +AAAAGT +AAAAGC +AAAAGG +AAATAA +AAATAT +AAATAC +AAATAG +AAATTA +AAATTT +AAATTC +AAATTG +AAATCA +AAATCT +AAATCC +AAATCG +AAATGA +AAATGT +AAATGC +AAATGG +AAACAA +AAACAT +AAACAC +AAACAG +AAACTA +AAACTT +AAACTC +AAACTG +AAACCA +AAACCT +AAACCC +AAACCG +AAACGA +AAACGT +AAACGC +AAACGG +AAAGAA +AAAGAT +AAAGAC +AAAGAG +AAAGTA +AAAGTT +AAAGTC +AAAGTG +AAAGCA +AAAGCT +AAAGCC +AAAGCG +AAAGGA +AAAGGT +AAAGGC +AAAGGG +AATAAA +AATAAT +AATAAC +AATAAG +AATATA +AATATT +AATATC +AATATG +AATACA +AATACT +AATACC +AATACG +AATAGA +AATAGT +AATAGC +AATAGG +AATTAA +AATTAT +AATTAC +AATTAG +AATTTA +AATTTT +AATTTC +AATTTG +AATTCA +AATTCT +AATTCC +AATTCG +AATTGA +AATTGT +AATTGC +AATTGG +AATCAA +AATCAT +AATCAC +AATCAG +AATCTA +AATCTT +AATCTC +AATCTG +AATCCA +AATCCT +AATCCC +AATCCG +AATCGA +AATCGT +AATCGC +AATCGG +AATGAA +AATGAT +AATGAC +AATGAG +AATGTA +AATGTT +AATGTC +AATGTG +AATGCA +AATGCT +AATGCC +AATGCG +AATGGA +AATGGT +AATGGC +AATGGG +AACAAA +AACAAT +AACAAC +AACAAG +AACATA +AACATT +AACATC +AACATG +AACACA +AACACT +AACACC +AACACG +AACAGA +AACAGT +AACAGC +AACAGG +AACTAA +AACTAT +AACTAC +AACTAG +AACTTA +AACTTT +AACTTC +AACTTG +AACTCA +AACTCT +AACTCC +AACTCG +AACTGA +AACTGT +AACTGC +AACTGG +AACCAA +AACCAT +AACCAC +AACCAG +AACCTA +AACCTT +AACCTC +AACCTG +AACCCA +AACCCT +AACCCC +AACCCG +AACCGA +AACCGT +AACCGC +AACCGG +AACGAA +AACGAT +AACGAC +AACGAG +AACGTA +AACGTT +AACGTC +AACGTG +AACGCA +AACGCT +AACGCC +AACGCG +AACGGA +AACGGT +AACGGC +AACGGG +AAGAAA +AAGAAT +AAGAAC +AAGAAG +AAGATA +AAGATT +AAGATC +AAGATG +AAGACA +AAGACT +AAGACC +AAGACG +AAGAGA +AAGAGT +AAGAGC +AAGAGG +AAGTAA +AAGTAT +AAGTAC +AAGTAG +AAGTTA +AAGTTT +AAGTTC +AAGTTG +AAGTCA +AAGTCT +AAGTCC +AAGTCG +AAGTGA +AAGTGT +AAGTGC +AAGTGG +AAGCAA +AAGCAT +AAGCAC +AAGCAG +AAGCTA +AAGCTT +AAGCTC +AAGCTG +AAGCCA +AAGCCT +AAGCCC +AAGCCG +AAGCGA +AAGCGT +AAGCGC +AAGCGG +AAGGAA +AAGGAT +AAGGAC +AAGGAG +AAGGTA +AAGGTT +AAGGTC +AAGGTG +AAGGCA +AAGGCT +AAGGCC +AAGGCG +AAGGGA +AAGGGT +AAGGGC +AAGGGG +ATAAAA +ATAAAT +ATAAAC +ATAAAG +ATAATA +ATAATT +ATAATC +ATAATG +ATAACA +ATAACT +ATAACC +ATAACG +ATAAGA +ATAAGT +ATAAGC +ATAAGG +ATATAA +ATATAT +ATATAC +ATATAG +ATATTA +ATATTT +ATATTC +ATATTG +ATATCA +ATATCT +ATATCC +ATATCG +ATATGA +ATATGT +ATATGC +ATATGG +ATACAA +ATACAT +ATACAC +ATACAG +ATACTA +ATACTT +ATACTC +ATACTG +ATACCA +ATACCT +ATACCC +ATACCG +ATACGA +ATACGT +ATACGC +ATACGG +ATAGAA +ATAGAT +ATAGAC +ATAGAG +ATAGTA +ATAGTT +ATAGTC +ATAGTG +ATAGCA +ATAGCT +ATAGCC +ATAGCG +ATAGGA +ATAGGT +ATAGGC +ATAGGG +ATTAAA +ATTAAT +ATTAAC +ATTAAG +ATTATA +ATTATT +ATTATC +ATTATG +ATTACA +ATTACT +ATTACC +ATTACG +ATTAGA +ATTAGT +ATTAGC +ATTAGG +ATTTAA +ATTTAT +ATTTAC +ATTTAG +ATTTTA +ATTTTT +ATTTTC +ATTTTG +ATTTCA +ATTTCT +ATTTCC +ATTTCG +ATTTGA +ATTTGT +ATTTGC +ATTTGG +ATTCAA +ATTCAT +ATTCAC +ATTCAG +ATTCTA +ATTCTT +ATTCTC +ATTCTG +ATTCCA +ATTCCT +ATTCCC +ATTCCG +ATTCGA +ATTCGT +ATTCGC +ATTCGG +ATTGAA +ATTGAT +ATTGAC +ATTGAG +ATTGTA +ATTGTT +ATTGTC +ATTGTG +ATTGCA +ATTGCT +ATTGCC +ATTGCG +ATTGGA +ATTGGT +ATTGGC +ATTGGG +ATCAAA +ATCAAT +ATCAAC +ATCAAG +ATCATA +ATCATT +ATCATC +ATCATG +ATCACA +ATCACT +ATCACC +ATCACG +ATCAGA +ATCAGT +ATCAGC +ATCAGG +ATCTAA +ATCTAT +ATCTAC +ATCTAG +ATCTTA +ATCTTT +ATCTTC +ATCTTG +ATCTCA +ATCTCT +ATCTCC +ATCTCG +ATCTGA +ATCTGT +ATCTGC +ATCTGG +ATCCAA +ATCCAT +ATCCAC +ATCCAG +ATCCTA +ATCCTT +ATCCTC +ATCCTG +ATCCCA +ATCCCT +ATCCCC +ATCCCG +ATCCGA +ATCCGT +ATCCGC +ATCCGG +ATCGAA +ATCGAT +ATCGAC +ATCGAG +ATCGTA +ATCGTT +ATCGTC +ATCGTG +ATCGCA +ATCGCT +ATCGCC +ATCGCG +ATCGGA +ATCGGT +ATCGGC +ATCGGG +ATGAAA +ATGAAT +ATGAAC +ATGAAG +ATGATA +ATGATT +ATGATC +ATGATG +ATGACA +ATGACT +ATGACC +ATGACG +ATGAGA +ATGAGT +ATGAGC +ATGAGG +ATGTAA +ATGTAT +ATGTAC +ATGTAG +ATGTTA +ATGTTT +ATGTTC +ATGTTG +ATGTCA +ATGTCT +ATGTCC +ATGTCG +ATGTGA +ATGTGT +ATGTGC +ATGTGG +ATGCAA +ATGCAT +ATGCAC +ATGCAG +ATGCTA +ATGCTT +ATGCTC +ATGCTG +ATGCCA +ATGCCT +ATGCCC +ATGCCG +ATGCGA +ATGCGT +ATGCGC +ATGCGG +ATGGAA +ATGGAT +ATGGAC +ATGGAG +ATGGTA +ATGGTT +ATGGTC +ATGGTG +ATGGCA +ATGGCT +ATGGCC +ATGGCG +ATGGGA +ATGGGT +ATGGGC +ATGGGG +ACAAAA +ACAAAT +ACAAAC +ACAAAG +ACAATA +ACAATT +ACAATC +ACAATG +ACAACA +ACAACT +ACAACC +ACAACG +ACAAGA +ACAAGT +ACAAGC +ACAAGG +ACATAA +ACATAT +ACATAC +ACATAG +ACATTA +ACATTT +ACATTC +ACATTG +ACATCA +ACATCT +ACATCC +ACATCG +ACATGA +ACATGT +ACATGC +ACATGG +ACACAA +ACACAT +ACACAC +ACACAG +ACACTA +ACACTT +ACACTC +ACACTG +ACACCA +ACACCT +ACACCC +ACACCG +ACACGA +ACACGT +ACACGC +ACACGG +ACAGAA +ACAGAT +ACAGAC +ACAGAG +ACAGTA +ACAGTT +ACAGTC +ACAGTG +ACAGCA +ACAGCT +ACAGCC +ACAGCG +ACAGGA +ACAGGT +ACAGGC +ACAGGG +ACTAAA +ACTAAT +ACTAAC +ACTAAG +ACTATA +ACTATT +ACTATC +ACTATG +ACTACA +ACTACT +ACTACC +ACTACG +ACTAGA +ACTAGT +ACTAGC +ACTAGG +ACTTAA +ACTTAT +ACTTAC +ACTTAG +ACTTTA +ACTTTT +ACTTTC +ACTTTG +ACTTCA +ACTTCT +ACTTCC +ACTTCG +ACTTGA +ACTTGT +ACTTGC +ACTTGG +ACTCAA +ACTCAT +ACTCAC +ACTCAG +ACTCTA +ACTCTT +ACTCTC +ACTCTG +ACTCCA +ACTCCT +ACTCCC +ACTCCG +ACTCGA +ACTCGT +ACTCGC +ACTCGG +ACTGAA +ACTGAT +ACTGAC +ACTGAG +ACTGTA +ACTGTT +ACTGTC +ACTGTG +ACTGCA +ACTGCT +ACTGCC +ACTGCG +ACTGGA +ACTGGT +ACTGGC +ACTGGG +ACCAAA +ACCAAT +ACCAAC +ACCAAG +ACCATA +ACCATT +ACCATC +ACCATG +ACCACA +ACCACT +ACCACC +ACCACG +ACCAGA +ACCAGT +ACCAGC +ACCAGG +ACCTAA +ACCTAT +ACCTAC +ACCTAG +ACCTTA +ACCTTT +ACCTTC +ACCTTG +ACCTCA +ACCTCT +ACCTCC +ACCTCG +ACCTGA +ACCTGT +ACCTGC +ACCTGG +ACCCAA +ACCCAT +ACCCAC +ACCCAG +ACCCTA +ACCCTT +ACCCTC +ACCCTG +ACCCCA +ACCCCT +ACCCCC +ACCCCG +ACCCGA +ACCCGT +ACCCGC +ACCCGG +ACCGAA +ACCGAT +ACCGAC +ACCGAG +ACCGTA +ACCGTT +ACCGTC +ACCGTG +ACCGCA +ACCGCT +ACCGCC +ACCGCG +ACCGGA +ACCGGT +ACCGGC +ACCGGG +ACGAAA +ACGAAT +ACGAAC +ACGAAG +ACGATA +ACGATT +ACGATC +ACGATG +ACGACA +ACGACT +ACGACC +ACGACG +ACGAGA +ACGAGT +ACGAGC +ACGAGG +ACGTAA +ACGTAT +ACGTAC +ACGTAG +ACGTTA +ACGTTT +ACGTTC +ACGTTG +ACGTCA +ACGTCT +ACGTCC +ACGTCG +ACGTGA +ACGTGT +ACGTGC +ACGTGG +ACGCAA +ACGCAT +ACGCAC +ACGCAG +ACGCTA +ACGCTT +ACGCTC +ACGCTG +ACGCCA +ACGCCT +ACGCCC +ACGCCG +ACGCGA +ACGCGT +ACGCGC +ACGCGG +ACGGAA +ACGGAT +ACGGAC +ACGGAG +ACGGTA +ACGGTT +ACGGTC +ACGGTG +ACGGCA +ACGGCT +ACGGCC +ACGGCG +ACGGGA +ACGGGT +ACGGGC +ACGGGG +AGAAAA +AGAAAT +AGAAAC +AGAAAG +AGAATA +AGAATT +AGAATC +AGAATG +AGAACA +AGAACT +AGAACC +AGAACG +AGAAGA +AGAAGT +AGAAGC +AGAAGG +AGATAA +AGATAT +AGATAC +AGATAG +AGATTA +AGATTT +AGATTC +AGATTG +AGATCA +AGATCT +AGATCC +AGATCG +AGATGA +AGATGT +AGATGC +AGATGG +AGACAA +AGACAT +AGACAC +AGACAG +AGACTA +AGACTT +AGACTC +AGACTG +AGACCA +AGACCT +AGACCC +AGACCG +AGACGA +AGACGT +AGACGC +AGACGG +AGAGAA +AGAGAT +AGAGAC +AGAGAG +AGAGTA +AGAGTT +AGAGTC +AGAGTG +AGAGCA +AGAGCT +AGAGCC +AGAGCG +AGAGGA +AGAGGT +AGAGGC +AGAGGG +AGTAAA +AGTAAT +AGTAAC +AGTAAG +AGTATA +AGTATT +AGTATC +AGTATG +AGTACA +AGTACT +AGTACC +AGTACG +AGTAGA +AGTAGT +AGTAGC +AGTAGG +AGTTAA +AGTTAT +AGTTAC +AGTTAG +AGTTTA +AGTTTT +AGTTTC +AGTTTG +AGTTCA +AGTTCT +AGTTCC +AGTTCG +AGTTGA +AGTTGT +AGTTGC +AGTTGG +AGTCAA +AGTCAT +AGTCAC +AGTCAG +AGTCTA +AGTCTT +AGTCTC +AGTCTG +AGTCCA +AGTCCT +AGTCCC +AGTCCG +AGTCGA +AGTCGT +AGTCGC +AGTCGG +AGTGAA +AGTGAT +AGTGAC +AGTGAG +AGTGTA +AGTGTT +AGTGTC +AGTGTG +AGTGCA +AGTGCT +AGTGCC +AGTGCG +AGTGGA +AGTGGT +AGTGGC +AGTGGG +AGCAAA +AGCAAT +AGCAAC +AGCAAG +AGCATA +AGCATT +AGCATC +AGCATG +AGCACA +AGCACT +AGCACC +AGCACG +AGCAGA +AGCAGT +AGCAGC +AGCAGG +AGCTAA +AGCTAT +AGCTAC +AGCTAG +AGCTTA +AGCTTT +AGCTTC +AGCTTG +AGCTCA +AGCTCT +AGCTCC +AGCTCG +AGCTGA +AGCTGT +AGCTGC +AGCTGG +AGCCAA +AGCCAT +AGCCAC +AGCCAG +AGCCTA +AGCCTT +AGCCTC +AGCCTG +AGCCCA +AGCCCT +AGCCCC +AGCCCG +AGCCGA +AGCCGT +AGCCGC +AGCCGG +AGCGAA +AGCGAT +AGCGAC +AGCGAG +AGCGTA +AGCGTT +AGCGTC +AGCGTG +AGCGCA +AGCGCT +AGCGCC +AGCGCG +AGCGGA +AGCGGT +AGCGGC +AGCGGG +AGGAAA +AGGAAT +AGGAAC +AGGAAG +AGGATA +AGGATT +AGGATC +AGGATG +AGGACA +AGGACT +AGGACC +AGGACG +AGGAGA +AGGAGT +AGGAGC +AGGAGG +AGGTAA +AGGTAT +AGGTAC +AGGTAG +AGGTTA +AGGTTT +AGGTTC +AGGTTG +AGGTCA +AGGTCT +AGGTCC +AGGTCG +AGGTGA +AGGTGT +AGGTGC +AGGTGG +AGGCAA +AGGCAT +AGGCAC +AGGCAG +AGGCTA +AGGCTT +AGGCTC +AGGCTG +AGGCCA +AGGCCT +AGGCCC +AGGCCG +AGGCGA +AGGCGT +AGGCGC +AGGCGG +AGGGAA +AGGGAT +AGGGAC +AGGGAG +AGGGTA +AGGGTT +AGGGTC +AGGGTG +AGGGCA +AGGGCT +AGGGCC +AGGGCG +AGGGGA +AGGGGT +AGGGGC +AGGGGG +TAAAAA +TAAAAT +TAAAAC +TAAAAG +TAAATA +TAAATT +TAAATC +TAAATG +TAAACA +TAAACT +TAAACC +TAAACG +TAAAGA +TAAAGT +TAAAGC +TAAAGG +TAATAA +TAATAT +TAATAC +TAATAG +TAATTA +TAATTT +TAATTC +TAATTG +TAATCA +TAATCT +TAATCC +TAATCG +TAATGA +TAATGT +TAATGC +TAATGG +TAACAA +TAACAT +TAACAC +TAACAG +TAACTA +TAACTT +TAACTC +TAACTG +TAACCA +TAACCT +TAACCC +TAACCG +TAACGA +TAACGT +TAACGC +TAACGG +TAAGAA +TAAGAT +TAAGAC +TAAGAG +TAAGTA +TAAGTT +TAAGTC +TAAGTG +TAAGCA +TAAGCT +TAAGCC +TAAGCG +TAAGGA +TAAGGT +TAAGGC +TAAGGG +TATAAA +TATAAT +TATAAC +TATAAG +TATATA +TATATT +TATATC +TATATG +TATACA +TATACT +TATACC +TATACG +TATAGA +TATAGT +TATAGC +TATAGG +TATTAA +TATTAT +TATTAC +TATTAG +TATTTA +TATTTT +TATTTC +TATTTG +TATTCA +TATTCT +TATTCC +TATTCG +TATTGA +TATTGT +TATTGC +TATTGG +TATCAA +TATCAT +TATCAC +TATCAG +TATCTA +TATCTT +TATCTC +TATCTG +TATCCA +TATCCT +TATCCC +TATCCG +TATCGA +TATCGT +TATCGC +TATCGG +TATGAA +TATGAT +TATGAC +TATGAG +TATGTA +TATGTT +TATGTC +TATGTG +TATGCA +TATGCT +TATGCC +TATGCG +TATGGA +TATGGT +TATGGC +TATGGG +TACAAA +TACAAT +TACAAC +TACAAG +TACATA +TACATT +TACATC +TACATG +TACACA +TACACT +TACACC +TACACG +TACAGA +TACAGT +TACAGC +TACAGG +TACTAA +TACTAT +TACTAC +TACTAG +TACTTA +TACTTT +TACTTC +TACTTG +TACTCA +TACTCT +TACTCC +TACTCG +TACTGA +TACTGT +TACTGC +TACTGG +TACCAA +TACCAT +TACCAC +TACCAG +TACCTA +TACCTT +TACCTC +TACCTG +TACCCA +TACCCT +TACCCC +TACCCG +TACCGA +TACCGT +TACCGC +TACCGG +TACGAA +TACGAT +TACGAC +TACGAG +TACGTA +TACGTT +TACGTC +TACGTG +TACGCA +TACGCT +TACGCC +TACGCG +TACGGA +TACGGT +TACGGC +TACGGG +TAGAAA +TAGAAT +TAGAAC +TAGAAG +TAGATA +TAGATT +TAGATC +TAGATG +TAGACA +TAGACT +TAGACC +TAGACG +TAGAGA +TAGAGT +TAGAGC +TAGAGG +TAGTAA +TAGTAT +TAGTAC +TAGTAG +TAGTTA +TAGTTT +TAGTTC +TAGTTG +TAGTCA +TAGTCT +TAGTCC +TAGTCG +TAGTGA +TAGTGT +TAGTGC +TAGTGG +TAGCAA +TAGCAT +TAGCAC +TAGCAG +TAGCTA +TAGCTT +TAGCTC +TAGCTG +TAGCCA +TAGCCT +TAGCCC +TAGCCG +TAGCGA +TAGCGT +TAGCGC +TAGCGG +TAGGAA +TAGGAT +TAGGAC +TAGGAG +TAGGTA +TAGGTT +TAGGTC +TAGGTG +TAGGCA +TAGGCT +TAGGCC +TAGGCG +TAGGGA +TAGGGT +TAGGGC +TAGGGG +TTAAAA +TTAAAT +TTAAAC +TTAAAG +TTAATA +TTAATT +TTAATC +TTAATG +TTAACA +TTAACT +TTAACC +TTAACG +TTAAGA +TTAAGT +TTAAGC +TTAAGG +TTATAA +TTATAT +TTATAC +TTATAG +TTATTA +TTATTT +TTATTC +TTATTG +TTATCA +TTATCT +TTATCC +TTATCG +TTATGA +TTATGT +TTATGC +TTATGG +TTACAA +TTACAT +TTACAC +TTACAG +TTACTA +TTACTT +TTACTC +TTACTG +TTACCA +TTACCT +TTACCC +TTACCG +TTACGA +TTACGT +TTACGC +TTACGG +TTAGAA +TTAGAT +TTAGAC +TTAGAG +TTAGTA +TTAGTT +TTAGTC +TTAGTG +TTAGCA +TTAGCT +TTAGCC +TTAGCG +TTAGGA +TTAGGT +TTAGGC +TTAGGG +TTTAAA +TTTAAT +TTTAAC +TTTAAG +TTTATA +TTTATT +TTTATC +TTTATG +TTTACA +TTTACT +TTTACC +TTTACG +TTTAGA +TTTAGT +TTTAGC +TTTAGG +TTTTAA +TTTTAT +TTTTAC +TTTTAG +TTTTTA +TTTTTT +TTTTTC +TTTTTG +TTTTCA +TTTTCT +TTTTCC +TTTTCG +TTTTGA +TTTTGT +TTTTGC +TTTTGG +TTTCAA +TTTCAT +TTTCAC +TTTCAG +TTTCTA +TTTCTT +TTTCTC +TTTCTG +TTTCCA +TTTCCT +TTTCCC +TTTCCG +TTTCGA +TTTCGT +TTTCGC +TTTCGG +TTTGAA +TTTGAT +TTTGAC +TTTGAG +TTTGTA +TTTGTT +TTTGTC +TTTGTG +TTTGCA +TTTGCT +TTTGCC +TTTGCG +TTTGGA +TTTGGT +TTTGGC +TTTGGG +TTCAAA +TTCAAT +TTCAAC +TTCAAG +TTCATA +TTCATT +TTCATC +TTCATG +TTCACA +TTCACT +TTCACC +TTCACG +TTCAGA +TTCAGT +TTCAGC +TTCAGG +TTCTAA +TTCTAT +TTCTAC +TTCTAG +TTCTTA +TTCTTT +TTCTTC +TTCTTG +TTCTCA +TTCTCT +TTCTCC +TTCTCG +TTCTGA +TTCTGT +TTCTGC +TTCTGG +TTCCAA +TTCCAT +TTCCAC +TTCCAG +TTCCTA +TTCCTT +TTCCTC +TTCCTG +TTCCCA +TTCCCT +TTCCCC +TTCCCG +TTCCGA +TTCCGT +TTCCGC +TTCCGG +TTCGAA +TTCGAT +TTCGAC +TTCGAG +TTCGTA +TTCGTT +TTCGTC +TTCGTG +TTCGCA +TTCGCT +TTCGCC +TTCGCG +TTCGGA +TTCGGT +TTCGGC +TTCGGG +TTGAAA +TTGAAT +TTGAAC +TTGAAG +TTGATA +TTGATT +TTGATC +TTGATG +TTGACA +TTGACT +TTGACC +TTGACG +TTGAGA +TTGAGT +TTGAGC +TTGAGG +TTGTAA +TTGTAT +TTGTAC +TTGTAG +TTGTTA +TTGTTT +TTGTTC +TTGTTG +TTGTCA +TTGTCT +TTGTCC +TTGTCG +TTGTGA +TTGTGT +TTGTGC +TTGTGG +TTGCAA +TTGCAT +TTGCAC +TTGCAG +TTGCTA +TTGCTT +TTGCTC +TTGCTG +TTGCCA +TTGCCT +TTGCCC +TTGCCG +TTGCGA +TTGCGT +TTGCGC +TTGCGG +TTGGAA +TTGGAT +TTGGAC +TTGGAG +TTGGTA +TTGGTT +TTGGTC +TTGGTG +TTGGCA +TTGGCT +TTGGCC +TTGGCG +TTGGGA +TTGGGT +TTGGGC +TTGGGG +TCAAAA +TCAAAT +TCAAAC +TCAAAG +TCAATA +TCAATT +TCAATC +TCAATG +TCAACA +TCAACT +TCAACC +TCAACG +TCAAGA +TCAAGT +TCAAGC +TCAAGG +TCATAA +TCATAT +TCATAC +TCATAG +TCATTA +TCATTT +TCATTC +TCATTG +TCATCA +TCATCT +TCATCC +TCATCG +TCATGA +TCATGT +TCATGC +TCATGG +TCACAA +TCACAT +TCACAC +TCACAG +TCACTA +TCACTT +TCACTC +TCACTG +TCACCA +TCACCT +TCACCC +TCACCG +TCACGA +TCACGT +TCACGC +TCACGG +TCAGAA +TCAGAT +TCAGAC +TCAGAG +TCAGTA +TCAGTT +TCAGTC +TCAGTG +TCAGCA +TCAGCT +TCAGCC +TCAGCG +TCAGGA +TCAGGT +TCAGGC +TCAGGG +TCTAAA +TCTAAT +TCTAAC +TCTAAG +TCTATA +TCTATT +TCTATC +TCTATG +TCTACA +TCTACT +TCTACC +TCTACG +TCTAGA +TCTAGT +TCTAGC +TCTAGG +TCTTAA +TCTTAT +TCTTAC +TCTTAG +TCTTTA +TCTTTT +TCTTTC +TCTTTG +TCTTCA +TCTTCT +TCTTCC +TCTTCG +TCTTGA +TCTTGT +TCTTGC +TCTTGG +TCTCAA +TCTCAT +TCTCAC +TCTCAG +TCTCTA +TCTCTT +TCTCTC +TCTCTG +TCTCCA +TCTCCT +TCTCCC +TCTCCG +TCTCGA +TCTCGT +TCTCGC +TCTCGG +TCTGAA +TCTGAT +TCTGAC +TCTGAG +TCTGTA +TCTGTT +TCTGTC +TCTGTG +TCTGCA +TCTGCT +TCTGCC +TCTGCG +TCTGGA +TCTGGT +TCTGGC +TCTGGG +TCCAAA +TCCAAT +TCCAAC +TCCAAG +TCCATA +TCCATT +TCCATC +TCCATG +TCCACA +TCCACT +TCCACC +TCCACG +TCCAGA +TCCAGT +TCCAGC +TCCAGG +TCCTAA +TCCTAT +TCCTAC +TCCTAG +TCCTTA +TCCTTT +TCCTTC +TCCTTG +TCCTCA +TCCTCT +TCCTCC +TCCTCG +TCCTGA +TCCTGT +TCCTGC +TCCTGG +TCCCAA +TCCCAT +TCCCAC +TCCCAG +TCCCTA +TCCCTT +TCCCTC +TCCCTG +TCCCCA +TCCCCT +TCCCCC +TCCCCG +TCCCGA +TCCCGT +TCCCGC +TCCCGG +TCCGAA +TCCGAT +TCCGAC +TCCGAG +TCCGTA +TCCGTT +TCCGTC +TCCGTG +TCCGCA +TCCGCT +TCCGCC +TCCGCG +TCCGGA +TCCGGT +TCCGGC +TCCGGG +TCGAAA +TCGAAT +TCGAAC +TCGAAG +TCGATA +TCGATT +TCGATC +TCGATG +TCGACA +TCGACT +TCGACC +TCGACG +TCGAGA +TCGAGT +TCGAGC +TCGAGG +TCGTAA +TCGTAT +TCGTAC +TCGTAG +TCGTTA +TCGTTT +TCGTTC +TCGTTG +TCGTCA +TCGTCT +TCGTCC +TCGTCG +TCGTGA +TCGTGT +TCGTGC +TCGTGG +TCGCAA +TCGCAT +TCGCAC +TCGCAG +TCGCTA +TCGCTT +TCGCTC +TCGCTG +TCGCCA +TCGCCT +TCGCCC +TCGCCG +TCGCGA +TCGCGT +TCGCGC +TCGCGG +TCGGAA +TCGGAT +TCGGAC +TCGGAG +TCGGTA +TCGGTT +TCGGTC +TCGGTG +TCGGCA +TCGGCT +TCGGCC +TCGGCG +TCGGGA +TCGGGT +TCGGGC +TCGGGG +TGAAAA +TGAAAT +TGAAAC +TGAAAG +TGAATA +TGAATT +TGAATC +TGAATG +TGAACA +TGAACT +TGAACC +TGAACG +TGAAGA +TGAAGT +TGAAGC +TGAAGG +TGATAA +TGATAT +TGATAC +TGATAG +TGATTA +TGATTT +TGATTC +TGATTG +TGATCA +TGATCT +TGATCC +TGATCG +TGATGA +TGATGT +TGATGC +TGATGG +TGACAA +TGACAT +TGACAC +TGACAG +TGACTA +TGACTT +TGACTC +TGACTG +TGACCA +TGACCT +TGACCC +TGACCG +TGACGA +TGACGT +TGACGC +TGACGG +TGAGAA +TGAGAT +TGAGAC +TGAGAG +TGAGTA +TGAGTT +TGAGTC +TGAGTG +TGAGCA +TGAGCT +TGAGCC +TGAGCG +TGAGGA +TGAGGT +TGAGGC +TGAGGG +TGTAAA +TGTAAT +TGTAAC +TGTAAG +TGTATA +TGTATT +TGTATC +TGTATG +TGTACA +TGTACT +TGTACC +TGTACG +TGTAGA +TGTAGT +TGTAGC +TGTAGG +TGTTAA +TGTTAT +TGTTAC +TGTTAG +TGTTTA +TGTTTT +TGTTTC +TGTTTG +TGTTCA +TGTTCT +TGTTCC +TGTTCG +TGTTGA +TGTTGT +TGTTGC +TGTTGG +TGTCAA +TGTCAT +TGTCAC +TGTCAG +TGTCTA +TGTCTT +TGTCTC +TGTCTG +TGTCCA +TGTCCT +TGTCCC +TGTCCG +TGTCGA +TGTCGT +TGTCGC +TGTCGG +TGTGAA +TGTGAT +TGTGAC +TGTGAG +TGTGTA +TGTGTT +TGTGTC +TGTGTG +TGTGCA +TGTGCT +TGTGCC +TGTGCG +TGTGGA +TGTGGT +TGTGGC +TGTGGG +TGCAAA +TGCAAT +TGCAAC +TGCAAG +TGCATA +TGCATT +TGCATC +TGCATG +TGCACA +TGCACT +TGCACC +TGCACG +TGCAGA +TGCAGT +TGCAGC +TGCAGG +TGCTAA +TGCTAT +TGCTAC +TGCTAG +TGCTTA +TGCTTT +TGCTTC +TGCTTG +TGCTCA +TGCTCT +TGCTCC +TGCTCG +TGCTGA +TGCTGT +TGCTGC +TGCTGG +TGCCAA +TGCCAT +TGCCAC +TGCCAG +TGCCTA +TGCCTT +TGCCTC +TGCCTG +TGCCCA +TGCCCT +TGCCCC +TGCCCG +TGCCGA +TGCCGT +TGCCGC +TGCCGG +TGCGAA +TGCGAT +TGCGAC +TGCGAG +TGCGTA +TGCGTT +TGCGTC +TGCGTG +TGCGCA +TGCGCT +TGCGCC +TGCGCG +TGCGGA +TGCGGT +TGCGGC +TGCGGG +TGGAAA +TGGAAT +TGGAAC +TGGAAG +TGGATA +TGGATT +TGGATC +TGGATG +TGGACA +TGGACT +TGGACC +TGGACG +TGGAGA +TGGAGT +TGGAGC +TGGAGG +TGGTAA +TGGTAT +TGGTAC +TGGTAG +TGGTTA +TGGTTT +TGGTTC +TGGTTG +TGGTCA +TGGTCT +TGGTCC +TGGTCG +TGGTGA +TGGTGT +TGGTGC +TGGTGG +TGGCAA +TGGCAT +TGGCAC +TGGCAG +TGGCTA +TGGCTT +TGGCTC +TGGCTG +TGGCCA +TGGCCT +TGGCCC +TGGCCG +TGGCGA +TGGCGT +TGGCGC +TGGCGG +TGGGAA +TGGGAT +TGGGAC +TGGGAG +TGGGTA +TGGGTT +TGGGTC +TGGGTG +TGGGCA +TGGGCT +TGGGCC +TGGGCG +TGGGGA +TGGGGT +TGGGGC +TGGGGG +CAAAAA +CAAAAT +CAAAAC +CAAAAG +CAAATA +CAAATT +CAAATC +CAAATG +CAAACA +CAAACT +CAAACC +CAAACG +CAAAGA +CAAAGT +CAAAGC +CAAAGG +CAATAA +CAATAT +CAATAC +CAATAG +CAATTA +CAATTT +CAATTC +CAATTG +CAATCA +CAATCT +CAATCC +CAATCG +CAATGA +CAATGT +CAATGC +CAATGG +CAACAA +CAACAT +CAACAC +CAACAG +CAACTA +CAACTT +CAACTC +CAACTG +CAACCA +CAACCT +CAACCC +CAACCG +CAACGA +CAACGT +CAACGC +CAACGG +CAAGAA +CAAGAT +CAAGAC +CAAGAG +CAAGTA +CAAGTT +CAAGTC +CAAGTG +CAAGCA +CAAGCT +CAAGCC +CAAGCG +CAAGGA +CAAGGT +CAAGGC +CAAGGG +CATAAA +CATAAT +CATAAC +CATAAG +CATATA +CATATT +CATATC +CATATG +CATACA +CATACT +CATACC +CATACG +CATAGA +CATAGT +CATAGC +CATAGG +CATTAA +CATTAT +CATTAC +CATTAG +CATTTA +CATTTT +CATTTC +CATTTG +CATTCA +CATTCT +CATTCC +CATTCG +CATTGA +CATTGT +CATTGC +CATTGG +CATCAA +CATCAT +CATCAC +CATCAG +CATCTA +CATCTT +CATCTC +CATCTG +CATCCA +CATCCT +CATCCC +CATCCG +CATCGA +CATCGT +CATCGC +CATCGG +CATGAA +CATGAT +CATGAC +CATGAG +CATGTA +CATGTT +CATGTC +CATGTG +CATGCA +CATGCT +CATGCC +CATGCG +CATGGA +CATGGT +CATGGC +CATGGG +CACAAA +CACAAT +CACAAC +CACAAG +CACATA +CACATT +CACATC +CACATG +CACACA +CACACT +CACACC +CACACG +CACAGA +CACAGT +CACAGC +CACAGG +CACTAA +CACTAT +CACTAC +CACTAG +CACTTA +CACTTT +CACTTC +CACTTG +CACTCA +CACTCT +CACTCC +CACTCG +CACTGA +CACTGT +CACTGC +CACTGG +CACCAA +CACCAT +CACCAC +CACCAG +CACCTA +CACCTT +CACCTC +CACCTG +CACCCA +CACCCT +CACCCC +CACCCG +CACCGA +CACCGT +CACCGC +CACCGG +CACGAA +CACGAT +CACGAC +CACGAG +CACGTA +CACGTT +CACGTC +CACGTG +CACGCA +CACGCT +CACGCC +CACGCG +CACGGA +CACGGT +CACGGC +CACGGG +CAGAAA +CAGAAT +CAGAAC +CAGAAG +CAGATA +CAGATT +CAGATC +CAGATG +CAGACA +CAGACT +CAGACC +CAGACG +CAGAGA +CAGAGT +CAGAGC +CAGAGG +CAGTAA +CAGTAT +CAGTAC +CAGTAG +CAGTTA +CAGTTT +CAGTTC +CAGTTG +CAGTCA +CAGTCT +CAGTCC +CAGTCG +CAGTGA +CAGTGT +CAGTGC +CAGTGG +CAGCAA +CAGCAT +CAGCAC +CAGCAG +CAGCTA +CAGCTT +CAGCTC +CAGCTG +CAGCCA +CAGCCT +CAGCCC +CAGCCG +CAGCGA +CAGCGT +CAGCGC +CAGCGG +CAGGAA +CAGGAT +CAGGAC +CAGGAG +CAGGTA +CAGGTT +CAGGTC +CAGGTG +CAGGCA +CAGGCT +CAGGCC +CAGGCG +CAGGGA +CAGGGT +CAGGGC +CAGGGG +CTAAAA +CTAAAT +CTAAAC +CTAAAG +CTAATA +CTAATT +CTAATC +CTAATG +CTAACA +CTAACT +CTAACC +CTAACG +CTAAGA +CTAAGT +CTAAGC +CTAAGG +CTATAA +CTATAT +CTATAC +CTATAG +CTATTA +CTATTT +CTATTC +CTATTG +CTATCA +CTATCT +CTATCC +CTATCG +CTATGA +CTATGT +CTATGC +CTATGG +CTACAA +CTACAT +CTACAC +CTACAG +CTACTA +CTACTT +CTACTC +CTACTG +CTACCA +CTACCT +CTACCC +CTACCG +CTACGA +CTACGT +CTACGC +CTACGG +CTAGAA +CTAGAT +CTAGAC +CTAGAG +CTAGTA +CTAGTT +CTAGTC +CTAGTG +CTAGCA +CTAGCT +CTAGCC +CTAGCG +CTAGGA +CTAGGT +CTAGGC +CTAGGG +CTTAAA +CTTAAT +CTTAAC +CTTAAG +CTTATA +CTTATT +CTTATC +CTTATG +CTTACA +CTTACT +CTTACC +CTTACG +CTTAGA +CTTAGT +CTTAGC +CTTAGG +CTTTAA +CTTTAT +CTTTAC +CTTTAG +CTTTTA +CTTTTT +CTTTTC +CTTTTG +CTTTCA +CTTTCT +CTTTCC +CTTTCG +CTTTGA +CTTTGT +CTTTGC +CTTTGG +CTTCAA +CTTCAT +CTTCAC +CTTCAG +CTTCTA +CTTCTT +CTTCTC +CTTCTG +CTTCCA +CTTCCT +CTTCCC +CTTCCG +CTTCGA +CTTCGT +CTTCGC +CTTCGG +CTTGAA +CTTGAT +CTTGAC +CTTGAG +CTTGTA +CTTGTT +CTTGTC +CTTGTG +CTTGCA +CTTGCT +CTTGCC +CTTGCG +CTTGGA +CTTGGT +CTTGGC +CTTGGG +CTCAAA +CTCAAT +CTCAAC +CTCAAG +CTCATA +CTCATT +CTCATC +CTCATG +CTCACA +CTCACT +CTCACC +CTCACG +CTCAGA +CTCAGT +CTCAGC +CTCAGG +CTCTAA +CTCTAT +CTCTAC +CTCTAG +CTCTTA +CTCTTT +CTCTTC +CTCTTG +CTCTCA +CTCTCT +CTCTCC +CTCTCG +CTCTGA +CTCTGT +CTCTGC +CTCTGG +CTCCAA +CTCCAT +CTCCAC +CTCCAG +CTCCTA +CTCCTT +CTCCTC +CTCCTG +CTCCCA +CTCCCT +CTCCCC +CTCCCG +CTCCGA +CTCCGT +CTCCGC +CTCCGG +CTCGAA +CTCGAT +CTCGAC +CTCGAG +CTCGTA +CTCGTT +CTCGTC +CTCGTG +CTCGCA +CTCGCT +CTCGCC +CTCGCG +CTCGGA +CTCGGT +CTCGGC +CTCGGG +CTGAAA +CTGAAT +CTGAAC +CTGAAG +CTGATA +CTGATT +CTGATC +CTGATG +CTGACA +CTGACT +CTGACC +CTGACG +CTGAGA +CTGAGT +CTGAGC +CTGAGG +CTGTAA +CTGTAT +CTGTAC +CTGTAG +CTGTTA +CTGTTT +CTGTTC +CTGTTG +CTGTCA +CTGTCT +CTGTCC +CTGTCG +CTGTGA +CTGTGT +CTGTGC +CTGTGG +CTGCAA +CTGCAT +CTGCAC +CTGCAG +CTGCTA +CTGCTT +CTGCTC +CTGCTG +CTGCCA +CTGCCT +CTGCCC +CTGCCG +CTGCGA +CTGCGT +CTGCGC +CTGCGG +CTGGAA +CTGGAT +CTGGAC +CTGGAG +CTGGTA +CTGGTT +CTGGTC +CTGGTG +CTGGCA +CTGGCT +CTGGCC +CTGGCG +CTGGGA +CTGGGT +CTGGGC +CTGGGG +CCAAAA +CCAAAT +CCAAAC +CCAAAG +CCAATA +CCAATT +CCAATC +CCAATG +CCAACA +CCAACT +CCAACC +CCAACG +CCAAGA +CCAAGT +CCAAGC +CCAAGG +CCATAA +CCATAT +CCATAC +CCATAG +CCATTA +CCATTT +CCATTC +CCATTG +CCATCA +CCATCT +CCATCC +CCATCG +CCATGA +CCATGT +CCATGC +CCATGG +CCACAA +CCACAT +CCACAC +CCACAG +CCACTA +CCACTT +CCACTC +CCACTG +CCACCA +CCACCT +CCACCC +CCACCG +CCACGA +CCACGT +CCACGC +CCACGG +CCAGAA +CCAGAT +CCAGAC +CCAGAG +CCAGTA +CCAGTT +CCAGTC +CCAGTG +CCAGCA +CCAGCT +CCAGCC +CCAGCG +CCAGGA +CCAGGT +CCAGGC +CCAGGG +CCTAAA +CCTAAT +CCTAAC +CCTAAG +CCTATA +CCTATT +CCTATC +CCTATG +CCTACA +CCTACT +CCTACC +CCTACG +CCTAGA +CCTAGT +CCTAGC +CCTAGG +CCTTAA +CCTTAT +CCTTAC +CCTTAG +CCTTTA +CCTTTT +CCTTTC +CCTTTG +CCTTCA +CCTTCT +CCTTCC +CCTTCG +CCTTGA +CCTTGT +CCTTGC +CCTTGG +CCTCAA +CCTCAT +CCTCAC +CCTCAG +CCTCTA +CCTCTT +CCTCTC +CCTCTG +CCTCCA +CCTCCT +CCTCCC +CCTCCG +CCTCGA +CCTCGT +CCTCGC +CCTCGG +CCTGAA +CCTGAT +CCTGAC +CCTGAG +CCTGTA +CCTGTT +CCTGTC +CCTGTG +CCTGCA +CCTGCT +CCTGCC +CCTGCG +CCTGGA +CCTGGT +CCTGGC +CCTGGG +CCCAAA +CCCAAT +CCCAAC +CCCAAG +CCCATA +CCCATT +CCCATC +CCCATG +CCCACA +CCCACT +CCCACC +CCCACG +CCCAGA +CCCAGT +CCCAGC +CCCAGG +CCCTAA +CCCTAT +CCCTAC +CCCTAG +CCCTTA +CCCTTT +CCCTTC +CCCTTG +CCCTCA +CCCTCT +CCCTCC +CCCTCG +CCCTGA +CCCTGT +CCCTGC +CCCTGG +CCCCAA +CCCCAT +CCCCAC +CCCCAG +CCCCTA +CCCCTT +CCCCTC +CCCCTG +CCCCCA +CCCCCT +CCCCCC +CCCCCG +CCCCGA +CCCCGT +CCCCGC +CCCCGG +CCCGAA +CCCGAT +CCCGAC +CCCGAG +CCCGTA +CCCGTT +CCCGTC +CCCGTG +CCCGCA +CCCGCT +CCCGCC +CCCGCG +CCCGGA +CCCGGT +CCCGGC +CCCGGG +CCGAAA +CCGAAT +CCGAAC +CCGAAG +CCGATA +CCGATT +CCGATC +CCGATG +CCGACA +CCGACT +CCGACC +CCGACG +CCGAGA +CCGAGT +CCGAGC +CCGAGG +CCGTAA +CCGTAT +CCGTAC +CCGTAG +CCGTTA +CCGTTT +CCGTTC +CCGTTG +CCGTCA +CCGTCT +CCGTCC +CCGTCG +CCGTGA +CCGTGT +CCGTGC +CCGTGG +CCGCAA +CCGCAT +CCGCAC +CCGCAG +CCGCTA +CCGCTT +CCGCTC +CCGCTG +CCGCCA +CCGCCT +CCGCCC +CCGCCG +CCGCGA +CCGCGT +CCGCGC +CCGCGG +CCGGAA +CCGGAT +CCGGAC +CCGGAG +CCGGTA +CCGGTT +CCGGTC +CCGGTG +CCGGCA +CCGGCT +CCGGCC +CCGGCG +CCGGGA +CCGGGT +CCGGGC +CCGGGG +CGAAAA +CGAAAT +CGAAAC +CGAAAG +CGAATA +CGAATT +CGAATC +CGAATG +CGAACA +CGAACT +CGAACC +CGAACG +CGAAGA +CGAAGT +CGAAGC +CGAAGG +CGATAA +CGATAT +CGATAC +CGATAG +CGATTA +CGATTT +CGATTC +CGATTG +CGATCA +CGATCT +CGATCC +CGATCG +CGATGA +CGATGT +CGATGC +CGATGG +CGACAA +CGACAT +CGACAC +CGACAG +CGACTA +CGACTT +CGACTC +CGACTG +CGACCA +CGACCT +CGACCC +CGACCG +CGACGA +CGACGT +CGACGC +CGACGG +CGAGAA +CGAGAT +CGAGAC +CGAGAG +CGAGTA +CGAGTT +CGAGTC +CGAGTG +CGAGCA +CGAGCT +CGAGCC +CGAGCG +CGAGGA +CGAGGT +CGAGGC +CGAGGG +CGTAAA +CGTAAT +CGTAAC +CGTAAG +CGTATA +CGTATT +CGTATC +CGTATG +CGTACA +CGTACT +CGTACC +CGTACG +CGTAGA +CGTAGT +CGTAGC +CGTAGG +CGTTAA +CGTTAT +CGTTAC +CGTTAG +CGTTTA +CGTTTT +CGTTTC +CGTTTG +CGTTCA +CGTTCT +CGTTCC +CGTTCG +CGTTGA +CGTTGT +CGTTGC +CGTTGG +CGTCAA +CGTCAT +CGTCAC +CGTCAG +CGTCTA +CGTCTT +CGTCTC +CGTCTG +CGTCCA +CGTCCT +CGTCCC +CGTCCG +CGTCGA +CGTCGT +CGTCGC +CGTCGG +CGTGAA +CGTGAT +CGTGAC +CGTGAG +CGTGTA +CGTGTT +CGTGTC +CGTGTG +CGTGCA +CGTGCT +CGTGCC +CGTGCG +CGTGGA +CGTGGT +CGTGGC +CGTGGG +CGCAAA +CGCAAT +CGCAAC +CGCAAG +CGCATA +CGCATT +CGCATC +CGCATG +CGCACA +CGCACT +CGCACC +CGCACG +CGCAGA +CGCAGT +CGCAGC +CGCAGG +CGCTAA +CGCTAT +CGCTAC +CGCTAG +CGCTTA +CGCTTT +CGCTTC +CGCTTG +CGCTCA +CGCTCT +CGCTCC +CGCTCG +CGCTGA +CGCTGT +CGCTGC +CGCTGG +CGCCAA +CGCCAT +CGCCAC +CGCCAG +CGCCTA +CGCCTT +CGCCTC +CGCCTG +CGCCCA +CGCCCT +CGCCCC +CGCCCG +CGCCGA +CGCCGT +CGCCGC +CGCCGG +CGCGAA +CGCGAT +CGCGAC +CGCGAG +CGCGTA +CGCGTT +CGCGTC +CGCGTG +CGCGCA +CGCGCT +CGCGCC +CGCGCG +CGCGGA +CGCGGT +CGCGGC +CGCGGG +CGGAAA +CGGAAT +CGGAAC +CGGAAG +CGGATA +CGGATT +CGGATC +CGGATG +CGGACA +CGGACT +CGGACC +CGGACG +CGGAGA +CGGAGT +CGGAGC +CGGAGG +CGGTAA +CGGTAT +CGGTAC +CGGTAG +CGGTTA +CGGTTT +CGGTTC +CGGTTG +CGGTCA +CGGTCT +CGGTCC +CGGTCG +CGGTGA +CGGTGT +CGGTGC +CGGTGG +CGGCAA +CGGCAT +CGGCAC +CGGCAG +CGGCTA +CGGCTT +CGGCTC +CGGCTG +CGGCCA +CGGCCT +CGGCCC +CGGCCG +CGGCGA +CGGCGT +CGGCGC +CGGCGG +CGGGAA +CGGGAT +CGGGAC +CGGGAG +CGGGTA +CGGGTT +CGGGTC +CGGGTG +CGGGCA +CGGGCT +CGGGCC +CGGGCG +CGGGGA +CGGGGT +CGGGGC +CGGGGG +GAAAAA +GAAAAT +GAAAAC +GAAAAG +GAAATA +GAAATT +GAAATC +GAAATG +GAAACA +GAAACT +GAAACC +GAAACG +GAAAGA +GAAAGT +GAAAGC +GAAAGG +GAATAA +GAATAT +GAATAC +GAATAG +GAATTA +GAATTT +GAATTC +GAATTG +GAATCA +GAATCT +GAATCC +GAATCG +GAATGA +GAATGT +GAATGC +GAATGG +GAACAA +GAACAT +GAACAC +GAACAG +GAACTA +GAACTT +GAACTC +GAACTG +GAACCA +GAACCT +GAACCC +GAACCG +GAACGA +GAACGT +GAACGC +GAACGG +GAAGAA +GAAGAT +GAAGAC +GAAGAG +GAAGTA +GAAGTT +GAAGTC +GAAGTG +GAAGCA +GAAGCT +GAAGCC +GAAGCG +GAAGGA +GAAGGT +GAAGGC +GAAGGG +GATAAA +GATAAT +GATAAC +GATAAG +GATATA +GATATT +GATATC +GATATG +GATACA +GATACT +GATACC +GATACG +GATAGA +GATAGT +GATAGC +GATAGG +GATTAA +GATTAT +GATTAC +GATTAG +GATTTA +GATTTT +GATTTC +GATTTG +GATTCA +GATTCT +GATTCC +GATTCG +GATTGA +GATTGT +GATTGC +GATTGG +GATCAA +GATCAT +GATCAC +GATCAG +GATCTA +GATCTT +GATCTC +GATCTG +GATCCA +GATCCT +GATCCC +GATCCG +GATCGA +GATCGT +GATCGC +GATCGG +GATGAA +GATGAT +GATGAC +GATGAG +GATGTA +GATGTT +GATGTC +GATGTG +GATGCA +GATGCT +GATGCC +GATGCG +GATGGA +GATGGT +GATGGC +GATGGG +GACAAA +GACAAT +GACAAC +GACAAG +GACATA +GACATT +GACATC +GACATG +GACACA +GACACT +GACACC +GACACG +GACAGA +GACAGT +GACAGC +GACAGG +GACTAA +GACTAT +GACTAC +GACTAG +GACTTA +GACTTT +GACTTC +GACTTG +GACTCA +GACTCT +GACTCC +GACTCG +GACTGA +GACTGT +GACTGC +GACTGG +GACCAA +GACCAT +GACCAC +GACCAG +GACCTA +GACCTT +GACCTC +GACCTG +GACCCA +GACCCT +GACCCC +GACCCG +GACCGA +GACCGT +GACCGC +GACCGG +GACGAA +GACGAT +GACGAC +GACGAG +GACGTA +GACGTT +GACGTC +GACGTG +GACGCA +GACGCT +GACGCC +GACGCG +GACGGA +GACGGT +GACGGC +GACGGG +GAGAAA +GAGAAT +GAGAAC +GAGAAG +GAGATA +GAGATT +GAGATC +GAGATG +GAGACA +GAGACT +GAGACC +GAGACG +GAGAGA +GAGAGT +GAGAGC +GAGAGG +GAGTAA +GAGTAT +GAGTAC +GAGTAG +GAGTTA +GAGTTT +GAGTTC +GAGTTG +GAGTCA +GAGTCT +GAGTCC +GAGTCG +GAGTGA +GAGTGT +GAGTGC +GAGTGG +GAGCAA +GAGCAT +GAGCAC +GAGCAG +GAGCTA +GAGCTT +GAGCTC +GAGCTG +GAGCCA +GAGCCT +GAGCCC +GAGCCG +GAGCGA +GAGCGT +GAGCGC +GAGCGG +GAGGAA +GAGGAT +GAGGAC +GAGGAG +GAGGTA +GAGGTT +GAGGTC +GAGGTG +GAGGCA +GAGGCT +GAGGCC +GAGGCG +GAGGGA +GAGGGT +GAGGGC +GAGGGG +GTAAAA +GTAAAT +GTAAAC +GTAAAG +GTAATA +GTAATT +GTAATC +GTAATG +GTAACA +GTAACT +GTAACC +GTAACG +GTAAGA +GTAAGT +GTAAGC +GTAAGG +GTATAA +GTATAT +GTATAC +GTATAG +GTATTA +GTATTT +GTATTC +GTATTG +GTATCA +GTATCT +GTATCC +GTATCG +GTATGA +GTATGT +GTATGC +GTATGG +GTACAA +GTACAT +GTACAC +GTACAG +GTACTA +GTACTT +GTACTC +GTACTG +GTACCA +GTACCT +GTACCC +GTACCG +GTACGA +GTACGT +GTACGC +GTACGG +GTAGAA +GTAGAT +GTAGAC +GTAGAG +GTAGTA +GTAGTT +GTAGTC +GTAGTG +GTAGCA +GTAGCT +GTAGCC +GTAGCG +GTAGGA +GTAGGT +GTAGGC +GTAGGG +GTTAAA +GTTAAT +GTTAAC +GTTAAG +GTTATA +GTTATT +GTTATC +GTTATG +GTTACA +GTTACT +GTTACC +GTTACG +GTTAGA +GTTAGT +GTTAGC +GTTAGG +GTTTAA +GTTTAT +GTTTAC +GTTTAG +GTTTTA +GTTTTT +GTTTTC +GTTTTG +GTTTCA +GTTTCT +GTTTCC +GTTTCG +GTTTGA +GTTTGT +GTTTGC +GTTTGG +GTTCAA +GTTCAT +GTTCAC +GTTCAG +GTTCTA +GTTCTT +GTTCTC +GTTCTG +GTTCCA +GTTCCT +GTTCCC +GTTCCG +GTTCGA +GTTCGT +GTTCGC +GTTCGG +GTTGAA +GTTGAT +GTTGAC +GTTGAG +GTTGTA +GTTGTT +GTTGTC +GTTGTG +GTTGCA +GTTGCT +GTTGCC +GTTGCG +GTTGGA +GTTGGT +GTTGGC +GTTGGG +GTCAAA +GTCAAT +GTCAAC +GTCAAG +GTCATA +GTCATT +GTCATC +GTCATG +GTCACA +GTCACT +GTCACC +GTCACG +GTCAGA +GTCAGT +GTCAGC +GTCAGG +GTCTAA +GTCTAT +GTCTAC +GTCTAG +GTCTTA +GTCTTT +GTCTTC +GTCTTG +GTCTCA +GTCTCT +GTCTCC +GTCTCG +GTCTGA +GTCTGT +GTCTGC +GTCTGG +GTCCAA +GTCCAT +GTCCAC +GTCCAG +GTCCTA +GTCCTT +GTCCTC +GTCCTG +GTCCCA +GTCCCT +GTCCCC +GTCCCG +GTCCGA +GTCCGT +GTCCGC +GTCCGG +GTCGAA +GTCGAT +GTCGAC +GTCGAG +GTCGTA +GTCGTT +GTCGTC +GTCGTG +GTCGCA +GTCGCT +GTCGCC +GTCGCG +GTCGGA +GTCGGT +GTCGGC +GTCGGG +GTGAAA +GTGAAT +GTGAAC +GTGAAG +GTGATA +GTGATT +GTGATC +GTGATG +GTGACA +GTGACT +GTGACC +GTGACG +GTGAGA +GTGAGT +GTGAGC +GTGAGG +GTGTAA +GTGTAT +GTGTAC +GTGTAG +GTGTTA +GTGTTT +GTGTTC +GTGTTG +GTGTCA +GTGTCT +GTGTCC +GTGTCG +GTGTGA +GTGTGT +GTGTGC +GTGTGG +GTGCAA +GTGCAT +GTGCAC +GTGCAG +GTGCTA +GTGCTT +GTGCTC +GTGCTG +GTGCCA +GTGCCT +GTGCCC +GTGCCG +GTGCGA +GTGCGT +GTGCGC +GTGCGG +GTGGAA +GTGGAT +GTGGAC +GTGGAG +GTGGTA +GTGGTT +GTGGTC +GTGGTG +GTGGCA +GTGGCT +GTGGCC +GTGGCG +GTGGGA +GTGGGT +GTGGGC +GTGGGG +GCAAAA +GCAAAT +GCAAAC +GCAAAG +GCAATA +GCAATT +GCAATC +GCAATG +GCAACA +GCAACT +GCAACC +GCAACG +GCAAGA +GCAAGT +GCAAGC +GCAAGG +GCATAA +GCATAT +GCATAC +GCATAG +GCATTA +GCATTT +GCATTC +GCATTG +GCATCA +GCATCT +GCATCC +GCATCG +GCATGA +GCATGT +GCATGC +GCATGG +GCACAA +GCACAT +GCACAC +GCACAG +GCACTA +GCACTT +GCACTC +GCACTG +GCACCA +GCACCT +GCACCC +GCACCG +GCACGA +GCACGT +GCACGC +GCACGG +GCAGAA +GCAGAT +GCAGAC +GCAGAG +GCAGTA +GCAGTT +GCAGTC +GCAGTG +GCAGCA +GCAGCT +GCAGCC +GCAGCG +GCAGGA +GCAGGT +GCAGGC +GCAGGG +GCTAAA +GCTAAT +GCTAAC +GCTAAG +GCTATA +GCTATT +GCTATC +GCTATG +GCTACA +GCTACT +GCTACC +GCTACG +GCTAGA +GCTAGT +GCTAGC +GCTAGG +GCTTAA +GCTTAT +GCTTAC +GCTTAG +GCTTTA +GCTTTT +GCTTTC +GCTTTG +GCTTCA +GCTTCT +GCTTCC +GCTTCG +GCTTGA +GCTTGT +GCTTGC +GCTTGG +GCTCAA +GCTCAT +GCTCAC +GCTCAG +GCTCTA +GCTCTT +GCTCTC +GCTCTG +GCTCCA +GCTCCT +GCTCCC +GCTCCG +GCTCGA +GCTCGT +GCTCGC +GCTCGG +GCTGAA +GCTGAT +GCTGAC +GCTGAG +GCTGTA +GCTGTT +GCTGTC +GCTGTG +GCTGCA +GCTGCT +GCTGCC +GCTGCG +GCTGGA +GCTGGT +GCTGGC +GCTGGG +GCCAAA +GCCAAT +GCCAAC +GCCAAG +GCCATA +GCCATT +GCCATC +GCCATG +GCCACA +GCCACT +GCCACC +GCCACG +GCCAGA +GCCAGT +GCCAGC +GCCAGG +GCCTAA +GCCTAT +GCCTAC +GCCTAG +GCCTTA +GCCTTT +GCCTTC +GCCTTG +GCCTCA +GCCTCT +GCCTCC +GCCTCG +GCCTGA +GCCTGT +GCCTGC +GCCTGG +GCCCAA +GCCCAT +GCCCAC +GCCCAG +GCCCTA +GCCCTT +GCCCTC +GCCCTG +GCCCCA +GCCCCT +GCCCCC +GCCCCG +GCCCGA +GCCCGT +GCCCGC +GCCCGG +GCCGAA +GCCGAT +GCCGAC +GCCGAG +GCCGTA +GCCGTT +GCCGTC +GCCGTG +GCCGCA +GCCGCT +GCCGCC +GCCGCG +GCCGGA +GCCGGT +GCCGGC +GCCGGG +GCGAAA +GCGAAT +GCGAAC +GCGAAG +GCGATA +GCGATT +GCGATC +GCGATG +GCGACA +GCGACT +GCGACC +GCGACG +GCGAGA +GCGAGT +GCGAGC +GCGAGG +GCGTAA +GCGTAT +GCGTAC +GCGTAG +GCGTTA +GCGTTT +GCGTTC +GCGTTG +GCGTCA +GCGTCT +GCGTCC +GCGTCG +GCGTGA +GCGTGT +GCGTGC +GCGTGG +GCGCAA +GCGCAT +GCGCAC +GCGCAG +GCGCTA +GCGCTT +GCGCTC +GCGCTG +GCGCCA +GCGCCT +GCGCCC +GCGCCG +GCGCGA +GCGCGT +GCGCGC +GCGCGG +GCGGAA +GCGGAT +GCGGAC +GCGGAG +GCGGTA +GCGGTT +GCGGTC +GCGGTG +GCGGCA +GCGGCT +GCGGCC +GCGGCG +GCGGGA +GCGGGT +GCGGGC +GCGGGG +GGAAAA +GGAAAT +GGAAAC +GGAAAG +GGAATA +GGAATT +GGAATC +GGAATG +GGAACA +GGAACT +GGAACC +GGAACG +GGAAGA +GGAAGT +GGAAGC +GGAAGG +GGATAA +GGATAT +GGATAC +GGATAG +GGATTA +GGATTT +GGATTC +GGATTG +GGATCA +GGATCT +GGATCC +GGATCG +GGATGA +GGATGT +GGATGC +GGATGG +GGACAA +GGACAT +GGACAC +GGACAG +GGACTA +GGACTT +GGACTC +GGACTG +GGACCA +GGACCT +GGACCC +GGACCG +GGACGA +GGACGT +GGACGC +GGACGG +GGAGAA +GGAGAT +GGAGAC +GGAGAG +GGAGTA +GGAGTT +GGAGTC +GGAGTG +GGAGCA +GGAGCT +GGAGCC +GGAGCG +GGAGGA +GGAGGT +GGAGGC +GGAGGG +GGTAAA +GGTAAT +GGTAAC +GGTAAG +GGTATA +GGTATT +GGTATC +GGTATG +GGTACA +GGTACT +GGTACC +GGTACG +GGTAGA +GGTAGT +GGTAGC +GGTAGG +GGTTAA +GGTTAT +GGTTAC +GGTTAG +GGTTTA +GGTTTT +GGTTTC +GGTTTG +GGTTCA +GGTTCT +GGTTCC +GGTTCG +GGTTGA +GGTTGT +GGTTGC +GGTTGG +GGTCAA +GGTCAT +GGTCAC +GGTCAG +GGTCTA +GGTCTT +GGTCTC +GGTCTG +GGTCCA +GGTCCT +GGTCCC +GGTCCG +GGTCGA +GGTCGT +GGTCGC +GGTCGG +GGTGAA +GGTGAT +GGTGAC +GGTGAG +GGTGTA +GGTGTT +GGTGTC +GGTGTG +GGTGCA +GGTGCT +GGTGCC +GGTGCG +GGTGGA +GGTGGT +GGTGGC +GGTGGG +GGCAAA +GGCAAT +GGCAAC +GGCAAG +GGCATA +GGCATT +GGCATC +GGCATG +GGCACA +GGCACT +GGCACC +GGCACG +GGCAGA +GGCAGT +GGCAGC +GGCAGG +GGCTAA +GGCTAT +GGCTAC +GGCTAG +GGCTTA +GGCTTT +GGCTTC +GGCTTG +GGCTCA +GGCTCT +GGCTCC +GGCTCG +GGCTGA +GGCTGT +GGCTGC +GGCTGG +GGCCAA +GGCCAT +GGCCAC +GGCCAG +GGCCTA +GGCCTT +GGCCTC +GGCCTG +GGCCCA +GGCCCT +GGCCCC +GGCCCG +GGCCGA +GGCCGT +GGCCGC +GGCCGG +GGCGAA +GGCGAT +GGCGAC +GGCGAG +GGCGTA +GGCGTT +GGCGTC +GGCGTG +GGCGCA +GGCGCT +GGCGCC +GGCGCG +GGCGGA +GGCGGT +GGCGGC +GGCGGG +GGGAAA +GGGAAT +GGGAAC +GGGAAG +GGGATA +GGGATT +GGGATC +GGGATG +GGGACA +GGGACT +GGGACC +GGGACG +GGGAGA +GGGAGT +GGGAGC +GGGAGG +GGGTAA +GGGTAT +GGGTAC +GGGTAG +GGGTTA +GGGTTT +GGGTTC +GGGTTG +GGGTCA +GGGTCT +GGGTCC +GGGTCG +GGGTGA +GGGTGT +GGGTGC +GGGTGG +GGGCAA +GGGCAT +GGGCAC +GGGCAG +GGGCTA +GGGCTT +GGGCTC +GGGCTG +GGGCCA +GGGCCT +GGGCCC +GGGCCG +GGGCGA +GGGCGT +GGGCGC +GGGCGG +GGGGAA +GGGGAT +GGGGAC +GGGGAG +GGGGTA +GGGGTT +GGGGTC +GGGGTG +GGGGCA +GGGGCT +GGGGCC +GGGGCG +GGGGGA +GGGGGT +GGGGGC +GGGGGG diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-3000/config.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..9b14e61d327bf586fcff6e817cf9d365bae18ad2 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-3000/config.json @@ -0,0 +1,26 @@ +{ + "_name_or_path": "fabihamakhdoomi/TinyDNABERT", + "architectures": [ + "BertForSequenceClassification" + ], + "attention_probs_dropout_prob": 0.1, + "classifier_dropout": null, + "hidden_act": "gelu", + "hidden_dropout_prob": 0.1, + "hidden_size": 128, + "initializer_range": 0.02, + "intermediate_size": 512, + "layer_norm_eps": 1e-12, + "max_position_embeddings": 512, + "model_type": "bert", + "num_attention_heads": 2, + "num_hidden_layers": 2, + "pad_token_id": 0, + "position_embedding_type": "absolute", + "problem_type": "single_label_classification", + "torch_dtype": "float32", + "transformers_version": "4.44.0", + "type_vocab_size": 2, + "use_cache": true, + "vocab_size": 4101 +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-3000/model.safetensors b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-3000/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..6aa3f471a0d3a0d0777bb9433be81d37d97c9334 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-3000/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cf4ab3d7dbfb42b8b0cc69e3339f508b471e0cf59512c7d486a86f6d4770222f +size 4021688 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-3000/optimizer.pt b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-3000/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..4d4846707bac2438f4361e213be84ab43ddc39c8 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-3000/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c882b9db4b5d7e66328d8edaaab6f02508112500cee1202f535465cd5913089b +size 8058700 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-3000/rng_state.pth b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-3000/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..6a219fb332ae976c7a91fd53a363fd52b1ca4044 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-3000/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9a88049e3e15cef9ae24c0067d91d8e8d16ed125ffa58bfc970554dbe3e0b701 +size 14244 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-3000/scheduler.pt b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-3000/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..6c655a9f3dd5d4d532f6c0a84b19b9ae5b766437 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-3000/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dbc2253b3cc1ddd11640ea56d3925f925017f9aa6e1577411dac42d629457af6 +size 1064 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-3000/special_tokens_map.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-3000/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..a8b3208c2884c4efb86e49300fdd3dc877220cdf --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-3000/special_tokens_map.json @@ -0,0 +1,7 @@ +{ + "cls_token": "[CLS]", + "mask_token": "[MASK]", + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-3000/tokenizer.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-3000/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..ebd782de6963b9032ce9374cfe30f5275be1a30b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-3000/tokenizer.json @@ -0,0 +1,4265 @@ +{ + "version": "1.0", + "truncation": { + "direction": "Right", + "max_length": 512, + "strategy": "LongestFirst", + "stride": 0 + }, + "padding": { + "strategy": { + "Fixed": 512 + }, + "direction": "Right", + "pad_to_multiple_of": null, + "pad_id": 0, + "pad_type_id": 0, + "pad_token": "[PAD]" + }, + "added_tokens": [ + { + "id": 0, + "content": "[PAD]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 1, + "content": "[UNK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 2, + "content": "[CLS]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 3, + "content": "[SEP]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 4, + "content": "[MASK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + } + ], + "normalizer": { + "type": "BertNormalizer", + "clean_text": true, + "handle_chinese_chars": true, + "strip_accents": null, + "lowercase": false + }, + "pre_tokenizer": { + "type": "BertPreTokenizer" + }, + "post_processor": { + "type": "TemplateProcessing", + "single": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + } + ], + "pair": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "B", + "type_id": 1 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 1 + } + } + ], + "special_tokens": { + "[CLS]": { + "id": "[CLS]", + "ids": [ + 2 + ], + "tokens": [ + "[CLS]" + ] + }, + "[SEP]": { + "id": "[SEP]", + "ids": [ + 3 + ], + "tokens": [ + "[SEP]" + ] + } + } + }, + "decoder": { + "type": "WordPiece", + "prefix": "##", + "cleanup": true + }, + "model": { + "type": "WordPiece", + "unk_token": "[UNK]", + "continuing_subword_prefix": "##", + "max_input_chars_per_word": 100, + "vocab": { + "[PAD]": 0, + "[UNK]": 1, + "[CLS]": 2, + "[SEP]": 3, + "[MASK]": 4, + "AAAAAA": 5, + "AAAAAT": 6, + "AAAAAC": 7, + "AAAAAG": 8, + "AAAATA": 9, + "AAAATT": 10, + "AAAATC": 11, + "AAAATG": 12, + "AAAACA": 13, + "AAAACT": 14, + "AAAACC": 15, + "AAAACG": 16, + "AAAAGA": 17, + "AAAAGT": 18, + "AAAAGC": 19, + "AAAAGG": 20, + "AAATAA": 21, + "AAATAT": 22, + "AAATAC": 23, + "AAATAG": 24, + "AAATTA": 25, + "AAATTT": 26, + "AAATTC": 27, + "AAATTG": 28, + "AAATCA": 29, + "AAATCT": 30, + "AAATCC": 31, + "AAATCG": 32, + "AAATGA": 33, + "AAATGT": 34, + "AAATGC": 35, + "AAATGG": 36, + "AAACAA": 37, + "AAACAT": 38, + "AAACAC": 39, + "AAACAG": 40, + "AAACTA": 41, + "AAACTT": 42, + "AAACTC": 43, + "AAACTG": 44, + "AAACCA": 45, + "AAACCT": 46, + "AAACCC": 47, + "AAACCG": 48, + "AAACGA": 49, + "AAACGT": 50, + "AAACGC": 51, + "AAACGG": 52, + "AAAGAA": 53, + "AAAGAT": 54, + "AAAGAC": 55, + "AAAGAG": 56, + "AAAGTA": 57, + "AAAGTT": 58, + "AAAGTC": 59, + "AAAGTG": 60, + "AAAGCA": 61, + "AAAGCT": 62, + "AAAGCC": 63, + "AAAGCG": 64, + "AAAGGA": 65, + "AAAGGT": 66, + "AAAGGC": 67, + "AAAGGG": 68, + "AATAAA": 69, + "AATAAT": 70, + "AATAAC": 71, + "AATAAG": 72, + "AATATA": 73, + "AATATT": 74, + "AATATC": 75, + "AATATG": 76, + "AATACA": 77, + "AATACT": 78, + "AATACC": 79, + "AATACG": 80, + "AATAGA": 81, + "AATAGT": 82, + "AATAGC": 83, + "AATAGG": 84, + "AATTAA": 85, + "AATTAT": 86, + "AATTAC": 87, + "AATTAG": 88, + "AATTTA": 89, + "AATTTT": 90, + "AATTTC": 91, + "AATTTG": 92, + "AATTCA": 93, + "AATTCT": 94, + "AATTCC": 95, + "AATTCG": 96, + "AATTGA": 97, + "AATTGT": 98, + "AATTGC": 99, + "AATTGG": 100, + "AATCAA": 101, + "AATCAT": 102, + "AATCAC": 103, + "AATCAG": 104, + "AATCTA": 105, + "AATCTT": 106, + "AATCTC": 107, + "AATCTG": 108, + "AATCCA": 109, + "AATCCT": 110, + "AATCCC": 111, + "AATCCG": 112, + "AATCGA": 113, + "AATCGT": 114, + "AATCGC": 115, + "AATCGG": 116, + "AATGAA": 117, + "AATGAT": 118, + "AATGAC": 119, + "AATGAG": 120, + "AATGTA": 121, + "AATGTT": 122, + "AATGTC": 123, + "AATGTG": 124, + "AATGCA": 125, + "AATGCT": 126, + "AATGCC": 127, + "AATGCG": 128, + "AATGGA": 129, + "AATGGT": 130, + "AATGGC": 131, + "AATGGG": 132, + "AACAAA": 133, + "AACAAT": 134, + "AACAAC": 135, + "AACAAG": 136, + "AACATA": 137, + "AACATT": 138, + "AACATC": 139, + "AACATG": 140, + "AACACA": 141, + "AACACT": 142, + "AACACC": 143, + "AACACG": 144, + "AACAGA": 145, + "AACAGT": 146, + "AACAGC": 147, + "AACAGG": 148, + "AACTAA": 149, + "AACTAT": 150, + "AACTAC": 151, + "AACTAG": 152, + "AACTTA": 153, + "AACTTT": 154, + "AACTTC": 155, + "AACTTG": 156, + "AACTCA": 157, + "AACTCT": 158, + "AACTCC": 159, + "AACTCG": 160, + "AACTGA": 161, + "AACTGT": 162, + "AACTGC": 163, + "AACTGG": 164, + "AACCAA": 165, + "AACCAT": 166, + "AACCAC": 167, + "AACCAG": 168, + "AACCTA": 169, + "AACCTT": 170, + "AACCTC": 171, + "AACCTG": 172, + "AACCCA": 173, + "AACCCT": 174, + "AACCCC": 175, + "AACCCG": 176, + "AACCGA": 177, + "AACCGT": 178, + "AACCGC": 179, + "AACCGG": 180, + "AACGAA": 181, + "AACGAT": 182, + "AACGAC": 183, + "AACGAG": 184, + "AACGTA": 185, + "AACGTT": 186, + "AACGTC": 187, + "AACGTG": 188, + "AACGCA": 189, + "AACGCT": 190, + "AACGCC": 191, + "AACGCG": 192, + "AACGGA": 193, + "AACGGT": 194, + "AACGGC": 195, + "AACGGG": 196, + "AAGAAA": 197, + "AAGAAT": 198, + "AAGAAC": 199, + "AAGAAG": 200, + "AAGATA": 201, + "AAGATT": 202, + "AAGATC": 203, + "AAGATG": 204, + "AAGACA": 205, + "AAGACT": 206, + "AAGACC": 207, + "AAGACG": 208, + "AAGAGA": 209, + "AAGAGT": 210, + "AAGAGC": 211, + "AAGAGG": 212, + "AAGTAA": 213, + "AAGTAT": 214, + "AAGTAC": 215, + "AAGTAG": 216, + "AAGTTA": 217, + "AAGTTT": 218, + "AAGTTC": 219, + "AAGTTG": 220, + "AAGTCA": 221, + "AAGTCT": 222, + "AAGTCC": 223, + "AAGTCG": 224, + "AAGTGA": 225, + "AAGTGT": 226, + "AAGTGC": 227, + "AAGTGG": 228, + "AAGCAA": 229, + "AAGCAT": 230, + "AAGCAC": 231, + "AAGCAG": 232, + "AAGCTA": 233, + "AAGCTT": 234, + "AAGCTC": 235, + "AAGCTG": 236, + "AAGCCA": 237, + "AAGCCT": 238, + "AAGCCC": 239, + "AAGCCG": 240, + "AAGCGA": 241, + "AAGCGT": 242, + "AAGCGC": 243, + "AAGCGG": 244, + "AAGGAA": 245, + "AAGGAT": 246, + "AAGGAC": 247, + "AAGGAG": 248, + "AAGGTA": 249, + "AAGGTT": 250, + "AAGGTC": 251, + "AAGGTG": 252, + "AAGGCA": 253, + "AAGGCT": 254, + "AAGGCC": 255, + "AAGGCG": 256, + "AAGGGA": 257, + "AAGGGT": 258, + "AAGGGC": 259, + "AAGGGG": 260, + "ATAAAA": 261, + "ATAAAT": 262, + "ATAAAC": 263, + "ATAAAG": 264, + "ATAATA": 265, + "ATAATT": 266, + "ATAATC": 267, + "ATAATG": 268, + "ATAACA": 269, + "ATAACT": 270, + "ATAACC": 271, + "ATAACG": 272, + "ATAAGA": 273, + "ATAAGT": 274, + "ATAAGC": 275, + "ATAAGG": 276, + "ATATAA": 277, + "ATATAT": 278, + "ATATAC": 279, + "ATATAG": 280, + "ATATTA": 281, + "ATATTT": 282, + "ATATTC": 283, + "ATATTG": 284, + "ATATCA": 285, + "ATATCT": 286, + "ATATCC": 287, + "ATATCG": 288, + "ATATGA": 289, + "ATATGT": 290, + "ATATGC": 291, + "ATATGG": 292, + "ATACAA": 293, + "ATACAT": 294, + "ATACAC": 295, + "ATACAG": 296, + "ATACTA": 297, + "ATACTT": 298, + "ATACTC": 299, + "ATACTG": 300, + "ATACCA": 301, + "ATACCT": 302, + "ATACCC": 303, + "ATACCG": 304, + "ATACGA": 305, + "ATACGT": 306, + "ATACGC": 307, + "ATACGG": 308, + "ATAGAA": 309, + "ATAGAT": 310, + "ATAGAC": 311, + "ATAGAG": 312, + "ATAGTA": 313, + "ATAGTT": 314, + "ATAGTC": 315, + "ATAGTG": 316, + "ATAGCA": 317, + "ATAGCT": 318, + "ATAGCC": 319, + "ATAGCG": 320, + "ATAGGA": 321, + "ATAGGT": 322, + "ATAGGC": 323, + "ATAGGG": 324, + "ATTAAA": 325, + "ATTAAT": 326, + "ATTAAC": 327, + "ATTAAG": 328, + "ATTATA": 329, + "ATTATT": 330, + "ATTATC": 331, + "ATTATG": 332, + "ATTACA": 333, + "ATTACT": 334, + "ATTACC": 335, + "ATTACG": 336, + "ATTAGA": 337, + "ATTAGT": 338, + "ATTAGC": 339, + "ATTAGG": 340, + "ATTTAA": 341, + "ATTTAT": 342, + "ATTTAC": 343, + "ATTTAG": 344, + "ATTTTA": 345, + "ATTTTT": 346, + "ATTTTC": 347, + "ATTTTG": 348, + "ATTTCA": 349, + "ATTTCT": 350, + "ATTTCC": 351, + "ATTTCG": 352, + "ATTTGA": 353, + "ATTTGT": 354, + "ATTTGC": 355, + "ATTTGG": 356, + "ATTCAA": 357, + "ATTCAT": 358, + "ATTCAC": 359, + "ATTCAG": 360, + "ATTCTA": 361, + "ATTCTT": 362, + "ATTCTC": 363, + "ATTCTG": 364, + "ATTCCA": 365, + "ATTCCT": 366, + "ATTCCC": 367, + "ATTCCG": 368, + "ATTCGA": 369, + "ATTCGT": 370, + "ATTCGC": 371, + "ATTCGG": 372, + "ATTGAA": 373, + "ATTGAT": 374, + "ATTGAC": 375, + "ATTGAG": 376, + "ATTGTA": 377, + "ATTGTT": 378, + "ATTGTC": 379, + "ATTGTG": 380, + "ATTGCA": 381, + "ATTGCT": 382, + "ATTGCC": 383, + "ATTGCG": 384, + "ATTGGA": 385, + "ATTGGT": 386, + "ATTGGC": 387, + "ATTGGG": 388, + "ATCAAA": 389, + "ATCAAT": 390, + "ATCAAC": 391, + "ATCAAG": 392, + "ATCATA": 393, + "ATCATT": 394, + "ATCATC": 395, + "ATCATG": 396, + "ATCACA": 397, + "ATCACT": 398, + "ATCACC": 399, + "ATCACG": 400, + "ATCAGA": 401, + "ATCAGT": 402, + "ATCAGC": 403, + "ATCAGG": 404, + "ATCTAA": 405, + "ATCTAT": 406, + "ATCTAC": 407, + "ATCTAG": 408, + "ATCTTA": 409, + "ATCTTT": 410, + "ATCTTC": 411, + "ATCTTG": 412, + "ATCTCA": 413, + "ATCTCT": 414, + "ATCTCC": 415, + "ATCTCG": 416, + "ATCTGA": 417, + "ATCTGT": 418, + "ATCTGC": 419, + "ATCTGG": 420, + "ATCCAA": 421, + "ATCCAT": 422, + "ATCCAC": 423, + "ATCCAG": 424, + "ATCCTA": 425, + "ATCCTT": 426, + "ATCCTC": 427, + "ATCCTG": 428, + "ATCCCA": 429, + "ATCCCT": 430, + "ATCCCC": 431, + "ATCCCG": 432, + "ATCCGA": 433, + "ATCCGT": 434, + "ATCCGC": 435, + "ATCCGG": 436, + "ATCGAA": 437, + "ATCGAT": 438, + "ATCGAC": 439, + "ATCGAG": 440, + "ATCGTA": 441, + "ATCGTT": 442, + "ATCGTC": 443, + "ATCGTG": 444, + "ATCGCA": 445, + "ATCGCT": 446, + "ATCGCC": 447, + "ATCGCG": 448, + "ATCGGA": 449, + "ATCGGT": 450, + "ATCGGC": 451, + "ATCGGG": 452, + "ATGAAA": 453, + "ATGAAT": 454, + "ATGAAC": 455, + "ATGAAG": 456, + "ATGATA": 457, + "ATGATT": 458, + "ATGATC": 459, + "ATGATG": 460, + "ATGACA": 461, + "ATGACT": 462, + "ATGACC": 463, + "ATGACG": 464, + "ATGAGA": 465, + "ATGAGT": 466, + "ATGAGC": 467, + "ATGAGG": 468, + "ATGTAA": 469, + "ATGTAT": 470, + "ATGTAC": 471, + "ATGTAG": 472, + "ATGTTA": 473, + "ATGTTT": 474, + "ATGTTC": 475, + "ATGTTG": 476, + "ATGTCA": 477, + "ATGTCT": 478, + "ATGTCC": 479, + "ATGTCG": 480, + "ATGTGA": 481, + "ATGTGT": 482, + "ATGTGC": 483, + "ATGTGG": 484, + "ATGCAA": 485, + "ATGCAT": 486, + "ATGCAC": 487, + "ATGCAG": 488, + "ATGCTA": 489, + "ATGCTT": 490, + "ATGCTC": 491, + "ATGCTG": 492, + "ATGCCA": 493, + "ATGCCT": 494, + "ATGCCC": 495, + "ATGCCG": 496, + "ATGCGA": 497, + "ATGCGT": 498, + "ATGCGC": 499, + "ATGCGG": 500, + "ATGGAA": 501, + "ATGGAT": 502, + "ATGGAC": 503, + "ATGGAG": 504, + "ATGGTA": 505, + "ATGGTT": 506, + "ATGGTC": 507, + "ATGGTG": 508, + "ATGGCA": 509, + "ATGGCT": 510, + "ATGGCC": 511, + "ATGGCG": 512, + "ATGGGA": 513, + "ATGGGT": 514, + "ATGGGC": 515, + "ATGGGG": 516, + "ACAAAA": 517, + "ACAAAT": 518, + "ACAAAC": 519, + "ACAAAG": 520, + "ACAATA": 521, + "ACAATT": 522, + "ACAATC": 523, + "ACAATG": 524, + "ACAACA": 525, + "ACAACT": 526, + "ACAACC": 527, + "ACAACG": 528, + "ACAAGA": 529, + "ACAAGT": 530, + "ACAAGC": 531, + "ACAAGG": 532, + "ACATAA": 533, + "ACATAT": 534, + "ACATAC": 535, + "ACATAG": 536, + "ACATTA": 537, + "ACATTT": 538, + "ACATTC": 539, + "ACATTG": 540, + "ACATCA": 541, + "ACATCT": 542, + "ACATCC": 543, + "ACATCG": 544, + "ACATGA": 545, + "ACATGT": 546, + "ACATGC": 547, + "ACATGG": 548, + "ACACAA": 549, + "ACACAT": 550, + "ACACAC": 551, + "ACACAG": 552, + "ACACTA": 553, + "ACACTT": 554, + "ACACTC": 555, + "ACACTG": 556, + "ACACCA": 557, + "ACACCT": 558, + "ACACCC": 559, + "ACACCG": 560, + "ACACGA": 561, + "ACACGT": 562, + "ACACGC": 563, + "ACACGG": 564, + "ACAGAA": 565, + "ACAGAT": 566, + "ACAGAC": 567, + "ACAGAG": 568, + "ACAGTA": 569, + "ACAGTT": 570, + "ACAGTC": 571, + "ACAGTG": 572, + "ACAGCA": 573, + "ACAGCT": 574, + "ACAGCC": 575, + "ACAGCG": 576, + "ACAGGA": 577, + "ACAGGT": 578, + "ACAGGC": 579, + "ACAGGG": 580, + "ACTAAA": 581, + "ACTAAT": 582, + "ACTAAC": 583, + "ACTAAG": 584, + "ACTATA": 585, + "ACTATT": 586, + "ACTATC": 587, + "ACTATG": 588, + "ACTACA": 589, + "ACTACT": 590, + "ACTACC": 591, + "ACTACG": 592, + "ACTAGA": 593, + "ACTAGT": 594, + "ACTAGC": 595, + "ACTAGG": 596, + "ACTTAA": 597, + "ACTTAT": 598, + "ACTTAC": 599, + "ACTTAG": 600, + "ACTTTA": 601, + "ACTTTT": 602, + "ACTTTC": 603, + "ACTTTG": 604, + "ACTTCA": 605, + "ACTTCT": 606, + "ACTTCC": 607, + "ACTTCG": 608, + "ACTTGA": 609, + "ACTTGT": 610, + "ACTTGC": 611, + "ACTTGG": 612, + "ACTCAA": 613, + "ACTCAT": 614, + "ACTCAC": 615, + "ACTCAG": 616, + "ACTCTA": 617, + "ACTCTT": 618, + "ACTCTC": 619, + "ACTCTG": 620, + "ACTCCA": 621, + "ACTCCT": 622, + "ACTCCC": 623, + "ACTCCG": 624, + "ACTCGA": 625, + "ACTCGT": 626, + "ACTCGC": 627, + "ACTCGG": 628, + "ACTGAA": 629, + "ACTGAT": 630, + "ACTGAC": 631, + "ACTGAG": 632, + "ACTGTA": 633, + "ACTGTT": 634, + "ACTGTC": 635, + "ACTGTG": 636, + "ACTGCA": 637, + "ACTGCT": 638, + "ACTGCC": 639, + "ACTGCG": 640, + "ACTGGA": 641, + "ACTGGT": 642, + "ACTGGC": 643, + "ACTGGG": 644, + "ACCAAA": 645, + "ACCAAT": 646, + "ACCAAC": 647, + "ACCAAG": 648, + "ACCATA": 649, + "ACCATT": 650, + "ACCATC": 651, + "ACCATG": 652, + "ACCACA": 653, + "ACCACT": 654, + "ACCACC": 655, + "ACCACG": 656, + "ACCAGA": 657, + "ACCAGT": 658, + "ACCAGC": 659, + "ACCAGG": 660, + "ACCTAA": 661, + "ACCTAT": 662, + "ACCTAC": 663, + "ACCTAG": 664, + "ACCTTA": 665, + "ACCTTT": 666, + "ACCTTC": 667, + "ACCTTG": 668, + "ACCTCA": 669, + "ACCTCT": 670, + "ACCTCC": 671, + "ACCTCG": 672, + "ACCTGA": 673, + "ACCTGT": 674, + "ACCTGC": 675, + "ACCTGG": 676, + "ACCCAA": 677, + "ACCCAT": 678, + "ACCCAC": 679, + "ACCCAG": 680, + "ACCCTA": 681, + "ACCCTT": 682, + "ACCCTC": 683, + "ACCCTG": 684, + "ACCCCA": 685, + "ACCCCT": 686, + "ACCCCC": 687, + "ACCCCG": 688, + "ACCCGA": 689, + "ACCCGT": 690, + "ACCCGC": 691, + "ACCCGG": 692, + "ACCGAA": 693, + "ACCGAT": 694, + "ACCGAC": 695, + "ACCGAG": 696, + "ACCGTA": 697, + "ACCGTT": 698, + "ACCGTC": 699, + "ACCGTG": 700, + "ACCGCA": 701, + "ACCGCT": 702, + "ACCGCC": 703, + "ACCGCG": 704, + "ACCGGA": 705, + "ACCGGT": 706, + "ACCGGC": 707, + "ACCGGG": 708, + "ACGAAA": 709, + "ACGAAT": 710, + "ACGAAC": 711, + "ACGAAG": 712, + "ACGATA": 713, + "ACGATT": 714, + "ACGATC": 715, + "ACGATG": 716, + "ACGACA": 717, + "ACGACT": 718, + "ACGACC": 719, + "ACGACG": 720, + "ACGAGA": 721, + "ACGAGT": 722, + "ACGAGC": 723, + "ACGAGG": 724, + "ACGTAA": 725, + "ACGTAT": 726, + "ACGTAC": 727, + "ACGTAG": 728, + "ACGTTA": 729, + "ACGTTT": 730, + "ACGTTC": 731, + "ACGTTG": 732, + "ACGTCA": 733, + "ACGTCT": 734, + "ACGTCC": 735, + "ACGTCG": 736, + "ACGTGA": 737, + "ACGTGT": 738, + "ACGTGC": 739, + "ACGTGG": 740, + "ACGCAA": 741, + "ACGCAT": 742, + "ACGCAC": 743, + "ACGCAG": 744, + "ACGCTA": 745, + "ACGCTT": 746, + "ACGCTC": 747, + "ACGCTG": 748, + "ACGCCA": 749, + "ACGCCT": 750, + "ACGCCC": 751, + "ACGCCG": 752, + "ACGCGA": 753, + "ACGCGT": 754, + "ACGCGC": 755, + "ACGCGG": 756, + "ACGGAA": 757, + "ACGGAT": 758, + "ACGGAC": 759, + "ACGGAG": 760, + "ACGGTA": 761, + "ACGGTT": 762, + "ACGGTC": 763, + "ACGGTG": 764, + "ACGGCA": 765, + "ACGGCT": 766, + "ACGGCC": 767, + "ACGGCG": 768, + "ACGGGA": 769, + "ACGGGT": 770, + "ACGGGC": 771, + "ACGGGG": 772, + "AGAAAA": 773, + "AGAAAT": 774, + "AGAAAC": 775, + "AGAAAG": 776, + "AGAATA": 777, + "AGAATT": 778, + "AGAATC": 779, + "AGAATG": 780, + "AGAACA": 781, + "AGAACT": 782, + "AGAACC": 783, + "AGAACG": 784, + "AGAAGA": 785, + "AGAAGT": 786, + "AGAAGC": 787, + "AGAAGG": 788, + "AGATAA": 789, + "AGATAT": 790, + "AGATAC": 791, + "AGATAG": 792, + "AGATTA": 793, + "AGATTT": 794, + "AGATTC": 795, + "AGATTG": 796, + "AGATCA": 797, + "AGATCT": 798, + "AGATCC": 799, + "AGATCG": 800, + "AGATGA": 801, + "AGATGT": 802, + "AGATGC": 803, + "AGATGG": 804, + "AGACAA": 805, + "AGACAT": 806, + "AGACAC": 807, + "AGACAG": 808, + "AGACTA": 809, + "AGACTT": 810, + "AGACTC": 811, + "AGACTG": 812, + "AGACCA": 813, + "AGACCT": 814, + "AGACCC": 815, + "AGACCG": 816, + "AGACGA": 817, + "AGACGT": 818, + "AGACGC": 819, + "AGACGG": 820, + "AGAGAA": 821, + "AGAGAT": 822, + "AGAGAC": 823, + "AGAGAG": 824, + "AGAGTA": 825, + "AGAGTT": 826, + "AGAGTC": 827, + "AGAGTG": 828, + "AGAGCA": 829, + "AGAGCT": 830, + "AGAGCC": 831, + "AGAGCG": 832, + "AGAGGA": 833, + "AGAGGT": 834, + "AGAGGC": 835, + "AGAGGG": 836, + "AGTAAA": 837, + "AGTAAT": 838, + "AGTAAC": 839, + "AGTAAG": 840, + "AGTATA": 841, + "AGTATT": 842, + "AGTATC": 843, + "AGTATG": 844, + "AGTACA": 845, + "AGTACT": 846, + "AGTACC": 847, + "AGTACG": 848, + "AGTAGA": 849, + "AGTAGT": 850, + "AGTAGC": 851, + "AGTAGG": 852, + "AGTTAA": 853, + "AGTTAT": 854, + "AGTTAC": 855, + "AGTTAG": 856, + "AGTTTA": 857, + "AGTTTT": 858, + "AGTTTC": 859, + "AGTTTG": 860, + "AGTTCA": 861, + "AGTTCT": 862, + "AGTTCC": 863, + "AGTTCG": 864, + "AGTTGA": 865, + "AGTTGT": 866, + "AGTTGC": 867, + "AGTTGG": 868, + "AGTCAA": 869, + "AGTCAT": 870, + "AGTCAC": 871, + "AGTCAG": 872, + "AGTCTA": 873, + "AGTCTT": 874, + "AGTCTC": 875, + "AGTCTG": 876, + "AGTCCA": 877, + "AGTCCT": 878, + "AGTCCC": 879, + "AGTCCG": 880, + "AGTCGA": 881, + "AGTCGT": 882, + "AGTCGC": 883, + "AGTCGG": 884, + "AGTGAA": 885, + "AGTGAT": 886, + "AGTGAC": 887, + "AGTGAG": 888, + "AGTGTA": 889, + "AGTGTT": 890, + "AGTGTC": 891, + "AGTGTG": 892, + "AGTGCA": 893, + "AGTGCT": 894, + "AGTGCC": 895, + "AGTGCG": 896, + "AGTGGA": 897, + "AGTGGT": 898, + "AGTGGC": 899, + "AGTGGG": 900, + "AGCAAA": 901, + "AGCAAT": 902, + "AGCAAC": 903, + "AGCAAG": 904, + "AGCATA": 905, + "AGCATT": 906, + "AGCATC": 907, + "AGCATG": 908, + "AGCACA": 909, + "AGCACT": 910, + "AGCACC": 911, + "AGCACG": 912, + "AGCAGA": 913, + "AGCAGT": 914, + "AGCAGC": 915, + "AGCAGG": 916, + "AGCTAA": 917, + "AGCTAT": 918, + "AGCTAC": 919, + "AGCTAG": 920, + "AGCTTA": 921, + "AGCTTT": 922, + "AGCTTC": 923, + "AGCTTG": 924, + "AGCTCA": 925, + "AGCTCT": 926, + "AGCTCC": 927, + "AGCTCG": 928, + "AGCTGA": 929, + "AGCTGT": 930, + "AGCTGC": 931, + "AGCTGG": 932, + "AGCCAA": 933, + "AGCCAT": 934, + "AGCCAC": 935, + "AGCCAG": 936, + "AGCCTA": 937, + "AGCCTT": 938, + "AGCCTC": 939, + "AGCCTG": 940, + "AGCCCA": 941, + "AGCCCT": 942, + "AGCCCC": 943, + "AGCCCG": 944, + "AGCCGA": 945, + "AGCCGT": 946, + "AGCCGC": 947, + "AGCCGG": 948, + "AGCGAA": 949, + "AGCGAT": 950, + "AGCGAC": 951, + "AGCGAG": 952, + "AGCGTA": 953, + "AGCGTT": 954, + "AGCGTC": 955, + "AGCGTG": 956, + "AGCGCA": 957, + "AGCGCT": 958, + "AGCGCC": 959, + "AGCGCG": 960, + "AGCGGA": 961, + "AGCGGT": 962, + "AGCGGC": 963, + "AGCGGG": 964, + "AGGAAA": 965, + "AGGAAT": 966, + "AGGAAC": 967, + "AGGAAG": 968, + "AGGATA": 969, + "AGGATT": 970, + "AGGATC": 971, + "AGGATG": 972, + "AGGACA": 973, + "AGGACT": 974, + "AGGACC": 975, + "AGGACG": 976, + "AGGAGA": 977, + "AGGAGT": 978, + "AGGAGC": 979, + "AGGAGG": 980, + "AGGTAA": 981, + "AGGTAT": 982, + "AGGTAC": 983, + "AGGTAG": 984, + "AGGTTA": 985, + "AGGTTT": 986, + "AGGTTC": 987, + "AGGTTG": 988, + "AGGTCA": 989, + "AGGTCT": 990, + "AGGTCC": 991, + "AGGTCG": 992, + "AGGTGA": 993, + "AGGTGT": 994, + "AGGTGC": 995, + "AGGTGG": 996, + "AGGCAA": 997, + "AGGCAT": 998, + "AGGCAC": 999, + "AGGCAG": 1000, + "AGGCTA": 1001, + "AGGCTT": 1002, + "AGGCTC": 1003, + "AGGCTG": 1004, + "AGGCCA": 1005, + "AGGCCT": 1006, + "AGGCCC": 1007, + "AGGCCG": 1008, + "AGGCGA": 1009, + "AGGCGT": 1010, + "AGGCGC": 1011, + "AGGCGG": 1012, + "AGGGAA": 1013, + "AGGGAT": 1014, + "AGGGAC": 1015, + "AGGGAG": 1016, + "AGGGTA": 1017, + "AGGGTT": 1018, + "AGGGTC": 1019, + "AGGGTG": 1020, + "AGGGCA": 1021, + "AGGGCT": 1022, + "AGGGCC": 1023, + "AGGGCG": 1024, + "AGGGGA": 1025, + "AGGGGT": 1026, + "AGGGGC": 1027, + "AGGGGG": 1028, + "TAAAAA": 1029, + "TAAAAT": 1030, + "TAAAAC": 1031, + "TAAAAG": 1032, + "TAAATA": 1033, + "TAAATT": 1034, + "TAAATC": 1035, + "TAAATG": 1036, + "TAAACA": 1037, + "TAAACT": 1038, + "TAAACC": 1039, + "TAAACG": 1040, + "TAAAGA": 1041, + "TAAAGT": 1042, + "TAAAGC": 1043, + "TAAAGG": 1044, + "TAATAA": 1045, + "TAATAT": 1046, + "TAATAC": 1047, + "TAATAG": 1048, + "TAATTA": 1049, + "TAATTT": 1050, + "TAATTC": 1051, + "TAATTG": 1052, + "TAATCA": 1053, + "TAATCT": 1054, + "TAATCC": 1055, + "TAATCG": 1056, + "TAATGA": 1057, + "TAATGT": 1058, + "TAATGC": 1059, + "TAATGG": 1060, + "TAACAA": 1061, + "TAACAT": 1062, + "TAACAC": 1063, + "TAACAG": 1064, + "TAACTA": 1065, + "TAACTT": 1066, + "TAACTC": 1067, + "TAACTG": 1068, + "TAACCA": 1069, + "TAACCT": 1070, + "TAACCC": 1071, + "TAACCG": 1072, + "TAACGA": 1073, + "TAACGT": 1074, + "TAACGC": 1075, + "TAACGG": 1076, + "TAAGAA": 1077, + "TAAGAT": 1078, + "TAAGAC": 1079, + "TAAGAG": 1080, + "TAAGTA": 1081, + "TAAGTT": 1082, + "TAAGTC": 1083, + "TAAGTG": 1084, + "TAAGCA": 1085, + "TAAGCT": 1086, + "TAAGCC": 1087, + "TAAGCG": 1088, + "TAAGGA": 1089, + "TAAGGT": 1090, + "TAAGGC": 1091, + "TAAGGG": 1092, + "TATAAA": 1093, + "TATAAT": 1094, + "TATAAC": 1095, + "TATAAG": 1096, + "TATATA": 1097, + "TATATT": 1098, + "TATATC": 1099, + "TATATG": 1100, + "TATACA": 1101, + "TATACT": 1102, + "TATACC": 1103, + "TATACG": 1104, + "TATAGA": 1105, + "TATAGT": 1106, + "TATAGC": 1107, + "TATAGG": 1108, + "TATTAA": 1109, + "TATTAT": 1110, + "TATTAC": 1111, + "TATTAG": 1112, + "TATTTA": 1113, + "TATTTT": 1114, + "TATTTC": 1115, + "TATTTG": 1116, + "TATTCA": 1117, + "TATTCT": 1118, + "TATTCC": 1119, + "TATTCG": 1120, + "TATTGA": 1121, + "TATTGT": 1122, + "TATTGC": 1123, + "TATTGG": 1124, + "TATCAA": 1125, + "TATCAT": 1126, + "TATCAC": 1127, + "TATCAG": 1128, + "TATCTA": 1129, + "TATCTT": 1130, + "TATCTC": 1131, + "TATCTG": 1132, + "TATCCA": 1133, + "TATCCT": 1134, + "TATCCC": 1135, + "TATCCG": 1136, + "TATCGA": 1137, + "TATCGT": 1138, + "TATCGC": 1139, + "TATCGG": 1140, + "TATGAA": 1141, + "TATGAT": 1142, + "TATGAC": 1143, + "TATGAG": 1144, + "TATGTA": 1145, + "TATGTT": 1146, + "TATGTC": 1147, + "TATGTG": 1148, + "TATGCA": 1149, + "TATGCT": 1150, + "TATGCC": 1151, + "TATGCG": 1152, + "TATGGA": 1153, + "TATGGT": 1154, + "TATGGC": 1155, + "TATGGG": 1156, + "TACAAA": 1157, + "TACAAT": 1158, + "TACAAC": 1159, + "TACAAG": 1160, + "TACATA": 1161, + "TACATT": 1162, + "TACATC": 1163, + "TACATG": 1164, + "TACACA": 1165, + "TACACT": 1166, + "TACACC": 1167, + "TACACG": 1168, + "TACAGA": 1169, + "TACAGT": 1170, + "TACAGC": 1171, + "TACAGG": 1172, + "TACTAA": 1173, + "TACTAT": 1174, + "TACTAC": 1175, + "TACTAG": 1176, + "TACTTA": 1177, + "TACTTT": 1178, + "TACTTC": 1179, + "TACTTG": 1180, + "TACTCA": 1181, + "TACTCT": 1182, + "TACTCC": 1183, + "TACTCG": 1184, + "TACTGA": 1185, + "TACTGT": 1186, + "TACTGC": 1187, + "TACTGG": 1188, + "TACCAA": 1189, + "TACCAT": 1190, + "TACCAC": 1191, + "TACCAG": 1192, + "TACCTA": 1193, + "TACCTT": 1194, + "TACCTC": 1195, + "TACCTG": 1196, + "TACCCA": 1197, + "TACCCT": 1198, + "TACCCC": 1199, + "TACCCG": 1200, + "TACCGA": 1201, + "TACCGT": 1202, + "TACCGC": 1203, + "TACCGG": 1204, + "TACGAA": 1205, + "TACGAT": 1206, + "TACGAC": 1207, + "TACGAG": 1208, + "TACGTA": 1209, + "TACGTT": 1210, + "TACGTC": 1211, + "TACGTG": 1212, + "TACGCA": 1213, + "TACGCT": 1214, + "TACGCC": 1215, + "TACGCG": 1216, + "TACGGA": 1217, + "TACGGT": 1218, + "TACGGC": 1219, + "TACGGG": 1220, + "TAGAAA": 1221, + "TAGAAT": 1222, + "TAGAAC": 1223, + "TAGAAG": 1224, + "TAGATA": 1225, + "TAGATT": 1226, + "TAGATC": 1227, + "TAGATG": 1228, + "TAGACA": 1229, + "TAGACT": 1230, + "TAGACC": 1231, + "TAGACG": 1232, + "TAGAGA": 1233, + "TAGAGT": 1234, + "TAGAGC": 1235, + "TAGAGG": 1236, + "TAGTAA": 1237, + "TAGTAT": 1238, + "TAGTAC": 1239, + "TAGTAG": 1240, + "TAGTTA": 1241, + "TAGTTT": 1242, + "TAGTTC": 1243, + "TAGTTG": 1244, + "TAGTCA": 1245, + "TAGTCT": 1246, + "TAGTCC": 1247, + "TAGTCG": 1248, + "TAGTGA": 1249, + "TAGTGT": 1250, + "TAGTGC": 1251, + "TAGTGG": 1252, + "TAGCAA": 1253, + "TAGCAT": 1254, + "TAGCAC": 1255, + "TAGCAG": 1256, + "TAGCTA": 1257, + "TAGCTT": 1258, + "TAGCTC": 1259, + "TAGCTG": 1260, + "TAGCCA": 1261, + "TAGCCT": 1262, + "TAGCCC": 1263, + "TAGCCG": 1264, + "TAGCGA": 1265, + "TAGCGT": 1266, + "TAGCGC": 1267, + "TAGCGG": 1268, + "TAGGAA": 1269, + "TAGGAT": 1270, + "TAGGAC": 1271, + "TAGGAG": 1272, + "TAGGTA": 1273, + "TAGGTT": 1274, + "TAGGTC": 1275, + "TAGGTG": 1276, + "TAGGCA": 1277, + "TAGGCT": 1278, + "TAGGCC": 1279, + "TAGGCG": 1280, + "TAGGGA": 1281, + "TAGGGT": 1282, + "TAGGGC": 1283, + "TAGGGG": 1284, + "TTAAAA": 1285, + "TTAAAT": 1286, + "TTAAAC": 1287, + "TTAAAG": 1288, + "TTAATA": 1289, + "TTAATT": 1290, + "TTAATC": 1291, + "TTAATG": 1292, + "TTAACA": 1293, + "TTAACT": 1294, + "TTAACC": 1295, + "TTAACG": 1296, + "TTAAGA": 1297, + "TTAAGT": 1298, + "TTAAGC": 1299, + "TTAAGG": 1300, + "TTATAA": 1301, + "TTATAT": 1302, + "TTATAC": 1303, + "TTATAG": 1304, + "TTATTA": 1305, + "TTATTT": 1306, + "TTATTC": 1307, + "TTATTG": 1308, + "TTATCA": 1309, + "TTATCT": 1310, + "TTATCC": 1311, + "TTATCG": 1312, + "TTATGA": 1313, + "TTATGT": 1314, + "TTATGC": 1315, + "TTATGG": 1316, + "TTACAA": 1317, + "TTACAT": 1318, + "TTACAC": 1319, + "TTACAG": 1320, + "TTACTA": 1321, + "TTACTT": 1322, + "TTACTC": 1323, + "TTACTG": 1324, + "TTACCA": 1325, + "TTACCT": 1326, + "TTACCC": 1327, + "TTACCG": 1328, + "TTACGA": 1329, + "TTACGT": 1330, + "TTACGC": 1331, + "TTACGG": 1332, + "TTAGAA": 1333, + "TTAGAT": 1334, + "TTAGAC": 1335, + "TTAGAG": 1336, + "TTAGTA": 1337, + "TTAGTT": 1338, + "TTAGTC": 1339, + "TTAGTG": 1340, + "TTAGCA": 1341, + "TTAGCT": 1342, + "TTAGCC": 1343, + "TTAGCG": 1344, + "TTAGGA": 1345, + "TTAGGT": 1346, + "TTAGGC": 1347, + "TTAGGG": 1348, + "TTTAAA": 1349, + "TTTAAT": 1350, + "TTTAAC": 1351, + "TTTAAG": 1352, + "TTTATA": 1353, + "TTTATT": 1354, + "TTTATC": 1355, + "TTTATG": 1356, + "TTTACA": 1357, + "TTTACT": 1358, + "TTTACC": 1359, + "TTTACG": 1360, + "TTTAGA": 1361, + "TTTAGT": 1362, + "TTTAGC": 1363, + "TTTAGG": 1364, + "TTTTAA": 1365, + "TTTTAT": 1366, + "TTTTAC": 1367, + "TTTTAG": 1368, + "TTTTTA": 1369, + "TTTTTT": 1370, + "TTTTTC": 1371, + "TTTTTG": 1372, + "TTTTCA": 1373, + "TTTTCT": 1374, + "TTTTCC": 1375, + "TTTTCG": 1376, + "TTTTGA": 1377, + "TTTTGT": 1378, + "TTTTGC": 1379, + "TTTTGG": 1380, + "TTTCAA": 1381, + "TTTCAT": 1382, + "TTTCAC": 1383, + "TTTCAG": 1384, + "TTTCTA": 1385, + "TTTCTT": 1386, + "TTTCTC": 1387, + "TTTCTG": 1388, + "TTTCCA": 1389, + "TTTCCT": 1390, + "TTTCCC": 1391, + "TTTCCG": 1392, + "TTTCGA": 1393, + "TTTCGT": 1394, + "TTTCGC": 1395, + "TTTCGG": 1396, + "TTTGAA": 1397, + "TTTGAT": 1398, + "TTTGAC": 1399, + "TTTGAG": 1400, + "TTTGTA": 1401, + "TTTGTT": 1402, + "TTTGTC": 1403, + "TTTGTG": 1404, + "TTTGCA": 1405, + "TTTGCT": 1406, + "TTTGCC": 1407, + "TTTGCG": 1408, + "TTTGGA": 1409, + "TTTGGT": 1410, + "TTTGGC": 1411, + "TTTGGG": 1412, + "TTCAAA": 1413, + "TTCAAT": 1414, + "TTCAAC": 1415, + "TTCAAG": 1416, + "TTCATA": 1417, + "TTCATT": 1418, + "TTCATC": 1419, + "TTCATG": 1420, + "TTCACA": 1421, + "TTCACT": 1422, + "TTCACC": 1423, + "TTCACG": 1424, + "TTCAGA": 1425, + "TTCAGT": 1426, + "TTCAGC": 1427, + "TTCAGG": 1428, + "TTCTAA": 1429, + "TTCTAT": 1430, + "TTCTAC": 1431, + "TTCTAG": 1432, + "TTCTTA": 1433, + "TTCTTT": 1434, + "TTCTTC": 1435, + "TTCTTG": 1436, + "TTCTCA": 1437, + "TTCTCT": 1438, + "TTCTCC": 1439, + "TTCTCG": 1440, + "TTCTGA": 1441, + "TTCTGT": 1442, + "TTCTGC": 1443, + "TTCTGG": 1444, + "TTCCAA": 1445, + "TTCCAT": 1446, + "TTCCAC": 1447, + "TTCCAG": 1448, + "TTCCTA": 1449, + "TTCCTT": 1450, + "TTCCTC": 1451, + "TTCCTG": 1452, + "TTCCCA": 1453, + "TTCCCT": 1454, + "TTCCCC": 1455, + "TTCCCG": 1456, + "TTCCGA": 1457, + "TTCCGT": 1458, + "TTCCGC": 1459, + "TTCCGG": 1460, + "TTCGAA": 1461, + "TTCGAT": 1462, + "TTCGAC": 1463, + "TTCGAG": 1464, + "TTCGTA": 1465, + "TTCGTT": 1466, + "TTCGTC": 1467, + "TTCGTG": 1468, + "TTCGCA": 1469, + "TTCGCT": 1470, + "TTCGCC": 1471, + "TTCGCG": 1472, + "TTCGGA": 1473, + "TTCGGT": 1474, + "TTCGGC": 1475, + "TTCGGG": 1476, + "TTGAAA": 1477, + "TTGAAT": 1478, + "TTGAAC": 1479, + "TTGAAG": 1480, + "TTGATA": 1481, + "TTGATT": 1482, + "TTGATC": 1483, + "TTGATG": 1484, + "TTGACA": 1485, + "TTGACT": 1486, + "TTGACC": 1487, + "TTGACG": 1488, + "TTGAGA": 1489, + "TTGAGT": 1490, + "TTGAGC": 1491, + "TTGAGG": 1492, + "TTGTAA": 1493, + "TTGTAT": 1494, + "TTGTAC": 1495, + "TTGTAG": 1496, + "TTGTTA": 1497, + "TTGTTT": 1498, + "TTGTTC": 1499, + "TTGTTG": 1500, + "TTGTCA": 1501, + "TTGTCT": 1502, + "TTGTCC": 1503, + "TTGTCG": 1504, + "TTGTGA": 1505, + "TTGTGT": 1506, + "TTGTGC": 1507, + "TTGTGG": 1508, + "TTGCAA": 1509, + "TTGCAT": 1510, + "TTGCAC": 1511, + "TTGCAG": 1512, + "TTGCTA": 1513, + "TTGCTT": 1514, + "TTGCTC": 1515, + "TTGCTG": 1516, + "TTGCCA": 1517, + "TTGCCT": 1518, + "TTGCCC": 1519, + "TTGCCG": 1520, + "TTGCGA": 1521, + "TTGCGT": 1522, + "TTGCGC": 1523, + "TTGCGG": 1524, + "TTGGAA": 1525, + "TTGGAT": 1526, + "TTGGAC": 1527, + "TTGGAG": 1528, + "TTGGTA": 1529, + "TTGGTT": 1530, + "TTGGTC": 1531, + "TTGGTG": 1532, + "TTGGCA": 1533, + "TTGGCT": 1534, + "TTGGCC": 1535, + "TTGGCG": 1536, + "TTGGGA": 1537, + "TTGGGT": 1538, + "TTGGGC": 1539, + "TTGGGG": 1540, + "TCAAAA": 1541, + "TCAAAT": 1542, + "TCAAAC": 1543, + "TCAAAG": 1544, + "TCAATA": 1545, + "TCAATT": 1546, + "TCAATC": 1547, + "TCAATG": 1548, + "TCAACA": 1549, + "TCAACT": 1550, + "TCAACC": 1551, + "TCAACG": 1552, + "TCAAGA": 1553, + "TCAAGT": 1554, + "TCAAGC": 1555, + "TCAAGG": 1556, + "TCATAA": 1557, + "TCATAT": 1558, + "TCATAC": 1559, + "TCATAG": 1560, + "TCATTA": 1561, + "TCATTT": 1562, + "TCATTC": 1563, + "TCATTG": 1564, + "TCATCA": 1565, + "TCATCT": 1566, + "TCATCC": 1567, + "TCATCG": 1568, + "TCATGA": 1569, + "TCATGT": 1570, + "TCATGC": 1571, + "TCATGG": 1572, + "TCACAA": 1573, + "TCACAT": 1574, + "TCACAC": 1575, + "TCACAG": 1576, + "TCACTA": 1577, + "TCACTT": 1578, + "TCACTC": 1579, + "TCACTG": 1580, + "TCACCA": 1581, + "TCACCT": 1582, + "TCACCC": 1583, + "TCACCG": 1584, + "TCACGA": 1585, + "TCACGT": 1586, + "TCACGC": 1587, + "TCACGG": 1588, + "TCAGAA": 1589, + "TCAGAT": 1590, + "TCAGAC": 1591, + "TCAGAG": 1592, + "TCAGTA": 1593, + "TCAGTT": 1594, + "TCAGTC": 1595, + "TCAGTG": 1596, + "TCAGCA": 1597, + "TCAGCT": 1598, + "TCAGCC": 1599, + "TCAGCG": 1600, + "TCAGGA": 1601, + "TCAGGT": 1602, + "TCAGGC": 1603, + "TCAGGG": 1604, + "TCTAAA": 1605, + "TCTAAT": 1606, + "TCTAAC": 1607, + "TCTAAG": 1608, + "TCTATA": 1609, + "TCTATT": 1610, + "TCTATC": 1611, + "TCTATG": 1612, + "TCTACA": 1613, + "TCTACT": 1614, + "TCTACC": 1615, + "TCTACG": 1616, + "TCTAGA": 1617, + "TCTAGT": 1618, + "TCTAGC": 1619, + "TCTAGG": 1620, + "TCTTAA": 1621, + "TCTTAT": 1622, + "TCTTAC": 1623, + "TCTTAG": 1624, + "TCTTTA": 1625, + "TCTTTT": 1626, + "TCTTTC": 1627, + "TCTTTG": 1628, + "TCTTCA": 1629, + "TCTTCT": 1630, + "TCTTCC": 1631, + "TCTTCG": 1632, + "TCTTGA": 1633, + "TCTTGT": 1634, + "TCTTGC": 1635, + "TCTTGG": 1636, + "TCTCAA": 1637, + "TCTCAT": 1638, + "TCTCAC": 1639, + "TCTCAG": 1640, + "TCTCTA": 1641, + "TCTCTT": 1642, + "TCTCTC": 1643, + "TCTCTG": 1644, + "TCTCCA": 1645, + "TCTCCT": 1646, + "TCTCCC": 1647, + "TCTCCG": 1648, + "TCTCGA": 1649, + "TCTCGT": 1650, + "TCTCGC": 1651, + "TCTCGG": 1652, + "TCTGAA": 1653, + "TCTGAT": 1654, + "TCTGAC": 1655, + "TCTGAG": 1656, + "TCTGTA": 1657, + "TCTGTT": 1658, + "TCTGTC": 1659, + "TCTGTG": 1660, + "TCTGCA": 1661, + "TCTGCT": 1662, + "TCTGCC": 1663, + "TCTGCG": 1664, + "TCTGGA": 1665, + "TCTGGT": 1666, + "TCTGGC": 1667, + "TCTGGG": 1668, + "TCCAAA": 1669, + "TCCAAT": 1670, + "TCCAAC": 1671, + "TCCAAG": 1672, + "TCCATA": 1673, + "TCCATT": 1674, + "TCCATC": 1675, + "TCCATG": 1676, + "TCCACA": 1677, + "TCCACT": 1678, + "TCCACC": 1679, + "TCCACG": 1680, + "TCCAGA": 1681, + "TCCAGT": 1682, + "TCCAGC": 1683, + "TCCAGG": 1684, + "TCCTAA": 1685, + "TCCTAT": 1686, + "TCCTAC": 1687, + "TCCTAG": 1688, + "TCCTTA": 1689, + "TCCTTT": 1690, + "TCCTTC": 1691, + "TCCTTG": 1692, + "TCCTCA": 1693, + "TCCTCT": 1694, + "TCCTCC": 1695, + "TCCTCG": 1696, + "TCCTGA": 1697, + "TCCTGT": 1698, + "TCCTGC": 1699, + "TCCTGG": 1700, + "TCCCAA": 1701, + "TCCCAT": 1702, + "TCCCAC": 1703, + "TCCCAG": 1704, + "TCCCTA": 1705, + "TCCCTT": 1706, + "TCCCTC": 1707, + "TCCCTG": 1708, + "TCCCCA": 1709, + "TCCCCT": 1710, + "TCCCCC": 1711, + "TCCCCG": 1712, + "TCCCGA": 1713, + "TCCCGT": 1714, + "TCCCGC": 1715, + "TCCCGG": 1716, + "TCCGAA": 1717, + "TCCGAT": 1718, + "TCCGAC": 1719, + "TCCGAG": 1720, + "TCCGTA": 1721, + "TCCGTT": 1722, + "TCCGTC": 1723, + "TCCGTG": 1724, + "TCCGCA": 1725, + "TCCGCT": 1726, + "TCCGCC": 1727, + "TCCGCG": 1728, + "TCCGGA": 1729, + "TCCGGT": 1730, + "TCCGGC": 1731, + "TCCGGG": 1732, + "TCGAAA": 1733, + "TCGAAT": 1734, + "TCGAAC": 1735, + "TCGAAG": 1736, + "TCGATA": 1737, + "TCGATT": 1738, + "TCGATC": 1739, + "TCGATG": 1740, + "TCGACA": 1741, + "TCGACT": 1742, + "TCGACC": 1743, + "TCGACG": 1744, + "TCGAGA": 1745, + "TCGAGT": 1746, + "TCGAGC": 1747, + "TCGAGG": 1748, + "TCGTAA": 1749, + "TCGTAT": 1750, + "TCGTAC": 1751, + "TCGTAG": 1752, + "TCGTTA": 1753, + "TCGTTT": 1754, + "TCGTTC": 1755, + "TCGTTG": 1756, + "TCGTCA": 1757, + "TCGTCT": 1758, + "TCGTCC": 1759, + "TCGTCG": 1760, + "TCGTGA": 1761, + "TCGTGT": 1762, + "TCGTGC": 1763, + "TCGTGG": 1764, + "TCGCAA": 1765, + "TCGCAT": 1766, + "TCGCAC": 1767, + "TCGCAG": 1768, + "TCGCTA": 1769, + "TCGCTT": 1770, + "TCGCTC": 1771, + "TCGCTG": 1772, + "TCGCCA": 1773, + "TCGCCT": 1774, + "TCGCCC": 1775, + "TCGCCG": 1776, + "TCGCGA": 1777, + "TCGCGT": 1778, + "TCGCGC": 1779, + "TCGCGG": 1780, + "TCGGAA": 1781, + "TCGGAT": 1782, + "TCGGAC": 1783, + "TCGGAG": 1784, + "TCGGTA": 1785, + "TCGGTT": 1786, + "TCGGTC": 1787, + "TCGGTG": 1788, + "TCGGCA": 1789, + "TCGGCT": 1790, + "TCGGCC": 1791, + "TCGGCG": 1792, + "TCGGGA": 1793, + "TCGGGT": 1794, + "TCGGGC": 1795, + "TCGGGG": 1796, + "TGAAAA": 1797, + "TGAAAT": 1798, + "TGAAAC": 1799, + "TGAAAG": 1800, + "TGAATA": 1801, + "TGAATT": 1802, + "TGAATC": 1803, + "TGAATG": 1804, + "TGAACA": 1805, + "TGAACT": 1806, + "TGAACC": 1807, + "TGAACG": 1808, + "TGAAGA": 1809, + "TGAAGT": 1810, + "TGAAGC": 1811, + "TGAAGG": 1812, + "TGATAA": 1813, + "TGATAT": 1814, + "TGATAC": 1815, + "TGATAG": 1816, + "TGATTA": 1817, + "TGATTT": 1818, + "TGATTC": 1819, + "TGATTG": 1820, + "TGATCA": 1821, + "TGATCT": 1822, + "TGATCC": 1823, + "TGATCG": 1824, + "TGATGA": 1825, + "TGATGT": 1826, + "TGATGC": 1827, + "TGATGG": 1828, + "TGACAA": 1829, + "TGACAT": 1830, + "TGACAC": 1831, + "TGACAG": 1832, + "TGACTA": 1833, + "TGACTT": 1834, + "TGACTC": 1835, + "TGACTG": 1836, + "TGACCA": 1837, + "TGACCT": 1838, + "TGACCC": 1839, + "TGACCG": 1840, + "TGACGA": 1841, + "TGACGT": 1842, + "TGACGC": 1843, + "TGACGG": 1844, + "TGAGAA": 1845, + "TGAGAT": 1846, + "TGAGAC": 1847, + "TGAGAG": 1848, + "TGAGTA": 1849, + "TGAGTT": 1850, + "TGAGTC": 1851, + "TGAGTG": 1852, + "TGAGCA": 1853, + "TGAGCT": 1854, + "TGAGCC": 1855, + "TGAGCG": 1856, + "TGAGGA": 1857, + "TGAGGT": 1858, + "TGAGGC": 1859, + "TGAGGG": 1860, + "TGTAAA": 1861, + "TGTAAT": 1862, + "TGTAAC": 1863, + "TGTAAG": 1864, + "TGTATA": 1865, + "TGTATT": 1866, + "TGTATC": 1867, + "TGTATG": 1868, + "TGTACA": 1869, + "TGTACT": 1870, + "TGTACC": 1871, + "TGTACG": 1872, + "TGTAGA": 1873, + "TGTAGT": 1874, + "TGTAGC": 1875, + "TGTAGG": 1876, + "TGTTAA": 1877, + "TGTTAT": 1878, + "TGTTAC": 1879, + "TGTTAG": 1880, + "TGTTTA": 1881, + "TGTTTT": 1882, + "TGTTTC": 1883, + "TGTTTG": 1884, + "TGTTCA": 1885, + "TGTTCT": 1886, + "TGTTCC": 1887, + "TGTTCG": 1888, + "TGTTGA": 1889, + "TGTTGT": 1890, + "TGTTGC": 1891, + "TGTTGG": 1892, + "TGTCAA": 1893, + "TGTCAT": 1894, + "TGTCAC": 1895, + "TGTCAG": 1896, + "TGTCTA": 1897, + "TGTCTT": 1898, + "TGTCTC": 1899, + "TGTCTG": 1900, + "TGTCCA": 1901, + "TGTCCT": 1902, + "TGTCCC": 1903, + "TGTCCG": 1904, + "TGTCGA": 1905, + "TGTCGT": 1906, + "TGTCGC": 1907, + "TGTCGG": 1908, + "TGTGAA": 1909, + "TGTGAT": 1910, + "TGTGAC": 1911, + "TGTGAG": 1912, + "TGTGTA": 1913, + "TGTGTT": 1914, + "TGTGTC": 1915, + "TGTGTG": 1916, + "TGTGCA": 1917, + "TGTGCT": 1918, + "TGTGCC": 1919, + "TGTGCG": 1920, + "TGTGGA": 1921, + "TGTGGT": 1922, + "TGTGGC": 1923, + "TGTGGG": 1924, + "TGCAAA": 1925, + "TGCAAT": 1926, + "TGCAAC": 1927, + "TGCAAG": 1928, + "TGCATA": 1929, + "TGCATT": 1930, + "TGCATC": 1931, + "TGCATG": 1932, + "TGCACA": 1933, + "TGCACT": 1934, + "TGCACC": 1935, + "TGCACG": 1936, + "TGCAGA": 1937, + "TGCAGT": 1938, + "TGCAGC": 1939, + "TGCAGG": 1940, + "TGCTAA": 1941, + "TGCTAT": 1942, + "TGCTAC": 1943, + "TGCTAG": 1944, + "TGCTTA": 1945, + "TGCTTT": 1946, + "TGCTTC": 1947, + "TGCTTG": 1948, + "TGCTCA": 1949, + "TGCTCT": 1950, + "TGCTCC": 1951, + "TGCTCG": 1952, + "TGCTGA": 1953, + "TGCTGT": 1954, + "TGCTGC": 1955, + "TGCTGG": 1956, + "TGCCAA": 1957, + "TGCCAT": 1958, + "TGCCAC": 1959, + "TGCCAG": 1960, + "TGCCTA": 1961, + "TGCCTT": 1962, + "TGCCTC": 1963, + "TGCCTG": 1964, + "TGCCCA": 1965, + "TGCCCT": 1966, + "TGCCCC": 1967, + "TGCCCG": 1968, + "TGCCGA": 1969, + "TGCCGT": 1970, + "TGCCGC": 1971, + "TGCCGG": 1972, + "TGCGAA": 1973, + "TGCGAT": 1974, + "TGCGAC": 1975, + "TGCGAG": 1976, + "TGCGTA": 1977, + "TGCGTT": 1978, + "TGCGTC": 1979, + "TGCGTG": 1980, + "TGCGCA": 1981, + "TGCGCT": 1982, + "TGCGCC": 1983, + "TGCGCG": 1984, + "TGCGGA": 1985, + "TGCGGT": 1986, + "TGCGGC": 1987, + "TGCGGG": 1988, + "TGGAAA": 1989, + "TGGAAT": 1990, + "TGGAAC": 1991, + "TGGAAG": 1992, + "TGGATA": 1993, + "TGGATT": 1994, + "TGGATC": 1995, + "TGGATG": 1996, + "TGGACA": 1997, + "TGGACT": 1998, + "TGGACC": 1999, + "TGGACG": 2000, + "TGGAGA": 2001, + "TGGAGT": 2002, + "TGGAGC": 2003, + "TGGAGG": 2004, + "TGGTAA": 2005, + "TGGTAT": 2006, + "TGGTAC": 2007, + "TGGTAG": 2008, + "TGGTTA": 2009, + "TGGTTT": 2010, + "TGGTTC": 2011, + "TGGTTG": 2012, + "TGGTCA": 2013, + "TGGTCT": 2014, + "TGGTCC": 2015, + "TGGTCG": 2016, + "TGGTGA": 2017, + "TGGTGT": 2018, + "TGGTGC": 2019, + "TGGTGG": 2020, + "TGGCAA": 2021, + "TGGCAT": 2022, + "TGGCAC": 2023, + "TGGCAG": 2024, + "TGGCTA": 2025, + "TGGCTT": 2026, + "TGGCTC": 2027, + "TGGCTG": 2028, + "TGGCCA": 2029, + "TGGCCT": 2030, + "TGGCCC": 2031, + "TGGCCG": 2032, + "TGGCGA": 2033, + "TGGCGT": 2034, + "TGGCGC": 2035, + "TGGCGG": 2036, + "TGGGAA": 2037, + "TGGGAT": 2038, + "TGGGAC": 2039, + "TGGGAG": 2040, + "TGGGTA": 2041, + "TGGGTT": 2042, + "TGGGTC": 2043, + "TGGGTG": 2044, + "TGGGCA": 2045, + "TGGGCT": 2046, + "TGGGCC": 2047, + "TGGGCG": 2048, + "TGGGGA": 2049, + "TGGGGT": 2050, + "TGGGGC": 2051, + "TGGGGG": 2052, + "CAAAAA": 2053, + "CAAAAT": 2054, + "CAAAAC": 2055, + "CAAAAG": 2056, + "CAAATA": 2057, + "CAAATT": 2058, + "CAAATC": 2059, + "CAAATG": 2060, + "CAAACA": 2061, + "CAAACT": 2062, + "CAAACC": 2063, + "CAAACG": 2064, + "CAAAGA": 2065, + "CAAAGT": 2066, + "CAAAGC": 2067, + "CAAAGG": 2068, + "CAATAA": 2069, + "CAATAT": 2070, + "CAATAC": 2071, + "CAATAG": 2072, + "CAATTA": 2073, + "CAATTT": 2074, + "CAATTC": 2075, + "CAATTG": 2076, + "CAATCA": 2077, + "CAATCT": 2078, + "CAATCC": 2079, + "CAATCG": 2080, + "CAATGA": 2081, + "CAATGT": 2082, + "CAATGC": 2083, + "CAATGG": 2084, + "CAACAA": 2085, + "CAACAT": 2086, + "CAACAC": 2087, + "CAACAG": 2088, + "CAACTA": 2089, + "CAACTT": 2090, + "CAACTC": 2091, + "CAACTG": 2092, + "CAACCA": 2093, + "CAACCT": 2094, + "CAACCC": 2095, + "CAACCG": 2096, + "CAACGA": 2097, + "CAACGT": 2098, + "CAACGC": 2099, + "CAACGG": 2100, + "CAAGAA": 2101, + "CAAGAT": 2102, + "CAAGAC": 2103, + "CAAGAG": 2104, + "CAAGTA": 2105, + "CAAGTT": 2106, + "CAAGTC": 2107, + "CAAGTG": 2108, + "CAAGCA": 2109, + "CAAGCT": 2110, + "CAAGCC": 2111, + "CAAGCG": 2112, + "CAAGGA": 2113, + "CAAGGT": 2114, + "CAAGGC": 2115, + "CAAGGG": 2116, + "CATAAA": 2117, + "CATAAT": 2118, + "CATAAC": 2119, + "CATAAG": 2120, + "CATATA": 2121, + "CATATT": 2122, + "CATATC": 2123, + "CATATG": 2124, + "CATACA": 2125, + "CATACT": 2126, + "CATACC": 2127, + "CATACG": 2128, + "CATAGA": 2129, + "CATAGT": 2130, + "CATAGC": 2131, + "CATAGG": 2132, + "CATTAA": 2133, + "CATTAT": 2134, + "CATTAC": 2135, + "CATTAG": 2136, + "CATTTA": 2137, + "CATTTT": 2138, + "CATTTC": 2139, + "CATTTG": 2140, + "CATTCA": 2141, + "CATTCT": 2142, + "CATTCC": 2143, + "CATTCG": 2144, + "CATTGA": 2145, + "CATTGT": 2146, + "CATTGC": 2147, + "CATTGG": 2148, + "CATCAA": 2149, + "CATCAT": 2150, + "CATCAC": 2151, + "CATCAG": 2152, + "CATCTA": 2153, + "CATCTT": 2154, + "CATCTC": 2155, + "CATCTG": 2156, + "CATCCA": 2157, + "CATCCT": 2158, + "CATCCC": 2159, + "CATCCG": 2160, + "CATCGA": 2161, + "CATCGT": 2162, + "CATCGC": 2163, + "CATCGG": 2164, + "CATGAA": 2165, + "CATGAT": 2166, + "CATGAC": 2167, + "CATGAG": 2168, + "CATGTA": 2169, + "CATGTT": 2170, + "CATGTC": 2171, + "CATGTG": 2172, + "CATGCA": 2173, + "CATGCT": 2174, + "CATGCC": 2175, + "CATGCG": 2176, + "CATGGA": 2177, + "CATGGT": 2178, + "CATGGC": 2179, + "CATGGG": 2180, + "CACAAA": 2181, + "CACAAT": 2182, + "CACAAC": 2183, + "CACAAG": 2184, + "CACATA": 2185, + "CACATT": 2186, + "CACATC": 2187, + "CACATG": 2188, + "CACACA": 2189, + "CACACT": 2190, + "CACACC": 2191, + "CACACG": 2192, + "CACAGA": 2193, + "CACAGT": 2194, + "CACAGC": 2195, + "CACAGG": 2196, + "CACTAA": 2197, + "CACTAT": 2198, + "CACTAC": 2199, + "CACTAG": 2200, + "CACTTA": 2201, + "CACTTT": 2202, + "CACTTC": 2203, + "CACTTG": 2204, + "CACTCA": 2205, + "CACTCT": 2206, + "CACTCC": 2207, + "CACTCG": 2208, + "CACTGA": 2209, + "CACTGT": 2210, + "CACTGC": 2211, + "CACTGG": 2212, + "CACCAA": 2213, + "CACCAT": 2214, + "CACCAC": 2215, + "CACCAG": 2216, + "CACCTA": 2217, + "CACCTT": 2218, + "CACCTC": 2219, + "CACCTG": 2220, + "CACCCA": 2221, + "CACCCT": 2222, + "CACCCC": 2223, + "CACCCG": 2224, + "CACCGA": 2225, + "CACCGT": 2226, + "CACCGC": 2227, + "CACCGG": 2228, + "CACGAA": 2229, + "CACGAT": 2230, + "CACGAC": 2231, + "CACGAG": 2232, + "CACGTA": 2233, + "CACGTT": 2234, + "CACGTC": 2235, + "CACGTG": 2236, + "CACGCA": 2237, + "CACGCT": 2238, + "CACGCC": 2239, + "CACGCG": 2240, + "CACGGA": 2241, + "CACGGT": 2242, + "CACGGC": 2243, + "CACGGG": 2244, + "CAGAAA": 2245, + "CAGAAT": 2246, + "CAGAAC": 2247, + "CAGAAG": 2248, + "CAGATA": 2249, + "CAGATT": 2250, + "CAGATC": 2251, + "CAGATG": 2252, + "CAGACA": 2253, + "CAGACT": 2254, + "CAGACC": 2255, + "CAGACG": 2256, + "CAGAGA": 2257, + "CAGAGT": 2258, + "CAGAGC": 2259, + "CAGAGG": 2260, + "CAGTAA": 2261, + "CAGTAT": 2262, + "CAGTAC": 2263, + "CAGTAG": 2264, + "CAGTTA": 2265, + "CAGTTT": 2266, + "CAGTTC": 2267, + "CAGTTG": 2268, + "CAGTCA": 2269, + "CAGTCT": 2270, + "CAGTCC": 2271, + "CAGTCG": 2272, + "CAGTGA": 2273, + "CAGTGT": 2274, + "CAGTGC": 2275, + "CAGTGG": 2276, + "CAGCAA": 2277, + "CAGCAT": 2278, + "CAGCAC": 2279, + "CAGCAG": 2280, + "CAGCTA": 2281, + "CAGCTT": 2282, + "CAGCTC": 2283, + "CAGCTG": 2284, + "CAGCCA": 2285, + "CAGCCT": 2286, + "CAGCCC": 2287, + "CAGCCG": 2288, + "CAGCGA": 2289, + "CAGCGT": 2290, + "CAGCGC": 2291, + "CAGCGG": 2292, + "CAGGAA": 2293, + "CAGGAT": 2294, + "CAGGAC": 2295, + "CAGGAG": 2296, + "CAGGTA": 2297, + "CAGGTT": 2298, + "CAGGTC": 2299, + "CAGGTG": 2300, + "CAGGCA": 2301, + "CAGGCT": 2302, + "CAGGCC": 2303, + "CAGGCG": 2304, + "CAGGGA": 2305, + "CAGGGT": 2306, + "CAGGGC": 2307, + "CAGGGG": 2308, + "CTAAAA": 2309, + "CTAAAT": 2310, + "CTAAAC": 2311, + "CTAAAG": 2312, + "CTAATA": 2313, + "CTAATT": 2314, + "CTAATC": 2315, + "CTAATG": 2316, + "CTAACA": 2317, + "CTAACT": 2318, + "CTAACC": 2319, + "CTAACG": 2320, + "CTAAGA": 2321, + "CTAAGT": 2322, + "CTAAGC": 2323, + "CTAAGG": 2324, + "CTATAA": 2325, + "CTATAT": 2326, + "CTATAC": 2327, + "CTATAG": 2328, + "CTATTA": 2329, + "CTATTT": 2330, + "CTATTC": 2331, + "CTATTG": 2332, + "CTATCA": 2333, + "CTATCT": 2334, + "CTATCC": 2335, + "CTATCG": 2336, + "CTATGA": 2337, + "CTATGT": 2338, + "CTATGC": 2339, + "CTATGG": 2340, + "CTACAA": 2341, + "CTACAT": 2342, + "CTACAC": 2343, + "CTACAG": 2344, + "CTACTA": 2345, + "CTACTT": 2346, + "CTACTC": 2347, + "CTACTG": 2348, + "CTACCA": 2349, + "CTACCT": 2350, + "CTACCC": 2351, + "CTACCG": 2352, + "CTACGA": 2353, + "CTACGT": 2354, + "CTACGC": 2355, + "CTACGG": 2356, + "CTAGAA": 2357, + "CTAGAT": 2358, + "CTAGAC": 2359, + "CTAGAG": 2360, + "CTAGTA": 2361, + "CTAGTT": 2362, + "CTAGTC": 2363, + "CTAGTG": 2364, + "CTAGCA": 2365, + "CTAGCT": 2366, + "CTAGCC": 2367, + "CTAGCG": 2368, + "CTAGGA": 2369, + "CTAGGT": 2370, + "CTAGGC": 2371, + "CTAGGG": 2372, + "CTTAAA": 2373, + "CTTAAT": 2374, + "CTTAAC": 2375, + "CTTAAG": 2376, + "CTTATA": 2377, + "CTTATT": 2378, + "CTTATC": 2379, + "CTTATG": 2380, + "CTTACA": 2381, + "CTTACT": 2382, + "CTTACC": 2383, + "CTTACG": 2384, + "CTTAGA": 2385, + "CTTAGT": 2386, + "CTTAGC": 2387, + "CTTAGG": 2388, + "CTTTAA": 2389, + "CTTTAT": 2390, + "CTTTAC": 2391, + "CTTTAG": 2392, + "CTTTTA": 2393, + "CTTTTT": 2394, + "CTTTTC": 2395, + "CTTTTG": 2396, + "CTTTCA": 2397, + "CTTTCT": 2398, + "CTTTCC": 2399, + "CTTTCG": 2400, + "CTTTGA": 2401, + "CTTTGT": 2402, + "CTTTGC": 2403, + "CTTTGG": 2404, + "CTTCAA": 2405, + "CTTCAT": 2406, + "CTTCAC": 2407, + "CTTCAG": 2408, + "CTTCTA": 2409, + "CTTCTT": 2410, + "CTTCTC": 2411, + "CTTCTG": 2412, + "CTTCCA": 2413, + "CTTCCT": 2414, + "CTTCCC": 2415, + "CTTCCG": 2416, + "CTTCGA": 2417, + "CTTCGT": 2418, + "CTTCGC": 2419, + "CTTCGG": 2420, + "CTTGAA": 2421, + "CTTGAT": 2422, + "CTTGAC": 2423, + "CTTGAG": 2424, + "CTTGTA": 2425, + "CTTGTT": 2426, + "CTTGTC": 2427, + "CTTGTG": 2428, + "CTTGCA": 2429, + "CTTGCT": 2430, + "CTTGCC": 2431, + "CTTGCG": 2432, + "CTTGGA": 2433, + "CTTGGT": 2434, + "CTTGGC": 2435, + "CTTGGG": 2436, + "CTCAAA": 2437, + "CTCAAT": 2438, + "CTCAAC": 2439, + "CTCAAG": 2440, + "CTCATA": 2441, + "CTCATT": 2442, + "CTCATC": 2443, + "CTCATG": 2444, + "CTCACA": 2445, + "CTCACT": 2446, + "CTCACC": 2447, + "CTCACG": 2448, + "CTCAGA": 2449, + "CTCAGT": 2450, + "CTCAGC": 2451, + "CTCAGG": 2452, + "CTCTAA": 2453, + "CTCTAT": 2454, + "CTCTAC": 2455, + "CTCTAG": 2456, + "CTCTTA": 2457, + "CTCTTT": 2458, + "CTCTTC": 2459, + "CTCTTG": 2460, + "CTCTCA": 2461, + "CTCTCT": 2462, + "CTCTCC": 2463, + "CTCTCG": 2464, + "CTCTGA": 2465, + "CTCTGT": 2466, + "CTCTGC": 2467, + "CTCTGG": 2468, + "CTCCAA": 2469, + "CTCCAT": 2470, + "CTCCAC": 2471, + "CTCCAG": 2472, + "CTCCTA": 2473, + "CTCCTT": 2474, + "CTCCTC": 2475, + "CTCCTG": 2476, + "CTCCCA": 2477, + "CTCCCT": 2478, + "CTCCCC": 2479, + "CTCCCG": 2480, + "CTCCGA": 2481, + "CTCCGT": 2482, + "CTCCGC": 2483, + "CTCCGG": 2484, + "CTCGAA": 2485, + "CTCGAT": 2486, + "CTCGAC": 2487, + "CTCGAG": 2488, + "CTCGTA": 2489, + "CTCGTT": 2490, + "CTCGTC": 2491, + "CTCGTG": 2492, + "CTCGCA": 2493, + "CTCGCT": 2494, + "CTCGCC": 2495, + "CTCGCG": 2496, + "CTCGGA": 2497, + "CTCGGT": 2498, + "CTCGGC": 2499, + "CTCGGG": 2500, + "CTGAAA": 2501, + "CTGAAT": 2502, + "CTGAAC": 2503, + "CTGAAG": 2504, + "CTGATA": 2505, + "CTGATT": 2506, + "CTGATC": 2507, + "CTGATG": 2508, + "CTGACA": 2509, + "CTGACT": 2510, + "CTGACC": 2511, + "CTGACG": 2512, + "CTGAGA": 2513, + "CTGAGT": 2514, + "CTGAGC": 2515, + "CTGAGG": 2516, + "CTGTAA": 2517, + "CTGTAT": 2518, + "CTGTAC": 2519, + "CTGTAG": 2520, + "CTGTTA": 2521, + "CTGTTT": 2522, + "CTGTTC": 2523, + "CTGTTG": 2524, + "CTGTCA": 2525, + "CTGTCT": 2526, + "CTGTCC": 2527, + "CTGTCG": 2528, + "CTGTGA": 2529, + "CTGTGT": 2530, + "CTGTGC": 2531, + "CTGTGG": 2532, + "CTGCAA": 2533, + "CTGCAT": 2534, + "CTGCAC": 2535, + "CTGCAG": 2536, + "CTGCTA": 2537, + "CTGCTT": 2538, + "CTGCTC": 2539, + "CTGCTG": 2540, + "CTGCCA": 2541, + "CTGCCT": 2542, + "CTGCCC": 2543, + "CTGCCG": 2544, + "CTGCGA": 2545, + "CTGCGT": 2546, + "CTGCGC": 2547, + "CTGCGG": 2548, + "CTGGAA": 2549, + "CTGGAT": 2550, + "CTGGAC": 2551, + "CTGGAG": 2552, + "CTGGTA": 2553, + "CTGGTT": 2554, + "CTGGTC": 2555, + "CTGGTG": 2556, + "CTGGCA": 2557, + "CTGGCT": 2558, + "CTGGCC": 2559, + "CTGGCG": 2560, + "CTGGGA": 2561, + "CTGGGT": 2562, + "CTGGGC": 2563, + "CTGGGG": 2564, + "CCAAAA": 2565, + "CCAAAT": 2566, + "CCAAAC": 2567, + "CCAAAG": 2568, + "CCAATA": 2569, + "CCAATT": 2570, + "CCAATC": 2571, + "CCAATG": 2572, + "CCAACA": 2573, + "CCAACT": 2574, + "CCAACC": 2575, + "CCAACG": 2576, + "CCAAGA": 2577, + "CCAAGT": 2578, + "CCAAGC": 2579, + "CCAAGG": 2580, + "CCATAA": 2581, + "CCATAT": 2582, + "CCATAC": 2583, + "CCATAG": 2584, + "CCATTA": 2585, + "CCATTT": 2586, + "CCATTC": 2587, + "CCATTG": 2588, + "CCATCA": 2589, + "CCATCT": 2590, + "CCATCC": 2591, + "CCATCG": 2592, + "CCATGA": 2593, + "CCATGT": 2594, + "CCATGC": 2595, + "CCATGG": 2596, + "CCACAA": 2597, + "CCACAT": 2598, + "CCACAC": 2599, + "CCACAG": 2600, + "CCACTA": 2601, + "CCACTT": 2602, + "CCACTC": 2603, + "CCACTG": 2604, + "CCACCA": 2605, + "CCACCT": 2606, + "CCACCC": 2607, + "CCACCG": 2608, + "CCACGA": 2609, + "CCACGT": 2610, + "CCACGC": 2611, + "CCACGG": 2612, + "CCAGAA": 2613, + "CCAGAT": 2614, + "CCAGAC": 2615, + "CCAGAG": 2616, + "CCAGTA": 2617, + "CCAGTT": 2618, + "CCAGTC": 2619, + "CCAGTG": 2620, + "CCAGCA": 2621, + "CCAGCT": 2622, + "CCAGCC": 2623, + "CCAGCG": 2624, + "CCAGGA": 2625, + "CCAGGT": 2626, + "CCAGGC": 2627, + "CCAGGG": 2628, + "CCTAAA": 2629, + "CCTAAT": 2630, + "CCTAAC": 2631, + "CCTAAG": 2632, + "CCTATA": 2633, + "CCTATT": 2634, + "CCTATC": 2635, + "CCTATG": 2636, + "CCTACA": 2637, + "CCTACT": 2638, + "CCTACC": 2639, + "CCTACG": 2640, + "CCTAGA": 2641, + "CCTAGT": 2642, + "CCTAGC": 2643, + "CCTAGG": 2644, + "CCTTAA": 2645, + "CCTTAT": 2646, + "CCTTAC": 2647, + "CCTTAG": 2648, + "CCTTTA": 2649, + "CCTTTT": 2650, + "CCTTTC": 2651, + "CCTTTG": 2652, + "CCTTCA": 2653, + "CCTTCT": 2654, + "CCTTCC": 2655, + "CCTTCG": 2656, + "CCTTGA": 2657, + "CCTTGT": 2658, + "CCTTGC": 2659, + "CCTTGG": 2660, + "CCTCAA": 2661, + "CCTCAT": 2662, + "CCTCAC": 2663, + "CCTCAG": 2664, + "CCTCTA": 2665, + "CCTCTT": 2666, + "CCTCTC": 2667, + "CCTCTG": 2668, + "CCTCCA": 2669, + "CCTCCT": 2670, + "CCTCCC": 2671, + "CCTCCG": 2672, + "CCTCGA": 2673, + "CCTCGT": 2674, + "CCTCGC": 2675, + "CCTCGG": 2676, + "CCTGAA": 2677, + "CCTGAT": 2678, + "CCTGAC": 2679, + "CCTGAG": 2680, + "CCTGTA": 2681, + "CCTGTT": 2682, + "CCTGTC": 2683, + "CCTGTG": 2684, + "CCTGCA": 2685, + "CCTGCT": 2686, + "CCTGCC": 2687, + "CCTGCG": 2688, + "CCTGGA": 2689, + "CCTGGT": 2690, + "CCTGGC": 2691, + "CCTGGG": 2692, + "CCCAAA": 2693, + "CCCAAT": 2694, + "CCCAAC": 2695, + "CCCAAG": 2696, + "CCCATA": 2697, + "CCCATT": 2698, + "CCCATC": 2699, + "CCCATG": 2700, + "CCCACA": 2701, + "CCCACT": 2702, + "CCCACC": 2703, + "CCCACG": 2704, + "CCCAGA": 2705, + "CCCAGT": 2706, + "CCCAGC": 2707, + "CCCAGG": 2708, + "CCCTAA": 2709, + "CCCTAT": 2710, + "CCCTAC": 2711, + "CCCTAG": 2712, + "CCCTTA": 2713, + "CCCTTT": 2714, + "CCCTTC": 2715, + "CCCTTG": 2716, + "CCCTCA": 2717, + "CCCTCT": 2718, + "CCCTCC": 2719, + "CCCTCG": 2720, + "CCCTGA": 2721, + "CCCTGT": 2722, + "CCCTGC": 2723, + "CCCTGG": 2724, + "CCCCAA": 2725, + "CCCCAT": 2726, + "CCCCAC": 2727, + "CCCCAG": 2728, + "CCCCTA": 2729, + "CCCCTT": 2730, + "CCCCTC": 2731, + "CCCCTG": 2732, + "CCCCCA": 2733, + "CCCCCT": 2734, + "CCCCCC": 2735, + "CCCCCG": 2736, + "CCCCGA": 2737, + "CCCCGT": 2738, + "CCCCGC": 2739, + "CCCCGG": 2740, + "CCCGAA": 2741, + "CCCGAT": 2742, + "CCCGAC": 2743, + "CCCGAG": 2744, + "CCCGTA": 2745, + "CCCGTT": 2746, + "CCCGTC": 2747, + "CCCGTG": 2748, + "CCCGCA": 2749, + "CCCGCT": 2750, + "CCCGCC": 2751, + "CCCGCG": 2752, + "CCCGGA": 2753, + "CCCGGT": 2754, + "CCCGGC": 2755, + "CCCGGG": 2756, + "CCGAAA": 2757, + "CCGAAT": 2758, + "CCGAAC": 2759, + "CCGAAG": 2760, + "CCGATA": 2761, + "CCGATT": 2762, + "CCGATC": 2763, + "CCGATG": 2764, + "CCGACA": 2765, + "CCGACT": 2766, + "CCGACC": 2767, + "CCGACG": 2768, + "CCGAGA": 2769, + "CCGAGT": 2770, + "CCGAGC": 2771, + "CCGAGG": 2772, + "CCGTAA": 2773, + "CCGTAT": 2774, + "CCGTAC": 2775, + "CCGTAG": 2776, + "CCGTTA": 2777, + "CCGTTT": 2778, + "CCGTTC": 2779, + "CCGTTG": 2780, + "CCGTCA": 2781, + "CCGTCT": 2782, + "CCGTCC": 2783, + "CCGTCG": 2784, + "CCGTGA": 2785, + "CCGTGT": 2786, + "CCGTGC": 2787, + "CCGTGG": 2788, + "CCGCAA": 2789, + "CCGCAT": 2790, + "CCGCAC": 2791, + "CCGCAG": 2792, + "CCGCTA": 2793, + "CCGCTT": 2794, + "CCGCTC": 2795, + "CCGCTG": 2796, + "CCGCCA": 2797, + "CCGCCT": 2798, + "CCGCCC": 2799, + "CCGCCG": 2800, + "CCGCGA": 2801, + "CCGCGT": 2802, + "CCGCGC": 2803, + "CCGCGG": 2804, + "CCGGAA": 2805, + "CCGGAT": 2806, + "CCGGAC": 2807, + "CCGGAG": 2808, + "CCGGTA": 2809, + "CCGGTT": 2810, + "CCGGTC": 2811, + "CCGGTG": 2812, + "CCGGCA": 2813, + "CCGGCT": 2814, + "CCGGCC": 2815, + "CCGGCG": 2816, + "CCGGGA": 2817, + "CCGGGT": 2818, + "CCGGGC": 2819, + "CCGGGG": 2820, + "CGAAAA": 2821, + "CGAAAT": 2822, + "CGAAAC": 2823, + "CGAAAG": 2824, + "CGAATA": 2825, + "CGAATT": 2826, + "CGAATC": 2827, + "CGAATG": 2828, + "CGAACA": 2829, + "CGAACT": 2830, + "CGAACC": 2831, + "CGAACG": 2832, + "CGAAGA": 2833, + "CGAAGT": 2834, + "CGAAGC": 2835, + "CGAAGG": 2836, + "CGATAA": 2837, + "CGATAT": 2838, + "CGATAC": 2839, + "CGATAG": 2840, + "CGATTA": 2841, + "CGATTT": 2842, + "CGATTC": 2843, + "CGATTG": 2844, + "CGATCA": 2845, + "CGATCT": 2846, + "CGATCC": 2847, + "CGATCG": 2848, + "CGATGA": 2849, + "CGATGT": 2850, + "CGATGC": 2851, + "CGATGG": 2852, + "CGACAA": 2853, + "CGACAT": 2854, + "CGACAC": 2855, + "CGACAG": 2856, + "CGACTA": 2857, + "CGACTT": 2858, + "CGACTC": 2859, + "CGACTG": 2860, + "CGACCA": 2861, + "CGACCT": 2862, + "CGACCC": 2863, + "CGACCG": 2864, + "CGACGA": 2865, + "CGACGT": 2866, + "CGACGC": 2867, + "CGACGG": 2868, + "CGAGAA": 2869, + "CGAGAT": 2870, + "CGAGAC": 2871, + "CGAGAG": 2872, + "CGAGTA": 2873, + "CGAGTT": 2874, + "CGAGTC": 2875, + "CGAGTG": 2876, + "CGAGCA": 2877, + "CGAGCT": 2878, + "CGAGCC": 2879, + "CGAGCG": 2880, + "CGAGGA": 2881, + "CGAGGT": 2882, + "CGAGGC": 2883, + "CGAGGG": 2884, + "CGTAAA": 2885, + "CGTAAT": 2886, + "CGTAAC": 2887, + "CGTAAG": 2888, + "CGTATA": 2889, + "CGTATT": 2890, + "CGTATC": 2891, + "CGTATG": 2892, + "CGTACA": 2893, + "CGTACT": 2894, + "CGTACC": 2895, + "CGTACG": 2896, + "CGTAGA": 2897, + "CGTAGT": 2898, + "CGTAGC": 2899, + "CGTAGG": 2900, + "CGTTAA": 2901, + "CGTTAT": 2902, + "CGTTAC": 2903, + "CGTTAG": 2904, + "CGTTTA": 2905, + "CGTTTT": 2906, + "CGTTTC": 2907, + "CGTTTG": 2908, + "CGTTCA": 2909, + "CGTTCT": 2910, + "CGTTCC": 2911, + "CGTTCG": 2912, + "CGTTGA": 2913, + "CGTTGT": 2914, + "CGTTGC": 2915, + "CGTTGG": 2916, + "CGTCAA": 2917, + "CGTCAT": 2918, + "CGTCAC": 2919, + "CGTCAG": 2920, + "CGTCTA": 2921, + "CGTCTT": 2922, + "CGTCTC": 2923, + "CGTCTG": 2924, + "CGTCCA": 2925, + "CGTCCT": 2926, + "CGTCCC": 2927, + "CGTCCG": 2928, + "CGTCGA": 2929, + "CGTCGT": 2930, + "CGTCGC": 2931, + "CGTCGG": 2932, + "CGTGAA": 2933, + "CGTGAT": 2934, + "CGTGAC": 2935, + "CGTGAG": 2936, + "CGTGTA": 2937, + "CGTGTT": 2938, + "CGTGTC": 2939, + "CGTGTG": 2940, + "CGTGCA": 2941, + "CGTGCT": 2942, + "CGTGCC": 2943, + "CGTGCG": 2944, + "CGTGGA": 2945, + "CGTGGT": 2946, + "CGTGGC": 2947, + "CGTGGG": 2948, + "CGCAAA": 2949, + "CGCAAT": 2950, + "CGCAAC": 2951, + "CGCAAG": 2952, + "CGCATA": 2953, + "CGCATT": 2954, + "CGCATC": 2955, + "CGCATG": 2956, + "CGCACA": 2957, + "CGCACT": 2958, + "CGCACC": 2959, + "CGCACG": 2960, + "CGCAGA": 2961, + "CGCAGT": 2962, + "CGCAGC": 2963, + "CGCAGG": 2964, + "CGCTAA": 2965, + "CGCTAT": 2966, + "CGCTAC": 2967, + "CGCTAG": 2968, + "CGCTTA": 2969, + "CGCTTT": 2970, + "CGCTTC": 2971, + "CGCTTG": 2972, + "CGCTCA": 2973, + "CGCTCT": 2974, + "CGCTCC": 2975, + "CGCTCG": 2976, + "CGCTGA": 2977, + "CGCTGT": 2978, + "CGCTGC": 2979, + "CGCTGG": 2980, + "CGCCAA": 2981, + "CGCCAT": 2982, + "CGCCAC": 2983, + "CGCCAG": 2984, + "CGCCTA": 2985, + "CGCCTT": 2986, + "CGCCTC": 2987, + "CGCCTG": 2988, + "CGCCCA": 2989, + "CGCCCT": 2990, + "CGCCCC": 2991, + "CGCCCG": 2992, + "CGCCGA": 2993, + "CGCCGT": 2994, + "CGCCGC": 2995, + "CGCCGG": 2996, + "CGCGAA": 2997, + "CGCGAT": 2998, + "CGCGAC": 2999, + "CGCGAG": 3000, + "CGCGTA": 3001, + "CGCGTT": 3002, + "CGCGTC": 3003, + "CGCGTG": 3004, + "CGCGCA": 3005, + "CGCGCT": 3006, + "CGCGCC": 3007, + "CGCGCG": 3008, + "CGCGGA": 3009, + "CGCGGT": 3010, + "CGCGGC": 3011, + "CGCGGG": 3012, + "CGGAAA": 3013, + "CGGAAT": 3014, + "CGGAAC": 3015, + "CGGAAG": 3016, + "CGGATA": 3017, + "CGGATT": 3018, + "CGGATC": 3019, + "CGGATG": 3020, + "CGGACA": 3021, + "CGGACT": 3022, + "CGGACC": 3023, + "CGGACG": 3024, + "CGGAGA": 3025, + "CGGAGT": 3026, + "CGGAGC": 3027, + "CGGAGG": 3028, + "CGGTAA": 3029, + "CGGTAT": 3030, + "CGGTAC": 3031, + "CGGTAG": 3032, + "CGGTTA": 3033, + "CGGTTT": 3034, + "CGGTTC": 3035, + "CGGTTG": 3036, + "CGGTCA": 3037, + "CGGTCT": 3038, + "CGGTCC": 3039, + "CGGTCG": 3040, + "CGGTGA": 3041, + "CGGTGT": 3042, + "CGGTGC": 3043, + "CGGTGG": 3044, + "CGGCAA": 3045, + "CGGCAT": 3046, + "CGGCAC": 3047, + "CGGCAG": 3048, + "CGGCTA": 3049, + "CGGCTT": 3050, + "CGGCTC": 3051, + "CGGCTG": 3052, + "CGGCCA": 3053, + "CGGCCT": 3054, + "CGGCCC": 3055, + "CGGCCG": 3056, + "CGGCGA": 3057, + "CGGCGT": 3058, + "CGGCGC": 3059, + "CGGCGG": 3060, + "CGGGAA": 3061, + "CGGGAT": 3062, + "CGGGAC": 3063, + "CGGGAG": 3064, + "CGGGTA": 3065, + "CGGGTT": 3066, + "CGGGTC": 3067, + "CGGGTG": 3068, + "CGGGCA": 3069, + "CGGGCT": 3070, + "CGGGCC": 3071, + "CGGGCG": 3072, + "CGGGGA": 3073, + "CGGGGT": 3074, + "CGGGGC": 3075, + "CGGGGG": 3076, + "GAAAAA": 3077, + "GAAAAT": 3078, + "GAAAAC": 3079, + "GAAAAG": 3080, + "GAAATA": 3081, + "GAAATT": 3082, + "GAAATC": 3083, + "GAAATG": 3084, + "GAAACA": 3085, + "GAAACT": 3086, + "GAAACC": 3087, + "GAAACG": 3088, + "GAAAGA": 3089, + "GAAAGT": 3090, + "GAAAGC": 3091, + "GAAAGG": 3092, + "GAATAA": 3093, + "GAATAT": 3094, + "GAATAC": 3095, + "GAATAG": 3096, + "GAATTA": 3097, + "GAATTT": 3098, + "GAATTC": 3099, + "GAATTG": 3100, + "GAATCA": 3101, + "GAATCT": 3102, + "GAATCC": 3103, + "GAATCG": 3104, + "GAATGA": 3105, + "GAATGT": 3106, + "GAATGC": 3107, + "GAATGG": 3108, + "GAACAA": 3109, + "GAACAT": 3110, + "GAACAC": 3111, + "GAACAG": 3112, + "GAACTA": 3113, + "GAACTT": 3114, + "GAACTC": 3115, + "GAACTG": 3116, + "GAACCA": 3117, + "GAACCT": 3118, + "GAACCC": 3119, + "GAACCG": 3120, + "GAACGA": 3121, + "GAACGT": 3122, + "GAACGC": 3123, + "GAACGG": 3124, + "GAAGAA": 3125, + "GAAGAT": 3126, + "GAAGAC": 3127, + "GAAGAG": 3128, + "GAAGTA": 3129, + "GAAGTT": 3130, + "GAAGTC": 3131, + "GAAGTG": 3132, + "GAAGCA": 3133, + "GAAGCT": 3134, + "GAAGCC": 3135, + "GAAGCG": 3136, + "GAAGGA": 3137, + "GAAGGT": 3138, + "GAAGGC": 3139, + "GAAGGG": 3140, + "GATAAA": 3141, + "GATAAT": 3142, + "GATAAC": 3143, + "GATAAG": 3144, + "GATATA": 3145, + "GATATT": 3146, + "GATATC": 3147, + "GATATG": 3148, + "GATACA": 3149, + "GATACT": 3150, + "GATACC": 3151, + "GATACG": 3152, + "GATAGA": 3153, + "GATAGT": 3154, + "GATAGC": 3155, + "GATAGG": 3156, + "GATTAA": 3157, + "GATTAT": 3158, + "GATTAC": 3159, + "GATTAG": 3160, + "GATTTA": 3161, + "GATTTT": 3162, + "GATTTC": 3163, + "GATTTG": 3164, + "GATTCA": 3165, + "GATTCT": 3166, + "GATTCC": 3167, + "GATTCG": 3168, + "GATTGA": 3169, + "GATTGT": 3170, + "GATTGC": 3171, + "GATTGG": 3172, + "GATCAA": 3173, + "GATCAT": 3174, + "GATCAC": 3175, + "GATCAG": 3176, + "GATCTA": 3177, + "GATCTT": 3178, + "GATCTC": 3179, + "GATCTG": 3180, + "GATCCA": 3181, + "GATCCT": 3182, + "GATCCC": 3183, + "GATCCG": 3184, + "GATCGA": 3185, + "GATCGT": 3186, + "GATCGC": 3187, + "GATCGG": 3188, + "GATGAA": 3189, + "GATGAT": 3190, + "GATGAC": 3191, + "GATGAG": 3192, + "GATGTA": 3193, + "GATGTT": 3194, + "GATGTC": 3195, + "GATGTG": 3196, + "GATGCA": 3197, + "GATGCT": 3198, + "GATGCC": 3199, + "GATGCG": 3200, + "GATGGA": 3201, + "GATGGT": 3202, + "GATGGC": 3203, + "GATGGG": 3204, + "GACAAA": 3205, + "GACAAT": 3206, + "GACAAC": 3207, + "GACAAG": 3208, + "GACATA": 3209, + "GACATT": 3210, + "GACATC": 3211, + "GACATG": 3212, + "GACACA": 3213, + "GACACT": 3214, + "GACACC": 3215, + "GACACG": 3216, + "GACAGA": 3217, + "GACAGT": 3218, + "GACAGC": 3219, + "GACAGG": 3220, + "GACTAA": 3221, + "GACTAT": 3222, + "GACTAC": 3223, + "GACTAG": 3224, + "GACTTA": 3225, + "GACTTT": 3226, + "GACTTC": 3227, + "GACTTG": 3228, + "GACTCA": 3229, + "GACTCT": 3230, + "GACTCC": 3231, + "GACTCG": 3232, + "GACTGA": 3233, + "GACTGT": 3234, + "GACTGC": 3235, + "GACTGG": 3236, + "GACCAA": 3237, + "GACCAT": 3238, + "GACCAC": 3239, + "GACCAG": 3240, + "GACCTA": 3241, + "GACCTT": 3242, + "GACCTC": 3243, + "GACCTG": 3244, + "GACCCA": 3245, + "GACCCT": 3246, + "GACCCC": 3247, + "GACCCG": 3248, + "GACCGA": 3249, + "GACCGT": 3250, + "GACCGC": 3251, + "GACCGG": 3252, + "GACGAA": 3253, + "GACGAT": 3254, + "GACGAC": 3255, + "GACGAG": 3256, + "GACGTA": 3257, + "GACGTT": 3258, + "GACGTC": 3259, + "GACGTG": 3260, + "GACGCA": 3261, + "GACGCT": 3262, + "GACGCC": 3263, + "GACGCG": 3264, + "GACGGA": 3265, + "GACGGT": 3266, + "GACGGC": 3267, + "GACGGG": 3268, + "GAGAAA": 3269, + "GAGAAT": 3270, + "GAGAAC": 3271, + "GAGAAG": 3272, + "GAGATA": 3273, + "GAGATT": 3274, + "GAGATC": 3275, + "GAGATG": 3276, + "GAGACA": 3277, + "GAGACT": 3278, + "GAGACC": 3279, + "GAGACG": 3280, + "GAGAGA": 3281, + "GAGAGT": 3282, + "GAGAGC": 3283, + "GAGAGG": 3284, + "GAGTAA": 3285, + "GAGTAT": 3286, + "GAGTAC": 3287, + "GAGTAG": 3288, + "GAGTTA": 3289, + "GAGTTT": 3290, + "GAGTTC": 3291, + "GAGTTG": 3292, + "GAGTCA": 3293, + "GAGTCT": 3294, + "GAGTCC": 3295, + "GAGTCG": 3296, + "GAGTGA": 3297, + "GAGTGT": 3298, + "GAGTGC": 3299, + "GAGTGG": 3300, + "GAGCAA": 3301, + "GAGCAT": 3302, + "GAGCAC": 3303, + "GAGCAG": 3304, + "GAGCTA": 3305, + "GAGCTT": 3306, + "GAGCTC": 3307, + "GAGCTG": 3308, + "GAGCCA": 3309, + "GAGCCT": 3310, + "GAGCCC": 3311, + "GAGCCG": 3312, + "GAGCGA": 3313, + "GAGCGT": 3314, + "GAGCGC": 3315, + "GAGCGG": 3316, + "GAGGAA": 3317, + "GAGGAT": 3318, + "GAGGAC": 3319, + "GAGGAG": 3320, + "GAGGTA": 3321, + "GAGGTT": 3322, + "GAGGTC": 3323, + "GAGGTG": 3324, + "GAGGCA": 3325, + "GAGGCT": 3326, + "GAGGCC": 3327, + "GAGGCG": 3328, + "GAGGGA": 3329, + "GAGGGT": 3330, + "GAGGGC": 3331, + "GAGGGG": 3332, + "GTAAAA": 3333, + "GTAAAT": 3334, + "GTAAAC": 3335, + "GTAAAG": 3336, + "GTAATA": 3337, + "GTAATT": 3338, + "GTAATC": 3339, + "GTAATG": 3340, + "GTAACA": 3341, + "GTAACT": 3342, + "GTAACC": 3343, + "GTAACG": 3344, + "GTAAGA": 3345, + "GTAAGT": 3346, + "GTAAGC": 3347, + "GTAAGG": 3348, + "GTATAA": 3349, + "GTATAT": 3350, + "GTATAC": 3351, + "GTATAG": 3352, + "GTATTA": 3353, + "GTATTT": 3354, + "GTATTC": 3355, + "GTATTG": 3356, + "GTATCA": 3357, + "GTATCT": 3358, + "GTATCC": 3359, + "GTATCG": 3360, + "GTATGA": 3361, + "GTATGT": 3362, + "GTATGC": 3363, + "GTATGG": 3364, + "GTACAA": 3365, + "GTACAT": 3366, + "GTACAC": 3367, + "GTACAG": 3368, + "GTACTA": 3369, + "GTACTT": 3370, + "GTACTC": 3371, + "GTACTG": 3372, + "GTACCA": 3373, + "GTACCT": 3374, + "GTACCC": 3375, + "GTACCG": 3376, + "GTACGA": 3377, + "GTACGT": 3378, + "GTACGC": 3379, + "GTACGG": 3380, + "GTAGAA": 3381, + "GTAGAT": 3382, + "GTAGAC": 3383, + "GTAGAG": 3384, + "GTAGTA": 3385, + "GTAGTT": 3386, + "GTAGTC": 3387, + "GTAGTG": 3388, + "GTAGCA": 3389, + "GTAGCT": 3390, + "GTAGCC": 3391, + "GTAGCG": 3392, + "GTAGGA": 3393, + "GTAGGT": 3394, + "GTAGGC": 3395, + "GTAGGG": 3396, + "GTTAAA": 3397, + "GTTAAT": 3398, + "GTTAAC": 3399, + "GTTAAG": 3400, + "GTTATA": 3401, + "GTTATT": 3402, + "GTTATC": 3403, + "GTTATG": 3404, + "GTTACA": 3405, + "GTTACT": 3406, + "GTTACC": 3407, + "GTTACG": 3408, + "GTTAGA": 3409, + "GTTAGT": 3410, + "GTTAGC": 3411, + "GTTAGG": 3412, + "GTTTAA": 3413, + "GTTTAT": 3414, + "GTTTAC": 3415, + "GTTTAG": 3416, + "GTTTTA": 3417, + "GTTTTT": 3418, + "GTTTTC": 3419, + "GTTTTG": 3420, + "GTTTCA": 3421, + "GTTTCT": 3422, + "GTTTCC": 3423, + "GTTTCG": 3424, + "GTTTGA": 3425, + "GTTTGT": 3426, + "GTTTGC": 3427, + "GTTTGG": 3428, + "GTTCAA": 3429, + "GTTCAT": 3430, + "GTTCAC": 3431, + "GTTCAG": 3432, + "GTTCTA": 3433, + "GTTCTT": 3434, + "GTTCTC": 3435, + "GTTCTG": 3436, + "GTTCCA": 3437, + "GTTCCT": 3438, + "GTTCCC": 3439, + "GTTCCG": 3440, + "GTTCGA": 3441, + "GTTCGT": 3442, + "GTTCGC": 3443, + "GTTCGG": 3444, + "GTTGAA": 3445, + "GTTGAT": 3446, + "GTTGAC": 3447, + "GTTGAG": 3448, + "GTTGTA": 3449, + "GTTGTT": 3450, + "GTTGTC": 3451, + "GTTGTG": 3452, + "GTTGCA": 3453, + "GTTGCT": 3454, + "GTTGCC": 3455, + "GTTGCG": 3456, + "GTTGGA": 3457, + "GTTGGT": 3458, + "GTTGGC": 3459, + "GTTGGG": 3460, + "GTCAAA": 3461, + "GTCAAT": 3462, + "GTCAAC": 3463, + "GTCAAG": 3464, + "GTCATA": 3465, + "GTCATT": 3466, + "GTCATC": 3467, + "GTCATG": 3468, + "GTCACA": 3469, + "GTCACT": 3470, + "GTCACC": 3471, + "GTCACG": 3472, + "GTCAGA": 3473, + "GTCAGT": 3474, + "GTCAGC": 3475, + "GTCAGG": 3476, + "GTCTAA": 3477, + "GTCTAT": 3478, + "GTCTAC": 3479, + "GTCTAG": 3480, + "GTCTTA": 3481, + "GTCTTT": 3482, + "GTCTTC": 3483, + "GTCTTG": 3484, + "GTCTCA": 3485, + "GTCTCT": 3486, + "GTCTCC": 3487, + "GTCTCG": 3488, + "GTCTGA": 3489, + "GTCTGT": 3490, + "GTCTGC": 3491, + "GTCTGG": 3492, + "GTCCAA": 3493, + "GTCCAT": 3494, + "GTCCAC": 3495, + "GTCCAG": 3496, + "GTCCTA": 3497, + "GTCCTT": 3498, + "GTCCTC": 3499, + "GTCCTG": 3500, + "GTCCCA": 3501, + "GTCCCT": 3502, + "GTCCCC": 3503, + "GTCCCG": 3504, + "GTCCGA": 3505, + "GTCCGT": 3506, + "GTCCGC": 3507, + "GTCCGG": 3508, + "GTCGAA": 3509, + "GTCGAT": 3510, + "GTCGAC": 3511, + "GTCGAG": 3512, + "GTCGTA": 3513, + "GTCGTT": 3514, + "GTCGTC": 3515, + "GTCGTG": 3516, + "GTCGCA": 3517, + "GTCGCT": 3518, + "GTCGCC": 3519, + "GTCGCG": 3520, + "GTCGGA": 3521, + "GTCGGT": 3522, + "GTCGGC": 3523, + "GTCGGG": 3524, + "GTGAAA": 3525, + "GTGAAT": 3526, + "GTGAAC": 3527, + "GTGAAG": 3528, + "GTGATA": 3529, + "GTGATT": 3530, + "GTGATC": 3531, + "GTGATG": 3532, + "GTGACA": 3533, + "GTGACT": 3534, + "GTGACC": 3535, + "GTGACG": 3536, + "GTGAGA": 3537, + "GTGAGT": 3538, + "GTGAGC": 3539, + "GTGAGG": 3540, + "GTGTAA": 3541, + "GTGTAT": 3542, + "GTGTAC": 3543, + "GTGTAG": 3544, + "GTGTTA": 3545, + "GTGTTT": 3546, + "GTGTTC": 3547, + "GTGTTG": 3548, + "GTGTCA": 3549, + "GTGTCT": 3550, + "GTGTCC": 3551, + "GTGTCG": 3552, + "GTGTGA": 3553, + "GTGTGT": 3554, + "GTGTGC": 3555, + "GTGTGG": 3556, + "GTGCAA": 3557, + "GTGCAT": 3558, + "GTGCAC": 3559, + "GTGCAG": 3560, + "GTGCTA": 3561, + "GTGCTT": 3562, + "GTGCTC": 3563, + "GTGCTG": 3564, + "GTGCCA": 3565, + "GTGCCT": 3566, + "GTGCCC": 3567, + "GTGCCG": 3568, + "GTGCGA": 3569, + "GTGCGT": 3570, + "GTGCGC": 3571, + "GTGCGG": 3572, + "GTGGAA": 3573, + "GTGGAT": 3574, + "GTGGAC": 3575, + "GTGGAG": 3576, + "GTGGTA": 3577, + "GTGGTT": 3578, + "GTGGTC": 3579, + "GTGGTG": 3580, + "GTGGCA": 3581, + "GTGGCT": 3582, + "GTGGCC": 3583, + "GTGGCG": 3584, + "GTGGGA": 3585, + "GTGGGT": 3586, + "GTGGGC": 3587, + "GTGGGG": 3588, + "GCAAAA": 3589, + "GCAAAT": 3590, + "GCAAAC": 3591, + "GCAAAG": 3592, + "GCAATA": 3593, + "GCAATT": 3594, + "GCAATC": 3595, + "GCAATG": 3596, + "GCAACA": 3597, + "GCAACT": 3598, + "GCAACC": 3599, + "GCAACG": 3600, + "GCAAGA": 3601, + "GCAAGT": 3602, + "GCAAGC": 3603, + "GCAAGG": 3604, + "GCATAA": 3605, + "GCATAT": 3606, + "GCATAC": 3607, + "GCATAG": 3608, + "GCATTA": 3609, + "GCATTT": 3610, + "GCATTC": 3611, + "GCATTG": 3612, + "GCATCA": 3613, + "GCATCT": 3614, + "GCATCC": 3615, + "GCATCG": 3616, + "GCATGA": 3617, + "GCATGT": 3618, + "GCATGC": 3619, + "GCATGG": 3620, + "GCACAA": 3621, + "GCACAT": 3622, + "GCACAC": 3623, + "GCACAG": 3624, + "GCACTA": 3625, + "GCACTT": 3626, + "GCACTC": 3627, + "GCACTG": 3628, + "GCACCA": 3629, + "GCACCT": 3630, + "GCACCC": 3631, + "GCACCG": 3632, + "GCACGA": 3633, + "GCACGT": 3634, + "GCACGC": 3635, + "GCACGG": 3636, + "GCAGAA": 3637, + "GCAGAT": 3638, + "GCAGAC": 3639, + "GCAGAG": 3640, + "GCAGTA": 3641, + "GCAGTT": 3642, + "GCAGTC": 3643, + "GCAGTG": 3644, + "GCAGCA": 3645, + "GCAGCT": 3646, + "GCAGCC": 3647, + "GCAGCG": 3648, + "GCAGGA": 3649, + "GCAGGT": 3650, + "GCAGGC": 3651, + "GCAGGG": 3652, + "GCTAAA": 3653, + "GCTAAT": 3654, + "GCTAAC": 3655, + "GCTAAG": 3656, + "GCTATA": 3657, + "GCTATT": 3658, + "GCTATC": 3659, + "GCTATG": 3660, + "GCTACA": 3661, + "GCTACT": 3662, + "GCTACC": 3663, + "GCTACG": 3664, + "GCTAGA": 3665, + "GCTAGT": 3666, + "GCTAGC": 3667, + "GCTAGG": 3668, + "GCTTAA": 3669, + "GCTTAT": 3670, + "GCTTAC": 3671, + "GCTTAG": 3672, + "GCTTTA": 3673, + "GCTTTT": 3674, + "GCTTTC": 3675, + "GCTTTG": 3676, + "GCTTCA": 3677, + "GCTTCT": 3678, + "GCTTCC": 3679, + "GCTTCG": 3680, + "GCTTGA": 3681, + "GCTTGT": 3682, + "GCTTGC": 3683, + "GCTTGG": 3684, + "GCTCAA": 3685, + "GCTCAT": 3686, + "GCTCAC": 3687, + "GCTCAG": 3688, + "GCTCTA": 3689, + "GCTCTT": 3690, + "GCTCTC": 3691, + "GCTCTG": 3692, + "GCTCCA": 3693, + "GCTCCT": 3694, + "GCTCCC": 3695, + "GCTCCG": 3696, + "GCTCGA": 3697, + "GCTCGT": 3698, + "GCTCGC": 3699, + "GCTCGG": 3700, + "GCTGAA": 3701, + "GCTGAT": 3702, + "GCTGAC": 3703, + "GCTGAG": 3704, + "GCTGTA": 3705, + "GCTGTT": 3706, + "GCTGTC": 3707, + "GCTGTG": 3708, + "GCTGCA": 3709, + "GCTGCT": 3710, + "GCTGCC": 3711, + "GCTGCG": 3712, + "GCTGGA": 3713, + "GCTGGT": 3714, + "GCTGGC": 3715, + "GCTGGG": 3716, + "GCCAAA": 3717, + "GCCAAT": 3718, + "GCCAAC": 3719, + "GCCAAG": 3720, + "GCCATA": 3721, + "GCCATT": 3722, + "GCCATC": 3723, + "GCCATG": 3724, + "GCCACA": 3725, + "GCCACT": 3726, + "GCCACC": 3727, + "GCCACG": 3728, + "GCCAGA": 3729, + "GCCAGT": 3730, + "GCCAGC": 3731, + "GCCAGG": 3732, + "GCCTAA": 3733, + "GCCTAT": 3734, + "GCCTAC": 3735, + "GCCTAG": 3736, + "GCCTTA": 3737, + "GCCTTT": 3738, + "GCCTTC": 3739, + "GCCTTG": 3740, + "GCCTCA": 3741, + "GCCTCT": 3742, + "GCCTCC": 3743, + "GCCTCG": 3744, + "GCCTGA": 3745, + "GCCTGT": 3746, + "GCCTGC": 3747, + "GCCTGG": 3748, + "GCCCAA": 3749, + "GCCCAT": 3750, + "GCCCAC": 3751, + "GCCCAG": 3752, + "GCCCTA": 3753, + "GCCCTT": 3754, + "GCCCTC": 3755, + "GCCCTG": 3756, + "GCCCCA": 3757, + "GCCCCT": 3758, + "GCCCCC": 3759, + "GCCCCG": 3760, + "GCCCGA": 3761, + "GCCCGT": 3762, + "GCCCGC": 3763, + "GCCCGG": 3764, + "GCCGAA": 3765, + "GCCGAT": 3766, + "GCCGAC": 3767, + "GCCGAG": 3768, + "GCCGTA": 3769, + "GCCGTT": 3770, + "GCCGTC": 3771, + "GCCGTG": 3772, + "GCCGCA": 3773, + "GCCGCT": 3774, + "GCCGCC": 3775, + "GCCGCG": 3776, + "GCCGGA": 3777, + "GCCGGT": 3778, + "GCCGGC": 3779, + "GCCGGG": 3780, + "GCGAAA": 3781, + "GCGAAT": 3782, + "GCGAAC": 3783, + "GCGAAG": 3784, + "GCGATA": 3785, + "GCGATT": 3786, + "GCGATC": 3787, + "GCGATG": 3788, + "GCGACA": 3789, + "GCGACT": 3790, + "GCGACC": 3791, + "GCGACG": 3792, + "GCGAGA": 3793, + "GCGAGT": 3794, + "GCGAGC": 3795, + "GCGAGG": 3796, + "GCGTAA": 3797, + "GCGTAT": 3798, + "GCGTAC": 3799, + "GCGTAG": 3800, + "GCGTTA": 3801, + "GCGTTT": 3802, + "GCGTTC": 3803, + "GCGTTG": 3804, + "GCGTCA": 3805, + "GCGTCT": 3806, + "GCGTCC": 3807, + "GCGTCG": 3808, + "GCGTGA": 3809, + "GCGTGT": 3810, + "GCGTGC": 3811, + "GCGTGG": 3812, + "GCGCAA": 3813, + "GCGCAT": 3814, + "GCGCAC": 3815, + "GCGCAG": 3816, + "GCGCTA": 3817, + "GCGCTT": 3818, + "GCGCTC": 3819, + "GCGCTG": 3820, + "GCGCCA": 3821, + "GCGCCT": 3822, + "GCGCCC": 3823, + "GCGCCG": 3824, + "GCGCGA": 3825, + "GCGCGT": 3826, + "GCGCGC": 3827, + "GCGCGG": 3828, + "GCGGAA": 3829, + "GCGGAT": 3830, + "GCGGAC": 3831, + "GCGGAG": 3832, + "GCGGTA": 3833, + "GCGGTT": 3834, + "GCGGTC": 3835, + "GCGGTG": 3836, + "GCGGCA": 3837, + "GCGGCT": 3838, + "GCGGCC": 3839, + "GCGGCG": 3840, + "GCGGGA": 3841, + "GCGGGT": 3842, + "GCGGGC": 3843, + "GCGGGG": 3844, + "GGAAAA": 3845, + "GGAAAT": 3846, + "GGAAAC": 3847, + "GGAAAG": 3848, + "GGAATA": 3849, + "GGAATT": 3850, + "GGAATC": 3851, + "GGAATG": 3852, + "GGAACA": 3853, + "GGAACT": 3854, + "GGAACC": 3855, + "GGAACG": 3856, + "GGAAGA": 3857, + "GGAAGT": 3858, + "GGAAGC": 3859, + "GGAAGG": 3860, + "GGATAA": 3861, + "GGATAT": 3862, + "GGATAC": 3863, + "GGATAG": 3864, + "GGATTA": 3865, + "GGATTT": 3866, + "GGATTC": 3867, + "GGATTG": 3868, + "GGATCA": 3869, + "GGATCT": 3870, + "GGATCC": 3871, + "GGATCG": 3872, + "GGATGA": 3873, + "GGATGT": 3874, + "GGATGC": 3875, + "GGATGG": 3876, + "GGACAA": 3877, + "GGACAT": 3878, + "GGACAC": 3879, + "GGACAG": 3880, + "GGACTA": 3881, + "GGACTT": 3882, + "GGACTC": 3883, + "GGACTG": 3884, + "GGACCA": 3885, + "GGACCT": 3886, + "GGACCC": 3887, + "GGACCG": 3888, + "GGACGA": 3889, + "GGACGT": 3890, + "GGACGC": 3891, + "GGACGG": 3892, + "GGAGAA": 3893, + "GGAGAT": 3894, + "GGAGAC": 3895, + "GGAGAG": 3896, + "GGAGTA": 3897, + "GGAGTT": 3898, + "GGAGTC": 3899, + "GGAGTG": 3900, + "GGAGCA": 3901, + "GGAGCT": 3902, + "GGAGCC": 3903, + "GGAGCG": 3904, + "GGAGGA": 3905, + "GGAGGT": 3906, + "GGAGGC": 3907, + "GGAGGG": 3908, + "GGTAAA": 3909, + "GGTAAT": 3910, + "GGTAAC": 3911, + "GGTAAG": 3912, + "GGTATA": 3913, + "GGTATT": 3914, + "GGTATC": 3915, + "GGTATG": 3916, + "GGTACA": 3917, + "GGTACT": 3918, + "GGTACC": 3919, + "GGTACG": 3920, + "GGTAGA": 3921, + "GGTAGT": 3922, + "GGTAGC": 3923, + "GGTAGG": 3924, + "GGTTAA": 3925, + "GGTTAT": 3926, + "GGTTAC": 3927, + "GGTTAG": 3928, + "GGTTTA": 3929, + "GGTTTT": 3930, + "GGTTTC": 3931, + "GGTTTG": 3932, + "GGTTCA": 3933, + "GGTTCT": 3934, + "GGTTCC": 3935, + "GGTTCG": 3936, + "GGTTGA": 3937, + "GGTTGT": 3938, + "GGTTGC": 3939, + "GGTTGG": 3940, + "GGTCAA": 3941, + "GGTCAT": 3942, + "GGTCAC": 3943, + "GGTCAG": 3944, + "GGTCTA": 3945, + "GGTCTT": 3946, + "GGTCTC": 3947, + "GGTCTG": 3948, + "GGTCCA": 3949, + "GGTCCT": 3950, + "GGTCCC": 3951, + "GGTCCG": 3952, + "GGTCGA": 3953, + "GGTCGT": 3954, + "GGTCGC": 3955, + "GGTCGG": 3956, + "GGTGAA": 3957, + "GGTGAT": 3958, + "GGTGAC": 3959, + "GGTGAG": 3960, + "GGTGTA": 3961, + "GGTGTT": 3962, + "GGTGTC": 3963, + "GGTGTG": 3964, + "GGTGCA": 3965, + "GGTGCT": 3966, + "GGTGCC": 3967, + "GGTGCG": 3968, + "GGTGGA": 3969, + "GGTGGT": 3970, + "GGTGGC": 3971, + "GGTGGG": 3972, + "GGCAAA": 3973, + "GGCAAT": 3974, + "GGCAAC": 3975, + "GGCAAG": 3976, + "GGCATA": 3977, + "GGCATT": 3978, + "GGCATC": 3979, + "GGCATG": 3980, + "GGCACA": 3981, + "GGCACT": 3982, + "GGCACC": 3983, + "GGCACG": 3984, + "GGCAGA": 3985, + "GGCAGT": 3986, + "GGCAGC": 3987, + "GGCAGG": 3988, + "GGCTAA": 3989, + "GGCTAT": 3990, + "GGCTAC": 3991, + "GGCTAG": 3992, + "GGCTTA": 3993, + "GGCTTT": 3994, + "GGCTTC": 3995, + "GGCTTG": 3996, + "GGCTCA": 3997, + "GGCTCT": 3998, + "GGCTCC": 3999, + "GGCTCG": 4000, + "GGCTGA": 4001, + "GGCTGT": 4002, + "GGCTGC": 4003, + "GGCTGG": 4004, + "GGCCAA": 4005, + "GGCCAT": 4006, + "GGCCAC": 4007, + "GGCCAG": 4008, + "GGCCTA": 4009, + "GGCCTT": 4010, + "GGCCTC": 4011, + "GGCCTG": 4012, + "GGCCCA": 4013, + "GGCCCT": 4014, + "GGCCCC": 4015, + "GGCCCG": 4016, + "GGCCGA": 4017, + "GGCCGT": 4018, + "GGCCGC": 4019, + "GGCCGG": 4020, + "GGCGAA": 4021, + "GGCGAT": 4022, + "GGCGAC": 4023, + "GGCGAG": 4024, + "GGCGTA": 4025, + "GGCGTT": 4026, + "GGCGTC": 4027, + "GGCGTG": 4028, + "GGCGCA": 4029, + "GGCGCT": 4030, + "GGCGCC": 4031, + "GGCGCG": 4032, + "GGCGGA": 4033, + "GGCGGT": 4034, + "GGCGGC": 4035, + "GGCGGG": 4036, + "GGGAAA": 4037, + "GGGAAT": 4038, + "GGGAAC": 4039, + "GGGAAG": 4040, + "GGGATA": 4041, + "GGGATT": 4042, + "GGGATC": 4043, + "GGGATG": 4044, + "GGGACA": 4045, + "GGGACT": 4046, + "GGGACC": 4047, + "GGGACG": 4048, + "GGGAGA": 4049, + "GGGAGT": 4050, + "GGGAGC": 4051, + "GGGAGG": 4052, + "GGGTAA": 4053, + "GGGTAT": 4054, + "GGGTAC": 4055, + "GGGTAG": 4056, + "GGGTTA": 4057, + "GGGTTT": 4058, + "GGGTTC": 4059, + "GGGTTG": 4060, + "GGGTCA": 4061, + "GGGTCT": 4062, + "GGGTCC": 4063, + "GGGTCG": 4064, + "GGGTGA": 4065, + "GGGTGT": 4066, + "GGGTGC": 4067, + "GGGTGG": 4068, + "GGGCAA": 4069, + "GGGCAT": 4070, + "GGGCAC": 4071, + "GGGCAG": 4072, + "GGGCTA": 4073, + "GGGCTT": 4074, + "GGGCTC": 4075, + "GGGCTG": 4076, + "GGGCCA": 4077, + "GGGCCT": 4078, + "GGGCCC": 4079, + "GGGCCG": 4080, + "GGGCGA": 4081, + "GGGCGT": 4082, + "GGGCGC": 4083, + "GGGCGG": 4084, + "GGGGAA": 4085, + "GGGGAT": 4086, + "GGGGAC": 4087, + "GGGGAG": 4088, + "GGGGTA": 4089, + "GGGGTT": 4090, + "GGGGTC": 4091, + "GGGGTG": 4092, + "GGGGCA": 4093, + "GGGGCT": 4094, + "GGGGCC": 4095, + "GGGGCG": 4096, + "GGGGGA": 4097, + "GGGGGT": 4098, + "GGGGGC": 4099, + "GGGGGG": 4100 + } + } +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-3000/tokenizer_config.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-3000/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..604e2b1b29af1dcb3920d2828b6a90276e5e3d02 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-3000/tokenizer_config.json @@ -0,0 +1,58 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "[UNK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "[CLS]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "[SEP]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "4": { + "content": "[MASK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "clean_up_tokenization_spaces": true, + "cls_token": "[CLS]", + "do_basic_tokenize": true, + "do_lower_case": false, + "mask_token": "[MASK]", + "max_len": 512, + "model_max_length": 512, + "never_split": null, + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "strip_accents": null, + "tokenize_chinese_chars": true, + "tokenizer_class": "BertTokenizer", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-3000/trainer_state.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-3000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..ae673f6f88fb86209120a95df7cf43d07e597230 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-3000/trainer_state.json @@ -0,0 +1,873 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 12.0, + "eval_steps": 500, + "global_step": 3000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.1, + "grad_norm": 1.1046091318130493, + "learning_rate": 5.319148936170213e-05, + "loss": 0.6562, + "step": 25 + }, + { + "epoch": 0.2, + "grad_norm": 0.4066573977470398, + "learning_rate": 0.00010638297872340425, + "loss": 0.5373, + "step": 50 + }, + { + "epoch": 0.3, + "grad_norm": 0.7340325713157654, + "learning_rate": 0.00015957446808510637, + "loss": 0.4995, + "step": 75 + }, + { + "epoch": 0.4, + "grad_norm": 2.134040117263794, + "learning_rate": 0.0002127659574468085, + "loss": 0.4871, + "step": 100 + }, + { + "epoch": 0.5, + "grad_norm": 2.911975860595703, + "learning_rate": 0.00026595744680851064, + "loss": 0.4507, + "step": 125 + }, + { + "epoch": 0.6, + "grad_norm": 2.040365219116211, + "learning_rate": 0.00031914893617021275, + "loss": 0.3944, + "step": 150 + }, + { + "epoch": 0.7, + "grad_norm": 5.643661022186279, + "learning_rate": 0.0003723404255319149, + "loss": 0.3863, + "step": 175 + }, + { + "epoch": 0.8, + "grad_norm": 3.882492780685425, + "learning_rate": 0.0003992081821181128, + "loss": 0.3875, + "step": 200 + }, + { + "epoch": 0.9, + "grad_norm": 5.87437105178833, + "learning_rate": 0.0003975585615308479, + "loss": 0.3856, + "step": 225 + }, + { + "epoch": 1.0, + "grad_norm": 4.611206531524658, + "learning_rate": 0.00039590894094358297, + "loss": 0.3866, + "step": 250 + }, + { + "epoch": 1.1, + "grad_norm": 3.6404929161071777, + "learning_rate": 0.0003942593203563181, + "loss": 0.3437, + "step": 275 + }, + { + "epoch": 1.2, + "grad_norm": 6.469444751739502, + "learning_rate": 0.0003926096997690532, + "loss": 0.3192, + "step": 300 + }, + { + "epoch": 1.3, + "grad_norm": 3.6251447200775146, + "learning_rate": 0.00039096007918178817, + "loss": 0.349, + "step": 325 + }, + { + "epoch": 1.4, + "grad_norm": 2.854794979095459, + "learning_rate": 0.0003893104585945233, + "loss": 0.3274, + "step": 350 + }, + { + "epoch": 1.5, + "grad_norm": 4.805097579956055, + "learning_rate": 0.0003876608380072583, + "loss": 0.3279, + "step": 375 + }, + { + "epoch": 1.6, + "grad_norm": 2.9891581535339355, + "learning_rate": 0.00038601121741999343, + "loss": 0.3166, + "step": 400 + }, + { + "epoch": 1.7, + "grad_norm": 5.333752155303955, + "learning_rate": 0.0003843615968327285, + "loss": 0.2856, + "step": 425 + }, + { + "epoch": 1.8, + "grad_norm": 2.615621566772461, + "learning_rate": 0.0003827119762454636, + "loss": 0.3286, + "step": 450 + }, + { + "epoch": 1.9, + "grad_norm": 2.4228415489196777, + "learning_rate": 0.00038106235565819863, + "loss": 0.2812, + "step": 475 + }, + { + "epoch": 2.0, + "grad_norm": 3.6015517711639404, + "learning_rate": 0.0003794127350709337, + "loss": 0.2884, + "step": 500 + }, + { + "epoch": 2.1, + "grad_norm": 2.7867860794067383, + "learning_rate": 0.0003777631144836688, + "loss": 0.2582, + "step": 525 + }, + { + "epoch": 2.2, + "grad_norm": 10.079645156860352, + "learning_rate": 0.00037611349389640383, + "loss": 0.245, + "step": 550 + }, + { + "epoch": 2.3, + "grad_norm": 2.475917100906372, + "learning_rate": 0.00037446387330913894, + "loss": 0.2615, + "step": 575 + }, + { + "epoch": 2.4, + "grad_norm": 8.098592758178711, + "learning_rate": 0.000372814252721874, + "loss": 0.2484, + "step": 600 + }, + { + "epoch": 2.5, + "grad_norm": 2.318723201751709, + "learning_rate": 0.00037116463213460903, + "loss": 0.2056, + "step": 625 + }, + { + "epoch": 2.6, + "grad_norm": 3.616283893585205, + "learning_rate": 0.00036951501154734414, + "loss": 0.2358, + "step": 650 + }, + { + "epoch": 2.7, + "grad_norm": 6.419492721557617, + "learning_rate": 0.0003678653909600792, + "loss": 0.2204, + "step": 675 + }, + { + "epoch": 2.8, + "grad_norm": 2.9333388805389404, + "learning_rate": 0.0003662157703728143, + "loss": 0.1857, + "step": 700 + }, + { + "epoch": 2.9, + "grad_norm": 7.691112995147705, + "learning_rate": 0.00036456614978554934, + "loss": 0.2054, + "step": 725 + }, + { + "epoch": 3.0, + "grad_norm": 4.908492565155029, + "learning_rate": 0.00036291652919828444, + "loss": 0.1956, + "step": 750 + }, + { + "epoch": 3.1, + "grad_norm": 1.8923287391662598, + "learning_rate": 0.00036126690861101944, + "loss": 0.1708, + "step": 775 + }, + { + "epoch": 3.2, + "grad_norm": 4.403504848480225, + "learning_rate": 0.00035961728802375454, + "loss": 0.1441, + "step": 800 + }, + { + "epoch": 3.3, + "grad_norm": 4.117386817932129, + "learning_rate": 0.0003579676674364896, + "loss": 0.1586, + "step": 825 + }, + { + "epoch": 3.4, + "grad_norm": 7.119816303253174, + "learning_rate": 0.0003563180468492247, + "loss": 0.1472, + "step": 850 + }, + { + "epoch": 3.5, + "grad_norm": 3.5796430110931396, + "learning_rate": 0.0003546684262619598, + "loss": 0.1802, + "step": 875 + }, + { + "epoch": 3.6, + "grad_norm": 2.97688889503479, + "learning_rate": 0.00035301880567469485, + "loss": 0.1578, + "step": 900 + }, + { + "epoch": 3.7, + "grad_norm": 3.716148614883423, + "learning_rate": 0.0003513691850874299, + "loss": 0.1662, + "step": 925 + }, + { + "epoch": 3.8, + "grad_norm": 6.1249566078186035, + "learning_rate": 0.00034971956450016495, + "loss": 0.159, + "step": 950 + }, + { + "epoch": 3.9, + "grad_norm": 3.0592427253723145, + "learning_rate": 0.00034806994391290005, + "loss": 0.148, + "step": 975 + }, + { + "epoch": 4.0, + "grad_norm": 4.467265605926514, + "learning_rate": 0.0003464203233256351, + "loss": 0.1442, + "step": 1000 + }, + { + "epoch": 4.1, + "grad_norm": 2.7223546504974365, + "learning_rate": 0.0003447707027383702, + "loss": 0.1339, + "step": 1025 + }, + { + "epoch": 4.2, + "grad_norm": 3.698854923248291, + "learning_rate": 0.00034312108215110525, + "loss": 0.1381, + "step": 1050 + }, + { + "epoch": 4.3, + "grad_norm": 7.418384552001953, + "learning_rate": 0.0003414714615638403, + "loss": 0.1158, + "step": 1075 + }, + { + "epoch": 4.4, + "grad_norm": 1.2887814044952393, + "learning_rate": 0.0003398218409765754, + "loss": 0.0855, + "step": 1100 + }, + { + "epoch": 4.5, + "grad_norm": 0.838731586933136, + "learning_rate": 0.00033817222038931045, + "loss": 0.1272, + "step": 1125 + }, + { + "epoch": 4.6, + "grad_norm": 5.912592887878418, + "learning_rate": 0.00033652259980204556, + "loss": 0.1156, + "step": 1150 + }, + { + "epoch": 4.7, + "grad_norm": 8.49521255493164, + "learning_rate": 0.0003348729792147806, + "loss": 0.1182, + "step": 1175 + }, + { + "epoch": 4.8, + "grad_norm": 10.278315544128418, + "learning_rate": 0.0003332233586275157, + "loss": 0.1071, + "step": 1200 + }, + { + "epoch": 4.9, + "grad_norm": 0.724703311920166, + "learning_rate": 0.00033157373804025076, + "loss": 0.11, + "step": 1225 + }, + { + "epoch": 5.0, + "grad_norm": 1.6808199882507324, + "learning_rate": 0.0003299241174529858, + "loss": 0.1101, + "step": 1250 + }, + { + "epoch": 5.1, + "grad_norm": 0.1038585901260376, + "learning_rate": 0.0003282744968657209, + "loss": 0.0791, + "step": 1275 + }, + { + "epoch": 5.2, + "grad_norm": 0.15866787731647491, + "learning_rate": 0.00032662487627845596, + "loss": 0.0638, + "step": 1300 + }, + { + "epoch": 5.3, + "grad_norm": 6.748287200927734, + "learning_rate": 0.00032497525569119106, + "loss": 0.0961, + "step": 1325 + }, + { + "epoch": 5.4, + "grad_norm": 1.5960336923599243, + "learning_rate": 0.0003233256351039261, + "loss": 0.1057, + "step": 1350 + }, + { + "epoch": 5.5, + "grad_norm": 3.3678526878356934, + "learning_rate": 0.00032167601451666116, + "loss": 0.1092, + "step": 1375 + }, + { + "epoch": 5.6, + "grad_norm": 4.743562698364258, + "learning_rate": 0.00032002639392939627, + "loss": 0.1054, + "step": 1400 + }, + { + "epoch": 5.7, + "grad_norm": 12.860156059265137, + "learning_rate": 0.0003183767733421313, + "loss": 0.1009, + "step": 1425 + }, + { + "epoch": 5.8, + "grad_norm": 0.8820792436599731, + "learning_rate": 0.0003167271527548664, + "loss": 0.1219, + "step": 1450 + }, + { + "epoch": 5.9, + "grad_norm": 2.732572317123413, + "learning_rate": 0.00031507753216760147, + "loss": 0.0682, + "step": 1475 + }, + { + "epoch": 6.0, + "grad_norm": 9.810711860656738, + "learning_rate": 0.00031342791158033657, + "loss": 0.1134, + "step": 1500 + }, + { + "epoch": 6.1, + "grad_norm": 6.699489593505859, + "learning_rate": 0.00031177829099307157, + "loss": 0.0557, + "step": 1525 + }, + { + "epoch": 6.2, + "grad_norm": 14.691123008728027, + "learning_rate": 0.00031012867040580667, + "loss": 0.0599, + "step": 1550 + }, + { + "epoch": 6.3, + "grad_norm": 24.44135093688965, + "learning_rate": 0.0003084790498185417, + "loss": 0.0818, + "step": 1575 + }, + { + "epoch": 6.4, + "grad_norm": 0.033995576202869415, + "learning_rate": 0.0003068294292312768, + "loss": 0.0692, + "step": 1600 + }, + { + "epoch": 6.5, + "grad_norm": 6.722524642944336, + "learning_rate": 0.0003051798086440119, + "loss": 0.0702, + "step": 1625 + }, + { + "epoch": 6.6, + "grad_norm": 2.9519436359405518, + "learning_rate": 0.000303530188056747, + "loss": 0.073, + "step": 1650 + }, + { + "epoch": 6.7, + "grad_norm": 1.5118328332901, + "learning_rate": 0.000301880567469482, + "loss": 0.0926, + "step": 1675 + }, + { + "epoch": 6.8, + "grad_norm": 2.1923201084136963, + "learning_rate": 0.0003002309468822171, + "loss": 0.0821, + "step": 1700 + }, + { + "epoch": 6.9, + "grad_norm": 3.34309458732605, + "learning_rate": 0.0002985813262949522, + "loss": 0.0827, + "step": 1725 + }, + { + "epoch": 7.0, + "grad_norm": 10.289947509765625, + "learning_rate": 0.00029693170570768723, + "loss": 0.1162, + "step": 1750 + }, + { + "epoch": 7.1, + "grad_norm": 0.051936667412519455, + "learning_rate": 0.00029528208512042233, + "loss": 0.0626, + "step": 1775 + }, + { + "epoch": 7.2, + "grad_norm": 1.8913339376449585, + "learning_rate": 0.0002936324645331574, + "loss": 0.0657, + "step": 1800 + }, + { + "epoch": 7.3, + "grad_norm": 15.073695182800293, + "learning_rate": 0.00029198284394589243, + "loss": 0.0539, + "step": 1825 + }, + { + "epoch": 7.4, + "grad_norm": 4.514949798583984, + "learning_rate": 0.00029033322335862753, + "loss": 0.056, + "step": 1850 + }, + { + "epoch": 7.5, + "grad_norm": 0.46094000339508057, + "learning_rate": 0.0002886836027713626, + "loss": 0.0667, + "step": 1875 + }, + { + "epoch": 7.6, + "grad_norm": 4.682873725891113, + "learning_rate": 0.0002870339821840977, + "loss": 0.0673, + "step": 1900 + }, + { + "epoch": 7.7, + "grad_norm": 3.581599235534668, + "learning_rate": 0.00028538436159683273, + "loss": 0.0699, + "step": 1925 + }, + { + "epoch": 7.8, + "grad_norm": 3.2518703937530518, + "learning_rate": 0.00028373474100956784, + "loss": 0.0628, + "step": 1950 + }, + { + "epoch": 7.9, + "grad_norm": 0.24071219563484192, + "learning_rate": 0.0002820851204223029, + "loss": 0.0747, + "step": 1975 + }, + { + "epoch": 8.0, + "grad_norm": 2.8545596599578857, + "learning_rate": 0.00028043549983503794, + "loss": 0.0688, + "step": 2000 + }, + { + "epoch": 8.1, + "grad_norm": 0.049499694257974625, + "learning_rate": 0.00027878587924777304, + "loss": 0.0563, + "step": 2025 + }, + { + "epoch": 8.2, + "grad_norm": 22.112478256225586, + "learning_rate": 0.0002771362586605081, + "loss": 0.0545, + "step": 2050 + }, + { + "epoch": 8.3, + "grad_norm": 20.87473487854004, + "learning_rate": 0.0002754866380732432, + "loss": 0.0256, + "step": 2075 + }, + { + "epoch": 8.4, + "grad_norm": 2.317768096923828, + "learning_rate": 0.00027383701748597824, + "loss": 0.0753, + "step": 2100 + }, + { + "epoch": 8.5, + "grad_norm": 20.241336822509766, + "learning_rate": 0.0002721873968987133, + "loss": 0.0257, + "step": 2125 + }, + { + "epoch": 8.6, + "grad_norm": 14.994575500488281, + "learning_rate": 0.00027053777631144834, + "loss": 0.0748, + "step": 2150 + }, + { + "epoch": 8.7, + "grad_norm": 0.8268639445304871, + "learning_rate": 0.00026888815572418344, + "loss": 0.0805, + "step": 2175 + }, + { + "epoch": 8.8, + "grad_norm": 0.031071001663804054, + "learning_rate": 0.00026723853513691855, + "loss": 0.0603, + "step": 2200 + }, + { + "epoch": 8.9, + "grad_norm": 2.4572932720184326, + "learning_rate": 0.0002655889145496536, + "loss": 0.071, + "step": 2225 + }, + { + "epoch": 9.0, + "grad_norm": 0.21708321571350098, + "learning_rate": 0.0002639392939623887, + "loss": 0.0606, + "step": 2250 + }, + { + "epoch": 9.1, + "grad_norm": 14.928559303283691, + "learning_rate": 0.0002622896733751237, + "loss": 0.0443, + "step": 2275 + }, + { + "epoch": 9.2, + "grad_norm": 1.108826994895935, + "learning_rate": 0.0002606400527878588, + "loss": 0.0509, + "step": 2300 + }, + { + "epoch": 9.3, + "grad_norm": 0.05721910670399666, + "learning_rate": 0.00025899043220059385, + "loss": 0.0459, + "step": 2325 + }, + { + "epoch": 9.4, + "grad_norm": 7.81903600692749, + "learning_rate": 0.00025734081161332895, + "loss": 0.0535, + "step": 2350 + }, + { + "epoch": 9.5, + "grad_norm": 0.01475981343537569, + "learning_rate": 0.000255691191026064, + "loss": 0.0399, + "step": 2375 + }, + { + "epoch": 9.6, + "grad_norm": 5.496493339538574, + "learning_rate": 0.0002540415704387991, + "loss": 0.0744, + "step": 2400 + }, + { + "epoch": 9.7, + "grad_norm": 0.05929339677095413, + "learning_rate": 0.00025239194985153415, + "loss": 0.0416, + "step": 2425 + }, + { + "epoch": 9.8, + "grad_norm": 0.06961135566234589, + "learning_rate": 0.0002507423292642692, + "loss": 0.04, + "step": 2450 + }, + { + "epoch": 9.9, + "grad_norm": 0.02165246568620205, + "learning_rate": 0.0002490927086770043, + "loss": 0.0984, + "step": 2475 + }, + { + "epoch": 10.0, + "grad_norm": 25.155675888061523, + "learning_rate": 0.00024744308808973936, + "loss": 0.042, + "step": 2500 + }, + { + "epoch": 10.1, + "grad_norm": 8.566617965698242, + "learning_rate": 0.00024579346750247446, + "loss": 0.0385, + "step": 2525 + }, + { + "epoch": 10.2, + "grad_norm": 0.024992404505610466, + "learning_rate": 0.0002441438469152095, + "loss": 0.0327, + "step": 2550 + }, + { + "epoch": 10.3, + "grad_norm": 0.22233247756958008, + "learning_rate": 0.00024249422632794456, + "loss": 0.0211, + "step": 2575 + }, + { + "epoch": 10.4, + "grad_norm": 13.307366371154785, + "learning_rate": 0.00024084460574067963, + "loss": 0.0472, + "step": 2600 + }, + { + "epoch": 10.5, + "grad_norm": 0.17787696421146393, + "learning_rate": 0.0002391949851534147, + "loss": 0.0346, + "step": 2625 + }, + { + "epoch": 10.6, + "grad_norm": 1.2845549583435059, + "learning_rate": 0.0002375453645661498, + "loss": 0.0587, + "step": 2650 + }, + { + "epoch": 10.7, + "grad_norm": 0.0603482760488987, + "learning_rate": 0.00023589574397888486, + "loss": 0.0394, + "step": 2675 + }, + { + "epoch": 10.8, + "grad_norm": 0.031819943338632584, + "learning_rate": 0.00023424612339161997, + "loss": 0.0434, + "step": 2700 + }, + { + "epoch": 10.9, + "grad_norm": 1.5738537311553955, + "learning_rate": 0.000232596502804355, + "loss": 0.0393, + "step": 2725 + }, + { + "epoch": 11.0, + "grad_norm": 0.6862583756446838, + "learning_rate": 0.00023094688221709007, + "loss": 0.0183, + "step": 2750 + }, + { + "epoch": 11.1, + "grad_norm": 0.0053489054553210735, + "learning_rate": 0.00022929726162982514, + "loss": 0.0277, + "step": 2775 + }, + { + "epoch": 11.2, + "grad_norm": 0.04365606606006622, + "learning_rate": 0.00022764764104256022, + "loss": 0.0337, + "step": 2800 + }, + { + "epoch": 11.3, + "grad_norm": 1.3348641395568848, + "learning_rate": 0.0002259980204552953, + "loss": 0.0219, + "step": 2825 + }, + { + "epoch": 11.4, + "grad_norm": 3.3695547580718994, + "learning_rate": 0.00022434839986803037, + "loss": 0.0227, + "step": 2850 + }, + { + "epoch": 11.5, + "grad_norm": 2.507455587387085, + "learning_rate": 0.00022269877928076542, + "loss": 0.0443, + "step": 2875 + }, + { + "epoch": 11.6, + "grad_norm": 0.013451912440359592, + "learning_rate": 0.0002210491586935005, + "loss": 0.0486, + "step": 2900 + }, + { + "epoch": 11.7, + "grad_norm": 0.031177405267953873, + "learning_rate": 0.00021939953810623557, + "loss": 0.0511, + "step": 2925 + }, + { + "epoch": 11.8, + "grad_norm": 0.02148960903286934, + "learning_rate": 0.00021774991751897065, + "loss": 0.0269, + "step": 2950 + }, + { + "epoch": 11.9, + "grad_norm": 0.01219563465565443, + "learning_rate": 0.00021610029693170573, + "loss": 0.042, + "step": 2975 + }, + { + "epoch": 12.0, + "grad_norm": 0.2062515765428543, + "learning_rate": 0.0002144506763444408, + "loss": 0.031, + "step": 3000 + } + ], + "logging_steps": 25, + "max_steps": 6250, + "num_input_tokens_seen": 0, + "num_train_epochs": 25, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 152458444800000.0, + "train_batch_size": 20, + "trial_name": null, + "trial_params": null +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-3000/training_args.bin b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-3000/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..ef3fc22e099df49861bca49d15da2b959cfcb6df --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-3000/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ec1d8ed36c45af65d67a23de6c5216a486932bfb09b720e1a4417b82eaaa215d +size 5304 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-3000/vocab.txt b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-3000/vocab.txt new file mode 100644 index 0000000000000000000000000000000000000000..cc9ad53732e21ea9180f18e1fb67a38cc5702d4c --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-3000/vocab.txt @@ -0,0 +1,4101 @@ +[PAD] +[UNK] +[CLS] +[SEP] +[MASK] +AAAAAA +AAAAAT +AAAAAC +AAAAAG +AAAATA +AAAATT +AAAATC +AAAATG +AAAACA +AAAACT +AAAACC +AAAACG +AAAAGA +AAAAGT +AAAAGC +AAAAGG +AAATAA +AAATAT +AAATAC +AAATAG +AAATTA +AAATTT +AAATTC +AAATTG +AAATCA +AAATCT +AAATCC +AAATCG +AAATGA +AAATGT +AAATGC +AAATGG +AAACAA +AAACAT +AAACAC +AAACAG +AAACTA +AAACTT +AAACTC +AAACTG +AAACCA +AAACCT +AAACCC +AAACCG +AAACGA +AAACGT +AAACGC +AAACGG +AAAGAA +AAAGAT +AAAGAC +AAAGAG +AAAGTA +AAAGTT +AAAGTC +AAAGTG +AAAGCA +AAAGCT +AAAGCC +AAAGCG +AAAGGA +AAAGGT +AAAGGC +AAAGGG +AATAAA +AATAAT +AATAAC +AATAAG +AATATA +AATATT +AATATC +AATATG +AATACA +AATACT +AATACC +AATACG +AATAGA +AATAGT +AATAGC +AATAGG +AATTAA +AATTAT +AATTAC +AATTAG +AATTTA +AATTTT +AATTTC +AATTTG +AATTCA +AATTCT +AATTCC +AATTCG +AATTGA +AATTGT +AATTGC +AATTGG +AATCAA +AATCAT +AATCAC +AATCAG +AATCTA +AATCTT +AATCTC +AATCTG +AATCCA +AATCCT +AATCCC +AATCCG +AATCGA +AATCGT +AATCGC +AATCGG +AATGAA +AATGAT +AATGAC +AATGAG +AATGTA +AATGTT +AATGTC +AATGTG +AATGCA +AATGCT +AATGCC +AATGCG +AATGGA +AATGGT +AATGGC +AATGGG +AACAAA +AACAAT +AACAAC +AACAAG +AACATA +AACATT +AACATC +AACATG +AACACA +AACACT +AACACC +AACACG +AACAGA +AACAGT +AACAGC +AACAGG +AACTAA +AACTAT +AACTAC +AACTAG +AACTTA +AACTTT +AACTTC +AACTTG +AACTCA +AACTCT +AACTCC +AACTCG +AACTGA +AACTGT +AACTGC +AACTGG +AACCAA +AACCAT +AACCAC +AACCAG +AACCTA +AACCTT +AACCTC +AACCTG +AACCCA +AACCCT +AACCCC +AACCCG +AACCGA +AACCGT +AACCGC +AACCGG +AACGAA +AACGAT +AACGAC +AACGAG +AACGTA +AACGTT +AACGTC +AACGTG +AACGCA +AACGCT +AACGCC +AACGCG +AACGGA +AACGGT +AACGGC +AACGGG +AAGAAA +AAGAAT +AAGAAC +AAGAAG +AAGATA +AAGATT +AAGATC +AAGATG +AAGACA +AAGACT +AAGACC +AAGACG +AAGAGA +AAGAGT +AAGAGC +AAGAGG +AAGTAA +AAGTAT +AAGTAC +AAGTAG +AAGTTA +AAGTTT +AAGTTC +AAGTTG +AAGTCA +AAGTCT +AAGTCC +AAGTCG +AAGTGA +AAGTGT +AAGTGC +AAGTGG +AAGCAA +AAGCAT +AAGCAC +AAGCAG +AAGCTA +AAGCTT +AAGCTC +AAGCTG +AAGCCA +AAGCCT +AAGCCC +AAGCCG +AAGCGA +AAGCGT +AAGCGC +AAGCGG +AAGGAA +AAGGAT +AAGGAC +AAGGAG +AAGGTA +AAGGTT +AAGGTC +AAGGTG +AAGGCA +AAGGCT +AAGGCC +AAGGCG +AAGGGA +AAGGGT +AAGGGC +AAGGGG +ATAAAA +ATAAAT +ATAAAC +ATAAAG +ATAATA +ATAATT +ATAATC +ATAATG +ATAACA +ATAACT +ATAACC +ATAACG +ATAAGA +ATAAGT +ATAAGC +ATAAGG +ATATAA +ATATAT +ATATAC +ATATAG +ATATTA +ATATTT +ATATTC +ATATTG +ATATCA +ATATCT +ATATCC +ATATCG +ATATGA +ATATGT +ATATGC +ATATGG +ATACAA +ATACAT +ATACAC +ATACAG +ATACTA +ATACTT +ATACTC +ATACTG +ATACCA +ATACCT +ATACCC +ATACCG +ATACGA +ATACGT +ATACGC +ATACGG +ATAGAA +ATAGAT +ATAGAC +ATAGAG +ATAGTA +ATAGTT +ATAGTC +ATAGTG +ATAGCA +ATAGCT +ATAGCC +ATAGCG +ATAGGA +ATAGGT +ATAGGC +ATAGGG +ATTAAA +ATTAAT +ATTAAC +ATTAAG +ATTATA +ATTATT +ATTATC +ATTATG +ATTACA +ATTACT +ATTACC +ATTACG +ATTAGA +ATTAGT +ATTAGC +ATTAGG +ATTTAA +ATTTAT +ATTTAC +ATTTAG +ATTTTA +ATTTTT +ATTTTC +ATTTTG +ATTTCA +ATTTCT +ATTTCC +ATTTCG +ATTTGA +ATTTGT +ATTTGC +ATTTGG +ATTCAA +ATTCAT +ATTCAC +ATTCAG +ATTCTA +ATTCTT +ATTCTC +ATTCTG +ATTCCA +ATTCCT +ATTCCC +ATTCCG +ATTCGA +ATTCGT +ATTCGC +ATTCGG +ATTGAA +ATTGAT +ATTGAC +ATTGAG +ATTGTA +ATTGTT +ATTGTC +ATTGTG +ATTGCA +ATTGCT +ATTGCC +ATTGCG +ATTGGA +ATTGGT +ATTGGC +ATTGGG +ATCAAA +ATCAAT +ATCAAC +ATCAAG +ATCATA +ATCATT +ATCATC +ATCATG +ATCACA +ATCACT +ATCACC +ATCACG +ATCAGA +ATCAGT +ATCAGC +ATCAGG +ATCTAA +ATCTAT +ATCTAC +ATCTAG +ATCTTA +ATCTTT +ATCTTC +ATCTTG +ATCTCA +ATCTCT +ATCTCC +ATCTCG +ATCTGA +ATCTGT +ATCTGC +ATCTGG +ATCCAA +ATCCAT +ATCCAC +ATCCAG +ATCCTA +ATCCTT +ATCCTC +ATCCTG +ATCCCA +ATCCCT +ATCCCC +ATCCCG +ATCCGA +ATCCGT +ATCCGC +ATCCGG +ATCGAA +ATCGAT +ATCGAC +ATCGAG +ATCGTA +ATCGTT +ATCGTC +ATCGTG +ATCGCA +ATCGCT +ATCGCC +ATCGCG +ATCGGA +ATCGGT +ATCGGC +ATCGGG +ATGAAA +ATGAAT +ATGAAC +ATGAAG +ATGATA +ATGATT +ATGATC +ATGATG +ATGACA +ATGACT +ATGACC +ATGACG +ATGAGA +ATGAGT +ATGAGC +ATGAGG +ATGTAA +ATGTAT +ATGTAC +ATGTAG +ATGTTA +ATGTTT +ATGTTC +ATGTTG +ATGTCA +ATGTCT +ATGTCC +ATGTCG +ATGTGA +ATGTGT +ATGTGC +ATGTGG +ATGCAA +ATGCAT +ATGCAC +ATGCAG +ATGCTA +ATGCTT +ATGCTC +ATGCTG +ATGCCA +ATGCCT +ATGCCC +ATGCCG +ATGCGA +ATGCGT +ATGCGC +ATGCGG +ATGGAA +ATGGAT +ATGGAC +ATGGAG +ATGGTA +ATGGTT +ATGGTC +ATGGTG +ATGGCA +ATGGCT +ATGGCC +ATGGCG +ATGGGA +ATGGGT +ATGGGC +ATGGGG +ACAAAA +ACAAAT +ACAAAC +ACAAAG +ACAATA +ACAATT +ACAATC +ACAATG +ACAACA +ACAACT +ACAACC +ACAACG +ACAAGA +ACAAGT +ACAAGC +ACAAGG +ACATAA +ACATAT +ACATAC +ACATAG +ACATTA +ACATTT +ACATTC +ACATTG +ACATCA +ACATCT +ACATCC +ACATCG +ACATGA +ACATGT +ACATGC +ACATGG +ACACAA +ACACAT +ACACAC +ACACAG +ACACTA +ACACTT +ACACTC +ACACTG +ACACCA +ACACCT +ACACCC +ACACCG +ACACGA +ACACGT +ACACGC +ACACGG +ACAGAA +ACAGAT +ACAGAC +ACAGAG +ACAGTA +ACAGTT +ACAGTC +ACAGTG +ACAGCA +ACAGCT +ACAGCC +ACAGCG +ACAGGA +ACAGGT +ACAGGC +ACAGGG +ACTAAA +ACTAAT +ACTAAC +ACTAAG +ACTATA +ACTATT +ACTATC +ACTATG +ACTACA +ACTACT +ACTACC +ACTACG +ACTAGA +ACTAGT +ACTAGC +ACTAGG +ACTTAA +ACTTAT +ACTTAC +ACTTAG +ACTTTA +ACTTTT +ACTTTC +ACTTTG +ACTTCA +ACTTCT +ACTTCC +ACTTCG +ACTTGA +ACTTGT +ACTTGC +ACTTGG +ACTCAA +ACTCAT +ACTCAC +ACTCAG +ACTCTA +ACTCTT +ACTCTC +ACTCTG +ACTCCA +ACTCCT +ACTCCC +ACTCCG +ACTCGA +ACTCGT +ACTCGC +ACTCGG +ACTGAA +ACTGAT +ACTGAC +ACTGAG +ACTGTA +ACTGTT +ACTGTC +ACTGTG +ACTGCA +ACTGCT +ACTGCC +ACTGCG +ACTGGA +ACTGGT +ACTGGC +ACTGGG +ACCAAA +ACCAAT +ACCAAC +ACCAAG +ACCATA +ACCATT +ACCATC +ACCATG +ACCACA +ACCACT +ACCACC +ACCACG +ACCAGA +ACCAGT +ACCAGC +ACCAGG +ACCTAA +ACCTAT +ACCTAC +ACCTAG +ACCTTA +ACCTTT +ACCTTC +ACCTTG +ACCTCA +ACCTCT +ACCTCC +ACCTCG +ACCTGA +ACCTGT +ACCTGC +ACCTGG +ACCCAA +ACCCAT +ACCCAC +ACCCAG +ACCCTA +ACCCTT +ACCCTC +ACCCTG +ACCCCA +ACCCCT +ACCCCC +ACCCCG +ACCCGA +ACCCGT +ACCCGC +ACCCGG +ACCGAA +ACCGAT +ACCGAC +ACCGAG +ACCGTA +ACCGTT +ACCGTC +ACCGTG +ACCGCA +ACCGCT +ACCGCC +ACCGCG +ACCGGA +ACCGGT +ACCGGC +ACCGGG +ACGAAA +ACGAAT +ACGAAC +ACGAAG +ACGATA +ACGATT +ACGATC +ACGATG +ACGACA +ACGACT +ACGACC +ACGACG +ACGAGA +ACGAGT +ACGAGC +ACGAGG +ACGTAA +ACGTAT +ACGTAC +ACGTAG +ACGTTA +ACGTTT +ACGTTC +ACGTTG +ACGTCA +ACGTCT +ACGTCC +ACGTCG +ACGTGA +ACGTGT +ACGTGC +ACGTGG +ACGCAA +ACGCAT +ACGCAC +ACGCAG +ACGCTA +ACGCTT +ACGCTC +ACGCTG +ACGCCA +ACGCCT +ACGCCC +ACGCCG +ACGCGA +ACGCGT +ACGCGC +ACGCGG +ACGGAA +ACGGAT +ACGGAC +ACGGAG +ACGGTA +ACGGTT +ACGGTC +ACGGTG +ACGGCA +ACGGCT +ACGGCC +ACGGCG +ACGGGA +ACGGGT +ACGGGC +ACGGGG +AGAAAA +AGAAAT +AGAAAC +AGAAAG +AGAATA +AGAATT +AGAATC +AGAATG +AGAACA +AGAACT +AGAACC +AGAACG +AGAAGA +AGAAGT +AGAAGC +AGAAGG +AGATAA +AGATAT +AGATAC +AGATAG +AGATTA +AGATTT +AGATTC +AGATTG +AGATCA +AGATCT +AGATCC +AGATCG +AGATGA +AGATGT +AGATGC +AGATGG +AGACAA +AGACAT +AGACAC +AGACAG +AGACTA +AGACTT +AGACTC +AGACTG +AGACCA +AGACCT +AGACCC +AGACCG +AGACGA +AGACGT +AGACGC +AGACGG +AGAGAA +AGAGAT +AGAGAC +AGAGAG +AGAGTA +AGAGTT +AGAGTC +AGAGTG +AGAGCA +AGAGCT +AGAGCC +AGAGCG +AGAGGA +AGAGGT +AGAGGC +AGAGGG +AGTAAA +AGTAAT +AGTAAC +AGTAAG +AGTATA +AGTATT +AGTATC +AGTATG +AGTACA +AGTACT +AGTACC +AGTACG +AGTAGA +AGTAGT +AGTAGC +AGTAGG +AGTTAA +AGTTAT +AGTTAC +AGTTAG +AGTTTA +AGTTTT +AGTTTC +AGTTTG +AGTTCA +AGTTCT +AGTTCC +AGTTCG +AGTTGA +AGTTGT +AGTTGC +AGTTGG +AGTCAA +AGTCAT +AGTCAC +AGTCAG +AGTCTA +AGTCTT +AGTCTC +AGTCTG +AGTCCA +AGTCCT +AGTCCC +AGTCCG +AGTCGA +AGTCGT +AGTCGC +AGTCGG +AGTGAA +AGTGAT +AGTGAC +AGTGAG +AGTGTA +AGTGTT +AGTGTC +AGTGTG +AGTGCA +AGTGCT +AGTGCC +AGTGCG +AGTGGA +AGTGGT +AGTGGC +AGTGGG +AGCAAA +AGCAAT +AGCAAC +AGCAAG +AGCATA +AGCATT +AGCATC +AGCATG +AGCACA +AGCACT +AGCACC +AGCACG +AGCAGA +AGCAGT +AGCAGC +AGCAGG +AGCTAA +AGCTAT +AGCTAC +AGCTAG +AGCTTA +AGCTTT +AGCTTC +AGCTTG +AGCTCA +AGCTCT +AGCTCC +AGCTCG +AGCTGA +AGCTGT +AGCTGC +AGCTGG +AGCCAA +AGCCAT +AGCCAC +AGCCAG +AGCCTA +AGCCTT +AGCCTC +AGCCTG +AGCCCA +AGCCCT +AGCCCC +AGCCCG +AGCCGA +AGCCGT +AGCCGC +AGCCGG +AGCGAA +AGCGAT +AGCGAC +AGCGAG +AGCGTA +AGCGTT +AGCGTC +AGCGTG +AGCGCA +AGCGCT +AGCGCC +AGCGCG +AGCGGA +AGCGGT +AGCGGC +AGCGGG +AGGAAA +AGGAAT +AGGAAC +AGGAAG +AGGATA +AGGATT +AGGATC +AGGATG +AGGACA +AGGACT +AGGACC +AGGACG +AGGAGA +AGGAGT +AGGAGC +AGGAGG +AGGTAA +AGGTAT +AGGTAC +AGGTAG +AGGTTA +AGGTTT +AGGTTC +AGGTTG +AGGTCA +AGGTCT +AGGTCC +AGGTCG +AGGTGA +AGGTGT +AGGTGC +AGGTGG +AGGCAA +AGGCAT +AGGCAC +AGGCAG +AGGCTA +AGGCTT +AGGCTC +AGGCTG +AGGCCA +AGGCCT +AGGCCC +AGGCCG +AGGCGA +AGGCGT +AGGCGC +AGGCGG +AGGGAA +AGGGAT +AGGGAC +AGGGAG +AGGGTA +AGGGTT +AGGGTC +AGGGTG +AGGGCA +AGGGCT +AGGGCC +AGGGCG +AGGGGA +AGGGGT +AGGGGC +AGGGGG +TAAAAA +TAAAAT +TAAAAC +TAAAAG +TAAATA +TAAATT +TAAATC +TAAATG +TAAACA +TAAACT +TAAACC +TAAACG +TAAAGA +TAAAGT +TAAAGC +TAAAGG +TAATAA +TAATAT +TAATAC +TAATAG +TAATTA +TAATTT +TAATTC +TAATTG +TAATCA +TAATCT +TAATCC +TAATCG +TAATGA +TAATGT +TAATGC +TAATGG +TAACAA +TAACAT +TAACAC +TAACAG +TAACTA +TAACTT +TAACTC +TAACTG +TAACCA +TAACCT +TAACCC +TAACCG +TAACGA +TAACGT +TAACGC +TAACGG +TAAGAA +TAAGAT +TAAGAC +TAAGAG +TAAGTA +TAAGTT +TAAGTC +TAAGTG +TAAGCA +TAAGCT +TAAGCC +TAAGCG +TAAGGA +TAAGGT +TAAGGC +TAAGGG +TATAAA +TATAAT +TATAAC +TATAAG +TATATA +TATATT +TATATC +TATATG +TATACA +TATACT +TATACC +TATACG +TATAGA +TATAGT +TATAGC +TATAGG +TATTAA +TATTAT +TATTAC +TATTAG +TATTTA +TATTTT +TATTTC +TATTTG +TATTCA +TATTCT +TATTCC +TATTCG +TATTGA +TATTGT +TATTGC +TATTGG +TATCAA +TATCAT +TATCAC +TATCAG +TATCTA +TATCTT +TATCTC +TATCTG +TATCCA +TATCCT +TATCCC +TATCCG +TATCGA +TATCGT +TATCGC +TATCGG +TATGAA +TATGAT +TATGAC +TATGAG +TATGTA +TATGTT +TATGTC +TATGTG +TATGCA +TATGCT +TATGCC +TATGCG +TATGGA +TATGGT +TATGGC +TATGGG +TACAAA +TACAAT +TACAAC +TACAAG +TACATA +TACATT +TACATC +TACATG +TACACA +TACACT +TACACC +TACACG +TACAGA +TACAGT +TACAGC +TACAGG +TACTAA +TACTAT +TACTAC +TACTAG +TACTTA +TACTTT +TACTTC +TACTTG +TACTCA +TACTCT +TACTCC +TACTCG +TACTGA +TACTGT +TACTGC +TACTGG +TACCAA +TACCAT +TACCAC +TACCAG +TACCTA +TACCTT +TACCTC +TACCTG +TACCCA +TACCCT +TACCCC +TACCCG +TACCGA +TACCGT +TACCGC +TACCGG +TACGAA +TACGAT +TACGAC +TACGAG +TACGTA +TACGTT +TACGTC +TACGTG +TACGCA +TACGCT +TACGCC +TACGCG +TACGGA +TACGGT +TACGGC +TACGGG +TAGAAA +TAGAAT +TAGAAC +TAGAAG +TAGATA +TAGATT +TAGATC +TAGATG +TAGACA +TAGACT +TAGACC +TAGACG +TAGAGA +TAGAGT +TAGAGC +TAGAGG +TAGTAA +TAGTAT +TAGTAC +TAGTAG +TAGTTA +TAGTTT +TAGTTC +TAGTTG +TAGTCA +TAGTCT +TAGTCC +TAGTCG +TAGTGA +TAGTGT +TAGTGC +TAGTGG +TAGCAA +TAGCAT +TAGCAC +TAGCAG +TAGCTA +TAGCTT +TAGCTC +TAGCTG +TAGCCA +TAGCCT +TAGCCC +TAGCCG +TAGCGA +TAGCGT +TAGCGC +TAGCGG +TAGGAA +TAGGAT +TAGGAC +TAGGAG +TAGGTA +TAGGTT +TAGGTC +TAGGTG +TAGGCA +TAGGCT +TAGGCC +TAGGCG +TAGGGA +TAGGGT +TAGGGC +TAGGGG +TTAAAA +TTAAAT +TTAAAC +TTAAAG +TTAATA +TTAATT +TTAATC +TTAATG +TTAACA +TTAACT +TTAACC +TTAACG +TTAAGA +TTAAGT +TTAAGC +TTAAGG +TTATAA +TTATAT +TTATAC +TTATAG +TTATTA +TTATTT +TTATTC +TTATTG +TTATCA +TTATCT +TTATCC +TTATCG +TTATGA +TTATGT +TTATGC +TTATGG +TTACAA +TTACAT +TTACAC +TTACAG +TTACTA +TTACTT +TTACTC +TTACTG +TTACCA +TTACCT +TTACCC +TTACCG +TTACGA +TTACGT +TTACGC +TTACGG +TTAGAA +TTAGAT +TTAGAC +TTAGAG +TTAGTA +TTAGTT +TTAGTC +TTAGTG +TTAGCA +TTAGCT +TTAGCC +TTAGCG +TTAGGA +TTAGGT +TTAGGC +TTAGGG +TTTAAA +TTTAAT +TTTAAC +TTTAAG +TTTATA +TTTATT +TTTATC +TTTATG +TTTACA +TTTACT +TTTACC +TTTACG +TTTAGA +TTTAGT +TTTAGC +TTTAGG +TTTTAA +TTTTAT +TTTTAC +TTTTAG +TTTTTA +TTTTTT +TTTTTC +TTTTTG +TTTTCA +TTTTCT +TTTTCC +TTTTCG +TTTTGA +TTTTGT +TTTTGC +TTTTGG +TTTCAA +TTTCAT +TTTCAC +TTTCAG +TTTCTA +TTTCTT +TTTCTC +TTTCTG +TTTCCA +TTTCCT +TTTCCC +TTTCCG +TTTCGA +TTTCGT +TTTCGC +TTTCGG +TTTGAA +TTTGAT +TTTGAC +TTTGAG +TTTGTA +TTTGTT +TTTGTC +TTTGTG +TTTGCA +TTTGCT +TTTGCC +TTTGCG +TTTGGA +TTTGGT +TTTGGC +TTTGGG +TTCAAA +TTCAAT +TTCAAC +TTCAAG +TTCATA +TTCATT +TTCATC +TTCATG +TTCACA +TTCACT +TTCACC +TTCACG +TTCAGA +TTCAGT +TTCAGC +TTCAGG +TTCTAA +TTCTAT +TTCTAC +TTCTAG +TTCTTA +TTCTTT +TTCTTC +TTCTTG +TTCTCA +TTCTCT +TTCTCC +TTCTCG +TTCTGA +TTCTGT +TTCTGC +TTCTGG +TTCCAA +TTCCAT +TTCCAC +TTCCAG +TTCCTA +TTCCTT +TTCCTC +TTCCTG +TTCCCA +TTCCCT +TTCCCC +TTCCCG +TTCCGA +TTCCGT +TTCCGC +TTCCGG +TTCGAA +TTCGAT +TTCGAC +TTCGAG +TTCGTA +TTCGTT +TTCGTC +TTCGTG +TTCGCA +TTCGCT +TTCGCC +TTCGCG +TTCGGA +TTCGGT +TTCGGC +TTCGGG +TTGAAA +TTGAAT +TTGAAC +TTGAAG +TTGATA +TTGATT +TTGATC +TTGATG +TTGACA +TTGACT +TTGACC +TTGACG +TTGAGA +TTGAGT +TTGAGC +TTGAGG +TTGTAA +TTGTAT +TTGTAC +TTGTAG +TTGTTA +TTGTTT +TTGTTC +TTGTTG +TTGTCA +TTGTCT +TTGTCC +TTGTCG +TTGTGA +TTGTGT +TTGTGC +TTGTGG +TTGCAA +TTGCAT +TTGCAC +TTGCAG +TTGCTA +TTGCTT +TTGCTC +TTGCTG +TTGCCA +TTGCCT +TTGCCC +TTGCCG +TTGCGA +TTGCGT +TTGCGC +TTGCGG +TTGGAA +TTGGAT +TTGGAC +TTGGAG +TTGGTA +TTGGTT +TTGGTC +TTGGTG +TTGGCA +TTGGCT +TTGGCC +TTGGCG +TTGGGA +TTGGGT +TTGGGC +TTGGGG +TCAAAA +TCAAAT +TCAAAC +TCAAAG +TCAATA +TCAATT +TCAATC +TCAATG +TCAACA +TCAACT +TCAACC +TCAACG +TCAAGA +TCAAGT +TCAAGC +TCAAGG +TCATAA +TCATAT +TCATAC +TCATAG +TCATTA +TCATTT +TCATTC +TCATTG +TCATCA +TCATCT +TCATCC +TCATCG +TCATGA +TCATGT +TCATGC +TCATGG +TCACAA +TCACAT +TCACAC +TCACAG +TCACTA +TCACTT +TCACTC +TCACTG +TCACCA +TCACCT +TCACCC +TCACCG +TCACGA +TCACGT +TCACGC +TCACGG +TCAGAA +TCAGAT +TCAGAC +TCAGAG +TCAGTA +TCAGTT +TCAGTC +TCAGTG +TCAGCA +TCAGCT +TCAGCC +TCAGCG +TCAGGA +TCAGGT +TCAGGC +TCAGGG +TCTAAA +TCTAAT +TCTAAC +TCTAAG +TCTATA +TCTATT +TCTATC +TCTATG +TCTACA +TCTACT +TCTACC +TCTACG +TCTAGA +TCTAGT +TCTAGC +TCTAGG +TCTTAA +TCTTAT +TCTTAC +TCTTAG +TCTTTA +TCTTTT +TCTTTC +TCTTTG +TCTTCA +TCTTCT +TCTTCC +TCTTCG +TCTTGA +TCTTGT +TCTTGC +TCTTGG +TCTCAA +TCTCAT +TCTCAC +TCTCAG +TCTCTA +TCTCTT +TCTCTC +TCTCTG +TCTCCA +TCTCCT +TCTCCC +TCTCCG +TCTCGA +TCTCGT +TCTCGC +TCTCGG +TCTGAA +TCTGAT +TCTGAC +TCTGAG +TCTGTA +TCTGTT +TCTGTC +TCTGTG +TCTGCA +TCTGCT +TCTGCC +TCTGCG +TCTGGA +TCTGGT +TCTGGC +TCTGGG +TCCAAA +TCCAAT +TCCAAC +TCCAAG +TCCATA +TCCATT +TCCATC +TCCATG +TCCACA +TCCACT +TCCACC +TCCACG +TCCAGA +TCCAGT +TCCAGC +TCCAGG +TCCTAA +TCCTAT +TCCTAC +TCCTAG +TCCTTA +TCCTTT +TCCTTC +TCCTTG +TCCTCA +TCCTCT +TCCTCC +TCCTCG +TCCTGA +TCCTGT +TCCTGC +TCCTGG +TCCCAA +TCCCAT +TCCCAC +TCCCAG +TCCCTA +TCCCTT +TCCCTC +TCCCTG +TCCCCA +TCCCCT +TCCCCC +TCCCCG +TCCCGA +TCCCGT +TCCCGC +TCCCGG +TCCGAA +TCCGAT +TCCGAC +TCCGAG +TCCGTA +TCCGTT +TCCGTC +TCCGTG +TCCGCA +TCCGCT +TCCGCC +TCCGCG +TCCGGA +TCCGGT +TCCGGC +TCCGGG +TCGAAA +TCGAAT +TCGAAC +TCGAAG +TCGATA +TCGATT +TCGATC +TCGATG +TCGACA +TCGACT +TCGACC +TCGACG +TCGAGA +TCGAGT +TCGAGC +TCGAGG +TCGTAA +TCGTAT +TCGTAC +TCGTAG +TCGTTA +TCGTTT +TCGTTC +TCGTTG +TCGTCA +TCGTCT +TCGTCC +TCGTCG +TCGTGA +TCGTGT +TCGTGC +TCGTGG +TCGCAA +TCGCAT +TCGCAC +TCGCAG +TCGCTA +TCGCTT +TCGCTC +TCGCTG +TCGCCA +TCGCCT +TCGCCC +TCGCCG +TCGCGA +TCGCGT +TCGCGC +TCGCGG +TCGGAA +TCGGAT +TCGGAC +TCGGAG +TCGGTA +TCGGTT +TCGGTC +TCGGTG +TCGGCA +TCGGCT +TCGGCC +TCGGCG +TCGGGA +TCGGGT +TCGGGC +TCGGGG +TGAAAA +TGAAAT +TGAAAC +TGAAAG +TGAATA +TGAATT +TGAATC +TGAATG +TGAACA +TGAACT +TGAACC +TGAACG +TGAAGA +TGAAGT +TGAAGC +TGAAGG +TGATAA +TGATAT +TGATAC +TGATAG +TGATTA +TGATTT +TGATTC +TGATTG +TGATCA +TGATCT +TGATCC +TGATCG +TGATGA +TGATGT +TGATGC +TGATGG +TGACAA +TGACAT +TGACAC +TGACAG +TGACTA +TGACTT +TGACTC +TGACTG +TGACCA +TGACCT +TGACCC +TGACCG +TGACGA +TGACGT +TGACGC +TGACGG +TGAGAA +TGAGAT +TGAGAC +TGAGAG +TGAGTA +TGAGTT +TGAGTC +TGAGTG +TGAGCA +TGAGCT +TGAGCC +TGAGCG +TGAGGA +TGAGGT +TGAGGC +TGAGGG +TGTAAA +TGTAAT +TGTAAC +TGTAAG +TGTATA +TGTATT +TGTATC +TGTATG +TGTACA +TGTACT +TGTACC +TGTACG +TGTAGA +TGTAGT +TGTAGC +TGTAGG +TGTTAA +TGTTAT +TGTTAC +TGTTAG +TGTTTA +TGTTTT +TGTTTC +TGTTTG +TGTTCA +TGTTCT +TGTTCC +TGTTCG +TGTTGA +TGTTGT +TGTTGC +TGTTGG +TGTCAA +TGTCAT +TGTCAC +TGTCAG +TGTCTA +TGTCTT +TGTCTC +TGTCTG +TGTCCA +TGTCCT +TGTCCC +TGTCCG +TGTCGA +TGTCGT +TGTCGC +TGTCGG +TGTGAA +TGTGAT +TGTGAC +TGTGAG +TGTGTA +TGTGTT +TGTGTC +TGTGTG +TGTGCA +TGTGCT +TGTGCC +TGTGCG +TGTGGA +TGTGGT +TGTGGC +TGTGGG +TGCAAA +TGCAAT +TGCAAC +TGCAAG +TGCATA +TGCATT +TGCATC +TGCATG +TGCACA +TGCACT +TGCACC +TGCACG +TGCAGA +TGCAGT +TGCAGC +TGCAGG +TGCTAA +TGCTAT +TGCTAC +TGCTAG +TGCTTA +TGCTTT +TGCTTC +TGCTTG +TGCTCA +TGCTCT +TGCTCC +TGCTCG +TGCTGA +TGCTGT +TGCTGC +TGCTGG +TGCCAA +TGCCAT +TGCCAC +TGCCAG +TGCCTA +TGCCTT +TGCCTC +TGCCTG +TGCCCA +TGCCCT +TGCCCC +TGCCCG +TGCCGA +TGCCGT +TGCCGC +TGCCGG +TGCGAA +TGCGAT +TGCGAC +TGCGAG +TGCGTA +TGCGTT +TGCGTC +TGCGTG +TGCGCA +TGCGCT +TGCGCC +TGCGCG +TGCGGA +TGCGGT +TGCGGC +TGCGGG +TGGAAA +TGGAAT +TGGAAC +TGGAAG +TGGATA +TGGATT +TGGATC +TGGATG +TGGACA +TGGACT +TGGACC +TGGACG +TGGAGA +TGGAGT +TGGAGC +TGGAGG +TGGTAA +TGGTAT +TGGTAC +TGGTAG +TGGTTA +TGGTTT +TGGTTC +TGGTTG +TGGTCA +TGGTCT +TGGTCC +TGGTCG +TGGTGA +TGGTGT +TGGTGC +TGGTGG +TGGCAA +TGGCAT +TGGCAC +TGGCAG +TGGCTA +TGGCTT +TGGCTC +TGGCTG +TGGCCA +TGGCCT +TGGCCC +TGGCCG +TGGCGA +TGGCGT +TGGCGC +TGGCGG +TGGGAA +TGGGAT +TGGGAC +TGGGAG +TGGGTA +TGGGTT +TGGGTC +TGGGTG +TGGGCA +TGGGCT +TGGGCC +TGGGCG +TGGGGA +TGGGGT +TGGGGC +TGGGGG +CAAAAA +CAAAAT +CAAAAC +CAAAAG +CAAATA +CAAATT +CAAATC +CAAATG +CAAACA +CAAACT +CAAACC +CAAACG +CAAAGA +CAAAGT +CAAAGC +CAAAGG +CAATAA +CAATAT +CAATAC +CAATAG +CAATTA +CAATTT +CAATTC +CAATTG +CAATCA +CAATCT +CAATCC +CAATCG +CAATGA +CAATGT +CAATGC +CAATGG +CAACAA +CAACAT +CAACAC +CAACAG +CAACTA +CAACTT +CAACTC +CAACTG +CAACCA +CAACCT +CAACCC +CAACCG +CAACGA +CAACGT +CAACGC +CAACGG +CAAGAA +CAAGAT +CAAGAC +CAAGAG +CAAGTA +CAAGTT +CAAGTC +CAAGTG +CAAGCA +CAAGCT +CAAGCC +CAAGCG +CAAGGA +CAAGGT +CAAGGC +CAAGGG +CATAAA +CATAAT +CATAAC +CATAAG +CATATA +CATATT +CATATC +CATATG +CATACA +CATACT +CATACC +CATACG +CATAGA +CATAGT +CATAGC +CATAGG +CATTAA +CATTAT +CATTAC +CATTAG +CATTTA +CATTTT +CATTTC +CATTTG +CATTCA +CATTCT +CATTCC +CATTCG +CATTGA +CATTGT +CATTGC +CATTGG +CATCAA +CATCAT +CATCAC +CATCAG +CATCTA +CATCTT +CATCTC +CATCTG +CATCCA +CATCCT +CATCCC +CATCCG +CATCGA +CATCGT +CATCGC +CATCGG +CATGAA +CATGAT +CATGAC +CATGAG +CATGTA +CATGTT +CATGTC +CATGTG +CATGCA +CATGCT +CATGCC +CATGCG +CATGGA +CATGGT +CATGGC +CATGGG +CACAAA +CACAAT +CACAAC +CACAAG +CACATA +CACATT +CACATC +CACATG +CACACA +CACACT +CACACC +CACACG +CACAGA +CACAGT +CACAGC +CACAGG +CACTAA +CACTAT +CACTAC +CACTAG +CACTTA +CACTTT +CACTTC +CACTTG +CACTCA +CACTCT +CACTCC +CACTCG +CACTGA +CACTGT +CACTGC +CACTGG +CACCAA +CACCAT +CACCAC +CACCAG +CACCTA +CACCTT +CACCTC +CACCTG +CACCCA +CACCCT +CACCCC +CACCCG +CACCGA +CACCGT +CACCGC +CACCGG +CACGAA +CACGAT +CACGAC +CACGAG +CACGTA +CACGTT +CACGTC +CACGTG +CACGCA +CACGCT +CACGCC +CACGCG +CACGGA +CACGGT +CACGGC +CACGGG +CAGAAA +CAGAAT +CAGAAC +CAGAAG +CAGATA +CAGATT +CAGATC +CAGATG +CAGACA +CAGACT +CAGACC +CAGACG +CAGAGA +CAGAGT +CAGAGC +CAGAGG +CAGTAA +CAGTAT +CAGTAC +CAGTAG +CAGTTA +CAGTTT +CAGTTC +CAGTTG +CAGTCA +CAGTCT +CAGTCC +CAGTCG +CAGTGA +CAGTGT +CAGTGC +CAGTGG +CAGCAA +CAGCAT +CAGCAC +CAGCAG +CAGCTA +CAGCTT +CAGCTC +CAGCTG +CAGCCA +CAGCCT +CAGCCC +CAGCCG +CAGCGA +CAGCGT +CAGCGC +CAGCGG +CAGGAA +CAGGAT +CAGGAC +CAGGAG +CAGGTA +CAGGTT +CAGGTC +CAGGTG +CAGGCA +CAGGCT +CAGGCC +CAGGCG +CAGGGA +CAGGGT +CAGGGC +CAGGGG +CTAAAA +CTAAAT +CTAAAC +CTAAAG +CTAATA +CTAATT +CTAATC +CTAATG +CTAACA +CTAACT +CTAACC +CTAACG +CTAAGA +CTAAGT +CTAAGC +CTAAGG +CTATAA +CTATAT +CTATAC +CTATAG +CTATTA +CTATTT +CTATTC +CTATTG +CTATCA +CTATCT +CTATCC +CTATCG +CTATGA +CTATGT +CTATGC +CTATGG +CTACAA +CTACAT +CTACAC +CTACAG +CTACTA +CTACTT +CTACTC +CTACTG +CTACCA +CTACCT +CTACCC +CTACCG +CTACGA +CTACGT +CTACGC +CTACGG +CTAGAA +CTAGAT +CTAGAC +CTAGAG +CTAGTA +CTAGTT +CTAGTC +CTAGTG +CTAGCA +CTAGCT +CTAGCC +CTAGCG +CTAGGA +CTAGGT +CTAGGC +CTAGGG +CTTAAA +CTTAAT +CTTAAC +CTTAAG +CTTATA +CTTATT +CTTATC +CTTATG +CTTACA +CTTACT +CTTACC +CTTACG +CTTAGA +CTTAGT +CTTAGC +CTTAGG +CTTTAA +CTTTAT +CTTTAC +CTTTAG +CTTTTA +CTTTTT +CTTTTC +CTTTTG +CTTTCA +CTTTCT +CTTTCC +CTTTCG +CTTTGA +CTTTGT +CTTTGC +CTTTGG +CTTCAA +CTTCAT +CTTCAC +CTTCAG +CTTCTA +CTTCTT +CTTCTC +CTTCTG +CTTCCA +CTTCCT +CTTCCC +CTTCCG +CTTCGA +CTTCGT +CTTCGC +CTTCGG +CTTGAA +CTTGAT +CTTGAC +CTTGAG +CTTGTA +CTTGTT +CTTGTC +CTTGTG +CTTGCA +CTTGCT +CTTGCC +CTTGCG +CTTGGA +CTTGGT +CTTGGC +CTTGGG +CTCAAA +CTCAAT +CTCAAC +CTCAAG +CTCATA +CTCATT +CTCATC +CTCATG +CTCACA +CTCACT +CTCACC +CTCACG +CTCAGA +CTCAGT +CTCAGC +CTCAGG +CTCTAA +CTCTAT +CTCTAC +CTCTAG +CTCTTA +CTCTTT +CTCTTC +CTCTTG +CTCTCA +CTCTCT +CTCTCC +CTCTCG +CTCTGA +CTCTGT +CTCTGC +CTCTGG +CTCCAA +CTCCAT +CTCCAC +CTCCAG +CTCCTA +CTCCTT +CTCCTC +CTCCTG +CTCCCA +CTCCCT +CTCCCC +CTCCCG +CTCCGA +CTCCGT +CTCCGC +CTCCGG +CTCGAA +CTCGAT +CTCGAC +CTCGAG +CTCGTA +CTCGTT +CTCGTC +CTCGTG +CTCGCA +CTCGCT +CTCGCC +CTCGCG +CTCGGA +CTCGGT +CTCGGC +CTCGGG +CTGAAA +CTGAAT +CTGAAC +CTGAAG +CTGATA +CTGATT +CTGATC +CTGATG +CTGACA +CTGACT +CTGACC +CTGACG +CTGAGA +CTGAGT +CTGAGC +CTGAGG +CTGTAA +CTGTAT +CTGTAC +CTGTAG +CTGTTA +CTGTTT +CTGTTC +CTGTTG +CTGTCA +CTGTCT +CTGTCC +CTGTCG +CTGTGA +CTGTGT +CTGTGC +CTGTGG +CTGCAA +CTGCAT +CTGCAC +CTGCAG +CTGCTA +CTGCTT +CTGCTC +CTGCTG +CTGCCA +CTGCCT +CTGCCC +CTGCCG +CTGCGA +CTGCGT +CTGCGC +CTGCGG +CTGGAA +CTGGAT +CTGGAC +CTGGAG +CTGGTA +CTGGTT +CTGGTC +CTGGTG +CTGGCA +CTGGCT +CTGGCC +CTGGCG +CTGGGA +CTGGGT +CTGGGC +CTGGGG +CCAAAA +CCAAAT +CCAAAC +CCAAAG +CCAATA +CCAATT +CCAATC +CCAATG +CCAACA +CCAACT +CCAACC +CCAACG +CCAAGA +CCAAGT +CCAAGC +CCAAGG +CCATAA +CCATAT +CCATAC +CCATAG +CCATTA +CCATTT +CCATTC +CCATTG +CCATCA +CCATCT +CCATCC +CCATCG +CCATGA +CCATGT +CCATGC +CCATGG +CCACAA +CCACAT +CCACAC +CCACAG +CCACTA +CCACTT +CCACTC +CCACTG +CCACCA +CCACCT +CCACCC +CCACCG +CCACGA +CCACGT +CCACGC +CCACGG +CCAGAA +CCAGAT +CCAGAC +CCAGAG +CCAGTA +CCAGTT +CCAGTC +CCAGTG +CCAGCA +CCAGCT +CCAGCC +CCAGCG +CCAGGA +CCAGGT +CCAGGC +CCAGGG +CCTAAA +CCTAAT +CCTAAC +CCTAAG +CCTATA +CCTATT +CCTATC +CCTATG +CCTACA +CCTACT +CCTACC +CCTACG +CCTAGA +CCTAGT +CCTAGC +CCTAGG +CCTTAA +CCTTAT +CCTTAC +CCTTAG +CCTTTA +CCTTTT +CCTTTC +CCTTTG +CCTTCA +CCTTCT +CCTTCC +CCTTCG +CCTTGA +CCTTGT +CCTTGC +CCTTGG +CCTCAA +CCTCAT +CCTCAC +CCTCAG +CCTCTA +CCTCTT +CCTCTC +CCTCTG +CCTCCA +CCTCCT +CCTCCC +CCTCCG +CCTCGA +CCTCGT +CCTCGC +CCTCGG +CCTGAA +CCTGAT +CCTGAC +CCTGAG +CCTGTA +CCTGTT +CCTGTC +CCTGTG +CCTGCA +CCTGCT +CCTGCC +CCTGCG +CCTGGA +CCTGGT +CCTGGC +CCTGGG +CCCAAA +CCCAAT +CCCAAC +CCCAAG +CCCATA +CCCATT +CCCATC +CCCATG +CCCACA +CCCACT +CCCACC +CCCACG +CCCAGA +CCCAGT +CCCAGC +CCCAGG +CCCTAA +CCCTAT +CCCTAC +CCCTAG +CCCTTA +CCCTTT +CCCTTC +CCCTTG +CCCTCA +CCCTCT +CCCTCC +CCCTCG +CCCTGA +CCCTGT +CCCTGC +CCCTGG +CCCCAA +CCCCAT +CCCCAC +CCCCAG +CCCCTA +CCCCTT +CCCCTC +CCCCTG +CCCCCA +CCCCCT +CCCCCC +CCCCCG +CCCCGA +CCCCGT +CCCCGC +CCCCGG +CCCGAA +CCCGAT +CCCGAC +CCCGAG +CCCGTA +CCCGTT +CCCGTC +CCCGTG +CCCGCA +CCCGCT +CCCGCC +CCCGCG +CCCGGA +CCCGGT +CCCGGC +CCCGGG +CCGAAA +CCGAAT +CCGAAC +CCGAAG +CCGATA +CCGATT +CCGATC +CCGATG +CCGACA +CCGACT +CCGACC +CCGACG +CCGAGA +CCGAGT +CCGAGC +CCGAGG +CCGTAA +CCGTAT +CCGTAC +CCGTAG +CCGTTA +CCGTTT +CCGTTC +CCGTTG +CCGTCA +CCGTCT +CCGTCC +CCGTCG +CCGTGA +CCGTGT +CCGTGC +CCGTGG +CCGCAA +CCGCAT +CCGCAC +CCGCAG +CCGCTA +CCGCTT +CCGCTC +CCGCTG +CCGCCA +CCGCCT +CCGCCC +CCGCCG +CCGCGA +CCGCGT +CCGCGC +CCGCGG +CCGGAA +CCGGAT +CCGGAC +CCGGAG +CCGGTA +CCGGTT +CCGGTC +CCGGTG +CCGGCA +CCGGCT +CCGGCC +CCGGCG +CCGGGA +CCGGGT +CCGGGC +CCGGGG +CGAAAA +CGAAAT +CGAAAC +CGAAAG +CGAATA +CGAATT +CGAATC +CGAATG +CGAACA +CGAACT +CGAACC +CGAACG +CGAAGA +CGAAGT +CGAAGC +CGAAGG +CGATAA +CGATAT +CGATAC +CGATAG +CGATTA +CGATTT +CGATTC +CGATTG +CGATCA +CGATCT +CGATCC +CGATCG +CGATGA +CGATGT +CGATGC +CGATGG +CGACAA +CGACAT +CGACAC +CGACAG +CGACTA +CGACTT +CGACTC +CGACTG +CGACCA +CGACCT +CGACCC +CGACCG +CGACGA +CGACGT +CGACGC +CGACGG +CGAGAA +CGAGAT +CGAGAC +CGAGAG +CGAGTA +CGAGTT +CGAGTC +CGAGTG +CGAGCA +CGAGCT +CGAGCC +CGAGCG +CGAGGA +CGAGGT +CGAGGC +CGAGGG +CGTAAA +CGTAAT +CGTAAC +CGTAAG +CGTATA +CGTATT +CGTATC +CGTATG +CGTACA +CGTACT +CGTACC +CGTACG +CGTAGA +CGTAGT +CGTAGC +CGTAGG +CGTTAA +CGTTAT +CGTTAC +CGTTAG +CGTTTA +CGTTTT +CGTTTC +CGTTTG +CGTTCA +CGTTCT +CGTTCC +CGTTCG +CGTTGA +CGTTGT +CGTTGC +CGTTGG +CGTCAA +CGTCAT +CGTCAC +CGTCAG +CGTCTA +CGTCTT +CGTCTC +CGTCTG +CGTCCA +CGTCCT +CGTCCC +CGTCCG +CGTCGA +CGTCGT +CGTCGC +CGTCGG +CGTGAA +CGTGAT +CGTGAC +CGTGAG +CGTGTA +CGTGTT +CGTGTC +CGTGTG +CGTGCA +CGTGCT +CGTGCC +CGTGCG +CGTGGA +CGTGGT +CGTGGC +CGTGGG +CGCAAA +CGCAAT +CGCAAC +CGCAAG +CGCATA +CGCATT +CGCATC +CGCATG +CGCACA +CGCACT +CGCACC +CGCACG +CGCAGA +CGCAGT +CGCAGC +CGCAGG +CGCTAA +CGCTAT +CGCTAC +CGCTAG +CGCTTA +CGCTTT +CGCTTC +CGCTTG +CGCTCA +CGCTCT +CGCTCC +CGCTCG +CGCTGA +CGCTGT +CGCTGC +CGCTGG +CGCCAA +CGCCAT +CGCCAC +CGCCAG +CGCCTA +CGCCTT +CGCCTC +CGCCTG +CGCCCA +CGCCCT +CGCCCC +CGCCCG +CGCCGA +CGCCGT +CGCCGC +CGCCGG +CGCGAA +CGCGAT +CGCGAC +CGCGAG +CGCGTA +CGCGTT +CGCGTC +CGCGTG +CGCGCA +CGCGCT +CGCGCC +CGCGCG +CGCGGA +CGCGGT +CGCGGC +CGCGGG +CGGAAA +CGGAAT +CGGAAC +CGGAAG +CGGATA +CGGATT +CGGATC +CGGATG +CGGACA +CGGACT +CGGACC +CGGACG +CGGAGA +CGGAGT +CGGAGC +CGGAGG +CGGTAA +CGGTAT +CGGTAC +CGGTAG +CGGTTA +CGGTTT +CGGTTC +CGGTTG +CGGTCA +CGGTCT +CGGTCC +CGGTCG +CGGTGA +CGGTGT +CGGTGC +CGGTGG +CGGCAA +CGGCAT +CGGCAC +CGGCAG +CGGCTA +CGGCTT +CGGCTC +CGGCTG +CGGCCA +CGGCCT +CGGCCC +CGGCCG +CGGCGA +CGGCGT +CGGCGC +CGGCGG +CGGGAA +CGGGAT +CGGGAC +CGGGAG +CGGGTA +CGGGTT +CGGGTC +CGGGTG +CGGGCA +CGGGCT +CGGGCC +CGGGCG +CGGGGA +CGGGGT +CGGGGC +CGGGGG +GAAAAA +GAAAAT +GAAAAC +GAAAAG +GAAATA +GAAATT +GAAATC +GAAATG +GAAACA +GAAACT +GAAACC +GAAACG +GAAAGA +GAAAGT +GAAAGC +GAAAGG +GAATAA +GAATAT +GAATAC +GAATAG +GAATTA +GAATTT +GAATTC +GAATTG +GAATCA +GAATCT +GAATCC +GAATCG +GAATGA +GAATGT +GAATGC +GAATGG +GAACAA +GAACAT +GAACAC +GAACAG +GAACTA +GAACTT +GAACTC +GAACTG +GAACCA +GAACCT +GAACCC +GAACCG +GAACGA +GAACGT +GAACGC +GAACGG +GAAGAA +GAAGAT +GAAGAC +GAAGAG +GAAGTA +GAAGTT +GAAGTC +GAAGTG +GAAGCA +GAAGCT +GAAGCC +GAAGCG +GAAGGA +GAAGGT +GAAGGC +GAAGGG +GATAAA +GATAAT +GATAAC +GATAAG +GATATA +GATATT +GATATC +GATATG +GATACA +GATACT +GATACC +GATACG +GATAGA +GATAGT +GATAGC +GATAGG +GATTAA +GATTAT +GATTAC +GATTAG +GATTTA +GATTTT +GATTTC +GATTTG +GATTCA +GATTCT +GATTCC +GATTCG +GATTGA +GATTGT +GATTGC +GATTGG +GATCAA +GATCAT +GATCAC +GATCAG +GATCTA +GATCTT +GATCTC +GATCTG +GATCCA +GATCCT +GATCCC +GATCCG +GATCGA +GATCGT +GATCGC +GATCGG +GATGAA +GATGAT +GATGAC +GATGAG +GATGTA +GATGTT +GATGTC +GATGTG +GATGCA +GATGCT +GATGCC +GATGCG +GATGGA +GATGGT +GATGGC +GATGGG +GACAAA +GACAAT +GACAAC +GACAAG +GACATA +GACATT +GACATC +GACATG +GACACA +GACACT +GACACC +GACACG +GACAGA +GACAGT +GACAGC +GACAGG +GACTAA +GACTAT +GACTAC +GACTAG +GACTTA +GACTTT +GACTTC +GACTTG +GACTCA +GACTCT +GACTCC +GACTCG +GACTGA +GACTGT +GACTGC +GACTGG +GACCAA +GACCAT +GACCAC +GACCAG +GACCTA +GACCTT +GACCTC +GACCTG +GACCCA +GACCCT +GACCCC +GACCCG +GACCGA +GACCGT +GACCGC +GACCGG +GACGAA +GACGAT +GACGAC +GACGAG +GACGTA +GACGTT +GACGTC +GACGTG +GACGCA +GACGCT +GACGCC +GACGCG +GACGGA +GACGGT +GACGGC +GACGGG +GAGAAA +GAGAAT +GAGAAC +GAGAAG +GAGATA +GAGATT +GAGATC +GAGATG +GAGACA +GAGACT +GAGACC +GAGACG +GAGAGA +GAGAGT +GAGAGC +GAGAGG +GAGTAA +GAGTAT +GAGTAC +GAGTAG +GAGTTA +GAGTTT +GAGTTC +GAGTTG +GAGTCA +GAGTCT +GAGTCC +GAGTCG +GAGTGA +GAGTGT +GAGTGC +GAGTGG +GAGCAA +GAGCAT +GAGCAC +GAGCAG +GAGCTA +GAGCTT +GAGCTC +GAGCTG +GAGCCA +GAGCCT +GAGCCC +GAGCCG +GAGCGA +GAGCGT +GAGCGC +GAGCGG +GAGGAA +GAGGAT +GAGGAC +GAGGAG +GAGGTA +GAGGTT +GAGGTC +GAGGTG +GAGGCA +GAGGCT +GAGGCC +GAGGCG +GAGGGA +GAGGGT +GAGGGC +GAGGGG +GTAAAA +GTAAAT +GTAAAC +GTAAAG +GTAATA +GTAATT +GTAATC +GTAATG +GTAACA +GTAACT +GTAACC +GTAACG +GTAAGA +GTAAGT +GTAAGC +GTAAGG +GTATAA +GTATAT +GTATAC +GTATAG +GTATTA +GTATTT +GTATTC +GTATTG +GTATCA +GTATCT +GTATCC +GTATCG +GTATGA +GTATGT +GTATGC +GTATGG +GTACAA +GTACAT +GTACAC +GTACAG +GTACTA +GTACTT +GTACTC +GTACTG +GTACCA +GTACCT +GTACCC +GTACCG +GTACGA +GTACGT +GTACGC +GTACGG +GTAGAA +GTAGAT +GTAGAC +GTAGAG +GTAGTA +GTAGTT +GTAGTC +GTAGTG +GTAGCA +GTAGCT +GTAGCC +GTAGCG +GTAGGA +GTAGGT +GTAGGC +GTAGGG +GTTAAA +GTTAAT +GTTAAC +GTTAAG +GTTATA +GTTATT +GTTATC +GTTATG +GTTACA +GTTACT +GTTACC +GTTACG +GTTAGA +GTTAGT +GTTAGC +GTTAGG +GTTTAA +GTTTAT +GTTTAC +GTTTAG +GTTTTA +GTTTTT +GTTTTC +GTTTTG +GTTTCA +GTTTCT +GTTTCC +GTTTCG +GTTTGA +GTTTGT +GTTTGC +GTTTGG +GTTCAA +GTTCAT +GTTCAC +GTTCAG +GTTCTA +GTTCTT +GTTCTC +GTTCTG +GTTCCA +GTTCCT +GTTCCC +GTTCCG +GTTCGA +GTTCGT +GTTCGC +GTTCGG +GTTGAA +GTTGAT +GTTGAC +GTTGAG +GTTGTA +GTTGTT +GTTGTC +GTTGTG +GTTGCA +GTTGCT +GTTGCC +GTTGCG +GTTGGA +GTTGGT +GTTGGC +GTTGGG +GTCAAA +GTCAAT +GTCAAC +GTCAAG +GTCATA +GTCATT +GTCATC +GTCATG +GTCACA +GTCACT +GTCACC +GTCACG +GTCAGA +GTCAGT +GTCAGC +GTCAGG +GTCTAA +GTCTAT +GTCTAC +GTCTAG +GTCTTA +GTCTTT +GTCTTC +GTCTTG +GTCTCA +GTCTCT +GTCTCC +GTCTCG +GTCTGA +GTCTGT +GTCTGC +GTCTGG +GTCCAA +GTCCAT +GTCCAC +GTCCAG +GTCCTA +GTCCTT +GTCCTC +GTCCTG +GTCCCA +GTCCCT +GTCCCC +GTCCCG +GTCCGA +GTCCGT +GTCCGC +GTCCGG +GTCGAA +GTCGAT +GTCGAC +GTCGAG +GTCGTA +GTCGTT +GTCGTC +GTCGTG +GTCGCA +GTCGCT +GTCGCC +GTCGCG +GTCGGA +GTCGGT +GTCGGC +GTCGGG +GTGAAA +GTGAAT +GTGAAC +GTGAAG +GTGATA +GTGATT +GTGATC +GTGATG +GTGACA +GTGACT +GTGACC +GTGACG +GTGAGA +GTGAGT +GTGAGC +GTGAGG +GTGTAA +GTGTAT +GTGTAC +GTGTAG +GTGTTA +GTGTTT +GTGTTC +GTGTTG +GTGTCA +GTGTCT +GTGTCC +GTGTCG +GTGTGA +GTGTGT +GTGTGC +GTGTGG +GTGCAA +GTGCAT +GTGCAC +GTGCAG +GTGCTA +GTGCTT +GTGCTC +GTGCTG +GTGCCA +GTGCCT +GTGCCC +GTGCCG +GTGCGA +GTGCGT +GTGCGC +GTGCGG +GTGGAA +GTGGAT +GTGGAC +GTGGAG +GTGGTA +GTGGTT +GTGGTC +GTGGTG +GTGGCA +GTGGCT +GTGGCC +GTGGCG +GTGGGA +GTGGGT +GTGGGC +GTGGGG +GCAAAA +GCAAAT +GCAAAC +GCAAAG +GCAATA +GCAATT +GCAATC +GCAATG +GCAACA +GCAACT +GCAACC +GCAACG +GCAAGA +GCAAGT +GCAAGC +GCAAGG +GCATAA +GCATAT +GCATAC +GCATAG +GCATTA +GCATTT +GCATTC +GCATTG +GCATCA +GCATCT +GCATCC +GCATCG +GCATGA +GCATGT +GCATGC +GCATGG +GCACAA +GCACAT +GCACAC +GCACAG +GCACTA +GCACTT +GCACTC +GCACTG +GCACCA +GCACCT +GCACCC +GCACCG +GCACGA +GCACGT +GCACGC +GCACGG +GCAGAA +GCAGAT +GCAGAC +GCAGAG +GCAGTA +GCAGTT +GCAGTC +GCAGTG +GCAGCA +GCAGCT +GCAGCC +GCAGCG +GCAGGA +GCAGGT +GCAGGC +GCAGGG +GCTAAA +GCTAAT +GCTAAC +GCTAAG +GCTATA +GCTATT +GCTATC +GCTATG +GCTACA +GCTACT +GCTACC +GCTACG +GCTAGA +GCTAGT +GCTAGC +GCTAGG +GCTTAA +GCTTAT +GCTTAC +GCTTAG +GCTTTA +GCTTTT +GCTTTC +GCTTTG +GCTTCA +GCTTCT +GCTTCC +GCTTCG +GCTTGA +GCTTGT +GCTTGC +GCTTGG +GCTCAA +GCTCAT +GCTCAC +GCTCAG +GCTCTA +GCTCTT +GCTCTC +GCTCTG +GCTCCA +GCTCCT +GCTCCC +GCTCCG +GCTCGA +GCTCGT +GCTCGC +GCTCGG +GCTGAA +GCTGAT +GCTGAC +GCTGAG +GCTGTA +GCTGTT +GCTGTC +GCTGTG +GCTGCA +GCTGCT +GCTGCC +GCTGCG +GCTGGA +GCTGGT +GCTGGC +GCTGGG +GCCAAA +GCCAAT +GCCAAC +GCCAAG +GCCATA +GCCATT +GCCATC +GCCATG +GCCACA +GCCACT +GCCACC +GCCACG +GCCAGA +GCCAGT +GCCAGC +GCCAGG +GCCTAA +GCCTAT +GCCTAC +GCCTAG +GCCTTA +GCCTTT +GCCTTC +GCCTTG +GCCTCA +GCCTCT +GCCTCC +GCCTCG +GCCTGA +GCCTGT +GCCTGC +GCCTGG +GCCCAA +GCCCAT +GCCCAC +GCCCAG +GCCCTA +GCCCTT +GCCCTC +GCCCTG +GCCCCA +GCCCCT +GCCCCC +GCCCCG +GCCCGA +GCCCGT +GCCCGC +GCCCGG +GCCGAA +GCCGAT +GCCGAC +GCCGAG +GCCGTA +GCCGTT +GCCGTC +GCCGTG +GCCGCA +GCCGCT +GCCGCC +GCCGCG +GCCGGA +GCCGGT +GCCGGC +GCCGGG +GCGAAA +GCGAAT +GCGAAC +GCGAAG +GCGATA +GCGATT +GCGATC +GCGATG +GCGACA +GCGACT +GCGACC +GCGACG +GCGAGA +GCGAGT +GCGAGC +GCGAGG +GCGTAA +GCGTAT +GCGTAC +GCGTAG +GCGTTA +GCGTTT +GCGTTC +GCGTTG +GCGTCA +GCGTCT +GCGTCC +GCGTCG +GCGTGA +GCGTGT +GCGTGC +GCGTGG +GCGCAA +GCGCAT +GCGCAC +GCGCAG +GCGCTA +GCGCTT +GCGCTC +GCGCTG +GCGCCA +GCGCCT +GCGCCC +GCGCCG +GCGCGA +GCGCGT +GCGCGC +GCGCGG +GCGGAA +GCGGAT +GCGGAC +GCGGAG +GCGGTA +GCGGTT +GCGGTC +GCGGTG +GCGGCA +GCGGCT +GCGGCC +GCGGCG +GCGGGA +GCGGGT +GCGGGC +GCGGGG +GGAAAA +GGAAAT +GGAAAC +GGAAAG +GGAATA +GGAATT +GGAATC +GGAATG +GGAACA +GGAACT +GGAACC +GGAACG +GGAAGA +GGAAGT +GGAAGC +GGAAGG +GGATAA +GGATAT +GGATAC +GGATAG +GGATTA +GGATTT +GGATTC +GGATTG +GGATCA +GGATCT +GGATCC +GGATCG +GGATGA +GGATGT +GGATGC +GGATGG +GGACAA +GGACAT +GGACAC +GGACAG +GGACTA +GGACTT +GGACTC +GGACTG +GGACCA +GGACCT +GGACCC +GGACCG +GGACGA +GGACGT +GGACGC +GGACGG +GGAGAA +GGAGAT +GGAGAC +GGAGAG +GGAGTA +GGAGTT +GGAGTC +GGAGTG +GGAGCA +GGAGCT +GGAGCC +GGAGCG +GGAGGA +GGAGGT +GGAGGC +GGAGGG +GGTAAA +GGTAAT +GGTAAC +GGTAAG +GGTATA +GGTATT +GGTATC +GGTATG +GGTACA +GGTACT +GGTACC +GGTACG +GGTAGA +GGTAGT +GGTAGC +GGTAGG +GGTTAA +GGTTAT +GGTTAC +GGTTAG +GGTTTA +GGTTTT +GGTTTC +GGTTTG +GGTTCA +GGTTCT +GGTTCC +GGTTCG +GGTTGA +GGTTGT +GGTTGC +GGTTGG +GGTCAA +GGTCAT +GGTCAC +GGTCAG +GGTCTA +GGTCTT +GGTCTC +GGTCTG +GGTCCA +GGTCCT +GGTCCC +GGTCCG +GGTCGA +GGTCGT +GGTCGC +GGTCGG +GGTGAA +GGTGAT +GGTGAC +GGTGAG +GGTGTA +GGTGTT +GGTGTC +GGTGTG +GGTGCA +GGTGCT +GGTGCC +GGTGCG +GGTGGA +GGTGGT +GGTGGC +GGTGGG +GGCAAA +GGCAAT +GGCAAC +GGCAAG +GGCATA +GGCATT +GGCATC +GGCATG +GGCACA +GGCACT +GGCACC +GGCACG +GGCAGA +GGCAGT +GGCAGC +GGCAGG +GGCTAA +GGCTAT +GGCTAC +GGCTAG +GGCTTA +GGCTTT +GGCTTC +GGCTTG +GGCTCA +GGCTCT +GGCTCC +GGCTCG +GGCTGA +GGCTGT +GGCTGC +GGCTGG +GGCCAA +GGCCAT +GGCCAC +GGCCAG +GGCCTA +GGCCTT +GGCCTC +GGCCTG +GGCCCA +GGCCCT +GGCCCC +GGCCCG +GGCCGA +GGCCGT +GGCCGC +GGCCGG +GGCGAA +GGCGAT +GGCGAC +GGCGAG +GGCGTA +GGCGTT +GGCGTC +GGCGTG +GGCGCA +GGCGCT +GGCGCC +GGCGCG +GGCGGA +GGCGGT +GGCGGC +GGCGGG +GGGAAA +GGGAAT +GGGAAC +GGGAAG +GGGATA +GGGATT +GGGATC +GGGATG +GGGACA +GGGACT +GGGACC +GGGACG +GGGAGA +GGGAGT +GGGAGC +GGGAGG +GGGTAA +GGGTAT +GGGTAC +GGGTAG +GGGTTA +GGGTTT +GGGTTC +GGGTTG +GGGTCA +GGGTCT +GGGTCC +GGGTCG +GGGTGA +GGGTGT +GGGTGC +GGGTGG +GGGCAA +GGGCAT +GGGCAC +GGGCAG +GGGCTA +GGGCTT +GGGCTC +GGGCTG +GGGCCA +GGGCCT +GGGCCC +GGGCCG +GGGCGA +GGGCGT +GGGCGC +GGGCGG +GGGGAA +GGGGAT +GGGGAC +GGGGAG +GGGGTA +GGGGTT +GGGGTC +GGGGTG +GGGGCA +GGGGCT +GGGGCC +GGGGCG +GGGGGA +GGGGGT +GGGGGC +GGGGGG diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-4000/config.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..9b14e61d327bf586fcff6e817cf9d365bae18ad2 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-4000/config.json @@ -0,0 +1,26 @@ +{ + "_name_or_path": "fabihamakhdoomi/TinyDNABERT", + "architectures": [ + "BertForSequenceClassification" + ], + "attention_probs_dropout_prob": 0.1, + "classifier_dropout": null, + "hidden_act": "gelu", + "hidden_dropout_prob": 0.1, + "hidden_size": 128, + "initializer_range": 0.02, + "intermediate_size": 512, + "layer_norm_eps": 1e-12, + "max_position_embeddings": 512, + "model_type": "bert", + "num_attention_heads": 2, + "num_hidden_layers": 2, + "pad_token_id": 0, + "position_embedding_type": "absolute", + "problem_type": "single_label_classification", + "torch_dtype": "float32", + "transformers_version": "4.44.0", + "type_vocab_size": 2, + "use_cache": true, + "vocab_size": 4101 +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-4000/model.safetensors b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-4000/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..4018f0deff6d4289e0da05cbe2891927fa686633 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-4000/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1e41a9e760852cfaacb4081a59019ba27352a275f46518d85b1531c77f47b063 +size 4021688 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-4000/optimizer.pt b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-4000/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..02aed7e5b262c686a42887742188578e35c155ca --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-4000/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:10c13f9e83353de21a86607f59de3779b6e80df301caf5b5e2756ba37c37df1a +size 8058700 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-4000/rng_state.pth b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-4000/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..e341a4a808ae24c21407bcc648bed6b2d6a16c6e --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-4000/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bc0eaa426d9fdf3c0f9bd6754ed553c3bff648a3431f959a858959a437717ed6 +size 14244 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-4000/scheduler.pt b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-4000/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..8c9a6adac55ac4d1afd6e1b61889589acfe31330 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-4000/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f3e454de3c0f52ddb8e5884c9783308658a6f61ede2ed6e377ca9a68d47c4515 +size 1064 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-4000/special_tokens_map.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-4000/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..a8b3208c2884c4efb86e49300fdd3dc877220cdf --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-4000/special_tokens_map.json @@ -0,0 +1,7 @@ +{ + "cls_token": "[CLS]", + "mask_token": "[MASK]", + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-4000/tokenizer.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-4000/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..ebd782de6963b9032ce9374cfe30f5275be1a30b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-4000/tokenizer.json @@ -0,0 +1,4265 @@ +{ + "version": "1.0", + "truncation": { + "direction": "Right", + "max_length": 512, + "strategy": "LongestFirst", + "stride": 0 + }, + "padding": { + "strategy": { + "Fixed": 512 + }, + "direction": "Right", + "pad_to_multiple_of": null, + "pad_id": 0, + "pad_type_id": 0, + "pad_token": "[PAD]" + }, + "added_tokens": [ + { + "id": 0, + "content": "[PAD]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 1, + "content": "[UNK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 2, + "content": "[CLS]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 3, + "content": "[SEP]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 4, + "content": "[MASK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + } + ], + "normalizer": { + "type": "BertNormalizer", + "clean_text": true, + "handle_chinese_chars": true, + "strip_accents": null, + "lowercase": false + }, + "pre_tokenizer": { + "type": "BertPreTokenizer" + }, + "post_processor": { + "type": "TemplateProcessing", + "single": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + } + ], + "pair": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "B", + "type_id": 1 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 1 + } + } + ], + "special_tokens": { + "[CLS]": { + "id": "[CLS]", + "ids": [ + 2 + ], + "tokens": [ + "[CLS]" + ] + }, + "[SEP]": { + "id": "[SEP]", + "ids": [ + 3 + ], + "tokens": [ + "[SEP]" + ] + } + } + }, + "decoder": { + "type": "WordPiece", + "prefix": "##", + "cleanup": true + }, + "model": { + "type": "WordPiece", + "unk_token": "[UNK]", + "continuing_subword_prefix": "##", + "max_input_chars_per_word": 100, + "vocab": { + "[PAD]": 0, + "[UNK]": 1, + "[CLS]": 2, + "[SEP]": 3, + "[MASK]": 4, + "AAAAAA": 5, + "AAAAAT": 6, + "AAAAAC": 7, + "AAAAAG": 8, + "AAAATA": 9, + "AAAATT": 10, + "AAAATC": 11, + "AAAATG": 12, + "AAAACA": 13, + "AAAACT": 14, + "AAAACC": 15, + "AAAACG": 16, + "AAAAGA": 17, + "AAAAGT": 18, + "AAAAGC": 19, + "AAAAGG": 20, + "AAATAA": 21, + "AAATAT": 22, + "AAATAC": 23, + "AAATAG": 24, + "AAATTA": 25, + "AAATTT": 26, + "AAATTC": 27, + "AAATTG": 28, + "AAATCA": 29, + "AAATCT": 30, + "AAATCC": 31, + "AAATCG": 32, + "AAATGA": 33, + "AAATGT": 34, + "AAATGC": 35, + "AAATGG": 36, + "AAACAA": 37, + "AAACAT": 38, + "AAACAC": 39, + "AAACAG": 40, + "AAACTA": 41, + "AAACTT": 42, + "AAACTC": 43, + "AAACTG": 44, + "AAACCA": 45, + "AAACCT": 46, + "AAACCC": 47, + "AAACCG": 48, + "AAACGA": 49, + "AAACGT": 50, + "AAACGC": 51, + "AAACGG": 52, + "AAAGAA": 53, + "AAAGAT": 54, + "AAAGAC": 55, + "AAAGAG": 56, + "AAAGTA": 57, + "AAAGTT": 58, + "AAAGTC": 59, + "AAAGTG": 60, + "AAAGCA": 61, + "AAAGCT": 62, + "AAAGCC": 63, + "AAAGCG": 64, + "AAAGGA": 65, + "AAAGGT": 66, + "AAAGGC": 67, + "AAAGGG": 68, + "AATAAA": 69, + "AATAAT": 70, + "AATAAC": 71, + "AATAAG": 72, + "AATATA": 73, + "AATATT": 74, + "AATATC": 75, + "AATATG": 76, + "AATACA": 77, + "AATACT": 78, + "AATACC": 79, + "AATACG": 80, + "AATAGA": 81, + "AATAGT": 82, + "AATAGC": 83, + "AATAGG": 84, + "AATTAA": 85, + "AATTAT": 86, + "AATTAC": 87, + "AATTAG": 88, + "AATTTA": 89, + "AATTTT": 90, + "AATTTC": 91, + "AATTTG": 92, + "AATTCA": 93, + "AATTCT": 94, + "AATTCC": 95, + "AATTCG": 96, + "AATTGA": 97, + "AATTGT": 98, + "AATTGC": 99, + "AATTGG": 100, + "AATCAA": 101, + "AATCAT": 102, + "AATCAC": 103, + "AATCAG": 104, + "AATCTA": 105, + "AATCTT": 106, + "AATCTC": 107, + "AATCTG": 108, + "AATCCA": 109, + "AATCCT": 110, + "AATCCC": 111, + "AATCCG": 112, + "AATCGA": 113, + "AATCGT": 114, + "AATCGC": 115, + "AATCGG": 116, + "AATGAA": 117, + "AATGAT": 118, + "AATGAC": 119, + "AATGAG": 120, + "AATGTA": 121, + "AATGTT": 122, + "AATGTC": 123, + "AATGTG": 124, + "AATGCA": 125, + "AATGCT": 126, + "AATGCC": 127, + "AATGCG": 128, + "AATGGA": 129, + "AATGGT": 130, + "AATGGC": 131, + "AATGGG": 132, + "AACAAA": 133, + "AACAAT": 134, + "AACAAC": 135, + "AACAAG": 136, + "AACATA": 137, + "AACATT": 138, + "AACATC": 139, + "AACATG": 140, + "AACACA": 141, + "AACACT": 142, + "AACACC": 143, + "AACACG": 144, + "AACAGA": 145, + "AACAGT": 146, + "AACAGC": 147, + "AACAGG": 148, + "AACTAA": 149, + "AACTAT": 150, + "AACTAC": 151, + "AACTAG": 152, + "AACTTA": 153, + "AACTTT": 154, + "AACTTC": 155, + "AACTTG": 156, + "AACTCA": 157, + "AACTCT": 158, + "AACTCC": 159, + "AACTCG": 160, + "AACTGA": 161, + "AACTGT": 162, + "AACTGC": 163, + "AACTGG": 164, + "AACCAA": 165, + "AACCAT": 166, + "AACCAC": 167, + "AACCAG": 168, + "AACCTA": 169, + "AACCTT": 170, + "AACCTC": 171, + "AACCTG": 172, + "AACCCA": 173, + "AACCCT": 174, + "AACCCC": 175, + "AACCCG": 176, + "AACCGA": 177, + "AACCGT": 178, + "AACCGC": 179, + "AACCGG": 180, + "AACGAA": 181, + "AACGAT": 182, + "AACGAC": 183, + "AACGAG": 184, + "AACGTA": 185, + "AACGTT": 186, + "AACGTC": 187, + "AACGTG": 188, + "AACGCA": 189, + "AACGCT": 190, + "AACGCC": 191, + "AACGCG": 192, + "AACGGA": 193, + "AACGGT": 194, + "AACGGC": 195, + "AACGGG": 196, + "AAGAAA": 197, + "AAGAAT": 198, + "AAGAAC": 199, + "AAGAAG": 200, + "AAGATA": 201, + "AAGATT": 202, + "AAGATC": 203, + "AAGATG": 204, + "AAGACA": 205, + "AAGACT": 206, + "AAGACC": 207, + "AAGACG": 208, + "AAGAGA": 209, + "AAGAGT": 210, + "AAGAGC": 211, + "AAGAGG": 212, + "AAGTAA": 213, + "AAGTAT": 214, + "AAGTAC": 215, + "AAGTAG": 216, + "AAGTTA": 217, + "AAGTTT": 218, + "AAGTTC": 219, + "AAGTTG": 220, + "AAGTCA": 221, + "AAGTCT": 222, + "AAGTCC": 223, + "AAGTCG": 224, + "AAGTGA": 225, + "AAGTGT": 226, + "AAGTGC": 227, + "AAGTGG": 228, + "AAGCAA": 229, + "AAGCAT": 230, + "AAGCAC": 231, + "AAGCAG": 232, + "AAGCTA": 233, + "AAGCTT": 234, + "AAGCTC": 235, + "AAGCTG": 236, + "AAGCCA": 237, + "AAGCCT": 238, + "AAGCCC": 239, + "AAGCCG": 240, + "AAGCGA": 241, + "AAGCGT": 242, + "AAGCGC": 243, + "AAGCGG": 244, + "AAGGAA": 245, + "AAGGAT": 246, + "AAGGAC": 247, + "AAGGAG": 248, + "AAGGTA": 249, + "AAGGTT": 250, + "AAGGTC": 251, + "AAGGTG": 252, + "AAGGCA": 253, + "AAGGCT": 254, + "AAGGCC": 255, + "AAGGCG": 256, + "AAGGGA": 257, + "AAGGGT": 258, + "AAGGGC": 259, + "AAGGGG": 260, + "ATAAAA": 261, + "ATAAAT": 262, + "ATAAAC": 263, + "ATAAAG": 264, + "ATAATA": 265, + "ATAATT": 266, + "ATAATC": 267, + "ATAATG": 268, + "ATAACA": 269, + "ATAACT": 270, + "ATAACC": 271, + "ATAACG": 272, + "ATAAGA": 273, + "ATAAGT": 274, + "ATAAGC": 275, + "ATAAGG": 276, + "ATATAA": 277, + "ATATAT": 278, + "ATATAC": 279, + "ATATAG": 280, + "ATATTA": 281, + "ATATTT": 282, + "ATATTC": 283, + "ATATTG": 284, + "ATATCA": 285, + "ATATCT": 286, + "ATATCC": 287, + "ATATCG": 288, + "ATATGA": 289, + "ATATGT": 290, + "ATATGC": 291, + "ATATGG": 292, + "ATACAA": 293, + "ATACAT": 294, + "ATACAC": 295, + "ATACAG": 296, + "ATACTA": 297, + "ATACTT": 298, + "ATACTC": 299, + "ATACTG": 300, + "ATACCA": 301, + "ATACCT": 302, + "ATACCC": 303, + "ATACCG": 304, + "ATACGA": 305, + "ATACGT": 306, + "ATACGC": 307, + "ATACGG": 308, + "ATAGAA": 309, + "ATAGAT": 310, + "ATAGAC": 311, + "ATAGAG": 312, + "ATAGTA": 313, + "ATAGTT": 314, + "ATAGTC": 315, + "ATAGTG": 316, + "ATAGCA": 317, + "ATAGCT": 318, + "ATAGCC": 319, + "ATAGCG": 320, + "ATAGGA": 321, + "ATAGGT": 322, + "ATAGGC": 323, + "ATAGGG": 324, + "ATTAAA": 325, + "ATTAAT": 326, + "ATTAAC": 327, + "ATTAAG": 328, + "ATTATA": 329, + "ATTATT": 330, + "ATTATC": 331, + "ATTATG": 332, + "ATTACA": 333, + "ATTACT": 334, + "ATTACC": 335, + "ATTACG": 336, + "ATTAGA": 337, + "ATTAGT": 338, + "ATTAGC": 339, + "ATTAGG": 340, + "ATTTAA": 341, + "ATTTAT": 342, + "ATTTAC": 343, + "ATTTAG": 344, + "ATTTTA": 345, + "ATTTTT": 346, + "ATTTTC": 347, + "ATTTTG": 348, + "ATTTCA": 349, + "ATTTCT": 350, + "ATTTCC": 351, + "ATTTCG": 352, + "ATTTGA": 353, + "ATTTGT": 354, + "ATTTGC": 355, + "ATTTGG": 356, + "ATTCAA": 357, + "ATTCAT": 358, + "ATTCAC": 359, + "ATTCAG": 360, + "ATTCTA": 361, + "ATTCTT": 362, + "ATTCTC": 363, + "ATTCTG": 364, + "ATTCCA": 365, + "ATTCCT": 366, + "ATTCCC": 367, + "ATTCCG": 368, + "ATTCGA": 369, + "ATTCGT": 370, + "ATTCGC": 371, + "ATTCGG": 372, + "ATTGAA": 373, + "ATTGAT": 374, + "ATTGAC": 375, + "ATTGAG": 376, + "ATTGTA": 377, + "ATTGTT": 378, + "ATTGTC": 379, + "ATTGTG": 380, + "ATTGCA": 381, + "ATTGCT": 382, + "ATTGCC": 383, + "ATTGCG": 384, + "ATTGGA": 385, + "ATTGGT": 386, + "ATTGGC": 387, + "ATTGGG": 388, + "ATCAAA": 389, + "ATCAAT": 390, + "ATCAAC": 391, + "ATCAAG": 392, + "ATCATA": 393, + "ATCATT": 394, + "ATCATC": 395, + "ATCATG": 396, + "ATCACA": 397, + "ATCACT": 398, + "ATCACC": 399, + "ATCACG": 400, + "ATCAGA": 401, + "ATCAGT": 402, + "ATCAGC": 403, + "ATCAGG": 404, + "ATCTAA": 405, + "ATCTAT": 406, + "ATCTAC": 407, + "ATCTAG": 408, + "ATCTTA": 409, + "ATCTTT": 410, + "ATCTTC": 411, + "ATCTTG": 412, + "ATCTCA": 413, + "ATCTCT": 414, + "ATCTCC": 415, + "ATCTCG": 416, + "ATCTGA": 417, + "ATCTGT": 418, + "ATCTGC": 419, + "ATCTGG": 420, + "ATCCAA": 421, + "ATCCAT": 422, + "ATCCAC": 423, + "ATCCAG": 424, + "ATCCTA": 425, + "ATCCTT": 426, + "ATCCTC": 427, + "ATCCTG": 428, + "ATCCCA": 429, + "ATCCCT": 430, + "ATCCCC": 431, + "ATCCCG": 432, + "ATCCGA": 433, + "ATCCGT": 434, + "ATCCGC": 435, + "ATCCGG": 436, + "ATCGAA": 437, + "ATCGAT": 438, + "ATCGAC": 439, + "ATCGAG": 440, + "ATCGTA": 441, + "ATCGTT": 442, + "ATCGTC": 443, + "ATCGTG": 444, + "ATCGCA": 445, + "ATCGCT": 446, + "ATCGCC": 447, + "ATCGCG": 448, + "ATCGGA": 449, + "ATCGGT": 450, + "ATCGGC": 451, + "ATCGGG": 452, + "ATGAAA": 453, + "ATGAAT": 454, + "ATGAAC": 455, + "ATGAAG": 456, + "ATGATA": 457, + "ATGATT": 458, + "ATGATC": 459, + "ATGATG": 460, + "ATGACA": 461, + "ATGACT": 462, + "ATGACC": 463, + "ATGACG": 464, + "ATGAGA": 465, + "ATGAGT": 466, + "ATGAGC": 467, + "ATGAGG": 468, + "ATGTAA": 469, + "ATGTAT": 470, + "ATGTAC": 471, + "ATGTAG": 472, + "ATGTTA": 473, + "ATGTTT": 474, + "ATGTTC": 475, + "ATGTTG": 476, + "ATGTCA": 477, + "ATGTCT": 478, + "ATGTCC": 479, + "ATGTCG": 480, + "ATGTGA": 481, + "ATGTGT": 482, + "ATGTGC": 483, + "ATGTGG": 484, + "ATGCAA": 485, + "ATGCAT": 486, + "ATGCAC": 487, + "ATGCAG": 488, + "ATGCTA": 489, + "ATGCTT": 490, + "ATGCTC": 491, + "ATGCTG": 492, + "ATGCCA": 493, + "ATGCCT": 494, + "ATGCCC": 495, + "ATGCCG": 496, + "ATGCGA": 497, + "ATGCGT": 498, + "ATGCGC": 499, + "ATGCGG": 500, + "ATGGAA": 501, + "ATGGAT": 502, + "ATGGAC": 503, + "ATGGAG": 504, + "ATGGTA": 505, + "ATGGTT": 506, + "ATGGTC": 507, + "ATGGTG": 508, + "ATGGCA": 509, + "ATGGCT": 510, + "ATGGCC": 511, + "ATGGCG": 512, + "ATGGGA": 513, + "ATGGGT": 514, + "ATGGGC": 515, + "ATGGGG": 516, + "ACAAAA": 517, + "ACAAAT": 518, + "ACAAAC": 519, + "ACAAAG": 520, + "ACAATA": 521, + "ACAATT": 522, + "ACAATC": 523, + "ACAATG": 524, + "ACAACA": 525, + "ACAACT": 526, + "ACAACC": 527, + "ACAACG": 528, + "ACAAGA": 529, + "ACAAGT": 530, + "ACAAGC": 531, + "ACAAGG": 532, + "ACATAA": 533, + "ACATAT": 534, + "ACATAC": 535, + "ACATAG": 536, + "ACATTA": 537, + "ACATTT": 538, + "ACATTC": 539, + "ACATTG": 540, + "ACATCA": 541, + "ACATCT": 542, + "ACATCC": 543, + "ACATCG": 544, + "ACATGA": 545, + "ACATGT": 546, + "ACATGC": 547, + "ACATGG": 548, + "ACACAA": 549, + "ACACAT": 550, + "ACACAC": 551, + "ACACAG": 552, + "ACACTA": 553, + "ACACTT": 554, + "ACACTC": 555, + "ACACTG": 556, + "ACACCA": 557, + "ACACCT": 558, + "ACACCC": 559, + "ACACCG": 560, + "ACACGA": 561, + "ACACGT": 562, + "ACACGC": 563, + "ACACGG": 564, + "ACAGAA": 565, + "ACAGAT": 566, + "ACAGAC": 567, + "ACAGAG": 568, + "ACAGTA": 569, + "ACAGTT": 570, + "ACAGTC": 571, + "ACAGTG": 572, + "ACAGCA": 573, + "ACAGCT": 574, + "ACAGCC": 575, + "ACAGCG": 576, + "ACAGGA": 577, + "ACAGGT": 578, + "ACAGGC": 579, + "ACAGGG": 580, + "ACTAAA": 581, + "ACTAAT": 582, + "ACTAAC": 583, + "ACTAAG": 584, + "ACTATA": 585, + "ACTATT": 586, + "ACTATC": 587, + "ACTATG": 588, + "ACTACA": 589, + "ACTACT": 590, + "ACTACC": 591, + "ACTACG": 592, + "ACTAGA": 593, + "ACTAGT": 594, + "ACTAGC": 595, + "ACTAGG": 596, + "ACTTAA": 597, + "ACTTAT": 598, + "ACTTAC": 599, + "ACTTAG": 600, + "ACTTTA": 601, + "ACTTTT": 602, + "ACTTTC": 603, + "ACTTTG": 604, + "ACTTCA": 605, + "ACTTCT": 606, + "ACTTCC": 607, + "ACTTCG": 608, + "ACTTGA": 609, + "ACTTGT": 610, + "ACTTGC": 611, + "ACTTGG": 612, + "ACTCAA": 613, + "ACTCAT": 614, + "ACTCAC": 615, + "ACTCAG": 616, + "ACTCTA": 617, + "ACTCTT": 618, + "ACTCTC": 619, + "ACTCTG": 620, + "ACTCCA": 621, + "ACTCCT": 622, + "ACTCCC": 623, + "ACTCCG": 624, + "ACTCGA": 625, + "ACTCGT": 626, + "ACTCGC": 627, + "ACTCGG": 628, + "ACTGAA": 629, + "ACTGAT": 630, + "ACTGAC": 631, + "ACTGAG": 632, + "ACTGTA": 633, + "ACTGTT": 634, + "ACTGTC": 635, + "ACTGTG": 636, + "ACTGCA": 637, + "ACTGCT": 638, + "ACTGCC": 639, + "ACTGCG": 640, + "ACTGGA": 641, + "ACTGGT": 642, + "ACTGGC": 643, + "ACTGGG": 644, + "ACCAAA": 645, + "ACCAAT": 646, + "ACCAAC": 647, + "ACCAAG": 648, + "ACCATA": 649, + "ACCATT": 650, + "ACCATC": 651, + "ACCATG": 652, + "ACCACA": 653, + "ACCACT": 654, + "ACCACC": 655, + "ACCACG": 656, + "ACCAGA": 657, + "ACCAGT": 658, + "ACCAGC": 659, + "ACCAGG": 660, + "ACCTAA": 661, + "ACCTAT": 662, + "ACCTAC": 663, + "ACCTAG": 664, + "ACCTTA": 665, + "ACCTTT": 666, + "ACCTTC": 667, + "ACCTTG": 668, + "ACCTCA": 669, + "ACCTCT": 670, + "ACCTCC": 671, + "ACCTCG": 672, + "ACCTGA": 673, + "ACCTGT": 674, + "ACCTGC": 675, + "ACCTGG": 676, + "ACCCAA": 677, + "ACCCAT": 678, + "ACCCAC": 679, + "ACCCAG": 680, + "ACCCTA": 681, + "ACCCTT": 682, + "ACCCTC": 683, + "ACCCTG": 684, + "ACCCCA": 685, + "ACCCCT": 686, + "ACCCCC": 687, + "ACCCCG": 688, + "ACCCGA": 689, + "ACCCGT": 690, + "ACCCGC": 691, + "ACCCGG": 692, + "ACCGAA": 693, + "ACCGAT": 694, + "ACCGAC": 695, + "ACCGAG": 696, + "ACCGTA": 697, + "ACCGTT": 698, + "ACCGTC": 699, + "ACCGTG": 700, + "ACCGCA": 701, + "ACCGCT": 702, + "ACCGCC": 703, + "ACCGCG": 704, + "ACCGGA": 705, + "ACCGGT": 706, + "ACCGGC": 707, + "ACCGGG": 708, + "ACGAAA": 709, + "ACGAAT": 710, + "ACGAAC": 711, + "ACGAAG": 712, + "ACGATA": 713, + "ACGATT": 714, + "ACGATC": 715, + "ACGATG": 716, + "ACGACA": 717, + "ACGACT": 718, + "ACGACC": 719, + "ACGACG": 720, + "ACGAGA": 721, + "ACGAGT": 722, + "ACGAGC": 723, + "ACGAGG": 724, + "ACGTAA": 725, + "ACGTAT": 726, + "ACGTAC": 727, + "ACGTAG": 728, + "ACGTTA": 729, + "ACGTTT": 730, + "ACGTTC": 731, + "ACGTTG": 732, + "ACGTCA": 733, + "ACGTCT": 734, + "ACGTCC": 735, + "ACGTCG": 736, + "ACGTGA": 737, + "ACGTGT": 738, + "ACGTGC": 739, + "ACGTGG": 740, + "ACGCAA": 741, + "ACGCAT": 742, + "ACGCAC": 743, + "ACGCAG": 744, + "ACGCTA": 745, + "ACGCTT": 746, + "ACGCTC": 747, + "ACGCTG": 748, + "ACGCCA": 749, + "ACGCCT": 750, + "ACGCCC": 751, + "ACGCCG": 752, + "ACGCGA": 753, + "ACGCGT": 754, + "ACGCGC": 755, + "ACGCGG": 756, + "ACGGAA": 757, + "ACGGAT": 758, + "ACGGAC": 759, + "ACGGAG": 760, + "ACGGTA": 761, + "ACGGTT": 762, + "ACGGTC": 763, + "ACGGTG": 764, + "ACGGCA": 765, + "ACGGCT": 766, + "ACGGCC": 767, + "ACGGCG": 768, + "ACGGGA": 769, + "ACGGGT": 770, + "ACGGGC": 771, + "ACGGGG": 772, + "AGAAAA": 773, + "AGAAAT": 774, + "AGAAAC": 775, + "AGAAAG": 776, + "AGAATA": 777, + "AGAATT": 778, + "AGAATC": 779, + "AGAATG": 780, + "AGAACA": 781, + "AGAACT": 782, + "AGAACC": 783, + "AGAACG": 784, + "AGAAGA": 785, + "AGAAGT": 786, + "AGAAGC": 787, + "AGAAGG": 788, + "AGATAA": 789, + "AGATAT": 790, + "AGATAC": 791, + "AGATAG": 792, + "AGATTA": 793, + "AGATTT": 794, + "AGATTC": 795, + "AGATTG": 796, + "AGATCA": 797, + "AGATCT": 798, + "AGATCC": 799, + "AGATCG": 800, + "AGATGA": 801, + "AGATGT": 802, + "AGATGC": 803, + "AGATGG": 804, + "AGACAA": 805, + "AGACAT": 806, + "AGACAC": 807, + "AGACAG": 808, + "AGACTA": 809, + "AGACTT": 810, + "AGACTC": 811, + "AGACTG": 812, + "AGACCA": 813, + "AGACCT": 814, + "AGACCC": 815, + "AGACCG": 816, + "AGACGA": 817, + "AGACGT": 818, + "AGACGC": 819, + "AGACGG": 820, + "AGAGAA": 821, + "AGAGAT": 822, + "AGAGAC": 823, + "AGAGAG": 824, + "AGAGTA": 825, + "AGAGTT": 826, + "AGAGTC": 827, + "AGAGTG": 828, + "AGAGCA": 829, + "AGAGCT": 830, + "AGAGCC": 831, + "AGAGCG": 832, + "AGAGGA": 833, + "AGAGGT": 834, + "AGAGGC": 835, + "AGAGGG": 836, + "AGTAAA": 837, + "AGTAAT": 838, + "AGTAAC": 839, + "AGTAAG": 840, + "AGTATA": 841, + "AGTATT": 842, + "AGTATC": 843, + "AGTATG": 844, + "AGTACA": 845, + "AGTACT": 846, + "AGTACC": 847, + "AGTACG": 848, + "AGTAGA": 849, + "AGTAGT": 850, + "AGTAGC": 851, + "AGTAGG": 852, + "AGTTAA": 853, + "AGTTAT": 854, + "AGTTAC": 855, + "AGTTAG": 856, + "AGTTTA": 857, + "AGTTTT": 858, + "AGTTTC": 859, + "AGTTTG": 860, + "AGTTCA": 861, + "AGTTCT": 862, + "AGTTCC": 863, + "AGTTCG": 864, + "AGTTGA": 865, + "AGTTGT": 866, + "AGTTGC": 867, + "AGTTGG": 868, + "AGTCAA": 869, + "AGTCAT": 870, + "AGTCAC": 871, + "AGTCAG": 872, + "AGTCTA": 873, + "AGTCTT": 874, + "AGTCTC": 875, + "AGTCTG": 876, + "AGTCCA": 877, + "AGTCCT": 878, + "AGTCCC": 879, + "AGTCCG": 880, + "AGTCGA": 881, + "AGTCGT": 882, + "AGTCGC": 883, + "AGTCGG": 884, + "AGTGAA": 885, + "AGTGAT": 886, + "AGTGAC": 887, + "AGTGAG": 888, + "AGTGTA": 889, + "AGTGTT": 890, + "AGTGTC": 891, + "AGTGTG": 892, + "AGTGCA": 893, + "AGTGCT": 894, + "AGTGCC": 895, + "AGTGCG": 896, + "AGTGGA": 897, + "AGTGGT": 898, + "AGTGGC": 899, + "AGTGGG": 900, + "AGCAAA": 901, + "AGCAAT": 902, + "AGCAAC": 903, + "AGCAAG": 904, + "AGCATA": 905, + "AGCATT": 906, + "AGCATC": 907, + "AGCATG": 908, + "AGCACA": 909, + "AGCACT": 910, + "AGCACC": 911, + "AGCACG": 912, + "AGCAGA": 913, + "AGCAGT": 914, + "AGCAGC": 915, + "AGCAGG": 916, + "AGCTAA": 917, + "AGCTAT": 918, + "AGCTAC": 919, + "AGCTAG": 920, + "AGCTTA": 921, + "AGCTTT": 922, + "AGCTTC": 923, + "AGCTTG": 924, + "AGCTCA": 925, + "AGCTCT": 926, + "AGCTCC": 927, + "AGCTCG": 928, + "AGCTGA": 929, + "AGCTGT": 930, + "AGCTGC": 931, + "AGCTGG": 932, + "AGCCAA": 933, + "AGCCAT": 934, + "AGCCAC": 935, + "AGCCAG": 936, + "AGCCTA": 937, + "AGCCTT": 938, + "AGCCTC": 939, + "AGCCTG": 940, + "AGCCCA": 941, + "AGCCCT": 942, + "AGCCCC": 943, + "AGCCCG": 944, + "AGCCGA": 945, + "AGCCGT": 946, + "AGCCGC": 947, + "AGCCGG": 948, + "AGCGAA": 949, + "AGCGAT": 950, + "AGCGAC": 951, + "AGCGAG": 952, + "AGCGTA": 953, + "AGCGTT": 954, + "AGCGTC": 955, + "AGCGTG": 956, + "AGCGCA": 957, + "AGCGCT": 958, + "AGCGCC": 959, + "AGCGCG": 960, + "AGCGGA": 961, + "AGCGGT": 962, + "AGCGGC": 963, + "AGCGGG": 964, + "AGGAAA": 965, + "AGGAAT": 966, + "AGGAAC": 967, + "AGGAAG": 968, + "AGGATA": 969, + "AGGATT": 970, + "AGGATC": 971, + "AGGATG": 972, + "AGGACA": 973, + "AGGACT": 974, + "AGGACC": 975, + "AGGACG": 976, + "AGGAGA": 977, + "AGGAGT": 978, + "AGGAGC": 979, + "AGGAGG": 980, + "AGGTAA": 981, + "AGGTAT": 982, + "AGGTAC": 983, + "AGGTAG": 984, + "AGGTTA": 985, + "AGGTTT": 986, + "AGGTTC": 987, + "AGGTTG": 988, + "AGGTCA": 989, + "AGGTCT": 990, + "AGGTCC": 991, + "AGGTCG": 992, + "AGGTGA": 993, + "AGGTGT": 994, + "AGGTGC": 995, + "AGGTGG": 996, + "AGGCAA": 997, + "AGGCAT": 998, + "AGGCAC": 999, + "AGGCAG": 1000, + "AGGCTA": 1001, + "AGGCTT": 1002, + "AGGCTC": 1003, + "AGGCTG": 1004, + "AGGCCA": 1005, + "AGGCCT": 1006, + "AGGCCC": 1007, + "AGGCCG": 1008, + "AGGCGA": 1009, + "AGGCGT": 1010, + "AGGCGC": 1011, + "AGGCGG": 1012, + "AGGGAA": 1013, + "AGGGAT": 1014, + "AGGGAC": 1015, + "AGGGAG": 1016, + "AGGGTA": 1017, + "AGGGTT": 1018, + "AGGGTC": 1019, + "AGGGTG": 1020, + "AGGGCA": 1021, + "AGGGCT": 1022, + "AGGGCC": 1023, + "AGGGCG": 1024, + "AGGGGA": 1025, + "AGGGGT": 1026, + "AGGGGC": 1027, + "AGGGGG": 1028, + "TAAAAA": 1029, + "TAAAAT": 1030, + "TAAAAC": 1031, + "TAAAAG": 1032, + "TAAATA": 1033, + "TAAATT": 1034, + "TAAATC": 1035, + "TAAATG": 1036, + "TAAACA": 1037, + "TAAACT": 1038, + "TAAACC": 1039, + "TAAACG": 1040, + "TAAAGA": 1041, + "TAAAGT": 1042, + "TAAAGC": 1043, + "TAAAGG": 1044, + "TAATAA": 1045, + "TAATAT": 1046, + "TAATAC": 1047, + "TAATAG": 1048, + "TAATTA": 1049, + "TAATTT": 1050, + "TAATTC": 1051, + "TAATTG": 1052, + "TAATCA": 1053, + "TAATCT": 1054, + "TAATCC": 1055, + "TAATCG": 1056, + "TAATGA": 1057, + "TAATGT": 1058, + "TAATGC": 1059, + "TAATGG": 1060, + "TAACAA": 1061, + "TAACAT": 1062, + "TAACAC": 1063, + "TAACAG": 1064, + "TAACTA": 1065, + "TAACTT": 1066, + "TAACTC": 1067, + "TAACTG": 1068, + "TAACCA": 1069, + "TAACCT": 1070, + "TAACCC": 1071, + "TAACCG": 1072, + "TAACGA": 1073, + "TAACGT": 1074, + "TAACGC": 1075, + "TAACGG": 1076, + "TAAGAA": 1077, + "TAAGAT": 1078, + "TAAGAC": 1079, + "TAAGAG": 1080, + "TAAGTA": 1081, + "TAAGTT": 1082, + "TAAGTC": 1083, + "TAAGTG": 1084, + "TAAGCA": 1085, + "TAAGCT": 1086, + "TAAGCC": 1087, + "TAAGCG": 1088, + "TAAGGA": 1089, + "TAAGGT": 1090, + "TAAGGC": 1091, + "TAAGGG": 1092, + "TATAAA": 1093, + "TATAAT": 1094, + "TATAAC": 1095, + "TATAAG": 1096, + "TATATA": 1097, + "TATATT": 1098, + "TATATC": 1099, + "TATATG": 1100, + "TATACA": 1101, + "TATACT": 1102, + "TATACC": 1103, + "TATACG": 1104, + "TATAGA": 1105, + "TATAGT": 1106, + "TATAGC": 1107, + "TATAGG": 1108, + "TATTAA": 1109, + "TATTAT": 1110, + "TATTAC": 1111, + "TATTAG": 1112, + "TATTTA": 1113, + "TATTTT": 1114, + "TATTTC": 1115, + "TATTTG": 1116, + "TATTCA": 1117, + "TATTCT": 1118, + "TATTCC": 1119, + "TATTCG": 1120, + "TATTGA": 1121, + "TATTGT": 1122, + "TATTGC": 1123, + "TATTGG": 1124, + "TATCAA": 1125, + "TATCAT": 1126, + "TATCAC": 1127, + "TATCAG": 1128, + "TATCTA": 1129, + "TATCTT": 1130, + "TATCTC": 1131, + "TATCTG": 1132, + "TATCCA": 1133, + "TATCCT": 1134, + "TATCCC": 1135, + "TATCCG": 1136, + "TATCGA": 1137, + "TATCGT": 1138, + "TATCGC": 1139, + "TATCGG": 1140, + "TATGAA": 1141, + "TATGAT": 1142, + "TATGAC": 1143, + "TATGAG": 1144, + "TATGTA": 1145, + "TATGTT": 1146, + "TATGTC": 1147, + "TATGTG": 1148, + "TATGCA": 1149, + "TATGCT": 1150, + "TATGCC": 1151, + "TATGCG": 1152, + "TATGGA": 1153, + "TATGGT": 1154, + "TATGGC": 1155, + "TATGGG": 1156, + "TACAAA": 1157, + "TACAAT": 1158, + "TACAAC": 1159, + "TACAAG": 1160, + "TACATA": 1161, + "TACATT": 1162, + "TACATC": 1163, + "TACATG": 1164, + "TACACA": 1165, + "TACACT": 1166, + "TACACC": 1167, + "TACACG": 1168, + "TACAGA": 1169, + "TACAGT": 1170, + "TACAGC": 1171, + "TACAGG": 1172, + "TACTAA": 1173, + "TACTAT": 1174, + "TACTAC": 1175, + "TACTAG": 1176, + "TACTTA": 1177, + "TACTTT": 1178, + "TACTTC": 1179, + "TACTTG": 1180, + "TACTCA": 1181, + "TACTCT": 1182, + "TACTCC": 1183, + "TACTCG": 1184, + "TACTGA": 1185, + "TACTGT": 1186, + "TACTGC": 1187, + "TACTGG": 1188, + "TACCAA": 1189, + "TACCAT": 1190, + "TACCAC": 1191, + "TACCAG": 1192, + "TACCTA": 1193, + "TACCTT": 1194, + "TACCTC": 1195, + "TACCTG": 1196, + "TACCCA": 1197, + "TACCCT": 1198, + "TACCCC": 1199, + "TACCCG": 1200, + "TACCGA": 1201, + "TACCGT": 1202, + "TACCGC": 1203, + "TACCGG": 1204, + "TACGAA": 1205, + "TACGAT": 1206, + "TACGAC": 1207, + "TACGAG": 1208, + "TACGTA": 1209, + "TACGTT": 1210, + "TACGTC": 1211, + "TACGTG": 1212, + "TACGCA": 1213, + "TACGCT": 1214, + "TACGCC": 1215, + "TACGCG": 1216, + "TACGGA": 1217, + "TACGGT": 1218, + "TACGGC": 1219, + "TACGGG": 1220, + "TAGAAA": 1221, + "TAGAAT": 1222, + "TAGAAC": 1223, + "TAGAAG": 1224, + "TAGATA": 1225, + "TAGATT": 1226, + "TAGATC": 1227, + "TAGATG": 1228, + "TAGACA": 1229, + "TAGACT": 1230, + "TAGACC": 1231, + "TAGACG": 1232, + "TAGAGA": 1233, + "TAGAGT": 1234, + "TAGAGC": 1235, + "TAGAGG": 1236, + "TAGTAA": 1237, + "TAGTAT": 1238, + "TAGTAC": 1239, + "TAGTAG": 1240, + "TAGTTA": 1241, + "TAGTTT": 1242, + "TAGTTC": 1243, + "TAGTTG": 1244, + "TAGTCA": 1245, + "TAGTCT": 1246, + "TAGTCC": 1247, + "TAGTCG": 1248, + "TAGTGA": 1249, + "TAGTGT": 1250, + "TAGTGC": 1251, + "TAGTGG": 1252, + "TAGCAA": 1253, + "TAGCAT": 1254, + "TAGCAC": 1255, + "TAGCAG": 1256, + "TAGCTA": 1257, + "TAGCTT": 1258, + "TAGCTC": 1259, + "TAGCTG": 1260, + "TAGCCA": 1261, + "TAGCCT": 1262, + "TAGCCC": 1263, + "TAGCCG": 1264, + "TAGCGA": 1265, + "TAGCGT": 1266, + "TAGCGC": 1267, + "TAGCGG": 1268, + "TAGGAA": 1269, + "TAGGAT": 1270, + "TAGGAC": 1271, + "TAGGAG": 1272, + "TAGGTA": 1273, + "TAGGTT": 1274, + "TAGGTC": 1275, + "TAGGTG": 1276, + "TAGGCA": 1277, + "TAGGCT": 1278, + "TAGGCC": 1279, + "TAGGCG": 1280, + "TAGGGA": 1281, + "TAGGGT": 1282, + "TAGGGC": 1283, + "TAGGGG": 1284, + "TTAAAA": 1285, + "TTAAAT": 1286, + "TTAAAC": 1287, + "TTAAAG": 1288, + "TTAATA": 1289, + "TTAATT": 1290, + "TTAATC": 1291, + "TTAATG": 1292, + "TTAACA": 1293, + "TTAACT": 1294, + "TTAACC": 1295, + "TTAACG": 1296, + "TTAAGA": 1297, + "TTAAGT": 1298, + "TTAAGC": 1299, + "TTAAGG": 1300, + "TTATAA": 1301, + "TTATAT": 1302, + "TTATAC": 1303, + "TTATAG": 1304, + "TTATTA": 1305, + "TTATTT": 1306, + "TTATTC": 1307, + "TTATTG": 1308, + "TTATCA": 1309, + "TTATCT": 1310, + "TTATCC": 1311, + "TTATCG": 1312, + "TTATGA": 1313, + "TTATGT": 1314, + "TTATGC": 1315, + "TTATGG": 1316, + "TTACAA": 1317, + "TTACAT": 1318, + "TTACAC": 1319, + "TTACAG": 1320, + "TTACTA": 1321, + "TTACTT": 1322, + "TTACTC": 1323, + "TTACTG": 1324, + "TTACCA": 1325, + "TTACCT": 1326, + "TTACCC": 1327, + "TTACCG": 1328, + "TTACGA": 1329, + "TTACGT": 1330, + "TTACGC": 1331, + "TTACGG": 1332, + "TTAGAA": 1333, + "TTAGAT": 1334, + "TTAGAC": 1335, + "TTAGAG": 1336, + "TTAGTA": 1337, + "TTAGTT": 1338, + "TTAGTC": 1339, + "TTAGTG": 1340, + "TTAGCA": 1341, + "TTAGCT": 1342, + "TTAGCC": 1343, + "TTAGCG": 1344, + "TTAGGA": 1345, + "TTAGGT": 1346, + "TTAGGC": 1347, + "TTAGGG": 1348, + "TTTAAA": 1349, + "TTTAAT": 1350, + "TTTAAC": 1351, + "TTTAAG": 1352, + "TTTATA": 1353, + "TTTATT": 1354, + "TTTATC": 1355, + "TTTATG": 1356, + "TTTACA": 1357, + "TTTACT": 1358, + "TTTACC": 1359, + "TTTACG": 1360, + "TTTAGA": 1361, + "TTTAGT": 1362, + "TTTAGC": 1363, + "TTTAGG": 1364, + "TTTTAA": 1365, + "TTTTAT": 1366, + "TTTTAC": 1367, + "TTTTAG": 1368, + "TTTTTA": 1369, + "TTTTTT": 1370, + "TTTTTC": 1371, + "TTTTTG": 1372, + "TTTTCA": 1373, + "TTTTCT": 1374, + "TTTTCC": 1375, + "TTTTCG": 1376, + "TTTTGA": 1377, + "TTTTGT": 1378, + "TTTTGC": 1379, + "TTTTGG": 1380, + "TTTCAA": 1381, + "TTTCAT": 1382, + "TTTCAC": 1383, + "TTTCAG": 1384, + "TTTCTA": 1385, + "TTTCTT": 1386, + "TTTCTC": 1387, + "TTTCTG": 1388, + "TTTCCA": 1389, + "TTTCCT": 1390, + "TTTCCC": 1391, + "TTTCCG": 1392, + "TTTCGA": 1393, + "TTTCGT": 1394, + "TTTCGC": 1395, + "TTTCGG": 1396, + "TTTGAA": 1397, + "TTTGAT": 1398, + "TTTGAC": 1399, + "TTTGAG": 1400, + "TTTGTA": 1401, + "TTTGTT": 1402, + "TTTGTC": 1403, + "TTTGTG": 1404, + "TTTGCA": 1405, + "TTTGCT": 1406, + "TTTGCC": 1407, + "TTTGCG": 1408, + "TTTGGA": 1409, + "TTTGGT": 1410, + "TTTGGC": 1411, + "TTTGGG": 1412, + "TTCAAA": 1413, + "TTCAAT": 1414, + "TTCAAC": 1415, + "TTCAAG": 1416, + "TTCATA": 1417, + "TTCATT": 1418, + "TTCATC": 1419, + "TTCATG": 1420, + "TTCACA": 1421, + "TTCACT": 1422, + "TTCACC": 1423, + "TTCACG": 1424, + "TTCAGA": 1425, + "TTCAGT": 1426, + "TTCAGC": 1427, + "TTCAGG": 1428, + "TTCTAA": 1429, + "TTCTAT": 1430, + "TTCTAC": 1431, + "TTCTAG": 1432, + "TTCTTA": 1433, + "TTCTTT": 1434, + "TTCTTC": 1435, + "TTCTTG": 1436, + "TTCTCA": 1437, + "TTCTCT": 1438, + "TTCTCC": 1439, + "TTCTCG": 1440, + "TTCTGA": 1441, + "TTCTGT": 1442, + "TTCTGC": 1443, + "TTCTGG": 1444, + "TTCCAA": 1445, + "TTCCAT": 1446, + "TTCCAC": 1447, + "TTCCAG": 1448, + "TTCCTA": 1449, + "TTCCTT": 1450, + "TTCCTC": 1451, + "TTCCTG": 1452, + "TTCCCA": 1453, + "TTCCCT": 1454, + "TTCCCC": 1455, + "TTCCCG": 1456, + "TTCCGA": 1457, + "TTCCGT": 1458, + "TTCCGC": 1459, + "TTCCGG": 1460, + "TTCGAA": 1461, + "TTCGAT": 1462, + "TTCGAC": 1463, + "TTCGAG": 1464, + "TTCGTA": 1465, + "TTCGTT": 1466, + "TTCGTC": 1467, + "TTCGTG": 1468, + "TTCGCA": 1469, + "TTCGCT": 1470, + "TTCGCC": 1471, + "TTCGCG": 1472, + "TTCGGA": 1473, + "TTCGGT": 1474, + "TTCGGC": 1475, + "TTCGGG": 1476, + "TTGAAA": 1477, + "TTGAAT": 1478, + "TTGAAC": 1479, + "TTGAAG": 1480, + "TTGATA": 1481, + "TTGATT": 1482, + "TTGATC": 1483, + "TTGATG": 1484, + "TTGACA": 1485, + "TTGACT": 1486, + "TTGACC": 1487, + "TTGACG": 1488, + "TTGAGA": 1489, + "TTGAGT": 1490, + "TTGAGC": 1491, + "TTGAGG": 1492, + "TTGTAA": 1493, + "TTGTAT": 1494, + "TTGTAC": 1495, + "TTGTAG": 1496, + "TTGTTA": 1497, + "TTGTTT": 1498, + "TTGTTC": 1499, + "TTGTTG": 1500, + "TTGTCA": 1501, + "TTGTCT": 1502, + "TTGTCC": 1503, + "TTGTCG": 1504, + "TTGTGA": 1505, + "TTGTGT": 1506, + "TTGTGC": 1507, + "TTGTGG": 1508, + "TTGCAA": 1509, + "TTGCAT": 1510, + "TTGCAC": 1511, + "TTGCAG": 1512, + "TTGCTA": 1513, + "TTGCTT": 1514, + "TTGCTC": 1515, + "TTGCTG": 1516, + "TTGCCA": 1517, + "TTGCCT": 1518, + "TTGCCC": 1519, + "TTGCCG": 1520, + "TTGCGA": 1521, + "TTGCGT": 1522, + "TTGCGC": 1523, + "TTGCGG": 1524, + "TTGGAA": 1525, + "TTGGAT": 1526, + "TTGGAC": 1527, + "TTGGAG": 1528, + "TTGGTA": 1529, + "TTGGTT": 1530, + "TTGGTC": 1531, + "TTGGTG": 1532, + "TTGGCA": 1533, + "TTGGCT": 1534, + "TTGGCC": 1535, + "TTGGCG": 1536, + "TTGGGA": 1537, + "TTGGGT": 1538, + "TTGGGC": 1539, + "TTGGGG": 1540, + "TCAAAA": 1541, + "TCAAAT": 1542, + "TCAAAC": 1543, + "TCAAAG": 1544, + "TCAATA": 1545, + "TCAATT": 1546, + "TCAATC": 1547, + "TCAATG": 1548, + "TCAACA": 1549, + "TCAACT": 1550, + "TCAACC": 1551, + "TCAACG": 1552, + "TCAAGA": 1553, + "TCAAGT": 1554, + "TCAAGC": 1555, + "TCAAGG": 1556, + "TCATAA": 1557, + "TCATAT": 1558, + "TCATAC": 1559, + "TCATAG": 1560, + "TCATTA": 1561, + "TCATTT": 1562, + "TCATTC": 1563, + "TCATTG": 1564, + "TCATCA": 1565, + "TCATCT": 1566, + "TCATCC": 1567, + "TCATCG": 1568, + "TCATGA": 1569, + "TCATGT": 1570, + "TCATGC": 1571, + "TCATGG": 1572, + "TCACAA": 1573, + "TCACAT": 1574, + "TCACAC": 1575, + "TCACAG": 1576, + "TCACTA": 1577, + "TCACTT": 1578, + "TCACTC": 1579, + "TCACTG": 1580, + "TCACCA": 1581, + "TCACCT": 1582, + "TCACCC": 1583, + "TCACCG": 1584, + "TCACGA": 1585, + "TCACGT": 1586, + "TCACGC": 1587, + "TCACGG": 1588, + "TCAGAA": 1589, + "TCAGAT": 1590, + "TCAGAC": 1591, + "TCAGAG": 1592, + "TCAGTA": 1593, + "TCAGTT": 1594, + "TCAGTC": 1595, + "TCAGTG": 1596, + "TCAGCA": 1597, + "TCAGCT": 1598, + "TCAGCC": 1599, + "TCAGCG": 1600, + "TCAGGA": 1601, + "TCAGGT": 1602, + "TCAGGC": 1603, + "TCAGGG": 1604, + "TCTAAA": 1605, + "TCTAAT": 1606, + "TCTAAC": 1607, + "TCTAAG": 1608, + "TCTATA": 1609, + "TCTATT": 1610, + "TCTATC": 1611, + "TCTATG": 1612, + "TCTACA": 1613, + "TCTACT": 1614, + "TCTACC": 1615, + "TCTACG": 1616, + "TCTAGA": 1617, + "TCTAGT": 1618, + "TCTAGC": 1619, + "TCTAGG": 1620, + "TCTTAA": 1621, + "TCTTAT": 1622, + "TCTTAC": 1623, + "TCTTAG": 1624, + "TCTTTA": 1625, + "TCTTTT": 1626, + "TCTTTC": 1627, + "TCTTTG": 1628, + "TCTTCA": 1629, + "TCTTCT": 1630, + "TCTTCC": 1631, + "TCTTCG": 1632, + "TCTTGA": 1633, + "TCTTGT": 1634, + "TCTTGC": 1635, + "TCTTGG": 1636, + "TCTCAA": 1637, + "TCTCAT": 1638, + "TCTCAC": 1639, + "TCTCAG": 1640, + "TCTCTA": 1641, + "TCTCTT": 1642, + "TCTCTC": 1643, + "TCTCTG": 1644, + "TCTCCA": 1645, + "TCTCCT": 1646, + "TCTCCC": 1647, + "TCTCCG": 1648, + "TCTCGA": 1649, + "TCTCGT": 1650, + "TCTCGC": 1651, + "TCTCGG": 1652, + "TCTGAA": 1653, + "TCTGAT": 1654, + "TCTGAC": 1655, + "TCTGAG": 1656, + "TCTGTA": 1657, + "TCTGTT": 1658, + "TCTGTC": 1659, + "TCTGTG": 1660, + "TCTGCA": 1661, + "TCTGCT": 1662, + "TCTGCC": 1663, + "TCTGCG": 1664, + "TCTGGA": 1665, + "TCTGGT": 1666, + "TCTGGC": 1667, + "TCTGGG": 1668, + "TCCAAA": 1669, + "TCCAAT": 1670, + "TCCAAC": 1671, + "TCCAAG": 1672, + "TCCATA": 1673, + "TCCATT": 1674, + "TCCATC": 1675, + "TCCATG": 1676, + "TCCACA": 1677, + "TCCACT": 1678, + "TCCACC": 1679, + "TCCACG": 1680, + "TCCAGA": 1681, + "TCCAGT": 1682, + "TCCAGC": 1683, + "TCCAGG": 1684, + "TCCTAA": 1685, + "TCCTAT": 1686, + "TCCTAC": 1687, + "TCCTAG": 1688, + "TCCTTA": 1689, + "TCCTTT": 1690, + "TCCTTC": 1691, + "TCCTTG": 1692, + "TCCTCA": 1693, + "TCCTCT": 1694, + "TCCTCC": 1695, + "TCCTCG": 1696, + "TCCTGA": 1697, + "TCCTGT": 1698, + "TCCTGC": 1699, + "TCCTGG": 1700, + "TCCCAA": 1701, + "TCCCAT": 1702, + "TCCCAC": 1703, + "TCCCAG": 1704, + "TCCCTA": 1705, + "TCCCTT": 1706, + "TCCCTC": 1707, + "TCCCTG": 1708, + "TCCCCA": 1709, + "TCCCCT": 1710, + "TCCCCC": 1711, + "TCCCCG": 1712, + "TCCCGA": 1713, + "TCCCGT": 1714, + "TCCCGC": 1715, + "TCCCGG": 1716, + "TCCGAA": 1717, + "TCCGAT": 1718, + "TCCGAC": 1719, + "TCCGAG": 1720, + "TCCGTA": 1721, + "TCCGTT": 1722, + "TCCGTC": 1723, + "TCCGTG": 1724, + "TCCGCA": 1725, + "TCCGCT": 1726, + "TCCGCC": 1727, + "TCCGCG": 1728, + "TCCGGA": 1729, + "TCCGGT": 1730, + "TCCGGC": 1731, + "TCCGGG": 1732, + "TCGAAA": 1733, + "TCGAAT": 1734, + "TCGAAC": 1735, + "TCGAAG": 1736, + "TCGATA": 1737, + "TCGATT": 1738, + "TCGATC": 1739, + "TCGATG": 1740, + "TCGACA": 1741, + "TCGACT": 1742, + "TCGACC": 1743, + "TCGACG": 1744, + "TCGAGA": 1745, + "TCGAGT": 1746, + "TCGAGC": 1747, + "TCGAGG": 1748, + "TCGTAA": 1749, + "TCGTAT": 1750, + "TCGTAC": 1751, + "TCGTAG": 1752, + "TCGTTA": 1753, + "TCGTTT": 1754, + "TCGTTC": 1755, + "TCGTTG": 1756, + "TCGTCA": 1757, + "TCGTCT": 1758, + "TCGTCC": 1759, + "TCGTCG": 1760, + "TCGTGA": 1761, + "TCGTGT": 1762, + "TCGTGC": 1763, + "TCGTGG": 1764, + "TCGCAA": 1765, + "TCGCAT": 1766, + "TCGCAC": 1767, + "TCGCAG": 1768, + "TCGCTA": 1769, + "TCGCTT": 1770, + "TCGCTC": 1771, + "TCGCTG": 1772, + "TCGCCA": 1773, + "TCGCCT": 1774, + "TCGCCC": 1775, + "TCGCCG": 1776, + "TCGCGA": 1777, + "TCGCGT": 1778, + "TCGCGC": 1779, + "TCGCGG": 1780, + "TCGGAA": 1781, + "TCGGAT": 1782, + "TCGGAC": 1783, + "TCGGAG": 1784, + "TCGGTA": 1785, + "TCGGTT": 1786, + "TCGGTC": 1787, + "TCGGTG": 1788, + "TCGGCA": 1789, + "TCGGCT": 1790, + "TCGGCC": 1791, + "TCGGCG": 1792, + "TCGGGA": 1793, + "TCGGGT": 1794, + "TCGGGC": 1795, + "TCGGGG": 1796, + "TGAAAA": 1797, + "TGAAAT": 1798, + "TGAAAC": 1799, + "TGAAAG": 1800, + "TGAATA": 1801, + "TGAATT": 1802, + "TGAATC": 1803, + "TGAATG": 1804, + "TGAACA": 1805, + "TGAACT": 1806, + "TGAACC": 1807, + "TGAACG": 1808, + "TGAAGA": 1809, + "TGAAGT": 1810, + "TGAAGC": 1811, + "TGAAGG": 1812, + "TGATAA": 1813, + "TGATAT": 1814, + "TGATAC": 1815, + "TGATAG": 1816, + "TGATTA": 1817, + "TGATTT": 1818, + "TGATTC": 1819, + "TGATTG": 1820, + "TGATCA": 1821, + "TGATCT": 1822, + "TGATCC": 1823, + "TGATCG": 1824, + "TGATGA": 1825, + "TGATGT": 1826, + "TGATGC": 1827, + "TGATGG": 1828, + "TGACAA": 1829, + "TGACAT": 1830, + "TGACAC": 1831, + "TGACAG": 1832, + "TGACTA": 1833, + "TGACTT": 1834, + "TGACTC": 1835, + "TGACTG": 1836, + "TGACCA": 1837, + "TGACCT": 1838, + "TGACCC": 1839, + "TGACCG": 1840, + "TGACGA": 1841, + "TGACGT": 1842, + "TGACGC": 1843, + "TGACGG": 1844, + "TGAGAA": 1845, + "TGAGAT": 1846, + "TGAGAC": 1847, + "TGAGAG": 1848, + "TGAGTA": 1849, + "TGAGTT": 1850, + "TGAGTC": 1851, + "TGAGTG": 1852, + "TGAGCA": 1853, + "TGAGCT": 1854, + "TGAGCC": 1855, + "TGAGCG": 1856, + "TGAGGA": 1857, + "TGAGGT": 1858, + "TGAGGC": 1859, + "TGAGGG": 1860, + "TGTAAA": 1861, + "TGTAAT": 1862, + "TGTAAC": 1863, + "TGTAAG": 1864, + "TGTATA": 1865, + "TGTATT": 1866, + "TGTATC": 1867, + "TGTATG": 1868, + "TGTACA": 1869, + "TGTACT": 1870, + "TGTACC": 1871, + "TGTACG": 1872, + "TGTAGA": 1873, + "TGTAGT": 1874, + "TGTAGC": 1875, + "TGTAGG": 1876, + "TGTTAA": 1877, + "TGTTAT": 1878, + "TGTTAC": 1879, + "TGTTAG": 1880, + "TGTTTA": 1881, + "TGTTTT": 1882, + "TGTTTC": 1883, + "TGTTTG": 1884, + "TGTTCA": 1885, + "TGTTCT": 1886, + "TGTTCC": 1887, + "TGTTCG": 1888, + "TGTTGA": 1889, + "TGTTGT": 1890, + "TGTTGC": 1891, + "TGTTGG": 1892, + "TGTCAA": 1893, + "TGTCAT": 1894, + "TGTCAC": 1895, + "TGTCAG": 1896, + "TGTCTA": 1897, + "TGTCTT": 1898, + "TGTCTC": 1899, + "TGTCTG": 1900, + "TGTCCA": 1901, + "TGTCCT": 1902, + "TGTCCC": 1903, + "TGTCCG": 1904, + "TGTCGA": 1905, + "TGTCGT": 1906, + "TGTCGC": 1907, + "TGTCGG": 1908, + "TGTGAA": 1909, + "TGTGAT": 1910, + "TGTGAC": 1911, + "TGTGAG": 1912, + "TGTGTA": 1913, + "TGTGTT": 1914, + "TGTGTC": 1915, + "TGTGTG": 1916, + "TGTGCA": 1917, + "TGTGCT": 1918, + "TGTGCC": 1919, + "TGTGCG": 1920, + "TGTGGA": 1921, + "TGTGGT": 1922, + "TGTGGC": 1923, + "TGTGGG": 1924, + "TGCAAA": 1925, + "TGCAAT": 1926, + "TGCAAC": 1927, + "TGCAAG": 1928, + "TGCATA": 1929, + "TGCATT": 1930, + "TGCATC": 1931, + "TGCATG": 1932, + "TGCACA": 1933, + "TGCACT": 1934, + "TGCACC": 1935, + "TGCACG": 1936, + "TGCAGA": 1937, + "TGCAGT": 1938, + "TGCAGC": 1939, + "TGCAGG": 1940, + "TGCTAA": 1941, + "TGCTAT": 1942, + "TGCTAC": 1943, + "TGCTAG": 1944, + "TGCTTA": 1945, + "TGCTTT": 1946, + "TGCTTC": 1947, + "TGCTTG": 1948, + "TGCTCA": 1949, + "TGCTCT": 1950, + "TGCTCC": 1951, + "TGCTCG": 1952, + "TGCTGA": 1953, + "TGCTGT": 1954, + "TGCTGC": 1955, + "TGCTGG": 1956, + "TGCCAA": 1957, + "TGCCAT": 1958, + "TGCCAC": 1959, + "TGCCAG": 1960, + "TGCCTA": 1961, + "TGCCTT": 1962, + "TGCCTC": 1963, + "TGCCTG": 1964, + "TGCCCA": 1965, + "TGCCCT": 1966, + "TGCCCC": 1967, + "TGCCCG": 1968, + "TGCCGA": 1969, + "TGCCGT": 1970, + "TGCCGC": 1971, + "TGCCGG": 1972, + "TGCGAA": 1973, + "TGCGAT": 1974, + "TGCGAC": 1975, + "TGCGAG": 1976, + "TGCGTA": 1977, + "TGCGTT": 1978, + "TGCGTC": 1979, + "TGCGTG": 1980, + "TGCGCA": 1981, + "TGCGCT": 1982, + "TGCGCC": 1983, + "TGCGCG": 1984, + "TGCGGA": 1985, + "TGCGGT": 1986, + "TGCGGC": 1987, + "TGCGGG": 1988, + "TGGAAA": 1989, + "TGGAAT": 1990, + "TGGAAC": 1991, + "TGGAAG": 1992, + "TGGATA": 1993, + "TGGATT": 1994, + "TGGATC": 1995, + "TGGATG": 1996, + "TGGACA": 1997, + "TGGACT": 1998, + "TGGACC": 1999, + "TGGACG": 2000, + "TGGAGA": 2001, + "TGGAGT": 2002, + "TGGAGC": 2003, + "TGGAGG": 2004, + "TGGTAA": 2005, + "TGGTAT": 2006, + "TGGTAC": 2007, + "TGGTAG": 2008, + "TGGTTA": 2009, + "TGGTTT": 2010, + "TGGTTC": 2011, + "TGGTTG": 2012, + "TGGTCA": 2013, + "TGGTCT": 2014, + "TGGTCC": 2015, + "TGGTCG": 2016, + "TGGTGA": 2017, + "TGGTGT": 2018, + "TGGTGC": 2019, + "TGGTGG": 2020, + "TGGCAA": 2021, + "TGGCAT": 2022, + "TGGCAC": 2023, + "TGGCAG": 2024, + "TGGCTA": 2025, + "TGGCTT": 2026, + "TGGCTC": 2027, + "TGGCTG": 2028, + "TGGCCA": 2029, + "TGGCCT": 2030, + "TGGCCC": 2031, + "TGGCCG": 2032, + "TGGCGA": 2033, + "TGGCGT": 2034, + "TGGCGC": 2035, + "TGGCGG": 2036, + "TGGGAA": 2037, + "TGGGAT": 2038, + "TGGGAC": 2039, + "TGGGAG": 2040, + "TGGGTA": 2041, + "TGGGTT": 2042, + "TGGGTC": 2043, + "TGGGTG": 2044, + "TGGGCA": 2045, + "TGGGCT": 2046, + "TGGGCC": 2047, + "TGGGCG": 2048, + "TGGGGA": 2049, + "TGGGGT": 2050, + "TGGGGC": 2051, + "TGGGGG": 2052, + "CAAAAA": 2053, + "CAAAAT": 2054, + "CAAAAC": 2055, + "CAAAAG": 2056, + "CAAATA": 2057, + "CAAATT": 2058, + "CAAATC": 2059, + "CAAATG": 2060, + "CAAACA": 2061, + "CAAACT": 2062, + "CAAACC": 2063, + "CAAACG": 2064, + "CAAAGA": 2065, + "CAAAGT": 2066, + "CAAAGC": 2067, + "CAAAGG": 2068, + "CAATAA": 2069, + "CAATAT": 2070, + "CAATAC": 2071, + "CAATAG": 2072, + "CAATTA": 2073, + "CAATTT": 2074, + "CAATTC": 2075, + "CAATTG": 2076, + "CAATCA": 2077, + "CAATCT": 2078, + "CAATCC": 2079, + "CAATCG": 2080, + "CAATGA": 2081, + "CAATGT": 2082, + "CAATGC": 2083, + "CAATGG": 2084, + "CAACAA": 2085, + "CAACAT": 2086, + "CAACAC": 2087, + "CAACAG": 2088, + "CAACTA": 2089, + "CAACTT": 2090, + "CAACTC": 2091, + "CAACTG": 2092, + "CAACCA": 2093, + "CAACCT": 2094, + "CAACCC": 2095, + "CAACCG": 2096, + "CAACGA": 2097, + "CAACGT": 2098, + "CAACGC": 2099, + "CAACGG": 2100, + "CAAGAA": 2101, + "CAAGAT": 2102, + "CAAGAC": 2103, + "CAAGAG": 2104, + "CAAGTA": 2105, + "CAAGTT": 2106, + "CAAGTC": 2107, + "CAAGTG": 2108, + "CAAGCA": 2109, + "CAAGCT": 2110, + "CAAGCC": 2111, + "CAAGCG": 2112, + "CAAGGA": 2113, + "CAAGGT": 2114, + "CAAGGC": 2115, + "CAAGGG": 2116, + "CATAAA": 2117, + "CATAAT": 2118, + "CATAAC": 2119, + "CATAAG": 2120, + "CATATA": 2121, + "CATATT": 2122, + "CATATC": 2123, + "CATATG": 2124, + "CATACA": 2125, + "CATACT": 2126, + "CATACC": 2127, + "CATACG": 2128, + "CATAGA": 2129, + "CATAGT": 2130, + "CATAGC": 2131, + "CATAGG": 2132, + "CATTAA": 2133, + "CATTAT": 2134, + "CATTAC": 2135, + "CATTAG": 2136, + "CATTTA": 2137, + "CATTTT": 2138, + "CATTTC": 2139, + "CATTTG": 2140, + "CATTCA": 2141, + "CATTCT": 2142, + "CATTCC": 2143, + "CATTCG": 2144, + "CATTGA": 2145, + "CATTGT": 2146, + "CATTGC": 2147, + "CATTGG": 2148, + "CATCAA": 2149, + "CATCAT": 2150, + "CATCAC": 2151, + "CATCAG": 2152, + "CATCTA": 2153, + "CATCTT": 2154, + "CATCTC": 2155, + "CATCTG": 2156, + "CATCCA": 2157, + "CATCCT": 2158, + "CATCCC": 2159, + "CATCCG": 2160, + "CATCGA": 2161, + "CATCGT": 2162, + "CATCGC": 2163, + "CATCGG": 2164, + "CATGAA": 2165, + "CATGAT": 2166, + "CATGAC": 2167, + "CATGAG": 2168, + "CATGTA": 2169, + "CATGTT": 2170, + "CATGTC": 2171, + "CATGTG": 2172, + "CATGCA": 2173, + "CATGCT": 2174, + "CATGCC": 2175, + "CATGCG": 2176, + "CATGGA": 2177, + "CATGGT": 2178, + "CATGGC": 2179, + "CATGGG": 2180, + "CACAAA": 2181, + "CACAAT": 2182, + "CACAAC": 2183, + "CACAAG": 2184, + "CACATA": 2185, + "CACATT": 2186, + "CACATC": 2187, + "CACATG": 2188, + "CACACA": 2189, + "CACACT": 2190, + "CACACC": 2191, + "CACACG": 2192, + "CACAGA": 2193, + "CACAGT": 2194, + "CACAGC": 2195, + "CACAGG": 2196, + "CACTAA": 2197, + "CACTAT": 2198, + "CACTAC": 2199, + "CACTAG": 2200, + "CACTTA": 2201, + "CACTTT": 2202, + "CACTTC": 2203, + "CACTTG": 2204, + "CACTCA": 2205, + "CACTCT": 2206, + "CACTCC": 2207, + "CACTCG": 2208, + "CACTGA": 2209, + "CACTGT": 2210, + "CACTGC": 2211, + "CACTGG": 2212, + "CACCAA": 2213, + "CACCAT": 2214, + "CACCAC": 2215, + "CACCAG": 2216, + "CACCTA": 2217, + "CACCTT": 2218, + "CACCTC": 2219, + "CACCTG": 2220, + "CACCCA": 2221, + "CACCCT": 2222, + "CACCCC": 2223, + "CACCCG": 2224, + "CACCGA": 2225, + "CACCGT": 2226, + "CACCGC": 2227, + "CACCGG": 2228, + "CACGAA": 2229, + "CACGAT": 2230, + "CACGAC": 2231, + "CACGAG": 2232, + "CACGTA": 2233, + "CACGTT": 2234, + "CACGTC": 2235, + "CACGTG": 2236, + "CACGCA": 2237, + "CACGCT": 2238, + "CACGCC": 2239, + "CACGCG": 2240, + "CACGGA": 2241, + "CACGGT": 2242, + "CACGGC": 2243, + "CACGGG": 2244, + "CAGAAA": 2245, + "CAGAAT": 2246, + "CAGAAC": 2247, + "CAGAAG": 2248, + "CAGATA": 2249, + "CAGATT": 2250, + "CAGATC": 2251, + "CAGATG": 2252, + "CAGACA": 2253, + "CAGACT": 2254, + "CAGACC": 2255, + "CAGACG": 2256, + "CAGAGA": 2257, + "CAGAGT": 2258, + "CAGAGC": 2259, + "CAGAGG": 2260, + "CAGTAA": 2261, + "CAGTAT": 2262, + "CAGTAC": 2263, + "CAGTAG": 2264, + "CAGTTA": 2265, + "CAGTTT": 2266, + "CAGTTC": 2267, + "CAGTTG": 2268, + "CAGTCA": 2269, + "CAGTCT": 2270, + "CAGTCC": 2271, + "CAGTCG": 2272, + "CAGTGA": 2273, + "CAGTGT": 2274, + "CAGTGC": 2275, + "CAGTGG": 2276, + "CAGCAA": 2277, + "CAGCAT": 2278, + "CAGCAC": 2279, + "CAGCAG": 2280, + "CAGCTA": 2281, + "CAGCTT": 2282, + "CAGCTC": 2283, + "CAGCTG": 2284, + "CAGCCA": 2285, + "CAGCCT": 2286, + "CAGCCC": 2287, + "CAGCCG": 2288, + "CAGCGA": 2289, + "CAGCGT": 2290, + "CAGCGC": 2291, + "CAGCGG": 2292, + "CAGGAA": 2293, + "CAGGAT": 2294, + "CAGGAC": 2295, + "CAGGAG": 2296, + "CAGGTA": 2297, + "CAGGTT": 2298, + "CAGGTC": 2299, + "CAGGTG": 2300, + "CAGGCA": 2301, + "CAGGCT": 2302, + "CAGGCC": 2303, + "CAGGCG": 2304, + "CAGGGA": 2305, + "CAGGGT": 2306, + "CAGGGC": 2307, + "CAGGGG": 2308, + "CTAAAA": 2309, + "CTAAAT": 2310, + "CTAAAC": 2311, + "CTAAAG": 2312, + "CTAATA": 2313, + "CTAATT": 2314, + "CTAATC": 2315, + "CTAATG": 2316, + "CTAACA": 2317, + "CTAACT": 2318, + "CTAACC": 2319, + "CTAACG": 2320, + "CTAAGA": 2321, + "CTAAGT": 2322, + "CTAAGC": 2323, + "CTAAGG": 2324, + "CTATAA": 2325, + "CTATAT": 2326, + "CTATAC": 2327, + "CTATAG": 2328, + "CTATTA": 2329, + "CTATTT": 2330, + "CTATTC": 2331, + "CTATTG": 2332, + "CTATCA": 2333, + "CTATCT": 2334, + "CTATCC": 2335, + "CTATCG": 2336, + "CTATGA": 2337, + "CTATGT": 2338, + "CTATGC": 2339, + "CTATGG": 2340, + "CTACAA": 2341, + "CTACAT": 2342, + "CTACAC": 2343, + "CTACAG": 2344, + "CTACTA": 2345, + "CTACTT": 2346, + "CTACTC": 2347, + "CTACTG": 2348, + "CTACCA": 2349, + "CTACCT": 2350, + "CTACCC": 2351, + "CTACCG": 2352, + "CTACGA": 2353, + "CTACGT": 2354, + "CTACGC": 2355, + "CTACGG": 2356, + "CTAGAA": 2357, + "CTAGAT": 2358, + "CTAGAC": 2359, + "CTAGAG": 2360, + "CTAGTA": 2361, + "CTAGTT": 2362, + "CTAGTC": 2363, + "CTAGTG": 2364, + "CTAGCA": 2365, + "CTAGCT": 2366, + "CTAGCC": 2367, + "CTAGCG": 2368, + "CTAGGA": 2369, + "CTAGGT": 2370, + "CTAGGC": 2371, + "CTAGGG": 2372, + "CTTAAA": 2373, + "CTTAAT": 2374, + "CTTAAC": 2375, + "CTTAAG": 2376, + "CTTATA": 2377, + "CTTATT": 2378, + "CTTATC": 2379, + "CTTATG": 2380, + "CTTACA": 2381, + "CTTACT": 2382, + "CTTACC": 2383, + "CTTACG": 2384, + "CTTAGA": 2385, + "CTTAGT": 2386, + "CTTAGC": 2387, + "CTTAGG": 2388, + "CTTTAA": 2389, + "CTTTAT": 2390, + "CTTTAC": 2391, + "CTTTAG": 2392, + "CTTTTA": 2393, + "CTTTTT": 2394, + "CTTTTC": 2395, + "CTTTTG": 2396, + "CTTTCA": 2397, + "CTTTCT": 2398, + "CTTTCC": 2399, + "CTTTCG": 2400, + "CTTTGA": 2401, + "CTTTGT": 2402, + "CTTTGC": 2403, + "CTTTGG": 2404, + "CTTCAA": 2405, + "CTTCAT": 2406, + "CTTCAC": 2407, + "CTTCAG": 2408, + "CTTCTA": 2409, + "CTTCTT": 2410, + "CTTCTC": 2411, + "CTTCTG": 2412, + "CTTCCA": 2413, + "CTTCCT": 2414, + "CTTCCC": 2415, + "CTTCCG": 2416, + "CTTCGA": 2417, + "CTTCGT": 2418, + "CTTCGC": 2419, + "CTTCGG": 2420, + "CTTGAA": 2421, + "CTTGAT": 2422, + "CTTGAC": 2423, + "CTTGAG": 2424, + "CTTGTA": 2425, + "CTTGTT": 2426, + "CTTGTC": 2427, + "CTTGTG": 2428, + "CTTGCA": 2429, + "CTTGCT": 2430, + "CTTGCC": 2431, + "CTTGCG": 2432, + "CTTGGA": 2433, + "CTTGGT": 2434, + "CTTGGC": 2435, + "CTTGGG": 2436, + "CTCAAA": 2437, + "CTCAAT": 2438, + "CTCAAC": 2439, + "CTCAAG": 2440, + "CTCATA": 2441, + "CTCATT": 2442, + "CTCATC": 2443, + "CTCATG": 2444, + "CTCACA": 2445, + "CTCACT": 2446, + "CTCACC": 2447, + "CTCACG": 2448, + "CTCAGA": 2449, + "CTCAGT": 2450, + "CTCAGC": 2451, + "CTCAGG": 2452, + "CTCTAA": 2453, + "CTCTAT": 2454, + "CTCTAC": 2455, + "CTCTAG": 2456, + "CTCTTA": 2457, + "CTCTTT": 2458, + "CTCTTC": 2459, + "CTCTTG": 2460, + "CTCTCA": 2461, + "CTCTCT": 2462, + "CTCTCC": 2463, + "CTCTCG": 2464, + "CTCTGA": 2465, + "CTCTGT": 2466, + "CTCTGC": 2467, + "CTCTGG": 2468, + "CTCCAA": 2469, + "CTCCAT": 2470, + "CTCCAC": 2471, + "CTCCAG": 2472, + "CTCCTA": 2473, + "CTCCTT": 2474, + "CTCCTC": 2475, + "CTCCTG": 2476, + "CTCCCA": 2477, + "CTCCCT": 2478, + "CTCCCC": 2479, + "CTCCCG": 2480, + "CTCCGA": 2481, + "CTCCGT": 2482, + "CTCCGC": 2483, + "CTCCGG": 2484, + "CTCGAA": 2485, + "CTCGAT": 2486, + "CTCGAC": 2487, + "CTCGAG": 2488, + "CTCGTA": 2489, + "CTCGTT": 2490, + "CTCGTC": 2491, + "CTCGTG": 2492, + "CTCGCA": 2493, + "CTCGCT": 2494, + "CTCGCC": 2495, + "CTCGCG": 2496, + "CTCGGA": 2497, + "CTCGGT": 2498, + "CTCGGC": 2499, + "CTCGGG": 2500, + "CTGAAA": 2501, + "CTGAAT": 2502, + "CTGAAC": 2503, + "CTGAAG": 2504, + "CTGATA": 2505, + "CTGATT": 2506, + "CTGATC": 2507, + "CTGATG": 2508, + "CTGACA": 2509, + "CTGACT": 2510, + "CTGACC": 2511, + "CTGACG": 2512, + "CTGAGA": 2513, + "CTGAGT": 2514, + "CTGAGC": 2515, + "CTGAGG": 2516, + "CTGTAA": 2517, + "CTGTAT": 2518, + "CTGTAC": 2519, + "CTGTAG": 2520, + "CTGTTA": 2521, + "CTGTTT": 2522, + "CTGTTC": 2523, + "CTGTTG": 2524, + "CTGTCA": 2525, + "CTGTCT": 2526, + "CTGTCC": 2527, + "CTGTCG": 2528, + "CTGTGA": 2529, + "CTGTGT": 2530, + "CTGTGC": 2531, + "CTGTGG": 2532, + "CTGCAA": 2533, + "CTGCAT": 2534, + "CTGCAC": 2535, + "CTGCAG": 2536, + "CTGCTA": 2537, + "CTGCTT": 2538, + "CTGCTC": 2539, + "CTGCTG": 2540, + "CTGCCA": 2541, + "CTGCCT": 2542, + "CTGCCC": 2543, + "CTGCCG": 2544, + "CTGCGA": 2545, + "CTGCGT": 2546, + "CTGCGC": 2547, + "CTGCGG": 2548, + "CTGGAA": 2549, + "CTGGAT": 2550, + "CTGGAC": 2551, + "CTGGAG": 2552, + "CTGGTA": 2553, + "CTGGTT": 2554, + "CTGGTC": 2555, + "CTGGTG": 2556, + "CTGGCA": 2557, + "CTGGCT": 2558, + "CTGGCC": 2559, + "CTGGCG": 2560, + "CTGGGA": 2561, + "CTGGGT": 2562, + "CTGGGC": 2563, + "CTGGGG": 2564, + "CCAAAA": 2565, + "CCAAAT": 2566, + "CCAAAC": 2567, + "CCAAAG": 2568, + "CCAATA": 2569, + "CCAATT": 2570, + "CCAATC": 2571, + "CCAATG": 2572, + "CCAACA": 2573, + "CCAACT": 2574, + "CCAACC": 2575, + "CCAACG": 2576, + "CCAAGA": 2577, + "CCAAGT": 2578, + "CCAAGC": 2579, + "CCAAGG": 2580, + "CCATAA": 2581, + "CCATAT": 2582, + "CCATAC": 2583, + "CCATAG": 2584, + "CCATTA": 2585, + "CCATTT": 2586, + "CCATTC": 2587, + "CCATTG": 2588, + "CCATCA": 2589, + "CCATCT": 2590, + "CCATCC": 2591, + "CCATCG": 2592, + "CCATGA": 2593, + "CCATGT": 2594, + "CCATGC": 2595, + "CCATGG": 2596, + "CCACAA": 2597, + "CCACAT": 2598, + "CCACAC": 2599, + "CCACAG": 2600, + "CCACTA": 2601, + "CCACTT": 2602, + "CCACTC": 2603, + "CCACTG": 2604, + "CCACCA": 2605, + "CCACCT": 2606, + "CCACCC": 2607, + "CCACCG": 2608, + "CCACGA": 2609, + "CCACGT": 2610, + "CCACGC": 2611, + "CCACGG": 2612, + "CCAGAA": 2613, + "CCAGAT": 2614, + "CCAGAC": 2615, + "CCAGAG": 2616, + "CCAGTA": 2617, + "CCAGTT": 2618, + "CCAGTC": 2619, + "CCAGTG": 2620, + "CCAGCA": 2621, + "CCAGCT": 2622, + "CCAGCC": 2623, + "CCAGCG": 2624, + "CCAGGA": 2625, + "CCAGGT": 2626, + "CCAGGC": 2627, + "CCAGGG": 2628, + "CCTAAA": 2629, + "CCTAAT": 2630, + "CCTAAC": 2631, + "CCTAAG": 2632, + "CCTATA": 2633, + "CCTATT": 2634, + "CCTATC": 2635, + "CCTATG": 2636, + "CCTACA": 2637, + "CCTACT": 2638, + "CCTACC": 2639, + "CCTACG": 2640, + "CCTAGA": 2641, + "CCTAGT": 2642, + "CCTAGC": 2643, + "CCTAGG": 2644, + "CCTTAA": 2645, + "CCTTAT": 2646, + "CCTTAC": 2647, + "CCTTAG": 2648, + "CCTTTA": 2649, + "CCTTTT": 2650, + "CCTTTC": 2651, + "CCTTTG": 2652, + "CCTTCA": 2653, + "CCTTCT": 2654, + "CCTTCC": 2655, + "CCTTCG": 2656, + "CCTTGA": 2657, + "CCTTGT": 2658, + "CCTTGC": 2659, + "CCTTGG": 2660, + "CCTCAA": 2661, + "CCTCAT": 2662, + "CCTCAC": 2663, + "CCTCAG": 2664, + "CCTCTA": 2665, + "CCTCTT": 2666, + "CCTCTC": 2667, + "CCTCTG": 2668, + "CCTCCA": 2669, + "CCTCCT": 2670, + "CCTCCC": 2671, + "CCTCCG": 2672, + "CCTCGA": 2673, + "CCTCGT": 2674, + "CCTCGC": 2675, + "CCTCGG": 2676, + "CCTGAA": 2677, + "CCTGAT": 2678, + "CCTGAC": 2679, + "CCTGAG": 2680, + "CCTGTA": 2681, + "CCTGTT": 2682, + "CCTGTC": 2683, + "CCTGTG": 2684, + "CCTGCA": 2685, + "CCTGCT": 2686, + "CCTGCC": 2687, + "CCTGCG": 2688, + "CCTGGA": 2689, + "CCTGGT": 2690, + "CCTGGC": 2691, + "CCTGGG": 2692, + "CCCAAA": 2693, + "CCCAAT": 2694, + "CCCAAC": 2695, + "CCCAAG": 2696, + "CCCATA": 2697, + "CCCATT": 2698, + "CCCATC": 2699, + "CCCATG": 2700, + "CCCACA": 2701, + "CCCACT": 2702, + "CCCACC": 2703, + "CCCACG": 2704, + "CCCAGA": 2705, + "CCCAGT": 2706, + "CCCAGC": 2707, + "CCCAGG": 2708, + "CCCTAA": 2709, + "CCCTAT": 2710, + "CCCTAC": 2711, + "CCCTAG": 2712, + "CCCTTA": 2713, + "CCCTTT": 2714, + "CCCTTC": 2715, + "CCCTTG": 2716, + "CCCTCA": 2717, + "CCCTCT": 2718, + "CCCTCC": 2719, + "CCCTCG": 2720, + "CCCTGA": 2721, + "CCCTGT": 2722, + "CCCTGC": 2723, + "CCCTGG": 2724, + "CCCCAA": 2725, + "CCCCAT": 2726, + "CCCCAC": 2727, + "CCCCAG": 2728, + "CCCCTA": 2729, + "CCCCTT": 2730, + "CCCCTC": 2731, + "CCCCTG": 2732, + "CCCCCA": 2733, + "CCCCCT": 2734, + "CCCCCC": 2735, + "CCCCCG": 2736, + "CCCCGA": 2737, + "CCCCGT": 2738, + "CCCCGC": 2739, + "CCCCGG": 2740, + "CCCGAA": 2741, + "CCCGAT": 2742, + "CCCGAC": 2743, + "CCCGAG": 2744, + "CCCGTA": 2745, + "CCCGTT": 2746, + "CCCGTC": 2747, + "CCCGTG": 2748, + "CCCGCA": 2749, + "CCCGCT": 2750, + "CCCGCC": 2751, + "CCCGCG": 2752, + "CCCGGA": 2753, + "CCCGGT": 2754, + "CCCGGC": 2755, + "CCCGGG": 2756, + "CCGAAA": 2757, + "CCGAAT": 2758, + "CCGAAC": 2759, + "CCGAAG": 2760, + "CCGATA": 2761, + "CCGATT": 2762, + "CCGATC": 2763, + "CCGATG": 2764, + "CCGACA": 2765, + "CCGACT": 2766, + "CCGACC": 2767, + "CCGACG": 2768, + "CCGAGA": 2769, + "CCGAGT": 2770, + "CCGAGC": 2771, + "CCGAGG": 2772, + "CCGTAA": 2773, + "CCGTAT": 2774, + "CCGTAC": 2775, + "CCGTAG": 2776, + "CCGTTA": 2777, + "CCGTTT": 2778, + "CCGTTC": 2779, + "CCGTTG": 2780, + "CCGTCA": 2781, + "CCGTCT": 2782, + "CCGTCC": 2783, + "CCGTCG": 2784, + "CCGTGA": 2785, + "CCGTGT": 2786, + "CCGTGC": 2787, + "CCGTGG": 2788, + "CCGCAA": 2789, + "CCGCAT": 2790, + "CCGCAC": 2791, + "CCGCAG": 2792, + "CCGCTA": 2793, + "CCGCTT": 2794, + "CCGCTC": 2795, + "CCGCTG": 2796, + "CCGCCA": 2797, + "CCGCCT": 2798, + "CCGCCC": 2799, + "CCGCCG": 2800, + "CCGCGA": 2801, + "CCGCGT": 2802, + "CCGCGC": 2803, + "CCGCGG": 2804, + "CCGGAA": 2805, + "CCGGAT": 2806, + "CCGGAC": 2807, + "CCGGAG": 2808, + "CCGGTA": 2809, + "CCGGTT": 2810, + "CCGGTC": 2811, + "CCGGTG": 2812, + "CCGGCA": 2813, + "CCGGCT": 2814, + "CCGGCC": 2815, + "CCGGCG": 2816, + "CCGGGA": 2817, + "CCGGGT": 2818, + "CCGGGC": 2819, + "CCGGGG": 2820, + "CGAAAA": 2821, + "CGAAAT": 2822, + "CGAAAC": 2823, + "CGAAAG": 2824, + "CGAATA": 2825, + "CGAATT": 2826, + "CGAATC": 2827, + "CGAATG": 2828, + "CGAACA": 2829, + "CGAACT": 2830, + "CGAACC": 2831, + "CGAACG": 2832, + "CGAAGA": 2833, + "CGAAGT": 2834, + "CGAAGC": 2835, + "CGAAGG": 2836, + "CGATAA": 2837, + "CGATAT": 2838, + "CGATAC": 2839, + "CGATAG": 2840, + "CGATTA": 2841, + "CGATTT": 2842, + "CGATTC": 2843, + "CGATTG": 2844, + "CGATCA": 2845, + "CGATCT": 2846, + "CGATCC": 2847, + "CGATCG": 2848, + "CGATGA": 2849, + "CGATGT": 2850, + "CGATGC": 2851, + "CGATGG": 2852, + "CGACAA": 2853, + "CGACAT": 2854, + "CGACAC": 2855, + "CGACAG": 2856, + "CGACTA": 2857, + "CGACTT": 2858, + "CGACTC": 2859, + "CGACTG": 2860, + "CGACCA": 2861, + "CGACCT": 2862, + "CGACCC": 2863, + "CGACCG": 2864, + "CGACGA": 2865, + "CGACGT": 2866, + "CGACGC": 2867, + "CGACGG": 2868, + "CGAGAA": 2869, + "CGAGAT": 2870, + "CGAGAC": 2871, + "CGAGAG": 2872, + "CGAGTA": 2873, + "CGAGTT": 2874, + "CGAGTC": 2875, + "CGAGTG": 2876, + "CGAGCA": 2877, + "CGAGCT": 2878, + "CGAGCC": 2879, + "CGAGCG": 2880, + "CGAGGA": 2881, + "CGAGGT": 2882, + "CGAGGC": 2883, + "CGAGGG": 2884, + "CGTAAA": 2885, + "CGTAAT": 2886, + "CGTAAC": 2887, + "CGTAAG": 2888, + "CGTATA": 2889, + "CGTATT": 2890, + "CGTATC": 2891, + "CGTATG": 2892, + "CGTACA": 2893, + "CGTACT": 2894, + "CGTACC": 2895, + "CGTACG": 2896, + "CGTAGA": 2897, + "CGTAGT": 2898, + "CGTAGC": 2899, + "CGTAGG": 2900, + "CGTTAA": 2901, + "CGTTAT": 2902, + "CGTTAC": 2903, + "CGTTAG": 2904, + "CGTTTA": 2905, + "CGTTTT": 2906, + "CGTTTC": 2907, + "CGTTTG": 2908, + "CGTTCA": 2909, + "CGTTCT": 2910, + "CGTTCC": 2911, + "CGTTCG": 2912, + "CGTTGA": 2913, + "CGTTGT": 2914, + "CGTTGC": 2915, + "CGTTGG": 2916, + "CGTCAA": 2917, + "CGTCAT": 2918, + "CGTCAC": 2919, + "CGTCAG": 2920, + "CGTCTA": 2921, + "CGTCTT": 2922, + "CGTCTC": 2923, + "CGTCTG": 2924, + "CGTCCA": 2925, + "CGTCCT": 2926, + "CGTCCC": 2927, + "CGTCCG": 2928, + "CGTCGA": 2929, + "CGTCGT": 2930, + "CGTCGC": 2931, + "CGTCGG": 2932, + "CGTGAA": 2933, + "CGTGAT": 2934, + "CGTGAC": 2935, + "CGTGAG": 2936, + "CGTGTA": 2937, + "CGTGTT": 2938, + "CGTGTC": 2939, + "CGTGTG": 2940, + "CGTGCA": 2941, + "CGTGCT": 2942, + "CGTGCC": 2943, + "CGTGCG": 2944, + "CGTGGA": 2945, + "CGTGGT": 2946, + "CGTGGC": 2947, + "CGTGGG": 2948, + "CGCAAA": 2949, + "CGCAAT": 2950, + "CGCAAC": 2951, + "CGCAAG": 2952, + "CGCATA": 2953, + "CGCATT": 2954, + "CGCATC": 2955, + "CGCATG": 2956, + "CGCACA": 2957, + "CGCACT": 2958, + "CGCACC": 2959, + "CGCACG": 2960, + "CGCAGA": 2961, + "CGCAGT": 2962, + "CGCAGC": 2963, + "CGCAGG": 2964, + "CGCTAA": 2965, + "CGCTAT": 2966, + "CGCTAC": 2967, + "CGCTAG": 2968, + "CGCTTA": 2969, + "CGCTTT": 2970, + "CGCTTC": 2971, + "CGCTTG": 2972, + "CGCTCA": 2973, + "CGCTCT": 2974, + "CGCTCC": 2975, + "CGCTCG": 2976, + "CGCTGA": 2977, + "CGCTGT": 2978, + "CGCTGC": 2979, + "CGCTGG": 2980, + "CGCCAA": 2981, + "CGCCAT": 2982, + "CGCCAC": 2983, + "CGCCAG": 2984, + "CGCCTA": 2985, + "CGCCTT": 2986, + "CGCCTC": 2987, + "CGCCTG": 2988, + "CGCCCA": 2989, + "CGCCCT": 2990, + "CGCCCC": 2991, + "CGCCCG": 2992, + "CGCCGA": 2993, + "CGCCGT": 2994, + "CGCCGC": 2995, + "CGCCGG": 2996, + "CGCGAA": 2997, + "CGCGAT": 2998, + "CGCGAC": 2999, + "CGCGAG": 3000, + "CGCGTA": 3001, + "CGCGTT": 3002, + "CGCGTC": 3003, + "CGCGTG": 3004, + "CGCGCA": 3005, + "CGCGCT": 3006, + "CGCGCC": 3007, + "CGCGCG": 3008, + "CGCGGA": 3009, + "CGCGGT": 3010, + "CGCGGC": 3011, + "CGCGGG": 3012, + "CGGAAA": 3013, + "CGGAAT": 3014, + "CGGAAC": 3015, + "CGGAAG": 3016, + "CGGATA": 3017, + "CGGATT": 3018, + "CGGATC": 3019, + "CGGATG": 3020, + "CGGACA": 3021, + "CGGACT": 3022, + "CGGACC": 3023, + "CGGACG": 3024, + "CGGAGA": 3025, + "CGGAGT": 3026, + "CGGAGC": 3027, + "CGGAGG": 3028, + "CGGTAA": 3029, + "CGGTAT": 3030, + "CGGTAC": 3031, + "CGGTAG": 3032, + "CGGTTA": 3033, + "CGGTTT": 3034, + "CGGTTC": 3035, + "CGGTTG": 3036, + "CGGTCA": 3037, + "CGGTCT": 3038, + "CGGTCC": 3039, + "CGGTCG": 3040, + "CGGTGA": 3041, + "CGGTGT": 3042, + "CGGTGC": 3043, + "CGGTGG": 3044, + "CGGCAA": 3045, + "CGGCAT": 3046, + "CGGCAC": 3047, + "CGGCAG": 3048, + "CGGCTA": 3049, + "CGGCTT": 3050, + "CGGCTC": 3051, + "CGGCTG": 3052, + "CGGCCA": 3053, + "CGGCCT": 3054, + "CGGCCC": 3055, + "CGGCCG": 3056, + "CGGCGA": 3057, + "CGGCGT": 3058, + "CGGCGC": 3059, + "CGGCGG": 3060, + "CGGGAA": 3061, + "CGGGAT": 3062, + "CGGGAC": 3063, + "CGGGAG": 3064, + "CGGGTA": 3065, + "CGGGTT": 3066, + "CGGGTC": 3067, + "CGGGTG": 3068, + "CGGGCA": 3069, + "CGGGCT": 3070, + "CGGGCC": 3071, + "CGGGCG": 3072, + "CGGGGA": 3073, + "CGGGGT": 3074, + "CGGGGC": 3075, + "CGGGGG": 3076, + "GAAAAA": 3077, + "GAAAAT": 3078, + "GAAAAC": 3079, + "GAAAAG": 3080, + "GAAATA": 3081, + "GAAATT": 3082, + "GAAATC": 3083, + "GAAATG": 3084, + "GAAACA": 3085, + "GAAACT": 3086, + "GAAACC": 3087, + "GAAACG": 3088, + "GAAAGA": 3089, + "GAAAGT": 3090, + "GAAAGC": 3091, + "GAAAGG": 3092, + "GAATAA": 3093, + "GAATAT": 3094, + "GAATAC": 3095, + "GAATAG": 3096, + "GAATTA": 3097, + "GAATTT": 3098, + "GAATTC": 3099, + "GAATTG": 3100, + "GAATCA": 3101, + "GAATCT": 3102, + "GAATCC": 3103, + "GAATCG": 3104, + "GAATGA": 3105, + "GAATGT": 3106, + "GAATGC": 3107, + "GAATGG": 3108, + "GAACAA": 3109, + "GAACAT": 3110, + "GAACAC": 3111, + "GAACAG": 3112, + "GAACTA": 3113, + "GAACTT": 3114, + "GAACTC": 3115, + "GAACTG": 3116, + "GAACCA": 3117, + "GAACCT": 3118, + "GAACCC": 3119, + "GAACCG": 3120, + "GAACGA": 3121, + "GAACGT": 3122, + "GAACGC": 3123, + "GAACGG": 3124, + "GAAGAA": 3125, + "GAAGAT": 3126, + "GAAGAC": 3127, + "GAAGAG": 3128, + "GAAGTA": 3129, + "GAAGTT": 3130, + "GAAGTC": 3131, + "GAAGTG": 3132, + "GAAGCA": 3133, + "GAAGCT": 3134, + "GAAGCC": 3135, + "GAAGCG": 3136, + "GAAGGA": 3137, + "GAAGGT": 3138, + "GAAGGC": 3139, + "GAAGGG": 3140, + "GATAAA": 3141, + "GATAAT": 3142, + "GATAAC": 3143, + "GATAAG": 3144, + "GATATA": 3145, + "GATATT": 3146, + "GATATC": 3147, + "GATATG": 3148, + "GATACA": 3149, + "GATACT": 3150, + "GATACC": 3151, + "GATACG": 3152, + "GATAGA": 3153, + "GATAGT": 3154, + "GATAGC": 3155, + "GATAGG": 3156, + "GATTAA": 3157, + "GATTAT": 3158, + "GATTAC": 3159, + "GATTAG": 3160, + "GATTTA": 3161, + "GATTTT": 3162, + "GATTTC": 3163, + "GATTTG": 3164, + "GATTCA": 3165, + "GATTCT": 3166, + "GATTCC": 3167, + "GATTCG": 3168, + "GATTGA": 3169, + "GATTGT": 3170, + "GATTGC": 3171, + "GATTGG": 3172, + "GATCAA": 3173, + "GATCAT": 3174, + "GATCAC": 3175, + "GATCAG": 3176, + "GATCTA": 3177, + "GATCTT": 3178, + "GATCTC": 3179, + "GATCTG": 3180, + "GATCCA": 3181, + "GATCCT": 3182, + "GATCCC": 3183, + "GATCCG": 3184, + "GATCGA": 3185, + "GATCGT": 3186, + "GATCGC": 3187, + "GATCGG": 3188, + "GATGAA": 3189, + "GATGAT": 3190, + "GATGAC": 3191, + "GATGAG": 3192, + "GATGTA": 3193, + "GATGTT": 3194, + "GATGTC": 3195, + "GATGTG": 3196, + "GATGCA": 3197, + "GATGCT": 3198, + "GATGCC": 3199, + "GATGCG": 3200, + "GATGGA": 3201, + "GATGGT": 3202, + "GATGGC": 3203, + "GATGGG": 3204, + "GACAAA": 3205, + "GACAAT": 3206, + "GACAAC": 3207, + "GACAAG": 3208, + "GACATA": 3209, + "GACATT": 3210, + "GACATC": 3211, + "GACATG": 3212, + "GACACA": 3213, + "GACACT": 3214, + "GACACC": 3215, + "GACACG": 3216, + "GACAGA": 3217, + "GACAGT": 3218, + "GACAGC": 3219, + "GACAGG": 3220, + "GACTAA": 3221, + "GACTAT": 3222, + "GACTAC": 3223, + "GACTAG": 3224, + "GACTTA": 3225, + "GACTTT": 3226, + "GACTTC": 3227, + "GACTTG": 3228, + "GACTCA": 3229, + "GACTCT": 3230, + "GACTCC": 3231, + "GACTCG": 3232, + "GACTGA": 3233, + "GACTGT": 3234, + "GACTGC": 3235, + "GACTGG": 3236, + "GACCAA": 3237, + "GACCAT": 3238, + "GACCAC": 3239, + "GACCAG": 3240, + "GACCTA": 3241, + "GACCTT": 3242, + "GACCTC": 3243, + "GACCTG": 3244, + "GACCCA": 3245, + "GACCCT": 3246, + "GACCCC": 3247, + "GACCCG": 3248, + "GACCGA": 3249, + "GACCGT": 3250, + "GACCGC": 3251, + "GACCGG": 3252, + "GACGAA": 3253, + "GACGAT": 3254, + "GACGAC": 3255, + "GACGAG": 3256, + "GACGTA": 3257, + "GACGTT": 3258, + "GACGTC": 3259, + "GACGTG": 3260, + "GACGCA": 3261, + "GACGCT": 3262, + "GACGCC": 3263, + "GACGCG": 3264, + "GACGGA": 3265, + "GACGGT": 3266, + "GACGGC": 3267, + "GACGGG": 3268, + "GAGAAA": 3269, + "GAGAAT": 3270, + "GAGAAC": 3271, + "GAGAAG": 3272, + "GAGATA": 3273, + "GAGATT": 3274, + "GAGATC": 3275, + "GAGATG": 3276, + "GAGACA": 3277, + "GAGACT": 3278, + "GAGACC": 3279, + "GAGACG": 3280, + "GAGAGA": 3281, + "GAGAGT": 3282, + "GAGAGC": 3283, + "GAGAGG": 3284, + "GAGTAA": 3285, + "GAGTAT": 3286, + "GAGTAC": 3287, + "GAGTAG": 3288, + "GAGTTA": 3289, + "GAGTTT": 3290, + "GAGTTC": 3291, + "GAGTTG": 3292, + "GAGTCA": 3293, + "GAGTCT": 3294, + "GAGTCC": 3295, + "GAGTCG": 3296, + "GAGTGA": 3297, + "GAGTGT": 3298, + "GAGTGC": 3299, + "GAGTGG": 3300, + "GAGCAA": 3301, + "GAGCAT": 3302, + "GAGCAC": 3303, + "GAGCAG": 3304, + "GAGCTA": 3305, + "GAGCTT": 3306, + "GAGCTC": 3307, + "GAGCTG": 3308, + "GAGCCA": 3309, + "GAGCCT": 3310, + "GAGCCC": 3311, + "GAGCCG": 3312, + "GAGCGA": 3313, + "GAGCGT": 3314, + "GAGCGC": 3315, + "GAGCGG": 3316, + "GAGGAA": 3317, + "GAGGAT": 3318, + "GAGGAC": 3319, + "GAGGAG": 3320, + "GAGGTA": 3321, + "GAGGTT": 3322, + "GAGGTC": 3323, + "GAGGTG": 3324, + "GAGGCA": 3325, + "GAGGCT": 3326, + "GAGGCC": 3327, + "GAGGCG": 3328, + "GAGGGA": 3329, + "GAGGGT": 3330, + "GAGGGC": 3331, + "GAGGGG": 3332, + "GTAAAA": 3333, + "GTAAAT": 3334, + "GTAAAC": 3335, + "GTAAAG": 3336, + "GTAATA": 3337, + "GTAATT": 3338, + "GTAATC": 3339, + "GTAATG": 3340, + "GTAACA": 3341, + "GTAACT": 3342, + "GTAACC": 3343, + "GTAACG": 3344, + "GTAAGA": 3345, + "GTAAGT": 3346, + "GTAAGC": 3347, + "GTAAGG": 3348, + "GTATAA": 3349, + "GTATAT": 3350, + "GTATAC": 3351, + "GTATAG": 3352, + "GTATTA": 3353, + "GTATTT": 3354, + "GTATTC": 3355, + "GTATTG": 3356, + "GTATCA": 3357, + "GTATCT": 3358, + "GTATCC": 3359, + "GTATCG": 3360, + "GTATGA": 3361, + "GTATGT": 3362, + "GTATGC": 3363, + "GTATGG": 3364, + "GTACAA": 3365, + "GTACAT": 3366, + "GTACAC": 3367, + "GTACAG": 3368, + "GTACTA": 3369, + "GTACTT": 3370, + "GTACTC": 3371, + "GTACTG": 3372, + "GTACCA": 3373, + "GTACCT": 3374, + "GTACCC": 3375, + "GTACCG": 3376, + "GTACGA": 3377, + "GTACGT": 3378, + "GTACGC": 3379, + "GTACGG": 3380, + "GTAGAA": 3381, + "GTAGAT": 3382, + "GTAGAC": 3383, + "GTAGAG": 3384, + "GTAGTA": 3385, + "GTAGTT": 3386, + "GTAGTC": 3387, + "GTAGTG": 3388, + "GTAGCA": 3389, + "GTAGCT": 3390, + "GTAGCC": 3391, + "GTAGCG": 3392, + "GTAGGA": 3393, + "GTAGGT": 3394, + "GTAGGC": 3395, + "GTAGGG": 3396, + "GTTAAA": 3397, + "GTTAAT": 3398, + "GTTAAC": 3399, + "GTTAAG": 3400, + "GTTATA": 3401, + "GTTATT": 3402, + "GTTATC": 3403, + "GTTATG": 3404, + "GTTACA": 3405, + "GTTACT": 3406, + "GTTACC": 3407, + "GTTACG": 3408, + "GTTAGA": 3409, + "GTTAGT": 3410, + "GTTAGC": 3411, + "GTTAGG": 3412, + "GTTTAA": 3413, + "GTTTAT": 3414, + "GTTTAC": 3415, + "GTTTAG": 3416, + "GTTTTA": 3417, + "GTTTTT": 3418, + "GTTTTC": 3419, + "GTTTTG": 3420, + "GTTTCA": 3421, + "GTTTCT": 3422, + "GTTTCC": 3423, + "GTTTCG": 3424, + "GTTTGA": 3425, + "GTTTGT": 3426, + "GTTTGC": 3427, + "GTTTGG": 3428, + "GTTCAA": 3429, + "GTTCAT": 3430, + "GTTCAC": 3431, + "GTTCAG": 3432, + "GTTCTA": 3433, + "GTTCTT": 3434, + "GTTCTC": 3435, + "GTTCTG": 3436, + "GTTCCA": 3437, + "GTTCCT": 3438, + "GTTCCC": 3439, + "GTTCCG": 3440, + "GTTCGA": 3441, + "GTTCGT": 3442, + "GTTCGC": 3443, + "GTTCGG": 3444, + "GTTGAA": 3445, + "GTTGAT": 3446, + "GTTGAC": 3447, + "GTTGAG": 3448, + "GTTGTA": 3449, + "GTTGTT": 3450, + "GTTGTC": 3451, + "GTTGTG": 3452, + "GTTGCA": 3453, + "GTTGCT": 3454, + "GTTGCC": 3455, + "GTTGCG": 3456, + "GTTGGA": 3457, + "GTTGGT": 3458, + "GTTGGC": 3459, + "GTTGGG": 3460, + "GTCAAA": 3461, + "GTCAAT": 3462, + "GTCAAC": 3463, + "GTCAAG": 3464, + "GTCATA": 3465, + "GTCATT": 3466, + "GTCATC": 3467, + "GTCATG": 3468, + "GTCACA": 3469, + "GTCACT": 3470, + "GTCACC": 3471, + "GTCACG": 3472, + "GTCAGA": 3473, + "GTCAGT": 3474, + "GTCAGC": 3475, + "GTCAGG": 3476, + "GTCTAA": 3477, + "GTCTAT": 3478, + "GTCTAC": 3479, + "GTCTAG": 3480, + "GTCTTA": 3481, + "GTCTTT": 3482, + "GTCTTC": 3483, + "GTCTTG": 3484, + "GTCTCA": 3485, + "GTCTCT": 3486, + "GTCTCC": 3487, + "GTCTCG": 3488, + "GTCTGA": 3489, + "GTCTGT": 3490, + "GTCTGC": 3491, + "GTCTGG": 3492, + "GTCCAA": 3493, + "GTCCAT": 3494, + "GTCCAC": 3495, + "GTCCAG": 3496, + "GTCCTA": 3497, + "GTCCTT": 3498, + "GTCCTC": 3499, + "GTCCTG": 3500, + "GTCCCA": 3501, + "GTCCCT": 3502, + "GTCCCC": 3503, + "GTCCCG": 3504, + "GTCCGA": 3505, + "GTCCGT": 3506, + "GTCCGC": 3507, + "GTCCGG": 3508, + "GTCGAA": 3509, + "GTCGAT": 3510, + "GTCGAC": 3511, + "GTCGAG": 3512, + "GTCGTA": 3513, + "GTCGTT": 3514, + "GTCGTC": 3515, + "GTCGTG": 3516, + "GTCGCA": 3517, + "GTCGCT": 3518, + "GTCGCC": 3519, + "GTCGCG": 3520, + "GTCGGA": 3521, + "GTCGGT": 3522, + "GTCGGC": 3523, + "GTCGGG": 3524, + "GTGAAA": 3525, + "GTGAAT": 3526, + "GTGAAC": 3527, + "GTGAAG": 3528, + "GTGATA": 3529, + "GTGATT": 3530, + "GTGATC": 3531, + "GTGATG": 3532, + "GTGACA": 3533, + "GTGACT": 3534, + "GTGACC": 3535, + "GTGACG": 3536, + "GTGAGA": 3537, + "GTGAGT": 3538, + "GTGAGC": 3539, + "GTGAGG": 3540, + "GTGTAA": 3541, + "GTGTAT": 3542, + "GTGTAC": 3543, + "GTGTAG": 3544, + "GTGTTA": 3545, + "GTGTTT": 3546, + "GTGTTC": 3547, + "GTGTTG": 3548, + "GTGTCA": 3549, + "GTGTCT": 3550, + "GTGTCC": 3551, + "GTGTCG": 3552, + "GTGTGA": 3553, + "GTGTGT": 3554, + "GTGTGC": 3555, + "GTGTGG": 3556, + "GTGCAA": 3557, + "GTGCAT": 3558, + "GTGCAC": 3559, + "GTGCAG": 3560, + "GTGCTA": 3561, + "GTGCTT": 3562, + "GTGCTC": 3563, + "GTGCTG": 3564, + "GTGCCA": 3565, + "GTGCCT": 3566, + "GTGCCC": 3567, + "GTGCCG": 3568, + "GTGCGA": 3569, + "GTGCGT": 3570, + "GTGCGC": 3571, + "GTGCGG": 3572, + "GTGGAA": 3573, + "GTGGAT": 3574, + "GTGGAC": 3575, + "GTGGAG": 3576, + "GTGGTA": 3577, + "GTGGTT": 3578, + "GTGGTC": 3579, + "GTGGTG": 3580, + "GTGGCA": 3581, + "GTGGCT": 3582, + "GTGGCC": 3583, + "GTGGCG": 3584, + "GTGGGA": 3585, + "GTGGGT": 3586, + "GTGGGC": 3587, + "GTGGGG": 3588, + "GCAAAA": 3589, + "GCAAAT": 3590, + "GCAAAC": 3591, + "GCAAAG": 3592, + "GCAATA": 3593, + "GCAATT": 3594, + "GCAATC": 3595, + "GCAATG": 3596, + "GCAACA": 3597, + "GCAACT": 3598, + "GCAACC": 3599, + "GCAACG": 3600, + "GCAAGA": 3601, + "GCAAGT": 3602, + "GCAAGC": 3603, + "GCAAGG": 3604, + "GCATAA": 3605, + "GCATAT": 3606, + "GCATAC": 3607, + "GCATAG": 3608, + "GCATTA": 3609, + "GCATTT": 3610, + "GCATTC": 3611, + "GCATTG": 3612, + "GCATCA": 3613, + "GCATCT": 3614, + "GCATCC": 3615, + "GCATCG": 3616, + "GCATGA": 3617, + "GCATGT": 3618, + "GCATGC": 3619, + "GCATGG": 3620, + "GCACAA": 3621, + "GCACAT": 3622, + "GCACAC": 3623, + "GCACAG": 3624, + "GCACTA": 3625, + "GCACTT": 3626, + "GCACTC": 3627, + "GCACTG": 3628, + "GCACCA": 3629, + "GCACCT": 3630, + "GCACCC": 3631, + "GCACCG": 3632, + "GCACGA": 3633, + "GCACGT": 3634, + "GCACGC": 3635, + "GCACGG": 3636, + "GCAGAA": 3637, + "GCAGAT": 3638, + "GCAGAC": 3639, + "GCAGAG": 3640, + "GCAGTA": 3641, + "GCAGTT": 3642, + "GCAGTC": 3643, + "GCAGTG": 3644, + "GCAGCA": 3645, + "GCAGCT": 3646, + "GCAGCC": 3647, + "GCAGCG": 3648, + "GCAGGA": 3649, + "GCAGGT": 3650, + "GCAGGC": 3651, + "GCAGGG": 3652, + "GCTAAA": 3653, + "GCTAAT": 3654, + "GCTAAC": 3655, + "GCTAAG": 3656, + "GCTATA": 3657, + "GCTATT": 3658, + "GCTATC": 3659, + "GCTATG": 3660, + "GCTACA": 3661, + "GCTACT": 3662, + "GCTACC": 3663, + "GCTACG": 3664, + "GCTAGA": 3665, + "GCTAGT": 3666, + "GCTAGC": 3667, + "GCTAGG": 3668, + "GCTTAA": 3669, + "GCTTAT": 3670, + "GCTTAC": 3671, + "GCTTAG": 3672, + "GCTTTA": 3673, + "GCTTTT": 3674, + "GCTTTC": 3675, + "GCTTTG": 3676, + "GCTTCA": 3677, + "GCTTCT": 3678, + "GCTTCC": 3679, + "GCTTCG": 3680, + "GCTTGA": 3681, + "GCTTGT": 3682, + "GCTTGC": 3683, + "GCTTGG": 3684, + "GCTCAA": 3685, + "GCTCAT": 3686, + "GCTCAC": 3687, + "GCTCAG": 3688, + "GCTCTA": 3689, + "GCTCTT": 3690, + "GCTCTC": 3691, + "GCTCTG": 3692, + "GCTCCA": 3693, + "GCTCCT": 3694, + "GCTCCC": 3695, + "GCTCCG": 3696, + "GCTCGA": 3697, + "GCTCGT": 3698, + "GCTCGC": 3699, + "GCTCGG": 3700, + "GCTGAA": 3701, + "GCTGAT": 3702, + "GCTGAC": 3703, + "GCTGAG": 3704, + "GCTGTA": 3705, + "GCTGTT": 3706, + "GCTGTC": 3707, + "GCTGTG": 3708, + "GCTGCA": 3709, + "GCTGCT": 3710, + "GCTGCC": 3711, + "GCTGCG": 3712, + "GCTGGA": 3713, + "GCTGGT": 3714, + "GCTGGC": 3715, + "GCTGGG": 3716, + "GCCAAA": 3717, + "GCCAAT": 3718, + "GCCAAC": 3719, + "GCCAAG": 3720, + "GCCATA": 3721, + "GCCATT": 3722, + "GCCATC": 3723, + "GCCATG": 3724, + "GCCACA": 3725, + "GCCACT": 3726, + "GCCACC": 3727, + "GCCACG": 3728, + "GCCAGA": 3729, + "GCCAGT": 3730, + "GCCAGC": 3731, + "GCCAGG": 3732, + "GCCTAA": 3733, + "GCCTAT": 3734, + "GCCTAC": 3735, + "GCCTAG": 3736, + "GCCTTA": 3737, + "GCCTTT": 3738, + "GCCTTC": 3739, + "GCCTTG": 3740, + "GCCTCA": 3741, + "GCCTCT": 3742, + "GCCTCC": 3743, + "GCCTCG": 3744, + "GCCTGA": 3745, + "GCCTGT": 3746, + "GCCTGC": 3747, + "GCCTGG": 3748, + "GCCCAA": 3749, + "GCCCAT": 3750, + "GCCCAC": 3751, + "GCCCAG": 3752, + "GCCCTA": 3753, + "GCCCTT": 3754, + "GCCCTC": 3755, + "GCCCTG": 3756, + "GCCCCA": 3757, + "GCCCCT": 3758, + "GCCCCC": 3759, + "GCCCCG": 3760, + "GCCCGA": 3761, + "GCCCGT": 3762, + "GCCCGC": 3763, + "GCCCGG": 3764, + "GCCGAA": 3765, + "GCCGAT": 3766, + "GCCGAC": 3767, + "GCCGAG": 3768, + "GCCGTA": 3769, + "GCCGTT": 3770, + "GCCGTC": 3771, + "GCCGTG": 3772, + "GCCGCA": 3773, + "GCCGCT": 3774, + "GCCGCC": 3775, + "GCCGCG": 3776, + "GCCGGA": 3777, + "GCCGGT": 3778, + "GCCGGC": 3779, + "GCCGGG": 3780, + "GCGAAA": 3781, + "GCGAAT": 3782, + "GCGAAC": 3783, + "GCGAAG": 3784, + "GCGATA": 3785, + "GCGATT": 3786, + "GCGATC": 3787, + "GCGATG": 3788, + "GCGACA": 3789, + "GCGACT": 3790, + "GCGACC": 3791, + "GCGACG": 3792, + "GCGAGA": 3793, + "GCGAGT": 3794, + "GCGAGC": 3795, + "GCGAGG": 3796, + "GCGTAA": 3797, + "GCGTAT": 3798, + "GCGTAC": 3799, + "GCGTAG": 3800, + "GCGTTA": 3801, + "GCGTTT": 3802, + "GCGTTC": 3803, + "GCGTTG": 3804, + "GCGTCA": 3805, + "GCGTCT": 3806, + "GCGTCC": 3807, + "GCGTCG": 3808, + "GCGTGA": 3809, + "GCGTGT": 3810, + "GCGTGC": 3811, + "GCGTGG": 3812, + "GCGCAA": 3813, + "GCGCAT": 3814, + "GCGCAC": 3815, + "GCGCAG": 3816, + "GCGCTA": 3817, + "GCGCTT": 3818, + "GCGCTC": 3819, + "GCGCTG": 3820, + "GCGCCA": 3821, + "GCGCCT": 3822, + "GCGCCC": 3823, + "GCGCCG": 3824, + "GCGCGA": 3825, + "GCGCGT": 3826, + "GCGCGC": 3827, + "GCGCGG": 3828, + "GCGGAA": 3829, + "GCGGAT": 3830, + "GCGGAC": 3831, + "GCGGAG": 3832, + "GCGGTA": 3833, + "GCGGTT": 3834, + "GCGGTC": 3835, + "GCGGTG": 3836, + "GCGGCA": 3837, + "GCGGCT": 3838, + "GCGGCC": 3839, + "GCGGCG": 3840, + "GCGGGA": 3841, + "GCGGGT": 3842, + "GCGGGC": 3843, + "GCGGGG": 3844, + "GGAAAA": 3845, + "GGAAAT": 3846, + "GGAAAC": 3847, + "GGAAAG": 3848, + "GGAATA": 3849, + "GGAATT": 3850, + "GGAATC": 3851, + "GGAATG": 3852, + "GGAACA": 3853, + "GGAACT": 3854, + "GGAACC": 3855, + "GGAACG": 3856, + "GGAAGA": 3857, + "GGAAGT": 3858, + "GGAAGC": 3859, + "GGAAGG": 3860, + "GGATAA": 3861, + "GGATAT": 3862, + "GGATAC": 3863, + "GGATAG": 3864, + "GGATTA": 3865, + "GGATTT": 3866, + "GGATTC": 3867, + "GGATTG": 3868, + "GGATCA": 3869, + "GGATCT": 3870, + "GGATCC": 3871, + "GGATCG": 3872, + "GGATGA": 3873, + "GGATGT": 3874, + "GGATGC": 3875, + "GGATGG": 3876, + "GGACAA": 3877, + "GGACAT": 3878, + "GGACAC": 3879, + "GGACAG": 3880, + "GGACTA": 3881, + "GGACTT": 3882, + "GGACTC": 3883, + "GGACTG": 3884, + "GGACCA": 3885, + "GGACCT": 3886, + "GGACCC": 3887, + "GGACCG": 3888, + "GGACGA": 3889, + "GGACGT": 3890, + "GGACGC": 3891, + "GGACGG": 3892, + "GGAGAA": 3893, + "GGAGAT": 3894, + "GGAGAC": 3895, + "GGAGAG": 3896, + "GGAGTA": 3897, + "GGAGTT": 3898, + "GGAGTC": 3899, + "GGAGTG": 3900, + "GGAGCA": 3901, + "GGAGCT": 3902, + "GGAGCC": 3903, + "GGAGCG": 3904, + "GGAGGA": 3905, + "GGAGGT": 3906, + "GGAGGC": 3907, + "GGAGGG": 3908, + "GGTAAA": 3909, + "GGTAAT": 3910, + "GGTAAC": 3911, + "GGTAAG": 3912, + "GGTATA": 3913, + "GGTATT": 3914, + "GGTATC": 3915, + "GGTATG": 3916, + "GGTACA": 3917, + "GGTACT": 3918, + "GGTACC": 3919, + "GGTACG": 3920, + "GGTAGA": 3921, + "GGTAGT": 3922, + "GGTAGC": 3923, + "GGTAGG": 3924, + "GGTTAA": 3925, + "GGTTAT": 3926, + "GGTTAC": 3927, + "GGTTAG": 3928, + "GGTTTA": 3929, + "GGTTTT": 3930, + "GGTTTC": 3931, + "GGTTTG": 3932, + "GGTTCA": 3933, + "GGTTCT": 3934, + "GGTTCC": 3935, + "GGTTCG": 3936, + "GGTTGA": 3937, + "GGTTGT": 3938, + "GGTTGC": 3939, + "GGTTGG": 3940, + "GGTCAA": 3941, + "GGTCAT": 3942, + "GGTCAC": 3943, + "GGTCAG": 3944, + "GGTCTA": 3945, + "GGTCTT": 3946, + "GGTCTC": 3947, + "GGTCTG": 3948, + "GGTCCA": 3949, + "GGTCCT": 3950, + "GGTCCC": 3951, + "GGTCCG": 3952, + "GGTCGA": 3953, + "GGTCGT": 3954, + "GGTCGC": 3955, + "GGTCGG": 3956, + "GGTGAA": 3957, + "GGTGAT": 3958, + "GGTGAC": 3959, + "GGTGAG": 3960, + "GGTGTA": 3961, + "GGTGTT": 3962, + "GGTGTC": 3963, + "GGTGTG": 3964, + "GGTGCA": 3965, + "GGTGCT": 3966, + "GGTGCC": 3967, + "GGTGCG": 3968, + "GGTGGA": 3969, + "GGTGGT": 3970, + "GGTGGC": 3971, + "GGTGGG": 3972, + "GGCAAA": 3973, + "GGCAAT": 3974, + "GGCAAC": 3975, + "GGCAAG": 3976, + "GGCATA": 3977, + "GGCATT": 3978, + "GGCATC": 3979, + "GGCATG": 3980, + "GGCACA": 3981, + "GGCACT": 3982, + "GGCACC": 3983, + "GGCACG": 3984, + "GGCAGA": 3985, + "GGCAGT": 3986, + "GGCAGC": 3987, + "GGCAGG": 3988, + "GGCTAA": 3989, + "GGCTAT": 3990, + "GGCTAC": 3991, + "GGCTAG": 3992, + "GGCTTA": 3993, + "GGCTTT": 3994, + "GGCTTC": 3995, + "GGCTTG": 3996, + "GGCTCA": 3997, + "GGCTCT": 3998, + "GGCTCC": 3999, + "GGCTCG": 4000, + "GGCTGA": 4001, + "GGCTGT": 4002, + "GGCTGC": 4003, + "GGCTGG": 4004, + "GGCCAA": 4005, + "GGCCAT": 4006, + "GGCCAC": 4007, + "GGCCAG": 4008, + "GGCCTA": 4009, + "GGCCTT": 4010, + "GGCCTC": 4011, + "GGCCTG": 4012, + "GGCCCA": 4013, + "GGCCCT": 4014, + "GGCCCC": 4015, + "GGCCCG": 4016, + "GGCCGA": 4017, + "GGCCGT": 4018, + "GGCCGC": 4019, + "GGCCGG": 4020, + "GGCGAA": 4021, + "GGCGAT": 4022, + "GGCGAC": 4023, + "GGCGAG": 4024, + "GGCGTA": 4025, + "GGCGTT": 4026, + "GGCGTC": 4027, + "GGCGTG": 4028, + "GGCGCA": 4029, + "GGCGCT": 4030, + "GGCGCC": 4031, + "GGCGCG": 4032, + "GGCGGA": 4033, + "GGCGGT": 4034, + "GGCGGC": 4035, + "GGCGGG": 4036, + "GGGAAA": 4037, + "GGGAAT": 4038, + "GGGAAC": 4039, + "GGGAAG": 4040, + "GGGATA": 4041, + "GGGATT": 4042, + "GGGATC": 4043, + "GGGATG": 4044, + "GGGACA": 4045, + "GGGACT": 4046, + "GGGACC": 4047, + "GGGACG": 4048, + "GGGAGA": 4049, + "GGGAGT": 4050, + "GGGAGC": 4051, + "GGGAGG": 4052, + "GGGTAA": 4053, + "GGGTAT": 4054, + "GGGTAC": 4055, + "GGGTAG": 4056, + "GGGTTA": 4057, + "GGGTTT": 4058, + "GGGTTC": 4059, + "GGGTTG": 4060, + "GGGTCA": 4061, + "GGGTCT": 4062, + "GGGTCC": 4063, + "GGGTCG": 4064, + "GGGTGA": 4065, + "GGGTGT": 4066, + "GGGTGC": 4067, + "GGGTGG": 4068, + "GGGCAA": 4069, + "GGGCAT": 4070, + "GGGCAC": 4071, + "GGGCAG": 4072, + "GGGCTA": 4073, + "GGGCTT": 4074, + "GGGCTC": 4075, + "GGGCTG": 4076, + "GGGCCA": 4077, + "GGGCCT": 4078, + "GGGCCC": 4079, + "GGGCCG": 4080, + "GGGCGA": 4081, + "GGGCGT": 4082, + "GGGCGC": 4083, + "GGGCGG": 4084, + "GGGGAA": 4085, + "GGGGAT": 4086, + "GGGGAC": 4087, + "GGGGAG": 4088, + "GGGGTA": 4089, + "GGGGTT": 4090, + "GGGGTC": 4091, + "GGGGTG": 4092, + "GGGGCA": 4093, + "GGGGCT": 4094, + "GGGGCC": 4095, + "GGGGCG": 4096, + "GGGGGA": 4097, + "GGGGGT": 4098, + "GGGGGC": 4099, + "GGGGGG": 4100 + } + } +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-4000/tokenizer_config.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-4000/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..604e2b1b29af1dcb3920d2828b6a90276e5e3d02 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-4000/tokenizer_config.json @@ -0,0 +1,58 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "[UNK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "[CLS]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "[SEP]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "4": { + "content": "[MASK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "clean_up_tokenization_spaces": true, + "cls_token": "[CLS]", + "do_basic_tokenize": true, + "do_lower_case": false, + "mask_token": "[MASK]", + "max_len": 512, + "model_max_length": 512, + "never_split": null, + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "strip_accents": null, + "tokenize_chinese_chars": true, + "tokenizer_class": "BertTokenizer", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-4000/trainer_state.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-4000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..4259c45d87787b836613cc72b0cb1c1c2f5b23f1 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-4000/trainer_state.json @@ -0,0 +1,1153 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 16.0, + "eval_steps": 500, + "global_step": 4000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.1, + "grad_norm": 1.1046091318130493, + "learning_rate": 5.319148936170213e-05, + "loss": 0.6562, + "step": 25 + }, + { + "epoch": 0.2, + "grad_norm": 0.4066573977470398, + "learning_rate": 0.00010638297872340425, + "loss": 0.5373, + "step": 50 + }, + { + "epoch": 0.3, + "grad_norm": 0.7340325713157654, + "learning_rate": 0.00015957446808510637, + "loss": 0.4995, + "step": 75 + }, + { + "epoch": 0.4, + "grad_norm": 2.134040117263794, + "learning_rate": 0.0002127659574468085, + "loss": 0.4871, + "step": 100 + }, + { + "epoch": 0.5, + "grad_norm": 2.911975860595703, + "learning_rate": 0.00026595744680851064, + "loss": 0.4507, + "step": 125 + }, + { + "epoch": 0.6, + "grad_norm": 2.040365219116211, + "learning_rate": 0.00031914893617021275, + "loss": 0.3944, + "step": 150 + }, + { + "epoch": 0.7, + "grad_norm": 5.643661022186279, + "learning_rate": 0.0003723404255319149, + "loss": 0.3863, + "step": 175 + }, + { + "epoch": 0.8, + "grad_norm": 3.882492780685425, + "learning_rate": 0.0003992081821181128, + "loss": 0.3875, + "step": 200 + }, + { + "epoch": 0.9, + "grad_norm": 5.87437105178833, + "learning_rate": 0.0003975585615308479, + "loss": 0.3856, + "step": 225 + }, + { + "epoch": 1.0, + "grad_norm": 4.611206531524658, + "learning_rate": 0.00039590894094358297, + "loss": 0.3866, + "step": 250 + }, + { + "epoch": 1.1, + "grad_norm": 3.6404929161071777, + "learning_rate": 0.0003942593203563181, + "loss": 0.3437, + "step": 275 + }, + { + "epoch": 1.2, + "grad_norm": 6.469444751739502, + "learning_rate": 0.0003926096997690532, + "loss": 0.3192, + "step": 300 + }, + { + "epoch": 1.3, + "grad_norm": 3.6251447200775146, + "learning_rate": 0.00039096007918178817, + "loss": 0.349, + "step": 325 + }, + { + "epoch": 1.4, + "grad_norm": 2.854794979095459, + "learning_rate": 0.0003893104585945233, + "loss": 0.3274, + "step": 350 + }, + { + "epoch": 1.5, + "grad_norm": 4.805097579956055, + "learning_rate": 0.0003876608380072583, + "loss": 0.3279, + "step": 375 + }, + { + "epoch": 1.6, + "grad_norm": 2.9891581535339355, + "learning_rate": 0.00038601121741999343, + "loss": 0.3166, + "step": 400 + }, + { + "epoch": 1.7, + "grad_norm": 5.333752155303955, + "learning_rate": 0.0003843615968327285, + "loss": 0.2856, + "step": 425 + }, + { + "epoch": 1.8, + "grad_norm": 2.615621566772461, + "learning_rate": 0.0003827119762454636, + "loss": 0.3286, + "step": 450 + }, + { + "epoch": 1.9, + "grad_norm": 2.4228415489196777, + "learning_rate": 0.00038106235565819863, + "loss": 0.2812, + "step": 475 + }, + { + "epoch": 2.0, + "grad_norm": 3.6015517711639404, + "learning_rate": 0.0003794127350709337, + "loss": 0.2884, + "step": 500 + }, + { + "epoch": 2.1, + "grad_norm": 2.7867860794067383, + "learning_rate": 0.0003777631144836688, + "loss": 0.2582, + "step": 525 + }, + { + "epoch": 2.2, + "grad_norm": 10.079645156860352, + "learning_rate": 0.00037611349389640383, + "loss": 0.245, + "step": 550 + }, + { + "epoch": 2.3, + "grad_norm": 2.475917100906372, + "learning_rate": 0.00037446387330913894, + "loss": 0.2615, + "step": 575 + }, + { + "epoch": 2.4, + "grad_norm": 8.098592758178711, + "learning_rate": 0.000372814252721874, + "loss": 0.2484, + "step": 600 + }, + { + "epoch": 2.5, + "grad_norm": 2.318723201751709, + "learning_rate": 0.00037116463213460903, + "loss": 0.2056, + "step": 625 + }, + { + "epoch": 2.6, + "grad_norm": 3.616283893585205, + "learning_rate": 0.00036951501154734414, + "loss": 0.2358, + "step": 650 + }, + { + "epoch": 2.7, + "grad_norm": 6.419492721557617, + "learning_rate": 0.0003678653909600792, + "loss": 0.2204, + "step": 675 + }, + { + "epoch": 2.8, + "grad_norm": 2.9333388805389404, + "learning_rate": 0.0003662157703728143, + "loss": 0.1857, + "step": 700 + }, + { + "epoch": 2.9, + "grad_norm": 7.691112995147705, + "learning_rate": 0.00036456614978554934, + "loss": 0.2054, + "step": 725 + }, + { + "epoch": 3.0, + "grad_norm": 4.908492565155029, + "learning_rate": 0.00036291652919828444, + "loss": 0.1956, + "step": 750 + }, + { + "epoch": 3.1, + "grad_norm": 1.8923287391662598, + "learning_rate": 0.00036126690861101944, + "loss": 0.1708, + "step": 775 + }, + { + "epoch": 3.2, + "grad_norm": 4.403504848480225, + "learning_rate": 0.00035961728802375454, + "loss": 0.1441, + "step": 800 + }, + { + "epoch": 3.3, + "grad_norm": 4.117386817932129, + "learning_rate": 0.0003579676674364896, + "loss": 0.1586, + "step": 825 + }, + { + "epoch": 3.4, + "grad_norm": 7.119816303253174, + "learning_rate": 0.0003563180468492247, + "loss": 0.1472, + "step": 850 + }, + { + "epoch": 3.5, + "grad_norm": 3.5796430110931396, + "learning_rate": 0.0003546684262619598, + "loss": 0.1802, + "step": 875 + }, + { + "epoch": 3.6, + "grad_norm": 2.97688889503479, + "learning_rate": 0.00035301880567469485, + "loss": 0.1578, + "step": 900 + }, + { + "epoch": 3.7, + "grad_norm": 3.716148614883423, + "learning_rate": 0.0003513691850874299, + "loss": 0.1662, + "step": 925 + }, + { + "epoch": 3.8, + "grad_norm": 6.1249566078186035, + "learning_rate": 0.00034971956450016495, + "loss": 0.159, + "step": 950 + }, + { + "epoch": 3.9, + "grad_norm": 3.0592427253723145, + "learning_rate": 0.00034806994391290005, + "loss": 0.148, + "step": 975 + }, + { + "epoch": 4.0, + "grad_norm": 4.467265605926514, + "learning_rate": 0.0003464203233256351, + "loss": 0.1442, + "step": 1000 + }, + { + "epoch": 4.1, + "grad_norm": 2.7223546504974365, + "learning_rate": 0.0003447707027383702, + "loss": 0.1339, + "step": 1025 + }, + { + "epoch": 4.2, + "grad_norm": 3.698854923248291, + "learning_rate": 0.00034312108215110525, + "loss": 0.1381, + "step": 1050 + }, + { + "epoch": 4.3, + "grad_norm": 7.418384552001953, + "learning_rate": 0.0003414714615638403, + "loss": 0.1158, + "step": 1075 + }, + { + "epoch": 4.4, + "grad_norm": 1.2887814044952393, + "learning_rate": 0.0003398218409765754, + "loss": 0.0855, + "step": 1100 + }, + { + "epoch": 4.5, + "grad_norm": 0.838731586933136, + "learning_rate": 0.00033817222038931045, + "loss": 0.1272, + "step": 1125 + }, + { + "epoch": 4.6, + "grad_norm": 5.912592887878418, + "learning_rate": 0.00033652259980204556, + "loss": 0.1156, + "step": 1150 + }, + { + "epoch": 4.7, + "grad_norm": 8.49521255493164, + "learning_rate": 0.0003348729792147806, + "loss": 0.1182, + "step": 1175 + }, + { + "epoch": 4.8, + "grad_norm": 10.278315544128418, + "learning_rate": 0.0003332233586275157, + "loss": 0.1071, + "step": 1200 + }, + { + "epoch": 4.9, + "grad_norm": 0.724703311920166, + "learning_rate": 0.00033157373804025076, + "loss": 0.11, + "step": 1225 + }, + { + "epoch": 5.0, + "grad_norm": 1.6808199882507324, + "learning_rate": 0.0003299241174529858, + "loss": 0.1101, + "step": 1250 + }, + { + "epoch": 5.1, + "grad_norm": 0.1038585901260376, + "learning_rate": 0.0003282744968657209, + "loss": 0.0791, + "step": 1275 + }, + { + "epoch": 5.2, + "grad_norm": 0.15866787731647491, + "learning_rate": 0.00032662487627845596, + "loss": 0.0638, + "step": 1300 + }, + { + "epoch": 5.3, + "grad_norm": 6.748287200927734, + "learning_rate": 0.00032497525569119106, + "loss": 0.0961, + "step": 1325 + }, + { + "epoch": 5.4, + "grad_norm": 1.5960336923599243, + "learning_rate": 0.0003233256351039261, + "loss": 0.1057, + "step": 1350 + }, + { + "epoch": 5.5, + "grad_norm": 3.3678526878356934, + "learning_rate": 0.00032167601451666116, + "loss": 0.1092, + "step": 1375 + }, + { + "epoch": 5.6, + "grad_norm": 4.743562698364258, + "learning_rate": 0.00032002639392939627, + "loss": 0.1054, + "step": 1400 + }, + { + "epoch": 5.7, + "grad_norm": 12.860156059265137, + "learning_rate": 0.0003183767733421313, + "loss": 0.1009, + "step": 1425 + }, + { + "epoch": 5.8, + "grad_norm": 0.8820792436599731, + "learning_rate": 0.0003167271527548664, + "loss": 0.1219, + "step": 1450 + }, + { + "epoch": 5.9, + "grad_norm": 2.732572317123413, + "learning_rate": 0.00031507753216760147, + "loss": 0.0682, + "step": 1475 + }, + { + "epoch": 6.0, + "grad_norm": 9.810711860656738, + "learning_rate": 0.00031342791158033657, + "loss": 0.1134, + "step": 1500 + }, + { + "epoch": 6.1, + "grad_norm": 6.699489593505859, + "learning_rate": 0.00031177829099307157, + "loss": 0.0557, + "step": 1525 + }, + { + "epoch": 6.2, + "grad_norm": 14.691123008728027, + "learning_rate": 0.00031012867040580667, + "loss": 0.0599, + "step": 1550 + }, + { + "epoch": 6.3, + "grad_norm": 24.44135093688965, + "learning_rate": 0.0003084790498185417, + "loss": 0.0818, + "step": 1575 + }, + { + "epoch": 6.4, + "grad_norm": 0.033995576202869415, + "learning_rate": 0.0003068294292312768, + "loss": 0.0692, + "step": 1600 + }, + { + "epoch": 6.5, + "grad_norm": 6.722524642944336, + "learning_rate": 0.0003051798086440119, + "loss": 0.0702, + "step": 1625 + }, + { + "epoch": 6.6, + "grad_norm": 2.9519436359405518, + "learning_rate": 0.000303530188056747, + "loss": 0.073, + "step": 1650 + }, + { + "epoch": 6.7, + "grad_norm": 1.5118328332901, + "learning_rate": 0.000301880567469482, + "loss": 0.0926, + "step": 1675 + }, + { + "epoch": 6.8, + "grad_norm": 2.1923201084136963, + "learning_rate": 0.0003002309468822171, + "loss": 0.0821, + "step": 1700 + }, + { + "epoch": 6.9, + "grad_norm": 3.34309458732605, + "learning_rate": 0.0002985813262949522, + "loss": 0.0827, + "step": 1725 + }, + { + "epoch": 7.0, + "grad_norm": 10.289947509765625, + "learning_rate": 0.00029693170570768723, + "loss": 0.1162, + "step": 1750 + }, + { + "epoch": 7.1, + "grad_norm": 0.051936667412519455, + "learning_rate": 0.00029528208512042233, + "loss": 0.0626, + "step": 1775 + }, + { + "epoch": 7.2, + "grad_norm": 1.8913339376449585, + "learning_rate": 0.0002936324645331574, + "loss": 0.0657, + "step": 1800 + }, + { + "epoch": 7.3, + "grad_norm": 15.073695182800293, + "learning_rate": 0.00029198284394589243, + "loss": 0.0539, + "step": 1825 + }, + { + "epoch": 7.4, + "grad_norm": 4.514949798583984, + "learning_rate": 0.00029033322335862753, + "loss": 0.056, + "step": 1850 + }, + { + "epoch": 7.5, + "grad_norm": 0.46094000339508057, + "learning_rate": 0.0002886836027713626, + "loss": 0.0667, + "step": 1875 + }, + { + "epoch": 7.6, + "grad_norm": 4.682873725891113, + "learning_rate": 0.0002870339821840977, + "loss": 0.0673, + "step": 1900 + }, + { + "epoch": 7.7, + "grad_norm": 3.581599235534668, + "learning_rate": 0.00028538436159683273, + "loss": 0.0699, + "step": 1925 + }, + { + "epoch": 7.8, + "grad_norm": 3.2518703937530518, + "learning_rate": 0.00028373474100956784, + "loss": 0.0628, + "step": 1950 + }, + { + "epoch": 7.9, + "grad_norm": 0.24071219563484192, + "learning_rate": 0.0002820851204223029, + "loss": 0.0747, + "step": 1975 + }, + { + "epoch": 8.0, + "grad_norm": 2.8545596599578857, + "learning_rate": 0.00028043549983503794, + "loss": 0.0688, + "step": 2000 + }, + { + "epoch": 8.1, + "grad_norm": 0.049499694257974625, + "learning_rate": 0.00027878587924777304, + "loss": 0.0563, + "step": 2025 + }, + { + "epoch": 8.2, + "grad_norm": 22.112478256225586, + "learning_rate": 0.0002771362586605081, + "loss": 0.0545, + "step": 2050 + }, + { + "epoch": 8.3, + "grad_norm": 20.87473487854004, + "learning_rate": 0.0002754866380732432, + "loss": 0.0256, + "step": 2075 + }, + { + "epoch": 8.4, + "grad_norm": 2.317768096923828, + "learning_rate": 0.00027383701748597824, + "loss": 0.0753, + "step": 2100 + }, + { + "epoch": 8.5, + "grad_norm": 20.241336822509766, + "learning_rate": 0.0002721873968987133, + "loss": 0.0257, + "step": 2125 + }, + { + "epoch": 8.6, + "grad_norm": 14.994575500488281, + "learning_rate": 0.00027053777631144834, + "loss": 0.0748, + "step": 2150 + }, + { + "epoch": 8.7, + "grad_norm": 0.8268639445304871, + "learning_rate": 0.00026888815572418344, + "loss": 0.0805, + "step": 2175 + }, + { + "epoch": 8.8, + "grad_norm": 0.031071001663804054, + "learning_rate": 0.00026723853513691855, + "loss": 0.0603, + "step": 2200 + }, + { + "epoch": 8.9, + "grad_norm": 2.4572932720184326, + "learning_rate": 0.0002655889145496536, + "loss": 0.071, + "step": 2225 + }, + { + "epoch": 9.0, + "grad_norm": 0.21708321571350098, + "learning_rate": 0.0002639392939623887, + "loss": 0.0606, + "step": 2250 + }, + { + "epoch": 9.1, + "grad_norm": 14.928559303283691, + "learning_rate": 0.0002622896733751237, + "loss": 0.0443, + "step": 2275 + }, + { + "epoch": 9.2, + "grad_norm": 1.108826994895935, + "learning_rate": 0.0002606400527878588, + "loss": 0.0509, + "step": 2300 + }, + { + "epoch": 9.3, + "grad_norm": 0.05721910670399666, + "learning_rate": 0.00025899043220059385, + "loss": 0.0459, + "step": 2325 + }, + { + "epoch": 9.4, + "grad_norm": 7.81903600692749, + "learning_rate": 0.00025734081161332895, + "loss": 0.0535, + "step": 2350 + }, + { + "epoch": 9.5, + "grad_norm": 0.01475981343537569, + "learning_rate": 0.000255691191026064, + "loss": 0.0399, + "step": 2375 + }, + { + "epoch": 9.6, + "grad_norm": 5.496493339538574, + "learning_rate": 0.0002540415704387991, + "loss": 0.0744, + "step": 2400 + }, + { + "epoch": 9.7, + "grad_norm": 0.05929339677095413, + "learning_rate": 0.00025239194985153415, + "loss": 0.0416, + "step": 2425 + }, + { + "epoch": 9.8, + "grad_norm": 0.06961135566234589, + "learning_rate": 0.0002507423292642692, + "loss": 0.04, + "step": 2450 + }, + { + "epoch": 9.9, + "grad_norm": 0.02165246568620205, + "learning_rate": 0.0002490927086770043, + "loss": 0.0984, + "step": 2475 + }, + { + "epoch": 10.0, + "grad_norm": 25.155675888061523, + "learning_rate": 0.00024744308808973936, + "loss": 0.042, + "step": 2500 + }, + { + "epoch": 10.1, + "grad_norm": 8.566617965698242, + "learning_rate": 0.00024579346750247446, + "loss": 0.0385, + "step": 2525 + }, + { + "epoch": 10.2, + "grad_norm": 0.024992404505610466, + "learning_rate": 0.0002441438469152095, + "loss": 0.0327, + "step": 2550 + }, + { + "epoch": 10.3, + "grad_norm": 0.22233247756958008, + "learning_rate": 0.00024249422632794456, + "loss": 0.0211, + "step": 2575 + }, + { + "epoch": 10.4, + "grad_norm": 13.307366371154785, + "learning_rate": 0.00024084460574067963, + "loss": 0.0472, + "step": 2600 + }, + { + "epoch": 10.5, + "grad_norm": 0.17787696421146393, + "learning_rate": 0.0002391949851534147, + "loss": 0.0346, + "step": 2625 + }, + { + "epoch": 10.6, + "grad_norm": 1.2845549583435059, + "learning_rate": 0.0002375453645661498, + "loss": 0.0587, + "step": 2650 + }, + { + "epoch": 10.7, + "grad_norm": 0.0603482760488987, + "learning_rate": 0.00023589574397888486, + "loss": 0.0394, + "step": 2675 + }, + { + "epoch": 10.8, + "grad_norm": 0.031819943338632584, + "learning_rate": 0.00023424612339161997, + "loss": 0.0434, + "step": 2700 + }, + { + "epoch": 10.9, + "grad_norm": 1.5738537311553955, + "learning_rate": 0.000232596502804355, + "loss": 0.0393, + "step": 2725 + }, + { + "epoch": 11.0, + "grad_norm": 0.6862583756446838, + "learning_rate": 0.00023094688221709007, + "loss": 0.0183, + "step": 2750 + }, + { + "epoch": 11.1, + "grad_norm": 0.0053489054553210735, + "learning_rate": 0.00022929726162982514, + "loss": 0.0277, + "step": 2775 + }, + { + "epoch": 11.2, + "grad_norm": 0.04365606606006622, + "learning_rate": 0.00022764764104256022, + "loss": 0.0337, + "step": 2800 + }, + { + "epoch": 11.3, + "grad_norm": 1.3348641395568848, + "learning_rate": 0.0002259980204552953, + "loss": 0.0219, + "step": 2825 + }, + { + "epoch": 11.4, + "grad_norm": 3.3695547580718994, + "learning_rate": 0.00022434839986803037, + "loss": 0.0227, + "step": 2850 + }, + { + "epoch": 11.5, + "grad_norm": 2.507455587387085, + "learning_rate": 0.00022269877928076542, + "loss": 0.0443, + "step": 2875 + }, + { + "epoch": 11.6, + "grad_norm": 0.013451912440359592, + "learning_rate": 0.0002210491586935005, + "loss": 0.0486, + "step": 2900 + }, + { + "epoch": 11.7, + "grad_norm": 0.031177405267953873, + "learning_rate": 0.00021939953810623557, + "loss": 0.0511, + "step": 2925 + }, + { + "epoch": 11.8, + "grad_norm": 0.02148960903286934, + "learning_rate": 0.00021774991751897065, + "loss": 0.0269, + "step": 2950 + }, + { + "epoch": 11.9, + "grad_norm": 0.01219563465565443, + "learning_rate": 0.00021610029693170573, + "loss": 0.042, + "step": 2975 + }, + { + "epoch": 12.0, + "grad_norm": 0.2062515765428543, + "learning_rate": 0.0002144506763444408, + "loss": 0.031, + "step": 3000 + }, + { + "epoch": 12.1, + "grad_norm": 3.235403060913086, + "learning_rate": 0.00021280105575717582, + "loss": 0.0414, + "step": 3025 + }, + { + "epoch": 12.2, + "grad_norm": 1.089821696281433, + "learning_rate": 0.00021115143516991093, + "loss": 0.0692, + "step": 3050 + }, + { + "epoch": 12.3, + "grad_norm": 0.013664484024047852, + "learning_rate": 0.000209501814582646, + "loss": 0.0367, + "step": 3075 + }, + { + "epoch": 12.4, + "grad_norm": 1.3101387023925781, + "learning_rate": 0.00020785219399538108, + "loss": 0.0277, + "step": 3100 + }, + { + "epoch": 12.5, + "grad_norm": 3.244853973388672, + "learning_rate": 0.00020620257340811616, + "loss": 0.0373, + "step": 3125 + }, + { + "epoch": 12.6, + "grad_norm": 0.013548053801059723, + "learning_rate": 0.00020455295282085123, + "loss": 0.027, + "step": 3150 + }, + { + "epoch": 12.7, + "grad_norm": 0.004670978989452124, + "learning_rate": 0.00020290333223358626, + "loss": 0.0252, + "step": 3175 + }, + { + "epoch": 12.8, + "grad_norm": 0.007080752402544022, + "learning_rate": 0.00020125371164632133, + "loss": 0.0187, + "step": 3200 + }, + { + "epoch": 12.9, + "grad_norm": 0.005918944254517555, + "learning_rate": 0.0001996040910590564, + "loss": 0.0423, + "step": 3225 + }, + { + "epoch": 13.0, + "grad_norm": 4.530269145965576, + "learning_rate": 0.00019795447047179148, + "loss": 0.0381, + "step": 3250 + }, + { + "epoch": 13.1, + "grad_norm": 0.00533739197999239, + "learning_rate": 0.0001963048498845266, + "loss": 0.0097, + "step": 3275 + }, + { + "epoch": 13.2, + "grad_norm": 0.7641226649284363, + "learning_rate": 0.00019465522929726164, + "loss": 0.0205, + "step": 3300 + }, + { + "epoch": 13.3, + "grad_norm": 2.16174054145813, + "learning_rate": 0.00019300560870999671, + "loss": 0.0174, + "step": 3325 + }, + { + "epoch": 13.4, + "grad_norm": 0.00814723875373602, + "learning_rate": 0.0001913559881227318, + "loss": 0.0215, + "step": 3350 + }, + { + "epoch": 13.5, + "grad_norm": 10.500170707702637, + "learning_rate": 0.00018970636753546684, + "loss": 0.0226, + "step": 3375 + }, + { + "epoch": 13.6, + "grad_norm": 2.612359046936035, + "learning_rate": 0.00018805674694820192, + "loss": 0.0314, + "step": 3400 + }, + { + "epoch": 13.7, + "grad_norm": 0.11265736818313599, + "learning_rate": 0.000186407126360937, + "loss": 0.0234, + "step": 3425 + }, + { + "epoch": 13.8, + "grad_norm": 0.06453213095664978, + "learning_rate": 0.00018475750577367207, + "loss": 0.0314, + "step": 3450 + }, + { + "epoch": 13.9, + "grad_norm": 0.01650763861835003, + "learning_rate": 0.00018310788518640715, + "loss": 0.0403, + "step": 3475 + }, + { + "epoch": 14.0, + "grad_norm": 0.002764373552054167, + "learning_rate": 0.00018145826459914222, + "loss": 0.0137, + "step": 3500 + }, + { + "epoch": 14.1, + "grad_norm": 0.0022430066019296646, + "learning_rate": 0.00017980864401187727, + "loss": 0.0084, + "step": 3525 + }, + { + "epoch": 14.2, + "grad_norm": 0.007267679553478956, + "learning_rate": 0.00017815902342461235, + "loss": 0.0161, + "step": 3550 + }, + { + "epoch": 14.3, + "grad_norm": 0.004844362381845713, + "learning_rate": 0.00017650940283734742, + "loss": 0.0399, + "step": 3575 + }, + { + "epoch": 14.4, + "grad_norm": 0.05529040843248367, + "learning_rate": 0.00017485978225008247, + "loss": 0.0192, + "step": 3600 + }, + { + "epoch": 14.5, + "grad_norm": 0.008248478174209595, + "learning_rate": 0.00017321016166281755, + "loss": 0.0079, + "step": 3625 + }, + { + "epoch": 14.6, + "grad_norm": 0.02617962658405304, + "learning_rate": 0.00017156054107555263, + "loss": 0.0244, + "step": 3650 + }, + { + "epoch": 14.7, + "grad_norm": 0.004497932270169258, + "learning_rate": 0.0001699109204882877, + "loss": 0.0464, + "step": 3675 + }, + { + "epoch": 14.8, + "grad_norm": 0.008586671203374863, + "learning_rate": 0.00016826129990102278, + "loss": 0.0354, + "step": 3700 + }, + { + "epoch": 14.9, + "grad_norm": 0.30177807807922363, + "learning_rate": 0.00016661167931375785, + "loss": 0.0147, + "step": 3725 + }, + { + "epoch": 15.0, + "grad_norm": 0.10390808433294296, + "learning_rate": 0.0001649620587264929, + "loss": 0.0078, + "step": 3750 + }, + { + "epoch": 15.1, + "grad_norm": 1.584868311882019, + "learning_rate": 0.00016331243813922798, + "loss": 0.0059, + "step": 3775 + }, + { + "epoch": 15.2, + "grad_norm": 0.0038078685756772757, + "learning_rate": 0.00016166281755196306, + "loss": 0.0143, + "step": 3800 + }, + { + "epoch": 15.3, + "grad_norm": 0.017795555293560028, + "learning_rate": 0.00016001319696469813, + "loss": 0.0003, + "step": 3825 + }, + { + "epoch": 15.4, + "grad_norm": 7.246474742889404, + "learning_rate": 0.0001583635763774332, + "loss": 0.0288, + "step": 3850 + }, + { + "epoch": 15.5, + "grad_norm": 0.0014716258738189936, + "learning_rate": 0.00015671395579016829, + "loss": 0.0113, + "step": 3875 + }, + { + "epoch": 15.6, + "grad_norm": 0.002159764291718602, + "learning_rate": 0.00015506433520290334, + "loss": 0.0138, + "step": 3900 + }, + { + "epoch": 15.7, + "grad_norm": 27.16730308532715, + "learning_rate": 0.0001534147146156384, + "loss": 0.0048, + "step": 3925 + }, + { + "epoch": 15.8, + "grad_norm": 0.004570333287119865, + "learning_rate": 0.0001517650940283735, + "loss": 0.0144, + "step": 3950 + }, + { + "epoch": 15.9, + "grad_norm": 0.008330491371452808, + "learning_rate": 0.00015011547344110854, + "loss": 0.0371, + "step": 3975 + }, + { + "epoch": 16.0, + "grad_norm": 15.75979232788086, + "learning_rate": 0.00014846585285384361, + "loss": 0.0207, + "step": 4000 + } + ], + "logging_steps": 25, + "max_steps": 6250, + "num_input_tokens_seen": 0, + "num_train_epochs": 25, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 203277926400000.0, + "train_batch_size": 20, + "trial_name": null, + "trial_params": null +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-4000/training_args.bin b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-4000/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..ef3fc22e099df49861bca49d15da2b959cfcb6df --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-4000/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ec1d8ed36c45af65d67a23de6c5216a486932bfb09b720e1a4417b82eaaa215d +size 5304 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-4000/vocab.txt b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-4000/vocab.txt new file mode 100644 index 0000000000000000000000000000000000000000..cc9ad53732e21ea9180f18e1fb67a38cc5702d4c --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-4000/vocab.txt @@ -0,0 +1,4101 @@ +[PAD] +[UNK] +[CLS] +[SEP] +[MASK] +AAAAAA +AAAAAT +AAAAAC +AAAAAG +AAAATA +AAAATT +AAAATC +AAAATG +AAAACA +AAAACT +AAAACC +AAAACG +AAAAGA +AAAAGT +AAAAGC +AAAAGG +AAATAA +AAATAT +AAATAC +AAATAG +AAATTA +AAATTT +AAATTC +AAATTG +AAATCA +AAATCT +AAATCC +AAATCG +AAATGA +AAATGT +AAATGC +AAATGG +AAACAA +AAACAT +AAACAC +AAACAG +AAACTA +AAACTT +AAACTC +AAACTG +AAACCA +AAACCT +AAACCC +AAACCG +AAACGA +AAACGT +AAACGC +AAACGG +AAAGAA +AAAGAT +AAAGAC +AAAGAG +AAAGTA +AAAGTT +AAAGTC +AAAGTG +AAAGCA +AAAGCT +AAAGCC +AAAGCG +AAAGGA +AAAGGT +AAAGGC +AAAGGG +AATAAA +AATAAT +AATAAC +AATAAG +AATATA +AATATT +AATATC +AATATG +AATACA +AATACT +AATACC +AATACG +AATAGA +AATAGT +AATAGC +AATAGG +AATTAA +AATTAT +AATTAC +AATTAG +AATTTA +AATTTT +AATTTC +AATTTG +AATTCA +AATTCT +AATTCC +AATTCG +AATTGA +AATTGT +AATTGC +AATTGG +AATCAA +AATCAT +AATCAC +AATCAG +AATCTA +AATCTT +AATCTC +AATCTG +AATCCA +AATCCT +AATCCC +AATCCG +AATCGA +AATCGT +AATCGC +AATCGG +AATGAA +AATGAT +AATGAC +AATGAG +AATGTA +AATGTT +AATGTC +AATGTG +AATGCA +AATGCT +AATGCC +AATGCG +AATGGA +AATGGT +AATGGC +AATGGG +AACAAA +AACAAT +AACAAC +AACAAG +AACATA +AACATT +AACATC +AACATG +AACACA +AACACT +AACACC +AACACG +AACAGA +AACAGT +AACAGC +AACAGG +AACTAA +AACTAT +AACTAC +AACTAG +AACTTA +AACTTT +AACTTC +AACTTG +AACTCA +AACTCT +AACTCC +AACTCG +AACTGA +AACTGT +AACTGC +AACTGG +AACCAA +AACCAT +AACCAC +AACCAG +AACCTA +AACCTT +AACCTC +AACCTG +AACCCA +AACCCT +AACCCC +AACCCG +AACCGA +AACCGT +AACCGC +AACCGG +AACGAA +AACGAT +AACGAC +AACGAG +AACGTA +AACGTT +AACGTC +AACGTG +AACGCA +AACGCT +AACGCC +AACGCG +AACGGA +AACGGT +AACGGC +AACGGG +AAGAAA +AAGAAT +AAGAAC +AAGAAG +AAGATA +AAGATT +AAGATC +AAGATG +AAGACA +AAGACT +AAGACC +AAGACG +AAGAGA +AAGAGT +AAGAGC +AAGAGG +AAGTAA +AAGTAT +AAGTAC +AAGTAG +AAGTTA +AAGTTT +AAGTTC +AAGTTG +AAGTCA +AAGTCT +AAGTCC +AAGTCG +AAGTGA +AAGTGT +AAGTGC +AAGTGG +AAGCAA +AAGCAT +AAGCAC +AAGCAG +AAGCTA +AAGCTT +AAGCTC +AAGCTG +AAGCCA +AAGCCT +AAGCCC +AAGCCG +AAGCGA +AAGCGT +AAGCGC +AAGCGG +AAGGAA +AAGGAT +AAGGAC +AAGGAG +AAGGTA +AAGGTT +AAGGTC +AAGGTG +AAGGCA +AAGGCT +AAGGCC +AAGGCG +AAGGGA +AAGGGT +AAGGGC +AAGGGG +ATAAAA +ATAAAT +ATAAAC +ATAAAG +ATAATA +ATAATT +ATAATC +ATAATG +ATAACA +ATAACT +ATAACC +ATAACG +ATAAGA +ATAAGT +ATAAGC +ATAAGG +ATATAA +ATATAT +ATATAC +ATATAG +ATATTA +ATATTT +ATATTC +ATATTG +ATATCA +ATATCT +ATATCC +ATATCG +ATATGA +ATATGT +ATATGC +ATATGG +ATACAA +ATACAT +ATACAC +ATACAG +ATACTA +ATACTT +ATACTC +ATACTG +ATACCA +ATACCT +ATACCC +ATACCG +ATACGA +ATACGT +ATACGC +ATACGG +ATAGAA +ATAGAT +ATAGAC +ATAGAG +ATAGTA +ATAGTT +ATAGTC +ATAGTG +ATAGCA +ATAGCT +ATAGCC +ATAGCG +ATAGGA +ATAGGT +ATAGGC +ATAGGG +ATTAAA +ATTAAT +ATTAAC +ATTAAG +ATTATA +ATTATT +ATTATC +ATTATG +ATTACA +ATTACT +ATTACC +ATTACG +ATTAGA +ATTAGT +ATTAGC +ATTAGG +ATTTAA +ATTTAT +ATTTAC +ATTTAG +ATTTTA +ATTTTT +ATTTTC +ATTTTG +ATTTCA +ATTTCT +ATTTCC +ATTTCG +ATTTGA +ATTTGT +ATTTGC +ATTTGG +ATTCAA +ATTCAT +ATTCAC +ATTCAG +ATTCTA +ATTCTT +ATTCTC +ATTCTG +ATTCCA +ATTCCT +ATTCCC +ATTCCG +ATTCGA +ATTCGT +ATTCGC +ATTCGG +ATTGAA +ATTGAT +ATTGAC +ATTGAG +ATTGTA +ATTGTT +ATTGTC +ATTGTG +ATTGCA +ATTGCT +ATTGCC +ATTGCG +ATTGGA +ATTGGT +ATTGGC +ATTGGG +ATCAAA +ATCAAT +ATCAAC +ATCAAG +ATCATA +ATCATT +ATCATC +ATCATG +ATCACA +ATCACT +ATCACC +ATCACG +ATCAGA +ATCAGT +ATCAGC +ATCAGG +ATCTAA +ATCTAT +ATCTAC +ATCTAG +ATCTTA +ATCTTT +ATCTTC +ATCTTG +ATCTCA +ATCTCT +ATCTCC +ATCTCG +ATCTGA +ATCTGT +ATCTGC +ATCTGG +ATCCAA +ATCCAT +ATCCAC +ATCCAG +ATCCTA +ATCCTT +ATCCTC +ATCCTG +ATCCCA +ATCCCT +ATCCCC +ATCCCG +ATCCGA +ATCCGT +ATCCGC +ATCCGG +ATCGAA +ATCGAT +ATCGAC +ATCGAG +ATCGTA +ATCGTT +ATCGTC +ATCGTG +ATCGCA +ATCGCT +ATCGCC +ATCGCG +ATCGGA +ATCGGT +ATCGGC +ATCGGG +ATGAAA +ATGAAT +ATGAAC +ATGAAG +ATGATA +ATGATT +ATGATC +ATGATG +ATGACA +ATGACT +ATGACC +ATGACG +ATGAGA +ATGAGT +ATGAGC +ATGAGG +ATGTAA +ATGTAT +ATGTAC +ATGTAG +ATGTTA +ATGTTT +ATGTTC +ATGTTG +ATGTCA +ATGTCT +ATGTCC +ATGTCG +ATGTGA +ATGTGT +ATGTGC +ATGTGG +ATGCAA +ATGCAT +ATGCAC +ATGCAG +ATGCTA +ATGCTT +ATGCTC +ATGCTG +ATGCCA +ATGCCT +ATGCCC +ATGCCG +ATGCGA +ATGCGT +ATGCGC +ATGCGG +ATGGAA +ATGGAT +ATGGAC +ATGGAG +ATGGTA +ATGGTT +ATGGTC +ATGGTG +ATGGCA +ATGGCT +ATGGCC +ATGGCG +ATGGGA +ATGGGT +ATGGGC +ATGGGG +ACAAAA +ACAAAT +ACAAAC +ACAAAG +ACAATA +ACAATT +ACAATC +ACAATG +ACAACA +ACAACT +ACAACC +ACAACG +ACAAGA +ACAAGT +ACAAGC +ACAAGG +ACATAA +ACATAT +ACATAC +ACATAG +ACATTA +ACATTT +ACATTC +ACATTG +ACATCA +ACATCT +ACATCC +ACATCG +ACATGA +ACATGT +ACATGC +ACATGG +ACACAA +ACACAT +ACACAC +ACACAG +ACACTA +ACACTT +ACACTC +ACACTG +ACACCA +ACACCT +ACACCC +ACACCG +ACACGA +ACACGT +ACACGC +ACACGG +ACAGAA +ACAGAT +ACAGAC +ACAGAG +ACAGTA +ACAGTT +ACAGTC +ACAGTG +ACAGCA +ACAGCT +ACAGCC +ACAGCG +ACAGGA +ACAGGT +ACAGGC +ACAGGG +ACTAAA +ACTAAT +ACTAAC +ACTAAG +ACTATA +ACTATT +ACTATC +ACTATG +ACTACA +ACTACT +ACTACC +ACTACG +ACTAGA +ACTAGT +ACTAGC +ACTAGG +ACTTAA +ACTTAT +ACTTAC +ACTTAG +ACTTTA +ACTTTT +ACTTTC +ACTTTG +ACTTCA +ACTTCT +ACTTCC +ACTTCG +ACTTGA +ACTTGT +ACTTGC +ACTTGG +ACTCAA +ACTCAT +ACTCAC +ACTCAG +ACTCTA +ACTCTT +ACTCTC +ACTCTG +ACTCCA +ACTCCT +ACTCCC +ACTCCG +ACTCGA +ACTCGT +ACTCGC +ACTCGG +ACTGAA +ACTGAT +ACTGAC +ACTGAG +ACTGTA +ACTGTT +ACTGTC +ACTGTG +ACTGCA +ACTGCT +ACTGCC +ACTGCG +ACTGGA +ACTGGT +ACTGGC +ACTGGG +ACCAAA +ACCAAT +ACCAAC +ACCAAG +ACCATA +ACCATT +ACCATC +ACCATG +ACCACA +ACCACT +ACCACC +ACCACG +ACCAGA +ACCAGT +ACCAGC +ACCAGG +ACCTAA +ACCTAT +ACCTAC +ACCTAG +ACCTTA +ACCTTT +ACCTTC +ACCTTG +ACCTCA +ACCTCT +ACCTCC +ACCTCG +ACCTGA +ACCTGT +ACCTGC +ACCTGG +ACCCAA +ACCCAT +ACCCAC +ACCCAG +ACCCTA +ACCCTT +ACCCTC +ACCCTG +ACCCCA +ACCCCT +ACCCCC +ACCCCG +ACCCGA +ACCCGT +ACCCGC +ACCCGG +ACCGAA +ACCGAT +ACCGAC +ACCGAG +ACCGTA +ACCGTT +ACCGTC +ACCGTG +ACCGCA +ACCGCT +ACCGCC +ACCGCG +ACCGGA +ACCGGT +ACCGGC +ACCGGG +ACGAAA +ACGAAT +ACGAAC +ACGAAG +ACGATA +ACGATT +ACGATC +ACGATG +ACGACA +ACGACT +ACGACC +ACGACG +ACGAGA +ACGAGT +ACGAGC +ACGAGG +ACGTAA +ACGTAT +ACGTAC +ACGTAG +ACGTTA +ACGTTT +ACGTTC +ACGTTG +ACGTCA +ACGTCT +ACGTCC +ACGTCG +ACGTGA +ACGTGT +ACGTGC +ACGTGG +ACGCAA +ACGCAT +ACGCAC +ACGCAG +ACGCTA +ACGCTT +ACGCTC +ACGCTG +ACGCCA +ACGCCT +ACGCCC +ACGCCG +ACGCGA +ACGCGT +ACGCGC +ACGCGG +ACGGAA +ACGGAT +ACGGAC +ACGGAG +ACGGTA +ACGGTT +ACGGTC +ACGGTG +ACGGCA +ACGGCT +ACGGCC +ACGGCG +ACGGGA +ACGGGT +ACGGGC +ACGGGG +AGAAAA +AGAAAT +AGAAAC +AGAAAG +AGAATA +AGAATT +AGAATC +AGAATG +AGAACA +AGAACT +AGAACC +AGAACG +AGAAGA +AGAAGT +AGAAGC +AGAAGG +AGATAA +AGATAT +AGATAC +AGATAG +AGATTA +AGATTT +AGATTC +AGATTG +AGATCA +AGATCT +AGATCC +AGATCG +AGATGA +AGATGT +AGATGC +AGATGG +AGACAA +AGACAT +AGACAC +AGACAG +AGACTA +AGACTT +AGACTC +AGACTG +AGACCA +AGACCT +AGACCC +AGACCG +AGACGA +AGACGT +AGACGC +AGACGG +AGAGAA +AGAGAT +AGAGAC +AGAGAG +AGAGTA +AGAGTT +AGAGTC +AGAGTG +AGAGCA +AGAGCT +AGAGCC +AGAGCG +AGAGGA +AGAGGT +AGAGGC +AGAGGG +AGTAAA +AGTAAT +AGTAAC +AGTAAG +AGTATA +AGTATT +AGTATC +AGTATG +AGTACA +AGTACT +AGTACC +AGTACG +AGTAGA +AGTAGT +AGTAGC +AGTAGG +AGTTAA +AGTTAT +AGTTAC +AGTTAG +AGTTTA +AGTTTT +AGTTTC +AGTTTG +AGTTCA +AGTTCT +AGTTCC +AGTTCG +AGTTGA +AGTTGT +AGTTGC +AGTTGG +AGTCAA +AGTCAT +AGTCAC +AGTCAG +AGTCTA +AGTCTT +AGTCTC +AGTCTG +AGTCCA +AGTCCT +AGTCCC +AGTCCG +AGTCGA +AGTCGT +AGTCGC +AGTCGG +AGTGAA +AGTGAT +AGTGAC +AGTGAG +AGTGTA +AGTGTT +AGTGTC +AGTGTG +AGTGCA +AGTGCT +AGTGCC +AGTGCG +AGTGGA +AGTGGT +AGTGGC +AGTGGG +AGCAAA +AGCAAT +AGCAAC +AGCAAG +AGCATA +AGCATT +AGCATC +AGCATG +AGCACA +AGCACT +AGCACC +AGCACG +AGCAGA +AGCAGT +AGCAGC +AGCAGG +AGCTAA +AGCTAT +AGCTAC +AGCTAG +AGCTTA +AGCTTT +AGCTTC +AGCTTG +AGCTCA +AGCTCT +AGCTCC +AGCTCG +AGCTGA +AGCTGT +AGCTGC +AGCTGG +AGCCAA +AGCCAT +AGCCAC +AGCCAG +AGCCTA +AGCCTT +AGCCTC +AGCCTG +AGCCCA +AGCCCT +AGCCCC +AGCCCG +AGCCGA +AGCCGT +AGCCGC +AGCCGG +AGCGAA +AGCGAT +AGCGAC +AGCGAG +AGCGTA +AGCGTT +AGCGTC +AGCGTG +AGCGCA +AGCGCT +AGCGCC +AGCGCG +AGCGGA +AGCGGT +AGCGGC +AGCGGG +AGGAAA +AGGAAT +AGGAAC +AGGAAG +AGGATA +AGGATT +AGGATC +AGGATG +AGGACA +AGGACT +AGGACC +AGGACG +AGGAGA +AGGAGT +AGGAGC +AGGAGG +AGGTAA +AGGTAT +AGGTAC +AGGTAG +AGGTTA +AGGTTT +AGGTTC +AGGTTG +AGGTCA +AGGTCT +AGGTCC +AGGTCG +AGGTGA +AGGTGT +AGGTGC +AGGTGG +AGGCAA +AGGCAT +AGGCAC +AGGCAG +AGGCTA +AGGCTT +AGGCTC +AGGCTG +AGGCCA +AGGCCT +AGGCCC +AGGCCG +AGGCGA +AGGCGT +AGGCGC +AGGCGG +AGGGAA +AGGGAT +AGGGAC +AGGGAG +AGGGTA +AGGGTT +AGGGTC +AGGGTG +AGGGCA +AGGGCT +AGGGCC +AGGGCG +AGGGGA +AGGGGT +AGGGGC +AGGGGG +TAAAAA +TAAAAT +TAAAAC +TAAAAG +TAAATA +TAAATT +TAAATC +TAAATG +TAAACA +TAAACT +TAAACC +TAAACG +TAAAGA +TAAAGT +TAAAGC +TAAAGG +TAATAA +TAATAT +TAATAC +TAATAG +TAATTA +TAATTT +TAATTC +TAATTG +TAATCA +TAATCT +TAATCC +TAATCG +TAATGA +TAATGT +TAATGC +TAATGG +TAACAA +TAACAT +TAACAC +TAACAG +TAACTA +TAACTT +TAACTC +TAACTG +TAACCA +TAACCT +TAACCC +TAACCG +TAACGA +TAACGT +TAACGC +TAACGG +TAAGAA +TAAGAT +TAAGAC +TAAGAG +TAAGTA +TAAGTT +TAAGTC +TAAGTG +TAAGCA +TAAGCT +TAAGCC +TAAGCG +TAAGGA +TAAGGT +TAAGGC +TAAGGG +TATAAA +TATAAT +TATAAC +TATAAG +TATATA +TATATT +TATATC +TATATG +TATACA +TATACT +TATACC +TATACG +TATAGA +TATAGT +TATAGC +TATAGG +TATTAA +TATTAT +TATTAC +TATTAG +TATTTA +TATTTT +TATTTC +TATTTG +TATTCA +TATTCT +TATTCC +TATTCG +TATTGA +TATTGT +TATTGC +TATTGG +TATCAA +TATCAT +TATCAC +TATCAG +TATCTA +TATCTT +TATCTC +TATCTG +TATCCA +TATCCT +TATCCC +TATCCG +TATCGA +TATCGT +TATCGC +TATCGG +TATGAA +TATGAT +TATGAC +TATGAG +TATGTA +TATGTT +TATGTC +TATGTG +TATGCA +TATGCT +TATGCC +TATGCG +TATGGA +TATGGT +TATGGC +TATGGG +TACAAA +TACAAT +TACAAC +TACAAG +TACATA +TACATT +TACATC +TACATG +TACACA +TACACT +TACACC +TACACG +TACAGA +TACAGT +TACAGC +TACAGG +TACTAA +TACTAT +TACTAC +TACTAG +TACTTA +TACTTT +TACTTC +TACTTG +TACTCA +TACTCT +TACTCC +TACTCG +TACTGA +TACTGT +TACTGC +TACTGG +TACCAA +TACCAT +TACCAC +TACCAG +TACCTA +TACCTT +TACCTC +TACCTG +TACCCA +TACCCT +TACCCC +TACCCG +TACCGA +TACCGT +TACCGC +TACCGG +TACGAA +TACGAT +TACGAC +TACGAG +TACGTA +TACGTT +TACGTC +TACGTG +TACGCA +TACGCT +TACGCC +TACGCG +TACGGA +TACGGT +TACGGC +TACGGG +TAGAAA +TAGAAT +TAGAAC +TAGAAG +TAGATA +TAGATT +TAGATC +TAGATG +TAGACA +TAGACT +TAGACC +TAGACG +TAGAGA +TAGAGT +TAGAGC +TAGAGG +TAGTAA +TAGTAT +TAGTAC +TAGTAG +TAGTTA +TAGTTT +TAGTTC +TAGTTG +TAGTCA +TAGTCT +TAGTCC +TAGTCG +TAGTGA +TAGTGT +TAGTGC +TAGTGG +TAGCAA +TAGCAT +TAGCAC +TAGCAG +TAGCTA +TAGCTT +TAGCTC +TAGCTG +TAGCCA +TAGCCT +TAGCCC +TAGCCG +TAGCGA +TAGCGT +TAGCGC +TAGCGG +TAGGAA +TAGGAT +TAGGAC +TAGGAG +TAGGTA +TAGGTT +TAGGTC +TAGGTG +TAGGCA +TAGGCT +TAGGCC +TAGGCG +TAGGGA +TAGGGT +TAGGGC +TAGGGG +TTAAAA +TTAAAT +TTAAAC +TTAAAG +TTAATA +TTAATT +TTAATC +TTAATG +TTAACA +TTAACT +TTAACC +TTAACG +TTAAGA +TTAAGT +TTAAGC +TTAAGG +TTATAA +TTATAT +TTATAC +TTATAG +TTATTA +TTATTT +TTATTC +TTATTG +TTATCA +TTATCT +TTATCC +TTATCG +TTATGA +TTATGT +TTATGC +TTATGG +TTACAA +TTACAT +TTACAC +TTACAG +TTACTA +TTACTT +TTACTC +TTACTG +TTACCA +TTACCT +TTACCC +TTACCG +TTACGA +TTACGT +TTACGC +TTACGG +TTAGAA +TTAGAT +TTAGAC +TTAGAG +TTAGTA +TTAGTT +TTAGTC +TTAGTG +TTAGCA +TTAGCT +TTAGCC +TTAGCG +TTAGGA +TTAGGT +TTAGGC +TTAGGG +TTTAAA +TTTAAT +TTTAAC +TTTAAG +TTTATA +TTTATT +TTTATC +TTTATG +TTTACA +TTTACT +TTTACC +TTTACG +TTTAGA +TTTAGT +TTTAGC +TTTAGG +TTTTAA +TTTTAT +TTTTAC +TTTTAG +TTTTTA +TTTTTT +TTTTTC +TTTTTG +TTTTCA +TTTTCT +TTTTCC +TTTTCG +TTTTGA +TTTTGT +TTTTGC +TTTTGG +TTTCAA +TTTCAT +TTTCAC +TTTCAG +TTTCTA +TTTCTT +TTTCTC +TTTCTG +TTTCCA +TTTCCT +TTTCCC +TTTCCG +TTTCGA +TTTCGT +TTTCGC +TTTCGG +TTTGAA +TTTGAT +TTTGAC +TTTGAG +TTTGTA +TTTGTT +TTTGTC +TTTGTG +TTTGCA +TTTGCT +TTTGCC +TTTGCG +TTTGGA +TTTGGT +TTTGGC +TTTGGG +TTCAAA +TTCAAT +TTCAAC +TTCAAG +TTCATA +TTCATT +TTCATC +TTCATG +TTCACA +TTCACT +TTCACC +TTCACG +TTCAGA +TTCAGT +TTCAGC +TTCAGG +TTCTAA +TTCTAT +TTCTAC +TTCTAG +TTCTTA +TTCTTT +TTCTTC +TTCTTG +TTCTCA +TTCTCT +TTCTCC +TTCTCG +TTCTGA +TTCTGT +TTCTGC +TTCTGG +TTCCAA +TTCCAT +TTCCAC +TTCCAG +TTCCTA +TTCCTT +TTCCTC +TTCCTG +TTCCCA +TTCCCT +TTCCCC +TTCCCG +TTCCGA +TTCCGT +TTCCGC +TTCCGG +TTCGAA +TTCGAT +TTCGAC +TTCGAG +TTCGTA +TTCGTT +TTCGTC +TTCGTG +TTCGCA +TTCGCT +TTCGCC +TTCGCG +TTCGGA +TTCGGT +TTCGGC +TTCGGG +TTGAAA +TTGAAT +TTGAAC +TTGAAG +TTGATA +TTGATT +TTGATC +TTGATG +TTGACA +TTGACT +TTGACC +TTGACG +TTGAGA +TTGAGT +TTGAGC +TTGAGG +TTGTAA +TTGTAT +TTGTAC +TTGTAG +TTGTTA +TTGTTT +TTGTTC +TTGTTG +TTGTCA +TTGTCT +TTGTCC +TTGTCG +TTGTGA +TTGTGT +TTGTGC +TTGTGG +TTGCAA +TTGCAT +TTGCAC +TTGCAG +TTGCTA +TTGCTT +TTGCTC +TTGCTG +TTGCCA +TTGCCT +TTGCCC +TTGCCG +TTGCGA +TTGCGT +TTGCGC +TTGCGG +TTGGAA +TTGGAT +TTGGAC +TTGGAG +TTGGTA +TTGGTT +TTGGTC +TTGGTG +TTGGCA +TTGGCT +TTGGCC +TTGGCG +TTGGGA +TTGGGT +TTGGGC +TTGGGG +TCAAAA +TCAAAT +TCAAAC +TCAAAG +TCAATA +TCAATT +TCAATC +TCAATG +TCAACA +TCAACT +TCAACC +TCAACG +TCAAGA +TCAAGT +TCAAGC +TCAAGG +TCATAA +TCATAT +TCATAC +TCATAG +TCATTA +TCATTT +TCATTC +TCATTG +TCATCA +TCATCT +TCATCC +TCATCG +TCATGA +TCATGT +TCATGC +TCATGG +TCACAA +TCACAT +TCACAC +TCACAG +TCACTA +TCACTT +TCACTC +TCACTG +TCACCA +TCACCT +TCACCC +TCACCG +TCACGA +TCACGT +TCACGC +TCACGG +TCAGAA +TCAGAT +TCAGAC +TCAGAG +TCAGTA +TCAGTT +TCAGTC +TCAGTG +TCAGCA +TCAGCT +TCAGCC +TCAGCG +TCAGGA +TCAGGT +TCAGGC +TCAGGG +TCTAAA +TCTAAT +TCTAAC +TCTAAG +TCTATA +TCTATT +TCTATC +TCTATG +TCTACA +TCTACT +TCTACC +TCTACG +TCTAGA +TCTAGT +TCTAGC +TCTAGG +TCTTAA +TCTTAT +TCTTAC +TCTTAG +TCTTTA +TCTTTT +TCTTTC +TCTTTG +TCTTCA +TCTTCT +TCTTCC +TCTTCG +TCTTGA +TCTTGT +TCTTGC +TCTTGG +TCTCAA +TCTCAT +TCTCAC +TCTCAG +TCTCTA +TCTCTT +TCTCTC +TCTCTG +TCTCCA +TCTCCT +TCTCCC +TCTCCG +TCTCGA +TCTCGT +TCTCGC +TCTCGG +TCTGAA +TCTGAT +TCTGAC +TCTGAG +TCTGTA +TCTGTT +TCTGTC +TCTGTG +TCTGCA +TCTGCT +TCTGCC +TCTGCG +TCTGGA +TCTGGT +TCTGGC +TCTGGG +TCCAAA +TCCAAT +TCCAAC +TCCAAG +TCCATA +TCCATT +TCCATC +TCCATG +TCCACA +TCCACT +TCCACC +TCCACG +TCCAGA +TCCAGT +TCCAGC +TCCAGG +TCCTAA +TCCTAT +TCCTAC +TCCTAG +TCCTTA +TCCTTT +TCCTTC +TCCTTG +TCCTCA +TCCTCT +TCCTCC +TCCTCG +TCCTGA +TCCTGT +TCCTGC +TCCTGG +TCCCAA +TCCCAT +TCCCAC +TCCCAG +TCCCTA +TCCCTT +TCCCTC +TCCCTG +TCCCCA +TCCCCT +TCCCCC +TCCCCG +TCCCGA +TCCCGT +TCCCGC +TCCCGG +TCCGAA +TCCGAT +TCCGAC +TCCGAG +TCCGTA +TCCGTT +TCCGTC +TCCGTG +TCCGCA +TCCGCT +TCCGCC +TCCGCG +TCCGGA +TCCGGT +TCCGGC +TCCGGG +TCGAAA +TCGAAT +TCGAAC +TCGAAG +TCGATA +TCGATT +TCGATC +TCGATG +TCGACA +TCGACT +TCGACC +TCGACG +TCGAGA +TCGAGT +TCGAGC +TCGAGG +TCGTAA +TCGTAT +TCGTAC +TCGTAG +TCGTTA +TCGTTT +TCGTTC +TCGTTG +TCGTCA +TCGTCT +TCGTCC +TCGTCG +TCGTGA +TCGTGT +TCGTGC +TCGTGG +TCGCAA +TCGCAT +TCGCAC +TCGCAG +TCGCTA +TCGCTT +TCGCTC +TCGCTG +TCGCCA +TCGCCT +TCGCCC +TCGCCG +TCGCGA +TCGCGT +TCGCGC +TCGCGG +TCGGAA +TCGGAT +TCGGAC +TCGGAG +TCGGTA +TCGGTT +TCGGTC +TCGGTG +TCGGCA +TCGGCT +TCGGCC +TCGGCG +TCGGGA +TCGGGT +TCGGGC +TCGGGG +TGAAAA +TGAAAT +TGAAAC +TGAAAG +TGAATA +TGAATT +TGAATC +TGAATG +TGAACA +TGAACT +TGAACC +TGAACG +TGAAGA +TGAAGT +TGAAGC +TGAAGG +TGATAA +TGATAT +TGATAC +TGATAG +TGATTA +TGATTT +TGATTC +TGATTG +TGATCA +TGATCT +TGATCC +TGATCG +TGATGA +TGATGT +TGATGC +TGATGG +TGACAA +TGACAT +TGACAC +TGACAG +TGACTA +TGACTT +TGACTC +TGACTG +TGACCA +TGACCT +TGACCC +TGACCG +TGACGA +TGACGT +TGACGC +TGACGG +TGAGAA +TGAGAT +TGAGAC +TGAGAG +TGAGTA +TGAGTT +TGAGTC +TGAGTG +TGAGCA +TGAGCT +TGAGCC +TGAGCG +TGAGGA +TGAGGT +TGAGGC +TGAGGG +TGTAAA +TGTAAT +TGTAAC +TGTAAG +TGTATA +TGTATT +TGTATC +TGTATG +TGTACA +TGTACT +TGTACC +TGTACG +TGTAGA +TGTAGT +TGTAGC +TGTAGG +TGTTAA +TGTTAT +TGTTAC +TGTTAG +TGTTTA +TGTTTT +TGTTTC +TGTTTG +TGTTCA +TGTTCT +TGTTCC +TGTTCG +TGTTGA +TGTTGT +TGTTGC +TGTTGG +TGTCAA +TGTCAT +TGTCAC +TGTCAG +TGTCTA +TGTCTT +TGTCTC +TGTCTG +TGTCCA +TGTCCT +TGTCCC +TGTCCG +TGTCGA +TGTCGT +TGTCGC +TGTCGG +TGTGAA +TGTGAT +TGTGAC +TGTGAG +TGTGTA +TGTGTT +TGTGTC +TGTGTG +TGTGCA +TGTGCT +TGTGCC +TGTGCG +TGTGGA +TGTGGT +TGTGGC +TGTGGG +TGCAAA +TGCAAT +TGCAAC +TGCAAG +TGCATA +TGCATT +TGCATC +TGCATG +TGCACA +TGCACT +TGCACC +TGCACG +TGCAGA +TGCAGT +TGCAGC +TGCAGG +TGCTAA +TGCTAT +TGCTAC +TGCTAG +TGCTTA +TGCTTT +TGCTTC +TGCTTG +TGCTCA +TGCTCT +TGCTCC +TGCTCG +TGCTGA +TGCTGT +TGCTGC +TGCTGG +TGCCAA +TGCCAT +TGCCAC +TGCCAG +TGCCTA +TGCCTT +TGCCTC +TGCCTG +TGCCCA +TGCCCT +TGCCCC +TGCCCG +TGCCGA +TGCCGT +TGCCGC +TGCCGG +TGCGAA +TGCGAT +TGCGAC +TGCGAG +TGCGTA +TGCGTT +TGCGTC +TGCGTG +TGCGCA +TGCGCT +TGCGCC +TGCGCG +TGCGGA +TGCGGT +TGCGGC +TGCGGG +TGGAAA +TGGAAT +TGGAAC +TGGAAG +TGGATA +TGGATT +TGGATC +TGGATG +TGGACA +TGGACT +TGGACC +TGGACG +TGGAGA +TGGAGT +TGGAGC +TGGAGG +TGGTAA +TGGTAT +TGGTAC +TGGTAG +TGGTTA +TGGTTT +TGGTTC +TGGTTG +TGGTCA +TGGTCT +TGGTCC +TGGTCG +TGGTGA +TGGTGT +TGGTGC +TGGTGG +TGGCAA +TGGCAT +TGGCAC +TGGCAG +TGGCTA +TGGCTT +TGGCTC +TGGCTG +TGGCCA +TGGCCT +TGGCCC +TGGCCG +TGGCGA +TGGCGT +TGGCGC +TGGCGG +TGGGAA +TGGGAT +TGGGAC +TGGGAG +TGGGTA +TGGGTT +TGGGTC +TGGGTG +TGGGCA +TGGGCT +TGGGCC +TGGGCG +TGGGGA +TGGGGT +TGGGGC +TGGGGG +CAAAAA +CAAAAT +CAAAAC +CAAAAG +CAAATA +CAAATT +CAAATC +CAAATG +CAAACA +CAAACT +CAAACC +CAAACG +CAAAGA +CAAAGT +CAAAGC +CAAAGG +CAATAA +CAATAT +CAATAC +CAATAG +CAATTA +CAATTT +CAATTC +CAATTG +CAATCA +CAATCT +CAATCC +CAATCG +CAATGA +CAATGT +CAATGC +CAATGG +CAACAA +CAACAT +CAACAC +CAACAG +CAACTA +CAACTT +CAACTC +CAACTG +CAACCA +CAACCT +CAACCC +CAACCG +CAACGA +CAACGT +CAACGC +CAACGG +CAAGAA +CAAGAT +CAAGAC +CAAGAG +CAAGTA +CAAGTT +CAAGTC +CAAGTG +CAAGCA +CAAGCT +CAAGCC +CAAGCG +CAAGGA +CAAGGT +CAAGGC +CAAGGG +CATAAA +CATAAT +CATAAC +CATAAG +CATATA +CATATT +CATATC +CATATG +CATACA +CATACT +CATACC +CATACG +CATAGA +CATAGT +CATAGC +CATAGG +CATTAA +CATTAT +CATTAC +CATTAG +CATTTA +CATTTT +CATTTC +CATTTG +CATTCA +CATTCT +CATTCC +CATTCG +CATTGA +CATTGT +CATTGC +CATTGG +CATCAA +CATCAT +CATCAC +CATCAG +CATCTA +CATCTT +CATCTC +CATCTG +CATCCA +CATCCT +CATCCC +CATCCG +CATCGA +CATCGT +CATCGC +CATCGG +CATGAA +CATGAT +CATGAC +CATGAG +CATGTA +CATGTT +CATGTC +CATGTG +CATGCA +CATGCT +CATGCC +CATGCG +CATGGA +CATGGT +CATGGC +CATGGG +CACAAA +CACAAT +CACAAC +CACAAG +CACATA +CACATT +CACATC +CACATG +CACACA +CACACT +CACACC +CACACG +CACAGA +CACAGT +CACAGC +CACAGG +CACTAA +CACTAT +CACTAC +CACTAG +CACTTA +CACTTT +CACTTC +CACTTG +CACTCA +CACTCT +CACTCC +CACTCG +CACTGA +CACTGT +CACTGC +CACTGG +CACCAA +CACCAT +CACCAC +CACCAG +CACCTA +CACCTT +CACCTC +CACCTG +CACCCA +CACCCT +CACCCC +CACCCG +CACCGA +CACCGT +CACCGC +CACCGG +CACGAA +CACGAT +CACGAC +CACGAG +CACGTA +CACGTT +CACGTC +CACGTG +CACGCA +CACGCT +CACGCC +CACGCG +CACGGA +CACGGT +CACGGC +CACGGG +CAGAAA +CAGAAT +CAGAAC +CAGAAG +CAGATA +CAGATT +CAGATC +CAGATG +CAGACA +CAGACT +CAGACC +CAGACG +CAGAGA +CAGAGT +CAGAGC +CAGAGG +CAGTAA +CAGTAT +CAGTAC +CAGTAG +CAGTTA +CAGTTT +CAGTTC +CAGTTG +CAGTCA +CAGTCT +CAGTCC +CAGTCG +CAGTGA +CAGTGT +CAGTGC +CAGTGG +CAGCAA +CAGCAT +CAGCAC +CAGCAG +CAGCTA +CAGCTT +CAGCTC +CAGCTG +CAGCCA +CAGCCT +CAGCCC +CAGCCG +CAGCGA +CAGCGT +CAGCGC +CAGCGG +CAGGAA +CAGGAT +CAGGAC +CAGGAG +CAGGTA +CAGGTT +CAGGTC +CAGGTG +CAGGCA +CAGGCT +CAGGCC +CAGGCG +CAGGGA +CAGGGT +CAGGGC +CAGGGG +CTAAAA +CTAAAT +CTAAAC +CTAAAG +CTAATA +CTAATT +CTAATC +CTAATG +CTAACA +CTAACT +CTAACC +CTAACG +CTAAGA +CTAAGT +CTAAGC +CTAAGG +CTATAA +CTATAT +CTATAC +CTATAG +CTATTA +CTATTT +CTATTC +CTATTG +CTATCA +CTATCT +CTATCC +CTATCG +CTATGA +CTATGT +CTATGC +CTATGG +CTACAA +CTACAT +CTACAC +CTACAG +CTACTA +CTACTT +CTACTC +CTACTG +CTACCA +CTACCT +CTACCC +CTACCG +CTACGA +CTACGT +CTACGC +CTACGG +CTAGAA +CTAGAT +CTAGAC +CTAGAG +CTAGTA +CTAGTT +CTAGTC +CTAGTG +CTAGCA +CTAGCT +CTAGCC +CTAGCG +CTAGGA +CTAGGT +CTAGGC +CTAGGG +CTTAAA +CTTAAT +CTTAAC +CTTAAG +CTTATA +CTTATT +CTTATC +CTTATG +CTTACA +CTTACT +CTTACC +CTTACG +CTTAGA +CTTAGT +CTTAGC +CTTAGG +CTTTAA +CTTTAT +CTTTAC +CTTTAG +CTTTTA +CTTTTT +CTTTTC +CTTTTG +CTTTCA +CTTTCT +CTTTCC +CTTTCG +CTTTGA +CTTTGT +CTTTGC +CTTTGG +CTTCAA +CTTCAT +CTTCAC +CTTCAG +CTTCTA +CTTCTT +CTTCTC +CTTCTG +CTTCCA +CTTCCT +CTTCCC +CTTCCG +CTTCGA +CTTCGT +CTTCGC +CTTCGG +CTTGAA +CTTGAT +CTTGAC +CTTGAG +CTTGTA +CTTGTT +CTTGTC +CTTGTG +CTTGCA +CTTGCT +CTTGCC +CTTGCG +CTTGGA +CTTGGT +CTTGGC +CTTGGG +CTCAAA +CTCAAT +CTCAAC +CTCAAG +CTCATA +CTCATT +CTCATC +CTCATG +CTCACA +CTCACT +CTCACC +CTCACG +CTCAGA +CTCAGT +CTCAGC +CTCAGG +CTCTAA +CTCTAT +CTCTAC +CTCTAG +CTCTTA +CTCTTT +CTCTTC +CTCTTG +CTCTCA +CTCTCT +CTCTCC +CTCTCG +CTCTGA +CTCTGT +CTCTGC +CTCTGG +CTCCAA +CTCCAT +CTCCAC +CTCCAG +CTCCTA +CTCCTT +CTCCTC +CTCCTG +CTCCCA +CTCCCT +CTCCCC +CTCCCG +CTCCGA +CTCCGT +CTCCGC +CTCCGG +CTCGAA +CTCGAT +CTCGAC +CTCGAG +CTCGTA +CTCGTT +CTCGTC +CTCGTG +CTCGCA +CTCGCT +CTCGCC +CTCGCG +CTCGGA +CTCGGT +CTCGGC +CTCGGG +CTGAAA +CTGAAT +CTGAAC +CTGAAG +CTGATA +CTGATT +CTGATC +CTGATG +CTGACA +CTGACT +CTGACC +CTGACG +CTGAGA +CTGAGT +CTGAGC +CTGAGG +CTGTAA +CTGTAT +CTGTAC +CTGTAG +CTGTTA +CTGTTT +CTGTTC +CTGTTG +CTGTCA +CTGTCT +CTGTCC +CTGTCG +CTGTGA +CTGTGT +CTGTGC +CTGTGG +CTGCAA +CTGCAT +CTGCAC +CTGCAG +CTGCTA +CTGCTT +CTGCTC +CTGCTG +CTGCCA +CTGCCT +CTGCCC +CTGCCG +CTGCGA +CTGCGT +CTGCGC +CTGCGG +CTGGAA +CTGGAT +CTGGAC +CTGGAG +CTGGTA +CTGGTT +CTGGTC +CTGGTG +CTGGCA +CTGGCT +CTGGCC +CTGGCG +CTGGGA +CTGGGT +CTGGGC +CTGGGG +CCAAAA +CCAAAT +CCAAAC +CCAAAG +CCAATA +CCAATT +CCAATC +CCAATG +CCAACA +CCAACT +CCAACC +CCAACG +CCAAGA +CCAAGT +CCAAGC +CCAAGG +CCATAA +CCATAT +CCATAC +CCATAG +CCATTA +CCATTT +CCATTC +CCATTG +CCATCA +CCATCT +CCATCC +CCATCG +CCATGA +CCATGT +CCATGC +CCATGG +CCACAA +CCACAT +CCACAC +CCACAG +CCACTA +CCACTT +CCACTC +CCACTG +CCACCA +CCACCT +CCACCC +CCACCG +CCACGA +CCACGT +CCACGC +CCACGG +CCAGAA +CCAGAT +CCAGAC +CCAGAG +CCAGTA +CCAGTT +CCAGTC +CCAGTG +CCAGCA +CCAGCT +CCAGCC +CCAGCG +CCAGGA +CCAGGT +CCAGGC +CCAGGG +CCTAAA +CCTAAT +CCTAAC +CCTAAG +CCTATA +CCTATT +CCTATC +CCTATG +CCTACA +CCTACT +CCTACC +CCTACG +CCTAGA +CCTAGT +CCTAGC +CCTAGG +CCTTAA +CCTTAT +CCTTAC +CCTTAG +CCTTTA +CCTTTT +CCTTTC +CCTTTG +CCTTCA +CCTTCT +CCTTCC +CCTTCG +CCTTGA +CCTTGT +CCTTGC +CCTTGG +CCTCAA +CCTCAT +CCTCAC +CCTCAG +CCTCTA +CCTCTT +CCTCTC +CCTCTG +CCTCCA +CCTCCT +CCTCCC +CCTCCG +CCTCGA +CCTCGT +CCTCGC +CCTCGG +CCTGAA +CCTGAT +CCTGAC +CCTGAG +CCTGTA +CCTGTT +CCTGTC +CCTGTG +CCTGCA +CCTGCT +CCTGCC +CCTGCG +CCTGGA +CCTGGT +CCTGGC +CCTGGG +CCCAAA +CCCAAT +CCCAAC +CCCAAG +CCCATA +CCCATT +CCCATC +CCCATG +CCCACA +CCCACT +CCCACC +CCCACG +CCCAGA +CCCAGT +CCCAGC +CCCAGG +CCCTAA +CCCTAT +CCCTAC +CCCTAG +CCCTTA +CCCTTT +CCCTTC +CCCTTG +CCCTCA +CCCTCT +CCCTCC +CCCTCG +CCCTGA +CCCTGT +CCCTGC +CCCTGG +CCCCAA +CCCCAT +CCCCAC +CCCCAG +CCCCTA +CCCCTT +CCCCTC +CCCCTG +CCCCCA +CCCCCT +CCCCCC +CCCCCG +CCCCGA +CCCCGT +CCCCGC +CCCCGG +CCCGAA +CCCGAT +CCCGAC +CCCGAG +CCCGTA +CCCGTT +CCCGTC +CCCGTG +CCCGCA +CCCGCT +CCCGCC +CCCGCG +CCCGGA +CCCGGT +CCCGGC +CCCGGG +CCGAAA +CCGAAT +CCGAAC +CCGAAG +CCGATA +CCGATT +CCGATC +CCGATG +CCGACA +CCGACT +CCGACC +CCGACG +CCGAGA +CCGAGT +CCGAGC +CCGAGG +CCGTAA +CCGTAT +CCGTAC +CCGTAG +CCGTTA +CCGTTT +CCGTTC +CCGTTG +CCGTCA +CCGTCT +CCGTCC +CCGTCG +CCGTGA +CCGTGT +CCGTGC +CCGTGG +CCGCAA +CCGCAT +CCGCAC +CCGCAG +CCGCTA +CCGCTT +CCGCTC +CCGCTG +CCGCCA +CCGCCT +CCGCCC +CCGCCG +CCGCGA +CCGCGT +CCGCGC +CCGCGG +CCGGAA +CCGGAT +CCGGAC +CCGGAG +CCGGTA +CCGGTT +CCGGTC +CCGGTG +CCGGCA +CCGGCT +CCGGCC +CCGGCG +CCGGGA +CCGGGT +CCGGGC +CCGGGG +CGAAAA +CGAAAT +CGAAAC +CGAAAG +CGAATA +CGAATT +CGAATC +CGAATG +CGAACA +CGAACT +CGAACC +CGAACG +CGAAGA +CGAAGT +CGAAGC +CGAAGG +CGATAA +CGATAT +CGATAC +CGATAG +CGATTA +CGATTT +CGATTC +CGATTG +CGATCA +CGATCT +CGATCC +CGATCG +CGATGA +CGATGT +CGATGC +CGATGG +CGACAA +CGACAT +CGACAC +CGACAG +CGACTA +CGACTT +CGACTC +CGACTG +CGACCA +CGACCT +CGACCC +CGACCG +CGACGA +CGACGT +CGACGC +CGACGG +CGAGAA +CGAGAT +CGAGAC +CGAGAG +CGAGTA +CGAGTT +CGAGTC +CGAGTG +CGAGCA +CGAGCT +CGAGCC +CGAGCG +CGAGGA +CGAGGT +CGAGGC +CGAGGG +CGTAAA +CGTAAT +CGTAAC +CGTAAG +CGTATA +CGTATT +CGTATC +CGTATG +CGTACA +CGTACT +CGTACC +CGTACG +CGTAGA +CGTAGT +CGTAGC +CGTAGG +CGTTAA +CGTTAT +CGTTAC +CGTTAG +CGTTTA +CGTTTT +CGTTTC +CGTTTG +CGTTCA +CGTTCT +CGTTCC +CGTTCG +CGTTGA +CGTTGT +CGTTGC +CGTTGG +CGTCAA +CGTCAT +CGTCAC +CGTCAG +CGTCTA +CGTCTT +CGTCTC +CGTCTG +CGTCCA +CGTCCT +CGTCCC +CGTCCG +CGTCGA +CGTCGT +CGTCGC +CGTCGG +CGTGAA +CGTGAT +CGTGAC +CGTGAG +CGTGTA +CGTGTT +CGTGTC +CGTGTG +CGTGCA +CGTGCT +CGTGCC +CGTGCG +CGTGGA +CGTGGT +CGTGGC +CGTGGG +CGCAAA +CGCAAT +CGCAAC +CGCAAG +CGCATA +CGCATT +CGCATC +CGCATG +CGCACA +CGCACT +CGCACC +CGCACG +CGCAGA +CGCAGT +CGCAGC +CGCAGG +CGCTAA +CGCTAT +CGCTAC +CGCTAG +CGCTTA +CGCTTT +CGCTTC +CGCTTG +CGCTCA +CGCTCT +CGCTCC +CGCTCG +CGCTGA +CGCTGT +CGCTGC +CGCTGG +CGCCAA +CGCCAT +CGCCAC +CGCCAG +CGCCTA +CGCCTT +CGCCTC +CGCCTG +CGCCCA +CGCCCT +CGCCCC +CGCCCG +CGCCGA +CGCCGT +CGCCGC +CGCCGG +CGCGAA +CGCGAT +CGCGAC +CGCGAG +CGCGTA +CGCGTT +CGCGTC +CGCGTG +CGCGCA +CGCGCT +CGCGCC +CGCGCG +CGCGGA +CGCGGT +CGCGGC +CGCGGG +CGGAAA +CGGAAT +CGGAAC +CGGAAG +CGGATA +CGGATT +CGGATC +CGGATG +CGGACA +CGGACT +CGGACC +CGGACG +CGGAGA +CGGAGT +CGGAGC +CGGAGG +CGGTAA +CGGTAT +CGGTAC +CGGTAG +CGGTTA +CGGTTT +CGGTTC +CGGTTG +CGGTCA +CGGTCT +CGGTCC +CGGTCG +CGGTGA +CGGTGT +CGGTGC +CGGTGG +CGGCAA +CGGCAT +CGGCAC +CGGCAG +CGGCTA +CGGCTT +CGGCTC +CGGCTG +CGGCCA +CGGCCT +CGGCCC +CGGCCG +CGGCGA +CGGCGT +CGGCGC +CGGCGG +CGGGAA +CGGGAT +CGGGAC +CGGGAG +CGGGTA +CGGGTT +CGGGTC +CGGGTG +CGGGCA +CGGGCT +CGGGCC +CGGGCG +CGGGGA +CGGGGT +CGGGGC +CGGGGG +GAAAAA +GAAAAT +GAAAAC +GAAAAG +GAAATA +GAAATT +GAAATC +GAAATG +GAAACA +GAAACT +GAAACC +GAAACG +GAAAGA +GAAAGT +GAAAGC +GAAAGG +GAATAA +GAATAT +GAATAC +GAATAG +GAATTA +GAATTT +GAATTC +GAATTG +GAATCA +GAATCT +GAATCC +GAATCG +GAATGA +GAATGT +GAATGC +GAATGG +GAACAA +GAACAT +GAACAC +GAACAG +GAACTA +GAACTT +GAACTC +GAACTG +GAACCA +GAACCT +GAACCC +GAACCG +GAACGA +GAACGT +GAACGC +GAACGG +GAAGAA +GAAGAT +GAAGAC +GAAGAG +GAAGTA +GAAGTT +GAAGTC +GAAGTG +GAAGCA +GAAGCT +GAAGCC +GAAGCG +GAAGGA +GAAGGT +GAAGGC +GAAGGG +GATAAA +GATAAT +GATAAC +GATAAG +GATATA +GATATT +GATATC +GATATG +GATACA +GATACT +GATACC +GATACG +GATAGA +GATAGT +GATAGC +GATAGG +GATTAA +GATTAT +GATTAC +GATTAG +GATTTA +GATTTT +GATTTC +GATTTG +GATTCA +GATTCT +GATTCC +GATTCG +GATTGA +GATTGT +GATTGC +GATTGG +GATCAA +GATCAT +GATCAC +GATCAG +GATCTA +GATCTT +GATCTC +GATCTG +GATCCA +GATCCT +GATCCC +GATCCG +GATCGA +GATCGT +GATCGC +GATCGG +GATGAA +GATGAT +GATGAC +GATGAG +GATGTA +GATGTT +GATGTC +GATGTG +GATGCA +GATGCT +GATGCC +GATGCG +GATGGA +GATGGT +GATGGC +GATGGG +GACAAA +GACAAT +GACAAC +GACAAG +GACATA +GACATT +GACATC +GACATG +GACACA +GACACT +GACACC +GACACG +GACAGA +GACAGT +GACAGC +GACAGG +GACTAA +GACTAT +GACTAC +GACTAG +GACTTA +GACTTT +GACTTC +GACTTG +GACTCA +GACTCT +GACTCC +GACTCG +GACTGA +GACTGT +GACTGC +GACTGG +GACCAA +GACCAT +GACCAC +GACCAG +GACCTA +GACCTT +GACCTC +GACCTG +GACCCA +GACCCT +GACCCC +GACCCG +GACCGA +GACCGT +GACCGC +GACCGG +GACGAA +GACGAT +GACGAC +GACGAG +GACGTA +GACGTT +GACGTC +GACGTG +GACGCA +GACGCT +GACGCC +GACGCG +GACGGA +GACGGT +GACGGC +GACGGG +GAGAAA +GAGAAT +GAGAAC +GAGAAG +GAGATA +GAGATT +GAGATC +GAGATG +GAGACA +GAGACT +GAGACC +GAGACG +GAGAGA +GAGAGT +GAGAGC +GAGAGG +GAGTAA +GAGTAT +GAGTAC +GAGTAG +GAGTTA +GAGTTT +GAGTTC +GAGTTG +GAGTCA +GAGTCT +GAGTCC +GAGTCG +GAGTGA +GAGTGT +GAGTGC +GAGTGG +GAGCAA +GAGCAT +GAGCAC +GAGCAG +GAGCTA +GAGCTT +GAGCTC +GAGCTG +GAGCCA +GAGCCT +GAGCCC +GAGCCG +GAGCGA +GAGCGT +GAGCGC +GAGCGG +GAGGAA +GAGGAT +GAGGAC +GAGGAG +GAGGTA +GAGGTT +GAGGTC +GAGGTG +GAGGCA +GAGGCT +GAGGCC +GAGGCG +GAGGGA +GAGGGT +GAGGGC +GAGGGG +GTAAAA +GTAAAT +GTAAAC +GTAAAG +GTAATA +GTAATT +GTAATC +GTAATG +GTAACA +GTAACT +GTAACC +GTAACG +GTAAGA +GTAAGT +GTAAGC +GTAAGG +GTATAA +GTATAT +GTATAC +GTATAG +GTATTA +GTATTT +GTATTC +GTATTG +GTATCA +GTATCT +GTATCC +GTATCG +GTATGA +GTATGT +GTATGC +GTATGG +GTACAA +GTACAT +GTACAC +GTACAG +GTACTA +GTACTT +GTACTC +GTACTG +GTACCA +GTACCT +GTACCC +GTACCG +GTACGA +GTACGT +GTACGC +GTACGG +GTAGAA +GTAGAT +GTAGAC +GTAGAG +GTAGTA +GTAGTT +GTAGTC +GTAGTG +GTAGCA +GTAGCT +GTAGCC +GTAGCG +GTAGGA +GTAGGT +GTAGGC +GTAGGG +GTTAAA +GTTAAT +GTTAAC +GTTAAG +GTTATA +GTTATT +GTTATC +GTTATG +GTTACA +GTTACT +GTTACC +GTTACG +GTTAGA +GTTAGT +GTTAGC +GTTAGG +GTTTAA +GTTTAT +GTTTAC +GTTTAG +GTTTTA +GTTTTT +GTTTTC +GTTTTG +GTTTCA +GTTTCT +GTTTCC +GTTTCG +GTTTGA +GTTTGT +GTTTGC +GTTTGG +GTTCAA +GTTCAT +GTTCAC +GTTCAG +GTTCTA +GTTCTT +GTTCTC +GTTCTG +GTTCCA +GTTCCT +GTTCCC +GTTCCG +GTTCGA +GTTCGT +GTTCGC +GTTCGG +GTTGAA +GTTGAT +GTTGAC +GTTGAG +GTTGTA +GTTGTT +GTTGTC +GTTGTG +GTTGCA +GTTGCT +GTTGCC +GTTGCG +GTTGGA +GTTGGT +GTTGGC +GTTGGG +GTCAAA +GTCAAT +GTCAAC +GTCAAG +GTCATA +GTCATT +GTCATC +GTCATG +GTCACA +GTCACT +GTCACC +GTCACG +GTCAGA +GTCAGT +GTCAGC +GTCAGG +GTCTAA +GTCTAT +GTCTAC +GTCTAG +GTCTTA +GTCTTT +GTCTTC +GTCTTG +GTCTCA +GTCTCT +GTCTCC +GTCTCG +GTCTGA +GTCTGT +GTCTGC +GTCTGG +GTCCAA +GTCCAT +GTCCAC +GTCCAG +GTCCTA +GTCCTT +GTCCTC +GTCCTG +GTCCCA +GTCCCT +GTCCCC +GTCCCG +GTCCGA +GTCCGT +GTCCGC +GTCCGG +GTCGAA +GTCGAT +GTCGAC +GTCGAG +GTCGTA +GTCGTT +GTCGTC +GTCGTG +GTCGCA +GTCGCT +GTCGCC +GTCGCG +GTCGGA +GTCGGT +GTCGGC +GTCGGG +GTGAAA +GTGAAT +GTGAAC +GTGAAG +GTGATA +GTGATT +GTGATC +GTGATG +GTGACA +GTGACT +GTGACC +GTGACG +GTGAGA +GTGAGT +GTGAGC +GTGAGG +GTGTAA +GTGTAT +GTGTAC +GTGTAG +GTGTTA +GTGTTT +GTGTTC +GTGTTG +GTGTCA +GTGTCT +GTGTCC +GTGTCG +GTGTGA +GTGTGT +GTGTGC +GTGTGG +GTGCAA +GTGCAT +GTGCAC +GTGCAG +GTGCTA +GTGCTT +GTGCTC +GTGCTG +GTGCCA +GTGCCT +GTGCCC +GTGCCG +GTGCGA +GTGCGT +GTGCGC +GTGCGG +GTGGAA +GTGGAT +GTGGAC +GTGGAG +GTGGTA +GTGGTT +GTGGTC +GTGGTG +GTGGCA +GTGGCT +GTGGCC +GTGGCG +GTGGGA +GTGGGT +GTGGGC +GTGGGG +GCAAAA +GCAAAT +GCAAAC +GCAAAG +GCAATA +GCAATT +GCAATC +GCAATG +GCAACA +GCAACT +GCAACC +GCAACG +GCAAGA +GCAAGT +GCAAGC +GCAAGG +GCATAA +GCATAT +GCATAC +GCATAG +GCATTA +GCATTT +GCATTC +GCATTG +GCATCA +GCATCT +GCATCC +GCATCG +GCATGA +GCATGT +GCATGC +GCATGG +GCACAA +GCACAT +GCACAC +GCACAG +GCACTA +GCACTT +GCACTC +GCACTG +GCACCA +GCACCT +GCACCC +GCACCG +GCACGA +GCACGT +GCACGC +GCACGG +GCAGAA +GCAGAT +GCAGAC +GCAGAG +GCAGTA +GCAGTT +GCAGTC +GCAGTG +GCAGCA +GCAGCT +GCAGCC +GCAGCG +GCAGGA +GCAGGT +GCAGGC +GCAGGG +GCTAAA +GCTAAT +GCTAAC +GCTAAG +GCTATA +GCTATT +GCTATC +GCTATG +GCTACA +GCTACT +GCTACC +GCTACG +GCTAGA +GCTAGT +GCTAGC +GCTAGG +GCTTAA +GCTTAT +GCTTAC +GCTTAG +GCTTTA +GCTTTT +GCTTTC +GCTTTG +GCTTCA +GCTTCT +GCTTCC +GCTTCG +GCTTGA +GCTTGT +GCTTGC +GCTTGG +GCTCAA +GCTCAT +GCTCAC +GCTCAG +GCTCTA +GCTCTT +GCTCTC +GCTCTG +GCTCCA +GCTCCT +GCTCCC +GCTCCG +GCTCGA +GCTCGT +GCTCGC +GCTCGG +GCTGAA +GCTGAT +GCTGAC +GCTGAG +GCTGTA +GCTGTT +GCTGTC +GCTGTG +GCTGCA +GCTGCT +GCTGCC +GCTGCG +GCTGGA +GCTGGT +GCTGGC +GCTGGG +GCCAAA +GCCAAT +GCCAAC +GCCAAG +GCCATA +GCCATT +GCCATC +GCCATG +GCCACA +GCCACT +GCCACC +GCCACG +GCCAGA +GCCAGT +GCCAGC +GCCAGG +GCCTAA +GCCTAT +GCCTAC +GCCTAG +GCCTTA +GCCTTT +GCCTTC +GCCTTG +GCCTCA +GCCTCT +GCCTCC +GCCTCG +GCCTGA +GCCTGT +GCCTGC +GCCTGG +GCCCAA +GCCCAT +GCCCAC +GCCCAG +GCCCTA +GCCCTT +GCCCTC +GCCCTG +GCCCCA +GCCCCT +GCCCCC +GCCCCG +GCCCGA +GCCCGT +GCCCGC +GCCCGG +GCCGAA +GCCGAT +GCCGAC +GCCGAG +GCCGTA +GCCGTT +GCCGTC +GCCGTG +GCCGCA +GCCGCT +GCCGCC +GCCGCG +GCCGGA +GCCGGT +GCCGGC +GCCGGG +GCGAAA +GCGAAT +GCGAAC +GCGAAG +GCGATA +GCGATT +GCGATC +GCGATG +GCGACA +GCGACT +GCGACC +GCGACG +GCGAGA +GCGAGT +GCGAGC +GCGAGG +GCGTAA +GCGTAT +GCGTAC +GCGTAG +GCGTTA +GCGTTT +GCGTTC +GCGTTG +GCGTCA +GCGTCT +GCGTCC +GCGTCG +GCGTGA +GCGTGT +GCGTGC +GCGTGG +GCGCAA +GCGCAT +GCGCAC +GCGCAG +GCGCTA +GCGCTT +GCGCTC +GCGCTG +GCGCCA +GCGCCT +GCGCCC +GCGCCG +GCGCGA +GCGCGT +GCGCGC +GCGCGG +GCGGAA +GCGGAT +GCGGAC +GCGGAG +GCGGTA +GCGGTT +GCGGTC +GCGGTG +GCGGCA +GCGGCT +GCGGCC +GCGGCG +GCGGGA +GCGGGT +GCGGGC +GCGGGG +GGAAAA +GGAAAT +GGAAAC +GGAAAG +GGAATA +GGAATT +GGAATC +GGAATG +GGAACA +GGAACT +GGAACC +GGAACG +GGAAGA +GGAAGT +GGAAGC +GGAAGG +GGATAA +GGATAT +GGATAC +GGATAG +GGATTA +GGATTT +GGATTC +GGATTG +GGATCA +GGATCT +GGATCC +GGATCG +GGATGA +GGATGT +GGATGC +GGATGG +GGACAA +GGACAT +GGACAC +GGACAG +GGACTA +GGACTT +GGACTC +GGACTG +GGACCA +GGACCT +GGACCC +GGACCG +GGACGA +GGACGT +GGACGC +GGACGG +GGAGAA +GGAGAT +GGAGAC +GGAGAG +GGAGTA +GGAGTT +GGAGTC +GGAGTG +GGAGCA +GGAGCT +GGAGCC +GGAGCG +GGAGGA +GGAGGT +GGAGGC +GGAGGG +GGTAAA +GGTAAT +GGTAAC +GGTAAG +GGTATA +GGTATT +GGTATC +GGTATG +GGTACA +GGTACT +GGTACC +GGTACG +GGTAGA +GGTAGT +GGTAGC +GGTAGG +GGTTAA +GGTTAT +GGTTAC +GGTTAG +GGTTTA +GGTTTT +GGTTTC +GGTTTG +GGTTCA +GGTTCT +GGTTCC +GGTTCG +GGTTGA +GGTTGT +GGTTGC +GGTTGG +GGTCAA +GGTCAT +GGTCAC +GGTCAG +GGTCTA +GGTCTT +GGTCTC +GGTCTG +GGTCCA +GGTCCT +GGTCCC +GGTCCG +GGTCGA +GGTCGT +GGTCGC +GGTCGG +GGTGAA +GGTGAT +GGTGAC +GGTGAG +GGTGTA +GGTGTT +GGTGTC +GGTGTG +GGTGCA +GGTGCT +GGTGCC +GGTGCG +GGTGGA +GGTGGT +GGTGGC +GGTGGG +GGCAAA +GGCAAT +GGCAAC +GGCAAG +GGCATA +GGCATT +GGCATC +GGCATG +GGCACA +GGCACT +GGCACC +GGCACG +GGCAGA +GGCAGT +GGCAGC +GGCAGG +GGCTAA +GGCTAT +GGCTAC +GGCTAG +GGCTTA +GGCTTT +GGCTTC +GGCTTG +GGCTCA +GGCTCT +GGCTCC +GGCTCG +GGCTGA +GGCTGT +GGCTGC +GGCTGG +GGCCAA +GGCCAT +GGCCAC +GGCCAG +GGCCTA +GGCCTT +GGCCTC +GGCCTG +GGCCCA +GGCCCT +GGCCCC +GGCCCG +GGCCGA +GGCCGT +GGCCGC +GGCCGG +GGCGAA +GGCGAT +GGCGAC +GGCGAG +GGCGTA +GGCGTT +GGCGTC +GGCGTG +GGCGCA +GGCGCT +GGCGCC +GGCGCG +GGCGGA +GGCGGT +GGCGGC +GGCGGG +GGGAAA +GGGAAT +GGGAAC +GGGAAG +GGGATA +GGGATT +GGGATC +GGGATG +GGGACA +GGGACT +GGGACC +GGGACG +GGGAGA +GGGAGT +GGGAGC +GGGAGG +GGGTAA +GGGTAT +GGGTAC +GGGTAG +GGGTTA +GGGTTT +GGGTTC +GGGTTG +GGGTCA +GGGTCT +GGGTCC +GGGTCG +GGGTGA +GGGTGT +GGGTGC +GGGTGG +GGGCAA +GGGCAT +GGGCAC +GGGCAG +GGGCTA +GGGCTT +GGGCTC +GGGCTG +GGGCCA +GGGCCT +GGGCCC +GGGCCG +GGGCGA +GGGCGT +GGGCGC +GGGCGG +GGGGAA +GGGGAT +GGGGAC +GGGGAG +GGGGTA +GGGGTT +GGGGTC +GGGGTG +GGGGCA +GGGGCT +GGGGCC +GGGGCG +GGGGGA +GGGGGT +GGGGGC +GGGGGG diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-5000/config.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-5000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..9b14e61d327bf586fcff6e817cf9d365bae18ad2 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-5000/config.json @@ -0,0 +1,26 @@ +{ + "_name_or_path": "fabihamakhdoomi/TinyDNABERT", + "architectures": [ + "BertForSequenceClassification" + ], + "attention_probs_dropout_prob": 0.1, + "classifier_dropout": null, + "hidden_act": "gelu", + "hidden_dropout_prob": 0.1, + "hidden_size": 128, + "initializer_range": 0.02, + "intermediate_size": 512, + "layer_norm_eps": 1e-12, + "max_position_embeddings": 512, + "model_type": "bert", + "num_attention_heads": 2, + "num_hidden_layers": 2, + "pad_token_id": 0, + "position_embedding_type": "absolute", + "problem_type": "single_label_classification", + "torch_dtype": "float32", + "transformers_version": "4.44.0", + "type_vocab_size": 2, + "use_cache": true, + "vocab_size": 4101 +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-5000/model.safetensors b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-5000/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..8718bcaedf57291b502e2422809cfb62cf6f3644 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-5000/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0e1f43e424dd4cee0708f12e35c217456e13a07dd76490c6de515539e8bf7fb2 +size 4021688 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-5000/optimizer.pt b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-5000/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..533ea25bebd9effa83372ee276ea5a7f88ec6b85 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-5000/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8d22d84f84a7a7e7ecbda4b108fea9528e2eef402641b6fb6f827b6064395456 +size 8058700 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-5000/rng_state.pth b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-5000/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..053218cb7d47f7e442d909ea8ed9bc9b286adfe8 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-5000/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:104246ae2bc9d8b31d442f0b88ec2d93b536dbfb9e32b9ff5076aee3a117c13d +size 14244 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-5000/scheduler.pt b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-5000/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..14aa799e6f5c93344d1b233b91d0bda9d8eb59e7 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-5000/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:eca4da5723e14faf16c3be96d23919390da5288d0d9db40021d332ee1df25b93 +size 1064 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-5000/special_tokens_map.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-5000/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..a8b3208c2884c4efb86e49300fdd3dc877220cdf --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-5000/special_tokens_map.json @@ -0,0 +1,7 @@ +{ + "cls_token": "[CLS]", + "mask_token": "[MASK]", + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-5000/tokenizer.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-5000/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..ebd782de6963b9032ce9374cfe30f5275be1a30b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-5000/tokenizer.json @@ -0,0 +1,4265 @@ +{ + "version": "1.0", + "truncation": { + "direction": "Right", + "max_length": 512, + "strategy": "LongestFirst", + "stride": 0 + }, + "padding": { + "strategy": { + "Fixed": 512 + }, + "direction": "Right", + "pad_to_multiple_of": null, + "pad_id": 0, + "pad_type_id": 0, + "pad_token": "[PAD]" + }, + "added_tokens": [ + { + "id": 0, + "content": "[PAD]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 1, + "content": "[UNK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 2, + "content": "[CLS]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 3, + "content": "[SEP]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 4, + "content": "[MASK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + } + ], + "normalizer": { + "type": "BertNormalizer", + "clean_text": true, + "handle_chinese_chars": true, + "strip_accents": null, + "lowercase": false + }, + "pre_tokenizer": { + "type": "BertPreTokenizer" + }, + "post_processor": { + "type": "TemplateProcessing", + "single": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + } + ], + "pair": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "B", + "type_id": 1 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 1 + } + } + ], + "special_tokens": { + "[CLS]": { + "id": "[CLS]", + "ids": [ + 2 + ], + "tokens": [ + "[CLS]" + ] + }, + "[SEP]": { + "id": "[SEP]", + "ids": [ + 3 + ], + "tokens": [ + "[SEP]" + ] + } + } + }, + "decoder": { + "type": "WordPiece", + "prefix": "##", + "cleanup": true + }, + "model": { + "type": "WordPiece", + "unk_token": "[UNK]", + "continuing_subword_prefix": "##", + "max_input_chars_per_word": 100, + "vocab": { + "[PAD]": 0, + "[UNK]": 1, + "[CLS]": 2, + "[SEP]": 3, + "[MASK]": 4, + "AAAAAA": 5, + "AAAAAT": 6, + "AAAAAC": 7, + "AAAAAG": 8, + "AAAATA": 9, + "AAAATT": 10, + "AAAATC": 11, + "AAAATG": 12, + "AAAACA": 13, + "AAAACT": 14, + "AAAACC": 15, + "AAAACG": 16, + "AAAAGA": 17, + "AAAAGT": 18, + "AAAAGC": 19, + "AAAAGG": 20, + "AAATAA": 21, + "AAATAT": 22, + "AAATAC": 23, + "AAATAG": 24, + "AAATTA": 25, + "AAATTT": 26, + "AAATTC": 27, + "AAATTG": 28, + "AAATCA": 29, + "AAATCT": 30, + "AAATCC": 31, + "AAATCG": 32, + "AAATGA": 33, + "AAATGT": 34, + "AAATGC": 35, + "AAATGG": 36, + "AAACAA": 37, + "AAACAT": 38, + "AAACAC": 39, + "AAACAG": 40, + "AAACTA": 41, + "AAACTT": 42, + "AAACTC": 43, + "AAACTG": 44, + "AAACCA": 45, + "AAACCT": 46, + "AAACCC": 47, + "AAACCG": 48, + "AAACGA": 49, + "AAACGT": 50, + "AAACGC": 51, + "AAACGG": 52, + "AAAGAA": 53, + "AAAGAT": 54, + "AAAGAC": 55, + "AAAGAG": 56, + "AAAGTA": 57, + "AAAGTT": 58, + "AAAGTC": 59, + "AAAGTG": 60, + "AAAGCA": 61, + "AAAGCT": 62, + "AAAGCC": 63, + "AAAGCG": 64, + "AAAGGA": 65, + "AAAGGT": 66, + "AAAGGC": 67, + "AAAGGG": 68, + "AATAAA": 69, + "AATAAT": 70, + "AATAAC": 71, + "AATAAG": 72, + "AATATA": 73, + "AATATT": 74, + "AATATC": 75, + "AATATG": 76, + "AATACA": 77, + "AATACT": 78, + "AATACC": 79, + "AATACG": 80, + "AATAGA": 81, + "AATAGT": 82, + "AATAGC": 83, + "AATAGG": 84, + "AATTAA": 85, + "AATTAT": 86, + "AATTAC": 87, + "AATTAG": 88, + "AATTTA": 89, + "AATTTT": 90, + "AATTTC": 91, + "AATTTG": 92, + "AATTCA": 93, + "AATTCT": 94, + "AATTCC": 95, + "AATTCG": 96, + "AATTGA": 97, + "AATTGT": 98, + "AATTGC": 99, + "AATTGG": 100, + "AATCAA": 101, + "AATCAT": 102, + "AATCAC": 103, + "AATCAG": 104, + "AATCTA": 105, + "AATCTT": 106, + "AATCTC": 107, + "AATCTG": 108, + "AATCCA": 109, + "AATCCT": 110, + "AATCCC": 111, + "AATCCG": 112, + "AATCGA": 113, + "AATCGT": 114, + "AATCGC": 115, + "AATCGG": 116, + "AATGAA": 117, + "AATGAT": 118, + "AATGAC": 119, + "AATGAG": 120, + "AATGTA": 121, + "AATGTT": 122, + "AATGTC": 123, + "AATGTG": 124, + "AATGCA": 125, + "AATGCT": 126, + "AATGCC": 127, + "AATGCG": 128, + "AATGGA": 129, + "AATGGT": 130, + "AATGGC": 131, + "AATGGG": 132, + "AACAAA": 133, + "AACAAT": 134, + "AACAAC": 135, + "AACAAG": 136, + "AACATA": 137, + "AACATT": 138, + "AACATC": 139, + "AACATG": 140, + "AACACA": 141, + "AACACT": 142, + "AACACC": 143, + "AACACG": 144, + "AACAGA": 145, + "AACAGT": 146, + "AACAGC": 147, + "AACAGG": 148, + "AACTAA": 149, + "AACTAT": 150, + "AACTAC": 151, + "AACTAG": 152, + "AACTTA": 153, + "AACTTT": 154, + "AACTTC": 155, + "AACTTG": 156, + "AACTCA": 157, + "AACTCT": 158, + "AACTCC": 159, + "AACTCG": 160, + "AACTGA": 161, + "AACTGT": 162, + "AACTGC": 163, + "AACTGG": 164, + "AACCAA": 165, + "AACCAT": 166, + "AACCAC": 167, + "AACCAG": 168, + "AACCTA": 169, + "AACCTT": 170, + "AACCTC": 171, + "AACCTG": 172, + "AACCCA": 173, + "AACCCT": 174, + "AACCCC": 175, + "AACCCG": 176, + "AACCGA": 177, + "AACCGT": 178, + "AACCGC": 179, + "AACCGG": 180, + "AACGAA": 181, + "AACGAT": 182, + "AACGAC": 183, + "AACGAG": 184, + "AACGTA": 185, + "AACGTT": 186, + "AACGTC": 187, + "AACGTG": 188, + "AACGCA": 189, + "AACGCT": 190, + "AACGCC": 191, + "AACGCG": 192, + "AACGGA": 193, + "AACGGT": 194, + "AACGGC": 195, + "AACGGG": 196, + "AAGAAA": 197, + "AAGAAT": 198, + "AAGAAC": 199, + "AAGAAG": 200, + "AAGATA": 201, + "AAGATT": 202, + "AAGATC": 203, + "AAGATG": 204, + "AAGACA": 205, + "AAGACT": 206, + "AAGACC": 207, + "AAGACG": 208, + "AAGAGA": 209, + "AAGAGT": 210, + "AAGAGC": 211, + "AAGAGG": 212, + "AAGTAA": 213, + "AAGTAT": 214, + "AAGTAC": 215, + "AAGTAG": 216, + "AAGTTA": 217, + "AAGTTT": 218, + "AAGTTC": 219, + "AAGTTG": 220, + "AAGTCA": 221, + "AAGTCT": 222, + "AAGTCC": 223, + "AAGTCG": 224, + "AAGTGA": 225, + "AAGTGT": 226, + "AAGTGC": 227, + "AAGTGG": 228, + "AAGCAA": 229, + "AAGCAT": 230, + "AAGCAC": 231, + "AAGCAG": 232, + "AAGCTA": 233, + "AAGCTT": 234, + "AAGCTC": 235, + "AAGCTG": 236, + "AAGCCA": 237, + "AAGCCT": 238, + "AAGCCC": 239, + "AAGCCG": 240, + "AAGCGA": 241, + "AAGCGT": 242, + "AAGCGC": 243, + "AAGCGG": 244, + "AAGGAA": 245, + "AAGGAT": 246, + "AAGGAC": 247, + "AAGGAG": 248, + "AAGGTA": 249, + "AAGGTT": 250, + "AAGGTC": 251, + "AAGGTG": 252, + "AAGGCA": 253, + "AAGGCT": 254, + "AAGGCC": 255, + "AAGGCG": 256, + "AAGGGA": 257, + "AAGGGT": 258, + "AAGGGC": 259, + "AAGGGG": 260, + "ATAAAA": 261, + "ATAAAT": 262, + "ATAAAC": 263, + "ATAAAG": 264, + "ATAATA": 265, + "ATAATT": 266, + "ATAATC": 267, + "ATAATG": 268, + "ATAACA": 269, + "ATAACT": 270, + "ATAACC": 271, + "ATAACG": 272, + "ATAAGA": 273, + "ATAAGT": 274, + "ATAAGC": 275, + "ATAAGG": 276, + "ATATAA": 277, + "ATATAT": 278, + "ATATAC": 279, + "ATATAG": 280, + "ATATTA": 281, + "ATATTT": 282, + "ATATTC": 283, + "ATATTG": 284, + "ATATCA": 285, + "ATATCT": 286, + "ATATCC": 287, + "ATATCG": 288, + "ATATGA": 289, + "ATATGT": 290, + "ATATGC": 291, + "ATATGG": 292, + "ATACAA": 293, + "ATACAT": 294, + "ATACAC": 295, + "ATACAG": 296, + "ATACTA": 297, + "ATACTT": 298, + "ATACTC": 299, + "ATACTG": 300, + "ATACCA": 301, + "ATACCT": 302, + "ATACCC": 303, + "ATACCG": 304, + "ATACGA": 305, + "ATACGT": 306, + "ATACGC": 307, + "ATACGG": 308, + "ATAGAA": 309, + "ATAGAT": 310, + "ATAGAC": 311, + "ATAGAG": 312, + "ATAGTA": 313, + "ATAGTT": 314, + "ATAGTC": 315, + "ATAGTG": 316, + "ATAGCA": 317, + "ATAGCT": 318, + "ATAGCC": 319, + "ATAGCG": 320, + "ATAGGA": 321, + "ATAGGT": 322, + "ATAGGC": 323, + "ATAGGG": 324, + "ATTAAA": 325, + "ATTAAT": 326, + "ATTAAC": 327, + "ATTAAG": 328, + "ATTATA": 329, + "ATTATT": 330, + "ATTATC": 331, + "ATTATG": 332, + "ATTACA": 333, + "ATTACT": 334, + "ATTACC": 335, + "ATTACG": 336, + "ATTAGA": 337, + "ATTAGT": 338, + "ATTAGC": 339, + "ATTAGG": 340, + "ATTTAA": 341, + "ATTTAT": 342, + "ATTTAC": 343, + "ATTTAG": 344, + "ATTTTA": 345, + "ATTTTT": 346, + "ATTTTC": 347, + "ATTTTG": 348, + "ATTTCA": 349, + "ATTTCT": 350, + "ATTTCC": 351, + "ATTTCG": 352, + "ATTTGA": 353, + "ATTTGT": 354, + "ATTTGC": 355, + "ATTTGG": 356, + "ATTCAA": 357, + "ATTCAT": 358, + "ATTCAC": 359, + "ATTCAG": 360, + "ATTCTA": 361, + "ATTCTT": 362, + "ATTCTC": 363, + "ATTCTG": 364, + "ATTCCA": 365, + "ATTCCT": 366, + "ATTCCC": 367, + "ATTCCG": 368, + "ATTCGA": 369, + "ATTCGT": 370, + "ATTCGC": 371, + "ATTCGG": 372, + "ATTGAA": 373, + "ATTGAT": 374, + "ATTGAC": 375, + "ATTGAG": 376, + "ATTGTA": 377, + "ATTGTT": 378, + "ATTGTC": 379, + "ATTGTG": 380, + "ATTGCA": 381, + "ATTGCT": 382, + "ATTGCC": 383, + "ATTGCG": 384, + "ATTGGA": 385, + "ATTGGT": 386, + "ATTGGC": 387, + "ATTGGG": 388, + "ATCAAA": 389, + "ATCAAT": 390, + "ATCAAC": 391, + "ATCAAG": 392, + "ATCATA": 393, + "ATCATT": 394, + "ATCATC": 395, + "ATCATG": 396, + "ATCACA": 397, + "ATCACT": 398, + "ATCACC": 399, + "ATCACG": 400, + "ATCAGA": 401, + "ATCAGT": 402, + "ATCAGC": 403, + "ATCAGG": 404, + "ATCTAA": 405, + "ATCTAT": 406, + "ATCTAC": 407, + "ATCTAG": 408, + "ATCTTA": 409, + "ATCTTT": 410, + "ATCTTC": 411, + "ATCTTG": 412, + "ATCTCA": 413, + "ATCTCT": 414, + "ATCTCC": 415, + "ATCTCG": 416, + "ATCTGA": 417, + "ATCTGT": 418, + "ATCTGC": 419, + "ATCTGG": 420, + "ATCCAA": 421, + "ATCCAT": 422, + "ATCCAC": 423, + "ATCCAG": 424, + "ATCCTA": 425, + "ATCCTT": 426, + "ATCCTC": 427, + "ATCCTG": 428, + "ATCCCA": 429, + "ATCCCT": 430, + "ATCCCC": 431, + "ATCCCG": 432, + "ATCCGA": 433, + "ATCCGT": 434, + "ATCCGC": 435, + "ATCCGG": 436, + "ATCGAA": 437, + "ATCGAT": 438, + "ATCGAC": 439, + "ATCGAG": 440, + "ATCGTA": 441, + "ATCGTT": 442, + "ATCGTC": 443, + "ATCGTG": 444, + "ATCGCA": 445, + "ATCGCT": 446, + "ATCGCC": 447, + "ATCGCG": 448, + "ATCGGA": 449, + "ATCGGT": 450, + "ATCGGC": 451, + "ATCGGG": 452, + "ATGAAA": 453, + "ATGAAT": 454, + "ATGAAC": 455, + "ATGAAG": 456, + "ATGATA": 457, + "ATGATT": 458, + "ATGATC": 459, + "ATGATG": 460, + "ATGACA": 461, + "ATGACT": 462, + "ATGACC": 463, + "ATGACG": 464, + "ATGAGA": 465, + "ATGAGT": 466, + "ATGAGC": 467, + "ATGAGG": 468, + "ATGTAA": 469, + "ATGTAT": 470, + "ATGTAC": 471, + "ATGTAG": 472, + "ATGTTA": 473, + "ATGTTT": 474, + "ATGTTC": 475, + "ATGTTG": 476, + "ATGTCA": 477, + "ATGTCT": 478, + "ATGTCC": 479, + "ATGTCG": 480, + "ATGTGA": 481, + "ATGTGT": 482, + "ATGTGC": 483, + "ATGTGG": 484, + "ATGCAA": 485, + "ATGCAT": 486, + "ATGCAC": 487, + "ATGCAG": 488, + "ATGCTA": 489, + "ATGCTT": 490, + "ATGCTC": 491, + "ATGCTG": 492, + "ATGCCA": 493, + "ATGCCT": 494, + "ATGCCC": 495, + "ATGCCG": 496, + "ATGCGA": 497, + "ATGCGT": 498, + "ATGCGC": 499, + "ATGCGG": 500, + "ATGGAA": 501, + "ATGGAT": 502, + "ATGGAC": 503, + "ATGGAG": 504, + "ATGGTA": 505, + "ATGGTT": 506, + "ATGGTC": 507, + "ATGGTG": 508, + "ATGGCA": 509, + "ATGGCT": 510, + "ATGGCC": 511, + "ATGGCG": 512, + "ATGGGA": 513, + "ATGGGT": 514, + "ATGGGC": 515, + "ATGGGG": 516, + "ACAAAA": 517, + "ACAAAT": 518, + "ACAAAC": 519, + "ACAAAG": 520, + "ACAATA": 521, + "ACAATT": 522, + "ACAATC": 523, + "ACAATG": 524, + "ACAACA": 525, + "ACAACT": 526, + "ACAACC": 527, + "ACAACG": 528, + "ACAAGA": 529, + "ACAAGT": 530, + "ACAAGC": 531, + "ACAAGG": 532, + "ACATAA": 533, + "ACATAT": 534, + "ACATAC": 535, + "ACATAG": 536, + "ACATTA": 537, + "ACATTT": 538, + "ACATTC": 539, + "ACATTG": 540, + "ACATCA": 541, + "ACATCT": 542, + "ACATCC": 543, + "ACATCG": 544, + "ACATGA": 545, + "ACATGT": 546, + "ACATGC": 547, + "ACATGG": 548, + "ACACAA": 549, + "ACACAT": 550, + "ACACAC": 551, + "ACACAG": 552, + "ACACTA": 553, + "ACACTT": 554, + "ACACTC": 555, + "ACACTG": 556, + "ACACCA": 557, + "ACACCT": 558, + "ACACCC": 559, + "ACACCG": 560, + "ACACGA": 561, + "ACACGT": 562, + "ACACGC": 563, + "ACACGG": 564, + "ACAGAA": 565, + "ACAGAT": 566, + "ACAGAC": 567, + "ACAGAG": 568, + "ACAGTA": 569, + "ACAGTT": 570, + "ACAGTC": 571, + "ACAGTG": 572, + "ACAGCA": 573, + "ACAGCT": 574, + "ACAGCC": 575, + "ACAGCG": 576, + "ACAGGA": 577, + "ACAGGT": 578, + "ACAGGC": 579, + "ACAGGG": 580, + "ACTAAA": 581, + "ACTAAT": 582, + "ACTAAC": 583, + "ACTAAG": 584, + "ACTATA": 585, + "ACTATT": 586, + "ACTATC": 587, + "ACTATG": 588, + "ACTACA": 589, + "ACTACT": 590, + "ACTACC": 591, + "ACTACG": 592, + "ACTAGA": 593, + "ACTAGT": 594, + "ACTAGC": 595, + "ACTAGG": 596, + "ACTTAA": 597, + "ACTTAT": 598, + "ACTTAC": 599, + "ACTTAG": 600, + "ACTTTA": 601, + "ACTTTT": 602, + "ACTTTC": 603, + "ACTTTG": 604, + "ACTTCA": 605, + "ACTTCT": 606, + "ACTTCC": 607, + "ACTTCG": 608, + "ACTTGA": 609, + "ACTTGT": 610, + "ACTTGC": 611, + "ACTTGG": 612, + "ACTCAA": 613, + "ACTCAT": 614, + "ACTCAC": 615, + "ACTCAG": 616, + "ACTCTA": 617, + "ACTCTT": 618, + "ACTCTC": 619, + "ACTCTG": 620, + "ACTCCA": 621, + "ACTCCT": 622, + "ACTCCC": 623, + "ACTCCG": 624, + "ACTCGA": 625, + "ACTCGT": 626, + "ACTCGC": 627, + "ACTCGG": 628, + "ACTGAA": 629, + "ACTGAT": 630, + "ACTGAC": 631, + "ACTGAG": 632, + "ACTGTA": 633, + "ACTGTT": 634, + "ACTGTC": 635, + "ACTGTG": 636, + "ACTGCA": 637, + "ACTGCT": 638, + "ACTGCC": 639, + "ACTGCG": 640, + "ACTGGA": 641, + "ACTGGT": 642, + "ACTGGC": 643, + "ACTGGG": 644, + "ACCAAA": 645, + "ACCAAT": 646, + "ACCAAC": 647, + "ACCAAG": 648, + "ACCATA": 649, + "ACCATT": 650, + "ACCATC": 651, + "ACCATG": 652, + "ACCACA": 653, + "ACCACT": 654, + "ACCACC": 655, + "ACCACG": 656, + "ACCAGA": 657, + "ACCAGT": 658, + "ACCAGC": 659, + "ACCAGG": 660, + "ACCTAA": 661, + "ACCTAT": 662, + "ACCTAC": 663, + "ACCTAG": 664, + "ACCTTA": 665, + "ACCTTT": 666, + "ACCTTC": 667, + "ACCTTG": 668, + "ACCTCA": 669, + "ACCTCT": 670, + "ACCTCC": 671, + "ACCTCG": 672, + "ACCTGA": 673, + "ACCTGT": 674, + "ACCTGC": 675, + "ACCTGG": 676, + "ACCCAA": 677, + "ACCCAT": 678, + "ACCCAC": 679, + "ACCCAG": 680, + "ACCCTA": 681, + "ACCCTT": 682, + "ACCCTC": 683, + "ACCCTG": 684, + "ACCCCA": 685, + "ACCCCT": 686, + "ACCCCC": 687, + "ACCCCG": 688, + "ACCCGA": 689, + "ACCCGT": 690, + "ACCCGC": 691, + "ACCCGG": 692, + "ACCGAA": 693, + "ACCGAT": 694, + "ACCGAC": 695, + "ACCGAG": 696, + "ACCGTA": 697, + "ACCGTT": 698, + "ACCGTC": 699, + "ACCGTG": 700, + "ACCGCA": 701, + "ACCGCT": 702, + "ACCGCC": 703, + "ACCGCG": 704, + "ACCGGA": 705, + "ACCGGT": 706, + "ACCGGC": 707, + "ACCGGG": 708, + "ACGAAA": 709, + "ACGAAT": 710, + "ACGAAC": 711, + "ACGAAG": 712, + "ACGATA": 713, + "ACGATT": 714, + "ACGATC": 715, + "ACGATG": 716, + "ACGACA": 717, + "ACGACT": 718, + "ACGACC": 719, + "ACGACG": 720, + "ACGAGA": 721, + "ACGAGT": 722, + "ACGAGC": 723, + "ACGAGG": 724, + "ACGTAA": 725, + "ACGTAT": 726, + "ACGTAC": 727, + "ACGTAG": 728, + "ACGTTA": 729, + "ACGTTT": 730, + "ACGTTC": 731, + "ACGTTG": 732, + "ACGTCA": 733, + "ACGTCT": 734, + "ACGTCC": 735, + "ACGTCG": 736, + "ACGTGA": 737, + "ACGTGT": 738, + "ACGTGC": 739, + "ACGTGG": 740, + "ACGCAA": 741, + "ACGCAT": 742, + "ACGCAC": 743, + "ACGCAG": 744, + "ACGCTA": 745, + "ACGCTT": 746, + "ACGCTC": 747, + "ACGCTG": 748, + "ACGCCA": 749, + "ACGCCT": 750, + "ACGCCC": 751, + "ACGCCG": 752, + "ACGCGA": 753, + "ACGCGT": 754, + "ACGCGC": 755, + "ACGCGG": 756, + "ACGGAA": 757, + "ACGGAT": 758, + "ACGGAC": 759, + "ACGGAG": 760, + "ACGGTA": 761, + "ACGGTT": 762, + "ACGGTC": 763, + "ACGGTG": 764, + "ACGGCA": 765, + "ACGGCT": 766, + "ACGGCC": 767, + "ACGGCG": 768, + "ACGGGA": 769, + "ACGGGT": 770, + "ACGGGC": 771, + "ACGGGG": 772, + "AGAAAA": 773, + "AGAAAT": 774, + "AGAAAC": 775, + "AGAAAG": 776, + "AGAATA": 777, + "AGAATT": 778, + "AGAATC": 779, + "AGAATG": 780, + "AGAACA": 781, + "AGAACT": 782, + "AGAACC": 783, + "AGAACG": 784, + "AGAAGA": 785, + "AGAAGT": 786, + "AGAAGC": 787, + "AGAAGG": 788, + "AGATAA": 789, + "AGATAT": 790, + "AGATAC": 791, + "AGATAG": 792, + "AGATTA": 793, + "AGATTT": 794, + "AGATTC": 795, + "AGATTG": 796, + "AGATCA": 797, + "AGATCT": 798, + "AGATCC": 799, + "AGATCG": 800, + "AGATGA": 801, + "AGATGT": 802, + "AGATGC": 803, + "AGATGG": 804, + "AGACAA": 805, + "AGACAT": 806, + "AGACAC": 807, + "AGACAG": 808, + "AGACTA": 809, + "AGACTT": 810, + "AGACTC": 811, + "AGACTG": 812, + "AGACCA": 813, + "AGACCT": 814, + "AGACCC": 815, + "AGACCG": 816, + "AGACGA": 817, + "AGACGT": 818, + "AGACGC": 819, + "AGACGG": 820, + "AGAGAA": 821, + "AGAGAT": 822, + "AGAGAC": 823, + "AGAGAG": 824, + "AGAGTA": 825, + "AGAGTT": 826, + "AGAGTC": 827, + "AGAGTG": 828, + "AGAGCA": 829, + "AGAGCT": 830, + "AGAGCC": 831, + "AGAGCG": 832, + "AGAGGA": 833, + "AGAGGT": 834, + "AGAGGC": 835, + "AGAGGG": 836, + "AGTAAA": 837, + "AGTAAT": 838, + "AGTAAC": 839, + "AGTAAG": 840, + "AGTATA": 841, + "AGTATT": 842, + "AGTATC": 843, + "AGTATG": 844, + "AGTACA": 845, + "AGTACT": 846, + "AGTACC": 847, + "AGTACG": 848, + "AGTAGA": 849, + "AGTAGT": 850, + "AGTAGC": 851, + "AGTAGG": 852, + "AGTTAA": 853, + "AGTTAT": 854, + "AGTTAC": 855, + "AGTTAG": 856, + "AGTTTA": 857, + "AGTTTT": 858, + "AGTTTC": 859, + "AGTTTG": 860, + "AGTTCA": 861, + "AGTTCT": 862, + "AGTTCC": 863, + "AGTTCG": 864, + "AGTTGA": 865, + "AGTTGT": 866, + "AGTTGC": 867, + "AGTTGG": 868, + "AGTCAA": 869, + "AGTCAT": 870, + "AGTCAC": 871, + "AGTCAG": 872, + "AGTCTA": 873, + "AGTCTT": 874, + "AGTCTC": 875, + "AGTCTG": 876, + "AGTCCA": 877, + "AGTCCT": 878, + "AGTCCC": 879, + "AGTCCG": 880, + "AGTCGA": 881, + "AGTCGT": 882, + "AGTCGC": 883, + "AGTCGG": 884, + "AGTGAA": 885, + "AGTGAT": 886, + "AGTGAC": 887, + "AGTGAG": 888, + "AGTGTA": 889, + "AGTGTT": 890, + "AGTGTC": 891, + "AGTGTG": 892, + "AGTGCA": 893, + "AGTGCT": 894, + "AGTGCC": 895, + "AGTGCG": 896, + "AGTGGA": 897, + "AGTGGT": 898, + "AGTGGC": 899, + "AGTGGG": 900, + "AGCAAA": 901, + "AGCAAT": 902, + "AGCAAC": 903, + "AGCAAG": 904, + "AGCATA": 905, + "AGCATT": 906, + "AGCATC": 907, + "AGCATG": 908, + "AGCACA": 909, + "AGCACT": 910, + "AGCACC": 911, + "AGCACG": 912, + "AGCAGA": 913, + "AGCAGT": 914, + "AGCAGC": 915, + "AGCAGG": 916, + "AGCTAA": 917, + "AGCTAT": 918, + "AGCTAC": 919, + "AGCTAG": 920, + "AGCTTA": 921, + "AGCTTT": 922, + "AGCTTC": 923, + "AGCTTG": 924, + "AGCTCA": 925, + "AGCTCT": 926, + "AGCTCC": 927, + "AGCTCG": 928, + "AGCTGA": 929, + "AGCTGT": 930, + "AGCTGC": 931, + "AGCTGG": 932, + "AGCCAA": 933, + "AGCCAT": 934, + "AGCCAC": 935, + "AGCCAG": 936, + "AGCCTA": 937, + "AGCCTT": 938, + "AGCCTC": 939, + "AGCCTG": 940, + "AGCCCA": 941, + "AGCCCT": 942, + "AGCCCC": 943, + "AGCCCG": 944, + "AGCCGA": 945, + "AGCCGT": 946, + "AGCCGC": 947, + "AGCCGG": 948, + "AGCGAA": 949, + "AGCGAT": 950, + "AGCGAC": 951, + "AGCGAG": 952, + "AGCGTA": 953, + "AGCGTT": 954, + "AGCGTC": 955, + "AGCGTG": 956, + "AGCGCA": 957, + "AGCGCT": 958, + "AGCGCC": 959, + "AGCGCG": 960, + "AGCGGA": 961, + "AGCGGT": 962, + "AGCGGC": 963, + "AGCGGG": 964, + "AGGAAA": 965, + "AGGAAT": 966, + "AGGAAC": 967, + "AGGAAG": 968, + "AGGATA": 969, + "AGGATT": 970, + "AGGATC": 971, + "AGGATG": 972, + "AGGACA": 973, + "AGGACT": 974, + "AGGACC": 975, + "AGGACG": 976, + "AGGAGA": 977, + "AGGAGT": 978, + "AGGAGC": 979, + "AGGAGG": 980, + "AGGTAA": 981, + "AGGTAT": 982, + "AGGTAC": 983, + "AGGTAG": 984, + "AGGTTA": 985, + "AGGTTT": 986, + "AGGTTC": 987, + "AGGTTG": 988, + "AGGTCA": 989, + "AGGTCT": 990, + "AGGTCC": 991, + "AGGTCG": 992, + "AGGTGA": 993, + "AGGTGT": 994, + "AGGTGC": 995, + "AGGTGG": 996, + "AGGCAA": 997, + "AGGCAT": 998, + "AGGCAC": 999, + "AGGCAG": 1000, + "AGGCTA": 1001, + "AGGCTT": 1002, + "AGGCTC": 1003, + "AGGCTG": 1004, + "AGGCCA": 1005, + "AGGCCT": 1006, + "AGGCCC": 1007, + "AGGCCG": 1008, + "AGGCGA": 1009, + "AGGCGT": 1010, + "AGGCGC": 1011, + "AGGCGG": 1012, + "AGGGAA": 1013, + "AGGGAT": 1014, + "AGGGAC": 1015, + "AGGGAG": 1016, + "AGGGTA": 1017, + "AGGGTT": 1018, + "AGGGTC": 1019, + "AGGGTG": 1020, + "AGGGCA": 1021, + "AGGGCT": 1022, + "AGGGCC": 1023, + "AGGGCG": 1024, + "AGGGGA": 1025, + "AGGGGT": 1026, + "AGGGGC": 1027, + "AGGGGG": 1028, + "TAAAAA": 1029, + "TAAAAT": 1030, + "TAAAAC": 1031, + "TAAAAG": 1032, + "TAAATA": 1033, + "TAAATT": 1034, + "TAAATC": 1035, + "TAAATG": 1036, + "TAAACA": 1037, + "TAAACT": 1038, + "TAAACC": 1039, + "TAAACG": 1040, + "TAAAGA": 1041, + "TAAAGT": 1042, + "TAAAGC": 1043, + "TAAAGG": 1044, + "TAATAA": 1045, + "TAATAT": 1046, + "TAATAC": 1047, + "TAATAG": 1048, + "TAATTA": 1049, + "TAATTT": 1050, + "TAATTC": 1051, + "TAATTG": 1052, + "TAATCA": 1053, + "TAATCT": 1054, + "TAATCC": 1055, + "TAATCG": 1056, + "TAATGA": 1057, + "TAATGT": 1058, + "TAATGC": 1059, + "TAATGG": 1060, + "TAACAA": 1061, + "TAACAT": 1062, + "TAACAC": 1063, + "TAACAG": 1064, + "TAACTA": 1065, + "TAACTT": 1066, + "TAACTC": 1067, + "TAACTG": 1068, + "TAACCA": 1069, + "TAACCT": 1070, + "TAACCC": 1071, + "TAACCG": 1072, + "TAACGA": 1073, + "TAACGT": 1074, + "TAACGC": 1075, + "TAACGG": 1076, + "TAAGAA": 1077, + "TAAGAT": 1078, + "TAAGAC": 1079, + "TAAGAG": 1080, + "TAAGTA": 1081, + "TAAGTT": 1082, + "TAAGTC": 1083, + "TAAGTG": 1084, + "TAAGCA": 1085, + "TAAGCT": 1086, + "TAAGCC": 1087, + "TAAGCG": 1088, + "TAAGGA": 1089, + "TAAGGT": 1090, + "TAAGGC": 1091, + "TAAGGG": 1092, + "TATAAA": 1093, + "TATAAT": 1094, + "TATAAC": 1095, + "TATAAG": 1096, + "TATATA": 1097, + "TATATT": 1098, + "TATATC": 1099, + "TATATG": 1100, + "TATACA": 1101, + "TATACT": 1102, + "TATACC": 1103, + "TATACG": 1104, + "TATAGA": 1105, + "TATAGT": 1106, + "TATAGC": 1107, + "TATAGG": 1108, + "TATTAA": 1109, + "TATTAT": 1110, + "TATTAC": 1111, + "TATTAG": 1112, + "TATTTA": 1113, + "TATTTT": 1114, + "TATTTC": 1115, + "TATTTG": 1116, + "TATTCA": 1117, + "TATTCT": 1118, + "TATTCC": 1119, + "TATTCG": 1120, + "TATTGA": 1121, + "TATTGT": 1122, + "TATTGC": 1123, + "TATTGG": 1124, + "TATCAA": 1125, + "TATCAT": 1126, + "TATCAC": 1127, + "TATCAG": 1128, + "TATCTA": 1129, + "TATCTT": 1130, + "TATCTC": 1131, + "TATCTG": 1132, + "TATCCA": 1133, + "TATCCT": 1134, + "TATCCC": 1135, + "TATCCG": 1136, + "TATCGA": 1137, + "TATCGT": 1138, + "TATCGC": 1139, + "TATCGG": 1140, + "TATGAA": 1141, + "TATGAT": 1142, + "TATGAC": 1143, + "TATGAG": 1144, + "TATGTA": 1145, + "TATGTT": 1146, + "TATGTC": 1147, + "TATGTG": 1148, + "TATGCA": 1149, + "TATGCT": 1150, + "TATGCC": 1151, + "TATGCG": 1152, + "TATGGA": 1153, + "TATGGT": 1154, + "TATGGC": 1155, + "TATGGG": 1156, + "TACAAA": 1157, + "TACAAT": 1158, + "TACAAC": 1159, + "TACAAG": 1160, + "TACATA": 1161, + "TACATT": 1162, + "TACATC": 1163, + "TACATG": 1164, + "TACACA": 1165, + "TACACT": 1166, + "TACACC": 1167, + "TACACG": 1168, + "TACAGA": 1169, + "TACAGT": 1170, + "TACAGC": 1171, + "TACAGG": 1172, + "TACTAA": 1173, + "TACTAT": 1174, + "TACTAC": 1175, + "TACTAG": 1176, + "TACTTA": 1177, + "TACTTT": 1178, + "TACTTC": 1179, + "TACTTG": 1180, + "TACTCA": 1181, + "TACTCT": 1182, + "TACTCC": 1183, + "TACTCG": 1184, + "TACTGA": 1185, + "TACTGT": 1186, + "TACTGC": 1187, + "TACTGG": 1188, + "TACCAA": 1189, + "TACCAT": 1190, + "TACCAC": 1191, + "TACCAG": 1192, + "TACCTA": 1193, + "TACCTT": 1194, + "TACCTC": 1195, + "TACCTG": 1196, + "TACCCA": 1197, + "TACCCT": 1198, + "TACCCC": 1199, + "TACCCG": 1200, + "TACCGA": 1201, + "TACCGT": 1202, + "TACCGC": 1203, + "TACCGG": 1204, + "TACGAA": 1205, + "TACGAT": 1206, + "TACGAC": 1207, + "TACGAG": 1208, + "TACGTA": 1209, + "TACGTT": 1210, + "TACGTC": 1211, + "TACGTG": 1212, + "TACGCA": 1213, + "TACGCT": 1214, + "TACGCC": 1215, + "TACGCG": 1216, + "TACGGA": 1217, + "TACGGT": 1218, + "TACGGC": 1219, + "TACGGG": 1220, + "TAGAAA": 1221, + "TAGAAT": 1222, + "TAGAAC": 1223, + "TAGAAG": 1224, + "TAGATA": 1225, + "TAGATT": 1226, + "TAGATC": 1227, + "TAGATG": 1228, + "TAGACA": 1229, + "TAGACT": 1230, + "TAGACC": 1231, + "TAGACG": 1232, + "TAGAGA": 1233, + "TAGAGT": 1234, + "TAGAGC": 1235, + "TAGAGG": 1236, + "TAGTAA": 1237, + "TAGTAT": 1238, + "TAGTAC": 1239, + "TAGTAG": 1240, + "TAGTTA": 1241, + "TAGTTT": 1242, + "TAGTTC": 1243, + "TAGTTG": 1244, + "TAGTCA": 1245, + "TAGTCT": 1246, + "TAGTCC": 1247, + "TAGTCG": 1248, + "TAGTGA": 1249, + "TAGTGT": 1250, + "TAGTGC": 1251, + "TAGTGG": 1252, + "TAGCAA": 1253, + "TAGCAT": 1254, + "TAGCAC": 1255, + "TAGCAG": 1256, + "TAGCTA": 1257, + "TAGCTT": 1258, + "TAGCTC": 1259, + "TAGCTG": 1260, + "TAGCCA": 1261, + "TAGCCT": 1262, + "TAGCCC": 1263, + "TAGCCG": 1264, + "TAGCGA": 1265, + "TAGCGT": 1266, + "TAGCGC": 1267, + "TAGCGG": 1268, + "TAGGAA": 1269, + "TAGGAT": 1270, + "TAGGAC": 1271, + "TAGGAG": 1272, + "TAGGTA": 1273, + "TAGGTT": 1274, + "TAGGTC": 1275, + "TAGGTG": 1276, + "TAGGCA": 1277, + "TAGGCT": 1278, + "TAGGCC": 1279, + "TAGGCG": 1280, + "TAGGGA": 1281, + "TAGGGT": 1282, + "TAGGGC": 1283, + "TAGGGG": 1284, + "TTAAAA": 1285, + "TTAAAT": 1286, + "TTAAAC": 1287, + "TTAAAG": 1288, + "TTAATA": 1289, + "TTAATT": 1290, + "TTAATC": 1291, + "TTAATG": 1292, + "TTAACA": 1293, + "TTAACT": 1294, + "TTAACC": 1295, + "TTAACG": 1296, + "TTAAGA": 1297, + "TTAAGT": 1298, + "TTAAGC": 1299, + "TTAAGG": 1300, + "TTATAA": 1301, + "TTATAT": 1302, + "TTATAC": 1303, + "TTATAG": 1304, + "TTATTA": 1305, + "TTATTT": 1306, + "TTATTC": 1307, + "TTATTG": 1308, + "TTATCA": 1309, + "TTATCT": 1310, + "TTATCC": 1311, + "TTATCG": 1312, + "TTATGA": 1313, + "TTATGT": 1314, + "TTATGC": 1315, + "TTATGG": 1316, + "TTACAA": 1317, + "TTACAT": 1318, + "TTACAC": 1319, + "TTACAG": 1320, + "TTACTA": 1321, + "TTACTT": 1322, + "TTACTC": 1323, + "TTACTG": 1324, + "TTACCA": 1325, + "TTACCT": 1326, + "TTACCC": 1327, + "TTACCG": 1328, + "TTACGA": 1329, + "TTACGT": 1330, + "TTACGC": 1331, + "TTACGG": 1332, + "TTAGAA": 1333, + "TTAGAT": 1334, + "TTAGAC": 1335, + "TTAGAG": 1336, + "TTAGTA": 1337, + "TTAGTT": 1338, + "TTAGTC": 1339, + "TTAGTG": 1340, + "TTAGCA": 1341, + "TTAGCT": 1342, + "TTAGCC": 1343, + "TTAGCG": 1344, + "TTAGGA": 1345, + "TTAGGT": 1346, + "TTAGGC": 1347, + "TTAGGG": 1348, + "TTTAAA": 1349, + "TTTAAT": 1350, + "TTTAAC": 1351, + "TTTAAG": 1352, + "TTTATA": 1353, + "TTTATT": 1354, + "TTTATC": 1355, + "TTTATG": 1356, + "TTTACA": 1357, + "TTTACT": 1358, + "TTTACC": 1359, + "TTTACG": 1360, + "TTTAGA": 1361, + "TTTAGT": 1362, + "TTTAGC": 1363, + "TTTAGG": 1364, + "TTTTAA": 1365, + "TTTTAT": 1366, + "TTTTAC": 1367, + "TTTTAG": 1368, + "TTTTTA": 1369, + "TTTTTT": 1370, + "TTTTTC": 1371, + "TTTTTG": 1372, + "TTTTCA": 1373, + "TTTTCT": 1374, + "TTTTCC": 1375, + "TTTTCG": 1376, + "TTTTGA": 1377, + "TTTTGT": 1378, + "TTTTGC": 1379, + "TTTTGG": 1380, + "TTTCAA": 1381, + "TTTCAT": 1382, + "TTTCAC": 1383, + "TTTCAG": 1384, + "TTTCTA": 1385, + "TTTCTT": 1386, + "TTTCTC": 1387, + "TTTCTG": 1388, + "TTTCCA": 1389, + "TTTCCT": 1390, + "TTTCCC": 1391, + "TTTCCG": 1392, + "TTTCGA": 1393, + "TTTCGT": 1394, + "TTTCGC": 1395, + "TTTCGG": 1396, + "TTTGAA": 1397, + "TTTGAT": 1398, + "TTTGAC": 1399, + "TTTGAG": 1400, + "TTTGTA": 1401, + "TTTGTT": 1402, + "TTTGTC": 1403, + "TTTGTG": 1404, + "TTTGCA": 1405, + "TTTGCT": 1406, + "TTTGCC": 1407, + "TTTGCG": 1408, + "TTTGGA": 1409, + "TTTGGT": 1410, + "TTTGGC": 1411, + "TTTGGG": 1412, + "TTCAAA": 1413, + "TTCAAT": 1414, + "TTCAAC": 1415, + "TTCAAG": 1416, + "TTCATA": 1417, + "TTCATT": 1418, + "TTCATC": 1419, + "TTCATG": 1420, + "TTCACA": 1421, + "TTCACT": 1422, + "TTCACC": 1423, + "TTCACG": 1424, + "TTCAGA": 1425, + "TTCAGT": 1426, + "TTCAGC": 1427, + "TTCAGG": 1428, + "TTCTAA": 1429, + "TTCTAT": 1430, + "TTCTAC": 1431, + "TTCTAG": 1432, + "TTCTTA": 1433, + "TTCTTT": 1434, + "TTCTTC": 1435, + "TTCTTG": 1436, + "TTCTCA": 1437, + "TTCTCT": 1438, + "TTCTCC": 1439, + "TTCTCG": 1440, + "TTCTGA": 1441, + "TTCTGT": 1442, + "TTCTGC": 1443, + "TTCTGG": 1444, + "TTCCAA": 1445, + "TTCCAT": 1446, + "TTCCAC": 1447, + "TTCCAG": 1448, + "TTCCTA": 1449, + "TTCCTT": 1450, + "TTCCTC": 1451, + "TTCCTG": 1452, + "TTCCCA": 1453, + "TTCCCT": 1454, + "TTCCCC": 1455, + "TTCCCG": 1456, + "TTCCGA": 1457, + "TTCCGT": 1458, + "TTCCGC": 1459, + "TTCCGG": 1460, + "TTCGAA": 1461, + "TTCGAT": 1462, + "TTCGAC": 1463, + "TTCGAG": 1464, + "TTCGTA": 1465, + "TTCGTT": 1466, + "TTCGTC": 1467, + "TTCGTG": 1468, + "TTCGCA": 1469, + "TTCGCT": 1470, + "TTCGCC": 1471, + "TTCGCG": 1472, + "TTCGGA": 1473, + "TTCGGT": 1474, + "TTCGGC": 1475, + "TTCGGG": 1476, + "TTGAAA": 1477, + "TTGAAT": 1478, + "TTGAAC": 1479, + "TTGAAG": 1480, + "TTGATA": 1481, + "TTGATT": 1482, + "TTGATC": 1483, + "TTGATG": 1484, + "TTGACA": 1485, + "TTGACT": 1486, + "TTGACC": 1487, + "TTGACG": 1488, + "TTGAGA": 1489, + "TTGAGT": 1490, + "TTGAGC": 1491, + "TTGAGG": 1492, + "TTGTAA": 1493, + "TTGTAT": 1494, + "TTGTAC": 1495, + "TTGTAG": 1496, + "TTGTTA": 1497, + "TTGTTT": 1498, + "TTGTTC": 1499, + "TTGTTG": 1500, + "TTGTCA": 1501, + "TTGTCT": 1502, + "TTGTCC": 1503, + "TTGTCG": 1504, + "TTGTGA": 1505, + "TTGTGT": 1506, + "TTGTGC": 1507, + "TTGTGG": 1508, + "TTGCAA": 1509, + "TTGCAT": 1510, + "TTGCAC": 1511, + "TTGCAG": 1512, + "TTGCTA": 1513, + "TTGCTT": 1514, + "TTGCTC": 1515, + "TTGCTG": 1516, + "TTGCCA": 1517, + "TTGCCT": 1518, + "TTGCCC": 1519, + "TTGCCG": 1520, + "TTGCGA": 1521, + "TTGCGT": 1522, + "TTGCGC": 1523, + "TTGCGG": 1524, + "TTGGAA": 1525, + "TTGGAT": 1526, + "TTGGAC": 1527, + "TTGGAG": 1528, + "TTGGTA": 1529, + "TTGGTT": 1530, + "TTGGTC": 1531, + "TTGGTG": 1532, + "TTGGCA": 1533, + "TTGGCT": 1534, + "TTGGCC": 1535, + "TTGGCG": 1536, + "TTGGGA": 1537, + "TTGGGT": 1538, + "TTGGGC": 1539, + "TTGGGG": 1540, + "TCAAAA": 1541, + "TCAAAT": 1542, + "TCAAAC": 1543, + "TCAAAG": 1544, + "TCAATA": 1545, + "TCAATT": 1546, + "TCAATC": 1547, + "TCAATG": 1548, + "TCAACA": 1549, + "TCAACT": 1550, + "TCAACC": 1551, + "TCAACG": 1552, + "TCAAGA": 1553, + "TCAAGT": 1554, + "TCAAGC": 1555, + "TCAAGG": 1556, + "TCATAA": 1557, + "TCATAT": 1558, + "TCATAC": 1559, + "TCATAG": 1560, + "TCATTA": 1561, + "TCATTT": 1562, + "TCATTC": 1563, + "TCATTG": 1564, + "TCATCA": 1565, + "TCATCT": 1566, + "TCATCC": 1567, + "TCATCG": 1568, + "TCATGA": 1569, + "TCATGT": 1570, + "TCATGC": 1571, + "TCATGG": 1572, + "TCACAA": 1573, + "TCACAT": 1574, + "TCACAC": 1575, + "TCACAG": 1576, + "TCACTA": 1577, + "TCACTT": 1578, + "TCACTC": 1579, + "TCACTG": 1580, + "TCACCA": 1581, + "TCACCT": 1582, + "TCACCC": 1583, + "TCACCG": 1584, + "TCACGA": 1585, + "TCACGT": 1586, + "TCACGC": 1587, + "TCACGG": 1588, + "TCAGAA": 1589, + "TCAGAT": 1590, + "TCAGAC": 1591, + "TCAGAG": 1592, + "TCAGTA": 1593, + "TCAGTT": 1594, + "TCAGTC": 1595, + "TCAGTG": 1596, + "TCAGCA": 1597, + "TCAGCT": 1598, + "TCAGCC": 1599, + "TCAGCG": 1600, + "TCAGGA": 1601, + "TCAGGT": 1602, + "TCAGGC": 1603, + "TCAGGG": 1604, + "TCTAAA": 1605, + "TCTAAT": 1606, + "TCTAAC": 1607, + "TCTAAG": 1608, + "TCTATA": 1609, + "TCTATT": 1610, + "TCTATC": 1611, + "TCTATG": 1612, + "TCTACA": 1613, + "TCTACT": 1614, + "TCTACC": 1615, + "TCTACG": 1616, + "TCTAGA": 1617, + "TCTAGT": 1618, + "TCTAGC": 1619, + "TCTAGG": 1620, + "TCTTAA": 1621, + "TCTTAT": 1622, + "TCTTAC": 1623, + "TCTTAG": 1624, + "TCTTTA": 1625, + "TCTTTT": 1626, + "TCTTTC": 1627, + "TCTTTG": 1628, + "TCTTCA": 1629, + "TCTTCT": 1630, + "TCTTCC": 1631, + "TCTTCG": 1632, + "TCTTGA": 1633, + "TCTTGT": 1634, + "TCTTGC": 1635, + "TCTTGG": 1636, + "TCTCAA": 1637, + "TCTCAT": 1638, + "TCTCAC": 1639, + "TCTCAG": 1640, + "TCTCTA": 1641, + "TCTCTT": 1642, + "TCTCTC": 1643, + "TCTCTG": 1644, + "TCTCCA": 1645, + "TCTCCT": 1646, + "TCTCCC": 1647, + "TCTCCG": 1648, + "TCTCGA": 1649, + "TCTCGT": 1650, + "TCTCGC": 1651, + "TCTCGG": 1652, + "TCTGAA": 1653, + "TCTGAT": 1654, + "TCTGAC": 1655, + "TCTGAG": 1656, + "TCTGTA": 1657, + "TCTGTT": 1658, + "TCTGTC": 1659, + "TCTGTG": 1660, + "TCTGCA": 1661, + "TCTGCT": 1662, + "TCTGCC": 1663, + "TCTGCG": 1664, + "TCTGGA": 1665, + "TCTGGT": 1666, + "TCTGGC": 1667, + "TCTGGG": 1668, + "TCCAAA": 1669, + "TCCAAT": 1670, + "TCCAAC": 1671, + "TCCAAG": 1672, + "TCCATA": 1673, + "TCCATT": 1674, + "TCCATC": 1675, + "TCCATG": 1676, + "TCCACA": 1677, + "TCCACT": 1678, + "TCCACC": 1679, + "TCCACG": 1680, + "TCCAGA": 1681, + "TCCAGT": 1682, + "TCCAGC": 1683, + "TCCAGG": 1684, + "TCCTAA": 1685, + "TCCTAT": 1686, + "TCCTAC": 1687, + "TCCTAG": 1688, + "TCCTTA": 1689, + "TCCTTT": 1690, + "TCCTTC": 1691, + "TCCTTG": 1692, + "TCCTCA": 1693, + "TCCTCT": 1694, + "TCCTCC": 1695, + "TCCTCG": 1696, + "TCCTGA": 1697, + "TCCTGT": 1698, + "TCCTGC": 1699, + "TCCTGG": 1700, + "TCCCAA": 1701, + "TCCCAT": 1702, + "TCCCAC": 1703, + "TCCCAG": 1704, + "TCCCTA": 1705, + "TCCCTT": 1706, + "TCCCTC": 1707, + "TCCCTG": 1708, + "TCCCCA": 1709, + "TCCCCT": 1710, + "TCCCCC": 1711, + "TCCCCG": 1712, + "TCCCGA": 1713, + "TCCCGT": 1714, + "TCCCGC": 1715, + "TCCCGG": 1716, + "TCCGAA": 1717, + "TCCGAT": 1718, + "TCCGAC": 1719, + "TCCGAG": 1720, + "TCCGTA": 1721, + "TCCGTT": 1722, + "TCCGTC": 1723, + "TCCGTG": 1724, + "TCCGCA": 1725, + "TCCGCT": 1726, + "TCCGCC": 1727, + "TCCGCG": 1728, + "TCCGGA": 1729, + "TCCGGT": 1730, + "TCCGGC": 1731, + "TCCGGG": 1732, + "TCGAAA": 1733, + "TCGAAT": 1734, + "TCGAAC": 1735, + "TCGAAG": 1736, + "TCGATA": 1737, + "TCGATT": 1738, + "TCGATC": 1739, + "TCGATG": 1740, + "TCGACA": 1741, + "TCGACT": 1742, + "TCGACC": 1743, + "TCGACG": 1744, + "TCGAGA": 1745, + "TCGAGT": 1746, + "TCGAGC": 1747, + "TCGAGG": 1748, + "TCGTAA": 1749, + "TCGTAT": 1750, + "TCGTAC": 1751, + "TCGTAG": 1752, + "TCGTTA": 1753, + "TCGTTT": 1754, + "TCGTTC": 1755, + "TCGTTG": 1756, + "TCGTCA": 1757, + "TCGTCT": 1758, + "TCGTCC": 1759, + "TCGTCG": 1760, + "TCGTGA": 1761, + "TCGTGT": 1762, + "TCGTGC": 1763, + "TCGTGG": 1764, + "TCGCAA": 1765, + "TCGCAT": 1766, + "TCGCAC": 1767, + "TCGCAG": 1768, + "TCGCTA": 1769, + "TCGCTT": 1770, + "TCGCTC": 1771, + "TCGCTG": 1772, + "TCGCCA": 1773, + "TCGCCT": 1774, + "TCGCCC": 1775, + "TCGCCG": 1776, + "TCGCGA": 1777, + "TCGCGT": 1778, + "TCGCGC": 1779, + "TCGCGG": 1780, + "TCGGAA": 1781, + "TCGGAT": 1782, + "TCGGAC": 1783, + "TCGGAG": 1784, + "TCGGTA": 1785, + "TCGGTT": 1786, + "TCGGTC": 1787, + "TCGGTG": 1788, + "TCGGCA": 1789, + "TCGGCT": 1790, + "TCGGCC": 1791, + "TCGGCG": 1792, + "TCGGGA": 1793, + "TCGGGT": 1794, + "TCGGGC": 1795, + "TCGGGG": 1796, + "TGAAAA": 1797, + "TGAAAT": 1798, + "TGAAAC": 1799, + "TGAAAG": 1800, + "TGAATA": 1801, + "TGAATT": 1802, + "TGAATC": 1803, + "TGAATG": 1804, + "TGAACA": 1805, + "TGAACT": 1806, + "TGAACC": 1807, + "TGAACG": 1808, + "TGAAGA": 1809, + "TGAAGT": 1810, + "TGAAGC": 1811, + "TGAAGG": 1812, + "TGATAA": 1813, + "TGATAT": 1814, + "TGATAC": 1815, + "TGATAG": 1816, + "TGATTA": 1817, + "TGATTT": 1818, + "TGATTC": 1819, + "TGATTG": 1820, + "TGATCA": 1821, + "TGATCT": 1822, + "TGATCC": 1823, + "TGATCG": 1824, + "TGATGA": 1825, + "TGATGT": 1826, + "TGATGC": 1827, + "TGATGG": 1828, + "TGACAA": 1829, + "TGACAT": 1830, + "TGACAC": 1831, + "TGACAG": 1832, + "TGACTA": 1833, + "TGACTT": 1834, + "TGACTC": 1835, + "TGACTG": 1836, + "TGACCA": 1837, + "TGACCT": 1838, + "TGACCC": 1839, + "TGACCG": 1840, + "TGACGA": 1841, + "TGACGT": 1842, + "TGACGC": 1843, + "TGACGG": 1844, + "TGAGAA": 1845, + "TGAGAT": 1846, + "TGAGAC": 1847, + "TGAGAG": 1848, + "TGAGTA": 1849, + "TGAGTT": 1850, + "TGAGTC": 1851, + "TGAGTG": 1852, + "TGAGCA": 1853, + "TGAGCT": 1854, + "TGAGCC": 1855, + "TGAGCG": 1856, + "TGAGGA": 1857, + "TGAGGT": 1858, + "TGAGGC": 1859, + "TGAGGG": 1860, + "TGTAAA": 1861, + "TGTAAT": 1862, + "TGTAAC": 1863, + "TGTAAG": 1864, + "TGTATA": 1865, + "TGTATT": 1866, + "TGTATC": 1867, + "TGTATG": 1868, + "TGTACA": 1869, + "TGTACT": 1870, + "TGTACC": 1871, + "TGTACG": 1872, + "TGTAGA": 1873, + "TGTAGT": 1874, + "TGTAGC": 1875, + "TGTAGG": 1876, + "TGTTAA": 1877, + "TGTTAT": 1878, + "TGTTAC": 1879, + "TGTTAG": 1880, + "TGTTTA": 1881, + "TGTTTT": 1882, + "TGTTTC": 1883, + "TGTTTG": 1884, + "TGTTCA": 1885, + "TGTTCT": 1886, + "TGTTCC": 1887, + "TGTTCG": 1888, + "TGTTGA": 1889, + "TGTTGT": 1890, + "TGTTGC": 1891, + "TGTTGG": 1892, + "TGTCAA": 1893, + "TGTCAT": 1894, + "TGTCAC": 1895, + "TGTCAG": 1896, + "TGTCTA": 1897, + "TGTCTT": 1898, + "TGTCTC": 1899, + "TGTCTG": 1900, + "TGTCCA": 1901, + "TGTCCT": 1902, + "TGTCCC": 1903, + "TGTCCG": 1904, + "TGTCGA": 1905, + "TGTCGT": 1906, + "TGTCGC": 1907, + "TGTCGG": 1908, + "TGTGAA": 1909, + "TGTGAT": 1910, + "TGTGAC": 1911, + "TGTGAG": 1912, + "TGTGTA": 1913, + "TGTGTT": 1914, + "TGTGTC": 1915, + "TGTGTG": 1916, + "TGTGCA": 1917, + "TGTGCT": 1918, + "TGTGCC": 1919, + "TGTGCG": 1920, + "TGTGGA": 1921, + "TGTGGT": 1922, + "TGTGGC": 1923, + "TGTGGG": 1924, + "TGCAAA": 1925, + "TGCAAT": 1926, + "TGCAAC": 1927, + "TGCAAG": 1928, + "TGCATA": 1929, + "TGCATT": 1930, + "TGCATC": 1931, + "TGCATG": 1932, + "TGCACA": 1933, + "TGCACT": 1934, + "TGCACC": 1935, + "TGCACG": 1936, + "TGCAGA": 1937, + "TGCAGT": 1938, + "TGCAGC": 1939, + "TGCAGG": 1940, + "TGCTAA": 1941, + "TGCTAT": 1942, + "TGCTAC": 1943, + "TGCTAG": 1944, + "TGCTTA": 1945, + "TGCTTT": 1946, + "TGCTTC": 1947, + "TGCTTG": 1948, + "TGCTCA": 1949, + "TGCTCT": 1950, + "TGCTCC": 1951, + "TGCTCG": 1952, + "TGCTGA": 1953, + "TGCTGT": 1954, + "TGCTGC": 1955, + "TGCTGG": 1956, + "TGCCAA": 1957, + "TGCCAT": 1958, + "TGCCAC": 1959, + "TGCCAG": 1960, + "TGCCTA": 1961, + "TGCCTT": 1962, + "TGCCTC": 1963, + "TGCCTG": 1964, + "TGCCCA": 1965, + "TGCCCT": 1966, + "TGCCCC": 1967, + "TGCCCG": 1968, + "TGCCGA": 1969, + "TGCCGT": 1970, + "TGCCGC": 1971, + "TGCCGG": 1972, + "TGCGAA": 1973, + "TGCGAT": 1974, + "TGCGAC": 1975, + "TGCGAG": 1976, + "TGCGTA": 1977, + "TGCGTT": 1978, + "TGCGTC": 1979, + "TGCGTG": 1980, + "TGCGCA": 1981, + "TGCGCT": 1982, + "TGCGCC": 1983, + "TGCGCG": 1984, + "TGCGGA": 1985, + "TGCGGT": 1986, + "TGCGGC": 1987, + "TGCGGG": 1988, + "TGGAAA": 1989, + "TGGAAT": 1990, + "TGGAAC": 1991, + "TGGAAG": 1992, + "TGGATA": 1993, + "TGGATT": 1994, + "TGGATC": 1995, + "TGGATG": 1996, + "TGGACA": 1997, + "TGGACT": 1998, + "TGGACC": 1999, + "TGGACG": 2000, + "TGGAGA": 2001, + "TGGAGT": 2002, + "TGGAGC": 2003, + "TGGAGG": 2004, + "TGGTAA": 2005, + "TGGTAT": 2006, + "TGGTAC": 2007, + "TGGTAG": 2008, + "TGGTTA": 2009, + "TGGTTT": 2010, + "TGGTTC": 2011, + "TGGTTG": 2012, + "TGGTCA": 2013, + "TGGTCT": 2014, + "TGGTCC": 2015, + "TGGTCG": 2016, + "TGGTGA": 2017, + "TGGTGT": 2018, + "TGGTGC": 2019, + "TGGTGG": 2020, + "TGGCAA": 2021, + "TGGCAT": 2022, + "TGGCAC": 2023, + "TGGCAG": 2024, + "TGGCTA": 2025, + "TGGCTT": 2026, + "TGGCTC": 2027, + "TGGCTG": 2028, + "TGGCCA": 2029, + "TGGCCT": 2030, + "TGGCCC": 2031, + "TGGCCG": 2032, + "TGGCGA": 2033, + "TGGCGT": 2034, + "TGGCGC": 2035, + "TGGCGG": 2036, + "TGGGAA": 2037, + "TGGGAT": 2038, + "TGGGAC": 2039, + "TGGGAG": 2040, + "TGGGTA": 2041, + "TGGGTT": 2042, + "TGGGTC": 2043, + "TGGGTG": 2044, + "TGGGCA": 2045, + "TGGGCT": 2046, + "TGGGCC": 2047, + "TGGGCG": 2048, + "TGGGGA": 2049, + "TGGGGT": 2050, + "TGGGGC": 2051, + "TGGGGG": 2052, + "CAAAAA": 2053, + "CAAAAT": 2054, + "CAAAAC": 2055, + "CAAAAG": 2056, + "CAAATA": 2057, + "CAAATT": 2058, + "CAAATC": 2059, + "CAAATG": 2060, + "CAAACA": 2061, + "CAAACT": 2062, + "CAAACC": 2063, + "CAAACG": 2064, + "CAAAGA": 2065, + "CAAAGT": 2066, + "CAAAGC": 2067, + "CAAAGG": 2068, + "CAATAA": 2069, + "CAATAT": 2070, + "CAATAC": 2071, + "CAATAG": 2072, + "CAATTA": 2073, + "CAATTT": 2074, + "CAATTC": 2075, + "CAATTG": 2076, + "CAATCA": 2077, + "CAATCT": 2078, + "CAATCC": 2079, + "CAATCG": 2080, + "CAATGA": 2081, + "CAATGT": 2082, + "CAATGC": 2083, + "CAATGG": 2084, + "CAACAA": 2085, + "CAACAT": 2086, + "CAACAC": 2087, + "CAACAG": 2088, + "CAACTA": 2089, + "CAACTT": 2090, + "CAACTC": 2091, + "CAACTG": 2092, + "CAACCA": 2093, + "CAACCT": 2094, + "CAACCC": 2095, + "CAACCG": 2096, + "CAACGA": 2097, + "CAACGT": 2098, + "CAACGC": 2099, + "CAACGG": 2100, + "CAAGAA": 2101, + "CAAGAT": 2102, + "CAAGAC": 2103, + "CAAGAG": 2104, + "CAAGTA": 2105, + "CAAGTT": 2106, + "CAAGTC": 2107, + "CAAGTG": 2108, + "CAAGCA": 2109, + "CAAGCT": 2110, + "CAAGCC": 2111, + "CAAGCG": 2112, + "CAAGGA": 2113, + "CAAGGT": 2114, + "CAAGGC": 2115, + "CAAGGG": 2116, + "CATAAA": 2117, + "CATAAT": 2118, + "CATAAC": 2119, + "CATAAG": 2120, + "CATATA": 2121, + "CATATT": 2122, + "CATATC": 2123, + "CATATG": 2124, + "CATACA": 2125, + "CATACT": 2126, + "CATACC": 2127, + "CATACG": 2128, + "CATAGA": 2129, + "CATAGT": 2130, + "CATAGC": 2131, + "CATAGG": 2132, + "CATTAA": 2133, + "CATTAT": 2134, + "CATTAC": 2135, + "CATTAG": 2136, + "CATTTA": 2137, + "CATTTT": 2138, + "CATTTC": 2139, + "CATTTG": 2140, + "CATTCA": 2141, + "CATTCT": 2142, + "CATTCC": 2143, + "CATTCG": 2144, + "CATTGA": 2145, + "CATTGT": 2146, + "CATTGC": 2147, + "CATTGG": 2148, + "CATCAA": 2149, + "CATCAT": 2150, + "CATCAC": 2151, + "CATCAG": 2152, + "CATCTA": 2153, + "CATCTT": 2154, + "CATCTC": 2155, + "CATCTG": 2156, + "CATCCA": 2157, + "CATCCT": 2158, + "CATCCC": 2159, + "CATCCG": 2160, + "CATCGA": 2161, + "CATCGT": 2162, + "CATCGC": 2163, + "CATCGG": 2164, + "CATGAA": 2165, + "CATGAT": 2166, + "CATGAC": 2167, + "CATGAG": 2168, + "CATGTA": 2169, + "CATGTT": 2170, + "CATGTC": 2171, + "CATGTG": 2172, + "CATGCA": 2173, + "CATGCT": 2174, + "CATGCC": 2175, + "CATGCG": 2176, + "CATGGA": 2177, + "CATGGT": 2178, + "CATGGC": 2179, + "CATGGG": 2180, + "CACAAA": 2181, + "CACAAT": 2182, + "CACAAC": 2183, + "CACAAG": 2184, + "CACATA": 2185, + "CACATT": 2186, + "CACATC": 2187, + "CACATG": 2188, + "CACACA": 2189, + "CACACT": 2190, + "CACACC": 2191, + "CACACG": 2192, + "CACAGA": 2193, + "CACAGT": 2194, + "CACAGC": 2195, + "CACAGG": 2196, + "CACTAA": 2197, + "CACTAT": 2198, + "CACTAC": 2199, + "CACTAG": 2200, + "CACTTA": 2201, + "CACTTT": 2202, + "CACTTC": 2203, + "CACTTG": 2204, + "CACTCA": 2205, + "CACTCT": 2206, + "CACTCC": 2207, + "CACTCG": 2208, + "CACTGA": 2209, + "CACTGT": 2210, + "CACTGC": 2211, + "CACTGG": 2212, + "CACCAA": 2213, + "CACCAT": 2214, + "CACCAC": 2215, + "CACCAG": 2216, + "CACCTA": 2217, + "CACCTT": 2218, + "CACCTC": 2219, + "CACCTG": 2220, + "CACCCA": 2221, + "CACCCT": 2222, + "CACCCC": 2223, + "CACCCG": 2224, + "CACCGA": 2225, + "CACCGT": 2226, + "CACCGC": 2227, + "CACCGG": 2228, + "CACGAA": 2229, + "CACGAT": 2230, + "CACGAC": 2231, + "CACGAG": 2232, + "CACGTA": 2233, + "CACGTT": 2234, + "CACGTC": 2235, + "CACGTG": 2236, + "CACGCA": 2237, + "CACGCT": 2238, + "CACGCC": 2239, + "CACGCG": 2240, + "CACGGA": 2241, + "CACGGT": 2242, + "CACGGC": 2243, + "CACGGG": 2244, + "CAGAAA": 2245, + "CAGAAT": 2246, + "CAGAAC": 2247, + "CAGAAG": 2248, + "CAGATA": 2249, + "CAGATT": 2250, + "CAGATC": 2251, + "CAGATG": 2252, + "CAGACA": 2253, + "CAGACT": 2254, + "CAGACC": 2255, + "CAGACG": 2256, + "CAGAGA": 2257, + "CAGAGT": 2258, + "CAGAGC": 2259, + "CAGAGG": 2260, + "CAGTAA": 2261, + "CAGTAT": 2262, + "CAGTAC": 2263, + "CAGTAG": 2264, + "CAGTTA": 2265, + "CAGTTT": 2266, + "CAGTTC": 2267, + "CAGTTG": 2268, + "CAGTCA": 2269, + "CAGTCT": 2270, + "CAGTCC": 2271, + "CAGTCG": 2272, + "CAGTGA": 2273, + "CAGTGT": 2274, + "CAGTGC": 2275, + "CAGTGG": 2276, + "CAGCAA": 2277, + "CAGCAT": 2278, + "CAGCAC": 2279, + "CAGCAG": 2280, + "CAGCTA": 2281, + "CAGCTT": 2282, + "CAGCTC": 2283, + "CAGCTG": 2284, + "CAGCCA": 2285, + "CAGCCT": 2286, + "CAGCCC": 2287, + "CAGCCG": 2288, + "CAGCGA": 2289, + "CAGCGT": 2290, + "CAGCGC": 2291, + "CAGCGG": 2292, + "CAGGAA": 2293, + "CAGGAT": 2294, + "CAGGAC": 2295, + "CAGGAG": 2296, + "CAGGTA": 2297, + "CAGGTT": 2298, + "CAGGTC": 2299, + "CAGGTG": 2300, + "CAGGCA": 2301, + "CAGGCT": 2302, + "CAGGCC": 2303, + "CAGGCG": 2304, + "CAGGGA": 2305, + "CAGGGT": 2306, + "CAGGGC": 2307, + "CAGGGG": 2308, + "CTAAAA": 2309, + "CTAAAT": 2310, + "CTAAAC": 2311, + "CTAAAG": 2312, + "CTAATA": 2313, + "CTAATT": 2314, + "CTAATC": 2315, + "CTAATG": 2316, + "CTAACA": 2317, + "CTAACT": 2318, + "CTAACC": 2319, + "CTAACG": 2320, + "CTAAGA": 2321, + "CTAAGT": 2322, + "CTAAGC": 2323, + "CTAAGG": 2324, + "CTATAA": 2325, + "CTATAT": 2326, + "CTATAC": 2327, + "CTATAG": 2328, + "CTATTA": 2329, + "CTATTT": 2330, + "CTATTC": 2331, + "CTATTG": 2332, + "CTATCA": 2333, + "CTATCT": 2334, + "CTATCC": 2335, + "CTATCG": 2336, + "CTATGA": 2337, + "CTATGT": 2338, + "CTATGC": 2339, + "CTATGG": 2340, + "CTACAA": 2341, + "CTACAT": 2342, + "CTACAC": 2343, + "CTACAG": 2344, + "CTACTA": 2345, + "CTACTT": 2346, + "CTACTC": 2347, + "CTACTG": 2348, + "CTACCA": 2349, + "CTACCT": 2350, + "CTACCC": 2351, + "CTACCG": 2352, + "CTACGA": 2353, + "CTACGT": 2354, + "CTACGC": 2355, + "CTACGG": 2356, + "CTAGAA": 2357, + "CTAGAT": 2358, + "CTAGAC": 2359, + "CTAGAG": 2360, + "CTAGTA": 2361, + "CTAGTT": 2362, + "CTAGTC": 2363, + "CTAGTG": 2364, + "CTAGCA": 2365, + "CTAGCT": 2366, + "CTAGCC": 2367, + "CTAGCG": 2368, + "CTAGGA": 2369, + "CTAGGT": 2370, + "CTAGGC": 2371, + "CTAGGG": 2372, + "CTTAAA": 2373, + "CTTAAT": 2374, + "CTTAAC": 2375, + "CTTAAG": 2376, + "CTTATA": 2377, + "CTTATT": 2378, + "CTTATC": 2379, + "CTTATG": 2380, + "CTTACA": 2381, + "CTTACT": 2382, + "CTTACC": 2383, + "CTTACG": 2384, + "CTTAGA": 2385, + "CTTAGT": 2386, + "CTTAGC": 2387, + "CTTAGG": 2388, + "CTTTAA": 2389, + "CTTTAT": 2390, + "CTTTAC": 2391, + "CTTTAG": 2392, + "CTTTTA": 2393, + "CTTTTT": 2394, + "CTTTTC": 2395, + "CTTTTG": 2396, + "CTTTCA": 2397, + "CTTTCT": 2398, + "CTTTCC": 2399, + "CTTTCG": 2400, + "CTTTGA": 2401, + "CTTTGT": 2402, + "CTTTGC": 2403, + "CTTTGG": 2404, + "CTTCAA": 2405, + "CTTCAT": 2406, + "CTTCAC": 2407, + "CTTCAG": 2408, + "CTTCTA": 2409, + "CTTCTT": 2410, + "CTTCTC": 2411, + "CTTCTG": 2412, + "CTTCCA": 2413, + "CTTCCT": 2414, + "CTTCCC": 2415, + "CTTCCG": 2416, + "CTTCGA": 2417, + "CTTCGT": 2418, + "CTTCGC": 2419, + "CTTCGG": 2420, + "CTTGAA": 2421, + "CTTGAT": 2422, + "CTTGAC": 2423, + "CTTGAG": 2424, + "CTTGTA": 2425, + "CTTGTT": 2426, + "CTTGTC": 2427, + "CTTGTG": 2428, + "CTTGCA": 2429, + "CTTGCT": 2430, + "CTTGCC": 2431, + "CTTGCG": 2432, + "CTTGGA": 2433, + "CTTGGT": 2434, + "CTTGGC": 2435, + "CTTGGG": 2436, + "CTCAAA": 2437, + "CTCAAT": 2438, + "CTCAAC": 2439, + "CTCAAG": 2440, + "CTCATA": 2441, + "CTCATT": 2442, + "CTCATC": 2443, + "CTCATG": 2444, + "CTCACA": 2445, + "CTCACT": 2446, + "CTCACC": 2447, + "CTCACG": 2448, + "CTCAGA": 2449, + "CTCAGT": 2450, + "CTCAGC": 2451, + "CTCAGG": 2452, + "CTCTAA": 2453, + "CTCTAT": 2454, + "CTCTAC": 2455, + "CTCTAG": 2456, + "CTCTTA": 2457, + "CTCTTT": 2458, + "CTCTTC": 2459, + "CTCTTG": 2460, + "CTCTCA": 2461, + "CTCTCT": 2462, + "CTCTCC": 2463, + "CTCTCG": 2464, + "CTCTGA": 2465, + "CTCTGT": 2466, + "CTCTGC": 2467, + "CTCTGG": 2468, + "CTCCAA": 2469, + "CTCCAT": 2470, + "CTCCAC": 2471, + "CTCCAG": 2472, + "CTCCTA": 2473, + "CTCCTT": 2474, + "CTCCTC": 2475, + "CTCCTG": 2476, + "CTCCCA": 2477, + "CTCCCT": 2478, + "CTCCCC": 2479, + "CTCCCG": 2480, + "CTCCGA": 2481, + "CTCCGT": 2482, + "CTCCGC": 2483, + "CTCCGG": 2484, + "CTCGAA": 2485, + "CTCGAT": 2486, + "CTCGAC": 2487, + "CTCGAG": 2488, + "CTCGTA": 2489, + "CTCGTT": 2490, + "CTCGTC": 2491, + "CTCGTG": 2492, + "CTCGCA": 2493, + "CTCGCT": 2494, + "CTCGCC": 2495, + "CTCGCG": 2496, + "CTCGGA": 2497, + "CTCGGT": 2498, + "CTCGGC": 2499, + "CTCGGG": 2500, + "CTGAAA": 2501, + "CTGAAT": 2502, + "CTGAAC": 2503, + "CTGAAG": 2504, + "CTGATA": 2505, + "CTGATT": 2506, + "CTGATC": 2507, + "CTGATG": 2508, + "CTGACA": 2509, + "CTGACT": 2510, + "CTGACC": 2511, + "CTGACG": 2512, + "CTGAGA": 2513, + "CTGAGT": 2514, + "CTGAGC": 2515, + "CTGAGG": 2516, + "CTGTAA": 2517, + "CTGTAT": 2518, + "CTGTAC": 2519, + "CTGTAG": 2520, + "CTGTTA": 2521, + "CTGTTT": 2522, + "CTGTTC": 2523, + "CTGTTG": 2524, + "CTGTCA": 2525, + "CTGTCT": 2526, + "CTGTCC": 2527, + "CTGTCG": 2528, + "CTGTGA": 2529, + "CTGTGT": 2530, + "CTGTGC": 2531, + "CTGTGG": 2532, + "CTGCAA": 2533, + "CTGCAT": 2534, + "CTGCAC": 2535, + "CTGCAG": 2536, + "CTGCTA": 2537, + "CTGCTT": 2538, + "CTGCTC": 2539, + "CTGCTG": 2540, + "CTGCCA": 2541, + "CTGCCT": 2542, + "CTGCCC": 2543, + "CTGCCG": 2544, + "CTGCGA": 2545, + "CTGCGT": 2546, + "CTGCGC": 2547, + "CTGCGG": 2548, + "CTGGAA": 2549, + "CTGGAT": 2550, + "CTGGAC": 2551, + "CTGGAG": 2552, + "CTGGTA": 2553, + "CTGGTT": 2554, + "CTGGTC": 2555, + "CTGGTG": 2556, + "CTGGCA": 2557, + "CTGGCT": 2558, + "CTGGCC": 2559, + "CTGGCG": 2560, + "CTGGGA": 2561, + "CTGGGT": 2562, + "CTGGGC": 2563, + "CTGGGG": 2564, + "CCAAAA": 2565, + "CCAAAT": 2566, + "CCAAAC": 2567, + "CCAAAG": 2568, + "CCAATA": 2569, + "CCAATT": 2570, + "CCAATC": 2571, + "CCAATG": 2572, + "CCAACA": 2573, + "CCAACT": 2574, + "CCAACC": 2575, + "CCAACG": 2576, + "CCAAGA": 2577, + "CCAAGT": 2578, + "CCAAGC": 2579, + "CCAAGG": 2580, + "CCATAA": 2581, + "CCATAT": 2582, + "CCATAC": 2583, + "CCATAG": 2584, + "CCATTA": 2585, + "CCATTT": 2586, + "CCATTC": 2587, + "CCATTG": 2588, + "CCATCA": 2589, + "CCATCT": 2590, + "CCATCC": 2591, + "CCATCG": 2592, + "CCATGA": 2593, + "CCATGT": 2594, + "CCATGC": 2595, + "CCATGG": 2596, + "CCACAA": 2597, + "CCACAT": 2598, + "CCACAC": 2599, + "CCACAG": 2600, + "CCACTA": 2601, + "CCACTT": 2602, + "CCACTC": 2603, + "CCACTG": 2604, + "CCACCA": 2605, + "CCACCT": 2606, + "CCACCC": 2607, + "CCACCG": 2608, + "CCACGA": 2609, + "CCACGT": 2610, + "CCACGC": 2611, + "CCACGG": 2612, + "CCAGAA": 2613, + "CCAGAT": 2614, + "CCAGAC": 2615, + "CCAGAG": 2616, + "CCAGTA": 2617, + "CCAGTT": 2618, + "CCAGTC": 2619, + "CCAGTG": 2620, + "CCAGCA": 2621, + "CCAGCT": 2622, + "CCAGCC": 2623, + "CCAGCG": 2624, + "CCAGGA": 2625, + "CCAGGT": 2626, + "CCAGGC": 2627, + "CCAGGG": 2628, + "CCTAAA": 2629, + "CCTAAT": 2630, + "CCTAAC": 2631, + "CCTAAG": 2632, + "CCTATA": 2633, + "CCTATT": 2634, + "CCTATC": 2635, + "CCTATG": 2636, + "CCTACA": 2637, + "CCTACT": 2638, + "CCTACC": 2639, + "CCTACG": 2640, + "CCTAGA": 2641, + "CCTAGT": 2642, + "CCTAGC": 2643, + "CCTAGG": 2644, + "CCTTAA": 2645, + "CCTTAT": 2646, + "CCTTAC": 2647, + "CCTTAG": 2648, + "CCTTTA": 2649, + "CCTTTT": 2650, + "CCTTTC": 2651, + "CCTTTG": 2652, + "CCTTCA": 2653, + "CCTTCT": 2654, + "CCTTCC": 2655, + "CCTTCG": 2656, + "CCTTGA": 2657, + "CCTTGT": 2658, + "CCTTGC": 2659, + "CCTTGG": 2660, + "CCTCAA": 2661, + "CCTCAT": 2662, + "CCTCAC": 2663, + "CCTCAG": 2664, + "CCTCTA": 2665, + "CCTCTT": 2666, + "CCTCTC": 2667, + "CCTCTG": 2668, + "CCTCCA": 2669, + "CCTCCT": 2670, + "CCTCCC": 2671, + "CCTCCG": 2672, + "CCTCGA": 2673, + "CCTCGT": 2674, + "CCTCGC": 2675, + "CCTCGG": 2676, + "CCTGAA": 2677, + "CCTGAT": 2678, + "CCTGAC": 2679, + "CCTGAG": 2680, + "CCTGTA": 2681, + "CCTGTT": 2682, + "CCTGTC": 2683, + "CCTGTG": 2684, + "CCTGCA": 2685, + "CCTGCT": 2686, + "CCTGCC": 2687, + "CCTGCG": 2688, + "CCTGGA": 2689, + "CCTGGT": 2690, + "CCTGGC": 2691, + "CCTGGG": 2692, + "CCCAAA": 2693, + "CCCAAT": 2694, + "CCCAAC": 2695, + "CCCAAG": 2696, + "CCCATA": 2697, + "CCCATT": 2698, + "CCCATC": 2699, + "CCCATG": 2700, + "CCCACA": 2701, + "CCCACT": 2702, + "CCCACC": 2703, + "CCCACG": 2704, + "CCCAGA": 2705, + "CCCAGT": 2706, + "CCCAGC": 2707, + "CCCAGG": 2708, + "CCCTAA": 2709, + "CCCTAT": 2710, + "CCCTAC": 2711, + "CCCTAG": 2712, + "CCCTTA": 2713, + "CCCTTT": 2714, + "CCCTTC": 2715, + "CCCTTG": 2716, + "CCCTCA": 2717, + "CCCTCT": 2718, + "CCCTCC": 2719, + "CCCTCG": 2720, + "CCCTGA": 2721, + "CCCTGT": 2722, + "CCCTGC": 2723, + "CCCTGG": 2724, + "CCCCAA": 2725, + "CCCCAT": 2726, + "CCCCAC": 2727, + "CCCCAG": 2728, + "CCCCTA": 2729, + "CCCCTT": 2730, + "CCCCTC": 2731, + "CCCCTG": 2732, + "CCCCCA": 2733, + "CCCCCT": 2734, + "CCCCCC": 2735, + "CCCCCG": 2736, + "CCCCGA": 2737, + "CCCCGT": 2738, + "CCCCGC": 2739, + "CCCCGG": 2740, + "CCCGAA": 2741, + "CCCGAT": 2742, + "CCCGAC": 2743, + "CCCGAG": 2744, + "CCCGTA": 2745, + "CCCGTT": 2746, + "CCCGTC": 2747, + "CCCGTG": 2748, + "CCCGCA": 2749, + "CCCGCT": 2750, + "CCCGCC": 2751, + "CCCGCG": 2752, + "CCCGGA": 2753, + "CCCGGT": 2754, + "CCCGGC": 2755, + "CCCGGG": 2756, + "CCGAAA": 2757, + "CCGAAT": 2758, + "CCGAAC": 2759, + "CCGAAG": 2760, + "CCGATA": 2761, + "CCGATT": 2762, + "CCGATC": 2763, + "CCGATG": 2764, + "CCGACA": 2765, + "CCGACT": 2766, + "CCGACC": 2767, + "CCGACG": 2768, + "CCGAGA": 2769, + "CCGAGT": 2770, + "CCGAGC": 2771, + "CCGAGG": 2772, + "CCGTAA": 2773, + "CCGTAT": 2774, + "CCGTAC": 2775, + "CCGTAG": 2776, + "CCGTTA": 2777, + "CCGTTT": 2778, + "CCGTTC": 2779, + "CCGTTG": 2780, + "CCGTCA": 2781, + "CCGTCT": 2782, + "CCGTCC": 2783, + "CCGTCG": 2784, + "CCGTGA": 2785, + "CCGTGT": 2786, + "CCGTGC": 2787, + "CCGTGG": 2788, + "CCGCAA": 2789, + "CCGCAT": 2790, + "CCGCAC": 2791, + "CCGCAG": 2792, + "CCGCTA": 2793, + "CCGCTT": 2794, + "CCGCTC": 2795, + "CCGCTG": 2796, + "CCGCCA": 2797, + "CCGCCT": 2798, + "CCGCCC": 2799, + "CCGCCG": 2800, + "CCGCGA": 2801, + "CCGCGT": 2802, + "CCGCGC": 2803, + "CCGCGG": 2804, + "CCGGAA": 2805, + "CCGGAT": 2806, + "CCGGAC": 2807, + "CCGGAG": 2808, + "CCGGTA": 2809, + "CCGGTT": 2810, + "CCGGTC": 2811, + "CCGGTG": 2812, + "CCGGCA": 2813, + "CCGGCT": 2814, + "CCGGCC": 2815, + "CCGGCG": 2816, + "CCGGGA": 2817, + "CCGGGT": 2818, + "CCGGGC": 2819, + "CCGGGG": 2820, + "CGAAAA": 2821, + "CGAAAT": 2822, + "CGAAAC": 2823, + "CGAAAG": 2824, + "CGAATA": 2825, + "CGAATT": 2826, + "CGAATC": 2827, + "CGAATG": 2828, + "CGAACA": 2829, + "CGAACT": 2830, + "CGAACC": 2831, + "CGAACG": 2832, + "CGAAGA": 2833, + "CGAAGT": 2834, + "CGAAGC": 2835, + "CGAAGG": 2836, + "CGATAA": 2837, + "CGATAT": 2838, + "CGATAC": 2839, + "CGATAG": 2840, + "CGATTA": 2841, + "CGATTT": 2842, + "CGATTC": 2843, + "CGATTG": 2844, + "CGATCA": 2845, + "CGATCT": 2846, + "CGATCC": 2847, + "CGATCG": 2848, + "CGATGA": 2849, + "CGATGT": 2850, + "CGATGC": 2851, + "CGATGG": 2852, + "CGACAA": 2853, + "CGACAT": 2854, + "CGACAC": 2855, + "CGACAG": 2856, + "CGACTA": 2857, + "CGACTT": 2858, + "CGACTC": 2859, + "CGACTG": 2860, + "CGACCA": 2861, + "CGACCT": 2862, + "CGACCC": 2863, + "CGACCG": 2864, + "CGACGA": 2865, + "CGACGT": 2866, + "CGACGC": 2867, + "CGACGG": 2868, + "CGAGAA": 2869, + "CGAGAT": 2870, + "CGAGAC": 2871, + "CGAGAG": 2872, + "CGAGTA": 2873, + "CGAGTT": 2874, + "CGAGTC": 2875, + "CGAGTG": 2876, + "CGAGCA": 2877, + "CGAGCT": 2878, + "CGAGCC": 2879, + "CGAGCG": 2880, + "CGAGGA": 2881, + "CGAGGT": 2882, + "CGAGGC": 2883, + "CGAGGG": 2884, + "CGTAAA": 2885, + "CGTAAT": 2886, + "CGTAAC": 2887, + "CGTAAG": 2888, + "CGTATA": 2889, + "CGTATT": 2890, + "CGTATC": 2891, + "CGTATG": 2892, + "CGTACA": 2893, + "CGTACT": 2894, + "CGTACC": 2895, + "CGTACG": 2896, + "CGTAGA": 2897, + "CGTAGT": 2898, + "CGTAGC": 2899, + "CGTAGG": 2900, + "CGTTAA": 2901, + "CGTTAT": 2902, + "CGTTAC": 2903, + "CGTTAG": 2904, + "CGTTTA": 2905, + "CGTTTT": 2906, + "CGTTTC": 2907, + "CGTTTG": 2908, + "CGTTCA": 2909, + "CGTTCT": 2910, + "CGTTCC": 2911, + "CGTTCG": 2912, + "CGTTGA": 2913, + "CGTTGT": 2914, + "CGTTGC": 2915, + "CGTTGG": 2916, + "CGTCAA": 2917, + "CGTCAT": 2918, + "CGTCAC": 2919, + "CGTCAG": 2920, + "CGTCTA": 2921, + "CGTCTT": 2922, + "CGTCTC": 2923, + "CGTCTG": 2924, + "CGTCCA": 2925, + "CGTCCT": 2926, + "CGTCCC": 2927, + "CGTCCG": 2928, + "CGTCGA": 2929, + "CGTCGT": 2930, + "CGTCGC": 2931, + "CGTCGG": 2932, + "CGTGAA": 2933, + "CGTGAT": 2934, + "CGTGAC": 2935, + "CGTGAG": 2936, + "CGTGTA": 2937, + "CGTGTT": 2938, + "CGTGTC": 2939, + "CGTGTG": 2940, + "CGTGCA": 2941, + "CGTGCT": 2942, + "CGTGCC": 2943, + "CGTGCG": 2944, + "CGTGGA": 2945, + "CGTGGT": 2946, + "CGTGGC": 2947, + "CGTGGG": 2948, + "CGCAAA": 2949, + "CGCAAT": 2950, + "CGCAAC": 2951, + "CGCAAG": 2952, + "CGCATA": 2953, + "CGCATT": 2954, + "CGCATC": 2955, + "CGCATG": 2956, + "CGCACA": 2957, + "CGCACT": 2958, + "CGCACC": 2959, + "CGCACG": 2960, + "CGCAGA": 2961, + "CGCAGT": 2962, + "CGCAGC": 2963, + "CGCAGG": 2964, + "CGCTAA": 2965, + "CGCTAT": 2966, + "CGCTAC": 2967, + "CGCTAG": 2968, + "CGCTTA": 2969, + "CGCTTT": 2970, + "CGCTTC": 2971, + "CGCTTG": 2972, + "CGCTCA": 2973, + "CGCTCT": 2974, + "CGCTCC": 2975, + "CGCTCG": 2976, + "CGCTGA": 2977, + "CGCTGT": 2978, + "CGCTGC": 2979, + "CGCTGG": 2980, + "CGCCAA": 2981, + "CGCCAT": 2982, + "CGCCAC": 2983, + "CGCCAG": 2984, + "CGCCTA": 2985, + "CGCCTT": 2986, + "CGCCTC": 2987, + "CGCCTG": 2988, + "CGCCCA": 2989, + "CGCCCT": 2990, + "CGCCCC": 2991, + "CGCCCG": 2992, + "CGCCGA": 2993, + "CGCCGT": 2994, + "CGCCGC": 2995, + "CGCCGG": 2996, + "CGCGAA": 2997, + "CGCGAT": 2998, + "CGCGAC": 2999, + "CGCGAG": 3000, + "CGCGTA": 3001, + "CGCGTT": 3002, + "CGCGTC": 3003, + "CGCGTG": 3004, + "CGCGCA": 3005, + "CGCGCT": 3006, + "CGCGCC": 3007, + "CGCGCG": 3008, + "CGCGGA": 3009, + "CGCGGT": 3010, + "CGCGGC": 3011, + "CGCGGG": 3012, + "CGGAAA": 3013, + "CGGAAT": 3014, + "CGGAAC": 3015, + "CGGAAG": 3016, + "CGGATA": 3017, + "CGGATT": 3018, + "CGGATC": 3019, + "CGGATG": 3020, + "CGGACA": 3021, + "CGGACT": 3022, + "CGGACC": 3023, + "CGGACG": 3024, + "CGGAGA": 3025, + "CGGAGT": 3026, + "CGGAGC": 3027, + "CGGAGG": 3028, + "CGGTAA": 3029, + "CGGTAT": 3030, + "CGGTAC": 3031, + "CGGTAG": 3032, + "CGGTTA": 3033, + "CGGTTT": 3034, + "CGGTTC": 3035, + "CGGTTG": 3036, + "CGGTCA": 3037, + "CGGTCT": 3038, + "CGGTCC": 3039, + "CGGTCG": 3040, + "CGGTGA": 3041, + "CGGTGT": 3042, + "CGGTGC": 3043, + "CGGTGG": 3044, + "CGGCAA": 3045, + "CGGCAT": 3046, + "CGGCAC": 3047, + "CGGCAG": 3048, + "CGGCTA": 3049, + "CGGCTT": 3050, + "CGGCTC": 3051, + "CGGCTG": 3052, + "CGGCCA": 3053, + "CGGCCT": 3054, + "CGGCCC": 3055, + "CGGCCG": 3056, + "CGGCGA": 3057, + "CGGCGT": 3058, + "CGGCGC": 3059, + "CGGCGG": 3060, + "CGGGAA": 3061, + "CGGGAT": 3062, + "CGGGAC": 3063, + "CGGGAG": 3064, + "CGGGTA": 3065, + "CGGGTT": 3066, + "CGGGTC": 3067, + "CGGGTG": 3068, + "CGGGCA": 3069, + "CGGGCT": 3070, + "CGGGCC": 3071, + "CGGGCG": 3072, + "CGGGGA": 3073, + "CGGGGT": 3074, + "CGGGGC": 3075, + "CGGGGG": 3076, + "GAAAAA": 3077, + "GAAAAT": 3078, + "GAAAAC": 3079, + "GAAAAG": 3080, + "GAAATA": 3081, + "GAAATT": 3082, + "GAAATC": 3083, + "GAAATG": 3084, + "GAAACA": 3085, + "GAAACT": 3086, + "GAAACC": 3087, + "GAAACG": 3088, + "GAAAGA": 3089, + "GAAAGT": 3090, + "GAAAGC": 3091, + "GAAAGG": 3092, + "GAATAA": 3093, + "GAATAT": 3094, + "GAATAC": 3095, + "GAATAG": 3096, + "GAATTA": 3097, + "GAATTT": 3098, + "GAATTC": 3099, + "GAATTG": 3100, + "GAATCA": 3101, + "GAATCT": 3102, + "GAATCC": 3103, + "GAATCG": 3104, + "GAATGA": 3105, + "GAATGT": 3106, + "GAATGC": 3107, + "GAATGG": 3108, + "GAACAA": 3109, + "GAACAT": 3110, + "GAACAC": 3111, + "GAACAG": 3112, + "GAACTA": 3113, + "GAACTT": 3114, + "GAACTC": 3115, + "GAACTG": 3116, + "GAACCA": 3117, + "GAACCT": 3118, + "GAACCC": 3119, + "GAACCG": 3120, + "GAACGA": 3121, + "GAACGT": 3122, + "GAACGC": 3123, + "GAACGG": 3124, + "GAAGAA": 3125, + "GAAGAT": 3126, + "GAAGAC": 3127, + "GAAGAG": 3128, + "GAAGTA": 3129, + "GAAGTT": 3130, + "GAAGTC": 3131, + "GAAGTG": 3132, + "GAAGCA": 3133, + "GAAGCT": 3134, + "GAAGCC": 3135, + "GAAGCG": 3136, + "GAAGGA": 3137, + "GAAGGT": 3138, + "GAAGGC": 3139, + "GAAGGG": 3140, + "GATAAA": 3141, + "GATAAT": 3142, + "GATAAC": 3143, + "GATAAG": 3144, + "GATATA": 3145, + "GATATT": 3146, + "GATATC": 3147, + "GATATG": 3148, + "GATACA": 3149, + "GATACT": 3150, + "GATACC": 3151, + "GATACG": 3152, + "GATAGA": 3153, + "GATAGT": 3154, + "GATAGC": 3155, + "GATAGG": 3156, + "GATTAA": 3157, + "GATTAT": 3158, + "GATTAC": 3159, + "GATTAG": 3160, + "GATTTA": 3161, + "GATTTT": 3162, + "GATTTC": 3163, + "GATTTG": 3164, + "GATTCA": 3165, + "GATTCT": 3166, + "GATTCC": 3167, + "GATTCG": 3168, + "GATTGA": 3169, + "GATTGT": 3170, + "GATTGC": 3171, + "GATTGG": 3172, + "GATCAA": 3173, + "GATCAT": 3174, + "GATCAC": 3175, + "GATCAG": 3176, + "GATCTA": 3177, + "GATCTT": 3178, + "GATCTC": 3179, + "GATCTG": 3180, + "GATCCA": 3181, + "GATCCT": 3182, + "GATCCC": 3183, + "GATCCG": 3184, + "GATCGA": 3185, + "GATCGT": 3186, + "GATCGC": 3187, + "GATCGG": 3188, + "GATGAA": 3189, + "GATGAT": 3190, + "GATGAC": 3191, + "GATGAG": 3192, + "GATGTA": 3193, + "GATGTT": 3194, + "GATGTC": 3195, + "GATGTG": 3196, + "GATGCA": 3197, + "GATGCT": 3198, + "GATGCC": 3199, + "GATGCG": 3200, + "GATGGA": 3201, + "GATGGT": 3202, + "GATGGC": 3203, + "GATGGG": 3204, + "GACAAA": 3205, + "GACAAT": 3206, + "GACAAC": 3207, + "GACAAG": 3208, + "GACATA": 3209, + "GACATT": 3210, + "GACATC": 3211, + "GACATG": 3212, + "GACACA": 3213, + "GACACT": 3214, + "GACACC": 3215, + "GACACG": 3216, + "GACAGA": 3217, + "GACAGT": 3218, + "GACAGC": 3219, + "GACAGG": 3220, + "GACTAA": 3221, + "GACTAT": 3222, + "GACTAC": 3223, + "GACTAG": 3224, + "GACTTA": 3225, + "GACTTT": 3226, + "GACTTC": 3227, + "GACTTG": 3228, + "GACTCA": 3229, + "GACTCT": 3230, + "GACTCC": 3231, + "GACTCG": 3232, + "GACTGA": 3233, + "GACTGT": 3234, + "GACTGC": 3235, + "GACTGG": 3236, + "GACCAA": 3237, + "GACCAT": 3238, + "GACCAC": 3239, + "GACCAG": 3240, + "GACCTA": 3241, + "GACCTT": 3242, + "GACCTC": 3243, + "GACCTG": 3244, + "GACCCA": 3245, + "GACCCT": 3246, + "GACCCC": 3247, + "GACCCG": 3248, + "GACCGA": 3249, + "GACCGT": 3250, + "GACCGC": 3251, + "GACCGG": 3252, + "GACGAA": 3253, + "GACGAT": 3254, + "GACGAC": 3255, + "GACGAG": 3256, + "GACGTA": 3257, + "GACGTT": 3258, + "GACGTC": 3259, + "GACGTG": 3260, + "GACGCA": 3261, + "GACGCT": 3262, + "GACGCC": 3263, + "GACGCG": 3264, + "GACGGA": 3265, + "GACGGT": 3266, + "GACGGC": 3267, + "GACGGG": 3268, + "GAGAAA": 3269, + "GAGAAT": 3270, + "GAGAAC": 3271, + "GAGAAG": 3272, + "GAGATA": 3273, + "GAGATT": 3274, + "GAGATC": 3275, + "GAGATG": 3276, + "GAGACA": 3277, + "GAGACT": 3278, + "GAGACC": 3279, + "GAGACG": 3280, + "GAGAGA": 3281, + "GAGAGT": 3282, + "GAGAGC": 3283, + "GAGAGG": 3284, + "GAGTAA": 3285, + "GAGTAT": 3286, + "GAGTAC": 3287, + "GAGTAG": 3288, + "GAGTTA": 3289, + "GAGTTT": 3290, + "GAGTTC": 3291, + "GAGTTG": 3292, + "GAGTCA": 3293, + "GAGTCT": 3294, + "GAGTCC": 3295, + "GAGTCG": 3296, + "GAGTGA": 3297, + "GAGTGT": 3298, + "GAGTGC": 3299, + "GAGTGG": 3300, + "GAGCAA": 3301, + "GAGCAT": 3302, + "GAGCAC": 3303, + "GAGCAG": 3304, + "GAGCTA": 3305, + "GAGCTT": 3306, + "GAGCTC": 3307, + "GAGCTG": 3308, + "GAGCCA": 3309, + "GAGCCT": 3310, + "GAGCCC": 3311, + "GAGCCG": 3312, + "GAGCGA": 3313, + "GAGCGT": 3314, + "GAGCGC": 3315, + "GAGCGG": 3316, + "GAGGAA": 3317, + "GAGGAT": 3318, + "GAGGAC": 3319, + "GAGGAG": 3320, + "GAGGTA": 3321, + "GAGGTT": 3322, + "GAGGTC": 3323, + "GAGGTG": 3324, + "GAGGCA": 3325, + "GAGGCT": 3326, + "GAGGCC": 3327, + "GAGGCG": 3328, + "GAGGGA": 3329, + "GAGGGT": 3330, + "GAGGGC": 3331, + "GAGGGG": 3332, + "GTAAAA": 3333, + "GTAAAT": 3334, + "GTAAAC": 3335, + "GTAAAG": 3336, + "GTAATA": 3337, + "GTAATT": 3338, + "GTAATC": 3339, + "GTAATG": 3340, + "GTAACA": 3341, + "GTAACT": 3342, + "GTAACC": 3343, + "GTAACG": 3344, + "GTAAGA": 3345, + "GTAAGT": 3346, + "GTAAGC": 3347, + "GTAAGG": 3348, + "GTATAA": 3349, + "GTATAT": 3350, + "GTATAC": 3351, + "GTATAG": 3352, + "GTATTA": 3353, + "GTATTT": 3354, + "GTATTC": 3355, + "GTATTG": 3356, + "GTATCA": 3357, + "GTATCT": 3358, + "GTATCC": 3359, + "GTATCG": 3360, + "GTATGA": 3361, + "GTATGT": 3362, + "GTATGC": 3363, + "GTATGG": 3364, + "GTACAA": 3365, + "GTACAT": 3366, + "GTACAC": 3367, + "GTACAG": 3368, + "GTACTA": 3369, + "GTACTT": 3370, + "GTACTC": 3371, + "GTACTG": 3372, + "GTACCA": 3373, + "GTACCT": 3374, + "GTACCC": 3375, + "GTACCG": 3376, + "GTACGA": 3377, + "GTACGT": 3378, + "GTACGC": 3379, + "GTACGG": 3380, + "GTAGAA": 3381, + "GTAGAT": 3382, + "GTAGAC": 3383, + "GTAGAG": 3384, + "GTAGTA": 3385, + "GTAGTT": 3386, + "GTAGTC": 3387, + "GTAGTG": 3388, + "GTAGCA": 3389, + "GTAGCT": 3390, + "GTAGCC": 3391, + "GTAGCG": 3392, + "GTAGGA": 3393, + "GTAGGT": 3394, + "GTAGGC": 3395, + "GTAGGG": 3396, + "GTTAAA": 3397, + "GTTAAT": 3398, + "GTTAAC": 3399, + "GTTAAG": 3400, + "GTTATA": 3401, + "GTTATT": 3402, + "GTTATC": 3403, + "GTTATG": 3404, + "GTTACA": 3405, + "GTTACT": 3406, + "GTTACC": 3407, + "GTTACG": 3408, + "GTTAGA": 3409, + "GTTAGT": 3410, + "GTTAGC": 3411, + "GTTAGG": 3412, + "GTTTAA": 3413, + "GTTTAT": 3414, + "GTTTAC": 3415, + "GTTTAG": 3416, + "GTTTTA": 3417, + "GTTTTT": 3418, + "GTTTTC": 3419, + "GTTTTG": 3420, + "GTTTCA": 3421, + "GTTTCT": 3422, + "GTTTCC": 3423, + "GTTTCG": 3424, + "GTTTGA": 3425, + "GTTTGT": 3426, + "GTTTGC": 3427, + "GTTTGG": 3428, + "GTTCAA": 3429, + "GTTCAT": 3430, + "GTTCAC": 3431, + "GTTCAG": 3432, + "GTTCTA": 3433, + "GTTCTT": 3434, + "GTTCTC": 3435, + "GTTCTG": 3436, + "GTTCCA": 3437, + "GTTCCT": 3438, + "GTTCCC": 3439, + "GTTCCG": 3440, + "GTTCGA": 3441, + "GTTCGT": 3442, + "GTTCGC": 3443, + "GTTCGG": 3444, + "GTTGAA": 3445, + "GTTGAT": 3446, + "GTTGAC": 3447, + "GTTGAG": 3448, + "GTTGTA": 3449, + "GTTGTT": 3450, + "GTTGTC": 3451, + "GTTGTG": 3452, + "GTTGCA": 3453, + "GTTGCT": 3454, + "GTTGCC": 3455, + "GTTGCG": 3456, + "GTTGGA": 3457, + "GTTGGT": 3458, + "GTTGGC": 3459, + "GTTGGG": 3460, + "GTCAAA": 3461, + "GTCAAT": 3462, + "GTCAAC": 3463, + "GTCAAG": 3464, + "GTCATA": 3465, + "GTCATT": 3466, + "GTCATC": 3467, + "GTCATG": 3468, + "GTCACA": 3469, + "GTCACT": 3470, + "GTCACC": 3471, + "GTCACG": 3472, + "GTCAGA": 3473, + "GTCAGT": 3474, + "GTCAGC": 3475, + "GTCAGG": 3476, + "GTCTAA": 3477, + "GTCTAT": 3478, + "GTCTAC": 3479, + "GTCTAG": 3480, + "GTCTTA": 3481, + "GTCTTT": 3482, + "GTCTTC": 3483, + "GTCTTG": 3484, + "GTCTCA": 3485, + "GTCTCT": 3486, + "GTCTCC": 3487, + "GTCTCG": 3488, + "GTCTGA": 3489, + "GTCTGT": 3490, + "GTCTGC": 3491, + "GTCTGG": 3492, + "GTCCAA": 3493, + "GTCCAT": 3494, + "GTCCAC": 3495, + "GTCCAG": 3496, + "GTCCTA": 3497, + "GTCCTT": 3498, + "GTCCTC": 3499, + "GTCCTG": 3500, + "GTCCCA": 3501, + "GTCCCT": 3502, + "GTCCCC": 3503, + "GTCCCG": 3504, + "GTCCGA": 3505, + "GTCCGT": 3506, + "GTCCGC": 3507, + "GTCCGG": 3508, + "GTCGAA": 3509, + "GTCGAT": 3510, + "GTCGAC": 3511, + "GTCGAG": 3512, + "GTCGTA": 3513, + "GTCGTT": 3514, + "GTCGTC": 3515, + "GTCGTG": 3516, + "GTCGCA": 3517, + "GTCGCT": 3518, + "GTCGCC": 3519, + "GTCGCG": 3520, + "GTCGGA": 3521, + "GTCGGT": 3522, + "GTCGGC": 3523, + "GTCGGG": 3524, + "GTGAAA": 3525, + "GTGAAT": 3526, + "GTGAAC": 3527, + "GTGAAG": 3528, + "GTGATA": 3529, + "GTGATT": 3530, + "GTGATC": 3531, + "GTGATG": 3532, + "GTGACA": 3533, + "GTGACT": 3534, + "GTGACC": 3535, + "GTGACG": 3536, + "GTGAGA": 3537, + "GTGAGT": 3538, + "GTGAGC": 3539, + "GTGAGG": 3540, + "GTGTAA": 3541, + "GTGTAT": 3542, + "GTGTAC": 3543, + "GTGTAG": 3544, + "GTGTTA": 3545, + "GTGTTT": 3546, + "GTGTTC": 3547, + "GTGTTG": 3548, + "GTGTCA": 3549, + "GTGTCT": 3550, + "GTGTCC": 3551, + "GTGTCG": 3552, + "GTGTGA": 3553, + "GTGTGT": 3554, + "GTGTGC": 3555, + "GTGTGG": 3556, + "GTGCAA": 3557, + "GTGCAT": 3558, + "GTGCAC": 3559, + "GTGCAG": 3560, + "GTGCTA": 3561, + "GTGCTT": 3562, + "GTGCTC": 3563, + "GTGCTG": 3564, + "GTGCCA": 3565, + "GTGCCT": 3566, + "GTGCCC": 3567, + "GTGCCG": 3568, + "GTGCGA": 3569, + "GTGCGT": 3570, + "GTGCGC": 3571, + "GTGCGG": 3572, + "GTGGAA": 3573, + "GTGGAT": 3574, + "GTGGAC": 3575, + "GTGGAG": 3576, + "GTGGTA": 3577, + "GTGGTT": 3578, + "GTGGTC": 3579, + "GTGGTG": 3580, + "GTGGCA": 3581, + "GTGGCT": 3582, + "GTGGCC": 3583, + "GTGGCG": 3584, + "GTGGGA": 3585, + "GTGGGT": 3586, + "GTGGGC": 3587, + "GTGGGG": 3588, + "GCAAAA": 3589, + "GCAAAT": 3590, + "GCAAAC": 3591, + "GCAAAG": 3592, + "GCAATA": 3593, + "GCAATT": 3594, + "GCAATC": 3595, + "GCAATG": 3596, + "GCAACA": 3597, + "GCAACT": 3598, + "GCAACC": 3599, + "GCAACG": 3600, + "GCAAGA": 3601, + "GCAAGT": 3602, + "GCAAGC": 3603, + "GCAAGG": 3604, + "GCATAA": 3605, + "GCATAT": 3606, + "GCATAC": 3607, + "GCATAG": 3608, + "GCATTA": 3609, + "GCATTT": 3610, + "GCATTC": 3611, + "GCATTG": 3612, + "GCATCA": 3613, + "GCATCT": 3614, + "GCATCC": 3615, + "GCATCG": 3616, + "GCATGA": 3617, + "GCATGT": 3618, + "GCATGC": 3619, + "GCATGG": 3620, + "GCACAA": 3621, + "GCACAT": 3622, + "GCACAC": 3623, + "GCACAG": 3624, + "GCACTA": 3625, + "GCACTT": 3626, + "GCACTC": 3627, + "GCACTG": 3628, + "GCACCA": 3629, + "GCACCT": 3630, + "GCACCC": 3631, + "GCACCG": 3632, + "GCACGA": 3633, + "GCACGT": 3634, + "GCACGC": 3635, + "GCACGG": 3636, + "GCAGAA": 3637, + "GCAGAT": 3638, + "GCAGAC": 3639, + "GCAGAG": 3640, + "GCAGTA": 3641, + "GCAGTT": 3642, + "GCAGTC": 3643, + "GCAGTG": 3644, + "GCAGCA": 3645, + "GCAGCT": 3646, + "GCAGCC": 3647, + "GCAGCG": 3648, + "GCAGGA": 3649, + "GCAGGT": 3650, + "GCAGGC": 3651, + "GCAGGG": 3652, + "GCTAAA": 3653, + "GCTAAT": 3654, + "GCTAAC": 3655, + "GCTAAG": 3656, + "GCTATA": 3657, + "GCTATT": 3658, + "GCTATC": 3659, + "GCTATG": 3660, + "GCTACA": 3661, + "GCTACT": 3662, + "GCTACC": 3663, + "GCTACG": 3664, + "GCTAGA": 3665, + "GCTAGT": 3666, + "GCTAGC": 3667, + "GCTAGG": 3668, + "GCTTAA": 3669, + "GCTTAT": 3670, + "GCTTAC": 3671, + "GCTTAG": 3672, + "GCTTTA": 3673, + "GCTTTT": 3674, + "GCTTTC": 3675, + "GCTTTG": 3676, + "GCTTCA": 3677, + "GCTTCT": 3678, + "GCTTCC": 3679, + "GCTTCG": 3680, + "GCTTGA": 3681, + "GCTTGT": 3682, + "GCTTGC": 3683, + "GCTTGG": 3684, + "GCTCAA": 3685, + "GCTCAT": 3686, + "GCTCAC": 3687, + "GCTCAG": 3688, + "GCTCTA": 3689, + "GCTCTT": 3690, + "GCTCTC": 3691, + "GCTCTG": 3692, + "GCTCCA": 3693, + "GCTCCT": 3694, + "GCTCCC": 3695, + "GCTCCG": 3696, + "GCTCGA": 3697, + "GCTCGT": 3698, + "GCTCGC": 3699, + "GCTCGG": 3700, + "GCTGAA": 3701, + "GCTGAT": 3702, + "GCTGAC": 3703, + "GCTGAG": 3704, + "GCTGTA": 3705, + "GCTGTT": 3706, + "GCTGTC": 3707, + "GCTGTG": 3708, + "GCTGCA": 3709, + "GCTGCT": 3710, + "GCTGCC": 3711, + "GCTGCG": 3712, + "GCTGGA": 3713, + "GCTGGT": 3714, + "GCTGGC": 3715, + "GCTGGG": 3716, + "GCCAAA": 3717, + "GCCAAT": 3718, + "GCCAAC": 3719, + "GCCAAG": 3720, + "GCCATA": 3721, + "GCCATT": 3722, + "GCCATC": 3723, + "GCCATG": 3724, + "GCCACA": 3725, + "GCCACT": 3726, + "GCCACC": 3727, + "GCCACG": 3728, + "GCCAGA": 3729, + "GCCAGT": 3730, + "GCCAGC": 3731, + "GCCAGG": 3732, + "GCCTAA": 3733, + "GCCTAT": 3734, + "GCCTAC": 3735, + "GCCTAG": 3736, + "GCCTTA": 3737, + "GCCTTT": 3738, + "GCCTTC": 3739, + "GCCTTG": 3740, + "GCCTCA": 3741, + "GCCTCT": 3742, + "GCCTCC": 3743, + "GCCTCG": 3744, + "GCCTGA": 3745, + "GCCTGT": 3746, + "GCCTGC": 3747, + "GCCTGG": 3748, + "GCCCAA": 3749, + "GCCCAT": 3750, + "GCCCAC": 3751, + "GCCCAG": 3752, + "GCCCTA": 3753, + "GCCCTT": 3754, + "GCCCTC": 3755, + "GCCCTG": 3756, + "GCCCCA": 3757, + "GCCCCT": 3758, + "GCCCCC": 3759, + "GCCCCG": 3760, + "GCCCGA": 3761, + "GCCCGT": 3762, + "GCCCGC": 3763, + "GCCCGG": 3764, + "GCCGAA": 3765, + "GCCGAT": 3766, + "GCCGAC": 3767, + "GCCGAG": 3768, + "GCCGTA": 3769, + "GCCGTT": 3770, + "GCCGTC": 3771, + "GCCGTG": 3772, + "GCCGCA": 3773, + "GCCGCT": 3774, + "GCCGCC": 3775, + "GCCGCG": 3776, + "GCCGGA": 3777, + "GCCGGT": 3778, + "GCCGGC": 3779, + "GCCGGG": 3780, + "GCGAAA": 3781, + "GCGAAT": 3782, + "GCGAAC": 3783, + "GCGAAG": 3784, + "GCGATA": 3785, + "GCGATT": 3786, + "GCGATC": 3787, + "GCGATG": 3788, + "GCGACA": 3789, + "GCGACT": 3790, + "GCGACC": 3791, + "GCGACG": 3792, + "GCGAGA": 3793, + "GCGAGT": 3794, + "GCGAGC": 3795, + "GCGAGG": 3796, + "GCGTAA": 3797, + "GCGTAT": 3798, + "GCGTAC": 3799, + "GCGTAG": 3800, + "GCGTTA": 3801, + "GCGTTT": 3802, + "GCGTTC": 3803, + "GCGTTG": 3804, + "GCGTCA": 3805, + "GCGTCT": 3806, + "GCGTCC": 3807, + "GCGTCG": 3808, + "GCGTGA": 3809, + "GCGTGT": 3810, + "GCGTGC": 3811, + "GCGTGG": 3812, + "GCGCAA": 3813, + "GCGCAT": 3814, + "GCGCAC": 3815, + "GCGCAG": 3816, + "GCGCTA": 3817, + "GCGCTT": 3818, + "GCGCTC": 3819, + "GCGCTG": 3820, + "GCGCCA": 3821, + "GCGCCT": 3822, + "GCGCCC": 3823, + "GCGCCG": 3824, + "GCGCGA": 3825, + "GCGCGT": 3826, + "GCGCGC": 3827, + "GCGCGG": 3828, + "GCGGAA": 3829, + "GCGGAT": 3830, + "GCGGAC": 3831, + "GCGGAG": 3832, + "GCGGTA": 3833, + "GCGGTT": 3834, + "GCGGTC": 3835, + "GCGGTG": 3836, + "GCGGCA": 3837, + "GCGGCT": 3838, + "GCGGCC": 3839, + "GCGGCG": 3840, + "GCGGGA": 3841, + "GCGGGT": 3842, + "GCGGGC": 3843, + "GCGGGG": 3844, + "GGAAAA": 3845, + "GGAAAT": 3846, + "GGAAAC": 3847, + "GGAAAG": 3848, + "GGAATA": 3849, + "GGAATT": 3850, + "GGAATC": 3851, + "GGAATG": 3852, + "GGAACA": 3853, + "GGAACT": 3854, + "GGAACC": 3855, + "GGAACG": 3856, + "GGAAGA": 3857, + "GGAAGT": 3858, + "GGAAGC": 3859, + "GGAAGG": 3860, + "GGATAA": 3861, + "GGATAT": 3862, + "GGATAC": 3863, + "GGATAG": 3864, + "GGATTA": 3865, + "GGATTT": 3866, + "GGATTC": 3867, + "GGATTG": 3868, + "GGATCA": 3869, + "GGATCT": 3870, + "GGATCC": 3871, + "GGATCG": 3872, + "GGATGA": 3873, + "GGATGT": 3874, + "GGATGC": 3875, + "GGATGG": 3876, + "GGACAA": 3877, + "GGACAT": 3878, + "GGACAC": 3879, + "GGACAG": 3880, + "GGACTA": 3881, + "GGACTT": 3882, + "GGACTC": 3883, + "GGACTG": 3884, + "GGACCA": 3885, + "GGACCT": 3886, + "GGACCC": 3887, + "GGACCG": 3888, + "GGACGA": 3889, + "GGACGT": 3890, + "GGACGC": 3891, + "GGACGG": 3892, + "GGAGAA": 3893, + "GGAGAT": 3894, + "GGAGAC": 3895, + "GGAGAG": 3896, + "GGAGTA": 3897, + "GGAGTT": 3898, + "GGAGTC": 3899, + "GGAGTG": 3900, + "GGAGCA": 3901, + "GGAGCT": 3902, + "GGAGCC": 3903, + "GGAGCG": 3904, + "GGAGGA": 3905, + "GGAGGT": 3906, + "GGAGGC": 3907, + "GGAGGG": 3908, + "GGTAAA": 3909, + "GGTAAT": 3910, + "GGTAAC": 3911, + "GGTAAG": 3912, + "GGTATA": 3913, + "GGTATT": 3914, + "GGTATC": 3915, + "GGTATG": 3916, + "GGTACA": 3917, + "GGTACT": 3918, + "GGTACC": 3919, + "GGTACG": 3920, + "GGTAGA": 3921, + "GGTAGT": 3922, + "GGTAGC": 3923, + "GGTAGG": 3924, + "GGTTAA": 3925, + "GGTTAT": 3926, + "GGTTAC": 3927, + "GGTTAG": 3928, + "GGTTTA": 3929, + "GGTTTT": 3930, + "GGTTTC": 3931, + "GGTTTG": 3932, + "GGTTCA": 3933, + "GGTTCT": 3934, + "GGTTCC": 3935, + "GGTTCG": 3936, + "GGTTGA": 3937, + "GGTTGT": 3938, + "GGTTGC": 3939, + "GGTTGG": 3940, + "GGTCAA": 3941, + "GGTCAT": 3942, + "GGTCAC": 3943, + "GGTCAG": 3944, + "GGTCTA": 3945, + "GGTCTT": 3946, + "GGTCTC": 3947, + "GGTCTG": 3948, + "GGTCCA": 3949, + "GGTCCT": 3950, + "GGTCCC": 3951, + "GGTCCG": 3952, + "GGTCGA": 3953, + "GGTCGT": 3954, + "GGTCGC": 3955, + "GGTCGG": 3956, + "GGTGAA": 3957, + "GGTGAT": 3958, + "GGTGAC": 3959, + "GGTGAG": 3960, + "GGTGTA": 3961, + "GGTGTT": 3962, + "GGTGTC": 3963, + "GGTGTG": 3964, + "GGTGCA": 3965, + "GGTGCT": 3966, + "GGTGCC": 3967, + "GGTGCG": 3968, + "GGTGGA": 3969, + "GGTGGT": 3970, + "GGTGGC": 3971, + "GGTGGG": 3972, + "GGCAAA": 3973, + "GGCAAT": 3974, + "GGCAAC": 3975, + "GGCAAG": 3976, + "GGCATA": 3977, + "GGCATT": 3978, + "GGCATC": 3979, + "GGCATG": 3980, + "GGCACA": 3981, + "GGCACT": 3982, + "GGCACC": 3983, + "GGCACG": 3984, + "GGCAGA": 3985, + "GGCAGT": 3986, + "GGCAGC": 3987, + "GGCAGG": 3988, + "GGCTAA": 3989, + "GGCTAT": 3990, + "GGCTAC": 3991, + "GGCTAG": 3992, + "GGCTTA": 3993, + "GGCTTT": 3994, + "GGCTTC": 3995, + "GGCTTG": 3996, + "GGCTCA": 3997, + "GGCTCT": 3998, + "GGCTCC": 3999, + "GGCTCG": 4000, + "GGCTGA": 4001, + "GGCTGT": 4002, + "GGCTGC": 4003, + "GGCTGG": 4004, + "GGCCAA": 4005, + "GGCCAT": 4006, + "GGCCAC": 4007, + "GGCCAG": 4008, + "GGCCTA": 4009, + "GGCCTT": 4010, + "GGCCTC": 4011, + "GGCCTG": 4012, + "GGCCCA": 4013, + "GGCCCT": 4014, + "GGCCCC": 4015, + "GGCCCG": 4016, + "GGCCGA": 4017, + "GGCCGT": 4018, + "GGCCGC": 4019, + "GGCCGG": 4020, + "GGCGAA": 4021, + "GGCGAT": 4022, + "GGCGAC": 4023, + "GGCGAG": 4024, + "GGCGTA": 4025, + "GGCGTT": 4026, + "GGCGTC": 4027, + "GGCGTG": 4028, + "GGCGCA": 4029, + "GGCGCT": 4030, + "GGCGCC": 4031, + "GGCGCG": 4032, + "GGCGGA": 4033, + "GGCGGT": 4034, + "GGCGGC": 4035, + "GGCGGG": 4036, + "GGGAAA": 4037, + "GGGAAT": 4038, + "GGGAAC": 4039, + "GGGAAG": 4040, + "GGGATA": 4041, + "GGGATT": 4042, + "GGGATC": 4043, + "GGGATG": 4044, + "GGGACA": 4045, + "GGGACT": 4046, + "GGGACC": 4047, + "GGGACG": 4048, + "GGGAGA": 4049, + "GGGAGT": 4050, + "GGGAGC": 4051, + "GGGAGG": 4052, + "GGGTAA": 4053, + "GGGTAT": 4054, + "GGGTAC": 4055, + "GGGTAG": 4056, + "GGGTTA": 4057, + "GGGTTT": 4058, + "GGGTTC": 4059, + "GGGTTG": 4060, + "GGGTCA": 4061, + "GGGTCT": 4062, + "GGGTCC": 4063, + "GGGTCG": 4064, + "GGGTGA": 4065, + "GGGTGT": 4066, + "GGGTGC": 4067, + "GGGTGG": 4068, + "GGGCAA": 4069, + "GGGCAT": 4070, + "GGGCAC": 4071, + "GGGCAG": 4072, + "GGGCTA": 4073, + "GGGCTT": 4074, + "GGGCTC": 4075, + "GGGCTG": 4076, + "GGGCCA": 4077, + "GGGCCT": 4078, + "GGGCCC": 4079, + "GGGCCG": 4080, + "GGGCGA": 4081, + "GGGCGT": 4082, + "GGGCGC": 4083, + "GGGCGG": 4084, + "GGGGAA": 4085, + "GGGGAT": 4086, + "GGGGAC": 4087, + "GGGGAG": 4088, + "GGGGTA": 4089, + "GGGGTT": 4090, + "GGGGTC": 4091, + "GGGGTG": 4092, + "GGGGCA": 4093, + "GGGGCT": 4094, + "GGGGCC": 4095, + "GGGGCG": 4096, + "GGGGGA": 4097, + "GGGGGT": 4098, + "GGGGGC": 4099, + "GGGGGG": 4100 + } + } +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-5000/tokenizer_config.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-5000/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..604e2b1b29af1dcb3920d2828b6a90276e5e3d02 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-5000/tokenizer_config.json @@ -0,0 +1,58 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "[UNK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "[CLS]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "[SEP]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "4": { + "content": "[MASK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "clean_up_tokenization_spaces": true, + "cls_token": "[CLS]", + "do_basic_tokenize": true, + "do_lower_case": false, + "mask_token": "[MASK]", + "max_len": 512, + "model_max_length": 512, + "never_split": null, + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "strip_accents": null, + "tokenize_chinese_chars": true, + "tokenizer_class": "BertTokenizer", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-5000/trainer_state.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-5000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..5c6237365009550c7635b796ed96efe67d28af5e --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-5000/trainer_state.json @@ -0,0 +1,1433 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 20.0, + "eval_steps": 500, + "global_step": 5000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.1, + "grad_norm": 1.1046091318130493, + "learning_rate": 5.319148936170213e-05, + "loss": 0.6562, + "step": 25 + }, + { + "epoch": 0.2, + "grad_norm": 0.4066573977470398, + "learning_rate": 0.00010638297872340425, + "loss": 0.5373, + "step": 50 + }, + { + "epoch": 0.3, + "grad_norm": 0.7340325713157654, + "learning_rate": 0.00015957446808510637, + "loss": 0.4995, + "step": 75 + }, + { + "epoch": 0.4, + "grad_norm": 2.134040117263794, + "learning_rate": 0.0002127659574468085, + "loss": 0.4871, + "step": 100 + }, + { + "epoch": 0.5, + "grad_norm": 2.911975860595703, + "learning_rate": 0.00026595744680851064, + "loss": 0.4507, + "step": 125 + }, + { + "epoch": 0.6, + "grad_norm": 2.040365219116211, + "learning_rate": 0.00031914893617021275, + "loss": 0.3944, + "step": 150 + }, + { + "epoch": 0.7, + "grad_norm": 5.643661022186279, + "learning_rate": 0.0003723404255319149, + "loss": 0.3863, + "step": 175 + }, + { + "epoch": 0.8, + "grad_norm": 3.882492780685425, + "learning_rate": 0.0003992081821181128, + "loss": 0.3875, + "step": 200 + }, + { + "epoch": 0.9, + "grad_norm": 5.87437105178833, + "learning_rate": 0.0003975585615308479, + "loss": 0.3856, + "step": 225 + }, + { + "epoch": 1.0, + "grad_norm": 4.611206531524658, + "learning_rate": 0.00039590894094358297, + "loss": 0.3866, + "step": 250 + }, + { + "epoch": 1.1, + "grad_norm": 3.6404929161071777, + "learning_rate": 0.0003942593203563181, + "loss": 0.3437, + "step": 275 + }, + { + "epoch": 1.2, + "grad_norm": 6.469444751739502, + "learning_rate": 0.0003926096997690532, + "loss": 0.3192, + "step": 300 + }, + { + "epoch": 1.3, + "grad_norm": 3.6251447200775146, + "learning_rate": 0.00039096007918178817, + "loss": 0.349, + "step": 325 + }, + { + "epoch": 1.4, + "grad_norm": 2.854794979095459, + "learning_rate": 0.0003893104585945233, + "loss": 0.3274, + "step": 350 + }, + { + "epoch": 1.5, + "grad_norm": 4.805097579956055, + "learning_rate": 0.0003876608380072583, + "loss": 0.3279, + "step": 375 + }, + { + "epoch": 1.6, + "grad_norm": 2.9891581535339355, + "learning_rate": 0.00038601121741999343, + "loss": 0.3166, + "step": 400 + }, + { + "epoch": 1.7, + "grad_norm": 5.333752155303955, + "learning_rate": 0.0003843615968327285, + "loss": 0.2856, + "step": 425 + }, + { + "epoch": 1.8, + "grad_norm": 2.615621566772461, + "learning_rate": 0.0003827119762454636, + "loss": 0.3286, + "step": 450 + }, + { + "epoch": 1.9, + "grad_norm": 2.4228415489196777, + "learning_rate": 0.00038106235565819863, + "loss": 0.2812, + "step": 475 + }, + { + "epoch": 2.0, + "grad_norm": 3.6015517711639404, + "learning_rate": 0.0003794127350709337, + "loss": 0.2884, + "step": 500 + }, + { + "epoch": 2.1, + "grad_norm": 2.7867860794067383, + "learning_rate": 0.0003777631144836688, + "loss": 0.2582, + "step": 525 + }, + { + "epoch": 2.2, + "grad_norm": 10.079645156860352, + "learning_rate": 0.00037611349389640383, + "loss": 0.245, + "step": 550 + }, + { + "epoch": 2.3, + "grad_norm": 2.475917100906372, + "learning_rate": 0.00037446387330913894, + "loss": 0.2615, + "step": 575 + }, + { + "epoch": 2.4, + "grad_norm": 8.098592758178711, + "learning_rate": 0.000372814252721874, + "loss": 0.2484, + "step": 600 + }, + { + "epoch": 2.5, + "grad_norm": 2.318723201751709, + "learning_rate": 0.00037116463213460903, + "loss": 0.2056, + "step": 625 + }, + { + "epoch": 2.6, + "grad_norm": 3.616283893585205, + "learning_rate": 0.00036951501154734414, + "loss": 0.2358, + "step": 650 + }, + { + "epoch": 2.7, + "grad_norm": 6.419492721557617, + "learning_rate": 0.0003678653909600792, + "loss": 0.2204, + "step": 675 + }, + { + "epoch": 2.8, + "grad_norm": 2.9333388805389404, + "learning_rate": 0.0003662157703728143, + "loss": 0.1857, + "step": 700 + }, + { + "epoch": 2.9, + "grad_norm": 7.691112995147705, + "learning_rate": 0.00036456614978554934, + "loss": 0.2054, + "step": 725 + }, + { + "epoch": 3.0, + "grad_norm": 4.908492565155029, + "learning_rate": 0.00036291652919828444, + "loss": 0.1956, + "step": 750 + }, + { + "epoch": 3.1, + "grad_norm": 1.8923287391662598, + "learning_rate": 0.00036126690861101944, + "loss": 0.1708, + "step": 775 + }, + { + "epoch": 3.2, + "grad_norm": 4.403504848480225, + "learning_rate": 0.00035961728802375454, + "loss": 0.1441, + "step": 800 + }, + { + "epoch": 3.3, + "grad_norm": 4.117386817932129, + "learning_rate": 0.0003579676674364896, + "loss": 0.1586, + "step": 825 + }, + { + "epoch": 3.4, + "grad_norm": 7.119816303253174, + "learning_rate": 0.0003563180468492247, + "loss": 0.1472, + "step": 850 + }, + { + "epoch": 3.5, + "grad_norm": 3.5796430110931396, + "learning_rate": 0.0003546684262619598, + "loss": 0.1802, + "step": 875 + }, + { + "epoch": 3.6, + "grad_norm": 2.97688889503479, + "learning_rate": 0.00035301880567469485, + "loss": 0.1578, + "step": 900 + }, + { + "epoch": 3.7, + "grad_norm": 3.716148614883423, + "learning_rate": 0.0003513691850874299, + "loss": 0.1662, + "step": 925 + }, + { + "epoch": 3.8, + "grad_norm": 6.1249566078186035, + "learning_rate": 0.00034971956450016495, + "loss": 0.159, + "step": 950 + }, + { + "epoch": 3.9, + "grad_norm": 3.0592427253723145, + "learning_rate": 0.00034806994391290005, + "loss": 0.148, + "step": 975 + }, + { + "epoch": 4.0, + "grad_norm": 4.467265605926514, + "learning_rate": 0.0003464203233256351, + "loss": 0.1442, + "step": 1000 + }, + { + "epoch": 4.1, + "grad_norm": 2.7223546504974365, + "learning_rate": 0.0003447707027383702, + "loss": 0.1339, + "step": 1025 + }, + { + "epoch": 4.2, + "grad_norm": 3.698854923248291, + "learning_rate": 0.00034312108215110525, + "loss": 0.1381, + "step": 1050 + }, + { + "epoch": 4.3, + "grad_norm": 7.418384552001953, + "learning_rate": 0.0003414714615638403, + "loss": 0.1158, + "step": 1075 + }, + { + "epoch": 4.4, + "grad_norm": 1.2887814044952393, + "learning_rate": 0.0003398218409765754, + "loss": 0.0855, + "step": 1100 + }, + { + "epoch": 4.5, + "grad_norm": 0.838731586933136, + "learning_rate": 0.00033817222038931045, + "loss": 0.1272, + "step": 1125 + }, + { + "epoch": 4.6, + "grad_norm": 5.912592887878418, + "learning_rate": 0.00033652259980204556, + "loss": 0.1156, + "step": 1150 + }, + { + "epoch": 4.7, + "grad_norm": 8.49521255493164, + "learning_rate": 0.0003348729792147806, + "loss": 0.1182, + "step": 1175 + }, + { + "epoch": 4.8, + "grad_norm": 10.278315544128418, + "learning_rate": 0.0003332233586275157, + "loss": 0.1071, + "step": 1200 + }, + { + "epoch": 4.9, + "grad_norm": 0.724703311920166, + "learning_rate": 0.00033157373804025076, + "loss": 0.11, + "step": 1225 + }, + { + "epoch": 5.0, + "grad_norm": 1.6808199882507324, + "learning_rate": 0.0003299241174529858, + "loss": 0.1101, + "step": 1250 + }, + { + "epoch": 5.1, + "grad_norm": 0.1038585901260376, + "learning_rate": 0.0003282744968657209, + "loss": 0.0791, + "step": 1275 + }, + { + "epoch": 5.2, + "grad_norm": 0.15866787731647491, + "learning_rate": 0.00032662487627845596, + "loss": 0.0638, + "step": 1300 + }, + { + "epoch": 5.3, + "grad_norm": 6.748287200927734, + "learning_rate": 0.00032497525569119106, + "loss": 0.0961, + "step": 1325 + }, + { + "epoch": 5.4, + "grad_norm": 1.5960336923599243, + "learning_rate": 0.0003233256351039261, + "loss": 0.1057, + "step": 1350 + }, + { + "epoch": 5.5, + "grad_norm": 3.3678526878356934, + "learning_rate": 0.00032167601451666116, + "loss": 0.1092, + "step": 1375 + }, + { + "epoch": 5.6, + "grad_norm": 4.743562698364258, + "learning_rate": 0.00032002639392939627, + "loss": 0.1054, + "step": 1400 + }, + { + "epoch": 5.7, + "grad_norm": 12.860156059265137, + "learning_rate": 0.0003183767733421313, + "loss": 0.1009, + "step": 1425 + }, + { + "epoch": 5.8, + "grad_norm": 0.8820792436599731, + "learning_rate": 0.0003167271527548664, + "loss": 0.1219, + "step": 1450 + }, + { + "epoch": 5.9, + "grad_norm": 2.732572317123413, + "learning_rate": 0.00031507753216760147, + "loss": 0.0682, + "step": 1475 + }, + { + "epoch": 6.0, + "grad_norm": 9.810711860656738, + "learning_rate": 0.00031342791158033657, + "loss": 0.1134, + "step": 1500 + }, + { + "epoch": 6.1, + "grad_norm": 6.699489593505859, + "learning_rate": 0.00031177829099307157, + "loss": 0.0557, + "step": 1525 + }, + { + "epoch": 6.2, + "grad_norm": 14.691123008728027, + "learning_rate": 0.00031012867040580667, + "loss": 0.0599, + "step": 1550 + }, + { + "epoch": 6.3, + "grad_norm": 24.44135093688965, + "learning_rate": 0.0003084790498185417, + "loss": 0.0818, + "step": 1575 + }, + { + "epoch": 6.4, + "grad_norm": 0.033995576202869415, + "learning_rate": 0.0003068294292312768, + "loss": 0.0692, + "step": 1600 + }, + { + "epoch": 6.5, + "grad_norm": 6.722524642944336, + "learning_rate": 0.0003051798086440119, + "loss": 0.0702, + "step": 1625 + }, + { + "epoch": 6.6, + "grad_norm": 2.9519436359405518, + "learning_rate": 0.000303530188056747, + "loss": 0.073, + "step": 1650 + }, + { + "epoch": 6.7, + "grad_norm": 1.5118328332901, + "learning_rate": 0.000301880567469482, + "loss": 0.0926, + "step": 1675 + }, + { + "epoch": 6.8, + "grad_norm": 2.1923201084136963, + "learning_rate": 0.0003002309468822171, + "loss": 0.0821, + "step": 1700 + }, + { + "epoch": 6.9, + "grad_norm": 3.34309458732605, + "learning_rate": 0.0002985813262949522, + "loss": 0.0827, + "step": 1725 + }, + { + "epoch": 7.0, + "grad_norm": 10.289947509765625, + "learning_rate": 0.00029693170570768723, + "loss": 0.1162, + "step": 1750 + }, + { + "epoch": 7.1, + "grad_norm": 0.051936667412519455, + "learning_rate": 0.00029528208512042233, + "loss": 0.0626, + "step": 1775 + }, + { + "epoch": 7.2, + "grad_norm": 1.8913339376449585, + "learning_rate": 0.0002936324645331574, + "loss": 0.0657, + "step": 1800 + }, + { + "epoch": 7.3, + "grad_norm": 15.073695182800293, + "learning_rate": 0.00029198284394589243, + "loss": 0.0539, + "step": 1825 + }, + { + "epoch": 7.4, + "grad_norm": 4.514949798583984, + "learning_rate": 0.00029033322335862753, + "loss": 0.056, + "step": 1850 + }, + { + "epoch": 7.5, + "grad_norm": 0.46094000339508057, + "learning_rate": 0.0002886836027713626, + "loss": 0.0667, + "step": 1875 + }, + { + "epoch": 7.6, + "grad_norm": 4.682873725891113, + "learning_rate": 0.0002870339821840977, + "loss": 0.0673, + "step": 1900 + }, + { + "epoch": 7.7, + "grad_norm": 3.581599235534668, + "learning_rate": 0.00028538436159683273, + "loss": 0.0699, + "step": 1925 + }, + { + "epoch": 7.8, + "grad_norm": 3.2518703937530518, + "learning_rate": 0.00028373474100956784, + "loss": 0.0628, + "step": 1950 + }, + { + "epoch": 7.9, + "grad_norm": 0.24071219563484192, + "learning_rate": 0.0002820851204223029, + "loss": 0.0747, + "step": 1975 + }, + { + "epoch": 8.0, + "grad_norm": 2.8545596599578857, + "learning_rate": 0.00028043549983503794, + "loss": 0.0688, + "step": 2000 + }, + { + "epoch": 8.1, + "grad_norm": 0.049499694257974625, + "learning_rate": 0.00027878587924777304, + "loss": 0.0563, + "step": 2025 + }, + { + "epoch": 8.2, + "grad_norm": 22.112478256225586, + "learning_rate": 0.0002771362586605081, + "loss": 0.0545, + "step": 2050 + }, + { + "epoch": 8.3, + "grad_norm": 20.87473487854004, + "learning_rate": 0.0002754866380732432, + "loss": 0.0256, + "step": 2075 + }, + { + "epoch": 8.4, + "grad_norm": 2.317768096923828, + "learning_rate": 0.00027383701748597824, + "loss": 0.0753, + "step": 2100 + }, + { + "epoch": 8.5, + "grad_norm": 20.241336822509766, + "learning_rate": 0.0002721873968987133, + "loss": 0.0257, + "step": 2125 + }, + { + "epoch": 8.6, + "grad_norm": 14.994575500488281, + "learning_rate": 0.00027053777631144834, + "loss": 0.0748, + "step": 2150 + }, + { + "epoch": 8.7, + "grad_norm": 0.8268639445304871, + "learning_rate": 0.00026888815572418344, + "loss": 0.0805, + "step": 2175 + }, + { + "epoch": 8.8, + "grad_norm": 0.031071001663804054, + "learning_rate": 0.00026723853513691855, + "loss": 0.0603, + "step": 2200 + }, + { + "epoch": 8.9, + "grad_norm": 2.4572932720184326, + "learning_rate": 0.0002655889145496536, + "loss": 0.071, + "step": 2225 + }, + { + "epoch": 9.0, + "grad_norm": 0.21708321571350098, + "learning_rate": 0.0002639392939623887, + "loss": 0.0606, + "step": 2250 + }, + { + "epoch": 9.1, + "grad_norm": 14.928559303283691, + "learning_rate": 0.0002622896733751237, + "loss": 0.0443, + "step": 2275 + }, + { + "epoch": 9.2, + "grad_norm": 1.108826994895935, + "learning_rate": 0.0002606400527878588, + "loss": 0.0509, + "step": 2300 + }, + { + "epoch": 9.3, + "grad_norm": 0.05721910670399666, + "learning_rate": 0.00025899043220059385, + "loss": 0.0459, + "step": 2325 + }, + { + "epoch": 9.4, + "grad_norm": 7.81903600692749, + "learning_rate": 0.00025734081161332895, + "loss": 0.0535, + "step": 2350 + }, + { + "epoch": 9.5, + "grad_norm": 0.01475981343537569, + "learning_rate": 0.000255691191026064, + "loss": 0.0399, + "step": 2375 + }, + { + "epoch": 9.6, + "grad_norm": 5.496493339538574, + "learning_rate": 0.0002540415704387991, + "loss": 0.0744, + "step": 2400 + }, + { + "epoch": 9.7, + "grad_norm": 0.05929339677095413, + "learning_rate": 0.00025239194985153415, + "loss": 0.0416, + "step": 2425 + }, + { + "epoch": 9.8, + "grad_norm": 0.06961135566234589, + "learning_rate": 0.0002507423292642692, + "loss": 0.04, + "step": 2450 + }, + { + "epoch": 9.9, + "grad_norm": 0.02165246568620205, + "learning_rate": 0.0002490927086770043, + "loss": 0.0984, + "step": 2475 + }, + { + "epoch": 10.0, + "grad_norm": 25.155675888061523, + "learning_rate": 0.00024744308808973936, + "loss": 0.042, + "step": 2500 + }, + { + "epoch": 10.1, + "grad_norm": 8.566617965698242, + "learning_rate": 0.00024579346750247446, + "loss": 0.0385, + "step": 2525 + }, + { + "epoch": 10.2, + "grad_norm": 0.024992404505610466, + "learning_rate": 0.0002441438469152095, + "loss": 0.0327, + "step": 2550 + }, + { + "epoch": 10.3, + "grad_norm": 0.22233247756958008, + "learning_rate": 0.00024249422632794456, + "loss": 0.0211, + "step": 2575 + }, + { + "epoch": 10.4, + "grad_norm": 13.307366371154785, + "learning_rate": 0.00024084460574067963, + "loss": 0.0472, + "step": 2600 + }, + { + "epoch": 10.5, + "grad_norm": 0.17787696421146393, + "learning_rate": 0.0002391949851534147, + "loss": 0.0346, + "step": 2625 + }, + { + "epoch": 10.6, + "grad_norm": 1.2845549583435059, + "learning_rate": 0.0002375453645661498, + "loss": 0.0587, + "step": 2650 + }, + { + "epoch": 10.7, + "grad_norm": 0.0603482760488987, + "learning_rate": 0.00023589574397888486, + "loss": 0.0394, + "step": 2675 + }, + { + "epoch": 10.8, + "grad_norm": 0.031819943338632584, + "learning_rate": 0.00023424612339161997, + "loss": 0.0434, + "step": 2700 + }, + { + "epoch": 10.9, + "grad_norm": 1.5738537311553955, + "learning_rate": 0.000232596502804355, + "loss": 0.0393, + "step": 2725 + }, + { + "epoch": 11.0, + "grad_norm": 0.6862583756446838, + "learning_rate": 0.00023094688221709007, + "loss": 0.0183, + "step": 2750 + }, + { + "epoch": 11.1, + "grad_norm": 0.0053489054553210735, + "learning_rate": 0.00022929726162982514, + "loss": 0.0277, + "step": 2775 + }, + { + "epoch": 11.2, + "grad_norm": 0.04365606606006622, + "learning_rate": 0.00022764764104256022, + "loss": 0.0337, + "step": 2800 + }, + { + "epoch": 11.3, + "grad_norm": 1.3348641395568848, + "learning_rate": 0.0002259980204552953, + "loss": 0.0219, + "step": 2825 + }, + { + "epoch": 11.4, + "grad_norm": 3.3695547580718994, + "learning_rate": 0.00022434839986803037, + "loss": 0.0227, + "step": 2850 + }, + { + "epoch": 11.5, + "grad_norm": 2.507455587387085, + "learning_rate": 0.00022269877928076542, + "loss": 0.0443, + "step": 2875 + }, + { + "epoch": 11.6, + "grad_norm": 0.013451912440359592, + "learning_rate": 0.0002210491586935005, + "loss": 0.0486, + "step": 2900 + }, + { + "epoch": 11.7, + "grad_norm": 0.031177405267953873, + "learning_rate": 0.00021939953810623557, + "loss": 0.0511, + "step": 2925 + }, + { + "epoch": 11.8, + "grad_norm": 0.02148960903286934, + "learning_rate": 0.00021774991751897065, + "loss": 0.0269, + "step": 2950 + }, + { + "epoch": 11.9, + "grad_norm": 0.01219563465565443, + "learning_rate": 0.00021610029693170573, + "loss": 0.042, + "step": 2975 + }, + { + "epoch": 12.0, + "grad_norm": 0.2062515765428543, + "learning_rate": 0.0002144506763444408, + "loss": 0.031, + "step": 3000 + }, + { + "epoch": 12.1, + "grad_norm": 3.235403060913086, + "learning_rate": 0.00021280105575717582, + "loss": 0.0414, + "step": 3025 + }, + { + "epoch": 12.2, + "grad_norm": 1.089821696281433, + "learning_rate": 0.00021115143516991093, + "loss": 0.0692, + "step": 3050 + }, + { + "epoch": 12.3, + "grad_norm": 0.013664484024047852, + "learning_rate": 0.000209501814582646, + "loss": 0.0367, + "step": 3075 + }, + { + "epoch": 12.4, + "grad_norm": 1.3101387023925781, + "learning_rate": 0.00020785219399538108, + "loss": 0.0277, + "step": 3100 + }, + { + "epoch": 12.5, + "grad_norm": 3.244853973388672, + "learning_rate": 0.00020620257340811616, + "loss": 0.0373, + "step": 3125 + }, + { + "epoch": 12.6, + "grad_norm": 0.013548053801059723, + "learning_rate": 0.00020455295282085123, + "loss": 0.027, + "step": 3150 + }, + { + "epoch": 12.7, + "grad_norm": 0.004670978989452124, + "learning_rate": 0.00020290333223358626, + "loss": 0.0252, + "step": 3175 + }, + { + "epoch": 12.8, + "grad_norm": 0.007080752402544022, + "learning_rate": 0.00020125371164632133, + "loss": 0.0187, + "step": 3200 + }, + { + "epoch": 12.9, + "grad_norm": 0.005918944254517555, + "learning_rate": 0.0001996040910590564, + "loss": 0.0423, + "step": 3225 + }, + { + "epoch": 13.0, + "grad_norm": 4.530269145965576, + "learning_rate": 0.00019795447047179148, + "loss": 0.0381, + "step": 3250 + }, + { + "epoch": 13.1, + "grad_norm": 0.00533739197999239, + "learning_rate": 0.0001963048498845266, + "loss": 0.0097, + "step": 3275 + }, + { + "epoch": 13.2, + "grad_norm": 0.7641226649284363, + "learning_rate": 0.00019465522929726164, + "loss": 0.0205, + "step": 3300 + }, + { + "epoch": 13.3, + "grad_norm": 2.16174054145813, + "learning_rate": 0.00019300560870999671, + "loss": 0.0174, + "step": 3325 + }, + { + "epoch": 13.4, + "grad_norm": 0.00814723875373602, + "learning_rate": 0.0001913559881227318, + "loss": 0.0215, + "step": 3350 + }, + { + "epoch": 13.5, + "grad_norm": 10.500170707702637, + "learning_rate": 0.00018970636753546684, + "loss": 0.0226, + "step": 3375 + }, + { + "epoch": 13.6, + "grad_norm": 2.612359046936035, + "learning_rate": 0.00018805674694820192, + "loss": 0.0314, + "step": 3400 + }, + { + "epoch": 13.7, + "grad_norm": 0.11265736818313599, + "learning_rate": 0.000186407126360937, + "loss": 0.0234, + "step": 3425 + }, + { + "epoch": 13.8, + "grad_norm": 0.06453213095664978, + "learning_rate": 0.00018475750577367207, + "loss": 0.0314, + "step": 3450 + }, + { + "epoch": 13.9, + "grad_norm": 0.01650763861835003, + "learning_rate": 0.00018310788518640715, + "loss": 0.0403, + "step": 3475 + }, + { + "epoch": 14.0, + "grad_norm": 0.002764373552054167, + "learning_rate": 0.00018145826459914222, + "loss": 0.0137, + "step": 3500 + }, + { + "epoch": 14.1, + "grad_norm": 0.0022430066019296646, + "learning_rate": 0.00017980864401187727, + "loss": 0.0084, + "step": 3525 + }, + { + "epoch": 14.2, + "grad_norm": 0.007267679553478956, + "learning_rate": 0.00017815902342461235, + "loss": 0.0161, + "step": 3550 + }, + { + "epoch": 14.3, + "grad_norm": 0.004844362381845713, + "learning_rate": 0.00017650940283734742, + "loss": 0.0399, + "step": 3575 + }, + { + "epoch": 14.4, + "grad_norm": 0.05529040843248367, + "learning_rate": 0.00017485978225008247, + "loss": 0.0192, + "step": 3600 + }, + { + "epoch": 14.5, + "grad_norm": 0.008248478174209595, + "learning_rate": 0.00017321016166281755, + "loss": 0.0079, + "step": 3625 + }, + { + "epoch": 14.6, + "grad_norm": 0.02617962658405304, + "learning_rate": 0.00017156054107555263, + "loss": 0.0244, + "step": 3650 + }, + { + "epoch": 14.7, + "grad_norm": 0.004497932270169258, + "learning_rate": 0.0001699109204882877, + "loss": 0.0464, + "step": 3675 + }, + { + "epoch": 14.8, + "grad_norm": 0.008586671203374863, + "learning_rate": 0.00016826129990102278, + "loss": 0.0354, + "step": 3700 + }, + { + "epoch": 14.9, + "grad_norm": 0.30177807807922363, + "learning_rate": 0.00016661167931375785, + "loss": 0.0147, + "step": 3725 + }, + { + "epoch": 15.0, + "grad_norm": 0.10390808433294296, + "learning_rate": 0.0001649620587264929, + "loss": 0.0078, + "step": 3750 + }, + { + "epoch": 15.1, + "grad_norm": 1.584868311882019, + "learning_rate": 0.00016331243813922798, + "loss": 0.0059, + "step": 3775 + }, + { + "epoch": 15.2, + "grad_norm": 0.0038078685756772757, + "learning_rate": 0.00016166281755196306, + "loss": 0.0143, + "step": 3800 + }, + { + "epoch": 15.3, + "grad_norm": 0.017795555293560028, + "learning_rate": 0.00016001319696469813, + "loss": 0.0003, + "step": 3825 + }, + { + "epoch": 15.4, + "grad_norm": 7.246474742889404, + "learning_rate": 0.0001583635763774332, + "loss": 0.0288, + "step": 3850 + }, + { + "epoch": 15.5, + "grad_norm": 0.0014716258738189936, + "learning_rate": 0.00015671395579016829, + "loss": 0.0113, + "step": 3875 + }, + { + "epoch": 15.6, + "grad_norm": 0.002159764291718602, + "learning_rate": 0.00015506433520290334, + "loss": 0.0138, + "step": 3900 + }, + { + "epoch": 15.7, + "grad_norm": 27.16730308532715, + "learning_rate": 0.0001534147146156384, + "loss": 0.0048, + "step": 3925 + }, + { + "epoch": 15.8, + "grad_norm": 0.004570333287119865, + "learning_rate": 0.0001517650940283735, + "loss": 0.0144, + "step": 3950 + }, + { + "epoch": 15.9, + "grad_norm": 0.008330491371452808, + "learning_rate": 0.00015011547344110854, + "loss": 0.0371, + "step": 3975 + }, + { + "epoch": 16.0, + "grad_norm": 15.75979232788086, + "learning_rate": 0.00014846585285384361, + "loss": 0.0207, + "step": 4000 + }, + { + "epoch": 16.1, + "grad_norm": 14.096012115478516, + "learning_rate": 0.0001468162322665787, + "loss": 0.018, + "step": 4025 + }, + { + "epoch": 16.2, + "grad_norm": 0.0015319391386583447, + "learning_rate": 0.00014516661167931377, + "loss": 0.0109, + "step": 4050 + }, + { + "epoch": 16.3, + "grad_norm": 0.003954900894314051, + "learning_rate": 0.00014351699109204884, + "loss": 0.012, + "step": 4075 + }, + { + "epoch": 16.4, + "grad_norm": 37.79055404663086, + "learning_rate": 0.00014186737050478392, + "loss": 0.0301, + "step": 4100 + }, + { + "epoch": 16.5, + "grad_norm": 0.0014516041846945882, + "learning_rate": 0.00014021774991751897, + "loss": 0.0116, + "step": 4125 + }, + { + "epoch": 16.6, + "grad_norm": 0.009134252555668354, + "learning_rate": 0.00013856812933025404, + "loss": 0.0126, + "step": 4150 + }, + { + "epoch": 16.7, + "grad_norm": 19.0325927734375, + "learning_rate": 0.00013691850874298912, + "loss": 0.0398, + "step": 4175 + }, + { + "epoch": 16.8, + "grad_norm": 0.41503193974494934, + "learning_rate": 0.00013526888815572417, + "loss": 0.001, + "step": 4200 + }, + { + "epoch": 16.9, + "grad_norm": 0.0014813834568485618, + "learning_rate": 0.00013361926756845927, + "loss": 0.0077, + "step": 4225 + }, + { + "epoch": 17.0, + "grad_norm": 0.007494773715734482, + "learning_rate": 0.00013196964698119435, + "loss": 0.0052, + "step": 4250 + }, + { + "epoch": 17.1, + "grad_norm": 0.002411301014944911, + "learning_rate": 0.0001303200263939294, + "loss": 0.0038, + "step": 4275 + }, + { + "epoch": 17.2, + "grad_norm": 0.01521711703389883, + "learning_rate": 0.00012867040580666448, + "loss": 0.0051, + "step": 4300 + }, + { + "epoch": 17.3, + "grad_norm": 0.0010520346695557237, + "learning_rate": 0.00012702078521939955, + "loss": 0.0167, + "step": 4325 + }, + { + "epoch": 17.4, + "grad_norm": 0.054758671671152115, + "learning_rate": 0.0001253711646321346, + "loss": 0.011, + "step": 4350 + }, + { + "epoch": 17.5, + "grad_norm": 0.0012187627144157887, + "learning_rate": 0.00012372154404486968, + "loss": 0.0239, + "step": 4375 + }, + { + "epoch": 17.6, + "grad_norm": 58.37099838256836, + "learning_rate": 0.00012207192345760475, + "loss": 0.0055, + "step": 4400 + }, + { + "epoch": 17.7, + "grad_norm": 0.27519506216049194, + "learning_rate": 0.00012042230287033982, + "loss": 0.0332, + "step": 4425 + }, + { + "epoch": 17.8, + "grad_norm": 0.0011721524642780423, + "learning_rate": 0.0001187726822830749, + "loss": 0.0048, + "step": 4450 + }, + { + "epoch": 17.9, + "grad_norm": 0.0015529991360381246, + "learning_rate": 0.00011712306169580998, + "loss": 0.0134, + "step": 4475 + }, + { + "epoch": 18.0, + "grad_norm": 0.0008489767787978053, + "learning_rate": 0.00011547344110854503, + "loss": 0.0148, + "step": 4500 + }, + { + "epoch": 18.1, + "grad_norm": 0.0010399922030046582, + "learning_rate": 0.00011382382052128011, + "loss": 0.0046, + "step": 4525 + }, + { + "epoch": 18.2, + "grad_norm": 0.0017199565190821886, + "learning_rate": 0.00011217419993401519, + "loss": 0.0123, + "step": 4550 + }, + { + "epoch": 18.3, + "grad_norm": 0.004095083102583885, + "learning_rate": 0.00011052457934675025, + "loss": 0.0078, + "step": 4575 + }, + { + "epoch": 18.4, + "grad_norm": 0.005569108761847019, + "learning_rate": 0.00010887495875948532, + "loss": 0.0001, + "step": 4600 + }, + { + "epoch": 18.5, + "grad_norm": 0.0013379952870309353, + "learning_rate": 0.0001072253381722204, + "loss": 0.0123, + "step": 4625 + }, + { + "epoch": 18.6, + "grad_norm": 0.13633179664611816, + "learning_rate": 0.00010557571758495546, + "loss": 0.016, + "step": 4650 + }, + { + "epoch": 18.7, + "grad_norm": 0.00383689533919096, + "learning_rate": 0.00010392609699769054, + "loss": 0.0238, + "step": 4675 + }, + { + "epoch": 18.8, + "grad_norm": 0.011090466752648354, + "learning_rate": 0.00010227647641042562, + "loss": 0.0299, + "step": 4700 + }, + { + "epoch": 18.9, + "grad_norm": 0.01791750080883503, + "learning_rate": 0.00010062685582316067, + "loss": 0.0006, + "step": 4725 + }, + { + "epoch": 19.0, + "grad_norm": 20.31785774230957, + "learning_rate": 9.897723523589574e-05, + "loss": 0.0067, + "step": 4750 + }, + { + "epoch": 19.1, + "grad_norm": 0.0011455218773335218, + "learning_rate": 9.732761464863082e-05, + "loss": 0.0138, + "step": 4775 + }, + { + "epoch": 19.2, + "grad_norm": 0.0014105259906500578, + "learning_rate": 9.56779940613659e-05, + "loss": 0.0005, + "step": 4800 + }, + { + "epoch": 19.3, + "grad_norm": 0.002618618542328477, + "learning_rate": 9.402837347410096e-05, + "loss": 0.0112, + "step": 4825 + }, + { + "epoch": 19.4, + "grad_norm": 0.001108857337385416, + "learning_rate": 9.237875288683603e-05, + "loss": 0.0164, + "step": 4850 + }, + { + "epoch": 19.5, + "grad_norm": 0.0010457661701366305, + "learning_rate": 9.072913229957111e-05, + "loss": 0.0104, + "step": 4875 + }, + { + "epoch": 19.6, + "grad_norm": 0.0012097499566152692, + "learning_rate": 8.907951171230617e-05, + "loss": 0.0039, + "step": 4900 + }, + { + "epoch": 19.7, + "grad_norm": 0.04536284878849983, + "learning_rate": 8.742989112504124e-05, + "loss": 0.0154, + "step": 4925 + }, + { + "epoch": 19.8, + "grad_norm": 29.97249984741211, + "learning_rate": 8.578027053777631e-05, + "loss": 0.0246, + "step": 4950 + }, + { + "epoch": 19.9, + "grad_norm": 0.01911325752735138, + "learning_rate": 8.413064995051139e-05, + "loss": 0.0176, + "step": 4975 + }, + { + "epoch": 20.0, + "grad_norm": 0.3129942715167999, + "learning_rate": 8.248102936324645e-05, + "loss": 0.0029, + "step": 5000 + } + ], + "logging_steps": 25, + "max_steps": 6250, + "num_input_tokens_seen": 0, + "num_train_epochs": 25, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 254097408000000.0, + "train_batch_size": 20, + "trial_name": null, + "trial_params": null +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-5000/training_args.bin b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-5000/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..ef3fc22e099df49861bca49d15da2b959cfcb6df --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-5000/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ec1d8ed36c45af65d67a23de6c5216a486932bfb09b720e1a4417b82eaaa215d +size 5304 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-5000/vocab.txt b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-5000/vocab.txt new file mode 100644 index 0000000000000000000000000000000000000000..cc9ad53732e21ea9180f18e1fb67a38cc5702d4c --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-5000/vocab.txt @@ -0,0 +1,4101 @@ +[PAD] +[UNK] +[CLS] +[SEP] +[MASK] +AAAAAA +AAAAAT +AAAAAC +AAAAAG +AAAATA +AAAATT +AAAATC +AAAATG +AAAACA +AAAACT +AAAACC +AAAACG +AAAAGA +AAAAGT +AAAAGC +AAAAGG +AAATAA +AAATAT +AAATAC +AAATAG +AAATTA +AAATTT +AAATTC +AAATTG +AAATCA +AAATCT +AAATCC +AAATCG +AAATGA +AAATGT +AAATGC +AAATGG +AAACAA +AAACAT +AAACAC +AAACAG +AAACTA +AAACTT +AAACTC +AAACTG +AAACCA +AAACCT +AAACCC +AAACCG +AAACGA +AAACGT +AAACGC +AAACGG +AAAGAA +AAAGAT +AAAGAC +AAAGAG +AAAGTA +AAAGTT +AAAGTC +AAAGTG +AAAGCA +AAAGCT +AAAGCC +AAAGCG +AAAGGA +AAAGGT +AAAGGC +AAAGGG +AATAAA +AATAAT +AATAAC +AATAAG +AATATA +AATATT +AATATC +AATATG +AATACA +AATACT +AATACC +AATACG +AATAGA +AATAGT +AATAGC +AATAGG +AATTAA +AATTAT +AATTAC +AATTAG +AATTTA +AATTTT +AATTTC +AATTTG +AATTCA +AATTCT +AATTCC +AATTCG +AATTGA +AATTGT +AATTGC +AATTGG +AATCAA +AATCAT +AATCAC +AATCAG +AATCTA +AATCTT +AATCTC +AATCTG +AATCCA +AATCCT +AATCCC +AATCCG +AATCGA +AATCGT +AATCGC +AATCGG +AATGAA +AATGAT +AATGAC +AATGAG +AATGTA +AATGTT +AATGTC +AATGTG +AATGCA +AATGCT +AATGCC +AATGCG +AATGGA +AATGGT +AATGGC +AATGGG +AACAAA +AACAAT +AACAAC +AACAAG +AACATA +AACATT +AACATC +AACATG +AACACA +AACACT +AACACC +AACACG +AACAGA +AACAGT +AACAGC +AACAGG +AACTAA +AACTAT +AACTAC +AACTAG +AACTTA +AACTTT +AACTTC +AACTTG +AACTCA +AACTCT +AACTCC +AACTCG +AACTGA +AACTGT +AACTGC +AACTGG +AACCAA +AACCAT +AACCAC +AACCAG +AACCTA +AACCTT +AACCTC +AACCTG +AACCCA +AACCCT +AACCCC +AACCCG +AACCGA +AACCGT +AACCGC +AACCGG +AACGAA +AACGAT +AACGAC +AACGAG +AACGTA +AACGTT +AACGTC +AACGTG +AACGCA +AACGCT +AACGCC +AACGCG +AACGGA +AACGGT +AACGGC +AACGGG +AAGAAA +AAGAAT +AAGAAC +AAGAAG +AAGATA +AAGATT +AAGATC +AAGATG +AAGACA +AAGACT +AAGACC +AAGACG +AAGAGA +AAGAGT +AAGAGC +AAGAGG +AAGTAA +AAGTAT +AAGTAC +AAGTAG +AAGTTA +AAGTTT +AAGTTC +AAGTTG +AAGTCA +AAGTCT +AAGTCC +AAGTCG +AAGTGA +AAGTGT +AAGTGC +AAGTGG +AAGCAA +AAGCAT +AAGCAC +AAGCAG +AAGCTA +AAGCTT +AAGCTC +AAGCTG +AAGCCA +AAGCCT +AAGCCC +AAGCCG +AAGCGA +AAGCGT +AAGCGC +AAGCGG +AAGGAA +AAGGAT +AAGGAC +AAGGAG +AAGGTA +AAGGTT +AAGGTC +AAGGTG +AAGGCA +AAGGCT +AAGGCC +AAGGCG +AAGGGA +AAGGGT +AAGGGC +AAGGGG +ATAAAA +ATAAAT +ATAAAC +ATAAAG +ATAATA +ATAATT +ATAATC +ATAATG +ATAACA +ATAACT +ATAACC +ATAACG +ATAAGA +ATAAGT +ATAAGC +ATAAGG +ATATAA +ATATAT +ATATAC +ATATAG +ATATTA +ATATTT +ATATTC +ATATTG +ATATCA +ATATCT +ATATCC +ATATCG +ATATGA +ATATGT +ATATGC +ATATGG +ATACAA +ATACAT +ATACAC +ATACAG +ATACTA +ATACTT +ATACTC +ATACTG +ATACCA +ATACCT +ATACCC +ATACCG +ATACGA +ATACGT +ATACGC +ATACGG +ATAGAA +ATAGAT +ATAGAC +ATAGAG +ATAGTA +ATAGTT +ATAGTC +ATAGTG +ATAGCA +ATAGCT +ATAGCC +ATAGCG +ATAGGA +ATAGGT +ATAGGC +ATAGGG +ATTAAA +ATTAAT +ATTAAC +ATTAAG +ATTATA +ATTATT +ATTATC +ATTATG +ATTACA +ATTACT +ATTACC +ATTACG +ATTAGA +ATTAGT +ATTAGC +ATTAGG +ATTTAA +ATTTAT +ATTTAC +ATTTAG +ATTTTA +ATTTTT +ATTTTC +ATTTTG +ATTTCA +ATTTCT +ATTTCC +ATTTCG +ATTTGA +ATTTGT +ATTTGC +ATTTGG +ATTCAA +ATTCAT +ATTCAC +ATTCAG +ATTCTA +ATTCTT +ATTCTC +ATTCTG +ATTCCA +ATTCCT +ATTCCC +ATTCCG +ATTCGA +ATTCGT +ATTCGC +ATTCGG +ATTGAA +ATTGAT +ATTGAC +ATTGAG +ATTGTA +ATTGTT +ATTGTC +ATTGTG +ATTGCA +ATTGCT +ATTGCC +ATTGCG +ATTGGA +ATTGGT +ATTGGC +ATTGGG +ATCAAA +ATCAAT +ATCAAC +ATCAAG +ATCATA +ATCATT +ATCATC +ATCATG +ATCACA +ATCACT +ATCACC +ATCACG +ATCAGA +ATCAGT +ATCAGC +ATCAGG +ATCTAA +ATCTAT +ATCTAC +ATCTAG +ATCTTA +ATCTTT +ATCTTC +ATCTTG +ATCTCA +ATCTCT +ATCTCC +ATCTCG +ATCTGA +ATCTGT +ATCTGC +ATCTGG +ATCCAA +ATCCAT +ATCCAC +ATCCAG +ATCCTA +ATCCTT +ATCCTC +ATCCTG +ATCCCA +ATCCCT +ATCCCC +ATCCCG +ATCCGA +ATCCGT +ATCCGC +ATCCGG +ATCGAA +ATCGAT +ATCGAC +ATCGAG +ATCGTA +ATCGTT +ATCGTC +ATCGTG +ATCGCA +ATCGCT +ATCGCC +ATCGCG +ATCGGA +ATCGGT +ATCGGC +ATCGGG +ATGAAA +ATGAAT +ATGAAC +ATGAAG +ATGATA +ATGATT +ATGATC +ATGATG +ATGACA +ATGACT +ATGACC +ATGACG +ATGAGA +ATGAGT +ATGAGC +ATGAGG +ATGTAA +ATGTAT +ATGTAC +ATGTAG +ATGTTA +ATGTTT +ATGTTC +ATGTTG +ATGTCA +ATGTCT +ATGTCC +ATGTCG +ATGTGA +ATGTGT +ATGTGC +ATGTGG +ATGCAA +ATGCAT +ATGCAC +ATGCAG +ATGCTA +ATGCTT +ATGCTC +ATGCTG +ATGCCA +ATGCCT +ATGCCC +ATGCCG +ATGCGA +ATGCGT +ATGCGC +ATGCGG +ATGGAA +ATGGAT +ATGGAC +ATGGAG +ATGGTA +ATGGTT +ATGGTC +ATGGTG +ATGGCA +ATGGCT +ATGGCC +ATGGCG +ATGGGA +ATGGGT +ATGGGC +ATGGGG +ACAAAA +ACAAAT +ACAAAC +ACAAAG +ACAATA +ACAATT +ACAATC +ACAATG +ACAACA +ACAACT +ACAACC +ACAACG +ACAAGA +ACAAGT +ACAAGC +ACAAGG +ACATAA +ACATAT +ACATAC +ACATAG +ACATTA +ACATTT +ACATTC +ACATTG +ACATCA +ACATCT +ACATCC +ACATCG +ACATGA +ACATGT +ACATGC +ACATGG +ACACAA +ACACAT +ACACAC +ACACAG +ACACTA +ACACTT +ACACTC +ACACTG +ACACCA +ACACCT +ACACCC +ACACCG +ACACGA +ACACGT +ACACGC +ACACGG +ACAGAA +ACAGAT +ACAGAC +ACAGAG +ACAGTA +ACAGTT +ACAGTC +ACAGTG +ACAGCA +ACAGCT +ACAGCC +ACAGCG +ACAGGA +ACAGGT +ACAGGC +ACAGGG +ACTAAA +ACTAAT +ACTAAC +ACTAAG +ACTATA +ACTATT +ACTATC +ACTATG +ACTACA +ACTACT +ACTACC +ACTACG +ACTAGA +ACTAGT +ACTAGC +ACTAGG +ACTTAA +ACTTAT +ACTTAC +ACTTAG +ACTTTA +ACTTTT +ACTTTC +ACTTTG +ACTTCA +ACTTCT +ACTTCC +ACTTCG +ACTTGA +ACTTGT +ACTTGC +ACTTGG +ACTCAA +ACTCAT +ACTCAC +ACTCAG +ACTCTA +ACTCTT +ACTCTC +ACTCTG +ACTCCA +ACTCCT +ACTCCC +ACTCCG +ACTCGA +ACTCGT +ACTCGC +ACTCGG +ACTGAA +ACTGAT +ACTGAC +ACTGAG +ACTGTA +ACTGTT +ACTGTC +ACTGTG +ACTGCA +ACTGCT +ACTGCC +ACTGCG +ACTGGA +ACTGGT +ACTGGC +ACTGGG +ACCAAA +ACCAAT +ACCAAC +ACCAAG +ACCATA +ACCATT +ACCATC +ACCATG +ACCACA +ACCACT +ACCACC +ACCACG +ACCAGA +ACCAGT +ACCAGC +ACCAGG +ACCTAA +ACCTAT +ACCTAC +ACCTAG +ACCTTA +ACCTTT +ACCTTC +ACCTTG +ACCTCA +ACCTCT +ACCTCC +ACCTCG +ACCTGA +ACCTGT +ACCTGC +ACCTGG +ACCCAA +ACCCAT +ACCCAC +ACCCAG +ACCCTA +ACCCTT +ACCCTC +ACCCTG +ACCCCA +ACCCCT +ACCCCC +ACCCCG +ACCCGA +ACCCGT +ACCCGC +ACCCGG +ACCGAA +ACCGAT +ACCGAC +ACCGAG +ACCGTA +ACCGTT +ACCGTC +ACCGTG +ACCGCA +ACCGCT +ACCGCC +ACCGCG +ACCGGA +ACCGGT +ACCGGC +ACCGGG +ACGAAA +ACGAAT +ACGAAC +ACGAAG +ACGATA +ACGATT +ACGATC +ACGATG +ACGACA +ACGACT +ACGACC +ACGACG +ACGAGA +ACGAGT +ACGAGC +ACGAGG +ACGTAA +ACGTAT +ACGTAC +ACGTAG +ACGTTA +ACGTTT +ACGTTC +ACGTTG +ACGTCA +ACGTCT +ACGTCC +ACGTCG +ACGTGA +ACGTGT +ACGTGC +ACGTGG +ACGCAA +ACGCAT +ACGCAC +ACGCAG +ACGCTA +ACGCTT +ACGCTC +ACGCTG +ACGCCA +ACGCCT +ACGCCC +ACGCCG +ACGCGA +ACGCGT +ACGCGC +ACGCGG +ACGGAA +ACGGAT +ACGGAC +ACGGAG +ACGGTA +ACGGTT +ACGGTC +ACGGTG +ACGGCA +ACGGCT +ACGGCC +ACGGCG +ACGGGA +ACGGGT +ACGGGC +ACGGGG +AGAAAA +AGAAAT +AGAAAC +AGAAAG +AGAATA +AGAATT +AGAATC +AGAATG +AGAACA +AGAACT +AGAACC +AGAACG +AGAAGA +AGAAGT +AGAAGC +AGAAGG +AGATAA +AGATAT +AGATAC +AGATAG +AGATTA +AGATTT +AGATTC +AGATTG +AGATCA +AGATCT +AGATCC +AGATCG +AGATGA +AGATGT +AGATGC +AGATGG +AGACAA +AGACAT +AGACAC +AGACAG +AGACTA +AGACTT +AGACTC +AGACTG +AGACCA +AGACCT +AGACCC +AGACCG +AGACGA +AGACGT +AGACGC +AGACGG +AGAGAA +AGAGAT +AGAGAC +AGAGAG +AGAGTA +AGAGTT +AGAGTC +AGAGTG +AGAGCA +AGAGCT +AGAGCC +AGAGCG +AGAGGA +AGAGGT +AGAGGC +AGAGGG +AGTAAA +AGTAAT +AGTAAC +AGTAAG +AGTATA +AGTATT +AGTATC +AGTATG +AGTACA +AGTACT +AGTACC +AGTACG +AGTAGA +AGTAGT +AGTAGC +AGTAGG +AGTTAA +AGTTAT +AGTTAC +AGTTAG +AGTTTA +AGTTTT +AGTTTC +AGTTTG +AGTTCA +AGTTCT +AGTTCC +AGTTCG +AGTTGA +AGTTGT +AGTTGC +AGTTGG +AGTCAA +AGTCAT +AGTCAC +AGTCAG +AGTCTA +AGTCTT +AGTCTC +AGTCTG +AGTCCA +AGTCCT +AGTCCC +AGTCCG +AGTCGA +AGTCGT +AGTCGC +AGTCGG +AGTGAA +AGTGAT +AGTGAC +AGTGAG +AGTGTA +AGTGTT +AGTGTC +AGTGTG +AGTGCA +AGTGCT +AGTGCC +AGTGCG +AGTGGA +AGTGGT +AGTGGC +AGTGGG +AGCAAA +AGCAAT +AGCAAC +AGCAAG +AGCATA +AGCATT +AGCATC +AGCATG +AGCACA +AGCACT +AGCACC +AGCACG +AGCAGA +AGCAGT +AGCAGC +AGCAGG +AGCTAA +AGCTAT +AGCTAC +AGCTAG +AGCTTA +AGCTTT +AGCTTC +AGCTTG +AGCTCA +AGCTCT +AGCTCC +AGCTCG +AGCTGA +AGCTGT +AGCTGC +AGCTGG +AGCCAA +AGCCAT +AGCCAC +AGCCAG +AGCCTA +AGCCTT +AGCCTC +AGCCTG +AGCCCA +AGCCCT +AGCCCC +AGCCCG +AGCCGA +AGCCGT +AGCCGC +AGCCGG +AGCGAA +AGCGAT +AGCGAC +AGCGAG +AGCGTA +AGCGTT +AGCGTC +AGCGTG +AGCGCA +AGCGCT +AGCGCC +AGCGCG +AGCGGA +AGCGGT +AGCGGC +AGCGGG +AGGAAA +AGGAAT +AGGAAC +AGGAAG +AGGATA +AGGATT +AGGATC +AGGATG +AGGACA +AGGACT +AGGACC +AGGACG +AGGAGA +AGGAGT +AGGAGC +AGGAGG +AGGTAA +AGGTAT +AGGTAC +AGGTAG +AGGTTA +AGGTTT +AGGTTC +AGGTTG +AGGTCA +AGGTCT +AGGTCC +AGGTCG +AGGTGA +AGGTGT +AGGTGC +AGGTGG +AGGCAA +AGGCAT +AGGCAC +AGGCAG +AGGCTA +AGGCTT +AGGCTC +AGGCTG +AGGCCA +AGGCCT +AGGCCC +AGGCCG +AGGCGA +AGGCGT +AGGCGC +AGGCGG +AGGGAA +AGGGAT +AGGGAC +AGGGAG +AGGGTA +AGGGTT +AGGGTC +AGGGTG +AGGGCA +AGGGCT +AGGGCC +AGGGCG +AGGGGA +AGGGGT +AGGGGC +AGGGGG +TAAAAA +TAAAAT +TAAAAC +TAAAAG +TAAATA +TAAATT +TAAATC +TAAATG +TAAACA +TAAACT +TAAACC +TAAACG +TAAAGA +TAAAGT +TAAAGC +TAAAGG +TAATAA +TAATAT +TAATAC +TAATAG +TAATTA +TAATTT +TAATTC +TAATTG +TAATCA +TAATCT +TAATCC +TAATCG +TAATGA +TAATGT +TAATGC +TAATGG +TAACAA +TAACAT +TAACAC +TAACAG +TAACTA +TAACTT +TAACTC +TAACTG +TAACCA +TAACCT +TAACCC +TAACCG +TAACGA +TAACGT +TAACGC +TAACGG +TAAGAA +TAAGAT +TAAGAC +TAAGAG +TAAGTA +TAAGTT +TAAGTC +TAAGTG +TAAGCA +TAAGCT +TAAGCC +TAAGCG +TAAGGA +TAAGGT +TAAGGC +TAAGGG +TATAAA +TATAAT +TATAAC +TATAAG +TATATA +TATATT +TATATC +TATATG +TATACA +TATACT +TATACC +TATACG +TATAGA +TATAGT +TATAGC +TATAGG +TATTAA +TATTAT +TATTAC +TATTAG +TATTTA +TATTTT +TATTTC +TATTTG +TATTCA +TATTCT +TATTCC +TATTCG +TATTGA +TATTGT +TATTGC +TATTGG +TATCAA +TATCAT +TATCAC +TATCAG +TATCTA +TATCTT +TATCTC +TATCTG +TATCCA +TATCCT +TATCCC +TATCCG +TATCGA +TATCGT +TATCGC +TATCGG +TATGAA +TATGAT +TATGAC +TATGAG +TATGTA +TATGTT +TATGTC +TATGTG +TATGCA +TATGCT +TATGCC +TATGCG +TATGGA +TATGGT +TATGGC +TATGGG +TACAAA +TACAAT +TACAAC +TACAAG +TACATA +TACATT +TACATC +TACATG +TACACA +TACACT +TACACC +TACACG +TACAGA +TACAGT +TACAGC +TACAGG +TACTAA +TACTAT +TACTAC +TACTAG +TACTTA +TACTTT +TACTTC +TACTTG +TACTCA +TACTCT +TACTCC +TACTCG +TACTGA +TACTGT +TACTGC +TACTGG +TACCAA +TACCAT +TACCAC +TACCAG +TACCTA +TACCTT +TACCTC +TACCTG +TACCCA +TACCCT +TACCCC +TACCCG +TACCGA +TACCGT +TACCGC +TACCGG +TACGAA +TACGAT +TACGAC +TACGAG +TACGTA +TACGTT +TACGTC +TACGTG +TACGCA +TACGCT +TACGCC +TACGCG +TACGGA +TACGGT +TACGGC +TACGGG +TAGAAA +TAGAAT +TAGAAC +TAGAAG +TAGATA +TAGATT +TAGATC +TAGATG +TAGACA +TAGACT +TAGACC +TAGACG +TAGAGA +TAGAGT +TAGAGC +TAGAGG +TAGTAA +TAGTAT +TAGTAC +TAGTAG +TAGTTA +TAGTTT +TAGTTC +TAGTTG +TAGTCA +TAGTCT +TAGTCC +TAGTCG +TAGTGA +TAGTGT +TAGTGC +TAGTGG +TAGCAA +TAGCAT +TAGCAC +TAGCAG +TAGCTA +TAGCTT +TAGCTC +TAGCTG +TAGCCA +TAGCCT +TAGCCC +TAGCCG +TAGCGA +TAGCGT +TAGCGC +TAGCGG +TAGGAA +TAGGAT +TAGGAC +TAGGAG +TAGGTA +TAGGTT +TAGGTC +TAGGTG +TAGGCA +TAGGCT +TAGGCC +TAGGCG +TAGGGA +TAGGGT +TAGGGC +TAGGGG +TTAAAA +TTAAAT +TTAAAC +TTAAAG +TTAATA +TTAATT +TTAATC +TTAATG +TTAACA +TTAACT +TTAACC +TTAACG +TTAAGA +TTAAGT +TTAAGC +TTAAGG +TTATAA +TTATAT +TTATAC +TTATAG +TTATTA +TTATTT +TTATTC +TTATTG +TTATCA +TTATCT +TTATCC +TTATCG +TTATGA +TTATGT +TTATGC +TTATGG +TTACAA +TTACAT +TTACAC +TTACAG +TTACTA +TTACTT +TTACTC +TTACTG +TTACCA +TTACCT +TTACCC +TTACCG +TTACGA +TTACGT +TTACGC +TTACGG +TTAGAA +TTAGAT +TTAGAC +TTAGAG +TTAGTA +TTAGTT +TTAGTC +TTAGTG +TTAGCA +TTAGCT +TTAGCC +TTAGCG +TTAGGA +TTAGGT +TTAGGC +TTAGGG +TTTAAA +TTTAAT +TTTAAC +TTTAAG +TTTATA +TTTATT +TTTATC +TTTATG +TTTACA +TTTACT +TTTACC +TTTACG +TTTAGA +TTTAGT +TTTAGC +TTTAGG +TTTTAA +TTTTAT +TTTTAC +TTTTAG +TTTTTA +TTTTTT +TTTTTC +TTTTTG +TTTTCA +TTTTCT +TTTTCC +TTTTCG +TTTTGA +TTTTGT +TTTTGC +TTTTGG +TTTCAA +TTTCAT +TTTCAC +TTTCAG +TTTCTA +TTTCTT +TTTCTC +TTTCTG +TTTCCA +TTTCCT +TTTCCC +TTTCCG +TTTCGA +TTTCGT +TTTCGC +TTTCGG +TTTGAA +TTTGAT +TTTGAC +TTTGAG +TTTGTA +TTTGTT +TTTGTC +TTTGTG +TTTGCA +TTTGCT +TTTGCC +TTTGCG +TTTGGA +TTTGGT +TTTGGC +TTTGGG +TTCAAA +TTCAAT +TTCAAC +TTCAAG +TTCATA +TTCATT +TTCATC +TTCATG +TTCACA +TTCACT +TTCACC +TTCACG +TTCAGA +TTCAGT +TTCAGC +TTCAGG +TTCTAA +TTCTAT +TTCTAC +TTCTAG +TTCTTA +TTCTTT +TTCTTC +TTCTTG +TTCTCA +TTCTCT +TTCTCC +TTCTCG +TTCTGA +TTCTGT +TTCTGC +TTCTGG +TTCCAA +TTCCAT +TTCCAC +TTCCAG +TTCCTA +TTCCTT +TTCCTC +TTCCTG +TTCCCA +TTCCCT +TTCCCC +TTCCCG +TTCCGA +TTCCGT +TTCCGC +TTCCGG +TTCGAA +TTCGAT +TTCGAC +TTCGAG +TTCGTA +TTCGTT +TTCGTC +TTCGTG +TTCGCA +TTCGCT +TTCGCC +TTCGCG +TTCGGA +TTCGGT +TTCGGC +TTCGGG +TTGAAA +TTGAAT +TTGAAC +TTGAAG +TTGATA +TTGATT +TTGATC +TTGATG +TTGACA +TTGACT +TTGACC +TTGACG +TTGAGA +TTGAGT +TTGAGC +TTGAGG +TTGTAA +TTGTAT +TTGTAC +TTGTAG +TTGTTA +TTGTTT +TTGTTC +TTGTTG +TTGTCA +TTGTCT +TTGTCC +TTGTCG +TTGTGA +TTGTGT +TTGTGC +TTGTGG +TTGCAA +TTGCAT +TTGCAC +TTGCAG +TTGCTA +TTGCTT +TTGCTC +TTGCTG +TTGCCA +TTGCCT +TTGCCC +TTGCCG +TTGCGA +TTGCGT +TTGCGC +TTGCGG +TTGGAA +TTGGAT +TTGGAC +TTGGAG +TTGGTA +TTGGTT +TTGGTC +TTGGTG +TTGGCA +TTGGCT +TTGGCC +TTGGCG +TTGGGA +TTGGGT +TTGGGC +TTGGGG +TCAAAA +TCAAAT +TCAAAC +TCAAAG +TCAATA +TCAATT +TCAATC +TCAATG +TCAACA +TCAACT +TCAACC +TCAACG +TCAAGA +TCAAGT +TCAAGC +TCAAGG +TCATAA +TCATAT +TCATAC +TCATAG +TCATTA +TCATTT +TCATTC +TCATTG +TCATCA +TCATCT +TCATCC +TCATCG +TCATGA +TCATGT +TCATGC +TCATGG +TCACAA +TCACAT +TCACAC +TCACAG +TCACTA +TCACTT +TCACTC +TCACTG +TCACCA +TCACCT +TCACCC +TCACCG +TCACGA +TCACGT +TCACGC +TCACGG +TCAGAA +TCAGAT +TCAGAC +TCAGAG +TCAGTA +TCAGTT +TCAGTC +TCAGTG +TCAGCA +TCAGCT +TCAGCC +TCAGCG +TCAGGA +TCAGGT +TCAGGC +TCAGGG +TCTAAA +TCTAAT +TCTAAC +TCTAAG +TCTATA +TCTATT +TCTATC +TCTATG +TCTACA +TCTACT +TCTACC +TCTACG +TCTAGA +TCTAGT +TCTAGC +TCTAGG +TCTTAA +TCTTAT +TCTTAC +TCTTAG +TCTTTA +TCTTTT +TCTTTC +TCTTTG +TCTTCA +TCTTCT +TCTTCC +TCTTCG +TCTTGA +TCTTGT +TCTTGC +TCTTGG +TCTCAA +TCTCAT +TCTCAC +TCTCAG +TCTCTA +TCTCTT +TCTCTC +TCTCTG +TCTCCA +TCTCCT +TCTCCC +TCTCCG +TCTCGA +TCTCGT +TCTCGC +TCTCGG +TCTGAA +TCTGAT +TCTGAC +TCTGAG +TCTGTA +TCTGTT +TCTGTC +TCTGTG +TCTGCA +TCTGCT +TCTGCC +TCTGCG +TCTGGA +TCTGGT +TCTGGC +TCTGGG +TCCAAA +TCCAAT +TCCAAC +TCCAAG +TCCATA +TCCATT +TCCATC +TCCATG +TCCACA +TCCACT +TCCACC +TCCACG +TCCAGA +TCCAGT +TCCAGC +TCCAGG +TCCTAA +TCCTAT +TCCTAC +TCCTAG +TCCTTA +TCCTTT +TCCTTC +TCCTTG +TCCTCA +TCCTCT +TCCTCC +TCCTCG +TCCTGA +TCCTGT +TCCTGC +TCCTGG +TCCCAA +TCCCAT +TCCCAC +TCCCAG +TCCCTA +TCCCTT +TCCCTC +TCCCTG +TCCCCA +TCCCCT +TCCCCC +TCCCCG +TCCCGA +TCCCGT +TCCCGC +TCCCGG +TCCGAA +TCCGAT +TCCGAC +TCCGAG +TCCGTA +TCCGTT +TCCGTC +TCCGTG +TCCGCA +TCCGCT +TCCGCC +TCCGCG +TCCGGA +TCCGGT +TCCGGC +TCCGGG +TCGAAA +TCGAAT +TCGAAC +TCGAAG +TCGATA +TCGATT +TCGATC +TCGATG +TCGACA +TCGACT +TCGACC +TCGACG +TCGAGA +TCGAGT +TCGAGC +TCGAGG +TCGTAA +TCGTAT +TCGTAC +TCGTAG +TCGTTA +TCGTTT +TCGTTC +TCGTTG +TCGTCA +TCGTCT +TCGTCC +TCGTCG +TCGTGA +TCGTGT +TCGTGC +TCGTGG +TCGCAA +TCGCAT +TCGCAC +TCGCAG +TCGCTA +TCGCTT +TCGCTC +TCGCTG +TCGCCA +TCGCCT +TCGCCC +TCGCCG +TCGCGA +TCGCGT +TCGCGC +TCGCGG +TCGGAA +TCGGAT +TCGGAC +TCGGAG +TCGGTA +TCGGTT +TCGGTC +TCGGTG +TCGGCA +TCGGCT +TCGGCC +TCGGCG +TCGGGA +TCGGGT +TCGGGC +TCGGGG +TGAAAA +TGAAAT +TGAAAC +TGAAAG +TGAATA +TGAATT +TGAATC +TGAATG +TGAACA +TGAACT +TGAACC +TGAACG +TGAAGA +TGAAGT +TGAAGC +TGAAGG +TGATAA +TGATAT +TGATAC +TGATAG +TGATTA +TGATTT +TGATTC +TGATTG +TGATCA +TGATCT +TGATCC +TGATCG +TGATGA +TGATGT +TGATGC +TGATGG +TGACAA +TGACAT +TGACAC +TGACAG +TGACTA +TGACTT +TGACTC +TGACTG +TGACCA +TGACCT +TGACCC +TGACCG +TGACGA +TGACGT +TGACGC +TGACGG +TGAGAA +TGAGAT +TGAGAC +TGAGAG +TGAGTA +TGAGTT +TGAGTC +TGAGTG +TGAGCA +TGAGCT +TGAGCC +TGAGCG +TGAGGA +TGAGGT +TGAGGC +TGAGGG +TGTAAA +TGTAAT +TGTAAC +TGTAAG +TGTATA +TGTATT +TGTATC +TGTATG +TGTACA +TGTACT +TGTACC +TGTACG +TGTAGA +TGTAGT +TGTAGC +TGTAGG +TGTTAA +TGTTAT +TGTTAC +TGTTAG +TGTTTA +TGTTTT +TGTTTC +TGTTTG +TGTTCA +TGTTCT +TGTTCC +TGTTCG +TGTTGA +TGTTGT +TGTTGC +TGTTGG +TGTCAA +TGTCAT +TGTCAC +TGTCAG +TGTCTA +TGTCTT +TGTCTC +TGTCTG +TGTCCA +TGTCCT +TGTCCC +TGTCCG +TGTCGA +TGTCGT +TGTCGC +TGTCGG +TGTGAA +TGTGAT +TGTGAC +TGTGAG +TGTGTA +TGTGTT +TGTGTC +TGTGTG +TGTGCA +TGTGCT +TGTGCC +TGTGCG +TGTGGA +TGTGGT +TGTGGC +TGTGGG +TGCAAA +TGCAAT +TGCAAC +TGCAAG +TGCATA +TGCATT +TGCATC +TGCATG +TGCACA +TGCACT +TGCACC +TGCACG +TGCAGA +TGCAGT +TGCAGC +TGCAGG +TGCTAA +TGCTAT +TGCTAC +TGCTAG +TGCTTA +TGCTTT +TGCTTC +TGCTTG +TGCTCA +TGCTCT +TGCTCC +TGCTCG +TGCTGA +TGCTGT +TGCTGC +TGCTGG +TGCCAA +TGCCAT +TGCCAC +TGCCAG +TGCCTA +TGCCTT +TGCCTC +TGCCTG +TGCCCA +TGCCCT +TGCCCC +TGCCCG +TGCCGA +TGCCGT +TGCCGC +TGCCGG +TGCGAA +TGCGAT +TGCGAC +TGCGAG +TGCGTA +TGCGTT +TGCGTC +TGCGTG +TGCGCA +TGCGCT +TGCGCC +TGCGCG +TGCGGA +TGCGGT +TGCGGC +TGCGGG +TGGAAA +TGGAAT +TGGAAC +TGGAAG +TGGATA +TGGATT +TGGATC +TGGATG +TGGACA +TGGACT +TGGACC +TGGACG +TGGAGA +TGGAGT +TGGAGC +TGGAGG +TGGTAA +TGGTAT +TGGTAC +TGGTAG +TGGTTA +TGGTTT +TGGTTC +TGGTTG +TGGTCA +TGGTCT +TGGTCC +TGGTCG +TGGTGA +TGGTGT +TGGTGC +TGGTGG +TGGCAA +TGGCAT +TGGCAC +TGGCAG +TGGCTA +TGGCTT +TGGCTC +TGGCTG +TGGCCA +TGGCCT +TGGCCC +TGGCCG +TGGCGA +TGGCGT +TGGCGC +TGGCGG +TGGGAA +TGGGAT +TGGGAC +TGGGAG +TGGGTA +TGGGTT +TGGGTC +TGGGTG +TGGGCA +TGGGCT +TGGGCC +TGGGCG +TGGGGA +TGGGGT +TGGGGC +TGGGGG +CAAAAA +CAAAAT +CAAAAC +CAAAAG +CAAATA +CAAATT +CAAATC +CAAATG +CAAACA +CAAACT +CAAACC +CAAACG +CAAAGA +CAAAGT +CAAAGC +CAAAGG +CAATAA +CAATAT +CAATAC +CAATAG +CAATTA +CAATTT +CAATTC +CAATTG +CAATCA +CAATCT +CAATCC +CAATCG +CAATGA +CAATGT +CAATGC +CAATGG +CAACAA +CAACAT +CAACAC +CAACAG +CAACTA +CAACTT +CAACTC +CAACTG +CAACCA +CAACCT +CAACCC +CAACCG +CAACGA +CAACGT +CAACGC +CAACGG +CAAGAA +CAAGAT +CAAGAC +CAAGAG +CAAGTA +CAAGTT +CAAGTC +CAAGTG +CAAGCA +CAAGCT +CAAGCC +CAAGCG +CAAGGA +CAAGGT +CAAGGC +CAAGGG +CATAAA +CATAAT +CATAAC +CATAAG +CATATA +CATATT +CATATC +CATATG +CATACA +CATACT +CATACC +CATACG +CATAGA +CATAGT +CATAGC +CATAGG +CATTAA +CATTAT +CATTAC +CATTAG +CATTTA +CATTTT +CATTTC +CATTTG +CATTCA +CATTCT +CATTCC +CATTCG +CATTGA +CATTGT +CATTGC +CATTGG +CATCAA +CATCAT +CATCAC +CATCAG +CATCTA +CATCTT +CATCTC +CATCTG +CATCCA +CATCCT +CATCCC +CATCCG +CATCGA +CATCGT +CATCGC +CATCGG +CATGAA +CATGAT +CATGAC +CATGAG +CATGTA +CATGTT +CATGTC +CATGTG +CATGCA +CATGCT +CATGCC +CATGCG +CATGGA +CATGGT +CATGGC +CATGGG +CACAAA +CACAAT +CACAAC +CACAAG +CACATA +CACATT +CACATC +CACATG +CACACA +CACACT +CACACC +CACACG +CACAGA +CACAGT +CACAGC +CACAGG +CACTAA +CACTAT +CACTAC +CACTAG +CACTTA +CACTTT +CACTTC +CACTTG +CACTCA +CACTCT +CACTCC +CACTCG +CACTGA +CACTGT +CACTGC +CACTGG +CACCAA +CACCAT +CACCAC +CACCAG +CACCTA +CACCTT +CACCTC +CACCTG +CACCCA +CACCCT +CACCCC +CACCCG +CACCGA +CACCGT +CACCGC +CACCGG +CACGAA +CACGAT +CACGAC +CACGAG +CACGTA +CACGTT +CACGTC +CACGTG +CACGCA +CACGCT +CACGCC +CACGCG +CACGGA +CACGGT +CACGGC +CACGGG +CAGAAA +CAGAAT +CAGAAC +CAGAAG +CAGATA +CAGATT +CAGATC +CAGATG +CAGACA +CAGACT +CAGACC +CAGACG +CAGAGA +CAGAGT +CAGAGC +CAGAGG +CAGTAA +CAGTAT +CAGTAC +CAGTAG +CAGTTA +CAGTTT +CAGTTC +CAGTTG +CAGTCA +CAGTCT +CAGTCC +CAGTCG +CAGTGA +CAGTGT +CAGTGC +CAGTGG +CAGCAA +CAGCAT +CAGCAC +CAGCAG +CAGCTA +CAGCTT +CAGCTC +CAGCTG +CAGCCA +CAGCCT +CAGCCC +CAGCCG +CAGCGA +CAGCGT +CAGCGC +CAGCGG +CAGGAA +CAGGAT +CAGGAC +CAGGAG +CAGGTA +CAGGTT +CAGGTC +CAGGTG +CAGGCA +CAGGCT +CAGGCC +CAGGCG +CAGGGA +CAGGGT +CAGGGC +CAGGGG +CTAAAA +CTAAAT +CTAAAC +CTAAAG +CTAATA +CTAATT +CTAATC +CTAATG +CTAACA +CTAACT +CTAACC +CTAACG +CTAAGA +CTAAGT +CTAAGC +CTAAGG +CTATAA +CTATAT +CTATAC +CTATAG +CTATTA +CTATTT +CTATTC +CTATTG +CTATCA +CTATCT +CTATCC +CTATCG +CTATGA +CTATGT +CTATGC +CTATGG +CTACAA +CTACAT +CTACAC +CTACAG +CTACTA +CTACTT +CTACTC +CTACTG +CTACCA +CTACCT +CTACCC +CTACCG +CTACGA +CTACGT +CTACGC +CTACGG +CTAGAA +CTAGAT +CTAGAC +CTAGAG +CTAGTA +CTAGTT +CTAGTC +CTAGTG +CTAGCA +CTAGCT +CTAGCC +CTAGCG +CTAGGA +CTAGGT +CTAGGC +CTAGGG +CTTAAA +CTTAAT +CTTAAC +CTTAAG +CTTATA +CTTATT +CTTATC +CTTATG +CTTACA +CTTACT +CTTACC +CTTACG +CTTAGA +CTTAGT +CTTAGC +CTTAGG +CTTTAA +CTTTAT +CTTTAC +CTTTAG +CTTTTA +CTTTTT +CTTTTC +CTTTTG +CTTTCA +CTTTCT +CTTTCC +CTTTCG +CTTTGA +CTTTGT +CTTTGC +CTTTGG +CTTCAA +CTTCAT +CTTCAC +CTTCAG +CTTCTA +CTTCTT +CTTCTC +CTTCTG +CTTCCA +CTTCCT +CTTCCC +CTTCCG +CTTCGA +CTTCGT +CTTCGC +CTTCGG +CTTGAA +CTTGAT +CTTGAC +CTTGAG +CTTGTA +CTTGTT +CTTGTC +CTTGTG +CTTGCA +CTTGCT +CTTGCC +CTTGCG +CTTGGA +CTTGGT +CTTGGC +CTTGGG +CTCAAA +CTCAAT +CTCAAC +CTCAAG +CTCATA +CTCATT +CTCATC +CTCATG +CTCACA +CTCACT +CTCACC +CTCACG +CTCAGA +CTCAGT +CTCAGC +CTCAGG +CTCTAA +CTCTAT +CTCTAC +CTCTAG +CTCTTA +CTCTTT +CTCTTC +CTCTTG +CTCTCA +CTCTCT +CTCTCC +CTCTCG +CTCTGA +CTCTGT +CTCTGC +CTCTGG +CTCCAA +CTCCAT +CTCCAC +CTCCAG +CTCCTA +CTCCTT +CTCCTC +CTCCTG +CTCCCA +CTCCCT +CTCCCC +CTCCCG +CTCCGA +CTCCGT +CTCCGC +CTCCGG +CTCGAA +CTCGAT +CTCGAC +CTCGAG +CTCGTA +CTCGTT +CTCGTC +CTCGTG +CTCGCA +CTCGCT +CTCGCC +CTCGCG +CTCGGA +CTCGGT +CTCGGC +CTCGGG +CTGAAA +CTGAAT +CTGAAC +CTGAAG +CTGATA +CTGATT +CTGATC +CTGATG +CTGACA +CTGACT +CTGACC +CTGACG +CTGAGA +CTGAGT +CTGAGC +CTGAGG +CTGTAA +CTGTAT +CTGTAC +CTGTAG +CTGTTA +CTGTTT +CTGTTC +CTGTTG +CTGTCA +CTGTCT +CTGTCC +CTGTCG +CTGTGA +CTGTGT +CTGTGC +CTGTGG +CTGCAA +CTGCAT +CTGCAC +CTGCAG +CTGCTA +CTGCTT +CTGCTC +CTGCTG +CTGCCA +CTGCCT +CTGCCC +CTGCCG +CTGCGA +CTGCGT +CTGCGC +CTGCGG +CTGGAA +CTGGAT +CTGGAC +CTGGAG +CTGGTA +CTGGTT +CTGGTC +CTGGTG +CTGGCA +CTGGCT +CTGGCC +CTGGCG +CTGGGA +CTGGGT +CTGGGC +CTGGGG +CCAAAA +CCAAAT +CCAAAC +CCAAAG +CCAATA +CCAATT +CCAATC +CCAATG +CCAACA +CCAACT +CCAACC +CCAACG +CCAAGA +CCAAGT +CCAAGC +CCAAGG +CCATAA +CCATAT +CCATAC +CCATAG +CCATTA +CCATTT +CCATTC +CCATTG +CCATCA +CCATCT +CCATCC +CCATCG +CCATGA +CCATGT +CCATGC +CCATGG +CCACAA +CCACAT +CCACAC +CCACAG +CCACTA +CCACTT +CCACTC +CCACTG +CCACCA +CCACCT +CCACCC +CCACCG +CCACGA +CCACGT +CCACGC +CCACGG +CCAGAA +CCAGAT +CCAGAC +CCAGAG +CCAGTA +CCAGTT +CCAGTC +CCAGTG +CCAGCA +CCAGCT +CCAGCC +CCAGCG +CCAGGA +CCAGGT +CCAGGC +CCAGGG +CCTAAA +CCTAAT +CCTAAC +CCTAAG +CCTATA +CCTATT +CCTATC +CCTATG +CCTACA +CCTACT +CCTACC +CCTACG +CCTAGA +CCTAGT +CCTAGC +CCTAGG +CCTTAA +CCTTAT +CCTTAC +CCTTAG +CCTTTA +CCTTTT +CCTTTC +CCTTTG +CCTTCA +CCTTCT +CCTTCC +CCTTCG +CCTTGA +CCTTGT +CCTTGC +CCTTGG +CCTCAA +CCTCAT +CCTCAC +CCTCAG +CCTCTA +CCTCTT +CCTCTC +CCTCTG +CCTCCA +CCTCCT +CCTCCC +CCTCCG +CCTCGA +CCTCGT +CCTCGC +CCTCGG +CCTGAA +CCTGAT +CCTGAC +CCTGAG +CCTGTA +CCTGTT +CCTGTC +CCTGTG +CCTGCA +CCTGCT +CCTGCC +CCTGCG +CCTGGA +CCTGGT +CCTGGC +CCTGGG +CCCAAA +CCCAAT +CCCAAC +CCCAAG +CCCATA +CCCATT +CCCATC +CCCATG +CCCACA +CCCACT +CCCACC +CCCACG +CCCAGA +CCCAGT +CCCAGC +CCCAGG +CCCTAA +CCCTAT +CCCTAC +CCCTAG +CCCTTA +CCCTTT +CCCTTC +CCCTTG +CCCTCA +CCCTCT +CCCTCC +CCCTCG +CCCTGA +CCCTGT +CCCTGC +CCCTGG +CCCCAA +CCCCAT +CCCCAC +CCCCAG +CCCCTA +CCCCTT +CCCCTC +CCCCTG +CCCCCA +CCCCCT +CCCCCC +CCCCCG +CCCCGA +CCCCGT +CCCCGC +CCCCGG +CCCGAA +CCCGAT +CCCGAC +CCCGAG +CCCGTA +CCCGTT +CCCGTC +CCCGTG +CCCGCA +CCCGCT +CCCGCC +CCCGCG +CCCGGA +CCCGGT +CCCGGC +CCCGGG +CCGAAA +CCGAAT +CCGAAC +CCGAAG +CCGATA +CCGATT +CCGATC +CCGATG +CCGACA +CCGACT +CCGACC +CCGACG +CCGAGA +CCGAGT +CCGAGC +CCGAGG +CCGTAA +CCGTAT +CCGTAC +CCGTAG +CCGTTA +CCGTTT +CCGTTC +CCGTTG +CCGTCA +CCGTCT +CCGTCC +CCGTCG +CCGTGA +CCGTGT +CCGTGC +CCGTGG +CCGCAA +CCGCAT +CCGCAC +CCGCAG +CCGCTA +CCGCTT +CCGCTC +CCGCTG +CCGCCA +CCGCCT +CCGCCC +CCGCCG +CCGCGA +CCGCGT +CCGCGC +CCGCGG +CCGGAA +CCGGAT +CCGGAC +CCGGAG +CCGGTA +CCGGTT +CCGGTC +CCGGTG +CCGGCA +CCGGCT +CCGGCC +CCGGCG +CCGGGA +CCGGGT +CCGGGC +CCGGGG +CGAAAA +CGAAAT +CGAAAC +CGAAAG +CGAATA +CGAATT +CGAATC +CGAATG +CGAACA +CGAACT +CGAACC +CGAACG +CGAAGA +CGAAGT +CGAAGC +CGAAGG +CGATAA +CGATAT +CGATAC +CGATAG +CGATTA +CGATTT +CGATTC +CGATTG +CGATCA +CGATCT +CGATCC +CGATCG +CGATGA +CGATGT +CGATGC +CGATGG +CGACAA +CGACAT +CGACAC +CGACAG +CGACTA +CGACTT +CGACTC +CGACTG +CGACCA +CGACCT +CGACCC +CGACCG +CGACGA +CGACGT +CGACGC +CGACGG +CGAGAA +CGAGAT +CGAGAC +CGAGAG +CGAGTA +CGAGTT +CGAGTC +CGAGTG +CGAGCA +CGAGCT +CGAGCC +CGAGCG +CGAGGA +CGAGGT +CGAGGC +CGAGGG +CGTAAA +CGTAAT +CGTAAC +CGTAAG +CGTATA +CGTATT +CGTATC +CGTATG +CGTACA +CGTACT +CGTACC +CGTACG +CGTAGA +CGTAGT +CGTAGC +CGTAGG +CGTTAA +CGTTAT +CGTTAC +CGTTAG +CGTTTA +CGTTTT +CGTTTC +CGTTTG +CGTTCA +CGTTCT +CGTTCC +CGTTCG +CGTTGA +CGTTGT +CGTTGC +CGTTGG +CGTCAA +CGTCAT +CGTCAC +CGTCAG +CGTCTA +CGTCTT +CGTCTC +CGTCTG +CGTCCA +CGTCCT +CGTCCC +CGTCCG +CGTCGA +CGTCGT +CGTCGC +CGTCGG +CGTGAA +CGTGAT +CGTGAC +CGTGAG +CGTGTA +CGTGTT +CGTGTC +CGTGTG +CGTGCA +CGTGCT +CGTGCC +CGTGCG +CGTGGA +CGTGGT +CGTGGC +CGTGGG +CGCAAA +CGCAAT +CGCAAC +CGCAAG +CGCATA +CGCATT +CGCATC +CGCATG +CGCACA +CGCACT +CGCACC +CGCACG +CGCAGA +CGCAGT +CGCAGC +CGCAGG +CGCTAA +CGCTAT +CGCTAC +CGCTAG +CGCTTA +CGCTTT +CGCTTC +CGCTTG +CGCTCA +CGCTCT +CGCTCC +CGCTCG +CGCTGA +CGCTGT +CGCTGC +CGCTGG +CGCCAA +CGCCAT +CGCCAC +CGCCAG +CGCCTA +CGCCTT +CGCCTC +CGCCTG +CGCCCA +CGCCCT +CGCCCC +CGCCCG +CGCCGA +CGCCGT +CGCCGC +CGCCGG +CGCGAA +CGCGAT +CGCGAC +CGCGAG +CGCGTA +CGCGTT +CGCGTC +CGCGTG +CGCGCA +CGCGCT +CGCGCC +CGCGCG +CGCGGA +CGCGGT +CGCGGC +CGCGGG +CGGAAA +CGGAAT +CGGAAC +CGGAAG +CGGATA +CGGATT +CGGATC +CGGATG +CGGACA +CGGACT +CGGACC +CGGACG +CGGAGA +CGGAGT +CGGAGC +CGGAGG +CGGTAA +CGGTAT +CGGTAC +CGGTAG +CGGTTA +CGGTTT +CGGTTC +CGGTTG +CGGTCA +CGGTCT +CGGTCC +CGGTCG +CGGTGA +CGGTGT +CGGTGC +CGGTGG +CGGCAA +CGGCAT +CGGCAC +CGGCAG +CGGCTA +CGGCTT +CGGCTC +CGGCTG +CGGCCA +CGGCCT +CGGCCC +CGGCCG +CGGCGA +CGGCGT +CGGCGC +CGGCGG +CGGGAA +CGGGAT +CGGGAC +CGGGAG +CGGGTA +CGGGTT +CGGGTC +CGGGTG +CGGGCA +CGGGCT +CGGGCC +CGGGCG +CGGGGA +CGGGGT +CGGGGC +CGGGGG +GAAAAA +GAAAAT +GAAAAC +GAAAAG +GAAATA +GAAATT +GAAATC +GAAATG +GAAACA +GAAACT +GAAACC +GAAACG +GAAAGA +GAAAGT +GAAAGC +GAAAGG +GAATAA +GAATAT +GAATAC +GAATAG +GAATTA +GAATTT +GAATTC +GAATTG +GAATCA +GAATCT +GAATCC +GAATCG +GAATGA +GAATGT +GAATGC +GAATGG +GAACAA +GAACAT +GAACAC +GAACAG +GAACTA +GAACTT +GAACTC +GAACTG +GAACCA +GAACCT +GAACCC +GAACCG +GAACGA +GAACGT +GAACGC +GAACGG +GAAGAA +GAAGAT +GAAGAC +GAAGAG +GAAGTA +GAAGTT +GAAGTC +GAAGTG +GAAGCA +GAAGCT +GAAGCC +GAAGCG +GAAGGA +GAAGGT +GAAGGC +GAAGGG +GATAAA +GATAAT +GATAAC +GATAAG +GATATA +GATATT +GATATC +GATATG +GATACA +GATACT +GATACC +GATACG +GATAGA +GATAGT +GATAGC +GATAGG +GATTAA +GATTAT +GATTAC +GATTAG +GATTTA +GATTTT +GATTTC +GATTTG +GATTCA +GATTCT +GATTCC +GATTCG +GATTGA +GATTGT +GATTGC +GATTGG +GATCAA +GATCAT +GATCAC +GATCAG +GATCTA +GATCTT +GATCTC +GATCTG +GATCCA +GATCCT +GATCCC +GATCCG +GATCGA +GATCGT +GATCGC +GATCGG +GATGAA +GATGAT +GATGAC +GATGAG +GATGTA +GATGTT +GATGTC +GATGTG +GATGCA +GATGCT +GATGCC +GATGCG +GATGGA +GATGGT +GATGGC +GATGGG +GACAAA +GACAAT +GACAAC +GACAAG +GACATA +GACATT +GACATC +GACATG +GACACA +GACACT +GACACC +GACACG +GACAGA +GACAGT +GACAGC +GACAGG +GACTAA +GACTAT +GACTAC +GACTAG +GACTTA +GACTTT +GACTTC +GACTTG +GACTCA +GACTCT +GACTCC +GACTCG +GACTGA +GACTGT +GACTGC +GACTGG +GACCAA +GACCAT +GACCAC +GACCAG +GACCTA +GACCTT +GACCTC +GACCTG +GACCCA +GACCCT +GACCCC +GACCCG +GACCGA +GACCGT +GACCGC +GACCGG +GACGAA +GACGAT +GACGAC +GACGAG +GACGTA +GACGTT +GACGTC +GACGTG +GACGCA +GACGCT +GACGCC +GACGCG +GACGGA +GACGGT +GACGGC +GACGGG +GAGAAA +GAGAAT +GAGAAC +GAGAAG +GAGATA +GAGATT +GAGATC +GAGATG +GAGACA +GAGACT +GAGACC +GAGACG +GAGAGA +GAGAGT +GAGAGC +GAGAGG +GAGTAA +GAGTAT +GAGTAC +GAGTAG +GAGTTA +GAGTTT +GAGTTC +GAGTTG +GAGTCA +GAGTCT +GAGTCC +GAGTCG +GAGTGA +GAGTGT +GAGTGC +GAGTGG +GAGCAA +GAGCAT +GAGCAC +GAGCAG +GAGCTA +GAGCTT +GAGCTC +GAGCTG +GAGCCA +GAGCCT +GAGCCC +GAGCCG +GAGCGA +GAGCGT +GAGCGC +GAGCGG +GAGGAA +GAGGAT +GAGGAC +GAGGAG +GAGGTA +GAGGTT +GAGGTC +GAGGTG +GAGGCA +GAGGCT +GAGGCC +GAGGCG +GAGGGA +GAGGGT +GAGGGC +GAGGGG +GTAAAA +GTAAAT +GTAAAC +GTAAAG +GTAATA +GTAATT +GTAATC +GTAATG +GTAACA +GTAACT +GTAACC +GTAACG +GTAAGA +GTAAGT +GTAAGC +GTAAGG +GTATAA +GTATAT +GTATAC +GTATAG +GTATTA +GTATTT +GTATTC +GTATTG +GTATCA +GTATCT +GTATCC +GTATCG +GTATGA +GTATGT +GTATGC +GTATGG +GTACAA +GTACAT +GTACAC +GTACAG +GTACTA +GTACTT +GTACTC +GTACTG +GTACCA +GTACCT +GTACCC +GTACCG +GTACGA +GTACGT +GTACGC +GTACGG +GTAGAA +GTAGAT +GTAGAC +GTAGAG +GTAGTA +GTAGTT +GTAGTC +GTAGTG +GTAGCA +GTAGCT +GTAGCC +GTAGCG +GTAGGA +GTAGGT +GTAGGC +GTAGGG +GTTAAA +GTTAAT +GTTAAC +GTTAAG +GTTATA +GTTATT +GTTATC +GTTATG +GTTACA +GTTACT +GTTACC +GTTACG +GTTAGA +GTTAGT +GTTAGC +GTTAGG +GTTTAA +GTTTAT +GTTTAC +GTTTAG +GTTTTA +GTTTTT +GTTTTC +GTTTTG +GTTTCA +GTTTCT +GTTTCC +GTTTCG +GTTTGA +GTTTGT +GTTTGC +GTTTGG +GTTCAA +GTTCAT +GTTCAC +GTTCAG +GTTCTA +GTTCTT +GTTCTC +GTTCTG +GTTCCA +GTTCCT +GTTCCC +GTTCCG +GTTCGA +GTTCGT +GTTCGC +GTTCGG +GTTGAA +GTTGAT +GTTGAC +GTTGAG +GTTGTA +GTTGTT +GTTGTC +GTTGTG +GTTGCA +GTTGCT +GTTGCC +GTTGCG +GTTGGA +GTTGGT +GTTGGC +GTTGGG +GTCAAA +GTCAAT +GTCAAC +GTCAAG +GTCATA +GTCATT +GTCATC +GTCATG +GTCACA +GTCACT +GTCACC +GTCACG +GTCAGA +GTCAGT +GTCAGC +GTCAGG +GTCTAA +GTCTAT +GTCTAC +GTCTAG +GTCTTA +GTCTTT +GTCTTC +GTCTTG +GTCTCA +GTCTCT +GTCTCC +GTCTCG +GTCTGA +GTCTGT +GTCTGC +GTCTGG +GTCCAA +GTCCAT +GTCCAC +GTCCAG +GTCCTA +GTCCTT +GTCCTC +GTCCTG +GTCCCA +GTCCCT +GTCCCC +GTCCCG +GTCCGA +GTCCGT +GTCCGC +GTCCGG +GTCGAA +GTCGAT +GTCGAC +GTCGAG +GTCGTA +GTCGTT +GTCGTC +GTCGTG +GTCGCA +GTCGCT +GTCGCC +GTCGCG +GTCGGA +GTCGGT +GTCGGC +GTCGGG +GTGAAA +GTGAAT +GTGAAC +GTGAAG +GTGATA +GTGATT +GTGATC +GTGATG +GTGACA +GTGACT +GTGACC +GTGACG +GTGAGA +GTGAGT +GTGAGC +GTGAGG +GTGTAA +GTGTAT +GTGTAC +GTGTAG +GTGTTA +GTGTTT +GTGTTC +GTGTTG +GTGTCA +GTGTCT +GTGTCC +GTGTCG +GTGTGA +GTGTGT +GTGTGC +GTGTGG +GTGCAA +GTGCAT +GTGCAC +GTGCAG +GTGCTA +GTGCTT +GTGCTC +GTGCTG +GTGCCA +GTGCCT +GTGCCC +GTGCCG +GTGCGA +GTGCGT +GTGCGC +GTGCGG +GTGGAA +GTGGAT +GTGGAC +GTGGAG +GTGGTA +GTGGTT +GTGGTC +GTGGTG +GTGGCA +GTGGCT +GTGGCC +GTGGCG +GTGGGA +GTGGGT +GTGGGC +GTGGGG +GCAAAA +GCAAAT +GCAAAC +GCAAAG +GCAATA +GCAATT +GCAATC +GCAATG +GCAACA +GCAACT +GCAACC +GCAACG +GCAAGA +GCAAGT +GCAAGC +GCAAGG +GCATAA +GCATAT +GCATAC +GCATAG +GCATTA +GCATTT +GCATTC +GCATTG +GCATCA +GCATCT +GCATCC +GCATCG +GCATGA +GCATGT +GCATGC +GCATGG +GCACAA +GCACAT +GCACAC +GCACAG +GCACTA +GCACTT +GCACTC +GCACTG +GCACCA +GCACCT +GCACCC +GCACCG +GCACGA +GCACGT +GCACGC +GCACGG +GCAGAA +GCAGAT +GCAGAC +GCAGAG +GCAGTA +GCAGTT +GCAGTC +GCAGTG +GCAGCA +GCAGCT +GCAGCC +GCAGCG +GCAGGA +GCAGGT +GCAGGC +GCAGGG +GCTAAA +GCTAAT +GCTAAC +GCTAAG +GCTATA +GCTATT +GCTATC +GCTATG +GCTACA +GCTACT +GCTACC +GCTACG +GCTAGA +GCTAGT +GCTAGC +GCTAGG +GCTTAA +GCTTAT +GCTTAC +GCTTAG +GCTTTA +GCTTTT +GCTTTC +GCTTTG +GCTTCA +GCTTCT +GCTTCC +GCTTCG +GCTTGA +GCTTGT +GCTTGC +GCTTGG +GCTCAA +GCTCAT +GCTCAC +GCTCAG +GCTCTA +GCTCTT +GCTCTC +GCTCTG +GCTCCA +GCTCCT +GCTCCC +GCTCCG +GCTCGA +GCTCGT +GCTCGC +GCTCGG +GCTGAA +GCTGAT +GCTGAC +GCTGAG +GCTGTA +GCTGTT +GCTGTC +GCTGTG +GCTGCA +GCTGCT +GCTGCC +GCTGCG +GCTGGA +GCTGGT +GCTGGC +GCTGGG +GCCAAA +GCCAAT +GCCAAC +GCCAAG +GCCATA +GCCATT +GCCATC +GCCATG +GCCACA +GCCACT +GCCACC +GCCACG +GCCAGA +GCCAGT +GCCAGC +GCCAGG +GCCTAA +GCCTAT +GCCTAC +GCCTAG +GCCTTA +GCCTTT +GCCTTC +GCCTTG +GCCTCA +GCCTCT +GCCTCC +GCCTCG +GCCTGA +GCCTGT +GCCTGC +GCCTGG +GCCCAA +GCCCAT +GCCCAC +GCCCAG +GCCCTA +GCCCTT +GCCCTC +GCCCTG +GCCCCA +GCCCCT +GCCCCC +GCCCCG +GCCCGA +GCCCGT +GCCCGC +GCCCGG +GCCGAA +GCCGAT +GCCGAC +GCCGAG +GCCGTA +GCCGTT +GCCGTC +GCCGTG +GCCGCA +GCCGCT +GCCGCC +GCCGCG +GCCGGA +GCCGGT +GCCGGC +GCCGGG +GCGAAA +GCGAAT +GCGAAC +GCGAAG +GCGATA +GCGATT +GCGATC +GCGATG +GCGACA +GCGACT +GCGACC +GCGACG +GCGAGA +GCGAGT +GCGAGC +GCGAGG +GCGTAA +GCGTAT +GCGTAC +GCGTAG +GCGTTA +GCGTTT +GCGTTC +GCGTTG +GCGTCA +GCGTCT +GCGTCC +GCGTCG +GCGTGA +GCGTGT +GCGTGC +GCGTGG +GCGCAA +GCGCAT +GCGCAC +GCGCAG +GCGCTA +GCGCTT +GCGCTC +GCGCTG +GCGCCA +GCGCCT +GCGCCC +GCGCCG +GCGCGA +GCGCGT +GCGCGC +GCGCGG +GCGGAA +GCGGAT +GCGGAC +GCGGAG +GCGGTA +GCGGTT +GCGGTC +GCGGTG +GCGGCA +GCGGCT +GCGGCC +GCGGCG +GCGGGA +GCGGGT +GCGGGC +GCGGGG +GGAAAA +GGAAAT +GGAAAC +GGAAAG +GGAATA +GGAATT +GGAATC +GGAATG +GGAACA +GGAACT +GGAACC +GGAACG +GGAAGA +GGAAGT +GGAAGC +GGAAGG +GGATAA +GGATAT +GGATAC +GGATAG +GGATTA +GGATTT +GGATTC +GGATTG +GGATCA +GGATCT +GGATCC +GGATCG +GGATGA +GGATGT +GGATGC +GGATGG +GGACAA +GGACAT +GGACAC +GGACAG +GGACTA +GGACTT +GGACTC +GGACTG +GGACCA +GGACCT +GGACCC +GGACCG +GGACGA +GGACGT +GGACGC +GGACGG +GGAGAA +GGAGAT +GGAGAC +GGAGAG +GGAGTA +GGAGTT +GGAGTC +GGAGTG +GGAGCA +GGAGCT +GGAGCC +GGAGCG +GGAGGA +GGAGGT +GGAGGC +GGAGGG +GGTAAA +GGTAAT +GGTAAC +GGTAAG +GGTATA +GGTATT +GGTATC +GGTATG +GGTACA +GGTACT +GGTACC +GGTACG +GGTAGA +GGTAGT +GGTAGC +GGTAGG +GGTTAA +GGTTAT +GGTTAC +GGTTAG +GGTTTA +GGTTTT +GGTTTC +GGTTTG +GGTTCA +GGTTCT +GGTTCC +GGTTCG +GGTTGA +GGTTGT +GGTTGC +GGTTGG +GGTCAA +GGTCAT +GGTCAC +GGTCAG +GGTCTA +GGTCTT +GGTCTC +GGTCTG +GGTCCA +GGTCCT +GGTCCC +GGTCCG +GGTCGA +GGTCGT +GGTCGC +GGTCGG +GGTGAA +GGTGAT +GGTGAC +GGTGAG +GGTGTA +GGTGTT +GGTGTC +GGTGTG +GGTGCA +GGTGCT +GGTGCC +GGTGCG +GGTGGA +GGTGGT +GGTGGC +GGTGGG +GGCAAA +GGCAAT +GGCAAC +GGCAAG +GGCATA +GGCATT +GGCATC +GGCATG +GGCACA +GGCACT +GGCACC +GGCACG +GGCAGA +GGCAGT +GGCAGC +GGCAGG +GGCTAA +GGCTAT +GGCTAC +GGCTAG +GGCTTA +GGCTTT +GGCTTC +GGCTTG +GGCTCA +GGCTCT +GGCTCC +GGCTCG +GGCTGA +GGCTGT +GGCTGC +GGCTGG +GGCCAA +GGCCAT +GGCCAC +GGCCAG +GGCCTA +GGCCTT +GGCCTC +GGCCTG +GGCCCA +GGCCCT +GGCCCC +GGCCCG +GGCCGA +GGCCGT +GGCCGC +GGCCGG +GGCGAA +GGCGAT +GGCGAC +GGCGAG +GGCGTA +GGCGTT +GGCGTC +GGCGTG +GGCGCA +GGCGCT +GGCGCC +GGCGCG +GGCGGA +GGCGGT +GGCGGC +GGCGGG +GGGAAA +GGGAAT +GGGAAC +GGGAAG +GGGATA +GGGATT +GGGATC +GGGATG +GGGACA +GGGACT +GGGACC +GGGACG +GGGAGA +GGGAGT +GGGAGC +GGGAGG +GGGTAA +GGGTAT +GGGTAC +GGGTAG +GGGTTA +GGGTTT +GGGTTC +GGGTTG +GGGTCA +GGGTCT +GGGTCC +GGGTCG +GGGTGA +GGGTGT +GGGTGC +GGGTGG +GGGCAA +GGGCAT +GGGCAC +GGGCAG +GGGCTA +GGGCTT +GGGCTC +GGGCTG +GGGCCA +GGGCCT +GGGCCC +GGGCCG +GGGCGA +GGGCGT +GGGCGC +GGGCGG +GGGGAA +GGGGAT +GGGGAC +GGGGAG +GGGGTA +GGGGTT +GGGGTC +GGGGTG +GGGGCA +GGGGCT +GGGGCC +GGGGCG +GGGGGA +GGGGGT +GGGGGC +GGGGGG diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6000/config.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..9b14e61d327bf586fcff6e817cf9d365bae18ad2 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6000/config.json @@ -0,0 +1,26 @@ +{ + "_name_or_path": "fabihamakhdoomi/TinyDNABERT", + "architectures": [ + "BertForSequenceClassification" + ], + "attention_probs_dropout_prob": 0.1, + "classifier_dropout": null, + "hidden_act": "gelu", + "hidden_dropout_prob": 0.1, + "hidden_size": 128, + "initializer_range": 0.02, + "intermediate_size": 512, + "layer_norm_eps": 1e-12, + "max_position_embeddings": 512, + "model_type": "bert", + "num_attention_heads": 2, + "num_hidden_layers": 2, + "pad_token_id": 0, + "position_embedding_type": "absolute", + "problem_type": "single_label_classification", + "torch_dtype": "float32", + "transformers_version": "4.44.0", + "type_vocab_size": 2, + "use_cache": true, + "vocab_size": 4101 +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6000/model.safetensors b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6000/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..107fa38a34cb48e93ed9631cf6493bdfe8b9312a --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6000/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4fb33efe9a5082c6e5703149593c85ca75536d751002eb55b8ab1e239293d75e +size 4021688 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6000/optimizer.pt b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6000/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..edffeab6cf2e8f1299dec39e7f99024f3505f6e0 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6000/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8e09dcace3e2d96ff4d8eab18d1c362a1bf1a88bfea8076eb616df3750a0869c +size 8058700 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6000/rng_state.pth b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6000/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..c78d09ac076b437860c4e8c07df963c522b10249 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6000/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d44e448e2a0dd138d067ef3ddf4b7abb608f898bed2f0d0880205ed4bbbe2136 +size 14244 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6000/scheduler.pt b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6000/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..f3ebc22eebe7fac9f20742118527e91760861e6d --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6000/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7e153e480933b4bdcc7d98226c2fdfde72db931809685729c8748251ae3ecae9 +size 1064 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6000/special_tokens_map.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6000/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..a8b3208c2884c4efb86e49300fdd3dc877220cdf --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6000/special_tokens_map.json @@ -0,0 +1,7 @@ +{ + "cls_token": "[CLS]", + "mask_token": "[MASK]", + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6000/tokenizer.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6000/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..ebd782de6963b9032ce9374cfe30f5275be1a30b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6000/tokenizer.json @@ -0,0 +1,4265 @@ +{ + "version": "1.0", + "truncation": { + "direction": "Right", + "max_length": 512, + "strategy": "LongestFirst", + "stride": 0 + }, + "padding": { + "strategy": { + "Fixed": 512 + }, + "direction": "Right", + "pad_to_multiple_of": null, + "pad_id": 0, + "pad_type_id": 0, + "pad_token": "[PAD]" + }, + "added_tokens": [ + { + "id": 0, + "content": "[PAD]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 1, + "content": "[UNK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 2, + "content": "[CLS]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 3, + "content": "[SEP]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 4, + "content": "[MASK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + } + ], + "normalizer": { + "type": "BertNormalizer", + "clean_text": true, + "handle_chinese_chars": true, + "strip_accents": null, + "lowercase": false + }, + "pre_tokenizer": { + "type": "BertPreTokenizer" + }, + "post_processor": { + "type": "TemplateProcessing", + "single": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + } + ], + "pair": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "B", + "type_id": 1 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 1 + } + } + ], + "special_tokens": { + "[CLS]": { + "id": "[CLS]", + "ids": [ + 2 + ], + "tokens": [ + "[CLS]" + ] + }, + "[SEP]": { + "id": "[SEP]", + "ids": [ + 3 + ], + "tokens": [ + "[SEP]" + ] + } + } + }, + "decoder": { + "type": "WordPiece", + "prefix": "##", + "cleanup": true + }, + "model": { + "type": "WordPiece", + "unk_token": "[UNK]", + "continuing_subword_prefix": "##", + "max_input_chars_per_word": 100, + "vocab": { + "[PAD]": 0, + "[UNK]": 1, + "[CLS]": 2, + "[SEP]": 3, + "[MASK]": 4, + "AAAAAA": 5, + "AAAAAT": 6, + "AAAAAC": 7, + "AAAAAG": 8, + "AAAATA": 9, + "AAAATT": 10, + "AAAATC": 11, + "AAAATG": 12, + "AAAACA": 13, + "AAAACT": 14, + "AAAACC": 15, + "AAAACG": 16, + "AAAAGA": 17, + "AAAAGT": 18, + "AAAAGC": 19, + "AAAAGG": 20, + "AAATAA": 21, + "AAATAT": 22, + "AAATAC": 23, + "AAATAG": 24, + "AAATTA": 25, + "AAATTT": 26, + "AAATTC": 27, + "AAATTG": 28, + "AAATCA": 29, + "AAATCT": 30, + "AAATCC": 31, + "AAATCG": 32, + "AAATGA": 33, + "AAATGT": 34, + "AAATGC": 35, + "AAATGG": 36, + "AAACAA": 37, + "AAACAT": 38, + "AAACAC": 39, + "AAACAG": 40, + "AAACTA": 41, + "AAACTT": 42, + "AAACTC": 43, + "AAACTG": 44, + "AAACCA": 45, + "AAACCT": 46, + "AAACCC": 47, + "AAACCG": 48, + "AAACGA": 49, + "AAACGT": 50, + "AAACGC": 51, + "AAACGG": 52, + "AAAGAA": 53, + "AAAGAT": 54, + "AAAGAC": 55, + "AAAGAG": 56, + "AAAGTA": 57, + "AAAGTT": 58, + "AAAGTC": 59, + "AAAGTG": 60, + "AAAGCA": 61, + "AAAGCT": 62, + "AAAGCC": 63, + "AAAGCG": 64, + "AAAGGA": 65, + "AAAGGT": 66, + "AAAGGC": 67, + "AAAGGG": 68, + "AATAAA": 69, + "AATAAT": 70, + "AATAAC": 71, + "AATAAG": 72, + "AATATA": 73, + "AATATT": 74, + "AATATC": 75, + "AATATG": 76, + "AATACA": 77, + "AATACT": 78, + "AATACC": 79, + "AATACG": 80, + "AATAGA": 81, + "AATAGT": 82, + "AATAGC": 83, + "AATAGG": 84, + "AATTAA": 85, + "AATTAT": 86, + "AATTAC": 87, + "AATTAG": 88, + "AATTTA": 89, + "AATTTT": 90, + "AATTTC": 91, + "AATTTG": 92, + "AATTCA": 93, + "AATTCT": 94, + "AATTCC": 95, + "AATTCG": 96, + "AATTGA": 97, + "AATTGT": 98, + "AATTGC": 99, + "AATTGG": 100, + "AATCAA": 101, + "AATCAT": 102, + "AATCAC": 103, + "AATCAG": 104, + "AATCTA": 105, + "AATCTT": 106, + "AATCTC": 107, + "AATCTG": 108, + "AATCCA": 109, + "AATCCT": 110, + "AATCCC": 111, + "AATCCG": 112, + "AATCGA": 113, + "AATCGT": 114, + "AATCGC": 115, + "AATCGG": 116, + "AATGAA": 117, + "AATGAT": 118, + "AATGAC": 119, + "AATGAG": 120, + "AATGTA": 121, + "AATGTT": 122, + "AATGTC": 123, + "AATGTG": 124, + "AATGCA": 125, + "AATGCT": 126, + "AATGCC": 127, + "AATGCG": 128, + "AATGGA": 129, + "AATGGT": 130, + "AATGGC": 131, + "AATGGG": 132, + "AACAAA": 133, + "AACAAT": 134, + "AACAAC": 135, + "AACAAG": 136, + "AACATA": 137, + "AACATT": 138, + "AACATC": 139, + "AACATG": 140, + "AACACA": 141, + "AACACT": 142, + "AACACC": 143, + "AACACG": 144, + "AACAGA": 145, + "AACAGT": 146, + "AACAGC": 147, + "AACAGG": 148, + "AACTAA": 149, + "AACTAT": 150, + "AACTAC": 151, + "AACTAG": 152, + "AACTTA": 153, + "AACTTT": 154, + "AACTTC": 155, + "AACTTG": 156, + "AACTCA": 157, + "AACTCT": 158, + "AACTCC": 159, + "AACTCG": 160, + "AACTGA": 161, + "AACTGT": 162, + "AACTGC": 163, + "AACTGG": 164, + "AACCAA": 165, + "AACCAT": 166, + "AACCAC": 167, + "AACCAG": 168, + "AACCTA": 169, + "AACCTT": 170, + "AACCTC": 171, + "AACCTG": 172, + "AACCCA": 173, + "AACCCT": 174, + "AACCCC": 175, + "AACCCG": 176, + "AACCGA": 177, + "AACCGT": 178, + "AACCGC": 179, + "AACCGG": 180, + "AACGAA": 181, + "AACGAT": 182, + "AACGAC": 183, + "AACGAG": 184, + "AACGTA": 185, + "AACGTT": 186, + "AACGTC": 187, + "AACGTG": 188, + "AACGCA": 189, + "AACGCT": 190, + "AACGCC": 191, + "AACGCG": 192, + "AACGGA": 193, + "AACGGT": 194, + "AACGGC": 195, + "AACGGG": 196, + "AAGAAA": 197, + "AAGAAT": 198, + "AAGAAC": 199, + "AAGAAG": 200, + "AAGATA": 201, + "AAGATT": 202, + "AAGATC": 203, + "AAGATG": 204, + "AAGACA": 205, + "AAGACT": 206, + "AAGACC": 207, + "AAGACG": 208, + "AAGAGA": 209, + "AAGAGT": 210, + "AAGAGC": 211, + "AAGAGG": 212, + "AAGTAA": 213, + "AAGTAT": 214, + "AAGTAC": 215, + "AAGTAG": 216, + "AAGTTA": 217, + "AAGTTT": 218, + "AAGTTC": 219, + "AAGTTG": 220, + "AAGTCA": 221, + "AAGTCT": 222, + "AAGTCC": 223, + "AAGTCG": 224, + "AAGTGA": 225, + "AAGTGT": 226, + "AAGTGC": 227, + "AAGTGG": 228, + "AAGCAA": 229, + "AAGCAT": 230, + "AAGCAC": 231, + "AAGCAG": 232, + "AAGCTA": 233, + "AAGCTT": 234, + "AAGCTC": 235, + "AAGCTG": 236, + "AAGCCA": 237, + "AAGCCT": 238, + "AAGCCC": 239, + "AAGCCG": 240, + "AAGCGA": 241, + "AAGCGT": 242, + "AAGCGC": 243, + "AAGCGG": 244, + "AAGGAA": 245, + "AAGGAT": 246, + "AAGGAC": 247, + "AAGGAG": 248, + "AAGGTA": 249, + "AAGGTT": 250, + "AAGGTC": 251, + "AAGGTG": 252, + "AAGGCA": 253, + "AAGGCT": 254, + "AAGGCC": 255, + "AAGGCG": 256, + "AAGGGA": 257, + "AAGGGT": 258, + "AAGGGC": 259, + "AAGGGG": 260, + "ATAAAA": 261, + "ATAAAT": 262, + "ATAAAC": 263, + "ATAAAG": 264, + "ATAATA": 265, + "ATAATT": 266, + "ATAATC": 267, + "ATAATG": 268, + "ATAACA": 269, + "ATAACT": 270, + "ATAACC": 271, + "ATAACG": 272, + "ATAAGA": 273, + "ATAAGT": 274, + "ATAAGC": 275, + "ATAAGG": 276, + "ATATAA": 277, + "ATATAT": 278, + "ATATAC": 279, + "ATATAG": 280, + "ATATTA": 281, + "ATATTT": 282, + "ATATTC": 283, + "ATATTG": 284, + "ATATCA": 285, + "ATATCT": 286, + "ATATCC": 287, + "ATATCG": 288, + "ATATGA": 289, + "ATATGT": 290, + "ATATGC": 291, + "ATATGG": 292, + "ATACAA": 293, + "ATACAT": 294, + "ATACAC": 295, + "ATACAG": 296, + "ATACTA": 297, + "ATACTT": 298, + "ATACTC": 299, + "ATACTG": 300, + "ATACCA": 301, + "ATACCT": 302, + "ATACCC": 303, + "ATACCG": 304, + "ATACGA": 305, + "ATACGT": 306, + "ATACGC": 307, + "ATACGG": 308, + "ATAGAA": 309, + "ATAGAT": 310, + "ATAGAC": 311, + "ATAGAG": 312, + "ATAGTA": 313, + "ATAGTT": 314, + "ATAGTC": 315, + "ATAGTG": 316, + "ATAGCA": 317, + "ATAGCT": 318, + "ATAGCC": 319, + "ATAGCG": 320, + "ATAGGA": 321, + "ATAGGT": 322, + "ATAGGC": 323, + "ATAGGG": 324, + "ATTAAA": 325, + "ATTAAT": 326, + "ATTAAC": 327, + "ATTAAG": 328, + "ATTATA": 329, + "ATTATT": 330, + "ATTATC": 331, + "ATTATG": 332, + "ATTACA": 333, + "ATTACT": 334, + "ATTACC": 335, + "ATTACG": 336, + "ATTAGA": 337, + "ATTAGT": 338, + "ATTAGC": 339, + "ATTAGG": 340, + "ATTTAA": 341, + "ATTTAT": 342, + "ATTTAC": 343, + "ATTTAG": 344, + "ATTTTA": 345, + "ATTTTT": 346, + "ATTTTC": 347, + "ATTTTG": 348, + "ATTTCA": 349, + "ATTTCT": 350, + "ATTTCC": 351, + "ATTTCG": 352, + "ATTTGA": 353, + "ATTTGT": 354, + "ATTTGC": 355, + "ATTTGG": 356, + "ATTCAA": 357, + "ATTCAT": 358, + "ATTCAC": 359, + "ATTCAG": 360, + "ATTCTA": 361, + "ATTCTT": 362, + "ATTCTC": 363, + "ATTCTG": 364, + "ATTCCA": 365, + "ATTCCT": 366, + "ATTCCC": 367, + "ATTCCG": 368, + "ATTCGA": 369, + "ATTCGT": 370, + "ATTCGC": 371, + "ATTCGG": 372, + "ATTGAA": 373, + "ATTGAT": 374, + "ATTGAC": 375, + "ATTGAG": 376, + "ATTGTA": 377, + "ATTGTT": 378, + "ATTGTC": 379, + "ATTGTG": 380, + "ATTGCA": 381, + "ATTGCT": 382, + "ATTGCC": 383, + "ATTGCG": 384, + "ATTGGA": 385, + "ATTGGT": 386, + "ATTGGC": 387, + "ATTGGG": 388, + "ATCAAA": 389, + "ATCAAT": 390, + "ATCAAC": 391, + "ATCAAG": 392, + "ATCATA": 393, + "ATCATT": 394, + "ATCATC": 395, + "ATCATG": 396, + "ATCACA": 397, + "ATCACT": 398, + "ATCACC": 399, + "ATCACG": 400, + "ATCAGA": 401, + "ATCAGT": 402, + "ATCAGC": 403, + "ATCAGG": 404, + "ATCTAA": 405, + "ATCTAT": 406, + "ATCTAC": 407, + "ATCTAG": 408, + "ATCTTA": 409, + "ATCTTT": 410, + "ATCTTC": 411, + "ATCTTG": 412, + "ATCTCA": 413, + "ATCTCT": 414, + "ATCTCC": 415, + "ATCTCG": 416, + "ATCTGA": 417, + "ATCTGT": 418, + "ATCTGC": 419, + "ATCTGG": 420, + "ATCCAA": 421, + "ATCCAT": 422, + "ATCCAC": 423, + "ATCCAG": 424, + "ATCCTA": 425, + "ATCCTT": 426, + "ATCCTC": 427, + "ATCCTG": 428, + "ATCCCA": 429, + "ATCCCT": 430, + "ATCCCC": 431, + "ATCCCG": 432, + "ATCCGA": 433, + "ATCCGT": 434, + "ATCCGC": 435, + "ATCCGG": 436, + "ATCGAA": 437, + "ATCGAT": 438, + "ATCGAC": 439, + "ATCGAG": 440, + "ATCGTA": 441, + "ATCGTT": 442, + "ATCGTC": 443, + "ATCGTG": 444, + "ATCGCA": 445, + "ATCGCT": 446, + "ATCGCC": 447, + "ATCGCG": 448, + "ATCGGA": 449, + "ATCGGT": 450, + "ATCGGC": 451, + "ATCGGG": 452, + "ATGAAA": 453, + "ATGAAT": 454, + "ATGAAC": 455, + "ATGAAG": 456, + "ATGATA": 457, + "ATGATT": 458, + "ATGATC": 459, + "ATGATG": 460, + "ATGACA": 461, + "ATGACT": 462, + "ATGACC": 463, + "ATGACG": 464, + "ATGAGA": 465, + "ATGAGT": 466, + "ATGAGC": 467, + "ATGAGG": 468, + "ATGTAA": 469, + "ATGTAT": 470, + "ATGTAC": 471, + "ATGTAG": 472, + "ATGTTA": 473, + "ATGTTT": 474, + "ATGTTC": 475, + "ATGTTG": 476, + "ATGTCA": 477, + "ATGTCT": 478, + "ATGTCC": 479, + "ATGTCG": 480, + "ATGTGA": 481, + "ATGTGT": 482, + "ATGTGC": 483, + "ATGTGG": 484, + "ATGCAA": 485, + "ATGCAT": 486, + "ATGCAC": 487, + "ATGCAG": 488, + "ATGCTA": 489, + "ATGCTT": 490, + "ATGCTC": 491, + "ATGCTG": 492, + "ATGCCA": 493, + "ATGCCT": 494, + "ATGCCC": 495, + "ATGCCG": 496, + "ATGCGA": 497, + "ATGCGT": 498, + "ATGCGC": 499, + "ATGCGG": 500, + "ATGGAA": 501, + "ATGGAT": 502, + "ATGGAC": 503, + "ATGGAG": 504, + "ATGGTA": 505, + "ATGGTT": 506, + "ATGGTC": 507, + "ATGGTG": 508, + "ATGGCA": 509, + "ATGGCT": 510, + "ATGGCC": 511, + "ATGGCG": 512, + "ATGGGA": 513, + "ATGGGT": 514, + "ATGGGC": 515, + "ATGGGG": 516, + "ACAAAA": 517, + "ACAAAT": 518, + "ACAAAC": 519, + "ACAAAG": 520, + "ACAATA": 521, + "ACAATT": 522, + "ACAATC": 523, + "ACAATG": 524, + "ACAACA": 525, + "ACAACT": 526, + "ACAACC": 527, + "ACAACG": 528, + "ACAAGA": 529, + "ACAAGT": 530, + "ACAAGC": 531, + "ACAAGG": 532, + "ACATAA": 533, + "ACATAT": 534, + "ACATAC": 535, + "ACATAG": 536, + "ACATTA": 537, + "ACATTT": 538, + "ACATTC": 539, + "ACATTG": 540, + "ACATCA": 541, + "ACATCT": 542, + "ACATCC": 543, + "ACATCG": 544, + "ACATGA": 545, + "ACATGT": 546, + "ACATGC": 547, + "ACATGG": 548, + "ACACAA": 549, + "ACACAT": 550, + "ACACAC": 551, + "ACACAG": 552, + "ACACTA": 553, + "ACACTT": 554, + "ACACTC": 555, + "ACACTG": 556, + "ACACCA": 557, + "ACACCT": 558, + "ACACCC": 559, + "ACACCG": 560, + "ACACGA": 561, + "ACACGT": 562, + "ACACGC": 563, + "ACACGG": 564, + "ACAGAA": 565, + "ACAGAT": 566, + "ACAGAC": 567, + "ACAGAG": 568, + "ACAGTA": 569, + "ACAGTT": 570, + "ACAGTC": 571, + "ACAGTG": 572, + "ACAGCA": 573, + "ACAGCT": 574, + "ACAGCC": 575, + "ACAGCG": 576, + "ACAGGA": 577, + "ACAGGT": 578, + "ACAGGC": 579, + "ACAGGG": 580, + "ACTAAA": 581, + "ACTAAT": 582, + "ACTAAC": 583, + "ACTAAG": 584, + "ACTATA": 585, + "ACTATT": 586, + "ACTATC": 587, + "ACTATG": 588, + "ACTACA": 589, + "ACTACT": 590, + "ACTACC": 591, + "ACTACG": 592, + "ACTAGA": 593, + "ACTAGT": 594, + "ACTAGC": 595, + "ACTAGG": 596, + "ACTTAA": 597, + "ACTTAT": 598, + "ACTTAC": 599, + "ACTTAG": 600, + "ACTTTA": 601, + "ACTTTT": 602, + "ACTTTC": 603, + "ACTTTG": 604, + "ACTTCA": 605, + "ACTTCT": 606, + "ACTTCC": 607, + "ACTTCG": 608, + "ACTTGA": 609, + "ACTTGT": 610, + "ACTTGC": 611, + "ACTTGG": 612, + "ACTCAA": 613, + "ACTCAT": 614, + "ACTCAC": 615, + "ACTCAG": 616, + "ACTCTA": 617, + "ACTCTT": 618, + "ACTCTC": 619, + "ACTCTG": 620, + "ACTCCA": 621, + "ACTCCT": 622, + "ACTCCC": 623, + "ACTCCG": 624, + "ACTCGA": 625, + "ACTCGT": 626, + "ACTCGC": 627, + "ACTCGG": 628, + "ACTGAA": 629, + "ACTGAT": 630, + "ACTGAC": 631, + "ACTGAG": 632, + "ACTGTA": 633, + "ACTGTT": 634, + "ACTGTC": 635, + "ACTGTG": 636, + "ACTGCA": 637, + "ACTGCT": 638, + "ACTGCC": 639, + "ACTGCG": 640, + "ACTGGA": 641, + "ACTGGT": 642, + "ACTGGC": 643, + "ACTGGG": 644, + "ACCAAA": 645, + "ACCAAT": 646, + "ACCAAC": 647, + "ACCAAG": 648, + "ACCATA": 649, + "ACCATT": 650, + "ACCATC": 651, + "ACCATG": 652, + "ACCACA": 653, + "ACCACT": 654, + "ACCACC": 655, + "ACCACG": 656, + "ACCAGA": 657, + "ACCAGT": 658, + "ACCAGC": 659, + "ACCAGG": 660, + "ACCTAA": 661, + "ACCTAT": 662, + "ACCTAC": 663, + "ACCTAG": 664, + "ACCTTA": 665, + "ACCTTT": 666, + "ACCTTC": 667, + "ACCTTG": 668, + "ACCTCA": 669, + "ACCTCT": 670, + "ACCTCC": 671, + "ACCTCG": 672, + "ACCTGA": 673, + "ACCTGT": 674, + "ACCTGC": 675, + "ACCTGG": 676, + "ACCCAA": 677, + "ACCCAT": 678, + "ACCCAC": 679, + "ACCCAG": 680, + "ACCCTA": 681, + "ACCCTT": 682, + "ACCCTC": 683, + "ACCCTG": 684, + "ACCCCA": 685, + "ACCCCT": 686, + "ACCCCC": 687, + "ACCCCG": 688, + "ACCCGA": 689, + "ACCCGT": 690, + "ACCCGC": 691, + "ACCCGG": 692, + "ACCGAA": 693, + "ACCGAT": 694, + "ACCGAC": 695, + "ACCGAG": 696, + "ACCGTA": 697, + "ACCGTT": 698, + "ACCGTC": 699, + "ACCGTG": 700, + "ACCGCA": 701, + "ACCGCT": 702, + "ACCGCC": 703, + "ACCGCG": 704, + "ACCGGA": 705, + "ACCGGT": 706, + "ACCGGC": 707, + "ACCGGG": 708, + "ACGAAA": 709, + "ACGAAT": 710, + "ACGAAC": 711, + "ACGAAG": 712, + "ACGATA": 713, + "ACGATT": 714, + "ACGATC": 715, + "ACGATG": 716, + "ACGACA": 717, + "ACGACT": 718, + "ACGACC": 719, + "ACGACG": 720, + "ACGAGA": 721, + "ACGAGT": 722, + "ACGAGC": 723, + "ACGAGG": 724, + "ACGTAA": 725, + "ACGTAT": 726, + "ACGTAC": 727, + "ACGTAG": 728, + "ACGTTA": 729, + "ACGTTT": 730, + "ACGTTC": 731, + "ACGTTG": 732, + "ACGTCA": 733, + "ACGTCT": 734, + "ACGTCC": 735, + "ACGTCG": 736, + "ACGTGA": 737, + "ACGTGT": 738, + "ACGTGC": 739, + "ACGTGG": 740, + "ACGCAA": 741, + "ACGCAT": 742, + "ACGCAC": 743, + "ACGCAG": 744, + "ACGCTA": 745, + "ACGCTT": 746, + "ACGCTC": 747, + "ACGCTG": 748, + "ACGCCA": 749, + "ACGCCT": 750, + "ACGCCC": 751, + "ACGCCG": 752, + "ACGCGA": 753, + "ACGCGT": 754, + "ACGCGC": 755, + "ACGCGG": 756, + "ACGGAA": 757, + "ACGGAT": 758, + "ACGGAC": 759, + "ACGGAG": 760, + "ACGGTA": 761, + "ACGGTT": 762, + "ACGGTC": 763, + "ACGGTG": 764, + "ACGGCA": 765, + "ACGGCT": 766, + "ACGGCC": 767, + "ACGGCG": 768, + "ACGGGA": 769, + "ACGGGT": 770, + "ACGGGC": 771, + "ACGGGG": 772, + "AGAAAA": 773, + "AGAAAT": 774, + "AGAAAC": 775, + "AGAAAG": 776, + "AGAATA": 777, + "AGAATT": 778, + "AGAATC": 779, + "AGAATG": 780, + "AGAACA": 781, + "AGAACT": 782, + "AGAACC": 783, + "AGAACG": 784, + "AGAAGA": 785, + "AGAAGT": 786, + "AGAAGC": 787, + "AGAAGG": 788, + "AGATAA": 789, + "AGATAT": 790, + "AGATAC": 791, + "AGATAG": 792, + "AGATTA": 793, + "AGATTT": 794, + "AGATTC": 795, + "AGATTG": 796, + "AGATCA": 797, + "AGATCT": 798, + "AGATCC": 799, + "AGATCG": 800, + "AGATGA": 801, + "AGATGT": 802, + "AGATGC": 803, + "AGATGG": 804, + "AGACAA": 805, + "AGACAT": 806, + "AGACAC": 807, + "AGACAG": 808, + "AGACTA": 809, + "AGACTT": 810, + "AGACTC": 811, + "AGACTG": 812, + "AGACCA": 813, + "AGACCT": 814, + "AGACCC": 815, + "AGACCG": 816, + "AGACGA": 817, + "AGACGT": 818, + "AGACGC": 819, + "AGACGG": 820, + "AGAGAA": 821, + "AGAGAT": 822, + "AGAGAC": 823, + "AGAGAG": 824, + "AGAGTA": 825, + "AGAGTT": 826, + "AGAGTC": 827, + "AGAGTG": 828, + "AGAGCA": 829, + "AGAGCT": 830, + "AGAGCC": 831, + "AGAGCG": 832, + "AGAGGA": 833, + "AGAGGT": 834, + "AGAGGC": 835, + "AGAGGG": 836, + "AGTAAA": 837, + "AGTAAT": 838, + "AGTAAC": 839, + "AGTAAG": 840, + "AGTATA": 841, + "AGTATT": 842, + "AGTATC": 843, + "AGTATG": 844, + "AGTACA": 845, + "AGTACT": 846, + "AGTACC": 847, + "AGTACG": 848, + "AGTAGA": 849, + "AGTAGT": 850, + "AGTAGC": 851, + "AGTAGG": 852, + "AGTTAA": 853, + "AGTTAT": 854, + "AGTTAC": 855, + "AGTTAG": 856, + "AGTTTA": 857, + "AGTTTT": 858, + "AGTTTC": 859, + "AGTTTG": 860, + "AGTTCA": 861, + "AGTTCT": 862, + "AGTTCC": 863, + "AGTTCG": 864, + "AGTTGA": 865, + "AGTTGT": 866, + "AGTTGC": 867, + "AGTTGG": 868, + "AGTCAA": 869, + "AGTCAT": 870, + "AGTCAC": 871, + "AGTCAG": 872, + "AGTCTA": 873, + "AGTCTT": 874, + "AGTCTC": 875, + "AGTCTG": 876, + "AGTCCA": 877, + "AGTCCT": 878, + "AGTCCC": 879, + "AGTCCG": 880, + "AGTCGA": 881, + "AGTCGT": 882, + "AGTCGC": 883, + "AGTCGG": 884, + "AGTGAA": 885, + "AGTGAT": 886, + "AGTGAC": 887, + "AGTGAG": 888, + "AGTGTA": 889, + "AGTGTT": 890, + "AGTGTC": 891, + "AGTGTG": 892, + "AGTGCA": 893, + "AGTGCT": 894, + "AGTGCC": 895, + "AGTGCG": 896, + "AGTGGA": 897, + "AGTGGT": 898, + "AGTGGC": 899, + "AGTGGG": 900, + "AGCAAA": 901, + "AGCAAT": 902, + "AGCAAC": 903, + "AGCAAG": 904, + "AGCATA": 905, + "AGCATT": 906, + "AGCATC": 907, + "AGCATG": 908, + "AGCACA": 909, + "AGCACT": 910, + "AGCACC": 911, + "AGCACG": 912, + "AGCAGA": 913, + "AGCAGT": 914, + "AGCAGC": 915, + "AGCAGG": 916, + "AGCTAA": 917, + "AGCTAT": 918, + "AGCTAC": 919, + "AGCTAG": 920, + "AGCTTA": 921, + "AGCTTT": 922, + "AGCTTC": 923, + "AGCTTG": 924, + "AGCTCA": 925, + "AGCTCT": 926, + "AGCTCC": 927, + "AGCTCG": 928, + "AGCTGA": 929, + "AGCTGT": 930, + "AGCTGC": 931, + "AGCTGG": 932, + "AGCCAA": 933, + "AGCCAT": 934, + "AGCCAC": 935, + "AGCCAG": 936, + "AGCCTA": 937, + "AGCCTT": 938, + "AGCCTC": 939, + "AGCCTG": 940, + "AGCCCA": 941, + "AGCCCT": 942, + "AGCCCC": 943, + "AGCCCG": 944, + "AGCCGA": 945, + "AGCCGT": 946, + "AGCCGC": 947, + "AGCCGG": 948, + "AGCGAA": 949, + "AGCGAT": 950, + "AGCGAC": 951, + "AGCGAG": 952, + "AGCGTA": 953, + "AGCGTT": 954, + "AGCGTC": 955, + "AGCGTG": 956, + "AGCGCA": 957, + "AGCGCT": 958, + "AGCGCC": 959, + "AGCGCG": 960, + "AGCGGA": 961, + "AGCGGT": 962, + "AGCGGC": 963, + "AGCGGG": 964, + "AGGAAA": 965, + "AGGAAT": 966, + "AGGAAC": 967, + "AGGAAG": 968, + "AGGATA": 969, + "AGGATT": 970, + "AGGATC": 971, + "AGGATG": 972, + "AGGACA": 973, + "AGGACT": 974, + "AGGACC": 975, + "AGGACG": 976, + "AGGAGA": 977, + "AGGAGT": 978, + "AGGAGC": 979, + "AGGAGG": 980, + "AGGTAA": 981, + "AGGTAT": 982, + "AGGTAC": 983, + "AGGTAG": 984, + "AGGTTA": 985, + "AGGTTT": 986, + "AGGTTC": 987, + "AGGTTG": 988, + "AGGTCA": 989, + "AGGTCT": 990, + "AGGTCC": 991, + "AGGTCG": 992, + "AGGTGA": 993, + "AGGTGT": 994, + "AGGTGC": 995, + "AGGTGG": 996, + "AGGCAA": 997, + "AGGCAT": 998, + "AGGCAC": 999, + "AGGCAG": 1000, + "AGGCTA": 1001, + "AGGCTT": 1002, + "AGGCTC": 1003, + "AGGCTG": 1004, + "AGGCCA": 1005, + "AGGCCT": 1006, + "AGGCCC": 1007, + "AGGCCG": 1008, + "AGGCGA": 1009, + "AGGCGT": 1010, + "AGGCGC": 1011, + "AGGCGG": 1012, + "AGGGAA": 1013, + "AGGGAT": 1014, + "AGGGAC": 1015, + "AGGGAG": 1016, + "AGGGTA": 1017, + "AGGGTT": 1018, + "AGGGTC": 1019, + "AGGGTG": 1020, + "AGGGCA": 1021, + "AGGGCT": 1022, + "AGGGCC": 1023, + "AGGGCG": 1024, + "AGGGGA": 1025, + "AGGGGT": 1026, + "AGGGGC": 1027, + "AGGGGG": 1028, + "TAAAAA": 1029, + "TAAAAT": 1030, + "TAAAAC": 1031, + "TAAAAG": 1032, + "TAAATA": 1033, + "TAAATT": 1034, + "TAAATC": 1035, + "TAAATG": 1036, + "TAAACA": 1037, + "TAAACT": 1038, + "TAAACC": 1039, + "TAAACG": 1040, + "TAAAGA": 1041, + "TAAAGT": 1042, + "TAAAGC": 1043, + "TAAAGG": 1044, + "TAATAA": 1045, + "TAATAT": 1046, + "TAATAC": 1047, + "TAATAG": 1048, + "TAATTA": 1049, + "TAATTT": 1050, + "TAATTC": 1051, + "TAATTG": 1052, + "TAATCA": 1053, + "TAATCT": 1054, + "TAATCC": 1055, + "TAATCG": 1056, + "TAATGA": 1057, + "TAATGT": 1058, + "TAATGC": 1059, + "TAATGG": 1060, + "TAACAA": 1061, + "TAACAT": 1062, + "TAACAC": 1063, + "TAACAG": 1064, + "TAACTA": 1065, + "TAACTT": 1066, + "TAACTC": 1067, + "TAACTG": 1068, + "TAACCA": 1069, + "TAACCT": 1070, + "TAACCC": 1071, + "TAACCG": 1072, + "TAACGA": 1073, + "TAACGT": 1074, + "TAACGC": 1075, + "TAACGG": 1076, + "TAAGAA": 1077, + "TAAGAT": 1078, + "TAAGAC": 1079, + "TAAGAG": 1080, + "TAAGTA": 1081, + "TAAGTT": 1082, + "TAAGTC": 1083, + "TAAGTG": 1084, + "TAAGCA": 1085, + "TAAGCT": 1086, + "TAAGCC": 1087, + "TAAGCG": 1088, + "TAAGGA": 1089, + "TAAGGT": 1090, + "TAAGGC": 1091, + "TAAGGG": 1092, + "TATAAA": 1093, + "TATAAT": 1094, + "TATAAC": 1095, + "TATAAG": 1096, + "TATATA": 1097, + "TATATT": 1098, + "TATATC": 1099, + "TATATG": 1100, + "TATACA": 1101, + "TATACT": 1102, + "TATACC": 1103, + "TATACG": 1104, + "TATAGA": 1105, + "TATAGT": 1106, + "TATAGC": 1107, + "TATAGG": 1108, + "TATTAA": 1109, + "TATTAT": 1110, + "TATTAC": 1111, + "TATTAG": 1112, + "TATTTA": 1113, + "TATTTT": 1114, + "TATTTC": 1115, + "TATTTG": 1116, + "TATTCA": 1117, + "TATTCT": 1118, + "TATTCC": 1119, + "TATTCG": 1120, + "TATTGA": 1121, + "TATTGT": 1122, + "TATTGC": 1123, + "TATTGG": 1124, + "TATCAA": 1125, + "TATCAT": 1126, + "TATCAC": 1127, + "TATCAG": 1128, + "TATCTA": 1129, + "TATCTT": 1130, + "TATCTC": 1131, + "TATCTG": 1132, + "TATCCA": 1133, + "TATCCT": 1134, + "TATCCC": 1135, + "TATCCG": 1136, + "TATCGA": 1137, + "TATCGT": 1138, + "TATCGC": 1139, + "TATCGG": 1140, + "TATGAA": 1141, + "TATGAT": 1142, + "TATGAC": 1143, + "TATGAG": 1144, + "TATGTA": 1145, + "TATGTT": 1146, + "TATGTC": 1147, + "TATGTG": 1148, + "TATGCA": 1149, + "TATGCT": 1150, + "TATGCC": 1151, + "TATGCG": 1152, + "TATGGA": 1153, + "TATGGT": 1154, + "TATGGC": 1155, + "TATGGG": 1156, + "TACAAA": 1157, + "TACAAT": 1158, + "TACAAC": 1159, + "TACAAG": 1160, + "TACATA": 1161, + "TACATT": 1162, + "TACATC": 1163, + "TACATG": 1164, + "TACACA": 1165, + "TACACT": 1166, + "TACACC": 1167, + "TACACG": 1168, + "TACAGA": 1169, + "TACAGT": 1170, + "TACAGC": 1171, + "TACAGG": 1172, + "TACTAA": 1173, + "TACTAT": 1174, + "TACTAC": 1175, + "TACTAG": 1176, + "TACTTA": 1177, + "TACTTT": 1178, + "TACTTC": 1179, + "TACTTG": 1180, + "TACTCA": 1181, + "TACTCT": 1182, + "TACTCC": 1183, + "TACTCG": 1184, + "TACTGA": 1185, + "TACTGT": 1186, + "TACTGC": 1187, + "TACTGG": 1188, + "TACCAA": 1189, + "TACCAT": 1190, + "TACCAC": 1191, + "TACCAG": 1192, + "TACCTA": 1193, + "TACCTT": 1194, + "TACCTC": 1195, + "TACCTG": 1196, + "TACCCA": 1197, + "TACCCT": 1198, + "TACCCC": 1199, + "TACCCG": 1200, + "TACCGA": 1201, + "TACCGT": 1202, + "TACCGC": 1203, + "TACCGG": 1204, + "TACGAA": 1205, + "TACGAT": 1206, + "TACGAC": 1207, + "TACGAG": 1208, + "TACGTA": 1209, + "TACGTT": 1210, + "TACGTC": 1211, + "TACGTG": 1212, + "TACGCA": 1213, + "TACGCT": 1214, + "TACGCC": 1215, + "TACGCG": 1216, + "TACGGA": 1217, + "TACGGT": 1218, + "TACGGC": 1219, + "TACGGG": 1220, + "TAGAAA": 1221, + "TAGAAT": 1222, + "TAGAAC": 1223, + "TAGAAG": 1224, + "TAGATA": 1225, + "TAGATT": 1226, + "TAGATC": 1227, + "TAGATG": 1228, + "TAGACA": 1229, + "TAGACT": 1230, + "TAGACC": 1231, + "TAGACG": 1232, + "TAGAGA": 1233, + "TAGAGT": 1234, + "TAGAGC": 1235, + "TAGAGG": 1236, + "TAGTAA": 1237, + "TAGTAT": 1238, + "TAGTAC": 1239, + "TAGTAG": 1240, + "TAGTTA": 1241, + "TAGTTT": 1242, + "TAGTTC": 1243, + "TAGTTG": 1244, + "TAGTCA": 1245, + "TAGTCT": 1246, + "TAGTCC": 1247, + "TAGTCG": 1248, + "TAGTGA": 1249, + "TAGTGT": 1250, + "TAGTGC": 1251, + "TAGTGG": 1252, + "TAGCAA": 1253, + "TAGCAT": 1254, + "TAGCAC": 1255, + "TAGCAG": 1256, + "TAGCTA": 1257, + "TAGCTT": 1258, + "TAGCTC": 1259, + "TAGCTG": 1260, + "TAGCCA": 1261, + "TAGCCT": 1262, + "TAGCCC": 1263, + "TAGCCG": 1264, + "TAGCGA": 1265, + "TAGCGT": 1266, + "TAGCGC": 1267, + "TAGCGG": 1268, + "TAGGAA": 1269, + "TAGGAT": 1270, + "TAGGAC": 1271, + "TAGGAG": 1272, + "TAGGTA": 1273, + "TAGGTT": 1274, + "TAGGTC": 1275, + "TAGGTG": 1276, + "TAGGCA": 1277, + "TAGGCT": 1278, + "TAGGCC": 1279, + "TAGGCG": 1280, + "TAGGGA": 1281, + "TAGGGT": 1282, + "TAGGGC": 1283, + "TAGGGG": 1284, + "TTAAAA": 1285, + "TTAAAT": 1286, + "TTAAAC": 1287, + "TTAAAG": 1288, + "TTAATA": 1289, + "TTAATT": 1290, + "TTAATC": 1291, + "TTAATG": 1292, + "TTAACA": 1293, + "TTAACT": 1294, + "TTAACC": 1295, + "TTAACG": 1296, + "TTAAGA": 1297, + "TTAAGT": 1298, + "TTAAGC": 1299, + "TTAAGG": 1300, + "TTATAA": 1301, + "TTATAT": 1302, + "TTATAC": 1303, + "TTATAG": 1304, + "TTATTA": 1305, + "TTATTT": 1306, + "TTATTC": 1307, + "TTATTG": 1308, + "TTATCA": 1309, + "TTATCT": 1310, + "TTATCC": 1311, + "TTATCG": 1312, + "TTATGA": 1313, + "TTATGT": 1314, + "TTATGC": 1315, + "TTATGG": 1316, + "TTACAA": 1317, + "TTACAT": 1318, + "TTACAC": 1319, + "TTACAG": 1320, + "TTACTA": 1321, + "TTACTT": 1322, + "TTACTC": 1323, + "TTACTG": 1324, + "TTACCA": 1325, + "TTACCT": 1326, + "TTACCC": 1327, + "TTACCG": 1328, + "TTACGA": 1329, + "TTACGT": 1330, + "TTACGC": 1331, + "TTACGG": 1332, + "TTAGAA": 1333, + "TTAGAT": 1334, + "TTAGAC": 1335, + "TTAGAG": 1336, + "TTAGTA": 1337, + "TTAGTT": 1338, + "TTAGTC": 1339, + "TTAGTG": 1340, + "TTAGCA": 1341, + "TTAGCT": 1342, + "TTAGCC": 1343, + "TTAGCG": 1344, + "TTAGGA": 1345, + "TTAGGT": 1346, + "TTAGGC": 1347, + "TTAGGG": 1348, + "TTTAAA": 1349, + "TTTAAT": 1350, + "TTTAAC": 1351, + "TTTAAG": 1352, + "TTTATA": 1353, + "TTTATT": 1354, + "TTTATC": 1355, + "TTTATG": 1356, + "TTTACA": 1357, + "TTTACT": 1358, + "TTTACC": 1359, + "TTTACG": 1360, + "TTTAGA": 1361, + "TTTAGT": 1362, + "TTTAGC": 1363, + "TTTAGG": 1364, + "TTTTAA": 1365, + "TTTTAT": 1366, + "TTTTAC": 1367, + "TTTTAG": 1368, + "TTTTTA": 1369, + "TTTTTT": 1370, + "TTTTTC": 1371, + "TTTTTG": 1372, + "TTTTCA": 1373, + "TTTTCT": 1374, + "TTTTCC": 1375, + "TTTTCG": 1376, + "TTTTGA": 1377, + "TTTTGT": 1378, + "TTTTGC": 1379, + "TTTTGG": 1380, + "TTTCAA": 1381, + "TTTCAT": 1382, + "TTTCAC": 1383, + "TTTCAG": 1384, + "TTTCTA": 1385, + "TTTCTT": 1386, + "TTTCTC": 1387, + "TTTCTG": 1388, + "TTTCCA": 1389, + "TTTCCT": 1390, + "TTTCCC": 1391, + "TTTCCG": 1392, + "TTTCGA": 1393, + "TTTCGT": 1394, + "TTTCGC": 1395, + "TTTCGG": 1396, + "TTTGAA": 1397, + "TTTGAT": 1398, + "TTTGAC": 1399, + "TTTGAG": 1400, + "TTTGTA": 1401, + "TTTGTT": 1402, + "TTTGTC": 1403, + "TTTGTG": 1404, + "TTTGCA": 1405, + "TTTGCT": 1406, + "TTTGCC": 1407, + "TTTGCG": 1408, + "TTTGGA": 1409, + "TTTGGT": 1410, + "TTTGGC": 1411, + "TTTGGG": 1412, + "TTCAAA": 1413, + "TTCAAT": 1414, + "TTCAAC": 1415, + "TTCAAG": 1416, + "TTCATA": 1417, + "TTCATT": 1418, + "TTCATC": 1419, + "TTCATG": 1420, + "TTCACA": 1421, + "TTCACT": 1422, + "TTCACC": 1423, + "TTCACG": 1424, + "TTCAGA": 1425, + "TTCAGT": 1426, + "TTCAGC": 1427, + "TTCAGG": 1428, + "TTCTAA": 1429, + "TTCTAT": 1430, + "TTCTAC": 1431, + "TTCTAG": 1432, + "TTCTTA": 1433, + "TTCTTT": 1434, + "TTCTTC": 1435, + "TTCTTG": 1436, + "TTCTCA": 1437, + "TTCTCT": 1438, + "TTCTCC": 1439, + "TTCTCG": 1440, + "TTCTGA": 1441, + "TTCTGT": 1442, + "TTCTGC": 1443, + "TTCTGG": 1444, + "TTCCAA": 1445, + "TTCCAT": 1446, + "TTCCAC": 1447, + "TTCCAG": 1448, + "TTCCTA": 1449, + "TTCCTT": 1450, + "TTCCTC": 1451, + "TTCCTG": 1452, + "TTCCCA": 1453, + "TTCCCT": 1454, + "TTCCCC": 1455, + "TTCCCG": 1456, + "TTCCGA": 1457, + "TTCCGT": 1458, + "TTCCGC": 1459, + "TTCCGG": 1460, + "TTCGAA": 1461, + "TTCGAT": 1462, + "TTCGAC": 1463, + "TTCGAG": 1464, + "TTCGTA": 1465, + "TTCGTT": 1466, + "TTCGTC": 1467, + "TTCGTG": 1468, + "TTCGCA": 1469, + "TTCGCT": 1470, + "TTCGCC": 1471, + "TTCGCG": 1472, + "TTCGGA": 1473, + "TTCGGT": 1474, + "TTCGGC": 1475, + "TTCGGG": 1476, + "TTGAAA": 1477, + "TTGAAT": 1478, + "TTGAAC": 1479, + "TTGAAG": 1480, + "TTGATA": 1481, + "TTGATT": 1482, + "TTGATC": 1483, + "TTGATG": 1484, + "TTGACA": 1485, + "TTGACT": 1486, + "TTGACC": 1487, + "TTGACG": 1488, + "TTGAGA": 1489, + "TTGAGT": 1490, + "TTGAGC": 1491, + "TTGAGG": 1492, + "TTGTAA": 1493, + "TTGTAT": 1494, + "TTGTAC": 1495, + "TTGTAG": 1496, + "TTGTTA": 1497, + "TTGTTT": 1498, + "TTGTTC": 1499, + "TTGTTG": 1500, + "TTGTCA": 1501, + "TTGTCT": 1502, + "TTGTCC": 1503, + "TTGTCG": 1504, + "TTGTGA": 1505, + "TTGTGT": 1506, + "TTGTGC": 1507, + "TTGTGG": 1508, + "TTGCAA": 1509, + "TTGCAT": 1510, + "TTGCAC": 1511, + "TTGCAG": 1512, + "TTGCTA": 1513, + "TTGCTT": 1514, + "TTGCTC": 1515, + "TTGCTG": 1516, + "TTGCCA": 1517, + "TTGCCT": 1518, + "TTGCCC": 1519, + "TTGCCG": 1520, + "TTGCGA": 1521, + "TTGCGT": 1522, + "TTGCGC": 1523, + "TTGCGG": 1524, + "TTGGAA": 1525, + "TTGGAT": 1526, + "TTGGAC": 1527, + "TTGGAG": 1528, + "TTGGTA": 1529, + "TTGGTT": 1530, + "TTGGTC": 1531, + "TTGGTG": 1532, + "TTGGCA": 1533, + "TTGGCT": 1534, + "TTGGCC": 1535, + "TTGGCG": 1536, + "TTGGGA": 1537, + "TTGGGT": 1538, + "TTGGGC": 1539, + "TTGGGG": 1540, + "TCAAAA": 1541, + "TCAAAT": 1542, + "TCAAAC": 1543, + "TCAAAG": 1544, + "TCAATA": 1545, + "TCAATT": 1546, + "TCAATC": 1547, + "TCAATG": 1548, + "TCAACA": 1549, + "TCAACT": 1550, + "TCAACC": 1551, + "TCAACG": 1552, + "TCAAGA": 1553, + "TCAAGT": 1554, + "TCAAGC": 1555, + "TCAAGG": 1556, + "TCATAA": 1557, + "TCATAT": 1558, + "TCATAC": 1559, + "TCATAG": 1560, + "TCATTA": 1561, + "TCATTT": 1562, + "TCATTC": 1563, + "TCATTG": 1564, + "TCATCA": 1565, + "TCATCT": 1566, + "TCATCC": 1567, + "TCATCG": 1568, + "TCATGA": 1569, + "TCATGT": 1570, + "TCATGC": 1571, + "TCATGG": 1572, + "TCACAA": 1573, + "TCACAT": 1574, + "TCACAC": 1575, + "TCACAG": 1576, + "TCACTA": 1577, + "TCACTT": 1578, + "TCACTC": 1579, + "TCACTG": 1580, + "TCACCA": 1581, + "TCACCT": 1582, + "TCACCC": 1583, + "TCACCG": 1584, + "TCACGA": 1585, + "TCACGT": 1586, + "TCACGC": 1587, + "TCACGG": 1588, + "TCAGAA": 1589, + "TCAGAT": 1590, + "TCAGAC": 1591, + "TCAGAG": 1592, + "TCAGTA": 1593, + "TCAGTT": 1594, + "TCAGTC": 1595, + "TCAGTG": 1596, + "TCAGCA": 1597, + "TCAGCT": 1598, + "TCAGCC": 1599, + "TCAGCG": 1600, + "TCAGGA": 1601, + "TCAGGT": 1602, + "TCAGGC": 1603, + "TCAGGG": 1604, + "TCTAAA": 1605, + "TCTAAT": 1606, + "TCTAAC": 1607, + "TCTAAG": 1608, + "TCTATA": 1609, + "TCTATT": 1610, + "TCTATC": 1611, + "TCTATG": 1612, + "TCTACA": 1613, + "TCTACT": 1614, + "TCTACC": 1615, + "TCTACG": 1616, + "TCTAGA": 1617, + "TCTAGT": 1618, + "TCTAGC": 1619, + "TCTAGG": 1620, + "TCTTAA": 1621, + "TCTTAT": 1622, + "TCTTAC": 1623, + "TCTTAG": 1624, + "TCTTTA": 1625, + "TCTTTT": 1626, + "TCTTTC": 1627, + "TCTTTG": 1628, + "TCTTCA": 1629, + "TCTTCT": 1630, + "TCTTCC": 1631, + "TCTTCG": 1632, + "TCTTGA": 1633, + "TCTTGT": 1634, + "TCTTGC": 1635, + "TCTTGG": 1636, + "TCTCAA": 1637, + "TCTCAT": 1638, + "TCTCAC": 1639, + "TCTCAG": 1640, + "TCTCTA": 1641, + "TCTCTT": 1642, + "TCTCTC": 1643, + "TCTCTG": 1644, + "TCTCCA": 1645, + "TCTCCT": 1646, + "TCTCCC": 1647, + "TCTCCG": 1648, + "TCTCGA": 1649, + "TCTCGT": 1650, + "TCTCGC": 1651, + "TCTCGG": 1652, + "TCTGAA": 1653, + "TCTGAT": 1654, + "TCTGAC": 1655, + "TCTGAG": 1656, + "TCTGTA": 1657, + "TCTGTT": 1658, + "TCTGTC": 1659, + "TCTGTG": 1660, + "TCTGCA": 1661, + "TCTGCT": 1662, + "TCTGCC": 1663, + "TCTGCG": 1664, + "TCTGGA": 1665, + "TCTGGT": 1666, + "TCTGGC": 1667, + "TCTGGG": 1668, + "TCCAAA": 1669, + "TCCAAT": 1670, + "TCCAAC": 1671, + "TCCAAG": 1672, + "TCCATA": 1673, + "TCCATT": 1674, + "TCCATC": 1675, + "TCCATG": 1676, + "TCCACA": 1677, + "TCCACT": 1678, + "TCCACC": 1679, + "TCCACG": 1680, + "TCCAGA": 1681, + "TCCAGT": 1682, + "TCCAGC": 1683, + "TCCAGG": 1684, + "TCCTAA": 1685, + "TCCTAT": 1686, + "TCCTAC": 1687, + "TCCTAG": 1688, + "TCCTTA": 1689, + "TCCTTT": 1690, + "TCCTTC": 1691, + "TCCTTG": 1692, + "TCCTCA": 1693, + "TCCTCT": 1694, + "TCCTCC": 1695, + "TCCTCG": 1696, + "TCCTGA": 1697, + "TCCTGT": 1698, + "TCCTGC": 1699, + "TCCTGG": 1700, + "TCCCAA": 1701, + "TCCCAT": 1702, + "TCCCAC": 1703, + "TCCCAG": 1704, + "TCCCTA": 1705, + "TCCCTT": 1706, + "TCCCTC": 1707, + "TCCCTG": 1708, + "TCCCCA": 1709, + "TCCCCT": 1710, + "TCCCCC": 1711, + "TCCCCG": 1712, + "TCCCGA": 1713, + "TCCCGT": 1714, + "TCCCGC": 1715, + "TCCCGG": 1716, + "TCCGAA": 1717, + "TCCGAT": 1718, + "TCCGAC": 1719, + "TCCGAG": 1720, + "TCCGTA": 1721, + "TCCGTT": 1722, + "TCCGTC": 1723, + "TCCGTG": 1724, + "TCCGCA": 1725, + "TCCGCT": 1726, + "TCCGCC": 1727, + "TCCGCG": 1728, + "TCCGGA": 1729, + "TCCGGT": 1730, + "TCCGGC": 1731, + "TCCGGG": 1732, + "TCGAAA": 1733, + "TCGAAT": 1734, + "TCGAAC": 1735, + "TCGAAG": 1736, + "TCGATA": 1737, + "TCGATT": 1738, + "TCGATC": 1739, + "TCGATG": 1740, + "TCGACA": 1741, + "TCGACT": 1742, + "TCGACC": 1743, + "TCGACG": 1744, + "TCGAGA": 1745, + "TCGAGT": 1746, + "TCGAGC": 1747, + "TCGAGG": 1748, + "TCGTAA": 1749, + "TCGTAT": 1750, + "TCGTAC": 1751, + "TCGTAG": 1752, + "TCGTTA": 1753, + "TCGTTT": 1754, + "TCGTTC": 1755, + "TCGTTG": 1756, + "TCGTCA": 1757, + "TCGTCT": 1758, + "TCGTCC": 1759, + "TCGTCG": 1760, + "TCGTGA": 1761, + "TCGTGT": 1762, + "TCGTGC": 1763, + "TCGTGG": 1764, + "TCGCAA": 1765, + "TCGCAT": 1766, + "TCGCAC": 1767, + "TCGCAG": 1768, + "TCGCTA": 1769, + "TCGCTT": 1770, + "TCGCTC": 1771, + "TCGCTG": 1772, + "TCGCCA": 1773, + "TCGCCT": 1774, + "TCGCCC": 1775, + "TCGCCG": 1776, + "TCGCGA": 1777, + "TCGCGT": 1778, + "TCGCGC": 1779, + "TCGCGG": 1780, + "TCGGAA": 1781, + "TCGGAT": 1782, + "TCGGAC": 1783, + "TCGGAG": 1784, + "TCGGTA": 1785, + "TCGGTT": 1786, + "TCGGTC": 1787, + "TCGGTG": 1788, + "TCGGCA": 1789, + "TCGGCT": 1790, + "TCGGCC": 1791, + "TCGGCG": 1792, + "TCGGGA": 1793, + "TCGGGT": 1794, + "TCGGGC": 1795, + "TCGGGG": 1796, + "TGAAAA": 1797, + "TGAAAT": 1798, + "TGAAAC": 1799, + "TGAAAG": 1800, + "TGAATA": 1801, + "TGAATT": 1802, + "TGAATC": 1803, + "TGAATG": 1804, + "TGAACA": 1805, + "TGAACT": 1806, + "TGAACC": 1807, + "TGAACG": 1808, + "TGAAGA": 1809, + "TGAAGT": 1810, + "TGAAGC": 1811, + "TGAAGG": 1812, + "TGATAA": 1813, + "TGATAT": 1814, + "TGATAC": 1815, + "TGATAG": 1816, + "TGATTA": 1817, + "TGATTT": 1818, + "TGATTC": 1819, + "TGATTG": 1820, + "TGATCA": 1821, + "TGATCT": 1822, + "TGATCC": 1823, + "TGATCG": 1824, + "TGATGA": 1825, + "TGATGT": 1826, + "TGATGC": 1827, + "TGATGG": 1828, + "TGACAA": 1829, + "TGACAT": 1830, + "TGACAC": 1831, + "TGACAG": 1832, + "TGACTA": 1833, + "TGACTT": 1834, + "TGACTC": 1835, + "TGACTG": 1836, + "TGACCA": 1837, + "TGACCT": 1838, + "TGACCC": 1839, + "TGACCG": 1840, + "TGACGA": 1841, + "TGACGT": 1842, + "TGACGC": 1843, + "TGACGG": 1844, + "TGAGAA": 1845, + "TGAGAT": 1846, + "TGAGAC": 1847, + "TGAGAG": 1848, + "TGAGTA": 1849, + "TGAGTT": 1850, + "TGAGTC": 1851, + "TGAGTG": 1852, + "TGAGCA": 1853, + "TGAGCT": 1854, + "TGAGCC": 1855, + "TGAGCG": 1856, + "TGAGGA": 1857, + "TGAGGT": 1858, + "TGAGGC": 1859, + "TGAGGG": 1860, + "TGTAAA": 1861, + "TGTAAT": 1862, + "TGTAAC": 1863, + "TGTAAG": 1864, + "TGTATA": 1865, + "TGTATT": 1866, + "TGTATC": 1867, + "TGTATG": 1868, + "TGTACA": 1869, + "TGTACT": 1870, + "TGTACC": 1871, + "TGTACG": 1872, + "TGTAGA": 1873, + "TGTAGT": 1874, + "TGTAGC": 1875, + "TGTAGG": 1876, + "TGTTAA": 1877, + "TGTTAT": 1878, + "TGTTAC": 1879, + "TGTTAG": 1880, + "TGTTTA": 1881, + "TGTTTT": 1882, + "TGTTTC": 1883, + "TGTTTG": 1884, + "TGTTCA": 1885, + "TGTTCT": 1886, + "TGTTCC": 1887, + "TGTTCG": 1888, + "TGTTGA": 1889, + "TGTTGT": 1890, + "TGTTGC": 1891, + "TGTTGG": 1892, + "TGTCAA": 1893, + "TGTCAT": 1894, + "TGTCAC": 1895, + "TGTCAG": 1896, + "TGTCTA": 1897, + "TGTCTT": 1898, + "TGTCTC": 1899, + "TGTCTG": 1900, + "TGTCCA": 1901, + "TGTCCT": 1902, + "TGTCCC": 1903, + "TGTCCG": 1904, + "TGTCGA": 1905, + "TGTCGT": 1906, + "TGTCGC": 1907, + "TGTCGG": 1908, + "TGTGAA": 1909, + "TGTGAT": 1910, + "TGTGAC": 1911, + "TGTGAG": 1912, + "TGTGTA": 1913, + "TGTGTT": 1914, + "TGTGTC": 1915, + "TGTGTG": 1916, + "TGTGCA": 1917, + "TGTGCT": 1918, + "TGTGCC": 1919, + "TGTGCG": 1920, + "TGTGGA": 1921, + "TGTGGT": 1922, + "TGTGGC": 1923, + "TGTGGG": 1924, + "TGCAAA": 1925, + "TGCAAT": 1926, + "TGCAAC": 1927, + "TGCAAG": 1928, + "TGCATA": 1929, + "TGCATT": 1930, + "TGCATC": 1931, + "TGCATG": 1932, + "TGCACA": 1933, + "TGCACT": 1934, + "TGCACC": 1935, + "TGCACG": 1936, + "TGCAGA": 1937, + "TGCAGT": 1938, + "TGCAGC": 1939, + "TGCAGG": 1940, + "TGCTAA": 1941, + "TGCTAT": 1942, + "TGCTAC": 1943, + "TGCTAG": 1944, + "TGCTTA": 1945, + "TGCTTT": 1946, + "TGCTTC": 1947, + "TGCTTG": 1948, + "TGCTCA": 1949, + "TGCTCT": 1950, + "TGCTCC": 1951, + "TGCTCG": 1952, + "TGCTGA": 1953, + "TGCTGT": 1954, + "TGCTGC": 1955, + "TGCTGG": 1956, + "TGCCAA": 1957, + "TGCCAT": 1958, + "TGCCAC": 1959, + "TGCCAG": 1960, + "TGCCTA": 1961, + "TGCCTT": 1962, + "TGCCTC": 1963, + "TGCCTG": 1964, + "TGCCCA": 1965, + "TGCCCT": 1966, + "TGCCCC": 1967, + "TGCCCG": 1968, + "TGCCGA": 1969, + "TGCCGT": 1970, + "TGCCGC": 1971, + "TGCCGG": 1972, + "TGCGAA": 1973, + "TGCGAT": 1974, + "TGCGAC": 1975, + "TGCGAG": 1976, + "TGCGTA": 1977, + "TGCGTT": 1978, + "TGCGTC": 1979, + "TGCGTG": 1980, + "TGCGCA": 1981, + "TGCGCT": 1982, + "TGCGCC": 1983, + "TGCGCG": 1984, + "TGCGGA": 1985, + "TGCGGT": 1986, + "TGCGGC": 1987, + "TGCGGG": 1988, + "TGGAAA": 1989, + "TGGAAT": 1990, + "TGGAAC": 1991, + "TGGAAG": 1992, + "TGGATA": 1993, + "TGGATT": 1994, + "TGGATC": 1995, + "TGGATG": 1996, + "TGGACA": 1997, + "TGGACT": 1998, + "TGGACC": 1999, + "TGGACG": 2000, + "TGGAGA": 2001, + "TGGAGT": 2002, + "TGGAGC": 2003, + "TGGAGG": 2004, + "TGGTAA": 2005, + "TGGTAT": 2006, + "TGGTAC": 2007, + "TGGTAG": 2008, + "TGGTTA": 2009, + "TGGTTT": 2010, + "TGGTTC": 2011, + "TGGTTG": 2012, + "TGGTCA": 2013, + "TGGTCT": 2014, + "TGGTCC": 2015, + "TGGTCG": 2016, + "TGGTGA": 2017, + "TGGTGT": 2018, + "TGGTGC": 2019, + "TGGTGG": 2020, + "TGGCAA": 2021, + "TGGCAT": 2022, + "TGGCAC": 2023, + "TGGCAG": 2024, + "TGGCTA": 2025, + "TGGCTT": 2026, + "TGGCTC": 2027, + "TGGCTG": 2028, + "TGGCCA": 2029, + "TGGCCT": 2030, + "TGGCCC": 2031, + "TGGCCG": 2032, + "TGGCGA": 2033, + "TGGCGT": 2034, + "TGGCGC": 2035, + "TGGCGG": 2036, + "TGGGAA": 2037, + "TGGGAT": 2038, + "TGGGAC": 2039, + "TGGGAG": 2040, + "TGGGTA": 2041, + "TGGGTT": 2042, + "TGGGTC": 2043, + "TGGGTG": 2044, + "TGGGCA": 2045, + "TGGGCT": 2046, + "TGGGCC": 2047, + "TGGGCG": 2048, + "TGGGGA": 2049, + "TGGGGT": 2050, + "TGGGGC": 2051, + "TGGGGG": 2052, + "CAAAAA": 2053, + "CAAAAT": 2054, + "CAAAAC": 2055, + "CAAAAG": 2056, + "CAAATA": 2057, + "CAAATT": 2058, + "CAAATC": 2059, + "CAAATG": 2060, + "CAAACA": 2061, + "CAAACT": 2062, + "CAAACC": 2063, + "CAAACG": 2064, + "CAAAGA": 2065, + "CAAAGT": 2066, + "CAAAGC": 2067, + "CAAAGG": 2068, + "CAATAA": 2069, + "CAATAT": 2070, + "CAATAC": 2071, + "CAATAG": 2072, + "CAATTA": 2073, + "CAATTT": 2074, + "CAATTC": 2075, + "CAATTG": 2076, + "CAATCA": 2077, + "CAATCT": 2078, + "CAATCC": 2079, + "CAATCG": 2080, + "CAATGA": 2081, + "CAATGT": 2082, + "CAATGC": 2083, + "CAATGG": 2084, + "CAACAA": 2085, + "CAACAT": 2086, + "CAACAC": 2087, + "CAACAG": 2088, + "CAACTA": 2089, + "CAACTT": 2090, + "CAACTC": 2091, + "CAACTG": 2092, + "CAACCA": 2093, + "CAACCT": 2094, + "CAACCC": 2095, + "CAACCG": 2096, + "CAACGA": 2097, + "CAACGT": 2098, + "CAACGC": 2099, + "CAACGG": 2100, + "CAAGAA": 2101, + "CAAGAT": 2102, + "CAAGAC": 2103, + "CAAGAG": 2104, + "CAAGTA": 2105, + "CAAGTT": 2106, + "CAAGTC": 2107, + "CAAGTG": 2108, + "CAAGCA": 2109, + "CAAGCT": 2110, + "CAAGCC": 2111, + "CAAGCG": 2112, + "CAAGGA": 2113, + "CAAGGT": 2114, + "CAAGGC": 2115, + "CAAGGG": 2116, + "CATAAA": 2117, + "CATAAT": 2118, + "CATAAC": 2119, + "CATAAG": 2120, + "CATATA": 2121, + "CATATT": 2122, + "CATATC": 2123, + "CATATG": 2124, + "CATACA": 2125, + "CATACT": 2126, + "CATACC": 2127, + "CATACG": 2128, + "CATAGA": 2129, + "CATAGT": 2130, + "CATAGC": 2131, + "CATAGG": 2132, + "CATTAA": 2133, + "CATTAT": 2134, + "CATTAC": 2135, + "CATTAG": 2136, + "CATTTA": 2137, + "CATTTT": 2138, + "CATTTC": 2139, + "CATTTG": 2140, + "CATTCA": 2141, + "CATTCT": 2142, + "CATTCC": 2143, + "CATTCG": 2144, + "CATTGA": 2145, + "CATTGT": 2146, + "CATTGC": 2147, + "CATTGG": 2148, + "CATCAA": 2149, + "CATCAT": 2150, + "CATCAC": 2151, + "CATCAG": 2152, + "CATCTA": 2153, + "CATCTT": 2154, + "CATCTC": 2155, + "CATCTG": 2156, + "CATCCA": 2157, + "CATCCT": 2158, + "CATCCC": 2159, + "CATCCG": 2160, + "CATCGA": 2161, + "CATCGT": 2162, + "CATCGC": 2163, + "CATCGG": 2164, + "CATGAA": 2165, + "CATGAT": 2166, + "CATGAC": 2167, + "CATGAG": 2168, + "CATGTA": 2169, + "CATGTT": 2170, + "CATGTC": 2171, + "CATGTG": 2172, + "CATGCA": 2173, + "CATGCT": 2174, + "CATGCC": 2175, + "CATGCG": 2176, + "CATGGA": 2177, + "CATGGT": 2178, + "CATGGC": 2179, + "CATGGG": 2180, + "CACAAA": 2181, + "CACAAT": 2182, + "CACAAC": 2183, + "CACAAG": 2184, + "CACATA": 2185, + "CACATT": 2186, + "CACATC": 2187, + "CACATG": 2188, + "CACACA": 2189, + "CACACT": 2190, + "CACACC": 2191, + "CACACG": 2192, + "CACAGA": 2193, + "CACAGT": 2194, + "CACAGC": 2195, + "CACAGG": 2196, + "CACTAA": 2197, + "CACTAT": 2198, + "CACTAC": 2199, + "CACTAG": 2200, + "CACTTA": 2201, + "CACTTT": 2202, + "CACTTC": 2203, + "CACTTG": 2204, + "CACTCA": 2205, + "CACTCT": 2206, + "CACTCC": 2207, + "CACTCG": 2208, + "CACTGA": 2209, + "CACTGT": 2210, + "CACTGC": 2211, + "CACTGG": 2212, + "CACCAA": 2213, + "CACCAT": 2214, + "CACCAC": 2215, + "CACCAG": 2216, + "CACCTA": 2217, + "CACCTT": 2218, + "CACCTC": 2219, + "CACCTG": 2220, + "CACCCA": 2221, + "CACCCT": 2222, + "CACCCC": 2223, + "CACCCG": 2224, + "CACCGA": 2225, + "CACCGT": 2226, + "CACCGC": 2227, + "CACCGG": 2228, + "CACGAA": 2229, + "CACGAT": 2230, + "CACGAC": 2231, + "CACGAG": 2232, + "CACGTA": 2233, + "CACGTT": 2234, + "CACGTC": 2235, + "CACGTG": 2236, + "CACGCA": 2237, + "CACGCT": 2238, + "CACGCC": 2239, + "CACGCG": 2240, + "CACGGA": 2241, + "CACGGT": 2242, + "CACGGC": 2243, + "CACGGG": 2244, + "CAGAAA": 2245, + "CAGAAT": 2246, + "CAGAAC": 2247, + "CAGAAG": 2248, + "CAGATA": 2249, + "CAGATT": 2250, + "CAGATC": 2251, + "CAGATG": 2252, + "CAGACA": 2253, + "CAGACT": 2254, + "CAGACC": 2255, + "CAGACG": 2256, + "CAGAGA": 2257, + "CAGAGT": 2258, + "CAGAGC": 2259, + "CAGAGG": 2260, + "CAGTAA": 2261, + "CAGTAT": 2262, + "CAGTAC": 2263, + "CAGTAG": 2264, + "CAGTTA": 2265, + "CAGTTT": 2266, + "CAGTTC": 2267, + "CAGTTG": 2268, + "CAGTCA": 2269, + "CAGTCT": 2270, + "CAGTCC": 2271, + "CAGTCG": 2272, + "CAGTGA": 2273, + "CAGTGT": 2274, + "CAGTGC": 2275, + "CAGTGG": 2276, + "CAGCAA": 2277, + "CAGCAT": 2278, + "CAGCAC": 2279, + "CAGCAG": 2280, + "CAGCTA": 2281, + "CAGCTT": 2282, + "CAGCTC": 2283, + "CAGCTG": 2284, + "CAGCCA": 2285, + "CAGCCT": 2286, + "CAGCCC": 2287, + "CAGCCG": 2288, + "CAGCGA": 2289, + "CAGCGT": 2290, + "CAGCGC": 2291, + "CAGCGG": 2292, + "CAGGAA": 2293, + "CAGGAT": 2294, + "CAGGAC": 2295, + "CAGGAG": 2296, + "CAGGTA": 2297, + "CAGGTT": 2298, + "CAGGTC": 2299, + "CAGGTG": 2300, + "CAGGCA": 2301, + "CAGGCT": 2302, + "CAGGCC": 2303, + "CAGGCG": 2304, + "CAGGGA": 2305, + "CAGGGT": 2306, + "CAGGGC": 2307, + "CAGGGG": 2308, + "CTAAAA": 2309, + "CTAAAT": 2310, + "CTAAAC": 2311, + "CTAAAG": 2312, + "CTAATA": 2313, + "CTAATT": 2314, + "CTAATC": 2315, + "CTAATG": 2316, + "CTAACA": 2317, + "CTAACT": 2318, + "CTAACC": 2319, + "CTAACG": 2320, + "CTAAGA": 2321, + "CTAAGT": 2322, + "CTAAGC": 2323, + "CTAAGG": 2324, + "CTATAA": 2325, + "CTATAT": 2326, + "CTATAC": 2327, + "CTATAG": 2328, + "CTATTA": 2329, + "CTATTT": 2330, + "CTATTC": 2331, + "CTATTG": 2332, + "CTATCA": 2333, + "CTATCT": 2334, + "CTATCC": 2335, + "CTATCG": 2336, + "CTATGA": 2337, + "CTATGT": 2338, + "CTATGC": 2339, + "CTATGG": 2340, + "CTACAA": 2341, + "CTACAT": 2342, + "CTACAC": 2343, + "CTACAG": 2344, + "CTACTA": 2345, + "CTACTT": 2346, + "CTACTC": 2347, + "CTACTG": 2348, + "CTACCA": 2349, + "CTACCT": 2350, + "CTACCC": 2351, + "CTACCG": 2352, + "CTACGA": 2353, + "CTACGT": 2354, + "CTACGC": 2355, + "CTACGG": 2356, + "CTAGAA": 2357, + "CTAGAT": 2358, + "CTAGAC": 2359, + "CTAGAG": 2360, + "CTAGTA": 2361, + "CTAGTT": 2362, + "CTAGTC": 2363, + "CTAGTG": 2364, + "CTAGCA": 2365, + "CTAGCT": 2366, + "CTAGCC": 2367, + "CTAGCG": 2368, + "CTAGGA": 2369, + "CTAGGT": 2370, + "CTAGGC": 2371, + "CTAGGG": 2372, + "CTTAAA": 2373, + "CTTAAT": 2374, + "CTTAAC": 2375, + "CTTAAG": 2376, + "CTTATA": 2377, + "CTTATT": 2378, + "CTTATC": 2379, + "CTTATG": 2380, + "CTTACA": 2381, + "CTTACT": 2382, + "CTTACC": 2383, + "CTTACG": 2384, + "CTTAGA": 2385, + "CTTAGT": 2386, + "CTTAGC": 2387, + "CTTAGG": 2388, + "CTTTAA": 2389, + "CTTTAT": 2390, + "CTTTAC": 2391, + "CTTTAG": 2392, + "CTTTTA": 2393, + "CTTTTT": 2394, + "CTTTTC": 2395, + "CTTTTG": 2396, + "CTTTCA": 2397, + "CTTTCT": 2398, + "CTTTCC": 2399, + "CTTTCG": 2400, + "CTTTGA": 2401, + "CTTTGT": 2402, + "CTTTGC": 2403, + "CTTTGG": 2404, + "CTTCAA": 2405, + "CTTCAT": 2406, + "CTTCAC": 2407, + "CTTCAG": 2408, + "CTTCTA": 2409, + "CTTCTT": 2410, + "CTTCTC": 2411, + "CTTCTG": 2412, + "CTTCCA": 2413, + "CTTCCT": 2414, + "CTTCCC": 2415, + "CTTCCG": 2416, + "CTTCGA": 2417, + "CTTCGT": 2418, + "CTTCGC": 2419, + "CTTCGG": 2420, + "CTTGAA": 2421, + "CTTGAT": 2422, + "CTTGAC": 2423, + "CTTGAG": 2424, + "CTTGTA": 2425, + "CTTGTT": 2426, + "CTTGTC": 2427, + "CTTGTG": 2428, + "CTTGCA": 2429, + "CTTGCT": 2430, + "CTTGCC": 2431, + "CTTGCG": 2432, + "CTTGGA": 2433, + "CTTGGT": 2434, + "CTTGGC": 2435, + "CTTGGG": 2436, + "CTCAAA": 2437, + "CTCAAT": 2438, + "CTCAAC": 2439, + "CTCAAG": 2440, + "CTCATA": 2441, + "CTCATT": 2442, + "CTCATC": 2443, + "CTCATG": 2444, + "CTCACA": 2445, + "CTCACT": 2446, + "CTCACC": 2447, + "CTCACG": 2448, + "CTCAGA": 2449, + "CTCAGT": 2450, + "CTCAGC": 2451, + "CTCAGG": 2452, + "CTCTAA": 2453, + "CTCTAT": 2454, + "CTCTAC": 2455, + "CTCTAG": 2456, + "CTCTTA": 2457, + "CTCTTT": 2458, + "CTCTTC": 2459, + "CTCTTG": 2460, + "CTCTCA": 2461, + "CTCTCT": 2462, + "CTCTCC": 2463, + "CTCTCG": 2464, + "CTCTGA": 2465, + "CTCTGT": 2466, + "CTCTGC": 2467, + "CTCTGG": 2468, + "CTCCAA": 2469, + "CTCCAT": 2470, + "CTCCAC": 2471, + "CTCCAG": 2472, + "CTCCTA": 2473, + "CTCCTT": 2474, + "CTCCTC": 2475, + "CTCCTG": 2476, + "CTCCCA": 2477, + "CTCCCT": 2478, + "CTCCCC": 2479, + "CTCCCG": 2480, + "CTCCGA": 2481, + "CTCCGT": 2482, + "CTCCGC": 2483, + "CTCCGG": 2484, + "CTCGAA": 2485, + "CTCGAT": 2486, + "CTCGAC": 2487, + "CTCGAG": 2488, + "CTCGTA": 2489, + "CTCGTT": 2490, + "CTCGTC": 2491, + "CTCGTG": 2492, + "CTCGCA": 2493, + "CTCGCT": 2494, + "CTCGCC": 2495, + "CTCGCG": 2496, + "CTCGGA": 2497, + "CTCGGT": 2498, + "CTCGGC": 2499, + "CTCGGG": 2500, + "CTGAAA": 2501, + "CTGAAT": 2502, + "CTGAAC": 2503, + "CTGAAG": 2504, + "CTGATA": 2505, + "CTGATT": 2506, + "CTGATC": 2507, + "CTGATG": 2508, + "CTGACA": 2509, + "CTGACT": 2510, + "CTGACC": 2511, + "CTGACG": 2512, + "CTGAGA": 2513, + "CTGAGT": 2514, + "CTGAGC": 2515, + "CTGAGG": 2516, + "CTGTAA": 2517, + "CTGTAT": 2518, + "CTGTAC": 2519, + "CTGTAG": 2520, + "CTGTTA": 2521, + "CTGTTT": 2522, + "CTGTTC": 2523, + "CTGTTG": 2524, + "CTGTCA": 2525, + "CTGTCT": 2526, + "CTGTCC": 2527, + "CTGTCG": 2528, + "CTGTGA": 2529, + "CTGTGT": 2530, + "CTGTGC": 2531, + "CTGTGG": 2532, + "CTGCAA": 2533, + "CTGCAT": 2534, + "CTGCAC": 2535, + "CTGCAG": 2536, + "CTGCTA": 2537, + "CTGCTT": 2538, + "CTGCTC": 2539, + "CTGCTG": 2540, + "CTGCCA": 2541, + "CTGCCT": 2542, + "CTGCCC": 2543, + "CTGCCG": 2544, + "CTGCGA": 2545, + "CTGCGT": 2546, + "CTGCGC": 2547, + "CTGCGG": 2548, + "CTGGAA": 2549, + "CTGGAT": 2550, + "CTGGAC": 2551, + "CTGGAG": 2552, + "CTGGTA": 2553, + "CTGGTT": 2554, + "CTGGTC": 2555, + "CTGGTG": 2556, + "CTGGCA": 2557, + "CTGGCT": 2558, + "CTGGCC": 2559, + "CTGGCG": 2560, + "CTGGGA": 2561, + "CTGGGT": 2562, + "CTGGGC": 2563, + "CTGGGG": 2564, + "CCAAAA": 2565, + "CCAAAT": 2566, + "CCAAAC": 2567, + "CCAAAG": 2568, + "CCAATA": 2569, + "CCAATT": 2570, + "CCAATC": 2571, + "CCAATG": 2572, + "CCAACA": 2573, + "CCAACT": 2574, + "CCAACC": 2575, + "CCAACG": 2576, + "CCAAGA": 2577, + "CCAAGT": 2578, + "CCAAGC": 2579, + "CCAAGG": 2580, + "CCATAA": 2581, + "CCATAT": 2582, + "CCATAC": 2583, + "CCATAG": 2584, + "CCATTA": 2585, + "CCATTT": 2586, + "CCATTC": 2587, + "CCATTG": 2588, + "CCATCA": 2589, + "CCATCT": 2590, + "CCATCC": 2591, + "CCATCG": 2592, + "CCATGA": 2593, + "CCATGT": 2594, + "CCATGC": 2595, + "CCATGG": 2596, + "CCACAA": 2597, + "CCACAT": 2598, + "CCACAC": 2599, + "CCACAG": 2600, + "CCACTA": 2601, + "CCACTT": 2602, + "CCACTC": 2603, + "CCACTG": 2604, + "CCACCA": 2605, + "CCACCT": 2606, + "CCACCC": 2607, + "CCACCG": 2608, + "CCACGA": 2609, + "CCACGT": 2610, + "CCACGC": 2611, + "CCACGG": 2612, + "CCAGAA": 2613, + "CCAGAT": 2614, + "CCAGAC": 2615, + "CCAGAG": 2616, + "CCAGTA": 2617, + "CCAGTT": 2618, + "CCAGTC": 2619, + "CCAGTG": 2620, + "CCAGCA": 2621, + "CCAGCT": 2622, + "CCAGCC": 2623, + "CCAGCG": 2624, + "CCAGGA": 2625, + "CCAGGT": 2626, + "CCAGGC": 2627, + "CCAGGG": 2628, + "CCTAAA": 2629, + "CCTAAT": 2630, + "CCTAAC": 2631, + "CCTAAG": 2632, + "CCTATA": 2633, + "CCTATT": 2634, + "CCTATC": 2635, + "CCTATG": 2636, + "CCTACA": 2637, + "CCTACT": 2638, + "CCTACC": 2639, + "CCTACG": 2640, + "CCTAGA": 2641, + "CCTAGT": 2642, + "CCTAGC": 2643, + "CCTAGG": 2644, + "CCTTAA": 2645, + "CCTTAT": 2646, + "CCTTAC": 2647, + "CCTTAG": 2648, + "CCTTTA": 2649, + "CCTTTT": 2650, + "CCTTTC": 2651, + "CCTTTG": 2652, + "CCTTCA": 2653, + "CCTTCT": 2654, + "CCTTCC": 2655, + "CCTTCG": 2656, + "CCTTGA": 2657, + "CCTTGT": 2658, + "CCTTGC": 2659, + "CCTTGG": 2660, + "CCTCAA": 2661, + "CCTCAT": 2662, + "CCTCAC": 2663, + "CCTCAG": 2664, + "CCTCTA": 2665, + "CCTCTT": 2666, + "CCTCTC": 2667, + "CCTCTG": 2668, + "CCTCCA": 2669, + "CCTCCT": 2670, + "CCTCCC": 2671, + "CCTCCG": 2672, + "CCTCGA": 2673, + "CCTCGT": 2674, + "CCTCGC": 2675, + "CCTCGG": 2676, + "CCTGAA": 2677, + "CCTGAT": 2678, + "CCTGAC": 2679, + "CCTGAG": 2680, + "CCTGTA": 2681, + "CCTGTT": 2682, + "CCTGTC": 2683, + "CCTGTG": 2684, + "CCTGCA": 2685, + "CCTGCT": 2686, + "CCTGCC": 2687, + "CCTGCG": 2688, + "CCTGGA": 2689, + "CCTGGT": 2690, + "CCTGGC": 2691, + "CCTGGG": 2692, + "CCCAAA": 2693, + "CCCAAT": 2694, + "CCCAAC": 2695, + "CCCAAG": 2696, + "CCCATA": 2697, + "CCCATT": 2698, + "CCCATC": 2699, + "CCCATG": 2700, + "CCCACA": 2701, + "CCCACT": 2702, + "CCCACC": 2703, + "CCCACG": 2704, + "CCCAGA": 2705, + "CCCAGT": 2706, + "CCCAGC": 2707, + "CCCAGG": 2708, + "CCCTAA": 2709, + "CCCTAT": 2710, + "CCCTAC": 2711, + "CCCTAG": 2712, + "CCCTTA": 2713, + "CCCTTT": 2714, + "CCCTTC": 2715, + "CCCTTG": 2716, + "CCCTCA": 2717, + "CCCTCT": 2718, + "CCCTCC": 2719, + "CCCTCG": 2720, + "CCCTGA": 2721, + "CCCTGT": 2722, + "CCCTGC": 2723, + "CCCTGG": 2724, + "CCCCAA": 2725, + "CCCCAT": 2726, + "CCCCAC": 2727, + "CCCCAG": 2728, + "CCCCTA": 2729, + "CCCCTT": 2730, + "CCCCTC": 2731, + "CCCCTG": 2732, + "CCCCCA": 2733, + "CCCCCT": 2734, + "CCCCCC": 2735, + "CCCCCG": 2736, + "CCCCGA": 2737, + "CCCCGT": 2738, + "CCCCGC": 2739, + "CCCCGG": 2740, + "CCCGAA": 2741, + "CCCGAT": 2742, + "CCCGAC": 2743, + "CCCGAG": 2744, + "CCCGTA": 2745, + "CCCGTT": 2746, + "CCCGTC": 2747, + "CCCGTG": 2748, + "CCCGCA": 2749, + "CCCGCT": 2750, + "CCCGCC": 2751, + "CCCGCG": 2752, + "CCCGGA": 2753, + "CCCGGT": 2754, + "CCCGGC": 2755, + "CCCGGG": 2756, + "CCGAAA": 2757, + "CCGAAT": 2758, + "CCGAAC": 2759, + "CCGAAG": 2760, + "CCGATA": 2761, + "CCGATT": 2762, + "CCGATC": 2763, + "CCGATG": 2764, + "CCGACA": 2765, + "CCGACT": 2766, + "CCGACC": 2767, + "CCGACG": 2768, + "CCGAGA": 2769, + "CCGAGT": 2770, + "CCGAGC": 2771, + "CCGAGG": 2772, + "CCGTAA": 2773, + "CCGTAT": 2774, + "CCGTAC": 2775, + "CCGTAG": 2776, + "CCGTTA": 2777, + "CCGTTT": 2778, + "CCGTTC": 2779, + "CCGTTG": 2780, + "CCGTCA": 2781, + "CCGTCT": 2782, + "CCGTCC": 2783, + "CCGTCG": 2784, + "CCGTGA": 2785, + "CCGTGT": 2786, + "CCGTGC": 2787, + "CCGTGG": 2788, + "CCGCAA": 2789, + "CCGCAT": 2790, + "CCGCAC": 2791, + "CCGCAG": 2792, + "CCGCTA": 2793, + "CCGCTT": 2794, + "CCGCTC": 2795, + "CCGCTG": 2796, + "CCGCCA": 2797, + "CCGCCT": 2798, + "CCGCCC": 2799, + "CCGCCG": 2800, + "CCGCGA": 2801, + "CCGCGT": 2802, + "CCGCGC": 2803, + "CCGCGG": 2804, + "CCGGAA": 2805, + "CCGGAT": 2806, + "CCGGAC": 2807, + "CCGGAG": 2808, + "CCGGTA": 2809, + "CCGGTT": 2810, + "CCGGTC": 2811, + "CCGGTG": 2812, + "CCGGCA": 2813, + "CCGGCT": 2814, + "CCGGCC": 2815, + "CCGGCG": 2816, + "CCGGGA": 2817, + "CCGGGT": 2818, + "CCGGGC": 2819, + "CCGGGG": 2820, + "CGAAAA": 2821, + "CGAAAT": 2822, + "CGAAAC": 2823, + "CGAAAG": 2824, + "CGAATA": 2825, + "CGAATT": 2826, + "CGAATC": 2827, + "CGAATG": 2828, + "CGAACA": 2829, + "CGAACT": 2830, + "CGAACC": 2831, + "CGAACG": 2832, + "CGAAGA": 2833, + "CGAAGT": 2834, + "CGAAGC": 2835, + "CGAAGG": 2836, + "CGATAA": 2837, + "CGATAT": 2838, + "CGATAC": 2839, + "CGATAG": 2840, + "CGATTA": 2841, + "CGATTT": 2842, + "CGATTC": 2843, + "CGATTG": 2844, + "CGATCA": 2845, + "CGATCT": 2846, + "CGATCC": 2847, + "CGATCG": 2848, + "CGATGA": 2849, + "CGATGT": 2850, + "CGATGC": 2851, + "CGATGG": 2852, + "CGACAA": 2853, + "CGACAT": 2854, + "CGACAC": 2855, + "CGACAG": 2856, + "CGACTA": 2857, + "CGACTT": 2858, + "CGACTC": 2859, + "CGACTG": 2860, + "CGACCA": 2861, + "CGACCT": 2862, + "CGACCC": 2863, + "CGACCG": 2864, + "CGACGA": 2865, + "CGACGT": 2866, + "CGACGC": 2867, + "CGACGG": 2868, + "CGAGAA": 2869, + "CGAGAT": 2870, + "CGAGAC": 2871, + "CGAGAG": 2872, + "CGAGTA": 2873, + "CGAGTT": 2874, + "CGAGTC": 2875, + "CGAGTG": 2876, + "CGAGCA": 2877, + "CGAGCT": 2878, + "CGAGCC": 2879, + "CGAGCG": 2880, + "CGAGGA": 2881, + "CGAGGT": 2882, + "CGAGGC": 2883, + "CGAGGG": 2884, + "CGTAAA": 2885, + "CGTAAT": 2886, + "CGTAAC": 2887, + "CGTAAG": 2888, + "CGTATA": 2889, + "CGTATT": 2890, + "CGTATC": 2891, + "CGTATG": 2892, + "CGTACA": 2893, + "CGTACT": 2894, + "CGTACC": 2895, + "CGTACG": 2896, + "CGTAGA": 2897, + "CGTAGT": 2898, + "CGTAGC": 2899, + "CGTAGG": 2900, + "CGTTAA": 2901, + "CGTTAT": 2902, + "CGTTAC": 2903, + "CGTTAG": 2904, + "CGTTTA": 2905, + "CGTTTT": 2906, + "CGTTTC": 2907, + "CGTTTG": 2908, + "CGTTCA": 2909, + "CGTTCT": 2910, + "CGTTCC": 2911, + "CGTTCG": 2912, + "CGTTGA": 2913, + "CGTTGT": 2914, + "CGTTGC": 2915, + "CGTTGG": 2916, + "CGTCAA": 2917, + "CGTCAT": 2918, + "CGTCAC": 2919, + "CGTCAG": 2920, + "CGTCTA": 2921, + "CGTCTT": 2922, + "CGTCTC": 2923, + "CGTCTG": 2924, + "CGTCCA": 2925, + "CGTCCT": 2926, + "CGTCCC": 2927, + "CGTCCG": 2928, + "CGTCGA": 2929, + "CGTCGT": 2930, + "CGTCGC": 2931, + "CGTCGG": 2932, + "CGTGAA": 2933, + "CGTGAT": 2934, + "CGTGAC": 2935, + "CGTGAG": 2936, + "CGTGTA": 2937, + "CGTGTT": 2938, + "CGTGTC": 2939, + "CGTGTG": 2940, + "CGTGCA": 2941, + "CGTGCT": 2942, + "CGTGCC": 2943, + "CGTGCG": 2944, + "CGTGGA": 2945, + "CGTGGT": 2946, + "CGTGGC": 2947, + "CGTGGG": 2948, + "CGCAAA": 2949, + "CGCAAT": 2950, + "CGCAAC": 2951, + "CGCAAG": 2952, + "CGCATA": 2953, + "CGCATT": 2954, + "CGCATC": 2955, + "CGCATG": 2956, + "CGCACA": 2957, + "CGCACT": 2958, + "CGCACC": 2959, + "CGCACG": 2960, + "CGCAGA": 2961, + "CGCAGT": 2962, + "CGCAGC": 2963, + "CGCAGG": 2964, + "CGCTAA": 2965, + "CGCTAT": 2966, + "CGCTAC": 2967, + "CGCTAG": 2968, + "CGCTTA": 2969, + "CGCTTT": 2970, + "CGCTTC": 2971, + "CGCTTG": 2972, + "CGCTCA": 2973, + "CGCTCT": 2974, + "CGCTCC": 2975, + "CGCTCG": 2976, + "CGCTGA": 2977, + "CGCTGT": 2978, + "CGCTGC": 2979, + "CGCTGG": 2980, + "CGCCAA": 2981, + "CGCCAT": 2982, + "CGCCAC": 2983, + "CGCCAG": 2984, + "CGCCTA": 2985, + "CGCCTT": 2986, + "CGCCTC": 2987, + "CGCCTG": 2988, + "CGCCCA": 2989, + "CGCCCT": 2990, + "CGCCCC": 2991, + "CGCCCG": 2992, + "CGCCGA": 2993, + "CGCCGT": 2994, + "CGCCGC": 2995, + "CGCCGG": 2996, + "CGCGAA": 2997, + "CGCGAT": 2998, + "CGCGAC": 2999, + "CGCGAG": 3000, + "CGCGTA": 3001, + "CGCGTT": 3002, + "CGCGTC": 3003, + "CGCGTG": 3004, + "CGCGCA": 3005, + "CGCGCT": 3006, + "CGCGCC": 3007, + "CGCGCG": 3008, + "CGCGGA": 3009, + "CGCGGT": 3010, + "CGCGGC": 3011, + "CGCGGG": 3012, + "CGGAAA": 3013, + "CGGAAT": 3014, + "CGGAAC": 3015, + "CGGAAG": 3016, + "CGGATA": 3017, + "CGGATT": 3018, + "CGGATC": 3019, + "CGGATG": 3020, + "CGGACA": 3021, + "CGGACT": 3022, + "CGGACC": 3023, + "CGGACG": 3024, + "CGGAGA": 3025, + "CGGAGT": 3026, + "CGGAGC": 3027, + "CGGAGG": 3028, + "CGGTAA": 3029, + "CGGTAT": 3030, + "CGGTAC": 3031, + "CGGTAG": 3032, + "CGGTTA": 3033, + "CGGTTT": 3034, + "CGGTTC": 3035, + "CGGTTG": 3036, + "CGGTCA": 3037, + "CGGTCT": 3038, + "CGGTCC": 3039, + "CGGTCG": 3040, + "CGGTGA": 3041, + "CGGTGT": 3042, + "CGGTGC": 3043, + "CGGTGG": 3044, + "CGGCAA": 3045, + "CGGCAT": 3046, + "CGGCAC": 3047, + "CGGCAG": 3048, + "CGGCTA": 3049, + "CGGCTT": 3050, + "CGGCTC": 3051, + "CGGCTG": 3052, + "CGGCCA": 3053, + "CGGCCT": 3054, + "CGGCCC": 3055, + "CGGCCG": 3056, + "CGGCGA": 3057, + "CGGCGT": 3058, + "CGGCGC": 3059, + "CGGCGG": 3060, + "CGGGAA": 3061, + "CGGGAT": 3062, + "CGGGAC": 3063, + "CGGGAG": 3064, + "CGGGTA": 3065, + "CGGGTT": 3066, + "CGGGTC": 3067, + "CGGGTG": 3068, + "CGGGCA": 3069, + "CGGGCT": 3070, + "CGGGCC": 3071, + "CGGGCG": 3072, + "CGGGGA": 3073, + "CGGGGT": 3074, + "CGGGGC": 3075, + "CGGGGG": 3076, + "GAAAAA": 3077, + "GAAAAT": 3078, + "GAAAAC": 3079, + "GAAAAG": 3080, + "GAAATA": 3081, + "GAAATT": 3082, + "GAAATC": 3083, + "GAAATG": 3084, + "GAAACA": 3085, + "GAAACT": 3086, + "GAAACC": 3087, + "GAAACG": 3088, + "GAAAGA": 3089, + "GAAAGT": 3090, + "GAAAGC": 3091, + "GAAAGG": 3092, + "GAATAA": 3093, + "GAATAT": 3094, + "GAATAC": 3095, + "GAATAG": 3096, + "GAATTA": 3097, + "GAATTT": 3098, + "GAATTC": 3099, + "GAATTG": 3100, + "GAATCA": 3101, + "GAATCT": 3102, + "GAATCC": 3103, + "GAATCG": 3104, + "GAATGA": 3105, + "GAATGT": 3106, + "GAATGC": 3107, + "GAATGG": 3108, + "GAACAA": 3109, + "GAACAT": 3110, + "GAACAC": 3111, + "GAACAG": 3112, + "GAACTA": 3113, + "GAACTT": 3114, + "GAACTC": 3115, + "GAACTG": 3116, + "GAACCA": 3117, + "GAACCT": 3118, + "GAACCC": 3119, + "GAACCG": 3120, + "GAACGA": 3121, + "GAACGT": 3122, + "GAACGC": 3123, + "GAACGG": 3124, + "GAAGAA": 3125, + "GAAGAT": 3126, + "GAAGAC": 3127, + "GAAGAG": 3128, + "GAAGTA": 3129, + "GAAGTT": 3130, + "GAAGTC": 3131, + "GAAGTG": 3132, + "GAAGCA": 3133, + "GAAGCT": 3134, + "GAAGCC": 3135, + "GAAGCG": 3136, + "GAAGGA": 3137, + "GAAGGT": 3138, + "GAAGGC": 3139, + "GAAGGG": 3140, + "GATAAA": 3141, + "GATAAT": 3142, + "GATAAC": 3143, + "GATAAG": 3144, + "GATATA": 3145, + "GATATT": 3146, + "GATATC": 3147, + "GATATG": 3148, + "GATACA": 3149, + "GATACT": 3150, + "GATACC": 3151, + "GATACG": 3152, + "GATAGA": 3153, + "GATAGT": 3154, + "GATAGC": 3155, + "GATAGG": 3156, + "GATTAA": 3157, + "GATTAT": 3158, + "GATTAC": 3159, + "GATTAG": 3160, + "GATTTA": 3161, + "GATTTT": 3162, + "GATTTC": 3163, + "GATTTG": 3164, + "GATTCA": 3165, + "GATTCT": 3166, + "GATTCC": 3167, + "GATTCG": 3168, + "GATTGA": 3169, + "GATTGT": 3170, + "GATTGC": 3171, + "GATTGG": 3172, + "GATCAA": 3173, + "GATCAT": 3174, + "GATCAC": 3175, + "GATCAG": 3176, + "GATCTA": 3177, + "GATCTT": 3178, + "GATCTC": 3179, + "GATCTG": 3180, + "GATCCA": 3181, + "GATCCT": 3182, + "GATCCC": 3183, + "GATCCG": 3184, + "GATCGA": 3185, + "GATCGT": 3186, + "GATCGC": 3187, + "GATCGG": 3188, + "GATGAA": 3189, + "GATGAT": 3190, + "GATGAC": 3191, + "GATGAG": 3192, + "GATGTA": 3193, + "GATGTT": 3194, + "GATGTC": 3195, + "GATGTG": 3196, + "GATGCA": 3197, + "GATGCT": 3198, + "GATGCC": 3199, + "GATGCG": 3200, + "GATGGA": 3201, + "GATGGT": 3202, + "GATGGC": 3203, + "GATGGG": 3204, + "GACAAA": 3205, + "GACAAT": 3206, + "GACAAC": 3207, + "GACAAG": 3208, + "GACATA": 3209, + "GACATT": 3210, + "GACATC": 3211, + "GACATG": 3212, + "GACACA": 3213, + "GACACT": 3214, + "GACACC": 3215, + "GACACG": 3216, + "GACAGA": 3217, + "GACAGT": 3218, + "GACAGC": 3219, + "GACAGG": 3220, + "GACTAA": 3221, + "GACTAT": 3222, + "GACTAC": 3223, + "GACTAG": 3224, + "GACTTA": 3225, + "GACTTT": 3226, + "GACTTC": 3227, + "GACTTG": 3228, + "GACTCA": 3229, + "GACTCT": 3230, + "GACTCC": 3231, + "GACTCG": 3232, + "GACTGA": 3233, + "GACTGT": 3234, + "GACTGC": 3235, + "GACTGG": 3236, + "GACCAA": 3237, + "GACCAT": 3238, + "GACCAC": 3239, + "GACCAG": 3240, + "GACCTA": 3241, + "GACCTT": 3242, + "GACCTC": 3243, + "GACCTG": 3244, + "GACCCA": 3245, + "GACCCT": 3246, + "GACCCC": 3247, + "GACCCG": 3248, + "GACCGA": 3249, + "GACCGT": 3250, + "GACCGC": 3251, + "GACCGG": 3252, + "GACGAA": 3253, + "GACGAT": 3254, + "GACGAC": 3255, + "GACGAG": 3256, + "GACGTA": 3257, + "GACGTT": 3258, + "GACGTC": 3259, + "GACGTG": 3260, + "GACGCA": 3261, + "GACGCT": 3262, + "GACGCC": 3263, + "GACGCG": 3264, + "GACGGA": 3265, + "GACGGT": 3266, + "GACGGC": 3267, + "GACGGG": 3268, + "GAGAAA": 3269, + "GAGAAT": 3270, + "GAGAAC": 3271, + "GAGAAG": 3272, + "GAGATA": 3273, + "GAGATT": 3274, + "GAGATC": 3275, + "GAGATG": 3276, + "GAGACA": 3277, + "GAGACT": 3278, + "GAGACC": 3279, + "GAGACG": 3280, + "GAGAGA": 3281, + "GAGAGT": 3282, + "GAGAGC": 3283, + "GAGAGG": 3284, + "GAGTAA": 3285, + "GAGTAT": 3286, + "GAGTAC": 3287, + "GAGTAG": 3288, + "GAGTTA": 3289, + "GAGTTT": 3290, + "GAGTTC": 3291, + "GAGTTG": 3292, + "GAGTCA": 3293, + "GAGTCT": 3294, + "GAGTCC": 3295, + "GAGTCG": 3296, + "GAGTGA": 3297, + "GAGTGT": 3298, + "GAGTGC": 3299, + "GAGTGG": 3300, + "GAGCAA": 3301, + "GAGCAT": 3302, + "GAGCAC": 3303, + "GAGCAG": 3304, + "GAGCTA": 3305, + "GAGCTT": 3306, + "GAGCTC": 3307, + "GAGCTG": 3308, + "GAGCCA": 3309, + "GAGCCT": 3310, + "GAGCCC": 3311, + "GAGCCG": 3312, + "GAGCGA": 3313, + "GAGCGT": 3314, + "GAGCGC": 3315, + "GAGCGG": 3316, + "GAGGAA": 3317, + "GAGGAT": 3318, + "GAGGAC": 3319, + "GAGGAG": 3320, + "GAGGTA": 3321, + "GAGGTT": 3322, + "GAGGTC": 3323, + "GAGGTG": 3324, + "GAGGCA": 3325, + "GAGGCT": 3326, + "GAGGCC": 3327, + "GAGGCG": 3328, + "GAGGGA": 3329, + "GAGGGT": 3330, + "GAGGGC": 3331, + "GAGGGG": 3332, + "GTAAAA": 3333, + "GTAAAT": 3334, + "GTAAAC": 3335, + "GTAAAG": 3336, + "GTAATA": 3337, + "GTAATT": 3338, + "GTAATC": 3339, + "GTAATG": 3340, + "GTAACA": 3341, + "GTAACT": 3342, + "GTAACC": 3343, + "GTAACG": 3344, + "GTAAGA": 3345, + "GTAAGT": 3346, + "GTAAGC": 3347, + "GTAAGG": 3348, + "GTATAA": 3349, + "GTATAT": 3350, + "GTATAC": 3351, + "GTATAG": 3352, + "GTATTA": 3353, + "GTATTT": 3354, + "GTATTC": 3355, + "GTATTG": 3356, + "GTATCA": 3357, + "GTATCT": 3358, + "GTATCC": 3359, + "GTATCG": 3360, + "GTATGA": 3361, + "GTATGT": 3362, + "GTATGC": 3363, + "GTATGG": 3364, + "GTACAA": 3365, + "GTACAT": 3366, + "GTACAC": 3367, + "GTACAG": 3368, + "GTACTA": 3369, + "GTACTT": 3370, + "GTACTC": 3371, + "GTACTG": 3372, + "GTACCA": 3373, + "GTACCT": 3374, + "GTACCC": 3375, + "GTACCG": 3376, + "GTACGA": 3377, + "GTACGT": 3378, + "GTACGC": 3379, + "GTACGG": 3380, + "GTAGAA": 3381, + "GTAGAT": 3382, + "GTAGAC": 3383, + "GTAGAG": 3384, + "GTAGTA": 3385, + "GTAGTT": 3386, + "GTAGTC": 3387, + "GTAGTG": 3388, + "GTAGCA": 3389, + "GTAGCT": 3390, + "GTAGCC": 3391, + "GTAGCG": 3392, + "GTAGGA": 3393, + "GTAGGT": 3394, + "GTAGGC": 3395, + "GTAGGG": 3396, + "GTTAAA": 3397, + "GTTAAT": 3398, + "GTTAAC": 3399, + "GTTAAG": 3400, + "GTTATA": 3401, + "GTTATT": 3402, + "GTTATC": 3403, + "GTTATG": 3404, + "GTTACA": 3405, + "GTTACT": 3406, + "GTTACC": 3407, + "GTTACG": 3408, + "GTTAGA": 3409, + "GTTAGT": 3410, + "GTTAGC": 3411, + "GTTAGG": 3412, + "GTTTAA": 3413, + "GTTTAT": 3414, + "GTTTAC": 3415, + "GTTTAG": 3416, + "GTTTTA": 3417, + "GTTTTT": 3418, + "GTTTTC": 3419, + "GTTTTG": 3420, + "GTTTCA": 3421, + "GTTTCT": 3422, + "GTTTCC": 3423, + "GTTTCG": 3424, + "GTTTGA": 3425, + "GTTTGT": 3426, + "GTTTGC": 3427, + "GTTTGG": 3428, + "GTTCAA": 3429, + "GTTCAT": 3430, + "GTTCAC": 3431, + "GTTCAG": 3432, + "GTTCTA": 3433, + "GTTCTT": 3434, + "GTTCTC": 3435, + "GTTCTG": 3436, + "GTTCCA": 3437, + "GTTCCT": 3438, + "GTTCCC": 3439, + "GTTCCG": 3440, + "GTTCGA": 3441, + "GTTCGT": 3442, + "GTTCGC": 3443, + "GTTCGG": 3444, + "GTTGAA": 3445, + "GTTGAT": 3446, + "GTTGAC": 3447, + "GTTGAG": 3448, + "GTTGTA": 3449, + "GTTGTT": 3450, + "GTTGTC": 3451, + "GTTGTG": 3452, + "GTTGCA": 3453, + "GTTGCT": 3454, + "GTTGCC": 3455, + "GTTGCG": 3456, + "GTTGGA": 3457, + "GTTGGT": 3458, + "GTTGGC": 3459, + "GTTGGG": 3460, + "GTCAAA": 3461, + "GTCAAT": 3462, + "GTCAAC": 3463, + "GTCAAG": 3464, + "GTCATA": 3465, + "GTCATT": 3466, + "GTCATC": 3467, + "GTCATG": 3468, + "GTCACA": 3469, + "GTCACT": 3470, + "GTCACC": 3471, + "GTCACG": 3472, + "GTCAGA": 3473, + "GTCAGT": 3474, + "GTCAGC": 3475, + "GTCAGG": 3476, + "GTCTAA": 3477, + "GTCTAT": 3478, + "GTCTAC": 3479, + "GTCTAG": 3480, + "GTCTTA": 3481, + "GTCTTT": 3482, + "GTCTTC": 3483, + "GTCTTG": 3484, + "GTCTCA": 3485, + "GTCTCT": 3486, + "GTCTCC": 3487, + "GTCTCG": 3488, + "GTCTGA": 3489, + "GTCTGT": 3490, + "GTCTGC": 3491, + "GTCTGG": 3492, + "GTCCAA": 3493, + "GTCCAT": 3494, + "GTCCAC": 3495, + "GTCCAG": 3496, + "GTCCTA": 3497, + "GTCCTT": 3498, + "GTCCTC": 3499, + "GTCCTG": 3500, + "GTCCCA": 3501, + "GTCCCT": 3502, + "GTCCCC": 3503, + "GTCCCG": 3504, + "GTCCGA": 3505, + "GTCCGT": 3506, + "GTCCGC": 3507, + "GTCCGG": 3508, + "GTCGAA": 3509, + "GTCGAT": 3510, + "GTCGAC": 3511, + "GTCGAG": 3512, + "GTCGTA": 3513, + "GTCGTT": 3514, + "GTCGTC": 3515, + "GTCGTG": 3516, + "GTCGCA": 3517, + "GTCGCT": 3518, + "GTCGCC": 3519, + "GTCGCG": 3520, + "GTCGGA": 3521, + "GTCGGT": 3522, + "GTCGGC": 3523, + "GTCGGG": 3524, + "GTGAAA": 3525, + "GTGAAT": 3526, + "GTGAAC": 3527, + "GTGAAG": 3528, + "GTGATA": 3529, + "GTGATT": 3530, + "GTGATC": 3531, + "GTGATG": 3532, + "GTGACA": 3533, + "GTGACT": 3534, + "GTGACC": 3535, + "GTGACG": 3536, + "GTGAGA": 3537, + "GTGAGT": 3538, + "GTGAGC": 3539, + "GTGAGG": 3540, + "GTGTAA": 3541, + "GTGTAT": 3542, + "GTGTAC": 3543, + "GTGTAG": 3544, + "GTGTTA": 3545, + "GTGTTT": 3546, + "GTGTTC": 3547, + "GTGTTG": 3548, + "GTGTCA": 3549, + "GTGTCT": 3550, + "GTGTCC": 3551, + "GTGTCG": 3552, + "GTGTGA": 3553, + "GTGTGT": 3554, + "GTGTGC": 3555, + "GTGTGG": 3556, + "GTGCAA": 3557, + "GTGCAT": 3558, + "GTGCAC": 3559, + "GTGCAG": 3560, + "GTGCTA": 3561, + "GTGCTT": 3562, + "GTGCTC": 3563, + "GTGCTG": 3564, + "GTGCCA": 3565, + "GTGCCT": 3566, + "GTGCCC": 3567, + "GTGCCG": 3568, + "GTGCGA": 3569, + "GTGCGT": 3570, + "GTGCGC": 3571, + "GTGCGG": 3572, + "GTGGAA": 3573, + "GTGGAT": 3574, + "GTGGAC": 3575, + "GTGGAG": 3576, + "GTGGTA": 3577, + "GTGGTT": 3578, + "GTGGTC": 3579, + "GTGGTG": 3580, + "GTGGCA": 3581, + "GTGGCT": 3582, + "GTGGCC": 3583, + "GTGGCG": 3584, + "GTGGGA": 3585, + "GTGGGT": 3586, + "GTGGGC": 3587, + "GTGGGG": 3588, + "GCAAAA": 3589, + "GCAAAT": 3590, + "GCAAAC": 3591, + "GCAAAG": 3592, + "GCAATA": 3593, + "GCAATT": 3594, + "GCAATC": 3595, + "GCAATG": 3596, + "GCAACA": 3597, + "GCAACT": 3598, + "GCAACC": 3599, + "GCAACG": 3600, + "GCAAGA": 3601, + "GCAAGT": 3602, + "GCAAGC": 3603, + "GCAAGG": 3604, + "GCATAA": 3605, + "GCATAT": 3606, + "GCATAC": 3607, + "GCATAG": 3608, + "GCATTA": 3609, + "GCATTT": 3610, + "GCATTC": 3611, + "GCATTG": 3612, + "GCATCA": 3613, + "GCATCT": 3614, + "GCATCC": 3615, + "GCATCG": 3616, + "GCATGA": 3617, + "GCATGT": 3618, + "GCATGC": 3619, + "GCATGG": 3620, + "GCACAA": 3621, + "GCACAT": 3622, + "GCACAC": 3623, + "GCACAG": 3624, + "GCACTA": 3625, + "GCACTT": 3626, + "GCACTC": 3627, + "GCACTG": 3628, + "GCACCA": 3629, + "GCACCT": 3630, + "GCACCC": 3631, + "GCACCG": 3632, + "GCACGA": 3633, + "GCACGT": 3634, + "GCACGC": 3635, + "GCACGG": 3636, + "GCAGAA": 3637, + "GCAGAT": 3638, + "GCAGAC": 3639, + "GCAGAG": 3640, + "GCAGTA": 3641, + "GCAGTT": 3642, + "GCAGTC": 3643, + "GCAGTG": 3644, + "GCAGCA": 3645, + "GCAGCT": 3646, + "GCAGCC": 3647, + "GCAGCG": 3648, + "GCAGGA": 3649, + "GCAGGT": 3650, + "GCAGGC": 3651, + "GCAGGG": 3652, + "GCTAAA": 3653, + "GCTAAT": 3654, + "GCTAAC": 3655, + "GCTAAG": 3656, + "GCTATA": 3657, + "GCTATT": 3658, + "GCTATC": 3659, + "GCTATG": 3660, + "GCTACA": 3661, + "GCTACT": 3662, + "GCTACC": 3663, + "GCTACG": 3664, + "GCTAGA": 3665, + "GCTAGT": 3666, + "GCTAGC": 3667, + "GCTAGG": 3668, + "GCTTAA": 3669, + "GCTTAT": 3670, + "GCTTAC": 3671, + "GCTTAG": 3672, + "GCTTTA": 3673, + "GCTTTT": 3674, + "GCTTTC": 3675, + "GCTTTG": 3676, + "GCTTCA": 3677, + "GCTTCT": 3678, + "GCTTCC": 3679, + "GCTTCG": 3680, + "GCTTGA": 3681, + "GCTTGT": 3682, + "GCTTGC": 3683, + "GCTTGG": 3684, + "GCTCAA": 3685, + "GCTCAT": 3686, + "GCTCAC": 3687, + "GCTCAG": 3688, + "GCTCTA": 3689, + "GCTCTT": 3690, + "GCTCTC": 3691, + "GCTCTG": 3692, + "GCTCCA": 3693, + "GCTCCT": 3694, + "GCTCCC": 3695, + "GCTCCG": 3696, + "GCTCGA": 3697, + "GCTCGT": 3698, + "GCTCGC": 3699, + "GCTCGG": 3700, + "GCTGAA": 3701, + "GCTGAT": 3702, + "GCTGAC": 3703, + "GCTGAG": 3704, + "GCTGTA": 3705, + "GCTGTT": 3706, + "GCTGTC": 3707, + "GCTGTG": 3708, + "GCTGCA": 3709, + "GCTGCT": 3710, + "GCTGCC": 3711, + "GCTGCG": 3712, + "GCTGGA": 3713, + "GCTGGT": 3714, + "GCTGGC": 3715, + "GCTGGG": 3716, + "GCCAAA": 3717, + "GCCAAT": 3718, + "GCCAAC": 3719, + "GCCAAG": 3720, + "GCCATA": 3721, + "GCCATT": 3722, + "GCCATC": 3723, + "GCCATG": 3724, + "GCCACA": 3725, + "GCCACT": 3726, + "GCCACC": 3727, + "GCCACG": 3728, + "GCCAGA": 3729, + "GCCAGT": 3730, + "GCCAGC": 3731, + "GCCAGG": 3732, + "GCCTAA": 3733, + "GCCTAT": 3734, + "GCCTAC": 3735, + "GCCTAG": 3736, + "GCCTTA": 3737, + "GCCTTT": 3738, + "GCCTTC": 3739, + "GCCTTG": 3740, + "GCCTCA": 3741, + "GCCTCT": 3742, + "GCCTCC": 3743, + "GCCTCG": 3744, + "GCCTGA": 3745, + "GCCTGT": 3746, + "GCCTGC": 3747, + "GCCTGG": 3748, + "GCCCAA": 3749, + "GCCCAT": 3750, + "GCCCAC": 3751, + "GCCCAG": 3752, + "GCCCTA": 3753, + "GCCCTT": 3754, + "GCCCTC": 3755, + "GCCCTG": 3756, + "GCCCCA": 3757, + "GCCCCT": 3758, + "GCCCCC": 3759, + "GCCCCG": 3760, + "GCCCGA": 3761, + "GCCCGT": 3762, + "GCCCGC": 3763, + "GCCCGG": 3764, + "GCCGAA": 3765, + "GCCGAT": 3766, + "GCCGAC": 3767, + "GCCGAG": 3768, + "GCCGTA": 3769, + "GCCGTT": 3770, + "GCCGTC": 3771, + "GCCGTG": 3772, + "GCCGCA": 3773, + "GCCGCT": 3774, + "GCCGCC": 3775, + "GCCGCG": 3776, + "GCCGGA": 3777, + "GCCGGT": 3778, + "GCCGGC": 3779, + "GCCGGG": 3780, + "GCGAAA": 3781, + "GCGAAT": 3782, + "GCGAAC": 3783, + "GCGAAG": 3784, + "GCGATA": 3785, + "GCGATT": 3786, + "GCGATC": 3787, + "GCGATG": 3788, + "GCGACA": 3789, + "GCGACT": 3790, + "GCGACC": 3791, + "GCGACG": 3792, + "GCGAGA": 3793, + "GCGAGT": 3794, + "GCGAGC": 3795, + "GCGAGG": 3796, + "GCGTAA": 3797, + "GCGTAT": 3798, + "GCGTAC": 3799, + "GCGTAG": 3800, + "GCGTTA": 3801, + "GCGTTT": 3802, + "GCGTTC": 3803, + "GCGTTG": 3804, + "GCGTCA": 3805, + "GCGTCT": 3806, + "GCGTCC": 3807, + "GCGTCG": 3808, + "GCGTGA": 3809, + "GCGTGT": 3810, + "GCGTGC": 3811, + "GCGTGG": 3812, + "GCGCAA": 3813, + "GCGCAT": 3814, + "GCGCAC": 3815, + "GCGCAG": 3816, + "GCGCTA": 3817, + "GCGCTT": 3818, + "GCGCTC": 3819, + "GCGCTG": 3820, + "GCGCCA": 3821, + "GCGCCT": 3822, + "GCGCCC": 3823, + "GCGCCG": 3824, + "GCGCGA": 3825, + "GCGCGT": 3826, + "GCGCGC": 3827, + "GCGCGG": 3828, + "GCGGAA": 3829, + "GCGGAT": 3830, + "GCGGAC": 3831, + "GCGGAG": 3832, + "GCGGTA": 3833, + "GCGGTT": 3834, + "GCGGTC": 3835, + "GCGGTG": 3836, + "GCGGCA": 3837, + "GCGGCT": 3838, + "GCGGCC": 3839, + "GCGGCG": 3840, + "GCGGGA": 3841, + "GCGGGT": 3842, + "GCGGGC": 3843, + "GCGGGG": 3844, + "GGAAAA": 3845, + "GGAAAT": 3846, + "GGAAAC": 3847, + "GGAAAG": 3848, + "GGAATA": 3849, + "GGAATT": 3850, + "GGAATC": 3851, + "GGAATG": 3852, + "GGAACA": 3853, + "GGAACT": 3854, + "GGAACC": 3855, + "GGAACG": 3856, + "GGAAGA": 3857, + "GGAAGT": 3858, + "GGAAGC": 3859, + "GGAAGG": 3860, + "GGATAA": 3861, + "GGATAT": 3862, + "GGATAC": 3863, + "GGATAG": 3864, + "GGATTA": 3865, + "GGATTT": 3866, + "GGATTC": 3867, + "GGATTG": 3868, + "GGATCA": 3869, + "GGATCT": 3870, + "GGATCC": 3871, + "GGATCG": 3872, + "GGATGA": 3873, + "GGATGT": 3874, + "GGATGC": 3875, + "GGATGG": 3876, + "GGACAA": 3877, + "GGACAT": 3878, + "GGACAC": 3879, + "GGACAG": 3880, + "GGACTA": 3881, + "GGACTT": 3882, + "GGACTC": 3883, + "GGACTG": 3884, + "GGACCA": 3885, + "GGACCT": 3886, + "GGACCC": 3887, + "GGACCG": 3888, + "GGACGA": 3889, + "GGACGT": 3890, + "GGACGC": 3891, + "GGACGG": 3892, + "GGAGAA": 3893, + "GGAGAT": 3894, + "GGAGAC": 3895, + "GGAGAG": 3896, + "GGAGTA": 3897, + "GGAGTT": 3898, + "GGAGTC": 3899, + "GGAGTG": 3900, + "GGAGCA": 3901, + "GGAGCT": 3902, + "GGAGCC": 3903, + "GGAGCG": 3904, + "GGAGGA": 3905, + "GGAGGT": 3906, + "GGAGGC": 3907, + "GGAGGG": 3908, + "GGTAAA": 3909, + "GGTAAT": 3910, + "GGTAAC": 3911, + "GGTAAG": 3912, + "GGTATA": 3913, + "GGTATT": 3914, + "GGTATC": 3915, + "GGTATG": 3916, + "GGTACA": 3917, + "GGTACT": 3918, + "GGTACC": 3919, + "GGTACG": 3920, + "GGTAGA": 3921, + "GGTAGT": 3922, + "GGTAGC": 3923, + "GGTAGG": 3924, + "GGTTAA": 3925, + "GGTTAT": 3926, + "GGTTAC": 3927, + "GGTTAG": 3928, + "GGTTTA": 3929, + "GGTTTT": 3930, + "GGTTTC": 3931, + "GGTTTG": 3932, + "GGTTCA": 3933, + "GGTTCT": 3934, + "GGTTCC": 3935, + "GGTTCG": 3936, + "GGTTGA": 3937, + "GGTTGT": 3938, + "GGTTGC": 3939, + "GGTTGG": 3940, + "GGTCAA": 3941, + "GGTCAT": 3942, + "GGTCAC": 3943, + "GGTCAG": 3944, + "GGTCTA": 3945, + "GGTCTT": 3946, + "GGTCTC": 3947, + "GGTCTG": 3948, + "GGTCCA": 3949, + "GGTCCT": 3950, + "GGTCCC": 3951, + "GGTCCG": 3952, + "GGTCGA": 3953, + "GGTCGT": 3954, + "GGTCGC": 3955, + "GGTCGG": 3956, + "GGTGAA": 3957, + "GGTGAT": 3958, + "GGTGAC": 3959, + "GGTGAG": 3960, + "GGTGTA": 3961, + "GGTGTT": 3962, + "GGTGTC": 3963, + "GGTGTG": 3964, + "GGTGCA": 3965, + "GGTGCT": 3966, + "GGTGCC": 3967, + "GGTGCG": 3968, + "GGTGGA": 3969, + "GGTGGT": 3970, + "GGTGGC": 3971, + "GGTGGG": 3972, + "GGCAAA": 3973, + "GGCAAT": 3974, + "GGCAAC": 3975, + "GGCAAG": 3976, + "GGCATA": 3977, + "GGCATT": 3978, + "GGCATC": 3979, + "GGCATG": 3980, + "GGCACA": 3981, + "GGCACT": 3982, + "GGCACC": 3983, + "GGCACG": 3984, + "GGCAGA": 3985, + "GGCAGT": 3986, + "GGCAGC": 3987, + "GGCAGG": 3988, + "GGCTAA": 3989, + "GGCTAT": 3990, + "GGCTAC": 3991, + "GGCTAG": 3992, + "GGCTTA": 3993, + "GGCTTT": 3994, + "GGCTTC": 3995, + "GGCTTG": 3996, + "GGCTCA": 3997, + "GGCTCT": 3998, + "GGCTCC": 3999, + "GGCTCG": 4000, + "GGCTGA": 4001, + "GGCTGT": 4002, + "GGCTGC": 4003, + "GGCTGG": 4004, + "GGCCAA": 4005, + "GGCCAT": 4006, + "GGCCAC": 4007, + "GGCCAG": 4008, + "GGCCTA": 4009, + "GGCCTT": 4010, + "GGCCTC": 4011, + "GGCCTG": 4012, + "GGCCCA": 4013, + "GGCCCT": 4014, + "GGCCCC": 4015, + "GGCCCG": 4016, + "GGCCGA": 4017, + "GGCCGT": 4018, + "GGCCGC": 4019, + "GGCCGG": 4020, + "GGCGAA": 4021, + "GGCGAT": 4022, + "GGCGAC": 4023, + "GGCGAG": 4024, + "GGCGTA": 4025, + "GGCGTT": 4026, + "GGCGTC": 4027, + "GGCGTG": 4028, + "GGCGCA": 4029, + "GGCGCT": 4030, + "GGCGCC": 4031, + "GGCGCG": 4032, + "GGCGGA": 4033, + "GGCGGT": 4034, + "GGCGGC": 4035, + "GGCGGG": 4036, + "GGGAAA": 4037, + "GGGAAT": 4038, + "GGGAAC": 4039, + "GGGAAG": 4040, + "GGGATA": 4041, + "GGGATT": 4042, + "GGGATC": 4043, + "GGGATG": 4044, + "GGGACA": 4045, + "GGGACT": 4046, + "GGGACC": 4047, + "GGGACG": 4048, + "GGGAGA": 4049, + "GGGAGT": 4050, + "GGGAGC": 4051, + "GGGAGG": 4052, + "GGGTAA": 4053, + "GGGTAT": 4054, + "GGGTAC": 4055, + "GGGTAG": 4056, + "GGGTTA": 4057, + "GGGTTT": 4058, + "GGGTTC": 4059, + "GGGTTG": 4060, + "GGGTCA": 4061, + "GGGTCT": 4062, + "GGGTCC": 4063, + "GGGTCG": 4064, + "GGGTGA": 4065, + "GGGTGT": 4066, + "GGGTGC": 4067, + "GGGTGG": 4068, + "GGGCAA": 4069, + "GGGCAT": 4070, + "GGGCAC": 4071, + "GGGCAG": 4072, + "GGGCTA": 4073, + "GGGCTT": 4074, + "GGGCTC": 4075, + "GGGCTG": 4076, + "GGGCCA": 4077, + "GGGCCT": 4078, + "GGGCCC": 4079, + "GGGCCG": 4080, + "GGGCGA": 4081, + "GGGCGT": 4082, + "GGGCGC": 4083, + "GGGCGG": 4084, + "GGGGAA": 4085, + "GGGGAT": 4086, + "GGGGAC": 4087, + "GGGGAG": 4088, + "GGGGTA": 4089, + "GGGGTT": 4090, + "GGGGTC": 4091, + "GGGGTG": 4092, + "GGGGCA": 4093, + "GGGGCT": 4094, + "GGGGCC": 4095, + "GGGGCG": 4096, + "GGGGGA": 4097, + "GGGGGT": 4098, + "GGGGGC": 4099, + "GGGGGG": 4100 + } + } +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6000/tokenizer_config.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6000/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..604e2b1b29af1dcb3920d2828b6a90276e5e3d02 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6000/tokenizer_config.json @@ -0,0 +1,58 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "[UNK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "[CLS]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "[SEP]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "4": { + "content": "[MASK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "clean_up_tokenization_spaces": true, + "cls_token": "[CLS]", + "do_basic_tokenize": true, + "do_lower_case": false, + "mask_token": "[MASK]", + "max_len": 512, + "model_max_length": 512, + "never_split": null, + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "strip_accents": null, + "tokenize_chinese_chars": true, + "tokenizer_class": "BertTokenizer", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6000/trainer_state.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..5b08e7c776ec9a0a6f4c5945fe46d8d95984137c --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6000/trainer_state.json @@ -0,0 +1,1713 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 24.0, + "eval_steps": 500, + "global_step": 6000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.1, + "grad_norm": 1.1046091318130493, + "learning_rate": 5.319148936170213e-05, + "loss": 0.6562, + "step": 25 + }, + { + "epoch": 0.2, + "grad_norm": 0.4066573977470398, + "learning_rate": 0.00010638297872340425, + "loss": 0.5373, + "step": 50 + }, + { + "epoch": 0.3, + "grad_norm": 0.7340325713157654, + "learning_rate": 0.00015957446808510637, + "loss": 0.4995, + "step": 75 + }, + { + "epoch": 0.4, + "grad_norm": 2.134040117263794, + "learning_rate": 0.0002127659574468085, + "loss": 0.4871, + "step": 100 + }, + { + "epoch": 0.5, + "grad_norm": 2.911975860595703, + "learning_rate": 0.00026595744680851064, + "loss": 0.4507, + "step": 125 + }, + { + "epoch": 0.6, + "grad_norm": 2.040365219116211, + "learning_rate": 0.00031914893617021275, + "loss": 0.3944, + "step": 150 + }, + { + "epoch": 0.7, + "grad_norm": 5.643661022186279, + "learning_rate": 0.0003723404255319149, + "loss": 0.3863, + "step": 175 + }, + { + "epoch": 0.8, + "grad_norm": 3.882492780685425, + "learning_rate": 0.0003992081821181128, + "loss": 0.3875, + "step": 200 + }, + { + "epoch": 0.9, + "grad_norm": 5.87437105178833, + "learning_rate": 0.0003975585615308479, + "loss": 0.3856, + "step": 225 + }, + { + "epoch": 1.0, + "grad_norm": 4.611206531524658, + "learning_rate": 0.00039590894094358297, + "loss": 0.3866, + "step": 250 + }, + { + "epoch": 1.1, + "grad_norm": 3.6404929161071777, + "learning_rate": 0.0003942593203563181, + "loss": 0.3437, + "step": 275 + }, + { + "epoch": 1.2, + "grad_norm": 6.469444751739502, + "learning_rate": 0.0003926096997690532, + "loss": 0.3192, + "step": 300 + }, + { + "epoch": 1.3, + "grad_norm": 3.6251447200775146, + "learning_rate": 0.00039096007918178817, + "loss": 0.349, + "step": 325 + }, + { + "epoch": 1.4, + "grad_norm": 2.854794979095459, + "learning_rate": 0.0003893104585945233, + "loss": 0.3274, + "step": 350 + }, + { + "epoch": 1.5, + "grad_norm": 4.805097579956055, + "learning_rate": 0.0003876608380072583, + "loss": 0.3279, + "step": 375 + }, + { + "epoch": 1.6, + "grad_norm": 2.9891581535339355, + "learning_rate": 0.00038601121741999343, + "loss": 0.3166, + "step": 400 + }, + { + "epoch": 1.7, + "grad_norm": 5.333752155303955, + "learning_rate": 0.0003843615968327285, + "loss": 0.2856, + "step": 425 + }, + { + "epoch": 1.8, + "grad_norm": 2.615621566772461, + "learning_rate": 0.0003827119762454636, + "loss": 0.3286, + "step": 450 + }, + { + "epoch": 1.9, + "grad_norm": 2.4228415489196777, + "learning_rate": 0.00038106235565819863, + "loss": 0.2812, + "step": 475 + }, + { + "epoch": 2.0, + "grad_norm": 3.6015517711639404, + "learning_rate": 0.0003794127350709337, + "loss": 0.2884, + "step": 500 + }, + { + "epoch": 2.1, + "grad_norm": 2.7867860794067383, + "learning_rate": 0.0003777631144836688, + "loss": 0.2582, + "step": 525 + }, + { + "epoch": 2.2, + "grad_norm": 10.079645156860352, + "learning_rate": 0.00037611349389640383, + "loss": 0.245, + "step": 550 + }, + { + "epoch": 2.3, + "grad_norm": 2.475917100906372, + "learning_rate": 0.00037446387330913894, + "loss": 0.2615, + "step": 575 + }, + { + "epoch": 2.4, + "grad_norm": 8.098592758178711, + "learning_rate": 0.000372814252721874, + "loss": 0.2484, + "step": 600 + }, + { + "epoch": 2.5, + "grad_norm": 2.318723201751709, + "learning_rate": 0.00037116463213460903, + "loss": 0.2056, + "step": 625 + }, + { + "epoch": 2.6, + "grad_norm": 3.616283893585205, + "learning_rate": 0.00036951501154734414, + "loss": 0.2358, + "step": 650 + }, + { + "epoch": 2.7, + "grad_norm": 6.419492721557617, + "learning_rate": 0.0003678653909600792, + "loss": 0.2204, + "step": 675 + }, + { + "epoch": 2.8, + "grad_norm": 2.9333388805389404, + "learning_rate": 0.0003662157703728143, + "loss": 0.1857, + "step": 700 + }, + { + "epoch": 2.9, + "grad_norm": 7.691112995147705, + "learning_rate": 0.00036456614978554934, + "loss": 0.2054, + "step": 725 + }, + { + "epoch": 3.0, + "grad_norm": 4.908492565155029, + "learning_rate": 0.00036291652919828444, + "loss": 0.1956, + "step": 750 + }, + { + "epoch": 3.1, + "grad_norm": 1.8923287391662598, + "learning_rate": 0.00036126690861101944, + "loss": 0.1708, + "step": 775 + }, + { + "epoch": 3.2, + "grad_norm": 4.403504848480225, + "learning_rate": 0.00035961728802375454, + "loss": 0.1441, + "step": 800 + }, + { + "epoch": 3.3, + "grad_norm": 4.117386817932129, + "learning_rate": 0.0003579676674364896, + "loss": 0.1586, + "step": 825 + }, + { + "epoch": 3.4, + "grad_norm": 7.119816303253174, + "learning_rate": 0.0003563180468492247, + "loss": 0.1472, + "step": 850 + }, + { + "epoch": 3.5, + "grad_norm": 3.5796430110931396, + "learning_rate": 0.0003546684262619598, + "loss": 0.1802, + "step": 875 + }, + { + "epoch": 3.6, + "grad_norm": 2.97688889503479, + "learning_rate": 0.00035301880567469485, + "loss": 0.1578, + "step": 900 + }, + { + "epoch": 3.7, + "grad_norm": 3.716148614883423, + "learning_rate": 0.0003513691850874299, + "loss": 0.1662, + "step": 925 + }, + { + "epoch": 3.8, + "grad_norm": 6.1249566078186035, + "learning_rate": 0.00034971956450016495, + "loss": 0.159, + "step": 950 + }, + { + "epoch": 3.9, + "grad_norm": 3.0592427253723145, + "learning_rate": 0.00034806994391290005, + "loss": 0.148, + "step": 975 + }, + { + "epoch": 4.0, + "grad_norm": 4.467265605926514, + "learning_rate": 0.0003464203233256351, + "loss": 0.1442, + "step": 1000 + }, + { + "epoch": 4.1, + "grad_norm": 2.7223546504974365, + "learning_rate": 0.0003447707027383702, + "loss": 0.1339, + "step": 1025 + }, + { + "epoch": 4.2, + "grad_norm": 3.698854923248291, + "learning_rate": 0.00034312108215110525, + "loss": 0.1381, + "step": 1050 + }, + { + "epoch": 4.3, + "grad_norm": 7.418384552001953, + "learning_rate": 0.0003414714615638403, + "loss": 0.1158, + "step": 1075 + }, + { + "epoch": 4.4, + "grad_norm": 1.2887814044952393, + "learning_rate": 0.0003398218409765754, + "loss": 0.0855, + "step": 1100 + }, + { + "epoch": 4.5, + "grad_norm": 0.838731586933136, + "learning_rate": 0.00033817222038931045, + "loss": 0.1272, + "step": 1125 + }, + { + "epoch": 4.6, + "grad_norm": 5.912592887878418, + "learning_rate": 0.00033652259980204556, + "loss": 0.1156, + "step": 1150 + }, + { + "epoch": 4.7, + "grad_norm": 8.49521255493164, + "learning_rate": 0.0003348729792147806, + "loss": 0.1182, + "step": 1175 + }, + { + "epoch": 4.8, + "grad_norm": 10.278315544128418, + "learning_rate": 0.0003332233586275157, + "loss": 0.1071, + "step": 1200 + }, + { + "epoch": 4.9, + "grad_norm": 0.724703311920166, + "learning_rate": 0.00033157373804025076, + "loss": 0.11, + "step": 1225 + }, + { + "epoch": 5.0, + "grad_norm": 1.6808199882507324, + "learning_rate": 0.0003299241174529858, + "loss": 0.1101, + "step": 1250 + }, + { + "epoch": 5.1, + "grad_norm": 0.1038585901260376, + "learning_rate": 0.0003282744968657209, + "loss": 0.0791, + "step": 1275 + }, + { + "epoch": 5.2, + "grad_norm": 0.15866787731647491, + "learning_rate": 0.00032662487627845596, + "loss": 0.0638, + "step": 1300 + }, + { + "epoch": 5.3, + "grad_norm": 6.748287200927734, + "learning_rate": 0.00032497525569119106, + "loss": 0.0961, + "step": 1325 + }, + { + "epoch": 5.4, + "grad_norm": 1.5960336923599243, + "learning_rate": 0.0003233256351039261, + "loss": 0.1057, + "step": 1350 + }, + { + "epoch": 5.5, + "grad_norm": 3.3678526878356934, + "learning_rate": 0.00032167601451666116, + "loss": 0.1092, + "step": 1375 + }, + { + "epoch": 5.6, + "grad_norm": 4.743562698364258, + "learning_rate": 0.00032002639392939627, + "loss": 0.1054, + "step": 1400 + }, + { + "epoch": 5.7, + "grad_norm": 12.860156059265137, + "learning_rate": 0.0003183767733421313, + "loss": 0.1009, + "step": 1425 + }, + { + "epoch": 5.8, + "grad_norm": 0.8820792436599731, + "learning_rate": 0.0003167271527548664, + "loss": 0.1219, + "step": 1450 + }, + { + "epoch": 5.9, + "grad_norm": 2.732572317123413, + "learning_rate": 0.00031507753216760147, + "loss": 0.0682, + "step": 1475 + }, + { + "epoch": 6.0, + "grad_norm": 9.810711860656738, + "learning_rate": 0.00031342791158033657, + "loss": 0.1134, + "step": 1500 + }, + { + "epoch": 6.1, + "grad_norm": 6.699489593505859, + "learning_rate": 0.00031177829099307157, + "loss": 0.0557, + "step": 1525 + }, + { + "epoch": 6.2, + "grad_norm": 14.691123008728027, + "learning_rate": 0.00031012867040580667, + "loss": 0.0599, + "step": 1550 + }, + { + "epoch": 6.3, + "grad_norm": 24.44135093688965, + "learning_rate": 0.0003084790498185417, + "loss": 0.0818, + "step": 1575 + }, + { + "epoch": 6.4, + "grad_norm": 0.033995576202869415, + "learning_rate": 0.0003068294292312768, + "loss": 0.0692, + "step": 1600 + }, + { + "epoch": 6.5, + "grad_norm": 6.722524642944336, + "learning_rate": 0.0003051798086440119, + "loss": 0.0702, + "step": 1625 + }, + { + "epoch": 6.6, + "grad_norm": 2.9519436359405518, + "learning_rate": 0.000303530188056747, + "loss": 0.073, + "step": 1650 + }, + { + "epoch": 6.7, + "grad_norm": 1.5118328332901, + "learning_rate": 0.000301880567469482, + "loss": 0.0926, + "step": 1675 + }, + { + "epoch": 6.8, + "grad_norm": 2.1923201084136963, + "learning_rate": 0.0003002309468822171, + "loss": 0.0821, + "step": 1700 + }, + { + "epoch": 6.9, + "grad_norm": 3.34309458732605, + "learning_rate": 0.0002985813262949522, + "loss": 0.0827, + "step": 1725 + }, + { + "epoch": 7.0, + "grad_norm": 10.289947509765625, + "learning_rate": 0.00029693170570768723, + "loss": 0.1162, + "step": 1750 + }, + { + "epoch": 7.1, + "grad_norm": 0.051936667412519455, + "learning_rate": 0.00029528208512042233, + "loss": 0.0626, + "step": 1775 + }, + { + "epoch": 7.2, + "grad_norm": 1.8913339376449585, + "learning_rate": 0.0002936324645331574, + "loss": 0.0657, + "step": 1800 + }, + { + "epoch": 7.3, + "grad_norm": 15.073695182800293, + "learning_rate": 0.00029198284394589243, + "loss": 0.0539, + "step": 1825 + }, + { + "epoch": 7.4, + "grad_norm": 4.514949798583984, + "learning_rate": 0.00029033322335862753, + "loss": 0.056, + "step": 1850 + }, + { + "epoch": 7.5, + "grad_norm": 0.46094000339508057, + "learning_rate": 0.0002886836027713626, + "loss": 0.0667, + "step": 1875 + }, + { + "epoch": 7.6, + "grad_norm": 4.682873725891113, + "learning_rate": 0.0002870339821840977, + "loss": 0.0673, + "step": 1900 + }, + { + "epoch": 7.7, + "grad_norm": 3.581599235534668, + "learning_rate": 0.00028538436159683273, + "loss": 0.0699, + "step": 1925 + }, + { + "epoch": 7.8, + "grad_norm": 3.2518703937530518, + "learning_rate": 0.00028373474100956784, + "loss": 0.0628, + "step": 1950 + }, + { + "epoch": 7.9, + "grad_norm": 0.24071219563484192, + "learning_rate": 0.0002820851204223029, + "loss": 0.0747, + "step": 1975 + }, + { + "epoch": 8.0, + "grad_norm": 2.8545596599578857, + "learning_rate": 0.00028043549983503794, + "loss": 0.0688, + "step": 2000 + }, + { + "epoch": 8.1, + "grad_norm": 0.049499694257974625, + "learning_rate": 0.00027878587924777304, + "loss": 0.0563, + "step": 2025 + }, + { + "epoch": 8.2, + "grad_norm": 22.112478256225586, + "learning_rate": 0.0002771362586605081, + "loss": 0.0545, + "step": 2050 + }, + { + "epoch": 8.3, + "grad_norm": 20.87473487854004, + "learning_rate": 0.0002754866380732432, + "loss": 0.0256, + "step": 2075 + }, + { + "epoch": 8.4, + "grad_norm": 2.317768096923828, + "learning_rate": 0.00027383701748597824, + "loss": 0.0753, + "step": 2100 + }, + { + "epoch": 8.5, + "grad_norm": 20.241336822509766, + "learning_rate": 0.0002721873968987133, + "loss": 0.0257, + "step": 2125 + }, + { + "epoch": 8.6, + "grad_norm": 14.994575500488281, + "learning_rate": 0.00027053777631144834, + "loss": 0.0748, + "step": 2150 + }, + { + "epoch": 8.7, + "grad_norm": 0.8268639445304871, + "learning_rate": 0.00026888815572418344, + "loss": 0.0805, + "step": 2175 + }, + { + "epoch": 8.8, + "grad_norm": 0.031071001663804054, + "learning_rate": 0.00026723853513691855, + "loss": 0.0603, + "step": 2200 + }, + { + "epoch": 8.9, + "grad_norm": 2.4572932720184326, + "learning_rate": 0.0002655889145496536, + "loss": 0.071, + "step": 2225 + }, + { + "epoch": 9.0, + "grad_norm": 0.21708321571350098, + "learning_rate": 0.0002639392939623887, + "loss": 0.0606, + "step": 2250 + }, + { + "epoch": 9.1, + "grad_norm": 14.928559303283691, + "learning_rate": 0.0002622896733751237, + "loss": 0.0443, + "step": 2275 + }, + { + "epoch": 9.2, + "grad_norm": 1.108826994895935, + "learning_rate": 0.0002606400527878588, + "loss": 0.0509, + "step": 2300 + }, + { + "epoch": 9.3, + "grad_norm": 0.05721910670399666, + "learning_rate": 0.00025899043220059385, + "loss": 0.0459, + "step": 2325 + }, + { + "epoch": 9.4, + "grad_norm": 7.81903600692749, + "learning_rate": 0.00025734081161332895, + "loss": 0.0535, + "step": 2350 + }, + { + "epoch": 9.5, + "grad_norm": 0.01475981343537569, + "learning_rate": 0.000255691191026064, + "loss": 0.0399, + "step": 2375 + }, + { + "epoch": 9.6, + "grad_norm": 5.496493339538574, + "learning_rate": 0.0002540415704387991, + "loss": 0.0744, + "step": 2400 + }, + { + "epoch": 9.7, + "grad_norm": 0.05929339677095413, + "learning_rate": 0.00025239194985153415, + "loss": 0.0416, + "step": 2425 + }, + { + "epoch": 9.8, + "grad_norm": 0.06961135566234589, + "learning_rate": 0.0002507423292642692, + "loss": 0.04, + "step": 2450 + }, + { + "epoch": 9.9, + "grad_norm": 0.02165246568620205, + "learning_rate": 0.0002490927086770043, + "loss": 0.0984, + "step": 2475 + }, + { + "epoch": 10.0, + "grad_norm": 25.155675888061523, + "learning_rate": 0.00024744308808973936, + "loss": 0.042, + "step": 2500 + }, + { + "epoch": 10.1, + "grad_norm": 8.566617965698242, + "learning_rate": 0.00024579346750247446, + "loss": 0.0385, + "step": 2525 + }, + { + "epoch": 10.2, + "grad_norm": 0.024992404505610466, + "learning_rate": 0.0002441438469152095, + "loss": 0.0327, + "step": 2550 + }, + { + "epoch": 10.3, + "grad_norm": 0.22233247756958008, + "learning_rate": 0.00024249422632794456, + "loss": 0.0211, + "step": 2575 + }, + { + "epoch": 10.4, + "grad_norm": 13.307366371154785, + "learning_rate": 0.00024084460574067963, + "loss": 0.0472, + "step": 2600 + }, + { + "epoch": 10.5, + "grad_norm": 0.17787696421146393, + "learning_rate": 0.0002391949851534147, + "loss": 0.0346, + "step": 2625 + }, + { + "epoch": 10.6, + "grad_norm": 1.2845549583435059, + "learning_rate": 0.0002375453645661498, + "loss": 0.0587, + "step": 2650 + }, + { + "epoch": 10.7, + "grad_norm": 0.0603482760488987, + "learning_rate": 0.00023589574397888486, + "loss": 0.0394, + "step": 2675 + }, + { + "epoch": 10.8, + "grad_norm": 0.031819943338632584, + "learning_rate": 0.00023424612339161997, + "loss": 0.0434, + "step": 2700 + }, + { + "epoch": 10.9, + "grad_norm": 1.5738537311553955, + "learning_rate": 0.000232596502804355, + "loss": 0.0393, + "step": 2725 + }, + { + "epoch": 11.0, + "grad_norm": 0.6862583756446838, + "learning_rate": 0.00023094688221709007, + "loss": 0.0183, + "step": 2750 + }, + { + "epoch": 11.1, + "grad_norm": 0.0053489054553210735, + "learning_rate": 0.00022929726162982514, + "loss": 0.0277, + "step": 2775 + }, + { + "epoch": 11.2, + "grad_norm": 0.04365606606006622, + "learning_rate": 0.00022764764104256022, + "loss": 0.0337, + "step": 2800 + }, + { + "epoch": 11.3, + "grad_norm": 1.3348641395568848, + "learning_rate": 0.0002259980204552953, + "loss": 0.0219, + "step": 2825 + }, + { + "epoch": 11.4, + "grad_norm": 3.3695547580718994, + "learning_rate": 0.00022434839986803037, + "loss": 0.0227, + "step": 2850 + }, + { + "epoch": 11.5, + "grad_norm": 2.507455587387085, + "learning_rate": 0.00022269877928076542, + "loss": 0.0443, + "step": 2875 + }, + { + "epoch": 11.6, + "grad_norm": 0.013451912440359592, + "learning_rate": 0.0002210491586935005, + "loss": 0.0486, + "step": 2900 + }, + { + "epoch": 11.7, + "grad_norm": 0.031177405267953873, + "learning_rate": 0.00021939953810623557, + "loss": 0.0511, + "step": 2925 + }, + { + "epoch": 11.8, + "grad_norm": 0.02148960903286934, + "learning_rate": 0.00021774991751897065, + "loss": 0.0269, + "step": 2950 + }, + { + "epoch": 11.9, + "grad_norm": 0.01219563465565443, + "learning_rate": 0.00021610029693170573, + "loss": 0.042, + "step": 2975 + }, + { + "epoch": 12.0, + "grad_norm": 0.2062515765428543, + "learning_rate": 0.0002144506763444408, + "loss": 0.031, + "step": 3000 + }, + { + "epoch": 12.1, + "grad_norm": 3.235403060913086, + "learning_rate": 0.00021280105575717582, + "loss": 0.0414, + "step": 3025 + }, + { + "epoch": 12.2, + "grad_norm": 1.089821696281433, + "learning_rate": 0.00021115143516991093, + "loss": 0.0692, + "step": 3050 + }, + { + "epoch": 12.3, + "grad_norm": 0.013664484024047852, + "learning_rate": 0.000209501814582646, + "loss": 0.0367, + "step": 3075 + }, + { + "epoch": 12.4, + "grad_norm": 1.3101387023925781, + "learning_rate": 0.00020785219399538108, + "loss": 0.0277, + "step": 3100 + }, + { + "epoch": 12.5, + "grad_norm": 3.244853973388672, + "learning_rate": 0.00020620257340811616, + "loss": 0.0373, + "step": 3125 + }, + { + "epoch": 12.6, + "grad_norm": 0.013548053801059723, + "learning_rate": 0.00020455295282085123, + "loss": 0.027, + "step": 3150 + }, + { + "epoch": 12.7, + "grad_norm": 0.004670978989452124, + "learning_rate": 0.00020290333223358626, + "loss": 0.0252, + "step": 3175 + }, + { + "epoch": 12.8, + "grad_norm": 0.007080752402544022, + "learning_rate": 0.00020125371164632133, + "loss": 0.0187, + "step": 3200 + }, + { + "epoch": 12.9, + "grad_norm": 0.005918944254517555, + "learning_rate": 0.0001996040910590564, + "loss": 0.0423, + "step": 3225 + }, + { + "epoch": 13.0, + "grad_norm": 4.530269145965576, + "learning_rate": 0.00019795447047179148, + "loss": 0.0381, + "step": 3250 + }, + { + "epoch": 13.1, + "grad_norm": 0.00533739197999239, + "learning_rate": 0.0001963048498845266, + "loss": 0.0097, + "step": 3275 + }, + { + "epoch": 13.2, + "grad_norm": 0.7641226649284363, + "learning_rate": 0.00019465522929726164, + "loss": 0.0205, + "step": 3300 + }, + { + "epoch": 13.3, + "grad_norm": 2.16174054145813, + "learning_rate": 0.00019300560870999671, + "loss": 0.0174, + "step": 3325 + }, + { + "epoch": 13.4, + "grad_norm": 0.00814723875373602, + "learning_rate": 0.0001913559881227318, + "loss": 0.0215, + "step": 3350 + }, + { + "epoch": 13.5, + "grad_norm": 10.500170707702637, + "learning_rate": 0.00018970636753546684, + "loss": 0.0226, + "step": 3375 + }, + { + "epoch": 13.6, + "grad_norm": 2.612359046936035, + "learning_rate": 0.00018805674694820192, + "loss": 0.0314, + "step": 3400 + }, + { + "epoch": 13.7, + "grad_norm": 0.11265736818313599, + "learning_rate": 0.000186407126360937, + "loss": 0.0234, + "step": 3425 + }, + { + "epoch": 13.8, + "grad_norm": 0.06453213095664978, + "learning_rate": 0.00018475750577367207, + "loss": 0.0314, + "step": 3450 + }, + { + "epoch": 13.9, + "grad_norm": 0.01650763861835003, + "learning_rate": 0.00018310788518640715, + "loss": 0.0403, + "step": 3475 + }, + { + "epoch": 14.0, + "grad_norm": 0.002764373552054167, + "learning_rate": 0.00018145826459914222, + "loss": 0.0137, + "step": 3500 + }, + { + "epoch": 14.1, + "grad_norm": 0.0022430066019296646, + "learning_rate": 0.00017980864401187727, + "loss": 0.0084, + "step": 3525 + }, + { + "epoch": 14.2, + "grad_norm": 0.007267679553478956, + "learning_rate": 0.00017815902342461235, + "loss": 0.0161, + "step": 3550 + }, + { + "epoch": 14.3, + "grad_norm": 0.004844362381845713, + "learning_rate": 0.00017650940283734742, + "loss": 0.0399, + "step": 3575 + }, + { + "epoch": 14.4, + "grad_norm": 0.05529040843248367, + "learning_rate": 0.00017485978225008247, + "loss": 0.0192, + "step": 3600 + }, + { + "epoch": 14.5, + "grad_norm": 0.008248478174209595, + "learning_rate": 0.00017321016166281755, + "loss": 0.0079, + "step": 3625 + }, + { + "epoch": 14.6, + "grad_norm": 0.02617962658405304, + "learning_rate": 0.00017156054107555263, + "loss": 0.0244, + "step": 3650 + }, + { + "epoch": 14.7, + "grad_norm": 0.004497932270169258, + "learning_rate": 0.0001699109204882877, + "loss": 0.0464, + "step": 3675 + }, + { + "epoch": 14.8, + "grad_norm": 0.008586671203374863, + "learning_rate": 0.00016826129990102278, + "loss": 0.0354, + "step": 3700 + }, + { + "epoch": 14.9, + "grad_norm": 0.30177807807922363, + "learning_rate": 0.00016661167931375785, + "loss": 0.0147, + "step": 3725 + }, + { + "epoch": 15.0, + "grad_norm": 0.10390808433294296, + "learning_rate": 0.0001649620587264929, + "loss": 0.0078, + "step": 3750 + }, + { + "epoch": 15.1, + "grad_norm": 1.584868311882019, + "learning_rate": 0.00016331243813922798, + "loss": 0.0059, + "step": 3775 + }, + { + "epoch": 15.2, + "grad_norm": 0.0038078685756772757, + "learning_rate": 0.00016166281755196306, + "loss": 0.0143, + "step": 3800 + }, + { + "epoch": 15.3, + "grad_norm": 0.017795555293560028, + "learning_rate": 0.00016001319696469813, + "loss": 0.0003, + "step": 3825 + }, + { + "epoch": 15.4, + "grad_norm": 7.246474742889404, + "learning_rate": 0.0001583635763774332, + "loss": 0.0288, + "step": 3850 + }, + { + "epoch": 15.5, + "grad_norm": 0.0014716258738189936, + "learning_rate": 0.00015671395579016829, + "loss": 0.0113, + "step": 3875 + }, + { + "epoch": 15.6, + "grad_norm": 0.002159764291718602, + "learning_rate": 0.00015506433520290334, + "loss": 0.0138, + "step": 3900 + }, + { + "epoch": 15.7, + "grad_norm": 27.16730308532715, + "learning_rate": 0.0001534147146156384, + "loss": 0.0048, + "step": 3925 + }, + { + "epoch": 15.8, + "grad_norm": 0.004570333287119865, + "learning_rate": 0.0001517650940283735, + "loss": 0.0144, + "step": 3950 + }, + { + "epoch": 15.9, + "grad_norm": 0.008330491371452808, + "learning_rate": 0.00015011547344110854, + "loss": 0.0371, + "step": 3975 + }, + { + "epoch": 16.0, + "grad_norm": 15.75979232788086, + "learning_rate": 0.00014846585285384361, + "loss": 0.0207, + "step": 4000 + }, + { + "epoch": 16.1, + "grad_norm": 14.096012115478516, + "learning_rate": 0.0001468162322665787, + "loss": 0.018, + "step": 4025 + }, + { + "epoch": 16.2, + "grad_norm": 0.0015319391386583447, + "learning_rate": 0.00014516661167931377, + "loss": 0.0109, + "step": 4050 + }, + { + "epoch": 16.3, + "grad_norm": 0.003954900894314051, + "learning_rate": 0.00014351699109204884, + "loss": 0.012, + "step": 4075 + }, + { + "epoch": 16.4, + "grad_norm": 37.79055404663086, + "learning_rate": 0.00014186737050478392, + "loss": 0.0301, + "step": 4100 + }, + { + "epoch": 16.5, + "grad_norm": 0.0014516041846945882, + "learning_rate": 0.00014021774991751897, + "loss": 0.0116, + "step": 4125 + }, + { + "epoch": 16.6, + "grad_norm": 0.009134252555668354, + "learning_rate": 0.00013856812933025404, + "loss": 0.0126, + "step": 4150 + }, + { + "epoch": 16.7, + "grad_norm": 19.0325927734375, + "learning_rate": 0.00013691850874298912, + "loss": 0.0398, + "step": 4175 + }, + { + "epoch": 16.8, + "grad_norm": 0.41503193974494934, + "learning_rate": 0.00013526888815572417, + "loss": 0.001, + "step": 4200 + }, + { + "epoch": 16.9, + "grad_norm": 0.0014813834568485618, + "learning_rate": 0.00013361926756845927, + "loss": 0.0077, + "step": 4225 + }, + { + "epoch": 17.0, + "grad_norm": 0.007494773715734482, + "learning_rate": 0.00013196964698119435, + "loss": 0.0052, + "step": 4250 + }, + { + "epoch": 17.1, + "grad_norm": 0.002411301014944911, + "learning_rate": 0.0001303200263939294, + "loss": 0.0038, + "step": 4275 + }, + { + "epoch": 17.2, + "grad_norm": 0.01521711703389883, + "learning_rate": 0.00012867040580666448, + "loss": 0.0051, + "step": 4300 + }, + { + "epoch": 17.3, + "grad_norm": 0.0010520346695557237, + "learning_rate": 0.00012702078521939955, + "loss": 0.0167, + "step": 4325 + }, + { + "epoch": 17.4, + "grad_norm": 0.054758671671152115, + "learning_rate": 0.0001253711646321346, + "loss": 0.011, + "step": 4350 + }, + { + "epoch": 17.5, + "grad_norm": 0.0012187627144157887, + "learning_rate": 0.00012372154404486968, + "loss": 0.0239, + "step": 4375 + }, + { + "epoch": 17.6, + "grad_norm": 58.37099838256836, + "learning_rate": 0.00012207192345760475, + "loss": 0.0055, + "step": 4400 + }, + { + "epoch": 17.7, + "grad_norm": 0.27519506216049194, + "learning_rate": 0.00012042230287033982, + "loss": 0.0332, + "step": 4425 + }, + { + "epoch": 17.8, + "grad_norm": 0.0011721524642780423, + "learning_rate": 0.0001187726822830749, + "loss": 0.0048, + "step": 4450 + }, + { + "epoch": 17.9, + "grad_norm": 0.0015529991360381246, + "learning_rate": 0.00011712306169580998, + "loss": 0.0134, + "step": 4475 + }, + { + "epoch": 18.0, + "grad_norm": 0.0008489767787978053, + "learning_rate": 0.00011547344110854503, + "loss": 0.0148, + "step": 4500 + }, + { + "epoch": 18.1, + "grad_norm": 0.0010399922030046582, + "learning_rate": 0.00011382382052128011, + "loss": 0.0046, + "step": 4525 + }, + { + "epoch": 18.2, + "grad_norm": 0.0017199565190821886, + "learning_rate": 0.00011217419993401519, + "loss": 0.0123, + "step": 4550 + }, + { + "epoch": 18.3, + "grad_norm": 0.004095083102583885, + "learning_rate": 0.00011052457934675025, + "loss": 0.0078, + "step": 4575 + }, + { + "epoch": 18.4, + "grad_norm": 0.005569108761847019, + "learning_rate": 0.00010887495875948532, + "loss": 0.0001, + "step": 4600 + }, + { + "epoch": 18.5, + "grad_norm": 0.0013379952870309353, + "learning_rate": 0.0001072253381722204, + "loss": 0.0123, + "step": 4625 + }, + { + "epoch": 18.6, + "grad_norm": 0.13633179664611816, + "learning_rate": 0.00010557571758495546, + "loss": 0.016, + "step": 4650 + }, + { + "epoch": 18.7, + "grad_norm": 0.00383689533919096, + "learning_rate": 0.00010392609699769054, + "loss": 0.0238, + "step": 4675 + }, + { + "epoch": 18.8, + "grad_norm": 0.011090466752648354, + "learning_rate": 0.00010227647641042562, + "loss": 0.0299, + "step": 4700 + }, + { + "epoch": 18.9, + "grad_norm": 0.01791750080883503, + "learning_rate": 0.00010062685582316067, + "loss": 0.0006, + "step": 4725 + }, + { + "epoch": 19.0, + "grad_norm": 20.31785774230957, + "learning_rate": 9.897723523589574e-05, + "loss": 0.0067, + "step": 4750 + }, + { + "epoch": 19.1, + "grad_norm": 0.0011455218773335218, + "learning_rate": 9.732761464863082e-05, + "loss": 0.0138, + "step": 4775 + }, + { + "epoch": 19.2, + "grad_norm": 0.0014105259906500578, + "learning_rate": 9.56779940613659e-05, + "loss": 0.0005, + "step": 4800 + }, + { + "epoch": 19.3, + "grad_norm": 0.002618618542328477, + "learning_rate": 9.402837347410096e-05, + "loss": 0.0112, + "step": 4825 + }, + { + "epoch": 19.4, + "grad_norm": 0.001108857337385416, + "learning_rate": 9.237875288683603e-05, + "loss": 0.0164, + "step": 4850 + }, + { + "epoch": 19.5, + "grad_norm": 0.0010457661701366305, + "learning_rate": 9.072913229957111e-05, + "loss": 0.0104, + "step": 4875 + }, + { + "epoch": 19.6, + "grad_norm": 0.0012097499566152692, + "learning_rate": 8.907951171230617e-05, + "loss": 0.0039, + "step": 4900 + }, + { + "epoch": 19.7, + "grad_norm": 0.04536284878849983, + "learning_rate": 8.742989112504124e-05, + "loss": 0.0154, + "step": 4925 + }, + { + "epoch": 19.8, + "grad_norm": 29.97249984741211, + "learning_rate": 8.578027053777631e-05, + "loss": 0.0246, + "step": 4950 + }, + { + "epoch": 19.9, + "grad_norm": 0.01911325752735138, + "learning_rate": 8.413064995051139e-05, + "loss": 0.0176, + "step": 4975 + }, + { + "epoch": 20.0, + "grad_norm": 0.3129942715167999, + "learning_rate": 8.248102936324645e-05, + "loss": 0.0029, + "step": 5000 + }, + { + "epoch": 20.1, + "grad_norm": 0.04958798363804817, + "learning_rate": 8.083140877598153e-05, + "loss": 0.0139, + "step": 5025 + }, + { + "epoch": 20.2, + "grad_norm": 0.003336989786475897, + "learning_rate": 7.91817881887166e-05, + "loss": 0.0159, + "step": 5050 + }, + { + "epoch": 20.3, + "grad_norm": 0.001012949156574905, + "learning_rate": 7.753216760145167e-05, + "loss": 0.0095, + "step": 5075 + }, + { + "epoch": 20.4, + "grad_norm": 0.0019106008112430573, + "learning_rate": 7.588254701418674e-05, + "loss": 0.0113, + "step": 5100 + }, + { + "epoch": 20.5, + "grad_norm": 0.034077223390340805, + "learning_rate": 7.423292642692181e-05, + "loss": 0.0101, + "step": 5125 + }, + { + "epoch": 20.6, + "grad_norm": 0.0011715433793142438, + "learning_rate": 7.258330583965688e-05, + "loss": 0.0111, + "step": 5150 + }, + { + "epoch": 20.7, + "grad_norm": 0.06765557825565338, + "learning_rate": 7.093368525239196e-05, + "loss": 0.0066, + "step": 5175 + }, + { + "epoch": 20.8, + "grad_norm": 0.0008974962402135134, + "learning_rate": 6.928406466512702e-05, + "loss": 0.0058, + "step": 5200 + }, + { + "epoch": 20.9, + "grad_norm": 0.00472952239215374, + "learning_rate": 6.763444407786209e-05, + "loss": 0.0027, + "step": 5225 + }, + { + "epoch": 21.0, + "grad_norm": 0.0006938530132174492, + "learning_rate": 6.598482349059718e-05, + "loss": 0.0009, + "step": 5250 + }, + { + "epoch": 21.1, + "grad_norm": 0.0007072632433846593, + "learning_rate": 6.433520290333224e-05, + "loss": 0.0015, + "step": 5275 + }, + { + "epoch": 21.2, + "grad_norm": 0.016313398256897926, + "learning_rate": 6.26855823160673e-05, + "loss": 0.0084, + "step": 5300 + }, + { + "epoch": 21.3, + "grad_norm": 0.0006648830021731555, + "learning_rate": 6.103596172880238e-05, + "loss": 0.0093, + "step": 5325 + }, + { + "epoch": 21.4, + "grad_norm": 0.0005273331771604717, + "learning_rate": 5.938634114153745e-05, + "loss": 0.0096, + "step": 5350 + }, + { + "epoch": 21.5, + "grad_norm": 0.0022026444785296917, + "learning_rate": 5.7736720554272516e-05, + "loss": 0.0107, + "step": 5375 + }, + { + "epoch": 21.6, + "grad_norm": 0.0009771068580448627, + "learning_rate": 5.608709996700759e-05, + "loss": 0.0017, + "step": 5400 + }, + { + "epoch": 21.7, + "grad_norm": 0.0006267099524848163, + "learning_rate": 5.443747937974266e-05, + "loss": 0.0032, + "step": 5425 + }, + { + "epoch": 21.8, + "grad_norm": 0.0006058421568013728, + "learning_rate": 5.278785879247773e-05, + "loss": 0.004, + "step": 5450 + }, + { + "epoch": 21.9, + "grad_norm": 0.0007795215351507068, + "learning_rate": 5.113823820521281e-05, + "loss": 0.0004, + "step": 5475 + }, + { + "epoch": 22.0, + "grad_norm": 0.013498655520379543, + "learning_rate": 4.948861761794787e-05, + "loss": 0.0035, + "step": 5500 + }, + { + "epoch": 22.1, + "grad_norm": 0.0007379506714642048, + "learning_rate": 4.783899703068295e-05, + "loss": 0.0085, + "step": 5525 + }, + { + "epoch": 22.2, + "grad_norm": 0.0010062415385618806, + "learning_rate": 4.618937644341802e-05, + "loss": 0.0001, + "step": 5550 + }, + { + "epoch": 22.3, + "grad_norm": 0.002069108886644244, + "learning_rate": 4.453975585615309e-05, + "loss": 0.0153, + "step": 5575 + }, + { + "epoch": 22.4, + "grad_norm": 0.0005407587741501629, + "learning_rate": 4.2890135268888156e-05, + "loss": 0.0088, + "step": 5600 + }, + { + "epoch": 22.5, + "grad_norm": 0.010531233623623848, + "learning_rate": 4.1240514681623226e-05, + "loss": 0.0002, + "step": 5625 + }, + { + "epoch": 22.6, + "grad_norm": 0.004741353914141655, + "learning_rate": 3.95908940943583e-05, + "loss": 0.0078, + "step": 5650 + }, + { + "epoch": 22.7, + "grad_norm": 0.0008237269939854741, + "learning_rate": 3.794127350709337e-05, + "loss": 0.0111, + "step": 5675 + }, + { + "epoch": 22.8, + "grad_norm": 0.0007508064154535532, + "learning_rate": 3.629165291982844e-05, + "loss": 0.0052, + "step": 5700 + }, + { + "epoch": 22.9, + "grad_norm": 0.0015650958521291614, + "learning_rate": 3.464203233256351e-05, + "loss": 0.0088, + "step": 5725 + }, + { + "epoch": 23.0, + "grad_norm": 0.0010173724731430411, + "learning_rate": 3.299241174529859e-05, + "loss": 0.0124, + "step": 5750 + }, + { + "epoch": 23.1, + "grad_norm": 43.15919494628906, + "learning_rate": 3.134279115803365e-05, + "loss": 0.004, + "step": 5775 + }, + { + "epoch": 23.2, + "grad_norm": 0.0005111348582431674, + "learning_rate": 2.9693170570768723e-05, + "loss": 0.0008, + "step": 5800 + }, + { + "epoch": 23.3, + "grad_norm": 0.009832415729761124, + "learning_rate": 2.8043549983503796e-05, + "loss": 0.0022, + "step": 5825 + }, + { + "epoch": 23.4, + "grad_norm": 0.0006390413036569953, + "learning_rate": 2.6393929396238866e-05, + "loss": 0.0001, + "step": 5850 + }, + { + "epoch": 23.5, + "grad_norm": 0.00047043984523043036, + "learning_rate": 2.4744308808973936e-05, + "loss": 0.0048, + "step": 5875 + }, + { + "epoch": 23.6, + "grad_norm": 0.0007061742944642901, + "learning_rate": 2.309468822170901e-05, + "loss": 0.0088, + "step": 5900 + }, + { + "epoch": 23.7, + "grad_norm": 0.0004057694459334016, + "learning_rate": 2.1445067634444078e-05, + "loss": 0.0, + "step": 5925 + }, + { + "epoch": 23.8, + "grad_norm": 0.0005023757694289088, + "learning_rate": 1.979544704717915e-05, + "loss": 0.001, + "step": 5950 + }, + { + "epoch": 23.9, + "grad_norm": 0.001338923815637827, + "learning_rate": 1.814582645991422e-05, + "loss": 0.0187, + "step": 5975 + }, + { + "epoch": 24.0, + "grad_norm": 0.018048524856567383, + "learning_rate": 1.6496205872649294e-05, + "loss": 0.0001, + "step": 6000 + } + ], + "logging_steps": 25, + "max_steps": 6250, + "num_input_tokens_seen": 0, + "num_train_epochs": 25, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 304916889600000.0, + "train_batch_size": 20, + "trial_name": null, + "trial_params": null +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6000/training_args.bin b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6000/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..ef3fc22e099df49861bca49d15da2b959cfcb6df --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6000/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ec1d8ed36c45af65d67a23de6c5216a486932bfb09b720e1a4417b82eaaa215d +size 5304 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6000/vocab.txt b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6000/vocab.txt new file mode 100644 index 0000000000000000000000000000000000000000..cc9ad53732e21ea9180f18e1fb67a38cc5702d4c --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6000/vocab.txt @@ -0,0 +1,4101 @@ +[PAD] +[UNK] +[CLS] +[SEP] +[MASK] +AAAAAA +AAAAAT +AAAAAC +AAAAAG +AAAATA +AAAATT +AAAATC +AAAATG +AAAACA +AAAACT +AAAACC +AAAACG +AAAAGA +AAAAGT +AAAAGC +AAAAGG +AAATAA +AAATAT +AAATAC +AAATAG +AAATTA +AAATTT +AAATTC +AAATTG +AAATCA +AAATCT +AAATCC +AAATCG +AAATGA +AAATGT +AAATGC +AAATGG +AAACAA +AAACAT +AAACAC +AAACAG +AAACTA +AAACTT +AAACTC +AAACTG +AAACCA +AAACCT +AAACCC +AAACCG +AAACGA +AAACGT +AAACGC +AAACGG +AAAGAA +AAAGAT +AAAGAC +AAAGAG +AAAGTA +AAAGTT +AAAGTC +AAAGTG +AAAGCA +AAAGCT +AAAGCC +AAAGCG +AAAGGA +AAAGGT +AAAGGC +AAAGGG +AATAAA +AATAAT +AATAAC +AATAAG +AATATA +AATATT +AATATC +AATATG +AATACA +AATACT +AATACC +AATACG +AATAGA +AATAGT +AATAGC +AATAGG +AATTAA +AATTAT +AATTAC +AATTAG +AATTTA +AATTTT +AATTTC +AATTTG +AATTCA +AATTCT +AATTCC +AATTCG +AATTGA +AATTGT +AATTGC +AATTGG +AATCAA +AATCAT +AATCAC +AATCAG +AATCTA +AATCTT +AATCTC +AATCTG +AATCCA +AATCCT +AATCCC +AATCCG +AATCGA +AATCGT +AATCGC +AATCGG +AATGAA +AATGAT +AATGAC +AATGAG +AATGTA +AATGTT +AATGTC +AATGTG +AATGCA +AATGCT +AATGCC +AATGCG +AATGGA +AATGGT +AATGGC +AATGGG +AACAAA +AACAAT +AACAAC +AACAAG +AACATA +AACATT +AACATC +AACATG +AACACA +AACACT +AACACC +AACACG +AACAGA +AACAGT +AACAGC +AACAGG +AACTAA +AACTAT +AACTAC +AACTAG +AACTTA +AACTTT +AACTTC +AACTTG +AACTCA +AACTCT +AACTCC +AACTCG +AACTGA +AACTGT +AACTGC +AACTGG +AACCAA +AACCAT +AACCAC +AACCAG +AACCTA +AACCTT +AACCTC +AACCTG +AACCCA +AACCCT +AACCCC +AACCCG +AACCGA +AACCGT +AACCGC +AACCGG +AACGAA +AACGAT +AACGAC +AACGAG +AACGTA +AACGTT +AACGTC +AACGTG +AACGCA +AACGCT +AACGCC +AACGCG +AACGGA +AACGGT +AACGGC +AACGGG +AAGAAA +AAGAAT +AAGAAC +AAGAAG +AAGATA +AAGATT +AAGATC +AAGATG +AAGACA +AAGACT +AAGACC +AAGACG +AAGAGA +AAGAGT +AAGAGC +AAGAGG +AAGTAA +AAGTAT +AAGTAC +AAGTAG +AAGTTA +AAGTTT +AAGTTC +AAGTTG +AAGTCA +AAGTCT +AAGTCC +AAGTCG +AAGTGA +AAGTGT +AAGTGC +AAGTGG +AAGCAA +AAGCAT +AAGCAC +AAGCAG +AAGCTA +AAGCTT +AAGCTC +AAGCTG +AAGCCA +AAGCCT +AAGCCC +AAGCCG +AAGCGA +AAGCGT +AAGCGC +AAGCGG +AAGGAA +AAGGAT +AAGGAC +AAGGAG +AAGGTA +AAGGTT +AAGGTC +AAGGTG +AAGGCA +AAGGCT +AAGGCC +AAGGCG +AAGGGA +AAGGGT +AAGGGC +AAGGGG +ATAAAA +ATAAAT +ATAAAC +ATAAAG +ATAATA +ATAATT +ATAATC +ATAATG +ATAACA +ATAACT +ATAACC +ATAACG +ATAAGA +ATAAGT +ATAAGC +ATAAGG +ATATAA +ATATAT +ATATAC +ATATAG +ATATTA +ATATTT +ATATTC +ATATTG +ATATCA +ATATCT +ATATCC +ATATCG +ATATGA +ATATGT +ATATGC +ATATGG +ATACAA +ATACAT +ATACAC +ATACAG +ATACTA +ATACTT +ATACTC +ATACTG +ATACCA +ATACCT +ATACCC +ATACCG +ATACGA +ATACGT +ATACGC +ATACGG +ATAGAA +ATAGAT +ATAGAC +ATAGAG +ATAGTA +ATAGTT +ATAGTC +ATAGTG +ATAGCA +ATAGCT +ATAGCC +ATAGCG +ATAGGA +ATAGGT +ATAGGC +ATAGGG +ATTAAA +ATTAAT +ATTAAC +ATTAAG +ATTATA +ATTATT +ATTATC +ATTATG +ATTACA +ATTACT +ATTACC +ATTACG +ATTAGA +ATTAGT +ATTAGC +ATTAGG +ATTTAA +ATTTAT +ATTTAC +ATTTAG +ATTTTA +ATTTTT +ATTTTC +ATTTTG +ATTTCA +ATTTCT +ATTTCC +ATTTCG +ATTTGA +ATTTGT +ATTTGC +ATTTGG +ATTCAA +ATTCAT +ATTCAC +ATTCAG +ATTCTA +ATTCTT +ATTCTC +ATTCTG +ATTCCA +ATTCCT +ATTCCC +ATTCCG +ATTCGA +ATTCGT +ATTCGC +ATTCGG +ATTGAA +ATTGAT +ATTGAC +ATTGAG +ATTGTA +ATTGTT +ATTGTC +ATTGTG +ATTGCA +ATTGCT +ATTGCC +ATTGCG +ATTGGA +ATTGGT +ATTGGC +ATTGGG +ATCAAA +ATCAAT +ATCAAC +ATCAAG +ATCATA +ATCATT +ATCATC +ATCATG +ATCACA +ATCACT +ATCACC +ATCACG +ATCAGA +ATCAGT +ATCAGC +ATCAGG +ATCTAA +ATCTAT +ATCTAC +ATCTAG +ATCTTA +ATCTTT +ATCTTC +ATCTTG +ATCTCA +ATCTCT +ATCTCC +ATCTCG +ATCTGA +ATCTGT +ATCTGC +ATCTGG +ATCCAA +ATCCAT +ATCCAC +ATCCAG +ATCCTA +ATCCTT +ATCCTC +ATCCTG +ATCCCA +ATCCCT +ATCCCC +ATCCCG +ATCCGA +ATCCGT +ATCCGC +ATCCGG +ATCGAA +ATCGAT +ATCGAC +ATCGAG +ATCGTA +ATCGTT +ATCGTC +ATCGTG +ATCGCA +ATCGCT +ATCGCC +ATCGCG +ATCGGA +ATCGGT +ATCGGC +ATCGGG +ATGAAA +ATGAAT +ATGAAC +ATGAAG +ATGATA +ATGATT +ATGATC +ATGATG +ATGACA +ATGACT +ATGACC +ATGACG +ATGAGA +ATGAGT +ATGAGC +ATGAGG +ATGTAA +ATGTAT +ATGTAC +ATGTAG +ATGTTA +ATGTTT +ATGTTC +ATGTTG +ATGTCA +ATGTCT +ATGTCC +ATGTCG +ATGTGA +ATGTGT +ATGTGC +ATGTGG +ATGCAA +ATGCAT +ATGCAC +ATGCAG +ATGCTA +ATGCTT +ATGCTC +ATGCTG +ATGCCA +ATGCCT +ATGCCC +ATGCCG +ATGCGA +ATGCGT +ATGCGC +ATGCGG +ATGGAA +ATGGAT +ATGGAC +ATGGAG +ATGGTA +ATGGTT +ATGGTC +ATGGTG +ATGGCA +ATGGCT +ATGGCC +ATGGCG +ATGGGA +ATGGGT +ATGGGC +ATGGGG +ACAAAA +ACAAAT +ACAAAC +ACAAAG +ACAATA +ACAATT +ACAATC +ACAATG +ACAACA +ACAACT +ACAACC +ACAACG +ACAAGA +ACAAGT +ACAAGC +ACAAGG +ACATAA +ACATAT +ACATAC +ACATAG +ACATTA +ACATTT +ACATTC +ACATTG +ACATCA +ACATCT +ACATCC +ACATCG +ACATGA +ACATGT +ACATGC +ACATGG +ACACAA +ACACAT +ACACAC +ACACAG +ACACTA +ACACTT +ACACTC +ACACTG +ACACCA +ACACCT +ACACCC +ACACCG +ACACGA +ACACGT +ACACGC +ACACGG +ACAGAA +ACAGAT +ACAGAC +ACAGAG +ACAGTA +ACAGTT +ACAGTC +ACAGTG +ACAGCA +ACAGCT +ACAGCC +ACAGCG +ACAGGA +ACAGGT +ACAGGC +ACAGGG +ACTAAA +ACTAAT +ACTAAC +ACTAAG +ACTATA +ACTATT +ACTATC +ACTATG +ACTACA +ACTACT +ACTACC +ACTACG +ACTAGA +ACTAGT +ACTAGC +ACTAGG +ACTTAA +ACTTAT +ACTTAC +ACTTAG +ACTTTA +ACTTTT +ACTTTC +ACTTTG +ACTTCA +ACTTCT +ACTTCC +ACTTCG +ACTTGA +ACTTGT +ACTTGC +ACTTGG +ACTCAA +ACTCAT +ACTCAC +ACTCAG +ACTCTA +ACTCTT +ACTCTC +ACTCTG +ACTCCA +ACTCCT +ACTCCC +ACTCCG +ACTCGA +ACTCGT +ACTCGC +ACTCGG +ACTGAA +ACTGAT +ACTGAC +ACTGAG +ACTGTA +ACTGTT +ACTGTC +ACTGTG +ACTGCA +ACTGCT +ACTGCC +ACTGCG +ACTGGA +ACTGGT +ACTGGC +ACTGGG +ACCAAA +ACCAAT +ACCAAC +ACCAAG +ACCATA +ACCATT +ACCATC +ACCATG +ACCACA +ACCACT +ACCACC +ACCACG +ACCAGA +ACCAGT +ACCAGC +ACCAGG +ACCTAA +ACCTAT +ACCTAC +ACCTAG +ACCTTA +ACCTTT +ACCTTC +ACCTTG +ACCTCA +ACCTCT +ACCTCC +ACCTCG +ACCTGA +ACCTGT +ACCTGC +ACCTGG +ACCCAA +ACCCAT +ACCCAC +ACCCAG +ACCCTA +ACCCTT +ACCCTC +ACCCTG +ACCCCA +ACCCCT +ACCCCC +ACCCCG +ACCCGA +ACCCGT +ACCCGC +ACCCGG +ACCGAA +ACCGAT +ACCGAC +ACCGAG +ACCGTA +ACCGTT +ACCGTC +ACCGTG +ACCGCA +ACCGCT +ACCGCC +ACCGCG +ACCGGA +ACCGGT +ACCGGC +ACCGGG +ACGAAA +ACGAAT +ACGAAC +ACGAAG +ACGATA +ACGATT +ACGATC +ACGATG +ACGACA +ACGACT +ACGACC +ACGACG +ACGAGA +ACGAGT +ACGAGC +ACGAGG +ACGTAA +ACGTAT +ACGTAC +ACGTAG +ACGTTA +ACGTTT +ACGTTC +ACGTTG +ACGTCA +ACGTCT +ACGTCC +ACGTCG +ACGTGA +ACGTGT +ACGTGC +ACGTGG +ACGCAA +ACGCAT +ACGCAC +ACGCAG +ACGCTA +ACGCTT +ACGCTC +ACGCTG +ACGCCA +ACGCCT +ACGCCC +ACGCCG +ACGCGA +ACGCGT +ACGCGC +ACGCGG +ACGGAA +ACGGAT +ACGGAC +ACGGAG +ACGGTA +ACGGTT +ACGGTC +ACGGTG +ACGGCA +ACGGCT +ACGGCC +ACGGCG +ACGGGA +ACGGGT +ACGGGC +ACGGGG +AGAAAA +AGAAAT +AGAAAC +AGAAAG +AGAATA +AGAATT +AGAATC +AGAATG +AGAACA +AGAACT +AGAACC +AGAACG +AGAAGA +AGAAGT +AGAAGC +AGAAGG +AGATAA +AGATAT +AGATAC +AGATAG +AGATTA +AGATTT +AGATTC +AGATTG +AGATCA +AGATCT +AGATCC +AGATCG +AGATGA +AGATGT +AGATGC +AGATGG +AGACAA +AGACAT +AGACAC +AGACAG +AGACTA +AGACTT +AGACTC +AGACTG +AGACCA +AGACCT +AGACCC +AGACCG +AGACGA +AGACGT +AGACGC +AGACGG +AGAGAA +AGAGAT +AGAGAC +AGAGAG +AGAGTA +AGAGTT +AGAGTC +AGAGTG +AGAGCA +AGAGCT +AGAGCC +AGAGCG +AGAGGA +AGAGGT +AGAGGC +AGAGGG +AGTAAA +AGTAAT +AGTAAC +AGTAAG +AGTATA +AGTATT +AGTATC +AGTATG +AGTACA +AGTACT +AGTACC +AGTACG +AGTAGA +AGTAGT +AGTAGC +AGTAGG +AGTTAA +AGTTAT +AGTTAC +AGTTAG +AGTTTA +AGTTTT +AGTTTC +AGTTTG +AGTTCA +AGTTCT +AGTTCC +AGTTCG +AGTTGA +AGTTGT +AGTTGC +AGTTGG +AGTCAA +AGTCAT +AGTCAC +AGTCAG +AGTCTA +AGTCTT +AGTCTC +AGTCTG +AGTCCA +AGTCCT +AGTCCC +AGTCCG +AGTCGA +AGTCGT +AGTCGC +AGTCGG +AGTGAA +AGTGAT +AGTGAC +AGTGAG +AGTGTA +AGTGTT +AGTGTC +AGTGTG +AGTGCA +AGTGCT +AGTGCC +AGTGCG +AGTGGA +AGTGGT +AGTGGC +AGTGGG +AGCAAA +AGCAAT +AGCAAC +AGCAAG +AGCATA +AGCATT +AGCATC +AGCATG +AGCACA +AGCACT +AGCACC +AGCACG +AGCAGA +AGCAGT +AGCAGC +AGCAGG +AGCTAA +AGCTAT +AGCTAC +AGCTAG +AGCTTA +AGCTTT +AGCTTC +AGCTTG +AGCTCA +AGCTCT +AGCTCC +AGCTCG +AGCTGA +AGCTGT +AGCTGC +AGCTGG +AGCCAA +AGCCAT +AGCCAC +AGCCAG +AGCCTA +AGCCTT +AGCCTC +AGCCTG +AGCCCA +AGCCCT +AGCCCC +AGCCCG +AGCCGA +AGCCGT +AGCCGC +AGCCGG +AGCGAA +AGCGAT +AGCGAC +AGCGAG +AGCGTA +AGCGTT +AGCGTC +AGCGTG +AGCGCA +AGCGCT +AGCGCC +AGCGCG +AGCGGA +AGCGGT +AGCGGC +AGCGGG +AGGAAA +AGGAAT +AGGAAC +AGGAAG +AGGATA +AGGATT +AGGATC +AGGATG +AGGACA +AGGACT +AGGACC +AGGACG +AGGAGA +AGGAGT +AGGAGC +AGGAGG +AGGTAA +AGGTAT +AGGTAC +AGGTAG +AGGTTA +AGGTTT +AGGTTC +AGGTTG +AGGTCA +AGGTCT +AGGTCC +AGGTCG +AGGTGA +AGGTGT +AGGTGC +AGGTGG +AGGCAA +AGGCAT +AGGCAC +AGGCAG +AGGCTA +AGGCTT +AGGCTC +AGGCTG +AGGCCA +AGGCCT +AGGCCC +AGGCCG +AGGCGA +AGGCGT +AGGCGC +AGGCGG +AGGGAA +AGGGAT +AGGGAC +AGGGAG +AGGGTA +AGGGTT +AGGGTC +AGGGTG +AGGGCA +AGGGCT +AGGGCC +AGGGCG +AGGGGA +AGGGGT +AGGGGC +AGGGGG +TAAAAA +TAAAAT +TAAAAC +TAAAAG +TAAATA +TAAATT +TAAATC +TAAATG +TAAACA +TAAACT +TAAACC +TAAACG +TAAAGA +TAAAGT +TAAAGC +TAAAGG +TAATAA +TAATAT +TAATAC +TAATAG +TAATTA +TAATTT +TAATTC +TAATTG +TAATCA +TAATCT +TAATCC +TAATCG +TAATGA +TAATGT +TAATGC +TAATGG +TAACAA +TAACAT +TAACAC +TAACAG +TAACTA +TAACTT +TAACTC +TAACTG +TAACCA +TAACCT +TAACCC +TAACCG +TAACGA +TAACGT +TAACGC +TAACGG +TAAGAA +TAAGAT +TAAGAC +TAAGAG +TAAGTA +TAAGTT +TAAGTC +TAAGTG +TAAGCA +TAAGCT +TAAGCC +TAAGCG +TAAGGA +TAAGGT +TAAGGC +TAAGGG +TATAAA +TATAAT +TATAAC +TATAAG +TATATA +TATATT +TATATC +TATATG +TATACA +TATACT +TATACC +TATACG +TATAGA +TATAGT +TATAGC +TATAGG +TATTAA +TATTAT +TATTAC +TATTAG +TATTTA +TATTTT +TATTTC +TATTTG +TATTCA +TATTCT +TATTCC +TATTCG +TATTGA +TATTGT +TATTGC +TATTGG +TATCAA +TATCAT +TATCAC +TATCAG +TATCTA +TATCTT +TATCTC +TATCTG +TATCCA +TATCCT +TATCCC +TATCCG +TATCGA +TATCGT +TATCGC +TATCGG +TATGAA +TATGAT +TATGAC +TATGAG +TATGTA +TATGTT +TATGTC +TATGTG +TATGCA +TATGCT +TATGCC +TATGCG +TATGGA +TATGGT +TATGGC +TATGGG +TACAAA +TACAAT +TACAAC +TACAAG +TACATA +TACATT +TACATC +TACATG +TACACA +TACACT +TACACC +TACACG +TACAGA +TACAGT +TACAGC +TACAGG +TACTAA +TACTAT +TACTAC +TACTAG +TACTTA +TACTTT +TACTTC +TACTTG +TACTCA +TACTCT +TACTCC +TACTCG +TACTGA +TACTGT +TACTGC +TACTGG +TACCAA +TACCAT +TACCAC +TACCAG +TACCTA +TACCTT +TACCTC +TACCTG +TACCCA +TACCCT +TACCCC +TACCCG +TACCGA +TACCGT +TACCGC +TACCGG +TACGAA +TACGAT +TACGAC +TACGAG +TACGTA +TACGTT +TACGTC +TACGTG +TACGCA +TACGCT +TACGCC +TACGCG +TACGGA +TACGGT +TACGGC +TACGGG +TAGAAA +TAGAAT +TAGAAC +TAGAAG +TAGATA +TAGATT +TAGATC +TAGATG +TAGACA +TAGACT +TAGACC +TAGACG +TAGAGA +TAGAGT +TAGAGC +TAGAGG +TAGTAA +TAGTAT +TAGTAC +TAGTAG +TAGTTA +TAGTTT +TAGTTC +TAGTTG +TAGTCA +TAGTCT +TAGTCC +TAGTCG +TAGTGA +TAGTGT +TAGTGC +TAGTGG +TAGCAA +TAGCAT +TAGCAC +TAGCAG +TAGCTA +TAGCTT +TAGCTC +TAGCTG +TAGCCA +TAGCCT +TAGCCC +TAGCCG +TAGCGA +TAGCGT +TAGCGC +TAGCGG +TAGGAA +TAGGAT +TAGGAC +TAGGAG +TAGGTA +TAGGTT +TAGGTC +TAGGTG +TAGGCA +TAGGCT +TAGGCC +TAGGCG +TAGGGA +TAGGGT +TAGGGC +TAGGGG +TTAAAA +TTAAAT +TTAAAC +TTAAAG +TTAATA +TTAATT +TTAATC +TTAATG +TTAACA +TTAACT +TTAACC +TTAACG +TTAAGA +TTAAGT +TTAAGC +TTAAGG +TTATAA +TTATAT +TTATAC +TTATAG +TTATTA +TTATTT +TTATTC +TTATTG +TTATCA +TTATCT +TTATCC +TTATCG +TTATGA +TTATGT +TTATGC +TTATGG +TTACAA +TTACAT +TTACAC +TTACAG +TTACTA +TTACTT +TTACTC +TTACTG +TTACCA +TTACCT +TTACCC +TTACCG +TTACGA +TTACGT +TTACGC +TTACGG +TTAGAA +TTAGAT +TTAGAC +TTAGAG +TTAGTA +TTAGTT +TTAGTC +TTAGTG +TTAGCA +TTAGCT +TTAGCC +TTAGCG +TTAGGA +TTAGGT +TTAGGC +TTAGGG +TTTAAA +TTTAAT +TTTAAC +TTTAAG +TTTATA +TTTATT +TTTATC +TTTATG +TTTACA +TTTACT +TTTACC +TTTACG +TTTAGA +TTTAGT +TTTAGC +TTTAGG +TTTTAA +TTTTAT +TTTTAC +TTTTAG +TTTTTA +TTTTTT +TTTTTC +TTTTTG +TTTTCA +TTTTCT +TTTTCC +TTTTCG +TTTTGA +TTTTGT +TTTTGC +TTTTGG +TTTCAA +TTTCAT +TTTCAC +TTTCAG +TTTCTA +TTTCTT +TTTCTC +TTTCTG +TTTCCA +TTTCCT +TTTCCC +TTTCCG +TTTCGA +TTTCGT +TTTCGC +TTTCGG +TTTGAA +TTTGAT +TTTGAC +TTTGAG +TTTGTA +TTTGTT +TTTGTC +TTTGTG +TTTGCA +TTTGCT +TTTGCC +TTTGCG +TTTGGA +TTTGGT +TTTGGC +TTTGGG +TTCAAA +TTCAAT +TTCAAC +TTCAAG +TTCATA +TTCATT +TTCATC +TTCATG +TTCACA +TTCACT +TTCACC +TTCACG +TTCAGA +TTCAGT +TTCAGC +TTCAGG +TTCTAA +TTCTAT +TTCTAC +TTCTAG +TTCTTA +TTCTTT +TTCTTC +TTCTTG +TTCTCA +TTCTCT +TTCTCC +TTCTCG +TTCTGA +TTCTGT +TTCTGC +TTCTGG +TTCCAA +TTCCAT +TTCCAC +TTCCAG +TTCCTA +TTCCTT +TTCCTC +TTCCTG +TTCCCA +TTCCCT +TTCCCC +TTCCCG +TTCCGA +TTCCGT +TTCCGC +TTCCGG +TTCGAA +TTCGAT +TTCGAC +TTCGAG +TTCGTA +TTCGTT +TTCGTC +TTCGTG +TTCGCA +TTCGCT +TTCGCC +TTCGCG +TTCGGA +TTCGGT +TTCGGC +TTCGGG +TTGAAA +TTGAAT +TTGAAC +TTGAAG +TTGATA +TTGATT +TTGATC +TTGATG +TTGACA +TTGACT +TTGACC +TTGACG +TTGAGA +TTGAGT +TTGAGC +TTGAGG +TTGTAA +TTGTAT +TTGTAC +TTGTAG +TTGTTA +TTGTTT +TTGTTC +TTGTTG +TTGTCA +TTGTCT +TTGTCC +TTGTCG +TTGTGA +TTGTGT +TTGTGC +TTGTGG +TTGCAA +TTGCAT +TTGCAC +TTGCAG +TTGCTA +TTGCTT +TTGCTC +TTGCTG +TTGCCA +TTGCCT +TTGCCC +TTGCCG +TTGCGA +TTGCGT +TTGCGC +TTGCGG +TTGGAA +TTGGAT +TTGGAC +TTGGAG +TTGGTA +TTGGTT +TTGGTC +TTGGTG +TTGGCA +TTGGCT +TTGGCC +TTGGCG +TTGGGA +TTGGGT +TTGGGC +TTGGGG +TCAAAA +TCAAAT +TCAAAC +TCAAAG +TCAATA +TCAATT +TCAATC +TCAATG +TCAACA +TCAACT +TCAACC +TCAACG +TCAAGA +TCAAGT +TCAAGC +TCAAGG +TCATAA +TCATAT +TCATAC +TCATAG +TCATTA +TCATTT +TCATTC +TCATTG +TCATCA +TCATCT +TCATCC +TCATCG +TCATGA +TCATGT +TCATGC +TCATGG +TCACAA +TCACAT +TCACAC +TCACAG +TCACTA +TCACTT +TCACTC +TCACTG +TCACCA +TCACCT +TCACCC +TCACCG +TCACGA +TCACGT +TCACGC +TCACGG +TCAGAA +TCAGAT +TCAGAC +TCAGAG +TCAGTA +TCAGTT +TCAGTC +TCAGTG +TCAGCA +TCAGCT +TCAGCC +TCAGCG +TCAGGA +TCAGGT +TCAGGC +TCAGGG +TCTAAA +TCTAAT +TCTAAC +TCTAAG +TCTATA +TCTATT +TCTATC +TCTATG +TCTACA +TCTACT +TCTACC +TCTACG +TCTAGA +TCTAGT +TCTAGC +TCTAGG +TCTTAA +TCTTAT +TCTTAC +TCTTAG +TCTTTA +TCTTTT +TCTTTC +TCTTTG +TCTTCA +TCTTCT +TCTTCC +TCTTCG +TCTTGA +TCTTGT +TCTTGC +TCTTGG +TCTCAA +TCTCAT +TCTCAC +TCTCAG +TCTCTA +TCTCTT +TCTCTC +TCTCTG +TCTCCA +TCTCCT +TCTCCC +TCTCCG +TCTCGA +TCTCGT +TCTCGC +TCTCGG +TCTGAA +TCTGAT +TCTGAC +TCTGAG +TCTGTA +TCTGTT +TCTGTC +TCTGTG +TCTGCA +TCTGCT +TCTGCC +TCTGCG +TCTGGA +TCTGGT +TCTGGC +TCTGGG +TCCAAA +TCCAAT +TCCAAC +TCCAAG +TCCATA +TCCATT +TCCATC +TCCATG +TCCACA +TCCACT +TCCACC +TCCACG +TCCAGA +TCCAGT +TCCAGC +TCCAGG +TCCTAA +TCCTAT +TCCTAC +TCCTAG +TCCTTA +TCCTTT +TCCTTC +TCCTTG +TCCTCA +TCCTCT +TCCTCC +TCCTCG +TCCTGA +TCCTGT +TCCTGC +TCCTGG +TCCCAA +TCCCAT +TCCCAC +TCCCAG +TCCCTA +TCCCTT +TCCCTC +TCCCTG +TCCCCA +TCCCCT +TCCCCC +TCCCCG +TCCCGA +TCCCGT +TCCCGC +TCCCGG +TCCGAA +TCCGAT +TCCGAC +TCCGAG +TCCGTA +TCCGTT +TCCGTC +TCCGTG +TCCGCA +TCCGCT +TCCGCC +TCCGCG +TCCGGA +TCCGGT +TCCGGC +TCCGGG +TCGAAA +TCGAAT +TCGAAC +TCGAAG +TCGATA +TCGATT +TCGATC +TCGATG +TCGACA +TCGACT +TCGACC +TCGACG +TCGAGA +TCGAGT +TCGAGC +TCGAGG +TCGTAA +TCGTAT +TCGTAC +TCGTAG +TCGTTA +TCGTTT +TCGTTC +TCGTTG +TCGTCA +TCGTCT +TCGTCC +TCGTCG +TCGTGA +TCGTGT +TCGTGC +TCGTGG +TCGCAA +TCGCAT +TCGCAC +TCGCAG +TCGCTA +TCGCTT +TCGCTC +TCGCTG +TCGCCA +TCGCCT +TCGCCC +TCGCCG +TCGCGA +TCGCGT +TCGCGC +TCGCGG +TCGGAA +TCGGAT +TCGGAC +TCGGAG +TCGGTA +TCGGTT +TCGGTC +TCGGTG +TCGGCA +TCGGCT +TCGGCC +TCGGCG +TCGGGA +TCGGGT +TCGGGC +TCGGGG +TGAAAA +TGAAAT +TGAAAC +TGAAAG +TGAATA +TGAATT +TGAATC +TGAATG +TGAACA +TGAACT +TGAACC +TGAACG +TGAAGA +TGAAGT +TGAAGC +TGAAGG +TGATAA +TGATAT +TGATAC +TGATAG +TGATTA +TGATTT +TGATTC +TGATTG +TGATCA +TGATCT +TGATCC +TGATCG +TGATGA +TGATGT +TGATGC +TGATGG +TGACAA +TGACAT +TGACAC +TGACAG +TGACTA +TGACTT +TGACTC +TGACTG +TGACCA +TGACCT +TGACCC +TGACCG +TGACGA +TGACGT +TGACGC +TGACGG +TGAGAA +TGAGAT +TGAGAC +TGAGAG +TGAGTA +TGAGTT +TGAGTC +TGAGTG +TGAGCA +TGAGCT +TGAGCC +TGAGCG +TGAGGA +TGAGGT +TGAGGC +TGAGGG +TGTAAA +TGTAAT +TGTAAC +TGTAAG +TGTATA +TGTATT +TGTATC +TGTATG +TGTACA +TGTACT +TGTACC +TGTACG +TGTAGA +TGTAGT +TGTAGC +TGTAGG +TGTTAA +TGTTAT +TGTTAC +TGTTAG +TGTTTA +TGTTTT +TGTTTC +TGTTTG +TGTTCA +TGTTCT +TGTTCC +TGTTCG +TGTTGA +TGTTGT +TGTTGC +TGTTGG +TGTCAA +TGTCAT +TGTCAC +TGTCAG +TGTCTA +TGTCTT +TGTCTC +TGTCTG +TGTCCA +TGTCCT +TGTCCC +TGTCCG +TGTCGA +TGTCGT +TGTCGC +TGTCGG +TGTGAA +TGTGAT +TGTGAC +TGTGAG +TGTGTA +TGTGTT +TGTGTC +TGTGTG +TGTGCA +TGTGCT +TGTGCC +TGTGCG +TGTGGA +TGTGGT +TGTGGC +TGTGGG +TGCAAA +TGCAAT +TGCAAC +TGCAAG +TGCATA +TGCATT +TGCATC +TGCATG +TGCACA +TGCACT +TGCACC +TGCACG +TGCAGA +TGCAGT +TGCAGC +TGCAGG +TGCTAA +TGCTAT +TGCTAC +TGCTAG +TGCTTA +TGCTTT +TGCTTC +TGCTTG +TGCTCA +TGCTCT +TGCTCC +TGCTCG +TGCTGA +TGCTGT +TGCTGC +TGCTGG +TGCCAA +TGCCAT +TGCCAC +TGCCAG +TGCCTA +TGCCTT +TGCCTC +TGCCTG +TGCCCA +TGCCCT +TGCCCC +TGCCCG +TGCCGA +TGCCGT +TGCCGC +TGCCGG +TGCGAA +TGCGAT +TGCGAC +TGCGAG +TGCGTA +TGCGTT +TGCGTC +TGCGTG +TGCGCA +TGCGCT +TGCGCC +TGCGCG +TGCGGA +TGCGGT +TGCGGC +TGCGGG +TGGAAA +TGGAAT +TGGAAC +TGGAAG +TGGATA +TGGATT +TGGATC +TGGATG +TGGACA +TGGACT +TGGACC +TGGACG +TGGAGA +TGGAGT +TGGAGC +TGGAGG +TGGTAA +TGGTAT +TGGTAC +TGGTAG +TGGTTA +TGGTTT +TGGTTC +TGGTTG +TGGTCA +TGGTCT +TGGTCC +TGGTCG +TGGTGA +TGGTGT +TGGTGC +TGGTGG +TGGCAA +TGGCAT +TGGCAC +TGGCAG +TGGCTA +TGGCTT +TGGCTC +TGGCTG +TGGCCA +TGGCCT +TGGCCC +TGGCCG +TGGCGA +TGGCGT +TGGCGC +TGGCGG +TGGGAA +TGGGAT +TGGGAC +TGGGAG +TGGGTA +TGGGTT +TGGGTC +TGGGTG +TGGGCA +TGGGCT +TGGGCC +TGGGCG +TGGGGA +TGGGGT +TGGGGC +TGGGGG +CAAAAA +CAAAAT +CAAAAC +CAAAAG +CAAATA +CAAATT +CAAATC +CAAATG +CAAACA +CAAACT +CAAACC +CAAACG +CAAAGA +CAAAGT +CAAAGC +CAAAGG +CAATAA +CAATAT +CAATAC +CAATAG +CAATTA +CAATTT +CAATTC +CAATTG +CAATCA +CAATCT +CAATCC +CAATCG +CAATGA +CAATGT +CAATGC +CAATGG +CAACAA +CAACAT +CAACAC +CAACAG +CAACTA +CAACTT +CAACTC +CAACTG +CAACCA +CAACCT +CAACCC +CAACCG +CAACGA +CAACGT +CAACGC +CAACGG +CAAGAA +CAAGAT +CAAGAC +CAAGAG +CAAGTA +CAAGTT +CAAGTC +CAAGTG +CAAGCA +CAAGCT +CAAGCC +CAAGCG +CAAGGA +CAAGGT +CAAGGC +CAAGGG +CATAAA +CATAAT +CATAAC +CATAAG +CATATA +CATATT +CATATC +CATATG +CATACA +CATACT +CATACC +CATACG +CATAGA +CATAGT +CATAGC +CATAGG +CATTAA +CATTAT +CATTAC +CATTAG +CATTTA +CATTTT +CATTTC +CATTTG +CATTCA +CATTCT +CATTCC +CATTCG +CATTGA +CATTGT +CATTGC +CATTGG +CATCAA +CATCAT +CATCAC +CATCAG +CATCTA +CATCTT +CATCTC +CATCTG +CATCCA +CATCCT +CATCCC +CATCCG +CATCGA +CATCGT +CATCGC +CATCGG +CATGAA +CATGAT +CATGAC +CATGAG +CATGTA +CATGTT +CATGTC +CATGTG +CATGCA +CATGCT +CATGCC +CATGCG +CATGGA +CATGGT +CATGGC +CATGGG +CACAAA +CACAAT +CACAAC +CACAAG +CACATA +CACATT +CACATC +CACATG +CACACA +CACACT +CACACC +CACACG +CACAGA +CACAGT +CACAGC +CACAGG +CACTAA +CACTAT +CACTAC +CACTAG +CACTTA +CACTTT +CACTTC +CACTTG +CACTCA +CACTCT +CACTCC +CACTCG +CACTGA +CACTGT +CACTGC +CACTGG +CACCAA +CACCAT +CACCAC +CACCAG +CACCTA +CACCTT +CACCTC +CACCTG +CACCCA +CACCCT +CACCCC +CACCCG +CACCGA +CACCGT +CACCGC +CACCGG +CACGAA +CACGAT +CACGAC +CACGAG +CACGTA +CACGTT +CACGTC +CACGTG +CACGCA +CACGCT +CACGCC +CACGCG +CACGGA +CACGGT +CACGGC +CACGGG +CAGAAA +CAGAAT +CAGAAC +CAGAAG +CAGATA +CAGATT +CAGATC +CAGATG +CAGACA +CAGACT +CAGACC +CAGACG +CAGAGA +CAGAGT +CAGAGC +CAGAGG +CAGTAA +CAGTAT +CAGTAC +CAGTAG +CAGTTA +CAGTTT +CAGTTC +CAGTTG +CAGTCA +CAGTCT +CAGTCC +CAGTCG +CAGTGA +CAGTGT +CAGTGC +CAGTGG +CAGCAA +CAGCAT +CAGCAC +CAGCAG +CAGCTA +CAGCTT +CAGCTC +CAGCTG +CAGCCA +CAGCCT +CAGCCC +CAGCCG +CAGCGA +CAGCGT +CAGCGC +CAGCGG +CAGGAA +CAGGAT +CAGGAC +CAGGAG +CAGGTA +CAGGTT +CAGGTC +CAGGTG +CAGGCA +CAGGCT +CAGGCC +CAGGCG +CAGGGA +CAGGGT +CAGGGC +CAGGGG +CTAAAA +CTAAAT +CTAAAC +CTAAAG +CTAATA +CTAATT +CTAATC +CTAATG +CTAACA +CTAACT +CTAACC +CTAACG +CTAAGA +CTAAGT +CTAAGC +CTAAGG +CTATAA +CTATAT +CTATAC +CTATAG +CTATTA +CTATTT +CTATTC +CTATTG +CTATCA +CTATCT +CTATCC +CTATCG +CTATGA +CTATGT +CTATGC +CTATGG +CTACAA +CTACAT +CTACAC +CTACAG +CTACTA +CTACTT +CTACTC +CTACTG +CTACCA +CTACCT +CTACCC +CTACCG +CTACGA +CTACGT +CTACGC +CTACGG +CTAGAA +CTAGAT +CTAGAC +CTAGAG +CTAGTA +CTAGTT +CTAGTC +CTAGTG +CTAGCA +CTAGCT +CTAGCC +CTAGCG +CTAGGA +CTAGGT +CTAGGC +CTAGGG +CTTAAA +CTTAAT +CTTAAC +CTTAAG +CTTATA +CTTATT +CTTATC +CTTATG +CTTACA +CTTACT +CTTACC +CTTACG +CTTAGA +CTTAGT +CTTAGC +CTTAGG +CTTTAA +CTTTAT +CTTTAC +CTTTAG +CTTTTA +CTTTTT +CTTTTC +CTTTTG +CTTTCA +CTTTCT +CTTTCC +CTTTCG +CTTTGA +CTTTGT +CTTTGC +CTTTGG +CTTCAA +CTTCAT +CTTCAC +CTTCAG +CTTCTA +CTTCTT +CTTCTC +CTTCTG +CTTCCA +CTTCCT +CTTCCC +CTTCCG +CTTCGA +CTTCGT +CTTCGC +CTTCGG +CTTGAA +CTTGAT +CTTGAC +CTTGAG +CTTGTA +CTTGTT +CTTGTC +CTTGTG +CTTGCA +CTTGCT +CTTGCC +CTTGCG +CTTGGA +CTTGGT +CTTGGC +CTTGGG +CTCAAA +CTCAAT +CTCAAC +CTCAAG +CTCATA +CTCATT +CTCATC +CTCATG +CTCACA +CTCACT +CTCACC +CTCACG +CTCAGA +CTCAGT +CTCAGC +CTCAGG +CTCTAA +CTCTAT +CTCTAC +CTCTAG +CTCTTA +CTCTTT +CTCTTC +CTCTTG +CTCTCA +CTCTCT +CTCTCC +CTCTCG +CTCTGA +CTCTGT +CTCTGC +CTCTGG +CTCCAA +CTCCAT +CTCCAC +CTCCAG +CTCCTA +CTCCTT +CTCCTC +CTCCTG +CTCCCA +CTCCCT +CTCCCC +CTCCCG +CTCCGA +CTCCGT +CTCCGC +CTCCGG +CTCGAA +CTCGAT +CTCGAC +CTCGAG +CTCGTA +CTCGTT +CTCGTC +CTCGTG +CTCGCA +CTCGCT +CTCGCC +CTCGCG +CTCGGA +CTCGGT +CTCGGC +CTCGGG +CTGAAA +CTGAAT +CTGAAC +CTGAAG +CTGATA +CTGATT +CTGATC +CTGATG +CTGACA +CTGACT +CTGACC +CTGACG +CTGAGA +CTGAGT +CTGAGC +CTGAGG +CTGTAA +CTGTAT +CTGTAC +CTGTAG +CTGTTA +CTGTTT +CTGTTC +CTGTTG +CTGTCA +CTGTCT +CTGTCC +CTGTCG +CTGTGA +CTGTGT +CTGTGC +CTGTGG +CTGCAA +CTGCAT +CTGCAC +CTGCAG +CTGCTA +CTGCTT +CTGCTC +CTGCTG +CTGCCA +CTGCCT +CTGCCC +CTGCCG +CTGCGA +CTGCGT +CTGCGC +CTGCGG +CTGGAA +CTGGAT +CTGGAC +CTGGAG +CTGGTA +CTGGTT +CTGGTC +CTGGTG +CTGGCA +CTGGCT +CTGGCC +CTGGCG +CTGGGA +CTGGGT +CTGGGC +CTGGGG +CCAAAA +CCAAAT +CCAAAC +CCAAAG +CCAATA +CCAATT +CCAATC +CCAATG +CCAACA +CCAACT +CCAACC +CCAACG +CCAAGA +CCAAGT +CCAAGC +CCAAGG +CCATAA +CCATAT +CCATAC +CCATAG +CCATTA +CCATTT +CCATTC +CCATTG +CCATCA +CCATCT +CCATCC +CCATCG +CCATGA +CCATGT +CCATGC +CCATGG +CCACAA +CCACAT +CCACAC +CCACAG +CCACTA +CCACTT +CCACTC +CCACTG +CCACCA +CCACCT +CCACCC +CCACCG +CCACGA +CCACGT +CCACGC +CCACGG +CCAGAA +CCAGAT +CCAGAC +CCAGAG +CCAGTA +CCAGTT +CCAGTC +CCAGTG +CCAGCA +CCAGCT +CCAGCC +CCAGCG +CCAGGA +CCAGGT +CCAGGC +CCAGGG +CCTAAA +CCTAAT +CCTAAC +CCTAAG +CCTATA +CCTATT +CCTATC +CCTATG +CCTACA +CCTACT +CCTACC +CCTACG +CCTAGA +CCTAGT +CCTAGC +CCTAGG +CCTTAA +CCTTAT +CCTTAC +CCTTAG +CCTTTA +CCTTTT +CCTTTC +CCTTTG +CCTTCA +CCTTCT +CCTTCC +CCTTCG +CCTTGA +CCTTGT +CCTTGC +CCTTGG +CCTCAA +CCTCAT +CCTCAC +CCTCAG +CCTCTA +CCTCTT +CCTCTC +CCTCTG +CCTCCA +CCTCCT +CCTCCC +CCTCCG +CCTCGA +CCTCGT +CCTCGC +CCTCGG +CCTGAA +CCTGAT +CCTGAC +CCTGAG +CCTGTA +CCTGTT +CCTGTC +CCTGTG +CCTGCA +CCTGCT +CCTGCC +CCTGCG +CCTGGA +CCTGGT +CCTGGC +CCTGGG +CCCAAA +CCCAAT +CCCAAC +CCCAAG +CCCATA +CCCATT +CCCATC +CCCATG +CCCACA +CCCACT +CCCACC +CCCACG +CCCAGA +CCCAGT +CCCAGC +CCCAGG +CCCTAA +CCCTAT +CCCTAC +CCCTAG +CCCTTA +CCCTTT +CCCTTC +CCCTTG +CCCTCA +CCCTCT +CCCTCC +CCCTCG +CCCTGA +CCCTGT +CCCTGC +CCCTGG +CCCCAA +CCCCAT +CCCCAC +CCCCAG +CCCCTA +CCCCTT +CCCCTC +CCCCTG +CCCCCA +CCCCCT +CCCCCC +CCCCCG +CCCCGA +CCCCGT +CCCCGC +CCCCGG +CCCGAA +CCCGAT +CCCGAC +CCCGAG +CCCGTA +CCCGTT +CCCGTC +CCCGTG +CCCGCA +CCCGCT +CCCGCC +CCCGCG +CCCGGA +CCCGGT +CCCGGC +CCCGGG +CCGAAA +CCGAAT +CCGAAC +CCGAAG +CCGATA +CCGATT +CCGATC +CCGATG +CCGACA +CCGACT +CCGACC +CCGACG +CCGAGA +CCGAGT +CCGAGC +CCGAGG +CCGTAA +CCGTAT +CCGTAC +CCGTAG +CCGTTA +CCGTTT +CCGTTC +CCGTTG +CCGTCA +CCGTCT +CCGTCC +CCGTCG +CCGTGA +CCGTGT +CCGTGC +CCGTGG +CCGCAA +CCGCAT +CCGCAC +CCGCAG +CCGCTA +CCGCTT +CCGCTC +CCGCTG +CCGCCA +CCGCCT +CCGCCC +CCGCCG +CCGCGA +CCGCGT +CCGCGC +CCGCGG +CCGGAA +CCGGAT +CCGGAC +CCGGAG +CCGGTA +CCGGTT +CCGGTC +CCGGTG +CCGGCA +CCGGCT +CCGGCC +CCGGCG +CCGGGA +CCGGGT +CCGGGC +CCGGGG +CGAAAA +CGAAAT +CGAAAC +CGAAAG +CGAATA +CGAATT +CGAATC +CGAATG +CGAACA +CGAACT +CGAACC +CGAACG +CGAAGA +CGAAGT +CGAAGC +CGAAGG +CGATAA +CGATAT +CGATAC +CGATAG +CGATTA +CGATTT +CGATTC +CGATTG +CGATCA +CGATCT +CGATCC +CGATCG +CGATGA +CGATGT +CGATGC +CGATGG +CGACAA +CGACAT +CGACAC +CGACAG +CGACTA +CGACTT +CGACTC +CGACTG +CGACCA +CGACCT +CGACCC +CGACCG +CGACGA +CGACGT +CGACGC +CGACGG +CGAGAA +CGAGAT +CGAGAC +CGAGAG +CGAGTA +CGAGTT +CGAGTC +CGAGTG +CGAGCA +CGAGCT +CGAGCC +CGAGCG +CGAGGA +CGAGGT +CGAGGC +CGAGGG +CGTAAA +CGTAAT +CGTAAC +CGTAAG +CGTATA +CGTATT +CGTATC +CGTATG +CGTACA +CGTACT +CGTACC +CGTACG +CGTAGA +CGTAGT +CGTAGC +CGTAGG +CGTTAA +CGTTAT +CGTTAC +CGTTAG +CGTTTA +CGTTTT +CGTTTC +CGTTTG +CGTTCA +CGTTCT +CGTTCC +CGTTCG +CGTTGA +CGTTGT +CGTTGC +CGTTGG +CGTCAA +CGTCAT +CGTCAC +CGTCAG +CGTCTA +CGTCTT +CGTCTC +CGTCTG +CGTCCA +CGTCCT +CGTCCC +CGTCCG +CGTCGA +CGTCGT +CGTCGC +CGTCGG +CGTGAA +CGTGAT +CGTGAC +CGTGAG +CGTGTA +CGTGTT +CGTGTC +CGTGTG +CGTGCA +CGTGCT +CGTGCC +CGTGCG +CGTGGA +CGTGGT +CGTGGC +CGTGGG +CGCAAA +CGCAAT +CGCAAC +CGCAAG +CGCATA +CGCATT +CGCATC +CGCATG +CGCACA +CGCACT +CGCACC +CGCACG +CGCAGA +CGCAGT +CGCAGC +CGCAGG +CGCTAA +CGCTAT +CGCTAC +CGCTAG +CGCTTA +CGCTTT +CGCTTC +CGCTTG +CGCTCA +CGCTCT +CGCTCC +CGCTCG +CGCTGA +CGCTGT +CGCTGC +CGCTGG +CGCCAA +CGCCAT +CGCCAC +CGCCAG +CGCCTA +CGCCTT +CGCCTC +CGCCTG +CGCCCA +CGCCCT +CGCCCC +CGCCCG +CGCCGA +CGCCGT +CGCCGC +CGCCGG +CGCGAA +CGCGAT +CGCGAC +CGCGAG +CGCGTA +CGCGTT +CGCGTC +CGCGTG +CGCGCA +CGCGCT +CGCGCC +CGCGCG +CGCGGA +CGCGGT +CGCGGC +CGCGGG +CGGAAA +CGGAAT +CGGAAC +CGGAAG +CGGATA +CGGATT +CGGATC +CGGATG +CGGACA +CGGACT +CGGACC +CGGACG +CGGAGA +CGGAGT +CGGAGC +CGGAGG +CGGTAA +CGGTAT +CGGTAC +CGGTAG +CGGTTA +CGGTTT +CGGTTC +CGGTTG +CGGTCA +CGGTCT +CGGTCC +CGGTCG +CGGTGA +CGGTGT +CGGTGC +CGGTGG +CGGCAA +CGGCAT +CGGCAC +CGGCAG +CGGCTA +CGGCTT +CGGCTC +CGGCTG +CGGCCA +CGGCCT +CGGCCC +CGGCCG +CGGCGA +CGGCGT +CGGCGC +CGGCGG +CGGGAA +CGGGAT +CGGGAC +CGGGAG +CGGGTA +CGGGTT +CGGGTC +CGGGTG +CGGGCA +CGGGCT +CGGGCC +CGGGCG +CGGGGA +CGGGGT +CGGGGC +CGGGGG +GAAAAA +GAAAAT +GAAAAC +GAAAAG +GAAATA +GAAATT +GAAATC +GAAATG +GAAACA +GAAACT +GAAACC +GAAACG +GAAAGA +GAAAGT +GAAAGC +GAAAGG +GAATAA +GAATAT +GAATAC +GAATAG +GAATTA +GAATTT +GAATTC +GAATTG +GAATCA +GAATCT +GAATCC +GAATCG +GAATGA +GAATGT +GAATGC +GAATGG +GAACAA +GAACAT +GAACAC +GAACAG +GAACTA +GAACTT +GAACTC +GAACTG +GAACCA +GAACCT +GAACCC +GAACCG +GAACGA +GAACGT +GAACGC +GAACGG +GAAGAA +GAAGAT +GAAGAC +GAAGAG +GAAGTA +GAAGTT +GAAGTC +GAAGTG +GAAGCA +GAAGCT +GAAGCC +GAAGCG +GAAGGA +GAAGGT +GAAGGC +GAAGGG +GATAAA +GATAAT +GATAAC +GATAAG +GATATA +GATATT +GATATC +GATATG +GATACA +GATACT +GATACC +GATACG +GATAGA +GATAGT +GATAGC +GATAGG +GATTAA +GATTAT +GATTAC +GATTAG +GATTTA +GATTTT +GATTTC +GATTTG +GATTCA +GATTCT +GATTCC +GATTCG +GATTGA +GATTGT +GATTGC +GATTGG +GATCAA +GATCAT +GATCAC +GATCAG +GATCTA +GATCTT +GATCTC +GATCTG +GATCCA +GATCCT +GATCCC +GATCCG +GATCGA +GATCGT +GATCGC +GATCGG +GATGAA +GATGAT +GATGAC +GATGAG +GATGTA +GATGTT +GATGTC +GATGTG +GATGCA +GATGCT +GATGCC +GATGCG +GATGGA +GATGGT +GATGGC +GATGGG +GACAAA +GACAAT +GACAAC +GACAAG +GACATA +GACATT +GACATC +GACATG +GACACA +GACACT +GACACC +GACACG +GACAGA +GACAGT +GACAGC +GACAGG +GACTAA +GACTAT +GACTAC +GACTAG +GACTTA +GACTTT +GACTTC +GACTTG +GACTCA +GACTCT +GACTCC +GACTCG +GACTGA +GACTGT +GACTGC +GACTGG +GACCAA +GACCAT +GACCAC +GACCAG +GACCTA +GACCTT +GACCTC +GACCTG +GACCCA +GACCCT +GACCCC +GACCCG +GACCGA +GACCGT +GACCGC +GACCGG +GACGAA +GACGAT +GACGAC +GACGAG +GACGTA +GACGTT +GACGTC +GACGTG +GACGCA +GACGCT +GACGCC +GACGCG +GACGGA +GACGGT +GACGGC +GACGGG +GAGAAA +GAGAAT +GAGAAC +GAGAAG +GAGATA +GAGATT +GAGATC +GAGATG +GAGACA +GAGACT +GAGACC +GAGACG +GAGAGA +GAGAGT +GAGAGC +GAGAGG +GAGTAA +GAGTAT +GAGTAC +GAGTAG +GAGTTA +GAGTTT +GAGTTC +GAGTTG +GAGTCA +GAGTCT +GAGTCC +GAGTCG +GAGTGA +GAGTGT +GAGTGC +GAGTGG +GAGCAA +GAGCAT +GAGCAC +GAGCAG +GAGCTA +GAGCTT +GAGCTC +GAGCTG +GAGCCA +GAGCCT +GAGCCC +GAGCCG +GAGCGA +GAGCGT +GAGCGC +GAGCGG +GAGGAA +GAGGAT +GAGGAC +GAGGAG +GAGGTA +GAGGTT +GAGGTC +GAGGTG +GAGGCA +GAGGCT +GAGGCC +GAGGCG +GAGGGA +GAGGGT +GAGGGC +GAGGGG +GTAAAA +GTAAAT +GTAAAC +GTAAAG +GTAATA +GTAATT +GTAATC +GTAATG +GTAACA +GTAACT +GTAACC +GTAACG +GTAAGA +GTAAGT +GTAAGC +GTAAGG +GTATAA +GTATAT +GTATAC +GTATAG +GTATTA +GTATTT +GTATTC +GTATTG +GTATCA +GTATCT +GTATCC +GTATCG +GTATGA +GTATGT +GTATGC +GTATGG +GTACAA +GTACAT +GTACAC +GTACAG +GTACTA +GTACTT +GTACTC +GTACTG +GTACCA +GTACCT +GTACCC +GTACCG +GTACGA +GTACGT +GTACGC +GTACGG +GTAGAA +GTAGAT +GTAGAC +GTAGAG +GTAGTA +GTAGTT +GTAGTC +GTAGTG +GTAGCA +GTAGCT +GTAGCC +GTAGCG +GTAGGA +GTAGGT +GTAGGC +GTAGGG +GTTAAA +GTTAAT +GTTAAC +GTTAAG +GTTATA +GTTATT +GTTATC +GTTATG +GTTACA +GTTACT +GTTACC +GTTACG +GTTAGA +GTTAGT +GTTAGC +GTTAGG +GTTTAA +GTTTAT +GTTTAC +GTTTAG +GTTTTA +GTTTTT +GTTTTC +GTTTTG +GTTTCA +GTTTCT +GTTTCC +GTTTCG +GTTTGA +GTTTGT +GTTTGC +GTTTGG +GTTCAA +GTTCAT +GTTCAC +GTTCAG +GTTCTA +GTTCTT +GTTCTC +GTTCTG +GTTCCA +GTTCCT +GTTCCC +GTTCCG +GTTCGA +GTTCGT +GTTCGC +GTTCGG +GTTGAA +GTTGAT +GTTGAC +GTTGAG +GTTGTA +GTTGTT +GTTGTC +GTTGTG +GTTGCA +GTTGCT +GTTGCC +GTTGCG +GTTGGA +GTTGGT +GTTGGC +GTTGGG +GTCAAA +GTCAAT +GTCAAC +GTCAAG +GTCATA +GTCATT +GTCATC +GTCATG +GTCACA +GTCACT +GTCACC +GTCACG +GTCAGA +GTCAGT +GTCAGC +GTCAGG +GTCTAA +GTCTAT +GTCTAC +GTCTAG +GTCTTA +GTCTTT +GTCTTC +GTCTTG +GTCTCA +GTCTCT +GTCTCC +GTCTCG +GTCTGA +GTCTGT +GTCTGC +GTCTGG +GTCCAA +GTCCAT +GTCCAC +GTCCAG +GTCCTA +GTCCTT +GTCCTC +GTCCTG +GTCCCA +GTCCCT +GTCCCC +GTCCCG +GTCCGA +GTCCGT +GTCCGC +GTCCGG +GTCGAA +GTCGAT +GTCGAC +GTCGAG +GTCGTA +GTCGTT +GTCGTC +GTCGTG +GTCGCA +GTCGCT +GTCGCC +GTCGCG +GTCGGA +GTCGGT +GTCGGC +GTCGGG +GTGAAA +GTGAAT +GTGAAC +GTGAAG +GTGATA +GTGATT +GTGATC +GTGATG +GTGACA +GTGACT +GTGACC +GTGACG +GTGAGA +GTGAGT +GTGAGC +GTGAGG +GTGTAA +GTGTAT +GTGTAC +GTGTAG +GTGTTA +GTGTTT +GTGTTC +GTGTTG +GTGTCA +GTGTCT +GTGTCC +GTGTCG +GTGTGA +GTGTGT +GTGTGC +GTGTGG +GTGCAA +GTGCAT +GTGCAC +GTGCAG +GTGCTA +GTGCTT +GTGCTC +GTGCTG +GTGCCA +GTGCCT +GTGCCC +GTGCCG +GTGCGA +GTGCGT +GTGCGC +GTGCGG +GTGGAA +GTGGAT +GTGGAC +GTGGAG +GTGGTA +GTGGTT +GTGGTC +GTGGTG +GTGGCA +GTGGCT +GTGGCC +GTGGCG +GTGGGA +GTGGGT +GTGGGC +GTGGGG +GCAAAA +GCAAAT +GCAAAC +GCAAAG +GCAATA +GCAATT +GCAATC +GCAATG +GCAACA +GCAACT +GCAACC +GCAACG +GCAAGA +GCAAGT +GCAAGC +GCAAGG +GCATAA +GCATAT +GCATAC +GCATAG +GCATTA +GCATTT +GCATTC +GCATTG +GCATCA +GCATCT +GCATCC +GCATCG +GCATGA +GCATGT +GCATGC +GCATGG +GCACAA +GCACAT +GCACAC +GCACAG +GCACTA +GCACTT +GCACTC +GCACTG +GCACCA +GCACCT +GCACCC +GCACCG +GCACGA +GCACGT +GCACGC +GCACGG +GCAGAA +GCAGAT +GCAGAC +GCAGAG +GCAGTA +GCAGTT +GCAGTC +GCAGTG +GCAGCA +GCAGCT +GCAGCC +GCAGCG +GCAGGA +GCAGGT +GCAGGC +GCAGGG +GCTAAA +GCTAAT +GCTAAC +GCTAAG +GCTATA +GCTATT +GCTATC +GCTATG +GCTACA +GCTACT +GCTACC +GCTACG +GCTAGA +GCTAGT +GCTAGC +GCTAGG +GCTTAA +GCTTAT +GCTTAC +GCTTAG +GCTTTA +GCTTTT +GCTTTC +GCTTTG +GCTTCA +GCTTCT +GCTTCC +GCTTCG +GCTTGA +GCTTGT +GCTTGC +GCTTGG +GCTCAA +GCTCAT +GCTCAC +GCTCAG +GCTCTA +GCTCTT +GCTCTC +GCTCTG +GCTCCA +GCTCCT +GCTCCC +GCTCCG +GCTCGA +GCTCGT +GCTCGC +GCTCGG +GCTGAA +GCTGAT +GCTGAC +GCTGAG +GCTGTA +GCTGTT +GCTGTC +GCTGTG +GCTGCA +GCTGCT +GCTGCC +GCTGCG +GCTGGA +GCTGGT +GCTGGC +GCTGGG +GCCAAA +GCCAAT +GCCAAC +GCCAAG +GCCATA +GCCATT +GCCATC +GCCATG +GCCACA +GCCACT +GCCACC +GCCACG +GCCAGA +GCCAGT +GCCAGC +GCCAGG +GCCTAA +GCCTAT +GCCTAC +GCCTAG +GCCTTA +GCCTTT +GCCTTC +GCCTTG +GCCTCA +GCCTCT +GCCTCC +GCCTCG +GCCTGA +GCCTGT +GCCTGC +GCCTGG +GCCCAA +GCCCAT +GCCCAC +GCCCAG +GCCCTA +GCCCTT +GCCCTC +GCCCTG +GCCCCA +GCCCCT +GCCCCC +GCCCCG +GCCCGA +GCCCGT +GCCCGC +GCCCGG +GCCGAA +GCCGAT +GCCGAC +GCCGAG +GCCGTA +GCCGTT +GCCGTC +GCCGTG +GCCGCA +GCCGCT +GCCGCC +GCCGCG +GCCGGA +GCCGGT +GCCGGC +GCCGGG +GCGAAA +GCGAAT +GCGAAC +GCGAAG +GCGATA +GCGATT +GCGATC +GCGATG +GCGACA +GCGACT +GCGACC +GCGACG +GCGAGA +GCGAGT +GCGAGC +GCGAGG +GCGTAA +GCGTAT +GCGTAC +GCGTAG +GCGTTA +GCGTTT +GCGTTC +GCGTTG +GCGTCA +GCGTCT +GCGTCC +GCGTCG +GCGTGA +GCGTGT +GCGTGC +GCGTGG +GCGCAA +GCGCAT +GCGCAC +GCGCAG +GCGCTA +GCGCTT +GCGCTC +GCGCTG +GCGCCA +GCGCCT +GCGCCC +GCGCCG +GCGCGA +GCGCGT +GCGCGC +GCGCGG +GCGGAA +GCGGAT +GCGGAC +GCGGAG +GCGGTA +GCGGTT +GCGGTC +GCGGTG +GCGGCA +GCGGCT +GCGGCC +GCGGCG +GCGGGA +GCGGGT +GCGGGC +GCGGGG +GGAAAA +GGAAAT +GGAAAC +GGAAAG +GGAATA +GGAATT +GGAATC +GGAATG +GGAACA +GGAACT +GGAACC +GGAACG +GGAAGA +GGAAGT +GGAAGC +GGAAGG +GGATAA +GGATAT +GGATAC +GGATAG +GGATTA +GGATTT +GGATTC +GGATTG +GGATCA +GGATCT +GGATCC +GGATCG +GGATGA +GGATGT +GGATGC +GGATGG +GGACAA +GGACAT +GGACAC +GGACAG +GGACTA +GGACTT +GGACTC +GGACTG +GGACCA +GGACCT +GGACCC +GGACCG +GGACGA +GGACGT +GGACGC +GGACGG +GGAGAA +GGAGAT +GGAGAC +GGAGAG +GGAGTA +GGAGTT +GGAGTC +GGAGTG +GGAGCA +GGAGCT +GGAGCC +GGAGCG +GGAGGA +GGAGGT +GGAGGC +GGAGGG +GGTAAA +GGTAAT +GGTAAC +GGTAAG +GGTATA +GGTATT +GGTATC +GGTATG +GGTACA +GGTACT +GGTACC +GGTACG +GGTAGA +GGTAGT +GGTAGC +GGTAGG +GGTTAA +GGTTAT +GGTTAC +GGTTAG +GGTTTA +GGTTTT +GGTTTC +GGTTTG +GGTTCA +GGTTCT +GGTTCC +GGTTCG +GGTTGA +GGTTGT +GGTTGC +GGTTGG +GGTCAA +GGTCAT +GGTCAC +GGTCAG +GGTCTA +GGTCTT +GGTCTC +GGTCTG +GGTCCA +GGTCCT +GGTCCC +GGTCCG +GGTCGA +GGTCGT +GGTCGC +GGTCGG +GGTGAA +GGTGAT +GGTGAC +GGTGAG +GGTGTA +GGTGTT +GGTGTC +GGTGTG +GGTGCA +GGTGCT +GGTGCC +GGTGCG +GGTGGA +GGTGGT +GGTGGC +GGTGGG +GGCAAA +GGCAAT +GGCAAC +GGCAAG +GGCATA +GGCATT +GGCATC +GGCATG +GGCACA +GGCACT +GGCACC +GGCACG +GGCAGA +GGCAGT +GGCAGC +GGCAGG +GGCTAA +GGCTAT +GGCTAC +GGCTAG +GGCTTA +GGCTTT +GGCTTC +GGCTTG +GGCTCA +GGCTCT +GGCTCC +GGCTCG +GGCTGA +GGCTGT +GGCTGC +GGCTGG +GGCCAA +GGCCAT +GGCCAC +GGCCAG +GGCCTA +GGCCTT +GGCCTC +GGCCTG +GGCCCA +GGCCCT +GGCCCC +GGCCCG +GGCCGA +GGCCGT +GGCCGC +GGCCGG +GGCGAA +GGCGAT +GGCGAC +GGCGAG +GGCGTA +GGCGTT +GGCGTC +GGCGTG +GGCGCA +GGCGCT +GGCGCC +GGCGCG +GGCGGA +GGCGGT +GGCGGC +GGCGGG +GGGAAA +GGGAAT +GGGAAC +GGGAAG +GGGATA +GGGATT +GGGATC +GGGATG +GGGACA +GGGACT +GGGACC +GGGACG +GGGAGA +GGGAGT +GGGAGC +GGGAGG +GGGTAA +GGGTAT +GGGTAC +GGGTAG +GGGTTA +GGGTTT +GGGTTC +GGGTTG +GGGTCA +GGGTCT +GGGTCC +GGGTCG +GGGTGA +GGGTGT +GGGTGC +GGGTGG +GGGCAA +GGGCAT +GGGCAC +GGGCAG +GGGCTA +GGGCTT +GGGCTC +GGGCTG +GGGCCA +GGGCCT +GGGCCC +GGGCCG +GGGCGA +GGGCGT +GGGCGC +GGGCGG +GGGGAA +GGGGAT +GGGGAC +GGGGAG +GGGGTA +GGGGTT +GGGGTC +GGGGTG +GGGGCA +GGGGCT +GGGGCC +GGGGCG +GGGGGA +GGGGGT +GGGGGC +GGGGGG diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6250/config.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6250/config.json new file mode 100644 index 0000000000000000000000000000000000000000..9b14e61d327bf586fcff6e817cf9d365bae18ad2 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6250/config.json @@ -0,0 +1,26 @@ +{ + "_name_or_path": "fabihamakhdoomi/TinyDNABERT", + "architectures": [ + "BertForSequenceClassification" + ], + "attention_probs_dropout_prob": 0.1, + "classifier_dropout": null, + "hidden_act": "gelu", + "hidden_dropout_prob": 0.1, + "hidden_size": 128, + "initializer_range": 0.02, + "intermediate_size": 512, + "layer_norm_eps": 1e-12, + "max_position_embeddings": 512, + "model_type": "bert", + "num_attention_heads": 2, + "num_hidden_layers": 2, + "pad_token_id": 0, + "position_embedding_type": "absolute", + "problem_type": "single_label_classification", + "torch_dtype": "float32", + "transformers_version": "4.44.0", + "type_vocab_size": 2, + "use_cache": true, + "vocab_size": 4101 +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6250/model.safetensors b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6250/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..adfc6dc126c0f0cf1e9fb672c81314f0966da850 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6250/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:89734419d84a104c06e9f3dcf425097e8ee231e3dd545e761eb4891ea4978cf9 +size 4021688 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6250/optimizer.pt b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6250/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..caa7d174fb4261af8867bcdd7b145d7fd49c91b7 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6250/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:84b0eda33f737701af9d8d3255ee7a1dcb6d2960f4ed869ace62f4957776347d +size 8058700 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6250/rng_state.pth b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6250/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..e666d96a1e7d953ad715b4f92cb4f8713cf02642 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6250/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:539fe12b01f7e37cbd1c254718c50ce0226ac2f3489b5b349456040c507b4a44 +size 14244 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6250/scheduler.pt b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6250/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..4c34d9dcafc0eb1507a84da1af7c2c5228be082e --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6250/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:20611d193eaf6658e6192005c679e5aaa561ac33254f734047a2032edc3e8f9b +size 1064 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6250/special_tokens_map.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6250/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..a8b3208c2884c4efb86e49300fdd3dc877220cdf --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6250/special_tokens_map.json @@ -0,0 +1,7 @@ +{ + "cls_token": "[CLS]", + "mask_token": "[MASK]", + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6250/tokenizer.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6250/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..ebd782de6963b9032ce9374cfe30f5275be1a30b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6250/tokenizer.json @@ -0,0 +1,4265 @@ +{ + "version": "1.0", + "truncation": { + "direction": "Right", + "max_length": 512, + "strategy": "LongestFirst", + "stride": 0 + }, + "padding": { + "strategy": { + "Fixed": 512 + }, + "direction": "Right", + "pad_to_multiple_of": null, + "pad_id": 0, + "pad_type_id": 0, + "pad_token": "[PAD]" + }, + "added_tokens": [ + { + "id": 0, + "content": "[PAD]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 1, + "content": "[UNK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 2, + "content": "[CLS]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 3, + "content": "[SEP]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 4, + "content": "[MASK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + } + ], + "normalizer": { + "type": "BertNormalizer", + "clean_text": true, + "handle_chinese_chars": true, + "strip_accents": null, + "lowercase": false + }, + "pre_tokenizer": { + "type": "BertPreTokenizer" + }, + "post_processor": { + "type": "TemplateProcessing", + "single": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + } + ], + "pair": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "B", + "type_id": 1 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 1 + } + } + ], + "special_tokens": { + "[CLS]": { + "id": "[CLS]", + "ids": [ + 2 + ], + "tokens": [ + "[CLS]" + ] + }, + "[SEP]": { + "id": "[SEP]", + "ids": [ + 3 + ], + "tokens": [ + "[SEP]" + ] + } + } + }, + "decoder": { + "type": "WordPiece", + "prefix": "##", + "cleanup": true + }, + "model": { + "type": "WordPiece", + "unk_token": "[UNK]", + "continuing_subword_prefix": "##", + "max_input_chars_per_word": 100, + "vocab": { + "[PAD]": 0, + "[UNK]": 1, + "[CLS]": 2, + "[SEP]": 3, + "[MASK]": 4, + "AAAAAA": 5, + "AAAAAT": 6, + "AAAAAC": 7, + "AAAAAG": 8, + "AAAATA": 9, + "AAAATT": 10, + "AAAATC": 11, + "AAAATG": 12, + "AAAACA": 13, + "AAAACT": 14, + "AAAACC": 15, + "AAAACG": 16, + "AAAAGA": 17, + "AAAAGT": 18, + "AAAAGC": 19, + "AAAAGG": 20, + "AAATAA": 21, + "AAATAT": 22, + "AAATAC": 23, + "AAATAG": 24, + "AAATTA": 25, + "AAATTT": 26, + "AAATTC": 27, + "AAATTG": 28, + "AAATCA": 29, + "AAATCT": 30, + "AAATCC": 31, + "AAATCG": 32, + "AAATGA": 33, + "AAATGT": 34, + "AAATGC": 35, + "AAATGG": 36, + "AAACAA": 37, + "AAACAT": 38, + "AAACAC": 39, + "AAACAG": 40, + "AAACTA": 41, + "AAACTT": 42, + "AAACTC": 43, + "AAACTG": 44, + "AAACCA": 45, + "AAACCT": 46, + "AAACCC": 47, + "AAACCG": 48, + "AAACGA": 49, + "AAACGT": 50, + "AAACGC": 51, + "AAACGG": 52, + "AAAGAA": 53, + "AAAGAT": 54, + "AAAGAC": 55, + "AAAGAG": 56, + "AAAGTA": 57, + "AAAGTT": 58, + "AAAGTC": 59, + "AAAGTG": 60, + "AAAGCA": 61, + "AAAGCT": 62, + "AAAGCC": 63, + "AAAGCG": 64, + "AAAGGA": 65, + "AAAGGT": 66, + "AAAGGC": 67, + "AAAGGG": 68, + "AATAAA": 69, + "AATAAT": 70, + "AATAAC": 71, + "AATAAG": 72, + "AATATA": 73, + "AATATT": 74, + "AATATC": 75, + "AATATG": 76, + "AATACA": 77, + "AATACT": 78, + "AATACC": 79, + "AATACG": 80, + "AATAGA": 81, + "AATAGT": 82, + "AATAGC": 83, + "AATAGG": 84, + "AATTAA": 85, + "AATTAT": 86, + "AATTAC": 87, + "AATTAG": 88, + "AATTTA": 89, + "AATTTT": 90, + "AATTTC": 91, + "AATTTG": 92, + "AATTCA": 93, + "AATTCT": 94, + "AATTCC": 95, + "AATTCG": 96, + "AATTGA": 97, + "AATTGT": 98, + "AATTGC": 99, + "AATTGG": 100, + "AATCAA": 101, + "AATCAT": 102, + "AATCAC": 103, + "AATCAG": 104, + "AATCTA": 105, + "AATCTT": 106, + "AATCTC": 107, + "AATCTG": 108, + "AATCCA": 109, + "AATCCT": 110, + "AATCCC": 111, + "AATCCG": 112, + "AATCGA": 113, + "AATCGT": 114, + "AATCGC": 115, + "AATCGG": 116, + "AATGAA": 117, + "AATGAT": 118, + "AATGAC": 119, + "AATGAG": 120, + "AATGTA": 121, + "AATGTT": 122, + "AATGTC": 123, + "AATGTG": 124, + "AATGCA": 125, + "AATGCT": 126, + "AATGCC": 127, + "AATGCG": 128, + "AATGGA": 129, + "AATGGT": 130, + "AATGGC": 131, + "AATGGG": 132, + "AACAAA": 133, + "AACAAT": 134, + "AACAAC": 135, + "AACAAG": 136, + "AACATA": 137, + "AACATT": 138, + "AACATC": 139, + "AACATG": 140, + "AACACA": 141, + "AACACT": 142, + "AACACC": 143, + "AACACG": 144, + "AACAGA": 145, + "AACAGT": 146, + "AACAGC": 147, + "AACAGG": 148, + "AACTAA": 149, + "AACTAT": 150, + "AACTAC": 151, + "AACTAG": 152, + "AACTTA": 153, + "AACTTT": 154, + "AACTTC": 155, + "AACTTG": 156, + "AACTCA": 157, + "AACTCT": 158, + "AACTCC": 159, + "AACTCG": 160, + "AACTGA": 161, + "AACTGT": 162, + "AACTGC": 163, + "AACTGG": 164, + "AACCAA": 165, + "AACCAT": 166, + "AACCAC": 167, + "AACCAG": 168, + "AACCTA": 169, + "AACCTT": 170, + "AACCTC": 171, + "AACCTG": 172, + "AACCCA": 173, + "AACCCT": 174, + "AACCCC": 175, + "AACCCG": 176, + "AACCGA": 177, + "AACCGT": 178, + "AACCGC": 179, + "AACCGG": 180, + "AACGAA": 181, + "AACGAT": 182, + "AACGAC": 183, + "AACGAG": 184, + "AACGTA": 185, + "AACGTT": 186, + "AACGTC": 187, + "AACGTG": 188, + "AACGCA": 189, + "AACGCT": 190, + "AACGCC": 191, + "AACGCG": 192, + "AACGGA": 193, + "AACGGT": 194, + "AACGGC": 195, + "AACGGG": 196, + "AAGAAA": 197, + "AAGAAT": 198, + "AAGAAC": 199, + "AAGAAG": 200, + "AAGATA": 201, + "AAGATT": 202, + "AAGATC": 203, + "AAGATG": 204, + "AAGACA": 205, + "AAGACT": 206, + "AAGACC": 207, + "AAGACG": 208, + "AAGAGA": 209, + "AAGAGT": 210, + "AAGAGC": 211, + "AAGAGG": 212, + "AAGTAA": 213, + "AAGTAT": 214, + "AAGTAC": 215, + "AAGTAG": 216, + "AAGTTA": 217, + "AAGTTT": 218, + "AAGTTC": 219, + "AAGTTG": 220, + "AAGTCA": 221, + "AAGTCT": 222, + "AAGTCC": 223, + "AAGTCG": 224, + "AAGTGA": 225, + "AAGTGT": 226, + "AAGTGC": 227, + "AAGTGG": 228, + "AAGCAA": 229, + "AAGCAT": 230, + "AAGCAC": 231, + "AAGCAG": 232, + "AAGCTA": 233, + "AAGCTT": 234, + "AAGCTC": 235, + "AAGCTG": 236, + "AAGCCA": 237, + "AAGCCT": 238, + "AAGCCC": 239, + "AAGCCG": 240, + "AAGCGA": 241, + "AAGCGT": 242, + "AAGCGC": 243, + "AAGCGG": 244, + "AAGGAA": 245, + "AAGGAT": 246, + "AAGGAC": 247, + "AAGGAG": 248, + "AAGGTA": 249, + "AAGGTT": 250, + "AAGGTC": 251, + "AAGGTG": 252, + "AAGGCA": 253, + "AAGGCT": 254, + "AAGGCC": 255, + "AAGGCG": 256, + "AAGGGA": 257, + "AAGGGT": 258, + "AAGGGC": 259, + "AAGGGG": 260, + "ATAAAA": 261, + "ATAAAT": 262, + "ATAAAC": 263, + "ATAAAG": 264, + "ATAATA": 265, + "ATAATT": 266, + "ATAATC": 267, + "ATAATG": 268, + "ATAACA": 269, + "ATAACT": 270, + "ATAACC": 271, + "ATAACG": 272, + "ATAAGA": 273, + "ATAAGT": 274, + "ATAAGC": 275, + "ATAAGG": 276, + "ATATAA": 277, + "ATATAT": 278, + "ATATAC": 279, + "ATATAG": 280, + "ATATTA": 281, + "ATATTT": 282, + "ATATTC": 283, + "ATATTG": 284, + "ATATCA": 285, + "ATATCT": 286, + "ATATCC": 287, + "ATATCG": 288, + "ATATGA": 289, + "ATATGT": 290, + "ATATGC": 291, + "ATATGG": 292, + "ATACAA": 293, + "ATACAT": 294, + "ATACAC": 295, + "ATACAG": 296, + "ATACTA": 297, + "ATACTT": 298, + "ATACTC": 299, + "ATACTG": 300, + "ATACCA": 301, + "ATACCT": 302, + "ATACCC": 303, + "ATACCG": 304, + "ATACGA": 305, + "ATACGT": 306, + "ATACGC": 307, + "ATACGG": 308, + "ATAGAA": 309, + "ATAGAT": 310, + "ATAGAC": 311, + "ATAGAG": 312, + "ATAGTA": 313, + "ATAGTT": 314, + "ATAGTC": 315, + "ATAGTG": 316, + "ATAGCA": 317, + "ATAGCT": 318, + "ATAGCC": 319, + "ATAGCG": 320, + "ATAGGA": 321, + "ATAGGT": 322, + "ATAGGC": 323, + "ATAGGG": 324, + "ATTAAA": 325, + "ATTAAT": 326, + "ATTAAC": 327, + "ATTAAG": 328, + "ATTATA": 329, + "ATTATT": 330, + "ATTATC": 331, + "ATTATG": 332, + "ATTACA": 333, + "ATTACT": 334, + "ATTACC": 335, + "ATTACG": 336, + "ATTAGA": 337, + "ATTAGT": 338, + "ATTAGC": 339, + "ATTAGG": 340, + "ATTTAA": 341, + "ATTTAT": 342, + "ATTTAC": 343, + "ATTTAG": 344, + "ATTTTA": 345, + "ATTTTT": 346, + "ATTTTC": 347, + "ATTTTG": 348, + "ATTTCA": 349, + "ATTTCT": 350, + "ATTTCC": 351, + "ATTTCG": 352, + "ATTTGA": 353, + "ATTTGT": 354, + "ATTTGC": 355, + "ATTTGG": 356, + "ATTCAA": 357, + "ATTCAT": 358, + "ATTCAC": 359, + "ATTCAG": 360, + "ATTCTA": 361, + "ATTCTT": 362, + "ATTCTC": 363, + "ATTCTG": 364, + "ATTCCA": 365, + "ATTCCT": 366, + "ATTCCC": 367, + "ATTCCG": 368, + "ATTCGA": 369, + "ATTCGT": 370, + "ATTCGC": 371, + "ATTCGG": 372, + "ATTGAA": 373, + "ATTGAT": 374, + "ATTGAC": 375, + "ATTGAG": 376, + "ATTGTA": 377, + "ATTGTT": 378, + "ATTGTC": 379, + "ATTGTG": 380, + "ATTGCA": 381, + "ATTGCT": 382, + "ATTGCC": 383, + "ATTGCG": 384, + "ATTGGA": 385, + "ATTGGT": 386, + "ATTGGC": 387, + "ATTGGG": 388, + "ATCAAA": 389, + "ATCAAT": 390, + "ATCAAC": 391, + "ATCAAG": 392, + "ATCATA": 393, + "ATCATT": 394, + "ATCATC": 395, + "ATCATG": 396, + "ATCACA": 397, + "ATCACT": 398, + "ATCACC": 399, + "ATCACG": 400, + "ATCAGA": 401, + "ATCAGT": 402, + "ATCAGC": 403, + "ATCAGG": 404, + "ATCTAA": 405, + "ATCTAT": 406, + "ATCTAC": 407, + "ATCTAG": 408, + "ATCTTA": 409, + "ATCTTT": 410, + "ATCTTC": 411, + "ATCTTG": 412, + "ATCTCA": 413, + "ATCTCT": 414, + "ATCTCC": 415, + "ATCTCG": 416, + "ATCTGA": 417, + "ATCTGT": 418, + "ATCTGC": 419, + "ATCTGG": 420, + "ATCCAA": 421, + "ATCCAT": 422, + "ATCCAC": 423, + "ATCCAG": 424, + "ATCCTA": 425, + "ATCCTT": 426, + "ATCCTC": 427, + "ATCCTG": 428, + "ATCCCA": 429, + "ATCCCT": 430, + "ATCCCC": 431, + "ATCCCG": 432, + "ATCCGA": 433, + "ATCCGT": 434, + "ATCCGC": 435, + "ATCCGG": 436, + "ATCGAA": 437, + "ATCGAT": 438, + "ATCGAC": 439, + "ATCGAG": 440, + "ATCGTA": 441, + "ATCGTT": 442, + "ATCGTC": 443, + "ATCGTG": 444, + "ATCGCA": 445, + "ATCGCT": 446, + "ATCGCC": 447, + "ATCGCG": 448, + "ATCGGA": 449, + "ATCGGT": 450, + "ATCGGC": 451, + "ATCGGG": 452, + "ATGAAA": 453, + "ATGAAT": 454, + "ATGAAC": 455, + "ATGAAG": 456, + "ATGATA": 457, + "ATGATT": 458, + "ATGATC": 459, + "ATGATG": 460, + "ATGACA": 461, + "ATGACT": 462, + "ATGACC": 463, + "ATGACG": 464, + "ATGAGA": 465, + "ATGAGT": 466, + "ATGAGC": 467, + "ATGAGG": 468, + "ATGTAA": 469, + "ATGTAT": 470, + "ATGTAC": 471, + "ATGTAG": 472, + "ATGTTA": 473, + "ATGTTT": 474, + "ATGTTC": 475, + "ATGTTG": 476, + "ATGTCA": 477, + "ATGTCT": 478, + "ATGTCC": 479, + "ATGTCG": 480, + "ATGTGA": 481, + "ATGTGT": 482, + "ATGTGC": 483, + "ATGTGG": 484, + "ATGCAA": 485, + "ATGCAT": 486, + "ATGCAC": 487, + "ATGCAG": 488, + "ATGCTA": 489, + "ATGCTT": 490, + "ATGCTC": 491, + "ATGCTG": 492, + "ATGCCA": 493, + "ATGCCT": 494, + "ATGCCC": 495, + "ATGCCG": 496, + "ATGCGA": 497, + "ATGCGT": 498, + "ATGCGC": 499, + "ATGCGG": 500, + "ATGGAA": 501, + "ATGGAT": 502, + "ATGGAC": 503, + "ATGGAG": 504, + "ATGGTA": 505, + "ATGGTT": 506, + "ATGGTC": 507, + "ATGGTG": 508, + "ATGGCA": 509, + "ATGGCT": 510, + "ATGGCC": 511, + "ATGGCG": 512, + "ATGGGA": 513, + "ATGGGT": 514, + "ATGGGC": 515, + "ATGGGG": 516, + "ACAAAA": 517, + "ACAAAT": 518, + "ACAAAC": 519, + "ACAAAG": 520, + "ACAATA": 521, + "ACAATT": 522, + "ACAATC": 523, + "ACAATG": 524, + "ACAACA": 525, + "ACAACT": 526, + "ACAACC": 527, + "ACAACG": 528, + "ACAAGA": 529, + "ACAAGT": 530, + "ACAAGC": 531, + "ACAAGG": 532, + "ACATAA": 533, + "ACATAT": 534, + "ACATAC": 535, + "ACATAG": 536, + "ACATTA": 537, + "ACATTT": 538, + "ACATTC": 539, + "ACATTG": 540, + "ACATCA": 541, + "ACATCT": 542, + "ACATCC": 543, + "ACATCG": 544, + "ACATGA": 545, + "ACATGT": 546, + "ACATGC": 547, + "ACATGG": 548, + "ACACAA": 549, + "ACACAT": 550, + "ACACAC": 551, + "ACACAG": 552, + "ACACTA": 553, + "ACACTT": 554, + "ACACTC": 555, + "ACACTG": 556, + "ACACCA": 557, + "ACACCT": 558, + "ACACCC": 559, + "ACACCG": 560, + "ACACGA": 561, + "ACACGT": 562, + "ACACGC": 563, + "ACACGG": 564, + "ACAGAA": 565, + "ACAGAT": 566, + "ACAGAC": 567, + "ACAGAG": 568, + "ACAGTA": 569, + "ACAGTT": 570, + "ACAGTC": 571, + "ACAGTG": 572, + "ACAGCA": 573, + "ACAGCT": 574, + "ACAGCC": 575, + "ACAGCG": 576, + "ACAGGA": 577, + "ACAGGT": 578, + "ACAGGC": 579, + "ACAGGG": 580, + "ACTAAA": 581, + "ACTAAT": 582, + "ACTAAC": 583, + "ACTAAG": 584, + "ACTATA": 585, + "ACTATT": 586, + "ACTATC": 587, + "ACTATG": 588, + "ACTACA": 589, + "ACTACT": 590, + "ACTACC": 591, + "ACTACG": 592, + "ACTAGA": 593, + "ACTAGT": 594, + "ACTAGC": 595, + "ACTAGG": 596, + "ACTTAA": 597, + "ACTTAT": 598, + "ACTTAC": 599, + "ACTTAG": 600, + "ACTTTA": 601, + "ACTTTT": 602, + "ACTTTC": 603, + "ACTTTG": 604, + "ACTTCA": 605, + "ACTTCT": 606, + "ACTTCC": 607, + "ACTTCG": 608, + "ACTTGA": 609, + "ACTTGT": 610, + "ACTTGC": 611, + "ACTTGG": 612, + "ACTCAA": 613, + "ACTCAT": 614, + "ACTCAC": 615, + "ACTCAG": 616, + "ACTCTA": 617, + "ACTCTT": 618, + "ACTCTC": 619, + "ACTCTG": 620, + "ACTCCA": 621, + "ACTCCT": 622, + "ACTCCC": 623, + "ACTCCG": 624, + "ACTCGA": 625, + "ACTCGT": 626, + "ACTCGC": 627, + "ACTCGG": 628, + "ACTGAA": 629, + "ACTGAT": 630, + "ACTGAC": 631, + "ACTGAG": 632, + "ACTGTA": 633, + "ACTGTT": 634, + "ACTGTC": 635, + "ACTGTG": 636, + "ACTGCA": 637, + "ACTGCT": 638, + "ACTGCC": 639, + "ACTGCG": 640, + "ACTGGA": 641, + "ACTGGT": 642, + "ACTGGC": 643, + "ACTGGG": 644, + "ACCAAA": 645, + "ACCAAT": 646, + "ACCAAC": 647, + "ACCAAG": 648, + "ACCATA": 649, + "ACCATT": 650, + "ACCATC": 651, + "ACCATG": 652, + "ACCACA": 653, + "ACCACT": 654, + "ACCACC": 655, + "ACCACG": 656, + "ACCAGA": 657, + "ACCAGT": 658, + "ACCAGC": 659, + "ACCAGG": 660, + "ACCTAA": 661, + "ACCTAT": 662, + "ACCTAC": 663, + "ACCTAG": 664, + "ACCTTA": 665, + "ACCTTT": 666, + "ACCTTC": 667, + "ACCTTG": 668, + "ACCTCA": 669, + "ACCTCT": 670, + "ACCTCC": 671, + "ACCTCG": 672, + "ACCTGA": 673, + "ACCTGT": 674, + "ACCTGC": 675, + "ACCTGG": 676, + "ACCCAA": 677, + "ACCCAT": 678, + "ACCCAC": 679, + "ACCCAG": 680, + "ACCCTA": 681, + "ACCCTT": 682, + "ACCCTC": 683, + "ACCCTG": 684, + "ACCCCA": 685, + "ACCCCT": 686, + "ACCCCC": 687, + "ACCCCG": 688, + "ACCCGA": 689, + "ACCCGT": 690, + "ACCCGC": 691, + "ACCCGG": 692, + "ACCGAA": 693, + "ACCGAT": 694, + "ACCGAC": 695, + "ACCGAG": 696, + "ACCGTA": 697, + "ACCGTT": 698, + "ACCGTC": 699, + "ACCGTG": 700, + "ACCGCA": 701, + "ACCGCT": 702, + "ACCGCC": 703, + "ACCGCG": 704, + "ACCGGA": 705, + "ACCGGT": 706, + "ACCGGC": 707, + "ACCGGG": 708, + "ACGAAA": 709, + "ACGAAT": 710, + "ACGAAC": 711, + "ACGAAG": 712, + "ACGATA": 713, + "ACGATT": 714, + "ACGATC": 715, + "ACGATG": 716, + "ACGACA": 717, + "ACGACT": 718, + "ACGACC": 719, + "ACGACG": 720, + "ACGAGA": 721, + "ACGAGT": 722, + "ACGAGC": 723, + "ACGAGG": 724, + "ACGTAA": 725, + "ACGTAT": 726, + "ACGTAC": 727, + "ACGTAG": 728, + "ACGTTA": 729, + "ACGTTT": 730, + "ACGTTC": 731, + "ACGTTG": 732, + "ACGTCA": 733, + "ACGTCT": 734, + "ACGTCC": 735, + "ACGTCG": 736, + "ACGTGA": 737, + "ACGTGT": 738, + "ACGTGC": 739, + "ACGTGG": 740, + "ACGCAA": 741, + "ACGCAT": 742, + "ACGCAC": 743, + "ACGCAG": 744, + "ACGCTA": 745, + "ACGCTT": 746, + "ACGCTC": 747, + "ACGCTG": 748, + "ACGCCA": 749, + "ACGCCT": 750, + "ACGCCC": 751, + "ACGCCG": 752, + "ACGCGA": 753, + "ACGCGT": 754, + "ACGCGC": 755, + "ACGCGG": 756, + "ACGGAA": 757, + "ACGGAT": 758, + "ACGGAC": 759, + "ACGGAG": 760, + "ACGGTA": 761, + "ACGGTT": 762, + "ACGGTC": 763, + "ACGGTG": 764, + "ACGGCA": 765, + "ACGGCT": 766, + "ACGGCC": 767, + "ACGGCG": 768, + "ACGGGA": 769, + "ACGGGT": 770, + "ACGGGC": 771, + "ACGGGG": 772, + "AGAAAA": 773, + "AGAAAT": 774, + "AGAAAC": 775, + "AGAAAG": 776, + "AGAATA": 777, + "AGAATT": 778, + "AGAATC": 779, + "AGAATG": 780, + "AGAACA": 781, + "AGAACT": 782, + "AGAACC": 783, + "AGAACG": 784, + "AGAAGA": 785, + "AGAAGT": 786, + "AGAAGC": 787, + "AGAAGG": 788, + "AGATAA": 789, + "AGATAT": 790, + "AGATAC": 791, + "AGATAG": 792, + "AGATTA": 793, + "AGATTT": 794, + "AGATTC": 795, + "AGATTG": 796, + "AGATCA": 797, + "AGATCT": 798, + "AGATCC": 799, + "AGATCG": 800, + "AGATGA": 801, + "AGATGT": 802, + "AGATGC": 803, + "AGATGG": 804, + "AGACAA": 805, + "AGACAT": 806, + "AGACAC": 807, + "AGACAG": 808, + "AGACTA": 809, + "AGACTT": 810, + "AGACTC": 811, + "AGACTG": 812, + "AGACCA": 813, + "AGACCT": 814, + "AGACCC": 815, + "AGACCG": 816, + "AGACGA": 817, + "AGACGT": 818, + "AGACGC": 819, + "AGACGG": 820, + "AGAGAA": 821, + "AGAGAT": 822, + "AGAGAC": 823, + "AGAGAG": 824, + "AGAGTA": 825, + "AGAGTT": 826, + "AGAGTC": 827, + "AGAGTG": 828, + "AGAGCA": 829, + "AGAGCT": 830, + "AGAGCC": 831, + "AGAGCG": 832, + "AGAGGA": 833, + "AGAGGT": 834, + "AGAGGC": 835, + "AGAGGG": 836, + "AGTAAA": 837, + "AGTAAT": 838, + "AGTAAC": 839, + "AGTAAG": 840, + "AGTATA": 841, + "AGTATT": 842, + "AGTATC": 843, + "AGTATG": 844, + "AGTACA": 845, + "AGTACT": 846, + "AGTACC": 847, + "AGTACG": 848, + "AGTAGA": 849, + "AGTAGT": 850, + "AGTAGC": 851, + "AGTAGG": 852, + "AGTTAA": 853, + "AGTTAT": 854, + "AGTTAC": 855, + "AGTTAG": 856, + "AGTTTA": 857, + "AGTTTT": 858, + "AGTTTC": 859, + "AGTTTG": 860, + "AGTTCA": 861, + "AGTTCT": 862, + "AGTTCC": 863, + "AGTTCG": 864, + "AGTTGA": 865, + "AGTTGT": 866, + "AGTTGC": 867, + "AGTTGG": 868, + "AGTCAA": 869, + "AGTCAT": 870, + "AGTCAC": 871, + "AGTCAG": 872, + "AGTCTA": 873, + "AGTCTT": 874, + "AGTCTC": 875, + "AGTCTG": 876, + "AGTCCA": 877, + "AGTCCT": 878, + "AGTCCC": 879, + "AGTCCG": 880, + "AGTCGA": 881, + "AGTCGT": 882, + "AGTCGC": 883, + "AGTCGG": 884, + "AGTGAA": 885, + "AGTGAT": 886, + "AGTGAC": 887, + "AGTGAG": 888, + "AGTGTA": 889, + "AGTGTT": 890, + "AGTGTC": 891, + "AGTGTG": 892, + "AGTGCA": 893, + "AGTGCT": 894, + "AGTGCC": 895, + "AGTGCG": 896, + "AGTGGA": 897, + "AGTGGT": 898, + "AGTGGC": 899, + "AGTGGG": 900, + "AGCAAA": 901, + "AGCAAT": 902, + "AGCAAC": 903, + "AGCAAG": 904, + "AGCATA": 905, + "AGCATT": 906, + "AGCATC": 907, + "AGCATG": 908, + "AGCACA": 909, + "AGCACT": 910, + "AGCACC": 911, + "AGCACG": 912, + "AGCAGA": 913, + "AGCAGT": 914, + "AGCAGC": 915, + "AGCAGG": 916, + "AGCTAA": 917, + "AGCTAT": 918, + "AGCTAC": 919, + "AGCTAG": 920, + "AGCTTA": 921, + "AGCTTT": 922, + "AGCTTC": 923, + "AGCTTG": 924, + "AGCTCA": 925, + "AGCTCT": 926, + "AGCTCC": 927, + "AGCTCG": 928, + "AGCTGA": 929, + "AGCTGT": 930, + "AGCTGC": 931, + "AGCTGG": 932, + "AGCCAA": 933, + "AGCCAT": 934, + "AGCCAC": 935, + "AGCCAG": 936, + "AGCCTA": 937, + "AGCCTT": 938, + "AGCCTC": 939, + "AGCCTG": 940, + "AGCCCA": 941, + "AGCCCT": 942, + "AGCCCC": 943, + "AGCCCG": 944, + "AGCCGA": 945, + "AGCCGT": 946, + "AGCCGC": 947, + "AGCCGG": 948, + "AGCGAA": 949, + "AGCGAT": 950, + "AGCGAC": 951, + "AGCGAG": 952, + "AGCGTA": 953, + "AGCGTT": 954, + "AGCGTC": 955, + "AGCGTG": 956, + "AGCGCA": 957, + "AGCGCT": 958, + "AGCGCC": 959, + "AGCGCG": 960, + "AGCGGA": 961, + "AGCGGT": 962, + "AGCGGC": 963, + "AGCGGG": 964, + "AGGAAA": 965, + "AGGAAT": 966, + "AGGAAC": 967, + "AGGAAG": 968, + "AGGATA": 969, + "AGGATT": 970, + "AGGATC": 971, + "AGGATG": 972, + "AGGACA": 973, + "AGGACT": 974, + "AGGACC": 975, + "AGGACG": 976, + "AGGAGA": 977, + "AGGAGT": 978, + "AGGAGC": 979, + "AGGAGG": 980, + "AGGTAA": 981, + "AGGTAT": 982, + "AGGTAC": 983, + "AGGTAG": 984, + "AGGTTA": 985, + "AGGTTT": 986, + "AGGTTC": 987, + "AGGTTG": 988, + "AGGTCA": 989, + "AGGTCT": 990, + "AGGTCC": 991, + "AGGTCG": 992, + "AGGTGA": 993, + "AGGTGT": 994, + "AGGTGC": 995, + "AGGTGG": 996, + "AGGCAA": 997, + "AGGCAT": 998, + "AGGCAC": 999, + "AGGCAG": 1000, + "AGGCTA": 1001, + "AGGCTT": 1002, + "AGGCTC": 1003, + "AGGCTG": 1004, + "AGGCCA": 1005, + "AGGCCT": 1006, + "AGGCCC": 1007, + "AGGCCG": 1008, + "AGGCGA": 1009, + "AGGCGT": 1010, + "AGGCGC": 1011, + "AGGCGG": 1012, + "AGGGAA": 1013, + "AGGGAT": 1014, + "AGGGAC": 1015, + "AGGGAG": 1016, + "AGGGTA": 1017, + "AGGGTT": 1018, + "AGGGTC": 1019, + "AGGGTG": 1020, + "AGGGCA": 1021, + "AGGGCT": 1022, + "AGGGCC": 1023, + "AGGGCG": 1024, + "AGGGGA": 1025, + "AGGGGT": 1026, + "AGGGGC": 1027, + "AGGGGG": 1028, + "TAAAAA": 1029, + "TAAAAT": 1030, + "TAAAAC": 1031, + "TAAAAG": 1032, + "TAAATA": 1033, + "TAAATT": 1034, + "TAAATC": 1035, + "TAAATG": 1036, + "TAAACA": 1037, + "TAAACT": 1038, + "TAAACC": 1039, + "TAAACG": 1040, + "TAAAGA": 1041, + "TAAAGT": 1042, + "TAAAGC": 1043, + "TAAAGG": 1044, + "TAATAA": 1045, + "TAATAT": 1046, + "TAATAC": 1047, + "TAATAG": 1048, + "TAATTA": 1049, + "TAATTT": 1050, + "TAATTC": 1051, + "TAATTG": 1052, + "TAATCA": 1053, + "TAATCT": 1054, + "TAATCC": 1055, + "TAATCG": 1056, + "TAATGA": 1057, + "TAATGT": 1058, + "TAATGC": 1059, + "TAATGG": 1060, + "TAACAA": 1061, + "TAACAT": 1062, + "TAACAC": 1063, + "TAACAG": 1064, + "TAACTA": 1065, + "TAACTT": 1066, + "TAACTC": 1067, + "TAACTG": 1068, + "TAACCA": 1069, + "TAACCT": 1070, + "TAACCC": 1071, + "TAACCG": 1072, + "TAACGA": 1073, + "TAACGT": 1074, + "TAACGC": 1075, + "TAACGG": 1076, + "TAAGAA": 1077, + "TAAGAT": 1078, + "TAAGAC": 1079, + "TAAGAG": 1080, + "TAAGTA": 1081, + "TAAGTT": 1082, + "TAAGTC": 1083, + "TAAGTG": 1084, + "TAAGCA": 1085, + "TAAGCT": 1086, + "TAAGCC": 1087, + "TAAGCG": 1088, + "TAAGGA": 1089, + "TAAGGT": 1090, + "TAAGGC": 1091, + "TAAGGG": 1092, + "TATAAA": 1093, + "TATAAT": 1094, + "TATAAC": 1095, + "TATAAG": 1096, + "TATATA": 1097, + "TATATT": 1098, + "TATATC": 1099, + "TATATG": 1100, + "TATACA": 1101, + "TATACT": 1102, + "TATACC": 1103, + "TATACG": 1104, + "TATAGA": 1105, + "TATAGT": 1106, + "TATAGC": 1107, + "TATAGG": 1108, + "TATTAA": 1109, + "TATTAT": 1110, + "TATTAC": 1111, + "TATTAG": 1112, + "TATTTA": 1113, + "TATTTT": 1114, + "TATTTC": 1115, + "TATTTG": 1116, + "TATTCA": 1117, + "TATTCT": 1118, + "TATTCC": 1119, + "TATTCG": 1120, + "TATTGA": 1121, + "TATTGT": 1122, + "TATTGC": 1123, + "TATTGG": 1124, + "TATCAA": 1125, + "TATCAT": 1126, + "TATCAC": 1127, + "TATCAG": 1128, + "TATCTA": 1129, + "TATCTT": 1130, + "TATCTC": 1131, + "TATCTG": 1132, + "TATCCA": 1133, + "TATCCT": 1134, + "TATCCC": 1135, + "TATCCG": 1136, + "TATCGA": 1137, + "TATCGT": 1138, + "TATCGC": 1139, + "TATCGG": 1140, + "TATGAA": 1141, + "TATGAT": 1142, + "TATGAC": 1143, + "TATGAG": 1144, + "TATGTA": 1145, + "TATGTT": 1146, + "TATGTC": 1147, + "TATGTG": 1148, + "TATGCA": 1149, + "TATGCT": 1150, + "TATGCC": 1151, + "TATGCG": 1152, + "TATGGA": 1153, + "TATGGT": 1154, + "TATGGC": 1155, + "TATGGG": 1156, + "TACAAA": 1157, + "TACAAT": 1158, + "TACAAC": 1159, + "TACAAG": 1160, + "TACATA": 1161, + "TACATT": 1162, + "TACATC": 1163, + "TACATG": 1164, + "TACACA": 1165, + "TACACT": 1166, + "TACACC": 1167, + "TACACG": 1168, + "TACAGA": 1169, + "TACAGT": 1170, + "TACAGC": 1171, + "TACAGG": 1172, + "TACTAA": 1173, + "TACTAT": 1174, + "TACTAC": 1175, + "TACTAG": 1176, + "TACTTA": 1177, + "TACTTT": 1178, + "TACTTC": 1179, + "TACTTG": 1180, + "TACTCA": 1181, + "TACTCT": 1182, + "TACTCC": 1183, + "TACTCG": 1184, + "TACTGA": 1185, + "TACTGT": 1186, + "TACTGC": 1187, + "TACTGG": 1188, + "TACCAA": 1189, + "TACCAT": 1190, + "TACCAC": 1191, + "TACCAG": 1192, + "TACCTA": 1193, + "TACCTT": 1194, + "TACCTC": 1195, + "TACCTG": 1196, + "TACCCA": 1197, + "TACCCT": 1198, + "TACCCC": 1199, + "TACCCG": 1200, + "TACCGA": 1201, + "TACCGT": 1202, + "TACCGC": 1203, + "TACCGG": 1204, + "TACGAA": 1205, + "TACGAT": 1206, + "TACGAC": 1207, + "TACGAG": 1208, + "TACGTA": 1209, + "TACGTT": 1210, + "TACGTC": 1211, + "TACGTG": 1212, + "TACGCA": 1213, + "TACGCT": 1214, + "TACGCC": 1215, + "TACGCG": 1216, + "TACGGA": 1217, + "TACGGT": 1218, + "TACGGC": 1219, + "TACGGG": 1220, + "TAGAAA": 1221, + "TAGAAT": 1222, + "TAGAAC": 1223, + "TAGAAG": 1224, + "TAGATA": 1225, + "TAGATT": 1226, + "TAGATC": 1227, + "TAGATG": 1228, + "TAGACA": 1229, + "TAGACT": 1230, + "TAGACC": 1231, + "TAGACG": 1232, + "TAGAGA": 1233, + "TAGAGT": 1234, + "TAGAGC": 1235, + "TAGAGG": 1236, + "TAGTAA": 1237, + "TAGTAT": 1238, + "TAGTAC": 1239, + "TAGTAG": 1240, + "TAGTTA": 1241, + "TAGTTT": 1242, + "TAGTTC": 1243, + "TAGTTG": 1244, + "TAGTCA": 1245, + "TAGTCT": 1246, + "TAGTCC": 1247, + "TAGTCG": 1248, + "TAGTGA": 1249, + "TAGTGT": 1250, + "TAGTGC": 1251, + "TAGTGG": 1252, + "TAGCAA": 1253, + "TAGCAT": 1254, + "TAGCAC": 1255, + "TAGCAG": 1256, + "TAGCTA": 1257, + "TAGCTT": 1258, + "TAGCTC": 1259, + "TAGCTG": 1260, + "TAGCCA": 1261, + "TAGCCT": 1262, + "TAGCCC": 1263, + "TAGCCG": 1264, + "TAGCGA": 1265, + "TAGCGT": 1266, + "TAGCGC": 1267, + "TAGCGG": 1268, + "TAGGAA": 1269, + "TAGGAT": 1270, + "TAGGAC": 1271, + "TAGGAG": 1272, + "TAGGTA": 1273, + "TAGGTT": 1274, + "TAGGTC": 1275, + "TAGGTG": 1276, + "TAGGCA": 1277, + "TAGGCT": 1278, + "TAGGCC": 1279, + "TAGGCG": 1280, + "TAGGGA": 1281, + "TAGGGT": 1282, + "TAGGGC": 1283, + "TAGGGG": 1284, + "TTAAAA": 1285, + "TTAAAT": 1286, + "TTAAAC": 1287, + "TTAAAG": 1288, + "TTAATA": 1289, + "TTAATT": 1290, + "TTAATC": 1291, + "TTAATG": 1292, + "TTAACA": 1293, + "TTAACT": 1294, + "TTAACC": 1295, + "TTAACG": 1296, + "TTAAGA": 1297, + "TTAAGT": 1298, + "TTAAGC": 1299, + "TTAAGG": 1300, + "TTATAA": 1301, + "TTATAT": 1302, + "TTATAC": 1303, + "TTATAG": 1304, + "TTATTA": 1305, + "TTATTT": 1306, + "TTATTC": 1307, + "TTATTG": 1308, + "TTATCA": 1309, + "TTATCT": 1310, + "TTATCC": 1311, + "TTATCG": 1312, + "TTATGA": 1313, + "TTATGT": 1314, + "TTATGC": 1315, + "TTATGG": 1316, + "TTACAA": 1317, + "TTACAT": 1318, + "TTACAC": 1319, + "TTACAG": 1320, + "TTACTA": 1321, + "TTACTT": 1322, + "TTACTC": 1323, + "TTACTG": 1324, + "TTACCA": 1325, + "TTACCT": 1326, + "TTACCC": 1327, + "TTACCG": 1328, + "TTACGA": 1329, + "TTACGT": 1330, + "TTACGC": 1331, + "TTACGG": 1332, + "TTAGAA": 1333, + "TTAGAT": 1334, + "TTAGAC": 1335, + "TTAGAG": 1336, + "TTAGTA": 1337, + "TTAGTT": 1338, + "TTAGTC": 1339, + "TTAGTG": 1340, + "TTAGCA": 1341, + "TTAGCT": 1342, + "TTAGCC": 1343, + "TTAGCG": 1344, + "TTAGGA": 1345, + "TTAGGT": 1346, + "TTAGGC": 1347, + "TTAGGG": 1348, + "TTTAAA": 1349, + "TTTAAT": 1350, + "TTTAAC": 1351, + "TTTAAG": 1352, + "TTTATA": 1353, + "TTTATT": 1354, + "TTTATC": 1355, + "TTTATG": 1356, + "TTTACA": 1357, + "TTTACT": 1358, + "TTTACC": 1359, + "TTTACG": 1360, + "TTTAGA": 1361, + "TTTAGT": 1362, + "TTTAGC": 1363, + "TTTAGG": 1364, + "TTTTAA": 1365, + "TTTTAT": 1366, + "TTTTAC": 1367, + "TTTTAG": 1368, + "TTTTTA": 1369, + "TTTTTT": 1370, + "TTTTTC": 1371, + "TTTTTG": 1372, + "TTTTCA": 1373, + "TTTTCT": 1374, + "TTTTCC": 1375, + "TTTTCG": 1376, + "TTTTGA": 1377, + "TTTTGT": 1378, + "TTTTGC": 1379, + "TTTTGG": 1380, + "TTTCAA": 1381, + "TTTCAT": 1382, + "TTTCAC": 1383, + "TTTCAG": 1384, + "TTTCTA": 1385, + "TTTCTT": 1386, + "TTTCTC": 1387, + "TTTCTG": 1388, + "TTTCCA": 1389, + "TTTCCT": 1390, + "TTTCCC": 1391, + "TTTCCG": 1392, + "TTTCGA": 1393, + "TTTCGT": 1394, + "TTTCGC": 1395, + "TTTCGG": 1396, + "TTTGAA": 1397, + "TTTGAT": 1398, + "TTTGAC": 1399, + "TTTGAG": 1400, + "TTTGTA": 1401, + "TTTGTT": 1402, + "TTTGTC": 1403, + "TTTGTG": 1404, + "TTTGCA": 1405, + "TTTGCT": 1406, + "TTTGCC": 1407, + "TTTGCG": 1408, + "TTTGGA": 1409, + "TTTGGT": 1410, + "TTTGGC": 1411, + "TTTGGG": 1412, + "TTCAAA": 1413, + "TTCAAT": 1414, + "TTCAAC": 1415, + "TTCAAG": 1416, + "TTCATA": 1417, + "TTCATT": 1418, + "TTCATC": 1419, + "TTCATG": 1420, + "TTCACA": 1421, + "TTCACT": 1422, + "TTCACC": 1423, + "TTCACG": 1424, + "TTCAGA": 1425, + "TTCAGT": 1426, + "TTCAGC": 1427, + "TTCAGG": 1428, + "TTCTAA": 1429, + "TTCTAT": 1430, + "TTCTAC": 1431, + "TTCTAG": 1432, + "TTCTTA": 1433, + "TTCTTT": 1434, + "TTCTTC": 1435, + "TTCTTG": 1436, + "TTCTCA": 1437, + "TTCTCT": 1438, + "TTCTCC": 1439, + "TTCTCG": 1440, + "TTCTGA": 1441, + "TTCTGT": 1442, + "TTCTGC": 1443, + "TTCTGG": 1444, + "TTCCAA": 1445, + "TTCCAT": 1446, + "TTCCAC": 1447, + "TTCCAG": 1448, + "TTCCTA": 1449, + "TTCCTT": 1450, + "TTCCTC": 1451, + "TTCCTG": 1452, + "TTCCCA": 1453, + "TTCCCT": 1454, + "TTCCCC": 1455, + "TTCCCG": 1456, + "TTCCGA": 1457, + "TTCCGT": 1458, + "TTCCGC": 1459, + "TTCCGG": 1460, + "TTCGAA": 1461, + "TTCGAT": 1462, + "TTCGAC": 1463, + "TTCGAG": 1464, + "TTCGTA": 1465, + "TTCGTT": 1466, + "TTCGTC": 1467, + "TTCGTG": 1468, + "TTCGCA": 1469, + "TTCGCT": 1470, + "TTCGCC": 1471, + "TTCGCG": 1472, + "TTCGGA": 1473, + "TTCGGT": 1474, + "TTCGGC": 1475, + "TTCGGG": 1476, + "TTGAAA": 1477, + "TTGAAT": 1478, + "TTGAAC": 1479, + "TTGAAG": 1480, + "TTGATA": 1481, + "TTGATT": 1482, + "TTGATC": 1483, + "TTGATG": 1484, + "TTGACA": 1485, + "TTGACT": 1486, + "TTGACC": 1487, + "TTGACG": 1488, + "TTGAGA": 1489, + "TTGAGT": 1490, + "TTGAGC": 1491, + "TTGAGG": 1492, + "TTGTAA": 1493, + "TTGTAT": 1494, + "TTGTAC": 1495, + "TTGTAG": 1496, + "TTGTTA": 1497, + "TTGTTT": 1498, + "TTGTTC": 1499, + "TTGTTG": 1500, + "TTGTCA": 1501, + "TTGTCT": 1502, + "TTGTCC": 1503, + "TTGTCG": 1504, + "TTGTGA": 1505, + "TTGTGT": 1506, + "TTGTGC": 1507, + "TTGTGG": 1508, + "TTGCAA": 1509, + "TTGCAT": 1510, + "TTGCAC": 1511, + "TTGCAG": 1512, + "TTGCTA": 1513, + "TTGCTT": 1514, + "TTGCTC": 1515, + "TTGCTG": 1516, + "TTGCCA": 1517, + "TTGCCT": 1518, + "TTGCCC": 1519, + "TTGCCG": 1520, + "TTGCGA": 1521, + "TTGCGT": 1522, + "TTGCGC": 1523, + "TTGCGG": 1524, + "TTGGAA": 1525, + "TTGGAT": 1526, + "TTGGAC": 1527, + "TTGGAG": 1528, + "TTGGTA": 1529, + "TTGGTT": 1530, + "TTGGTC": 1531, + "TTGGTG": 1532, + "TTGGCA": 1533, + "TTGGCT": 1534, + "TTGGCC": 1535, + "TTGGCG": 1536, + "TTGGGA": 1537, + "TTGGGT": 1538, + "TTGGGC": 1539, + "TTGGGG": 1540, + "TCAAAA": 1541, + "TCAAAT": 1542, + "TCAAAC": 1543, + "TCAAAG": 1544, + "TCAATA": 1545, + "TCAATT": 1546, + "TCAATC": 1547, + "TCAATG": 1548, + "TCAACA": 1549, + "TCAACT": 1550, + "TCAACC": 1551, + "TCAACG": 1552, + "TCAAGA": 1553, + "TCAAGT": 1554, + "TCAAGC": 1555, + "TCAAGG": 1556, + "TCATAA": 1557, + "TCATAT": 1558, + "TCATAC": 1559, + "TCATAG": 1560, + "TCATTA": 1561, + "TCATTT": 1562, + "TCATTC": 1563, + "TCATTG": 1564, + "TCATCA": 1565, + "TCATCT": 1566, + "TCATCC": 1567, + "TCATCG": 1568, + "TCATGA": 1569, + "TCATGT": 1570, + "TCATGC": 1571, + "TCATGG": 1572, + "TCACAA": 1573, + "TCACAT": 1574, + "TCACAC": 1575, + "TCACAG": 1576, + "TCACTA": 1577, + "TCACTT": 1578, + "TCACTC": 1579, + "TCACTG": 1580, + "TCACCA": 1581, + "TCACCT": 1582, + "TCACCC": 1583, + "TCACCG": 1584, + "TCACGA": 1585, + "TCACGT": 1586, + "TCACGC": 1587, + "TCACGG": 1588, + "TCAGAA": 1589, + "TCAGAT": 1590, + "TCAGAC": 1591, + "TCAGAG": 1592, + "TCAGTA": 1593, + "TCAGTT": 1594, + "TCAGTC": 1595, + "TCAGTG": 1596, + "TCAGCA": 1597, + "TCAGCT": 1598, + "TCAGCC": 1599, + "TCAGCG": 1600, + "TCAGGA": 1601, + "TCAGGT": 1602, + "TCAGGC": 1603, + "TCAGGG": 1604, + "TCTAAA": 1605, + "TCTAAT": 1606, + "TCTAAC": 1607, + "TCTAAG": 1608, + "TCTATA": 1609, + "TCTATT": 1610, + "TCTATC": 1611, + "TCTATG": 1612, + "TCTACA": 1613, + "TCTACT": 1614, + "TCTACC": 1615, + "TCTACG": 1616, + "TCTAGA": 1617, + "TCTAGT": 1618, + "TCTAGC": 1619, + "TCTAGG": 1620, + "TCTTAA": 1621, + "TCTTAT": 1622, + "TCTTAC": 1623, + "TCTTAG": 1624, + "TCTTTA": 1625, + "TCTTTT": 1626, + "TCTTTC": 1627, + "TCTTTG": 1628, + "TCTTCA": 1629, + "TCTTCT": 1630, + "TCTTCC": 1631, + "TCTTCG": 1632, + "TCTTGA": 1633, + "TCTTGT": 1634, + "TCTTGC": 1635, + "TCTTGG": 1636, + "TCTCAA": 1637, + "TCTCAT": 1638, + "TCTCAC": 1639, + "TCTCAG": 1640, + "TCTCTA": 1641, + "TCTCTT": 1642, + "TCTCTC": 1643, + "TCTCTG": 1644, + "TCTCCA": 1645, + "TCTCCT": 1646, + "TCTCCC": 1647, + "TCTCCG": 1648, + "TCTCGA": 1649, + "TCTCGT": 1650, + "TCTCGC": 1651, + "TCTCGG": 1652, + "TCTGAA": 1653, + "TCTGAT": 1654, + "TCTGAC": 1655, + "TCTGAG": 1656, + "TCTGTA": 1657, + "TCTGTT": 1658, + "TCTGTC": 1659, + "TCTGTG": 1660, + "TCTGCA": 1661, + "TCTGCT": 1662, + "TCTGCC": 1663, + "TCTGCG": 1664, + "TCTGGA": 1665, + "TCTGGT": 1666, + "TCTGGC": 1667, + "TCTGGG": 1668, + "TCCAAA": 1669, + "TCCAAT": 1670, + "TCCAAC": 1671, + "TCCAAG": 1672, + "TCCATA": 1673, + "TCCATT": 1674, + "TCCATC": 1675, + "TCCATG": 1676, + "TCCACA": 1677, + "TCCACT": 1678, + "TCCACC": 1679, + "TCCACG": 1680, + "TCCAGA": 1681, + "TCCAGT": 1682, + "TCCAGC": 1683, + "TCCAGG": 1684, + "TCCTAA": 1685, + "TCCTAT": 1686, + "TCCTAC": 1687, + "TCCTAG": 1688, + "TCCTTA": 1689, + "TCCTTT": 1690, + "TCCTTC": 1691, + "TCCTTG": 1692, + "TCCTCA": 1693, + "TCCTCT": 1694, + "TCCTCC": 1695, + "TCCTCG": 1696, + "TCCTGA": 1697, + "TCCTGT": 1698, + "TCCTGC": 1699, + "TCCTGG": 1700, + "TCCCAA": 1701, + "TCCCAT": 1702, + "TCCCAC": 1703, + "TCCCAG": 1704, + "TCCCTA": 1705, + "TCCCTT": 1706, + "TCCCTC": 1707, + "TCCCTG": 1708, + "TCCCCA": 1709, + "TCCCCT": 1710, + "TCCCCC": 1711, + "TCCCCG": 1712, + "TCCCGA": 1713, + "TCCCGT": 1714, + "TCCCGC": 1715, + "TCCCGG": 1716, + "TCCGAA": 1717, + "TCCGAT": 1718, + "TCCGAC": 1719, + "TCCGAG": 1720, + "TCCGTA": 1721, + "TCCGTT": 1722, + "TCCGTC": 1723, + "TCCGTG": 1724, + "TCCGCA": 1725, + "TCCGCT": 1726, + "TCCGCC": 1727, + "TCCGCG": 1728, + "TCCGGA": 1729, + "TCCGGT": 1730, + "TCCGGC": 1731, + "TCCGGG": 1732, + "TCGAAA": 1733, + "TCGAAT": 1734, + "TCGAAC": 1735, + "TCGAAG": 1736, + "TCGATA": 1737, + "TCGATT": 1738, + "TCGATC": 1739, + "TCGATG": 1740, + "TCGACA": 1741, + "TCGACT": 1742, + "TCGACC": 1743, + "TCGACG": 1744, + "TCGAGA": 1745, + "TCGAGT": 1746, + "TCGAGC": 1747, + "TCGAGG": 1748, + "TCGTAA": 1749, + "TCGTAT": 1750, + "TCGTAC": 1751, + "TCGTAG": 1752, + "TCGTTA": 1753, + "TCGTTT": 1754, + "TCGTTC": 1755, + "TCGTTG": 1756, + "TCGTCA": 1757, + "TCGTCT": 1758, + "TCGTCC": 1759, + "TCGTCG": 1760, + "TCGTGA": 1761, + "TCGTGT": 1762, + "TCGTGC": 1763, + "TCGTGG": 1764, + "TCGCAA": 1765, + "TCGCAT": 1766, + "TCGCAC": 1767, + "TCGCAG": 1768, + "TCGCTA": 1769, + "TCGCTT": 1770, + "TCGCTC": 1771, + "TCGCTG": 1772, + "TCGCCA": 1773, + "TCGCCT": 1774, + "TCGCCC": 1775, + "TCGCCG": 1776, + "TCGCGA": 1777, + "TCGCGT": 1778, + "TCGCGC": 1779, + "TCGCGG": 1780, + "TCGGAA": 1781, + "TCGGAT": 1782, + "TCGGAC": 1783, + "TCGGAG": 1784, + "TCGGTA": 1785, + "TCGGTT": 1786, + "TCGGTC": 1787, + "TCGGTG": 1788, + "TCGGCA": 1789, + "TCGGCT": 1790, + "TCGGCC": 1791, + "TCGGCG": 1792, + "TCGGGA": 1793, + "TCGGGT": 1794, + "TCGGGC": 1795, + "TCGGGG": 1796, + "TGAAAA": 1797, + "TGAAAT": 1798, + "TGAAAC": 1799, + "TGAAAG": 1800, + "TGAATA": 1801, + "TGAATT": 1802, + "TGAATC": 1803, + "TGAATG": 1804, + "TGAACA": 1805, + "TGAACT": 1806, + "TGAACC": 1807, + "TGAACG": 1808, + "TGAAGA": 1809, + "TGAAGT": 1810, + "TGAAGC": 1811, + "TGAAGG": 1812, + "TGATAA": 1813, + "TGATAT": 1814, + "TGATAC": 1815, + "TGATAG": 1816, + "TGATTA": 1817, + "TGATTT": 1818, + "TGATTC": 1819, + "TGATTG": 1820, + "TGATCA": 1821, + "TGATCT": 1822, + "TGATCC": 1823, + "TGATCG": 1824, + "TGATGA": 1825, + "TGATGT": 1826, + "TGATGC": 1827, + "TGATGG": 1828, + "TGACAA": 1829, + "TGACAT": 1830, + "TGACAC": 1831, + "TGACAG": 1832, + "TGACTA": 1833, + "TGACTT": 1834, + "TGACTC": 1835, + "TGACTG": 1836, + "TGACCA": 1837, + "TGACCT": 1838, + "TGACCC": 1839, + "TGACCG": 1840, + "TGACGA": 1841, + "TGACGT": 1842, + "TGACGC": 1843, + "TGACGG": 1844, + "TGAGAA": 1845, + "TGAGAT": 1846, + "TGAGAC": 1847, + "TGAGAG": 1848, + "TGAGTA": 1849, + "TGAGTT": 1850, + "TGAGTC": 1851, + "TGAGTG": 1852, + "TGAGCA": 1853, + "TGAGCT": 1854, + "TGAGCC": 1855, + "TGAGCG": 1856, + "TGAGGA": 1857, + "TGAGGT": 1858, + "TGAGGC": 1859, + "TGAGGG": 1860, + "TGTAAA": 1861, + "TGTAAT": 1862, + "TGTAAC": 1863, + "TGTAAG": 1864, + "TGTATA": 1865, + "TGTATT": 1866, + "TGTATC": 1867, + "TGTATG": 1868, + "TGTACA": 1869, + "TGTACT": 1870, + "TGTACC": 1871, + "TGTACG": 1872, + "TGTAGA": 1873, + "TGTAGT": 1874, + "TGTAGC": 1875, + "TGTAGG": 1876, + "TGTTAA": 1877, + "TGTTAT": 1878, + "TGTTAC": 1879, + "TGTTAG": 1880, + "TGTTTA": 1881, + "TGTTTT": 1882, + "TGTTTC": 1883, + "TGTTTG": 1884, + "TGTTCA": 1885, + "TGTTCT": 1886, + "TGTTCC": 1887, + "TGTTCG": 1888, + "TGTTGA": 1889, + "TGTTGT": 1890, + "TGTTGC": 1891, + "TGTTGG": 1892, + "TGTCAA": 1893, + "TGTCAT": 1894, + "TGTCAC": 1895, + "TGTCAG": 1896, + "TGTCTA": 1897, + "TGTCTT": 1898, + "TGTCTC": 1899, + "TGTCTG": 1900, + "TGTCCA": 1901, + "TGTCCT": 1902, + "TGTCCC": 1903, + "TGTCCG": 1904, + "TGTCGA": 1905, + "TGTCGT": 1906, + "TGTCGC": 1907, + "TGTCGG": 1908, + "TGTGAA": 1909, + "TGTGAT": 1910, + "TGTGAC": 1911, + "TGTGAG": 1912, + "TGTGTA": 1913, + "TGTGTT": 1914, + "TGTGTC": 1915, + "TGTGTG": 1916, + "TGTGCA": 1917, + "TGTGCT": 1918, + "TGTGCC": 1919, + "TGTGCG": 1920, + "TGTGGA": 1921, + "TGTGGT": 1922, + "TGTGGC": 1923, + "TGTGGG": 1924, + "TGCAAA": 1925, + "TGCAAT": 1926, + "TGCAAC": 1927, + "TGCAAG": 1928, + "TGCATA": 1929, + "TGCATT": 1930, + "TGCATC": 1931, + "TGCATG": 1932, + "TGCACA": 1933, + "TGCACT": 1934, + "TGCACC": 1935, + "TGCACG": 1936, + "TGCAGA": 1937, + "TGCAGT": 1938, + "TGCAGC": 1939, + "TGCAGG": 1940, + "TGCTAA": 1941, + "TGCTAT": 1942, + "TGCTAC": 1943, + "TGCTAG": 1944, + "TGCTTA": 1945, + "TGCTTT": 1946, + "TGCTTC": 1947, + "TGCTTG": 1948, + "TGCTCA": 1949, + "TGCTCT": 1950, + "TGCTCC": 1951, + "TGCTCG": 1952, + "TGCTGA": 1953, + "TGCTGT": 1954, + "TGCTGC": 1955, + "TGCTGG": 1956, + "TGCCAA": 1957, + "TGCCAT": 1958, + "TGCCAC": 1959, + "TGCCAG": 1960, + "TGCCTA": 1961, + "TGCCTT": 1962, + "TGCCTC": 1963, + "TGCCTG": 1964, + "TGCCCA": 1965, + "TGCCCT": 1966, + "TGCCCC": 1967, + "TGCCCG": 1968, + "TGCCGA": 1969, + "TGCCGT": 1970, + "TGCCGC": 1971, + "TGCCGG": 1972, + "TGCGAA": 1973, + "TGCGAT": 1974, + "TGCGAC": 1975, + "TGCGAG": 1976, + "TGCGTA": 1977, + "TGCGTT": 1978, + "TGCGTC": 1979, + "TGCGTG": 1980, + "TGCGCA": 1981, + "TGCGCT": 1982, + "TGCGCC": 1983, + "TGCGCG": 1984, + "TGCGGA": 1985, + "TGCGGT": 1986, + "TGCGGC": 1987, + "TGCGGG": 1988, + "TGGAAA": 1989, + "TGGAAT": 1990, + "TGGAAC": 1991, + "TGGAAG": 1992, + "TGGATA": 1993, + "TGGATT": 1994, + "TGGATC": 1995, + "TGGATG": 1996, + "TGGACA": 1997, + "TGGACT": 1998, + "TGGACC": 1999, + "TGGACG": 2000, + "TGGAGA": 2001, + "TGGAGT": 2002, + "TGGAGC": 2003, + "TGGAGG": 2004, + "TGGTAA": 2005, + "TGGTAT": 2006, + "TGGTAC": 2007, + "TGGTAG": 2008, + "TGGTTA": 2009, + "TGGTTT": 2010, + "TGGTTC": 2011, + "TGGTTG": 2012, + "TGGTCA": 2013, + "TGGTCT": 2014, + "TGGTCC": 2015, + "TGGTCG": 2016, + "TGGTGA": 2017, + "TGGTGT": 2018, + "TGGTGC": 2019, + "TGGTGG": 2020, + "TGGCAA": 2021, + "TGGCAT": 2022, + "TGGCAC": 2023, + "TGGCAG": 2024, + "TGGCTA": 2025, + "TGGCTT": 2026, + "TGGCTC": 2027, + "TGGCTG": 2028, + "TGGCCA": 2029, + "TGGCCT": 2030, + "TGGCCC": 2031, + "TGGCCG": 2032, + "TGGCGA": 2033, + "TGGCGT": 2034, + "TGGCGC": 2035, + "TGGCGG": 2036, + "TGGGAA": 2037, + "TGGGAT": 2038, + "TGGGAC": 2039, + "TGGGAG": 2040, + "TGGGTA": 2041, + "TGGGTT": 2042, + "TGGGTC": 2043, + "TGGGTG": 2044, + "TGGGCA": 2045, + "TGGGCT": 2046, + "TGGGCC": 2047, + "TGGGCG": 2048, + "TGGGGA": 2049, + "TGGGGT": 2050, + "TGGGGC": 2051, + "TGGGGG": 2052, + "CAAAAA": 2053, + "CAAAAT": 2054, + "CAAAAC": 2055, + "CAAAAG": 2056, + "CAAATA": 2057, + "CAAATT": 2058, + "CAAATC": 2059, + "CAAATG": 2060, + "CAAACA": 2061, + "CAAACT": 2062, + "CAAACC": 2063, + "CAAACG": 2064, + "CAAAGA": 2065, + "CAAAGT": 2066, + "CAAAGC": 2067, + "CAAAGG": 2068, + "CAATAA": 2069, + "CAATAT": 2070, + "CAATAC": 2071, + "CAATAG": 2072, + "CAATTA": 2073, + "CAATTT": 2074, + "CAATTC": 2075, + "CAATTG": 2076, + "CAATCA": 2077, + "CAATCT": 2078, + "CAATCC": 2079, + "CAATCG": 2080, + "CAATGA": 2081, + "CAATGT": 2082, + "CAATGC": 2083, + "CAATGG": 2084, + "CAACAA": 2085, + "CAACAT": 2086, + "CAACAC": 2087, + "CAACAG": 2088, + "CAACTA": 2089, + "CAACTT": 2090, + "CAACTC": 2091, + "CAACTG": 2092, + "CAACCA": 2093, + "CAACCT": 2094, + "CAACCC": 2095, + "CAACCG": 2096, + "CAACGA": 2097, + "CAACGT": 2098, + "CAACGC": 2099, + "CAACGG": 2100, + "CAAGAA": 2101, + "CAAGAT": 2102, + "CAAGAC": 2103, + "CAAGAG": 2104, + "CAAGTA": 2105, + "CAAGTT": 2106, + "CAAGTC": 2107, + "CAAGTG": 2108, + "CAAGCA": 2109, + "CAAGCT": 2110, + "CAAGCC": 2111, + "CAAGCG": 2112, + "CAAGGA": 2113, + "CAAGGT": 2114, + "CAAGGC": 2115, + "CAAGGG": 2116, + "CATAAA": 2117, + "CATAAT": 2118, + "CATAAC": 2119, + "CATAAG": 2120, + "CATATA": 2121, + "CATATT": 2122, + "CATATC": 2123, + "CATATG": 2124, + "CATACA": 2125, + "CATACT": 2126, + "CATACC": 2127, + "CATACG": 2128, + "CATAGA": 2129, + "CATAGT": 2130, + "CATAGC": 2131, + "CATAGG": 2132, + "CATTAA": 2133, + "CATTAT": 2134, + "CATTAC": 2135, + "CATTAG": 2136, + "CATTTA": 2137, + "CATTTT": 2138, + "CATTTC": 2139, + "CATTTG": 2140, + "CATTCA": 2141, + "CATTCT": 2142, + "CATTCC": 2143, + "CATTCG": 2144, + "CATTGA": 2145, + "CATTGT": 2146, + "CATTGC": 2147, + "CATTGG": 2148, + "CATCAA": 2149, + "CATCAT": 2150, + "CATCAC": 2151, + "CATCAG": 2152, + "CATCTA": 2153, + "CATCTT": 2154, + "CATCTC": 2155, + "CATCTG": 2156, + "CATCCA": 2157, + "CATCCT": 2158, + "CATCCC": 2159, + "CATCCG": 2160, + "CATCGA": 2161, + "CATCGT": 2162, + "CATCGC": 2163, + "CATCGG": 2164, + "CATGAA": 2165, + "CATGAT": 2166, + "CATGAC": 2167, + "CATGAG": 2168, + "CATGTA": 2169, + "CATGTT": 2170, + "CATGTC": 2171, + "CATGTG": 2172, + "CATGCA": 2173, + "CATGCT": 2174, + "CATGCC": 2175, + "CATGCG": 2176, + "CATGGA": 2177, + "CATGGT": 2178, + "CATGGC": 2179, + "CATGGG": 2180, + "CACAAA": 2181, + "CACAAT": 2182, + "CACAAC": 2183, + "CACAAG": 2184, + "CACATA": 2185, + "CACATT": 2186, + "CACATC": 2187, + "CACATG": 2188, + "CACACA": 2189, + "CACACT": 2190, + "CACACC": 2191, + "CACACG": 2192, + "CACAGA": 2193, + "CACAGT": 2194, + "CACAGC": 2195, + "CACAGG": 2196, + "CACTAA": 2197, + "CACTAT": 2198, + "CACTAC": 2199, + "CACTAG": 2200, + "CACTTA": 2201, + "CACTTT": 2202, + "CACTTC": 2203, + "CACTTG": 2204, + "CACTCA": 2205, + "CACTCT": 2206, + "CACTCC": 2207, + "CACTCG": 2208, + "CACTGA": 2209, + "CACTGT": 2210, + "CACTGC": 2211, + "CACTGG": 2212, + "CACCAA": 2213, + "CACCAT": 2214, + "CACCAC": 2215, + "CACCAG": 2216, + "CACCTA": 2217, + "CACCTT": 2218, + "CACCTC": 2219, + "CACCTG": 2220, + "CACCCA": 2221, + "CACCCT": 2222, + "CACCCC": 2223, + "CACCCG": 2224, + "CACCGA": 2225, + "CACCGT": 2226, + "CACCGC": 2227, + "CACCGG": 2228, + "CACGAA": 2229, + "CACGAT": 2230, + "CACGAC": 2231, + "CACGAG": 2232, + "CACGTA": 2233, + "CACGTT": 2234, + "CACGTC": 2235, + "CACGTG": 2236, + "CACGCA": 2237, + "CACGCT": 2238, + "CACGCC": 2239, + "CACGCG": 2240, + "CACGGA": 2241, + "CACGGT": 2242, + "CACGGC": 2243, + "CACGGG": 2244, + "CAGAAA": 2245, + "CAGAAT": 2246, + "CAGAAC": 2247, + "CAGAAG": 2248, + "CAGATA": 2249, + "CAGATT": 2250, + "CAGATC": 2251, + "CAGATG": 2252, + "CAGACA": 2253, + "CAGACT": 2254, + "CAGACC": 2255, + "CAGACG": 2256, + "CAGAGA": 2257, + "CAGAGT": 2258, + "CAGAGC": 2259, + "CAGAGG": 2260, + "CAGTAA": 2261, + "CAGTAT": 2262, + "CAGTAC": 2263, + "CAGTAG": 2264, + "CAGTTA": 2265, + "CAGTTT": 2266, + "CAGTTC": 2267, + "CAGTTG": 2268, + "CAGTCA": 2269, + "CAGTCT": 2270, + "CAGTCC": 2271, + "CAGTCG": 2272, + "CAGTGA": 2273, + "CAGTGT": 2274, + "CAGTGC": 2275, + "CAGTGG": 2276, + "CAGCAA": 2277, + "CAGCAT": 2278, + "CAGCAC": 2279, + "CAGCAG": 2280, + "CAGCTA": 2281, + "CAGCTT": 2282, + "CAGCTC": 2283, + "CAGCTG": 2284, + "CAGCCA": 2285, + "CAGCCT": 2286, + "CAGCCC": 2287, + "CAGCCG": 2288, + "CAGCGA": 2289, + "CAGCGT": 2290, + "CAGCGC": 2291, + "CAGCGG": 2292, + "CAGGAA": 2293, + "CAGGAT": 2294, + "CAGGAC": 2295, + "CAGGAG": 2296, + "CAGGTA": 2297, + "CAGGTT": 2298, + "CAGGTC": 2299, + "CAGGTG": 2300, + "CAGGCA": 2301, + "CAGGCT": 2302, + "CAGGCC": 2303, + "CAGGCG": 2304, + "CAGGGA": 2305, + "CAGGGT": 2306, + "CAGGGC": 2307, + "CAGGGG": 2308, + "CTAAAA": 2309, + "CTAAAT": 2310, + "CTAAAC": 2311, + "CTAAAG": 2312, + "CTAATA": 2313, + "CTAATT": 2314, + "CTAATC": 2315, + "CTAATG": 2316, + "CTAACA": 2317, + "CTAACT": 2318, + "CTAACC": 2319, + "CTAACG": 2320, + "CTAAGA": 2321, + "CTAAGT": 2322, + "CTAAGC": 2323, + "CTAAGG": 2324, + "CTATAA": 2325, + "CTATAT": 2326, + "CTATAC": 2327, + "CTATAG": 2328, + "CTATTA": 2329, + "CTATTT": 2330, + "CTATTC": 2331, + "CTATTG": 2332, + "CTATCA": 2333, + "CTATCT": 2334, + "CTATCC": 2335, + "CTATCG": 2336, + "CTATGA": 2337, + "CTATGT": 2338, + "CTATGC": 2339, + "CTATGG": 2340, + "CTACAA": 2341, + "CTACAT": 2342, + "CTACAC": 2343, + "CTACAG": 2344, + "CTACTA": 2345, + "CTACTT": 2346, + "CTACTC": 2347, + "CTACTG": 2348, + "CTACCA": 2349, + "CTACCT": 2350, + "CTACCC": 2351, + "CTACCG": 2352, + "CTACGA": 2353, + "CTACGT": 2354, + "CTACGC": 2355, + "CTACGG": 2356, + "CTAGAA": 2357, + "CTAGAT": 2358, + "CTAGAC": 2359, + "CTAGAG": 2360, + "CTAGTA": 2361, + "CTAGTT": 2362, + "CTAGTC": 2363, + "CTAGTG": 2364, + "CTAGCA": 2365, + "CTAGCT": 2366, + "CTAGCC": 2367, + "CTAGCG": 2368, + "CTAGGA": 2369, + "CTAGGT": 2370, + "CTAGGC": 2371, + "CTAGGG": 2372, + "CTTAAA": 2373, + "CTTAAT": 2374, + "CTTAAC": 2375, + "CTTAAG": 2376, + "CTTATA": 2377, + "CTTATT": 2378, + "CTTATC": 2379, + "CTTATG": 2380, + "CTTACA": 2381, + "CTTACT": 2382, + "CTTACC": 2383, + "CTTACG": 2384, + "CTTAGA": 2385, + "CTTAGT": 2386, + "CTTAGC": 2387, + "CTTAGG": 2388, + "CTTTAA": 2389, + "CTTTAT": 2390, + "CTTTAC": 2391, + "CTTTAG": 2392, + "CTTTTA": 2393, + "CTTTTT": 2394, + "CTTTTC": 2395, + "CTTTTG": 2396, + "CTTTCA": 2397, + "CTTTCT": 2398, + "CTTTCC": 2399, + "CTTTCG": 2400, + "CTTTGA": 2401, + "CTTTGT": 2402, + "CTTTGC": 2403, + "CTTTGG": 2404, + "CTTCAA": 2405, + "CTTCAT": 2406, + "CTTCAC": 2407, + "CTTCAG": 2408, + "CTTCTA": 2409, + "CTTCTT": 2410, + "CTTCTC": 2411, + "CTTCTG": 2412, + "CTTCCA": 2413, + "CTTCCT": 2414, + "CTTCCC": 2415, + "CTTCCG": 2416, + "CTTCGA": 2417, + "CTTCGT": 2418, + "CTTCGC": 2419, + "CTTCGG": 2420, + "CTTGAA": 2421, + "CTTGAT": 2422, + "CTTGAC": 2423, + "CTTGAG": 2424, + "CTTGTA": 2425, + "CTTGTT": 2426, + "CTTGTC": 2427, + "CTTGTG": 2428, + "CTTGCA": 2429, + "CTTGCT": 2430, + "CTTGCC": 2431, + "CTTGCG": 2432, + "CTTGGA": 2433, + "CTTGGT": 2434, + "CTTGGC": 2435, + "CTTGGG": 2436, + "CTCAAA": 2437, + "CTCAAT": 2438, + "CTCAAC": 2439, + "CTCAAG": 2440, + "CTCATA": 2441, + "CTCATT": 2442, + "CTCATC": 2443, + "CTCATG": 2444, + "CTCACA": 2445, + "CTCACT": 2446, + "CTCACC": 2447, + "CTCACG": 2448, + "CTCAGA": 2449, + "CTCAGT": 2450, + "CTCAGC": 2451, + "CTCAGG": 2452, + "CTCTAA": 2453, + "CTCTAT": 2454, + "CTCTAC": 2455, + "CTCTAG": 2456, + "CTCTTA": 2457, + "CTCTTT": 2458, + "CTCTTC": 2459, + "CTCTTG": 2460, + "CTCTCA": 2461, + "CTCTCT": 2462, + "CTCTCC": 2463, + "CTCTCG": 2464, + "CTCTGA": 2465, + "CTCTGT": 2466, + "CTCTGC": 2467, + "CTCTGG": 2468, + "CTCCAA": 2469, + "CTCCAT": 2470, + "CTCCAC": 2471, + "CTCCAG": 2472, + "CTCCTA": 2473, + "CTCCTT": 2474, + "CTCCTC": 2475, + "CTCCTG": 2476, + "CTCCCA": 2477, + "CTCCCT": 2478, + "CTCCCC": 2479, + "CTCCCG": 2480, + "CTCCGA": 2481, + "CTCCGT": 2482, + "CTCCGC": 2483, + "CTCCGG": 2484, + "CTCGAA": 2485, + "CTCGAT": 2486, + "CTCGAC": 2487, + "CTCGAG": 2488, + "CTCGTA": 2489, + "CTCGTT": 2490, + "CTCGTC": 2491, + "CTCGTG": 2492, + "CTCGCA": 2493, + "CTCGCT": 2494, + "CTCGCC": 2495, + "CTCGCG": 2496, + "CTCGGA": 2497, + "CTCGGT": 2498, + "CTCGGC": 2499, + "CTCGGG": 2500, + "CTGAAA": 2501, + "CTGAAT": 2502, + "CTGAAC": 2503, + "CTGAAG": 2504, + "CTGATA": 2505, + "CTGATT": 2506, + "CTGATC": 2507, + "CTGATG": 2508, + "CTGACA": 2509, + "CTGACT": 2510, + "CTGACC": 2511, + "CTGACG": 2512, + "CTGAGA": 2513, + "CTGAGT": 2514, + "CTGAGC": 2515, + "CTGAGG": 2516, + "CTGTAA": 2517, + "CTGTAT": 2518, + "CTGTAC": 2519, + "CTGTAG": 2520, + "CTGTTA": 2521, + "CTGTTT": 2522, + "CTGTTC": 2523, + "CTGTTG": 2524, + "CTGTCA": 2525, + "CTGTCT": 2526, + "CTGTCC": 2527, + "CTGTCG": 2528, + "CTGTGA": 2529, + "CTGTGT": 2530, + "CTGTGC": 2531, + "CTGTGG": 2532, + "CTGCAA": 2533, + "CTGCAT": 2534, + "CTGCAC": 2535, + "CTGCAG": 2536, + "CTGCTA": 2537, + "CTGCTT": 2538, + "CTGCTC": 2539, + "CTGCTG": 2540, + "CTGCCA": 2541, + "CTGCCT": 2542, + "CTGCCC": 2543, + "CTGCCG": 2544, + "CTGCGA": 2545, + "CTGCGT": 2546, + "CTGCGC": 2547, + "CTGCGG": 2548, + "CTGGAA": 2549, + "CTGGAT": 2550, + "CTGGAC": 2551, + "CTGGAG": 2552, + "CTGGTA": 2553, + "CTGGTT": 2554, + "CTGGTC": 2555, + "CTGGTG": 2556, + "CTGGCA": 2557, + "CTGGCT": 2558, + "CTGGCC": 2559, + "CTGGCG": 2560, + "CTGGGA": 2561, + "CTGGGT": 2562, + "CTGGGC": 2563, + "CTGGGG": 2564, + "CCAAAA": 2565, + "CCAAAT": 2566, + "CCAAAC": 2567, + "CCAAAG": 2568, + "CCAATA": 2569, + "CCAATT": 2570, + "CCAATC": 2571, + "CCAATG": 2572, + "CCAACA": 2573, + "CCAACT": 2574, + "CCAACC": 2575, + "CCAACG": 2576, + "CCAAGA": 2577, + "CCAAGT": 2578, + "CCAAGC": 2579, + "CCAAGG": 2580, + "CCATAA": 2581, + "CCATAT": 2582, + "CCATAC": 2583, + "CCATAG": 2584, + "CCATTA": 2585, + "CCATTT": 2586, + "CCATTC": 2587, + "CCATTG": 2588, + "CCATCA": 2589, + "CCATCT": 2590, + "CCATCC": 2591, + "CCATCG": 2592, + "CCATGA": 2593, + "CCATGT": 2594, + "CCATGC": 2595, + "CCATGG": 2596, + "CCACAA": 2597, + "CCACAT": 2598, + "CCACAC": 2599, + "CCACAG": 2600, + "CCACTA": 2601, + "CCACTT": 2602, + "CCACTC": 2603, + "CCACTG": 2604, + "CCACCA": 2605, + "CCACCT": 2606, + "CCACCC": 2607, + "CCACCG": 2608, + "CCACGA": 2609, + "CCACGT": 2610, + "CCACGC": 2611, + "CCACGG": 2612, + "CCAGAA": 2613, + "CCAGAT": 2614, + "CCAGAC": 2615, + "CCAGAG": 2616, + "CCAGTA": 2617, + "CCAGTT": 2618, + "CCAGTC": 2619, + "CCAGTG": 2620, + "CCAGCA": 2621, + "CCAGCT": 2622, + "CCAGCC": 2623, + "CCAGCG": 2624, + "CCAGGA": 2625, + "CCAGGT": 2626, + "CCAGGC": 2627, + "CCAGGG": 2628, + "CCTAAA": 2629, + "CCTAAT": 2630, + "CCTAAC": 2631, + "CCTAAG": 2632, + "CCTATA": 2633, + "CCTATT": 2634, + "CCTATC": 2635, + "CCTATG": 2636, + "CCTACA": 2637, + "CCTACT": 2638, + "CCTACC": 2639, + "CCTACG": 2640, + "CCTAGA": 2641, + "CCTAGT": 2642, + "CCTAGC": 2643, + "CCTAGG": 2644, + "CCTTAA": 2645, + "CCTTAT": 2646, + "CCTTAC": 2647, + "CCTTAG": 2648, + "CCTTTA": 2649, + "CCTTTT": 2650, + "CCTTTC": 2651, + "CCTTTG": 2652, + "CCTTCA": 2653, + "CCTTCT": 2654, + "CCTTCC": 2655, + "CCTTCG": 2656, + "CCTTGA": 2657, + "CCTTGT": 2658, + "CCTTGC": 2659, + "CCTTGG": 2660, + "CCTCAA": 2661, + "CCTCAT": 2662, + "CCTCAC": 2663, + "CCTCAG": 2664, + "CCTCTA": 2665, + "CCTCTT": 2666, + "CCTCTC": 2667, + "CCTCTG": 2668, + "CCTCCA": 2669, + "CCTCCT": 2670, + "CCTCCC": 2671, + "CCTCCG": 2672, + "CCTCGA": 2673, + "CCTCGT": 2674, + "CCTCGC": 2675, + "CCTCGG": 2676, + "CCTGAA": 2677, + "CCTGAT": 2678, + "CCTGAC": 2679, + "CCTGAG": 2680, + "CCTGTA": 2681, + "CCTGTT": 2682, + "CCTGTC": 2683, + "CCTGTG": 2684, + "CCTGCA": 2685, + "CCTGCT": 2686, + "CCTGCC": 2687, + "CCTGCG": 2688, + "CCTGGA": 2689, + "CCTGGT": 2690, + "CCTGGC": 2691, + "CCTGGG": 2692, + "CCCAAA": 2693, + "CCCAAT": 2694, + "CCCAAC": 2695, + "CCCAAG": 2696, + "CCCATA": 2697, + "CCCATT": 2698, + "CCCATC": 2699, + "CCCATG": 2700, + "CCCACA": 2701, + "CCCACT": 2702, + "CCCACC": 2703, + "CCCACG": 2704, + "CCCAGA": 2705, + "CCCAGT": 2706, + "CCCAGC": 2707, + "CCCAGG": 2708, + "CCCTAA": 2709, + "CCCTAT": 2710, + "CCCTAC": 2711, + "CCCTAG": 2712, + "CCCTTA": 2713, + "CCCTTT": 2714, + "CCCTTC": 2715, + "CCCTTG": 2716, + "CCCTCA": 2717, + "CCCTCT": 2718, + "CCCTCC": 2719, + "CCCTCG": 2720, + "CCCTGA": 2721, + "CCCTGT": 2722, + "CCCTGC": 2723, + "CCCTGG": 2724, + "CCCCAA": 2725, + "CCCCAT": 2726, + "CCCCAC": 2727, + "CCCCAG": 2728, + "CCCCTA": 2729, + "CCCCTT": 2730, + "CCCCTC": 2731, + "CCCCTG": 2732, + "CCCCCA": 2733, + "CCCCCT": 2734, + "CCCCCC": 2735, + "CCCCCG": 2736, + "CCCCGA": 2737, + "CCCCGT": 2738, + "CCCCGC": 2739, + "CCCCGG": 2740, + "CCCGAA": 2741, + "CCCGAT": 2742, + "CCCGAC": 2743, + "CCCGAG": 2744, + "CCCGTA": 2745, + "CCCGTT": 2746, + "CCCGTC": 2747, + "CCCGTG": 2748, + "CCCGCA": 2749, + "CCCGCT": 2750, + "CCCGCC": 2751, + "CCCGCG": 2752, + "CCCGGA": 2753, + "CCCGGT": 2754, + "CCCGGC": 2755, + "CCCGGG": 2756, + "CCGAAA": 2757, + "CCGAAT": 2758, + "CCGAAC": 2759, + "CCGAAG": 2760, + "CCGATA": 2761, + "CCGATT": 2762, + "CCGATC": 2763, + "CCGATG": 2764, + "CCGACA": 2765, + "CCGACT": 2766, + "CCGACC": 2767, + "CCGACG": 2768, + "CCGAGA": 2769, + "CCGAGT": 2770, + "CCGAGC": 2771, + "CCGAGG": 2772, + "CCGTAA": 2773, + "CCGTAT": 2774, + "CCGTAC": 2775, + "CCGTAG": 2776, + "CCGTTA": 2777, + "CCGTTT": 2778, + "CCGTTC": 2779, + "CCGTTG": 2780, + "CCGTCA": 2781, + "CCGTCT": 2782, + "CCGTCC": 2783, + "CCGTCG": 2784, + "CCGTGA": 2785, + "CCGTGT": 2786, + "CCGTGC": 2787, + "CCGTGG": 2788, + "CCGCAA": 2789, + "CCGCAT": 2790, + "CCGCAC": 2791, + "CCGCAG": 2792, + "CCGCTA": 2793, + "CCGCTT": 2794, + "CCGCTC": 2795, + "CCGCTG": 2796, + "CCGCCA": 2797, + "CCGCCT": 2798, + "CCGCCC": 2799, + "CCGCCG": 2800, + "CCGCGA": 2801, + "CCGCGT": 2802, + "CCGCGC": 2803, + "CCGCGG": 2804, + "CCGGAA": 2805, + "CCGGAT": 2806, + "CCGGAC": 2807, + "CCGGAG": 2808, + "CCGGTA": 2809, + "CCGGTT": 2810, + "CCGGTC": 2811, + "CCGGTG": 2812, + "CCGGCA": 2813, + "CCGGCT": 2814, + "CCGGCC": 2815, + "CCGGCG": 2816, + "CCGGGA": 2817, + "CCGGGT": 2818, + "CCGGGC": 2819, + "CCGGGG": 2820, + "CGAAAA": 2821, + "CGAAAT": 2822, + "CGAAAC": 2823, + "CGAAAG": 2824, + "CGAATA": 2825, + "CGAATT": 2826, + "CGAATC": 2827, + "CGAATG": 2828, + "CGAACA": 2829, + "CGAACT": 2830, + "CGAACC": 2831, + "CGAACG": 2832, + "CGAAGA": 2833, + "CGAAGT": 2834, + "CGAAGC": 2835, + "CGAAGG": 2836, + "CGATAA": 2837, + "CGATAT": 2838, + "CGATAC": 2839, + "CGATAG": 2840, + "CGATTA": 2841, + "CGATTT": 2842, + "CGATTC": 2843, + "CGATTG": 2844, + "CGATCA": 2845, + "CGATCT": 2846, + "CGATCC": 2847, + "CGATCG": 2848, + "CGATGA": 2849, + "CGATGT": 2850, + "CGATGC": 2851, + "CGATGG": 2852, + "CGACAA": 2853, + "CGACAT": 2854, + "CGACAC": 2855, + "CGACAG": 2856, + "CGACTA": 2857, + "CGACTT": 2858, + "CGACTC": 2859, + "CGACTG": 2860, + "CGACCA": 2861, + "CGACCT": 2862, + "CGACCC": 2863, + "CGACCG": 2864, + "CGACGA": 2865, + "CGACGT": 2866, + "CGACGC": 2867, + "CGACGG": 2868, + "CGAGAA": 2869, + "CGAGAT": 2870, + "CGAGAC": 2871, + "CGAGAG": 2872, + "CGAGTA": 2873, + "CGAGTT": 2874, + "CGAGTC": 2875, + "CGAGTG": 2876, + "CGAGCA": 2877, + "CGAGCT": 2878, + "CGAGCC": 2879, + "CGAGCG": 2880, + "CGAGGA": 2881, + "CGAGGT": 2882, + "CGAGGC": 2883, + "CGAGGG": 2884, + "CGTAAA": 2885, + "CGTAAT": 2886, + "CGTAAC": 2887, + "CGTAAG": 2888, + "CGTATA": 2889, + "CGTATT": 2890, + "CGTATC": 2891, + "CGTATG": 2892, + "CGTACA": 2893, + "CGTACT": 2894, + "CGTACC": 2895, + "CGTACG": 2896, + "CGTAGA": 2897, + "CGTAGT": 2898, + "CGTAGC": 2899, + "CGTAGG": 2900, + "CGTTAA": 2901, + "CGTTAT": 2902, + "CGTTAC": 2903, + "CGTTAG": 2904, + "CGTTTA": 2905, + "CGTTTT": 2906, + "CGTTTC": 2907, + "CGTTTG": 2908, + "CGTTCA": 2909, + "CGTTCT": 2910, + "CGTTCC": 2911, + "CGTTCG": 2912, + "CGTTGA": 2913, + "CGTTGT": 2914, + "CGTTGC": 2915, + "CGTTGG": 2916, + "CGTCAA": 2917, + "CGTCAT": 2918, + "CGTCAC": 2919, + "CGTCAG": 2920, + "CGTCTA": 2921, + "CGTCTT": 2922, + "CGTCTC": 2923, + "CGTCTG": 2924, + "CGTCCA": 2925, + "CGTCCT": 2926, + "CGTCCC": 2927, + "CGTCCG": 2928, + "CGTCGA": 2929, + "CGTCGT": 2930, + "CGTCGC": 2931, + "CGTCGG": 2932, + "CGTGAA": 2933, + "CGTGAT": 2934, + "CGTGAC": 2935, + "CGTGAG": 2936, + "CGTGTA": 2937, + "CGTGTT": 2938, + "CGTGTC": 2939, + "CGTGTG": 2940, + "CGTGCA": 2941, + "CGTGCT": 2942, + "CGTGCC": 2943, + "CGTGCG": 2944, + "CGTGGA": 2945, + "CGTGGT": 2946, + "CGTGGC": 2947, + "CGTGGG": 2948, + "CGCAAA": 2949, + "CGCAAT": 2950, + "CGCAAC": 2951, + "CGCAAG": 2952, + "CGCATA": 2953, + "CGCATT": 2954, + "CGCATC": 2955, + "CGCATG": 2956, + "CGCACA": 2957, + "CGCACT": 2958, + "CGCACC": 2959, + "CGCACG": 2960, + "CGCAGA": 2961, + "CGCAGT": 2962, + "CGCAGC": 2963, + "CGCAGG": 2964, + "CGCTAA": 2965, + "CGCTAT": 2966, + "CGCTAC": 2967, + "CGCTAG": 2968, + "CGCTTA": 2969, + "CGCTTT": 2970, + "CGCTTC": 2971, + "CGCTTG": 2972, + "CGCTCA": 2973, + "CGCTCT": 2974, + "CGCTCC": 2975, + "CGCTCG": 2976, + "CGCTGA": 2977, + "CGCTGT": 2978, + "CGCTGC": 2979, + "CGCTGG": 2980, + "CGCCAA": 2981, + "CGCCAT": 2982, + "CGCCAC": 2983, + "CGCCAG": 2984, + "CGCCTA": 2985, + "CGCCTT": 2986, + "CGCCTC": 2987, + "CGCCTG": 2988, + "CGCCCA": 2989, + "CGCCCT": 2990, + "CGCCCC": 2991, + "CGCCCG": 2992, + "CGCCGA": 2993, + "CGCCGT": 2994, + "CGCCGC": 2995, + "CGCCGG": 2996, + "CGCGAA": 2997, + "CGCGAT": 2998, + "CGCGAC": 2999, + "CGCGAG": 3000, + "CGCGTA": 3001, + "CGCGTT": 3002, + "CGCGTC": 3003, + "CGCGTG": 3004, + "CGCGCA": 3005, + "CGCGCT": 3006, + "CGCGCC": 3007, + "CGCGCG": 3008, + "CGCGGA": 3009, + "CGCGGT": 3010, + "CGCGGC": 3011, + "CGCGGG": 3012, + "CGGAAA": 3013, + "CGGAAT": 3014, + "CGGAAC": 3015, + "CGGAAG": 3016, + "CGGATA": 3017, + "CGGATT": 3018, + "CGGATC": 3019, + "CGGATG": 3020, + "CGGACA": 3021, + "CGGACT": 3022, + "CGGACC": 3023, + "CGGACG": 3024, + "CGGAGA": 3025, + "CGGAGT": 3026, + "CGGAGC": 3027, + "CGGAGG": 3028, + "CGGTAA": 3029, + "CGGTAT": 3030, + "CGGTAC": 3031, + "CGGTAG": 3032, + "CGGTTA": 3033, + "CGGTTT": 3034, + "CGGTTC": 3035, + "CGGTTG": 3036, + "CGGTCA": 3037, + "CGGTCT": 3038, + "CGGTCC": 3039, + "CGGTCG": 3040, + "CGGTGA": 3041, + "CGGTGT": 3042, + "CGGTGC": 3043, + "CGGTGG": 3044, + "CGGCAA": 3045, + "CGGCAT": 3046, + "CGGCAC": 3047, + "CGGCAG": 3048, + "CGGCTA": 3049, + "CGGCTT": 3050, + "CGGCTC": 3051, + "CGGCTG": 3052, + "CGGCCA": 3053, + "CGGCCT": 3054, + "CGGCCC": 3055, + "CGGCCG": 3056, + "CGGCGA": 3057, + "CGGCGT": 3058, + "CGGCGC": 3059, + "CGGCGG": 3060, + "CGGGAA": 3061, + "CGGGAT": 3062, + "CGGGAC": 3063, + "CGGGAG": 3064, + "CGGGTA": 3065, + "CGGGTT": 3066, + "CGGGTC": 3067, + "CGGGTG": 3068, + "CGGGCA": 3069, + "CGGGCT": 3070, + "CGGGCC": 3071, + "CGGGCG": 3072, + "CGGGGA": 3073, + "CGGGGT": 3074, + "CGGGGC": 3075, + "CGGGGG": 3076, + "GAAAAA": 3077, + "GAAAAT": 3078, + "GAAAAC": 3079, + "GAAAAG": 3080, + "GAAATA": 3081, + "GAAATT": 3082, + "GAAATC": 3083, + "GAAATG": 3084, + "GAAACA": 3085, + "GAAACT": 3086, + "GAAACC": 3087, + "GAAACG": 3088, + "GAAAGA": 3089, + "GAAAGT": 3090, + "GAAAGC": 3091, + "GAAAGG": 3092, + "GAATAA": 3093, + "GAATAT": 3094, + "GAATAC": 3095, + "GAATAG": 3096, + "GAATTA": 3097, + "GAATTT": 3098, + "GAATTC": 3099, + "GAATTG": 3100, + "GAATCA": 3101, + "GAATCT": 3102, + "GAATCC": 3103, + "GAATCG": 3104, + "GAATGA": 3105, + "GAATGT": 3106, + "GAATGC": 3107, + "GAATGG": 3108, + "GAACAA": 3109, + "GAACAT": 3110, + "GAACAC": 3111, + "GAACAG": 3112, + "GAACTA": 3113, + "GAACTT": 3114, + "GAACTC": 3115, + "GAACTG": 3116, + "GAACCA": 3117, + "GAACCT": 3118, + "GAACCC": 3119, + "GAACCG": 3120, + "GAACGA": 3121, + "GAACGT": 3122, + "GAACGC": 3123, + "GAACGG": 3124, + "GAAGAA": 3125, + "GAAGAT": 3126, + "GAAGAC": 3127, + "GAAGAG": 3128, + "GAAGTA": 3129, + "GAAGTT": 3130, + "GAAGTC": 3131, + "GAAGTG": 3132, + "GAAGCA": 3133, + "GAAGCT": 3134, + "GAAGCC": 3135, + "GAAGCG": 3136, + "GAAGGA": 3137, + "GAAGGT": 3138, + "GAAGGC": 3139, + "GAAGGG": 3140, + "GATAAA": 3141, + "GATAAT": 3142, + "GATAAC": 3143, + "GATAAG": 3144, + "GATATA": 3145, + "GATATT": 3146, + "GATATC": 3147, + "GATATG": 3148, + "GATACA": 3149, + "GATACT": 3150, + "GATACC": 3151, + "GATACG": 3152, + "GATAGA": 3153, + "GATAGT": 3154, + "GATAGC": 3155, + "GATAGG": 3156, + "GATTAA": 3157, + "GATTAT": 3158, + "GATTAC": 3159, + "GATTAG": 3160, + "GATTTA": 3161, + "GATTTT": 3162, + "GATTTC": 3163, + "GATTTG": 3164, + "GATTCA": 3165, + "GATTCT": 3166, + "GATTCC": 3167, + "GATTCG": 3168, + "GATTGA": 3169, + "GATTGT": 3170, + "GATTGC": 3171, + "GATTGG": 3172, + "GATCAA": 3173, + "GATCAT": 3174, + "GATCAC": 3175, + "GATCAG": 3176, + "GATCTA": 3177, + "GATCTT": 3178, + "GATCTC": 3179, + "GATCTG": 3180, + "GATCCA": 3181, + "GATCCT": 3182, + "GATCCC": 3183, + "GATCCG": 3184, + "GATCGA": 3185, + "GATCGT": 3186, + "GATCGC": 3187, + "GATCGG": 3188, + "GATGAA": 3189, + "GATGAT": 3190, + "GATGAC": 3191, + "GATGAG": 3192, + "GATGTA": 3193, + "GATGTT": 3194, + "GATGTC": 3195, + "GATGTG": 3196, + "GATGCA": 3197, + "GATGCT": 3198, + "GATGCC": 3199, + "GATGCG": 3200, + "GATGGA": 3201, + "GATGGT": 3202, + "GATGGC": 3203, + "GATGGG": 3204, + "GACAAA": 3205, + "GACAAT": 3206, + "GACAAC": 3207, + "GACAAG": 3208, + "GACATA": 3209, + "GACATT": 3210, + "GACATC": 3211, + "GACATG": 3212, + "GACACA": 3213, + "GACACT": 3214, + "GACACC": 3215, + "GACACG": 3216, + "GACAGA": 3217, + "GACAGT": 3218, + "GACAGC": 3219, + "GACAGG": 3220, + "GACTAA": 3221, + "GACTAT": 3222, + "GACTAC": 3223, + "GACTAG": 3224, + "GACTTA": 3225, + "GACTTT": 3226, + "GACTTC": 3227, + "GACTTG": 3228, + "GACTCA": 3229, + "GACTCT": 3230, + "GACTCC": 3231, + "GACTCG": 3232, + "GACTGA": 3233, + "GACTGT": 3234, + "GACTGC": 3235, + "GACTGG": 3236, + "GACCAA": 3237, + "GACCAT": 3238, + "GACCAC": 3239, + "GACCAG": 3240, + "GACCTA": 3241, + "GACCTT": 3242, + "GACCTC": 3243, + "GACCTG": 3244, + "GACCCA": 3245, + "GACCCT": 3246, + "GACCCC": 3247, + "GACCCG": 3248, + "GACCGA": 3249, + "GACCGT": 3250, + "GACCGC": 3251, + "GACCGG": 3252, + "GACGAA": 3253, + "GACGAT": 3254, + "GACGAC": 3255, + "GACGAG": 3256, + "GACGTA": 3257, + "GACGTT": 3258, + "GACGTC": 3259, + "GACGTG": 3260, + "GACGCA": 3261, + "GACGCT": 3262, + "GACGCC": 3263, + "GACGCG": 3264, + "GACGGA": 3265, + "GACGGT": 3266, + "GACGGC": 3267, + "GACGGG": 3268, + "GAGAAA": 3269, + "GAGAAT": 3270, + "GAGAAC": 3271, + "GAGAAG": 3272, + "GAGATA": 3273, + "GAGATT": 3274, + "GAGATC": 3275, + "GAGATG": 3276, + "GAGACA": 3277, + "GAGACT": 3278, + "GAGACC": 3279, + "GAGACG": 3280, + "GAGAGA": 3281, + "GAGAGT": 3282, + "GAGAGC": 3283, + "GAGAGG": 3284, + "GAGTAA": 3285, + "GAGTAT": 3286, + "GAGTAC": 3287, + "GAGTAG": 3288, + "GAGTTA": 3289, + "GAGTTT": 3290, + "GAGTTC": 3291, + "GAGTTG": 3292, + "GAGTCA": 3293, + "GAGTCT": 3294, + "GAGTCC": 3295, + "GAGTCG": 3296, + "GAGTGA": 3297, + "GAGTGT": 3298, + "GAGTGC": 3299, + "GAGTGG": 3300, + "GAGCAA": 3301, + "GAGCAT": 3302, + "GAGCAC": 3303, + "GAGCAG": 3304, + "GAGCTA": 3305, + "GAGCTT": 3306, + "GAGCTC": 3307, + "GAGCTG": 3308, + "GAGCCA": 3309, + "GAGCCT": 3310, + "GAGCCC": 3311, + "GAGCCG": 3312, + "GAGCGA": 3313, + "GAGCGT": 3314, + "GAGCGC": 3315, + "GAGCGG": 3316, + "GAGGAA": 3317, + "GAGGAT": 3318, + "GAGGAC": 3319, + "GAGGAG": 3320, + "GAGGTA": 3321, + "GAGGTT": 3322, + "GAGGTC": 3323, + "GAGGTG": 3324, + "GAGGCA": 3325, + "GAGGCT": 3326, + "GAGGCC": 3327, + "GAGGCG": 3328, + "GAGGGA": 3329, + "GAGGGT": 3330, + "GAGGGC": 3331, + "GAGGGG": 3332, + "GTAAAA": 3333, + "GTAAAT": 3334, + "GTAAAC": 3335, + "GTAAAG": 3336, + "GTAATA": 3337, + "GTAATT": 3338, + "GTAATC": 3339, + "GTAATG": 3340, + "GTAACA": 3341, + "GTAACT": 3342, + "GTAACC": 3343, + "GTAACG": 3344, + "GTAAGA": 3345, + "GTAAGT": 3346, + "GTAAGC": 3347, + "GTAAGG": 3348, + "GTATAA": 3349, + "GTATAT": 3350, + "GTATAC": 3351, + "GTATAG": 3352, + "GTATTA": 3353, + "GTATTT": 3354, + "GTATTC": 3355, + "GTATTG": 3356, + "GTATCA": 3357, + "GTATCT": 3358, + "GTATCC": 3359, + "GTATCG": 3360, + "GTATGA": 3361, + "GTATGT": 3362, + "GTATGC": 3363, + "GTATGG": 3364, + "GTACAA": 3365, + "GTACAT": 3366, + "GTACAC": 3367, + "GTACAG": 3368, + "GTACTA": 3369, + "GTACTT": 3370, + "GTACTC": 3371, + "GTACTG": 3372, + "GTACCA": 3373, + "GTACCT": 3374, + "GTACCC": 3375, + "GTACCG": 3376, + "GTACGA": 3377, + "GTACGT": 3378, + "GTACGC": 3379, + "GTACGG": 3380, + "GTAGAA": 3381, + "GTAGAT": 3382, + "GTAGAC": 3383, + "GTAGAG": 3384, + "GTAGTA": 3385, + "GTAGTT": 3386, + "GTAGTC": 3387, + "GTAGTG": 3388, + "GTAGCA": 3389, + "GTAGCT": 3390, + "GTAGCC": 3391, + "GTAGCG": 3392, + "GTAGGA": 3393, + "GTAGGT": 3394, + "GTAGGC": 3395, + "GTAGGG": 3396, + "GTTAAA": 3397, + "GTTAAT": 3398, + "GTTAAC": 3399, + "GTTAAG": 3400, + "GTTATA": 3401, + "GTTATT": 3402, + "GTTATC": 3403, + "GTTATG": 3404, + "GTTACA": 3405, + "GTTACT": 3406, + "GTTACC": 3407, + "GTTACG": 3408, + "GTTAGA": 3409, + "GTTAGT": 3410, + "GTTAGC": 3411, + "GTTAGG": 3412, + "GTTTAA": 3413, + "GTTTAT": 3414, + "GTTTAC": 3415, + "GTTTAG": 3416, + "GTTTTA": 3417, + "GTTTTT": 3418, + "GTTTTC": 3419, + "GTTTTG": 3420, + "GTTTCA": 3421, + "GTTTCT": 3422, + "GTTTCC": 3423, + "GTTTCG": 3424, + "GTTTGA": 3425, + "GTTTGT": 3426, + "GTTTGC": 3427, + "GTTTGG": 3428, + "GTTCAA": 3429, + "GTTCAT": 3430, + "GTTCAC": 3431, + "GTTCAG": 3432, + "GTTCTA": 3433, + "GTTCTT": 3434, + "GTTCTC": 3435, + "GTTCTG": 3436, + "GTTCCA": 3437, + "GTTCCT": 3438, + "GTTCCC": 3439, + "GTTCCG": 3440, + "GTTCGA": 3441, + "GTTCGT": 3442, + "GTTCGC": 3443, + "GTTCGG": 3444, + "GTTGAA": 3445, + "GTTGAT": 3446, + "GTTGAC": 3447, + "GTTGAG": 3448, + "GTTGTA": 3449, + "GTTGTT": 3450, + "GTTGTC": 3451, + "GTTGTG": 3452, + "GTTGCA": 3453, + "GTTGCT": 3454, + "GTTGCC": 3455, + "GTTGCG": 3456, + "GTTGGA": 3457, + "GTTGGT": 3458, + "GTTGGC": 3459, + "GTTGGG": 3460, + "GTCAAA": 3461, + "GTCAAT": 3462, + "GTCAAC": 3463, + "GTCAAG": 3464, + "GTCATA": 3465, + "GTCATT": 3466, + "GTCATC": 3467, + "GTCATG": 3468, + "GTCACA": 3469, + "GTCACT": 3470, + "GTCACC": 3471, + "GTCACG": 3472, + "GTCAGA": 3473, + "GTCAGT": 3474, + "GTCAGC": 3475, + "GTCAGG": 3476, + "GTCTAA": 3477, + "GTCTAT": 3478, + "GTCTAC": 3479, + "GTCTAG": 3480, + "GTCTTA": 3481, + "GTCTTT": 3482, + "GTCTTC": 3483, + "GTCTTG": 3484, + "GTCTCA": 3485, + "GTCTCT": 3486, + "GTCTCC": 3487, + "GTCTCG": 3488, + "GTCTGA": 3489, + "GTCTGT": 3490, + "GTCTGC": 3491, + "GTCTGG": 3492, + "GTCCAA": 3493, + "GTCCAT": 3494, + "GTCCAC": 3495, + "GTCCAG": 3496, + "GTCCTA": 3497, + "GTCCTT": 3498, + "GTCCTC": 3499, + "GTCCTG": 3500, + "GTCCCA": 3501, + "GTCCCT": 3502, + "GTCCCC": 3503, + "GTCCCG": 3504, + "GTCCGA": 3505, + "GTCCGT": 3506, + "GTCCGC": 3507, + "GTCCGG": 3508, + "GTCGAA": 3509, + "GTCGAT": 3510, + "GTCGAC": 3511, + "GTCGAG": 3512, + "GTCGTA": 3513, + "GTCGTT": 3514, + "GTCGTC": 3515, + "GTCGTG": 3516, + "GTCGCA": 3517, + "GTCGCT": 3518, + "GTCGCC": 3519, + "GTCGCG": 3520, + "GTCGGA": 3521, + "GTCGGT": 3522, + "GTCGGC": 3523, + "GTCGGG": 3524, + "GTGAAA": 3525, + "GTGAAT": 3526, + "GTGAAC": 3527, + "GTGAAG": 3528, + "GTGATA": 3529, + "GTGATT": 3530, + "GTGATC": 3531, + "GTGATG": 3532, + "GTGACA": 3533, + "GTGACT": 3534, + "GTGACC": 3535, + "GTGACG": 3536, + "GTGAGA": 3537, + "GTGAGT": 3538, + "GTGAGC": 3539, + "GTGAGG": 3540, + "GTGTAA": 3541, + "GTGTAT": 3542, + "GTGTAC": 3543, + "GTGTAG": 3544, + "GTGTTA": 3545, + "GTGTTT": 3546, + "GTGTTC": 3547, + "GTGTTG": 3548, + "GTGTCA": 3549, + "GTGTCT": 3550, + "GTGTCC": 3551, + "GTGTCG": 3552, + "GTGTGA": 3553, + "GTGTGT": 3554, + "GTGTGC": 3555, + "GTGTGG": 3556, + "GTGCAA": 3557, + "GTGCAT": 3558, + "GTGCAC": 3559, + "GTGCAG": 3560, + "GTGCTA": 3561, + "GTGCTT": 3562, + "GTGCTC": 3563, + "GTGCTG": 3564, + "GTGCCA": 3565, + "GTGCCT": 3566, + "GTGCCC": 3567, + "GTGCCG": 3568, + "GTGCGA": 3569, + "GTGCGT": 3570, + "GTGCGC": 3571, + "GTGCGG": 3572, + "GTGGAA": 3573, + "GTGGAT": 3574, + "GTGGAC": 3575, + "GTGGAG": 3576, + "GTGGTA": 3577, + "GTGGTT": 3578, + "GTGGTC": 3579, + "GTGGTG": 3580, + "GTGGCA": 3581, + "GTGGCT": 3582, + "GTGGCC": 3583, + "GTGGCG": 3584, + "GTGGGA": 3585, + "GTGGGT": 3586, + "GTGGGC": 3587, + "GTGGGG": 3588, + "GCAAAA": 3589, + "GCAAAT": 3590, + "GCAAAC": 3591, + "GCAAAG": 3592, + "GCAATA": 3593, + "GCAATT": 3594, + "GCAATC": 3595, + "GCAATG": 3596, + "GCAACA": 3597, + "GCAACT": 3598, + "GCAACC": 3599, + "GCAACG": 3600, + "GCAAGA": 3601, + "GCAAGT": 3602, + "GCAAGC": 3603, + "GCAAGG": 3604, + "GCATAA": 3605, + "GCATAT": 3606, + "GCATAC": 3607, + "GCATAG": 3608, + "GCATTA": 3609, + "GCATTT": 3610, + "GCATTC": 3611, + "GCATTG": 3612, + "GCATCA": 3613, + "GCATCT": 3614, + "GCATCC": 3615, + "GCATCG": 3616, + "GCATGA": 3617, + "GCATGT": 3618, + "GCATGC": 3619, + "GCATGG": 3620, + "GCACAA": 3621, + "GCACAT": 3622, + "GCACAC": 3623, + "GCACAG": 3624, + "GCACTA": 3625, + "GCACTT": 3626, + "GCACTC": 3627, + "GCACTG": 3628, + "GCACCA": 3629, + "GCACCT": 3630, + "GCACCC": 3631, + "GCACCG": 3632, + "GCACGA": 3633, + "GCACGT": 3634, + "GCACGC": 3635, + "GCACGG": 3636, + "GCAGAA": 3637, + "GCAGAT": 3638, + "GCAGAC": 3639, + "GCAGAG": 3640, + "GCAGTA": 3641, + "GCAGTT": 3642, + "GCAGTC": 3643, + "GCAGTG": 3644, + "GCAGCA": 3645, + "GCAGCT": 3646, + "GCAGCC": 3647, + "GCAGCG": 3648, + "GCAGGA": 3649, + "GCAGGT": 3650, + "GCAGGC": 3651, + "GCAGGG": 3652, + "GCTAAA": 3653, + "GCTAAT": 3654, + "GCTAAC": 3655, + "GCTAAG": 3656, + "GCTATA": 3657, + "GCTATT": 3658, + "GCTATC": 3659, + "GCTATG": 3660, + "GCTACA": 3661, + "GCTACT": 3662, + "GCTACC": 3663, + "GCTACG": 3664, + "GCTAGA": 3665, + "GCTAGT": 3666, + "GCTAGC": 3667, + "GCTAGG": 3668, + "GCTTAA": 3669, + "GCTTAT": 3670, + "GCTTAC": 3671, + "GCTTAG": 3672, + "GCTTTA": 3673, + "GCTTTT": 3674, + "GCTTTC": 3675, + "GCTTTG": 3676, + "GCTTCA": 3677, + "GCTTCT": 3678, + "GCTTCC": 3679, + "GCTTCG": 3680, + "GCTTGA": 3681, + "GCTTGT": 3682, + "GCTTGC": 3683, + "GCTTGG": 3684, + "GCTCAA": 3685, + "GCTCAT": 3686, + "GCTCAC": 3687, + "GCTCAG": 3688, + "GCTCTA": 3689, + "GCTCTT": 3690, + "GCTCTC": 3691, + "GCTCTG": 3692, + "GCTCCA": 3693, + "GCTCCT": 3694, + "GCTCCC": 3695, + "GCTCCG": 3696, + "GCTCGA": 3697, + "GCTCGT": 3698, + "GCTCGC": 3699, + "GCTCGG": 3700, + "GCTGAA": 3701, + "GCTGAT": 3702, + "GCTGAC": 3703, + "GCTGAG": 3704, + "GCTGTA": 3705, + "GCTGTT": 3706, + "GCTGTC": 3707, + "GCTGTG": 3708, + "GCTGCA": 3709, + "GCTGCT": 3710, + "GCTGCC": 3711, + "GCTGCG": 3712, + "GCTGGA": 3713, + "GCTGGT": 3714, + "GCTGGC": 3715, + "GCTGGG": 3716, + "GCCAAA": 3717, + "GCCAAT": 3718, + "GCCAAC": 3719, + "GCCAAG": 3720, + "GCCATA": 3721, + "GCCATT": 3722, + "GCCATC": 3723, + "GCCATG": 3724, + "GCCACA": 3725, + "GCCACT": 3726, + "GCCACC": 3727, + "GCCACG": 3728, + "GCCAGA": 3729, + "GCCAGT": 3730, + "GCCAGC": 3731, + "GCCAGG": 3732, + "GCCTAA": 3733, + "GCCTAT": 3734, + "GCCTAC": 3735, + "GCCTAG": 3736, + "GCCTTA": 3737, + "GCCTTT": 3738, + "GCCTTC": 3739, + "GCCTTG": 3740, + "GCCTCA": 3741, + "GCCTCT": 3742, + "GCCTCC": 3743, + "GCCTCG": 3744, + "GCCTGA": 3745, + "GCCTGT": 3746, + "GCCTGC": 3747, + "GCCTGG": 3748, + "GCCCAA": 3749, + "GCCCAT": 3750, + "GCCCAC": 3751, + "GCCCAG": 3752, + "GCCCTA": 3753, + "GCCCTT": 3754, + "GCCCTC": 3755, + "GCCCTG": 3756, + "GCCCCA": 3757, + "GCCCCT": 3758, + "GCCCCC": 3759, + "GCCCCG": 3760, + "GCCCGA": 3761, + "GCCCGT": 3762, + "GCCCGC": 3763, + "GCCCGG": 3764, + "GCCGAA": 3765, + "GCCGAT": 3766, + "GCCGAC": 3767, + "GCCGAG": 3768, + "GCCGTA": 3769, + "GCCGTT": 3770, + "GCCGTC": 3771, + "GCCGTG": 3772, + "GCCGCA": 3773, + "GCCGCT": 3774, + "GCCGCC": 3775, + "GCCGCG": 3776, + "GCCGGA": 3777, + "GCCGGT": 3778, + "GCCGGC": 3779, + "GCCGGG": 3780, + "GCGAAA": 3781, + "GCGAAT": 3782, + "GCGAAC": 3783, + "GCGAAG": 3784, + "GCGATA": 3785, + "GCGATT": 3786, + "GCGATC": 3787, + "GCGATG": 3788, + "GCGACA": 3789, + "GCGACT": 3790, + "GCGACC": 3791, + "GCGACG": 3792, + "GCGAGA": 3793, + "GCGAGT": 3794, + "GCGAGC": 3795, + "GCGAGG": 3796, + "GCGTAA": 3797, + "GCGTAT": 3798, + "GCGTAC": 3799, + "GCGTAG": 3800, + "GCGTTA": 3801, + "GCGTTT": 3802, + "GCGTTC": 3803, + "GCGTTG": 3804, + "GCGTCA": 3805, + "GCGTCT": 3806, + "GCGTCC": 3807, + "GCGTCG": 3808, + "GCGTGA": 3809, + "GCGTGT": 3810, + "GCGTGC": 3811, + "GCGTGG": 3812, + "GCGCAA": 3813, + "GCGCAT": 3814, + "GCGCAC": 3815, + "GCGCAG": 3816, + "GCGCTA": 3817, + "GCGCTT": 3818, + "GCGCTC": 3819, + "GCGCTG": 3820, + "GCGCCA": 3821, + "GCGCCT": 3822, + "GCGCCC": 3823, + "GCGCCG": 3824, + "GCGCGA": 3825, + "GCGCGT": 3826, + "GCGCGC": 3827, + "GCGCGG": 3828, + "GCGGAA": 3829, + "GCGGAT": 3830, + "GCGGAC": 3831, + "GCGGAG": 3832, + "GCGGTA": 3833, + "GCGGTT": 3834, + "GCGGTC": 3835, + "GCGGTG": 3836, + "GCGGCA": 3837, + "GCGGCT": 3838, + "GCGGCC": 3839, + "GCGGCG": 3840, + "GCGGGA": 3841, + "GCGGGT": 3842, + "GCGGGC": 3843, + "GCGGGG": 3844, + "GGAAAA": 3845, + "GGAAAT": 3846, + "GGAAAC": 3847, + "GGAAAG": 3848, + "GGAATA": 3849, + "GGAATT": 3850, + "GGAATC": 3851, + "GGAATG": 3852, + "GGAACA": 3853, + "GGAACT": 3854, + "GGAACC": 3855, + "GGAACG": 3856, + "GGAAGA": 3857, + "GGAAGT": 3858, + "GGAAGC": 3859, + "GGAAGG": 3860, + "GGATAA": 3861, + "GGATAT": 3862, + "GGATAC": 3863, + "GGATAG": 3864, + "GGATTA": 3865, + "GGATTT": 3866, + "GGATTC": 3867, + "GGATTG": 3868, + "GGATCA": 3869, + "GGATCT": 3870, + "GGATCC": 3871, + "GGATCG": 3872, + "GGATGA": 3873, + "GGATGT": 3874, + "GGATGC": 3875, + "GGATGG": 3876, + "GGACAA": 3877, + "GGACAT": 3878, + "GGACAC": 3879, + "GGACAG": 3880, + "GGACTA": 3881, + "GGACTT": 3882, + "GGACTC": 3883, + "GGACTG": 3884, + "GGACCA": 3885, + "GGACCT": 3886, + "GGACCC": 3887, + "GGACCG": 3888, + "GGACGA": 3889, + "GGACGT": 3890, + "GGACGC": 3891, + "GGACGG": 3892, + "GGAGAA": 3893, + "GGAGAT": 3894, + "GGAGAC": 3895, + "GGAGAG": 3896, + "GGAGTA": 3897, + "GGAGTT": 3898, + "GGAGTC": 3899, + "GGAGTG": 3900, + "GGAGCA": 3901, + "GGAGCT": 3902, + "GGAGCC": 3903, + "GGAGCG": 3904, + "GGAGGA": 3905, + "GGAGGT": 3906, + "GGAGGC": 3907, + "GGAGGG": 3908, + "GGTAAA": 3909, + "GGTAAT": 3910, + "GGTAAC": 3911, + "GGTAAG": 3912, + "GGTATA": 3913, + "GGTATT": 3914, + "GGTATC": 3915, + "GGTATG": 3916, + "GGTACA": 3917, + "GGTACT": 3918, + "GGTACC": 3919, + "GGTACG": 3920, + "GGTAGA": 3921, + "GGTAGT": 3922, + "GGTAGC": 3923, + "GGTAGG": 3924, + "GGTTAA": 3925, + "GGTTAT": 3926, + "GGTTAC": 3927, + "GGTTAG": 3928, + "GGTTTA": 3929, + "GGTTTT": 3930, + "GGTTTC": 3931, + "GGTTTG": 3932, + "GGTTCA": 3933, + "GGTTCT": 3934, + "GGTTCC": 3935, + "GGTTCG": 3936, + "GGTTGA": 3937, + "GGTTGT": 3938, + "GGTTGC": 3939, + "GGTTGG": 3940, + "GGTCAA": 3941, + "GGTCAT": 3942, + "GGTCAC": 3943, + "GGTCAG": 3944, + "GGTCTA": 3945, + "GGTCTT": 3946, + "GGTCTC": 3947, + "GGTCTG": 3948, + "GGTCCA": 3949, + "GGTCCT": 3950, + "GGTCCC": 3951, + "GGTCCG": 3952, + "GGTCGA": 3953, + "GGTCGT": 3954, + "GGTCGC": 3955, + "GGTCGG": 3956, + "GGTGAA": 3957, + "GGTGAT": 3958, + "GGTGAC": 3959, + "GGTGAG": 3960, + "GGTGTA": 3961, + "GGTGTT": 3962, + "GGTGTC": 3963, + "GGTGTG": 3964, + "GGTGCA": 3965, + "GGTGCT": 3966, + "GGTGCC": 3967, + "GGTGCG": 3968, + "GGTGGA": 3969, + "GGTGGT": 3970, + "GGTGGC": 3971, + "GGTGGG": 3972, + "GGCAAA": 3973, + "GGCAAT": 3974, + "GGCAAC": 3975, + "GGCAAG": 3976, + "GGCATA": 3977, + "GGCATT": 3978, + "GGCATC": 3979, + "GGCATG": 3980, + "GGCACA": 3981, + "GGCACT": 3982, + "GGCACC": 3983, + "GGCACG": 3984, + "GGCAGA": 3985, + "GGCAGT": 3986, + "GGCAGC": 3987, + "GGCAGG": 3988, + "GGCTAA": 3989, + "GGCTAT": 3990, + "GGCTAC": 3991, + "GGCTAG": 3992, + "GGCTTA": 3993, + "GGCTTT": 3994, + "GGCTTC": 3995, + "GGCTTG": 3996, + "GGCTCA": 3997, + "GGCTCT": 3998, + "GGCTCC": 3999, + "GGCTCG": 4000, + "GGCTGA": 4001, + "GGCTGT": 4002, + "GGCTGC": 4003, + "GGCTGG": 4004, + "GGCCAA": 4005, + "GGCCAT": 4006, + "GGCCAC": 4007, + "GGCCAG": 4008, + "GGCCTA": 4009, + "GGCCTT": 4010, + "GGCCTC": 4011, + "GGCCTG": 4012, + "GGCCCA": 4013, + "GGCCCT": 4014, + "GGCCCC": 4015, + "GGCCCG": 4016, + "GGCCGA": 4017, + "GGCCGT": 4018, + "GGCCGC": 4019, + "GGCCGG": 4020, + "GGCGAA": 4021, + "GGCGAT": 4022, + "GGCGAC": 4023, + "GGCGAG": 4024, + "GGCGTA": 4025, + "GGCGTT": 4026, + "GGCGTC": 4027, + "GGCGTG": 4028, + "GGCGCA": 4029, + "GGCGCT": 4030, + "GGCGCC": 4031, + "GGCGCG": 4032, + "GGCGGA": 4033, + "GGCGGT": 4034, + "GGCGGC": 4035, + "GGCGGG": 4036, + "GGGAAA": 4037, + "GGGAAT": 4038, + "GGGAAC": 4039, + "GGGAAG": 4040, + "GGGATA": 4041, + "GGGATT": 4042, + "GGGATC": 4043, + "GGGATG": 4044, + "GGGACA": 4045, + "GGGACT": 4046, + "GGGACC": 4047, + "GGGACG": 4048, + "GGGAGA": 4049, + "GGGAGT": 4050, + "GGGAGC": 4051, + "GGGAGG": 4052, + "GGGTAA": 4053, + "GGGTAT": 4054, + "GGGTAC": 4055, + "GGGTAG": 4056, + "GGGTTA": 4057, + "GGGTTT": 4058, + "GGGTTC": 4059, + "GGGTTG": 4060, + "GGGTCA": 4061, + "GGGTCT": 4062, + "GGGTCC": 4063, + "GGGTCG": 4064, + "GGGTGA": 4065, + "GGGTGT": 4066, + "GGGTGC": 4067, + "GGGTGG": 4068, + "GGGCAA": 4069, + "GGGCAT": 4070, + "GGGCAC": 4071, + "GGGCAG": 4072, + "GGGCTA": 4073, + "GGGCTT": 4074, + "GGGCTC": 4075, + "GGGCTG": 4076, + "GGGCCA": 4077, + "GGGCCT": 4078, + "GGGCCC": 4079, + "GGGCCG": 4080, + "GGGCGA": 4081, + "GGGCGT": 4082, + "GGGCGC": 4083, + "GGGCGG": 4084, + "GGGGAA": 4085, + "GGGGAT": 4086, + "GGGGAC": 4087, + "GGGGAG": 4088, + "GGGGTA": 4089, + "GGGGTT": 4090, + "GGGGTC": 4091, + "GGGGTG": 4092, + "GGGGCA": 4093, + "GGGGCT": 4094, + "GGGGCC": 4095, + "GGGGCG": 4096, + "GGGGGA": 4097, + "GGGGGT": 4098, + "GGGGGC": 4099, + "GGGGGG": 4100 + } + } +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6250/tokenizer_config.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6250/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..604e2b1b29af1dcb3920d2828b6a90276e5e3d02 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6250/tokenizer_config.json @@ -0,0 +1,58 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "[UNK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "[CLS]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "[SEP]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "4": { + "content": "[MASK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "clean_up_tokenization_spaces": true, + "cls_token": "[CLS]", + "do_basic_tokenize": true, + "do_lower_case": false, + "mask_token": "[MASK]", + "max_len": 512, + "model_max_length": 512, + "never_split": null, + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "strip_accents": null, + "tokenize_chinese_chars": true, + "tokenizer_class": "BertTokenizer", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6250/trainer_state.json b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6250/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..80f0506b2f49ca22a98bb8e4aeb38402bb1c9383 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6250/trainer_state.json @@ -0,0 +1,1783 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 25.0, + "eval_steps": 500, + "global_step": 6250, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.1, + "grad_norm": 1.1046091318130493, + "learning_rate": 5.319148936170213e-05, + "loss": 0.6562, + "step": 25 + }, + { + "epoch": 0.2, + "grad_norm": 0.4066573977470398, + "learning_rate": 0.00010638297872340425, + "loss": 0.5373, + "step": 50 + }, + { + "epoch": 0.3, + "grad_norm": 0.7340325713157654, + "learning_rate": 0.00015957446808510637, + "loss": 0.4995, + "step": 75 + }, + { + "epoch": 0.4, + "grad_norm": 2.134040117263794, + "learning_rate": 0.0002127659574468085, + "loss": 0.4871, + "step": 100 + }, + { + "epoch": 0.5, + "grad_norm": 2.911975860595703, + "learning_rate": 0.00026595744680851064, + "loss": 0.4507, + "step": 125 + }, + { + "epoch": 0.6, + "grad_norm": 2.040365219116211, + "learning_rate": 0.00031914893617021275, + "loss": 0.3944, + "step": 150 + }, + { + "epoch": 0.7, + "grad_norm": 5.643661022186279, + "learning_rate": 0.0003723404255319149, + "loss": 0.3863, + "step": 175 + }, + { + "epoch": 0.8, + "grad_norm": 3.882492780685425, + "learning_rate": 0.0003992081821181128, + "loss": 0.3875, + "step": 200 + }, + { + "epoch": 0.9, + "grad_norm": 5.87437105178833, + "learning_rate": 0.0003975585615308479, + "loss": 0.3856, + "step": 225 + }, + { + "epoch": 1.0, + "grad_norm": 4.611206531524658, + "learning_rate": 0.00039590894094358297, + "loss": 0.3866, + "step": 250 + }, + { + "epoch": 1.1, + "grad_norm": 3.6404929161071777, + "learning_rate": 0.0003942593203563181, + "loss": 0.3437, + "step": 275 + }, + { + "epoch": 1.2, + "grad_norm": 6.469444751739502, + "learning_rate": 0.0003926096997690532, + "loss": 0.3192, + "step": 300 + }, + { + "epoch": 1.3, + "grad_norm": 3.6251447200775146, + "learning_rate": 0.00039096007918178817, + "loss": 0.349, + "step": 325 + }, + { + "epoch": 1.4, + "grad_norm": 2.854794979095459, + "learning_rate": 0.0003893104585945233, + "loss": 0.3274, + "step": 350 + }, + { + "epoch": 1.5, + "grad_norm": 4.805097579956055, + "learning_rate": 0.0003876608380072583, + "loss": 0.3279, + "step": 375 + }, + { + "epoch": 1.6, + "grad_norm": 2.9891581535339355, + "learning_rate": 0.00038601121741999343, + "loss": 0.3166, + "step": 400 + }, + { + "epoch": 1.7, + "grad_norm": 5.333752155303955, + "learning_rate": 0.0003843615968327285, + "loss": 0.2856, + "step": 425 + }, + { + "epoch": 1.8, + "grad_norm": 2.615621566772461, + "learning_rate": 0.0003827119762454636, + "loss": 0.3286, + "step": 450 + }, + { + "epoch": 1.9, + "grad_norm": 2.4228415489196777, + "learning_rate": 0.00038106235565819863, + "loss": 0.2812, + "step": 475 + }, + { + "epoch": 2.0, + "grad_norm": 3.6015517711639404, + "learning_rate": 0.0003794127350709337, + "loss": 0.2884, + "step": 500 + }, + { + "epoch": 2.1, + "grad_norm": 2.7867860794067383, + "learning_rate": 0.0003777631144836688, + "loss": 0.2582, + "step": 525 + }, + { + "epoch": 2.2, + "grad_norm": 10.079645156860352, + "learning_rate": 0.00037611349389640383, + "loss": 0.245, + "step": 550 + }, + { + "epoch": 2.3, + "grad_norm": 2.475917100906372, + "learning_rate": 0.00037446387330913894, + "loss": 0.2615, + "step": 575 + }, + { + "epoch": 2.4, + "grad_norm": 8.098592758178711, + "learning_rate": 0.000372814252721874, + "loss": 0.2484, + "step": 600 + }, + { + "epoch": 2.5, + "grad_norm": 2.318723201751709, + "learning_rate": 0.00037116463213460903, + "loss": 0.2056, + "step": 625 + }, + { + "epoch": 2.6, + "grad_norm": 3.616283893585205, + "learning_rate": 0.00036951501154734414, + "loss": 0.2358, + "step": 650 + }, + { + "epoch": 2.7, + "grad_norm": 6.419492721557617, + "learning_rate": 0.0003678653909600792, + "loss": 0.2204, + "step": 675 + }, + { + "epoch": 2.8, + "grad_norm": 2.9333388805389404, + "learning_rate": 0.0003662157703728143, + "loss": 0.1857, + "step": 700 + }, + { + "epoch": 2.9, + "grad_norm": 7.691112995147705, + "learning_rate": 0.00036456614978554934, + "loss": 0.2054, + "step": 725 + }, + { + "epoch": 3.0, + "grad_norm": 4.908492565155029, + "learning_rate": 0.00036291652919828444, + "loss": 0.1956, + "step": 750 + }, + { + "epoch": 3.1, + "grad_norm": 1.8923287391662598, + "learning_rate": 0.00036126690861101944, + "loss": 0.1708, + "step": 775 + }, + { + "epoch": 3.2, + "grad_norm": 4.403504848480225, + "learning_rate": 0.00035961728802375454, + "loss": 0.1441, + "step": 800 + }, + { + "epoch": 3.3, + "grad_norm": 4.117386817932129, + "learning_rate": 0.0003579676674364896, + "loss": 0.1586, + "step": 825 + }, + { + "epoch": 3.4, + "grad_norm": 7.119816303253174, + "learning_rate": 0.0003563180468492247, + "loss": 0.1472, + "step": 850 + }, + { + "epoch": 3.5, + "grad_norm": 3.5796430110931396, + "learning_rate": 0.0003546684262619598, + "loss": 0.1802, + "step": 875 + }, + { + "epoch": 3.6, + "grad_norm": 2.97688889503479, + "learning_rate": 0.00035301880567469485, + "loss": 0.1578, + "step": 900 + }, + { + "epoch": 3.7, + "grad_norm": 3.716148614883423, + "learning_rate": 0.0003513691850874299, + "loss": 0.1662, + "step": 925 + }, + { + "epoch": 3.8, + "grad_norm": 6.1249566078186035, + "learning_rate": 0.00034971956450016495, + "loss": 0.159, + "step": 950 + }, + { + "epoch": 3.9, + "grad_norm": 3.0592427253723145, + "learning_rate": 0.00034806994391290005, + "loss": 0.148, + "step": 975 + }, + { + "epoch": 4.0, + "grad_norm": 4.467265605926514, + "learning_rate": 0.0003464203233256351, + "loss": 0.1442, + "step": 1000 + }, + { + "epoch": 4.1, + "grad_norm": 2.7223546504974365, + "learning_rate": 0.0003447707027383702, + "loss": 0.1339, + "step": 1025 + }, + { + "epoch": 4.2, + "grad_norm": 3.698854923248291, + "learning_rate": 0.00034312108215110525, + "loss": 0.1381, + "step": 1050 + }, + { + "epoch": 4.3, + "grad_norm": 7.418384552001953, + "learning_rate": 0.0003414714615638403, + "loss": 0.1158, + "step": 1075 + }, + { + "epoch": 4.4, + "grad_norm": 1.2887814044952393, + "learning_rate": 0.0003398218409765754, + "loss": 0.0855, + "step": 1100 + }, + { + "epoch": 4.5, + "grad_norm": 0.838731586933136, + "learning_rate": 0.00033817222038931045, + "loss": 0.1272, + "step": 1125 + }, + { + "epoch": 4.6, + "grad_norm": 5.912592887878418, + "learning_rate": 0.00033652259980204556, + "loss": 0.1156, + "step": 1150 + }, + { + "epoch": 4.7, + "grad_norm": 8.49521255493164, + "learning_rate": 0.0003348729792147806, + "loss": 0.1182, + "step": 1175 + }, + { + "epoch": 4.8, + "grad_norm": 10.278315544128418, + "learning_rate": 0.0003332233586275157, + "loss": 0.1071, + "step": 1200 + }, + { + "epoch": 4.9, + "grad_norm": 0.724703311920166, + "learning_rate": 0.00033157373804025076, + "loss": 0.11, + "step": 1225 + }, + { + "epoch": 5.0, + "grad_norm": 1.6808199882507324, + "learning_rate": 0.0003299241174529858, + "loss": 0.1101, + "step": 1250 + }, + { + "epoch": 5.1, + "grad_norm": 0.1038585901260376, + "learning_rate": 0.0003282744968657209, + "loss": 0.0791, + "step": 1275 + }, + { + "epoch": 5.2, + "grad_norm": 0.15866787731647491, + "learning_rate": 0.00032662487627845596, + "loss": 0.0638, + "step": 1300 + }, + { + "epoch": 5.3, + "grad_norm": 6.748287200927734, + "learning_rate": 0.00032497525569119106, + "loss": 0.0961, + "step": 1325 + }, + { + "epoch": 5.4, + "grad_norm": 1.5960336923599243, + "learning_rate": 0.0003233256351039261, + "loss": 0.1057, + "step": 1350 + }, + { + "epoch": 5.5, + "grad_norm": 3.3678526878356934, + "learning_rate": 0.00032167601451666116, + "loss": 0.1092, + "step": 1375 + }, + { + "epoch": 5.6, + "grad_norm": 4.743562698364258, + "learning_rate": 0.00032002639392939627, + "loss": 0.1054, + "step": 1400 + }, + { + "epoch": 5.7, + "grad_norm": 12.860156059265137, + "learning_rate": 0.0003183767733421313, + "loss": 0.1009, + "step": 1425 + }, + { + "epoch": 5.8, + "grad_norm": 0.8820792436599731, + "learning_rate": 0.0003167271527548664, + "loss": 0.1219, + "step": 1450 + }, + { + "epoch": 5.9, + "grad_norm": 2.732572317123413, + "learning_rate": 0.00031507753216760147, + "loss": 0.0682, + "step": 1475 + }, + { + "epoch": 6.0, + "grad_norm": 9.810711860656738, + "learning_rate": 0.00031342791158033657, + "loss": 0.1134, + "step": 1500 + }, + { + "epoch": 6.1, + "grad_norm": 6.699489593505859, + "learning_rate": 0.00031177829099307157, + "loss": 0.0557, + "step": 1525 + }, + { + "epoch": 6.2, + "grad_norm": 14.691123008728027, + "learning_rate": 0.00031012867040580667, + "loss": 0.0599, + "step": 1550 + }, + { + "epoch": 6.3, + "grad_norm": 24.44135093688965, + "learning_rate": 0.0003084790498185417, + "loss": 0.0818, + "step": 1575 + }, + { + "epoch": 6.4, + "grad_norm": 0.033995576202869415, + "learning_rate": 0.0003068294292312768, + "loss": 0.0692, + "step": 1600 + }, + { + "epoch": 6.5, + "grad_norm": 6.722524642944336, + "learning_rate": 0.0003051798086440119, + "loss": 0.0702, + "step": 1625 + }, + { + "epoch": 6.6, + "grad_norm": 2.9519436359405518, + "learning_rate": 0.000303530188056747, + "loss": 0.073, + "step": 1650 + }, + { + "epoch": 6.7, + "grad_norm": 1.5118328332901, + "learning_rate": 0.000301880567469482, + "loss": 0.0926, + "step": 1675 + }, + { + "epoch": 6.8, + "grad_norm": 2.1923201084136963, + "learning_rate": 0.0003002309468822171, + "loss": 0.0821, + "step": 1700 + }, + { + "epoch": 6.9, + "grad_norm": 3.34309458732605, + "learning_rate": 0.0002985813262949522, + "loss": 0.0827, + "step": 1725 + }, + { + "epoch": 7.0, + "grad_norm": 10.289947509765625, + "learning_rate": 0.00029693170570768723, + "loss": 0.1162, + "step": 1750 + }, + { + "epoch": 7.1, + "grad_norm": 0.051936667412519455, + "learning_rate": 0.00029528208512042233, + "loss": 0.0626, + "step": 1775 + }, + { + "epoch": 7.2, + "grad_norm": 1.8913339376449585, + "learning_rate": 0.0002936324645331574, + "loss": 0.0657, + "step": 1800 + }, + { + "epoch": 7.3, + "grad_norm": 15.073695182800293, + "learning_rate": 0.00029198284394589243, + "loss": 0.0539, + "step": 1825 + }, + { + "epoch": 7.4, + "grad_norm": 4.514949798583984, + "learning_rate": 0.00029033322335862753, + "loss": 0.056, + "step": 1850 + }, + { + "epoch": 7.5, + "grad_norm": 0.46094000339508057, + "learning_rate": 0.0002886836027713626, + "loss": 0.0667, + "step": 1875 + }, + { + "epoch": 7.6, + "grad_norm": 4.682873725891113, + "learning_rate": 0.0002870339821840977, + "loss": 0.0673, + "step": 1900 + }, + { + "epoch": 7.7, + "grad_norm": 3.581599235534668, + "learning_rate": 0.00028538436159683273, + "loss": 0.0699, + "step": 1925 + }, + { + "epoch": 7.8, + "grad_norm": 3.2518703937530518, + "learning_rate": 0.00028373474100956784, + "loss": 0.0628, + "step": 1950 + }, + { + "epoch": 7.9, + "grad_norm": 0.24071219563484192, + "learning_rate": 0.0002820851204223029, + "loss": 0.0747, + "step": 1975 + }, + { + "epoch": 8.0, + "grad_norm": 2.8545596599578857, + "learning_rate": 0.00028043549983503794, + "loss": 0.0688, + "step": 2000 + }, + { + "epoch": 8.1, + "grad_norm": 0.049499694257974625, + "learning_rate": 0.00027878587924777304, + "loss": 0.0563, + "step": 2025 + }, + { + "epoch": 8.2, + "grad_norm": 22.112478256225586, + "learning_rate": 0.0002771362586605081, + "loss": 0.0545, + "step": 2050 + }, + { + "epoch": 8.3, + "grad_norm": 20.87473487854004, + "learning_rate": 0.0002754866380732432, + "loss": 0.0256, + "step": 2075 + }, + { + "epoch": 8.4, + "grad_norm": 2.317768096923828, + "learning_rate": 0.00027383701748597824, + "loss": 0.0753, + "step": 2100 + }, + { + "epoch": 8.5, + "grad_norm": 20.241336822509766, + "learning_rate": 0.0002721873968987133, + "loss": 0.0257, + "step": 2125 + }, + { + "epoch": 8.6, + "grad_norm": 14.994575500488281, + "learning_rate": 0.00027053777631144834, + "loss": 0.0748, + "step": 2150 + }, + { + "epoch": 8.7, + "grad_norm": 0.8268639445304871, + "learning_rate": 0.00026888815572418344, + "loss": 0.0805, + "step": 2175 + }, + { + "epoch": 8.8, + "grad_norm": 0.031071001663804054, + "learning_rate": 0.00026723853513691855, + "loss": 0.0603, + "step": 2200 + }, + { + "epoch": 8.9, + "grad_norm": 2.4572932720184326, + "learning_rate": 0.0002655889145496536, + "loss": 0.071, + "step": 2225 + }, + { + "epoch": 9.0, + "grad_norm": 0.21708321571350098, + "learning_rate": 0.0002639392939623887, + "loss": 0.0606, + "step": 2250 + }, + { + "epoch": 9.1, + "grad_norm": 14.928559303283691, + "learning_rate": 0.0002622896733751237, + "loss": 0.0443, + "step": 2275 + }, + { + "epoch": 9.2, + "grad_norm": 1.108826994895935, + "learning_rate": 0.0002606400527878588, + "loss": 0.0509, + "step": 2300 + }, + { + "epoch": 9.3, + "grad_norm": 0.05721910670399666, + "learning_rate": 0.00025899043220059385, + "loss": 0.0459, + "step": 2325 + }, + { + "epoch": 9.4, + "grad_norm": 7.81903600692749, + "learning_rate": 0.00025734081161332895, + "loss": 0.0535, + "step": 2350 + }, + { + "epoch": 9.5, + "grad_norm": 0.01475981343537569, + "learning_rate": 0.000255691191026064, + "loss": 0.0399, + "step": 2375 + }, + { + "epoch": 9.6, + "grad_norm": 5.496493339538574, + "learning_rate": 0.0002540415704387991, + "loss": 0.0744, + "step": 2400 + }, + { + "epoch": 9.7, + "grad_norm": 0.05929339677095413, + "learning_rate": 0.00025239194985153415, + "loss": 0.0416, + "step": 2425 + }, + { + "epoch": 9.8, + "grad_norm": 0.06961135566234589, + "learning_rate": 0.0002507423292642692, + "loss": 0.04, + "step": 2450 + }, + { + "epoch": 9.9, + "grad_norm": 0.02165246568620205, + "learning_rate": 0.0002490927086770043, + "loss": 0.0984, + "step": 2475 + }, + { + "epoch": 10.0, + "grad_norm": 25.155675888061523, + "learning_rate": 0.00024744308808973936, + "loss": 0.042, + "step": 2500 + }, + { + "epoch": 10.1, + "grad_norm": 8.566617965698242, + "learning_rate": 0.00024579346750247446, + "loss": 0.0385, + "step": 2525 + }, + { + "epoch": 10.2, + "grad_norm": 0.024992404505610466, + "learning_rate": 0.0002441438469152095, + "loss": 0.0327, + "step": 2550 + }, + { + "epoch": 10.3, + "grad_norm": 0.22233247756958008, + "learning_rate": 0.00024249422632794456, + "loss": 0.0211, + "step": 2575 + }, + { + "epoch": 10.4, + "grad_norm": 13.307366371154785, + "learning_rate": 0.00024084460574067963, + "loss": 0.0472, + "step": 2600 + }, + { + "epoch": 10.5, + "grad_norm": 0.17787696421146393, + "learning_rate": 0.0002391949851534147, + "loss": 0.0346, + "step": 2625 + }, + { + "epoch": 10.6, + "grad_norm": 1.2845549583435059, + "learning_rate": 0.0002375453645661498, + "loss": 0.0587, + "step": 2650 + }, + { + "epoch": 10.7, + "grad_norm": 0.0603482760488987, + "learning_rate": 0.00023589574397888486, + "loss": 0.0394, + "step": 2675 + }, + { + "epoch": 10.8, + "grad_norm": 0.031819943338632584, + "learning_rate": 0.00023424612339161997, + "loss": 0.0434, + "step": 2700 + }, + { + "epoch": 10.9, + "grad_norm": 1.5738537311553955, + "learning_rate": 0.000232596502804355, + "loss": 0.0393, + "step": 2725 + }, + { + "epoch": 11.0, + "grad_norm": 0.6862583756446838, + "learning_rate": 0.00023094688221709007, + "loss": 0.0183, + "step": 2750 + }, + { + "epoch": 11.1, + "grad_norm": 0.0053489054553210735, + "learning_rate": 0.00022929726162982514, + "loss": 0.0277, + "step": 2775 + }, + { + "epoch": 11.2, + "grad_norm": 0.04365606606006622, + "learning_rate": 0.00022764764104256022, + "loss": 0.0337, + "step": 2800 + }, + { + "epoch": 11.3, + "grad_norm": 1.3348641395568848, + "learning_rate": 0.0002259980204552953, + "loss": 0.0219, + "step": 2825 + }, + { + "epoch": 11.4, + "grad_norm": 3.3695547580718994, + "learning_rate": 0.00022434839986803037, + "loss": 0.0227, + "step": 2850 + }, + { + "epoch": 11.5, + "grad_norm": 2.507455587387085, + "learning_rate": 0.00022269877928076542, + "loss": 0.0443, + "step": 2875 + }, + { + "epoch": 11.6, + "grad_norm": 0.013451912440359592, + "learning_rate": 0.0002210491586935005, + "loss": 0.0486, + "step": 2900 + }, + { + "epoch": 11.7, + "grad_norm": 0.031177405267953873, + "learning_rate": 0.00021939953810623557, + "loss": 0.0511, + "step": 2925 + }, + { + "epoch": 11.8, + "grad_norm": 0.02148960903286934, + "learning_rate": 0.00021774991751897065, + "loss": 0.0269, + "step": 2950 + }, + { + "epoch": 11.9, + "grad_norm": 0.01219563465565443, + "learning_rate": 0.00021610029693170573, + "loss": 0.042, + "step": 2975 + }, + { + "epoch": 12.0, + "grad_norm": 0.2062515765428543, + "learning_rate": 0.0002144506763444408, + "loss": 0.031, + "step": 3000 + }, + { + "epoch": 12.1, + "grad_norm": 3.235403060913086, + "learning_rate": 0.00021280105575717582, + "loss": 0.0414, + "step": 3025 + }, + { + "epoch": 12.2, + "grad_norm": 1.089821696281433, + "learning_rate": 0.00021115143516991093, + "loss": 0.0692, + "step": 3050 + }, + { + "epoch": 12.3, + "grad_norm": 0.013664484024047852, + "learning_rate": 0.000209501814582646, + "loss": 0.0367, + "step": 3075 + }, + { + "epoch": 12.4, + "grad_norm": 1.3101387023925781, + "learning_rate": 0.00020785219399538108, + "loss": 0.0277, + "step": 3100 + }, + { + "epoch": 12.5, + "grad_norm": 3.244853973388672, + "learning_rate": 0.00020620257340811616, + "loss": 0.0373, + "step": 3125 + }, + { + "epoch": 12.6, + "grad_norm": 0.013548053801059723, + "learning_rate": 0.00020455295282085123, + "loss": 0.027, + "step": 3150 + }, + { + "epoch": 12.7, + "grad_norm": 0.004670978989452124, + "learning_rate": 0.00020290333223358626, + "loss": 0.0252, + "step": 3175 + }, + { + "epoch": 12.8, + "grad_norm": 0.007080752402544022, + "learning_rate": 0.00020125371164632133, + "loss": 0.0187, + "step": 3200 + }, + { + "epoch": 12.9, + "grad_norm": 0.005918944254517555, + "learning_rate": 0.0001996040910590564, + "loss": 0.0423, + "step": 3225 + }, + { + "epoch": 13.0, + "grad_norm": 4.530269145965576, + "learning_rate": 0.00019795447047179148, + "loss": 0.0381, + "step": 3250 + }, + { + "epoch": 13.1, + "grad_norm": 0.00533739197999239, + "learning_rate": 0.0001963048498845266, + "loss": 0.0097, + "step": 3275 + }, + { + "epoch": 13.2, + "grad_norm": 0.7641226649284363, + "learning_rate": 0.00019465522929726164, + "loss": 0.0205, + "step": 3300 + }, + { + "epoch": 13.3, + "grad_norm": 2.16174054145813, + "learning_rate": 0.00019300560870999671, + "loss": 0.0174, + "step": 3325 + }, + { + "epoch": 13.4, + "grad_norm": 0.00814723875373602, + "learning_rate": 0.0001913559881227318, + "loss": 0.0215, + "step": 3350 + }, + { + "epoch": 13.5, + "grad_norm": 10.500170707702637, + "learning_rate": 0.00018970636753546684, + "loss": 0.0226, + "step": 3375 + }, + { + "epoch": 13.6, + "grad_norm": 2.612359046936035, + "learning_rate": 0.00018805674694820192, + "loss": 0.0314, + "step": 3400 + }, + { + "epoch": 13.7, + "grad_norm": 0.11265736818313599, + "learning_rate": 0.000186407126360937, + "loss": 0.0234, + "step": 3425 + }, + { + "epoch": 13.8, + "grad_norm": 0.06453213095664978, + "learning_rate": 0.00018475750577367207, + "loss": 0.0314, + "step": 3450 + }, + { + "epoch": 13.9, + "grad_norm": 0.01650763861835003, + "learning_rate": 0.00018310788518640715, + "loss": 0.0403, + "step": 3475 + }, + { + "epoch": 14.0, + "grad_norm": 0.002764373552054167, + "learning_rate": 0.00018145826459914222, + "loss": 0.0137, + "step": 3500 + }, + { + "epoch": 14.1, + "grad_norm": 0.0022430066019296646, + "learning_rate": 0.00017980864401187727, + "loss": 0.0084, + "step": 3525 + }, + { + "epoch": 14.2, + "grad_norm": 0.007267679553478956, + "learning_rate": 0.00017815902342461235, + "loss": 0.0161, + "step": 3550 + }, + { + "epoch": 14.3, + "grad_norm": 0.004844362381845713, + "learning_rate": 0.00017650940283734742, + "loss": 0.0399, + "step": 3575 + }, + { + "epoch": 14.4, + "grad_norm": 0.05529040843248367, + "learning_rate": 0.00017485978225008247, + "loss": 0.0192, + "step": 3600 + }, + { + "epoch": 14.5, + "grad_norm": 0.008248478174209595, + "learning_rate": 0.00017321016166281755, + "loss": 0.0079, + "step": 3625 + }, + { + "epoch": 14.6, + "grad_norm": 0.02617962658405304, + "learning_rate": 0.00017156054107555263, + "loss": 0.0244, + "step": 3650 + }, + { + "epoch": 14.7, + "grad_norm": 0.004497932270169258, + "learning_rate": 0.0001699109204882877, + "loss": 0.0464, + "step": 3675 + }, + { + "epoch": 14.8, + "grad_norm": 0.008586671203374863, + "learning_rate": 0.00016826129990102278, + "loss": 0.0354, + "step": 3700 + }, + { + "epoch": 14.9, + "grad_norm": 0.30177807807922363, + "learning_rate": 0.00016661167931375785, + "loss": 0.0147, + "step": 3725 + }, + { + "epoch": 15.0, + "grad_norm": 0.10390808433294296, + "learning_rate": 0.0001649620587264929, + "loss": 0.0078, + "step": 3750 + }, + { + "epoch": 15.1, + "grad_norm": 1.584868311882019, + "learning_rate": 0.00016331243813922798, + "loss": 0.0059, + "step": 3775 + }, + { + "epoch": 15.2, + "grad_norm": 0.0038078685756772757, + "learning_rate": 0.00016166281755196306, + "loss": 0.0143, + "step": 3800 + }, + { + "epoch": 15.3, + "grad_norm": 0.017795555293560028, + "learning_rate": 0.00016001319696469813, + "loss": 0.0003, + "step": 3825 + }, + { + "epoch": 15.4, + "grad_norm": 7.246474742889404, + "learning_rate": 0.0001583635763774332, + "loss": 0.0288, + "step": 3850 + }, + { + "epoch": 15.5, + "grad_norm": 0.0014716258738189936, + "learning_rate": 0.00015671395579016829, + "loss": 0.0113, + "step": 3875 + }, + { + "epoch": 15.6, + "grad_norm": 0.002159764291718602, + "learning_rate": 0.00015506433520290334, + "loss": 0.0138, + "step": 3900 + }, + { + "epoch": 15.7, + "grad_norm": 27.16730308532715, + "learning_rate": 0.0001534147146156384, + "loss": 0.0048, + "step": 3925 + }, + { + "epoch": 15.8, + "grad_norm": 0.004570333287119865, + "learning_rate": 0.0001517650940283735, + "loss": 0.0144, + "step": 3950 + }, + { + "epoch": 15.9, + "grad_norm": 0.008330491371452808, + "learning_rate": 0.00015011547344110854, + "loss": 0.0371, + "step": 3975 + }, + { + "epoch": 16.0, + "grad_norm": 15.75979232788086, + "learning_rate": 0.00014846585285384361, + "loss": 0.0207, + "step": 4000 + }, + { + "epoch": 16.1, + "grad_norm": 14.096012115478516, + "learning_rate": 0.0001468162322665787, + "loss": 0.018, + "step": 4025 + }, + { + "epoch": 16.2, + "grad_norm": 0.0015319391386583447, + "learning_rate": 0.00014516661167931377, + "loss": 0.0109, + "step": 4050 + }, + { + "epoch": 16.3, + "grad_norm": 0.003954900894314051, + "learning_rate": 0.00014351699109204884, + "loss": 0.012, + "step": 4075 + }, + { + "epoch": 16.4, + "grad_norm": 37.79055404663086, + "learning_rate": 0.00014186737050478392, + "loss": 0.0301, + "step": 4100 + }, + { + "epoch": 16.5, + "grad_norm": 0.0014516041846945882, + "learning_rate": 0.00014021774991751897, + "loss": 0.0116, + "step": 4125 + }, + { + "epoch": 16.6, + "grad_norm": 0.009134252555668354, + "learning_rate": 0.00013856812933025404, + "loss": 0.0126, + "step": 4150 + }, + { + "epoch": 16.7, + "grad_norm": 19.0325927734375, + "learning_rate": 0.00013691850874298912, + "loss": 0.0398, + "step": 4175 + }, + { + "epoch": 16.8, + "grad_norm": 0.41503193974494934, + "learning_rate": 0.00013526888815572417, + "loss": 0.001, + "step": 4200 + }, + { + "epoch": 16.9, + "grad_norm": 0.0014813834568485618, + "learning_rate": 0.00013361926756845927, + "loss": 0.0077, + "step": 4225 + }, + { + "epoch": 17.0, + "grad_norm": 0.007494773715734482, + "learning_rate": 0.00013196964698119435, + "loss": 0.0052, + "step": 4250 + }, + { + "epoch": 17.1, + "grad_norm": 0.002411301014944911, + "learning_rate": 0.0001303200263939294, + "loss": 0.0038, + "step": 4275 + }, + { + "epoch": 17.2, + "grad_norm": 0.01521711703389883, + "learning_rate": 0.00012867040580666448, + "loss": 0.0051, + "step": 4300 + }, + { + "epoch": 17.3, + "grad_norm": 0.0010520346695557237, + "learning_rate": 0.00012702078521939955, + "loss": 0.0167, + "step": 4325 + }, + { + "epoch": 17.4, + "grad_norm": 0.054758671671152115, + "learning_rate": 0.0001253711646321346, + "loss": 0.011, + "step": 4350 + }, + { + "epoch": 17.5, + "grad_norm": 0.0012187627144157887, + "learning_rate": 0.00012372154404486968, + "loss": 0.0239, + "step": 4375 + }, + { + "epoch": 17.6, + "grad_norm": 58.37099838256836, + "learning_rate": 0.00012207192345760475, + "loss": 0.0055, + "step": 4400 + }, + { + "epoch": 17.7, + "grad_norm": 0.27519506216049194, + "learning_rate": 0.00012042230287033982, + "loss": 0.0332, + "step": 4425 + }, + { + "epoch": 17.8, + "grad_norm": 0.0011721524642780423, + "learning_rate": 0.0001187726822830749, + "loss": 0.0048, + "step": 4450 + }, + { + "epoch": 17.9, + "grad_norm": 0.0015529991360381246, + "learning_rate": 0.00011712306169580998, + "loss": 0.0134, + "step": 4475 + }, + { + "epoch": 18.0, + "grad_norm": 0.0008489767787978053, + "learning_rate": 0.00011547344110854503, + "loss": 0.0148, + "step": 4500 + }, + { + "epoch": 18.1, + "grad_norm": 0.0010399922030046582, + "learning_rate": 0.00011382382052128011, + "loss": 0.0046, + "step": 4525 + }, + { + "epoch": 18.2, + "grad_norm": 0.0017199565190821886, + "learning_rate": 0.00011217419993401519, + "loss": 0.0123, + "step": 4550 + }, + { + "epoch": 18.3, + "grad_norm": 0.004095083102583885, + "learning_rate": 0.00011052457934675025, + "loss": 0.0078, + "step": 4575 + }, + { + "epoch": 18.4, + "grad_norm": 0.005569108761847019, + "learning_rate": 0.00010887495875948532, + "loss": 0.0001, + "step": 4600 + }, + { + "epoch": 18.5, + "grad_norm": 0.0013379952870309353, + "learning_rate": 0.0001072253381722204, + "loss": 0.0123, + "step": 4625 + }, + { + "epoch": 18.6, + "grad_norm": 0.13633179664611816, + "learning_rate": 0.00010557571758495546, + "loss": 0.016, + "step": 4650 + }, + { + "epoch": 18.7, + "grad_norm": 0.00383689533919096, + "learning_rate": 0.00010392609699769054, + "loss": 0.0238, + "step": 4675 + }, + { + "epoch": 18.8, + "grad_norm": 0.011090466752648354, + "learning_rate": 0.00010227647641042562, + "loss": 0.0299, + "step": 4700 + }, + { + "epoch": 18.9, + "grad_norm": 0.01791750080883503, + "learning_rate": 0.00010062685582316067, + "loss": 0.0006, + "step": 4725 + }, + { + "epoch": 19.0, + "grad_norm": 20.31785774230957, + "learning_rate": 9.897723523589574e-05, + "loss": 0.0067, + "step": 4750 + }, + { + "epoch": 19.1, + "grad_norm": 0.0011455218773335218, + "learning_rate": 9.732761464863082e-05, + "loss": 0.0138, + "step": 4775 + }, + { + "epoch": 19.2, + "grad_norm": 0.0014105259906500578, + "learning_rate": 9.56779940613659e-05, + "loss": 0.0005, + "step": 4800 + }, + { + "epoch": 19.3, + "grad_norm": 0.002618618542328477, + "learning_rate": 9.402837347410096e-05, + "loss": 0.0112, + "step": 4825 + }, + { + "epoch": 19.4, + "grad_norm": 0.001108857337385416, + "learning_rate": 9.237875288683603e-05, + "loss": 0.0164, + "step": 4850 + }, + { + "epoch": 19.5, + "grad_norm": 0.0010457661701366305, + "learning_rate": 9.072913229957111e-05, + "loss": 0.0104, + "step": 4875 + }, + { + "epoch": 19.6, + "grad_norm": 0.0012097499566152692, + "learning_rate": 8.907951171230617e-05, + "loss": 0.0039, + "step": 4900 + }, + { + "epoch": 19.7, + "grad_norm": 0.04536284878849983, + "learning_rate": 8.742989112504124e-05, + "loss": 0.0154, + "step": 4925 + }, + { + "epoch": 19.8, + "grad_norm": 29.97249984741211, + "learning_rate": 8.578027053777631e-05, + "loss": 0.0246, + "step": 4950 + }, + { + "epoch": 19.9, + "grad_norm": 0.01911325752735138, + "learning_rate": 8.413064995051139e-05, + "loss": 0.0176, + "step": 4975 + }, + { + "epoch": 20.0, + "grad_norm": 0.3129942715167999, + "learning_rate": 8.248102936324645e-05, + "loss": 0.0029, + "step": 5000 + }, + { + "epoch": 20.1, + "grad_norm": 0.04958798363804817, + "learning_rate": 8.083140877598153e-05, + "loss": 0.0139, + "step": 5025 + }, + { + "epoch": 20.2, + "grad_norm": 0.003336989786475897, + "learning_rate": 7.91817881887166e-05, + "loss": 0.0159, + "step": 5050 + }, + { + "epoch": 20.3, + "grad_norm": 0.001012949156574905, + "learning_rate": 7.753216760145167e-05, + "loss": 0.0095, + "step": 5075 + }, + { + "epoch": 20.4, + "grad_norm": 0.0019106008112430573, + "learning_rate": 7.588254701418674e-05, + "loss": 0.0113, + "step": 5100 + }, + { + "epoch": 20.5, + "grad_norm": 0.034077223390340805, + "learning_rate": 7.423292642692181e-05, + "loss": 0.0101, + "step": 5125 + }, + { + "epoch": 20.6, + "grad_norm": 0.0011715433793142438, + "learning_rate": 7.258330583965688e-05, + "loss": 0.0111, + "step": 5150 + }, + { + "epoch": 20.7, + "grad_norm": 0.06765557825565338, + "learning_rate": 7.093368525239196e-05, + "loss": 0.0066, + "step": 5175 + }, + { + "epoch": 20.8, + "grad_norm": 0.0008974962402135134, + "learning_rate": 6.928406466512702e-05, + "loss": 0.0058, + "step": 5200 + }, + { + "epoch": 20.9, + "grad_norm": 0.00472952239215374, + "learning_rate": 6.763444407786209e-05, + "loss": 0.0027, + "step": 5225 + }, + { + "epoch": 21.0, + "grad_norm": 0.0006938530132174492, + "learning_rate": 6.598482349059718e-05, + "loss": 0.0009, + "step": 5250 + }, + { + "epoch": 21.1, + "grad_norm": 0.0007072632433846593, + "learning_rate": 6.433520290333224e-05, + "loss": 0.0015, + "step": 5275 + }, + { + "epoch": 21.2, + "grad_norm": 0.016313398256897926, + "learning_rate": 6.26855823160673e-05, + "loss": 0.0084, + "step": 5300 + }, + { + "epoch": 21.3, + "grad_norm": 0.0006648830021731555, + "learning_rate": 6.103596172880238e-05, + "loss": 0.0093, + "step": 5325 + }, + { + "epoch": 21.4, + "grad_norm": 0.0005273331771604717, + "learning_rate": 5.938634114153745e-05, + "loss": 0.0096, + "step": 5350 + }, + { + "epoch": 21.5, + "grad_norm": 0.0022026444785296917, + "learning_rate": 5.7736720554272516e-05, + "loss": 0.0107, + "step": 5375 + }, + { + "epoch": 21.6, + "grad_norm": 0.0009771068580448627, + "learning_rate": 5.608709996700759e-05, + "loss": 0.0017, + "step": 5400 + }, + { + "epoch": 21.7, + "grad_norm": 0.0006267099524848163, + "learning_rate": 5.443747937974266e-05, + "loss": 0.0032, + "step": 5425 + }, + { + "epoch": 21.8, + "grad_norm": 0.0006058421568013728, + "learning_rate": 5.278785879247773e-05, + "loss": 0.004, + "step": 5450 + }, + { + "epoch": 21.9, + "grad_norm": 0.0007795215351507068, + "learning_rate": 5.113823820521281e-05, + "loss": 0.0004, + "step": 5475 + }, + { + "epoch": 22.0, + "grad_norm": 0.013498655520379543, + "learning_rate": 4.948861761794787e-05, + "loss": 0.0035, + "step": 5500 + }, + { + "epoch": 22.1, + "grad_norm": 0.0007379506714642048, + "learning_rate": 4.783899703068295e-05, + "loss": 0.0085, + "step": 5525 + }, + { + "epoch": 22.2, + "grad_norm": 0.0010062415385618806, + "learning_rate": 4.618937644341802e-05, + "loss": 0.0001, + "step": 5550 + }, + { + "epoch": 22.3, + "grad_norm": 0.002069108886644244, + "learning_rate": 4.453975585615309e-05, + "loss": 0.0153, + "step": 5575 + }, + { + "epoch": 22.4, + "grad_norm": 0.0005407587741501629, + "learning_rate": 4.2890135268888156e-05, + "loss": 0.0088, + "step": 5600 + }, + { + "epoch": 22.5, + "grad_norm": 0.010531233623623848, + "learning_rate": 4.1240514681623226e-05, + "loss": 0.0002, + "step": 5625 + }, + { + "epoch": 22.6, + "grad_norm": 0.004741353914141655, + "learning_rate": 3.95908940943583e-05, + "loss": 0.0078, + "step": 5650 + }, + { + "epoch": 22.7, + "grad_norm": 0.0008237269939854741, + "learning_rate": 3.794127350709337e-05, + "loss": 0.0111, + "step": 5675 + }, + { + "epoch": 22.8, + "grad_norm": 0.0007508064154535532, + "learning_rate": 3.629165291982844e-05, + "loss": 0.0052, + "step": 5700 + }, + { + "epoch": 22.9, + "grad_norm": 0.0015650958521291614, + "learning_rate": 3.464203233256351e-05, + "loss": 0.0088, + "step": 5725 + }, + { + "epoch": 23.0, + "grad_norm": 0.0010173724731430411, + "learning_rate": 3.299241174529859e-05, + "loss": 0.0124, + "step": 5750 + }, + { + "epoch": 23.1, + "grad_norm": 43.15919494628906, + "learning_rate": 3.134279115803365e-05, + "loss": 0.004, + "step": 5775 + }, + { + "epoch": 23.2, + "grad_norm": 0.0005111348582431674, + "learning_rate": 2.9693170570768723e-05, + "loss": 0.0008, + "step": 5800 + }, + { + "epoch": 23.3, + "grad_norm": 0.009832415729761124, + "learning_rate": 2.8043549983503796e-05, + "loss": 0.0022, + "step": 5825 + }, + { + "epoch": 23.4, + "grad_norm": 0.0006390413036569953, + "learning_rate": 2.6393929396238866e-05, + "loss": 0.0001, + "step": 5850 + }, + { + "epoch": 23.5, + "grad_norm": 0.00047043984523043036, + "learning_rate": 2.4744308808973936e-05, + "loss": 0.0048, + "step": 5875 + }, + { + "epoch": 23.6, + "grad_norm": 0.0007061742944642901, + "learning_rate": 2.309468822170901e-05, + "loss": 0.0088, + "step": 5900 + }, + { + "epoch": 23.7, + "grad_norm": 0.0004057694459334016, + "learning_rate": 2.1445067634444078e-05, + "loss": 0.0, + "step": 5925 + }, + { + "epoch": 23.8, + "grad_norm": 0.0005023757694289088, + "learning_rate": 1.979544704717915e-05, + "loss": 0.001, + "step": 5950 + }, + { + "epoch": 23.9, + "grad_norm": 0.001338923815637827, + "learning_rate": 1.814582645991422e-05, + "loss": 0.0187, + "step": 5975 + }, + { + "epoch": 24.0, + "grad_norm": 0.018048524856567383, + "learning_rate": 1.6496205872649294e-05, + "loss": 0.0001, + "step": 6000 + }, + { + "epoch": 24.1, + "grad_norm": 0.0005618903669528663, + "learning_rate": 1.4846585285384362e-05, + "loss": 0.013, + "step": 6025 + }, + { + "epoch": 24.2, + "grad_norm": 0.0005956863169558346, + "learning_rate": 1.3196964698119433e-05, + "loss": 0.0139, + "step": 6050 + }, + { + "epoch": 24.3, + "grad_norm": 0.000518847955390811, + "learning_rate": 1.1547344110854504e-05, + "loss": 0.0, + "step": 6075 + }, + { + "epoch": 24.4, + "grad_norm": 0.006304449401795864, + "learning_rate": 9.897723523589576e-06, + "loss": 0.0002, + "step": 6100 + }, + { + "epoch": 24.5, + "grad_norm": 0.00045549788046628237, + "learning_rate": 8.248102936324647e-06, + "loss": 0.011, + "step": 6125 + }, + { + "epoch": 24.6, + "grad_norm": 0.005512393545359373, + "learning_rate": 6.5984823490597165e-06, + "loss": 0.0001, + "step": 6150 + }, + { + "epoch": 24.7, + "grad_norm": 0.0026837822515517473, + "learning_rate": 4.948861761794788e-06, + "loss": 0.0105, + "step": 6175 + }, + { + "epoch": 24.8, + "grad_norm": 0.0005764334346167743, + "learning_rate": 3.2992411745298582e-06, + "loss": 0.0003, + "step": 6200 + }, + { + "epoch": 24.9, + "grad_norm": 0.0006994607392698526, + "learning_rate": 1.6496205872649291e-06, + "loss": 0.0084, + "step": 6225 + }, + { + "epoch": 25.0, + "grad_norm": 0.0030341825913637877, + "learning_rate": 0.0, + "loss": 0.009, + "step": 6250 + } + ], + "logging_steps": 25, + "max_steps": 6250, + "num_input_tokens_seen": 0, + "num_train_epochs": 25, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 317621760000000.0, + "train_batch_size": 20, + "trial_name": null, + "trial_params": null +} diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6250/training_args.bin b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6250/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..ef3fc22e099df49861bca49d15da2b959cfcb6df --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6250/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ec1d8ed36c45af65d67a23de6c5216a486932bfb09b720e1a4417b82eaaa215d +size 5304 diff --git a/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6250/vocab.txt b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6250/vocab.txt new file mode 100644 index 0000000000000000000000000000000000000000..cc9ad53732e21ea9180f18e1fb67a38cc5702d4c --- /dev/null +++ b/Finetuned Models/finetuning_outputs/dnasplice/TinyDNABERT_base_model/checkpoint-6250/vocab.txt @@ -0,0 +1,4101 @@ +[PAD] +[UNK] +[CLS] +[SEP] +[MASK] +AAAAAA +AAAAAT +AAAAAC +AAAAAG +AAAATA +AAAATT +AAAATC +AAAATG +AAAACA +AAAACT +AAAACC +AAAACG +AAAAGA +AAAAGT +AAAAGC +AAAAGG +AAATAA +AAATAT +AAATAC +AAATAG +AAATTA +AAATTT +AAATTC +AAATTG +AAATCA +AAATCT +AAATCC +AAATCG +AAATGA +AAATGT +AAATGC +AAATGG +AAACAA +AAACAT +AAACAC +AAACAG +AAACTA +AAACTT +AAACTC +AAACTG +AAACCA +AAACCT +AAACCC +AAACCG +AAACGA +AAACGT +AAACGC +AAACGG +AAAGAA +AAAGAT +AAAGAC +AAAGAG +AAAGTA +AAAGTT +AAAGTC +AAAGTG +AAAGCA +AAAGCT +AAAGCC +AAAGCG +AAAGGA +AAAGGT +AAAGGC +AAAGGG +AATAAA +AATAAT +AATAAC +AATAAG +AATATA +AATATT +AATATC +AATATG +AATACA +AATACT +AATACC +AATACG +AATAGA +AATAGT +AATAGC +AATAGG +AATTAA +AATTAT +AATTAC +AATTAG +AATTTA +AATTTT +AATTTC +AATTTG +AATTCA +AATTCT +AATTCC +AATTCG +AATTGA +AATTGT +AATTGC +AATTGG +AATCAA +AATCAT +AATCAC +AATCAG +AATCTA +AATCTT +AATCTC +AATCTG +AATCCA +AATCCT +AATCCC +AATCCG +AATCGA +AATCGT +AATCGC +AATCGG +AATGAA +AATGAT +AATGAC +AATGAG +AATGTA +AATGTT +AATGTC +AATGTG +AATGCA +AATGCT +AATGCC +AATGCG +AATGGA +AATGGT +AATGGC +AATGGG +AACAAA +AACAAT +AACAAC +AACAAG +AACATA +AACATT +AACATC +AACATG +AACACA +AACACT +AACACC +AACACG +AACAGA +AACAGT +AACAGC +AACAGG +AACTAA +AACTAT +AACTAC +AACTAG +AACTTA +AACTTT +AACTTC +AACTTG +AACTCA +AACTCT +AACTCC +AACTCG +AACTGA +AACTGT +AACTGC +AACTGG +AACCAA +AACCAT +AACCAC +AACCAG +AACCTA +AACCTT +AACCTC +AACCTG +AACCCA +AACCCT +AACCCC +AACCCG +AACCGA +AACCGT +AACCGC +AACCGG +AACGAA +AACGAT +AACGAC +AACGAG +AACGTA +AACGTT +AACGTC +AACGTG +AACGCA +AACGCT +AACGCC +AACGCG +AACGGA +AACGGT +AACGGC +AACGGG +AAGAAA +AAGAAT +AAGAAC +AAGAAG +AAGATA +AAGATT +AAGATC +AAGATG +AAGACA +AAGACT +AAGACC +AAGACG +AAGAGA +AAGAGT +AAGAGC +AAGAGG +AAGTAA +AAGTAT +AAGTAC +AAGTAG +AAGTTA +AAGTTT +AAGTTC +AAGTTG +AAGTCA +AAGTCT +AAGTCC +AAGTCG +AAGTGA +AAGTGT +AAGTGC +AAGTGG +AAGCAA +AAGCAT +AAGCAC +AAGCAG +AAGCTA +AAGCTT +AAGCTC +AAGCTG +AAGCCA +AAGCCT +AAGCCC +AAGCCG +AAGCGA +AAGCGT +AAGCGC +AAGCGG +AAGGAA +AAGGAT +AAGGAC +AAGGAG +AAGGTA +AAGGTT +AAGGTC +AAGGTG +AAGGCA +AAGGCT +AAGGCC +AAGGCG +AAGGGA +AAGGGT +AAGGGC +AAGGGG +ATAAAA +ATAAAT +ATAAAC +ATAAAG +ATAATA +ATAATT +ATAATC +ATAATG +ATAACA +ATAACT +ATAACC +ATAACG +ATAAGA +ATAAGT +ATAAGC +ATAAGG +ATATAA +ATATAT +ATATAC +ATATAG +ATATTA +ATATTT +ATATTC +ATATTG +ATATCA +ATATCT +ATATCC +ATATCG +ATATGA +ATATGT +ATATGC +ATATGG +ATACAA +ATACAT +ATACAC +ATACAG +ATACTA +ATACTT +ATACTC +ATACTG +ATACCA +ATACCT +ATACCC +ATACCG +ATACGA +ATACGT +ATACGC +ATACGG +ATAGAA +ATAGAT +ATAGAC +ATAGAG +ATAGTA +ATAGTT +ATAGTC +ATAGTG +ATAGCA +ATAGCT +ATAGCC +ATAGCG +ATAGGA +ATAGGT +ATAGGC +ATAGGG +ATTAAA +ATTAAT +ATTAAC +ATTAAG +ATTATA +ATTATT +ATTATC +ATTATG +ATTACA +ATTACT +ATTACC +ATTACG +ATTAGA +ATTAGT +ATTAGC +ATTAGG +ATTTAA +ATTTAT +ATTTAC +ATTTAG +ATTTTA +ATTTTT +ATTTTC +ATTTTG +ATTTCA +ATTTCT +ATTTCC +ATTTCG +ATTTGA +ATTTGT +ATTTGC +ATTTGG +ATTCAA +ATTCAT +ATTCAC +ATTCAG +ATTCTA +ATTCTT +ATTCTC +ATTCTG +ATTCCA +ATTCCT +ATTCCC +ATTCCG +ATTCGA +ATTCGT +ATTCGC +ATTCGG +ATTGAA +ATTGAT +ATTGAC +ATTGAG +ATTGTA +ATTGTT +ATTGTC +ATTGTG +ATTGCA +ATTGCT +ATTGCC +ATTGCG +ATTGGA +ATTGGT +ATTGGC +ATTGGG +ATCAAA +ATCAAT +ATCAAC +ATCAAG +ATCATA +ATCATT +ATCATC +ATCATG +ATCACA +ATCACT +ATCACC +ATCACG +ATCAGA +ATCAGT +ATCAGC +ATCAGG +ATCTAA +ATCTAT +ATCTAC +ATCTAG +ATCTTA +ATCTTT +ATCTTC +ATCTTG +ATCTCA +ATCTCT +ATCTCC +ATCTCG +ATCTGA +ATCTGT +ATCTGC +ATCTGG +ATCCAA +ATCCAT +ATCCAC +ATCCAG +ATCCTA +ATCCTT +ATCCTC +ATCCTG +ATCCCA +ATCCCT +ATCCCC +ATCCCG +ATCCGA +ATCCGT +ATCCGC +ATCCGG +ATCGAA +ATCGAT +ATCGAC +ATCGAG +ATCGTA +ATCGTT +ATCGTC +ATCGTG +ATCGCA +ATCGCT +ATCGCC +ATCGCG +ATCGGA +ATCGGT +ATCGGC +ATCGGG +ATGAAA +ATGAAT +ATGAAC +ATGAAG +ATGATA +ATGATT +ATGATC +ATGATG +ATGACA +ATGACT +ATGACC +ATGACG +ATGAGA +ATGAGT +ATGAGC +ATGAGG +ATGTAA +ATGTAT +ATGTAC +ATGTAG +ATGTTA +ATGTTT +ATGTTC +ATGTTG +ATGTCA +ATGTCT +ATGTCC +ATGTCG +ATGTGA +ATGTGT +ATGTGC +ATGTGG +ATGCAA +ATGCAT +ATGCAC +ATGCAG +ATGCTA +ATGCTT +ATGCTC +ATGCTG +ATGCCA +ATGCCT +ATGCCC +ATGCCG +ATGCGA +ATGCGT +ATGCGC +ATGCGG +ATGGAA +ATGGAT +ATGGAC +ATGGAG +ATGGTA +ATGGTT +ATGGTC +ATGGTG +ATGGCA +ATGGCT +ATGGCC +ATGGCG +ATGGGA +ATGGGT +ATGGGC +ATGGGG +ACAAAA +ACAAAT +ACAAAC +ACAAAG +ACAATA +ACAATT +ACAATC +ACAATG +ACAACA +ACAACT +ACAACC +ACAACG +ACAAGA +ACAAGT +ACAAGC +ACAAGG +ACATAA +ACATAT +ACATAC +ACATAG +ACATTA +ACATTT +ACATTC +ACATTG +ACATCA +ACATCT +ACATCC +ACATCG +ACATGA +ACATGT +ACATGC +ACATGG +ACACAA +ACACAT +ACACAC +ACACAG +ACACTA +ACACTT +ACACTC +ACACTG +ACACCA +ACACCT +ACACCC +ACACCG +ACACGA +ACACGT +ACACGC +ACACGG +ACAGAA +ACAGAT +ACAGAC +ACAGAG +ACAGTA +ACAGTT +ACAGTC +ACAGTG +ACAGCA +ACAGCT +ACAGCC +ACAGCG +ACAGGA +ACAGGT +ACAGGC +ACAGGG +ACTAAA +ACTAAT +ACTAAC +ACTAAG +ACTATA +ACTATT +ACTATC +ACTATG +ACTACA +ACTACT +ACTACC +ACTACG +ACTAGA +ACTAGT +ACTAGC +ACTAGG +ACTTAA +ACTTAT +ACTTAC +ACTTAG +ACTTTA +ACTTTT +ACTTTC +ACTTTG +ACTTCA +ACTTCT +ACTTCC +ACTTCG +ACTTGA +ACTTGT +ACTTGC +ACTTGG +ACTCAA +ACTCAT +ACTCAC +ACTCAG +ACTCTA +ACTCTT +ACTCTC +ACTCTG +ACTCCA +ACTCCT +ACTCCC +ACTCCG +ACTCGA +ACTCGT +ACTCGC +ACTCGG +ACTGAA +ACTGAT +ACTGAC +ACTGAG +ACTGTA +ACTGTT +ACTGTC +ACTGTG +ACTGCA +ACTGCT +ACTGCC +ACTGCG +ACTGGA +ACTGGT +ACTGGC +ACTGGG +ACCAAA +ACCAAT +ACCAAC +ACCAAG +ACCATA +ACCATT +ACCATC +ACCATG +ACCACA +ACCACT +ACCACC +ACCACG +ACCAGA +ACCAGT +ACCAGC +ACCAGG +ACCTAA +ACCTAT +ACCTAC +ACCTAG +ACCTTA +ACCTTT +ACCTTC +ACCTTG +ACCTCA +ACCTCT +ACCTCC +ACCTCG +ACCTGA +ACCTGT +ACCTGC +ACCTGG +ACCCAA +ACCCAT +ACCCAC +ACCCAG +ACCCTA +ACCCTT +ACCCTC +ACCCTG +ACCCCA +ACCCCT +ACCCCC +ACCCCG +ACCCGA +ACCCGT +ACCCGC +ACCCGG +ACCGAA +ACCGAT +ACCGAC +ACCGAG +ACCGTA +ACCGTT +ACCGTC +ACCGTG +ACCGCA +ACCGCT +ACCGCC +ACCGCG +ACCGGA +ACCGGT +ACCGGC +ACCGGG +ACGAAA +ACGAAT +ACGAAC +ACGAAG +ACGATA +ACGATT +ACGATC +ACGATG +ACGACA +ACGACT +ACGACC +ACGACG +ACGAGA +ACGAGT +ACGAGC +ACGAGG +ACGTAA +ACGTAT +ACGTAC +ACGTAG +ACGTTA +ACGTTT +ACGTTC +ACGTTG +ACGTCA +ACGTCT +ACGTCC +ACGTCG +ACGTGA +ACGTGT +ACGTGC +ACGTGG +ACGCAA +ACGCAT +ACGCAC +ACGCAG +ACGCTA +ACGCTT +ACGCTC +ACGCTG +ACGCCA +ACGCCT +ACGCCC +ACGCCG +ACGCGA +ACGCGT +ACGCGC +ACGCGG +ACGGAA +ACGGAT +ACGGAC +ACGGAG +ACGGTA +ACGGTT +ACGGTC +ACGGTG +ACGGCA +ACGGCT +ACGGCC +ACGGCG +ACGGGA +ACGGGT +ACGGGC +ACGGGG +AGAAAA +AGAAAT +AGAAAC +AGAAAG +AGAATA +AGAATT +AGAATC +AGAATG +AGAACA +AGAACT +AGAACC +AGAACG +AGAAGA +AGAAGT +AGAAGC +AGAAGG +AGATAA +AGATAT +AGATAC +AGATAG +AGATTA +AGATTT +AGATTC +AGATTG +AGATCA +AGATCT +AGATCC +AGATCG +AGATGA +AGATGT +AGATGC +AGATGG +AGACAA +AGACAT +AGACAC +AGACAG +AGACTA +AGACTT +AGACTC +AGACTG +AGACCA +AGACCT +AGACCC +AGACCG +AGACGA +AGACGT +AGACGC +AGACGG +AGAGAA +AGAGAT +AGAGAC +AGAGAG +AGAGTA +AGAGTT +AGAGTC +AGAGTG +AGAGCA +AGAGCT +AGAGCC +AGAGCG +AGAGGA +AGAGGT +AGAGGC +AGAGGG +AGTAAA +AGTAAT +AGTAAC +AGTAAG +AGTATA +AGTATT +AGTATC +AGTATG +AGTACA +AGTACT +AGTACC +AGTACG +AGTAGA +AGTAGT +AGTAGC +AGTAGG +AGTTAA +AGTTAT +AGTTAC +AGTTAG +AGTTTA +AGTTTT +AGTTTC +AGTTTG +AGTTCA +AGTTCT +AGTTCC +AGTTCG +AGTTGA +AGTTGT +AGTTGC +AGTTGG +AGTCAA +AGTCAT +AGTCAC +AGTCAG +AGTCTA +AGTCTT +AGTCTC +AGTCTG +AGTCCA +AGTCCT +AGTCCC +AGTCCG +AGTCGA +AGTCGT +AGTCGC +AGTCGG +AGTGAA +AGTGAT +AGTGAC +AGTGAG +AGTGTA +AGTGTT +AGTGTC +AGTGTG +AGTGCA +AGTGCT +AGTGCC +AGTGCG +AGTGGA +AGTGGT +AGTGGC +AGTGGG +AGCAAA +AGCAAT +AGCAAC +AGCAAG +AGCATA +AGCATT +AGCATC +AGCATG +AGCACA +AGCACT +AGCACC +AGCACG +AGCAGA +AGCAGT +AGCAGC +AGCAGG +AGCTAA +AGCTAT +AGCTAC +AGCTAG +AGCTTA +AGCTTT +AGCTTC +AGCTTG +AGCTCA +AGCTCT +AGCTCC +AGCTCG +AGCTGA +AGCTGT +AGCTGC +AGCTGG +AGCCAA +AGCCAT +AGCCAC +AGCCAG +AGCCTA +AGCCTT +AGCCTC +AGCCTG +AGCCCA +AGCCCT +AGCCCC +AGCCCG +AGCCGA +AGCCGT +AGCCGC +AGCCGG +AGCGAA +AGCGAT +AGCGAC +AGCGAG +AGCGTA +AGCGTT +AGCGTC +AGCGTG +AGCGCA +AGCGCT +AGCGCC +AGCGCG +AGCGGA +AGCGGT +AGCGGC +AGCGGG +AGGAAA +AGGAAT +AGGAAC +AGGAAG +AGGATA +AGGATT +AGGATC +AGGATG +AGGACA +AGGACT +AGGACC +AGGACG +AGGAGA +AGGAGT +AGGAGC +AGGAGG +AGGTAA +AGGTAT +AGGTAC +AGGTAG +AGGTTA +AGGTTT +AGGTTC +AGGTTG +AGGTCA +AGGTCT +AGGTCC +AGGTCG +AGGTGA +AGGTGT +AGGTGC +AGGTGG +AGGCAA +AGGCAT +AGGCAC +AGGCAG +AGGCTA +AGGCTT +AGGCTC +AGGCTG +AGGCCA +AGGCCT +AGGCCC +AGGCCG +AGGCGA +AGGCGT +AGGCGC +AGGCGG +AGGGAA +AGGGAT +AGGGAC +AGGGAG +AGGGTA +AGGGTT +AGGGTC +AGGGTG +AGGGCA +AGGGCT +AGGGCC +AGGGCG +AGGGGA +AGGGGT +AGGGGC +AGGGGG +TAAAAA +TAAAAT +TAAAAC +TAAAAG +TAAATA +TAAATT +TAAATC +TAAATG +TAAACA +TAAACT +TAAACC +TAAACG +TAAAGA +TAAAGT +TAAAGC +TAAAGG +TAATAA +TAATAT +TAATAC +TAATAG +TAATTA +TAATTT +TAATTC +TAATTG +TAATCA +TAATCT +TAATCC +TAATCG +TAATGA +TAATGT +TAATGC +TAATGG +TAACAA +TAACAT +TAACAC +TAACAG +TAACTA +TAACTT +TAACTC +TAACTG +TAACCA +TAACCT +TAACCC +TAACCG +TAACGA +TAACGT +TAACGC +TAACGG +TAAGAA +TAAGAT +TAAGAC +TAAGAG +TAAGTA +TAAGTT +TAAGTC +TAAGTG +TAAGCA +TAAGCT +TAAGCC +TAAGCG +TAAGGA +TAAGGT +TAAGGC +TAAGGG +TATAAA +TATAAT +TATAAC +TATAAG +TATATA +TATATT +TATATC +TATATG +TATACA +TATACT +TATACC +TATACG +TATAGA +TATAGT +TATAGC +TATAGG +TATTAA +TATTAT +TATTAC +TATTAG +TATTTA +TATTTT +TATTTC +TATTTG +TATTCA +TATTCT +TATTCC +TATTCG +TATTGA +TATTGT +TATTGC +TATTGG +TATCAA +TATCAT +TATCAC +TATCAG +TATCTA +TATCTT +TATCTC +TATCTG +TATCCA +TATCCT +TATCCC +TATCCG +TATCGA +TATCGT +TATCGC +TATCGG +TATGAA +TATGAT +TATGAC +TATGAG +TATGTA +TATGTT +TATGTC +TATGTG +TATGCA +TATGCT +TATGCC +TATGCG +TATGGA +TATGGT +TATGGC +TATGGG +TACAAA +TACAAT +TACAAC +TACAAG +TACATA +TACATT +TACATC +TACATG +TACACA +TACACT +TACACC +TACACG +TACAGA +TACAGT +TACAGC +TACAGG +TACTAA +TACTAT +TACTAC +TACTAG +TACTTA +TACTTT +TACTTC +TACTTG +TACTCA +TACTCT +TACTCC +TACTCG +TACTGA +TACTGT +TACTGC +TACTGG +TACCAA +TACCAT +TACCAC +TACCAG +TACCTA +TACCTT +TACCTC +TACCTG +TACCCA +TACCCT +TACCCC +TACCCG +TACCGA +TACCGT +TACCGC +TACCGG +TACGAA +TACGAT +TACGAC +TACGAG +TACGTA +TACGTT +TACGTC +TACGTG +TACGCA +TACGCT +TACGCC +TACGCG +TACGGA +TACGGT +TACGGC +TACGGG +TAGAAA +TAGAAT +TAGAAC +TAGAAG +TAGATA +TAGATT +TAGATC +TAGATG +TAGACA +TAGACT +TAGACC +TAGACG +TAGAGA +TAGAGT +TAGAGC +TAGAGG +TAGTAA +TAGTAT +TAGTAC +TAGTAG +TAGTTA +TAGTTT +TAGTTC +TAGTTG +TAGTCA +TAGTCT +TAGTCC +TAGTCG +TAGTGA +TAGTGT +TAGTGC +TAGTGG +TAGCAA +TAGCAT +TAGCAC +TAGCAG +TAGCTA +TAGCTT +TAGCTC +TAGCTG +TAGCCA +TAGCCT +TAGCCC +TAGCCG +TAGCGA +TAGCGT +TAGCGC +TAGCGG +TAGGAA +TAGGAT +TAGGAC +TAGGAG +TAGGTA +TAGGTT +TAGGTC +TAGGTG +TAGGCA +TAGGCT +TAGGCC +TAGGCG +TAGGGA +TAGGGT +TAGGGC +TAGGGG +TTAAAA +TTAAAT +TTAAAC +TTAAAG +TTAATA +TTAATT +TTAATC +TTAATG +TTAACA +TTAACT +TTAACC +TTAACG +TTAAGA +TTAAGT +TTAAGC +TTAAGG +TTATAA +TTATAT +TTATAC +TTATAG +TTATTA +TTATTT +TTATTC +TTATTG +TTATCA +TTATCT +TTATCC +TTATCG +TTATGA +TTATGT +TTATGC +TTATGG +TTACAA +TTACAT +TTACAC +TTACAG +TTACTA +TTACTT +TTACTC +TTACTG +TTACCA +TTACCT +TTACCC +TTACCG +TTACGA +TTACGT +TTACGC +TTACGG +TTAGAA +TTAGAT +TTAGAC +TTAGAG +TTAGTA +TTAGTT +TTAGTC +TTAGTG +TTAGCA +TTAGCT +TTAGCC +TTAGCG +TTAGGA +TTAGGT +TTAGGC +TTAGGG +TTTAAA +TTTAAT +TTTAAC +TTTAAG +TTTATA +TTTATT +TTTATC +TTTATG +TTTACA +TTTACT +TTTACC +TTTACG +TTTAGA +TTTAGT +TTTAGC +TTTAGG +TTTTAA +TTTTAT +TTTTAC +TTTTAG +TTTTTA +TTTTTT +TTTTTC +TTTTTG +TTTTCA +TTTTCT +TTTTCC +TTTTCG +TTTTGA +TTTTGT +TTTTGC +TTTTGG +TTTCAA +TTTCAT +TTTCAC +TTTCAG +TTTCTA +TTTCTT +TTTCTC +TTTCTG +TTTCCA +TTTCCT +TTTCCC +TTTCCG +TTTCGA +TTTCGT +TTTCGC +TTTCGG +TTTGAA +TTTGAT +TTTGAC +TTTGAG +TTTGTA +TTTGTT +TTTGTC +TTTGTG +TTTGCA +TTTGCT +TTTGCC +TTTGCG +TTTGGA +TTTGGT +TTTGGC +TTTGGG +TTCAAA +TTCAAT +TTCAAC +TTCAAG +TTCATA +TTCATT +TTCATC +TTCATG +TTCACA +TTCACT +TTCACC +TTCACG +TTCAGA +TTCAGT +TTCAGC +TTCAGG +TTCTAA +TTCTAT +TTCTAC +TTCTAG +TTCTTA +TTCTTT +TTCTTC +TTCTTG +TTCTCA +TTCTCT +TTCTCC +TTCTCG +TTCTGA +TTCTGT +TTCTGC +TTCTGG +TTCCAA +TTCCAT +TTCCAC +TTCCAG +TTCCTA +TTCCTT +TTCCTC +TTCCTG +TTCCCA +TTCCCT +TTCCCC +TTCCCG +TTCCGA +TTCCGT +TTCCGC +TTCCGG +TTCGAA +TTCGAT +TTCGAC +TTCGAG +TTCGTA +TTCGTT +TTCGTC +TTCGTG +TTCGCA +TTCGCT +TTCGCC +TTCGCG +TTCGGA +TTCGGT +TTCGGC +TTCGGG +TTGAAA +TTGAAT +TTGAAC +TTGAAG +TTGATA +TTGATT +TTGATC +TTGATG +TTGACA +TTGACT +TTGACC +TTGACG +TTGAGA +TTGAGT +TTGAGC +TTGAGG +TTGTAA +TTGTAT +TTGTAC +TTGTAG +TTGTTA +TTGTTT +TTGTTC +TTGTTG +TTGTCA +TTGTCT +TTGTCC +TTGTCG +TTGTGA +TTGTGT +TTGTGC +TTGTGG +TTGCAA +TTGCAT +TTGCAC +TTGCAG +TTGCTA +TTGCTT +TTGCTC +TTGCTG +TTGCCA +TTGCCT +TTGCCC +TTGCCG +TTGCGA +TTGCGT +TTGCGC +TTGCGG +TTGGAA +TTGGAT +TTGGAC +TTGGAG +TTGGTA +TTGGTT +TTGGTC +TTGGTG +TTGGCA +TTGGCT +TTGGCC +TTGGCG +TTGGGA +TTGGGT +TTGGGC +TTGGGG +TCAAAA +TCAAAT +TCAAAC +TCAAAG +TCAATA +TCAATT +TCAATC +TCAATG +TCAACA +TCAACT +TCAACC +TCAACG +TCAAGA +TCAAGT +TCAAGC +TCAAGG +TCATAA +TCATAT +TCATAC +TCATAG +TCATTA +TCATTT +TCATTC +TCATTG +TCATCA +TCATCT +TCATCC +TCATCG +TCATGA +TCATGT +TCATGC +TCATGG +TCACAA +TCACAT +TCACAC +TCACAG +TCACTA +TCACTT +TCACTC +TCACTG +TCACCA +TCACCT +TCACCC +TCACCG +TCACGA +TCACGT +TCACGC +TCACGG +TCAGAA +TCAGAT +TCAGAC +TCAGAG +TCAGTA +TCAGTT +TCAGTC +TCAGTG +TCAGCA +TCAGCT +TCAGCC +TCAGCG +TCAGGA +TCAGGT +TCAGGC +TCAGGG +TCTAAA +TCTAAT +TCTAAC +TCTAAG +TCTATA +TCTATT +TCTATC +TCTATG +TCTACA +TCTACT +TCTACC +TCTACG +TCTAGA +TCTAGT +TCTAGC +TCTAGG +TCTTAA +TCTTAT +TCTTAC +TCTTAG +TCTTTA +TCTTTT +TCTTTC +TCTTTG +TCTTCA +TCTTCT +TCTTCC +TCTTCG +TCTTGA +TCTTGT +TCTTGC +TCTTGG +TCTCAA +TCTCAT +TCTCAC +TCTCAG +TCTCTA +TCTCTT +TCTCTC +TCTCTG +TCTCCA +TCTCCT +TCTCCC +TCTCCG +TCTCGA +TCTCGT +TCTCGC +TCTCGG +TCTGAA +TCTGAT +TCTGAC +TCTGAG +TCTGTA +TCTGTT +TCTGTC +TCTGTG +TCTGCA +TCTGCT +TCTGCC +TCTGCG +TCTGGA +TCTGGT +TCTGGC +TCTGGG +TCCAAA +TCCAAT +TCCAAC +TCCAAG +TCCATA +TCCATT +TCCATC +TCCATG +TCCACA +TCCACT +TCCACC +TCCACG +TCCAGA +TCCAGT +TCCAGC +TCCAGG +TCCTAA +TCCTAT +TCCTAC +TCCTAG +TCCTTA +TCCTTT +TCCTTC +TCCTTG +TCCTCA +TCCTCT +TCCTCC +TCCTCG +TCCTGA +TCCTGT +TCCTGC +TCCTGG +TCCCAA +TCCCAT +TCCCAC +TCCCAG +TCCCTA +TCCCTT +TCCCTC +TCCCTG +TCCCCA +TCCCCT +TCCCCC +TCCCCG +TCCCGA +TCCCGT +TCCCGC +TCCCGG +TCCGAA +TCCGAT +TCCGAC +TCCGAG +TCCGTA +TCCGTT +TCCGTC +TCCGTG +TCCGCA +TCCGCT +TCCGCC +TCCGCG +TCCGGA +TCCGGT +TCCGGC +TCCGGG +TCGAAA +TCGAAT +TCGAAC +TCGAAG +TCGATA +TCGATT +TCGATC +TCGATG +TCGACA +TCGACT +TCGACC +TCGACG +TCGAGA +TCGAGT +TCGAGC +TCGAGG +TCGTAA +TCGTAT +TCGTAC +TCGTAG +TCGTTA +TCGTTT +TCGTTC +TCGTTG +TCGTCA +TCGTCT +TCGTCC +TCGTCG +TCGTGA +TCGTGT +TCGTGC +TCGTGG +TCGCAA +TCGCAT +TCGCAC +TCGCAG +TCGCTA +TCGCTT +TCGCTC +TCGCTG +TCGCCA +TCGCCT +TCGCCC +TCGCCG +TCGCGA +TCGCGT +TCGCGC +TCGCGG +TCGGAA +TCGGAT +TCGGAC +TCGGAG +TCGGTA +TCGGTT +TCGGTC +TCGGTG +TCGGCA +TCGGCT +TCGGCC +TCGGCG +TCGGGA +TCGGGT +TCGGGC +TCGGGG +TGAAAA +TGAAAT +TGAAAC +TGAAAG +TGAATA +TGAATT +TGAATC +TGAATG +TGAACA +TGAACT +TGAACC +TGAACG +TGAAGA +TGAAGT +TGAAGC +TGAAGG +TGATAA +TGATAT +TGATAC +TGATAG +TGATTA +TGATTT +TGATTC +TGATTG +TGATCA +TGATCT +TGATCC +TGATCG +TGATGA +TGATGT +TGATGC +TGATGG +TGACAA +TGACAT +TGACAC +TGACAG +TGACTA +TGACTT +TGACTC +TGACTG +TGACCA +TGACCT +TGACCC +TGACCG +TGACGA +TGACGT +TGACGC +TGACGG +TGAGAA +TGAGAT +TGAGAC +TGAGAG +TGAGTA +TGAGTT +TGAGTC +TGAGTG +TGAGCA +TGAGCT +TGAGCC +TGAGCG +TGAGGA +TGAGGT +TGAGGC +TGAGGG +TGTAAA +TGTAAT +TGTAAC +TGTAAG +TGTATA +TGTATT +TGTATC +TGTATG +TGTACA +TGTACT +TGTACC +TGTACG +TGTAGA +TGTAGT +TGTAGC +TGTAGG +TGTTAA +TGTTAT +TGTTAC +TGTTAG +TGTTTA +TGTTTT +TGTTTC +TGTTTG +TGTTCA +TGTTCT +TGTTCC +TGTTCG +TGTTGA +TGTTGT +TGTTGC +TGTTGG +TGTCAA +TGTCAT +TGTCAC +TGTCAG +TGTCTA +TGTCTT +TGTCTC +TGTCTG +TGTCCA +TGTCCT +TGTCCC +TGTCCG +TGTCGA +TGTCGT +TGTCGC +TGTCGG +TGTGAA +TGTGAT +TGTGAC +TGTGAG +TGTGTA +TGTGTT +TGTGTC +TGTGTG +TGTGCA +TGTGCT +TGTGCC +TGTGCG +TGTGGA +TGTGGT +TGTGGC +TGTGGG +TGCAAA +TGCAAT +TGCAAC +TGCAAG +TGCATA +TGCATT +TGCATC +TGCATG +TGCACA +TGCACT +TGCACC +TGCACG +TGCAGA +TGCAGT +TGCAGC +TGCAGG +TGCTAA +TGCTAT +TGCTAC +TGCTAG +TGCTTA +TGCTTT +TGCTTC +TGCTTG +TGCTCA +TGCTCT +TGCTCC +TGCTCG +TGCTGA +TGCTGT +TGCTGC +TGCTGG +TGCCAA +TGCCAT +TGCCAC +TGCCAG +TGCCTA +TGCCTT +TGCCTC +TGCCTG +TGCCCA +TGCCCT +TGCCCC +TGCCCG +TGCCGA +TGCCGT +TGCCGC +TGCCGG +TGCGAA +TGCGAT +TGCGAC +TGCGAG +TGCGTA +TGCGTT +TGCGTC +TGCGTG +TGCGCA +TGCGCT +TGCGCC +TGCGCG +TGCGGA +TGCGGT +TGCGGC +TGCGGG +TGGAAA +TGGAAT +TGGAAC +TGGAAG +TGGATA +TGGATT +TGGATC +TGGATG +TGGACA +TGGACT +TGGACC +TGGACG +TGGAGA +TGGAGT +TGGAGC +TGGAGG +TGGTAA +TGGTAT +TGGTAC +TGGTAG +TGGTTA +TGGTTT +TGGTTC +TGGTTG +TGGTCA +TGGTCT +TGGTCC +TGGTCG +TGGTGA +TGGTGT +TGGTGC +TGGTGG +TGGCAA +TGGCAT +TGGCAC +TGGCAG +TGGCTA +TGGCTT +TGGCTC +TGGCTG +TGGCCA +TGGCCT +TGGCCC +TGGCCG +TGGCGA +TGGCGT +TGGCGC +TGGCGG +TGGGAA +TGGGAT +TGGGAC +TGGGAG +TGGGTA +TGGGTT +TGGGTC +TGGGTG +TGGGCA +TGGGCT +TGGGCC +TGGGCG +TGGGGA +TGGGGT +TGGGGC +TGGGGG +CAAAAA +CAAAAT +CAAAAC +CAAAAG +CAAATA +CAAATT +CAAATC +CAAATG +CAAACA +CAAACT +CAAACC +CAAACG +CAAAGA +CAAAGT +CAAAGC +CAAAGG +CAATAA +CAATAT +CAATAC +CAATAG +CAATTA +CAATTT +CAATTC +CAATTG +CAATCA +CAATCT +CAATCC +CAATCG +CAATGA +CAATGT +CAATGC +CAATGG +CAACAA +CAACAT +CAACAC +CAACAG +CAACTA +CAACTT +CAACTC +CAACTG +CAACCA +CAACCT +CAACCC +CAACCG +CAACGA +CAACGT +CAACGC +CAACGG +CAAGAA +CAAGAT +CAAGAC +CAAGAG +CAAGTA +CAAGTT +CAAGTC +CAAGTG +CAAGCA +CAAGCT +CAAGCC +CAAGCG +CAAGGA +CAAGGT +CAAGGC +CAAGGG +CATAAA +CATAAT +CATAAC +CATAAG +CATATA +CATATT +CATATC +CATATG +CATACA +CATACT +CATACC +CATACG +CATAGA +CATAGT +CATAGC +CATAGG +CATTAA +CATTAT +CATTAC +CATTAG +CATTTA +CATTTT +CATTTC +CATTTG +CATTCA +CATTCT +CATTCC +CATTCG +CATTGA +CATTGT +CATTGC +CATTGG +CATCAA +CATCAT +CATCAC +CATCAG +CATCTA +CATCTT +CATCTC +CATCTG +CATCCA +CATCCT +CATCCC +CATCCG +CATCGA +CATCGT +CATCGC +CATCGG +CATGAA +CATGAT +CATGAC +CATGAG +CATGTA +CATGTT +CATGTC +CATGTG +CATGCA +CATGCT +CATGCC +CATGCG +CATGGA +CATGGT +CATGGC +CATGGG +CACAAA +CACAAT +CACAAC +CACAAG +CACATA +CACATT +CACATC +CACATG +CACACA +CACACT +CACACC +CACACG +CACAGA +CACAGT +CACAGC +CACAGG +CACTAA +CACTAT +CACTAC +CACTAG +CACTTA +CACTTT +CACTTC +CACTTG +CACTCA +CACTCT +CACTCC +CACTCG +CACTGA +CACTGT +CACTGC +CACTGG +CACCAA +CACCAT +CACCAC +CACCAG +CACCTA +CACCTT +CACCTC +CACCTG +CACCCA +CACCCT +CACCCC +CACCCG +CACCGA +CACCGT +CACCGC +CACCGG +CACGAA +CACGAT +CACGAC +CACGAG +CACGTA +CACGTT +CACGTC +CACGTG +CACGCA +CACGCT +CACGCC +CACGCG +CACGGA +CACGGT +CACGGC +CACGGG +CAGAAA +CAGAAT +CAGAAC +CAGAAG +CAGATA +CAGATT +CAGATC +CAGATG +CAGACA +CAGACT +CAGACC +CAGACG +CAGAGA +CAGAGT +CAGAGC +CAGAGG +CAGTAA +CAGTAT +CAGTAC +CAGTAG +CAGTTA +CAGTTT +CAGTTC +CAGTTG +CAGTCA +CAGTCT +CAGTCC +CAGTCG +CAGTGA +CAGTGT +CAGTGC +CAGTGG +CAGCAA +CAGCAT +CAGCAC +CAGCAG +CAGCTA +CAGCTT +CAGCTC +CAGCTG +CAGCCA +CAGCCT +CAGCCC +CAGCCG +CAGCGA +CAGCGT +CAGCGC +CAGCGG +CAGGAA +CAGGAT +CAGGAC +CAGGAG +CAGGTA +CAGGTT +CAGGTC +CAGGTG +CAGGCA +CAGGCT +CAGGCC +CAGGCG +CAGGGA +CAGGGT +CAGGGC +CAGGGG +CTAAAA +CTAAAT +CTAAAC +CTAAAG +CTAATA +CTAATT +CTAATC +CTAATG +CTAACA +CTAACT +CTAACC +CTAACG +CTAAGA +CTAAGT +CTAAGC +CTAAGG +CTATAA +CTATAT +CTATAC +CTATAG +CTATTA +CTATTT +CTATTC +CTATTG +CTATCA +CTATCT +CTATCC +CTATCG +CTATGA +CTATGT +CTATGC +CTATGG +CTACAA +CTACAT +CTACAC +CTACAG +CTACTA +CTACTT +CTACTC +CTACTG +CTACCA +CTACCT +CTACCC +CTACCG +CTACGA +CTACGT +CTACGC +CTACGG +CTAGAA +CTAGAT +CTAGAC +CTAGAG +CTAGTA +CTAGTT +CTAGTC +CTAGTG +CTAGCA +CTAGCT +CTAGCC +CTAGCG +CTAGGA +CTAGGT +CTAGGC +CTAGGG +CTTAAA +CTTAAT +CTTAAC +CTTAAG +CTTATA +CTTATT +CTTATC +CTTATG +CTTACA +CTTACT +CTTACC +CTTACG +CTTAGA +CTTAGT +CTTAGC +CTTAGG +CTTTAA +CTTTAT +CTTTAC +CTTTAG +CTTTTA +CTTTTT +CTTTTC +CTTTTG +CTTTCA +CTTTCT +CTTTCC +CTTTCG +CTTTGA +CTTTGT +CTTTGC +CTTTGG +CTTCAA +CTTCAT +CTTCAC +CTTCAG +CTTCTA +CTTCTT +CTTCTC +CTTCTG +CTTCCA +CTTCCT +CTTCCC +CTTCCG +CTTCGA +CTTCGT +CTTCGC +CTTCGG +CTTGAA +CTTGAT +CTTGAC +CTTGAG +CTTGTA +CTTGTT +CTTGTC +CTTGTG +CTTGCA +CTTGCT +CTTGCC +CTTGCG +CTTGGA +CTTGGT +CTTGGC +CTTGGG +CTCAAA +CTCAAT +CTCAAC +CTCAAG +CTCATA +CTCATT +CTCATC +CTCATG +CTCACA +CTCACT +CTCACC +CTCACG +CTCAGA +CTCAGT +CTCAGC +CTCAGG +CTCTAA +CTCTAT +CTCTAC +CTCTAG +CTCTTA +CTCTTT +CTCTTC +CTCTTG +CTCTCA +CTCTCT +CTCTCC +CTCTCG +CTCTGA +CTCTGT +CTCTGC +CTCTGG +CTCCAA +CTCCAT +CTCCAC +CTCCAG +CTCCTA +CTCCTT +CTCCTC +CTCCTG +CTCCCA +CTCCCT +CTCCCC +CTCCCG +CTCCGA +CTCCGT +CTCCGC +CTCCGG +CTCGAA +CTCGAT +CTCGAC +CTCGAG +CTCGTA +CTCGTT +CTCGTC +CTCGTG +CTCGCA +CTCGCT +CTCGCC +CTCGCG +CTCGGA +CTCGGT +CTCGGC +CTCGGG +CTGAAA +CTGAAT +CTGAAC +CTGAAG +CTGATA +CTGATT +CTGATC +CTGATG +CTGACA +CTGACT +CTGACC +CTGACG +CTGAGA +CTGAGT +CTGAGC +CTGAGG +CTGTAA +CTGTAT +CTGTAC +CTGTAG +CTGTTA +CTGTTT +CTGTTC +CTGTTG +CTGTCA +CTGTCT +CTGTCC +CTGTCG +CTGTGA +CTGTGT +CTGTGC +CTGTGG +CTGCAA +CTGCAT +CTGCAC +CTGCAG +CTGCTA +CTGCTT +CTGCTC +CTGCTG +CTGCCA +CTGCCT +CTGCCC +CTGCCG +CTGCGA +CTGCGT +CTGCGC +CTGCGG +CTGGAA +CTGGAT +CTGGAC +CTGGAG +CTGGTA +CTGGTT +CTGGTC +CTGGTG +CTGGCA +CTGGCT +CTGGCC +CTGGCG +CTGGGA +CTGGGT +CTGGGC +CTGGGG +CCAAAA +CCAAAT +CCAAAC +CCAAAG +CCAATA +CCAATT +CCAATC +CCAATG +CCAACA +CCAACT +CCAACC +CCAACG +CCAAGA +CCAAGT +CCAAGC +CCAAGG +CCATAA +CCATAT +CCATAC +CCATAG +CCATTA +CCATTT +CCATTC +CCATTG +CCATCA +CCATCT +CCATCC +CCATCG +CCATGA +CCATGT +CCATGC +CCATGG +CCACAA +CCACAT +CCACAC +CCACAG +CCACTA +CCACTT +CCACTC +CCACTG +CCACCA +CCACCT +CCACCC +CCACCG +CCACGA +CCACGT +CCACGC +CCACGG +CCAGAA +CCAGAT +CCAGAC +CCAGAG +CCAGTA +CCAGTT +CCAGTC +CCAGTG +CCAGCA +CCAGCT +CCAGCC +CCAGCG +CCAGGA +CCAGGT +CCAGGC +CCAGGG +CCTAAA +CCTAAT +CCTAAC +CCTAAG +CCTATA +CCTATT +CCTATC +CCTATG +CCTACA +CCTACT +CCTACC +CCTACG +CCTAGA +CCTAGT +CCTAGC +CCTAGG +CCTTAA +CCTTAT +CCTTAC +CCTTAG +CCTTTA +CCTTTT +CCTTTC +CCTTTG +CCTTCA +CCTTCT +CCTTCC +CCTTCG +CCTTGA +CCTTGT +CCTTGC +CCTTGG +CCTCAA +CCTCAT +CCTCAC +CCTCAG +CCTCTA +CCTCTT +CCTCTC +CCTCTG +CCTCCA +CCTCCT +CCTCCC +CCTCCG +CCTCGA +CCTCGT +CCTCGC +CCTCGG +CCTGAA +CCTGAT +CCTGAC +CCTGAG +CCTGTA +CCTGTT +CCTGTC +CCTGTG +CCTGCA +CCTGCT +CCTGCC +CCTGCG +CCTGGA +CCTGGT +CCTGGC +CCTGGG +CCCAAA +CCCAAT +CCCAAC +CCCAAG +CCCATA +CCCATT +CCCATC +CCCATG +CCCACA +CCCACT +CCCACC +CCCACG +CCCAGA +CCCAGT +CCCAGC +CCCAGG +CCCTAA +CCCTAT +CCCTAC +CCCTAG +CCCTTA +CCCTTT +CCCTTC +CCCTTG +CCCTCA +CCCTCT +CCCTCC +CCCTCG +CCCTGA +CCCTGT +CCCTGC +CCCTGG +CCCCAA +CCCCAT +CCCCAC +CCCCAG +CCCCTA +CCCCTT +CCCCTC +CCCCTG +CCCCCA +CCCCCT +CCCCCC +CCCCCG +CCCCGA +CCCCGT +CCCCGC +CCCCGG +CCCGAA +CCCGAT +CCCGAC +CCCGAG +CCCGTA +CCCGTT +CCCGTC +CCCGTG +CCCGCA +CCCGCT +CCCGCC +CCCGCG +CCCGGA +CCCGGT +CCCGGC +CCCGGG +CCGAAA +CCGAAT +CCGAAC +CCGAAG +CCGATA +CCGATT +CCGATC +CCGATG +CCGACA +CCGACT +CCGACC +CCGACG +CCGAGA +CCGAGT +CCGAGC +CCGAGG +CCGTAA +CCGTAT +CCGTAC +CCGTAG +CCGTTA +CCGTTT +CCGTTC +CCGTTG +CCGTCA +CCGTCT +CCGTCC +CCGTCG +CCGTGA +CCGTGT +CCGTGC +CCGTGG +CCGCAA +CCGCAT +CCGCAC +CCGCAG +CCGCTA +CCGCTT +CCGCTC +CCGCTG +CCGCCA +CCGCCT +CCGCCC +CCGCCG +CCGCGA +CCGCGT +CCGCGC +CCGCGG +CCGGAA +CCGGAT +CCGGAC +CCGGAG +CCGGTA +CCGGTT +CCGGTC +CCGGTG +CCGGCA +CCGGCT +CCGGCC +CCGGCG +CCGGGA +CCGGGT +CCGGGC +CCGGGG +CGAAAA +CGAAAT +CGAAAC +CGAAAG +CGAATA +CGAATT +CGAATC +CGAATG +CGAACA +CGAACT +CGAACC +CGAACG +CGAAGA +CGAAGT +CGAAGC +CGAAGG +CGATAA +CGATAT +CGATAC +CGATAG +CGATTA +CGATTT +CGATTC +CGATTG +CGATCA +CGATCT +CGATCC +CGATCG +CGATGA +CGATGT +CGATGC +CGATGG +CGACAA +CGACAT +CGACAC +CGACAG +CGACTA +CGACTT +CGACTC +CGACTG +CGACCA +CGACCT +CGACCC +CGACCG +CGACGA +CGACGT +CGACGC +CGACGG +CGAGAA +CGAGAT +CGAGAC +CGAGAG +CGAGTA +CGAGTT +CGAGTC +CGAGTG +CGAGCA +CGAGCT +CGAGCC +CGAGCG +CGAGGA +CGAGGT +CGAGGC +CGAGGG +CGTAAA +CGTAAT +CGTAAC +CGTAAG +CGTATA +CGTATT +CGTATC +CGTATG +CGTACA +CGTACT +CGTACC +CGTACG +CGTAGA +CGTAGT +CGTAGC +CGTAGG +CGTTAA +CGTTAT +CGTTAC +CGTTAG +CGTTTA +CGTTTT +CGTTTC +CGTTTG +CGTTCA +CGTTCT +CGTTCC +CGTTCG +CGTTGA +CGTTGT +CGTTGC +CGTTGG +CGTCAA +CGTCAT +CGTCAC +CGTCAG +CGTCTA +CGTCTT +CGTCTC +CGTCTG +CGTCCA +CGTCCT +CGTCCC +CGTCCG +CGTCGA +CGTCGT +CGTCGC +CGTCGG +CGTGAA +CGTGAT +CGTGAC +CGTGAG +CGTGTA +CGTGTT +CGTGTC +CGTGTG +CGTGCA +CGTGCT +CGTGCC +CGTGCG +CGTGGA +CGTGGT +CGTGGC +CGTGGG +CGCAAA +CGCAAT +CGCAAC +CGCAAG +CGCATA +CGCATT +CGCATC +CGCATG +CGCACA +CGCACT +CGCACC +CGCACG +CGCAGA +CGCAGT +CGCAGC +CGCAGG +CGCTAA +CGCTAT +CGCTAC +CGCTAG +CGCTTA +CGCTTT +CGCTTC +CGCTTG +CGCTCA +CGCTCT +CGCTCC +CGCTCG +CGCTGA +CGCTGT +CGCTGC +CGCTGG +CGCCAA +CGCCAT +CGCCAC +CGCCAG +CGCCTA +CGCCTT +CGCCTC +CGCCTG +CGCCCA +CGCCCT +CGCCCC +CGCCCG +CGCCGA +CGCCGT +CGCCGC +CGCCGG +CGCGAA +CGCGAT +CGCGAC +CGCGAG +CGCGTA +CGCGTT +CGCGTC +CGCGTG +CGCGCA +CGCGCT +CGCGCC +CGCGCG +CGCGGA +CGCGGT +CGCGGC +CGCGGG +CGGAAA +CGGAAT +CGGAAC +CGGAAG +CGGATA +CGGATT +CGGATC +CGGATG +CGGACA +CGGACT +CGGACC +CGGACG +CGGAGA +CGGAGT +CGGAGC +CGGAGG +CGGTAA +CGGTAT +CGGTAC +CGGTAG +CGGTTA +CGGTTT +CGGTTC +CGGTTG +CGGTCA +CGGTCT +CGGTCC +CGGTCG +CGGTGA +CGGTGT +CGGTGC +CGGTGG +CGGCAA +CGGCAT +CGGCAC +CGGCAG +CGGCTA +CGGCTT +CGGCTC +CGGCTG +CGGCCA +CGGCCT +CGGCCC +CGGCCG +CGGCGA +CGGCGT +CGGCGC +CGGCGG +CGGGAA +CGGGAT +CGGGAC +CGGGAG +CGGGTA +CGGGTT +CGGGTC +CGGGTG +CGGGCA +CGGGCT +CGGGCC +CGGGCG +CGGGGA +CGGGGT +CGGGGC +CGGGGG +GAAAAA +GAAAAT +GAAAAC +GAAAAG +GAAATA +GAAATT +GAAATC +GAAATG +GAAACA +GAAACT +GAAACC +GAAACG +GAAAGA +GAAAGT +GAAAGC +GAAAGG +GAATAA +GAATAT +GAATAC +GAATAG +GAATTA +GAATTT +GAATTC +GAATTG +GAATCA +GAATCT +GAATCC +GAATCG +GAATGA +GAATGT +GAATGC +GAATGG +GAACAA +GAACAT +GAACAC +GAACAG +GAACTA +GAACTT +GAACTC +GAACTG +GAACCA +GAACCT +GAACCC +GAACCG +GAACGA +GAACGT +GAACGC +GAACGG +GAAGAA +GAAGAT +GAAGAC +GAAGAG +GAAGTA +GAAGTT +GAAGTC +GAAGTG +GAAGCA +GAAGCT +GAAGCC +GAAGCG +GAAGGA +GAAGGT +GAAGGC +GAAGGG +GATAAA +GATAAT +GATAAC +GATAAG +GATATA +GATATT +GATATC +GATATG +GATACA +GATACT +GATACC +GATACG +GATAGA +GATAGT +GATAGC +GATAGG +GATTAA +GATTAT +GATTAC +GATTAG +GATTTA +GATTTT +GATTTC +GATTTG +GATTCA +GATTCT +GATTCC +GATTCG +GATTGA +GATTGT +GATTGC +GATTGG +GATCAA +GATCAT +GATCAC +GATCAG +GATCTA +GATCTT +GATCTC +GATCTG +GATCCA +GATCCT +GATCCC +GATCCG +GATCGA +GATCGT +GATCGC +GATCGG +GATGAA +GATGAT +GATGAC +GATGAG +GATGTA +GATGTT +GATGTC +GATGTG +GATGCA +GATGCT +GATGCC +GATGCG +GATGGA +GATGGT +GATGGC +GATGGG +GACAAA +GACAAT +GACAAC +GACAAG +GACATA +GACATT +GACATC +GACATG +GACACA +GACACT +GACACC +GACACG +GACAGA +GACAGT +GACAGC +GACAGG +GACTAA +GACTAT +GACTAC +GACTAG +GACTTA +GACTTT +GACTTC +GACTTG +GACTCA +GACTCT +GACTCC +GACTCG +GACTGA +GACTGT +GACTGC +GACTGG +GACCAA +GACCAT +GACCAC +GACCAG +GACCTA +GACCTT +GACCTC +GACCTG +GACCCA +GACCCT +GACCCC +GACCCG +GACCGA +GACCGT +GACCGC +GACCGG +GACGAA +GACGAT +GACGAC +GACGAG +GACGTA +GACGTT +GACGTC +GACGTG +GACGCA +GACGCT +GACGCC +GACGCG +GACGGA +GACGGT +GACGGC +GACGGG +GAGAAA +GAGAAT +GAGAAC +GAGAAG +GAGATA +GAGATT +GAGATC +GAGATG +GAGACA +GAGACT +GAGACC +GAGACG +GAGAGA +GAGAGT +GAGAGC +GAGAGG +GAGTAA +GAGTAT +GAGTAC +GAGTAG +GAGTTA +GAGTTT +GAGTTC +GAGTTG +GAGTCA +GAGTCT +GAGTCC +GAGTCG +GAGTGA +GAGTGT +GAGTGC +GAGTGG +GAGCAA +GAGCAT +GAGCAC +GAGCAG +GAGCTA +GAGCTT +GAGCTC +GAGCTG +GAGCCA +GAGCCT +GAGCCC +GAGCCG +GAGCGA +GAGCGT +GAGCGC +GAGCGG +GAGGAA +GAGGAT +GAGGAC +GAGGAG +GAGGTA +GAGGTT +GAGGTC +GAGGTG +GAGGCA +GAGGCT +GAGGCC +GAGGCG +GAGGGA +GAGGGT +GAGGGC +GAGGGG +GTAAAA +GTAAAT +GTAAAC +GTAAAG +GTAATA +GTAATT +GTAATC +GTAATG +GTAACA +GTAACT +GTAACC +GTAACG +GTAAGA +GTAAGT +GTAAGC +GTAAGG +GTATAA +GTATAT +GTATAC +GTATAG +GTATTA +GTATTT +GTATTC +GTATTG +GTATCA +GTATCT +GTATCC +GTATCG +GTATGA +GTATGT +GTATGC +GTATGG +GTACAA +GTACAT +GTACAC +GTACAG +GTACTA +GTACTT +GTACTC +GTACTG +GTACCA +GTACCT +GTACCC +GTACCG +GTACGA +GTACGT +GTACGC +GTACGG +GTAGAA +GTAGAT +GTAGAC +GTAGAG +GTAGTA +GTAGTT +GTAGTC +GTAGTG +GTAGCA +GTAGCT +GTAGCC +GTAGCG +GTAGGA +GTAGGT +GTAGGC +GTAGGG +GTTAAA +GTTAAT +GTTAAC +GTTAAG +GTTATA +GTTATT +GTTATC +GTTATG +GTTACA +GTTACT +GTTACC +GTTACG +GTTAGA +GTTAGT +GTTAGC +GTTAGG +GTTTAA +GTTTAT +GTTTAC +GTTTAG +GTTTTA +GTTTTT +GTTTTC +GTTTTG +GTTTCA +GTTTCT +GTTTCC +GTTTCG +GTTTGA +GTTTGT +GTTTGC +GTTTGG +GTTCAA +GTTCAT +GTTCAC +GTTCAG +GTTCTA +GTTCTT +GTTCTC +GTTCTG +GTTCCA +GTTCCT +GTTCCC +GTTCCG +GTTCGA +GTTCGT +GTTCGC +GTTCGG +GTTGAA +GTTGAT +GTTGAC +GTTGAG +GTTGTA +GTTGTT +GTTGTC +GTTGTG +GTTGCA +GTTGCT +GTTGCC +GTTGCG +GTTGGA +GTTGGT +GTTGGC +GTTGGG +GTCAAA +GTCAAT +GTCAAC +GTCAAG +GTCATA +GTCATT +GTCATC +GTCATG +GTCACA +GTCACT +GTCACC +GTCACG +GTCAGA +GTCAGT +GTCAGC +GTCAGG +GTCTAA +GTCTAT +GTCTAC +GTCTAG +GTCTTA +GTCTTT +GTCTTC +GTCTTG +GTCTCA +GTCTCT +GTCTCC +GTCTCG +GTCTGA +GTCTGT +GTCTGC +GTCTGG +GTCCAA +GTCCAT +GTCCAC +GTCCAG +GTCCTA +GTCCTT +GTCCTC +GTCCTG +GTCCCA +GTCCCT +GTCCCC +GTCCCG +GTCCGA +GTCCGT +GTCCGC +GTCCGG +GTCGAA +GTCGAT +GTCGAC +GTCGAG +GTCGTA +GTCGTT +GTCGTC +GTCGTG +GTCGCA +GTCGCT +GTCGCC +GTCGCG +GTCGGA +GTCGGT +GTCGGC +GTCGGG +GTGAAA +GTGAAT +GTGAAC +GTGAAG +GTGATA +GTGATT +GTGATC +GTGATG +GTGACA +GTGACT +GTGACC +GTGACG +GTGAGA +GTGAGT +GTGAGC +GTGAGG +GTGTAA +GTGTAT +GTGTAC +GTGTAG +GTGTTA +GTGTTT +GTGTTC +GTGTTG +GTGTCA +GTGTCT +GTGTCC +GTGTCG +GTGTGA +GTGTGT +GTGTGC +GTGTGG +GTGCAA +GTGCAT +GTGCAC +GTGCAG +GTGCTA +GTGCTT +GTGCTC +GTGCTG +GTGCCA +GTGCCT +GTGCCC +GTGCCG +GTGCGA +GTGCGT +GTGCGC +GTGCGG +GTGGAA +GTGGAT +GTGGAC +GTGGAG +GTGGTA +GTGGTT +GTGGTC +GTGGTG +GTGGCA +GTGGCT +GTGGCC +GTGGCG +GTGGGA +GTGGGT +GTGGGC +GTGGGG +GCAAAA +GCAAAT +GCAAAC +GCAAAG +GCAATA +GCAATT +GCAATC +GCAATG +GCAACA +GCAACT +GCAACC +GCAACG +GCAAGA +GCAAGT +GCAAGC +GCAAGG +GCATAA +GCATAT +GCATAC +GCATAG +GCATTA +GCATTT +GCATTC +GCATTG +GCATCA +GCATCT +GCATCC +GCATCG +GCATGA +GCATGT +GCATGC +GCATGG +GCACAA +GCACAT +GCACAC +GCACAG +GCACTA +GCACTT +GCACTC +GCACTG +GCACCA +GCACCT +GCACCC +GCACCG +GCACGA +GCACGT +GCACGC +GCACGG +GCAGAA +GCAGAT +GCAGAC +GCAGAG +GCAGTA +GCAGTT +GCAGTC +GCAGTG +GCAGCA +GCAGCT +GCAGCC +GCAGCG +GCAGGA +GCAGGT +GCAGGC +GCAGGG +GCTAAA +GCTAAT +GCTAAC +GCTAAG +GCTATA +GCTATT +GCTATC +GCTATG +GCTACA +GCTACT +GCTACC +GCTACG +GCTAGA +GCTAGT +GCTAGC +GCTAGG +GCTTAA +GCTTAT +GCTTAC +GCTTAG +GCTTTA +GCTTTT +GCTTTC +GCTTTG +GCTTCA +GCTTCT +GCTTCC +GCTTCG +GCTTGA +GCTTGT +GCTTGC +GCTTGG +GCTCAA +GCTCAT +GCTCAC +GCTCAG +GCTCTA +GCTCTT +GCTCTC +GCTCTG +GCTCCA +GCTCCT +GCTCCC +GCTCCG +GCTCGA +GCTCGT +GCTCGC +GCTCGG +GCTGAA +GCTGAT +GCTGAC +GCTGAG +GCTGTA +GCTGTT +GCTGTC +GCTGTG +GCTGCA +GCTGCT +GCTGCC +GCTGCG +GCTGGA +GCTGGT +GCTGGC +GCTGGG +GCCAAA +GCCAAT +GCCAAC +GCCAAG +GCCATA +GCCATT +GCCATC +GCCATG +GCCACA +GCCACT +GCCACC +GCCACG +GCCAGA +GCCAGT +GCCAGC +GCCAGG +GCCTAA +GCCTAT +GCCTAC +GCCTAG +GCCTTA +GCCTTT +GCCTTC +GCCTTG +GCCTCA +GCCTCT +GCCTCC +GCCTCG +GCCTGA +GCCTGT +GCCTGC +GCCTGG +GCCCAA +GCCCAT +GCCCAC +GCCCAG +GCCCTA +GCCCTT +GCCCTC +GCCCTG +GCCCCA +GCCCCT +GCCCCC +GCCCCG +GCCCGA +GCCCGT +GCCCGC +GCCCGG +GCCGAA +GCCGAT +GCCGAC +GCCGAG +GCCGTA +GCCGTT +GCCGTC +GCCGTG +GCCGCA +GCCGCT +GCCGCC +GCCGCG +GCCGGA +GCCGGT +GCCGGC +GCCGGG +GCGAAA +GCGAAT +GCGAAC +GCGAAG +GCGATA +GCGATT +GCGATC +GCGATG +GCGACA +GCGACT +GCGACC +GCGACG +GCGAGA +GCGAGT +GCGAGC +GCGAGG +GCGTAA +GCGTAT +GCGTAC +GCGTAG +GCGTTA +GCGTTT +GCGTTC +GCGTTG +GCGTCA +GCGTCT +GCGTCC +GCGTCG +GCGTGA +GCGTGT +GCGTGC +GCGTGG +GCGCAA +GCGCAT +GCGCAC +GCGCAG +GCGCTA +GCGCTT +GCGCTC +GCGCTG +GCGCCA +GCGCCT +GCGCCC +GCGCCG +GCGCGA +GCGCGT +GCGCGC +GCGCGG +GCGGAA +GCGGAT +GCGGAC +GCGGAG +GCGGTA +GCGGTT +GCGGTC +GCGGTG +GCGGCA +GCGGCT +GCGGCC +GCGGCG +GCGGGA +GCGGGT +GCGGGC +GCGGGG +GGAAAA +GGAAAT +GGAAAC +GGAAAG +GGAATA +GGAATT +GGAATC +GGAATG +GGAACA +GGAACT +GGAACC +GGAACG +GGAAGA +GGAAGT +GGAAGC +GGAAGG +GGATAA +GGATAT +GGATAC +GGATAG +GGATTA +GGATTT +GGATTC +GGATTG +GGATCA +GGATCT +GGATCC +GGATCG +GGATGA +GGATGT +GGATGC +GGATGG +GGACAA +GGACAT +GGACAC +GGACAG +GGACTA +GGACTT +GGACTC +GGACTG +GGACCA +GGACCT +GGACCC +GGACCG +GGACGA +GGACGT +GGACGC +GGACGG +GGAGAA +GGAGAT +GGAGAC +GGAGAG +GGAGTA +GGAGTT +GGAGTC +GGAGTG +GGAGCA +GGAGCT +GGAGCC +GGAGCG +GGAGGA +GGAGGT +GGAGGC +GGAGGG +GGTAAA +GGTAAT +GGTAAC +GGTAAG +GGTATA +GGTATT +GGTATC +GGTATG +GGTACA +GGTACT +GGTACC +GGTACG +GGTAGA +GGTAGT +GGTAGC +GGTAGG +GGTTAA +GGTTAT +GGTTAC +GGTTAG +GGTTTA +GGTTTT +GGTTTC +GGTTTG +GGTTCA +GGTTCT +GGTTCC +GGTTCG +GGTTGA +GGTTGT +GGTTGC +GGTTGG +GGTCAA +GGTCAT +GGTCAC +GGTCAG +GGTCTA +GGTCTT +GGTCTC +GGTCTG +GGTCCA +GGTCCT +GGTCCC +GGTCCG +GGTCGA +GGTCGT +GGTCGC +GGTCGG +GGTGAA +GGTGAT +GGTGAC +GGTGAG +GGTGTA +GGTGTT +GGTGTC +GGTGTG +GGTGCA +GGTGCT +GGTGCC +GGTGCG +GGTGGA +GGTGGT +GGTGGC +GGTGGG +GGCAAA +GGCAAT +GGCAAC +GGCAAG +GGCATA +GGCATT +GGCATC +GGCATG +GGCACA +GGCACT +GGCACC +GGCACG +GGCAGA +GGCAGT +GGCAGC +GGCAGG +GGCTAA +GGCTAT +GGCTAC +GGCTAG +GGCTTA +GGCTTT +GGCTTC +GGCTTG +GGCTCA +GGCTCT +GGCTCC +GGCTCG +GGCTGA +GGCTGT +GGCTGC +GGCTGG +GGCCAA +GGCCAT +GGCCAC +GGCCAG +GGCCTA +GGCCTT +GGCCTC +GGCCTG +GGCCCA +GGCCCT +GGCCCC +GGCCCG +GGCCGA +GGCCGT +GGCCGC +GGCCGG +GGCGAA +GGCGAT +GGCGAC +GGCGAG +GGCGTA +GGCGTT +GGCGTC +GGCGTG +GGCGCA +GGCGCT +GGCGCC +GGCGCG +GGCGGA +GGCGGT +GGCGGC +GGCGGG +GGGAAA +GGGAAT +GGGAAC +GGGAAG +GGGATA +GGGATT +GGGATC +GGGATG +GGGACA +GGGACT +GGGACC +GGGACG +GGGAGA +GGGAGT +GGGAGC +GGGAGG +GGGTAA +GGGTAT +GGGTAC +GGGTAG +GGGTTA +GGGTTT +GGGTTC +GGGTTG +GGGTCA +GGGTCT +GGGTCC +GGGTCG +GGGTGA +GGGTGT +GGGTGC +GGGTGG +GGGCAA +GGGCAT +GGGCAC +GGGCAG +GGGCTA +GGGCTT +GGGCTC +GGGCTG +GGGCCA +GGGCCT +GGGCCC +GGGCCG +GGGCGA +GGGCGT +GGGCGC +GGGCGG +GGGGAA +GGGGAT +GGGGAC +GGGGAG +GGGGTA +GGGGTT +GGGGTC +GGGGTG +GGGGCA +GGGGCT +GGGGCC +GGGGCG +GGGGGA +GGGGGT +GGGGGC +GGGGGG diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/README.md b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..500fd976d92b217455f181bc98c77e612d9a18d9 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/README.md @@ -0,0 +1,202 @@ +--- +base_model: fabihamakhdoomi/TinyDNABERT +library_name: peft +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.12.0 \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/adapter_config.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..04c63e99bbc6b810449dece367c0afb165125b92 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/adapter_config.json @@ -0,0 +1,42 @@ +{ + "alpha_pattern": {}, + "auto_mapping": null, + "base_model_name_or_path": "fabihamakhdoomi/TinyDNABERT", + "bias": "none", + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 16, + "lora_dropout": 0.2, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": [ + "classifier", + "score" + ], + "peft_type": "LORA", + "r": 8, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "bert.encoder.layer.1.attention.self.query", + "bert.encoder.layer.0.output.dense", + "bert.encoder.layer.1.output.dense", + "bert.encoder.layer.0.intermediate.dense", + "bert.encoder.layer.1.intermediate.dense", + "bert.encoder.layer.0.attention.output.dense", + "bert.encoder.layer.1.attention.output.dense", + "bert.encoder.layer.0.attention.self.query", + "bert.encoder.layer.0.attention.self.value", + "bert.encoder.layer.1.attention.self.key", + "bert.encoder.layer.1.attention.self.value", + "bert.encoder.layer.0.attention.self.key" + ], + "task_type": "SEQ_CLS", + "use_dora": false, + "use_rslora": false +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/adapter_model.safetensors b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/adapter_model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..4d83d1a2f4fbe349d31b719cf2689f13481e43f3 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/adapter_model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:45c456f162a3426d826c14614c4dbbbe1e8b43b55e74a32b9021a3deae82dae5 +size 151952 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/optimizer.pt b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..ef8edcd352fa506cab5c95ba1411d6e737c1d078 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e10d8b7dfc3acd8d751dfa23647146900482612ef7cdb39f44cbfae80d182a57 +size 312682 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/rng_state.pth b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..d769758d72ed276350b7475a3c6ab31d03d6d464 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:af365fb722ff3640b9c3cbda0c33c89309e74956fafe530249e5f878f489cae3 +size 14244 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/scheduler.pt b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..e21fdaa14c288938c4b2c1277194110b759d8724 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7480edaf7ecb5003556fedbf146cf26ec75a9ee19931ce031bbb3b509f9927ce +size 1064 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/special_tokens_map.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..a8b3208c2884c4efb86e49300fdd3dc877220cdf --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/special_tokens_map.json @@ -0,0 +1,7 @@ +{ + "cls_token": "[CLS]", + "mask_token": "[MASK]", + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/tokenizer.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..ebd782de6963b9032ce9374cfe30f5275be1a30b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/tokenizer.json @@ -0,0 +1,4265 @@ +{ + "version": "1.0", + "truncation": { + "direction": "Right", + "max_length": 512, + "strategy": "LongestFirst", + "stride": 0 + }, + "padding": { + "strategy": { + "Fixed": 512 + }, + "direction": "Right", + "pad_to_multiple_of": null, + "pad_id": 0, + "pad_type_id": 0, + "pad_token": "[PAD]" + }, + "added_tokens": [ + { + "id": 0, + "content": "[PAD]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 1, + "content": "[UNK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 2, + "content": "[CLS]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 3, + "content": "[SEP]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 4, + "content": "[MASK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + } + ], + "normalizer": { + "type": "BertNormalizer", + "clean_text": true, + "handle_chinese_chars": true, + "strip_accents": null, + "lowercase": false + }, + "pre_tokenizer": { + "type": "BertPreTokenizer" + }, + "post_processor": { + "type": "TemplateProcessing", + "single": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + } + ], + "pair": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "B", + "type_id": 1 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 1 + } + } + ], + "special_tokens": { + "[CLS]": { + "id": "[CLS]", + "ids": [ + 2 + ], + "tokens": [ + "[CLS]" + ] + }, + "[SEP]": { + "id": "[SEP]", + "ids": [ + 3 + ], + "tokens": [ + "[SEP]" + ] + } + } + }, + "decoder": { + "type": "WordPiece", + "prefix": "##", + "cleanup": true + }, + "model": { + "type": "WordPiece", + "unk_token": "[UNK]", + "continuing_subword_prefix": "##", + "max_input_chars_per_word": 100, + "vocab": { + "[PAD]": 0, + "[UNK]": 1, + "[CLS]": 2, + "[SEP]": 3, + "[MASK]": 4, + "AAAAAA": 5, + "AAAAAT": 6, + "AAAAAC": 7, + "AAAAAG": 8, + "AAAATA": 9, + "AAAATT": 10, + "AAAATC": 11, + "AAAATG": 12, + "AAAACA": 13, + "AAAACT": 14, + "AAAACC": 15, + "AAAACG": 16, + "AAAAGA": 17, + "AAAAGT": 18, + "AAAAGC": 19, + "AAAAGG": 20, + "AAATAA": 21, + "AAATAT": 22, + "AAATAC": 23, + "AAATAG": 24, + "AAATTA": 25, + "AAATTT": 26, + "AAATTC": 27, + "AAATTG": 28, + "AAATCA": 29, + "AAATCT": 30, + "AAATCC": 31, + "AAATCG": 32, + "AAATGA": 33, + "AAATGT": 34, + "AAATGC": 35, + "AAATGG": 36, + "AAACAA": 37, + "AAACAT": 38, + "AAACAC": 39, + "AAACAG": 40, + "AAACTA": 41, + "AAACTT": 42, + "AAACTC": 43, + "AAACTG": 44, + "AAACCA": 45, + "AAACCT": 46, + "AAACCC": 47, + "AAACCG": 48, + "AAACGA": 49, + "AAACGT": 50, + "AAACGC": 51, + "AAACGG": 52, + "AAAGAA": 53, + "AAAGAT": 54, + "AAAGAC": 55, + "AAAGAG": 56, + "AAAGTA": 57, + "AAAGTT": 58, + "AAAGTC": 59, + "AAAGTG": 60, + "AAAGCA": 61, + "AAAGCT": 62, + "AAAGCC": 63, + "AAAGCG": 64, + "AAAGGA": 65, + "AAAGGT": 66, + "AAAGGC": 67, + "AAAGGG": 68, + "AATAAA": 69, + "AATAAT": 70, + "AATAAC": 71, + "AATAAG": 72, + "AATATA": 73, + "AATATT": 74, + "AATATC": 75, + "AATATG": 76, + "AATACA": 77, + "AATACT": 78, + "AATACC": 79, + "AATACG": 80, + "AATAGA": 81, + "AATAGT": 82, + "AATAGC": 83, + "AATAGG": 84, + "AATTAA": 85, + "AATTAT": 86, + "AATTAC": 87, + "AATTAG": 88, + "AATTTA": 89, + "AATTTT": 90, + "AATTTC": 91, + "AATTTG": 92, + "AATTCA": 93, + "AATTCT": 94, + "AATTCC": 95, + "AATTCG": 96, + "AATTGA": 97, + "AATTGT": 98, + "AATTGC": 99, + "AATTGG": 100, + "AATCAA": 101, + "AATCAT": 102, + "AATCAC": 103, + "AATCAG": 104, + "AATCTA": 105, + "AATCTT": 106, + "AATCTC": 107, + "AATCTG": 108, + "AATCCA": 109, + "AATCCT": 110, + "AATCCC": 111, + "AATCCG": 112, + "AATCGA": 113, + "AATCGT": 114, + "AATCGC": 115, + "AATCGG": 116, + "AATGAA": 117, + "AATGAT": 118, + "AATGAC": 119, + "AATGAG": 120, + "AATGTA": 121, + "AATGTT": 122, + "AATGTC": 123, + "AATGTG": 124, + "AATGCA": 125, + "AATGCT": 126, + "AATGCC": 127, + "AATGCG": 128, + "AATGGA": 129, + "AATGGT": 130, + "AATGGC": 131, + "AATGGG": 132, + "AACAAA": 133, + "AACAAT": 134, + "AACAAC": 135, + "AACAAG": 136, + "AACATA": 137, + "AACATT": 138, + "AACATC": 139, + "AACATG": 140, + "AACACA": 141, + "AACACT": 142, + "AACACC": 143, + "AACACG": 144, + "AACAGA": 145, + "AACAGT": 146, + "AACAGC": 147, + "AACAGG": 148, + "AACTAA": 149, + "AACTAT": 150, + "AACTAC": 151, + "AACTAG": 152, + "AACTTA": 153, + "AACTTT": 154, + "AACTTC": 155, + "AACTTG": 156, + "AACTCA": 157, + "AACTCT": 158, + "AACTCC": 159, + "AACTCG": 160, + "AACTGA": 161, + "AACTGT": 162, + "AACTGC": 163, + "AACTGG": 164, + "AACCAA": 165, + "AACCAT": 166, + "AACCAC": 167, + "AACCAG": 168, + "AACCTA": 169, + "AACCTT": 170, + "AACCTC": 171, + "AACCTG": 172, + "AACCCA": 173, + "AACCCT": 174, + "AACCCC": 175, + "AACCCG": 176, + "AACCGA": 177, + "AACCGT": 178, + "AACCGC": 179, + "AACCGG": 180, + "AACGAA": 181, + "AACGAT": 182, + "AACGAC": 183, + "AACGAG": 184, + "AACGTA": 185, + "AACGTT": 186, + "AACGTC": 187, + "AACGTG": 188, + "AACGCA": 189, + "AACGCT": 190, + "AACGCC": 191, + "AACGCG": 192, + "AACGGA": 193, + "AACGGT": 194, + "AACGGC": 195, + "AACGGG": 196, + "AAGAAA": 197, + "AAGAAT": 198, + "AAGAAC": 199, + "AAGAAG": 200, + "AAGATA": 201, + "AAGATT": 202, + "AAGATC": 203, + "AAGATG": 204, + "AAGACA": 205, + "AAGACT": 206, + "AAGACC": 207, + "AAGACG": 208, + "AAGAGA": 209, + "AAGAGT": 210, + "AAGAGC": 211, + "AAGAGG": 212, + "AAGTAA": 213, + "AAGTAT": 214, + "AAGTAC": 215, + "AAGTAG": 216, + "AAGTTA": 217, + "AAGTTT": 218, + "AAGTTC": 219, + "AAGTTG": 220, + "AAGTCA": 221, + "AAGTCT": 222, + "AAGTCC": 223, + "AAGTCG": 224, + "AAGTGA": 225, + "AAGTGT": 226, + "AAGTGC": 227, + "AAGTGG": 228, + "AAGCAA": 229, + "AAGCAT": 230, + "AAGCAC": 231, + "AAGCAG": 232, + "AAGCTA": 233, + "AAGCTT": 234, + "AAGCTC": 235, + "AAGCTG": 236, + "AAGCCA": 237, + "AAGCCT": 238, + "AAGCCC": 239, + "AAGCCG": 240, + "AAGCGA": 241, + "AAGCGT": 242, + "AAGCGC": 243, + "AAGCGG": 244, + "AAGGAA": 245, + "AAGGAT": 246, + "AAGGAC": 247, + "AAGGAG": 248, + "AAGGTA": 249, + "AAGGTT": 250, + "AAGGTC": 251, + "AAGGTG": 252, + "AAGGCA": 253, + "AAGGCT": 254, + "AAGGCC": 255, + "AAGGCG": 256, + "AAGGGA": 257, + "AAGGGT": 258, + "AAGGGC": 259, + "AAGGGG": 260, + "ATAAAA": 261, + "ATAAAT": 262, + "ATAAAC": 263, + "ATAAAG": 264, + "ATAATA": 265, + "ATAATT": 266, + "ATAATC": 267, + "ATAATG": 268, + "ATAACA": 269, + "ATAACT": 270, + "ATAACC": 271, + "ATAACG": 272, + "ATAAGA": 273, + "ATAAGT": 274, + "ATAAGC": 275, + "ATAAGG": 276, + "ATATAA": 277, + "ATATAT": 278, + "ATATAC": 279, + "ATATAG": 280, + "ATATTA": 281, + "ATATTT": 282, + "ATATTC": 283, + "ATATTG": 284, + "ATATCA": 285, + "ATATCT": 286, + "ATATCC": 287, + "ATATCG": 288, + "ATATGA": 289, + "ATATGT": 290, + "ATATGC": 291, + "ATATGG": 292, + "ATACAA": 293, + "ATACAT": 294, + "ATACAC": 295, + "ATACAG": 296, + "ATACTA": 297, + "ATACTT": 298, + "ATACTC": 299, + "ATACTG": 300, + "ATACCA": 301, + "ATACCT": 302, + "ATACCC": 303, + "ATACCG": 304, + "ATACGA": 305, + "ATACGT": 306, + "ATACGC": 307, + "ATACGG": 308, + "ATAGAA": 309, + "ATAGAT": 310, + "ATAGAC": 311, + "ATAGAG": 312, + "ATAGTA": 313, + "ATAGTT": 314, + "ATAGTC": 315, + "ATAGTG": 316, + "ATAGCA": 317, + "ATAGCT": 318, + "ATAGCC": 319, + "ATAGCG": 320, + "ATAGGA": 321, + "ATAGGT": 322, + "ATAGGC": 323, + "ATAGGG": 324, + "ATTAAA": 325, + "ATTAAT": 326, + "ATTAAC": 327, + "ATTAAG": 328, + "ATTATA": 329, + "ATTATT": 330, + "ATTATC": 331, + "ATTATG": 332, + "ATTACA": 333, + "ATTACT": 334, + "ATTACC": 335, + "ATTACG": 336, + "ATTAGA": 337, + "ATTAGT": 338, + "ATTAGC": 339, + "ATTAGG": 340, + "ATTTAA": 341, + "ATTTAT": 342, + "ATTTAC": 343, + "ATTTAG": 344, + "ATTTTA": 345, + "ATTTTT": 346, + "ATTTTC": 347, + "ATTTTG": 348, + "ATTTCA": 349, + "ATTTCT": 350, + "ATTTCC": 351, + "ATTTCG": 352, + "ATTTGA": 353, + "ATTTGT": 354, + "ATTTGC": 355, + "ATTTGG": 356, + "ATTCAA": 357, + "ATTCAT": 358, + "ATTCAC": 359, + "ATTCAG": 360, + "ATTCTA": 361, + "ATTCTT": 362, + "ATTCTC": 363, + "ATTCTG": 364, + "ATTCCA": 365, + "ATTCCT": 366, + "ATTCCC": 367, + "ATTCCG": 368, + "ATTCGA": 369, + "ATTCGT": 370, + "ATTCGC": 371, + "ATTCGG": 372, + "ATTGAA": 373, + "ATTGAT": 374, + "ATTGAC": 375, + "ATTGAG": 376, + "ATTGTA": 377, + "ATTGTT": 378, + "ATTGTC": 379, + "ATTGTG": 380, + "ATTGCA": 381, + "ATTGCT": 382, + "ATTGCC": 383, + "ATTGCG": 384, + "ATTGGA": 385, + "ATTGGT": 386, + "ATTGGC": 387, + "ATTGGG": 388, + "ATCAAA": 389, + "ATCAAT": 390, + "ATCAAC": 391, + "ATCAAG": 392, + "ATCATA": 393, + "ATCATT": 394, + "ATCATC": 395, + "ATCATG": 396, + "ATCACA": 397, + "ATCACT": 398, + "ATCACC": 399, + "ATCACG": 400, + "ATCAGA": 401, + "ATCAGT": 402, + "ATCAGC": 403, + "ATCAGG": 404, + "ATCTAA": 405, + "ATCTAT": 406, + "ATCTAC": 407, + "ATCTAG": 408, + "ATCTTA": 409, + "ATCTTT": 410, + "ATCTTC": 411, + "ATCTTG": 412, + "ATCTCA": 413, + "ATCTCT": 414, + "ATCTCC": 415, + "ATCTCG": 416, + "ATCTGA": 417, + "ATCTGT": 418, + "ATCTGC": 419, + "ATCTGG": 420, + "ATCCAA": 421, + "ATCCAT": 422, + "ATCCAC": 423, + "ATCCAG": 424, + "ATCCTA": 425, + "ATCCTT": 426, + "ATCCTC": 427, + "ATCCTG": 428, + "ATCCCA": 429, + "ATCCCT": 430, + "ATCCCC": 431, + "ATCCCG": 432, + "ATCCGA": 433, + "ATCCGT": 434, + "ATCCGC": 435, + "ATCCGG": 436, + "ATCGAA": 437, + "ATCGAT": 438, + "ATCGAC": 439, + "ATCGAG": 440, + "ATCGTA": 441, + "ATCGTT": 442, + "ATCGTC": 443, + "ATCGTG": 444, + "ATCGCA": 445, + "ATCGCT": 446, + "ATCGCC": 447, + "ATCGCG": 448, + "ATCGGA": 449, + "ATCGGT": 450, + "ATCGGC": 451, + "ATCGGG": 452, + "ATGAAA": 453, + "ATGAAT": 454, + "ATGAAC": 455, + "ATGAAG": 456, + "ATGATA": 457, + "ATGATT": 458, + "ATGATC": 459, + "ATGATG": 460, + "ATGACA": 461, + "ATGACT": 462, + "ATGACC": 463, + "ATGACG": 464, + "ATGAGA": 465, + "ATGAGT": 466, + "ATGAGC": 467, + "ATGAGG": 468, + "ATGTAA": 469, + "ATGTAT": 470, + "ATGTAC": 471, + "ATGTAG": 472, + "ATGTTA": 473, + "ATGTTT": 474, + "ATGTTC": 475, + "ATGTTG": 476, + "ATGTCA": 477, + "ATGTCT": 478, + "ATGTCC": 479, + "ATGTCG": 480, + "ATGTGA": 481, + "ATGTGT": 482, + "ATGTGC": 483, + "ATGTGG": 484, + "ATGCAA": 485, + "ATGCAT": 486, + "ATGCAC": 487, + "ATGCAG": 488, + "ATGCTA": 489, + "ATGCTT": 490, + "ATGCTC": 491, + "ATGCTG": 492, + "ATGCCA": 493, + "ATGCCT": 494, + "ATGCCC": 495, + "ATGCCG": 496, + "ATGCGA": 497, + "ATGCGT": 498, + "ATGCGC": 499, + "ATGCGG": 500, + "ATGGAA": 501, + "ATGGAT": 502, + "ATGGAC": 503, + "ATGGAG": 504, + "ATGGTA": 505, + "ATGGTT": 506, + "ATGGTC": 507, + "ATGGTG": 508, + "ATGGCA": 509, + "ATGGCT": 510, + "ATGGCC": 511, + "ATGGCG": 512, + "ATGGGA": 513, + "ATGGGT": 514, + "ATGGGC": 515, + "ATGGGG": 516, + "ACAAAA": 517, + "ACAAAT": 518, + "ACAAAC": 519, + "ACAAAG": 520, + "ACAATA": 521, + "ACAATT": 522, + "ACAATC": 523, + "ACAATG": 524, + "ACAACA": 525, + "ACAACT": 526, + "ACAACC": 527, + "ACAACG": 528, + "ACAAGA": 529, + "ACAAGT": 530, + "ACAAGC": 531, + "ACAAGG": 532, + "ACATAA": 533, + "ACATAT": 534, + "ACATAC": 535, + "ACATAG": 536, + "ACATTA": 537, + "ACATTT": 538, + "ACATTC": 539, + "ACATTG": 540, + "ACATCA": 541, + "ACATCT": 542, + "ACATCC": 543, + "ACATCG": 544, + "ACATGA": 545, + "ACATGT": 546, + "ACATGC": 547, + "ACATGG": 548, + "ACACAA": 549, + "ACACAT": 550, + "ACACAC": 551, + "ACACAG": 552, + "ACACTA": 553, + "ACACTT": 554, + "ACACTC": 555, + "ACACTG": 556, + "ACACCA": 557, + "ACACCT": 558, + "ACACCC": 559, + "ACACCG": 560, + "ACACGA": 561, + "ACACGT": 562, + "ACACGC": 563, + "ACACGG": 564, + "ACAGAA": 565, + "ACAGAT": 566, + "ACAGAC": 567, + "ACAGAG": 568, + "ACAGTA": 569, + "ACAGTT": 570, + "ACAGTC": 571, + "ACAGTG": 572, + "ACAGCA": 573, + "ACAGCT": 574, + "ACAGCC": 575, + "ACAGCG": 576, + "ACAGGA": 577, + "ACAGGT": 578, + "ACAGGC": 579, + "ACAGGG": 580, + "ACTAAA": 581, + "ACTAAT": 582, + "ACTAAC": 583, + "ACTAAG": 584, + "ACTATA": 585, + "ACTATT": 586, + "ACTATC": 587, + "ACTATG": 588, + "ACTACA": 589, + "ACTACT": 590, + "ACTACC": 591, + "ACTACG": 592, + "ACTAGA": 593, + "ACTAGT": 594, + "ACTAGC": 595, + "ACTAGG": 596, + "ACTTAA": 597, + "ACTTAT": 598, + "ACTTAC": 599, + "ACTTAG": 600, + "ACTTTA": 601, + "ACTTTT": 602, + "ACTTTC": 603, + "ACTTTG": 604, + "ACTTCA": 605, + "ACTTCT": 606, + "ACTTCC": 607, + "ACTTCG": 608, + "ACTTGA": 609, + "ACTTGT": 610, + "ACTTGC": 611, + "ACTTGG": 612, + "ACTCAA": 613, + "ACTCAT": 614, + "ACTCAC": 615, + "ACTCAG": 616, + "ACTCTA": 617, + "ACTCTT": 618, + "ACTCTC": 619, + "ACTCTG": 620, + "ACTCCA": 621, + "ACTCCT": 622, + "ACTCCC": 623, + "ACTCCG": 624, + "ACTCGA": 625, + "ACTCGT": 626, + "ACTCGC": 627, + "ACTCGG": 628, + "ACTGAA": 629, + "ACTGAT": 630, + "ACTGAC": 631, + "ACTGAG": 632, + "ACTGTA": 633, + "ACTGTT": 634, + "ACTGTC": 635, + "ACTGTG": 636, + "ACTGCA": 637, + "ACTGCT": 638, + "ACTGCC": 639, + "ACTGCG": 640, + "ACTGGA": 641, + "ACTGGT": 642, + "ACTGGC": 643, + "ACTGGG": 644, + "ACCAAA": 645, + "ACCAAT": 646, + "ACCAAC": 647, + "ACCAAG": 648, + "ACCATA": 649, + "ACCATT": 650, + "ACCATC": 651, + "ACCATG": 652, + "ACCACA": 653, + "ACCACT": 654, + "ACCACC": 655, + "ACCACG": 656, + "ACCAGA": 657, + "ACCAGT": 658, + "ACCAGC": 659, + "ACCAGG": 660, + "ACCTAA": 661, + "ACCTAT": 662, + "ACCTAC": 663, + "ACCTAG": 664, + "ACCTTA": 665, + "ACCTTT": 666, + "ACCTTC": 667, + "ACCTTG": 668, + "ACCTCA": 669, + "ACCTCT": 670, + "ACCTCC": 671, + "ACCTCG": 672, + "ACCTGA": 673, + "ACCTGT": 674, + "ACCTGC": 675, + "ACCTGG": 676, + "ACCCAA": 677, + "ACCCAT": 678, + "ACCCAC": 679, + "ACCCAG": 680, + "ACCCTA": 681, + "ACCCTT": 682, + "ACCCTC": 683, + "ACCCTG": 684, + "ACCCCA": 685, + "ACCCCT": 686, + "ACCCCC": 687, + "ACCCCG": 688, + "ACCCGA": 689, + "ACCCGT": 690, + "ACCCGC": 691, + "ACCCGG": 692, + "ACCGAA": 693, + "ACCGAT": 694, + "ACCGAC": 695, + "ACCGAG": 696, + "ACCGTA": 697, + "ACCGTT": 698, + "ACCGTC": 699, + "ACCGTG": 700, + "ACCGCA": 701, + "ACCGCT": 702, + "ACCGCC": 703, + "ACCGCG": 704, + "ACCGGA": 705, + "ACCGGT": 706, + "ACCGGC": 707, + "ACCGGG": 708, + "ACGAAA": 709, + "ACGAAT": 710, + "ACGAAC": 711, + "ACGAAG": 712, + "ACGATA": 713, + "ACGATT": 714, + "ACGATC": 715, + "ACGATG": 716, + "ACGACA": 717, + "ACGACT": 718, + "ACGACC": 719, + "ACGACG": 720, + "ACGAGA": 721, + "ACGAGT": 722, + "ACGAGC": 723, + "ACGAGG": 724, + "ACGTAA": 725, + "ACGTAT": 726, + "ACGTAC": 727, + "ACGTAG": 728, + "ACGTTA": 729, + "ACGTTT": 730, + "ACGTTC": 731, + "ACGTTG": 732, + "ACGTCA": 733, + "ACGTCT": 734, + "ACGTCC": 735, + "ACGTCG": 736, + "ACGTGA": 737, + "ACGTGT": 738, + "ACGTGC": 739, + "ACGTGG": 740, + "ACGCAA": 741, + "ACGCAT": 742, + "ACGCAC": 743, + "ACGCAG": 744, + "ACGCTA": 745, + "ACGCTT": 746, + "ACGCTC": 747, + "ACGCTG": 748, + "ACGCCA": 749, + "ACGCCT": 750, + "ACGCCC": 751, + "ACGCCG": 752, + "ACGCGA": 753, + "ACGCGT": 754, + "ACGCGC": 755, + "ACGCGG": 756, + "ACGGAA": 757, + "ACGGAT": 758, + "ACGGAC": 759, + "ACGGAG": 760, + "ACGGTA": 761, + "ACGGTT": 762, + "ACGGTC": 763, + "ACGGTG": 764, + "ACGGCA": 765, + "ACGGCT": 766, + "ACGGCC": 767, + "ACGGCG": 768, + "ACGGGA": 769, + "ACGGGT": 770, + "ACGGGC": 771, + "ACGGGG": 772, + "AGAAAA": 773, + "AGAAAT": 774, + "AGAAAC": 775, + "AGAAAG": 776, + "AGAATA": 777, + "AGAATT": 778, + "AGAATC": 779, + "AGAATG": 780, + "AGAACA": 781, + "AGAACT": 782, + "AGAACC": 783, + "AGAACG": 784, + "AGAAGA": 785, + "AGAAGT": 786, + "AGAAGC": 787, + "AGAAGG": 788, + "AGATAA": 789, + "AGATAT": 790, + "AGATAC": 791, + "AGATAG": 792, + "AGATTA": 793, + "AGATTT": 794, + "AGATTC": 795, + "AGATTG": 796, + "AGATCA": 797, + "AGATCT": 798, + "AGATCC": 799, + "AGATCG": 800, + "AGATGA": 801, + "AGATGT": 802, + "AGATGC": 803, + "AGATGG": 804, + "AGACAA": 805, + "AGACAT": 806, + "AGACAC": 807, + "AGACAG": 808, + "AGACTA": 809, + "AGACTT": 810, + "AGACTC": 811, + "AGACTG": 812, + "AGACCA": 813, + "AGACCT": 814, + "AGACCC": 815, + "AGACCG": 816, + "AGACGA": 817, + "AGACGT": 818, + "AGACGC": 819, + "AGACGG": 820, + "AGAGAA": 821, + "AGAGAT": 822, + "AGAGAC": 823, + "AGAGAG": 824, + "AGAGTA": 825, + "AGAGTT": 826, + "AGAGTC": 827, + "AGAGTG": 828, + "AGAGCA": 829, + "AGAGCT": 830, + "AGAGCC": 831, + "AGAGCG": 832, + "AGAGGA": 833, + "AGAGGT": 834, + "AGAGGC": 835, + "AGAGGG": 836, + "AGTAAA": 837, + "AGTAAT": 838, + "AGTAAC": 839, + "AGTAAG": 840, + "AGTATA": 841, + "AGTATT": 842, + "AGTATC": 843, + "AGTATG": 844, + "AGTACA": 845, + "AGTACT": 846, + "AGTACC": 847, + "AGTACG": 848, + "AGTAGA": 849, + "AGTAGT": 850, + "AGTAGC": 851, + "AGTAGG": 852, + "AGTTAA": 853, + "AGTTAT": 854, + "AGTTAC": 855, + "AGTTAG": 856, + "AGTTTA": 857, + "AGTTTT": 858, + "AGTTTC": 859, + "AGTTTG": 860, + "AGTTCA": 861, + "AGTTCT": 862, + "AGTTCC": 863, + "AGTTCG": 864, + "AGTTGA": 865, + "AGTTGT": 866, + "AGTTGC": 867, + "AGTTGG": 868, + "AGTCAA": 869, + "AGTCAT": 870, + "AGTCAC": 871, + "AGTCAG": 872, + "AGTCTA": 873, + "AGTCTT": 874, + "AGTCTC": 875, + "AGTCTG": 876, + "AGTCCA": 877, + "AGTCCT": 878, + "AGTCCC": 879, + "AGTCCG": 880, + "AGTCGA": 881, + "AGTCGT": 882, + "AGTCGC": 883, + "AGTCGG": 884, + "AGTGAA": 885, + "AGTGAT": 886, + "AGTGAC": 887, + "AGTGAG": 888, + "AGTGTA": 889, + "AGTGTT": 890, + "AGTGTC": 891, + "AGTGTG": 892, + "AGTGCA": 893, + "AGTGCT": 894, + "AGTGCC": 895, + "AGTGCG": 896, + "AGTGGA": 897, + "AGTGGT": 898, + "AGTGGC": 899, + "AGTGGG": 900, + "AGCAAA": 901, + "AGCAAT": 902, + "AGCAAC": 903, + "AGCAAG": 904, + "AGCATA": 905, + "AGCATT": 906, + "AGCATC": 907, + "AGCATG": 908, + "AGCACA": 909, + "AGCACT": 910, + "AGCACC": 911, + "AGCACG": 912, + "AGCAGA": 913, + "AGCAGT": 914, + "AGCAGC": 915, + "AGCAGG": 916, + "AGCTAA": 917, + "AGCTAT": 918, + "AGCTAC": 919, + "AGCTAG": 920, + "AGCTTA": 921, + "AGCTTT": 922, + "AGCTTC": 923, + "AGCTTG": 924, + "AGCTCA": 925, + "AGCTCT": 926, + "AGCTCC": 927, + "AGCTCG": 928, + "AGCTGA": 929, + "AGCTGT": 930, + "AGCTGC": 931, + "AGCTGG": 932, + "AGCCAA": 933, + "AGCCAT": 934, + "AGCCAC": 935, + "AGCCAG": 936, + "AGCCTA": 937, + "AGCCTT": 938, + "AGCCTC": 939, + "AGCCTG": 940, + "AGCCCA": 941, + "AGCCCT": 942, + "AGCCCC": 943, + "AGCCCG": 944, + "AGCCGA": 945, + "AGCCGT": 946, + "AGCCGC": 947, + "AGCCGG": 948, + "AGCGAA": 949, + "AGCGAT": 950, + "AGCGAC": 951, + "AGCGAG": 952, + "AGCGTA": 953, + "AGCGTT": 954, + "AGCGTC": 955, + "AGCGTG": 956, + "AGCGCA": 957, + "AGCGCT": 958, + "AGCGCC": 959, + "AGCGCG": 960, + "AGCGGA": 961, + "AGCGGT": 962, + "AGCGGC": 963, + "AGCGGG": 964, + "AGGAAA": 965, + "AGGAAT": 966, + "AGGAAC": 967, + "AGGAAG": 968, + "AGGATA": 969, + "AGGATT": 970, + "AGGATC": 971, + "AGGATG": 972, + "AGGACA": 973, + "AGGACT": 974, + "AGGACC": 975, + "AGGACG": 976, + "AGGAGA": 977, + "AGGAGT": 978, + "AGGAGC": 979, + "AGGAGG": 980, + "AGGTAA": 981, + "AGGTAT": 982, + "AGGTAC": 983, + "AGGTAG": 984, + "AGGTTA": 985, + "AGGTTT": 986, + "AGGTTC": 987, + "AGGTTG": 988, + "AGGTCA": 989, + "AGGTCT": 990, + "AGGTCC": 991, + "AGGTCG": 992, + "AGGTGA": 993, + "AGGTGT": 994, + "AGGTGC": 995, + "AGGTGG": 996, + "AGGCAA": 997, + "AGGCAT": 998, + "AGGCAC": 999, + "AGGCAG": 1000, + "AGGCTA": 1001, + "AGGCTT": 1002, + "AGGCTC": 1003, + "AGGCTG": 1004, + "AGGCCA": 1005, + "AGGCCT": 1006, + "AGGCCC": 1007, + "AGGCCG": 1008, + "AGGCGA": 1009, + "AGGCGT": 1010, + "AGGCGC": 1011, + "AGGCGG": 1012, + "AGGGAA": 1013, + "AGGGAT": 1014, + "AGGGAC": 1015, + "AGGGAG": 1016, + "AGGGTA": 1017, + "AGGGTT": 1018, + "AGGGTC": 1019, + "AGGGTG": 1020, + "AGGGCA": 1021, + "AGGGCT": 1022, + "AGGGCC": 1023, + "AGGGCG": 1024, + "AGGGGA": 1025, + "AGGGGT": 1026, + "AGGGGC": 1027, + "AGGGGG": 1028, + "TAAAAA": 1029, + "TAAAAT": 1030, + "TAAAAC": 1031, + "TAAAAG": 1032, + "TAAATA": 1033, + "TAAATT": 1034, + "TAAATC": 1035, + "TAAATG": 1036, + "TAAACA": 1037, + "TAAACT": 1038, + "TAAACC": 1039, + "TAAACG": 1040, + "TAAAGA": 1041, + "TAAAGT": 1042, + "TAAAGC": 1043, + "TAAAGG": 1044, + "TAATAA": 1045, + "TAATAT": 1046, + "TAATAC": 1047, + "TAATAG": 1048, + "TAATTA": 1049, + "TAATTT": 1050, + "TAATTC": 1051, + "TAATTG": 1052, + "TAATCA": 1053, + "TAATCT": 1054, + "TAATCC": 1055, + "TAATCG": 1056, + "TAATGA": 1057, + "TAATGT": 1058, + "TAATGC": 1059, + "TAATGG": 1060, + "TAACAA": 1061, + "TAACAT": 1062, + "TAACAC": 1063, + "TAACAG": 1064, + "TAACTA": 1065, + "TAACTT": 1066, + "TAACTC": 1067, + "TAACTG": 1068, + "TAACCA": 1069, + "TAACCT": 1070, + "TAACCC": 1071, + "TAACCG": 1072, + "TAACGA": 1073, + "TAACGT": 1074, + "TAACGC": 1075, + "TAACGG": 1076, + "TAAGAA": 1077, + "TAAGAT": 1078, + "TAAGAC": 1079, + "TAAGAG": 1080, + "TAAGTA": 1081, + "TAAGTT": 1082, + "TAAGTC": 1083, + "TAAGTG": 1084, + "TAAGCA": 1085, + "TAAGCT": 1086, + "TAAGCC": 1087, + "TAAGCG": 1088, + "TAAGGA": 1089, + "TAAGGT": 1090, + "TAAGGC": 1091, + "TAAGGG": 1092, + "TATAAA": 1093, + "TATAAT": 1094, + "TATAAC": 1095, + "TATAAG": 1096, + "TATATA": 1097, + "TATATT": 1098, + "TATATC": 1099, + "TATATG": 1100, + "TATACA": 1101, + "TATACT": 1102, + "TATACC": 1103, + "TATACG": 1104, + "TATAGA": 1105, + "TATAGT": 1106, + "TATAGC": 1107, + "TATAGG": 1108, + "TATTAA": 1109, + "TATTAT": 1110, + "TATTAC": 1111, + "TATTAG": 1112, + "TATTTA": 1113, + "TATTTT": 1114, + "TATTTC": 1115, + "TATTTG": 1116, + "TATTCA": 1117, + "TATTCT": 1118, + "TATTCC": 1119, + "TATTCG": 1120, + "TATTGA": 1121, + "TATTGT": 1122, + "TATTGC": 1123, + "TATTGG": 1124, + "TATCAA": 1125, + "TATCAT": 1126, + "TATCAC": 1127, + "TATCAG": 1128, + "TATCTA": 1129, + "TATCTT": 1130, + "TATCTC": 1131, + "TATCTG": 1132, + "TATCCA": 1133, + "TATCCT": 1134, + "TATCCC": 1135, + "TATCCG": 1136, + "TATCGA": 1137, + "TATCGT": 1138, + "TATCGC": 1139, + "TATCGG": 1140, + "TATGAA": 1141, + "TATGAT": 1142, + "TATGAC": 1143, + "TATGAG": 1144, + "TATGTA": 1145, + "TATGTT": 1146, + "TATGTC": 1147, + "TATGTG": 1148, + "TATGCA": 1149, + "TATGCT": 1150, + "TATGCC": 1151, + "TATGCG": 1152, + "TATGGA": 1153, + "TATGGT": 1154, + "TATGGC": 1155, + "TATGGG": 1156, + "TACAAA": 1157, + "TACAAT": 1158, + "TACAAC": 1159, + "TACAAG": 1160, + "TACATA": 1161, + "TACATT": 1162, + "TACATC": 1163, + "TACATG": 1164, + "TACACA": 1165, + "TACACT": 1166, + "TACACC": 1167, + "TACACG": 1168, + "TACAGA": 1169, + "TACAGT": 1170, + "TACAGC": 1171, + "TACAGG": 1172, + "TACTAA": 1173, + "TACTAT": 1174, + "TACTAC": 1175, + "TACTAG": 1176, + "TACTTA": 1177, + "TACTTT": 1178, + "TACTTC": 1179, + "TACTTG": 1180, + "TACTCA": 1181, + "TACTCT": 1182, + "TACTCC": 1183, + "TACTCG": 1184, + "TACTGA": 1185, + "TACTGT": 1186, + "TACTGC": 1187, + "TACTGG": 1188, + "TACCAA": 1189, + "TACCAT": 1190, + "TACCAC": 1191, + "TACCAG": 1192, + "TACCTA": 1193, + "TACCTT": 1194, + "TACCTC": 1195, + "TACCTG": 1196, + "TACCCA": 1197, + "TACCCT": 1198, + "TACCCC": 1199, + "TACCCG": 1200, + "TACCGA": 1201, + "TACCGT": 1202, + "TACCGC": 1203, + "TACCGG": 1204, + "TACGAA": 1205, + "TACGAT": 1206, + "TACGAC": 1207, + "TACGAG": 1208, + "TACGTA": 1209, + "TACGTT": 1210, + "TACGTC": 1211, + "TACGTG": 1212, + "TACGCA": 1213, + "TACGCT": 1214, + "TACGCC": 1215, + "TACGCG": 1216, + "TACGGA": 1217, + "TACGGT": 1218, + "TACGGC": 1219, + "TACGGG": 1220, + "TAGAAA": 1221, + "TAGAAT": 1222, + "TAGAAC": 1223, + "TAGAAG": 1224, + "TAGATA": 1225, + "TAGATT": 1226, + "TAGATC": 1227, + "TAGATG": 1228, + "TAGACA": 1229, + "TAGACT": 1230, + "TAGACC": 1231, + "TAGACG": 1232, + "TAGAGA": 1233, + "TAGAGT": 1234, + "TAGAGC": 1235, + "TAGAGG": 1236, + "TAGTAA": 1237, + "TAGTAT": 1238, + "TAGTAC": 1239, + "TAGTAG": 1240, + "TAGTTA": 1241, + "TAGTTT": 1242, + "TAGTTC": 1243, + "TAGTTG": 1244, + "TAGTCA": 1245, + "TAGTCT": 1246, + "TAGTCC": 1247, + "TAGTCG": 1248, + "TAGTGA": 1249, + "TAGTGT": 1250, + "TAGTGC": 1251, + "TAGTGG": 1252, + "TAGCAA": 1253, + "TAGCAT": 1254, + "TAGCAC": 1255, + "TAGCAG": 1256, + "TAGCTA": 1257, + "TAGCTT": 1258, + "TAGCTC": 1259, + "TAGCTG": 1260, + "TAGCCA": 1261, + "TAGCCT": 1262, + "TAGCCC": 1263, + "TAGCCG": 1264, + "TAGCGA": 1265, + "TAGCGT": 1266, + "TAGCGC": 1267, + "TAGCGG": 1268, + "TAGGAA": 1269, + "TAGGAT": 1270, + "TAGGAC": 1271, + "TAGGAG": 1272, + "TAGGTA": 1273, + "TAGGTT": 1274, + "TAGGTC": 1275, + "TAGGTG": 1276, + "TAGGCA": 1277, + "TAGGCT": 1278, + "TAGGCC": 1279, + "TAGGCG": 1280, + "TAGGGA": 1281, + "TAGGGT": 1282, + "TAGGGC": 1283, + "TAGGGG": 1284, + "TTAAAA": 1285, + "TTAAAT": 1286, + "TTAAAC": 1287, + "TTAAAG": 1288, + "TTAATA": 1289, + "TTAATT": 1290, + "TTAATC": 1291, + "TTAATG": 1292, + "TTAACA": 1293, + "TTAACT": 1294, + "TTAACC": 1295, + "TTAACG": 1296, + "TTAAGA": 1297, + "TTAAGT": 1298, + "TTAAGC": 1299, + "TTAAGG": 1300, + "TTATAA": 1301, + "TTATAT": 1302, + "TTATAC": 1303, + "TTATAG": 1304, + "TTATTA": 1305, + "TTATTT": 1306, + "TTATTC": 1307, + "TTATTG": 1308, + "TTATCA": 1309, + "TTATCT": 1310, + "TTATCC": 1311, + "TTATCG": 1312, + "TTATGA": 1313, + "TTATGT": 1314, + "TTATGC": 1315, + "TTATGG": 1316, + "TTACAA": 1317, + "TTACAT": 1318, + "TTACAC": 1319, + "TTACAG": 1320, + "TTACTA": 1321, + "TTACTT": 1322, + "TTACTC": 1323, + "TTACTG": 1324, + "TTACCA": 1325, + "TTACCT": 1326, + "TTACCC": 1327, + "TTACCG": 1328, + "TTACGA": 1329, + "TTACGT": 1330, + "TTACGC": 1331, + "TTACGG": 1332, + "TTAGAA": 1333, + "TTAGAT": 1334, + "TTAGAC": 1335, + "TTAGAG": 1336, + "TTAGTA": 1337, + "TTAGTT": 1338, + "TTAGTC": 1339, + "TTAGTG": 1340, + "TTAGCA": 1341, + "TTAGCT": 1342, + "TTAGCC": 1343, + "TTAGCG": 1344, + "TTAGGA": 1345, + "TTAGGT": 1346, + "TTAGGC": 1347, + "TTAGGG": 1348, + "TTTAAA": 1349, + "TTTAAT": 1350, + "TTTAAC": 1351, + "TTTAAG": 1352, + "TTTATA": 1353, + "TTTATT": 1354, + "TTTATC": 1355, + "TTTATG": 1356, + "TTTACA": 1357, + "TTTACT": 1358, + "TTTACC": 1359, + "TTTACG": 1360, + "TTTAGA": 1361, + "TTTAGT": 1362, + "TTTAGC": 1363, + "TTTAGG": 1364, + "TTTTAA": 1365, + "TTTTAT": 1366, + "TTTTAC": 1367, + "TTTTAG": 1368, + "TTTTTA": 1369, + "TTTTTT": 1370, + "TTTTTC": 1371, + "TTTTTG": 1372, + "TTTTCA": 1373, + "TTTTCT": 1374, + "TTTTCC": 1375, + "TTTTCG": 1376, + "TTTTGA": 1377, + "TTTTGT": 1378, + "TTTTGC": 1379, + "TTTTGG": 1380, + "TTTCAA": 1381, + "TTTCAT": 1382, + "TTTCAC": 1383, + "TTTCAG": 1384, + "TTTCTA": 1385, + "TTTCTT": 1386, + "TTTCTC": 1387, + "TTTCTG": 1388, + "TTTCCA": 1389, + "TTTCCT": 1390, + "TTTCCC": 1391, + "TTTCCG": 1392, + "TTTCGA": 1393, + "TTTCGT": 1394, + "TTTCGC": 1395, + "TTTCGG": 1396, + "TTTGAA": 1397, + "TTTGAT": 1398, + "TTTGAC": 1399, + "TTTGAG": 1400, + "TTTGTA": 1401, + "TTTGTT": 1402, + "TTTGTC": 1403, + "TTTGTG": 1404, + "TTTGCA": 1405, + "TTTGCT": 1406, + "TTTGCC": 1407, + "TTTGCG": 1408, + "TTTGGA": 1409, + "TTTGGT": 1410, + "TTTGGC": 1411, + "TTTGGG": 1412, + "TTCAAA": 1413, + "TTCAAT": 1414, + "TTCAAC": 1415, + "TTCAAG": 1416, + "TTCATA": 1417, + "TTCATT": 1418, + "TTCATC": 1419, + "TTCATG": 1420, + "TTCACA": 1421, + "TTCACT": 1422, + "TTCACC": 1423, + "TTCACG": 1424, + "TTCAGA": 1425, + "TTCAGT": 1426, + "TTCAGC": 1427, + "TTCAGG": 1428, + "TTCTAA": 1429, + "TTCTAT": 1430, + "TTCTAC": 1431, + "TTCTAG": 1432, + "TTCTTA": 1433, + "TTCTTT": 1434, + "TTCTTC": 1435, + "TTCTTG": 1436, + "TTCTCA": 1437, + "TTCTCT": 1438, + "TTCTCC": 1439, + "TTCTCG": 1440, + "TTCTGA": 1441, + "TTCTGT": 1442, + "TTCTGC": 1443, + "TTCTGG": 1444, + "TTCCAA": 1445, + "TTCCAT": 1446, + "TTCCAC": 1447, + "TTCCAG": 1448, + "TTCCTA": 1449, + "TTCCTT": 1450, + "TTCCTC": 1451, + "TTCCTG": 1452, + "TTCCCA": 1453, + "TTCCCT": 1454, + "TTCCCC": 1455, + "TTCCCG": 1456, + "TTCCGA": 1457, + "TTCCGT": 1458, + "TTCCGC": 1459, + "TTCCGG": 1460, + "TTCGAA": 1461, + "TTCGAT": 1462, + "TTCGAC": 1463, + "TTCGAG": 1464, + "TTCGTA": 1465, + "TTCGTT": 1466, + "TTCGTC": 1467, + "TTCGTG": 1468, + "TTCGCA": 1469, + "TTCGCT": 1470, + "TTCGCC": 1471, + "TTCGCG": 1472, + "TTCGGA": 1473, + "TTCGGT": 1474, + "TTCGGC": 1475, + "TTCGGG": 1476, + "TTGAAA": 1477, + "TTGAAT": 1478, + "TTGAAC": 1479, + "TTGAAG": 1480, + "TTGATA": 1481, + "TTGATT": 1482, + "TTGATC": 1483, + "TTGATG": 1484, + "TTGACA": 1485, + "TTGACT": 1486, + "TTGACC": 1487, + "TTGACG": 1488, + "TTGAGA": 1489, + "TTGAGT": 1490, + "TTGAGC": 1491, + "TTGAGG": 1492, + "TTGTAA": 1493, + "TTGTAT": 1494, + "TTGTAC": 1495, + "TTGTAG": 1496, + "TTGTTA": 1497, + "TTGTTT": 1498, + "TTGTTC": 1499, + "TTGTTG": 1500, + "TTGTCA": 1501, + "TTGTCT": 1502, + "TTGTCC": 1503, + "TTGTCG": 1504, + "TTGTGA": 1505, + "TTGTGT": 1506, + "TTGTGC": 1507, + "TTGTGG": 1508, + "TTGCAA": 1509, + "TTGCAT": 1510, + "TTGCAC": 1511, + "TTGCAG": 1512, + "TTGCTA": 1513, + "TTGCTT": 1514, + "TTGCTC": 1515, + "TTGCTG": 1516, + "TTGCCA": 1517, + "TTGCCT": 1518, + "TTGCCC": 1519, + "TTGCCG": 1520, + "TTGCGA": 1521, + "TTGCGT": 1522, + "TTGCGC": 1523, + "TTGCGG": 1524, + "TTGGAA": 1525, + "TTGGAT": 1526, + "TTGGAC": 1527, + "TTGGAG": 1528, + "TTGGTA": 1529, + "TTGGTT": 1530, + "TTGGTC": 1531, + "TTGGTG": 1532, + "TTGGCA": 1533, + "TTGGCT": 1534, + "TTGGCC": 1535, + "TTGGCG": 1536, + "TTGGGA": 1537, + "TTGGGT": 1538, + "TTGGGC": 1539, + "TTGGGG": 1540, + "TCAAAA": 1541, + "TCAAAT": 1542, + "TCAAAC": 1543, + "TCAAAG": 1544, + "TCAATA": 1545, + "TCAATT": 1546, + "TCAATC": 1547, + "TCAATG": 1548, + "TCAACA": 1549, + "TCAACT": 1550, + "TCAACC": 1551, + "TCAACG": 1552, + "TCAAGA": 1553, + "TCAAGT": 1554, + "TCAAGC": 1555, + "TCAAGG": 1556, + "TCATAA": 1557, + "TCATAT": 1558, + "TCATAC": 1559, + "TCATAG": 1560, + "TCATTA": 1561, + "TCATTT": 1562, + "TCATTC": 1563, + "TCATTG": 1564, + "TCATCA": 1565, + "TCATCT": 1566, + "TCATCC": 1567, + "TCATCG": 1568, + "TCATGA": 1569, + "TCATGT": 1570, + "TCATGC": 1571, + "TCATGG": 1572, + "TCACAA": 1573, + "TCACAT": 1574, + "TCACAC": 1575, + "TCACAG": 1576, + "TCACTA": 1577, + "TCACTT": 1578, + "TCACTC": 1579, + "TCACTG": 1580, + "TCACCA": 1581, + "TCACCT": 1582, + "TCACCC": 1583, + "TCACCG": 1584, + "TCACGA": 1585, + "TCACGT": 1586, + "TCACGC": 1587, + "TCACGG": 1588, + "TCAGAA": 1589, + "TCAGAT": 1590, + "TCAGAC": 1591, + "TCAGAG": 1592, + "TCAGTA": 1593, + "TCAGTT": 1594, + "TCAGTC": 1595, + "TCAGTG": 1596, + "TCAGCA": 1597, + "TCAGCT": 1598, + "TCAGCC": 1599, + "TCAGCG": 1600, + "TCAGGA": 1601, + "TCAGGT": 1602, + "TCAGGC": 1603, + "TCAGGG": 1604, + "TCTAAA": 1605, + "TCTAAT": 1606, + "TCTAAC": 1607, + "TCTAAG": 1608, + "TCTATA": 1609, + "TCTATT": 1610, + "TCTATC": 1611, + "TCTATG": 1612, + "TCTACA": 1613, + "TCTACT": 1614, + "TCTACC": 1615, + "TCTACG": 1616, + "TCTAGA": 1617, + "TCTAGT": 1618, + "TCTAGC": 1619, + "TCTAGG": 1620, + "TCTTAA": 1621, + "TCTTAT": 1622, + "TCTTAC": 1623, + "TCTTAG": 1624, + "TCTTTA": 1625, + "TCTTTT": 1626, + "TCTTTC": 1627, + "TCTTTG": 1628, + "TCTTCA": 1629, + "TCTTCT": 1630, + "TCTTCC": 1631, + "TCTTCG": 1632, + "TCTTGA": 1633, + "TCTTGT": 1634, + "TCTTGC": 1635, + "TCTTGG": 1636, + "TCTCAA": 1637, + "TCTCAT": 1638, + "TCTCAC": 1639, + "TCTCAG": 1640, + "TCTCTA": 1641, + "TCTCTT": 1642, + "TCTCTC": 1643, + "TCTCTG": 1644, + "TCTCCA": 1645, + "TCTCCT": 1646, + "TCTCCC": 1647, + "TCTCCG": 1648, + "TCTCGA": 1649, + "TCTCGT": 1650, + "TCTCGC": 1651, + "TCTCGG": 1652, + "TCTGAA": 1653, + "TCTGAT": 1654, + "TCTGAC": 1655, + "TCTGAG": 1656, + "TCTGTA": 1657, + "TCTGTT": 1658, + "TCTGTC": 1659, + "TCTGTG": 1660, + "TCTGCA": 1661, + "TCTGCT": 1662, + "TCTGCC": 1663, + "TCTGCG": 1664, + "TCTGGA": 1665, + "TCTGGT": 1666, + "TCTGGC": 1667, + "TCTGGG": 1668, + "TCCAAA": 1669, + "TCCAAT": 1670, + "TCCAAC": 1671, + "TCCAAG": 1672, + "TCCATA": 1673, + "TCCATT": 1674, + "TCCATC": 1675, + "TCCATG": 1676, + "TCCACA": 1677, + "TCCACT": 1678, + "TCCACC": 1679, + "TCCACG": 1680, + "TCCAGA": 1681, + "TCCAGT": 1682, + "TCCAGC": 1683, + "TCCAGG": 1684, + "TCCTAA": 1685, + "TCCTAT": 1686, + "TCCTAC": 1687, + "TCCTAG": 1688, + "TCCTTA": 1689, + "TCCTTT": 1690, + "TCCTTC": 1691, + "TCCTTG": 1692, + "TCCTCA": 1693, + "TCCTCT": 1694, + "TCCTCC": 1695, + "TCCTCG": 1696, + "TCCTGA": 1697, + "TCCTGT": 1698, + "TCCTGC": 1699, + "TCCTGG": 1700, + "TCCCAA": 1701, + "TCCCAT": 1702, + "TCCCAC": 1703, + "TCCCAG": 1704, + "TCCCTA": 1705, + "TCCCTT": 1706, + "TCCCTC": 1707, + "TCCCTG": 1708, + "TCCCCA": 1709, + "TCCCCT": 1710, + "TCCCCC": 1711, + "TCCCCG": 1712, + "TCCCGA": 1713, + "TCCCGT": 1714, + "TCCCGC": 1715, + "TCCCGG": 1716, + "TCCGAA": 1717, + "TCCGAT": 1718, + "TCCGAC": 1719, + "TCCGAG": 1720, + "TCCGTA": 1721, + "TCCGTT": 1722, + "TCCGTC": 1723, + "TCCGTG": 1724, + "TCCGCA": 1725, + "TCCGCT": 1726, + "TCCGCC": 1727, + "TCCGCG": 1728, + "TCCGGA": 1729, + "TCCGGT": 1730, + "TCCGGC": 1731, + "TCCGGG": 1732, + "TCGAAA": 1733, + "TCGAAT": 1734, + "TCGAAC": 1735, + "TCGAAG": 1736, + "TCGATA": 1737, + "TCGATT": 1738, + "TCGATC": 1739, + "TCGATG": 1740, + "TCGACA": 1741, + "TCGACT": 1742, + "TCGACC": 1743, + "TCGACG": 1744, + "TCGAGA": 1745, + "TCGAGT": 1746, + "TCGAGC": 1747, + "TCGAGG": 1748, + "TCGTAA": 1749, + "TCGTAT": 1750, + "TCGTAC": 1751, + "TCGTAG": 1752, + "TCGTTA": 1753, + "TCGTTT": 1754, + "TCGTTC": 1755, + "TCGTTG": 1756, + "TCGTCA": 1757, + "TCGTCT": 1758, + "TCGTCC": 1759, + "TCGTCG": 1760, + "TCGTGA": 1761, + "TCGTGT": 1762, + "TCGTGC": 1763, + "TCGTGG": 1764, + "TCGCAA": 1765, + "TCGCAT": 1766, + "TCGCAC": 1767, + "TCGCAG": 1768, + "TCGCTA": 1769, + "TCGCTT": 1770, + "TCGCTC": 1771, + "TCGCTG": 1772, + "TCGCCA": 1773, + "TCGCCT": 1774, + "TCGCCC": 1775, + "TCGCCG": 1776, + "TCGCGA": 1777, + "TCGCGT": 1778, + "TCGCGC": 1779, + "TCGCGG": 1780, + "TCGGAA": 1781, + "TCGGAT": 1782, + "TCGGAC": 1783, + "TCGGAG": 1784, + "TCGGTA": 1785, + "TCGGTT": 1786, + "TCGGTC": 1787, + "TCGGTG": 1788, + "TCGGCA": 1789, + "TCGGCT": 1790, + "TCGGCC": 1791, + "TCGGCG": 1792, + "TCGGGA": 1793, + "TCGGGT": 1794, + "TCGGGC": 1795, + "TCGGGG": 1796, + "TGAAAA": 1797, + "TGAAAT": 1798, + "TGAAAC": 1799, + "TGAAAG": 1800, + "TGAATA": 1801, + "TGAATT": 1802, + "TGAATC": 1803, + "TGAATG": 1804, + "TGAACA": 1805, + "TGAACT": 1806, + "TGAACC": 1807, + "TGAACG": 1808, + "TGAAGA": 1809, + "TGAAGT": 1810, + "TGAAGC": 1811, + "TGAAGG": 1812, + "TGATAA": 1813, + "TGATAT": 1814, + "TGATAC": 1815, + "TGATAG": 1816, + "TGATTA": 1817, + "TGATTT": 1818, + "TGATTC": 1819, + "TGATTG": 1820, + "TGATCA": 1821, + "TGATCT": 1822, + "TGATCC": 1823, + "TGATCG": 1824, + "TGATGA": 1825, + "TGATGT": 1826, + "TGATGC": 1827, + "TGATGG": 1828, + "TGACAA": 1829, + "TGACAT": 1830, + "TGACAC": 1831, + "TGACAG": 1832, + "TGACTA": 1833, + "TGACTT": 1834, + "TGACTC": 1835, + "TGACTG": 1836, + "TGACCA": 1837, + "TGACCT": 1838, + "TGACCC": 1839, + "TGACCG": 1840, + "TGACGA": 1841, + "TGACGT": 1842, + "TGACGC": 1843, + "TGACGG": 1844, + "TGAGAA": 1845, + "TGAGAT": 1846, + "TGAGAC": 1847, + "TGAGAG": 1848, + "TGAGTA": 1849, + "TGAGTT": 1850, + "TGAGTC": 1851, + "TGAGTG": 1852, + "TGAGCA": 1853, + "TGAGCT": 1854, + "TGAGCC": 1855, + "TGAGCG": 1856, + "TGAGGA": 1857, + "TGAGGT": 1858, + "TGAGGC": 1859, + "TGAGGG": 1860, + "TGTAAA": 1861, + "TGTAAT": 1862, + "TGTAAC": 1863, + "TGTAAG": 1864, + "TGTATA": 1865, + "TGTATT": 1866, + "TGTATC": 1867, + "TGTATG": 1868, + "TGTACA": 1869, + "TGTACT": 1870, + "TGTACC": 1871, + "TGTACG": 1872, + "TGTAGA": 1873, + "TGTAGT": 1874, + "TGTAGC": 1875, + "TGTAGG": 1876, + "TGTTAA": 1877, + "TGTTAT": 1878, + "TGTTAC": 1879, + "TGTTAG": 1880, + "TGTTTA": 1881, + "TGTTTT": 1882, + "TGTTTC": 1883, + "TGTTTG": 1884, + "TGTTCA": 1885, + "TGTTCT": 1886, + "TGTTCC": 1887, + "TGTTCG": 1888, + "TGTTGA": 1889, + "TGTTGT": 1890, + "TGTTGC": 1891, + "TGTTGG": 1892, + "TGTCAA": 1893, + "TGTCAT": 1894, + "TGTCAC": 1895, + "TGTCAG": 1896, + "TGTCTA": 1897, + "TGTCTT": 1898, + "TGTCTC": 1899, + "TGTCTG": 1900, + "TGTCCA": 1901, + "TGTCCT": 1902, + "TGTCCC": 1903, + "TGTCCG": 1904, + "TGTCGA": 1905, + "TGTCGT": 1906, + "TGTCGC": 1907, + "TGTCGG": 1908, + "TGTGAA": 1909, + "TGTGAT": 1910, + "TGTGAC": 1911, + "TGTGAG": 1912, + "TGTGTA": 1913, + "TGTGTT": 1914, + "TGTGTC": 1915, + "TGTGTG": 1916, + "TGTGCA": 1917, + "TGTGCT": 1918, + "TGTGCC": 1919, + "TGTGCG": 1920, + "TGTGGA": 1921, + "TGTGGT": 1922, + "TGTGGC": 1923, + "TGTGGG": 1924, + "TGCAAA": 1925, + "TGCAAT": 1926, + "TGCAAC": 1927, + "TGCAAG": 1928, + "TGCATA": 1929, + "TGCATT": 1930, + "TGCATC": 1931, + "TGCATG": 1932, + "TGCACA": 1933, + "TGCACT": 1934, + "TGCACC": 1935, + "TGCACG": 1936, + "TGCAGA": 1937, + "TGCAGT": 1938, + "TGCAGC": 1939, + "TGCAGG": 1940, + "TGCTAA": 1941, + "TGCTAT": 1942, + "TGCTAC": 1943, + "TGCTAG": 1944, + "TGCTTA": 1945, + "TGCTTT": 1946, + "TGCTTC": 1947, + "TGCTTG": 1948, + "TGCTCA": 1949, + "TGCTCT": 1950, + "TGCTCC": 1951, + "TGCTCG": 1952, + "TGCTGA": 1953, + "TGCTGT": 1954, + "TGCTGC": 1955, + "TGCTGG": 1956, + "TGCCAA": 1957, + "TGCCAT": 1958, + "TGCCAC": 1959, + "TGCCAG": 1960, + "TGCCTA": 1961, + "TGCCTT": 1962, + "TGCCTC": 1963, + "TGCCTG": 1964, + "TGCCCA": 1965, + "TGCCCT": 1966, + "TGCCCC": 1967, + "TGCCCG": 1968, + "TGCCGA": 1969, + "TGCCGT": 1970, + "TGCCGC": 1971, + "TGCCGG": 1972, + "TGCGAA": 1973, + "TGCGAT": 1974, + "TGCGAC": 1975, + "TGCGAG": 1976, + "TGCGTA": 1977, + "TGCGTT": 1978, + "TGCGTC": 1979, + "TGCGTG": 1980, + "TGCGCA": 1981, + "TGCGCT": 1982, + "TGCGCC": 1983, + "TGCGCG": 1984, + "TGCGGA": 1985, + "TGCGGT": 1986, + "TGCGGC": 1987, + "TGCGGG": 1988, + "TGGAAA": 1989, + "TGGAAT": 1990, + "TGGAAC": 1991, + "TGGAAG": 1992, + "TGGATA": 1993, + "TGGATT": 1994, + "TGGATC": 1995, + "TGGATG": 1996, + "TGGACA": 1997, + "TGGACT": 1998, + "TGGACC": 1999, + "TGGACG": 2000, + "TGGAGA": 2001, + "TGGAGT": 2002, + "TGGAGC": 2003, + "TGGAGG": 2004, + "TGGTAA": 2005, + "TGGTAT": 2006, + "TGGTAC": 2007, + "TGGTAG": 2008, + "TGGTTA": 2009, + "TGGTTT": 2010, + "TGGTTC": 2011, + "TGGTTG": 2012, + "TGGTCA": 2013, + "TGGTCT": 2014, + "TGGTCC": 2015, + "TGGTCG": 2016, + "TGGTGA": 2017, + "TGGTGT": 2018, + "TGGTGC": 2019, + "TGGTGG": 2020, + "TGGCAA": 2021, + "TGGCAT": 2022, + "TGGCAC": 2023, + "TGGCAG": 2024, + "TGGCTA": 2025, + "TGGCTT": 2026, + "TGGCTC": 2027, + "TGGCTG": 2028, + "TGGCCA": 2029, + "TGGCCT": 2030, + "TGGCCC": 2031, + "TGGCCG": 2032, + "TGGCGA": 2033, + "TGGCGT": 2034, + "TGGCGC": 2035, + "TGGCGG": 2036, + "TGGGAA": 2037, + "TGGGAT": 2038, + "TGGGAC": 2039, + "TGGGAG": 2040, + "TGGGTA": 2041, + "TGGGTT": 2042, + "TGGGTC": 2043, + "TGGGTG": 2044, + "TGGGCA": 2045, + "TGGGCT": 2046, + "TGGGCC": 2047, + "TGGGCG": 2048, + "TGGGGA": 2049, + "TGGGGT": 2050, + "TGGGGC": 2051, + "TGGGGG": 2052, + "CAAAAA": 2053, + "CAAAAT": 2054, + "CAAAAC": 2055, + "CAAAAG": 2056, + "CAAATA": 2057, + "CAAATT": 2058, + "CAAATC": 2059, + "CAAATG": 2060, + "CAAACA": 2061, + "CAAACT": 2062, + "CAAACC": 2063, + "CAAACG": 2064, + "CAAAGA": 2065, + "CAAAGT": 2066, + "CAAAGC": 2067, + "CAAAGG": 2068, + "CAATAA": 2069, + "CAATAT": 2070, + "CAATAC": 2071, + "CAATAG": 2072, + "CAATTA": 2073, + "CAATTT": 2074, + "CAATTC": 2075, + "CAATTG": 2076, + "CAATCA": 2077, + "CAATCT": 2078, + "CAATCC": 2079, + "CAATCG": 2080, + "CAATGA": 2081, + "CAATGT": 2082, + "CAATGC": 2083, + "CAATGG": 2084, + "CAACAA": 2085, + "CAACAT": 2086, + "CAACAC": 2087, + "CAACAG": 2088, + "CAACTA": 2089, + "CAACTT": 2090, + "CAACTC": 2091, + "CAACTG": 2092, + "CAACCA": 2093, + "CAACCT": 2094, + "CAACCC": 2095, + "CAACCG": 2096, + "CAACGA": 2097, + "CAACGT": 2098, + "CAACGC": 2099, + "CAACGG": 2100, + "CAAGAA": 2101, + "CAAGAT": 2102, + "CAAGAC": 2103, + "CAAGAG": 2104, + "CAAGTA": 2105, + "CAAGTT": 2106, + "CAAGTC": 2107, + "CAAGTG": 2108, + "CAAGCA": 2109, + "CAAGCT": 2110, + "CAAGCC": 2111, + "CAAGCG": 2112, + "CAAGGA": 2113, + "CAAGGT": 2114, + "CAAGGC": 2115, + "CAAGGG": 2116, + "CATAAA": 2117, + "CATAAT": 2118, + "CATAAC": 2119, + "CATAAG": 2120, + "CATATA": 2121, + "CATATT": 2122, + "CATATC": 2123, + "CATATG": 2124, + "CATACA": 2125, + "CATACT": 2126, + "CATACC": 2127, + "CATACG": 2128, + "CATAGA": 2129, + "CATAGT": 2130, + "CATAGC": 2131, + "CATAGG": 2132, + "CATTAA": 2133, + "CATTAT": 2134, + "CATTAC": 2135, + "CATTAG": 2136, + "CATTTA": 2137, + "CATTTT": 2138, + "CATTTC": 2139, + "CATTTG": 2140, + "CATTCA": 2141, + "CATTCT": 2142, + "CATTCC": 2143, + "CATTCG": 2144, + "CATTGA": 2145, + "CATTGT": 2146, + "CATTGC": 2147, + "CATTGG": 2148, + "CATCAA": 2149, + "CATCAT": 2150, + "CATCAC": 2151, + "CATCAG": 2152, + "CATCTA": 2153, + "CATCTT": 2154, + "CATCTC": 2155, + "CATCTG": 2156, + "CATCCA": 2157, + "CATCCT": 2158, + "CATCCC": 2159, + "CATCCG": 2160, + "CATCGA": 2161, + "CATCGT": 2162, + "CATCGC": 2163, + "CATCGG": 2164, + "CATGAA": 2165, + "CATGAT": 2166, + "CATGAC": 2167, + "CATGAG": 2168, + "CATGTA": 2169, + "CATGTT": 2170, + "CATGTC": 2171, + "CATGTG": 2172, + "CATGCA": 2173, + "CATGCT": 2174, + "CATGCC": 2175, + "CATGCG": 2176, + "CATGGA": 2177, + "CATGGT": 2178, + "CATGGC": 2179, + "CATGGG": 2180, + "CACAAA": 2181, + "CACAAT": 2182, + "CACAAC": 2183, + "CACAAG": 2184, + "CACATA": 2185, + "CACATT": 2186, + "CACATC": 2187, + "CACATG": 2188, + "CACACA": 2189, + "CACACT": 2190, + "CACACC": 2191, + "CACACG": 2192, + "CACAGA": 2193, + "CACAGT": 2194, + "CACAGC": 2195, + "CACAGG": 2196, + "CACTAA": 2197, + "CACTAT": 2198, + "CACTAC": 2199, + "CACTAG": 2200, + "CACTTA": 2201, + "CACTTT": 2202, + "CACTTC": 2203, + "CACTTG": 2204, + "CACTCA": 2205, + "CACTCT": 2206, + "CACTCC": 2207, + "CACTCG": 2208, + "CACTGA": 2209, + "CACTGT": 2210, + "CACTGC": 2211, + "CACTGG": 2212, + "CACCAA": 2213, + "CACCAT": 2214, + "CACCAC": 2215, + "CACCAG": 2216, + "CACCTA": 2217, + "CACCTT": 2218, + "CACCTC": 2219, + "CACCTG": 2220, + "CACCCA": 2221, + "CACCCT": 2222, + "CACCCC": 2223, + "CACCCG": 2224, + "CACCGA": 2225, + "CACCGT": 2226, + "CACCGC": 2227, + "CACCGG": 2228, + "CACGAA": 2229, + "CACGAT": 2230, + "CACGAC": 2231, + "CACGAG": 2232, + "CACGTA": 2233, + "CACGTT": 2234, + "CACGTC": 2235, + "CACGTG": 2236, + "CACGCA": 2237, + "CACGCT": 2238, + "CACGCC": 2239, + "CACGCG": 2240, + "CACGGA": 2241, + "CACGGT": 2242, + "CACGGC": 2243, + "CACGGG": 2244, + "CAGAAA": 2245, + "CAGAAT": 2246, + "CAGAAC": 2247, + "CAGAAG": 2248, + "CAGATA": 2249, + "CAGATT": 2250, + "CAGATC": 2251, + "CAGATG": 2252, + "CAGACA": 2253, + "CAGACT": 2254, + "CAGACC": 2255, + "CAGACG": 2256, + "CAGAGA": 2257, + "CAGAGT": 2258, + "CAGAGC": 2259, + "CAGAGG": 2260, + "CAGTAA": 2261, + "CAGTAT": 2262, + "CAGTAC": 2263, + "CAGTAG": 2264, + "CAGTTA": 2265, + "CAGTTT": 2266, + "CAGTTC": 2267, + "CAGTTG": 2268, + "CAGTCA": 2269, + "CAGTCT": 2270, + "CAGTCC": 2271, + "CAGTCG": 2272, + "CAGTGA": 2273, + "CAGTGT": 2274, + "CAGTGC": 2275, + "CAGTGG": 2276, + "CAGCAA": 2277, + "CAGCAT": 2278, + "CAGCAC": 2279, + "CAGCAG": 2280, + "CAGCTA": 2281, + "CAGCTT": 2282, + "CAGCTC": 2283, + "CAGCTG": 2284, + "CAGCCA": 2285, + "CAGCCT": 2286, + "CAGCCC": 2287, + "CAGCCG": 2288, + "CAGCGA": 2289, + "CAGCGT": 2290, + "CAGCGC": 2291, + "CAGCGG": 2292, + "CAGGAA": 2293, + "CAGGAT": 2294, + "CAGGAC": 2295, + "CAGGAG": 2296, + "CAGGTA": 2297, + "CAGGTT": 2298, + "CAGGTC": 2299, + "CAGGTG": 2300, + "CAGGCA": 2301, + "CAGGCT": 2302, + "CAGGCC": 2303, + "CAGGCG": 2304, + "CAGGGA": 2305, + "CAGGGT": 2306, + "CAGGGC": 2307, + "CAGGGG": 2308, + "CTAAAA": 2309, + "CTAAAT": 2310, + "CTAAAC": 2311, + "CTAAAG": 2312, + "CTAATA": 2313, + "CTAATT": 2314, + "CTAATC": 2315, + "CTAATG": 2316, + "CTAACA": 2317, + "CTAACT": 2318, + "CTAACC": 2319, + "CTAACG": 2320, + "CTAAGA": 2321, + "CTAAGT": 2322, + "CTAAGC": 2323, + "CTAAGG": 2324, + "CTATAA": 2325, + "CTATAT": 2326, + "CTATAC": 2327, + "CTATAG": 2328, + "CTATTA": 2329, + "CTATTT": 2330, + "CTATTC": 2331, + "CTATTG": 2332, + "CTATCA": 2333, + "CTATCT": 2334, + "CTATCC": 2335, + "CTATCG": 2336, + "CTATGA": 2337, + "CTATGT": 2338, + "CTATGC": 2339, + "CTATGG": 2340, + "CTACAA": 2341, + "CTACAT": 2342, + "CTACAC": 2343, + "CTACAG": 2344, + "CTACTA": 2345, + "CTACTT": 2346, + "CTACTC": 2347, + "CTACTG": 2348, + "CTACCA": 2349, + "CTACCT": 2350, + "CTACCC": 2351, + "CTACCG": 2352, + "CTACGA": 2353, + "CTACGT": 2354, + "CTACGC": 2355, + "CTACGG": 2356, + "CTAGAA": 2357, + "CTAGAT": 2358, + "CTAGAC": 2359, + "CTAGAG": 2360, + "CTAGTA": 2361, + "CTAGTT": 2362, + "CTAGTC": 2363, + "CTAGTG": 2364, + "CTAGCA": 2365, + "CTAGCT": 2366, + "CTAGCC": 2367, + "CTAGCG": 2368, + "CTAGGA": 2369, + "CTAGGT": 2370, + "CTAGGC": 2371, + "CTAGGG": 2372, + "CTTAAA": 2373, + "CTTAAT": 2374, + "CTTAAC": 2375, + "CTTAAG": 2376, + "CTTATA": 2377, + "CTTATT": 2378, + "CTTATC": 2379, + "CTTATG": 2380, + "CTTACA": 2381, + "CTTACT": 2382, + "CTTACC": 2383, + "CTTACG": 2384, + "CTTAGA": 2385, + "CTTAGT": 2386, + "CTTAGC": 2387, + "CTTAGG": 2388, + "CTTTAA": 2389, + "CTTTAT": 2390, + "CTTTAC": 2391, + "CTTTAG": 2392, + "CTTTTA": 2393, + "CTTTTT": 2394, + "CTTTTC": 2395, + "CTTTTG": 2396, + "CTTTCA": 2397, + "CTTTCT": 2398, + "CTTTCC": 2399, + "CTTTCG": 2400, + "CTTTGA": 2401, + "CTTTGT": 2402, + "CTTTGC": 2403, + "CTTTGG": 2404, + "CTTCAA": 2405, + "CTTCAT": 2406, + "CTTCAC": 2407, + "CTTCAG": 2408, + "CTTCTA": 2409, + "CTTCTT": 2410, + "CTTCTC": 2411, + "CTTCTG": 2412, + "CTTCCA": 2413, + "CTTCCT": 2414, + "CTTCCC": 2415, + "CTTCCG": 2416, + "CTTCGA": 2417, + "CTTCGT": 2418, + "CTTCGC": 2419, + "CTTCGG": 2420, + "CTTGAA": 2421, + "CTTGAT": 2422, + "CTTGAC": 2423, + "CTTGAG": 2424, + "CTTGTA": 2425, + "CTTGTT": 2426, + "CTTGTC": 2427, + "CTTGTG": 2428, + "CTTGCA": 2429, + "CTTGCT": 2430, + "CTTGCC": 2431, + "CTTGCG": 2432, + "CTTGGA": 2433, + "CTTGGT": 2434, + "CTTGGC": 2435, + "CTTGGG": 2436, + "CTCAAA": 2437, + "CTCAAT": 2438, + "CTCAAC": 2439, + "CTCAAG": 2440, + "CTCATA": 2441, + "CTCATT": 2442, + "CTCATC": 2443, + "CTCATG": 2444, + "CTCACA": 2445, + "CTCACT": 2446, + "CTCACC": 2447, + "CTCACG": 2448, + "CTCAGA": 2449, + "CTCAGT": 2450, + "CTCAGC": 2451, + "CTCAGG": 2452, + "CTCTAA": 2453, + "CTCTAT": 2454, + "CTCTAC": 2455, + "CTCTAG": 2456, + "CTCTTA": 2457, + "CTCTTT": 2458, + "CTCTTC": 2459, + "CTCTTG": 2460, + "CTCTCA": 2461, + "CTCTCT": 2462, + "CTCTCC": 2463, + "CTCTCG": 2464, + "CTCTGA": 2465, + "CTCTGT": 2466, + "CTCTGC": 2467, + "CTCTGG": 2468, + "CTCCAA": 2469, + "CTCCAT": 2470, + "CTCCAC": 2471, + "CTCCAG": 2472, + "CTCCTA": 2473, + "CTCCTT": 2474, + "CTCCTC": 2475, + "CTCCTG": 2476, + "CTCCCA": 2477, + "CTCCCT": 2478, + "CTCCCC": 2479, + "CTCCCG": 2480, + "CTCCGA": 2481, + "CTCCGT": 2482, + "CTCCGC": 2483, + "CTCCGG": 2484, + "CTCGAA": 2485, + "CTCGAT": 2486, + "CTCGAC": 2487, + "CTCGAG": 2488, + "CTCGTA": 2489, + "CTCGTT": 2490, + "CTCGTC": 2491, + "CTCGTG": 2492, + "CTCGCA": 2493, + "CTCGCT": 2494, + "CTCGCC": 2495, + "CTCGCG": 2496, + "CTCGGA": 2497, + "CTCGGT": 2498, + "CTCGGC": 2499, + "CTCGGG": 2500, + "CTGAAA": 2501, + "CTGAAT": 2502, + "CTGAAC": 2503, + "CTGAAG": 2504, + "CTGATA": 2505, + "CTGATT": 2506, + "CTGATC": 2507, + "CTGATG": 2508, + "CTGACA": 2509, + "CTGACT": 2510, + "CTGACC": 2511, + "CTGACG": 2512, + "CTGAGA": 2513, + "CTGAGT": 2514, + "CTGAGC": 2515, + "CTGAGG": 2516, + "CTGTAA": 2517, + "CTGTAT": 2518, + "CTGTAC": 2519, + "CTGTAG": 2520, + "CTGTTA": 2521, + "CTGTTT": 2522, + "CTGTTC": 2523, + "CTGTTG": 2524, + "CTGTCA": 2525, + "CTGTCT": 2526, + "CTGTCC": 2527, + "CTGTCG": 2528, + "CTGTGA": 2529, + "CTGTGT": 2530, + "CTGTGC": 2531, + "CTGTGG": 2532, + "CTGCAA": 2533, + "CTGCAT": 2534, + "CTGCAC": 2535, + "CTGCAG": 2536, + "CTGCTA": 2537, + "CTGCTT": 2538, + "CTGCTC": 2539, + "CTGCTG": 2540, + "CTGCCA": 2541, + "CTGCCT": 2542, + "CTGCCC": 2543, + "CTGCCG": 2544, + "CTGCGA": 2545, + "CTGCGT": 2546, + "CTGCGC": 2547, + "CTGCGG": 2548, + "CTGGAA": 2549, + "CTGGAT": 2550, + "CTGGAC": 2551, + "CTGGAG": 2552, + "CTGGTA": 2553, + "CTGGTT": 2554, + "CTGGTC": 2555, + "CTGGTG": 2556, + "CTGGCA": 2557, + "CTGGCT": 2558, + "CTGGCC": 2559, + "CTGGCG": 2560, + "CTGGGA": 2561, + "CTGGGT": 2562, + "CTGGGC": 2563, + "CTGGGG": 2564, + "CCAAAA": 2565, + "CCAAAT": 2566, + "CCAAAC": 2567, + "CCAAAG": 2568, + "CCAATA": 2569, + "CCAATT": 2570, + "CCAATC": 2571, + "CCAATG": 2572, + "CCAACA": 2573, + "CCAACT": 2574, + "CCAACC": 2575, + "CCAACG": 2576, + "CCAAGA": 2577, + "CCAAGT": 2578, + "CCAAGC": 2579, + "CCAAGG": 2580, + "CCATAA": 2581, + "CCATAT": 2582, + "CCATAC": 2583, + "CCATAG": 2584, + "CCATTA": 2585, + "CCATTT": 2586, + "CCATTC": 2587, + "CCATTG": 2588, + "CCATCA": 2589, + "CCATCT": 2590, + "CCATCC": 2591, + "CCATCG": 2592, + "CCATGA": 2593, + "CCATGT": 2594, + "CCATGC": 2595, + "CCATGG": 2596, + "CCACAA": 2597, + "CCACAT": 2598, + "CCACAC": 2599, + "CCACAG": 2600, + "CCACTA": 2601, + "CCACTT": 2602, + "CCACTC": 2603, + "CCACTG": 2604, + "CCACCA": 2605, + "CCACCT": 2606, + "CCACCC": 2607, + "CCACCG": 2608, + "CCACGA": 2609, + "CCACGT": 2610, + "CCACGC": 2611, + "CCACGG": 2612, + "CCAGAA": 2613, + "CCAGAT": 2614, + "CCAGAC": 2615, + "CCAGAG": 2616, + "CCAGTA": 2617, + "CCAGTT": 2618, + "CCAGTC": 2619, + "CCAGTG": 2620, + "CCAGCA": 2621, + "CCAGCT": 2622, + "CCAGCC": 2623, + "CCAGCG": 2624, + "CCAGGA": 2625, + "CCAGGT": 2626, + "CCAGGC": 2627, + "CCAGGG": 2628, + "CCTAAA": 2629, + "CCTAAT": 2630, + "CCTAAC": 2631, + "CCTAAG": 2632, + "CCTATA": 2633, + "CCTATT": 2634, + "CCTATC": 2635, + "CCTATG": 2636, + "CCTACA": 2637, + "CCTACT": 2638, + "CCTACC": 2639, + "CCTACG": 2640, + "CCTAGA": 2641, + "CCTAGT": 2642, + "CCTAGC": 2643, + "CCTAGG": 2644, + "CCTTAA": 2645, + "CCTTAT": 2646, + "CCTTAC": 2647, + "CCTTAG": 2648, + "CCTTTA": 2649, + "CCTTTT": 2650, + "CCTTTC": 2651, + "CCTTTG": 2652, + "CCTTCA": 2653, + "CCTTCT": 2654, + "CCTTCC": 2655, + "CCTTCG": 2656, + "CCTTGA": 2657, + "CCTTGT": 2658, + "CCTTGC": 2659, + "CCTTGG": 2660, + "CCTCAA": 2661, + "CCTCAT": 2662, + "CCTCAC": 2663, + "CCTCAG": 2664, + "CCTCTA": 2665, + "CCTCTT": 2666, + "CCTCTC": 2667, + "CCTCTG": 2668, + "CCTCCA": 2669, + "CCTCCT": 2670, + "CCTCCC": 2671, + "CCTCCG": 2672, + "CCTCGA": 2673, + "CCTCGT": 2674, + "CCTCGC": 2675, + "CCTCGG": 2676, + "CCTGAA": 2677, + "CCTGAT": 2678, + "CCTGAC": 2679, + "CCTGAG": 2680, + "CCTGTA": 2681, + "CCTGTT": 2682, + "CCTGTC": 2683, + "CCTGTG": 2684, + "CCTGCA": 2685, + "CCTGCT": 2686, + "CCTGCC": 2687, + "CCTGCG": 2688, + "CCTGGA": 2689, + "CCTGGT": 2690, + "CCTGGC": 2691, + "CCTGGG": 2692, + "CCCAAA": 2693, + "CCCAAT": 2694, + "CCCAAC": 2695, + "CCCAAG": 2696, + "CCCATA": 2697, + "CCCATT": 2698, + "CCCATC": 2699, + "CCCATG": 2700, + "CCCACA": 2701, + "CCCACT": 2702, + "CCCACC": 2703, + "CCCACG": 2704, + "CCCAGA": 2705, + "CCCAGT": 2706, + "CCCAGC": 2707, + "CCCAGG": 2708, + "CCCTAA": 2709, + "CCCTAT": 2710, + "CCCTAC": 2711, + "CCCTAG": 2712, + "CCCTTA": 2713, + "CCCTTT": 2714, + "CCCTTC": 2715, + "CCCTTG": 2716, + "CCCTCA": 2717, + "CCCTCT": 2718, + "CCCTCC": 2719, + "CCCTCG": 2720, + "CCCTGA": 2721, + "CCCTGT": 2722, + "CCCTGC": 2723, + "CCCTGG": 2724, + "CCCCAA": 2725, + "CCCCAT": 2726, + "CCCCAC": 2727, + "CCCCAG": 2728, + "CCCCTA": 2729, + "CCCCTT": 2730, + "CCCCTC": 2731, + "CCCCTG": 2732, + "CCCCCA": 2733, + "CCCCCT": 2734, + "CCCCCC": 2735, + "CCCCCG": 2736, + "CCCCGA": 2737, + "CCCCGT": 2738, + "CCCCGC": 2739, + "CCCCGG": 2740, + "CCCGAA": 2741, + "CCCGAT": 2742, + "CCCGAC": 2743, + "CCCGAG": 2744, + "CCCGTA": 2745, + "CCCGTT": 2746, + "CCCGTC": 2747, + "CCCGTG": 2748, + "CCCGCA": 2749, + "CCCGCT": 2750, + "CCCGCC": 2751, + "CCCGCG": 2752, + "CCCGGA": 2753, + "CCCGGT": 2754, + "CCCGGC": 2755, + "CCCGGG": 2756, + "CCGAAA": 2757, + "CCGAAT": 2758, + "CCGAAC": 2759, + "CCGAAG": 2760, + "CCGATA": 2761, + "CCGATT": 2762, + "CCGATC": 2763, + "CCGATG": 2764, + "CCGACA": 2765, + "CCGACT": 2766, + "CCGACC": 2767, + "CCGACG": 2768, + "CCGAGA": 2769, + "CCGAGT": 2770, + "CCGAGC": 2771, + "CCGAGG": 2772, + "CCGTAA": 2773, + "CCGTAT": 2774, + "CCGTAC": 2775, + "CCGTAG": 2776, + "CCGTTA": 2777, + "CCGTTT": 2778, + "CCGTTC": 2779, + "CCGTTG": 2780, + "CCGTCA": 2781, + "CCGTCT": 2782, + "CCGTCC": 2783, + "CCGTCG": 2784, + "CCGTGA": 2785, + "CCGTGT": 2786, + "CCGTGC": 2787, + "CCGTGG": 2788, + "CCGCAA": 2789, + "CCGCAT": 2790, + "CCGCAC": 2791, + "CCGCAG": 2792, + "CCGCTA": 2793, + "CCGCTT": 2794, + "CCGCTC": 2795, + "CCGCTG": 2796, + "CCGCCA": 2797, + "CCGCCT": 2798, + "CCGCCC": 2799, + "CCGCCG": 2800, + "CCGCGA": 2801, + "CCGCGT": 2802, + "CCGCGC": 2803, + "CCGCGG": 2804, + "CCGGAA": 2805, + "CCGGAT": 2806, + "CCGGAC": 2807, + "CCGGAG": 2808, + "CCGGTA": 2809, + "CCGGTT": 2810, + "CCGGTC": 2811, + "CCGGTG": 2812, + "CCGGCA": 2813, + "CCGGCT": 2814, + "CCGGCC": 2815, + "CCGGCG": 2816, + "CCGGGA": 2817, + "CCGGGT": 2818, + "CCGGGC": 2819, + "CCGGGG": 2820, + "CGAAAA": 2821, + "CGAAAT": 2822, + "CGAAAC": 2823, + "CGAAAG": 2824, + "CGAATA": 2825, + "CGAATT": 2826, + "CGAATC": 2827, + "CGAATG": 2828, + "CGAACA": 2829, + "CGAACT": 2830, + "CGAACC": 2831, + "CGAACG": 2832, + "CGAAGA": 2833, + "CGAAGT": 2834, + "CGAAGC": 2835, + "CGAAGG": 2836, + "CGATAA": 2837, + "CGATAT": 2838, + "CGATAC": 2839, + "CGATAG": 2840, + "CGATTA": 2841, + "CGATTT": 2842, + "CGATTC": 2843, + "CGATTG": 2844, + "CGATCA": 2845, + "CGATCT": 2846, + "CGATCC": 2847, + "CGATCG": 2848, + "CGATGA": 2849, + "CGATGT": 2850, + "CGATGC": 2851, + "CGATGG": 2852, + "CGACAA": 2853, + "CGACAT": 2854, + "CGACAC": 2855, + "CGACAG": 2856, + "CGACTA": 2857, + "CGACTT": 2858, + "CGACTC": 2859, + "CGACTG": 2860, + "CGACCA": 2861, + "CGACCT": 2862, + "CGACCC": 2863, + "CGACCG": 2864, + "CGACGA": 2865, + "CGACGT": 2866, + "CGACGC": 2867, + "CGACGG": 2868, + "CGAGAA": 2869, + "CGAGAT": 2870, + "CGAGAC": 2871, + "CGAGAG": 2872, + "CGAGTA": 2873, + "CGAGTT": 2874, + "CGAGTC": 2875, + "CGAGTG": 2876, + "CGAGCA": 2877, + "CGAGCT": 2878, + "CGAGCC": 2879, + "CGAGCG": 2880, + "CGAGGA": 2881, + "CGAGGT": 2882, + "CGAGGC": 2883, + "CGAGGG": 2884, + "CGTAAA": 2885, + "CGTAAT": 2886, + "CGTAAC": 2887, + "CGTAAG": 2888, + "CGTATA": 2889, + "CGTATT": 2890, + "CGTATC": 2891, + "CGTATG": 2892, + "CGTACA": 2893, + "CGTACT": 2894, + "CGTACC": 2895, + "CGTACG": 2896, + "CGTAGA": 2897, + "CGTAGT": 2898, + "CGTAGC": 2899, + "CGTAGG": 2900, + "CGTTAA": 2901, + "CGTTAT": 2902, + "CGTTAC": 2903, + "CGTTAG": 2904, + "CGTTTA": 2905, + "CGTTTT": 2906, + "CGTTTC": 2907, + "CGTTTG": 2908, + "CGTTCA": 2909, + "CGTTCT": 2910, + "CGTTCC": 2911, + "CGTTCG": 2912, + "CGTTGA": 2913, + "CGTTGT": 2914, + "CGTTGC": 2915, + "CGTTGG": 2916, + "CGTCAA": 2917, + "CGTCAT": 2918, + "CGTCAC": 2919, + "CGTCAG": 2920, + "CGTCTA": 2921, + "CGTCTT": 2922, + "CGTCTC": 2923, + "CGTCTG": 2924, + "CGTCCA": 2925, + "CGTCCT": 2926, + "CGTCCC": 2927, + "CGTCCG": 2928, + "CGTCGA": 2929, + "CGTCGT": 2930, + "CGTCGC": 2931, + "CGTCGG": 2932, + "CGTGAA": 2933, + "CGTGAT": 2934, + "CGTGAC": 2935, + "CGTGAG": 2936, + "CGTGTA": 2937, + "CGTGTT": 2938, + "CGTGTC": 2939, + "CGTGTG": 2940, + "CGTGCA": 2941, + "CGTGCT": 2942, + "CGTGCC": 2943, + "CGTGCG": 2944, + "CGTGGA": 2945, + "CGTGGT": 2946, + "CGTGGC": 2947, + "CGTGGG": 2948, + "CGCAAA": 2949, + "CGCAAT": 2950, + "CGCAAC": 2951, + "CGCAAG": 2952, + "CGCATA": 2953, + "CGCATT": 2954, + "CGCATC": 2955, + "CGCATG": 2956, + "CGCACA": 2957, + "CGCACT": 2958, + "CGCACC": 2959, + "CGCACG": 2960, + "CGCAGA": 2961, + "CGCAGT": 2962, + "CGCAGC": 2963, + "CGCAGG": 2964, + "CGCTAA": 2965, + "CGCTAT": 2966, + "CGCTAC": 2967, + "CGCTAG": 2968, + "CGCTTA": 2969, + "CGCTTT": 2970, + "CGCTTC": 2971, + "CGCTTG": 2972, + "CGCTCA": 2973, + "CGCTCT": 2974, + "CGCTCC": 2975, + "CGCTCG": 2976, + "CGCTGA": 2977, + "CGCTGT": 2978, + "CGCTGC": 2979, + "CGCTGG": 2980, + "CGCCAA": 2981, + "CGCCAT": 2982, + "CGCCAC": 2983, + "CGCCAG": 2984, + "CGCCTA": 2985, + "CGCCTT": 2986, + "CGCCTC": 2987, + "CGCCTG": 2988, + "CGCCCA": 2989, + "CGCCCT": 2990, + "CGCCCC": 2991, + "CGCCCG": 2992, + "CGCCGA": 2993, + "CGCCGT": 2994, + "CGCCGC": 2995, + "CGCCGG": 2996, + "CGCGAA": 2997, + "CGCGAT": 2998, + "CGCGAC": 2999, + "CGCGAG": 3000, + "CGCGTA": 3001, + "CGCGTT": 3002, + "CGCGTC": 3003, + "CGCGTG": 3004, + "CGCGCA": 3005, + "CGCGCT": 3006, + "CGCGCC": 3007, + "CGCGCG": 3008, + "CGCGGA": 3009, + "CGCGGT": 3010, + "CGCGGC": 3011, + "CGCGGG": 3012, + "CGGAAA": 3013, + "CGGAAT": 3014, + "CGGAAC": 3015, + "CGGAAG": 3016, + "CGGATA": 3017, + "CGGATT": 3018, + "CGGATC": 3019, + "CGGATG": 3020, + "CGGACA": 3021, + "CGGACT": 3022, + "CGGACC": 3023, + "CGGACG": 3024, + "CGGAGA": 3025, + "CGGAGT": 3026, + "CGGAGC": 3027, + "CGGAGG": 3028, + "CGGTAA": 3029, + "CGGTAT": 3030, + "CGGTAC": 3031, + "CGGTAG": 3032, + "CGGTTA": 3033, + "CGGTTT": 3034, + "CGGTTC": 3035, + "CGGTTG": 3036, + "CGGTCA": 3037, + "CGGTCT": 3038, + "CGGTCC": 3039, + "CGGTCG": 3040, + "CGGTGA": 3041, + "CGGTGT": 3042, + "CGGTGC": 3043, + "CGGTGG": 3044, + "CGGCAA": 3045, + "CGGCAT": 3046, + "CGGCAC": 3047, + "CGGCAG": 3048, + "CGGCTA": 3049, + "CGGCTT": 3050, + "CGGCTC": 3051, + "CGGCTG": 3052, + "CGGCCA": 3053, + "CGGCCT": 3054, + "CGGCCC": 3055, + "CGGCCG": 3056, + "CGGCGA": 3057, + "CGGCGT": 3058, + "CGGCGC": 3059, + "CGGCGG": 3060, + "CGGGAA": 3061, + "CGGGAT": 3062, + "CGGGAC": 3063, + "CGGGAG": 3064, + "CGGGTA": 3065, + "CGGGTT": 3066, + "CGGGTC": 3067, + "CGGGTG": 3068, + "CGGGCA": 3069, + "CGGGCT": 3070, + "CGGGCC": 3071, + "CGGGCG": 3072, + "CGGGGA": 3073, + "CGGGGT": 3074, + "CGGGGC": 3075, + "CGGGGG": 3076, + "GAAAAA": 3077, + "GAAAAT": 3078, + "GAAAAC": 3079, + "GAAAAG": 3080, + "GAAATA": 3081, + "GAAATT": 3082, + "GAAATC": 3083, + "GAAATG": 3084, + "GAAACA": 3085, + "GAAACT": 3086, + "GAAACC": 3087, + "GAAACG": 3088, + "GAAAGA": 3089, + "GAAAGT": 3090, + "GAAAGC": 3091, + "GAAAGG": 3092, + "GAATAA": 3093, + "GAATAT": 3094, + "GAATAC": 3095, + "GAATAG": 3096, + "GAATTA": 3097, + "GAATTT": 3098, + "GAATTC": 3099, + "GAATTG": 3100, + "GAATCA": 3101, + "GAATCT": 3102, + "GAATCC": 3103, + "GAATCG": 3104, + "GAATGA": 3105, + "GAATGT": 3106, + "GAATGC": 3107, + "GAATGG": 3108, + "GAACAA": 3109, + "GAACAT": 3110, + "GAACAC": 3111, + "GAACAG": 3112, + "GAACTA": 3113, + "GAACTT": 3114, + "GAACTC": 3115, + "GAACTG": 3116, + "GAACCA": 3117, + "GAACCT": 3118, + "GAACCC": 3119, + "GAACCG": 3120, + "GAACGA": 3121, + "GAACGT": 3122, + "GAACGC": 3123, + "GAACGG": 3124, + "GAAGAA": 3125, + "GAAGAT": 3126, + "GAAGAC": 3127, + "GAAGAG": 3128, + "GAAGTA": 3129, + "GAAGTT": 3130, + "GAAGTC": 3131, + "GAAGTG": 3132, + "GAAGCA": 3133, + "GAAGCT": 3134, + "GAAGCC": 3135, + "GAAGCG": 3136, + "GAAGGA": 3137, + "GAAGGT": 3138, + "GAAGGC": 3139, + "GAAGGG": 3140, + "GATAAA": 3141, + "GATAAT": 3142, + "GATAAC": 3143, + "GATAAG": 3144, + "GATATA": 3145, + "GATATT": 3146, + "GATATC": 3147, + "GATATG": 3148, + "GATACA": 3149, + "GATACT": 3150, + "GATACC": 3151, + "GATACG": 3152, + "GATAGA": 3153, + "GATAGT": 3154, + "GATAGC": 3155, + "GATAGG": 3156, + "GATTAA": 3157, + "GATTAT": 3158, + "GATTAC": 3159, + "GATTAG": 3160, + "GATTTA": 3161, + "GATTTT": 3162, + "GATTTC": 3163, + "GATTTG": 3164, + "GATTCA": 3165, + "GATTCT": 3166, + "GATTCC": 3167, + "GATTCG": 3168, + "GATTGA": 3169, + "GATTGT": 3170, + "GATTGC": 3171, + "GATTGG": 3172, + "GATCAA": 3173, + "GATCAT": 3174, + "GATCAC": 3175, + "GATCAG": 3176, + "GATCTA": 3177, + "GATCTT": 3178, + "GATCTC": 3179, + "GATCTG": 3180, + "GATCCA": 3181, + "GATCCT": 3182, + "GATCCC": 3183, + "GATCCG": 3184, + "GATCGA": 3185, + "GATCGT": 3186, + "GATCGC": 3187, + "GATCGG": 3188, + "GATGAA": 3189, + "GATGAT": 3190, + "GATGAC": 3191, + "GATGAG": 3192, + "GATGTA": 3193, + "GATGTT": 3194, + "GATGTC": 3195, + "GATGTG": 3196, + "GATGCA": 3197, + "GATGCT": 3198, + "GATGCC": 3199, + "GATGCG": 3200, + "GATGGA": 3201, + "GATGGT": 3202, + "GATGGC": 3203, + "GATGGG": 3204, + "GACAAA": 3205, + "GACAAT": 3206, + "GACAAC": 3207, + "GACAAG": 3208, + "GACATA": 3209, + "GACATT": 3210, + "GACATC": 3211, + "GACATG": 3212, + "GACACA": 3213, + "GACACT": 3214, + "GACACC": 3215, + "GACACG": 3216, + "GACAGA": 3217, + "GACAGT": 3218, + "GACAGC": 3219, + "GACAGG": 3220, + "GACTAA": 3221, + "GACTAT": 3222, + "GACTAC": 3223, + "GACTAG": 3224, + "GACTTA": 3225, + "GACTTT": 3226, + "GACTTC": 3227, + "GACTTG": 3228, + "GACTCA": 3229, + "GACTCT": 3230, + "GACTCC": 3231, + "GACTCG": 3232, + "GACTGA": 3233, + "GACTGT": 3234, + "GACTGC": 3235, + "GACTGG": 3236, + "GACCAA": 3237, + "GACCAT": 3238, + "GACCAC": 3239, + "GACCAG": 3240, + "GACCTA": 3241, + "GACCTT": 3242, + "GACCTC": 3243, + "GACCTG": 3244, + "GACCCA": 3245, + "GACCCT": 3246, + "GACCCC": 3247, + "GACCCG": 3248, + "GACCGA": 3249, + "GACCGT": 3250, + "GACCGC": 3251, + "GACCGG": 3252, + "GACGAA": 3253, + "GACGAT": 3254, + "GACGAC": 3255, + "GACGAG": 3256, + "GACGTA": 3257, + "GACGTT": 3258, + "GACGTC": 3259, + "GACGTG": 3260, + "GACGCA": 3261, + "GACGCT": 3262, + "GACGCC": 3263, + "GACGCG": 3264, + "GACGGA": 3265, + "GACGGT": 3266, + "GACGGC": 3267, + "GACGGG": 3268, + "GAGAAA": 3269, + "GAGAAT": 3270, + "GAGAAC": 3271, + "GAGAAG": 3272, + "GAGATA": 3273, + "GAGATT": 3274, + "GAGATC": 3275, + "GAGATG": 3276, + "GAGACA": 3277, + "GAGACT": 3278, + "GAGACC": 3279, + "GAGACG": 3280, + "GAGAGA": 3281, + "GAGAGT": 3282, + "GAGAGC": 3283, + "GAGAGG": 3284, + "GAGTAA": 3285, + "GAGTAT": 3286, + "GAGTAC": 3287, + "GAGTAG": 3288, + "GAGTTA": 3289, + "GAGTTT": 3290, + "GAGTTC": 3291, + "GAGTTG": 3292, + "GAGTCA": 3293, + "GAGTCT": 3294, + "GAGTCC": 3295, + "GAGTCG": 3296, + "GAGTGA": 3297, + "GAGTGT": 3298, + "GAGTGC": 3299, + "GAGTGG": 3300, + "GAGCAA": 3301, + "GAGCAT": 3302, + "GAGCAC": 3303, + "GAGCAG": 3304, + "GAGCTA": 3305, + "GAGCTT": 3306, + "GAGCTC": 3307, + "GAGCTG": 3308, + "GAGCCA": 3309, + "GAGCCT": 3310, + "GAGCCC": 3311, + "GAGCCG": 3312, + "GAGCGA": 3313, + "GAGCGT": 3314, + "GAGCGC": 3315, + "GAGCGG": 3316, + "GAGGAA": 3317, + "GAGGAT": 3318, + "GAGGAC": 3319, + "GAGGAG": 3320, + "GAGGTA": 3321, + "GAGGTT": 3322, + "GAGGTC": 3323, + "GAGGTG": 3324, + "GAGGCA": 3325, + "GAGGCT": 3326, + "GAGGCC": 3327, + "GAGGCG": 3328, + "GAGGGA": 3329, + "GAGGGT": 3330, + "GAGGGC": 3331, + "GAGGGG": 3332, + "GTAAAA": 3333, + "GTAAAT": 3334, + "GTAAAC": 3335, + "GTAAAG": 3336, + "GTAATA": 3337, + "GTAATT": 3338, + "GTAATC": 3339, + "GTAATG": 3340, + "GTAACA": 3341, + "GTAACT": 3342, + "GTAACC": 3343, + "GTAACG": 3344, + "GTAAGA": 3345, + "GTAAGT": 3346, + "GTAAGC": 3347, + "GTAAGG": 3348, + "GTATAA": 3349, + "GTATAT": 3350, + "GTATAC": 3351, + "GTATAG": 3352, + "GTATTA": 3353, + "GTATTT": 3354, + "GTATTC": 3355, + "GTATTG": 3356, + "GTATCA": 3357, + "GTATCT": 3358, + "GTATCC": 3359, + "GTATCG": 3360, + "GTATGA": 3361, + "GTATGT": 3362, + "GTATGC": 3363, + "GTATGG": 3364, + "GTACAA": 3365, + "GTACAT": 3366, + "GTACAC": 3367, + "GTACAG": 3368, + "GTACTA": 3369, + "GTACTT": 3370, + "GTACTC": 3371, + "GTACTG": 3372, + "GTACCA": 3373, + "GTACCT": 3374, + "GTACCC": 3375, + "GTACCG": 3376, + "GTACGA": 3377, + "GTACGT": 3378, + "GTACGC": 3379, + "GTACGG": 3380, + "GTAGAA": 3381, + "GTAGAT": 3382, + "GTAGAC": 3383, + "GTAGAG": 3384, + "GTAGTA": 3385, + "GTAGTT": 3386, + "GTAGTC": 3387, + "GTAGTG": 3388, + "GTAGCA": 3389, + "GTAGCT": 3390, + "GTAGCC": 3391, + "GTAGCG": 3392, + "GTAGGA": 3393, + "GTAGGT": 3394, + "GTAGGC": 3395, + "GTAGGG": 3396, + "GTTAAA": 3397, + "GTTAAT": 3398, + "GTTAAC": 3399, + "GTTAAG": 3400, + "GTTATA": 3401, + "GTTATT": 3402, + "GTTATC": 3403, + "GTTATG": 3404, + "GTTACA": 3405, + "GTTACT": 3406, + "GTTACC": 3407, + "GTTACG": 3408, + "GTTAGA": 3409, + "GTTAGT": 3410, + "GTTAGC": 3411, + "GTTAGG": 3412, + "GTTTAA": 3413, + "GTTTAT": 3414, + "GTTTAC": 3415, + "GTTTAG": 3416, + "GTTTTA": 3417, + "GTTTTT": 3418, + "GTTTTC": 3419, + "GTTTTG": 3420, + "GTTTCA": 3421, + "GTTTCT": 3422, + "GTTTCC": 3423, + "GTTTCG": 3424, + "GTTTGA": 3425, + "GTTTGT": 3426, + "GTTTGC": 3427, + "GTTTGG": 3428, + "GTTCAA": 3429, + "GTTCAT": 3430, + "GTTCAC": 3431, + "GTTCAG": 3432, + "GTTCTA": 3433, + "GTTCTT": 3434, + "GTTCTC": 3435, + "GTTCTG": 3436, + "GTTCCA": 3437, + "GTTCCT": 3438, + "GTTCCC": 3439, + "GTTCCG": 3440, + "GTTCGA": 3441, + "GTTCGT": 3442, + "GTTCGC": 3443, + "GTTCGG": 3444, + "GTTGAA": 3445, + "GTTGAT": 3446, + "GTTGAC": 3447, + "GTTGAG": 3448, + "GTTGTA": 3449, + "GTTGTT": 3450, + "GTTGTC": 3451, + "GTTGTG": 3452, + "GTTGCA": 3453, + "GTTGCT": 3454, + "GTTGCC": 3455, + "GTTGCG": 3456, + "GTTGGA": 3457, + "GTTGGT": 3458, + "GTTGGC": 3459, + "GTTGGG": 3460, + "GTCAAA": 3461, + "GTCAAT": 3462, + "GTCAAC": 3463, + "GTCAAG": 3464, + "GTCATA": 3465, + "GTCATT": 3466, + "GTCATC": 3467, + "GTCATG": 3468, + "GTCACA": 3469, + "GTCACT": 3470, + "GTCACC": 3471, + "GTCACG": 3472, + "GTCAGA": 3473, + "GTCAGT": 3474, + "GTCAGC": 3475, + "GTCAGG": 3476, + "GTCTAA": 3477, + "GTCTAT": 3478, + "GTCTAC": 3479, + "GTCTAG": 3480, + "GTCTTA": 3481, + "GTCTTT": 3482, + "GTCTTC": 3483, + "GTCTTG": 3484, + "GTCTCA": 3485, + "GTCTCT": 3486, + "GTCTCC": 3487, + "GTCTCG": 3488, + "GTCTGA": 3489, + "GTCTGT": 3490, + "GTCTGC": 3491, + "GTCTGG": 3492, + "GTCCAA": 3493, + "GTCCAT": 3494, + "GTCCAC": 3495, + "GTCCAG": 3496, + "GTCCTA": 3497, + "GTCCTT": 3498, + "GTCCTC": 3499, + "GTCCTG": 3500, + "GTCCCA": 3501, + "GTCCCT": 3502, + "GTCCCC": 3503, + "GTCCCG": 3504, + "GTCCGA": 3505, + "GTCCGT": 3506, + "GTCCGC": 3507, + "GTCCGG": 3508, + "GTCGAA": 3509, + "GTCGAT": 3510, + "GTCGAC": 3511, + "GTCGAG": 3512, + "GTCGTA": 3513, + "GTCGTT": 3514, + "GTCGTC": 3515, + "GTCGTG": 3516, + "GTCGCA": 3517, + "GTCGCT": 3518, + "GTCGCC": 3519, + "GTCGCG": 3520, + "GTCGGA": 3521, + "GTCGGT": 3522, + "GTCGGC": 3523, + "GTCGGG": 3524, + "GTGAAA": 3525, + "GTGAAT": 3526, + "GTGAAC": 3527, + "GTGAAG": 3528, + "GTGATA": 3529, + "GTGATT": 3530, + "GTGATC": 3531, + "GTGATG": 3532, + "GTGACA": 3533, + "GTGACT": 3534, + "GTGACC": 3535, + "GTGACG": 3536, + "GTGAGA": 3537, + "GTGAGT": 3538, + "GTGAGC": 3539, + "GTGAGG": 3540, + "GTGTAA": 3541, + "GTGTAT": 3542, + "GTGTAC": 3543, + "GTGTAG": 3544, + "GTGTTA": 3545, + "GTGTTT": 3546, + "GTGTTC": 3547, + "GTGTTG": 3548, + "GTGTCA": 3549, + "GTGTCT": 3550, + "GTGTCC": 3551, + "GTGTCG": 3552, + "GTGTGA": 3553, + "GTGTGT": 3554, + "GTGTGC": 3555, + "GTGTGG": 3556, + "GTGCAA": 3557, + "GTGCAT": 3558, + "GTGCAC": 3559, + "GTGCAG": 3560, + "GTGCTA": 3561, + "GTGCTT": 3562, + "GTGCTC": 3563, + "GTGCTG": 3564, + "GTGCCA": 3565, + "GTGCCT": 3566, + "GTGCCC": 3567, + "GTGCCG": 3568, + "GTGCGA": 3569, + "GTGCGT": 3570, + "GTGCGC": 3571, + "GTGCGG": 3572, + "GTGGAA": 3573, + "GTGGAT": 3574, + "GTGGAC": 3575, + "GTGGAG": 3576, + "GTGGTA": 3577, + "GTGGTT": 3578, + "GTGGTC": 3579, + "GTGGTG": 3580, + "GTGGCA": 3581, + "GTGGCT": 3582, + "GTGGCC": 3583, + "GTGGCG": 3584, + "GTGGGA": 3585, + "GTGGGT": 3586, + "GTGGGC": 3587, + "GTGGGG": 3588, + "GCAAAA": 3589, + "GCAAAT": 3590, + "GCAAAC": 3591, + "GCAAAG": 3592, + "GCAATA": 3593, + "GCAATT": 3594, + "GCAATC": 3595, + "GCAATG": 3596, + "GCAACA": 3597, + "GCAACT": 3598, + "GCAACC": 3599, + "GCAACG": 3600, + "GCAAGA": 3601, + "GCAAGT": 3602, + "GCAAGC": 3603, + "GCAAGG": 3604, + "GCATAA": 3605, + "GCATAT": 3606, + "GCATAC": 3607, + "GCATAG": 3608, + "GCATTA": 3609, + "GCATTT": 3610, + "GCATTC": 3611, + "GCATTG": 3612, + "GCATCA": 3613, + "GCATCT": 3614, + "GCATCC": 3615, + "GCATCG": 3616, + "GCATGA": 3617, + "GCATGT": 3618, + "GCATGC": 3619, + "GCATGG": 3620, + "GCACAA": 3621, + "GCACAT": 3622, + "GCACAC": 3623, + "GCACAG": 3624, + "GCACTA": 3625, + "GCACTT": 3626, + "GCACTC": 3627, + "GCACTG": 3628, + "GCACCA": 3629, + "GCACCT": 3630, + "GCACCC": 3631, + "GCACCG": 3632, + "GCACGA": 3633, + "GCACGT": 3634, + "GCACGC": 3635, + "GCACGG": 3636, + "GCAGAA": 3637, + "GCAGAT": 3638, + "GCAGAC": 3639, + "GCAGAG": 3640, + "GCAGTA": 3641, + "GCAGTT": 3642, + "GCAGTC": 3643, + "GCAGTG": 3644, + "GCAGCA": 3645, + "GCAGCT": 3646, + "GCAGCC": 3647, + "GCAGCG": 3648, + "GCAGGA": 3649, + "GCAGGT": 3650, + "GCAGGC": 3651, + "GCAGGG": 3652, + "GCTAAA": 3653, + "GCTAAT": 3654, + "GCTAAC": 3655, + "GCTAAG": 3656, + "GCTATA": 3657, + "GCTATT": 3658, + "GCTATC": 3659, + "GCTATG": 3660, + "GCTACA": 3661, + "GCTACT": 3662, + "GCTACC": 3663, + "GCTACG": 3664, + "GCTAGA": 3665, + "GCTAGT": 3666, + "GCTAGC": 3667, + "GCTAGG": 3668, + "GCTTAA": 3669, + "GCTTAT": 3670, + "GCTTAC": 3671, + "GCTTAG": 3672, + "GCTTTA": 3673, + "GCTTTT": 3674, + "GCTTTC": 3675, + "GCTTTG": 3676, + "GCTTCA": 3677, + "GCTTCT": 3678, + "GCTTCC": 3679, + "GCTTCG": 3680, + "GCTTGA": 3681, + "GCTTGT": 3682, + "GCTTGC": 3683, + "GCTTGG": 3684, + "GCTCAA": 3685, + "GCTCAT": 3686, + "GCTCAC": 3687, + "GCTCAG": 3688, + "GCTCTA": 3689, + "GCTCTT": 3690, + "GCTCTC": 3691, + "GCTCTG": 3692, + "GCTCCA": 3693, + "GCTCCT": 3694, + "GCTCCC": 3695, + "GCTCCG": 3696, + "GCTCGA": 3697, + "GCTCGT": 3698, + "GCTCGC": 3699, + "GCTCGG": 3700, + "GCTGAA": 3701, + "GCTGAT": 3702, + "GCTGAC": 3703, + "GCTGAG": 3704, + "GCTGTA": 3705, + "GCTGTT": 3706, + "GCTGTC": 3707, + "GCTGTG": 3708, + "GCTGCA": 3709, + "GCTGCT": 3710, + "GCTGCC": 3711, + "GCTGCG": 3712, + "GCTGGA": 3713, + "GCTGGT": 3714, + "GCTGGC": 3715, + "GCTGGG": 3716, + "GCCAAA": 3717, + "GCCAAT": 3718, + "GCCAAC": 3719, + "GCCAAG": 3720, + "GCCATA": 3721, + "GCCATT": 3722, + "GCCATC": 3723, + "GCCATG": 3724, + "GCCACA": 3725, + "GCCACT": 3726, + "GCCACC": 3727, + "GCCACG": 3728, + "GCCAGA": 3729, + "GCCAGT": 3730, + "GCCAGC": 3731, + "GCCAGG": 3732, + "GCCTAA": 3733, + "GCCTAT": 3734, + "GCCTAC": 3735, + "GCCTAG": 3736, + "GCCTTA": 3737, + "GCCTTT": 3738, + "GCCTTC": 3739, + "GCCTTG": 3740, + "GCCTCA": 3741, + "GCCTCT": 3742, + "GCCTCC": 3743, + "GCCTCG": 3744, + "GCCTGA": 3745, + "GCCTGT": 3746, + "GCCTGC": 3747, + "GCCTGG": 3748, + "GCCCAA": 3749, + "GCCCAT": 3750, + "GCCCAC": 3751, + "GCCCAG": 3752, + "GCCCTA": 3753, + "GCCCTT": 3754, + "GCCCTC": 3755, + "GCCCTG": 3756, + "GCCCCA": 3757, + "GCCCCT": 3758, + "GCCCCC": 3759, + "GCCCCG": 3760, + "GCCCGA": 3761, + "GCCCGT": 3762, + "GCCCGC": 3763, + "GCCCGG": 3764, + "GCCGAA": 3765, + "GCCGAT": 3766, + "GCCGAC": 3767, + "GCCGAG": 3768, + "GCCGTA": 3769, + "GCCGTT": 3770, + "GCCGTC": 3771, + "GCCGTG": 3772, + "GCCGCA": 3773, + "GCCGCT": 3774, + "GCCGCC": 3775, + "GCCGCG": 3776, + "GCCGGA": 3777, + "GCCGGT": 3778, + "GCCGGC": 3779, + "GCCGGG": 3780, + "GCGAAA": 3781, + "GCGAAT": 3782, + "GCGAAC": 3783, + "GCGAAG": 3784, + "GCGATA": 3785, + "GCGATT": 3786, + "GCGATC": 3787, + "GCGATG": 3788, + "GCGACA": 3789, + "GCGACT": 3790, + "GCGACC": 3791, + "GCGACG": 3792, + "GCGAGA": 3793, + "GCGAGT": 3794, + "GCGAGC": 3795, + "GCGAGG": 3796, + "GCGTAA": 3797, + "GCGTAT": 3798, + "GCGTAC": 3799, + "GCGTAG": 3800, + "GCGTTA": 3801, + "GCGTTT": 3802, + "GCGTTC": 3803, + "GCGTTG": 3804, + "GCGTCA": 3805, + "GCGTCT": 3806, + "GCGTCC": 3807, + "GCGTCG": 3808, + "GCGTGA": 3809, + "GCGTGT": 3810, + "GCGTGC": 3811, + "GCGTGG": 3812, + "GCGCAA": 3813, + "GCGCAT": 3814, + "GCGCAC": 3815, + "GCGCAG": 3816, + "GCGCTA": 3817, + "GCGCTT": 3818, + "GCGCTC": 3819, + "GCGCTG": 3820, + "GCGCCA": 3821, + "GCGCCT": 3822, + "GCGCCC": 3823, + "GCGCCG": 3824, + "GCGCGA": 3825, + "GCGCGT": 3826, + "GCGCGC": 3827, + "GCGCGG": 3828, + "GCGGAA": 3829, + "GCGGAT": 3830, + "GCGGAC": 3831, + "GCGGAG": 3832, + "GCGGTA": 3833, + "GCGGTT": 3834, + "GCGGTC": 3835, + "GCGGTG": 3836, + "GCGGCA": 3837, + "GCGGCT": 3838, + "GCGGCC": 3839, + "GCGGCG": 3840, + "GCGGGA": 3841, + "GCGGGT": 3842, + "GCGGGC": 3843, + "GCGGGG": 3844, + "GGAAAA": 3845, + "GGAAAT": 3846, + "GGAAAC": 3847, + "GGAAAG": 3848, + "GGAATA": 3849, + "GGAATT": 3850, + "GGAATC": 3851, + "GGAATG": 3852, + "GGAACA": 3853, + "GGAACT": 3854, + "GGAACC": 3855, + "GGAACG": 3856, + "GGAAGA": 3857, + "GGAAGT": 3858, + "GGAAGC": 3859, + "GGAAGG": 3860, + "GGATAA": 3861, + "GGATAT": 3862, + "GGATAC": 3863, + "GGATAG": 3864, + "GGATTA": 3865, + "GGATTT": 3866, + "GGATTC": 3867, + "GGATTG": 3868, + "GGATCA": 3869, + "GGATCT": 3870, + "GGATCC": 3871, + "GGATCG": 3872, + "GGATGA": 3873, + "GGATGT": 3874, + "GGATGC": 3875, + "GGATGG": 3876, + "GGACAA": 3877, + "GGACAT": 3878, + "GGACAC": 3879, + "GGACAG": 3880, + "GGACTA": 3881, + "GGACTT": 3882, + "GGACTC": 3883, + "GGACTG": 3884, + "GGACCA": 3885, + "GGACCT": 3886, + "GGACCC": 3887, + "GGACCG": 3888, + "GGACGA": 3889, + "GGACGT": 3890, + "GGACGC": 3891, + "GGACGG": 3892, + "GGAGAA": 3893, + "GGAGAT": 3894, + "GGAGAC": 3895, + "GGAGAG": 3896, + "GGAGTA": 3897, + "GGAGTT": 3898, + "GGAGTC": 3899, + "GGAGTG": 3900, + "GGAGCA": 3901, + "GGAGCT": 3902, + "GGAGCC": 3903, + "GGAGCG": 3904, + "GGAGGA": 3905, + "GGAGGT": 3906, + "GGAGGC": 3907, + "GGAGGG": 3908, + "GGTAAA": 3909, + "GGTAAT": 3910, + "GGTAAC": 3911, + "GGTAAG": 3912, + "GGTATA": 3913, + "GGTATT": 3914, + "GGTATC": 3915, + "GGTATG": 3916, + "GGTACA": 3917, + "GGTACT": 3918, + "GGTACC": 3919, + "GGTACG": 3920, + "GGTAGA": 3921, + "GGTAGT": 3922, + "GGTAGC": 3923, + "GGTAGG": 3924, + "GGTTAA": 3925, + "GGTTAT": 3926, + "GGTTAC": 3927, + "GGTTAG": 3928, + "GGTTTA": 3929, + "GGTTTT": 3930, + "GGTTTC": 3931, + "GGTTTG": 3932, + "GGTTCA": 3933, + "GGTTCT": 3934, + "GGTTCC": 3935, + "GGTTCG": 3936, + "GGTTGA": 3937, + "GGTTGT": 3938, + "GGTTGC": 3939, + "GGTTGG": 3940, + "GGTCAA": 3941, + "GGTCAT": 3942, + "GGTCAC": 3943, + "GGTCAG": 3944, + "GGTCTA": 3945, + "GGTCTT": 3946, + "GGTCTC": 3947, + "GGTCTG": 3948, + "GGTCCA": 3949, + "GGTCCT": 3950, + "GGTCCC": 3951, + "GGTCCG": 3952, + "GGTCGA": 3953, + "GGTCGT": 3954, + "GGTCGC": 3955, + "GGTCGG": 3956, + "GGTGAA": 3957, + "GGTGAT": 3958, + "GGTGAC": 3959, + "GGTGAG": 3960, + "GGTGTA": 3961, + "GGTGTT": 3962, + "GGTGTC": 3963, + "GGTGTG": 3964, + "GGTGCA": 3965, + "GGTGCT": 3966, + "GGTGCC": 3967, + "GGTGCG": 3968, + "GGTGGA": 3969, + "GGTGGT": 3970, + "GGTGGC": 3971, + "GGTGGG": 3972, + "GGCAAA": 3973, + "GGCAAT": 3974, + "GGCAAC": 3975, + "GGCAAG": 3976, + "GGCATA": 3977, + "GGCATT": 3978, + "GGCATC": 3979, + "GGCATG": 3980, + "GGCACA": 3981, + "GGCACT": 3982, + "GGCACC": 3983, + "GGCACG": 3984, + "GGCAGA": 3985, + "GGCAGT": 3986, + "GGCAGC": 3987, + "GGCAGG": 3988, + "GGCTAA": 3989, + "GGCTAT": 3990, + "GGCTAC": 3991, + "GGCTAG": 3992, + "GGCTTA": 3993, + "GGCTTT": 3994, + "GGCTTC": 3995, + "GGCTTG": 3996, + "GGCTCA": 3997, + "GGCTCT": 3998, + "GGCTCC": 3999, + "GGCTCG": 4000, + "GGCTGA": 4001, + "GGCTGT": 4002, + "GGCTGC": 4003, + "GGCTGG": 4004, + "GGCCAA": 4005, + "GGCCAT": 4006, + "GGCCAC": 4007, + "GGCCAG": 4008, + "GGCCTA": 4009, + "GGCCTT": 4010, + "GGCCTC": 4011, + "GGCCTG": 4012, + "GGCCCA": 4013, + "GGCCCT": 4014, + "GGCCCC": 4015, + "GGCCCG": 4016, + "GGCCGA": 4017, + "GGCCGT": 4018, + "GGCCGC": 4019, + "GGCCGG": 4020, + "GGCGAA": 4021, + "GGCGAT": 4022, + "GGCGAC": 4023, + "GGCGAG": 4024, + "GGCGTA": 4025, + "GGCGTT": 4026, + "GGCGTC": 4027, + "GGCGTG": 4028, + "GGCGCA": 4029, + "GGCGCT": 4030, + "GGCGCC": 4031, + "GGCGCG": 4032, + "GGCGGA": 4033, + "GGCGGT": 4034, + "GGCGGC": 4035, + "GGCGGG": 4036, + "GGGAAA": 4037, + "GGGAAT": 4038, + "GGGAAC": 4039, + "GGGAAG": 4040, + "GGGATA": 4041, + "GGGATT": 4042, + "GGGATC": 4043, + "GGGATG": 4044, + "GGGACA": 4045, + "GGGACT": 4046, + "GGGACC": 4047, + "GGGACG": 4048, + "GGGAGA": 4049, + "GGGAGT": 4050, + "GGGAGC": 4051, + "GGGAGG": 4052, + "GGGTAA": 4053, + "GGGTAT": 4054, + "GGGTAC": 4055, + "GGGTAG": 4056, + "GGGTTA": 4057, + "GGGTTT": 4058, + "GGGTTC": 4059, + "GGGTTG": 4060, + "GGGTCA": 4061, + "GGGTCT": 4062, + "GGGTCC": 4063, + "GGGTCG": 4064, + "GGGTGA": 4065, + "GGGTGT": 4066, + "GGGTGC": 4067, + "GGGTGG": 4068, + "GGGCAA": 4069, + "GGGCAT": 4070, + "GGGCAC": 4071, + "GGGCAG": 4072, + "GGGCTA": 4073, + "GGGCTT": 4074, + "GGGCTC": 4075, + "GGGCTG": 4076, + "GGGCCA": 4077, + "GGGCCT": 4078, + "GGGCCC": 4079, + "GGGCCG": 4080, + "GGGCGA": 4081, + "GGGCGT": 4082, + "GGGCGC": 4083, + "GGGCGG": 4084, + "GGGGAA": 4085, + "GGGGAT": 4086, + "GGGGAC": 4087, + "GGGGAG": 4088, + "GGGGTA": 4089, + "GGGGTT": 4090, + "GGGGTC": 4091, + "GGGGTG": 4092, + "GGGGCA": 4093, + "GGGGCT": 4094, + "GGGGCC": 4095, + "GGGGCG": 4096, + "GGGGGA": 4097, + "GGGGGT": 4098, + "GGGGGC": 4099, + "GGGGGG": 4100 + } + } +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/tokenizer_config.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..604e2b1b29af1dcb3920d2828b6a90276e5e3d02 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/tokenizer_config.json @@ -0,0 +1,58 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "[UNK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "[CLS]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "[SEP]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "4": { + "content": "[MASK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "clean_up_tokenization_spaces": true, + "cls_token": "[CLS]", + "do_basic_tokenize": true, + "do_lower_case": false, + "mask_token": "[MASK]", + "max_len": 512, + "model_max_length": 512, + "never_split": null, + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "strip_accents": null, + "tokenize_chinese_chars": true, + "tokenizer_class": "BertTokenizer", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/trainer_state.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..327101821141f8b19a184445ebf735ce9247fcef --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/trainer_state.json @@ -0,0 +1,313 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 4.0, + "eval_steps": 500, + "global_step": 1000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.1, + "grad_norm": 0.2218121588230133, + "learning_rate": 5.319148936170213e-05, + "loss": 0.6927, + "step": 25 + }, + { + "epoch": 0.2, + "grad_norm": 0.44851255416870117, + "learning_rate": 0.00010638297872340425, + "loss": 0.6929, + "step": 50 + }, + { + "epoch": 0.3, + "grad_norm": 0.2440175861120224, + "learning_rate": 0.00015957446808510637, + "loss": 0.6924, + "step": 75 + }, + { + "epoch": 0.4, + "grad_norm": 0.3422006666660309, + "learning_rate": 0.0002127659574468085, + "loss": 0.6922, + "step": 100 + }, + { + "epoch": 0.5, + "grad_norm": 0.267333447933197, + "learning_rate": 0.00026595744680851064, + "loss": 0.6913, + "step": 125 + }, + { + "epoch": 0.6, + "grad_norm": 0.45699772238731384, + "learning_rate": 0.00031914893617021275, + "loss": 0.6896, + "step": 150 + }, + { + "epoch": 0.7, + "grad_norm": 0.3004947304725647, + "learning_rate": 0.0003723404255319149, + "loss": 0.6887, + "step": 175 + }, + { + "epoch": 0.8, + "grad_norm": 0.33440759778022766, + "learning_rate": 0.0003992081821181128, + "loss": 0.6857, + "step": 200 + }, + { + "epoch": 0.9, + "grad_norm": 0.2918928861618042, + "learning_rate": 0.0003975585615308479, + "loss": 0.6882, + "step": 225 + }, + { + "epoch": 1.0, + "grad_norm": 0.2934085726737976, + "learning_rate": 0.00039590894094358297, + "loss": 0.6807, + "step": 250 + }, + { + "epoch": 1.1, + "grad_norm": 0.3364830017089844, + "learning_rate": 0.0003942593203563181, + "loss": 0.6848, + "step": 275 + }, + { + "epoch": 1.2, + "grad_norm": 0.8379159569740295, + "learning_rate": 0.0003926096997690532, + "loss": 0.677, + "step": 300 + }, + { + "epoch": 1.3, + "grad_norm": 0.37671706080436707, + "learning_rate": 0.00039096007918178817, + "loss": 0.6744, + "step": 325 + }, + { + "epoch": 1.4, + "grad_norm": 0.34329330921173096, + "learning_rate": 0.0003893104585945233, + "loss": 0.6733, + "step": 350 + }, + { + "epoch": 1.5, + "grad_norm": 0.4286085367202759, + "learning_rate": 0.0003876608380072583, + "loss": 0.6772, + "step": 375 + }, + { + "epoch": 1.6, + "grad_norm": 0.3576306998729706, + "learning_rate": 0.00038601121741999343, + "loss": 0.6799, + "step": 400 + }, + { + "epoch": 1.7, + "grad_norm": 0.4636435806751251, + "learning_rate": 0.0003843615968327285, + "loss": 0.6819, + "step": 425 + }, + { + "epoch": 1.8, + "grad_norm": 0.4552135765552521, + "learning_rate": 0.0003827119762454636, + "loss": 0.6731, + "step": 450 + }, + { + "epoch": 1.9, + "grad_norm": 0.3565954566001892, + "learning_rate": 0.00038106235565819863, + "loss": 0.6718, + "step": 475 + }, + { + "epoch": 2.0, + "grad_norm": 0.5451686978340149, + "learning_rate": 0.0003794127350709337, + "loss": 0.6858, + "step": 500 + }, + { + "epoch": 2.1, + "grad_norm": 0.3730587959289551, + "learning_rate": 0.0003777631144836688, + "loss": 0.667, + "step": 525 + }, + { + "epoch": 2.2, + "grad_norm": 0.6642417907714844, + "learning_rate": 0.00037611349389640383, + "loss": 0.6662, + "step": 550 + }, + { + "epoch": 2.3, + "grad_norm": 0.3889809250831604, + "learning_rate": 0.00037446387330913894, + "loss": 0.6718, + "step": 575 + }, + { + "epoch": 2.4, + "grad_norm": 0.47203826904296875, + "learning_rate": 0.000372814252721874, + "loss": 0.6687, + "step": 600 + }, + { + "epoch": 2.5, + "grad_norm": 0.5293862819671631, + "learning_rate": 0.00037116463213460903, + "loss": 0.6655, + "step": 625 + }, + { + "epoch": 2.6, + "grad_norm": 0.36724531650543213, + "learning_rate": 0.00036951501154734414, + "loss": 0.672, + "step": 650 + }, + { + "epoch": 2.7, + "grad_norm": 0.5733087062835693, + "learning_rate": 0.0003678653909600792, + "loss": 0.6683, + "step": 675 + }, + { + "epoch": 2.8, + "grad_norm": 0.6385205984115601, + "learning_rate": 0.0003662157703728143, + "loss": 0.67, + "step": 700 + }, + { + "epoch": 2.9, + "grad_norm": 0.6146281957626343, + "learning_rate": 0.00036456614978554934, + "loss": 0.6606, + "step": 725 + }, + { + "epoch": 3.0, + "grad_norm": 0.6535729169845581, + "learning_rate": 0.00036291652919828444, + "loss": 0.6691, + "step": 750 + }, + { + "epoch": 3.1, + "grad_norm": 0.5194367170333862, + "learning_rate": 0.00036126690861101944, + "loss": 0.6517, + "step": 775 + }, + { + "epoch": 3.2, + "grad_norm": 0.7116692066192627, + "learning_rate": 0.00035961728802375454, + "loss": 0.6501, + "step": 800 + }, + { + "epoch": 3.3, + "grad_norm": 0.5338636040687561, + "learning_rate": 0.0003579676674364896, + "loss": 0.6682, + "step": 825 + }, + { + "epoch": 3.4, + "grad_norm": 0.6912877559661865, + "learning_rate": 0.0003563180468492247, + "loss": 0.6632, + "step": 850 + }, + { + "epoch": 3.5, + "grad_norm": 0.6925108432769775, + "learning_rate": 0.0003546684262619598, + "loss": 0.653, + "step": 875 + }, + { + "epoch": 3.6, + "grad_norm": 0.5360299944877625, + "learning_rate": 0.00035301880567469485, + "loss": 0.6613, + "step": 900 + }, + { + "epoch": 3.7, + "grad_norm": 0.6487969160079956, + "learning_rate": 0.0003513691850874299, + "loss": 0.6627, + "step": 925 + }, + { + "epoch": 3.8, + "grad_norm": 1.7260371446609497, + "learning_rate": 0.00034971956450016495, + "loss": 0.6436, + "step": 950 + }, + { + "epoch": 3.9, + "grad_norm": 0.5852354168891907, + "learning_rate": 0.00034806994391290005, + "loss": 0.6613, + "step": 975 + }, + { + "epoch": 4.0, + "grad_norm": 0.7802712917327881, + "learning_rate": 0.0003464203233256351, + "loss": 0.658, + "step": 1000 + } + ], + "logging_steps": 25, + "max_steps": 6250, + "num_input_tokens_seen": 0, + "num_train_epochs": 25, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 55381032960000.0, + "train_batch_size": 20, + "trial_name": null, + "trial_params": null +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/training_args.bin b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..540db3fbee786996d9705d768d31ba6c371187be --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dc7816b5f3fd38f01f9ba38a197ce070cde06e4c90e5260cb4814f34f628907d +size 5304 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/vocab.txt b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/vocab.txt new file mode 100644 index 0000000000000000000000000000000000000000..cc9ad53732e21ea9180f18e1fb67a38cc5702d4c --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-1000/vocab.txt @@ -0,0 +1,4101 @@ +[PAD] +[UNK] +[CLS] +[SEP] +[MASK] +AAAAAA +AAAAAT +AAAAAC +AAAAAG +AAAATA +AAAATT +AAAATC +AAAATG +AAAACA +AAAACT +AAAACC +AAAACG +AAAAGA +AAAAGT +AAAAGC +AAAAGG +AAATAA +AAATAT +AAATAC +AAATAG +AAATTA +AAATTT +AAATTC +AAATTG +AAATCA +AAATCT +AAATCC +AAATCG +AAATGA +AAATGT +AAATGC +AAATGG +AAACAA +AAACAT +AAACAC +AAACAG +AAACTA +AAACTT +AAACTC +AAACTG +AAACCA +AAACCT +AAACCC +AAACCG +AAACGA +AAACGT +AAACGC +AAACGG +AAAGAA +AAAGAT +AAAGAC +AAAGAG +AAAGTA +AAAGTT +AAAGTC +AAAGTG +AAAGCA +AAAGCT +AAAGCC +AAAGCG +AAAGGA +AAAGGT +AAAGGC +AAAGGG +AATAAA +AATAAT +AATAAC +AATAAG +AATATA +AATATT +AATATC +AATATG +AATACA +AATACT +AATACC +AATACG +AATAGA +AATAGT +AATAGC +AATAGG +AATTAA +AATTAT +AATTAC +AATTAG +AATTTA +AATTTT +AATTTC +AATTTG +AATTCA +AATTCT +AATTCC +AATTCG +AATTGA +AATTGT +AATTGC +AATTGG +AATCAA +AATCAT +AATCAC +AATCAG +AATCTA +AATCTT +AATCTC +AATCTG +AATCCA +AATCCT +AATCCC +AATCCG +AATCGA +AATCGT +AATCGC +AATCGG +AATGAA +AATGAT +AATGAC +AATGAG +AATGTA +AATGTT +AATGTC +AATGTG +AATGCA +AATGCT +AATGCC +AATGCG +AATGGA +AATGGT +AATGGC +AATGGG +AACAAA +AACAAT +AACAAC +AACAAG +AACATA +AACATT +AACATC +AACATG +AACACA +AACACT +AACACC +AACACG +AACAGA +AACAGT +AACAGC +AACAGG +AACTAA +AACTAT +AACTAC +AACTAG +AACTTA +AACTTT +AACTTC +AACTTG +AACTCA +AACTCT +AACTCC +AACTCG +AACTGA +AACTGT +AACTGC +AACTGG +AACCAA +AACCAT +AACCAC +AACCAG +AACCTA +AACCTT +AACCTC +AACCTG +AACCCA +AACCCT +AACCCC +AACCCG +AACCGA +AACCGT +AACCGC +AACCGG +AACGAA +AACGAT +AACGAC +AACGAG +AACGTA +AACGTT +AACGTC +AACGTG +AACGCA +AACGCT +AACGCC +AACGCG +AACGGA +AACGGT +AACGGC +AACGGG +AAGAAA +AAGAAT +AAGAAC +AAGAAG +AAGATA +AAGATT +AAGATC +AAGATG +AAGACA +AAGACT +AAGACC +AAGACG +AAGAGA +AAGAGT +AAGAGC +AAGAGG +AAGTAA +AAGTAT +AAGTAC +AAGTAG +AAGTTA +AAGTTT +AAGTTC +AAGTTG +AAGTCA +AAGTCT +AAGTCC +AAGTCG +AAGTGA +AAGTGT +AAGTGC +AAGTGG +AAGCAA +AAGCAT +AAGCAC +AAGCAG +AAGCTA +AAGCTT +AAGCTC +AAGCTG +AAGCCA +AAGCCT +AAGCCC +AAGCCG +AAGCGA +AAGCGT +AAGCGC +AAGCGG +AAGGAA +AAGGAT +AAGGAC +AAGGAG +AAGGTA +AAGGTT +AAGGTC +AAGGTG +AAGGCA +AAGGCT +AAGGCC +AAGGCG +AAGGGA +AAGGGT +AAGGGC +AAGGGG +ATAAAA +ATAAAT +ATAAAC +ATAAAG +ATAATA +ATAATT +ATAATC +ATAATG +ATAACA +ATAACT +ATAACC +ATAACG +ATAAGA +ATAAGT +ATAAGC +ATAAGG +ATATAA +ATATAT +ATATAC +ATATAG +ATATTA +ATATTT +ATATTC +ATATTG +ATATCA +ATATCT +ATATCC +ATATCG +ATATGA +ATATGT +ATATGC +ATATGG +ATACAA +ATACAT +ATACAC +ATACAG +ATACTA +ATACTT +ATACTC +ATACTG +ATACCA +ATACCT +ATACCC +ATACCG +ATACGA +ATACGT +ATACGC +ATACGG +ATAGAA +ATAGAT +ATAGAC +ATAGAG +ATAGTA +ATAGTT +ATAGTC +ATAGTG +ATAGCA +ATAGCT +ATAGCC +ATAGCG +ATAGGA +ATAGGT +ATAGGC +ATAGGG +ATTAAA +ATTAAT +ATTAAC +ATTAAG +ATTATA +ATTATT +ATTATC +ATTATG +ATTACA +ATTACT +ATTACC +ATTACG +ATTAGA +ATTAGT +ATTAGC +ATTAGG +ATTTAA +ATTTAT +ATTTAC +ATTTAG +ATTTTA +ATTTTT +ATTTTC +ATTTTG +ATTTCA +ATTTCT +ATTTCC +ATTTCG +ATTTGA +ATTTGT +ATTTGC +ATTTGG +ATTCAA +ATTCAT +ATTCAC +ATTCAG +ATTCTA +ATTCTT +ATTCTC +ATTCTG +ATTCCA +ATTCCT +ATTCCC +ATTCCG +ATTCGA +ATTCGT +ATTCGC +ATTCGG +ATTGAA +ATTGAT +ATTGAC +ATTGAG +ATTGTA +ATTGTT +ATTGTC +ATTGTG +ATTGCA +ATTGCT +ATTGCC +ATTGCG +ATTGGA +ATTGGT +ATTGGC +ATTGGG +ATCAAA +ATCAAT +ATCAAC +ATCAAG +ATCATA +ATCATT +ATCATC +ATCATG +ATCACA +ATCACT +ATCACC +ATCACG +ATCAGA +ATCAGT +ATCAGC +ATCAGG +ATCTAA +ATCTAT +ATCTAC +ATCTAG +ATCTTA +ATCTTT +ATCTTC +ATCTTG +ATCTCA +ATCTCT +ATCTCC +ATCTCG +ATCTGA +ATCTGT +ATCTGC +ATCTGG +ATCCAA +ATCCAT +ATCCAC +ATCCAG +ATCCTA +ATCCTT +ATCCTC +ATCCTG +ATCCCA +ATCCCT +ATCCCC +ATCCCG +ATCCGA +ATCCGT +ATCCGC +ATCCGG +ATCGAA +ATCGAT +ATCGAC +ATCGAG +ATCGTA +ATCGTT +ATCGTC +ATCGTG +ATCGCA +ATCGCT +ATCGCC +ATCGCG +ATCGGA +ATCGGT +ATCGGC +ATCGGG +ATGAAA +ATGAAT +ATGAAC +ATGAAG +ATGATA +ATGATT +ATGATC +ATGATG +ATGACA +ATGACT +ATGACC +ATGACG +ATGAGA +ATGAGT +ATGAGC +ATGAGG +ATGTAA +ATGTAT +ATGTAC +ATGTAG +ATGTTA +ATGTTT +ATGTTC +ATGTTG +ATGTCA +ATGTCT +ATGTCC +ATGTCG +ATGTGA +ATGTGT +ATGTGC +ATGTGG +ATGCAA +ATGCAT +ATGCAC +ATGCAG +ATGCTA +ATGCTT +ATGCTC +ATGCTG +ATGCCA +ATGCCT +ATGCCC +ATGCCG +ATGCGA +ATGCGT +ATGCGC +ATGCGG +ATGGAA +ATGGAT +ATGGAC +ATGGAG +ATGGTA +ATGGTT +ATGGTC +ATGGTG +ATGGCA +ATGGCT +ATGGCC +ATGGCG +ATGGGA +ATGGGT +ATGGGC +ATGGGG +ACAAAA +ACAAAT +ACAAAC +ACAAAG +ACAATA +ACAATT +ACAATC +ACAATG +ACAACA +ACAACT +ACAACC +ACAACG +ACAAGA +ACAAGT +ACAAGC +ACAAGG +ACATAA +ACATAT +ACATAC +ACATAG +ACATTA +ACATTT +ACATTC +ACATTG +ACATCA +ACATCT +ACATCC +ACATCG +ACATGA +ACATGT +ACATGC +ACATGG +ACACAA +ACACAT +ACACAC +ACACAG +ACACTA +ACACTT +ACACTC +ACACTG +ACACCA +ACACCT +ACACCC +ACACCG +ACACGA +ACACGT +ACACGC +ACACGG +ACAGAA +ACAGAT +ACAGAC +ACAGAG +ACAGTA +ACAGTT +ACAGTC +ACAGTG +ACAGCA +ACAGCT +ACAGCC +ACAGCG +ACAGGA +ACAGGT +ACAGGC +ACAGGG +ACTAAA +ACTAAT +ACTAAC +ACTAAG +ACTATA +ACTATT +ACTATC +ACTATG +ACTACA +ACTACT +ACTACC +ACTACG +ACTAGA +ACTAGT +ACTAGC +ACTAGG +ACTTAA +ACTTAT +ACTTAC +ACTTAG +ACTTTA +ACTTTT +ACTTTC +ACTTTG +ACTTCA +ACTTCT +ACTTCC +ACTTCG +ACTTGA +ACTTGT +ACTTGC +ACTTGG +ACTCAA +ACTCAT +ACTCAC +ACTCAG +ACTCTA +ACTCTT +ACTCTC +ACTCTG +ACTCCA +ACTCCT +ACTCCC +ACTCCG +ACTCGA +ACTCGT +ACTCGC +ACTCGG +ACTGAA +ACTGAT +ACTGAC +ACTGAG +ACTGTA +ACTGTT +ACTGTC +ACTGTG +ACTGCA +ACTGCT +ACTGCC +ACTGCG +ACTGGA +ACTGGT +ACTGGC +ACTGGG +ACCAAA +ACCAAT +ACCAAC +ACCAAG +ACCATA +ACCATT +ACCATC +ACCATG +ACCACA +ACCACT +ACCACC +ACCACG +ACCAGA +ACCAGT +ACCAGC +ACCAGG +ACCTAA +ACCTAT +ACCTAC +ACCTAG +ACCTTA +ACCTTT +ACCTTC +ACCTTG +ACCTCA +ACCTCT +ACCTCC +ACCTCG +ACCTGA +ACCTGT +ACCTGC +ACCTGG +ACCCAA +ACCCAT +ACCCAC +ACCCAG +ACCCTA +ACCCTT +ACCCTC +ACCCTG +ACCCCA +ACCCCT +ACCCCC +ACCCCG +ACCCGA +ACCCGT +ACCCGC +ACCCGG +ACCGAA +ACCGAT +ACCGAC +ACCGAG +ACCGTA +ACCGTT +ACCGTC +ACCGTG +ACCGCA +ACCGCT +ACCGCC +ACCGCG +ACCGGA +ACCGGT +ACCGGC +ACCGGG +ACGAAA +ACGAAT +ACGAAC +ACGAAG +ACGATA +ACGATT +ACGATC +ACGATG +ACGACA +ACGACT +ACGACC +ACGACG +ACGAGA +ACGAGT +ACGAGC +ACGAGG +ACGTAA +ACGTAT +ACGTAC +ACGTAG +ACGTTA +ACGTTT +ACGTTC +ACGTTG +ACGTCA +ACGTCT +ACGTCC +ACGTCG +ACGTGA +ACGTGT +ACGTGC +ACGTGG +ACGCAA +ACGCAT +ACGCAC +ACGCAG +ACGCTA +ACGCTT +ACGCTC +ACGCTG +ACGCCA +ACGCCT +ACGCCC +ACGCCG +ACGCGA +ACGCGT +ACGCGC +ACGCGG +ACGGAA +ACGGAT +ACGGAC +ACGGAG +ACGGTA +ACGGTT +ACGGTC +ACGGTG +ACGGCA +ACGGCT +ACGGCC +ACGGCG +ACGGGA +ACGGGT +ACGGGC +ACGGGG +AGAAAA +AGAAAT +AGAAAC +AGAAAG +AGAATA +AGAATT +AGAATC +AGAATG +AGAACA +AGAACT +AGAACC +AGAACG +AGAAGA +AGAAGT +AGAAGC +AGAAGG +AGATAA +AGATAT +AGATAC +AGATAG +AGATTA +AGATTT +AGATTC +AGATTG +AGATCA +AGATCT +AGATCC +AGATCG +AGATGA +AGATGT +AGATGC +AGATGG +AGACAA +AGACAT +AGACAC +AGACAG +AGACTA +AGACTT +AGACTC +AGACTG +AGACCA +AGACCT +AGACCC +AGACCG +AGACGA +AGACGT +AGACGC +AGACGG +AGAGAA +AGAGAT +AGAGAC +AGAGAG +AGAGTA +AGAGTT +AGAGTC +AGAGTG +AGAGCA +AGAGCT +AGAGCC +AGAGCG +AGAGGA +AGAGGT +AGAGGC +AGAGGG +AGTAAA +AGTAAT +AGTAAC +AGTAAG +AGTATA +AGTATT +AGTATC +AGTATG +AGTACA +AGTACT +AGTACC +AGTACG +AGTAGA +AGTAGT +AGTAGC +AGTAGG +AGTTAA +AGTTAT +AGTTAC +AGTTAG +AGTTTA +AGTTTT +AGTTTC +AGTTTG +AGTTCA +AGTTCT +AGTTCC +AGTTCG +AGTTGA +AGTTGT +AGTTGC +AGTTGG +AGTCAA +AGTCAT +AGTCAC +AGTCAG +AGTCTA +AGTCTT +AGTCTC +AGTCTG +AGTCCA +AGTCCT +AGTCCC +AGTCCG +AGTCGA +AGTCGT +AGTCGC +AGTCGG +AGTGAA +AGTGAT +AGTGAC +AGTGAG +AGTGTA +AGTGTT +AGTGTC +AGTGTG +AGTGCA +AGTGCT +AGTGCC +AGTGCG +AGTGGA +AGTGGT +AGTGGC +AGTGGG +AGCAAA +AGCAAT +AGCAAC +AGCAAG +AGCATA +AGCATT +AGCATC +AGCATG +AGCACA +AGCACT +AGCACC +AGCACG +AGCAGA +AGCAGT +AGCAGC +AGCAGG +AGCTAA +AGCTAT +AGCTAC +AGCTAG +AGCTTA +AGCTTT +AGCTTC +AGCTTG +AGCTCA +AGCTCT +AGCTCC +AGCTCG +AGCTGA +AGCTGT +AGCTGC +AGCTGG +AGCCAA +AGCCAT +AGCCAC +AGCCAG +AGCCTA +AGCCTT +AGCCTC +AGCCTG +AGCCCA +AGCCCT +AGCCCC +AGCCCG +AGCCGA +AGCCGT +AGCCGC +AGCCGG +AGCGAA +AGCGAT +AGCGAC +AGCGAG +AGCGTA +AGCGTT +AGCGTC +AGCGTG +AGCGCA +AGCGCT +AGCGCC +AGCGCG +AGCGGA +AGCGGT +AGCGGC +AGCGGG +AGGAAA +AGGAAT +AGGAAC +AGGAAG +AGGATA +AGGATT +AGGATC +AGGATG +AGGACA +AGGACT +AGGACC +AGGACG +AGGAGA +AGGAGT +AGGAGC +AGGAGG +AGGTAA +AGGTAT +AGGTAC +AGGTAG +AGGTTA +AGGTTT +AGGTTC +AGGTTG +AGGTCA +AGGTCT +AGGTCC +AGGTCG +AGGTGA +AGGTGT +AGGTGC +AGGTGG +AGGCAA +AGGCAT +AGGCAC +AGGCAG +AGGCTA +AGGCTT +AGGCTC +AGGCTG +AGGCCA +AGGCCT +AGGCCC +AGGCCG +AGGCGA +AGGCGT +AGGCGC +AGGCGG +AGGGAA +AGGGAT +AGGGAC +AGGGAG +AGGGTA +AGGGTT +AGGGTC +AGGGTG +AGGGCA +AGGGCT +AGGGCC +AGGGCG +AGGGGA +AGGGGT +AGGGGC +AGGGGG +TAAAAA +TAAAAT +TAAAAC +TAAAAG +TAAATA +TAAATT +TAAATC +TAAATG +TAAACA +TAAACT +TAAACC +TAAACG +TAAAGA +TAAAGT +TAAAGC +TAAAGG +TAATAA +TAATAT +TAATAC +TAATAG +TAATTA +TAATTT +TAATTC +TAATTG +TAATCA +TAATCT +TAATCC +TAATCG +TAATGA +TAATGT +TAATGC +TAATGG +TAACAA +TAACAT +TAACAC +TAACAG +TAACTA +TAACTT +TAACTC +TAACTG +TAACCA +TAACCT +TAACCC +TAACCG +TAACGA +TAACGT +TAACGC +TAACGG +TAAGAA +TAAGAT +TAAGAC +TAAGAG +TAAGTA +TAAGTT +TAAGTC +TAAGTG +TAAGCA +TAAGCT +TAAGCC +TAAGCG +TAAGGA +TAAGGT +TAAGGC +TAAGGG +TATAAA +TATAAT +TATAAC +TATAAG +TATATA +TATATT +TATATC +TATATG +TATACA +TATACT +TATACC +TATACG +TATAGA +TATAGT +TATAGC +TATAGG +TATTAA +TATTAT +TATTAC +TATTAG +TATTTA +TATTTT +TATTTC +TATTTG +TATTCA +TATTCT +TATTCC +TATTCG +TATTGA +TATTGT +TATTGC +TATTGG +TATCAA +TATCAT +TATCAC +TATCAG +TATCTA +TATCTT +TATCTC +TATCTG +TATCCA +TATCCT +TATCCC +TATCCG +TATCGA +TATCGT +TATCGC +TATCGG +TATGAA +TATGAT +TATGAC +TATGAG +TATGTA +TATGTT +TATGTC +TATGTG +TATGCA +TATGCT +TATGCC +TATGCG +TATGGA +TATGGT +TATGGC +TATGGG +TACAAA +TACAAT +TACAAC +TACAAG +TACATA +TACATT +TACATC +TACATG +TACACA +TACACT +TACACC +TACACG +TACAGA +TACAGT +TACAGC +TACAGG +TACTAA +TACTAT +TACTAC +TACTAG +TACTTA +TACTTT +TACTTC +TACTTG +TACTCA +TACTCT +TACTCC +TACTCG +TACTGA +TACTGT +TACTGC +TACTGG +TACCAA +TACCAT +TACCAC +TACCAG +TACCTA +TACCTT +TACCTC +TACCTG +TACCCA +TACCCT +TACCCC +TACCCG +TACCGA +TACCGT +TACCGC +TACCGG +TACGAA +TACGAT +TACGAC +TACGAG +TACGTA +TACGTT +TACGTC +TACGTG +TACGCA +TACGCT +TACGCC +TACGCG +TACGGA +TACGGT +TACGGC +TACGGG +TAGAAA +TAGAAT +TAGAAC +TAGAAG +TAGATA +TAGATT +TAGATC +TAGATG +TAGACA +TAGACT +TAGACC +TAGACG +TAGAGA +TAGAGT +TAGAGC +TAGAGG +TAGTAA +TAGTAT +TAGTAC +TAGTAG +TAGTTA +TAGTTT +TAGTTC +TAGTTG +TAGTCA +TAGTCT +TAGTCC +TAGTCG +TAGTGA +TAGTGT +TAGTGC +TAGTGG +TAGCAA +TAGCAT +TAGCAC +TAGCAG +TAGCTA +TAGCTT +TAGCTC +TAGCTG +TAGCCA +TAGCCT +TAGCCC +TAGCCG +TAGCGA +TAGCGT +TAGCGC +TAGCGG +TAGGAA +TAGGAT +TAGGAC +TAGGAG +TAGGTA +TAGGTT +TAGGTC +TAGGTG +TAGGCA +TAGGCT +TAGGCC +TAGGCG +TAGGGA +TAGGGT +TAGGGC +TAGGGG +TTAAAA +TTAAAT +TTAAAC +TTAAAG +TTAATA +TTAATT +TTAATC +TTAATG +TTAACA +TTAACT +TTAACC +TTAACG +TTAAGA +TTAAGT +TTAAGC +TTAAGG +TTATAA +TTATAT +TTATAC +TTATAG +TTATTA +TTATTT +TTATTC +TTATTG +TTATCA +TTATCT +TTATCC +TTATCG +TTATGA +TTATGT +TTATGC +TTATGG +TTACAA +TTACAT +TTACAC +TTACAG +TTACTA +TTACTT +TTACTC +TTACTG +TTACCA +TTACCT +TTACCC +TTACCG +TTACGA +TTACGT +TTACGC +TTACGG +TTAGAA +TTAGAT +TTAGAC +TTAGAG +TTAGTA +TTAGTT +TTAGTC +TTAGTG +TTAGCA +TTAGCT +TTAGCC +TTAGCG +TTAGGA +TTAGGT +TTAGGC +TTAGGG +TTTAAA +TTTAAT +TTTAAC +TTTAAG +TTTATA +TTTATT +TTTATC +TTTATG +TTTACA +TTTACT +TTTACC +TTTACG +TTTAGA +TTTAGT +TTTAGC +TTTAGG +TTTTAA +TTTTAT +TTTTAC +TTTTAG +TTTTTA +TTTTTT +TTTTTC +TTTTTG +TTTTCA +TTTTCT +TTTTCC +TTTTCG +TTTTGA +TTTTGT +TTTTGC +TTTTGG +TTTCAA +TTTCAT +TTTCAC +TTTCAG +TTTCTA +TTTCTT +TTTCTC +TTTCTG +TTTCCA +TTTCCT +TTTCCC +TTTCCG +TTTCGA +TTTCGT +TTTCGC +TTTCGG +TTTGAA +TTTGAT +TTTGAC +TTTGAG +TTTGTA +TTTGTT +TTTGTC +TTTGTG +TTTGCA +TTTGCT +TTTGCC +TTTGCG +TTTGGA +TTTGGT +TTTGGC +TTTGGG +TTCAAA +TTCAAT +TTCAAC +TTCAAG +TTCATA +TTCATT +TTCATC +TTCATG +TTCACA +TTCACT +TTCACC +TTCACG +TTCAGA +TTCAGT +TTCAGC +TTCAGG +TTCTAA +TTCTAT +TTCTAC +TTCTAG +TTCTTA +TTCTTT +TTCTTC +TTCTTG +TTCTCA +TTCTCT +TTCTCC +TTCTCG +TTCTGA +TTCTGT +TTCTGC +TTCTGG +TTCCAA +TTCCAT +TTCCAC +TTCCAG +TTCCTA +TTCCTT +TTCCTC +TTCCTG +TTCCCA +TTCCCT +TTCCCC +TTCCCG +TTCCGA +TTCCGT +TTCCGC +TTCCGG +TTCGAA +TTCGAT +TTCGAC +TTCGAG +TTCGTA +TTCGTT +TTCGTC +TTCGTG +TTCGCA +TTCGCT +TTCGCC +TTCGCG +TTCGGA +TTCGGT +TTCGGC +TTCGGG +TTGAAA +TTGAAT +TTGAAC +TTGAAG +TTGATA +TTGATT +TTGATC +TTGATG +TTGACA +TTGACT +TTGACC +TTGACG +TTGAGA +TTGAGT +TTGAGC +TTGAGG +TTGTAA +TTGTAT +TTGTAC +TTGTAG +TTGTTA +TTGTTT +TTGTTC +TTGTTG +TTGTCA +TTGTCT +TTGTCC +TTGTCG +TTGTGA +TTGTGT +TTGTGC +TTGTGG +TTGCAA +TTGCAT +TTGCAC +TTGCAG +TTGCTA +TTGCTT +TTGCTC +TTGCTG +TTGCCA +TTGCCT +TTGCCC +TTGCCG +TTGCGA +TTGCGT +TTGCGC +TTGCGG +TTGGAA +TTGGAT +TTGGAC +TTGGAG +TTGGTA +TTGGTT +TTGGTC +TTGGTG +TTGGCA +TTGGCT +TTGGCC +TTGGCG +TTGGGA +TTGGGT +TTGGGC +TTGGGG +TCAAAA +TCAAAT +TCAAAC +TCAAAG +TCAATA +TCAATT +TCAATC +TCAATG +TCAACA +TCAACT +TCAACC +TCAACG +TCAAGA +TCAAGT +TCAAGC +TCAAGG +TCATAA +TCATAT +TCATAC +TCATAG +TCATTA +TCATTT +TCATTC +TCATTG +TCATCA +TCATCT +TCATCC +TCATCG +TCATGA +TCATGT +TCATGC +TCATGG +TCACAA +TCACAT +TCACAC +TCACAG +TCACTA +TCACTT +TCACTC +TCACTG +TCACCA +TCACCT +TCACCC +TCACCG +TCACGA +TCACGT +TCACGC +TCACGG +TCAGAA +TCAGAT +TCAGAC +TCAGAG +TCAGTA +TCAGTT +TCAGTC +TCAGTG +TCAGCA +TCAGCT +TCAGCC +TCAGCG +TCAGGA +TCAGGT +TCAGGC +TCAGGG +TCTAAA +TCTAAT +TCTAAC +TCTAAG +TCTATA +TCTATT +TCTATC +TCTATG +TCTACA +TCTACT +TCTACC +TCTACG +TCTAGA +TCTAGT +TCTAGC +TCTAGG +TCTTAA +TCTTAT +TCTTAC +TCTTAG +TCTTTA +TCTTTT +TCTTTC +TCTTTG +TCTTCA +TCTTCT +TCTTCC +TCTTCG +TCTTGA +TCTTGT +TCTTGC +TCTTGG +TCTCAA +TCTCAT +TCTCAC +TCTCAG +TCTCTA +TCTCTT +TCTCTC +TCTCTG +TCTCCA +TCTCCT +TCTCCC +TCTCCG +TCTCGA +TCTCGT +TCTCGC +TCTCGG +TCTGAA +TCTGAT +TCTGAC +TCTGAG +TCTGTA +TCTGTT +TCTGTC +TCTGTG +TCTGCA +TCTGCT +TCTGCC +TCTGCG +TCTGGA +TCTGGT +TCTGGC +TCTGGG +TCCAAA +TCCAAT +TCCAAC +TCCAAG +TCCATA +TCCATT +TCCATC +TCCATG +TCCACA +TCCACT +TCCACC +TCCACG +TCCAGA +TCCAGT +TCCAGC +TCCAGG +TCCTAA +TCCTAT +TCCTAC +TCCTAG +TCCTTA +TCCTTT +TCCTTC +TCCTTG +TCCTCA +TCCTCT +TCCTCC +TCCTCG +TCCTGA +TCCTGT +TCCTGC +TCCTGG +TCCCAA +TCCCAT +TCCCAC +TCCCAG +TCCCTA +TCCCTT +TCCCTC +TCCCTG +TCCCCA +TCCCCT +TCCCCC +TCCCCG +TCCCGA +TCCCGT +TCCCGC +TCCCGG +TCCGAA +TCCGAT +TCCGAC +TCCGAG +TCCGTA +TCCGTT +TCCGTC +TCCGTG +TCCGCA +TCCGCT +TCCGCC +TCCGCG +TCCGGA +TCCGGT +TCCGGC +TCCGGG +TCGAAA +TCGAAT +TCGAAC +TCGAAG +TCGATA +TCGATT +TCGATC +TCGATG +TCGACA +TCGACT +TCGACC +TCGACG +TCGAGA +TCGAGT +TCGAGC +TCGAGG +TCGTAA +TCGTAT +TCGTAC +TCGTAG +TCGTTA +TCGTTT +TCGTTC +TCGTTG +TCGTCA +TCGTCT +TCGTCC +TCGTCG +TCGTGA +TCGTGT +TCGTGC +TCGTGG +TCGCAA +TCGCAT +TCGCAC +TCGCAG +TCGCTA +TCGCTT +TCGCTC +TCGCTG +TCGCCA +TCGCCT +TCGCCC +TCGCCG +TCGCGA +TCGCGT +TCGCGC +TCGCGG +TCGGAA +TCGGAT +TCGGAC +TCGGAG +TCGGTA +TCGGTT +TCGGTC +TCGGTG +TCGGCA +TCGGCT +TCGGCC +TCGGCG +TCGGGA +TCGGGT +TCGGGC +TCGGGG +TGAAAA +TGAAAT +TGAAAC +TGAAAG +TGAATA +TGAATT +TGAATC +TGAATG +TGAACA +TGAACT +TGAACC +TGAACG +TGAAGA +TGAAGT +TGAAGC +TGAAGG +TGATAA +TGATAT +TGATAC +TGATAG +TGATTA +TGATTT +TGATTC +TGATTG +TGATCA +TGATCT +TGATCC +TGATCG +TGATGA +TGATGT +TGATGC +TGATGG +TGACAA +TGACAT +TGACAC +TGACAG +TGACTA +TGACTT +TGACTC +TGACTG +TGACCA +TGACCT +TGACCC +TGACCG +TGACGA +TGACGT +TGACGC +TGACGG +TGAGAA +TGAGAT +TGAGAC +TGAGAG +TGAGTA +TGAGTT +TGAGTC +TGAGTG +TGAGCA +TGAGCT +TGAGCC +TGAGCG +TGAGGA +TGAGGT +TGAGGC +TGAGGG +TGTAAA +TGTAAT +TGTAAC +TGTAAG +TGTATA +TGTATT +TGTATC +TGTATG +TGTACA +TGTACT +TGTACC +TGTACG +TGTAGA +TGTAGT +TGTAGC +TGTAGG +TGTTAA +TGTTAT +TGTTAC +TGTTAG +TGTTTA +TGTTTT +TGTTTC +TGTTTG +TGTTCA +TGTTCT +TGTTCC +TGTTCG +TGTTGA +TGTTGT +TGTTGC +TGTTGG +TGTCAA +TGTCAT +TGTCAC +TGTCAG +TGTCTA +TGTCTT +TGTCTC +TGTCTG +TGTCCA +TGTCCT +TGTCCC +TGTCCG +TGTCGA +TGTCGT +TGTCGC +TGTCGG +TGTGAA +TGTGAT +TGTGAC +TGTGAG +TGTGTA +TGTGTT +TGTGTC +TGTGTG +TGTGCA +TGTGCT +TGTGCC +TGTGCG +TGTGGA +TGTGGT +TGTGGC +TGTGGG +TGCAAA +TGCAAT +TGCAAC +TGCAAG +TGCATA +TGCATT +TGCATC +TGCATG +TGCACA +TGCACT +TGCACC +TGCACG +TGCAGA +TGCAGT +TGCAGC +TGCAGG +TGCTAA +TGCTAT +TGCTAC +TGCTAG +TGCTTA +TGCTTT +TGCTTC +TGCTTG +TGCTCA +TGCTCT +TGCTCC +TGCTCG +TGCTGA +TGCTGT +TGCTGC +TGCTGG +TGCCAA +TGCCAT +TGCCAC +TGCCAG +TGCCTA +TGCCTT +TGCCTC +TGCCTG +TGCCCA +TGCCCT +TGCCCC +TGCCCG +TGCCGA +TGCCGT +TGCCGC +TGCCGG +TGCGAA +TGCGAT +TGCGAC +TGCGAG +TGCGTA +TGCGTT +TGCGTC +TGCGTG +TGCGCA +TGCGCT +TGCGCC +TGCGCG +TGCGGA +TGCGGT +TGCGGC +TGCGGG +TGGAAA +TGGAAT +TGGAAC +TGGAAG +TGGATA +TGGATT +TGGATC +TGGATG +TGGACA +TGGACT +TGGACC +TGGACG +TGGAGA +TGGAGT +TGGAGC +TGGAGG +TGGTAA +TGGTAT +TGGTAC +TGGTAG +TGGTTA +TGGTTT +TGGTTC +TGGTTG +TGGTCA +TGGTCT +TGGTCC +TGGTCG +TGGTGA +TGGTGT +TGGTGC +TGGTGG +TGGCAA +TGGCAT +TGGCAC +TGGCAG +TGGCTA +TGGCTT +TGGCTC +TGGCTG +TGGCCA +TGGCCT +TGGCCC +TGGCCG +TGGCGA +TGGCGT +TGGCGC +TGGCGG +TGGGAA +TGGGAT +TGGGAC +TGGGAG +TGGGTA +TGGGTT +TGGGTC +TGGGTG +TGGGCA +TGGGCT +TGGGCC +TGGGCG +TGGGGA +TGGGGT +TGGGGC +TGGGGG +CAAAAA +CAAAAT +CAAAAC +CAAAAG +CAAATA +CAAATT +CAAATC +CAAATG +CAAACA +CAAACT +CAAACC +CAAACG +CAAAGA +CAAAGT +CAAAGC +CAAAGG +CAATAA +CAATAT +CAATAC +CAATAG +CAATTA +CAATTT +CAATTC +CAATTG +CAATCA +CAATCT +CAATCC +CAATCG +CAATGA +CAATGT +CAATGC +CAATGG +CAACAA +CAACAT +CAACAC +CAACAG +CAACTA +CAACTT +CAACTC +CAACTG +CAACCA +CAACCT +CAACCC +CAACCG +CAACGA +CAACGT +CAACGC +CAACGG +CAAGAA +CAAGAT +CAAGAC +CAAGAG +CAAGTA +CAAGTT +CAAGTC +CAAGTG +CAAGCA +CAAGCT +CAAGCC +CAAGCG +CAAGGA +CAAGGT +CAAGGC +CAAGGG +CATAAA +CATAAT +CATAAC +CATAAG +CATATA +CATATT +CATATC +CATATG +CATACA +CATACT +CATACC +CATACG +CATAGA +CATAGT +CATAGC +CATAGG +CATTAA +CATTAT +CATTAC +CATTAG +CATTTA +CATTTT +CATTTC +CATTTG +CATTCA +CATTCT +CATTCC +CATTCG +CATTGA +CATTGT +CATTGC +CATTGG +CATCAA +CATCAT +CATCAC +CATCAG +CATCTA +CATCTT +CATCTC +CATCTG +CATCCA +CATCCT +CATCCC +CATCCG +CATCGA +CATCGT +CATCGC +CATCGG +CATGAA +CATGAT +CATGAC +CATGAG +CATGTA +CATGTT +CATGTC +CATGTG +CATGCA +CATGCT +CATGCC +CATGCG +CATGGA +CATGGT +CATGGC +CATGGG +CACAAA +CACAAT +CACAAC +CACAAG +CACATA +CACATT +CACATC +CACATG +CACACA +CACACT +CACACC +CACACG +CACAGA +CACAGT +CACAGC +CACAGG +CACTAA +CACTAT +CACTAC +CACTAG +CACTTA +CACTTT +CACTTC +CACTTG +CACTCA +CACTCT +CACTCC +CACTCG +CACTGA +CACTGT +CACTGC +CACTGG +CACCAA +CACCAT +CACCAC +CACCAG +CACCTA +CACCTT +CACCTC +CACCTG +CACCCA +CACCCT +CACCCC +CACCCG +CACCGA +CACCGT +CACCGC +CACCGG +CACGAA +CACGAT +CACGAC +CACGAG +CACGTA +CACGTT +CACGTC +CACGTG +CACGCA +CACGCT +CACGCC +CACGCG +CACGGA +CACGGT +CACGGC +CACGGG +CAGAAA +CAGAAT +CAGAAC +CAGAAG +CAGATA +CAGATT +CAGATC +CAGATG +CAGACA +CAGACT +CAGACC +CAGACG +CAGAGA +CAGAGT +CAGAGC +CAGAGG +CAGTAA +CAGTAT +CAGTAC +CAGTAG +CAGTTA +CAGTTT +CAGTTC +CAGTTG +CAGTCA +CAGTCT +CAGTCC +CAGTCG +CAGTGA +CAGTGT +CAGTGC +CAGTGG +CAGCAA +CAGCAT +CAGCAC +CAGCAG +CAGCTA +CAGCTT +CAGCTC +CAGCTG +CAGCCA +CAGCCT +CAGCCC +CAGCCG +CAGCGA +CAGCGT +CAGCGC +CAGCGG +CAGGAA +CAGGAT +CAGGAC +CAGGAG +CAGGTA +CAGGTT +CAGGTC +CAGGTG +CAGGCA +CAGGCT +CAGGCC +CAGGCG +CAGGGA +CAGGGT +CAGGGC +CAGGGG +CTAAAA +CTAAAT +CTAAAC +CTAAAG +CTAATA +CTAATT +CTAATC +CTAATG +CTAACA +CTAACT +CTAACC +CTAACG +CTAAGA +CTAAGT +CTAAGC +CTAAGG +CTATAA +CTATAT +CTATAC +CTATAG +CTATTA +CTATTT +CTATTC +CTATTG +CTATCA +CTATCT +CTATCC +CTATCG +CTATGA +CTATGT +CTATGC +CTATGG +CTACAA +CTACAT +CTACAC +CTACAG +CTACTA +CTACTT +CTACTC +CTACTG +CTACCA +CTACCT +CTACCC +CTACCG +CTACGA +CTACGT +CTACGC +CTACGG +CTAGAA +CTAGAT +CTAGAC +CTAGAG +CTAGTA +CTAGTT +CTAGTC +CTAGTG +CTAGCA +CTAGCT +CTAGCC +CTAGCG +CTAGGA +CTAGGT +CTAGGC +CTAGGG +CTTAAA +CTTAAT +CTTAAC +CTTAAG +CTTATA +CTTATT +CTTATC +CTTATG +CTTACA +CTTACT +CTTACC +CTTACG +CTTAGA +CTTAGT +CTTAGC +CTTAGG +CTTTAA +CTTTAT +CTTTAC +CTTTAG +CTTTTA +CTTTTT +CTTTTC +CTTTTG +CTTTCA +CTTTCT +CTTTCC +CTTTCG +CTTTGA +CTTTGT +CTTTGC +CTTTGG +CTTCAA +CTTCAT +CTTCAC +CTTCAG +CTTCTA +CTTCTT +CTTCTC +CTTCTG +CTTCCA +CTTCCT +CTTCCC +CTTCCG +CTTCGA +CTTCGT +CTTCGC +CTTCGG +CTTGAA +CTTGAT +CTTGAC +CTTGAG +CTTGTA +CTTGTT +CTTGTC +CTTGTG +CTTGCA +CTTGCT +CTTGCC +CTTGCG +CTTGGA +CTTGGT +CTTGGC +CTTGGG +CTCAAA +CTCAAT +CTCAAC +CTCAAG +CTCATA +CTCATT +CTCATC +CTCATG +CTCACA +CTCACT +CTCACC +CTCACG +CTCAGA +CTCAGT +CTCAGC +CTCAGG +CTCTAA +CTCTAT +CTCTAC +CTCTAG +CTCTTA +CTCTTT +CTCTTC +CTCTTG +CTCTCA +CTCTCT +CTCTCC +CTCTCG +CTCTGA +CTCTGT +CTCTGC +CTCTGG +CTCCAA +CTCCAT +CTCCAC +CTCCAG +CTCCTA +CTCCTT +CTCCTC +CTCCTG +CTCCCA +CTCCCT +CTCCCC +CTCCCG +CTCCGA +CTCCGT +CTCCGC +CTCCGG +CTCGAA +CTCGAT +CTCGAC +CTCGAG +CTCGTA +CTCGTT +CTCGTC +CTCGTG +CTCGCA +CTCGCT +CTCGCC +CTCGCG +CTCGGA +CTCGGT +CTCGGC +CTCGGG +CTGAAA +CTGAAT +CTGAAC +CTGAAG +CTGATA +CTGATT +CTGATC +CTGATG +CTGACA +CTGACT +CTGACC +CTGACG +CTGAGA +CTGAGT +CTGAGC +CTGAGG +CTGTAA +CTGTAT +CTGTAC +CTGTAG +CTGTTA +CTGTTT +CTGTTC +CTGTTG +CTGTCA +CTGTCT +CTGTCC +CTGTCG +CTGTGA +CTGTGT +CTGTGC +CTGTGG +CTGCAA +CTGCAT +CTGCAC +CTGCAG +CTGCTA +CTGCTT +CTGCTC +CTGCTG +CTGCCA +CTGCCT +CTGCCC +CTGCCG +CTGCGA +CTGCGT +CTGCGC +CTGCGG +CTGGAA +CTGGAT +CTGGAC +CTGGAG +CTGGTA +CTGGTT +CTGGTC +CTGGTG +CTGGCA +CTGGCT +CTGGCC +CTGGCG +CTGGGA +CTGGGT +CTGGGC +CTGGGG +CCAAAA +CCAAAT +CCAAAC +CCAAAG +CCAATA +CCAATT +CCAATC +CCAATG +CCAACA +CCAACT +CCAACC +CCAACG +CCAAGA +CCAAGT +CCAAGC +CCAAGG +CCATAA +CCATAT +CCATAC +CCATAG +CCATTA +CCATTT +CCATTC +CCATTG +CCATCA +CCATCT +CCATCC +CCATCG +CCATGA +CCATGT +CCATGC +CCATGG +CCACAA +CCACAT +CCACAC +CCACAG +CCACTA +CCACTT +CCACTC +CCACTG +CCACCA +CCACCT +CCACCC +CCACCG +CCACGA +CCACGT +CCACGC +CCACGG +CCAGAA +CCAGAT +CCAGAC +CCAGAG +CCAGTA +CCAGTT +CCAGTC +CCAGTG +CCAGCA +CCAGCT +CCAGCC +CCAGCG +CCAGGA +CCAGGT +CCAGGC +CCAGGG +CCTAAA +CCTAAT +CCTAAC +CCTAAG +CCTATA +CCTATT +CCTATC +CCTATG +CCTACA +CCTACT +CCTACC +CCTACG +CCTAGA +CCTAGT +CCTAGC +CCTAGG +CCTTAA +CCTTAT +CCTTAC +CCTTAG +CCTTTA +CCTTTT +CCTTTC +CCTTTG +CCTTCA +CCTTCT +CCTTCC +CCTTCG +CCTTGA +CCTTGT +CCTTGC +CCTTGG +CCTCAA +CCTCAT +CCTCAC +CCTCAG +CCTCTA +CCTCTT +CCTCTC +CCTCTG +CCTCCA +CCTCCT +CCTCCC +CCTCCG +CCTCGA +CCTCGT +CCTCGC +CCTCGG +CCTGAA +CCTGAT +CCTGAC +CCTGAG +CCTGTA +CCTGTT +CCTGTC +CCTGTG +CCTGCA +CCTGCT +CCTGCC +CCTGCG +CCTGGA +CCTGGT +CCTGGC +CCTGGG +CCCAAA +CCCAAT +CCCAAC +CCCAAG +CCCATA +CCCATT +CCCATC +CCCATG +CCCACA +CCCACT +CCCACC +CCCACG +CCCAGA +CCCAGT +CCCAGC +CCCAGG +CCCTAA +CCCTAT +CCCTAC +CCCTAG +CCCTTA +CCCTTT +CCCTTC +CCCTTG +CCCTCA +CCCTCT +CCCTCC +CCCTCG +CCCTGA +CCCTGT +CCCTGC +CCCTGG +CCCCAA +CCCCAT +CCCCAC +CCCCAG +CCCCTA +CCCCTT +CCCCTC +CCCCTG +CCCCCA +CCCCCT +CCCCCC +CCCCCG +CCCCGA +CCCCGT +CCCCGC +CCCCGG +CCCGAA +CCCGAT +CCCGAC +CCCGAG +CCCGTA +CCCGTT +CCCGTC +CCCGTG +CCCGCA +CCCGCT +CCCGCC +CCCGCG +CCCGGA +CCCGGT +CCCGGC +CCCGGG +CCGAAA +CCGAAT +CCGAAC +CCGAAG +CCGATA +CCGATT +CCGATC +CCGATG +CCGACA +CCGACT +CCGACC +CCGACG +CCGAGA +CCGAGT +CCGAGC +CCGAGG +CCGTAA +CCGTAT +CCGTAC +CCGTAG +CCGTTA +CCGTTT +CCGTTC +CCGTTG +CCGTCA +CCGTCT +CCGTCC +CCGTCG +CCGTGA +CCGTGT +CCGTGC +CCGTGG +CCGCAA +CCGCAT +CCGCAC +CCGCAG +CCGCTA +CCGCTT +CCGCTC +CCGCTG +CCGCCA +CCGCCT +CCGCCC +CCGCCG +CCGCGA +CCGCGT +CCGCGC +CCGCGG +CCGGAA +CCGGAT +CCGGAC +CCGGAG +CCGGTA +CCGGTT +CCGGTC +CCGGTG +CCGGCA +CCGGCT +CCGGCC +CCGGCG +CCGGGA +CCGGGT +CCGGGC +CCGGGG +CGAAAA +CGAAAT +CGAAAC +CGAAAG +CGAATA +CGAATT +CGAATC +CGAATG +CGAACA +CGAACT +CGAACC +CGAACG +CGAAGA +CGAAGT +CGAAGC +CGAAGG +CGATAA +CGATAT +CGATAC +CGATAG +CGATTA +CGATTT +CGATTC +CGATTG +CGATCA +CGATCT +CGATCC +CGATCG +CGATGA +CGATGT +CGATGC +CGATGG +CGACAA +CGACAT +CGACAC +CGACAG +CGACTA +CGACTT +CGACTC +CGACTG +CGACCA +CGACCT +CGACCC +CGACCG +CGACGA +CGACGT +CGACGC +CGACGG +CGAGAA +CGAGAT +CGAGAC +CGAGAG +CGAGTA +CGAGTT +CGAGTC +CGAGTG +CGAGCA +CGAGCT +CGAGCC +CGAGCG +CGAGGA +CGAGGT +CGAGGC +CGAGGG +CGTAAA +CGTAAT +CGTAAC +CGTAAG +CGTATA +CGTATT +CGTATC +CGTATG +CGTACA +CGTACT +CGTACC +CGTACG +CGTAGA +CGTAGT +CGTAGC +CGTAGG +CGTTAA +CGTTAT +CGTTAC +CGTTAG +CGTTTA +CGTTTT +CGTTTC +CGTTTG +CGTTCA +CGTTCT +CGTTCC +CGTTCG +CGTTGA +CGTTGT +CGTTGC +CGTTGG +CGTCAA +CGTCAT +CGTCAC +CGTCAG +CGTCTA +CGTCTT +CGTCTC +CGTCTG +CGTCCA +CGTCCT +CGTCCC +CGTCCG +CGTCGA +CGTCGT +CGTCGC +CGTCGG +CGTGAA +CGTGAT +CGTGAC +CGTGAG +CGTGTA +CGTGTT +CGTGTC +CGTGTG +CGTGCA +CGTGCT +CGTGCC +CGTGCG +CGTGGA +CGTGGT +CGTGGC +CGTGGG +CGCAAA +CGCAAT +CGCAAC +CGCAAG +CGCATA +CGCATT +CGCATC +CGCATG +CGCACA +CGCACT +CGCACC +CGCACG +CGCAGA +CGCAGT +CGCAGC +CGCAGG +CGCTAA +CGCTAT +CGCTAC +CGCTAG +CGCTTA +CGCTTT +CGCTTC +CGCTTG +CGCTCA +CGCTCT +CGCTCC +CGCTCG +CGCTGA +CGCTGT +CGCTGC +CGCTGG +CGCCAA +CGCCAT +CGCCAC +CGCCAG +CGCCTA +CGCCTT +CGCCTC +CGCCTG +CGCCCA +CGCCCT +CGCCCC +CGCCCG +CGCCGA +CGCCGT +CGCCGC +CGCCGG +CGCGAA +CGCGAT +CGCGAC +CGCGAG +CGCGTA +CGCGTT +CGCGTC +CGCGTG +CGCGCA +CGCGCT +CGCGCC +CGCGCG +CGCGGA +CGCGGT +CGCGGC +CGCGGG +CGGAAA +CGGAAT +CGGAAC +CGGAAG +CGGATA +CGGATT +CGGATC +CGGATG +CGGACA +CGGACT +CGGACC +CGGACG +CGGAGA +CGGAGT +CGGAGC +CGGAGG +CGGTAA +CGGTAT +CGGTAC +CGGTAG +CGGTTA +CGGTTT +CGGTTC +CGGTTG +CGGTCA +CGGTCT +CGGTCC +CGGTCG +CGGTGA +CGGTGT +CGGTGC +CGGTGG +CGGCAA +CGGCAT +CGGCAC +CGGCAG +CGGCTA +CGGCTT +CGGCTC +CGGCTG +CGGCCA +CGGCCT +CGGCCC +CGGCCG +CGGCGA +CGGCGT +CGGCGC +CGGCGG +CGGGAA +CGGGAT +CGGGAC +CGGGAG +CGGGTA +CGGGTT +CGGGTC +CGGGTG +CGGGCA +CGGGCT +CGGGCC +CGGGCG +CGGGGA +CGGGGT +CGGGGC +CGGGGG +GAAAAA +GAAAAT +GAAAAC +GAAAAG +GAAATA +GAAATT +GAAATC +GAAATG +GAAACA +GAAACT +GAAACC +GAAACG +GAAAGA +GAAAGT +GAAAGC +GAAAGG +GAATAA +GAATAT +GAATAC +GAATAG +GAATTA +GAATTT +GAATTC +GAATTG +GAATCA +GAATCT +GAATCC +GAATCG +GAATGA +GAATGT +GAATGC +GAATGG +GAACAA +GAACAT +GAACAC +GAACAG +GAACTA +GAACTT +GAACTC +GAACTG +GAACCA +GAACCT +GAACCC +GAACCG +GAACGA +GAACGT +GAACGC +GAACGG +GAAGAA +GAAGAT +GAAGAC +GAAGAG +GAAGTA +GAAGTT +GAAGTC +GAAGTG +GAAGCA +GAAGCT +GAAGCC +GAAGCG +GAAGGA +GAAGGT +GAAGGC +GAAGGG +GATAAA +GATAAT +GATAAC +GATAAG +GATATA +GATATT +GATATC +GATATG +GATACA +GATACT +GATACC +GATACG +GATAGA +GATAGT +GATAGC +GATAGG +GATTAA +GATTAT +GATTAC +GATTAG +GATTTA +GATTTT +GATTTC +GATTTG +GATTCA +GATTCT +GATTCC +GATTCG +GATTGA +GATTGT +GATTGC +GATTGG +GATCAA +GATCAT +GATCAC +GATCAG +GATCTA +GATCTT +GATCTC +GATCTG +GATCCA +GATCCT +GATCCC +GATCCG +GATCGA +GATCGT +GATCGC +GATCGG +GATGAA +GATGAT +GATGAC +GATGAG +GATGTA +GATGTT +GATGTC +GATGTG +GATGCA +GATGCT +GATGCC +GATGCG +GATGGA +GATGGT +GATGGC +GATGGG +GACAAA +GACAAT +GACAAC +GACAAG +GACATA +GACATT +GACATC +GACATG +GACACA +GACACT +GACACC +GACACG +GACAGA +GACAGT +GACAGC +GACAGG +GACTAA +GACTAT +GACTAC +GACTAG +GACTTA +GACTTT +GACTTC +GACTTG +GACTCA +GACTCT +GACTCC +GACTCG +GACTGA +GACTGT +GACTGC +GACTGG +GACCAA +GACCAT +GACCAC +GACCAG +GACCTA +GACCTT +GACCTC +GACCTG +GACCCA +GACCCT +GACCCC +GACCCG +GACCGA +GACCGT +GACCGC +GACCGG +GACGAA +GACGAT +GACGAC +GACGAG +GACGTA +GACGTT +GACGTC +GACGTG +GACGCA +GACGCT +GACGCC +GACGCG +GACGGA +GACGGT +GACGGC +GACGGG +GAGAAA +GAGAAT +GAGAAC +GAGAAG +GAGATA +GAGATT +GAGATC +GAGATG +GAGACA +GAGACT +GAGACC +GAGACG +GAGAGA +GAGAGT +GAGAGC +GAGAGG +GAGTAA +GAGTAT +GAGTAC +GAGTAG +GAGTTA +GAGTTT +GAGTTC +GAGTTG +GAGTCA +GAGTCT +GAGTCC +GAGTCG +GAGTGA +GAGTGT +GAGTGC +GAGTGG +GAGCAA +GAGCAT +GAGCAC +GAGCAG +GAGCTA +GAGCTT +GAGCTC +GAGCTG +GAGCCA +GAGCCT +GAGCCC +GAGCCG +GAGCGA +GAGCGT +GAGCGC +GAGCGG +GAGGAA +GAGGAT +GAGGAC +GAGGAG +GAGGTA +GAGGTT +GAGGTC +GAGGTG +GAGGCA +GAGGCT +GAGGCC +GAGGCG +GAGGGA +GAGGGT +GAGGGC +GAGGGG +GTAAAA +GTAAAT +GTAAAC +GTAAAG +GTAATA +GTAATT +GTAATC +GTAATG +GTAACA +GTAACT +GTAACC +GTAACG +GTAAGA +GTAAGT +GTAAGC +GTAAGG +GTATAA +GTATAT +GTATAC +GTATAG +GTATTA +GTATTT +GTATTC +GTATTG +GTATCA +GTATCT +GTATCC +GTATCG +GTATGA +GTATGT +GTATGC +GTATGG +GTACAA +GTACAT +GTACAC +GTACAG +GTACTA +GTACTT +GTACTC +GTACTG +GTACCA +GTACCT +GTACCC +GTACCG +GTACGA +GTACGT +GTACGC +GTACGG +GTAGAA +GTAGAT +GTAGAC +GTAGAG +GTAGTA +GTAGTT +GTAGTC +GTAGTG +GTAGCA +GTAGCT +GTAGCC +GTAGCG +GTAGGA +GTAGGT +GTAGGC +GTAGGG +GTTAAA +GTTAAT +GTTAAC +GTTAAG +GTTATA +GTTATT +GTTATC +GTTATG +GTTACA +GTTACT +GTTACC +GTTACG +GTTAGA +GTTAGT +GTTAGC +GTTAGG +GTTTAA +GTTTAT +GTTTAC +GTTTAG +GTTTTA +GTTTTT +GTTTTC +GTTTTG +GTTTCA +GTTTCT +GTTTCC +GTTTCG +GTTTGA +GTTTGT +GTTTGC +GTTTGG +GTTCAA +GTTCAT +GTTCAC +GTTCAG +GTTCTA +GTTCTT +GTTCTC +GTTCTG +GTTCCA +GTTCCT +GTTCCC +GTTCCG +GTTCGA +GTTCGT +GTTCGC +GTTCGG +GTTGAA +GTTGAT +GTTGAC +GTTGAG +GTTGTA +GTTGTT +GTTGTC +GTTGTG +GTTGCA +GTTGCT +GTTGCC +GTTGCG +GTTGGA +GTTGGT +GTTGGC +GTTGGG +GTCAAA +GTCAAT +GTCAAC +GTCAAG +GTCATA +GTCATT +GTCATC +GTCATG +GTCACA +GTCACT +GTCACC +GTCACG +GTCAGA +GTCAGT +GTCAGC +GTCAGG +GTCTAA +GTCTAT +GTCTAC +GTCTAG +GTCTTA +GTCTTT +GTCTTC +GTCTTG +GTCTCA +GTCTCT +GTCTCC +GTCTCG +GTCTGA +GTCTGT +GTCTGC +GTCTGG +GTCCAA +GTCCAT +GTCCAC +GTCCAG +GTCCTA +GTCCTT +GTCCTC +GTCCTG +GTCCCA +GTCCCT +GTCCCC +GTCCCG +GTCCGA +GTCCGT +GTCCGC +GTCCGG +GTCGAA +GTCGAT +GTCGAC +GTCGAG +GTCGTA +GTCGTT +GTCGTC +GTCGTG +GTCGCA +GTCGCT +GTCGCC +GTCGCG +GTCGGA +GTCGGT +GTCGGC +GTCGGG +GTGAAA +GTGAAT +GTGAAC +GTGAAG +GTGATA +GTGATT +GTGATC +GTGATG +GTGACA +GTGACT +GTGACC +GTGACG +GTGAGA +GTGAGT +GTGAGC +GTGAGG +GTGTAA +GTGTAT +GTGTAC +GTGTAG +GTGTTA +GTGTTT +GTGTTC +GTGTTG +GTGTCA +GTGTCT +GTGTCC +GTGTCG +GTGTGA +GTGTGT +GTGTGC +GTGTGG +GTGCAA +GTGCAT +GTGCAC +GTGCAG +GTGCTA +GTGCTT +GTGCTC +GTGCTG +GTGCCA +GTGCCT +GTGCCC +GTGCCG +GTGCGA +GTGCGT +GTGCGC +GTGCGG +GTGGAA +GTGGAT +GTGGAC +GTGGAG +GTGGTA +GTGGTT +GTGGTC +GTGGTG +GTGGCA +GTGGCT +GTGGCC +GTGGCG +GTGGGA +GTGGGT +GTGGGC +GTGGGG +GCAAAA +GCAAAT +GCAAAC +GCAAAG +GCAATA +GCAATT +GCAATC +GCAATG +GCAACA +GCAACT +GCAACC +GCAACG +GCAAGA +GCAAGT +GCAAGC +GCAAGG +GCATAA +GCATAT +GCATAC +GCATAG +GCATTA +GCATTT +GCATTC +GCATTG +GCATCA +GCATCT +GCATCC +GCATCG +GCATGA +GCATGT +GCATGC +GCATGG +GCACAA +GCACAT +GCACAC +GCACAG +GCACTA +GCACTT +GCACTC +GCACTG +GCACCA +GCACCT +GCACCC +GCACCG +GCACGA +GCACGT +GCACGC +GCACGG +GCAGAA +GCAGAT +GCAGAC +GCAGAG +GCAGTA +GCAGTT +GCAGTC +GCAGTG +GCAGCA +GCAGCT +GCAGCC +GCAGCG +GCAGGA +GCAGGT +GCAGGC +GCAGGG +GCTAAA +GCTAAT +GCTAAC +GCTAAG +GCTATA +GCTATT +GCTATC +GCTATG +GCTACA +GCTACT +GCTACC +GCTACG +GCTAGA +GCTAGT +GCTAGC +GCTAGG +GCTTAA +GCTTAT +GCTTAC +GCTTAG +GCTTTA +GCTTTT +GCTTTC +GCTTTG +GCTTCA +GCTTCT +GCTTCC +GCTTCG +GCTTGA +GCTTGT +GCTTGC +GCTTGG +GCTCAA +GCTCAT +GCTCAC +GCTCAG +GCTCTA +GCTCTT +GCTCTC +GCTCTG +GCTCCA +GCTCCT +GCTCCC +GCTCCG +GCTCGA +GCTCGT +GCTCGC +GCTCGG +GCTGAA +GCTGAT +GCTGAC +GCTGAG +GCTGTA +GCTGTT +GCTGTC +GCTGTG +GCTGCA +GCTGCT +GCTGCC +GCTGCG +GCTGGA +GCTGGT +GCTGGC +GCTGGG +GCCAAA +GCCAAT +GCCAAC +GCCAAG +GCCATA +GCCATT +GCCATC +GCCATG +GCCACA +GCCACT +GCCACC +GCCACG +GCCAGA +GCCAGT +GCCAGC +GCCAGG +GCCTAA +GCCTAT +GCCTAC +GCCTAG +GCCTTA +GCCTTT +GCCTTC +GCCTTG +GCCTCA +GCCTCT +GCCTCC +GCCTCG +GCCTGA +GCCTGT +GCCTGC +GCCTGG +GCCCAA +GCCCAT +GCCCAC +GCCCAG +GCCCTA +GCCCTT +GCCCTC +GCCCTG +GCCCCA +GCCCCT +GCCCCC +GCCCCG +GCCCGA +GCCCGT +GCCCGC +GCCCGG +GCCGAA +GCCGAT +GCCGAC +GCCGAG +GCCGTA +GCCGTT +GCCGTC +GCCGTG +GCCGCA +GCCGCT +GCCGCC +GCCGCG +GCCGGA +GCCGGT +GCCGGC +GCCGGG +GCGAAA +GCGAAT +GCGAAC +GCGAAG +GCGATA +GCGATT +GCGATC +GCGATG +GCGACA +GCGACT +GCGACC +GCGACG +GCGAGA +GCGAGT +GCGAGC +GCGAGG +GCGTAA +GCGTAT +GCGTAC +GCGTAG +GCGTTA +GCGTTT +GCGTTC +GCGTTG +GCGTCA +GCGTCT +GCGTCC +GCGTCG +GCGTGA +GCGTGT +GCGTGC +GCGTGG +GCGCAA +GCGCAT +GCGCAC +GCGCAG +GCGCTA +GCGCTT +GCGCTC +GCGCTG +GCGCCA +GCGCCT +GCGCCC +GCGCCG +GCGCGA +GCGCGT +GCGCGC +GCGCGG +GCGGAA +GCGGAT +GCGGAC +GCGGAG +GCGGTA +GCGGTT +GCGGTC +GCGGTG +GCGGCA +GCGGCT +GCGGCC +GCGGCG +GCGGGA +GCGGGT +GCGGGC +GCGGGG +GGAAAA +GGAAAT +GGAAAC +GGAAAG +GGAATA +GGAATT +GGAATC +GGAATG +GGAACA +GGAACT +GGAACC +GGAACG +GGAAGA +GGAAGT +GGAAGC +GGAAGG +GGATAA +GGATAT +GGATAC +GGATAG +GGATTA +GGATTT +GGATTC +GGATTG +GGATCA +GGATCT +GGATCC +GGATCG +GGATGA +GGATGT +GGATGC +GGATGG +GGACAA +GGACAT +GGACAC +GGACAG +GGACTA +GGACTT +GGACTC +GGACTG +GGACCA +GGACCT +GGACCC +GGACCG +GGACGA +GGACGT +GGACGC +GGACGG +GGAGAA +GGAGAT +GGAGAC +GGAGAG +GGAGTA +GGAGTT +GGAGTC +GGAGTG +GGAGCA +GGAGCT +GGAGCC +GGAGCG +GGAGGA +GGAGGT +GGAGGC +GGAGGG +GGTAAA +GGTAAT +GGTAAC +GGTAAG +GGTATA +GGTATT +GGTATC +GGTATG +GGTACA +GGTACT +GGTACC +GGTACG +GGTAGA +GGTAGT +GGTAGC +GGTAGG +GGTTAA +GGTTAT +GGTTAC +GGTTAG +GGTTTA +GGTTTT +GGTTTC +GGTTTG +GGTTCA +GGTTCT +GGTTCC +GGTTCG +GGTTGA +GGTTGT +GGTTGC +GGTTGG +GGTCAA +GGTCAT +GGTCAC +GGTCAG +GGTCTA +GGTCTT +GGTCTC +GGTCTG +GGTCCA +GGTCCT +GGTCCC +GGTCCG +GGTCGA +GGTCGT +GGTCGC +GGTCGG +GGTGAA +GGTGAT +GGTGAC +GGTGAG +GGTGTA +GGTGTT +GGTGTC +GGTGTG +GGTGCA +GGTGCT +GGTGCC +GGTGCG +GGTGGA +GGTGGT +GGTGGC +GGTGGG +GGCAAA +GGCAAT +GGCAAC +GGCAAG +GGCATA +GGCATT +GGCATC +GGCATG +GGCACA +GGCACT +GGCACC +GGCACG +GGCAGA +GGCAGT +GGCAGC +GGCAGG +GGCTAA +GGCTAT +GGCTAC +GGCTAG +GGCTTA +GGCTTT +GGCTTC +GGCTTG +GGCTCA +GGCTCT +GGCTCC +GGCTCG +GGCTGA +GGCTGT +GGCTGC +GGCTGG +GGCCAA +GGCCAT +GGCCAC +GGCCAG +GGCCTA +GGCCTT +GGCCTC +GGCCTG +GGCCCA +GGCCCT +GGCCCC +GGCCCG +GGCCGA +GGCCGT +GGCCGC +GGCCGG +GGCGAA +GGCGAT +GGCGAC +GGCGAG +GGCGTA +GGCGTT +GGCGTC +GGCGTG +GGCGCA +GGCGCT +GGCGCC +GGCGCG +GGCGGA +GGCGGT +GGCGGC +GGCGGG +GGGAAA +GGGAAT +GGGAAC +GGGAAG +GGGATA +GGGATT +GGGATC +GGGATG +GGGACA +GGGACT +GGGACC +GGGACG +GGGAGA +GGGAGT +GGGAGC +GGGAGG +GGGTAA +GGGTAT +GGGTAC +GGGTAG +GGGTTA +GGGTTT +GGGTTC +GGGTTG +GGGTCA +GGGTCT +GGGTCC +GGGTCG +GGGTGA +GGGTGT +GGGTGC +GGGTGG +GGGCAA +GGGCAT +GGGCAC +GGGCAG +GGGCTA +GGGCTT +GGGCTC +GGGCTG +GGGCCA +GGGCCT +GGGCCC +GGGCCG +GGGCGA +GGGCGT +GGGCGC +GGGCGG +GGGGAA +GGGGAT +GGGGAC +GGGGAG +GGGGTA +GGGGTT +GGGGTC +GGGGTG +GGGGCA +GGGGCT +GGGGCC +GGGGCG +GGGGGA +GGGGGT +GGGGGC +GGGGGG diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/README.md b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..500fd976d92b217455f181bc98c77e612d9a18d9 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/README.md @@ -0,0 +1,202 @@ +--- +base_model: fabihamakhdoomi/TinyDNABERT +library_name: peft +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.12.0 \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/adapter_config.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..04c63e99bbc6b810449dece367c0afb165125b92 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/adapter_config.json @@ -0,0 +1,42 @@ +{ + "alpha_pattern": {}, + "auto_mapping": null, + "base_model_name_or_path": "fabihamakhdoomi/TinyDNABERT", + "bias": "none", + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 16, + "lora_dropout": 0.2, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": [ + "classifier", + "score" + ], + "peft_type": "LORA", + "r": 8, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "bert.encoder.layer.1.attention.self.query", + "bert.encoder.layer.0.output.dense", + "bert.encoder.layer.1.output.dense", + "bert.encoder.layer.0.intermediate.dense", + "bert.encoder.layer.1.intermediate.dense", + "bert.encoder.layer.0.attention.output.dense", + "bert.encoder.layer.1.attention.output.dense", + "bert.encoder.layer.0.attention.self.query", + "bert.encoder.layer.0.attention.self.value", + "bert.encoder.layer.1.attention.self.key", + "bert.encoder.layer.1.attention.self.value", + "bert.encoder.layer.0.attention.self.key" + ], + "task_type": "SEQ_CLS", + "use_dora": false, + "use_rslora": false +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/adapter_model.safetensors b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/adapter_model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..f9e87070aabaa8d8a971c0601d6cc03042977aa6 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/adapter_model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:90c0bda524a3174254d76c06807032eaeccb06847800920f92a03845d63e7cdf +size 151952 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/optimizer.pt b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..d2d97f8907b9852388452b4dafd1fa2d417415ff --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1196fc38989d464cbf69db906084a8365f010c49d03ed0d44d9656e78c555614 +size 312682 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/rng_state.pth b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..f70e226d95e8a928fcaafa095fe05a0982a8afb3 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fed19b5a5643714b6623a6032bf908ff668a150f2ea673a45992e0ae4072b338 +size 14244 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/scheduler.pt b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..2287c4e5e02bffea6969638a53c3d5402c6dc2b0 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:da119a7607c7d2d96a92ebe57fbaf5fb22ebfce3399b1702b2209f460c0bfe35 +size 1064 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/special_tokens_map.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..a8b3208c2884c4efb86e49300fdd3dc877220cdf --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/special_tokens_map.json @@ -0,0 +1,7 @@ +{ + "cls_token": "[CLS]", + "mask_token": "[MASK]", + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/tokenizer.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..ebd782de6963b9032ce9374cfe30f5275be1a30b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/tokenizer.json @@ -0,0 +1,4265 @@ +{ + "version": "1.0", + "truncation": { + "direction": "Right", + "max_length": 512, + "strategy": "LongestFirst", + "stride": 0 + }, + "padding": { + "strategy": { + "Fixed": 512 + }, + "direction": "Right", + "pad_to_multiple_of": null, + "pad_id": 0, + "pad_type_id": 0, + "pad_token": "[PAD]" + }, + "added_tokens": [ + { + "id": 0, + "content": "[PAD]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 1, + "content": "[UNK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 2, + "content": "[CLS]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 3, + "content": "[SEP]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 4, + "content": "[MASK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + } + ], + "normalizer": { + "type": "BertNormalizer", + "clean_text": true, + "handle_chinese_chars": true, + "strip_accents": null, + "lowercase": false + }, + "pre_tokenizer": { + "type": "BertPreTokenizer" + }, + "post_processor": { + "type": "TemplateProcessing", + "single": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + } + ], + "pair": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "B", + "type_id": 1 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 1 + } + } + ], + "special_tokens": { + "[CLS]": { + "id": "[CLS]", + "ids": [ + 2 + ], + "tokens": [ + "[CLS]" + ] + }, + "[SEP]": { + "id": "[SEP]", + "ids": [ + 3 + ], + "tokens": [ + "[SEP]" + ] + } + } + }, + "decoder": { + "type": "WordPiece", + "prefix": "##", + "cleanup": true + }, + "model": { + "type": "WordPiece", + "unk_token": "[UNK]", + "continuing_subword_prefix": "##", + "max_input_chars_per_word": 100, + "vocab": { + "[PAD]": 0, + "[UNK]": 1, + "[CLS]": 2, + "[SEP]": 3, + "[MASK]": 4, + "AAAAAA": 5, + "AAAAAT": 6, + "AAAAAC": 7, + "AAAAAG": 8, + "AAAATA": 9, + "AAAATT": 10, + "AAAATC": 11, + "AAAATG": 12, + "AAAACA": 13, + "AAAACT": 14, + "AAAACC": 15, + "AAAACG": 16, + "AAAAGA": 17, + "AAAAGT": 18, + "AAAAGC": 19, + "AAAAGG": 20, + "AAATAA": 21, + "AAATAT": 22, + "AAATAC": 23, + "AAATAG": 24, + "AAATTA": 25, + "AAATTT": 26, + "AAATTC": 27, + "AAATTG": 28, + "AAATCA": 29, + "AAATCT": 30, + "AAATCC": 31, + "AAATCG": 32, + "AAATGA": 33, + "AAATGT": 34, + "AAATGC": 35, + "AAATGG": 36, + "AAACAA": 37, + "AAACAT": 38, + "AAACAC": 39, + "AAACAG": 40, + "AAACTA": 41, + "AAACTT": 42, + "AAACTC": 43, + "AAACTG": 44, + "AAACCA": 45, + "AAACCT": 46, + "AAACCC": 47, + "AAACCG": 48, + "AAACGA": 49, + "AAACGT": 50, + "AAACGC": 51, + "AAACGG": 52, + "AAAGAA": 53, + "AAAGAT": 54, + "AAAGAC": 55, + "AAAGAG": 56, + "AAAGTA": 57, + "AAAGTT": 58, + "AAAGTC": 59, + "AAAGTG": 60, + "AAAGCA": 61, + "AAAGCT": 62, + "AAAGCC": 63, + "AAAGCG": 64, + "AAAGGA": 65, + "AAAGGT": 66, + "AAAGGC": 67, + "AAAGGG": 68, + "AATAAA": 69, + "AATAAT": 70, + "AATAAC": 71, + "AATAAG": 72, + "AATATA": 73, + "AATATT": 74, + "AATATC": 75, + "AATATG": 76, + "AATACA": 77, + "AATACT": 78, + "AATACC": 79, + "AATACG": 80, + "AATAGA": 81, + "AATAGT": 82, + "AATAGC": 83, + "AATAGG": 84, + "AATTAA": 85, + "AATTAT": 86, + "AATTAC": 87, + "AATTAG": 88, + "AATTTA": 89, + "AATTTT": 90, + "AATTTC": 91, + "AATTTG": 92, + "AATTCA": 93, + "AATTCT": 94, + "AATTCC": 95, + "AATTCG": 96, + "AATTGA": 97, + "AATTGT": 98, + "AATTGC": 99, + "AATTGG": 100, + "AATCAA": 101, + "AATCAT": 102, + "AATCAC": 103, + "AATCAG": 104, + "AATCTA": 105, + "AATCTT": 106, + "AATCTC": 107, + "AATCTG": 108, + "AATCCA": 109, + "AATCCT": 110, + "AATCCC": 111, + "AATCCG": 112, + "AATCGA": 113, + "AATCGT": 114, + "AATCGC": 115, + "AATCGG": 116, + "AATGAA": 117, + "AATGAT": 118, + "AATGAC": 119, + "AATGAG": 120, + "AATGTA": 121, + "AATGTT": 122, + "AATGTC": 123, + "AATGTG": 124, + "AATGCA": 125, + "AATGCT": 126, + "AATGCC": 127, + "AATGCG": 128, + "AATGGA": 129, + "AATGGT": 130, + "AATGGC": 131, + "AATGGG": 132, + "AACAAA": 133, + "AACAAT": 134, + "AACAAC": 135, + "AACAAG": 136, + "AACATA": 137, + "AACATT": 138, + "AACATC": 139, + "AACATG": 140, + "AACACA": 141, + "AACACT": 142, + "AACACC": 143, + "AACACG": 144, + "AACAGA": 145, + "AACAGT": 146, + "AACAGC": 147, + "AACAGG": 148, + "AACTAA": 149, + "AACTAT": 150, + "AACTAC": 151, + "AACTAG": 152, + "AACTTA": 153, + "AACTTT": 154, + "AACTTC": 155, + "AACTTG": 156, + "AACTCA": 157, + "AACTCT": 158, + "AACTCC": 159, + "AACTCG": 160, + "AACTGA": 161, + "AACTGT": 162, + "AACTGC": 163, + "AACTGG": 164, + "AACCAA": 165, + "AACCAT": 166, + "AACCAC": 167, + "AACCAG": 168, + "AACCTA": 169, + "AACCTT": 170, + "AACCTC": 171, + "AACCTG": 172, + "AACCCA": 173, + "AACCCT": 174, + "AACCCC": 175, + "AACCCG": 176, + "AACCGA": 177, + "AACCGT": 178, + "AACCGC": 179, + "AACCGG": 180, + "AACGAA": 181, + "AACGAT": 182, + "AACGAC": 183, + "AACGAG": 184, + "AACGTA": 185, + "AACGTT": 186, + "AACGTC": 187, + "AACGTG": 188, + "AACGCA": 189, + "AACGCT": 190, + "AACGCC": 191, + "AACGCG": 192, + "AACGGA": 193, + "AACGGT": 194, + "AACGGC": 195, + "AACGGG": 196, + "AAGAAA": 197, + "AAGAAT": 198, + "AAGAAC": 199, + "AAGAAG": 200, + "AAGATA": 201, + "AAGATT": 202, + "AAGATC": 203, + "AAGATG": 204, + "AAGACA": 205, + "AAGACT": 206, + "AAGACC": 207, + "AAGACG": 208, + "AAGAGA": 209, + "AAGAGT": 210, + "AAGAGC": 211, + "AAGAGG": 212, + "AAGTAA": 213, + "AAGTAT": 214, + "AAGTAC": 215, + "AAGTAG": 216, + "AAGTTA": 217, + "AAGTTT": 218, + "AAGTTC": 219, + "AAGTTG": 220, + "AAGTCA": 221, + "AAGTCT": 222, + "AAGTCC": 223, + "AAGTCG": 224, + "AAGTGA": 225, + "AAGTGT": 226, + "AAGTGC": 227, + "AAGTGG": 228, + "AAGCAA": 229, + "AAGCAT": 230, + "AAGCAC": 231, + "AAGCAG": 232, + "AAGCTA": 233, + "AAGCTT": 234, + "AAGCTC": 235, + "AAGCTG": 236, + "AAGCCA": 237, + "AAGCCT": 238, + "AAGCCC": 239, + "AAGCCG": 240, + "AAGCGA": 241, + "AAGCGT": 242, + "AAGCGC": 243, + "AAGCGG": 244, + "AAGGAA": 245, + "AAGGAT": 246, + "AAGGAC": 247, + "AAGGAG": 248, + "AAGGTA": 249, + "AAGGTT": 250, + "AAGGTC": 251, + "AAGGTG": 252, + "AAGGCA": 253, + "AAGGCT": 254, + "AAGGCC": 255, + "AAGGCG": 256, + "AAGGGA": 257, + "AAGGGT": 258, + "AAGGGC": 259, + "AAGGGG": 260, + "ATAAAA": 261, + "ATAAAT": 262, + "ATAAAC": 263, + "ATAAAG": 264, + "ATAATA": 265, + "ATAATT": 266, + "ATAATC": 267, + "ATAATG": 268, + "ATAACA": 269, + "ATAACT": 270, + "ATAACC": 271, + "ATAACG": 272, + "ATAAGA": 273, + "ATAAGT": 274, + "ATAAGC": 275, + "ATAAGG": 276, + "ATATAA": 277, + "ATATAT": 278, + "ATATAC": 279, + "ATATAG": 280, + "ATATTA": 281, + "ATATTT": 282, + "ATATTC": 283, + "ATATTG": 284, + "ATATCA": 285, + "ATATCT": 286, + "ATATCC": 287, + "ATATCG": 288, + "ATATGA": 289, + "ATATGT": 290, + "ATATGC": 291, + "ATATGG": 292, + "ATACAA": 293, + "ATACAT": 294, + "ATACAC": 295, + "ATACAG": 296, + "ATACTA": 297, + "ATACTT": 298, + "ATACTC": 299, + "ATACTG": 300, + "ATACCA": 301, + "ATACCT": 302, + "ATACCC": 303, + "ATACCG": 304, + "ATACGA": 305, + "ATACGT": 306, + "ATACGC": 307, + "ATACGG": 308, + "ATAGAA": 309, + "ATAGAT": 310, + "ATAGAC": 311, + "ATAGAG": 312, + "ATAGTA": 313, + "ATAGTT": 314, + "ATAGTC": 315, + "ATAGTG": 316, + "ATAGCA": 317, + "ATAGCT": 318, + "ATAGCC": 319, + "ATAGCG": 320, + "ATAGGA": 321, + "ATAGGT": 322, + "ATAGGC": 323, + "ATAGGG": 324, + "ATTAAA": 325, + "ATTAAT": 326, + "ATTAAC": 327, + "ATTAAG": 328, + "ATTATA": 329, + "ATTATT": 330, + "ATTATC": 331, + "ATTATG": 332, + "ATTACA": 333, + "ATTACT": 334, + "ATTACC": 335, + "ATTACG": 336, + "ATTAGA": 337, + "ATTAGT": 338, + "ATTAGC": 339, + "ATTAGG": 340, + "ATTTAA": 341, + "ATTTAT": 342, + "ATTTAC": 343, + "ATTTAG": 344, + "ATTTTA": 345, + "ATTTTT": 346, + "ATTTTC": 347, + "ATTTTG": 348, + "ATTTCA": 349, + "ATTTCT": 350, + "ATTTCC": 351, + "ATTTCG": 352, + "ATTTGA": 353, + "ATTTGT": 354, + "ATTTGC": 355, + "ATTTGG": 356, + "ATTCAA": 357, + "ATTCAT": 358, + "ATTCAC": 359, + "ATTCAG": 360, + "ATTCTA": 361, + "ATTCTT": 362, + "ATTCTC": 363, + "ATTCTG": 364, + "ATTCCA": 365, + "ATTCCT": 366, + "ATTCCC": 367, + "ATTCCG": 368, + "ATTCGA": 369, + "ATTCGT": 370, + "ATTCGC": 371, + "ATTCGG": 372, + "ATTGAA": 373, + "ATTGAT": 374, + "ATTGAC": 375, + "ATTGAG": 376, + "ATTGTA": 377, + "ATTGTT": 378, + "ATTGTC": 379, + "ATTGTG": 380, + "ATTGCA": 381, + "ATTGCT": 382, + "ATTGCC": 383, + "ATTGCG": 384, + "ATTGGA": 385, + "ATTGGT": 386, + "ATTGGC": 387, + "ATTGGG": 388, + "ATCAAA": 389, + "ATCAAT": 390, + "ATCAAC": 391, + "ATCAAG": 392, + "ATCATA": 393, + "ATCATT": 394, + "ATCATC": 395, + "ATCATG": 396, + "ATCACA": 397, + "ATCACT": 398, + "ATCACC": 399, + "ATCACG": 400, + "ATCAGA": 401, + "ATCAGT": 402, + "ATCAGC": 403, + "ATCAGG": 404, + "ATCTAA": 405, + "ATCTAT": 406, + "ATCTAC": 407, + "ATCTAG": 408, + "ATCTTA": 409, + "ATCTTT": 410, + "ATCTTC": 411, + "ATCTTG": 412, + "ATCTCA": 413, + "ATCTCT": 414, + "ATCTCC": 415, + "ATCTCG": 416, + "ATCTGA": 417, + "ATCTGT": 418, + "ATCTGC": 419, + "ATCTGG": 420, + "ATCCAA": 421, + "ATCCAT": 422, + "ATCCAC": 423, + "ATCCAG": 424, + "ATCCTA": 425, + "ATCCTT": 426, + "ATCCTC": 427, + "ATCCTG": 428, + "ATCCCA": 429, + "ATCCCT": 430, + "ATCCCC": 431, + "ATCCCG": 432, + "ATCCGA": 433, + "ATCCGT": 434, + "ATCCGC": 435, + "ATCCGG": 436, + "ATCGAA": 437, + "ATCGAT": 438, + "ATCGAC": 439, + "ATCGAG": 440, + "ATCGTA": 441, + "ATCGTT": 442, + "ATCGTC": 443, + "ATCGTG": 444, + "ATCGCA": 445, + "ATCGCT": 446, + "ATCGCC": 447, + "ATCGCG": 448, + "ATCGGA": 449, + "ATCGGT": 450, + "ATCGGC": 451, + "ATCGGG": 452, + "ATGAAA": 453, + "ATGAAT": 454, + "ATGAAC": 455, + "ATGAAG": 456, + "ATGATA": 457, + "ATGATT": 458, + "ATGATC": 459, + "ATGATG": 460, + "ATGACA": 461, + "ATGACT": 462, + "ATGACC": 463, + "ATGACG": 464, + "ATGAGA": 465, + "ATGAGT": 466, + "ATGAGC": 467, + "ATGAGG": 468, + "ATGTAA": 469, + "ATGTAT": 470, + "ATGTAC": 471, + "ATGTAG": 472, + "ATGTTA": 473, + "ATGTTT": 474, + "ATGTTC": 475, + "ATGTTG": 476, + "ATGTCA": 477, + "ATGTCT": 478, + "ATGTCC": 479, + "ATGTCG": 480, + "ATGTGA": 481, + "ATGTGT": 482, + "ATGTGC": 483, + "ATGTGG": 484, + "ATGCAA": 485, + "ATGCAT": 486, + "ATGCAC": 487, + "ATGCAG": 488, + "ATGCTA": 489, + "ATGCTT": 490, + "ATGCTC": 491, + "ATGCTG": 492, + "ATGCCA": 493, + "ATGCCT": 494, + "ATGCCC": 495, + "ATGCCG": 496, + "ATGCGA": 497, + "ATGCGT": 498, + "ATGCGC": 499, + "ATGCGG": 500, + "ATGGAA": 501, + "ATGGAT": 502, + "ATGGAC": 503, + "ATGGAG": 504, + "ATGGTA": 505, + "ATGGTT": 506, + "ATGGTC": 507, + "ATGGTG": 508, + "ATGGCA": 509, + "ATGGCT": 510, + "ATGGCC": 511, + "ATGGCG": 512, + "ATGGGA": 513, + "ATGGGT": 514, + "ATGGGC": 515, + "ATGGGG": 516, + "ACAAAA": 517, + "ACAAAT": 518, + "ACAAAC": 519, + "ACAAAG": 520, + "ACAATA": 521, + "ACAATT": 522, + "ACAATC": 523, + "ACAATG": 524, + "ACAACA": 525, + "ACAACT": 526, + "ACAACC": 527, + "ACAACG": 528, + "ACAAGA": 529, + "ACAAGT": 530, + "ACAAGC": 531, + "ACAAGG": 532, + "ACATAA": 533, + "ACATAT": 534, + "ACATAC": 535, + "ACATAG": 536, + "ACATTA": 537, + "ACATTT": 538, + "ACATTC": 539, + "ACATTG": 540, + "ACATCA": 541, + "ACATCT": 542, + "ACATCC": 543, + "ACATCG": 544, + "ACATGA": 545, + "ACATGT": 546, + "ACATGC": 547, + "ACATGG": 548, + "ACACAA": 549, + "ACACAT": 550, + "ACACAC": 551, + "ACACAG": 552, + "ACACTA": 553, + "ACACTT": 554, + "ACACTC": 555, + "ACACTG": 556, + "ACACCA": 557, + "ACACCT": 558, + "ACACCC": 559, + "ACACCG": 560, + "ACACGA": 561, + "ACACGT": 562, + "ACACGC": 563, + "ACACGG": 564, + "ACAGAA": 565, + "ACAGAT": 566, + "ACAGAC": 567, + "ACAGAG": 568, + "ACAGTA": 569, + "ACAGTT": 570, + "ACAGTC": 571, + "ACAGTG": 572, + "ACAGCA": 573, + "ACAGCT": 574, + "ACAGCC": 575, + "ACAGCG": 576, + "ACAGGA": 577, + "ACAGGT": 578, + "ACAGGC": 579, + "ACAGGG": 580, + "ACTAAA": 581, + "ACTAAT": 582, + "ACTAAC": 583, + "ACTAAG": 584, + "ACTATA": 585, + "ACTATT": 586, + "ACTATC": 587, + "ACTATG": 588, + "ACTACA": 589, + "ACTACT": 590, + "ACTACC": 591, + "ACTACG": 592, + "ACTAGA": 593, + "ACTAGT": 594, + "ACTAGC": 595, + "ACTAGG": 596, + "ACTTAA": 597, + "ACTTAT": 598, + "ACTTAC": 599, + "ACTTAG": 600, + "ACTTTA": 601, + "ACTTTT": 602, + "ACTTTC": 603, + "ACTTTG": 604, + "ACTTCA": 605, + "ACTTCT": 606, + "ACTTCC": 607, + "ACTTCG": 608, + "ACTTGA": 609, + "ACTTGT": 610, + "ACTTGC": 611, + "ACTTGG": 612, + "ACTCAA": 613, + "ACTCAT": 614, + "ACTCAC": 615, + "ACTCAG": 616, + "ACTCTA": 617, + "ACTCTT": 618, + "ACTCTC": 619, + "ACTCTG": 620, + "ACTCCA": 621, + "ACTCCT": 622, + "ACTCCC": 623, + "ACTCCG": 624, + "ACTCGA": 625, + "ACTCGT": 626, + "ACTCGC": 627, + "ACTCGG": 628, + "ACTGAA": 629, + "ACTGAT": 630, + "ACTGAC": 631, + "ACTGAG": 632, + "ACTGTA": 633, + "ACTGTT": 634, + "ACTGTC": 635, + "ACTGTG": 636, + "ACTGCA": 637, + "ACTGCT": 638, + "ACTGCC": 639, + "ACTGCG": 640, + "ACTGGA": 641, + "ACTGGT": 642, + "ACTGGC": 643, + "ACTGGG": 644, + "ACCAAA": 645, + "ACCAAT": 646, + "ACCAAC": 647, + "ACCAAG": 648, + "ACCATA": 649, + "ACCATT": 650, + "ACCATC": 651, + "ACCATG": 652, + "ACCACA": 653, + "ACCACT": 654, + "ACCACC": 655, + "ACCACG": 656, + "ACCAGA": 657, + "ACCAGT": 658, + "ACCAGC": 659, + "ACCAGG": 660, + "ACCTAA": 661, + "ACCTAT": 662, + "ACCTAC": 663, + "ACCTAG": 664, + "ACCTTA": 665, + "ACCTTT": 666, + "ACCTTC": 667, + "ACCTTG": 668, + "ACCTCA": 669, + "ACCTCT": 670, + "ACCTCC": 671, + "ACCTCG": 672, + "ACCTGA": 673, + "ACCTGT": 674, + "ACCTGC": 675, + "ACCTGG": 676, + "ACCCAA": 677, + "ACCCAT": 678, + "ACCCAC": 679, + "ACCCAG": 680, + "ACCCTA": 681, + "ACCCTT": 682, + "ACCCTC": 683, + "ACCCTG": 684, + "ACCCCA": 685, + "ACCCCT": 686, + "ACCCCC": 687, + "ACCCCG": 688, + "ACCCGA": 689, + "ACCCGT": 690, + "ACCCGC": 691, + "ACCCGG": 692, + "ACCGAA": 693, + "ACCGAT": 694, + "ACCGAC": 695, + "ACCGAG": 696, + "ACCGTA": 697, + "ACCGTT": 698, + "ACCGTC": 699, + "ACCGTG": 700, + "ACCGCA": 701, + "ACCGCT": 702, + "ACCGCC": 703, + "ACCGCG": 704, + "ACCGGA": 705, + "ACCGGT": 706, + "ACCGGC": 707, + "ACCGGG": 708, + "ACGAAA": 709, + "ACGAAT": 710, + "ACGAAC": 711, + "ACGAAG": 712, + "ACGATA": 713, + "ACGATT": 714, + "ACGATC": 715, + "ACGATG": 716, + "ACGACA": 717, + "ACGACT": 718, + "ACGACC": 719, + "ACGACG": 720, + "ACGAGA": 721, + "ACGAGT": 722, + "ACGAGC": 723, + "ACGAGG": 724, + "ACGTAA": 725, + "ACGTAT": 726, + "ACGTAC": 727, + "ACGTAG": 728, + "ACGTTA": 729, + "ACGTTT": 730, + "ACGTTC": 731, + "ACGTTG": 732, + "ACGTCA": 733, + "ACGTCT": 734, + "ACGTCC": 735, + "ACGTCG": 736, + "ACGTGA": 737, + "ACGTGT": 738, + "ACGTGC": 739, + "ACGTGG": 740, + "ACGCAA": 741, + "ACGCAT": 742, + "ACGCAC": 743, + "ACGCAG": 744, + "ACGCTA": 745, + "ACGCTT": 746, + "ACGCTC": 747, + "ACGCTG": 748, + "ACGCCA": 749, + "ACGCCT": 750, + "ACGCCC": 751, + "ACGCCG": 752, + "ACGCGA": 753, + "ACGCGT": 754, + "ACGCGC": 755, + "ACGCGG": 756, + "ACGGAA": 757, + "ACGGAT": 758, + "ACGGAC": 759, + "ACGGAG": 760, + "ACGGTA": 761, + "ACGGTT": 762, + "ACGGTC": 763, + "ACGGTG": 764, + "ACGGCA": 765, + "ACGGCT": 766, + "ACGGCC": 767, + "ACGGCG": 768, + "ACGGGA": 769, + "ACGGGT": 770, + "ACGGGC": 771, + "ACGGGG": 772, + "AGAAAA": 773, + "AGAAAT": 774, + "AGAAAC": 775, + "AGAAAG": 776, + "AGAATA": 777, + "AGAATT": 778, + "AGAATC": 779, + "AGAATG": 780, + "AGAACA": 781, + "AGAACT": 782, + "AGAACC": 783, + "AGAACG": 784, + "AGAAGA": 785, + "AGAAGT": 786, + "AGAAGC": 787, + "AGAAGG": 788, + "AGATAA": 789, + "AGATAT": 790, + "AGATAC": 791, + "AGATAG": 792, + "AGATTA": 793, + "AGATTT": 794, + "AGATTC": 795, + "AGATTG": 796, + "AGATCA": 797, + "AGATCT": 798, + "AGATCC": 799, + "AGATCG": 800, + "AGATGA": 801, + "AGATGT": 802, + "AGATGC": 803, + "AGATGG": 804, + "AGACAA": 805, + "AGACAT": 806, + "AGACAC": 807, + "AGACAG": 808, + "AGACTA": 809, + "AGACTT": 810, + "AGACTC": 811, + "AGACTG": 812, + "AGACCA": 813, + "AGACCT": 814, + "AGACCC": 815, + "AGACCG": 816, + "AGACGA": 817, + "AGACGT": 818, + "AGACGC": 819, + "AGACGG": 820, + "AGAGAA": 821, + "AGAGAT": 822, + "AGAGAC": 823, + "AGAGAG": 824, + "AGAGTA": 825, + "AGAGTT": 826, + "AGAGTC": 827, + "AGAGTG": 828, + "AGAGCA": 829, + "AGAGCT": 830, + "AGAGCC": 831, + "AGAGCG": 832, + "AGAGGA": 833, + "AGAGGT": 834, + "AGAGGC": 835, + "AGAGGG": 836, + "AGTAAA": 837, + "AGTAAT": 838, + "AGTAAC": 839, + "AGTAAG": 840, + "AGTATA": 841, + "AGTATT": 842, + "AGTATC": 843, + "AGTATG": 844, + "AGTACA": 845, + "AGTACT": 846, + "AGTACC": 847, + "AGTACG": 848, + "AGTAGA": 849, + "AGTAGT": 850, + "AGTAGC": 851, + "AGTAGG": 852, + "AGTTAA": 853, + "AGTTAT": 854, + "AGTTAC": 855, + "AGTTAG": 856, + "AGTTTA": 857, + "AGTTTT": 858, + "AGTTTC": 859, + "AGTTTG": 860, + "AGTTCA": 861, + "AGTTCT": 862, + "AGTTCC": 863, + "AGTTCG": 864, + "AGTTGA": 865, + "AGTTGT": 866, + "AGTTGC": 867, + "AGTTGG": 868, + "AGTCAA": 869, + "AGTCAT": 870, + "AGTCAC": 871, + "AGTCAG": 872, + "AGTCTA": 873, + "AGTCTT": 874, + "AGTCTC": 875, + "AGTCTG": 876, + "AGTCCA": 877, + "AGTCCT": 878, + "AGTCCC": 879, + "AGTCCG": 880, + "AGTCGA": 881, + "AGTCGT": 882, + "AGTCGC": 883, + "AGTCGG": 884, + "AGTGAA": 885, + "AGTGAT": 886, + "AGTGAC": 887, + "AGTGAG": 888, + "AGTGTA": 889, + "AGTGTT": 890, + "AGTGTC": 891, + "AGTGTG": 892, + "AGTGCA": 893, + "AGTGCT": 894, + "AGTGCC": 895, + "AGTGCG": 896, + "AGTGGA": 897, + "AGTGGT": 898, + "AGTGGC": 899, + "AGTGGG": 900, + "AGCAAA": 901, + "AGCAAT": 902, + "AGCAAC": 903, + "AGCAAG": 904, + "AGCATA": 905, + "AGCATT": 906, + "AGCATC": 907, + "AGCATG": 908, + "AGCACA": 909, + "AGCACT": 910, + "AGCACC": 911, + "AGCACG": 912, + "AGCAGA": 913, + "AGCAGT": 914, + "AGCAGC": 915, + "AGCAGG": 916, + "AGCTAA": 917, + "AGCTAT": 918, + "AGCTAC": 919, + "AGCTAG": 920, + "AGCTTA": 921, + "AGCTTT": 922, + "AGCTTC": 923, + "AGCTTG": 924, + "AGCTCA": 925, + "AGCTCT": 926, + "AGCTCC": 927, + "AGCTCG": 928, + "AGCTGA": 929, + "AGCTGT": 930, + "AGCTGC": 931, + "AGCTGG": 932, + "AGCCAA": 933, + "AGCCAT": 934, + "AGCCAC": 935, + "AGCCAG": 936, + "AGCCTA": 937, + "AGCCTT": 938, + "AGCCTC": 939, + "AGCCTG": 940, + "AGCCCA": 941, + "AGCCCT": 942, + "AGCCCC": 943, + "AGCCCG": 944, + "AGCCGA": 945, + "AGCCGT": 946, + "AGCCGC": 947, + "AGCCGG": 948, + "AGCGAA": 949, + "AGCGAT": 950, + "AGCGAC": 951, + "AGCGAG": 952, + "AGCGTA": 953, + "AGCGTT": 954, + "AGCGTC": 955, + "AGCGTG": 956, + "AGCGCA": 957, + "AGCGCT": 958, + "AGCGCC": 959, + "AGCGCG": 960, + "AGCGGA": 961, + "AGCGGT": 962, + "AGCGGC": 963, + "AGCGGG": 964, + "AGGAAA": 965, + "AGGAAT": 966, + "AGGAAC": 967, + "AGGAAG": 968, + "AGGATA": 969, + "AGGATT": 970, + "AGGATC": 971, + "AGGATG": 972, + "AGGACA": 973, + "AGGACT": 974, + "AGGACC": 975, + "AGGACG": 976, + "AGGAGA": 977, + "AGGAGT": 978, + "AGGAGC": 979, + "AGGAGG": 980, + "AGGTAA": 981, + "AGGTAT": 982, + "AGGTAC": 983, + "AGGTAG": 984, + "AGGTTA": 985, + "AGGTTT": 986, + "AGGTTC": 987, + "AGGTTG": 988, + "AGGTCA": 989, + "AGGTCT": 990, + "AGGTCC": 991, + "AGGTCG": 992, + "AGGTGA": 993, + "AGGTGT": 994, + "AGGTGC": 995, + "AGGTGG": 996, + "AGGCAA": 997, + "AGGCAT": 998, + "AGGCAC": 999, + "AGGCAG": 1000, + "AGGCTA": 1001, + "AGGCTT": 1002, + "AGGCTC": 1003, + "AGGCTG": 1004, + "AGGCCA": 1005, + "AGGCCT": 1006, + "AGGCCC": 1007, + "AGGCCG": 1008, + "AGGCGA": 1009, + "AGGCGT": 1010, + "AGGCGC": 1011, + "AGGCGG": 1012, + "AGGGAA": 1013, + "AGGGAT": 1014, + "AGGGAC": 1015, + "AGGGAG": 1016, + "AGGGTA": 1017, + "AGGGTT": 1018, + "AGGGTC": 1019, + "AGGGTG": 1020, + "AGGGCA": 1021, + "AGGGCT": 1022, + "AGGGCC": 1023, + "AGGGCG": 1024, + "AGGGGA": 1025, + "AGGGGT": 1026, + "AGGGGC": 1027, + "AGGGGG": 1028, + "TAAAAA": 1029, + "TAAAAT": 1030, + "TAAAAC": 1031, + "TAAAAG": 1032, + "TAAATA": 1033, + "TAAATT": 1034, + "TAAATC": 1035, + "TAAATG": 1036, + "TAAACA": 1037, + "TAAACT": 1038, + "TAAACC": 1039, + "TAAACG": 1040, + "TAAAGA": 1041, + "TAAAGT": 1042, + "TAAAGC": 1043, + "TAAAGG": 1044, + "TAATAA": 1045, + "TAATAT": 1046, + "TAATAC": 1047, + "TAATAG": 1048, + "TAATTA": 1049, + "TAATTT": 1050, + "TAATTC": 1051, + "TAATTG": 1052, + "TAATCA": 1053, + "TAATCT": 1054, + "TAATCC": 1055, + "TAATCG": 1056, + "TAATGA": 1057, + "TAATGT": 1058, + "TAATGC": 1059, + "TAATGG": 1060, + "TAACAA": 1061, + "TAACAT": 1062, + "TAACAC": 1063, + "TAACAG": 1064, + "TAACTA": 1065, + "TAACTT": 1066, + "TAACTC": 1067, + "TAACTG": 1068, + "TAACCA": 1069, + "TAACCT": 1070, + "TAACCC": 1071, + "TAACCG": 1072, + "TAACGA": 1073, + "TAACGT": 1074, + "TAACGC": 1075, + "TAACGG": 1076, + "TAAGAA": 1077, + "TAAGAT": 1078, + "TAAGAC": 1079, + "TAAGAG": 1080, + "TAAGTA": 1081, + "TAAGTT": 1082, + "TAAGTC": 1083, + "TAAGTG": 1084, + "TAAGCA": 1085, + "TAAGCT": 1086, + "TAAGCC": 1087, + "TAAGCG": 1088, + "TAAGGA": 1089, + "TAAGGT": 1090, + "TAAGGC": 1091, + "TAAGGG": 1092, + "TATAAA": 1093, + "TATAAT": 1094, + "TATAAC": 1095, + "TATAAG": 1096, + "TATATA": 1097, + "TATATT": 1098, + "TATATC": 1099, + "TATATG": 1100, + "TATACA": 1101, + "TATACT": 1102, + "TATACC": 1103, + "TATACG": 1104, + "TATAGA": 1105, + "TATAGT": 1106, + "TATAGC": 1107, + "TATAGG": 1108, + "TATTAA": 1109, + "TATTAT": 1110, + "TATTAC": 1111, + "TATTAG": 1112, + "TATTTA": 1113, + "TATTTT": 1114, + "TATTTC": 1115, + "TATTTG": 1116, + "TATTCA": 1117, + "TATTCT": 1118, + "TATTCC": 1119, + "TATTCG": 1120, + "TATTGA": 1121, + "TATTGT": 1122, + "TATTGC": 1123, + "TATTGG": 1124, + "TATCAA": 1125, + "TATCAT": 1126, + "TATCAC": 1127, + "TATCAG": 1128, + "TATCTA": 1129, + "TATCTT": 1130, + "TATCTC": 1131, + "TATCTG": 1132, + "TATCCA": 1133, + "TATCCT": 1134, + "TATCCC": 1135, + "TATCCG": 1136, + "TATCGA": 1137, + "TATCGT": 1138, + "TATCGC": 1139, + "TATCGG": 1140, + "TATGAA": 1141, + "TATGAT": 1142, + "TATGAC": 1143, + "TATGAG": 1144, + "TATGTA": 1145, + "TATGTT": 1146, + "TATGTC": 1147, + "TATGTG": 1148, + "TATGCA": 1149, + "TATGCT": 1150, + "TATGCC": 1151, + "TATGCG": 1152, + "TATGGA": 1153, + "TATGGT": 1154, + "TATGGC": 1155, + "TATGGG": 1156, + "TACAAA": 1157, + "TACAAT": 1158, + "TACAAC": 1159, + "TACAAG": 1160, + "TACATA": 1161, + "TACATT": 1162, + "TACATC": 1163, + "TACATG": 1164, + "TACACA": 1165, + "TACACT": 1166, + "TACACC": 1167, + "TACACG": 1168, + "TACAGA": 1169, + "TACAGT": 1170, + "TACAGC": 1171, + "TACAGG": 1172, + "TACTAA": 1173, + "TACTAT": 1174, + "TACTAC": 1175, + "TACTAG": 1176, + "TACTTA": 1177, + "TACTTT": 1178, + "TACTTC": 1179, + "TACTTG": 1180, + "TACTCA": 1181, + "TACTCT": 1182, + "TACTCC": 1183, + "TACTCG": 1184, + "TACTGA": 1185, + "TACTGT": 1186, + "TACTGC": 1187, + "TACTGG": 1188, + "TACCAA": 1189, + "TACCAT": 1190, + "TACCAC": 1191, + "TACCAG": 1192, + "TACCTA": 1193, + "TACCTT": 1194, + "TACCTC": 1195, + "TACCTG": 1196, + "TACCCA": 1197, + "TACCCT": 1198, + "TACCCC": 1199, + "TACCCG": 1200, + "TACCGA": 1201, + "TACCGT": 1202, + "TACCGC": 1203, + "TACCGG": 1204, + "TACGAA": 1205, + "TACGAT": 1206, + "TACGAC": 1207, + "TACGAG": 1208, + "TACGTA": 1209, + "TACGTT": 1210, + "TACGTC": 1211, + "TACGTG": 1212, + "TACGCA": 1213, + "TACGCT": 1214, + "TACGCC": 1215, + "TACGCG": 1216, + "TACGGA": 1217, + "TACGGT": 1218, + "TACGGC": 1219, + "TACGGG": 1220, + "TAGAAA": 1221, + "TAGAAT": 1222, + "TAGAAC": 1223, + "TAGAAG": 1224, + "TAGATA": 1225, + "TAGATT": 1226, + "TAGATC": 1227, + "TAGATG": 1228, + "TAGACA": 1229, + "TAGACT": 1230, + "TAGACC": 1231, + "TAGACG": 1232, + "TAGAGA": 1233, + "TAGAGT": 1234, + "TAGAGC": 1235, + "TAGAGG": 1236, + "TAGTAA": 1237, + "TAGTAT": 1238, + "TAGTAC": 1239, + "TAGTAG": 1240, + "TAGTTA": 1241, + "TAGTTT": 1242, + "TAGTTC": 1243, + "TAGTTG": 1244, + "TAGTCA": 1245, + "TAGTCT": 1246, + "TAGTCC": 1247, + "TAGTCG": 1248, + "TAGTGA": 1249, + "TAGTGT": 1250, + "TAGTGC": 1251, + "TAGTGG": 1252, + "TAGCAA": 1253, + "TAGCAT": 1254, + "TAGCAC": 1255, + "TAGCAG": 1256, + "TAGCTA": 1257, + "TAGCTT": 1258, + "TAGCTC": 1259, + "TAGCTG": 1260, + "TAGCCA": 1261, + "TAGCCT": 1262, + "TAGCCC": 1263, + "TAGCCG": 1264, + "TAGCGA": 1265, + "TAGCGT": 1266, + "TAGCGC": 1267, + "TAGCGG": 1268, + "TAGGAA": 1269, + "TAGGAT": 1270, + "TAGGAC": 1271, + "TAGGAG": 1272, + "TAGGTA": 1273, + "TAGGTT": 1274, + "TAGGTC": 1275, + "TAGGTG": 1276, + "TAGGCA": 1277, + "TAGGCT": 1278, + "TAGGCC": 1279, + "TAGGCG": 1280, + "TAGGGA": 1281, + "TAGGGT": 1282, + "TAGGGC": 1283, + "TAGGGG": 1284, + "TTAAAA": 1285, + "TTAAAT": 1286, + "TTAAAC": 1287, + "TTAAAG": 1288, + "TTAATA": 1289, + "TTAATT": 1290, + "TTAATC": 1291, + "TTAATG": 1292, + "TTAACA": 1293, + "TTAACT": 1294, + "TTAACC": 1295, + "TTAACG": 1296, + "TTAAGA": 1297, + "TTAAGT": 1298, + "TTAAGC": 1299, + "TTAAGG": 1300, + "TTATAA": 1301, + "TTATAT": 1302, + "TTATAC": 1303, + "TTATAG": 1304, + "TTATTA": 1305, + "TTATTT": 1306, + "TTATTC": 1307, + "TTATTG": 1308, + "TTATCA": 1309, + "TTATCT": 1310, + "TTATCC": 1311, + "TTATCG": 1312, + "TTATGA": 1313, + "TTATGT": 1314, + "TTATGC": 1315, + "TTATGG": 1316, + "TTACAA": 1317, + "TTACAT": 1318, + "TTACAC": 1319, + "TTACAG": 1320, + "TTACTA": 1321, + "TTACTT": 1322, + "TTACTC": 1323, + "TTACTG": 1324, + "TTACCA": 1325, + "TTACCT": 1326, + "TTACCC": 1327, + "TTACCG": 1328, + "TTACGA": 1329, + "TTACGT": 1330, + "TTACGC": 1331, + "TTACGG": 1332, + "TTAGAA": 1333, + "TTAGAT": 1334, + "TTAGAC": 1335, + "TTAGAG": 1336, + "TTAGTA": 1337, + "TTAGTT": 1338, + "TTAGTC": 1339, + "TTAGTG": 1340, + "TTAGCA": 1341, + "TTAGCT": 1342, + "TTAGCC": 1343, + "TTAGCG": 1344, + "TTAGGA": 1345, + "TTAGGT": 1346, + "TTAGGC": 1347, + "TTAGGG": 1348, + "TTTAAA": 1349, + "TTTAAT": 1350, + "TTTAAC": 1351, + "TTTAAG": 1352, + "TTTATA": 1353, + "TTTATT": 1354, + "TTTATC": 1355, + "TTTATG": 1356, + "TTTACA": 1357, + "TTTACT": 1358, + "TTTACC": 1359, + "TTTACG": 1360, + "TTTAGA": 1361, + "TTTAGT": 1362, + "TTTAGC": 1363, + "TTTAGG": 1364, + "TTTTAA": 1365, + "TTTTAT": 1366, + "TTTTAC": 1367, + "TTTTAG": 1368, + "TTTTTA": 1369, + "TTTTTT": 1370, + "TTTTTC": 1371, + "TTTTTG": 1372, + "TTTTCA": 1373, + "TTTTCT": 1374, + "TTTTCC": 1375, + "TTTTCG": 1376, + "TTTTGA": 1377, + "TTTTGT": 1378, + "TTTTGC": 1379, + "TTTTGG": 1380, + "TTTCAA": 1381, + "TTTCAT": 1382, + "TTTCAC": 1383, + "TTTCAG": 1384, + "TTTCTA": 1385, + "TTTCTT": 1386, + "TTTCTC": 1387, + "TTTCTG": 1388, + "TTTCCA": 1389, + "TTTCCT": 1390, + "TTTCCC": 1391, + "TTTCCG": 1392, + "TTTCGA": 1393, + "TTTCGT": 1394, + "TTTCGC": 1395, + "TTTCGG": 1396, + "TTTGAA": 1397, + "TTTGAT": 1398, + "TTTGAC": 1399, + "TTTGAG": 1400, + "TTTGTA": 1401, + "TTTGTT": 1402, + "TTTGTC": 1403, + "TTTGTG": 1404, + "TTTGCA": 1405, + "TTTGCT": 1406, + "TTTGCC": 1407, + "TTTGCG": 1408, + "TTTGGA": 1409, + "TTTGGT": 1410, + "TTTGGC": 1411, + "TTTGGG": 1412, + "TTCAAA": 1413, + "TTCAAT": 1414, + "TTCAAC": 1415, + "TTCAAG": 1416, + "TTCATA": 1417, + "TTCATT": 1418, + "TTCATC": 1419, + "TTCATG": 1420, + "TTCACA": 1421, + "TTCACT": 1422, + "TTCACC": 1423, + "TTCACG": 1424, + "TTCAGA": 1425, + "TTCAGT": 1426, + "TTCAGC": 1427, + "TTCAGG": 1428, + "TTCTAA": 1429, + "TTCTAT": 1430, + "TTCTAC": 1431, + "TTCTAG": 1432, + "TTCTTA": 1433, + "TTCTTT": 1434, + "TTCTTC": 1435, + "TTCTTG": 1436, + "TTCTCA": 1437, + "TTCTCT": 1438, + "TTCTCC": 1439, + "TTCTCG": 1440, + "TTCTGA": 1441, + "TTCTGT": 1442, + "TTCTGC": 1443, + "TTCTGG": 1444, + "TTCCAA": 1445, + "TTCCAT": 1446, + "TTCCAC": 1447, + "TTCCAG": 1448, + "TTCCTA": 1449, + "TTCCTT": 1450, + "TTCCTC": 1451, + "TTCCTG": 1452, + "TTCCCA": 1453, + "TTCCCT": 1454, + "TTCCCC": 1455, + "TTCCCG": 1456, + "TTCCGA": 1457, + "TTCCGT": 1458, + "TTCCGC": 1459, + "TTCCGG": 1460, + "TTCGAA": 1461, + "TTCGAT": 1462, + "TTCGAC": 1463, + "TTCGAG": 1464, + "TTCGTA": 1465, + "TTCGTT": 1466, + "TTCGTC": 1467, + "TTCGTG": 1468, + "TTCGCA": 1469, + "TTCGCT": 1470, + "TTCGCC": 1471, + "TTCGCG": 1472, + "TTCGGA": 1473, + "TTCGGT": 1474, + "TTCGGC": 1475, + "TTCGGG": 1476, + "TTGAAA": 1477, + "TTGAAT": 1478, + "TTGAAC": 1479, + "TTGAAG": 1480, + "TTGATA": 1481, + "TTGATT": 1482, + "TTGATC": 1483, + "TTGATG": 1484, + "TTGACA": 1485, + "TTGACT": 1486, + "TTGACC": 1487, + "TTGACG": 1488, + "TTGAGA": 1489, + "TTGAGT": 1490, + "TTGAGC": 1491, + "TTGAGG": 1492, + "TTGTAA": 1493, + "TTGTAT": 1494, + "TTGTAC": 1495, + "TTGTAG": 1496, + "TTGTTA": 1497, + "TTGTTT": 1498, + "TTGTTC": 1499, + "TTGTTG": 1500, + "TTGTCA": 1501, + "TTGTCT": 1502, + "TTGTCC": 1503, + "TTGTCG": 1504, + "TTGTGA": 1505, + "TTGTGT": 1506, + "TTGTGC": 1507, + "TTGTGG": 1508, + "TTGCAA": 1509, + "TTGCAT": 1510, + "TTGCAC": 1511, + "TTGCAG": 1512, + "TTGCTA": 1513, + "TTGCTT": 1514, + "TTGCTC": 1515, + "TTGCTG": 1516, + "TTGCCA": 1517, + "TTGCCT": 1518, + "TTGCCC": 1519, + "TTGCCG": 1520, + "TTGCGA": 1521, + "TTGCGT": 1522, + "TTGCGC": 1523, + "TTGCGG": 1524, + "TTGGAA": 1525, + "TTGGAT": 1526, + "TTGGAC": 1527, + "TTGGAG": 1528, + "TTGGTA": 1529, + "TTGGTT": 1530, + "TTGGTC": 1531, + "TTGGTG": 1532, + "TTGGCA": 1533, + "TTGGCT": 1534, + "TTGGCC": 1535, + "TTGGCG": 1536, + "TTGGGA": 1537, + "TTGGGT": 1538, + "TTGGGC": 1539, + "TTGGGG": 1540, + "TCAAAA": 1541, + "TCAAAT": 1542, + "TCAAAC": 1543, + "TCAAAG": 1544, + "TCAATA": 1545, + "TCAATT": 1546, + "TCAATC": 1547, + "TCAATG": 1548, + "TCAACA": 1549, + "TCAACT": 1550, + "TCAACC": 1551, + "TCAACG": 1552, + "TCAAGA": 1553, + "TCAAGT": 1554, + "TCAAGC": 1555, + "TCAAGG": 1556, + "TCATAA": 1557, + "TCATAT": 1558, + "TCATAC": 1559, + "TCATAG": 1560, + "TCATTA": 1561, + "TCATTT": 1562, + "TCATTC": 1563, + "TCATTG": 1564, + "TCATCA": 1565, + "TCATCT": 1566, + "TCATCC": 1567, + "TCATCG": 1568, + "TCATGA": 1569, + "TCATGT": 1570, + "TCATGC": 1571, + "TCATGG": 1572, + "TCACAA": 1573, + "TCACAT": 1574, + "TCACAC": 1575, + "TCACAG": 1576, + "TCACTA": 1577, + "TCACTT": 1578, + "TCACTC": 1579, + "TCACTG": 1580, + "TCACCA": 1581, + "TCACCT": 1582, + "TCACCC": 1583, + "TCACCG": 1584, + "TCACGA": 1585, + "TCACGT": 1586, + "TCACGC": 1587, + "TCACGG": 1588, + "TCAGAA": 1589, + "TCAGAT": 1590, + "TCAGAC": 1591, + "TCAGAG": 1592, + "TCAGTA": 1593, + "TCAGTT": 1594, + "TCAGTC": 1595, + "TCAGTG": 1596, + "TCAGCA": 1597, + "TCAGCT": 1598, + "TCAGCC": 1599, + "TCAGCG": 1600, + "TCAGGA": 1601, + "TCAGGT": 1602, + "TCAGGC": 1603, + "TCAGGG": 1604, + "TCTAAA": 1605, + "TCTAAT": 1606, + "TCTAAC": 1607, + "TCTAAG": 1608, + "TCTATA": 1609, + "TCTATT": 1610, + "TCTATC": 1611, + "TCTATG": 1612, + "TCTACA": 1613, + "TCTACT": 1614, + "TCTACC": 1615, + "TCTACG": 1616, + "TCTAGA": 1617, + "TCTAGT": 1618, + "TCTAGC": 1619, + "TCTAGG": 1620, + "TCTTAA": 1621, + "TCTTAT": 1622, + "TCTTAC": 1623, + "TCTTAG": 1624, + "TCTTTA": 1625, + "TCTTTT": 1626, + "TCTTTC": 1627, + "TCTTTG": 1628, + "TCTTCA": 1629, + "TCTTCT": 1630, + "TCTTCC": 1631, + "TCTTCG": 1632, + "TCTTGA": 1633, + "TCTTGT": 1634, + "TCTTGC": 1635, + "TCTTGG": 1636, + "TCTCAA": 1637, + "TCTCAT": 1638, + "TCTCAC": 1639, + "TCTCAG": 1640, + "TCTCTA": 1641, + "TCTCTT": 1642, + "TCTCTC": 1643, + "TCTCTG": 1644, + "TCTCCA": 1645, + "TCTCCT": 1646, + "TCTCCC": 1647, + "TCTCCG": 1648, + "TCTCGA": 1649, + "TCTCGT": 1650, + "TCTCGC": 1651, + "TCTCGG": 1652, + "TCTGAA": 1653, + "TCTGAT": 1654, + "TCTGAC": 1655, + "TCTGAG": 1656, + "TCTGTA": 1657, + "TCTGTT": 1658, + "TCTGTC": 1659, + "TCTGTG": 1660, + "TCTGCA": 1661, + "TCTGCT": 1662, + "TCTGCC": 1663, + "TCTGCG": 1664, + "TCTGGA": 1665, + "TCTGGT": 1666, + "TCTGGC": 1667, + "TCTGGG": 1668, + "TCCAAA": 1669, + "TCCAAT": 1670, + "TCCAAC": 1671, + "TCCAAG": 1672, + "TCCATA": 1673, + "TCCATT": 1674, + "TCCATC": 1675, + "TCCATG": 1676, + "TCCACA": 1677, + "TCCACT": 1678, + "TCCACC": 1679, + "TCCACG": 1680, + "TCCAGA": 1681, + "TCCAGT": 1682, + "TCCAGC": 1683, + "TCCAGG": 1684, + "TCCTAA": 1685, + "TCCTAT": 1686, + "TCCTAC": 1687, + "TCCTAG": 1688, + "TCCTTA": 1689, + "TCCTTT": 1690, + "TCCTTC": 1691, + "TCCTTG": 1692, + "TCCTCA": 1693, + "TCCTCT": 1694, + "TCCTCC": 1695, + "TCCTCG": 1696, + "TCCTGA": 1697, + "TCCTGT": 1698, + "TCCTGC": 1699, + "TCCTGG": 1700, + "TCCCAA": 1701, + "TCCCAT": 1702, + "TCCCAC": 1703, + "TCCCAG": 1704, + "TCCCTA": 1705, + "TCCCTT": 1706, + "TCCCTC": 1707, + "TCCCTG": 1708, + "TCCCCA": 1709, + "TCCCCT": 1710, + "TCCCCC": 1711, + "TCCCCG": 1712, + "TCCCGA": 1713, + "TCCCGT": 1714, + "TCCCGC": 1715, + "TCCCGG": 1716, + "TCCGAA": 1717, + "TCCGAT": 1718, + "TCCGAC": 1719, + "TCCGAG": 1720, + "TCCGTA": 1721, + "TCCGTT": 1722, + "TCCGTC": 1723, + "TCCGTG": 1724, + "TCCGCA": 1725, + "TCCGCT": 1726, + "TCCGCC": 1727, + "TCCGCG": 1728, + "TCCGGA": 1729, + "TCCGGT": 1730, + "TCCGGC": 1731, + "TCCGGG": 1732, + "TCGAAA": 1733, + "TCGAAT": 1734, + "TCGAAC": 1735, + "TCGAAG": 1736, + "TCGATA": 1737, + "TCGATT": 1738, + "TCGATC": 1739, + "TCGATG": 1740, + "TCGACA": 1741, + "TCGACT": 1742, + "TCGACC": 1743, + "TCGACG": 1744, + "TCGAGA": 1745, + "TCGAGT": 1746, + "TCGAGC": 1747, + "TCGAGG": 1748, + "TCGTAA": 1749, + "TCGTAT": 1750, + "TCGTAC": 1751, + "TCGTAG": 1752, + "TCGTTA": 1753, + "TCGTTT": 1754, + "TCGTTC": 1755, + "TCGTTG": 1756, + "TCGTCA": 1757, + "TCGTCT": 1758, + "TCGTCC": 1759, + "TCGTCG": 1760, + "TCGTGA": 1761, + "TCGTGT": 1762, + "TCGTGC": 1763, + "TCGTGG": 1764, + "TCGCAA": 1765, + "TCGCAT": 1766, + "TCGCAC": 1767, + "TCGCAG": 1768, + "TCGCTA": 1769, + "TCGCTT": 1770, + "TCGCTC": 1771, + "TCGCTG": 1772, + "TCGCCA": 1773, + "TCGCCT": 1774, + "TCGCCC": 1775, + "TCGCCG": 1776, + "TCGCGA": 1777, + "TCGCGT": 1778, + "TCGCGC": 1779, + "TCGCGG": 1780, + "TCGGAA": 1781, + "TCGGAT": 1782, + "TCGGAC": 1783, + "TCGGAG": 1784, + "TCGGTA": 1785, + "TCGGTT": 1786, + "TCGGTC": 1787, + "TCGGTG": 1788, + "TCGGCA": 1789, + "TCGGCT": 1790, + "TCGGCC": 1791, + "TCGGCG": 1792, + "TCGGGA": 1793, + "TCGGGT": 1794, + "TCGGGC": 1795, + "TCGGGG": 1796, + "TGAAAA": 1797, + "TGAAAT": 1798, + "TGAAAC": 1799, + "TGAAAG": 1800, + "TGAATA": 1801, + "TGAATT": 1802, + "TGAATC": 1803, + "TGAATG": 1804, + "TGAACA": 1805, + "TGAACT": 1806, + "TGAACC": 1807, + "TGAACG": 1808, + "TGAAGA": 1809, + "TGAAGT": 1810, + "TGAAGC": 1811, + "TGAAGG": 1812, + "TGATAA": 1813, + "TGATAT": 1814, + "TGATAC": 1815, + "TGATAG": 1816, + "TGATTA": 1817, + "TGATTT": 1818, + "TGATTC": 1819, + "TGATTG": 1820, + "TGATCA": 1821, + "TGATCT": 1822, + "TGATCC": 1823, + "TGATCG": 1824, + "TGATGA": 1825, + "TGATGT": 1826, + "TGATGC": 1827, + "TGATGG": 1828, + "TGACAA": 1829, + "TGACAT": 1830, + "TGACAC": 1831, + "TGACAG": 1832, + "TGACTA": 1833, + "TGACTT": 1834, + "TGACTC": 1835, + "TGACTG": 1836, + "TGACCA": 1837, + "TGACCT": 1838, + "TGACCC": 1839, + "TGACCG": 1840, + "TGACGA": 1841, + "TGACGT": 1842, + "TGACGC": 1843, + "TGACGG": 1844, + "TGAGAA": 1845, + "TGAGAT": 1846, + "TGAGAC": 1847, + "TGAGAG": 1848, + "TGAGTA": 1849, + "TGAGTT": 1850, + "TGAGTC": 1851, + "TGAGTG": 1852, + "TGAGCA": 1853, + "TGAGCT": 1854, + "TGAGCC": 1855, + "TGAGCG": 1856, + "TGAGGA": 1857, + "TGAGGT": 1858, + "TGAGGC": 1859, + "TGAGGG": 1860, + "TGTAAA": 1861, + "TGTAAT": 1862, + "TGTAAC": 1863, + "TGTAAG": 1864, + "TGTATA": 1865, + "TGTATT": 1866, + "TGTATC": 1867, + "TGTATG": 1868, + "TGTACA": 1869, + "TGTACT": 1870, + "TGTACC": 1871, + "TGTACG": 1872, + "TGTAGA": 1873, + "TGTAGT": 1874, + "TGTAGC": 1875, + "TGTAGG": 1876, + "TGTTAA": 1877, + "TGTTAT": 1878, + "TGTTAC": 1879, + "TGTTAG": 1880, + "TGTTTA": 1881, + "TGTTTT": 1882, + "TGTTTC": 1883, + "TGTTTG": 1884, + "TGTTCA": 1885, + "TGTTCT": 1886, + "TGTTCC": 1887, + "TGTTCG": 1888, + "TGTTGA": 1889, + "TGTTGT": 1890, + "TGTTGC": 1891, + "TGTTGG": 1892, + "TGTCAA": 1893, + "TGTCAT": 1894, + "TGTCAC": 1895, + "TGTCAG": 1896, + "TGTCTA": 1897, + "TGTCTT": 1898, + "TGTCTC": 1899, + "TGTCTG": 1900, + "TGTCCA": 1901, + "TGTCCT": 1902, + "TGTCCC": 1903, + "TGTCCG": 1904, + "TGTCGA": 1905, + "TGTCGT": 1906, + "TGTCGC": 1907, + "TGTCGG": 1908, + "TGTGAA": 1909, + "TGTGAT": 1910, + "TGTGAC": 1911, + "TGTGAG": 1912, + "TGTGTA": 1913, + "TGTGTT": 1914, + "TGTGTC": 1915, + "TGTGTG": 1916, + "TGTGCA": 1917, + "TGTGCT": 1918, + "TGTGCC": 1919, + "TGTGCG": 1920, + "TGTGGA": 1921, + "TGTGGT": 1922, + "TGTGGC": 1923, + "TGTGGG": 1924, + "TGCAAA": 1925, + "TGCAAT": 1926, + "TGCAAC": 1927, + "TGCAAG": 1928, + "TGCATA": 1929, + "TGCATT": 1930, + "TGCATC": 1931, + "TGCATG": 1932, + "TGCACA": 1933, + "TGCACT": 1934, + "TGCACC": 1935, + "TGCACG": 1936, + "TGCAGA": 1937, + "TGCAGT": 1938, + "TGCAGC": 1939, + "TGCAGG": 1940, + "TGCTAA": 1941, + "TGCTAT": 1942, + "TGCTAC": 1943, + "TGCTAG": 1944, + "TGCTTA": 1945, + "TGCTTT": 1946, + "TGCTTC": 1947, + "TGCTTG": 1948, + "TGCTCA": 1949, + "TGCTCT": 1950, + "TGCTCC": 1951, + "TGCTCG": 1952, + "TGCTGA": 1953, + "TGCTGT": 1954, + "TGCTGC": 1955, + "TGCTGG": 1956, + "TGCCAA": 1957, + "TGCCAT": 1958, + "TGCCAC": 1959, + "TGCCAG": 1960, + "TGCCTA": 1961, + "TGCCTT": 1962, + "TGCCTC": 1963, + "TGCCTG": 1964, + "TGCCCA": 1965, + "TGCCCT": 1966, + "TGCCCC": 1967, + "TGCCCG": 1968, + "TGCCGA": 1969, + "TGCCGT": 1970, + "TGCCGC": 1971, + "TGCCGG": 1972, + "TGCGAA": 1973, + "TGCGAT": 1974, + "TGCGAC": 1975, + "TGCGAG": 1976, + "TGCGTA": 1977, + "TGCGTT": 1978, + "TGCGTC": 1979, + "TGCGTG": 1980, + "TGCGCA": 1981, + "TGCGCT": 1982, + "TGCGCC": 1983, + "TGCGCG": 1984, + "TGCGGA": 1985, + "TGCGGT": 1986, + "TGCGGC": 1987, + "TGCGGG": 1988, + "TGGAAA": 1989, + "TGGAAT": 1990, + "TGGAAC": 1991, + "TGGAAG": 1992, + "TGGATA": 1993, + "TGGATT": 1994, + "TGGATC": 1995, + "TGGATG": 1996, + "TGGACA": 1997, + "TGGACT": 1998, + "TGGACC": 1999, + "TGGACG": 2000, + "TGGAGA": 2001, + "TGGAGT": 2002, + "TGGAGC": 2003, + "TGGAGG": 2004, + "TGGTAA": 2005, + "TGGTAT": 2006, + "TGGTAC": 2007, + "TGGTAG": 2008, + "TGGTTA": 2009, + "TGGTTT": 2010, + "TGGTTC": 2011, + "TGGTTG": 2012, + "TGGTCA": 2013, + "TGGTCT": 2014, + "TGGTCC": 2015, + "TGGTCG": 2016, + "TGGTGA": 2017, + "TGGTGT": 2018, + "TGGTGC": 2019, + "TGGTGG": 2020, + "TGGCAA": 2021, + "TGGCAT": 2022, + "TGGCAC": 2023, + "TGGCAG": 2024, + "TGGCTA": 2025, + "TGGCTT": 2026, + "TGGCTC": 2027, + "TGGCTG": 2028, + "TGGCCA": 2029, + "TGGCCT": 2030, + "TGGCCC": 2031, + "TGGCCG": 2032, + "TGGCGA": 2033, + "TGGCGT": 2034, + "TGGCGC": 2035, + "TGGCGG": 2036, + "TGGGAA": 2037, + "TGGGAT": 2038, + "TGGGAC": 2039, + "TGGGAG": 2040, + "TGGGTA": 2041, + "TGGGTT": 2042, + "TGGGTC": 2043, + "TGGGTG": 2044, + "TGGGCA": 2045, + "TGGGCT": 2046, + "TGGGCC": 2047, + "TGGGCG": 2048, + "TGGGGA": 2049, + "TGGGGT": 2050, + "TGGGGC": 2051, + "TGGGGG": 2052, + "CAAAAA": 2053, + "CAAAAT": 2054, + "CAAAAC": 2055, + "CAAAAG": 2056, + "CAAATA": 2057, + "CAAATT": 2058, + "CAAATC": 2059, + "CAAATG": 2060, + "CAAACA": 2061, + "CAAACT": 2062, + "CAAACC": 2063, + "CAAACG": 2064, + "CAAAGA": 2065, + "CAAAGT": 2066, + "CAAAGC": 2067, + "CAAAGG": 2068, + "CAATAA": 2069, + "CAATAT": 2070, + "CAATAC": 2071, + "CAATAG": 2072, + "CAATTA": 2073, + "CAATTT": 2074, + "CAATTC": 2075, + "CAATTG": 2076, + "CAATCA": 2077, + "CAATCT": 2078, + "CAATCC": 2079, + "CAATCG": 2080, + "CAATGA": 2081, + "CAATGT": 2082, + "CAATGC": 2083, + "CAATGG": 2084, + "CAACAA": 2085, + "CAACAT": 2086, + "CAACAC": 2087, + "CAACAG": 2088, + "CAACTA": 2089, + "CAACTT": 2090, + "CAACTC": 2091, + "CAACTG": 2092, + "CAACCA": 2093, + "CAACCT": 2094, + "CAACCC": 2095, + "CAACCG": 2096, + "CAACGA": 2097, + "CAACGT": 2098, + "CAACGC": 2099, + "CAACGG": 2100, + "CAAGAA": 2101, + "CAAGAT": 2102, + "CAAGAC": 2103, + "CAAGAG": 2104, + "CAAGTA": 2105, + "CAAGTT": 2106, + "CAAGTC": 2107, + "CAAGTG": 2108, + "CAAGCA": 2109, + "CAAGCT": 2110, + "CAAGCC": 2111, + "CAAGCG": 2112, + "CAAGGA": 2113, + "CAAGGT": 2114, + "CAAGGC": 2115, + "CAAGGG": 2116, + "CATAAA": 2117, + "CATAAT": 2118, + "CATAAC": 2119, + "CATAAG": 2120, + "CATATA": 2121, + "CATATT": 2122, + "CATATC": 2123, + "CATATG": 2124, + "CATACA": 2125, + "CATACT": 2126, + "CATACC": 2127, + "CATACG": 2128, + "CATAGA": 2129, + "CATAGT": 2130, + "CATAGC": 2131, + "CATAGG": 2132, + "CATTAA": 2133, + "CATTAT": 2134, + "CATTAC": 2135, + "CATTAG": 2136, + "CATTTA": 2137, + "CATTTT": 2138, + "CATTTC": 2139, + "CATTTG": 2140, + "CATTCA": 2141, + "CATTCT": 2142, + "CATTCC": 2143, + "CATTCG": 2144, + "CATTGA": 2145, + "CATTGT": 2146, + "CATTGC": 2147, + "CATTGG": 2148, + "CATCAA": 2149, + "CATCAT": 2150, + "CATCAC": 2151, + "CATCAG": 2152, + "CATCTA": 2153, + "CATCTT": 2154, + "CATCTC": 2155, + "CATCTG": 2156, + "CATCCA": 2157, + "CATCCT": 2158, + "CATCCC": 2159, + "CATCCG": 2160, + "CATCGA": 2161, + "CATCGT": 2162, + "CATCGC": 2163, + "CATCGG": 2164, + "CATGAA": 2165, + "CATGAT": 2166, + "CATGAC": 2167, + "CATGAG": 2168, + "CATGTA": 2169, + "CATGTT": 2170, + "CATGTC": 2171, + "CATGTG": 2172, + "CATGCA": 2173, + "CATGCT": 2174, + "CATGCC": 2175, + "CATGCG": 2176, + "CATGGA": 2177, + "CATGGT": 2178, + "CATGGC": 2179, + "CATGGG": 2180, + "CACAAA": 2181, + "CACAAT": 2182, + "CACAAC": 2183, + "CACAAG": 2184, + "CACATA": 2185, + "CACATT": 2186, + "CACATC": 2187, + "CACATG": 2188, + "CACACA": 2189, + "CACACT": 2190, + "CACACC": 2191, + "CACACG": 2192, + "CACAGA": 2193, + "CACAGT": 2194, + "CACAGC": 2195, + "CACAGG": 2196, + "CACTAA": 2197, + "CACTAT": 2198, + "CACTAC": 2199, + "CACTAG": 2200, + "CACTTA": 2201, + "CACTTT": 2202, + "CACTTC": 2203, + "CACTTG": 2204, + "CACTCA": 2205, + "CACTCT": 2206, + "CACTCC": 2207, + "CACTCG": 2208, + "CACTGA": 2209, + "CACTGT": 2210, + "CACTGC": 2211, + "CACTGG": 2212, + "CACCAA": 2213, + "CACCAT": 2214, + "CACCAC": 2215, + "CACCAG": 2216, + "CACCTA": 2217, + "CACCTT": 2218, + "CACCTC": 2219, + "CACCTG": 2220, + "CACCCA": 2221, + "CACCCT": 2222, + "CACCCC": 2223, + "CACCCG": 2224, + "CACCGA": 2225, + "CACCGT": 2226, + "CACCGC": 2227, + "CACCGG": 2228, + "CACGAA": 2229, + "CACGAT": 2230, + "CACGAC": 2231, + "CACGAG": 2232, + "CACGTA": 2233, + "CACGTT": 2234, + "CACGTC": 2235, + "CACGTG": 2236, + "CACGCA": 2237, + "CACGCT": 2238, + "CACGCC": 2239, + "CACGCG": 2240, + "CACGGA": 2241, + "CACGGT": 2242, + "CACGGC": 2243, + "CACGGG": 2244, + "CAGAAA": 2245, + "CAGAAT": 2246, + "CAGAAC": 2247, + "CAGAAG": 2248, + "CAGATA": 2249, + "CAGATT": 2250, + "CAGATC": 2251, + "CAGATG": 2252, + "CAGACA": 2253, + "CAGACT": 2254, + "CAGACC": 2255, + "CAGACG": 2256, + "CAGAGA": 2257, + "CAGAGT": 2258, + "CAGAGC": 2259, + "CAGAGG": 2260, + "CAGTAA": 2261, + "CAGTAT": 2262, + "CAGTAC": 2263, + "CAGTAG": 2264, + "CAGTTA": 2265, + "CAGTTT": 2266, + "CAGTTC": 2267, + "CAGTTG": 2268, + "CAGTCA": 2269, + "CAGTCT": 2270, + "CAGTCC": 2271, + "CAGTCG": 2272, + "CAGTGA": 2273, + "CAGTGT": 2274, + "CAGTGC": 2275, + "CAGTGG": 2276, + "CAGCAA": 2277, + "CAGCAT": 2278, + "CAGCAC": 2279, + "CAGCAG": 2280, + "CAGCTA": 2281, + "CAGCTT": 2282, + "CAGCTC": 2283, + "CAGCTG": 2284, + "CAGCCA": 2285, + "CAGCCT": 2286, + "CAGCCC": 2287, + "CAGCCG": 2288, + "CAGCGA": 2289, + "CAGCGT": 2290, + "CAGCGC": 2291, + "CAGCGG": 2292, + "CAGGAA": 2293, + "CAGGAT": 2294, + "CAGGAC": 2295, + "CAGGAG": 2296, + "CAGGTA": 2297, + "CAGGTT": 2298, + "CAGGTC": 2299, + "CAGGTG": 2300, + "CAGGCA": 2301, + "CAGGCT": 2302, + "CAGGCC": 2303, + "CAGGCG": 2304, + "CAGGGA": 2305, + "CAGGGT": 2306, + "CAGGGC": 2307, + "CAGGGG": 2308, + "CTAAAA": 2309, + "CTAAAT": 2310, + "CTAAAC": 2311, + "CTAAAG": 2312, + "CTAATA": 2313, + "CTAATT": 2314, + "CTAATC": 2315, + "CTAATG": 2316, + "CTAACA": 2317, + "CTAACT": 2318, + "CTAACC": 2319, + "CTAACG": 2320, + "CTAAGA": 2321, + "CTAAGT": 2322, + "CTAAGC": 2323, + "CTAAGG": 2324, + "CTATAA": 2325, + "CTATAT": 2326, + "CTATAC": 2327, + "CTATAG": 2328, + "CTATTA": 2329, + "CTATTT": 2330, + "CTATTC": 2331, + "CTATTG": 2332, + "CTATCA": 2333, + "CTATCT": 2334, + "CTATCC": 2335, + "CTATCG": 2336, + "CTATGA": 2337, + "CTATGT": 2338, + "CTATGC": 2339, + "CTATGG": 2340, + "CTACAA": 2341, + "CTACAT": 2342, + "CTACAC": 2343, + "CTACAG": 2344, + "CTACTA": 2345, + "CTACTT": 2346, + "CTACTC": 2347, + "CTACTG": 2348, + "CTACCA": 2349, + "CTACCT": 2350, + "CTACCC": 2351, + "CTACCG": 2352, + "CTACGA": 2353, + "CTACGT": 2354, + "CTACGC": 2355, + "CTACGG": 2356, + "CTAGAA": 2357, + "CTAGAT": 2358, + "CTAGAC": 2359, + "CTAGAG": 2360, + "CTAGTA": 2361, + "CTAGTT": 2362, + "CTAGTC": 2363, + "CTAGTG": 2364, + "CTAGCA": 2365, + "CTAGCT": 2366, + "CTAGCC": 2367, + "CTAGCG": 2368, + "CTAGGA": 2369, + "CTAGGT": 2370, + "CTAGGC": 2371, + "CTAGGG": 2372, + "CTTAAA": 2373, + "CTTAAT": 2374, + "CTTAAC": 2375, + "CTTAAG": 2376, + "CTTATA": 2377, + "CTTATT": 2378, + "CTTATC": 2379, + "CTTATG": 2380, + "CTTACA": 2381, + "CTTACT": 2382, + "CTTACC": 2383, + "CTTACG": 2384, + "CTTAGA": 2385, + "CTTAGT": 2386, + "CTTAGC": 2387, + "CTTAGG": 2388, + "CTTTAA": 2389, + "CTTTAT": 2390, + "CTTTAC": 2391, + "CTTTAG": 2392, + "CTTTTA": 2393, + "CTTTTT": 2394, + "CTTTTC": 2395, + "CTTTTG": 2396, + "CTTTCA": 2397, + "CTTTCT": 2398, + "CTTTCC": 2399, + "CTTTCG": 2400, + "CTTTGA": 2401, + "CTTTGT": 2402, + "CTTTGC": 2403, + "CTTTGG": 2404, + "CTTCAA": 2405, + "CTTCAT": 2406, + "CTTCAC": 2407, + "CTTCAG": 2408, + "CTTCTA": 2409, + "CTTCTT": 2410, + "CTTCTC": 2411, + "CTTCTG": 2412, + "CTTCCA": 2413, + "CTTCCT": 2414, + "CTTCCC": 2415, + "CTTCCG": 2416, + "CTTCGA": 2417, + "CTTCGT": 2418, + "CTTCGC": 2419, + "CTTCGG": 2420, + "CTTGAA": 2421, + "CTTGAT": 2422, + "CTTGAC": 2423, + "CTTGAG": 2424, + "CTTGTA": 2425, + "CTTGTT": 2426, + "CTTGTC": 2427, + "CTTGTG": 2428, + "CTTGCA": 2429, + "CTTGCT": 2430, + "CTTGCC": 2431, + "CTTGCG": 2432, + "CTTGGA": 2433, + "CTTGGT": 2434, + "CTTGGC": 2435, + "CTTGGG": 2436, + "CTCAAA": 2437, + "CTCAAT": 2438, + "CTCAAC": 2439, + "CTCAAG": 2440, + "CTCATA": 2441, + "CTCATT": 2442, + "CTCATC": 2443, + "CTCATG": 2444, + "CTCACA": 2445, + "CTCACT": 2446, + "CTCACC": 2447, + "CTCACG": 2448, + "CTCAGA": 2449, + "CTCAGT": 2450, + "CTCAGC": 2451, + "CTCAGG": 2452, + "CTCTAA": 2453, + "CTCTAT": 2454, + "CTCTAC": 2455, + "CTCTAG": 2456, + "CTCTTA": 2457, + "CTCTTT": 2458, + "CTCTTC": 2459, + "CTCTTG": 2460, + "CTCTCA": 2461, + "CTCTCT": 2462, + "CTCTCC": 2463, + "CTCTCG": 2464, + "CTCTGA": 2465, + "CTCTGT": 2466, + "CTCTGC": 2467, + "CTCTGG": 2468, + "CTCCAA": 2469, + "CTCCAT": 2470, + "CTCCAC": 2471, + "CTCCAG": 2472, + "CTCCTA": 2473, + "CTCCTT": 2474, + "CTCCTC": 2475, + "CTCCTG": 2476, + "CTCCCA": 2477, + "CTCCCT": 2478, + "CTCCCC": 2479, + "CTCCCG": 2480, + "CTCCGA": 2481, + "CTCCGT": 2482, + "CTCCGC": 2483, + "CTCCGG": 2484, + "CTCGAA": 2485, + "CTCGAT": 2486, + "CTCGAC": 2487, + "CTCGAG": 2488, + "CTCGTA": 2489, + "CTCGTT": 2490, + "CTCGTC": 2491, + "CTCGTG": 2492, + "CTCGCA": 2493, + "CTCGCT": 2494, + "CTCGCC": 2495, + "CTCGCG": 2496, + "CTCGGA": 2497, + "CTCGGT": 2498, + "CTCGGC": 2499, + "CTCGGG": 2500, + "CTGAAA": 2501, + "CTGAAT": 2502, + "CTGAAC": 2503, + "CTGAAG": 2504, + "CTGATA": 2505, + "CTGATT": 2506, + "CTGATC": 2507, + "CTGATG": 2508, + "CTGACA": 2509, + "CTGACT": 2510, + "CTGACC": 2511, + "CTGACG": 2512, + "CTGAGA": 2513, + "CTGAGT": 2514, + "CTGAGC": 2515, + "CTGAGG": 2516, + "CTGTAA": 2517, + "CTGTAT": 2518, + "CTGTAC": 2519, + "CTGTAG": 2520, + "CTGTTA": 2521, + "CTGTTT": 2522, + "CTGTTC": 2523, + "CTGTTG": 2524, + "CTGTCA": 2525, + "CTGTCT": 2526, + "CTGTCC": 2527, + "CTGTCG": 2528, + "CTGTGA": 2529, + "CTGTGT": 2530, + "CTGTGC": 2531, + "CTGTGG": 2532, + "CTGCAA": 2533, + "CTGCAT": 2534, + "CTGCAC": 2535, + "CTGCAG": 2536, + "CTGCTA": 2537, + "CTGCTT": 2538, + "CTGCTC": 2539, + "CTGCTG": 2540, + "CTGCCA": 2541, + "CTGCCT": 2542, + "CTGCCC": 2543, + "CTGCCG": 2544, + "CTGCGA": 2545, + "CTGCGT": 2546, + "CTGCGC": 2547, + "CTGCGG": 2548, + "CTGGAA": 2549, + "CTGGAT": 2550, + "CTGGAC": 2551, + "CTGGAG": 2552, + "CTGGTA": 2553, + "CTGGTT": 2554, + "CTGGTC": 2555, + "CTGGTG": 2556, + "CTGGCA": 2557, + "CTGGCT": 2558, + "CTGGCC": 2559, + "CTGGCG": 2560, + "CTGGGA": 2561, + "CTGGGT": 2562, + "CTGGGC": 2563, + "CTGGGG": 2564, + "CCAAAA": 2565, + "CCAAAT": 2566, + "CCAAAC": 2567, + "CCAAAG": 2568, + "CCAATA": 2569, + "CCAATT": 2570, + "CCAATC": 2571, + "CCAATG": 2572, + "CCAACA": 2573, + "CCAACT": 2574, + "CCAACC": 2575, + "CCAACG": 2576, + "CCAAGA": 2577, + "CCAAGT": 2578, + "CCAAGC": 2579, + "CCAAGG": 2580, + "CCATAA": 2581, + "CCATAT": 2582, + "CCATAC": 2583, + "CCATAG": 2584, + "CCATTA": 2585, + "CCATTT": 2586, + "CCATTC": 2587, + "CCATTG": 2588, + "CCATCA": 2589, + "CCATCT": 2590, + "CCATCC": 2591, + "CCATCG": 2592, + "CCATGA": 2593, + "CCATGT": 2594, + "CCATGC": 2595, + "CCATGG": 2596, + "CCACAA": 2597, + "CCACAT": 2598, + "CCACAC": 2599, + "CCACAG": 2600, + "CCACTA": 2601, + "CCACTT": 2602, + "CCACTC": 2603, + "CCACTG": 2604, + "CCACCA": 2605, + "CCACCT": 2606, + "CCACCC": 2607, + "CCACCG": 2608, + "CCACGA": 2609, + "CCACGT": 2610, + "CCACGC": 2611, + "CCACGG": 2612, + "CCAGAA": 2613, + "CCAGAT": 2614, + "CCAGAC": 2615, + "CCAGAG": 2616, + "CCAGTA": 2617, + "CCAGTT": 2618, + "CCAGTC": 2619, + "CCAGTG": 2620, + "CCAGCA": 2621, + "CCAGCT": 2622, + "CCAGCC": 2623, + "CCAGCG": 2624, + "CCAGGA": 2625, + "CCAGGT": 2626, + "CCAGGC": 2627, + "CCAGGG": 2628, + "CCTAAA": 2629, + "CCTAAT": 2630, + "CCTAAC": 2631, + "CCTAAG": 2632, + "CCTATA": 2633, + "CCTATT": 2634, + "CCTATC": 2635, + "CCTATG": 2636, + "CCTACA": 2637, + "CCTACT": 2638, + "CCTACC": 2639, + "CCTACG": 2640, + "CCTAGA": 2641, + "CCTAGT": 2642, + "CCTAGC": 2643, + "CCTAGG": 2644, + "CCTTAA": 2645, + "CCTTAT": 2646, + "CCTTAC": 2647, + "CCTTAG": 2648, + "CCTTTA": 2649, + "CCTTTT": 2650, + "CCTTTC": 2651, + "CCTTTG": 2652, + "CCTTCA": 2653, + "CCTTCT": 2654, + "CCTTCC": 2655, + "CCTTCG": 2656, + "CCTTGA": 2657, + "CCTTGT": 2658, + "CCTTGC": 2659, + "CCTTGG": 2660, + "CCTCAA": 2661, + "CCTCAT": 2662, + "CCTCAC": 2663, + "CCTCAG": 2664, + "CCTCTA": 2665, + "CCTCTT": 2666, + "CCTCTC": 2667, + "CCTCTG": 2668, + "CCTCCA": 2669, + "CCTCCT": 2670, + "CCTCCC": 2671, + "CCTCCG": 2672, + "CCTCGA": 2673, + "CCTCGT": 2674, + "CCTCGC": 2675, + "CCTCGG": 2676, + "CCTGAA": 2677, + "CCTGAT": 2678, + "CCTGAC": 2679, + "CCTGAG": 2680, + "CCTGTA": 2681, + "CCTGTT": 2682, + "CCTGTC": 2683, + "CCTGTG": 2684, + "CCTGCA": 2685, + "CCTGCT": 2686, + "CCTGCC": 2687, + "CCTGCG": 2688, + "CCTGGA": 2689, + "CCTGGT": 2690, + "CCTGGC": 2691, + "CCTGGG": 2692, + "CCCAAA": 2693, + "CCCAAT": 2694, + "CCCAAC": 2695, + "CCCAAG": 2696, + "CCCATA": 2697, + "CCCATT": 2698, + "CCCATC": 2699, + "CCCATG": 2700, + "CCCACA": 2701, + "CCCACT": 2702, + "CCCACC": 2703, + "CCCACG": 2704, + "CCCAGA": 2705, + "CCCAGT": 2706, + "CCCAGC": 2707, + "CCCAGG": 2708, + "CCCTAA": 2709, + "CCCTAT": 2710, + "CCCTAC": 2711, + "CCCTAG": 2712, + "CCCTTA": 2713, + "CCCTTT": 2714, + "CCCTTC": 2715, + "CCCTTG": 2716, + "CCCTCA": 2717, + "CCCTCT": 2718, + "CCCTCC": 2719, + "CCCTCG": 2720, + "CCCTGA": 2721, + "CCCTGT": 2722, + "CCCTGC": 2723, + "CCCTGG": 2724, + "CCCCAA": 2725, + "CCCCAT": 2726, + "CCCCAC": 2727, + "CCCCAG": 2728, + "CCCCTA": 2729, + "CCCCTT": 2730, + "CCCCTC": 2731, + "CCCCTG": 2732, + "CCCCCA": 2733, + "CCCCCT": 2734, + "CCCCCC": 2735, + "CCCCCG": 2736, + "CCCCGA": 2737, + "CCCCGT": 2738, + "CCCCGC": 2739, + "CCCCGG": 2740, + "CCCGAA": 2741, + "CCCGAT": 2742, + "CCCGAC": 2743, + "CCCGAG": 2744, + "CCCGTA": 2745, + "CCCGTT": 2746, + "CCCGTC": 2747, + "CCCGTG": 2748, + "CCCGCA": 2749, + "CCCGCT": 2750, + "CCCGCC": 2751, + "CCCGCG": 2752, + "CCCGGA": 2753, + "CCCGGT": 2754, + "CCCGGC": 2755, + "CCCGGG": 2756, + "CCGAAA": 2757, + "CCGAAT": 2758, + "CCGAAC": 2759, + "CCGAAG": 2760, + "CCGATA": 2761, + "CCGATT": 2762, + "CCGATC": 2763, + "CCGATG": 2764, + "CCGACA": 2765, + "CCGACT": 2766, + "CCGACC": 2767, + "CCGACG": 2768, + "CCGAGA": 2769, + "CCGAGT": 2770, + "CCGAGC": 2771, + "CCGAGG": 2772, + "CCGTAA": 2773, + "CCGTAT": 2774, + "CCGTAC": 2775, + "CCGTAG": 2776, + "CCGTTA": 2777, + "CCGTTT": 2778, + "CCGTTC": 2779, + "CCGTTG": 2780, + "CCGTCA": 2781, + "CCGTCT": 2782, + "CCGTCC": 2783, + "CCGTCG": 2784, + "CCGTGA": 2785, + "CCGTGT": 2786, + "CCGTGC": 2787, + "CCGTGG": 2788, + "CCGCAA": 2789, + "CCGCAT": 2790, + "CCGCAC": 2791, + "CCGCAG": 2792, + "CCGCTA": 2793, + "CCGCTT": 2794, + "CCGCTC": 2795, + "CCGCTG": 2796, + "CCGCCA": 2797, + "CCGCCT": 2798, + "CCGCCC": 2799, + "CCGCCG": 2800, + "CCGCGA": 2801, + "CCGCGT": 2802, + "CCGCGC": 2803, + "CCGCGG": 2804, + "CCGGAA": 2805, + "CCGGAT": 2806, + "CCGGAC": 2807, + "CCGGAG": 2808, + "CCGGTA": 2809, + "CCGGTT": 2810, + "CCGGTC": 2811, + "CCGGTG": 2812, + "CCGGCA": 2813, + "CCGGCT": 2814, + "CCGGCC": 2815, + "CCGGCG": 2816, + "CCGGGA": 2817, + "CCGGGT": 2818, + "CCGGGC": 2819, + "CCGGGG": 2820, + "CGAAAA": 2821, + "CGAAAT": 2822, + "CGAAAC": 2823, + "CGAAAG": 2824, + "CGAATA": 2825, + "CGAATT": 2826, + "CGAATC": 2827, + "CGAATG": 2828, + "CGAACA": 2829, + "CGAACT": 2830, + "CGAACC": 2831, + "CGAACG": 2832, + "CGAAGA": 2833, + "CGAAGT": 2834, + "CGAAGC": 2835, + "CGAAGG": 2836, + "CGATAA": 2837, + "CGATAT": 2838, + "CGATAC": 2839, + "CGATAG": 2840, + "CGATTA": 2841, + "CGATTT": 2842, + "CGATTC": 2843, + "CGATTG": 2844, + "CGATCA": 2845, + "CGATCT": 2846, + "CGATCC": 2847, + "CGATCG": 2848, + "CGATGA": 2849, + "CGATGT": 2850, + "CGATGC": 2851, + "CGATGG": 2852, + "CGACAA": 2853, + "CGACAT": 2854, + "CGACAC": 2855, + "CGACAG": 2856, + "CGACTA": 2857, + "CGACTT": 2858, + "CGACTC": 2859, + "CGACTG": 2860, + "CGACCA": 2861, + "CGACCT": 2862, + "CGACCC": 2863, + "CGACCG": 2864, + "CGACGA": 2865, + "CGACGT": 2866, + "CGACGC": 2867, + "CGACGG": 2868, + "CGAGAA": 2869, + "CGAGAT": 2870, + "CGAGAC": 2871, + "CGAGAG": 2872, + "CGAGTA": 2873, + "CGAGTT": 2874, + "CGAGTC": 2875, + "CGAGTG": 2876, + "CGAGCA": 2877, + "CGAGCT": 2878, + "CGAGCC": 2879, + "CGAGCG": 2880, + "CGAGGA": 2881, + "CGAGGT": 2882, + "CGAGGC": 2883, + "CGAGGG": 2884, + "CGTAAA": 2885, + "CGTAAT": 2886, + "CGTAAC": 2887, + "CGTAAG": 2888, + "CGTATA": 2889, + "CGTATT": 2890, + "CGTATC": 2891, + "CGTATG": 2892, + "CGTACA": 2893, + "CGTACT": 2894, + "CGTACC": 2895, + "CGTACG": 2896, + "CGTAGA": 2897, + "CGTAGT": 2898, + "CGTAGC": 2899, + "CGTAGG": 2900, + "CGTTAA": 2901, + "CGTTAT": 2902, + "CGTTAC": 2903, + "CGTTAG": 2904, + "CGTTTA": 2905, + "CGTTTT": 2906, + "CGTTTC": 2907, + "CGTTTG": 2908, + "CGTTCA": 2909, + "CGTTCT": 2910, + "CGTTCC": 2911, + "CGTTCG": 2912, + "CGTTGA": 2913, + "CGTTGT": 2914, + "CGTTGC": 2915, + "CGTTGG": 2916, + "CGTCAA": 2917, + "CGTCAT": 2918, + "CGTCAC": 2919, + "CGTCAG": 2920, + "CGTCTA": 2921, + "CGTCTT": 2922, + "CGTCTC": 2923, + "CGTCTG": 2924, + "CGTCCA": 2925, + "CGTCCT": 2926, + "CGTCCC": 2927, + "CGTCCG": 2928, + "CGTCGA": 2929, + "CGTCGT": 2930, + "CGTCGC": 2931, + "CGTCGG": 2932, + "CGTGAA": 2933, + "CGTGAT": 2934, + "CGTGAC": 2935, + "CGTGAG": 2936, + "CGTGTA": 2937, + "CGTGTT": 2938, + "CGTGTC": 2939, + "CGTGTG": 2940, + "CGTGCA": 2941, + "CGTGCT": 2942, + "CGTGCC": 2943, + "CGTGCG": 2944, + "CGTGGA": 2945, + "CGTGGT": 2946, + "CGTGGC": 2947, + "CGTGGG": 2948, + "CGCAAA": 2949, + "CGCAAT": 2950, + "CGCAAC": 2951, + "CGCAAG": 2952, + "CGCATA": 2953, + "CGCATT": 2954, + "CGCATC": 2955, + "CGCATG": 2956, + "CGCACA": 2957, + "CGCACT": 2958, + "CGCACC": 2959, + "CGCACG": 2960, + "CGCAGA": 2961, + "CGCAGT": 2962, + "CGCAGC": 2963, + "CGCAGG": 2964, + "CGCTAA": 2965, + "CGCTAT": 2966, + "CGCTAC": 2967, + "CGCTAG": 2968, + "CGCTTA": 2969, + "CGCTTT": 2970, + "CGCTTC": 2971, + "CGCTTG": 2972, + "CGCTCA": 2973, + "CGCTCT": 2974, + "CGCTCC": 2975, + "CGCTCG": 2976, + "CGCTGA": 2977, + "CGCTGT": 2978, + "CGCTGC": 2979, + "CGCTGG": 2980, + "CGCCAA": 2981, + "CGCCAT": 2982, + "CGCCAC": 2983, + "CGCCAG": 2984, + "CGCCTA": 2985, + "CGCCTT": 2986, + "CGCCTC": 2987, + "CGCCTG": 2988, + "CGCCCA": 2989, + "CGCCCT": 2990, + "CGCCCC": 2991, + "CGCCCG": 2992, + "CGCCGA": 2993, + "CGCCGT": 2994, + "CGCCGC": 2995, + "CGCCGG": 2996, + "CGCGAA": 2997, + "CGCGAT": 2998, + "CGCGAC": 2999, + "CGCGAG": 3000, + "CGCGTA": 3001, + "CGCGTT": 3002, + "CGCGTC": 3003, + "CGCGTG": 3004, + "CGCGCA": 3005, + "CGCGCT": 3006, + "CGCGCC": 3007, + "CGCGCG": 3008, + "CGCGGA": 3009, + "CGCGGT": 3010, + "CGCGGC": 3011, + "CGCGGG": 3012, + "CGGAAA": 3013, + "CGGAAT": 3014, + "CGGAAC": 3015, + "CGGAAG": 3016, + "CGGATA": 3017, + "CGGATT": 3018, + "CGGATC": 3019, + "CGGATG": 3020, + "CGGACA": 3021, + "CGGACT": 3022, + "CGGACC": 3023, + "CGGACG": 3024, + "CGGAGA": 3025, + "CGGAGT": 3026, + "CGGAGC": 3027, + "CGGAGG": 3028, + "CGGTAA": 3029, + "CGGTAT": 3030, + "CGGTAC": 3031, + "CGGTAG": 3032, + "CGGTTA": 3033, + "CGGTTT": 3034, + "CGGTTC": 3035, + "CGGTTG": 3036, + "CGGTCA": 3037, + "CGGTCT": 3038, + "CGGTCC": 3039, + "CGGTCG": 3040, + "CGGTGA": 3041, + "CGGTGT": 3042, + "CGGTGC": 3043, + "CGGTGG": 3044, + "CGGCAA": 3045, + "CGGCAT": 3046, + "CGGCAC": 3047, + "CGGCAG": 3048, + "CGGCTA": 3049, + "CGGCTT": 3050, + "CGGCTC": 3051, + "CGGCTG": 3052, + "CGGCCA": 3053, + "CGGCCT": 3054, + "CGGCCC": 3055, + "CGGCCG": 3056, + "CGGCGA": 3057, + "CGGCGT": 3058, + "CGGCGC": 3059, + "CGGCGG": 3060, + "CGGGAA": 3061, + "CGGGAT": 3062, + "CGGGAC": 3063, + "CGGGAG": 3064, + "CGGGTA": 3065, + "CGGGTT": 3066, + "CGGGTC": 3067, + "CGGGTG": 3068, + "CGGGCA": 3069, + "CGGGCT": 3070, + "CGGGCC": 3071, + "CGGGCG": 3072, + "CGGGGA": 3073, + "CGGGGT": 3074, + "CGGGGC": 3075, + "CGGGGG": 3076, + "GAAAAA": 3077, + "GAAAAT": 3078, + "GAAAAC": 3079, + "GAAAAG": 3080, + "GAAATA": 3081, + "GAAATT": 3082, + "GAAATC": 3083, + "GAAATG": 3084, + "GAAACA": 3085, + "GAAACT": 3086, + "GAAACC": 3087, + "GAAACG": 3088, + "GAAAGA": 3089, + "GAAAGT": 3090, + "GAAAGC": 3091, + "GAAAGG": 3092, + "GAATAA": 3093, + "GAATAT": 3094, + "GAATAC": 3095, + "GAATAG": 3096, + "GAATTA": 3097, + "GAATTT": 3098, + "GAATTC": 3099, + "GAATTG": 3100, + "GAATCA": 3101, + "GAATCT": 3102, + "GAATCC": 3103, + "GAATCG": 3104, + "GAATGA": 3105, + "GAATGT": 3106, + "GAATGC": 3107, + "GAATGG": 3108, + "GAACAA": 3109, + "GAACAT": 3110, + "GAACAC": 3111, + "GAACAG": 3112, + "GAACTA": 3113, + "GAACTT": 3114, + "GAACTC": 3115, + "GAACTG": 3116, + "GAACCA": 3117, + "GAACCT": 3118, + "GAACCC": 3119, + "GAACCG": 3120, + "GAACGA": 3121, + "GAACGT": 3122, + "GAACGC": 3123, + "GAACGG": 3124, + "GAAGAA": 3125, + "GAAGAT": 3126, + "GAAGAC": 3127, + "GAAGAG": 3128, + "GAAGTA": 3129, + "GAAGTT": 3130, + "GAAGTC": 3131, + "GAAGTG": 3132, + "GAAGCA": 3133, + "GAAGCT": 3134, + "GAAGCC": 3135, + "GAAGCG": 3136, + "GAAGGA": 3137, + "GAAGGT": 3138, + "GAAGGC": 3139, + "GAAGGG": 3140, + "GATAAA": 3141, + "GATAAT": 3142, + "GATAAC": 3143, + "GATAAG": 3144, + "GATATA": 3145, + "GATATT": 3146, + "GATATC": 3147, + "GATATG": 3148, + "GATACA": 3149, + "GATACT": 3150, + "GATACC": 3151, + "GATACG": 3152, + "GATAGA": 3153, + "GATAGT": 3154, + "GATAGC": 3155, + "GATAGG": 3156, + "GATTAA": 3157, + "GATTAT": 3158, + "GATTAC": 3159, + "GATTAG": 3160, + "GATTTA": 3161, + "GATTTT": 3162, + "GATTTC": 3163, + "GATTTG": 3164, + "GATTCA": 3165, + "GATTCT": 3166, + "GATTCC": 3167, + "GATTCG": 3168, + "GATTGA": 3169, + "GATTGT": 3170, + "GATTGC": 3171, + "GATTGG": 3172, + "GATCAA": 3173, + "GATCAT": 3174, + "GATCAC": 3175, + "GATCAG": 3176, + "GATCTA": 3177, + "GATCTT": 3178, + "GATCTC": 3179, + "GATCTG": 3180, + "GATCCA": 3181, + "GATCCT": 3182, + "GATCCC": 3183, + "GATCCG": 3184, + "GATCGA": 3185, + "GATCGT": 3186, + "GATCGC": 3187, + "GATCGG": 3188, + "GATGAA": 3189, + "GATGAT": 3190, + "GATGAC": 3191, + "GATGAG": 3192, + "GATGTA": 3193, + "GATGTT": 3194, + "GATGTC": 3195, + "GATGTG": 3196, + "GATGCA": 3197, + "GATGCT": 3198, + "GATGCC": 3199, + "GATGCG": 3200, + "GATGGA": 3201, + "GATGGT": 3202, + "GATGGC": 3203, + "GATGGG": 3204, + "GACAAA": 3205, + "GACAAT": 3206, + "GACAAC": 3207, + "GACAAG": 3208, + "GACATA": 3209, + "GACATT": 3210, + "GACATC": 3211, + "GACATG": 3212, + "GACACA": 3213, + "GACACT": 3214, + "GACACC": 3215, + "GACACG": 3216, + "GACAGA": 3217, + "GACAGT": 3218, + "GACAGC": 3219, + "GACAGG": 3220, + "GACTAA": 3221, + "GACTAT": 3222, + "GACTAC": 3223, + "GACTAG": 3224, + "GACTTA": 3225, + "GACTTT": 3226, + "GACTTC": 3227, + "GACTTG": 3228, + "GACTCA": 3229, + "GACTCT": 3230, + "GACTCC": 3231, + "GACTCG": 3232, + "GACTGA": 3233, + "GACTGT": 3234, + "GACTGC": 3235, + "GACTGG": 3236, + "GACCAA": 3237, + "GACCAT": 3238, + "GACCAC": 3239, + "GACCAG": 3240, + "GACCTA": 3241, + "GACCTT": 3242, + "GACCTC": 3243, + "GACCTG": 3244, + "GACCCA": 3245, + "GACCCT": 3246, + "GACCCC": 3247, + "GACCCG": 3248, + "GACCGA": 3249, + "GACCGT": 3250, + "GACCGC": 3251, + "GACCGG": 3252, + "GACGAA": 3253, + "GACGAT": 3254, + "GACGAC": 3255, + "GACGAG": 3256, + "GACGTA": 3257, + "GACGTT": 3258, + "GACGTC": 3259, + "GACGTG": 3260, + "GACGCA": 3261, + "GACGCT": 3262, + "GACGCC": 3263, + "GACGCG": 3264, + "GACGGA": 3265, + "GACGGT": 3266, + "GACGGC": 3267, + "GACGGG": 3268, + "GAGAAA": 3269, + "GAGAAT": 3270, + "GAGAAC": 3271, + "GAGAAG": 3272, + "GAGATA": 3273, + "GAGATT": 3274, + "GAGATC": 3275, + "GAGATG": 3276, + "GAGACA": 3277, + "GAGACT": 3278, + "GAGACC": 3279, + "GAGACG": 3280, + "GAGAGA": 3281, + "GAGAGT": 3282, + "GAGAGC": 3283, + "GAGAGG": 3284, + "GAGTAA": 3285, + "GAGTAT": 3286, + "GAGTAC": 3287, + "GAGTAG": 3288, + "GAGTTA": 3289, + "GAGTTT": 3290, + "GAGTTC": 3291, + "GAGTTG": 3292, + "GAGTCA": 3293, + "GAGTCT": 3294, + "GAGTCC": 3295, + "GAGTCG": 3296, + "GAGTGA": 3297, + "GAGTGT": 3298, + "GAGTGC": 3299, + "GAGTGG": 3300, + "GAGCAA": 3301, + "GAGCAT": 3302, + "GAGCAC": 3303, + "GAGCAG": 3304, + "GAGCTA": 3305, + "GAGCTT": 3306, + "GAGCTC": 3307, + "GAGCTG": 3308, + "GAGCCA": 3309, + "GAGCCT": 3310, + "GAGCCC": 3311, + "GAGCCG": 3312, + "GAGCGA": 3313, + "GAGCGT": 3314, + "GAGCGC": 3315, + "GAGCGG": 3316, + "GAGGAA": 3317, + "GAGGAT": 3318, + "GAGGAC": 3319, + "GAGGAG": 3320, + "GAGGTA": 3321, + "GAGGTT": 3322, + "GAGGTC": 3323, + "GAGGTG": 3324, + "GAGGCA": 3325, + "GAGGCT": 3326, + "GAGGCC": 3327, + "GAGGCG": 3328, + "GAGGGA": 3329, + "GAGGGT": 3330, + "GAGGGC": 3331, + "GAGGGG": 3332, + "GTAAAA": 3333, + "GTAAAT": 3334, + "GTAAAC": 3335, + "GTAAAG": 3336, + "GTAATA": 3337, + "GTAATT": 3338, + "GTAATC": 3339, + "GTAATG": 3340, + "GTAACA": 3341, + "GTAACT": 3342, + "GTAACC": 3343, + "GTAACG": 3344, + "GTAAGA": 3345, + "GTAAGT": 3346, + "GTAAGC": 3347, + "GTAAGG": 3348, + "GTATAA": 3349, + "GTATAT": 3350, + "GTATAC": 3351, + "GTATAG": 3352, + "GTATTA": 3353, + "GTATTT": 3354, + "GTATTC": 3355, + "GTATTG": 3356, + "GTATCA": 3357, + "GTATCT": 3358, + "GTATCC": 3359, + "GTATCG": 3360, + "GTATGA": 3361, + "GTATGT": 3362, + "GTATGC": 3363, + "GTATGG": 3364, + "GTACAA": 3365, + "GTACAT": 3366, + "GTACAC": 3367, + "GTACAG": 3368, + "GTACTA": 3369, + "GTACTT": 3370, + "GTACTC": 3371, + "GTACTG": 3372, + "GTACCA": 3373, + "GTACCT": 3374, + "GTACCC": 3375, + "GTACCG": 3376, + "GTACGA": 3377, + "GTACGT": 3378, + "GTACGC": 3379, + "GTACGG": 3380, + "GTAGAA": 3381, + "GTAGAT": 3382, + "GTAGAC": 3383, + "GTAGAG": 3384, + "GTAGTA": 3385, + "GTAGTT": 3386, + "GTAGTC": 3387, + "GTAGTG": 3388, + "GTAGCA": 3389, + "GTAGCT": 3390, + "GTAGCC": 3391, + "GTAGCG": 3392, + "GTAGGA": 3393, + "GTAGGT": 3394, + "GTAGGC": 3395, + "GTAGGG": 3396, + "GTTAAA": 3397, + "GTTAAT": 3398, + "GTTAAC": 3399, + "GTTAAG": 3400, + "GTTATA": 3401, + "GTTATT": 3402, + "GTTATC": 3403, + "GTTATG": 3404, + "GTTACA": 3405, + "GTTACT": 3406, + "GTTACC": 3407, + "GTTACG": 3408, + "GTTAGA": 3409, + "GTTAGT": 3410, + "GTTAGC": 3411, + "GTTAGG": 3412, + "GTTTAA": 3413, + "GTTTAT": 3414, + "GTTTAC": 3415, + "GTTTAG": 3416, + "GTTTTA": 3417, + "GTTTTT": 3418, + "GTTTTC": 3419, + "GTTTTG": 3420, + "GTTTCA": 3421, + "GTTTCT": 3422, + "GTTTCC": 3423, + "GTTTCG": 3424, + "GTTTGA": 3425, + "GTTTGT": 3426, + "GTTTGC": 3427, + "GTTTGG": 3428, + "GTTCAA": 3429, + "GTTCAT": 3430, + "GTTCAC": 3431, + "GTTCAG": 3432, + "GTTCTA": 3433, + "GTTCTT": 3434, + "GTTCTC": 3435, + "GTTCTG": 3436, + "GTTCCA": 3437, + "GTTCCT": 3438, + "GTTCCC": 3439, + "GTTCCG": 3440, + "GTTCGA": 3441, + "GTTCGT": 3442, + "GTTCGC": 3443, + "GTTCGG": 3444, + "GTTGAA": 3445, + "GTTGAT": 3446, + "GTTGAC": 3447, + "GTTGAG": 3448, + "GTTGTA": 3449, + "GTTGTT": 3450, + "GTTGTC": 3451, + "GTTGTG": 3452, + "GTTGCA": 3453, + "GTTGCT": 3454, + "GTTGCC": 3455, + "GTTGCG": 3456, + "GTTGGA": 3457, + "GTTGGT": 3458, + "GTTGGC": 3459, + "GTTGGG": 3460, + "GTCAAA": 3461, + "GTCAAT": 3462, + "GTCAAC": 3463, + "GTCAAG": 3464, + "GTCATA": 3465, + "GTCATT": 3466, + "GTCATC": 3467, + "GTCATG": 3468, + "GTCACA": 3469, + "GTCACT": 3470, + "GTCACC": 3471, + "GTCACG": 3472, + "GTCAGA": 3473, + "GTCAGT": 3474, + "GTCAGC": 3475, + "GTCAGG": 3476, + "GTCTAA": 3477, + "GTCTAT": 3478, + "GTCTAC": 3479, + "GTCTAG": 3480, + "GTCTTA": 3481, + "GTCTTT": 3482, + "GTCTTC": 3483, + "GTCTTG": 3484, + "GTCTCA": 3485, + "GTCTCT": 3486, + "GTCTCC": 3487, + "GTCTCG": 3488, + "GTCTGA": 3489, + "GTCTGT": 3490, + "GTCTGC": 3491, + "GTCTGG": 3492, + "GTCCAA": 3493, + "GTCCAT": 3494, + "GTCCAC": 3495, + "GTCCAG": 3496, + "GTCCTA": 3497, + "GTCCTT": 3498, + "GTCCTC": 3499, + "GTCCTG": 3500, + "GTCCCA": 3501, + "GTCCCT": 3502, + "GTCCCC": 3503, + "GTCCCG": 3504, + "GTCCGA": 3505, + "GTCCGT": 3506, + "GTCCGC": 3507, + "GTCCGG": 3508, + "GTCGAA": 3509, + "GTCGAT": 3510, + "GTCGAC": 3511, + "GTCGAG": 3512, + "GTCGTA": 3513, + "GTCGTT": 3514, + "GTCGTC": 3515, + "GTCGTG": 3516, + "GTCGCA": 3517, + "GTCGCT": 3518, + "GTCGCC": 3519, + "GTCGCG": 3520, + "GTCGGA": 3521, + "GTCGGT": 3522, + "GTCGGC": 3523, + "GTCGGG": 3524, + "GTGAAA": 3525, + "GTGAAT": 3526, + "GTGAAC": 3527, + "GTGAAG": 3528, + "GTGATA": 3529, + "GTGATT": 3530, + "GTGATC": 3531, + "GTGATG": 3532, + "GTGACA": 3533, + "GTGACT": 3534, + "GTGACC": 3535, + "GTGACG": 3536, + "GTGAGA": 3537, + "GTGAGT": 3538, + "GTGAGC": 3539, + "GTGAGG": 3540, + "GTGTAA": 3541, + "GTGTAT": 3542, + "GTGTAC": 3543, + "GTGTAG": 3544, + "GTGTTA": 3545, + "GTGTTT": 3546, + "GTGTTC": 3547, + "GTGTTG": 3548, + "GTGTCA": 3549, + "GTGTCT": 3550, + "GTGTCC": 3551, + "GTGTCG": 3552, + "GTGTGA": 3553, + "GTGTGT": 3554, + "GTGTGC": 3555, + "GTGTGG": 3556, + "GTGCAA": 3557, + "GTGCAT": 3558, + "GTGCAC": 3559, + "GTGCAG": 3560, + "GTGCTA": 3561, + "GTGCTT": 3562, + "GTGCTC": 3563, + "GTGCTG": 3564, + "GTGCCA": 3565, + "GTGCCT": 3566, + "GTGCCC": 3567, + "GTGCCG": 3568, + "GTGCGA": 3569, + "GTGCGT": 3570, + "GTGCGC": 3571, + "GTGCGG": 3572, + "GTGGAA": 3573, + "GTGGAT": 3574, + "GTGGAC": 3575, + "GTGGAG": 3576, + "GTGGTA": 3577, + "GTGGTT": 3578, + "GTGGTC": 3579, + "GTGGTG": 3580, + "GTGGCA": 3581, + "GTGGCT": 3582, + "GTGGCC": 3583, + "GTGGCG": 3584, + "GTGGGA": 3585, + "GTGGGT": 3586, + "GTGGGC": 3587, + "GTGGGG": 3588, + "GCAAAA": 3589, + "GCAAAT": 3590, + "GCAAAC": 3591, + "GCAAAG": 3592, + "GCAATA": 3593, + "GCAATT": 3594, + "GCAATC": 3595, + "GCAATG": 3596, + "GCAACA": 3597, + "GCAACT": 3598, + "GCAACC": 3599, + "GCAACG": 3600, + "GCAAGA": 3601, + "GCAAGT": 3602, + "GCAAGC": 3603, + "GCAAGG": 3604, + "GCATAA": 3605, + "GCATAT": 3606, + "GCATAC": 3607, + "GCATAG": 3608, + "GCATTA": 3609, + "GCATTT": 3610, + "GCATTC": 3611, + "GCATTG": 3612, + "GCATCA": 3613, + "GCATCT": 3614, + "GCATCC": 3615, + "GCATCG": 3616, + "GCATGA": 3617, + "GCATGT": 3618, + "GCATGC": 3619, + "GCATGG": 3620, + "GCACAA": 3621, + "GCACAT": 3622, + "GCACAC": 3623, + "GCACAG": 3624, + "GCACTA": 3625, + "GCACTT": 3626, + "GCACTC": 3627, + "GCACTG": 3628, + "GCACCA": 3629, + "GCACCT": 3630, + "GCACCC": 3631, + "GCACCG": 3632, + "GCACGA": 3633, + "GCACGT": 3634, + "GCACGC": 3635, + "GCACGG": 3636, + "GCAGAA": 3637, + "GCAGAT": 3638, + "GCAGAC": 3639, + "GCAGAG": 3640, + "GCAGTA": 3641, + "GCAGTT": 3642, + "GCAGTC": 3643, + "GCAGTG": 3644, + "GCAGCA": 3645, + "GCAGCT": 3646, + "GCAGCC": 3647, + "GCAGCG": 3648, + "GCAGGA": 3649, + "GCAGGT": 3650, + "GCAGGC": 3651, + "GCAGGG": 3652, + "GCTAAA": 3653, + "GCTAAT": 3654, + "GCTAAC": 3655, + "GCTAAG": 3656, + "GCTATA": 3657, + "GCTATT": 3658, + "GCTATC": 3659, + "GCTATG": 3660, + "GCTACA": 3661, + "GCTACT": 3662, + "GCTACC": 3663, + "GCTACG": 3664, + "GCTAGA": 3665, + "GCTAGT": 3666, + "GCTAGC": 3667, + "GCTAGG": 3668, + "GCTTAA": 3669, + "GCTTAT": 3670, + "GCTTAC": 3671, + "GCTTAG": 3672, + "GCTTTA": 3673, + "GCTTTT": 3674, + "GCTTTC": 3675, + "GCTTTG": 3676, + "GCTTCA": 3677, + "GCTTCT": 3678, + "GCTTCC": 3679, + "GCTTCG": 3680, + "GCTTGA": 3681, + "GCTTGT": 3682, + "GCTTGC": 3683, + "GCTTGG": 3684, + "GCTCAA": 3685, + "GCTCAT": 3686, + "GCTCAC": 3687, + "GCTCAG": 3688, + "GCTCTA": 3689, + "GCTCTT": 3690, + "GCTCTC": 3691, + "GCTCTG": 3692, + "GCTCCA": 3693, + "GCTCCT": 3694, + "GCTCCC": 3695, + "GCTCCG": 3696, + "GCTCGA": 3697, + "GCTCGT": 3698, + "GCTCGC": 3699, + "GCTCGG": 3700, + "GCTGAA": 3701, + "GCTGAT": 3702, + "GCTGAC": 3703, + "GCTGAG": 3704, + "GCTGTA": 3705, + "GCTGTT": 3706, + "GCTGTC": 3707, + "GCTGTG": 3708, + "GCTGCA": 3709, + "GCTGCT": 3710, + "GCTGCC": 3711, + "GCTGCG": 3712, + "GCTGGA": 3713, + "GCTGGT": 3714, + "GCTGGC": 3715, + "GCTGGG": 3716, + "GCCAAA": 3717, + "GCCAAT": 3718, + "GCCAAC": 3719, + "GCCAAG": 3720, + "GCCATA": 3721, + "GCCATT": 3722, + "GCCATC": 3723, + "GCCATG": 3724, + "GCCACA": 3725, + "GCCACT": 3726, + "GCCACC": 3727, + "GCCACG": 3728, + "GCCAGA": 3729, + "GCCAGT": 3730, + "GCCAGC": 3731, + "GCCAGG": 3732, + "GCCTAA": 3733, + "GCCTAT": 3734, + "GCCTAC": 3735, + "GCCTAG": 3736, + "GCCTTA": 3737, + "GCCTTT": 3738, + "GCCTTC": 3739, + "GCCTTG": 3740, + "GCCTCA": 3741, + "GCCTCT": 3742, + "GCCTCC": 3743, + "GCCTCG": 3744, + "GCCTGA": 3745, + "GCCTGT": 3746, + "GCCTGC": 3747, + "GCCTGG": 3748, + "GCCCAA": 3749, + "GCCCAT": 3750, + "GCCCAC": 3751, + "GCCCAG": 3752, + "GCCCTA": 3753, + "GCCCTT": 3754, + "GCCCTC": 3755, + "GCCCTG": 3756, + "GCCCCA": 3757, + "GCCCCT": 3758, + "GCCCCC": 3759, + "GCCCCG": 3760, + "GCCCGA": 3761, + "GCCCGT": 3762, + "GCCCGC": 3763, + "GCCCGG": 3764, + "GCCGAA": 3765, + "GCCGAT": 3766, + "GCCGAC": 3767, + "GCCGAG": 3768, + "GCCGTA": 3769, + "GCCGTT": 3770, + "GCCGTC": 3771, + "GCCGTG": 3772, + "GCCGCA": 3773, + "GCCGCT": 3774, + "GCCGCC": 3775, + "GCCGCG": 3776, + "GCCGGA": 3777, + "GCCGGT": 3778, + "GCCGGC": 3779, + "GCCGGG": 3780, + "GCGAAA": 3781, + "GCGAAT": 3782, + "GCGAAC": 3783, + "GCGAAG": 3784, + "GCGATA": 3785, + "GCGATT": 3786, + "GCGATC": 3787, + "GCGATG": 3788, + "GCGACA": 3789, + "GCGACT": 3790, + "GCGACC": 3791, + "GCGACG": 3792, + "GCGAGA": 3793, + "GCGAGT": 3794, + "GCGAGC": 3795, + "GCGAGG": 3796, + "GCGTAA": 3797, + "GCGTAT": 3798, + "GCGTAC": 3799, + "GCGTAG": 3800, + "GCGTTA": 3801, + "GCGTTT": 3802, + "GCGTTC": 3803, + "GCGTTG": 3804, + "GCGTCA": 3805, + "GCGTCT": 3806, + "GCGTCC": 3807, + "GCGTCG": 3808, + "GCGTGA": 3809, + "GCGTGT": 3810, + "GCGTGC": 3811, + "GCGTGG": 3812, + "GCGCAA": 3813, + "GCGCAT": 3814, + "GCGCAC": 3815, + "GCGCAG": 3816, + "GCGCTA": 3817, + "GCGCTT": 3818, + "GCGCTC": 3819, + "GCGCTG": 3820, + "GCGCCA": 3821, + "GCGCCT": 3822, + "GCGCCC": 3823, + "GCGCCG": 3824, + "GCGCGA": 3825, + "GCGCGT": 3826, + "GCGCGC": 3827, + "GCGCGG": 3828, + "GCGGAA": 3829, + "GCGGAT": 3830, + "GCGGAC": 3831, + "GCGGAG": 3832, + "GCGGTA": 3833, + "GCGGTT": 3834, + "GCGGTC": 3835, + "GCGGTG": 3836, + "GCGGCA": 3837, + "GCGGCT": 3838, + "GCGGCC": 3839, + "GCGGCG": 3840, + "GCGGGA": 3841, + "GCGGGT": 3842, + "GCGGGC": 3843, + "GCGGGG": 3844, + "GGAAAA": 3845, + "GGAAAT": 3846, + "GGAAAC": 3847, + "GGAAAG": 3848, + "GGAATA": 3849, + "GGAATT": 3850, + "GGAATC": 3851, + "GGAATG": 3852, + "GGAACA": 3853, + "GGAACT": 3854, + "GGAACC": 3855, + "GGAACG": 3856, + "GGAAGA": 3857, + "GGAAGT": 3858, + "GGAAGC": 3859, + "GGAAGG": 3860, + "GGATAA": 3861, + "GGATAT": 3862, + "GGATAC": 3863, + "GGATAG": 3864, + "GGATTA": 3865, + "GGATTT": 3866, + "GGATTC": 3867, + "GGATTG": 3868, + "GGATCA": 3869, + "GGATCT": 3870, + "GGATCC": 3871, + "GGATCG": 3872, + "GGATGA": 3873, + "GGATGT": 3874, + "GGATGC": 3875, + "GGATGG": 3876, + "GGACAA": 3877, + "GGACAT": 3878, + "GGACAC": 3879, + "GGACAG": 3880, + "GGACTA": 3881, + "GGACTT": 3882, + "GGACTC": 3883, + "GGACTG": 3884, + "GGACCA": 3885, + "GGACCT": 3886, + "GGACCC": 3887, + "GGACCG": 3888, + "GGACGA": 3889, + "GGACGT": 3890, + "GGACGC": 3891, + "GGACGG": 3892, + "GGAGAA": 3893, + "GGAGAT": 3894, + "GGAGAC": 3895, + "GGAGAG": 3896, + "GGAGTA": 3897, + "GGAGTT": 3898, + "GGAGTC": 3899, + "GGAGTG": 3900, + "GGAGCA": 3901, + "GGAGCT": 3902, + "GGAGCC": 3903, + "GGAGCG": 3904, + "GGAGGA": 3905, + "GGAGGT": 3906, + "GGAGGC": 3907, + "GGAGGG": 3908, + "GGTAAA": 3909, + "GGTAAT": 3910, + "GGTAAC": 3911, + "GGTAAG": 3912, + "GGTATA": 3913, + "GGTATT": 3914, + "GGTATC": 3915, + "GGTATG": 3916, + "GGTACA": 3917, + "GGTACT": 3918, + "GGTACC": 3919, + "GGTACG": 3920, + "GGTAGA": 3921, + "GGTAGT": 3922, + "GGTAGC": 3923, + "GGTAGG": 3924, + "GGTTAA": 3925, + "GGTTAT": 3926, + "GGTTAC": 3927, + "GGTTAG": 3928, + "GGTTTA": 3929, + "GGTTTT": 3930, + "GGTTTC": 3931, + "GGTTTG": 3932, + "GGTTCA": 3933, + "GGTTCT": 3934, + "GGTTCC": 3935, + "GGTTCG": 3936, + "GGTTGA": 3937, + "GGTTGT": 3938, + "GGTTGC": 3939, + "GGTTGG": 3940, + "GGTCAA": 3941, + "GGTCAT": 3942, + "GGTCAC": 3943, + "GGTCAG": 3944, + "GGTCTA": 3945, + "GGTCTT": 3946, + "GGTCTC": 3947, + "GGTCTG": 3948, + "GGTCCA": 3949, + "GGTCCT": 3950, + "GGTCCC": 3951, + "GGTCCG": 3952, + "GGTCGA": 3953, + "GGTCGT": 3954, + "GGTCGC": 3955, + "GGTCGG": 3956, + "GGTGAA": 3957, + "GGTGAT": 3958, + "GGTGAC": 3959, + "GGTGAG": 3960, + "GGTGTA": 3961, + "GGTGTT": 3962, + "GGTGTC": 3963, + "GGTGTG": 3964, + "GGTGCA": 3965, + "GGTGCT": 3966, + "GGTGCC": 3967, + "GGTGCG": 3968, + "GGTGGA": 3969, + "GGTGGT": 3970, + "GGTGGC": 3971, + "GGTGGG": 3972, + "GGCAAA": 3973, + "GGCAAT": 3974, + "GGCAAC": 3975, + "GGCAAG": 3976, + "GGCATA": 3977, + "GGCATT": 3978, + "GGCATC": 3979, + "GGCATG": 3980, + "GGCACA": 3981, + "GGCACT": 3982, + "GGCACC": 3983, + "GGCACG": 3984, + "GGCAGA": 3985, + "GGCAGT": 3986, + "GGCAGC": 3987, + "GGCAGG": 3988, + "GGCTAA": 3989, + "GGCTAT": 3990, + "GGCTAC": 3991, + "GGCTAG": 3992, + "GGCTTA": 3993, + "GGCTTT": 3994, + "GGCTTC": 3995, + "GGCTTG": 3996, + "GGCTCA": 3997, + "GGCTCT": 3998, + "GGCTCC": 3999, + "GGCTCG": 4000, + "GGCTGA": 4001, + "GGCTGT": 4002, + "GGCTGC": 4003, + "GGCTGG": 4004, + "GGCCAA": 4005, + "GGCCAT": 4006, + "GGCCAC": 4007, + "GGCCAG": 4008, + "GGCCTA": 4009, + "GGCCTT": 4010, + "GGCCTC": 4011, + "GGCCTG": 4012, + "GGCCCA": 4013, + "GGCCCT": 4014, + "GGCCCC": 4015, + "GGCCCG": 4016, + "GGCCGA": 4017, + "GGCCGT": 4018, + "GGCCGC": 4019, + "GGCCGG": 4020, + "GGCGAA": 4021, + "GGCGAT": 4022, + "GGCGAC": 4023, + "GGCGAG": 4024, + "GGCGTA": 4025, + "GGCGTT": 4026, + "GGCGTC": 4027, + "GGCGTG": 4028, + "GGCGCA": 4029, + "GGCGCT": 4030, + "GGCGCC": 4031, + "GGCGCG": 4032, + "GGCGGA": 4033, + "GGCGGT": 4034, + "GGCGGC": 4035, + "GGCGGG": 4036, + "GGGAAA": 4037, + "GGGAAT": 4038, + "GGGAAC": 4039, + "GGGAAG": 4040, + "GGGATA": 4041, + "GGGATT": 4042, + "GGGATC": 4043, + "GGGATG": 4044, + "GGGACA": 4045, + "GGGACT": 4046, + "GGGACC": 4047, + "GGGACG": 4048, + "GGGAGA": 4049, + "GGGAGT": 4050, + "GGGAGC": 4051, + "GGGAGG": 4052, + "GGGTAA": 4053, + "GGGTAT": 4054, + "GGGTAC": 4055, + "GGGTAG": 4056, + "GGGTTA": 4057, + "GGGTTT": 4058, + "GGGTTC": 4059, + "GGGTTG": 4060, + "GGGTCA": 4061, + "GGGTCT": 4062, + "GGGTCC": 4063, + "GGGTCG": 4064, + "GGGTGA": 4065, + "GGGTGT": 4066, + "GGGTGC": 4067, + "GGGTGG": 4068, + "GGGCAA": 4069, + "GGGCAT": 4070, + "GGGCAC": 4071, + "GGGCAG": 4072, + "GGGCTA": 4073, + "GGGCTT": 4074, + "GGGCTC": 4075, + "GGGCTG": 4076, + "GGGCCA": 4077, + "GGGCCT": 4078, + "GGGCCC": 4079, + "GGGCCG": 4080, + "GGGCGA": 4081, + "GGGCGT": 4082, + "GGGCGC": 4083, + "GGGCGG": 4084, + "GGGGAA": 4085, + "GGGGAT": 4086, + "GGGGAC": 4087, + "GGGGAG": 4088, + "GGGGTA": 4089, + "GGGGTT": 4090, + "GGGGTC": 4091, + "GGGGTG": 4092, + "GGGGCA": 4093, + "GGGGCT": 4094, + "GGGGCC": 4095, + "GGGGCG": 4096, + "GGGGGA": 4097, + "GGGGGT": 4098, + "GGGGGC": 4099, + "GGGGGG": 4100 + } + } +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/tokenizer_config.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..604e2b1b29af1dcb3920d2828b6a90276e5e3d02 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/tokenizer_config.json @@ -0,0 +1,58 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "[UNK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "[CLS]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "[SEP]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "4": { + "content": "[MASK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "clean_up_tokenization_spaces": true, + "cls_token": "[CLS]", + "do_basic_tokenize": true, + "do_lower_case": false, + "mask_token": "[MASK]", + "max_len": 512, + "model_max_length": 512, + "never_split": null, + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "strip_accents": null, + "tokenize_chinese_chars": true, + "tokenizer_class": "BertTokenizer", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/trainer_state.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..d80b7f92746d6504d7e36e5ebde8896613f21008 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/trainer_state.json @@ -0,0 +1,593 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 8.0, + "eval_steps": 500, + "global_step": 2000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.1, + "grad_norm": 0.2218121588230133, + "learning_rate": 5.319148936170213e-05, + "loss": 0.6927, + "step": 25 + }, + { + "epoch": 0.2, + "grad_norm": 0.44851255416870117, + "learning_rate": 0.00010638297872340425, + "loss": 0.6929, + "step": 50 + }, + { + "epoch": 0.3, + "grad_norm": 0.2440175861120224, + "learning_rate": 0.00015957446808510637, + "loss": 0.6924, + "step": 75 + }, + { + "epoch": 0.4, + "grad_norm": 0.3422006666660309, + "learning_rate": 0.0002127659574468085, + "loss": 0.6922, + "step": 100 + }, + { + "epoch": 0.5, + "grad_norm": 0.267333447933197, + "learning_rate": 0.00026595744680851064, + "loss": 0.6913, + "step": 125 + }, + { + "epoch": 0.6, + "grad_norm": 0.45699772238731384, + "learning_rate": 0.00031914893617021275, + "loss": 0.6896, + "step": 150 + }, + { + "epoch": 0.7, + "grad_norm": 0.3004947304725647, + "learning_rate": 0.0003723404255319149, + "loss": 0.6887, + "step": 175 + }, + { + "epoch": 0.8, + "grad_norm": 0.33440759778022766, + "learning_rate": 0.0003992081821181128, + "loss": 0.6857, + "step": 200 + }, + { + "epoch": 0.9, + "grad_norm": 0.2918928861618042, + "learning_rate": 0.0003975585615308479, + "loss": 0.6882, + "step": 225 + }, + { + "epoch": 1.0, + "grad_norm": 0.2934085726737976, + "learning_rate": 0.00039590894094358297, + "loss": 0.6807, + "step": 250 + }, + { + "epoch": 1.1, + "grad_norm": 0.3364830017089844, + "learning_rate": 0.0003942593203563181, + "loss": 0.6848, + "step": 275 + }, + { + "epoch": 1.2, + "grad_norm": 0.8379159569740295, + "learning_rate": 0.0003926096997690532, + "loss": 0.677, + "step": 300 + }, + { + "epoch": 1.3, + "grad_norm": 0.37671706080436707, + "learning_rate": 0.00039096007918178817, + "loss": 0.6744, + "step": 325 + }, + { + "epoch": 1.4, + "grad_norm": 0.34329330921173096, + "learning_rate": 0.0003893104585945233, + "loss": 0.6733, + "step": 350 + }, + { + "epoch": 1.5, + "grad_norm": 0.4286085367202759, + "learning_rate": 0.0003876608380072583, + "loss": 0.6772, + "step": 375 + }, + { + "epoch": 1.6, + "grad_norm": 0.3576306998729706, + "learning_rate": 0.00038601121741999343, + "loss": 0.6799, + "step": 400 + }, + { + "epoch": 1.7, + "grad_norm": 0.4636435806751251, + "learning_rate": 0.0003843615968327285, + "loss": 0.6819, + "step": 425 + }, + { + "epoch": 1.8, + "grad_norm": 0.4552135765552521, + "learning_rate": 0.0003827119762454636, + "loss": 0.6731, + "step": 450 + }, + { + "epoch": 1.9, + "grad_norm": 0.3565954566001892, + "learning_rate": 0.00038106235565819863, + "loss": 0.6718, + "step": 475 + }, + { + "epoch": 2.0, + "grad_norm": 0.5451686978340149, + "learning_rate": 0.0003794127350709337, + "loss": 0.6858, + "step": 500 + }, + { + "epoch": 2.1, + "grad_norm": 0.3730587959289551, + "learning_rate": 0.0003777631144836688, + "loss": 0.667, + "step": 525 + }, + { + "epoch": 2.2, + "grad_norm": 0.6642417907714844, + "learning_rate": 0.00037611349389640383, + "loss": 0.6662, + "step": 550 + }, + { + "epoch": 2.3, + "grad_norm": 0.3889809250831604, + "learning_rate": 0.00037446387330913894, + "loss": 0.6718, + "step": 575 + }, + { + "epoch": 2.4, + "grad_norm": 0.47203826904296875, + "learning_rate": 0.000372814252721874, + "loss": 0.6687, + "step": 600 + }, + { + "epoch": 2.5, + "grad_norm": 0.5293862819671631, + "learning_rate": 0.00037116463213460903, + "loss": 0.6655, + "step": 625 + }, + { + "epoch": 2.6, + "grad_norm": 0.36724531650543213, + "learning_rate": 0.00036951501154734414, + "loss": 0.672, + "step": 650 + }, + { + "epoch": 2.7, + "grad_norm": 0.5733087062835693, + "learning_rate": 0.0003678653909600792, + "loss": 0.6683, + "step": 675 + }, + { + "epoch": 2.8, + "grad_norm": 0.6385205984115601, + "learning_rate": 0.0003662157703728143, + "loss": 0.67, + "step": 700 + }, + { + "epoch": 2.9, + "grad_norm": 0.6146281957626343, + "learning_rate": 0.00036456614978554934, + "loss": 0.6606, + "step": 725 + }, + { + "epoch": 3.0, + "grad_norm": 0.6535729169845581, + "learning_rate": 0.00036291652919828444, + "loss": 0.6691, + "step": 750 + }, + { + "epoch": 3.1, + "grad_norm": 0.5194367170333862, + "learning_rate": 0.00036126690861101944, + "loss": 0.6517, + "step": 775 + }, + { + "epoch": 3.2, + "grad_norm": 0.7116692066192627, + "learning_rate": 0.00035961728802375454, + "loss": 0.6501, + "step": 800 + }, + { + "epoch": 3.3, + "grad_norm": 0.5338636040687561, + "learning_rate": 0.0003579676674364896, + "loss": 0.6682, + "step": 825 + }, + { + "epoch": 3.4, + "grad_norm": 0.6912877559661865, + "learning_rate": 0.0003563180468492247, + "loss": 0.6632, + "step": 850 + }, + { + "epoch": 3.5, + "grad_norm": 0.6925108432769775, + "learning_rate": 0.0003546684262619598, + "loss": 0.653, + "step": 875 + }, + { + "epoch": 3.6, + "grad_norm": 0.5360299944877625, + "learning_rate": 0.00035301880567469485, + "loss": 0.6613, + "step": 900 + }, + { + "epoch": 3.7, + "grad_norm": 0.6487969160079956, + "learning_rate": 0.0003513691850874299, + "loss": 0.6627, + "step": 925 + }, + { + "epoch": 3.8, + "grad_norm": 1.7260371446609497, + "learning_rate": 0.00034971956450016495, + "loss": 0.6436, + "step": 950 + }, + { + "epoch": 3.9, + "grad_norm": 0.5852354168891907, + "learning_rate": 0.00034806994391290005, + "loss": 0.6613, + "step": 975 + }, + { + "epoch": 4.0, + "grad_norm": 0.7802712917327881, + "learning_rate": 0.0003464203233256351, + "loss": 0.658, + "step": 1000 + }, + { + "epoch": 4.1, + "grad_norm": 0.7335640788078308, + "learning_rate": 0.0003447707027383702, + "loss": 0.6476, + "step": 1025 + }, + { + "epoch": 4.2, + "grad_norm": 0.9540443420410156, + "learning_rate": 0.00034312108215110525, + "loss": 0.6334, + "step": 1050 + }, + { + "epoch": 4.3, + "grad_norm": 1.3698925971984863, + "learning_rate": 0.0003414714615638403, + "loss": 0.6512, + "step": 1075 + }, + { + "epoch": 4.4, + "grad_norm": 0.8282333016395569, + "learning_rate": 0.0003398218409765754, + "loss": 0.6556, + "step": 1100 + }, + { + "epoch": 4.5, + "grad_norm": 1.4414807558059692, + "learning_rate": 0.00033817222038931045, + "loss": 0.6628, + "step": 1125 + }, + { + "epoch": 4.6, + "grad_norm": 0.5493957996368408, + "learning_rate": 0.00033652259980204556, + "loss": 0.6461, + "step": 1150 + }, + { + "epoch": 4.7, + "grad_norm": 1.2343393564224243, + "learning_rate": 0.0003348729792147806, + "loss": 0.6464, + "step": 1175 + }, + { + "epoch": 4.8, + "grad_norm": 0.7163398265838623, + "learning_rate": 0.0003332233586275157, + "loss": 0.6518, + "step": 1200 + }, + { + "epoch": 4.9, + "grad_norm": 0.7878938317298889, + "learning_rate": 0.00033157373804025076, + "loss": 0.6502, + "step": 1225 + }, + { + "epoch": 5.0, + "grad_norm": 0.952571451663971, + "learning_rate": 0.0003299241174529858, + "loss": 0.6463, + "step": 1250 + }, + { + "epoch": 5.1, + "grad_norm": 0.6535393595695496, + "learning_rate": 0.0003282744968657209, + "loss": 0.6361, + "step": 1275 + }, + { + "epoch": 5.2, + "grad_norm": 0.8239464163780212, + "learning_rate": 0.00032662487627845596, + "loss": 0.6264, + "step": 1300 + }, + { + "epoch": 5.3, + "grad_norm": 1.4500653743743896, + "learning_rate": 0.00032497525569119106, + "loss": 0.6544, + "step": 1325 + }, + { + "epoch": 5.4, + "grad_norm": 1.1515296697616577, + "learning_rate": 0.0003233256351039261, + "loss": 0.6629, + "step": 1350 + }, + { + "epoch": 5.5, + "grad_norm": 0.6899295449256897, + "learning_rate": 0.00032167601451666116, + "loss": 0.6453, + "step": 1375 + }, + { + "epoch": 5.6, + "grad_norm": 1.1782217025756836, + "learning_rate": 0.00032002639392939627, + "loss": 0.6362, + "step": 1400 + }, + { + "epoch": 5.7, + "grad_norm": 0.6971372961997986, + "learning_rate": 0.0003183767733421313, + "loss": 0.6593, + "step": 1425 + }, + { + "epoch": 5.8, + "grad_norm": 0.7293921709060669, + "learning_rate": 0.0003167271527548664, + "loss": 0.6318, + "step": 1450 + }, + { + "epoch": 5.9, + "grad_norm": 0.6860108375549316, + "learning_rate": 0.00031507753216760147, + "loss": 0.6402, + "step": 1475 + }, + { + "epoch": 6.0, + "grad_norm": 1.330344319343567, + "learning_rate": 0.00031342791158033657, + "loss": 0.6276, + "step": 1500 + }, + { + "epoch": 6.1, + "grad_norm": 0.7261562943458557, + "learning_rate": 0.00031177829099307157, + "loss": 0.6369, + "step": 1525 + }, + { + "epoch": 6.2, + "grad_norm": 1.1324542760849, + "learning_rate": 0.00031012867040580667, + "loss": 0.6228, + "step": 1550 + }, + { + "epoch": 6.3, + "grad_norm": 0.7002120614051819, + "learning_rate": 0.0003084790498185417, + "loss": 0.6389, + "step": 1575 + }, + { + "epoch": 6.4, + "grad_norm": 1.3143397569656372, + "learning_rate": 0.0003068294292312768, + "loss": 0.6362, + "step": 1600 + }, + { + "epoch": 6.5, + "grad_norm": 1.6375401020050049, + "learning_rate": 0.0003051798086440119, + "loss": 0.6348, + "step": 1625 + }, + { + "epoch": 6.6, + "grad_norm": 1.4343605041503906, + "learning_rate": 0.000303530188056747, + "loss": 0.6496, + "step": 1650 + }, + { + "epoch": 6.7, + "grad_norm": 0.815762996673584, + "learning_rate": 0.000301880567469482, + "loss": 0.6407, + "step": 1675 + }, + { + "epoch": 6.8, + "grad_norm": 1.068608045578003, + "learning_rate": 0.0003002309468822171, + "loss": 0.6417, + "step": 1700 + }, + { + "epoch": 6.9, + "grad_norm": 0.8775753974914551, + "learning_rate": 0.0002985813262949522, + "loss": 0.6281, + "step": 1725 + }, + { + "epoch": 7.0, + "grad_norm": 1.041817545890808, + "learning_rate": 0.00029693170570768723, + "loss": 0.6485, + "step": 1750 + }, + { + "epoch": 7.1, + "grad_norm": 0.7135840058326721, + "learning_rate": 0.00029528208512042233, + "loss": 0.6192, + "step": 1775 + }, + { + "epoch": 7.2, + "grad_norm": 1.5523930788040161, + "learning_rate": 0.0002936324645331574, + "loss": 0.636, + "step": 1800 + }, + { + "epoch": 7.3, + "grad_norm": 0.7677174210548401, + "learning_rate": 0.00029198284394589243, + "loss": 0.6213, + "step": 1825 + }, + { + "epoch": 7.4, + "grad_norm": 0.8086565732955933, + "learning_rate": 0.00029033322335862753, + "loss": 0.6292, + "step": 1850 + }, + { + "epoch": 7.5, + "grad_norm": 3.1128971576690674, + "learning_rate": 0.0002886836027713626, + "loss": 0.648, + "step": 1875 + }, + { + "epoch": 7.6, + "grad_norm": 0.9243882894515991, + "learning_rate": 0.0002870339821840977, + "loss": 0.6286, + "step": 1900 + }, + { + "epoch": 7.7, + "grad_norm": 1.8203461170196533, + "learning_rate": 0.00028538436159683273, + "loss": 0.6301, + "step": 1925 + }, + { + "epoch": 7.8, + "grad_norm": 1.0924643278121948, + "learning_rate": 0.00028373474100956784, + "loss": 0.6378, + "step": 1950 + }, + { + "epoch": 7.9, + "grad_norm": 1.0839996337890625, + "learning_rate": 0.0002820851204223029, + "loss": 0.6265, + "step": 1975 + }, + { + "epoch": 8.0, + "grad_norm": 0.7143226861953735, + "learning_rate": 0.00028043549983503794, + "loss": 0.649, + "step": 2000 + } + ], + "logging_steps": 25, + "max_steps": 6250, + "num_input_tokens_seen": 0, + "num_train_epochs": 25, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 110762065920000.0, + "train_batch_size": 20, + "trial_name": null, + "trial_params": null +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/training_args.bin b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..540db3fbee786996d9705d768d31ba6c371187be --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dc7816b5f3fd38f01f9ba38a197ce070cde06e4c90e5260cb4814f34f628907d +size 5304 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/vocab.txt b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/vocab.txt new file mode 100644 index 0000000000000000000000000000000000000000..cc9ad53732e21ea9180f18e1fb67a38cc5702d4c --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-2000/vocab.txt @@ -0,0 +1,4101 @@ +[PAD] +[UNK] +[CLS] +[SEP] +[MASK] +AAAAAA +AAAAAT +AAAAAC +AAAAAG +AAAATA +AAAATT +AAAATC +AAAATG +AAAACA +AAAACT +AAAACC +AAAACG +AAAAGA +AAAAGT +AAAAGC +AAAAGG +AAATAA +AAATAT +AAATAC +AAATAG +AAATTA +AAATTT +AAATTC +AAATTG +AAATCA +AAATCT +AAATCC +AAATCG +AAATGA +AAATGT +AAATGC +AAATGG +AAACAA +AAACAT +AAACAC +AAACAG +AAACTA +AAACTT +AAACTC +AAACTG +AAACCA +AAACCT +AAACCC +AAACCG +AAACGA +AAACGT +AAACGC +AAACGG +AAAGAA +AAAGAT +AAAGAC +AAAGAG +AAAGTA +AAAGTT +AAAGTC +AAAGTG +AAAGCA +AAAGCT +AAAGCC +AAAGCG +AAAGGA +AAAGGT +AAAGGC +AAAGGG +AATAAA +AATAAT +AATAAC +AATAAG +AATATA +AATATT +AATATC +AATATG +AATACA +AATACT +AATACC +AATACG +AATAGA +AATAGT +AATAGC +AATAGG +AATTAA +AATTAT +AATTAC +AATTAG +AATTTA +AATTTT +AATTTC +AATTTG +AATTCA +AATTCT +AATTCC +AATTCG +AATTGA +AATTGT +AATTGC +AATTGG +AATCAA +AATCAT +AATCAC +AATCAG +AATCTA +AATCTT +AATCTC +AATCTG +AATCCA +AATCCT +AATCCC +AATCCG +AATCGA +AATCGT +AATCGC +AATCGG +AATGAA +AATGAT +AATGAC +AATGAG +AATGTA +AATGTT +AATGTC +AATGTG +AATGCA +AATGCT +AATGCC +AATGCG +AATGGA +AATGGT +AATGGC +AATGGG +AACAAA +AACAAT +AACAAC +AACAAG +AACATA +AACATT +AACATC +AACATG +AACACA +AACACT +AACACC +AACACG +AACAGA +AACAGT +AACAGC +AACAGG +AACTAA +AACTAT +AACTAC +AACTAG +AACTTA +AACTTT +AACTTC +AACTTG +AACTCA +AACTCT +AACTCC +AACTCG +AACTGA +AACTGT +AACTGC +AACTGG +AACCAA +AACCAT +AACCAC +AACCAG +AACCTA +AACCTT +AACCTC +AACCTG +AACCCA +AACCCT +AACCCC +AACCCG +AACCGA +AACCGT +AACCGC +AACCGG +AACGAA +AACGAT +AACGAC +AACGAG +AACGTA +AACGTT +AACGTC +AACGTG +AACGCA +AACGCT +AACGCC +AACGCG +AACGGA +AACGGT +AACGGC +AACGGG +AAGAAA +AAGAAT +AAGAAC +AAGAAG +AAGATA +AAGATT +AAGATC +AAGATG +AAGACA +AAGACT +AAGACC +AAGACG +AAGAGA +AAGAGT +AAGAGC +AAGAGG +AAGTAA +AAGTAT +AAGTAC +AAGTAG +AAGTTA +AAGTTT +AAGTTC +AAGTTG +AAGTCA +AAGTCT +AAGTCC +AAGTCG +AAGTGA +AAGTGT +AAGTGC +AAGTGG +AAGCAA +AAGCAT +AAGCAC +AAGCAG +AAGCTA +AAGCTT +AAGCTC +AAGCTG +AAGCCA +AAGCCT +AAGCCC +AAGCCG +AAGCGA +AAGCGT +AAGCGC +AAGCGG +AAGGAA +AAGGAT +AAGGAC +AAGGAG +AAGGTA +AAGGTT +AAGGTC +AAGGTG +AAGGCA +AAGGCT +AAGGCC +AAGGCG +AAGGGA +AAGGGT +AAGGGC +AAGGGG +ATAAAA +ATAAAT +ATAAAC +ATAAAG +ATAATA +ATAATT +ATAATC +ATAATG +ATAACA +ATAACT +ATAACC +ATAACG +ATAAGA +ATAAGT +ATAAGC +ATAAGG +ATATAA +ATATAT +ATATAC +ATATAG +ATATTA +ATATTT +ATATTC +ATATTG +ATATCA +ATATCT +ATATCC +ATATCG +ATATGA +ATATGT +ATATGC +ATATGG +ATACAA +ATACAT +ATACAC +ATACAG +ATACTA +ATACTT +ATACTC +ATACTG +ATACCA +ATACCT +ATACCC +ATACCG +ATACGA +ATACGT +ATACGC +ATACGG +ATAGAA +ATAGAT +ATAGAC +ATAGAG +ATAGTA +ATAGTT +ATAGTC +ATAGTG +ATAGCA +ATAGCT +ATAGCC +ATAGCG +ATAGGA +ATAGGT +ATAGGC +ATAGGG +ATTAAA +ATTAAT +ATTAAC +ATTAAG +ATTATA +ATTATT +ATTATC +ATTATG +ATTACA +ATTACT +ATTACC +ATTACG +ATTAGA +ATTAGT +ATTAGC +ATTAGG +ATTTAA +ATTTAT +ATTTAC +ATTTAG +ATTTTA +ATTTTT +ATTTTC +ATTTTG +ATTTCA +ATTTCT +ATTTCC +ATTTCG +ATTTGA +ATTTGT +ATTTGC +ATTTGG +ATTCAA +ATTCAT +ATTCAC +ATTCAG +ATTCTA +ATTCTT +ATTCTC +ATTCTG +ATTCCA +ATTCCT +ATTCCC +ATTCCG +ATTCGA +ATTCGT +ATTCGC +ATTCGG +ATTGAA +ATTGAT +ATTGAC +ATTGAG +ATTGTA +ATTGTT +ATTGTC +ATTGTG +ATTGCA +ATTGCT +ATTGCC +ATTGCG +ATTGGA +ATTGGT +ATTGGC +ATTGGG +ATCAAA +ATCAAT +ATCAAC +ATCAAG +ATCATA +ATCATT +ATCATC +ATCATG +ATCACA +ATCACT +ATCACC +ATCACG +ATCAGA +ATCAGT +ATCAGC +ATCAGG +ATCTAA +ATCTAT +ATCTAC +ATCTAG +ATCTTA +ATCTTT +ATCTTC +ATCTTG +ATCTCA +ATCTCT +ATCTCC +ATCTCG +ATCTGA +ATCTGT +ATCTGC +ATCTGG +ATCCAA +ATCCAT +ATCCAC +ATCCAG +ATCCTA +ATCCTT +ATCCTC +ATCCTG +ATCCCA +ATCCCT +ATCCCC +ATCCCG +ATCCGA +ATCCGT +ATCCGC +ATCCGG +ATCGAA +ATCGAT +ATCGAC +ATCGAG +ATCGTA +ATCGTT +ATCGTC +ATCGTG +ATCGCA +ATCGCT +ATCGCC +ATCGCG +ATCGGA +ATCGGT +ATCGGC +ATCGGG +ATGAAA +ATGAAT +ATGAAC +ATGAAG +ATGATA +ATGATT +ATGATC +ATGATG +ATGACA +ATGACT +ATGACC +ATGACG +ATGAGA +ATGAGT +ATGAGC +ATGAGG +ATGTAA +ATGTAT +ATGTAC +ATGTAG +ATGTTA +ATGTTT +ATGTTC +ATGTTG +ATGTCA +ATGTCT +ATGTCC +ATGTCG +ATGTGA +ATGTGT +ATGTGC +ATGTGG +ATGCAA +ATGCAT +ATGCAC +ATGCAG +ATGCTA +ATGCTT +ATGCTC +ATGCTG +ATGCCA +ATGCCT +ATGCCC +ATGCCG +ATGCGA +ATGCGT +ATGCGC +ATGCGG +ATGGAA +ATGGAT +ATGGAC +ATGGAG +ATGGTA +ATGGTT +ATGGTC +ATGGTG +ATGGCA +ATGGCT +ATGGCC +ATGGCG +ATGGGA +ATGGGT +ATGGGC +ATGGGG +ACAAAA +ACAAAT +ACAAAC +ACAAAG +ACAATA +ACAATT +ACAATC +ACAATG +ACAACA +ACAACT +ACAACC +ACAACG +ACAAGA +ACAAGT +ACAAGC +ACAAGG +ACATAA +ACATAT +ACATAC +ACATAG +ACATTA +ACATTT +ACATTC +ACATTG +ACATCA +ACATCT +ACATCC +ACATCG +ACATGA +ACATGT +ACATGC +ACATGG +ACACAA +ACACAT +ACACAC +ACACAG +ACACTA +ACACTT +ACACTC +ACACTG +ACACCA +ACACCT +ACACCC +ACACCG +ACACGA +ACACGT +ACACGC +ACACGG +ACAGAA +ACAGAT +ACAGAC +ACAGAG +ACAGTA +ACAGTT +ACAGTC +ACAGTG +ACAGCA +ACAGCT +ACAGCC +ACAGCG +ACAGGA +ACAGGT +ACAGGC +ACAGGG +ACTAAA +ACTAAT +ACTAAC +ACTAAG +ACTATA +ACTATT +ACTATC +ACTATG +ACTACA +ACTACT +ACTACC +ACTACG +ACTAGA +ACTAGT +ACTAGC +ACTAGG +ACTTAA +ACTTAT +ACTTAC +ACTTAG +ACTTTA +ACTTTT +ACTTTC +ACTTTG +ACTTCA +ACTTCT +ACTTCC +ACTTCG +ACTTGA +ACTTGT +ACTTGC +ACTTGG +ACTCAA +ACTCAT +ACTCAC +ACTCAG +ACTCTA +ACTCTT +ACTCTC +ACTCTG +ACTCCA +ACTCCT +ACTCCC +ACTCCG +ACTCGA +ACTCGT +ACTCGC +ACTCGG +ACTGAA +ACTGAT +ACTGAC +ACTGAG +ACTGTA +ACTGTT +ACTGTC +ACTGTG +ACTGCA +ACTGCT +ACTGCC +ACTGCG +ACTGGA +ACTGGT +ACTGGC +ACTGGG +ACCAAA +ACCAAT +ACCAAC +ACCAAG +ACCATA +ACCATT +ACCATC +ACCATG +ACCACA +ACCACT +ACCACC +ACCACG +ACCAGA +ACCAGT +ACCAGC +ACCAGG +ACCTAA +ACCTAT +ACCTAC +ACCTAG +ACCTTA +ACCTTT +ACCTTC +ACCTTG +ACCTCA +ACCTCT +ACCTCC +ACCTCG +ACCTGA +ACCTGT +ACCTGC +ACCTGG +ACCCAA +ACCCAT +ACCCAC +ACCCAG +ACCCTA +ACCCTT +ACCCTC +ACCCTG +ACCCCA +ACCCCT +ACCCCC +ACCCCG +ACCCGA +ACCCGT +ACCCGC +ACCCGG +ACCGAA +ACCGAT +ACCGAC +ACCGAG +ACCGTA +ACCGTT +ACCGTC +ACCGTG +ACCGCA +ACCGCT +ACCGCC +ACCGCG +ACCGGA +ACCGGT +ACCGGC +ACCGGG +ACGAAA +ACGAAT +ACGAAC +ACGAAG +ACGATA +ACGATT +ACGATC +ACGATG +ACGACA +ACGACT +ACGACC +ACGACG +ACGAGA +ACGAGT +ACGAGC +ACGAGG +ACGTAA +ACGTAT +ACGTAC +ACGTAG +ACGTTA +ACGTTT +ACGTTC +ACGTTG +ACGTCA +ACGTCT +ACGTCC +ACGTCG +ACGTGA +ACGTGT +ACGTGC +ACGTGG +ACGCAA +ACGCAT +ACGCAC +ACGCAG +ACGCTA +ACGCTT +ACGCTC +ACGCTG +ACGCCA +ACGCCT +ACGCCC +ACGCCG +ACGCGA +ACGCGT +ACGCGC +ACGCGG +ACGGAA +ACGGAT +ACGGAC +ACGGAG +ACGGTA +ACGGTT +ACGGTC +ACGGTG +ACGGCA +ACGGCT +ACGGCC +ACGGCG +ACGGGA +ACGGGT +ACGGGC +ACGGGG +AGAAAA +AGAAAT +AGAAAC +AGAAAG +AGAATA +AGAATT +AGAATC +AGAATG +AGAACA +AGAACT +AGAACC +AGAACG +AGAAGA +AGAAGT +AGAAGC +AGAAGG +AGATAA +AGATAT +AGATAC +AGATAG +AGATTA +AGATTT +AGATTC +AGATTG +AGATCA +AGATCT +AGATCC +AGATCG +AGATGA +AGATGT +AGATGC +AGATGG +AGACAA +AGACAT +AGACAC +AGACAG +AGACTA +AGACTT +AGACTC +AGACTG +AGACCA +AGACCT +AGACCC +AGACCG +AGACGA +AGACGT +AGACGC +AGACGG +AGAGAA +AGAGAT +AGAGAC +AGAGAG +AGAGTA +AGAGTT +AGAGTC +AGAGTG +AGAGCA +AGAGCT +AGAGCC +AGAGCG +AGAGGA +AGAGGT +AGAGGC +AGAGGG +AGTAAA +AGTAAT +AGTAAC +AGTAAG +AGTATA +AGTATT +AGTATC +AGTATG +AGTACA +AGTACT +AGTACC +AGTACG +AGTAGA +AGTAGT +AGTAGC +AGTAGG +AGTTAA +AGTTAT +AGTTAC +AGTTAG +AGTTTA +AGTTTT +AGTTTC +AGTTTG +AGTTCA +AGTTCT +AGTTCC +AGTTCG +AGTTGA +AGTTGT +AGTTGC +AGTTGG +AGTCAA +AGTCAT +AGTCAC +AGTCAG +AGTCTA +AGTCTT +AGTCTC +AGTCTG +AGTCCA +AGTCCT +AGTCCC +AGTCCG +AGTCGA +AGTCGT +AGTCGC +AGTCGG +AGTGAA +AGTGAT +AGTGAC +AGTGAG +AGTGTA +AGTGTT +AGTGTC +AGTGTG +AGTGCA +AGTGCT +AGTGCC +AGTGCG +AGTGGA +AGTGGT +AGTGGC +AGTGGG +AGCAAA +AGCAAT +AGCAAC +AGCAAG +AGCATA +AGCATT +AGCATC +AGCATG +AGCACA +AGCACT +AGCACC +AGCACG +AGCAGA +AGCAGT +AGCAGC +AGCAGG +AGCTAA +AGCTAT +AGCTAC +AGCTAG +AGCTTA +AGCTTT +AGCTTC +AGCTTG +AGCTCA +AGCTCT +AGCTCC +AGCTCG +AGCTGA +AGCTGT +AGCTGC +AGCTGG +AGCCAA +AGCCAT +AGCCAC +AGCCAG +AGCCTA +AGCCTT +AGCCTC +AGCCTG +AGCCCA +AGCCCT +AGCCCC +AGCCCG +AGCCGA +AGCCGT +AGCCGC +AGCCGG +AGCGAA +AGCGAT +AGCGAC +AGCGAG +AGCGTA +AGCGTT +AGCGTC +AGCGTG +AGCGCA +AGCGCT +AGCGCC +AGCGCG +AGCGGA +AGCGGT +AGCGGC +AGCGGG +AGGAAA +AGGAAT +AGGAAC +AGGAAG +AGGATA +AGGATT +AGGATC +AGGATG +AGGACA +AGGACT +AGGACC +AGGACG +AGGAGA +AGGAGT +AGGAGC +AGGAGG +AGGTAA +AGGTAT +AGGTAC +AGGTAG +AGGTTA +AGGTTT +AGGTTC +AGGTTG +AGGTCA +AGGTCT +AGGTCC +AGGTCG +AGGTGA +AGGTGT +AGGTGC +AGGTGG +AGGCAA +AGGCAT +AGGCAC +AGGCAG +AGGCTA +AGGCTT +AGGCTC +AGGCTG +AGGCCA +AGGCCT +AGGCCC +AGGCCG +AGGCGA +AGGCGT +AGGCGC +AGGCGG +AGGGAA +AGGGAT +AGGGAC +AGGGAG +AGGGTA +AGGGTT +AGGGTC +AGGGTG +AGGGCA +AGGGCT +AGGGCC +AGGGCG +AGGGGA +AGGGGT +AGGGGC +AGGGGG +TAAAAA +TAAAAT +TAAAAC +TAAAAG +TAAATA +TAAATT +TAAATC +TAAATG +TAAACA +TAAACT +TAAACC +TAAACG +TAAAGA +TAAAGT +TAAAGC +TAAAGG +TAATAA +TAATAT +TAATAC +TAATAG +TAATTA +TAATTT +TAATTC +TAATTG +TAATCA +TAATCT +TAATCC +TAATCG +TAATGA +TAATGT +TAATGC +TAATGG +TAACAA +TAACAT +TAACAC +TAACAG +TAACTA +TAACTT +TAACTC +TAACTG +TAACCA +TAACCT +TAACCC +TAACCG +TAACGA +TAACGT +TAACGC +TAACGG +TAAGAA +TAAGAT +TAAGAC +TAAGAG +TAAGTA +TAAGTT +TAAGTC +TAAGTG +TAAGCA +TAAGCT +TAAGCC +TAAGCG +TAAGGA +TAAGGT +TAAGGC +TAAGGG +TATAAA +TATAAT +TATAAC +TATAAG +TATATA +TATATT +TATATC +TATATG +TATACA +TATACT +TATACC +TATACG +TATAGA +TATAGT +TATAGC +TATAGG +TATTAA +TATTAT +TATTAC +TATTAG +TATTTA +TATTTT +TATTTC +TATTTG +TATTCA +TATTCT +TATTCC +TATTCG +TATTGA +TATTGT +TATTGC +TATTGG +TATCAA +TATCAT +TATCAC +TATCAG +TATCTA +TATCTT +TATCTC +TATCTG +TATCCA +TATCCT +TATCCC +TATCCG +TATCGA +TATCGT +TATCGC +TATCGG +TATGAA +TATGAT +TATGAC +TATGAG +TATGTA +TATGTT +TATGTC +TATGTG +TATGCA +TATGCT +TATGCC +TATGCG +TATGGA +TATGGT +TATGGC +TATGGG +TACAAA +TACAAT +TACAAC +TACAAG +TACATA +TACATT +TACATC +TACATG +TACACA +TACACT +TACACC +TACACG +TACAGA +TACAGT +TACAGC +TACAGG +TACTAA +TACTAT +TACTAC +TACTAG +TACTTA +TACTTT +TACTTC +TACTTG +TACTCA +TACTCT +TACTCC +TACTCG +TACTGA +TACTGT +TACTGC +TACTGG +TACCAA +TACCAT +TACCAC +TACCAG +TACCTA +TACCTT +TACCTC +TACCTG +TACCCA +TACCCT +TACCCC +TACCCG +TACCGA +TACCGT +TACCGC +TACCGG +TACGAA +TACGAT +TACGAC +TACGAG +TACGTA +TACGTT +TACGTC +TACGTG +TACGCA +TACGCT +TACGCC +TACGCG +TACGGA +TACGGT +TACGGC +TACGGG +TAGAAA +TAGAAT +TAGAAC +TAGAAG +TAGATA +TAGATT +TAGATC +TAGATG +TAGACA +TAGACT +TAGACC +TAGACG +TAGAGA +TAGAGT +TAGAGC +TAGAGG +TAGTAA +TAGTAT +TAGTAC +TAGTAG +TAGTTA +TAGTTT +TAGTTC +TAGTTG +TAGTCA +TAGTCT +TAGTCC +TAGTCG +TAGTGA +TAGTGT +TAGTGC +TAGTGG +TAGCAA +TAGCAT +TAGCAC +TAGCAG +TAGCTA +TAGCTT +TAGCTC +TAGCTG +TAGCCA +TAGCCT +TAGCCC +TAGCCG +TAGCGA +TAGCGT +TAGCGC +TAGCGG +TAGGAA +TAGGAT +TAGGAC +TAGGAG +TAGGTA +TAGGTT +TAGGTC +TAGGTG +TAGGCA +TAGGCT +TAGGCC +TAGGCG +TAGGGA +TAGGGT +TAGGGC +TAGGGG +TTAAAA +TTAAAT +TTAAAC +TTAAAG +TTAATA +TTAATT +TTAATC +TTAATG +TTAACA +TTAACT +TTAACC +TTAACG +TTAAGA +TTAAGT +TTAAGC +TTAAGG +TTATAA +TTATAT +TTATAC +TTATAG +TTATTA +TTATTT +TTATTC +TTATTG +TTATCA +TTATCT +TTATCC +TTATCG +TTATGA +TTATGT +TTATGC +TTATGG +TTACAA +TTACAT +TTACAC +TTACAG +TTACTA +TTACTT +TTACTC +TTACTG +TTACCA +TTACCT +TTACCC +TTACCG +TTACGA +TTACGT +TTACGC +TTACGG +TTAGAA +TTAGAT +TTAGAC +TTAGAG +TTAGTA +TTAGTT +TTAGTC +TTAGTG +TTAGCA +TTAGCT +TTAGCC +TTAGCG +TTAGGA +TTAGGT +TTAGGC +TTAGGG +TTTAAA +TTTAAT +TTTAAC +TTTAAG +TTTATA +TTTATT +TTTATC +TTTATG +TTTACA +TTTACT +TTTACC +TTTACG +TTTAGA +TTTAGT +TTTAGC +TTTAGG +TTTTAA +TTTTAT +TTTTAC +TTTTAG +TTTTTA +TTTTTT +TTTTTC +TTTTTG +TTTTCA +TTTTCT +TTTTCC +TTTTCG +TTTTGA +TTTTGT +TTTTGC +TTTTGG +TTTCAA +TTTCAT +TTTCAC +TTTCAG +TTTCTA +TTTCTT +TTTCTC +TTTCTG +TTTCCA +TTTCCT +TTTCCC +TTTCCG +TTTCGA +TTTCGT +TTTCGC +TTTCGG +TTTGAA +TTTGAT +TTTGAC +TTTGAG +TTTGTA +TTTGTT +TTTGTC +TTTGTG +TTTGCA +TTTGCT +TTTGCC +TTTGCG +TTTGGA +TTTGGT +TTTGGC +TTTGGG +TTCAAA +TTCAAT +TTCAAC +TTCAAG +TTCATA +TTCATT +TTCATC +TTCATG +TTCACA +TTCACT +TTCACC +TTCACG +TTCAGA +TTCAGT +TTCAGC +TTCAGG +TTCTAA +TTCTAT +TTCTAC +TTCTAG +TTCTTA +TTCTTT +TTCTTC +TTCTTG +TTCTCA +TTCTCT +TTCTCC +TTCTCG +TTCTGA +TTCTGT +TTCTGC +TTCTGG +TTCCAA +TTCCAT +TTCCAC +TTCCAG +TTCCTA +TTCCTT +TTCCTC +TTCCTG +TTCCCA +TTCCCT +TTCCCC +TTCCCG +TTCCGA +TTCCGT +TTCCGC +TTCCGG +TTCGAA +TTCGAT +TTCGAC +TTCGAG +TTCGTA +TTCGTT +TTCGTC +TTCGTG +TTCGCA +TTCGCT +TTCGCC +TTCGCG +TTCGGA +TTCGGT +TTCGGC +TTCGGG +TTGAAA +TTGAAT +TTGAAC +TTGAAG +TTGATA +TTGATT +TTGATC +TTGATG +TTGACA +TTGACT +TTGACC +TTGACG +TTGAGA +TTGAGT +TTGAGC +TTGAGG +TTGTAA +TTGTAT +TTGTAC +TTGTAG +TTGTTA +TTGTTT +TTGTTC +TTGTTG +TTGTCA +TTGTCT +TTGTCC +TTGTCG +TTGTGA +TTGTGT +TTGTGC +TTGTGG +TTGCAA +TTGCAT +TTGCAC +TTGCAG +TTGCTA +TTGCTT +TTGCTC +TTGCTG +TTGCCA +TTGCCT +TTGCCC +TTGCCG +TTGCGA +TTGCGT +TTGCGC +TTGCGG +TTGGAA +TTGGAT +TTGGAC +TTGGAG +TTGGTA +TTGGTT +TTGGTC +TTGGTG +TTGGCA +TTGGCT +TTGGCC +TTGGCG +TTGGGA +TTGGGT +TTGGGC +TTGGGG +TCAAAA +TCAAAT +TCAAAC +TCAAAG +TCAATA +TCAATT +TCAATC +TCAATG +TCAACA +TCAACT +TCAACC +TCAACG +TCAAGA +TCAAGT +TCAAGC +TCAAGG +TCATAA +TCATAT +TCATAC +TCATAG +TCATTA +TCATTT +TCATTC +TCATTG +TCATCA +TCATCT +TCATCC +TCATCG +TCATGA +TCATGT +TCATGC +TCATGG +TCACAA +TCACAT +TCACAC +TCACAG +TCACTA +TCACTT +TCACTC +TCACTG +TCACCA +TCACCT +TCACCC +TCACCG +TCACGA +TCACGT +TCACGC +TCACGG +TCAGAA +TCAGAT +TCAGAC +TCAGAG +TCAGTA +TCAGTT +TCAGTC +TCAGTG +TCAGCA +TCAGCT +TCAGCC +TCAGCG +TCAGGA +TCAGGT +TCAGGC +TCAGGG +TCTAAA +TCTAAT +TCTAAC +TCTAAG +TCTATA +TCTATT +TCTATC +TCTATG +TCTACA +TCTACT +TCTACC +TCTACG +TCTAGA +TCTAGT +TCTAGC +TCTAGG +TCTTAA +TCTTAT +TCTTAC +TCTTAG +TCTTTA +TCTTTT +TCTTTC +TCTTTG +TCTTCA +TCTTCT +TCTTCC +TCTTCG +TCTTGA +TCTTGT +TCTTGC +TCTTGG +TCTCAA +TCTCAT +TCTCAC +TCTCAG +TCTCTA +TCTCTT +TCTCTC +TCTCTG +TCTCCA +TCTCCT +TCTCCC +TCTCCG +TCTCGA +TCTCGT +TCTCGC +TCTCGG +TCTGAA +TCTGAT +TCTGAC +TCTGAG +TCTGTA +TCTGTT +TCTGTC +TCTGTG +TCTGCA +TCTGCT +TCTGCC +TCTGCG +TCTGGA +TCTGGT +TCTGGC +TCTGGG +TCCAAA +TCCAAT +TCCAAC +TCCAAG +TCCATA +TCCATT +TCCATC +TCCATG +TCCACA +TCCACT +TCCACC +TCCACG +TCCAGA +TCCAGT +TCCAGC +TCCAGG +TCCTAA +TCCTAT +TCCTAC +TCCTAG +TCCTTA +TCCTTT +TCCTTC +TCCTTG +TCCTCA +TCCTCT +TCCTCC +TCCTCG +TCCTGA +TCCTGT +TCCTGC +TCCTGG +TCCCAA +TCCCAT +TCCCAC +TCCCAG +TCCCTA +TCCCTT +TCCCTC +TCCCTG +TCCCCA +TCCCCT +TCCCCC +TCCCCG +TCCCGA +TCCCGT +TCCCGC +TCCCGG +TCCGAA +TCCGAT +TCCGAC +TCCGAG +TCCGTA +TCCGTT +TCCGTC +TCCGTG +TCCGCA +TCCGCT +TCCGCC +TCCGCG +TCCGGA +TCCGGT +TCCGGC +TCCGGG +TCGAAA +TCGAAT +TCGAAC +TCGAAG +TCGATA +TCGATT +TCGATC +TCGATG +TCGACA +TCGACT +TCGACC +TCGACG +TCGAGA +TCGAGT +TCGAGC +TCGAGG +TCGTAA +TCGTAT +TCGTAC +TCGTAG +TCGTTA +TCGTTT +TCGTTC +TCGTTG +TCGTCA +TCGTCT +TCGTCC +TCGTCG +TCGTGA +TCGTGT +TCGTGC +TCGTGG +TCGCAA +TCGCAT +TCGCAC +TCGCAG +TCGCTA +TCGCTT +TCGCTC +TCGCTG +TCGCCA +TCGCCT +TCGCCC +TCGCCG +TCGCGA +TCGCGT +TCGCGC +TCGCGG +TCGGAA +TCGGAT +TCGGAC +TCGGAG +TCGGTA +TCGGTT +TCGGTC +TCGGTG +TCGGCA +TCGGCT +TCGGCC +TCGGCG +TCGGGA +TCGGGT +TCGGGC +TCGGGG +TGAAAA +TGAAAT +TGAAAC +TGAAAG +TGAATA +TGAATT +TGAATC +TGAATG +TGAACA +TGAACT +TGAACC +TGAACG +TGAAGA +TGAAGT +TGAAGC +TGAAGG +TGATAA +TGATAT +TGATAC +TGATAG +TGATTA +TGATTT +TGATTC +TGATTG +TGATCA +TGATCT +TGATCC +TGATCG +TGATGA +TGATGT +TGATGC +TGATGG +TGACAA +TGACAT +TGACAC +TGACAG +TGACTA +TGACTT +TGACTC +TGACTG +TGACCA +TGACCT +TGACCC +TGACCG +TGACGA +TGACGT +TGACGC +TGACGG +TGAGAA +TGAGAT +TGAGAC +TGAGAG +TGAGTA +TGAGTT +TGAGTC +TGAGTG +TGAGCA +TGAGCT +TGAGCC +TGAGCG +TGAGGA +TGAGGT +TGAGGC +TGAGGG +TGTAAA +TGTAAT +TGTAAC +TGTAAG +TGTATA +TGTATT +TGTATC +TGTATG +TGTACA +TGTACT +TGTACC +TGTACG +TGTAGA +TGTAGT +TGTAGC +TGTAGG +TGTTAA +TGTTAT +TGTTAC +TGTTAG +TGTTTA +TGTTTT +TGTTTC +TGTTTG +TGTTCA +TGTTCT +TGTTCC +TGTTCG +TGTTGA +TGTTGT +TGTTGC +TGTTGG +TGTCAA +TGTCAT +TGTCAC +TGTCAG +TGTCTA +TGTCTT +TGTCTC +TGTCTG +TGTCCA +TGTCCT +TGTCCC +TGTCCG +TGTCGA +TGTCGT +TGTCGC +TGTCGG +TGTGAA +TGTGAT +TGTGAC +TGTGAG +TGTGTA +TGTGTT +TGTGTC +TGTGTG +TGTGCA +TGTGCT +TGTGCC +TGTGCG +TGTGGA +TGTGGT +TGTGGC +TGTGGG +TGCAAA +TGCAAT +TGCAAC +TGCAAG +TGCATA +TGCATT +TGCATC +TGCATG +TGCACA +TGCACT +TGCACC +TGCACG +TGCAGA +TGCAGT +TGCAGC +TGCAGG +TGCTAA +TGCTAT +TGCTAC +TGCTAG +TGCTTA +TGCTTT +TGCTTC +TGCTTG +TGCTCA +TGCTCT +TGCTCC +TGCTCG +TGCTGA +TGCTGT +TGCTGC +TGCTGG +TGCCAA +TGCCAT +TGCCAC +TGCCAG +TGCCTA +TGCCTT +TGCCTC +TGCCTG +TGCCCA +TGCCCT +TGCCCC +TGCCCG +TGCCGA +TGCCGT +TGCCGC +TGCCGG +TGCGAA +TGCGAT +TGCGAC +TGCGAG +TGCGTA +TGCGTT +TGCGTC +TGCGTG +TGCGCA +TGCGCT +TGCGCC +TGCGCG +TGCGGA +TGCGGT +TGCGGC +TGCGGG +TGGAAA +TGGAAT +TGGAAC +TGGAAG +TGGATA +TGGATT +TGGATC +TGGATG +TGGACA +TGGACT +TGGACC +TGGACG +TGGAGA +TGGAGT +TGGAGC +TGGAGG +TGGTAA +TGGTAT +TGGTAC +TGGTAG +TGGTTA +TGGTTT +TGGTTC +TGGTTG +TGGTCA +TGGTCT +TGGTCC +TGGTCG +TGGTGA +TGGTGT +TGGTGC +TGGTGG +TGGCAA +TGGCAT +TGGCAC +TGGCAG +TGGCTA +TGGCTT +TGGCTC +TGGCTG +TGGCCA +TGGCCT +TGGCCC +TGGCCG +TGGCGA +TGGCGT +TGGCGC +TGGCGG +TGGGAA +TGGGAT +TGGGAC +TGGGAG +TGGGTA +TGGGTT +TGGGTC +TGGGTG +TGGGCA +TGGGCT +TGGGCC +TGGGCG +TGGGGA +TGGGGT +TGGGGC +TGGGGG +CAAAAA +CAAAAT +CAAAAC +CAAAAG +CAAATA +CAAATT +CAAATC +CAAATG +CAAACA +CAAACT +CAAACC +CAAACG +CAAAGA +CAAAGT +CAAAGC +CAAAGG +CAATAA +CAATAT +CAATAC +CAATAG +CAATTA +CAATTT +CAATTC +CAATTG +CAATCA +CAATCT +CAATCC +CAATCG +CAATGA +CAATGT +CAATGC +CAATGG +CAACAA +CAACAT +CAACAC +CAACAG +CAACTA +CAACTT +CAACTC +CAACTG +CAACCA +CAACCT +CAACCC +CAACCG +CAACGA +CAACGT +CAACGC +CAACGG +CAAGAA +CAAGAT +CAAGAC +CAAGAG +CAAGTA +CAAGTT +CAAGTC +CAAGTG +CAAGCA +CAAGCT +CAAGCC +CAAGCG +CAAGGA +CAAGGT +CAAGGC +CAAGGG +CATAAA +CATAAT +CATAAC +CATAAG +CATATA +CATATT +CATATC +CATATG +CATACA +CATACT +CATACC +CATACG +CATAGA +CATAGT +CATAGC +CATAGG +CATTAA +CATTAT +CATTAC +CATTAG +CATTTA +CATTTT +CATTTC +CATTTG +CATTCA +CATTCT +CATTCC +CATTCG +CATTGA +CATTGT +CATTGC +CATTGG +CATCAA +CATCAT +CATCAC +CATCAG +CATCTA +CATCTT +CATCTC +CATCTG +CATCCA +CATCCT +CATCCC +CATCCG +CATCGA +CATCGT +CATCGC +CATCGG +CATGAA +CATGAT +CATGAC +CATGAG +CATGTA +CATGTT +CATGTC +CATGTG +CATGCA +CATGCT +CATGCC +CATGCG +CATGGA +CATGGT +CATGGC +CATGGG +CACAAA +CACAAT +CACAAC +CACAAG +CACATA +CACATT +CACATC +CACATG +CACACA +CACACT +CACACC +CACACG +CACAGA +CACAGT +CACAGC +CACAGG +CACTAA +CACTAT +CACTAC +CACTAG +CACTTA +CACTTT +CACTTC +CACTTG +CACTCA +CACTCT +CACTCC +CACTCG +CACTGA +CACTGT +CACTGC +CACTGG +CACCAA +CACCAT +CACCAC +CACCAG +CACCTA +CACCTT +CACCTC +CACCTG +CACCCA +CACCCT +CACCCC +CACCCG +CACCGA +CACCGT +CACCGC +CACCGG +CACGAA +CACGAT +CACGAC +CACGAG +CACGTA +CACGTT +CACGTC +CACGTG +CACGCA +CACGCT +CACGCC +CACGCG +CACGGA +CACGGT +CACGGC +CACGGG +CAGAAA +CAGAAT +CAGAAC +CAGAAG +CAGATA +CAGATT +CAGATC +CAGATG +CAGACA +CAGACT +CAGACC +CAGACG +CAGAGA +CAGAGT +CAGAGC +CAGAGG +CAGTAA +CAGTAT +CAGTAC +CAGTAG +CAGTTA +CAGTTT +CAGTTC +CAGTTG +CAGTCA +CAGTCT +CAGTCC +CAGTCG +CAGTGA +CAGTGT +CAGTGC +CAGTGG +CAGCAA +CAGCAT +CAGCAC +CAGCAG +CAGCTA +CAGCTT +CAGCTC +CAGCTG +CAGCCA +CAGCCT +CAGCCC +CAGCCG +CAGCGA +CAGCGT +CAGCGC +CAGCGG +CAGGAA +CAGGAT +CAGGAC +CAGGAG +CAGGTA +CAGGTT +CAGGTC +CAGGTG +CAGGCA +CAGGCT +CAGGCC +CAGGCG +CAGGGA +CAGGGT +CAGGGC +CAGGGG +CTAAAA +CTAAAT +CTAAAC +CTAAAG +CTAATA +CTAATT +CTAATC +CTAATG +CTAACA +CTAACT +CTAACC +CTAACG +CTAAGA +CTAAGT +CTAAGC +CTAAGG +CTATAA +CTATAT +CTATAC +CTATAG +CTATTA +CTATTT +CTATTC +CTATTG +CTATCA +CTATCT +CTATCC +CTATCG +CTATGA +CTATGT +CTATGC +CTATGG +CTACAA +CTACAT +CTACAC +CTACAG +CTACTA +CTACTT +CTACTC +CTACTG +CTACCA +CTACCT +CTACCC +CTACCG +CTACGA +CTACGT +CTACGC +CTACGG +CTAGAA +CTAGAT +CTAGAC +CTAGAG +CTAGTA +CTAGTT +CTAGTC +CTAGTG +CTAGCA +CTAGCT +CTAGCC +CTAGCG +CTAGGA +CTAGGT +CTAGGC +CTAGGG +CTTAAA +CTTAAT +CTTAAC +CTTAAG +CTTATA +CTTATT +CTTATC +CTTATG +CTTACA +CTTACT +CTTACC +CTTACG +CTTAGA +CTTAGT +CTTAGC +CTTAGG +CTTTAA +CTTTAT +CTTTAC +CTTTAG +CTTTTA +CTTTTT +CTTTTC +CTTTTG +CTTTCA +CTTTCT +CTTTCC +CTTTCG +CTTTGA +CTTTGT +CTTTGC +CTTTGG +CTTCAA +CTTCAT +CTTCAC +CTTCAG +CTTCTA +CTTCTT +CTTCTC +CTTCTG +CTTCCA +CTTCCT +CTTCCC +CTTCCG +CTTCGA +CTTCGT +CTTCGC +CTTCGG +CTTGAA +CTTGAT +CTTGAC +CTTGAG +CTTGTA +CTTGTT +CTTGTC +CTTGTG +CTTGCA +CTTGCT +CTTGCC +CTTGCG +CTTGGA +CTTGGT +CTTGGC +CTTGGG +CTCAAA +CTCAAT +CTCAAC +CTCAAG +CTCATA +CTCATT +CTCATC +CTCATG +CTCACA +CTCACT +CTCACC +CTCACG +CTCAGA +CTCAGT +CTCAGC +CTCAGG +CTCTAA +CTCTAT +CTCTAC +CTCTAG +CTCTTA +CTCTTT +CTCTTC +CTCTTG +CTCTCA +CTCTCT +CTCTCC +CTCTCG +CTCTGA +CTCTGT +CTCTGC +CTCTGG +CTCCAA +CTCCAT +CTCCAC +CTCCAG +CTCCTA +CTCCTT +CTCCTC +CTCCTG +CTCCCA +CTCCCT +CTCCCC +CTCCCG +CTCCGA +CTCCGT +CTCCGC +CTCCGG +CTCGAA +CTCGAT +CTCGAC +CTCGAG +CTCGTA +CTCGTT +CTCGTC +CTCGTG +CTCGCA +CTCGCT +CTCGCC +CTCGCG +CTCGGA +CTCGGT +CTCGGC +CTCGGG +CTGAAA +CTGAAT +CTGAAC +CTGAAG +CTGATA +CTGATT +CTGATC +CTGATG +CTGACA +CTGACT +CTGACC +CTGACG +CTGAGA +CTGAGT +CTGAGC +CTGAGG +CTGTAA +CTGTAT +CTGTAC +CTGTAG +CTGTTA +CTGTTT +CTGTTC +CTGTTG +CTGTCA +CTGTCT +CTGTCC +CTGTCG +CTGTGA +CTGTGT +CTGTGC +CTGTGG +CTGCAA +CTGCAT +CTGCAC +CTGCAG +CTGCTA +CTGCTT +CTGCTC +CTGCTG +CTGCCA +CTGCCT +CTGCCC +CTGCCG +CTGCGA +CTGCGT +CTGCGC +CTGCGG +CTGGAA +CTGGAT +CTGGAC +CTGGAG +CTGGTA +CTGGTT +CTGGTC +CTGGTG +CTGGCA +CTGGCT +CTGGCC +CTGGCG +CTGGGA +CTGGGT +CTGGGC +CTGGGG +CCAAAA +CCAAAT +CCAAAC +CCAAAG +CCAATA +CCAATT +CCAATC +CCAATG +CCAACA +CCAACT +CCAACC +CCAACG +CCAAGA +CCAAGT +CCAAGC +CCAAGG +CCATAA +CCATAT +CCATAC +CCATAG +CCATTA +CCATTT +CCATTC +CCATTG +CCATCA +CCATCT +CCATCC +CCATCG +CCATGA +CCATGT +CCATGC +CCATGG +CCACAA +CCACAT +CCACAC +CCACAG +CCACTA +CCACTT +CCACTC +CCACTG +CCACCA +CCACCT +CCACCC +CCACCG +CCACGA +CCACGT +CCACGC +CCACGG +CCAGAA +CCAGAT +CCAGAC +CCAGAG +CCAGTA +CCAGTT +CCAGTC +CCAGTG +CCAGCA +CCAGCT +CCAGCC +CCAGCG +CCAGGA +CCAGGT +CCAGGC +CCAGGG +CCTAAA +CCTAAT +CCTAAC +CCTAAG +CCTATA +CCTATT +CCTATC +CCTATG +CCTACA +CCTACT +CCTACC +CCTACG +CCTAGA +CCTAGT +CCTAGC +CCTAGG +CCTTAA +CCTTAT +CCTTAC +CCTTAG +CCTTTA +CCTTTT +CCTTTC +CCTTTG +CCTTCA +CCTTCT +CCTTCC +CCTTCG +CCTTGA +CCTTGT +CCTTGC +CCTTGG +CCTCAA +CCTCAT +CCTCAC +CCTCAG +CCTCTA +CCTCTT +CCTCTC +CCTCTG +CCTCCA +CCTCCT +CCTCCC +CCTCCG +CCTCGA +CCTCGT +CCTCGC +CCTCGG +CCTGAA +CCTGAT +CCTGAC +CCTGAG +CCTGTA +CCTGTT +CCTGTC +CCTGTG +CCTGCA +CCTGCT +CCTGCC +CCTGCG +CCTGGA +CCTGGT +CCTGGC +CCTGGG +CCCAAA +CCCAAT +CCCAAC +CCCAAG +CCCATA +CCCATT +CCCATC +CCCATG +CCCACA +CCCACT +CCCACC +CCCACG +CCCAGA +CCCAGT +CCCAGC +CCCAGG +CCCTAA +CCCTAT +CCCTAC +CCCTAG +CCCTTA +CCCTTT +CCCTTC +CCCTTG +CCCTCA +CCCTCT +CCCTCC +CCCTCG +CCCTGA +CCCTGT +CCCTGC +CCCTGG +CCCCAA +CCCCAT +CCCCAC +CCCCAG +CCCCTA +CCCCTT +CCCCTC +CCCCTG +CCCCCA +CCCCCT +CCCCCC +CCCCCG +CCCCGA +CCCCGT +CCCCGC +CCCCGG +CCCGAA +CCCGAT +CCCGAC +CCCGAG +CCCGTA +CCCGTT +CCCGTC +CCCGTG +CCCGCA +CCCGCT +CCCGCC +CCCGCG +CCCGGA +CCCGGT +CCCGGC +CCCGGG +CCGAAA +CCGAAT +CCGAAC +CCGAAG +CCGATA +CCGATT +CCGATC +CCGATG +CCGACA +CCGACT +CCGACC +CCGACG +CCGAGA +CCGAGT +CCGAGC +CCGAGG +CCGTAA +CCGTAT +CCGTAC +CCGTAG +CCGTTA +CCGTTT +CCGTTC +CCGTTG +CCGTCA +CCGTCT +CCGTCC +CCGTCG +CCGTGA +CCGTGT +CCGTGC +CCGTGG +CCGCAA +CCGCAT +CCGCAC +CCGCAG +CCGCTA +CCGCTT +CCGCTC +CCGCTG +CCGCCA +CCGCCT +CCGCCC +CCGCCG +CCGCGA +CCGCGT +CCGCGC +CCGCGG +CCGGAA +CCGGAT +CCGGAC +CCGGAG +CCGGTA +CCGGTT +CCGGTC +CCGGTG +CCGGCA +CCGGCT +CCGGCC +CCGGCG +CCGGGA +CCGGGT +CCGGGC +CCGGGG +CGAAAA +CGAAAT +CGAAAC +CGAAAG +CGAATA +CGAATT +CGAATC +CGAATG +CGAACA +CGAACT +CGAACC +CGAACG +CGAAGA +CGAAGT +CGAAGC +CGAAGG +CGATAA +CGATAT +CGATAC +CGATAG +CGATTA +CGATTT +CGATTC +CGATTG +CGATCA +CGATCT +CGATCC +CGATCG +CGATGA +CGATGT +CGATGC +CGATGG +CGACAA +CGACAT +CGACAC +CGACAG +CGACTA +CGACTT +CGACTC +CGACTG +CGACCA +CGACCT +CGACCC +CGACCG +CGACGA +CGACGT +CGACGC +CGACGG +CGAGAA +CGAGAT +CGAGAC +CGAGAG +CGAGTA +CGAGTT +CGAGTC +CGAGTG +CGAGCA +CGAGCT +CGAGCC +CGAGCG +CGAGGA +CGAGGT +CGAGGC +CGAGGG +CGTAAA +CGTAAT +CGTAAC +CGTAAG +CGTATA +CGTATT +CGTATC +CGTATG +CGTACA +CGTACT +CGTACC +CGTACG +CGTAGA +CGTAGT +CGTAGC +CGTAGG +CGTTAA +CGTTAT +CGTTAC +CGTTAG +CGTTTA +CGTTTT +CGTTTC +CGTTTG +CGTTCA +CGTTCT +CGTTCC +CGTTCG +CGTTGA +CGTTGT +CGTTGC +CGTTGG +CGTCAA +CGTCAT +CGTCAC +CGTCAG +CGTCTA +CGTCTT +CGTCTC +CGTCTG +CGTCCA +CGTCCT +CGTCCC +CGTCCG +CGTCGA +CGTCGT +CGTCGC +CGTCGG +CGTGAA +CGTGAT +CGTGAC +CGTGAG +CGTGTA +CGTGTT +CGTGTC +CGTGTG +CGTGCA +CGTGCT +CGTGCC +CGTGCG +CGTGGA +CGTGGT +CGTGGC +CGTGGG +CGCAAA +CGCAAT +CGCAAC +CGCAAG +CGCATA +CGCATT +CGCATC +CGCATG +CGCACA +CGCACT +CGCACC +CGCACG +CGCAGA +CGCAGT +CGCAGC +CGCAGG +CGCTAA +CGCTAT +CGCTAC +CGCTAG +CGCTTA +CGCTTT +CGCTTC +CGCTTG +CGCTCA +CGCTCT +CGCTCC +CGCTCG +CGCTGA +CGCTGT +CGCTGC +CGCTGG +CGCCAA +CGCCAT +CGCCAC +CGCCAG +CGCCTA +CGCCTT +CGCCTC +CGCCTG +CGCCCA +CGCCCT +CGCCCC +CGCCCG +CGCCGA +CGCCGT +CGCCGC +CGCCGG +CGCGAA +CGCGAT +CGCGAC +CGCGAG +CGCGTA +CGCGTT +CGCGTC +CGCGTG +CGCGCA +CGCGCT +CGCGCC +CGCGCG +CGCGGA +CGCGGT +CGCGGC +CGCGGG +CGGAAA +CGGAAT +CGGAAC +CGGAAG +CGGATA +CGGATT +CGGATC +CGGATG +CGGACA +CGGACT +CGGACC +CGGACG +CGGAGA +CGGAGT +CGGAGC +CGGAGG +CGGTAA +CGGTAT +CGGTAC +CGGTAG +CGGTTA +CGGTTT +CGGTTC +CGGTTG +CGGTCA +CGGTCT +CGGTCC +CGGTCG +CGGTGA +CGGTGT +CGGTGC +CGGTGG +CGGCAA +CGGCAT +CGGCAC +CGGCAG +CGGCTA +CGGCTT +CGGCTC +CGGCTG +CGGCCA +CGGCCT +CGGCCC +CGGCCG +CGGCGA +CGGCGT +CGGCGC +CGGCGG +CGGGAA +CGGGAT +CGGGAC +CGGGAG +CGGGTA +CGGGTT +CGGGTC +CGGGTG +CGGGCA +CGGGCT +CGGGCC +CGGGCG +CGGGGA +CGGGGT +CGGGGC +CGGGGG +GAAAAA +GAAAAT +GAAAAC +GAAAAG +GAAATA +GAAATT +GAAATC +GAAATG +GAAACA +GAAACT +GAAACC +GAAACG +GAAAGA +GAAAGT +GAAAGC +GAAAGG +GAATAA +GAATAT +GAATAC +GAATAG +GAATTA +GAATTT +GAATTC +GAATTG +GAATCA +GAATCT +GAATCC +GAATCG +GAATGA +GAATGT +GAATGC +GAATGG +GAACAA +GAACAT +GAACAC +GAACAG +GAACTA +GAACTT +GAACTC +GAACTG +GAACCA +GAACCT +GAACCC +GAACCG +GAACGA +GAACGT +GAACGC +GAACGG +GAAGAA +GAAGAT +GAAGAC +GAAGAG +GAAGTA +GAAGTT +GAAGTC +GAAGTG +GAAGCA +GAAGCT +GAAGCC +GAAGCG +GAAGGA +GAAGGT +GAAGGC +GAAGGG +GATAAA +GATAAT +GATAAC +GATAAG +GATATA +GATATT +GATATC +GATATG +GATACA +GATACT +GATACC +GATACG +GATAGA +GATAGT +GATAGC +GATAGG +GATTAA +GATTAT +GATTAC +GATTAG +GATTTA +GATTTT +GATTTC +GATTTG +GATTCA +GATTCT +GATTCC +GATTCG +GATTGA +GATTGT +GATTGC +GATTGG +GATCAA +GATCAT +GATCAC +GATCAG +GATCTA +GATCTT +GATCTC +GATCTG +GATCCA +GATCCT +GATCCC +GATCCG +GATCGA +GATCGT +GATCGC +GATCGG +GATGAA +GATGAT +GATGAC +GATGAG +GATGTA +GATGTT +GATGTC +GATGTG +GATGCA +GATGCT +GATGCC +GATGCG +GATGGA +GATGGT +GATGGC +GATGGG +GACAAA +GACAAT +GACAAC +GACAAG +GACATA +GACATT +GACATC +GACATG +GACACA +GACACT +GACACC +GACACG +GACAGA +GACAGT +GACAGC +GACAGG +GACTAA +GACTAT +GACTAC +GACTAG +GACTTA +GACTTT +GACTTC +GACTTG +GACTCA +GACTCT +GACTCC +GACTCG +GACTGA +GACTGT +GACTGC +GACTGG +GACCAA +GACCAT +GACCAC +GACCAG +GACCTA +GACCTT +GACCTC +GACCTG +GACCCA +GACCCT +GACCCC +GACCCG +GACCGA +GACCGT +GACCGC +GACCGG +GACGAA +GACGAT +GACGAC +GACGAG +GACGTA +GACGTT +GACGTC +GACGTG +GACGCA +GACGCT +GACGCC +GACGCG +GACGGA +GACGGT +GACGGC +GACGGG +GAGAAA +GAGAAT +GAGAAC +GAGAAG +GAGATA +GAGATT +GAGATC +GAGATG +GAGACA +GAGACT +GAGACC +GAGACG +GAGAGA +GAGAGT +GAGAGC +GAGAGG +GAGTAA +GAGTAT +GAGTAC +GAGTAG +GAGTTA +GAGTTT +GAGTTC +GAGTTG +GAGTCA +GAGTCT +GAGTCC +GAGTCG +GAGTGA +GAGTGT +GAGTGC +GAGTGG +GAGCAA +GAGCAT +GAGCAC +GAGCAG +GAGCTA +GAGCTT +GAGCTC +GAGCTG +GAGCCA +GAGCCT +GAGCCC +GAGCCG +GAGCGA +GAGCGT +GAGCGC +GAGCGG +GAGGAA +GAGGAT +GAGGAC +GAGGAG +GAGGTA +GAGGTT +GAGGTC +GAGGTG +GAGGCA +GAGGCT +GAGGCC +GAGGCG +GAGGGA +GAGGGT +GAGGGC +GAGGGG +GTAAAA +GTAAAT +GTAAAC +GTAAAG +GTAATA +GTAATT +GTAATC +GTAATG +GTAACA +GTAACT +GTAACC +GTAACG +GTAAGA +GTAAGT +GTAAGC +GTAAGG +GTATAA +GTATAT +GTATAC +GTATAG +GTATTA +GTATTT +GTATTC +GTATTG +GTATCA +GTATCT +GTATCC +GTATCG +GTATGA +GTATGT +GTATGC +GTATGG +GTACAA +GTACAT +GTACAC +GTACAG +GTACTA +GTACTT +GTACTC +GTACTG +GTACCA +GTACCT +GTACCC +GTACCG +GTACGA +GTACGT +GTACGC +GTACGG +GTAGAA +GTAGAT +GTAGAC +GTAGAG +GTAGTA +GTAGTT +GTAGTC +GTAGTG +GTAGCA +GTAGCT +GTAGCC +GTAGCG +GTAGGA +GTAGGT +GTAGGC +GTAGGG +GTTAAA +GTTAAT +GTTAAC +GTTAAG +GTTATA +GTTATT +GTTATC +GTTATG +GTTACA +GTTACT +GTTACC +GTTACG +GTTAGA +GTTAGT +GTTAGC +GTTAGG +GTTTAA +GTTTAT +GTTTAC +GTTTAG +GTTTTA +GTTTTT +GTTTTC +GTTTTG +GTTTCA +GTTTCT +GTTTCC +GTTTCG +GTTTGA +GTTTGT +GTTTGC +GTTTGG +GTTCAA +GTTCAT +GTTCAC +GTTCAG +GTTCTA +GTTCTT +GTTCTC +GTTCTG +GTTCCA +GTTCCT +GTTCCC +GTTCCG +GTTCGA +GTTCGT +GTTCGC +GTTCGG +GTTGAA +GTTGAT +GTTGAC +GTTGAG +GTTGTA +GTTGTT +GTTGTC +GTTGTG +GTTGCA +GTTGCT +GTTGCC +GTTGCG +GTTGGA +GTTGGT +GTTGGC +GTTGGG +GTCAAA +GTCAAT +GTCAAC +GTCAAG +GTCATA +GTCATT +GTCATC +GTCATG +GTCACA +GTCACT +GTCACC +GTCACG +GTCAGA +GTCAGT +GTCAGC +GTCAGG +GTCTAA +GTCTAT +GTCTAC +GTCTAG +GTCTTA +GTCTTT +GTCTTC +GTCTTG +GTCTCA +GTCTCT +GTCTCC +GTCTCG +GTCTGA +GTCTGT +GTCTGC +GTCTGG +GTCCAA +GTCCAT +GTCCAC +GTCCAG +GTCCTA +GTCCTT +GTCCTC +GTCCTG +GTCCCA +GTCCCT +GTCCCC +GTCCCG +GTCCGA +GTCCGT +GTCCGC +GTCCGG +GTCGAA +GTCGAT +GTCGAC +GTCGAG +GTCGTA +GTCGTT +GTCGTC +GTCGTG +GTCGCA +GTCGCT +GTCGCC +GTCGCG +GTCGGA +GTCGGT +GTCGGC +GTCGGG +GTGAAA +GTGAAT +GTGAAC +GTGAAG +GTGATA +GTGATT +GTGATC +GTGATG +GTGACA +GTGACT +GTGACC +GTGACG +GTGAGA +GTGAGT +GTGAGC +GTGAGG +GTGTAA +GTGTAT +GTGTAC +GTGTAG +GTGTTA +GTGTTT +GTGTTC +GTGTTG +GTGTCA +GTGTCT +GTGTCC +GTGTCG +GTGTGA +GTGTGT +GTGTGC +GTGTGG +GTGCAA +GTGCAT +GTGCAC +GTGCAG +GTGCTA +GTGCTT +GTGCTC +GTGCTG +GTGCCA +GTGCCT +GTGCCC +GTGCCG +GTGCGA +GTGCGT +GTGCGC +GTGCGG +GTGGAA +GTGGAT +GTGGAC +GTGGAG +GTGGTA +GTGGTT +GTGGTC +GTGGTG +GTGGCA +GTGGCT +GTGGCC +GTGGCG +GTGGGA +GTGGGT +GTGGGC +GTGGGG +GCAAAA +GCAAAT +GCAAAC +GCAAAG +GCAATA +GCAATT +GCAATC +GCAATG +GCAACA +GCAACT +GCAACC +GCAACG +GCAAGA +GCAAGT +GCAAGC +GCAAGG +GCATAA +GCATAT +GCATAC +GCATAG +GCATTA +GCATTT +GCATTC +GCATTG +GCATCA +GCATCT +GCATCC +GCATCG +GCATGA +GCATGT +GCATGC +GCATGG +GCACAA +GCACAT +GCACAC +GCACAG +GCACTA +GCACTT +GCACTC +GCACTG +GCACCA +GCACCT +GCACCC +GCACCG +GCACGA +GCACGT +GCACGC +GCACGG +GCAGAA +GCAGAT +GCAGAC +GCAGAG +GCAGTA +GCAGTT +GCAGTC +GCAGTG +GCAGCA +GCAGCT +GCAGCC +GCAGCG +GCAGGA +GCAGGT +GCAGGC +GCAGGG +GCTAAA +GCTAAT +GCTAAC +GCTAAG +GCTATA +GCTATT +GCTATC +GCTATG +GCTACA +GCTACT +GCTACC +GCTACG +GCTAGA +GCTAGT +GCTAGC +GCTAGG +GCTTAA +GCTTAT +GCTTAC +GCTTAG +GCTTTA +GCTTTT +GCTTTC +GCTTTG +GCTTCA +GCTTCT +GCTTCC +GCTTCG +GCTTGA +GCTTGT +GCTTGC +GCTTGG +GCTCAA +GCTCAT +GCTCAC +GCTCAG +GCTCTA +GCTCTT +GCTCTC +GCTCTG +GCTCCA +GCTCCT +GCTCCC +GCTCCG +GCTCGA +GCTCGT +GCTCGC +GCTCGG +GCTGAA +GCTGAT +GCTGAC +GCTGAG +GCTGTA +GCTGTT +GCTGTC +GCTGTG +GCTGCA +GCTGCT +GCTGCC +GCTGCG +GCTGGA +GCTGGT +GCTGGC +GCTGGG +GCCAAA +GCCAAT +GCCAAC +GCCAAG +GCCATA +GCCATT +GCCATC +GCCATG +GCCACA +GCCACT +GCCACC +GCCACG +GCCAGA +GCCAGT +GCCAGC +GCCAGG +GCCTAA +GCCTAT +GCCTAC +GCCTAG +GCCTTA +GCCTTT +GCCTTC +GCCTTG +GCCTCA +GCCTCT +GCCTCC +GCCTCG +GCCTGA +GCCTGT +GCCTGC +GCCTGG +GCCCAA +GCCCAT +GCCCAC +GCCCAG +GCCCTA +GCCCTT +GCCCTC +GCCCTG +GCCCCA +GCCCCT +GCCCCC +GCCCCG +GCCCGA +GCCCGT +GCCCGC +GCCCGG +GCCGAA +GCCGAT +GCCGAC +GCCGAG +GCCGTA +GCCGTT +GCCGTC +GCCGTG +GCCGCA +GCCGCT +GCCGCC +GCCGCG +GCCGGA +GCCGGT +GCCGGC +GCCGGG +GCGAAA +GCGAAT +GCGAAC +GCGAAG +GCGATA +GCGATT +GCGATC +GCGATG +GCGACA +GCGACT +GCGACC +GCGACG +GCGAGA +GCGAGT +GCGAGC +GCGAGG +GCGTAA +GCGTAT +GCGTAC +GCGTAG +GCGTTA +GCGTTT +GCGTTC +GCGTTG +GCGTCA +GCGTCT +GCGTCC +GCGTCG +GCGTGA +GCGTGT +GCGTGC +GCGTGG +GCGCAA +GCGCAT +GCGCAC +GCGCAG +GCGCTA +GCGCTT +GCGCTC +GCGCTG +GCGCCA +GCGCCT +GCGCCC +GCGCCG +GCGCGA +GCGCGT +GCGCGC +GCGCGG +GCGGAA +GCGGAT +GCGGAC +GCGGAG +GCGGTA +GCGGTT +GCGGTC +GCGGTG +GCGGCA +GCGGCT +GCGGCC +GCGGCG +GCGGGA +GCGGGT +GCGGGC +GCGGGG +GGAAAA +GGAAAT +GGAAAC +GGAAAG +GGAATA +GGAATT +GGAATC +GGAATG +GGAACA +GGAACT +GGAACC +GGAACG +GGAAGA +GGAAGT +GGAAGC +GGAAGG +GGATAA +GGATAT +GGATAC +GGATAG +GGATTA +GGATTT +GGATTC +GGATTG +GGATCA +GGATCT +GGATCC +GGATCG +GGATGA +GGATGT +GGATGC +GGATGG +GGACAA +GGACAT +GGACAC +GGACAG +GGACTA +GGACTT +GGACTC +GGACTG +GGACCA +GGACCT +GGACCC +GGACCG +GGACGA +GGACGT +GGACGC +GGACGG +GGAGAA +GGAGAT +GGAGAC +GGAGAG +GGAGTA +GGAGTT +GGAGTC +GGAGTG +GGAGCA +GGAGCT +GGAGCC +GGAGCG +GGAGGA +GGAGGT +GGAGGC +GGAGGG +GGTAAA +GGTAAT +GGTAAC +GGTAAG +GGTATA +GGTATT +GGTATC +GGTATG +GGTACA +GGTACT +GGTACC +GGTACG +GGTAGA +GGTAGT +GGTAGC +GGTAGG +GGTTAA +GGTTAT +GGTTAC +GGTTAG +GGTTTA +GGTTTT +GGTTTC +GGTTTG +GGTTCA +GGTTCT +GGTTCC +GGTTCG +GGTTGA +GGTTGT +GGTTGC +GGTTGG +GGTCAA +GGTCAT +GGTCAC +GGTCAG +GGTCTA +GGTCTT +GGTCTC +GGTCTG +GGTCCA +GGTCCT +GGTCCC +GGTCCG +GGTCGA +GGTCGT +GGTCGC +GGTCGG +GGTGAA +GGTGAT +GGTGAC +GGTGAG +GGTGTA +GGTGTT +GGTGTC +GGTGTG +GGTGCA +GGTGCT +GGTGCC +GGTGCG +GGTGGA +GGTGGT +GGTGGC +GGTGGG +GGCAAA +GGCAAT +GGCAAC +GGCAAG +GGCATA +GGCATT +GGCATC +GGCATG +GGCACA +GGCACT +GGCACC +GGCACG +GGCAGA +GGCAGT +GGCAGC +GGCAGG +GGCTAA +GGCTAT +GGCTAC +GGCTAG +GGCTTA +GGCTTT +GGCTTC +GGCTTG +GGCTCA +GGCTCT +GGCTCC +GGCTCG +GGCTGA +GGCTGT +GGCTGC +GGCTGG +GGCCAA +GGCCAT +GGCCAC +GGCCAG +GGCCTA +GGCCTT +GGCCTC +GGCCTG +GGCCCA +GGCCCT +GGCCCC +GGCCCG +GGCCGA +GGCCGT +GGCCGC +GGCCGG +GGCGAA +GGCGAT +GGCGAC +GGCGAG +GGCGTA +GGCGTT +GGCGTC +GGCGTG +GGCGCA +GGCGCT +GGCGCC +GGCGCG +GGCGGA +GGCGGT +GGCGGC +GGCGGG +GGGAAA +GGGAAT +GGGAAC +GGGAAG +GGGATA +GGGATT +GGGATC +GGGATG +GGGACA +GGGACT +GGGACC +GGGACG +GGGAGA +GGGAGT +GGGAGC +GGGAGG +GGGTAA +GGGTAT +GGGTAC +GGGTAG +GGGTTA +GGGTTT +GGGTTC +GGGTTG +GGGTCA +GGGTCT +GGGTCC +GGGTCG +GGGTGA +GGGTGT +GGGTGC +GGGTGG +GGGCAA +GGGCAT +GGGCAC +GGGCAG +GGGCTA +GGGCTT +GGGCTC +GGGCTG +GGGCCA +GGGCCT +GGGCCC +GGGCCG +GGGCGA +GGGCGT +GGGCGC +GGGCGG +GGGGAA +GGGGAT +GGGGAC +GGGGAG +GGGGTA +GGGGTT +GGGGTC +GGGGTG +GGGGCA +GGGGCT +GGGGCC +GGGGCG +GGGGGA +GGGGGT +GGGGGC +GGGGGG diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/README.md b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..500fd976d92b217455f181bc98c77e612d9a18d9 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/README.md @@ -0,0 +1,202 @@ +--- +base_model: fabihamakhdoomi/TinyDNABERT +library_name: peft +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.12.0 \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/adapter_config.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..04c63e99bbc6b810449dece367c0afb165125b92 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/adapter_config.json @@ -0,0 +1,42 @@ +{ + "alpha_pattern": {}, + "auto_mapping": null, + "base_model_name_or_path": "fabihamakhdoomi/TinyDNABERT", + "bias": "none", + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 16, + "lora_dropout": 0.2, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": [ + "classifier", + "score" + ], + "peft_type": "LORA", + "r": 8, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "bert.encoder.layer.1.attention.self.query", + "bert.encoder.layer.0.output.dense", + "bert.encoder.layer.1.output.dense", + "bert.encoder.layer.0.intermediate.dense", + "bert.encoder.layer.1.intermediate.dense", + "bert.encoder.layer.0.attention.output.dense", + "bert.encoder.layer.1.attention.output.dense", + "bert.encoder.layer.0.attention.self.query", + "bert.encoder.layer.0.attention.self.value", + "bert.encoder.layer.1.attention.self.key", + "bert.encoder.layer.1.attention.self.value", + "bert.encoder.layer.0.attention.self.key" + ], + "task_type": "SEQ_CLS", + "use_dora": false, + "use_rslora": false +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/adapter_model.safetensors b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/adapter_model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..78d23f9dabfc7e8c12ea36948ebea9eb240506cf --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/adapter_model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f17a88989611d8c24478d064c497271a7c4affa0732e1eebd07b38d2d247fc3c +size 151952 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/optimizer.pt b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..9186f366a5df1797b2756d45f3155b31658e7d96 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0f103ad30b408fd30ebfd3762e84f2a0006eb4262cf6156ffdb9f9c63b6bb8af +size 312682 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/rng_state.pth b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..b07d245718b57ab2fe28aa8fcd87fd566f5ca37a --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a7b08a9b6cab3ce1ea404a1560c73adefdab4cd5f0d8619baa241f06062472ab +size 14244 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/scheduler.pt b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..6c655a9f3dd5d4d532f6c0a84b19b9ae5b766437 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dbc2253b3cc1ddd11640ea56d3925f925017f9aa6e1577411dac42d629457af6 +size 1064 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/special_tokens_map.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..a8b3208c2884c4efb86e49300fdd3dc877220cdf --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/special_tokens_map.json @@ -0,0 +1,7 @@ +{ + "cls_token": "[CLS]", + "mask_token": "[MASK]", + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/tokenizer.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..ebd782de6963b9032ce9374cfe30f5275be1a30b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/tokenizer.json @@ -0,0 +1,4265 @@ +{ + "version": "1.0", + "truncation": { + "direction": "Right", + "max_length": 512, + "strategy": "LongestFirst", + "stride": 0 + }, + "padding": { + "strategy": { + "Fixed": 512 + }, + "direction": "Right", + "pad_to_multiple_of": null, + "pad_id": 0, + "pad_type_id": 0, + "pad_token": "[PAD]" + }, + "added_tokens": [ + { + "id": 0, + "content": "[PAD]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 1, + "content": "[UNK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 2, + "content": "[CLS]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 3, + "content": "[SEP]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 4, + "content": "[MASK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + } + ], + "normalizer": { + "type": "BertNormalizer", + "clean_text": true, + "handle_chinese_chars": true, + "strip_accents": null, + "lowercase": false + }, + "pre_tokenizer": { + "type": "BertPreTokenizer" + }, + "post_processor": { + "type": "TemplateProcessing", + "single": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + } + ], + "pair": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "B", + "type_id": 1 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 1 + } + } + ], + "special_tokens": { + "[CLS]": { + "id": "[CLS]", + "ids": [ + 2 + ], + "tokens": [ + "[CLS]" + ] + }, + "[SEP]": { + "id": "[SEP]", + "ids": [ + 3 + ], + "tokens": [ + "[SEP]" + ] + } + } + }, + "decoder": { + "type": "WordPiece", + "prefix": "##", + "cleanup": true + }, + "model": { + "type": "WordPiece", + "unk_token": "[UNK]", + "continuing_subword_prefix": "##", + "max_input_chars_per_word": 100, + "vocab": { + "[PAD]": 0, + "[UNK]": 1, + "[CLS]": 2, + "[SEP]": 3, + "[MASK]": 4, + "AAAAAA": 5, + "AAAAAT": 6, + "AAAAAC": 7, + "AAAAAG": 8, + "AAAATA": 9, + "AAAATT": 10, + "AAAATC": 11, + "AAAATG": 12, + "AAAACA": 13, + "AAAACT": 14, + "AAAACC": 15, + "AAAACG": 16, + "AAAAGA": 17, + "AAAAGT": 18, + "AAAAGC": 19, + "AAAAGG": 20, + "AAATAA": 21, + "AAATAT": 22, + "AAATAC": 23, + "AAATAG": 24, + "AAATTA": 25, + "AAATTT": 26, + "AAATTC": 27, + "AAATTG": 28, + "AAATCA": 29, + "AAATCT": 30, + "AAATCC": 31, + "AAATCG": 32, + "AAATGA": 33, + "AAATGT": 34, + "AAATGC": 35, + "AAATGG": 36, + "AAACAA": 37, + "AAACAT": 38, + "AAACAC": 39, + "AAACAG": 40, + "AAACTA": 41, + "AAACTT": 42, + "AAACTC": 43, + "AAACTG": 44, + "AAACCA": 45, + "AAACCT": 46, + "AAACCC": 47, + "AAACCG": 48, + "AAACGA": 49, + "AAACGT": 50, + "AAACGC": 51, + "AAACGG": 52, + "AAAGAA": 53, + "AAAGAT": 54, + "AAAGAC": 55, + "AAAGAG": 56, + "AAAGTA": 57, + "AAAGTT": 58, + "AAAGTC": 59, + "AAAGTG": 60, + "AAAGCA": 61, + "AAAGCT": 62, + "AAAGCC": 63, + "AAAGCG": 64, + "AAAGGA": 65, + "AAAGGT": 66, + "AAAGGC": 67, + "AAAGGG": 68, + "AATAAA": 69, + "AATAAT": 70, + "AATAAC": 71, + "AATAAG": 72, + "AATATA": 73, + "AATATT": 74, + "AATATC": 75, + "AATATG": 76, + "AATACA": 77, + "AATACT": 78, + "AATACC": 79, + "AATACG": 80, + "AATAGA": 81, + "AATAGT": 82, + "AATAGC": 83, + "AATAGG": 84, + "AATTAA": 85, + "AATTAT": 86, + "AATTAC": 87, + "AATTAG": 88, + "AATTTA": 89, + "AATTTT": 90, + "AATTTC": 91, + "AATTTG": 92, + "AATTCA": 93, + "AATTCT": 94, + "AATTCC": 95, + "AATTCG": 96, + "AATTGA": 97, + "AATTGT": 98, + "AATTGC": 99, + "AATTGG": 100, + "AATCAA": 101, + "AATCAT": 102, + "AATCAC": 103, + "AATCAG": 104, + "AATCTA": 105, + "AATCTT": 106, + "AATCTC": 107, + "AATCTG": 108, + "AATCCA": 109, + "AATCCT": 110, + "AATCCC": 111, + "AATCCG": 112, + "AATCGA": 113, + "AATCGT": 114, + "AATCGC": 115, + "AATCGG": 116, + "AATGAA": 117, + "AATGAT": 118, + "AATGAC": 119, + "AATGAG": 120, + "AATGTA": 121, + "AATGTT": 122, + "AATGTC": 123, + "AATGTG": 124, + "AATGCA": 125, + "AATGCT": 126, + "AATGCC": 127, + "AATGCG": 128, + "AATGGA": 129, + "AATGGT": 130, + "AATGGC": 131, + "AATGGG": 132, + "AACAAA": 133, + "AACAAT": 134, + "AACAAC": 135, + "AACAAG": 136, + "AACATA": 137, + "AACATT": 138, + "AACATC": 139, + "AACATG": 140, + "AACACA": 141, + "AACACT": 142, + "AACACC": 143, + "AACACG": 144, + "AACAGA": 145, + "AACAGT": 146, + "AACAGC": 147, + "AACAGG": 148, + "AACTAA": 149, + "AACTAT": 150, + "AACTAC": 151, + "AACTAG": 152, + "AACTTA": 153, + "AACTTT": 154, + "AACTTC": 155, + "AACTTG": 156, + "AACTCA": 157, + "AACTCT": 158, + "AACTCC": 159, + "AACTCG": 160, + "AACTGA": 161, + "AACTGT": 162, + "AACTGC": 163, + "AACTGG": 164, + "AACCAA": 165, + "AACCAT": 166, + "AACCAC": 167, + "AACCAG": 168, + "AACCTA": 169, + "AACCTT": 170, + "AACCTC": 171, + "AACCTG": 172, + "AACCCA": 173, + "AACCCT": 174, + "AACCCC": 175, + "AACCCG": 176, + "AACCGA": 177, + "AACCGT": 178, + "AACCGC": 179, + "AACCGG": 180, + "AACGAA": 181, + "AACGAT": 182, + "AACGAC": 183, + "AACGAG": 184, + "AACGTA": 185, + "AACGTT": 186, + "AACGTC": 187, + "AACGTG": 188, + "AACGCA": 189, + "AACGCT": 190, + "AACGCC": 191, + "AACGCG": 192, + "AACGGA": 193, + "AACGGT": 194, + "AACGGC": 195, + "AACGGG": 196, + "AAGAAA": 197, + "AAGAAT": 198, + "AAGAAC": 199, + "AAGAAG": 200, + "AAGATA": 201, + "AAGATT": 202, + "AAGATC": 203, + "AAGATG": 204, + "AAGACA": 205, + "AAGACT": 206, + "AAGACC": 207, + "AAGACG": 208, + "AAGAGA": 209, + "AAGAGT": 210, + "AAGAGC": 211, + "AAGAGG": 212, + "AAGTAA": 213, + "AAGTAT": 214, + "AAGTAC": 215, + "AAGTAG": 216, + "AAGTTA": 217, + "AAGTTT": 218, + "AAGTTC": 219, + "AAGTTG": 220, + "AAGTCA": 221, + "AAGTCT": 222, + "AAGTCC": 223, + "AAGTCG": 224, + "AAGTGA": 225, + "AAGTGT": 226, + "AAGTGC": 227, + "AAGTGG": 228, + "AAGCAA": 229, + "AAGCAT": 230, + "AAGCAC": 231, + "AAGCAG": 232, + "AAGCTA": 233, + "AAGCTT": 234, + "AAGCTC": 235, + "AAGCTG": 236, + "AAGCCA": 237, + "AAGCCT": 238, + "AAGCCC": 239, + "AAGCCG": 240, + "AAGCGA": 241, + "AAGCGT": 242, + "AAGCGC": 243, + "AAGCGG": 244, + "AAGGAA": 245, + "AAGGAT": 246, + "AAGGAC": 247, + "AAGGAG": 248, + "AAGGTA": 249, + "AAGGTT": 250, + "AAGGTC": 251, + "AAGGTG": 252, + "AAGGCA": 253, + "AAGGCT": 254, + "AAGGCC": 255, + "AAGGCG": 256, + "AAGGGA": 257, + "AAGGGT": 258, + "AAGGGC": 259, + "AAGGGG": 260, + "ATAAAA": 261, + "ATAAAT": 262, + "ATAAAC": 263, + "ATAAAG": 264, + "ATAATA": 265, + "ATAATT": 266, + "ATAATC": 267, + "ATAATG": 268, + "ATAACA": 269, + "ATAACT": 270, + "ATAACC": 271, + "ATAACG": 272, + "ATAAGA": 273, + "ATAAGT": 274, + "ATAAGC": 275, + "ATAAGG": 276, + "ATATAA": 277, + "ATATAT": 278, + "ATATAC": 279, + "ATATAG": 280, + "ATATTA": 281, + "ATATTT": 282, + "ATATTC": 283, + "ATATTG": 284, + "ATATCA": 285, + "ATATCT": 286, + "ATATCC": 287, + "ATATCG": 288, + "ATATGA": 289, + "ATATGT": 290, + "ATATGC": 291, + "ATATGG": 292, + "ATACAA": 293, + "ATACAT": 294, + "ATACAC": 295, + "ATACAG": 296, + "ATACTA": 297, + "ATACTT": 298, + "ATACTC": 299, + "ATACTG": 300, + "ATACCA": 301, + "ATACCT": 302, + "ATACCC": 303, + "ATACCG": 304, + "ATACGA": 305, + "ATACGT": 306, + "ATACGC": 307, + "ATACGG": 308, + "ATAGAA": 309, + "ATAGAT": 310, + "ATAGAC": 311, + "ATAGAG": 312, + "ATAGTA": 313, + "ATAGTT": 314, + "ATAGTC": 315, + "ATAGTG": 316, + "ATAGCA": 317, + "ATAGCT": 318, + "ATAGCC": 319, + "ATAGCG": 320, + "ATAGGA": 321, + "ATAGGT": 322, + "ATAGGC": 323, + "ATAGGG": 324, + "ATTAAA": 325, + "ATTAAT": 326, + "ATTAAC": 327, + "ATTAAG": 328, + "ATTATA": 329, + "ATTATT": 330, + "ATTATC": 331, + "ATTATG": 332, + "ATTACA": 333, + "ATTACT": 334, + "ATTACC": 335, + "ATTACG": 336, + "ATTAGA": 337, + "ATTAGT": 338, + "ATTAGC": 339, + "ATTAGG": 340, + "ATTTAA": 341, + "ATTTAT": 342, + "ATTTAC": 343, + "ATTTAG": 344, + "ATTTTA": 345, + "ATTTTT": 346, + "ATTTTC": 347, + "ATTTTG": 348, + "ATTTCA": 349, + "ATTTCT": 350, + "ATTTCC": 351, + "ATTTCG": 352, + "ATTTGA": 353, + "ATTTGT": 354, + "ATTTGC": 355, + "ATTTGG": 356, + "ATTCAA": 357, + "ATTCAT": 358, + "ATTCAC": 359, + "ATTCAG": 360, + "ATTCTA": 361, + "ATTCTT": 362, + "ATTCTC": 363, + "ATTCTG": 364, + "ATTCCA": 365, + "ATTCCT": 366, + "ATTCCC": 367, + "ATTCCG": 368, + "ATTCGA": 369, + "ATTCGT": 370, + "ATTCGC": 371, + "ATTCGG": 372, + "ATTGAA": 373, + "ATTGAT": 374, + "ATTGAC": 375, + "ATTGAG": 376, + "ATTGTA": 377, + "ATTGTT": 378, + "ATTGTC": 379, + "ATTGTG": 380, + "ATTGCA": 381, + "ATTGCT": 382, + "ATTGCC": 383, + "ATTGCG": 384, + "ATTGGA": 385, + "ATTGGT": 386, + "ATTGGC": 387, + "ATTGGG": 388, + "ATCAAA": 389, + "ATCAAT": 390, + "ATCAAC": 391, + "ATCAAG": 392, + "ATCATA": 393, + "ATCATT": 394, + "ATCATC": 395, + "ATCATG": 396, + "ATCACA": 397, + "ATCACT": 398, + "ATCACC": 399, + "ATCACG": 400, + "ATCAGA": 401, + "ATCAGT": 402, + "ATCAGC": 403, + "ATCAGG": 404, + "ATCTAA": 405, + "ATCTAT": 406, + "ATCTAC": 407, + "ATCTAG": 408, + "ATCTTA": 409, + "ATCTTT": 410, + "ATCTTC": 411, + "ATCTTG": 412, + "ATCTCA": 413, + "ATCTCT": 414, + "ATCTCC": 415, + "ATCTCG": 416, + "ATCTGA": 417, + "ATCTGT": 418, + "ATCTGC": 419, + "ATCTGG": 420, + "ATCCAA": 421, + "ATCCAT": 422, + "ATCCAC": 423, + "ATCCAG": 424, + "ATCCTA": 425, + "ATCCTT": 426, + "ATCCTC": 427, + "ATCCTG": 428, + "ATCCCA": 429, + "ATCCCT": 430, + "ATCCCC": 431, + "ATCCCG": 432, + "ATCCGA": 433, + "ATCCGT": 434, + "ATCCGC": 435, + "ATCCGG": 436, + "ATCGAA": 437, + "ATCGAT": 438, + "ATCGAC": 439, + "ATCGAG": 440, + "ATCGTA": 441, + "ATCGTT": 442, + "ATCGTC": 443, + "ATCGTG": 444, + "ATCGCA": 445, + "ATCGCT": 446, + "ATCGCC": 447, + "ATCGCG": 448, + "ATCGGA": 449, + "ATCGGT": 450, + "ATCGGC": 451, + "ATCGGG": 452, + "ATGAAA": 453, + "ATGAAT": 454, + "ATGAAC": 455, + "ATGAAG": 456, + "ATGATA": 457, + "ATGATT": 458, + "ATGATC": 459, + "ATGATG": 460, + "ATGACA": 461, + "ATGACT": 462, + "ATGACC": 463, + "ATGACG": 464, + "ATGAGA": 465, + "ATGAGT": 466, + "ATGAGC": 467, + "ATGAGG": 468, + "ATGTAA": 469, + "ATGTAT": 470, + "ATGTAC": 471, + "ATGTAG": 472, + "ATGTTA": 473, + "ATGTTT": 474, + "ATGTTC": 475, + "ATGTTG": 476, + "ATGTCA": 477, + "ATGTCT": 478, + "ATGTCC": 479, + "ATGTCG": 480, + "ATGTGA": 481, + "ATGTGT": 482, + "ATGTGC": 483, + "ATGTGG": 484, + "ATGCAA": 485, + "ATGCAT": 486, + "ATGCAC": 487, + "ATGCAG": 488, + "ATGCTA": 489, + "ATGCTT": 490, + "ATGCTC": 491, + "ATGCTG": 492, + "ATGCCA": 493, + "ATGCCT": 494, + "ATGCCC": 495, + "ATGCCG": 496, + "ATGCGA": 497, + "ATGCGT": 498, + "ATGCGC": 499, + "ATGCGG": 500, + "ATGGAA": 501, + "ATGGAT": 502, + "ATGGAC": 503, + "ATGGAG": 504, + "ATGGTA": 505, + "ATGGTT": 506, + "ATGGTC": 507, + "ATGGTG": 508, + "ATGGCA": 509, + "ATGGCT": 510, + "ATGGCC": 511, + "ATGGCG": 512, + "ATGGGA": 513, + "ATGGGT": 514, + "ATGGGC": 515, + "ATGGGG": 516, + "ACAAAA": 517, + "ACAAAT": 518, + "ACAAAC": 519, + "ACAAAG": 520, + "ACAATA": 521, + "ACAATT": 522, + "ACAATC": 523, + "ACAATG": 524, + "ACAACA": 525, + "ACAACT": 526, + "ACAACC": 527, + "ACAACG": 528, + "ACAAGA": 529, + "ACAAGT": 530, + "ACAAGC": 531, + "ACAAGG": 532, + "ACATAA": 533, + "ACATAT": 534, + "ACATAC": 535, + "ACATAG": 536, + "ACATTA": 537, + "ACATTT": 538, + "ACATTC": 539, + "ACATTG": 540, + "ACATCA": 541, + "ACATCT": 542, + "ACATCC": 543, + "ACATCG": 544, + "ACATGA": 545, + "ACATGT": 546, + "ACATGC": 547, + "ACATGG": 548, + "ACACAA": 549, + "ACACAT": 550, + "ACACAC": 551, + "ACACAG": 552, + "ACACTA": 553, + "ACACTT": 554, + "ACACTC": 555, + "ACACTG": 556, + "ACACCA": 557, + "ACACCT": 558, + "ACACCC": 559, + "ACACCG": 560, + "ACACGA": 561, + "ACACGT": 562, + "ACACGC": 563, + "ACACGG": 564, + "ACAGAA": 565, + "ACAGAT": 566, + "ACAGAC": 567, + "ACAGAG": 568, + "ACAGTA": 569, + "ACAGTT": 570, + "ACAGTC": 571, + "ACAGTG": 572, + "ACAGCA": 573, + "ACAGCT": 574, + "ACAGCC": 575, + "ACAGCG": 576, + "ACAGGA": 577, + "ACAGGT": 578, + "ACAGGC": 579, + "ACAGGG": 580, + "ACTAAA": 581, + "ACTAAT": 582, + "ACTAAC": 583, + "ACTAAG": 584, + "ACTATA": 585, + "ACTATT": 586, + "ACTATC": 587, + "ACTATG": 588, + "ACTACA": 589, + "ACTACT": 590, + "ACTACC": 591, + "ACTACG": 592, + "ACTAGA": 593, + "ACTAGT": 594, + "ACTAGC": 595, + "ACTAGG": 596, + "ACTTAA": 597, + "ACTTAT": 598, + "ACTTAC": 599, + "ACTTAG": 600, + "ACTTTA": 601, + "ACTTTT": 602, + "ACTTTC": 603, + "ACTTTG": 604, + "ACTTCA": 605, + "ACTTCT": 606, + "ACTTCC": 607, + "ACTTCG": 608, + "ACTTGA": 609, + "ACTTGT": 610, + "ACTTGC": 611, + "ACTTGG": 612, + "ACTCAA": 613, + "ACTCAT": 614, + "ACTCAC": 615, + "ACTCAG": 616, + "ACTCTA": 617, + "ACTCTT": 618, + "ACTCTC": 619, + "ACTCTG": 620, + "ACTCCA": 621, + "ACTCCT": 622, + "ACTCCC": 623, + "ACTCCG": 624, + "ACTCGA": 625, + "ACTCGT": 626, + "ACTCGC": 627, + "ACTCGG": 628, + "ACTGAA": 629, + "ACTGAT": 630, + "ACTGAC": 631, + "ACTGAG": 632, + "ACTGTA": 633, + "ACTGTT": 634, + "ACTGTC": 635, + "ACTGTG": 636, + "ACTGCA": 637, + "ACTGCT": 638, + "ACTGCC": 639, + "ACTGCG": 640, + "ACTGGA": 641, + "ACTGGT": 642, + "ACTGGC": 643, + "ACTGGG": 644, + "ACCAAA": 645, + "ACCAAT": 646, + "ACCAAC": 647, + "ACCAAG": 648, + "ACCATA": 649, + "ACCATT": 650, + "ACCATC": 651, + "ACCATG": 652, + "ACCACA": 653, + "ACCACT": 654, + "ACCACC": 655, + "ACCACG": 656, + "ACCAGA": 657, + "ACCAGT": 658, + "ACCAGC": 659, + "ACCAGG": 660, + "ACCTAA": 661, + "ACCTAT": 662, + "ACCTAC": 663, + "ACCTAG": 664, + "ACCTTA": 665, + "ACCTTT": 666, + "ACCTTC": 667, + "ACCTTG": 668, + "ACCTCA": 669, + "ACCTCT": 670, + "ACCTCC": 671, + "ACCTCG": 672, + "ACCTGA": 673, + "ACCTGT": 674, + "ACCTGC": 675, + "ACCTGG": 676, + "ACCCAA": 677, + "ACCCAT": 678, + "ACCCAC": 679, + "ACCCAG": 680, + "ACCCTA": 681, + "ACCCTT": 682, + "ACCCTC": 683, + "ACCCTG": 684, + "ACCCCA": 685, + "ACCCCT": 686, + "ACCCCC": 687, + "ACCCCG": 688, + "ACCCGA": 689, + "ACCCGT": 690, + "ACCCGC": 691, + "ACCCGG": 692, + "ACCGAA": 693, + "ACCGAT": 694, + "ACCGAC": 695, + "ACCGAG": 696, + "ACCGTA": 697, + "ACCGTT": 698, + "ACCGTC": 699, + "ACCGTG": 700, + "ACCGCA": 701, + "ACCGCT": 702, + "ACCGCC": 703, + "ACCGCG": 704, + "ACCGGA": 705, + "ACCGGT": 706, + "ACCGGC": 707, + "ACCGGG": 708, + "ACGAAA": 709, + "ACGAAT": 710, + "ACGAAC": 711, + "ACGAAG": 712, + "ACGATA": 713, + "ACGATT": 714, + "ACGATC": 715, + "ACGATG": 716, + "ACGACA": 717, + "ACGACT": 718, + "ACGACC": 719, + "ACGACG": 720, + "ACGAGA": 721, + "ACGAGT": 722, + "ACGAGC": 723, + "ACGAGG": 724, + "ACGTAA": 725, + "ACGTAT": 726, + "ACGTAC": 727, + "ACGTAG": 728, + "ACGTTA": 729, + "ACGTTT": 730, + "ACGTTC": 731, + "ACGTTG": 732, + "ACGTCA": 733, + "ACGTCT": 734, + "ACGTCC": 735, + "ACGTCG": 736, + "ACGTGA": 737, + "ACGTGT": 738, + "ACGTGC": 739, + "ACGTGG": 740, + "ACGCAA": 741, + "ACGCAT": 742, + "ACGCAC": 743, + "ACGCAG": 744, + "ACGCTA": 745, + "ACGCTT": 746, + "ACGCTC": 747, + "ACGCTG": 748, + "ACGCCA": 749, + "ACGCCT": 750, + "ACGCCC": 751, + "ACGCCG": 752, + "ACGCGA": 753, + "ACGCGT": 754, + "ACGCGC": 755, + "ACGCGG": 756, + "ACGGAA": 757, + "ACGGAT": 758, + "ACGGAC": 759, + "ACGGAG": 760, + "ACGGTA": 761, + "ACGGTT": 762, + "ACGGTC": 763, + "ACGGTG": 764, + "ACGGCA": 765, + "ACGGCT": 766, + "ACGGCC": 767, + "ACGGCG": 768, + "ACGGGA": 769, + "ACGGGT": 770, + "ACGGGC": 771, + "ACGGGG": 772, + "AGAAAA": 773, + "AGAAAT": 774, + "AGAAAC": 775, + "AGAAAG": 776, + "AGAATA": 777, + "AGAATT": 778, + "AGAATC": 779, + "AGAATG": 780, + "AGAACA": 781, + "AGAACT": 782, + "AGAACC": 783, + "AGAACG": 784, + "AGAAGA": 785, + "AGAAGT": 786, + "AGAAGC": 787, + "AGAAGG": 788, + "AGATAA": 789, + "AGATAT": 790, + "AGATAC": 791, + "AGATAG": 792, + "AGATTA": 793, + "AGATTT": 794, + "AGATTC": 795, + "AGATTG": 796, + "AGATCA": 797, + "AGATCT": 798, + "AGATCC": 799, + "AGATCG": 800, + "AGATGA": 801, + "AGATGT": 802, + "AGATGC": 803, + "AGATGG": 804, + "AGACAA": 805, + "AGACAT": 806, + "AGACAC": 807, + "AGACAG": 808, + "AGACTA": 809, + "AGACTT": 810, + "AGACTC": 811, + "AGACTG": 812, + "AGACCA": 813, + "AGACCT": 814, + "AGACCC": 815, + "AGACCG": 816, + "AGACGA": 817, + "AGACGT": 818, + "AGACGC": 819, + "AGACGG": 820, + "AGAGAA": 821, + "AGAGAT": 822, + "AGAGAC": 823, + "AGAGAG": 824, + "AGAGTA": 825, + "AGAGTT": 826, + "AGAGTC": 827, + "AGAGTG": 828, + "AGAGCA": 829, + "AGAGCT": 830, + "AGAGCC": 831, + "AGAGCG": 832, + "AGAGGA": 833, + "AGAGGT": 834, + "AGAGGC": 835, + "AGAGGG": 836, + "AGTAAA": 837, + "AGTAAT": 838, + "AGTAAC": 839, + "AGTAAG": 840, + "AGTATA": 841, + "AGTATT": 842, + "AGTATC": 843, + "AGTATG": 844, + "AGTACA": 845, + "AGTACT": 846, + "AGTACC": 847, + "AGTACG": 848, + "AGTAGA": 849, + "AGTAGT": 850, + "AGTAGC": 851, + "AGTAGG": 852, + "AGTTAA": 853, + "AGTTAT": 854, + "AGTTAC": 855, + "AGTTAG": 856, + "AGTTTA": 857, + "AGTTTT": 858, + "AGTTTC": 859, + "AGTTTG": 860, + "AGTTCA": 861, + "AGTTCT": 862, + "AGTTCC": 863, + "AGTTCG": 864, + "AGTTGA": 865, + "AGTTGT": 866, + "AGTTGC": 867, + "AGTTGG": 868, + "AGTCAA": 869, + "AGTCAT": 870, + "AGTCAC": 871, + "AGTCAG": 872, + "AGTCTA": 873, + "AGTCTT": 874, + "AGTCTC": 875, + "AGTCTG": 876, + "AGTCCA": 877, + "AGTCCT": 878, + "AGTCCC": 879, + "AGTCCG": 880, + "AGTCGA": 881, + "AGTCGT": 882, + "AGTCGC": 883, + "AGTCGG": 884, + "AGTGAA": 885, + "AGTGAT": 886, + "AGTGAC": 887, + "AGTGAG": 888, + "AGTGTA": 889, + "AGTGTT": 890, + "AGTGTC": 891, + "AGTGTG": 892, + "AGTGCA": 893, + "AGTGCT": 894, + "AGTGCC": 895, + "AGTGCG": 896, + "AGTGGA": 897, + "AGTGGT": 898, + "AGTGGC": 899, + "AGTGGG": 900, + "AGCAAA": 901, + "AGCAAT": 902, + "AGCAAC": 903, + "AGCAAG": 904, + "AGCATA": 905, + "AGCATT": 906, + "AGCATC": 907, + "AGCATG": 908, + "AGCACA": 909, + "AGCACT": 910, + "AGCACC": 911, + "AGCACG": 912, + "AGCAGA": 913, + "AGCAGT": 914, + "AGCAGC": 915, + "AGCAGG": 916, + "AGCTAA": 917, + "AGCTAT": 918, + "AGCTAC": 919, + "AGCTAG": 920, + "AGCTTA": 921, + "AGCTTT": 922, + "AGCTTC": 923, + "AGCTTG": 924, + "AGCTCA": 925, + "AGCTCT": 926, + "AGCTCC": 927, + "AGCTCG": 928, + "AGCTGA": 929, + "AGCTGT": 930, + "AGCTGC": 931, + "AGCTGG": 932, + "AGCCAA": 933, + "AGCCAT": 934, + "AGCCAC": 935, + "AGCCAG": 936, + "AGCCTA": 937, + "AGCCTT": 938, + "AGCCTC": 939, + "AGCCTG": 940, + "AGCCCA": 941, + "AGCCCT": 942, + "AGCCCC": 943, + "AGCCCG": 944, + "AGCCGA": 945, + "AGCCGT": 946, + "AGCCGC": 947, + "AGCCGG": 948, + "AGCGAA": 949, + "AGCGAT": 950, + "AGCGAC": 951, + "AGCGAG": 952, + "AGCGTA": 953, + "AGCGTT": 954, + "AGCGTC": 955, + "AGCGTG": 956, + "AGCGCA": 957, + "AGCGCT": 958, + "AGCGCC": 959, + "AGCGCG": 960, + "AGCGGA": 961, + "AGCGGT": 962, + "AGCGGC": 963, + "AGCGGG": 964, + "AGGAAA": 965, + "AGGAAT": 966, + "AGGAAC": 967, + "AGGAAG": 968, + "AGGATA": 969, + "AGGATT": 970, + "AGGATC": 971, + "AGGATG": 972, + "AGGACA": 973, + "AGGACT": 974, + "AGGACC": 975, + "AGGACG": 976, + "AGGAGA": 977, + "AGGAGT": 978, + "AGGAGC": 979, + "AGGAGG": 980, + "AGGTAA": 981, + "AGGTAT": 982, + "AGGTAC": 983, + "AGGTAG": 984, + "AGGTTA": 985, + "AGGTTT": 986, + "AGGTTC": 987, + "AGGTTG": 988, + "AGGTCA": 989, + "AGGTCT": 990, + "AGGTCC": 991, + "AGGTCG": 992, + "AGGTGA": 993, + "AGGTGT": 994, + "AGGTGC": 995, + "AGGTGG": 996, + "AGGCAA": 997, + "AGGCAT": 998, + "AGGCAC": 999, + "AGGCAG": 1000, + "AGGCTA": 1001, + "AGGCTT": 1002, + "AGGCTC": 1003, + "AGGCTG": 1004, + "AGGCCA": 1005, + "AGGCCT": 1006, + "AGGCCC": 1007, + "AGGCCG": 1008, + "AGGCGA": 1009, + "AGGCGT": 1010, + "AGGCGC": 1011, + "AGGCGG": 1012, + "AGGGAA": 1013, + "AGGGAT": 1014, + "AGGGAC": 1015, + "AGGGAG": 1016, + "AGGGTA": 1017, + "AGGGTT": 1018, + "AGGGTC": 1019, + "AGGGTG": 1020, + "AGGGCA": 1021, + "AGGGCT": 1022, + "AGGGCC": 1023, + "AGGGCG": 1024, + "AGGGGA": 1025, + "AGGGGT": 1026, + "AGGGGC": 1027, + "AGGGGG": 1028, + "TAAAAA": 1029, + "TAAAAT": 1030, + "TAAAAC": 1031, + "TAAAAG": 1032, + "TAAATA": 1033, + "TAAATT": 1034, + "TAAATC": 1035, + "TAAATG": 1036, + "TAAACA": 1037, + "TAAACT": 1038, + "TAAACC": 1039, + "TAAACG": 1040, + "TAAAGA": 1041, + "TAAAGT": 1042, + "TAAAGC": 1043, + "TAAAGG": 1044, + "TAATAA": 1045, + "TAATAT": 1046, + "TAATAC": 1047, + "TAATAG": 1048, + "TAATTA": 1049, + "TAATTT": 1050, + "TAATTC": 1051, + "TAATTG": 1052, + "TAATCA": 1053, + "TAATCT": 1054, + "TAATCC": 1055, + "TAATCG": 1056, + "TAATGA": 1057, + "TAATGT": 1058, + "TAATGC": 1059, + "TAATGG": 1060, + "TAACAA": 1061, + "TAACAT": 1062, + "TAACAC": 1063, + "TAACAG": 1064, + "TAACTA": 1065, + "TAACTT": 1066, + "TAACTC": 1067, + "TAACTG": 1068, + "TAACCA": 1069, + "TAACCT": 1070, + "TAACCC": 1071, + "TAACCG": 1072, + "TAACGA": 1073, + "TAACGT": 1074, + "TAACGC": 1075, + "TAACGG": 1076, + "TAAGAA": 1077, + "TAAGAT": 1078, + "TAAGAC": 1079, + "TAAGAG": 1080, + "TAAGTA": 1081, + "TAAGTT": 1082, + "TAAGTC": 1083, + "TAAGTG": 1084, + "TAAGCA": 1085, + "TAAGCT": 1086, + "TAAGCC": 1087, + "TAAGCG": 1088, + "TAAGGA": 1089, + "TAAGGT": 1090, + "TAAGGC": 1091, + "TAAGGG": 1092, + "TATAAA": 1093, + "TATAAT": 1094, + "TATAAC": 1095, + "TATAAG": 1096, + "TATATA": 1097, + "TATATT": 1098, + "TATATC": 1099, + "TATATG": 1100, + "TATACA": 1101, + "TATACT": 1102, + "TATACC": 1103, + "TATACG": 1104, + "TATAGA": 1105, + "TATAGT": 1106, + "TATAGC": 1107, + "TATAGG": 1108, + "TATTAA": 1109, + "TATTAT": 1110, + "TATTAC": 1111, + "TATTAG": 1112, + "TATTTA": 1113, + "TATTTT": 1114, + "TATTTC": 1115, + "TATTTG": 1116, + "TATTCA": 1117, + "TATTCT": 1118, + "TATTCC": 1119, + "TATTCG": 1120, + "TATTGA": 1121, + "TATTGT": 1122, + "TATTGC": 1123, + "TATTGG": 1124, + "TATCAA": 1125, + "TATCAT": 1126, + "TATCAC": 1127, + "TATCAG": 1128, + "TATCTA": 1129, + "TATCTT": 1130, + "TATCTC": 1131, + "TATCTG": 1132, + "TATCCA": 1133, + "TATCCT": 1134, + "TATCCC": 1135, + "TATCCG": 1136, + "TATCGA": 1137, + "TATCGT": 1138, + "TATCGC": 1139, + "TATCGG": 1140, + "TATGAA": 1141, + "TATGAT": 1142, + "TATGAC": 1143, + "TATGAG": 1144, + "TATGTA": 1145, + "TATGTT": 1146, + "TATGTC": 1147, + "TATGTG": 1148, + "TATGCA": 1149, + "TATGCT": 1150, + "TATGCC": 1151, + "TATGCG": 1152, + "TATGGA": 1153, + "TATGGT": 1154, + "TATGGC": 1155, + "TATGGG": 1156, + "TACAAA": 1157, + "TACAAT": 1158, + "TACAAC": 1159, + "TACAAG": 1160, + "TACATA": 1161, + "TACATT": 1162, + "TACATC": 1163, + "TACATG": 1164, + "TACACA": 1165, + "TACACT": 1166, + "TACACC": 1167, + "TACACG": 1168, + "TACAGA": 1169, + "TACAGT": 1170, + "TACAGC": 1171, + "TACAGG": 1172, + "TACTAA": 1173, + "TACTAT": 1174, + "TACTAC": 1175, + "TACTAG": 1176, + "TACTTA": 1177, + "TACTTT": 1178, + "TACTTC": 1179, + "TACTTG": 1180, + "TACTCA": 1181, + "TACTCT": 1182, + "TACTCC": 1183, + "TACTCG": 1184, + "TACTGA": 1185, + "TACTGT": 1186, + "TACTGC": 1187, + "TACTGG": 1188, + "TACCAA": 1189, + "TACCAT": 1190, + "TACCAC": 1191, + "TACCAG": 1192, + "TACCTA": 1193, + "TACCTT": 1194, + "TACCTC": 1195, + "TACCTG": 1196, + "TACCCA": 1197, + "TACCCT": 1198, + "TACCCC": 1199, + "TACCCG": 1200, + "TACCGA": 1201, + "TACCGT": 1202, + "TACCGC": 1203, + "TACCGG": 1204, + "TACGAA": 1205, + "TACGAT": 1206, + "TACGAC": 1207, + "TACGAG": 1208, + "TACGTA": 1209, + "TACGTT": 1210, + "TACGTC": 1211, + "TACGTG": 1212, + "TACGCA": 1213, + "TACGCT": 1214, + "TACGCC": 1215, + "TACGCG": 1216, + "TACGGA": 1217, + "TACGGT": 1218, + "TACGGC": 1219, + "TACGGG": 1220, + "TAGAAA": 1221, + "TAGAAT": 1222, + "TAGAAC": 1223, + "TAGAAG": 1224, + "TAGATA": 1225, + "TAGATT": 1226, + "TAGATC": 1227, + "TAGATG": 1228, + "TAGACA": 1229, + "TAGACT": 1230, + "TAGACC": 1231, + "TAGACG": 1232, + "TAGAGA": 1233, + "TAGAGT": 1234, + "TAGAGC": 1235, + "TAGAGG": 1236, + "TAGTAA": 1237, + "TAGTAT": 1238, + "TAGTAC": 1239, + "TAGTAG": 1240, + "TAGTTA": 1241, + "TAGTTT": 1242, + "TAGTTC": 1243, + "TAGTTG": 1244, + "TAGTCA": 1245, + "TAGTCT": 1246, + "TAGTCC": 1247, + "TAGTCG": 1248, + "TAGTGA": 1249, + "TAGTGT": 1250, + "TAGTGC": 1251, + "TAGTGG": 1252, + "TAGCAA": 1253, + "TAGCAT": 1254, + "TAGCAC": 1255, + "TAGCAG": 1256, + "TAGCTA": 1257, + "TAGCTT": 1258, + "TAGCTC": 1259, + "TAGCTG": 1260, + "TAGCCA": 1261, + "TAGCCT": 1262, + "TAGCCC": 1263, + "TAGCCG": 1264, + "TAGCGA": 1265, + "TAGCGT": 1266, + "TAGCGC": 1267, + "TAGCGG": 1268, + "TAGGAA": 1269, + "TAGGAT": 1270, + "TAGGAC": 1271, + "TAGGAG": 1272, + "TAGGTA": 1273, + "TAGGTT": 1274, + "TAGGTC": 1275, + "TAGGTG": 1276, + "TAGGCA": 1277, + "TAGGCT": 1278, + "TAGGCC": 1279, + "TAGGCG": 1280, + "TAGGGA": 1281, + "TAGGGT": 1282, + "TAGGGC": 1283, + "TAGGGG": 1284, + "TTAAAA": 1285, + "TTAAAT": 1286, + "TTAAAC": 1287, + "TTAAAG": 1288, + "TTAATA": 1289, + "TTAATT": 1290, + "TTAATC": 1291, + "TTAATG": 1292, + "TTAACA": 1293, + "TTAACT": 1294, + "TTAACC": 1295, + "TTAACG": 1296, + "TTAAGA": 1297, + "TTAAGT": 1298, + "TTAAGC": 1299, + "TTAAGG": 1300, + "TTATAA": 1301, + "TTATAT": 1302, + "TTATAC": 1303, + "TTATAG": 1304, + "TTATTA": 1305, + "TTATTT": 1306, + "TTATTC": 1307, + "TTATTG": 1308, + "TTATCA": 1309, + "TTATCT": 1310, + "TTATCC": 1311, + "TTATCG": 1312, + "TTATGA": 1313, + "TTATGT": 1314, + "TTATGC": 1315, + "TTATGG": 1316, + "TTACAA": 1317, + "TTACAT": 1318, + "TTACAC": 1319, + "TTACAG": 1320, + "TTACTA": 1321, + "TTACTT": 1322, + "TTACTC": 1323, + "TTACTG": 1324, + "TTACCA": 1325, + "TTACCT": 1326, + "TTACCC": 1327, + "TTACCG": 1328, + "TTACGA": 1329, + "TTACGT": 1330, + "TTACGC": 1331, + "TTACGG": 1332, + "TTAGAA": 1333, + "TTAGAT": 1334, + "TTAGAC": 1335, + "TTAGAG": 1336, + "TTAGTA": 1337, + "TTAGTT": 1338, + "TTAGTC": 1339, + "TTAGTG": 1340, + "TTAGCA": 1341, + "TTAGCT": 1342, + "TTAGCC": 1343, + "TTAGCG": 1344, + "TTAGGA": 1345, + "TTAGGT": 1346, + "TTAGGC": 1347, + "TTAGGG": 1348, + "TTTAAA": 1349, + "TTTAAT": 1350, + "TTTAAC": 1351, + "TTTAAG": 1352, + "TTTATA": 1353, + "TTTATT": 1354, + "TTTATC": 1355, + "TTTATG": 1356, + "TTTACA": 1357, + "TTTACT": 1358, + "TTTACC": 1359, + "TTTACG": 1360, + "TTTAGA": 1361, + "TTTAGT": 1362, + "TTTAGC": 1363, + "TTTAGG": 1364, + "TTTTAA": 1365, + "TTTTAT": 1366, + "TTTTAC": 1367, + "TTTTAG": 1368, + "TTTTTA": 1369, + "TTTTTT": 1370, + "TTTTTC": 1371, + "TTTTTG": 1372, + "TTTTCA": 1373, + "TTTTCT": 1374, + "TTTTCC": 1375, + "TTTTCG": 1376, + "TTTTGA": 1377, + "TTTTGT": 1378, + "TTTTGC": 1379, + "TTTTGG": 1380, + "TTTCAA": 1381, + "TTTCAT": 1382, + "TTTCAC": 1383, + "TTTCAG": 1384, + "TTTCTA": 1385, + "TTTCTT": 1386, + "TTTCTC": 1387, + "TTTCTG": 1388, + "TTTCCA": 1389, + "TTTCCT": 1390, + "TTTCCC": 1391, + "TTTCCG": 1392, + "TTTCGA": 1393, + "TTTCGT": 1394, + "TTTCGC": 1395, + "TTTCGG": 1396, + "TTTGAA": 1397, + "TTTGAT": 1398, + "TTTGAC": 1399, + "TTTGAG": 1400, + "TTTGTA": 1401, + "TTTGTT": 1402, + "TTTGTC": 1403, + "TTTGTG": 1404, + "TTTGCA": 1405, + "TTTGCT": 1406, + "TTTGCC": 1407, + "TTTGCG": 1408, + "TTTGGA": 1409, + "TTTGGT": 1410, + "TTTGGC": 1411, + "TTTGGG": 1412, + "TTCAAA": 1413, + "TTCAAT": 1414, + "TTCAAC": 1415, + "TTCAAG": 1416, + "TTCATA": 1417, + "TTCATT": 1418, + "TTCATC": 1419, + "TTCATG": 1420, + "TTCACA": 1421, + "TTCACT": 1422, + "TTCACC": 1423, + "TTCACG": 1424, + "TTCAGA": 1425, + "TTCAGT": 1426, + "TTCAGC": 1427, + "TTCAGG": 1428, + "TTCTAA": 1429, + "TTCTAT": 1430, + "TTCTAC": 1431, + "TTCTAG": 1432, + "TTCTTA": 1433, + "TTCTTT": 1434, + "TTCTTC": 1435, + "TTCTTG": 1436, + "TTCTCA": 1437, + "TTCTCT": 1438, + "TTCTCC": 1439, + "TTCTCG": 1440, + "TTCTGA": 1441, + "TTCTGT": 1442, + "TTCTGC": 1443, + "TTCTGG": 1444, + "TTCCAA": 1445, + "TTCCAT": 1446, + "TTCCAC": 1447, + "TTCCAG": 1448, + "TTCCTA": 1449, + "TTCCTT": 1450, + "TTCCTC": 1451, + "TTCCTG": 1452, + "TTCCCA": 1453, + "TTCCCT": 1454, + "TTCCCC": 1455, + "TTCCCG": 1456, + "TTCCGA": 1457, + "TTCCGT": 1458, + "TTCCGC": 1459, + "TTCCGG": 1460, + "TTCGAA": 1461, + "TTCGAT": 1462, + "TTCGAC": 1463, + "TTCGAG": 1464, + "TTCGTA": 1465, + "TTCGTT": 1466, + "TTCGTC": 1467, + "TTCGTG": 1468, + "TTCGCA": 1469, + "TTCGCT": 1470, + "TTCGCC": 1471, + "TTCGCG": 1472, + "TTCGGA": 1473, + "TTCGGT": 1474, + "TTCGGC": 1475, + "TTCGGG": 1476, + "TTGAAA": 1477, + "TTGAAT": 1478, + "TTGAAC": 1479, + "TTGAAG": 1480, + "TTGATA": 1481, + "TTGATT": 1482, + "TTGATC": 1483, + "TTGATG": 1484, + "TTGACA": 1485, + "TTGACT": 1486, + "TTGACC": 1487, + "TTGACG": 1488, + "TTGAGA": 1489, + "TTGAGT": 1490, + "TTGAGC": 1491, + "TTGAGG": 1492, + "TTGTAA": 1493, + "TTGTAT": 1494, + "TTGTAC": 1495, + "TTGTAG": 1496, + "TTGTTA": 1497, + "TTGTTT": 1498, + "TTGTTC": 1499, + "TTGTTG": 1500, + "TTGTCA": 1501, + "TTGTCT": 1502, + "TTGTCC": 1503, + "TTGTCG": 1504, + "TTGTGA": 1505, + "TTGTGT": 1506, + "TTGTGC": 1507, + "TTGTGG": 1508, + "TTGCAA": 1509, + "TTGCAT": 1510, + "TTGCAC": 1511, + "TTGCAG": 1512, + "TTGCTA": 1513, + "TTGCTT": 1514, + "TTGCTC": 1515, + "TTGCTG": 1516, + "TTGCCA": 1517, + "TTGCCT": 1518, + "TTGCCC": 1519, + "TTGCCG": 1520, + "TTGCGA": 1521, + "TTGCGT": 1522, + "TTGCGC": 1523, + "TTGCGG": 1524, + "TTGGAA": 1525, + "TTGGAT": 1526, + "TTGGAC": 1527, + "TTGGAG": 1528, + "TTGGTA": 1529, + "TTGGTT": 1530, + "TTGGTC": 1531, + "TTGGTG": 1532, + "TTGGCA": 1533, + "TTGGCT": 1534, + "TTGGCC": 1535, + "TTGGCG": 1536, + "TTGGGA": 1537, + "TTGGGT": 1538, + "TTGGGC": 1539, + "TTGGGG": 1540, + "TCAAAA": 1541, + "TCAAAT": 1542, + "TCAAAC": 1543, + "TCAAAG": 1544, + "TCAATA": 1545, + "TCAATT": 1546, + "TCAATC": 1547, + "TCAATG": 1548, + "TCAACA": 1549, + "TCAACT": 1550, + "TCAACC": 1551, + "TCAACG": 1552, + "TCAAGA": 1553, + "TCAAGT": 1554, + "TCAAGC": 1555, + "TCAAGG": 1556, + "TCATAA": 1557, + "TCATAT": 1558, + "TCATAC": 1559, + "TCATAG": 1560, + "TCATTA": 1561, + "TCATTT": 1562, + "TCATTC": 1563, + "TCATTG": 1564, + "TCATCA": 1565, + "TCATCT": 1566, + "TCATCC": 1567, + "TCATCG": 1568, + "TCATGA": 1569, + "TCATGT": 1570, + "TCATGC": 1571, + "TCATGG": 1572, + "TCACAA": 1573, + "TCACAT": 1574, + "TCACAC": 1575, + "TCACAG": 1576, + "TCACTA": 1577, + "TCACTT": 1578, + "TCACTC": 1579, + "TCACTG": 1580, + "TCACCA": 1581, + "TCACCT": 1582, + "TCACCC": 1583, + "TCACCG": 1584, + "TCACGA": 1585, + "TCACGT": 1586, + "TCACGC": 1587, + "TCACGG": 1588, + "TCAGAA": 1589, + "TCAGAT": 1590, + "TCAGAC": 1591, + "TCAGAG": 1592, + "TCAGTA": 1593, + "TCAGTT": 1594, + "TCAGTC": 1595, + "TCAGTG": 1596, + "TCAGCA": 1597, + "TCAGCT": 1598, + "TCAGCC": 1599, + "TCAGCG": 1600, + "TCAGGA": 1601, + "TCAGGT": 1602, + "TCAGGC": 1603, + "TCAGGG": 1604, + "TCTAAA": 1605, + "TCTAAT": 1606, + "TCTAAC": 1607, + "TCTAAG": 1608, + "TCTATA": 1609, + "TCTATT": 1610, + "TCTATC": 1611, + "TCTATG": 1612, + "TCTACA": 1613, + "TCTACT": 1614, + "TCTACC": 1615, + "TCTACG": 1616, + "TCTAGA": 1617, + "TCTAGT": 1618, + "TCTAGC": 1619, + "TCTAGG": 1620, + "TCTTAA": 1621, + "TCTTAT": 1622, + "TCTTAC": 1623, + "TCTTAG": 1624, + "TCTTTA": 1625, + "TCTTTT": 1626, + "TCTTTC": 1627, + "TCTTTG": 1628, + "TCTTCA": 1629, + "TCTTCT": 1630, + "TCTTCC": 1631, + "TCTTCG": 1632, + "TCTTGA": 1633, + "TCTTGT": 1634, + "TCTTGC": 1635, + "TCTTGG": 1636, + "TCTCAA": 1637, + "TCTCAT": 1638, + "TCTCAC": 1639, + "TCTCAG": 1640, + "TCTCTA": 1641, + "TCTCTT": 1642, + "TCTCTC": 1643, + "TCTCTG": 1644, + "TCTCCA": 1645, + "TCTCCT": 1646, + "TCTCCC": 1647, + "TCTCCG": 1648, + "TCTCGA": 1649, + "TCTCGT": 1650, + "TCTCGC": 1651, + "TCTCGG": 1652, + "TCTGAA": 1653, + "TCTGAT": 1654, + "TCTGAC": 1655, + "TCTGAG": 1656, + "TCTGTA": 1657, + "TCTGTT": 1658, + "TCTGTC": 1659, + "TCTGTG": 1660, + "TCTGCA": 1661, + "TCTGCT": 1662, + "TCTGCC": 1663, + "TCTGCG": 1664, + "TCTGGA": 1665, + "TCTGGT": 1666, + "TCTGGC": 1667, + "TCTGGG": 1668, + "TCCAAA": 1669, + "TCCAAT": 1670, + "TCCAAC": 1671, + "TCCAAG": 1672, + "TCCATA": 1673, + "TCCATT": 1674, + "TCCATC": 1675, + "TCCATG": 1676, + "TCCACA": 1677, + "TCCACT": 1678, + "TCCACC": 1679, + "TCCACG": 1680, + "TCCAGA": 1681, + "TCCAGT": 1682, + "TCCAGC": 1683, + "TCCAGG": 1684, + "TCCTAA": 1685, + "TCCTAT": 1686, + "TCCTAC": 1687, + "TCCTAG": 1688, + "TCCTTA": 1689, + "TCCTTT": 1690, + "TCCTTC": 1691, + "TCCTTG": 1692, + "TCCTCA": 1693, + "TCCTCT": 1694, + "TCCTCC": 1695, + "TCCTCG": 1696, + "TCCTGA": 1697, + "TCCTGT": 1698, + "TCCTGC": 1699, + "TCCTGG": 1700, + "TCCCAA": 1701, + "TCCCAT": 1702, + "TCCCAC": 1703, + "TCCCAG": 1704, + "TCCCTA": 1705, + "TCCCTT": 1706, + "TCCCTC": 1707, + "TCCCTG": 1708, + "TCCCCA": 1709, + "TCCCCT": 1710, + "TCCCCC": 1711, + "TCCCCG": 1712, + "TCCCGA": 1713, + "TCCCGT": 1714, + "TCCCGC": 1715, + "TCCCGG": 1716, + "TCCGAA": 1717, + "TCCGAT": 1718, + "TCCGAC": 1719, + "TCCGAG": 1720, + "TCCGTA": 1721, + "TCCGTT": 1722, + "TCCGTC": 1723, + "TCCGTG": 1724, + "TCCGCA": 1725, + "TCCGCT": 1726, + "TCCGCC": 1727, + "TCCGCG": 1728, + "TCCGGA": 1729, + "TCCGGT": 1730, + "TCCGGC": 1731, + "TCCGGG": 1732, + "TCGAAA": 1733, + "TCGAAT": 1734, + "TCGAAC": 1735, + "TCGAAG": 1736, + "TCGATA": 1737, + "TCGATT": 1738, + "TCGATC": 1739, + "TCGATG": 1740, + "TCGACA": 1741, + "TCGACT": 1742, + "TCGACC": 1743, + "TCGACG": 1744, + "TCGAGA": 1745, + "TCGAGT": 1746, + "TCGAGC": 1747, + "TCGAGG": 1748, + "TCGTAA": 1749, + "TCGTAT": 1750, + "TCGTAC": 1751, + "TCGTAG": 1752, + "TCGTTA": 1753, + "TCGTTT": 1754, + "TCGTTC": 1755, + "TCGTTG": 1756, + "TCGTCA": 1757, + "TCGTCT": 1758, + "TCGTCC": 1759, + "TCGTCG": 1760, + "TCGTGA": 1761, + "TCGTGT": 1762, + "TCGTGC": 1763, + "TCGTGG": 1764, + "TCGCAA": 1765, + "TCGCAT": 1766, + "TCGCAC": 1767, + "TCGCAG": 1768, + "TCGCTA": 1769, + "TCGCTT": 1770, + "TCGCTC": 1771, + "TCGCTG": 1772, + "TCGCCA": 1773, + "TCGCCT": 1774, + "TCGCCC": 1775, + "TCGCCG": 1776, + "TCGCGA": 1777, + "TCGCGT": 1778, + "TCGCGC": 1779, + "TCGCGG": 1780, + "TCGGAA": 1781, + "TCGGAT": 1782, + "TCGGAC": 1783, + "TCGGAG": 1784, + "TCGGTA": 1785, + "TCGGTT": 1786, + "TCGGTC": 1787, + "TCGGTG": 1788, + "TCGGCA": 1789, + "TCGGCT": 1790, + "TCGGCC": 1791, + "TCGGCG": 1792, + "TCGGGA": 1793, + "TCGGGT": 1794, + "TCGGGC": 1795, + "TCGGGG": 1796, + "TGAAAA": 1797, + "TGAAAT": 1798, + "TGAAAC": 1799, + "TGAAAG": 1800, + "TGAATA": 1801, + "TGAATT": 1802, + "TGAATC": 1803, + "TGAATG": 1804, + "TGAACA": 1805, + "TGAACT": 1806, + "TGAACC": 1807, + "TGAACG": 1808, + "TGAAGA": 1809, + "TGAAGT": 1810, + "TGAAGC": 1811, + "TGAAGG": 1812, + "TGATAA": 1813, + "TGATAT": 1814, + "TGATAC": 1815, + "TGATAG": 1816, + "TGATTA": 1817, + "TGATTT": 1818, + "TGATTC": 1819, + "TGATTG": 1820, + "TGATCA": 1821, + "TGATCT": 1822, + "TGATCC": 1823, + "TGATCG": 1824, + "TGATGA": 1825, + "TGATGT": 1826, + "TGATGC": 1827, + "TGATGG": 1828, + "TGACAA": 1829, + "TGACAT": 1830, + "TGACAC": 1831, + "TGACAG": 1832, + "TGACTA": 1833, + "TGACTT": 1834, + "TGACTC": 1835, + "TGACTG": 1836, + "TGACCA": 1837, + "TGACCT": 1838, + "TGACCC": 1839, + "TGACCG": 1840, + "TGACGA": 1841, + "TGACGT": 1842, + "TGACGC": 1843, + "TGACGG": 1844, + "TGAGAA": 1845, + "TGAGAT": 1846, + "TGAGAC": 1847, + "TGAGAG": 1848, + "TGAGTA": 1849, + "TGAGTT": 1850, + "TGAGTC": 1851, + "TGAGTG": 1852, + "TGAGCA": 1853, + "TGAGCT": 1854, + "TGAGCC": 1855, + "TGAGCG": 1856, + "TGAGGA": 1857, + "TGAGGT": 1858, + "TGAGGC": 1859, + "TGAGGG": 1860, + "TGTAAA": 1861, + "TGTAAT": 1862, + "TGTAAC": 1863, + "TGTAAG": 1864, + "TGTATA": 1865, + "TGTATT": 1866, + "TGTATC": 1867, + "TGTATG": 1868, + "TGTACA": 1869, + "TGTACT": 1870, + "TGTACC": 1871, + "TGTACG": 1872, + "TGTAGA": 1873, + "TGTAGT": 1874, + "TGTAGC": 1875, + "TGTAGG": 1876, + "TGTTAA": 1877, + "TGTTAT": 1878, + "TGTTAC": 1879, + "TGTTAG": 1880, + "TGTTTA": 1881, + "TGTTTT": 1882, + "TGTTTC": 1883, + "TGTTTG": 1884, + "TGTTCA": 1885, + "TGTTCT": 1886, + "TGTTCC": 1887, + "TGTTCG": 1888, + "TGTTGA": 1889, + "TGTTGT": 1890, + "TGTTGC": 1891, + "TGTTGG": 1892, + "TGTCAA": 1893, + "TGTCAT": 1894, + "TGTCAC": 1895, + "TGTCAG": 1896, + "TGTCTA": 1897, + "TGTCTT": 1898, + "TGTCTC": 1899, + "TGTCTG": 1900, + "TGTCCA": 1901, + "TGTCCT": 1902, + "TGTCCC": 1903, + "TGTCCG": 1904, + "TGTCGA": 1905, + "TGTCGT": 1906, + "TGTCGC": 1907, + "TGTCGG": 1908, + "TGTGAA": 1909, + "TGTGAT": 1910, + "TGTGAC": 1911, + "TGTGAG": 1912, + "TGTGTA": 1913, + "TGTGTT": 1914, + "TGTGTC": 1915, + "TGTGTG": 1916, + "TGTGCA": 1917, + "TGTGCT": 1918, + "TGTGCC": 1919, + "TGTGCG": 1920, + "TGTGGA": 1921, + "TGTGGT": 1922, + "TGTGGC": 1923, + "TGTGGG": 1924, + "TGCAAA": 1925, + "TGCAAT": 1926, + "TGCAAC": 1927, + "TGCAAG": 1928, + "TGCATA": 1929, + "TGCATT": 1930, + "TGCATC": 1931, + "TGCATG": 1932, + "TGCACA": 1933, + "TGCACT": 1934, + "TGCACC": 1935, + "TGCACG": 1936, + "TGCAGA": 1937, + "TGCAGT": 1938, + "TGCAGC": 1939, + "TGCAGG": 1940, + "TGCTAA": 1941, + "TGCTAT": 1942, + "TGCTAC": 1943, + "TGCTAG": 1944, + "TGCTTA": 1945, + "TGCTTT": 1946, + "TGCTTC": 1947, + "TGCTTG": 1948, + "TGCTCA": 1949, + "TGCTCT": 1950, + "TGCTCC": 1951, + "TGCTCG": 1952, + "TGCTGA": 1953, + "TGCTGT": 1954, + "TGCTGC": 1955, + "TGCTGG": 1956, + "TGCCAA": 1957, + "TGCCAT": 1958, + "TGCCAC": 1959, + "TGCCAG": 1960, + "TGCCTA": 1961, + "TGCCTT": 1962, + "TGCCTC": 1963, + "TGCCTG": 1964, + "TGCCCA": 1965, + "TGCCCT": 1966, + "TGCCCC": 1967, + "TGCCCG": 1968, + "TGCCGA": 1969, + "TGCCGT": 1970, + "TGCCGC": 1971, + "TGCCGG": 1972, + "TGCGAA": 1973, + "TGCGAT": 1974, + "TGCGAC": 1975, + "TGCGAG": 1976, + "TGCGTA": 1977, + "TGCGTT": 1978, + "TGCGTC": 1979, + "TGCGTG": 1980, + "TGCGCA": 1981, + "TGCGCT": 1982, + "TGCGCC": 1983, + "TGCGCG": 1984, + "TGCGGA": 1985, + "TGCGGT": 1986, + "TGCGGC": 1987, + "TGCGGG": 1988, + "TGGAAA": 1989, + "TGGAAT": 1990, + "TGGAAC": 1991, + "TGGAAG": 1992, + "TGGATA": 1993, + "TGGATT": 1994, + "TGGATC": 1995, + "TGGATG": 1996, + "TGGACA": 1997, + "TGGACT": 1998, + "TGGACC": 1999, + "TGGACG": 2000, + "TGGAGA": 2001, + "TGGAGT": 2002, + "TGGAGC": 2003, + "TGGAGG": 2004, + "TGGTAA": 2005, + "TGGTAT": 2006, + "TGGTAC": 2007, + "TGGTAG": 2008, + "TGGTTA": 2009, + "TGGTTT": 2010, + "TGGTTC": 2011, + "TGGTTG": 2012, + "TGGTCA": 2013, + "TGGTCT": 2014, + "TGGTCC": 2015, + "TGGTCG": 2016, + "TGGTGA": 2017, + "TGGTGT": 2018, + "TGGTGC": 2019, + "TGGTGG": 2020, + "TGGCAA": 2021, + "TGGCAT": 2022, + "TGGCAC": 2023, + "TGGCAG": 2024, + "TGGCTA": 2025, + "TGGCTT": 2026, + "TGGCTC": 2027, + "TGGCTG": 2028, + "TGGCCA": 2029, + "TGGCCT": 2030, + "TGGCCC": 2031, + "TGGCCG": 2032, + "TGGCGA": 2033, + "TGGCGT": 2034, + "TGGCGC": 2035, + "TGGCGG": 2036, + "TGGGAA": 2037, + "TGGGAT": 2038, + "TGGGAC": 2039, + "TGGGAG": 2040, + "TGGGTA": 2041, + "TGGGTT": 2042, + "TGGGTC": 2043, + "TGGGTG": 2044, + "TGGGCA": 2045, + "TGGGCT": 2046, + "TGGGCC": 2047, + "TGGGCG": 2048, + "TGGGGA": 2049, + "TGGGGT": 2050, + "TGGGGC": 2051, + "TGGGGG": 2052, + "CAAAAA": 2053, + "CAAAAT": 2054, + "CAAAAC": 2055, + "CAAAAG": 2056, + "CAAATA": 2057, + "CAAATT": 2058, + "CAAATC": 2059, + "CAAATG": 2060, + "CAAACA": 2061, + "CAAACT": 2062, + "CAAACC": 2063, + "CAAACG": 2064, + "CAAAGA": 2065, + "CAAAGT": 2066, + "CAAAGC": 2067, + "CAAAGG": 2068, + "CAATAA": 2069, + "CAATAT": 2070, + "CAATAC": 2071, + "CAATAG": 2072, + "CAATTA": 2073, + "CAATTT": 2074, + "CAATTC": 2075, + "CAATTG": 2076, + "CAATCA": 2077, + "CAATCT": 2078, + "CAATCC": 2079, + "CAATCG": 2080, + "CAATGA": 2081, + "CAATGT": 2082, + "CAATGC": 2083, + "CAATGG": 2084, + "CAACAA": 2085, + "CAACAT": 2086, + "CAACAC": 2087, + "CAACAG": 2088, + "CAACTA": 2089, + "CAACTT": 2090, + "CAACTC": 2091, + "CAACTG": 2092, + "CAACCA": 2093, + "CAACCT": 2094, + "CAACCC": 2095, + "CAACCG": 2096, + "CAACGA": 2097, + "CAACGT": 2098, + "CAACGC": 2099, + "CAACGG": 2100, + "CAAGAA": 2101, + "CAAGAT": 2102, + "CAAGAC": 2103, + "CAAGAG": 2104, + "CAAGTA": 2105, + "CAAGTT": 2106, + "CAAGTC": 2107, + "CAAGTG": 2108, + "CAAGCA": 2109, + "CAAGCT": 2110, + "CAAGCC": 2111, + "CAAGCG": 2112, + "CAAGGA": 2113, + "CAAGGT": 2114, + "CAAGGC": 2115, + "CAAGGG": 2116, + "CATAAA": 2117, + "CATAAT": 2118, + "CATAAC": 2119, + "CATAAG": 2120, + "CATATA": 2121, + "CATATT": 2122, + "CATATC": 2123, + "CATATG": 2124, + "CATACA": 2125, + "CATACT": 2126, + "CATACC": 2127, + "CATACG": 2128, + "CATAGA": 2129, + "CATAGT": 2130, + "CATAGC": 2131, + "CATAGG": 2132, + "CATTAA": 2133, + "CATTAT": 2134, + "CATTAC": 2135, + "CATTAG": 2136, + "CATTTA": 2137, + "CATTTT": 2138, + "CATTTC": 2139, + "CATTTG": 2140, + "CATTCA": 2141, + "CATTCT": 2142, + "CATTCC": 2143, + "CATTCG": 2144, + "CATTGA": 2145, + "CATTGT": 2146, + "CATTGC": 2147, + "CATTGG": 2148, + "CATCAA": 2149, + "CATCAT": 2150, + "CATCAC": 2151, + "CATCAG": 2152, + "CATCTA": 2153, + "CATCTT": 2154, + "CATCTC": 2155, + "CATCTG": 2156, + "CATCCA": 2157, + "CATCCT": 2158, + "CATCCC": 2159, + "CATCCG": 2160, + "CATCGA": 2161, + "CATCGT": 2162, + "CATCGC": 2163, + "CATCGG": 2164, + "CATGAA": 2165, + "CATGAT": 2166, + "CATGAC": 2167, + "CATGAG": 2168, + "CATGTA": 2169, + "CATGTT": 2170, + "CATGTC": 2171, + "CATGTG": 2172, + "CATGCA": 2173, + "CATGCT": 2174, + "CATGCC": 2175, + "CATGCG": 2176, + "CATGGA": 2177, + "CATGGT": 2178, + "CATGGC": 2179, + "CATGGG": 2180, + "CACAAA": 2181, + "CACAAT": 2182, + "CACAAC": 2183, + "CACAAG": 2184, + "CACATA": 2185, + "CACATT": 2186, + "CACATC": 2187, + "CACATG": 2188, + "CACACA": 2189, + "CACACT": 2190, + "CACACC": 2191, + "CACACG": 2192, + "CACAGA": 2193, + "CACAGT": 2194, + "CACAGC": 2195, + "CACAGG": 2196, + "CACTAA": 2197, + "CACTAT": 2198, + "CACTAC": 2199, + "CACTAG": 2200, + "CACTTA": 2201, + "CACTTT": 2202, + "CACTTC": 2203, + "CACTTG": 2204, + "CACTCA": 2205, + "CACTCT": 2206, + "CACTCC": 2207, + "CACTCG": 2208, + "CACTGA": 2209, + "CACTGT": 2210, + "CACTGC": 2211, + "CACTGG": 2212, + "CACCAA": 2213, + "CACCAT": 2214, + "CACCAC": 2215, + "CACCAG": 2216, + "CACCTA": 2217, + "CACCTT": 2218, + "CACCTC": 2219, + "CACCTG": 2220, + "CACCCA": 2221, + "CACCCT": 2222, + "CACCCC": 2223, + "CACCCG": 2224, + "CACCGA": 2225, + "CACCGT": 2226, + "CACCGC": 2227, + "CACCGG": 2228, + "CACGAA": 2229, + "CACGAT": 2230, + "CACGAC": 2231, + "CACGAG": 2232, + "CACGTA": 2233, + "CACGTT": 2234, + "CACGTC": 2235, + "CACGTG": 2236, + "CACGCA": 2237, + "CACGCT": 2238, + "CACGCC": 2239, + "CACGCG": 2240, + "CACGGA": 2241, + "CACGGT": 2242, + "CACGGC": 2243, + "CACGGG": 2244, + "CAGAAA": 2245, + "CAGAAT": 2246, + "CAGAAC": 2247, + "CAGAAG": 2248, + "CAGATA": 2249, + "CAGATT": 2250, + "CAGATC": 2251, + "CAGATG": 2252, + "CAGACA": 2253, + "CAGACT": 2254, + "CAGACC": 2255, + "CAGACG": 2256, + "CAGAGA": 2257, + "CAGAGT": 2258, + "CAGAGC": 2259, + "CAGAGG": 2260, + "CAGTAA": 2261, + "CAGTAT": 2262, + "CAGTAC": 2263, + "CAGTAG": 2264, + "CAGTTA": 2265, + "CAGTTT": 2266, + "CAGTTC": 2267, + "CAGTTG": 2268, + "CAGTCA": 2269, + "CAGTCT": 2270, + "CAGTCC": 2271, + "CAGTCG": 2272, + "CAGTGA": 2273, + "CAGTGT": 2274, + "CAGTGC": 2275, + "CAGTGG": 2276, + "CAGCAA": 2277, + "CAGCAT": 2278, + "CAGCAC": 2279, + "CAGCAG": 2280, + "CAGCTA": 2281, + "CAGCTT": 2282, + "CAGCTC": 2283, + "CAGCTG": 2284, + "CAGCCA": 2285, + "CAGCCT": 2286, + "CAGCCC": 2287, + "CAGCCG": 2288, + "CAGCGA": 2289, + "CAGCGT": 2290, + "CAGCGC": 2291, + "CAGCGG": 2292, + "CAGGAA": 2293, + "CAGGAT": 2294, + "CAGGAC": 2295, + "CAGGAG": 2296, + "CAGGTA": 2297, + "CAGGTT": 2298, + "CAGGTC": 2299, + "CAGGTG": 2300, + "CAGGCA": 2301, + "CAGGCT": 2302, + "CAGGCC": 2303, + "CAGGCG": 2304, + "CAGGGA": 2305, + "CAGGGT": 2306, + "CAGGGC": 2307, + "CAGGGG": 2308, + "CTAAAA": 2309, + "CTAAAT": 2310, + "CTAAAC": 2311, + "CTAAAG": 2312, + "CTAATA": 2313, + "CTAATT": 2314, + "CTAATC": 2315, + "CTAATG": 2316, + "CTAACA": 2317, + "CTAACT": 2318, + "CTAACC": 2319, + "CTAACG": 2320, + "CTAAGA": 2321, + "CTAAGT": 2322, + "CTAAGC": 2323, + "CTAAGG": 2324, + "CTATAA": 2325, + "CTATAT": 2326, + "CTATAC": 2327, + "CTATAG": 2328, + "CTATTA": 2329, + "CTATTT": 2330, + "CTATTC": 2331, + "CTATTG": 2332, + "CTATCA": 2333, + "CTATCT": 2334, + "CTATCC": 2335, + "CTATCG": 2336, + "CTATGA": 2337, + "CTATGT": 2338, + "CTATGC": 2339, + "CTATGG": 2340, + "CTACAA": 2341, + "CTACAT": 2342, + "CTACAC": 2343, + "CTACAG": 2344, + "CTACTA": 2345, + "CTACTT": 2346, + "CTACTC": 2347, + "CTACTG": 2348, + "CTACCA": 2349, + "CTACCT": 2350, + "CTACCC": 2351, + "CTACCG": 2352, + "CTACGA": 2353, + "CTACGT": 2354, + "CTACGC": 2355, + "CTACGG": 2356, + "CTAGAA": 2357, + "CTAGAT": 2358, + "CTAGAC": 2359, + "CTAGAG": 2360, + "CTAGTA": 2361, + "CTAGTT": 2362, + "CTAGTC": 2363, + "CTAGTG": 2364, + "CTAGCA": 2365, + "CTAGCT": 2366, + "CTAGCC": 2367, + "CTAGCG": 2368, + "CTAGGA": 2369, + "CTAGGT": 2370, + "CTAGGC": 2371, + "CTAGGG": 2372, + "CTTAAA": 2373, + "CTTAAT": 2374, + "CTTAAC": 2375, + "CTTAAG": 2376, + "CTTATA": 2377, + "CTTATT": 2378, + "CTTATC": 2379, + "CTTATG": 2380, + "CTTACA": 2381, + "CTTACT": 2382, + "CTTACC": 2383, + "CTTACG": 2384, + "CTTAGA": 2385, + "CTTAGT": 2386, + "CTTAGC": 2387, + "CTTAGG": 2388, + "CTTTAA": 2389, + "CTTTAT": 2390, + "CTTTAC": 2391, + "CTTTAG": 2392, + "CTTTTA": 2393, + "CTTTTT": 2394, + "CTTTTC": 2395, + "CTTTTG": 2396, + "CTTTCA": 2397, + "CTTTCT": 2398, + "CTTTCC": 2399, + "CTTTCG": 2400, + "CTTTGA": 2401, + "CTTTGT": 2402, + "CTTTGC": 2403, + "CTTTGG": 2404, + "CTTCAA": 2405, + "CTTCAT": 2406, + "CTTCAC": 2407, + "CTTCAG": 2408, + "CTTCTA": 2409, + "CTTCTT": 2410, + "CTTCTC": 2411, + "CTTCTG": 2412, + "CTTCCA": 2413, + "CTTCCT": 2414, + "CTTCCC": 2415, + "CTTCCG": 2416, + "CTTCGA": 2417, + "CTTCGT": 2418, + "CTTCGC": 2419, + "CTTCGG": 2420, + "CTTGAA": 2421, + "CTTGAT": 2422, + "CTTGAC": 2423, + "CTTGAG": 2424, + "CTTGTA": 2425, + "CTTGTT": 2426, + "CTTGTC": 2427, + "CTTGTG": 2428, + "CTTGCA": 2429, + "CTTGCT": 2430, + "CTTGCC": 2431, + "CTTGCG": 2432, + "CTTGGA": 2433, + "CTTGGT": 2434, + "CTTGGC": 2435, + "CTTGGG": 2436, + "CTCAAA": 2437, + "CTCAAT": 2438, + "CTCAAC": 2439, + "CTCAAG": 2440, + "CTCATA": 2441, + "CTCATT": 2442, + "CTCATC": 2443, + "CTCATG": 2444, + "CTCACA": 2445, + "CTCACT": 2446, + "CTCACC": 2447, + "CTCACG": 2448, + "CTCAGA": 2449, + "CTCAGT": 2450, + "CTCAGC": 2451, + "CTCAGG": 2452, + "CTCTAA": 2453, + "CTCTAT": 2454, + "CTCTAC": 2455, + "CTCTAG": 2456, + "CTCTTA": 2457, + "CTCTTT": 2458, + "CTCTTC": 2459, + "CTCTTG": 2460, + "CTCTCA": 2461, + "CTCTCT": 2462, + "CTCTCC": 2463, + "CTCTCG": 2464, + "CTCTGA": 2465, + "CTCTGT": 2466, + "CTCTGC": 2467, + "CTCTGG": 2468, + "CTCCAA": 2469, + "CTCCAT": 2470, + "CTCCAC": 2471, + "CTCCAG": 2472, + "CTCCTA": 2473, + "CTCCTT": 2474, + "CTCCTC": 2475, + "CTCCTG": 2476, + "CTCCCA": 2477, + "CTCCCT": 2478, + "CTCCCC": 2479, + "CTCCCG": 2480, + "CTCCGA": 2481, + "CTCCGT": 2482, + "CTCCGC": 2483, + "CTCCGG": 2484, + "CTCGAA": 2485, + "CTCGAT": 2486, + "CTCGAC": 2487, + "CTCGAG": 2488, + "CTCGTA": 2489, + "CTCGTT": 2490, + "CTCGTC": 2491, + "CTCGTG": 2492, + "CTCGCA": 2493, + "CTCGCT": 2494, + "CTCGCC": 2495, + "CTCGCG": 2496, + "CTCGGA": 2497, + "CTCGGT": 2498, + "CTCGGC": 2499, + "CTCGGG": 2500, + "CTGAAA": 2501, + "CTGAAT": 2502, + "CTGAAC": 2503, + "CTGAAG": 2504, + "CTGATA": 2505, + "CTGATT": 2506, + "CTGATC": 2507, + "CTGATG": 2508, + "CTGACA": 2509, + "CTGACT": 2510, + "CTGACC": 2511, + "CTGACG": 2512, + "CTGAGA": 2513, + "CTGAGT": 2514, + "CTGAGC": 2515, + "CTGAGG": 2516, + "CTGTAA": 2517, + "CTGTAT": 2518, + "CTGTAC": 2519, + "CTGTAG": 2520, + "CTGTTA": 2521, + "CTGTTT": 2522, + "CTGTTC": 2523, + "CTGTTG": 2524, + "CTGTCA": 2525, + "CTGTCT": 2526, + "CTGTCC": 2527, + "CTGTCG": 2528, + "CTGTGA": 2529, + "CTGTGT": 2530, + "CTGTGC": 2531, + "CTGTGG": 2532, + "CTGCAA": 2533, + "CTGCAT": 2534, + "CTGCAC": 2535, + "CTGCAG": 2536, + "CTGCTA": 2537, + "CTGCTT": 2538, + "CTGCTC": 2539, + "CTGCTG": 2540, + "CTGCCA": 2541, + "CTGCCT": 2542, + "CTGCCC": 2543, + "CTGCCG": 2544, + "CTGCGA": 2545, + "CTGCGT": 2546, + "CTGCGC": 2547, + "CTGCGG": 2548, + "CTGGAA": 2549, + "CTGGAT": 2550, + "CTGGAC": 2551, + "CTGGAG": 2552, + "CTGGTA": 2553, + "CTGGTT": 2554, + "CTGGTC": 2555, + "CTGGTG": 2556, + "CTGGCA": 2557, + "CTGGCT": 2558, + "CTGGCC": 2559, + "CTGGCG": 2560, + "CTGGGA": 2561, + "CTGGGT": 2562, + "CTGGGC": 2563, + "CTGGGG": 2564, + "CCAAAA": 2565, + "CCAAAT": 2566, + "CCAAAC": 2567, + "CCAAAG": 2568, + "CCAATA": 2569, + "CCAATT": 2570, + "CCAATC": 2571, + "CCAATG": 2572, + "CCAACA": 2573, + "CCAACT": 2574, + "CCAACC": 2575, + "CCAACG": 2576, + "CCAAGA": 2577, + "CCAAGT": 2578, + "CCAAGC": 2579, + "CCAAGG": 2580, + "CCATAA": 2581, + "CCATAT": 2582, + "CCATAC": 2583, + "CCATAG": 2584, + "CCATTA": 2585, + "CCATTT": 2586, + "CCATTC": 2587, + "CCATTG": 2588, + "CCATCA": 2589, + "CCATCT": 2590, + "CCATCC": 2591, + "CCATCG": 2592, + "CCATGA": 2593, + "CCATGT": 2594, + "CCATGC": 2595, + "CCATGG": 2596, + "CCACAA": 2597, + "CCACAT": 2598, + "CCACAC": 2599, + "CCACAG": 2600, + "CCACTA": 2601, + "CCACTT": 2602, + "CCACTC": 2603, + "CCACTG": 2604, + "CCACCA": 2605, + "CCACCT": 2606, + "CCACCC": 2607, + "CCACCG": 2608, + "CCACGA": 2609, + "CCACGT": 2610, + "CCACGC": 2611, + "CCACGG": 2612, + "CCAGAA": 2613, + "CCAGAT": 2614, + "CCAGAC": 2615, + "CCAGAG": 2616, + "CCAGTA": 2617, + "CCAGTT": 2618, + "CCAGTC": 2619, + "CCAGTG": 2620, + "CCAGCA": 2621, + "CCAGCT": 2622, + "CCAGCC": 2623, + "CCAGCG": 2624, + "CCAGGA": 2625, + "CCAGGT": 2626, + "CCAGGC": 2627, + "CCAGGG": 2628, + "CCTAAA": 2629, + "CCTAAT": 2630, + "CCTAAC": 2631, + "CCTAAG": 2632, + "CCTATA": 2633, + "CCTATT": 2634, + "CCTATC": 2635, + "CCTATG": 2636, + "CCTACA": 2637, + "CCTACT": 2638, + "CCTACC": 2639, + "CCTACG": 2640, + "CCTAGA": 2641, + "CCTAGT": 2642, + "CCTAGC": 2643, + "CCTAGG": 2644, + "CCTTAA": 2645, + "CCTTAT": 2646, + "CCTTAC": 2647, + "CCTTAG": 2648, + "CCTTTA": 2649, + "CCTTTT": 2650, + "CCTTTC": 2651, + "CCTTTG": 2652, + "CCTTCA": 2653, + "CCTTCT": 2654, + "CCTTCC": 2655, + "CCTTCG": 2656, + "CCTTGA": 2657, + "CCTTGT": 2658, + "CCTTGC": 2659, + "CCTTGG": 2660, + "CCTCAA": 2661, + "CCTCAT": 2662, + "CCTCAC": 2663, + "CCTCAG": 2664, + "CCTCTA": 2665, + "CCTCTT": 2666, + "CCTCTC": 2667, + "CCTCTG": 2668, + "CCTCCA": 2669, + "CCTCCT": 2670, + "CCTCCC": 2671, + "CCTCCG": 2672, + "CCTCGA": 2673, + "CCTCGT": 2674, + "CCTCGC": 2675, + "CCTCGG": 2676, + "CCTGAA": 2677, + "CCTGAT": 2678, + "CCTGAC": 2679, + "CCTGAG": 2680, + "CCTGTA": 2681, + "CCTGTT": 2682, + "CCTGTC": 2683, + "CCTGTG": 2684, + "CCTGCA": 2685, + "CCTGCT": 2686, + "CCTGCC": 2687, + "CCTGCG": 2688, + "CCTGGA": 2689, + "CCTGGT": 2690, + "CCTGGC": 2691, + "CCTGGG": 2692, + "CCCAAA": 2693, + "CCCAAT": 2694, + "CCCAAC": 2695, + "CCCAAG": 2696, + "CCCATA": 2697, + "CCCATT": 2698, + "CCCATC": 2699, + "CCCATG": 2700, + "CCCACA": 2701, + "CCCACT": 2702, + "CCCACC": 2703, + "CCCACG": 2704, + "CCCAGA": 2705, + "CCCAGT": 2706, + "CCCAGC": 2707, + "CCCAGG": 2708, + "CCCTAA": 2709, + "CCCTAT": 2710, + "CCCTAC": 2711, + "CCCTAG": 2712, + "CCCTTA": 2713, + "CCCTTT": 2714, + "CCCTTC": 2715, + "CCCTTG": 2716, + "CCCTCA": 2717, + "CCCTCT": 2718, + "CCCTCC": 2719, + "CCCTCG": 2720, + "CCCTGA": 2721, + "CCCTGT": 2722, + "CCCTGC": 2723, + "CCCTGG": 2724, + "CCCCAA": 2725, + "CCCCAT": 2726, + "CCCCAC": 2727, + "CCCCAG": 2728, + "CCCCTA": 2729, + "CCCCTT": 2730, + "CCCCTC": 2731, + "CCCCTG": 2732, + "CCCCCA": 2733, + "CCCCCT": 2734, + "CCCCCC": 2735, + "CCCCCG": 2736, + "CCCCGA": 2737, + "CCCCGT": 2738, + "CCCCGC": 2739, + "CCCCGG": 2740, + "CCCGAA": 2741, + "CCCGAT": 2742, + "CCCGAC": 2743, + "CCCGAG": 2744, + "CCCGTA": 2745, + "CCCGTT": 2746, + "CCCGTC": 2747, + "CCCGTG": 2748, + "CCCGCA": 2749, + "CCCGCT": 2750, + "CCCGCC": 2751, + "CCCGCG": 2752, + "CCCGGA": 2753, + "CCCGGT": 2754, + "CCCGGC": 2755, + "CCCGGG": 2756, + "CCGAAA": 2757, + "CCGAAT": 2758, + "CCGAAC": 2759, + "CCGAAG": 2760, + "CCGATA": 2761, + "CCGATT": 2762, + "CCGATC": 2763, + "CCGATG": 2764, + "CCGACA": 2765, + "CCGACT": 2766, + "CCGACC": 2767, + "CCGACG": 2768, + "CCGAGA": 2769, + "CCGAGT": 2770, + "CCGAGC": 2771, + "CCGAGG": 2772, + "CCGTAA": 2773, + "CCGTAT": 2774, + "CCGTAC": 2775, + "CCGTAG": 2776, + "CCGTTA": 2777, + "CCGTTT": 2778, + "CCGTTC": 2779, + "CCGTTG": 2780, + "CCGTCA": 2781, + "CCGTCT": 2782, + "CCGTCC": 2783, + "CCGTCG": 2784, + "CCGTGA": 2785, + "CCGTGT": 2786, + "CCGTGC": 2787, + "CCGTGG": 2788, + "CCGCAA": 2789, + "CCGCAT": 2790, + "CCGCAC": 2791, + "CCGCAG": 2792, + "CCGCTA": 2793, + "CCGCTT": 2794, + "CCGCTC": 2795, + "CCGCTG": 2796, + "CCGCCA": 2797, + "CCGCCT": 2798, + "CCGCCC": 2799, + "CCGCCG": 2800, + "CCGCGA": 2801, + "CCGCGT": 2802, + "CCGCGC": 2803, + "CCGCGG": 2804, + "CCGGAA": 2805, + "CCGGAT": 2806, + "CCGGAC": 2807, + "CCGGAG": 2808, + "CCGGTA": 2809, + "CCGGTT": 2810, + "CCGGTC": 2811, + "CCGGTG": 2812, + "CCGGCA": 2813, + "CCGGCT": 2814, + "CCGGCC": 2815, + "CCGGCG": 2816, + "CCGGGA": 2817, + "CCGGGT": 2818, + "CCGGGC": 2819, + "CCGGGG": 2820, + "CGAAAA": 2821, + "CGAAAT": 2822, + "CGAAAC": 2823, + "CGAAAG": 2824, + "CGAATA": 2825, + "CGAATT": 2826, + "CGAATC": 2827, + "CGAATG": 2828, + "CGAACA": 2829, + "CGAACT": 2830, + "CGAACC": 2831, + "CGAACG": 2832, + "CGAAGA": 2833, + "CGAAGT": 2834, + "CGAAGC": 2835, + "CGAAGG": 2836, + "CGATAA": 2837, + "CGATAT": 2838, + "CGATAC": 2839, + "CGATAG": 2840, + "CGATTA": 2841, + "CGATTT": 2842, + "CGATTC": 2843, + "CGATTG": 2844, + "CGATCA": 2845, + "CGATCT": 2846, + "CGATCC": 2847, + "CGATCG": 2848, + "CGATGA": 2849, + "CGATGT": 2850, + "CGATGC": 2851, + "CGATGG": 2852, + "CGACAA": 2853, + "CGACAT": 2854, + "CGACAC": 2855, + "CGACAG": 2856, + "CGACTA": 2857, + "CGACTT": 2858, + "CGACTC": 2859, + "CGACTG": 2860, + "CGACCA": 2861, + "CGACCT": 2862, + "CGACCC": 2863, + "CGACCG": 2864, + "CGACGA": 2865, + "CGACGT": 2866, + "CGACGC": 2867, + "CGACGG": 2868, + "CGAGAA": 2869, + "CGAGAT": 2870, + "CGAGAC": 2871, + "CGAGAG": 2872, + "CGAGTA": 2873, + "CGAGTT": 2874, + "CGAGTC": 2875, + "CGAGTG": 2876, + "CGAGCA": 2877, + "CGAGCT": 2878, + "CGAGCC": 2879, + "CGAGCG": 2880, + "CGAGGA": 2881, + "CGAGGT": 2882, + "CGAGGC": 2883, + "CGAGGG": 2884, + "CGTAAA": 2885, + "CGTAAT": 2886, + "CGTAAC": 2887, + "CGTAAG": 2888, + "CGTATA": 2889, + "CGTATT": 2890, + "CGTATC": 2891, + "CGTATG": 2892, + "CGTACA": 2893, + "CGTACT": 2894, + "CGTACC": 2895, + "CGTACG": 2896, + "CGTAGA": 2897, + "CGTAGT": 2898, + "CGTAGC": 2899, + "CGTAGG": 2900, + "CGTTAA": 2901, + "CGTTAT": 2902, + "CGTTAC": 2903, + "CGTTAG": 2904, + "CGTTTA": 2905, + "CGTTTT": 2906, + "CGTTTC": 2907, + "CGTTTG": 2908, + "CGTTCA": 2909, + "CGTTCT": 2910, + "CGTTCC": 2911, + "CGTTCG": 2912, + "CGTTGA": 2913, + "CGTTGT": 2914, + "CGTTGC": 2915, + "CGTTGG": 2916, + "CGTCAA": 2917, + "CGTCAT": 2918, + "CGTCAC": 2919, + "CGTCAG": 2920, + "CGTCTA": 2921, + "CGTCTT": 2922, + "CGTCTC": 2923, + "CGTCTG": 2924, + "CGTCCA": 2925, + "CGTCCT": 2926, + "CGTCCC": 2927, + "CGTCCG": 2928, + "CGTCGA": 2929, + "CGTCGT": 2930, + "CGTCGC": 2931, + "CGTCGG": 2932, + "CGTGAA": 2933, + "CGTGAT": 2934, + "CGTGAC": 2935, + "CGTGAG": 2936, + "CGTGTA": 2937, + "CGTGTT": 2938, + "CGTGTC": 2939, + "CGTGTG": 2940, + "CGTGCA": 2941, + "CGTGCT": 2942, + "CGTGCC": 2943, + "CGTGCG": 2944, + "CGTGGA": 2945, + "CGTGGT": 2946, + "CGTGGC": 2947, + "CGTGGG": 2948, + "CGCAAA": 2949, + "CGCAAT": 2950, + "CGCAAC": 2951, + "CGCAAG": 2952, + "CGCATA": 2953, + "CGCATT": 2954, + "CGCATC": 2955, + "CGCATG": 2956, + "CGCACA": 2957, + "CGCACT": 2958, + "CGCACC": 2959, + "CGCACG": 2960, + "CGCAGA": 2961, + "CGCAGT": 2962, + "CGCAGC": 2963, + "CGCAGG": 2964, + "CGCTAA": 2965, + "CGCTAT": 2966, + "CGCTAC": 2967, + "CGCTAG": 2968, + "CGCTTA": 2969, + "CGCTTT": 2970, + "CGCTTC": 2971, + "CGCTTG": 2972, + "CGCTCA": 2973, + "CGCTCT": 2974, + "CGCTCC": 2975, + "CGCTCG": 2976, + "CGCTGA": 2977, + "CGCTGT": 2978, + "CGCTGC": 2979, + "CGCTGG": 2980, + "CGCCAA": 2981, + "CGCCAT": 2982, + "CGCCAC": 2983, + "CGCCAG": 2984, + "CGCCTA": 2985, + "CGCCTT": 2986, + "CGCCTC": 2987, + "CGCCTG": 2988, + "CGCCCA": 2989, + "CGCCCT": 2990, + "CGCCCC": 2991, + "CGCCCG": 2992, + "CGCCGA": 2993, + "CGCCGT": 2994, + "CGCCGC": 2995, + "CGCCGG": 2996, + "CGCGAA": 2997, + "CGCGAT": 2998, + "CGCGAC": 2999, + "CGCGAG": 3000, + "CGCGTA": 3001, + "CGCGTT": 3002, + "CGCGTC": 3003, + "CGCGTG": 3004, + "CGCGCA": 3005, + "CGCGCT": 3006, + "CGCGCC": 3007, + "CGCGCG": 3008, + "CGCGGA": 3009, + "CGCGGT": 3010, + "CGCGGC": 3011, + "CGCGGG": 3012, + "CGGAAA": 3013, + "CGGAAT": 3014, + "CGGAAC": 3015, + "CGGAAG": 3016, + "CGGATA": 3017, + "CGGATT": 3018, + "CGGATC": 3019, + "CGGATG": 3020, + "CGGACA": 3021, + "CGGACT": 3022, + "CGGACC": 3023, + "CGGACG": 3024, + "CGGAGA": 3025, + "CGGAGT": 3026, + "CGGAGC": 3027, + "CGGAGG": 3028, + "CGGTAA": 3029, + "CGGTAT": 3030, + "CGGTAC": 3031, + "CGGTAG": 3032, + "CGGTTA": 3033, + "CGGTTT": 3034, + "CGGTTC": 3035, + "CGGTTG": 3036, + "CGGTCA": 3037, + "CGGTCT": 3038, + "CGGTCC": 3039, + "CGGTCG": 3040, + "CGGTGA": 3041, + "CGGTGT": 3042, + "CGGTGC": 3043, + "CGGTGG": 3044, + "CGGCAA": 3045, + "CGGCAT": 3046, + "CGGCAC": 3047, + "CGGCAG": 3048, + "CGGCTA": 3049, + "CGGCTT": 3050, + "CGGCTC": 3051, + "CGGCTG": 3052, + "CGGCCA": 3053, + "CGGCCT": 3054, + "CGGCCC": 3055, + "CGGCCG": 3056, + "CGGCGA": 3057, + "CGGCGT": 3058, + "CGGCGC": 3059, + "CGGCGG": 3060, + "CGGGAA": 3061, + "CGGGAT": 3062, + "CGGGAC": 3063, + "CGGGAG": 3064, + "CGGGTA": 3065, + "CGGGTT": 3066, + "CGGGTC": 3067, + "CGGGTG": 3068, + "CGGGCA": 3069, + "CGGGCT": 3070, + "CGGGCC": 3071, + "CGGGCG": 3072, + "CGGGGA": 3073, + "CGGGGT": 3074, + "CGGGGC": 3075, + "CGGGGG": 3076, + "GAAAAA": 3077, + "GAAAAT": 3078, + "GAAAAC": 3079, + "GAAAAG": 3080, + "GAAATA": 3081, + "GAAATT": 3082, + "GAAATC": 3083, + "GAAATG": 3084, + "GAAACA": 3085, + "GAAACT": 3086, + "GAAACC": 3087, + "GAAACG": 3088, + "GAAAGA": 3089, + "GAAAGT": 3090, + "GAAAGC": 3091, + "GAAAGG": 3092, + "GAATAA": 3093, + "GAATAT": 3094, + "GAATAC": 3095, + "GAATAG": 3096, + "GAATTA": 3097, + "GAATTT": 3098, + "GAATTC": 3099, + "GAATTG": 3100, + "GAATCA": 3101, + "GAATCT": 3102, + "GAATCC": 3103, + "GAATCG": 3104, + "GAATGA": 3105, + "GAATGT": 3106, + "GAATGC": 3107, + "GAATGG": 3108, + "GAACAA": 3109, + "GAACAT": 3110, + "GAACAC": 3111, + "GAACAG": 3112, + "GAACTA": 3113, + "GAACTT": 3114, + "GAACTC": 3115, + "GAACTG": 3116, + "GAACCA": 3117, + "GAACCT": 3118, + "GAACCC": 3119, + "GAACCG": 3120, + "GAACGA": 3121, + "GAACGT": 3122, + "GAACGC": 3123, + "GAACGG": 3124, + "GAAGAA": 3125, + "GAAGAT": 3126, + "GAAGAC": 3127, + "GAAGAG": 3128, + "GAAGTA": 3129, + "GAAGTT": 3130, + "GAAGTC": 3131, + "GAAGTG": 3132, + "GAAGCA": 3133, + "GAAGCT": 3134, + "GAAGCC": 3135, + "GAAGCG": 3136, + "GAAGGA": 3137, + "GAAGGT": 3138, + "GAAGGC": 3139, + "GAAGGG": 3140, + "GATAAA": 3141, + "GATAAT": 3142, + "GATAAC": 3143, + "GATAAG": 3144, + "GATATA": 3145, + "GATATT": 3146, + "GATATC": 3147, + "GATATG": 3148, + "GATACA": 3149, + "GATACT": 3150, + "GATACC": 3151, + "GATACG": 3152, + "GATAGA": 3153, + "GATAGT": 3154, + "GATAGC": 3155, + "GATAGG": 3156, + "GATTAA": 3157, + "GATTAT": 3158, + "GATTAC": 3159, + "GATTAG": 3160, + "GATTTA": 3161, + "GATTTT": 3162, + "GATTTC": 3163, + "GATTTG": 3164, + "GATTCA": 3165, + "GATTCT": 3166, + "GATTCC": 3167, + "GATTCG": 3168, + "GATTGA": 3169, + "GATTGT": 3170, + "GATTGC": 3171, + "GATTGG": 3172, + "GATCAA": 3173, + "GATCAT": 3174, + "GATCAC": 3175, + "GATCAG": 3176, + "GATCTA": 3177, + "GATCTT": 3178, + "GATCTC": 3179, + "GATCTG": 3180, + "GATCCA": 3181, + "GATCCT": 3182, + "GATCCC": 3183, + "GATCCG": 3184, + "GATCGA": 3185, + "GATCGT": 3186, + "GATCGC": 3187, + "GATCGG": 3188, + "GATGAA": 3189, + "GATGAT": 3190, + "GATGAC": 3191, + "GATGAG": 3192, + "GATGTA": 3193, + "GATGTT": 3194, + "GATGTC": 3195, + "GATGTG": 3196, + "GATGCA": 3197, + "GATGCT": 3198, + "GATGCC": 3199, + "GATGCG": 3200, + "GATGGA": 3201, + "GATGGT": 3202, + "GATGGC": 3203, + "GATGGG": 3204, + "GACAAA": 3205, + "GACAAT": 3206, + "GACAAC": 3207, + "GACAAG": 3208, + "GACATA": 3209, + "GACATT": 3210, + "GACATC": 3211, + "GACATG": 3212, + "GACACA": 3213, + "GACACT": 3214, + "GACACC": 3215, + "GACACG": 3216, + "GACAGA": 3217, + "GACAGT": 3218, + "GACAGC": 3219, + "GACAGG": 3220, + "GACTAA": 3221, + "GACTAT": 3222, + "GACTAC": 3223, + "GACTAG": 3224, + "GACTTA": 3225, + "GACTTT": 3226, + "GACTTC": 3227, + "GACTTG": 3228, + "GACTCA": 3229, + "GACTCT": 3230, + "GACTCC": 3231, + "GACTCG": 3232, + "GACTGA": 3233, + "GACTGT": 3234, + "GACTGC": 3235, + "GACTGG": 3236, + "GACCAA": 3237, + "GACCAT": 3238, + "GACCAC": 3239, + "GACCAG": 3240, + "GACCTA": 3241, + "GACCTT": 3242, + "GACCTC": 3243, + "GACCTG": 3244, + "GACCCA": 3245, + "GACCCT": 3246, + "GACCCC": 3247, + "GACCCG": 3248, + "GACCGA": 3249, + "GACCGT": 3250, + "GACCGC": 3251, + "GACCGG": 3252, + "GACGAA": 3253, + "GACGAT": 3254, + "GACGAC": 3255, + "GACGAG": 3256, + "GACGTA": 3257, + "GACGTT": 3258, + "GACGTC": 3259, + "GACGTG": 3260, + "GACGCA": 3261, + "GACGCT": 3262, + "GACGCC": 3263, + "GACGCG": 3264, + "GACGGA": 3265, + "GACGGT": 3266, + "GACGGC": 3267, + "GACGGG": 3268, + "GAGAAA": 3269, + "GAGAAT": 3270, + "GAGAAC": 3271, + "GAGAAG": 3272, + "GAGATA": 3273, + "GAGATT": 3274, + "GAGATC": 3275, + "GAGATG": 3276, + "GAGACA": 3277, + "GAGACT": 3278, + "GAGACC": 3279, + "GAGACG": 3280, + "GAGAGA": 3281, + "GAGAGT": 3282, + "GAGAGC": 3283, + "GAGAGG": 3284, + "GAGTAA": 3285, + "GAGTAT": 3286, + "GAGTAC": 3287, + "GAGTAG": 3288, + "GAGTTA": 3289, + "GAGTTT": 3290, + "GAGTTC": 3291, + "GAGTTG": 3292, + "GAGTCA": 3293, + "GAGTCT": 3294, + "GAGTCC": 3295, + "GAGTCG": 3296, + "GAGTGA": 3297, + "GAGTGT": 3298, + "GAGTGC": 3299, + "GAGTGG": 3300, + "GAGCAA": 3301, + "GAGCAT": 3302, + "GAGCAC": 3303, + "GAGCAG": 3304, + "GAGCTA": 3305, + "GAGCTT": 3306, + "GAGCTC": 3307, + "GAGCTG": 3308, + "GAGCCA": 3309, + "GAGCCT": 3310, + "GAGCCC": 3311, + "GAGCCG": 3312, + "GAGCGA": 3313, + "GAGCGT": 3314, + "GAGCGC": 3315, + "GAGCGG": 3316, + "GAGGAA": 3317, + "GAGGAT": 3318, + "GAGGAC": 3319, + "GAGGAG": 3320, + "GAGGTA": 3321, + "GAGGTT": 3322, + "GAGGTC": 3323, + "GAGGTG": 3324, + "GAGGCA": 3325, + "GAGGCT": 3326, + "GAGGCC": 3327, + "GAGGCG": 3328, + "GAGGGA": 3329, + "GAGGGT": 3330, + "GAGGGC": 3331, + "GAGGGG": 3332, + "GTAAAA": 3333, + "GTAAAT": 3334, + "GTAAAC": 3335, + "GTAAAG": 3336, + "GTAATA": 3337, + "GTAATT": 3338, + "GTAATC": 3339, + "GTAATG": 3340, + "GTAACA": 3341, + "GTAACT": 3342, + "GTAACC": 3343, + "GTAACG": 3344, + "GTAAGA": 3345, + "GTAAGT": 3346, + "GTAAGC": 3347, + "GTAAGG": 3348, + "GTATAA": 3349, + "GTATAT": 3350, + "GTATAC": 3351, + "GTATAG": 3352, + "GTATTA": 3353, + "GTATTT": 3354, + "GTATTC": 3355, + "GTATTG": 3356, + "GTATCA": 3357, + "GTATCT": 3358, + "GTATCC": 3359, + "GTATCG": 3360, + "GTATGA": 3361, + "GTATGT": 3362, + "GTATGC": 3363, + "GTATGG": 3364, + "GTACAA": 3365, + "GTACAT": 3366, + "GTACAC": 3367, + "GTACAG": 3368, + "GTACTA": 3369, + "GTACTT": 3370, + "GTACTC": 3371, + "GTACTG": 3372, + "GTACCA": 3373, + "GTACCT": 3374, + "GTACCC": 3375, + "GTACCG": 3376, + "GTACGA": 3377, + "GTACGT": 3378, + "GTACGC": 3379, + "GTACGG": 3380, + "GTAGAA": 3381, + "GTAGAT": 3382, + "GTAGAC": 3383, + "GTAGAG": 3384, + "GTAGTA": 3385, + "GTAGTT": 3386, + "GTAGTC": 3387, + "GTAGTG": 3388, + "GTAGCA": 3389, + "GTAGCT": 3390, + "GTAGCC": 3391, + "GTAGCG": 3392, + "GTAGGA": 3393, + "GTAGGT": 3394, + "GTAGGC": 3395, + "GTAGGG": 3396, + "GTTAAA": 3397, + "GTTAAT": 3398, + "GTTAAC": 3399, + "GTTAAG": 3400, + "GTTATA": 3401, + "GTTATT": 3402, + "GTTATC": 3403, + "GTTATG": 3404, + "GTTACA": 3405, + "GTTACT": 3406, + "GTTACC": 3407, + "GTTACG": 3408, + "GTTAGA": 3409, + "GTTAGT": 3410, + "GTTAGC": 3411, + "GTTAGG": 3412, + "GTTTAA": 3413, + "GTTTAT": 3414, + "GTTTAC": 3415, + "GTTTAG": 3416, + "GTTTTA": 3417, + "GTTTTT": 3418, + "GTTTTC": 3419, + "GTTTTG": 3420, + "GTTTCA": 3421, + "GTTTCT": 3422, + "GTTTCC": 3423, + "GTTTCG": 3424, + "GTTTGA": 3425, + "GTTTGT": 3426, + "GTTTGC": 3427, + "GTTTGG": 3428, + "GTTCAA": 3429, + "GTTCAT": 3430, + "GTTCAC": 3431, + "GTTCAG": 3432, + "GTTCTA": 3433, + "GTTCTT": 3434, + "GTTCTC": 3435, + "GTTCTG": 3436, + "GTTCCA": 3437, + "GTTCCT": 3438, + "GTTCCC": 3439, + "GTTCCG": 3440, + "GTTCGA": 3441, + "GTTCGT": 3442, + "GTTCGC": 3443, + "GTTCGG": 3444, + "GTTGAA": 3445, + "GTTGAT": 3446, + "GTTGAC": 3447, + "GTTGAG": 3448, + "GTTGTA": 3449, + "GTTGTT": 3450, + "GTTGTC": 3451, + "GTTGTG": 3452, + "GTTGCA": 3453, + "GTTGCT": 3454, + "GTTGCC": 3455, + "GTTGCG": 3456, + "GTTGGA": 3457, + "GTTGGT": 3458, + "GTTGGC": 3459, + "GTTGGG": 3460, + "GTCAAA": 3461, + "GTCAAT": 3462, + "GTCAAC": 3463, + "GTCAAG": 3464, + "GTCATA": 3465, + "GTCATT": 3466, + "GTCATC": 3467, + "GTCATG": 3468, + "GTCACA": 3469, + "GTCACT": 3470, + "GTCACC": 3471, + "GTCACG": 3472, + "GTCAGA": 3473, + "GTCAGT": 3474, + "GTCAGC": 3475, + "GTCAGG": 3476, + "GTCTAA": 3477, + "GTCTAT": 3478, + "GTCTAC": 3479, + "GTCTAG": 3480, + "GTCTTA": 3481, + "GTCTTT": 3482, + "GTCTTC": 3483, + "GTCTTG": 3484, + "GTCTCA": 3485, + "GTCTCT": 3486, + "GTCTCC": 3487, + "GTCTCG": 3488, + "GTCTGA": 3489, + "GTCTGT": 3490, + "GTCTGC": 3491, + "GTCTGG": 3492, + "GTCCAA": 3493, + "GTCCAT": 3494, + "GTCCAC": 3495, + "GTCCAG": 3496, + "GTCCTA": 3497, + "GTCCTT": 3498, + "GTCCTC": 3499, + "GTCCTG": 3500, + "GTCCCA": 3501, + "GTCCCT": 3502, + "GTCCCC": 3503, + "GTCCCG": 3504, + "GTCCGA": 3505, + "GTCCGT": 3506, + "GTCCGC": 3507, + "GTCCGG": 3508, + "GTCGAA": 3509, + "GTCGAT": 3510, + "GTCGAC": 3511, + "GTCGAG": 3512, + "GTCGTA": 3513, + "GTCGTT": 3514, + "GTCGTC": 3515, + "GTCGTG": 3516, + "GTCGCA": 3517, + "GTCGCT": 3518, + "GTCGCC": 3519, + "GTCGCG": 3520, + "GTCGGA": 3521, + "GTCGGT": 3522, + "GTCGGC": 3523, + "GTCGGG": 3524, + "GTGAAA": 3525, + "GTGAAT": 3526, + "GTGAAC": 3527, + "GTGAAG": 3528, + "GTGATA": 3529, + "GTGATT": 3530, + "GTGATC": 3531, + "GTGATG": 3532, + "GTGACA": 3533, + "GTGACT": 3534, + "GTGACC": 3535, + "GTGACG": 3536, + "GTGAGA": 3537, + "GTGAGT": 3538, + "GTGAGC": 3539, + "GTGAGG": 3540, + "GTGTAA": 3541, + "GTGTAT": 3542, + "GTGTAC": 3543, + "GTGTAG": 3544, + "GTGTTA": 3545, + "GTGTTT": 3546, + "GTGTTC": 3547, + "GTGTTG": 3548, + "GTGTCA": 3549, + "GTGTCT": 3550, + "GTGTCC": 3551, + "GTGTCG": 3552, + "GTGTGA": 3553, + "GTGTGT": 3554, + "GTGTGC": 3555, + "GTGTGG": 3556, + "GTGCAA": 3557, + "GTGCAT": 3558, + "GTGCAC": 3559, + "GTGCAG": 3560, + "GTGCTA": 3561, + "GTGCTT": 3562, + "GTGCTC": 3563, + "GTGCTG": 3564, + "GTGCCA": 3565, + "GTGCCT": 3566, + "GTGCCC": 3567, + "GTGCCG": 3568, + "GTGCGA": 3569, + "GTGCGT": 3570, + "GTGCGC": 3571, + "GTGCGG": 3572, + "GTGGAA": 3573, + "GTGGAT": 3574, + "GTGGAC": 3575, + "GTGGAG": 3576, + "GTGGTA": 3577, + "GTGGTT": 3578, + "GTGGTC": 3579, + "GTGGTG": 3580, + "GTGGCA": 3581, + "GTGGCT": 3582, + "GTGGCC": 3583, + "GTGGCG": 3584, + "GTGGGA": 3585, + "GTGGGT": 3586, + "GTGGGC": 3587, + "GTGGGG": 3588, + "GCAAAA": 3589, + "GCAAAT": 3590, + "GCAAAC": 3591, + "GCAAAG": 3592, + "GCAATA": 3593, + "GCAATT": 3594, + "GCAATC": 3595, + "GCAATG": 3596, + "GCAACA": 3597, + "GCAACT": 3598, + "GCAACC": 3599, + "GCAACG": 3600, + "GCAAGA": 3601, + "GCAAGT": 3602, + "GCAAGC": 3603, + "GCAAGG": 3604, + "GCATAA": 3605, + "GCATAT": 3606, + "GCATAC": 3607, + "GCATAG": 3608, + "GCATTA": 3609, + "GCATTT": 3610, + "GCATTC": 3611, + "GCATTG": 3612, + "GCATCA": 3613, + "GCATCT": 3614, + "GCATCC": 3615, + "GCATCG": 3616, + "GCATGA": 3617, + "GCATGT": 3618, + "GCATGC": 3619, + "GCATGG": 3620, + "GCACAA": 3621, + "GCACAT": 3622, + "GCACAC": 3623, + "GCACAG": 3624, + "GCACTA": 3625, + "GCACTT": 3626, + "GCACTC": 3627, + "GCACTG": 3628, + "GCACCA": 3629, + "GCACCT": 3630, + "GCACCC": 3631, + "GCACCG": 3632, + "GCACGA": 3633, + "GCACGT": 3634, + "GCACGC": 3635, + "GCACGG": 3636, + "GCAGAA": 3637, + "GCAGAT": 3638, + "GCAGAC": 3639, + "GCAGAG": 3640, + "GCAGTA": 3641, + "GCAGTT": 3642, + "GCAGTC": 3643, + "GCAGTG": 3644, + "GCAGCA": 3645, + "GCAGCT": 3646, + "GCAGCC": 3647, + "GCAGCG": 3648, + "GCAGGA": 3649, + "GCAGGT": 3650, + "GCAGGC": 3651, + "GCAGGG": 3652, + "GCTAAA": 3653, + "GCTAAT": 3654, + "GCTAAC": 3655, + "GCTAAG": 3656, + "GCTATA": 3657, + "GCTATT": 3658, + "GCTATC": 3659, + "GCTATG": 3660, + "GCTACA": 3661, + "GCTACT": 3662, + "GCTACC": 3663, + "GCTACG": 3664, + "GCTAGA": 3665, + "GCTAGT": 3666, + "GCTAGC": 3667, + "GCTAGG": 3668, + "GCTTAA": 3669, + "GCTTAT": 3670, + "GCTTAC": 3671, + "GCTTAG": 3672, + "GCTTTA": 3673, + "GCTTTT": 3674, + "GCTTTC": 3675, + "GCTTTG": 3676, + "GCTTCA": 3677, + "GCTTCT": 3678, + "GCTTCC": 3679, + "GCTTCG": 3680, + "GCTTGA": 3681, + "GCTTGT": 3682, + "GCTTGC": 3683, + "GCTTGG": 3684, + "GCTCAA": 3685, + "GCTCAT": 3686, + "GCTCAC": 3687, + "GCTCAG": 3688, + "GCTCTA": 3689, + "GCTCTT": 3690, + "GCTCTC": 3691, + "GCTCTG": 3692, + "GCTCCA": 3693, + "GCTCCT": 3694, + "GCTCCC": 3695, + "GCTCCG": 3696, + "GCTCGA": 3697, + "GCTCGT": 3698, + "GCTCGC": 3699, + "GCTCGG": 3700, + "GCTGAA": 3701, + "GCTGAT": 3702, + "GCTGAC": 3703, + "GCTGAG": 3704, + "GCTGTA": 3705, + "GCTGTT": 3706, + "GCTGTC": 3707, + "GCTGTG": 3708, + "GCTGCA": 3709, + "GCTGCT": 3710, + "GCTGCC": 3711, + "GCTGCG": 3712, + "GCTGGA": 3713, + "GCTGGT": 3714, + "GCTGGC": 3715, + "GCTGGG": 3716, + "GCCAAA": 3717, + "GCCAAT": 3718, + "GCCAAC": 3719, + "GCCAAG": 3720, + "GCCATA": 3721, + "GCCATT": 3722, + "GCCATC": 3723, + "GCCATG": 3724, + "GCCACA": 3725, + "GCCACT": 3726, + "GCCACC": 3727, + "GCCACG": 3728, + "GCCAGA": 3729, + "GCCAGT": 3730, + "GCCAGC": 3731, + "GCCAGG": 3732, + "GCCTAA": 3733, + "GCCTAT": 3734, + "GCCTAC": 3735, + "GCCTAG": 3736, + "GCCTTA": 3737, + "GCCTTT": 3738, + "GCCTTC": 3739, + "GCCTTG": 3740, + "GCCTCA": 3741, + "GCCTCT": 3742, + "GCCTCC": 3743, + "GCCTCG": 3744, + "GCCTGA": 3745, + "GCCTGT": 3746, + "GCCTGC": 3747, + "GCCTGG": 3748, + "GCCCAA": 3749, + "GCCCAT": 3750, + "GCCCAC": 3751, + "GCCCAG": 3752, + "GCCCTA": 3753, + "GCCCTT": 3754, + "GCCCTC": 3755, + "GCCCTG": 3756, + "GCCCCA": 3757, + "GCCCCT": 3758, + "GCCCCC": 3759, + "GCCCCG": 3760, + "GCCCGA": 3761, + "GCCCGT": 3762, + "GCCCGC": 3763, + "GCCCGG": 3764, + "GCCGAA": 3765, + "GCCGAT": 3766, + "GCCGAC": 3767, + "GCCGAG": 3768, + "GCCGTA": 3769, + "GCCGTT": 3770, + "GCCGTC": 3771, + "GCCGTG": 3772, + "GCCGCA": 3773, + "GCCGCT": 3774, + "GCCGCC": 3775, + "GCCGCG": 3776, + "GCCGGA": 3777, + "GCCGGT": 3778, + "GCCGGC": 3779, + "GCCGGG": 3780, + "GCGAAA": 3781, + "GCGAAT": 3782, + "GCGAAC": 3783, + "GCGAAG": 3784, + "GCGATA": 3785, + "GCGATT": 3786, + "GCGATC": 3787, + "GCGATG": 3788, + "GCGACA": 3789, + "GCGACT": 3790, + "GCGACC": 3791, + "GCGACG": 3792, + "GCGAGA": 3793, + "GCGAGT": 3794, + "GCGAGC": 3795, + "GCGAGG": 3796, + "GCGTAA": 3797, + "GCGTAT": 3798, + "GCGTAC": 3799, + "GCGTAG": 3800, + "GCGTTA": 3801, + "GCGTTT": 3802, + "GCGTTC": 3803, + "GCGTTG": 3804, + "GCGTCA": 3805, + "GCGTCT": 3806, + "GCGTCC": 3807, + "GCGTCG": 3808, + "GCGTGA": 3809, + "GCGTGT": 3810, + "GCGTGC": 3811, + "GCGTGG": 3812, + "GCGCAA": 3813, + "GCGCAT": 3814, + "GCGCAC": 3815, + "GCGCAG": 3816, + "GCGCTA": 3817, + "GCGCTT": 3818, + "GCGCTC": 3819, + "GCGCTG": 3820, + "GCGCCA": 3821, + "GCGCCT": 3822, + "GCGCCC": 3823, + "GCGCCG": 3824, + "GCGCGA": 3825, + "GCGCGT": 3826, + "GCGCGC": 3827, + "GCGCGG": 3828, + "GCGGAA": 3829, + "GCGGAT": 3830, + "GCGGAC": 3831, + "GCGGAG": 3832, + "GCGGTA": 3833, + "GCGGTT": 3834, + "GCGGTC": 3835, + "GCGGTG": 3836, + "GCGGCA": 3837, + "GCGGCT": 3838, + "GCGGCC": 3839, + "GCGGCG": 3840, + "GCGGGA": 3841, + "GCGGGT": 3842, + "GCGGGC": 3843, + "GCGGGG": 3844, + "GGAAAA": 3845, + "GGAAAT": 3846, + "GGAAAC": 3847, + "GGAAAG": 3848, + "GGAATA": 3849, + "GGAATT": 3850, + "GGAATC": 3851, + "GGAATG": 3852, + "GGAACA": 3853, + "GGAACT": 3854, + "GGAACC": 3855, + "GGAACG": 3856, + "GGAAGA": 3857, + "GGAAGT": 3858, + "GGAAGC": 3859, + "GGAAGG": 3860, + "GGATAA": 3861, + "GGATAT": 3862, + "GGATAC": 3863, + "GGATAG": 3864, + "GGATTA": 3865, + "GGATTT": 3866, + "GGATTC": 3867, + "GGATTG": 3868, + "GGATCA": 3869, + "GGATCT": 3870, + "GGATCC": 3871, + "GGATCG": 3872, + "GGATGA": 3873, + "GGATGT": 3874, + "GGATGC": 3875, + "GGATGG": 3876, + "GGACAA": 3877, + "GGACAT": 3878, + "GGACAC": 3879, + "GGACAG": 3880, + "GGACTA": 3881, + "GGACTT": 3882, + "GGACTC": 3883, + "GGACTG": 3884, + "GGACCA": 3885, + "GGACCT": 3886, + "GGACCC": 3887, + "GGACCG": 3888, + "GGACGA": 3889, + "GGACGT": 3890, + "GGACGC": 3891, + "GGACGG": 3892, + "GGAGAA": 3893, + "GGAGAT": 3894, + "GGAGAC": 3895, + "GGAGAG": 3896, + "GGAGTA": 3897, + "GGAGTT": 3898, + "GGAGTC": 3899, + "GGAGTG": 3900, + "GGAGCA": 3901, + "GGAGCT": 3902, + "GGAGCC": 3903, + "GGAGCG": 3904, + "GGAGGA": 3905, + "GGAGGT": 3906, + "GGAGGC": 3907, + "GGAGGG": 3908, + "GGTAAA": 3909, + "GGTAAT": 3910, + "GGTAAC": 3911, + "GGTAAG": 3912, + "GGTATA": 3913, + "GGTATT": 3914, + "GGTATC": 3915, + "GGTATG": 3916, + "GGTACA": 3917, + "GGTACT": 3918, + "GGTACC": 3919, + "GGTACG": 3920, + "GGTAGA": 3921, + "GGTAGT": 3922, + "GGTAGC": 3923, + "GGTAGG": 3924, + "GGTTAA": 3925, + "GGTTAT": 3926, + "GGTTAC": 3927, + "GGTTAG": 3928, + "GGTTTA": 3929, + "GGTTTT": 3930, + "GGTTTC": 3931, + "GGTTTG": 3932, + "GGTTCA": 3933, + "GGTTCT": 3934, + "GGTTCC": 3935, + "GGTTCG": 3936, + "GGTTGA": 3937, + "GGTTGT": 3938, + "GGTTGC": 3939, + "GGTTGG": 3940, + "GGTCAA": 3941, + "GGTCAT": 3942, + "GGTCAC": 3943, + "GGTCAG": 3944, + "GGTCTA": 3945, + "GGTCTT": 3946, + "GGTCTC": 3947, + "GGTCTG": 3948, + "GGTCCA": 3949, + "GGTCCT": 3950, + "GGTCCC": 3951, + "GGTCCG": 3952, + "GGTCGA": 3953, + "GGTCGT": 3954, + "GGTCGC": 3955, + "GGTCGG": 3956, + "GGTGAA": 3957, + "GGTGAT": 3958, + "GGTGAC": 3959, + "GGTGAG": 3960, + "GGTGTA": 3961, + "GGTGTT": 3962, + "GGTGTC": 3963, + "GGTGTG": 3964, + "GGTGCA": 3965, + "GGTGCT": 3966, + "GGTGCC": 3967, + "GGTGCG": 3968, + "GGTGGA": 3969, + "GGTGGT": 3970, + "GGTGGC": 3971, + "GGTGGG": 3972, + "GGCAAA": 3973, + "GGCAAT": 3974, + "GGCAAC": 3975, + "GGCAAG": 3976, + "GGCATA": 3977, + "GGCATT": 3978, + "GGCATC": 3979, + "GGCATG": 3980, + "GGCACA": 3981, + "GGCACT": 3982, + "GGCACC": 3983, + "GGCACG": 3984, + "GGCAGA": 3985, + "GGCAGT": 3986, + "GGCAGC": 3987, + "GGCAGG": 3988, + "GGCTAA": 3989, + "GGCTAT": 3990, + "GGCTAC": 3991, + "GGCTAG": 3992, + "GGCTTA": 3993, + "GGCTTT": 3994, + "GGCTTC": 3995, + "GGCTTG": 3996, + "GGCTCA": 3997, + "GGCTCT": 3998, + "GGCTCC": 3999, + "GGCTCG": 4000, + "GGCTGA": 4001, + "GGCTGT": 4002, + "GGCTGC": 4003, + "GGCTGG": 4004, + "GGCCAA": 4005, + "GGCCAT": 4006, + "GGCCAC": 4007, + "GGCCAG": 4008, + "GGCCTA": 4009, + "GGCCTT": 4010, + "GGCCTC": 4011, + "GGCCTG": 4012, + "GGCCCA": 4013, + "GGCCCT": 4014, + "GGCCCC": 4015, + "GGCCCG": 4016, + "GGCCGA": 4017, + "GGCCGT": 4018, + "GGCCGC": 4019, + "GGCCGG": 4020, + "GGCGAA": 4021, + "GGCGAT": 4022, + "GGCGAC": 4023, + "GGCGAG": 4024, + "GGCGTA": 4025, + "GGCGTT": 4026, + "GGCGTC": 4027, + "GGCGTG": 4028, + "GGCGCA": 4029, + "GGCGCT": 4030, + "GGCGCC": 4031, + "GGCGCG": 4032, + "GGCGGA": 4033, + "GGCGGT": 4034, + "GGCGGC": 4035, + "GGCGGG": 4036, + "GGGAAA": 4037, + "GGGAAT": 4038, + "GGGAAC": 4039, + "GGGAAG": 4040, + "GGGATA": 4041, + "GGGATT": 4042, + "GGGATC": 4043, + "GGGATG": 4044, + "GGGACA": 4045, + "GGGACT": 4046, + "GGGACC": 4047, + "GGGACG": 4048, + "GGGAGA": 4049, + "GGGAGT": 4050, + "GGGAGC": 4051, + "GGGAGG": 4052, + "GGGTAA": 4053, + "GGGTAT": 4054, + "GGGTAC": 4055, + "GGGTAG": 4056, + "GGGTTA": 4057, + "GGGTTT": 4058, + "GGGTTC": 4059, + "GGGTTG": 4060, + "GGGTCA": 4061, + "GGGTCT": 4062, + "GGGTCC": 4063, + "GGGTCG": 4064, + "GGGTGA": 4065, + "GGGTGT": 4066, + "GGGTGC": 4067, + "GGGTGG": 4068, + "GGGCAA": 4069, + "GGGCAT": 4070, + "GGGCAC": 4071, + "GGGCAG": 4072, + "GGGCTA": 4073, + "GGGCTT": 4074, + "GGGCTC": 4075, + "GGGCTG": 4076, + "GGGCCA": 4077, + "GGGCCT": 4078, + "GGGCCC": 4079, + "GGGCCG": 4080, + "GGGCGA": 4081, + "GGGCGT": 4082, + "GGGCGC": 4083, + "GGGCGG": 4084, + "GGGGAA": 4085, + "GGGGAT": 4086, + "GGGGAC": 4087, + "GGGGAG": 4088, + "GGGGTA": 4089, + "GGGGTT": 4090, + "GGGGTC": 4091, + "GGGGTG": 4092, + "GGGGCA": 4093, + "GGGGCT": 4094, + "GGGGCC": 4095, + "GGGGCG": 4096, + "GGGGGA": 4097, + "GGGGGT": 4098, + "GGGGGC": 4099, + "GGGGGG": 4100 + } + } +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/tokenizer_config.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..604e2b1b29af1dcb3920d2828b6a90276e5e3d02 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/tokenizer_config.json @@ -0,0 +1,58 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "[UNK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "[CLS]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "[SEP]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "4": { + "content": "[MASK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "clean_up_tokenization_spaces": true, + "cls_token": "[CLS]", + "do_basic_tokenize": true, + "do_lower_case": false, + "mask_token": "[MASK]", + "max_len": 512, + "model_max_length": 512, + "never_split": null, + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "strip_accents": null, + "tokenize_chinese_chars": true, + "tokenizer_class": "BertTokenizer", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/trainer_state.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..89282fa193808a19ccc05637442decedfdffdb11 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/trainer_state.json @@ -0,0 +1,873 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 12.0, + "eval_steps": 500, + "global_step": 3000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.1, + "grad_norm": 0.2218121588230133, + "learning_rate": 5.319148936170213e-05, + "loss": 0.6927, + "step": 25 + }, + { + "epoch": 0.2, + "grad_norm": 0.44851255416870117, + "learning_rate": 0.00010638297872340425, + "loss": 0.6929, + "step": 50 + }, + { + "epoch": 0.3, + "grad_norm": 0.2440175861120224, + "learning_rate": 0.00015957446808510637, + "loss": 0.6924, + "step": 75 + }, + { + "epoch": 0.4, + "grad_norm": 0.3422006666660309, + "learning_rate": 0.0002127659574468085, + "loss": 0.6922, + "step": 100 + }, + { + "epoch": 0.5, + "grad_norm": 0.267333447933197, + "learning_rate": 0.00026595744680851064, + "loss": 0.6913, + "step": 125 + }, + { + "epoch": 0.6, + "grad_norm": 0.45699772238731384, + "learning_rate": 0.00031914893617021275, + "loss": 0.6896, + "step": 150 + }, + { + "epoch": 0.7, + "grad_norm": 0.3004947304725647, + "learning_rate": 0.0003723404255319149, + "loss": 0.6887, + "step": 175 + }, + { + "epoch": 0.8, + "grad_norm": 0.33440759778022766, + "learning_rate": 0.0003992081821181128, + "loss": 0.6857, + "step": 200 + }, + { + "epoch": 0.9, + "grad_norm": 0.2918928861618042, + "learning_rate": 0.0003975585615308479, + "loss": 0.6882, + "step": 225 + }, + { + "epoch": 1.0, + "grad_norm": 0.2934085726737976, + "learning_rate": 0.00039590894094358297, + "loss": 0.6807, + "step": 250 + }, + { + "epoch": 1.1, + "grad_norm": 0.3364830017089844, + "learning_rate": 0.0003942593203563181, + "loss": 0.6848, + "step": 275 + }, + { + "epoch": 1.2, + "grad_norm": 0.8379159569740295, + "learning_rate": 0.0003926096997690532, + "loss": 0.677, + "step": 300 + }, + { + "epoch": 1.3, + "grad_norm": 0.37671706080436707, + "learning_rate": 0.00039096007918178817, + "loss": 0.6744, + "step": 325 + }, + { + "epoch": 1.4, + "grad_norm": 0.34329330921173096, + "learning_rate": 0.0003893104585945233, + "loss": 0.6733, + "step": 350 + }, + { + "epoch": 1.5, + "grad_norm": 0.4286085367202759, + "learning_rate": 0.0003876608380072583, + "loss": 0.6772, + "step": 375 + }, + { + "epoch": 1.6, + "grad_norm": 0.3576306998729706, + "learning_rate": 0.00038601121741999343, + "loss": 0.6799, + "step": 400 + }, + { + "epoch": 1.7, + "grad_norm": 0.4636435806751251, + "learning_rate": 0.0003843615968327285, + "loss": 0.6819, + "step": 425 + }, + { + "epoch": 1.8, + "grad_norm": 0.4552135765552521, + "learning_rate": 0.0003827119762454636, + "loss": 0.6731, + "step": 450 + }, + { + "epoch": 1.9, + "grad_norm": 0.3565954566001892, + "learning_rate": 0.00038106235565819863, + "loss": 0.6718, + "step": 475 + }, + { + "epoch": 2.0, + "grad_norm": 0.5451686978340149, + "learning_rate": 0.0003794127350709337, + "loss": 0.6858, + "step": 500 + }, + { + "epoch": 2.1, + "grad_norm": 0.3730587959289551, + "learning_rate": 0.0003777631144836688, + "loss": 0.667, + "step": 525 + }, + { + "epoch": 2.2, + "grad_norm": 0.6642417907714844, + "learning_rate": 0.00037611349389640383, + "loss": 0.6662, + "step": 550 + }, + { + "epoch": 2.3, + "grad_norm": 0.3889809250831604, + "learning_rate": 0.00037446387330913894, + "loss": 0.6718, + "step": 575 + }, + { + "epoch": 2.4, + "grad_norm": 0.47203826904296875, + "learning_rate": 0.000372814252721874, + "loss": 0.6687, + "step": 600 + }, + { + "epoch": 2.5, + "grad_norm": 0.5293862819671631, + "learning_rate": 0.00037116463213460903, + "loss": 0.6655, + "step": 625 + }, + { + "epoch": 2.6, + "grad_norm": 0.36724531650543213, + "learning_rate": 0.00036951501154734414, + "loss": 0.672, + "step": 650 + }, + { + "epoch": 2.7, + "grad_norm": 0.5733087062835693, + "learning_rate": 0.0003678653909600792, + "loss": 0.6683, + "step": 675 + }, + { + "epoch": 2.8, + "grad_norm": 0.6385205984115601, + "learning_rate": 0.0003662157703728143, + "loss": 0.67, + "step": 700 + }, + { + "epoch": 2.9, + "grad_norm": 0.6146281957626343, + "learning_rate": 0.00036456614978554934, + "loss": 0.6606, + "step": 725 + }, + { + "epoch": 3.0, + "grad_norm": 0.6535729169845581, + "learning_rate": 0.00036291652919828444, + "loss": 0.6691, + "step": 750 + }, + { + "epoch": 3.1, + "grad_norm": 0.5194367170333862, + "learning_rate": 0.00036126690861101944, + "loss": 0.6517, + "step": 775 + }, + { + "epoch": 3.2, + "grad_norm": 0.7116692066192627, + "learning_rate": 0.00035961728802375454, + "loss": 0.6501, + "step": 800 + }, + { + "epoch": 3.3, + "grad_norm": 0.5338636040687561, + "learning_rate": 0.0003579676674364896, + "loss": 0.6682, + "step": 825 + }, + { + "epoch": 3.4, + "grad_norm": 0.6912877559661865, + "learning_rate": 0.0003563180468492247, + "loss": 0.6632, + "step": 850 + }, + { + "epoch": 3.5, + "grad_norm": 0.6925108432769775, + "learning_rate": 0.0003546684262619598, + "loss": 0.653, + "step": 875 + }, + { + "epoch": 3.6, + "grad_norm": 0.5360299944877625, + "learning_rate": 0.00035301880567469485, + "loss": 0.6613, + "step": 900 + }, + { + "epoch": 3.7, + "grad_norm": 0.6487969160079956, + "learning_rate": 0.0003513691850874299, + "loss": 0.6627, + "step": 925 + }, + { + "epoch": 3.8, + "grad_norm": 1.7260371446609497, + "learning_rate": 0.00034971956450016495, + "loss": 0.6436, + "step": 950 + }, + { + "epoch": 3.9, + "grad_norm": 0.5852354168891907, + "learning_rate": 0.00034806994391290005, + "loss": 0.6613, + "step": 975 + }, + { + "epoch": 4.0, + "grad_norm": 0.7802712917327881, + "learning_rate": 0.0003464203233256351, + "loss": 0.658, + "step": 1000 + }, + { + "epoch": 4.1, + "grad_norm": 0.7335640788078308, + "learning_rate": 0.0003447707027383702, + "loss": 0.6476, + "step": 1025 + }, + { + "epoch": 4.2, + "grad_norm": 0.9540443420410156, + "learning_rate": 0.00034312108215110525, + "loss": 0.6334, + "step": 1050 + }, + { + "epoch": 4.3, + "grad_norm": 1.3698925971984863, + "learning_rate": 0.0003414714615638403, + "loss": 0.6512, + "step": 1075 + }, + { + "epoch": 4.4, + "grad_norm": 0.8282333016395569, + "learning_rate": 0.0003398218409765754, + "loss": 0.6556, + "step": 1100 + }, + { + "epoch": 4.5, + "grad_norm": 1.4414807558059692, + "learning_rate": 0.00033817222038931045, + "loss": 0.6628, + "step": 1125 + }, + { + "epoch": 4.6, + "grad_norm": 0.5493957996368408, + "learning_rate": 0.00033652259980204556, + "loss": 0.6461, + "step": 1150 + }, + { + "epoch": 4.7, + "grad_norm": 1.2343393564224243, + "learning_rate": 0.0003348729792147806, + "loss": 0.6464, + "step": 1175 + }, + { + "epoch": 4.8, + "grad_norm": 0.7163398265838623, + "learning_rate": 0.0003332233586275157, + "loss": 0.6518, + "step": 1200 + }, + { + "epoch": 4.9, + "grad_norm": 0.7878938317298889, + "learning_rate": 0.00033157373804025076, + "loss": 0.6502, + "step": 1225 + }, + { + "epoch": 5.0, + "grad_norm": 0.952571451663971, + "learning_rate": 0.0003299241174529858, + "loss": 0.6463, + "step": 1250 + }, + { + "epoch": 5.1, + "grad_norm": 0.6535393595695496, + "learning_rate": 0.0003282744968657209, + "loss": 0.6361, + "step": 1275 + }, + { + "epoch": 5.2, + "grad_norm": 0.8239464163780212, + "learning_rate": 0.00032662487627845596, + "loss": 0.6264, + "step": 1300 + }, + { + "epoch": 5.3, + "grad_norm": 1.4500653743743896, + "learning_rate": 0.00032497525569119106, + "loss": 0.6544, + "step": 1325 + }, + { + "epoch": 5.4, + "grad_norm": 1.1515296697616577, + "learning_rate": 0.0003233256351039261, + "loss": 0.6629, + "step": 1350 + }, + { + "epoch": 5.5, + "grad_norm": 0.6899295449256897, + "learning_rate": 0.00032167601451666116, + "loss": 0.6453, + "step": 1375 + }, + { + "epoch": 5.6, + "grad_norm": 1.1782217025756836, + "learning_rate": 0.00032002639392939627, + "loss": 0.6362, + "step": 1400 + }, + { + "epoch": 5.7, + "grad_norm": 0.6971372961997986, + "learning_rate": 0.0003183767733421313, + "loss": 0.6593, + "step": 1425 + }, + { + "epoch": 5.8, + "grad_norm": 0.7293921709060669, + "learning_rate": 0.0003167271527548664, + "loss": 0.6318, + "step": 1450 + }, + { + "epoch": 5.9, + "grad_norm": 0.6860108375549316, + "learning_rate": 0.00031507753216760147, + "loss": 0.6402, + "step": 1475 + }, + { + "epoch": 6.0, + "grad_norm": 1.330344319343567, + "learning_rate": 0.00031342791158033657, + "loss": 0.6276, + "step": 1500 + }, + { + "epoch": 6.1, + "grad_norm": 0.7261562943458557, + "learning_rate": 0.00031177829099307157, + "loss": 0.6369, + "step": 1525 + }, + { + "epoch": 6.2, + "grad_norm": 1.1324542760849, + "learning_rate": 0.00031012867040580667, + "loss": 0.6228, + "step": 1550 + }, + { + "epoch": 6.3, + "grad_norm": 0.7002120614051819, + "learning_rate": 0.0003084790498185417, + "loss": 0.6389, + "step": 1575 + }, + { + "epoch": 6.4, + "grad_norm": 1.3143397569656372, + "learning_rate": 0.0003068294292312768, + "loss": 0.6362, + "step": 1600 + }, + { + "epoch": 6.5, + "grad_norm": 1.6375401020050049, + "learning_rate": 0.0003051798086440119, + "loss": 0.6348, + "step": 1625 + }, + { + "epoch": 6.6, + "grad_norm": 1.4343605041503906, + "learning_rate": 0.000303530188056747, + "loss": 0.6496, + "step": 1650 + }, + { + "epoch": 6.7, + "grad_norm": 0.815762996673584, + "learning_rate": 0.000301880567469482, + "loss": 0.6407, + "step": 1675 + }, + { + "epoch": 6.8, + "grad_norm": 1.068608045578003, + "learning_rate": 0.0003002309468822171, + "loss": 0.6417, + "step": 1700 + }, + { + "epoch": 6.9, + "grad_norm": 0.8775753974914551, + "learning_rate": 0.0002985813262949522, + "loss": 0.6281, + "step": 1725 + }, + { + "epoch": 7.0, + "grad_norm": 1.041817545890808, + "learning_rate": 0.00029693170570768723, + "loss": 0.6485, + "step": 1750 + }, + { + "epoch": 7.1, + "grad_norm": 0.7135840058326721, + "learning_rate": 0.00029528208512042233, + "loss": 0.6192, + "step": 1775 + }, + { + "epoch": 7.2, + "grad_norm": 1.5523930788040161, + "learning_rate": 0.0002936324645331574, + "loss": 0.636, + "step": 1800 + }, + { + "epoch": 7.3, + "grad_norm": 0.7677174210548401, + "learning_rate": 0.00029198284394589243, + "loss": 0.6213, + "step": 1825 + }, + { + "epoch": 7.4, + "grad_norm": 0.8086565732955933, + "learning_rate": 0.00029033322335862753, + "loss": 0.6292, + "step": 1850 + }, + { + "epoch": 7.5, + "grad_norm": 3.1128971576690674, + "learning_rate": 0.0002886836027713626, + "loss": 0.648, + "step": 1875 + }, + { + "epoch": 7.6, + "grad_norm": 0.9243882894515991, + "learning_rate": 0.0002870339821840977, + "loss": 0.6286, + "step": 1900 + }, + { + "epoch": 7.7, + "grad_norm": 1.8203461170196533, + "learning_rate": 0.00028538436159683273, + "loss": 0.6301, + "step": 1925 + }, + { + "epoch": 7.8, + "grad_norm": 1.0924643278121948, + "learning_rate": 0.00028373474100956784, + "loss": 0.6378, + "step": 1950 + }, + { + "epoch": 7.9, + "grad_norm": 1.0839996337890625, + "learning_rate": 0.0002820851204223029, + "loss": 0.6265, + "step": 1975 + }, + { + "epoch": 8.0, + "grad_norm": 0.7143226861953735, + "learning_rate": 0.00028043549983503794, + "loss": 0.649, + "step": 2000 + }, + { + "epoch": 8.1, + "grad_norm": 0.8778955936431885, + "learning_rate": 0.00027878587924777304, + "loss": 0.6411, + "step": 2025 + }, + { + "epoch": 8.2, + "grad_norm": 1.644904613494873, + "learning_rate": 0.0002771362586605081, + "loss": 0.6168, + "step": 2050 + }, + { + "epoch": 8.3, + "grad_norm": 0.7915159463882446, + "learning_rate": 0.0002754866380732432, + "loss": 0.6324, + "step": 2075 + }, + { + "epoch": 8.4, + "grad_norm": 1.1212801933288574, + "learning_rate": 0.00027383701748597824, + "loss": 0.6232, + "step": 2100 + }, + { + "epoch": 8.5, + "grad_norm": 0.7717731595039368, + "learning_rate": 0.0002721873968987133, + "loss": 0.624, + "step": 2125 + }, + { + "epoch": 8.6, + "grad_norm": 1.0577210187911987, + "learning_rate": 0.00027053777631144834, + "loss": 0.6296, + "step": 2150 + }, + { + "epoch": 8.7, + "grad_norm": 1.0531567335128784, + "learning_rate": 0.00026888815572418344, + "loss": 0.6345, + "step": 2175 + }, + { + "epoch": 8.8, + "grad_norm": 2.8983640670776367, + "learning_rate": 0.00026723853513691855, + "loss": 0.6142, + "step": 2200 + }, + { + "epoch": 8.9, + "grad_norm": 0.7773396968841553, + "learning_rate": 0.0002655889145496536, + "loss": 0.634, + "step": 2225 + }, + { + "epoch": 9.0, + "grad_norm": 1.0587151050567627, + "learning_rate": 0.0002639392939623887, + "loss": 0.6326, + "step": 2250 + }, + { + "epoch": 9.1, + "grad_norm": 0.8009796738624573, + "learning_rate": 0.0002622896733751237, + "loss": 0.6175, + "step": 2275 + }, + { + "epoch": 9.2, + "grad_norm": 1.2530399560928345, + "learning_rate": 0.0002606400527878588, + "loss": 0.6239, + "step": 2300 + }, + { + "epoch": 9.3, + "grad_norm": 0.7976368069648743, + "learning_rate": 0.00025899043220059385, + "loss": 0.6452, + "step": 2325 + }, + { + "epoch": 9.4, + "grad_norm": 1.455032229423523, + "learning_rate": 0.00025734081161332895, + "loss": 0.6351, + "step": 2350 + }, + { + "epoch": 9.5, + "grad_norm": 0.8916760087013245, + "learning_rate": 0.000255691191026064, + "loss": 0.5986, + "step": 2375 + }, + { + "epoch": 9.6, + "grad_norm": 1.104746699333191, + "learning_rate": 0.0002540415704387991, + "loss": 0.6104, + "step": 2400 + }, + { + "epoch": 9.7, + "grad_norm": 0.9449218511581421, + "learning_rate": 0.00025239194985153415, + "loss": 0.6262, + "step": 2425 + }, + { + "epoch": 9.8, + "grad_norm": 2.0286591053009033, + "learning_rate": 0.0002507423292642692, + "loss": 0.6124, + "step": 2450 + }, + { + "epoch": 9.9, + "grad_norm": 2.273597002029419, + "learning_rate": 0.0002490927086770043, + "loss": 0.6164, + "step": 2475 + }, + { + "epoch": 10.0, + "grad_norm": 0.8174235224723816, + "learning_rate": 0.00024744308808973936, + "loss": 0.6254, + "step": 2500 + }, + { + "epoch": 10.1, + "grad_norm": 1.0447920560836792, + "learning_rate": 0.00024579346750247446, + "loss": 0.6187, + "step": 2525 + }, + { + "epoch": 10.2, + "grad_norm": 1.0433655977249146, + "learning_rate": 0.0002441438469152095, + "loss": 0.6183, + "step": 2550 + }, + { + "epoch": 10.3, + "grad_norm": 1.6479028463363647, + "learning_rate": 0.00024249422632794456, + "loss": 0.6353, + "step": 2575 + }, + { + "epoch": 10.4, + "grad_norm": 0.8032572865486145, + "learning_rate": 0.00024084460574067963, + "loss": 0.6206, + "step": 2600 + }, + { + "epoch": 10.5, + "grad_norm": 1.0675058364868164, + "learning_rate": 0.0002391949851534147, + "loss": 0.6168, + "step": 2625 + }, + { + "epoch": 10.6, + "grad_norm": 1.2228609323501587, + "learning_rate": 0.0002375453645661498, + "loss": 0.6007, + "step": 2650 + }, + { + "epoch": 10.7, + "grad_norm": 1.1027027368545532, + "learning_rate": 0.00023589574397888486, + "loss": 0.6227, + "step": 2675 + }, + { + "epoch": 10.8, + "grad_norm": 1.2949633598327637, + "learning_rate": 0.00023424612339161997, + "loss": 0.6179, + "step": 2700 + }, + { + "epoch": 10.9, + "grad_norm": 1.235032320022583, + "learning_rate": 0.000232596502804355, + "loss": 0.6216, + "step": 2725 + }, + { + "epoch": 11.0, + "grad_norm": 1.0087182521820068, + "learning_rate": 0.00023094688221709007, + "loss": 0.6286, + "step": 2750 + }, + { + "epoch": 11.1, + "grad_norm": 1.301207423210144, + "learning_rate": 0.00022929726162982514, + "loss": 0.5953, + "step": 2775 + }, + { + "epoch": 11.2, + "grad_norm": 1.6798988580703735, + "learning_rate": 0.00022764764104256022, + "loss": 0.6331, + "step": 2800 + }, + { + "epoch": 11.3, + "grad_norm": 1.1364004611968994, + "learning_rate": 0.0002259980204552953, + "loss": 0.6047, + "step": 2825 + }, + { + "epoch": 11.4, + "grad_norm": 0.9963313341140747, + "learning_rate": 0.00022434839986803037, + "loss": 0.6192, + "step": 2850 + }, + { + "epoch": 11.5, + "grad_norm": 1.5672775506973267, + "learning_rate": 0.00022269877928076542, + "loss": 0.6063, + "step": 2875 + }, + { + "epoch": 11.6, + "grad_norm": 1.4239299297332764, + "learning_rate": 0.0002210491586935005, + "loss": 0.6058, + "step": 2900 + }, + { + "epoch": 11.7, + "grad_norm": 1.5440959930419922, + "learning_rate": 0.00021939953810623557, + "loss": 0.6239, + "step": 2925 + }, + { + "epoch": 11.8, + "grad_norm": 1.35140860080719, + "learning_rate": 0.00021774991751897065, + "loss": 0.6241, + "step": 2950 + }, + { + "epoch": 11.9, + "grad_norm": 1.9355136156082153, + "learning_rate": 0.00021610029693170573, + "loss": 0.6135, + "step": 2975 + }, + { + "epoch": 12.0, + "grad_norm": 1.134037733078003, + "learning_rate": 0.0002144506763444408, + "loss": 0.6142, + "step": 3000 + } + ], + "logging_steps": 25, + "max_steps": 6250, + "num_input_tokens_seen": 0, + "num_train_epochs": 25, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 166143098880000.0, + "train_batch_size": 20, + "trial_name": null, + "trial_params": null +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/training_args.bin b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..540db3fbee786996d9705d768d31ba6c371187be --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dc7816b5f3fd38f01f9ba38a197ce070cde06e4c90e5260cb4814f34f628907d +size 5304 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/vocab.txt b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/vocab.txt new file mode 100644 index 0000000000000000000000000000000000000000..cc9ad53732e21ea9180f18e1fb67a38cc5702d4c --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-3000/vocab.txt @@ -0,0 +1,4101 @@ +[PAD] +[UNK] +[CLS] +[SEP] +[MASK] +AAAAAA +AAAAAT +AAAAAC +AAAAAG +AAAATA +AAAATT +AAAATC +AAAATG +AAAACA +AAAACT +AAAACC +AAAACG +AAAAGA +AAAAGT +AAAAGC +AAAAGG +AAATAA +AAATAT +AAATAC +AAATAG +AAATTA +AAATTT +AAATTC +AAATTG +AAATCA +AAATCT +AAATCC +AAATCG +AAATGA +AAATGT +AAATGC +AAATGG +AAACAA +AAACAT +AAACAC +AAACAG +AAACTA +AAACTT +AAACTC +AAACTG +AAACCA +AAACCT +AAACCC +AAACCG +AAACGA +AAACGT +AAACGC +AAACGG +AAAGAA +AAAGAT +AAAGAC +AAAGAG +AAAGTA +AAAGTT +AAAGTC +AAAGTG +AAAGCA +AAAGCT +AAAGCC +AAAGCG +AAAGGA +AAAGGT +AAAGGC +AAAGGG +AATAAA +AATAAT +AATAAC +AATAAG +AATATA +AATATT +AATATC +AATATG +AATACA +AATACT +AATACC +AATACG +AATAGA +AATAGT +AATAGC +AATAGG +AATTAA +AATTAT +AATTAC +AATTAG +AATTTA +AATTTT +AATTTC +AATTTG +AATTCA +AATTCT +AATTCC +AATTCG +AATTGA +AATTGT +AATTGC +AATTGG +AATCAA +AATCAT +AATCAC +AATCAG +AATCTA +AATCTT +AATCTC +AATCTG +AATCCA +AATCCT +AATCCC +AATCCG +AATCGA +AATCGT +AATCGC +AATCGG +AATGAA +AATGAT +AATGAC +AATGAG +AATGTA +AATGTT +AATGTC +AATGTG +AATGCA +AATGCT +AATGCC +AATGCG +AATGGA +AATGGT +AATGGC +AATGGG +AACAAA +AACAAT +AACAAC +AACAAG +AACATA +AACATT +AACATC +AACATG +AACACA +AACACT +AACACC +AACACG +AACAGA +AACAGT +AACAGC +AACAGG +AACTAA +AACTAT +AACTAC +AACTAG +AACTTA +AACTTT +AACTTC +AACTTG +AACTCA +AACTCT +AACTCC +AACTCG +AACTGA +AACTGT +AACTGC +AACTGG +AACCAA +AACCAT +AACCAC +AACCAG +AACCTA +AACCTT +AACCTC +AACCTG +AACCCA +AACCCT +AACCCC +AACCCG +AACCGA +AACCGT +AACCGC +AACCGG +AACGAA +AACGAT +AACGAC +AACGAG +AACGTA +AACGTT +AACGTC +AACGTG +AACGCA +AACGCT +AACGCC +AACGCG +AACGGA +AACGGT +AACGGC +AACGGG +AAGAAA +AAGAAT +AAGAAC +AAGAAG +AAGATA +AAGATT +AAGATC +AAGATG +AAGACA +AAGACT +AAGACC +AAGACG +AAGAGA +AAGAGT +AAGAGC +AAGAGG +AAGTAA +AAGTAT +AAGTAC +AAGTAG +AAGTTA +AAGTTT +AAGTTC +AAGTTG +AAGTCA +AAGTCT +AAGTCC +AAGTCG +AAGTGA +AAGTGT +AAGTGC +AAGTGG +AAGCAA +AAGCAT +AAGCAC +AAGCAG +AAGCTA +AAGCTT +AAGCTC +AAGCTG +AAGCCA +AAGCCT +AAGCCC +AAGCCG +AAGCGA +AAGCGT +AAGCGC +AAGCGG +AAGGAA +AAGGAT +AAGGAC +AAGGAG +AAGGTA +AAGGTT +AAGGTC +AAGGTG +AAGGCA +AAGGCT +AAGGCC +AAGGCG +AAGGGA +AAGGGT +AAGGGC +AAGGGG +ATAAAA +ATAAAT +ATAAAC +ATAAAG +ATAATA +ATAATT +ATAATC +ATAATG +ATAACA +ATAACT +ATAACC +ATAACG +ATAAGA +ATAAGT +ATAAGC +ATAAGG +ATATAA +ATATAT +ATATAC +ATATAG +ATATTA +ATATTT +ATATTC +ATATTG +ATATCA +ATATCT +ATATCC +ATATCG +ATATGA +ATATGT +ATATGC +ATATGG +ATACAA +ATACAT +ATACAC +ATACAG +ATACTA +ATACTT +ATACTC +ATACTG +ATACCA +ATACCT +ATACCC +ATACCG +ATACGA +ATACGT +ATACGC +ATACGG +ATAGAA +ATAGAT +ATAGAC +ATAGAG +ATAGTA +ATAGTT +ATAGTC +ATAGTG +ATAGCA +ATAGCT +ATAGCC +ATAGCG +ATAGGA +ATAGGT +ATAGGC +ATAGGG +ATTAAA +ATTAAT +ATTAAC +ATTAAG +ATTATA +ATTATT +ATTATC +ATTATG +ATTACA +ATTACT +ATTACC +ATTACG +ATTAGA +ATTAGT +ATTAGC +ATTAGG +ATTTAA +ATTTAT +ATTTAC +ATTTAG +ATTTTA +ATTTTT +ATTTTC +ATTTTG +ATTTCA +ATTTCT +ATTTCC +ATTTCG +ATTTGA +ATTTGT +ATTTGC +ATTTGG +ATTCAA +ATTCAT +ATTCAC +ATTCAG +ATTCTA +ATTCTT +ATTCTC +ATTCTG +ATTCCA +ATTCCT +ATTCCC +ATTCCG +ATTCGA +ATTCGT +ATTCGC +ATTCGG +ATTGAA +ATTGAT +ATTGAC +ATTGAG +ATTGTA +ATTGTT +ATTGTC +ATTGTG +ATTGCA +ATTGCT +ATTGCC +ATTGCG +ATTGGA +ATTGGT +ATTGGC +ATTGGG +ATCAAA +ATCAAT +ATCAAC +ATCAAG +ATCATA +ATCATT +ATCATC +ATCATG +ATCACA +ATCACT +ATCACC +ATCACG +ATCAGA +ATCAGT +ATCAGC +ATCAGG +ATCTAA +ATCTAT +ATCTAC +ATCTAG +ATCTTA +ATCTTT +ATCTTC +ATCTTG +ATCTCA +ATCTCT +ATCTCC +ATCTCG +ATCTGA +ATCTGT +ATCTGC +ATCTGG +ATCCAA +ATCCAT +ATCCAC +ATCCAG +ATCCTA +ATCCTT +ATCCTC +ATCCTG +ATCCCA +ATCCCT +ATCCCC +ATCCCG +ATCCGA +ATCCGT +ATCCGC +ATCCGG +ATCGAA +ATCGAT +ATCGAC +ATCGAG +ATCGTA +ATCGTT +ATCGTC +ATCGTG +ATCGCA +ATCGCT +ATCGCC +ATCGCG +ATCGGA +ATCGGT +ATCGGC +ATCGGG +ATGAAA +ATGAAT +ATGAAC +ATGAAG +ATGATA +ATGATT +ATGATC +ATGATG +ATGACA +ATGACT +ATGACC +ATGACG +ATGAGA +ATGAGT +ATGAGC +ATGAGG +ATGTAA +ATGTAT +ATGTAC +ATGTAG +ATGTTA +ATGTTT +ATGTTC +ATGTTG +ATGTCA +ATGTCT +ATGTCC +ATGTCG +ATGTGA +ATGTGT +ATGTGC +ATGTGG +ATGCAA +ATGCAT +ATGCAC +ATGCAG +ATGCTA +ATGCTT +ATGCTC +ATGCTG +ATGCCA +ATGCCT +ATGCCC +ATGCCG +ATGCGA +ATGCGT +ATGCGC +ATGCGG +ATGGAA +ATGGAT +ATGGAC +ATGGAG +ATGGTA +ATGGTT +ATGGTC +ATGGTG +ATGGCA +ATGGCT +ATGGCC +ATGGCG +ATGGGA +ATGGGT +ATGGGC +ATGGGG +ACAAAA +ACAAAT +ACAAAC +ACAAAG +ACAATA +ACAATT +ACAATC +ACAATG +ACAACA +ACAACT +ACAACC +ACAACG +ACAAGA +ACAAGT +ACAAGC +ACAAGG +ACATAA +ACATAT +ACATAC +ACATAG +ACATTA +ACATTT +ACATTC +ACATTG +ACATCA +ACATCT +ACATCC +ACATCG +ACATGA +ACATGT +ACATGC +ACATGG +ACACAA +ACACAT +ACACAC +ACACAG +ACACTA +ACACTT +ACACTC +ACACTG +ACACCA +ACACCT +ACACCC +ACACCG +ACACGA +ACACGT +ACACGC +ACACGG +ACAGAA +ACAGAT +ACAGAC +ACAGAG +ACAGTA +ACAGTT +ACAGTC +ACAGTG +ACAGCA +ACAGCT +ACAGCC +ACAGCG +ACAGGA +ACAGGT +ACAGGC +ACAGGG +ACTAAA +ACTAAT +ACTAAC +ACTAAG +ACTATA +ACTATT +ACTATC +ACTATG +ACTACA +ACTACT +ACTACC +ACTACG +ACTAGA +ACTAGT +ACTAGC +ACTAGG +ACTTAA +ACTTAT +ACTTAC +ACTTAG +ACTTTA +ACTTTT +ACTTTC +ACTTTG +ACTTCA +ACTTCT +ACTTCC +ACTTCG +ACTTGA +ACTTGT +ACTTGC +ACTTGG +ACTCAA +ACTCAT +ACTCAC +ACTCAG +ACTCTA +ACTCTT +ACTCTC +ACTCTG +ACTCCA +ACTCCT +ACTCCC +ACTCCG +ACTCGA +ACTCGT +ACTCGC +ACTCGG +ACTGAA +ACTGAT +ACTGAC +ACTGAG +ACTGTA +ACTGTT +ACTGTC +ACTGTG +ACTGCA +ACTGCT +ACTGCC +ACTGCG +ACTGGA +ACTGGT +ACTGGC +ACTGGG +ACCAAA +ACCAAT +ACCAAC +ACCAAG +ACCATA +ACCATT +ACCATC +ACCATG +ACCACA +ACCACT +ACCACC +ACCACG +ACCAGA +ACCAGT +ACCAGC +ACCAGG +ACCTAA +ACCTAT +ACCTAC +ACCTAG +ACCTTA +ACCTTT +ACCTTC +ACCTTG +ACCTCA +ACCTCT +ACCTCC +ACCTCG +ACCTGA +ACCTGT +ACCTGC +ACCTGG +ACCCAA +ACCCAT +ACCCAC +ACCCAG +ACCCTA +ACCCTT +ACCCTC +ACCCTG +ACCCCA +ACCCCT +ACCCCC +ACCCCG +ACCCGA +ACCCGT +ACCCGC +ACCCGG +ACCGAA +ACCGAT +ACCGAC +ACCGAG +ACCGTA +ACCGTT +ACCGTC +ACCGTG +ACCGCA +ACCGCT +ACCGCC +ACCGCG +ACCGGA +ACCGGT +ACCGGC +ACCGGG +ACGAAA +ACGAAT +ACGAAC +ACGAAG +ACGATA +ACGATT +ACGATC +ACGATG +ACGACA +ACGACT +ACGACC +ACGACG +ACGAGA +ACGAGT +ACGAGC +ACGAGG +ACGTAA +ACGTAT +ACGTAC +ACGTAG +ACGTTA +ACGTTT +ACGTTC +ACGTTG +ACGTCA +ACGTCT +ACGTCC +ACGTCG +ACGTGA +ACGTGT +ACGTGC +ACGTGG +ACGCAA +ACGCAT +ACGCAC +ACGCAG +ACGCTA +ACGCTT +ACGCTC +ACGCTG +ACGCCA +ACGCCT +ACGCCC +ACGCCG +ACGCGA +ACGCGT +ACGCGC +ACGCGG +ACGGAA +ACGGAT +ACGGAC +ACGGAG +ACGGTA +ACGGTT +ACGGTC +ACGGTG +ACGGCA +ACGGCT +ACGGCC +ACGGCG +ACGGGA +ACGGGT +ACGGGC +ACGGGG +AGAAAA +AGAAAT +AGAAAC +AGAAAG +AGAATA +AGAATT +AGAATC +AGAATG +AGAACA +AGAACT +AGAACC +AGAACG +AGAAGA +AGAAGT +AGAAGC +AGAAGG +AGATAA +AGATAT +AGATAC +AGATAG +AGATTA +AGATTT +AGATTC +AGATTG +AGATCA +AGATCT +AGATCC +AGATCG +AGATGA +AGATGT +AGATGC +AGATGG +AGACAA +AGACAT +AGACAC +AGACAG +AGACTA +AGACTT +AGACTC +AGACTG +AGACCA +AGACCT +AGACCC +AGACCG +AGACGA +AGACGT +AGACGC +AGACGG +AGAGAA +AGAGAT +AGAGAC +AGAGAG +AGAGTA +AGAGTT +AGAGTC +AGAGTG +AGAGCA +AGAGCT +AGAGCC +AGAGCG +AGAGGA +AGAGGT +AGAGGC +AGAGGG +AGTAAA +AGTAAT +AGTAAC +AGTAAG +AGTATA +AGTATT +AGTATC +AGTATG +AGTACA +AGTACT +AGTACC +AGTACG +AGTAGA +AGTAGT +AGTAGC +AGTAGG +AGTTAA +AGTTAT +AGTTAC +AGTTAG +AGTTTA +AGTTTT +AGTTTC +AGTTTG +AGTTCA +AGTTCT +AGTTCC +AGTTCG +AGTTGA +AGTTGT +AGTTGC +AGTTGG +AGTCAA +AGTCAT +AGTCAC +AGTCAG +AGTCTA +AGTCTT +AGTCTC +AGTCTG +AGTCCA +AGTCCT +AGTCCC +AGTCCG +AGTCGA +AGTCGT +AGTCGC +AGTCGG +AGTGAA +AGTGAT +AGTGAC +AGTGAG +AGTGTA +AGTGTT +AGTGTC +AGTGTG +AGTGCA +AGTGCT +AGTGCC +AGTGCG +AGTGGA +AGTGGT +AGTGGC +AGTGGG +AGCAAA +AGCAAT +AGCAAC +AGCAAG +AGCATA +AGCATT +AGCATC +AGCATG +AGCACA +AGCACT +AGCACC +AGCACG +AGCAGA +AGCAGT +AGCAGC +AGCAGG +AGCTAA +AGCTAT +AGCTAC +AGCTAG +AGCTTA +AGCTTT +AGCTTC +AGCTTG +AGCTCA +AGCTCT +AGCTCC +AGCTCG +AGCTGA +AGCTGT +AGCTGC +AGCTGG +AGCCAA +AGCCAT +AGCCAC +AGCCAG +AGCCTA +AGCCTT +AGCCTC +AGCCTG +AGCCCA +AGCCCT +AGCCCC +AGCCCG +AGCCGA +AGCCGT +AGCCGC +AGCCGG +AGCGAA +AGCGAT +AGCGAC +AGCGAG +AGCGTA +AGCGTT +AGCGTC +AGCGTG +AGCGCA +AGCGCT +AGCGCC +AGCGCG +AGCGGA +AGCGGT +AGCGGC +AGCGGG +AGGAAA +AGGAAT +AGGAAC +AGGAAG +AGGATA +AGGATT +AGGATC +AGGATG +AGGACA +AGGACT +AGGACC +AGGACG +AGGAGA +AGGAGT +AGGAGC +AGGAGG +AGGTAA +AGGTAT +AGGTAC +AGGTAG +AGGTTA +AGGTTT +AGGTTC +AGGTTG +AGGTCA +AGGTCT +AGGTCC +AGGTCG +AGGTGA +AGGTGT +AGGTGC +AGGTGG +AGGCAA +AGGCAT +AGGCAC +AGGCAG +AGGCTA +AGGCTT +AGGCTC +AGGCTG +AGGCCA +AGGCCT +AGGCCC +AGGCCG +AGGCGA +AGGCGT +AGGCGC +AGGCGG +AGGGAA +AGGGAT +AGGGAC +AGGGAG +AGGGTA +AGGGTT +AGGGTC +AGGGTG +AGGGCA +AGGGCT +AGGGCC +AGGGCG +AGGGGA +AGGGGT +AGGGGC +AGGGGG +TAAAAA +TAAAAT +TAAAAC +TAAAAG +TAAATA +TAAATT +TAAATC +TAAATG +TAAACA +TAAACT +TAAACC +TAAACG +TAAAGA +TAAAGT +TAAAGC +TAAAGG +TAATAA +TAATAT +TAATAC +TAATAG +TAATTA +TAATTT +TAATTC +TAATTG +TAATCA +TAATCT +TAATCC +TAATCG +TAATGA +TAATGT +TAATGC +TAATGG +TAACAA +TAACAT +TAACAC +TAACAG +TAACTA +TAACTT +TAACTC +TAACTG +TAACCA +TAACCT +TAACCC +TAACCG +TAACGA +TAACGT +TAACGC +TAACGG +TAAGAA +TAAGAT +TAAGAC +TAAGAG +TAAGTA +TAAGTT +TAAGTC +TAAGTG +TAAGCA +TAAGCT +TAAGCC +TAAGCG +TAAGGA +TAAGGT +TAAGGC +TAAGGG +TATAAA +TATAAT +TATAAC +TATAAG +TATATA +TATATT +TATATC +TATATG +TATACA +TATACT +TATACC +TATACG +TATAGA +TATAGT +TATAGC +TATAGG +TATTAA +TATTAT +TATTAC +TATTAG +TATTTA +TATTTT +TATTTC +TATTTG +TATTCA +TATTCT +TATTCC +TATTCG +TATTGA +TATTGT +TATTGC +TATTGG +TATCAA +TATCAT +TATCAC +TATCAG +TATCTA +TATCTT +TATCTC +TATCTG +TATCCA +TATCCT +TATCCC +TATCCG +TATCGA +TATCGT +TATCGC +TATCGG +TATGAA +TATGAT +TATGAC +TATGAG +TATGTA +TATGTT +TATGTC +TATGTG +TATGCA +TATGCT +TATGCC +TATGCG +TATGGA +TATGGT +TATGGC +TATGGG +TACAAA +TACAAT +TACAAC +TACAAG +TACATA +TACATT +TACATC +TACATG +TACACA +TACACT +TACACC +TACACG +TACAGA +TACAGT +TACAGC +TACAGG +TACTAA +TACTAT +TACTAC +TACTAG +TACTTA +TACTTT +TACTTC +TACTTG +TACTCA +TACTCT +TACTCC +TACTCG +TACTGA +TACTGT +TACTGC +TACTGG +TACCAA +TACCAT +TACCAC +TACCAG +TACCTA +TACCTT +TACCTC +TACCTG +TACCCA +TACCCT +TACCCC +TACCCG +TACCGA +TACCGT +TACCGC +TACCGG +TACGAA +TACGAT +TACGAC +TACGAG +TACGTA +TACGTT +TACGTC +TACGTG +TACGCA +TACGCT +TACGCC +TACGCG +TACGGA +TACGGT +TACGGC +TACGGG +TAGAAA +TAGAAT +TAGAAC +TAGAAG +TAGATA +TAGATT +TAGATC +TAGATG +TAGACA +TAGACT +TAGACC +TAGACG +TAGAGA +TAGAGT +TAGAGC +TAGAGG +TAGTAA +TAGTAT +TAGTAC +TAGTAG +TAGTTA +TAGTTT +TAGTTC +TAGTTG +TAGTCA +TAGTCT +TAGTCC +TAGTCG +TAGTGA +TAGTGT +TAGTGC +TAGTGG +TAGCAA +TAGCAT +TAGCAC +TAGCAG +TAGCTA +TAGCTT +TAGCTC +TAGCTG +TAGCCA +TAGCCT +TAGCCC +TAGCCG +TAGCGA +TAGCGT +TAGCGC +TAGCGG +TAGGAA +TAGGAT +TAGGAC +TAGGAG +TAGGTA +TAGGTT +TAGGTC +TAGGTG +TAGGCA +TAGGCT +TAGGCC +TAGGCG +TAGGGA +TAGGGT +TAGGGC +TAGGGG +TTAAAA +TTAAAT +TTAAAC +TTAAAG +TTAATA +TTAATT +TTAATC +TTAATG +TTAACA +TTAACT +TTAACC +TTAACG +TTAAGA +TTAAGT +TTAAGC +TTAAGG +TTATAA +TTATAT +TTATAC +TTATAG +TTATTA +TTATTT +TTATTC +TTATTG +TTATCA +TTATCT +TTATCC +TTATCG +TTATGA +TTATGT +TTATGC +TTATGG +TTACAA +TTACAT +TTACAC +TTACAG +TTACTA +TTACTT +TTACTC +TTACTG +TTACCA +TTACCT +TTACCC +TTACCG +TTACGA +TTACGT +TTACGC +TTACGG +TTAGAA +TTAGAT +TTAGAC +TTAGAG +TTAGTA +TTAGTT +TTAGTC +TTAGTG +TTAGCA +TTAGCT +TTAGCC +TTAGCG +TTAGGA +TTAGGT +TTAGGC +TTAGGG +TTTAAA +TTTAAT +TTTAAC +TTTAAG +TTTATA +TTTATT +TTTATC +TTTATG +TTTACA +TTTACT +TTTACC +TTTACG +TTTAGA +TTTAGT +TTTAGC +TTTAGG +TTTTAA +TTTTAT +TTTTAC +TTTTAG +TTTTTA +TTTTTT +TTTTTC +TTTTTG +TTTTCA +TTTTCT +TTTTCC +TTTTCG +TTTTGA +TTTTGT +TTTTGC +TTTTGG +TTTCAA +TTTCAT +TTTCAC +TTTCAG +TTTCTA +TTTCTT +TTTCTC +TTTCTG +TTTCCA +TTTCCT +TTTCCC +TTTCCG +TTTCGA +TTTCGT +TTTCGC +TTTCGG +TTTGAA +TTTGAT +TTTGAC +TTTGAG +TTTGTA +TTTGTT +TTTGTC +TTTGTG +TTTGCA +TTTGCT +TTTGCC +TTTGCG +TTTGGA +TTTGGT +TTTGGC +TTTGGG +TTCAAA +TTCAAT +TTCAAC +TTCAAG +TTCATA +TTCATT +TTCATC +TTCATG +TTCACA +TTCACT +TTCACC +TTCACG +TTCAGA +TTCAGT +TTCAGC +TTCAGG +TTCTAA +TTCTAT +TTCTAC +TTCTAG +TTCTTA +TTCTTT +TTCTTC +TTCTTG +TTCTCA +TTCTCT +TTCTCC +TTCTCG +TTCTGA +TTCTGT +TTCTGC +TTCTGG +TTCCAA +TTCCAT +TTCCAC +TTCCAG +TTCCTA +TTCCTT +TTCCTC +TTCCTG +TTCCCA +TTCCCT +TTCCCC +TTCCCG +TTCCGA +TTCCGT +TTCCGC +TTCCGG +TTCGAA +TTCGAT +TTCGAC +TTCGAG +TTCGTA +TTCGTT +TTCGTC +TTCGTG +TTCGCA +TTCGCT +TTCGCC +TTCGCG +TTCGGA +TTCGGT +TTCGGC +TTCGGG +TTGAAA +TTGAAT +TTGAAC +TTGAAG +TTGATA +TTGATT +TTGATC +TTGATG +TTGACA +TTGACT +TTGACC +TTGACG +TTGAGA +TTGAGT +TTGAGC +TTGAGG +TTGTAA +TTGTAT +TTGTAC +TTGTAG +TTGTTA +TTGTTT +TTGTTC +TTGTTG +TTGTCA +TTGTCT +TTGTCC +TTGTCG +TTGTGA +TTGTGT +TTGTGC +TTGTGG +TTGCAA +TTGCAT +TTGCAC +TTGCAG +TTGCTA +TTGCTT +TTGCTC +TTGCTG +TTGCCA +TTGCCT +TTGCCC +TTGCCG +TTGCGA +TTGCGT +TTGCGC +TTGCGG +TTGGAA +TTGGAT +TTGGAC +TTGGAG +TTGGTA +TTGGTT +TTGGTC +TTGGTG +TTGGCA +TTGGCT +TTGGCC +TTGGCG +TTGGGA +TTGGGT +TTGGGC +TTGGGG +TCAAAA +TCAAAT +TCAAAC +TCAAAG +TCAATA +TCAATT +TCAATC +TCAATG +TCAACA +TCAACT +TCAACC +TCAACG +TCAAGA +TCAAGT +TCAAGC +TCAAGG +TCATAA +TCATAT +TCATAC +TCATAG +TCATTA +TCATTT +TCATTC +TCATTG +TCATCA +TCATCT +TCATCC +TCATCG +TCATGA +TCATGT +TCATGC +TCATGG +TCACAA +TCACAT +TCACAC +TCACAG +TCACTA +TCACTT +TCACTC +TCACTG +TCACCA +TCACCT +TCACCC +TCACCG +TCACGA +TCACGT +TCACGC +TCACGG +TCAGAA +TCAGAT +TCAGAC +TCAGAG +TCAGTA +TCAGTT +TCAGTC +TCAGTG +TCAGCA +TCAGCT +TCAGCC +TCAGCG +TCAGGA +TCAGGT +TCAGGC +TCAGGG +TCTAAA +TCTAAT +TCTAAC +TCTAAG +TCTATA +TCTATT +TCTATC +TCTATG +TCTACA +TCTACT +TCTACC +TCTACG +TCTAGA +TCTAGT +TCTAGC +TCTAGG +TCTTAA +TCTTAT +TCTTAC +TCTTAG +TCTTTA +TCTTTT +TCTTTC +TCTTTG +TCTTCA +TCTTCT +TCTTCC +TCTTCG +TCTTGA +TCTTGT +TCTTGC +TCTTGG +TCTCAA +TCTCAT +TCTCAC +TCTCAG +TCTCTA +TCTCTT +TCTCTC +TCTCTG +TCTCCA +TCTCCT +TCTCCC +TCTCCG +TCTCGA +TCTCGT +TCTCGC +TCTCGG +TCTGAA +TCTGAT +TCTGAC +TCTGAG +TCTGTA +TCTGTT +TCTGTC +TCTGTG +TCTGCA +TCTGCT +TCTGCC +TCTGCG +TCTGGA +TCTGGT +TCTGGC +TCTGGG +TCCAAA +TCCAAT +TCCAAC +TCCAAG +TCCATA +TCCATT +TCCATC +TCCATG +TCCACA +TCCACT +TCCACC +TCCACG +TCCAGA +TCCAGT +TCCAGC +TCCAGG +TCCTAA +TCCTAT +TCCTAC +TCCTAG +TCCTTA +TCCTTT +TCCTTC +TCCTTG +TCCTCA +TCCTCT +TCCTCC +TCCTCG +TCCTGA +TCCTGT +TCCTGC +TCCTGG +TCCCAA +TCCCAT +TCCCAC +TCCCAG +TCCCTA +TCCCTT +TCCCTC +TCCCTG +TCCCCA +TCCCCT +TCCCCC +TCCCCG +TCCCGA +TCCCGT +TCCCGC +TCCCGG +TCCGAA +TCCGAT +TCCGAC +TCCGAG +TCCGTA +TCCGTT +TCCGTC +TCCGTG +TCCGCA +TCCGCT +TCCGCC +TCCGCG +TCCGGA +TCCGGT +TCCGGC +TCCGGG +TCGAAA +TCGAAT +TCGAAC +TCGAAG +TCGATA +TCGATT +TCGATC +TCGATG +TCGACA +TCGACT +TCGACC +TCGACG +TCGAGA +TCGAGT +TCGAGC +TCGAGG +TCGTAA +TCGTAT +TCGTAC +TCGTAG +TCGTTA +TCGTTT +TCGTTC +TCGTTG +TCGTCA +TCGTCT +TCGTCC +TCGTCG +TCGTGA +TCGTGT +TCGTGC +TCGTGG +TCGCAA +TCGCAT +TCGCAC +TCGCAG +TCGCTA +TCGCTT +TCGCTC +TCGCTG +TCGCCA +TCGCCT +TCGCCC +TCGCCG +TCGCGA +TCGCGT +TCGCGC +TCGCGG +TCGGAA +TCGGAT +TCGGAC +TCGGAG +TCGGTA +TCGGTT +TCGGTC +TCGGTG +TCGGCA +TCGGCT +TCGGCC +TCGGCG +TCGGGA +TCGGGT +TCGGGC +TCGGGG +TGAAAA +TGAAAT +TGAAAC +TGAAAG +TGAATA +TGAATT +TGAATC +TGAATG +TGAACA +TGAACT +TGAACC +TGAACG +TGAAGA +TGAAGT +TGAAGC +TGAAGG +TGATAA +TGATAT +TGATAC +TGATAG +TGATTA +TGATTT +TGATTC +TGATTG +TGATCA +TGATCT +TGATCC +TGATCG +TGATGA +TGATGT +TGATGC +TGATGG +TGACAA +TGACAT +TGACAC +TGACAG +TGACTA +TGACTT +TGACTC +TGACTG +TGACCA +TGACCT +TGACCC +TGACCG +TGACGA +TGACGT +TGACGC +TGACGG +TGAGAA +TGAGAT +TGAGAC +TGAGAG +TGAGTA +TGAGTT +TGAGTC +TGAGTG +TGAGCA +TGAGCT +TGAGCC +TGAGCG +TGAGGA +TGAGGT +TGAGGC +TGAGGG +TGTAAA +TGTAAT +TGTAAC +TGTAAG +TGTATA +TGTATT +TGTATC +TGTATG +TGTACA +TGTACT +TGTACC +TGTACG +TGTAGA +TGTAGT +TGTAGC +TGTAGG +TGTTAA +TGTTAT +TGTTAC +TGTTAG +TGTTTA +TGTTTT +TGTTTC +TGTTTG +TGTTCA +TGTTCT +TGTTCC +TGTTCG +TGTTGA +TGTTGT +TGTTGC +TGTTGG +TGTCAA +TGTCAT +TGTCAC +TGTCAG +TGTCTA +TGTCTT +TGTCTC +TGTCTG +TGTCCA +TGTCCT +TGTCCC +TGTCCG +TGTCGA +TGTCGT +TGTCGC +TGTCGG +TGTGAA +TGTGAT +TGTGAC +TGTGAG +TGTGTA +TGTGTT +TGTGTC +TGTGTG +TGTGCA +TGTGCT +TGTGCC +TGTGCG +TGTGGA +TGTGGT +TGTGGC +TGTGGG +TGCAAA +TGCAAT +TGCAAC +TGCAAG +TGCATA +TGCATT +TGCATC +TGCATG +TGCACA +TGCACT +TGCACC +TGCACG +TGCAGA +TGCAGT +TGCAGC +TGCAGG +TGCTAA +TGCTAT +TGCTAC +TGCTAG +TGCTTA +TGCTTT +TGCTTC +TGCTTG +TGCTCA +TGCTCT +TGCTCC +TGCTCG +TGCTGA +TGCTGT +TGCTGC +TGCTGG +TGCCAA +TGCCAT +TGCCAC +TGCCAG +TGCCTA +TGCCTT +TGCCTC +TGCCTG +TGCCCA +TGCCCT +TGCCCC +TGCCCG +TGCCGA +TGCCGT +TGCCGC +TGCCGG +TGCGAA +TGCGAT +TGCGAC +TGCGAG +TGCGTA +TGCGTT +TGCGTC +TGCGTG +TGCGCA +TGCGCT +TGCGCC +TGCGCG +TGCGGA +TGCGGT +TGCGGC +TGCGGG +TGGAAA +TGGAAT +TGGAAC +TGGAAG +TGGATA +TGGATT +TGGATC +TGGATG +TGGACA +TGGACT +TGGACC +TGGACG +TGGAGA +TGGAGT +TGGAGC +TGGAGG +TGGTAA +TGGTAT +TGGTAC +TGGTAG +TGGTTA +TGGTTT +TGGTTC +TGGTTG +TGGTCA +TGGTCT +TGGTCC +TGGTCG +TGGTGA +TGGTGT +TGGTGC +TGGTGG +TGGCAA +TGGCAT +TGGCAC +TGGCAG +TGGCTA +TGGCTT +TGGCTC +TGGCTG +TGGCCA +TGGCCT +TGGCCC +TGGCCG +TGGCGA +TGGCGT +TGGCGC +TGGCGG +TGGGAA +TGGGAT +TGGGAC +TGGGAG +TGGGTA +TGGGTT +TGGGTC +TGGGTG +TGGGCA +TGGGCT +TGGGCC +TGGGCG +TGGGGA +TGGGGT +TGGGGC +TGGGGG +CAAAAA +CAAAAT +CAAAAC +CAAAAG +CAAATA +CAAATT +CAAATC +CAAATG +CAAACA +CAAACT +CAAACC +CAAACG +CAAAGA +CAAAGT +CAAAGC +CAAAGG +CAATAA +CAATAT +CAATAC +CAATAG +CAATTA +CAATTT +CAATTC +CAATTG +CAATCA +CAATCT +CAATCC +CAATCG +CAATGA +CAATGT +CAATGC +CAATGG +CAACAA +CAACAT +CAACAC +CAACAG +CAACTA +CAACTT +CAACTC +CAACTG +CAACCA +CAACCT +CAACCC +CAACCG +CAACGA +CAACGT +CAACGC +CAACGG +CAAGAA +CAAGAT +CAAGAC +CAAGAG +CAAGTA +CAAGTT +CAAGTC +CAAGTG +CAAGCA +CAAGCT +CAAGCC +CAAGCG +CAAGGA +CAAGGT +CAAGGC +CAAGGG +CATAAA +CATAAT +CATAAC +CATAAG +CATATA +CATATT +CATATC +CATATG +CATACA +CATACT +CATACC +CATACG +CATAGA +CATAGT +CATAGC +CATAGG +CATTAA +CATTAT +CATTAC +CATTAG +CATTTA +CATTTT +CATTTC +CATTTG +CATTCA +CATTCT +CATTCC +CATTCG +CATTGA +CATTGT +CATTGC +CATTGG +CATCAA +CATCAT +CATCAC +CATCAG +CATCTA +CATCTT +CATCTC +CATCTG +CATCCA +CATCCT +CATCCC +CATCCG +CATCGA +CATCGT +CATCGC +CATCGG +CATGAA +CATGAT +CATGAC +CATGAG +CATGTA +CATGTT +CATGTC +CATGTG +CATGCA +CATGCT +CATGCC +CATGCG +CATGGA +CATGGT +CATGGC +CATGGG +CACAAA +CACAAT +CACAAC +CACAAG +CACATA +CACATT +CACATC +CACATG +CACACA +CACACT +CACACC +CACACG +CACAGA +CACAGT +CACAGC +CACAGG +CACTAA +CACTAT +CACTAC +CACTAG +CACTTA +CACTTT +CACTTC +CACTTG +CACTCA +CACTCT +CACTCC +CACTCG +CACTGA +CACTGT +CACTGC +CACTGG +CACCAA +CACCAT +CACCAC +CACCAG +CACCTA +CACCTT +CACCTC +CACCTG +CACCCA +CACCCT +CACCCC +CACCCG +CACCGA +CACCGT +CACCGC +CACCGG +CACGAA +CACGAT +CACGAC +CACGAG +CACGTA +CACGTT +CACGTC +CACGTG +CACGCA +CACGCT +CACGCC +CACGCG +CACGGA +CACGGT +CACGGC +CACGGG +CAGAAA +CAGAAT +CAGAAC +CAGAAG +CAGATA +CAGATT +CAGATC +CAGATG +CAGACA +CAGACT +CAGACC +CAGACG +CAGAGA +CAGAGT +CAGAGC +CAGAGG +CAGTAA +CAGTAT +CAGTAC +CAGTAG +CAGTTA +CAGTTT +CAGTTC +CAGTTG +CAGTCA +CAGTCT +CAGTCC +CAGTCG +CAGTGA +CAGTGT +CAGTGC +CAGTGG +CAGCAA +CAGCAT +CAGCAC +CAGCAG +CAGCTA +CAGCTT +CAGCTC +CAGCTG +CAGCCA +CAGCCT +CAGCCC +CAGCCG +CAGCGA +CAGCGT +CAGCGC +CAGCGG +CAGGAA +CAGGAT +CAGGAC +CAGGAG +CAGGTA +CAGGTT +CAGGTC +CAGGTG +CAGGCA +CAGGCT +CAGGCC +CAGGCG +CAGGGA +CAGGGT +CAGGGC +CAGGGG +CTAAAA +CTAAAT +CTAAAC +CTAAAG +CTAATA +CTAATT +CTAATC +CTAATG +CTAACA +CTAACT +CTAACC +CTAACG +CTAAGA +CTAAGT +CTAAGC +CTAAGG +CTATAA +CTATAT +CTATAC +CTATAG +CTATTA +CTATTT +CTATTC +CTATTG +CTATCA +CTATCT +CTATCC +CTATCG +CTATGA +CTATGT +CTATGC +CTATGG +CTACAA +CTACAT +CTACAC +CTACAG +CTACTA +CTACTT +CTACTC +CTACTG +CTACCA +CTACCT +CTACCC +CTACCG +CTACGA +CTACGT +CTACGC +CTACGG +CTAGAA +CTAGAT +CTAGAC +CTAGAG +CTAGTA +CTAGTT +CTAGTC +CTAGTG +CTAGCA +CTAGCT +CTAGCC +CTAGCG +CTAGGA +CTAGGT +CTAGGC +CTAGGG +CTTAAA +CTTAAT +CTTAAC +CTTAAG +CTTATA +CTTATT +CTTATC +CTTATG +CTTACA +CTTACT +CTTACC +CTTACG +CTTAGA +CTTAGT +CTTAGC +CTTAGG +CTTTAA +CTTTAT +CTTTAC +CTTTAG +CTTTTA +CTTTTT +CTTTTC +CTTTTG +CTTTCA +CTTTCT +CTTTCC +CTTTCG +CTTTGA +CTTTGT +CTTTGC +CTTTGG +CTTCAA +CTTCAT +CTTCAC +CTTCAG +CTTCTA +CTTCTT +CTTCTC +CTTCTG +CTTCCA +CTTCCT +CTTCCC +CTTCCG +CTTCGA +CTTCGT +CTTCGC +CTTCGG +CTTGAA +CTTGAT +CTTGAC +CTTGAG +CTTGTA +CTTGTT +CTTGTC +CTTGTG +CTTGCA +CTTGCT +CTTGCC +CTTGCG +CTTGGA +CTTGGT +CTTGGC +CTTGGG +CTCAAA +CTCAAT +CTCAAC +CTCAAG +CTCATA +CTCATT +CTCATC +CTCATG +CTCACA +CTCACT +CTCACC +CTCACG +CTCAGA +CTCAGT +CTCAGC +CTCAGG +CTCTAA +CTCTAT +CTCTAC +CTCTAG +CTCTTA +CTCTTT +CTCTTC +CTCTTG +CTCTCA +CTCTCT +CTCTCC +CTCTCG +CTCTGA +CTCTGT +CTCTGC +CTCTGG +CTCCAA +CTCCAT +CTCCAC +CTCCAG +CTCCTA +CTCCTT +CTCCTC +CTCCTG +CTCCCA +CTCCCT +CTCCCC +CTCCCG +CTCCGA +CTCCGT +CTCCGC +CTCCGG +CTCGAA +CTCGAT +CTCGAC +CTCGAG +CTCGTA +CTCGTT +CTCGTC +CTCGTG +CTCGCA +CTCGCT +CTCGCC +CTCGCG +CTCGGA +CTCGGT +CTCGGC +CTCGGG +CTGAAA +CTGAAT +CTGAAC +CTGAAG +CTGATA +CTGATT +CTGATC +CTGATG +CTGACA +CTGACT +CTGACC +CTGACG +CTGAGA +CTGAGT +CTGAGC +CTGAGG +CTGTAA +CTGTAT +CTGTAC +CTGTAG +CTGTTA +CTGTTT +CTGTTC +CTGTTG +CTGTCA +CTGTCT +CTGTCC +CTGTCG +CTGTGA +CTGTGT +CTGTGC +CTGTGG +CTGCAA +CTGCAT +CTGCAC +CTGCAG +CTGCTA +CTGCTT +CTGCTC +CTGCTG +CTGCCA +CTGCCT +CTGCCC +CTGCCG +CTGCGA +CTGCGT +CTGCGC +CTGCGG +CTGGAA +CTGGAT +CTGGAC +CTGGAG +CTGGTA +CTGGTT +CTGGTC +CTGGTG +CTGGCA +CTGGCT +CTGGCC +CTGGCG +CTGGGA +CTGGGT +CTGGGC +CTGGGG +CCAAAA +CCAAAT +CCAAAC +CCAAAG +CCAATA +CCAATT +CCAATC +CCAATG +CCAACA +CCAACT +CCAACC +CCAACG +CCAAGA +CCAAGT +CCAAGC +CCAAGG +CCATAA +CCATAT +CCATAC +CCATAG +CCATTA +CCATTT +CCATTC +CCATTG +CCATCA +CCATCT +CCATCC +CCATCG +CCATGA +CCATGT +CCATGC +CCATGG +CCACAA +CCACAT +CCACAC +CCACAG +CCACTA +CCACTT +CCACTC +CCACTG +CCACCA +CCACCT +CCACCC +CCACCG +CCACGA +CCACGT +CCACGC +CCACGG +CCAGAA +CCAGAT +CCAGAC +CCAGAG +CCAGTA +CCAGTT +CCAGTC +CCAGTG +CCAGCA +CCAGCT +CCAGCC +CCAGCG +CCAGGA +CCAGGT +CCAGGC +CCAGGG +CCTAAA +CCTAAT +CCTAAC +CCTAAG +CCTATA +CCTATT +CCTATC +CCTATG +CCTACA +CCTACT +CCTACC +CCTACG +CCTAGA +CCTAGT +CCTAGC +CCTAGG +CCTTAA +CCTTAT +CCTTAC +CCTTAG +CCTTTA +CCTTTT +CCTTTC +CCTTTG +CCTTCA +CCTTCT +CCTTCC +CCTTCG +CCTTGA +CCTTGT +CCTTGC +CCTTGG +CCTCAA +CCTCAT +CCTCAC +CCTCAG +CCTCTA +CCTCTT +CCTCTC +CCTCTG +CCTCCA +CCTCCT +CCTCCC +CCTCCG +CCTCGA +CCTCGT +CCTCGC +CCTCGG +CCTGAA +CCTGAT +CCTGAC +CCTGAG +CCTGTA +CCTGTT +CCTGTC +CCTGTG +CCTGCA +CCTGCT +CCTGCC +CCTGCG +CCTGGA +CCTGGT +CCTGGC +CCTGGG +CCCAAA +CCCAAT +CCCAAC +CCCAAG +CCCATA +CCCATT +CCCATC +CCCATG +CCCACA +CCCACT +CCCACC +CCCACG +CCCAGA +CCCAGT +CCCAGC +CCCAGG +CCCTAA +CCCTAT +CCCTAC +CCCTAG +CCCTTA +CCCTTT +CCCTTC +CCCTTG +CCCTCA +CCCTCT +CCCTCC +CCCTCG +CCCTGA +CCCTGT +CCCTGC +CCCTGG +CCCCAA +CCCCAT +CCCCAC +CCCCAG +CCCCTA +CCCCTT +CCCCTC +CCCCTG +CCCCCA +CCCCCT +CCCCCC +CCCCCG +CCCCGA +CCCCGT +CCCCGC +CCCCGG +CCCGAA +CCCGAT +CCCGAC +CCCGAG +CCCGTA +CCCGTT +CCCGTC +CCCGTG +CCCGCA +CCCGCT +CCCGCC +CCCGCG +CCCGGA +CCCGGT +CCCGGC +CCCGGG +CCGAAA +CCGAAT +CCGAAC +CCGAAG +CCGATA +CCGATT +CCGATC +CCGATG +CCGACA +CCGACT +CCGACC +CCGACG +CCGAGA +CCGAGT +CCGAGC +CCGAGG +CCGTAA +CCGTAT +CCGTAC +CCGTAG +CCGTTA +CCGTTT +CCGTTC +CCGTTG +CCGTCA +CCGTCT +CCGTCC +CCGTCG +CCGTGA +CCGTGT +CCGTGC +CCGTGG +CCGCAA +CCGCAT +CCGCAC +CCGCAG +CCGCTA +CCGCTT +CCGCTC +CCGCTG +CCGCCA +CCGCCT +CCGCCC +CCGCCG +CCGCGA +CCGCGT +CCGCGC +CCGCGG +CCGGAA +CCGGAT +CCGGAC +CCGGAG +CCGGTA +CCGGTT +CCGGTC +CCGGTG +CCGGCA +CCGGCT +CCGGCC +CCGGCG +CCGGGA +CCGGGT +CCGGGC +CCGGGG +CGAAAA +CGAAAT +CGAAAC +CGAAAG +CGAATA +CGAATT +CGAATC +CGAATG +CGAACA +CGAACT +CGAACC +CGAACG +CGAAGA +CGAAGT +CGAAGC +CGAAGG +CGATAA +CGATAT +CGATAC +CGATAG +CGATTA +CGATTT +CGATTC +CGATTG +CGATCA +CGATCT +CGATCC +CGATCG +CGATGA +CGATGT +CGATGC +CGATGG +CGACAA +CGACAT +CGACAC +CGACAG +CGACTA +CGACTT +CGACTC +CGACTG +CGACCA +CGACCT +CGACCC +CGACCG +CGACGA +CGACGT +CGACGC +CGACGG +CGAGAA +CGAGAT +CGAGAC +CGAGAG +CGAGTA +CGAGTT +CGAGTC +CGAGTG +CGAGCA +CGAGCT +CGAGCC +CGAGCG +CGAGGA +CGAGGT +CGAGGC +CGAGGG +CGTAAA +CGTAAT +CGTAAC +CGTAAG +CGTATA +CGTATT +CGTATC +CGTATG +CGTACA +CGTACT +CGTACC +CGTACG +CGTAGA +CGTAGT +CGTAGC +CGTAGG +CGTTAA +CGTTAT +CGTTAC +CGTTAG +CGTTTA +CGTTTT +CGTTTC +CGTTTG +CGTTCA +CGTTCT +CGTTCC +CGTTCG +CGTTGA +CGTTGT +CGTTGC +CGTTGG +CGTCAA +CGTCAT +CGTCAC +CGTCAG +CGTCTA +CGTCTT +CGTCTC +CGTCTG +CGTCCA +CGTCCT +CGTCCC +CGTCCG +CGTCGA +CGTCGT +CGTCGC +CGTCGG +CGTGAA +CGTGAT +CGTGAC +CGTGAG +CGTGTA +CGTGTT +CGTGTC +CGTGTG +CGTGCA +CGTGCT +CGTGCC +CGTGCG +CGTGGA +CGTGGT +CGTGGC +CGTGGG +CGCAAA +CGCAAT +CGCAAC +CGCAAG +CGCATA +CGCATT +CGCATC +CGCATG +CGCACA +CGCACT +CGCACC +CGCACG +CGCAGA +CGCAGT +CGCAGC +CGCAGG +CGCTAA +CGCTAT +CGCTAC +CGCTAG +CGCTTA +CGCTTT +CGCTTC +CGCTTG +CGCTCA +CGCTCT +CGCTCC +CGCTCG +CGCTGA +CGCTGT +CGCTGC +CGCTGG +CGCCAA +CGCCAT +CGCCAC +CGCCAG +CGCCTA +CGCCTT +CGCCTC +CGCCTG +CGCCCA +CGCCCT +CGCCCC +CGCCCG +CGCCGA +CGCCGT +CGCCGC +CGCCGG +CGCGAA +CGCGAT +CGCGAC +CGCGAG +CGCGTA +CGCGTT +CGCGTC +CGCGTG +CGCGCA +CGCGCT +CGCGCC +CGCGCG +CGCGGA +CGCGGT +CGCGGC +CGCGGG +CGGAAA +CGGAAT +CGGAAC +CGGAAG +CGGATA +CGGATT +CGGATC +CGGATG +CGGACA +CGGACT +CGGACC +CGGACG +CGGAGA +CGGAGT +CGGAGC +CGGAGG +CGGTAA +CGGTAT +CGGTAC +CGGTAG +CGGTTA +CGGTTT +CGGTTC +CGGTTG +CGGTCA +CGGTCT +CGGTCC +CGGTCG +CGGTGA +CGGTGT +CGGTGC +CGGTGG +CGGCAA +CGGCAT +CGGCAC +CGGCAG +CGGCTA +CGGCTT +CGGCTC +CGGCTG +CGGCCA +CGGCCT +CGGCCC +CGGCCG +CGGCGA +CGGCGT +CGGCGC +CGGCGG +CGGGAA +CGGGAT +CGGGAC +CGGGAG +CGGGTA +CGGGTT +CGGGTC +CGGGTG +CGGGCA +CGGGCT +CGGGCC +CGGGCG +CGGGGA +CGGGGT +CGGGGC +CGGGGG +GAAAAA +GAAAAT +GAAAAC +GAAAAG +GAAATA +GAAATT +GAAATC +GAAATG +GAAACA +GAAACT +GAAACC +GAAACG +GAAAGA +GAAAGT +GAAAGC +GAAAGG +GAATAA +GAATAT +GAATAC +GAATAG +GAATTA +GAATTT +GAATTC +GAATTG +GAATCA +GAATCT +GAATCC +GAATCG +GAATGA +GAATGT +GAATGC +GAATGG +GAACAA +GAACAT +GAACAC +GAACAG +GAACTA +GAACTT +GAACTC +GAACTG +GAACCA +GAACCT +GAACCC +GAACCG +GAACGA +GAACGT +GAACGC +GAACGG +GAAGAA +GAAGAT +GAAGAC +GAAGAG +GAAGTA +GAAGTT +GAAGTC +GAAGTG +GAAGCA +GAAGCT +GAAGCC +GAAGCG +GAAGGA +GAAGGT +GAAGGC +GAAGGG +GATAAA +GATAAT +GATAAC +GATAAG +GATATA +GATATT +GATATC +GATATG +GATACA +GATACT +GATACC +GATACG +GATAGA +GATAGT +GATAGC +GATAGG +GATTAA +GATTAT +GATTAC +GATTAG +GATTTA +GATTTT +GATTTC +GATTTG +GATTCA +GATTCT +GATTCC +GATTCG +GATTGA +GATTGT +GATTGC +GATTGG +GATCAA +GATCAT +GATCAC +GATCAG +GATCTA +GATCTT +GATCTC +GATCTG +GATCCA +GATCCT +GATCCC +GATCCG +GATCGA +GATCGT +GATCGC +GATCGG +GATGAA +GATGAT +GATGAC +GATGAG +GATGTA +GATGTT +GATGTC +GATGTG +GATGCA +GATGCT +GATGCC +GATGCG +GATGGA +GATGGT +GATGGC +GATGGG +GACAAA +GACAAT +GACAAC +GACAAG +GACATA +GACATT +GACATC +GACATG +GACACA +GACACT +GACACC +GACACG +GACAGA +GACAGT +GACAGC +GACAGG +GACTAA +GACTAT +GACTAC +GACTAG +GACTTA +GACTTT +GACTTC +GACTTG +GACTCA +GACTCT +GACTCC +GACTCG +GACTGA +GACTGT +GACTGC +GACTGG +GACCAA +GACCAT +GACCAC +GACCAG +GACCTA +GACCTT +GACCTC +GACCTG +GACCCA +GACCCT +GACCCC +GACCCG +GACCGA +GACCGT +GACCGC +GACCGG +GACGAA +GACGAT +GACGAC +GACGAG +GACGTA +GACGTT +GACGTC +GACGTG +GACGCA +GACGCT +GACGCC +GACGCG +GACGGA +GACGGT +GACGGC +GACGGG +GAGAAA +GAGAAT +GAGAAC +GAGAAG +GAGATA +GAGATT +GAGATC +GAGATG +GAGACA +GAGACT +GAGACC +GAGACG +GAGAGA +GAGAGT +GAGAGC +GAGAGG +GAGTAA +GAGTAT +GAGTAC +GAGTAG +GAGTTA +GAGTTT +GAGTTC +GAGTTG +GAGTCA +GAGTCT +GAGTCC +GAGTCG +GAGTGA +GAGTGT +GAGTGC +GAGTGG +GAGCAA +GAGCAT +GAGCAC +GAGCAG +GAGCTA +GAGCTT +GAGCTC +GAGCTG +GAGCCA +GAGCCT +GAGCCC +GAGCCG +GAGCGA +GAGCGT +GAGCGC +GAGCGG +GAGGAA +GAGGAT +GAGGAC +GAGGAG +GAGGTA +GAGGTT +GAGGTC +GAGGTG +GAGGCA +GAGGCT +GAGGCC +GAGGCG +GAGGGA +GAGGGT +GAGGGC +GAGGGG +GTAAAA +GTAAAT +GTAAAC +GTAAAG +GTAATA +GTAATT +GTAATC +GTAATG +GTAACA +GTAACT +GTAACC +GTAACG +GTAAGA +GTAAGT +GTAAGC +GTAAGG +GTATAA +GTATAT +GTATAC +GTATAG +GTATTA +GTATTT +GTATTC +GTATTG +GTATCA +GTATCT +GTATCC +GTATCG +GTATGA +GTATGT +GTATGC +GTATGG +GTACAA +GTACAT +GTACAC +GTACAG +GTACTA +GTACTT +GTACTC +GTACTG +GTACCA +GTACCT +GTACCC +GTACCG +GTACGA +GTACGT +GTACGC +GTACGG +GTAGAA +GTAGAT +GTAGAC +GTAGAG +GTAGTA +GTAGTT +GTAGTC +GTAGTG +GTAGCA +GTAGCT +GTAGCC +GTAGCG +GTAGGA +GTAGGT +GTAGGC +GTAGGG +GTTAAA +GTTAAT +GTTAAC +GTTAAG +GTTATA +GTTATT +GTTATC +GTTATG +GTTACA +GTTACT +GTTACC +GTTACG +GTTAGA +GTTAGT +GTTAGC +GTTAGG +GTTTAA +GTTTAT +GTTTAC +GTTTAG +GTTTTA +GTTTTT +GTTTTC +GTTTTG +GTTTCA +GTTTCT +GTTTCC +GTTTCG +GTTTGA +GTTTGT +GTTTGC +GTTTGG +GTTCAA +GTTCAT +GTTCAC +GTTCAG +GTTCTA +GTTCTT +GTTCTC +GTTCTG +GTTCCA +GTTCCT +GTTCCC +GTTCCG +GTTCGA +GTTCGT +GTTCGC +GTTCGG +GTTGAA +GTTGAT +GTTGAC +GTTGAG +GTTGTA +GTTGTT +GTTGTC +GTTGTG +GTTGCA +GTTGCT +GTTGCC +GTTGCG +GTTGGA +GTTGGT +GTTGGC +GTTGGG +GTCAAA +GTCAAT +GTCAAC +GTCAAG +GTCATA +GTCATT +GTCATC +GTCATG +GTCACA +GTCACT +GTCACC +GTCACG +GTCAGA +GTCAGT +GTCAGC +GTCAGG +GTCTAA +GTCTAT +GTCTAC +GTCTAG +GTCTTA +GTCTTT +GTCTTC +GTCTTG +GTCTCA +GTCTCT +GTCTCC +GTCTCG +GTCTGA +GTCTGT +GTCTGC +GTCTGG +GTCCAA +GTCCAT +GTCCAC +GTCCAG +GTCCTA +GTCCTT +GTCCTC +GTCCTG +GTCCCA +GTCCCT +GTCCCC +GTCCCG +GTCCGA +GTCCGT +GTCCGC +GTCCGG +GTCGAA +GTCGAT +GTCGAC +GTCGAG +GTCGTA +GTCGTT +GTCGTC +GTCGTG +GTCGCA +GTCGCT +GTCGCC +GTCGCG +GTCGGA +GTCGGT +GTCGGC +GTCGGG +GTGAAA +GTGAAT +GTGAAC +GTGAAG +GTGATA +GTGATT +GTGATC +GTGATG +GTGACA +GTGACT +GTGACC +GTGACG +GTGAGA +GTGAGT +GTGAGC +GTGAGG +GTGTAA +GTGTAT +GTGTAC +GTGTAG +GTGTTA +GTGTTT +GTGTTC +GTGTTG +GTGTCA +GTGTCT +GTGTCC +GTGTCG +GTGTGA +GTGTGT +GTGTGC +GTGTGG +GTGCAA +GTGCAT +GTGCAC +GTGCAG +GTGCTA +GTGCTT +GTGCTC +GTGCTG +GTGCCA +GTGCCT +GTGCCC +GTGCCG +GTGCGA +GTGCGT +GTGCGC +GTGCGG +GTGGAA +GTGGAT +GTGGAC +GTGGAG +GTGGTA +GTGGTT +GTGGTC +GTGGTG +GTGGCA +GTGGCT +GTGGCC +GTGGCG +GTGGGA +GTGGGT +GTGGGC +GTGGGG +GCAAAA +GCAAAT +GCAAAC +GCAAAG +GCAATA +GCAATT +GCAATC +GCAATG +GCAACA +GCAACT +GCAACC +GCAACG +GCAAGA +GCAAGT +GCAAGC +GCAAGG +GCATAA +GCATAT +GCATAC +GCATAG +GCATTA +GCATTT +GCATTC +GCATTG +GCATCA +GCATCT +GCATCC +GCATCG +GCATGA +GCATGT +GCATGC +GCATGG +GCACAA +GCACAT +GCACAC +GCACAG +GCACTA +GCACTT +GCACTC +GCACTG +GCACCA +GCACCT +GCACCC +GCACCG +GCACGA +GCACGT +GCACGC +GCACGG +GCAGAA +GCAGAT +GCAGAC +GCAGAG +GCAGTA +GCAGTT +GCAGTC +GCAGTG +GCAGCA +GCAGCT +GCAGCC +GCAGCG +GCAGGA +GCAGGT +GCAGGC +GCAGGG +GCTAAA +GCTAAT +GCTAAC +GCTAAG +GCTATA +GCTATT +GCTATC +GCTATG +GCTACA +GCTACT +GCTACC +GCTACG +GCTAGA +GCTAGT +GCTAGC +GCTAGG +GCTTAA +GCTTAT +GCTTAC +GCTTAG +GCTTTA +GCTTTT +GCTTTC +GCTTTG +GCTTCA +GCTTCT +GCTTCC +GCTTCG +GCTTGA +GCTTGT +GCTTGC +GCTTGG +GCTCAA +GCTCAT +GCTCAC +GCTCAG +GCTCTA +GCTCTT +GCTCTC +GCTCTG +GCTCCA +GCTCCT +GCTCCC +GCTCCG +GCTCGA +GCTCGT +GCTCGC +GCTCGG +GCTGAA +GCTGAT +GCTGAC +GCTGAG +GCTGTA +GCTGTT +GCTGTC +GCTGTG +GCTGCA +GCTGCT +GCTGCC +GCTGCG +GCTGGA +GCTGGT +GCTGGC +GCTGGG +GCCAAA +GCCAAT +GCCAAC +GCCAAG +GCCATA +GCCATT +GCCATC +GCCATG +GCCACA +GCCACT +GCCACC +GCCACG +GCCAGA +GCCAGT +GCCAGC +GCCAGG +GCCTAA +GCCTAT +GCCTAC +GCCTAG +GCCTTA +GCCTTT +GCCTTC +GCCTTG +GCCTCA +GCCTCT +GCCTCC +GCCTCG +GCCTGA +GCCTGT +GCCTGC +GCCTGG +GCCCAA +GCCCAT +GCCCAC +GCCCAG +GCCCTA +GCCCTT +GCCCTC +GCCCTG +GCCCCA +GCCCCT +GCCCCC +GCCCCG +GCCCGA +GCCCGT +GCCCGC +GCCCGG +GCCGAA +GCCGAT +GCCGAC +GCCGAG +GCCGTA +GCCGTT +GCCGTC +GCCGTG +GCCGCA +GCCGCT +GCCGCC +GCCGCG +GCCGGA +GCCGGT +GCCGGC +GCCGGG +GCGAAA +GCGAAT +GCGAAC +GCGAAG +GCGATA +GCGATT +GCGATC +GCGATG +GCGACA +GCGACT +GCGACC +GCGACG +GCGAGA +GCGAGT +GCGAGC +GCGAGG +GCGTAA +GCGTAT +GCGTAC +GCGTAG +GCGTTA +GCGTTT +GCGTTC +GCGTTG +GCGTCA +GCGTCT +GCGTCC +GCGTCG +GCGTGA +GCGTGT +GCGTGC +GCGTGG +GCGCAA +GCGCAT +GCGCAC +GCGCAG +GCGCTA +GCGCTT +GCGCTC +GCGCTG +GCGCCA +GCGCCT +GCGCCC +GCGCCG +GCGCGA +GCGCGT +GCGCGC +GCGCGG +GCGGAA +GCGGAT +GCGGAC +GCGGAG +GCGGTA +GCGGTT +GCGGTC +GCGGTG +GCGGCA +GCGGCT +GCGGCC +GCGGCG +GCGGGA +GCGGGT +GCGGGC +GCGGGG +GGAAAA +GGAAAT +GGAAAC +GGAAAG +GGAATA +GGAATT +GGAATC +GGAATG +GGAACA +GGAACT +GGAACC +GGAACG +GGAAGA +GGAAGT +GGAAGC +GGAAGG +GGATAA +GGATAT +GGATAC +GGATAG +GGATTA +GGATTT +GGATTC +GGATTG +GGATCA +GGATCT +GGATCC +GGATCG +GGATGA +GGATGT +GGATGC +GGATGG +GGACAA +GGACAT +GGACAC +GGACAG +GGACTA +GGACTT +GGACTC +GGACTG +GGACCA +GGACCT +GGACCC +GGACCG +GGACGA +GGACGT +GGACGC +GGACGG +GGAGAA +GGAGAT +GGAGAC +GGAGAG +GGAGTA +GGAGTT +GGAGTC +GGAGTG +GGAGCA +GGAGCT +GGAGCC +GGAGCG +GGAGGA +GGAGGT +GGAGGC +GGAGGG +GGTAAA +GGTAAT +GGTAAC +GGTAAG +GGTATA +GGTATT +GGTATC +GGTATG +GGTACA +GGTACT +GGTACC +GGTACG +GGTAGA +GGTAGT +GGTAGC +GGTAGG +GGTTAA +GGTTAT +GGTTAC +GGTTAG +GGTTTA +GGTTTT +GGTTTC +GGTTTG +GGTTCA +GGTTCT +GGTTCC +GGTTCG +GGTTGA +GGTTGT +GGTTGC +GGTTGG +GGTCAA +GGTCAT +GGTCAC +GGTCAG +GGTCTA +GGTCTT +GGTCTC +GGTCTG +GGTCCA +GGTCCT +GGTCCC +GGTCCG +GGTCGA +GGTCGT +GGTCGC +GGTCGG +GGTGAA +GGTGAT +GGTGAC +GGTGAG +GGTGTA +GGTGTT +GGTGTC +GGTGTG +GGTGCA +GGTGCT +GGTGCC +GGTGCG +GGTGGA +GGTGGT +GGTGGC +GGTGGG +GGCAAA +GGCAAT +GGCAAC +GGCAAG +GGCATA +GGCATT +GGCATC +GGCATG +GGCACA +GGCACT +GGCACC +GGCACG +GGCAGA +GGCAGT +GGCAGC +GGCAGG +GGCTAA +GGCTAT +GGCTAC +GGCTAG +GGCTTA +GGCTTT +GGCTTC +GGCTTG +GGCTCA +GGCTCT +GGCTCC +GGCTCG +GGCTGA +GGCTGT +GGCTGC +GGCTGG +GGCCAA +GGCCAT +GGCCAC +GGCCAG +GGCCTA +GGCCTT +GGCCTC +GGCCTG +GGCCCA +GGCCCT +GGCCCC +GGCCCG +GGCCGA +GGCCGT +GGCCGC +GGCCGG +GGCGAA +GGCGAT +GGCGAC +GGCGAG +GGCGTA +GGCGTT +GGCGTC +GGCGTG +GGCGCA +GGCGCT +GGCGCC +GGCGCG +GGCGGA +GGCGGT +GGCGGC +GGCGGG +GGGAAA +GGGAAT +GGGAAC +GGGAAG +GGGATA +GGGATT +GGGATC +GGGATG +GGGACA +GGGACT +GGGACC +GGGACG +GGGAGA +GGGAGT +GGGAGC +GGGAGG +GGGTAA +GGGTAT +GGGTAC +GGGTAG +GGGTTA +GGGTTT +GGGTTC +GGGTTG +GGGTCA +GGGTCT +GGGTCC +GGGTCG +GGGTGA +GGGTGT +GGGTGC +GGGTGG +GGGCAA +GGGCAT +GGGCAC +GGGCAG +GGGCTA +GGGCTT +GGGCTC +GGGCTG +GGGCCA +GGGCCT +GGGCCC +GGGCCG +GGGCGA +GGGCGT +GGGCGC +GGGCGG +GGGGAA +GGGGAT +GGGGAC +GGGGAG +GGGGTA +GGGGTT +GGGGTC +GGGGTG +GGGGCA +GGGGCT +GGGGCC +GGGGCG +GGGGGA +GGGGGT +GGGGGC +GGGGGG diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/README.md b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..500fd976d92b217455f181bc98c77e612d9a18d9 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/README.md @@ -0,0 +1,202 @@ +--- +base_model: fabihamakhdoomi/TinyDNABERT +library_name: peft +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.12.0 \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/adapter_config.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..04c63e99bbc6b810449dece367c0afb165125b92 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/adapter_config.json @@ -0,0 +1,42 @@ +{ + "alpha_pattern": {}, + "auto_mapping": null, + "base_model_name_or_path": "fabihamakhdoomi/TinyDNABERT", + "bias": "none", + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 16, + "lora_dropout": 0.2, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": [ + "classifier", + "score" + ], + "peft_type": "LORA", + "r": 8, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "bert.encoder.layer.1.attention.self.query", + "bert.encoder.layer.0.output.dense", + "bert.encoder.layer.1.output.dense", + "bert.encoder.layer.0.intermediate.dense", + "bert.encoder.layer.1.intermediate.dense", + "bert.encoder.layer.0.attention.output.dense", + "bert.encoder.layer.1.attention.output.dense", + "bert.encoder.layer.0.attention.self.query", + "bert.encoder.layer.0.attention.self.value", + "bert.encoder.layer.1.attention.self.key", + "bert.encoder.layer.1.attention.self.value", + "bert.encoder.layer.0.attention.self.key" + ], + "task_type": "SEQ_CLS", + "use_dora": false, + "use_rslora": false +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/adapter_model.safetensors b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/adapter_model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..68c42bf3759d9e17d49c6ba28e96c33e17b0fe84 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/adapter_model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c63f37368e9d95220569ef0da46e49f03c91793ca1a0211616efc532b7f48deb +size 151952 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/optimizer.pt b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..5f5d895cadf9a189be5bc07292fee76f94ec08b2 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6ea872a9ffd82db3d109983975be5ff975ce844e054cb60a18604c70d42f8d68 +size 312682 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/rng_state.pth b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..ab07d6382b5d2f619b3927fc83d6571e1bcccc54 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f0f0d7ddc5f130ac1babf491a193bd5c163e7a1b572c7a4eac20de52cd6b2bd7 +size 14244 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/scheduler.pt b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..8c9a6adac55ac4d1afd6e1b61889589acfe31330 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f3e454de3c0f52ddb8e5884c9783308658a6f61ede2ed6e377ca9a68d47c4515 +size 1064 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/special_tokens_map.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..a8b3208c2884c4efb86e49300fdd3dc877220cdf --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/special_tokens_map.json @@ -0,0 +1,7 @@ +{ + "cls_token": "[CLS]", + "mask_token": "[MASK]", + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/tokenizer.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..ebd782de6963b9032ce9374cfe30f5275be1a30b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/tokenizer.json @@ -0,0 +1,4265 @@ +{ + "version": "1.0", + "truncation": { + "direction": "Right", + "max_length": 512, + "strategy": "LongestFirst", + "stride": 0 + }, + "padding": { + "strategy": { + "Fixed": 512 + }, + "direction": "Right", + "pad_to_multiple_of": null, + "pad_id": 0, + "pad_type_id": 0, + "pad_token": "[PAD]" + }, + "added_tokens": [ + { + "id": 0, + "content": "[PAD]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 1, + "content": "[UNK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 2, + "content": "[CLS]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 3, + "content": "[SEP]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 4, + "content": "[MASK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + } + ], + "normalizer": { + "type": "BertNormalizer", + "clean_text": true, + "handle_chinese_chars": true, + "strip_accents": null, + "lowercase": false + }, + "pre_tokenizer": { + "type": "BertPreTokenizer" + }, + "post_processor": { + "type": "TemplateProcessing", + "single": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + } + ], + "pair": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "B", + "type_id": 1 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 1 + } + } + ], + "special_tokens": { + "[CLS]": { + "id": "[CLS]", + "ids": [ + 2 + ], + "tokens": [ + "[CLS]" + ] + }, + "[SEP]": { + "id": "[SEP]", + "ids": [ + 3 + ], + "tokens": [ + "[SEP]" + ] + } + } + }, + "decoder": { + "type": "WordPiece", + "prefix": "##", + "cleanup": true + }, + "model": { + "type": "WordPiece", + "unk_token": "[UNK]", + "continuing_subword_prefix": "##", + "max_input_chars_per_word": 100, + "vocab": { + "[PAD]": 0, + "[UNK]": 1, + "[CLS]": 2, + "[SEP]": 3, + "[MASK]": 4, + "AAAAAA": 5, + "AAAAAT": 6, + "AAAAAC": 7, + "AAAAAG": 8, + "AAAATA": 9, + "AAAATT": 10, + "AAAATC": 11, + "AAAATG": 12, + "AAAACA": 13, + "AAAACT": 14, + "AAAACC": 15, + "AAAACG": 16, + "AAAAGA": 17, + "AAAAGT": 18, + "AAAAGC": 19, + "AAAAGG": 20, + "AAATAA": 21, + "AAATAT": 22, + "AAATAC": 23, + "AAATAG": 24, + "AAATTA": 25, + "AAATTT": 26, + "AAATTC": 27, + "AAATTG": 28, + "AAATCA": 29, + "AAATCT": 30, + "AAATCC": 31, + "AAATCG": 32, + "AAATGA": 33, + "AAATGT": 34, + "AAATGC": 35, + "AAATGG": 36, + "AAACAA": 37, + "AAACAT": 38, + "AAACAC": 39, + "AAACAG": 40, + "AAACTA": 41, + "AAACTT": 42, + "AAACTC": 43, + "AAACTG": 44, + "AAACCA": 45, + "AAACCT": 46, + "AAACCC": 47, + "AAACCG": 48, + "AAACGA": 49, + "AAACGT": 50, + "AAACGC": 51, + "AAACGG": 52, + "AAAGAA": 53, + "AAAGAT": 54, + "AAAGAC": 55, + "AAAGAG": 56, + "AAAGTA": 57, + "AAAGTT": 58, + "AAAGTC": 59, + "AAAGTG": 60, + "AAAGCA": 61, + "AAAGCT": 62, + "AAAGCC": 63, + "AAAGCG": 64, + "AAAGGA": 65, + "AAAGGT": 66, + "AAAGGC": 67, + "AAAGGG": 68, + "AATAAA": 69, + "AATAAT": 70, + "AATAAC": 71, + "AATAAG": 72, + "AATATA": 73, + "AATATT": 74, + "AATATC": 75, + "AATATG": 76, + "AATACA": 77, + "AATACT": 78, + "AATACC": 79, + "AATACG": 80, + "AATAGA": 81, + "AATAGT": 82, + "AATAGC": 83, + "AATAGG": 84, + "AATTAA": 85, + "AATTAT": 86, + "AATTAC": 87, + "AATTAG": 88, + "AATTTA": 89, + "AATTTT": 90, + "AATTTC": 91, + "AATTTG": 92, + "AATTCA": 93, + "AATTCT": 94, + "AATTCC": 95, + "AATTCG": 96, + "AATTGA": 97, + "AATTGT": 98, + "AATTGC": 99, + "AATTGG": 100, + "AATCAA": 101, + "AATCAT": 102, + "AATCAC": 103, + "AATCAG": 104, + "AATCTA": 105, + "AATCTT": 106, + "AATCTC": 107, + "AATCTG": 108, + "AATCCA": 109, + "AATCCT": 110, + "AATCCC": 111, + "AATCCG": 112, + "AATCGA": 113, + "AATCGT": 114, + "AATCGC": 115, + "AATCGG": 116, + "AATGAA": 117, + "AATGAT": 118, + "AATGAC": 119, + "AATGAG": 120, + "AATGTA": 121, + "AATGTT": 122, + "AATGTC": 123, + "AATGTG": 124, + "AATGCA": 125, + "AATGCT": 126, + "AATGCC": 127, + "AATGCG": 128, + "AATGGA": 129, + "AATGGT": 130, + "AATGGC": 131, + "AATGGG": 132, + "AACAAA": 133, + "AACAAT": 134, + "AACAAC": 135, + "AACAAG": 136, + "AACATA": 137, + "AACATT": 138, + "AACATC": 139, + "AACATG": 140, + "AACACA": 141, + "AACACT": 142, + "AACACC": 143, + "AACACG": 144, + "AACAGA": 145, + "AACAGT": 146, + "AACAGC": 147, + "AACAGG": 148, + "AACTAA": 149, + "AACTAT": 150, + "AACTAC": 151, + "AACTAG": 152, + "AACTTA": 153, + "AACTTT": 154, + "AACTTC": 155, + "AACTTG": 156, + "AACTCA": 157, + "AACTCT": 158, + "AACTCC": 159, + "AACTCG": 160, + "AACTGA": 161, + "AACTGT": 162, + "AACTGC": 163, + "AACTGG": 164, + "AACCAA": 165, + "AACCAT": 166, + "AACCAC": 167, + "AACCAG": 168, + "AACCTA": 169, + "AACCTT": 170, + "AACCTC": 171, + "AACCTG": 172, + "AACCCA": 173, + "AACCCT": 174, + "AACCCC": 175, + "AACCCG": 176, + "AACCGA": 177, + "AACCGT": 178, + "AACCGC": 179, + "AACCGG": 180, + "AACGAA": 181, + "AACGAT": 182, + "AACGAC": 183, + "AACGAG": 184, + "AACGTA": 185, + "AACGTT": 186, + "AACGTC": 187, + "AACGTG": 188, + "AACGCA": 189, + "AACGCT": 190, + "AACGCC": 191, + "AACGCG": 192, + "AACGGA": 193, + "AACGGT": 194, + "AACGGC": 195, + "AACGGG": 196, + "AAGAAA": 197, + "AAGAAT": 198, + "AAGAAC": 199, + "AAGAAG": 200, + "AAGATA": 201, + "AAGATT": 202, + "AAGATC": 203, + "AAGATG": 204, + "AAGACA": 205, + "AAGACT": 206, + "AAGACC": 207, + "AAGACG": 208, + "AAGAGA": 209, + "AAGAGT": 210, + "AAGAGC": 211, + "AAGAGG": 212, + "AAGTAA": 213, + "AAGTAT": 214, + "AAGTAC": 215, + "AAGTAG": 216, + "AAGTTA": 217, + "AAGTTT": 218, + "AAGTTC": 219, + "AAGTTG": 220, + "AAGTCA": 221, + "AAGTCT": 222, + "AAGTCC": 223, + "AAGTCG": 224, + "AAGTGA": 225, + "AAGTGT": 226, + "AAGTGC": 227, + "AAGTGG": 228, + "AAGCAA": 229, + "AAGCAT": 230, + "AAGCAC": 231, + "AAGCAG": 232, + "AAGCTA": 233, + "AAGCTT": 234, + "AAGCTC": 235, + "AAGCTG": 236, + "AAGCCA": 237, + "AAGCCT": 238, + "AAGCCC": 239, + "AAGCCG": 240, + "AAGCGA": 241, + "AAGCGT": 242, + "AAGCGC": 243, + "AAGCGG": 244, + "AAGGAA": 245, + "AAGGAT": 246, + "AAGGAC": 247, + "AAGGAG": 248, + "AAGGTA": 249, + "AAGGTT": 250, + "AAGGTC": 251, + "AAGGTG": 252, + "AAGGCA": 253, + "AAGGCT": 254, + "AAGGCC": 255, + "AAGGCG": 256, + "AAGGGA": 257, + "AAGGGT": 258, + "AAGGGC": 259, + "AAGGGG": 260, + "ATAAAA": 261, + "ATAAAT": 262, + "ATAAAC": 263, + "ATAAAG": 264, + "ATAATA": 265, + "ATAATT": 266, + "ATAATC": 267, + "ATAATG": 268, + "ATAACA": 269, + "ATAACT": 270, + "ATAACC": 271, + "ATAACG": 272, + "ATAAGA": 273, + "ATAAGT": 274, + "ATAAGC": 275, + "ATAAGG": 276, + "ATATAA": 277, + "ATATAT": 278, + "ATATAC": 279, + "ATATAG": 280, + "ATATTA": 281, + "ATATTT": 282, + "ATATTC": 283, + "ATATTG": 284, + "ATATCA": 285, + "ATATCT": 286, + "ATATCC": 287, + "ATATCG": 288, + "ATATGA": 289, + "ATATGT": 290, + "ATATGC": 291, + "ATATGG": 292, + "ATACAA": 293, + "ATACAT": 294, + "ATACAC": 295, + "ATACAG": 296, + "ATACTA": 297, + "ATACTT": 298, + "ATACTC": 299, + "ATACTG": 300, + "ATACCA": 301, + "ATACCT": 302, + "ATACCC": 303, + "ATACCG": 304, + "ATACGA": 305, + "ATACGT": 306, + "ATACGC": 307, + "ATACGG": 308, + "ATAGAA": 309, + "ATAGAT": 310, + "ATAGAC": 311, + "ATAGAG": 312, + "ATAGTA": 313, + "ATAGTT": 314, + "ATAGTC": 315, + "ATAGTG": 316, + "ATAGCA": 317, + "ATAGCT": 318, + "ATAGCC": 319, + "ATAGCG": 320, + "ATAGGA": 321, + "ATAGGT": 322, + "ATAGGC": 323, + "ATAGGG": 324, + "ATTAAA": 325, + "ATTAAT": 326, + "ATTAAC": 327, + "ATTAAG": 328, + "ATTATA": 329, + "ATTATT": 330, + "ATTATC": 331, + "ATTATG": 332, + "ATTACA": 333, + "ATTACT": 334, + "ATTACC": 335, + "ATTACG": 336, + "ATTAGA": 337, + "ATTAGT": 338, + "ATTAGC": 339, + "ATTAGG": 340, + "ATTTAA": 341, + "ATTTAT": 342, + "ATTTAC": 343, + "ATTTAG": 344, + "ATTTTA": 345, + "ATTTTT": 346, + "ATTTTC": 347, + "ATTTTG": 348, + "ATTTCA": 349, + "ATTTCT": 350, + "ATTTCC": 351, + "ATTTCG": 352, + "ATTTGA": 353, + "ATTTGT": 354, + "ATTTGC": 355, + "ATTTGG": 356, + "ATTCAA": 357, + "ATTCAT": 358, + "ATTCAC": 359, + "ATTCAG": 360, + "ATTCTA": 361, + "ATTCTT": 362, + "ATTCTC": 363, + "ATTCTG": 364, + "ATTCCA": 365, + "ATTCCT": 366, + "ATTCCC": 367, + "ATTCCG": 368, + "ATTCGA": 369, + "ATTCGT": 370, + "ATTCGC": 371, + "ATTCGG": 372, + "ATTGAA": 373, + "ATTGAT": 374, + "ATTGAC": 375, + "ATTGAG": 376, + "ATTGTA": 377, + "ATTGTT": 378, + "ATTGTC": 379, + "ATTGTG": 380, + "ATTGCA": 381, + "ATTGCT": 382, + "ATTGCC": 383, + "ATTGCG": 384, + "ATTGGA": 385, + "ATTGGT": 386, + "ATTGGC": 387, + "ATTGGG": 388, + "ATCAAA": 389, + "ATCAAT": 390, + "ATCAAC": 391, + "ATCAAG": 392, + "ATCATA": 393, + "ATCATT": 394, + "ATCATC": 395, + "ATCATG": 396, + "ATCACA": 397, + "ATCACT": 398, + "ATCACC": 399, + "ATCACG": 400, + "ATCAGA": 401, + "ATCAGT": 402, + "ATCAGC": 403, + "ATCAGG": 404, + "ATCTAA": 405, + "ATCTAT": 406, + "ATCTAC": 407, + "ATCTAG": 408, + "ATCTTA": 409, + "ATCTTT": 410, + "ATCTTC": 411, + "ATCTTG": 412, + "ATCTCA": 413, + "ATCTCT": 414, + "ATCTCC": 415, + "ATCTCG": 416, + "ATCTGA": 417, + "ATCTGT": 418, + "ATCTGC": 419, + "ATCTGG": 420, + "ATCCAA": 421, + "ATCCAT": 422, + "ATCCAC": 423, + "ATCCAG": 424, + "ATCCTA": 425, + "ATCCTT": 426, + "ATCCTC": 427, + "ATCCTG": 428, + "ATCCCA": 429, + "ATCCCT": 430, + "ATCCCC": 431, + "ATCCCG": 432, + "ATCCGA": 433, + "ATCCGT": 434, + "ATCCGC": 435, + "ATCCGG": 436, + "ATCGAA": 437, + "ATCGAT": 438, + "ATCGAC": 439, + "ATCGAG": 440, + "ATCGTA": 441, + "ATCGTT": 442, + "ATCGTC": 443, + "ATCGTG": 444, + "ATCGCA": 445, + "ATCGCT": 446, + "ATCGCC": 447, + "ATCGCG": 448, + "ATCGGA": 449, + "ATCGGT": 450, + "ATCGGC": 451, + "ATCGGG": 452, + "ATGAAA": 453, + "ATGAAT": 454, + "ATGAAC": 455, + "ATGAAG": 456, + "ATGATA": 457, + "ATGATT": 458, + "ATGATC": 459, + "ATGATG": 460, + "ATGACA": 461, + "ATGACT": 462, + "ATGACC": 463, + "ATGACG": 464, + "ATGAGA": 465, + "ATGAGT": 466, + "ATGAGC": 467, + "ATGAGG": 468, + "ATGTAA": 469, + "ATGTAT": 470, + "ATGTAC": 471, + "ATGTAG": 472, + "ATGTTA": 473, + "ATGTTT": 474, + "ATGTTC": 475, + "ATGTTG": 476, + "ATGTCA": 477, + "ATGTCT": 478, + "ATGTCC": 479, + "ATGTCG": 480, + "ATGTGA": 481, + "ATGTGT": 482, + "ATGTGC": 483, + "ATGTGG": 484, + "ATGCAA": 485, + "ATGCAT": 486, + "ATGCAC": 487, + "ATGCAG": 488, + "ATGCTA": 489, + "ATGCTT": 490, + "ATGCTC": 491, + "ATGCTG": 492, + "ATGCCA": 493, + "ATGCCT": 494, + "ATGCCC": 495, + "ATGCCG": 496, + "ATGCGA": 497, + "ATGCGT": 498, + "ATGCGC": 499, + "ATGCGG": 500, + "ATGGAA": 501, + "ATGGAT": 502, + "ATGGAC": 503, + "ATGGAG": 504, + "ATGGTA": 505, + "ATGGTT": 506, + "ATGGTC": 507, + "ATGGTG": 508, + "ATGGCA": 509, + "ATGGCT": 510, + "ATGGCC": 511, + "ATGGCG": 512, + "ATGGGA": 513, + "ATGGGT": 514, + "ATGGGC": 515, + "ATGGGG": 516, + "ACAAAA": 517, + "ACAAAT": 518, + "ACAAAC": 519, + "ACAAAG": 520, + "ACAATA": 521, + "ACAATT": 522, + "ACAATC": 523, + "ACAATG": 524, + "ACAACA": 525, + "ACAACT": 526, + "ACAACC": 527, + "ACAACG": 528, + "ACAAGA": 529, + "ACAAGT": 530, + "ACAAGC": 531, + "ACAAGG": 532, + "ACATAA": 533, + "ACATAT": 534, + "ACATAC": 535, + "ACATAG": 536, + "ACATTA": 537, + "ACATTT": 538, + "ACATTC": 539, + "ACATTG": 540, + "ACATCA": 541, + "ACATCT": 542, + "ACATCC": 543, + "ACATCG": 544, + "ACATGA": 545, + "ACATGT": 546, + "ACATGC": 547, + "ACATGG": 548, + "ACACAA": 549, + "ACACAT": 550, + "ACACAC": 551, + "ACACAG": 552, + "ACACTA": 553, + "ACACTT": 554, + "ACACTC": 555, + "ACACTG": 556, + "ACACCA": 557, + "ACACCT": 558, + "ACACCC": 559, + "ACACCG": 560, + "ACACGA": 561, + "ACACGT": 562, + "ACACGC": 563, + "ACACGG": 564, + "ACAGAA": 565, + "ACAGAT": 566, + "ACAGAC": 567, + "ACAGAG": 568, + "ACAGTA": 569, + "ACAGTT": 570, + "ACAGTC": 571, + "ACAGTG": 572, + "ACAGCA": 573, + "ACAGCT": 574, + "ACAGCC": 575, + "ACAGCG": 576, + "ACAGGA": 577, + "ACAGGT": 578, + "ACAGGC": 579, + "ACAGGG": 580, + "ACTAAA": 581, + "ACTAAT": 582, + "ACTAAC": 583, + "ACTAAG": 584, + "ACTATA": 585, + "ACTATT": 586, + "ACTATC": 587, + "ACTATG": 588, + "ACTACA": 589, + "ACTACT": 590, + "ACTACC": 591, + "ACTACG": 592, + "ACTAGA": 593, + "ACTAGT": 594, + "ACTAGC": 595, + "ACTAGG": 596, + "ACTTAA": 597, + "ACTTAT": 598, + "ACTTAC": 599, + "ACTTAG": 600, + "ACTTTA": 601, + "ACTTTT": 602, + "ACTTTC": 603, + "ACTTTG": 604, + "ACTTCA": 605, + "ACTTCT": 606, + "ACTTCC": 607, + "ACTTCG": 608, + "ACTTGA": 609, + "ACTTGT": 610, + "ACTTGC": 611, + "ACTTGG": 612, + "ACTCAA": 613, + "ACTCAT": 614, + "ACTCAC": 615, + "ACTCAG": 616, + "ACTCTA": 617, + "ACTCTT": 618, + "ACTCTC": 619, + "ACTCTG": 620, + "ACTCCA": 621, + "ACTCCT": 622, + "ACTCCC": 623, + "ACTCCG": 624, + "ACTCGA": 625, + "ACTCGT": 626, + "ACTCGC": 627, + "ACTCGG": 628, + "ACTGAA": 629, + "ACTGAT": 630, + "ACTGAC": 631, + "ACTGAG": 632, + "ACTGTA": 633, + "ACTGTT": 634, + "ACTGTC": 635, + "ACTGTG": 636, + "ACTGCA": 637, + "ACTGCT": 638, + "ACTGCC": 639, + "ACTGCG": 640, + "ACTGGA": 641, + "ACTGGT": 642, + "ACTGGC": 643, + "ACTGGG": 644, + "ACCAAA": 645, + "ACCAAT": 646, + "ACCAAC": 647, + "ACCAAG": 648, + "ACCATA": 649, + "ACCATT": 650, + "ACCATC": 651, + "ACCATG": 652, + "ACCACA": 653, + "ACCACT": 654, + "ACCACC": 655, + "ACCACG": 656, + "ACCAGA": 657, + "ACCAGT": 658, + "ACCAGC": 659, + "ACCAGG": 660, + "ACCTAA": 661, + "ACCTAT": 662, + "ACCTAC": 663, + "ACCTAG": 664, + "ACCTTA": 665, + "ACCTTT": 666, + "ACCTTC": 667, + "ACCTTG": 668, + "ACCTCA": 669, + "ACCTCT": 670, + "ACCTCC": 671, + "ACCTCG": 672, + "ACCTGA": 673, + "ACCTGT": 674, + "ACCTGC": 675, + "ACCTGG": 676, + "ACCCAA": 677, + "ACCCAT": 678, + "ACCCAC": 679, + "ACCCAG": 680, + "ACCCTA": 681, + "ACCCTT": 682, + "ACCCTC": 683, + "ACCCTG": 684, + "ACCCCA": 685, + "ACCCCT": 686, + "ACCCCC": 687, + "ACCCCG": 688, + "ACCCGA": 689, + "ACCCGT": 690, + "ACCCGC": 691, + "ACCCGG": 692, + "ACCGAA": 693, + "ACCGAT": 694, + "ACCGAC": 695, + "ACCGAG": 696, + "ACCGTA": 697, + "ACCGTT": 698, + "ACCGTC": 699, + "ACCGTG": 700, + "ACCGCA": 701, + "ACCGCT": 702, + "ACCGCC": 703, + "ACCGCG": 704, + "ACCGGA": 705, + "ACCGGT": 706, + "ACCGGC": 707, + "ACCGGG": 708, + "ACGAAA": 709, + "ACGAAT": 710, + "ACGAAC": 711, + "ACGAAG": 712, + "ACGATA": 713, + "ACGATT": 714, + "ACGATC": 715, + "ACGATG": 716, + "ACGACA": 717, + "ACGACT": 718, + "ACGACC": 719, + "ACGACG": 720, + "ACGAGA": 721, + "ACGAGT": 722, + "ACGAGC": 723, + "ACGAGG": 724, + "ACGTAA": 725, + "ACGTAT": 726, + "ACGTAC": 727, + "ACGTAG": 728, + "ACGTTA": 729, + "ACGTTT": 730, + "ACGTTC": 731, + "ACGTTG": 732, + "ACGTCA": 733, + "ACGTCT": 734, + "ACGTCC": 735, + "ACGTCG": 736, + "ACGTGA": 737, + "ACGTGT": 738, + "ACGTGC": 739, + "ACGTGG": 740, + "ACGCAA": 741, + "ACGCAT": 742, + "ACGCAC": 743, + "ACGCAG": 744, + "ACGCTA": 745, + "ACGCTT": 746, + "ACGCTC": 747, + "ACGCTG": 748, + "ACGCCA": 749, + "ACGCCT": 750, + "ACGCCC": 751, + "ACGCCG": 752, + "ACGCGA": 753, + "ACGCGT": 754, + "ACGCGC": 755, + "ACGCGG": 756, + "ACGGAA": 757, + "ACGGAT": 758, + "ACGGAC": 759, + "ACGGAG": 760, + "ACGGTA": 761, + "ACGGTT": 762, + "ACGGTC": 763, + "ACGGTG": 764, + "ACGGCA": 765, + "ACGGCT": 766, + "ACGGCC": 767, + "ACGGCG": 768, + "ACGGGA": 769, + "ACGGGT": 770, + "ACGGGC": 771, + "ACGGGG": 772, + "AGAAAA": 773, + "AGAAAT": 774, + "AGAAAC": 775, + "AGAAAG": 776, + "AGAATA": 777, + "AGAATT": 778, + "AGAATC": 779, + "AGAATG": 780, + "AGAACA": 781, + "AGAACT": 782, + "AGAACC": 783, + "AGAACG": 784, + "AGAAGA": 785, + "AGAAGT": 786, + "AGAAGC": 787, + "AGAAGG": 788, + "AGATAA": 789, + "AGATAT": 790, + "AGATAC": 791, + "AGATAG": 792, + "AGATTA": 793, + "AGATTT": 794, + "AGATTC": 795, + "AGATTG": 796, + "AGATCA": 797, + "AGATCT": 798, + "AGATCC": 799, + "AGATCG": 800, + "AGATGA": 801, + "AGATGT": 802, + "AGATGC": 803, + "AGATGG": 804, + "AGACAA": 805, + "AGACAT": 806, + "AGACAC": 807, + "AGACAG": 808, + "AGACTA": 809, + "AGACTT": 810, + "AGACTC": 811, + "AGACTG": 812, + "AGACCA": 813, + "AGACCT": 814, + "AGACCC": 815, + "AGACCG": 816, + "AGACGA": 817, + "AGACGT": 818, + "AGACGC": 819, + "AGACGG": 820, + "AGAGAA": 821, + "AGAGAT": 822, + "AGAGAC": 823, + "AGAGAG": 824, + "AGAGTA": 825, + "AGAGTT": 826, + "AGAGTC": 827, + "AGAGTG": 828, + "AGAGCA": 829, + "AGAGCT": 830, + "AGAGCC": 831, + "AGAGCG": 832, + "AGAGGA": 833, + "AGAGGT": 834, + "AGAGGC": 835, + "AGAGGG": 836, + "AGTAAA": 837, + "AGTAAT": 838, + "AGTAAC": 839, + "AGTAAG": 840, + "AGTATA": 841, + "AGTATT": 842, + "AGTATC": 843, + "AGTATG": 844, + "AGTACA": 845, + "AGTACT": 846, + "AGTACC": 847, + "AGTACG": 848, + "AGTAGA": 849, + "AGTAGT": 850, + "AGTAGC": 851, + "AGTAGG": 852, + "AGTTAA": 853, + "AGTTAT": 854, + "AGTTAC": 855, + "AGTTAG": 856, + "AGTTTA": 857, + "AGTTTT": 858, + "AGTTTC": 859, + "AGTTTG": 860, + "AGTTCA": 861, + "AGTTCT": 862, + "AGTTCC": 863, + "AGTTCG": 864, + "AGTTGA": 865, + "AGTTGT": 866, + "AGTTGC": 867, + "AGTTGG": 868, + "AGTCAA": 869, + "AGTCAT": 870, + "AGTCAC": 871, + "AGTCAG": 872, + "AGTCTA": 873, + "AGTCTT": 874, + "AGTCTC": 875, + "AGTCTG": 876, + "AGTCCA": 877, + "AGTCCT": 878, + "AGTCCC": 879, + "AGTCCG": 880, + "AGTCGA": 881, + "AGTCGT": 882, + "AGTCGC": 883, + "AGTCGG": 884, + "AGTGAA": 885, + "AGTGAT": 886, + "AGTGAC": 887, + "AGTGAG": 888, + "AGTGTA": 889, + "AGTGTT": 890, + "AGTGTC": 891, + "AGTGTG": 892, + "AGTGCA": 893, + "AGTGCT": 894, + "AGTGCC": 895, + "AGTGCG": 896, + "AGTGGA": 897, + "AGTGGT": 898, + "AGTGGC": 899, + "AGTGGG": 900, + "AGCAAA": 901, + "AGCAAT": 902, + "AGCAAC": 903, + "AGCAAG": 904, + "AGCATA": 905, + "AGCATT": 906, + "AGCATC": 907, + "AGCATG": 908, + "AGCACA": 909, + "AGCACT": 910, + "AGCACC": 911, + "AGCACG": 912, + "AGCAGA": 913, + "AGCAGT": 914, + "AGCAGC": 915, + "AGCAGG": 916, + "AGCTAA": 917, + "AGCTAT": 918, + "AGCTAC": 919, + "AGCTAG": 920, + "AGCTTA": 921, + "AGCTTT": 922, + "AGCTTC": 923, + "AGCTTG": 924, + "AGCTCA": 925, + "AGCTCT": 926, + "AGCTCC": 927, + "AGCTCG": 928, + "AGCTGA": 929, + "AGCTGT": 930, + "AGCTGC": 931, + "AGCTGG": 932, + "AGCCAA": 933, + "AGCCAT": 934, + "AGCCAC": 935, + "AGCCAG": 936, + "AGCCTA": 937, + "AGCCTT": 938, + "AGCCTC": 939, + "AGCCTG": 940, + "AGCCCA": 941, + "AGCCCT": 942, + "AGCCCC": 943, + "AGCCCG": 944, + "AGCCGA": 945, + "AGCCGT": 946, + "AGCCGC": 947, + "AGCCGG": 948, + "AGCGAA": 949, + "AGCGAT": 950, + "AGCGAC": 951, + "AGCGAG": 952, + "AGCGTA": 953, + "AGCGTT": 954, + "AGCGTC": 955, + "AGCGTG": 956, + "AGCGCA": 957, + "AGCGCT": 958, + "AGCGCC": 959, + "AGCGCG": 960, + "AGCGGA": 961, + "AGCGGT": 962, + "AGCGGC": 963, + "AGCGGG": 964, + "AGGAAA": 965, + "AGGAAT": 966, + "AGGAAC": 967, + "AGGAAG": 968, + "AGGATA": 969, + "AGGATT": 970, + "AGGATC": 971, + "AGGATG": 972, + "AGGACA": 973, + "AGGACT": 974, + "AGGACC": 975, + "AGGACG": 976, + "AGGAGA": 977, + "AGGAGT": 978, + "AGGAGC": 979, + "AGGAGG": 980, + "AGGTAA": 981, + "AGGTAT": 982, + "AGGTAC": 983, + "AGGTAG": 984, + "AGGTTA": 985, + "AGGTTT": 986, + "AGGTTC": 987, + "AGGTTG": 988, + "AGGTCA": 989, + "AGGTCT": 990, + "AGGTCC": 991, + "AGGTCG": 992, + "AGGTGA": 993, + "AGGTGT": 994, + "AGGTGC": 995, + "AGGTGG": 996, + "AGGCAA": 997, + "AGGCAT": 998, + "AGGCAC": 999, + "AGGCAG": 1000, + "AGGCTA": 1001, + "AGGCTT": 1002, + "AGGCTC": 1003, + "AGGCTG": 1004, + "AGGCCA": 1005, + "AGGCCT": 1006, + "AGGCCC": 1007, + "AGGCCG": 1008, + "AGGCGA": 1009, + "AGGCGT": 1010, + "AGGCGC": 1011, + "AGGCGG": 1012, + "AGGGAA": 1013, + "AGGGAT": 1014, + "AGGGAC": 1015, + "AGGGAG": 1016, + "AGGGTA": 1017, + "AGGGTT": 1018, + "AGGGTC": 1019, + "AGGGTG": 1020, + "AGGGCA": 1021, + "AGGGCT": 1022, + "AGGGCC": 1023, + "AGGGCG": 1024, + "AGGGGA": 1025, + "AGGGGT": 1026, + "AGGGGC": 1027, + "AGGGGG": 1028, + "TAAAAA": 1029, + "TAAAAT": 1030, + "TAAAAC": 1031, + "TAAAAG": 1032, + "TAAATA": 1033, + "TAAATT": 1034, + "TAAATC": 1035, + "TAAATG": 1036, + "TAAACA": 1037, + "TAAACT": 1038, + "TAAACC": 1039, + "TAAACG": 1040, + "TAAAGA": 1041, + "TAAAGT": 1042, + "TAAAGC": 1043, + "TAAAGG": 1044, + "TAATAA": 1045, + "TAATAT": 1046, + "TAATAC": 1047, + "TAATAG": 1048, + "TAATTA": 1049, + "TAATTT": 1050, + "TAATTC": 1051, + "TAATTG": 1052, + "TAATCA": 1053, + "TAATCT": 1054, + "TAATCC": 1055, + "TAATCG": 1056, + "TAATGA": 1057, + "TAATGT": 1058, + "TAATGC": 1059, + "TAATGG": 1060, + "TAACAA": 1061, + "TAACAT": 1062, + "TAACAC": 1063, + "TAACAG": 1064, + "TAACTA": 1065, + "TAACTT": 1066, + "TAACTC": 1067, + "TAACTG": 1068, + "TAACCA": 1069, + "TAACCT": 1070, + "TAACCC": 1071, + "TAACCG": 1072, + "TAACGA": 1073, + "TAACGT": 1074, + "TAACGC": 1075, + "TAACGG": 1076, + "TAAGAA": 1077, + "TAAGAT": 1078, + "TAAGAC": 1079, + "TAAGAG": 1080, + "TAAGTA": 1081, + "TAAGTT": 1082, + "TAAGTC": 1083, + "TAAGTG": 1084, + "TAAGCA": 1085, + "TAAGCT": 1086, + "TAAGCC": 1087, + "TAAGCG": 1088, + "TAAGGA": 1089, + "TAAGGT": 1090, + "TAAGGC": 1091, + "TAAGGG": 1092, + "TATAAA": 1093, + "TATAAT": 1094, + "TATAAC": 1095, + "TATAAG": 1096, + "TATATA": 1097, + "TATATT": 1098, + "TATATC": 1099, + "TATATG": 1100, + "TATACA": 1101, + "TATACT": 1102, + "TATACC": 1103, + "TATACG": 1104, + "TATAGA": 1105, + "TATAGT": 1106, + "TATAGC": 1107, + "TATAGG": 1108, + "TATTAA": 1109, + "TATTAT": 1110, + "TATTAC": 1111, + "TATTAG": 1112, + "TATTTA": 1113, + "TATTTT": 1114, + "TATTTC": 1115, + "TATTTG": 1116, + "TATTCA": 1117, + "TATTCT": 1118, + "TATTCC": 1119, + "TATTCG": 1120, + "TATTGA": 1121, + "TATTGT": 1122, + "TATTGC": 1123, + "TATTGG": 1124, + "TATCAA": 1125, + "TATCAT": 1126, + "TATCAC": 1127, + "TATCAG": 1128, + "TATCTA": 1129, + "TATCTT": 1130, + "TATCTC": 1131, + "TATCTG": 1132, + "TATCCA": 1133, + "TATCCT": 1134, + "TATCCC": 1135, + "TATCCG": 1136, + "TATCGA": 1137, + "TATCGT": 1138, + "TATCGC": 1139, + "TATCGG": 1140, + "TATGAA": 1141, + "TATGAT": 1142, + "TATGAC": 1143, + "TATGAG": 1144, + "TATGTA": 1145, + "TATGTT": 1146, + "TATGTC": 1147, + "TATGTG": 1148, + "TATGCA": 1149, + "TATGCT": 1150, + "TATGCC": 1151, + "TATGCG": 1152, + "TATGGA": 1153, + "TATGGT": 1154, + "TATGGC": 1155, + "TATGGG": 1156, + "TACAAA": 1157, + "TACAAT": 1158, + "TACAAC": 1159, + "TACAAG": 1160, + "TACATA": 1161, + "TACATT": 1162, + "TACATC": 1163, + "TACATG": 1164, + "TACACA": 1165, + "TACACT": 1166, + "TACACC": 1167, + "TACACG": 1168, + "TACAGA": 1169, + "TACAGT": 1170, + "TACAGC": 1171, + "TACAGG": 1172, + "TACTAA": 1173, + "TACTAT": 1174, + "TACTAC": 1175, + "TACTAG": 1176, + "TACTTA": 1177, + "TACTTT": 1178, + "TACTTC": 1179, + "TACTTG": 1180, + "TACTCA": 1181, + "TACTCT": 1182, + "TACTCC": 1183, + "TACTCG": 1184, + "TACTGA": 1185, + "TACTGT": 1186, + "TACTGC": 1187, + "TACTGG": 1188, + "TACCAA": 1189, + "TACCAT": 1190, + "TACCAC": 1191, + "TACCAG": 1192, + "TACCTA": 1193, + "TACCTT": 1194, + "TACCTC": 1195, + "TACCTG": 1196, + "TACCCA": 1197, + "TACCCT": 1198, + "TACCCC": 1199, + "TACCCG": 1200, + "TACCGA": 1201, + "TACCGT": 1202, + "TACCGC": 1203, + "TACCGG": 1204, + "TACGAA": 1205, + "TACGAT": 1206, + "TACGAC": 1207, + "TACGAG": 1208, + "TACGTA": 1209, + "TACGTT": 1210, + "TACGTC": 1211, + "TACGTG": 1212, + "TACGCA": 1213, + "TACGCT": 1214, + "TACGCC": 1215, + "TACGCG": 1216, + "TACGGA": 1217, + "TACGGT": 1218, + "TACGGC": 1219, + "TACGGG": 1220, + "TAGAAA": 1221, + "TAGAAT": 1222, + "TAGAAC": 1223, + "TAGAAG": 1224, + "TAGATA": 1225, + "TAGATT": 1226, + "TAGATC": 1227, + "TAGATG": 1228, + "TAGACA": 1229, + "TAGACT": 1230, + "TAGACC": 1231, + "TAGACG": 1232, + "TAGAGA": 1233, + "TAGAGT": 1234, + "TAGAGC": 1235, + "TAGAGG": 1236, + "TAGTAA": 1237, + "TAGTAT": 1238, + "TAGTAC": 1239, + "TAGTAG": 1240, + "TAGTTA": 1241, + "TAGTTT": 1242, + "TAGTTC": 1243, + "TAGTTG": 1244, + "TAGTCA": 1245, + "TAGTCT": 1246, + "TAGTCC": 1247, + "TAGTCG": 1248, + "TAGTGA": 1249, + "TAGTGT": 1250, + "TAGTGC": 1251, + "TAGTGG": 1252, + "TAGCAA": 1253, + "TAGCAT": 1254, + "TAGCAC": 1255, + "TAGCAG": 1256, + "TAGCTA": 1257, + "TAGCTT": 1258, + "TAGCTC": 1259, + "TAGCTG": 1260, + "TAGCCA": 1261, + "TAGCCT": 1262, + "TAGCCC": 1263, + "TAGCCG": 1264, + "TAGCGA": 1265, + "TAGCGT": 1266, + "TAGCGC": 1267, + "TAGCGG": 1268, + "TAGGAA": 1269, + "TAGGAT": 1270, + "TAGGAC": 1271, + "TAGGAG": 1272, + "TAGGTA": 1273, + "TAGGTT": 1274, + "TAGGTC": 1275, + "TAGGTG": 1276, + "TAGGCA": 1277, + "TAGGCT": 1278, + "TAGGCC": 1279, + "TAGGCG": 1280, + "TAGGGA": 1281, + "TAGGGT": 1282, + "TAGGGC": 1283, + "TAGGGG": 1284, + "TTAAAA": 1285, + "TTAAAT": 1286, + "TTAAAC": 1287, + "TTAAAG": 1288, + "TTAATA": 1289, + "TTAATT": 1290, + "TTAATC": 1291, + "TTAATG": 1292, + "TTAACA": 1293, + "TTAACT": 1294, + "TTAACC": 1295, + "TTAACG": 1296, + "TTAAGA": 1297, + "TTAAGT": 1298, + "TTAAGC": 1299, + "TTAAGG": 1300, + "TTATAA": 1301, + "TTATAT": 1302, + "TTATAC": 1303, + "TTATAG": 1304, + "TTATTA": 1305, + "TTATTT": 1306, + "TTATTC": 1307, + "TTATTG": 1308, + "TTATCA": 1309, + "TTATCT": 1310, + "TTATCC": 1311, + "TTATCG": 1312, + "TTATGA": 1313, + "TTATGT": 1314, + "TTATGC": 1315, + "TTATGG": 1316, + "TTACAA": 1317, + "TTACAT": 1318, + "TTACAC": 1319, + "TTACAG": 1320, + "TTACTA": 1321, + "TTACTT": 1322, + "TTACTC": 1323, + "TTACTG": 1324, + "TTACCA": 1325, + "TTACCT": 1326, + "TTACCC": 1327, + "TTACCG": 1328, + "TTACGA": 1329, + "TTACGT": 1330, + "TTACGC": 1331, + "TTACGG": 1332, + "TTAGAA": 1333, + "TTAGAT": 1334, + "TTAGAC": 1335, + "TTAGAG": 1336, + "TTAGTA": 1337, + "TTAGTT": 1338, + "TTAGTC": 1339, + "TTAGTG": 1340, + "TTAGCA": 1341, + "TTAGCT": 1342, + "TTAGCC": 1343, + "TTAGCG": 1344, + "TTAGGA": 1345, + "TTAGGT": 1346, + "TTAGGC": 1347, + "TTAGGG": 1348, + "TTTAAA": 1349, + "TTTAAT": 1350, + "TTTAAC": 1351, + "TTTAAG": 1352, + "TTTATA": 1353, + "TTTATT": 1354, + "TTTATC": 1355, + "TTTATG": 1356, + "TTTACA": 1357, + "TTTACT": 1358, + "TTTACC": 1359, + "TTTACG": 1360, + "TTTAGA": 1361, + "TTTAGT": 1362, + "TTTAGC": 1363, + "TTTAGG": 1364, + "TTTTAA": 1365, + "TTTTAT": 1366, + "TTTTAC": 1367, + "TTTTAG": 1368, + "TTTTTA": 1369, + "TTTTTT": 1370, + "TTTTTC": 1371, + "TTTTTG": 1372, + "TTTTCA": 1373, + "TTTTCT": 1374, + "TTTTCC": 1375, + "TTTTCG": 1376, + "TTTTGA": 1377, + "TTTTGT": 1378, + "TTTTGC": 1379, + "TTTTGG": 1380, + "TTTCAA": 1381, + "TTTCAT": 1382, + "TTTCAC": 1383, + "TTTCAG": 1384, + "TTTCTA": 1385, + "TTTCTT": 1386, + "TTTCTC": 1387, + "TTTCTG": 1388, + "TTTCCA": 1389, + "TTTCCT": 1390, + "TTTCCC": 1391, + "TTTCCG": 1392, + "TTTCGA": 1393, + "TTTCGT": 1394, + "TTTCGC": 1395, + "TTTCGG": 1396, + "TTTGAA": 1397, + "TTTGAT": 1398, + "TTTGAC": 1399, + "TTTGAG": 1400, + "TTTGTA": 1401, + "TTTGTT": 1402, + "TTTGTC": 1403, + "TTTGTG": 1404, + "TTTGCA": 1405, + "TTTGCT": 1406, + "TTTGCC": 1407, + "TTTGCG": 1408, + "TTTGGA": 1409, + "TTTGGT": 1410, + "TTTGGC": 1411, + "TTTGGG": 1412, + "TTCAAA": 1413, + "TTCAAT": 1414, + "TTCAAC": 1415, + "TTCAAG": 1416, + "TTCATA": 1417, + "TTCATT": 1418, + "TTCATC": 1419, + "TTCATG": 1420, + "TTCACA": 1421, + "TTCACT": 1422, + "TTCACC": 1423, + "TTCACG": 1424, + "TTCAGA": 1425, + "TTCAGT": 1426, + "TTCAGC": 1427, + "TTCAGG": 1428, + "TTCTAA": 1429, + "TTCTAT": 1430, + "TTCTAC": 1431, + "TTCTAG": 1432, + "TTCTTA": 1433, + "TTCTTT": 1434, + "TTCTTC": 1435, + "TTCTTG": 1436, + "TTCTCA": 1437, + "TTCTCT": 1438, + "TTCTCC": 1439, + "TTCTCG": 1440, + "TTCTGA": 1441, + "TTCTGT": 1442, + "TTCTGC": 1443, + "TTCTGG": 1444, + "TTCCAA": 1445, + "TTCCAT": 1446, + "TTCCAC": 1447, + "TTCCAG": 1448, + "TTCCTA": 1449, + "TTCCTT": 1450, + "TTCCTC": 1451, + "TTCCTG": 1452, + "TTCCCA": 1453, + "TTCCCT": 1454, + "TTCCCC": 1455, + "TTCCCG": 1456, + "TTCCGA": 1457, + "TTCCGT": 1458, + "TTCCGC": 1459, + "TTCCGG": 1460, + "TTCGAA": 1461, + "TTCGAT": 1462, + "TTCGAC": 1463, + "TTCGAG": 1464, + "TTCGTA": 1465, + "TTCGTT": 1466, + "TTCGTC": 1467, + "TTCGTG": 1468, + "TTCGCA": 1469, + "TTCGCT": 1470, + "TTCGCC": 1471, + "TTCGCG": 1472, + "TTCGGA": 1473, + "TTCGGT": 1474, + "TTCGGC": 1475, + "TTCGGG": 1476, + "TTGAAA": 1477, + "TTGAAT": 1478, + "TTGAAC": 1479, + "TTGAAG": 1480, + "TTGATA": 1481, + "TTGATT": 1482, + "TTGATC": 1483, + "TTGATG": 1484, + "TTGACA": 1485, + "TTGACT": 1486, + "TTGACC": 1487, + "TTGACG": 1488, + "TTGAGA": 1489, + "TTGAGT": 1490, + "TTGAGC": 1491, + "TTGAGG": 1492, + "TTGTAA": 1493, + "TTGTAT": 1494, + "TTGTAC": 1495, + "TTGTAG": 1496, + "TTGTTA": 1497, + "TTGTTT": 1498, + "TTGTTC": 1499, + "TTGTTG": 1500, + "TTGTCA": 1501, + "TTGTCT": 1502, + "TTGTCC": 1503, + "TTGTCG": 1504, + "TTGTGA": 1505, + "TTGTGT": 1506, + "TTGTGC": 1507, + "TTGTGG": 1508, + "TTGCAA": 1509, + "TTGCAT": 1510, + "TTGCAC": 1511, + "TTGCAG": 1512, + "TTGCTA": 1513, + "TTGCTT": 1514, + "TTGCTC": 1515, + "TTGCTG": 1516, + "TTGCCA": 1517, + "TTGCCT": 1518, + "TTGCCC": 1519, + "TTGCCG": 1520, + "TTGCGA": 1521, + "TTGCGT": 1522, + "TTGCGC": 1523, + "TTGCGG": 1524, + "TTGGAA": 1525, + "TTGGAT": 1526, + "TTGGAC": 1527, + "TTGGAG": 1528, + "TTGGTA": 1529, + "TTGGTT": 1530, + "TTGGTC": 1531, + "TTGGTG": 1532, + "TTGGCA": 1533, + "TTGGCT": 1534, + "TTGGCC": 1535, + "TTGGCG": 1536, + "TTGGGA": 1537, + "TTGGGT": 1538, + "TTGGGC": 1539, + "TTGGGG": 1540, + "TCAAAA": 1541, + "TCAAAT": 1542, + "TCAAAC": 1543, + "TCAAAG": 1544, + "TCAATA": 1545, + "TCAATT": 1546, + "TCAATC": 1547, + "TCAATG": 1548, + "TCAACA": 1549, + "TCAACT": 1550, + "TCAACC": 1551, + "TCAACG": 1552, + "TCAAGA": 1553, + "TCAAGT": 1554, + "TCAAGC": 1555, + "TCAAGG": 1556, + "TCATAA": 1557, + "TCATAT": 1558, + "TCATAC": 1559, + "TCATAG": 1560, + "TCATTA": 1561, + "TCATTT": 1562, + "TCATTC": 1563, + "TCATTG": 1564, + "TCATCA": 1565, + "TCATCT": 1566, + "TCATCC": 1567, + "TCATCG": 1568, + "TCATGA": 1569, + "TCATGT": 1570, + "TCATGC": 1571, + "TCATGG": 1572, + "TCACAA": 1573, + "TCACAT": 1574, + "TCACAC": 1575, + "TCACAG": 1576, + "TCACTA": 1577, + "TCACTT": 1578, + "TCACTC": 1579, + "TCACTG": 1580, + "TCACCA": 1581, + "TCACCT": 1582, + "TCACCC": 1583, + "TCACCG": 1584, + "TCACGA": 1585, + "TCACGT": 1586, + "TCACGC": 1587, + "TCACGG": 1588, + "TCAGAA": 1589, + "TCAGAT": 1590, + "TCAGAC": 1591, + "TCAGAG": 1592, + "TCAGTA": 1593, + "TCAGTT": 1594, + "TCAGTC": 1595, + "TCAGTG": 1596, + "TCAGCA": 1597, + "TCAGCT": 1598, + "TCAGCC": 1599, + "TCAGCG": 1600, + "TCAGGA": 1601, + "TCAGGT": 1602, + "TCAGGC": 1603, + "TCAGGG": 1604, + "TCTAAA": 1605, + "TCTAAT": 1606, + "TCTAAC": 1607, + "TCTAAG": 1608, + "TCTATA": 1609, + "TCTATT": 1610, + "TCTATC": 1611, + "TCTATG": 1612, + "TCTACA": 1613, + "TCTACT": 1614, + "TCTACC": 1615, + "TCTACG": 1616, + "TCTAGA": 1617, + "TCTAGT": 1618, + "TCTAGC": 1619, + "TCTAGG": 1620, + "TCTTAA": 1621, + "TCTTAT": 1622, + "TCTTAC": 1623, + "TCTTAG": 1624, + "TCTTTA": 1625, + "TCTTTT": 1626, + "TCTTTC": 1627, + "TCTTTG": 1628, + "TCTTCA": 1629, + "TCTTCT": 1630, + "TCTTCC": 1631, + "TCTTCG": 1632, + "TCTTGA": 1633, + "TCTTGT": 1634, + "TCTTGC": 1635, + "TCTTGG": 1636, + "TCTCAA": 1637, + "TCTCAT": 1638, + "TCTCAC": 1639, + "TCTCAG": 1640, + "TCTCTA": 1641, + "TCTCTT": 1642, + "TCTCTC": 1643, + "TCTCTG": 1644, + "TCTCCA": 1645, + "TCTCCT": 1646, + "TCTCCC": 1647, + "TCTCCG": 1648, + "TCTCGA": 1649, + "TCTCGT": 1650, + "TCTCGC": 1651, + "TCTCGG": 1652, + "TCTGAA": 1653, + "TCTGAT": 1654, + "TCTGAC": 1655, + "TCTGAG": 1656, + "TCTGTA": 1657, + "TCTGTT": 1658, + "TCTGTC": 1659, + "TCTGTG": 1660, + "TCTGCA": 1661, + "TCTGCT": 1662, + "TCTGCC": 1663, + "TCTGCG": 1664, + "TCTGGA": 1665, + "TCTGGT": 1666, + "TCTGGC": 1667, + "TCTGGG": 1668, + "TCCAAA": 1669, + "TCCAAT": 1670, + "TCCAAC": 1671, + "TCCAAG": 1672, + "TCCATA": 1673, + "TCCATT": 1674, + "TCCATC": 1675, + "TCCATG": 1676, + "TCCACA": 1677, + "TCCACT": 1678, + "TCCACC": 1679, + "TCCACG": 1680, + "TCCAGA": 1681, + "TCCAGT": 1682, + "TCCAGC": 1683, + "TCCAGG": 1684, + "TCCTAA": 1685, + "TCCTAT": 1686, + "TCCTAC": 1687, + "TCCTAG": 1688, + "TCCTTA": 1689, + "TCCTTT": 1690, + "TCCTTC": 1691, + "TCCTTG": 1692, + "TCCTCA": 1693, + "TCCTCT": 1694, + "TCCTCC": 1695, + "TCCTCG": 1696, + "TCCTGA": 1697, + "TCCTGT": 1698, + "TCCTGC": 1699, + "TCCTGG": 1700, + "TCCCAA": 1701, + "TCCCAT": 1702, + "TCCCAC": 1703, + "TCCCAG": 1704, + "TCCCTA": 1705, + "TCCCTT": 1706, + "TCCCTC": 1707, + "TCCCTG": 1708, + "TCCCCA": 1709, + "TCCCCT": 1710, + "TCCCCC": 1711, + "TCCCCG": 1712, + "TCCCGA": 1713, + "TCCCGT": 1714, + "TCCCGC": 1715, + "TCCCGG": 1716, + "TCCGAA": 1717, + "TCCGAT": 1718, + "TCCGAC": 1719, + "TCCGAG": 1720, + "TCCGTA": 1721, + "TCCGTT": 1722, + "TCCGTC": 1723, + "TCCGTG": 1724, + "TCCGCA": 1725, + "TCCGCT": 1726, + "TCCGCC": 1727, + "TCCGCG": 1728, + "TCCGGA": 1729, + "TCCGGT": 1730, + "TCCGGC": 1731, + "TCCGGG": 1732, + "TCGAAA": 1733, + "TCGAAT": 1734, + "TCGAAC": 1735, + "TCGAAG": 1736, + "TCGATA": 1737, + "TCGATT": 1738, + "TCGATC": 1739, + "TCGATG": 1740, + "TCGACA": 1741, + "TCGACT": 1742, + "TCGACC": 1743, + "TCGACG": 1744, + "TCGAGA": 1745, + "TCGAGT": 1746, + "TCGAGC": 1747, + "TCGAGG": 1748, + "TCGTAA": 1749, + "TCGTAT": 1750, + "TCGTAC": 1751, + "TCGTAG": 1752, + "TCGTTA": 1753, + "TCGTTT": 1754, + "TCGTTC": 1755, + "TCGTTG": 1756, + "TCGTCA": 1757, + "TCGTCT": 1758, + "TCGTCC": 1759, + "TCGTCG": 1760, + "TCGTGA": 1761, + "TCGTGT": 1762, + "TCGTGC": 1763, + "TCGTGG": 1764, + "TCGCAA": 1765, + "TCGCAT": 1766, + "TCGCAC": 1767, + "TCGCAG": 1768, + "TCGCTA": 1769, + "TCGCTT": 1770, + "TCGCTC": 1771, + "TCGCTG": 1772, + "TCGCCA": 1773, + "TCGCCT": 1774, + "TCGCCC": 1775, + "TCGCCG": 1776, + "TCGCGA": 1777, + "TCGCGT": 1778, + "TCGCGC": 1779, + "TCGCGG": 1780, + "TCGGAA": 1781, + "TCGGAT": 1782, + "TCGGAC": 1783, + "TCGGAG": 1784, + "TCGGTA": 1785, + "TCGGTT": 1786, + "TCGGTC": 1787, + "TCGGTG": 1788, + "TCGGCA": 1789, + "TCGGCT": 1790, + "TCGGCC": 1791, + "TCGGCG": 1792, + "TCGGGA": 1793, + "TCGGGT": 1794, + "TCGGGC": 1795, + "TCGGGG": 1796, + "TGAAAA": 1797, + "TGAAAT": 1798, + "TGAAAC": 1799, + "TGAAAG": 1800, + "TGAATA": 1801, + "TGAATT": 1802, + "TGAATC": 1803, + "TGAATG": 1804, + "TGAACA": 1805, + "TGAACT": 1806, + "TGAACC": 1807, + "TGAACG": 1808, + "TGAAGA": 1809, + "TGAAGT": 1810, + "TGAAGC": 1811, + "TGAAGG": 1812, + "TGATAA": 1813, + "TGATAT": 1814, + "TGATAC": 1815, + "TGATAG": 1816, + "TGATTA": 1817, + "TGATTT": 1818, + "TGATTC": 1819, + "TGATTG": 1820, + "TGATCA": 1821, + "TGATCT": 1822, + "TGATCC": 1823, + "TGATCG": 1824, + "TGATGA": 1825, + "TGATGT": 1826, + "TGATGC": 1827, + "TGATGG": 1828, + "TGACAA": 1829, + "TGACAT": 1830, + "TGACAC": 1831, + "TGACAG": 1832, + "TGACTA": 1833, + "TGACTT": 1834, + "TGACTC": 1835, + "TGACTG": 1836, + "TGACCA": 1837, + "TGACCT": 1838, + "TGACCC": 1839, + "TGACCG": 1840, + "TGACGA": 1841, + "TGACGT": 1842, + "TGACGC": 1843, + "TGACGG": 1844, + "TGAGAA": 1845, + "TGAGAT": 1846, + "TGAGAC": 1847, + "TGAGAG": 1848, + "TGAGTA": 1849, + "TGAGTT": 1850, + "TGAGTC": 1851, + "TGAGTG": 1852, + "TGAGCA": 1853, + "TGAGCT": 1854, + "TGAGCC": 1855, + "TGAGCG": 1856, + "TGAGGA": 1857, + "TGAGGT": 1858, + "TGAGGC": 1859, + "TGAGGG": 1860, + "TGTAAA": 1861, + "TGTAAT": 1862, + "TGTAAC": 1863, + "TGTAAG": 1864, + "TGTATA": 1865, + "TGTATT": 1866, + "TGTATC": 1867, + "TGTATG": 1868, + "TGTACA": 1869, + "TGTACT": 1870, + "TGTACC": 1871, + "TGTACG": 1872, + "TGTAGA": 1873, + "TGTAGT": 1874, + "TGTAGC": 1875, + "TGTAGG": 1876, + "TGTTAA": 1877, + "TGTTAT": 1878, + "TGTTAC": 1879, + "TGTTAG": 1880, + "TGTTTA": 1881, + "TGTTTT": 1882, + "TGTTTC": 1883, + "TGTTTG": 1884, + "TGTTCA": 1885, + "TGTTCT": 1886, + "TGTTCC": 1887, + "TGTTCG": 1888, + "TGTTGA": 1889, + "TGTTGT": 1890, + "TGTTGC": 1891, + "TGTTGG": 1892, + "TGTCAA": 1893, + "TGTCAT": 1894, + "TGTCAC": 1895, + "TGTCAG": 1896, + "TGTCTA": 1897, + "TGTCTT": 1898, + "TGTCTC": 1899, + "TGTCTG": 1900, + "TGTCCA": 1901, + "TGTCCT": 1902, + "TGTCCC": 1903, + "TGTCCG": 1904, + "TGTCGA": 1905, + "TGTCGT": 1906, + "TGTCGC": 1907, + "TGTCGG": 1908, + "TGTGAA": 1909, + "TGTGAT": 1910, + "TGTGAC": 1911, + "TGTGAG": 1912, + "TGTGTA": 1913, + "TGTGTT": 1914, + "TGTGTC": 1915, + "TGTGTG": 1916, + "TGTGCA": 1917, + "TGTGCT": 1918, + "TGTGCC": 1919, + "TGTGCG": 1920, + "TGTGGA": 1921, + "TGTGGT": 1922, + "TGTGGC": 1923, + "TGTGGG": 1924, + "TGCAAA": 1925, + "TGCAAT": 1926, + "TGCAAC": 1927, + "TGCAAG": 1928, + "TGCATA": 1929, + "TGCATT": 1930, + "TGCATC": 1931, + "TGCATG": 1932, + "TGCACA": 1933, + "TGCACT": 1934, + "TGCACC": 1935, + "TGCACG": 1936, + "TGCAGA": 1937, + "TGCAGT": 1938, + "TGCAGC": 1939, + "TGCAGG": 1940, + "TGCTAA": 1941, + "TGCTAT": 1942, + "TGCTAC": 1943, + "TGCTAG": 1944, + "TGCTTA": 1945, + "TGCTTT": 1946, + "TGCTTC": 1947, + "TGCTTG": 1948, + "TGCTCA": 1949, + "TGCTCT": 1950, + "TGCTCC": 1951, + "TGCTCG": 1952, + "TGCTGA": 1953, + "TGCTGT": 1954, + "TGCTGC": 1955, + "TGCTGG": 1956, + "TGCCAA": 1957, + "TGCCAT": 1958, + "TGCCAC": 1959, + "TGCCAG": 1960, + "TGCCTA": 1961, + "TGCCTT": 1962, + "TGCCTC": 1963, + "TGCCTG": 1964, + "TGCCCA": 1965, + "TGCCCT": 1966, + "TGCCCC": 1967, + "TGCCCG": 1968, + "TGCCGA": 1969, + "TGCCGT": 1970, + "TGCCGC": 1971, + "TGCCGG": 1972, + "TGCGAA": 1973, + "TGCGAT": 1974, + "TGCGAC": 1975, + "TGCGAG": 1976, + "TGCGTA": 1977, + "TGCGTT": 1978, + "TGCGTC": 1979, + "TGCGTG": 1980, + "TGCGCA": 1981, + "TGCGCT": 1982, + "TGCGCC": 1983, + "TGCGCG": 1984, + "TGCGGA": 1985, + "TGCGGT": 1986, + "TGCGGC": 1987, + "TGCGGG": 1988, + "TGGAAA": 1989, + "TGGAAT": 1990, + "TGGAAC": 1991, + "TGGAAG": 1992, + "TGGATA": 1993, + "TGGATT": 1994, + "TGGATC": 1995, + "TGGATG": 1996, + "TGGACA": 1997, + "TGGACT": 1998, + "TGGACC": 1999, + "TGGACG": 2000, + "TGGAGA": 2001, + "TGGAGT": 2002, + "TGGAGC": 2003, + "TGGAGG": 2004, + "TGGTAA": 2005, + "TGGTAT": 2006, + "TGGTAC": 2007, + "TGGTAG": 2008, + "TGGTTA": 2009, + "TGGTTT": 2010, + "TGGTTC": 2011, + "TGGTTG": 2012, + "TGGTCA": 2013, + "TGGTCT": 2014, + "TGGTCC": 2015, + "TGGTCG": 2016, + "TGGTGA": 2017, + "TGGTGT": 2018, + "TGGTGC": 2019, + "TGGTGG": 2020, + "TGGCAA": 2021, + "TGGCAT": 2022, + "TGGCAC": 2023, + "TGGCAG": 2024, + "TGGCTA": 2025, + "TGGCTT": 2026, + "TGGCTC": 2027, + "TGGCTG": 2028, + "TGGCCA": 2029, + "TGGCCT": 2030, + "TGGCCC": 2031, + "TGGCCG": 2032, + "TGGCGA": 2033, + "TGGCGT": 2034, + "TGGCGC": 2035, + "TGGCGG": 2036, + "TGGGAA": 2037, + "TGGGAT": 2038, + "TGGGAC": 2039, + "TGGGAG": 2040, + "TGGGTA": 2041, + "TGGGTT": 2042, + "TGGGTC": 2043, + "TGGGTG": 2044, + "TGGGCA": 2045, + "TGGGCT": 2046, + "TGGGCC": 2047, + "TGGGCG": 2048, + "TGGGGA": 2049, + "TGGGGT": 2050, + "TGGGGC": 2051, + "TGGGGG": 2052, + "CAAAAA": 2053, + "CAAAAT": 2054, + "CAAAAC": 2055, + "CAAAAG": 2056, + "CAAATA": 2057, + "CAAATT": 2058, + "CAAATC": 2059, + "CAAATG": 2060, + "CAAACA": 2061, + "CAAACT": 2062, + "CAAACC": 2063, + "CAAACG": 2064, + "CAAAGA": 2065, + "CAAAGT": 2066, + "CAAAGC": 2067, + "CAAAGG": 2068, + "CAATAA": 2069, + "CAATAT": 2070, + "CAATAC": 2071, + "CAATAG": 2072, + "CAATTA": 2073, + "CAATTT": 2074, + "CAATTC": 2075, + "CAATTG": 2076, + "CAATCA": 2077, + "CAATCT": 2078, + "CAATCC": 2079, + "CAATCG": 2080, + "CAATGA": 2081, + "CAATGT": 2082, + "CAATGC": 2083, + "CAATGG": 2084, + "CAACAA": 2085, + "CAACAT": 2086, + "CAACAC": 2087, + "CAACAG": 2088, + "CAACTA": 2089, + "CAACTT": 2090, + "CAACTC": 2091, + "CAACTG": 2092, + "CAACCA": 2093, + "CAACCT": 2094, + "CAACCC": 2095, + "CAACCG": 2096, + "CAACGA": 2097, + "CAACGT": 2098, + "CAACGC": 2099, + "CAACGG": 2100, + "CAAGAA": 2101, + "CAAGAT": 2102, + "CAAGAC": 2103, + "CAAGAG": 2104, + "CAAGTA": 2105, + "CAAGTT": 2106, + "CAAGTC": 2107, + "CAAGTG": 2108, + "CAAGCA": 2109, + "CAAGCT": 2110, + "CAAGCC": 2111, + "CAAGCG": 2112, + "CAAGGA": 2113, + "CAAGGT": 2114, + "CAAGGC": 2115, + "CAAGGG": 2116, + "CATAAA": 2117, + "CATAAT": 2118, + "CATAAC": 2119, + "CATAAG": 2120, + "CATATA": 2121, + "CATATT": 2122, + "CATATC": 2123, + "CATATG": 2124, + "CATACA": 2125, + "CATACT": 2126, + "CATACC": 2127, + "CATACG": 2128, + "CATAGA": 2129, + "CATAGT": 2130, + "CATAGC": 2131, + "CATAGG": 2132, + "CATTAA": 2133, + "CATTAT": 2134, + "CATTAC": 2135, + "CATTAG": 2136, + "CATTTA": 2137, + "CATTTT": 2138, + "CATTTC": 2139, + "CATTTG": 2140, + "CATTCA": 2141, + "CATTCT": 2142, + "CATTCC": 2143, + "CATTCG": 2144, + "CATTGA": 2145, + "CATTGT": 2146, + "CATTGC": 2147, + "CATTGG": 2148, + "CATCAA": 2149, + "CATCAT": 2150, + "CATCAC": 2151, + "CATCAG": 2152, + "CATCTA": 2153, + "CATCTT": 2154, + "CATCTC": 2155, + "CATCTG": 2156, + "CATCCA": 2157, + "CATCCT": 2158, + "CATCCC": 2159, + "CATCCG": 2160, + "CATCGA": 2161, + "CATCGT": 2162, + "CATCGC": 2163, + "CATCGG": 2164, + "CATGAA": 2165, + "CATGAT": 2166, + "CATGAC": 2167, + "CATGAG": 2168, + "CATGTA": 2169, + "CATGTT": 2170, + "CATGTC": 2171, + "CATGTG": 2172, + "CATGCA": 2173, + "CATGCT": 2174, + "CATGCC": 2175, + "CATGCG": 2176, + "CATGGA": 2177, + "CATGGT": 2178, + "CATGGC": 2179, + "CATGGG": 2180, + "CACAAA": 2181, + "CACAAT": 2182, + "CACAAC": 2183, + "CACAAG": 2184, + "CACATA": 2185, + "CACATT": 2186, + "CACATC": 2187, + "CACATG": 2188, + "CACACA": 2189, + "CACACT": 2190, + "CACACC": 2191, + "CACACG": 2192, + "CACAGA": 2193, + "CACAGT": 2194, + "CACAGC": 2195, + "CACAGG": 2196, + "CACTAA": 2197, + "CACTAT": 2198, + "CACTAC": 2199, + "CACTAG": 2200, + "CACTTA": 2201, + "CACTTT": 2202, + "CACTTC": 2203, + "CACTTG": 2204, + "CACTCA": 2205, + "CACTCT": 2206, + "CACTCC": 2207, + "CACTCG": 2208, + "CACTGA": 2209, + "CACTGT": 2210, + "CACTGC": 2211, + "CACTGG": 2212, + "CACCAA": 2213, + "CACCAT": 2214, + "CACCAC": 2215, + "CACCAG": 2216, + "CACCTA": 2217, + "CACCTT": 2218, + "CACCTC": 2219, + "CACCTG": 2220, + "CACCCA": 2221, + "CACCCT": 2222, + "CACCCC": 2223, + "CACCCG": 2224, + "CACCGA": 2225, + "CACCGT": 2226, + "CACCGC": 2227, + "CACCGG": 2228, + "CACGAA": 2229, + "CACGAT": 2230, + "CACGAC": 2231, + "CACGAG": 2232, + "CACGTA": 2233, + "CACGTT": 2234, + "CACGTC": 2235, + "CACGTG": 2236, + "CACGCA": 2237, + "CACGCT": 2238, + "CACGCC": 2239, + "CACGCG": 2240, + "CACGGA": 2241, + "CACGGT": 2242, + "CACGGC": 2243, + "CACGGG": 2244, + "CAGAAA": 2245, + "CAGAAT": 2246, + "CAGAAC": 2247, + "CAGAAG": 2248, + "CAGATA": 2249, + "CAGATT": 2250, + "CAGATC": 2251, + "CAGATG": 2252, + "CAGACA": 2253, + "CAGACT": 2254, + "CAGACC": 2255, + "CAGACG": 2256, + "CAGAGA": 2257, + "CAGAGT": 2258, + "CAGAGC": 2259, + "CAGAGG": 2260, + "CAGTAA": 2261, + "CAGTAT": 2262, + "CAGTAC": 2263, + "CAGTAG": 2264, + "CAGTTA": 2265, + "CAGTTT": 2266, + "CAGTTC": 2267, + "CAGTTG": 2268, + "CAGTCA": 2269, + "CAGTCT": 2270, + "CAGTCC": 2271, + "CAGTCG": 2272, + "CAGTGA": 2273, + "CAGTGT": 2274, + "CAGTGC": 2275, + "CAGTGG": 2276, + "CAGCAA": 2277, + "CAGCAT": 2278, + "CAGCAC": 2279, + "CAGCAG": 2280, + "CAGCTA": 2281, + "CAGCTT": 2282, + "CAGCTC": 2283, + "CAGCTG": 2284, + "CAGCCA": 2285, + "CAGCCT": 2286, + "CAGCCC": 2287, + "CAGCCG": 2288, + "CAGCGA": 2289, + "CAGCGT": 2290, + "CAGCGC": 2291, + "CAGCGG": 2292, + "CAGGAA": 2293, + "CAGGAT": 2294, + "CAGGAC": 2295, + "CAGGAG": 2296, + "CAGGTA": 2297, + "CAGGTT": 2298, + "CAGGTC": 2299, + "CAGGTG": 2300, + "CAGGCA": 2301, + "CAGGCT": 2302, + "CAGGCC": 2303, + "CAGGCG": 2304, + "CAGGGA": 2305, + "CAGGGT": 2306, + "CAGGGC": 2307, + "CAGGGG": 2308, + "CTAAAA": 2309, + "CTAAAT": 2310, + "CTAAAC": 2311, + "CTAAAG": 2312, + "CTAATA": 2313, + "CTAATT": 2314, + "CTAATC": 2315, + "CTAATG": 2316, + "CTAACA": 2317, + "CTAACT": 2318, + "CTAACC": 2319, + "CTAACG": 2320, + "CTAAGA": 2321, + "CTAAGT": 2322, + "CTAAGC": 2323, + "CTAAGG": 2324, + "CTATAA": 2325, + "CTATAT": 2326, + "CTATAC": 2327, + "CTATAG": 2328, + "CTATTA": 2329, + "CTATTT": 2330, + "CTATTC": 2331, + "CTATTG": 2332, + "CTATCA": 2333, + "CTATCT": 2334, + "CTATCC": 2335, + "CTATCG": 2336, + "CTATGA": 2337, + "CTATGT": 2338, + "CTATGC": 2339, + "CTATGG": 2340, + "CTACAA": 2341, + "CTACAT": 2342, + "CTACAC": 2343, + "CTACAG": 2344, + "CTACTA": 2345, + "CTACTT": 2346, + "CTACTC": 2347, + "CTACTG": 2348, + "CTACCA": 2349, + "CTACCT": 2350, + "CTACCC": 2351, + "CTACCG": 2352, + "CTACGA": 2353, + "CTACGT": 2354, + "CTACGC": 2355, + "CTACGG": 2356, + "CTAGAA": 2357, + "CTAGAT": 2358, + "CTAGAC": 2359, + "CTAGAG": 2360, + "CTAGTA": 2361, + "CTAGTT": 2362, + "CTAGTC": 2363, + "CTAGTG": 2364, + "CTAGCA": 2365, + "CTAGCT": 2366, + "CTAGCC": 2367, + "CTAGCG": 2368, + "CTAGGA": 2369, + "CTAGGT": 2370, + "CTAGGC": 2371, + "CTAGGG": 2372, + "CTTAAA": 2373, + "CTTAAT": 2374, + "CTTAAC": 2375, + "CTTAAG": 2376, + "CTTATA": 2377, + "CTTATT": 2378, + "CTTATC": 2379, + "CTTATG": 2380, + "CTTACA": 2381, + "CTTACT": 2382, + "CTTACC": 2383, + "CTTACG": 2384, + "CTTAGA": 2385, + "CTTAGT": 2386, + "CTTAGC": 2387, + "CTTAGG": 2388, + "CTTTAA": 2389, + "CTTTAT": 2390, + "CTTTAC": 2391, + "CTTTAG": 2392, + "CTTTTA": 2393, + "CTTTTT": 2394, + "CTTTTC": 2395, + "CTTTTG": 2396, + "CTTTCA": 2397, + "CTTTCT": 2398, + "CTTTCC": 2399, + "CTTTCG": 2400, + "CTTTGA": 2401, + "CTTTGT": 2402, + "CTTTGC": 2403, + "CTTTGG": 2404, + "CTTCAA": 2405, + "CTTCAT": 2406, + "CTTCAC": 2407, + "CTTCAG": 2408, + "CTTCTA": 2409, + "CTTCTT": 2410, + "CTTCTC": 2411, + "CTTCTG": 2412, + "CTTCCA": 2413, + "CTTCCT": 2414, + "CTTCCC": 2415, + "CTTCCG": 2416, + "CTTCGA": 2417, + "CTTCGT": 2418, + "CTTCGC": 2419, + "CTTCGG": 2420, + "CTTGAA": 2421, + "CTTGAT": 2422, + "CTTGAC": 2423, + "CTTGAG": 2424, + "CTTGTA": 2425, + "CTTGTT": 2426, + "CTTGTC": 2427, + "CTTGTG": 2428, + "CTTGCA": 2429, + "CTTGCT": 2430, + "CTTGCC": 2431, + "CTTGCG": 2432, + "CTTGGA": 2433, + "CTTGGT": 2434, + "CTTGGC": 2435, + "CTTGGG": 2436, + "CTCAAA": 2437, + "CTCAAT": 2438, + "CTCAAC": 2439, + "CTCAAG": 2440, + "CTCATA": 2441, + "CTCATT": 2442, + "CTCATC": 2443, + "CTCATG": 2444, + "CTCACA": 2445, + "CTCACT": 2446, + "CTCACC": 2447, + "CTCACG": 2448, + "CTCAGA": 2449, + "CTCAGT": 2450, + "CTCAGC": 2451, + "CTCAGG": 2452, + "CTCTAA": 2453, + "CTCTAT": 2454, + "CTCTAC": 2455, + "CTCTAG": 2456, + "CTCTTA": 2457, + "CTCTTT": 2458, + "CTCTTC": 2459, + "CTCTTG": 2460, + "CTCTCA": 2461, + "CTCTCT": 2462, + "CTCTCC": 2463, + "CTCTCG": 2464, + "CTCTGA": 2465, + "CTCTGT": 2466, + "CTCTGC": 2467, + "CTCTGG": 2468, + "CTCCAA": 2469, + "CTCCAT": 2470, + "CTCCAC": 2471, + "CTCCAG": 2472, + "CTCCTA": 2473, + "CTCCTT": 2474, + "CTCCTC": 2475, + "CTCCTG": 2476, + "CTCCCA": 2477, + "CTCCCT": 2478, + "CTCCCC": 2479, + "CTCCCG": 2480, + "CTCCGA": 2481, + "CTCCGT": 2482, + "CTCCGC": 2483, + "CTCCGG": 2484, + "CTCGAA": 2485, + "CTCGAT": 2486, + "CTCGAC": 2487, + "CTCGAG": 2488, + "CTCGTA": 2489, + "CTCGTT": 2490, + "CTCGTC": 2491, + "CTCGTG": 2492, + "CTCGCA": 2493, + "CTCGCT": 2494, + "CTCGCC": 2495, + "CTCGCG": 2496, + "CTCGGA": 2497, + "CTCGGT": 2498, + "CTCGGC": 2499, + "CTCGGG": 2500, + "CTGAAA": 2501, + "CTGAAT": 2502, + "CTGAAC": 2503, + "CTGAAG": 2504, + "CTGATA": 2505, + "CTGATT": 2506, + "CTGATC": 2507, + "CTGATG": 2508, + "CTGACA": 2509, + "CTGACT": 2510, + "CTGACC": 2511, + "CTGACG": 2512, + "CTGAGA": 2513, + "CTGAGT": 2514, + "CTGAGC": 2515, + "CTGAGG": 2516, + "CTGTAA": 2517, + "CTGTAT": 2518, + "CTGTAC": 2519, + "CTGTAG": 2520, + "CTGTTA": 2521, + "CTGTTT": 2522, + "CTGTTC": 2523, + "CTGTTG": 2524, + "CTGTCA": 2525, + "CTGTCT": 2526, + "CTGTCC": 2527, + "CTGTCG": 2528, + "CTGTGA": 2529, + "CTGTGT": 2530, + "CTGTGC": 2531, + "CTGTGG": 2532, + "CTGCAA": 2533, + "CTGCAT": 2534, + "CTGCAC": 2535, + "CTGCAG": 2536, + "CTGCTA": 2537, + "CTGCTT": 2538, + "CTGCTC": 2539, + "CTGCTG": 2540, + "CTGCCA": 2541, + "CTGCCT": 2542, + "CTGCCC": 2543, + "CTGCCG": 2544, + "CTGCGA": 2545, + "CTGCGT": 2546, + "CTGCGC": 2547, + "CTGCGG": 2548, + "CTGGAA": 2549, + "CTGGAT": 2550, + "CTGGAC": 2551, + "CTGGAG": 2552, + "CTGGTA": 2553, + "CTGGTT": 2554, + "CTGGTC": 2555, + "CTGGTG": 2556, + "CTGGCA": 2557, + "CTGGCT": 2558, + "CTGGCC": 2559, + "CTGGCG": 2560, + "CTGGGA": 2561, + "CTGGGT": 2562, + "CTGGGC": 2563, + "CTGGGG": 2564, + "CCAAAA": 2565, + "CCAAAT": 2566, + "CCAAAC": 2567, + "CCAAAG": 2568, + "CCAATA": 2569, + "CCAATT": 2570, + "CCAATC": 2571, + "CCAATG": 2572, + "CCAACA": 2573, + "CCAACT": 2574, + "CCAACC": 2575, + "CCAACG": 2576, + "CCAAGA": 2577, + "CCAAGT": 2578, + "CCAAGC": 2579, + "CCAAGG": 2580, + "CCATAA": 2581, + "CCATAT": 2582, + "CCATAC": 2583, + "CCATAG": 2584, + "CCATTA": 2585, + "CCATTT": 2586, + "CCATTC": 2587, + "CCATTG": 2588, + "CCATCA": 2589, + "CCATCT": 2590, + "CCATCC": 2591, + "CCATCG": 2592, + "CCATGA": 2593, + "CCATGT": 2594, + "CCATGC": 2595, + "CCATGG": 2596, + "CCACAA": 2597, + "CCACAT": 2598, + "CCACAC": 2599, + "CCACAG": 2600, + "CCACTA": 2601, + "CCACTT": 2602, + "CCACTC": 2603, + "CCACTG": 2604, + "CCACCA": 2605, + "CCACCT": 2606, + "CCACCC": 2607, + "CCACCG": 2608, + "CCACGA": 2609, + "CCACGT": 2610, + "CCACGC": 2611, + "CCACGG": 2612, + "CCAGAA": 2613, + "CCAGAT": 2614, + "CCAGAC": 2615, + "CCAGAG": 2616, + "CCAGTA": 2617, + "CCAGTT": 2618, + "CCAGTC": 2619, + "CCAGTG": 2620, + "CCAGCA": 2621, + "CCAGCT": 2622, + "CCAGCC": 2623, + "CCAGCG": 2624, + "CCAGGA": 2625, + "CCAGGT": 2626, + "CCAGGC": 2627, + "CCAGGG": 2628, + "CCTAAA": 2629, + "CCTAAT": 2630, + "CCTAAC": 2631, + "CCTAAG": 2632, + "CCTATA": 2633, + "CCTATT": 2634, + "CCTATC": 2635, + "CCTATG": 2636, + "CCTACA": 2637, + "CCTACT": 2638, + "CCTACC": 2639, + "CCTACG": 2640, + "CCTAGA": 2641, + "CCTAGT": 2642, + "CCTAGC": 2643, + "CCTAGG": 2644, + "CCTTAA": 2645, + "CCTTAT": 2646, + "CCTTAC": 2647, + "CCTTAG": 2648, + "CCTTTA": 2649, + "CCTTTT": 2650, + "CCTTTC": 2651, + "CCTTTG": 2652, + "CCTTCA": 2653, + "CCTTCT": 2654, + "CCTTCC": 2655, + "CCTTCG": 2656, + "CCTTGA": 2657, + "CCTTGT": 2658, + "CCTTGC": 2659, + "CCTTGG": 2660, + "CCTCAA": 2661, + "CCTCAT": 2662, + "CCTCAC": 2663, + "CCTCAG": 2664, + "CCTCTA": 2665, + "CCTCTT": 2666, + "CCTCTC": 2667, + "CCTCTG": 2668, + "CCTCCA": 2669, + "CCTCCT": 2670, + "CCTCCC": 2671, + "CCTCCG": 2672, + "CCTCGA": 2673, + "CCTCGT": 2674, + "CCTCGC": 2675, + "CCTCGG": 2676, + "CCTGAA": 2677, + "CCTGAT": 2678, + "CCTGAC": 2679, + "CCTGAG": 2680, + "CCTGTA": 2681, + "CCTGTT": 2682, + "CCTGTC": 2683, + "CCTGTG": 2684, + "CCTGCA": 2685, + "CCTGCT": 2686, + "CCTGCC": 2687, + "CCTGCG": 2688, + "CCTGGA": 2689, + "CCTGGT": 2690, + "CCTGGC": 2691, + "CCTGGG": 2692, + "CCCAAA": 2693, + "CCCAAT": 2694, + "CCCAAC": 2695, + "CCCAAG": 2696, + "CCCATA": 2697, + "CCCATT": 2698, + "CCCATC": 2699, + "CCCATG": 2700, + "CCCACA": 2701, + "CCCACT": 2702, + "CCCACC": 2703, + "CCCACG": 2704, + "CCCAGA": 2705, + "CCCAGT": 2706, + "CCCAGC": 2707, + "CCCAGG": 2708, + "CCCTAA": 2709, + "CCCTAT": 2710, + "CCCTAC": 2711, + "CCCTAG": 2712, + "CCCTTA": 2713, + "CCCTTT": 2714, + "CCCTTC": 2715, + "CCCTTG": 2716, + "CCCTCA": 2717, + "CCCTCT": 2718, + "CCCTCC": 2719, + "CCCTCG": 2720, + "CCCTGA": 2721, + "CCCTGT": 2722, + "CCCTGC": 2723, + "CCCTGG": 2724, + "CCCCAA": 2725, + "CCCCAT": 2726, + "CCCCAC": 2727, + "CCCCAG": 2728, + "CCCCTA": 2729, + "CCCCTT": 2730, + "CCCCTC": 2731, + "CCCCTG": 2732, + "CCCCCA": 2733, + "CCCCCT": 2734, + "CCCCCC": 2735, + "CCCCCG": 2736, + "CCCCGA": 2737, + "CCCCGT": 2738, + "CCCCGC": 2739, + "CCCCGG": 2740, + "CCCGAA": 2741, + "CCCGAT": 2742, + "CCCGAC": 2743, + "CCCGAG": 2744, + "CCCGTA": 2745, + "CCCGTT": 2746, + "CCCGTC": 2747, + "CCCGTG": 2748, + "CCCGCA": 2749, + "CCCGCT": 2750, + "CCCGCC": 2751, + "CCCGCG": 2752, + "CCCGGA": 2753, + "CCCGGT": 2754, + "CCCGGC": 2755, + "CCCGGG": 2756, + "CCGAAA": 2757, + "CCGAAT": 2758, + "CCGAAC": 2759, + "CCGAAG": 2760, + "CCGATA": 2761, + "CCGATT": 2762, + "CCGATC": 2763, + "CCGATG": 2764, + "CCGACA": 2765, + "CCGACT": 2766, + "CCGACC": 2767, + "CCGACG": 2768, + "CCGAGA": 2769, + "CCGAGT": 2770, + "CCGAGC": 2771, + "CCGAGG": 2772, + "CCGTAA": 2773, + "CCGTAT": 2774, + "CCGTAC": 2775, + "CCGTAG": 2776, + "CCGTTA": 2777, + "CCGTTT": 2778, + "CCGTTC": 2779, + "CCGTTG": 2780, + "CCGTCA": 2781, + "CCGTCT": 2782, + "CCGTCC": 2783, + "CCGTCG": 2784, + "CCGTGA": 2785, + "CCGTGT": 2786, + "CCGTGC": 2787, + "CCGTGG": 2788, + "CCGCAA": 2789, + "CCGCAT": 2790, + "CCGCAC": 2791, + "CCGCAG": 2792, + "CCGCTA": 2793, + "CCGCTT": 2794, + "CCGCTC": 2795, + "CCGCTG": 2796, + "CCGCCA": 2797, + "CCGCCT": 2798, + "CCGCCC": 2799, + "CCGCCG": 2800, + "CCGCGA": 2801, + "CCGCGT": 2802, + "CCGCGC": 2803, + "CCGCGG": 2804, + "CCGGAA": 2805, + "CCGGAT": 2806, + "CCGGAC": 2807, + "CCGGAG": 2808, + "CCGGTA": 2809, + "CCGGTT": 2810, + "CCGGTC": 2811, + "CCGGTG": 2812, + "CCGGCA": 2813, + "CCGGCT": 2814, + "CCGGCC": 2815, + "CCGGCG": 2816, + "CCGGGA": 2817, + "CCGGGT": 2818, + "CCGGGC": 2819, + "CCGGGG": 2820, + "CGAAAA": 2821, + "CGAAAT": 2822, + "CGAAAC": 2823, + "CGAAAG": 2824, + "CGAATA": 2825, + "CGAATT": 2826, + "CGAATC": 2827, + "CGAATG": 2828, + "CGAACA": 2829, + "CGAACT": 2830, + "CGAACC": 2831, + "CGAACG": 2832, + "CGAAGA": 2833, + "CGAAGT": 2834, + "CGAAGC": 2835, + "CGAAGG": 2836, + "CGATAA": 2837, + "CGATAT": 2838, + "CGATAC": 2839, + "CGATAG": 2840, + "CGATTA": 2841, + "CGATTT": 2842, + "CGATTC": 2843, + "CGATTG": 2844, + "CGATCA": 2845, + "CGATCT": 2846, + "CGATCC": 2847, + "CGATCG": 2848, + "CGATGA": 2849, + "CGATGT": 2850, + "CGATGC": 2851, + "CGATGG": 2852, + "CGACAA": 2853, + "CGACAT": 2854, + "CGACAC": 2855, + "CGACAG": 2856, + "CGACTA": 2857, + "CGACTT": 2858, + "CGACTC": 2859, + "CGACTG": 2860, + "CGACCA": 2861, + "CGACCT": 2862, + "CGACCC": 2863, + "CGACCG": 2864, + "CGACGA": 2865, + "CGACGT": 2866, + "CGACGC": 2867, + "CGACGG": 2868, + "CGAGAA": 2869, + "CGAGAT": 2870, + "CGAGAC": 2871, + "CGAGAG": 2872, + "CGAGTA": 2873, + "CGAGTT": 2874, + "CGAGTC": 2875, + "CGAGTG": 2876, + "CGAGCA": 2877, + "CGAGCT": 2878, + "CGAGCC": 2879, + "CGAGCG": 2880, + "CGAGGA": 2881, + "CGAGGT": 2882, + "CGAGGC": 2883, + "CGAGGG": 2884, + "CGTAAA": 2885, + "CGTAAT": 2886, + "CGTAAC": 2887, + "CGTAAG": 2888, + "CGTATA": 2889, + "CGTATT": 2890, + "CGTATC": 2891, + "CGTATG": 2892, + "CGTACA": 2893, + "CGTACT": 2894, + "CGTACC": 2895, + "CGTACG": 2896, + "CGTAGA": 2897, + "CGTAGT": 2898, + "CGTAGC": 2899, + "CGTAGG": 2900, + "CGTTAA": 2901, + "CGTTAT": 2902, + "CGTTAC": 2903, + "CGTTAG": 2904, + "CGTTTA": 2905, + "CGTTTT": 2906, + "CGTTTC": 2907, + "CGTTTG": 2908, + "CGTTCA": 2909, + "CGTTCT": 2910, + "CGTTCC": 2911, + "CGTTCG": 2912, + "CGTTGA": 2913, + "CGTTGT": 2914, + "CGTTGC": 2915, + "CGTTGG": 2916, + "CGTCAA": 2917, + "CGTCAT": 2918, + "CGTCAC": 2919, + "CGTCAG": 2920, + "CGTCTA": 2921, + "CGTCTT": 2922, + "CGTCTC": 2923, + "CGTCTG": 2924, + "CGTCCA": 2925, + "CGTCCT": 2926, + "CGTCCC": 2927, + "CGTCCG": 2928, + "CGTCGA": 2929, + "CGTCGT": 2930, + "CGTCGC": 2931, + "CGTCGG": 2932, + "CGTGAA": 2933, + "CGTGAT": 2934, + "CGTGAC": 2935, + "CGTGAG": 2936, + "CGTGTA": 2937, + "CGTGTT": 2938, + "CGTGTC": 2939, + "CGTGTG": 2940, + "CGTGCA": 2941, + "CGTGCT": 2942, + "CGTGCC": 2943, + "CGTGCG": 2944, + "CGTGGA": 2945, + "CGTGGT": 2946, + "CGTGGC": 2947, + "CGTGGG": 2948, + "CGCAAA": 2949, + "CGCAAT": 2950, + "CGCAAC": 2951, + "CGCAAG": 2952, + "CGCATA": 2953, + "CGCATT": 2954, + "CGCATC": 2955, + "CGCATG": 2956, + "CGCACA": 2957, + "CGCACT": 2958, + "CGCACC": 2959, + "CGCACG": 2960, + "CGCAGA": 2961, + "CGCAGT": 2962, + "CGCAGC": 2963, + "CGCAGG": 2964, + "CGCTAA": 2965, + "CGCTAT": 2966, + "CGCTAC": 2967, + "CGCTAG": 2968, + "CGCTTA": 2969, + "CGCTTT": 2970, + "CGCTTC": 2971, + "CGCTTG": 2972, + "CGCTCA": 2973, + "CGCTCT": 2974, + "CGCTCC": 2975, + "CGCTCG": 2976, + "CGCTGA": 2977, + "CGCTGT": 2978, + "CGCTGC": 2979, + "CGCTGG": 2980, + "CGCCAA": 2981, + "CGCCAT": 2982, + "CGCCAC": 2983, + "CGCCAG": 2984, + "CGCCTA": 2985, + "CGCCTT": 2986, + "CGCCTC": 2987, + "CGCCTG": 2988, + "CGCCCA": 2989, + "CGCCCT": 2990, + "CGCCCC": 2991, + "CGCCCG": 2992, + "CGCCGA": 2993, + "CGCCGT": 2994, + "CGCCGC": 2995, + "CGCCGG": 2996, + "CGCGAA": 2997, + "CGCGAT": 2998, + "CGCGAC": 2999, + "CGCGAG": 3000, + "CGCGTA": 3001, + "CGCGTT": 3002, + "CGCGTC": 3003, + "CGCGTG": 3004, + "CGCGCA": 3005, + "CGCGCT": 3006, + "CGCGCC": 3007, + "CGCGCG": 3008, + "CGCGGA": 3009, + "CGCGGT": 3010, + "CGCGGC": 3011, + "CGCGGG": 3012, + "CGGAAA": 3013, + "CGGAAT": 3014, + "CGGAAC": 3015, + "CGGAAG": 3016, + "CGGATA": 3017, + "CGGATT": 3018, + "CGGATC": 3019, + "CGGATG": 3020, + "CGGACA": 3021, + "CGGACT": 3022, + "CGGACC": 3023, + "CGGACG": 3024, + "CGGAGA": 3025, + "CGGAGT": 3026, + "CGGAGC": 3027, + "CGGAGG": 3028, + "CGGTAA": 3029, + "CGGTAT": 3030, + "CGGTAC": 3031, + "CGGTAG": 3032, + "CGGTTA": 3033, + "CGGTTT": 3034, + "CGGTTC": 3035, + "CGGTTG": 3036, + "CGGTCA": 3037, + "CGGTCT": 3038, + "CGGTCC": 3039, + "CGGTCG": 3040, + "CGGTGA": 3041, + "CGGTGT": 3042, + "CGGTGC": 3043, + "CGGTGG": 3044, + "CGGCAA": 3045, + "CGGCAT": 3046, + "CGGCAC": 3047, + "CGGCAG": 3048, + "CGGCTA": 3049, + "CGGCTT": 3050, + "CGGCTC": 3051, + "CGGCTG": 3052, + "CGGCCA": 3053, + "CGGCCT": 3054, + "CGGCCC": 3055, + "CGGCCG": 3056, + "CGGCGA": 3057, + "CGGCGT": 3058, + "CGGCGC": 3059, + "CGGCGG": 3060, + "CGGGAA": 3061, + "CGGGAT": 3062, + "CGGGAC": 3063, + "CGGGAG": 3064, + "CGGGTA": 3065, + "CGGGTT": 3066, + "CGGGTC": 3067, + "CGGGTG": 3068, + "CGGGCA": 3069, + "CGGGCT": 3070, + "CGGGCC": 3071, + "CGGGCG": 3072, + "CGGGGA": 3073, + "CGGGGT": 3074, + "CGGGGC": 3075, + "CGGGGG": 3076, + "GAAAAA": 3077, + "GAAAAT": 3078, + "GAAAAC": 3079, + "GAAAAG": 3080, + "GAAATA": 3081, + "GAAATT": 3082, + "GAAATC": 3083, + "GAAATG": 3084, + "GAAACA": 3085, + "GAAACT": 3086, + "GAAACC": 3087, + "GAAACG": 3088, + "GAAAGA": 3089, + "GAAAGT": 3090, + "GAAAGC": 3091, + "GAAAGG": 3092, + "GAATAA": 3093, + "GAATAT": 3094, + "GAATAC": 3095, + "GAATAG": 3096, + "GAATTA": 3097, + "GAATTT": 3098, + "GAATTC": 3099, + "GAATTG": 3100, + "GAATCA": 3101, + "GAATCT": 3102, + "GAATCC": 3103, + "GAATCG": 3104, + "GAATGA": 3105, + "GAATGT": 3106, + "GAATGC": 3107, + "GAATGG": 3108, + "GAACAA": 3109, + "GAACAT": 3110, + "GAACAC": 3111, + "GAACAG": 3112, + "GAACTA": 3113, + "GAACTT": 3114, + "GAACTC": 3115, + "GAACTG": 3116, + "GAACCA": 3117, + "GAACCT": 3118, + "GAACCC": 3119, + "GAACCG": 3120, + "GAACGA": 3121, + "GAACGT": 3122, + "GAACGC": 3123, + "GAACGG": 3124, + "GAAGAA": 3125, + "GAAGAT": 3126, + "GAAGAC": 3127, + "GAAGAG": 3128, + "GAAGTA": 3129, + "GAAGTT": 3130, + "GAAGTC": 3131, + "GAAGTG": 3132, + "GAAGCA": 3133, + "GAAGCT": 3134, + "GAAGCC": 3135, + "GAAGCG": 3136, + "GAAGGA": 3137, + "GAAGGT": 3138, + "GAAGGC": 3139, + "GAAGGG": 3140, + "GATAAA": 3141, + "GATAAT": 3142, + "GATAAC": 3143, + "GATAAG": 3144, + "GATATA": 3145, + "GATATT": 3146, + "GATATC": 3147, + "GATATG": 3148, + "GATACA": 3149, + "GATACT": 3150, + "GATACC": 3151, + "GATACG": 3152, + "GATAGA": 3153, + "GATAGT": 3154, + "GATAGC": 3155, + "GATAGG": 3156, + "GATTAA": 3157, + "GATTAT": 3158, + "GATTAC": 3159, + "GATTAG": 3160, + "GATTTA": 3161, + "GATTTT": 3162, + "GATTTC": 3163, + "GATTTG": 3164, + "GATTCA": 3165, + "GATTCT": 3166, + "GATTCC": 3167, + "GATTCG": 3168, + "GATTGA": 3169, + "GATTGT": 3170, + "GATTGC": 3171, + "GATTGG": 3172, + "GATCAA": 3173, + "GATCAT": 3174, + "GATCAC": 3175, + "GATCAG": 3176, + "GATCTA": 3177, + "GATCTT": 3178, + "GATCTC": 3179, + "GATCTG": 3180, + "GATCCA": 3181, + "GATCCT": 3182, + "GATCCC": 3183, + "GATCCG": 3184, + "GATCGA": 3185, + "GATCGT": 3186, + "GATCGC": 3187, + "GATCGG": 3188, + "GATGAA": 3189, + "GATGAT": 3190, + "GATGAC": 3191, + "GATGAG": 3192, + "GATGTA": 3193, + "GATGTT": 3194, + "GATGTC": 3195, + "GATGTG": 3196, + "GATGCA": 3197, + "GATGCT": 3198, + "GATGCC": 3199, + "GATGCG": 3200, + "GATGGA": 3201, + "GATGGT": 3202, + "GATGGC": 3203, + "GATGGG": 3204, + "GACAAA": 3205, + "GACAAT": 3206, + "GACAAC": 3207, + "GACAAG": 3208, + "GACATA": 3209, + "GACATT": 3210, + "GACATC": 3211, + "GACATG": 3212, + "GACACA": 3213, + "GACACT": 3214, + "GACACC": 3215, + "GACACG": 3216, + "GACAGA": 3217, + "GACAGT": 3218, + "GACAGC": 3219, + "GACAGG": 3220, + "GACTAA": 3221, + "GACTAT": 3222, + "GACTAC": 3223, + "GACTAG": 3224, + "GACTTA": 3225, + "GACTTT": 3226, + "GACTTC": 3227, + "GACTTG": 3228, + "GACTCA": 3229, + "GACTCT": 3230, + "GACTCC": 3231, + "GACTCG": 3232, + "GACTGA": 3233, + "GACTGT": 3234, + "GACTGC": 3235, + "GACTGG": 3236, + "GACCAA": 3237, + "GACCAT": 3238, + "GACCAC": 3239, + "GACCAG": 3240, + "GACCTA": 3241, + "GACCTT": 3242, + "GACCTC": 3243, + "GACCTG": 3244, + "GACCCA": 3245, + "GACCCT": 3246, + "GACCCC": 3247, + "GACCCG": 3248, + "GACCGA": 3249, + "GACCGT": 3250, + "GACCGC": 3251, + "GACCGG": 3252, + "GACGAA": 3253, + "GACGAT": 3254, + "GACGAC": 3255, + "GACGAG": 3256, + "GACGTA": 3257, + "GACGTT": 3258, + "GACGTC": 3259, + "GACGTG": 3260, + "GACGCA": 3261, + "GACGCT": 3262, + "GACGCC": 3263, + "GACGCG": 3264, + "GACGGA": 3265, + "GACGGT": 3266, + "GACGGC": 3267, + "GACGGG": 3268, + "GAGAAA": 3269, + "GAGAAT": 3270, + "GAGAAC": 3271, + "GAGAAG": 3272, + "GAGATA": 3273, + "GAGATT": 3274, + "GAGATC": 3275, + "GAGATG": 3276, + "GAGACA": 3277, + "GAGACT": 3278, + "GAGACC": 3279, + "GAGACG": 3280, + "GAGAGA": 3281, + "GAGAGT": 3282, + "GAGAGC": 3283, + "GAGAGG": 3284, + "GAGTAA": 3285, + "GAGTAT": 3286, + "GAGTAC": 3287, + "GAGTAG": 3288, + "GAGTTA": 3289, + "GAGTTT": 3290, + "GAGTTC": 3291, + "GAGTTG": 3292, + "GAGTCA": 3293, + "GAGTCT": 3294, + "GAGTCC": 3295, + "GAGTCG": 3296, + "GAGTGA": 3297, + "GAGTGT": 3298, + "GAGTGC": 3299, + "GAGTGG": 3300, + "GAGCAA": 3301, + "GAGCAT": 3302, + "GAGCAC": 3303, + "GAGCAG": 3304, + "GAGCTA": 3305, + "GAGCTT": 3306, + "GAGCTC": 3307, + "GAGCTG": 3308, + "GAGCCA": 3309, + "GAGCCT": 3310, + "GAGCCC": 3311, + "GAGCCG": 3312, + "GAGCGA": 3313, + "GAGCGT": 3314, + "GAGCGC": 3315, + "GAGCGG": 3316, + "GAGGAA": 3317, + "GAGGAT": 3318, + "GAGGAC": 3319, + "GAGGAG": 3320, + "GAGGTA": 3321, + "GAGGTT": 3322, + "GAGGTC": 3323, + "GAGGTG": 3324, + "GAGGCA": 3325, + "GAGGCT": 3326, + "GAGGCC": 3327, + "GAGGCG": 3328, + "GAGGGA": 3329, + "GAGGGT": 3330, + "GAGGGC": 3331, + "GAGGGG": 3332, + "GTAAAA": 3333, + "GTAAAT": 3334, + "GTAAAC": 3335, + "GTAAAG": 3336, + "GTAATA": 3337, + "GTAATT": 3338, + "GTAATC": 3339, + "GTAATG": 3340, + "GTAACA": 3341, + "GTAACT": 3342, + "GTAACC": 3343, + "GTAACG": 3344, + "GTAAGA": 3345, + "GTAAGT": 3346, + "GTAAGC": 3347, + "GTAAGG": 3348, + "GTATAA": 3349, + "GTATAT": 3350, + "GTATAC": 3351, + "GTATAG": 3352, + "GTATTA": 3353, + "GTATTT": 3354, + "GTATTC": 3355, + "GTATTG": 3356, + "GTATCA": 3357, + "GTATCT": 3358, + "GTATCC": 3359, + "GTATCG": 3360, + "GTATGA": 3361, + "GTATGT": 3362, + "GTATGC": 3363, + "GTATGG": 3364, + "GTACAA": 3365, + "GTACAT": 3366, + "GTACAC": 3367, + "GTACAG": 3368, + "GTACTA": 3369, + "GTACTT": 3370, + "GTACTC": 3371, + "GTACTG": 3372, + "GTACCA": 3373, + "GTACCT": 3374, + "GTACCC": 3375, + "GTACCG": 3376, + "GTACGA": 3377, + "GTACGT": 3378, + "GTACGC": 3379, + "GTACGG": 3380, + "GTAGAA": 3381, + "GTAGAT": 3382, + "GTAGAC": 3383, + "GTAGAG": 3384, + "GTAGTA": 3385, + "GTAGTT": 3386, + "GTAGTC": 3387, + "GTAGTG": 3388, + "GTAGCA": 3389, + "GTAGCT": 3390, + "GTAGCC": 3391, + "GTAGCG": 3392, + "GTAGGA": 3393, + "GTAGGT": 3394, + "GTAGGC": 3395, + "GTAGGG": 3396, + "GTTAAA": 3397, + "GTTAAT": 3398, + "GTTAAC": 3399, + "GTTAAG": 3400, + "GTTATA": 3401, + "GTTATT": 3402, + "GTTATC": 3403, + "GTTATG": 3404, + "GTTACA": 3405, + "GTTACT": 3406, + "GTTACC": 3407, + "GTTACG": 3408, + "GTTAGA": 3409, + "GTTAGT": 3410, + "GTTAGC": 3411, + "GTTAGG": 3412, + "GTTTAA": 3413, + "GTTTAT": 3414, + "GTTTAC": 3415, + "GTTTAG": 3416, + "GTTTTA": 3417, + "GTTTTT": 3418, + "GTTTTC": 3419, + "GTTTTG": 3420, + "GTTTCA": 3421, + "GTTTCT": 3422, + "GTTTCC": 3423, + "GTTTCG": 3424, + "GTTTGA": 3425, + "GTTTGT": 3426, + "GTTTGC": 3427, + "GTTTGG": 3428, + "GTTCAA": 3429, + "GTTCAT": 3430, + "GTTCAC": 3431, + "GTTCAG": 3432, + "GTTCTA": 3433, + "GTTCTT": 3434, + "GTTCTC": 3435, + "GTTCTG": 3436, + "GTTCCA": 3437, + "GTTCCT": 3438, + "GTTCCC": 3439, + "GTTCCG": 3440, + "GTTCGA": 3441, + "GTTCGT": 3442, + "GTTCGC": 3443, + "GTTCGG": 3444, + "GTTGAA": 3445, + "GTTGAT": 3446, + "GTTGAC": 3447, + "GTTGAG": 3448, + "GTTGTA": 3449, + "GTTGTT": 3450, + "GTTGTC": 3451, + "GTTGTG": 3452, + "GTTGCA": 3453, + "GTTGCT": 3454, + "GTTGCC": 3455, + "GTTGCG": 3456, + "GTTGGA": 3457, + "GTTGGT": 3458, + "GTTGGC": 3459, + "GTTGGG": 3460, + "GTCAAA": 3461, + "GTCAAT": 3462, + "GTCAAC": 3463, + "GTCAAG": 3464, + "GTCATA": 3465, + "GTCATT": 3466, + "GTCATC": 3467, + "GTCATG": 3468, + "GTCACA": 3469, + "GTCACT": 3470, + "GTCACC": 3471, + "GTCACG": 3472, + "GTCAGA": 3473, + "GTCAGT": 3474, + "GTCAGC": 3475, + "GTCAGG": 3476, + "GTCTAA": 3477, + "GTCTAT": 3478, + "GTCTAC": 3479, + "GTCTAG": 3480, + "GTCTTA": 3481, + "GTCTTT": 3482, + "GTCTTC": 3483, + "GTCTTG": 3484, + "GTCTCA": 3485, + "GTCTCT": 3486, + "GTCTCC": 3487, + "GTCTCG": 3488, + "GTCTGA": 3489, + "GTCTGT": 3490, + "GTCTGC": 3491, + "GTCTGG": 3492, + "GTCCAA": 3493, + "GTCCAT": 3494, + "GTCCAC": 3495, + "GTCCAG": 3496, + "GTCCTA": 3497, + "GTCCTT": 3498, + "GTCCTC": 3499, + "GTCCTG": 3500, + "GTCCCA": 3501, + "GTCCCT": 3502, + "GTCCCC": 3503, + "GTCCCG": 3504, + "GTCCGA": 3505, + "GTCCGT": 3506, + "GTCCGC": 3507, + "GTCCGG": 3508, + "GTCGAA": 3509, + "GTCGAT": 3510, + "GTCGAC": 3511, + "GTCGAG": 3512, + "GTCGTA": 3513, + "GTCGTT": 3514, + "GTCGTC": 3515, + "GTCGTG": 3516, + "GTCGCA": 3517, + "GTCGCT": 3518, + "GTCGCC": 3519, + "GTCGCG": 3520, + "GTCGGA": 3521, + "GTCGGT": 3522, + "GTCGGC": 3523, + "GTCGGG": 3524, + "GTGAAA": 3525, + "GTGAAT": 3526, + "GTGAAC": 3527, + "GTGAAG": 3528, + "GTGATA": 3529, + "GTGATT": 3530, + "GTGATC": 3531, + "GTGATG": 3532, + "GTGACA": 3533, + "GTGACT": 3534, + "GTGACC": 3535, + "GTGACG": 3536, + "GTGAGA": 3537, + "GTGAGT": 3538, + "GTGAGC": 3539, + "GTGAGG": 3540, + "GTGTAA": 3541, + "GTGTAT": 3542, + "GTGTAC": 3543, + "GTGTAG": 3544, + "GTGTTA": 3545, + "GTGTTT": 3546, + "GTGTTC": 3547, + "GTGTTG": 3548, + "GTGTCA": 3549, + "GTGTCT": 3550, + "GTGTCC": 3551, + "GTGTCG": 3552, + "GTGTGA": 3553, + "GTGTGT": 3554, + "GTGTGC": 3555, + "GTGTGG": 3556, + "GTGCAA": 3557, + "GTGCAT": 3558, + "GTGCAC": 3559, + "GTGCAG": 3560, + "GTGCTA": 3561, + "GTGCTT": 3562, + "GTGCTC": 3563, + "GTGCTG": 3564, + "GTGCCA": 3565, + "GTGCCT": 3566, + "GTGCCC": 3567, + "GTGCCG": 3568, + "GTGCGA": 3569, + "GTGCGT": 3570, + "GTGCGC": 3571, + "GTGCGG": 3572, + "GTGGAA": 3573, + "GTGGAT": 3574, + "GTGGAC": 3575, + "GTGGAG": 3576, + "GTGGTA": 3577, + "GTGGTT": 3578, + "GTGGTC": 3579, + "GTGGTG": 3580, + "GTGGCA": 3581, + "GTGGCT": 3582, + "GTGGCC": 3583, + "GTGGCG": 3584, + "GTGGGA": 3585, + "GTGGGT": 3586, + "GTGGGC": 3587, + "GTGGGG": 3588, + "GCAAAA": 3589, + "GCAAAT": 3590, + "GCAAAC": 3591, + "GCAAAG": 3592, + "GCAATA": 3593, + "GCAATT": 3594, + "GCAATC": 3595, + "GCAATG": 3596, + "GCAACA": 3597, + "GCAACT": 3598, + "GCAACC": 3599, + "GCAACG": 3600, + "GCAAGA": 3601, + "GCAAGT": 3602, + "GCAAGC": 3603, + "GCAAGG": 3604, + "GCATAA": 3605, + "GCATAT": 3606, + "GCATAC": 3607, + "GCATAG": 3608, + "GCATTA": 3609, + "GCATTT": 3610, + "GCATTC": 3611, + "GCATTG": 3612, + "GCATCA": 3613, + "GCATCT": 3614, + "GCATCC": 3615, + "GCATCG": 3616, + "GCATGA": 3617, + "GCATGT": 3618, + "GCATGC": 3619, + "GCATGG": 3620, + "GCACAA": 3621, + "GCACAT": 3622, + "GCACAC": 3623, + "GCACAG": 3624, + "GCACTA": 3625, + "GCACTT": 3626, + "GCACTC": 3627, + "GCACTG": 3628, + "GCACCA": 3629, + "GCACCT": 3630, + "GCACCC": 3631, + "GCACCG": 3632, + "GCACGA": 3633, + "GCACGT": 3634, + "GCACGC": 3635, + "GCACGG": 3636, + "GCAGAA": 3637, + "GCAGAT": 3638, + "GCAGAC": 3639, + "GCAGAG": 3640, + "GCAGTA": 3641, + "GCAGTT": 3642, + "GCAGTC": 3643, + "GCAGTG": 3644, + "GCAGCA": 3645, + "GCAGCT": 3646, + "GCAGCC": 3647, + "GCAGCG": 3648, + "GCAGGA": 3649, + "GCAGGT": 3650, + "GCAGGC": 3651, + "GCAGGG": 3652, + "GCTAAA": 3653, + "GCTAAT": 3654, + "GCTAAC": 3655, + "GCTAAG": 3656, + "GCTATA": 3657, + "GCTATT": 3658, + "GCTATC": 3659, + "GCTATG": 3660, + "GCTACA": 3661, + "GCTACT": 3662, + "GCTACC": 3663, + "GCTACG": 3664, + "GCTAGA": 3665, + "GCTAGT": 3666, + "GCTAGC": 3667, + "GCTAGG": 3668, + "GCTTAA": 3669, + "GCTTAT": 3670, + "GCTTAC": 3671, + "GCTTAG": 3672, + "GCTTTA": 3673, + "GCTTTT": 3674, + "GCTTTC": 3675, + "GCTTTG": 3676, + "GCTTCA": 3677, + "GCTTCT": 3678, + "GCTTCC": 3679, + "GCTTCG": 3680, + "GCTTGA": 3681, + "GCTTGT": 3682, + "GCTTGC": 3683, + "GCTTGG": 3684, + "GCTCAA": 3685, + "GCTCAT": 3686, + "GCTCAC": 3687, + "GCTCAG": 3688, + "GCTCTA": 3689, + "GCTCTT": 3690, + "GCTCTC": 3691, + "GCTCTG": 3692, + "GCTCCA": 3693, + "GCTCCT": 3694, + "GCTCCC": 3695, + "GCTCCG": 3696, + "GCTCGA": 3697, + "GCTCGT": 3698, + "GCTCGC": 3699, + "GCTCGG": 3700, + "GCTGAA": 3701, + "GCTGAT": 3702, + "GCTGAC": 3703, + "GCTGAG": 3704, + "GCTGTA": 3705, + "GCTGTT": 3706, + "GCTGTC": 3707, + "GCTGTG": 3708, + "GCTGCA": 3709, + "GCTGCT": 3710, + "GCTGCC": 3711, + "GCTGCG": 3712, + "GCTGGA": 3713, + "GCTGGT": 3714, + "GCTGGC": 3715, + "GCTGGG": 3716, + "GCCAAA": 3717, + "GCCAAT": 3718, + "GCCAAC": 3719, + "GCCAAG": 3720, + "GCCATA": 3721, + "GCCATT": 3722, + "GCCATC": 3723, + "GCCATG": 3724, + "GCCACA": 3725, + "GCCACT": 3726, + "GCCACC": 3727, + "GCCACG": 3728, + "GCCAGA": 3729, + "GCCAGT": 3730, + "GCCAGC": 3731, + "GCCAGG": 3732, + "GCCTAA": 3733, + "GCCTAT": 3734, + "GCCTAC": 3735, + "GCCTAG": 3736, + "GCCTTA": 3737, + "GCCTTT": 3738, + "GCCTTC": 3739, + "GCCTTG": 3740, + "GCCTCA": 3741, + "GCCTCT": 3742, + "GCCTCC": 3743, + "GCCTCG": 3744, + "GCCTGA": 3745, + "GCCTGT": 3746, + "GCCTGC": 3747, + "GCCTGG": 3748, + "GCCCAA": 3749, + "GCCCAT": 3750, + "GCCCAC": 3751, + "GCCCAG": 3752, + "GCCCTA": 3753, + "GCCCTT": 3754, + "GCCCTC": 3755, + "GCCCTG": 3756, + "GCCCCA": 3757, + "GCCCCT": 3758, + "GCCCCC": 3759, + "GCCCCG": 3760, + "GCCCGA": 3761, + "GCCCGT": 3762, + "GCCCGC": 3763, + "GCCCGG": 3764, + "GCCGAA": 3765, + "GCCGAT": 3766, + "GCCGAC": 3767, + "GCCGAG": 3768, + "GCCGTA": 3769, + "GCCGTT": 3770, + "GCCGTC": 3771, + "GCCGTG": 3772, + "GCCGCA": 3773, + "GCCGCT": 3774, + "GCCGCC": 3775, + "GCCGCG": 3776, + "GCCGGA": 3777, + "GCCGGT": 3778, + "GCCGGC": 3779, + "GCCGGG": 3780, + "GCGAAA": 3781, + "GCGAAT": 3782, + "GCGAAC": 3783, + "GCGAAG": 3784, + "GCGATA": 3785, + "GCGATT": 3786, + "GCGATC": 3787, + "GCGATG": 3788, + "GCGACA": 3789, + "GCGACT": 3790, + "GCGACC": 3791, + "GCGACG": 3792, + "GCGAGA": 3793, + "GCGAGT": 3794, + "GCGAGC": 3795, + "GCGAGG": 3796, + "GCGTAA": 3797, + "GCGTAT": 3798, + "GCGTAC": 3799, + "GCGTAG": 3800, + "GCGTTA": 3801, + "GCGTTT": 3802, + "GCGTTC": 3803, + "GCGTTG": 3804, + "GCGTCA": 3805, + "GCGTCT": 3806, + "GCGTCC": 3807, + "GCGTCG": 3808, + "GCGTGA": 3809, + "GCGTGT": 3810, + "GCGTGC": 3811, + "GCGTGG": 3812, + "GCGCAA": 3813, + "GCGCAT": 3814, + "GCGCAC": 3815, + "GCGCAG": 3816, + "GCGCTA": 3817, + "GCGCTT": 3818, + "GCGCTC": 3819, + "GCGCTG": 3820, + "GCGCCA": 3821, + "GCGCCT": 3822, + "GCGCCC": 3823, + "GCGCCG": 3824, + "GCGCGA": 3825, + "GCGCGT": 3826, + "GCGCGC": 3827, + "GCGCGG": 3828, + "GCGGAA": 3829, + "GCGGAT": 3830, + "GCGGAC": 3831, + "GCGGAG": 3832, + "GCGGTA": 3833, + "GCGGTT": 3834, + "GCGGTC": 3835, + "GCGGTG": 3836, + "GCGGCA": 3837, + "GCGGCT": 3838, + "GCGGCC": 3839, + "GCGGCG": 3840, + "GCGGGA": 3841, + "GCGGGT": 3842, + "GCGGGC": 3843, + "GCGGGG": 3844, + "GGAAAA": 3845, + "GGAAAT": 3846, + "GGAAAC": 3847, + "GGAAAG": 3848, + "GGAATA": 3849, + "GGAATT": 3850, + "GGAATC": 3851, + "GGAATG": 3852, + "GGAACA": 3853, + "GGAACT": 3854, + "GGAACC": 3855, + "GGAACG": 3856, + "GGAAGA": 3857, + "GGAAGT": 3858, + "GGAAGC": 3859, + "GGAAGG": 3860, + "GGATAA": 3861, + "GGATAT": 3862, + "GGATAC": 3863, + "GGATAG": 3864, + "GGATTA": 3865, + "GGATTT": 3866, + "GGATTC": 3867, + "GGATTG": 3868, + "GGATCA": 3869, + "GGATCT": 3870, + "GGATCC": 3871, + "GGATCG": 3872, + "GGATGA": 3873, + "GGATGT": 3874, + "GGATGC": 3875, + "GGATGG": 3876, + "GGACAA": 3877, + "GGACAT": 3878, + "GGACAC": 3879, + "GGACAG": 3880, + "GGACTA": 3881, + "GGACTT": 3882, + "GGACTC": 3883, + "GGACTG": 3884, + "GGACCA": 3885, + "GGACCT": 3886, + "GGACCC": 3887, + "GGACCG": 3888, + "GGACGA": 3889, + "GGACGT": 3890, + "GGACGC": 3891, + "GGACGG": 3892, + "GGAGAA": 3893, + "GGAGAT": 3894, + "GGAGAC": 3895, + "GGAGAG": 3896, + "GGAGTA": 3897, + "GGAGTT": 3898, + "GGAGTC": 3899, + "GGAGTG": 3900, + "GGAGCA": 3901, + "GGAGCT": 3902, + "GGAGCC": 3903, + "GGAGCG": 3904, + "GGAGGA": 3905, + "GGAGGT": 3906, + "GGAGGC": 3907, + "GGAGGG": 3908, + "GGTAAA": 3909, + "GGTAAT": 3910, + "GGTAAC": 3911, + "GGTAAG": 3912, + "GGTATA": 3913, + "GGTATT": 3914, + "GGTATC": 3915, + "GGTATG": 3916, + "GGTACA": 3917, + "GGTACT": 3918, + "GGTACC": 3919, + "GGTACG": 3920, + "GGTAGA": 3921, + "GGTAGT": 3922, + "GGTAGC": 3923, + "GGTAGG": 3924, + "GGTTAA": 3925, + "GGTTAT": 3926, + "GGTTAC": 3927, + "GGTTAG": 3928, + "GGTTTA": 3929, + "GGTTTT": 3930, + "GGTTTC": 3931, + "GGTTTG": 3932, + "GGTTCA": 3933, + "GGTTCT": 3934, + "GGTTCC": 3935, + "GGTTCG": 3936, + "GGTTGA": 3937, + "GGTTGT": 3938, + "GGTTGC": 3939, + "GGTTGG": 3940, + "GGTCAA": 3941, + "GGTCAT": 3942, + "GGTCAC": 3943, + "GGTCAG": 3944, + "GGTCTA": 3945, + "GGTCTT": 3946, + "GGTCTC": 3947, + "GGTCTG": 3948, + "GGTCCA": 3949, + "GGTCCT": 3950, + "GGTCCC": 3951, + "GGTCCG": 3952, + "GGTCGA": 3953, + "GGTCGT": 3954, + "GGTCGC": 3955, + "GGTCGG": 3956, + "GGTGAA": 3957, + "GGTGAT": 3958, + "GGTGAC": 3959, + "GGTGAG": 3960, + "GGTGTA": 3961, + "GGTGTT": 3962, + "GGTGTC": 3963, + "GGTGTG": 3964, + "GGTGCA": 3965, + "GGTGCT": 3966, + "GGTGCC": 3967, + "GGTGCG": 3968, + "GGTGGA": 3969, + "GGTGGT": 3970, + "GGTGGC": 3971, + "GGTGGG": 3972, + "GGCAAA": 3973, + "GGCAAT": 3974, + "GGCAAC": 3975, + "GGCAAG": 3976, + "GGCATA": 3977, + "GGCATT": 3978, + "GGCATC": 3979, + "GGCATG": 3980, + "GGCACA": 3981, + "GGCACT": 3982, + "GGCACC": 3983, + "GGCACG": 3984, + "GGCAGA": 3985, + "GGCAGT": 3986, + "GGCAGC": 3987, + "GGCAGG": 3988, + "GGCTAA": 3989, + "GGCTAT": 3990, + "GGCTAC": 3991, + "GGCTAG": 3992, + "GGCTTA": 3993, + "GGCTTT": 3994, + "GGCTTC": 3995, + "GGCTTG": 3996, + "GGCTCA": 3997, + "GGCTCT": 3998, + "GGCTCC": 3999, + "GGCTCG": 4000, + "GGCTGA": 4001, + "GGCTGT": 4002, + "GGCTGC": 4003, + "GGCTGG": 4004, + "GGCCAA": 4005, + "GGCCAT": 4006, + "GGCCAC": 4007, + "GGCCAG": 4008, + "GGCCTA": 4009, + "GGCCTT": 4010, + "GGCCTC": 4011, + "GGCCTG": 4012, + "GGCCCA": 4013, + "GGCCCT": 4014, + "GGCCCC": 4015, + "GGCCCG": 4016, + "GGCCGA": 4017, + "GGCCGT": 4018, + "GGCCGC": 4019, + "GGCCGG": 4020, + "GGCGAA": 4021, + "GGCGAT": 4022, + "GGCGAC": 4023, + "GGCGAG": 4024, + "GGCGTA": 4025, + "GGCGTT": 4026, + "GGCGTC": 4027, + "GGCGTG": 4028, + "GGCGCA": 4029, + "GGCGCT": 4030, + "GGCGCC": 4031, + "GGCGCG": 4032, + "GGCGGA": 4033, + "GGCGGT": 4034, + "GGCGGC": 4035, + "GGCGGG": 4036, + "GGGAAA": 4037, + "GGGAAT": 4038, + "GGGAAC": 4039, + "GGGAAG": 4040, + "GGGATA": 4041, + "GGGATT": 4042, + "GGGATC": 4043, + "GGGATG": 4044, + "GGGACA": 4045, + "GGGACT": 4046, + "GGGACC": 4047, + "GGGACG": 4048, + "GGGAGA": 4049, + "GGGAGT": 4050, + "GGGAGC": 4051, + "GGGAGG": 4052, + "GGGTAA": 4053, + "GGGTAT": 4054, + "GGGTAC": 4055, + "GGGTAG": 4056, + "GGGTTA": 4057, + "GGGTTT": 4058, + "GGGTTC": 4059, + "GGGTTG": 4060, + "GGGTCA": 4061, + "GGGTCT": 4062, + "GGGTCC": 4063, + "GGGTCG": 4064, + "GGGTGA": 4065, + "GGGTGT": 4066, + "GGGTGC": 4067, + "GGGTGG": 4068, + "GGGCAA": 4069, + "GGGCAT": 4070, + "GGGCAC": 4071, + "GGGCAG": 4072, + "GGGCTA": 4073, + "GGGCTT": 4074, + "GGGCTC": 4075, + "GGGCTG": 4076, + "GGGCCA": 4077, + "GGGCCT": 4078, + "GGGCCC": 4079, + "GGGCCG": 4080, + "GGGCGA": 4081, + "GGGCGT": 4082, + "GGGCGC": 4083, + "GGGCGG": 4084, + "GGGGAA": 4085, + "GGGGAT": 4086, + "GGGGAC": 4087, + "GGGGAG": 4088, + "GGGGTA": 4089, + "GGGGTT": 4090, + "GGGGTC": 4091, + "GGGGTG": 4092, + "GGGGCA": 4093, + "GGGGCT": 4094, + "GGGGCC": 4095, + "GGGGCG": 4096, + "GGGGGA": 4097, + "GGGGGT": 4098, + "GGGGGC": 4099, + "GGGGGG": 4100 + } + } +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/tokenizer_config.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..604e2b1b29af1dcb3920d2828b6a90276e5e3d02 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/tokenizer_config.json @@ -0,0 +1,58 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "[UNK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "[CLS]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "[SEP]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "4": { + "content": "[MASK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "clean_up_tokenization_spaces": true, + "cls_token": "[CLS]", + "do_basic_tokenize": true, + "do_lower_case": false, + "mask_token": "[MASK]", + "max_len": 512, + "model_max_length": 512, + "never_split": null, + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "strip_accents": null, + "tokenize_chinese_chars": true, + "tokenizer_class": "BertTokenizer", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/trainer_state.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..d2b7976e20f2002926af4059e6e137b45d393f1a --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/trainer_state.json @@ -0,0 +1,1153 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 16.0, + "eval_steps": 500, + "global_step": 4000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.1, + "grad_norm": 0.2218121588230133, + "learning_rate": 5.319148936170213e-05, + "loss": 0.6927, + "step": 25 + }, + { + "epoch": 0.2, + "grad_norm": 0.44851255416870117, + "learning_rate": 0.00010638297872340425, + "loss": 0.6929, + "step": 50 + }, + { + "epoch": 0.3, + "grad_norm": 0.2440175861120224, + "learning_rate": 0.00015957446808510637, + "loss": 0.6924, + "step": 75 + }, + { + "epoch": 0.4, + "grad_norm": 0.3422006666660309, + "learning_rate": 0.0002127659574468085, + "loss": 0.6922, + "step": 100 + }, + { + "epoch": 0.5, + "grad_norm": 0.267333447933197, + "learning_rate": 0.00026595744680851064, + "loss": 0.6913, + "step": 125 + }, + { + "epoch": 0.6, + "grad_norm": 0.45699772238731384, + "learning_rate": 0.00031914893617021275, + "loss": 0.6896, + "step": 150 + }, + { + "epoch": 0.7, + "grad_norm": 0.3004947304725647, + "learning_rate": 0.0003723404255319149, + "loss": 0.6887, + "step": 175 + }, + { + "epoch": 0.8, + "grad_norm": 0.33440759778022766, + "learning_rate": 0.0003992081821181128, + "loss": 0.6857, + "step": 200 + }, + { + "epoch": 0.9, + "grad_norm": 0.2918928861618042, + "learning_rate": 0.0003975585615308479, + "loss": 0.6882, + "step": 225 + }, + { + "epoch": 1.0, + "grad_norm": 0.2934085726737976, + "learning_rate": 0.00039590894094358297, + "loss": 0.6807, + "step": 250 + }, + { + "epoch": 1.1, + "grad_norm": 0.3364830017089844, + "learning_rate": 0.0003942593203563181, + "loss": 0.6848, + "step": 275 + }, + { + "epoch": 1.2, + "grad_norm": 0.8379159569740295, + "learning_rate": 0.0003926096997690532, + "loss": 0.677, + "step": 300 + }, + { + "epoch": 1.3, + "grad_norm": 0.37671706080436707, + "learning_rate": 0.00039096007918178817, + "loss": 0.6744, + "step": 325 + }, + { + "epoch": 1.4, + "grad_norm": 0.34329330921173096, + "learning_rate": 0.0003893104585945233, + "loss": 0.6733, + "step": 350 + }, + { + "epoch": 1.5, + "grad_norm": 0.4286085367202759, + "learning_rate": 0.0003876608380072583, + "loss": 0.6772, + "step": 375 + }, + { + "epoch": 1.6, + "grad_norm": 0.3576306998729706, + "learning_rate": 0.00038601121741999343, + "loss": 0.6799, + "step": 400 + }, + { + "epoch": 1.7, + "grad_norm": 0.4636435806751251, + "learning_rate": 0.0003843615968327285, + "loss": 0.6819, + "step": 425 + }, + { + "epoch": 1.8, + "grad_norm": 0.4552135765552521, + "learning_rate": 0.0003827119762454636, + "loss": 0.6731, + "step": 450 + }, + { + "epoch": 1.9, + "grad_norm": 0.3565954566001892, + "learning_rate": 0.00038106235565819863, + "loss": 0.6718, + "step": 475 + }, + { + "epoch": 2.0, + "grad_norm": 0.5451686978340149, + "learning_rate": 0.0003794127350709337, + "loss": 0.6858, + "step": 500 + }, + { + "epoch": 2.1, + "grad_norm": 0.3730587959289551, + "learning_rate": 0.0003777631144836688, + "loss": 0.667, + "step": 525 + }, + { + "epoch": 2.2, + "grad_norm": 0.6642417907714844, + "learning_rate": 0.00037611349389640383, + "loss": 0.6662, + "step": 550 + }, + { + "epoch": 2.3, + "grad_norm": 0.3889809250831604, + "learning_rate": 0.00037446387330913894, + "loss": 0.6718, + "step": 575 + }, + { + "epoch": 2.4, + "grad_norm": 0.47203826904296875, + "learning_rate": 0.000372814252721874, + "loss": 0.6687, + "step": 600 + }, + { + "epoch": 2.5, + "grad_norm": 0.5293862819671631, + "learning_rate": 0.00037116463213460903, + "loss": 0.6655, + "step": 625 + }, + { + "epoch": 2.6, + "grad_norm": 0.36724531650543213, + "learning_rate": 0.00036951501154734414, + "loss": 0.672, + "step": 650 + }, + { + "epoch": 2.7, + "grad_norm": 0.5733087062835693, + "learning_rate": 0.0003678653909600792, + "loss": 0.6683, + "step": 675 + }, + { + "epoch": 2.8, + "grad_norm": 0.6385205984115601, + "learning_rate": 0.0003662157703728143, + "loss": 0.67, + "step": 700 + }, + { + "epoch": 2.9, + "grad_norm": 0.6146281957626343, + "learning_rate": 0.00036456614978554934, + "loss": 0.6606, + "step": 725 + }, + { + "epoch": 3.0, + "grad_norm": 0.6535729169845581, + "learning_rate": 0.00036291652919828444, + "loss": 0.6691, + "step": 750 + }, + { + "epoch": 3.1, + "grad_norm": 0.5194367170333862, + "learning_rate": 0.00036126690861101944, + "loss": 0.6517, + "step": 775 + }, + { + "epoch": 3.2, + "grad_norm": 0.7116692066192627, + "learning_rate": 0.00035961728802375454, + "loss": 0.6501, + "step": 800 + }, + { + "epoch": 3.3, + "grad_norm": 0.5338636040687561, + "learning_rate": 0.0003579676674364896, + "loss": 0.6682, + "step": 825 + }, + { + "epoch": 3.4, + "grad_norm": 0.6912877559661865, + "learning_rate": 0.0003563180468492247, + "loss": 0.6632, + "step": 850 + }, + { + "epoch": 3.5, + "grad_norm": 0.6925108432769775, + "learning_rate": 0.0003546684262619598, + "loss": 0.653, + "step": 875 + }, + { + "epoch": 3.6, + "grad_norm": 0.5360299944877625, + "learning_rate": 0.00035301880567469485, + "loss": 0.6613, + "step": 900 + }, + { + "epoch": 3.7, + "grad_norm": 0.6487969160079956, + "learning_rate": 0.0003513691850874299, + "loss": 0.6627, + "step": 925 + }, + { + "epoch": 3.8, + "grad_norm": 1.7260371446609497, + "learning_rate": 0.00034971956450016495, + "loss": 0.6436, + "step": 950 + }, + { + "epoch": 3.9, + "grad_norm": 0.5852354168891907, + "learning_rate": 0.00034806994391290005, + "loss": 0.6613, + "step": 975 + }, + { + "epoch": 4.0, + "grad_norm": 0.7802712917327881, + "learning_rate": 0.0003464203233256351, + "loss": 0.658, + "step": 1000 + }, + { + "epoch": 4.1, + "grad_norm": 0.7335640788078308, + "learning_rate": 0.0003447707027383702, + "loss": 0.6476, + "step": 1025 + }, + { + "epoch": 4.2, + "grad_norm": 0.9540443420410156, + "learning_rate": 0.00034312108215110525, + "loss": 0.6334, + "step": 1050 + }, + { + "epoch": 4.3, + "grad_norm": 1.3698925971984863, + "learning_rate": 0.0003414714615638403, + "loss": 0.6512, + "step": 1075 + }, + { + "epoch": 4.4, + "grad_norm": 0.8282333016395569, + "learning_rate": 0.0003398218409765754, + "loss": 0.6556, + "step": 1100 + }, + { + "epoch": 4.5, + "grad_norm": 1.4414807558059692, + "learning_rate": 0.00033817222038931045, + "loss": 0.6628, + "step": 1125 + }, + { + "epoch": 4.6, + "grad_norm": 0.5493957996368408, + "learning_rate": 0.00033652259980204556, + "loss": 0.6461, + "step": 1150 + }, + { + "epoch": 4.7, + "grad_norm": 1.2343393564224243, + "learning_rate": 0.0003348729792147806, + "loss": 0.6464, + "step": 1175 + }, + { + "epoch": 4.8, + "grad_norm": 0.7163398265838623, + "learning_rate": 0.0003332233586275157, + "loss": 0.6518, + "step": 1200 + }, + { + "epoch": 4.9, + "grad_norm": 0.7878938317298889, + "learning_rate": 0.00033157373804025076, + "loss": 0.6502, + "step": 1225 + }, + { + "epoch": 5.0, + "grad_norm": 0.952571451663971, + "learning_rate": 0.0003299241174529858, + "loss": 0.6463, + "step": 1250 + }, + { + "epoch": 5.1, + "grad_norm": 0.6535393595695496, + "learning_rate": 0.0003282744968657209, + "loss": 0.6361, + "step": 1275 + }, + { + "epoch": 5.2, + "grad_norm": 0.8239464163780212, + "learning_rate": 0.00032662487627845596, + "loss": 0.6264, + "step": 1300 + }, + { + "epoch": 5.3, + "grad_norm": 1.4500653743743896, + "learning_rate": 0.00032497525569119106, + "loss": 0.6544, + "step": 1325 + }, + { + "epoch": 5.4, + "grad_norm": 1.1515296697616577, + "learning_rate": 0.0003233256351039261, + "loss": 0.6629, + "step": 1350 + }, + { + "epoch": 5.5, + "grad_norm": 0.6899295449256897, + "learning_rate": 0.00032167601451666116, + "loss": 0.6453, + "step": 1375 + }, + { + "epoch": 5.6, + "grad_norm": 1.1782217025756836, + "learning_rate": 0.00032002639392939627, + "loss": 0.6362, + "step": 1400 + }, + { + "epoch": 5.7, + "grad_norm": 0.6971372961997986, + "learning_rate": 0.0003183767733421313, + "loss": 0.6593, + "step": 1425 + }, + { + "epoch": 5.8, + "grad_norm": 0.7293921709060669, + "learning_rate": 0.0003167271527548664, + "loss": 0.6318, + "step": 1450 + }, + { + "epoch": 5.9, + "grad_norm": 0.6860108375549316, + "learning_rate": 0.00031507753216760147, + "loss": 0.6402, + "step": 1475 + }, + { + "epoch": 6.0, + "grad_norm": 1.330344319343567, + "learning_rate": 0.00031342791158033657, + "loss": 0.6276, + "step": 1500 + }, + { + "epoch": 6.1, + "grad_norm": 0.7261562943458557, + "learning_rate": 0.00031177829099307157, + "loss": 0.6369, + "step": 1525 + }, + { + "epoch": 6.2, + "grad_norm": 1.1324542760849, + "learning_rate": 0.00031012867040580667, + "loss": 0.6228, + "step": 1550 + }, + { + "epoch": 6.3, + "grad_norm": 0.7002120614051819, + "learning_rate": 0.0003084790498185417, + "loss": 0.6389, + "step": 1575 + }, + { + "epoch": 6.4, + "grad_norm": 1.3143397569656372, + "learning_rate": 0.0003068294292312768, + "loss": 0.6362, + "step": 1600 + }, + { + "epoch": 6.5, + "grad_norm": 1.6375401020050049, + "learning_rate": 0.0003051798086440119, + "loss": 0.6348, + "step": 1625 + }, + { + "epoch": 6.6, + "grad_norm": 1.4343605041503906, + "learning_rate": 0.000303530188056747, + "loss": 0.6496, + "step": 1650 + }, + { + "epoch": 6.7, + "grad_norm": 0.815762996673584, + "learning_rate": 0.000301880567469482, + "loss": 0.6407, + "step": 1675 + }, + { + "epoch": 6.8, + "grad_norm": 1.068608045578003, + "learning_rate": 0.0003002309468822171, + "loss": 0.6417, + "step": 1700 + }, + { + "epoch": 6.9, + "grad_norm": 0.8775753974914551, + "learning_rate": 0.0002985813262949522, + "loss": 0.6281, + "step": 1725 + }, + { + "epoch": 7.0, + "grad_norm": 1.041817545890808, + "learning_rate": 0.00029693170570768723, + "loss": 0.6485, + "step": 1750 + }, + { + "epoch": 7.1, + "grad_norm": 0.7135840058326721, + "learning_rate": 0.00029528208512042233, + "loss": 0.6192, + "step": 1775 + }, + { + "epoch": 7.2, + "grad_norm": 1.5523930788040161, + "learning_rate": 0.0002936324645331574, + "loss": 0.636, + "step": 1800 + }, + { + "epoch": 7.3, + "grad_norm": 0.7677174210548401, + "learning_rate": 0.00029198284394589243, + "loss": 0.6213, + "step": 1825 + }, + { + "epoch": 7.4, + "grad_norm": 0.8086565732955933, + "learning_rate": 0.00029033322335862753, + "loss": 0.6292, + "step": 1850 + }, + { + "epoch": 7.5, + "grad_norm": 3.1128971576690674, + "learning_rate": 0.0002886836027713626, + "loss": 0.648, + "step": 1875 + }, + { + "epoch": 7.6, + "grad_norm": 0.9243882894515991, + "learning_rate": 0.0002870339821840977, + "loss": 0.6286, + "step": 1900 + }, + { + "epoch": 7.7, + "grad_norm": 1.8203461170196533, + "learning_rate": 0.00028538436159683273, + "loss": 0.6301, + "step": 1925 + }, + { + "epoch": 7.8, + "grad_norm": 1.0924643278121948, + "learning_rate": 0.00028373474100956784, + "loss": 0.6378, + "step": 1950 + }, + { + "epoch": 7.9, + "grad_norm": 1.0839996337890625, + "learning_rate": 0.0002820851204223029, + "loss": 0.6265, + "step": 1975 + }, + { + "epoch": 8.0, + "grad_norm": 0.7143226861953735, + "learning_rate": 0.00028043549983503794, + "loss": 0.649, + "step": 2000 + }, + { + "epoch": 8.1, + "grad_norm": 0.8778955936431885, + "learning_rate": 0.00027878587924777304, + "loss": 0.6411, + "step": 2025 + }, + { + "epoch": 8.2, + "grad_norm": 1.644904613494873, + "learning_rate": 0.0002771362586605081, + "loss": 0.6168, + "step": 2050 + }, + { + "epoch": 8.3, + "grad_norm": 0.7915159463882446, + "learning_rate": 0.0002754866380732432, + "loss": 0.6324, + "step": 2075 + }, + { + "epoch": 8.4, + "grad_norm": 1.1212801933288574, + "learning_rate": 0.00027383701748597824, + "loss": 0.6232, + "step": 2100 + }, + { + "epoch": 8.5, + "grad_norm": 0.7717731595039368, + "learning_rate": 0.0002721873968987133, + "loss": 0.624, + "step": 2125 + }, + { + "epoch": 8.6, + "grad_norm": 1.0577210187911987, + "learning_rate": 0.00027053777631144834, + "loss": 0.6296, + "step": 2150 + }, + { + "epoch": 8.7, + "grad_norm": 1.0531567335128784, + "learning_rate": 0.00026888815572418344, + "loss": 0.6345, + "step": 2175 + }, + { + "epoch": 8.8, + "grad_norm": 2.8983640670776367, + "learning_rate": 0.00026723853513691855, + "loss": 0.6142, + "step": 2200 + }, + { + "epoch": 8.9, + "grad_norm": 0.7773396968841553, + "learning_rate": 0.0002655889145496536, + "loss": 0.634, + "step": 2225 + }, + { + "epoch": 9.0, + "grad_norm": 1.0587151050567627, + "learning_rate": 0.0002639392939623887, + "loss": 0.6326, + "step": 2250 + }, + { + "epoch": 9.1, + "grad_norm": 0.8009796738624573, + "learning_rate": 0.0002622896733751237, + "loss": 0.6175, + "step": 2275 + }, + { + "epoch": 9.2, + "grad_norm": 1.2530399560928345, + "learning_rate": 0.0002606400527878588, + "loss": 0.6239, + "step": 2300 + }, + { + "epoch": 9.3, + "grad_norm": 0.7976368069648743, + "learning_rate": 0.00025899043220059385, + "loss": 0.6452, + "step": 2325 + }, + { + "epoch": 9.4, + "grad_norm": 1.455032229423523, + "learning_rate": 0.00025734081161332895, + "loss": 0.6351, + "step": 2350 + }, + { + "epoch": 9.5, + "grad_norm": 0.8916760087013245, + "learning_rate": 0.000255691191026064, + "loss": 0.5986, + "step": 2375 + }, + { + "epoch": 9.6, + "grad_norm": 1.104746699333191, + "learning_rate": 0.0002540415704387991, + "loss": 0.6104, + "step": 2400 + }, + { + "epoch": 9.7, + "grad_norm": 0.9449218511581421, + "learning_rate": 0.00025239194985153415, + "loss": 0.6262, + "step": 2425 + }, + { + "epoch": 9.8, + "grad_norm": 2.0286591053009033, + "learning_rate": 0.0002507423292642692, + "loss": 0.6124, + "step": 2450 + }, + { + "epoch": 9.9, + "grad_norm": 2.273597002029419, + "learning_rate": 0.0002490927086770043, + "loss": 0.6164, + "step": 2475 + }, + { + "epoch": 10.0, + "grad_norm": 0.8174235224723816, + "learning_rate": 0.00024744308808973936, + "loss": 0.6254, + "step": 2500 + }, + { + "epoch": 10.1, + "grad_norm": 1.0447920560836792, + "learning_rate": 0.00024579346750247446, + "loss": 0.6187, + "step": 2525 + }, + { + "epoch": 10.2, + "grad_norm": 1.0433655977249146, + "learning_rate": 0.0002441438469152095, + "loss": 0.6183, + "step": 2550 + }, + { + "epoch": 10.3, + "grad_norm": 1.6479028463363647, + "learning_rate": 0.00024249422632794456, + "loss": 0.6353, + "step": 2575 + }, + { + "epoch": 10.4, + "grad_norm": 0.8032572865486145, + "learning_rate": 0.00024084460574067963, + "loss": 0.6206, + "step": 2600 + }, + { + "epoch": 10.5, + "grad_norm": 1.0675058364868164, + "learning_rate": 0.0002391949851534147, + "loss": 0.6168, + "step": 2625 + }, + { + "epoch": 10.6, + "grad_norm": 1.2228609323501587, + "learning_rate": 0.0002375453645661498, + "loss": 0.6007, + "step": 2650 + }, + { + "epoch": 10.7, + "grad_norm": 1.1027027368545532, + "learning_rate": 0.00023589574397888486, + "loss": 0.6227, + "step": 2675 + }, + { + "epoch": 10.8, + "grad_norm": 1.2949633598327637, + "learning_rate": 0.00023424612339161997, + "loss": 0.6179, + "step": 2700 + }, + { + "epoch": 10.9, + "grad_norm": 1.235032320022583, + "learning_rate": 0.000232596502804355, + "loss": 0.6216, + "step": 2725 + }, + { + "epoch": 11.0, + "grad_norm": 1.0087182521820068, + "learning_rate": 0.00023094688221709007, + "loss": 0.6286, + "step": 2750 + }, + { + "epoch": 11.1, + "grad_norm": 1.301207423210144, + "learning_rate": 0.00022929726162982514, + "loss": 0.5953, + "step": 2775 + }, + { + "epoch": 11.2, + "grad_norm": 1.6798988580703735, + "learning_rate": 0.00022764764104256022, + "loss": 0.6331, + "step": 2800 + }, + { + "epoch": 11.3, + "grad_norm": 1.1364004611968994, + "learning_rate": 0.0002259980204552953, + "loss": 0.6047, + "step": 2825 + }, + { + "epoch": 11.4, + "grad_norm": 0.9963313341140747, + "learning_rate": 0.00022434839986803037, + "loss": 0.6192, + "step": 2850 + }, + { + "epoch": 11.5, + "grad_norm": 1.5672775506973267, + "learning_rate": 0.00022269877928076542, + "loss": 0.6063, + "step": 2875 + }, + { + "epoch": 11.6, + "grad_norm": 1.4239299297332764, + "learning_rate": 0.0002210491586935005, + "loss": 0.6058, + "step": 2900 + }, + { + "epoch": 11.7, + "grad_norm": 1.5440959930419922, + "learning_rate": 0.00021939953810623557, + "loss": 0.6239, + "step": 2925 + }, + { + "epoch": 11.8, + "grad_norm": 1.35140860080719, + "learning_rate": 0.00021774991751897065, + "loss": 0.6241, + "step": 2950 + }, + { + "epoch": 11.9, + "grad_norm": 1.9355136156082153, + "learning_rate": 0.00021610029693170573, + "loss": 0.6135, + "step": 2975 + }, + { + "epoch": 12.0, + "grad_norm": 1.134037733078003, + "learning_rate": 0.0002144506763444408, + "loss": 0.6142, + "step": 3000 + }, + { + "epoch": 12.1, + "grad_norm": 1.094366192817688, + "learning_rate": 0.00021280105575717582, + "loss": 0.6226, + "step": 3025 + }, + { + "epoch": 12.2, + "grad_norm": 1.1197398900985718, + "learning_rate": 0.00021115143516991093, + "loss": 0.6073, + "step": 3050 + }, + { + "epoch": 12.3, + "grad_norm": 1.240188717842102, + "learning_rate": 0.000209501814582646, + "loss": 0.6265, + "step": 3075 + }, + { + "epoch": 12.4, + "grad_norm": 0.9774200916290283, + "learning_rate": 0.00020785219399538108, + "loss": 0.6361, + "step": 3100 + }, + { + "epoch": 12.5, + "grad_norm": 0.902847409248352, + "learning_rate": 0.00020620257340811616, + "loss": 0.6129, + "step": 3125 + }, + { + "epoch": 12.6, + "grad_norm": 0.8658815622329712, + "learning_rate": 0.00020455295282085123, + "loss": 0.5953, + "step": 3150 + }, + { + "epoch": 12.7, + "grad_norm": 1.4172385931015015, + "learning_rate": 0.00020290333223358626, + "loss": 0.6158, + "step": 3175 + }, + { + "epoch": 12.8, + "grad_norm": 2.6393818855285645, + "learning_rate": 0.00020125371164632133, + "loss": 0.5978, + "step": 3200 + }, + { + "epoch": 12.9, + "grad_norm": 2.4407801628112793, + "learning_rate": 0.0001996040910590564, + "loss": 0.6178, + "step": 3225 + }, + { + "epoch": 13.0, + "grad_norm": 1.2797659635543823, + "learning_rate": 0.00019795447047179148, + "loss": 0.5849, + "step": 3250 + }, + { + "epoch": 13.1, + "grad_norm": 1.590224266052246, + "learning_rate": 0.0001963048498845266, + "loss": 0.5983, + "step": 3275 + }, + { + "epoch": 13.2, + "grad_norm": 1.155685544013977, + "learning_rate": 0.00019465522929726164, + "loss": 0.6301, + "step": 3300 + }, + { + "epoch": 13.3, + "grad_norm": 1.701331377029419, + "learning_rate": 0.00019300560870999671, + "loss": 0.5865, + "step": 3325 + }, + { + "epoch": 13.4, + "grad_norm": 1.2553105354309082, + "learning_rate": 0.0001913559881227318, + "loss": 0.6113, + "step": 3350 + }, + { + "epoch": 13.5, + "grad_norm": 1.495653510093689, + "learning_rate": 0.00018970636753546684, + "loss": 0.6001, + "step": 3375 + }, + { + "epoch": 13.6, + "grad_norm": 1.2800766229629517, + "learning_rate": 0.00018805674694820192, + "loss": 0.6013, + "step": 3400 + }, + { + "epoch": 13.7, + "grad_norm": 1.0035955905914307, + "learning_rate": 0.000186407126360937, + "loss": 0.621, + "step": 3425 + }, + { + "epoch": 13.8, + "grad_norm": 1.0515904426574707, + "learning_rate": 0.00018475750577367207, + "loss": 0.6297, + "step": 3450 + }, + { + "epoch": 13.9, + "grad_norm": 2.071884870529175, + "learning_rate": 0.00018310788518640715, + "loss": 0.6004, + "step": 3475 + }, + { + "epoch": 14.0, + "grad_norm": 1.4622681140899658, + "learning_rate": 0.00018145826459914222, + "loss": 0.6021, + "step": 3500 + }, + { + "epoch": 14.1, + "grad_norm": 1.1860116720199585, + "learning_rate": 0.00017980864401187727, + "loss": 0.602, + "step": 3525 + }, + { + "epoch": 14.2, + "grad_norm": 1.0958752632141113, + "learning_rate": 0.00017815902342461235, + "loss": 0.6084, + "step": 3550 + }, + { + "epoch": 14.3, + "grad_norm": 1.4344367980957031, + "learning_rate": 0.00017650940283734742, + "loss": 0.6332, + "step": 3575 + }, + { + "epoch": 14.4, + "grad_norm": 1.1095603704452515, + "learning_rate": 0.00017485978225008247, + "loss": 0.5945, + "step": 3600 + }, + { + "epoch": 14.5, + "grad_norm": 1.416272521018982, + "learning_rate": 0.00017321016166281755, + "loss": 0.5964, + "step": 3625 + }, + { + "epoch": 14.6, + "grad_norm": 0.9727759957313538, + "learning_rate": 0.00017156054107555263, + "loss": 0.6231, + "step": 3650 + }, + { + "epoch": 14.7, + "grad_norm": 0.9979771375656128, + "learning_rate": 0.0001699109204882877, + "loss": 0.5849, + "step": 3675 + }, + { + "epoch": 14.8, + "grad_norm": 1.2574740648269653, + "learning_rate": 0.00016826129990102278, + "loss": 0.6076, + "step": 3700 + }, + { + "epoch": 14.9, + "grad_norm": 1.1657873392105103, + "learning_rate": 0.00016661167931375785, + "loss": 0.6111, + "step": 3725 + }, + { + "epoch": 15.0, + "grad_norm": 1.767681360244751, + "learning_rate": 0.0001649620587264929, + "loss": 0.6025, + "step": 3750 + }, + { + "epoch": 15.1, + "grad_norm": 1.3131200075149536, + "learning_rate": 0.00016331243813922798, + "loss": 0.6039, + "step": 3775 + }, + { + "epoch": 15.2, + "grad_norm": 1.1376745700836182, + "learning_rate": 0.00016166281755196306, + "loss": 0.5995, + "step": 3800 + }, + { + "epoch": 15.3, + "grad_norm": 1.2298001050949097, + "learning_rate": 0.00016001319696469813, + "loss": 0.602, + "step": 3825 + }, + { + "epoch": 15.4, + "grad_norm": 2.735769271850586, + "learning_rate": 0.0001583635763774332, + "loss": 0.5902, + "step": 3850 + }, + { + "epoch": 15.5, + "grad_norm": 1.3570619821548462, + "learning_rate": 0.00015671395579016829, + "loss": 0.5963, + "step": 3875 + }, + { + "epoch": 15.6, + "grad_norm": 1.6737253665924072, + "learning_rate": 0.00015506433520290334, + "loss": 0.6002, + "step": 3900 + }, + { + "epoch": 15.7, + "grad_norm": 1.386838674545288, + "learning_rate": 0.0001534147146156384, + "loss": 0.5905, + "step": 3925 + }, + { + "epoch": 15.8, + "grad_norm": 1.670160174369812, + "learning_rate": 0.0001517650940283735, + "loss": 0.6, + "step": 3950 + }, + { + "epoch": 15.9, + "grad_norm": 1.1611871719360352, + "learning_rate": 0.00015011547344110854, + "loss": 0.616, + "step": 3975 + }, + { + "epoch": 16.0, + "grad_norm": 0.9084995985031128, + "learning_rate": 0.00014846585285384361, + "loss": 0.6036, + "step": 4000 + } + ], + "logging_steps": 25, + "max_steps": 6250, + "num_input_tokens_seen": 0, + "num_train_epochs": 25, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 221524131840000.0, + "train_batch_size": 20, + "trial_name": null, + "trial_params": null +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/training_args.bin b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..540db3fbee786996d9705d768d31ba6c371187be --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dc7816b5f3fd38f01f9ba38a197ce070cde06e4c90e5260cb4814f34f628907d +size 5304 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/vocab.txt b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/vocab.txt new file mode 100644 index 0000000000000000000000000000000000000000..cc9ad53732e21ea9180f18e1fb67a38cc5702d4c --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-4000/vocab.txt @@ -0,0 +1,4101 @@ +[PAD] +[UNK] +[CLS] +[SEP] +[MASK] +AAAAAA +AAAAAT +AAAAAC +AAAAAG +AAAATA +AAAATT +AAAATC +AAAATG +AAAACA +AAAACT +AAAACC +AAAACG +AAAAGA +AAAAGT +AAAAGC +AAAAGG +AAATAA +AAATAT +AAATAC +AAATAG +AAATTA +AAATTT +AAATTC +AAATTG +AAATCA +AAATCT +AAATCC +AAATCG +AAATGA +AAATGT +AAATGC +AAATGG +AAACAA +AAACAT +AAACAC +AAACAG +AAACTA +AAACTT +AAACTC +AAACTG +AAACCA +AAACCT +AAACCC +AAACCG +AAACGA +AAACGT +AAACGC +AAACGG +AAAGAA +AAAGAT +AAAGAC +AAAGAG +AAAGTA +AAAGTT +AAAGTC +AAAGTG +AAAGCA +AAAGCT +AAAGCC +AAAGCG +AAAGGA +AAAGGT +AAAGGC +AAAGGG +AATAAA +AATAAT +AATAAC +AATAAG +AATATA +AATATT +AATATC +AATATG +AATACA +AATACT +AATACC +AATACG +AATAGA +AATAGT +AATAGC +AATAGG +AATTAA +AATTAT +AATTAC +AATTAG +AATTTA +AATTTT +AATTTC +AATTTG +AATTCA +AATTCT +AATTCC +AATTCG +AATTGA +AATTGT +AATTGC +AATTGG +AATCAA +AATCAT +AATCAC +AATCAG +AATCTA +AATCTT +AATCTC +AATCTG +AATCCA +AATCCT +AATCCC +AATCCG +AATCGA +AATCGT +AATCGC +AATCGG +AATGAA +AATGAT +AATGAC +AATGAG +AATGTA +AATGTT +AATGTC +AATGTG +AATGCA +AATGCT +AATGCC +AATGCG +AATGGA +AATGGT +AATGGC +AATGGG +AACAAA +AACAAT +AACAAC +AACAAG +AACATA +AACATT +AACATC +AACATG +AACACA +AACACT +AACACC +AACACG +AACAGA +AACAGT +AACAGC +AACAGG +AACTAA +AACTAT +AACTAC +AACTAG +AACTTA +AACTTT +AACTTC +AACTTG +AACTCA +AACTCT +AACTCC +AACTCG +AACTGA +AACTGT +AACTGC +AACTGG +AACCAA +AACCAT +AACCAC +AACCAG +AACCTA +AACCTT +AACCTC +AACCTG +AACCCA +AACCCT +AACCCC +AACCCG +AACCGA +AACCGT +AACCGC +AACCGG +AACGAA +AACGAT +AACGAC +AACGAG +AACGTA +AACGTT +AACGTC +AACGTG +AACGCA +AACGCT +AACGCC +AACGCG +AACGGA +AACGGT +AACGGC +AACGGG +AAGAAA +AAGAAT +AAGAAC +AAGAAG +AAGATA +AAGATT +AAGATC +AAGATG +AAGACA +AAGACT +AAGACC +AAGACG +AAGAGA +AAGAGT +AAGAGC +AAGAGG +AAGTAA +AAGTAT +AAGTAC +AAGTAG +AAGTTA +AAGTTT +AAGTTC +AAGTTG +AAGTCA +AAGTCT +AAGTCC +AAGTCG +AAGTGA +AAGTGT +AAGTGC +AAGTGG +AAGCAA +AAGCAT +AAGCAC +AAGCAG +AAGCTA +AAGCTT +AAGCTC +AAGCTG +AAGCCA +AAGCCT +AAGCCC +AAGCCG +AAGCGA +AAGCGT +AAGCGC +AAGCGG +AAGGAA +AAGGAT +AAGGAC +AAGGAG +AAGGTA +AAGGTT +AAGGTC +AAGGTG +AAGGCA +AAGGCT +AAGGCC +AAGGCG +AAGGGA +AAGGGT +AAGGGC +AAGGGG +ATAAAA +ATAAAT +ATAAAC +ATAAAG +ATAATA +ATAATT +ATAATC +ATAATG +ATAACA +ATAACT +ATAACC +ATAACG +ATAAGA +ATAAGT +ATAAGC +ATAAGG +ATATAA +ATATAT +ATATAC +ATATAG +ATATTA +ATATTT +ATATTC +ATATTG +ATATCA +ATATCT +ATATCC +ATATCG +ATATGA +ATATGT +ATATGC +ATATGG +ATACAA +ATACAT +ATACAC +ATACAG +ATACTA +ATACTT +ATACTC +ATACTG +ATACCA +ATACCT +ATACCC +ATACCG +ATACGA +ATACGT +ATACGC +ATACGG +ATAGAA +ATAGAT +ATAGAC +ATAGAG +ATAGTA +ATAGTT +ATAGTC +ATAGTG +ATAGCA +ATAGCT +ATAGCC +ATAGCG +ATAGGA +ATAGGT +ATAGGC +ATAGGG +ATTAAA +ATTAAT +ATTAAC +ATTAAG +ATTATA +ATTATT +ATTATC +ATTATG +ATTACA +ATTACT +ATTACC +ATTACG +ATTAGA +ATTAGT +ATTAGC +ATTAGG +ATTTAA +ATTTAT +ATTTAC +ATTTAG +ATTTTA +ATTTTT +ATTTTC +ATTTTG +ATTTCA +ATTTCT +ATTTCC +ATTTCG +ATTTGA +ATTTGT +ATTTGC +ATTTGG +ATTCAA +ATTCAT +ATTCAC +ATTCAG +ATTCTA +ATTCTT +ATTCTC +ATTCTG +ATTCCA +ATTCCT +ATTCCC +ATTCCG +ATTCGA +ATTCGT +ATTCGC +ATTCGG +ATTGAA +ATTGAT +ATTGAC +ATTGAG +ATTGTA +ATTGTT +ATTGTC +ATTGTG +ATTGCA +ATTGCT +ATTGCC +ATTGCG +ATTGGA +ATTGGT +ATTGGC +ATTGGG +ATCAAA +ATCAAT +ATCAAC +ATCAAG +ATCATA +ATCATT +ATCATC +ATCATG +ATCACA +ATCACT +ATCACC +ATCACG +ATCAGA +ATCAGT +ATCAGC +ATCAGG +ATCTAA +ATCTAT +ATCTAC +ATCTAG +ATCTTA +ATCTTT +ATCTTC +ATCTTG +ATCTCA +ATCTCT +ATCTCC +ATCTCG +ATCTGA +ATCTGT +ATCTGC +ATCTGG +ATCCAA +ATCCAT +ATCCAC +ATCCAG +ATCCTA +ATCCTT +ATCCTC +ATCCTG +ATCCCA +ATCCCT +ATCCCC +ATCCCG +ATCCGA +ATCCGT +ATCCGC +ATCCGG +ATCGAA +ATCGAT +ATCGAC +ATCGAG +ATCGTA +ATCGTT +ATCGTC +ATCGTG +ATCGCA +ATCGCT +ATCGCC +ATCGCG +ATCGGA +ATCGGT +ATCGGC +ATCGGG +ATGAAA +ATGAAT +ATGAAC +ATGAAG +ATGATA +ATGATT +ATGATC +ATGATG +ATGACA +ATGACT +ATGACC +ATGACG +ATGAGA +ATGAGT +ATGAGC +ATGAGG +ATGTAA +ATGTAT +ATGTAC +ATGTAG +ATGTTA +ATGTTT +ATGTTC +ATGTTG +ATGTCA +ATGTCT +ATGTCC +ATGTCG +ATGTGA +ATGTGT +ATGTGC +ATGTGG +ATGCAA +ATGCAT +ATGCAC +ATGCAG +ATGCTA +ATGCTT +ATGCTC +ATGCTG +ATGCCA +ATGCCT +ATGCCC +ATGCCG +ATGCGA +ATGCGT +ATGCGC +ATGCGG +ATGGAA +ATGGAT +ATGGAC +ATGGAG +ATGGTA +ATGGTT +ATGGTC +ATGGTG +ATGGCA +ATGGCT +ATGGCC +ATGGCG +ATGGGA +ATGGGT +ATGGGC +ATGGGG +ACAAAA +ACAAAT +ACAAAC +ACAAAG +ACAATA +ACAATT +ACAATC +ACAATG +ACAACA +ACAACT +ACAACC +ACAACG +ACAAGA +ACAAGT +ACAAGC +ACAAGG +ACATAA +ACATAT +ACATAC +ACATAG +ACATTA +ACATTT +ACATTC +ACATTG +ACATCA +ACATCT +ACATCC +ACATCG +ACATGA +ACATGT +ACATGC +ACATGG +ACACAA +ACACAT +ACACAC +ACACAG +ACACTA +ACACTT +ACACTC +ACACTG +ACACCA +ACACCT +ACACCC +ACACCG +ACACGA +ACACGT +ACACGC +ACACGG +ACAGAA +ACAGAT +ACAGAC +ACAGAG +ACAGTA +ACAGTT +ACAGTC +ACAGTG +ACAGCA +ACAGCT +ACAGCC +ACAGCG +ACAGGA +ACAGGT +ACAGGC +ACAGGG +ACTAAA +ACTAAT +ACTAAC +ACTAAG +ACTATA +ACTATT +ACTATC +ACTATG +ACTACA +ACTACT +ACTACC +ACTACG +ACTAGA +ACTAGT +ACTAGC +ACTAGG +ACTTAA +ACTTAT +ACTTAC +ACTTAG +ACTTTA +ACTTTT +ACTTTC +ACTTTG +ACTTCA +ACTTCT +ACTTCC +ACTTCG +ACTTGA +ACTTGT +ACTTGC +ACTTGG +ACTCAA +ACTCAT +ACTCAC +ACTCAG +ACTCTA +ACTCTT +ACTCTC +ACTCTG +ACTCCA +ACTCCT +ACTCCC +ACTCCG +ACTCGA +ACTCGT +ACTCGC +ACTCGG +ACTGAA +ACTGAT +ACTGAC +ACTGAG +ACTGTA +ACTGTT +ACTGTC +ACTGTG +ACTGCA +ACTGCT +ACTGCC +ACTGCG +ACTGGA +ACTGGT +ACTGGC +ACTGGG +ACCAAA +ACCAAT +ACCAAC +ACCAAG +ACCATA +ACCATT +ACCATC +ACCATG +ACCACA +ACCACT +ACCACC +ACCACG +ACCAGA +ACCAGT +ACCAGC +ACCAGG +ACCTAA +ACCTAT +ACCTAC +ACCTAG +ACCTTA +ACCTTT +ACCTTC +ACCTTG +ACCTCA +ACCTCT +ACCTCC +ACCTCG +ACCTGA +ACCTGT +ACCTGC +ACCTGG +ACCCAA +ACCCAT +ACCCAC +ACCCAG +ACCCTA +ACCCTT +ACCCTC +ACCCTG +ACCCCA +ACCCCT +ACCCCC +ACCCCG +ACCCGA +ACCCGT +ACCCGC +ACCCGG +ACCGAA +ACCGAT +ACCGAC +ACCGAG +ACCGTA +ACCGTT +ACCGTC +ACCGTG +ACCGCA +ACCGCT +ACCGCC +ACCGCG +ACCGGA +ACCGGT +ACCGGC +ACCGGG +ACGAAA +ACGAAT +ACGAAC +ACGAAG +ACGATA +ACGATT +ACGATC +ACGATG +ACGACA +ACGACT +ACGACC +ACGACG +ACGAGA +ACGAGT +ACGAGC +ACGAGG +ACGTAA +ACGTAT +ACGTAC +ACGTAG +ACGTTA +ACGTTT +ACGTTC +ACGTTG +ACGTCA +ACGTCT +ACGTCC +ACGTCG +ACGTGA +ACGTGT +ACGTGC +ACGTGG +ACGCAA +ACGCAT +ACGCAC +ACGCAG +ACGCTA +ACGCTT +ACGCTC +ACGCTG +ACGCCA +ACGCCT +ACGCCC +ACGCCG +ACGCGA +ACGCGT +ACGCGC +ACGCGG +ACGGAA +ACGGAT +ACGGAC +ACGGAG +ACGGTA +ACGGTT +ACGGTC +ACGGTG +ACGGCA +ACGGCT +ACGGCC +ACGGCG +ACGGGA +ACGGGT +ACGGGC +ACGGGG +AGAAAA +AGAAAT +AGAAAC +AGAAAG +AGAATA +AGAATT +AGAATC +AGAATG +AGAACA +AGAACT +AGAACC +AGAACG +AGAAGA +AGAAGT +AGAAGC +AGAAGG +AGATAA +AGATAT +AGATAC +AGATAG +AGATTA +AGATTT +AGATTC +AGATTG +AGATCA +AGATCT +AGATCC +AGATCG +AGATGA +AGATGT +AGATGC +AGATGG +AGACAA +AGACAT +AGACAC +AGACAG +AGACTA +AGACTT +AGACTC +AGACTG +AGACCA +AGACCT +AGACCC +AGACCG +AGACGA +AGACGT +AGACGC +AGACGG +AGAGAA +AGAGAT +AGAGAC +AGAGAG +AGAGTA +AGAGTT +AGAGTC +AGAGTG +AGAGCA +AGAGCT +AGAGCC +AGAGCG +AGAGGA +AGAGGT +AGAGGC +AGAGGG +AGTAAA +AGTAAT +AGTAAC +AGTAAG +AGTATA +AGTATT +AGTATC +AGTATG +AGTACA +AGTACT +AGTACC +AGTACG +AGTAGA +AGTAGT +AGTAGC +AGTAGG +AGTTAA +AGTTAT +AGTTAC +AGTTAG +AGTTTA +AGTTTT +AGTTTC +AGTTTG +AGTTCA +AGTTCT +AGTTCC +AGTTCG +AGTTGA +AGTTGT +AGTTGC +AGTTGG +AGTCAA +AGTCAT +AGTCAC +AGTCAG +AGTCTA +AGTCTT +AGTCTC +AGTCTG +AGTCCA +AGTCCT +AGTCCC +AGTCCG +AGTCGA +AGTCGT +AGTCGC +AGTCGG +AGTGAA +AGTGAT +AGTGAC +AGTGAG +AGTGTA +AGTGTT +AGTGTC +AGTGTG +AGTGCA +AGTGCT +AGTGCC +AGTGCG +AGTGGA +AGTGGT +AGTGGC +AGTGGG +AGCAAA +AGCAAT +AGCAAC +AGCAAG +AGCATA +AGCATT +AGCATC +AGCATG +AGCACA +AGCACT +AGCACC +AGCACG +AGCAGA +AGCAGT +AGCAGC +AGCAGG +AGCTAA +AGCTAT +AGCTAC +AGCTAG +AGCTTA +AGCTTT +AGCTTC +AGCTTG +AGCTCA +AGCTCT +AGCTCC +AGCTCG +AGCTGA +AGCTGT +AGCTGC +AGCTGG +AGCCAA +AGCCAT +AGCCAC +AGCCAG +AGCCTA +AGCCTT +AGCCTC +AGCCTG +AGCCCA +AGCCCT +AGCCCC +AGCCCG +AGCCGA +AGCCGT +AGCCGC +AGCCGG +AGCGAA +AGCGAT +AGCGAC +AGCGAG +AGCGTA +AGCGTT +AGCGTC +AGCGTG +AGCGCA +AGCGCT +AGCGCC +AGCGCG +AGCGGA +AGCGGT +AGCGGC +AGCGGG +AGGAAA +AGGAAT +AGGAAC +AGGAAG +AGGATA +AGGATT +AGGATC +AGGATG +AGGACA +AGGACT +AGGACC +AGGACG +AGGAGA +AGGAGT +AGGAGC +AGGAGG +AGGTAA +AGGTAT +AGGTAC +AGGTAG +AGGTTA +AGGTTT +AGGTTC +AGGTTG +AGGTCA +AGGTCT +AGGTCC +AGGTCG +AGGTGA +AGGTGT +AGGTGC +AGGTGG +AGGCAA +AGGCAT +AGGCAC +AGGCAG +AGGCTA +AGGCTT +AGGCTC +AGGCTG +AGGCCA +AGGCCT +AGGCCC +AGGCCG +AGGCGA +AGGCGT +AGGCGC +AGGCGG +AGGGAA +AGGGAT +AGGGAC +AGGGAG +AGGGTA +AGGGTT +AGGGTC +AGGGTG +AGGGCA +AGGGCT +AGGGCC +AGGGCG +AGGGGA +AGGGGT +AGGGGC +AGGGGG +TAAAAA +TAAAAT +TAAAAC +TAAAAG +TAAATA +TAAATT +TAAATC +TAAATG +TAAACA +TAAACT +TAAACC +TAAACG +TAAAGA +TAAAGT +TAAAGC +TAAAGG +TAATAA +TAATAT +TAATAC +TAATAG +TAATTA +TAATTT +TAATTC +TAATTG +TAATCA +TAATCT +TAATCC +TAATCG +TAATGA +TAATGT +TAATGC +TAATGG +TAACAA +TAACAT +TAACAC +TAACAG +TAACTA +TAACTT +TAACTC +TAACTG +TAACCA +TAACCT +TAACCC +TAACCG +TAACGA +TAACGT +TAACGC +TAACGG +TAAGAA +TAAGAT +TAAGAC +TAAGAG +TAAGTA +TAAGTT +TAAGTC +TAAGTG +TAAGCA +TAAGCT +TAAGCC +TAAGCG +TAAGGA +TAAGGT +TAAGGC +TAAGGG +TATAAA +TATAAT +TATAAC +TATAAG +TATATA +TATATT +TATATC +TATATG +TATACA +TATACT +TATACC +TATACG +TATAGA +TATAGT +TATAGC +TATAGG +TATTAA +TATTAT +TATTAC +TATTAG +TATTTA +TATTTT +TATTTC +TATTTG +TATTCA +TATTCT +TATTCC +TATTCG +TATTGA +TATTGT +TATTGC +TATTGG +TATCAA +TATCAT +TATCAC +TATCAG +TATCTA +TATCTT +TATCTC +TATCTG +TATCCA +TATCCT +TATCCC +TATCCG +TATCGA +TATCGT +TATCGC +TATCGG +TATGAA +TATGAT +TATGAC +TATGAG +TATGTA +TATGTT +TATGTC +TATGTG +TATGCA +TATGCT +TATGCC +TATGCG +TATGGA +TATGGT +TATGGC +TATGGG +TACAAA +TACAAT +TACAAC +TACAAG +TACATA +TACATT +TACATC +TACATG +TACACA +TACACT +TACACC +TACACG +TACAGA +TACAGT +TACAGC +TACAGG +TACTAA +TACTAT +TACTAC +TACTAG +TACTTA +TACTTT +TACTTC +TACTTG +TACTCA +TACTCT +TACTCC +TACTCG +TACTGA +TACTGT +TACTGC +TACTGG +TACCAA +TACCAT +TACCAC +TACCAG +TACCTA +TACCTT +TACCTC +TACCTG +TACCCA +TACCCT +TACCCC +TACCCG +TACCGA +TACCGT +TACCGC +TACCGG +TACGAA +TACGAT +TACGAC +TACGAG +TACGTA +TACGTT +TACGTC +TACGTG +TACGCA +TACGCT +TACGCC +TACGCG +TACGGA +TACGGT +TACGGC +TACGGG +TAGAAA +TAGAAT +TAGAAC +TAGAAG +TAGATA +TAGATT +TAGATC +TAGATG +TAGACA +TAGACT +TAGACC +TAGACG +TAGAGA +TAGAGT +TAGAGC +TAGAGG +TAGTAA +TAGTAT +TAGTAC +TAGTAG +TAGTTA +TAGTTT +TAGTTC +TAGTTG +TAGTCA +TAGTCT +TAGTCC +TAGTCG +TAGTGA +TAGTGT +TAGTGC +TAGTGG +TAGCAA +TAGCAT +TAGCAC +TAGCAG +TAGCTA +TAGCTT +TAGCTC +TAGCTG +TAGCCA +TAGCCT +TAGCCC +TAGCCG +TAGCGA +TAGCGT +TAGCGC +TAGCGG +TAGGAA +TAGGAT +TAGGAC +TAGGAG +TAGGTA +TAGGTT +TAGGTC +TAGGTG +TAGGCA +TAGGCT +TAGGCC +TAGGCG +TAGGGA +TAGGGT +TAGGGC +TAGGGG +TTAAAA +TTAAAT +TTAAAC +TTAAAG +TTAATA +TTAATT +TTAATC +TTAATG +TTAACA +TTAACT +TTAACC +TTAACG +TTAAGA +TTAAGT +TTAAGC +TTAAGG +TTATAA +TTATAT +TTATAC +TTATAG +TTATTA +TTATTT +TTATTC +TTATTG +TTATCA +TTATCT +TTATCC +TTATCG +TTATGA +TTATGT +TTATGC +TTATGG +TTACAA +TTACAT +TTACAC +TTACAG +TTACTA +TTACTT +TTACTC +TTACTG +TTACCA +TTACCT +TTACCC +TTACCG +TTACGA +TTACGT +TTACGC +TTACGG +TTAGAA +TTAGAT +TTAGAC +TTAGAG +TTAGTA +TTAGTT +TTAGTC +TTAGTG +TTAGCA +TTAGCT +TTAGCC +TTAGCG +TTAGGA +TTAGGT +TTAGGC +TTAGGG +TTTAAA +TTTAAT +TTTAAC +TTTAAG +TTTATA +TTTATT +TTTATC +TTTATG +TTTACA +TTTACT +TTTACC +TTTACG +TTTAGA +TTTAGT +TTTAGC +TTTAGG +TTTTAA +TTTTAT +TTTTAC +TTTTAG +TTTTTA +TTTTTT +TTTTTC +TTTTTG +TTTTCA +TTTTCT +TTTTCC +TTTTCG +TTTTGA +TTTTGT +TTTTGC +TTTTGG +TTTCAA +TTTCAT +TTTCAC +TTTCAG +TTTCTA +TTTCTT +TTTCTC +TTTCTG +TTTCCA +TTTCCT +TTTCCC +TTTCCG +TTTCGA +TTTCGT +TTTCGC +TTTCGG +TTTGAA +TTTGAT +TTTGAC +TTTGAG +TTTGTA +TTTGTT +TTTGTC +TTTGTG +TTTGCA +TTTGCT +TTTGCC +TTTGCG +TTTGGA +TTTGGT +TTTGGC +TTTGGG +TTCAAA +TTCAAT +TTCAAC +TTCAAG +TTCATA +TTCATT +TTCATC +TTCATG +TTCACA +TTCACT +TTCACC +TTCACG +TTCAGA +TTCAGT +TTCAGC +TTCAGG +TTCTAA +TTCTAT +TTCTAC +TTCTAG +TTCTTA +TTCTTT +TTCTTC +TTCTTG +TTCTCA +TTCTCT +TTCTCC +TTCTCG +TTCTGA +TTCTGT +TTCTGC +TTCTGG +TTCCAA +TTCCAT +TTCCAC +TTCCAG +TTCCTA +TTCCTT +TTCCTC +TTCCTG +TTCCCA +TTCCCT +TTCCCC +TTCCCG +TTCCGA +TTCCGT +TTCCGC +TTCCGG +TTCGAA +TTCGAT +TTCGAC +TTCGAG +TTCGTA +TTCGTT +TTCGTC +TTCGTG +TTCGCA +TTCGCT +TTCGCC +TTCGCG +TTCGGA +TTCGGT +TTCGGC +TTCGGG +TTGAAA +TTGAAT +TTGAAC +TTGAAG +TTGATA +TTGATT +TTGATC +TTGATG +TTGACA +TTGACT +TTGACC +TTGACG +TTGAGA +TTGAGT +TTGAGC +TTGAGG +TTGTAA +TTGTAT +TTGTAC +TTGTAG +TTGTTA +TTGTTT +TTGTTC +TTGTTG +TTGTCA +TTGTCT +TTGTCC +TTGTCG +TTGTGA +TTGTGT +TTGTGC +TTGTGG +TTGCAA +TTGCAT +TTGCAC +TTGCAG +TTGCTA +TTGCTT +TTGCTC +TTGCTG +TTGCCA +TTGCCT +TTGCCC +TTGCCG +TTGCGA +TTGCGT +TTGCGC +TTGCGG +TTGGAA +TTGGAT +TTGGAC +TTGGAG +TTGGTA +TTGGTT +TTGGTC +TTGGTG +TTGGCA +TTGGCT +TTGGCC +TTGGCG +TTGGGA +TTGGGT +TTGGGC +TTGGGG +TCAAAA +TCAAAT +TCAAAC +TCAAAG +TCAATA +TCAATT +TCAATC +TCAATG +TCAACA +TCAACT +TCAACC +TCAACG +TCAAGA +TCAAGT +TCAAGC +TCAAGG +TCATAA +TCATAT +TCATAC +TCATAG +TCATTA +TCATTT +TCATTC +TCATTG +TCATCA +TCATCT +TCATCC +TCATCG +TCATGA +TCATGT +TCATGC +TCATGG +TCACAA +TCACAT +TCACAC +TCACAG +TCACTA +TCACTT +TCACTC +TCACTG +TCACCA +TCACCT +TCACCC +TCACCG +TCACGA +TCACGT +TCACGC +TCACGG +TCAGAA +TCAGAT +TCAGAC +TCAGAG +TCAGTA +TCAGTT +TCAGTC +TCAGTG +TCAGCA +TCAGCT +TCAGCC +TCAGCG +TCAGGA +TCAGGT +TCAGGC +TCAGGG +TCTAAA +TCTAAT +TCTAAC +TCTAAG +TCTATA +TCTATT +TCTATC +TCTATG +TCTACA +TCTACT +TCTACC +TCTACG +TCTAGA +TCTAGT +TCTAGC +TCTAGG +TCTTAA +TCTTAT +TCTTAC +TCTTAG +TCTTTA +TCTTTT +TCTTTC +TCTTTG +TCTTCA +TCTTCT +TCTTCC +TCTTCG +TCTTGA +TCTTGT +TCTTGC +TCTTGG +TCTCAA +TCTCAT +TCTCAC +TCTCAG +TCTCTA +TCTCTT +TCTCTC +TCTCTG +TCTCCA +TCTCCT +TCTCCC +TCTCCG +TCTCGA +TCTCGT +TCTCGC +TCTCGG +TCTGAA +TCTGAT +TCTGAC +TCTGAG +TCTGTA +TCTGTT +TCTGTC +TCTGTG +TCTGCA +TCTGCT +TCTGCC +TCTGCG +TCTGGA +TCTGGT +TCTGGC +TCTGGG +TCCAAA +TCCAAT +TCCAAC +TCCAAG +TCCATA +TCCATT +TCCATC +TCCATG +TCCACA +TCCACT +TCCACC +TCCACG +TCCAGA +TCCAGT +TCCAGC +TCCAGG +TCCTAA +TCCTAT +TCCTAC +TCCTAG +TCCTTA +TCCTTT +TCCTTC +TCCTTG +TCCTCA +TCCTCT +TCCTCC +TCCTCG +TCCTGA +TCCTGT +TCCTGC +TCCTGG +TCCCAA +TCCCAT +TCCCAC +TCCCAG +TCCCTA +TCCCTT +TCCCTC +TCCCTG +TCCCCA +TCCCCT +TCCCCC +TCCCCG +TCCCGA +TCCCGT +TCCCGC +TCCCGG +TCCGAA +TCCGAT +TCCGAC +TCCGAG +TCCGTA +TCCGTT +TCCGTC +TCCGTG +TCCGCA +TCCGCT +TCCGCC +TCCGCG +TCCGGA +TCCGGT +TCCGGC +TCCGGG +TCGAAA +TCGAAT +TCGAAC +TCGAAG +TCGATA +TCGATT +TCGATC +TCGATG +TCGACA +TCGACT +TCGACC +TCGACG +TCGAGA +TCGAGT +TCGAGC +TCGAGG +TCGTAA +TCGTAT +TCGTAC +TCGTAG +TCGTTA +TCGTTT +TCGTTC +TCGTTG +TCGTCA +TCGTCT +TCGTCC +TCGTCG +TCGTGA +TCGTGT +TCGTGC +TCGTGG +TCGCAA +TCGCAT +TCGCAC +TCGCAG +TCGCTA +TCGCTT +TCGCTC +TCGCTG +TCGCCA +TCGCCT +TCGCCC +TCGCCG +TCGCGA +TCGCGT +TCGCGC +TCGCGG +TCGGAA +TCGGAT +TCGGAC +TCGGAG +TCGGTA +TCGGTT +TCGGTC +TCGGTG +TCGGCA +TCGGCT +TCGGCC +TCGGCG +TCGGGA +TCGGGT +TCGGGC +TCGGGG +TGAAAA +TGAAAT +TGAAAC +TGAAAG +TGAATA +TGAATT +TGAATC +TGAATG +TGAACA +TGAACT +TGAACC +TGAACG +TGAAGA +TGAAGT +TGAAGC +TGAAGG +TGATAA +TGATAT +TGATAC +TGATAG +TGATTA +TGATTT +TGATTC +TGATTG +TGATCA +TGATCT +TGATCC +TGATCG +TGATGA +TGATGT +TGATGC +TGATGG +TGACAA +TGACAT +TGACAC +TGACAG +TGACTA +TGACTT +TGACTC +TGACTG +TGACCA +TGACCT +TGACCC +TGACCG +TGACGA +TGACGT +TGACGC +TGACGG +TGAGAA +TGAGAT +TGAGAC +TGAGAG +TGAGTA +TGAGTT +TGAGTC +TGAGTG +TGAGCA +TGAGCT +TGAGCC +TGAGCG +TGAGGA +TGAGGT +TGAGGC +TGAGGG +TGTAAA +TGTAAT +TGTAAC +TGTAAG +TGTATA +TGTATT +TGTATC +TGTATG +TGTACA +TGTACT +TGTACC +TGTACG +TGTAGA +TGTAGT +TGTAGC +TGTAGG +TGTTAA +TGTTAT +TGTTAC +TGTTAG +TGTTTA +TGTTTT +TGTTTC +TGTTTG +TGTTCA +TGTTCT +TGTTCC +TGTTCG +TGTTGA +TGTTGT +TGTTGC +TGTTGG +TGTCAA +TGTCAT +TGTCAC +TGTCAG +TGTCTA +TGTCTT +TGTCTC +TGTCTG +TGTCCA +TGTCCT +TGTCCC +TGTCCG +TGTCGA +TGTCGT +TGTCGC +TGTCGG +TGTGAA +TGTGAT +TGTGAC +TGTGAG +TGTGTA +TGTGTT +TGTGTC +TGTGTG +TGTGCA +TGTGCT +TGTGCC +TGTGCG +TGTGGA +TGTGGT +TGTGGC +TGTGGG +TGCAAA +TGCAAT +TGCAAC +TGCAAG +TGCATA +TGCATT +TGCATC +TGCATG +TGCACA +TGCACT +TGCACC +TGCACG +TGCAGA +TGCAGT +TGCAGC +TGCAGG +TGCTAA +TGCTAT +TGCTAC +TGCTAG +TGCTTA +TGCTTT +TGCTTC +TGCTTG +TGCTCA +TGCTCT +TGCTCC +TGCTCG +TGCTGA +TGCTGT +TGCTGC +TGCTGG +TGCCAA +TGCCAT +TGCCAC +TGCCAG +TGCCTA +TGCCTT +TGCCTC +TGCCTG +TGCCCA +TGCCCT +TGCCCC +TGCCCG +TGCCGA +TGCCGT +TGCCGC +TGCCGG +TGCGAA +TGCGAT +TGCGAC +TGCGAG +TGCGTA +TGCGTT +TGCGTC +TGCGTG +TGCGCA +TGCGCT +TGCGCC +TGCGCG +TGCGGA +TGCGGT +TGCGGC +TGCGGG +TGGAAA +TGGAAT +TGGAAC +TGGAAG +TGGATA +TGGATT +TGGATC +TGGATG +TGGACA +TGGACT +TGGACC +TGGACG +TGGAGA +TGGAGT +TGGAGC +TGGAGG +TGGTAA +TGGTAT +TGGTAC +TGGTAG +TGGTTA +TGGTTT +TGGTTC +TGGTTG +TGGTCA +TGGTCT +TGGTCC +TGGTCG +TGGTGA +TGGTGT +TGGTGC +TGGTGG +TGGCAA +TGGCAT +TGGCAC +TGGCAG +TGGCTA +TGGCTT +TGGCTC +TGGCTG +TGGCCA +TGGCCT +TGGCCC +TGGCCG +TGGCGA +TGGCGT +TGGCGC +TGGCGG +TGGGAA +TGGGAT +TGGGAC +TGGGAG +TGGGTA +TGGGTT +TGGGTC +TGGGTG +TGGGCA +TGGGCT +TGGGCC +TGGGCG +TGGGGA +TGGGGT +TGGGGC +TGGGGG +CAAAAA +CAAAAT +CAAAAC +CAAAAG +CAAATA +CAAATT +CAAATC +CAAATG +CAAACA +CAAACT +CAAACC +CAAACG +CAAAGA +CAAAGT +CAAAGC +CAAAGG +CAATAA +CAATAT +CAATAC +CAATAG +CAATTA +CAATTT +CAATTC +CAATTG +CAATCA +CAATCT +CAATCC +CAATCG +CAATGA +CAATGT +CAATGC +CAATGG +CAACAA +CAACAT +CAACAC +CAACAG +CAACTA +CAACTT +CAACTC +CAACTG +CAACCA +CAACCT +CAACCC +CAACCG +CAACGA +CAACGT +CAACGC +CAACGG +CAAGAA +CAAGAT +CAAGAC +CAAGAG +CAAGTA +CAAGTT +CAAGTC +CAAGTG +CAAGCA +CAAGCT +CAAGCC +CAAGCG +CAAGGA +CAAGGT +CAAGGC +CAAGGG +CATAAA +CATAAT +CATAAC +CATAAG +CATATA +CATATT +CATATC +CATATG +CATACA +CATACT +CATACC +CATACG +CATAGA +CATAGT +CATAGC +CATAGG +CATTAA +CATTAT +CATTAC +CATTAG +CATTTA +CATTTT +CATTTC +CATTTG +CATTCA +CATTCT +CATTCC +CATTCG +CATTGA +CATTGT +CATTGC +CATTGG +CATCAA +CATCAT +CATCAC +CATCAG +CATCTA +CATCTT +CATCTC +CATCTG +CATCCA +CATCCT +CATCCC +CATCCG +CATCGA +CATCGT +CATCGC +CATCGG +CATGAA +CATGAT +CATGAC +CATGAG +CATGTA +CATGTT +CATGTC +CATGTG +CATGCA +CATGCT +CATGCC +CATGCG +CATGGA +CATGGT +CATGGC +CATGGG +CACAAA +CACAAT +CACAAC +CACAAG +CACATA +CACATT +CACATC +CACATG +CACACA +CACACT +CACACC +CACACG +CACAGA +CACAGT +CACAGC +CACAGG +CACTAA +CACTAT +CACTAC +CACTAG +CACTTA +CACTTT +CACTTC +CACTTG +CACTCA +CACTCT +CACTCC +CACTCG +CACTGA +CACTGT +CACTGC +CACTGG +CACCAA +CACCAT +CACCAC +CACCAG +CACCTA +CACCTT +CACCTC +CACCTG +CACCCA +CACCCT +CACCCC +CACCCG +CACCGA +CACCGT +CACCGC +CACCGG +CACGAA +CACGAT +CACGAC +CACGAG +CACGTA +CACGTT +CACGTC +CACGTG +CACGCA +CACGCT +CACGCC +CACGCG +CACGGA +CACGGT +CACGGC +CACGGG +CAGAAA +CAGAAT +CAGAAC +CAGAAG +CAGATA +CAGATT +CAGATC +CAGATG +CAGACA +CAGACT +CAGACC +CAGACG +CAGAGA +CAGAGT +CAGAGC +CAGAGG +CAGTAA +CAGTAT +CAGTAC +CAGTAG +CAGTTA +CAGTTT +CAGTTC +CAGTTG +CAGTCA +CAGTCT +CAGTCC +CAGTCG +CAGTGA +CAGTGT +CAGTGC +CAGTGG +CAGCAA +CAGCAT +CAGCAC +CAGCAG +CAGCTA +CAGCTT +CAGCTC +CAGCTG +CAGCCA +CAGCCT +CAGCCC +CAGCCG +CAGCGA +CAGCGT +CAGCGC +CAGCGG +CAGGAA +CAGGAT +CAGGAC +CAGGAG +CAGGTA +CAGGTT +CAGGTC +CAGGTG +CAGGCA +CAGGCT +CAGGCC +CAGGCG +CAGGGA +CAGGGT +CAGGGC +CAGGGG +CTAAAA +CTAAAT +CTAAAC +CTAAAG +CTAATA +CTAATT +CTAATC +CTAATG +CTAACA +CTAACT +CTAACC +CTAACG +CTAAGA +CTAAGT +CTAAGC +CTAAGG +CTATAA +CTATAT +CTATAC +CTATAG +CTATTA +CTATTT +CTATTC +CTATTG +CTATCA +CTATCT +CTATCC +CTATCG +CTATGA +CTATGT +CTATGC +CTATGG +CTACAA +CTACAT +CTACAC +CTACAG +CTACTA +CTACTT +CTACTC +CTACTG +CTACCA +CTACCT +CTACCC +CTACCG +CTACGA +CTACGT +CTACGC +CTACGG +CTAGAA +CTAGAT +CTAGAC +CTAGAG +CTAGTA +CTAGTT +CTAGTC +CTAGTG +CTAGCA +CTAGCT +CTAGCC +CTAGCG +CTAGGA +CTAGGT +CTAGGC +CTAGGG +CTTAAA +CTTAAT +CTTAAC +CTTAAG +CTTATA +CTTATT +CTTATC +CTTATG +CTTACA +CTTACT +CTTACC +CTTACG +CTTAGA +CTTAGT +CTTAGC +CTTAGG +CTTTAA +CTTTAT +CTTTAC +CTTTAG +CTTTTA +CTTTTT +CTTTTC +CTTTTG +CTTTCA +CTTTCT +CTTTCC +CTTTCG +CTTTGA +CTTTGT +CTTTGC +CTTTGG +CTTCAA +CTTCAT +CTTCAC +CTTCAG +CTTCTA +CTTCTT +CTTCTC +CTTCTG +CTTCCA +CTTCCT +CTTCCC +CTTCCG +CTTCGA +CTTCGT +CTTCGC +CTTCGG +CTTGAA +CTTGAT +CTTGAC +CTTGAG +CTTGTA +CTTGTT +CTTGTC +CTTGTG +CTTGCA +CTTGCT +CTTGCC +CTTGCG +CTTGGA +CTTGGT +CTTGGC +CTTGGG +CTCAAA +CTCAAT +CTCAAC +CTCAAG +CTCATA +CTCATT +CTCATC +CTCATG +CTCACA +CTCACT +CTCACC +CTCACG +CTCAGA +CTCAGT +CTCAGC +CTCAGG +CTCTAA +CTCTAT +CTCTAC +CTCTAG +CTCTTA +CTCTTT +CTCTTC +CTCTTG +CTCTCA +CTCTCT +CTCTCC +CTCTCG +CTCTGA +CTCTGT +CTCTGC +CTCTGG +CTCCAA +CTCCAT +CTCCAC +CTCCAG +CTCCTA +CTCCTT +CTCCTC +CTCCTG +CTCCCA +CTCCCT +CTCCCC +CTCCCG +CTCCGA +CTCCGT +CTCCGC +CTCCGG +CTCGAA +CTCGAT +CTCGAC +CTCGAG +CTCGTA +CTCGTT +CTCGTC +CTCGTG +CTCGCA +CTCGCT +CTCGCC +CTCGCG +CTCGGA +CTCGGT +CTCGGC +CTCGGG +CTGAAA +CTGAAT +CTGAAC +CTGAAG +CTGATA +CTGATT +CTGATC +CTGATG +CTGACA +CTGACT +CTGACC +CTGACG +CTGAGA +CTGAGT +CTGAGC +CTGAGG +CTGTAA +CTGTAT +CTGTAC +CTGTAG +CTGTTA +CTGTTT +CTGTTC +CTGTTG +CTGTCA +CTGTCT +CTGTCC +CTGTCG +CTGTGA +CTGTGT +CTGTGC +CTGTGG +CTGCAA +CTGCAT +CTGCAC +CTGCAG +CTGCTA +CTGCTT +CTGCTC +CTGCTG +CTGCCA +CTGCCT +CTGCCC +CTGCCG +CTGCGA +CTGCGT +CTGCGC +CTGCGG +CTGGAA +CTGGAT +CTGGAC +CTGGAG +CTGGTA +CTGGTT +CTGGTC +CTGGTG +CTGGCA +CTGGCT +CTGGCC +CTGGCG +CTGGGA +CTGGGT +CTGGGC +CTGGGG +CCAAAA +CCAAAT +CCAAAC +CCAAAG +CCAATA +CCAATT +CCAATC +CCAATG +CCAACA +CCAACT +CCAACC +CCAACG +CCAAGA +CCAAGT +CCAAGC +CCAAGG +CCATAA +CCATAT +CCATAC +CCATAG +CCATTA +CCATTT +CCATTC +CCATTG +CCATCA +CCATCT +CCATCC +CCATCG +CCATGA +CCATGT +CCATGC +CCATGG +CCACAA +CCACAT +CCACAC +CCACAG +CCACTA +CCACTT +CCACTC +CCACTG +CCACCA +CCACCT +CCACCC +CCACCG +CCACGA +CCACGT +CCACGC +CCACGG +CCAGAA +CCAGAT +CCAGAC +CCAGAG +CCAGTA +CCAGTT +CCAGTC +CCAGTG +CCAGCA +CCAGCT +CCAGCC +CCAGCG +CCAGGA +CCAGGT +CCAGGC +CCAGGG +CCTAAA +CCTAAT +CCTAAC +CCTAAG +CCTATA +CCTATT +CCTATC +CCTATG +CCTACA +CCTACT +CCTACC +CCTACG +CCTAGA +CCTAGT +CCTAGC +CCTAGG +CCTTAA +CCTTAT +CCTTAC +CCTTAG +CCTTTA +CCTTTT +CCTTTC +CCTTTG +CCTTCA +CCTTCT +CCTTCC +CCTTCG +CCTTGA +CCTTGT +CCTTGC +CCTTGG +CCTCAA +CCTCAT +CCTCAC +CCTCAG +CCTCTA +CCTCTT +CCTCTC +CCTCTG +CCTCCA +CCTCCT +CCTCCC +CCTCCG +CCTCGA +CCTCGT +CCTCGC +CCTCGG +CCTGAA +CCTGAT +CCTGAC +CCTGAG +CCTGTA +CCTGTT +CCTGTC +CCTGTG +CCTGCA +CCTGCT +CCTGCC +CCTGCG +CCTGGA +CCTGGT +CCTGGC +CCTGGG +CCCAAA +CCCAAT +CCCAAC +CCCAAG +CCCATA +CCCATT +CCCATC +CCCATG +CCCACA +CCCACT +CCCACC +CCCACG +CCCAGA +CCCAGT +CCCAGC +CCCAGG +CCCTAA +CCCTAT +CCCTAC +CCCTAG +CCCTTA +CCCTTT +CCCTTC +CCCTTG +CCCTCA +CCCTCT +CCCTCC +CCCTCG +CCCTGA +CCCTGT +CCCTGC +CCCTGG +CCCCAA +CCCCAT +CCCCAC +CCCCAG +CCCCTA +CCCCTT +CCCCTC +CCCCTG +CCCCCA +CCCCCT +CCCCCC +CCCCCG +CCCCGA +CCCCGT +CCCCGC +CCCCGG +CCCGAA +CCCGAT +CCCGAC +CCCGAG +CCCGTA +CCCGTT +CCCGTC +CCCGTG +CCCGCA +CCCGCT +CCCGCC +CCCGCG +CCCGGA +CCCGGT +CCCGGC +CCCGGG +CCGAAA +CCGAAT +CCGAAC +CCGAAG +CCGATA +CCGATT +CCGATC +CCGATG +CCGACA +CCGACT +CCGACC +CCGACG +CCGAGA +CCGAGT +CCGAGC +CCGAGG +CCGTAA +CCGTAT +CCGTAC +CCGTAG +CCGTTA +CCGTTT +CCGTTC +CCGTTG +CCGTCA +CCGTCT +CCGTCC +CCGTCG +CCGTGA +CCGTGT +CCGTGC +CCGTGG +CCGCAA +CCGCAT +CCGCAC +CCGCAG +CCGCTA +CCGCTT +CCGCTC +CCGCTG +CCGCCA +CCGCCT +CCGCCC +CCGCCG +CCGCGA +CCGCGT +CCGCGC +CCGCGG +CCGGAA +CCGGAT +CCGGAC +CCGGAG +CCGGTA +CCGGTT +CCGGTC +CCGGTG +CCGGCA +CCGGCT +CCGGCC +CCGGCG +CCGGGA +CCGGGT +CCGGGC +CCGGGG +CGAAAA +CGAAAT +CGAAAC +CGAAAG +CGAATA +CGAATT +CGAATC +CGAATG +CGAACA +CGAACT +CGAACC +CGAACG +CGAAGA +CGAAGT +CGAAGC +CGAAGG +CGATAA +CGATAT +CGATAC +CGATAG +CGATTA +CGATTT +CGATTC +CGATTG +CGATCA +CGATCT +CGATCC +CGATCG +CGATGA +CGATGT +CGATGC +CGATGG +CGACAA +CGACAT +CGACAC +CGACAG +CGACTA +CGACTT +CGACTC +CGACTG +CGACCA +CGACCT +CGACCC +CGACCG +CGACGA +CGACGT +CGACGC +CGACGG +CGAGAA +CGAGAT +CGAGAC +CGAGAG +CGAGTA +CGAGTT +CGAGTC +CGAGTG +CGAGCA +CGAGCT +CGAGCC +CGAGCG +CGAGGA +CGAGGT +CGAGGC +CGAGGG +CGTAAA +CGTAAT +CGTAAC +CGTAAG +CGTATA +CGTATT +CGTATC +CGTATG +CGTACA +CGTACT +CGTACC +CGTACG +CGTAGA +CGTAGT +CGTAGC +CGTAGG +CGTTAA +CGTTAT +CGTTAC +CGTTAG +CGTTTA +CGTTTT +CGTTTC +CGTTTG +CGTTCA +CGTTCT +CGTTCC +CGTTCG +CGTTGA +CGTTGT +CGTTGC +CGTTGG +CGTCAA +CGTCAT +CGTCAC +CGTCAG +CGTCTA +CGTCTT +CGTCTC +CGTCTG +CGTCCA +CGTCCT +CGTCCC +CGTCCG +CGTCGA +CGTCGT +CGTCGC +CGTCGG +CGTGAA +CGTGAT +CGTGAC +CGTGAG +CGTGTA +CGTGTT +CGTGTC +CGTGTG +CGTGCA +CGTGCT +CGTGCC +CGTGCG +CGTGGA +CGTGGT +CGTGGC +CGTGGG +CGCAAA +CGCAAT +CGCAAC +CGCAAG +CGCATA +CGCATT +CGCATC +CGCATG +CGCACA +CGCACT +CGCACC +CGCACG +CGCAGA +CGCAGT +CGCAGC +CGCAGG +CGCTAA +CGCTAT +CGCTAC +CGCTAG +CGCTTA +CGCTTT +CGCTTC +CGCTTG +CGCTCA +CGCTCT +CGCTCC +CGCTCG +CGCTGA +CGCTGT +CGCTGC +CGCTGG +CGCCAA +CGCCAT +CGCCAC +CGCCAG +CGCCTA +CGCCTT +CGCCTC +CGCCTG +CGCCCA +CGCCCT +CGCCCC +CGCCCG +CGCCGA +CGCCGT +CGCCGC +CGCCGG +CGCGAA +CGCGAT +CGCGAC +CGCGAG +CGCGTA +CGCGTT +CGCGTC +CGCGTG +CGCGCA +CGCGCT +CGCGCC +CGCGCG +CGCGGA +CGCGGT +CGCGGC +CGCGGG +CGGAAA +CGGAAT +CGGAAC +CGGAAG +CGGATA +CGGATT +CGGATC +CGGATG +CGGACA +CGGACT +CGGACC +CGGACG +CGGAGA +CGGAGT +CGGAGC +CGGAGG +CGGTAA +CGGTAT +CGGTAC +CGGTAG +CGGTTA +CGGTTT +CGGTTC +CGGTTG +CGGTCA +CGGTCT +CGGTCC +CGGTCG +CGGTGA +CGGTGT +CGGTGC +CGGTGG +CGGCAA +CGGCAT +CGGCAC +CGGCAG +CGGCTA +CGGCTT +CGGCTC +CGGCTG +CGGCCA +CGGCCT +CGGCCC +CGGCCG +CGGCGA +CGGCGT +CGGCGC +CGGCGG +CGGGAA +CGGGAT +CGGGAC +CGGGAG +CGGGTA +CGGGTT +CGGGTC +CGGGTG +CGGGCA +CGGGCT +CGGGCC +CGGGCG +CGGGGA +CGGGGT +CGGGGC +CGGGGG +GAAAAA +GAAAAT +GAAAAC +GAAAAG +GAAATA +GAAATT +GAAATC +GAAATG +GAAACA +GAAACT +GAAACC +GAAACG +GAAAGA +GAAAGT +GAAAGC +GAAAGG +GAATAA +GAATAT +GAATAC +GAATAG +GAATTA +GAATTT +GAATTC +GAATTG +GAATCA +GAATCT +GAATCC +GAATCG +GAATGA +GAATGT +GAATGC +GAATGG +GAACAA +GAACAT +GAACAC +GAACAG +GAACTA +GAACTT +GAACTC +GAACTG +GAACCA +GAACCT +GAACCC +GAACCG +GAACGA +GAACGT +GAACGC +GAACGG +GAAGAA +GAAGAT +GAAGAC +GAAGAG +GAAGTA +GAAGTT +GAAGTC +GAAGTG +GAAGCA +GAAGCT +GAAGCC +GAAGCG +GAAGGA +GAAGGT +GAAGGC +GAAGGG +GATAAA +GATAAT +GATAAC +GATAAG +GATATA +GATATT +GATATC +GATATG +GATACA +GATACT +GATACC +GATACG +GATAGA +GATAGT +GATAGC +GATAGG +GATTAA +GATTAT +GATTAC +GATTAG +GATTTA +GATTTT +GATTTC +GATTTG +GATTCA +GATTCT +GATTCC +GATTCG +GATTGA +GATTGT +GATTGC +GATTGG +GATCAA +GATCAT +GATCAC +GATCAG +GATCTA +GATCTT +GATCTC +GATCTG +GATCCA +GATCCT +GATCCC +GATCCG +GATCGA +GATCGT +GATCGC +GATCGG +GATGAA +GATGAT +GATGAC +GATGAG +GATGTA +GATGTT +GATGTC +GATGTG +GATGCA +GATGCT +GATGCC +GATGCG +GATGGA +GATGGT +GATGGC +GATGGG +GACAAA +GACAAT +GACAAC +GACAAG +GACATA +GACATT +GACATC +GACATG +GACACA +GACACT +GACACC +GACACG +GACAGA +GACAGT +GACAGC +GACAGG +GACTAA +GACTAT +GACTAC +GACTAG +GACTTA +GACTTT +GACTTC +GACTTG +GACTCA +GACTCT +GACTCC +GACTCG +GACTGA +GACTGT +GACTGC +GACTGG +GACCAA +GACCAT +GACCAC +GACCAG +GACCTA +GACCTT +GACCTC +GACCTG +GACCCA +GACCCT +GACCCC +GACCCG +GACCGA +GACCGT +GACCGC +GACCGG +GACGAA +GACGAT +GACGAC +GACGAG +GACGTA +GACGTT +GACGTC +GACGTG +GACGCA +GACGCT +GACGCC +GACGCG +GACGGA +GACGGT +GACGGC +GACGGG +GAGAAA +GAGAAT +GAGAAC +GAGAAG +GAGATA +GAGATT +GAGATC +GAGATG +GAGACA +GAGACT +GAGACC +GAGACG +GAGAGA +GAGAGT +GAGAGC +GAGAGG +GAGTAA +GAGTAT +GAGTAC +GAGTAG +GAGTTA +GAGTTT +GAGTTC +GAGTTG +GAGTCA +GAGTCT +GAGTCC +GAGTCG +GAGTGA +GAGTGT +GAGTGC +GAGTGG +GAGCAA +GAGCAT +GAGCAC +GAGCAG +GAGCTA +GAGCTT +GAGCTC +GAGCTG +GAGCCA +GAGCCT +GAGCCC +GAGCCG +GAGCGA +GAGCGT +GAGCGC +GAGCGG +GAGGAA +GAGGAT +GAGGAC +GAGGAG +GAGGTA +GAGGTT +GAGGTC +GAGGTG +GAGGCA +GAGGCT +GAGGCC +GAGGCG +GAGGGA +GAGGGT +GAGGGC +GAGGGG +GTAAAA +GTAAAT +GTAAAC +GTAAAG +GTAATA +GTAATT +GTAATC +GTAATG +GTAACA +GTAACT +GTAACC +GTAACG +GTAAGA +GTAAGT +GTAAGC +GTAAGG +GTATAA +GTATAT +GTATAC +GTATAG +GTATTA +GTATTT +GTATTC +GTATTG +GTATCA +GTATCT +GTATCC +GTATCG +GTATGA +GTATGT +GTATGC +GTATGG +GTACAA +GTACAT +GTACAC +GTACAG +GTACTA +GTACTT +GTACTC +GTACTG +GTACCA +GTACCT +GTACCC +GTACCG +GTACGA +GTACGT +GTACGC +GTACGG +GTAGAA +GTAGAT +GTAGAC +GTAGAG +GTAGTA +GTAGTT +GTAGTC +GTAGTG +GTAGCA +GTAGCT +GTAGCC +GTAGCG +GTAGGA +GTAGGT +GTAGGC +GTAGGG +GTTAAA +GTTAAT +GTTAAC +GTTAAG +GTTATA +GTTATT +GTTATC +GTTATG +GTTACA +GTTACT +GTTACC +GTTACG +GTTAGA +GTTAGT +GTTAGC +GTTAGG +GTTTAA +GTTTAT +GTTTAC +GTTTAG +GTTTTA +GTTTTT +GTTTTC +GTTTTG +GTTTCA +GTTTCT +GTTTCC +GTTTCG +GTTTGA +GTTTGT +GTTTGC +GTTTGG +GTTCAA +GTTCAT +GTTCAC +GTTCAG +GTTCTA +GTTCTT +GTTCTC +GTTCTG +GTTCCA +GTTCCT +GTTCCC +GTTCCG +GTTCGA +GTTCGT +GTTCGC +GTTCGG +GTTGAA +GTTGAT +GTTGAC +GTTGAG +GTTGTA +GTTGTT +GTTGTC +GTTGTG +GTTGCA +GTTGCT +GTTGCC +GTTGCG +GTTGGA +GTTGGT +GTTGGC +GTTGGG +GTCAAA +GTCAAT +GTCAAC +GTCAAG +GTCATA +GTCATT +GTCATC +GTCATG +GTCACA +GTCACT +GTCACC +GTCACG +GTCAGA +GTCAGT +GTCAGC +GTCAGG +GTCTAA +GTCTAT +GTCTAC +GTCTAG +GTCTTA +GTCTTT +GTCTTC +GTCTTG +GTCTCA +GTCTCT +GTCTCC +GTCTCG +GTCTGA +GTCTGT +GTCTGC +GTCTGG +GTCCAA +GTCCAT +GTCCAC +GTCCAG +GTCCTA +GTCCTT +GTCCTC +GTCCTG +GTCCCA +GTCCCT +GTCCCC +GTCCCG +GTCCGA +GTCCGT +GTCCGC +GTCCGG +GTCGAA +GTCGAT +GTCGAC +GTCGAG +GTCGTA +GTCGTT +GTCGTC +GTCGTG +GTCGCA +GTCGCT +GTCGCC +GTCGCG +GTCGGA +GTCGGT +GTCGGC +GTCGGG +GTGAAA +GTGAAT +GTGAAC +GTGAAG +GTGATA +GTGATT +GTGATC +GTGATG +GTGACA +GTGACT +GTGACC +GTGACG +GTGAGA +GTGAGT +GTGAGC +GTGAGG +GTGTAA +GTGTAT +GTGTAC +GTGTAG +GTGTTA +GTGTTT +GTGTTC +GTGTTG +GTGTCA +GTGTCT +GTGTCC +GTGTCG +GTGTGA +GTGTGT +GTGTGC +GTGTGG +GTGCAA +GTGCAT +GTGCAC +GTGCAG +GTGCTA +GTGCTT +GTGCTC +GTGCTG +GTGCCA +GTGCCT +GTGCCC +GTGCCG +GTGCGA +GTGCGT +GTGCGC +GTGCGG +GTGGAA +GTGGAT +GTGGAC +GTGGAG +GTGGTA +GTGGTT +GTGGTC +GTGGTG +GTGGCA +GTGGCT +GTGGCC +GTGGCG +GTGGGA +GTGGGT +GTGGGC +GTGGGG +GCAAAA +GCAAAT +GCAAAC +GCAAAG +GCAATA +GCAATT +GCAATC +GCAATG +GCAACA +GCAACT +GCAACC +GCAACG +GCAAGA +GCAAGT +GCAAGC +GCAAGG +GCATAA +GCATAT +GCATAC +GCATAG +GCATTA +GCATTT +GCATTC +GCATTG +GCATCA +GCATCT +GCATCC +GCATCG +GCATGA +GCATGT +GCATGC +GCATGG +GCACAA +GCACAT +GCACAC +GCACAG +GCACTA +GCACTT +GCACTC +GCACTG +GCACCA +GCACCT +GCACCC +GCACCG +GCACGA +GCACGT +GCACGC +GCACGG +GCAGAA +GCAGAT +GCAGAC +GCAGAG +GCAGTA +GCAGTT +GCAGTC +GCAGTG +GCAGCA +GCAGCT +GCAGCC +GCAGCG +GCAGGA +GCAGGT +GCAGGC +GCAGGG +GCTAAA +GCTAAT +GCTAAC +GCTAAG +GCTATA +GCTATT +GCTATC +GCTATG +GCTACA +GCTACT +GCTACC +GCTACG +GCTAGA +GCTAGT +GCTAGC +GCTAGG +GCTTAA +GCTTAT +GCTTAC +GCTTAG +GCTTTA +GCTTTT +GCTTTC +GCTTTG +GCTTCA +GCTTCT +GCTTCC +GCTTCG +GCTTGA +GCTTGT +GCTTGC +GCTTGG +GCTCAA +GCTCAT +GCTCAC +GCTCAG +GCTCTA +GCTCTT +GCTCTC +GCTCTG +GCTCCA +GCTCCT +GCTCCC +GCTCCG +GCTCGA +GCTCGT +GCTCGC +GCTCGG +GCTGAA +GCTGAT +GCTGAC +GCTGAG +GCTGTA +GCTGTT +GCTGTC +GCTGTG +GCTGCA +GCTGCT +GCTGCC +GCTGCG +GCTGGA +GCTGGT +GCTGGC +GCTGGG +GCCAAA +GCCAAT +GCCAAC +GCCAAG +GCCATA +GCCATT +GCCATC +GCCATG +GCCACA +GCCACT +GCCACC +GCCACG +GCCAGA +GCCAGT +GCCAGC +GCCAGG +GCCTAA +GCCTAT +GCCTAC +GCCTAG +GCCTTA +GCCTTT +GCCTTC +GCCTTG +GCCTCA +GCCTCT +GCCTCC +GCCTCG +GCCTGA +GCCTGT +GCCTGC +GCCTGG +GCCCAA +GCCCAT +GCCCAC +GCCCAG +GCCCTA +GCCCTT +GCCCTC +GCCCTG +GCCCCA +GCCCCT +GCCCCC +GCCCCG +GCCCGA +GCCCGT +GCCCGC +GCCCGG +GCCGAA +GCCGAT +GCCGAC +GCCGAG +GCCGTA +GCCGTT +GCCGTC +GCCGTG +GCCGCA +GCCGCT +GCCGCC +GCCGCG +GCCGGA +GCCGGT +GCCGGC +GCCGGG +GCGAAA +GCGAAT +GCGAAC +GCGAAG +GCGATA +GCGATT +GCGATC +GCGATG +GCGACA +GCGACT +GCGACC +GCGACG +GCGAGA +GCGAGT +GCGAGC +GCGAGG +GCGTAA +GCGTAT +GCGTAC +GCGTAG +GCGTTA +GCGTTT +GCGTTC +GCGTTG +GCGTCA +GCGTCT +GCGTCC +GCGTCG +GCGTGA +GCGTGT +GCGTGC +GCGTGG +GCGCAA +GCGCAT +GCGCAC +GCGCAG +GCGCTA +GCGCTT +GCGCTC +GCGCTG +GCGCCA +GCGCCT +GCGCCC +GCGCCG +GCGCGA +GCGCGT +GCGCGC +GCGCGG +GCGGAA +GCGGAT +GCGGAC +GCGGAG +GCGGTA +GCGGTT +GCGGTC +GCGGTG +GCGGCA +GCGGCT +GCGGCC +GCGGCG +GCGGGA +GCGGGT +GCGGGC +GCGGGG +GGAAAA +GGAAAT +GGAAAC +GGAAAG +GGAATA +GGAATT +GGAATC +GGAATG +GGAACA +GGAACT +GGAACC +GGAACG +GGAAGA +GGAAGT +GGAAGC +GGAAGG +GGATAA +GGATAT +GGATAC +GGATAG +GGATTA +GGATTT +GGATTC +GGATTG +GGATCA +GGATCT +GGATCC +GGATCG +GGATGA +GGATGT +GGATGC +GGATGG +GGACAA +GGACAT +GGACAC +GGACAG +GGACTA +GGACTT +GGACTC +GGACTG +GGACCA +GGACCT +GGACCC +GGACCG +GGACGA +GGACGT +GGACGC +GGACGG +GGAGAA +GGAGAT +GGAGAC +GGAGAG +GGAGTA +GGAGTT +GGAGTC +GGAGTG +GGAGCA +GGAGCT +GGAGCC +GGAGCG +GGAGGA +GGAGGT +GGAGGC +GGAGGG +GGTAAA +GGTAAT +GGTAAC +GGTAAG +GGTATA +GGTATT +GGTATC +GGTATG +GGTACA +GGTACT +GGTACC +GGTACG +GGTAGA +GGTAGT +GGTAGC +GGTAGG +GGTTAA +GGTTAT +GGTTAC +GGTTAG +GGTTTA +GGTTTT +GGTTTC +GGTTTG +GGTTCA +GGTTCT +GGTTCC +GGTTCG +GGTTGA +GGTTGT +GGTTGC +GGTTGG +GGTCAA +GGTCAT +GGTCAC +GGTCAG +GGTCTA +GGTCTT +GGTCTC +GGTCTG +GGTCCA +GGTCCT +GGTCCC +GGTCCG +GGTCGA +GGTCGT +GGTCGC +GGTCGG +GGTGAA +GGTGAT +GGTGAC +GGTGAG +GGTGTA +GGTGTT +GGTGTC +GGTGTG +GGTGCA +GGTGCT +GGTGCC +GGTGCG +GGTGGA +GGTGGT +GGTGGC +GGTGGG +GGCAAA +GGCAAT +GGCAAC +GGCAAG +GGCATA +GGCATT +GGCATC +GGCATG +GGCACA +GGCACT +GGCACC +GGCACG +GGCAGA +GGCAGT +GGCAGC +GGCAGG +GGCTAA +GGCTAT +GGCTAC +GGCTAG +GGCTTA +GGCTTT +GGCTTC +GGCTTG +GGCTCA +GGCTCT +GGCTCC +GGCTCG +GGCTGA +GGCTGT +GGCTGC +GGCTGG +GGCCAA +GGCCAT +GGCCAC +GGCCAG +GGCCTA +GGCCTT +GGCCTC +GGCCTG +GGCCCA +GGCCCT +GGCCCC +GGCCCG +GGCCGA +GGCCGT +GGCCGC +GGCCGG +GGCGAA +GGCGAT +GGCGAC +GGCGAG +GGCGTA +GGCGTT +GGCGTC +GGCGTG +GGCGCA +GGCGCT +GGCGCC +GGCGCG +GGCGGA +GGCGGT +GGCGGC +GGCGGG +GGGAAA +GGGAAT +GGGAAC +GGGAAG +GGGATA +GGGATT +GGGATC +GGGATG +GGGACA +GGGACT +GGGACC +GGGACG +GGGAGA +GGGAGT +GGGAGC +GGGAGG +GGGTAA +GGGTAT +GGGTAC +GGGTAG +GGGTTA +GGGTTT +GGGTTC +GGGTTG +GGGTCA +GGGTCT +GGGTCC +GGGTCG +GGGTGA +GGGTGT +GGGTGC +GGGTGG +GGGCAA +GGGCAT +GGGCAC +GGGCAG +GGGCTA +GGGCTT +GGGCTC +GGGCTG +GGGCCA +GGGCCT +GGGCCC +GGGCCG +GGGCGA +GGGCGT +GGGCGC +GGGCGG +GGGGAA +GGGGAT +GGGGAC +GGGGAG +GGGGTA +GGGGTT +GGGGTC +GGGGTG +GGGGCA +GGGGCT +GGGGCC +GGGGCG +GGGGGA +GGGGGT +GGGGGC +GGGGGG diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/README.md b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..500fd976d92b217455f181bc98c77e612d9a18d9 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/README.md @@ -0,0 +1,202 @@ +--- +base_model: fabihamakhdoomi/TinyDNABERT +library_name: peft +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.12.0 \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/adapter_config.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..04c63e99bbc6b810449dece367c0afb165125b92 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/adapter_config.json @@ -0,0 +1,42 @@ +{ + "alpha_pattern": {}, + "auto_mapping": null, + "base_model_name_or_path": "fabihamakhdoomi/TinyDNABERT", + "bias": "none", + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 16, + "lora_dropout": 0.2, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": [ + "classifier", + "score" + ], + "peft_type": "LORA", + "r": 8, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "bert.encoder.layer.1.attention.self.query", + "bert.encoder.layer.0.output.dense", + "bert.encoder.layer.1.output.dense", + "bert.encoder.layer.0.intermediate.dense", + "bert.encoder.layer.1.intermediate.dense", + "bert.encoder.layer.0.attention.output.dense", + "bert.encoder.layer.1.attention.output.dense", + "bert.encoder.layer.0.attention.self.query", + "bert.encoder.layer.0.attention.self.value", + "bert.encoder.layer.1.attention.self.key", + "bert.encoder.layer.1.attention.self.value", + "bert.encoder.layer.0.attention.self.key" + ], + "task_type": "SEQ_CLS", + "use_dora": false, + "use_rslora": false +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/adapter_model.safetensors b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/adapter_model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..9b63d8364ad37945584951a86f16ac73752ce726 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/adapter_model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:513028e503a5cf3a4c7f99cb00b7c2a06f02650d0c97197e6fdf8d8fec5a62c0 +size 151952 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/optimizer.pt b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..6e7015f29028197d0f74e29c374b5f0c4289ca08 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bf41d052b17b36dd8bb6383e2e950c7e3b7846f3359dcd172aa202d9d9948431 +size 312682 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/rng_state.pth b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..bd3d1cad75d68826b1d075775d8e5141764e5725 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:855054099544841bdc3a4d2a62131897c2be69ce849e4324026c75100489f631 +size 14244 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/scheduler.pt b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..14aa799e6f5c93344d1b233b91d0bda9d8eb59e7 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:eca4da5723e14faf16c3be96d23919390da5288d0d9db40021d332ee1df25b93 +size 1064 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/special_tokens_map.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..a8b3208c2884c4efb86e49300fdd3dc877220cdf --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/special_tokens_map.json @@ -0,0 +1,7 @@ +{ + "cls_token": "[CLS]", + "mask_token": "[MASK]", + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/tokenizer.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..ebd782de6963b9032ce9374cfe30f5275be1a30b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/tokenizer.json @@ -0,0 +1,4265 @@ +{ + "version": "1.0", + "truncation": { + "direction": "Right", + "max_length": 512, + "strategy": "LongestFirst", + "stride": 0 + }, + "padding": { + "strategy": { + "Fixed": 512 + }, + "direction": "Right", + "pad_to_multiple_of": null, + "pad_id": 0, + "pad_type_id": 0, + "pad_token": "[PAD]" + }, + "added_tokens": [ + { + "id": 0, + "content": "[PAD]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 1, + "content": "[UNK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 2, + "content": "[CLS]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 3, + "content": "[SEP]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 4, + "content": "[MASK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + } + ], + "normalizer": { + "type": "BertNormalizer", + "clean_text": true, + "handle_chinese_chars": true, + "strip_accents": null, + "lowercase": false + }, + "pre_tokenizer": { + "type": "BertPreTokenizer" + }, + "post_processor": { + "type": "TemplateProcessing", + "single": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + } + ], + "pair": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "B", + "type_id": 1 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 1 + } + } + ], + "special_tokens": { + "[CLS]": { + "id": "[CLS]", + "ids": [ + 2 + ], + "tokens": [ + "[CLS]" + ] + }, + "[SEP]": { + "id": "[SEP]", + "ids": [ + 3 + ], + "tokens": [ + "[SEP]" + ] + } + } + }, + "decoder": { + "type": "WordPiece", + "prefix": "##", + "cleanup": true + }, + "model": { + "type": "WordPiece", + "unk_token": "[UNK]", + "continuing_subword_prefix": "##", + "max_input_chars_per_word": 100, + "vocab": { + "[PAD]": 0, + "[UNK]": 1, + "[CLS]": 2, + "[SEP]": 3, + "[MASK]": 4, + "AAAAAA": 5, + "AAAAAT": 6, + "AAAAAC": 7, + "AAAAAG": 8, + "AAAATA": 9, + "AAAATT": 10, + "AAAATC": 11, + "AAAATG": 12, + "AAAACA": 13, + "AAAACT": 14, + "AAAACC": 15, + "AAAACG": 16, + "AAAAGA": 17, + "AAAAGT": 18, + "AAAAGC": 19, + "AAAAGG": 20, + "AAATAA": 21, + "AAATAT": 22, + "AAATAC": 23, + "AAATAG": 24, + "AAATTA": 25, + "AAATTT": 26, + "AAATTC": 27, + "AAATTG": 28, + "AAATCA": 29, + "AAATCT": 30, + "AAATCC": 31, + "AAATCG": 32, + "AAATGA": 33, + "AAATGT": 34, + "AAATGC": 35, + "AAATGG": 36, + "AAACAA": 37, + "AAACAT": 38, + "AAACAC": 39, + "AAACAG": 40, + "AAACTA": 41, + "AAACTT": 42, + "AAACTC": 43, + "AAACTG": 44, + "AAACCA": 45, + "AAACCT": 46, + "AAACCC": 47, + "AAACCG": 48, + "AAACGA": 49, + "AAACGT": 50, + "AAACGC": 51, + "AAACGG": 52, + "AAAGAA": 53, + "AAAGAT": 54, + "AAAGAC": 55, + "AAAGAG": 56, + "AAAGTA": 57, + "AAAGTT": 58, + "AAAGTC": 59, + "AAAGTG": 60, + "AAAGCA": 61, + "AAAGCT": 62, + "AAAGCC": 63, + "AAAGCG": 64, + "AAAGGA": 65, + "AAAGGT": 66, + "AAAGGC": 67, + "AAAGGG": 68, + "AATAAA": 69, + "AATAAT": 70, + "AATAAC": 71, + "AATAAG": 72, + "AATATA": 73, + "AATATT": 74, + "AATATC": 75, + "AATATG": 76, + "AATACA": 77, + "AATACT": 78, + "AATACC": 79, + "AATACG": 80, + "AATAGA": 81, + "AATAGT": 82, + "AATAGC": 83, + "AATAGG": 84, + "AATTAA": 85, + "AATTAT": 86, + "AATTAC": 87, + "AATTAG": 88, + "AATTTA": 89, + "AATTTT": 90, + "AATTTC": 91, + "AATTTG": 92, + "AATTCA": 93, + "AATTCT": 94, + "AATTCC": 95, + "AATTCG": 96, + "AATTGA": 97, + "AATTGT": 98, + "AATTGC": 99, + "AATTGG": 100, + "AATCAA": 101, + "AATCAT": 102, + "AATCAC": 103, + "AATCAG": 104, + "AATCTA": 105, + "AATCTT": 106, + "AATCTC": 107, + "AATCTG": 108, + "AATCCA": 109, + "AATCCT": 110, + "AATCCC": 111, + "AATCCG": 112, + "AATCGA": 113, + "AATCGT": 114, + "AATCGC": 115, + "AATCGG": 116, + "AATGAA": 117, + "AATGAT": 118, + "AATGAC": 119, + "AATGAG": 120, + "AATGTA": 121, + "AATGTT": 122, + "AATGTC": 123, + "AATGTG": 124, + "AATGCA": 125, + "AATGCT": 126, + "AATGCC": 127, + "AATGCG": 128, + "AATGGA": 129, + "AATGGT": 130, + "AATGGC": 131, + "AATGGG": 132, + "AACAAA": 133, + "AACAAT": 134, + "AACAAC": 135, + "AACAAG": 136, + "AACATA": 137, + "AACATT": 138, + "AACATC": 139, + "AACATG": 140, + "AACACA": 141, + "AACACT": 142, + "AACACC": 143, + "AACACG": 144, + "AACAGA": 145, + "AACAGT": 146, + "AACAGC": 147, + "AACAGG": 148, + "AACTAA": 149, + "AACTAT": 150, + "AACTAC": 151, + "AACTAG": 152, + "AACTTA": 153, + "AACTTT": 154, + "AACTTC": 155, + "AACTTG": 156, + "AACTCA": 157, + "AACTCT": 158, + "AACTCC": 159, + "AACTCG": 160, + "AACTGA": 161, + "AACTGT": 162, + "AACTGC": 163, + "AACTGG": 164, + "AACCAA": 165, + "AACCAT": 166, + "AACCAC": 167, + "AACCAG": 168, + "AACCTA": 169, + "AACCTT": 170, + "AACCTC": 171, + "AACCTG": 172, + "AACCCA": 173, + "AACCCT": 174, + "AACCCC": 175, + "AACCCG": 176, + "AACCGA": 177, + "AACCGT": 178, + "AACCGC": 179, + "AACCGG": 180, + "AACGAA": 181, + "AACGAT": 182, + "AACGAC": 183, + "AACGAG": 184, + "AACGTA": 185, + "AACGTT": 186, + "AACGTC": 187, + "AACGTG": 188, + "AACGCA": 189, + "AACGCT": 190, + "AACGCC": 191, + "AACGCG": 192, + "AACGGA": 193, + "AACGGT": 194, + "AACGGC": 195, + "AACGGG": 196, + "AAGAAA": 197, + "AAGAAT": 198, + "AAGAAC": 199, + "AAGAAG": 200, + "AAGATA": 201, + "AAGATT": 202, + "AAGATC": 203, + "AAGATG": 204, + "AAGACA": 205, + "AAGACT": 206, + "AAGACC": 207, + "AAGACG": 208, + "AAGAGA": 209, + "AAGAGT": 210, + "AAGAGC": 211, + "AAGAGG": 212, + "AAGTAA": 213, + "AAGTAT": 214, + "AAGTAC": 215, + "AAGTAG": 216, + "AAGTTA": 217, + "AAGTTT": 218, + "AAGTTC": 219, + "AAGTTG": 220, + "AAGTCA": 221, + "AAGTCT": 222, + "AAGTCC": 223, + "AAGTCG": 224, + "AAGTGA": 225, + "AAGTGT": 226, + "AAGTGC": 227, + "AAGTGG": 228, + "AAGCAA": 229, + "AAGCAT": 230, + "AAGCAC": 231, + "AAGCAG": 232, + "AAGCTA": 233, + "AAGCTT": 234, + "AAGCTC": 235, + "AAGCTG": 236, + "AAGCCA": 237, + "AAGCCT": 238, + "AAGCCC": 239, + "AAGCCG": 240, + "AAGCGA": 241, + "AAGCGT": 242, + "AAGCGC": 243, + "AAGCGG": 244, + "AAGGAA": 245, + "AAGGAT": 246, + "AAGGAC": 247, + "AAGGAG": 248, + "AAGGTA": 249, + "AAGGTT": 250, + "AAGGTC": 251, + "AAGGTG": 252, + "AAGGCA": 253, + "AAGGCT": 254, + "AAGGCC": 255, + "AAGGCG": 256, + "AAGGGA": 257, + "AAGGGT": 258, + "AAGGGC": 259, + "AAGGGG": 260, + "ATAAAA": 261, + "ATAAAT": 262, + "ATAAAC": 263, + "ATAAAG": 264, + "ATAATA": 265, + "ATAATT": 266, + "ATAATC": 267, + "ATAATG": 268, + "ATAACA": 269, + "ATAACT": 270, + "ATAACC": 271, + "ATAACG": 272, + "ATAAGA": 273, + "ATAAGT": 274, + "ATAAGC": 275, + "ATAAGG": 276, + "ATATAA": 277, + "ATATAT": 278, + "ATATAC": 279, + "ATATAG": 280, + "ATATTA": 281, + "ATATTT": 282, + "ATATTC": 283, + "ATATTG": 284, + "ATATCA": 285, + "ATATCT": 286, + "ATATCC": 287, + "ATATCG": 288, + "ATATGA": 289, + "ATATGT": 290, + "ATATGC": 291, + "ATATGG": 292, + "ATACAA": 293, + "ATACAT": 294, + "ATACAC": 295, + "ATACAG": 296, + "ATACTA": 297, + "ATACTT": 298, + "ATACTC": 299, + "ATACTG": 300, + "ATACCA": 301, + "ATACCT": 302, + "ATACCC": 303, + "ATACCG": 304, + "ATACGA": 305, + "ATACGT": 306, + "ATACGC": 307, + "ATACGG": 308, + "ATAGAA": 309, + "ATAGAT": 310, + "ATAGAC": 311, + "ATAGAG": 312, + "ATAGTA": 313, + "ATAGTT": 314, + "ATAGTC": 315, + "ATAGTG": 316, + "ATAGCA": 317, + "ATAGCT": 318, + "ATAGCC": 319, + "ATAGCG": 320, + "ATAGGA": 321, + "ATAGGT": 322, + "ATAGGC": 323, + "ATAGGG": 324, + "ATTAAA": 325, + "ATTAAT": 326, + "ATTAAC": 327, + "ATTAAG": 328, + "ATTATA": 329, + "ATTATT": 330, + "ATTATC": 331, + "ATTATG": 332, + "ATTACA": 333, + "ATTACT": 334, + "ATTACC": 335, + "ATTACG": 336, + "ATTAGA": 337, + "ATTAGT": 338, + "ATTAGC": 339, + "ATTAGG": 340, + "ATTTAA": 341, + "ATTTAT": 342, + "ATTTAC": 343, + "ATTTAG": 344, + "ATTTTA": 345, + "ATTTTT": 346, + "ATTTTC": 347, + "ATTTTG": 348, + "ATTTCA": 349, + "ATTTCT": 350, + "ATTTCC": 351, + "ATTTCG": 352, + "ATTTGA": 353, + "ATTTGT": 354, + "ATTTGC": 355, + "ATTTGG": 356, + "ATTCAA": 357, + "ATTCAT": 358, + "ATTCAC": 359, + "ATTCAG": 360, + "ATTCTA": 361, + "ATTCTT": 362, + "ATTCTC": 363, + "ATTCTG": 364, + "ATTCCA": 365, + "ATTCCT": 366, + "ATTCCC": 367, + "ATTCCG": 368, + "ATTCGA": 369, + "ATTCGT": 370, + "ATTCGC": 371, + "ATTCGG": 372, + "ATTGAA": 373, + "ATTGAT": 374, + "ATTGAC": 375, + "ATTGAG": 376, + "ATTGTA": 377, + "ATTGTT": 378, + "ATTGTC": 379, + "ATTGTG": 380, + "ATTGCA": 381, + "ATTGCT": 382, + "ATTGCC": 383, + "ATTGCG": 384, + "ATTGGA": 385, + "ATTGGT": 386, + "ATTGGC": 387, + "ATTGGG": 388, + "ATCAAA": 389, + "ATCAAT": 390, + "ATCAAC": 391, + "ATCAAG": 392, + "ATCATA": 393, + "ATCATT": 394, + "ATCATC": 395, + "ATCATG": 396, + "ATCACA": 397, + "ATCACT": 398, + "ATCACC": 399, + "ATCACG": 400, + "ATCAGA": 401, + "ATCAGT": 402, + "ATCAGC": 403, + "ATCAGG": 404, + "ATCTAA": 405, + "ATCTAT": 406, + "ATCTAC": 407, + "ATCTAG": 408, + "ATCTTA": 409, + "ATCTTT": 410, + "ATCTTC": 411, + "ATCTTG": 412, + "ATCTCA": 413, + "ATCTCT": 414, + "ATCTCC": 415, + "ATCTCG": 416, + "ATCTGA": 417, + "ATCTGT": 418, + "ATCTGC": 419, + "ATCTGG": 420, + "ATCCAA": 421, + "ATCCAT": 422, + "ATCCAC": 423, + "ATCCAG": 424, + "ATCCTA": 425, + "ATCCTT": 426, + "ATCCTC": 427, + "ATCCTG": 428, + "ATCCCA": 429, + "ATCCCT": 430, + "ATCCCC": 431, + "ATCCCG": 432, + "ATCCGA": 433, + "ATCCGT": 434, + "ATCCGC": 435, + "ATCCGG": 436, + "ATCGAA": 437, + "ATCGAT": 438, + "ATCGAC": 439, + "ATCGAG": 440, + "ATCGTA": 441, + "ATCGTT": 442, + "ATCGTC": 443, + "ATCGTG": 444, + "ATCGCA": 445, + "ATCGCT": 446, + "ATCGCC": 447, + "ATCGCG": 448, + "ATCGGA": 449, + "ATCGGT": 450, + "ATCGGC": 451, + "ATCGGG": 452, + "ATGAAA": 453, + "ATGAAT": 454, + "ATGAAC": 455, + "ATGAAG": 456, + "ATGATA": 457, + "ATGATT": 458, + "ATGATC": 459, + "ATGATG": 460, + "ATGACA": 461, + "ATGACT": 462, + "ATGACC": 463, + "ATGACG": 464, + "ATGAGA": 465, + "ATGAGT": 466, + "ATGAGC": 467, + "ATGAGG": 468, + "ATGTAA": 469, + "ATGTAT": 470, + "ATGTAC": 471, + "ATGTAG": 472, + "ATGTTA": 473, + "ATGTTT": 474, + "ATGTTC": 475, + "ATGTTG": 476, + "ATGTCA": 477, + "ATGTCT": 478, + "ATGTCC": 479, + "ATGTCG": 480, + "ATGTGA": 481, + "ATGTGT": 482, + "ATGTGC": 483, + "ATGTGG": 484, + "ATGCAA": 485, + "ATGCAT": 486, + "ATGCAC": 487, + "ATGCAG": 488, + "ATGCTA": 489, + "ATGCTT": 490, + "ATGCTC": 491, + "ATGCTG": 492, + "ATGCCA": 493, + "ATGCCT": 494, + "ATGCCC": 495, + "ATGCCG": 496, + "ATGCGA": 497, + "ATGCGT": 498, + "ATGCGC": 499, + "ATGCGG": 500, + "ATGGAA": 501, + "ATGGAT": 502, + "ATGGAC": 503, + "ATGGAG": 504, + "ATGGTA": 505, + "ATGGTT": 506, + "ATGGTC": 507, + "ATGGTG": 508, + "ATGGCA": 509, + "ATGGCT": 510, + "ATGGCC": 511, + "ATGGCG": 512, + "ATGGGA": 513, + "ATGGGT": 514, + "ATGGGC": 515, + "ATGGGG": 516, + "ACAAAA": 517, + "ACAAAT": 518, + "ACAAAC": 519, + "ACAAAG": 520, + "ACAATA": 521, + "ACAATT": 522, + "ACAATC": 523, + "ACAATG": 524, + "ACAACA": 525, + "ACAACT": 526, + "ACAACC": 527, + "ACAACG": 528, + "ACAAGA": 529, + "ACAAGT": 530, + "ACAAGC": 531, + "ACAAGG": 532, + "ACATAA": 533, + "ACATAT": 534, + "ACATAC": 535, + "ACATAG": 536, + "ACATTA": 537, + "ACATTT": 538, + "ACATTC": 539, + "ACATTG": 540, + "ACATCA": 541, + "ACATCT": 542, + "ACATCC": 543, + "ACATCG": 544, + "ACATGA": 545, + "ACATGT": 546, + "ACATGC": 547, + "ACATGG": 548, + "ACACAA": 549, + "ACACAT": 550, + "ACACAC": 551, + "ACACAG": 552, + "ACACTA": 553, + "ACACTT": 554, + "ACACTC": 555, + "ACACTG": 556, + "ACACCA": 557, + "ACACCT": 558, + "ACACCC": 559, + "ACACCG": 560, + "ACACGA": 561, + "ACACGT": 562, + "ACACGC": 563, + "ACACGG": 564, + "ACAGAA": 565, + "ACAGAT": 566, + "ACAGAC": 567, + "ACAGAG": 568, + "ACAGTA": 569, + "ACAGTT": 570, + "ACAGTC": 571, + "ACAGTG": 572, + "ACAGCA": 573, + "ACAGCT": 574, + "ACAGCC": 575, + "ACAGCG": 576, + "ACAGGA": 577, + "ACAGGT": 578, + "ACAGGC": 579, + "ACAGGG": 580, + "ACTAAA": 581, + "ACTAAT": 582, + "ACTAAC": 583, + "ACTAAG": 584, + "ACTATA": 585, + "ACTATT": 586, + "ACTATC": 587, + "ACTATG": 588, + "ACTACA": 589, + "ACTACT": 590, + "ACTACC": 591, + "ACTACG": 592, + "ACTAGA": 593, + "ACTAGT": 594, + "ACTAGC": 595, + "ACTAGG": 596, + "ACTTAA": 597, + "ACTTAT": 598, + "ACTTAC": 599, + "ACTTAG": 600, + "ACTTTA": 601, + "ACTTTT": 602, + "ACTTTC": 603, + "ACTTTG": 604, + "ACTTCA": 605, + "ACTTCT": 606, + "ACTTCC": 607, + "ACTTCG": 608, + "ACTTGA": 609, + "ACTTGT": 610, + "ACTTGC": 611, + "ACTTGG": 612, + "ACTCAA": 613, + "ACTCAT": 614, + "ACTCAC": 615, + "ACTCAG": 616, + "ACTCTA": 617, + "ACTCTT": 618, + "ACTCTC": 619, + "ACTCTG": 620, + "ACTCCA": 621, + "ACTCCT": 622, + "ACTCCC": 623, + "ACTCCG": 624, + "ACTCGA": 625, + "ACTCGT": 626, + "ACTCGC": 627, + "ACTCGG": 628, + "ACTGAA": 629, + "ACTGAT": 630, + "ACTGAC": 631, + "ACTGAG": 632, + "ACTGTA": 633, + "ACTGTT": 634, + "ACTGTC": 635, + "ACTGTG": 636, + "ACTGCA": 637, + "ACTGCT": 638, + "ACTGCC": 639, + "ACTGCG": 640, + "ACTGGA": 641, + "ACTGGT": 642, + "ACTGGC": 643, + "ACTGGG": 644, + "ACCAAA": 645, + "ACCAAT": 646, + "ACCAAC": 647, + "ACCAAG": 648, + "ACCATA": 649, + "ACCATT": 650, + "ACCATC": 651, + "ACCATG": 652, + "ACCACA": 653, + "ACCACT": 654, + "ACCACC": 655, + "ACCACG": 656, + "ACCAGA": 657, + "ACCAGT": 658, + "ACCAGC": 659, + "ACCAGG": 660, + "ACCTAA": 661, + "ACCTAT": 662, + "ACCTAC": 663, + "ACCTAG": 664, + "ACCTTA": 665, + "ACCTTT": 666, + "ACCTTC": 667, + "ACCTTG": 668, + "ACCTCA": 669, + "ACCTCT": 670, + "ACCTCC": 671, + "ACCTCG": 672, + "ACCTGA": 673, + "ACCTGT": 674, + "ACCTGC": 675, + "ACCTGG": 676, + "ACCCAA": 677, + "ACCCAT": 678, + "ACCCAC": 679, + "ACCCAG": 680, + "ACCCTA": 681, + "ACCCTT": 682, + "ACCCTC": 683, + "ACCCTG": 684, + "ACCCCA": 685, + "ACCCCT": 686, + "ACCCCC": 687, + "ACCCCG": 688, + "ACCCGA": 689, + "ACCCGT": 690, + "ACCCGC": 691, + "ACCCGG": 692, + "ACCGAA": 693, + "ACCGAT": 694, + "ACCGAC": 695, + "ACCGAG": 696, + "ACCGTA": 697, + "ACCGTT": 698, + "ACCGTC": 699, + "ACCGTG": 700, + "ACCGCA": 701, + "ACCGCT": 702, + "ACCGCC": 703, + "ACCGCG": 704, + "ACCGGA": 705, + "ACCGGT": 706, + "ACCGGC": 707, + "ACCGGG": 708, + "ACGAAA": 709, + "ACGAAT": 710, + "ACGAAC": 711, + "ACGAAG": 712, + "ACGATA": 713, + "ACGATT": 714, + "ACGATC": 715, + "ACGATG": 716, + "ACGACA": 717, + "ACGACT": 718, + "ACGACC": 719, + "ACGACG": 720, + "ACGAGA": 721, + "ACGAGT": 722, + "ACGAGC": 723, + "ACGAGG": 724, + "ACGTAA": 725, + "ACGTAT": 726, + "ACGTAC": 727, + "ACGTAG": 728, + "ACGTTA": 729, + "ACGTTT": 730, + "ACGTTC": 731, + "ACGTTG": 732, + "ACGTCA": 733, + "ACGTCT": 734, + "ACGTCC": 735, + "ACGTCG": 736, + "ACGTGA": 737, + "ACGTGT": 738, + "ACGTGC": 739, + "ACGTGG": 740, + "ACGCAA": 741, + "ACGCAT": 742, + "ACGCAC": 743, + "ACGCAG": 744, + "ACGCTA": 745, + "ACGCTT": 746, + "ACGCTC": 747, + "ACGCTG": 748, + "ACGCCA": 749, + "ACGCCT": 750, + "ACGCCC": 751, + "ACGCCG": 752, + "ACGCGA": 753, + "ACGCGT": 754, + "ACGCGC": 755, + "ACGCGG": 756, + "ACGGAA": 757, + "ACGGAT": 758, + "ACGGAC": 759, + "ACGGAG": 760, + "ACGGTA": 761, + "ACGGTT": 762, + "ACGGTC": 763, + "ACGGTG": 764, + "ACGGCA": 765, + "ACGGCT": 766, + "ACGGCC": 767, + "ACGGCG": 768, + "ACGGGA": 769, + "ACGGGT": 770, + "ACGGGC": 771, + "ACGGGG": 772, + "AGAAAA": 773, + "AGAAAT": 774, + "AGAAAC": 775, + "AGAAAG": 776, + "AGAATA": 777, + "AGAATT": 778, + "AGAATC": 779, + "AGAATG": 780, + "AGAACA": 781, + "AGAACT": 782, + "AGAACC": 783, + "AGAACG": 784, + "AGAAGA": 785, + "AGAAGT": 786, + "AGAAGC": 787, + "AGAAGG": 788, + "AGATAA": 789, + "AGATAT": 790, + "AGATAC": 791, + "AGATAG": 792, + "AGATTA": 793, + "AGATTT": 794, + "AGATTC": 795, + "AGATTG": 796, + "AGATCA": 797, + "AGATCT": 798, + "AGATCC": 799, + "AGATCG": 800, + "AGATGA": 801, + "AGATGT": 802, + "AGATGC": 803, + "AGATGG": 804, + "AGACAA": 805, + "AGACAT": 806, + "AGACAC": 807, + "AGACAG": 808, + "AGACTA": 809, + "AGACTT": 810, + "AGACTC": 811, + "AGACTG": 812, + "AGACCA": 813, + "AGACCT": 814, + "AGACCC": 815, + "AGACCG": 816, + "AGACGA": 817, + "AGACGT": 818, + "AGACGC": 819, + "AGACGG": 820, + "AGAGAA": 821, + "AGAGAT": 822, + "AGAGAC": 823, + "AGAGAG": 824, + "AGAGTA": 825, + "AGAGTT": 826, + "AGAGTC": 827, + "AGAGTG": 828, + "AGAGCA": 829, + "AGAGCT": 830, + "AGAGCC": 831, + "AGAGCG": 832, + "AGAGGA": 833, + "AGAGGT": 834, + "AGAGGC": 835, + "AGAGGG": 836, + "AGTAAA": 837, + "AGTAAT": 838, + "AGTAAC": 839, + "AGTAAG": 840, + "AGTATA": 841, + "AGTATT": 842, + "AGTATC": 843, + "AGTATG": 844, + "AGTACA": 845, + "AGTACT": 846, + "AGTACC": 847, + "AGTACG": 848, + "AGTAGA": 849, + "AGTAGT": 850, + "AGTAGC": 851, + "AGTAGG": 852, + "AGTTAA": 853, + "AGTTAT": 854, + "AGTTAC": 855, + "AGTTAG": 856, + "AGTTTA": 857, + "AGTTTT": 858, + "AGTTTC": 859, + "AGTTTG": 860, + "AGTTCA": 861, + "AGTTCT": 862, + "AGTTCC": 863, + "AGTTCG": 864, + "AGTTGA": 865, + "AGTTGT": 866, + "AGTTGC": 867, + "AGTTGG": 868, + "AGTCAA": 869, + "AGTCAT": 870, + "AGTCAC": 871, + "AGTCAG": 872, + "AGTCTA": 873, + "AGTCTT": 874, + "AGTCTC": 875, + "AGTCTG": 876, + "AGTCCA": 877, + "AGTCCT": 878, + "AGTCCC": 879, + "AGTCCG": 880, + "AGTCGA": 881, + "AGTCGT": 882, + "AGTCGC": 883, + "AGTCGG": 884, + "AGTGAA": 885, + "AGTGAT": 886, + "AGTGAC": 887, + "AGTGAG": 888, + "AGTGTA": 889, + "AGTGTT": 890, + "AGTGTC": 891, + "AGTGTG": 892, + "AGTGCA": 893, + "AGTGCT": 894, + "AGTGCC": 895, + "AGTGCG": 896, + "AGTGGA": 897, + "AGTGGT": 898, + "AGTGGC": 899, + "AGTGGG": 900, + "AGCAAA": 901, + "AGCAAT": 902, + "AGCAAC": 903, + "AGCAAG": 904, + "AGCATA": 905, + "AGCATT": 906, + "AGCATC": 907, + "AGCATG": 908, + "AGCACA": 909, + "AGCACT": 910, + "AGCACC": 911, + "AGCACG": 912, + "AGCAGA": 913, + "AGCAGT": 914, + "AGCAGC": 915, + "AGCAGG": 916, + "AGCTAA": 917, + "AGCTAT": 918, + "AGCTAC": 919, + "AGCTAG": 920, + "AGCTTA": 921, + "AGCTTT": 922, + "AGCTTC": 923, + "AGCTTG": 924, + "AGCTCA": 925, + "AGCTCT": 926, + "AGCTCC": 927, + "AGCTCG": 928, + "AGCTGA": 929, + "AGCTGT": 930, + "AGCTGC": 931, + "AGCTGG": 932, + "AGCCAA": 933, + "AGCCAT": 934, + "AGCCAC": 935, + "AGCCAG": 936, + "AGCCTA": 937, + "AGCCTT": 938, + "AGCCTC": 939, + "AGCCTG": 940, + "AGCCCA": 941, + "AGCCCT": 942, + "AGCCCC": 943, + "AGCCCG": 944, + "AGCCGA": 945, + "AGCCGT": 946, + "AGCCGC": 947, + "AGCCGG": 948, + "AGCGAA": 949, + "AGCGAT": 950, + "AGCGAC": 951, + "AGCGAG": 952, + "AGCGTA": 953, + "AGCGTT": 954, + "AGCGTC": 955, + "AGCGTG": 956, + "AGCGCA": 957, + "AGCGCT": 958, + "AGCGCC": 959, + "AGCGCG": 960, + "AGCGGA": 961, + "AGCGGT": 962, + "AGCGGC": 963, + "AGCGGG": 964, + "AGGAAA": 965, + "AGGAAT": 966, + "AGGAAC": 967, + "AGGAAG": 968, + "AGGATA": 969, + "AGGATT": 970, + "AGGATC": 971, + "AGGATG": 972, + "AGGACA": 973, + "AGGACT": 974, + "AGGACC": 975, + "AGGACG": 976, + "AGGAGA": 977, + "AGGAGT": 978, + "AGGAGC": 979, + "AGGAGG": 980, + "AGGTAA": 981, + "AGGTAT": 982, + "AGGTAC": 983, + "AGGTAG": 984, + "AGGTTA": 985, + "AGGTTT": 986, + "AGGTTC": 987, + "AGGTTG": 988, + "AGGTCA": 989, + "AGGTCT": 990, + "AGGTCC": 991, + "AGGTCG": 992, + "AGGTGA": 993, + "AGGTGT": 994, + "AGGTGC": 995, + "AGGTGG": 996, + "AGGCAA": 997, + "AGGCAT": 998, + "AGGCAC": 999, + "AGGCAG": 1000, + "AGGCTA": 1001, + "AGGCTT": 1002, + "AGGCTC": 1003, + "AGGCTG": 1004, + "AGGCCA": 1005, + "AGGCCT": 1006, + "AGGCCC": 1007, + "AGGCCG": 1008, + "AGGCGA": 1009, + "AGGCGT": 1010, + "AGGCGC": 1011, + "AGGCGG": 1012, + "AGGGAA": 1013, + "AGGGAT": 1014, + "AGGGAC": 1015, + "AGGGAG": 1016, + "AGGGTA": 1017, + "AGGGTT": 1018, + "AGGGTC": 1019, + "AGGGTG": 1020, + "AGGGCA": 1021, + "AGGGCT": 1022, + "AGGGCC": 1023, + "AGGGCG": 1024, + "AGGGGA": 1025, + "AGGGGT": 1026, + "AGGGGC": 1027, + "AGGGGG": 1028, + "TAAAAA": 1029, + "TAAAAT": 1030, + "TAAAAC": 1031, + "TAAAAG": 1032, + "TAAATA": 1033, + "TAAATT": 1034, + "TAAATC": 1035, + "TAAATG": 1036, + "TAAACA": 1037, + "TAAACT": 1038, + "TAAACC": 1039, + "TAAACG": 1040, + "TAAAGA": 1041, + "TAAAGT": 1042, + "TAAAGC": 1043, + "TAAAGG": 1044, + "TAATAA": 1045, + "TAATAT": 1046, + "TAATAC": 1047, + "TAATAG": 1048, + "TAATTA": 1049, + "TAATTT": 1050, + "TAATTC": 1051, + "TAATTG": 1052, + "TAATCA": 1053, + "TAATCT": 1054, + "TAATCC": 1055, + "TAATCG": 1056, + "TAATGA": 1057, + "TAATGT": 1058, + "TAATGC": 1059, + "TAATGG": 1060, + "TAACAA": 1061, + "TAACAT": 1062, + "TAACAC": 1063, + "TAACAG": 1064, + "TAACTA": 1065, + "TAACTT": 1066, + "TAACTC": 1067, + "TAACTG": 1068, + "TAACCA": 1069, + "TAACCT": 1070, + "TAACCC": 1071, + "TAACCG": 1072, + "TAACGA": 1073, + "TAACGT": 1074, + "TAACGC": 1075, + "TAACGG": 1076, + "TAAGAA": 1077, + "TAAGAT": 1078, + "TAAGAC": 1079, + "TAAGAG": 1080, + "TAAGTA": 1081, + "TAAGTT": 1082, + "TAAGTC": 1083, + "TAAGTG": 1084, + "TAAGCA": 1085, + "TAAGCT": 1086, + "TAAGCC": 1087, + "TAAGCG": 1088, + "TAAGGA": 1089, + "TAAGGT": 1090, + "TAAGGC": 1091, + "TAAGGG": 1092, + "TATAAA": 1093, + "TATAAT": 1094, + "TATAAC": 1095, + "TATAAG": 1096, + "TATATA": 1097, + "TATATT": 1098, + "TATATC": 1099, + "TATATG": 1100, + "TATACA": 1101, + "TATACT": 1102, + "TATACC": 1103, + "TATACG": 1104, + "TATAGA": 1105, + "TATAGT": 1106, + "TATAGC": 1107, + "TATAGG": 1108, + "TATTAA": 1109, + "TATTAT": 1110, + "TATTAC": 1111, + "TATTAG": 1112, + "TATTTA": 1113, + "TATTTT": 1114, + "TATTTC": 1115, + "TATTTG": 1116, + "TATTCA": 1117, + "TATTCT": 1118, + "TATTCC": 1119, + "TATTCG": 1120, + "TATTGA": 1121, + "TATTGT": 1122, + "TATTGC": 1123, + "TATTGG": 1124, + "TATCAA": 1125, + "TATCAT": 1126, + "TATCAC": 1127, + "TATCAG": 1128, + "TATCTA": 1129, + "TATCTT": 1130, + "TATCTC": 1131, + "TATCTG": 1132, + "TATCCA": 1133, + "TATCCT": 1134, + "TATCCC": 1135, + "TATCCG": 1136, + "TATCGA": 1137, + "TATCGT": 1138, + "TATCGC": 1139, + "TATCGG": 1140, + "TATGAA": 1141, + "TATGAT": 1142, + "TATGAC": 1143, + "TATGAG": 1144, + "TATGTA": 1145, + "TATGTT": 1146, + "TATGTC": 1147, + "TATGTG": 1148, + "TATGCA": 1149, + "TATGCT": 1150, + "TATGCC": 1151, + "TATGCG": 1152, + "TATGGA": 1153, + "TATGGT": 1154, + "TATGGC": 1155, + "TATGGG": 1156, + "TACAAA": 1157, + "TACAAT": 1158, + "TACAAC": 1159, + "TACAAG": 1160, + "TACATA": 1161, + "TACATT": 1162, + "TACATC": 1163, + "TACATG": 1164, + "TACACA": 1165, + "TACACT": 1166, + "TACACC": 1167, + "TACACG": 1168, + "TACAGA": 1169, + "TACAGT": 1170, + "TACAGC": 1171, + "TACAGG": 1172, + "TACTAA": 1173, + "TACTAT": 1174, + "TACTAC": 1175, + "TACTAG": 1176, + "TACTTA": 1177, + "TACTTT": 1178, + "TACTTC": 1179, + "TACTTG": 1180, + "TACTCA": 1181, + "TACTCT": 1182, + "TACTCC": 1183, + "TACTCG": 1184, + "TACTGA": 1185, + "TACTGT": 1186, + "TACTGC": 1187, + "TACTGG": 1188, + "TACCAA": 1189, + "TACCAT": 1190, + "TACCAC": 1191, + "TACCAG": 1192, + "TACCTA": 1193, + "TACCTT": 1194, + "TACCTC": 1195, + "TACCTG": 1196, + "TACCCA": 1197, + "TACCCT": 1198, + "TACCCC": 1199, + "TACCCG": 1200, + "TACCGA": 1201, + "TACCGT": 1202, + "TACCGC": 1203, + "TACCGG": 1204, + "TACGAA": 1205, + "TACGAT": 1206, + "TACGAC": 1207, + "TACGAG": 1208, + "TACGTA": 1209, + "TACGTT": 1210, + "TACGTC": 1211, + "TACGTG": 1212, + "TACGCA": 1213, + "TACGCT": 1214, + "TACGCC": 1215, + "TACGCG": 1216, + "TACGGA": 1217, + "TACGGT": 1218, + "TACGGC": 1219, + "TACGGG": 1220, + "TAGAAA": 1221, + "TAGAAT": 1222, + "TAGAAC": 1223, + "TAGAAG": 1224, + "TAGATA": 1225, + "TAGATT": 1226, + "TAGATC": 1227, + "TAGATG": 1228, + "TAGACA": 1229, + "TAGACT": 1230, + "TAGACC": 1231, + "TAGACG": 1232, + "TAGAGA": 1233, + "TAGAGT": 1234, + "TAGAGC": 1235, + "TAGAGG": 1236, + "TAGTAA": 1237, + "TAGTAT": 1238, + "TAGTAC": 1239, + "TAGTAG": 1240, + "TAGTTA": 1241, + "TAGTTT": 1242, + "TAGTTC": 1243, + "TAGTTG": 1244, + "TAGTCA": 1245, + "TAGTCT": 1246, + "TAGTCC": 1247, + "TAGTCG": 1248, + "TAGTGA": 1249, + "TAGTGT": 1250, + "TAGTGC": 1251, + "TAGTGG": 1252, + "TAGCAA": 1253, + "TAGCAT": 1254, + "TAGCAC": 1255, + "TAGCAG": 1256, + "TAGCTA": 1257, + "TAGCTT": 1258, + "TAGCTC": 1259, + "TAGCTG": 1260, + "TAGCCA": 1261, + "TAGCCT": 1262, + "TAGCCC": 1263, + "TAGCCG": 1264, + "TAGCGA": 1265, + "TAGCGT": 1266, + "TAGCGC": 1267, + "TAGCGG": 1268, + "TAGGAA": 1269, + "TAGGAT": 1270, + "TAGGAC": 1271, + "TAGGAG": 1272, + "TAGGTA": 1273, + "TAGGTT": 1274, + "TAGGTC": 1275, + "TAGGTG": 1276, + "TAGGCA": 1277, + "TAGGCT": 1278, + "TAGGCC": 1279, + "TAGGCG": 1280, + "TAGGGA": 1281, + "TAGGGT": 1282, + "TAGGGC": 1283, + "TAGGGG": 1284, + "TTAAAA": 1285, + "TTAAAT": 1286, + "TTAAAC": 1287, + "TTAAAG": 1288, + "TTAATA": 1289, + "TTAATT": 1290, + "TTAATC": 1291, + "TTAATG": 1292, + "TTAACA": 1293, + "TTAACT": 1294, + "TTAACC": 1295, + "TTAACG": 1296, + "TTAAGA": 1297, + "TTAAGT": 1298, + "TTAAGC": 1299, + "TTAAGG": 1300, + "TTATAA": 1301, + "TTATAT": 1302, + "TTATAC": 1303, + "TTATAG": 1304, + "TTATTA": 1305, + "TTATTT": 1306, + "TTATTC": 1307, + "TTATTG": 1308, + "TTATCA": 1309, + "TTATCT": 1310, + "TTATCC": 1311, + "TTATCG": 1312, + "TTATGA": 1313, + "TTATGT": 1314, + "TTATGC": 1315, + "TTATGG": 1316, + "TTACAA": 1317, + "TTACAT": 1318, + "TTACAC": 1319, + "TTACAG": 1320, + "TTACTA": 1321, + "TTACTT": 1322, + "TTACTC": 1323, + "TTACTG": 1324, + "TTACCA": 1325, + "TTACCT": 1326, + "TTACCC": 1327, + "TTACCG": 1328, + "TTACGA": 1329, + "TTACGT": 1330, + "TTACGC": 1331, + "TTACGG": 1332, + "TTAGAA": 1333, + "TTAGAT": 1334, + "TTAGAC": 1335, + "TTAGAG": 1336, + "TTAGTA": 1337, + "TTAGTT": 1338, + "TTAGTC": 1339, + "TTAGTG": 1340, + "TTAGCA": 1341, + "TTAGCT": 1342, + "TTAGCC": 1343, + "TTAGCG": 1344, + "TTAGGA": 1345, + "TTAGGT": 1346, + "TTAGGC": 1347, + "TTAGGG": 1348, + "TTTAAA": 1349, + "TTTAAT": 1350, + "TTTAAC": 1351, + "TTTAAG": 1352, + "TTTATA": 1353, + "TTTATT": 1354, + "TTTATC": 1355, + "TTTATG": 1356, + "TTTACA": 1357, + "TTTACT": 1358, + "TTTACC": 1359, + "TTTACG": 1360, + "TTTAGA": 1361, + "TTTAGT": 1362, + "TTTAGC": 1363, + "TTTAGG": 1364, + "TTTTAA": 1365, + "TTTTAT": 1366, + "TTTTAC": 1367, + "TTTTAG": 1368, + "TTTTTA": 1369, + "TTTTTT": 1370, + "TTTTTC": 1371, + "TTTTTG": 1372, + "TTTTCA": 1373, + "TTTTCT": 1374, + "TTTTCC": 1375, + "TTTTCG": 1376, + "TTTTGA": 1377, + "TTTTGT": 1378, + "TTTTGC": 1379, + "TTTTGG": 1380, + "TTTCAA": 1381, + "TTTCAT": 1382, + "TTTCAC": 1383, + "TTTCAG": 1384, + "TTTCTA": 1385, + "TTTCTT": 1386, + "TTTCTC": 1387, + "TTTCTG": 1388, + "TTTCCA": 1389, + "TTTCCT": 1390, + "TTTCCC": 1391, + "TTTCCG": 1392, + "TTTCGA": 1393, + "TTTCGT": 1394, + "TTTCGC": 1395, + "TTTCGG": 1396, + "TTTGAA": 1397, + "TTTGAT": 1398, + "TTTGAC": 1399, + "TTTGAG": 1400, + "TTTGTA": 1401, + "TTTGTT": 1402, + "TTTGTC": 1403, + "TTTGTG": 1404, + "TTTGCA": 1405, + "TTTGCT": 1406, + "TTTGCC": 1407, + "TTTGCG": 1408, + "TTTGGA": 1409, + "TTTGGT": 1410, + "TTTGGC": 1411, + "TTTGGG": 1412, + "TTCAAA": 1413, + "TTCAAT": 1414, + "TTCAAC": 1415, + "TTCAAG": 1416, + "TTCATA": 1417, + "TTCATT": 1418, + "TTCATC": 1419, + "TTCATG": 1420, + "TTCACA": 1421, + "TTCACT": 1422, + "TTCACC": 1423, + "TTCACG": 1424, + "TTCAGA": 1425, + "TTCAGT": 1426, + "TTCAGC": 1427, + "TTCAGG": 1428, + "TTCTAA": 1429, + "TTCTAT": 1430, + "TTCTAC": 1431, + "TTCTAG": 1432, + "TTCTTA": 1433, + "TTCTTT": 1434, + "TTCTTC": 1435, + "TTCTTG": 1436, + "TTCTCA": 1437, + "TTCTCT": 1438, + "TTCTCC": 1439, + "TTCTCG": 1440, + "TTCTGA": 1441, + "TTCTGT": 1442, + "TTCTGC": 1443, + "TTCTGG": 1444, + "TTCCAA": 1445, + "TTCCAT": 1446, + "TTCCAC": 1447, + "TTCCAG": 1448, + "TTCCTA": 1449, + "TTCCTT": 1450, + "TTCCTC": 1451, + "TTCCTG": 1452, + "TTCCCA": 1453, + "TTCCCT": 1454, + "TTCCCC": 1455, + "TTCCCG": 1456, + "TTCCGA": 1457, + "TTCCGT": 1458, + "TTCCGC": 1459, + "TTCCGG": 1460, + "TTCGAA": 1461, + "TTCGAT": 1462, + "TTCGAC": 1463, + "TTCGAG": 1464, + "TTCGTA": 1465, + "TTCGTT": 1466, + "TTCGTC": 1467, + "TTCGTG": 1468, + "TTCGCA": 1469, + "TTCGCT": 1470, + "TTCGCC": 1471, + "TTCGCG": 1472, + "TTCGGA": 1473, + "TTCGGT": 1474, + "TTCGGC": 1475, + "TTCGGG": 1476, + "TTGAAA": 1477, + "TTGAAT": 1478, + "TTGAAC": 1479, + "TTGAAG": 1480, + "TTGATA": 1481, + "TTGATT": 1482, + "TTGATC": 1483, + "TTGATG": 1484, + "TTGACA": 1485, + "TTGACT": 1486, + "TTGACC": 1487, + "TTGACG": 1488, + "TTGAGA": 1489, + "TTGAGT": 1490, + "TTGAGC": 1491, + "TTGAGG": 1492, + "TTGTAA": 1493, + "TTGTAT": 1494, + "TTGTAC": 1495, + "TTGTAG": 1496, + "TTGTTA": 1497, + "TTGTTT": 1498, + "TTGTTC": 1499, + "TTGTTG": 1500, + "TTGTCA": 1501, + "TTGTCT": 1502, + "TTGTCC": 1503, + "TTGTCG": 1504, + "TTGTGA": 1505, + "TTGTGT": 1506, + "TTGTGC": 1507, + "TTGTGG": 1508, + "TTGCAA": 1509, + "TTGCAT": 1510, + "TTGCAC": 1511, + "TTGCAG": 1512, + "TTGCTA": 1513, + "TTGCTT": 1514, + "TTGCTC": 1515, + "TTGCTG": 1516, + "TTGCCA": 1517, + "TTGCCT": 1518, + "TTGCCC": 1519, + "TTGCCG": 1520, + "TTGCGA": 1521, + "TTGCGT": 1522, + "TTGCGC": 1523, + "TTGCGG": 1524, + "TTGGAA": 1525, + "TTGGAT": 1526, + "TTGGAC": 1527, + "TTGGAG": 1528, + "TTGGTA": 1529, + "TTGGTT": 1530, + "TTGGTC": 1531, + "TTGGTG": 1532, + "TTGGCA": 1533, + "TTGGCT": 1534, + "TTGGCC": 1535, + "TTGGCG": 1536, + "TTGGGA": 1537, + "TTGGGT": 1538, + "TTGGGC": 1539, + "TTGGGG": 1540, + "TCAAAA": 1541, + "TCAAAT": 1542, + "TCAAAC": 1543, + "TCAAAG": 1544, + "TCAATA": 1545, + "TCAATT": 1546, + "TCAATC": 1547, + "TCAATG": 1548, + "TCAACA": 1549, + "TCAACT": 1550, + "TCAACC": 1551, + "TCAACG": 1552, + "TCAAGA": 1553, + "TCAAGT": 1554, + "TCAAGC": 1555, + "TCAAGG": 1556, + "TCATAA": 1557, + "TCATAT": 1558, + "TCATAC": 1559, + "TCATAG": 1560, + "TCATTA": 1561, + "TCATTT": 1562, + "TCATTC": 1563, + "TCATTG": 1564, + "TCATCA": 1565, + "TCATCT": 1566, + "TCATCC": 1567, + "TCATCG": 1568, + "TCATGA": 1569, + "TCATGT": 1570, + "TCATGC": 1571, + "TCATGG": 1572, + "TCACAA": 1573, + "TCACAT": 1574, + "TCACAC": 1575, + "TCACAG": 1576, + "TCACTA": 1577, + "TCACTT": 1578, + "TCACTC": 1579, + "TCACTG": 1580, + "TCACCA": 1581, + "TCACCT": 1582, + "TCACCC": 1583, + "TCACCG": 1584, + "TCACGA": 1585, + "TCACGT": 1586, + "TCACGC": 1587, + "TCACGG": 1588, + "TCAGAA": 1589, + "TCAGAT": 1590, + "TCAGAC": 1591, + "TCAGAG": 1592, + "TCAGTA": 1593, + "TCAGTT": 1594, + "TCAGTC": 1595, + "TCAGTG": 1596, + "TCAGCA": 1597, + "TCAGCT": 1598, + "TCAGCC": 1599, + "TCAGCG": 1600, + "TCAGGA": 1601, + "TCAGGT": 1602, + "TCAGGC": 1603, + "TCAGGG": 1604, + "TCTAAA": 1605, + "TCTAAT": 1606, + "TCTAAC": 1607, + "TCTAAG": 1608, + "TCTATA": 1609, + "TCTATT": 1610, + "TCTATC": 1611, + "TCTATG": 1612, + "TCTACA": 1613, + "TCTACT": 1614, + "TCTACC": 1615, + "TCTACG": 1616, + "TCTAGA": 1617, + "TCTAGT": 1618, + "TCTAGC": 1619, + "TCTAGG": 1620, + "TCTTAA": 1621, + "TCTTAT": 1622, + "TCTTAC": 1623, + "TCTTAG": 1624, + "TCTTTA": 1625, + "TCTTTT": 1626, + "TCTTTC": 1627, + "TCTTTG": 1628, + "TCTTCA": 1629, + "TCTTCT": 1630, + "TCTTCC": 1631, + "TCTTCG": 1632, + "TCTTGA": 1633, + "TCTTGT": 1634, + "TCTTGC": 1635, + "TCTTGG": 1636, + "TCTCAA": 1637, + "TCTCAT": 1638, + "TCTCAC": 1639, + "TCTCAG": 1640, + "TCTCTA": 1641, + "TCTCTT": 1642, + "TCTCTC": 1643, + "TCTCTG": 1644, + "TCTCCA": 1645, + "TCTCCT": 1646, + "TCTCCC": 1647, + "TCTCCG": 1648, + "TCTCGA": 1649, + "TCTCGT": 1650, + "TCTCGC": 1651, + "TCTCGG": 1652, + "TCTGAA": 1653, + "TCTGAT": 1654, + "TCTGAC": 1655, + "TCTGAG": 1656, + "TCTGTA": 1657, + "TCTGTT": 1658, + "TCTGTC": 1659, + "TCTGTG": 1660, + "TCTGCA": 1661, + "TCTGCT": 1662, + "TCTGCC": 1663, + "TCTGCG": 1664, + "TCTGGA": 1665, + "TCTGGT": 1666, + "TCTGGC": 1667, + "TCTGGG": 1668, + "TCCAAA": 1669, + "TCCAAT": 1670, + "TCCAAC": 1671, + "TCCAAG": 1672, + "TCCATA": 1673, + "TCCATT": 1674, + "TCCATC": 1675, + "TCCATG": 1676, + "TCCACA": 1677, + "TCCACT": 1678, + "TCCACC": 1679, + "TCCACG": 1680, + "TCCAGA": 1681, + "TCCAGT": 1682, + "TCCAGC": 1683, + "TCCAGG": 1684, + "TCCTAA": 1685, + "TCCTAT": 1686, + "TCCTAC": 1687, + "TCCTAG": 1688, + "TCCTTA": 1689, + "TCCTTT": 1690, + "TCCTTC": 1691, + "TCCTTG": 1692, + "TCCTCA": 1693, + "TCCTCT": 1694, + "TCCTCC": 1695, + "TCCTCG": 1696, + "TCCTGA": 1697, + "TCCTGT": 1698, + "TCCTGC": 1699, + "TCCTGG": 1700, + "TCCCAA": 1701, + "TCCCAT": 1702, + "TCCCAC": 1703, + "TCCCAG": 1704, + "TCCCTA": 1705, + "TCCCTT": 1706, + "TCCCTC": 1707, + "TCCCTG": 1708, + "TCCCCA": 1709, + "TCCCCT": 1710, + "TCCCCC": 1711, + "TCCCCG": 1712, + "TCCCGA": 1713, + "TCCCGT": 1714, + "TCCCGC": 1715, + "TCCCGG": 1716, + "TCCGAA": 1717, + "TCCGAT": 1718, + "TCCGAC": 1719, + "TCCGAG": 1720, + "TCCGTA": 1721, + "TCCGTT": 1722, + "TCCGTC": 1723, + "TCCGTG": 1724, + "TCCGCA": 1725, + "TCCGCT": 1726, + "TCCGCC": 1727, + "TCCGCG": 1728, + "TCCGGA": 1729, + "TCCGGT": 1730, + "TCCGGC": 1731, + "TCCGGG": 1732, + "TCGAAA": 1733, + "TCGAAT": 1734, + "TCGAAC": 1735, + "TCGAAG": 1736, + "TCGATA": 1737, + "TCGATT": 1738, + "TCGATC": 1739, + "TCGATG": 1740, + "TCGACA": 1741, + "TCGACT": 1742, + "TCGACC": 1743, + "TCGACG": 1744, + "TCGAGA": 1745, + "TCGAGT": 1746, + "TCGAGC": 1747, + "TCGAGG": 1748, + "TCGTAA": 1749, + "TCGTAT": 1750, + "TCGTAC": 1751, + "TCGTAG": 1752, + "TCGTTA": 1753, + "TCGTTT": 1754, + "TCGTTC": 1755, + "TCGTTG": 1756, + "TCGTCA": 1757, + "TCGTCT": 1758, + "TCGTCC": 1759, + "TCGTCG": 1760, + "TCGTGA": 1761, + "TCGTGT": 1762, + "TCGTGC": 1763, + "TCGTGG": 1764, + "TCGCAA": 1765, + "TCGCAT": 1766, + "TCGCAC": 1767, + "TCGCAG": 1768, + "TCGCTA": 1769, + "TCGCTT": 1770, + "TCGCTC": 1771, + "TCGCTG": 1772, + "TCGCCA": 1773, + "TCGCCT": 1774, + "TCGCCC": 1775, + "TCGCCG": 1776, + "TCGCGA": 1777, + "TCGCGT": 1778, + "TCGCGC": 1779, + "TCGCGG": 1780, + "TCGGAA": 1781, + "TCGGAT": 1782, + "TCGGAC": 1783, + "TCGGAG": 1784, + "TCGGTA": 1785, + "TCGGTT": 1786, + "TCGGTC": 1787, + "TCGGTG": 1788, + "TCGGCA": 1789, + "TCGGCT": 1790, + "TCGGCC": 1791, + "TCGGCG": 1792, + "TCGGGA": 1793, + "TCGGGT": 1794, + "TCGGGC": 1795, + "TCGGGG": 1796, + "TGAAAA": 1797, + "TGAAAT": 1798, + "TGAAAC": 1799, + "TGAAAG": 1800, + "TGAATA": 1801, + "TGAATT": 1802, + "TGAATC": 1803, + "TGAATG": 1804, + "TGAACA": 1805, + "TGAACT": 1806, + "TGAACC": 1807, + "TGAACG": 1808, + "TGAAGA": 1809, + "TGAAGT": 1810, + "TGAAGC": 1811, + "TGAAGG": 1812, + "TGATAA": 1813, + "TGATAT": 1814, + "TGATAC": 1815, + "TGATAG": 1816, + "TGATTA": 1817, + "TGATTT": 1818, + "TGATTC": 1819, + "TGATTG": 1820, + "TGATCA": 1821, + "TGATCT": 1822, + "TGATCC": 1823, + "TGATCG": 1824, + "TGATGA": 1825, + "TGATGT": 1826, + "TGATGC": 1827, + "TGATGG": 1828, + "TGACAA": 1829, + "TGACAT": 1830, + "TGACAC": 1831, + "TGACAG": 1832, + "TGACTA": 1833, + "TGACTT": 1834, + "TGACTC": 1835, + "TGACTG": 1836, + "TGACCA": 1837, + "TGACCT": 1838, + "TGACCC": 1839, + "TGACCG": 1840, + "TGACGA": 1841, + "TGACGT": 1842, + "TGACGC": 1843, + "TGACGG": 1844, + "TGAGAA": 1845, + "TGAGAT": 1846, + "TGAGAC": 1847, + "TGAGAG": 1848, + "TGAGTA": 1849, + "TGAGTT": 1850, + "TGAGTC": 1851, + "TGAGTG": 1852, + "TGAGCA": 1853, + "TGAGCT": 1854, + "TGAGCC": 1855, + "TGAGCG": 1856, + "TGAGGA": 1857, + "TGAGGT": 1858, + "TGAGGC": 1859, + "TGAGGG": 1860, + "TGTAAA": 1861, + "TGTAAT": 1862, + "TGTAAC": 1863, + "TGTAAG": 1864, + "TGTATA": 1865, + "TGTATT": 1866, + "TGTATC": 1867, + "TGTATG": 1868, + "TGTACA": 1869, + "TGTACT": 1870, + "TGTACC": 1871, + "TGTACG": 1872, + "TGTAGA": 1873, + "TGTAGT": 1874, + "TGTAGC": 1875, + "TGTAGG": 1876, + "TGTTAA": 1877, + "TGTTAT": 1878, + "TGTTAC": 1879, + "TGTTAG": 1880, + "TGTTTA": 1881, + "TGTTTT": 1882, + "TGTTTC": 1883, + "TGTTTG": 1884, + "TGTTCA": 1885, + "TGTTCT": 1886, + "TGTTCC": 1887, + "TGTTCG": 1888, + "TGTTGA": 1889, + "TGTTGT": 1890, + "TGTTGC": 1891, + "TGTTGG": 1892, + "TGTCAA": 1893, + "TGTCAT": 1894, + "TGTCAC": 1895, + "TGTCAG": 1896, + "TGTCTA": 1897, + "TGTCTT": 1898, + "TGTCTC": 1899, + "TGTCTG": 1900, + "TGTCCA": 1901, + "TGTCCT": 1902, + "TGTCCC": 1903, + "TGTCCG": 1904, + "TGTCGA": 1905, + "TGTCGT": 1906, + "TGTCGC": 1907, + "TGTCGG": 1908, + "TGTGAA": 1909, + "TGTGAT": 1910, + "TGTGAC": 1911, + "TGTGAG": 1912, + "TGTGTA": 1913, + "TGTGTT": 1914, + "TGTGTC": 1915, + "TGTGTG": 1916, + "TGTGCA": 1917, + "TGTGCT": 1918, + "TGTGCC": 1919, + "TGTGCG": 1920, + "TGTGGA": 1921, + "TGTGGT": 1922, + "TGTGGC": 1923, + "TGTGGG": 1924, + "TGCAAA": 1925, + "TGCAAT": 1926, + "TGCAAC": 1927, + "TGCAAG": 1928, + "TGCATA": 1929, + "TGCATT": 1930, + "TGCATC": 1931, + "TGCATG": 1932, + "TGCACA": 1933, + "TGCACT": 1934, + "TGCACC": 1935, + "TGCACG": 1936, + "TGCAGA": 1937, + "TGCAGT": 1938, + "TGCAGC": 1939, + "TGCAGG": 1940, + "TGCTAA": 1941, + "TGCTAT": 1942, + "TGCTAC": 1943, + "TGCTAG": 1944, + "TGCTTA": 1945, + "TGCTTT": 1946, + "TGCTTC": 1947, + "TGCTTG": 1948, + "TGCTCA": 1949, + "TGCTCT": 1950, + "TGCTCC": 1951, + "TGCTCG": 1952, + "TGCTGA": 1953, + "TGCTGT": 1954, + "TGCTGC": 1955, + "TGCTGG": 1956, + "TGCCAA": 1957, + "TGCCAT": 1958, + "TGCCAC": 1959, + "TGCCAG": 1960, + "TGCCTA": 1961, + "TGCCTT": 1962, + "TGCCTC": 1963, + "TGCCTG": 1964, + "TGCCCA": 1965, + "TGCCCT": 1966, + "TGCCCC": 1967, + "TGCCCG": 1968, + "TGCCGA": 1969, + "TGCCGT": 1970, + "TGCCGC": 1971, + "TGCCGG": 1972, + "TGCGAA": 1973, + "TGCGAT": 1974, + "TGCGAC": 1975, + "TGCGAG": 1976, + "TGCGTA": 1977, + "TGCGTT": 1978, + "TGCGTC": 1979, + "TGCGTG": 1980, + "TGCGCA": 1981, + "TGCGCT": 1982, + "TGCGCC": 1983, + "TGCGCG": 1984, + "TGCGGA": 1985, + "TGCGGT": 1986, + "TGCGGC": 1987, + "TGCGGG": 1988, + "TGGAAA": 1989, + "TGGAAT": 1990, + "TGGAAC": 1991, + "TGGAAG": 1992, + "TGGATA": 1993, + "TGGATT": 1994, + "TGGATC": 1995, + "TGGATG": 1996, + "TGGACA": 1997, + "TGGACT": 1998, + "TGGACC": 1999, + "TGGACG": 2000, + "TGGAGA": 2001, + "TGGAGT": 2002, + "TGGAGC": 2003, + "TGGAGG": 2004, + "TGGTAA": 2005, + "TGGTAT": 2006, + "TGGTAC": 2007, + "TGGTAG": 2008, + "TGGTTA": 2009, + "TGGTTT": 2010, + "TGGTTC": 2011, + "TGGTTG": 2012, + "TGGTCA": 2013, + "TGGTCT": 2014, + "TGGTCC": 2015, + "TGGTCG": 2016, + "TGGTGA": 2017, + "TGGTGT": 2018, + "TGGTGC": 2019, + "TGGTGG": 2020, + "TGGCAA": 2021, + "TGGCAT": 2022, + "TGGCAC": 2023, + "TGGCAG": 2024, + "TGGCTA": 2025, + "TGGCTT": 2026, + "TGGCTC": 2027, + "TGGCTG": 2028, + "TGGCCA": 2029, + "TGGCCT": 2030, + "TGGCCC": 2031, + "TGGCCG": 2032, + "TGGCGA": 2033, + "TGGCGT": 2034, + "TGGCGC": 2035, + "TGGCGG": 2036, + "TGGGAA": 2037, + "TGGGAT": 2038, + "TGGGAC": 2039, + "TGGGAG": 2040, + "TGGGTA": 2041, + "TGGGTT": 2042, + "TGGGTC": 2043, + "TGGGTG": 2044, + "TGGGCA": 2045, + "TGGGCT": 2046, + "TGGGCC": 2047, + "TGGGCG": 2048, + "TGGGGA": 2049, + "TGGGGT": 2050, + "TGGGGC": 2051, + "TGGGGG": 2052, + "CAAAAA": 2053, + "CAAAAT": 2054, + "CAAAAC": 2055, + "CAAAAG": 2056, + "CAAATA": 2057, + "CAAATT": 2058, + "CAAATC": 2059, + "CAAATG": 2060, + "CAAACA": 2061, + "CAAACT": 2062, + "CAAACC": 2063, + "CAAACG": 2064, + "CAAAGA": 2065, + "CAAAGT": 2066, + "CAAAGC": 2067, + "CAAAGG": 2068, + "CAATAA": 2069, + "CAATAT": 2070, + "CAATAC": 2071, + "CAATAG": 2072, + "CAATTA": 2073, + "CAATTT": 2074, + "CAATTC": 2075, + "CAATTG": 2076, + "CAATCA": 2077, + "CAATCT": 2078, + "CAATCC": 2079, + "CAATCG": 2080, + "CAATGA": 2081, + "CAATGT": 2082, + "CAATGC": 2083, + "CAATGG": 2084, + "CAACAA": 2085, + "CAACAT": 2086, + "CAACAC": 2087, + "CAACAG": 2088, + "CAACTA": 2089, + "CAACTT": 2090, + "CAACTC": 2091, + "CAACTG": 2092, + "CAACCA": 2093, + "CAACCT": 2094, + "CAACCC": 2095, + "CAACCG": 2096, + "CAACGA": 2097, + "CAACGT": 2098, + "CAACGC": 2099, + "CAACGG": 2100, + "CAAGAA": 2101, + "CAAGAT": 2102, + "CAAGAC": 2103, + "CAAGAG": 2104, + "CAAGTA": 2105, + "CAAGTT": 2106, + "CAAGTC": 2107, + "CAAGTG": 2108, + "CAAGCA": 2109, + "CAAGCT": 2110, + "CAAGCC": 2111, + "CAAGCG": 2112, + "CAAGGA": 2113, + "CAAGGT": 2114, + "CAAGGC": 2115, + "CAAGGG": 2116, + "CATAAA": 2117, + "CATAAT": 2118, + "CATAAC": 2119, + "CATAAG": 2120, + "CATATA": 2121, + "CATATT": 2122, + "CATATC": 2123, + "CATATG": 2124, + "CATACA": 2125, + "CATACT": 2126, + "CATACC": 2127, + "CATACG": 2128, + "CATAGA": 2129, + "CATAGT": 2130, + "CATAGC": 2131, + "CATAGG": 2132, + "CATTAA": 2133, + "CATTAT": 2134, + "CATTAC": 2135, + "CATTAG": 2136, + "CATTTA": 2137, + "CATTTT": 2138, + "CATTTC": 2139, + "CATTTG": 2140, + "CATTCA": 2141, + "CATTCT": 2142, + "CATTCC": 2143, + "CATTCG": 2144, + "CATTGA": 2145, + "CATTGT": 2146, + "CATTGC": 2147, + "CATTGG": 2148, + "CATCAA": 2149, + "CATCAT": 2150, + "CATCAC": 2151, + "CATCAG": 2152, + "CATCTA": 2153, + "CATCTT": 2154, + "CATCTC": 2155, + "CATCTG": 2156, + "CATCCA": 2157, + "CATCCT": 2158, + "CATCCC": 2159, + "CATCCG": 2160, + "CATCGA": 2161, + "CATCGT": 2162, + "CATCGC": 2163, + "CATCGG": 2164, + "CATGAA": 2165, + "CATGAT": 2166, + "CATGAC": 2167, + "CATGAG": 2168, + "CATGTA": 2169, + "CATGTT": 2170, + "CATGTC": 2171, + "CATGTG": 2172, + "CATGCA": 2173, + "CATGCT": 2174, + "CATGCC": 2175, + "CATGCG": 2176, + "CATGGA": 2177, + "CATGGT": 2178, + "CATGGC": 2179, + "CATGGG": 2180, + "CACAAA": 2181, + "CACAAT": 2182, + "CACAAC": 2183, + "CACAAG": 2184, + "CACATA": 2185, + "CACATT": 2186, + "CACATC": 2187, + "CACATG": 2188, + "CACACA": 2189, + "CACACT": 2190, + "CACACC": 2191, + "CACACG": 2192, + "CACAGA": 2193, + "CACAGT": 2194, + "CACAGC": 2195, + "CACAGG": 2196, + "CACTAA": 2197, + "CACTAT": 2198, + "CACTAC": 2199, + "CACTAG": 2200, + "CACTTA": 2201, + "CACTTT": 2202, + "CACTTC": 2203, + "CACTTG": 2204, + "CACTCA": 2205, + "CACTCT": 2206, + "CACTCC": 2207, + "CACTCG": 2208, + "CACTGA": 2209, + "CACTGT": 2210, + "CACTGC": 2211, + "CACTGG": 2212, + "CACCAA": 2213, + "CACCAT": 2214, + "CACCAC": 2215, + "CACCAG": 2216, + "CACCTA": 2217, + "CACCTT": 2218, + "CACCTC": 2219, + "CACCTG": 2220, + "CACCCA": 2221, + "CACCCT": 2222, + "CACCCC": 2223, + "CACCCG": 2224, + "CACCGA": 2225, + "CACCGT": 2226, + "CACCGC": 2227, + "CACCGG": 2228, + "CACGAA": 2229, + "CACGAT": 2230, + "CACGAC": 2231, + "CACGAG": 2232, + "CACGTA": 2233, + "CACGTT": 2234, + "CACGTC": 2235, + "CACGTG": 2236, + "CACGCA": 2237, + "CACGCT": 2238, + "CACGCC": 2239, + "CACGCG": 2240, + "CACGGA": 2241, + "CACGGT": 2242, + "CACGGC": 2243, + "CACGGG": 2244, + "CAGAAA": 2245, + "CAGAAT": 2246, + "CAGAAC": 2247, + "CAGAAG": 2248, + "CAGATA": 2249, + "CAGATT": 2250, + "CAGATC": 2251, + "CAGATG": 2252, + "CAGACA": 2253, + "CAGACT": 2254, + "CAGACC": 2255, + "CAGACG": 2256, + "CAGAGA": 2257, + "CAGAGT": 2258, + "CAGAGC": 2259, + "CAGAGG": 2260, + "CAGTAA": 2261, + "CAGTAT": 2262, + "CAGTAC": 2263, + "CAGTAG": 2264, + "CAGTTA": 2265, + "CAGTTT": 2266, + "CAGTTC": 2267, + "CAGTTG": 2268, + "CAGTCA": 2269, + "CAGTCT": 2270, + "CAGTCC": 2271, + "CAGTCG": 2272, + "CAGTGA": 2273, + "CAGTGT": 2274, + "CAGTGC": 2275, + "CAGTGG": 2276, + "CAGCAA": 2277, + "CAGCAT": 2278, + "CAGCAC": 2279, + "CAGCAG": 2280, + "CAGCTA": 2281, + "CAGCTT": 2282, + "CAGCTC": 2283, + "CAGCTG": 2284, + "CAGCCA": 2285, + "CAGCCT": 2286, + "CAGCCC": 2287, + "CAGCCG": 2288, + "CAGCGA": 2289, + "CAGCGT": 2290, + "CAGCGC": 2291, + "CAGCGG": 2292, + "CAGGAA": 2293, + "CAGGAT": 2294, + "CAGGAC": 2295, + "CAGGAG": 2296, + "CAGGTA": 2297, + "CAGGTT": 2298, + "CAGGTC": 2299, + "CAGGTG": 2300, + "CAGGCA": 2301, + "CAGGCT": 2302, + "CAGGCC": 2303, + "CAGGCG": 2304, + "CAGGGA": 2305, + "CAGGGT": 2306, + "CAGGGC": 2307, + "CAGGGG": 2308, + "CTAAAA": 2309, + "CTAAAT": 2310, + "CTAAAC": 2311, + "CTAAAG": 2312, + "CTAATA": 2313, + "CTAATT": 2314, + "CTAATC": 2315, + "CTAATG": 2316, + "CTAACA": 2317, + "CTAACT": 2318, + "CTAACC": 2319, + "CTAACG": 2320, + "CTAAGA": 2321, + "CTAAGT": 2322, + "CTAAGC": 2323, + "CTAAGG": 2324, + "CTATAA": 2325, + "CTATAT": 2326, + "CTATAC": 2327, + "CTATAG": 2328, + "CTATTA": 2329, + "CTATTT": 2330, + "CTATTC": 2331, + "CTATTG": 2332, + "CTATCA": 2333, + "CTATCT": 2334, + "CTATCC": 2335, + "CTATCG": 2336, + "CTATGA": 2337, + "CTATGT": 2338, + "CTATGC": 2339, + "CTATGG": 2340, + "CTACAA": 2341, + "CTACAT": 2342, + "CTACAC": 2343, + "CTACAG": 2344, + "CTACTA": 2345, + "CTACTT": 2346, + "CTACTC": 2347, + "CTACTG": 2348, + "CTACCA": 2349, + "CTACCT": 2350, + "CTACCC": 2351, + "CTACCG": 2352, + "CTACGA": 2353, + "CTACGT": 2354, + "CTACGC": 2355, + "CTACGG": 2356, + "CTAGAA": 2357, + "CTAGAT": 2358, + "CTAGAC": 2359, + "CTAGAG": 2360, + "CTAGTA": 2361, + "CTAGTT": 2362, + "CTAGTC": 2363, + "CTAGTG": 2364, + "CTAGCA": 2365, + "CTAGCT": 2366, + "CTAGCC": 2367, + "CTAGCG": 2368, + "CTAGGA": 2369, + "CTAGGT": 2370, + "CTAGGC": 2371, + "CTAGGG": 2372, + "CTTAAA": 2373, + "CTTAAT": 2374, + "CTTAAC": 2375, + "CTTAAG": 2376, + "CTTATA": 2377, + "CTTATT": 2378, + "CTTATC": 2379, + "CTTATG": 2380, + "CTTACA": 2381, + "CTTACT": 2382, + "CTTACC": 2383, + "CTTACG": 2384, + "CTTAGA": 2385, + "CTTAGT": 2386, + "CTTAGC": 2387, + "CTTAGG": 2388, + "CTTTAA": 2389, + "CTTTAT": 2390, + "CTTTAC": 2391, + "CTTTAG": 2392, + "CTTTTA": 2393, + "CTTTTT": 2394, + "CTTTTC": 2395, + "CTTTTG": 2396, + "CTTTCA": 2397, + "CTTTCT": 2398, + "CTTTCC": 2399, + "CTTTCG": 2400, + "CTTTGA": 2401, + "CTTTGT": 2402, + "CTTTGC": 2403, + "CTTTGG": 2404, + "CTTCAA": 2405, + "CTTCAT": 2406, + "CTTCAC": 2407, + "CTTCAG": 2408, + "CTTCTA": 2409, + "CTTCTT": 2410, + "CTTCTC": 2411, + "CTTCTG": 2412, + "CTTCCA": 2413, + "CTTCCT": 2414, + "CTTCCC": 2415, + "CTTCCG": 2416, + "CTTCGA": 2417, + "CTTCGT": 2418, + "CTTCGC": 2419, + "CTTCGG": 2420, + "CTTGAA": 2421, + "CTTGAT": 2422, + "CTTGAC": 2423, + "CTTGAG": 2424, + "CTTGTA": 2425, + "CTTGTT": 2426, + "CTTGTC": 2427, + "CTTGTG": 2428, + "CTTGCA": 2429, + "CTTGCT": 2430, + "CTTGCC": 2431, + "CTTGCG": 2432, + "CTTGGA": 2433, + "CTTGGT": 2434, + "CTTGGC": 2435, + "CTTGGG": 2436, + "CTCAAA": 2437, + "CTCAAT": 2438, + "CTCAAC": 2439, + "CTCAAG": 2440, + "CTCATA": 2441, + "CTCATT": 2442, + "CTCATC": 2443, + "CTCATG": 2444, + "CTCACA": 2445, + "CTCACT": 2446, + "CTCACC": 2447, + "CTCACG": 2448, + "CTCAGA": 2449, + "CTCAGT": 2450, + "CTCAGC": 2451, + "CTCAGG": 2452, + "CTCTAA": 2453, + "CTCTAT": 2454, + "CTCTAC": 2455, + "CTCTAG": 2456, + "CTCTTA": 2457, + "CTCTTT": 2458, + "CTCTTC": 2459, + "CTCTTG": 2460, + "CTCTCA": 2461, + "CTCTCT": 2462, + "CTCTCC": 2463, + "CTCTCG": 2464, + "CTCTGA": 2465, + "CTCTGT": 2466, + "CTCTGC": 2467, + "CTCTGG": 2468, + "CTCCAA": 2469, + "CTCCAT": 2470, + "CTCCAC": 2471, + "CTCCAG": 2472, + "CTCCTA": 2473, + "CTCCTT": 2474, + "CTCCTC": 2475, + "CTCCTG": 2476, + "CTCCCA": 2477, + "CTCCCT": 2478, + "CTCCCC": 2479, + "CTCCCG": 2480, + "CTCCGA": 2481, + "CTCCGT": 2482, + "CTCCGC": 2483, + "CTCCGG": 2484, + "CTCGAA": 2485, + "CTCGAT": 2486, + "CTCGAC": 2487, + "CTCGAG": 2488, + "CTCGTA": 2489, + "CTCGTT": 2490, + "CTCGTC": 2491, + "CTCGTG": 2492, + "CTCGCA": 2493, + "CTCGCT": 2494, + "CTCGCC": 2495, + "CTCGCG": 2496, + "CTCGGA": 2497, + "CTCGGT": 2498, + "CTCGGC": 2499, + "CTCGGG": 2500, + "CTGAAA": 2501, + "CTGAAT": 2502, + "CTGAAC": 2503, + "CTGAAG": 2504, + "CTGATA": 2505, + "CTGATT": 2506, + "CTGATC": 2507, + "CTGATG": 2508, + "CTGACA": 2509, + "CTGACT": 2510, + "CTGACC": 2511, + "CTGACG": 2512, + "CTGAGA": 2513, + "CTGAGT": 2514, + "CTGAGC": 2515, + "CTGAGG": 2516, + "CTGTAA": 2517, + "CTGTAT": 2518, + "CTGTAC": 2519, + "CTGTAG": 2520, + "CTGTTA": 2521, + "CTGTTT": 2522, + "CTGTTC": 2523, + "CTGTTG": 2524, + "CTGTCA": 2525, + "CTGTCT": 2526, + "CTGTCC": 2527, + "CTGTCG": 2528, + "CTGTGA": 2529, + "CTGTGT": 2530, + "CTGTGC": 2531, + "CTGTGG": 2532, + "CTGCAA": 2533, + "CTGCAT": 2534, + "CTGCAC": 2535, + "CTGCAG": 2536, + "CTGCTA": 2537, + "CTGCTT": 2538, + "CTGCTC": 2539, + "CTGCTG": 2540, + "CTGCCA": 2541, + "CTGCCT": 2542, + "CTGCCC": 2543, + "CTGCCG": 2544, + "CTGCGA": 2545, + "CTGCGT": 2546, + "CTGCGC": 2547, + "CTGCGG": 2548, + "CTGGAA": 2549, + "CTGGAT": 2550, + "CTGGAC": 2551, + "CTGGAG": 2552, + "CTGGTA": 2553, + "CTGGTT": 2554, + "CTGGTC": 2555, + "CTGGTG": 2556, + "CTGGCA": 2557, + "CTGGCT": 2558, + "CTGGCC": 2559, + "CTGGCG": 2560, + "CTGGGA": 2561, + "CTGGGT": 2562, + "CTGGGC": 2563, + "CTGGGG": 2564, + "CCAAAA": 2565, + "CCAAAT": 2566, + "CCAAAC": 2567, + "CCAAAG": 2568, + "CCAATA": 2569, + "CCAATT": 2570, + "CCAATC": 2571, + "CCAATG": 2572, + "CCAACA": 2573, + "CCAACT": 2574, + "CCAACC": 2575, + "CCAACG": 2576, + "CCAAGA": 2577, + "CCAAGT": 2578, + "CCAAGC": 2579, + "CCAAGG": 2580, + "CCATAA": 2581, + "CCATAT": 2582, + "CCATAC": 2583, + "CCATAG": 2584, + "CCATTA": 2585, + "CCATTT": 2586, + "CCATTC": 2587, + "CCATTG": 2588, + "CCATCA": 2589, + "CCATCT": 2590, + "CCATCC": 2591, + "CCATCG": 2592, + "CCATGA": 2593, + "CCATGT": 2594, + "CCATGC": 2595, + "CCATGG": 2596, + "CCACAA": 2597, + "CCACAT": 2598, + "CCACAC": 2599, + "CCACAG": 2600, + "CCACTA": 2601, + "CCACTT": 2602, + "CCACTC": 2603, + "CCACTG": 2604, + "CCACCA": 2605, + "CCACCT": 2606, + "CCACCC": 2607, + "CCACCG": 2608, + "CCACGA": 2609, + "CCACGT": 2610, + "CCACGC": 2611, + "CCACGG": 2612, + "CCAGAA": 2613, + "CCAGAT": 2614, + "CCAGAC": 2615, + "CCAGAG": 2616, + "CCAGTA": 2617, + "CCAGTT": 2618, + "CCAGTC": 2619, + "CCAGTG": 2620, + "CCAGCA": 2621, + "CCAGCT": 2622, + "CCAGCC": 2623, + "CCAGCG": 2624, + "CCAGGA": 2625, + "CCAGGT": 2626, + "CCAGGC": 2627, + "CCAGGG": 2628, + "CCTAAA": 2629, + "CCTAAT": 2630, + "CCTAAC": 2631, + "CCTAAG": 2632, + "CCTATA": 2633, + "CCTATT": 2634, + "CCTATC": 2635, + "CCTATG": 2636, + "CCTACA": 2637, + "CCTACT": 2638, + "CCTACC": 2639, + "CCTACG": 2640, + "CCTAGA": 2641, + "CCTAGT": 2642, + "CCTAGC": 2643, + "CCTAGG": 2644, + "CCTTAA": 2645, + "CCTTAT": 2646, + "CCTTAC": 2647, + "CCTTAG": 2648, + "CCTTTA": 2649, + "CCTTTT": 2650, + "CCTTTC": 2651, + "CCTTTG": 2652, + "CCTTCA": 2653, + "CCTTCT": 2654, + "CCTTCC": 2655, + "CCTTCG": 2656, + "CCTTGA": 2657, + "CCTTGT": 2658, + "CCTTGC": 2659, + "CCTTGG": 2660, + "CCTCAA": 2661, + "CCTCAT": 2662, + "CCTCAC": 2663, + "CCTCAG": 2664, + "CCTCTA": 2665, + "CCTCTT": 2666, + "CCTCTC": 2667, + "CCTCTG": 2668, + "CCTCCA": 2669, + "CCTCCT": 2670, + "CCTCCC": 2671, + "CCTCCG": 2672, + "CCTCGA": 2673, + "CCTCGT": 2674, + "CCTCGC": 2675, + "CCTCGG": 2676, + "CCTGAA": 2677, + "CCTGAT": 2678, + "CCTGAC": 2679, + "CCTGAG": 2680, + "CCTGTA": 2681, + "CCTGTT": 2682, + "CCTGTC": 2683, + "CCTGTG": 2684, + "CCTGCA": 2685, + "CCTGCT": 2686, + "CCTGCC": 2687, + "CCTGCG": 2688, + "CCTGGA": 2689, + "CCTGGT": 2690, + "CCTGGC": 2691, + "CCTGGG": 2692, + "CCCAAA": 2693, + "CCCAAT": 2694, + "CCCAAC": 2695, + "CCCAAG": 2696, + "CCCATA": 2697, + "CCCATT": 2698, + "CCCATC": 2699, + "CCCATG": 2700, + "CCCACA": 2701, + "CCCACT": 2702, + "CCCACC": 2703, + "CCCACG": 2704, + "CCCAGA": 2705, + "CCCAGT": 2706, + "CCCAGC": 2707, + "CCCAGG": 2708, + "CCCTAA": 2709, + "CCCTAT": 2710, + "CCCTAC": 2711, + "CCCTAG": 2712, + "CCCTTA": 2713, + "CCCTTT": 2714, + "CCCTTC": 2715, + "CCCTTG": 2716, + "CCCTCA": 2717, + "CCCTCT": 2718, + "CCCTCC": 2719, + "CCCTCG": 2720, + "CCCTGA": 2721, + "CCCTGT": 2722, + "CCCTGC": 2723, + "CCCTGG": 2724, + "CCCCAA": 2725, + "CCCCAT": 2726, + "CCCCAC": 2727, + "CCCCAG": 2728, + "CCCCTA": 2729, + "CCCCTT": 2730, + "CCCCTC": 2731, + "CCCCTG": 2732, + "CCCCCA": 2733, + "CCCCCT": 2734, + "CCCCCC": 2735, + "CCCCCG": 2736, + "CCCCGA": 2737, + "CCCCGT": 2738, + "CCCCGC": 2739, + "CCCCGG": 2740, + "CCCGAA": 2741, + "CCCGAT": 2742, + "CCCGAC": 2743, + "CCCGAG": 2744, + "CCCGTA": 2745, + "CCCGTT": 2746, + "CCCGTC": 2747, + "CCCGTG": 2748, + "CCCGCA": 2749, + "CCCGCT": 2750, + "CCCGCC": 2751, + "CCCGCG": 2752, + "CCCGGA": 2753, + "CCCGGT": 2754, + "CCCGGC": 2755, + "CCCGGG": 2756, + "CCGAAA": 2757, + "CCGAAT": 2758, + "CCGAAC": 2759, + "CCGAAG": 2760, + "CCGATA": 2761, + "CCGATT": 2762, + "CCGATC": 2763, + "CCGATG": 2764, + "CCGACA": 2765, + "CCGACT": 2766, + "CCGACC": 2767, + "CCGACG": 2768, + "CCGAGA": 2769, + "CCGAGT": 2770, + "CCGAGC": 2771, + "CCGAGG": 2772, + "CCGTAA": 2773, + "CCGTAT": 2774, + "CCGTAC": 2775, + "CCGTAG": 2776, + "CCGTTA": 2777, + "CCGTTT": 2778, + "CCGTTC": 2779, + "CCGTTG": 2780, + "CCGTCA": 2781, + "CCGTCT": 2782, + "CCGTCC": 2783, + "CCGTCG": 2784, + "CCGTGA": 2785, + "CCGTGT": 2786, + "CCGTGC": 2787, + "CCGTGG": 2788, + "CCGCAA": 2789, + "CCGCAT": 2790, + "CCGCAC": 2791, + "CCGCAG": 2792, + "CCGCTA": 2793, + "CCGCTT": 2794, + "CCGCTC": 2795, + "CCGCTG": 2796, + "CCGCCA": 2797, + "CCGCCT": 2798, + "CCGCCC": 2799, + "CCGCCG": 2800, + "CCGCGA": 2801, + "CCGCGT": 2802, + "CCGCGC": 2803, + "CCGCGG": 2804, + "CCGGAA": 2805, + "CCGGAT": 2806, + "CCGGAC": 2807, + "CCGGAG": 2808, + "CCGGTA": 2809, + "CCGGTT": 2810, + "CCGGTC": 2811, + "CCGGTG": 2812, + "CCGGCA": 2813, + "CCGGCT": 2814, + "CCGGCC": 2815, + "CCGGCG": 2816, + "CCGGGA": 2817, + "CCGGGT": 2818, + "CCGGGC": 2819, + "CCGGGG": 2820, + "CGAAAA": 2821, + "CGAAAT": 2822, + "CGAAAC": 2823, + "CGAAAG": 2824, + "CGAATA": 2825, + "CGAATT": 2826, + "CGAATC": 2827, + "CGAATG": 2828, + "CGAACA": 2829, + "CGAACT": 2830, + "CGAACC": 2831, + "CGAACG": 2832, + "CGAAGA": 2833, + "CGAAGT": 2834, + "CGAAGC": 2835, + "CGAAGG": 2836, + "CGATAA": 2837, + "CGATAT": 2838, + "CGATAC": 2839, + "CGATAG": 2840, + "CGATTA": 2841, + "CGATTT": 2842, + "CGATTC": 2843, + "CGATTG": 2844, + "CGATCA": 2845, + "CGATCT": 2846, + "CGATCC": 2847, + "CGATCG": 2848, + "CGATGA": 2849, + "CGATGT": 2850, + "CGATGC": 2851, + "CGATGG": 2852, + "CGACAA": 2853, + "CGACAT": 2854, + "CGACAC": 2855, + "CGACAG": 2856, + "CGACTA": 2857, + "CGACTT": 2858, + "CGACTC": 2859, + "CGACTG": 2860, + "CGACCA": 2861, + "CGACCT": 2862, + "CGACCC": 2863, + "CGACCG": 2864, + "CGACGA": 2865, + "CGACGT": 2866, + "CGACGC": 2867, + "CGACGG": 2868, + "CGAGAA": 2869, + "CGAGAT": 2870, + "CGAGAC": 2871, + "CGAGAG": 2872, + "CGAGTA": 2873, + "CGAGTT": 2874, + "CGAGTC": 2875, + "CGAGTG": 2876, + "CGAGCA": 2877, + "CGAGCT": 2878, + "CGAGCC": 2879, + "CGAGCG": 2880, + "CGAGGA": 2881, + "CGAGGT": 2882, + "CGAGGC": 2883, + "CGAGGG": 2884, + "CGTAAA": 2885, + "CGTAAT": 2886, + "CGTAAC": 2887, + "CGTAAG": 2888, + "CGTATA": 2889, + "CGTATT": 2890, + "CGTATC": 2891, + "CGTATG": 2892, + "CGTACA": 2893, + "CGTACT": 2894, + "CGTACC": 2895, + "CGTACG": 2896, + "CGTAGA": 2897, + "CGTAGT": 2898, + "CGTAGC": 2899, + "CGTAGG": 2900, + "CGTTAA": 2901, + "CGTTAT": 2902, + "CGTTAC": 2903, + "CGTTAG": 2904, + "CGTTTA": 2905, + "CGTTTT": 2906, + "CGTTTC": 2907, + "CGTTTG": 2908, + "CGTTCA": 2909, + "CGTTCT": 2910, + "CGTTCC": 2911, + "CGTTCG": 2912, + "CGTTGA": 2913, + "CGTTGT": 2914, + "CGTTGC": 2915, + "CGTTGG": 2916, + "CGTCAA": 2917, + "CGTCAT": 2918, + "CGTCAC": 2919, + "CGTCAG": 2920, + "CGTCTA": 2921, + "CGTCTT": 2922, + "CGTCTC": 2923, + "CGTCTG": 2924, + "CGTCCA": 2925, + "CGTCCT": 2926, + "CGTCCC": 2927, + "CGTCCG": 2928, + "CGTCGA": 2929, + "CGTCGT": 2930, + "CGTCGC": 2931, + "CGTCGG": 2932, + "CGTGAA": 2933, + "CGTGAT": 2934, + "CGTGAC": 2935, + "CGTGAG": 2936, + "CGTGTA": 2937, + "CGTGTT": 2938, + "CGTGTC": 2939, + "CGTGTG": 2940, + "CGTGCA": 2941, + "CGTGCT": 2942, + "CGTGCC": 2943, + "CGTGCG": 2944, + "CGTGGA": 2945, + "CGTGGT": 2946, + "CGTGGC": 2947, + "CGTGGG": 2948, + "CGCAAA": 2949, + "CGCAAT": 2950, + "CGCAAC": 2951, + "CGCAAG": 2952, + "CGCATA": 2953, + "CGCATT": 2954, + "CGCATC": 2955, + "CGCATG": 2956, + "CGCACA": 2957, + "CGCACT": 2958, + "CGCACC": 2959, + "CGCACG": 2960, + "CGCAGA": 2961, + "CGCAGT": 2962, + "CGCAGC": 2963, + "CGCAGG": 2964, + "CGCTAA": 2965, + "CGCTAT": 2966, + "CGCTAC": 2967, + "CGCTAG": 2968, + "CGCTTA": 2969, + "CGCTTT": 2970, + "CGCTTC": 2971, + "CGCTTG": 2972, + "CGCTCA": 2973, + "CGCTCT": 2974, + "CGCTCC": 2975, + "CGCTCG": 2976, + "CGCTGA": 2977, + "CGCTGT": 2978, + "CGCTGC": 2979, + "CGCTGG": 2980, + "CGCCAA": 2981, + "CGCCAT": 2982, + "CGCCAC": 2983, + "CGCCAG": 2984, + "CGCCTA": 2985, + "CGCCTT": 2986, + "CGCCTC": 2987, + "CGCCTG": 2988, + "CGCCCA": 2989, + "CGCCCT": 2990, + "CGCCCC": 2991, + "CGCCCG": 2992, + "CGCCGA": 2993, + "CGCCGT": 2994, + "CGCCGC": 2995, + "CGCCGG": 2996, + "CGCGAA": 2997, + "CGCGAT": 2998, + "CGCGAC": 2999, + "CGCGAG": 3000, + "CGCGTA": 3001, + "CGCGTT": 3002, + "CGCGTC": 3003, + "CGCGTG": 3004, + "CGCGCA": 3005, + "CGCGCT": 3006, + "CGCGCC": 3007, + "CGCGCG": 3008, + "CGCGGA": 3009, + "CGCGGT": 3010, + "CGCGGC": 3011, + "CGCGGG": 3012, + "CGGAAA": 3013, + "CGGAAT": 3014, + "CGGAAC": 3015, + "CGGAAG": 3016, + "CGGATA": 3017, + "CGGATT": 3018, + "CGGATC": 3019, + "CGGATG": 3020, + "CGGACA": 3021, + "CGGACT": 3022, + "CGGACC": 3023, + "CGGACG": 3024, + "CGGAGA": 3025, + "CGGAGT": 3026, + "CGGAGC": 3027, + "CGGAGG": 3028, + "CGGTAA": 3029, + "CGGTAT": 3030, + "CGGTAC": 3031, + "CGGTAG": 3032, + "CGGTTA": 3033, + "CGGTTT": 3034, + "CGGTTC": 3035, + "CGGTTG": 3036, + "CGGTCA": 3037, + "CGGTCT": 3038, + "CGGTCC": 3039, + "CGGTCG": 3040, + "CGGTGA": 3041, + "CGGTGT": 3042, + "CGGTGC": 3043, + "CGGTGG": 3044, + "CGGCAA": 3045, + "CGGCAT": 3046, + "CGGCAC": 3047, + "CGGCAG": 3048, + "CGGCTA": 3049, + "CGGCTT": 3050, + "CGGCTC": 3051, + "CGGCTG": 3052, + "CGGCCA": 3053, + "CGGCCT": 3054, + "CGGCCC": 3055, + "CGGCCG": 3056, + "CGGCGA": 3057, + "CGGCGT": 3058, + "CGGCGC": 3059, + "CGGCGG": 3060, + "CGGGAA": 3061, + "CGGGAT": 3062, + "CGGGAC": 3063, + "CGGGAG": 3064, + "CGGGTA": 3065, + "CGGGTT": 3066, + "CGGGTC": 3067, + "CGGGTG": 3068, + "CGGGCA": 3069, + "CGGGCT": 3070, + "CGGGCC": 3071, + "CGGGCG": 3072, + "CGGGGA": 3073, + "CGGGGT": 3074, + "CGGGGC": 3075, + "CGGGGG": 3076, + "GAAAAA": 3077, + "GAAAAT": 3078, + "GAAAAC": 3079, + "GAAAAG": 3080, + "GAAATA": 3081, + "GAAATT": 3082, + "GAAATC": 3083, + "GAAATG": 3084, + "GAAACA": 3085, + "GAAACT": 3086, + "GAAACC": 3087, + "GAAACG": 3088, + "GAAAGA": 3089, + "GAAAGT": 3090, + "GAAAGC": 3091, + "GAAAGG": 3092, + "GAATAA": 3093, + "GAATAT": 3094, + "GAATAC": 3095, + "GAATAG": 3096, + "GAATTA": 3097, + "GAATTT": 3098, + "GAATTC": 3099, + "GAATTG": 3100, + "GAATCA": 3101, + "GAATCT": 3102, + "GAATCC": 3103, + "GAATCG": 3104, + "GAATGA": 3105, + "GAATGT": 3106, + "GAATGC": 3107, + "GAATGG": 3108, + "GAACAA": 3109, + "GAACAT": 3110, + "GAACAC": 3111, + "GAACAG": 3112, + "GAACTA": 3113, + "GAACTT": 3114, + "GAACTC": 3115, + "GAACTG": 3116, + "GAACCA": 3117, + "GAACCT": 3118, + "GAACCC": 3119, + "GAACCG": 3120, + "GAACGA": 3121, + "GAACGT": 3122, + "GAACGC": 3123, + "GAACGG": 3124, + "GAAGAA": 3125, + "GAAGAT": 3126, + "GAAGAC": 3127, + "GAAGAG": 3128, + "GAAGTA": 3129, + "GAAGTT": 3130, + "GAAGTC": 3131, + "GAAGTG": 3132, + "GAAGCA": 3133, + "GAAGCT": 3134, + "GAAGCC": 3135, + "GAAGCG": 3136, + "GAAGGA": 3137, + "GAAGGT": 3138, + "GAAGGC": 3139, + "GAAGGG": 3140, + "GATAAA": 3141, + "GATAAT": 3142, + "GATAAC": 3143, + "GATAAG": 3144, + "GATATA": 3145, + "GATATT": 3146, + "GATATC": 3147, + "GATATG": 3148, + "GATACA": 3149, + "GATACT": 3150, + "GATACC": 3151, + "GATACG": 3152, + "GATAGA": 3153, + "GATAGT": 3154, + "GATAGC": 3155, + "GATAGG": 3156, + "GATTAA": 3157, + "GATTAT": 3158, + "GATTAC": 3159, + "GATTAG": 3160, + "GATTTA": 3161, + "GATTTT": 3162, + "GATTTC": 3163, + "GATTTG": 3164, + "GATTCA": 3165, + "GATTCT": 3166, + "GATTCC": 3167, + "GATTCG": 3168, + "GATTGA": 3169, + "GATTGT": 3170, + "GATTGC": 3171, + "GATTGG": 3172, + "GATCAA": 3173, + "GATCAT": 3174, + "GATCAC": 3175, + "GATCAG": 3176, + "GATCTA": 3177, + "GATCTT": 3178, + "GATCTC": 3179, + "GATCTG": 3180, + "GATCCA": 3181, + "GATCCT": 3182, + "GATCCC": 3183, + "GATCCG": 3184, + "GATCGA": 3185, + "GATCGT": 3186, + "GATCGC": 3187, + "GATCGG": 3188, + "GATGAA": 3189, + "GATGAT": 3190, + "GATGAC": 3191, + "GATGAG": 3192, + "GATGTA": 3193, + "GATGTT": 3194, + "GATGTC": 3195, + "GATGTG": 3196, + "GATGCA": 3197, + "GATGCT": 3198, + "GATGCC": 3199, + "GATGCG": 3200, + "GATGGA": 3201, + "GATGGT": 3202, + "GATGGC": 3203, + "GATGGG": 3204, + "GACAAA": 3205, + "GACAAT": 3206, + "GACAAC": 3207, + "GACAAG": 3208, + "GACATA": 3209, + "GACATT": 3210, + "GACATC": 3211, + "GACATG": 3212, + "GACACA": 3213, + "GACACT": 3214, + "GACACC": 3215, + "GACACG": 3216, + "GACAGA": 3217, + "GACAGT": 3218, + "GACAGC": 3219, + "GACAGG": 3220, + "GACTAA": 3221, + "GACTAT": 3222, + "GACTAC": 3223, + "GACTAG": 3224, + "GACTTA": 3225, + "GACTTT": 3226, + "GACTTC": 3227, + "GACTTG": 3228, + "GACTCA": 3229, + "GACTCT": 3230, + "GACTCC": 3231, + "GACTCG": 3232, + "GACTGA": 3233, + "GACTGT": 3234, + "GACTGC": 3235, + "GACTGG": 3236, + "GACCAA": 3237, + "GACCAT": 3238, + "GACCAC": 3239, + "GACCAG": 3240, + "GACCTA": 3241, + "GACCTT": 3242, + "GACCTC": 3243, + "GACCTG": 3244, + "GACCCA": 3245, + "GACCCT": 3246, + "GACCCC": 3247, + "GACCCG": 3248, + "GACCGA": 3249, + "GACCGT": 3250, + "GACCGC": 3251, + "GACCGG": 3252, + "GACGAA": 3253, + "GACGAT": 3254, + "GACGAC": 3255, + "GACGAG": 3256, + "GACGTA": 3257, + "GACGTT": 3258, + "GACGTC": 3259, + "GACGTG": 3260, + "GACGCA": 3261, + "GACGCT": 3262, + "GACGCC": 3263, + "GACGCG": 3264, + "GACGGA": 3265, + "GACGGT": 3266, + "GACGGC": 3267, + "GACGGG": 3268, + "GAGAAA": 3269, + "GAGAAT": 3270, + "GAGAAC": 3271, + "GAGAAG": 3272, + "GAGATA": 3273, + "GAGATT": 3274, + "GAGATC": 3275, + "GAGATG": 3276, + "GAGACA": 3277, + "GAGACT": 3278, + "GAGACC": 3279, + "GAGACG": 3280, + "GAGAGA": 3281, + "GAGAGT": 3282, + "GAGAGC": 3283, + "GAGAGG": 3284, + "GAGTAA": 3285, + "GAGTAT": 3286, + "GAGTAC": 3287, + "GAGTAG": 3288, + "GAGTTA": 3289, + "GAGTTT": 3290, + "GAGTTC": 3291, + "GAGTTG": 3292, + "GAGTCA": 3293, + "GAGTCT": 3294, + "GAGTCC": 3295, + "GAGTCG": 3296, + "GAGTGA": 3297, + "GAGTGT": 3298, + "GAGTGC": 3299, + "GAGTGG": 3300, + "GAGCAA": 3301, + "GAGCAT": 3302, + "GAGCAC": 3303, + "GAGCAG": 3304, + "GAGCTA": 3305, + "GAGCTT": 3306, + "GAGCTC": 3307, + "GAGCTG": 3308, + "GAGCCA": 3309, + "GAGCCT": 3310, + "GAGCCC": 3311, + "GAGCCG": 3312, + "GAGCGA": 3313, + "GAGCGT": 3314, + "GAGCGC": 3315, + "GAGCGG": 3316, + "GAGGAA": 3317, + "GAGGAT": 3318, + "GAGGAC": 3319, + "GAGGAG": 3320, + "GAGGTA": 3321, + "GAGGTT": 3322, + "GAGGTC": 3323, + "GAGGTG": 3324, + "GAGGCA": 3325, + "GAGGCT": 3326, + "GAGGCC": 3327, + "GAGGCG": 3328, + "GAGGGA": 3329, + "GAGGGT": 3330, + "GAGGGC": 3331, + "GAGGGG": 3332, + "GTAAAA": 3333, + "GTAAAT": 3334, + "GTAAAC": 3335, + "GTAAAG": 3336, + "GTAATA": 3337, + "GTAATT": 3338, + "GTAATC": 3339, + "GTAATG": 3340, + "GTAACA": 3341, + "GTAACT": 3342, + "GTAACC": 3343, + "GTAACG": 3344, + "GTAAGA": 3345, + "GTAAGT": 3346, + "GTAAGC": 3347, + "GTAAGG": 3348, + "GTATAA": 3349, + "GTATAT": 3350, + "GTATAC": 3351, + "GTATAG": 3352, + "GTATTA": 3353, + "GTATTT": 3354, + "GTATTC": 3355, + "GTATTG": 3356, + "GTATCA": 3357, + "GTATCT": 3358, + "GTATCC": 3359, + "GTATCG": 3360, + "GTATGA": 3361, + "GTATGT": 3362, + "GTATGC": 3363, + "GTATGG": 3364, + "GTACAA": 3365, + "GTACAT": 3366, + "GTACAC": 3367, + "GTACAG": 3368, + "GTACTA": 3369, + "GTACTT": 3370, + "GTACTC": 3371, + "GTACTG": 3372, + "GTACCA": 3373, + "GTACCT": 3374, + "GTACCC": 3375, + "GTACCG": 3376, + "GTACGA": 3377, + "GTACGT": 3378, + "GTACGC": 3379, + "GTACGG": 3380, + "GTAGAA": 3381, + "GTAGAT": 3382, + "GTAGAC": 3383, + "GTAGAG": 3384, + "GTAGTA": 3385, + "GTAGTT": 3386, + "GTAGTC": 3387, + "GTAGTG": 3388, + "GTAGCA": 3389, + "GTAGCT": 3390, + "GTAGCC": 3391, + "GTAGCG": 3392, + "GTAGGA": 3393, + "GTAGGT": 3394, + "GTAGGC": 3395, + "GTAGGG": 3396, + "GTTAAA": 3397, + "GTTAAT": 3398, + "GTTAAC": 3399, + "GTTAAG": 3400, + "GTTATA": 3401, + "GTTATT": 3402, + "GTTATC": 3403, + "GTTATG": 3404, + "GTTACA": 3405, + "GTTACT": 3406, + "GTTACC": 3407, + "GTTACG": 3408, + "GTTAGA": 3409, + "GTTAGT": 3410, + "GTTAGC": 3411, + "GTTAGG": 3412, + "GTTTAA": 3413, + "GTTTAT": 3414, + "GTTTAC": 3415, + "GTTTAG": 3416, + "GTTTTA": 3417, + "GTTTTT": 3418, + "GTTTTC": 3419, + "GTTTTG": 3420, + "GTTTCA": 3421, + "GTTTCT": 3422, + "GTTTCC": 3423, + "GTTTCG": 3424, + "GTTTGA": 3425, + "GTTTGT": 3426, + "GTTTGC": 3427, + "GTTTGG": 3428, + "GTTCAA": 3429, + "GTTCAT": 3430, + "GTTCAC": 3431, + "GTTCAG": 3432, + "GTTCTA": 3433, + "GTTCTT": 3434, + "GTTCTC": 3435, + "GTTCTG": 3436, + "GTTCCA": 3437, + "GTTCCT": 3438, + "GTTCCC": 3439, + "GTTCCG": 3440, + "GTTCGA": 3441, + "GTTCGT": 3442, + "GTTCGC": 3443, + "GTTCGG": 3444, + "GTTGAA": 3445, + "GTTGAT": 3446, + "GTTGAC": 3447, + "GTTGAG": 3448, + "GTTGTA": 3449, + "GTTGTT": 3450, + "GTTGTC": 3451, + "GTTGTG": 3452, + "GTTGCA": 3453, + "GTTGCT": 3454, + "GTTGCC": 3455, + "GTTGCG": 3456, + "GTTGGA": 3457, + "GTTGGT": 3458, + "GTTGGC": 3459, + "GTTGGG": 3460, + "GTCAAA": 3461, + "GTCAAT": 3462, + "GTCAAC": 3463, + "GTCAAG": 3464, + "GTCATA": 3465, + "GTCATT": 3466, + "GTCATC": 3467, + "GTCATG": 3468, + "GTCACA": 3469, + "GTCACT": 3470, + "GTCACC": 3471, + "GTCACG": 3472, + "GTCAGA": 3473, + "GTCAGT": 3474, + "GTCAGC": 3475, + "GTCAGG": 3476, + "GTCTAA": 3477, + "GTCTAT": 3478, + "GTCTAC": 3479, + "GTCTAG": 3480, + "GTCTTA": 3481, + "GTCTTT": 3482, + "GTCTTC": 3483, + "GTCTTG": 3484, + "GTCTCA": 3485, + "GTCTCT": 3486, + "GTCTCC": 3487, + "GTCTCG": 3488, + "GTCTGA": 3489, + "GTCTGT": 3490, + "GTCTGC": 3491, + "GTCTGG": 3492, + "GTCCAA": 3493, + "GTCCAT": 3494, + "GTCCAC": 3495, + "GTCCAG": 3496, + "GTCCTA": 3497, + "GTCCTT": 3498, + "GTCCTC": 3499, + "GTCCTG": 3500, + "GTCCCA": 3501, + "GTCCCT": 3502, + "GTCCCC": 3503, + "GTCCCG": 3504, + "GTCCGA": 3505, + "GTCCGT": 3506, + "GTCCGC": 3507, + "GTCCGG": 3508, + "GTCGAA": 3509, + "GTCGAT": 3510, + "GTCGAC": 3511, + "GTCGAG": 3512, + "GTCGTA": 3513, + "GTCGTT": 3514, + "GTCGTC": 3515, + "GTCGTG": 3516, + "GTCGCA": 3517, + "GTCGCT": 3518, + "GTCGCC": 3519, + "GTCGCG": 3520, + "GTCGGA": 3521, + "GTCGGT": 3522, + "GTCGGC": 3523, + "GTCGGG": 3524, + "GTGAAA": 3525, + "GTGAAT": 3526, + "GTGAAC": 3527, + "GTGAAG": 3528, + "GTGATA": 3529, + "GTGATT": 3530, + "GTGATC": 3531, + "GTGATG": 3532, + "GTGACA": 3533, + "GTGACT": 3534, + "GTGACC": 3535, + "GTGACG": 3536, + "GTGAGA": 3537, + "GTGAGT": 3538, + "GTGAGC": 3539, + "GTGAGG": 3540, + "GTGTAA": 3541, + "GTGTAT": 3542, + "GTGTAC": 3543, + "GTGTAG": 3544, + "GTGTTA": 3545, + "GTGTTT": 3546, + "GTGTTC": 3547, + "GTGTTG": 3548, + "GTGTCA": 3549, + "GTGTCT": 3550, + "GTGTCC": 3551, + "GTGTCG": 3552, + "GTGTGA": 3553, + "GTGTGT": 3554, + "GTGTGC": 3555, + "GTGTGG": 3556, + "GTGCAA": 3557, + "GTGCAT": 3558, + "GTGCAC": 3559, + "GTGCAG": 3560, + "GTGCTA": 3561, + "GTGCTT": 3562, + "GTGCTC": 3563, + "GTGCTG": 3564, + "GTGCCA": 3565, + "GTGCCT": 3566, + "GTGCCC": 3567, + "GTGCCG": 3568, + "GTGCGA": 3569, + "GTGCGT": 3570, + "GTGCGC": 3571, + "GTGCGG": 3572, + "GTGGAA": 3573, + "GTGGAT": 3574, + "GTGGAC": 3575, + "GTGGAG": 3576, + "GTGGTA": 3577, + "GTGGTT": 3578, + "GTGGTC": 3579, + "GTGGTG": 3580, + "GTGGCA": 3581, + "GTGGCT": 3582, + "GTGGCC": 3583, + "GTGGCG": 3584, + "GTGGGA": 3585, + "GTGGGT": 3586, + "GTGGGC": 3587, + "GTGGGG": 3588, + "GCAAAA": 3589, + "GCAAAT": 3590, + "GCAAAC": 3591, + "GCAAAG": 3592, + "GCAATA": 3593, + "GCAATT": 3594, + "GCAATC": 3595, + "GCAATG": 3596, + "GCAACA": 3597, + "GCAACT": 3598, + "GCAACC": 3599, + "GCAACG": 3600, + "GCAAGA": 3601, + "GCAAGT": 3602, + "GCAAGC": 3603, + "GCAAGG": 3604, + "GCATAA": 3605, + "GCATAT": 3606, + "GCATAC": 3607, + "GCATAG": 3608, + "GCATTA": 3609, + "GCATTT": 3610, + "GCATTC": 3611, + "GCATTG": 3612, + "GCATCA": 3613, + "GCATCT": 3614, + "GCATCC": 3615, + "GCATCG": 3616, + "GCATGA": 3617, + "GCATGT": 3618, + "GCATGC": 3619, + "GCATGG": 3620, + "GCACAA": 3621, + "GCACAT": 3622, + "GCACAC": 3623, + "GCACAG": 3624, + "GCACTA": 3625, + "GCACTT": 3626, + "GCACTC": 3627, + "GCACTG": 3628, + "GCACCA": 3629, + "GCACCT": 3630, + "GCACCC": 3631, + "GCACCG": 3632, + "GCACGA": 3633, + "GCACGT": 3634, + "GCACGC": 3635, + "GCACGG": 3636, + "GCAGAA": 3637, + "GCAGAT": 3638, + "GCAGAC": 3639, + "GCAGAG": 3640, + "GCAGTA": 3641, + "GCAGTT": 3642, + "GCAGTC": 3643, + "GCAGTG": 3644, + "GCAGCA": 3645, + "GCAGCT": 3646, + "GCAGCC": 3647, + "GCAGCG": 3648, + "GCAGGA": 3649, + "GCAGGT": 3650, + "GCAGGC": 3651, + "GCAGGG": 3652, + "GCTAAA": 3653, + "GCTAAT": 3654, + "GCTAAC": 3655, + "GCTAAG": 3656, + "GCTATA": 3657, + "GCTATT": 3658, + "GCTATC": 3659, + "GCTATG": 3660, + "GCTACA": 3661, + "GCTACT": 3662, + "GCTACC": 3663, + "GCTACG": 3664, + "GCTAGA": 3665, + "GCTAGT": 3666, + "GCTAGC": 3667, + "GCTAGG": 3668, + "GCTTAA": 3669, + "GCTTAT": 3670, + "GCTTAC": 3671, + "GCTTAG": 3672, + "GCTTTA": 3673, + "GCTTTT": 3674, + "GCTTTC": 3675, + "GCTTTG": 3676, + "GCTTCA": 3677, + "GCTTCT": 3678, + "GCTTCC": 3679, + "GCTTCG": 3680, + "GCTTGA": 3681, + "GCTTGT": 3682, + "GCTTGC": 3683, + "GCTTGG": 3684, + "GCTCAA": 3685, + "GCTCAT": 3686, + "GCTCAC": 3687, + "GCTCAG": 3688, + "GCTCTA": 3689, + "GCTCTT": 3690, + "GCTCTC": 3691, + "GCTCTG": 3692, + "GCTCCA": 3693, + "GCTCCT": 3694, + "GCTCCC": 3695, + "GCTCCG": 3696, + "GCTCGA": 3697, + "GCTCGT": 3698, + "GCTCGC": 3699, + "GCTCGG": 3700, + "GCTGAA": 3701, + "GCTGAT": 3702, + "GCTGAC": 3703, + "GCTGAG": 3704, + "GCTGTA": 3705, + "GCTGTT": 3706, + "GCTGTC": 3707, + "GCTGTG": 3708, + "GCTGCA": 3709, + "GCTGCT": 3710, + "GCTGCC": 3711, + "GCTGCG": 3712, + "GCTGGA": 3713, + "GCTGGT": 3714, + "GCTGGC": 3715, + "GCTGGG": 3716, + "GCCAAA": 3717, + "GCCAAT": 3718, + "GCCAAC": 3719, + "GCCAAG": 3720, + "GCCATA": 3721, + "GCCATT": 3722, + "GCCATC": 3723, + "GCCATG": 3724, + "GCCACA": 3725, + "GCCACT": 3726, + "GCCACC": 3727, + "GCCACG": 3728, + "GCCAGA": 3729, + "GCCAGT": 3730, + "GCCAGC": 3731, + "GCCAGG": 3732, + "GCCTAA": 3733, + "GCCTAT": 3734, + "GCCTAC": 3735, + "GCCTAG": 3736, + "GCCTTA": 3737, + "GCCTTT": 3738, + "GCCTTC": 3739, + "GCCTTG": 3740, + "GCCTCA": 3741, + "GCCTCT": 3742, + "GCCTCC": 3743, + "GCCTCG": 3744, + "GCCTGA": 3745, + "GCCTGT": 3746, + "GCCTGC": 3747, + "GCCTGG": 3748, + "GCCCAA": 3749, + "GCCCAT": 3750, + "GCCCAC": 3751, + "GCCCAG": 3752, + "GCCCTA": 3753, + "GCCCTT": 3754, + "GCCCTC": 3755, + "GCCCTG": 3756, + "GCCCCA": 3757, + "GCCCCT": 3758, + "GCCCCC": 3759, + "GCCCCG": 3760, + "GCCCGA": 3761, + "GCCCGT": 3762, + "GCCCGC": 3763, + "GCCCGG": 3764, + "GCCGAA": 3765, + "GCCGAT": 3766, + "GCCGAC": 3767, + "GCCGAG": 3768, + "GCCGTA": 3769, + "GCCGTT": 3770, + "GCCGTC": 3771, + "GCCGTG": 3772, + "GCCGCA": 3773, + "GCCGCT": 3774, + "GCCGCC": 3775, + "GCCGCG": 3776, + "GCCGGA": 3777, + "GCCGGT": 3778, + "GCCGGC": 3779, + "GCCGGG": 3780, + "GCGAAA": 3781, + "GCGAAT": 3782, + "GCGAAC": 3783, + "GCGAAG": 3784, + "GCGATA": 3785, + "GCGATT": 3786, + "GCGATC": 3787, + "GCGATG": 3788, + "GCGACA": 3789, + "GCGACT": 3790, + "GCGACC": 3791, + "GCGACG": 3792, + "GCGAGA": 3793, + "GCGAGT": 3794, + "GCGAGC": 3795, + "GCGAGG": 3796, + "GCGTAA": 3797, + "GCGTAT": 3798, + "GCGTAC": 3799, + "GCGTAG": 3800, + "GCGTTA": 3801, + "GCGTTT": 3802, + "GCGTTC": 3803, + "GCGTTG": 3804, + "GCGTCA": 3805, + "GCGTCT": 3806, + "GCGTCC": 3807, + "GCGTCG": 3808, + "GCGTGA": 3809, + "GCGTGT": 3810, + "GCGTGC": 3811, + "GCGTGG": 3812, + "GCGCAA": 3813, + "GCGCAT": 3814, + "GCGCAC": 3815, + "GCGCAG": 3816, + "GCGCTA": 3817, + "GCGCTT": 3818, + "GCGCTC": 3819, + "GCGCTG": 3820, + "GCGCCA": 3821, + "GCGCCT": 3822, + "GCGCCC": 3823, + "GCGCCG": 3824, + "GCGCGA": 3825, + "GCGCGT": 3826, + "GCGCGC": 3827, + "GCGCGG": 3828, + "GCGGAA": 3829, + "GCGGAT": 3830, + "GCGGAC": 3831, + "GCGGAG": 3832, + "GCGGTA": 3833, + "GCGGTT": 3834, + "GCGGTC": 3835, + "GCGGTG": 3836, + "GCGGCA": 3837, + "GCGGCT": 3838, + "GCGGCC": 3839, + "GCGGCG": 3840, + "GCGGGA": 3841, + "GCGGGT": 3842, + "GCGGGC": 3843, + "GCGGGG": 3844, + "GGAAAA": 3845, + "GGAAAT": 3846, + "GGAAAC": 3847, + "GGAAAG": 3848, + "GGAATA": 3849, + "GGAATT": 3850, + "GGAATC": 3851, + "GGAATG": 3852, + "GGAACA": 3853, + "GGAACT": 3854, + "GGAACC": 3855, + "GGAACG": 3856, + "GGAAGA": 3857, + "GGAAGT": 3858, + "GGAAGC": 3859, + "GGAAGG": 3860, + "GGATAA": 3861, + "GGATAT": 3862, + "GGATAC": 3863, + "GGATAG": 3864, + "GGATTA": 3865, + "GGATTT": 3866, + "GGATTC": 3867, + "GGATTG": 3868, + "GGATCA": 3869, + "GGATCT": 3870, + "GGATCC": 3871, + "GGATCG": 3872, + "GGATGA": 3873, + "GGATGT": 3874, + "GGATGC": 3875, + "GGATGG": 3876, + "GGACAA": 3877, + "GGACAT": 3878, + "GGACAC": 3879, + "GGACAG": 3880, + "GGACTA": 3881, + "GGACTT": 3882, + "GGACTC": 3883, + "GGACTG": 3884, + "GGACCA": 3885, + "GGACCT": 3886, + "GGACCC": 3887, + "GGACCG": 3888, + "GGACGA": 3889, + "GGACGT": 3890, + "GGACGC": 3891, + "GGACGG": 3892, + "GGAGAA": 3893, + "GGAGAT": 3894, + "GGAGAC": 3895, + "GGAGAG": 3896, + "GGAGTA": 3897, + "GGAGTT": 3898, + "GGAGTC": 3899, + "GGAGTG": 3900, + "GGAGCA": 3901, + "GGAGCT": 3902, + "GGAGCC": 3903, + "GGAGCG": 3904, + "GGAGGA": 3905, + "GGAGGT": 3906, + "GGAGGC": 3907, + "GGAGGG": 3908, + "GGTAAA": 3909, + "GGTAAT": 3910, + "GGTAAC": 3911, + "GGTAAG": 3912, + "GGTATA": 3913, + "GGTATT": 3914, + "GGTATC": 3915, + "GGTATG": 3916, + "GGTACA": 3917, + "GGTACT": 3918, + "GGTACC": 3919, + "GGTACG": 3920, + "GGTAGA": 3921, + "GGTAGT": 3922, + "GGTAGC": 3923, + "GGTAGG": 3924, + "GGTTAA": 3925, + "GGTTAT": 3926, + "GGTTAC": 3927, + "GGTTAG": 3928, + "GGTTTA": 3929, + "GGTTTT": 3930, + "GGTTTC": 3931, + "GGTTTG": 3932, + "GGTTCA": 3933, + "GGTTCT": 3934, + "GGTTCC": 3935, + "GGTTCG": 3936, + "GGTTGA": 3937, + "GGTTGT": 3938, + "GGTTGC": 3939, + "GGTTGG": 3940, + "GGTCAA": 3941, + "GGTCAT": 3942, + "GGTCAC": 3943, + "GGTCAG": 3944, + "GGTCTA": 3945, + "GGTCTT": 3946, + "GGTCTC": 3947, + "GGTCTG": 3948, + "GGTCCA": 3949, + "GGTCCT": 3950, + "GGTCCC": 3951, + "GGTCCG": 3952, + "GGTCGA": 3953, + "GGTCGT": 3954, + "GGTCGC": 3955, + "GGTCGG": 3956, + "GGTGAA": 3957, + "GGTGAT": 3958, + "GGTGAC": 3959, + "GGTGAG": 3960, + "GGTGTA": 3961, + "GGTGTT": 3962, + "GGTGTC": 3963, + "GGTGTG": 3964, + "GGTGCA": 3965, + "GGTGCT": 3966, + "GGTGCC": 3967, + "GGTGCG": 3968, + "GGTGGA": 3969, + "GGTGGT": 3970, + "GGTGGC": 3971, + "GGTGGG": 3972, + "GGCAAA": 3973, + "GGCAAT": 3974, + "GGCAAC": 3975, + "GGCAAG": 3976, + "GGCATA": 3977, + "GGCATT": 3978, + "GGCATC": 3979, + "GGCATG": 3980, + "GGCACA": 3981, + "GGCACT": 3982, + "GGCACC": 3983, + "GGCACG": 3984, + "GGCAGA": 3985, + "GGCAGT": 3986, + "GGCAGC": 3987, + "GGCAGG": 3988, + "GGCTAA": 3989, + "GGCTAT": 3990, + "GGCTAC": 3991, + "GGCTAG": 3992, + "GGCTTA": 3993, + "GGCTTT": 3994, + "GGCTTC": 3995, + "GGCTTG": 3996, + "GGCTCA": 3997, + "GGCTCT": 3998, + "GGCTCC": 3999, + "GGCTCG": 4000, + "GGCTGA": 4001, + "GGCTGT": 4002, + "GGCTGC": 4003, + "GGCTGG": 4004, + "GGCCAA": 4005, + "GGCCAT": 4006, + "GGCCAC": 4007, + "GGCCAG": 4008, + "GGCCTA": 4009, + "GGCCTT": 4010, + "GGCCTC": 4011, + "GGCCTG": 4012, + "GGCCCA": 4013, + "GGCCCT": 4014, + "GGCCCC": 4015, + "GGCCCG": 4016, + "GGCCGA": 4017, + "GGCCGT": 4018, + "GGCCGC": 4019, + "GGCCGG": 4020, + "GGCGAA": 4021, + "GGCGAT": 4022, + "GGCGAC": 4023, + "GGCGAG": 4024, + "GGCGTA": 4025, + "GGCGTT": 4026, + "GGCGTC": 4027, + "GGCGTG": 4028, + "GGCGCA": 4029, + "GGCGCT": 4030, + "GGCGCC": 4031, + "GGCGCG": 4032, + "GGCGGA": 4033, + "GGCGGT": 4034, + "GGCGGC": 4035, + "GGCGGG": 4036, + "GGGAAA": 4037, + "GGGAAT": 4038, + "GGGAAC": 4039, + "GGGAAG": 4040, + "GGGATA": 4041, + "GGGATT": 4042, + "GGGATC": 4043, + "GGGATG": 4044, + "GGGACA": 4045, + "GGGACT": 4046, + "GGGACC": 4047, + "GGGACG": 4048, + "GGGAGA": 4049, + "GGGAGT": 4050, + "GGGAGC": 4051, + "GGGAGG": 4052, + "GGGTAA": 4053, + "GGGTAT": 4054, + "GGGTAC": 4055, + "GGGTAG": 4056, + "GGGTTA": 4057, + "GGGTTT": 4058, + "GGGTTC": 4059, + "GGGTTG": 4060, + "GGGTCA": 4061, + "GGGTCT": 4062, + "GGGTCC": 4063, + "GGGTCG": 4064, + "GGGTGA": 4065, + "GGGTGT": 4066, + "GGGTGC": 4067, + "GGGTGG": 4068, + "GGGCAA": 4069, + "GGGCAT": 4070, + "GGGCAC": 4071, + "GGGCAG": 4072, + "GGGCTA": 4073, + "GGGCTT": 4074, + "GGGCTC": 4075, + "GGGCTG": 4076, + "GGGCCA": 4077, + "GGGCCT": 4078, + "GGGCCC": 4079, + "GGGCCG": 4080, + "GGGCGA": 4081, + "GGGCGT": 4082, + "GGGCGC": 4083, + "GGGCGG": 4084, + "GGGGAA": 4085, + "GGGGAT": 4086, + "GGGGAC": 4087, + "GGGGAG": 4088, + "GGGGTA": 4089, + "GGGGTT": 4090, + "GGGGTC": 4091, + "GGGGTG": 4092, + "GGGGCA": 4093, + "GGGGCT": 4094, + "GGGGCC": 4095, + "GGGGCG": 4096, + "GGGGGA": 4097, + "GGGGGT": 4098, + "GGGGGC": 4099, + "GGGGGG": 4100 + } + } +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/tokenizer_config.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..604e2b1b29af1dcb3920d2828b6a90276e5e3d02 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/tokenizer_config.json @@ -0,0 +1,58 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "[UNK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "[CLS]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "[SEP]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "4": { + "content": "[MASK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "clean_up_tokenization_spaces": true, + "cls_token": "[CLS]", + "do_basic_tokenize": true, + "do_lower_case": false, + "mask_token": "[MASK]", + "max_len": 512, + "model_max_length": 512, + "never_split": null, + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "strip_accents": null, + "tokenize_chinese_chars": true, + "tokenizer_class": "BertTokenizer", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/trainer_state.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..8d1699810ba11c482a918365d9ab3cba85ec3b85 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/trainer_state.json @@ -0,0 +1,1433 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 20.0, + "eval_steps": 500, + "global_step": 5000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.1, + "grad_norm": 0.2218121588230133, + "learning_rate": 5.319148936170213e-05, + "loss": 0.6927, + "step": 25 + }, + { + "epoch": 0.2, + "grad_norm": 0.44851255416870117, + "learning_rate": 0.00010638297872340425, + "loss": 0.6929, + "step": 50 + }, + { + "epoch": 0.3, + "grad_norm": 0.2440175861120224, + "learning_rate": 0.00015957446808510637, + "loss": 0.6924, + "step": 75 + }, + { + "epoch": 0.4, + "grad_norm": 0.3422006666660309, + "learning_rate": 0.0002127659574468085, + "loss": 0.6922, + "step": 100 + }, + { + "epoch": 0.5, + "grad_norm": 0.267333447933197, + "learning_rate": 0.00026595744680851064, + "loss": 0.6913, + "step": 125 + }, + { + "epoch": 0.6, + "grad_norm": 0.45699772238731384, + "learning_rate": 0.00031914893617021275, + "loss": 0.6896, + "step": 150 + }, + { + "epoch": 0.7, + "grad_norm": 0.3004947304725647, + "learning_rate": 0.0003723404255319149, + "loss": 0.6887, + "step": 175 + }, + { + "epoch": 0.8, + "grad_norm": 0.33440759778022766, + "learning_rate": 0.0003992081821181128, + "loss": 0.6857, + "step": 200 + }, + { + "epoch": 0.9, + "grad_norm": 0.2918928861618042, + "learning_rate": 0.0003975585615308479, + "loss": 0.6882, + "step": 225 + }, + { + "epoch": 1.0, + "grad_norm": 0.2934085726737976, + "learning_rate": 0.00039590894094358297, + "loss": 0.6807, + "step": 250 + }, + { + "epoch": 1.1, + "grad_norm": 0.3364830017089844, + "learning_rate": 0.0003942593203563181, + "loss": 0.6848, + "step": 275 + }, + { + "epoch": 1.2, + "grad_norm": 0.8379159569740295, + "learning_rate": 0.0003926096997690532, + "loss": 0.677, + "step": 300 + }, + { + "epoch": 1.3, + "grad_norm": 0.37671706080436707, + "learning_rate": 0.00039096007918178817, + "loss": 0.6744, + "step": 325 + }, + { + "epoch": 1.4, + "grad_norm": 0.34329330921173096, + "learning_rate": 0.0003893104585945233, + "loss": 0.6733, + "step": 350 + }, + { + "epoch": 1.5, + "grad_norm": 0.4286085367202759, + "learning_rate": 0.0003876608380072583, + "loss": 0.6772, + "step": 375 + }, + { + "epoch": 1.6, + "grad_norm": 0.3576306998729706, + "learning_rate": 0.00038601121741999343, + "loss": 0.6799, + "step": 400 + }, + { + "epoch": 1.7, + "grad_norm": 0.4636435806751251, + "learning_rate": 0.0003843615968327285, + "loss": 0.6819, + "step": 425 + }, + { + "epoch": 1.8, + "grad_norm": 0.4552135765552521, + "learning_rate": 0.0003827119762454636, + "loss": 0.6731, + "step": 450 + }, + { + "epoch": 1.9, + "grad_norm": 0.3565954566001892, + "learning_rate": 0.00038106235565819863, + "loss": 0.6718, + "step": 475 + }, + { + "epoch": 2.0, + "grad_norm": 0.5451686978340149, + "learning_rate": 0.0003794127350709337, + "loss": 0.6858, + "step": 500 + }, + { + "epoch": 2.1, + "grad_norm": 0.3730587959289551, + "learning_rate": 0.0003777631144836688, + "loss": 0.667, + "step": 525 + }, + { + "epoch": 2.2, + "grad_norm": 0.6642417907714844, + "learning_rate": 0.00037611349389640383, + "loss": 0.6662, + "step": 550 + }, + { + "epoch": 2.3, + "grad_norm": 0.3889809250831604, + "learning_rate": 0.00037446387330913894, + "loss": 0.6718, + "step": 575 + }, + { + "epoch": 2.4, + "grad_norm": 0.47203826904296875, + "learning_rate": 0.000372814252721874, + "loss": 0.6687, + "step": 600 + }, + { + "epoch": 2.5, + "grad_norm": 0.5293862819671631, + "learning_rate": 0.00037116463213460903, + "loss": 0.6655, + "step": 625 + }, + { + "epoch": 2.6, + "grad_norm": 0.36724531650543213, + "learning_rate": 0.00036951501154734414, + "loss": 0.672, + "step": 650 + }, + { + "epoch": 2.7, + "grad_norm": 0.5733087062835693, + "learning_rate": 0.0003678653909600792, + "loss": 0.6683, + "step": 675 + }, + { + "epoch": 2.8, + "grad_norm": 0.6385205984115601, + "learning_rate": 0.0003662157703728143, + "loss": 0.67, + "step": 700 + }, + { + "epoch": 2.9, + "grad_norm": 0.6146281957626343, + "learning_rate": 0.00036456614978554934, + "loss": 0.6606, + "step": 725 + }, + { + "epoch": 3.0, + "grad_norm": 0.6535729169845581, + "learning_rate": 0.00036291652919828444, + "loss": 0.6691, + "step": 750 + }, + { + "epoch": 3.1, + "grad_norm": 0.5194367170333862, + "learning_rate": 0.00036126690861101944, + "loss": 0.6517, + "step": 775 + }, + { + "epoch": 3.2, + "grad_norm": 0.7116692066192627, + "learning_rate": 0.00035961728802375454, + "loss": 0.6501, + "step": 800 + }, + { + "epoch": 3.3, + "grad_norm": 0.5338636040687561, + "learning_rate": 0.0003579676674364896, + "loss": 0.6682, + "step": 825 + }, + { + "epoch": 3.4, + "grad_norm": 0.6912877559661865, + "learning_rate": 0.0003563180468492247, + "loss": 0.6632, + "step": 850 + }, + { + "epoch": 3.5, + "grad_norm": 0.6925108432769775, + "learning_rate": 0.0003546684262619598, + "loss": 0.653, + "step": 875 + }, + { + "epoch": 3.6, + "grad_norm": 0.5360299944877625, + "learning_rate": 0.00035301880567469485, + "loss": 0.6613, + "step": 900 + }, + { + "epoch": 3.7, + "grad_norm": 0.6487969160079956, + "learning_rate": 0.0003513691850874299, + "loss": 0.6627, + "step": 925 + }, + { + "epoch": 3.8, + "grad_norm": 1.7260371446609497, + "learning_rate": 0.00034971956450016495, + "loss": 0.6436, + "step": 950 + }, + { + "epoch": 3.9, + "grad_norm": 0.5852354168891907, + "learning_rate": 0.00034806994391290005, + "loss": 0.6613, + "step": 975 + }, + { + "epoch": 4.0, + "grad_norm": 0.7802712917327881, + "learning_rate": 0.0003464203233256351, + "loss": 0.658, + "step": 1000 + }, + { + "epoch": 4.1, + "grad_norm": 0.7335640788078308, + "learning_rate": 0.0003447707027383702, + "loss": 0.6476, + "step": 1025 + }, + { + "epoch": 4.2, + "grad_norm": 0.9540443420410156, + "learning_rate": 0.00034312108215110525, + "loss": 0.6334, + "step": 1050 + }, + { + "epoch": 4.3, + "grad_norm": 1.3698925971984863, + "learning_rate": 0.0003414714615638403, + "loss": 0.6512, + "step": 1075 + }, + { + "epoch": 4.4, + "grad_norm": 0.8282333016395569, + "learning_rate": 0.0003398218409765754, + "loss": 0.6556, + "step": 1100 + }, + { + "epoch": 4.5, + "grad_norm": 1.4414807558059692, + "learning_rate": 0.00033817222038931045, + "loss": 0.6628, + "step": 1125 + }, + { + "epoch": 4.6, + "grad_norm": 0.5493957996368408, + "learning_rate": 0.00033652259980204556, + "loss": 0.6461, + "step": 1150 + }, + { + "epoch": 4.7, + "grad_norm": 1.2343393564224243, + "learning_rate": 0.0003348729792147806, + "loss": 0.6464, + "step": 1175 + }, + { + "epoch": 4.8, + "grad_norm": 0.7163398265838623, + "learning_rate": 0.0003332233586275157, + "loss": 0.6518, + "step": 1200 + }, + { + "epoch": 4.9, + "grad_norm": 0.7878938317298889, + "learning_rate": 0.00033157373804025076, + "loss": 0.6502, + "step": 1225 + }, + { + "epoch": 5.0, + "grad_norm": 0.952571451663971, + "learning_rate": 0.0003299241174529858, + "loss": 0.6463, + "step": 1250 + }, + { + "epoch": 5.1, + "grad_norm": 0.6535393595695496, + "learning_rate": 0.0003282744968657209, + "loss": 0.6361, + "step": 1275 + }, + { + "epoch": 5.2, + "grad_norm": 0.8239464163780212, + "learning_rate": 0.00032662487627845596, + "loss": 0.6264, + "step": 1300 + }, + { + "epoch": 5.3, + "grad_norm": 1.4500653743743896, + "learning_rate": 0.00032497525569119106, + "loss": 0.6544, + "step": 1325 + }, + { + "epoch": 5.4, + "grad_norm": 1.1515296697616577, + "learning_rate": 0.0003233256351039261, + "loss": 0.6629, + "step": 1350 + }, + { + "epoch": 5.5, + "grad_norm": 0.6899295449256897, + "learning_rate": 0.00032167601451666116, + "loss": 0.6453, + "step": 1375 + }, + { + "epoch": 5.6, + "grad_norm": 1.1782217025756836, + "learning_rate": 0.00032002639392939627, + "loss": 0.6362, + "step": 1400 + }, + { + "epoch": 5.7, + "grad_norm": 0.6971372961997986, + "learning_rate": 0.0003183767733421313, + "loss": 0.6593, + "step": 1425 + }, + { + "epoch": 5.8, + "grad_norm": 0.7293921709060669, + "learning_rate": 0.0003167271527548664, + "loss": 0.6318, + "step": 1450 + }, + { + "epoch": 5.9, + "grad_norm": 0.6860108375549316, + "learning_rate": 0.00031507753216760147, + "loss": 0.6402, + "step": 1475 + }, + { + "epoch": 6.0, + "grad_norm": 1.330344319343567, + "learning_rate": 0.00031342791158033657, + "loss": 0.6276, + "step": 1500 + }, + { + "epoch": 6.1, + "grad_norm": 0.7261562943458557, + "learning_rate": 0.00031177829099307157, + "loss": 0.6369, + "step": 1525 + }, + { + "epoch": 6.2, + "grad_norm": 1.1324542760849, + "learning_rate": 0.00031012867040580667, + "loss": 0.6228, + "step": 1550 + }, + { + "epoch": 6.3, + "grad_norm": 0.7002120614051819, + "learning_rate": 0.0003084790498185417, + "loss": 0.6389, + "step": 1575 + }, + { + "epoch": 6.4, + "grad_norm": 1.3143397569656372, + "learning_rate": 0.0003068294292312768, + "loss": 0.6362, + "step": 1600 + }, + { + "epoch": 6.5, + "grad_norm": 1.6375401020050049, + "learning_rate": 0.0003051798086440119, + "loss": 0.6348, + "step": 1625 + }, + { + "epoch": 6.6, + "grad_norm": 1.4343605041503906, + "learning_rate": 0.000303530188056747, + "loss": 0.6496, + "step": 1650 + }, + { + "epoch": 6.7, + "grad_norm": 0.815762996673584, + "learning_rate": 0.000301880567469482, + "loss": 0.6407, + "step": 1675 + }, + { + "epoch": 6.8, + "grad_norm": 1.068608045578003, + "learning_rate": 0.0003002309468822171, + "loss": 0.6417, + "step": 1700 + }, + { + "epoch": 6.9, + "grad_norm": 0.8775753974914551, + "learning_rate": 0.0002985813262949522, + "loss": 0.6281, + "step": 1725 + }, + { + "epoch": 7.0, + "grad_norm": 1.041817545890808, + "learning_rate": 0.00029693170570768723, + "loss": 0.6485, + "step": 1750 + }, + { + "epoch": 7.1, + "grad_norm": 0.7135840058326721, + "learning_rate": 0.00029528208512042233, + "loss": 0.6192, + "step": 1775 + }, + { + "epoch": 7.2, + "grad_norm": 1.5523930788040161, + "learning_rate": 0.0002936324645331574, + "loss": 0.636, + "step": 1800 + }, + { + "epoch": 7.3, + "grad_norm": 0.7677174210548401, + "learning_rate": 0.00029198284394589243, + "loss": 0.6213, + "step": 1825 + }, + { + "epoch": 7.4, + "grad_norm": 0.8086565732955933, + "learning_rate": 0.00029033322335862753, + "loss": 0.6292, + "step": 1850 + }, + { + "epoch": 7.5, + "grad_norm": 3.1128971576690674, + "learning_rate": 0.0002886836027713626, + "loss": 0.648, + "step": 1875 + }, + { + "epoch": 7.6, + "grad_norm": 0.9243882894515991, + "learning_rate": 0.0002870339821840977, + "loss": 0.6286, + "step": 1900 + }, + { + "epoch": 7.7, + "grad_norm": 1.8203461170196533, + "learning_rate": 0.00028538436159683273, + "loss": 0.6301, + "step": 1925 + }, + { + "epoch": 7.8, + "grad_norm": 1.0924643278121948, + "learning_rate": 0.00028373474100956784, + "loss": 0.6378, + "step": 1950 + }, + { + "epoch": 7.9, + "grad_norm": 1.0839996337890625, + "learning_rate": 0.0002820851204223029, + "loss": 0.6265, + "step": 1975 + }, + { + "epoch": 8.0, + "grad_norm": 0.7143226861953735, + "learning_rate": 0.00028043549983503794, + "loss": 0.649, + "step": 2000 + }, + { + "epoch": 8.1, + "grad_norm": 0.8778955936431885, + "learning_rate": 0.00027878587924777304, + "loss": 0.6411, + "step": 2025 + }, + { + "epoch": 8.2, + "grad_norm": 1.644904613494873, + "learning_rate": 0.0002771362586605081, + "loss": 0.6168, + "step": 2050 + }, + { + "epoch": 8.3, + "grad_norm": 0.7915159463882446, + "learning_rate": 0.0002754866380732432, + "loss": 0.6324, + "step": 2075 + }, + { + "epoch": 8.4, + "grad_norm": 1.1212801933288574, + "learning_rate": 0.00027383701748597824, + "loss": 0.6232, + "step": 2100 + }, + { + "epoch": 8.5, + "grad_norm": 0.7717731595039368, + "learning_rate": 0.0002721873968987133, + "loss": 0.624, + "step": 2125 + }, + { + "epoch": 8.6, + "grad_norm": 1.0577210187911987, + "learning_rate": 0.00027053777631144834, + "loss": 0.6296, + "step": 2150 + }, + { + "epoch": 8.7, + "grad_norm": 1.0531567335128784, + "learning_rate": 0.00026888815572418344, + "loss": 0.6345, + "step": 2175 + }, + { + "epoch": 8.8, + "grad_norm": 2.8983640670776367, + "learning_rate": 0.00026723853513691855, + "loss": 0.6142, + "step": 2200 + }, + { + "epoch": 8.9, + "grad_norm": 0.7773396968841553, + "learning_rate": 0.0002655889145496536, + "loss": 0.634, + "step": 2225 + }, + { + "epoch": 9.0, + "grad_norm": 1.0587151050567627, + "learning_rate": 0.0002639392939623887, + "loss": 0.6326, + "step": 2250 + }, + { + "epoch": 9.1, + "grad_norm": 0.8009796738624573, + "learning_rate": 0.0002622896733751237, + "loss": 0.6175, + "step": 2275 + }, + { + "epoch": 9.2, + "grad_norm": 1.2530399560928345, + "learning_rate": 0.0002606400527878588, + "loss": 0.6239, + "step": 2300 + }, + { + "epoch": 9.3, + "grad_norm": 0.7976368069648743, + "learning_rate": 0.00025899043220059385, + "loss": 0.6452, + "step": 2325 + }, + { + "epoch": 9.4, + "grad_norm": 1.455032229423523, + "learning_rate": 0.00025734081161332895, + "loss": 0.6351, + "step": 2350 + }, + { + "epoch": 9.5, + "grad_norm": 0.8916760087013245, + "learning_rate": 0.000255691191026064, + "loss": 0.5986, + "step": 2375 + }, + { + "epoch": 9.6, + "grad_norm": 1.104746699333191, + "learning_rate": 0.0002540415704387991, + "loss": 0.6104, + "step": 2400 + }, + { + "epoch": 9.7, + "grad_norm": 0.9449218511581421, + "learning_rate": 0.00025239194985153415, + "loss": 0.6262, + "step": 2425 + }, + { + "epoch": 9.8, + "grad_norm": 2.0286591053009033, + "learning_rate": 0.0002507423292642692, + "loss": 0.6124, + "step": 2450 + }, + { + "epoch": 9.9, + "grad_norm": 2.273597002029419, + "learning_rate": 0.0002490927086770043, + "loss": 0.6164, + "step": 2475 + }, + { + "epoch": 10.0, + "grad_norm": 0.8174235224723816, + "learning_rate": 0.00024744308808973936, + "loss": 0.6254, + "step": 2500 + }, + { + "epoch": 10.1, + "grad_norm": 1.0447920560836792, + "learning_rate": 0.00024579346750247446, + "loss": 0.6187, + "step": 2525 + }, + { + "epoch": 10.2, + "grad_norm": 1.0433655977249146, + "learning_rate": 0.0002441438469152095, + "loss": 0.6183, + "step": 2550 + }, + { + "epoch": 10.3, + "grad_norm": 1.6479028463363647, + "learning_rate": 0.00024249422632794456, + "loss": 0.6353, + "step": 2575 + }, + { + "epoch": 10.4, + "grad_norm": 0.8032572865486145, + "learning_rate": 0.00024084460574067963, + "loss": 0.6206, + "step": 2600 + }, + { + "epoch": 10.5, + "grad_norm": 1.0675058364868164, + "learning_rate": 0.0002391949851534147, + "loss": 0.6168, + "step": 2625 + }, + { + "epoch": 10.6, + "grad_norm": 1.2228609323501587, + "learning_rate": 0.0002375453645661498, + "loss": 0.6007, + "step": 2650 + }, + { + "epoch": 10.7, + "grad_norm": 1.1027027368545532, + "learning_rate": 0.00023589574397888486, + "loss": 0.6227, + "step": 2675 + }, + { + "epoch": 10.8, + "grad_norm": 1.2949633598327637, + "learning_rate": 0.00023424612339161997, + "loss": 0.6179, + "step": 2700 + }, + { + "epoch": 10.9, + "grad_norm": 1.235032320022583, + "learning_rate": 0.000232596502804355, + "loss": 0.6216, + "step": 2725 + }, + { + "epoch": 11.0, + "grad_norm": 1.0087182521820068, + "learning_rate": 0.00023094688221709007, + "loss": 0.6286, + "step": 2750 + }, + { + "epoch": 11.1, + "grad_norm": 1.301207423210144, + "learning_rate": 0.00022929726162982514, + "loss": 0.5953, + "step": 2775 + }, + { + "epoch": 11.2, + "grad_norm": 1.6798988580703735, + "learning_rate": 0.00022764764104256022, + "loss": 0.6331, + "step": 2800 + }, + { + "epoch": 11.3, + "grad_norm": 1.1364004611968994, + "learning_rate": 0.0002259980204552953, + "loss": 0.6047, + "step": 2825 + }, + { + "epoch": 11.4, + "grad_norm": 0.9963313341140747, + "learning_rate": 0.00022434839986803037, + "loss": 0.6192, + "step": 2850 + }, + { + "epoch": 11.5, + "grad_norm": 1.5672775506973267, + "learning_rate": 0.00022269877928076542, + "loss": 0.6063, + "step": 2875 + }, + { + "epoch": 11.6, + "grad_norm": 1.4239299297332764, + "learning_rate": 0.0002210491586935005, + "loss": 0.6058, + "step": 2900 + }, + { + "epoch": 11.7, + "grad_norm": 1.5440959930419922, + "learning_rate": 0.00021939953810623557, + "loss": 0.6239, + "step": 2925 + }, + { + "epoch": 11.8, + "grad_norm": 1.35140860080719, + "learning_rate": 0.00021774991751897065, + "loss": 0.6241, + "step": 2950 + }, + { + "epoch": 11.9, + "grad_norm": 1.9355136156082153, + "learning_rate": 0.00021610029693170573, + "loss": 0.6135, + "step": 2975 + }, + { + "epoch": 12.0, + "grad_norm": 1.134037733078003, + "learning_rate": 0.0002144506763444408, + "loss": 0.6142, + "step": 3000 + }, + { + "epoch": 12.1, + "grad_norm": 1.094366192817688, + "learning_rate": 0.00021280105575717582, + "loss": 0.6226, + "step": 3025 + }, + { + "epoch": 12.2, + "grad_norm": 1.1197398900985718, + "learning_rate": 0.00021115143516991093, + "loss": 0.6073, + "step": 3050 + }, + { + "epoch": 12.3, + "grad_norm": 1.240188717842102, + "learning_rate": 0.000209501814582646, + "loss": 0.6265, + "step": 3075 + }, + { + "epoch": 12.4, + "grad_norm": 0.9774200916290283, + "learning_rate": 0.00020785219399538108, + "loss": 0.6361, + "step": 3100 + }, + { + "epoch": 12.5, + "grad_norm": 0.902847409248352, + "learning_rate": 0.00020620257340811616, + "loss": 0.6129, + "step": 3125 + }, + { + "epoch": 12.6, + "grad_norm": 0.8658815622329712, + "learning_rate": 0.00020455295282085123, + "loss": 0.5953, + "step": 3150 + }, + { + "epoch": 12.7, + "grad_norm": 1.4172385931015015, + "learning_rate": 0.00020290333223358626, + "loss": 0.6158, + "step": 3175 + }, + { + "epoch": 12.8, + "grad_norm": 2.6393818855285645, + "learning_rate": 0.00020125371164632133, + "loss": 0.5978, + "step": 3200 + }, + { + "epoch": 12.9, + "grad_norm": 2.4407801628112793, + "learning_rate": 0.0001996040910590564, + "loss": 0.6178, + "step": 3225 + }, + { + "epoch": 13.0, + "grad_norm": 1.2797659635543823, + "learning_rate": 0.00019795447047179148, + "loss": 0.5849, + "step": 3250 + }, + { + "epoch": 13.1, + "grad_norm": 1.590224266052246, + "learning_rate": 0.0001963048498845266, + "loss": 0.5983, + "step": 3275 + }, + { + "epoch": 13.2, + "grad_norm": 1.155685544013977, + "learning_rate": 0.00019465522929726164, + "loss": 0.6301, + "step": 3300 + }, + { + "epoch": 13.3, + "grad_norm": 1.701331377029419, + "learning_rate": 0.00019300560870999671, + "loss": 0.5865, + "step": 3325 + }, + { + "epoch": 13.4, + "grad_norm": 1.2553105354309082, + "learning_rate": 0.0001913559881227318, + "loss": 0.6113, + "step": 3350 + }, + { + "epoch": 13.5, + "grad_norm": 1.495653510093689, + "learning_rate": 0.00018970636753546684, + "loss": 0.6001, + "step": 3375 + }, + { + "epoch": 13.6, + "grad_norm": 1.2800766229629517, + "learning_rate": 0.00018805674694820192, + "loss": 0.6013, + "step": 3400 + }, + { + "epoch": 13.7, + "grad_norm": 1.0035955905914307, + "learning_rate": 0.000186407126360937, + "loss": 0.621, + "step": 3425 + }, + { + "epoch": 13.8, + "grad_norm": 1.0515904426574707, + "learning_rate": 0.00018475750577367207, + "loss": 0.6297, + "step": 3450 + }, + { + "epoch": 13.9, + "grad_norm": 2.071884870529175, + "learning_rate": 0.00018310788518640715, + "loss": 0.6004, + "step": 3475 + }, + { + "epoch": 14.0, + "grad_norm": 1.4622681140899658, + "learning_rate": 0.00018145826459914222, + "loss": 0.6021, + "step": 3500 + }, + { + "epoch": 14.1, + "grad_norm": 1.1860116720199585, + "learning_rate": 0.00017980864401187727, + "loss": 0.602, + "step": 3525 + }, + { + "epoch": 14.2, + "grad_norm": 1.0958752632141113, + "learning_rate": 0.00017815902342461235, + "loss": 0.6084, + "step": 3550 + }, + { + "epoch": 14.3, + "grad_norm": 1.4344367980957031, + "learning_rate": 0.00017650940283734742, + "loss": 0.6332, + "step": 3575 + }, + { + "epoch": 14.4, + "grad_norm": 1.1095603704452515, + "learning_rate": 0.00017485978225008247, + "loss": 0.5945, + "step": 3600 + }, + { + "epoch": 14.5, + "grad_norm": 1.416272521018982, + "learning_rate": 0.00017321016166281755, + "loss": 0.5964, + "step": 3625 + }, + { + "epoch": 14.6, + "grad_norm": 0.9727759957313538, + "learning_rate": 0.00017156054107555263, + "loss": 0.6231, + "step": 3650 + }, + { + "epoch": 14.7, + "grad_norm": 0.9979771375656128, + "learning_rate": 0.0001699109204882877, + "loss": 0.5849, + "step": 3675 + }, + { + "epoch": 14.8, + "grad_norm": 1.2574740648269653, + "learning_rate": 0.00016826129990102278, + "loss": 0.6076, + "step": 3700 + }, + { + "epoch": 14.9, + "grad_norm": 1.1657873392105103, + "learning_rate": 0.00016661167931375785, + "loss": 0.6111, + "step": 3725 + }, + { + "epoch": 15.0, + "grad_norm": 1.767681360244751, + "learning_rate": 0.0001649620587264929, + "loss": 0.6025, + "step": 3750 + }, + { + "epoch": 15.1, + "grad_norm": 1.3131200075149536, + "learning_rate": 0.00016331243813922798, + "loss": 0.6039, + "step": 3775 + }, + { + "epoch": 15.2, + "grad_norm": 1.1376745700836182, + "learning_rate": 0.00016166281755196306, + "loss": 0.5995, + "step": 3800 + }, + { + "epoch": 15.3, + "grad_norm": 1.2298001050949097, + "learning_rate": 0.00016001319696469813, + "loss": 0.602, + "step": 3825 + }, + { + "epoch": 15.4, + "grad_norm": 2.735769271850586, + "learning_rate": 0.0001583635763774332, + "loss": 0.5902, + "step": 3850 + }, + { + "epoch": 15.5, + "grad_norm": 1.3570619821548462, + "learning_rate": 0.00015671395579016829, + "loss": 0.5963, + "step": 3875 + }, + { + "epoch": 15.6, + "grad_norm": 1.6737253665924072, + "learning_rate": 0.00015506433520290334, + "loss": 0.6002, + "step": 3900 + }, + { + "epoch": 15.7, + "grad_norm": 1.386838674545288, + "learning_rate": 0.0001534147146156384, + "loss": 0.5905, + "step": 3925 + }, + { + "epoch": 15.8, + "grad_norm": 1.670160174369812, + "learning_rate": 0.0001517650940283735, + "loss": 0.6, + "step": 3950 + }, + { + "epoch": 15.9, + "grad_norm": 1.1611871719360352, + "learning_rate": 0.00015011547344110854, + "loss": 0.616, + "step": 3975 + }, + { + "epoch": 16.0, + "grad_norm": 0.9084995985031128, + "learning_rate": 0.00014846585285384361, + "loss": 0.6036, + "step": 4000 + }, + { + "epoch": 16.1, + "grad_norm": 2.1180953979492188, + "learning_rate": 0.0001468162322665787, + "loss": 0.5896, + "step": 4025 + }, + { + "epoch": 16.2, + "grad_norm": 1.9384114742279053, + "learning_rate": 0.00014516661167931377, + "loss": 0.584, + "step": 4050 + }, + { + "epoch": 16.3, + "grad_norm": 1.9472033977508545, + "learning_rate": 0.00014351699109204884, + "loss": 0.6059, + "step": 4075 + }, + { + "epoch": 16.4, + "grad_norm": 1.6221973896026611, + "learning_rate": 0.00014186737050478392, + "loss": 0.5934, + "step": 4100 + }, + { + "epoch": 16.5, + "grad_norm": 2.1453840732574463, + "learning_rate": 0.00014021774991751897, + "loss": 0.5846, + "step": 4125 + }, + { + "epoch": 16.6, + "grad_norm": 2.0782151222229004, + "learning_rate": 0.00013856812933025404, + "loss": 0.6131, + "step": 4150 + }, + { + "epoch": 16.7, + "grad_norm": 1.437988519668579, + "learning_rate": 0.00013691850874298912, + "loss": 0.6082, + "step": 4175 + }, + { + "epoch": 16.8, + "grad_norm": 1.7626062631607056, + "learning_rate": 0.00013526888815572417, + "loss": 0.5928, + "step": 4200 + }, + { + "epoch": 16.9, + "grad_norm": 1.0071998834609985, + "learning_rate": 0.00013361926756845927, + "loss": 0.6019, + "step": 4225 + }, + { + "epoch": 17.0, + "grad_norm": 1.1568998098373413, + "learning_rate": 0.00013196964698119435, + "loss": 0.6196, + "step": 4250 + }, + { + "epoch": 17.1, + "grad_norm": 1.9417431354522705, + "learning_rate": 0.0001303200263939294, + "loss": 0.607, + "step": 4275 + }, + { + "epoch": 17.2, + "grad_norm": 1.0894469022750854, + "learning_rate": 0.00012867040580666448, + "loss": 0.5874, + "step": 4300 + }, + { + "epoch": 17.3, + "grad_norm": 1.3170983791351318, + "learning_rate": 0.00012702078521939955, + "loss": 0.5988, + "step": 4325 + }, + { + "epoch": 17.4, + "grad_norm": 1.0083190202713013, + "learning_rate": 0.0001253711646321346, + "loss": 0.6021, + "step": 4350 + }, + { + "epoch": 17.5, + "grad_norm": 1.24309504032135, + "learning_rate": 0.00012372154404486968, + "loss": 0.6097, + "step": 4375 + }, + { + "epoch": 17.6, + "grad_norm": 1.5258588790893555, + "learning_rate": 0.00012207192345760475, + "loss": 0.5891, + "step": 4400 + }, + { + "epoch": 17.7, + "grad_norm": 2.491304636001587, + "learning_rate": 0.00012042230287033982, + "loss": 0.6049, + "step": 4425 + }, + { + "epoch": 17.8, + "grad_norm": 1.737715482711792, + "learning_rate": 0.0001187726822830749, + "loss": 0.5888, + "step": 4450 + }, + { + "epoch": 17.9, + "grad_norm": 2.217761754989624, + "learning_rate": 0.00011712306169580998, + "loss": 0.6075, + "step": 4475 + }, + { + "epoch": 18.0, + "grad_norm": 1.4966678619384766, + "learning_rate": 0.00011547344110854503, + "loss": 0.5962, + "step": 4500 + }, + { + "epoch": 18.1, + "grad_norm": 1.1560686826705933, + "learning_rate": 0.00011382382052128011, + "loss": 0.5654, + "step": 4525 + }, + { + "epoch": 18.2, + "grad_norm": 1.1144013404846191, + "learning_rate": 0.00011217419993401519, + "loss": 0.6001, + "step": 4550 + }, + { + "epoch": 18.3, + "grad_norm": 5.605530738830566, + "learning_rate": 0.00011052457934675025, + "loss": 0.5977, + "step": 4575 + }, + { + "epoch": 18.4, + "grad_norm": 1.0827083587646484, + "learning_rate": 0.00010887495875948532, + "loss": 0.5954, + "step": 4600 + }, + { + "epoch": 18.5, + "grad_norm": 3.0124075412750244, + "learning_rate": 0.0001072253381722204, + "loss": 0.5921, + "step": 4625 + }, + { + "epoch": 18.6, + "grad_norm": 1.5420926809310913, + "learning_rate": 0.00010557571758495546, + "loss": 0.5832, + "step": 4650 + }, + { + "epoch": 18.7, + "grad_norm": 1.5770180225372314, + "learning_rate": 0.00010392609699769054, + "loss": 0.6138, + "step": 4675 + }, + { + "epoch": 18.8, + "grad_norm": 1.9862656593322754, + "learning_rate": 0.00010227647641042562, + "loss": 0.595, + "step": 4700 + }, + { + "epoch": 18.9, + "grad_norm": 1.6087626218795776, + "learning_rate": 0.00010062685582316067, + "loss": 0.5965, + "step": 4725 + }, + { + "epoch": 19.0, + "grad_norm": 0.9878010749816895, + "learning_rate": 9.897723523589574e-05, + "loss": 0.5807, + "step": 4750 + }, + { + "epoch": 19.1, + "grad_norm": 1.3701050281524658, + "learning_rate": 9.732761464863082e-05, + "loss": 0.5953, + "step": 4775 + }, + { + "epoch": 19.2, + "grad_norm": 1.4488855600357056, + "learning_rate": 9.56779940613659e-05, + "loss": 0.578, + "step": 4800 + }, + { + "epoch": 19.3, + "grad_norm": 1.8450454473495483, + "learning_rate": 9.402837347410096e-05, + "loss": 0.5734, + "step": 4825 + }, + { + "epoch": 19.4, + "grad_norm": 1.2953295707702637, + "learning_rate": 9.237875288683603e-05, + "loss": 0.6188, + "step": 4850 + }, + { + "epoch": 19.5, + "grad_norm": 4.279994010925293, + "learning_rate": 9.072913229957111e-05, + "loss": 0.5712, + "step": 4875 + }, + { + "epoch": 19.6, + "grad_norm": 1.4677122831344604, + "learning_rate": 8.907951171230617e-05, + "loss": 0.5921, + "step": 4900 + }, + { + "epoch": 19.7, + "grad_norm": 2.559695243835449, + "learning_rate": 8.742989112504124e-05, + "loss": 0.599, + "step": 4925 + }, + { + "epoch": 19.8, + "grad_norm": 1.063706874847412, + "learning_rate": 8.578027053777631e-05, + "loss": 0.603, + "step": 4950 + }, + { + "epoch": 19.9, + "grad_norm": 1.1021010875701904, + "learning_rate": 8.413064995051139e-05, + "loss": 0.5841, + "step": 4975 + }, + { + "epoch": 20.0, + "grad_norm": 1.2500898838043213, + "learning_rate": 8.248102936324645e-05, + "loss": 0.5946, + "step": 5000 + } + ], + "logging_steps": 25, + "max_steps": 6250, + "num_input_tokens_seen": 0, + "num_train_epochs": 25, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 276905164800000.0, + "train_batch_size": 20, + "trial_name": null, + "trial_params": null +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/training_args.bin b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..540db3fbee786996d9705d768d31ba6c371187be --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dc7816b5f3fd38f01f9ba38a197ce070cde06e4c90e5260cb4814f34f628907d +size 5304 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/vocab.txt b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/vocab.txt new file mode 100644 index 0000000000000000000000000000000000000000..cc9ad53732e21ea9180f18e1fb67a38cc5702d4c --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-5000/vocab.txt @@ -0,0 +1,4101 @@ +[PAD] +[UNK] +[CLS] +[SEP] +[MASK] +AAAAAA +AAAAAT +AAAAAC +AAAAAG +AAAATA +AAAATT +AAAATC +AAAATG +AAAACA +AAAACT +AAAACC +AAAACG +AAAAGA +AAAAGT +AAAAGC +AAAAGG +AAATAA +AAATAT +AAATAC +AAATAG +AAATTA +AAATTT +AAATTC +AAATTG +AAATCA +AAATCT +AAATCC +AAATCG +AAATGA +AAATGT +AAATGC +AAATGG +AAACAA +AAACAT +AAACAC +AAACAG +AAACTA +AAACTT +AAACTC +AAACTG +AAACCA +AAACCT +AAACCC +AAACCG +AAACGA +AAACGT +AAACGC +AAACGG +AAAGAA +AAAGAT +AAAGAC +AAAGAG +AAAGTA +AAAGTT +AAAGTC +AAAGTG +AAAGCA +AAAGCT +AAAGCC +AAAGCG +AAAGGA +AAAGGT +AAAGGC +AAAGGG +AATAAA +AATAAT +AATAAC +AATAAG +AATATA +AATATT +AATATC +AATATG +AATACA +AATACT +AATACC +AATACG +AATAGA +AATAGT +AATAGC +AATAGG +AATTAA +AATTAT +AATTAC +AATTAG +AATTTA +AATTTT +AATTTC +AATTTG +AATTCA +AATTCT +AATTCC +AATTCG +AATTGA +AATTGT +AATTGC +AATTGG +AATCAA +AATCAT +AATCAC +AATCAG +AATCTA +AATCTT +AATCTC +AATCTG +AATCCA +AATCCT +AATCCC +AATCCG +AATCGA +AATCGT +AATCGC +AATCGG +AATGAA +AATGAT +AATGAC +AATGAG +AATGTA +AATGTT +AATGTC +AATGTG +AATGCA +AATGCT +AATGCC +AATGCG +AATGGA +AATGGT +AATGGC +AATGGG +AACAAA +AACAAT +AACAAC +AACAAG +AACATA +AACATT +AACATC +AACATG +AACACA +AACACT +AACACC +AACACG +AACAGA +AACAGT +AACAGC +AACAGG +AACTAA +AACTAT +AACTAC +AACTAG +AACTTA +AACTTT +AACTTC +AACTTG +AACTCA +AACTCT +AACTCC +AACTCG +AACTGA +AACTGT +AACTGC +AACTGG +AACCAA +AACCAT +AACCAC +AACCAG +AACCTA +AACCTT +AACCTC +AACCTG +AACCCA +AACCCT +AACCCC +AACCCG +AACCGA +AACCGT +AACCGC +AACCGG +AACGAA +AACGAT +AACGAC +AACGAG +AACGTA +AACGTT +AACGTC +AACGTG +AACGCA +AACGCT +AACGCC +AACGCG +AACGGA +AACGGT +AACGGC +AACGGG +AAGAAA +AAGAAT +AAGAAC +AAGAAG +AAGATA +AAGATT +AAGATC +AAGATG +AAGACA +AAGACT +AAGACC +AAGACG +AAGAGA +AAGAGT +AAGAGC +AAGAGG +AAGTAA +AAGTAT +AAGTAC +AAGTAG +AAGTTA +AAGTTT +AAGTTC +AAGTTG +AAGTCA +AAGTCT +AAGTCC +AAGTCG +AAGTGA +AAGTGT +AAGTGC +AAGTGG +AAGCAA +AAGCAT +AAGCAC +AAGCAG +AAGCTA +AAGCTT +AAGCTC +AAGCTG +AAGCCA +AAGCCT +AAGCCC +AAGCCG +AAGCGA +AAGCGT +AAGCGC +AAGCGG +AAGGAA +AAGGAT +AAGGAC +AAGGAG +AAGGTA +AAGGTT +AAGGTC +AAGGTG +AAGGCA +AAGGCT +AAGGCC +AAGGCG +AAGGGA +AAGGGT +AAGGGC +AAGGGG +ATAAAA +ATAAAT +ATAAAC +ATAAAG +ATAATA +ATAATT +ATAATC +ATAATG +ATAACA +ATAACT +ATAACC +ATAACG +ATAAGA +ATAAGT +ATAAGC +ATAAGG +ATATAA +ATATAT +ATATAC +ATATAG +ATATTA +ATATTT +ATATTC +ATATTG +ATATCA +ATATCT +ATATCC +ATATCG +ATATGA +ATATGT +ATATGC +ATATGG +ATACAA +ATACAT +ATACAC +ATACAG +ATACTA +ATACTT +ATACTC +ATACTG +ATACCA +ATACCT +ATACCC +ATACCG +ATACGA +ATACGT +ATACGC +ATACGG +ATAGAA +ATAGAT +ATAGAC +ATAGAG +ATAGTA +ATAGTT +ATAGTC +ATAGTG +ATAGCA +ATAGCT +ATAGCC +ATAGCG +ATAGGA +ATAGGT +ATAGGC +ATAGGG +ATTAAA +ATTAAT +ATTAAC +ATTAAG +ATTATA +ATTATT +ATTATC +ATTATG +ATTACA +ATTACT +ATTACC +ATTACG +ATTAGA +ATTAGT +ATTAGC +ATTAGG +ATTTAA +ATTTAT +ATTTAC +ATTTAG +ATTTTA +ATTTTT +ATTTTC +ATTTTG +ATTTCA +ATTTCT +ATTTCC +ATTTCG +ATTTGA +ATTTGT +ATTTGC +ATTTGG +ATTCAA +ATTCAT +ATTCAC +ATTCAG +ATTCTA +ATTCTT +ATTCTC +ATTCTG +ATTCCA +ATTCCT +ATTCCC +ATTCCG +ATTCGA +ATTCGT +ATTCGC +ATTCGG +ATTGAA +ATTGAT +ATTGAC +ATTGAG +ATTGTA +ATTGTT +ATTGTC +ATTGTG +ATTGCA +ATTGCT +ATTGCC +ATTGCG +ATTGGA +ATTGGT +ATTGGC +ATTGGG +ATCAAA +ATCAAT +ATCAAC +ATCAAG +ATCATA +ATCATT +ATCATC +ATCATG +ATCACA +ATCACT +ATCACC +ATCACG +ATCAGA +ATCAGT +ATCAGC +ATCAGG +ATCTAA +ATCTAT +ATCTAC +ATCTAG +ATCTTA +ATCTTT +ATCTTC +ATCTTG +ATCTCA +ATCTCT +ATCTCC +ATCTCG +ATCTGA +ATCTGT +ATCTGC +ATCTGG +ATCCAA +ATCCAT +ATCCAC +ATCCAG +ATCCTA +ATCCTT +ATCCTC +ATCCTG +ATCCCA +ATCCCT +ATCCCC +ATCCCG +ATCCGA +ATCCGT +ATCCGC +ATCCGG +ATCGAA +ATCGAT +ATCGAC +ATCGAG +ATCGTA +ATCGTT +ATCGTC +ATCGTG +ATCGCA +ATCGCT +ATCGCC +ATCGCG +ATCGGA +ATCGGT +ATCGGC +ATCGGG +ATGAAA +ATGAAT +ATGAAC +ATGAAG +ATGATA +ATGATT +ATGATC +ATGATG +ATGACA +ATGACT +ATGACC +ATGACG +ATGAGA +ATGAGT +ATGAGC +ATGAGG +ATGTAA +ATGTAT +ATGTAC +ATGTAG +ATGTTA +ATGTTT +ATGTTC +ATGTTG +ATGTCA +ATGTCT +ATGTCC +ATGTCG +ATGTGA +ATGTGT +ATGTGC +ATGTGG +ATGCAA +ATGCAT +ATGCAC +ATGCAG +ATGCTA +ATGCTT +ATGCTC +ATGCTG +ATGCCA +ATGCCT +ATGCCC +ATGCCG +ATGCGA +ATGCGT +ATGCGC +ATGCGG +ATGGAA +ATGGAT +ATGGAC +ATGGAG +ATGGTA +ATGGTT +ATGGTC +ATGGTG +ATGGCA +ATGGCT +ATGGCC +ATGGCG +ATGGGA +ATGGGT +ATGGGC +ATGGGG +ACAAAA +ACAAAT +ACAAAC +ACAAAG +ACAATA +ACAATT +ACAATC +ACAATG +ACAACA +ACAACT +ACAACC +ACAACG +ACAAGA +ACAAGT +ACAAGC +ACAAGG +ACATAA +ACATAT +ACATAC +ACATAG +ACATTA +ACATTT +ACATTC +ACATTG +ACATCA +ACATCT +ACATCC +ACATCG +ACATGA +ACATGT +ACATGC +ACATGG +ACACAA +ACACAT +ACACAC +ACACAG +ACACTA +ACACTT +ACACTC +ACACTG +ACACCA +ACACCT +ACACCC +ACACCG +ACACGA +ACACGT +ACACGC +ACACGG +ACAGAA +ACAGAT +ACAGAC +ACAGAG +ACAGTA +ACAGTT +ACAGTC +ACAGTG +ACAGCA +ACAGCT +ACAGCC +ACAGCG +ACAGGA +ACAGGT +ACAGGC +ACAGGG +ACTAAA +ACTAAT +ACTAAC +ACTAAG +ACTATA +ACTATT +ACTATC +ACTATG +ACTACA +ACTACT +ACTACC +ACTACG +ACTAGA +ACTAGT +ACTAGC +ACTAGG +ACTTAA +ACTTAT +ACTTAC +ACTTAG +ACTTTA +ACTTTT +ACTTTC +ACTTTG +ACTTCA +ACTTCT +ACTTCC +ACTTCG +ACTTGA +ACTTGT +ACTTGC +ACTTGG +ACTCAA +ACTCAT +ACTCAC +ACTCAG +ACTCTA +ACTCTT +ACTCTC +ACTCTG +ACTCCA +ACTCCT +ACTCCC +ACTCCG +ACTCGA +ACTCGT +ACTCGC +ACTCGG +ACTGAA +ACTGAT +ACTGAC +ACTGAG +ACTGTA +ACTGTT +ACTGTC +ACTGTG +ACTGCA +ACTGCT +ACTGCC +ACTGCG +ACTGGA +ACTGGT +ACTGGC +ACTGGG +ACCAAA +ACCAAT +ACCAAC +ACCAAG +ACCATA +ACCATT +ACCATC +ACCATG +ACCACA +ACCACT +ACCACC +ACCACG +ACCAGA +ACCAGT +ACCAGC +ACCAGG +ACCTAA +ACCTAT +ACCTAC +ACCTAG +ACCTTA +ACCTTT +ACCTTC +ACCTTG +ACCTCA +ACCTCT +ACCTCC +ACCTCG +ACCTGA +ACCTGT +ACCTGC +ACCTGG +ACCCAA +ACCCAT +ACCCAC +ACCCAG +ACCCTA +ACCCTT +ACCCTC +ACCCTG +ACCCCA +ACCCCT +ACCCCC +ACCCCG +ACCCGA +ACCCGT +ACCCGC +ACCCGG +ACCGAA +ACCGAT +ACCGAC +ACCGAG +ACCGTA +ACCGTT +ACCGTC +ACCGTG +ACCGCA +ACCGCT +ACCGCC +ACCGCG +ACCGGA +ACCGGT +ACCGGC +ACCGGG +ACGAAA +ACGAAT +ACGAAC +ACGAAG +ACGATA +ACGATT +ACGATC +ACGATG +ACGACA +ACGACT +ACGACC +ACGACG +ACGAGA +ACGAGT +ACGAGC +ACGAGG +ACGTAA +ACGTAT +ACGTAC +ACGTAG +ACGTTA +ACGTTT +ACGTTC +ACGTTG +ACGTCA +ACGTCT +ACGTCC +ACGTCG +ACGTGA +ACGTGT +ACGTGC +ACGTGG +ACGCAA +ACGCAT +ACGCAC +ACGCAG +ACGCTA +ACGCTT +ACGCTC +ACGCTG +ACGCCA +ACGCCT +ACGCCC +ACGCCG +ACGCGA +ACGCGT +ACGCGC +ACGCGG +ACGGAA +ACGGAT +ACGGAC +ACGGAG +ACGGTA +ACGGTT +ACGGTC +ACGGTG +ACGGCA +ACGGCT +ACGGCC +ACGGCG +ACGGGA +ACGGGT +ACGGGC +ACGGGG +AGAAAA +AGAAAT +AGAAAC +AGAAAG +AGAATA +AGAATT +AGAATC +AGAATG +AGAACA +AGAACT +AGAACC +AGAACG +AGAAGA +AGAAGT +AGAAGC +AGAAGG +AGATAA +AGATAT +AGATAC +AGATAG +AGATTA +AGATTT +AGATTC +AGATTG +AGATCA +AGATCT +AGATCC +AGATCG +AGATGA +AGATGT +AGATGC +AGATGG +AGACAA +AGACAT +AGACAC +AGACAG +AGACTA +AGACTT +AGACTC +AGACTG +AGACCA +AGACCT +AGACCC +AGACCG +AGACGA +AGACGT +AGACGC +AGACGG +AGAGAA +AGAGAT +AGAGAC +AGAGAG +AGAGTA +AGAGTT +AGAGTC +AGAGTG +AGAGCA +AGAGCT +AGAGCC +AGAGCG +AGAGGA +AGAGGT +AGAGGC +AGAGGG +AGTAAA +AGTAAT +AGTAAC +AGTAAG +AGTATA +AGTATT +AGTATC +AGTATG +AGTACA +AGTACT +AGTACC +AGTACG +AGTAGA +AGTAGT +AGTAGC +AGTAGG +AGTTAA +AGTTAT +AGTTAC +AGTTAG +AGTTTA +AGTTTT +AGTTTC +AGTTTG +AGTTCA +AGTTCT +AGTTCC +AGTTCG +AGTTGA +AGTTGT +AGTTGC +AGTTGG +AGTCAA +AGTCAT +AGTCAC +AGTCAG +AGTCTA +AGTCTT +AGTCTC +AGTCTG +AGTCCA +AGTCCT +AGTCCC +AGTCCG +AGTCGA +AGTCGT +AGTCGC +AGTCGG +AGTGAA +AGTGAT +AGTGAC +AGTGAG +AGTGTA +AGTGTT +AGTGTC +AGTGTG +AGTGCA +AGTGCT +AGTGCC +AGTGCG +AGTGGA +AGTGGT +AGTGGC +AGTGGG +AGCAAA +AGCAAT +AGCAAC +AGCAAG +AGCATA +AGCATT +AGCATC +AGCATG +AGCACA +AGCACT +AGCACC +AGCACG +AGCAGA +AGCAGT +AGCAGC +AGCAGG +AGCTAA +AGCTAT +AGCTAC +AGCTAG +AGCTTA +AGCTTT +AGCTTC +AGCTTG +AGCTCA +AGCTCT +AGCTCC +AGCTCG +AGCTGA +AGCTGT +AGCTGC +AGCTGG +AGCCAA +AGCCAT +AGCCAC +AGCCAG +AGCCTA +AGCCTT +AGCCTC +AGCCTG +AGCCCA +AGCCCT +AGCCCC +AGCCCG +AGCCGA +AGCCGT +AGCCGC +AGCCGG +AGCGAA +AGCGAT +AGCGAC +AGCGAG +AGCGTA +AGCGTT +AGCGTC +AGCGTG +AGCGCA +AGCGCT +AGCGCC +AGCGCG +AGCGGA +AGCGGT +AGCGGC +AGCGGG +AGGAAA +AGGAAT +AGGAAC +AGGAAG +AGGATA +AGGATT +AGGATC +AGGATG +AGGACA +AGGACT +AGGACC +AGGACG +AGGAGA +AGGAGT +AGGAGC +AGGAGG +AGGTAA +AGGTAT +AGGTAC +AGGTAG +AGGTTA +AGGTTT +AGGTTC +AGGTTG +AGGTCA +AGGTCT +AGGTCC +AGGTCG +AGGTGA +AGGTGT +AGGTGC +AGGTGG +AGGCAA +AGGCAT +AGGCAC +AGGCAG +AGGCTA +AGGCTT +AGGCTC +AGGCTG +AGGCCA +AGGCCT +AGGCCC +AGGCCG +AGGCGA +AGGCGT +AGGCGC +AGGCGG +AGGGAA +AGGGAT +AGGGAC +AGGGAG +AGGGTA +AGGGTT +AGGGTC +AGGGTG +AGGGCA +AGGGCT +AGGGCC +AGGGCG +AGGGGA +AGGGGT +AGGGGC +AGGGGG +TAAAAA +TAAAAT +TAAAAC +TAAAAG +TAAATA +TAAATT +TAAATC +TAAATG +TAAACA +TAAACT +TAAACC +TAAACG +TAAAGA +TAAAGT +TAAAGC +TAAAGG +TAATAA +TAATAT +TAATAC +TAATAG +TAATTA +TAATTT +TAATTC +TAATTG +TAATCA +TAATCT +TAATCC +TAATCG +TAATGA +TAATGT +TAATGC +TAATGG +TAACAA +TAACAT +TAACAC +TAACAG +TAACTA +TAACTT +TAACTC +TAACTG +TAACCA +TAACCT +TAACCC +TAACCG +TAACGA +TAACGT +TAACGC +TAACGG +TAAGAA +TAAGAT +TAAGAC +TAAGAG +TAAGTA +TAAGTT +TAAGTC +TAAGTG +TAAGCA +TAAGCT +TAAGCC +TAAGCG +TAAGGA +TAAGGT +TAAGGC +TAAGGG +TATAAA +TATAAT +TATAAC +TATAAG +TATATA +TATATT +TATATC +TATATG +TATACA +TATACT +TATACC +TATACG +TATAGA +TATAGT +TATAGC +TATAGG +TATTAA +TATTAT +TATTAC +TATTAG +TATTTA +TATTTT +TATTTC +TATTTG +TATTCA +TATTCT +TATTCC +TATTCG +TATTGA +TATTGT +TATTGC +TATTGG +TATCAA +TATCAT +TATCAC +TATCAG +TATCTA +TATCTT +TATCTC +TATCTG +TATCCA +TATCCT +TATCCC +TATCCG +TATCGA +TATCGT +TATCGC +TATCGG +TATGAA +TATGAT +TATGAC +TATGAG +TATGTA +TATGTT +TATGTC +TATGTG +TATGCA +TATGCT +TATGCC +TATGCG +TATGGA +TATGGT +TATGGC +TATGGG +TACAAA +TACAAT +TACAAC +TACAAG +TACATA +TACATT +TACATC +TACATG +TACACA +TACACT +TACACC +TACACG +TACAGA +TACAGT +TACAGC +TACAGG +TACTAA +TACTAT +TACTAC +TACTAG +TACTTA +TACTTT +TACTTC +TACTTG +TACTCA +TACTCT +TACTCC +TACTCG +TACTGA +TACTGT +TACTGC +TACTGG +TACCAA +TACCAT +TACCAC +TACCAG +TACCTA +TACCTT +TACCTC +TACCTG +TACCCA +TACCCT +TACCCC +TACCCG +TACCGA +TACCGT +TACCGC +TACCGG +TACGAA +TACGAT +TACGAC +TACGAG +TACGTA +TACGTT +TACGTC +TACGTG +TACGCA +TACGCT +TACGCC +TACGCG +TACGGA +TACGGT +TACGGC +TACGGG +TAGAAA +TAGAAT +TAGAAC +TAGAAG +TAGATA +TAGATT +TAGATC +TAGATG +TAGACA +TAGACT +TAGACC +TAGACG +TAGAGA +TAGAGT +TAGAGC +TAGAGG +TAGTAA +TAGTAT +TAGTAC +TAGTAG +TAGTTA +TAGTTT +TAGTTC +TAGTTG +TAGTCA +TAGTCT +TAGTCC +TAGTCG +TAGTGA +TAGTGT +TAGTGC +TAGTGG +TAGCAA +TAGCAT +TAGCAC +TAGCAG +TAGCTA +TAGCTT +TAGCTC +TAGCTG +TAGCCA +TAGCCT +TAGCCC +TAGCCG +TAGCGA +TAGCGT +TAGCGC +TAGCGG +TAGGAA +TAGGAT +TAGGAC +TAGGAG +TAGGTA +TAGGTT +TAGGTC +TAGGTG +TAGGCA +TAGGCT +TAGGCC +TAGGCG +TAGGGA +TAGGGT +TAGGGC +TAGGGG +TTAAAA +TTAAAT +TTAAAC +TTAAAG +TTAATA +TTAATT +TTAATC +TTAATG +TTAACA +TTAACT +TTAACC +TTAACG +TTAAGA +TTAAGT +TTAAGC +TTAAGG +TTATAA +TTATAT +TTATAC +TTATAG +TTATTA +TTATTT +TTATTC +TTATTG +TTATCA +TTATCT +TTATCC +TTATCG +TTATGA +TTATGT +TTATGC +TTATGG +TTACAA +TTACAT +TTACAC +TTACAG +TTACTA +TTACTT +TTACTC +TTACTG +TTACCA +TTACCT +TTACCC +TTACCG +TTACGA +TTACGT +TTACGC +TTACGG +TTAGAA +TTAGAT +TTAGAC +TTAGAG +TTAGTA +TTAGTT +TTAGTC +TTAGTG +TTAGCA +TTAGCT +TTAGCC +TTAGCG +TTAGGA +TTAGGT +TTAGGC +TTAGGG +TTTAAA +TTTAAT +TTTAAC +TTTAAG +TTTATA +TTTATT +TTTATC +TTTATG +TTTACA +TTTACT +TTTACC +TTTACG +TTTAGA +TTTAGT +TTTAGC +TTTAGG +TTTTAA +TTTTAT +TTTTAC +TTTTAG +TTTTTA +TTTTTT +TTTTTC +TTTTTG +TTTTCA +TTTTCT +TTTTCC +TTTTCG +TTTTGA +TTTTGT +TTTTGC +TTTTGG +TTTCAA +TTTCAT +TTTCAC +TTTCAG +TTTCTA +TTTCTT +TTTCTC +TTTCTG +TTTCCA +TTTCCT +TTTCCC +TTTCCG +TTTCGA +TTTCGT +TTTCGC +TTTCGG +TTTGAA +TTTGAT +TTTGAC +TTTGAG +TTTGTA +TTTGTT +TTTGTC +TTTGTG +TTTGCA +TTTGCT +TTTGCC +TTTGCG +TTTGGA +TTTGGT +TTTGGC +TTTGGG +TTCAAA +TTCAAT +TTCAAC +TTCAAG +TTCATA +TTCATT +TTCATC +TTCATG +TTCACA +TTCACT +TTCACC +TTCACG +TTCAGA +TTCAGT +TTCAGC +TTCAGG +TTCTAA +TTCTAT +TTCTAC +TTCTAG +TTCTTA +TTCTTT +TTCTTC +TTCTTG +TTCTCA +TTCTCT +TTCTCC +TTCTCG +TTCTGA +TTCTGT +TTCTGC +TTCTGG +TTCCAA +TTCCAT +TTCCAC +TTCCAG +TTCCTA +TTCCTT +TTCCTC +TTCCTG +TTCCCA +TTCCCT +TTCCCC +TTCCCG +TTCCGA +TTCCGT +TTCCGC +TTCCGG +TTCGAA +TTCGAT +TTCGAC +TTCGAG +TTCGTA +TTCGTT +TTCGTC +TTCGTG +TTCGCA +TTCGCT +TTCGCC +TTCGCG +TTCGGA +TTCGGT +TTCGGC +TTCGGG +TTGAAA +TTGAAT +TTGAAC +TTGAAG +TTGATA +TTGATT +TTGATC +TTGATG +TTGACA +TTGACT +TTGACC +TTGACG +TTGAGA +TTGAGT +TTGAGC +TTGAGG +TTGTAA +TTGTAT +TTGTAC +TTGTAG +TTGTTA +TTGTTT +TTGTTC +TTGTTG +TTGTCA +TTGTCT +TTGTCC +TTGTCG +TTGTGA +TTGTGT +TTGTGC +TTGTGG +TTGCAA +TTGCAT +TTGCAC +TTGCAG +TTGCTA +TTGCTT +TTGCTC +TTGCTG +TTGCCA +TTGCCT +TTGCCC +TTGCCG +TTGCGA +TTGCGT +TTGCGC +TTGCGG +TTGGAA +TTGGAT +TTGGAC +TTGGAG +TTGGTA +TTGGTT +TTGGTC +TTGGTG +TTGGCA +TTGGCT +TTGGCC +TTGGCG +TTGGGA +TTGGGT +TTGGGC +TTGGGG +TCAAAA +TCAAAT +TCAAAC +TCAAAG +TCAATA +TCAATT +TCAATC +TCAATG +TCAACA +TCAACT +TCAACC +TCAACG +TCAAGA +TCAAGT +TCAAGC +TCAAGG +TCATAA +TCATAT +TCATAC +TCATAG +TCATTA +TCATTT +TCATTC +TCATTG +TCATCA +TCATCT +TCATCC +TCATCG +TCATGA +TCATGT +TCATGC +TCATGG +TCACAA +TCACAT +TCACAC +TCACAG +TCACTA +TCACTT +TCACTC +TCACTG +TCACCA +TCACCT +TCACCC +TCACCG +TCACGA +TCACGT +TCACGC +TCACGG +TCAGAA +TCAGAT +TCAGAC +TCAGAG +TCAGTA +TCAGTT +TCAGTC +TCAGTG +TCAGCA +TCAGCT +TCAGCC +TCAGCG +TCAGGA +TCAGGT +TCAGGC +TCAGGG +TCTAAA +TCTAAT +TCTAAC +TCTAAG +TCTATA +TCTATT +TCTATC +TCTATG +TCTACA +TCTACT +TCTACC +TCTACG +TCTAGA +TCTAGT +TCTAGC +TCTAGG +TCTTAA +TCTTAT +TCTTAC +TCTTAG +TCTTTA +TCTTTT +TCTTTC +TCTTTG +TCTTCA +TCTTCT +TCTTCC +TCTTCG +TCTTGA +TCTTGT +TCTTGC +TCTTGG +TCTCAA +TCTCAT +TCTCAC +TCTCAG +TCTCTA +TCTCTT +TCTCTC +TCTCTG +TCTCCA +TCTCCT +TCTCCC +TCTCCG +TCTCGA +TCTCGT +TCTCGC +TCTCGG +TCTGAA +TCTGAT +TCTGAC +TCTGAG +TCTGTA +TCTGTT +TCTGTC +TCTGTG +TCTGCA +TCTGCT +TCTGCC +TCTGCG +TCTGGA +TCTGGT +TCTGGC +TCTGGG +TCCAAA +TCCAAT +TCCAAC +TCCAAG +TCCATA +TCCATT +TCCATC +TCCATG +TCCACA +TCCACT +TCCACC +TCCACG +TCCAGA +TCCAGT +TCCAGC +TCCAGG +TCCTAA +TCCTAT +TCCTAC +TCCTAG +TCCTTA +TCCTTT +TCCTTC +TCCTTG +TCCTCA +TCCTCT +TCCTCC +TCCTCG +TCCTGA +TCCTGT +TCCTGC +TCCTGG +TCCCAA +TCCCAT +TCCCAC +TCCCAG +TCCCTA +TCCCTT +TCCCTC +TCCCTG +TCCCCA +TCCCCT +TCCCCC +TCCCCG +TCCCGA +TCCCGT +TCCCGC +TCCCGG +TCCGAA +TCCGAT +TCCGAC +TCCGAG +TCCGTA +TCCGTT +TCCGTC +TCCGTG +TCCGCA +TCCGCT +TCCGCC +TCCGCG +TCCGGA +TCCGGT +TCCGGC +TCCGGG +TCGAAA +TCGAAT +TCGAAC +TCGAAG +TCGATA +TCGATT +TCGATC +TCGATG +TCGACA +TCGACT +TCGACC +TCGACG +TCGAGA +TCGAGT +TCGAGC +TCGAGG +TCGTAA +TCGTAT +TCGTAC +TCGTAG +TCGTTA +TCGTTT +TCGTTC +TCGTTG +TCGTCA +TCGTCT +TCGTCC +TCGTCG +TCGTGA +TCGTGT +TCGTGC +TCGTGG +TCGCAA +TCGCAT +TCGCAC +TCGCAG +TCGCTA +TCGCTT +TCGCTC +TCGCTG +TCGCCA +TCGCCT +TCGCCC +TCGCCG +TCGCGA +TCGCGT +TCGCGC +TCGCGG +TCGGAA +TCGGAT +TCGGAC +TCGGAG +TCGGTA +TCGGTT +TCGGTC +TCGGTG +TCGGCA +TCGGCT +TCGGCC +TCGGCG +TCGGGA +TCGGGT +TCGGGC +TCGGGG +TGAAAA +TGAAAT +TGAAAC +TGAAAG +TGAATA +TGAATT +TGAATC +TGAATG +TGAACA +TGAACT +TGAACC +TGAACG +TGAAGA +TGAAGT +TGAAGC +TGAAGG +TGATAA +TGATAT +TGATAC +TGATAG +TGATTA +TGATTT +TGATTC +TGATTG +TGATCA +TGATCT +TGATCC +TGATCG +TGATGA +TGATGT +TGATGC +TGATGG +TGACAA +TGACAT +TGACAC +TGACAG +TGACTA +TGACTT +TGACTC +TGACTG +TGACCA +TGACCT +TGACCC +TGACCG +TGACGA +TGACGT +TGACGC +TGACGG +TGAGAA +TGAGAT +TGAGAC +TGAGAG +TGAGTA +TGAGTT +TGAGTC +TGAGTG +TGAGCA +TGAGCT +TGAGCC +TGAGCG +TGAGGA +TGAGGT +TGAGGC +TGAGGG +TGTAAA +TGTAAT +TGTAAC +TGTAAG +TGTATA +TGTATT +TGTATC +TGTATG +TGTACA +TGTACT +TGTACC +TGTACG +TGTAGA +TGTAGT +TGTAGC +TGTAGG +TGTTAA +TGTTAT +TGTTAC +TGTTAG +TGTTTA +TGTTTT +TGTTTC +TGTTTG +TGTTCA +TGTTCT +TGTTCC +TGTTCG +TGTTGA +TGTTGT +TGTTGC +TGTTGG +TGTCAA +TGTCAT +TGTCAC +TGTCAG +TGTCTA +TGTCTT +TGTCTC +TGTCTG +TGTCCA +TGTCCT +TGTCCC +TGTCCG +TGTCGA +TGTCGT +TGTCGC +TGTCGG +TGTGAA +TGTGAT +TGTGAC +TGTGAG +TGTGTA +TGTGTT +TGTGTC +TGTGTG +TGTGCA +TGTGCT +TGTGCC +TGTGCG +TGTGGA +TGTGGT +TGTGGC +TGTGGG +TGCAAA +TGCAAT +TGCAAC +TGCAAG +TGCATA +TGCATT +TGCATC +TGCATG +TGCACA +TGCACT +TGCACC +TGCACG +TGCAGA +TGCAGT +TGCAGC +TGCAGG +TGCTAA +TGCTAT +TGCTAC +TGCTAG +TGCTTA +TGCTTT +TGCTTC +TGCTTG +TGCTCA +TGCTCT +TGCTCC +TGCTCG +TGCTGA +TGCTGT +TGCTGC +TGCTGG +TGCCAA +TGCCAT +TGCCAC +TGCCAG +TGCCTA +TGCCTT +TGCCTC +TGCCTG +TGCCCA +TGCCCT +TGCCCC +TGCCCG +TGCCGA +TGCCGT +TGCCGC +TGCCGG +TGCGAA +TGCGAT +TGCGAC +TGCGAG +TGCGTA +TGCGTT +TGCGTC +TGCGTG +TGCGCA +TGCGCT +TGCGCC +TGCGCG +TGCGGA +TGCGGT +TGCGGC +TGCGGG +TGGAAA +TGGAAT +TGGAAC +TGGAAG +TGGATA +TGGATT +TGGATC +TGGATG +TGGACA +TGGACT +TGGACC +TGGACG +TGGAGA +TGGAGT +TGGAGC +TGGAGG +TGGTAA +TGGTAT +TGGTAC +TGGTAG +TGGTTA +TGGTTT +TGGTTC +TGGTTG +TGGTCA +TGGTCT +TGGTCC +TGGTCG +TGGTGA +TGGTGT +TGGTGC +TGGTGG +TGGCAA +TGGCAT +TGGCAC +TGGCAG +TGGCTA +TGGCTT +TGGCTC +TGGCTG +TGGCCA +TGGCCT +TGGCCC +TGGCCG +TGGCGA +TGGCGT +TGGCGC +TGGCGG +TGGGAA +TGGGAT +TGGGAC +TGGGAG +TGGGTA +TGGGTT +TGGGTC +TGGGTG +TGGGCA +TGGGCT +TGGGCC +TGGGCG +TGGGGA +TGGGGT +TGGGGC +TGGGGG +CAAAAA +CAAAAT +CAAAAC +CAAAAG +CAAATA +CAAATT +CAAATC +CAAATG +CAAACA +CAAACT +CAAACC +CAAACG +CAAAGA +CAAAGT +CAAAGC +CAAAGG +CAATAA +CAATAT +CAATAC +CAATAG +CAATTA +CAATTT +CAATTC +CAATTG +CAATCA +CAATCT +CAATCC +CAATCG +CAATGA +CAATGT +CAATGC +CAATGG +CAACAA +CAACAT +CAACAC +CAACAG +CAACTA +CAACTT +CAACTC +CAACTG +CAACCA +CAACCT +CAACCC +CAACCG +CAACGA +CAACGT +CAACGC +CAACGG +CAAGAA +CAAGAT +CAAGAC +CAAGAG +CAAGTA +CAAGTT +CAAGTC +CAAGTG +CAAGCA +CAAGCT +CAAGCC +CAAGCG +CAAGGA +CAAGGT +CAAGGC +CAAGGG +CATAAA +CATAAT +CATAAC +CATAAG +CATATA +CATATT +CATATC +CATATG +CATACA +CATACT +CATACC +CATACG +CATAGA +CATAGT +CATAGC +CATAGG +CATTAA +CATTAT +CATTAC +CATTAG +CATTTA +CATTTT +CATTTC +CATTTG +CATTCA +CATTCT +CATTCC +CATTCG +CATTGA +CATTGT +CATTGC +CATTGG +CATCAA +CATCAT +CATCAC +CATCAG +CATCTA +CATCTT +CATCTC +CATCTG +CATCCA +CATCCT +CATCCC +CATCCG +CATCGA +CATCGT +CATCGC +CATCGG +CATGAA +CATGAT +CATGAC +CATGAG +CATGTA +CATGTT +CATGTC +CATGTG +CATGCA +CATGCT +CATGCC +CATGCG +CATGGA +CATGGT +CATGGC +CATGGG +CACAAA +CACAAT +CACAAC +CACAAG +CACATA +CACATT +CACATC +CACATG +CACACA +CACACT +CACACC +CACACG +CACAGA +CACAGT +CACAGC +CACAGG +CACTAA +CACTAT +CACTAC +CACTAG +CACTTA +CACTTT +CACTTC +CACTTG +CACTCA +CACTCT +CACTCC +CACTCG +CACTGA +CACTGT +CACTGC +CACTGG +CACCAA +CACCAT +CACCAC +CACCAG +CACCTA +CACCTT +CACCTC +CACCTG +CACCCA +CACCCT +CACCCC +CACCCG +CACCGA +CACCGT +CACCGC +CACCGG +CACGAA +CACGAT +CACGAC +CACGAG +CACGTA +CACGTT +CACGTC +CACGTG +CACGCA +CACGCT +CACGCC +CACGCG +CACGGA +CACGGT +CACGGC +CACGGG +CAGAAA +CAGAAT +CAGAAC +CAGAAG +CAGATA +CAGATT +CAGATC +CAGATG +CAGACA +CAGACT +CAGACC +CAGACG +CAGAGA +CAGAGT +CAGAGC +CAGAGG +CAGTAA +CAGTAT +CAGTAC +CAGTAG +CAGTTA +CAGTTT +CAGTTC +CAGTTG +CAGTCA +CAGTCT +CAGTCC +CAGTCG +CAGTGA +CAGTGT +CAGTGC +CAGTGG +CAGCAA +CAGCAT +CAGCAC +CAGCAG +CAGCTA +CAGCTT +CAGCTC +CAGCTG +CAGCCA +CAGCCT +CAGCCC +CAGCCG +CAGCGA +CAGCGT +CAGCGC +CAGCGG +CAGGAA +CAGGAT +CAGGAC +CAGGAG +CAGGTA +CAGGTT +CAGGTC +CAGGTG +CAGGCA +CAGGCT +CAGGCC +CAGGCG +CAGGGA +CAGGGT +CAGGGC +CAGGGG +CTAAAA +CTAAAT +CTAAAC +CTAAAG +CTAATA +CTAATT +CTAATC +CTAATG +CTAACA +CTAACT +CTAACC +CTAACG +CTAAGA +CTAAGT +CTAAGC +CTAAGG +CTATAA +CTATAT +CTATAC +CTATAG +CTATTA +CTATTT +CTATTC +CTATTG +CTATCA +CTATCT +CTATCC +CTATCG +CTATGA +CTATGT +CTATGC +CTATGG +CTACAA +CTACAT +CTACAC +CTACAG +CTACTA +CTACTT +CTACTC +CTACTG +CTACCA +CTACCT +CTACCC +CTACCG +CTACGA +CTACGT +CTACGC +CTACGG +CTAGAA +CTAGAT +CTAGAC +CTAGAG +CTAGTA +CTAGTT +CTAGTC +CTAGTG +CTAGCA +CTAGCT +CTAGCC +CTAGCG +CTAGGA +CTAGGT +CTAGGC +CTAGGG +CTTAAA +CTTAAT +CTTAAC +CTTAAG +CTTATA +CTTATT +CTTATC +CTTATG +CTTACA +CTTACT +CTTACC +CTTACG +CTTAGA +CTTAGT +CTTAGC +CTTAGG +CTTTAA +CTTTAT +CTTTAC +CTTTAG +CTTTTA +CTTTTT +CTTTTC +CTTTTG +CTTTCA +CTTTCT +CTTTCC +CTTTCG +CTTTGA +CTTTGT +CTTTGC +CTTTGG +CTTCAA +CTTCAT +CTTCAC +CTTCAG +CTTCTA +CTTCTT +CTTCTC +CTTCTG +CTTCCA +CTTCCT +CTTCCC +CTTCCG +CTTCGA +CTTCGT +CTTCGC +CTTCGG +CTTGAA +CTTGAT +CTTGAC +CTTGAG +CTTGTA +CTTGTT +CTTGTC +CTTGTG +CTTGCA +CTTGCT +CTTGCC +CTTGCG +CTTGGA +CTTGGT +CTTGGC +CTTGGG +CTCAAA +CTCAAT +CTCAAC +CTCAAG +CTCATA +CTCATT +CTCATC +CTCATG +CTCACA +CTCACT +CTCACC +CTCACG +CTCAGA +CTCAGT +CTCAGC +CTCAGG +CTCTAA +CTCTAT +CTCTAC +CTCTAG +CTCTTA +CTCTTT +CTCTTC +CTCTTG +CTCTCA +CTCTCT +CTCTCC +CTCTCG +CTCTGA +CTCTGT +CTCTGC +CTCTGG +CTCCAA +CTCCAT +CTCCAC +CTCCAG +CTCCTA +CTCCTT +CTCCTC +CTCCTG +CTCCCA +CTCCCT +CTCCCC +CTCCCG +CTCCGA +CTCCGT +CTCCGC +CTCCGG +CTCGAA +CTCGAT +CTCGAC +CTCGAG +CTCGTA +CTCGTT +CTCGTC +CTCGTG +CTCGCA +CTCGCT +CTCGCC +CTCGCG +CTCGGA +CTCGGT +CTCGGC +CTCGGG +CTGAAA +CTGAAT +CTGAAC +CTGAAG +CTGATA +CTGATT +CTGATC +CTGATG +CTGACA +CTGACT +CTGACC +CTGACG +CTGAGA +CTGAGT +CTGAGC +CTGAGG +CTGTAA +CTGTAT +CTGTAC +CTGTAG +CTGTTA +CTGTTT +CTGTTC +CTGTTG +CTGTCA +CTGTCT +CTGTCC +CTGTCG +CTGTGA +CTGTGT +CTGTGC +CTGTGG +CTGCAA +CTGCAT +CTGCAC +CTGCAG +CTGCTA +CTGCTT +CTGCTC +CTGCTG +CTGCCA +CTGCCT +CTGCCC +CTGCCG +CTGCGA +CTGCGT +CTGCGC +CTGCGG +CTGGAA +CTGGAT +CTGGAC +CTGGAG +CTGGTA +CTGGTT +CTGGTC +CTGGTG +CTGGCA +CTGGCT +CTGGCC +CTGGCG +CTGGGA +CTGGGT +CTGGGC +CTGGGG +CCAAAA +CCAAAT +CCAAAC +CCAAAG +CCAATA +CCAATT +CCAATC +CCAATG +CCAACA +CCAACT +CCAACC +CCAACG +CCAAGA +CCAAGT +CCAAGC +CCAAGG +CCATAA +CCATAT +CCATAC +CCATAG +CCATTA +CCATTT +CCATTC +CCATTG +CCATCA +CCATCT +CCATCC +CCATCG +CCATGA +CCATGT +CCATGC +CCATGG +CCACAA +CCACAT +CCACAC +CCACAG +CCACTA +CCACTT +CCACTC +CCACTG +CCACCA +CCACCT +CCACCC +CCACCG +CCACGA +CCACGT +CCACGC +CCACGG +CCAGAA +CCAGAT +CCAGAC +CCAGAG +CCAGTA +CCAGTT +CCAGTC +CCAGTG +CCAGCA +CCAGCT +CCAGCC +CCAGCG +CCAGGA +CCAGGT +CCAGGC +CCAGGG +CCTAAA +CCTAAT +CCTAAC +CCTAAG +CCTATA +CCTATT +CCTATC +CCTATG +CCTACA +CCTACT +CCTACC +CCTACG +CCTAGA +CCTAGT +CCTAGC +CCTAGG +CCTTAA +CCTTAT +CCTTAC +CCTTAG +CCTTTA +CCTTTT +CCTTTC +CCTTTG +CCTTCA +CCTTCT +CCTTCC +CCTTCG +CCTTGA +CCTTGT +CCTTGC +CCTTGG +CCTCAA +CCTCAT +CCTCAC +CCTCAG +CCTCTA +CCTCTT +CCTCTC +CCTCTG +CCTCCA +CCTCCT +CCTCCC +CCTCCG +CCTCGA +CCTCGT +CCTCGC +CCTCGG +CCTGAA +CCTGAT +CCTGAC +CCTGAG +CCTGTA +CCTGTT +CCTGTC +CCTGTG +CCTGCA +CCTGCT +CCTGCC +CCTGCG +CCTGGA +CCTGGT +CCTGGC +CCTGGG +CCCAAA +CCCAAT +CCCAAC +CCCAAG +CCCATA +CCCATT +CCCATC +CCCATG +CCCACA +CCCACT +CCCACC +CCCACG +CCCAGA +CCCAGT +CCCAGC +CCCAGG +CCCTAA +CCCTAT +CCCTAC +CCCTAG +CCCTTA +CCCTTT +CCCTTC +CCCTTG +CCCTCA +CCCTCT +CCCTCC +CCCTCG +CCCTGA +CCCTGT +CCCTGC +CCCTGG +CCCCAA +CCCCAT +CCCCAC +CCCCAG +CCCCTA +CCCCTT +CCCCTC +CCCCTG +CCCCCA +CCCCCT +CCCCCC +CCCCCG +CCCCGA +CCCCGT +CCCCGC +CCCCGG +CCCGAA +CCCGAT +CCCGAC +CCCGAG +CCCGTA +CCCGTT +CCCGTC +CCCGTG +CCCGCA +CCCGCT +CCCGCC +CCCGCG +CCCGGA +CCCGGT +CCCGGC +CCCGGG +CCGAAA +CCGAAT +CCGAAC +CCGAAG +CCGATA +CCGATT +CCGATC +CCGATG +CCGACA +CCGACT +CCGACC +CCGACG +CCGAGA +CCGAGT +CCGAGC +CCGAGG +CCGTAA +CCGTAT +CCGTAC +CCGTAG +CCGTTA +CCGTTT +CCGTTC +CCGTTG +CCGTCA +CCGTCT +CCGTCC +CCGTCG +CCGTGA +CCGTGT +CCGTGC +CCGTGG +CCGCAA +CCGCAT +CCGCAC +CCGCAG +CCGCTA +CCGCTT +CCGCTC +CCGCTG +CCGCCA +CCGCCT +CCGCCC +CCGCCG +CCGCGA +CCGCGT +CCGCGC +CCGCGG +CCGGAA +CCGGAT +CCGGAC +CCGGAG +CCGGTA +CCGGTT +CCGGTC +CCGGTG +CCGGCA +CCGGCT +CCGGCC +CCGGCG +CCGGGA +CCGGGT +CCGGGC +CCGGGG +CGAAAA +CGAAAT +CGAAAC +CGAAAG +CGAATA +CGAATT +CGAATC +CGAATG +CGAACA +CGAACT +CGAACC +CGAACG +CGAAGA +CGAAGT +CGAAGC +CGAAGG +CGATAA +CGATAT +CGATAC +CGATAG +CGATTA +CGATTT +CGATTC +CGATTG +CGATCA +CGATCT +CGATCC +CGATCG +CGATGA +CGATGT +CGATGC +CGATGG +CGACAA +CGACAT +CGACAC +CGACAG +CGACTA +CGACTT +CGACTC +CGACTG +CGACCA +CGACCT +CGACCC +CGACCG +CGACGA +CGACGT +CGACGC +CGACGG +CGAGAA +CGAGAT +CGAGAC +CGAGAG +CGAGTA +CGAGTT +CGAGTC +CGAGTG +CGAGCA +CGAGCT +CGAGCC +CGAGCG +CGAGGA +CGAGGT +CGAGGC +CGAGGG +CGTAAA +CGTAAT +CGTAAC +CGTAAG +CGTATA +CGTATT +CGTATC +CGTATG +CGTACA +CGTACT +CGTACC +CGTACG +CGTAGA +CGTAGT +CGTAGC +CGTAGG +CGTTAA +CGTTAT +CGTTAC +CGTTAG +CGTTTA +CGTTTT +CGTTTC +CGTTTG +CGTTCA +CGTTCT +CGTTCC +CGTTCG +CGTTGA +CGTTGT +CGTTGC +CGTTGG +CGTCAA +CGTCAT +CGTCAC +CGTCAG +CGTCTA +CGTCTT +CGTCTC +CGTCTG +CGTCCA +CGTCCT +CGTCCC +CGTCCG +CGTCGA +CGTCGT +CGTCGC +CGTCGG +CGTGAA +CGTGAT +CGTGAC +CGTGAG +CGTGTA +CGTGTT +CGTGTC +CGTGTG +CGTGCA +CGTGCT +CGTGCC +CGTGCG +CGTGGA +CGTGGT +CGTGGC +CGTGGG +CGCAAA +CGCAAT +CGCAAC +CGCAAG +CGCATA +CGCATT +CGCATC +CGCATG +CGCACA +CGCACT +CGCACC +CGCACG +CGCAGA +CGCAGT +CGCAGC +CGCAGG +CGCTAA +CGCTAT +CGCTAC +CGCTAG +CGCTTA +CGCTTT +CGCTTC +CGCTTG +CGCTCA +CGCTCT +CGCTCC +CGCTCG +CGCTGA +CGCTGT +CGCTGC +CGCTGG +CGCCAA +CGCCAT +CGCCAC +CGCCAG +CGCCTA +CGCCTT +CGCCTC +CGCCTG +CGCCCA +CGCCCT +CGCCCC +CGCCCG +CGCCGA +CGCCGT +CGCCGC +CGCCGG +CGCGAA +CGCGAT +CGCGAC +CGCGAG +CGCGTA +CGCGTT +CGCGTC +CGCGTG +CGCGCA +CGCGCT +CGCGCC +CGCGCG +CGCGGA +CGCGGT +CGCGGC +CGCGGG +CGGAAA +CGGAAT +CGGAAC +CGGAAG +CGGATA +CGGATT +CGGATC +CGGATG +CGGACA +CGGACT +CGGACC +CGGACG +CGGAGA +CGGAGT +CGGAGC +CGGAGG +CGGTAA +CGGTAT +CGGTAC +CGGTAG +CGGTTA +CGGTTT +CGGTTC +CGGTTG +CGGTCA +CGGTCT +CGGTCC +CGGTCG +CGGTGA +CGGTGT +CGGTGC +CGGTGG +CGGCAA +CGGCAT +CGGCAC +CGGCAG +CGGCTA +CGGCTT +CGGCTC +CGGCTG +CGGCCA +CGGCCT +CGGCCC +CGGCCG +CGGCGA +CGGCGT +CGGCGC +CGGCGG +CGGGAA +CGGGAT +CGGGAC +CGGGAG +CGGGTA +CGGGTT +CGGGTC +CGGGTG +CGGGCA +CGGGCT +CGGGCC +CGGGCG +CGGGGA +CGGGGT +CGGGGC +CGGGGG +GAAAAA +GAAAAT +GAAAAC +GAAAAG +GAAATA +GAAATT +GAAATC +GAAATG +GAAACA +GAAACT +GAAACC +GAAACG +GAAAGA +GAAAGT +GAAAGC +GAAAGG +GAATAA +GAATAT +GAATAC +GAATAG +GAATTA +GAATTT +GAATTC +GAATTG +GAATCA +GAATCT +GAATCC +GAATCG +GAATGA +GAATGT +GAATGC +GAATGG +GAACAA +GAACAT +GAACAC +GAACAG +GAACTA +GAACTT +GAACTC +GAACTG +GAACCA +GAACCT +GAACCC +GAACCG +GAACGA +GAACGT +GAACGC +GAACGG +GAAGAA +GAAGAT +GAAGAC +GAAGAG +GAAGTA +GAAGTT +GAAGTC +GAAGTG +GAAGCA +GAAGCT +GAAGCC +GAAGCG +GAAGGA +GAAGGT +GAAGGC +GAAGGG +GATAAA +GATAAT +GATAAC +GATAAG +GATATA +GATATT +GATATC +GATATG +GATACA +GATACT +GATACC +GATACG +GATAGA +GATAGT +GATAGC +GATAGG +GATTAA +GATTAT +GATTAC +GATTAG +GATTTA +GATTTT +GATTTC +GATTTG +GATTCA +GATTCT +GATTCC +GATTCG +GATTGA +GATTGT +GATTGC +GATTGG +GATCAA +GATCAT +GATCAC +GATCAG +GATCTA +GATCTT +GATCTC +GATCTG +GATCCA +GATCCT +GATCCC +GATCCG +GATCGA +GATCGT +GATCGC +GATCGG +GATGAA +GATGAT +GATGAC +GATGAG +GATGTA +GATGTT +GATGTC +GATGTG +GATGCA +GATGCT +GATGCC +GATGCG +GATGGA +GATGGT +GATGGC +GATGGG +GACAAA +GACAAT +GACAAC +GACAAG +GACATA +GACATT +GACATC +GACATG +GACACA +GACACT +GACACC +GACACG +GACAGA +GACAGT +GACAGC +GACAGG +GACTAA +GACTAT +GACTAC +GACTAG +GACTTA +GACTTT +GACTTC +GACTTG +GACTCA +GACTCT +GACTCC +GACTCG +GACTGA +GACTGT +GACTGC +GACTGG +GACCAA +GACCAT +GACCAC +GACCAG +GACCTA +GACCTT +GACCTC +GACCTG +GACCCA +GACCCT +GACCCC +GACCCG +GACCGA +GACCGT +GACCGC +GACCGG +GACGAA +GACGAT +GACGAC +GACGAG +GACGTA +GACGTT +GACGTC +GACGTG +GACGCA +GACGCT +GACGCC +GACGCG +GACGGA +GACGGT +GACGGC +GACGGG +GAGAAA +GAGAAT +GAGAAC +GAGAAG +GAGATA +GAGATT +GAGATC +GAGATG +GAGACA +GAGACT +GAGACC +GAGACG +GAGAGA +GAGAGT +GAGAGC +GAGAGG +GAGTAA +GAGTAT +GAGTAC +GAGTAG +GAGTTA +GAGTTT +GAGTTC +GAGTTG +GAGTCA +GAGTCT +GAGTCC +GAGTCG +GAGTGA +GAGTGT +GAGTGC +GAGTGG +GAGCAA +GAGCAT +GAGCAC +GAGCAG +GAGCTA +GAGCTT +GAGCTC +GAGCTG +GAGCCA +GAGCCT +GAGCCC +GAGCCG +GAGCGA +GAGCGT +GAGCGC +GAGCGG +GAGGAA +GAGGAT +GAGGAC +GAGGAG +GAGGTA +GAGGTT +GAGGTC +GAGGTG +GAGGCA +GAGGCT +GAGGCC +GAGGCG +GAGGGA +GAGGGT +GAGGGC +GAGGGG +GTAAAA +GTAAAT +GTAAAC +GTAAAG +GTAATA +GTAATT +GTAATC +GTAATG +GTAACA +GTAACT +GTAACC +GTAACG +GTAAGA +GTAAGT +GTAAGC +GTAAGG +GTATAA +GTATAT +GTATAC +GTATAG +GTATTA +GTATTT +GTATTC +GTATTG +GTATCA +GTATCT +GTATCC +GTATCG +GTATGA +GTATGT +GTATGC +GTATGG +GTACAA +GTACAT +GTACAC +GTACAG +GTACTA +GTACTT +GTACTC +GTACTG +GTACCA +GTACCT +GTACCC +GTACCG +GTACGA +GTACGT +GTACGC +GTACGG +GTAGAA +GTAGAT +GTAGAC +GTAGAG +GTAGTA +GTAGTT +GTAGTC +GTAGTG +GTAGCA +GTAGCT +GTAGCC +GTAGCG +GTAGGA +GTAGGT +GTAGGC +GTAGGG +GTTAAA +GTTAAT +GTTAAC +GTTAAG +GTTATA +GTTATT +GTTATC +GTTATG +GTTACA +GTTACT +GTTACC +GTTACG +GTTAGA +GTTAGT +GTTAGC +GTTAGG +GTTTAA +GTTTAT +GTTTAC +GTTTAG +GTTTTA +GTTTTT +GTTTTC +GTTTTG +GTTTCA +GTTTCT +GTTTCC +GTTTCG +GTTTGA +GTTTGT +GTTTGC +GTTTGG +GTTCAA +GTTCAT +GTTCAC +GTTCAG +GTTCTA +GTTCTT +GTTCTC +GTTCTG +GTTCCA +GTTCCT +GTTCCC +GTTCCG +GTTCGA +GTTCGT +GTTCGC +GTTCGG +GTTGAA +GTTGAT +GTTGAC +GTTGAG +GTTGTA +GTTGTT +GTTGTC +GTTGTG +GTTGCA +GTTGCT +GTTGCC +GTTGCG +GTTGGA +GTTGGT +GTTGGC +GTTGGG +GTCAAA +GTCAAT +GTCAAC +GTCAAG +GTCATA +GTCATT +GTCATC +GTCATG +GTCACA +GTCACT +GTCACC +GTCACG +GTCAGA +GTCAGT +GTCAGC +GTCAGG +GTCTAA +GTCTAT +GTCTAC +GTCTAG +GTCTTA +GTCTTT +GTCTTC +GTCTTG +GTCTCA +GTCTCT +GTCTCC +GTCTCG +GTCTGA +GTCTGT +GTCTGC +GTCTGG +GTCCAA +GTCCAT +GTCCAC +GTCCAG +GTCCTA +GTCCTT +GTCCTC +GTCCTG +GTCCCA +GTCCCT +GTCCCC +GTCCCG +GTCCGA +GTCCGT +GTCCGC +GTCCGG +GTCGAA +GTCGAT +GTCGAC +GTCGAG +GTCGTA +GTCGTT +GTCGTC +GTCGTG +GTCGCA +GTCGCT +GTCGCC +GTCGCG +GTCGGA +GTCGGT +GTCGGC +GTCGGG +GTGAAA +GTGAAT +GTGAAC +GTGAAG +GTGATA +GTGATT +GTGATC +GTGATG +GTGACA +GTGACT +GTGACC +GTGACG +GTGAGA +GTGAGT +GTGAGC +GTGAGG +GTGTAA +GTGTAT +GTGTAC +GTGTAG +GTGTTA +GTGTTT +GTGTTC +GTGTTG +GTGTCA +GTGTCT +GTGTCC +GTGTCG +GTGTGA +GTGTGT +GTGTGC +GTGTGG +GTGCAA +GTGCAT +GTGCAC +GTGCAG +GTGCTA +GTGCTT +GTGCTC +GTGCTG +GTGCCA +GTGCCT +GTGCCC +GTGCCG +GTGCGA +GTGCGT +GTGCGC +GTGCGG +GTGGAA +GTGGAT +GTGGAC +GTGGAG +GTGGTA +GTGGTT +GTGGTC +GTGGTG +GTGGCA +GTGGCT +GTGGCC +GTGGCG +GTGGGA +GTGGGT +GTGGGC +GTGGGG +GCAAAA +GCAAAT +GCAAAC +GCAAAG +GCAATA +GCAATT +GCAATC +GCAATG +GCAACA +GCAACT +GCAACC +GCAACG +GCAAGA +GCAAGT +GCAAGC +GCAAGG +GCATAA +GCATAT +GCATAC +GCATAG +GCATTA +GCATTT +GCATTC +GCATTG +GCATCA +GCATCT +GCATCC +GCATCG +GCATGA +GCATGT +GCATGC +GCATGG +GCACAA +GCACAT +GCACAC +GCACAG +GCACTA +GCACTT +GCACTC +GCACTG +GCACCA +GCACCT +GCACCC +GCACCG +GCACGA +GCACGT +GCACGC +GCACGG +GCAGAA +GCAGAT +GCAGAC +GCAGAG +GCAGTA +GCAGTT +GCAGTC +GCAGTG +GCAGCA +GCAGCT +GCAGCC +GCAGCG +GCAGGA +GCAGGT +GCAGGC +GCAGGG +GCTAAA +GCTAAT +GCTAAC +GCTAAG +GCTATA +GCTATT +GCTATC +GCTATG +GCTACA +GCTACT +GCTACC +GCTACG +GCTAGA +GCTAGT +GCTAGC +GCTAGG +GCTTAA +GCTTAT +GCTTAC +GCTTAG +GCTTTA +GCTTTT +GCTTTC +GCTTTG +GCTTCA +GCTTCT +GCTTCC +GCTTCG +GCTTGA +GCTTGT +GCTTGC +GCTTGG +GCTCAA +GCTCAT +GCTCAC +GCTCAG +GCTCTA +GCTCTT +GCTCTC +GCTCTG +GCTCCA +GCTCCT +GCTCCC +GCTCCG +GCTCGA +GCTCGT +GCTCGC +GCTCGG +GCTGAA +GCTGAT +GCTGAC +GCTGAG +GCTGTA +GCTGTT +GCTGTC +GCTGTG +GCTGCA +GCTGCT +GCTGCC +GCTGCG +GCTGGA +GCTGGT +GCTGGC +GCTGGG +GCCAAA +GCCAAT +GCCAAC +GCCAAG +GCCATA +GCCATT +GCCATC +GCCATG +GCCACA +GCCACT +GCCACC +GCCACG +GCCAGA +GCCAGT +GCCAGC +GCCAGG +GCCTAA +GCCTAT +GCCTAC +GCCTAG +GCCTTA +GCCTTT +GCCTTC +GCCTTG +GCCTCA +GCCTCT +GCCTCC +GCCTCG +GCCTGA +GCCTGT +GCCTGC +GCCTGG +GCCCAA +GCCCAT +GCCCAC +GCCCAG +GCCCTA +GCCCTT +GCCCTC +GCCCTG +GCCCCA +GCCCCT +GCCCCC +GCCCCG +GCCCGA +GCCCGT +GCCCGC +GCCCGG +GCCGAA +GCCGAT +GCCGAC +GCCGAG +GCCGTA +GCCGTT +GCCGTC +GCCGTG +GCCGCA +GCCGCT +GCCGCC +GCCGCG +GCCGGA +GCCGGT +GCCGGC +GCCGGG +GCGAAA +GCGAAT +GCGAAC +GCGAAG +GCGATA +GCGATT +GCGATC +GCGATG +GCGACA +GCGACT +GCGACC +GCGACG +GCGAGA +GCGAGT +GCGAGC +GCGAGG +GCGTAA +GCGTAT +GCGTAC +GCGTAG +GCGTTA +GCGTTT +GCGTTC +GCGTTG +GCGTCA +GCGTCT +GCGTCC +GCGTCG +GCGTGA +GCGTGT +GCGTGC +GCGTGG +GCGCAA +GCGCAT +GCGCAC +GCGCAG +GCGCTA +GCGCTT +GCGCTC +GCGCTG +GCGCCA +GCGCCT +GCGCCC +GCGCCG +GCGCGA +GCGCGT +GCGCGC +GCGCGG +GCGGAA +GCGGAT +GCGGAC +GCGGAG +GCGGTA +GCGGTT +GCGGTC +GCGGTG +GCGGCA +GCGGCT +GCGGCC +GCGGCG +GCGGGA +GCGGGT +GCGGGC +GCGGGG +GGAAAA +GGAAAT +GGAAAC +GGAAAG +GGAATA +GGAATT +GGAATC +GGAATG +GGAACA +GGAACT +GGAACC +GGAACG +GGAAGA +GGAAGT +GGAAGC +GGAAGG +GGATAA +GGATAT +GGATAC +GGATAG +GGATTA +GGATTT +GGATTC +GGATTG +GGATCA +GGATCT +GGATCC +GGATCG +GGATGA +GGATGT +GGATGC +GGATGG +GGACAA +GGACAT +GGACAC +GGACAG +GGACTA +GGACTT +GGACTC +GGACTG +GGACCA +GGACCT +GGACCC +GGACCG +GGACGA +GGACGT +GGACGC +GGACGG +GGAGAA +GGAGAT +GGAGAC +GGAGAG +GGAGTA +GGAGTT +GGAGTC +GGAGTG +GGAGCA +GGAGCT +GGAGCC +GGAGCG +GGAGGA +GGAGGT +GGAGGC +GGAGGG +GGTAAA +GGTAAT +GGTAAC +GGTAAG +GGTATA +GGTATT +GGTATC +GGTATG +GGTACA +GGTACT +GGTACC +GGTACG +GGTAGA +GGTAGT +GGTAGC +GGTAGG +GGTTAA +GGTTAT +GGTTAC +GGTTAG +GGTTTA +GGTTTT +GGTTTC +GGTTTG +GGTTCA +GGTTCT +GGTTCC +GGTTCG +GGTTGA +GGTTGT +GGTTGC +GGTTGG +GGTCAA +GGTCAT +GGTCAC +GGTCAG +GGTCTA +GGTCTT +GGTCTC +GGTCTG +GGTCCA +GGTCCT +GGTCCC +GGTCCG +GGTCGA +GGTCGT +GGTCGC +GGTCGG +GGTGAA +GGTGAT +GGTGAC +GGTGAG +GGTGTA +GGTGTT +GGTGTC +GGTGTG +GGTGCA +GGTGCT +GGTGCC +GGTGCG +GGTGGA +GGTGGT +GGTGGC +GGTGGG +GGCAAA +GGCAAT +GGCAAC +GGCAAG +GGCATA +GGCATT +GGCATC +GGCATG +GGCACA +GGCACT +GGCACC +GGCACG +GGCAGA +GGCAGT +GGCAGC +GGCAGG +GGCTAA +GGCTAT +GGCTAC +GGCTAG +GGCTTA +GGCTTT +GGCTTC +GGCTTG +GGCTCA +GGCTCT +GGCTCC +GGCTCG +GGCTGA +GGCTGT +GGCTGC +GGCTGG +GGCCAA +GGCCAT +GGCCAC +GGCCAG +GGCCTA +GGCCTT +GGCCTC +GGCCTG +GGCCCA +GGCCCT +GGCCCC +GGCCCG +GGCCGA +GGCCGT +GGCCGC +GGCCGG +GGCGAA +GGCGAT +GGCGAC +GGCGAG +GGCGTA +GGCGTT +GGCGTC +GGCGTG +GGCGCA +GGCGCT +GGCGCC +GGCGCG +GGCGGA +GGCGGT +GGCGGC +GGCGGG +GGGAAA +GGGAAT +GGGAAC +GGGAAG +GGGATA +GGGATT +GGGATC +GGGATG +GGGACA +GGGACT +GGGACC +GGGACG +GGGAGA +GGGAGT +GGGAGC +GGGAGG +GGGTAA +GGGTAT +GGGTAC +GGGTAG +GGGTTA +GGGTTT +GGGTTC +GGGTTG +GGGTCA +GGGTCT +GGGTCC +GGGTCG +GGGTGA +GGGTGT +GGGTGC +GGGTGG +GGGCAA +GGGCAT +GGGCAC +GGGCAG +GGGCTA +GGGCTT +GGGCTC +GGGCTG +GGGCCA +GGGCCT +GGGCCC +GGGCCG +GGGCGA +GGGCGT +GGGCGC +GGGCGG +GGGGAA +GGGGAT +GGGGAC +GGGGAG +GGGGTA +GGGGTT +GGGGTC +GGGGTG +GGGGCA +GGGGCT +GGGGCC +GGGGCG +GGGGGA +GGGGGT +GGGGGC +GGGGGG diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/README.md b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/README.md new file mode 100644 index 0000000000000000000000000000000000000000..500fd976d92b217455f181bc98c77e612d9a18d9 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/README.md @@ -0,0 +1,202 @@ +--- +base_model: fabihamakhdoomi/TinyDNABERT +library_name: peft +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.12.0 \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/adapter_config.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..04c63e99bbc6b810449dece367c0afb165125b92 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/adapter_config.json @@ -0,0 +1,42 @@ +{ + "alpha_pattern": {}, + "auto_mapping": null, + "base_model_name_or_path": "fabihamakhdoomi/TinyDNABERT", + "bias": "none", + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 16, + "lora_dropout": 0.2, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": [ + "classifier", + "score" + ], + "peft_type": "LORA", + "r": 8, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "bert.encoder.layer.1.attention.self.query", + "bert.encoder.layer.0.output.dense", + "bert.encoder.layer.1.output.dense", + "bert.encoder.layer.0.intermediate.dense", + "bert.encoder.layer.1.intermediate.dense", + "bert.encoder.layer.0.attention.output.dense", + "bert.encoder.layer.1.attention.output.dense", + "bert.encoder.layer.0.attention.self.query", + "bert.encoder.layer.0.attention.self.value", + "bert.encoder.layer.1.attention.self.key", + "bert.encoder.layer.1.attention.self.value", + "bert.encoder.layer.0.attention.self.key" + ], + "task_type": "SEQ_CLS", + "use_dora": false, + "use_rslora": false +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/adapter_model.safetensors b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/adapter_model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..3635fab2757d66847f53e9cb26e813e46719dcda --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/adapter_model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a53f024b07b5b444bba5bef85d38042f1f4d77b27edca47efe7a615eff910783 +size 151952 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/optimizer.pt b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..162a1be6e69f8963ca7177590133b274df642772 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:65ef592591f7b58ab53f6f263f89d60651c70fc0e9a017ae548f171209254721 +size 312682 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/rng_state.pth b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..7d59ecb6c9b7d6ec50b060f3843e47cd8ba9b667 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:84106ec8b7c96df034e207cd31f8dd32759188fe95dd21501975d40030684bf4 +size 14244 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/scheduler.pt b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..f3ebc22eebe7fac9f20742118527e91760861e6d --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7e153e480933b4bdcc7d98226c2fdfde72db931809685729c8748251ae3ecae9 +size 1064 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/special_tokens_map.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..a8b3208c2884c4efb86e49300fdd3dc877220cdf --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/special_tokens_map.json @@ -0,0 +1,7 @@ +{ + "cls_token": "[CLS]", + "mask_token": "[MASK]", + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/tokenizer.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..ebd782de6963b9032ce9374cfe30f5275be1a30b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/tokenizer.json @@ -0,0 +1,4265 @@ +{ + "version": "1.0", + "truncation": { + "direction": "Right", + "max_length": 512, + "strategy": "LongestFirst", + "stride": 0 + }, + "padding": { + "strategy": { + "Fixed": 512 + }, + "direction": "Right", + "pad_to_multiple_of": null, + "pad_id": 0, + "pad_type_id": 0, + "pad_token": "[PAD]" + }, + "added_tokens": [ + { + "id": 0, + "content": "[PAD]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 1, + "content": "[UNK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 2, + "content": "[CLS]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 3, + "content": "[SEP]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 4, + "content": "[MASK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + } + ], + "normalizer": { + "type": "BertNormalizer", + "clean_text": true, + "handle_chinese_chars": true, + "strip_accents": null, + "lowercase": false + }, + "pre_tokenizer": { + "type": "BertPreTokenizer" + }, + "post_processor": { + "type": "TemplateProcessing", + "single": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + } + ], + "pair": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "B", + "type_id": 1 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 1 + } + } + ], + "special_tokens": { + "[CLS]": { + "id": "[CLS]", + "ids": [ + 2 + ], + "tokens": [ + "[CLS]" + ] + }, + "[SEP]": { + "id": "[SEP]", + "ids": [ + 3 + ], + "tokens": [ + "[SEP]" + ] + } + } + }, + "decoder": { + "type": "WordPiece", + "prefix": "##", + "cleanup": true + }, + "model": { + "type": "WordPiece", + "unk_token": "[UNK]", + "continuing_subword_prefix": "##", + "max_input_chars_per_word": 100, + "vocab": { + "[PAD]": 0, + "[UNK]": 1, + "[CLS]": 2, + "[SEP]": 3, + "[MASK]": 4, + "AAAAAA": 5, + "AAAAAT": 6, + "AAAAAC": 7, + "AAAAAG": 8, + "AAAATA": 9, + "AAAATT": 10, + "AAAATC": 11, + "AAAATG": 12, + "AAAACA": 13, + "AAAACT": 14, + "AAAACC": 15, + "AAAACG": 16, + "AAAAGA": 17, + "AAAAGT": 18, + "AAAAGC": 19, + "AAAAGG": 20, + "AAATAA": 21, + "AAATAT": 22, + "AAATAC": 23, + "AAATAG": 24, + "AAATTA": 25, + "AAATTT": 26, + "AAATTC": 27, + "AAATTG": 28, + "AAATCA": 29, + "AAATCT": 30, + "AAATCC": 31, + "AAATCG": 32, + "AAATGA": 33, + "AAATGT": 34, + "AAATGC": 35, + "AAATGG": 36, + "AAACAA": 37, + "AAACAT": 38, + "AAACAC": 39, + "AAACAG": 40, + "AAACTA": 41, + "AAACTT": 42, + "AAACTC": 43, + "AAACTG": 44, + "AAACCA": 45, + "AAACCT": 46, + "AAACCC": 47, + "AAACCG": 48, + "AAACGA": 49, + "AAACGT": 50, + "AAACGC": 51, + "AAACGG": 52, + "AAAGAA": 53, + "AAAGAT": 54, + "AAAGAC": 55, + "AAAGAG": 56, + "AAAGTA": 57, + "AAAGTT": 58, + "AAAGTC": 59, + "AAAGTG": 60, + "AAAGCA": 61, + "AAAGCT": 62, + "AAAGCC": 63, + "AAAGCG": 64, + "AAAGGA": 65, + "AAAGGT": 66, + "AAAGGC": 67, + "AAAGGG": 68, + "AATAAA": 69, + "AATAAT": 70, + "AATAAC": 71, + "AATAAG": 72, + "AATATA": 73, + "AATATT": 74, + "AATATC": 75, + "AATATG": 76, + "AATACA": 77, + "AATACT": 78, + "AATACC": 79, + "AATACG": 80, + "AATAGA": 81, + "AATAGT": 82, + "AATAGC": 83, + "AATAGG": 84, + "AATTAA": 85, + "AATTAT": 86, + "AATTAC": 87, + "AATTAG": 88, + "AATTTA": 89, + "AATTTT": 90, + "AATTTC": 91, + "AATTTG": 92, + "AATTCA": 93, + "AATTCT": 94, + "AATTCC": 95, + "AATTCG": 96, + "AATTGA": 97, + "AATTGT": 98, + "AATTGC": 99, + "AATTGG": 100, + "AATCAA": 101, + "AATCAT": 102, + "AATCAC": 103, + "AATCAG": 104, + "AATCTA": 105, + "AATCTT": 106, + "AATCTC": 107, + "AATCTG": 108, + "AATCCA": 109, + "AATCCT": 110, + "AATCCC": 111, + "AATCCG": 112, + "AATCGA": 113, + "AATCGT": 114, + "AATCGC": 115, + "AATCGG": 116, + "AATGAA": 117, + "AATGAT": 118, + "AATGAC": 119, + "AATGAG": 120, + "AATGTA": 121, + "AATGTT": 122, + "AATGTC": 123, + "AATGTG": 124, + "AATGCA": 125, + "AATGCT": 126, + "AATGCC": 127, + "AATGCG": 128, + "AATGGA": 129, + "AATGGT": 130, + "AATGGC": 131, + "AATGGG": 132, + "AACAAA": 133, + "AACAAT": 134, + "AACAAC": 135, + "AACAAG": 136, + "AACATA": 137, + "AACATT": 138, + "AACATC": 139, + "AACATG": 140, + "AACACA": 141, + "AACACT": 142, + "AACACC": 143, + "AACACG": 144, + "AACAGA": 145, + "AACAGT": 146, + "AACAGC": 147, + "AACAGG": 148, + "AACTAA": 149, + "AACTAT": 150, + "AACTAC": 151, + "AACTAG": 152, + "AACTTA": 153, + "AACTTT": 154, + "AACTTC": 155, + "AACTTG": 156, + "AACTCA": 157, + "AACTCT": 158, + "AACTCC": 159, + "AACTCG": 160, + "AACTGA": 161, + "AACTGT": 162, + "AACTGC": 163, + "AACTGG": 164, + "AACCAA": 165, + "AACCAT": 166, + "AACCAC": 167, + "AACCAG": 168, + "AACCTA": 169, + "AACCTT": 170, + "AACCTC": 171, + "AACCTG": 172, + "AACCCA": 173, + "AACCCT": 174, + "AACCCC": 175, + "AACCCG": 176, + "AACCGA": 177, + "AACCGT": 178, + "AACCGC": 179, + "AACCGG": 180, + "AACGAA": 181, + "AACGAT": 182, + "AACGAC": 183, + "AACGAG": 184, + "AACGTA": 185, + "AACGTT": 186, + "AACGTC": 187, + "AACGTG": 188, + "AACGCA": 189, + "AACGCT": 190, + "AACGCC": 191, + "AACGCG": 192, + "AACGGA": 193, + "AACGGT": 194, + "AACGGC": 195, + "AACGGG": 196, + "AAGAAA": 197, + "AAGAAT": 198, + "AAGAAC": 199, + "AAGAAG": 200, + "AAGATA": 201, + "AAGATT": 202, + "AAGATC": 203, + "AAGATG": 204, + "AAGACA": 205, + "AAGACT": 206, + "AAGACC": 207, + "AAGACG": 208, + "AAGAGA": 209, + "AAGAGT": 210, + "AAGAGC": 211, + "AAGAGG": 212, + "AAGTAA": 213, + "AAGTAT": 214, + "AAGTAC": 215, + "AAGTAG": 216, + "AAGTTA": 217, + "AAGTTT": 218, + "AAGTTC": 219, + "AAGTTG": 220, + "AAGTCA": 221, + "AAGTCT": 222, + "AAGTCC": 223, + "AAGTCG": 224, + "AAGTGA": 225, + "AAGTGT": 226, + "AAGTGC": 227, + "AAGTGG": 228, + "AAGCAA": 229, + "AAGCAT": 230, + "AAGCAC": 231, + "AAGCAG": 232, + "AAGCTA": 233, + "AAGCTT": 234, + "AAGCTC": 235, + "AAGCTG": 236, + "AAGCCA": 237, + "AAGCCT": 238, + "AAGCCC": 239, + "AAGCCG": 240, + "AAGCGA": 241, + "AAGCGT": 242, + "AAGCGC": 243, + "AAGCGG": 244, + "AAGGAA": 245, + "AAGGAT": 246, + "AAGGAC": 247, + "AAGGAG": 248, + "AAGGTA": 249, + "AAGGTT": 250, + "AAGGTC": 251, + "AAGGTG": 252, + "AAGGCA": 253, + "AAGGCT": 254, + "AAGGCC": 255, + "AAGGCG": 256, + "AAGGGA": 257, + "AAGGGT": 258, + "AAGGGC": 259, + "AAGGGG": 260, + "ATAAAA": 261, + "ATAAAT": 262, + "ATAAAC": 263, + "ATAAAG": 264, + "ATAATA": 265, + "ATAATT": 266, + "ATAATC": 267, + "ATAATG": 268, + "ATAACA": 269, + "ATAACT": 270, + "ATAACC": 271, + "ATAACG": 272, + "ATAAGA": 273, + "ATAAGT": 274, + "ATAAGC": 275, + "ATAAGG": 276, + "ATATAA": 277, + "ATATAT": 278, + "ATATAC": 279, + "ATATAG": 280, + "ATATTA": 281, + "ATATTT": 282, + "ATATTC": 283, + "ATATTG": 284, + "ATATCA": 285, + "ATATCT": 286, + "ATATCC": 287, + "ATATCG": 288, + "ATATGA": 289, + "ATATGT": 290, + "ATATGC": 291, + "ATATGG": 292, + "ATACAA": 293, + "ATACAT": 294, + "ATACAC": 295, + "ATACAG": 296, + "ATACTA": 297, + "ATACTT": 298, + "ATACTC": 299, + "ATACTG": 300, + "ATACCA": 301, + "ATACCT": 302, + "ATACCC": 303, + "ATACCG": 304, + "ATACGA": 305, + "ATACGT": 306, + "ATACGC": 307, + "ATACGG": 308, + "ATAGAA": 309, + "ATAGAT": 310, + "ATAGAC": 311, + "ATAGAG": 312, + "ATAGTA": 313, + "ATAGTT": 314, + "ATAGTC": 315, + "ATAGTG": 316, + "ATAGCA": 317, + "ATAGCT": 318, + "ATAGCC": 319, + "ATAGCG": 320, + "ATAGGA": 321, + "ATAGGT": 322, + "ATAGGC": 323, + "ATAGGG": 324, + "ATTAAA": 325, + "ATTAAT": 326, + "ATTAAC": 327, + "ATTAAG": 328, + "ATTATA": 329, + "ATTATT": 330, + "ATTATC": 331, + "ATTATG": 332, + "ATTACA": 333, + "ATTACT": 334, + "ATTACC": 335, + "ATTACG": 336, + "ATTAGA": 337, + "ATTAGT": 338, + "ATTAGC": 339, + "ATTAGG": 340, + "ATTTAA": 341, + "ATTTAT": 342, + "ATTTAC": 343, + "ATTTAG": 344, + "ATTTTA": 345, + "ATTTTT": 346, + "ATTTTC": 347, + "ATTTTG": 348, + "ATTTCA": 349, + "ATTTCT": 350, + "ATTTCC": 351, + "ATTTCG": 352, + "ATTTGA": 353, + "ATTTGT": 354, + "ATTTGC": 355, + "ATTTGG": 356, + "ATTCAA": 357, + "ATTCAT": 358, + "ATTCAC": 359, + "ATTCAG": 360, + "ATTCTA": 361, + "ATTCTT": 362, + "ATTCTC": 363, + "ATTCTG": 364, + "ATTCCA": 365, + "ATTCCT": 366, + "ATTCCC": 367, + "ATTCCG": 368, + "ATTCGA": 369, + "ATTCGT": 370, + "ATTCGC": 371, + "ATTCGG": 372, + "ATTGAA": 373, + "ATTGAT": 374, + "ATTGAC": 375, + "ATTGAG": 376, + "ATTGTA": 377, + "ATTGTT": 378, + "ATTGTC": 379, + "ATTGTG": 380, + "ATTGCA": 381, + "ATTGCT": 382, + "ATTGCC": 383, + "ATTGCG": 384, + "ATTGGA": 385, + "ATTGGT": 386, + "ATTGGC": 387, + "ATTGGG": 388, + "ATCAAA": 389, + "ATCAAT": 390, + "ATCAAC": 391, + "ATCAAG": 392, + "ATCATA": 393, + "ATCATT": 394, + "ATCATC": 395, + "ATCATG": 396, + "ATCACA": 397, + "ATCACT": 398, + "ATCACC": 399, + "ATCACG": 400, + "ATCAGA": 401, + "ATCAGT": 402, + "ATCAGC": 403, + "ATCAGG": 404, + "ATCTAA": 405, + "ATCTAT": 406, + "ATCTAC": 407, + "ATCTAG": 408, + "ATCTTA": 409, + "ATCTTT": 410, + "ATCTTC": 411, + "ATCTTG": 412, + "ATCTCA": 413, + "ATCTCT": 414, + "ATCTCC": 415, + "ATCTCG": 416, + "ATCTGA": 417, + "ATCTGT": 418, + "ATCTGC": 419, + "ATCTGG": 420, + "ATCCAA": 421, + "ATCCAT": 422, + "ATCCAC": 423, + "ATCCAG": 424, + "ATCCTA": 425, + "ATCCTT": 426, + "ATCCTC": 427, + "ATCCTG": 428, + "ATCCCA": 429, + "ATCCCT": 430, + "ATCCCC": 431, + "ATCCCG": 432, + "ATCCGA": 433, + "ATCCGT": 434, + "ATCCGC": 435, + "ATCCGG": 436, + "ATCGAA": 437, + "ATCGAT": 438, + "ATCGAC": 439, + "ATCGAG": 440, + "ATCGTA": 441, + "ATCGTT": 442, + "ATCGTC": 443, + "ATCGTG": 444, + "ATCGCA": 445, + "ATCGCT": 446, + "ATCGCC": 447, + "ATCGCG": 448, + "ATCGGA": 449, + "ATCGGT": 450, + "ATCGGC": 451, + "ATCGGG": 452, + "ATGAAA": 453, + "ATGAAT": 454, + "ATGAAC": 455, + "ATGAAG": 456, + "ATGATA": 457, + "ATGATT": 458, + "ATGATC": 459, + "ATGATG": 460, + "ATGACA": 461, + "ATGACT": 462, + "ATGACC": 463, + "ATGACG": 464, + "ATGAGA": 465, + "ATGAGT": 466, + "ATGAGC": 467, + "ATGAGG": 468, + "ATGTAA": 469, + "ATGTAT": 470, + "ATGTAC": 471, + "ATGTAG": 472, + "ATGTTA": 473, + "ATGTTT": 474, + "ATGTTC": 475, + "ATGTTG": 476, + "ATGTCA": 477, + "ATGTCT": 478, + "ATGTCC": 479, + "ATGTCG": 480, + "ATGTGA": 481, + "ATGTGT": 482, + "ATGTGC": 483, + "ATGTGG": 484, + "ATGCAA": 485, + "ATGCAT": 486, + "ATGCAC": 487, + "ATGCAG": 488, + "ATGCTA": 489, + "ATGCTT": 490, + "ATGCTC": 491, + "ATGCTG": 492, + "ATGCCA": 493, + "ATGCCT": 494, + "ATGCCC": 495, + "ATGCCG": 496, + "ATGCGA": 497, + "ATGCGT": 498, + "ATGCGC": 499, + "ATGCGG": 500, + "ATGGAA": 501, + "ATGGAT": 502, + "ATGGAC": 503, + "ATGGAG": 504, + "ATGGTA": 505, + "ATGGTT": 506, + "ATGGTC": 507, + "ATGGTG": 508, + "ATGGCA": 509, + "ATGGCT": 510, + "ATGGCC": 511, + "ATGGCG": 512, + "ATGGGA": 513, + "ATGGGT": 514, + "ATGGGC": 515, + "ATGGGG": 516, + "ACAAAA": 517, + "ACAAAT": 518, + "ACAAAC": 519, + "ACAAAG": 520, + "ACAATA": 521, + "ACAATT": 522, + "ACAATC": 523, + "ACAATG": 524, + "ACAACA": 525, + "ACAACT": 526, + "ACAACC": 527, + "ACAACG": 528, + "ACAAGA": 529, + "ACAAGT": 530, + "ACAAGC": 531, + "ACAAGG": 532, + "ACATAA": 533, + "ACATAT": 534, + "ACATAC": 535, + "ACATAG": 536, + "ACATTA": 537, + "ACATTT": 538, + "ACATTC": 539, + "ACATTG": 540, + "ACATCA": 541, + "ACATCT": 542, + "ACATCC": 543, + "ACATCG": 544, + "ACATGA": 545, + "ACATGT": 546, + "ACATGC": 547, + "ACATGG": 548, + "ACACAA": 549, + "ACACAT": 550, + "ACACAC": 551, + "ACACAG": 552, + "ACACTA": 553, + "ACACTT": 554, + "ACACTC": 555, + "ACACTG": 556, + "ACACCA": 557, + "ACACCT": 558, + "ACACCC": 559, + "ACACCG": 560, + "ACACGA": 561, + "ACACGT": 562, + "ACACGC": 563, + "ACACGG": 564, + "ACAGAA": 565, + "ACAGAT": 566, + "ACAGAC": 567, + "ACAGAG": 568, + "ACAGTA": 569, + "ACAGTT": 570, + "ACAGTC": 571, + "ACAGTG": 572, + "ACAGCA": 573, + "ACAGCT": 574, + "ACAGCC": 575, + "ACAGCG": 576, + "ACAGGA": 577, + "ACAGGT": 578, + "ACAGGC": 579, + "ACAGGG": 580, + "ACTAAA": 581, + "ACTAAT": 582, + "ACTAAC": 583, + "ACTAAG": 584, + "ACTATA": 585, + "ACTATT": 586, + "ACTATC": 587, + "ACTATG": 588, + "ACTACA": 589, + "ACTACT": 590, + "ACTACC": 591, + "ACTACG": 592, + "ACTAGA": 593, + "ACTAGT": 594, + "ACTAGC": 595, + "ACTAGG": 596, + "ACTTAA": 597, + "ACTTAT": 598, + "ACTTAC": 599, + "ACTTAG": 600, + "ACTTTA": 601, + "ACTTTT": 602, + "ACTTTC": 603, + "ACTTTG": 604, + "ACTTCA": 605, + "ACTTCT": 606, + "ACTTCC": 607, + "ACTTCG": 608, + "ACTTGA": 609, + "ACTTGT": 610, + "ACTTGC": 611, + "ACTTGG": 612, + "ACTCAA": 613, + "ACTCAT": 614, + "ACTCAC": 615, + "ACTCAG": 616, + "ACTCTA": 617, + "ACTCTT": 618, + "ACTCTC": 619, + "ACTCTG": 620, + "ACTCCA": 621, + "ACTCCT": 622, + "ACTCCC": 623, + "ACTCCG": 624, + "ACTCGA": 625, + "ACTCGT": 626, + "ACTCGC": 627, + "ACTCGG": 628, + "ACTGAA": 629, + "ACTGAT": 630, + "ACTGAC": 631, + "ACTGAG": 632, + "ACTGTA": 633, + "ACTGTT": 634, + "ACTGTC": 635, + "ACTGTG": 636, + "ACTGCA": 637, + "ACTGCT": 638, + "ACTGCC": 639, + "ACTGCG": 640, + "ACTGGA": 641, + "ACTGGT": 642, + "ACTGGC": 643, + "ACTGGG": 644, + "ACCAAA": 645, + "ACCAAT": 646, + "ACCAAC": 647, + "ACCAAG": 648, + "ACCATA": 649, + "ACCATT": 650, + "ACCATC": 651, + "ACCATG": 652, + "ACCACA": 653, + "ACCACT": 654, + "ACCACC": 655, + "ACCACG": 656, + "ACCAGA": 657, + "ACCAGT": 658, + "ACCAGC": 659, + "ACCAGG": 660, + "ACCTAA": 661, + "ACCTAT": 662, + "ACCTAC": 663, + "ACCTAG": 664, + "ACCTTA": 665, + "ACCTTT": 666, + "ACCTTC": 667, + "ACCTTG": 668, + "ACCTCA": 669, + "ACCTCT": 670, + "ACCTCC": 671, + "ACCTCG": 672, + "ACCTGA": 673, + "ACCTGT": 674, + "ACCTGC": 675, + "ACCTGG": 676, + "ACCCAA": 677, + "ACCCAT": 678, + "ACCCAC": 679, + "ACCCAG": 680, + "ACCCTA": 681, + "ACCCTT": 682, + "ACCCTC": 683, + "ACCCTG": 684, + "ACCCCA": 685, + "ACCCCT": 686, + "ACCCCC": 687, + "ACCCCG": 688, + "ACCCGA": 689, + "ACCCGT": 690, + "ACCCGC": 691, + "ACCCGG": 692, + "ACCGAA": 693, + "ACCGAT": 694, + "ACCGAC": 695, + "ACCGAG": 696, + "ACCGTA": 697, + "ACCGTT": 698, + "ACCGTC": 699, + "ACCGTG": 700, + "ACCGCA": 701, + "ACCGCT": 702, + "ACCGCC": 703, + "ACCGCG": 704, + "ACCGGA": 705, + "ACCGGT": 706, + "ACCGGC": 707, + "ACCGGG": 708, + "ACGAAA": 709, + "ACGAAT": 710, + "ACGAAC": 711, + "ACGAAG": 712, + "ACGATA": 713, + "ACGATT": 714, + "ACGATC": 715, + "ACGATG": 716, + "ACGACA": 717, + "ACGACT": 718, + "ACGACC": 719, + "ACGACG": 720, + "ACGAGA": 721, + "ACGAGT": 722, + "ACGAGC": 723, + "ACGAGG": 724, + "ACGTAA": 725, + "ACGTAT": 726, + "ACGTAC": 727, + "ACGTAG": 728, + "ACGTTA": 729, + "ACGTTT": 730, + "ACGTTC": 731, + "ACGTTG": 732, + "ACGTCA": 733, + "ACGTCT": 734, + "ACGTCC": 735, + "ACGTCG": 736, + "ACGTGA": 737, + "ACGTGT": 738, + "ACGTGC": 739, + "ACGTGG": 740, + "ACGCAA": 741, + "ACGCAT": 742, + "ACGCAC": 743, + "ACGCAG": 744, + "ACGCTA": 745, + "ACGCTT": 746, + "ACGCTC": 747, + "ACGCTG": 748, + "ACGCCA": 749, + "ACGCCT": 750, + "ACGCCC": 751, + "ACGCCG": 752, + "ACGCGA": 753, + "ACGCGT": 754, + "ACGCGC": 755, + "ACGCGG": 756, + "ACGGAA": 757, + "ACGGAT": 758, + "ACGGAC": 759, + "ACGGAG": 760, + "ACGGTA": 761, + "ACGGTT": 762, + "ACGGTC": 763, + "ACGGTG": 764, + "ACGGCA": 765, + "ACGGCT": 766, + "ACGGCC": 767, + "ACGGCG": 768, + "ACGGGA": 769, + "ACGGGT": 770, + "ACGGGC": 771, + "ACGGGG": 772, + "AGAAAA": 773, + "AGAAAT": 774, + "AGAAAC": 775, + "AGAAAG": 776, + "AGAATA": 777, + "AGAATT": 778, + "AGAATC": 779, + "AGAATG": 780, + "AGAACA": 781, + "AGAACT": 782, + "AGAACC": 783, + "AGAACG": 784, + "AGAAGA": 785, + "AGAAGT": 786, + "AGAAGC": 787, + "AGAAGG": 788, + "AGATAA": 789, + "AGATAT": 790, + "AGATAC": 791, + "AGATAG": 792, + "AGATTA": 793, + "AGATTT": 794, + "AGATTC": 795, + "AGATTG": 796, + "AGATCA": 797, + "AGATCT": 798, + "AGATCC": 799, + "AGATCG": 800, + "AGATGA": 801, + "AGATGT": 802, + "AGATGC": 803, + "AGATGG": 804, + "AGACAA": 805, + "AGACAT": 806, + "AGACAC": 807, + "AGACAG": 808, + "AGACTA": 809, + "AGACTT": 810, + "AGACTC": 811, + "AGACTG": 812, + "AGACCA": 813, + "AGACCT": 814, + "AGACCC": 815, + "AGACCG": 816, + "AGACGA": 817, + "AGACGT": 818, + "AGACGC": 819, + "AGACGG": 820, + "AGAGAA": 821, + "AGAGAT": 822, + "AGAGAC": 823, + "AGAGAG": 824, + "AGAGTA": 825, + "AGAGTT": 826, + "AGAGTC": 827, + "AGAGTG": 828, + "AGAGCA": 829, + "AGAGCT": 830, + "AGAGCC": 831, + "AGAGCG": 832, + "AGAGGA": 833, + "AGAGGT": 834, + "AGAGGC": 835, + "AGAGGG": 836, + "AGTAAA": 837, + "AGTAAT": 838, + "AGTAAC": 839, + "AGTAAG": 840, + "AGTATA": 841, + "AGTATT": 842, + "AGTATC": 843, + "AGTATG": 844, + "AGTACA": 845, + "AGTACT": 846, + "AGTACC": 847, + "AGTACG": 848, + "AGTAGA": 849, + "AGTAGT": 850, + "AGTAGC": 851, + "AGTAGG": 852, + "AGTTAA": 853, + "AGTTAT": 854, + "AGTTAC": 855, + "AGTTAG": 856, + "AGTTTA": 857, + "AGTTTT": 858, + "AGTTTC": 859, + "AGTTTG": 860, + "AGTTCA": 861, + "AGTTCT": 862, + "AGTTCC": 863, + "AGTTCG": 864, + "AGTTGA": 865, + "AGTTGT": 866, + "AGTTGC": 867, + "AGTTGG": 868, + "AGTCAA": 869, + "AGTCAT": 870, + "AGTCAC": 871, + "AGTCAG": 872, + "AGTCTA": 873, + "AGTCTT": 874, + "AGTCTC": 875, + "AGTCTG": 876, + "AGTCCA": 877, + "AGTCCT": 878, + "AGTCCC": 879, + "AGTCCG": 880, + "AGTCGA": 881, + "AGTCGT": 882, + "AGTCGC": 883, + "AGTCGG": 884, + "AGTGAA": 885, + "AGTGAT": 886, + "AGTGAC": 887, + "AGTGAG": 888, + "AGTGTA": 889, + "AGTGTT": 890, + "AGTGTC": 891, + "AGTGTG": 892, + "AGTGCA": 893, + "AGTGCT": 894, + "AGTGCC": 895, + "AGTGCG": 896, + "AGTGGA": 897, + "AGTGGT": 898, + "AGTGGC": 899, + "AGTGGG": 900, + "AGCAAA": 901, + "AGCAAT": 902, + "AGCAAC": 903, + "AGCAAG": 904, + "AGCATA": 905, + "AGCATT": 906, + "AGCATC": 907, + "AGCATG": 908, + "AGCACA": 909, + "AGCACT": 910, + "AGCACC": 911, + "AGCACG": 912, + "AGCAGA": 913, + "AGCAGT": 914, + "AGCAGC": 915, + "AGCAGG": 916, + "AGCTAA": 917, + "AGCTAT": 918, + "AGCTAC": 919, + "AGCTAG": 920, + "AGCTTA": 921, + "AGCTTT": 922, + "AGCTTC": 923, + "AGCTTG": 924, + "AGCTCA": 925, + "AGCTCT": 926, + "AGCTCC": 927, + "AGCTCG": 928, + "AGCTGA": 929, + "AGCTGT": 930, + "AGCTGC": 931, + "AGCTGG": 932, + "AGCCAA": 933, + "AGCCAT": 934, + "AGCCAC": 935, + "AGCCAG": 936, + "AGCCTA": 937, + "AGCCTT": 938, + "AGCCTC": 939, + "AGCCTG": 940, + "AGCCCA": 941, + "AGCCCT": 942, + "AGCCCC": 943, + "AGCCCG": 944, + "AGCCGA": 945, + "AGCCGT": 946, + "AGCCGC": 947, + "AGCCGG": 948, + "AGCGAA": 949, + "AGCGAT": 950, + "AGCGAC": 951, + "AGCGAG": 952, + "AGCGTA": 953, + "AGCGTT": 954, + "AGCGTC": 955, + "AGCGTG": 956, + "AGCGCA": 957, + "AGCGCT": 958, + "AGCGCC": 959, + "AGCGCG": 960, + "AGCGGA": 961, + "AGCGGT": 962, + "AGCGGC": 963, + "AGCGGG": 964, + "AGGAAA": 965, + "AGGAAT": 966, + "AGGAAC": 967, + "AGGAAG": 968, + "AGGATA": 969, + "AGGATT": 970, + "AGGATC": 971, + "AGGATG": 972, + "AGGACA": 973, + "AGGACT": 974, + "AGGACC": 975, + "AGGACG": 976, + "AGGAGA": 977, + "AGGAGT": 978, + "AGGAGC": 979, + "AGGAGG": 980, + "AGGTAA": 981, + "AGGTAT": 982, + "AGGTAC": 983, + "AGGTAG": 984, + "AGGTTA": 985, + "AGGTTT": 986, + "AGGTTC": 987, + "AGGTTG": 988, + "AGGTCA": 989, + "AGGTCT": 990, + "AGGTCC": 991, + "AGGTCG": 992, + "AGGTGA": 993, + "AGGTGT": 994, + "AGGTGC": 995, + "AGGTGG": 996, + "AGGCAA": 997, + "AGGCAT": 998, + "AGGCAC": 999, + "AGGCAG": 1000, + "AGGCTA": 1001, + "AGGCTT": 1002, + "AGGCTC": 1003, + "AGGCTG": 1004, + "AGGCCA": 1005, + "AGGCCT": 1006, + "AGGCCC": 1007, + "AGGCCG": 1008, + "AGGCGA": 1009, + "AGGCGT": 1010, + "AGGCGC": 1011, + "AGGCGG": 1012, + "AGGGAA": 1013, + "AGGGAT": 1014, + "AGGGAC": 1015, + "AGGGAG": 1016, + "AGGGTA": 1017, + "AGGGTT": 1018, + "AGGGTC": 1019, + "AGGGTG": 1020, + "AGGGCA": 1021, + "AGGGCT": 1022, + "AGGGCC": 1023, + "AGGGCG": 1024, + "AGGGGA": 1025, + "AGGGGT": 1026, + "AGGGGC": 1027, + "AGGGGG": 1028, + "TAAAAA": 1029, + "TAAAAT": 1030, + "TAAAAC": 1031, + "TAAAAG": 1032, + "TAAATA": 1033, + "TAAATT": 1034, + "TAAATC": 1035, + "TAAATG": 1036, + "TAAACA": 1037, + "TAAACT": 1038, + "TAAACC": 1039, + "TAAACG": 1040, + "TAAAGA": 1041, + "TAAAGT": 1042, + "TAAAGC": 1043, + "TAAAGG": 1044, + "TAATAA": 1045, + "TAATAT": 1046, + "TAATAC": 1047, + "TAATAG": 1048, + "TAATTA": 1049, + "TAATTT": 1050, + "TAATTC": 1051, + "TAATTG": 1052, + "TAATCA": 1053, + "TAATCT": 1054, + "TAATCC": 1055, + "TAATCG": 1056, + "TAATGA": 1057, + "TAATGT": 1058, + "TAATGC": 1059, + "TAATGG": 1060, + "TAACAA": 1061, + "TAACAT": 1062, + "TAACAC": 1063, + "TAACAG": 1064, + "TAACTA": 1065, + "TAACTT": 1066, + "TAACTC": 1067, + "TAACTG": 1068, + "TAACCA": 1069, + "TAACCT": 1070, + "TAACCC": 1071, + "TAACCG": 1072, + "TAACGA": 1073, + "TAACGT": 1074, + "TAACGC": 1075, + "TAACGG": 1076, + "TAAGAA": 1077, + "TAAGAT": 1078, + "TAAGAC": 1079, + "TAAGAG": 1080, + "TAAGTA": 1081, + "TAAGTT": 1082, + "TAAGTC": 1083, + "TAAGTG": 1084, + "TAAGCA": 1085, + "TAAGCT": 1086, + "TAAGCC": 1087, + "TAAGCG": 1088, + "TAAGGA": 1089, + "TAAGGT": 1090, + "TAAGGC": 1091, + "TAAGGG": 1092, + "TATAAA": 1093, + "TATAAT": 1094, + "TATAAC": 1095, + "TATAAG": 1096, + "TATATA": 1097, + "TATATT": 1098, + "TATATC": 1099, + "TATATG": 1100, + "TATACA": 1101, + "TATACT": 1102, + "TATACC": 1103, + "TATACG": 1104, + "TATAGA": 1105, + "TATAGT": 1106, + "TATAGC": 1107, + "TATAGG": 1108, + "TATTAA": 1109, + "TATTAT": 1110, + "TATTAC": 1111, + "TATTAG": 1112, + "TATTTA": 1113, + "TATTTT": 1114, + "TATTTC": 1115, + "TATTTG": 1116, + "TATTCA": 1117, + "TATTCT": 1118, + "TATTCC": 1119, + "TATTCG": 1120, + "TATTGA": 1121, + "TATTGT": 1122, + "TATTGC": 1123, + "TATTGG": 1124, + "TATCAA": 1125, + "TATCAT": 1126, + "TATCAC": 1127, + "TATCAG": 1128, + "TATCTA": 1129, + "TATCTT": 1130, + "TATCTC": 1131, + "TATCTG": 1132, + "TATCCA": 1133, + "TATCCT": 1134, + "TATCCC": 1135, + "TATCCG": 1136, + "TATCGA": 1137, + "TATCGT": 1138, + "TATCGC": 1139, + "TATCGG": 1140, + "TATGAA": 1141, + "TATGAT": 1142, + "TATGAC": 1143, + "TATGAG": 1144, + "TATGTA": 1145, + "TATGTT": 1146, + "TATGTC": 1147, + "TATGTG": 1148, + "TATGCA": 1149, + "TATGCT": 1150, + "TATGCC": 1151, + "TATGCG": 1152, + "TATGGA": 1153, + "TATGGT": 1154, + "TATGGC": 1155, + "TATGGG": 1156, + "TACAAA": 1157, + "TACAAT": 1158, + "TACAAC": 1159, + "TACAAG": 1160, + "TACATA": 1161, + "TACATT": 1162, + "TACATC": 1163, + "TACATG": 1164, + "TACACA": 1165, + "TACACT": 1166, + "TACACC": 1167, + "TACACG": 1168, + "TACAGA": 1169, + "TACAGT": 1170, + "TACAGC": 1171, + "TACAGG": 1172, + "TACTAA": 1173, + "TACTAT": 1174, + "TACTAC": 1175, + "TACTAG": 1176, + "TACTTA": 1177, + "TACTTT": 1178, + "TACTTC": 1179, + "TACTTG": 1180, + "TACTCA": 1181, + "TACTCT": 1182, + "TACTCC": 1183, + "TACTCG": 1184, + "TACTGA": 1185, + "TACTGT": 1186, + "TACTGC": 1187, + "TACTGG": 1188, + "TACCAA": 1189, + "TACCAT": 1190, + "TACCAC": 1191, + "TACCAG": 1192, + "TACCTA": 1193, + "TACCTT": 1194, + "TACCTC": 1195, + "TACCTG": 1196, + "TACCCA": 1197, + "TACCCT": 1198, + "TACCCC": 1199, + "TACCCG": 1200, + "TACCGA": 1201, + "TACCGT": 1202, + "TACCGC": 1203, + "TACCGG": 1204, + "TACGAA": 1205, + "TACGAT": 1206, + "TACGAC": 1207, + "TACGAG": 1208, + "TACGTA": 1209, + "TACGTT": 1210, + "TACGTC": 1211, + "TACGTG": 1212, + "TACGCA": 1213, + "TACGCT": 1214, + "TACGCC": 1215, + "TACGCG": 1216, + "TACGGA": 1217, + "TACGGT": 1218, + "TACGGC": 1219, + "TACGGG": 1220, + "TAGAAA": 1221, + "TAGAAT": 1222, + "TAGAAC": 1223, + "TAGAAG": 1224, + "TAGATA": 1225, + "TAGATT": 1226, + "TAGATC": 1227, + "TAGATG": 1228, + "TAGACA": 1229, + "TAGACT": 1230, + "TAGACC": 1231, + "TAGACG": 1232, + "TAGAGA": 1233, + "TAGAGT": 1234, + "TAGAGC": 1235, + "TAGAGG": 1236, + "TAGTAA": 1237, + "TAGTAT": 1238, + "TAGTAC": 1239, + "TAGTAG": 1240, + "TAGTTA": 1241, + "TAGTTT": 1242, + "TAGTTC": 1243, + "TAGTTG": 1244, + "TAGTCA": 1245, + "TAGTCT": 1246, + "TAGTCC": 1247, + "TAGTCG": 1248, + "TAGTGA": 1249, + "TAGTGT": 1250, + "TAGTGC": 1251, + "TAGTGG": 1252, + "TAGCAA": 1253, + "TAGCAT": 1254, + "TAGCAC": 1255, + "TAGCAG": 1256, + "TAGCTA": 1257, + "TAGCTT": 1258, + "TAGCTC": 1259, + "TAGCTG": 1260, + "TAGCCA": 1261, + "TAGCCT": 1262, + "TAGCCC": 1263, + "TAGCCG": 1264, + "TAGCGA": 1265, + "TAGCGT": 1266, + "TAGCGC": 1267, + "TAGCGG": 1268, + "TAGGAA": 1269, + "TAGGAT": 1270, + "TAGGAC": 1271, + "TAGGAG": 1272, + "TAGGTA": 1273, + "TAGGTT": 1274, + "TAGGTC": 1275, + "TAGGTG": 1276, + "TAGGCA": 1277, + "TAGGCT": 1278, + "TAGGCC": 1279, + "TAGGCG": 1280, + "TAGGGA": 1281, + "TAGGGT": 1282, + "TAGGGC": 1283, + "TAGGGG": 1284, + "TTAAAA": 1285, + "TTAAAT": 1286, + "TTAAAC": 1287, + "TTAAAG": 1288, + "TTAATA": 1289, + "TTAATT": 1290, + "TTAATC": 1291, + "TTAATG": 1292, + "TTAACA": 1293, + "TTAACT": 1294, + "TTAACC": 1295, + "TTAACG": 1296, + "TTAAGA": 1297, + "TTAAGT": 1298, + "TTAAGC": 1299, + "TTAAGG": 1300, + "TTATAA": 1301, + "TTATAT": 1302, + "TTATAC": 1303, + "TTATAG": 1304, + "TTATTA": 1305, + "TTATTT": 1306, + "TTATTC": 1307, + "TTATTG": 1308, + "TTATCA": 1309, + "TTATCT": 1310, + "TTATCC": 1311, + "TTATCG": 1312, + "TTATGA": 1313, + "TTATGT": 1314, + "TTATGC": 1315, + "TTATGG": 1316, + "TTACAA": 1317, + "TTACAT": 1318, + "TTACAC": 1319, + "TTACAG": 1320, + "TTACTA": 1321, + "TTACTT": 1322, + "TTACTC": 1323, + "TTACTG": 1324, + "TTACCA": 1325, + "TTACCT": 1326, + "TTACCC": 1327, + "TTACCG": 1328, + "TTACGA": 1329, + "TTACGT": 1330, + "TTACGC": 1331, + "TTACGG": 1332, + "TTAGAA": 1333, + "TTAGAT": 1334, + "TTAGAC": 1335, + "TTAGAG": 1336, + "TTAGTA": 1337, + "TTAGTT": 1338, + "TTAGTC": 1339, + "TTAGTG": 1340, + "TTAGCA": 1341, + "TTAGCT": 1342, + "TTAGCC": 1343, + "TTAGCG": 1344, + "TTAGGA": 1345, + "TTAGGT": 1346, + "TTAGGC": 1347, + "TTAGGG": 1348, + "TTTAAA": 1349, + "TTTAAT": 1350, + "TTTAAC": 1351, + "TTTAAG": 1352, + "TTTATA": 1353, + "TTTATT": 1354, + "TTTATC": 1355, + "TTTATG": 1356, + "TTTACA": 1357, + "TTTACT": 1358, + "TTTACC": 1359, + "TTTACG": 1360, + "TTTAGA": 1361, + "TTTAGT": 1362, + "TTTAGC": 1363, + "TTTAGG": 1364, + "TTTTAA": 1365, + "TTTTAT": 1366, + "TTTTAC": 1367, + "TTTTAG": 1368, + "TTTTTA": 1369, + "TTTTTT": 1370, + "TTTTTC": 1371, + "TTTTTG": 1372, + "TTTTCA": 1373, + "TTTTCT": 1374, + "TTTTCC": 1375, + "TTTTCG": 1376, + "TTTTGA": 1377, + "TTTTGT": 1378, + "TTTTGC": 1379, + "TTTTGG": 1380, + "TTTCAA": 1381, + "TTTCAT": 1382, + "TTTCAC": 1383, + "TTTCAG": 1384, + "TTTCTA": 1385, + "TTTCTT": 1386, + "TTTCTC": 1387, + "TTTCTG": 1388, + "TTTCCA": 1389, + "TTTCCT": 1390, + "TTTCCC": 1391, + "TTTCCG": 1392, + "TTTCGA": 1393, + "TTTCGT": 1394, + "TTTCGC": 1395, + "TTTCGG": 1396, + "TTTGAA": 1397, + "TTTGAT": 1398, + "TTTGAC": 1399, + "TTTGAG": 1400, + "TTTGTA": 1401, + "TTTGTT": 1402, + "TTTGTC": 1403, + "TTTGTG": 1404, + "TTTGCA": 1405, + "TTTGCT": 1406, + "TTTGCC": 1407, + "TTTGCG": 1408, + "TTTGGA": 1409, + "TTTGGT": 1410, + "TTTGGC": 1411, + "TTTGGG": 1412, + "TTCAAA": 1413, + "TTCAAT": 1414, + "TTCAAC": 1415, + "TTCAAG": 1416, + "TTCATA": 1417, + "TTCATT": 1418, + "TTCATC": 1419, + "TTCATG": 1420, + "TTCACA": 1421, + "TTCACT": 1422, + "TTCACC": 1423, + "TTCACG": 1424, + "TTCAGA": 1425, + "TTCAGT": 1426, + "TTCAGC": 1427, + "TTCAGG": 1428, + "TTCTAA": 1429, + "TTCTAT": 1430, + "TTCTAC": 1431, + "TTCTAG": 1432, + "TTCTTA": 1433, + "TTCTTT": 1434, + "TTCTTC": 1435, + "TTCTTG": 1436, + "TTCTCA": 1437, + "TTCTCT": 1438, + "TTCTCC": 1439, + "TTCTCG": 1440, + "TTCTGA": 1441, + "TTCTGT": 1442, + "TTCTGC": 1443, + "TTCTGG": 1444, + "TTCCAA": 1445, + "TTCCAT": 1446, + "TTCCAC": 1447, + "TTCCAG": 1448, + "TTCCTA": 1449, + "TTCCTT": 1450, + "TTCCTC": 1451, + "TTCCTG": 1452, + "TTCCCA": 1453, + "TTCCCT": 1454, + "TTCCCC": 1455, + "TTCCCG": 1456, + "TTCCGA": 1457, + "TTCCGT": 1458, + "TTCCGC": 1459, + "TTCCGG": 1460, + "TTCGAA": 1461, + "TTCGAT": 1462, + "TTCGAC": 1463, + "TTCGAG": 1464, + "TTCGTA": 1465, + "TTCGTT": 1466, + "TTCGTC": 1467, + "TTCGTG": 1468, + "TTCGCA": 1469, + "TTCGCT": 1470, + "TTCGCC": 1471, + "TTCGCG": 1472, + "TTCGGA": 1473, + "TTCGGT": 1474, + "TTCGGC": 1475, + "TTCGGG": 1476, + "TTGAAA": 1477, + "TTGAAT": 1478, + "TTGAAC": 1479, + "TTGAAG": 1480, + "TTGATA": 1481, + "TTGATT": 1482, + "TTGATC": 1483, + "TTGATG": 1484, + "TTGACA": 1485, + "TTGACT": 1486, + "TTGACC": 1487, + "TTGACG": 1488, + "TTGAGA": 1489, + "TTGAGT": 1490, + "TTGAGC": 1491, + "TTGAGG": 1492, + "TTGTAA": 1493, + "TTGTAT": 1494, + "TTGTAC": 1495, + "TTGTAG": 1496, + "TTGTTA": 1497, + "TTGTTT": 1498, + "TTGTTC": 1499, + "TTGTTG": 1500, + "TTGTCA": 1501, + "TTGTCT": 1502, + "TTGTCC": 1503, + "TTGTCG": 1504, + "TTGTGA": 1505, + "TTGTGT": 1506, + "TTGTGC": 1507, + "TTGTGG": 1508, + "TTGCAA": 1509, + "TTGCAT": 1510, + "TTGCAC": 1511, + "TTGCAG": 1512, + "TTGCTA": 1513, + "TTGCTT": 1514, + "TTGCTC": 1515, + "TTGCTG": 1516, + "TTGCCA": 1517, + "TTGCCT": 1518, + "TTGCCC": 1519, + "TTGCCG": 1520, + "TTGCGA": 1521, + "TTGCGT": 1522, + "TTGCGC": 1523, + "TTGCGG": 1524, + "TTGGAA": 1525, + "TTGGAT": 1526, + "TTGGAC": 1527, + "TTGGAG": 1528, + "TTGGTA": 1529, + "TTGGTT": 1530, + "TTGGTC": 1531, + "TTGGTG": 1532, + "TTGGCA": 1533, + "TTGGCT": 1534, + "TTGGCC": 1535, + "TTGGCG": 1536, + "TTGGGA": 1537, + "TTGGGT": 1538, + "TTGGGC": 1539, + "TTGGGG": 1540, + "TCAAAA": 1541, + "TCAAAT": 1542, + "TCAAAC": 1543, + "TCAAAG": 1544, + "TCAATA": 1545, + "TCAATT": 1546, + "TCAATC": 1547, + "TCAATG": 1548, + "TCAACA": 1549, + "TCAACT": 1550, + "TCAACC": 1551, + "TCAACG": 1552, + "TCAAGA": 1553, + "TCAAGT": 1554, + "TCAAGC": 1555, + "TCAAGG": 1556, + "TCATAA": 1557, + "TCATAT": 1558, + "TCATAC": 1559, + "TCATAG": 1560, + "TCATTA": 1561, + "TCATTT": 1562, + "TCATTC": 1563, + "TCATTG": 1564, + "TCATCA": 1565, + "TCATCT": 1566, + "TCATCC": 1567, + "TCATCG": 1568, + "TCATGA": 1569, + "TCATGT": 1570, + "TCATGC": 1571, + "TCATGG": 1572, + "TCACAA": 1573, + "TCACAT": 1574, + "TCACAC": 1575, + "TCACAG": 1576, + "TCACTA": 1577, + "TCACTT": 1578, + "TCACTC": 1579, + "TCACTG": 1580, + "TCACCA": 1581, + "TCACCT": 1582, + "TCACCC": 1583, + "TCACCG": 1584, + "TCACGA": 1585, + "TCACGT": 1586, + "TCACGC": 1587, + "TCACGG": 1588, + "TCAGAA": 1589, + "TCAGAT": 1590, + "TCAGAC": 1591, + "TCAGAG": 1592, + "TCAGTA": 1593, + "TCAGTT": 1594, + "TCAGTC": 1595, + "TCAGTG": 1596, + "TCAGCA": 1597, + "TCAGCT": 1598, + "TCAGCC": 1599, + "TCAGCG": 1600, + "TCAGGA": 1601, + "TCAGGT": 1602, + "TCAGGC": 1603, + "TCAGGG": 1604, + "TCTAAA": 1605, + "TCTAAT": 1606, + "TCTAAC": 1607, + "TCTAAG": 1608, + "TCTATA": 1609, + "TCTATT": 1610, + "TCTATC": 1611, + "TCTATG": 1612, + "TCTACA": 1613, + "TCTACT": 1614, + "TCTACC": 1615, + "TCTACG": 1616, + "TCTAGA": 1617, + "TCTAGT": 1618, + "TCTAGC": 1619, + "TCTAGG": 1620, + "TCTTAA": 1621, + "TCTTAT": 1622, + "TCTTAC": 1623, + "TCTTAG": 1624, + "TCTTTA": 1625, + "TCTTTT": 1626, + "TCTTTC": 1627, + "TCTTTG": 1628, + "TCTTCA": 1629, + "TCTTCT": 1630, + "TCTTCC": 1631, + "TCTTCG": 1632, + "TCTTGA": 1633, + "TCTTGT": 1634, + "TCTTGC": 1635, + "TCTTGG": 1636, + "TCTCAA": 1637, + "TCTCAT": 1638, + "TCTCAC": 1639, + "TCTCAG": 1640, + "TCTCTA": 1641, + "TCTCTT": 1642, + "TCTCTC": 1643, + "TCTCTG": 1644, + "TCTCCA": 1645, + "TCTCCT": 1646, + "TCTCCC": 1647, + "TCTCCG": 1648, + "TCTCGA": 1649, + "TCTCGT": 1650, + "TCTCGC": 1651, + "TCTCGG": 1652, + "TCTGAA": 1653, + "TCTGAT": 1654, + "TCTGAC": 1655, + "TCTGAG": 1656, + "TCTGTA": 1657, + "TCTGTT": 1658, + "TCTGTC": 1659, + "TCTGTG": 1660, + "TCTGCA": 1661, + "TCTGCT": 1662, + "TCTGCC": 1663, + "TCTGCG": 1664, + "TCTGGA": 1665, + "TCTGGT": 1666, + "TCTGGC": 1667, + "TCTGGG": 1668, + "TCCAAA": 1669, + "TCCAAT": 1670, + "TCCAAC": 1671, + "TCCAAG": 1672, + "TCCATA": 1673, + "TCCATT": 1674, + "TCCATC": 1675, + "TCCATG": 1676, + "TCCACA": 1677, + "TCCACT": 1678, + "TCCACC": 1679, + "TCCACG": 1680, + "TCCAGA": 1681, + "TCCAGT": 1682, + "TCCAGC": 1683, + "TCCAGG": 1684, + "TCCTAA": 1685, + "TCCTAT": 1686, + "TCCTAC": 1687, + "TCCTAG": 1688, + "TCCTTA": 1689, + "TCCTTT": 1690, + "TCCTTC": 1691, + "TCCTTG": 1692, + "TCCTCA": 1693, + "TCCTCT": 1694, + "TCCTCC": 1695, + "TCCTCG": 1696, + "TCCTGA": 1697, + "TCCTGT": 1698, + "TCCTGC": 1699, + "TCCTGG": 1700, + "TCCCAA": 1701, + "TCCCAT": 1702, + "TCCCAC": 1703, + "TCCCAG": 1704, + "TCCCTA": 1705, + "TCCCTT": 1706, + "TCCCTC": 1707, + "TCCCTG": 1708, + "TCCCCA": 1709, + "TCCCCT": 1710, + "TCCCCC": 1711, + "TCCCCG": 1712, + "TCCCGA": 1713, + "TCCCGT": 1714, + "TCCCGC": 1715, + "TCCCGG": 1716, + "TCCGAA": 1717, + "TCCGAT": 1718, + "TCCGAC": 1719, + "TCCGAG": 1720, + "TCCGTA": 1721, + "TCCGTT": 1722, + "TCCGTC": 1723, + "TCCGTG": 1724, + "TCCGCA": 1725, + "TCCGCT": 1726, + "TCCGCC": 1727, + "TCCGCG": 1728, + "TCCGGA": 1729, + "TCCGGT": 1730, + "TCCGGC": 1731, + "TCCGGG": 1732, + "TCGAAA": 1733, + "TCGAAT": 1734, + "TCGAAC": 1735, + "TCGAAG": 1736, + "TCGATA": 1737, + "TCGATT": 1738, + "TCGATC": 1739, + "TCGATG": 1740, + "TCGACA": 1741, + "TCGACT": 1742, + "TCGACC": 1743, + "TCGACG": 1744, + "TCGAGA": 1745, + "TCGAGT": 1746, + "TCGAGC": 1747, + "TCGAGG": 1748, + "TCGTAA": 1749, + "TCGTAT": 1750, + "TCGTAC": 1751, + "TCGTAG": 1752, + "TCGTTA": 1753, + "TCGTTT": 1754, + "TCGTTC": 1755, + "TCGTTG": 1756, + "TCGTCA": 1757, + "TCGTCT": 1758, + "TCGTCC": 1759, + "TCGTCG": 1760, + "TCGTGA": 1761, + "TCGTGT": 1762, + "TCGTGC": 1763, + "TCGTGG": 1764, + "TCGCAA": 1765, + "TCGCAT": 1766, + "TCGCAC": 1767, + "TCGCAG": 1768, + "TCGCTA": 1769, + "TCGCTT": 1770, + "TCGCTC": 1771, + "TCGCTG": 1772, + "TCGCCA": 1773, + "TCGCCT": 1774, + "TCGCCC": 1775, + "TCGCCG": 1776, + "TCGCGA": 1777, + "TCGCGT": 1778, + "TCGCGC": 1779, + "TCGCGG": 1780, + "TCGGAA": 1781, + "TCGGAT": 1782, + "TCGGAC": 1783, + "TCGGAG": 1784, + "TCGGTA": 1785, + "TCGGTT": 1786, + "TCGGTC": 1787, + "TCGGTG": 1788, + "TCGGCA": 1789, + "TCGGCT": 1790, + "TCGGCC": 1791, + "TCGGCG": 1792, + "TCGGGA": 1793, + "TCGGGT": 1794, + "TCGGGC": 1795, + "TCGGGG": 1796, + "TGAAAA": 1797, + "TGAAAT": 1798, + "TGAAAC": 1799, + "TGAAAG": 1800, + "TGAATA": 1801, + "TGAATT": 1802, + "TGAATC": 1803, + "TGAATG": 1804, + "TGAACA": 1805, + "TGAACT": 1806, + "TGAACC": 1807, + "TGAACG": 1808, + "TGAAGA": 1809, + "TGAAGT": 1810, + "TGAAGC": 1811, + "TGAAGG": 1812, + "TGATAA": 1813, + "TGATAT": 1814, + "TGATAC": 1815, + "TGATAG": 1816, + "TGATTA": 1817, + "TGATTT": 1818, + "TGATTC": 1819, + "TGATTG": 1820, + "TGATCA": 1821, + "TGATCT": 1822, + "TGATCC": 1823, + "TGATCG": 1824, + "TGATGA": 1825, + "TGATGT": 1826, + "TGATGC": 1827, + "TGATGG": 1828, + "TGACAA": 1829, + "TGACAT": 1830, + "TGACAC": 1831, + "TGACAG": 1832, + "TGACTA": 1833, + "TGACTT": 1834, + "TGACTC": 1835, + "TGACTG": 1836, + "TGACCA": 1837, + "TGACCT": 1838, + "TGACCC": 1839, + "TGACCG": 1840, + "TGACGA": 1841, + "TGACGT": 1842, + "TGACGC": 1843, + "TGACGG": 1844, + "TGAGAA": 1845, + "TGAGAT": 1846, + "TGAGAC": 1847, + "TGAGAG": 1848, + "TGAGTA": 1849, + "TGAGTT": 1850, + "TGAGTC": 1851, + "TGAGTG": 1852, + "TGAGCA": 1853, + "TGAGCT": 1854, + "TGAGCC": 1855, + "TGAGCG": 1856, + "TGAGGA": 1857, + "TGAGGT": 1858, + "TGAGGC": 1859, + "TGAGGG": 1860, + "TGTAAA": 1861, + "TGTAAT": 1862, + "TGTAAC": 1863, + "TGTAAG": 1864, + "TGTATA": 1865, + "TGTATT": 1866, + "TGTATC": 1867, + "TGTATG": 1868, + "TGTACA": 1869, + "TGTACT": 1870, + "TGTACC": 1871, + "TGTACG": 1872, + "TGTAGA": 1873, + "TGTAGT": 1874, + "TGTAGC": 1875, + "TGTAGG": 1876, + "TGTTAA": 1877, + "TGTTAT": 1878, + "TGTTAC": 1879, + "TGTTAG": 1880, + "TGTTTA": 1881, + "TGTTTT": 1882, + "TGTTTC": 1883, + "TGTTTG": 1884, + "TGTTCA": 1885, + "TGTTCT": 1886, + "TGTTCC": 1887, + "TGTTCG": 1888, + "TGTTGA": 1889, + "TGTTGT": 1890, + "TGTTGC": 1891, + "TGTTGG": 1892, + "TGTCAA": 1893, + "TGTCAT": 1894, + "TGTCAC": 1895, + "TGTCAG": 1896, + "TGTCTA": 1897, + "TGTCTT": 1898, + "TGTCTC": 1899, + "TGTCTG": 1900, + "TGTCCA": 1901, + "TGTCCT": 1902, + "TGTCCC": 1903, + "TGTCCG": 1904, + "TGTCGA": 1905, + "TGTCGT": 1906, + "TGTCGC": 1907, + "TGTCGG": 1908, + "TGTGAA": 1909, + "TGTGAT": 1910, + "TGTGAC": 1911, + "TGTGAG": 1912, + "TGTGTA": 1913, + "TGTGTT": 1914, + "TGTGTC": 1915, + "TGTGTG": 1916, + "TGTGCA": 1917, + "TGTGCT": 1918, + "TGTGCC": 1919, + "TGTGCG": 1920, + "TGTGGA": 1921, + "TGTGGT": 1922, + "TGTGGC": 1923, + "TGTGGG": 1924, + "TGCAAA": 1925, + "TGCAAT": 1926, + "TGCAAC": 1927, + "TGCAAG": 1928, + "TGCATA": 1929, + "TGCATT": 1930, + "TGCATC": 1931, + "TGCATG": 1932, + "TGCACA": 1933, + "TGCACT": 1934, + "TGCACC": 1935, + "TGCACG": 1936, + "TGCAGA": 1937, + "TGCAGT": 1938, + "TGCAGC": 1939, + "TGCAGG": 1940, + "TGCTAA": 1941, + "TGCTAT": 1942, + "TGCTAC": 1943, + "TGCTAG": 1944, + "TGCTTA": 1945, + "TGCTTT": 1946, + "TGCTTC": 1947, + "TGCTTG": 1948, + "TGCTCA": 1949, + "TGCTCT": 1950, + "TGCTCC": 1951, + "TGCTCG": 1952, + "TGCTGA": 1953, + "TGCTGT": 1954, + "TGCTGC": 1955, + "TGCTGG": 1956, + "TGCCAA": 1957, + "TGCCAT": 1958, + "TGCCAC": 1959, + "TGCCAG": 1960, + "TGCCTA": 1961, + "TGCCTT": 1962, + "TGCCTC": 1963, + "TGCCTG": 1964, + "TGCCCA": 1965, + "TGCCCT": 1966, + "TGCCCC": 1967, + "TGCCCG": 1968, + "TGCCGA": 1969, + "TGCCGT": 1970, + "TGCCGC": 1971, + "TGCCGG": 1972, + "TGCGAA": 1973, + "TGCGAT": 1974, + "TGCGAC": 1975, + "TGCGAG": 1976, + "TGCGTA": 1977, + "TGCGTT": 1978, + "TGCGTC": 1979, + "TGCGTG": 1980, + "TGCGCA": 1981, + "TGCGCT": 1982, + "TGCGCC": 1983, + "TGCGCG": 1984, + "TGCGGA": 1985, + "TGCGGT": 1986, + "TGCGGC": 1987, + "TGCGGG": 1988, + "TGGAAA": 1989, + "TGGAAT": 1990, + "TGGAAC": 1991, + "TGGAAG": 1992, + "TGGATA": 1993, + "TGGATT": 1994, + "TGGATC": 1995, + "TGGATG": 1996, + "TGGACA": 1997, + "TGGACT": 1998, + "TGGACC": 1999, + "TGGACG": 2000, + "TGGAGA": 2001, + "TGGAGT": 2002, + "TGGAGC": 2003, + "TGGAGG": 2004, + "TGGTAA": 2005, + "TGGTAT": 2006, + "TGGTAC": 2007, + "TGGTAG": 2008, + "TGGTTA": 2009, + "TGGTTT": 2010, + "TGGTTC": 2011, + "TGGTTG": 2012, + "TGGTCA": 2013, + "TGGTCT": 2014, + "TGGTCC": 2015, + "TGGTCG": 2016, + "TGGTGA": 2017, + "TGGTGT": 2018, + "TGGTGC": 2019, + "TGGTGG": 2020, + "TGGCAA": 2021, + "TGGCAT": 2022, + "TGGCAC": 2023, + "TGGCAG": 2024, + "TGGCTA": 2025, + "TGGCTT": 2026, + "TGGCTC": 2027, + "TGGCTG": 2028, + "TGGCCA": 2029, + "TGGCCT": 2030, + "TGGCCC": 2031, + "TGGCCG": 2032, + "TGGCGA": 2033, + "TGGCGT": 2034, + "TGGCGC": 2035, + "TGGCGG": 2036, + "TGGGAA": 2037, + "TGGGAT": 2038, + "TGGGAC": 2039, + "TGGGAG": 2040, + "TGGGTA": 2041, + "TGGGTT": 2042, + "TGGGTC": 2043, + "TGGGTG": 2044, + "TGGGCA": 2045, + "TGGGCT": 2046, + "TGGGCC": 2047, + "TGGGCG": 2048, + "TGGGGA": 2049, + "TGGGGT": 2050, + "TGGGGC": 2051, + "TGGGGG": 2052, + "CAAAAA": 2053, + "CAAAAT": 2054, + "CAAAAC": 2055, + "CAAAAG": 2056, + "CAAATA": 2057, + "CAAATT": 2058, + "CAAATC": 2059, + "CAAATG": 2060, + "CAAACA": 2061, + "CAAACT": 2062, + "CAAACC": 2063, + "CAAACG": 2064, + "CAAAGA": 2065, + "CAAAGT": 2066, + "CAAAGC": 2067, + "CAAAGG": 2068, + "CAATAA": 2069, + "CAATAT": 2070, + "CAATAC": 2071, + "CAATAG": 2072, + "CAATTA": 2073, + "CAATTT": 2074, + "CAATTC": 2075, + "CAATTG": 2076, + "CAATCA": 2077, + "CAATCT": 2078, + "CAATCC": 2079, + "CAATCG": 2080, + "CAATGA": 2081, + "CAATGT": 2082, + "CAATGC": 2083, + "CAATGG": 2084, + "CAACAA": 2085, + "CAACAT": 2086, + "CAACAC": 2087, + "CAACAG": 2088, + "CAACTA": 2089, + "CAACTT": 2090, + "CAACTC": 2091, + "CAACTG": 2092, + "CAACCA": 2093, + "CAACCT": 2094, + "CAACCC": 2095, + "CAACCG": 2096, + "CAACGA": 2097, + "CAACGT": 2098, + "CAACGC": 2099, + "CAACGG": 2100, + "CAAGAA": 2101, + "CAAGAT": 2102, + "CAAGAC": 2103, + "CAAGAG": 2104, + "CAAGTA": 2105, + "CAAGTT": 2106, + "CAAGTC": 2107, + "CAAGTG": 2108, + "CAAGCA": 2109, + "CAAGCT": 2110, + "CAAGCC": 2111, + "CAAGCG": 2112, + "CAAGGA": 2113, + "CAAGGT": 2114, + "CAAGGC": 2115, + "CAAGGG": 2116, + "CATAAA": 2117, + "CATAAT": 2118, + "CATAAC": 2119, + "CATAAG": 2120, + "CATATA": 2121, + "CATATT": 2122, + "CATATC": 2123, + "CATATG": 2124, + "CATACA": 2125, + "CATACT": 2126, + "CATACC": 2127, + "CATACG": 2128, + "CATAGA": 2129, + "CATAGT": 2130, + "CATAGC": 2131, + "CATAGG": 2132, + "CATTAA": 2133, + "CATTAT": 2134, + "CATTAC": 2135, + "CATTAG": 2136, + "CATTTA": 2137, + "CATTTT": 2138, + "CATTTC": 2139, + "CATTTG": 2140, + "CATTCA": 2141, + "CATTCT": 2142, + "CATTCC": 2143, + "CATTCG": 2144, + "CATTGA": 2145, + "CATTGT": 2146, + "CATTGC": 2147, + "CATTGG": 2148, + "CATCAA": 2149, + "CATCAT": 2150, + "CATCAC": 2151, + "CATCAG": 2152, + "CATCTA": 2153, + "CATCTT": 2154, + "CATCTC": 2155, + "CATCTG": 2156, + "CATCCA": 2157, + "CATCCT": 2158, + "CATCCC": 2159, + "CATCCG": 2160, + "CATCGA": 2161, + "CATCGT": 2162, + "CATCGC": 2163, + "CATCGG": 2164, + "CATGAA": 2165, + "CATGAT": 2166, + "CATGAC": 2167, + "CATGAG": 2168, + "CATGTA": 2169, + "CATGTT": 2170, + "CATGTC": 2171, + "CATGTG": 2172, + "CATGCA": 2173, + "CATGCT": 2174, + "CATGCC": 2175, + "CATGCG": 2176, + "CATGGA": 2177, + "CATGGT": 2178, + "CATGGC": 2179, + "CATGGG": 2180, + "CACAAA": 2181, + "CACAAT": 2182, + "CACAAC": 2183, + "CACAAG": 2184, + "CACATA": 2185, + "CACATT": 2186, + "CACATC": 2187, + "CACATG": 2188, + "CACACA": 2189, + "CACACT": 2190, + "CACACC": 2191, + "CACACG": 2192, + "CACAGA": 2193, + "CACAGT": 2194, + "CACAGC": 2195, + "CACAGG": 2196, + "CACTAA": 2197, + "CACTAT": 2198, + "CACTAC": 2199, + "CACTAG": 2200, + "CACTTA": 2201, + "CACTTT": 2202, + "CACTTC": 2203, + "CACTTG": 2204, + "CACTCA": 2205, + "CACTCT": 2206, + "CACTCC": 2207, + "CACTCG": 2208, + "CACTGA": 2209, + "CACTGT": 2210, + "CACTGC": 2211, + "CACTGG": 2212, + "CACCAA": 2213, + "CACCAT": 2214, + "CACCAC": 2215, + "CACCAG": 2216, + "CACCTA": 2217, + "CACCTT": 2218, + "CACCTC": 2219, + "CACCTG": 2220, + "CACCCA": 2221, + "CACCCT": 2222, + "CACCCC": 2223, + "CACCCG": 2224, + "CACCGA": 2225, + "CACCGT": 2226, + "CACCGC": 2227, + "CACCGG": 2228, + "CACGAA": 2229, + "CACGAT": 2230, + "CACGAC": 2231, + "CACGAG": 2232, + "CACGTA": 2233, + "CACGTT": 2234, + "CACGTC": 2235, + "CACGTG": 2236, + "CACGCA": 2237, + "CACGCT": 2238, + "CACGCC": 2239, + "CACGCG": 2240, + "CACGGA": 2241, + "CACGGT": 2242, + "CACGGC": 2243, + "CACGGG": 2244, + "CAGAAA": 2245, + "CAGAAT": 2246, + "CAGAAC": 2247, + "CAGAAG": 2248, + "CAGATA": 2249, + "CAGATT": 2250, + "CAGATC": 2251, + "CAGATG": 2252, + "CAGACA": 2253, + "CAGACT": 2254, + "CAGACC": 2255, + "CAGACG": 2256, + "CAGAGA": 2257, + "CAGAGT": 2258, + "CAGAGC": 2259, + "CAGAGG": 2260, + "CAGTAA": 2261, + "CAGTAT": 2262, + "CAGTAC": 2263, + "CAGTAG": 2264, + "CAGTTA": 2265, + "CAGTTT": 2266, + "CAGTTC": 2267, + "CAGTTG": 2268, + "CAGTCA": 2269, + "CAGTCT": 2270, + "CAGTCC": 2271, + "CAGTCG": 2272, + "CAGTGA": 2273, + "CAGTGT": 2274, + "CAGTGC": 2275, + "CAGTGG": 2276, + "CAGCAA": 2277, + "CAGCAT": 2278, + "CAGCAC": 2279, + "CAGCAG": 2280, + "CAGCTA": 2281, + "CAGCTT": 2282, + "CAGCTC": 2283, + "CAGCTG": 2284, + "CAGCCA": 2285, + "CAGCCT": 2286, + "CAGCCC": 2287, + "CAGCCG": 2288, + "CAGCGA": 2289, + "CAGCGT": 2290, + "CAGCGC": 2291, + "CAGCGG": 2292, + "CAGGAA": 2293, + "CAGGAT": 2294, + "CAGGAC": 2295, + "CAGGAG": 2296, + "CAGGTA": 2297, + "CAGGTT": 2298, + "CAGGTC": 2299, + "CAGGTG": 2300, + "CAGGCA": 2301, + "CAGGCT": 2302, + "CAGGCC": 2303, + "CAGGCG": 2304, + "CAGGGA": 2305, + "CAGGGT": 2306, + "CAGGGC": 2307, + "CAGGGG": 2308, + "CTAAAA": 2309, + "CTAAAT": 2310, + "CTAAAC": 2311, + "CTAAAG": 2312, + "CTAATA": 2313, + "CTAATT": 2314, + "CTAATC": 2315, + "CTAATG": 2316, + "CTAACA": 2317, + "CTAACT": 2318, + "CTAACC": 2319, + "CTAACG": 2320, + "CTAAGA": 2321, + "CTAAGT": 2322, + "CTAAGC": 2323, + "CTAAGG": 2324, + "CTATAA": 2325, + "CTATAT": 2326, + "CTATAC": 2327, + "CTATAG": 2328, + "CTATTA": 2329, + "CTATTT": 2330, + "CTATTC": 2331, + "CTATTG": 2332, + "CTATCA": 2333, + "CTATCT": 2334, + "CTATCC": 2335, + "CTATCG": 2336, + "CTATGA": 2337, + "CTATGT": 2338, + "CTATGC": 2339, + "CTATGG": 2340, + "CTACAA": 2341, + "CTACAT": 2342, + "CTACAC": 2343, + "CTACAG": 2344, + "CTACTA": 2345, + "CTACTT": 2346, + "CTACTC": 2347, + "CTACTG": 2348, + "CTACCA": 2349, + "CTACCT": 2350, + "CTACCC": 2351, + "CTACCG": 2352, + "CTACGA": 2353, + "CTACGT": 2354, + "CTACGC": 2355, + "CTACGG": 2356, + "CTAGAA": 2357, + "CTAGAT": 2358, + "CTAGAC": 2359, + "CTAGAG": 2360, + "CTAGTA": 2361, + "CTAGTT": 2362, + "CTAGTC": 2363, + "CTAGTG": 2364, + "CTAGCA": 2365, + "CTAGCT": 2366, + "CTAGCC": 2367, + "CTAGCG": 2368, + "CTAGGA": 2369, + "CTAGGT": 2370, + "CTAGGC": 2371, + "CTAGGG": 2372, + "CTTAAA": 2373, + "CTTAAT": 2374, + "CTTAAC": 2375, + "CTTAAG": 2376, + "CTTATA": 2377, + "CTTATT": 2378, + "CTTATC": 2379, + "CTTATG": 2380, + "CTTACA": 2381, + "CTTACT": 2382, + "CTTACC": 2383, + "CTTACG": 2384, + "CTTAGA": 2385, + "CTTAGT": 2386, + "CTTAGC": 2387, + "CTTAGG": 2388, + "CTTTAA": 2389, + "CTTTAT": 2390, + "CTTTAC": 2391, + "CTTTAG": 2392, + "CTTTTA": 2393, + "CTTTTT": 2394, + "CTTTTC": 2395, + "CTTTTG": 2396, + "CTTTCA": 2397, + "CTTTCT": 2398, + "CTTTCC": 2399, + "CTTTCG": 2400, + "CTTTGA": 2401, + "CTTTGT": 2402, + "CTTTGC": 2403, + "CTTTGG": 2404, + "CTTCAA": 2405, + "CTTCAT": 2406, + "CTTCAC": 2407, + "CTTCAG": 2408, + "CTTCTA": 2409, + "CTTCTT": 2410, + "CTTCTC": 2411, + "CTTCTG": 2412, + "CTTCCA": 2413, + "CTTCCT": 2414, + "CTTCCC": 2415, + "CTTCCG": 2416, + "CTTCGA": 2417, + "CTTCGT": 2418, + "CTTCGC": 2419, + "CTTCGG": 2420, + "CTTGAA": 2421, + "CTTGAT": 2422, + "CTTGAC": 2423, + "CTTGAG": 2424, + "CTTGTA": 2425, + "CTTGTT": 2426, + "CTTGTC": 2427, + "CTTGTG": 2428, + "CTTGCA": 2429, + "CTTGCT": 2430, + "CTTGCC": 2431, + "CTTGCG": 2432, + "CTTGGA": 2433, + "CTTGGT": 2434, + "CTTGGC": 2435, + "CTTGGG": 2436, + "CTCAAA": 2437, + "CTCAAT": 2438, + "CTCAAC": 2439, + "CTCAAG": 2440, + "CTCATA": 2441, + "CTCATT": 2442, + "CTCATC": 2443, + "CTCATG": 2444, + "CTCACA": 2445, + "CTCACT": 2446, + "CTCACC": 2447, + "CTCACG": 2448, + "CTCAGA": 2449, + "CTCAGT": 2450, + "CTCAGC": 2451, + "CTCAGG": 2452, + "CTCTAA": 2453, + "CTCTAT": 2454, + "CTCTAC": 2455, + "CTCTAG": 2456, + "CTCTTA": 2457, + "CTCTTT": 2458, + "CTCTTC": 2459, + "CTCTTG": 2460, + "CTCTCA": 2461, + "CTCTCT": 2462, + "CTCTCC": 2463, + "CTCTCG": 2464, + "CTCTGA": 2465, + "CTCTGT": 2466, + "CTCTGC": 2467, + "CTCTGG": 2468, + "CTCCAA": 2469, + "CTCCAT": 2470, + "CTCCAC": 2471, + "CTCCAG": 2472, + "CTCCTA": 2473, + "CTCCTT": 2474, + "CTCCTC": 2475, + "CTCCTG": 2476, + "CTCCCA": 2477, + "CTCCCT": 2478, + "CTCCCC": 2479, + "CTCCCG": 2480, + "CTCCGA": 2481, + "CTCCGT": 2482, + "CTCCGC": 2483, + "CTCCGG": 2484, + "CTCGAA": 2485, + "CTCGAT": 2486, + "CTCGAC": 2487, + "CTCGAG": 2488, + "CTCGTA": 2489, + "CTCGTT": 2490, + "CTCGTC": 2491, + "CTCGTG": 2492, + "CTCGCA": 2493, + "CTCGCT": 2494, + "CTCGCC": 2495, + "CTCGCG": 2496, + "CTCGGA": 2497, + "CTCGGT": 2498, + "CTCGGC": 2499, + "CTCGGG": 2500, + "CTGAAA": 2501, + "CTGAAT": 2502, + "CTGAAC": 2503, + "CTGAAG": 2504, + "CTGATA": 2505, + "CTGATT": 2506, + "CTGATC": 2507, + "CTGATG": 2508, + "CTGACA": 2509, + "CTGACT": 2510, + "CTGACC": 2511, + "CTGACG": 2512, + "CTGAGA": 2513, + "CTGAGT": 2514, + "CTGAGC": 2515, + "CTGAGG": 2516, + "CTGTAA": 2517, + "CTGTAT": 2518, + "CTGTAC": 2519, + "CTGTAG": 2520, + "CTGTTA": 2521, + "CTGTTT": 2522, + "CTGTTC": 2523, + "CTGTTG": 2524, + "CTGTCA": 2525, + "CTGTCT": 2526, + "CTGTCC": 2527, + "CTGTCG": 2528, + "CTGTGA": 2529, + "CTGTGT": 2530, + "CTGTGC": 2531, + "CTGTGG": 2532, + "CTGCAA": 2533, + "CTGCAT": 2534, + "CTGCAC": 2535, + "CTGCAG": 2536, + "CTGCTA": 2537, + "CTGCTT": 2538, + "CTGCTC": 2539, + "CTGCTG": 2540, + "CTGCCA": 2541, + "CTGCCT": 2542, + "CTGCCC": 2543, + "CTGCCG": 2544, + "CTGCGA": 2545, + "CTGCGT": 2546, + "CTGCGC": 2547, + "CTGCGG": 2548, + "CTGGAA": 2549, + "CTGGAT": 2550, + "CTGGAC": 2551, + "CTGGAG": 2552, + "CTGGTA": 2553, + "CTGGTT": 2554, + "CTGGTC": 2555, + "CTGGTG": 2556, + "CTGGCA": 2557, + "CTGGCT": 2558, + "CTGGCC": 2559, + "CTGGCG": 2560, + "CTGGGA": 2561, + "CTGGGT": 2562, + "CTGGGC": 2563, + "CTGGGG": 2564, + "CCAAAA": 2565, + "CCAAAT": 2566, + "CCAAAC": 2567, + "CCAAAG": 2568, + "CCAATA": 2569, + "CCAATT": 2570, + "CCAATC": 2571, + "CCAATG": 2572, + "CCAACA": 2573, + "CCAACT": 2574, + "CCAACC": 2575, + "CCAACG": 2576, + "CCAAGA": 2577, + "CCAAGT": 2578, + "CCAAGC": 2579, + "CCAAGG": 2580, + "CCATAA": 2581, + "CCATAT": 2582, + "CCATAC": 2583, + "CCATAG": 2584, + "CCATTA": 2585, + "CCATTT": 2586, + "CCATTC": 2587, + "CCATTG": 2588, + "CCATCA": 2589, + "CCATCT": 2590, + "CCATCC": 2591, + "CCATCG": 2592, + "CCATGA": 2593, + "CCATGT": 2594, + "CCATGC": 2595, + "CCATGG": 2596, + "CCACAA": 2597, + "CCACAT": 2598, + "CCACAC": 2599, + "CCACAG": 2600, + "CCACTA": 2601, + "CCACTT": 2602, + "CCACTC": 2603, + "CCACTG": 2604, + "CCACCA": 2605, + "CCACCT": 2606, + "CCACCC": 2607, + "CCACCG": 2608, + "CCACGA": 2609, + "CCACGT": 2610, + "CCACGC": 2611, + "CCACGG": 2612, + "CCAGAA": 2613, + "CCAGAT": 2614, + "CCAGAC": 2615, + "CCAGAG": 2616, + "CCAGTA": 2617, + "CCAGTT": 2618, + "CCAGTC": 2619, + "CCAGTG": 2620, + "CCAGCA": 2621, + "CCAGCT": 2622, + "CCAGCC": 2623, + "CCAGCG": 2624, + "CCAGGA": 2625, + "CCAGGT": 2626, + "CCAGGC": 2627, + "CCAGGG": 2628, + "CCTAAA": 2629, + "CCTAAT": 2630, + "CCTAAC": 2631, + "CCTAAG": 2632, + "CCTATA": 2633, + "CCTATT": 2634, + "CCTATC": 2635, + "CCTATG": 2636, + "CCTACA": 2637, + "CCTACT": 2638, + "CCTACC": 2639, + "CCTACG": 2640, + "CCTAGA": 2641, + "CCTAGT": 2642, + "CCTAGC": 2643, + "CCTAGG": 2644, + "CCTTAA": 2645, + "CCTTAT": 2646, + "CCTTAC": 2647, + "CCTTAG": 2648, + "CCTTTA": 2649, + "CCTTTT": 2650, + "CCTTTC": 2651, + "CCTTTG": 2652, + "CCTTCA": 2653, + "CCTTCT": 2654, + "CCTTCC": 2655, + "CCTTCG": 2656, + "CCTTGA": 2657, + "CCTTGT": 2658, + "CCTTGC": 2659, + "CCTTGG": 2660, + "CCTCAA": 2661, + "CCTCAT": 2662, + "CCTCAC": 2663, + "CCTCAG": 2664, + "CCTCTA": 2665, + "CCTCTT": 2666, + "CCTCTC": 2667, + "CCTCTG": 2668, + "CCTCCA": 2669, + "CCTCCT": 2670, + "CCTCCC": 2671, + "CCTCCG": 2672, + "CCTCGA": 2673, + "CCTCGT": 2674, + "CCTCGC": 2675, + "CCTCGG": 2676, + "CCTGAA": 2677, + "CCTGAT": 2678, + "CCTGAC": 2679, + "CCTGAG": 2680, + "CCTGTA": 2681, + "CCTGTT": 2682, + "CCTGTC": 2683, + "CCTGTG": 2684, + "CCTGCA": 2685, + "CCTGCT": 2686, + "CCTGCC": 2687, + "CCTGCG": 2688, + "CCTGGA": 2689, + "CCTGGT": 2690, + "CCTGGC": 2691, + "CCTGGG": 2692, + "CCCAAA": 2693, + "CCCAAT": 2694, + "CCCAAC": 2695, + "CCCAAG": 2696, + "CCCATA": 2697, + "CCCATT": 2698, + "CCCATC": 2699, + "CCCATG": 2700, + "CCCACA": 2701, + "CCCACT": 2702, + "CCCACC": 2703, + "CCCACG": 2704, + "CCCAGA": 2705, + "CCCAGT": 2706, + "CCCAGC": 2707, + "CCCAGG": 2708, + "CCCTAA": 2709, + "CCCTAT": 2710, + "CCCTAC": 2711, + "CCCTAG": 2712, + "CCCTTA": 2713, + "CCCTTT": 2714, + "CCCTTC": 2715, + "CCCTTG": 2716, + "CCCTCA": 2717, + "CCCTCT": 2718, + "CCCTCC": 2719, + "CCCTCG": 2720, + "CCCTGA": 2721, + "CCCTGT": 2722, + "CCCTGC": 2723, + "CCCTGG": 2724, + "CCCCAA": 2725, + "CCCCAT": 2726, + "CCCCAC": 2727, + "CCCCAG": 2728, + "CCCCTA": 2729, + "CCCCTT": 2730, + "CCCCTC": 2731, + "CCCCTG": 2732, + "CCCCCA": 2733, + "CCCCCT": 2734, + "CCCCCC": 2735, + "CCCCCG": 2736, + "CCCCGA": 2737, + "CCCCGT": 2738, + "CCCCGC": 2739, + "CCCCGG": 2740, + "CCCGAA": 2741, + "CCCGAT": 2742, + "CCCGAC": 2743, + "CCCGAG": 2744, + "CCCGTA": 2745, + "CCCGTT": 2746, + "CCCGTC": 2747, + "CCCGTG": 2748, + "CCCGCA": 2749, + "CCCGCT": 2750, + "CCCGCC": 2751, + "CCCGCG": 2752, + "CCCGGA": 2753, + "CCCGGT": 2754, + "CCCGGC": 2755, + "CCCGGG": 2756, + "CCGAAA": 2757, + "CCGAAT": 2758, + "CCGAAC": 2759, + "CCGAAG": 2760, + "CCGATA": 2761, + "CCGATT": 2762, + "CCGATC": 2763, + "CCGATG": 2764, + "CCGACA": 2765, + "CCGACT": 2766, + "CCGACC": 2767, + "CCGACG": 2768, + "CCGAGA": 2769, + "CCGAGT": 2770, + "CCGAGC": 2771, + "CCGAGG": 2772, + "CCGTAA": 2773, + "CCGTAT": 2774, + "CCGTAC": 2775, + "CCGTAG": 2776, + "CCGTTA": 2777, + "CCGTTT": 2778, + "CCGTTC": 2779, + "CCGTTG": 2780, + "CCGTCA": 2781, + "CCGTCT": 2782, + "CCGTCC": 2783, + "CCGTCG": 2784, + "CCGTGA": 2785, + "CCGTGT": 2786, + "CCGTGC": 2787, + "CCGTGG": 2788, + "CCGCAA": 2789, + "CCGCAT": 2790, + "CCGCAC": 2791, + "CCGCAG": 2792, + "CCGCTA": 2793, + "CCGCTT": 2794, + "CCGCTC": 2795, + "CCGCTG": 2796, + "CCGCCA": 2797, + "CCGCCT": 2798, + "CCGCCC": 2799, + "CCGCCG": 2800, + "CCGCGA": 2801, + "CCGCGT": 2802, + "CCGCGC": 2803, + "CCGCGG": 2804, + "CCGGAA": 2805, + "CCGGAT": 2806, + "CCGGAC": 2807, + "CCGGAG": 2808, + "CCGGTA": 2809, + "CCGGTT": 2810, + "CCGGTC": 2811, + "CCGGTG": 2812, + "CCGGCA": 2813, + "CCGGCT": 2814, + "CCGGCC": 2815, + "CCGGCG": 2816, + "CCGGGA": 2817, + "CCGGGT": 2818, + "CCGGGC": 2819, + "CCGGGG": 2820, + "CGAAAA": 2821, + "CGAAAT": 2822, + "CGAAAC": 2823, + "CGAAAG": 2824, + "CGAATA": 2825, + "CGAATT": 2826, + "CGAATC": 2827, + "CGAATG": 2828, + "CGAACA": 2829, + "CGAACT": 2830, + "CGAACC": 2831, + "CGAACG": 2832, + "CGAAGA": 2833, + "CGAAGT": 2834, + "CGAAGC": 2835, + "CGAAGG": 2836, + "CGATAA": 2837, + "CGATAT": 2838, + "CGATAC": 2839, + "CGATAG": 2840, + "CGATTA": 2841, + "CGATTT": 2842, + "CGATTC": 2843, + "CGATTG": 2844, + "CGATCA": 2845, + "CGATCT": 2846, + "CGATCC": 2847, + "CGATCG": 2848, + "CGATGA": 2849, + "CGATGT": 2850, + "CGATGC": 2851, + "CGATGG": 2852, + "CGACAA": 2853, + "CGACAT": 2854, + "CGACAC": 2855, + "CGACAG": 2856, + "CGACTA": 2857, + "CGACTT": 2858, + "CGACTC": 2859, + "CGACTG": 2860, + "CGACCA": 2861, + "CGACCT": 2862, + "CGACCC": 2863, + "CGACCG": 2864, + "CGACGA": 2865, + "CGACGT": 2866, + "CGACGC": 2867, + "CGACGG": 2868, + "CGAGAA": 2869, + "CGAGAT": 2870, + "CGAGAC": 2871, + "CGAGAG": 2872, + "CGAGTA": 2873, + "CGAGTT": 2874, + "CGAGTC": 2875, + "CGAGTG": 2876, + "CGAGCA": 2877, + "CGAGCT": 2878, + "CGAGCC": 2879, + "CGAGCG": 2880, + "CGAGGA": 2881, + "CGAGGT": 2882, + "CGAGGC": 2883, + "CGAGGG": 2884, + "CGTAAA": 2885, + "CGTAAT": 2886, + "CGTAAC": 2887, + "CGTAAG": 2888, + "CGTATA": 2889, + "CGTATT": 2890, + "CGTATC": 2891, + "CGTATG": 2892, + "CGTACA": 2893, + "CGTACT": 2894, + "CGTACC": 2895, + "CGTACG": 2896, + "CGTAGA": 2897, + "CGTAGT": 2898, + "CGTAGC": 2899, + "CGTAGG": 2900, + "CGTTAA": 2901, + "CGTTAT": 2902, + "CGTTAC": 2903, + "CGTTAG": 2904, + "CGTTTA": 2905, + "CGTTTT": 2906, + "CGTTTC": 2907, + "CGTTTG": 2908, + "CGTTCA": 2909, + "CGTTCT": 2910, + "CGTTCC": 2911, + "CGTTCG": 2912, + "CGTTGA": 2913, + "CGTTGT": 2914, + "CGTTGC": 2915, + "CGTTGG": 2916, + "CGTCAA": 2917, + "CGTCAT": 2918, + "CGTCAC": 2919, + "CGTCAG": 2920, + "CGTCTA": 2921, + "CGTCTT": 2922, + "CGTCTC": 2923, + "CGTCTG": 2924, + "CGTCCA": 2925, + "CGTCCT": 2926, + "CGTCCC": 2927, + "CGTCCG": 2928, + "CGTCGA": 2929, + "CGTCGT": 2930, + "CGTCGC": 2931, + "CGTCGG": 2932, + "CGTGAA": 2933, + "CGTGAT": 2934, + "CGTGAC": 2935, + "CGTGAG": 2936, + "CGTGTA": 2937, + "CGTGTT": 2938, + "CGTGTC": 2939, + "CGTGTG": 2940, + "CGTGCA": 2941, + "CGTGCT": 2942, + "CGTGCC": 2943, + "CGTGCG": 2944, + "CGTGGA": 2945, + "CGTGGT": 2946, + "CGTGGC": 2947, + "CGTGGG": 2948, + "CGCAAA": 2949, + "CGCAAT": 2950, + "CGCAAC": 2951, + "CGCAAG": 2952, + "CGCATA": 2953, + "CGCATT": 2954, + "CGCATC": 2955, + "CGCATG": 2956, + "CGCACA": 2957, + "CGCACT": 2958, + "CGCACC": 2959, + "CGCACG": 2960, + "CGCAGA": 2961, + "CGCAGT": 2962, + "CGCAGC": 2963, + "CGCAGG": 2964, + "CGCTAA": 2965, + "CGCTAT": 2966, + "CGCTAC": 2967, + "CGCTAG": 2968, + "CGCTTA": 2969, + "CGCTTT": 2970, + "CGCTTC": 2971, + "CGCTTG": 2972, + "CGCTCA": 2973, + "CGCTCT": 2974, + "CGCTCC": 2975, + "CGCTCG": 2976, + "CGCTGA": 2977, + "CGCTGT": 2978, + "CGCTGC": 2979, + "CGCTGG": 2980, + "CGCCAA": 2981, + "CGCCAT": 2982, + "CGCCAC": 2983, + "CGCCAG": 2984, + "CGCCTA": 2985, + "CGCCTT": 2986, + "CGCCTC": 2987, + "CGCCTG": 2988, + "CGCCCA": 2989, + "CGCCCT": 2990, + "CGCCCC": 2991, + "CGCCCG": 2992, + "CGCCGA": 2993, + "CGCCGT": 2994, + "CGCCGC": 2995, + "CGCCGG": 2996, + "CGCGAA": 2997, + "CGCGAT": 2998, + "CGCGAC": 2999, + "CGCGAG": 3000, + "CGCGTA": 3001, + "CGCGTT": 3002, + "CGCGTC": 3003, + "CGCGTG": 3004, + "CGCGCA": 3005, + "CGCGCT": 3006, + "CGCGCC": 3007, + "CGCGCG": 3008, + "CGCGGA": 3009, + "CGCGGT": 3010, + "CGCGGC": 3011, + "CGCGGG": 3012, + "CGGAAA": 3013, + "CGGAAT": 3014, + "CGGAAC": 3015, + "CGGAAG": 3016, + "CGGATA": 3017, + "CGGATT": 3018, + "CGGATC": 3019, + "CGGATG": 3020, + "CGGACA": 3021, + "CGGACT": 3022, + "CGGACC": 3023, + "CGGACG": 3024, + "CGGAGA": 3025, + "CGGAGT": 3026, + "CGGAGC": 3027, + "CGGAGG": 3028, + "CGGTAA": 3029, + "CGGTAT": 3030, + "CGGTAC": 3031, + "CGGTAG": 3032, + "CGGTTA": 3033, + "CGGTTT": 3034, + "CGGTTC": 3035, + "CGGTTG": 3036, + "CGGTCA": 3037, + "CGGTCT": 3038, + "CGGTCC": 3039, + "CGGTCG": 3040, + "CGGTGA": 3041, + "CGGTGT": 3042, + "CGGTGC": 3043, + "CGGTGG": 3044, + "CGGCAA": 3045, + "CGGCAT": 3046, + "CGGCAC": 3047, + "CGGCAG": 3048, + "CGGCTA": 3049, + "CGGCTT": 3050, + "CGGCTC": 3051, + "CGGCTG": 3052, + "CGGCCA": 3053, + "CGGCCT": 3054, + "CGGCCC": 3055, + "CGGCCG": 3056, + "CGGCGA": 3057, + "CGGCGT": 3058, + "CGGCGC": 3059, + "CGGCGG": 3060, + "CGGGAA": 3061, + "CGGGAT": 3062, + "CGGGAC": 3063, + "CGGGAG": 3064, + "CGGGTA": 3065, + "CGGGTT": 3066, + "CGGGTC": 3067, + "CGGGTG": 3068, + "CGGGCA": 3069, + "CGGGCT": 3070, + "CGGGCC": 3071, + "CGGGCG": 3072, + "CGGGGA": 3073, + "CGGGGT": 3074, + "CGGGGC": 3075, + "CGGGGG": 3076, + "GAAAAA": 3077, + "GAAAAT": 3078, + "GAAAAC": 3079, + "GAAAAG": 3080, + "GAAATA": 3081, + "GAAATT": 3082, + "GAAATC": 3083, + "GAAATG": 3084, + "GAAACA": 3085, + "GAAACT": 3086, + "GAAACC": 3087, + "GAAACG": 3088, + "GAAAGA": 3089, + "GAAAGT": 3090, + "GAAAGC": 3091, + "GAAAGG": 3092, + "GAATAA": 3093, + "GAATAT": 3094, + "GAATAC": 3095, + "GAATAG": 3096, + "GAATTA": 3097, + "GAATTT": 3098, + "GAATTC": 3099, + "GAATTG": 3100, + "GAATCA": 3101, + "GAATCT": 3102, + "GAATCC": 3103, + "GAATCG": 3104, + "GAATGA": 3105, + "GAATGT": 3106, + "GAATGC": 3107, + "GAATGG": 3108, + "GAACAA": 3109, + "GAACAT": 3110, + "GAACAC": 3111, + "GAACAG": 3112, + "GAACTA": 3113, + "GAACTT": 3114, + "GAACTC": 3115, + "GAACTG": 3116, + "GAACCA": 3117, + "GAACCT": 3118, + "GAACCC": 3119, + "GAACCG": 3120, + "GAACGA": 3121, + "GAACGT": 3122, + "GAACGC": 3123, + "GAACGG": 3124, + "GAAGAA": 3125, + "GAAGAT": 3126, + "GAAGAC": 3127, + "GAAGAG": 3128, + "GAAGTA": 3129, + "GAAGTT": 3130, + "GAAGTC": 3131, + "GAAGTG": 3132, + "GAAGCA": 3133, + "GAAGCT": 3134, + "GAAGCC": 3135, + "GAAGCG": 3136, + "GAAGGA": 3137, + "GAAGGT": 3138, + "GAAGGC": 3139, + "GAAGGG": 3140, + "GATAAA": 3141, + "GATAAT": 3142, + "GATAAC": 3143, + "GATAAG": 3144, + "GATATA": 3145, + "GATATT": 3146, + "GATATC": 3147, + "GATATG": 3148, + "GATACA": 3149, + "GATACT": 3150, + "GATACC": 3151, + "GATACG": 3152, + "GATAGA": 3153, + "GATAGT": 3154, + "GATAGC": 3155, + "GATAGG": 3156, + "GATTAA": 3157, + "GATTAT": 3158, + "GATTAC": 3159, + "GATTAG": 3160, + "GATTTA": 3161, + "GATTTT": 3162, + "GATTTC": 3163, + "GATTTG": 3164, + "GATTCA": 3165, + "GATTCT": 3166, + "GATTCC": 3167, + "GATTCG": 3168, + "GATTGA": 3169, + "GATTGT": 3170, + "GATTGC": 3171, + "GATTGG": 3172, + "GATCAA": 3173, + "GATCAT": 3174, + "GATCAC": 3175, + "GATCAG": 3176, + "GATCTA": 3177, + "GATCTT": 3178, + "GATCTC": 3179, + "GATCTG": 3180, + "GATCCA": 3181, + "GATCCT": 3182, + "GATCCC": 3183, + "GATCCG": 3184, + "GATCGA": 3185, + "GATCGT": 3186, + "GATCGC": 3187, + "GATCGG": 3188, + "GATGAA": 3189, + "GATGAT": 3190, + "GATGAC": 3191, + "GATGAG": 3192, + "GATGTA": 3193, + "GATGTT": 3194, + "GATGTC": 3195, + "GATGTG": 3196, + "GATGCA": 3197, + "GATGCT": 3198, + "GATGCC": 3199, + "GATGCG": 3200, + "GATGGA": 3201, + "GATGGT": 3202, + "GATGGC": 3203, + "GATGGG": 3204, + "GACAAA": 3205, + "GACAAT": 3206, + "GACAAC": 3207, + "GACAAG": 3208, + "GACATA": 3209, + "GACATT": 3210, + "GACATC": 3211, + "GACATG": 3212, + "GACACA": 3213, + "GACACT": 3214, + "GACACC": 3215, + "GACACG": 3216, + "GACAGA": 3217, + "GACAGT": 3218, + "GACAGC": 3219, + "GACAGG": 3220, + "GACTAA": 3221, + "GACTAT": 3222, + "GACTAC": 3223, + "GACTAG": 3224, + "GACTTA": 3225, + "GACTTT": 3226, + "GACTTC": 3227, + "GACTTG": 3228, + "GACTCA": 3229, + "GACTCT": 3230, + "GACTCC": 3231, + "GACTCG": 3232, + "GACTGA": 3233, + "GACTGT": 3234, + "GACTGC": 3235, + "GACTGG": 3236, + "GACCAA": 3237, + "GACCAT": 3238, + "GACCAC": 3239, + "GACCAG": 3240, + "GACCTA": 3241, + "GACCTT": 3242, + "GACCTC": 3243, + "GACCTG": 3244, + "GACCCA": 3245, + "GACCCT": 3246, + "GACCCC": 3247, + "GACCCG": 3248, + "GACCGA": 3249, + "GACCGT": 3250, + "GACCGC": 3251, + "GACCGG": 3252, + "GACGAA": 3253, + "GACGAT": 3254, + "GACGAC": 3255, + "GACGAG": 3256, + "GACGTA": 3257, + "GACGTT": 3258, + "GACGTC": 3259, + "GACGTG": 3260, + "GACGCA": 3261, + "GACGCT": 3262, + "GACGCC": 3263, + "GACGCG": 3264, + "GACGGA": 3265, + "GACGGT": 3266, + "GACGGC": 3267, + "GACGGG": 3268, + "GAGAAA": 3269, + "GAGAAT": 3270, + "GAGAAC": 3271, + "GAGAAG": 3272, + "GAGATA": 3273, + "GAGATT": 3274, + "GAGATC": 3275, + "GAGATG": 3276, + "GAGACA": 3277, + "GAGACT": 3278, + "GAGACC": 3279, + "GAGACG": 3280, + "GAGAGA": 3281, + "GAGAGT": 3282, + "GAGAGC": 3283, + "GAGAGG": 3284, + "GAGTAA": 3285, + "GAGTAT": 3286, + "GAGTAC": 3287, + "GAGTAG": 3288, + "GAGTTA": 3289, + "GAGTTT": 3290, + "GAGTTC": 3291, + "GAGTTG": 3292, + "GAGTCA": 3293, + "GAGTCT": 3294, + "GAGTCC": 3295, + "GAGTCG": 3296, + "GAGTGA": 3297, + "GAGTGT": 3298, + "GAGTGC": 3299, + "GAGTGG": 3300, + "GAGCAA": 3301, + "GAGCAT": 3302, + "GAGCAC": 3303, + "GAGCAG": 3304, + "GAGCTA": 3305, + "GAGCTT": 3306, + "GAGCTC": 3307, + "GAGCTG": 3308, + "GAGCCA": 3309, + "GAGCCT": 3310, + "GAGCCC": 3311, + "GAGCCG": 3312, + "GAGCGA": 3313, + "GAGCGT": 3314, + "GAGCGC": 3315, + "GAGCGG": 3316, + "GAGGAA": 3317, + "GAGGAT": 3318, + "GAGGAC": 3319, + "GAGGAG": 3320, + "GAGGTA": 3321, + "GAGGTT": 3322, + "GAGGTC": 3323, + "GAGGTG": 3324, + "GAGGCA": 3325, + "GAGGCT": 3326, + "GAGGCC": 3327, + "GAGGCG": 3328, + "GAGGGA": 3329, + "GAGGGT": 3330, + "GAGGGC": 3331, + "GAGGGG": 3332, + "GTAAAA": 3333, + "GTAAAT": 3334, + "GTAAAC": 3335, + "GTAAAG": 3336, + "GTAATA": 3337, + "GTAATT": 3338, + "GTAATC": 3339, + "GTAATG": 3340, + "GTAACA": 3341, + "GTAACT": 3342, + "GTAACC": 3343, + "GTAACG": 3344, + "GTAAGA": 3345, + "GTAAGT": 3346, + "GTAAGC": 3347, + "GTAAGG": 3348, + "GTATAA": 3349, + "GTATAT": 3350, + "GTATAC": 3351, + "GTATAG": 3352, + "GTATTA": 3353, + "GTATTT": 3354, + "GTATTC": 3355, + "GTATTG": 3356, + "GTATCA": 3357, + "GTATCT": 3358, + "GTATCC": 3359, + "GTATCG": 3360, + "GTATGA": 3361, + "GTATGT": 3362, + "GTATGC": 3363, + "GTATGG": 3364, + "GTACAA": 3365, + "GTACAT": 3366, + "GTACAC": 3367, + "GTACAG": 3368, + "GTACTA": 3369, + "GTACTT": 3370, + "GTACTC": 3371, + "GTACTG": 3372, + "GTACCA": 3373, + "GTACCT": 3374, + "GTACCC": 3375, + "GTACCG": 3376, + "GTACGA": 3377, + "GTACGT": 3378, + "GTACGC": 3379, + "GTACGG": 3380, + "GTAGAA": 3381, + "GTAGAT": 3382, + "GTAGAC": 3383, + "GTAGAG": 3384, + "GTAGTA": 3385, + "GTAGTT": 3386, + "GTAGTC": 3387, + "GTAGTG": 3388, + "GTAGCA": 3389, + "GTAGCT": 3390, + "GTAGCC": 3391, + "GTAGCG": 3392, + "GTAGGA": 3393, + "GTAGGT": 3394, + "GTAGGC": 3395, + "GTAGGG": 3396, + "GTTAAA": 3397, + "GTTAAT": 3398, + "GTTAAC": 3399, + "GTTAAG": 3400, + "GTTATA": 3401, + "GTTATT": 3402, + "GTTATC": 3403, + "GTTATG": 3404, + "GTTACA": 3405, + "GTTACT": 3406, + "GTTACC": 3407, + "GTTACG": 3408, + "GTTAGA": 3409, + "GTTAGT": 3410, + "GTTAGC": 3411, + "GTTAGG": 3412, + "GTTTAA": 3413, + "GTTTAT": 3414, + "GTTTAC": 3415, + "GTTTAG": 3416, + "GTTTTA": 3417, + "GTTTTT": 3418, + "GTTTTC": 3419, + "GTTTTG": 3420, + "GTTTCA": 3421, + "GTTTCT": 3422, + "GTTTCC": 3423, + "GTTTCG": 3424, + "GTTTGA": 3425, + "GTTTGT": 3426, + "GTTTGC": 3427, + "GTTTGG": 3428, + "GTTCAA": 3429, + "GTTCAT": 3430, + "GTTCAC": 3431, + "GTTCAG": 3432, + "GTTCTA": 3433, + "GTTCTT": 3434, + "GTTCTC": 3435, + "GTTCTG": 3436, + "GTTCCA": 3437, + "GTTCCT": 3438, + "GTTCCC": 3439, + "GTTCCG": 3440, + "GTTCGA": 3441, + "GTTCGT": 3442, + "GTTCGC": 3443, + "GTTCGG": 3444, + "GTTGAA": 3445, + "GTTGAT": 3446, + "GTTGAC": 3447, + "GTTGAG": 3448, + "GTTGTA": 3449, + "GTTGTT": 3450, + "GTTGTC": 3451, + "GTTGTG": 3452, + "GTTGCA": 3453, + "GTTGCT": 3454, + "GTTGCC": 3455, + "GTTGCG": 3456, + "GTTGGA": 3457, + "GTTGGT": 3458, + "GTTGGC": 3459, + "GTTGGG": 3460, + "GTCAAA": 3461, + "GTCAAT": 3462, + "GTCAAC": 3463, + "GTCAAG": 3464, + "GTCATA": 3465, + "GTCATT": 3466, + "GTCATC": 3467, + "GTCATG": 3468, + "GTCACA": 3469, + "GTCACT": 3470, + "GTCACC": 3471, + "GTCACG": 3472, + "GTCAGA": 3473, + "GTCAGT": 3474, + "GTCAGC": 3475, + "GTCAGG": 3476, + "GTCTAA": 3477, + "GTCTAT": 3478, + "GTCTAC": 3479, + "GTCTAG": 3480, + "GTCTTA": 3481, + "GTCTTT": 3482, + "GTCTTC": 3483, + "GTCTTG": 3484, + "GTCTCA": 3485, + "GTCTCT": 3486, + "GTCTCC": 3487, + "GTCTCG": 3488, + "GTCTGA": 3489, + "GTCTGT": 3490, + "GTCTGC": 3491, + "GTCTGG": 3492, + "GTCCAA": 3493, + "GTCCAT": 3494, + "GTCCAC": 3495, + "GTCCAG": 3496, + "GTCCTA": 3497, + "GTCCTT": 3498, + "GTCCTC": 3499, + "GTCCTG": 3500, + "GTCCCA": 3501, + "GTCCCT": 3502, + "GTCCCC": 3503, + "GTCCCG": 3504, + "GTCCGA": 3505, + "GTCCGT": 3506, + "GTCCGC": 3507, + "GTCCGG": 3508, + "GTCGAA": 3509, + "GTCGAT": 3510, + "GTCGAC": 3511, + "GTCGAG": 3512, + "GTCGTA": 3513, + "GTCGTT": 3514, + "GTCGTC": 3515, + "GTCGTG": 3516, + "GTCGCA": 3517, + "GTCGCT": 3518, + "GTCGCC": 3519, + "GTCGCG": 3520, + "GTCGGA": 3521, + "GTCGGT": 3522, + "GTCGGC": 3523, + "GTCGGG": 3524, + "GTGAAA": 3525, + "GTGAAT": 3526, + "GTGAAC": 3527, + "GTGAAG": 3528, + "GTGATA": 3529, + "GTGATT": 3530, + "GTGATC": 3531, + "GTGATG": 3532, + "GTGACA": 3533, + "GTGACT": 3534, + "GTGACC": 3535, + "GTGACG": 3536, + "GTGAGA": 3537, + "GTGAGT": 3538, + "GTGAGC": 3539, + "GTGAGG": 3540, + "GTGTAA": 3541, + "GTGTAT": 3542, + "GTGTAC": 3543, + "GTGTAG": 3544, + "GTGTTA": 3545, + "GTGTTT": 3546, + "GTGTTC": 3547, + "GTGTTG": 3548, + "GTGTCA": 3549, + "GTGTCT": 3550, + "GTGTCC": 3551, + "GTGTCG": 3552, + "GTGTGA": 3553, + "GTGTGT": 3554, + "GTGTGC": 3555, + "GTGTGG": 3556, + "GTGCAA": 3557, + "GTGCAT": 3558, + "GTGCAC": 3559, + "GTGCAG": 3560, + "GTGCTA": 3561, + "GTGCTT": 3562, + "GTGCTC": 3563, + "GTGCTG": 3564, + "GTGCCA": 3565, + "GTGCCT": 3566, + "GTGCCC": 3567, + "GTGCCG": 3568, + "GTGCGA": 3569, + "GTGCGT": 3570, + "GTGCGC": 3571, + "GTGCGG": 3572, + "GTGGAA": 3573, + "GTGGAT": 3574, + "GTGGAC": 3575, + "GTGGAG": 3576, + "GTGGTA": 3577, + "GTGGTT": 3578, + "GTGGTC": 3579, + "GTGGTG": 3580, + "GTGGCA": 3581, + "GTGGCT": 3582, + "GTGGCC": 3583, + "GTGGCG": 3584, + "GTGGGA": 3585, + "GTGGGT": 3586, + "GTGGGC": 3587, + "GTGGGG": 3588, + "GCAAAA": 3589, + "GCAAAT": 3590, + "GCAAAC": 3591, + "GCAAAG": 3592, + "GCAATA": 3593, + "GCAATT": 3594, + "GCAATC": 3595, + "GCAATG": 3596, + "GCAACA": 3597, + "GCAACT": 3598, + "GCAACC": 3599, + "GCAACG": 3600, + "GCAAGA": 3601, + "GCAAGT": 3602, + "GCAAGC": 3603, + "GCAAGG": 3604, + "GCATAA": 3605, + "GCATAT": 3606, + "GCATAC": 3607, + "GCATAG": 3608, + "GCATTA": 3609, + "GCATTT": 3610, + "GCATTC": 3611, + "GCATTG": 3612, + "GCATCA": 3613, + "GCATCT": 3614, + "GCATCC": 3615, + "GCATCG": 3616, + "GCATGA": 3617, + "GCATGT": 3618, + "GCATGC": 3619, + "GCATGG": 3620, + "GCACAA": 3621, + "GCACAT": 3622, + "GCACAC": 3623, + "GCACAG": 3624, + "GCACTA": 3625, + "GCACTT": 3626, + "GCACTC": 3627, + "GCACTG": 3628, + "GCACCA": 3629, + "GCACCT": 3630, + "GCACCC": 3631, + "GCACCG": 3632, + "GCACGA": 3633, + "GCACGT": 3634, + "GCACGC": 3635, + "GCACGG": 3636, + "GCAGAA": 3637, + "GCAGAT": 3638, + "GCAGAC": 3639, + "GCAGAG": 3640, + "GCAGTA": 3641, + "GCAGTT": 3642, + "GCAGTC": 3643, + "GCAGTG": 3644, + "GCAGCA": 3645, + "GCAGCT": 3646, + "GCAGCC": 3647, + "GCAGCG": 3648, + "GCAGGA": 3649, + "GCAGGT": 3650, + "GCAGGC": 3651, + "GCAGGG": 3652, + "GCTAAA": 3653, + "GCTAAT": 3654, + "GCTAAC": 3655, + "GCTAAG": 3656, + "GCTATA": 3657, + "GCTATT": 3658, + "GCTATC": 3659, + "GCTATG": 3660, + "GCTACA": 3661, + "GCTACT": 3662, + "GCTACC": 3663, + "GCTACG": 3664, + "GCTAGA": 3665, + "GCTAGT": 3666, + "GCTAGC": 3667, + "GCTAGG": 3668, + "GCTTAA": 3669, + "GCTTAT": 3670, + "GCTTAC": 3671, + "GCTTAG": 3672, + "GCTTTA": 3673, + "GCTTTT": 3674, + "GCTTTC": 3675, + "GCTTTG": 3676, + "GCTTCA": 3677, + "GCTTCT": 3678, + "GCTTCC": 3679, + "GCTTCG": 3680, + "GCTTGA": 3681, + "GCTTGT": 3682, + "GCTTGC": 3683, + "GCTTGG": 3684, + "GCTCAA": 3685, + "GCTCAT": 3686, + "GCTCAC": 3687, + "GCTCAG": 3688, + "GCTCTA": 3689, + "GCTCTT": 3690, + "GCTCTC": 3691, + "GCTCTG": 3692, + "GCTCCA": 3693, + "GCTCCT": 3694, + "GCTCCC": 3695, + "GCTCCG": 3696, + "GCTCGA": 3697, + "GCTCGT": 3698, + "GCTCGC": 3699, + "GCTCGG": 3700, + "GCTGAA": 3701, + "GCTGAT": 3702, + "GCTGAC": 3703, + "GCTGAG": 3704, + "GCTGTA": 3705, + "GCTGTT": 3706, + "GCTGTC": 3707, + "GCTGTG": 3708, + "GCTGCA": 3709, + "GCTGCT": 3710, + "GCTGCC": 3711, + "GCTGCG": 3712, + "GCTGGA": 3713, + "GCTGGT": 3714, + "GCTGGC": 3715, + "GCTGGG": 3716, + "GCCAAA": 3717, + "GCCAAT": 3718, + "GCCAAC": 3719, + "GCCAAG": 3720, + "GCCATA": 3721, + "GCCATT": 3722, + "GCCATC": 3723, + "GCCATG": 3724, + "GCCACA": 3725, + "GCCACT": 3726, + "GCCACC": 3727, + "GCCACG": 3728, + "GCCAGA": 3729, + "GCCAGT": 3730, + "GCCAGC": 3731, + "GCCAGG": 3732, + "GCCTAA": 3733, + "GCCTAT": 3734, + "GCCTAC": 3735, + "GCCTAG": 3736, + "GCCTTA": 3737, + "GCCTTT": 3738, + "GCCTTC": 3739, + "GCCTTG": 3740, + "GCCTCA": 3741, + "GCCTCT": 3742, + "GCCTCC": 3743, + "GCCTCG": 3744, + "GCCTGA": 3745, + "GCCTGT": 3746, + "GCCTGC": 3747, + "GCCTGG": 3748, + "GCCCAA": 3749, + "GCCCAT": 3750, + "GCCCAC": 3751, + "GCCCAG": 3752, + "GCCCTA": 3753, + "GCCCTT": 3754, + "GCCCTC": 3755, + "GCCCTG": 3756, + "GCCCCA": 3757, + "GCCCCT": 3758, + "GCCCCC": 3759, + "GCCCCG": 3760, + "GCCCGA": 3761, + "GCCCGT": 3762, + "GCCCGC": 3763, + "GCCCGG": 3764, + "GCCGAA": 3765, + "GCCGAT": 3766, + "GCCGAC": 3767, + "GCCGAG": 3768, + "GCCGTA": 3769, + "GCCGTT": 3770, + "GCCGTC": 3771, + "GCCGTG": 3772, + "GCCGCA": 3773, + "GCCGCT": 3774, + "GCCGCC": 3775, + "GCCGCG": 3776, + "GCCGGA": 3777, + "GCCGGT": 3778, + "GCCGGC": 3779, + "GCCGGG": 3780, + "GCGAAA": 3781, + "GCGAAT": 3782, + "GCGAAC": 3783, + "GCGAAG": 3784, + "GCGATA": 3785, + "GCGATT": 3786, + "GCGATC": 3787, + "GCGATG": 3788, + "GCGACA": 3789, + "GCGACT": 3790, + "GCGACC": 3791, + "GCGACG": 3792, + "GCGAGA": 3793, + "GCGAGT": 3794, + "GCGAGC": 3795, + "GCGAGG": 3796, + "GCGTAA": 3797, + "GCGTAT": 3798, + "GCGTAC": 3799, + "GCGTAG": 3800, + "GCGTTA": 3801, + "GCGTTT": 3802, + "GCGTTC": 3803, + "GCGTTG": 3804, + "GCGTCA": 3805, + "GCGTCT": 3806, + "GCGTCC": 3807, + "GCGTCG": 3808, + "GCGTGA": 3809, + "GCGTGT": 3810, + "GCGTGC": 3811, + "GCGTGG": 3812, + "GCGCAA": 3813, + "GCGCAT": 3814, + "GCGCAC": 3815, + "GCGCAG": 3816, + "GCGCTA": 3817, + "GCGCTT": 3818, + "GCGCTC": 3819, + "GCGCTG": 3820, + "GCGCCA": 3821, + "GCGCCT": 3822, + "GCGCCC": 3823, + "GCGCCG": 3824, + "GCGCGA": 3825, + "GCGCGT": 3826, + "GCGCGC": 3827, + "GCGCGG": 3828, + "GCGGAA": 3829, + "GCGGAT": 3830, + "GCGGAC": 3831, + "GCGGAG": 3832, + "GCGGTA": 3833, + "GCGGTT": 3834, + "GCGGTC": 3835, + "GCGGTG": 3836, + "GCGGCA": 3837, + "GCGGCT": 3838, + "GCGGCC": 3839, + "GCGGCG": 3840, + "GCGGGA": 3841, + "GCGGGT": 3842, + "GCGGGC": 3843, + "GCGGGG": 3844, + "GGAAAA": 3845, + "GGAAAT": 3846, + "GGAAAC": 3847, + "GGAAAG": 3848, + "GGAATA": 3849, + "GGAATT": 3850, + "GGAATC": 3851, + "GGAATG": 3852, + "GGAACA": 3853, + "GGAACT": 3854, + "GGAACC": 3855, + "GGAACG": 3856, + "GGAAGA": 3857, + "GGAAGT": 3858, + "GGAAGC": 3859, + "GGAAGG": 3860, + "GGATAA": 3861, + "GGATAT": 3862, + "GGATAC": 3863, + "GGATAG": 3864, + "GGATTA": 3865, + "GGATTT": 3866, + "GGATTC": 3867, + "GGATTG": 3868, + "GGATCA": 3869, + "GGATCT": 3870, + "GGATCC": 3871, + "GGATCG": 3872, + "GGATGA": 3873, + "GGATGT": 3874, + "GGATGC": 3875, + "GGATGG": 3876, + "GGACAA": 3877, + "GGACAT": 3878, + "GGACAC": 3879, + "GGACAG": 3880, + "GGACTA": 3881, + "GGACTT": 3882, + "GGACTC": 3883, + "GGACTG": 3884, + "GGACCA": 3885, + "GGACCT": 3886, + "GGACCC": 3887, + "GGACCG": 3888, + "GGACGA": 3889, + "GGACGT": 3890, + "GGACGC": 3891, + "GGACGG": 3892, + "GGAGAA": 3893, + "GGAGAT": 3894, + "GGAGAC": 3895, + "GGAGAG": 3896, + "GGAGTA": 3897, + "GGAGTT": 3898, + "GGAGTC": 3899, + "GGAGTG": 3900, + "GGAGCA": 3901, + "GGAGCT": 3902, + "GGAGCC": 3903, + "GGAGCG": 3904, + "GGAGGA": 3905, + "GGAGGT": 3906, + "GGAGGC": 3907, + "GGAGGG": 3908, + "GGTAAA": 3909, + "GGTAAT": 3910, + "GGTAAC": 3911, + "GGTAAG": 3912, + "GGTATA": 3913, + "GGTATT": 3914, + "GGTATC": 3915, + "GGTATG": 3916, + "GGTACA": 3917, + "GGTACT": 3918, + "GGTACC": 3919, + "GGTACG": 3920, + "GGTAGA": 3921, + "GGTAGT": 3922, + "GGTAGC": 3923, + "GGTAGG": 3924, + "GGTTAA": 3925, + "GGTTAT": 3926, + "GGTTAC": 3927, + "GGTTAG": 3928, + "GGTTTA": 3929, + "GGTTTT": 3930, + "GGTTTC": 3931, + "GGTTTG": 3932, + "GGTTCA": 3933, + "GGTTCT": 3934, + "GGTTCC": 3935, + "GGTTCG": 3936, + "GGTTGA": 3937, + "GGTTGT": 3938, + "GGTTGC": 3939, + "GGTTGG": 3940, + "GGTCAA": 3941, + "GGTCAT": 3942, + "GGTCAC": 3943, + "GGTCAG": 3944, + "GGTCTA": 3945, + "GGTCTT": 3946, + "GGTCTC": 3947, + "GGTCTG": 3948, + "GGTCCA": 3949, + "GGTCCT": 3950, + "GGTCCC": 3951, + "GGTCCG": 3952, + "GGTCGA": 3953, + "GGTCGT": 3954, + "GGTCGC": 3955, + "GGTCGG": 3956, + "GGTGAA": 3957, + "GGTGAT": 3958, + "GGTGAC": 3959, + "GGTGAG": 3960, + "GGTGTA": 3961, + "GGTGTT": 3962, + "GGTGTC": 3963, + "GGTGTG": 3964, + "GGTGCA": 3965, + "GGTGCT": 3966, + "GGTGCC": 3967, + "GGTGCG": 3968, + "GGTGGA": 3969, + "GGTGGT": 3970, + "GGTGGC": 3971, + "GGTGGG": 3972, + "GGCAAA": 3973, + "GGCAAT": 3974, + "GGCAAC": 3975, + "GGCAAG": 3976, + "GGCATA": 3977, + "GGCATT": 3978, + "GGCATC": 3979, + "GGCATG": 3980, + "GGCACA": 3981, + "GGCACT": 3982, + "GGCACC": 3983, + "GGCACG": 3984, + "GGCAGA": 3985, + "GGCAGT": 3986, + "GGCAGC": 3987, + "GGCAGG": 3988, + "GGCTAA": 3989, + "GGCTAT": 3990, + "GGCTAC": 3991, + "GGCTAG": 3992, + "GGCTTA": 3993, + "GGCTTT": 3994, + "GGCTTC": 3995, + "GGCTTG": 3996, + "GGCTCA": 3997, + "GGCTCT": 3998, + "GGCTCC": 3999, + "GGCTCG": 4000, + "GGCTGA": 4001, + "GGCTGT": 4002, + "GGCTGC": 4003, + "GGCTGG": 4004, + "GGCCAA": 4005, + "GGCCAT": 4006, + "GGCCAC": 4007, + "GGCCAG": 4008, + "GGCCTA": 4009, + "GGCCTT": 4010, + "GGCCTC": 4011, + "GGCCTG": 4012, + "GGCCCA": 4013, + "GGCCCT": 4014, + "GGCCCC": 4015, + "GGCCCG": 4016, + "GGCCGA": 4017, + "GGCCGT": 4018, + "GGCCGC": 4019, + "GGCCGG": 4020, + "GGCGAA": 4021, + "GGCGAT": 4022, + "GGCGAC": 4023, + "GGCGAG": 4024, + "GGCGTA": 4025, + "GGCGTT": 4026, + "GGCGTC": 4027, + "GGCGTG": 4028, + "GGCGCA": 4029, + "GGCGCT": 4030, + "GGCGCC": 4031, + "GGCGCG": 4032, + "GGCGGA": 4033, + "GGCGGT": 4034, + "GGCGGC": 4035, + "GGCGGG": 4036, + "GGGAAA": 4037, + "GGGAAT": 4038, + "GGGAAC": 4039, + "GGGAAG": 4040, + "GGGATA": 4041, + "GGGATT": 4042, + "GGGATC": 4043, + "GGGATG": 4044, + "GGGACA": 4045, + "GGGACT": 4046, + "GGGACC": 4047, + "GGGACG": 4048, + "GGGAGA": 4049, + "GGGAGT": 4050, + "GGGAGC": 4051, + "GGGAGG": 4052, + "GGGTAA": 4053, + "GGGTAT": 4054, + "GGGTAC": 4055, + "GGGTAG": 4056, + "GGGTTA": 4057, + "GGGTTT": 4058, + "GGGTTC": 4059, + "GGGTTG": 4060, + "GGGTCA": 4061, + "GGGTCT": 4062, + "GGGTCC": 4063, + "GGGTCG": 4064, + "GGGTGA": 4065, + "GGGTGT": 4066, + "GGGTGC": 4067, + "GGGTGG": 4068, + "GGGCAA": 4069, + "GGGCAT": 4070, + "GGGCAC": 4071, + "GGGCAG": 4072, + "GGGCTA": 4073, + "GGGCTT": 4074, + "GGGCTC": 4075, + "GGGCTG": 4076, + "GGGCCA": 4077, + "GGGCCT": 4078, + "GGGCCC": 4079, + "GGGCCG": 4080, + "GGGCGA": 4081, + "GGGCGT": 4082, + "GGGCGC": 4083, + "GGGCGG": 4084, + "GGGGAA": 4085, + "GGGGAT": 4086, + "GGGGAC": 4087, + "GGGGAG": 4088, + "GGGGTA": 4089, + "GGGGTT": 4090, + "GGGGTC": 4091, + "GGGGTG": 4092, + "GGGGCA": 4093, + "GGGGCT": 4094, + "GGGGCC": 4095, + "GGGGCG": 4096, + "GGGGGA": 4097, + "GGGGGT": 4098, + "GGGGGC": 4099, + "GGGGGG": 4100 + } + } +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/tokenizer_config.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..604e2b1b29af1dcb3920d2828b6a90276e5e3d02 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/tokenizer_config.json @@ -0,0 +1,58 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "[UNK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "[CLS]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "[SEP]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "4": { + "content": "[MASK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "clean_up_tokenization_spaces": true, + "cls_token": "[CLS]", + "do_basic_tokenize": true, + "do_lower_case": false, + "mask_token": "[MASK]", + "max_len": 512, + "model_max_length": 512, + "never_split": null, + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "strip_accents": null, + "tokenize_chinese_chars": true, + "tokenizer_class": "BertTokenizer", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/trainer_state.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..0c269196d2ab8d1a9852c2924979e6b0c34bc17c --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/trainer_state.json @@ -0,0 +1,1713 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 24.0, + "eval_steps": 500, + "global_step": 6000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.1, + "grad_norm": 0.2218121588230133, + "learning_rate": 5.319148936170213e-05, + "loss": 0.6927, + "step": 25 + }, + { + "epoch": 0.2, + "grad_norm": 0.44851255416870117, + "learning_rate": 0.00010638297872340425, + "loss": 0.6929, + "step": 50 + }, + { + "epoch": 0.3, + "grad_norm": 0.2440175861120224, + "learning_rate": 0.00015957446808510637, + "loss": 0.6924, + "step": 75 + }, + { + "epoch": 0.4, + "grad_norm": 0.3422006666660309, + "learning_rate": 0.0002127659574468085, + "loss": 0.6922, + "step": 100 + }, + { + "epoch": 0.5, + "grad_norm": 0.267333447933197, + "learning_rate": 0.00026595744680851064, + "loss": 0.6913, + "step": 125 + }, + { + "epoch": 0.6, + "grad_norm": 0.45699772238731384, + "learning_rate": 0.00031914893617021275, + "loss": 0.6896, + "step": 150 + }, + { + "epoch": 0.7, + "grad_norm": 0.3004947304725647, + "learning_rate": 0.0003723404255319149, + "loss": 0.6887, + "step": 175 + }, + { + "epoch": 0.8, + "grad_norm": 0.33440759778022766, + "learning_rate": 0.0003992081821181128, + "loss": 0.6857, + "step": 200 + }, + { + "epoch": 0.9, + "grad_norm": 0.2918928861618042, + "learning_rate": 0.0003975585615308479, + "loss": 0.6882, + "step": 225 + }, + { + "epoch": 1.0, + "grad_norm": 0.2934085726737976, + "learning_rate": 0.00039590894094358297, + "loss": 0.6807, + "step": 250 + }, + { + "epoch": 1.1, + "grad_norm": 0.3364830017089844, + "learning_rate": 0.0003942593203563181, + "loss": 0.6848, + "step": 275 + }, + { + "epoch": 1.2, + "grad_norm": 0.8379159569740295, + "learning_rate": 0.0003926096997690532, + "loss": 0.677, + "step": 300 + }, + { + "epoch": 1.3, + "grad_norm": 0.37671706080436707, + "learning_rate": 0.00039096007918178817, + "loss": 0.6744, + "step": 325 + }, + { + "epoch": 1.4, + "grad_norm": 0.34329330921173096, + "learning_rate": 0.0003893104585945233, + "loss": 0.6733, + "step": 350 + }, + { + "epoch": 1.5, + "grad_norm": 0.4286085367202759, + "learning_rate": 0.0003876608380072583, + "loss": 0.6772, + "step": 375 + }, + { + "epoch": 1.6, + "grad_norm": 0.3576306998729706, + "learning_rate": 0.00038601121741999343, + "loss": 0.6799, + "step": 400 + }, + { + "epoch": 1.7, + "grad_norm": 0.4636435806751251, + "learning_rate": 0.0003843615968327285, + "loss": 0.6819, + "step": 425 + }, + { + "epoch": 1.8, + "grad_norm": 0.4552135765552521, + "learning_rate": 0.0003827119762454636, + "loss": 0.6731, + "step": 450 + }, + { + "epoch": 1.9, + "grad_norm": 0.3565954566001892, + "learning_rate": 0.00038106235565819863, + "loss": 0.6718, + "step": 475 + }, + { + "epoch": 2.0, + "grad_norm": 0.5451686978340149, + "learning_rate": 0.0003794127350709337, + "loss": 0.6858, + "step": 500 + }, + { + "epoch": 2.1, + "grad_norm": 0.3730587959289551, + "learning_rate": 0.0003777631144836688, + "loss": 0.667, + "step": 525 + }, + { + "epoch": 2.2, + "grad_norm": 0.6642417907714844, + "learning_rate": 0.00037611349389640383, + "loss": 0.6662, + "step": 550 + }, + { + "epoch": 2.3, + "grad_norm": 0.3889809250831604, + "learning_rate": 0.00037446387330913894, + "loss": 0.6718, + "step": 575 + }, + { + "epoch": 2.4, + "grad_norm": 0.47203826904296875, + "learning_rate": 0.000372814252721874, + "loss": 0.6687, + "step": 600 + }, + { + "epoch": 2.5, + "grad_norm": 0.5293862819671631, + "learning_rate": 0.00037116463213460903, + "loss": 0.6655, + "step": 625 + }, + { + "epoch": 2.6, + "grad_norm": 0.36724531650543213, + "learning_rate": 0.00036951501154734414, + "loss": 0.672, + "step": 650 + }, + { + "epoch": 2.7, + "grad_norm": 0.5733087062835693, + "learning_rate": 0.0003678653909600792, + "loss": 0.6683, + "step": 675 + }, + { + "epoch": 2.8, + "grad_norm": 0.6385205984115601, + "learning_rate": 0.0003662157703728143, + "loss": 0.67, + "step": 700 + }, + { + "epoch": 2.9, + "grad_norm": 0.6146281957626343, + "learning_rate": 0.00036456614978554934, + "loss": 0.6606, + "step": 725 + }, + { + "epoch": 3.0, + "grad_norm": 0.6535729169845581, + "learning_rate": 0.00036291652919828444, + "loss": 0.6691, + "step": 750 + }, + { + "epoch": 3.1, + "grad_norm": 0.5194367170333862, + "learning_rate": 0.00036126690861101944, + "loss": 0.6517, + "step": 775 + }, + { + "epoch": 3.2, + "grad_norm": 0.7116692066192627, + "learning_rate": 0.00035961728802375454, + "loss": 0.6501, + "step": 800 + }, + { + "epoch": 3.3, + "grad_norm": 0.5338636040687561, + "learning_rate": 0.0003579676674364896, + "loss": 0.6682, + "step": 825 + }, + { + "epoch": 3.4, + "grad_norm": 0.6912877559661865, + "learning_rate": 0.0003563180468492247, + "loss": 0.6632, + "step": 850 + }, + { + "epoch": 3.5, + "grad_norm": 0.6925108432769775, + "learning_rate": 0.0003546684262619598, + "loss": 0.653, + "step": 875 + }, + { + "epoch": 3.6, + "grad_norm": 0.5360299944877625, + "learning_rate": 0.00035301880567469485, + "loss": 0.6613, + "step": 900 + }, + { + "epoch": 3.7, + "grad_norm": 0.6487969160079956, + "learning_rate": 0.0003513691850874299, + "loss": 0.6627, + "step": 925 + }, + { + "epoch": 3.8, + "grad_norm": 1.7260371446609497, + "learning_rate": 0.00034971956450016495, + "loss": 0.6436, + "step": 950 + }, + { + "epoch": 3.9, + "grad_norm": 0.5852354168891907, + "learning_rate": 0.00034806994391290005, + "loss": 0.6613, + "step": 975 + }, + { + "epoch": 4.0, + "grad_norm": 0.7802712917327881, + "learning_rate": 0.0003464203233256351, + "loss": 0.658, + "step": 1000 + }, + { + "epoch": 4.1, + "grad_norm": 0.7335640788078308, + "learning_rate": 0.0003447707027383702, + "loss": 0.6476, + "step": 1025 + }, + { + "epoch": 4.2, + "grad_norm": 0.9540443420410156, + "learning_rate": 0.00034312108215110525, + "loss": 0.6334, + "step": 1050 + }, + { + "epoch": 4.3, + "grad_norm": 1.3698925971984863, + "learning_rate": 0.0003414714615638403, + "loss": 0.6512, + "step": 1075 + }, + { + "epoch": 4.4, + "grad_norm": 0.8282333016395569, + "learning_rate": 0.0003398218409765754, + "loss": 0.6556, + "step": 1100 + }, + { + "epoch": 4.5, + "grad_norm": 1.4414807558059692, + "learning_rate": 0.00033817222038931045, + "loss": 0.6628, + "step": 1125 + }, + { + "epoch": 4.6, + "grad_norm": 0.5493957996368408, + "learning_rate": 0.00033652259980204556, + "loss": 0.6461, + "step": 1150 + }, + { + "epoch": 4.7, + "grad_norm": 1.2343393564224243, + "learning_rate": 0.0003348729792147806, + "loss": 0.6464, + "step": 1175 + }, + { + "epoch": 4.8, + "grad_norm": 0.7163398265838623, + "learning_rate": 0.0003332233586275157, + "loss": 0.6518, + "step": 1200 + }, + { + "epoch": 4.9, + "grad_norm": 0.7878938317298889, + "learning_rate": 0.00033157373804025076, + "loss": 0.6502, + "step": 1225 + }, + { + "epoch": 5.0, + "grad_norm": 0.952571451663971, + "learning_rate": 0.0003299241174529858, + "loss": 0.6463, + "step": 1250 + }, + { + "epoch": 5.1, + "grad_norm": 0.6535393595695496, + "learning_rate": 0.0003282744968657209, + "loss": 0.6361, + "step": 1275 + }, + { + "epoch": 5.2, + "grad_norm": 0.8239464163780212, + "learning_rate": 0.00032662487627845596, + "loss": 0.6264, + "step": 1300 + }, + { + "epoch": 5.3, + "grad_norm": 1.4500653743743896, + "learning_rate": 0.00032497525569119106, + "loss": 0.6544, + "step": 1325 + }, + { + "epoch": 5.4, + "grad_norm": 1.1515296697616577, + "learning_rate": 0.0003233256351039261, + "loss": 0.6629, + "step": 1350 + }, + { + "epoch": 5.5, + "grad_norm": 0.6899295449256897, + "learning_rate": 0.00032167601451666116, + "loss": 0.6453, + "step": 1375 + }, + { + "epoch": 5.6, + "grad_norm": 1.1782217025756836, + "learning_rate": 0.00032002639392939627, + "loss": 0.6362, + "step": 1400 + }, + { + "epoch": 5.7, + "grad_norm": 0.6971372961997986, + "learning_rate": 0.0003183767733421313, + "loss": 0.6593, + "step": 1425 + }, + { + "epoch": 5.8, + "grad_norm": 0.7293921709060669, + "learning_rate": 0.0003167271527548664, + "loss": 0.6318, + "step": 1450 + }, + { + "epoch": 5.9, + "grad_norm": 0.6860108375549316, + "learning_rate": 0.00031507753216760147, + "loss": 0.6402, + "step": 1475 + }, + { + "epoch": 6.0, + "grad_norm": 1.330344319343567, + "learning_rate": 0.00031342791158033657, + "loss": 0.6276, + "step": 1500 + }, + { + "epoch": 6.1, + "grad_norm": 0.7261562943458557, + "learning_rate": 0.00031177829099307157, + "loss": 0.6369, + "step": 1525 + }, + { + "epoch": 6.2, + "grad_norm": 1.1324542760849, + "learning_rate": 0.00031012867040580667, + "loss": 0.6228, + "step": 1550 + }, + { + "epoch": 6.3, + "grad_norm": 0.7002120614051819, + "learning_rate": 0.0003084790498185417, + "loss": 0.6389, + "step": 1575 + }, + { + "epoch": 6.4, + "grad_norm": 1.3143397569656372, + "learning_rate": 0.0003068294292312768, + "loss": 0.6362, + "step": 1600 + }, + { + "epoch": 6.5, + "grad_norm": 1.6375401020050049, + "learning_rate": 0.0003051798086440119, + "loss": 0.6348, + "step": 1625 + }, + { + "epoch": 6.6, + "grad_norm": 1.4343605041503906, + "learning_rate": 0.000303530188056747, + "loss": 0.6496, + "step": 1650 + }, + { + "epoch": 6.7, + "grad_norm": 0.815762996673584, + "learning_rate": 0.000301880567469482, + "loss": 0.6407, + "step": 1675 + }, + { + "epoch": 6.8, + "grad_norm": 1.068608045578003, + "learning_rate": 0.0003002309468822171, + "loss": 0.6417, + "step": 1700 + }, + { + "epoch": 6.9, + "grad_norm": 0.8775753974914551, + "learning_rate": 0.0002985813262949522, + "loss": 0.6281, + "step": 1725 + }, + { + "epoch": 7.0, + "grad_norm": 1.041817545890808, + "learning_rate": 0.00029693170570768723, + "loss": 0.6485, + "step": 1750 + }, + { + "epoch": 7.1, + "grad_norm": 0.7135840058326721, + "learning_rate": 0.00029528208512042233, + "loss": 0.6192, + "step": 1775 + }, + { + "epoch": 7.2, + "grad_norm": 1.5523930788040161, + "learning_rate": 0.0002936324645331574, + "loss": 0.636, + "step": 1800 + }, + { + "epoch": 7.3, + "grad_norm": 0.7677174210548401, + "learning_rate": 0.00029198284394589243, + "loss": 0.6213, + "step": 1825 + }, + { + "epoch": 7.4, + "grad_norm": 0.8086565732955933, + "learning_rate": 0.00029033322335862753, + "loss": 0.6292, + "step": 1850 + }, + { + "epoch": 7.5, + "grad_norm": 3.1128971576690674, + "learning_rate": 0.0002886836027713626, + "loss": 0.648, + "step": 1875 + }, + { + "epoch": 7.6, + "grad_norm": 0.9243882894515991, + "learning_rate": 0.0002870339821840977, + "loss": 0.6286, + "step": 1900 + }, + { + "epoch": 7.7, + "grad_norm": 1.8203461170196533, + "learning_rate": 0.00028538436159683273, + "loss": 0.6301, + "step": 1925 + }, + { + "epoch": 7.8, + "grad_norm": 1.0924643278121948, + "learning_rate": 0.00028373474100956784, + "loss": 0.6378, + "step": 1950 + }, + { + "epoch": 7.9, + "grad_norm": 1.0839996337890625, + "learning_rate": 0.0002820851204223029, + "loss": 0.6265, + "step": 1975 + }, + { + "epoch": 8.0, + "grad_norm": 0.7143226861953735, + "learning_rate": 0.00028043549983503794, + "loss": 0.649, + "step": 2000 + }, + { + "epoch": 8.1, + "grad_norm": 0.8778955936431885, + "learning_rate": 0.00027878587924777304, + "loss": 0.6411, + "step": 2025 + }, + { + "epoch": 8.2, + "grad_norm": 1.644904613494873, + "learning_rate": 0.0002771362586605081, + "loss": 0.6168, + "step": 2050 + }, + { + "epoch": 8.3, + "grad_norm": 0.7915159463882446, + "learning_rate": 0.0002754866380732432, + "loss": 0.6324, + "step": 2075 + }, + { + "epoch": 8.4, + "grad_norm": 1.1212801933288574, + "learning_rate": 0.00027383701748597824, + "loss": 0.6232, + "step": 2100 + }, + { + "epoch": 8.5, + "grad_norm": 0.7717731595039368, + "learning_rate": 0.0002721873968987133, + "loss": 0.624, + "step": 2125 + }, + { + "epoch": 8.6, + "grad_norm": 1.0577210187911987, + "learning_rate": 0.00027053777631144834, + "loss": 0.6296, + "step": 2150 + }, + { + "epoch": 8.7, + "grad_norm": 1.0531567335128784, + "learning_rate": 0.00026888815572418344, + "loss": 0.6345, + "step": 2175 + }, + { + "epoch": 8.8, + "grad_norm": 2.8983640670776367, + "learning_rate": 0.00026723853513691855, + "loss": 0.6142, + "step": 2200 + }, + { + "epoch": 8.9, + "grad_norm": 0.7773396968841553, + "learning_rate": 0.0002655889145496536, + "loss": 0.634, + "step": 2225 + }, + { + "epoch": 9.0, + "grad_norm": 1.0587151050567627, + "learning_rate": 0.0002639392939623887, + "loss": 0.6326, + "step": 2250 + }, + { + "epoch": 9.1, + "grad_norm": 0.8009796738624573, + "learning_rate": 0.0002622896733751237, + "loss": 0.6175, + "step": 2275 + }, + { + "epoch": 9.2, + "grad_norm": 1.2530399560928345, + "learning_rate": 0.0002606400527878588, + "loss": 0.6239, + "step": 2300 + }, + { + "epoch": 9.3, + "grad_norm": 0.7976368069648743, + "learning_rate": 0.00025899043220059385, + "loss": 0.6452, + "step": 2325 + }, + { + "epoch": 9.4, + "grad_norm": 1.455032229423523, + "learning_rate": 0.00025734081161332895, + "loss": 0.6351, + "step": 2350 + }, + { + "epoch": 9.5, + "grad_norm": 0.8916760087013245, + "learning_rate": 0.000255691191026064, + "loss": 0.5986, + "step": 2375 + }, + { + "epoch": 9.6, + "grad_norm": 1.104746699333191, + "learning_rate": 0.0002540415704387991, + "loss": 0.6104, + "step": 2400 + }, + { + "epoch": 9.7, + "grad_norm": 0.9449218511581421, + "learning_rate": 0.00025239194985153415, + "loss": 0.6262, + "step": 2425 + }, + { + "epoch": 9.8, + "grad_norm": 2.0286591053009033, + "learning_rate": 0.0002507423292642692, + "loss": 0.6124, + "step": 2450 + }, + { + "epoch": 9.9, + "grad_norm": 2.273597002029419, + "learning_rate": 0.0002490927086770043, + "loss": 0.6164, + "step": 2475 + }, + { + "epoch": 10.0, + "grad_norm": 0.8174235224723816, + "learning_rate": 0.00024744308808973936, + "loss": 0.6254, + "step": 2500 + }, + { + "epoch": 10.1, + "grad_norm": 1.0447920560836792, + "learning_rate": 0.00024579346750247446, + "loss": 0.6187, + "step": 2525 + }, + { + "epoch": 10.2, + "grad_norm": 1.0433655977249146, + "learning_rate": 0.0002441438469152095, + "loss": 0.6183, + "step": 2550 + }, + { + "epoch": 10.3, + "grad_norm": 1.6479028463363647, + "learning_rate": 0.00024249422632794456, + "loss": 0.6353, + "step": 2575 + }, + { + "epoch": 10.4, + "grad_norm": 0.8032572865486145, + "learning_rate": 0.00024084460574067963, + "loss": 0.6206, + "step": 2600 + }, + { + "epoch": 10.5, + "grad_norm": 1.0675058364868164, + "learning_rate": 0.0002391949851534147, + "loss": 0.6168, + "step": 2625 + }, + { + "epoch": 10.6, + "grad_norm": 1.2228609323501587, + "learning_rate": 0.0002375453645661498, + "loss": 0.6007, + "step": 2650 + }, + { + "epoch": 10.7, + "grad_norm": 1.1027027368545532, + "learning_rate": 0.00023589574397888486, + "loss": 0.6227, + "step": 2675 + }, + { + "epoch": 10.8, + "grad_norm": 1.2949633598327637, + "learning_rate": 0.00023424612339161997, + "loss": 0.6179, + "step": 2700 + }, + { + "epoch": 10.9, + "grad_norm": 1.235032320022583, + "learning_rate": 0.000232596502804355, + "loss": 0.6216, + "step": 2725 + }, + { + "epoch": 11.0, + "grad_norm": 1.0087182521820068, + "learning_rate": 0.00023094688221709007, + "loss": 0.6286, + "step": 2750 + }, + { + "epoch": 11.1, + "grad_norm": 1.301207423210144, + "learning_rate": 0.00022929726162982514, + "loss": 0.5953, + "step": 2775 + }, + { + "epoch": 11.2, + "grad_norm": 1.6798988580703735, + "learning_rate": 0.00022764764104256022, + "loss": 0.6331, + "step": 2800 + }, + { + "epoch": 11.3, + "grad_norm": 1.1364004611968994, + "learning_rate": 0.0002259980204552953, + "loss": 0.6047, + "step": 2825 + }, + { + "epoch": 11.4, + "grad_norm": 0.9963313341140747, + "learning_rate": 0.00022434839986803037, + "loss": 0.6192, + "step": 2850 + }, + { + "epoch": 11.5, + "grad_norm": 1.5672775506973267, + "learning_rate": 0.00022269877928076542, + "loss": 0.6063, + "step": 2875 + }, + { + "epoch": 11.6, + "grad_norm": 1.4239299297332764, + "learning_rate": 0.0002210491586935005, + "loss": 0.6058, + "step": 2900 + }, + { + "epoch": 11.7, + "grad_norm": 1.5440959930419922, + "learning_rate": 0.00021939953810623557, + "loss": 0.6239, + "step": 2925 + }, + { + "epoch": 11.8, + "grad_norm": 1.35140860080719, + "learning_rate": 0.00021774991751897065, + "loss": 0.6241, + "step": 2950 + }, + { + "epoch": 11.9, + "grad_norm": 1.9355136156082153, + "learning_rate": 0.00021610029693170573, + "loss": 0.6135, + "step": 2975 + }, + { + "epoch": 12.0, + "grad_norm": 1.134037733078003, + "learning_rate": 0.0002144506763444408, + "loss": 0.6142, + "step": 3000 + }, + { + "epoch": 12.1, + "grad_norm": 1.094366192817688, + "learning_rate": 0.00021280105575717582, + "loss": 0.6226, + "step": 3025 + }, + { + "epoch": 12.2, + "grad_norm": 1.1197398900985718, + "learning_rate": 0.00021115143516991093, + "loss": 0.6073, + "step": 3050 + }, + { + "epoch": 12.3, + "grad_norm": 1.240188717842102, + "learning_rate": 0.000209501814582646, + "loss": 0.6265, + "step": 3075 + }, + { + "epoch": 12.4, + "grad_norm": 0.9774200916290283, + "learning_rate": 0.00020785219399538108, + "loss": 0.6361, + "step": 3100 + }, + { + "epoch": 12.5, + "grad_norm": 0.902847409248352, + "learning_rate": 0.00020620257340811616, + "loss": 0.6129, + "step": 3125 + }, + { + "epoch": 12.6, + "grad_norm": 0.8658815622329712, + "learning_rate": 0.00020455295282085123, + "loss": 0.5953, + "step": 3150 + }, + { + "epoch": 12.7, + "grad_norm": 1.4172385931015015, + "learning_rate": 0.00020290333223358626, + "loss": 0.6158, + "step": 3175 + }, + { + "epoch": 12.8, + "grad_norm": 2.6393818855285645, + "learning_rate": 0.00020125371164632133, + "loss": 0.5978, + "step": 3200 + }, + { + "epoch": 12.9, + "grad_norm": 2.4407801628112793, + "learning_rate": 0.0001996040910590564, + "loss": 0.6178, + "step": 3225 + }, + { + "epoch": 13.0, + "grad_norm": 1.2797659635543823, + "learning_rate": 0.00019795447047179148, + "loss": 0.5849, + "step": 3250 + }, + { + "epoch": 13.1, + "grad_norm": 1.590224266052246, + "learning_rate": 0.0001963048498845266, + "loss": 0.5983, + "step": 3275 + }, + { + "epoch": 13.2, + "grad_norm": 1.155685544013977, + "learning_rate": 0.00019465522929726164, + "loss": 0.6301, + "step": 3300 + }, + { + "epoch": 13.3, + "grad_norm": 1.701331377029419, + "learning_rate": 0.00019300560870999671, + "loss": 0.5865, + "step": 3325 + }, + { + "epoch": 13.4, + "grad_norm": 1.2553105354309082, + "learning_rate": 0.0001913559881227318, + "loss": 0.6113, + "step": 3350 + }, + { + "epoch": 13.5, + "grad_norm": 1.495653510093689, + "learning_rate": 0.00018970636753546684, + "loss": 0.6001, + "step": 3375 + }, + { + "epoch": 13.6, + "grad_norm": 1.2800766229629517, + "learning_rate": 0.00018805674694820192, + "loss": 0.6013, + "step": 3400 + }, + { + "epoch": 13.7, + "grad_norm": 1.0035955905914307, + "learning_rate": 0.000186407126360937, + "loss": 0.621, + "step": 3425 + }, + { + "epoch": 13.8, + "grad_norm": 1.0515904426574707, + "learning_rate": 0.00018475750577367207, + "loss": 0.6297, + "step": 3450 + }, + { + "epoch": 13.9, + "grad_norm": 2.071884870529175, + "learning_rate": 0.00018310788518640715, + "loss": 0.6004, + "step": 3475 + }, + { + "epoch": 14.0, + "grad_norm": 1.4622681140899658, + "learning_rate": 0.00018145826459914222, + "loss": 0.6021, + "step": 3500 + }, + { + "epoch": 14.1, + "grad_norm": 1.1860116720199585, + "learning_rate": 0.00017980864401187727, + "loss": 0.602, + "step": 3525 + }, + { + "epoch": 14.2, + "grad_norm": 1.0958752632141113, + "learning_rate": 0.00017815902342461235, + "loss": 0.6084, + "step": 3550 + }, + { + "epoch": 14.3, + "grad_norm": 1.4344367980957031, + "learning_rate": 0.00017650940283734742, + "loss": 0.6332, + "step": 3575 + }, + { + "epoch": 14.4, + "grad_norm": 1.1095603704452515, + "learning_rate": 0.00017485978225008247, + "loss": 0.5945, + "step": 3600 + }, + { + "epoch": 14.5, + "grad_norm": 1.416272521018982, + "learning_rate": 0.00017321016166281755, + "loss": 0.5964, + "step": 3625 + }, + { + "epoch": 14.6, + "grad_norm": 0.9727759957313538, + "learning_rate": 0.00017156054107555263, + "loss": 0.6231, + "step": 3650 + }, + { + "epoch": 14.7, + "grad_norm": 0.9979771375656128, + "learning_rate": 0.0001699109204882877, + "loss": 0.5849, + "step": 3675 + }, + { + "epoch": 14.8, + "grad_norm": 1.2574740648269653, + "learning_rate": 0.00016826129990102278, + "loss": 0.6076, + "step": 3700 + }, + { + "epoch": 14.9, + "grad_norm": 1.1657873392105103, + "learning_rate": 0.00016661167931375785, + "loss": 0.6111, + "step": 3725 + }, + { + "epoch": 15.0, + "grad_norm": 1.767681360244751, + "learning_rate": 0.0001649620587264929, + "loss": 0.6025, + "step": 3750 + }, + { + "epoch": 15.1, + "grad_norm": 1.3131200075149536, + "learning_rate": 0.00016331243813922798, + "loss": 0.6039, + "step": 3775 + }, + { + "epoch": 15.2, + "grad_norm": 1.1376745700836182, + "learning_rate": 0.00016166281755196306, + "loss": 0.5995, + "step": 3800 + }, + { + "epoch": 15.3, + "grad_norm": 1.2298001050949097, + "learning_rate": 0.00016001319696469813, + "loss": 0.602, + "step": 3825 + }, + { + "epoch": 15.4, + "grad_norm": 2.735769271850586, + "learning_rate": 0.0001583635763774332, + "loss": 0.5902, + "step": 3850 + }, + { + "epoch": 15.5, + "grad_norm": 1.3570619821548462, + "learning_rate": 0.00015671395579016829, + "loss": 0.5963, + "step": 3875 + }, + { + "epoch": 15.6, + "grad_norm": 1.6737253665924072, + "learning_rate": 0.00015506433520290334, + "loss": 0.6002, + "step": 3900 + }, + { + "epoch": 15.7, + "grad_norm": 1.386838674545288, + "learning_rate": 0.0001534147146156384, + "loss": 0.5905, + "step": 3925 + }, + { + "epoch": 15.8, + "grad_norm": 1.670160174369812, + "learning_rate": 0.0001517650940283735, + "loss": 0.6, + "step": 3950 + }, + { + "epoch": 15.9, + "grad_norm": 1.1611871719360352, + "learning_rate": 0.00015011547344110854, + "loss": 0.616, + "step": 3975 + }, + { + "epoch": 16.0, + "grad_norm": 0.9084995985031128, + "learning_rate": 0.00014846585285384361, + "loss": 0.6036, + "step": 4000 + }, + { + "epoch": 16.1, + "grad_norm": 2.1180953979492188, + "learning_rate": 0.0001468162322665787, + "loss": 0.5896, + "step": 4025 + }, + { + "epoch": 16.2, + "grad_norm": 1.9384114742279053, + "learning_rate": 0.00014516661167931377, + "loss": 0.584, + "step": 4050 + }, + { + "epoch": 16.3, + "grad_norm": 1.9472033977508545, + "learning_rate": 0.00014351699109204884, + "loss": 0.6059, + "step": 4075 + }, + { + "epoch": 16.4, + "grad_norm": 1.6221973896026611, + "learning_rate": 0.00014186737050478392, + "loss": 0.5934, + "step": 4100 + }, + { + "epoch": 16.5, + "grad_norm": 2.1453840732574463, + "learning_rate": 0.00014021774991751897, + "loss": 0.5846, + "step": 4125 + }, + { + "epoch": 16.6, + "grad_norm": 2.0782151222229004, + "learning_rate": 0.00013856812933025404, + "loss": 0.6131, + "step": 4150 + }, + { + "epoch": 16.7, + "grad_norm": 1.437988519668579, + "learning_rate": 0.00013691850874298912, + "loss": 0.6082, + "step": 4175 + }, + { + "epoch": 16.8, + "grad_norm": 1.7626062631607056, + "learning_rate": 0.00013526888815572417, + "loss": 0.5928, + "step": 4200 + }, + { + "epoch": 16.9, + "grad_norm": 1.0071998834609985, + "learning_rate": 0.00013361926756845927, + "loss": 0.6019, + "step": 4225 + }, + { + "epoch": 17.0, + "grad_norm": 1.1568998098373413, + "learning_rate": 0.00013196964698119435, + "loss": 0.6196, + "step": 4250 + }, + { + "epoch": 17.1, + "grad_norm": 1.9417431354522705, + "learning_rate": 0.0001303200263939294, + "loss": 0.607, + "step": 4275 + }, + { + "epoch": 17.2, + "grad_norm": 1.0894469022750854, + "learning_rate": 0.00012867040580666448, + "loss": 0.5874, + "step": 4300 + }, + { + "epoch": 17.3, + "grad_norm": 1.3170983791351318, + "learning_rate": 0.00012702078521939955, + "loss": 0.5988, + "step": 4325 + }, + { + "epoch": 17.4, + "grad_norm": 1.0083190202713013, + "learning_rate": 0.0001253711646321346, + "loss": 0.6021, + "step": 4350 + }, + { + "epoch": 17.5, + "grad_norm": 1.24309504032135, + "learning_rate": 0.00012372154404486968, + "loss": 0.6097, + "step": 4375 + }, + { + "epoch": 17.6, + "grad_norm": 1.5258588790893555, + "learning_rate": 0.00012207192345760475, + "loss": 0.5891, + "step": 4400 + }, + { + "epoch": 17.7, + "grad_norm": 2.491304636001587, + "learning_rate": 0.00012042230287033982, + "loss": 0.6049, + "step": 4425 + }, + { + "epoch": 17.8, + "grad_norm": 1.737715482711792, + "learning_rate": 0.0001187726822830749, + "loss": 0.5888, + "step": 4450 + }, + { + "epoch": 17.9, + "grad_norm": 2.217761754989624, + "learning_rate": 0.00011712306169580998, + "loss": 0.6075, + "step": 4475 + }, + { + "epoch": 18.0, + "grad_norm": 1.4966678619384766, + "learning_rate": 0.00011547344110854503, + "loss": 0.5962, + "step": 4500 + }, + { + "epoch": 18.1, + "grad_norm": 1.1560686826705933, + "learning_rate": 0.00011382382052128011, + "loss": 0.5654, + "step": 4525 + }, + { + "epoch": 18.2, + "grad_norm": 1.1144013404846191, + "learning_rate": 0.00011217419993401519, + "loss": 0.6001, + "step": 4550 + }, + { + "epoch": 18.3, + "grad_norm": 5.605530738830566, + "learning_rate": 0.00011052457934675025, + "loss": 0.5977, + "step": 4575 + }, + { + "epoch": 18.4, + "grad_norm": 1.0827083587646484, + "learning_rate": 0.00010887495875948532, + "loss": 0.5954, + "step": 4600 + }, + { + "epoch": 18.5, + "grad_norm": 3.0124075412750244, + "learning_rate": 0.0001072253381722204, + "loss": 0.5921, + "step": 4625 + }, + { + "epoch": 18.6, + "grad_norm": 1.5420926809310913, + "learning_rate": 0.00010557571758495546, + "loss": 0.5832, + "step": 4650 + }, + { + "epoch": 18.7, + "grad_norm": 1.5770180225372314, + "learning_rate": 0.00010392609699769054, + "loss": 0.6138, + "step": 4675 + }, + { + "epoch": 18.8, + "grad_norm": 1.9862656593322754, + "learning_rate": 0.00010227647641042562, + "loss": 0.595, + "step": 4700 + }, + { + "epoch": 18.9, + "grad_norm": 1.6087626218795776, + "learning_rate": 0.00010062685582316067, + "loss": 0.5965, + "step": 4725 + }, + { + "epoch": 19.0, + "grad_norm": 0.9878010749816895, + "learning_rate": 9.897723523589574e-05, + "loss": 0.5807, + "step": 4750 + }, + { + "epoch": 19.1, + "grad_norm": 1.3701050281524658, + "learning_rate": 9.732761464863082e-05, + "loss": 0.5953, + "step": 4775 + }, + { + "epoch": 19.2, + "grad_norm": 1.4488855600357056, + "learning_rate": 9.56779940613659e-05, + "loss": 0.578, + "step": 4800 + }, + { + "epoch": 19.3, + "grad_norm": 1.8450454473495483, + "learning_rate": 9.402837347410096e-05, + "loss": 0.5734, + "step": 4825 + }, + { + "epoch": 19.4, + "grad_norm": 1.2953295707702637, + "learning_rate": 9.237875288683603e-05, + "loss": 0.6188, + "step": 4850 + }, + { + "epoch": 19.5, + "grad_norm": 4.279994010925293, + "learning_rate": 9.072913229957111e-05, + "loss": 0.5712, + "step": 4875 + }, + { + "epoch": 19.6, + "grad_norm": 1.4677122831344604, + "learning_rate": 8.907951171230617e-05, + "loss": 0.5921, + "step": 4900 + }, + { + "epoch": 19.7, + "grad_norm": 2.559695243835449, + "learning_rate": 8.742989112504124e-05, + "loss": 0.599, + "step": 4925 + }, + { + "epoch": 19.8, + "grad_norm": 1.063706874847412, + "learning_rate": 8.578027053777631e-05, + "loss": 0.603, + "step": 4950 + }, + { + "epoch": 19.9, + "grad_norm": 1.1021010875701904, + "learning_rate": 8.413064995051139e-05, + "loss": 0.5841, + "step": 4975 + }, + { + "epoch": 20.0, + "grad_norm": 1.2500898838043213, + "learning_rate": 8.248102936324645e-05, + "loss": 0.5946, + "step": 5000 + }, + { + "epoch": 20.1, + "grad_norm": 1.2104499340057373, + "learning_rate": 8.083140877598153e-05, + "loss": 0.5906, + "step": 5025 + }, + { + "epoch": 20.2, + "grad_norm": 2.4167215824127197, + "learning_rate": 7.91817881887166e-05, + "loss": 0.5843, + "step": 5050 + }, + { + "epoch": 20.3, + "grad_norm": 1.2537106275558472, + "learning_rate": 7.753216760145167e-05, + "loss": 0.5924, + "step": 5075 + }, + { + "epoch": 20.4, + "grad_norm": 1.244513750076294, + "learning_rate": 7.588254701418674e-05, + "loss": 0.6031, + "step": 5100 + }, + { + "epoch": 20.5, + "grad_norm": 2.094501256942749, + "learning_rate": 7.423292642692181e-05, + "loss": 0.5909, + "step": 5125 + }, + { + "epoch": 20.6, + "grad_norm": 0.9758375287055969, + "learning_rate": 7.258330583965688e-05, + "loss": 0.5736, + "step": 5150 + }, + { + "epoch": 20.7, + "grad_norm": 1.5794312953948975, + "learning_rate": 7.093368525239196e-05, + "loss": 0.5817, + "step": 5175 + }, + { + "epoch": 20.8, + "grad_norm": 2.009289264678955, + "learning_rate": 6.928406466512702e-05, + "loss": 0.5912, + "step": 5200 + }, + { + "epoch": 20.9, + "grad_norm": 1.850706696510315, + "learning_rate": 6.763444407786209e-05, + "loss": 0.5851, + "step": 5225 + }, + { + "epoch": 21.0, + "grad_norm": 1.3678967952728271, + "learning_rate": 6.598482349059718e-05, + "loss": 0.5884, + "step": 5250 + }, + { + "epoch": 21.1, + "grad_norm": 1.3373990058898926, + "learning_rate": 6.433520290333224e-05, + "loss": 0.6052, + "step": 5275 + }, + { + "epoch": 21.2, + "grad_norm": 1.510580062866211, + "learning_rate": 6.26855823160673e-05, + "loss": 0.6049, + "step": 5300 + }, + { + "epoch": 21.3, + "grad_norm": 1.514527678489685, + "learning_rate": 6.103596172880238e-05, + "loss": 0.5617, + "step": 5325 + }, + { + "epoch": 21.4, + "grad_norm": 1.1603538990020752, + "learning_rate": 5.938634114153745e-05, + "loss": 0.5852, + "step": 5350 + }, + { + "epoch": 21.5, + "grad_norm": 2.2692604064941406, + "learning_rate": 5.7736720554272516e-05, + "loss": 0.5863, + "step": 5375 + }, + { + "epoch": 21.6, + "grad_norm": 2.5274813175201416, + "learning_rate": 5.608709996700759e-05, + "loss": 0.5857, + "step": 5400 + }, + { + "epoch": 21.7, + "grad_norm": 2.518244504928589, + "learning_rate": 5.443747937974266e-05, + "loss": 0.5899, + "step": 5425 + }, + { + "epoch": 21.8, + "grad_norm": 3.0091376304626465, + "learning_rate": 5.278785879247773e-05, + "loss": 0.5882, + "step": 5450 + }, + { + "epoch": 21.9, + "grad_norm": 1.8518242835998535, + "learning_rate": 5.113823820521281e-05, + "loss": 0.5515, + "step": 5475 + }, + { + "epoch": 22.0, + "grad_norm": 1.4333727359771729, + "learning_rate": 4.948861761794787e-05, + "loss": 0.5748, + "step": 5500 + }, + { + "epoch": 22.1, + "grad_norm": 1.7678550481796265, + "learning_rate": 4.783899703068295e-05, + "loss": 0.5745, + "step": 5525 + }, + { + "epoch": 22.2, + "grad_norm": 1.4223382472991943, + "learning_rate": 4.618937644341802e-05, + "loss": 0.5795, + "step": 5550 + }, + { + "epoch": 22.3, + "grad_norm": 2.587275505065918, + "learning_rate": 4.453975585615309e-05, + "loss": 0.5797, + "step": 5575 + }, + { + "epoch": 22.4, + "grad_norm": 1.3082376718521118, + "learning_rate": 4.2890135268888156e-05, + "loss": 0.6038, + "step": 5600 + }, + { + "epoch": 22.5, + "grad_norm": 1.264905333518982, + "learning_rate": 4.1240514681623226e-05, + "loss": 0.5558, + "step": 5625 + }, + { + "epoch": 22.6, + "grad_norm": 1.301184058189392, + "learning_rate": 3.95908940943583e-05, + "loss": 0.6024, + "step": 5650 + }, + { + "epoch": 22.7, + "grad_norm": 3.1338558197021484, + "learning_rate": 3.794127350709337e-05, + "loss": 0.5887, + "step": 5675 + }, + { + "epoch": 22.8, + "grad_norm": 1.5282409191131592, + "learning_rate": 3.629165291982844e-05, + "loss": 0.5874, + "step": 5700 + }, + { + "epoch": 22.9, + "grad_norm": 1.4762812852859497, + "learning_rate": 3.464203233256351e-05, + "loss": 0.5764, + "step": 5725 + }, + { + "epoch": 23.0, + "grad_norm": 4.048628330230713, + "learning_rate": 3.299241174529859e-05, + "loss": 0.571, + "step": 5750 + }, + { + "epoch": 23.1, + "grad_norm": 1.2434217929840088, + "learning_rate": 3.134279115803365e-05, + "loss": 0.5735, + "step": 5775 + }, + { + "epoch": 23.2, + "grad_norm": 1.442742943763733, + "learning_rate": 2.9693170570768723e-05, + "loss": 0.5957, + "step": 5800 + }, + { + "epoch": 23.3, + "grad_norm": 2.088374137878418, + "learning_rate": 2.8043549983503796e-05, + "loss": 0.6164, + "step": 5825 + }, + { + "epoch": 23.4, + "grad_norm": 1.1105170249938965, + "learning_rate": 2.6393929396238866e-05, + "loss": 0.5732, + "step": 5850 + }, + { + "epoch": 23.5, + "grad_norm": 1.6173477172851562, + "learning_rate": 2.4744308808973936e-05, + "loss": 0.5698, + "step": 5875 + }, + { + "epoch": 23.6, + "grad_norm": 1.760382056236267, + "learning_rate": 2.309468822170901e-05, + "loss": 0.565, + "step": 5900 + }, + { + "epoch": 23.7, + "grad_norm": 2.0124967098236084, + "learning_rate": 2.1445067634444078e-05, + "loss": 0.5837, + "step": 5925 + }, + { + "epoch": 23.8, + "grad_norm": 1.8501543998718262, + "learning_rate": 1.979544704717915e-05, + "loss": 0.555, + "step": 5950 + }, + { + "epoch": 23.9, + "grad_norm": 1.4653395414352417, + "learning_rate": 1.814582645991422e-05, + "loss": 0.5742, + "step": 5975 + }, + { + "epoch": 24.0, + "grad_norm": 2.650632381439209, + "learning_rate": 1.6496205872649294e-05, + "loss": 0.5917, + "step": 6000 + } + ], + "logging_steps": 25, + "max_steps": 6250, + "num_input_tokens_seen": 0, + "num_train_epochs": 25, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 332286197760000.0, + "train_batch_size": 20, + "trial_name": null, + "trial_params": null +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/training_args.bin b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..540db3fbee786996d9705d768d31ba6c371187be --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dc7816b5f3fd38f01f9ba38a197ce070cde06e4c90e5260cb4814f34f628907d +size 5304 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/vocab.txt b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/vocab.txt new file mode 100644 index 0000000000000000000000000000000000000000..cc9ad53732e21ea9180f18e1fb67a38cc5702d4c --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6000/vocab.txt @@ -0,0 +1,4101 @@ +[PAD] +[UNK] +[CLS] +[SEP] +[MASK] +AAAAAA +AAAAAT +AAAAAC +AAAAAG +AAAATA +AAAATT +AAAATC +AAAATG +AAAACA +AAAACT +AAAACC +AAAACG +AAAAGA +AAAAGT +AAAAGC +AAAAGG +AAATAA +AAATAT +AAATAC +AAATAG +AAATTA +AAATTT +AAATTC +AAATTG +AAATCA +AAATCT +AAATCC +AAATCG +AAATGA +AAATGT +AAATGC +AAATGG +AAACAA +AAACAT +AAACAC +AAACAG +AAACTA +AAACTT +AAACTC +AAACTG +AAACCA +AAACCT +AAACCC +AAACCG +AAACGA +AAACGT +AAACGC +AAACGG +AAAGAA +AAAGAT +AAAGAC +AAAGAG +AAAGTA +AAAGTT +AAAGTC +AAAGTG +AAAGCA +AAAGCT +AAAGCC +AAAGCG +AAAGGA +AAAGGT +AAAGGC +AAAGGG +AATAAA +AATAAT +AATAAC +AATAAG +AATATA +AATATT +AATATC +AATATG +AATACA +AATACT +AATACC +AATACG +AATAGA +AATAGT +AATAGC +AATAGG +AATTAA +AATTAT +AATTAC +AATTAG +AATTTA +AATTTT +AATTTC +AATTTG +AATTCA +AATTCT +AATTCC +AATTCG +AATTGA +AATTGT +AATTGC +AATTGG +AATCAA +AATCAT +AATCAC +AATCAG +AATCTA +AATCTT +AATCTC +AATCTG +AATCCA +AATCCT +AATCCC +AATCCG +AATCGA +AATCGT +AATCGC +AATCGG +AATGAA +AATGAT +AATGAC +AATGAG +AATGTA +AATGTT +AATGTC +AATGTG +AATGCA +AATGCT +AATGCC +AATGCG +AATGGA +AATGGT +AATGGC +AATGGG +AACAAA +AACAAT +AACAAC +AACAAG +AACATA +AACATT +AACATC +AACATG +AACACA +AACACT +AACACC +AACACG +AACAGA +AACAGT +AACAGC +AACAGG +AACTAA +AACTAT +AACTAC +AACTAG +AACTTA +AACTTT +AACTTC +AACTTG +AACTCA +AACTCT +AACTCC +AACTCG +AACTGA +AACTGT +AACTGC +AACTGG +AACCAA +AACCAT +AACCAC +AACCAG +AACCTA +AACCTT +AACCTC +AACCTG +AACCCA +AACCCT +AACCCC +AACCCG +AACCGA +AACCGT +AACCGC +AACCGG +AACGAA +AACGAT +AACGAC +AACGAG +AACGTA +AACGTT +AACGTC +AACGTG +AACGCA +AACGCT +AACGCC +AACGCG +AACGGA +AACGGT +AACGGC +AACGGG +AAGAAA +AAGAAT +AAGAAC +AAGAAG +AAGATA +AAGATT +AAGATC +AAGATG +AAGACA +AAGACT +AAGACC +AAGACG +AAGAGA +AAGAGT +AAGAGC +AAGAGG +AAGTAA +AAGTAT +AAGTAC +AAGTAG +AAGTTA +AAGTTT +AAGTTC +AAGTTG +AAGTCA +AAGTCT +AAGTCC +AAGTCG +AAGTGA +AAGTGT +AAGTGC +AAGTGG +AAGCAA +AAGCAT +AAGCAC +AAGCAG +AAGCTA +AAGCTT +AAGCTC +AAGCTG +AAGCCA +AAGCCT +AAGCCC +AAGCCG +AAGCGA +AAGCGT +AAGCGC +AAGCGG +AAGGAA +AAGGAT +AAGGAC +AAGGAG +AAGGTA +AAGGTT +AAGGTC +AAGGTG +AAGGCA +AAGGCT +AAGGCC +AAGGCG +AAGGGA +AAGGGT +AAGGGC +AAGGGG +ATAAAA +ATAAAT +ATAAAC +ATAAAG +ATAATA +ATAATT +ATAATC +ATAATG +ATAACA +ATAACT +ATAACC +ATAACG +ATAAGA +ATAAGT +ATAAGC +ATAAGG +ATATAA +ATATAT +ATATAC +ATATAG +ATATTA +ATATTT +ATATTC +ATATTG +ATATCA +ATATCT +ATATCC +ATATCG +ATATGA +ATATGT +ATATGC +ATATGG +ATACAA +ATACAT +ATACAC +ATACAG +ATACTA +ATACTT +ATACTC +ATACTG +ATACCA +ATACCT +ATACCC +ATACCG +ATACGA +ATACGT +ATACGC +ATACGG +ATAGAA +ATAGAT +ATAGAC +ATAGAG +ATAGTA +ATAGTT +ATAGTC +ATAGTG +ATAGCA +ATAGCT +ATAGCC +ATAGCG +ATAGGA +ATAGGT +ATAGGC +ATAGGG +ATTAAA +ATTAAT +ATTAAC +ATTAAG +ATTATA +ATTATT +ATTATC +ATTATG +ATTACA +ATTACT +ATTACC +ATTACG +ATTAGA +ATTAGT +ATTAGC +ATTAGG +ATTTAA +ATTTAT +ATTTAC +ATTTAG +ATTTTA +ATTTTT +ATTTTC +ATTTTG +ATTTCA +ATTTCT +ATTTCC +ATTTCG +ATTTGA +ATTTGT +ATTTGC +ATTTGG +ATTCAA +ATTCAT +ATTCAC +ATTCAG +ATTCTA +ATTCTT +ATTCTC +ATTCTG +ATTCCA +ATTCCT +ATTCCC +ATTCCG +ATTCGA +ATTCGT +ATTCGC +ATTCGG +ATTGAA +ATTGAT +ATTGAC +ATTGAG +ATTGTA +ATTGTT +ATTGTC +ATTGTG +ATTGCA +ATTGCT +ATTGCC +ATTGCG +ATTGGA +ATTGGT +ATTGGC +ATTGGG +ATCAAA +ATCAAT +ATCAAC +ATCAAG +ATCATA +ATCATT +ATCATC +ATCATG +ATCACA +ATCACT +ATCACC +ATCACG +ATCAGA +ATCAGT +ATCAGC +ATCAGG +ATCTAA +ATCTAT +ATCTAC +ATCTAG +ATCTTA +ATCTTT +ATCTTC +ATCTTG +ATCTCA +ATCTCT +ATCTCC +ATCTCG +ATCTGA +ATCTGT +ATCTGC +ATCTGG +ATCCAA +ATCCAT +ATCCAC +ATCCAG +ATCCTA +ATCCTT +ATCCTC +ATCCTG +ATCCCA +ATCCCT +ATCCCC +ATCCCG +ATCCGA +ATCCGT +ATCCGC +ATCCGG +ATCGAA +ATCGAT +ATCGAC +ATCGAG +ATCGTA +ATCGTT +ATCGTC +ATCGTG +ATCGCA +ATCGCT +ATCGCC +ATCGCG +ATCGGA +ATCGGT +ATCGGC +ATCGGG +ATGAAA +ATGAAT +ATGAAC +ATGAAG +ATGATA +ATGATT +ATGATC +ATGATG +ATGACA +ATGACT +ATGACC +ATGACG +ATGAGA +ATGAGT +ATGAGC +ATGAGG +ATGTAA +ATGTAT +ATGTAC +ATGTAG +ATGTTA +ATGTTT +ATGTTC +ATGTTG +ATGTCA +ATGTCT +ATGTCC +ATGTCG +ATGTGA +ATGTGT +ATGTGC +ATGTGG +ATGCAA +ATGCAT +ATGCAC +ATGCAG +ATGCTA +ATGCTT +ATGCTC +ATGCTG +ATGCCA +ATGCCT +ATGCCC +ATGCCG +ATGCGA +ATGCGT +ATGCGC +ATGCGG +ATGGAA +ATGGAT +ATGGAC +ATGGAG +ATGGTA +ATGGTT +ATGGTC +ATGGTG +ATGGCA +ATGGCT +ATGGCC +ATGGCG +ATGGGA +ATGGGT +ATGGGC +ATGGGG +ACAAAA +ACAAAT +ACAAAC +ACAAAG +ACAATA +ACAATT +ACAATC +ACAATG +ACAACA +ACAACT +ACAACC +ACAACG +ACAAGA +ACAAGT +ACAAGC +ACAAGG +ACATAA +ACATAT +ACATAC +ACATAG +ACATTA +ACATTT +ACATTC +ACATTG +ACATCA +ACATCT +ACATCC +ACATCG +ACATGA +ACATGT +ACATGC +ACATGG +ACACAA +ACACAT +ACACAC +ACACAG +ACACTA +ACACTT +ACACTC +ACACTG +ACACCA +ACACCT +ACACCC +ACACCG +ACACGA +ACACGT +ACACGC +ACACGG +ACAGAA +ACAGAT +ACAGAC +ACAGAG +ACAGTA +ACAGTT +ACAGTC +ACAGTG +ACAGCA +ACAGCT +ACAGCC +ACAGCG +ACAGGA +ACAGGT +ACAGGC +ACAGGG +ACTAAA +ACTAAT +ACTAAC +ACTAAG +ACTATA +ACTATT +ACTATC +ACTATG +ACTACA +ACTACT +ACTACC +ACTACG +ACTAGA +ACTAGT +ACTAGC +ACTAGG +ACTTAA +ACTTAT +ACTTAC +ACTTAG +ACTTTA +ACTTTT +ACTTTC +ACTTTG +ACTTCA +ACTTCT +ACTTCC +ACTTCG +ACTTGA +ACTTGT +ACTTGC +ACTTGG +ACTCAA +ACTCAT +ACTCAC +ACTCAG +ACTCTA +ACTCTT +ACTCTC +ACTCTG +ACTCCA +ACTCCT +ACTCCC +ACTCCG +ACTCGA +ACTCGT +ACTCGC +ACTCGG +ACTGAA +ACTGAT +ACTGAC +ACTGAG +ACTGTA +ACTGTT +ACTGTC +ACTGTG +ACTGCA +ACTGCT +ACTGCC +ACTGCG +ACTGGA +ACTGGT +ACTGGC +ACTGGG +ACCAAA +ACCAAT +ACCAAC +ACCAAG +ACCATA +ACCATT +ACCATC +ACCATG +ACCACA +ACCACT +ACCACC +ACCACG +ACCAGA +ACCAGT +ACCAGC +ACCAGG +ACCTAA +ACCTAT +ACCTAC +ACCTAG +ACCTTA +ACCTTT +ACCTTC +ACCTTG +ACCTCA +ACCTCT +ACCTCC +ACCTCG +ACCTGA +ACCTGT +ACCTGC +ACCTGG +ACCCAA +ACCCAT +ACCCAC +ACCCAG +ACCCTA +ACCCTT +ACCCTC +ACCCTG +ACCCCA +ACCCCT +ACCCCC +ACCCCG +ACCCGA +ACCCGT +ACCCGC +ACCCGG +ACCGAA +ACCGAT +ACCGAC +ACCGAG +ACCGTA +ACCGTT +ACCGTC +ACCGTG +ACCGCA +ACCGCT +ACCGCC +ACCGCG +ACCGGA +ACCGGT +ACCGGC +ACCGGG +ACGAAA +ACGAAT +ACGAAC +ACGAAG +ACGATA +ACGATT +ACGATC +ACGATG +ACGACA +ACGACT +ACGACC +ACGACG +ACGAGA +ACGAGT +ACGAGC +ACGAGG +ACGTAA +ACGTAT +ACGTAC +ACGTAG +ACGTTA +ACGTTT +ACGTTC +ACGTTG +ACGTCA +ACGTCT +ACGTCC +ACGTCG +ACGTGA +ACGTGT +ACGTGC +ACGTGG +ACGCAA +ACGCAT +ACGCAC +ACGCAG +ACGCTA +ACGCTT +ACGCTC +ACGCTG +ACGCCA +ACGCCT +ACGCCC +ACGCCG +ACGCGA +ACGCGT +ACGCGC +ACGCGG +ACGGAA +ACGGAT +ACGGAC +ACGGAG +ACGGTA +ACGGTT +ACGGTC +ACGGTG +ACGGCA +ACGGCT +ACGGCC +ACGGCG +ACGGGA +ACGGGT +ACGGGC +ACGGGG +AGAAAA +AGAAAT +AGAAAC +AGAAAG +AGAATA +AGAATT +AGAATC +AGAATG +AGAACA +AGAACT +AGAACC +AGAACG +AGAAGA +AGAAGT +AGAAGC +AGAAGG +AGATAA +AGATAT +AGATAC +AGATAG +AGATTA +AGATTT +AGATTC +AGATTG +AGATCA +AGATCT +AGATCC +AGATCG +AGATGA +AGATGT +AGATGC +AGATGG +AGACAA +AGACAT +AGACAC +AGACAG +AGACTA +AGACTT +AGACTC +AGACTG +AGACCA +AGACCT +AGACCC +AGACCG +AGACGA +AGACGT +AGACGC +AGACGG +AGAGAA +AGAGAT +AGAGAC +AGAGAG +AGAGTA +AGAGTT +AGAGTC +AGAGTG +AGAGCA +AGAGCT +AGAGCC +AGAGCG +AGAGGA +AGAGGT +AGAGGC +AGAGGG +AGTAAA +AGTAAT +AGTAAC +AGTAAG +AGTATA +AGTATT +AGTATC +AGTATG +AGTACA +AGTACT +AGTACC +AGTACG +AGTAGA +AGTAGT +AGTAGC +AGTAGG +AGTTAA +AGTTAT +AGTTAC +AGTTAG +AGTTTA +AGTTTT +AGTTTC +AGTTTG +AGTTCA +AGTTCT +AGTTCC +AGTTCG +AGTTGA +AGTTGT +AGTTGC +AGTTGG +AGTCAA +AGTCAT +AGTCAC +AGTCAG +AGTCTA +AGTCTT +AGTCTC +AGTCTG +AGTCCA +AGTCCT +AGTCCC +AGTCCG +AGTCGA +AGTCGT +AGTCGC +AGTCGG +AGTGAA +AGTGAT +AGTGAC +AGTGAG +AGTGTA +AGTGTT +AGTGTC +AGTGTG +AGTGCA +AGTGCT +AGTGCC +AGTGCG +AGTGGA +AGTGGT +AGTGGC +AGTGGG +AGCAAA +AGCAAT +AGCAAC +AGCAAG +AGCATA +AGCATT +AGCATC +AGCATG +AGCACA +AGCACT +AGCACC +AGCACG +AGCAGA +AGCAGT +AGCAGC +AGCAGG +AGCTAA +AGCTAT +AGCTAC +AGCTAG +AGCTTA +AGCTTT +AGCTTC +AGCTTG +AGCTCA +AGCTCT +AGCTCC +AGCTCG +AGCTGA +AGCTGT +AGCTGC +AGCTGG +AGCCAA +AGCCAT +AGCCAC +AGCCAG +AGCCTA +AGCCTT +AGCCTC +AGCCTG +AGCCCA +AGCCCT +AGCCCC +AGCCCG +AGCCGA +AGCCGT +AGCCGC +AGCCGG +AGCGAA +AGCGAT +AGCGAC +AGCGAG +AGCGTA +AGCGTT +AGCGTC +AGCGTG +AGCGCA +AGCGCT +AGCGCC +AGCGCG +AGCGGA +AGCGGT +AGCGGC +AGCGGG +AGGAAA +AGGAAT +AGGAAC +AGGAAG +AGGATA +AGGATT +AGGATC +AGGATG +AGGACA +AGGACT +AGGACC +AGGACG +AGGAGA +AGGAGT +AGGAGC +AGGAGG +AGGTAA +AGGTAT +AGGTAC +AGGTAG +AGGTTA +AGGTTT +AGGTTC +AGGTTG +AGGTCA +AGGTCT +AGGTCC +AGGTCG +AGGTGA +AGGTGT +AGGTGC +AGGTGG +AGGCAA +AGGCAT +AGGCAC +AGGCAG +AGGCTA +AGGCTT +AGGCTC +AGGCTG +AGGCCA +AGGCCT +AGGCCC +AGGCCG +AGGCGA +AGGCGT +AGGCGC +AGGCGG +AGGGAA +AGGGAT +AGGGAC +AGGGAG +AGGGTA +AGGGTT +AGGGTC +AGGGTG +AGGGCA +AGGGCT +AGGGCC +AGGGCG +AGGGGA +AGGGGT +AGGGGC +AGGGGG +TAAAAA +TAAAAT +TAAAAC +TAAAAG +TAAATA +TAAATT +TAAATC +TAAATG +TAAACA +TAAACT +TAAACC +TAAACG +TAAAGA +TAAAGT +TAAAGC +TAAAGG +TAATAA +TAATAT +TAATAC +TAATAG +TAATTA +TAATTT +TAATTC +TAATTG +TAATCA +TAATCT +TAATCC +TAATCG +TAATGA +TAATGT +TAATGC +TAATGG +TAACAA +TAACAT +TAACAC +TAACAG +TAACTA +TAACTT +TAACTC +TAACTG +TAACCA +TAACCT +TAACCC +TAACCG +TAACGA +TAACGT +TAACGC +TAACGG +TAAGAA +TAAGAT +TAAGAC +TAAGAG +TAAGTA +TAAGTT +TAAGTC +TAAGTG +TAAGCA +TAAGCT +TAAGCC +TAAGCG +TAAGGA +TAAGGT +TAAGGC +TAAGGG +TATAAA +TATAAT +TATAAC +TATAAG +TATATA +TATATT +TATATC +TATATG +TATACA +TATACT +TATACC +TATACG +TATAGA +TATAGT +TATAGC +TATAGG +TATTAA +TATTAT +TATTAC +TATTAG +TATTTA +TATTTT +TATTTC +TATTTG +TATTCA +TATTCT +TATTCC +TATTCG +TATTGA +TATTGT +TATTGC +TATTGG +TATCAA +TATCAT +TATCAC +TATCAG +TATCTA +TATCTT +TATCTC +TATCTG +TATCCA +TATCCT +TATCCC +TATCCG +TATCGA +TATCGT +TATCGC +TATCGG +TATGAA +TATGAT +TATGAC +TATGAG +TATGTA +TATGTT +TATGTC +TATGTG +TATGCA +TATGCT +TATGCC +TATGCG +TATGGA +TATGGT +TATGGC +TATGGG +TACAAA +TACAAT +TACAAC +TACAAG +TACATA +TACATT +TACATC +TACATG +TACACA +TACACT +TACACC +TACACG +TACAGA +TACAGT +TACAGC +TACAGG +TACTAA +TACTAT +TACTAC +TACTAG +TACTTA +TACTTT +TACTTC +TACTTG +TACTCA +TACTCT +TACTCC +TACTCG +TACTGA +TACTGT +TACTGC +TACTGG +TACCAA +TACCAT +TACCAC +TACCAG +TACCTA +TACCTT +TACCTC +TACCTG +TACCCA +TACCCT +TACCCC +TACCCG +TACCGA +TACCGT +TACCGC +TACCGG +TACGAA +TACGAT +TACGAC +TACGAG +TACGTA +TACGTT +TACGTC +TACGTG +TACGCA +TACGCT +TACGCC +TACGCG +TACGGA +TACGGT +TACGGC +TACGGG +TAGAAA +TAGAAT +TAGAAC +TAGAAG +TAGATA +TAGATT +TAGATC +TAGATG +TAGACA +TAGACT +TAGACC +TAGACG +TAGAGA +TAGAGT +TAGAGC +TAGAGG +TAGTAA +TAGTAT +TAGTAC +TAGTAG +TAGTTA +TAGTTT +TAGTTC +TAGTTG +TAGTCA +TAGTCT +TAGTCC +TAGTCG +TAGTGA +TAGTGT +TAGTGC +TAGTGG +TAGCAA +TAGCAT +TAGCAC +TAGCAG +TAGCTA +TAGCTT +TAGCTC +TAGCTG +TAGCCA +TAGCCT +TAGCCC +TAGCCG +TAGCGA +TAGCGT +TAGCGC +TAGCGG +TAGGAA +TAGGAT +TAGGAC +TAGGAG +TAGGTA +TAGGTT +TAGGTC +TAGGTG +TAGGCA +TAGGCT +TAGGCC +TAGGCG +TAGGGA +TAGGGT +TAGGGC +TAGGGG +TTAAAA +TTAAAT +TTAAAC +TTAAAG +TTAATA +TTAATT +TTAATC +TTAATG +TTAACA +TTAACT +TTAACC +TTAACG +TTAAGA +TTAAGT +TTAAGC +TTAAGG +TTATAA +TTATAT +TTATAC +TTATAG +TTATTA +TTATTT +TTATTC +TTATTG +TTATCA +TTATCT +TTATCC +TTATCG +TTATGA +TTATGT +TTATGC +TTATGG +TTACAA +TTACAT +TTACAC +TTACAG +TTACTA +TTACTT +TTACTC +TTACTG +TTACCA +TTACCT +TTACCC +TTACCG +TTACGA +TTACGT +TTACGC +TTACGG +TTAGAA +TTAGAT +TTAGAC +TTAGAG +TTAGTA +TTAGTT +TTAGTC +TTAGTG +TTAGCA +TTAGCT +TTAGCC +TTAGCG +TTAGGA +TTAGGT +TTAGGC +TTAGGG +TTTAAA +TTTAAT +TTTAAC +TTTAAG +TTTATA +TTTATT +TTTATC +TTTATG +TTTACA +TTTACT +TTTACC +TTTACG +TTTAGA +TTTAGT +TTTAGC +TTTAGG +TTTTAA +TTTTAT +TTTTAC +TTTTAG +TTTTTA +TTTTTT +TTTTTC +TTTTTG +TTTTCA +TTTTCT +TTTTCC +TTTTCG +TTTTGA +TTTTGT +TTTTGC +TTTTGG +TTTCAA +TTTCAT +TTTCAC +TTTCAG +TTTCTA +TTTCTT +TTTCTC +TTTCTG +TTTCCA +TTTCCT +TTTCCC +TTTCCG +TTTCGA +TTTCGT +TTTCGC +TTTCGG +TTTGAA +TTTGAT +TTTGAC +TTTGAG +TTTGTA +TTTGTT +TTTGTC +TTTGTG +TTTGCA +TTTGCT +TTTGCC +TTTGCG +TTTGGA +TTTGGT +TTTGGC +TTTGGG +TTCAAA +TTCAAT +TTCAAC +TTCAAG +TTCATA +TTCATT +TTCATC +TTCATG +TTCACA +TTCACT +TTCACC +TTCACG +TTCAGA +TTCAGT +TTCAGC +TTCAGG +TTCTAA +TTCTAT +TTCTAC +TTCTAG +TTCTTA +TTCTTT +TTCTTC +TTCTTG +TTCTCA +TTCTCT +TTCTCC +TTCTCG +TTCTGA +TTCTGT +TTCTGC +TTCTGG +TTCCAA +TTCCAT +TTCCAC +TTCCAG +TTCCTA +TTCCTT +TTCCTC +TTCCTG +TTCCCA +TTCCCT +TTCCCC +TTCCCG +TTCCGA +TTCCGT +TTCCGC +TTCCGG +TTCGAA +TTCGAT +TTCGAC +TTCGAG +TTCGTA +TTCGTT +TTCGTC +TTCGTG +TTCGCA +TTCGCT +TTCGCC +TTCGCG +TTCGGA +TTCGGT +TTCGGC +TTCGGG +TTGAAA +TTGAAT +TTGAAC +TTGAAG +TTGATA +TTGATT +TTGATC +TTGATG +TTGACA +TTGACT +TTGACC +TTGACG +TTGAGA +TTGAGT +TTGAGC +TTGAGG +TTGTAA +TTGTAT +TTGTAC +TTGTAG +TTGTTA +TTGTTT +TTGTTC +TTGTTG +TTGTCA +TTGTCT +TTGTCC +TTGTCG +TTGTGA +TTGTGT +TTGTGC +TTGTGG +TTGCAA +TTGCAT +TTGCAC +TTGCAG +TTGCTA +TTGCTT +TTGCTC +TTGCTG +TTGCCA +TTGCCT +TTGCCC +TTGCCG +TTGCGA +TTGCGT +TTGCGC +TTGCGG +TTGGAA +TTGGAT +TTGGAC +TTGGAG +TTGGTA +TTGGTT +TTGGTC +TTGGTG +TTGGCA +TTGGCT +TTGGCC +TTGGCG +TTGGGA +TTGGGT +TTGGGC +TTGGGG +TCAAAA +TCAAAT +TCAAAC +TCAAAG +TCAATA +TCAATT +TCAATC +TCAATG +TCAACA +TCAACT +TCAACC +TCAACG +TCAAGA +TCAAGT +TCAAGC +TCAAGG +TCATAA +TCATAT +TCATAC +TCATAG +TCATTA +TCATTT +TCATTC +TCATTG +TCATCA +TCATCT +TCATCC +TCATCG +TCATGA +TCATGT +TCATGC +TCATGG +TCACAA +TCACAT +TCACAC +TCACAG +TCACTA +TCACTT +TCACTC +TCACTG +TCACCA +TCACCT +TCACCC +TCACCG +TCACGA +TCACGT +TCACGC +TCACGG +TCAGAA +TCAGAT +TCAGAC +TCAGAG +TCAGTA +TCAGTT +TCAGTC +TCAGTG +TCAGCA +TCAGCT +TCAGCC +TCAGCG +TCAGGA +TCAGGT +TCAGGC +TCAGGG +TCTAAA +TCTAAT +TCTAAC +TCTAAG +TCTATA +TCTATT +TCTATC +TCTATG +TCTACA +TCTACT +TCTACC +TCTACG +TCTAGA +TCTAGT +TCTAGC +TCTAGG +TCTTAA +TCTTAT +TCTTAC +TCTTAG +TCTTTA +TCTTTT +TCTTTC +TCTTTG +TCTTCA +TCTTCT +TCTTCC +TCTTCG +TCTTGA +TCTTGT +TCTTGC +TCTTGG +TCTCAA +TCTCAT +TCTCAC +TCTCAG +TCTCTA +TCTCTT +TCTCTC +TCTCTG +TCTCCA +TCTCCT +TCTCCC +TCTCCG +TCTCGA +TCTCGT +TCTCGC +TCTCGG +TCTGAA +TCTGAT +TCTGAC +TCTGAG +TCTGTA +TCTGTT +TCTGTC +TCTGTG +TCTGCA +TCTGCT +TCTGCC +TCTGCG +TCTGGA +TCTGGT +TCTGGC +TCTGGG +TCCAAA +TCCAAT +TCCAAC +TCCAAG +TCCATA +TCCATT +TCCATC +TCCATG +TCCACA +TCCACT +TCCACC +TCCACG +TCCAGA +TCCAGT +TCCAGC +TCCAGG +TCCTAA +TCCTAT +TCCTAC +TCCTAG +TCCTTA +TCCTTT +TCCTTC +TCCTTG +TCCTCA +TCCTCT +TCCTCC +TCCTCG +TCCTGA +TCCTGT +TCCTGC +TCCTGG +TCCCAA +TCCCAT +TCCCAC +TCCCAG +TCCCTA +TCCCTT +TCCCTC +TCCCTG +TCCCCA +TCCCCT +TCCCCC +TCCCCG +TCCCGA +TCCCGT +TCCCGC +TCCCGG +TCCGAA +TCCGAT +TCCGAC +TCCGAG +TCCGTA +TCCGTT +TCCGTC +TCCGTG +TCCGCA +TCCGCT +TCCGCC +TCCGCG +TCCGGA +TCCGGT +TCCGGC +TCCGGG +TCGAAA +TCGAAT +TCGAAC +TCGAAG +TCGATA +TCGATT +TCGATC +TCGATG +TCGACA +TCGACT +TCGACC +TCGACG +TCGAGA +TCGAGT +TCGAGC +TCGAGG +TCGTAA +TCGTAT +TCGTAC +TCGTAG +TCGTTA +TCGTTT +TCGTTC +TCGTTG +TCGTCA +TCGTCT +TCGTCC +TCGTCG +TCGTGA +TCGTGT +TCGTGC +TCGTGG +TCGCAA +TCGCAT +TCGCAC +TCGCAG +TCGCTA +TCGCTT +TCGCTC +TCGCTG +TCGCCA +TCGCCT +TCGCCC +TCGCCG +TCGCGA +TCGCGT +TCGCGC +TCGCGG +TCGGAA +TCGGAT +TCGGAC +TCGGAG +TCGGTA +TCGGTT +TCGGTC +TCGGTG +TCGGCA +TCGGCT +TCGGCC +TCGGCG +TCGGGA +TCGGGT +TCGGGC +TCGGGG +TGAAAA +TGAAAT +TGAAAC +TGAAAG +TGAATA +TGAATT +TGAATC +TGAATG +TGAACA +TGAACT +TGAACC +TGAACG +TGAAGA +TGAAGT +TGAAGC +TGAAGG +TGATAA +TGATAT +TGATAC +TGATAG +TGATTA +TGATTT +TGATTC +TGATTG +TGATCA +TGATCT +TGATCC +TGATCG +TGATGA +TGATGT +TGATGC +TGATGG +TGACAA +TGACAT +TGACAC +TGACAG +TGACTA +TGACTT +TGACTC +TGACTG +TGACCA +TGACCT +TGACCC +TGACCG +TGACGA +TGACGT +TGACGC +TGACGG +TGAGAA +TGAGAT +TGAGAC +TGAGAG +TGAGTA +TGAGTT +TGAGTC +TGAGTG +TGAGCA +TGAGCT +TGAGCC +TGAGCG +TGAGGA +TGAGGT +TGAGGC +TGAGGG +TGTAAA +TGTAAT +TGTAAC +TGTAAG +TGTATA +TGTATT +TGTATC +TGTATG +TGTACA +TGTACT +TGTACC +TGTACG +TGTAGA +TGTAGT +TGTAGC +TGTAGG +TGTTAA +TGTTAT +TGTTAC +TGTTAG +TGTTTA +TGTTTT +TGTTTC +TGTTTG +TGTTCA +TGTTCT +TGTTCC +TGTTCG +TGTTGA +TGTTGT +TGTTGC +TGTTGG +TGTCAA +TGTCAT +TGTCAC +TGTCAG +TGTCTA +TGTCTT +TGTCTC +TGTCTG +TGTCCA +TGTCCT +TGTCCC +TGTCCG +TGTCGA +TGTCGT +TGTCGC +TGTCGG +TGTGAA +TGTGAT +TGTGAC +TGTGAG +TGTGTA +TGTGTT +TGTGTC +TGTGTG +TGTGCA +TGTGCT +TGTGCC +TGTGCG +TGTGGA +TGTGGT +TGTGGC +TGTGGG +TGCAAA +TGCAAT +TGCAAC +TGCAAG +TGCATA +TGCATT +TGCATC +TGCATG +TGCACA +TGCACT +TGCACC +TGCACG +TGCAGA +TGCAGT +TGCAGC +TGCAGG +TGCTAA +TGCTAT +TGCTAC +TGCTAG +TGCTTA +TGCTTT +TGCTTC +TGCTTG +TGCTCA +TGCTCT +TGCTCC +TGCTCG +TGCTGA +TGCTGT +TGCTGC +TGCTGG +TGCCAA +TGCCAT +TGCCAC +TGCCAG +TGCCTA +TGCCTT +TGCCTC +TGCCTG +TGCCCA +TGCCCT +TGCCCC +TGCCCG +TGCCGA +TGCCGT +TGCCGC +TGCCGG +TGCGAA +TGCGAT +TGCGAC +TGCGAG +TGCGTA +TGCGTT +TGCGTC +TGCGTG +TGCGCA +TGCGCT +TGCGCC +TGCGCG +TGCGGA +TGCGGT +TGCGGC +TGCGGG +TGGAAA +TGGAAT +TGGAAC +TGGAAG +TGGATA +TGGATT +TGGATC +TGGATG +TGGACA +TGGACT +TGGACC +TGGACG +TGGAGA +TGGAGT +TGGAGC +TGGAGG +TGGTAA +TGGTAT +TGGTAC +TGGTAG +TGGTTA +TGGTTT +TGGTTC +TGGTTG +TGGTCA +TGGTCT +TGGTCC +TGGTCG +TGGTGA +TGGTGT +TGGTGC +TGGTGG +TGGCAA +TGGCAT +TGGCAC +TGGCAG +TGGCTA +TGGCTT +TGGCTC +TGGCTG +TGGCCA +TGGCCT +TGGCCC +TGGCCG +TGGCGA +TGGCGT +TGGCGC +TGGCGG +TGGGAA +TGGGAT +TGGGAC +TGGGAG +TGGGTA +TGGGTT +TGGGTC +TGGGTG +TGGGCA +TGGGCT +TGGGCC +TGGGCG +TGGGGA +TGGGGT +TGGGGC +TGGGGG +CAAAAA +CAAAAT +CAAAAC +CAAAAG +CAAATA +CAAATT +CAAATC +CAAATG +CAAACA +CAAACT +CAAACC +CAAACG +CAAAGA +CAAAGT +CAAAGC +CAAAGG +CAATAA +CAATAT +CAATAC +CAATAG +CAATTA +CAATTT +CAATTC +CAATTG +CAATCA +CAATCT +CAATCC +CAATCG +CAATGA +CAATGT +CAATGC +CAATGG +CAACAA +CAACAT +CAACAC +CAACAG +CAACTA +CAACTT +CAACTC +CAACTG +CAACCA +CAACCT +CAACCC +CAACCG +CAACGA +CAACGT +CAACGC +CAACGG +CAAGAA +CAAGAT +CAAGAC +CAAGAG +CAAGTA +CAAGTT +CAAGTC +CAAGTG +CAAGCA +CAAGCT +CAAGCC +CAAGCG +CAAGGA +CAAGGT +CAAGGC +CAAGGG +CATAAA +CATAAT +CATAAC +CATAAG +CATATA +CATATT +CATATC +CATATG +CATACA +CATACT +CATACC +CATACG +CATAGA +CATAGT +CATAGC +CATAGG +CATTAA +CATTAT +CATTAC +CATTAG +CATTTA +CATTTT +CATTTC +CATTTG +CATTCA +CATTCT +CATTCC +CATTCG +CATTGA +CATTGT +CATTGC +CATTGG +CATCAA +CATCAT +CATCAC +CATCAG +CATCTA +CATCTT +CATCTC +CATCTG +CATCCA +CATCCT +CATCCC +CATCCG +CATCGA +CATCGT +CATCGC +CATCGG +CATGAA +CATGAT +CATGAC +CATGAG +CATGTA +CATGTT +CATGTC +CATGTG +CATGCA +CATGCT +CATGCC +CATGCG +CATGGA +CATGGT +CATGGC +CATGGG +CACAAA +CACAAT +CACAAC +CACAAG +CACATA +CACATT +CACATC +CACATG +CACACA +CACACT +CACACC +CACACG +CACAGA +CACAGT +CACAGC +CACAGG +CACTAA +CACTAT +CACTAC +CACTAG +CACTTA +CACTTT +CACTTC +CACTTG +CACTCA +CACTCT +CACTCC +CACTCG +CACTGA +CACTGT +CACTGC +CACTGG +CACCAA +CACCAT +CACCAC +CACCAG +CACCTA +CACCTT +CACCTC +CACCTG +CACCCA +CACCCT +CACCCC +CACCCG +CACCGA +CACCGT +CACCGC +CACCGG +CACGAA +CACGAT +CACGAC +CACGAG +CACGTA +CACGTT +CACGTC +CACGTG +CACGCA +CACGCT +CACGCC +CACGCG +CACGGA +CACGGT +CACGGC +CACGGG +CAGAAA +CAGAAT +CAGAAC +CAGAAG +CAGATA +CAGATT +CAGATC +CAGATG +CAGACA +CAGACT +CAGACC +CAGACG +CAGAGA +CAGAGT +CAGAGC +CAGAGG +CAGTAA +CAGTAT +CAGTAC +CAGTAG +CAGTTA +CAGTTT +CAGTTC +CAGTTG +CAGTCA +CAGTCT +CAGTCC +CAGTCG +CAGTGA +CAGTGT +CAGTGC +CAGTGG +CAGCAA +CAGCAT +CAGCAC +CAGCAG +CAGCTA +CAGCTT +CAGCTC +CAGCTG +CAGCCA +CAGCCT +CAGCCC +CAGCCG +CAGCGA +CAGCGT +CAGCGC +CAGCGG +CAGGAA +CAGGAT +CAGGAC +CAGGAG +CAGGTA +CAGGTT +CAGGTC +CAGGTG +CAGGCA +CAGGCT +CAGGCC +CAGGCG +CAGGGA +CAGGGT +CAGGGC +CAGGGG +CTAAAA +CTAAAT +CTAAAC +CTAAAG +CTAATA +CTAATT +CTAATC +CTAATG +CTAACA +CTAACT +CTAACC +CTAACG +CTAAGA +CTAAGT +CTAAGC +CTAAGG +CTATAA +CTATAT +CTATAC +CTATAG +CTATTA +CTATTT +CTATTC +CTATTG +CTATCA +CTATCT +CTATCC +CTATCG +CTATGA +CTATGT +CTATGC +CTATGG +CTACAA +CTACAT +CTACAC +CTACAG +CTACTA +CTACTT +CTACTC +CTACTG +CTACCA +CTACCT +CTACCC +CTACCG +CTACGA +CTACGT +CTACGC +CTACGG +CTAGAA +CTAGAT +CTAGAC +CTAGAG +CTAGTA +CTAGTT +CTAGTC +CTAGTG +CTAGCA +CTAGCT +CTAGCC +CTAGCG +CTAGGA +CTAGGT +CTAGGC +CTAGGG +CTTAAA +CTTAAT +CTTAAC +CTTAAG +CTTATA +CTTATT +CTTATC +CTTATG +CTTACA +CTTACT +CTTACC +CTTACG +CTTAGA +CTTAGT +CTTAGC +CTTAGG +CTTTAA +CTTTAT +CTTTAC +CTTTAG +CTTTTA +CTTTTT +CTTTTC +CTTTTG +CTTTCA +CTTTCT +CTTTCC +CTTTCG +CTTTGA +CTTTGT +CTTTGC +CTTTGG +CTTCAA +CTTCAT +CTTCAC +CTTCAG +CTTCTA +CTTCTT +CTTCTC +CTTCTG +CTTCCA +CTTCCT +CTTCCC +CTTCCG +CTTCGA +CTTCGT +CTTCGC +CTTCGG +CTTGAA +CTTGAT +CTTGAC +CTTGAG +CTTGTA +CTTGTT +CTTGTC +CTTGTG +CTTGCA +CTTGCT +CTTGCC +CTTGCG +CTTGGA +CTTGGT +CTTGGC +CTTGGG +CTCAAA +CTCAAT +CTCAAC +CTCAAG +CTCATA +CTCATT +CTCATC +CTCATG +CTCACA +CTCACT +CTCACC +CTCACG +CTCAGA +CTCAGT +CTCAGC +CTCAGG +CTCTAA +CTCTAT +CTCTAC +CTCTAG +CTCTTA +CTCTTT +CTCTTC +CTCTTG +CTCTCA +CTCTCT +CTCTCC +CTCTCG +CTCTGA +CTCTGT +CTCTGC +CTCTGG +CTCCAA +CTCCAT +CTCCAC +CTCCAG +CTCCTA +CTCCTT +CTCCTC +CTCCTG +CTCCCA +CTCCCT +CTCCCC +CTCCCG +CTCCGA +CTCCGT +CTCCGC +CTCCGG +CTCGAA +CTCGAT +CTCGAC +CTCGAG +CTCGTA +CTCGTT +CTCGTC +CTCGTG +CTCGCA +CTCGCT +CTCGCC +CTCGCG +CTCGGA +CTCGGT +CTCGGC +CTCGGG +CTGAAA +CTGAAT +CTGAAC +CTGAAG +CTGATA +CTGATT +CTGATC +CTGATG +CTGACA +CTGACT +CTGACC +CTGACG +CTGAGA +CTGAGT +CTGAGC +CTGAGG +CTGTAA +CTGTAT +CTGTAC +CTGTAG +CTGTTA +CTGTTT +CTGTTC +CTGTTG +CTGTCA +CTGTCT +CTGTCC +CTGTCG +CTGTGA +CTGTGT +CTGTGC +CTGTGG +CTGCAA +CTGCAT +CTGCAC +CTGCAG +CTGCTA +CTGCTT +CTGCTC +CTGCTG +CTGCCA +CTGCCT +CTGCCC +CTGCCG +CTGCGA +CTGCGT +CTGCGC +CTGCGG +CTGGAA +CTGGAT +CTGGAC +CTGGAG +CTGGTA +CTGGTT +CTGGTC +CTGGTG +CTGGCA +CTGGCT +CTGGCC +CTGGCG +CTGGGA +CTGGGT +CTGGGC +CTGGGG +CCAAAA +CCAAAT +CCAAAC +CCAAAG +CCAATA +CCAATT +CCAATC +CCAATG +CCAACA +CCAACT +CCAACC +CCAACG +CCAAGA +CCAAGT +CCAAGC +CCAAGG +CCATAA +CCATAT +CCATAC +CCATAG +CCATTA +CCATTT +CCATTC +CCATTG +CCATCA +CCATCT +CCATCC +CCATCG +CCATGA +CCATGT +CCATGC +CCATGG +CCACAA +CCACAT +CCACAC +CCACAG +CCACTA +CCACTT +CCACTC +CCACTG +CCACCA +CCACCT +CCACCC +CCACCG +CCACGA +CCACGT +CCACGC +CCACGG +CCAGAA +CCAGAT +CCAGAC +CCAGAG +CCAGTA +CCAGTT +CCAGTC +CCAGTG +CCAGCA +CCAGCT +CCAGCC +CCAGCG +CCAGGA +CCAGGT +CCAGGC +CCAGGG +CCTAAA +CCTAAT +CCTAAC +CCTAAG +CCTATA +CCTATT +CCTATC +CCTATG +CCTACA +CCTACT +CCTACC +CCTACG +CCTAGA +CCTAGT +CCTAGC +CCTAGG +CCTTAA +CCTTAT +CCTTAC +CCTTAG +CCTTTA +CCTTTT +CCTTTC +CCTTTG +CCTTCA +CCTTCT +CCTTCC +CCTTCG +CCTTGA +CCTTGT +CCTTGC +CCTTGG +CCTCAA +CCTCAT +CCTCAC +CCTCAG +CCTCTA +CCTCTT +CCTCTC +CCTCTG +CCTCCA +CCTCCT +CCTCCC +CCTCCG +CCTCGA +CCTCGT +CCTCGC +CCTCGG +CCTGAA +CCTGAT +CCTGAC +CCTGAG +CCTGTA +CCTGTT +CCTGTC +CCTGTG +CCTGCA +CCTGCT +CCTGCC +CCTGCG +CCTGGA +CCTGGT +CCTGGC +CCTGGG +CCCAAA +CCCAAT +CCCAAC +CCCAAG +CCCATA +CCCATT +CCCATC +CCCATG +CCCACA +CCCACT +CCCACC +CCCACG +CCCAGA +CCCAGT +CCCAGC +CCCAGG +CCCTAA +CCCTAT +CCCTAC +CCCTAG +CCCTTA +CCCTTT +CCCTTC +CCCTTG +CCCTCA +CCCTCT +CCCTCC +CCCTCG +CCCTGA +CCCTGT +CCCTGC +CCCTGG +CCCCAA +CCCCAT +CCCCAC +CCCCAG +CCCCTA +CCCCTT +CCCCTC +CCCCTG +CCCCCA +CCCCCT +CCCCCC +CCCCCG +CCCCGA +CCCCGT +CCCCGC +CCCCGG +CCCGAA +CCCGAT +CCCGAC +CCCGAG +CCCGTA +CCCGTT +CCCGTC +CCCGTG +CCCGCA +CCCGCT +CCCGCC +CCCGCG +CCCGGA +CCCGGT +CCCGGC +CCCGGG +CCGAAA +CCGAAT +CCGAAC +CCGAAG +CCGATA +CCGATT +CCGATC +CCGATG +CCGACA +CCGACT +CCGACC +CCGACG +CCGAGA +CCGAGT +CCGAGC +CCGAGG +CCGTAA +CCGTAT +CCGTAC +CCGTAG +CCGTTA +CCGTTT +CCGTTC +CCGTTG +CCGTCA +CCGTCT +CCGTCC +CCGTCG +CCGTGA +CCGTGT +CCGTGC +CCGTGG +CCGCAA +CCGCAT +CCGCAC +CCGCAG +CCGCTA +CCGCTT +CCGCTC +CCGCTG +CCGCCA +CCGCCT +CCGCCC +CCGCCG +CCGCGA +CCGCGT +CCGCGC +CCGCGG +CCGGAA +CCGGAT +CCGGAC +CCGGAG +CCGGTA +CCGGTT +CCGGTC +CCGGTG +CCGGCA +CCGGCT +CCGGCC +CCGGCG +CCGGGA +CCGGGT +CCGGGC +CCGGGG +CGAAAA +CGAAAT +CGAAAC +CGAAAG +CGAATA +CGAATT +CGAATC +CGAATG +CGAACA +CGAACT +CGAACC +CGAACG +CGAAGA +CGAAGT +CGAAGC +CGAAGG +CGATAA +CGATAT +CGATAC +CGATAG +CGATTA +CGATTT +CGATTC +CGATTG +CGATCA +CGATCT +CGATCC +CGATCG +CGATGA +CGATGT +CGATGC +CGATGG +CGACAA +CGACAT +CGACAC +CGACAG +CGACTA +CGACTT +CGACTC +CGACTG +CGACCA +CGACCT +CGACCC +CGACCG +CGACGA +CGACGT +CGACGC +CGACGG +CGAGAA +CGAGAT +CGAGAC +CGAGAG +CGAGTA +CGAGTT +CGAGTC +CGAGTG +CGAGCA +CGAGCT +CGAGCC +CGAGCG +CGAGGA +CGAGGT +CGAGGC +CGAGGG +CGTAAA +CGTAAT +CGTAAC +CGTAAG +CGTATA +CGTATT +CGTATC +CGTATG +CGTACA +CGTACT +CGTACC +CGTACG +CGTAGA +CGTAGT +CGTAGC +CGTAGG +CGTTAA +CGTTAT +CGTTAC +CGTTAG +CGTTTA +CGTTTT +CGTTTC +CGTTTG +CGTTCA +CGTTCT +CGTTCC +CGTTCG +CGTTGA +CGTTGT +CGTTGC +CGTTGG +CGTCAA +CGTCAT +CGTCAC +CGTCAG +CGTCTA +CGTCTT +CGTCTC +CGTCTG +CGTCCA +CGTCCT +CGTCCC +CGTCCG +CGTCGA +CGTCGT +CGTCGC +CGTCGG +CGTGAA +CGTGAT +CGTGAC +CGTGAG +CGTGTA +CGTGTT +CGTGTC +CGTGTG +CGTGCA +CGTGCT +CGTGCC +CGTGCG +CGTGGA +CGTGGT +CGTGGC +CGTGGG +CGCAAA +CGCAAT +CGCAAC +CGCAAG +CGCATA +CGCATT +CGCATC +CGCATG +CGCACA +CGCACT +CGCACC +CGCACG +CGCAGA +CGCAGT +CGCAGC +CGCAGG +CGCTAA +CGCTAT +CGCTAC +CGCTAG +CGCTTA +CGCTTT +CGCTTC +CGCTTG +CGCTCA +CGCTCT +CGCTCC +CGCTCG +CGCTGA +CGCTGT +CGCTGC +CGCTGG +CGCCAA +CGCCAT +CGCCAC +CGCCAG +CGCCTA +CGCCTT +CGCCTC +CGCCTG +CGCCCA +CGCCCT +CGCCCC +CGCCCG +CGCCGA +CGCCGT +CGCCGC +CGCCGG +CGCGAA +CGCGAT +CGCGAC +CGCGAG +CGCGTA +CGCGTT +CGCGTC +CGCGTG +CGCGCA +CGCGCT +CGCGCC +CGCGCG +CGCGGA +CGCGGT +CGCGGC +CGCGGG +CGGAAA +CGGAAT +CGGAAC +CGGAAG +CGGATA +CGGATT +CGGATC +CGGATG +CGGACA +CGGACT +CGGACC +CGGACG +CGGAGA +CGGAGT +CGGAGC +CGGAGG +CGGTAA +CGGTAT +CGGTAC +CGGTAG +CGGTTA +CGGTTT +CGGTTC +CGGTTG +CGGTCA +CGGTCT +CGGTCC +CGGTCG +CGGTGA +CGGTGT +CGGTGC +CGGTGG +CGGCAA +CGGCAT +CGGCAC +CGGCAG +CGGCTA +CGGCTT +CGGCTC +CGGCTG +CGGCCA +CGGCCT +CGGCCC +CGGCCG +CGGCGA +CGGCGT +CGGCGC +CGGCGG +CGGGAA +CGGGAT +CGGGAC +CGGGAG +CGGGTA +CGGGTT +CGGGTC +CGGGTG +CGGGCA +CGGGCT +CGGGCC +CGGGCG +CGGGGA +CGGGGT +CGGGGC +CGGGGG +GAAAAA +GAAAAT +GAAAAC +GAAAAG +GAAATA +GAAATT +GAAATC +GAAATG +GAAACA +GAAACT +GAAACC +GAAACG +GAAAGA +GAAAGT +GAAAGC +GAAAGG +GAATAA +GAATAT +GAATAC +GAATAG +GAATTA +GAATTT +GAATTC +GAATTG +GAATCA +GAATCT +GAATCC +GAATCG +GAATGA +GAATGT +GAATGC +GAATGG +GAACAA +GAACAT +GAACAC +GAACAG +GAACTA +GAACTT +GAACTC +GAACTG +GAACCA +GAACCT +GAACCC +GAACCG +GAACGA +GAACGT +GAACGC +GAACGG +GAAGAA +GAAGAT +GAAGAC +GAAGAG +GAAGTA +GAAGTT +GAAGTC +GAAGTG +GAAGCA +GAAGCT +GAAGCC +GAAGCG +GAAGGA +GAAGGT +GAAGGC +GAAGGG +GATAAA +GATAAT +GATAAC +GATAAG +GATATA +GATATT +GATATC +GATATG +GATACA +GATACT +GATACC +GATACG +GATAGA +GATAGT +GATAGC +GATAGG +GATTAA +GATTAT +GATTAC +GATTAG +GATTTA +GATTTT +GATTTC +GATTTG +GATTCA +GATTCT +GATTCC +GATTCG +GATTGA +GATTGT +GATTGC +GATTGG +GATCAA +GATCAT +GATCAC +GATCAG +GATCTA +GATCTT +GATCTC +GATCTG +GATCCA +GATCCT +GATCCC +GATCCG +GATCGA +GATCGT +GATCGC +GATCGG +GATGAA +GATGAT +GATGAC +GATGAG +GATGTA +GATGTT +GATGTC +GATGTG +GATGCA +GATGCT +GATGCC +GATGCG +GATGGA +GATGGT +GATGGC +GATGGG +GACAAA +GACAAT +GACAAC +GACAAG +GACATA +GACATT +GACATC +GACATG +GACACA +GACACT +GACACC +GACACG +GACAGA +GACAGT +GACAGC +GACAGG +GACTAA +GACTAT +GACTAC +GACTAG +GACTTA +GACTTT +GACTTC +GACTTG +GACTCA +GACTCT +GACTCC +GACTCG +GACTGA +GACTGT +GACTGC +GACTGG +GACCAA +GACCAT +GACCAC +GACCAG +GACCTA +GACCTT +GACCTC +GACCTG +GACCCA +GACCCT +GACCCC +GACCCG +GACCGA +GACCGT +GACCGC +GACCGG +GACGAA +GACGAT +GACGAC +GACGAG +GACGTA +GACGTT +GACGTC +GACGTG +GACGCA +GACGCT +GACGCC +GACGCG +GACGGA +GACGGT +GACGGC +GACGGG +GAGAAA +GAGAAT +GAGAAC +GAGAAG +GAGATA +GAGATT +GAGATC +GAGATG +GAGACA +GAGACT +GAGACC +GAGACG +GAGAGA +GAGAGT +GAGAGC +GAGAGG +GAGTAA +GAGTAT +GAGTAC +GAGTAG +GAGTTA +GAGTTT +GAGTTC +GAGTTG +GAGTCA +GAGTCT +GAGTCC +GAGTCG +GAGTGA +GAGTGT +GAGTGC +GAGTGG +GAGCAA +GAGCAT +GAGCAC +GAGCAG +GAGCTA +GAGCTT +GAGCTC +GAGCTG +GAGCCA +GAGCCT +GAGCCC +GAGCCG +GAGCGA +GAGCGT +GAGCGC +GAGCGG +GAGGAA +GAGGAT +GAGGAC +GAGGAG +GAGGTA +GAGGTT +GAGGTC +GAGGTG +GAGGCA +GAGGCT +GAGGCC +GAGGCG +GAGGGA +GAGGGT +GAGGGC +GAGGGG +GTAAAA +GTAAAT +GTAAAC +GTAAAG +GTAATA +GTAATT +GTAATC +GTAATG +GTAACA +GTAACT +GTAACC +GTAACG +GTAAGA +GTAAGT +GTAAGC +GTAAGG +GTATAA +GTATAT +GTATAC +GTATAG +GTATTA +GTATTT +GTATTC +GTATTG +GTATCA +GTATCT +GTATCC +GTATCG +GTATGA +GTATGT +GTATGC +GTATGG +GTACAA +GTACAT +GTACAC +GTACAG +GTACTA +GTACTT +GTACTC +GTACTG +GTACCA +GTACCT +GTACCC +GTACCG +GTACGA +GTACGT +GTACGC +GTACGG +GTAGAA +GTAGAT +GTAGAC +GTAGAG +GTAGTA +GTAGTT +GTAGTC +GTAGTG +GTAGCA +GTAGCT +GTAGCC +GTAGCG +GTAGGA +GTAGGT +GTAGGC +GTAGGG +GTTAAA +GTTAAT +GTTAAC +GTTAAG +GTTATA +GTTATT +GTTATC +GTTATG +GTTACA +GTTACT +GTTACC +GTTACG +GTTAGA +GTTAGT +GTTAGC +GTTAGG +GTTTAA +GTTTAT +GTTTAC +GTTTAG +GTTTTA +GTTTTT +GTTTTC +GTTTTG +GTTTCA +GTTTCT +GTTTCC +GTTTCG +GTTTGA +GTTTGT +GTTTGC +GTTTGG +GTTCAA +GTTCAT +GTTCAC +GTTCAG +GTTCTA +GTTCTT +GTTCTC +GTTCTG +GTTCCA +GTTCCT +GTTCCC +GTTCCG +GTTCGA +GTTCGT +GTTCGC +GTTCGG +GTTGAA +GTTGAT +GTTGAC +GTTGAG +GTTGTA +GTTGTT +GTTGTC +GTTGTG +GTTGCA +GTTGCT +GTTGCC +GTTGCG +GTTGGA +GTTGGT +GTTGGC +GTTGGG +GTCAAA +GTCAAT +GTCAAC +GTCAAG +GTCATA +GTCATT +GTCATC +GTCATG +GTCACA +GTCACT +GTCACC +GTCACG +GTCAGA +GTCAGT +GTCAGC +GTCAGG +GTCTAA +GTCTAT +GTCTAC +GTCTAG +GTCTTA +GTCTTT +GTCTTC +GTCTTG +GTCTCA +GTCTCT +GTCTCC +GTCTCG +GTCTGA +GTCTGT +GTCTGC +GTCTGG +GTCCAA +GTCCAT +GTCCAC +GTCCAG +GTCCTA +GTCCTT +GTCCTC +GTCCTG +GTCCCA +GTCCCT +GTCCCC +GTCCCG +GTCCGA +GTCCGT +GTCCGC +GTCCGG +GTCGAA +GTCGAT +GTCGAC +GTCGAG +GTCGTA +GTCGTT +GTCGTC +GTCGTG +GTCGCA +GTCGCT +GTCGCC +GTCGCG +GTCGGA +GTCGGT +GTCGGC +GTCGGG +GTGAAA +GTGAAT +GTGAAC +GTGAAG +GTGATA +GTGATT +GTGATC +GTGATG +GTGACA +GTGACT +GTGACC +GTGACG +GTGAGA +GTGAGT +GTGAGC +GTGAGG +GTGTAA +GTGTAT +GTGTAC +GTGTAG +GTGTTA +GTGTTT +GTGTTC +GTGTTG +GTGTCA +GTGTCT +GTGTCC +GTGTCG +GTGTGA +GTGTGT +GTGTGC +GTGTGG +GTGCAA +GTGCAT +GTGCAC +GTGCAG +GTGCTA +GTGCTT +GTGCTC +GTGCTG +GTGCCA +GTGCCT +GTGCCC +GTGCCG +GTGCGA +GTGCGT +GTGCGC +GTGCGG +GTGGAA +GTGGAT +GTGGAC +GTGGAG +GTGGTA +GTGGTT +GTGGTC +GTGGTG +GTGGCA +GTGGCT +GTGGCC +GTGGCG +GTGGGA +GTGGGT +GTGGGC +GTGGGG +GCAAAA +GCAAAT +GCAAAC +GCAAAG +GCAATA +GCAATT +GCAATC +GCAATG +GCAACA +GCAACT +GCAACC +GCAACG +GCAAGA +GCAAGT +GCAAGC +GCAAGG +GCATAA +GCATAT +GCATAC +GCATAG +GCATTA +GCATTT +GCATTC +GCATTG +GCATCA +GCATCT +GCATCC +GCATCG +GCATGA +GCATGT +GCATGC +GCATGG +GCACAA +GCACAT +GCACAC +GCACAG +GCACTA +GCACTT +GCACTC +GCACTG +GCACCA +GCACCT +GCACCC +GCACCG +GCACGA +GCACGT +GCACGC +GCACGG +GCAGAA +GCAGAT +GCAGAC +GCAGAG +GCAGTA +GCAGTT +GCAGTC +GCAGTG +GCAGCA +GCAGCT +GCAGCC +GCAGCG +GCAGGA +GCAGGT +GCAGGC +GCAGGG +GCTAAA +GCTAAT +GCTAAC +GCTAAG +GCTATA +GCTATT +GCTATC +GCTATG +GCTACA +GCTACT +GCTACC +GCTACG +GCTAGA +GCTAGT +GCTAGC +GCTAGG +GCTTAA +GCTTAT +GCTTAC +GCTTAG +GCTTTA +GCTTTT +GCTTTC +GCTTTG +GCTTCA +GCTTCT +GCTTCC +GCTTCG +GCTTGA +GCTTGT +GCTTGC +GCTTGG +GCTCAA +GCTCAT +GCTCAC +GCTCAG +GCTCTA +GCTCTT +GCTCTC +GCTCTG +GCTCCA +GCTCCT +GCTCCC +GCTCCG +GCTCGA +GCTCGT +GCTCGC +GCTCGG +GCTGAA +GCTGAT +GCTGAC +GCTGAG +GCTGTA +GCTGTT +GCTGTC +GCTGTG +GCTGCA +GCTGCT +GCTGCC +GCTGCG +GCTGGA +GCTGGT +GCTGGC +GCTGGG +GCCAAA +GCCAAT +GCCAAC +GCCAAG +GCCATA +GCCATT +GCCATC +GCCATG +GCCACA +GCCACT +GCCACC +GCCACG +GCCAGA +GCCAGT +GCCAGC +GCCAGG +GCCTAA +GCCTAT +GCCTAC +GCCTAG +GCCTTA +GCCTTT +GCCTTC +GCCTTG +GCCTCA +GCCTCT +GCCTCC +GCCTCG +GCCTGA +GCCTGT +GCCTGC +GCCTGG +GCCCAA +GCCCAT +GCCCAC +GCCCAG +GCCCTA +GCCCTT +GCCCTC +GCCCTG +GCCCCA +GCCCCT +GCCCCC +GCCCCG +GCCCGA +GCCCGT +GCCCGC +GCCCGG +GCCGAA +GCCGAT +GCCGAC +GCCGAG +GCCGTA +GCCGTT +GCCGTC +GCCGTG +GCCGCA +GCCGCT +GCCGCC +GCCGCG +GCCGGA +GCCGGT +GCCGGC +GCCGGG +GCGAAA +GCGAAT +GCGAAC +GCGAAG +GCGATA +GCGATT +GCGATC +GCGATG +GCGACA +GCGACT +GCGACC +GCGACG +GCGAGA +GCGAGT +GCGAGC +GCGAGG +GCGTAA +GCGTAT +GCGTAC +GCGTAG +GCGTTA +GCGTTT +GCGTTC +GCGTTG +GCGTCA +GCGTCT +GCGTCC +GCGTCG +GCGTGA +GCGTGT +GCGTGC +GCGTGG +GCGCAA +GCGCAT +GCGCAC +GCGCAG +GCGCTA +GCGCTT +GCGCTC +GCGCTG +GCGCCA +GCGCCT +GCGCCC +GCGCCG +GCGCGA +GCGCGT +GCGCGC +GCGCGG +GCGGAA +GCGGAT +GCGGAC +GCGGAG +GCGGTA +GCGGTT +GCGGTC +GCGGTG +GCGGCA +GCGGCT +GCGGCC +GCGGCG +GCGGGA +GCGGGT +GCGGGC +GCGGGG +GGAAAA +GGAAAT +GGAAAC +GGAAAG +GGAATA +GGAATT +GGAATC +GGAATG +GGAACA +GGAACT +GGAACC +GGAACG +GGAAGA +GGAAGT +GGAAGC +GGAAGG +GGATAA +GGATAT +GGATAC +GGATAG +GGATTA +GGATTT +GGATTC +GGATTG +GGATCA +GGATCT +GGATCC +GGATCG +GGATGA +GGATGT +GGATGC +GGATGG +GGACAA +GGACAT +GGACAC +GGACAG +GGACTA +GGACTT +GGACTC +GGACTG +GGACCA +GGACCT +GGACCC +GGACCG +GGACGA +GGACGT +GGACGC +GGACGG +GGAGAA +GGAGAT +GGAGAC +GGAGAG +GGAGTA +GGAGTT +GGAGTC +GGAGTG +GGAGCA +GGAGCT +GGAGCC +GGAGCG +GGAGGA +GGAGGT +GGAGGC +GGAGGG +GGTAAA +GGTAAT +GGTAAC +GGTAAG +GGTATA +GGTATT +GGTATC +GGTATG +GGTACA +GGTACT +GGTACC +GGTACG +GGTAGA +GGTAGT +GGTAGC +GGTAGG +GGTTAA +GGTTAT +GGTTAC +GGTTAG +GGTTTA +GGTTTT +GGTTTC +GGTTTG +GGTTCA +GGTTCT +GGTTCC +GGTTCG +GGTTGA +GGTTGT +GGTTGC +GGTTGG +GGTCAA +GGTCAT +GGTCAC +GGTCAG +GGTCTA +GGTCTT +GGTCTC +GGTCTG +GGTCCA +GGTCCT +GGTCCC +GGTCCG +GGTCGA +GGTCGT +GGTCGC +GGTCGG +GGTGAA +GGTGAT +GGTGAC +GGTGAG +GGTGTA +GGTGTT +GGTGTC +GGTGTG +GGTGCA +GGTGCT +GGTGCC +GGTGCG +GGTGGA +GGTGGT +GGTGGC +GGTGGG +GGCAAA +GGCAAT +GGCAAC +GGCAAG +GGCATA +GGCATT +GGCATC +GGCATG +GGCACA +GGCACT +GGCACC +GGCACG +GGCAGA +GGCAGT +GGCAGC +GGCAGG +GGCTAA +GGCTAT +GGCTAC +GGCTAG +GGCTTA +GGCTTT +GGCTTC +GGCTTG +GGCTCA +GGCTCT +GGCTCC +GGCTCG +GGCTGA +GGCTGT +GGCTGC +GGCTGG +GGCCAA +GGCCAT +GGCCAC +GGCCAG +GGCCTA +GGCCTT +GGCCTC +GGCCTG +GGCCCA +GGCCCT +GGCCCC +GGCCCG +GGCCGA +GGCCGT +GGCCGC +GGCCGG +GGCGAA +GGCGAT +GGCGAC +GGCGAG +GGCGTA +GGCGTT +GGCGTC +GGCGTG +GGCGCA +GGCGCT +GGCGCC +GGCGCG +GGCGGA +GGCGGT +GGCGGC +GGCGGG +GGGAAA +GGGAAT +GGGAAC +GGGAAG +GGGATA +GGGATT +GGGATC +GGGATG +GGGACA +GGGACT +GGGACC +GGGACG +GGGAGA +GGGAGT +GGGAGC +GGGAGG +GGGTAA +GGGTAT +GGGTAC +GGGTAG +GGGTTA +GGGTTT +GGGTTC +GGGTTG +GGGTCA +GGGTCT +GGGTCC +GGGTCG +GGGTGA +GGGTGT +GGGTGC +GGGTGG +GGGCAA +GGGCAT +GGGCAC +GGGCAG +GGGCTA +GGGCTT +GGGCTC +GGGCTG +GGGCCA +GGGCCT +GGGCCC +GGGCCG +GGGCGA +GGGCGT +GGGCGC +GGGCGG +GGGGAA +GGGGAT +GGGGAC +GGGGAG +GGGGTA +GGGGTT +GGGGTC +GGGGTG +GGGGCA +GGGGCT +GGGGCC +GGGGCG +GGGGGA +GGGGGT +GGGGGC +GGGGGG diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/README.md b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/README.md new file mode 100644 index 0000000000000000000000000000000000000000..500fd976d92b217455f181bc98c77e612d9a18d9 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/README.md @@ -0,0 +1,202 @@ +--- +base_model: fabihamakhdoomi/TinyDNABERT +library_name: peft +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.12.0 \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/adapter_config.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..04c63e99bbc6b810449dece367c0afb165125b92 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/adapter_config.json @@ -0,0 +1,42 @@ +{ + "alpha_pattern": {}, + "auto_mapping": null, + "base_model_name_or_path": "fabihamakhdoomi/TinyDNABERT", + "bias": "none", + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 16, + "lora_dropout": 0.2, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": [ + "classifier", + "score" + ], + "peft_type": "LORA", + "r": 8, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "bert.encoder.layer.1.attention.self.query", + "bert.encoder.layer.0.output.dense", + "bert.encoder.layer.1.output.dense", + "bert.encoder.layer.0.intermediate.dense", + "bert.encoder.layer.1.intermediate.dense", + "bert.encoder.layer.0.attention.output.dense", + "bert.encoder.layer.1.attention.output.dense", + "bert.encoder.layer.0.attention.self.query", + "bert.encoder.layer.0.attention.self.value", + "bert.encoder.layer.1.attention.self.key", + "bert.encoder.layer.1.attention.self.value", + "bert.encoder.layer.0.attention.self.key" + ], + "task_type": "SEQ_CLS", + "use_dora": false, + "use_rslora": false +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/adapter_model.safetensors b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/adapter_model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..7af7dce16d645e3bfaa25908dff4ddd7099f39f1 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/adapter_model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:828399ca13a87b93a397a88d62f3f55f2112fc6849a383773ce0d8eb2ba96eb8 +size 151952 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/optimizer.pt b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..e54395a743a0e20527e87d54209143dc29305e59 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b9a70cec0135a92d17b0c5c04d6f61e534cba62cf611cfb7ad55e42e39fcf4d0 +size 312682 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/rng_state.pth b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..506b93da132609b3f461ebc7645f04db448a3371 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:77b9489769cad6c0f8fc553b5b129c47588998775afee5582baf3b7ef2be1cfa +size 14244 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/scheduler.pt b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..4c34d9dcafc0eb1507a84da1af7c2c5228be082e --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:20611d193eaf6658e6192005c679e5aaa561ac33254f734047a2032edc3e8f9b +size 1064 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/special_tokens_map.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..a8b3208c2884c4efb86e49300fdd3dc877220cdf --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/special_tokens_map.json @@ -0,0 +1,7 @@ +{ + "cls_token": "[CLS]", + "mask_token": "[MASK]", + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/tokenizer.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..ebd782de6963b9032ce9374cfe30f5275be1a30b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/tokenizer.json @@ -0,0 +1,4265 @@ +{ + "version": "1.0", + "truncation": { + "direction": "Right", + "max_length": 512, + "strategy": "LongestFirst", + "stride": 0 + }, + "padding": { + "strategy": { + "Fixed": 512 + }, + "direction": "Right", + "pad_to_multiple_of": null, + "pad_id": 0, + "pad_type_id": 0, + "pad_token": "[PAD]" + }, + "added_tokens": [ + { + "id": 0, + "content": "[PAD]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 1, + "content": "[UNK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 2, + "content": "[CLS]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 3, + "content": "[SEP]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 4, + "content": "[MASK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + } + ], + "normalizer": { + "type": "BertNormalizer", + "clean_text": true, + "handle_chinese_chars": true, + "strip_accents": null, + "lowercase": false + }, + "pre_tokenizer": { + "type": "BertPreTokenizer" + }, + "post_processor": { + "type": "TemplateProcessing", + "single": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + } + ], + "pair": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "B", + "type_id": 1 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 1 + } + } + ], + "special_tokens": { + "[CLS]": { + "id": "[CLS]", + "ids": [ + 2 + ], + "tokens": [ + "[CLS]" + ] + }, + "[SEP]": { + "id": "[SEP]", + "ids": [ + 3 + ], + "tokens": [ + "[SEP]" + ] + } + } + }, + "decoder": { + "type": "WordPiece", + "prefix": "##", + "cleanup": true + }, + "model": { + "type": "WordPiece", + "unk_token": "[UNK]", + "continuing_subword_prefix": "##", + "max_input_chars_per_word": 100, + "vocab": { + "[PAD]": 0, + "[UNK]": 1, + "[CLS]": 2, + "[SEP]": 3, + "[MASK]": 4, + "AAAAAA": 5, + "AAAAAT": 6, + "AAAAAC": 7, + "AAAAAG": 8, + "AAAATA": 9, + "AAAATT": 10, + "AAAATC": 11, + "AAAATG": 12, + "AAAACA": 13, + "AAAACT": 14, + "AAAACC": 15, + "AAAACG": 16, + "AAAAGA": 17, + "AAAAGT": 18, + "AAAAGC": 19, + "AAAAGG": 20, + "AAATAA": 21, + "AAATAT": 22, + "AAATAC": 23, + "AAATAG": 24, + "AAATTA": 25, + "AAATTT": 26, + "AAATTC": 27, + "AAATTG": 28, + "AAATCA": 29, + "AAATCT": 30, + "AAATCC": 31, + "AAATCG": 32, + "AAATGA": 33, + "AAATGT": 34, + "AAATGC": 35, + "AAATGG": 36, + "AAACAA": 37, + "AAACAT": 38, + "AAACAC": 39, + "AAACAG": 40, + "AAACTA": 41, + "AAACTT": 42, + "AAACTC": 43, + "AAACTG": 44, + "AAACCA": 45, + "AAACCT": 46, + "AAACCC": 47, + "AAACCG": 48, + "AAACGA": 49, + "AAACGT": 50, + "AAACGC": 51, + "AAACGG": 52, + "AAAGAA": 53, + "AAAGAT": 54, + "AAAGAC": 55, + "AAAGAG": 56, + "AAAGTA": 57, + "AAAGTT": 58, + "AAAGTC": 59, + "AAAGTG": 60, + "AAAGCA": 61, + "AAAGCT": 62, + "AAAGCC": 63, + "AAAGCG": 64, + "AAAGGA": 65, + "AAAGGT": 66, + "AAAGGC": 67, + "AAAGGG": 68, + "AATAAA": 69, + "AATAAT": 70, + "AATAAC": 71, + "AATAAG": 72, + "AATATA": 73, + "AATATT": 74, + "AATATC": 75, + "AATATG": 76, + "AATACA": 77, + "AATACT": 78, + "AATACC": 79, + "AATACG": 80, + "AATAGA": 81, + "AATAGT": 82, + "AATAGC": 83, + "AATAGG": 84, + "AATTAA": 85, + "AATTAT": 86, + "AATTAC": 87, + "AATTAG": 88, + "AATTTA": 89, + "AATTTT": 90, + "AATTTC": 91, + "AATTTG": 92, + "AATTCA": 93, + "AATTCT": 94, + "AATTCC": 95, + "AATTCG": 96, + "AATTGA": 97, + "AATTGT": 98, + "AATTGC": 99, + "AATTGG": 100, + "AATCAA": 101, + "AATCAT": 102, + "AATCAC": 103, + "AATCAG": 104, + "AATCTA": 105, + "AATCTT": 106, + "AATCTC": 107, + "AATCTG": 108, + "AATCCA": 109, + "AATCCT": 110, + "AATCCC": 111, + "AATCCG": 112, + "AATCGA": 113, + "AATCGT": 114, + "AATCGC": 115, + "AATCGG": 116, + "AATGAA": 117, + "AATGAT": 118, + "AATGAC": 119, + "AATGAG": 120, + "AATGTA": 121, + "AATGTT": 122, + "AATGTC": 123, + "AATGTG": 124, + "AATGCA": 125, + "AATGCT": 126, + "AATGCC": 127, + "AATGCG": 128, + "AATGGA": 129, + "AATGGT": 130, + "AATGGC": 131, + "AATGGG": 132, + "AACAAA": 133, + "AACAAT": 134, + "AACAAC": 135, + "AACAAG": 136, + "AACATA": 137, + "AACATT": 138, + "AACATC": 139, + "AACATG": 140, + "AACACA": 141, + "AACACT": 142, + "AACACC": 143, + "AACACG": 144, + "AACAGA": 145, + "AACAGT": 146, + "AACAGC": 147, + "AACAGG": 148, + "AACTAA": 149, + "AACTAT": 150, + "AACTAC": 151, + "AACTAG": 152, + "AACTTA": 153, + "AACTTT": 154, + "AACTTC": 155, + "AACTTG": 156, + "AACTCA": 157, + "AACTCT": 158, + "AACTCC": 159, + "AACTCG": 160, + "AACTGA": 161, + "AACTGT": 162, + "AACTGC": 163, + "AACTGG": 164, + "AACCAA": 165, + "AACCAT": 166, + "AACCAC": 167, + "AACCAG": 168, + "AACCTA": 169, + "AACCTT": 170, + "AACCTC": 171, + "AACCTG": 172, + "AACCCA": 173, + "AACCCT": 174, + "AACCCC": 175, + "AACCCG": 176, + "AACCGA": 177, + "AACCGT": 178, + "AACCGC": 179, + "AACCGG": 180, + "AACGAA": 181, + "AACGAT": 182, + "AACGAC": 183, + "AACGAG": 184, + "AACGTA": 185, + "AACGTT": 186, + "AACGTC": 187, + "AACGTG": 188, + "AACGCA": 189, + "AACGCT": 190, + "AACGCC": 191, + "AACGCG": 192, + "AACGGA": 193, + "AACGGT": 194, + "AACGGC": 195, + "AACGGG": 196, + "AAGAAA": 197, + "AAGAAT": 198, + "AAGAAC": 199, + "AAGAAG": 200, + "AAGATA": 201, + "AAGATT": 202, + "AAGATC": 203, + "AAGATG": 204, + "AAGACA": 205, + "AAGACT": 206, + "AAGACC": 207, + "AAGACG": 208, + "AAGAGA": 209, + "AAGAGT": 210, + "AAGAGC": 211, + "AAGAGG": 212, + "AAGTAA": 213, + "AAGTAT": 214, + "AAGTAC": 215, + "AAGTAG": 216, + "AAGTTA": 217, + "AAGTTT": 218, + "AAGTTC": 219, + "AAGTTG": 220, + "AAGTCA": 221, + "AAGTCT": 222, + "AAGTCC": 223, + "AAGTCG": 224, + "AAGTGA": 225, + "AAGTGT": 226, + "AAGTGC": 227, + "AAGTGG": 228, + "AAGCAA": 229, + "AAGCAT": 230, + "AAGCAC": 231, + "AAGCAG": 232, + "AAGCTA": 233, + "AAGCTT": 234, + "AAGCTC": 235, + "AAGCTG": 236, + "AAGCCA": 237, + "AAGCCT": 238, + "AAGCCC": 239, + "AAGCCG": 240, + "AAGCGA": 241, + "AAGCGT": 242, + "AAGCGC": 243, + "AAGCGG": 244, + "AAGGAA": 245, + "AAGGAT": 246, + "AAGGAC": 247, + "AAGGAG": 248, + "AAGGTA": 249, + "AAGGTT": 250, + "AAGGTC": 251, + "AAGGTG": 252, + "AAGGCA": 253, + "AAGGCT": 254, + "AAGGCC": 255, + "AAGGCG": 256, + "AAGGGA": 257, + "AAGGGT": 258, + "AAGGGC": 259, + "AAGGGG": 260, + "ATAAAA": 261, + "ATAAAT": 262, + "ATAAAC": 263, + "ATAAAG": 264, + "ATAATA": 265, + "ATAATT": 266, + "ATAATC": 267, + "ATAATG": 268, + "ATAACA": 269, + "ATAACT": 270, + "ATAACC": 271, + "ATAACG": 272, + "ATAAGA": 273, + "ATAAGT": 274, + "ATAAGC": 275, + "ATAAGG": 276, + "ATATAA": 277, + "ATATAT": 278, + "ATATAC": 279, + "ATATAG": 280, + "ATATTA": 281, + "ATATTT": 282, + "ATATTC": 283, + "ATATTG": 284, + "ATATCA": 285, + "ATATCT": 286, + "ATATCC": 287, + "ATATCG": 288, + "ATATGA": 289, + "ATATGT": 290, + "ATATGC": 291, + "ATATGG": 292, + "ATACAA": 293, + "ATACAT": 294, + "ATACAC": 295, + "ATACAG": 296, + "ATACTA": 297, + "ATACTT": 298, + "ATACTC": 299, + "ATACTG": 300, + "ATACCA": 301, + "ATACCT": 302, + "ATACCC": 303, + "ATACCG": 304, + "ATACGA": 305, + "ATACGT": 306, + "ATACGC": 307, + "ATACGG": 308, + "ATAGAA": 309, + "ATAGAT": 310, + "ATAGAC": 311, + "ATAGAG": 312, + "ATAGTA": 313, + "ATAGTT": 314, + "ATAGTC": 315, + "ATAGTG": 316, + "ATAGCA": 317, + "ATAGCT": 318, + "ATAGCC": 319, + "ATAGCG": 320, + "ATAGGA": 321, + "ATAGGT": 322, + "ATAGGC": 323, + "ATAGGG": 324, + "ATTAAA": 325, + "ATTAAT": 326, + "ATTAAC": 327, + "ATTAAG": 328, + "ATTATA": 329, + "ATTATT": 330, + "ATTATC": 331, + "ATTATG": 332, + "ATTACA": 333, + "ATTACT": 334, + "ATTACC": 335, + "ATTACG": 336, + "ATTAGA": 337, + "ATTAGT": 338, + "ATTAGC": 339, + "ATTAGG": 340, + "ATTTAA": 341, + "ATTTAT": 342, + "ATTTAC": 343, + "ATTTAG": 344, + "ATTTTA": 345, + "ATTTTT": 346, + "ATTTTC": 347, + "ATTTTG": 348, + "ATTTCA": 349, + "ATTTCT": 350, + "ATTTCC": 351, + "ATTTCG": 352, + "ATTTGA": 353, + "ATTTGT": 354, + "ATTTGC": 355, + "ATTTGG": 356, + "ATTCAA": 357, + "ATTCAT": 358, + "ATTCAC": 359, + "ATTCAG": 360, + "ATTCTA": 361, + "ATTCTT": 362, + "ATTCTC": 363, + "ATTCTG": 364, + "ATTCCA": 365, + "ATTCCT": 366, + "ATTCCC": 367, + "ATTCCG": 368, + "ATTCGA": 369, + "ATTCGT": 370, + "ATTCGC": 371, + "ATTCGG": 372, + "ATTGAA": 373, + "ATTGAT": 374, + "ATTGAC": 375, + "ATTGAG": 376, + "ATTGTA": 377, + "ATTGTT": 378, + "ATTGTC": 379, + "ATTGTG": 380, + "ATTGCA": 381, + "ATTGCT": 382, + "ATTGCC": 383, + "ATTGCG": 384, + "ATTGGA": 385, + "ATTGGT": 386, + "ATTGGC": 387, + "ATTGGG": 388, + "ATCAAA": 389, + "ATCAAT": 390, + "ATCAAC": 391, + "ATCAAG": 392, + "ATCATA": 393, + "ATCATT": 394, + "ATCATC": 395, + "ATCATG": 396, + "ATCACA": 397, + "ATCACT": 398, + "ATCACC": 399, + "ATCACG": 400, + "ATCAGA": 401, + "ATCAGT": 402, + "ATCAGC": 403, + "ATCAGG": 404, + "ATCTAA": 405, + "ATCTAT": 406, + "ATCTAC": 407, + "ATCTAG": 408, + "ATCTTA": 409, + "ATCTTT": 410, + "ATCTTC": 411, + "ATCTTG": 412, + "ATCTCA": 413, + "ATCTCT": 414, + "ATCTCC": 415, + "ATCTCG": 416, + "ATCTGA": 417, + "ATCTGT": 418, + "ATCTGC": 419, + "ATCTGG": 420, + "ATCCAA": 421, + "ATCCAT": 422, + "ATCCAC": 423, + "ATCCAG": 424, + "ATCCTA": 425, + "ATCCTT": 426, + "ATCCTC": 427, + "ATCCTG": 428, + "ATCCCA": 429, + "ATCCCT": 430, + "ATCCCC": 431, + "ATCCCG": 432, + "ATCCGA": 433, + "ATCCGT": 434, + "ATCCGC": 435, + "ATCCGG": 436, + "ATCGAA": 437, + "ATCGAT": 438, + "ATCGAC": 439, + "ATCGAG": 440, + "ATCGTA": 441, + "ATCGTT": 442, + "ATCGTC": 443, + "ATCGTG": 444, + "ATCGCA": 445, + "ATCGCT": 446, + "ATCGCC": 447, + "ATCGCG": 448, + "ATCGGA": 449, + "ATCGGT": 450, + "ATCGGC": 451, + "ATCGGG": 452, + "ATGAAA": 453, + "ATGAAT": 454, + "ATGAAC": 455, + "ATGAAG": 456, + "ATGATA": 457, + "ATGATT": 458, + "ATGATC": 459, + "ATGATG": 460, + "ATGACA": 461, + "ATGACT": 462, + "ATGACC": 463, + "ATGACG": 464, + "ATGAGA": 465, + "ATGAGT": 466, + "ATGAGC": 467, + "ATGAGG": 468, + "ATGTAA": 469, + "ATGTAT": 470, + "ATGTAC": 471, + "ATGTAG": 472, + "ATGTTA": 473, + "ATGTTT": 474, + "ATGTTC": 475, + "ATGTTG": 476, + "ATGTCA": 477, + "ATGTCT": 478, + "ATGTCC": 479, + "ATGTCG": 480, + "ATGTGA": 481, + "ATGTGT": 482, + "ATGTGC": 483, + "ATGTGG": 484, + "ATGCAA": 485, + "ATGCAT": 486, + "ATGCAC": 487, + "ATGCAG": 488, + "ATGCTA": 489, + "ATGCTT": 490, + "ATGCTC": 491, + "ATGCTG": 492, + "ATGCCA": 493, + "ATGCCT": 494, + "ATGCCC": 495, + "ATGCCG": 496, + "ATGCGA": 497, + "ATGCGT": 498, + "ATGCGC": 499, + "ATGCGG": 500, + "ATGGAA": 501, + "ATGGAT": 502, + "ATGGAC": 503, + "ATGGAG": 504, + "ATGGTA": 505, + "ATGGTT": 506, + "ATGGTC": 507, + "ATGGTG": 508, + "ATGGCA": 509, + "ATGGCT": 510, + "ATGGCC": 511, + "ATGGCG": 512, + "ATGGGA": 513, + "ATGGGT": 514, + "ATGGGC": 515, + "ATGGGG": 516, + "ACAAAA": 517, + "ACAAAT": 518, + "ACAAAC": 519, + "ACAAAG": 520, + "ACAATA": 521, + "ACAATT": 522, + "ACAATC": 523, + "ACAATG": 524, + "ACAACA": 525, + "ACAACT": 526, + "ACAACC": 527, + "ACAACG": 528, + "ACAAGA": 529, + "ACAAGT": 530, + "ACAAGC": 531, + "ACAAGG": 532, + "ACATAA": 533, + "ACATAT": 534, + "ACATAC": 535, + "ACATAG": 536, + "ACATTA": 537, + "ACATTT": 538, + "ACATTC": 539, + "ACATTG": 540, + "ACATCA": 541, + "ACATCT": 542, + "ACATCC": 543, + "ACATCG": 544, + "ACATGA": 545, + "ACATGT": 546, + "ACATGC": 547, + "ACATGG": 548, + "ACACAA": 549, + "ACACAT": 550, + "ACACAC": 551, + "ACACAG": 552, + "ACACTA": 553, + "ACACTT": 554, + "ACACTC": 555, + "ACACTG": 556, + "ACACCA": 557, + "ACACCT": 558, + "ACACCC": 559, + "ACACCG": 560, + "ACACGA": 561, + "ACACGT": 562, + "ACACGC": 563, + "ACACGG": 564, + "ACAGAA": 565, + "ACAGAT": 566, + "ACAGAC": 567, + "ACAGAG": 568, + "ACAGTA": 569, + "ACAGTT": 570, + "ACAGTC": 571, + "ACAGTG": 572, + "ACAGCA": 573, + "ACAGCT": 574, + "ACAGCC": 575, + "ACAGCG": 576, + "ACAGGA": 577, + "ACAGGT": 578, + "ACAGGC": 579, + "ACAGGG": 580, + "ACTAAA": 581, + "ACTAAT": 582, + "ACTAAC": 583, + "ACTAAG": 584, + "ACTATA": 585, + "ACTATT": 586, + "ACTATC": 587, + "ACTATG": 588, + "ACTACA": 589, + "ACTACT": 590, + "ACTACC": 591, + "ACTACG": 592, + "ACTAGA": 593, + "ACTAGT": 594, + "ACTAGC": 595, + "ACTAGG": 596, + "ACTTAA": 597, + "ACTTAT": 598, + "ACTTAC": 599, + "ACTTAG": 600, + "ACTTTA": 601, + "ACTTTT": 602, + "ACTTTC": 603, + "ACTTTG": 604, + "ACTTCA": 605, + "ACTTCT": 606, + "ACTTCC": 607, + "ACTTCG": 608, + "ACTTGA": 609, + "ACTTGT": 610, + "ACTTGC": 611, + "ACTTGG": 612, + "ACTCAA": 613, + "ACTCAT": 614, + "ACTCAC": 615, + "ACTCAG": 616, + "ACTCTA": 617, + "ACTCTT": 618, + "ACTCTC": 619, + "ACTCTG": 620, + "ACTCCA": 621, + "ACTCCT": 622, + "ACTCCC": 623, + "ACTCCG": 624, + "ACTCGA": 625, + "ACTCGT": 626, + "ACTCGC": 627, + "ACTCGG": 628, + "ACTGAA": 629, + "ACTGAT": 630, + "ACTGAC": 631, + "ACTGAG": 632, + "ACTGTA": 633, + "ACTGTT": 634, + "ACTGTC": 635, + "ACTGTG": 636, + "ACTGCA": 637, + "ACTGCT": 638, + "ACTGCC": 639, + "ACTGCG": 640, + "ACTGGA": 641, + "ACTGGT": 642, + "ACTGGC": 643, + "ACTGGG": 644, + "ACCAAA": 645, + "ACCAAT": 646, + "ACCAAC": 647, + "ACCAAG": 648, + "ACCATA": 649, + "ACCATT": 650, + "ACCATC": 651, + "ACCATG": 652, + "ACCACA": 653, + "ACCACT": 654, + "ACCACC": 655, + "ACCACG": 656, + "ACCAGA": 657, + "ACCAGT": 658, + "ACCAGC": 659, + "ACCAGG": 660, + "ACCTAA": 661, + "ACCTAT": 662, + "ACCTAC": 663, + "ACCTAG": 664, + "ACCTTA": 665, + "ACCTTT": 666, + "ACCTTC": 667, + "ACCTTG": 668, + "ACCTCA": 669, + "ACCTCT": 670, + "ACCTCC": 671, + "ACCTCG": 672, + "ACCTGA": 673, + "ACCTGT": 674, + "ACCTGC": 675, + "ACCTGG": 676, + "ACCCAA": 677, + "ACCCAT": 678, + "ACCCAC": 679, + "ACCCAG": 680, + "ACCCTA": 681, + "ACCCTT": 682, + "ACCCTC": 683, + "ACCCTG": 684, + "ACCCCA": 685, + "ACCCCT": 686, + "ACCCCC": 687, + "ACCCCG": 688, + "ACCCGA": 689, + "ACCCGT": 690, + "ACCCGC": 691, + "ACCCGG": 692, + "ACCGAA": 693, + "ACCGAT": 694, + "ACCGAC": 695, + "ACCGAG": 696, + "ACCGTA": 697, + "ACCGTT": 698, + "ACCGTC": 699, + "ACCGTG": 700, + "ACCGCA": 701, + "ACCGCT": 702, + "ACCGCC": 703, + "ACCGCG": 704, + "ACCGGA": 705, + "ACCGGT": 706, + "ACCGGC": 707, + "ACCGGG": 708, + "ACGAAA": 709, + "ACGAAT": 710, + "ACGAAC": 711, + "ACGAAG": 712, + "ACGATA": 713, + "ACGATT": 714, + "ACGATC": 715, + "ACGATG": 716, + "ACGACA": 717, + "ACGACT": 718, + "ACGACC": 719, + "ACGACG": 720, + "ACGAGA": 721, + "ACGAGT": 722, + "ACGAGC": 723, + "ACGAGG": 724, + "ACGTAA": 725, + "ACGTAT": 726, + "ACGTAC": 727, + "ACGTAG": 728, + "ACGTTA": 729, + "ACGTTT": 730, + "ACGTTC": 731, + "ACGTTG": 732, + "ACGTCA": 733, + "ACGTCT": 734, + "ACGTCC": 735, + "ACGTCG": 736, + "ACGTGA": 737, + "ACGTGT": 738, + "ACGTGC": 739, + "ACGTGG": 740, + "ACGCAA": 741, + "ACGCAT": 742, + "ACGCAC": 743, + "ACGCAG": 744, + "ACGCTA": 745, + "ACGCTT": 746, + "ACGCTC": 747, + "ACGCTG": 748, + "ACGCCA": 749, + "ACGCCT": 750, + "ACGCCC": 751, + "ACGCCG": 752, + "ACGCGA": 753, + "ACGCGT": 754, + "ACGCGC": 755, + "ACGCGG": 756, + "ACGGAA": 757, + "ACGGAT": 758, + "ACGGAC": 759, + "ACGGAG": 760, + "ACGGTA": 761, + "ACGGTT": 762, + "ACGGTC": 763, + "ACGGTG": 764, + "ACGGCA": 765, + "ACGGCT": 766, + "ACGGCC": 767, + "ACGGCG": 768, + "ACGGGA": 769, + "ACGGGT": 770, + "ACGGGC": 771, + "ACGGGG": 772, + "AGAAAA": 773, + "AGAAAT": 774, + "AGAAAC": 775, + "AGAAAG": 776, + "AGAATA": 777, + "AGAATT": 778, + "AGAATC": 779, + "AGAATG": 780, + "AGAACA": 781, + "AGAACT": 782, + "AGAACC": 783, + "AGAACG": 784, + "AGAAGA": 785, + "AGAAGT": 786, + "AGAAGC": 787, + "AGAAGG": 788, + "AGATAA": 789, + "AGATAT": 790, + "AGATAC": 791, + "AGATAG": 792, + "AGATTA": 793, + "AGATTT": 794, + "AGATTC": 795, + "AGATTG": 796, + "AGATCA": 797, + "AGATCT": 798, + "AGATCC": 799, + "AGATCG": 800, + "AGATGA": 801, + "AGATGT": 802, + "AGATGC": 803, + "AGATGG": 804, + "AGACAA": 805, + "AGACAT": 806, + "AGACAC": 807, + "AGACAG": 808, + "AGACTA": 809, + "AGACTT": 810, + "AGACTC": 811, + "AGACTG": 812, + "AGACCA": 813, + "AGACCT": 814, + "AGACCC": 815, + "AGACCG": 816, + "AGACGA": 817, + "AGACGT": 818, + "AGACGC": 819, + "AGACGG": 820, + "AGAGAA": 821, + "AGAGAT": 822, + "AGAGAC": 823, + "AGAGAG": 824, + "AGAGTA": 825, + "AGAGTT": 826, + "AGAGTC": 827, + "AGAGTG": 828, + "AGAGCA": 829, + "AGAGCT": 830, + "AGAGCC": 831, + "AGAGCG": 832, + "AGAGGA": 833, + "AGAGGT": 834, + "AGAGGC": 835, + "AGAGGG": 836, + "AGTAAA": 837, + "AGTAAT": 838, + "AGTAAC": 839, + "AGTAAG": 840, + "AGTATA": 841, + "AGTATT": 842, + "AGTATC": 843, + "AGTATG": 844, + "AGTACA": 845, + "AGTACT": 846, + "AGTACC": 847, + "AGTACG": 848, + "AGTAGA": 849, + "AGTAGT": 850, + "AGTAGC": 851, + "AGTAGG": 852, + "AGTTAA": 853, + "AGTTAT": 854, + "AGTTAC": 855, + "AGTTAG": 856, + "AGTTTA": 857, + "AGTTTT": 858, + "AGTTTC": 859, + "AGTTTG": 860, + "AGTTCA": 861, + "AGTTCT": 862, + "AGTTCC": 863, + "AGTTCG": 864, + "AGTTGA": 865, + "AGTTGT": 866, + "AGTTGC": 867, + "AGTTGG": 868, + "AGTCAA": 869, + "AGTCAT": 870, + "AGTCAC": 871, + "AGTCAG": 872, + "AGTCTA": 873, + "AGTCTT": 874, + "AGTCTC": 875, + "AGTCTG": 876, + "AGTCCA": 877, + "AGTCCT": 878, + "AGTCCC": 879, + "AGTCCG": 880, + "AGTCGA": 881, + "AGTCGT": 882, + "AGTCGC": 883, + "AGTCGG": 884, + "AGTGAA": 885, + "AGTGAT": 886, + "AGTGAC": 887, + "AGTGAG": 888, + "AGTGTA": 889, + "AGTGTT": 890, + "AGTGTC": 891, + "AGTGTG": 892, + "AGTGCA": 893, + "AGTGCT": 894, + "AGTGCC": 895, + "AGTGCG": 896, + "AGTGGA": 897, + "AGTGGT": 898, + "AGTGGC": 899, + "AGTGGG": 900, + "AGCAAA": 901, + "AGCAAT": 902, + "AGCAAC": 903, + "AGCAAG": 904, + "AGCATA": 905, + "AGCATT": 906, + "AGCATC": 907, + "AGCATG": 908, + "AGCACA": 909, + "AGCACT": 910, + "AGCACC": 911, + "AGCACG": 912, + "AGCAGA": 913, + "AGCAGT": 914, + "AGCAGC": 915, + "AGCAGG": 916, + "AGCTAA": 917, + "AGCTAT": 918, + "AGCTAC": 919, + "AGCTAG": 920, + "AGCTTA": 921, + "AGCTTT": 922, + "AGCTTC": 923, + "AGCTTG": 924, + "AGCTCA": 925, + "AGCTCT": 926, + "AGCTCC": 927, + "AGCTCG": 928, + "AGCTGA": 929, + "AGCTGT": 930, + "AGCTGC": 931, + "AGCTGG": 932, + "AGCCAA": 933, + "AGCCAT": 934, + "AGCCAC": 935, + "AGCCAG": 936, + "AGCCTA": 937, + "AGCCTT": 938, + "AGCCTC": 939, + "AGCCTG": 940, + "AGCCCA": 941, + "AGCCCT": 942, + "AGCCCC": 943, + "AGCCCG": 944, + "AGCCGA": 945, + "AGCCGT": 946, + "AGCCGC": 947, + "AGCCGG": 948, + "AGCGAA": 949, + "AGCGAT": 950, + "AGCGAC": 951, + "AGCGAG": 952, + "AGCGTA": 953, + "AGCGTT": 954, + "AGCGTC": 955, + "AGCGTG": 956, + "AGCGCA": 957, + "AGCGCT": 958, + "AGCGCC": 959, + "AGCGCG": 960, + "AGCGGA": 961, + "AGCGGT": 962, + "AGCGGC": 963, + "AGCGGG": 964, + "AGGAAA": 965, + "AGGAAT": 966, + "AGGAAC": 967, + "AGGAAG": 968, + "AGGATA": 969, + "AGGATT": 970, + "AGGATC": 971, + "AGGATG": 972, + "AGGACA": 973, + "AGGACT": 974, + "AGGACC": 975, + "AGGACG": 976, + "AGGAGA": 977, + "AGGAGT": 978, + "AGGAGC": 979, + "AGGAGG": 980, + "AGGTAA": 981, + "AGGTAT": 982, + "AGGTAC": 983, + "AGGTAG": 984, + "AGGTTA": 985, + "AGGTTT": 986, + "AGGTTC": 987, + "AGGTTG": 988, + "AGGTCA": 989, + "AGGTCT": 990, + "AGGTCC": 991, + "AGGTCG": 992, + "AGGTGA": 993, + "AGGTGT": 994, + "AGGTGC": 995, + "AGGTGG": 996, + "AGGCAA": 997, + "AGGCAT": 998, + "AGGCAC": 999, + "AGGCAG": 1000, + "AGGCTA": 1001, + "AGGCTT": 1002, + "AGGCTC": 1003, + "AGGCTG": 1004, + "AGGCCA": 1005, + "AGGCCT": 1006, + "AGGCCC": 1007, + "AGGCCG": 1008, + "AGGCGA": 1009, + "AGGCGT": 1010, + "AGGCGC": 1011, + "AGGCGG": 1012, + "AGGGAA": 1013, + "AGGGAT": 1014, + "AGGGAC": 1015, + "AGGGAG": 1016, + "AGGGTA": 1017, + "AGGGTT": 1018, + "AGGGTC": 1019, + "AGGGTG": 1020, + "AGGGCA": 1021, + "AGGGCT": 1022, + "AGGGCC": 1023, + "AGGGCG": 1024, + "AGGGGA": 1025, + "AGGGGT": 1026, + "AGGGGC": 1027, + "AGGGGG": 1028, + "TAAAAA": 1029, + "TAAAAT": 1030, + "TAAAAC": 1031, + "TAAAAG": 1032, + "TAAATA": 1033, + "TAAATT": 1034, + "TAAATC": 1035, + "TAAATG": 1036, + "TAAACA": 1037, + "TAAACT": 1038, + "TAAACC": 1039, + "TAAACG": 1040, + "TAAAGA": 1041, + "TAAAGT": 1042, + "TAAAGC": 1043, + "TAAAGG": 1044, + "TAATAA": 1045, + "TAATAT": 1046, + "TAATAC": 1047, + "TAATAG": 1048, + "TAATTA": 1049, + "TAATTT": 1050, + "TAATTC": 1051, + "TAATTG": 1052, + "TAATCA": 1053, + "TAATCT": 1054, + "TAATCC": 1055, + "TAATCG": 1056, + "TAATGA": 1057, + "TAATGT": 1058, + "TAATGC": 1059, + "TAATGG": 1060, + "TAACAA": 1061, + "TAACAT": 1062, + "TAACAC": 1063, + "TAACAG": 1064, + "TAACTA": 1065, + "TAACTT": 1066, + "TAACTC": 1067, + "TAACTG": 1068, + "TAACCA": 1069, + "TAACCT": 1070, + "TAACCC": 1071, + "TAACCG": 1072, + "TAACGA": 1073, + "TAACGT": 1074, + "TAACGC": 1075, + "TAACGG": 1076, + "TAAGAA": 1077, + "TAAGAT": 1078, + "TAAGAC": 1079, + "TAAGAG": 1080, + "TAAGTA": 1081, + "TAAGTT": 1082, + "TAAGTC": 1083, + "TAAGTG": 1084, + "TAAGCA": 1085, + "TAAGCT": 1086, + "TAAGCC": 1087, + "TAAGCG": 1088, + "TAAGGA": 1089, + "TAAGGT": 1090, + "TAAGGC": 1091, + "TAAGGG": 1092, + "TATAAA": 1093, + "TATAAT": 1094, + "TATAAC": 1095, + "TATAAG": 1096, + "TATATA": 1097, + "TATATT": 1098, + "TATATC": 1099, + "TATATG": 1100, + "TATACA": 1101, + "TATACT": 1102, + "TATACC": 1103, + "TATACG": 1104, + "TATAGA": 1105, + "TATAGT": 1106, + "TATAGC": 1107, + "TATAGG": 1108, + "TATTAA": 1109, + "TATTAT": 1110, + "TATTAC": 1111, + "TATTAG": 1112, + "TATTTA": 1113, + "TATTTT": 1114, + "TATTTC": 1115, + "TATTTG": 1116, + "TATTCA": 1117, + "TATTCT": 1118, + "TATTCC": 1119, + "TATTCG": 1120, + "TATTGA": 1121, + "TATTGT": 1122, + "TATTGC": 1123, + "TATTGG": 1124, + "TATCAA": 1125, + "TATCAT": 1126, + "TATCAC": 1127, + "TATCAG": 1128, + "TATCTA": 1129, + "TATCTT": 1130, + "TATCTC": 1131, + "TATCTG": 1132, + "TATCCA": 1133, + "TATCCT": 1134, + "TATCCC": 1135, + "TATCCG": 1136, + "TATCGA": 1137, + "TATCGT": 1138, + "TATCGC": 1139, + "TATCGG": 1140, + "TATGAA": 1141, + "TATGAT": 1142, + "TATGAC": 1143, + "TATGAG": 1144, + "TATGTA": 1145, + "TATGTT": 1146, + "TATGTC": 1147, + "TATGTG": 1148, + "TATGCA": 1149, + "TATGCT": 1150, + "TATGCC": 1151, + "TATGCG": 1152, + "TATGGA": 1153, + "TATGGT": 1154, + "TATGGC": 1155, + "TATGGG": 1156, + "TACAAA": 1157, + "TACAAT": 1158, + "TACAAC": 1159, + "TACAAG": 1160, + "TACATA": 1161, + "TACATT": 1162, + "TACATC": 1163, + "TACATG": 1164, + "TACACA": 1165, + "TACACT": 1166, + "TACACC": 1167, + "TACACG": 1168, + "TACAGA": 1169, + "TACAGT": 1170, + "TACAGC": 1171, + "TACAGG": 1172, + "TACTAA": 1173, + "TACTAT": 1174, + "TACTAC": 1175, + "TACTAG": 1176, + "TACTTA": 1177, + "TACTTT": 1178, + "TACTTC": 1179, + "TACTTG": 1180, + "TACTCA": 1181, + "TACTCT": 1182, + "TACTCC": 1183, + "TACTCG": 1184, + "TACTGA": 1185, + "TACTGT": 1186, + "TACTGC": 1187, + "TACTGG": 1188, + "TACCAA": 1189, + "TACCAT": 1190, + "TACCAC": 1191, + "TACCAG": 1192, + "TACCTA": 1193, + "TACCTT": 1194, + "TACCTC": 1195, + "TACCTG": 1196, + "TACCCA": 1197, + "TACCCT": 1198, + "TACCCC": 1199, + "TACCCG": 1200, + "TACCGA": 1201, + "TACCGT": 1202, + "TACCGC": 1203, + "TACCGG": 1204, + "TACGAA": 1205, + "TACGAT": 1206, + "TACGAC": 1207, + "TACGAG": 1208, + "TACGTA": 1209, + "TACGTT": 1210, + "TACGTC": 1211, + "TACGTG": 1212, + "TACGCA": 1213, + "TACGCT": 1214, + "TACGCC": 1215, + "TACGCG": 1216, + "TACGGA": 1217, + "TACGGT": 1218, + "TACGGC": 1219, + "TACGGG": 1220, + "TAGAAA": 1221, + "TAGAAT": 1222, + "TAGAAC": 1223, + "TAGAAG": 1224, + "TAGATA": 1225, + "TAGATT": 1226, + "TAGATC": 1227, + "TAGATG": 1228, + "TAGACA": 1229, + "TAGACT": 1230, + "TAGACC": 1231, + "TAGACG": 1232, + "TAGAGA": 1233, + "TAGAGT": 1234, + "TAGAGC": 1235, + "TAGAGG": 1236, + "TAGTAA": 1237, + "TAGTAT": 1238, + "TAGTAC": 1239, + "TAGTAG": 1240, + "TAGTTA": 1241, + "TAGTTT": 1242, + "TAGTTC": 1243, + "TAGTTG": 1244, + "TAGTCA": 1245, + "TAGTCT": 1246, + "TAGTCC": 1247, + "TAGTCG": 1248, + "TAGTGA": 1249, + "TAGTGT": 1250, + "TAGTGC": 1251, + "TAGTGG": 1252, + "TAGCAA": 1253, + "TAGCAT": 1254, + "TAGCAC": 1255, + "TAGCAG": 1256, + "TAGCTA": 1257, + "TAGCTT": 1258, + "TAGCTC": 1259, + "TAGCTG": 1260, + "TAGCCA": 1261, + "TAGCCT": 1262, + "TAGCCC": 1263, + "TAGCCG": 1264, + "TAGCGA": 1265, + "TAGCGT": 1266, + "TAGCGC": 1267, + "TAGCGG": 1268, + "TAGGAA": 1269, + "TAGGAT": 1270, + "TAGGAC": 1271, + "TAGGAG": 1272, + "TAGGTA": 1273, + "TAGGTT": 1274, + "TAGGTC": 1275, + "TAGGTG": 1276, + "TAGGCA": 1277, + "TAGGCT": 1278, + "TAGGCC": 1279, + "TAGGCG": 1280, + "TAGGGA": 1281, + "TAGGGT": 1282, + "TAGGGC": 1283, + "TAGGGG": 1284, + "TTAAAA": 1285, + "TTAAAT": 1286, + "TTAAAC": 1287, + "TTAAAG": 1288, + "TTAATA": 1289, + "TTAATT": 1290, + "TTAATC": 1291, + "TTAATG": 1292, + "TTAACA": 1293, + "TTAACT": 1294, + "TTAACC": 1295, + "TTAACG": 1296, + "TTAAGA": 1297, + "TTAAGT": 1298, + "TTAAGC": 1299, + "TTAAGG": 1300, + "TTATAA": 1301, + "TTATAT": 1302, + "TTATAC": 1303, + "TTATAG": 1304, + "TTATTA": 1305, + "TTATTT": 1306, + "TTATTC": 1307, + "TTATTG": 1308, + "TTATCA": 1309, + "TTATCT": 1310, + "TTATCC": 1311, + "TTATCG": 1312, + "TTATGA": 1313, + "TTATGT": 1314, + "TTATGC": 1315, + "TTATGG": 1316, + "TTACAA": 1317, + "TTACAT": 1318, + "TTACAC": 1319, + "TTACAG": 1320, + "TTACTA": 1321, + "TTACTT": 1322, + "TTACTC": 1323, + "TTACTG": 1324, + "TTACCA": 1325, + "TTACCT": 1326, + "TTACCC": 1327, + "TTACCG": 1328, + "TTACGA": 1329, + "TTACGT": 1330, + "TTACGC": 1331, + "TTACGG": 1332, + "TTAGAA": 1333, + "TTAGAT": 1334, + "TTAGAC": 1335, + "TTAGAG": 1336, + "TTAGTA": 1337, + "TTAGTT": 1338, + "TTAGTC": 1339, + "TTAGTG": 1340, + "TTAGCA": 1341, + "TTAGCT": 1342, + "TTAGCC": 1343, + "TTAGCG": 1344, + "TTAGGA": 1345, + "TTAGGT": 1346, + "TTAGGC": 1347, + "TTAGGG": 1348, + "TTTAAA": 1349, + "TTTAAT": 1350, + "TTTAAC": 1351, + "TTTAAG": 1352, + "TTTATA": 1353, + "TTTATT": 1354, + "TTTATC": 1355, + "TTTATG": 1356, + "TTTACA": 1357, + "TTTACT": 1358, + "TTTACC": 1359, + "TTTACG": 1360, + "TTTAGA": 1361, + "TTTAGT": 1362, + "TTTAGC": 1363, + "TTTAGG": 1364, + "TTTTAA": 1365, + "TTTTAT": 1366, + "TTTTAC": 1367, + "TTTTAG": 1368, + "TTTTTA": 1369, + "TTTTTT": 1370, + "TTTTTC": 1371, + "TTTTTG": 1372, + "TTTTCA": 1373, + "TTTTCT": 1374, + "TTTTCC": 1375, + "TTTTCG": 1376, + "TTTTGA": 1377, + "TTTTGT": 1378, + "TTTTGC": 1379, + "TTTTGG": 1380, + "TTTCAA": 1381, + "TTTCAT": 1382, + "TTTCAC": 1383, + "TTTCAG": 1384, + "TTTCTA": 1385, + "TTTCTT": 1386, + "TTTCTC": 1387, + "TTTCTG": 1388, + "TTTCCA": 1389, + "TTTCCT": 1390, + "TTTCCC": 1391, + "TTTCCG": 1392, + "TTTCGA": 1393, + "TTTCGT": 1394, + "TTTCGC": 1395, + "TTTCGG": 1396, + "TTTGAA": 1397, + "TTTGAT": 1398, + "TTTGAC": 1399, + "TTTGAG": 1400, + "TTTGTA": 1401, + "TTTGTT": 1402, + "TTTGTC": 1403, + "TTTGTG": 1404, + "TTTGCA": 1405, + "TTTGCT": 1406, + "TTTGCC": 1407, + "TTTGCG": 1408, + "TTTGGA": 1409, + "TTTGGT": 1410, + "TTTGGC": 1411, + "TTTGGG": 1412, + "TTCAAA": 1413, + "TTCAAT": 1414, + "TTCAAC": 1415, + "TTCAAG": 1416, + "TTCATA": 1417, + "TTCATT": 1418, + "TTCATC": 1419, + "TTCATG": 1420, + "TTCACA": 1421, + "TTCACT": 1422, + "TTCACC": 1423, + "TTCACG": 1424, + "TTCAGA": 1425, + "TTCAGT": 1426, + "TTCAGC": 1427, + "TTCAGG": 1428, + "TTCTAA": 1429, + "TTCTAT": 1430, + "TTCTAC": 1431, + "TTCTAG": 1432, + "TTCTTA": 1433, + "TTCTTT": 1434, + "TTCTTC": 1435, + "TTCTTG": 1436, + "TTCTCA": 1437, + "TTCTCT": 1438, + "TTCTCC": 1439, + "TTCTCG": 1440, + "TTCTGA": 1441, + "TTCTGT": 1442, + "TTCTGC": 1443, + "TTCTGG": 1444, + "TTCCAA": 1445, + "TTCCAT": 1446, + "TTCCAC": 1447, + "TTCCAG": 1448, + "TTCCTA": 1449, + "TTCCTT": 1450, + "TTCCTC": 1451, + "TTCCTG": 1452, + "TTCCCA": 1453, + "TTCCCT": 1454, + "TTCCCC": 1455, + "TTCCCG": 1456, + "TTCCGA": 1457, + "TTCCGT": 1458, + "TTCCGC": 1459, + "TTCCGG": 1460, + "TTCGAA": 1461, + "TTCGAT": 1462, + "TTCGAC": 1463, + "TTCGAG": 1464, + "TTCGTA": 1465, + "TTCGTT": 1466, + "TTCGTC": 1467, + "TTCGTG": 1468, + "TTCGCA": 1469, + "TTCGCT": 1470, + "TTCGCC": 1471, + "TTCGCG": 1472, + "TTCGGA": 1473, + "TTCGGT": 1474, + "TTCGGC": 1475, + "TTCGGG": 1476, + "TTGAAA": 1477, + "TTGAAT": 1478, + "TTGAAC": 1479, + "TTGAAG": 1480, + "TTGATA": 1481, + "TTGATT": 1482, + "TTGATC": 1483, + "TTGATG": 1484, + "TTGACA": 1485, + "TTGACT": 1486, + "TTGACC": 1487, + "TTGACG": 1488, + "TTGAGA": 1489, + "TTGAGT": 1490, + "TTGAGC": 1491, + "TTGAGG": 1492, + "TTGTAA": 1493, + "TTGTAT": 1494, + "TTGTAC": 1495, + "TTGTAG": 1496, + "TTGTTA": 1497, + "TTGTTT": 1498, + "TTGTTC": 1499, + "TTGTTG": 1500, + "TTGTCA": 1501, + "TTGTCT": 1502, + "TTGTCC": 1503, + "TTGTCG": 1504, + "TTGTGA": 1505, + "TTGTGT": 1506, + "TTGTGC": 1507, + "TTGTGG": 1508, + "TTGCAA": 1509, + "TTGCAT": 1510, + "TTGCAC": 1511, + "TTGCAG": 1512, + "TTGCTA": 1513, + "TTGCTT": 1514, + "TTGCTC": 1515, + "TTGCTG": 1516, + "TTGCCA": 1517, + "TTGCCT": 1518, + "TTGCCC": 1519, + "TTGCCG": 1520, + "TTGCGA": 1521, + "TTGCGT": 1522, + "TTGCGC": 1523, + "TTGCGG": 1524, + "TTGGAA": 1525, + "TTGGAT": 1526, + "TTGGAC": 1527, + "TTGGAG": 1528, + "TTGGTA": 1529, + "TTGGTT": 1530, + "TTGGTC": 1531, + "TTGGTG": 1532, + "TTGGCA": 1533, + "TTGGCT": 1534, + "TTGGCC": 1535, + "TTGGCG": 1536, + "TTGGGA": 1537, + "TTGGGT": 1538, + "TTGGGC": 1539, + "TTGGGG": 1540, + "TCAAAA": 1541, + "TCAAAT": 1542, + "TCAAAC": 1543, + "TCAAAG": 1544, + "TCAATA": 1545, + "TCAATT": 1546, + "TCAATC": 1547, + "TCAATG": 1548, + "TCAACA": 1549, + "TCAACT": 1550, + "TCAACC": 1551, + "TCAACG": 1552, + "TCAAGA": 1553, + "TCAAGT": 1554, + "TCAAGC": 1555, + "TCAAGG": 1556, + "TCATAA": 1557, + "TCATAT": 1558, + "TCATAC": 1559, + "TCATAG": 1560, + "TCATTA": 1561, + "TCATTT": 1562, + "TCATTC": 1563, + "TCATTG": 1564, + "TCATCA": 1565, + "TCATCT": 1566, + "TCATCC": 1567, + "TCATCG": 1568, + "TCATGA": 1569, + "TCATGT": 1570, + "TCATGC": 1571, + "TCATGG": 1572, + "TCACAA": 1573, + "TCACAT": 1574, + "TCACAC": 1575, + "TCACAG": 1576, + "TCACTA": 1577, + "TCACTT": 1578, + "TCACTC": 1579, + "TCACTG": 1580, + "TCACCA": 1581, + "TCACCT": 1582, + "TCACCC": 1583, + "TCACCG": 1584, + "TCACGA": 1585, + "TCACGT": 1586, + "TCACGC": 1587, + "TCACGG": 1588, + "TCAGAA": 1589, + "TCAGAT": 1590, + "TCAGAC": 1591, + "TCAGAG": 1592, + "TCAGTA": 1593, + "TCAGTT": 1594, + "TCAGTC": 1595, + "TCAGTG": 1596, + "TCAGCA": 1597, + "TCAGCT": 1598, + "TCAGCC": 1599, + "TCAGCG": 1600, + "TCAGGA": 1601, + "TCAGGT": 1602, + "TCAGGC": 1603, + "TCAGGG": 1604, + "TCTAAA": 1605, + "TCTAAT": 1606, + "TCTAAC": 1607, + "TCTAAG": 1608, + "TCTATA": 1609, + "TCTATT": 1610, + "TCTATC": 1611, + "TCTATG": 1612, + "TCTACA": 1613, + "TCTACT": 1614, + "TCTACC": 1615, + "TCTACG": 1616, + "TCTAGA": 1617, + "TCTAGT": 1618, + "TCTAGC": 1619, + "TCTAGG": 1620, + "TCTTAA": 1621, + "TCTTAT": 1622, + "TCTTAC": 1623, + "TCTTAG": 1624, + "TCTTTA": 1625, + "TCTTTT": 1626, + "TCTTTC": 1627, + "TCTTTG": 1628, + "TCTTCA": 1629, + "TCTTCT": 1630, + "TCTTCC": 1631, + "TCTTCG": 1632, + "TCTTGA": 1633, + "TCTTGT": 1634, + "TCTTGC": 1635, + "TCTTGG": 1636, + "TCTCAA": 1637, + "TCTCAT": 1638, + "TCTCAC": 1639, + "TCTCAG": 1640, + "TCTCTA": 1641, + "TCTCTT": 1642, + "TCTCTC": 1643, + "TCTCTG": 1644, + "TCTCCA": 1645, + "TCTCCT": 1646, + "TCTCCC": 1647, + "TCTCCG": 1648, + "TCTCGA": 1649, + "TCTCGT": 1650, + "TCTCGC": 1651, + "TCTCGG": 1652, + "TCTGAA": 1653, + "TCTGAT": 1654, + "TCTGAC": 1655, + "TCTGAG": 1656, + "TCTGTA": 1657, + "TCTGTT": 1658, + "TCTGTC": 1659, + "TCTGTG": 1660, + "TCTGCA": 1661, + "TCTGCT": 1662, + "TCTGCC": 1663, + "TCTGCG": 1664, + "TCTGGA": 1665, + "TCTGGT": 1666, + "TCTGGC": 1667, + "TCTGGG": 1668, + "TCCAAA": 1669, + "TCCAAT": 1670, + "TCCAAC": 1671, + "TCCAAG": 1672, + "TCCATA": 1673, + "TCCATT": 1674, + "TCCATC": 1675, + "TCCATG": 1676, + "TCCACA": 1677, + "TCCACT": 1678, + "TCCACC": 1679, + "TCCACG": 1680, + "TCCAGA": 1681, + "TCCAGT": 1682, + "TCCAGC": 1683, + "TCCAGG": 1684, + "TCCTAA": 1685, + "TCCTAT": 1686, + "TCCTAC": 1687, + "TCCTAG": 1688, + "TCCTTA": 1689, + "TCCTTT": 1690, + "TCCTTC": 1691, + "TCCTTG": 1692, + "TCCTCA": 1693, + "TCCTCT": 1694, + "TCCTCC": 1695, + "TCCTCG": 1696, + "TCCTGA": 1697, + "TCCTGT": 1698, + "TCCTGC": 1699, + "TCCTGG": 1700, + "TCCCAA": 1701, + "TCCCAT": 1702, + "TCCCAC": 1703, + "TCCCAG": 1704, + "TCCCTA": 1705, + "TCCCTT": 1706, + "TCCCTC": 1707, + "TCCCTG": 1708, + "TCCCCA": 1709, + "TCCCCT": 1710, + "TCCCCC": 1711, + "TCCCCG": 1712, + "TCCCGA": 1713, + "TCCCGT": 1714, + "TCCCGC": 1715, + "TCCCGG": 1716, + "TCCGAA": 1717, + "TCCGAT": 1718, + "TCCGAC": 1719, + "TCCGAG": 1720, + "TCCGTA": 1721, + "TCCGTT": 1722, + "TCCGTC": 1723, + "TCCGTG": 1724, + "TCCGCA": 1725, + "TCCGCT": 1726, + "TCCGCC": 1727, + "TCCGCG": 1728, + "TCCGGA": 1729, + "TCCGGT": 1730, + "TCCGGC": 1731, + "TCCGGG": 1732, + "TCGAAA": 1733, + "TCGAAT": 1734, + "TCGAAC": 1735, + "TCGAAG": 1736, + "TCGATA": 1737, + "TCGATT": 1738, + "TCGATC": 1739, + "TCGATG": 1740, + "TCGACA": 1741, + "TCGACT": 1742, + "TCGACC": 1743, + "TCGACG": 1744, + "TCGAGA": 1745, + "TCGAGT": 1746, + "TCGAGC": 1747, + "TCGAGG": 1748, + "TCGTAA": 1749, + "TCGTAT": 1750, + "TCGTAC": 1751, + "TCGTAG": 1752, + "TCGTTA": 1753, + "TCGTTT": 1754, + "TCGTTC": 1755, + "TCGTTG": 1756, + "TCGTCA": 1757, + "TCGTCT": 1758, + "TCGTCC": 1759, + "TCGTCG": 1760, + "TCGTGA": 1761, + "TCGTGT": 1762, + "TCGTGC": 1763, + "TCGTGG": 1764, + "TCGCAA": 1765, + "TCGCAT": 1766, + "TCGCAC": 1767, + "TCGCAG": 1768, + "TCGCTA": 1769, + "TCGCTT": 1770, + "TCGCTC": 1771, + "TCGCTG": 1772, + "TCGCCA": 1773, + "TCGCCT": 1774, + "TCGCCC": 1775, + "TCGCCG": 1776, + "TCGCGA": 1777, + "TCGCGT": 1778, + "TCGCGC": 1779, + "TCGCGG": 1780, + "TCGGAA": 1781, + "TCGGAT": 1782, + "TCGGAC": 1783, + "TCGGAG": 1784, + "TCGGTA": 1785, + "TCGGTT": 1786, + "TCGGTC": 1787, + "TCGGTG": 1788, + "TCGGCA": 1789, + "TCGGCT": 1790, + "TCGGCC": 1791, + "TCGGCG": 1792, + "TCGGGA": 1793, + "TCGGGT": 1794, + "TCGGGC": 1795, + "TCGGGG": 1796, + "TGAAAA": 1797, + "TGAAAT": 1798, + "TGAAAC": 1799, + "TGAAAG": 1800, + "TGAATA": 1801, + "TGAATT": 1802, + "TGAATC": 1803, + "TGAATG": 1804, + "TGAACA": 1805, + "TGAACT": 1806, + "TGAACC": 1807, + "TGAACG": 1808, + "TGAAGA": 1809, + "TGAAGT": 1810, + "TGAAGC": 1811, + "TGAAGG": 1812, + "TGATAA": 1813, + "TGATAT": 1814, + "TGATAC": 1815, + "TGATAG": 1816, + "TGATTA": 1817, + "TGATTT": 1818, + "TGATTC": 1819, + "TGATTG": 1820, + "TGATCA": 1821, + "TGATCT": 1822, + "TGATCC": 1823, + "TGATCG": 1824, + "TGATGA": 1825, + "TGATGT": 1826, + "TGATGC": 1827, + "TGATGG": 1828, + "TGACAA": 1829, + "TGACAT": 1830, + "TGACAC": 1831, + "TGACAG": 1832, + "TGACTA": 1833, + "TGACTT": 1834, + "TGACTC": 1835, + "TGACTG": 1836, + "TGACCA": 1837, + "TGACCT": 1838, + "TGACCC": 1839, + "TGACCG": 1840, + "TGACGA": 1841, + "TGACGT": 1842, + "TGACGC": 1843, + "TGACGG": 1844, + "TGAGAA": 1845, + "TGAGAT": 1846, + "TGAGAC": 1847, + "TGAGAG": 1848, + "TGAGTA": 1849, + "TGAGTT": 1850, + "TGAGTC": 1851, + "TGAGTG": 1852, + "TGAGCA": 1853, + "TGAGCT": 1854, + "TGAGCC": 1855, + "TGAGCG": 1856, + "TGAGGA": 1857, + "TGAGGT": 1858, + "TGAGGC": 1859, + "TGAGGG": 1860, + "TGTAAA": 1861, + "TGTAAT": 1862, + "TGTAAC": 1863, + "TGTAAG": 1864, + "TGTATA": 1865, + "TGTATT": 1866, + "TGTATC": 1867, + "TGTATG": 1868, + "TGTACA": 1869, + "TGTACT": 1870, + "TGTACC": 1871, + "TGTACG": 1872, + "TGTAGA": 1873, + "TGTAGT": 1874, + "TGTAGC": 1875, + "TGTAGG": 1876, + "TGTTAA": 1877, + "TGTTAT": 1878, + "TGTTAC": 1879, + "TGTTAG": 1880, + "TGTTTA": 1881, + "TGTTTT": 1882, + "TGTTTC": 1883, + "TGTTTG": 1884, + "TGTTCA": 1885, + "TGTTCT": 1886, + "TGTTCC": 1887, + "TGTTCG": 1888, + "TGTTGA": 1889, + "TGTTGT": 1890, + "TGTTGC": 1891, + "TGTTGG": 1892, + "TGTCAA": 1893, + "TGTCAT": 1894, + "TGTCAC": 1895, + "TGTCAG": 1896, + "TGTCTA": 1897, + "TGTCTT": 1898, + "TGTCTC": 1899, + "TGTCTG": 1900, + "TGTCCA": 1901, + "TGTCCT": 1902, + "TGTCCC": 1903, + "TGTCCG": 1904, + "TGTCGA": 1905, + "TGTCGT": 1906, + "TGTCGC": 1907, + "TGTCGG": 1908, + "TGTGAA": 1909, + "TGTGAT": 1910, + "TGTGAC": 1911, + "TGTGAG": 1912, + "TGTGTA": 1913, + "TGTGTT": 1914, + "TGTGTC": 1915, + "TGTGTG": 1916, + "TGTGCA": 1917, + "TGTGCT": 1918, + "TGTGCC": 1919, + "TGTGCG": 1920, + "TGTGGA": 1921, + "TGTGGT": 1922, + "TGTGGC": 1923, + "TGTGGG": 1924, + "TGCAAA": 1925, + "TGCAAT": 1926, + "TGCAAC": 1927, + "TGCAAG": 1928, + "TGCATA": 1929, + "TGCATT": 1930, + "TGCATC": 1931, + "TGCATG": 1932, + "TGCACA": 1933, + "TGCACT": 1934, + "TGCACC": 1935, + "TGCACG": 1936, + "TGCAGA": 1937, + "TGCAGT": 1938, + "TGCAGC": 1939, + "TGCAGG": 1940, + "TGCTAA": 1941, + "TGCTAT": 1942, + "TGCTAC": 1943, + "TGCTAG": 1944, + "TGCTTA": 1945, + "TGCTTT": 1946, + "TGCTTC": 1947, + "TGCTTG": 1948, + "TGCTCA": 1949, + "TGCTCT": 1950, + "TGCTCC": 1951, + "TGCTCG": 1952, + "TGCTGA": 1953, + "TGCTGT": 1954, + "TGCTGC": 1955, + "TGCTGG": 1956, + "TGCCAA": 1957, + "TGCCAT": 1958, + "TGCCAC": 1959, + "TGCCAG": 1960, + "TGCCTA": 1961, + "TGCCTT": 1962, + "TGCCTC": 1963, + "TGCCTG": 1964, + "TGCCCA": 1965, + "TGCCCT": 1966, + "TGCCCC": 1967, + "TGCCCG": 1968, + "TGCCGA": 1969, + "TGCCGT": 1970, + "TGCCGC": 1971, + "TGCCGG": 1972, + "TGCGAA": 1973, + "TGCGAT": 1974, + "TGCGAC": 1975, + "TGCGAG": 1976, + "TGCGTA": 1977, + "TGCGTT": 1978, + "TGCGTC": 1979, + "TGCGTG": 1980, + "TGCGCA": 1981, + "TGCGCT": 1982, + "TGCGCC": 1983, + "TGCGCG": 1984, + "TGCGGA": 1985, + "TGCGGT": 1986, + "TGCGGC": 1987, + "TGCGGG": 1988, + "TGGAAA": 1989, + "TGGAAT": 1990, + "TGGAAC": 1991, + "TGGAAG": 1992, + "TGGATA": 1993, + "TGGATT": 1994, + "TGGATC": 1995, + "TGGATG": 1996, + "TGGACA": 1997, + "TGGACT": 1998, + "TGGACC": 1999, + "TGGACG": 2000, + "TGGAGA": 2001, + "TGGAGT": 2002, + "TGGAGC": 2003, + "TGGAGG": 2004, + "TGGTAA": 2005, + "TGGTAT": 2006, + "TGGTAC": 2007, + "TGGTAG": 2008, + "TGGTTA": 2009, + "TGGTTT": 2010, + "TGGTTC": 2011, + "TGGTTG": 2012, + "TGGTCA": 2013, + "TGGTCT": 2014, + "TGGTCC": 2015, + "TGGTCG": 2016, + "TGGTGA": 2017, + "TGGTGT": 2018, + "TGGTGC": 2019, + "TGGTGG": 2020, + "TGGCAA": 2021, + "TGGCAT": 2022, + "TGGCAC": 2023, + "TGGCAG": 2024, + "TGGCTA": 2025, + "TGGCTT": 2026, + "TGGCTC": 2027, + "TGGCTG": 2028, + "TGGCCA": 2029, + "TGGCCT": 2030, + "TGGCCC": 2031, + "TGGCCG": 2032, + "TGGCGA": 2033, + "TGGCGT": 2034, + "TGGCGC": 2035, + "TGGCGG": 2036, + "TGGGAA": 2037, + "TGGGAT": 2038, + "TGGGAC": 2039, + "TGGGAG": 2040, + "TGGGTA": 2041, + "TGGGTT": 2042, + "TGGGTC": 2043, + "TGGGTG": 2044, + "TGGGCA": 2045, + "TGGGCT": 2046, + "TGGGCC": 2047, + "TGGGCG": 2048, + "TGGGGA": 2049, + "TGGGGT": 2050, + "TGGGGC": 2051, + "TGGGGG": 2052, + "CAAAAA": 2053, + "CAAAAT": 2054, + "CAAAAC": 2055, + "CAAAAG": 2056, + "CAAATA": 2057, + "CAAATT": 2058, + "CAAATC": 2059, + "CAAATG": 2060, + "CAAACA": 2061, + "CAAACT": 2062, + "CAAACC": 2063, + "CAAACG": 2064, + "CAAAGA": 2065, + "CAAAGT": 2066, + "CAAAGC": 2067, + "CAAAGG": 2068, + "CAATAA": 2069, + "CAATAT": 2070, + "CAATAC": 2071, + "CAATAG": 2072, + "CAATTA": 2073, + "CAATTT": 2074, + "CAATTC": 2075, + "CAATTG": 2076, + "CAATCA": 2077, + "CAATCT": 2078, + "CAATCC": 2079, + "CAATCG": 2080, + "CAATGA": 2081, + "CAATGT": 2082, + "CAATGC": 2083, + "CAATGG": 2084, + "CAACAA": 2085, + "CAACAT": 2086, + "CAACAC": 2087, + "CAACAG": 2088, + "CAACTA": 2089, + "CAACTT": 2090, + "CAACTC": 2091, + "CAACTG": 2092, + "CAACCA": 2093, + "CAACCT": 2094, + "CAACCC": 2095, + "CAACCG": 2096, + "CAACGA": 2097, + "CAACGT": 2098, + "CAACGC": 2099, + "CAACGG": 2100, + "CAAGAA": 2101, + "CAAGAT": 2102, + "CAAGAC": 2103, + "CAAGAG": 2104, + "CAAGTA": 2105, + "CAAGTT": 2106, + "CAAGTC": 2107, + "CAAGTG": 2108, + "CAAGCA": 2109, + "CAAGCT": 2110, + "CAAGCC": 2111, + "CAAGCG": 2112, + "CAAGGA": 2113, + "CAAGGT": 2114, + "CAAGGC": 2115, + "CAAGGG": 2116, + "CATAAA": 2117, + "CATAAT": 2118, + "CATAAC": 2119, + "CATAAG": 2120, + "CATATA": 2121, + "CATATT": 2122, + "CATATC": 2123, + "CATATG": 2124, + "CATACA": 2125, + "CATACT": 2126, + "CATACC": 2127, + "CATACG": 2128, + "CATAGA": 2129, + "CATAGT": 2130, + "CATAGC": 2131, + "CATAGG": 2132, + "CATTAA": 2133, + "CATTAT": 2134, + "CATTAC": 2135, + "CATTAG": 2136, + "CATTTA": 2137, + "CATTTT": 2138, + "CATTTC": 2139, + "CATTTG": 2140, + "CATTCA": 2141, + "CATTCT": 2142, + "CATTCC": 2143, + "CATTCG": 2144, + "CATTGA": 2145, + "CATTGT": 2146, + "CATTGC": 2147, + "CATTGG": 2148, + "CATCAA": 2149, + "CATCAT": 2150, + "CATCAC": 2151, + "CATCAG": 2152, + "CATCTA": 2153, + "CATCTT": 2154, + "CATCTC": 2155, + "CATCTG": 2156, + "CATCCA": 2157, + "CATCCT": 2158, + "CATCCC": 2159, + "CATCCG": 2160, + "CATCGA": 2161, + "CATCGT": 2162, + "CATCGC": 2163, + "CATCGG": 2164, + "CATGAA": 2165, + "CATGAT": 2166, + "CATGAC": 2167, + "CATGAG": 2168, + "CATGTA": 2169, + "CATGTT": 2170, + "CATGTC": 2171, + "CATGTG": 2172, + "CATGCA": 2173, + "CATGCT": 2174, + "CATGCC": 2175, + "CATGCG": 2176, + "CATGGA": 2177, + "CATGGT": 2178, + "CATGGC": 2179, + "CATGGG": 2180, + "CACAAA": 2181, + "CACAAT": 2182, + "CACAAC": 2183, + "CACAAG": 2184, + "CACATA": 2185, + "CACATT": 2186, + "CACATC": 2187, + "CACATG": 2188, + "CACACA": 2189, + "CACACT": 2190, + "CACACC": 2191, + "CACACG": 2192, + "CACAGA": 2193, + "CACAGT": 2194, + "CACAGC": 2195, + "CACAGG": 2196, + "CACTAA": 2197, + "CACTAT": 2198, + "CACTAC": 2199, + "CACTAG": 2200, + "CACTTA": 2201, + "CACTTT": 2202, + "CACTTC": 2203, + "CACTTG": 2204, + "CACTCA": 2205, + "CACTCT": 2206, + "CACTCC": 2207, + "CACTCG": 2208, + "CACTGA": 2209, + "CACTGT": 2210, + "CACTGC": 2211, + "CACTGG": 2212, + "CACCAA": 2213, + "CACCAT": 2214, + "CACCAC": 2215, + "CACCAG": 2216, + "CACCTA": 2217, + "CACCTT": 2218, + "CACCTC": 2219, + "CACCTG": 2220, + "CACCCA": 2221, + "CACCCT": 2222, + "CACCCC": 2223, + "CACCCG": 2224, + "CACCGA": 2225, + "CACCGT": 2226, + "CACCGC": 2227, + "CACCGG": 2228, + "CACGAA": 2229, + "CACGAT": 2230, + "CACGAC": 2231, + "CACGAG": 2232, + "CACGTA": 2233, + "CACGTT": 2234, + "CACGTC": 2235, + "CACGTG": 2236, + "CACGCA": 2237, + "CACGCT": 2238, + "CACGCC": 2239, + "CACGCG": 2240, + "CACGGA": 2241, + "CACGGT": 2242, + "CACGGC": 2243, + "CACGGG": 2244, + "CAGAAA": 2245, + "CAGAAT": 2246, + "CAGAAC": 2247, + "CAGAAG": 2248, + "CAGATA": 2249, + "CAGATT": 2250, + "CAGATC": 2251, + "CAGATG": 2252, + "CAGACA": 2253, + "CAGACT": 2254, + "CAGACC": 2255, + "CAGACG": 2256, + "CAGAGA": 2257, + "CAGAGT": 2258, + "CAGAGC": 2259, + "CAGAGG": 2260, + "CAGTAA": 2261, + "CAGTAT": 2262, + "CAGTAC": 2263, + "CAGTAG": 2264, + "CAGTTA": 2265, + "CAGTTT": 2266, + "CAGTTC": 2267, + "CAGTTG": 2268, + "CAGTCA": 2269, + "CAGTCT": 2270, + "CAGTCC": 2271, + "CAGTCG": 2272, + "CAGTGA": 2273, + "CAGTGT": 2274, + "CAGTGC": 2275, + "CAGTGG": 2276, + "CAGCAA": 2277, + "CAGCAT": 2278, + "CAGCAC": 2279, + "CAGCAG": 2280, + "CAGCTA": 2281, + "CAGCTT": 2282, + "CAGCTC": 2283, + "CAGCTG": 2284, + "CAGCCA": 2285, + "CAGCCT": 2286, + "CAGCCC": 2287, + "CAGCCG": 2288, + "CAGCGA": 2289, + "CAGCGT": 2290, + "CAGCGC": 2291, + "CAGCGG": 2292, + "CAGGAA": 2293, + "CAGGAT": 2294, + "CAGGAC": 2295, + "CAGGAG": 2296, + "CAGGTA": 2297, + "CAGGTT": 2298, + "CAGGTC": 2299, + "CAGGTG": 2300, + "CAGGCA": 2301, + "CAGGCT": 2302, + "CAGGCC": 2303, + "CAGGCG": 2304, + "CAGGGA": 2305, + "CAGGGT": 2306, + "CAGGGC": 2307, + "CAGGGG": 2308, + "CTAAAA": 2309, + "CTAAAT": 2310, + "CTAAAC": 2311, + "CTAAAG": 2312, + "CTAATA": 2313, + "CTAATT": 2314, + "CTAATC": 2315, + "CTAATG": 2316, + "CTAACA": 2317, + "CTAACT": 2318, + "CTAACC": 2319, + "CTAACG": 2320, + "CTAAGA": 2321, + "CTAAGT": 2322, + "CTAAGC": 2323, + "CTAAGG": 2324, + "CTATAA": 2325, + "CTATAT": 2326, + "CTATAC": 2327, + "CTATAG": 2328, + "CTATTA": 2329, + "CTATTT": 2330, + "CTATTC": 2331, + "CTATTG": 2332, + "CTATCA": 2333, + "CTATCT": 2334, + "CTATCC": 2335, + "CTATCG": 2336, + "CTATGA": 2337, + "CTATGT": 2338, + "CTATGC": 2339, + "CTATGG": 2340, + "CTACAA": 2341, + "CTACAT": 2342, + "CTACAC": 2343, + "CTACAG": 2344, + "CTACTA": 2345, + "CTACTT": 2346, + "CTACTC": 2347, + "CTACTG": 2348, + "CTACCA": 2349, + "CTACCT": 2350, + "CTACCC": 2351, + "CTACCG": 2352, + "CTACGA": 2353, + "CTACGT": 2354, + "CTACGC": 2355, + "CTACGG": 2356, + "CTAGAA": 2357, + "CTAGAT": 2358, + "CTAGAC": 2359, + "CTAGAG": 2360, + "CTAGTA": 2361, + "CTAGTT": 2362, + "CTAGTC": 2363, + "CTAGTG": 2364, + "CTAGCA": 2365, + "CTAGCT": 2366, + "CTAGCC": 2367, + "CTAGCG": 2368, + "CTAGGA": 2369, + "CTAGGT": 2370, + "CTAGGC": 2371, + "CTAGGG": 2372, + "CTTAAA": 2373, + "CTTAAT": 2374, + "CTTAAC": 2375, + "CTTAAG": 2376, + "CTTATA": 2377, + "CTTATT": 2378, + "CTTATC": 2379, + "CTTATG": 2380, + "CTTACA": 2381, + "CTTACT": 2382, + "CTTACC": 2383, + "CTTACG": 2384, + "CTTAGA": 2385, + "CTTAGT": 2386, + "CTTAGC": 2387, + "CTTAGG": 2388, + "CTTTAA": 2389, + "CTTTAT": 2390, + "CTTTAC": 2391, + "CTTTAG": 2392, + "CTTTTA": 2393, + "CTTTTT": 2394, + "CTTTTC": 2395, + "CTTTTG": 2396, + "CTTTCA": 2397, + "CTTTCT": 2398, + "CTTTCC": 2399, + "CTTTCG": 2400, + "CTTTGA": 2401, + "CTTTGT": 2402, + "CTTTGC": 2403, + "CTTTGG": 2404, + "CTTCAA": 2405, + "CTTCAT": 2406, + "CTTCAC": 2407, + "CTTCAG": 2408, + "CTTCTA": 2409, + "CTTCTT": 2410, + "CTTCTC": 2411, + "CTTCTG": 2412, + "CTTCCA": 2413, + "CTTCCT": 2414, + "CTTCCC": 2415, + "CTTCCG": 2416, + "CTTCGA": 2417, + "CTTCGT": 2418, + "CTTCGC": 2419, + "CTTCGG": 2420, + "CTTGAA": 2421, + "CTTGAT": 2422, + "CTTGAC": 2423, + "CTTGAG": 2424, + "CTTGTA": 2425, + "CTTGTT": 2426, + "CTTGTC": 2427, + "CTTGTG": 2428, + "CTTGCA": 2429, + "CTTGCT": 2430, + "CTTGCC": 2431, + "CTTGCG": 2432, + "CTTGGA": 2433, + "CTTGGT": 2434, + "CTTGGC": 2435, + "CTTGGG": 2436, + "CTCAAA": 2437, + "CTCAAT": 2438, + "CTCAAC": 2439, + "CTCAAG": 2440, + "CTCATA": 2441, + "CTCATT": 2442, + "CTCATC": 2443, + "CTCATG": 2444, + "CTCACA": 2445, + "CTCACT": 2446, + "CTCACC": 2447, + "CTCACG": 2448, + "CTCAGA": 2449, + "CTCAGT": 2450, + "CTCAGC": 2451, + "CTCAGG": 2452, + "CTCTAA": 2453, + "CTCTAT": 2454, + "CTCTAC": 2455, + "CTCTAG": 2456, + "CTCTTA": 2457, + "CTCTTT": 2458, + "CTCTTC": 2459, + "CTCTTG": 2460, + "CTCTCA": 2461, + "CTCTCT": 2462, + "CTCTCC": 2463, + "CTCTCG": 2464, + "CTCTGA": 2465, + "CTCTGT": 2466, + "CTCTGC": 2467, + "CTCTGG": 2468, + "CTCCAA": 2469, + "CTCCAT": 2470, + "CTCCAC": 2471, + "CTCCAG": 2472, + "CTCCTA": 2473, + "CTCCTT": 2474, + "CTCCTC": 2475, + "CTCCTG": 2476, + "CTCCCA": 2477, + "CTCCCT": 2478, + "CTCCCC": 2479, + "CTCCCG": 2480, + "CTCCGA": 2481, + "CTCCGT": 2482, + "CTCCGC": 2483, + "CTCCGG": 2484, + "CTCGAA": 2485, + "CTCGAT": 2486, + "CTCGAC": 2487, + "CTCGAG": 2488, + "CTCGTA": 2489, + "CTCGTT": 2490, + "CTCGTC": 2491, + "CTCGTG": 2492, + "CTCGCA": 2493, + "CTCGCT": 2494, + "CTCGCC": 2495, + "CTCGCG": 2496, + "CTCGGA": 2497, + "CTCGGT": 2498, + "CTCGGC": 2499, + "CTCGGG": 2500, + "CTGAAA": 2501, + "CTGAAT": 2502, + "CTGAAC": 2503, + "CTGAAG": 2504, + "CTGATA": 2505, + "CTGATT": 2506, + "CTGATC": 2507, + "CTGATG": 2508, + "CTGACA": 2509, + "CTGACT": 2510, + "CTGACC": 2511, + "CTGACG": 2512, + "CTGAGA": 2513, + "CTGAGT": 2514, + "CTGAGC": 2515, + "CTGAGG": 2516, + "CTGTAA": 2517, + "CTGTAT": 2518, + "CTGTAC": 2519, + "CTGTAG": 2520, + "CTGTTA": 2521, + "CTGTTT": 2522, + "CTGTTC": 2523, + "CTGTTG": 2524, + "CTGTCA": 2525, + "CTGTCT": 2526, + "CTGTCC": 2527, + "CTGTCG": 2528, + "CTGTGA": 2529, + "CTGTGT": 2530, + "CTGTGC": 2531, + "CTGTGG": 2532, + "CTGCAA": 2533, + "CTGCAT": 2534, + "CTGCAC": 2535, + "CTGCAG": 2536, + "CTGCTA": 2537, + "CTGCTT": 2538, + "CTGCTC": 2539, + "CTGCTG": 2540, + "CTGCCA": 2541, + "CTGCCT": 2542, + "CTGCCC": 2543, + "CTGCCG": 2544, + "CTGCGA": 2545, + "CTGCGT": 2546, + "CTGCGC": 2547, + "CTGCGG": 2548, + "CTGGAA": 2549, + "CTGGAT": 2550, + "CTGGAC": 2551, + "CTGGAG": 2552, + "CTGGTA": 2553, + "CTGGTT": 2554, + "CTGGTC": 2555, + "CTGGTG": 2556, + "CTGGCA": 2557, + "CTGGCT": 2558, + "CTGGCC": 2559, + "CTGGCG": 2560, + "CTGGGA": 2561, + "CTGGGT": 2562, + "CTGGGC": 2563, + "CTGGGG": 2564, + "CCAAAA": 2565, + "CCAAAT": 2566, + "CCAAAC": 2567, + "CCAAAG": 2568, + "CCAATA": 2569, + "CCAATT": 2570, + "CCAATC": 2571, + "CCAATG": 2572, + "CCAACA": 2573, + "CCAACT": 2574, + "CCAACC": 2575, + "CCAACG": 2576, + "CCAAGA": 2577, + "CCAAGT": 2578, + "CCAAGC": 2579, + "CCAAGG": 2580, + "CCATAA": 2581, + "CCATAT": 2582, + "CCATAC": 2583, + "CCATAG": 2584, + "CCATTA": 2585, + "CCATTT": 2586, + "CCATTC": 2587, + "CCATTG": 2588, + "CCATCA": 2589, + "CCATCT": 2590, + "CCATCC": 2591, + "CCATCG": 2592, + "CCATGA": 2593, + "CCATGT": 2594, + "CCATGC": 2595, + "CCATGG": 2596, + "CCACAA": 2597, + "CCACAT": 2598, + "CCACAC": 2599, + "CCACAG": 2600, + "CCACTA": 2601, + "CCACTT": 2602, + "CCACTC": 2603, + "CCACTG": 2604, + "CCACCA": 2605, + "CCACCT": 2606, + "CCACCC": 2607, + "CCACCG": 2608, + "CCACGA": 2609, + "CCACGT": 2610, + "CCACGC": 2611, + "CCACGG": 2612, + "CCAGAA": 2613, + "CCAGAT": 2614, + "CCAGAC": 2615, + "CCAGAG": 2616, + "CCAGTA": 2617, + "CCAGTT": 2618, + "CCAGTC": 2619, + "CCAGTG": 2620, + "CCAGCA": 2621, + "CCAGCT": 2622, + "CCAGCC": 2623, + "CCAGCG": 2624, + "CCAGGA": 2625, + "CCAGGT": 2626, + "CCAGGC": 2627, + "CCAGGG": 2628, + "CCTAAA": 2629, + "CCTAAT": 2630, + "CCTAAC": 2631, + "CCTAAG": 2632, + "CCTATA": 2633, + "CCTATT": 2634, + "CCTATC": 2635, + "CCTATG": 2636, + "CCTACA": 2637, + "CCTACT": 2638, + "CCTACC": 2639, + "CCTACG": 2640, + "CCTAGA": 2641, + "CCTAGT": 2642, + "CCTAGC": 2643, + "CCTAGG": 2644, + "CCTTAA": 2645, + "CCTTAT": 2646, + "CCTTAC": 2647, + "CCTTAG": 2648, + "CCTTTA": 2649, + "CCTTTT": 2650, + "CCTTTC": 2651, + "CCTTTG": 2652, + "CCTTCA": 2653, + "CCTTCT": 2654, + "CCTTCC": 2655, + "CCTTCG": 2656, + "CCTTGA": 2657, + "CCTTGT": 2658, + "CCTTGC": 2659, + "CCTTGG": 2660, + "CCTCAA": 2661, + "CCTCAT": 2662, + "CCTCAC": 2663, + "CCTCAG": 2664, + "CCTCTA": 2665, + "CCTCTT": 2666, + "CCTCTC": 2667, + "CCTCTG": 2668, + "CCTCCA": 2669, + "CCTCCT": 2670, + "CCTCCC": 2671, + "CCTCCG": 2672, + "CCTCGA": 2673, + "CCTCGT": 2674, + "CCTCGC": 2675, + "CCTCGG": 2676, + "CCTGAA": 2677, + "CCTGAT": 2678, + "CCTGAC": 2679, + "CCTGAG": 2680, + "CCTGTA": 2681, + "CCTGTT": 2682, + "CCTGTC": 2683, + "CCTGTG": 2684, + "CCTGCA": 2685, + "CCTGCT": 2686, + "CCTGCC": 2687, + "CCTGCG": 2688, + "CCTGGA": 2689, + "CCTGGT": 2690, + "CCTGGC": 2691, + "CCTGGG": 2692, + "CCCAAA": 2693, + "CCCAAT": 2694, + "CCCAAC": 2695, + "CCCAAG": 2696, + "CCCATA": 2697, + "CCCATT": 2698, + "CCCATC": 2699, + "CCCATG": 2700, + "CCCACA": 2701, + "CCCACT": 2702, + "CCCACC": 2703, + "CCCACG": 2704, + "CCCAGA": 2705, + "CCCAGT": 2706, + "CCCAGC": 2707, + "CCCAGG": 2708, + "CCCTAA": 2709, + "CCCTAT": 2710, + "CCCTAC": 2711, + "CCCTAG": 2712, + "CCCTTA": 2713, + "CCCTTT": 2714, + "CCCTTC": 2715, + "CCCTTG": 2716, + "CCCTCA": 2717, + "CCCTCT": 2718, + "CCCTCC": 2719, + "CCCTCG": 2720, + "CCCTGA": 2721, + "CCCTGT": 2722, + "CCCTGC": 2723, + "CCCTGG": 2724, + "CCCCAA": 2725, + "CCCCAT": 2726, + "CCCCAC": 2727, + "CCCCAG": 2728, + "CCCCTA": 2729, + "CCCCTT": 2730, + "CCCCTC": 2731, + "CCCCTG": 2732, + "CCCCCA": 2733, + "CCCCCT": 2734, + "CCCCCC": 2735, + "CCCCCG": 2736, + "CCCCGA": 2737, + "CCCCGT": 2738, + "CCCCGC": 2739, + "CCCCGG": 2740, + "CCCGAA": 2741, + "CCCGAT": 2742, + "CCCGAC": 2743, + "CCCGAG": 2744, + "CCCGTA": 2745, + "CCCGTT": 2746, + "CCCGTC": 2747, + "CCCGTG": 2748, + "CCCGCA": 2749, + "CCCGCT": 2750, + "CCCGCC": 2751, + "CCCGCG": 2752, + "CCCGGA": 2753, + "CCCGGT": 2754, + "CCCGGC": 2755, + "CCCGGG": 2756, + "CCGAAA": 2757, + "CCGAAT": 2758, + "CCGAAC": 2759, + "CCGAAG": 2760, + "CCGATA": 2761, + "CCGATT": 2762, + "CCGATC": 2763, + "CCGATG": 2764, + "CCGACA": 2765, + "CCGACT": 2766, + "CCGACC": 2767, + "CCGACG": 2768, + "CCGAGA": 2769, + "CCGAGT": 2770, + "CCGAGC": 2771, + "CCGAGG": 2772, + "CCGTAA": 2773, + "CCGTAT": 2774, + "CCGTAC": 2775, + "CCGTAG": 2776, + "CCGTTA": 2777, + "CCGTTT": 2778, + "CCGTTC": 2779, + "CCGTTG": 2780, + "CCGTCA": 2781, + "CCGTCT": 2782, + "CCGTCC": 2783, + "CCGTCG": 2784, + "CCGTGA": 2785, + "CCGTGT": 2786, + "CCGTGC": 2787, + "CCGTGG": 2788, + "CCGCAA": 2789, + "CCGCAT": 2790, + "CCGCAC": 2791, + "CCGCAG": 2792, + "CCGCTA": 2793, + "CCGCTT": 2794, + "CCGCTC": 2795, + "CCGCTG": 2796, + "CCGCCA": 2797, + "CCGCCT": 2798, + "CCGCCC": 2799, + "CCGCCG": 2800, + "CCGCGA": 2801, + "CCGCGT": 2802, + "CCGCGC": 2803, + "CCGCGG": 2804, + "CCGGAA": 2805, + "CCGGAT": 2806, + "CCGGAC": 2807, + "CCGGAG": 2808, + "CCGGTA": 2809, + "CCGGTT": 2810, + "CCGGTC": 2811, + "CCGGTG": 2812, + "CCGGCA": 2813, + "CCGGCT": 2814, + "CCGGCC": 2815, + "CCGGCG": 2816, + "CCGGGA": 2817, + "CCGGGT": 2818, + "CCGGGC": 2819, + "CCGGGG": 2820, + "CGAAAA": 2821, + "CGAAAT": 2822, + "CGAAAC": 2823, + "CGAAAG": 2824, + "CGAATA": 2825, + "CGAATT": 2826, + "CGAATC": 2827, + "CGAATG": 2828, + "CGAACA": 2829, + "CGAACT": 2830, + "CGAACC": 2831, + "CGAACG": 2832, + "CGAAGA": 2833, + "CGAAGT": 2834, + "CGAAGC": 2835, + "CGAAGG": 2836, + "CGATAA": 2837, + "CGATAT": 2838, + "CGATAC": 2839, + "CGATAG": 2840, + "CGATTA": 2841, + "CGATTT": 2842, + "CGATTC": 2843, + "CGATTG": 2844, + "CGATCA": 2845, + "CGATCT": 2846, + "CGATCC": 2847, + "CGATCG": 2848, + "CGATGA": 2849, + "CGATGT": 2850, + "CGATGC": 2851, + "CGATGG": 2852, + "CGACAA": 2853, + "CGACAT": 2854, + "CGACAC": 2855, + "CGACAG": 2856, + "CGACTA": 2857, + "CGACTT": 2858, + "CGACTC": 2859, + "CGACTG": 2860, + "CGACCA": 2861, + "CGACCT": 2862, + "CGACCC": 2863, + "CGACCG": 2864, + "CGACGA": 2865, + "CGACGT": 2866, + "CGACGC": 2867, + "CGACGG": 2868, + "CGAGAA": 2869, + "CGAGAT": 2870, + "CGAGAC": 2871, + "CGAGAG": 2872, + "CGAGTA": 2873, + "CGAGTT": 2874, + "CGAGTC": 2875, + "CGAGTG": 2876, + "CGAGCA": 2877, + "CGAGCT": 2878, + "CGAGCC": 2879, + "CGAGCG": 2880, + "CGAGGA": 2881, + "CGAGGT": 2882, + "CGAGGC": 2883, + "CGAGGG": 2884, + "CGTAAA": 2885, + "CGTAAT": 2886, + "CGTAAC": 2887, + "CGTAAG": 2888, + "CGTATA": 2889, + "CGTATT": 2890, + "CGTATC": 2891, + "CGTATG": 2892, + "CGTACA": 2893, + "CGTACT": 2894, + "CGTACC": 2895, + "CGTACG": 2896, + "CGTAGA": 2897, + "CGTAGT": 2898, + "CGTAGC": 2899, + "CGTAGG": 2900, + "CGTTAA": 2901, + "CGTTAT": 2902, + "CGTTAC": 2903, + "CGTTAG": 2904, + "CGTTTA": 2905, + "CGTTTT": 2906, + "CGTTTC": 2907, + "CGTTTG": 2908, + "CGTTCA": 2909, + "CGTTCT": 2910, + "CGTTCC": 2911, + "CGTTCG": 2912, + "CGTTGA": 2913, + "CGTTGT": 2914, + "CGTTGC": 2915, + "CGTTGG": 2916, + "CGTCAA": 2917, + "CGTCAT": 2918, + "CGTCAC": 2919, + "CGTCAG": 2920, + "CGTCTA": 2921, + "CGTCTT": 2922, + "CGTCTC": 2923, + "CGTCTG": 2924, + "CGTCCA": 2925, + "CGTCCT": 2926, + "CGTCCC": 2927, + "CGTCCG": 2928, + "CGTCGA": 2929, + "CGTCGT": 2930, + "CGTCGC": 2931, + "CGTCGG": 2932, + "CGTGAA": 2933, + "CGTGAT": 2934, + "CGTGAC": 2935, + "CGTGAG": 2936, + "CGTGTA": 2937, + "CGTGTT": 2938, + "CGTGTC": 2939, + "CGTGTG": 2940, + "CGTGCA": 2941, + "CGTGCT": 2942, + "CGTGCC": 2943, + "CGTGCG": 2944, + "CGTGGA": 2945, + "CGTGGT": 2946, + "CGTGGC": 2947, + "CGTGGG": 2948, + "CGCAAA": 2949, + "CGCAAT": 2950, + "CGCAAC": 2951, + "CGCAAG": 2952, + "CGCATA": 2953, + "CGCATT": 2954, + "CGCATC": 2955, + "CGCATG": 2956, + "CGCACA": 2957, + "CGCACT": 2958, + "CGCACC": 2959, + "CGCACG": 2960, + "CGCAGA": 2961, + "CGCAGT": 2962, + "CGCAGC": 2963, + "CGCAGG": 2964, + "CGCTAA": 2965, + "CGCTAT": 2966, + "CGCTAC": 2967, + "CGCTAG": 2968, + "CGCTTA": 2969, + "CGCTTT": 2970, + "CGCTTC": 2971, + "CGCTTG": 2972, + "CGCTCA": 2973, + "CGCTCT": 2974, + "CGCTCC": 2975, + "CGCTCG": 2976, + "CGCTGA": 2977, + "CGCTGT": 2978, + "CGCTGC": 2979, + "CGCTGG": 2980, + "CGCCAA": 2981, + "CGCCAT": 2982, + "CGCCAC": 2983, + "CGCCAG": 2984, + "CGCCTA": 2985, + "CGCCTT": 2986, + "CGCCTC": 2987, + "CGCCTG": 2988, + "CGCCCA": 2989, + "CGCCCT": 2990, + "CGCCCC": 2991, + "CGCCCG": 2992, + "CGCCGA": 2993, + "CGCCGT": 2994, + "CGCCGC": 2995, + "CGCCGG": 2996, + "CGCGAA": 2997, + "CGCGAT": 2998, + "CGCGAC": 2999, + "CGCGAG": 3000, + "CGCGTA": 3001, + "CGCGTT": 3002, + "CGCGTC": 3003, + "CGCGTG": 3004, + "CGCGCA": 3005, + "CGCGCT": 3006, + "CGCGCC": 3007, + "CGCGCG": 3008, + "CGCGGA": 3009, + "CGCGGT": 3010, + "CGCGGC": 3011, + "CGCGGG": 3012, + "CGGAAA": 3013, + "CGGAAT": 3014, + "CGGAAC": 3015, + "CGGAAG": 3016, + "CGGATA": 3017, + "CGGATT": 3018, + "CGGATC": 3019, + "CGGATG": 3020, + "CGGACA": 3021, + "CGGACT": 3022, + "CGGACC": 3023, + "CGGACG": 3024, + "CGGAGA": 3025, + "CGGAGT": 3026, + "CGGAGC": 3027, + "CGGAGG": 3028, + "CGGTAA": 3029, + "CGGTAT": 3030, + "CGGTAC": 3031, + "CGGTAG": 3032, + "CGGTTA": 3033, + "CGGTTT": 3034, + "CGGTTC": 3035, + "CGGTTG": 3036, + "CGGTCA": 3037, + "CGGTCT": 3038, + "CGGTCC": 3039, + "CGGTCG": 3040, + "CGGTGA": 3041, + "CGGTGT": 3042, + "CGGTGC": 3043, + "CGGTGG": 3044, + "CGGCAA": 3045, + "CGGCAT": 3046, + "CGGCAC": 3047, + "CGGCAG": 3048, + "CGGCTA": 3049, + "CGGCTT": 3050, + "CGGCTC": 3051, + "CGGCTG": 3052, + "CGGCCA": 3053, + "CGGCCT": 3054, + "CGGCCC": 3055, + "CGGCCG": 3056, + "CGGCGA": 3057, + "CGGCGT": 3058, + "CGGCGC": 3059, + "CGGCGG": 3060, + "CGGGAA": 3061, + "CGGGAT": 3062, + "CGGGAC": 3063, + "CGGGAG": 3064, + "CGGGTA": 3065, + "CGGGTT": 3066, + "CGGGTC": 3067, + "CGGGTG": 3068, + "CGGGCA": 3069, + "CGGGCT": 3070, + "CGGGCC": 3071, + "CGGGCG": 3072, + "CGGGGA": 3073, + "CGGGGT": 3074, + "CGGGGC": 3075, + "CGGGGG": 3076, + "GAAAAA": 3077, + "GAAAAT": 3078, + "GAAAAC": 3079, + "GAAAAG": 3080, + "GAAATA": 3081, + "GAAATT": 3082, + "GAAATC": 3083, + "GAAATG": 3084, + "GAAACA": 3085, + "GAAACT": 3086, + "GAAACC": 3087, + "GAAACG": 3088, + "GAAAGA": 3089, + "GAAAGT": 3090, + "GAAAGC": 3091, + "GAAAGG": 3092, + "GAATAA": 3093, + "GAATAT": 3094, + "GAATAC": 3095, + "GAATAG": 3096, + "GAATTA": 3097, + "GAATTT": 3098, + "GAATTC": 3099, + "GAATTG": 3100, + "GAATCA": 3101, + "GAATCT": 3102, + "GAATCC": 3103, + "GAATCG": 3104, + "GAATGA": 3105, + "GAATGT": 3106, + "GAATGC": 3107, + "GAATGG": 3108, + "GAACAA": 3109, + "GAACAT": 3110, + "GAACAC": 3111, + "GAACAG": 3112, + "GAACTA": 3113, + "GAACTT": 3114, + "GAACTC": 3115, + "GAACTG": 3116, + "GAACCA": 3117, + "GAACCT": 3118, + "GAACCC": 3119, + "GAACCG": 3120, + "GAACGA": 3121, + "GAACGT": 3122, + "GAACGC": 3123, + "GAACGG": 3124, + "GAAGAA": 3125, + "GAAGAT": 3126, + "GAAGAC": 3127, + "GAAGAG": 3128, + "GAAGTA": 3129, + "GAAGTT": 3130, + "GAAGTC": 3131, + "GAAGTG": 3132, + "GAAGCA": 3133, + "GAAGCT": 3134, + "GAAGCC": 3135, + "GAAGCG": 3136, + "GAAGGA": 3137, + "GAAGGT": 3138, + "GAAGGC": 3139, + "GAAGGG": 3140, + "GATAAA": 3141, + "GATAAT": 3142, + "GATAAC": 3143, + "GATAAG": 3144, + "GATATA": 3145, + "GATATT": 3146, + "GATATC": 3147, + "GATATG": 3148, + "GATACA": 3149, + "GATACT": 3150, + "GATACC": 3151, + "GATACG": 3152, + "GATAGA": 3153, + "GATAGT": 3154, + "GATAGC": 3155, + "GATAGG": 3156, + "GATTAA": 3157, + "GATTAT": 3158, + "GATTAC": 3159, + "GATTAG": 3160, + "GATTTA": 3161, + "GATTTT": 3162, + "GATTTC": 3163, + "GATTTG": 3164, + "GATTCA": 3165, + "GATTCT": 3166, + "GATTCC": 3167, + "GATTCG": 3168, + "GATTGA": 3169, + "GATTGT": 3170, + "GATTGC": 3171, + "GATTGG": 3172, + "GATCAA": 3173, + "GATCAT": 3174, + "GATCAC": 3175, + "GATCAG": 3176, + "GATCTA": 3177, + "GATCTT": 3178, + "GATCTC": 3179, + "GATCTG": 3180, + "GATCCA": 3181, + "GATCCT": 3182, + "GATCCC": 3183, + "GATCCG": 3184, + "GATCGA": 3185, + "GATCGT": 3186, + "GATCGC": 3187, + "GATCGG": 3188, + "GATGAA": 3189, + "GATGAT": 3190, + "GATGAC": 3191, + "GATGAG": 3192, + "GATGTA": 3193, + "GATGTT": 3194, + "GATGTC": 3195, + "GATGTG": 3196, + "GATGCA": 3197, + "GATGCT": 3198, + "GATGCC": 3199, + "GATGCG": 3200, + "GATGGA": 3201, + "GATGGT": 3202, + "GATGGC": 3203, + "GATGGG": 3204, + "GACAAA": 3205, + "GACAAT": 3206, + "GACAAC": 3207, + "GACAAG": 3208, + "GACATA": 3209, + "GACATT": 3210, + "GACATC": 3211, + "GACATG": 3212, + "GACACA": 3213, + "GACACT": 3214, + "GACACC": 3215, + "GACACG": 3216, + "GACAGA": 3217, + "GACAGT": 3218, + "GACAGC": 3219, + "GACAGG": 3220, + "GACTAA": 3221, + "GACTAT": 3222, + "GACTAC": 3223, + "GACTAG": 3224, + "GACTTA": 3225, + "GACTTT": 3226, + "GACTTC": 3227, + "GACTTG": 3228, + "GACTCA": 3229, + "GACTCT": 3230, + "GACTCC": 3231, + "GACTCG": 3232, + "GACTGA": 3233, + "GACTGT": 3234, + "GACTGC": 3235, + "GACTGG": 3236, + "GACCAA": 3237, + "GACCAT": 3238, + "GACCAC": 3239, + "GACCAG": 3240, + "GACCTA": 3241, + "GACCTT": 3242, + "GACCTC": 3243, + "GACCTG": 3244, + "GACCCA": 3245, + "GACCCT": 3246, + "GACCCC": 3247, + "GACCCG": 3248, + "GACCGA": 3249, + "GACCGT": 3250, + "GACCGC": 3251, + "GACCGG": 3252, + "GACGAA": 3253, + "GACGAT": 3254, + "GACGAC": 3255, + "GACGAG": 3256, + "GACGTA": 3257, + "GACGTT": 3258, + "GACGTC": 3259, + "GACGTG": 3260, + "GACGCA": 3261, + "GACGCT": 3262, + "GACGCC": 3263, + "GACGCG": 3264, + "GACGGA": 3265, + "GACGGT": 3266, + "GACGGC": 3267, + "GACGGG": 3268, + "GAGAAA": 3269, + "GAGAAT": 3270, + "GAGAAC": 3271, + "GAGAAG": 3272, + "GAGATA": 3273, + "GAGATT": 3274, + "GAGATC": 3275, + "GAGATG": 3276, + "GAGACA": 3277, + "GAGACT": 3278, + "GAGACC": 3279, + "GAGACG": 3280, + "GAGAGA": 3281, + "GAGAGT": 3282, + "GAGAGC": 3283, + "GAGAGG": 3284, + "GAGTAA": 3285, + "GAGTAT": 3286, + "GAGTAC": 3287, + "GAGTAG": 3288, + "GAGTTA": 3289, + "GAGTTT": 3290, + "GAGTTC": 3291, + "GAGTTG": 3292, + "GAGTCA": 3293, + "GAGTCT": 3294, + "GAGTCC": 3295, + "GAGTCG": 3296, + "GAGTGA": 3297, + "GAGTGT": 3298, + "GAGTGC": 3299, + "GAGTGG": 3300, + "GAGCAA": 3301, + "GAGCAT": 3302, + "GAGCAC": 3303, + "GAGCAG": 3304, + "GAGCTA": 3305, + "GAGCTT": 3306, + "GAGCTC": 3307, + "GAGCTG": 3308, + "GAGCCA": 3309, + "GAGCCT": 3310, + "GAGCCC": 3311, + "GAGCCG": 3312, + "GAGCGA": 3313, + "GAGCGT": 3314, + "GAGCGC": 3315, + "GAGCGG": 3316, + "GAGGAA": 3317, + "GAGGAT": 3318, + "GAGGAC": 3319, + "GAGGAG": 3320, + "GAGGTA": 3321, + "GAGGTT": 3322, + "GAGGTC": 3323, + "GAGGTG": 3324, + "GAGGCA": 3325, + "GAGGCT": 3326, + "GAGGCC": 3327, + "GAGGCG": 3328, + "GAGGGA": 3329, + "GAGGGT": 3330, + "GAGGGC": 3331, + "GAGGGG": 3332, + "GTAAAA": 3333, + "GTAAAT": 3334, + "GTAAAC": 3335, + "GTAAAG": 3336, + "GTAATA": 3337, + "GTAATT": 3338, + "GTAATC": 3339, + "GTAATG": 3340, + "GTAACA": 3341, + "GTAACT": 3342, + "GTAACC": 3343, + "GTAACG": 3344, + "GTAAGA": 3345, + "GTAAGT": 3346, + "GTAAGC": 3347, + "GTAAGG": 3348, + "GTATAA": 3349, + "GTATAT": 3350, + "GTATAC": 3351, + "GTATAG": 3352, + "GTATTA": 3353, + "GTATTT": 3354, + "GTATTC": 3355, + "GTATTG": 3356, + "GTATCA": 3357, + "GTATCT": 3358, + "GTATCC": 3359, + "GTATCG": 3360, + "GTATGA": 3361, + "GTATGT": 3362, + "GTATGC": 3363, + "GTATGG": 3364, + "GTACAA": 3365, + "GTACAT": 3366, + "GTACAC": 3367, + "GTACAG": 3368, + "GTACTA": 3369, + "GTACTT": 3370, + "GTACTC": 3371, + "GTACTG": 3372, + "GTACCA": 3373, + "GTACCT": 3374, + "GTACCC": 3375, + "GTACCG": 3376, + "GTACGA": 3377, + "GTACGT": 3378, + "GTACGC": 3379, + "GTACGG": 3380, + "GTAGAA": 3381, + "GTAGAT": 3382, + "GTAGAC": 3383, + "GTAGAG": 3384, + "GTAGTA": 3385, + "GTAGTT": 3386, + "GTAGTC": 3387, + "GTAGTG": 3388, + "GTAGCA": 3389, + "GTAGCT": 3390, + "GTAGCC": 3391, + "GTAGCG": 3392, + "GTAGGA": 3393, + "GTAGGT": 3394, + "GTAGGC": 3395, + "GTAGGG": 3396, + "GTTAAA": 3397, + "GTTAAT": 3398, + "GTTAAC": 3399, + "GTTAAG": 3400, + "GTTATA": 3401, + "GTTATT": 3402, + "GTTATC": 3403, + "GTTATG": 3404, + "GTTACA": 3405, + "GTTACT": 3406, + "GTTACC": 3407, + "GTTACG": 3408, + "GTTAGA": 3409, + "GTTAGT": 3410, + "GTTAGC": 3411, + "GTTAGG": 3412, + "GTTTAA": 3413, + "GTTTAT": 3414, + "GTTTAC": 3415, + "GTTTAG": 3416, + "GTTTTA": 3417, + "GTTTTT": 3418, + "GTTTTC": 3419, + "GTTTTG": 3420, + "GTTTCA": 3421, + "GTTTCT": 3422, + "GTTTCC": 3423, + "GTTTCG": 3424, + "GTTTGA": 3425, + "GTTTGT": 3426, + "GTTTGC": 3427, + "GTTTGG": 3428, + "GTTCAA": 3429, + "GTTCAT": 3430, + "GTTCAC": 3431, + "GTTCAG": 3432, + "GTTCTA": 3433, + "GTTCTT": 3434, + "GTTCTC": 3435, + "GTTCTG": 3436, + "GTTCCA": 3437, + "GTTCCT": 3438, + "GTTCCC": 3439, + "GTTCCG": 3440, + "GTTCGA": 3441, + "GTTCGT": 3442, + "GTTCGC": 3443, + "GTTCGG": 3444, + "GTTGAA": 3445, + "GTTGAT": 3446, + "GTTGAC": 3447, + "GTTGAG": 3448, + "GTTGTA": 3449, + "GTTGTT": 3450, + "GTTGTC": 3451, + "GTTGTG": 3452, + "GTTGCA": 3453, + "GTTGCT": 3454, + "GTTGCC": 3455, + "GTTGCG": 3456, + "GTTGGA": 3457, + "GTTGGT": 3458, + "GTTGGC": 3459, + "GTTGGG": 3460, + "GTCAAA": 3461, + "GTCAAT": 3462, + "GTCAAC": 3463, + "GTCAAG": 3464, + "GTCATA": 3465, + "GTCATT": 3466, + "GTCATC": 3467, + "GTCATG": 3468, + "GTCACA": 3469, + "GTCACT": 3470, + "GTCACC": 3471, + "GTCACG": 3472, + "GTCAGA": 3473, + "GTCAGT": 3474, + "GTCAGC": 3475, + "GTCAGG": 3476, + "GTCTAA": 3477, + "GTCTAT": 3478, + "GTCTAC": 3479, + "GTCTAG": 3480, + "GTCTTA": 3481, + "GTCTTT": 3482, + "GTCTTC": 3483, + "GTCTTG": 3484, + "GTCTCA": 3485, + "GTCTCT": 3486, + "GTCTCC": 3487, + "GTCTCG": 3488, + "GTCTGA": 3489, + "GTCTGT": 3490, + "GTCTGC": 3491, + "GTCTGG": 3492, + "GTCCAA": 3493, + "GTCCAT": 3494, + "GTCCAC": 3495, + "GTCCAG": 3496, + "GTCCTA": 3497, + "GTCCTT": 3498, + "GTCCTC": 3499, + "GTCCTG": 3500, + "GTCCCA": 3501, + "GTCCCT": 3502, + "GTCCCC": 3503, + "GTCCCG": 3504, + "GTCCGA": 3505, + "GTCCGT": 3506, + "GTCCGC": 3507, + "GTCCGG": 3508, + "GTCGAA": 3509, + "GTCGAT": 3510, + "GTCGAC": 3511, + "GTCGAG": 3512, + "GTCGTA": 3513, + "GTCGTT": 3514, + "GTCGTC": 3515, + "GTCGTG": 3516, + "GTCGCA": 3517, + "GTCGCT": 3518, + "GTCGCC": 3519, + "GTCGCG": 3520, + "GTCGGA": 3521, + "GTCGGT": 3522, + "GTCGGC": 3523, + "GTCGGG": 3524, + "GTGAAA": 3525, + "GTGAAT": 3526, + "GTGAAC": 3527, + "GTGAAG": 3528, + "GTGATA": 3529, + "GTGATT": 3530, + "GTGATC": 3531, + "GTGATG": 3532, + "GTGACA": 3533, + "GTGACT": 3534, + "GTGACC": 3535, + "GTGACG": 3536, + "GTGAGA": 3537, + "GTGAGT": 3538, + "GTGAGC": 3539, + "GTGAGG": 3540, + "GTGTAA": 3541, + "GTGTAT": 3542, + "GTGTAC": 3543, + "GTGTAG": 3544, + "GTGTTA": 3545, + "GTGTTT": 3546, + "GTGTTC": 3547, + "GTGTTG": 3548, + "GTGTCA": 3549, + "GTGTCT": 3550, + "GTGTCC": 3551, + "GTGTCG": 3552, + "GTGTGA": 3553, + "GTGTGT": 3554, + "GTGTGC": 3555, + "GTGTGG": 3556, + "GTGCAA": 3557, + "GTGCAT": 3558, + "GTGCAC": 3559, + "GTGCAG": 3560, + "GTGCTA": 3561, + "GTGCTT": 3562, + "GTGCTC": 3563, + "GTGCTG": 3564, + "GTGCCA": 3565, + "GTGCCT": 3566, + "GTGCCC": 3567, + "GTGCCG": 3568, + "GTGCGA": 3569, + "GTGCGT": 3570, + "GTGCGC": 3571, + "GTGCGG": 3572, + "GTGGAA": 3573, + "GTGGAT": 3574, + "GTGGAC": 3575, + "GTGGAG": 3576, + "GTGGTA": 3577, + "GTGGTT": 3578, + "GTGGTC": 3579, + "GTGGTG": 3580, + "GTGGCA": 3581, + "GTGGCT": 3582, + "GTGGCC": 3583, + "GTGGCG": 3584, + "GTGGGA": 3585, + "GTGGGT": 3586, + "GTGGGC": 3587, + "GTGGGG": 3588, + "GCAAAA": 3589, + "GCAAAT": 3590, + "GCAAAC": 3591, + "GCAAAG": 3592, + "GCAATA": 3593, + "GCAATT": 3594, + "GCAATC": 3595, + "GCAATG": 3596, + "GCAACA": 3597, + "GCAACT": 3598, + "GCAACC": 3599, + "GCAACG": 3600, + "GCAAGA": 3601, + "GCAAGT": 3602, + "GCAAGC": 3603, + "GCAAGG": 3604, + "GCATAA": 3605, + "GCATAT": 3606, + "GCATAC": 3607, + "GCATAG": 3608, + "GCATTA": 3609, + "GCATTT": 3610, + "GCATTC": 3611, + "GCATTG": 3612, + "GCATCA": 3613, + "GCATCT": 3614, + "GCATCC": 3615, + "GCATCG": 3616, + "GCATGA": 3617, + "GCATGT": 3618, + "GCATGC": 3619, + "GCATGG": 3620, + "GCACAA": 3621, + "GCACAT": 3622, + "GCACAC": 3623, + "GCACAG": 3624, + "GCACTA": 3625, + "GCACTT": 3626, + "GCACTC": 3627, + "GCACTG": 3628, + "GCACCA": 3629, + "GCACCT": 3630, + "GCACCC": 3631, + "GCACCG": 3632, + "GCACGA": 3633, + "GCACGT": 3634, + "GCACGC": 3635, + "GCACGG": 3636, + "GCAGAA": 3637, + "GCAGAT": 3638, + "GCAGAC": 3639, + "GCAGAG": 3640, + "GCAGTA": 3641, + "GCAGTT": 3642, + "GCAGTC": 3643, + "GCAGTG": 3644, + "GCAGCA": 3645, + "GCAGCT": 3646, + "GCAGCC": 3647, + "GCAGCG": 3648, + "GCAGGA": 3649, + "GCAGGT": 3650, + "GCAGGC": 3651, + "GCAGGG": 3652, + "GCTAAA": 3653, + "GCTAAT": 3654, + "GCTAAC": 3655, + "GCTAAG": 3656, + "GCTATA": 3657, + "GCTATT": 3658, + "GCTATC": 3659, + "GCTATG": 3660, + "GCTACA": 3661, + "GCTACT": 3662, + "GCTACC": 3663, + "GCTACG": 3664, + "GCTAGA": 3665, + "GCTAGT": 3666, + "GCTAGC": 3667, + "GCTAGG": 3668, + "GCTTAA": 3669, + "GCTTAT": 3670, + "GCTTAC": 3671, + "GCTTAG": 3672, + "GCTTTA": 3673, + "GCTTTT": 3674, + "GCTTTC": 3675, + "GCTTTG": 3676, + "GCTTCA": 3677, + "GCTTCT": 3678, + "GCTTCC": 3679, + "GCTTCG": 3680, + "GCTTGA": 3681, + "GCTTGT": 3682, + "GCTTGC": 3683, + "GCTTGG": 3684, + "GCTCAA": 3685, + "GCTCAT": 3686, + "GCTCAC": 3687, + "GCTCAG": 3688, + "GCTCTA": 3689, + "GCTCTT": 3690, + "GCTCTC": 3691, + "GCTCTG": 3692, + "GCTCCA": 3693, + "GCTCCT": 3694, + "GCTCCC": 3695, + "GCTCCG": 3696, + "GCTCGA": 3697, + "GCTCGT": 3698, + "GCTCGC": 3699, + "GCTCGG": 3700, + "GCTGAA": 3701, + "GCTGAT": 3702, + "GCTGAC": 3703, + "GCTGAG": 3704, + "GCTGTA": 3705, + "GCTGTT": 3706, + "GCTGTC": 3707, + "GCTGTG": 3708, + "GCTGCA": 3709, + "GCTGCT": 3710, + "GCTGCC": 3711, + "GCTGCG": 3712, + "GCTGGA": 3713, + "GCTGGT": 3714, + "GCTGGC": 3715, + "GCTGGG": 3716, + "GCCAAA": 3717, + "GCCAAT": 3718, + "GCCAAC": 3719, + "GCCAAG": 3720, + "GCCATA": 3721, + "GCCATT": 3722, + "GCCATC": 3723, + "GCCATG": 3724, + "GCCACA": 3725, + "GCCACT": 3726, + "GCCACC": 3727, + "GCCACG": 3728, + "GCCAGA": 3729, + "GCCAGT": 3730, + "GCCAGC": 3731, + "GCCAGG": 3732, + "GCCTAA": 3733, + "GCCTAT": 3734, + "GCCTAC": 3735, + "GCCTAG": 3736, + "GCCTTA": 3737, + "GCCTTT": 3738, + "GCCTTC": 3739, + "GCCTTG": 3740, + "GCCTCA": 3741, + "GCCTCT": 3742, + "GCCTCC": 3743, + "GCCTCG": 3744, + "GCCTGA": 3745, + "GCCTGT": 3746, + "GCCTGC": 3747, + "GCCTGG": 3748, + "GCCCAA": 3749, + "GCCCAT": 3750, + "GCCCAC": 3751, + "GCCCAG": 3752, + "GCCCTA": 3753, + "GCCCTT": 3754, + "GCCCTC": 3755, + "GCCCTG": 3756, + "GCCCCA": 3757, + "GCCCCT": 3758, + "GCCCCC": 3759, + "GCCCCG": 3760, + "GCCCGA": 3761, + "GCCCGT": 3762, + "GCCCGC": 3763, + "GCCCGG": 3764, + "GCCGAA": 3765, + "GCCGAT": 3766, + "GCCGAC": 3767, + "GCCGAG": 3768, + "GCCGTA": 3769, + "GCCGTT": 3770, + "GCCGTC": 3771, + "GCCGTG": 3772, + "GCCGCA": 3773, + "GCCGCT": 3774, + "GCCGCC": 3775, + "GCCGCG": 3776, + "GCCGGA": 3777, + "GCCGGT": 3778, + "GCCGGC": 3779, + "GCCGGG": 3780, + "GCGAAA": 3781, + "GCGAAT": 3782, + "GCGAAC": 3783, + "GCGAAG": 3784, + "GCGATA": 3785, + "GCGATT": 3786, + "GCGATC": 3787, + "GCGATG": 3788, + "GCGACA": 3789, + "GCGACT": 3790, + "GCGACC": 3791, + "GCGACG": 3792, + "GCGAGA": 3793, + "GCGAGT": 3794, + "GCGAGC": 3795, + "GCGAGG": 3796, + "GCGTAA": 3797, + "GCGTAT": 3798, + "GCGTAC": 3799, + "GCGTAG": 3800, + "GCGTTA": 3801, + "GCGTTT": 3802, + "GCGTTC": 3803, + "GCGTTG": 3804, + "GCGTCA": 3805, + "GCGTCT": 3806, + "GCGTCC": 3807, + "GCGTCG": 3808, + "GCGTGA": 3809, + "GCGTGT": 3810, + "GCGTGC": 3811, + "GCGTGG": 3812, + "GCGCAA": 3813, + "GCGCAT": 3814, + "GCGCAC": 3815, + "GCGCAG": 3816, + "GCGCTA": 3817, + "GCGCTT": 3818, + "GCGCTC": 3819, + "GCGCTG": 3820, + "GCGCCA": 3821, + "GCGCCT": 3822, + "GCGCCC": 3823, + "GCGCCG": 3824, + "GCGCGA": 3825, + "GCGCGT": 3826, + "GCGCGC": 3827, + "GCGCGG": 3828, + "GCGGAA": 3829, + "GCGGAT": 3830, + "GCGGAC": 3831, + "GCGGAG": 3832, + "GCGGTA": 3833, + "GCGGTT": 3834, + "GCGGTC": 3835, + "GCGGTG": 3836, + "GCGGCA": 3837, + "GCGGCT": 3838, + "GCGGCC": 3839, + "GCGGCG": 3840, + "GCGGGA": 3841, + "GCGGGT": 3842, + "GCGGGC": 3843, + "GCGGGG": 3844, + "GGAAAA": 3845, + "GGAAAT": 3846, + "GGAAAC": 3847, + "GGAAAG": 3848, + "GGAATA": 3849, + "GGAATT": 3850, + "GGAATC": 3851, + "GGAATG": 3852, + "GGAACA": 3853, + "GGAACT": 3854, + "GGAACC": 3855, + "GGAACG": 3856, + "GGAAGA": 3857, + "GGAAGT": 3858, + "GGAAGC": 3859, + "GGAAGG": 3860, + "GGATAA": 3861, + "GGATAT": 3862, + "GGATAC": 3863, + "GGATAG": 3864, + "GGATTA": 3865, + "GGATTT": 3866, + "GGATTC": 3867, + "GGATTG": 3868, + "GGATCA": 3869, + "GGATCT": 3870, + "GGATCC": 3871, + "GGATCG": 3872, + "GGATGA": 3873, + "GGATGT": 3874, + "GGATGC": 3875, + "GGATGG": 3876, + "GGACAA": 3877, + "GGACAT": 3878, + "GGACAC": 3879, + "GGACAG": 3880, + "GGACTA": 3881, + "GGACTT": 3882, + "GGACTC": 3883, + "GGACTG": 3884, + "GGACCA": 3885, + "GGACCT": 3886, + "GGACCC": 3887, + "GGACCG": 3888, + "GGACGA": 3889, + "GGACGT": 3890, + "GGACGC": 3891, + "GGACGG": 3892, + "GGAGAA": 3893, + "GGAGAT": 3894, + "GGAGAC": 3895, + "GGAGAG": 3896, + "GGAGTA": 3897, + "GGAGTT": 3898, + "GGAGTC": 3899, + "GGAGTG": 3900, + "GGAGCA": 3901, + "GGAGCT": 3902, + "GGAGCC": 3903, + "GGAGCG": 3904, + "GGAGGA": 3905, + "GGAGGT": 3906, + "GGAGGC": 3907, + "GGAGGG": 3908, + "GGTAAA": 3909, + "GGTAAT": 3910, + "GGTAAC": 3911, + "GGTAAG": 3912, + "GGTATA": 3913, + "GGTATT": 3914, + "GGTATC": 3915, + "GGTATG": 3916, + "GGTACA": 3917, + "GGTACT": 3918, + "GGTACC": 3919, + "GGTACG": 3920, + "GGTAGA": 3921, + "GGTAGT": 3922, + "GGTAGC": 3923, + "GGTAGG": 3924, + "GGTTAA": 3925, + "GGTTAT": 3926, + "GGTTAC": 3927, + "GGTTAG": 3928, + "GGTTTA": 3929, + "GGTTTT": 3930, + "GGTTTC": 3931, + "GGTTTG": 3932, + "GGTTCA": 3933, + "GGTTCT": 3934, + "GGTTCC": 3935, + "GGTTCG": 3936, + "GGTTGA": 3937, + "GGTTGT": 3938, + "GGTTGC": 3939, + "GGTTGG": 3940, + "GGTCAA": 3941, + "GGTCAT": 3942, + "GGTCAC": 3943, + "GGTCAG": 3944, + "GGTCTA": 3945, + "GGTCTT": 3946, + "GGTCTC": 3947, + "GGTCTG": 3948, + "GGTCCA": 3949, + "GGTCCT": 3950, + "GGTCCC": 3951, + "GGTCCG": 3952, + "GGTCGA": 3953, + "GGTCGT": 3954, + "GGTCGC": 3955, + "GGTCGG": 3956, + "GGTGAA": 3957, + "GGTGAT": 3958, + "GGTGAC": 3959, + "GGTGAG": 3960, + "GGTGTA": 3961, + "GGTGTT": 3962, + "GGTGTC": 3963, + "GGTGTG": 3964, + "GGTGCA": 3965, + "GGTGCT": 3966, + "GGTGCC": 3967, + "GGTGCG": 3968, + "GGTGGA": 3969, + "GGTGGT": 3970, + "GGTGGC": 3971, + "GGTGGG": 3972, + "GGCAAA": 3973, + "GGCAAT": 3974, + "GGCAAC": 3975, + "GGCAAG": 3976, + "GGCATA": 3977, + "GGCATT": 3978, + "GGCATC": 3979, + "GGCATG": 3980, + "GGCACA": 3981, + "GGCACT": 3982, + "GGCACC": 3983, + "GGCACG": 3984, + "GGCAGA": 3985, + "GGCAGT": 3986, + "GGCAGC": 3987, + "GGCAGG": 3988, + "GGCTAA": 3989, + "GGCTAT": 3990, + "GGCTAC": 3991, + "GGCTAG": 3992, + "GGCTTA": 3993, + "GGCTTT": 3994, + "GGCTTC": 3995, + "GGCTTG": 3996, + "GGCTCA": 3997, + "GGCTCT": 3998, + "GGCTCC": 3999, + "GGCTCG": 4000, + "GGCTGA": 4001, + "GGCTGT": 4002, + "GGCTGC": 4003, + "GGCTGG": 4004, + "GGCCAA": 4005, + "GGCCAT": 4006, + "GGCCAC": 4007, + "GGCCAG": 4008, + "GGCCTA": 4009, + "GGCCTT": 4010, + "GGCCTC": 4011, + "GGCCTG": 4012, + "GGCCCA": 4013, + "GGCCCT": 4014, + "GGCCCC": 4015, + "GGCCCG": 4016, + "GGCCGA": 4017, + "GGCCGT": 4018, + "GGCCGC": 4019, + "GGCCGG": 4020, + "GGCGAA": 4021, + "GGCGAT": 4022, + "GGCGAC": 4023, + "GGCGAG": 4024, + "GGCGTA": 4025, + "GGCGTT": 4026, + "GGCGTC": 4027, + "GGCGTG": 4028, + "GGCGCA": 4029, + "GGCGCT": 4030, + "GGCGCC": 4031, + "GGCGCG": 4032, + "GGCGGA": 4033, + "GGCGGT": 4034, + "GGCGGC": 4035, + "GGCGGG": 4036, + "GGGAAA": 4037, + "GGGAAT": 4038, + "GGGAAC": 4039, + "GGGAAG": 4040, + "GGGATA": 4041, + "GGGATT": 4042, + "GGGATC": 4043, + "GGGATG": 4044, + "GGGACA": 4045, + "GGGACT": 4046, + "GGGACC": 4047, + "GGGACG": 4048, + "GGGAGA": 4049, + "GGGAGT": 4050, + "GGGAGC": 4051, + "GGGAGG": 4052, + "GGGTAA": 4053, + "GGGTAT": 4054, + "GGGTAC": 4055, + "GGGTAG": 4056, + "GGGTTA": 4057, + "GGGTTT": 4058, + "GGGTTC": 4059, + "GGGTTG": 4060, + "GGGTCA": 4061, + "GGGTCT": 4062, + "GGGTCC": 4063, + "GGGTCG": 4064, + "GGGTGA": 4065, + "GGGTGT": 4066, + "GGGTGC": 4067, + "GGGTGG": 4068, + "GGGCAA": 4069, + "GGGCAT": 4070, + "GGGCAC": 4071, + "GGGCAG": 4072, + "GGGCTA": 4073, + "GGGCTT": 4074, + "GGGCTC": 4075, + "GGGCTG": 4076, + "GGGCCA": 4077, + "GGGCCT": 4078, + "GGGCCC": 4079, + "GGGCCG": 4080, + "GGGCGA": 4081, + "GGGCGT": 4082, + "GGGCGC": 4083, + "GGGCGG": 4084, + "GGGGAA": 4085, + "GGGGAT": 4086, + "GGGGAC": 4087, + "GGGGAG": 4088, + "GGGGTA": 4089, + "GGGGTT": 4090, + "GGGGTC": 4091, + "GGGGTG": 4092, + "GGGGCA": 4093, + "GGGGCT": 4094, + "GGGGCC": 4095, + "GGGGCG": 4096, + "GGGGGA": 4097, + "GGGGGT": 4098, + "GGGGGC": 4099, + "GGGGGG": 4100 + } + } +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/tokenizer_config.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..604e2b1b29af1dcb3920d2828b6a90276e5e3d02 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/tokenizer_config.json @@ -0,0 +1,58 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "[UNK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "[CLS]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "[SEP]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "4": { + "content": "[MASK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "clean_up_tokenization_spaces": true, + "cls_token": "[CLS]", + "do_basic_tokenize": true, + "do_lower_case": false, + "mask_token": "[MASK]", + "max_len": 512, + "model_max_length": 512, + "never_split": null, + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "strip_accents": null, + "tokenize_chinese_chars": true, + "tokenizer_class": "BertTokenizer", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/trainer_state.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..d85f28d01acb532510d2fa93ed8cc37764dd34f4 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/trainer_state.json @@ -0,0 +1,1783 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 25.0, + "eval_steps": 500, + "global_step": 6250, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.1, + "grad_norm": 0.2218121588230133, + "learning_rate": 5.319148936170213e-05, + "loss": 0.6927, + "step": 25 + }, + { + "epoch": 0.2, + "grad_norm": 0.44851255416870117, + "learning_rate": 0.00010638297872340425, + "loss": 0.6929, + "step": 50 + }, + { + "epoch": 0.3, + "grad_norm": 0.2440175861120224, + "learning_rate": 0.00015957446808510637, + "loss": 0.6924, + "step": 75 + }, + { + "epoch": 0.4, + "grad_norm": 0.3422006666660309, + "learning_rate": 0.0002127659574468085, + "loss": 0.6922, + "step": 100 + }, + { + "epoch": 0.5, + "grad_norm": 0.267333447933197, + "learning_rate": 0.00026595744680851064, + "loss": 0.6913, + "step": 125 + }, + { + "epoch": 0.6, + "grad_norm": 0.45699772238731384, + "learning_rate": 0.00031914893617021275, + "loss": 0.6896, + "step": 150 + }, + { + "epoch": 0.7, + "grad_norm": 0.3004947304725647, + "learning_rate": 0.0003723404255319149, + "loss": 0.6887, + "step": 175 + }, + { + "epoch": 0.8, + "grad_norm": 0.33440759778022766, + "learning_rate": 0.0003992081821181128, + "loss": 0.6857, + "step": 200 + }, + { + "epoch": 0.9, + "grad_norm": 0.2918928861618042, + "learning_rate": 0.0003975585615308479, + "loss": 0.6882, + "step": 225 + }, + { + "epoch": 1.0, + "grad_norm": 0.2934085726737976, + "learning_rate": 0.00039590894094358297, + "loss": 0.6807, + "step": 250 + }, + { + "epoch": 1.1, + "grad_norm": 0.3364830017089844, + "learning_rate": 0.0003942593203563181, + "loss": 0.6848, + "step": 275 + }, + { + "epoch": 1.2, + "grad_norm": 0.8379159569740295, + "learning_rate": 0.0003926096997690532, + "loss": 0.677, + "step": 300 + }, + { + "epoch": 1.3, + "grad_norm": 0.37671706080436707, + "learning_rate": 0.00039096007918178817, + "loss": 0.6744, + "step": 325 + }, + { + "epoch": 1.4, + "grad_norm": 0.34329330921173096, + "learning_rate": 0.0003893104585945233, + "loss": 0.6733, + "step": 350 + }, + { + "epoch": 1.5, + "grad_norm": 0.4286085367202759, + "learning_rate": 0.0003876608380072583, + "loss": 0.6772, + "step": 375 + }, + { + "epoch": 1.6, + "grad_norm": 0.3576306998729706, + "learning_rate": 0.00038601121741999343, + "loss": 0.6799, + "step": 400 + }, + { + "epoch": 1.7, + "grad_norm": 0.4636435806751251, + "learning_rate": 0.0003843615968327285, + "loss": 0.6819, + "step": 425 + }, + { + "epoch": 1.8, + "grad_norm": 0.4552135765552521, + "learning_rate": 0.0003827119762454636, + "loss": 0.6731, + "step": 450 + }, + { + "epoch": 1.9, + "grad_norm": 0.3565954566001892, + "learning_rate": 0.00038106235565819863, + "loss": 0.6718, + "step": 475 + }, + { + "epoch": 2.0, + "grad_norm": 0.5451686978340149, + "learning_rate": 0.0003794127350709337, + "loss": 0.6858, + "step": 500 + }, + { + "epoch": 2.1, + "grad_norm": 0.3730587959289551, + "learning_rate": 0.0003777631144836688, + "loss": 0.667, + "step": 525 + }, + { + "epoch": 2.2, + "grad_norm": 0.6642417907714844, + "learning_rate": 0.00037611349389640383, + "loss": 0.6662, + "step": 550 + }, + { + "epoch": 2.3, + "grad_norm": 0.3889809250831604, + "learning_rate": 0.00037446387330913894, + "loss": 0.6718, + "step": 575 + }, + { + "epoch": 2.4, + "grad_norm": 0.47203826904296875, + "learning_rate": 0.000372814252721874, + "loss": 0.6687, + "step": 600 + }, + { + "epoch": 2.5, + "grad_norm": 0.5293862819671631, + "learning_rate": 0.00037116463213460903, + "loss": 0.6655, + "step": 625 + }, + { + "epoch": 2.6, + "grad_norm": 0.36724531650543213, + "learning_rate": 0.00036951501154734414, + "loss": 0.672, + "step": 650 + }, + { + "epoch": 2.7, + "grad_norm": 0.5733087062835693, + "learning_rate": 0.0003678653909600792, + "loss": 0.6683, + "step": 675 + }, + { + "epoch": 2.8, + "grad_norm": 0.6385205984115601, + "learning_rate": 0.0003662157703728143, + "loss": 0.67, + "step": 700 + }, + { + "epoch": 2.9, + "grad_norm": 0.6146281957626343, + "learning_rate": 0.00036456614978554934, + "loss": 0.6606, + "step": 725 + }, + { + "epoch": 3.0, + "grad_norm": 0.6535729169845581, + "learning_rate": 0.00036291652919828444, + "loss": 0.6691, + "step": 750 + }, + { + "epoch": 3.1, + "grad_norm": 0.5194367170333862, + "learning_rate": 0.00036126690861101944, + "loss": 0.6517, + "step": 775 + }, + { + "epoch": 3.2, + "grad_norm": 0.7116692066192627, + "learning_rate": 0.00035961728802375454, + "loss": 0.6501, + "step": 800 + }, + { + "epoch": 3.3, + "grad_norm": 0.5338636040687561, + "learning_rate": 0.0003579676674364896, + "loss": 0.6682, + "step": 825 + }, + { + "epoch": 3.4, + "grad_norm": 0.6912877559661865, + "learning_rate": 0.0003563180468492247, + "loss": 0.6632, + "step": 850 + }, + { + "epoch": 3.5, + "grad_norm": 0.6925108432769775, + "learning_rate": 0.0003546684262619598, + "loss": 0.653, + "step": 875 + }, + { + "epoch": 3.6, + "grad_norm": 0.5360299944877625, + "learning_rate": 0.00035301880567469485, + "loss": 0.6613, + "step": 900 + }, + { + "epoch": 3.7, + "grad_norm": 0.6487969160079956, + "learning_rate": 0.0003513691850874299, + "loss": 0.6627, + "step": 925 + }, + { + "epoch": 3.8, + "grad_norm": 1.7260371446609497, + "learning_rate": 0.00034971956450016495, + "loss": 0.6436, + "step": 950 + }, + { + "epoch": 3.9, + "grad_norm": 0.5852354168891907, + "learning_rate": 0.00034806994391290005, + "loss": 0.6613, + "step": 975 + }, + { + "epoch": 4.0, + "grad_norm": 0.7802712917327881, + "learning_rate": 0.0003464203233256351, + "loss": 0.658, + "step": 1000 + }, + { + "epoch": 4.1, + "grad_norm": 0.7335640788078308, + "learning_rate": 0.0003447707027383702, + "loss": 0.6476, + "step": 1025 + }, + { + "epoch": 4.2, + "grad_norm": 0.9540443420410156, + "learning_rate": 0.00034312108215110525, + "loss": 0.6334, + "step": 1050 + }, + { + "epoch": 4.3, + "grad_norm": 1.3698925971984863, + "learning_rate": 0.0003414714615638403, + "loss": 0.6512, + "step": 1075 + }, + { + "epoch": 4.4, + "grad_norm": 0.8282333016395569, + "learning_rate": 0.0003398218409765754, + "loss": 0.6556, + "step": 1100 + }, + { + "epoch": 4.5, + "grad_norm": 1.4414807558059692, + "learning_rate": 0.00033817222038931045, + "loss": 0.6628, + "step": 1125 + }, + { + "epoch": 4.6, + "grad_norm": 0.5493957996368408, + "learning_rate": 0.00033652259980204556, + "loss": 0.6461, + "step": 1150 + }, + { + "epoch": 4.7, + "grad_norm": 1.2343393564224243, + "learning_rate": 0.0003348729792147806, + "loss": 0.6464, + "step": 1175 + }, + { + "epoch": 4.8, + "grad_norm": 0.7163398265838623, + "learning_rate": 0.0003332233586275157, + "loss": 0.6518, + "step": 1200 + }, + { + "epoch": 4.9, + "grad_norm": 0.7878938317298889, + "learning_rate": 0.00033157373804025076, + "loss": 0.6502, + "step": 1225 + }, + { + "epoch": 5.0, + "grad_norm": 0.952571451663971, + "learning_rate": 0.0003299241174529858, + "loss": 0.6463, + "step": 1250 + }, + { + "epoch": 5.1, + "grad_norm": 0.6535393595695496, + "learning_rate": 0.0003282744968657209, + "loss": 0.6361, + "step": 1275 + }, + { + "epoch": 5.2, + "grad_norm": 0.8239464163780212, + "learning_rate": 0.00032662487627845596, + "loss": 0.6264, + "step": 1300 + }, + { + "epoch": 5.3, + "grad_norm": 1.4500653743743896, + "learning_rate": 0.00032497525569119106, + "loss": 0.6544, + "step": 1325 + }, + { + "epoch": 5.4, + "grad_norm": 1.1515296697616577, + "learning_rate": 0.0003233256351039261, + "loss": 0.6629, + "step": 1350 + }, + { + "epoch": 5.5, + "grad_norm": 0.6899295449256897, + "learning_rate": 0.00032167601451666116, + "loss": 0.6453, + "step": 1375 + }, + { + "epoch": 5.6, + "grad_norm": 1.1782217025756836, + "learning_rate": 0.00032002639392939627, + "loss": 0.6362, + "step": 1400 + }, + { + "epoch": 5.7, + "grad_norm": 0.6971372961997986, + "learning_rate": 0.0003183767733421313, + "loss": 0.6593, + "step": 1425 + }, + { + "epoch": 5.8, + "grad_norm": 0.7293921709060669, + "learning_rate": 0.0003167271527548664, + "loss": 0.6318, + "step": 1450 + }, + { + "epoch": 5.9, + "grad_norm": 0.6860108375549316, + "learning_rate": 0.00031507753216760147, + "loss": 0.6402, + "step": 1475 + }, + { + "epoch": 6.0, + "grad_norm": 1.330344319343567, + "learning_rate": 0.00031342791158033657, + "loss": 0.6276, + "step": 1500 + }, + { + "epoch": 6.1, + "grad_norm": 0.7261562943458557, + "learning_rate": 0.00031177829099307157, + "loss": 0.6369, + "step": 1525 + }, + { + "epoch": 6.2, + "grad_norm": 1.1324542760849, + "learning_rate": 0.00031012867040580667, + "loss": 0.6228, + "step": 1550 + }, + { + "epoch": 6.3, + "grad_norm": 0.7002120614051819, + "learning_rate": 0.0003084790498185417, + "loss": 0.6389, + "step": 1575 + }, + { + "epoch": 6.4, + "grad_norm": 1.3143397569656372, + "learning_rate": 0.0003068294292312768, + "loss": 0.6362, + "step": 1600 + }, + { + "epoch": 6.5, + "grad_norm": 1.6375401020050049, + "learning_rate": 0.0003051798086440119, + "loss": 0.6348, + "step": 1625 + }, + { + "epoch": 6.6, + "grad_norm": 1.4343605041503906, + "learning_rate": 0.000303530188056747, + "loss": 0.6496, + "step": 1650 + }, + { + "epoch": 6.7, + "grad_norm": 0.815762996673584, + "learning_rate": 0.000301880567469482, + "loss": 0.6407, + "step": 1675 + }, + { + "epoch": 6.8, + "grad_norm": 1.068608045578003, + "learning_rate": 0.0003002309468822171, + "loss": 0.6417, + "step": 1700 + }, + { + "epoch": 6.9, + "grad_norm": 0.8775753974914551, + "learning_rate": 0.0002985813262949522, + "loss": 0.6281, + "step": 1725 + }, + { + "epoch": 7.0, + "grad_norm": 1.041817545890808, + "learning_rate": 0.00029693170570768723, + "loss": 0.6485, + "step": 1750 + }, + { + "epoch": 7.1, + "grad_norm": 0.7135840058326721, + "learning_rate": 0.00029528208512042233, + "loss": 0.6192, + "step": 1775 + }, + { + "epoch": 7.2, + "grad_norm": 1.5523930788040161, + "learning_rate": 0.0002936324645331574, + "loss": 0.636, + "step": 1800 + }, + { + "epoch": 7.3, + "grad_norm": 0.7677174210548401, + "learning_rate": 0.00029198284394589243, + "loss": 0.6213, + "step": 1825 + }, + { + "epoch": 7.4, + "grad_norm": 0.8086565732955933, + "learning_rate": 0.00029033322335862753, + "loss": 0.6292, + "step": 1850 + }, + { + "epoch": 7.5, + "grad_norm": 3.1128971576690674, + "learning_rate": 0.0002886836027713626, + "loss": 0.648, + "step": 1875 + }, + { + "epoch": 7.6, + "grad_norm": 0.9243882894515991, + "learning_rate": 0.0002870339821840977, + "loss": 0.6286, + "step": 1900 + }, + { + "epoch": 7.7, + "grad_norm": 1.8203461170196533, + "learning_rate": 0.00028538436159683273, + "loss": 0.6301, + "step": 1925 + }, + { + "epoch": 7.8, + "grad_norm": 1.0924643278121948, + "learning_rate": 0.00028373474100956784, + "loss": 0.6378, + "step": 1950 + }, + { + "epoch": 7.9, + "grad_norm": 1.0839996337890625, + "learning_rate": 0.0002820851204223029, + "loss": 0.6265, + "step": 1975 + }, + { + "epoch": 8.0, + "grad_norm": 0.7143226861953735, + "learning_rate": 0.00028043549983503794, + "loss": 0.649, + "step": 2000 + }, + { + "epoch": 8.1, + "grad_norm": 0.8778955936431885, + "learning_rate": 0.00027878587924777304, + "loss": 0.6411, + "step": 2025 + }, + { + "epoch": 8.2, + "grad_norm": 1.644904613494873, + "learning_rate": 0.0002771362586605081, + "loss": 0.6168, + "step": 2050 + }, + { + "epoch": 8.3, + "grad_norm": 0.7915159463882446, + "learning_rate": 0.0002754866380732432, + "loss": 0.6324, + "step": 2075 + }, + { + "epoch": 8.4, + "grad_norm": 1.1212801933288574, + "learning_rate": 0.00027383701748597824, + "loss": 0.6232, + "step": 2100 + }, + { + "epoch": 8.5, + "grad_norm": 0.7717731595039368, + "learning_rate": 0.0002721873968987133, + "loss": 0.624, + "step": 2125 + }, + { + "epoch": 8.6, + "grad_norm": 1.0577210187911987, + "learning_rate": 0.00027053777631144834, + "loss": 0.6296, + "step": 2150 + }, + { + "epoch": 8.7, + "grad_norm": 1.0531567335128784, + "learning_rate": 0.00026888815572418344, + "loss": 0.6345, + "step": 2175 + }, + { + "epoch": 8.8, + "grad_norm": 2.8983640670776367, + "learning_rate": 0.00026723853513691855, + "loss": 0.6142, + "step": 2200 + }, + { + "epoch": 8.9, + "grad_norm": 0.7773396968841553, + "learning_rate": 0.0002655889145496536, + "loss": 0.634, + "step": 2225 + }, + { + "epoch": 9.0, + "grad_norm": 1.0587151050567627, + "learning_rate": 0.0002639392939623887, + "loss": 0.6326, + "step": 2250 + }, + { + "epoch": 9.1, + "grad_norm": 0.8009796738624573, + "learning_rate": 0.0002622896733751237, + "loss": 0.6175, + "step": 2275 + }, + { + "epoch": 9.2, + "grad_norm": 1.2530399560928345, + "learning_rate": 0.0002606400527878588, + "loss": 0.6239, + "step": 2300 + }, + { + "epoch": 9.3, + "grad_norm": 0.7976368069648743, + "learning_rate": 0.00025899043220059385, + "loss": 0.6452, + "step": 2325 + }, + { + "epoch": 9.4, + "grad_norm": 1.455032229423523, + "learning_rate": 0.00025734081161332895, + "loss": 0.6351, + "step": 2350 + }, + { + "epoch": 9.5, + "grad_norm": 0.8916760087013245, + "learning_rate": 0.000255691191026064, + "loss": 0.5986, + "step": 2375 + }, + { + "epoch": 9.6, + "grad_norm": 1.104746699333191, + "learning_rate": 0.0002540415704387991, + "loss": 0.6104, + "step": 2400 + }, + { + "epoch": 9.7, + "grad_norm": 0.9449218511581421, + "learning_rate": 0.00025239194985153415, + "loss": 0.6262, + "step": 2425 + }, + { + "epoch": 9.8, + "grad_norm": 2.0286591053009033, + "learning_rate": 0.0002507423292642692, + "loss": 0.6124, + "step": 2450 + }, + { + "epoch": 9.9, + "grad_norm": 2.273597002029419, + "learning_rate": 0.0002490927086770043, + "loss": 0.6164, + "step": 2475 + }, + { + "epoch": 10.0, + "grad_norm": 0.8174235224723816, + "learning_rate": 0.00024744308808973936, + "loss": 0.6254, + "step": 2500 + }, + { + "epoch": 10.1, + "grad_norm": 1.0447920560836792, + "learning_rate": 0.00024579346750247446, + "loss": 0.6187, + "step": 2525 + }, + { + "epoch": 10.2, + "grad_norm": 1.0433655977249146, + "learning_rate": 0.0002441438469152095, + "loss": 0.6183, + "step": 2550 + }, + { + "epoch": 10.3, + "grad_norm": 1.6479028463363647, + "learning_rate": 0.00024249422632794456, + "loss": 0.6353, + "step": 2575 + }, + { + "epoch": 10.4, + "grad_norm": 0.8032572865486145, + "learning_rate": 0.00024084460574067963, + "loss": 0.6206, + "step": 2600 + }, + { + "epoch": 10.5, + "grad_norm": 1.0675058364868164, + "learning_rate": 0.0002391949851534147, + "loss": 0.6168, + "step": 2625 + }, + { + "epoch": 10.6, + "grad_norm": 1.2228609323501587, + "learning_rate": 0.0002375453645661498, + "loss": 0.6007, + "step": 2650 + }, + { + "epoch": 10.7, + "grad_norm": 1.1027027368545532, + "learning_rate": 0.00023589574397888486, + "loss": 0.6227, + "step": 2675 + }, + { + "epoch": 10.8, + "grad_norm": 1.2949633598327637, + "learning_rate": 0.00023424612339161997, + "loss": 0.6179, + "step": 2700 + }, + { + "epoch": 10.9, + "grad_norm": 1.235032320022583, + "learning_rate": 0.000232596502804355, + "loss": 0.6216, + "step": 2725 + }, + { + "epoch": 11.0, + "grad_norm": 1.0087182521820068, + "learning_rate": 0.00023094688221709007, + "loss": 0.6286, + "step": 2750 + }, + { + "epoch": 11.1, + "grad_norm": 1.301207423210144, + "learning_rate": 0.00022929726162982514, + "loss": 0.5953, + "step": 2775 + }, + { + "epoch": 11.2, + "grad_norm": 1.6798988580703735, + "learning_rate": 0.00022764764104256022, + "loss": 0.6331, + "step": 2800 + }, + { + "epoch": 11.3, + "grad_norm": 1.1364004611968994, + "learning_rate": 0.0002259980204552953, + "loss": 0.6047, + "step": 2825 + }, + { + "epoch": 11.4, + "grad_norm": 0.9963313341140747, + "learning_rate": 0.00022434839986803037, + "loss": 0.6192, + "step": 2850 + }, + { + "epoch": 11.5, + "grad_norm": 1.5672775506973267, + "learning_rate": 0.00022269877928076542, + "loss": 0.6063, + "step": 2875 + }, + { + "epoch": 11.6, + "grad_norm": 1.4239299297332764, + "learning_rate": 0.0002210491586935005, + "loss": 0.6058, + "step": 2900 + }, + { + "epoch": 11.7, + "grad_norm": 1.5440959930419922, + "learning_rate": 0.00021939953810623557, + "loss": 0.6239, + "step": 2925 + }, + { + "epoch": 11.8, + "grad_norm": 1.35140860080719, + "learning_rate": 0.00021774991751897065, + "loss": 0.6241, + "step": 2950 + }, + { + "epoch": 11.9, + "grad_norm": 1.9355136156082153, + "learning_rate": 0.00021610029693170573, + "loss": 0.6135, + "step": 2975 + }, + { + "epoch": 12.0, + "grad_norm": 1.134037733078003, + "learning_rate": 0.0002144506763444408, + "loss": 0.6142, + "step": 3000 + }, + { + "epoch": 12.1, + "grad_norm": 1.094366192817688, + "learning_rate": 0.00021280105575717582, + "loss": 0.6226, + "step": 3025 + }, + { + "epoch": 12.2, + "grad_norm": 1.1197398900985718, + "learning_rate": 0.00021115143516991093, + "loss": 0.6073, + "step": 3050 + }, + { + "epoch": 12.3, + "grad_norm": 1.240188717842102, + "learning_rate": 0.000209501814582646, + "loss": 0.6265, + "step": 3075 + }, + { + "epoch": 12.4, + "grad_norm": 0.9774200916290283, + "learning_rate": 0.00020785219399538108, + "loss": 0.6361, + "step": 3100 + }, + { + "epoch": 12.5, + "grad_norm": 0.902847409248352, + "learning_rate": 0.00020620257340811616, + "loss": 0.6129, + "step": 3125 + }, + { + "epoch": 12.6, + "grad_norm": 0.8658815622329712, + "learning_rate": 0.00020455295282085123, + "loss": 0.5953, + "step": 3150 + }, + { + "epoch": 12.7, + "grad_norm": 1.4172385931015015, + "learning_rate": 0.00020290333223358626, + "loss": 0.6158, + "step": 3175 + }, + { + "epoch": 12.8, + "grad_norm": 2.6393818855285645, + "learning_rate": 0.00020125371164632133, + "loss": 0.5978, + "step": 3200 + }, + { + "epoch": 12.9, + "grad_norm": 2.4407801628112793, + "learning_rate": 0.0001996040910590564, + "loss": 0.6178, + "step": 3225 + }, + { + "epoch": 13.0, + "grad_norm": 1.2797659635543823, + "learning_rate": 0.00019795447047179148, + "loss": 0.5849, + "step": 3250 + }, + { + "epoch": 13.1, + "grad_norm": 1.590224266052246, + "learning_rate": 0.0001963048498845266, + "loss": 0.5983, + "step": 3275 + }, + { + "epoch": 13.2, + "grad_norm": 1.155685544013977, + "learning_rate": 0.00019465522929726164, + "loss": 0.6301, + "step": 3300 + }, + { + "epoch": 13.3, + "grad_norm": 1.701331377029419, + "learning_rate": 0.00019300560870999671, + "loss": 0.5865, + "step": 3325 + }, + { + "epoch": 13.4, + "grad_norm": 1.2553105354309082, + "learning_rate": 0.0001913559881227318, + "loss": 0.6113, + "step": 3350 + }, + { + "epoch": 13.5, + "grad_norm": 1.495653510093689, + "learning_rate": 0.00018970636753546684, + "loss": 0.6001, + "step": 3375 + }, + { + "epoch": 13.6, + "grad_norm": 1.2800766229629517, + "learning_rate": 0.00018805674694820192, + "loss": 0.6013, + "step": 3400 + }, + { + "epoch": 13.7, + "grad_norm": 1.0035955905914307, + "learning_rate": 0.000186407126360937, + "loss": 0.621, + "step": 3425 + }, + { + "epoch": 13.8, + "grad_norm": 1.0515904426574707, + "learning_rate": 0.00018475750577367207, + "loss": 0.6297, + "step": 3450 + }, + { + "epoch": 13.9, + "grad_norm": 2.071884870529175, + "learning_rate": 0.00018310788518640715, + "loss": 0.6004, + "step": 3475 + }, + { + "epoch": 14.0, + "grad_norm": 1.4622681140899658, + "learning_rate": 0.00018145826459914222, + "loss": 0.6021, + "step": 3500 + }, + { + "epoch": 14.1, + "grad_norm": 1.1860116720199585, + "learning_rate": 0.00017980864401187727, + "loss": 0.602, + "step": 3525 + }, + { + "epoch": 14.2, + "grad_norm": 1.0958752632141113, + "learning_rate": 0.00017815902342461235, + "loss": 0.6084, + "step": 3550 + }, + { + "epoch": 14.3, + "grad_norm": 1.4344367980957031, + "learning_rate": 0.00017650940283734742, + "loss": 0.6332, + "step": 3575 + }, + { + "epoch": 14.4, + "grad_norm": 1.1095603704452515, + "learning_rate": 0.00017485978225008247, + "loss": 0.5945, + "step": 3600 + }, + { + "epoch": 14.5, + "grad_norm": 1.416272521018982, + "learning_rate": 0.00017321016166281755, + "loss": 0.5964, + "step": 3625 + }, + { + "epoch": 14.6, + "grad_norm": 0.9727759957313538, + "learning_rate": 0.00017156054107555263, + "loss": 0.6231, + "step": 3650 + }, + { + "epoch": 14.7, + "grad_norm": 0.9979771375656128, + "learning_rate": 0.0001699109204882877, + "loss": 0.5849, + "step": 3675 + }, + { + "epoch": 14.8, + "grad_norm": 1.2574740648269653, + "learning_rate": 0.00016826129990102278, + "loss": 0.6076, + "step": 3700 + }, + { + "epoch": 14.9, + "grad_norm": 1.1657873392105103, + "learning_rate": 0.00016661167931375785, + "loss": 0.6111, + "step": 3725 + }, + { + "epoch": 15.0, + "grad_norm": 1.767681360244751, + "learning_rate": 0.0001649620587264929, + "loss": 0.6025, + "step": 3750 + }, + { + "epoch": 15.1, + "grad_norm": 1.3131200075149536, + "learning_rate": 0.00016331243813922798, + "loss": 0.6039, + "step": 3775 + }, + { + "epoch": 15.2, + "grad_norm": 1.1376745700836182, + "learning_rate": 0.00016166281755196306, + "loss": 0.5995, + "step": 3800 + }, + { + "epoch": 15.3, + "grad_norm": 1.2298001050949097, + "learning_rate": 0.00016001319696469813, + "loss": 0.602, + "step": 3825 + }, + { + "epoch": 15.4, + "grad_norm": 2.735769271850586, + "learning_rate": 0.0001583635763774332, + "loss": 0.5902, + "step": 3850 + }, + { + "epoch": 15.5, + "grad_norm": 1.3570619821548462, + "learning_rate": 0.00015671395579016829, + "loss": 0.5963, + "step": 3875 + }, + { + "epoch": 15.6, + "grad_norm": 1.6737253665924072, + "learning_rate": 0.00015506433520290334, + "loss": 0.6002, + "step": 3900 + }, + { + "epoch": 15.7, + "grad_norm": 1.386838674545288, + "learning_rate": 0.0001534147146156384, + "loss": 0.5905, + "step": 3925 + }, + { + "epoch": 15.8, + "grad_norm": 1.670160174369812, + "learning_rate": 0.0001517650940283735, + "loss": 0.6, + "step": 3950 + }, + { + "epoch": 15.9, + "grad_norm": 1.1611871719360352, + "learning_rate": 0.00015011547344110854, + "loss": 0.616, + "step": 3975 + }, + { + "epoch": 16.0, + "grad_norm": 0.9084995985031128, + "learning_rate": 0.00014846585285384361, + "loss": 0.6036, + "step": 4000 + }, + { + "epoch": 16.1, + "grad_norm": 2.1180953979492188, + "learning_rate": 0.0001468162322665787, + "loss": 0.5896, + "step": 4025 + }, + { + "epoch": 16.2, + "grad_norm": 1.9384114742279053, + "learning_rate": 0.00014516661167931377, + "loss": 0.584, + "step": 4050 + }, + { + "epoch": 16.3, + "grad_norm": 1.9472033977508545, + "learning_rate": 0.00014351699109204884, + "loss": 0.6059, + "step": 4075 + }, + { + "epoch": 16.4, + "grad_norm": 1.6221973896026611, + "learning_rate": 0.00014186737050478392, + "loss": 0.5934, + "step": 4100 + }, + { + "epoch": 16.5, + "grad_norm": 2.1453840732574463, + "learning_rate": 0.00014021774991751897, + "loss": 0.5846, + "step": 4125 + }, + { + "epoch": 16.6, + "grad_norm": 2.0782151222229004, + "learning_rate": 0.00013856812933025404, + "loss": 0.6131, + "step": 4150 + }, + { + "epoch": 16.7, + "grad_norm": 1.437988519668579, + "learning_rate": 0.00013691850874298912, + "loss": 0.6082, + "step": 4175 + }, + { + "epoch": 16.8, + "grad_norm": 1.7626062631607056, + "learning_rate": 0.00013526888815572417, + "loss": 0.5928, + "step": 4200 + }, + { + "epoch": 16.9, + "grad_norm": 1.0071998834609985, + "learning_rate": 0.00013361926756845927, + "loss": 0.6019, + "step": 4225 + }, + { + "epoch": 17.0, + "grad_norm": 1.1568998098373413, + "learning_rate": 0.00013196964698119435, + "loss": 0.6196, + "step": 4250 + }, + { + "epoch": 17.1, + "grad_norm": 1.9417431354522705, + "learning_rate": 0.0001303200263939294, + "loss": 0.607, + "step": 4275 + }, + { + "epoch": 17.2, + "grad_norm": 1.0894469022750854, + "learning_rate": 0.00012867040580666448, + "loss": 0.5874, + "step": 4300 + }, + { + "epoch": 17.3, + "grad_norm": 1.3170983791351318, + "learning_rate": 0.00012702078521939955, + "loss": 0.5988, + "step": 4325 + }, + { + "epoch": 17.4, + "grad_norm": 1.0083190202713013, + "learning_rate": 0.0001253711646321346, + "loss": 0.6021, + "step": 4350 + }, + { + "epoch": 17.5, + "grad_norm": 1.24309504032135, + "learning_rate": 0.00012372154404486968, + "loss": 0.6097, + "step": 4375 + }, + { + "epoch": 17.6, + "grad_norm": 1.5258588790893555, + "learning_rate": 0.00012207192345760475, + "loss": 0.5891, + "step": 4400 + }, + { + "epoch": 17.7, + "grad_norm": 2.491304636001587, + "learning_rate": 0.00012042230287033982, + "loss": 0.6049, + "step": 4425 + }, + { + "epoch": 17.8, + "grad_norm": 1.737715482711792, + "learning_rate": 0.0001187726822830749, + "loss": 0.5888, + "step": 4450 + }, + { + "epoch": 17.9, + "grad_norm": 2.217761754989624, + "learning_rate": 0.00011712306169580998, + "loss": 0.6075, + "step": 4475 + }, + { + "epoch": 18.0, + "grad_norm": 1.4966678619384766, + "learning_rate": 0.00011547344110854503, + "loss": 0.5962, + "step": 4500 + }, + { + "epoch": 18.1, + "grad_norm": 1.1560686826705933, + "learning_rate": 0.00011382382052128011, + "loss": 0.5654, + "step": 4525 + }, + { + "epoch": 18.2, + "grad_norm": 1.1144013404846191, + "learning_rate": 0.00011217419993401519, + "loss": 0.6001, + "step": 4550 + }, + { + "epoch": 18.3, + "grad_norm": 5.605530738830566, + "learning_rate": 0.00011052457934675025, + "loss": 0.5977, + "step": 4575 + }, + { + "epoch": 18.4, + "grad_norm": 1.0827083587646484, + "learning_rate": 0.00010887495875948532, + "loss": 0.5954, + "step": 4600 + }, + { + "epoch": 18.5, + "grad_norm": 3.0124075412750244, + "learning_rate": 0.0001072253381722204, + "loss": 0.5921, + "step": 4625 + }, + { + "epoch": 18.6, + "grad_norm": 1.5420926809310913, + "learning_rate": 0.00010557571758495546, + "loss": 0.5832, + "step": 4650 + }, + { + "epoch": 18.7, + "grad_norm": 1.5770180225372314, + "learning_rate": 0.00010392609699769054, + "loss": 0.6138, + "step": 4675 + }, + { + "epoch": 18.8, + "grad_norm": 1.9862656593322754, + "learning_rate": 0.00010227647641042562, + "loss": 0.595, + "step": 4700 + }, + { + "epoch": 18.9, + "grad_norm": 1.6087626218795776, + "learning_rate": 0.00010062685582316067, + "loss": 0.5965, + "step": 4725 + }, + { + "epoch": 19.0, + "grad_norm": 0.9878010749816895, + "learning_rate": 9.897723523589574e-05, + "loss": 0.5807, + "step": 4750 + }, + { + "epoch": 19.1, + "grad_norm": 1.3701050281524658, + "learning_rate": 9.732761464863082e-05, + "loss": 0.5953, + "step": 4775 + }, + { + "epoch": 19.2, + "grad_norm": 1.4488855600357056, + "learning_rate": 9.56779940613659e-05, + "loss": 0.578, + "step": 4800 + }, + { + "epoch": 19.3, + "grad_norm": 1.8450454473495483, + "learning_rate": 9.402837347410096e-05, + "loss": 0.5734, + "step": 4825 + }, + { + "epoch": 19.4, + "grad_norm": 1.2953295707702637, + "learning_rate": 9.237875288683603e-05, + "loss": 0.6188, + "step": 4850 + }, + { + "epoch": 19.5, + "grad_norm": 4.279994010925293, + "learning_rate": 9.072913229957111e-05, + "loss": 0.5712, + "step": 4875 + }, + { + "epoch": 19.6, + "grad_norm": 1.4677122831344604, + "learning_rate": 8.907951171230617e-05, + "loss": 0.5921, + "step": 4900 + }, + { + "epoch": 19.7, + "grad_norm": 2.559695243835449, + "learning_rate": 8.742989112504124e-05, + "loss": 0.599, + "step": 4925 + }, + { + "epoch": 19.8, + "grad_norm": 1.063706874847412, + "learning_rate": 8.578027053777631e-05, + "loss": 0.603, + "step": 4950 + }, + { + "epoch": 19.9, + "grad_norm": 1.1021010875701904, + "learning_rate": 8.413064995051139e-05, + "loss": 0.5841, + "step": 4975 + }, + { + "epoch": 20.0, + "grad_norm": 1.2500898838043213, + "learning_rate": 8.248102936324645e-05, + "loss": 0.5946, + "step": 5000 + }, + { + "epoch": 20.1, + "grad_norm": 1.2104499340057373, + "learning_rate": 8.083140877598153e-05, + "loss": 0.5906, + "step": 5025 + }, + { + "epoch": 20.2, + "grad_norm": 2.4167215824127197, + "learning_rate": 7.91817881887166e-05, + "loss": 0.5843, + "step": 5050 + }, + { + "epoch": 20.3, + "grad_norm": 1.2537106275558472, + "learning_rate": 7.753216760145167e-05, + "loss": 0.5924, + "step": 5075 + }, + { + "epoch": 20.4, + "grad_norm": 1.244513750076294, + "learning_rate": 7.588254701418674e-05, + "loss": 0.6031, + "step": 5100 + }, + { + "epoch": 20.5, + "grad_norm": 2.094501256942749, + "learning_rate": 7.423292642692181e-05, + "loss": 0.5909, + "step": 5125 + }, + { + "epoch": 20.6, + "grad_norm": 0.9758375287055969, + "learning_rate": 7.258330583965688e-05, + "loss": 0.5736, + "step": 5150 + }, + { + "epoch": 20.7, + "grad_norm": 1.5794312953948975, + "learning_rate": 7.093368525239196e-05, + "loss": 0.5817, + "step": 5175 + }, + { + "epoch": 20.8, + "grad_norm": 2.009289264678955, + "learning_rate": 6.928406466512702e-05, + "loss": 0.5912, + "step": 5200 + }, + { + "epoch": 20.9, + "grad_norm": 1.850706696510315, + "learning_rate": 6.763444407786209e-05, + "loss": 0.5851, + "step": 5225 + }, + { + "epoch": 21.0, + "grad_norm": 1.3678967952728271, + "learning_rate": 6.598482349059718e-05, + "loss": 0.5884, + "step": 5250 + }, + { + "epoch": 21.1, + "grad_norm": 1.3373990058898926, + "learning_rate": 6.433520290333224e-05, + "loss": 0.6052, + "step": 5275 + }, + { + "epoch": 21.2, + "grad_norm": 1.510580062866211, + "learning_rate": 6.26855823160673e-05, + "loss": 0.6049, + "step": 5300 + }, + { + "epoch": 21.3, + "grad_norm": 1.514527678489685, + "learning_rate": 6.103596172880238e-05, + "loss": 0.5617, + "step": 5325 + }, + { + "epoch": 21.4, + "grad_norm": 1.1603538990020752, + "learning_rate": 5.938634114153745e-05, + "loss": 0.5852, + "step": 5350 + }, + { + "epoch": 21.5, + "grad_norm": 2.2692604064941406, + "learning_rate": 5.7736720554272516e-05, + "loss": 0.5863, + "step": 5375 + }, + { + "epoch": 21.6, + "grad_norm": 2.5274813175201416, + "learning_rate": 5.608709996700759e-05, + "loss": 0.5857, + "step": 5400 + }, + { + "epoch": 21.7, + "grad_norm": 2.518244504928589, + "learning_rate": 5.443747937974266e-05, + "loss": 0.5899, + "step": 5425 + }, + { + "epoch": 21.8, + "grad_norm": 3.0091376304626465, + "learning_rate": 5.278785879247773e-05, + "loss": 0.5882, + "step": 5450 + }, + { + "epoch": 21.9, + "grad_norm": 1.8518242835998535, + "learning_rate": 5.113823820521281e-05, + "loss": 0.5515, + "step": 5475 + }, + { + "epoch": 22.0, + "grad_norm": 1.4333727359771729, + "learning_rate": 4.948861761794787e-05, + "loss": 0.5748, + "step": 5500 + }, + { + "epoch": 22.1, + "grad_norm": 1.7678550481796265, + "learning_rate": 4.783899703068295e-05, + "loss": 0.5745, + "step": 5525 + }, + { + "epoch": 22.2, + "grad_norm": 1.4223382472991943, + "learning_rate": 4.618937644341802e-05, + "loss": 0.5795, + "step": 5550 + }, + { + "epoch": 22.3, + "grad_norm": 2.587275505065918, + "learning_rate": 4.453975585615309e-05, + "loss": 0.5797, + "step": 5575 + }, + { + "epoch": 22.4, + "grad_norm": 1.3082376718521118, + "learning_rate": 4.2890135268888156e-05, + "loss": 0.6038, + "step": 5600 + }, + { + "epoch": 22.5, + "grad_norm": 1.264905333518982, + "learning_rate": 4.1240514681623226e-05, + "loss": 0.5558, + "step": 5625 + }, + { + "epoch": 22.6, + "grad_norm": 1.301184058189392, + "learning_rate": 3.95908940943583e-05, + "loss": 0.6024, + "step": 5650 + }, + { + "epoch": 22.7, + "grad_norm": 3.1338558197021484, + "learning_rate": 3.794127350709337e-05, + "loss": 0.5887, + "step": 5675 + }, + { + "epoch": 22.8, + "grad_norm": 1.5282409191131592, + "learning_rate": 3.629165291982844e-05, + "loss": 0.5874, + "step": 5700 + }, + { + "epoch": 22.9, + "grad_norm": 1.4762812852859497, + "learning_rate": 3.464203233256351e-05, + "loss": 0.5764, + "step": 5725 + }, + { + "epoch": 23.0, + "grad_norm": 4.048628330230713, + "learning_rate": 3.299241174529859e-05, + "loss": 0.571, + "step": 5750 + }, + { + "epoch": 23.1, + "grad_norm": 1.2434217929840088, + "learning_rate": 3.134279115803365e-05, + "loss": 0.5735, + "step": 5775 + }, + { + "epoch": 23.2, + "grad_norm": 1.442742943763733, + "learning_rate": 2.9693170570768723e-05, + "loss": 0.5957, + "step": 5800 + }, + { + "epoch": 23.3, + "grad_norm": 2.088374137878418, + "learning_rate": 2.8043549983503796e-05, + "loss": 0.6164, + "step": 5825 + }, + { + "epoch": 23.4, + "grad_norm": 1.1105170249938965, + "learning_rate": 2.6393929396238866e-05, + "loss": 0.5732, + "step": 5850 + }, + { + "epoch": 23.5, + "grad_norm": 1.6173477172851562, + "learning_rate": 2.4744308808973936e-05, + "loss": 0.5698, + "step": 5875 + }, + { + "epoch": 23.6, + "grad_norm": 1.760382056236267, + "learning_rate": 2.309468822170901e-05, + "loss": 0.565, + "step": 5900 + }, + { + "epoch": 23.7, + "grad_norm": 2.0124967098236084, + "learning_rate": 2.1445067634444078e-05, + "loss": 0.5837, + "step": 5925 + }, + { + "epoch": 23.8, + "grad_norm": 1.8501543998718262, + "learning_rate": 1.979544704717915e-05, + "loss": 0.555, + "step": 5950 + }, + { + "epoch": 23.9, + "grad_norm": 1.4653395414352417, + "learning_rate": 1.814582645991422e-05, + "loss": 0.5742, + "step": 5975 + }, + { + "epoch": 24.0, + "grad_norm": 2.650632381439209, + "learning_rate": 1.6496205872649294e-05, + "loss": 0.5917, + "step": 6000 + }, + { + "epoch": 24.1, + "grad_norm": 1.68106210231781, + "learning_rate": 1.4846585285384362e-05, + "loss": 0.5842, + "step": 6025 + }, + { + "epoch": 24.2, + "grad_norm": 2.0407559871673584, + "learning_rate": 1.3196964698119433e-05, + "loss": 0.6026, + "step": 6050 + }, + { + "epoch": 24.3, + "grad_norm": 1.1904737949371338, + "learning_rate": 1.1547344110854504e-05, + "loss": 0.5827, + "step": 6075 + }, + { + "epoch": 24.4, + "grad_norm": 1.5794881582260132, + "learning_rate": 9.897723523589576e-06, + "loss": 0.5904, + "step": 6100 + }, + { + "epoch": 24.5, + "grad_norm": 1.7687537670135498, + "learning_rate": 8.248102936324647e-06, + "loss": 0.5957, + "step": 6125 + }, + { + "epoch": 24.6, + "grad_norm": 1.1084353923797607, + "learning_rate": 6.5984823490597165e-06, + "loss": 0.5739, + "step": 6150 + }, + { + "epoch": 24.7, + "grad_norm": 2.721799850463867, + "learning_rate": 4.948861761794788e-06, + "loss": 0.5836, + "step": 6175 + }, + { + "epoch": 24.8, + "grad_norm": 1.1731759309768677, + "learning_rate": 3.2992411745298582e-06, + "loss": 0.5806, + "step": 6200 + }, + { + "epoch": 24.9, + "grad_norm": 3.0228796005249023, + "learning_rate": 1.6496205872649291e-06, + "loss": 0.5862, + "step": 6225 + }, + { + "epoch": 25.0, + "grad_norm": 1.4986634254455566, + "learning_rate": 0.0, + "loss": 0.5917, + "step": 6250 + } + ], + "logging_steps": 25, + "max_steps": 6250, + "num_input_tokens_seen": 0, + "num_train_epochs": 25, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 346131456000000.0, + "train_batch_size": 20, + "trial_name": null, + "trial_params": null +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/training_args.bin b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..540db3fbee786996d9705d768d31ba6c371187be --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dc7816b5f3fd38f01f9ba38a197ce070cde06e4c90e5260cb4814f34f628907d +size 5304 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/vocab.txt b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/vocab.txt new file mode 100644 index 0000000000000000000000000000000000000000..cc9ad53732e21ea9180f18e1fb67a38cc5702d4c --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_attention_plus_feed_forward/checkpoint-6250/vocab.txt @@ -0,0 +1,4101 @@ +[PAD] +[UNK] +[CLS] +[SEP] +[MASK] +AAAAAA +AAAAAT +AAAAAC +AAAAAG +AAAATA +AAAATT +AAAATC +AAAATG +AAAACA +AAAACT +AAAACC +AAAACG +AAAAGA +AAAAGT +AAAAGC +AAAAGG +AAATAA +AAATAT +AAATAC +AAATAG +AAATTA +AAATTT +AAATTC +AAATTG +AAATCA +AAATCT +AAATCC +AAATCG +AAATGA +AAATGT +AAATGC +AAATGG +AAACAA +AAACAT +AAACAC +AAACAG +AAACTA +AAACTT +AAACTC +AAACTG +AAACCA +AAACCT +AAACCC +AAACCG +AAACGA +AAACGT +AAACGC +AAACGG +AAAGAA +AAAGAT +AAAGAC +AAAGAG +AAAGTA +AAAGTT +AAAGTC +AAAGTG +AAAGCA +AAAGCT +AAAGCC +AAAGCG +AAAGGA +AAAGGT +AAAGGC +AAAGGG +AATAAA +AATAAT +AATAAC +AATAAG +AATATA +AATATT +AATATC +AATATG +AATACA +AATACT +AATACC +AATACG +AATAGA +AATAGT +AATAGC +AATAGG +AATTAA +AATTAT +AATTAC +AATTAG +AATTTA +AATTTT +AATTTC +AATTTG +AATTCA +AATTCT +AATTCC +AATTCG +AATTGA +AATTGT +AATTGC +AATTGG +AATCAA +AATCAT +AATCAC +AATCAG +AATCTA +AATCTT +AATCTC +AATCTG +AATCCA +AATCCT +AATCCC +AATCCG +AATCGA +AATCGT +AATCGC +AATCGG +AATGAA +AATGAT +AATGAC +AATGAG +AATGTA +AATGTT +AATGTC +AATGTG +AATGCA +AATGCT +AATGCC +AATGCG +AATGGA +AATGGT +AATGGC +AATGGG +AACAAA +AACAAT +AACAAC +AACAAG +AACATA +AACATT +AACATC +AACATG +AACACA +AACACT +AACACC +AACACG +AACAGA +AACAGT +AACAGC +AACAGG +AACTAA +AACTAT +AACTAC +AACTAG +AACTTA +AACTTT +AACTTC +AACTTG +AACTCA +AACTCT +AACTCC +AACTCG +AACTGA +AACTGT +AACTGC +AACTGG +AACCAA +AACCAT +AACCAC +AACCAG +AACCTA +AACCTT +AACCTC +AACCTG +AACCCA +AACCCT +AACCCC +AACCCG +AACCGA +AACCGT +AACCGC +AACCGG +AACGAA +AACGAT +AACGAC +AACGAG +AACGTA +AACGTT +AACGTC +AACGTG +AACGCA +AACGCT +AACGCC +AACGCG +AACGGA +AACGGT +AACGGC +AACGGG +AAGAAA +AAGAAT +AAGAAC +AAGAAG +AAGATA +AAGATT +AAGATC +AAGATG +AAGACA +AAGACT +AAGACC +AAGACG +AAGAGA +AAGAGT +AAGAGC +AAGAGG +AAGTAA +AAGTAT +AAGTAC +AAGTAG +AAGTTA +AAGTTT +AAGTTC +AAGTTG +AAGTCA +AAGTCT +AAGTCC +AAGTCG +AAGTGA +AAGTGT +AAGTGC +AAGTGG +AAGCAA +AAGCAT +AAGCAC +AAGCAG +AAGCTA +AAGCTT +AAGCTC +AAGCTG +AAGCCA +AAGCCT +AAGCCC +AAGCCG +AAGCGA +AAGCGT +AAGCGC +AAGCGG +AAGGAA +AAGGAT +AAGGAC +AAGGAG +AAGGTA +AAGGTT +AAGGTC +AAGGTG +AAGGCA +AAGGCT +AAGGCC +AAGGCG +AAGGGA +AAGGGT +AAGGGC +AAGGGG +ATAAAA +ATAAAT +ATAAAC +ATAAAG +ATAATA +ATAATT +ATAATC +ATAATG +ATAACA +ATAACT +ATAACC +ATAACG +ATAAGA +ATAAGT +ATAAGC +ATAAGG +ATATAA +ATATAT +ATATAC +ATATAG +ATATTA +ATATTT +ATATTC +ATATTG +ATATCA +ATATCT +ATATCC +ATATCG +ATATGA +ATATGT +ATATGC +ATATGG +ATACAA +ATACAT +ATACAC +ATACAG +ATACTA +ATACTT +ATACTC +ATACTG +ATACCA +ATACCT +ATACCC +ATACCG +ATACGA +ATACGT +ATACGC +ATACGG +ATAGAA +ATAGAT +ATAGAC +ATAGAG +ATAGTA +ATAGTT +ATAGTC +ATAGTG +ATAGCA +ATAGCT +ATAGCC +ATAGCG +ATAGGA +ATAGGT +ATAGGC +ATAGGG +ATTAAA +ATTAAT +ATTAAC +ATTAAG +ATTATA +ATTATT +ATTATC +ATTATG +ATTACA +ATTACT +ATTACC +ATTACG +ATTAGA +ATTAGT +ATTAGC +ATTAGG +ATTTAA +ATTTAT +ATTTAC +ATTTAG +ATTTTA +ATTTTT +ATTTTC +ATTTTG +ATTTCA +ATTTCT +ATTTCC +ATTTCG +ATTTGA +ATTTGT +ATTTGC +ATTTGG +ATTCAA +ATTCAT +ATTCAC +ATTCAG +ATTCTA +ATTCTT +ATTCTC +ATTCTG +ATTCCA +ATTCCT +ATTCCC +ATTCCG +ATTCGA +ATTCGT +ATTCGC +ATTCGG +ATTGAA +ATTGAT +ATTGAC +ATTGAG +ATTGTA +ATTGTT +ATTGTC +ATTGTG +ATTGCA +ATTGCT +ATTGCC +ATTGCG +ATTGGA +ATTGGT +ATTGGC +ATTGGG +ATCAAA +ATCAAT +ATCAAC +ATCAAG +ATCATA +ATCATT +ATCATC +ATCATG +ATCACA +ATCACT +ATCACC +ATCACG +ATCAGA +ATCAGT +ATCAGC +ATCAGG +ATCTAA +ATCTAT +ATCTAC +ATCTAG +ATCTTA +ATCTTT +ATCTTC +ATCTTG +ATCTCA +ATCTCT +ATCTCC +ATCTCG +ATCTGA +ATCTGT +ATCTGC +ATCTGG +ATCCAA +ATCCAT +ATCCAC +ATCCAG +ATCCTA +ATCCTT +ATCCTC +ATCCTG +ATCCCA +ATCCCT +ATCCCC +ATCCCG +ATCCGA +ATCCGT +ATCCGC +ATCCGG +ATCGAA +ATCGAT +ATCGAC +ATCGAG +ATCGTA +ATCGTT +ATCGTC +ATCGTG +ATCGCA +ATCGCT +ATCGCC +ATCGCG +ATCGGA +ATCGGT +ATCGGC +ATCGGG +ATGAAA +ATGAAT +ATGAAC +ATGAAG +ATGATA +ATGATT +ATGATC +ATGATG +ATGACA +ATGACT +ATGACC +ATGACG +ATGAGA +ATGAGT +ATGAGC +ATGAGG +ATGTAA +ATGTAT +ATGTAC +ATGTAG +ATGTTA +ATGTTT +ATGTTC +ATGTTG +ATGTCA +ATGTCT +ATGTCC +ATGTCG +ATGTGA +ATGTGT +ATGTGC +ATGTGG +ATGCAA +ATGCAT +ATGCAC +ATGCAG +ATGCTA +ATGCTT +ATGCTC +ATGCTG +ATGCCA +ATGCCT +ATGCCC +ATGCCG +ATGCGA +ATGCGT +ATGCGC +ATGCGG +ATGGAA +ATGGAT +ATGGAC +ATGGAG +ATGGTA +ATGGTT +ATGGTC +ATGGTG +ATGGCA +ATGGCT +ATGGCC +ATGGCG +ATGGGA +ATGGGT +ATGGGC +ATGGGG +ACAAAA +ACAAAT +ACAAAC +ACAAAG +ACAATA +ACAATT +ACAATC +ACAATG +ACAACA +ACAACT +ACAACC +ACAACG +ACAAGA +ACAAGT +ACAAGC +ACAAGG +ACATAA +ACATAT +ACATAC +ACATAG +ACATTA +ACATTT +ACATTC +ACATTG +ACATCA +ACATCT +ACATCC +ACATCG +ACATGA +ACATGT +ACATGC +ACATGG +ACACAA +ACACAT +ACACAC +ACACAG +ACACTA +ACACTT +ACACTC +ACACTG +ACACCA +ACACCT +ACACCC +ACACCG +ACACGA +ACACGT +ACACGC +ACACGG +ACAGAA +ACAGAT +ACAGAC +ACAGAG +ACAGTA +ACAGTT +ACAGTC +ACAGTG +ACAGCA +ACAGCT +ACAGCC +ACAGCG +ACAGGA +ACAGGT +ACAGGC +ACAGGG +ACTAAA +ACTAAT +ACTAAC +ACTAAG +ACTATA +ACTATT +ACTATC +ACTATG +ACTACA +ACTACT +ACTACC +ACTACG +ACTAGA +ACTAGT +ACTAGC +ACTAGG +ACTTAA +ACTTAT +ACTTAC +ACTTAG +ACTTTA +ACTTTT +ACTTTC +ACTTTG +ACTTCA +ACTTCT +ACTTCC +ACTTCG +ACTTGA +ACTTGT +ACTTGC +ACTTGG +ACTCAA +ACTCAT +ACTCAC +ACTCAG +ACTCTA +ACTCTT +ACTCTC +ACTCTG +ACTCCA +ACTCCT +ACTCCC +ACTCCG +ACTCGA +ACTCGT +ACTCGC +ACTCGG +ACTGAA +ACTGAT +ACTGAC +ACTGAG +ACTGTA +ACTGTT +ACTGTC +ACTGTG +ACTGCA +ACTGCT +ACTGCC +ACTGCG +ACTGGA +ACTGGT +ACTGGC +ACTGGG +ACCAAA +ACCAAT +ACCAAC +ACCAAG +ACCATA +ACCATT +ACCATC +ACCATG +ACCACA +ACCACT +ACCACC +ACCACG +ACCAGA +ACCAGT +ACCAGC +ACCAGG +ACCTAA +ACCTAT +ACCTAC +ACCTAG +ACCTTA +ACCTTT +ACCTTC +ACCTTG +ACCTCA +ACCTCT +ACCTCC +ACCTCG +ACCTGA +ACCTGT +ACCTGC +ACCTGG +ACCCAA +ACCCAT +ACCCAC +ACCCAG +ACCCTA +ACCCTT +ACCCTC +ACCCTG +ACCCCA +ACCCCT +ACCCCC +ACCCCG +ACCCGA +ACCCGT +ACCCGC +ACCCGG +ACCGAA +ACCGAT +ACCGAC +ACCGAG +ACCGTA +ACCGTT +ACCGTC +ACCGTG +ACCGCA +ACCGCT +ACCGCC +ACCGCG +ACCGGA +ACCGGT +ACCGGC +ACCGGG +ACGAAA +ACGAAT +ACGAAC +ACGAAG +ACGATA +ACGATT +ACGATC +ACGATG +ACGACA +ACGACT +ACGACC +ACGACG +ACGAGA +ACGAGT +ACGAGC +ACGAGG +ACGTAA +ACGTAT +ACGTAC +ACGTAG +ACGTTA +ACGTTT +ACGTTC +ACGTTG +ACGTCA +ACGTCT +ACGTCC +ACGTCG +ACGTGA +ACGTGT +ACGTGC +ACGTGG +ACGCAA +ACGCAT +ACGCAC +ACGCAG +ACGCTA +ACGCTT +ACGCTC +ACGCTG +ACGCCA +ACGCCT +ACGCCC +ACGCCG +ACGCGA +ACGCGT +ACGCGC +ACGCGG +ACGGAA +ACGGAT +ACGGAC +ACGGAG +ACGGTA +ACGGTT +ACGGTC +ACGGTG +ACGGCA +ACGGCT +ACGGCC +ACGGCG +ACGGGA +ACGGGT +ACGGGC +ACGGGG +AGAAAA +AGAAAT +AGAAAC +AGAAAG +AGAATA +AGAATT +AGAATC +AGAATG +AGAACA +AGAACT +AGAACC +AGAACG +AGAAGA +AGAAGT +AGAAGC +AGAAGG +AGATAA +AGATAT +AGATAC +AGATAG +AGATTA +AGATTT +AGATTC +AGATTG +AGATCA +AGATCT +AGATCC +AGATCG +AGATGA +AGATGT +AGATGC +AGATGG +AGACAA +AGACAT +AGACAC +AGACAG +AGACTA +AGACTT +AGACTC +AGACTG +AGACCA +AGACCT +AGACCC +AGACCG +AGACGA +AGACGT +AGACGC +AGACGG +AGAGAA +AGAGAT +AGAGAC +AGAGAG +AGAGTA +AGAGTT +AGAGTC +AGAGTG +AGAGCA +AGAGCT +AGAGCC +AGAGCG +AGAGGA +AGAGGT +AGAGGC +AGAGGG +AGTAAA +AGTAAT +AGTAAC +AGTAAG +AGTATA +AGTATT +AGTATC +AGTATG +AGTACA +AGTACT +AGTACC +AGTACG +AGTAGA +AGTAGT +AGTAGC +AGTAGG +AGTTAA +AGTTAT +AGTTAC +AGTTAG +AGTTTA +AGTTTT +AGTTTC +AGTTTG +AGTTCA +AGTTCT +AGTTCC +AGTTCG +AGTTGA +AGTTGT +AGTTGC +AGTTGG +AGTCAA +AGTCAT +AGTCAC +AGTCAG +AGTCTA +AGTCTT +AGTCTC +AGTCTG +AGTCCA +AGTCCT +AGTCCC +AGTCCG +AGTCGA +AGTCGT +AGTCGC +AGTCGG +AGTGAA +AGTGAT +AGTGAC +AGTGAG +AGTGTA +AGTGTT +AGTGTC +AGTGTG +AGTGCA +AGTGCT +AGTGCC +AGTGCG +AGTGGA +AGTGGT +AGTGGC +AGTGGG +AGCAAA +AGCAAT +AGCAAC +AGCAAG +AGCATA +AGCATT +AGCATC +AGCATG +AGCACA +AGCACT +AGCACC +AGCACG +AGCAGA +AGCAGT +AGCAGC +AGCAGG +AGCTAA +AGCTAT +AGCTAC +AGCTAG +AGCTTA +AGCTTT +AGCTTC +AGCTTG +AGCTCA +AGCTCT +AGCTCC +AGCTCG +AGCTGA +AGCTGT +AGCTGC +AGCTGG +AGCCAA +AGCCAT +AGCCAC +AGCCAG +AGCCTA +AGCCTT +AGCCTC +AGCCTG +AGCCCA +AGCCCT +AGCCCC +AGCCCG +AGCCGA +AGCCGT +AGCCGC +AGCCGG +AGCGAA +AGCGAT +AGCGAC +AGCGAG +AGCGTA +AGCGTT +AGCGTC +AGCGTG +AGCGCA +AGCGCT +AGCGCC +AGCGCG +AGCGGA +AGCGGT +AGCGGC +AGCGGG +AGGAAA +AGGAAT +AGGAAC +AGGAAG +AGGATA +AGGATT +AGGATC +AGGATG +AGGACA +AGGACT +AGGACC +AGGACG +AGGAGA +AGGAGT +AGGAGC +AGGAGG +AGGTAA +AGGTAT +AGGTAC +AGGTAG +AGGTTA +AGGTTT +AGGTTC +AGGTTG +AGGTCA +AGGTCT +AGGTCC +AGGTCG +AGGTGA +AGGTGT +AGGTGC +AGGTGG +AGGCAA +AGGCAT +AGGCAC +AGGCAG +AGGCTA +AGGCTT +AGGCTC +AGGCTG +AGGCCA +AGGCCT +AGGCCC +AGGCCG +AGGCGA +AGGCGT +AGGCGC +AGGCGG +AGGGAA +AGGGAT +AGGGAC +AGGGAG +AGGGTA +AGGGTT +AGGGTC +AGGGTG +AGGGCA +AGGGCT +AGGGCC +AGGGCG +AGGGGA +AGGGGT +AGGGGC +AGGGGG +TAAAAA +TAAAAT +TAAAAC +TAAAAG +TAAATA +TAAATT +TAAATC +TAAATG +TAAACA +TAAACT +TAAACC +TAAACG +TAAAGA +TAAAGT +TAAAGC +TAAAGG +TAATAA +TAATAT +TAATAC +TAATAG +TAATTA +TAATTT +TAATTC +TAATTG +TAATCA +TAATCT +TAATCC +TAATCG +TAATGA +TAATGT +TAATGC +TAATGG +TAACAA +TAACAT +TAACAC +TAACAG +TAACTA +TAACTT +TAACTC +TAACTG +TAACCA +TAACCT +TAACCC +TAACCG +TAACGA +TAACGT +TAACGC +TAACGG +TAAGAA +TAAGAT +TAAGAC +TAAGAG +TAAGTA +TAAGTT +TAAGTC +TAAGTG +TAAGCA +TAAGCT +TAAGCC +TAAGCG +TAAGGA +TAAGGT +TAAGGC +TAAGGG +TATAAA +TATAAT +TATAAC +TATAAG +TATATA +TATATT +TATATC +TATATG +TATACA +TATACT +TATACC +TATACG +TATAGA +TATAGT +TATAGC +TATAGG +TATTAA +TATTAT +TATTAC +TATTAG +TATTTA +TATTTT +TATTTC +TATTTG +TATTCA +TATTCT +TATTCC +TATTCG +TATTGA +TATTGT +TATTGC +TATTGG +TATCAA +TATCAT +TATCAC +TATCAG +TATCTA +TATCTT +TATCTC +TATCTG +TATCCA +TATCCT +TATCCC +TATCCG +TATCGA +TATCGT +TATCGC +TATCGG +TATGAA +TATGAT +TATGAC +TATGAG +TATGTA +TATGTT +TATGTC +TATGTG +TATGCA +TATGCT +TATGCC +TATGCG +TATGGA +TATGGT +TATGGC +TATGGG +TACAAA +TACAAT +TACAAC +TACAAG +TACATA +TACATT +TACATC +TACATG +TACACA +TACACT +TACACC +TACACG +TACAGA +TACAGT +TACAGC +TACAGG +TACTAA +TACTAT +TACTAC +TACTAG +TACTTA +TACTTT +TACTTC +TACTTG +TACTCA +TACTCT +TACTCC +TACTCG +TACTGA +TACTGT +TACTGC +TACTGG +TACCAA +TACCAT +TACCAC +TACCAG +TACCTA +TACCTT +TACCTC +TACCTG +TACCCA +TACCCT +TACCCC +TACCCG +TACCGA +TACCGT +TACCGC +TACCGG +TACGAA +TACGAT +TACGAC +TACGAG +TACGTA +TACGTT +TACGTC +TACGTG +TACGCA +TACGCT +TACGCC +TACGCG +TACGGA +TACGGT +TACGGC +TACGGG +TAGAAA +TAGAAT +TAGAAC +TAGAAG +TAGATA +TAGATT +TAGATC +TAGATG +TAGACA +TAGACT +TAGACC +TAGACG +TAGAGA +TAGAGT +TAGAGC +TAGAGG +TAGTAA +TAGTAT +TAGTAC +TAGTAG +TAGTTA +TAGTTT +TAGTTC +TAGTTG +TAGTCA +TAGTCT +TAGTCC +TAGTCG +TAGTGA +TAGTGT +TAGTGC +TAGTGG +TAGCAA +TAGCAT +TAGCAC +TAGCAG +TAGCTA +TAGCTT +TAGCTC +TAGCTG +TAGCCA +TAGCCT +TAGCCC +TAGCCG +TAGCGA +TAGCGT +TAGCGC +TAGCGG +TAGGAA +TAGGAT +TAGGAC +TAGGAG +TAGGTA +TAGGTT +TAGGTC +TAGGTG +TAGGCA +TAGGCT +TAGGCC +TAGGCG +TAGGGA +TAGGGT +TAGGGC +TAGGGG +TTAAAA +TTAAAT +TTAAAC +TTAAAG +TTAATA +TTAATT +TTAATC +TTAATG +TTAACA +TTAACT +TTAACC +TTAACG +TTAAGA +TTAAGT +TTAAGC +TTAAGG +TTATAA +TTATAT +TTATAC +TTATAG +TTATTA +TTATTT +TTATTC +TTATTG +TTATCA +TTATCT +TTATCC +TTATCG +TTATGA +TTATGT +TTATGC +TTATGG +TTACAA +TTACAT +TTACAC +TTACAG +TTACTA +TTACTT +TTACTC +TTACTG +TTACCA +TTACCT +TTACCC +TTACCG +TTACGA +TTACGT +TTACGC +TTACGG +TTAGAA +TTAGAT +TTAGAC +TTAGAG +TTAGTA +TTAGTT +TTAGTC +TTAGTG +TTAGCA +TTAGCT +TTAGCC +TTAGCG +TTAGGA +TTAGGT +TTAGGC +TTAGGG +TTTAAA +TTTAAT +TTTAAC +TTTAAG +TTTATA +TTTATT +TTTATC +TTTATG +TTTACA +TTTACT +TTTACC +TTTACG +TTTAGA +TTTAGT +TTTAGC +TTTAGG +TTTTAA +TTTTAT +TTTTAC +TTTTAG +TTTTTA +TTTTTT +TTTTTC +TTTTTG +TTTTCA +TTTTCT +TTTTCC +TTTTCG +TTTTGA +TTTTGT +TTTTGC +TTTTGG +TTTCAA +TTTCAT +TTTCAC +TTTCAG +TTTCTA +TTTCTT +TTTCTC +TTTCTG +TTTCCA +TTTCCT +TTTCCC +TTTCCG +TTTCGA +TTTCGT +TTTCGC +TTTCGG +TTTGAA +TTTGAT +TTTGAC +TTTGAG +TTTGTA +TTTGTT +TTTGTC +TTTGTG +TTTGCA +TTTGCT +TTTGCC +TTTGCG +TTTGGA +TTTGGT +TTTGGC +TTTGGG +TTCAAA +TTCAAT +TTCAAC +TTCAAG +TTCATA +TTCATT +TTCATC +TTCATG +TTCACA +TTCACT +TTCACC +TTCACG +TTCAGA +TTCAGT +TTCAGC +TTCAGG +TTCTAA +TTCTAT +TTCTAC +TTCTAG +TTCTTA +TTCTTT +TTCTTC +TTCTTG +TTCTCA +TTCTCT +TTCTCC +TTCTCG +TTCTGA +TTCTGT +TTCTGC +TTCTGG +TTCCAA +TTCCAT +TTCCAC +TTCCAG +TTCCTA +TTCCTT +TTCCTC +TTCCTG +TTCCCA +TTCCCT +TTCCCC +TTCCCG +TTCCGA +TTCCGT +TTCCGC +TTCCGG +TTCGAA +TTCGAT +TTCGAC +TTCGAG +TTCGTA +TTCGTT +TTCGTC +TTCGTG +TTCGCA +TTCGCT +TTCGCC +TTCGCG +TTCGGA +TTCGGT +TTCGGC +TTCGGG +TTGAAA +TTGAAT +TTGAAC +TTGAAG +TTGATA +TTGATT +TTGATC +TTGATG +TTGACA +TTGACT +TTGACC +TTGACG +TTGAGA +TTGAGT +TTGAGC +TTGAGG +TTGTAA +TTGTAT +TTGTAC +TTGTAG +TTGTTA +TTGTTT +TTGTTC +TTGTTG +TTGTCA +TTGTCT +TTGTCC +TTGTCG +TTGTGA +TTGTGT +TTGTGC +TTGTGG +TTGCAA +TTGCAT +TTGCAC +TTGCAG +TTGCTA +TTGCTT +TTGCTC +TTGCTG +TTGCCA +TTGCCT +TTGCCC +TTGCCG +TTGCGA +TTGCGT +TTGCGC +TTGCGG +TTGGAA +TTGGAT +TTGGAC +TTGGAG +TTGGTA +TTGGTT +TTGGTC +TTGGTG +TTGGCA +TTGGCT +TTGGCC +TTGGCG +TTGGGA +TTGGGT +TTGGGC +TTGGGG +TCAAAA +TCAAAT +TCAAAC +TCAAAG +TCAATA +TCAATT +TCAATC +TCAATG +TCAACA +TCAACT +TCAACC +TCAACG +TCAAGA +TCAAGT +TCAAGC +TCAAGG +TCATAA +TCATAT +TCATAC +TCATAG +TCATTA +TCATTT +TCATTC +TCATTG +TCATCA +TCATCT +TCATCC +TCATCG +TCATGA +TCATGT +TCATGC +TCATGG +TCACAA +TCACAT +TCACAC +TCACAG +TCACTA +TCACTT +TCACTC +TCACTG +TCACCA +TCACCT +TCACCC +TCACCG +TCACGA +TCACGT +TCACGC +TCACGG +TCAGAA +TCAGAT +TCAGAC +TCAGAG +TCAGTA +TCAGTT +TCAGTC +TCAGTG +TCAGCA +TCAGCT +TCAGCC +TCAGCG +TCAGGA +TCAGGT +TCAGGC +TCAGGG +TCTAAA +TCTAAT +TCTAAC +TCTAAG +TCTATA +TCTATT +TCTATC +TCTATG +TCTACA +TCTACT +TCTACC +TCTACG +TCTAGA +TCTAGT +TCTAGC +TCTAGG +TCTTAA +TCTTAT +TCTTAC +TCTTAG +TCTTTA +TCTTTT +TCTTTC +TCTTTG +TCTTCA +TCTTCT +TCTTCC +TCTTCG +TCTTGA +TCTTGT +TCTTGC +TCTTGG +TCTCAA +TCTCAT +TCTCAC +TCTCAG +TCTCTA +TCTCTT +TCTCTC +TCTCTG +TCTCCA +TCTCCT +TCTCCC +TCTCCG +TCTCGA +TCTCGT +TCTCGC +TCTCGG +TCTGAA +TCTGAT +TCTGAC +TCTGAG +TCTGTA +TCTGTT +TCTGTC +TCTGTG +TCTGCA +TCTGCT +TCTGCC +TCTGCG +TCTGGA +TCTGGT +TCTGGC +TCTGGG +TCCAAA +TCCAAT +TCCAAC +TCCAAG +TCCATA +TCCATT +TCCATC +TCCATG +TCCACA +TCCACT +TCCACC +TCCACG +TCCAGA +TCCAGT +TCCAGC +TCCAGG +TCCTAA +TCCTAT +TCCTAC +TCCTAG +TCCTTA +TCCTTT +TCCTTC +TCCTTG +TCCTCA +TCCTCT +TCCTCC +TCCTCG +TCCTGA +TCCTGT +TCCTGC +TCCTGG +TCCCAA +TCCCAT +TCCCAC +TCCCAG +TCCCTA +TCCCTT +TCCCTC +TCCCTG +TCCCCA +TCCCCT +TCCCCC +TCCCCG +TCCCGA +TCCCGT +TCCCGC +TCCCGG +TCCGAA +TCCGAT +TCCGAC +TCCGAG +TCCGTA +TCCGTT +TCCGTC +TCCGTG +TCCGCA +TCCGCT +TCCGCC +TCCGCG +TCCGGA +TCCGGT +TCCGGC +TCCGGG +TCGAAA +TCGAAT +TCGAAC +TCGAAG +TCGATA +TCGATT +TCGATC +TCGATG +TCGACA +TCGACT +TCGACC +TCGACG +TCGAGA +TCGAGT +TCGAGC +TCGAGG +TCGTAA +TCGTAT +TCGTAC +TCGTAG +TCGTTA +TCGTTT +TCGTTC +TCGTTG +TCGTCA +TCGTCT +TCGTCC +TCGTCG +TCGTGA +TCGTGT +TCGTGC +TCGTGG +TCGCAA +TCGCAT +TCGCAC +TCGCAG +TCGCTA +TCGCTT +TCGCTC +TCGCTG +TCGCCA +TCGCCT +TCGCCC +TCGCCG +TCGCGA +TCGCGT +TCGCGC +TCGCGG +TCGGAA +TCGGAT +TCGGAC +TCGGAG +TCGGTA +TCGGTT +TCGGTC +TCGGTG +TCGGCA +TCGGCT +TCGGCC +TCGGCG +TCGGGA +TCGGGT +TCGGGC +TCGGGG +TGAAAA +TGAAAT +TGAAAC +TGAAAG +TGAATA +TGAATT +TGAATC +TGAATG +TGAACA +TGAACT +TGAACC +TGAACG +TGAAGA +TGAAGT +TGAAGC +TGAAGG +TGATAA +TGATAT +TGATAC +TGATAG +TGATTA +TGATTT +TGATTC +TGATTG +TGATCA +TGATCT +TGATCC +TGATCG +TGATGA +TGATGT +TGATGC +TGATGG +TGACAA +TGACAT +TGACAC +TGACAG +TGACTA +TGACTT +TGACTC +TGACTG +TGACCA +TGACCT +TGACCC +TGACCG +TGACGA +TGACGT +TGACGC +TGACGG +TGAGAA +TGAGAT +TGAGAC +TGAGAG +TGAGTA +TGAGTT +TGAGTC +TGAGTG +TGAGCA +TGAGCT +TGAGCC +TGAGCG +TGAGGA +TGAGGT +TGAGGC +TGAGGG +TGTAAA +TGTAAT +TGTAAC +TGTAAG +TGTATA +TGTATT +TGTATC +TGTATG +TGTACA +TGTACT +TGTACC +TGTACG +TGTAGA +TGTAGT +TGTAGC +TGTAGG +TGTTAA +TGTTAT +TGTTAC +TGTTAG +TGTTTA +TGTTTT +TGTTTC +TGTTTG +TGTTCA +TGTTCT +TGTTCC +TGTTCG +TGTTGA +TGTTGT +TGTTGC +TGTTGG +TGTCAA +TGTCAT +TGTCAC +TGTCAG +TGTCTA +TGTCTT +TGTCTC +TGTCTG +TGTCCA +TGTCCT +TGTCCC +TGTCCG +TGTCGA +TGTCGT +TGTCGC +TGTCGG +TGTGAA +TGTGAT +TGTGAC +TGTGAG +TGTGTA +TGTGTT +TGTGTC +TGTGTG +TGTGCA +TGTGCT +TGTGCC +TGTGCG +TGTGGA +TGTGGT +TGTGGC +TGTGGG +TGCAAA +TGCAAT +TGCAAC +TGCAAG +TGCATA +TGCATT +TGCATC +TGCATG +TGCACA +TGCACT +TGCACC +TGCACG +TGCAGA +TGCAGT +TGCAGC +TGCAGG +TGCTAA +TGCTAT +TGCTAC +TGCTAG +TGCTTA +TGCTTT +TGCTTC +TGCTTG +TGCTCA +TGCTCT +TGCTCC +TGCTCG +TGCTGA +TGCTGT +TGCTGC +TGCTGG +TGCCAA +TGCCAT +TGCCAC +TGCCAG +TGCCTA +TGCCTT +TGCCTC +TGCCTG +TGCCCA +TGCCCT +TGCCCC +TGCCCG +TGCCGA +TGCCGT +TGCCGC +TGCCGG +TGCGAA +TGCGAT +TGCGAC +TGCGAG +TGCGTA +TGCGTT +TGCGTC +TGCGTG +TGCGCA +TGCGCT +TGCGCC +TGCGCG +TGCGGA +TGCGGT +TGCGGC +TGCGGG +TGGAAA +TGGAAT +TGGAAC +TGGAAG +TGGATA +TGGATT +TGGATC +TGGATG +TGGACA +TGGACT +TGGACC +TGGACG +TGGAGA +TGGAGT +TGGAGC +TGGAGG +TGGTAA +TGGTAT +TGGTAC +TGGTAG +TGGTTA +TGGTTT +TGGTTC +TGGTTG +TGGTCA +TGGTCT +TGGTCC +TGGTCG +TGGTGA +TGGTGT +TGGTGC +TGGTGG +TGGCAA +TGGCAT +TGGCAC +TGGCAG +TGGCTA +TGGCTT +TGGCTC +TGGCTG +TGGCCA +TGGCCT +TGGCCC +TGGCCG +TGGCGA +TGGCGT +TGGCGC +TGGCGG +TGGGAA +TGGGAT +TGGGAC +TGGGAG +TGGGTA +TGGGTT +TGGGTC +TGGGTG +TGGGCA +TGGGCT +TGGGCC +TGGGCG +TGGGGA +TGGGGT +TGGGGC +TGGGGG +CAAAAA +CAAAAT +CAAAAC +CAAAAG +CAAATA +CAAATT +CAAATC +CAAATG +CAAACA +CAAACT +CAAACC +CAAACG +CAAAGA +CAAAGT +CAAAGC +CAAAGG +CAATAA +CAATAT +CAATAC +CAATAG +CAATTA +CAATTT +CAATTC +CAATTG +CAATCA +CAATCT +CAATCC +CAATCG +CAATGA +CAATGT +CAATGC +CAATGG +CAACAA +CAACAT +CAACAC +CAACAG +CAACTA +CAACTT +CAACTC +CAACTG +CAACCA +CAACCT +CAACCC +CAACCG +CAACGA +CAACGT +CAACGC +CAACGG +CAAGAA +CAAGAT +CAAGAC +CAAGAG +CAAGTA +CAAGTT +CAAGTC +CAAGTG +CAAGCA +CAAGCT +CAAGCC +CAAGCG +CAAGGA +CAAGGT +CAAGGC +CAAGGG +CATAAA +CATAAT +CATAAC +CATAAG +CATATA +CATATT +CATATC +CATATG +CATACA +CATACT +CATACC +CATACG +CATAGA +CATAGT +CATAGC +CATAGG +CATTAA +CATTAT +CATTAC +CATTAG +CATTTA +CATTTT +CATTTC +CATTTG +CATTCA +CATTCT +CATTCC +CATTCG +CATTGA +CATTGT +CATTGC +CATTGG +CATCAA +CATCAT +CATCAC +CATCAG +CATCTA +CATCTT +CATCTC +CATCTG +CATCCA +CATCCT +CATCCC +CATCCG +CATCGA +CATCGT +CATCGC +CATCGG +CATGAA +CATGAT +CATGAC +CATGAG +CATGTA +CATGTT +CATGTC +CATGTG +CATGCA +CATGCT +CATGCC +CATGCG +CATGGA +CATGGT +CATGGC +CATGGG +CACAAA +CACAAT +CACAAC +CACAAG +CACATA +CACATT +CACATC +CACATG +CACACA +CACACT +CACACC +CACACG +CACAGA +CACAGT +CACAGC +CACAGG +CACTAA +CACTAT +CACTAC +CACTAG +CACTTA +CACTTT +CACTTC +CACTTG +CACTCA +CACTCT +CACTCC +CACTCG +CACTGA +CACTGT +CACTGC +CACTGG +CACCAA +CACCAT +CACCAC +CACCAG +CACCTA +CACCTT +CACCTC +CACCTG +CACCCA +CACCCT +CACCCC +CACCCG +CACCGA +CACCGT +CACCGC +CACCGG +CACGAA +CACGAT +CACGAC +CACGAG +CACGTA +CACGTT +CACGTC +CACGTG +CACGCA +CACGCT +CACGCC +CACGCG +CACGGA +CACGGT +CACGGC +CACGGG +CAGAAA +CAGAAT +CAGAAC +CAGAAG +CAGATA +CAGATT +CAGATC +CAGATG +CAGACA +CAGACT +CAGACC +CAGACG +CAGAGA +CAGAGT +CAGAGC +CAGAGG +CAGTAA +CAGTAT +CAGTAC +CAGTAG +CAGTTA +CAGTTT +CAGTTC +CAGTTG +CAGTCA +CAGTCT +CAGTCC +CAGTCG +CAGTGA +CAGTGT +CAGTGC +CAGTGG +CAGCAA +CAGCAT +CAGCAC +CAGCAG +CAGCTA +CAGCTT +CAGCTC +CAGCTG +CAGCCA +CAGCCT +CAGCCC +CAGCCG +CAGCGA +CAGCGT +CAGCGC +CAGCGG +CAGGAA +CAGGAT +CAGGAC +CAGGAG +CAGGTA +CAGGTT +CAGGTC +CAGGTG +CAGGCA +CAGGCT +CAGGCC +CAGGCG +CAGGGA +CAGGGT +CAGGGC +CAGGGG +CTAAAA +CTAAAT +CTAAAC +CTAAAG +CTAATA +CTAATT +CTAATC +CTAATG +CTAACA +CTAACT +CTAACC +CTAACG +CTAAGA +CTAAGT +CTAAGC +CTAAGG +CTATAA +CTATAT +CTATAC +CTATAG +CTATTA +CTATTT +CTATTC +CTATTG +CTATCA +CTATCT +CTATCC +CTATCG +CTATGA +CTATGT +CTATGC +CTATGG +CTACAA +CTACAT +CTACAC +CTACAG +CTACTA +CTACTT +CTACTC +CTACTG +CTACCA +CTACCT +CTACCC +CTACCG +CTACGA +CTACGT +CTACGC +CTACGG +CTAGAA +CTAGAT +CTAGAC +CTAGAG +CTAGTA +CTAGTT +CTAGTC +CTAGTG +CTAGCA +CTAGCT +CTAGCC +CTAGCG +CTAGGA +CTAGGT +CTAGGC +CTAGGG +CTTAAA +CTTAAT +CTTAAC +CTTAAG +CTTATA +CTTATT +CTTATC +CTTATG +CTTACA +CTTACT +CTTACC +CTTACG +CTTAGA +CTTAGT +CTTAGC +CTTAGG +CTTTAA +CTTTAT +CTTTAC +CTTTAG +CTTTTA +CTTTTT +CTTTTC +CTTTTG +CTTTCA +CTTTCT +CTTTCC +CTTTCG +CTTTGA +CTTTGT +CTTTGC +CTTTGG +CTTCAA +CTTCAT +CTTCAC +CTTCAG +CTTCTA +CTTCTT +CTTCTC +CTTCTG +CTTCCA +CTTCCT +CTTCCC +CTTCCG +CTTCGA +CTTCGT +CTTCGC +CTTCGG +CTTGAA +CTTGAT +CTTGAC +CTTGAG +CTTGTA +CTTGTT +CTTGTC +CTTGTG +CTTGCA +CTTGCT +CTTGCC +CTTGCG +CTTGGA +CTTGGT +CTTGGC +CTTGGG +CTCAAA +CTCAAT +CTCAAC +CTCAAG +CTCATA +CTCATT +CTCATC +CTCATG +CTCACA +CTCACT +CTCACC +CTCACG +CTCAGA +CTCAGT +CTCAGC +CTCAGG +CTCTAA +CTCTAT +CTCTAC +CTCTAG +CTCTTA +CTCTTT +CTCTTC +CTCTTG +CTCTCA +CTCTCT +CTCTCC +CTCTCG +CTCTGA +CTCTGT +CTCTGC +CTCTGG +CTCCAA +CTCCAT +CTCCAC +CTCCAG +CTCCTA +CTCCTT +CTCCTC +CTCCTG +CTCCCA +CTCCCT +CTCCCC +CTCCCG +CTCCGA +CTCCGT +CTCCGC +CTCCGG +CTCGAA +CTCGAT +CTCGAC +CTCGAG +CTCGTA +CTCGTT +CTCGTC +CTCGTG +CTCGCA +CTCGCT +CTCGCC +CTCGCG +CTCGGA +CTCGGT +CTCGGC +CTCGGG +CTGAAA +CTGAAT +CTGAAC +CTGAAG +CTGATA +CTGATT +CTGATC +CTGATG +CTGACA +CTGACT +CTGACC +CTGACG +CTGAGA +CTGAGT +CTGAGC +CTGAGG +CTGTAA +CTGTAT +CTGTAC +CTGTAG +CTGTTA +CTGTTT +CTGTTC +CTGTTG +CTGTCA +CTGTCT +CTGTCC +CTGTCG +CTGTGA +CTGTGT +CTGTGC +CTGTGG +CTGCAA +CTGCAT +CTGCAC +CTGCAG +CTGCTA +CTGCTT +CTGCTC +CTGCTG +CTGCCA +CTGCCT +CTGCCC +CTGCCG +CTGCGA +CTGCGT +CTGCGC +CTGCGG +CTGGAA +CTGGAT +CTGGAC +CTGGAG +CTGGTA +CTGGTT +CTGGTC +CTGGTG +CTGGCA +CTGGCT +CTGGCC +CTGGCG +CTGGGA +CTGGGT +CTGGGC +CTGGGG +CCAAAA +CCAAAT +CCAAAC +CCAAAG +CCAATA +CCAATT +CCAATC +CCAATG +CCAACA +CCAACT +CCAACC +CCAACG +CCAAGA +CCAAGT +CCAAGC +CCAAGG +CCATAA +CCATAT +CCATAC +CCATAG +CCATTA +CCATTT +CCATTC +CCATTG +CCATCA +CCATCT +CCATCC +CCATCG +CCATGA +CCATGT +CCATGC +CCATGG +CCACAA +CCACAT +CCACAC +CCACAG +CCACTA +CCACTT +CCACTC +CCACTG +CCACCA +CCACCT +CCACCC +CCACCG +CCACGA +CCACGT +CCACGC +CCACGG +CCAGAA +CCAGAT +CCAGAC +CCAGAG +CCAGTA +CCAGTT +CCAGTC +CCAGTG +CCAGCA +CCAGCT +CCAGCC +CCAGCG +CCAGGA +CCAGGT +CCAGGC +CCAGGG +CCTAAA +CCTAAT +CCTAAC +CCTAAG +CCTATA +CCTATT +CCTATC +CCTATG +CCTACA +CCTACT +CCTACC +CCTACG +CCTAGA +CCTAGT +CCTAGC +CCTAGG +CCTTAA +CCTTAT +CCTTAC +CCTTAG +CCTTTA +CCTTTT +CCTTTC +CCTTTG +CCTTCA +CCTTCT +CCTTCC +CCTTCG +CCTTGA +CCTTGT +CCTTGC +CCTTGG +CCTCAA +CCTCAT +CCTCAC +CCTCAG +CCTCTA +CCTCTT +CCTCTC +CCTCTG +CCTCCA +CCTCCT +CCTCCC +CCTCCG +CCTCGA +CCTCGT +CCTCGC +CCTCGG +CCTGAA +CCTGAT +CCTGAC +CCTGAG +CCTGTA +CCTGTT +CCTGTC +CCTGTG +CCTGCA +CCTGCT +CCTGCC +CCTGCG +CCTGGA +CCTGGT +CCTGGC +CCTGGG +CCCAAA +CCCAAT +CCCAAC +CCCAAG +CCCATA +CCCATT +CCCATC +CCCATG +CCCACA +CCCACT +CCCACC +CCCACG +CCCAGA +CCCAGT +CCCAGC +CCCAGG +CCCTAA +CCCTAT +CCCTAC +CCCTAG +CCCTTA +CCCTTT +CCCTTC +CCCTTG +CCCTCA +CCCTCT +CCCTCC +CCCTCG +CCCTGA +CCCTGT +CCCTGC +CCCTGG +CCCCAA +CCCCAT +CCCCAC +CCCCAG +CCCCTA +CCCCTT +CCCCTC +CCCCTG +CCCCCA +CCCCCT +CCCCCC +CCCCCG +CCCCGA +CCCCGT +CCCCGC +CCCCGG +CCCGAA +CCCGAT +CCCGAC +CCCGAG +CCCGTA +CCCGTT +CCCGTC +CCCGTG +CCCGCA +CCCGCT +CCCGCC +CCCGCG +CCCGGA +CCCGGT +CCCGGC +CCCGGG +CCGAAA +CCGAAT +CCGAAC +CCGAAG +CCGATA +CCGATT +CCGATC +CCGATG +CCGACA +CCGACT +CCGACC +CCGACG +CCGAGA +CCGAGT +CCGAGC +CCGAGG +CCGTAA +CCGTAT +CCGTAC +CCGTAG +CCGTTA +CCGTTT +CCGTTC +CCGTTG +CCGTCA +CCGTCT +CCGTCC +CCGTCG +CCGTGA +CCGTGT +CCGTGC +CCGTGG +CCGCAA +CCGCAT +CCGCAC +CCGCAG +CCGCTA +CCGCTT +CCGCTC +CCGCTG +CCGCCA +CCGCCT +CCGCCC +CCGCCG +CCGCGA +CCGCGT +CCGCGC +CCGCGG +CCGGAA +CCGGAT +CCGGAC +CCGGAG +CCGGTA +CCGGTT +CCGGTC +CCGGTG +CCGGCA +CCGGCT +CCGGCC +CCGGCG +CCGGGA +CCGGGT +CCGGGC +CCGGGG +CGAAAA +CGAAAT +CGAAAC +CGAAAG +CGAATA +CGAATT +CGAATC +CGAATG +CGAACA +CGAACT +CGAACC +CGAACG +CGAAGA +CGAAGT +CGAAGC +CGAAGG +CGATAA +CGATAT +CGATAC +CGATAG +CGATTA +CGATTT +CGATTC +CGATTG +CGATCA +CGATCT +CGATCC +CGATCG +CGATGA +CGATGT +CGATGC +CGATGG +CGACAA +CGACAT +CGACAC +CGACAG +CGACTA +CGACTT +CGACTC +CGACTG +CGACCA +CGACCT +CGACCC +CGACCG +CGACGA +CGACGT +CGACGC +CGACGG +CGAGAA +CGAGAT +CGAGAC +CGAGAG +CGAGTA +CGAGTT +CGAGTC +CGAGTG +CGAGCA +CGAGCT +CGAGCC +CGAGCG +CGAGGA +CGAGGT +CGAGGC +CGAGGG +CGTAAA +CGTAAT +CGTAAC +CGTAAG +CGTATA +CGTATT +CGTATC +CGTATG +CGTACA +CGTACT +CGTACC +CGTACG +CGTAGA +CGTAGT +CGTAGC +CGTAGG +CGTTAA +CGTTAT +CGTTAC +CGTTAG +CGTTTA +CGTTTT +CGTTTC +CGTTTG +CGTTCA +CGTTCT +CGTTCC +CGTTCG +CGTTGA +CGTTGT +CGTTGC +CGTTGG +CGTCAA +CGTCAT +CGTCAC +CGTCAG +CGTCTA +CGTCTT +CGTCTC +CGTCTG +CGTCCA +CGTCCT +CGTCCC +CGTCCG +CGTCGA +CGTCGT +CGTCGC +CGTCGG +CGTGAA +CGTGAT +CGTGAC +CGTGAG +CGTGTA +CGTGTT +CGTGTC +CGTGTG +CGTGCA +CGTGCT +CGTGCC +CGTGCG +CGTGGA +CGTGGT +CGTGGC +CGTGGG +CGCAAA +CGCAAT +CGCAAC +CGCAAG +CGCATA +CGCATT +CGCATC +CGCATG +CGCACA +CGCACT +CGCACC +CGCACG +CGCAGA +CGCAGT +CGCAGC +CGCAGG +CGCTAA +CGCTAT +CGCTAC +CGCTAG +CGCTTA +CGCTTT +CGCTTC +CGCTTG +CGCTCA +CGCTCT +CGCTCC +CGCTCG +CGCTGA +CGCTGT +CGCTGC +CGCTGG +CGCCAA +CGCCAT +CGCCAC +CGCCAG +CGCCTA +CGCCTT +CGCCTC +CGCCTG +CGCCCA +CGCCCT +CGCCCC +CGCCCG +CGCCGA +CGCCGT +CGCCGC +CGCCGG +CGCGAA +CGCGAT +CGCGAC +CGCGAG +CGCGTA +CGCGTT +CGCGTC +CGCGTG +CGCGCA +CGCGCT +CGCGCC +CGCGCG +CGCGGA +CGCGGT +CGCGGC +CGCGGG +CGGAAA +CGGAAT +CGGAAC +CGGAAG +CGGATA +CGGATT +CGGATC +CGGATG +CGGACA +CGGACT +CGGACC +CGGACG +CGGAGA +CGGAGT +CGGAGC +CGGAGG +CGGTAA +CGGTAT +CGGTAC +CGGTAG +CGGTTA +CGGTTT +CGGTTC +CGGTTG +CGGTCA +CGGTCT +CGGTCC +CGGTCG +CGGTGA +CGGTGT +CGGTGC +CGGTGG +CGGCAA +CGGCAT +CGGCAC +CGGCAG +CGGCTA +CGGCTT +CGGCTC +CGGCTG +CGGCCA +CGGCCT +CGGCCC +CGGCCG +CGGCGA +CGGCGT +CGGCGC +CGGCGG +CGGGAA +CGGGAT +CGGGAC +CGGGAG +CGGGTA +CGGGTT +CGGGTC +CGGGTG +CGGGCA +CGGGCT +CGGGCC +CGGGCG +CGGGGA +CGGGGT +CGGGGC +CGGGGG +GAAAAA +GAAAAT +GAAAAC +GAAAAG +GAAATA +GAAATT +GAAATC +GAAATG +GAAACA +GAAACT +GAAACC +GAAACG +GAAAGA +GAAAGT +GAAAGC +GAAAGG +GAATAA +GAATAT +GAATAC +GAATAG +GAATTA +GAATTT +GAATTC +GAATTG +GAATCA +GAATCT +GAATCC +GAATCG +GAATGA +GAATGT +GAATGC +GAATGG +GAACAA +GAACAT +GAACAC +GAACAG +GAACTA +GAACTT +GAACTC +GAACTG +GAACCA +GAACCT +GAACCC +GAACCG +GAACGA +GAACGT +GAACGC +GAACGG +GAAGAA +GAAGAT +GAAGAC +GAAGAG +GAAGTA +GAAGTT +GAAGTC +GAAGTG +GAAGCA +GAAGCT +GAAGCC +GAAGCG +GAAGGA +GAAGGT +GAAGGC +GAAGGG +GATAAA +GATAAT +GATAAC +GATAAG +GATATA +GATATT +GATATC +GATATG +GATACA +GATACT +GATACC +GATACG +GATAGA +GATAGT +GATAGC +GATAGG +GATTAA +GATTAT +GATTAC +GATTAG +GATTTA +GATTTT +GATTTC +GATTTG +GATTCA +GATTCT +GATTCC +GATTCG +GATTGA +GATTGT +GATTGC +GATTGG +GATCAA +GATCAT +GATCAC +GATCAG +GATCTA +GATCTT +GATCTC +GATCTG +GATCCA +GATCCT +GATCCC +GATCCG +GATCGA +GATCGT +GATCGC +GATCGG +GATGAA +GATGAT +GATGAC +GATGAG +GATGTA +GATGTT +GATGTC +GATGTG +GATGCA +GATGCT +GATGCC +GATGCG +GATGGA +GATGGT +GATGGC +GATGGG +GACAAA +GACAAT +GACAAC +GACAAG +GACATA +GACATT +GACATC +GACATG +GACACA +GACACT +GACACC +GACACG +GACAGA +GACAGT +GACAGC +GACAGG +GACTAA +GACTAT +GACTAC +GACTAG +GACTTA +GACTTT +GACTTC +GACTTG +GACTCA +GACTCT +GACTCC +GACTCG +GACTGA +GACTGT +GACTGC +GACTGG +GACCAA +GACCAT +GACCAC +GACCAG +GACCTA +GACCTT +GACCTC +GACCTG +GACCCA +GACCCT +GACCCC +GACCCG +GACCGA +GACCGT +GACCGC +GACCGG +GACGAA +GACGAT +GACGAC +GACGAG +GACGTA +GACGTT +GACGTC +GACGTG +GACGCA +GACGCT +GACGCC +GACGCG +GACGGA +GACGGT +GACGGC +GACGGG +GAGAAA +GAGAAT +GAGAAC +GAGAAG +GAGATA +GAGATT +GAGATC +GAGATG +GAGACA +GAGACT +GAGACC +GAGACG +GAGAGA +GAGAGT +GAGAGC +GAGAGG +GAGTAA +GAGTAT +GAGTAC +GAGTAG +GAGTTA +GAGTTT +GAGTTC +GAGTTG +GAGTCA +GAGTCT +GAGTCC +GAGTCG +GAGTGA +GAGTGT +GAGTGC +GAGTGG +GAGCAA +GAGCAT +GAGCAC +GAGCAG +GAGCTA +GAGCTT +GAGCTC +GAGCTG +GAGCCA +GAGCCT +GAGCCC +GAGCCG +GAGCGA +GAGCGT +GAGCGC +GAGCGG +GAGGAA +GAGGAT +GAGGAC +GAGGAG +GAGGTA +GAGGTT +GAGGTC +GAGGTG +GAGGCA +GAGGCT +GAGGCC +GAGGCG +GAGGGA +GAGGGT +GAGGGC +GAGGGG +GTAAAA +GTAAAT +GTAAAC +GTAAAG +GTAATA +GTAATT +GTAATC +GTAATG +GTAACA +GTAACT +GTAACC +GTAACG +GTAAGA +GTAAGT +GTAAGC +GTAAGG +GTATAA +GTATAT +GTATAC +GTATAG +GTATTA +GTATTT +GTATTC +GTATTG +GTATCA +GTATCT +GTATCC +GTATCG +GTATGA +GTATGT +GTATGC +GTATGG +GTACAA +GTACAT +GTACAC +GTACAG +GTACTA +GTACTT +GTACTC +GTACTG +GTACCA +GTACCT +GTACCC +GTACCG +GTACGA +GTACGT +GTACGC +GTACGG +GTAGAA +GTAGAT +GTAGAC +GTAGAG +GTAGTA +GTAGTT +GTAGTC +GTAGTG +GTAGCA +GTAGCT +GTAGCC +GTAGCG +GTAGGA +GTAGGT +GTAGGC +GTAGGG +GTTAAA +GTTAAT +GTTAAC +GTTAAG +GTTATA +GTTATT +GTTATC +GTTATG +GTTACA +GTTACT +GTTACC +GTTACG +GTTAGA +GTTAGT +GTTAGC +GTTAGG +GTTTAA +GTTTAT +GTTTAC +GTTTAG +GTTTTA +GTTTTT +GTTTTC +GTTTTG +GTTTCA +GTTTCT +GTTTCC +GTTTCG +GTTTGA +GTTTGT +GTTTGC +GTTTGG +GTTCAA +GTTCAT +GTTCAC +GTTCAG +GTTCTA +GTTCTT +GTTCTC +GTTCTG +GTTCCA +GTTCCT +GTTCCC +GTTCCG +GTTCGA +GTTCGT +GTTCGC +GTTCGG +GTTGAA +GTTGAT +GTTGAC +GTTGAG +GTTGTA +GTTGTT +GTTGTC +GTTGTG +GTTGCA +GTTGCT +GTTGCC +GTTGCG +GTTGGA +GTTGGT +GTTGGC +GTTGGG +GTCAAA +GTCAAT +GTCAAC +GTCAAG +GTCATA +GTCATT +GTCATC +GTCATG +GTCACA +GTCACT +GTCACC +GTCACG +GTCAGA +GTCAGT +GTCAGC +GTCAGG +GTCTAA +GTCTAT +GTCTAC +GTCTAG +GTCTTA +GTCTTT +GTCTTC +GTCTTG +GTCTCA +GTCTCT +GTCTCC +GTCTCG +GTCTGA +GTCTGT +GTCTGC +GTCTGG +GTCCAA +GTCCAT +GTCCAC +GTCCAG +GTCCTA +GTCCTT +GTCCTC +GTCCTG +GTCCCA +GTCCCT +GTCCCC +GTCCCG +GTCCGA +GTCCGT +GTCCGC +GTCCGG +GTCGAA +GTCGAT +GTCGAC +GTCGAG +GTCGTA +GTCGTT +GTCGTC +GTCGTG +GTCGCA +GTCGCT +GTCGCC +GTCGCG +GTCGGA +GTCGGT +GTCGGC +GTCGGG +GTGAAA +GTGAAT +GTGAAC +GTGAAG +GTGATA +GTGATT +GTGATC +GTGATG +GTGACA +GTGACT +GTGACC +GTGACG +GTGAGA +GTGAGT +GTGAGC +GTGAGG +GTGTAA +GTGTAT +GTGTAC +GTGTAG +GTGTTA +GTGTTT +GTGTTC +GTGTTG +GTGTCA +GTGTCT +GTGTCC +GTGTCG +GTGTGA +GTGTGT +GTGTGC +GTGTGG +GTGCAA +GTGCAT +GTGCAC +GTGCAG +GTGCTA +GTGCTT +GTGCTC +GTGCTG +GTGCCA +GTGCCT +GTGCCC +GTGCCG +GTGCGA +GTGCGT +GTGCGC +GTGCGG +GTGGAA +GTGGAT +GTGGAC +GTGGAG +GTGGTA +GTGGTT +GTGGTC +GTGGTG +GTGGCA +GTGGCT +GTGGCC +GTGGCG +GTGGGA +GTGGGT +GTGGGC +GTGGGG +GCAAAA +GCAAAT +GCAAAC +GCAAAG +GCAATA +GCAATT +GCAATC +GCAATG +GCAACA +GCAACT +GCAACC +GCAACG +GCAAGA +GCAAGT +GCAAGC +GCAAGG +GCATAA +GCATAT +GCATAC +GCATAG +GCATTA +GCATTT +GCATTC +GCATTG +GCATCA +GCATCT +GCATCC +GCATCG +GCATGA +GCATGT +GCATGC +GCATGG +GCACAA +GCACAT +GCACAC +GCACAG +GCACTA +GCACTT +GCACTC +GCACTG +GCACCA +GCACCT +GCACCC +GCACCG +GCACGA +GCACGT +GCACGC +GCACGG +GCAGAA +GCAGAT +GCAGAC +GCAGAG +GCAGTA +GCAGTT +GCAGTC +GCAGTG +GCAGCA +GCAGCT +GCAGCC +GCAGCG +GCAGGA +GCAGGT +GCAGGC +GCAGGG +GCTAAA +GCTAAT +GCTAAC +GCTAAG +GCTATA +GCTATT +GCTATC +GCTATG +GCTACA +GCTACT +GCTACC +GCTACG +GCTAGA +GCTAGT +GCTAGC +GCTAGG +GCTTAA +GCTTAT +GCTTAC +GCTTAG +GCTTTA +GCTTTT +GCTTTC +GCTTTG +GCTTCA +GCTTCT +GCTTCC +GCTTCG +GCTTGA +GCTTGT +GCTTGC +GCTTGG +GCTCAA +GCTCAT +GCTCAC +GCTCAG +GCTCTA +GCTCTT +GCTCTC +GCTCTG +GCTCCA +GCTCCT +GCTCCC +GCTCCG +GCTCGA +GCTCGT +GCTCGC +GCTCGG +GCTGAA +GCTGAT +GCTGAC +GCTGAG +GCTGTA +GCTGTT +GCTGTC +GCTGTG +GCTGCA +GCTGCT +GCTGCC +GCTGCG +GCTGGA +GCTGGT +GCTGGC +GCTGGG +GCCAAA +GCCAAT +GCCAAC +GCCAAG +GCCATA +GCCATT +GCCATC +GCCATG +GCCACA +GCCACT +GCCACC +GCCACG +GCCAGA +GCCAGT +GCCAGC +GCCAGG +GCCTAA +GCCTAT +GCCTAC +GCCTAG +GCCTTA +GCCTTT +GCCTTC +GCCTTG +GCCTCA +GCCTCT +GCCTCC +GCCTCG +GCCTGA +GCCTGT +GCCTGC +GCCTGG +GCCCAA +GCCCAT +GCCCAC +GCCCAG +GCCCTA +GCCCTT +GCCCTC +GCCCTG +GCCCCA +GCCCCT +GCCCCC +GCCCCG +GCCCGA +GCCCGT +GCCCGC +GCCCGG +GCCGAA +GCCGAT +GCCGAC +GCCGAG +GCCGTA +GCCGTT +GCCGTC +GCCGTG +GCCGCA +GCCGCT +GCCGCC +GCCGCG +GCCGGA +GCCGGT +GCCGGC +GCCGGG +GCGAAA +GCGAAT +GCGAAC +GCGAAG +GCGATA +GCGATT +GCGATC +GCGATG +GCGACA +GCGACT +GCGACC +GCGACG +GCGAGA +GCGAGT +GCGAGC +GCGAGG +GCGTAA +GCGTAT +GCGTAC +GCGTAG +GCGTTA +GCGTTT +GCGTTC +GCGTTG +GCGTCA +GCGTCT +GCGTCC +GCGTCG +GCGTGA +GCGTGT +GCGTGC +GCGTGG +GCGCAA +GCGCAT +GCGCAC +GCGCAG +GCGCTA +GCGCTT +GCGCTC +GCGCTG +GCGCCA +GCGCCT +GCGCCC +GCGCCG +GCGCGA +GCGCGT +GCGCGC +GCGCGG +GCGGAA +GCGGAT +GCGGAC +GCGGAG +GCGGTA +GCGGTT +GCGGTC +GCGGTG +GCGGCA +GCGGCT +GCGGCC +GCGGCG +GCGGGA +GCGGGT +GCGGGC +GCGGGG +GGAAAA +GGAAAT +GGAAAC +GGAAAG +GGAATA +GGAATT +GGAATC +GGAATG +GGAACA +GGAACT +GGAACC +GGAACG +GGAAGA +GGAAGT +GGAAGC +GGAAGG +GGATAA +GGATAT +GGATAC +GGATAG +GGATTA +GGATTT +GGATTC +GGATTG +GGATCA +GGATCT +GGATCC +GGATCG +GGATGA +GGATGT +GGATGC +GGATGG +GGACAA +GGACAT +GGACAC +GGACAG +GGACTA +GGACTT +GGACTC +GGACTG +GGACCA +GGACCT +GGACCC +GGACCG +GGACGA +GGACGT +GGACGC +GGACGG +GGAGAA +GGAGAT +GGAGAC +GGAGAG +GGAGTA +GGAGTT +GGAGTC +GGAGTG +GGAGCA +GGAGCT +GGAGCC +GGAGCG +GGAGGA +GGAGGT +GGAGGC +GGAGGG +GGTAAA +GGTAAT +GGTAAC +GGTAAG +GGTATA +GGTATT +GGTATC +GGTATG +GGTACA +GGTACT +GGTACC +GGTACG +GGTAGA +GGTAGT +GGTAGC +GGTAGG +GGTTAA +GGTTAT +GGTTAC +GGTTAG +GGTTTA +GGTTTT +GGTTTC +GGTTTG +GGTTCA +GGTTCT +GGTTCC +GGTTCG +GGTTGA +GGTTGT +GGTTGC +GGTTGG +GGTCAA +GGTCAT +GGTCAC +GGTCAG +GGTCTA +GGTCTT +GGTCTC +GGTCTG +GGTCCA +GGTCCT +GGTCCC +GGTCCG +GGTCGA +GGTCGT +GGTCGC +GGTCGG +GGTGAA +GGTGAT +GGTGAC +GGTGAG +GGTGTA +GGTGTT +GGTGTC +GGTGTG +GGTGCA +GGTGCT +GGTGCC +GGTGCG +GGTGGA +GGTGGT +GGTGGC +GGTGGG +GGCAAA +GGCAAT +GGCAAC +GGCAAG +GGCATA +GGCATT +GGCATC +GGCATG +GGCACA +GGCACT +GGCACC +GGCACG +GGCAGA +GGCAGT +GGCAGC +GGCAGG +GGCTAA +GGCTAT +GGCTAC +GGCTAG +GGCTTA +GGCTTT +GGCTTC +GGCTTG +GGCTCA +GGCTCT +GGCTCC +GGCTCG +GGCTGA +GGCTGT +GGCTGC +GGCTGG +GGCCAA +GGCCAT +GGCCAC +GGCCAG +GGCCTA +GGCCTT +GGCCTC +GGCCTG +GGCCCA +GGCCCT +GGCCCC +GGCCCG +GGCCGA +GGCCGT +GGCCGC +GGCCGG +GGCGAA +GGCGAT +GGCGAC +GGCGAG +GGCGTA +GGCGTT +GGCGTC +GGCGTG +GGCGCA +GGCGCT +GGCGCC +GGCGCG +GGCGGA +GGCGGT +GGCGGC +GGCGGG +GGGAAA +GGGAAT +GGGAAC +GGGAAG +GGGATA +GGGATT +GGGATC +GGGATG +GGGACA +GGGACT +GGGACC +GGGACG +GGGAGA +GGGAGT +GGGAGC +GGGAGG +GGGTAA +GGGTAT +GGGTAC +GGGTAG +GGGTTA +GGGTTT +GGGTTC +GGGTTG +GGGTCA +GGGTCT +GGGTCC +GGGTCG +GGGTGA +GGGTGT +GGGTGC +GGGTGG +GGGCAA +GGGCAT +GGGCAC +GGGCAG +GGGCTA +GGGCTT +GGGCTC +GGGCTG +GGGCCA +GGGCCT +GGGCCC +GGGCCG +GGGCGA +GGGCGT +GGGCGC +GGGCGG +GGGGAA +GGGGAT +GGGGAC +GGGGAG +GGGGTA +GGGGTT +GGGGTC +GGGGTG +GGGGCA +GGGGCT +GGGGCC +GGGGCG +GGGGGA +GGGGGT +GGGGGC +GGGGGG diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-1000/config.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-1000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..9b14e61d327bf586fcff6e817cf9d365bae18ad2 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-1000/config.json @@ -0,0 +1,26 @@ +{ + "_name_or_path": "fabihamakhdoomi/TinyDNABERT", + "architectures": [ + "BertForSequenceClassification" + ], + "attention_probs_dropout_prob": 0.1, + "classifier_dropout": null, + "hidden_act": "gelu", + "hidden_dropout_prob": 0.1, + "hidden_size": 128, + "initializer_range": 0.02, + "intermediate_size": 512, + "layer_norm_eps": 1e-12, + "max_position_embeddings": 512, + "model_type": "bert", + "num_attention_heads": 2, + "num_hidden_layers": 2, + "pad_token_id": 0, + "position_embedding_type": "absolute", + "problem_type": "single_label_classification", + "torch_dtype": "float32", + "transformers_version": "4.44.0", + "type_vocab_size": 2, + "use_cache": true, + "vocab_size": 4101 +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-1000/model.safetensors b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-1000/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..c25ccb7e23d276e1814fec4364a46613fcaff421 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-1000/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1c55fb79957b6532642702d2dd6ddde50fd29aa0db58508dff44ed95762fd67c +size 4021688 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-1000/optimizer.pt b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-1000/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..260a40c46b9a6d1a2d0afc77f959022b64d2508a --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-1000/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2f6f07f7b9452142eee9a5d1012901cccf32657eafab537d520a524a5351e15a +size 8058700 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-1000/rng_state.pth b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-1000/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..031c7b1895cd8940e942badcb833c16894800a59 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-1000/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e570743f570544924d9d30f7a79905faa2dba99076a5c64ec39e4f6a500cf60e +size 14244 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-1000/scheduler.pt b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-1000/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..e21fdaa14c288938c4b2c1277194110b759d8724 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-1000/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7480edaf7ecb5003556fedbf146cf26ec75a9ee19931ce031bbb3b509f9927ce +size 1064 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-1000/special_tokens_map.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-1000/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..a8b3208c2884c4efb86e49300fdd3dc877220cdf --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-1000/special_tokens_map.json @@ -0,0 +1,7 @@ +{ + "cls_token": "[CLS]", + "mask_token": "[MASK]", + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-1000/tokenizer.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-1000/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..ebd782de6963b9032ce9374cfe30f5275be1a30b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-1000/tokenizer.json @@ -0,0 +1,4265 @@ +{ + "version": "1.0", + "truncation": { + "direction": "Right", + "max_length": 512, + "strategy": "LongestFirst", + "stride": 0 + }, + "padding": { + "strategy": { + "Fixed": 512 + }, + "direction": "Right", + "pad_to_multiple_of": null, + "pad_id": 0, + "pad_type_id": 0, + "pad_token": "[PAD]" + }, + "added_tokens": [ + { + "id": 0, + "content": "[PAD]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 1, + "content": "[UNK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 2, + "content": "[CLS]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 3, + "content": "[SEP]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 4, + "content": "[MASK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + } + ], + "normalizer": { + "type": "BertNormalizer", + "clean_text": true, + "handle_chinese_chars": true, + "strip_accents": null, + "lowercase": false + }, + "pre_tokenizer": { + "type": "BertPreTokenizer" + }, + "post_processor": { + "type": "TemplateProcessing", + "single": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + } + ], + "pair": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "B", + "type_id": 1 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 1 + } + } + ], + "special_tokens": { + "[CLS]": { + "id": "[CLS]", + "ids": [ + 2 + ], + "tokens": [ + "[CLS]" + ] + }, + "[SEP]": { + "id": "[SEP]", + "ids": [ + 3 + ], + "tokens": [ + "[SEP]" + ] + } + } + }, + "decoder": { + "type": "WordPiece", + "prefix": "##", + "cleanup": true + }, + "model": { + "type": "WordPiece", + "unk_token": "[UNK]", + "continuing_subword_prefix": "##", + "max_input_chars_per_word": 100, + "vocab": { + "[PAD]": 0, + "[UNK]": 1, + "[CLS]": 2, + "[SEP]": 3, + "[MASK]": 4, + "AAAAAA": 5, + "AAAAAT": 6, + "AAAAAC": 7, + "AAAAAG": 8, + "AAAATA": 9, + "AAAATT": 10, + "AAAATC": 11, + "AAAATG": 12, + "AAAACA": 13, + "AAAACT": 14, + "AAAACC": 15, + "AAAACG": 16, + "AAAAGA": 17, + "AAAAGT": 18, + "AAAAGC": 19, + "AAAAGG": 20, + "AAATAA": 21, + "AAATAT": 22, + "AAATAC": 23, + "AAATAG": 24, + "AAATTA": 25, + "AAATTT": 26, + "AAATTC": 27, + "AAATTG": 28, + "AAATCA": 29, + "AAATCT": 30, + "AAATCC": 31, + "AAATCG": 32, + "AAATGA": 33, + "AAATGT": 34, + "AAATGC": 35, + "AAATGG": 36, + "AAACAA": 37, + "AAACAT": 38, + "AAACAC": 39, + "AAACAG": 40, + "AAACTA": 41, + "AAACTT": 42, + "AAACTC": 43, + "AAACTG": 44, + "AAACCA": 45, + "AAACCT": 46, + "AAACCC": 47, + "AAACCG": 48, + "AAACGA": 49, + "AAACGT": 50, + "AAACGC": 51, + "AAACGG": 52, + "AAAGAA": 53, + "AAAGAT": 54, + "AAAGAC": 55, + "AAAGAG": 56, + "AAAGTA": 57, + "AAAGTT": 58, + "AAAGTC": 59, + "AAAGTG": 60, + "AAAGCA": 61, + "AAAGCT": 62, + "AAAGCC": 63, + "AAAGCG": 64, + "AAAGGA": 65, + "AAAGGT": 66, + "AAAGGC": 67, + "AAAGGG": 68, + "AATAAA": 69, + "AATAAT": 70, + "AATAAC": 71, + "AATAAG": 72, + "AATATA": 73, + "AATATT": 74, + "AATATC": 75, + "AATATG": 76, + "AATACA": 77, + "AATACT": 78, + "AATACC": 79, + "AATACG": 80, + "AATAGA": 81, + "AATAGT": 82, + "AATAGC": 83, + "AATAGG": 84, + "AATTAA": 85, + "AATTAT": 86, + "AATTAC": 87, + "AATTAG": 88, + "AATTTA": 89, + "AATTTT": 90, + "AATTTC": 91, + "AATTTG": 92, + "AATTCA": 93, + "AATTCT": 94, + "AATTCC": 95, + "AATTCG": 96, + "AATTGA": 97, + "AATTGT": 98, + "AATTGC": 99, + "AATTGG": 100, + "AATCAA": 101, + "AATCAT": 102, + "AATCAC": 103, + "AATCAG": 104, + "AATCTA": 105, + "AATCTT": 106, + "AATCTC": 107, + "AATCTG": 108, + "AATCCA": 109, + "AATCCT": 110, + "AATCCC": 111, + "AATCCG": 112, + "AATCGA": 113, + "AATCGT": 114, + "AATCGC": 115, + "AATCGG": 116, + "AATGAA": 117, + "AATGAT": 118, + "AATGAC": 119, + "AATGAG": 120, + "AATGTA": 121, + "AATGTT": 122, + "AATGTC": 123, + "AATGTG": 124, + "AATGCA": 125, + "AATGCT": 126, + "AATGCC": 127, + "AATGCG": 128, + "AATGGA": 129, + "AATGGT": 130, + "AATGGC": 131, + "AATGGG": 132, + "AACAAA": 133, + "AACAAT": 134, + "AACAAC": 135, + "AACAAG": 136, + "AACATA": 137, + "AACATT": 138, + "AACATC": 139, + "AACATG": 140, + "AACACA": 141, + "AACACT": 142, + "AACACC": 143, + "AACACG": 144, + "AACAGA": 145, + "AACAGT": 146, + "AACAGC": 147, + "AACAGG": 148, + "AACTAA": 149, + "AACTAT": 150, + "AACTAC": 151, + "AACTAG": 152, + "AACTTA": 153, + "AACTTT": 154, + "AACTTC": 155, + "AACTTG": 156, + "AACTCA": 157, + "AACTCT": 158, + "AACTCC": 159, + "AACTCG": 160, + "AACTGA": 161, + "AACTGT": 162, + "AACTGC": 163, + "AACTGG": 164, + "AACCAA": 165, + "AACCAT": 166, + "AACCAC": 167, + "AACCAG": 168, + "AACCTA": 169, + "AACCTT": 170, + "AACCTC": 171, + "AACCTG": 172, + "AACCCA": 173, + "AACCCT": 174, + "AACCCC": 175, + "AACCCG": 176, + "AACCGA": 177, + "AACCGT": 178, + "AACCGC": 179, + "AACCGG": 180, + "AACGAA": 181, + "AACGAT": 182, + "AACGAC": 183, + "AACGAG": 184, + "AACGTA": 185, + "AACGTT": 186, + "AACGTC": 187, + "AACGTG": 188, + "AACGCA": 189, + "AACGCT": 190, + "AACGCC": 191, + "AACGCG": 192, + "AACGGA": 193, + "AACGGT": 194, + "AACGGC": 195, + "AACGGG": 196, + "AAGAAA": 197, + "AAGAAT": 198, + "AAGAAC": 199, + "AAGAAG": 200, + "AAGATA": 201, + "AAGATT": 202, + "AAGATC": 203, + "AAGATG": 204, + "AAGACA": 205, + "AAGACT": 206, + "AAGACC": 207, + "AAGACG": 208, + "AAGAGA": 209, + "AAGAGT": 210, + "AAGAGC": 211, + "AAGAGG": 212, + "AAGTAA": 213, + "AAGTAT": 214, + "AAGTAC": 215, + "AAGTAG": 216, + "AAGTTA": 217, + "AAGTTT": 218, + "AAGTTC": 219, + "AAGTTG": 220, + "AAGTCA": 221, + "AAGTCT": 222, + "AAGTCC": 223, + "AAGTCG": 224, + "AAGTGA": 225, + "AAGTGT": 226, + "AAGTGC": 227, + "AAGTGG": 228, + "AAGCAA": 229, + "AAGCAT": 230, + "AAGCAC": 231, + "AAGCAG": 232, + "AAGCTA": 233, + "AAGCTT": 234, + "AAGCTC": 235, + "AAGCTG": 236, + "AAGCCA": 237, + "AAGCCT": 238, + "AAGCCC": 239, + "AAGCCG": 240, + "AAGCGA": 241, + "AAGCGT": 242, + "AAGCGC": 243, + "AAGCGG": 244, + "AAGGAA": 245, + "AAGGAT": 246, + "AAGGAC": 247, + "AAGGAG": 248, + "AAGGTA": 249, + "AAGGTT": 250, + "AAGGTC": 251, + "AAGGTG": 252, + "AAGGCA": 253, + "AAGGCT": 254, + "AAGGCC": 255, + "AAGGCG": 256, + "AAGGGA": 257, + "AAGGGT": 258, + "AAGGGC": 259, + "AAGGGG": 260, + "ATAAAA": 261, + "ATAAAT": 262, + "ATAAAC": 263, + "ATAAAG": 264, + "ATAATA": 265, + "ATAATT": 266, + "ATAATC": 267, + "ATAATG": 268, + "ATAACA": 269, + "ATAACT": 270, + "ATAACC": 271, + "ATAACG": 272, + "ATAAGA": 273, + "ATAAGT": 274, + "ATAAGC": 275, + "ATAAGG": 276, + "ATATAA": 277, + "ATATAT": 278, + "ATATAC": 279, + "ATATAG": 280, + "ATATTA": 281, + "ATATTT": 282, + "ATATTC": 283, + "ATATTG": 284, + "ATATCA": 285, + "ATATCT": 286, + "ATATCC": 287, + "ATATCG": 288, + "ATATGA": 289, + "ATATGT": 290, + "ATATGC": 291, + "ATATGG": 292, + "ATACAA": 293, + "ATACAT": 294, + "ATACAC": 295, + "ATACAG": 296, + "ATACTA": 297, + "ATACTT": 298, + "ATACTC": 299, + "ATACTG": 300, + "ATACCA": 301, + "ATACCT": 302, + "ATACCC": 303, + "ATACCG": 304, + "ATACGA": 305, + "ATACGT": 306, + "ATACGC": 307, + "ATACGG": 308, + "ATAGAA": 309, + "ATAGAT": 310, + "ATAGAC": 311, + "ATAGAG": 312, + "ATAGTA": 313, + "ATAGTT": 314, + "ATAGTC": 315, + "ATAGTG": 316, + "ATAGCA": 317, + "ATAGCT": 318, + "ATAGCC": 319, + "ATAGCG": 320, + "ATAGGA": 321, + "ATAGGT": 322, + "ATAGGC": 323, + "ATAGGG": 324, + "ATTAAA": 325, + "ATTAAT": 326, + "ATTAAC": 327, + "ATTAAG": 328, + "ATTATA": 329, + "ATTATT": 330, + "ATTATC": 331, + "ATTATG": 332, + "ATTACA": 333, + "ATTACT": 334, + "ATTACC": 335, + "ATTACG": 336, + "ATTAGA": 337, + "ATTAGT": 338, + "ATTAGC": 339, + "ATTAGG": 340, + "ATTTAA": 341, + "ATTTAT": 342, + "ATTTAC": 343, + "ATTTAG": 344, + "ATTTTA": 345, + "ATTTTT": 346, + "ATTTTC": 347, + "ATTTTG": 348, + "ATTTCA": 349, + "ATTTCT": 350, + "ATTTCC": 351, + "ATTTCG": 352, + "ATTTGA": 353, + "ATTTGT": 354, + "ATTTGC": 355, + "ATTTGG": 356, + "ATTCAA": 357, + "ATTCAT": 358, + "ATTCAC": 359, + "ATTCAG": 360, + "ATTCTA": 361, + "ATTCTT": 362, + "ATTCTC": 363, + "ATTCTG": 364, + "ATTCCA": 365, + "ATTCCT": 366, + "ATTCCC": 367, + "ATTCCG": 368, + "ATTCGA": 369, + "ATTCGT": 370, + "ATTCGC": 371, + "ATTCGG": 372, + "ATTGAA": 373, + "ATTGAT": 374, + "ATTGAC": 375, + "ATTGAG": 376, + "ATTGTA": 377, + "ATTGTT": 378, + "ATTGTC": 379, + "ATTGTG": 380, + "ATTGCA": 381, + "ATTGCT": 382, + "ATTGCC": 383, + "ATTGCG": 384, + "ATTGGA": 385, + "ATTGGT": 386, + "ATTGGC": 387, + "ATTGGG": 388, + "ATCAAA": 389, + "ATCAAT": 390, + "ATCAAC": 391, + "ATCAAG": 392, + "ATCATA": 393, + "ATCATT": 394, + "ATCATC": 395, + "ATCATG": 396, + "ATCACA": 397, + "ATCACT": 398, + "ATCACC": 399, + "ATCACG": 400, + "ATCAGA": 401, + "ATCAGT": 402, + "ATCAGC": 403, + "ATCAGG": 404, + "ATCTAA": 405, + "ATCTAT": 406, + "ATCTAC": 407, + "ATCTAG": 408, + "ATCTTA": 409, + "ATCTTT": 410, + "ATCTTC": 411, + "ATCTTG": 412, + "ATCTCA": 413, + "ATCTCT": 414, + "ATCTCC": 415, + "ATCTCG": 416, + "ATCTGA": 417, + "ATCTGT": 418, + "ATCTGC": 419, + "ATCTGG": 420, + "ATCCAA": 421, + "ATCCAT": 422, + "ATCCAC": 423, + "ATCCAG": 424, + "ATCCTA": 425, + "ATCCTT": 426, + "ATCCTC": 427, + "ATCCTG": 428, + "ATCCCA": 429, + "ATCCCT": 430, + "ATCCCC": 431, + "ATCCCG": 432, + "ATCCGA": 433, + "ATCCGT": 434, + "ATCCGC": 435, + "ATCCGG": 436, + "ATCGAA": 437, + "ATCGAT": 438, + "ATCGAC": 439, + "ATCGAG": 440, + "ATCGTA": 441, + "ATCGTT": 442, + "ATCGTC": 443, + "ATCGTG": 444, + "ATCGCA": 445, + "ATCGCT": 446, + "ATCGCC": 447, + "ATCGCG": 448, + "ATCGGA": 449, + "ATCGGT": 450, + "ATCGGC": 451, + "ATCGGG": 452, + "ATGAAA": 453, + "ATGAAT": 454, + "ATGAAC": 455, + "ATGAAG": 456, + "ATGATA": 457, + "ATGATT": 458, + "ATGATC": 459, + "ATGATG": 460, + "ATGACA": 461, + "ATGACT": 462, + "ATGACC": 463, + "ATGACG": 464, + "ATGAGA": 465, + "ATGAGT": 466, + "ATGAGC": 467, + "ATGAGG": 468, + "ATGTAA": 469, + "ATGTAT": 470, + "ATGTAC": 471, + "ATGTAG": 472, + "ATGTTA": 473, + "ATGTTT": 474, + "ATGTTC": 475, + "ATGTTG": 476, + "ATGTCA": 477, + "ATGTCT": 478, + "ATGTCC": 479, + "ATGTCG": 480, + "ATGTGA": 481, + "ATGTGT": 482, + "ATGTGC": 483, + "ATGTGG": 484, + "ATGCAA": 485, + "ATGCAT": 486, + "ATGCAC": 487, + "ATGCAG": 488, + "ATGCTA": 489, + "ATGCTT": 490, + "ATGCTC": 491, + "ATGCTG": 492, + "ATGCCA": 493, + "ATGCCT": 494, + "ATGCCC": 495, + "ATGCCG": 496, + "ATGCGA": 497, + "ATGCGT": 498, + "ATGCGC": 499, + "ATGCGG": 500, + "ATGGAA": 501, + "ATGGAT": 502, + "ATGGAC": 503, + "ATGGAG": 504, + "ATGGTA": 505, + "ATGGTT": 506, + "ATGGTC": 507, + "ATGGTG": 508, + "ATGGCA": 509, + "ATGGCT": 510, + "ATGGCC": 511, + "ATGGCG": 512, + "ATGGGA": 513, + "ATGGGT": 514, + "ATGGGC": 515, + "ATGGGG": 516, + "ACAAAA": 517, + "ACAAAT": 518, + "ACAAAC": 519, + "ACAAAG": 520, + "ACAATA": 521, + "ACAATT": 522, + "ACAATC": 523, + "ACAATG": 524, + "ACAACA": 525, + "ACAACT": 526, + "ACAACC": 527, + "ACAACG": 528, + "ACAAGA": 529, + "ACAAGT": 530, + "ACAAGC": 531, + "ACAAGG": 532, + "ACATAA": 533, + "ACATAT": 534, + "ACATAC": 535, + "ACATAG": 536, + "ACATTA": 537, + "ACATTT": 538, + "ACATTC": 539, + "ACATTG": 540, + "ACATCA": 541, + "ACATCT": 542, + "ACATCC": 543, + "ACATCG": 544, + "ACATGA": 545, + "ACATGT": 546, + "ACATGC": 547, + "ACATGG": 548, + "ACACAA": 549, + "ACACAT": 550, + "ACACAC": 551, + "ACACAG": 552, + "ACACTA": 553, + "ACACTT": 554, + "ACACTC": 555, + "ACACTG": 556, + "ACACCA": 557, + "ACACCT": 558, + "ACACCC": 559, + "ACACCG": 560, + "ACACGA": 561, + "ACACGT": 562, + "ACACGC": 563, + "ACACGG": 564, + "ACAGAA": 565, + "ACAGAT": 566, + "ACAGAC": 567, + "ACAGAG": 568, + "ACAGTA": 569, + "ACAGTT": 570, + "ACAGTC": 571, + "ACAGTG": 572, + "ACAGCA": 573, + "ACAGCT": 574, + "ACAGCC": 575, + "ACAGCG": 576, + "ACAGGA": 577, + "ACAGGT": 578, + "ACAGGC": 579, + "ACAGGG": 580, + "ACTAAA": 581, + "ACTAAT": 582, + "ACTAAC": 583, + "ACTAAG": 584, + "ACTATA": 585, + "ACTATT": 586, + "ACTATC": 587, + "ACTATG": 588, + "ACTACA": 589, + "ACTACT": 590, + "ACTACC": 591, + "ACTACG": 592, + "ACTAGA": 593, + "ACTAGT": 594, + "ACTAGC": 595, + "ACTAGG": 596, + "ACTTAA": 597, + "ACTTAT": 598, + "ACTTAC": 599, + "ACTTAG": 600, + "ACTTTA": 601, + "ACTTTT": 602, + "ACTTTC": 603, + "ACTTTG": 604, + "ACTTCA": 605, + "ACTTCT": 606, + "ACTTCC": 607, + "ACTTCG": 608, + "ACTTGA": 609, + "ACTTGT": 610, + "ACTTGC": 611, + "ACTTGG": 612, + "ACTCAA": 613, + "ACTCAT": 614, + "ACTCAC": 615, + "ACTCAG": 616, + "ACTCTA": 617, + "ACTCTT": 618, + "ACTCTC": 619, + "ACTCTG": 620, + "ACTCCA": 621, + "ACTCCT": 622, + "ACTCCC": 623, + "ACTCCG": 624, + "ACTCGA": 625, + "ACTCGT": 626, + "ACTCGC": 627, + "ACTCGG": 628, + "ACTGAA": 629, + "ACTGAT": 630, + "ACTGAC": 631, + "ACTGAG": 632, + "ACTGTA": 633, + "ACTGTT": 634, + "ACTGTC": 635, + "ACTGTG": 636, + "ACTGCA": 637, + "ACTGCT": 638, + "ACTGCC": 639, + "ACTGCG": 640, + "ACTGGA": 641, + "ACTGGT": 642, + "ACTGGC": 643, + "ACTGGG": 644, + "ACCAAA": 645, + "ACCAAT": 646, + "ACCAAC": 647, + "ACCAAG": 648, + "ACCATA": 649, + "ACCATT": 650, + "ACCATC": 651, + "ACCATG": 652, + "ACCACA": 653, + "ACCACT": 654, + "ACCACC": 655, + "ACCACG": 656, + "ACCAGA": 657, + "ACCAGT": 658, + "ACCAGC": 659, + "ACCAGG": 660, + "ACCTAA": 661, + "ACCTAT": 662, + "ACCTAC": 663, + "ACCTAG": 664, + "ACCTTA": 665, + "ACCTTT": 666, + "ACCTTC": 667, + "ACCTTG": 668, + "ACCTCA": 669, + "ACCTCT": 670, + "ACCTCC": 671, + "ACCTCG": 672, + "ACCTGA": 673, + "ACCTGT": 674, + "ACCTGC": 675, + "ACCTGG": 676, + "ACCCAA": 677, + "ACCCAT": 678, + "ACCCAC": 679, + "ACCCAG": 680, + "ACCCTA": 681, + "ACCCTT": 682, + "ACCCTC": 683, + "ACCCTG": 684, + "ACCCCA": 685, + "ACCCCT": 686, + "ACCCCC": 687, + "ACCCCG": 688, + "ACCCGA": 689, + "ACCCGT": 690, + "ACCCGC": 691, + "ACCCGG": 692, + "ACCGAA": 693, + "ACCGAT": 694, + "ACCGAC": 695, + "ACCGAG": 696, + "ACCGTA": 697, + "ACCGTT": 698, + "ACCGTC": 699, + "ACCGTG": 700, + "ACCGCA": 701, + "ACCGCT": 702, + "ACCGCC": 703, + "ACCGCG": 704, + "ACCGGA": 705, + "ACCGGT": 706, + "ACCGGC": 707, + "ACCGGG": 708, + "ACGAAA": 709, + "ACGAAT": 710, + "ACGAAC": 711, + "ACGAAG": 712, + "ACGATA": 713, + "ACGATT": 714, + "ACGATC": 715, + "ACGATG": 716, + "ACGACA": 717, + "ACGACT": 718, + "ACGACC": 719, + "ACGACG": 720, + "ACGAGA": 721, + "ACGAGT": 722, + "ACGAGC": 723, + "ACGAGG": 724, + "ACGTAA": 725, + "ACGTAT": 726, + "ACGTAC": 727, + "ACGTAG": 728, + "ACGTTA": 729, + "ACGTTT": 730, + "ACGTTC": 731, + "ACGTTG": 732, + "ACGTCA": 733, + "ACGTCT": 734, + "ACGTCC": 735, + "ACGTCG": 736, + "ACGTGA": 737, + "ACGTGT": 738, + "ACGTGC": 739, + "ACGTGG": 740, + "ACGCAA": 741, + "ACGCAT": 742, + "ACGCAC": 743, + "ACGCAG": 744, + "ACGCTA": 745, + "ACGCTT": 746, + "ACGCTC": 747, + "ACGCTG": 748, + "ACGCCA": 749, + "ACGCCT": 750, + "ACGCCC": 751, + "ACGCCG": 752, + "ACGCGA": 753, + "ACGCGT": 754, + "ACGCGC": 755, + "ACGCGG": 756, + "ACGGAA": 757, + "ACGGAT": 758, + "ACGGAC": 759, + "ACGGAG": 760, + "ACGGTA": 761, + "ACGGTT": 762, + "ACGGTC": 763, + "ACGGTG": 764, + "ACGGCA": 765, + "ACGGCT": 766, + "ACGGCC": 767, + "ACGGCG": 768, + "ACGGGA": 769, + "ACGGGT": 770, + "ACGGGC": 771, + "ACGGGG": 772, + "AGAAAA": 773, + "AGAAAT": 774, + "AGAAAC": 775, + "AGAAAG": 776, + "AGAATA": 777, + "AGAATT": 778, + "AGAATC": 779, + "AGAATG": 780, + "AGAACA": 781, + "AGAACT": 782, + "AGAACC": 783, + "AGAACG": 784, + "AGAAGA": 785, + "AGAAGT": 786, + "AGAAGC": 787, + "AGAAGG": 788, + "AGATAA": 789, + "AGATAT": 790, + "AGATAC": 791, + "AGATAG": 792, + "AGATTA": 793, + "AGATTT": 794, + "AGATTC": 795, + "AGATTG": 796, + "AGATCA": 797, + "AGATCT": 798, + "AGATCC": 799, + "AGATCG": 800, + "AGATGA": 801, + "AGATGT": 802, + "AGATGC": 803, + "AGATGG": 804, + "AGACAA": 805, + "AGACAT": 806, + "AGACAC": 807, + "AGACAG": 808, + "AGACTA": 809, + "AGACTT": 810, + "AGACTC": 811, + "AGACTG": 812, + "AGACCA": 813, + "AGACCT": 814, + "AGACCC": 815, + "AGACCG": 816, + "AGACGA": 817, + "AGACGT": 818, + "AGACGC": 819, + "AGACGG": 820, + "AGAGAA": 821, + "AGAGAT": 822, + "AGAGAC": 823, + "AGAGAG": 824, + "AGAGTA": 825, + "AGAGTT": 826, + "AGAGTC": 827, + "AGAGTG": 828, + "AGAGCA": 829, + "AGAGCT": 830, + "AGAGCC": 831, + "AGAGCG": 832, + "AGAGGA": 833, + "AGAGGT": 834, + "AGAGGC": 835, + "AGAGGG": 836, + "AGTAAA": 837, + "AGTAAT": 838, + "AGTAAC": 839, + "AGTAAG": 840, + "AGTATA": 841, + "AGTATT": 842, + "AGTATC": 843, + "AGTATG": 844, + "AGTACA": 845, + "AGTACT": 846, + "AGTACC": 847, + "AGTACG": 848, + "AGTAGA": 849, + "AGTAGT": 850, + "AGTAGC": 851, + "AGTAGG": 852, + "AGTTAA": 853, + "AGTTAT": 854, + "AGTTAC": 855, + "AGTTAG": 856, + "AGTTTA": 857, + "AGTTTT": 858, + "AGTTTC": 859, + "AGTTTG": 860, + "AGTTCA": 861, + "AGTTCT": 862, + "AGTTCC": 863, + "AGTTCG": 864, + "AGTTGA": 865, + "AGTTGT": 866, + "AGTTGC": 867, + "AGTTGG": 868, + "AGTCAA": 869, + "AGTCAT": 870, + "AGTCAC": 871, + "AGTCAG": 872, + "AGTCTA": 873, + "AGTCTT": 874, + "AGTCTC": 875, + "AGTCTG": 876, + "AGTCCA": 877, + "AGTCCT": 878, + "AGTCCC": 879, + "AGTCCG": 880, + "AGTCGA": 881, + "AGTCGT": 882, + "AGTCGC": 883, + "AGTCGG": 884, + "AGTGAA": 885, + "AGTGAT": 886, + "AGTGAC": 887, + "AGTGAG": 888, + "AGTGTA": 889, + "AGTGTT": 890, + "AGTGTC": 891, + "AGTGTG": 892, + "AGTGCA": 893, + "AGTGCT": 894, + "AGTGCC": 895, + "AGTGCG": 896, + "AGTGGA": 897, + "AGTGGT": 898, + "AGTGGC": 899, + "AGTGGG": 900, + "AGCAAA": 901, + "AGCAAT": 902, + "AGCAAC": 903, + "AGCAAG": 904, + "AGCATA": 905, + "AGCATT": 906, + "AGCATC": 907, + "AGCATG": 908, + "AGCACA": 909, + "AGCACT": 910, + "AGCACC": 911, + "AGCACG": 912, + "AGCAGA": 913, + "AGCAGT": 914, + "AGCAGC": 915, + "AGCAGG": 916, + "AGCTAA": 917, + "AGCTAT": 918, + "AGCTAC": 919, + "AGCTAG": 920, + "AGCTTA": 921, + "AGCTTT": 922, + "AGCTTC": 923, + "AGCTTG": 924, + "AGCTCA": 925, + "AGCTCT": 926, + "AGCTCC": 927, + "AGCTCG": 928, + "AGCTGA": 929, + "AGCTGT": 930, + "AGCTGC": 931, + "AGCTGG": 932, + "AGCCAA": 933, + "AGCCAT": 934, + "AGCCAC": 935, + "AGCCAG": 936, + "AGCCTA": 937, + "AGCCTT": 938, + "AGCCTC": 939, + "AGCCTG": 940, + "AGCCCA": 941, + "AGCCCT": 942, + "AGCCCC": 943, + "AGCCCG": 944, + "AGCCGA": 945, + "AGCCGT": 946, + "AGCCGC": 947, + "AGCCGG": 948, + "AGCGAA": 949, + "AGCGAT": 950, + "AGCGAC": 951, + "AGCGAG": 952, + "AGCGTA": 953, + "AGCGTT": 954, + "AGCGTC": 955, + "AGCGTG": 956, + "AGCGCA": 957, + "AGCGCT": 958, + "AGCGCC": 959, + "AGCGCG": 960, + "AGCGGA": 961, + "AGCGGT": 962, + "AGCGGC": 963, + "AGCGGG": 964, + "AGGAAA": 965, + "AGGAAT": 966, + "AGGAAC": 967, + "AGGAAG": 968, + "AGGATA": 969, + "AGGATT": 970, + "AGGATC": 971, + "AGGATG": 972, + "AGGACA": 973, + "AGGACT": 974, + "AGGACC": 975, + "AGGACG": 976, + "AGGAGA": 977, + "AGGAGT": 978, + "AGGAGC": 979, + "AGGAGG": 980, + "AGGTAA": 981, + "AGGTAT": 982, + "AGGTAC": 983, + "AGGTAG": 984, + "AGGTTA": 985, + "AGGTTT": 986, + "AGGTTC": 987, + "AGGTTG": 988, + "AGGTCA": 989, + "AGGTCT": 990, + "AGGTCC": 991, + "AGGTCG": 992, + "AGGTGA": 993, + "AGGTGT": 994, + "AGGTGC": 995, + "AGGTGG": 996, + "AGGCAA": 997, + "AGGCAT": 998, + "AGGCAC": 999, + "AGGCAG": 1000, + "AGGCTA": 1001, + "AGGCTT": 1002, + "AGGCTC": 1003, + "AGGCTG": 1004, + "AGGCCA": 1005, + "AGGCCT": 1006, + "AGGCCC": 1007, + "AGGCCG": 1008, + "AGGCGA": 1009, + "AGGCGT": 1010, + "AGGCGC": 1011, + "AGGCGG": 1012, + "AGGGAA": 1013, + "AGGGAT": 1014, + "AGGGAC": 1015, + "AGGGAG": 1016, + "AGGGTA": 1017, + "AGGGTT": 1018, + "AGGGTC": 1019, + "AGGGTG": 1020, + "AGGGCA": 1021, + "AGGGCT": 1022, + "AGGGCC": 1023, + "AGGGCG": 1024, + "AGGGGA": 1025, + "AGGGGT": 1026, + "AGGGGC": 1027, + "AGGGGG": 1028, + "TAAAAA": 1029, + "TAAAAT": 1030, + "TAAAAC": 1031, + "TAAAAG": 1032, + "TAAATA": 1033, + "TAAATT": 1034, + "TAAATC": 1035, + "TAAATG": 1036, + "TAAACA": 1037, + "TAAACT": 1038, + "TAAACC": 1039, + "TAAACG": 1040, + "TAAAGA": 1041, + "TAAAGT": 1042, + "TAAAGC": 1043, + "TAAAGG": 1044, + "TAATAA": 1045, + "TAATAT": 1046, + "TAATAC": 1047, + "TAATAG": 1048, + "TAATTA": 1049, + "TAATTT": 1050, + "TAATTC": 1051, + "TAATTG": 1052, + "TAATCA": 1053, + "TAATCT": 1054, + "TAATCC": 1055, + "TAATCG": 1056, + "TAATGA": 1057, + "TAATGT": 1058, + "TAATGC": 1059, + "TAATGG": 1060, + "TAACAA": 1061, + "TAACAT": 1062, + "TAACAC": 1063, + "TAACAG": 1064, + "TAACTA": 1065, + "TAACTT": 1066, + "TAACTC": 1067, + "TAACTG": 1068, + "TAACCA": 1069, + "TAACCT": 1070, + "TAACCC": 1071, + "TAACCG": 1072, + "TAACGA": 1073, + "TAACGT": 1074, + "TAACGC": 1075, + "TAACGG": 1076, + "TAAGAA": 1077, + "TAAGAT": 1078, + "TAAGAC": 1079, + "TAAGAG": 1080, + "TAAGTA": 1081, + "TAAGTT": 1082, + "TAAGTC": 1083, + "TAAGTG": 1084, + "TAAGCA": 1085, + "TAAGCT": 1086, + "TAAGCC": 1087, + "TAAGCG": 1088, + "TAAGGA": 1089, + "TAAGGT": 1090, + "TAAGGC": 1091, + "TAAGGG": 1092, + "TATAAA": 1093, + "TATAAT": 1094, + "TATAAC": 1095, + "TATAAG": 1096, + "TATATA": 1097, + "TATATT": 1098, + "TATATC": 1099, + "TATATG": 1100, + "TATACA": 1101, + "TATACT": 1102, + "TATACC": 1103, + "TATACG": 1104, + "TATAGA": 1105, + "TATAGT": 1106, + "TATAGC": 1107, + "TATAGG": 1108, + "TATTAA": 1109, + "TATTAT": 1110, + "TATTAC": 1111, + "TATTAG": 1112, + "TATTTA": 1113, + "TATTTT": 1114, + "TATTTC": 1115, + "TATTTG": 1116, + "TATTCA": 1117, + "TATTCT": 1118, + "TATTCC": 1119, + "TATTCG": 1120, + "TATTGA": 1121, + "TATTGT": 1122, + "TATTGC": 1123, + "TATTGG": 1124, + "TATCAA": 1125, + "TATCAT": 1126, + "TATCAC": 1127, + "TATCAG": 1128, + "TATCTA": 1129, + "TATCTT": 1130, + "TATCTC": 1131, + "TATCTG": 1132, + "TATCCA": 1133, + "TATCCT": 1134, + "TATCCC": 1135, + "TATCCG": 1136, + "TATCGA": 1137, + "TATCGT": 1138, + "TATCGC": 1139, + "TATCGG": 1140, + "TATGAA": 1141, + "TATGAT": 1142, + "TATGAC": 1143, + "TATGAG": 1144, + "TATGTA": 1145, + "TATGTT": 1146, + "TATGTC": 1147, + "TATGTG": 1148, + "TATGCA": 1149, + "TATGCT": 1150, + "TATGCC": 1151, + "TATGCG": 1152, + "TATGGA": 1153, + "TATGGT": 1154, + "TATGGC": 1155, + "TATGGG": 1156, + "TACAAA": 1157, + "TACAAT": 1158, + "TACAAC": 1159, + "TACAAG": 1160, + "TACATA": 1161, + "TACATT": 1162, + "TACATC": 1163, + "TACATG": 1164, + "TACACA": 1165, + "TACACT": 1166, + "TACACC": 1167, + "TACACG": 1168, + "TACAGA": 1169, + "TACAGT": 1170, + "TACAGC": 1171, + "TACAGG": 1172, + "TACTAA": 1173, + "TACTAT": 1174, + "TACTAC": 1175, + "TACTAG": 1176, + "TACTTA": 1177, + "TACTTT": 1178, + "TACTTC": 1179, + "TACTTG": 1180, + "TACTCA": 1181, + "TACTCT": 1182, + "TACTCC": 1183, + "TACTCG": 1184, + "TACTGA": 1185, + "TACTGT": 1186, + "TACTGC": 1187, + "TACTGG": 1188, + "TACCAA": 1189, + "TACCAT": 1190, + "TACCAC": 1191, + "TACCAG": 1192, + "TACCTA": 1193, + "TACCTT": 1194, + "TACCTC": 1195, + "TACCTG": 1196, + "TACCCA": 1197, + "TACCCT": 1198, + "TACCCC": 1199, + "TACCCG": 1200, + "TACCGA": 1201, + "TACCGT": 1202, + "TACCGC": 1203, + "TACCGG": 1204, + "TACGAA": 1205, + "TACGAT": 1206, + "TACGAC": 1207, + "TACGAG": 1208, + "TACGTA": 1209, + "TACGTT": 1210, + "TACGTC": 1211, + "TACGTG": 1212, + "TACGCA": 1213, + "TACGCT": 1214, + "TACGCC": 1215, + "TACGCG": 1216, + "TACGGA": 1217, + "TACGGT": 1218, + "TACGGC": 1219, + "TACGGG": 1220, + "TAGAAA": 1221, + "TAGAAT": 1222, + "TAGAAC": 1223, + "TAGAAG": 1224, + "TAGATA": 1225, + "TAGATT": 1226, + "TAGATC": 1227, + "TAGATG": 1228, + "TAGACA": 1229, + "TAGACT": 1230, + "TAGACC": 1231, + "TAGACG": 1232, + "TAGAGA": 1233, + "TAGAGT": 1234, + "TAGAGC": 1235, + "TAGAGG": 1236, + "TAGTAA": 1237, + "TAGTAT": 1238, + "TAGTAC": 1239, + "TAGTAG": 1240, + "TAGTTA": 1241, + "TAGTTT": 1242, + "TAGTTC": 1243, + "TAGTTG": 1244, + "TAGTCA": 1245, + "TAGTCT": 1246, + "TAGTCC": 1247, + "TAGTCG": 1248, + "TAGTGA": 1249, + "TAGTGT": 1250, + "TAGTGC": 1251, + "TAGTGG": 1252, + "TAGCAA": 1253, + "TAGCAT": 1254, + "TAGCAC": 1255, + "TAGCAG": 1256, + "TAGCTA": 1257, + "TAGCTT": 1258, + "TAGCTC": 1259, + "TAGCTG": 1260, + "TAGCCA": 1261, + "TAGCCT": 1262, + "TAGCCC": 1263, + "TAGCCG": 1264, + "TAGCGA": 1265, + "TAGCGT": 1266, + "TAGCGC": 1267, + "TAGCGG": 1268, + "TAGGAA": 1269, + "TAGGAT": 1270, + "TAGGAC": 1271, + "TAGGAG": 1272, + "TAGGTA": 1273, + "TAGGTT": 1274, + "TAGGTC": 1275, + "TAGGTG": 1276, + "TAGGCA": 1277, + "TAGGCT": 1278, + "TAGGCC": 1279, + "TAGGCG": 1280, + "TAGGGA": 1281, + "TAGGGT": 1282, + "TAGGGC": 1283, + "TAGGGG": 1284, + "TTAAAA": 1285, + "TTAAAT": 1286, + "TTAAAC": 1287, + "TTAAAG": 1288, + "TTAATA": 1289, + "TTAATT": 1290, + "TTAATC": 1291, + "TTAATG": 1292, + "TTAACA": 1293, + "TTAACT": 1294, + "TTAACC": 1295, + "TTAACG": 1296, + "TTAAGA": 1297, + "TTAAGT": 1298, + "TTAAGC": 1299, + "TTAAGG": 1300, + "TTATAA": 1301, + "TTATAT": 1302, + "TTATAC": 1303, + "TTATAG": 1304, + "TTATTA": 1305, + "TTATTT": 1306, + "TTATTC": 1307, + "TTATTG": 1308, + "TTATCA": 1309, + "TTATCT": 1310, + "TTATCC": 1311, + "TTATCG": 1312, + "TTATGA": 1313, + "TTATGT": 1314, + "TTATGC": 1315, + "TTATGG": 1316, + "TTACAA": 1317, + "TTACAT": 1318, + "TTACAC": 1319, + "TTACAG": 1320, + "TTACTA": 1321, + "TTACTT": 1322, + "TTACTC": 1323, + "TTACTG": 1324, + "TTACCA": 1325, + "TTACCT": 1326, + "TTACCC": 1327, + "TTACCG": 1328, + "TTACGA": 1329, + "TTACGT": 1330, + "TTACGC": 1331, + "TTACGG": 1332, + "TTAGAA": 1333, + "TTAGAT": 1334, + "TTAGAC": 1335, + "TTAGAG": 1336, + "TTAGTA": 1337, + "TTAGTT": 1338, + "TTAGTC": 1339, + "TTAGTG": 1340, + "TTAGCA": 1341, + "TTAGCT": 1342, + "TTAGCC": 1343, + "TTAGCG": 1344, + "TTAGGA": 1345, + "TTAGGT": 1346, + "TTAGGC": 1347, + "TTAGGG": 1348, + "TTTAAA": 1349, + "TTTAAT": 1350, + "TTTAAC": 1351, + "TTTAAG": 1352, + "TTTATA": 1353, + "TTTATT": 1354, + "TTTATC": 1355, + "TTTATG": 1356, + "TTTACA": 1357, + "TTTACT": 1358, + "TTTACC": 1359, + "TTTACG": 1360, + "TTTAGA": 1361, + "TTTAGT": 1362, + "TTTAGC": 1363, + "TTTAGG": 1364, + "TTTTAA": 1365, + "TTTTAT": 1366, + "TTTTAC": 1367, + "TTTTAG": 1368, + "TTTTTA": 1369, + "TTTTTT": 1370, + "TTTTTC": 1371, + "TTTTTG": 1372, + "TTTTCA": 1373, + "TTTTCT": 1374, + "TTTTCC": 1375, + "TTTTCG": 1376, + "TTTTGA": 1377, + "TTTTGT": 1378, + "TTTTGC": 1379, + "TTTTGG": 1380, + "TTTCAA": 1381, + "TTTCAT": 1382, + "TTTCAC": 1383, + "TTTCAG": 1384, + "TTTCTA": 1385, + "TTTCTT": 1386, + "TTTCTC": 1387, + "TTTCTG": 1388, + "TTTCCA": 1389, + "TTTCCT": 1390, + "TTTCCC": 1391, + "TTTCCG": 1392, + "TTTCGA": 1393, + "TTTCGT": 1394, + "TTTCGC": 1395, + "TTTCGG": 1396, + "TTTGAA": 1397, + "TTTGAT": 1398, + "TTTGAC": 1399, + "TTTGAG": 1400, + "TTTGTA": 1401, + "TTTGTT": 1402, + "TTTGTC": 1403, + "TTTGTG": 1404, + "TTTGCA": 1405, + "TTTGCT": 1406, + "TTTGCC": 1407, + "TTTGCG": 1408, + "TTTGGA": 1409, + "TTTGGT": 1410, + "TTTGGC": 1411, + "TTTGGG": 1412, + "TTCAAA": 1413, + "TTCAAT": 1414, + "TTCAAC": 1415, + "TTCAAG": 1416, + "TTCATA": 1417, + "TTCATT": 1418, + "TTCATC": 1419, + "TTCATG": 1420, + "TTCACA": 1421, + "TTCACT": 1422, + "TTCACC": 1423, + "TTCACG": 1424, + "TTCAGA": 1425, + "TTCAGT": 1426, + "TTCAGC": 1427, + "TTCAGG": 1428, + "TTCTAA": 1429, + "TTCTAT": 1430, + "TTCTAC": 1431, + "TTCTAG": 1432, + "TTCTTA": 1433, + "TTCTTT": 1434, + "TTCTTC": 1435, + "TTCTTG": 1436, + "TTCTCA": 1437, + "TTCTCT": 1438, + "TTCTCC": 1439, + "TTCTCG": 1440, + "TTCTGA": 1441, + "TTCTGT": 1442, + "TTCTGC": 1443, + "TTCTGG": 1444, + "TTCCAA": 1445, + "TTCCAT": 1446, + "TTCCAC": 1447, + "TTCCAG": 1448, + "TTCCTA": 1449, + "TTCCTT": 1450, + "TTCCTC": 1451, + "TTCCTG": 1452, + "TTCCCA": 1453, + "TTCCCT": 1454, + "TTCCCC": 1455, + "TTCCCG": 1456, + "TTCCGA": 1457, + "TTCCGT": 1458, + "TTCCGC": 1459, + "TTCCGG": 1460, + "TTCGAA": 1461, + "TTCGAT": 1462, + "TTCGAC": 1463, + "TTCGAG": 1464, + "TTCGTA": 1465, + "TTCGTT": 1466, + "TTCGTC": 1467, + "TTCGTG": 1468, + "TTCGCA": 1469, + "TTCGCT": 1470, + "TTCGCC": 1471, + "TTCGCG": 1472, + "TTCGGA": 1473, + "TTCGGT": 1474, + "TTCGGC": 1475, + "TTCGGG": 1476, + "TTGAAA": 1477, + "TTGAAT": 1478, + "TTGAAC": 1479, + "TTGAAG": 1480, + "TTGATA": 1481, + "TTGATT": 1482, + "TTGATC": 1483, + "TTGATG": 1484, + "TTGACA": 1485, + "TTGACT": 1486, + "TTGACC": 1487, + "TTGACG": 1488, + "TTGAGA": 1489, + "TTGAGT": 1490, + "TTGAGC": 1491, + "TTGAGG": 1492, + "TTGTAA": 1493, + "TTGTAT": 1494, + "TTGTAC": 1495, + "TTGTAG": 1496, + "TTGTTA": 1497, + "TTGTTT": 1498, + "TTGTTC": 1499, + "TTGTTG": 1500, + "TTGTCA": 1501, + "TTGTCT": 1502, + "TTGTCC": 1503, + "TTGTCG": 1504, + "TTGTGA": 1505, + "TTGTGT": 1506, + "TTGTGC": 1507, + "TTGTGG": 1508, + "TTGCAA": 1509, + "TTGCAT": 1510, + "TTGCAC": 1511, + "TTGCAG": 1512, + "TTGCTA": 1513, + "TTGCTT": 1514, + "TTGCTC": 1515, + "TTGCTG": 1516, + "TTGCCA": 1517, + "TTGCCT": 1518, + "TTGCCC": 1519, + "TTGCCG": 1520, + "TTGCGA": 1521, + "TTGCGT": 1522, + "TTGCGC": 1523, + "TTGCGG": 1524, + "TTGGAA": 1525, + "TTGGAT": 1526, + "TTGGAC": 1527, + "TTGGAG": 1528, + "TTGGTA": 1529, + "TTGGTT": 1530, + "TTGGTC": 1531, + "TTGGTG": 1532, + "TTGGCA": 1533, + "TTGGCT": 1534, + "TTGGCC": 1535, + "TTGGCG": 1536, + "TTGGGA": 1537, + "TTGGGT": 1538, + "TTGGGC": 1539, + "TTGGGG": 1540, + "TCAAAA": 1541, + "TCAAAT": 1542, + "TCAAAC": 1543, + "TCAAAG": 1544, + "TCAATA": 1545, + "TCAATT": 1546, + "TCAATC": 1547, + "TCAATG": 1548, + "TCAACA": 1549, + "TCAACT": 1550, + "TCAACC": 1551, + "TCAACG": 1552, + "TCAAGA": 1553, + "TCAAGT": 1554, + "TCAAGC": 1555, + "TCAAGG": 1556, + "TCATAA": 1557, + "TCATAT": 1558, + "TCATAC": 1559, + "TCATAG": 1560, + "TCATTA": 1561, + "TCATTT": 1562, + "TCATTC": 1563, + "TCATTG": 1564, + "TCATCA": 1565, + "TCATCT": 1566, + "TCATCC": 1567, + "TCATCG": 1568, + "TCATGA": 1569, + "TCATGT": 1570, + "TCATGC": 1571, + "TCATGG": 1572, + "TCACAA": 1573, + "TCACAT": 1574, + "TCACAC": 1575, + "TCACAG": 1576, + "TCACTA": 1577, + "TCACTT": 1578, + "TCACTC": 1579, + "TCACTG": 1580, + "TCACCA": 1581, + "TCACCT": 1582, + "TCACCC": 1583, + "TCACCG": 1584, + "TCACGA": 1585, + "TCACGT": 1586, + "TCACGC": 1587, + "TCACGG": 1588, + "TCAGAA": 1589, + "TCAGAT": 1590, + "TCAGAC": 1591, + "TCAGAG": 1592, + "TCAGTA": 1593, + "TCAGTT": 1594, + "TCAGTC": 1595, + "TCAGTG": 1596, + "TCAGCA": 1597, + "TCAGCT": 1598, + "TCAGCC": 1599, + "TCAGCG": 1600, + "TCAGGA": 1601, + "TCAGGT": 1602, + "TCAGGC": 1603, + "TCAGGG": 1604, + "TCTAAA": 1605, + "TCTAAT": 1606, + "TCTAAC": 1607, + "TCTAAG": 1608, + "TCTATA": 1609, + "TCTATT": 1610, + "TCTATC": 1611, + "TCTATG": 1612, + "TCTACA": 1613, + "TCTACT": 1614, + "TCTACC": 1615, + "TCTACG": 1616, + "TCTAGA": 1617, + "TCTAGT": 1618, + "TCTAGC": 1619, + "TCTAGG": 1620, + "TCTTAA": 1621, + "TCTTAT": 1622, + "TCTTAC": 1623, + "TCTTAG": 1624, + "TCTTTA": 1625, + "TCTTTT": 1626, + "TCTTTC": 1627, + "TCTTTG": 1628, + "TCTTCA": 1629, + "TCTTCT": 1630, + "TCTTCC": 1631, + "TCTTCG": 1632, + "TCTTGA": 1633, + "TCTTGT": 1634, + "TCTTGC": 1635, + "TCTTGG": 1636, + "TCTCAA": 1637, + "TCTCAT": 1638, + "TCTCAC": 1639, + "TCTCAG": 1640, + "TCTCTA": 1641, + "TCTCTT": 1642, + "TCTCTC": 1643, + "TCTCTG": 1644, + "TCTCCA": 1645, + "TCTCCT": 1646, + "TCTCCC": 1647, + "TCTCCG": 1648, + "TCTCGA": 1649, + "TCTCGT": 1650, + "TCTCGC": 1651, + "TCTCGG": 1652, + "TCTGAA": 1653, + "TCTGAT": 1654, + "TCTGAC": 1655, + "TCTGAG": 1656, + "TCTGTA": 1657, + "TCTGTT": 1658, + "TCTGTC": 1659, + "TCTGTG": 1660, + "TCTGCA": 1661, + "TCTGCT": 1662, + "TCTGCC": 1663, + "TCTGCG": 1664, + "TCTGGA": 1665, + "TCTGGT": 1666, + "TCTGGC": 1667, + "TCTGGG": 1668, + "TCCAAA": 1669, + "TCCAAT": 1670, + "TCCAAC": 1671, + "TCCAAG": 1672, + "TCCATA": 1673, + "TCCATT": 1674, + "TCCATC": 1675, + "TCCATG": 1676, + "TCCACA": 1677, + "TCCACT": 1678, + "TCCACC": 1679, + "TCCACG": 1680, + "TCCAGA": 1681, + "TCCAGT": 1682, + "TCCAGC": 1683, + "TCCAGG": 1684, + "TCCTAA": 1685, + "TCCTAT": 1686, + "TCCTAC": 1687, + "TCCTAG": 1688, + "TCCTTA": 1689, + "TCCTTT": 1690, + "TCCTTC": 1691, + "TCCTTG": 1692, + "TCCTCA": 1693, + "TCCTCT": 1694, + "TCCTCC": 1695, + "TCCTCG": 1696, + "TCCTGA": 1697, + "TCCTGT": 1698, + "TCCTGC": 1699, + "TCCTGG": 1700, + "TCCCAA": 1701, + "TCCCAT": 1702, + "TCCCAC": 1703, + "TCCCAG": 1704, + "TCCCTA": 1705, + "TCCCTT": 1706, + "TCCCTC": 1707, + "TCCCTG": 1708, + "TCCCCA": 1709, + "TCCCCT": 1710, + "TCCCCC": 1711, + "TCCCCG": 1712, + "TCCCGA": 1713, + "TCCCGT": 1714, + "TCCCGC": 1715, + "TCCCGG": 1716, + "TCCGAA": 1717, + "TCCGAT": 1718, + "TCCGAC": 1719, + "TCCGAG": 1720, + "TCCGTA": 1721, + "TCCGTT": 1722, + "TCCGTC": 1723, + "TCCGTG": 1724, + "TCCGCA": 1725, + "TCCGCT": 1726, + "TCCGCC": 1727, + "TCCGCG": 1728, + "TCCGGA": 1729, + "TCCGGT": 1730, + "TCCGGC": 1731, + "TCCGGG": 1732, + "TCGAAA": 1733, + "TCGAAT": 1734, + "TCGAAC": 1735, + "TCGAAG": 1736, + "TCGATA": 1737, + "TCGATT": 1738, + "TCGATC": 1739, + "TCGATG": 1740, + "TCGACA": 1741, + "TCGACT": 1742, + "TCGACC": 1743, + "TCGACG": 1744, + "TCGAGA": 1745, + "TCGAGT": 1746, + "TCGAGC": 1747, + "TCGAGG": 1748, + "TCGTAA": 1749, + "TCGTAT": 1750, + "TCGTAC": 1751, + "TCGTAG": 1752, + "TCGTTA": 1753, + "TCGTTT": 1754, + "TCGTTC": 1755, + "TCGTTG": 1756, + "TCGTCA": 1757, + "TCGTCT": 1758, + "TCGTCC": 1759, + "TCGTCG": 1760, + "TCGTGA": 1761, + "TCGTGT": 1762, + "TCGTGC": 1763, + "TCGTGG": 1764, + "TCGCAA": 1765, + "TCGCAT": 1766, + "TCGCAC": 1767, + "TCGCAG": 1768, + "TCGCTA": 1769, + "TCGCTT": 1770, + "TCGCTC": 1771, + "TCGCTG": 1772, + "TCGCCA": 1773, + "TCGCCT": 1774, + "TCGCCC": 1775, + "TCGCCG": 1776, + "TCGCGA": 1777, + "TCGCGT": 1778, + "TCGCGC": 1779, + "TCGCGG": 1780, + "TCGGAA": 1781, + "TCGGAT": 1782, + "TCGGAC": 1783, + "TCGGAG": 1784, + "TCGGTA": 1785, + "TCGGTT": 1786, + "TCGGTC": 1787, + "TCGGTG": 1788, + "TCGGCA": 1789, + "TCGGCT": 1790, + "TCGGCC": 1791, + "TCGGCG": 1792, + "TCGGGA": 1793, + "TCGGGT": 1794, + "TCGGGC": 1795, + "TCGGGG": 1796, + "TGAAAA": 1797, + "TGAAAT": 1798, + "TGAAAC": 1799, + "TGAAAG": 1800, + "TGAATA": 1801, + "TGAATT": 1802, + "TGAATC": 1803, + "TGAATG": 1804, + "TGAACA": 1805, + "TGAACT": 1806, + "TGAACC": 1807, + "TGAACG": 1808, + "TGAAGA": 1809, + "TGAAGT": 1810, + "TGAAGC": 1811, + "TGAAGG": 1812, + "TGATAA": 1813, + "TGATAT": 1814, + "TGATAC": 1815, + "TGATAG": 1816, + "TGATTA": 1817, + "TGATTT": 1818, + "TGATTC": 1819, + "TGATTG": 1820, + "TGATCA": 1821, + "TGATCT": 1822, + "TGATCC": 1823, + "TGATCG": 1824, + "TGATGA": 1825, + "TGATGT": 1826, + "TGATGC": 1827, + "TGATGG": 1828, + "TGACAA": 1829, + "TGACAT": 1830, + "TGACAC": 1831, + "TGACAG": 1832, + "TGACTA": 1833, + "TGACTT": 1834, + "TGACTC": 1835, + "TGACTG": 1836, + "TGACCA": 1837, + "TGACCT": 1838, + "TGACCC": 1839, + "TGACCG": 1840, + "TGACGA": 1841, + "TGACGT": 1842, + "TGACGC": 1843, + "TGACGG": 1844, + "TGAGAA": 1845, + "TGAGAT": 1846, + "TGAGAC": 1847, + "TGAGAG": 1848, + "TGAGTA": 1849, + "TGAGTT": 1850, + "TGAGTC": 1851, + "TGAGTG": 1852, + "TGAGCA": 1853, + "TGAGCT": 1854, + "TGAGCC": 1855, + "TGAGCG": 1856, + "TGAGGA": 1857, + "TGAGGT": 1858, + "TGAGGC": 1859, + "TGAGGG": 1860, + "TGTAAA": 1861, + "TGTAAT": 1862, + "TGTAAC": 1863, + "TGTAAG": 1864, + "TGTATA": 1865, + "TGTATT": 1866, + "TGTATC": 1867, + "TGTATG": 1868, + "TGTACA": 1869, + "TGTACT": 1870, + "TGTACC": 1871, + "TGTACG": 1872, + "TGTAGA": 1873, + "TGTAGT": 1874, + "TGTAGC": 1875, + "TGTAGG": 1876, + "TGTTAA": 1877, + "TGTTAT": 1878, + "TGTTAC": 1879, + "TGTTAG": 1880, + "TGTTTA": 1881, + "TGTTTT": 1882, + "TGTTTC": 1883, + "TGTTTG": 1884, + "TGTTCA": 1885, + "TGTTCT": 1886, + "TGTTCC": 1887, + "TGTTCG": 1888, + "TGTTGA": 1889, + "TGTTGT": 1890, + "TGTTGC": 1891, + "TGTTGG": 1892, + "TGTCAA": 1893, + "TGTCAT": 1894, + "TGTCAC": 1895, + "TGTCAG": 1896, + "TGTCTA": 1897, + "TGTCTT": 1898, + "TGTCTC": 1899, + "TGTCTG": 1900, + "TGTCCA": 1901, + "TGTCCT": 1902, + "TGTCCC": 1903, + "TGTCCG": 1904, + "TGTCGA": 1905, + "TGTCGT": 1906, + "TGTCGC": 1907, + "TGTCGG": 1908, + "TGTGAA": 1909, + "TGTGAT": 1910, + "TGTGAC": 1911, + "TGTGAG": 1912, + "TGTGTA": 1913, + "TGTGTT": 1914, + "TGTGTC": 1915, + "TGTGTG": 1916, + "TGTGCA": 1917, + "TGTGCT": 1918, + "TGTGCC": 1919, + "TGTGCG": 1920, + "TGTGGA": 1921, + "TGTGGT": 1922, + "TGTGGC": 1923, + "TGTGGG": 1924, + "TGCAAA": 1925, + "TGCAAT": 1926, + "TGCAAC": 1927, + "TGCAAG": 1928, + "TGCATA": 1929, + "TGCATT": 1930, + "TGCATC": 1931, + "TGCATG": 1932, + "TGCACA": 1933, + "TGCACT": 1934, + "TGCACC": 1935, + "TGCACG": 1936, + "TGCAGA": 1937, + "TGCAGT": 1938, + "TGCAGC": 1939, + "TGCAGG": 1940, + "TGCTAA": 1941, + "TGCTAT": 1942, + "TGCTAC": 1943, + "TGCTAG": 1944, + "TGCTTA": 1945, + "TGCTTT": 1946, + "TGCTTC": 1947, + "TGCTTG": 1948, + "TGCTCA": 1949, + "TGCTCT": 1950, + "TGCTCC": 1951, + "TGCTCG": 1952, + "TGCTGA": 1953, + "TGCTGT": 1954, + "TGCTGC": 1955, + "TGCTGG": 1956, + "TGCCAA": 1957, + "TGCCAT": 1958, + "TGCCAC": 1959, + "TGCCAG": 1960, + "TGCCTA": 1961, + "TGCCTT": 1962, + "TGCCTC": 1963, + "TGCCTG": 1964, + "TGCCCA": 1965, + "TGCCCT": 1966, + "TGCCCC": 1967, + "TGCCCG": 1968, + "TGCCGA": 1969, + "TGCCGT": 1970, + "TGCCGC": 1971, + "TGCCGG": 1972, + "TGCGAA": 1973, + "TGCGAT": 1974, + "TGCGAC": 1975, + "TGCGAG": 1976, + "TGCGTA": 1977, + "TGCGTT": 1978, + "TGCGTC": 1979, + "TGCGTG": 1980, + "TGCGCA": 1981, + "TGCGCT": 1982, + "TGCGCC": 1983, + "TGCGCG": 1984, + "TGCGGA": 1985, + "TGCGGT": 1986, + "TGCGGC": 1987, + "TGCGGG": 1988, + "TGGAAA": 1989, + "TGGAAT": 1990, + "TGGAAC": 1991, + "TGGAAG": 1992, + "TGGATA": 1993, + "TGGATT": 1994, + "TGGATC": 1995, + "TGGATG": 1996, + "TGGACA": 1997, + "TGGACT": 1998, + "TGGACC": 1999, + "TGGACG": 2000, + "TGGAGA": 2001, + "TGGAGT": 2002, + "TGGAGC": 2003, + "TGGAGG": 2004, + "TGGTAA": 2005, + "TGGTAT": 2006, + "TGGTAC": 2007, + "TGGTAG": 2008, + "TGGTTA": 2009, + "TGGTTT": 2010, + "TGGTTC": 2011, + "TGGTTG": 2012, + "TGGTCA": 2013, + "TGGTCT": 2014, + "TGGTCC": 2015, + "TGGTCG": 2016, + "TGGTGA": 2017, + "TGGTGT": 2018, + "TGGTGC": 2019, + "TGGTGG": 2020, + "TGGCAA": 2021, + "TGGCAT": 2022, + "TGGCAC": 2023, + "TGGCAG": 2024, + "TGGCTA": 2025, + "TGGCTT": 2026, + "TGGCTC": 2027, + "TGGCTG": 2028, + "TGGCCA": 2029, + "TGGCCT": 2030, + "TGGCCC": 2031, + "TGGCCG": 2032, + "TGGCGA": 2033, + "TGGCGT": 2034, + "TGGCGC": 2035, + "TGGCGG": 2036, + "TGGGAA": 2037, + "TGGGAT": 2038, + "TGGGAC": 2039, + "TGGGAG": 2040, + "TGGGTA": 2041, + "TGGGTT": 2042, + "TGGGTC": 2043, + "TGGGTG": 2044, + "TGGGCA": 2045, + "TGGGCT": 2046, + "TGGGCC": 2047, + "TGGGCG": 2048, + "TGGGGA": 2049, + "TGGGGT": 2050, + "TGGGGC": 2051, + "TGGGGG": 2052, + "CAAAAA": 2053, + "CAAAAT": 2054, + "CAAAAC": 2055, + "CAAAAG": 2056, + "CAAATA": 2057, + "CAAATT": 2058, + "CAAATC": 2059, + "CAAATG": 2060, + "CAAACA": 2061, + "CAAACT": 2062, + "CAAACC": 2063, + "CAAACG": 2064, + "CAAAGA": 2065, + "CAAAGT": 2066, + "CAAAGC": 2067, + "CAAAGG": 2068, + "CAATAA": 2069, + "CAATAT": 2070, + "CAATAC": 2071, + "CAATAG": 2072, + "CAATTA": 2073, + "CAATTT": 2074, + "CAATTC": 2075, + "CAATTG": 2076, + "CAATCA": 2077, + "CAATCT": 2078, + "CAATCC": 2079, + "CAATCG": 2080, + "CAATGA": 2081, + "CAATGT": 2082, + "CAATGC": 2083, + "CAATGG": 2084, + "CAACAA": 2085, + "CAACAT": 2086, + "CAACAC": 2087, + "CAACAG": 2088, + "CAACTA": 2089, + "CAACTT": 2090, + "CAACTC": 2091, + "CAACTG": 2092, + "CAACCA": 2093, + "CAACCT": 2094, + "CAACCC": 2095, + "CAACCG": 2096, + "CAACGA": 2097, + "CAACGT": 2098, + "CAACGC": 2099, + "CAACGG": 2100, + "CAAGAA": 2101, + "CAAGAT": 2102, + "CAAGAC": 2103, + "CAAGAG": 2104, + "CAAGTA": 2105, + "CAAGTT": 2106, + "CAAGTC": 2107, + "CAAGTG": 2108, + "CAAGCA": 2109, + "CAAGCT": 2110, + "CAAGCC": 2111, + "CAAGCG": 2112, + "CAAGGA": 2113, + "CAAGGT": 2114, + "CAAGGC": 2115, + "CAAGGG": 2116, + "CATAAA": 2117, + "CATAAT": 2118, + "CATAAC": 2119, + "CATAAG": 2120, + "CATATA": 2121, + "CATATT": 2122, + "CATATC": 2123, + "CATATG": 2124, + "CATACA": 2125, + "CATACT": 2126, + "CATACC": 2127, + "CATACG": 2128, + "CATAGA": 2129, + "CATAGT": 2130, + "CATAGC": 2131, + "CATAGG": 2132, + "CATTAA": 2133, + "CATTAT": 2134, + "CATTAC": 2135, + "CATTAG": 2136, + "CATTTA": 2137, + "CATTTT": 2138, + "CATTTC": 2139, + "CATTTG": 2140, + "CATTCA": 2141, + "CATTCT": 2142, + "CATTCC": 2143, + "CATTCG": 2144, + "CATTGA": 2145, + "CATTGT": 2146, + "CATTGC": 2147, + "CATTGG": 2148, + "CATCAA": 2149, + "CATCAT": 2150, + "CATCAC": 2151, + "CATCAG": 2152, + "CATCTA": 2153, + "CATCTT": 2154, + "CATCTC": 2155, + "CATCTG": 2156, + "CATCCA": 2157, + "CATCCT": 2158, + "CATCCC": 2159, + "CATCCG": 2160, + "CATCGA": 2161, + "CATCGT": 2162, + "CATCGC": 2163, + "CATCGG": 2164, + "CATGAA": 2165, + "CATGAT": 2166, + "CATGAC": 2167, + "CATGAG": 2168, + "CATGTA": 2169, + "CATGTT": 2170, + "CATGTC": 2171, + "CATGTG": 2172, + "CATGCA": 2173, + "CATGCT": 2174, + "CATGCC": 2175, + "CATGCG": 2176, + "CATGGA": 2177, + "CATGGT": 2178, + "CATGGC": 2179, + "CATGGG": 2180, + "CACAAA": 2181, + "CACAAT": 2182, + "CACAAC": 2183, + "CACAAG": 2184, + "CACATA": 2185, + "CACATT": 2186, + "CACATC": 2187, + "CACATG": 2188, + "CACACA": 2189, + "CACACT": 2190, + "CACACC": 2191, + "CACACG": 2192, + "CACAGA": 2193, + "CACAGT": 2194, + "CACAGC": 2195, + "CACAGG": 2196, + "CACTAA": 2197, + "CACTAT": 2198, + "CACTAC": 2199, + "CACTAG": 2200, + "CACTTA": 2201, + "CACTTT": 2202, + "CACTTC": 2203, + "CACTTG": 2204, + "CACTCA": 2205, + "CACTCT": 2206, + "CACTCC": 2207, + "CACTCG": 2208, + "CACTGA": 2209, + "CACTGT": 2210, + "CACTGC": 2211, + "CACTGG": 2212, + "CACCAA": 2213, + "CACCAT": 2214, + "CACCAC": 2215, + "CACCAG": 2216, + "CACCTA": 2217, + "CACCTT": 2218, + "CACCTC": 2219, + "CACCTG": 2220, + "CACCCA": 2221, + "CACCCT": 2222, + "CACCCC": 2223, + "CACCCG": 2224, + "CACCGA": 2225, + "CACCGT": 2226, + "CACCGC": 2227, + "CACCGG": 2228, + "CACGAA": 2229, + "CACGAT": 2230, + "CACGAC": 2231, + "CACGAG": 2232, + "CACGTA": 2233, + "CACGTT": 2234, + "CACGTC": 2235, + "CACGTG": 2236, + "CACGCA": 2237, + "CACGCT": 2238, + "CACGCC": 2239, + "CACGCG": 2240, + "CACGGA": 2241, + "CACGGT": 2242, + "CACGGC": 2243, + "CACGGG": 2244, + "CAGAAA": 2245, + "CAGAAT": 2246, + "CAGAAC": 2247, + "CAGAAG": 2248, + "CAGATA": 2249, + "CAGATT": 2250, + "CAGATC": 2251, + "CAGATG": 2252, + "CAGACA": 2253, + "CAGACT": 2254, + "CAGACC": 2255, + "CAGACG": 2256, + "CAGAGA": 2257, + "CAGAGT": 2258, + "CAGAGC": 2259, + "CAGAGG": 2260, + "CAGTAA": 2261, + "CAGTAT": 2262, + "CAGTAC": 2263, + "CAGTAG": 2264, + "CAGTTA": 2265, + "CAGTTT": 2266, + "CAGTTC": 2267, + "CAGTTG": 2268, + "CAGTCA": 2269, + "CAGTCT": 2270, + "CAGTCC": 2271, + "CAGTCG": 2272, + "CAGTGA": 2273, + "CAGTGT": 2274, + "CAGTGC": 2275, + "CAGTGG": 2276, + "CAGCAA": 2277, + "CAGCAT": 2278, + "CAGCAC": 2279, + "CAGCAG": 2280, + "CAGCTA": 2281, + "CAGCTT": 2282, + "CAGCTC": 2283, + "CAGCTG": 2284, + "CAGCCA": 2285, + "CAGCCT": 2286, + "CAGCCC": 2287, + "CAGCCG": 2288, + "CAGCGA": 2289, + "CAGCGT": 2290, + "CAGCGC": 2291, + "CAGCGG": 2292, + "CAGGAA": 2293, + "CAGGAT": 2294, + "CAGGAC": 2295, + "CAGGAG": 2296, + "CAGGTA": 2297, + "CAGGTT": 2298, + "CAGGTC": 2299, + "CAGGTG": 2300, + "CAGGCA": 2301, + "CAGGCT": 2302, + "CAGGCC": 2303, + "CAGGCG": 2304, + "CAGGGA": 2305, + "CAGGGT": 2306, + "CAGGGC": 2307, + "CAGGGG": 2308, + "CTAAAA": 2309, + "CTAAAT": 2310, + "CTAAAC": 2311, + "CTAAAG": 2312, + "CTAATA": 2313, + "CTAATT": 2314, + "CTAATC": 2315, + "CTAATG": 2316, + "CTAACA": 2317, + "CTAACT": 2318, + "CTAACC": 2319, + "CTAACG": 2320, + "CTAAGA": 2321, + "CTAAGT": 2322, + "CTAAGC": 2323, + "CTAAGG": 2324, + "CTATAA": 2325, + "CTATAT": 2326, + "CTATAC": 2327, + "CTATAG": 2328, + "CTATTA": 2329, + "CTATTT": 2330, + "CTATTC": 2331, + "CTATTG": 2332, + "CTATCA": 2333, + "CTATCT": 2334, + "CTATCC": 2335, + "CTATCG": 2336, + "CTATGA": 2337, + "CTATGT": 2338, + "CTATGC": 2339, + "CTATGG": 2340, + "CTACAA": 2341, + "CTACAT": 2342, + "CTACAC": 2343, + "CTACAG": 2344, + "CTACTA": 2345, + "CTACTT": 2346, + "CTACTC": 2347, + "CTACTG": 2348, + "CTACCA": 2349, + "CTACCT": 2350, + "CTACCC": 2351, + "CTACCG": 2352, + "CTACGA": 2353, + "CTACGT": 2354, + "CTACGC": 2355, + "CTACGG": 2356, + "CTAGAA": 2357, + "CTAGAT": 2358, + "CTAGAC": 2359, + "CTAGAG": 2360, + "CTAGTA": 2361, + "CTAGTT": 2362, + "CTAGTC": 2363, + "CTAGTG": 2364, + "CTAGCA": 2365, + "CTAGCT": 2366, + "CTAGCC": 2367, + "CTAGCG": 2368, + "CTAGGA": 2369, + "CTAGGT": 2370, + "CTAGGC": 2371, + "CTAGGG": 2372, + "CTTAAA": 2373, + "CTTAAT": 2374, + "CTTAAC": 2375, + "CTTAAG": 2376, + "CTTATA": 2377, + "CTTATT": 2378, + "CTTATC": 2379, + "CTTATG": 2380, + "CTTACA": 2381, + "CTTACT": 2382, + "CTTACC": 2383, + "CTTACG": 2384, + "CTTAGA": 2385, + "CTTAGT": 2386, + "CTTAGC": 2387, + "CTTAGG": 2388, + "CTTTAA": 2389, + "CTTTAT": 2390, + "CTTTAC": 2391, + "CTTTAG": 2392, + "CTTTTA": 2393, + "CTTTTT": 2394, + "CTTTTC": 2395, + "CTTTTG": 2396, + "CTTTCA": 2397, + "CTTTCT": 2398, + "CTTTCC": 2399, + "CTTTCG": 2400, + "CTTTGA": 2401, + "CTTTGT": 2402, + "CTTTGC": 2403, + "CTTTGG": 2404, + "CTTCAA": 2405, + "CTTCAT": 2406, + "CTTCAC": 2407, + "CTTCAG": 2408, + "CTTCTA": 2409, + "CTTCTT": 2410, + "CTTCTC": 2411, + "CTTCTG": 2412, + "CTTCCA": 2413, + "CTTCCT": 2414, + "CTTCCC": 2415, + "CTTCCG": 2416, + "CTTCGA": 2417, + "CTTCGT": 2418, + "CTTCGC": 2419, + "CTTCGG": 2420, + "CTTGAA": 2421, + "CTTGAT": 2422, + "CTTGAC": 2423, + "CTTGAG": 2424, + "CTTGTA": 2425, + "CTTGTT": 2426, + "CTTGTC": 2427, + "CTTGTG": 2428, + "CTTGCA": 2429, + "CTTGCT": 2430, + "CTTGCC": 2431, + "CTTGCG": 2432, + "CTTGGA": 2433, + "CTTGGT": 2434, + "CTTGGC": 2435, + "CTTGGG": 2436, + "CTCAAA": 2437, + "CTCAAT": 2438, + "CTCAAC": 2439, + "CTCAAG": 2440, + "CTCATA": 2441, + "CTCATT": 2442, + "CTCATC": 2443, + "CTCATG": 2444, + "CTCACA": 2445, + "CTCACT": 2446, + "CTCACC": 2447, + "CTCACG": 2448, + "CTCAGA": 2449, + "CTCAGT": 2450, + "CTCAGC": 2451, + "CTCAGG": 2452, + "CTCTAA": 2453, + "CTCTAT": 2454, + "CTCTAC": 2455, + "CTCTAG": 2456, + "CTCTTA": 2457, + "CTCTTT": 2458, + "CTCTTC": 2459, + "CTCTTG": 2460, + "CTCTCA": 2461, + "CTCTCT": 2462, + "CTCTCC": 2463, + "CTCTCG": 2464, + "CTCTGA": 2465, + "CTCTGT": 2466, + "CTCTGC": 2467, + "CTCTGG": 2468, + "CTCCAA": 2469, + "CTCCAT": 2470, + "CTCCAC": 2471, + "CTCCAG": 2472, + "CTCCTA": 2473, + "CTCCTT": 2474, + "CTCCTC": 2475, + "CTCCTG": 2476, + "CTCCCA": 2477, + "CTCCCT": 2478, + "CTCCCC": 2479, + "CTCCCG": 2480, + "CTCCGA": 2481, + "CTCCGT": 2482, + "CTCCGC": 2483, + "CTCCGG": 2484, + "CTCGAA": 2485, + "CTCGAT": 2486, + "CTCGAC": 2487, + "CTCGAG": 2488, + "CTCGTA": 2489, + "CTCGTT": 2490, + "CTCGTC": 2491, + "CTCGTG": 2492, + "CTCGCA": 2493, + "CTCGCT": 2494, + "CTCGCC": 2495, + "CTCGCG": 2496, + "CTCGGA": 2497, + "CTCGGT": 2498, + "CTCGGC": 2499, + "CTCGGG": 2500, + "CTGAAA": 2501, + "CTGAAT": 2502, + "CTGAAC": 2503, + "CTGAAG": 2504, + "CTGATA": 2505, + "CTGATT": 2506, + "CTGATC": 2507, + "CTGATG": 2508, + "CTGACA": 2509, + "CTGACT": 2510, + "CTGACC": 2511, + "CTGACG": 2512, + "CTGAGA": 2513, + "CTGAGT": 2514, + "CTGAGC": 2515, + "CTGAGG": 2516, + "CTGTAA": 2517, + "CTGTAT": 2518, + "CTGTAC": 2519, + "CTGTAG": 2520, + "CTGTTA": 2521, + "CTGTTT": 2522, + "CTGTTC": 2523, + "CTGTTG": 2524, + "CTGTCA": 2525, + "CTGTCT": 2526, + "CTGTCC": 2527, + "CTGTCG": 2528, + "CTGTGA": 2529, + "CTGTGT": 2530, + "CTGTGC": 2531, + "CTGTGG": 2532, + "CTGCAA": 2533, + "CTGCAT": 2534, + "CTGCAC": 2535, + "CTGCAG": 2536, + "CTGCTA": 2537, + "CTGCTT": 2538, + "CTGCTC": 2539, + "CTGCTG": 2540, + "CTGCCA": 2541, + "CTGCCT": 2542, + "CTGCCC": 2543, + "CTGCCG": 2544, + "CTGCGA": 2545, + "CTGCGT": 2546, + "CTGCGC": 2547, + "CTGCGG": 2548, + "CTGGAA": 2549, + "CTGGAT": 2550, + "CTGGAC": 2551, + "CTGGAG": 2552, + "CTGGTA": 2553, + "CTGGTT": 2554, + "CTGGTC": 2555, + "CTGGTG": 2556, + "CTGGCA": 2557, + "CTGGCT": 2558, + "CTGGCC": 2559, + "CTGGCG": 2560, + "CTGGGA": 2561, + "CTGGGT": 2562, + "CTGGGC": 2563, + "CTGGGG": 2564, + "CCAAAA": 2565, + "CCAAAT": 2566, + "CCAAAC": 2567, + "CCAAAG": 2568, + "CCAATA": 2569, + "CCAATT": 2570, + "CCAATC": 2571, + "CCAATG": 2572, + "CCAACA": 2573, + "CCAACT": 2574, + "CCAACC": 2575, + "CCAACG": 2576, + "CCAAGA": 2577, + "CCAAGT": 2578, + "CCAAGC": 2579, + "CCAAGG": 2580, + "CCATAA": 2581, + "CCATAT": 2582, + "CCATAC": 2583, + "CCATAG": 2584, + "CCATTA": 2585, + "CCATTT": 2586, + "CCATTC": 2587, + "CCATTG": 2588, + "CCATCA": 2589, + "CCATCT": 2590, + "CCATCC": 2591, + "CCATCG": 2592, + "CCATGA": 2593, + "CCATGT": 2594, + "CCATGC": 2595, + "CCATGG": 2596, + "CCACAA": 2597, + "CCACAT": 2598, + "CCACAC": 2599, + "CCACAG": 2600, + "CCACTA": 2601, + "CCACTT": 2602, + "CCACTC": 2603, + "CCACTG": 2604, + "CCACCA": 2605, + "CCACCT": 2606, + "CCACCC": 2607, + "CCACCG": 2608, + "CCACGA": 2609, + "CCACGT": 2610, + "CCACGC": 2611, + "CCACGG": 2612, + "CCAGAA": 2613, + "CCAGAT": 2614, + "CCAGAC": 2615, + "CCAGAG": 2616, + "CCAGTA": 2617, + "CCAGTT": 2618, + "CCAGTC": 2619, + "CCAGTG": 2620, + "CCAGCA": 2621, + "CCAGCT": 2622, + "CCAGCC": 2623, + "CCAGCG": 2624, + "CCAGGA": 2625, + "CCAGGT": 2626, + "CCAGGC": 2627, + "CCAGGG": 2628, + "CCTAAA": 2629, + "CCTAAT": 2630, + "CCTAAC": 2631, + "CCTAAG": 2632, + "CCTATA": 2633, + "CCTATT": 2634, + "CCTATC": 2635, + "CCTATG": 2636, + "CCTACA": 2637, + "CCTACT": 2638, + "CCTACC": 2639, + "CCTACG": 2640, + "CCTAGA": 2641, + "CCTAGT": 2642, + "CCTAGC": 2643, + "CCTAGG": 2644, + "CCTTAA": 2645, + "CCTTAT": 2646, + "CCTTAC": 2647, + "CCTTAG": 2648, + "CCTTTA": 2649, + "CCTTTT": 2650, + "CCTTTC": 2651, + "CCTTTG": 2652, + "CCTTCA": 2653, + "CCTTCT": 2654, + "CCTTCC": 2655, + "CCTTCG": 2656, + "CCTTGA": 2657, + "CCTTGT": 2658, + "CCTTGC": 2659, + "CCTTGG": 2660, + "CCTCAA": 2661, + "CCTCAT": 2662, + "CCTCAC": 2663, + "CCTCAG": 2664, + "CCTCTA": 2665, + "CCTCTT": 2666, + "CCTCTC": 2667, + "CCTCTG": 2668, + "CCTCCA": 2669, + "CCTCCT": 2670, + "CCTCCC": 2671, + "CCTCCG": 2672, + "CCTCGA": 2673, + "CCTCGT": 2674, + "CCTCGC": 2675, + "CCTCGG": 2676, + "CCTGAA": 2677, + "CCTGAT": 2678, + "CCTGAC": 2679, + "CCTGAG": 2680, + "CCTGTA": 2681, + "CCTGTT": 2682, + "CCTGTC": 2683, + "CCTGTG": 2684, + "CCTGCA": 2685, + "CCTGCT": 2686, + "CCTGCC": 2687, + "CCTGCG": 2688, + "CCTGGA": 2689, + "CCTGGT": 2690, + "CCTGGC": 2691, + "CCTGGG": 2692, + "CCCAAA": 2693, + "CCCAAT": 2694, + "CCCAAC": 2695, + "CCCAAG": 2696, + "CCCATA": 2697, + "CCCATT": 2698, + "CCCATC": 2699, + "CCCATG": 2700, + "CCCACA": 2701, + "CCCACT": 2702, + "CCCACC": 2703, + "CCCACG": 2704, + "CCCAGA": 2705, + "CCCAGT": 2706, + "CCCAGC": 2707, + "CCCAGG": 2708, + "CCCTAA": 2709, + "CCCTAT": 2710, + "CCCTAC": 2711, + "CCCTAG": 2712, + "CCCTTA": 2713, + "CCCTTT": 2714, + "CCCTTC": 2715, + "CCCTTG": 2716, + "CCCTCA": 2717, + "CCCTCT": 2718, + "CCCTCC": 2719, + "CCCTCG": 2720, + "CCCTGA": 2721, + "CCCTGT": 2722, + "CCCTGC": 2723, + "CCCTGG": 2724, + "CCCCAA": 2725, + "CCCCAT": 2726, + "CCCCAC": 2727, + "CCCCAG": 2728, + "CCCCTA": 2729, + "CCCCTT": 2730, + "CCCCTC": 2731, + "CCCCTG": 2732, + "CCCCCA": 2733, + "CCCCCT": 2734, + "CCCCCC": 2735, + "CCCCCG": 2736, + "CCCCGA": 2737, + "CCCCGT": 2738, + "CCCCGC": 2739, + "CCCCGG": 2740, + "CCCGAA": 2741, + "CCCGAT": 2742, + "CCCGAC": 2743, + "CCCGAG": 2744, + "CCCGTA": 2745, + "CCCGTT": 2746, + "CCCGTC": 2747, + "CCCGTG": 2748, + "CCCGCA": 2749, + "CCCGCT": 2750, + "CCCGCC": 2751, + "CCCGCG": 2752, + "CCCGGA": 2753, + "CCCGGT": 2754, + "CCCGGC": 2755, + "CCCGGG": 2756, + "CCGAAA": 2757, + "CCGAAT": 2758, + "CCGAAC": 2759, + "CCGAAG": 2760, + "CCGATA": 2761, + "CCGATT": 2762, + "CCGATC": 2763, + "CCGATG": 2764, + "CCGACA": 2765, + "CCGACT": 2766, + "CCGACC": 2767, + "CCGACG": 2768, + "CCGAGA": 2769, + "CCGAGT": 2770, + "CCGAGC": 2771, + "CCGAGG": 2772, + "CCGTAA": 2773, + "CCGTAT": 2774, + "CCGTAC": 2775, + "CCGTAG": 2776, + "CCGTTA": 2777, + "CCGTTT": 2778, + "CCGTTC": 2779, + "CCGTTG": 2780, + "CCGTCA": 2781, + "CCGTCT": 2782, + "CCGTCC": 2783, + "CCGTCG": 2784, + "CCGTGA": 2785, + "CCGTGT": 2786, + "CCGTGC": 2787, + "CCGTGG": 2788, + "CCGCAA": 2789, + "CCGCAT": 2790, + "CCGCAC": 2791, + "CCGCAG": 2792, + "CCGCTA": 2793, + "CCGCTT": 2794, + "CCGCTC": 2795, + "CCGCTG": 2796, + "CCGCCA": 2797, + "CCGCCT": 2798, + "CCGCCC": 2799, + "CCGCCG": 2800, + "CCGCGA": 2801, + "CCGCGT": 2802, + "CCGCGC": 2803, + "CCGCGG": 2804, + "CCGGAA": 2805, + "CCGGAT": 2806, + "CCGGAC": 2807, + "CCGGAG": 2808, + "CCGGTA": 2809, + "CCGGTT": 2810, + "CCGGTC": 2811, + "CCGGTG": 2812, + "CCGGCA": 2813, + "CCGGCT": 2814, + "CCGGCC": 2815, + "CCGGCG": 2816, + "CCGGGA": 2817, + "CCGGGT": 2818, + "CCGGGC": 2819, + "CCGGGG": 2820, + "CGAAAA": 2821, + "CGAAAT": 2822, + "CGAAAC": 2823, + "CGAAAG": 2824, + "CGAATA": 2825, + "CGAATT": 2826, + "CGAATC": 2827, + "CGAATG": 2828, + "CGAACA": 2829, + "CGAACT": 2830, + "CGAACC": 2831, + "CGAACG": 2832, + "CGAAGA": 2833, + "CGAAGT": 2834, + "CGAAGC": 2835, + "CGAAGG": 2836, + "CGATAA": 2837, + "CGATAT": 2838, + "CGATAC": 2839, + "CGATAG": 2840, + "CGATTA": 2841, + "CGATTT": 2842, + "CGATTC": 2843, + "CGATTG": 2844, + "CGATCA": 2845, + "CGATCT": 2846, + "CGATCC": 2847, + "CGATCG": 2848, + "CGATGA": 2849, + "CGATGT": 2850, + "CGATGC": 2851, + "CGATGG": 2852, + "CGACAA": 2853, + "CGACAT": 2854, + "CGACAC": 2855, + "CGACAG": 2856, + "CGACTA": 2857, + "CGACTT": 2858, + "CGACTC": 2859, + "CGACTG": 2860, + "CGACCA": 2861, + "CGACCT": 2862, + "CGACCC": 2863, + "CGACCG": 2864, + "CGACGA": 2865, + "CGACGT": 2866, + "CGACGC": 2867, + "CGACGG": 2868, + "CGAGAA": 2869, + "CGAGAT": 2870, + "CGAGAC": 2871, + "CGAGAG": 2872, + "CGAGTA": 2873, + "CGAGTT": 2874, + "CGAGTC": 2875, + "CGAGTG": 2876, + "CGAGCA": 2877, + "CGAGCT": 2878, + "CGAGCC": 2879, + "CGAGCG": 2880, + "CGAGGA": 2881, + "CGAGGT": 2882, + "CGAGGC": 2883, + "CGAGGG": 2884, + "CGTAAA": 2885, + "CGTAAT": 2886, + "CGTAAC": 2887, + "CGTAAG": 2888, + "CGTATA": 2889, + "CGTATT": 2890, + "CGTATC": 2891, + "CGTATG": 2892, + "CGTACA": 2893, + "CGTACT": 2894, + "CGTACC": 2895, + "CGTACG": 2896, + "CGTAGA": 2897, + "CGTAGT": 2898, + "CGTAGC": 2899, + "CGTAGG": 2900, + "CGTTAA": 2901, + "CGTTAT": 2902, + "CGTTAC": 2903, + "CGTTAG": 2904, + "CGTTTA": 2905, + "CGTTTT": 2906, + "CGTTTC": 2907, + "CGTTTG": 2908, + "CGTTCA": 2909, + "CGTTCT": 2910, + "CGTTCC": 2911, + "CGTTCG": 2912, + "CGTTGA": 2913, + "CGTTGT": 2914, + "CGTTGC": 2915, + "CGTTGG": 2916, + "CGTCAA": 2917, + "CGTCAT": 2918, + "CGTCAC": 2919, + "CGTCAG": 2920, + "CGTCTA": 2921, + "CGTCTT": 2922, + "CGTCTC": 2923, + "CGTCTG": 2924, + "CGTCCA": 2925, + "CGTCCT": 2926, + "CGTCCC": 2927, + "CGTCCG": 2928, + "CGTCGA": 2929, + "CGTCGT": 2930, + "CGTCGC": 2931, + "CGTCGG": 2932, + "CGTGAA": 2933, + "CGTGAT": 2934, + "CGTGAC": 2935, + "CGTGAG": 2936, + "CGTGTA": 2937, + "CGTGTT": 2938, + "CGTGTC": 2939, + "CGTGTG": 2940, + "CGTGCA": 2941, + "CGTGCT": 2942, + "CGTGCC": 2943, + "CGTGCG": 2944, + "CGTGGA": 2945, + "CGTGGT": 2946, + "CGTGGC": 2947, + "CGTGGG": 2948, + "CGCAAA": 2949, + "CGCAAT": 2950, + "CGCAAC": 2951, + "CGCAAG": 2952, + "CGCATA": 2953, + "CGCATT": 2954, + "CGCATC": 2955, + "CGCATG": 2956, + "CGCACA": 2957, + "CGCACT": 2958, + "CGCACC": 2959, + "CGCACG": 2960, + "CGCAGA": 2961, + "CGCAGT": 2962, + "CGCAGC": 2963, + "CGCAGG": 2964, + "CGCTAA": 2965, + "CGCTAT": 2966, + "CGCTAC": 2967, + "CGCTAG": 2968, + "CGCTTA": 2969, + "CGCTTT": 2970, + "CGCTTC": 2971, + "CGCTTG": 2972, + "CGCTCA": 2973, + "CGCTCT": 2974, + "CGCTCC": 2975, + "CGCTCG": 2976, + "CGCTGA": 2977, + "CGCTGT": 2978, + "CGCTGC": 2979, + "CGCTGG": 2980, + "CGCCAA": 2981, + "CGCCAT": 2982, + "CGCCAC": 2983, + "CGCCAG": 2984, + "CGCCTA": 2985, + "CGCCTT": 2986, + "CGCCTC": 2987, + "CGCCTG": 2988, + "CGCCCA": 2989, + "CGCCCT": 2990, + "CGCCCC": 2991, + "CGCCCG": 2992, + "CGCCGA": 2993, + "CGCCGT": 2994, + "CGCCGC": 2995, + "CGCCGG": 2996, + "CGCGAA": 2997, + "CGCGAT": 2998, + "CGCGAC": 2999, + "CGCGAG": 3000, + "CGCGTA": 3001, + "CGCGTT": 3002, + "CGCGTC": 3003, + "CGCGTG": 3004, + "CGCGCA": 3005, + "CGCGCT": 3006, + "CGCGCC": 3007, + "CGCGCG": 3008, + "CGCGGA": 3009, + "CGCGGT": 3010, + "CGCGGC": 3011, + "CGCGGG": 3012, + "CGGAAA": 3013, + "CGGAAT": 3014, + "CGGAAC": 3015, + "CGGAAG": 3016, + "CGGATA": 3017, + "CGGATT": 3018, + "CGGATC": 3019, + "CGGATG": 3020, + "CGGACA": 3021, + "CGGACT": 3022, + "CGGACC": 3023, + "CGGACG": 3024, + "CGGAGA": 3025, + "CGGAGT": 3026, + "CGGAGC": 3027, + "CGGAGG": 3028, + "CGGTAA": 3029, + "CGGTAT": 3030, + "CGGTAC": 3031, + "CGGTAG": 3032, + "CGGTTA": 3033, + "CGGTTT": 3034, + "CGGTTC": 3035, + "CGGTTG": 3036, + "CGGTCA": 3037, + "CGGTCT": 3038, + "CGGTCC": 3039, + "CGGTCG": 3040, + "CGGTGA": 3041, + "CGGTGT": 3042, + "CGGTGC": 3043, + "CGGTGG": 3044, + "CGGCAA": 3045, + "CGGCAT": 3046, + "CGGCAC": 3047, + "CGGCAG": 3048, + "CGGCTA": 3049, + "CGGCTT": 3050, + "CGGCTC": 3051, + "CGGCTG": 3052, + "CGGCCA": 3053, + "CGGCCT": 3054, + "CGGCCC": 3055, + "CGGCCG": 3056, + "CGGCGA": 3057, + "CGGCGT": 3058, + "CGGCGC": 3059, + "CGGCGG": 3060, + "CGGGAA": 3061, + "CGGGAT": 3062, + "CGGGAC": 3063, + "CGGGAG": 3064, + "CGGGTA": 3065, + "CGGGTT": 3066, + "CGGGTC": 3067, + "CGGGTG": 3068, + "CGGGCA": 3069, + "CGGGCT": 3070, + "CGGGCC": 3071, + "CGGGCG": 3072, + "CGGGGA": 3073, + "CGGGGT": 3074, + "CGGGGC": 3075, + "CGGGGG": 3076, + "GAAAAA": 3077, + "GAAAAT": 3078, + "GAAAAC": 3079, + "GAAAAG": 3080, + "GAAATA": 3081, + "GAAATT": 3082, + "GAAATC": 3083, + "GAAATG": 3084, + "GAAACA": 3085, + "GAAACT": 3086, + "GAAACC": 3087, + "GAAACG": 3088, + "GAAAGA": 3089, + "GAAAGT": 3090, + "GAAAGC": 3091, + "GAAAGG": 3092, + "GAATAA": 3093, + "GAATAT": 3094, + "GAATAC": 3095, + "GAATAG": 3096, + "GAATTA": 3097, + "GAATTT": 3098, + "GAATTC": 3099, + "GAATTG": 3100, + "GAATCA": 3101, + "GAATCT": 3102, + "GAATCC": 3103, + "GAATCG": 3104, + "GAATGA": 3105, + "GAATGT": 3106, + "GAATGC": 3107, + "GAATGG": 3108, + "GAACAA": 3109, + "GAACAT": 3110, + "GAACAC": 3111, + "GAACAG": 3112, + "GAACTA": 3113, + "GAACTT": 3114, + "GAACTC": 3115, + "GAACTG": 3116, + "GAACCA": 3117, + "GAACCT": 3118, + "GAACCC": 3119, + "GAACCG": 3120, + "GAACGA": 3121, + "GAACGT": 3122, + "GAACGC": 3123, + "GAACGG": 3124, + "GAAGAA": 3125, + "GAAGAT": 3126, + "GAAGAC": 3127, + "GAAGAG": 3128, + "GAAGTA": 3129, + "GAAGTT": 3130, + "GAAGTC": 3131, + "GAAGTG": 3132, + "GAAGCA": 3133, + "GAAGCT": 3134, + "GAAGCC": 3135, + "GAAGCG": 3136, + "GAAGGA": 3137, + "GAAGGT": 3138, + "GAAGGC": 3139, + "GAAGGG": 3140, + "GATAAA": 3141, + "GATAAT": 3142, + "GATAAC": 3143, + "GATAAG": 3144, + "GATATA": 3145, + "GATATT": 3146, + "GATATC": 3147, + "GATATG": 3148, + "GATACA": 3149, + "GATACT": 3150, + "GATACC": 3151, + "GATACG": 3152, + "GATAGA": 3153, + "GATAGT": 3154, + "GATAGC": 3155, + "GATAGG": 3156, + "GATTAA": 3157, + "GATTAT": 3158, + "GATTAC": 3159, + "GATTAG": 3160, + "GATTTA": 3161, + "GATTTT": 3162, + "GATTTC": 3163, + "GATTTG": 3164, + "GATTCA": 3165, + "GATTCT": 3166, + "GATTCC": 3167, + "GATTCG": 3168, + "GATTGA": 3169, + "GATTGT": 3170, + "GATTGC": 3171, + "GATTGG": 3172, + "GATCAA": 3173, + "GATCAT": 3174, + "GATCAC": 3175, + "GATCAG": 3176, + "GATCTA": 3177, + "GATCTT": 3178, + "GATCTC": 3179, + "GATCTG": 3180, + "GATCCA": 3181, + "GATCCT": 3182, + "GATCCC": 3183, + "GATCCG": 3184, + "GATCGA": 3185, + "GATCGT": 3186, + "GATCGC": 3187, + "GATCGG": 3188, + "GATGAA": 3189, + "GATGAT": 3190, + "GATGAC": 3191, + "GATGAG": 3192, + "GATGTA": 3193, + "GATGTT": 3194, + "GATGTC": 3195, + "GATGTG": 3196, + "GATGCA": 3197, + "GATGCT": 3198, + "GATGCC": 3199, + "GATGCG": 3200, + "GATGGA": 3201, + "GATGGT": 3202, + "GATGGC": 3203, + "GATGGG": 3204, + "GACAAA": 3205, + "GACAAT": 3206, + "GACAAC": 3207, + "GACAAG": 3208, + "GACATA": 3209, + "GACATT": 3210, + "GACATC": 3211, + "GACATG": 3212, + "GACACA": 3213, + "GACACT": 3214, + "GACACC": 3215, + "GACACG": 3216, + "GACAGA": 3217, + "GACAGT": 3218, + "GACAGC": 3219, + "GACAGG": 3220, + "GACTAA": 3221, + "GACTAT": 3222, + "GACTAC": 3223, + "GACTAG": 3224, + "GACTTA": 3225, + "GACTTT": 3226, + "GACTTC": 3227, + "GACTTG": 3228, + "GACTCA": 3229, + "GACTCT": 3230, + "GACTCC": 3231, + "GACTCG": 3232, + "GACTGA": 3233, + "GACTGT": 3234, + "GACTGC": 3235, + "GACTGG": 3236, + "GACCAA": 3237, + "GACCAT": 3238, + "GACCAC": 3239, + "GACCAG": 3240, + "GACCTA": 3241, + "GACCTT": 3242, + "GACCTC": 3243, + "GACCTG": 3244, + "GACCCA": 3245, + "GACCCT": 3246, + "GACCCC": 3247, + "GACCCG": 3248, + "GACCGA": 3249, + "GACCGT": 3250, + "GACCGC": 3251, + "GACCGG": 3252, + "GACGAA": 3253, + "GACGAT": 3254, + "GACGAC": 3255, + "GACGAG": 3256, + "GACGTA": 3257, + "GACGTT": 3258, + "GACGTC": 3259, + "GACGTG": 3260, + "GACGCA": 3261, + "GACGCT": 3262, + "GACGCC": 3263, + "GACGCG": 3264, + "GACGGA": 3265, + "GACGGT": 3266, + "GACGGC": 3267, + "GACGGG": 3268, + "GAGAAA": 3269, + "GAGAAT": 3270, + "GAGAAC": 3271, + "GAGAAG": 3272, + "GAGATA": 3273, + "GAGATT": 3274, + "GAGATC": 3275, + "GAGATG": 3276, + "GAGACA": 3277, + "GAGACT": 3278, + "GAGACC": 3279, + "GAGACG": 3280, + "GAGAGA": 3281, + "GAGAGT": 3282, + "GAGAGC": 3283, + "GAGAGG": 3284, + "GAGTAA": 3285, + "GAGTAT": 3286, + "GAGTAC": 3287, + "GAGTAG": 3288, + "GAGTTA": 3289, + "GAGTTT": 3290, + "GAGTTC": 3291, + "GAGTTG": 3292, + "GAGTCA": 3293, + "GAGTCT": 3294, + "GAGTCC": 3295, + "GAGTCG": 3296, + "GAGTGA": 3297, + "GAGTGT": 3298, + "GAGTGC": 3299, + "GAGTGG": 3300, + "GAGCAA": 3301, + "GAGCAT": 3302, + "GAGCAC": 3303, + "GAGCAG": 3304, + "GAGCTA": 3305, + "GAGCTT": 3306, + "GAGCTC": 3307, + "GAGCTG": 3308, + "GAGCCA": 3309, + "GAGCCT": 3310, + "GAGCCC": 3311, + "GAGCCG": 3312, + "GAGCGA": 3313, + "GAGCGT": 3314, + "GAGCGC": 3315, + "GAGCGG": 3316, + "GAGGAA": 3317, + "GAGGAT": 3318, + "GAGGAC": 3319, + "GAGGAG": 3320, + "GAGGTA": 3321, + "GAGGTT": 3322, + "GAGGTC": 3323, + "GAGGTG": 3324, + "GAGGCA": 3325, + "GAGGCT": 3326, + "GAGGCC": 3327, + "GAGGCG": 3328, + "GAGGGA": 3329, + "GAGGGT": 3330, + "GAGGGC": 3331, + "GAGGGG": 3332, + "GTAAAA": 3333, + "GTAAAT": 3334, + "GTAAAC": 3335, + "GTAAAG": 3336, + "GTAATA": 3337, + "GTAATT": 3338, + "GTAATC": 3339, + "GTAATG": 3340, + "GTAACA": 3341, + "GTAACT": 3342, + "GTAACC": 3343, + "GTAACG": 3344, + "GTAAGA": 3345, + "GTAAGT": 3346, + "GTAAGC": 3347, + "GTAAGG": 3348, + "GTATAA": 3349, + "GTATAT": 3350, + "GTATAC": 3351, + "GTATAG": 3352, + "GTATTA": 3353, + "GTATTT": 3354, + "GTATTC": 3355, + "GTATTG": 3356, + "GTATCA": 3357, + "GTATCT": 3358, + "GTATCC": 3359, + "GTATCG": 3360, + "GTATGA": 3361, + "GTATGT": 3362, + "GTATGC": 3363, + "GTATGG": 3364, + "GTACAA": 3365, + "GTACAT": 3366, + "GTACAC": 3367, + "GTACAG": 3368, + "GTACTA": 3369, + "GTACTT": 3370, + "GTACTC": 3371, + "GTACTG": 3372, + "GTACCA": 3373, + "GTACCT": 3374, + "GTACCC": 3375, + "GTACCG": 3376, + "GTACGA": 3377, + "GTACGT": 3378, + "GTACGC": 3379, + "GTACGG": 3380, + "GTAGAA": 3381, + "GTAGAT": 3382, + "GTAGAC": 3383, + "GTAGAG": 3384, + "GTAGTA": 3385, + "GTAGTT": 3386, + "GTAGTC": 3387, + "GTAGTG": 3388, + "GTAGCA": 3389, + "GTAGCT": 3390, + "GTAGCC": 3391, + "GTAGCG": 3392, + "GTAGGA": 3393, + "GTAGGT": 3394, + "GTAGGC": 3395, + "GTAGGG": 3396, + "GTTAAA": 3397, + "GTTAAT": 3398, + "GTTAAC": 3399, + "GTTAAG": 3400, + "GTTATA": 3401, + "GTTATT": 3402, + "GTTATC": 3403, + "GTTATG": 3404, + "GTTACA": 3405, + "GTTACT": 3406, + "GTTACC": 3407, + "GTTACG": 3408, + "GTTAGA": 3409, + "GTTAGT": 3410, + "GTTAGC": 3411, + "GTTAGG": 3412, + "GTTTAA": 3413, + "GTTTAT": 3414, + "GTTTAC": 3415, + "GTTTAG": 3416, + "GTTTTA": 3417, + "GTTTTT": 3418, + "GTTTTC": 3419, + "GTTTTG": 3420, + "GTTTCA": 3421, + "GTTTCT": 3422, + "GTTTCC": 3423, + "GTTTCG": 3424, + "GTTTGA": 3425, + "GTTTGT": 3426, + "GTTTGC": 3427, + "GTTTGG": 3428, + "GTTCAA": 3429, + "GTTCAT": 3430, + "GTTCAC": 3431, + "GTTCAG": 3432, + "GTTCTA": 3433, + "GTTCTT": 3434, + "GTTCTC": 3435, + "GTTCTG": 3436, + "GTTCCA": 3437, + "GTTCCT": 3438, + "GTTCCC": 3439, + "GTTCCG": 3440, + "GTTCGA": 3441, + "GTTCGT": 3442, + "GTTCGC": 3443, + "GTTCGG": 3444, + "GTTGAA": 3445, + "GTTGAT": 3446, + "GTTGAC": 3447, + "GTTGAG": 3448, + "GTTGTA": 3449, + "GTTGTT": 3450, + "GTTGTC": 3451, + "GTTGTG": 3452, + "GTTGCA": 3453, + "GTTGCT": 3454, + "GTTGCC": 3455, + "GTTGCG": 3456, + "GTTGGA": 3457, + "GTTGGT": 3458, + "GTTGGC": 3459, + "GTTGGG": 3460, + "GTCAAA": 3461, + "GTCAAT": 3462, + "GTCAAC": 3463, + "GTCAAG": 3464, + "GTCATA": 3465, + "GTCATT": 3466, + "GTCATC": 3467, + "GTCATG": 3468, + "GTCACA": 3469, + "GTCACT": 3470, + "GTCACC": 3471, + "GTCACG": 3472, + "GTCAGA": 3473, + "GTCAGT": 3474, + "GTCAGC": 3475, + "GTCAGG": 3476, + "GTCTAA": 3477, + "GTCTAT": 3478, + "GTCTAC": 3479, + "GTCTAG": 3480, + "GTCTTA": 3481, + "GTCTTT": 3482, + "GTCTTC": 3483, + "GTCTTG": 3484, + "GTCTCA": 3485, + "GTCTCT": 3486, + "GTCTCC": 3487, + "GTCTCG": 3488, + "GTCTGA": 3489, + "GTCTGT": 3490, + "GTCTGC": 3491, + "GTCTGG": 3492, + "GTCCAA": 3493, + "GTCCAT": 3494, + "GTCCAC": 3495, + "GTCCAG": 3496, + "GTCCTA": 3497, + "GTCCTT": 3498, + "GTCCTC": 3499, + "GTCCTG": 3500, + "GTCCCA": 3501, + "GTCCCT": 3502, + "GTCCCC": 3503, + "GTCCCG": 3504, + "GTCCGA": 3505, + "GTCCGT": 3506, + "GTCCGC": 3507, + "GTCCGG": 3508, + "GTCGAA": 3509, + "GTCGAT": 3510, + "GTCGAC": 3511, + "GTCGAG": 3512, + "GTCGTA": 3513, + "GTCGTT": 3514, + "GTCGTC": 3515, + "GTCGTG": 3516, + "GTCGCA": 3517, + "GTCGCT": 3518, + "GTCGCC": 3519, + "GTCGCG": 3520, + "GTCGGA": 3521, + "GTCGGT": 3522, + "GTCGGC": 3523, + "GTCGGG": 3524, + "GTGAAA": 3525, + "GTGAAT": 3526, + "GTGAAC": 3527, + "GTGAAG": 3528, + "GTGATA": 3529, + "GTGATT": 3530, + "GTGATC": 3531, + "GTGATG": 3532, + "GTGACA": 3533, + "GTGACT": 3534, + "GTGACC": 3535, + "GTGACG": 3536, + "GTGAGA": 3537, + "GTGAGT": 3538, + "GTGAGC": 3539, + "GTGAGG": 3540, + "GTGTAA": 3541, + "GTGTAT": 3542, + "GTGTAC": 3543, + "GTGTAG": 3544, + "GTGTTA": 3545, + "GTGTTT": 3546, + "GTGTTC": 3547, + "GTGTTG": 3548, + "GTGTCA": 3549, + "GTGTCT": 3550, + "GTGTCC": 3551, + "GTGTCG": 3552, + "GTGTGA": 3553, + "GTGTGT": 3554, + "GTGTGC": 3555, + "GTGTGG": 3556, + "GTGCAA": 3557, + "GTGCAT": 3558, + "GTGCAC": 3559, + "GTGCAG": 3560, + "GTGCTA": 3561, + "GTGCTT": 3562, + "GTGCTC": 3563, + "GTGCTG": 3564, + "GTGCCA": 3565, + "GTGCCT": 3566, + "GTGCCC": 3567, + "GTGCCG": 3568, + "GTGCGA": 3569, + "GTGCGT": 3570, + "GTGCGC": 3571, + "GTGCGG": 3572, + "GTGGAA": 3573, + "GTGGAT": 3574, + "GTGGAC": 3575, + "GTGGAG": 3576, + "GTGGTA": 3577, + "GTGGTT": 3578, + "GTGGTC": 3579, + "GTGGTG": 3580, + "GTGGCA": 3581, + "GTGGCT": 3582, + "GTGGCC": 3583, + "GTGGCG": 3584, + "GTGGGA": 3585, + "GTGGGT": 3586, + "GTGGGC": 3587, + "GTGGGG": 3588, + "GCAAAA": 3589, + "GCAAAT": 3590, + "GCAAAC": 3591, + "GCAAAG": 3592, + "GCAATA": 3593, + "GCAATT": 3594, + "GCAATC": 3595, + "GCAATG": 3596, + "GCAACA": 3597, + "GCAACT": 3598, + "GCAACC": 3599, + "GCAACG": 3600, + "GCAAGA": 3601, + "GCAAGT": 3602, + "GCAAGC": 3603, + "GCAAGG": 3604, + "GCATAA": 3605, + "GCATAT": 3606, + "GCATAC": 3607, + "GCATAG": 3608, + "GCATTA": 3609, + "GCATTT": 3610, + "GCATTC": 3611, + "GCATTG": 3612, + "GCATCA": 3613, + "GCATCT": 3614, + "GCATCC": 3615, + "GCATCG": 3616, + "GCATGA": 3617, + "GCATGT": 3618, + "GCATGC": 3619, + "GCATGG": 3620, + "GCACAA": 3621, + "GCACAT": 3622, + "GCACAC": 3623, + "GCACAG": 3624, + "GCACTA": 3625, + "GCACTT": 3626, + "GCACTC": 3627, + "GCACTG": 3628, + "GCACCA": 3629, + "GCACCT": 3630, + "GCACCC": 3631, + "GCACCG": 3632, + "GCACGA": 3633, + "GCACGT": 3634, + "GCACGC": 3635, + "GCACGG": 3636, + "GCAGAA": 3637, + "GCAGAT": 3638, + "GCAGAC": 3639, + "GCAGAG": 3640, + "GCAGTA": 3641, + "GCAGTT": 3642, + "GCAGTC": 3643, + "GCAGTG": 3644, + "GCAGCA": 3645, + "GCAGCT": 3646, + "GCAGCC": 3647, + "GCAGCG": 3648, + "GCAGGA": 3649, + "GCAGGT": 3650, + "GCAGGC": 3651, + "GCAGGG": 3652, + "GCTAAA": 3653, + "GCTAAT": 3654, + "GCTAAC": 3655, + "GCTAAG": 3656, + "GCTATA": 3657, + "GCTATT": 3658, + "GCTATC": 3659, + "GCTATG": 3660, + "GCTACA": 3661, + "GCTACT": 3662, + "GCTACC": 3663, + "GCTACG": 3664, + "GCTAGA": 3665, + "GCTAGT": 3666, + "GCTAGC": 3667, + "GCTAGG": 3668, + "GCTTAA": 3669, + "GCTTAT": 3670, + "GCTTAC": 3671, + "GCTTAG": 3672, + "GCTTTA": 3673, + "GCTTTT": 3674, + "GCTTTC": 3675, + "GCTTTG": 3676, + "GCTTCA": 3677, + "GCTTCT": 3678, + "GCTTCC": 3679, + "GCTTCG": 3680, + "GCTTGA": 3681, + "GCTTGT": 3682, + "GCTTGC": 3683, + "GCTTGG": 3684, + "GCTCAA": 3685, + "GCTCAT": 3686, + "GCTCAC": 3687, + "GCTCAG": 3688, + "GCTCTA": 3689, + "GCTCTT": 3690, + "GCTCTC": 3691, + "GCTCTG": 3692, + "GCTCCA": 3693, + "GCTCCT": 3694, + "GCTCCC": 3695, + "GCTCCG": 3696, + "GCTCGA": 3697, + "GCTCGT": 3698, + "GCTCGC": 3699, + "GCTCGG": 3700, + "GCTGAA": 3701, + "GCTGAT": 3702, + "GCTGAC": 3703, + "GCTGAG": 3704, + "GCTGTA": 3705, + "GCTGTT": 3706, + "GCTGTC": 3707, + "GCTGTG": 3708, + "GCTGCA": 3709, + "GCTGCT": 3710, + "GCTGCC": 3711, + "GCTGCG": 3712, + "GCTGGA": 3713, + "GCTGGT": 3714, + "GCTGGC": 3715, + "GCTGGG": 3716, + "GCCAAA": 3717, + "GCCAAT": 3718, + "GCCAAC": 3719, + "GCCAAG": 3720, + "GCCATA": 3721, + "GCCATT": 3722, + "GCCATC": 3723, + "GCCATG": 3724, + "GCCACA": 3725, + "GCCACT": 3726, + "GCCACC": 3727, + "GCCACG": 3728, + "GCCAGA": 3729, + "GCCAGT": 3730, + "GCCAGC": 3731, + "GCCAGG": 3732, + "GCCTAA": 3733, + "GCCTAT": 3734, + "GCCTAC": 3735, + "GCCTAG": 3736, + "GCCTTA": 3737, + "GCCTTT": 3738, + "GCCTTC": 3739, + "GCCTTG": 3740, + "GCCTCA": 3741, + "GCCTCT": 3742, + "GCCTCC": 3743, + "GCCTCG": 3744, + "GCCTGA": 3745, + "GCCTGT": 3746, + "GCCTGC": 3747, + "GCCTGG": 3748, + "GCCCAA": 3749, + "GCCCAT": 3750, + "GCCCAC": 3751, + "GCCCAG": 3752, + "GCCCTA": 3753, + "GCCCTT": 3754, + "GCCCTC": 3755, + "GCCCTG": 3756, + "GCCCCA": 3757, + "GCCCCT": 3758, + "GCCCCC": 3759, + "GCCCCG": 3760, + "GCCCGA": 3761, + "GCCCGT": 3762, + "GCCCGC": 3763, + "GCCCGG": 3764, + "GCCGAA": 3765, + "GCCGAT": 3766, + "GCCGAC": 3767, + "GCCGAG": 3768, + "GCCGTA": 3769, + "GCCGTT": 3770, + "GCCGTC": 3771, + "GCCGTG": 3772, + "GCCGCA": 3773, + "GCCGCT": 3774, + "GCCGCC": 3775, + "GCCGCG": 3776, + "GCCGGA": 3777, + "GCCGGT": 3778, + "GCCGGC": 3779, + "GCCGGG": 3780, + "GCGAAA": 3781, + "GCGAAT": 3782, + "GCGAAC": 3783, + "GCGAAG": 3784, + "GCGATA": 3785, + "GCGATT": 3786, + "GCGATC": 3787, + "GCGATG": 3788, + "GCGACA": 3789, + "GCGACT": 3790, + "GCGACC": 3791, + "GCGACG": 3792, + "GCGAGA": 3793, + "GCGAGT": 3794, + "GCGAGC": 3795, + "GCGAGG": 3796, + "GCGTAA": 3797, + "GCGTAT": 3798, + "GCGTAC": 3799, + "GCGTAG": 3800, + "GCGTTA": 3801, + "GCGTTT": 3802, + "GCGTTC": 3803, + "GCGTTG": 3804, + "GCGTCA": 3805, + "GCGTCT": 3806, + "GCGTCC": 3807, + "GCGTCG": 3808, + "GCGTGA": 3809, + "GCGTGT": 3810, + "GCGTGC": 3811, + "GCGTGG": 3812, + "GCGCAA": 3813, + "GCGCAT": 3814, + "GCGCAC": 3815, + "GCGCAG": 3816, + "GCGCTA": 3817, + "GCGCTT": 3818, + "GCGCTC": 3819, + "GCGCTG": 3820, + "GCGCCA": 3821, + "GCGCCT": 3822, + "GCGCCC": 3823, + "GCGCCG": 3824, + "GCGCGA": 3825, + "GCGCGT": 3826, + "GCGCGC": 3827, + "GCGCGG": 3828, + "GCGGAA": 3829, + "GCGGAT": 3830, + "GCGGAC": 3831, + "GCGGAG": 3832, + "GCGGTA": 3833, + "GCGGTT": 3834, + "GCGGTC": 3835, + "GCGGTG": 3836, + "GCGGCA": 3837, + "GCGGCT": 3838, + "GCGGCC": 3839, + "GCGGCG": 3840, + "GCGGGA": 3841, + "GCGGGT": 3842, + "GCGGGC": 3843, + "GCGGGG": 3844, + "GGAAAA": 3845, + "GGAAAT": 3846, + "GGAAAC": 3847, + "GGAAAG": 3848, + "GGAATA": 3849, + "GGAATT": 3850, + "GGAATC": 3851, + "GGAATG": 3852, + "GGAACA": 3853, + "GGAACT": 3854, + "GGAACC": 3855, + "GGAACG": 3856, + "GGAAGA": 3857, + "GGAAGT": 3858, + "GGAAGC": 3859, + "GGAAGG": 3860, + "GGATAA": 3861, + "GGATAT": 3862, + "GGATAC": 3863, + "GGATAG": 3864, + "GGATTA": 3865, + "GGATTT": 3866, + "GGATTC": 3867, + "GGATTG": 3868, + "GGATCA": 3869, + "GGATCT": 3870, + "GGATCC": 3871, + "GGATCG": 3872, + "GGATGA": 3873, + "GGATGT": 3874, + "GGATGC": 3875, + "GGATGG": 3876, + "GGACAA": 3877, + "GGACAT": 3878, + "GGACAC": 3879, + "GGACAG": 3880, + "GGACTA": 3881, + "GGACTT": 3882, + "GGACTC": 3883, + "GGACTG": 3884, + "GGACCA": 3885, + "GGACCT": 3886, + "GGACCC": 3887, + "GGACCG": 3888, + "GGACGA": 3889, + "GGACGT": 3890, + "GGACGC": 3891, + "GGACGG": 3892, + "GGAGAA": 3893, + "GGAGAT": 3894, + "GGAGAC": 3895, + "GGAGAG": 3896, + "GGAGTA": 3897, + "GGAGTT": 3898, + "GGAGTC": 3899, + "GGAGTG": 3900, + "GGAGCA": 3901, + "GGAGCT": 3902, + "GGAGCC": 3903, + "GGAGCG": 3904, + "GGAGGA": 3905, + "GGAGGT": 3906, + "GGAGGC": 3907, + "GGAGGG": 3908, + "GGTAAA": 3909, + "GGTAAT": 3910, + "GGTAAC": 3911, + "GGTAAG": 3912, + "GGTATA": 3913, + "GGTATT": 3914, + "GGTATC": 3915, + "GGTATG": 3916, + "GGTACA": 3917, + "GGTACT": 3918, + "GGTACC": 3919, + "GGTACG": 3920, + "GGTAGA": 3921, + "GGTAGT": 3922, + "GGTAGC": 3923, + "GGTAGG": 3924, + "GGTTAA": 3925, + "GGTTAT": 3926, + "GGTTAC": 3927, + "GGTTAG": 3928, + "GGTTTA": 3929, + "GGTTTT": 3930, + "GGTTTC": 3931, + "GGTTTG": 3932, + "GGTTCA": 3933, + "GGTTCT": 3934, + "GGTTCC": 3935, + "GGTTCG": 3936, + "GGTTGA": 3937, + "GGTTGT": 3938, + "GGTTGC": 3939, + "GGTTGG": 3940, + "GGTCAA": 3941, + "GGTCAT": 3942, + "GGTCAC": 3943, + "GGTCAG": 3944, + "GGTCTA": 3945, + "GGTCTT": 3946, + "GGTCTC": 3947, + "GGTCTG": 3948, + "GGTCCA": 3949, + "GGTCCT": 3950, + "GGTCCC": 3951, + "GGTCCG": 3952, + "GGTCGA": 3953, + "GGTCGT": 3954, + "GGTCGC": 3955, + "GGTCGG": 3956, + "GGTGAA": 3957, + "GGTGAT": 3958, + "GGTGAC": 3959, + "GGTGAG": 3960, + "GGTGTA": 3961, + "GGTGTT": 3962, + "GGTGTC": 3963, + "GGTGTG": 3964, + "GGTGCA": 3965, + "GGTGCT": 3966, + "GGTGCC": 3967, + "GGTGCG": 3968, + "GGTGGA": 3969, + "GGTGGT": 3970, + "GGTGGC": 3971, + "GGTGGG": 3972, + "GGCAAA": 3973, + "GGCAAT": 3974, + "GGCAAC": 3975, + "GGCAAG": 3976, + "GGCATA": 3977, + "GGCATT": 3978, + "GGCATC": 3979, + "GGCATG": 3980, + "GGCACA": 3981, + "GGCACT": 3982, + "GGCACC": 3983, + "GGCACG": 3984, + "GGCAGA": 3985, + "GGCAGT": 3986, + "GGCAGC": 3987, + "GGCAGG": 3988, + "GGCTAA": 3989, + "GGCTAT": 3990, + "GGCTAC": 3991, + "GGCTAG": 3992, + "GGCTTA": 3993, + "GGCTTT": 3994, + "GGCTTC": 3995, + "GGCTTG": 3996, + "GGCTCA": 3997, + "GGCTCT": 3998, + "GGCTCC": 3999, + "GGCTCG": 4000, + "GGCTGA": 4001, + "GGCTGT": 4002, + "GGCTGC": 4003, + "GGCTGG": 4004, + "GGCCAA": 4005, + "GGCCAT": 4006, + "GGCCAC": 4007, + "GGCCAG": 4008, + "GGCCTA": 4009, + "GGCCTT": 4010, + "GGCCTC": 4011, + "GGCCTG": 4012, + "GGCCCA": 4013, + "GGCCCT": 4014, + "GGCCCC": 4015, + "GGCCCG": 4016, + "GGCCGA": 4017, + "GGCCGT": 4018, + "GGCCGC": 4019, + "GGCCGG": 4020, + "GGCGAA": 4021, + "GGCGAT": 4022, + "GGCGAC": 4023, + "GGCGAG": 4024, + "GGCGTA": 4025, + "GGCGTT": 4026, + "GGCGTC": 4027, + "GGCGTG": 4028, + "GGCGCA": 4029, + "GGCGCT": 4030, + "GGCGCC": 4031, + "GGCGCG": 4032, + "GGCGGA": 4033, + "GGCGGT": 4034, + "GGCGGC": 4035, + "GGCGGG": 4036, + "GGGAAA": 4037, + "GGGAAT": 4038, + "GGGAAC": 4039, + "GGGAAG": 4040, + "GGGATA": 4041, + "GGGATT": 4042, + "GGGATC": 4043, + "GGGATG": 4044, + "GGGACA": 4045, + "GGGACT": 4046, + "GGGACC": 4047, + "GGGACG": 4048, + "GGGAGA": 4049, + "GGGAGT": 4050, + "GGGAGC": 4051, + "GGGAGG": 4052, + "GGGTAA": 4053, + "GGGTAT": 4054, + "GGGTAC": 4055, + "GGGTAG": 4056, + "GGGTTA": 4057, + "GGGTTT": 4058, + "GGGTTC": 4059, + "GGGTTG": 4060, + "GGGTCA": 4061, + "GGGTCT": 4062, + "GGGTCC": 4063, + "GGGTCG": 4064, + "GGGTGA": 4065, + "GGGTGT": 4066, + "GGGTGC": 4067, + "GGGTGG": 4068, + "GGGCAA": 4069, + "GGGCAT": 4070, + "GGGCAC": 4071, + "GGGCAG": 4072, + "GGGCTA": 4073, + "GGGCTT": 4074, + "GGGCTC": 4075, + "GGGCTG": 4076, + "GGGCCA": 4077, + "GGGCCT": 4078, + "GGGCCC": 4079, + "GGGCCG": 4080, + "GGGCGA": 4081, + "GGGCGT": 4082, + "GGGCGC": 4083, + "GGGCGG": 4084, + "GGGGAA": 4085, + "GGGGAT": 4086, + "GGGGAC": 4087, + "GGGGAG": 4088, + "GGGGTA": 4089, + "GGGGTT": 4090, + "GGGGTC": 4091, + "GGGGTG": 4092, + "GGGGCA": 4093, + "GGGGCT": 4094, + "GGGGCC": 4095, + "GGGGCG": 4096, + "GGGGGA": 4097, + "GGGGGT": 4098, + "GGGGGC": 4099, + "GGGGGG": 4100 + } + } +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-1000/tokenizer_config.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-1000/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..604e2b1b29af1dcb3920d2828b6a90276e5e3d02 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-1000/tokenizer_config.json @@ -0,0 +1,58 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "[UNK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "[CLS]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "[SEP]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "4": { + "content": "[MASK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "clean_up_tokenization_spaces": true, + "cls_token": "[CLS]", + "do_basic_tokenize": true, + "do_lower_case": false, + "mask_token": "[MASK]", + "max_len": 512, + "model_max_length": 512, + "never_split": null, + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "strip_accents": null, + "tokenize_chinese_chars": true, + "tokenizer_class": "BertTokenizer", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-1000/trainer_state.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-1000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..d929be8c15f3b131469cca3bbacb2cd28b9cbb6d --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-1000/trainer_state.json @@ -0,0 +1,313 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 4.0, + "eval_steps": 500, + "global_step": 1000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.1, + "grad_norm": 0.6611918807029724, + "learning_rate": 5.319148936170213e-05, + "loss": 0.6932, + "step": 25 + }, + { + "epoch": 0.2, + "grad_norm": 0.9467485547065735, + "learning_rate": 0.00010638297872340425, + "loss": 0.6912, + "step": 50 + }, + { + "epoch": 0.3, + "grad_norm": 0.9465051889419556, + "learning_rate": 0.00015957446808510637, + "loss": 0.6893, + "step": 75 + }, + { + "epoch": 0.4, + "grad_norm": 1.5430934429168701, + "learning_rate": 0.0002127659574468085, + "loss": 0.6804, + "step": 100 + }, + { + "epoch": 0.5, + "grad_norm": 1.6103107929229736, + "learning_rate": 0.00026595744680851064, + "loss": 0.6806, + "step": 125 + }, + { + "epoch": 0.6, + "grad_norm": 2.3773200511932373, + "learning_rate": 0.00031914893617021275, + "loss": 0.6801, + "step": 150 + }, + { + "epoch": 0.7, + "grad_norm": 1.8832203149795532, + "learning_rate": 0.0003723404255319149, + "loss": 0.6791, + "step": 175 + }, + { + "epoch": 0.8, + "grad_norm": 1.3350876569747925, + "learning_rate": 0.0003992081821181128, + "loss": 0.6746, + "step": 200 + }, + { + "epoch": 0.9, + "grad_norm": 1.5796219110488892, + "learning_rate": 0.0003975585615308479, + "loss": 0.6771, + "step": 225 + }, + { + "epoch": 1.0, + "grad_norm": 1.884006381034851, + "learning_rate": 0.00039590894094358297, + "loss": 0.6649, + "step": 250 + }, + { + "epoch": 1.1, + "grad_norm": 3.286440372467041, + "learning_rate": 0.0003942593203563181, + "loss": 0.6388, + "step": 275 + }, + { + "epoch": 1.2, + "grad_norm": 8.244946479797363, + "learning_rate": 0.0003926096997690532, + "loss": 0.6238, + "step": 300 + }, + { + "epoch": 1.3, + "grad_norm": 4.265683650970459, + "learning_rate": 0.00039096007918178817, + "loss": 0.6383, + "step": 325 + }, + { + "epoch": 1.4, + "grad_norm": 4.115826606750488, + "learning_rate": 0.0003893104585945233, + "loss": 0.6194, + "step": 350 + }, + { + "epoch": 1.5, + "grad_norm": 5.694250583648682, + "learning_rate": 0.0003876608380072583, + "loss": 0.6324, + "step": 375 + }, + { + "epoch": 1.6, + "grad_norm": 3.463121175765991, + "learning_rate": 0.00038601121741999343, + "loss": 0.621, + "step": 400 + }, + { + "epoch": 1.7, + "grad_norm": 4.582865238189697, + "learning_rate": 0.0003843615968327285, + "loss": 0.6116, + "step": 425 + }, + { + "epoch": 1.8, + "grad_norm": 11.996281623840332, + "learning_rate": 0.0003827119762454636, + "loss": 0.6393, + "step": 450 + }, + { + "epoch": 1.9, + "grad_norm": 3.0407373905181885, + "learning_rate": 0.00038106235565819863, + "loss": 0.628, + "step": 475 + }, + { + "epoch": 2.0, + "grad_norm": 2.917588233947754, + "learning_rate": 0.0003794127350709337, + "loss": 0.6078, + "step": 500 + }, + { + "epoch": 2.1, + "grad_norm": 4.748379707336426, + "learning_rate": 0.0003777631144836688, + "loss": 0.4899, + "step": 525 + }, + { + "epoch": 2.2, + "grad_norm": 3.8076977729797363, + "learning_rate": 0.00037611349389640383, + "loss": 0.5086, + "step": 550 + }, + { + "epoch": 2.3, + "grad_norm": 5.2440714836120605, + "learning_rate": 0.00037446387330913894, + "loss": 0.5327, + "step": 575 + }, + { + "epoch": 2.4, + "grad_norm": 7.110438346862793, + "learning_rate": 0.000372814252721874, + "loss": 0.5436, + "step": 600 + }, + { + "epoch": 2.5, + "grad_norm": 5.46150541305542, + "learning_rate": 0.00037116463213460903, + "loss": 0.5294, + "step": 625 + }, + { + "epoch": 2.6, + "grad_norm": 5.136163234710693, + "learning_rate": 0.00036951501154734414, + "loss": 0.5245, + "step": 650 + }, + { + "epoch": 2.7, + "grad_norm": 8.735346794128418, + "learning_rate": 0.0003678653909600792, + "loss": 0.5449, + "step": 675 + }, + { + "epoch": 2.8, + "grad_norm": 2.922825574874878, + "learning_rate": 0.0003662157703728143, + "loss": 0.5406, + "step": 700 + }, + { + "epoch": 2.9, + "grad_norm": 7.744819641113281, + "learning_rate": 0.00036456614978554934, + "loss": 0.5447, + "step": 725 + }, + { + "epoch": 3.0, + "grad_norm": 11.6185884475708, + "learning_rate": 0.00036291652919828444, + "loss": 0.5195, + "step": 750 + }, + { + "epoch": 3.1, + "grad_norm": 3.73836088180542, + "learning_rate": 0.00036126690861101944, + "loss": 0.3824, + "step": 775 + }, + { + "epoch": 3.2, + "grad_norm": 14.850343704223633, + "learning_rate": 0.00035961728802375454, + "loss": 0.4071, + "step": 800 + }, + { + "epoch": 3.3, + "grad_norm": 5.7157440185546875, + "learning_rate": 0.0003579676674364896, + "loss": 0.3986, + "step": 825 + }, + { + "epoch": 3.4, + "grad_norm": 12.418399810791016, + "learning_rate": 0.0003563180468492247, + "loss": 0.4282, + "step": 850 + }, + { + "epoch": 3.5, + "grad_norm": 12.793001174926758, + "learning_rate": 0.0003546684262619598, + "loss": 0.4822, + "step": 875 + }, + { + "epoch": 3.6, + "grad_norm": 6.489450931549072, + "learning_rate": 0.00035301880567469485, + "loss": 0.4239, + "step": 900 + }, + { + "epoch": 3.7, + "grad_norm": 5.365822792053223, + "learning_rate": 0.0003513691850874299, + "loss": 0.421, + "step": 925 + }, + { + "epoch": 3.8, + "grad_norm": 12.643745422363281, + "learning_rate": 0.00034971956450016495, + "loss": 0.3964, + "step": 950 + }, + { + "epoch": 3.9, + "grad_norm": 14.334024429321289, + "learning_rate": 0.00034806994391290005, + "loss": 0.4634, + "step": 975 + }, + { + "epoch": 4.0, + "grad_norm": 6.819091320037842, + "learning_rate": 0.0003464203233256351, + "loss": 0.4139, + "step": 1000 + } + ], + "logging_steps": 25, + "max_steps": 6250, + "num_input_tokens_seen": 0, + "num_train_epochs": 25, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 50819481600000.0, + "train_batch_size": 20, + "trial_name": null, + "trial_params": null +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-1000/training_args.bin b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-1000/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..798d8a9f3a9513692035320aba5484c89ac49e69 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-1000/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1cabf5db3e9d79ec0c1945933395ff57418204414cc81f77c90324d003fac782 +size 5304 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-1000/vocab.txt b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-1000/vocab.txt new file mode 100644 index 0000000000000000000000000000000000000000..cc9ad53732e21ea9180f18e1fb67a38cc5702d4c --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-1000/vocab.txt @@ -0,0 +1,4101 @@ +[PAD] +[UNK] +[CLS] +[SEP] +[MASK] +AAAAAA +AAAAAT +AAAAAC +AAAAAG +AAAATA +AAAATT +AAAATC +AAAATG +AAAACA +AAAACT +AAAACC +AAAACG +AAAAGA +AAAAGT +AAAAGC +AAAAGG +AAATAA +AAATAT +AAATAC +AAATAG +AAATTA +AAATTT +AAATTC +AAATTG +AAATCA +AAATCT +AAATCC +AAATCG +AAATGA +AAATGT +AAATGC +AAATGG +AAACAA +AAACAT +AAACAC +AAACAG +AAACTA +AAACTT +AAACTC +AAACTG +AAACCA +AAACCT +AAACCC +AAACCG +AAACGA +AAACGT +AAACGC +AAACGG +AAAGAA +AAAGAT +AAAGAC +AAAGAG +AAAGTA +AAAGTT +AAAGTC +AAAGTG +AAAGCA +AAAGCT +AAAGCC +AAAGCG +AAAGGA +AAAGGT +AAAGGC +AAAGGG +AATAAA +AATAAT +AATAAC +AATAAG +AATATA +AATATT +AATATC +AATATG +AATACA +AATACT +AATACC +AATACG +AATAGA +AATAGT +AATAGC +AATAGG +AATTAA +AATTAT +AATTAC +AATTAG +AATTTA +AATTTT +AATTTC +AATTTG +AATTCA +AATTCT +AATTCC +AATTCG +AATTGA +AATTGT +AATTGC +AATTGG +AATCAA +AATCAT +AATCAC +AATCAG +AATCTA +AATCTT +AATCTC +AATCTG +AATCCA +AATCCT +AATCCC +AATCCG +AATCGA +AATCGT +AATCGC +AATCGG +AATGAA +AATGAT +AATGAC +AATGAG +AATGTA +AATGTT +AATGTC +AATGTG +AATGCA +AATGCT +AATGCC +AATGCG +AATGGA +AATGGT +AATGGC +AATGGG +AACAAA +AACAAT +AACAAC +AACAAG +AACATA +AACATT +AACATC +AACATG +AACACA +AACACT +AACACC +AACACG +AACAGA +AACAGT +AACAGC +AACAGG +AACTAA +AACTAT +AACTAC +AACTAG +AACTTA +AACTTT +AACTTC +AACTTG +AACTCA +AACTCT +AACTCC +AACTCG +AACTGA +AACTGT +AACTGC +AACTGG +AACCAA +AACCAT +AACCAC +AACCAG +AACCTA +AACCTT +AACCTC +AACCTG +AACCCA +AACCCT +AACCCC +AACCCG +AACCGA +AACCGT +AACCGC +AACCGG +AACGAA +AACGAT +AACGAC +AACGAG +AACGTA +AACGTT +AACGTC +AACGTG +AACGCA +AACGCT +AACGCC +AACGCG +AACGGA +AACGGT +AACGGC +AACGGG +AAGAAA +AAGAAT +AAGAAC +AAGAAG +AAGATA +AAGATT +AAGATC +AAGATG +AAGACA +AAGACT +AAGACC +AAGACG +AAGAGA +AAGAGT +AAGAGC +AAGAGG +AAGTAA +AAGTAT +AAGTAC +AAGTAG +AAGTTA +AAGTTT +AAGTTC +AAGTTG +AAGTCA +AAGTCT +AAGTCC +AAGTCG +AAGTGA +AAGTGT +AAGTGC +AAGTGG +AAGCAA +AAGCAT +AAGCAC +AAGCAG +AAGCTA +AAGCTT +AAGCTC +AAGCTG +AAGCCA +AAGCCT +AAGCCC +AAGCCG +AAGCGA +AAGCGT +AAGCGC +AAGCGG +AAGGAA +AAGGAT +AAGGAC +AAGGAG +AAGGTA +AAGGTT +AAGGTC +AAGGTG +AAGGCA +AAGGCT +AAGGCC +AAGGCG +AAGGGA +AAGGGT +AAGGGC +AAGGGG +ATAAAA +ATAAAT +ATAAAC +ATAAAG +ATAATA +ATAATT +ATAATC +ATAATG +ATAACA +ATAACT +ATAACC +ATAACG +ATAAGA +ATAAGT +ATAAGC +ATAAGG +ATATAA +ATATAT +ATATAC +ATATAG +ATATTA +ATATTT +ATATTC +ATATTG +ATATCA +ATATCT +ATATCC +ATATCG +ATATGA +ATATGT +ATATGC +ATATGG +ATACAA +ATACAT +ATACAC +ATACAG +ATACTA +ATACTT +ATACTC +ATACTG +ATACCA +ATACCT +ATACCC +ATACCG +ATACGA +ATACGT +ATACGC +ATACGG +ATAGAA +ATAGAT +ATAGAC +ATAGAG +ATAGTA +ATAGTT +ATAGTC +ATAGTG +ATAGCA +ATAGCT +ATAGCC +ATAGCG +ATAGGA +ATAGGT +ATAGGC +ATAGGG +ATTAAA +ATTAAT +ATTAAC +ATTAAG +ATTATA +ATTATT +ATTATC +ATTATG +ATTACA +ATTACT +ATTACC +ATTACG +ATTAGA +ATTAGT +ATTAGC +ATTAGG +ATTTAA +ATTTAT +ATTTAC +ATTTAG +ATTTTA +ATTTTT +ATTTTC +ATTTTG +ATTTCA +ATTTCT +ATTTCC +ATTTCG +ATTTGA +ATTTGT +ATTTGC +ATTTGG +ATTCAA +ATTCAT +ATTCAC +ATTCAG +ATTCTA +ATTCTT +ATTCTC +ATTCTG +ATTCCA +ATTCCT +ATTCCC +ATTCCG +ATTCGA +ATTCGT +ATTCGC +ATTCGG +ATTGAA +ATTGAT +ATTGAC +ATTGAG +ATTGTA +ATTGTT +ATTGTC +ATTGTG +ATTGCA +ATTGCT +ATTGCC +ATTGCG +ATTGGA +ATTGGT +ATTGGC +ATTGGG +ATCAAA +ATCAAT +ATCAAC +ATCAAG +ATCATA +ATCATT +ATCATC +ATCATG +ATCACA +ATCACT +ATCACC +ATCACG +ATCAGA +ATCAGT +ATCAGC +ATCAGG +ATCTAA +ATCTAT +ATCTAC +ATCTAG +ATCTTA +ATCTTT +ATCTTC +ATCTTG +ATCTCA +ATCTCT +ATCTCC +ATCTCG +ATCTGA +ATCTGT +ATCTGC +ATCTGG +ATCCAA +ATCCAT +ATCCAC +ATCCAG +ATCCTA +ATCCTT +ATCCTC +ATCCTG +ATCCCA +ATCCCT +ATCCCC +ATCCCG +ATCCGA +ATCCGT +ATCCGC +ATCCGG +ATCGAA +ATCGAT +ATCGAC +ATCGAG +ATCGTA +ATCGTT +ATCGTC +ATCGTG +ATCGCA +ATCGCT +ATCGCC +ATCGCG +ATCGGA +ATCGGT +ATCGGC +ATCGGG +ATGAAA +ATGAAT +ATGAAC +ATGAAG +ATGATA +ATGATT +ATGATC +ATGATG +ATGACA +ATGACT +ATGACC +ATGACG +ATGAGA +ATGAGT +ATGAGC +ATGAGG +ATGTAA +ATGTAT +ATGTAC +ATGTAG +ATGTTA +ATGTTT +ATGTTC +ATGTTG +ATGTCA +ATGTCT +ATGTCC +ATGTCG +ATGTGA +ATGTGT +ATGTGC +ATGTGG +ATGCAA +ATGCAT +ATGCAC +ATGCAG +ATGCTA +ATGCTT +ATGCTC +ATGCTG +ATGCCA +ATGCCT +ATGCCC +ATGCCG +ATGCGA +ATGCGT +ATGCGC +ATGCGG +ATGGAA +ATGGAT +ATGGAC +ATGGAG +ATGGTA +ATGGTT +ATGGTC +ATGGTG +ATGGCA +ATGGCT +ATGGCC +ATGGCG +ATGGGA +ATGGGT +ATGGGC +ATGGGG +ACAAAA +ACAAAT +ACAAAC +ACAAAG +ACAATA +ACAATT +ACAATC +ACAATG +ACAACA +ACAACT +ACAACC +ACAACG +ACAAGA +ACAAGT +ACAAGC +ACAAGG +ACATAA +ACATAT +ACATAC +ACATAG +ACATTA +ACATTT +ACATTC +ACATTG +ACATCA +ACATCT +ACATCC +ACATCG +ACATGA +ACATGT +ACATGC +ACATGG +ACACAA +ACACAT +ACACAC +ACACAG +ACACTA +ACACTT +ACACTC +ACACTG +ACACCA +ACACCT +ACACCC +ACACCG +ACACGA +ACACGT +ACACGC +ACACGG +ACAGAA +ACAGAT +ACAGAC +ACAGAG +ACAGTA +ACAGTT +ACAGTC +ACAGTG +ACAGCA +ACAGCT +ACAGCC +ACAGCG +ACAGGA +ACAGGT +ACAGGC +ACAGGG +ACTAAA +ACTAAT +ACTAAC +ACTAAG +ACTATA +ACTATT +ACTATC +ACTATG +ACTACA +ACTACT +ACTACC +ACTACG +ACTAGA +ACTAGT +ACTAGC +ACTAGG +ACTTAA +ACTTAT +ACTTAC +ACTTAG +ACTTTA +ACTTTT +ACTTTC +ACTTTG +ACTTCA +ACTTCT +ACTTCC +ACTTCG +ACTTGA +ACTTGT +ACTTGC +ACTTGG +ACTCAA +ACTCAT +ACTCAC +ACTCAG +ACTCTA +ACTCTT +ACTCTC +ACTCTG +ACTCCA +ACTCCT +ACTCCC +ACTCCG +ACTCGA +ACTCGT +ACTCGC +ACTCGG +ACTGAA +ACTGAT +ACTGAC +ACTGAG +ACTGTA +ACTGTT +ACTGTC +ACTGTG +ACTGCA +ACTGCT +ACTGCC +ACTGCG +ACTGGA +ACTGGT +ACTGGC +ACTGGG +ACCAAA +ACCAAT +ACCAAC +ACCAAG +ACCATA +ACCATT +ACCATC +ACCATG +ACCACA +ACCACT +ACCACC +ACCACG +ACCAGA +ACCAGT +ACCAGC +ACCAGG +ACCTAA +ACCTAT +ACCTAC +ACCTAG +ACCTTA +ACCTTT +ACCTTC +ACCTTG +ACCTCA +ACCTCT +ACCTCC +ACCTCG +ACCTGA +ACCTGT +ACCTGC +ACCTGG +ACCCAA +ACCCAT +ACCCAC +ACCCAG +ACCCTA +ACCCTT +ACCCTC +ACCCTG +ACCCCA +ACCCCT +ACCCCC +ACCCCG +ACCCGA +ACCCGT +ACCCGC +ACCCGG +ACCGAA +ACCGAT +ACCGAC +ACCGAG +ACCGTA +ACCGTT +ACCGTC +ACCGTG +ACCGCA +ACCGCT +ACCGCC +ACCGCG +ACCGGA +ACCGGT +ACCGGC +ACCGGG +ACGAAA +ACGAAT +ACGAAC +ACGAAG +ACGATA +ACGATT +ACGATC +ACGATG +ACGACA +ACGACT +ACGACC +ACGACG +ACGAGA +ACGAGT +ACGAGC +ACGAGG +ACGTAA +ACGTAT +ACGTAC +ACGTAG +ACGTTA +ACGTTT +ACGTTC +ACGTTG +ACGTCA +ACGTCT +ACGTCC +ACGTCG +ACGTGA +ACGTGT +ACGTGC +ACGTGG +ACGCAA +ACGCAT +ACGCAC +ACGCAG +ACGCTA +ACGCTT +ACGCTC +ACGCTG +ACGCCA +ACGCCT +ACGCCC +ACGCCG +ACGCGA +ACGCGT +ACGCGC +ACGCGG +ACGGAA +ACGGAT +ACGGAC +ACGGAG +ACGGTA +ACGGTT +ACGGTC +ACGGTG +ACGGCA +ACGGCT +ACGGCC +ACGGCG +ACGGGA +ACGGGT +ACGGGC +ACGGGG +AGAAAA +AGAAAT +AGAAAC +AGAAAG +AGAATA +AGAATT +AGAATC +AGAATG +AGAACA +AGAACT +AGAACC +AGAACG +AGAAGA +AGAAGT +AGAAGC +AGAAGG +AGATAA +AGATAT +AGATAC +AGATAG +AGATTA +AGATTT +AGATTC +AGATTG +AGATCA +AGATCT +AGATCC +AGATCG +AGATGA +AGATGT +AGATGC +AGATGG +AGACAA +AGACAT +AGACAC +AGACAG +AGACTA +AGACTT +AGACTC +AGACTG +AGACCA +AGACCT +AGACCC +AGACCG +AGACGA +AGACGT +AGACGC +AGACGG +AGAGAA +AGAGAT +AGAGAC +AGAGAG +AGAGTA +AGAGTT +AGAGTC +AGAGTG +AGAGCA +AGAGCT +AGAGCC +AGAGCG +AGAGGA +AGAGGT +AGAGGC +AGAGGG +AGTAAA +AGTAAT +AGTAAC +AGTAAG +AGTATA +AGTATT +AGTATC +AGTATG +AGTACA +AGTACT +AGTACC +AGTACG +AGTAGA +AGTAGT +AGTAGC +AGTAGG +AGTTAA +AGTTAT +AGTTAC +AGTTAG +AGTTTA +AGTTTT +AGTTTC +AGTTTG +AGTTCA +AGTTCT +AGTTCC +AGTTCG +AGTTGA +AGTTGT +AGTTGC +AGTTGG +AGTCAA +AGTCAT +AGTCAC +AGTCAG +AGTCTA +AGTCTT +AGTCTC +AGTCTG +AGTCCA +AGTCCT +AGTCCC +AGTCCG +AGTCGA +AGTCGT +AGTCGC +AGTCGG +AGTGAA +AGTGAT +AGTGAC +AGTGAG +AGTGTA +AGTGTT +AGTGTC +AGTGTG +AGTGCA +AGTGCT +AGTGCC +AGTGCG +AGTGGA +AGTGGT +AGTGGC +AGTGGG +AGCAAA +AGCAAT +AGCAAC +AGCAAG +AGCATA +AGCATT +AGCATC +AGCATG +AGCACA +AGCACT +AGCACC +AGCACG +AGCAGA +AGCAGT +AGCAGC +AGCAGG +AGCTAA +AGCTAT +AGCTAC +AGCTAG +AGCTTA +AGCTTT +AGCTTC +AGCTTG +AGCTCA +AGCTCT +AGCTCC +AGCTCG +AGCTGA +AGCTGT +AGCTGC +AGCTGG +AGCCAA +AGCCAT +AGCCAC +AGCCAG +AGCCTA +AGCCTT +AGCCTC +AGCCTG +AGCCCA +AGCCCT +AGCCCC +AGCCCG +AGCCGA +AGCCGT +AGCCGC +AGCCGG +AGCGAA +AGCGAT +AGCGAC +AGCGAG +AGCGTA +AGCGTT +AGCGTC +AGCGTG +AGCGCA +AGCGCT +AGCGCC +AGCGCG +AGCGGA +AGCGGT +AGCGGC +AGCGGG +AGGAAA +AGGAAT +AGGAAC +AGGAAG +AGGATA +AGGATT +AGGATC +AGGATG +AGGACA +AGGACT +AGGACC +AGGACG +AGGAGA +AGGAGT +AGGAGC +AGGAGG +AGGTAA +AGGTAT +AGGTAC +AGGTAG +AGGTTA +AGGTTT +AGGTTC +AGGTTG +AGGTCA +AGGTCT +AGGTCC +AGGTCG +AGGTGA +AGGTGT +AGGTGC +AGGTGG +AGGCAA +AGGCAT +AGGCAC +AGGCAG +AGGCTA +AGGCTT +AGGCTC +AGGCTG +AGGCCA +AGGCCT +AGGCCC +AGGCCG +AGGCGA +AGGCGT +AGGCGC +AGGCGG +AGGGAA +AGGGAT +AGGGAC +AGGGAG +AGGGTA +AGGGTT +AGGGTC +AGGGTG +AGGGCA +AGGGCT +AGGGCC +AGGGCG +AGGGGA +AGGGGT +AGGGGC +AGGGGG +TAAAAA +TAAAAT +TAAAAC +TAAAAG +TAAATA +TAAATT +TAAATC +TAAATG +TAAACA +TAAACT +TAAACC +TAAACG +TAAAGA +TAAAGT +TAAAGC +TAAAGG +TAATAA +TAATAT +TAATAC +TAATAG +TAATTA +TAATTT +TAATTC +TAATTG +TAATCA +TAATCT +TAATCC +TAATCG +TAATGA +TAATGT +TAATGC +TAATGG +TAACAA +TAACAT +TAACAC +TAACAG +TAACTA +TAACTT +TAACTC +TAACTG +TAACCA +TAACCT +TAACCC +TAACCG +TAACGA +TAACGT +TAACGC +TAACGG +TAAGAA +TAAGAT +TAAGAC +TAAGAG +TAAGTA +TAAGTT +TAAGTC +TAAGTG +TAAGCA +TAAGCT +TAAGCC +TAAGCG +TAAGGA +TAAGGT +TAAGGC +TAAGGG +TATAAA +TATAAT +TATAAC +TATAAG +TATATA +TATATT +TATATC +TATATG +TATACA +TATACT +TATACC +TATACG +TATAGA +TATAGT +TATAGC +TATAGG +TATTAA +TATTAT +TATTAC +TATTAG +TATTTA +TATTTT +TATTTC +TATTTG +TATTCA +TATTCT +TATTCC +TATTCG +TATTGA +TATTGT +TATTGC +TATTGG +TATCAA +TATCAT +TATCAC +TATCAG +TATCTA +TATCTT +TATCTC +TATCTG +TATCCA +TATCCT +TATCCC +TATCCG +TATCGA +TATCGT +TATCGC +TATCGG +TATGAA +TATGAT +TATGAC +TATGAG +TATGTA +TATGTT +TATGTC +TATGTG +TATGCA +TATGCT +TATGCC +TATGCG +TATGGA +TATGGT +TATGGC +TATGGG +TACAAA +TACAAT +TACAAC +TACAAG +TACATA +TACATT +TACATC +TACATG +TACACA +TACACT +TACACC +TACACG +TACAGA +TACAGT +TACAGC +TACAGG +TACTAA +TACTAT +TACTAC +TACTAG +TACTTA +TACTTT +TACTTC +TACTTG +TACTCA +TACTCT +TACTCC +TACTCG +TACTGA +TACTGT +TACTGC +TACTGG +TACCAA +TACCAT +TACCAC +TACCAG +TACCTA +TACCTT +TACCTC +TACCTG +TACCCA +TACCCT +TACCCC +TACCCG +TACCGA +TACCGT +TACCGC +TACCGG +TACGAA +TACGAT +TACGAC +TACGAG +TACGTA +TACGTT +TACGTC +TACGTG +TACGCA +TACGCT +TACGCC +TACGCG +TACGGA +TACGGT +TACGGC +TACGGG +TAGAAA +TAGAAT +TAGAAC +TAGAAG +TAGATA +TAGATT +TAGATC +TAGATG +TAGACA +TAGACT +TAGACC +TAGACG +TAGAGA +TAGAGT +TAGAGC +TAGAGG +TAGTAA +TAGTAT +TAGTAC +TAGTAG +TAGTTA +TAGTTT +TAGTTC +TAGTTG +TAGTCA +TAGTCT +TAGTCC +TAGTCG +TAGTGA +TAGTGT +TAGTGC +TAGTGG +TAGCAA +TAGCAT +TAGCAC +TAGCAG +TAGCTA +TAGCTT +TAGCTC +TAGCTG +TAGCCA +TAGCCT +TAGCCC +TAGCCG +TAGCGA +TAGCGT +TAGCGC +TAGCGG +TAGGAA +TAGGAT +TAGGAC +TAGGAG +TAGGTA +TAGGTT +TAGGTC +TAGGTG +TAGGCA +TAGGCT +TAGGCC +TAGGCG +TAGGGA +TAGGGT +TAGGGC +TAGGGG +TTAAAA +TTAAAT +TTAAAC +TTAAAG +TTAATA +TTAATT +TTAATC +TTAATG +TTAACA +TTAACT +TTAACC +TTAACG +TTAAGA +TTAAGT +TTAAGC +TTAAGG +TTATAA +TTATAT +TTATAC +TTATAG +TTATTA +TTATTT +TTATTC +TTATTG +TTATCA +TTATCT +TTATCC +TTATCG +TTATGA +TTATGT +TTATGC +TTATGG +TTACAA +TTACAT +TTACAC +TTACAG +TTACTA +TTACTT +TTACTC +TTACTG +TTACCA +TTACCT +TTACCC +TTACCG +TTACGA +TTACGT +TTACGC +TTACGG +TTAGAA +TTAGAT +TTAGAC +TTAGAG +TTAGTA +TTAGTT +TTAGTC +TTAGTG +TTAGCA +TTAGCT +TTAGCC +TTAGCG +TTAGGA +TTAGGT +TTAGGC +TTAGGG +TTTAAA +TTTAAT +TTTAAC +TTTAAG +TTTATA +TTTATT +TTTATC +TTTATG +TTTACA +TTTACT +TTTACC +TTTACG +TTTAGA +TTTAGT +TTTAGC +TTTAGG +TTTTAA +TTTTAT +TTTTAC +TTTTAG +TTTTTA +TTTTTT +TTTTTC +TTTTTG +TTTTCA +TTTTCT +TTTTCC +TTTTCG +TTTTGA +TTTTGT +TTTTGC +TTTTGG +TTTCAA +TTTCAT +TTTCAC +TTTCAG +TTTCTA +TTTCTT +TTTCTC +TTTCTG +TTTCCA +TTTCCT +TTTCCC +TTTCCG +TTTCGA +TTTCGT +TTTCGC +TTTCGG +TTTGAA +TTTGAT +TTTGAC +TTTGAG +TTTGTA +TTTGTT +TTTGTC +TTTGTG +TTTGCA +TTTGCT +TTTGCC +TTTGCG +TTTGGA +TTTGGT +TTTGGC +TTTGGG +TTCAAA +TTCAAT +TTCAAC +TTCAAG +TTCATA +TTCATT +TTCATC +TTCATG +TTCACA +TTCACT +TTCACC +TTCACG +TTCAGA +TTCAGT +TTCAGC +TTCAGG +TTCTAA +TTCTAT +TTCTAC +TTCTAG +TTCTTA +TTCTTT +TTCTTC +TTCTTG +TTCTCA +TTCTCT +TTCTCC +TTCTCG +TTCTGA +TTCTGT +TTCTGC +TTCTGG +TTCCAA +TTCCAT +TTCCAC +TTCCAG +TTCCTA +TTCCTT +TTCCTC +TTCCTG +TTCCCA +TTCCCT +TTCCCC +TTCCCG +TTCCGA +TTCCGT +TTCCGC +TTCCGG +TTCGAA +TTCGAT +TTCGAC +TTCGAG +TTCGTA +TTCGTT +TTCGTC +TTCGTG +TTCGCA +TTCGCT +TTCGCC +TTCGCG +TTCGGA +TTCGGT +TTCGGC +TTCGGG +TTGAAA +TTGAAT +TTGAAC +TTGAAG +TTGATA +TTGATT +TTGATC +TTGATG +TTGACA +TTGACT +TTGACC +TTGACG +TTGAGA +TTGAGT +TTGAGC +TTGAGG +TTGTAA +TTGTAT +TTGTAC +TTGTAG +TTGTTA +TTGTTT +TTGTTC +TTGTTG +TTGTCA +TTGTCT +TTGTCC +TTGTCG +TTGTGA +TTGTGT +TTGTGC +TTGTGG +TTGCAA +TTGCAT +TTGCAC +TTGCAG +TTGCTA +TTGCTT +TTGCTC +TTGCTG +TTGCCA +TTGCCT +TTGCCC +TTGCCG +TTGCGA +TTGCGT +TTGCGC +TTGCGG +TTGGAA +TTGGAT +TTGGAC +TTGGAG +TTGGTA +TTGGTT +TTGGTC +TTGGTG +TTGGCA +TTGGCT +TTGGCC +TTGGCG +TTGGGA +TTGGGT +TTGGGC +TTGGGG +TCAAAA +TCAAAT +TCAAAC +TCAAAG +TCAATA +TCAATT +TCAATC +TCAATG +TCAACA +TCAACT +TCAACC +TCAACG +TCAAGA +TCAAGT +TCAAGC +TCAAGG +TCATAA +TCATAT +TCATAC +TCATAG +TCATTA +TCATTT +TCATTC +TCATTG +TCATCA +TCATCT +TCATCC +TCATCG +TCATGA +TCATGT +TCATGC +TCATGG +TCACAA +TCACAT +TCACAC +TCACAG +TCACTA +TCACTT +TCACTC +TCACTG +TCACCA +TCACCT +TCACCC +TCACCG +TCACGA +TCACGT +TCACGC +TCACGG +TCAGAA +TCAGAT +TCAGAC +TCAGAG +TCAGTA +TCAGTT +TCAGTC +TCAGTG +TCAGCA +TCAGCT +TCAGCC +TCAGCG +TCAGGA +TCAGGT +TCAGGC +TCAGGG +TCTAAA +TCTAAT +TCTAAC +TCTAAG +TCTATA +TCTATT +TCTATC +TCTATG +TCTACA +TCTACT +TCTACC +TCTACG +TCTAGA +TCTAGT +TCTAGC +TCTAGG +TCTTAA +TCTTAT +TCTTAC +TCTTAG +TCTTTA +TCTTTT +TCTTTC +TCTTTG +TCTTCA +TCTTCT +TCTTCC +TCTTCG +TCTTGA +TCTTGT +TCTTGC +TCTTGG +TCTCAA +TCTCAT +TCTCAC +TCTCAG +TCTCTA +TCTCTT +TCTCTC +TCTCTG +TCTCCA +TCTCCT +TCTCCC +TCTCCG +TCTCGA +TCTCGT +TCTCGC +TCTCGG +TCTGAA +TCTGAT +TCTGAC +TCTGAG +TCTGTA +TCTGTT +TCTGTC +TCTGTG +TCTGCA +TCTGCT +TCTGCC +TCTGCG +TCTGGA +TCTGGT +TCTGGC +TCTGGG +TCCAAA +TCCAAT +TCCAAC +TCCAAG +TCCATA +TCCATT +TCCATC +TCCATG +TCCACA +TCCACT +TCCACC +TCCACG +TCCAGA +TCCAGT +TCCAGC +TCCAGG +TCCTAA +TCCTAT +TCCTAC +TCCTAG +TCCTTA +TCCTTT +TCCTTC +TCCTTG +TCCTCA +TCCTCT +TCCTCC +TCCTCG +TCCTGA +TCCTGT +TCCTGC +TCCTGG +TCCCAA +TCCCAT +TCCCAC +TCCCAG +TCCCTA +TCCCTT +TCCCTC +TCCCTG +TCCCCA +TCCCCT +TCCCCC +TCCCCG +TCCCGA +TCCCGT +TCCCGC +TCCCGG +TCCGAA +TCCGAT +TCCGAC +TCCGAG +TCCGTA +TCCGTT +TCCGTC +TCCGTG +TCCGCA +TCCGCT +TCCGCC +TCCGCG +TCCGGA +TCCGGT +TCCGGC +TCCGGG +TCGAAA +TCGAAT +TCGAAC +TCGAAG +TCGATA +TCGATT +TCGATC +TCGATG +TCGACA +TCGACT +TCGACC +TCGACG +TCGAGA +TCGAGT +TCGAGC +TCGAGG +TCGTAA +TCGTAT +TCGTAC +TCGTAG +TCGTTA +TCGTTT +TCGTTC +TCGTTG +TCGTCA +TCGTCT +TCGTCC +TCGTCG +TCGTGA +TCGTGT +TCGTGC +TCGTGG +TCGCAA +TCGCAT +TCGCAC +TCGCAG +TCGCTA +TCGCTT +TCGCTC +TCGCTG +TCGCCA +TCGCCT +TCGCCC +TCGCCG +TCGCGA +TCGCGT +TCGCGC +TCGCGG +TCGGAA +TCGGAT +TCGGAC +TCGGAG +TCGGTA +TCGGTT +TCGGTC +TCGGTG +TCGGCA +TCGGCT +TCGGCC +TCGGCG +TCGGGA +TCGGGT +TCGGGC +TCGGGG +TGAAAA +TGAAAT +TGAAAC +TGAAAG +TGAATA +TGAATT +TGAATC +TGAATG +TGAACA +TGAACT +TGAACC +TGAACG +TGAAGA +TGAAGT +TGAAGC +TGAAGG +TGATAA +TGATAT +TGATAC +TGATAG +TGATTA +TGATTT +TGATTC +TGATTG +TGATCA +TGATCT +TGATCC +TGATCG +TGATGA +TGATGT +TGATGC +TGATGG +TGACAA +TGACAT +TGACAC +TGACAG +TGACTA +TGACTT +TGACTC +TGACTG +TGACCA +TGACCT +TGACCC +TGACCG +TGACGA +TGACGT +TGACGC +TGACGG +TGAGAA +TGAGAT +TGAGAC +TGAGAG +TGAGTA +TGAGTT +TGAGTC +TGAGTG +TGAGCA +TGAGCT +TGAGCC +TGAGCG +TGAGGA +TGAGGT +TGAGGC +TGAGGG +TGTAAA +TGTAAT +TGTAAC +TGTAAG +TGTATA +TGTATT +TGTATC +TGTATG +TGTACA +TGTACT +TGTACC +TGTACG +TGTAGA +TGTAGT +TGTAGC +TGTAGG +TGTTAA +TGTTAT +TGTTAC +TGTTAG +TGTTTA +TGTTTT +TGTTTC +TGTTTG +TGTTCA +TGTTCT +TGTTCC +TGTTCG +TGTTGA +TGTTGT +TGTTGC +TGTTGG +TGTCAA +TGTCAT +TGTCAC +TGTCAG +TGTCTA +TGTCTT +TGTCTC +TGTCTG +TGTCCA +TGTCCT +TGTCCC +TGTCCG +TGTCGA +TGTCGT +TGTCGC +TGTCGG +TGTGAA +TGTGAT +TGTGAC +TGTGAG +TGTGTA +TGTGTT +TGTGTC +TGTGTG +TGTGCA +TGTGCT +TGTGCC +TGTGCG +TGTGGA +TGTGGT +TGTGGC +TGTGGG +TGCAAA +TGCAAT +TGCAAC +TGCAAG +TGCATA +TGCATT +TGCATC +TGCATG +TGCACA +TGCACT +TGCACC +TGCACG +TGCAGA +TGCAGT +TGCAGC +TGCAGG +TGCTAA +TGCTAT +TGCTAC +TGCTAG +TGCTTA +TGCTTT +TGCTTC +TGCTTG +TGCTCA +TGCTCT +TGCTCC +TGCTCG +TGCTGA +TGCTGT +TGCTGC +TGCTGG +TGCCAA +TGCCAT +TGCCAC +TGCCAG +TGCCTA +TGCCTT +TGCCTC +TGCCTG +TGCCCA +TGCCCT +TGCCCC +TGCCCG +TGCCGA +TGCCGT +TGCCGC +TGCCGG +TGCGAA +TGCGAT +TGCGAC +TGCGAG +TGCGTA +TGCGTT +TGCGTC +TGCGTG +TGCGCA +TGCGCT +TGCGCC +TGCGCG +TGCGGA +TGCGGT +TGCGGC +TGCGGG +TGGAAA +TGGAAT +TGGAAC +TGGAAG +TGGATA +TGGATT +TGGATC +TGGATG +TGGACA +TGGACT +TGGACC +TGGACG +TGGAGA +TGGAGT +TGGAGC +TGGAGG +TGGTAA +TGGTAT +TGGTAC +TGGTAG +TGGTTA +TGGTTT +TGGTTC +TGGTTG +TGGTCA +TGGTCT +TGGTCC +TGGTCG +TGGTGA +TGGTGT +TGGTGC +TGGTGG +TGGCAA +TGGCAT +TGGCAC +TGGCAG +TGGCTA +TGGCTT +TGGCTC +TGGCTG +TGGCCA +TGGCCT +TGGCCC +TGGCCG +TGGCGA +TGGCGT +TGGCGC +TGGCGG +TGGGAA +TGGGAT +TGGGAC +TGGGAG +TGGGTA +TGGGTT +TGGGTC +TGGGTG +TGGGCA +TGGGCT +TGGGCC +TGGGCG +TGGGGA +TGGGGT +TGGGGC +TGGGGG +CAAAAA +CAAAAT +CAAAAC +CAAAAG +CAAATA +CAAATT +CAAATC +CAAATG +CAAACA +CAAACT +CAAACC +CAAACG +CAAAGA +CAAAGT +CAAAGC +CAAAGG +CAATAA +CAATAT +CAATAC +CAATAG +CAATTA +CAATTT +CAATTC +CAATTG +CAATCA +CAATCT +CAATCC +CAATCG +CAATGA +CAATGT +CAATGC +CAATGG +CAACAA +CAACAT +CAACAC +CAACAG +CAACTA +CAACTT +CAACTC +CAACTG +CAACCA +CAACCT +CAACCC +CAACCG +CAACGA +CAACGT +CAACGC +CAACGG +CAAGAA +CAAGAT +CAAGAC +CAAGAG +CAAGTA +CAAGTT +CAAGTC +CAAGTG +CAAGCA +CAAGCT +CAAGCC +CAAGCG +CAAGGA +CAAGGT +CAAGGC +CAAGGG +CATAAA +CATAAT +CATAAC +CATAAG +CATATA +CATATT +CATATC +CATATG +CATACA +CATACT +CATACC +CATACG +CATAGA +CATAGT +CATAGC +CATAGG +CATTAA +CATTAT +CATTAC +CATTAG +CATTTA +CATTTT +CATTTC +CATTTG +CATTCA +CATTCT +CATTCC +CATTCG +CATTGA +CATTGT +CATTGC +CATTGG +CATCAA +CATCAT +CATCAC +CATCAG +CATCTA +CATCTT +CATCTC +CATCTG +CATCCA +CATCCT +CATCCC +CATCCG +CATCGA +CATCGT +CATCGC +CATCGG +CATGAA +CATGAT +CATGAC +CATGAG +CATGTA +CATGTT +CATGTC +CATGTG +CATGCA +CATGCT +CATGCC +CATGCG +CATGGA +CATGGT +CATGGC +CATGGG +CACAAA +CACAAT +CACAAC +CACAAG +CACATA +CACATT +CACATC +CACATG +CACACA +CACACT +CACACC +CACACG +CACAGA +CACAGT +CACAGC +CACAGG +CACTAA +CACTAT +CACTAC +CACTAG +CACTTA +CACTTT +CACTTC +CACTTG +CACTCA +CACTCT +CACTCC +CACTCG +CACTGA +CACTGT +CACTGC +CACTGG +CACCAA +CACCAT +CACCAC +CACCAG +CACCTA +CACCTT +CACCTC +CACCTG +CACCCA +CACCCT +CACCCC +CACCCG +CACCGA +CACCGT +CACCGC +CACCGG +CACGAA +CACGAT +CACGAC +CACGAG +CACGTA +CACGTT +CACGTC +CACGTG +CACGCA +CACGCT +CACGCC +CACGCG +CACGGA +CACGGT +CACGGC +CACGGG +CAGAAA +CAGAAT +CAGAAC +CAGAAG +CAGATA +CAGATT +CAGATC +CAGATG +CAGACA +CAGACT +CAGACC +CAGACG +CAGAGA +CAGAGT +CAGAGC +CAGAGG +CAGTAA +CAGTAT +CAGTAC +CAGTAG +CAGTTA +CAGTTT +CAGTTC +CAGTTG +CAGTCA +CAGTCT +CAGTCC +CAGTCG +CAGTGA +CAGTGT +CAGTGC +CAGTGG +CAGCAA +CAGCAT +CAGCAC +CAGCAG +CAGCTA +CAGCTT +CAGCTC +CAGCTG +CAGCCA +CAGCCT +CAGCCC +CAGCCG +CAGCGA +CAGCGT +CAGCGC +CAGCGG +CAGGAA +CAGGAT +CAGGAC +CAGGAG +CAGGTA +CAGGTT +CAGGTC +CAGGTG +CAGGCA +CAGGCT +CAGGCC +CAGGCG +CAGGGA +CAGGGT +CAGGGC +CAGGGG +CTAAAA +CTAAAT +CTAAAC +CTAAAG +CTAATA +CTAATT +CTAATC +CTAATG +CTAACA +CTAACT +CTAACC +CTAACG +CTAAGA +CTAAGT +CTAAGC +CTAAGG +CTATAA +CTATAT +CTATAC +CTATAG +CTATTA +CTATTT +CTATTC +CTATTG +CTATCA +CTATCT +CTATCC +CTATCG +CTATGA +CTATGT +CTATGC +CTATGG +CTACAA +CTACAT +CTACAC +CTACAG +CTACTA +CTACTT +CTACTC +CTACTG +CTACCA +CTACCT +CTACCC +CTACCG +CTACGA +CTACGT +CTACGC +CTACGG +CTAGAA +CTAGAT +CTAGAC +CTAGAG +CTAGTA +CTAGTT +CTAGTC +CTAGTG +CTAGCA +CTAGCT +CTAGCC +CTAGCG +CTAGGA +CTAGGT +CTAGGC +CTAGGG +CTTAAA +CTTAAT +CTTAAC +CTTAAG +CTTATA +CTTATT +CTTATC +CTTATG +CTTACA +CTTACT +CTTACC +CTTACG +CTTAGA +CTTAGT +CTTAGC +CTTAGG +CTTTAA +CTTTAT +CTTTAC +CTTTAG +CTTTTA +CTTTTT +CTTTTC +CTTTTG +CTTTCA +CTTTCT +CTTTCC +CTTTCG +CTTTGA +CTTTGT +CTTTGC +CTTTGG +CTTCAA +CTTCAT +CTTCAC +CTTCAG +CTTCTA +CTTCTT +CTTCTC +CTTCTG +CTTCCA +CTTCCT +CTTCCC +CTTCCG +CTTCGA +CTTCGT +CTTCGC +CTTCGG +CTTGAA +CTTGAT +CTTGAC +CTTGAG +CTTGTA +CTTGTT +CTTGTC +CTTGTG +CTTGCA +CTTGCT +CTTGCC +CTTGCG +CTTGGA +CTTGGT +CTTGGC +CTTGGG +CTCAAA +CTCAAT +CTCAAC +CTCAAG +CTCATA +CTCATT +CTCATC +CTCATG +CTCACA +CTCACT +CTCACC +CTCACG +CTCAGA +CTCAGT +CTCAGC +CTCAGG +CTCTAA +CTCTAT +CTCTAC +CTCTAG +CTCTTA +CTCTTT +CTCTTC +CTCTTG +CTCTCA +CTCTCT +CTCTCC +CTCTCG +CTCTGA +CTCTGT +CTCTGC +CTCTGG +CTCCAA +CTCCAT +CTCCAC +CTCCAG +CTCCTA +CTCCTT +CTCCTC +CTCCTG +CTCCCA +CTCCCT +CTCCCC +CTCCCG +CTCCGA +CTCCGT +CTCCGC +CTCCGG +CTCGAA +CTCGAT +CTCGAC +CTCGAG +CTCGTA +CTCGTT +CTCGTC +CTCGTG +CTCGCA +CTCGCT +CTCGCC +CTCGCG +CTCGGA +CTCGGT +CTCGGC +CTCGGG +CTGAAA +CTGAAT +CTGAAC +CTGAAG +CTGATA +CTGATT +CTGATC +CTGATG +CTGACA +CTGACT +CTGACC +CTGACG +CTGAGA +CTGAGT +CTGAGC +CTGAGG +CTGTAA +CTGTAT +CTGTAC +CTGTAG +CTGTTA +CTGTTT +CTGTTC +CTGTTG +CTGTCA +CTGTCT +CTGTCC +CTGTCG +CTGTGA +CTGTGT +CTGTGC +CTGTGG +CTGCAA +CTGCAT +CTGCAC +CTGCAG +CTGCTA +CTGCTT +CTGCTC +CTGCTG +CTGCCA +CTGCCT +CTGCCC +CTGCCG +CTGCGA +CTGCGT +CTGCGC +CTGCGG +CTGGAA +CTGGAT +CTGGAC +CTGGAG +CTGGTA +CTGGTT +CTGGTC +CTGGTG +CTGGCA +CTGGCT +CTGGCC +CTGGCG +CTGGGA +CTGGGT +CTGGGC +CTGGGG +CCAAAA +CCAAAT +CCAAAC +CCAAAG +CCAATA +CCAATT +CCAATC +CCAATG +CCAACA +CCAACT +CCAACC +CCAACG +CCAAGA +CCAAGT +CCAAGC +CCAAGG +CCATAA +CCATAT +CCATAC +CCATAG +CCATTA +CCATTT +CCATTC +CCATTG +CCATCA +CCATCT +CCATCC +CCATCG +CCATGA +CCATGT +CCATGC +CCATGG +CCACAA +CCACAT +CCACAC +CCACAG +CCACTA +CCACTT +CCACTC +CCACTG +CCACCA +CCACCT +CCACCC +CCACCG +CCACGA +CCACGT +CCACGC +CCACGG +CCAGAA +CCAGAT +CCAGAC +CCAGAG +CCAGTA +CCAGTT +CCAGTC +CCAGTG +CCAGCA +CCAGCT +CCAGCC +CCAGCG +CCAGGA +CCAGGT +CCAGGC +CCAGGG +CCTAAA +CCTAAT +CCTAAC +CCTAAG +CCTATA +CCTATT +CCTATC +CCTATG +CCTACA +CCTACT +CCTACC +CCTACG +CCTAGA +CCTAGT +CCTAGC +CCTAGG +CCTTAA +CCTTAT +CCTTAC +CCTTAG +CCTTTA +CCTTTT +CCTTTC +CCTTTG +CCTTCA +CCTTCT +CCTTCC +CCTTCG +CCTTGA +CCTTGT +CCTTGC +CCTTGG +CCTCAA +CCTCAT +CCTCAC +CCTCAG +CCTCTA +CCTCTT +CCTCTC +CCTCTG +CCTCCA +CCTCCT +CCTCCC +CCTCCG +CCTCGA +CCTCGT +CCTCGC +CCTCGG +CCTGAA +CCTGAT +CCTGAC +CCTGAG +CCTGTA +CCTGTT +CCTGTC +CCTGTG +CCTGCA +CCTGCT +CCTGCC +CCTGCG +CCTGGA +CCTGGT +CCTGGC +CCTGGG +CCCAAA +CCCAAT +CCCAAC +CCCAAG +CCCATA +CCCATT +CCCATC +CCCATG +CCCACA +CCCACT +CCCACC +CCCACG +CCCAGA +CCCAGT +CCCAGC +CCCAGG +CCCTAA +CCCTAT +CCCTAC +CCCTAG +CCCTTA +CCCTTT +CCCTTC +CCCTTG +CCCTCA +CCCTCT +CCCTCC +CCCTCG +CCCTGA +CCCTGT +CCCTGC +CCCTGG +CCCCAA +CCCCAT +CCCCAC +CCCCAG +CCCCTA +CCCCTT +CCCCTC +CCCCTG +CCCCCA +CCCCCT +CCCCCC +CCCCCG +CCCCGA +CCCCGT +CCCCGC +CCCCGG +CCCGAA +CCCGAT +CCCGAC +CCCGAG +CCCGTA +CCCGTT +CCCGTC +CCCGTG +CCCGCA +CCCGCT +CCCGCC +CCCGCG +CCCGGA +CCCGGT +CCCGGC +CCCGGG +CCGAAA +CCGAAT +CCGAAC +CCGAAG +CCGATA +CCGATT +CCGATC +CCGATG +CCGACA +CCGACT +CCGACC +CCGACG +CCGAGA +CCGAGT +CCGAGC +CCGAGG +CCGTAA +CCGTAT +CCGTAC +CCGTAG +CCGTTA +CCGTTT +CCGTTC +CCGTTG +CCGTCA +CCGTCT +CCGTCC +CCGTCG +CCGTGA +CCGTGT +CCGTGC +CCGTGG +CCGCAA +CCGCAT +CCGCAC +CCGCAG +CCGCTA +CCGCTT +CCGCTC +CCGCTG +CCGCCA +CCGCCT +CCGCCC +CCGCCG +CCGCGA +CCGCGT +CCGCGC +CCGCGG +CCGGAA +CCGGAT +CCGGAC +CCGGAG +CCGGTA +CCGGTT +CCGGTC +CCGGTG +CCGGCA +CCGGCT +CCGGCC +CCGGCG +CCGGGA +CCGGGT +CCGGGC +CCGGGG +CGAAAA +CGAAAT +CGAAAC +CGAAAG +CGAATA +CGAATT +CGAATC +CGAATG +CGAACA +CGAACT +CGAACC +CGAACG +CGAAGA +CGAAGT +CGAAGC +CGAAGG +CGATAA +CGATAT +CGATAC +CGATAG +CGATTA +CGATTT +CGATTC +CGATTG +CGATCA +CGATCT +CGATCC +CGATCG +CGATGA +CGATGT +CGATGC +CGATGG +CGACAA +CGACAT +CGACAC +CGACAG +CGACTA +CGACTT +CGACTC +CGACTG +CGACCA +CGACCT +CGACCC +CGACCG +CGACGA +CGACGT +CGACGC +CGACGG +CGAGAA +CGAGAT +CGAGAC +CGAGAG +CGAGTA +CGAGTT +CGAGTC +CGAGTG +CGAGCA +CGAGCT +CGAGCC +CGAGCG +CGAGGA +CGAGGT +CGAGGC +CGAGGG +CGTAAA +CGTAAT +CGTAAC +CGTAAG +CGTATA +CGTATT +CGTATC +CGTATG +CGTACA +CGTACT +CGTACC +CGTACG +CGTAGA +CGTAGT +CGTAGC +CGTAGG +CGTTAA +CGTTAT +CGTTAC +CGTTAG +CGTTTA +CGTTTT +CGTTTC +CGTTTG +CGTTCA +CGTTCT +CGTTCC +CGTTCG +CGTTGA +CGTTGT +CGTTGC +CGTTGG +CGTCAA +CGTCAT +CGTCAC +CGTCAG +CGTCTA +CGTCTT +CGTCTC +CGTCTG +CGTCCA +CGTCCT +CGTCCC +CGTCCG +CGTCGA +CGTCGT +CGTCGC +CGTCGG +CGTGAA +CGTGAT +CGTGAC +CGTGAG +CGTGTA +CGTGTT +CGTGTC +CGTGTG +CGTGCA +CGTGCT +CGTGCC +CGTGCG +CGTGGA +CGTGGT +CGTGGC +CGTGGG +CGCAAA +CGCAAT +CGCAAC +CGCAAG +CGCATA +CGCATT +CGCATC +CGCATG +CGCACA +CGCACT +CGCACC +CGCACG +CGCAGA +CGCAGT +CGCAGC +CGCAGG +CGCTAA +CGCTAT +CGCTAC +CGCTAG +CGCTTA +CGCTTT +CGCTTC +CGCTTG +CGCTCA +CGCTCT +CGCTCC +CGCTCG +CGCTGA +CGCTGT +CGCTGC +CGCTGG +CGCCAA +CGCCAT +CGCCAC +CGCCAG +CGCCTA +CGCCTT +CGCCTC +CGCCTG +CGCCCA +CGCCCT +CGCCCC +CGCCCG +CGCCGA +CGCCGT +CGCCGC +CGCCGG +CGCGAA +CGCGAT +CGCGAC +CGCGAG +CGCGTA +CGCGTT +CGCGTC +CGCGTG +CGCGCA +CGCGCT +CGCGCC +CGCGCG +CGCGGA +CGCGGT +CGCGGC +CGCGGG +CGGAAA +CGGAAT +CGGAAC +CGGAAG +CGGATA +CGGATT +CGGATC +CGGATG +CGGACA +CGGACT +CGGACC +CGGACG +CGGAGA +CGGAGT +CGGAGC +CGGAGG +CGGTAA +CGGTAT +CGGTAC +CGGTAG +CGGTTA +CGGTTT +CGGTTC +CGGTTG +CGGTCA +CGGTCT +CGGTCC +CGGTCG +CGGTGA +CGGTGT +CGGTGC +CGGTGG +CGGCAA +CGGCAT +CGGCAC +CGGCAG +CGGCTA +CGGCTT +CGGCTC +CGGCTG +CGGCCA +CGGCCT +CGGCCC +CGGCCG +CGGCGA +CGGCGT +CGGCGC +CGGCGG +CGGGAA +CGGGAT +CGGGAC +CGGGAG +CGGGTA +CGGGTT +CGGGTC +CGGGTG +CGGGCA +CGGGCT +CGGGCC +CGGGCG +CGGGGA +CGGGGT +CGGGGC +CGGGGG +GAAAAA +GAAAAT +GAAAAC +GAAAAG +GAAATA +GAAATT +GAAATC +GAAATG +GAAACA +GAAACT +GAAACC +GAAACG +GAAAGA +GAAAGT +GAAAGC +GAAAGG +GAATAA +GAATAT +GAATAC +GAATAG +GAATTA +GAATTT +GAATTC +GAATTG +GAATCA +GAATCT +GAATCC +GAATCG +GAATGA +GAATGT +GAATGC +GAATGG +GAACAA +GAACAT +GAACAC +GAACAG +GAACTA +GAACTT +GAACTC +GAACTG +GAACCA +GAACCT +GAACCC +GAACCG +GAACGA +GAACGT +GAACGC +GAACGG +GAAGAA +GAAGAT +GAAGAC +GAAGAG +GAAGTA +GAAGTT +GAAGTC +GAAGTG +GAAGCA +GAAGCT +GAAGCC +GAAGCG +GAAGGA +GAAGGT +GAAGGC +GAAGGG +GATAAA +GATAAT +GATAAC +GATAAG +GATATA +GATATT +GATATC +GATATG +GATACA +GATACT +GATACC +GATACG +GATAGA +GATAGT +GATAGC +GATAGG +GATTAA +GATTAT +GATTAC +GATTAG +GATTTA +GATTTT +GATTTC +GATTTG +GATTCA +GATTCT +GATTCC +GATTCG +GATTGA +GATTGT +GATTGC +GATTGG +GATCAA +GATCAT +GATCAC +GATCAG +GATCTA +GATCTT +GATCTC +GATCTG +GATCCA +GATCCT +GATCCC +GATCCG +GATCGA +GATCGT +GATCGC +GATCGG +GATGAA +GATGAT +GATGAC +GATGAG +GATGTA +GATGTT +GATGTC +GATGTG +GATGCA +GATGCT +GATGCC +GATGCG +GATGGA +GATGGT +GATGGC +GATGGG +GACAAA +GACAAT +GACAAC +GACAAG +GACATA +GACATT +GACATC +GACATG +GACACA +GACACT +GACACC +GACACG +GACAGA +GACAGT +GACAGC +GACAGG +GACTAA +GACTAT +GACTAC +GACTAG +GACTTA +GACTTT +GACTTC +GACTTG +GACTCA +GACTCT +GACTCC +GACTCG +GACTGA +GACTGT +GACTGC +GACTGG +GACCAA +GACCAT +GACCAC +GACCAG +GACCTA +GACCTT +GACCTC +GACCTG +GACCCA +GACCCT +GACCCC +GACCCG +GACCGA +GACCGT +GACCGC +GACCGG +GACGAA +GACGAT +GACGAC +GACGAG +GACGTA +GACGTT +GACGTC +GACGTG +GACGCA +GACGCT +GACGCC +GACGCG +GACGGA +GACGGT +GACGGC +GACGGG +GAGAAA +GAGAAT +GAGAAC +GAGAAG +GAGATA +GAGATT +GAGATC +GAGATG +GAGACA +GAGACT +GAGACC +GAGACG +GAGAGA +GAGAGT +GAGAGC +GAGAGG +GAGTAA +GAGTAT +GAGTAC +GAGTAG +GAGTTA +GAGTTT +GAGTTC +GAGTTG +GAGTCA +GAGTCT +GAGTCC +GAGTCG +GAGTGA +GAGTGT +GAGTGC +GAGTGG +GAGCAA +GAGCAT +GAGCAC +GAGCAG +GAGCTA +GAGCTT +GAGCTC +GAGCTG +GAGCCA +GAGCCT +GAGCCC +GAGCCG +GAGCGA +GAGCGT +GAGCGC +GAGCGG +GAGGAA +GAGGAT +GAGGAC +GAGGAG +GAGGTA +GAGGTT +GAGGTC +GAGGTG +GAGGCA +GAGGCT +GAGGCC +GAGGCG +GAGGGA +GAGGGT +GAGGGC +GAGGGG +GTAAAA +GTAAAT +GTAAAC +GTAAAG +GTAATA +GTAATT +GTAATC +GTAATG +GTAACA +GTAACT +GTAACC +GTAACG +GTAAGA +GTAAGT +GTAAGC +GTAAGG +GTATAA +GTATAT +GTATAC +GTATAG +GTATTA +GTATTT +GTATTC +GTATTG +GTATCA +GTATCT +GTATCC +GTATCG +GTATGA +GTATGT +GTATGC +GTATGG +GTACAA +GTACAT +GTACAC +GTACAG +GTACTA +GTACTT +GTACTC +GTACTG +GTACCA +GTACCT +GTACCC +GTACCG +GTACGA +GTACGT +GTACGC +GTACGG +GTAGAA +GTAGAT +GTAGAC +GTAGAG +GTAGTA +GTAGTT +GTAGTC +GTAGTG +GTAGCA +GTAGCT +GTAGCC +GTAGCG +GTAGGA +GTAGGT +GTAGGC +GTAGGG +GTTAAA +GTTAAT +GTTAAC +GTTAAG +GTTATA +GTTATT +GTTATC +GTTATG +GTTACA +GTTACT +GTTACC +GTTACG +GTTAGA +GTTAGT +GTTAGC +GTTAGG +GTTTAA +GTTTAT +GTTTAC +GTTTAG +GTTTTA +GTTTTT +GTTTTC +GTTTTG +GTTTCA +GTTTCT +GTTTCC +GTTTCG +GTTTGA +GTTTGT +GTTTGC +GTTTGG +GTTCAA +GTTCAT +GTTCAC +GTTCAG +GTTCTA +GTTCTT +GTTCTC +GTTCTG +GTTCCA +GTTCCT +GTTCCC +GTTCCG +GTTCGA +GTTCGT +GTTCGC +GTTCGG +GTTGAA +GTTGAT +GTTGAC +GTTGAG +GTTGTA +GTTGTT +GTTGTC +GTTGTG +GTTGCA +GTTGCT +GTTGCC +GTTGCG +GTTGGA +GTTGGT +GTTGGC +GTTGGG +GTCAAA +GTCAAT +GTCAAC +GTCAAG +GTCATA +GTCATT +GTCATC +GTCATG +GTCACA +GTCACT +GTCACC +GTCACG +GTCAGA +GTCAGT +GTCAGC +GTCAGG +GTCTAA +GTCTAT +GTCTAC +GTCTAG +GTCTTA +GTCTTT +GTCTTC +GTCTTG +GTCTCA +GTCTCT +GTCTCC +GTCTCG +GTCTGA +GTCTGT +GTCTGC +GTCTGG +GTCCAA +GTCCAT +GTCCAC +GTCCAG +GTCCTA +GTCCTT +GTCCTC +GTCCTG +GTCCCA +GTCCCT +GTCCCC +GTCCCG +GTCCGA +GTCCGT +GTCCGC +GTCCGG +GTCGAA +GTCGAT +GTCGAC +GTCGAG +GTCGTA +GTCGTT +GTCGTC +GTCGTG +GTCGCA +GTCGCT +GTCGCC +GTCGCG +GTCGGA +GTCGGT +GTCGGC +GTCGGG +GTGAAA +GTGAAT +GTGAAC +GTGAAG +GTGATA +GTGATT +GTGATC +GTGATG +GTGACA +GTGACT +GTGACC +GTGACG +GTGAGA +GTGAGT +GTGAGC +GTGAGG +GTGTAA +GTGTAT +GTGTAC +GTGTAG +GTGTTA +GTGTTT +GTGTTC +GTGTTG +GTGTCA +GTGTCT +GTGTCC +GTGTCG +GTGTGA +GTGTGT +GTGTGC +GTGTGG +GTGCAA +GTGCAT +GTGCAC +GTGCAG +GTGCTA +GTGCTT +GTGCTC +GTGCTG +GTGCCA +GTGCCT +GTGCCC +GTGCCG +GTGCGA +GTGCGT +GTGCGC +GTGCGG +GTGGAA +GTGGAT +GTGGAC +GTGGAG +GTGGTA +GTGGTT +GTGGTC +GTGGTG +GTGGCA +GTGGCT +GTGGCC +GTGGCG +GTGGGA +GTGGGT +GTGGGC +GTGGGG +GCAAAA +GCAAAT +GCAAAC +GCAAAG +GCAATA +GCAATT +GCAATC +GCAATG +GCAACA +GCAACT +GCAACC +GCAACG +GCAAGA +GCAAGT +GCAAGC +GCAAGG +GCATAA +GCATAT +GCATAC +GCATAG +GCATTA +GCATTT +GCATTC +GCATTG +GCATCA +GCATCT +GCATCC +GCATCG +GCATGA +GCATGT +GCATGC +GCATGG +GCACAA +GCACAT +GCACAC +GCACAG +GCACTA +GCACTT +GCACTC +GCACTG +GCACCA +GCACCT +GCACCC +GCACCG +GCACGA +GCACGT +GCACGC +GCACGG +GCAGAA +GCAGAT +GCAGAC +GCAGAG +GCAGTA +GCAGTT +GCAGTC +GCAGTG +GCAGCA +GCAGCT +GCAGCC +GCAGCG +GCAGGA +GCAGGT +GCAGGC +GCAGGG +GCTAAA +GCTAAT +GCTAAC +GCTAAG +GCTATA +GCTATT +GCTATC +GCTATG +GCTACA +GCTACT +GCTACC +GCTACG +GCTAGA +GCTAGT +GCTAGC +GCTAGG +GCTTAA +GCTTAT +GCTTAC +GCTTAG +GCTTTA +GCTTTT +GCTTTC +GCTTTG +GCTTCA +GCTTCT +GCTTCC +GCTTCG +GCTTGA +GCTTGT +GCTTGC +GCTTGG +GCTCAA +GCTCAT +GCTCAC +GCTCAG +GCTCTA +GCTCTT +GCTCTC +GCTCTG +GCTCCA +GCTCCT +GCTCCC +GCTCCG +GCTCGA +GCTCGT +GCTCGC +GCTCGG +GCTGAA +GCTGAT +GCTGAC +GCTGAG +GCTGTA +GCTGTT +GCTGTC +GCTGTG +GCTGCA +GCTGCT +GCTGCC +GCTGCG +GCTGGA +GCTGGT +GCTGGC +GCTGGG +GCCAAA +GCCAAT +GCCAAC +GCCAAG +GCCATA +GCCATT +GCCATC +GCCATG +GCCACA +GCCACT +GCCACC +GCCACG +GCCAGA +GCCAGT +GCCAGC +GCCAGG +GCCTAA +GCCTAT +GCCTAC +GCCTAG +GCCTTA +GCCTTT +GCCTTC +GCCTTG +GCCTCA +GCCTCT +GCCTCC +GCCTCG +GCCTGA +GCCTGT +GCCTGC +GCCTGG +GCCCAA +GCCCAT +GCCCAC +GCCCAG +GCCCTA +GCCCTT +GCCCTC +GCCCTG +GCCCCA +GCCCCT +GCCCCC +GCCCCG +GCCCGA +GCCCGT +GCCCGC +GCCCGG +GCCGAA +GCCGAT +GCCGAC +GCCGAG +GCCGTA +GCCGTT +GCCGTC +GCCGTG +GCCGCA +GCCGCT +GCCGCC +GCCGCG +GCCGGA +GCCGGT +GCCGGC +GCCGGG +GCGAAA +GCGAAT +GCGAAC +GCGAAG +GCGATA +GCGATT +GCGATC +GCGATG +GCGACA +GCGACT +GCGACC +GCGACG +GCGAGA +GCGAGT +GCGAGC +GCGAGG +GCGTAA +GCGTAT +GCGTAC +GCGTAG +GCGTTA +GCGTTT +GCGTTC +GCGTTG +GCGTCA +GCGTCT +GCGTCC +GCGTCG +GCGTGA +GCGTGT +GCGTGC +GCGTGG +GCGCAA +GCGCAT +GCGCAC +GCGCAG +GCGCTA +GCGCTT +GCGCTC +GCGCTG +GCGCCA +GCGCCT +GCGCCC +GCGCCG +GCGCGA +GCGCGT +GCGCGC +GCGCGG +GCGGAA +GCGGAT +GCGGAC +GCGGAG +GCGGTA +GCGGTT +GCGGTC +GCGGTG +GCGGCA +GCGGCT +GCGGCC +GCGGCG +GCGGGA +GCGGGT +GCGGGC +GCGGGG +GGAAAA +GGAAAT +GGAAAC +GGAAAG +GGAATA +GGAATT +GGAATC +GGAATG +GGAACA +GGAACT +GGAACC +GGAACG +GGAAGA +GGAAGT +GGAAGC +GGAAGG +GGATAA +GGATAT +GGATAC +GGATAG +GGATTA +GGATTT +GGATTC +GGATTG +GGATCA +GGATCT +GGATCC +GGATCG +GGATGA +GGATGT +GGATGC +GGATGG +GGACAA +GGACAT +GGACAC +GGACAG +GGACTA +GGACTT +GGACTC +GGACTG +GGACCA +GGACCT +GGACCC +GGACCG +GGACGA +GGACGT +GGACGC +GGACGG +GGAGAA +GGAGAT +GGAGAC +GGAGAG +GGAGTA +GGAGTT +GGAGTC +GGAGTG +GGAGCA +GGAGCT +GGAGCC +GGAGCG +GGAGGA +GGAGGT +GGAGGC +GGAGGG +GGTAAA +GGTAAT +GGTAAC +GGTAAG +GGTATA +GGTATT +GGTATC +GGTATG +GGTACA +GGTACT +GGTACC +GGTACG +GGTAGA +GGTAGT +GGTAGC +GGTAGG +GGTTAA +GGTTAT +GGTTAC +GGTTAG +GGTTTA +GGTTTT +GGTTTC +GGTTTG +GGTTCA +GGTTCT +GGTTCC +GGTTCG +GGTTGA +GGTTGT +GGTTGC +GGTTGG +GGTCAA +GGTCAT +GGTCAC +GGTCAG +GGTCTA +GGTCTT +GGTCTC +GGTCTG +GGTCCA +GGTCCT +GGTCCC +GGTCCG +GGTCGA +GGTCGT +GGTCGC +GGTCGG +GGTGAA +GGTGAT +GGTGAC +GGTGAG +GGTGTA +GGTGTT +GGTGTC +GGTGTG +GGTGCA +GGTGCT +GGTGCC +GGTGCG +GGTGGA +GGTGGT +GGTGGC +GGTGGG +GGCAAA +GGCAAT +GGCAAC +GGCAAG +GGCATA +GGCATT +GGCATC +GGCATG +GGCACA +GGCACT +GGCACC +GGCACG +GGCAGA +GGCAGT +GGCAGC +GGCAGG +GGCTAA +GGCTAT +GGCTAC +GGCTAG +GGCTTA +GGCTTT +GGCTTC +GGCTTG +GGCTCA +GGCTCT +GGCTCC +GGCTCG +GGCTGA +GGCTGT +GGCTGC +GGCTGG +GGCCAA +GGCCAT +GGCCAC +GGCCAG +GGCCTA +GGCCTT +GGCCTC +GGCCTG +GGCCCA +GGCCCT +GGCCCC +GGCCCG +GGCCGA +GGCCGT +GGCCGC +GGCCGG +GGCGAA +GGCGAT +GGCGAC +GGCGAG +GGCGTA +GGCGTT +GGCGTC +GGCGTG +GGCGCA +GGCGCT +GGCGCC +GGCGCG +GGCGGA +GGCGGT +GGCGGC +GGCGGG +GGGAAA +GGGAAT +GGGAAC +GGGAAG +GGGATA +GGGATT +GGGATC +GGGATG +GGGACA +GGGACT +GGGACC +GGGACG +GGGAGA +GGGAGT +GGGAGC +GGGAGG +GGGTAA +GGGTAT +GGGTAC +GGGTAG +GGGTTA +GGGTTT +GGGTTC +GGGTTG +GGGTCA +GGGTCT +GGGTCC +GGGTCG +GGGTGA +GGGTGT +GGGTGC +GGGTGG +GGGCAA +GGGCAT +GGGCAC +GGGCAG +GGGCTA +GGGCTT +GGGCTC +GGGCTG +GGGCCA +GGGCCT +GGGCCC +GGGCCG +GGGCGA +GGGCGT +GGGCGC +GGGCGG +GGGGAA +GGGGAT +GGGGAC +GGGGAG +GGGGTA +GGGGTT +GGGGTC +GGGGTG +GGGGCA +GGGGCT +GGGGCC +GGGGCG +GGGGGA +GGGGGT +GGGGGC +GGGGGG diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-2000/config.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-2000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..9b14e61d327bf586fcff6e817cf9d365bae18ad2 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-2000/config.json @@ -0,0 +1,26 @@ +{ + "_name_or_path": "fabihamakhdoomi/TinyDNABERT", + "architectures": [ + "BertForSequenceClassification" + ], + "attention_probs_dropout_prob": 0.1, + "classifier_dropout": null, + "hidden_act": "gelu", + "hidden_dropout_prob": 0.1, + "hidden_size": 128, + "initializer_range": 0.02, + "intermediate_size": 512, + "layer_norm_eps": 1e-12, + "max_position_embeddings": 512, + "model_type": "bert", + "num_attention_heads": 2, + "num_hidden_layers": 2, + "pad_token_id": 0, + "position_embedding_type": "absolute", + "problem_type": "single_label_classification", + "torch_dtype": "float32", + "transformers_version": "4.44.0", + "type_vocab_size": 2, + "use_cache": true, + "vocab_size": 4101 +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-2000/model.safetensors b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-2000/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..b02b9856594004c1a651f3a21666b5e3f44c0f87 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-2000/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:818800f0d9a43c89c4fc60c39c93ad103aceb15c503d6f2e12190e111f2323ec +size 4021688 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-2000/optimizer.pt b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-2000/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..37012fac8c7c4a98507026ed140b82e2b481f641 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-2000/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:433801f96615b3478f7f831030021309b5b5f3e40574a76fbac6cef3a936f904 +size 8058700 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-2000/rng_state.pth b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-2000/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..dc62b940907551fbd6412f1b948afb4ecb3c061f --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-2000/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b1753d15a4935b9bbb100d316f8edf3879525b0e6b5b0c8edeffb1787132e3cd +size 14244 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-2000/scheduler.pt b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-2000/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..2287c4e5e02bffea6969638a53c3d5402c6dc2b0 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-2000/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:da119a7607c7d2d96a92ebe57fbaf5fb22ebfce3399b1702b2209f460c0bfe35 +size 1064 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-2000/special_tokens_map.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-2000/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..a8b3208c2884c4efb86e49300fdd3dc877220cdf --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-2000/special_tokens_map.json @@ -0,0 +1,7 @@ +{ + "cls_token": "[CLS]", + "mask_token": "[MASK]", + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-2000/tokenizer.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-2000/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..ebd782de6963b9032ce9374cfe30f5275be1a30b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-2000/tokenizer.json @@ -0,0 +1,4265 @@ +{ + "version": "1.0", + "truncation": { + "direction": "Right", + "max_length": 512, + "strategy": "LongestFirst", + "stride": 0 + }, + "padding": { + "strategy": { + "Fixed": 512 + }, + "direction": "Right", + "pad_to_multiple_of": null, + "pad_id": 0, + "pad_type_id": 0, + "pad_token": "[PAD]" + }, + "added_tokens": [ + { + "id": 0, + "content": "[PAD]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 1, + "content": "[UNK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 2, + "content": "[CLS]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 3, + "content": "[SEP]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 4, + "content": "[MASK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + } + ], + "normalizer": { + "type": "BertNormalizer", + "clean_text": true, + "handle_chinese_chars": true, + "strip_accents": null, + "lowercase": false + }, + "pre_tokenizer": { + "type": "BertPreTokenizer" + }, + "post_processor": { + "type": "TemplateProcessing", + "single": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + } + ], + "pair": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "B", + "type_id": 1 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 1 + } + } + ], + "special_tokens": { + "[CLS]": { + "id": "[CLS]", + "ids": [ + 2 + ], + "tokens": [ + "[CLS]" + ] + }, + "[SEP]": { + "id": "[SEP]", + "ids": [ + 3 + ], + "tokens": [ + "[SEP]" + ] + } + } + }, + "decoder": { + "type": "WordPiece", + "prefix": "##", + "cleanup": true + }, + "model": { + "type": "WordPiece", + "unk_token": "[UNK]", + "continuing_subword_prefix": "##", + "max_input_chars_per_word": 100, + "vocab": { + "[PAD]": 0, + "[UNK]": 1, + "[CLS]": 2, + "[SEP]": 3, + "[MASK]": 4, + "AAAAAA": 5, + "AAAAAT": 6, + "AAAAAC": 7, + "AAAAAG": 8, + "AAAATA": 9, + "AAAATT": 10, + "AAAATC": 11, + "AAAATG": 12, + "AAAACA": 13, + "AAAACT": 14, + "AAAACC": 15, + "AAAACG": 16, + "AAAAGA": 17, + "AAAAGT": 18, + "AAAAGC": 19, + "AAAAGG": 20, + "AAATAA": 21, + "AAATAT": 22, + "AAATAC": 23, + "AAATAG": 24, + "AAATTA": 25, + "AAATTT": 26, + "AAATTC": 27, + "AAATTG": 28, + "AAATCA": 29, + "AAATCT": 30, + "AAATCC": 31, + "AAATCG": 32, + "AAATGA": 33, + "AAATGT": 34, + "AAATGC": 35, + "AAATGG": 36, + "AAACAA": 37, + "AAACAT": 38, + "AAACAC": 39, + "AAACAG": 40, + "AAACTA": 41, + "AAACTT": 42, + "AAACTC": 43, + "AAACTG": 44, + "AAACCA": 45, + "AAACCT": 46, + "AAACCC": 47, + "AAACCG": 48, + "AAACGA": 49, + "AAACGT": 50, + "AAACGC": 51, + "AAACGG": 52, + "AAAGAA": 53, + "AAAGAT": 54, + "AAAGAC": 55, + "AAAGAG": 56, + "AAAGTA": 57, + "AAAGTT": 58, + "AAAGTC": 59, + "AAAGTG": 60, + "AAAGCA": 61, + "AAAGCT": 62, + "AAAGCC": 63, + "AAAGCG": 64, + "AAAGGA": 65, + "AAAGGT": 66, + "AAAGGC": 67, + "AAAGGG": 68, + "AATAAA": 69, + "AATAAT": 70, + "AATAAC": 71, + "AATAAG": 72, + "AATATA": 73, + "AATATT": 74, + "AATATC": 75, + "AATATG": 76, + "AATACA": 77, + "AATACT": 78, + "AATACC": 79, + "AATACG": 80, + "AATAGA": 81, + "AATAGT": 82, + "AATAGC": 83, + "AATAGG": 84, + "AATTAA": 85, + "AATTAT": 86, + "AATTAC": 87, + "AATTAG": 88, + "AATTTA": 89, + "AATTTT": 90, + "AATTTC": 91, + "AATTTG": 92, + "AATTCA": 93, + "AATTCT": 94, + "AATTCC": 95, + "AATTCG": 96, + "AATTGA": 97, + "AATTGT": 98, + "AATTGC": 99, + "AATTGG": 100, + "AATCAA": 101, + "AATCAT": 102, + "AATCAC": 103, + "AATCAG": 104, + "AATCTA": 105, + "AATCTT": 106, + "AATCTC": 107, + "AATCTG": 108, + "AATCCA": 109, + "AATCCT": 110, + "AATCCC": 111, + "AATCCG": 112, + "AATCGA": 113, + "AATCGT": 114, + "AATCGC": 115, + "AATCGG": 116, + "AATGAA": 117, + "AATGAT": 118, + "AATGAC": 119, + "AATGAG": 120, + "AATGTA": 121, + "AATGTT": 122, + "AATGTC": 123, + "AATGTG": 124, + "AATGCA": 125, + "AATGCT": 126, + "AATGCC": 127, + "AATGCG": 128, + "AATGGA": 129, + "AATGGT": 130, + "AATGGC": 131, + "AATGGG": 132, + "AACAAA": 133, + "AACAAT": 134, + "AACAAC": 135, + "AACAAG": 136, + "AACATA": 137, + "AACATT": 138, + "AACATC": 139, + "AACATG": 140, + "AACACA": 141, + "AACACT": 142, + "AACACC": 143, + "AACACG": 144, + "AACAGA": 145, + "AACAGT": 146, + "AACAGC": 147, + "AACAGG": 148, + "AACTAA": 149, + "AACTAT": 150, + "AACTAC": 151, + "AACTAG": 152, + "AACTTA": 153, + "AACTTT": 154, + "AACTTC": 155, + "AACTTG": 156, + "AACTCA": 157, + "AACTCT": 158, + "AACTCC": 159, + "AACTCG": 160, + "AACTGA": 161, + "AACTGT": 162, + "AACTGC": 163, + "AACTGG": 164, + "AACCAA": 165, + "AACCAT": 166, + "AACCAC": 167, + "AACCAG": 168, + "AACCTA": 169, + "AACCTT": 170, + "AACCTC": 171, + "AACCTG": 172, + "AACCCA": 173, + "AACCCT": 174, + "AACCCC": 175, + "AACCCG": 176, + "AACCGA": 177, + "AACCGT": 178, + "AACCGC": 179, + "AACCGG": 180, + "AACGAA": 181, + "AACGAT": 182, + "AACGAC": 183, + "AACGAG": 184, + "AACGTA": 185, + "AACGTT": 186, + "AACGTC": 187, + "AACGTG": 188, + "AACGCA": 189, + "AACGCT": 190, + "AACGCC": 191, + "AACGCG": 192, + "AACGGA": 193, + "AACGGT": 194, + "AACGGC": 195, + "AACGGG": 196, + "AAGAAA": 197, + "AAGAAT": 198, + "AAGAAC": 199, + "AAGAAG": 200, + "AAGATA": 201, + "AAGATT": 202, + "AAGATC": 203, + "AAGATG": 204, + "AAGACA": 205, + "AAGACT": 206, + "AAGACC": 207, + "AAGACG": 208, + "AAGAGA": 209, + "AAGAGT": 210, + "AAGAGC": 211, + "AAGAGG": 212, + "AAGTAA": 213, + "AAGTAT": 214, + "AAGTAC": 215, + "AAGTAG": 216, + "AAGTTA": 217, + "AAGTTT": 218, + "AAGTTC": 219, + "AAGTTG": 220, + "AAGTCA": 221, + "AAGTCT": 222, + "AAGTCC": 223, + "AAGTCG": 224, + "AAGTGA": 225, + "AAGTGT": 226, + "AAGTGC": 227, + "AAGTGG": 228, + "AAGCAA": 229, + "AAGCAT": 230, + "AAGCAC": 231, + "AAGCAG": 232, + "AAGCTA": 233, + "AAGCTT": 234, + "AAGCTC": 235, + "AAGCTG": 236, + "AAGCCA": 237, + "AAGCCT": 238, + "AAGCCC": 239, + "AAGCCG": 240, + "AAGCGA": 241, + "AAGCGT": 242, + "AAGCGC": 243, + "AAGCGG": 244, + "AAGGAA": 245, + "AAGGAT": 246, + "AAGGAC": 247, + "AAGGAG": 248, + "AAGGTA": 249, + "AAGGTT": 250, + "AAGGTC": 251, + "AAGGTG": 252, + "AAGGCA": 253, + "AAGGCT": 254, + "AAGGCC": 255, + "AAGGCG": 256, + "AAGGGA": 257, + "AAGGGT": 258, + "AAGGGC": 259, + "AAGGGG": 260, + "ATAAAA": 261, + "ATAAAT": 262, + "ATAAAC": 263, + "ATAAAG": 264, + "ATAATA": 265, + "ATAATT": 266, + "ATAATC": 267, + "ATAATG": 268, + "ATAACA": 269, + "ATAACT": 270, + "ATAACC": 271, + "ATAACG": 272, + "ATAAGA": 273, + "ATAAGT": 274, + "ATAAGC": 275, + "ATAAGG": 276, + "ATATAA": 277, + "ATATAT": 278, + "ATATAC": 279, + "ATATAG": 280, + "ATATTA": 281, + "ATATTT": 282, + "ATATTC": 283, + "ATATTG": 284, + "ATATCA": 285, + "ATATCT": 286, + "ATATCC": 287, + "ATATCG": 288, + "ATATGA": 289, + "ATATGT": 290, + "ATATGC": 291, + "ATATGG": 292, + "ATACAA": 293, + "ATACAT": 294, + "ATACAC": 295, + "ATACAG": 296, + "ATACTA": 297, + "ATACTT": 298, + "ATACTC": 299, + "ATACTG": 300, + "ATACCA": 301, + "ATACCT": 302, + "ATACCC": 303, + "ATACCG": 304, + "ATACGA": 305, + "ATACGT": 306, + "ATACGC": 307, + "ATACGG": 308, + "ATAGAA": 309, + "ATAGAT": 310, + "ATAGAC": 311, + "ATAGAG": 312, + "ATAGTA": 313, + "ATAGTT": 314, + "ATAGTC": 315, + "ATAGTG": 316, + "ATAGCA": 317, + "ATAGCT": 318, + "ATAGCC": 319, + "ATAGCG": 320, + "ATAGGA": 321, + "ATAGGT": 322, + "ATAGGC": 323, + "ATAGGG": 324, + "ATTAAA": 325, + "ATTAAT": 326, + "ATTAAC": 327, + "ATTAAG": 328, + "ATTATA": 329, + "ATTATT": 330, + "ATTATC": 331, + "ATTATG": 332, + "ATTACA": 333, + "ATTACT": 334, + "ATTACC": 335, + "ATTACG": 336, + "ATTAGA": 337, + "ATTAGT": 338, + "ATTAGC": 339, + "ATTAGG": 340, + "ATTTAA": 341, + "ATTTAT": 342, + "ATTTAC": 343, + "ATTTAG": 344, + "ATTTTA": 345, + "ATTTTT": 346, + "ATTTTC": 347, + "ATTTTG": 348, + "ATTTCA": 349, + "ATTTCT": 350, + "ATTTCC": 351, + "ATTTCG": 352, + "ATTTGA": 353, + "ATTTGT": 354, + "ATTTGC": 355, + "ATTTGG": 356, + "ATTCAA": 357, + "ATTCAT": 358, + "ATTCAC": 359, + "ATTCAG": 360, + "ATTCTA": 361, + "ATTCTT": 362, + "ATTCTC": 363, + "ATTCTG": 364, + "ATTCCA": 365, + "ATTCCT": 366, + "ATTCCC": 367, + "ATTCCG": 368, + "ATTCGA": 369, + "ATTCGT": 370, + "ATTCGC": 371, + "ATTCGG": 372, + "ATTGAA": 373, + "ATTGAT": 374, + "ATTGAC": 375, + "ATTGAG": 376, + "ATTGTA": 377, + "ATTGTT": 378, + "ATTGTC": 379, + "ATTGTG": 380, + "ATTGCA": 381, + "ATTGCT": 382, + "ATTGCC": 383, + "ATTGCG": 384, + "ATTGGA": 385, + "ATTGGT": 386, + "ATTGGC": 387, + "ATTGGG": 388, + "ATCAAA": 389, + "ATCAAT": 390, + "ATCAAC": 391, + "ATCAAG": 392, + "ATCATA": 393, + "ATCATT": 394, + "ATCATC": 395, + "ATCATG": 396, + "ATCACA": 397, + "ATCACT": 398, + "ATCACC": 399, + "ATCACG": 400, + "ATCAGA": 401, + "ATCAGT": 402, + "ATCAGC": 403, + "ATCAGG": 404, + "ATCTAA": 405, + "ATCTAT": 406, + "ATCTAC": 407, + "ATCTAG": 408, + "ATCTTA": 409, + "ATCTTT": 410, + "ATCTTC": 411, + "ATCTTG": 412, + "ATCTCA": 413, + "ATCTCT": 414, + "ATCTCC": 415, + "ATCTCG": 416, + "ATCTGA": 417, + "ATCTGT": 418, + "ATCTGC": 419, + "ATCTGG": 420, + "ATCCAA": 421, + "ATCCAT": 422, + "ATCCAC": 423, + "ATCCAG": 424, + "ATCCTA": 425, + "ATCCTT": 426, + "ATCCTC": 427, + "ATCCTG": 428, + "ATCCCA": 429, + "ATCCCT": 430, + "ATCCCC": 431, + "ATCCCG": 432, + "ATCCGA": 433, + "ATCCGT": 434, + "ATCCGC": 435, + "ATCCGG": 436, + "ATCGAA": 437, + "ATCGAT": 438, + "ATCGAC": 439, + "ATCGAG": 440, + "ATCGTA": 441, + "ATCGTT": 442, + "ATCGTC": 443, + "ATCGTG": 444, + "ATCGCA": 445, + "ATCGCT": 446, + "ATCGCC": 447, + "ATCGCG": 448, + "ATCGGA": 449, + "ATCGGT": 450, + "ATCGGC": 451, + "ATCGGG": 452, + "ATGAAA": 453, + "ATGAAT": 454, + "ATGAAC": 455, + "ATGAAG": 456, + "ATGATA": 457, + "ATGATT": 458, + "ATGATC": 459, + "ATGATG": 460, + "ATGACA": 461, + "ATGACT": 462, + "ATGACC": 463, + "ATGACG": 464, + "ATGAGA": 465, + "ATGAGT": 466, + "ATGAGC": 467, + "ATGAGG": 468, + "ATGTAA": 469, + "ATGTAT": 470, + "ATGTAC": 471, + "ATGTAG": 472, + "ATGTTA": 473, + "ATGTTT": 474, + "ATGTTC": 475, + "ATGTTG": 476, + "ATGTCA": 477, + "ATGTCT": 478, + "ATGTCC": 479, + "ATGTCG": 480, + "ATGTGA": 481, + "ATGTGT": 482, + "ATGTGC": 483, + "ATGTGG": 484, + "ATGCAA": 485, + "ATGCAT": 486, + "ATGCAC": 487, + "ATGCAG": 488, + "ATGCTA": 489, + "ATGCTT": 490, + "ATGCTC": 491, + "ATGCTG": 492, + "ATGCCA": 493, + "ATGCCT": 494, + "ATGCCC": 495, + "ATGCCG": 496, + "ATGCGA": 497, + "ATGCGT": 498, + "ATGCGC": 499, + "ATGCGG": 500, + "ATGGAA": 501, + "ATGGAT": 502, + "ATGGAC": 503, + "ATGGAG": 504, + "ATGGTA": 505, + "ATGGTT": 506, + "ATGGTC": 507, + "ATGGTG": 508, + "ATGGCA": 509, + "ATGGCT": 510, + "ATGGCC": 511, + "ATGGCG": 512, + "ATGGGA": 513, + "ATGGGT": 514, + "ATGGGC": 515, + "ATGGGG": 516, + "ACAAAA": 517, + "ACAAAT": 518, + "ACAAAC": 519, + "ACAAAG": 520, + "ACAATA": 521, + "ACAATT": 522, + "ACAATC": 523, + "ACAATG": 524, + "ACAACA": 525, + "ACAACT": 526, + "ACAACC": 527, + "ACAACG": 528, + "ACAAGA": 529, + "ACAAGT": 530, + "ACAAGC": 531, + "ACAAGG": 532, + "ACATAA": 533, + "ACATAT": 534, + "ACATAC": 535, + "ACATAG": 536, + "ACATTA": 537, + "ACATTT": 538, + "ACATTC": 539, + "ACATTG": 540, + "ACATCA": 541, + "ACATCT": 542, + "ACATCC": 543, + "ACATCG": 544, + "ACATGA": 545, + "ACATGT": 546, + "ACATGC": 547, + "ACATGG": 548, + "ACACAA": 549, + "ACACAT": 550, + "ACACAC": 551, + "ACACAG": 552, + "ACACTA": 553, + "ACACTT": 554, + "ACACTC": 555, + "ACACTG": 556, + "ACACCA": 557, + "ACACCT": 558, + "ACACCC": 559, + "ACACCG": 560, + "ACACGA": 561, + "ACACGT": 562, + "ACACGC": 563, + "ACACGG": 564, + "ACAGAA": 565, + "ACAGAT": 566, + "ACAGAC": 567, + "ACAGAG": 568, + "ACAGTA": 569, + "ACAGTT": 570, + "ACAGTC": 571, + "ACAGTG": 572, + "ACAGCA": 573, + "ACAGCT": 574, + "ACAGCC": 575, + "ACAGCG": 576, + "ACAGGA": 577, + "ACAGGT": 578, + "ACAGGC": 579, + "ACAGGG": 580, + "ACTAAA": 581, + "ACTAAT": 582, + "ACTAAC": 583, + "ACTAAG": 584, + "ACTATA": 585, + "ACTATT": 586, + "ACTATC": 587, + "ACTATG": 588, + "ACTACA": 589, + "ACTACT": 590, + "ACTACC": 591, + "ACTACG": 592, + "ACTAGA": 593, + "ACTAGT": 594, + "ACTAGC": 595, + "ACTAGG": 596, + "ACTTAA": 597, + "ACTTAT": 598, + "ACTTAC": 599, + "ACTTAG": 600, + "ACTTTA": 601, + "ACTTTT": 602, + "ACTTTC": 603, + "ACTTTG": 604, + "ACTTCA": 605, + "ACTTCT": 606, + "ACTTCC": 607, + "ACTTCG": 608, + "ACTTGA": 609, + "ACTTGT": 610, + "ACTTGC": 611, + "ACTTGG": 612, + "ACTCAA": 613, + "ACTCAT": 614, + "ACTCAC": 615, + "ACTCAG": 616, + "ACTCTA": 617, + "ACTCTT": 618, + "ACTCTC": 619, + "ACTCTG": 620, + "ACTCCA": 621, + "ACTCCT": 622, + "ACTCCC": 623, + "ACTCCG": 624, + "ACTCGA": 625, + "ACTCGT": 626, + "ACTCGC": 627, + "ACTCGG": 628, + "ACTGAA": 629, + "ACTGAT": 630, + "ACTGAC": 631, + "ACTGAG": 632, + "ACTGTA": 633, + "ACTGTT": 634, + "ACTGTC": 635, + "ACTGTG": 636, + "ACTGCA": 637, + "ACTGCT": 638, + "ACTGCC": 639, + "ACTGCG": 640, + "ACTGGA": 641, + "ACTGGT": 642, + "ACTGGC": 643, + "ACTGGG": 644, + "ACCAAA": 645, + "ACCAAT": 646, + "ACCAAC": 647, + "ACCAAG": 648, + "ACCATA": 649, + "ACCATT": 650, + "ACCATC": 651, + "ACCATG": 652, + "ACCACA": 653, + "ACCACT": 654, + "ACCACC": 655, + "ACCACG": 656, + "ACCAGA": 657, + "ACCAGT": 658, + "ACCAGC": 659, + "ACCAGG": 660, + "ACCTAA": 661, + "ACCTAT": 662, + "ACCTAC": 663, + "ACCTAG": 664, + "ACCTTA": 665, + "ACCTTT": 666, + "ACCTTC": 667, + "ACCTTG": 668, + "ACCTCA": 669, + "ACCTCT": 670, + "ACCTCC": 671, + "ACCTCG": 672, + "ACCTGA": 673, + "ACCTGT": 674, + "ACCTGC": 675, + "ACCTGG": 676, + "ACCCAA": 677, + "ACCCAT": 678, + "ACCCAC": 679, + "ACCCAG": 680, + "ACCCTA": 681, + "ACCCTT": 682, + "ACCCTC": 683, + "ACCCTG": 684, + "ACCCCA": 685, + "ACCCCT": 686, + "ACCCCC": 687, + "ACCCCG": 688, + "ACCCGA": 689, + "ACCCGT": 690, + "ACCCGC": 691, + "ACCCGG": 692, + "ACCGAA": 693, + "ACCGAT": 694, + "ACCGAC": 695, + "ACCGAG": 696, + "ACCGTA": 697, + "ACCGTT": 698, + "ACCGTC": 699, + "ACCGTG": 700, + "ACCGCA": 701, + "ACCGCT": 702, + "ACCGCC": 703, + "ACCGCG": 704, + "ACCGGA": 705, + "ACCGGT": 706, + "ACCGGC": 707, + "ACCGGG": 708, + "ACGAAA": 709, + "ACGAAT": 710, + "ACGAAC": 711, + "ACGAAG": 712, + "ACGATA": 713, + "ACGATT": 714, + "ACGATC": 715, + "ACGATG": 716, + "ACGACA": 717, + "ACGACT": 718, + "ACGACC": 719, + "ACGACG": 720, + "ACGAGA": 721, + "ACGAGT": 722, + "ACGAGC": 723, + "ACGAGG": 724, + "ACGTAA": 725, + "ACGTAT": 726, + "ACGTAC": 727, + "ACGTAG": 728, + "ACGTTA": 729, + "ACGTTT": 730, + "ACGTTC": 731, + "ACGTTG": 732, + "ACGTCA": 733, + "ACGTCT": 734, + "ACGTCC": 735, + "ACGTCG": 736, + "ACGTGA": 737, + "ACGTGT": 738, + "ACGTGC": 739, + "ACGTGG": 740, + "ACGCAA": 741, + "ACGCAT": 742, + "ACGCAC": 743, + "ACGCAG": 744, + "ACGCTA": 745, + "ACGCTT": 746, + "ACGCTC": 747, + "ACGCTG": 748, + "ACGCCA": 749, + "ACGCCT": 750, + "ACGCCC": 751, + "ACGCCG": 752, + "ACGCGA": 753, + "ACGCGT": 754, + "ACGCGC": 755, + "ACGCGG": 756, + "ACGGAA": 757, + "ACGGAT": 758, + "ACGGAC": 759, + "ACGGAG": 760, + "ACGGTA": 761, + "ACGGTT": 762, + "ACGGTC": 763, + "ACGGTG": 764, + "ACGGCA": 765, + "ACGGCT": 766, + "ACGGCC": 767, + "ACGGCG": 768, + "ACGGGA": 769, + "ACGGGT": 770, + "ACGGGC": 771, + "ACGGGG": 772, + "AGAAAA": 773, + "AGAAAT": 774, + "AGAAAC": 775, + "AGAAAG": 776, + "AGAATA": 777, + "AGAATT": 778, + "AGAATC": 779, + "AGAATG": 780, + "AGAACA": 781, + "AGAACT": 782, + "AGAACC": 783, + "AGAACG": 784, + "AGAAGA": 785, + "AGAAGT": 786, + "AGAAGC": 787, + "AGAAGG": 788, + "AGATAA": 789, + "AGATAT": 790, + "AGATAC": 791, + "AGATAG": 792, + "AGATTA": 793, + "AGATTT": 794, + "AGATTC": 795, + "AGATTG": 796, + "AGATCA": 797, + "AGATCT": 798, + "AGATCC": 799, + "AGATCG": 800, + "AGATGA": 801, + "AGATGT": 802, + "AGATGC": 803, + "AGATGG": 804, + "AGACAA": 805, + "AGACAT": 806, + "AGACAC": 807, + "AGACAG": 808, + "AGACTA": 809, + "AGACTT": 810, + "AGACTC": 811, + "AGACTG": 812, + "AGACCA": 813, + "AGACCT": 814, + "AGACCC": 815, + "AGACCG": 816, + "AGACGA": 817, + "AGACGT": 818, + "AGACGC": 819, + "AGACGG": 820, + "AGAGAA": 821, + "AGAGAT": 822, + "AGAGAC": 823, + "AGAGAG": 824, + "AGAGTA": 825, + "AGAGTT": 826, + "AGAGTC": 827, + "AGAGTG": 828, + "AGAGCA": 829, + "AGAGCT": 830, + "AGAGCC": 831, + "AGAGCG": 832, + "AGAGGA": 833, + "AGAGGT": 834, + "AGAGGC": 835, + "AGAGGG": 836, + "AGTAAA": 837, + "AGTAAT": 838, + "AGTAAC": 839, + "AGTAAG": 840, + "AGTATA": 841, + "AGTATT": 842, + "AGTATC": 843, + "AGTATG": 844, + "AGTACA": 845, + "AGTACT": 846, + "AGTACC": 847, + "AGTACG": 848, + "AGTAGA": 849, + "AGTAGT": 850, + "AGTAGC": 851, + "AGTAGG": 852, + "AGTTAA": 853, + "AGTTAT": 854, + "AGTTAC": 855, + "AGTTAG": 856, + "AGTTTA": 857, + "AGTTTT": 858, + "AGTTTC": 859, + "AGTTTG": 860, + "AGTTCA": 861, + "AGTTCT": 862, + "AGTTCC": 863, + "AGTTCG": 864, + "AGTTGA": 865, + "AGTTGT": 866, + "AGTTGC": 867, + "AGTTGG": 868, + "AGTCAA": 869, + "AGTCAT": 870, + "AGTCAC": 871, + "AGTCAG": 872, + "AGTCTA": 873, + "AGTCTT": 874, + "AGTCTC": 875, + "AGTCTG": 876, + "AGTCCA": 877, + "AGTCCT": 878, + "AGTCCC": 879, + "AGTCCG": 880, + "AGTCGA": 881, + "AGTCGT": 882, + "AGTCGC": 883, + "AGTCGG": 884, + "AGTGAA": 885, + "AGTGAT": 886, + "AGTGAC": 887, + "AGTGAG": 888, + "AGTGTA": 889, + "AGTGTT": 890, + "AGTGTC": 891, + "AGTGTG": 892, + "AGTGCA": 893, + "AGTGCT": 894, + "AGTGCC": 895, + "AGTGCG": 896, + "AGTGGA": 897, + "AGTGGT": 898, + "AGTGGC": 899, + "AGTGGG": 900, + "AGCAAA": 901, + "AGCAAT": 902, + "AGCAAC": 903, + "AGCAAG": 904, + "AGCATA": 905, + "AGCATT": 906, + "AGCATC": 907, + "AGCATG": 908, + "AGCACA": 909, + "AGCACT": 910, + "AGCACC": 911, + "AGCACG": 912, + "AGCAGA": 913, + "AGCAGT": 914, + "AGCAGC": 915, + "AGCAGG": 916, + "AGCTAA": 917, + "AGCTAT": 918, + "AGCTAC": 919, + "AGCTAG": 920, + "AGCTTA": 921, + "AGCTTT": 922, + "AGCTTC": 923, + "AGCTTG": 924, + "AGCTCA": 925, + "AGCTCT": 926, + "AGCTCC": 927, + "AGCTCG": 928, + "AGCTGA": 929, + "AGCTGT": 930, + "AGCTGC": 931, + "AGCTGG": 932, + "AGCCAA": 933, + "AGCCAT": 934, + "AGCCAC": 935, + "AGCCAG": 936, + "AGCCTA": 937, + "AGCCTT": 938, + "AGCCTC": 939, + "AGCCTG": 940, + "AGCCCA": 941, + "AGCCCT": 942, + "AGCCCC": 943, + "AGCCCG": 944, + "AGCCGA": 945, + "AGCCGT": 946, + "AGCCGC": 947, + "AGCCGG": 948, + "AGCGAA": 949, + "AGCGAT": 950, + "AGCGAC": 951, + "AGCGAG": 952, + "AGCGTA": 953, + "AGCGTT": 954, + "AGCGTC": 955, + "AGCGTG": 956, + "AGCGCA": 957, + "AGCGCT": 958, + "AGCGCC": 959, + "AGCGCG": 960, + "AGCGGA": 961, + "AGCGGT": 962, + "AGCGGC": 963, + "AGCGGG": 964, + "AGGAAA": 965, + "AGGAAT": 966, + "AGGAAC": 967, + "AGGAAG": 968, + "AGGATA": 969, + "AGGATT": 970, + "AGGATC": 971, + "AGGATG": 972, + "AGGACA": 973, + "AGGACT": 974, + "AGGACC": 975, + "AGGACG": 976, + "AGGAGA": 977, + "AGGAGT": 978, + "AGGAGC": 979, + "AGGAGG": 980, + "AGGTAA": 981, + "AGGTAT": 982, + "AGGTAC": 983, + "AGGTAG": 984, + "AGGTTA": 985, + "AGGTTT": 986, + "AGGTTC": 987, + "AGGTTG": 988, + "AGGTCA": 989, + "AGGTCT": 990, + "AGGTCC": 991, + "AGGTCG": 992, + "AGGTGA": 993, + "AGGTGT": 994, + "AGGTGC": 995, + "AGGTGG": 996, + "AGGCAA": 997, + "AGGCAT": 998, + "AGGCAC": 999, + "AGGCAG": 1000, + "AGGCTA": 1001, + "AGGCTT": 1002, + "AGGCTC": 1003, + "AGGCTG": 1004, + "AGGCCA": 1005, + "AGGCCT": 1006, + "AGGCCC": 1007, + "AGGCCG": 1008, + "AGGCGA": 1009, + "AGGCGT": 1010, + "AGGCGC": 1011, + "AGGCGG": 1012, + "AGGGAA": 1013, + "AGGGAT": 1014, + "AGGGAC": 1015, + "AGGGAG": 1016, + "AGGGTA": 1017, + "AGGGTT": 1018, + "AGGGTC": 1019, + "AGGGTG": 1020, + "AGGGCA": 1021, + "AGGGCT": 1022, + "AGGGCC": 1023, + "AGGGCG": 1024, + "AGGGGA": 1025, + "AGGGGT": 1026, + "AGGGGC": 1027, + "AGGGGG": 1028, + "TAAAAA": 1029, + "TAAAAT": 1030, + "TAAAAC": 1031, + "TAAAAG": 1032, + "TAAATA": 1033, + "TAAATT": 1034, + "TAAATC": 1035, + "TAAATG": 1036, + "TAAACA": 1037, + "TAAACT": 1038, + "TAAACC": 1039, + "TAAACG": 1040, + "TAAAGA": 1041, + "TAAAGT": 1042, + "TAAAGC": 1043, + "TAAAGG": 1044, + "TAATAA": 1045, + "TAATAT": 1046, + "TAATAC": 1047, + "TAATAG": 1048, + "TAATTA": 1049, + "TAATTT": 1050, + "TAATTC": 1051, + "TAATTG": 1052, + "TAATCA": 1053, + "TAATCT": 1054, + "TAATCC": 1055, + "TAATCG": 1056, + "TAATGA": 1057, + "TAATGT": 1058, + "TAATGC": 1059, + "TAATGG": 1060, + "TAACAA": 1061, + "TAACAT": 1062, + "TAACAC": 1063, + "TAACAG": 1064, + "TAACTA": 1065, + "TAACTT": 1066, + "TAACTC": 1067, + "TAACTG": 1068, + "TAACCA": 1069, + "TAACCT": 1070, + "TAACCC": 1071, + "TAACCG": 1072, + "TAACGA": 1073, + "TAACGT": 1074, + "TAACGC": 1075, + "TAACGG": 1076, + "TAAGAA": 1077, + "TAAGAT": 1078, + "TAAGAC": 1079, + "TAAGAG": 1080, + "TAAGTA": 1081, + "TAAGTT": 1082, + "TAAGTC": 1083, + "TAAGTG": 1084, + "TAAGCA": 1085, + "TAAGCT": 1086, + "TAAGCC": 1087, + "TAAGCG": 1088, + "TAAGGA": 1089, + "TAAGGT": 1090, + "TAAGGC": 1091, + "TAAGGG": 1092, + "TATAAA": 1093, + "TATAAT": 1094, + "TATAAC": 1095, + "TATAAG": 1096, + "TATATA": 1097, + "TATATT": 1098, + "TATATC": 1099, + "TATATG": 1100, + "TATACA": 1101, + "TATACT": 1102, + "TATACC": 1103, + "TATACG": 1104, + "TATAGA": 1105, + "TATAGT": 1106, + "TATAGC": 1107, + "TATAGG": 1108, + "TATTAA": 1109, + "TATTAT": 1110, + "TATTAC": 1111, + "TATTAG": 1112, + "TATTTA": 1113, + "TATTTT": 1114, + "TATTTC": 1115, + "TATTTG": 1116, + "TATTCA": 1117, + "TATTCT": 1118, + "TATTCC": 1119, + "TATTCG": 1120, + "TATTGA": 1121, + "TATTGT": 1122, + "TATTGC": 1123, + "TATTGG": 1124, + "TATCAA": 1125, + "TATCAT": 1126, + "TATCAC": 1127, + "TATCAG": 1128, + "TATCTA": 1129, + "TATCTT": 1130, + "TATCTC": 1131, + "TATCTG": 1132, + "TATCCA": 1133, + "TATCCT": 1134, + "TATCCC": 1135, + "TATCCG": 1136, + "TATCGA": 1137, + "TATCGT": 1138, + "TATCGC": 1139, + "TATCGG": 1140, + "TATGAA": 1141, + "TATGAT": 1142, + "TATGAC": 1143, + "TATGAG": 1144, + "TATGTA": 1145, + "TATGTT": 1146, + "TATGTC": 1147, + "TATGTG": 1148, + "TATGCA": 1149, + "TATGCT": 1150, + "TATGCC": 1151, + "TATGCG": 1152, + "TATGGA": 1153, + "TATGGT": 1154, + "TATGGC": 1155, + "TATGGG": 1156, + "TACAAA": 1157, + "TACAAT": 1158, + "TACAAC": 1159, + "TACAAG": 1160, + "TACATA": 1161, + "TACATT": 1162, + "TACATC": 1163, + "TACATG": 1164, + "TACACA": 1165, + "TACACT": 1166, + "TACACC": 1167, + "TACACG": 1168, + "TACAGA": 1169, + "TACAGT": 1170, + "TACAGC": 1171, + "TACAGG": 1172, + "TACTAA": 1173, + "TACTAT": 1174, + "TACTAC": 1175, + "TACTAG": 1176, + "TACTTA": 1177, + "TACTTT": 1178, + "TACTTC": 1179, + "TACTTG": 1180, + "TACTCA": 1181, + "TACTCT": 1182, + "TACTCC": 1183, + "TACTCG": 1184, + "TACTGA": 1185, + "TACTGT": 1186, + "TACTGC": 1187, + "TACTGG": 1188, + "TACCAA": 1189, + "TACCAT": 1190, + "TACCAC": 1191, + "TACCAG": 1192, + "TACCTA": 1193, + "TACCTT": 1194, + "TACCTC": 1195, + "TACCTG": 1196, + "TACCCA": 1197, + "TACCCT": 1198, + "TACCCC": 1199, + "TACCCG": 1200, + "TACCGA": 1201, + "TACCGT": 1202, + "TACCGC": 1203, + "TACCGG": 1204, + "TACGAA": 1205, + "TACGAT": 1206, + "TACGAC": 1207, + "TACGAG": 1208, + "TACGTA": 1209, + "TACGTT": 1210, + "TACGTC": 1211, + "TACGTG": 1212, + "TACGCA": 1213, + "TACGCT": 1214, + "TACGCC": 1215, + "TACGCG": 1216, + "TACGGA": 1217, + "TACGGT": 1218, + "TACGGC": 1219, + "TACGGG": 1220, + "TAGAAA": 1221, + "TAGAAT": 1222, + "TAGAAC": 1223, + "TAGAAG": 1224, + "TAGATA": 1225, + "TAGATT": 1226, + "TAGATC": 1227, + "TAGATG": 1228, + "TAGACA": 1229, + "TAGACT": 1230, + "TAGACC": 1231, + "TAGACG": 1232, + "TAGAGA": 1233, + "TAGAGT": 1234, + "TAGAGC": 1235, + "TAGAGG": 1236, + "TAGTAA": 1237, + "TAGTAT": 1238, + "TAGTAC": 1239, + "TAGTAG": 1240, + "TAGTTA": 1241, + "TAGTTT": 1242, + "TAGTTC": 1243, + "TAGTTG": 1244, + "TAGTCA": 1245, + "TAGTCT": 1246, + "TAGTCC": 1247, + "TAGTCG": 1248, + "TAGTGA": 1249, + "TAGTGT": 1250, + "TAGTGC": 1251, + "TAGTGG": 1252, + "TAGCAA": 1253, + "TAGCAT": 1254, + "TAGCAC": 1255, + "TAGCAG": 1256, + "TAGCTA": 1257, + "TAGCTT": 1258, + "TAGCTC": 1259, + "TAGCTG": 1260, + "TAGCCA": 1261, + "TAGCCT": 1262, + "TAGCCC": 1263, + "TAGCCG": 1264, + "TAGCGA": 1265, + "TAGCGT": 1266, + "TAGCGC": 1267, + "TAGCGG": 1268, + "TAGGAA": 1269, + "TAGGAT": 1270, + "TAGGAC": 1271, + "TAGGAG": 1272, + "TAGGTA": 1273, + "TAGGTT": 1274, + "TAGGTC": 1275, + "TAGGTG": 1276, + "TAGGCA": 1277, + "TAGGCT": 1278, + "TAGGCC": 1279, + "TAGGCG": 1280, + "TAGGGA": 1281, + "TAGGGT": 1282, + "TAGGGC": 1283, + "TAGGGG": 1284, + "TTAAAA": 1285, + "TTAAAT": 1286, + "TTAAAC": 1287, + "TTAAAG": 1288, + "TTAATA": 1289, + "TTAATT": 1290, + "TTAATC": 1291, + "TTAATG": 1292, + "TTAACA": 1293, + "TTAACT": 1294, + "TTAACC": 1295, + "TTAACG": 1296, + "TTAAGA": 1297, + "TTAAGT": 1298, + "TTAAGC": 1299, + "TTAAGG": 1300, + "TTATAA": 1301, + "TTATAT": 1302, + "TTATAC": 1303, + "TTATAG": 1304, + "TTATTA": 1305, + "TTATTT": 1306, + "TTATTC": 1307, + "TTATTG": 1308, + "TTATCA": 1309, + "TTATCT": 1310, + "TTATCC": 1311, + "TTATCG": 1312, + "TTATGA": 1313, + "TTATGT": 1314, + "TTATGC": 1315, + "TTATGG": 1316, + "TTACAA": 1317, + "TTACAT": 1318, + "TTACAC": 1319, + "TTACAG": 1320, + "TTACTA": 1321, + "TTACTT": 1322, + "TTACTC": 1323, + "TTACTG": 1324, + "TTACCA": 1325, + "TTACCT": 1326, + "TTACCC": 1327, + "TTACCG": 1328, + "TTACGA": 1329, + "TTACGT": 1330, + "TTACGC": 1331, + "TTACGG": 1332, + "TTAGAA": 1333, + "TTAGAT": 1334, + "TTAGAC": 1335, + "TTAGAG": 1336, + "TTAGTA": 1337, + "TTAGTT": 1338, + "TTAGTC": 1339, + "TTAGTG": 1340, + "TTAGCA": 1341, + "TTAGCT": 1342, + "TTAGCC": 1343, + "TTAGCG": 1344, + "TTAGGA": 1345, + "TTAGGT": 1346, + "TTAGGC": 1347, + "TTAGGG": 1348, + "TTTAAA": 1349, + "TTTAAT": 1350, + "TTTAAC": 1351, + "TTTAAG": 1352, + "TTTATA": 1353, + "TTTATT": 1354, + "TTTATC": 1355, + "TTTATG": 1356, + "TTTACA": 1357, + "TTTACT": 1358, + "TTTACC": 1359, + "TTTACG": 1360, + "TTTAGA": 1361, + "TTTAGT": 1362, + "TTTAGC": 1363, + "TTTAGG": 1364, + "TTTTAA": 1365, + "TTTTAT": 1366, + "TTTTAC": 1367, + "TTTTAG": 1368, + "TTTTTA": 1369, + "TTTTTT": 1370, + "TTTTTC": 1371, + "TTTTTG": 1372, + "TTTTCA": 1373, + "TTTTCT": 1374, + "TTTTCC": 1375, + "TTTTCG": 1376, + "TTTTGA": 1377, + "TTTTGT": 1378, + "TTTTGC": 1379, + "TTTTGG": 1380, + "TTTCAA": 1381, + "TTTCAT": 1382, + "TTTCAC": 1383, + "TTTCAG": 1384, + "TTTCTA": 1385, + "TTTCTT": 1386, + "TTTCTC": 1387, + "TTTCTG": 1388, + "TTTCCA": 1389, + "TTTCCT": 1390, + "TTTCCC": 1391, + "TTTCCG": 1392, + "TTTCGA": 1393, + "TTTCGT": 1394, + "TTTCGC": 1395, + "TTTCGG": 1396, + "TTTGAA": 1397, + "TTTGAT": 1398, + "TTTGAC": 1399, + "TTTGAG": 1400, + "TTTGTA": 1401, + "TTTGTT": 1402, + "TTTGTC": 1403, + "TTTGTG": 1404, + "TTTGCA": 1405, + "TTTGCT": 1406, + "TTTGCC": 1407, + "TTTGCG": 1408, + "TTTGGA": 1409, + "TTTGGT": 1410, + "TTTGGC": 1411, + "TTTGGG": 1412, + "TTCAAA": 1413, + "TTCAAT": 1414, + "TTCAAC": 1415, + "TTCAAG": 1416, + "TTCATA": 1417, + "TTCATT": 1418, + "TTCATC": 1419, + "TTCATG": 1420, + "TTCACA": 1421, + "TTCACT": 1422, + "TTCACC": 1423, + "TTCACG": 1424, + "TTCAGA": 1425, + "TTCAGT": 1426, + "TTCAGC": 1427, + "TTCAGG": 1428, + "TTCTAA": 1429, + "TTCTAT": 1430, + "TTCTAC": 1431, + "TTCTAG": 1432, + "TTCTTA": 1433, + "TTCTTT": 1434, + "TTCTTC": 1435, + "TTCTTG": 1436, + "TTCTCA": 1437, + "TTCTCT": 1438, + "TTCTCC": 1439, + "TTCTCG": 1440, + "TTCTGA": 1441, + "TTCTGT": 1442, + "TTCTGC": 1443, + "TTCTGG": 1444, + "TTCCAA": 1445, + "TTCCAT": 1446, + "TTCCAC": 1447, + "TTCCAG": 1448, + "TTCCTA": 1449, + "TTCCTT": 1450, + "TTCCTC": 1451, + "TTCCTG": 1452, + "TTCCCA": 1453, + "TTCCCT": 1454, + "TTCCCC": 1455, + "TTCCCG": 1456, + "TTCCGA": 1457, + "TTCCGT": 1458, + "TTCCGC": 1459, + "TTCCGG": 1460, + "TTCGAA": 1461, + "TTCGAT": 1462, + "TTCGAC": 1463, + "TTCGAG": 1464, + "TTCGTA": 1465, + "TTCGTT": 1466, + "TTCGTC": 1467, + "TTCGTG": 1468, + "TTCGCA": 1469, + "TTCGCT": 1470, + "TTCGCC": 1471, + "TTCGCG": 1472, + "TTCGGA": 1473, + "TTCGGT": 1474, + "TTCGGC": 1475, + "TTCGGG": 1476, + "TTGAAA": 1477, + "TTGAAT": 1478, + "TTGAAC": 1479, + "TTGAAG": 1480, + "TTGATA": 1481, + "TTGATT": 1482, + "TTGATC": 1483, + "TTGATG": 1484, + "TTGACA": 1485, + "TTGACT": 1486, + "TTGACC": 1487, + "TTGACG": 1488, + "TTGAGA": 1489, + "TTGAGT": 1490, + "TTGAGC": 1491, + "TTGAGG": 1492, + "TTGTAA": 1493, + "TTGTAT": 1494, + "TTGTAC": 1495, + "TTGTAG": 1496, + "TTGTTA": 1497, + "TTGTTT": 1498, + "TTGTTC": 1499, + "TTGTTG": 1500, + "TTGTCA": 1501, + "TTGTCT": 1502, + "TTGTCC": 1503, + "TTGTCG": 1504, + "TTGTGA": 1505, + "TTGTGT": 1506, + "TTGTGC": 1507, + "TTGTGG": 1508, + "TTGCAA": 1509, + "TTGCAT": 1510, + "TTGCAC": 1511, + "TTGCAG": 1512, + "TTGCTA": 1513, + "TTGCTT": 1514, + "TTGCTC": 1515, + "TTGCTG": 1516, + "TTGCCA": 1517, + "TTGCCT": 1518, + "TTGCCC": 1519, + "TTGCCG": 1520, + "TTGCGA": 1521, + "TTGCGT": 1522, + "TTGCGC": 1523, + "TTGCGG": 1524, + "TTGGAA": 1525, + "TTGGAT": 1526, + "TTGGAC": 1527, + "TTGGAG": 1528, + "TTGGTA": 1529, + "TTGGTT": 1530, + "TTGGTC": 1531, + "TTGGTG": 1532, + "TTGGCA": 1533, + "TTGGCT": 1534, + "TTGGCC": 1535, + "TTGGCG": 1536, + "TTGGGA": 1537, + "TTGGGT": 1538, + "TTGGGC": 1539, + "TTGGGG": 1540, + "TCAAAA": 1541, + "TCAAAT": 1542, + "TCAAAC": 1543, + "TCAAAG": 1544, + "TCAATA": 1545, + "TCAATT": 1546, + "TCAATC": 1547, + "TCAATG": 1548, + "TCAACA": 1549, + "TCAACT": 1550, + "TCAACC": 1551, + "TCAACG": 1552, + "TCAAGA": 1553, + "TCAAGT": 1554, + "TCAAGC": 1555, + "TCAAGG": 1556, + "TCATAA": 1557, + "TCATAT": 1558, + "TCATAC": 1559, + "TCATAG": 1560, + "TCATTA": 1561, + "TCATTT": 1562, + "TCATTC": 1563, + "TCATTG": 1564, + "TCATCA": 1565, + "TCATCT": 1566, + "TCATCC": 1567, + "TCATCG": 1568, + "TCATGA": 1569, + "TCATGT": 1570, + "TCATGC": 1571, + "TCATGG": 1572, + "TCACAA": 1573, + "TCACAT": 1574, + "TCACAC": 1575, + "TCACAG": 1576, + "TCACTA": 1577, + "TCACTT": 1578, + "TCACTC": 1579, + "TCACTG": 1580, + "TCACCA": 1581, + "TCACCT": 1582, + "TCACCC": 1583, + "TCACCG": 1584, + "TCACGA": 1585, + "TCACGT": 1586, + "TCACGC": 1587, + "TCACGG": 1588, + "TCAGAA": 1589, + "TCAGAT": 1590, + "TCAGAC": 1591, + "TCAGAG": 1592, + "TCAGTA": 1593, + "TCAGTT": 1594, + "TCAGTC": 1595, + "TCAGTG": 1596, + "TCAGCA": 1597, + "TCAGCT": 1598, + "TCAGCC": 1599, + "TCAGCG": 1600, + "TCAGGA": 1601, + "TCAGGT": 1602, + "TCAGGC": 1603, + "TCAGGG": 1604, + "TCTAAA": 1605, + "TCTAAT": 1606, + "TCTAAC": 1607, + "TCTAAG": 1608, + "TCTATA": 1609, + "TCTATT": 1610, + "TCTATC": 1611, + "TCTATG": 1612, + "TCTACA": 1613, + "TCTACT": 1614, + "TCTACC": 1615, + "TCTACG": 1616, + "TCTAGA": 1617, + "TCTAGT": 1618, + "TCTAGC": 1619, + "TCTAGG": 1620, + "TCTTAA": 1621, + "TCTTAT": 1622, + "TCTTAC": 1623, + "TCTTAG": 1624, + "TCTTTA": 1625, + "TCTTTT": 1626, + "TCTTTC": 1627, + "TCTTTG": 1628, + "TCTTCA": 1629, + "TCTTCT": 1630, + "TCTTCC": 1631, + "TCTTCG": 1632, + "TCTTGA": 1633, + "TCTTGT": 1634, + "TCTTGC": 1635, + "TCTTGG": 1636, + "TCTCAA": 1637, + "TCTCAT": 1638, + "TCTCAC": 1639, + "TCTCAG": 1640, + "TCTCTA": 1641, + "TCTCTT": 1642, + "TCTCTC": 1643, + "TCTCTG": 1644, + "TCTCCA": 1645, + "TCTCCT": 1646, + "TCTCCC": 1647, + "TCTCCG": 1648, + "TCTCGA": 1649, + "TCTCGT": 1650, + "TCTCGC": 1651, + "TCTCGG": 1652, + "TCTGAA": 1653, + "TCTGAT": 1654, + "TCTGAC": 1655, + "TCTGAG": 1656, + "TCTGTA": 1657, + "TCTGTT": 1658, + "TCTGTC": 1659, + "TCTGTG": 1660, + "TCTGCA": 1661, + "TCTGCT": 1662, + "TCTGCC": 1663, + "TCTGCG": 1664, + "TCTGGA": 1665, + "TCTGGT": 1666, + "TCTGGC": 1667, + "TCTGGG": 1668, + "TCCAAA": 1669, + "TCCAAT": 1670, + "TCCAAC": 1671, + "TCCAAG": 1672, + "TCCATA": 1673, + "TCCATT": 1674, + "TCCATC": 1675, + "TCCATG": 1676, + "TCCACA": 1677, + "TCCACT": 1678, + "TCCACC": 1679, + "TCCACG": 1680, + "TCCAGA": 1681, + "TCCAGT": 1682, + "TCCAGC": 1683, + "TCCAGG": 1684, + "TCCTAA": 1685, + "TCCTAT": 1686, + "TCCTAC": 1687, + "TCCTAG": 1688, + "TCCTTA": 1689, + "TCCTTT": 1690, + "TCCTTC": 1691, + "TCCTTG": 1692, + "TCCTCA": 1693, + "TCCTCT": 1694, + "TCCTCC": 1695, + "TCCTCG": 1696, + "TCCTGA": 1697, + "TCCTGT": 1698, + "TCCTGC": 1699, + "TCCTGG": 1700, + "TCCCAA": 1701, + "TCCCAT": 1702, + "TCCCAC": 1703, + "TCCCAG": 1704, + "TCCCTA": 1705, + "TCCCTT": 1706, + "TCCCTC": 1707, + "TCCCTG": 1708, + "TCCCCA": 1709, + "TCCCCT": 1710, + "TCCCCC": 1711, + "TCCCCG": 1712, + "TCCCGA": 1713, + "TCCCGT": 1714, + "TCCCGC": 1715, + "TCCCGG": 1716, + "TCCGAA": 1717, + "TCCGAT": 1718, + "TCCGAC": 1719, + "TCCGAG": 1720, + "TCCGTA": 1721, + "TCCGTT": 1722, + "TCCGTC": 1723, + "TCCGTG": 1724, + "TCCGCA": 1725, + "TCCGCT": 1726, + "TCCGCC": 1727, + "TCCGCG": 1728, + "TCCGGA": 1729, + "TCCGGT": 1730, + "TCCGGC": 1731, + "TCCGGG": 1732, + "TCGAAA": 1733, + "TCGAAT": 1734, + "TCGAAC": 1735, + "TCGAAG": 1736, + "TCGATA": 1737, + "TCGATT": 1738, + "TCGATC": 1739, + "TCGATG": 1740, + "TCGACA": 1741, + "TCGACT": 1742, + "TCGACC": 1743, + "TCGACG": 1744, + "TCGAGA": 1745, + "TCGAGT": 1746, + "TCGAGC": 1747, + "TCGAGG": 1748, + "TCGTAA": 1749, + "TCGTAT": 1750, + "TCGTAC": 1751, + "TCGTAG": 1752, + "TCGTTA": 1753, + "TCGTTT": 1754, + "TCGTTC": 1755, + "TCGTTG": 1756, + "TCGTCA": 1757, + "TCGTCT": 1758, + "TCGTCC": 1759, + "TCGTCG": 1760, + "TCGTGA": 1761, + "TCGTGT": 1762, + "TCGTGC": 1763, + "TCGTGG": 1764, + "TCGCAA": 1765, + "TCGCAT": 1766, + "TCGCAC": 1767, + "TCGCAG": 1768, + "TCGCTA": 1769, + "TCGCTT": 1770, + "TCGCTC": 1771, + "TCGCTG": 1772, + "TCGCCA": 1773, + "TCGCCT": 1774, + "TCGCCC": 1775, + "TCGCCG": 1776, + "TCGCGA": 1777, + "TCGCGT": 1778, + "TCGCGC": 1779, + "TCGCGG": 1780, + "TCGGAA": 1781, + "TCGGAT": 1782, + "TCGGAC": 1783, + "TCGGAG": 1784, + "TCGGTA": 1785, + "TCGGTT": 1786, + "TCGGTC": 1787, + "TCGGTG": 1788, + "TCGGCA": 1789, + "TCGGCT": 1790, + "TCGGCC": 1791, + "TCGGCG": 1792, + "TCGGGA": 1793, + "TCGGGT": 1794, + "TCGGGC": 1795, + "TCGGGG": 1796, + "TGAAAA": 1797, + "TGAAAT": 1798, + "TGAAAC": 1799, + "TGAAAG": 1800, + "TGAATA": 1801, + "TGAATT": 1802, + "TGAATC": 1803, + "TGAATG": 1804, + "TGAACA": 1805, + "TGAACT": 1806, + "TGAACC": 1807, + "TGAACG": 1808, + "TGAAGA": 1809, + "TGAAGT": 1810, + "TGAAGC": 1811, + "TGAAGG": 1812, + "TGATAA": 1813, + "TGATAT": 1814, + "TGATAC": 1815, + "TGATAG": 1816, + "TGATTA": 1817, + "TGATTT": 1818, + "TGATTC": 1819, + "TGATTG": 1820, + "TGATCA": 1821, + "TGATCT": 1822, + "TGATCC": 1823, + "TGATCG": 1824, + "TGATGA": 1825, + "TGATGT": 1826, + "TGATGC": 1827, + "TGATGG": 1828, + "TGACAA": 1829, + "TGACAT": 1830, + "TGACAC": 1831, + "TGACAG": 1832, + "TGACTA": 1833, + "TGACTT": 1834, + "TGACTC": 1835, + "TGACTG": 1836, + "TGACCA": 1837, + "TGACCT": 1838, + "TGACCC": 1839, + "TGACCG": 1840, + "TGACGA": 1841, + "TGACGT": 1842, + "TGACGC": 1843, + "TGACGG": 1844, + "TGAGAA": 1845, + "TGAGAT": 1846, + "TGAGAC": 1847, + "TGAGAG": 1848, + "TGAGTA": 1849, + "TGAGTT": 1850, + "TGAGTC": 1851, + "TGAGTG": 1852, + "TGAGCA": 1853, + "TGAGCT": 1854, + "TGAGCC": 1855, + "TGAGCG": 1856, + "TGAGGA": 1857, + "TGAGGT": 1858, + "TGAGGC": 1859, + "TGAGGG": 1860, + "TGTAAA": 1861, + "TGTAAT": 1862, + "TGTAAC": 1863, + "TGTAAG": 1864, + "TGTATA": 1865, + "TGTATT": 1866, + "TGTATC": 1867, + "TGTATG": 1868, + "TGTACA": 1869, + "TGTACT": 1870, + "TGTACC": 1871, + "TGTACG": 1872, + "TGTAGA": 1873, + "TGTAGT": 1874, + "TGTAGC": 1875, + "TGTAGG": 1876, + "TGTTAA": 1877, + "TGTTAT": 1878, + "TGTTAC": 1879, + "TGTTAG": 1880, + "TGTTTA": 1881, + "TGTTTT": 1882, + "TGTTTC": 1883, + "TGTTTG": 1884, + "TGTTCA": 1885, + "TGTTCT": 1886, + "TGTTCC": 1887, + "TGTTCG": 1888, + "TGTTGA": 1889, + "TGTTGT": 1890, + "TGTTGC": 1891, + "TGTTGG": 1892, + "TGTCAA": 1893, + "TGTCAT": 1894, + "TGTCAC": 1895, + "TGTCAG": 1896, + "TGTCTA": 1897, + "TGTCTT": 1898, + "TGTCTC": 1899, + "TGTCTG": 1900, + "TGTCCA": 1901, + "TGTCCT": 1902, + "TGTCCC": 1903, + "TGTCCG": 1904, + "TGTCGA": 1905, + "TGTCGT": 1906, + "TGTCGC": 1907, + "TGTCGG": 1908, + "TGTGAA": 1909, + "TGTGAT": 1910, + "TGTGAC": 1911, + "TGTGAG": 1912, + "TGTGTA": 1913, + "TGTGTT": 1914, + "TGTGTC": 1915, + "TGTGTG": 1916, + "TGTGCA": 1917, + "TGTGCT": 1918, + "TGTGCC": 1919, + "TGTGCG": 1920, + "TGTGGA": 1921, + "TGTGGT": 1922, + "TGTGGC": 1923, + "TGTGGG": 1924, + "TGCAAA": 1925, + "TGCAAT": 1926, + "TGCAAC": 1927, + "TGCAAG": 1928, + "TGCATA": 1929, + "TGCATT": 1930, + "TGCATC": 1931, + "TGCATG": 1932, + "TGCACA": 1933, + "TGCACT": 1934, + "TGCACC": 1935, + "TGCACG": 1936, + "TGCAGA": 1937, + "TGCAGT": 1938, + "TGCAGC": 1939, + "TGCAGG": 1940, + "TGCTAA": 1941, + "TGCTAT": 1942, + "TGCTAC": 1943, + "TGCTAG": 1944, + "TGCTTA": 1945, + "TGCTTT": 1946, + "TGCTTC": 1947, + "TGCTTG": 1948, + "TGCTCA": 1949, + "TGCTCT": 1950, + "TGCTCC": 1951, + "TGCTCG": 1952, + "TGCTGA": 1953, + "TGCTGT": 1954, + "TGCTGC": 1955, + "TGCTGG": 1956, + "TGCCAA": 1957, + "TGCCAT": 1958, + "TGCCAC": 1959, + "TGCCAG": 1960, + "TGCCTA": 1961, + "TGCCTT": 1962, + "TGCCTC": 1963, + "TGCCTG": 1964, + "TGCCCA": 1965, + "TGCCCT": 1966, + "TGCCCC": 1967, + "TGCCCG": 1968, + "TGCCGA": 1969, + "TGCCGT": 1970, + "TGCCGC": 1971, + "TGCCGG": 1972, + "TGCGAA": 1973, + "TGCGAT": 1974, + "TGCGAC": 1975, + "TGCGAG": 1976, + "TGCGTA": 1977, + "TGCGTT": 1978, + "TGCGTC": 1979, + "TGCGTG": 1980, + "TGCGCA": 1981, + "TGCGCT": 1982, + "TGCGCC": 1983, + "TGCGCG": 1984, + "TGCGGA": 1985, + "TGCGGT": 1986, + "TGCGGC": 1987, + "TGCGGG": 1988, + "TGGAAA": 1989, + "TGGAAT": 1990, + "TGGAAC": 1991, + "TGGAAG": 1992, + "TGGATA": 1993, + "TGGATT": 1994, + "TGGATC": 1995, + "TGGATG": 1996, + "TGGACA": 1997, + "TGGACT": 1998, + "TGGACC": 1999, + "TGGACG": 2000, + "TGGAGA": 2001, + "TGGAGT": 2002, + "TGGAGC": 2003, + "TGGAGG": 2004, + "TGGTAA": 2005, + "TGGTAT": 2006, + "TGGTAC": 2007, + "TGGTAG": 2008, + "TGGTTA": 2009, + "TGGTTT": 2010, + "TGGTTC": 2011, + "TGGTTG": 2012, + "TGGTCA": 2013, + "TGGTCT": 2014, + "TGGTCC": 2015, + "TGGTCG": 2016, + "TGGTGA": 2017, + "TGGTGT": 2018, + "TGGTGC": 2019, + "TGGTGG": 2020, + "TGGCAA": 2021, + "TGGCAT": 2022, + "TGGCAC": 2023, + "TGGCAG": 2024, + "TGGCTA": 2025, + "TGGCTT": 2026, + "TGGCTC": 2027, + "TGGCTG": 2028, + "TGGCCA": 2029, + "TGGCCT": 2030, + "TGGCCC": 2031, + "TGGCCG": 2032, + "TGGCGA": 2033, + "TGGCGT": 2034, + "TGGCGC": 2035, + "TGGCGG": 2036, + "TGGGAA": 2037, + "TGGGAT": 2038, + "TGGGAC": 2039, + "TGGGAG": 2040, + "TGGGTA": 2041, + "TGGGTT": 2042, + "TGGGTC": 2043, + "TGGGTG": 2044, + "TGGGCA": 2045, + "TGGGCT": 2046, + "TGGGCC": 2047, + "TGGGCG": 2048, + "TGGGGA": 2049, + "TGGGGT": 2050, + "TGGGGC": 2051, + "TGGGGG": 2052, + "CAAAAA": 2053, + "CAAAAT": 2054, + "CAAAAC": 2055, + "CAAAAG": 2056, + "CAAATA": 2057, + "CAAATT": 2058, + "CAAATC": 2059, + "CAAATG": 2060, + "CAAACA": 2061, + "CAAACT": 2062, + "CAAACC": 2063, + "CAAACG": 2064, + "CAAAGA": 2065, + "CAAAGT": 2066, + "CAAAGC": 2067, + "CAAAGG": 2068, + "CAATAA": 2069, + "CAATAT": 2070, + "CAATAC": 2071, + "CAATAG": 2072, + "CAATTA": 2073, + "CAATTT": 2074, + "CAATTC": 2075, + "CAATTG": 2076, + "CAATCA": 2077, + "CAATCT": 2078, + "CAATCC": 2079, + "CAATCG": 2080, + "CAATGA": 2081, + "CAATGT": 2082, + "CAATGC": 2083, + "CAATGG": 2084, + "CAACAA": 2085, + "CAACAT": 2086, + "CAACAC": 2087, + "CAACAG": 2088, + "CAACTA": 2089, + "CAACTT": 2090, + "CAACTC": 2091, + "CAACTG": 2092, + "CAACCA": 2093, + "CAACCT": 2094, + "CAACCC": 2095, + "CAACCG": 2096, + "CAACGA": 2097, + "CAACGT": 2098, + "CAACGC": 2099, + "CAACGG": 2100, + "CAAGAA": 2101, + "CAAGAT": 2102, + "CAAGAC": 2103, + "CAAGAG": 2104, + "CAAGTA": 2105, + "CAAGTT": 2106, + "CAAGTC": 2107, + "CAAGTG": 2108, + "CAAGCA": 2109, + "CAAGCT": 2110, + "CAAGCC": 2111, + "CAAGCG": 2112, + "CAAGGA": 2113, + "CAAGGT": 2114, + "CAAGGC": 2115, + "CAAGGG": 2116, + "CATAAA": 2117, + "CATAAT": 2118, + "CATAAC": 2119, + "CATAAG": 2120, + "CATATA": 2121, + "CATATT": 2122, + "CATATC": 2123, + "CATATG": 2124, + "CATACA": 2125, + "CATACT": 2126, + "CATACC": 2127, + "CATACG": 2128, + "CATAGA": 2129, + "CATAGT": 2130, + "CATAGC": 2131, + "CATAGG": 2132, + "CATTAA": 2133, + "CATTAT": 2134, + "CATTAC": 2135, + "CATTAG": 2136, + "CATTTA": 2137, + "CATTTT": 2138, + "CATTTC": 2139, + "CATTTG": 2140, + "CATTCA": 2141, + "CATTCT": 2142, + "CATTCC": 2143, + "CATTCG": 2144, + "CATTGA": 2145, + "CATTGT": 2146, + "CATTGC": 2147, + "CATTGG": 2148, + "CATCAA": 2149, + "CATCAT": 2150, + "CATCAC": 2151, + "CATCAG": 2152, + "CATCTA": 2153, + "CATCTT": 2154, + "CATCTC": 2155, + "CATCTG": 2156, + "CATCCA": 2157, + "CATCCT": 2158, + "CATCCC": 2159, + "CATCCG": 2160, + "CATCGA": 2161, + "CATCGT": 2162, + "CATCGC": 2163, + "CATCGG": 2164, + "CATGAA": 2165, + "CATGAT": 2166, + "CATGAC": 2167, + "CATGAG": 2168, + "CATGTA": 2169, + "CATGTT": 2170, + "CATGTC": 2171, + "CATGTG": 2172, + "CATGCA": 2173, + "CATGCT": 2174, + "CATGCC": 2175, + "CATGCG": 2176, + "CATGGA": 2177, + "CATGGT": 2178, + "CATGGC": 2179, + "CATGGG": 2180, + "CACAAA": 2181, + "CACAAT": 2182, + "CACAAC": 2183, + "CACAAG": 2184, + "CACATA": 2185, + "CACATT": 2186, + "CACATC": 2187, + "CACATG": 2188, + "CACACA": 2189, + "CACACT": 2190, + "CACACC": 2191, + "CACACG": 2192, + "CACAGA": 2193, + "CACAGT": 2194, + "CACAGC": 2195, + "CACAGG": 2196, + "CACTAA": 2197, + "CACTAT": 2198, + "CACTAC": 2199, + "CACTAG": 2200, + "CACTTA": 2201, + "CACTTT": 2202, + "CACTTC": 2203, + "CACTTG": 2204, + "CACTCA": 2205, + "CACTCT": 2206, + "CACTCC": 2207, + "CACTCG": 2208, + "CACTGA": 2209, + "CACTGT": 2210, + "CACTGC": 2211, + "CACTGG": 2212, + "CACCAA": 2213, + "CACCAT": 2214, + "CACCAC": 2215, + "CACCAG": 2216, + "CACCTA": 2217, + "CACCTT": 2218, + "CACCTC": 2219, + "CACCTG": 2220, + "CACCCA": 2221, + "CACCCT": 2222, + "CACCCC": 2223, + "CACCCG": 2224, + "CACCGA": 2225, + "CACCGT": 2226, + "CACCGC": 2227, + "CACCGG": 2228, + "CACGAA": 2229, + "CACGAT": 2230, + "CACGAC": 2231, + "CACGAG": 2232, + "CACGTA": 2233, + "CACGTT": 2234, + "CACGTC": 2235, + "CACGTG": 2236, + "CACGCA": 2237, + "CACGCT": 2238, + "CACGCC": 2239, + "CACGCG": 2240, + "CACGGA": 2241, + "CACGGT": 2242, + "CACGGC": 2243, + "CACGGG": 2244, + "CAGAAA": 2245, + "CAGAAT": 2246, + "CAGAAC": 2247, + "CAGAAG": 2248, + "CAGATA": 2249, + "CAGATT": 2250, + "CAGATC": 2251, + "CAGATG": 2252, + "CAGACA": 2253, + "CAGACT": 2254, + "CAGACC": 2255, + "CAGACG": 2256, + "CAGAGA": 2257, + "CAGAGT": 2258, + "CAGAGC": 2259, + "CAGAGG": 2260, + "CAGTAA": 2261, + "CAGTAT": 2262, + "CAGTAC": 2263, + "CAGTAG": 2264, + "CAGTTA": 2265, + "CAGTTT": 2266, + "CAGTTC": 2267, + "CAGTTG": 2268, + "CAGTCA": 2269, + "CAGTCT": 2270, + "CAGTCC": 2271, + "CAGTCG": 2272, + "CAGTGA": 2273, + "CAGTGT": 2274, + "CAGTGC": 2275, + "CAGTGG": 2276, + "CAGCAA": 2277, + "CAGCAT": 2278, + "CAGCAC": 2279, + "CAGCAG": 2280, + "CAGCTA": 2281, + "CAGCTT": 2282, + "CAGCTC": 2283, + "CAGCTG": 2284, + "CAGCCA": 2285, + "CAGCCT": 2286, + "CAGCCC": 2287, + "CAGCCG": 2288, + "CAGCGA": 2289, + "CAGCGT": 2290, + "CAGCGC": 2291, + "CAGCGG": 2292, + "CAGGAA": 2293, + "CAGGAT": 2294, + "CAGGAC": 2295, + "CAGGAG": 2296, + "CAGGTA": 2297, + "CAGGTT": 2298, + "CAGGTC": 2299, + "CAGGTG": 2300, + "CAGGCA": 2301, + "CAGGCT": 2302, + "CAGGCC": 2303, + "CAGGCG": 2304, + "CAGGGA": 2305, + "CAGGGT": 2306, + "CAGGGC": 2307, + "CAGGGG": 2308, + "CTAAAA": 2309, + "CTAAAT": 2310, + "CTAAAC": 2311, + "CTAAAG": 2312, + "CTAATA": 2313, + "CTAATT": 2314, + "CTAATC": 2315, + "CTAATG": 2316, + "CTAACA": 2317, + "CTAACT": 2318, + "CTAACC": 2319, + "CTAACG": 2320, + "CTAAGA": 2321, + "CTAAGT": 2322, + "CTAAGC": 2323, + "CTAAGG": 2324, + "CTATAA": 2325, + "CTATAT": 2326, + "CTATAC": 2327, + "CTATAG": 2328, + "CTATTA": 2329, + "CTATTT": 2330, + "CTATTC": 2331, + "CTATTG": 2332, + "CTATCA": 2333, + "CTATCT": 2334, + "CTATCC": 2335, + "CTATCG": 2336, + "CTATGA": 2337, + "CTATGT": 2338, + "CTATGC": 2339, + "CTATGG": 2340, + "CTACAA": 2341, + "CTACAT": 2342, + "CTACAC": 2343, + "CTACAG": 2344, + "CTACTA": 2345, + "CTACTT": 2346, + "CTACTC": 2347, + "CTACTG": 2348, + "CTACCA": 2349, + "CTACCT": 2350, + "CTACCC": 2351, + "CTACCG": 2352, + "CTACGA": 2353, + "CTACGT": 2354, + "CTACGC": 2355, + "CTACGG": 2356, + "CTAGAA": 2357, + "CTAGAT": 2358, + "CTAGAC": 2359, + "CTAGAG": 2360, + "CTAGTA": 2361, + "CTAGTT": 2362, + "CTAGTC": 2363, + "CTAGTG": 2364, + "CTAGCA": 2365, + "CTAGCT": 2366, + "CTAGCC": 2367, + "CTAGCG": 2368, + "CTAGGA": 2369, + "CTAGGT": 2370, + "CTAGGC": 2371, + "CTAGGG": 2372, + "CTTAAA": 2373, + "CTTAAT": 2374, + "CTTAAC": 2375, + "CTTAAG": 2376, + "CTTATA": 2377, + "CTTATT": 2378, + "CTTATC": 2379, + "CTTATG": 2380, + "CTTACA": 2381, + "CTTACT": 2382, + "CTTACC": 2383, + "CTTACG": 2384, + "CTTAGA": 2385, + "CTTAGT": 2386, + "CTTAGC": 2387, + "CTTAGG": 2388, + "CTTTAA": 2389, + "CTTTAT": 2390, + "CTTTAC": 2391, + "CTTTAG": 2392, + "CTTTTA": 2393, + "CTTTTT": 2394, + "CTTTTC": 2395, + "CTTTTG": 2396, + "CTTTCA": 2397, + "CTTTCT": 2398, + "CTTTCC": 2399, + "CTTTCG": 2400, + "CTTTGA": 2401, + "CTTTGT": 2402, + "CTTTGC": 2403, + "CTTTGG": 2404, + "CTTCAA": 2405, + "CTTCAT": 2406, + "CTTCAC": 2407, + "CTTCAG": 2408, + "CTTCTA": 2409, + "CTTCTT": 2410, + "CTTCTC": 2411, + "CTTCTG": 2412, + "CTTCCA": 2413, + "CTTCCT": 2414, + "CTTCCC": 2415, + "CTTCCG": 2416, + "CTTCGA": 2417, + "CTTCGT": 2418, + "CTTCGC": 2419, + "CTTCGG": 2420, + "CTTGAA": 2421, + "CTTGAT": 2422, + "CTTGAC": 2423, + "CTTGAG": 2424, + "CTTGTA": 2425, + "CTTGTT": 2426, + "CTTGTC": 2427, + "CTTGTG": 2428, + "CTTGCA": 2429, + "CTTGCT": 2430, + "CTTGCC": 2431, + "CTTGCG": 2432, + "CTTGGA": 2433, + "CTTGGT": 2434, + "CTTGGC": 2435, + "CTTGGG": 2436, + "CTCAAA": 2437, + "CTCAAT": 2438, + "CTCAAC": 2439, + "CTCAAG": 2440, + "CTCATA": 2441, + "CTCATT": 2442, + "CTCATC": 2443, + "CTCATG": 2444, + "CTCACA": 2445, + "CTCACT": 2446, + "CTCACC": 2447, + "CTCACG": 2448, + "CTCAGA": 2449, + "CTCAGT": 2450, + "CTCAGC": 2451, + "CTCAGG": 2452, + "CTCTAA": 2453, + "CTCTAT": 2454, + "CTCTAC": 2455, + "CTCTAG": 2456, + "CTCTTA": 2457, + "CTCTTT": 2458, + "CTCTTC": 2459, + "CTCTTG": 2460, + "CTCTCA": 2461, + "CTCTCT": 2462, + "CTCTCC": 2463, + "CTCTCG": 2464, + "CTCTGA": 2465, + "CTCTGT": 2466, + "CTCTGC": 2467, + "CTCTGG": 2468, + "CTCCAA": 2469, + "CTCCAT": 2470, + "CTCCAC": 2471, + "CTCCAG": 2472, + "CTCCTA": 2473, + "CTCCTT": 2474, + "CTCCTC": 2475, + "CTCCTG": 2476, + "CTCCCA": 2477, + "CTCCCT": 2478, + "CTCCCC": 2479, + "CTCCCG": 2480, + "CTCCGA": 2481, + "CTCCGT": 2482, + "CTCCGC": 2483, + "CTCCGG": 2484, + "CTCGAA": 2485, + "CTCGAT": 2486, + "CTCGAC": 2487, + "CTCGAG": 2488, + "CTCGTA": 2489, + "CTCGTT": 2490, + "CTCGTC": 2491, + "CTCGTG": 2492, + "CTCGCA": 2493, + "CTCGCT": 2494, + "CTCGCC": 2495, + "CTCGCG": 2496, + "CTCGGA": 2497, + "CTCGGT": 2498, + "CTCGGC": 2499, + "CTCGGG": 2500, + "CTGAAA": 2501, + "CTGAAT": 2502, + "CTGAAC": 2503, + "CTGAAG": 2504, + "CTGATA": 2505, + "CTGATT": 2506, + "CTGATC": 2507, + "CTGATG": 2508, + "CTGACA": 2509, + "CTGACT": 2510, + "CTGACC": 2511, + "CTGACG": 2512, + "CTGAGA": 2513, + "CTGAGT": 2514, + "CTGAGC": 2515, + "CTGAGG": 2516, + "CTGTAA": 2517, + "CTGTAT": 2518, + "CTGTAC": 2519, + "CTGTAG": 2520, + "CTGTTA": 2521, + "CTGTTT": 2522, + "CTGTTC": 2523, + "CTGTTG": 2524, + "CTGTCA": 2525, + "CTGTCT": 2526, + "CTGTCC": 2527, + "CTGTCG": 2528, + "CTGTGA": 2529, + "CTGTGT": 2530, + "CTGTGC": 2531, + "CTGTGG": 2532, + "CTGCAA": 2533, + "CTGCAT": 2534, + "CTGCAC": 2535, + "CTGCAG": 2536, + "CTGCTA": 2537, + "CTGCTT": 2538, + "CTGCTC": 2539, + "CTGCTG": 2540, + "CTGCCA": 2541, + "CTGCCT": 2542, + "CTGCCC": 2543, + "CTGCCG": 2544, + "CTGCGA": 2545, + "CTGCGT": 2546, + "CTGCGC": 2547, + "CTGCGG": 2548, + "CTGGAA": 2549, + "CTGGAT": 2550, + "CTGGAC": 2551, + "CTGGAG": 2552, + "CTGGTA": 2553, + "CTGGTT": 2554, + "CTGGTC": 2555, + "CTGGTG": 2556, + "CTGGCA": 2557, + "CTGGCT": 2558, + "CTGGCC": 2559, + "CTGGCG": 2560, + "CTGGGA": 2561, + "CTGGGT": 2562, + "CTGGGC": 2563, + "CTGGGG": 2564, + "CCAAAA": 2565, + "CCAAAT": 2566, + "CCAAAC": 2567, + "CCAAAG": 2568, + "CCAATA": 2569, + "CCAATT": 2570, + "CCAATC": 2571, + "CCAATG": 2572, + "CCAACA": 2573, + "CCAACT": 2574, + "CCAACC": 2575, + "CCAACG": 2576, + "CCAAGA": 2577, + "CCAAGT": 2578, + "CCAAGC": 2579, + "CCAAGG": 2580, + "CCATAA": 2581, + "CCATAT": 2582, + "CCATAC": 2583, + "CCATAG": 2584, + "CCATTA": 2585, + "CCATTT": 2586, + "CCATTC": 2587, + "CCATTG": 2588, + "CCATCA": 2589, + "CCATCT": 2590, + "CCATCC": 2591, + "CCATCG": 2592, + "CCATGA": 2593, + "CCATGT": 2594, + "CCATGC": 2595, + "CCATGG": 2596, + "CCACAA": 2597, + "CCACAT": 2598, + "CCACAC": 2599, + "CCACAG": 2600, + "CCACTA": 2601, + "CCACTT": 2602, + "CCACTC": 2603, + "CCACTG": 2604, + "CCACCA": 2605, + "CCACCT": 2606, + "CCACCC": 2607, + "CCACCG": 2608, + "CCACGA": 2609, + "CCACGT": 2610, + "CCACGC": 2611, + "CCACGG": 2612, + "CCAGAA": 2613, + "CCAGAT": 2614, + "CCAGAC": 2615, + "CCAGAG": 2616, + "CCAGTA": 2617, + "CCAGTT": 2618, + "CCAGTC": 2619, + "CCAGTG": 2620, + "CCAGCA": 2621, + "CCAGCT": 2622, + "CCAGCC": 2623, + "CCAGCG": 2624, + "CCAGGA": 2625, + "CCAGGT": 2626, + "CCAGGC": 2627, + "CCAGGG": 2628, + "CCTAAA": 2629, + "CCTAAT": 2630, + "CCTAAC": 2631, + "CCTAAG": 2632, + "CCTATA": 2633, + "CCTATT": 2634, + "CCTATC": 2635, + "CCTATG": 2636, + "CCTACA": 2637, + "CCTACT": 2638, + "CCTACC": 2639, + "CCTACG": 2640, + "CCTAGA": 2641, + "CCTAGT": 2642, + "CCTAGC": 2643, + "CCTAGG": 2644, + "CCTTAA": 2645, + "CCTTAT": 2646, + "CCTTAC": 2647, + "CCTTAG": 2648, + "CCTTTA": 2649, + "CCTTTT": 2650, + "CCTTTC": 2651, + "CCTTTG": 2652, + "CCTTCA": 2653, + "CCTTCT": 2654, + "CCTTCC": 2655, + "CCTTCG": 2656, + "CCTTGA": 2657, + "CCTTGT": 2658, + "CCTTGC": 2659, + "CCTTGG": 2660, + "CCTCAA": 2661, + "CCTCAT": 2662, + "CCTCAC": 2663, + "CCTCAG": 2664, + "CCTCTA": 2665, + "CCTCTT": 2666, + "CCTCTC": 2667, + "CCTCTG": 2668, + "CCTCCA": 2669, + "CCTCCT": 2670, + "CCTCCC": 2671, + "CCTCCG": 2672, + "CCTCGA": 2673, + "CCTCGT": 2674, + "CCTCGC": 2675, + "CCTCGG": 2676, + "CCTGAA": 2677, + "CCTGAT": 2678, + "CCTGAC": 2679, + "CCTGAG": 2680, + "CCTGTA": 2681, + "CCTGTT": 2682, + "CCTGTC": 2683, + "CCTGTG": 2684, + "CCTGCA": 2685, + "CCTGCT": 2686, + "CCTGCC": 2687, + "CCTGCG": 2688, + "CCTGGA": 2689, + "CCTGGT": 2690, + "CCTGGC": 2691, + "CCTGGG": 2692, + "CCCAAA": 2693, + "CCCAAT": 2694, + "CCCAAC": 2695, + "CCCAAG": 2696, + "CCCATA": 2697, + "CCCATT": 2698, + "CCCATC": 2699, + "CCCATG": 2700, + "CCCACA": 2701, + "CCCACT": 2702, + "CCCACC": 2703, + "CCCACG": 2704, + "CCCAGA": 2705, + "CCCAGT": 2706, + "CCCAGC": 2707, + "CCCAGG": 2708, + "CCCTAA": 2709, + "CCCTAT": 2710, + "CCCTAC": 2711, + "CCCTAG": 2712, + "CCCTTA": 2713, + "CCCTTT": 2714, + "CCCTTC": 2715, + "CCCTTG": 2716, + "CCCTCA": 2717, + "CCCTCT": 2718, + "CCCTCC": 2719, + "CCCTCG": 2720, + "CCCTGA": 2721, + "CCCTGT": 2722, + "CCCTGC": 2723, + "CCCTGG": 2724, + "CCCCAA": 2725, + "CCCCAT": 2726, + "CCCCAC": 2727, + "CCCCAG": 2728, + "CCCCTA": 2729, + "CCCCTT": 2730, + "CCCCTC": 2731, + "CCCCTG": 2732, + "CCCCCA": 2733, + "CCCCCT": 2734, + "CCCCCC": 2735, + "CCCCCG": 2736, + "CCCCGA": 2737, + "CCCCGT": 2738, + "CCCCGC": 2739, + "CCCCGG": 2740, + "CCCGAA": 2741, + "CCCGAT": 2742, + "CCCGAC": 2743, + "CCCGAG": 2744, + "CCCGTA": 2745, + "CCCGTT": 2746, + "CCCGTC": 2747, + "CCCGTG": 2748, + "CCCGCA": 2749, + "CCCGCT": 2750, + "CCCGCC": 2751, + "CCCGCG": 2752, + "CCCGGA": 2753, + "CCCGGT": 2754, + "CCCGGC": 2755, + "CCCGGG": 2756, + "CCGAAA": 2757, + "CCGAAT": 2758, + "CCGAAC": 2759, + "CCGAAG": 2760, + "CCGATA": 2761, + "CCGATT": 2762, + "CCGATC": 2763, + "CCGATG": 2764, + "CCGACA": 2765, + "CCGACT": 2766, + "CCGACC": 2767, + "CCGACG": 2768, + "CCGAGA": 2769, + "CCGAGT": 2770, + "CCGAGC": 2771, + "CCGAGG": 2772, + "CCGTAA": 2773, + "CCGTAT": 2774, + "CCGTAC": 2775, + "CCGTAG": 2776, + "CCGTTA": 2777, + "CCGTTT": 2778, + "CCGTTC": 2779, + "CCGTTG": 2780, + "CCGTCA": 2781, + "CCGTCT": 2782, + "CCGTCC": 2783, + "CCGTCG": 2784, + "CCGTGA": 2785, + "CCGTGT": 2786, + "CCGTGC": 2787, + "CCGTGG": 2788, + "CCGCAA": 2789, + "CCGCAT": 2790, + "CCGCAC": 2791, + "CCGCAG": 2792, + "CCGCTA": 2793, + "CCGCTT": 2794, + "CCGCTC": 2795, + "CCGCTG": 2796, + "CCGCCA": 2797, + "CCGCCT": 2798, + "CCGCCC": 2799, + "CCGCCG": 2800, + "CCGCGA": 2801, + "CCGCGT": 2802, + "CCGCGC": 2803, + "CCGCGG": 2804, + "CCGGAA": 2805, + "CCGGAT": 2806, + "CCGGAC": 2807, + "CCGGAG": 2808, + "CCGGTA": 2809, + "CCGGTT": 2810, + "CCGGTC": 2811, + "CCGGTG": 2812, + "CCGGCA": 2813, + "CCGGCT": 2814, + "CCGGCC": 2815, + "CCGGCG": 2816, + "CCGGGA": 2817, + "CCGGGT": 2818, + "CCGGGC": 2819, + "CCGGGG": 2820, + "CGAAAA": 2821, + "CGAAAT": 2822, + "CGAAAC": 2823, + "CGAAAG": 2824, + "CGAATA": 2825, + "CGAATT": 2826, + "CGAATC": 2827, + "CGAATG": 2828, + "CGAACA": 2829, + "CGAACT": 2830, + "CGAACC": 2831, + "CGAACG": 2832, + "CGAAGA": 2833, + "CGAAGT": 2834, + "CGAAGC": 2835, + "CGAAGG": 2836, + "CGATAA": 2837, + "CGATAT": 2838, + "CGATAC": 2839, + "CGATAG": 2840, + "CGATTA": 2841, + "CGATTT": 2842, + "CGATTC": 2843, + "CGATTG": 2844, + "CGATCA": 2845, + "CGATCT": 2846, + "CGATCC": 2847, + "CGATCG": 2848, + "CGATGA": 2849, + "CGATGT": 2850, + "CGATGC": 2851, + "CGATGG": 2852, + "CGACAA": 2853, + "CGACAT": 2854, + "CGACAC": 2855, + "CGACAG": 2856, + "CGACTA": 2857, + "CGACTT": 2858, + "CGACTC": 2859, + "CGACTG": 2860, + "CGACCA": 2861, + "CGACCT": 2862, + "CGACCC": 2863, + "CGACCG": 2864, + "CGACGA": 2865, + "CGACGT": 2866, + "CGACGC": 2867, + "CGACGG": 2868, + "CGAGAA": 2869, + "CGAGAT": 2870, + "CGAGAC": 2871, + "CGAGAG": 2872, + "CGAGTA": 2873, + "CGAGTT": 2874, + "CGAGTC": 2875, + "CGAGTG": 2876, + "CGAGCA": 2877, + "CGAGCT": 2878, + "CGAGCC": 2879, + "CGAGCG": 2880, + "CGAGGA": 2881, + "CGAGGT": 2882, + "CGAGGC": 2883, + "CGAGGG": 2884, + "CGTAAA": 2885, + "CGTAAT": 2886, + "CGTAAC": 2887, + "CGTAAG": 2888, + "CGTATA": 2889, + "CGTATT": 2890, + "CGTATC": 2891, + "CGTATG": 2892, + "CGTACA": 2893, + "CGTACT": 2894, + "CGTACC": 2895, + "CGTACG": 2896, + "CGTAGA": 2897, + "CGTAGT": 2898, + "CGTAGC": 2899, + "CGTAGG": 2900, + "CGTTAA": 2901, + "CGTTAT": 2902, + "CGTTAC": 2903, + "CGTTAG": 2904, + "CGTTTA": 2905, + "CGTTTT": 2906, + "CGTTTC": 2907, + "CGTTTG": 2908, + "CGTTCA": 2909, + "CGTTCT": 2910, + "CGTTCC": 2911, + "CGTTCG": 2912, + "CGTTGA": 2913, + "CGTTGT": 2914, + "CGTTGC": 2915, + "CGTTGG": 2916, + "CGTCAA": 2917, + "CGTCAT": 2918, + "CGTCAC": 2919, + "CGTCAG": 2920, + "CGTCTA": 2921, + "CGTCTT": 2922, + "CGTCTC": 2923, + "CGTCTG": 2924, + "CGTCCA": 2925, + "CGTCCT": 2926, + "CGTCCC": 2927, + "CGTCCG": 2928, + "CGTCGA": 2929, + "CGTCGT": 2930, + "CGTCGC": 2931, + "CGTCGG": 2932, + "CGTGAA": 2933, + "CGTGAT": 2934, + "CGTGAC": 2935, + "CGTGAG": 2936, + "CGTGTA": 2937, + "CGTGTT": 2938, + "CGTGTC": 2939, + "CGTGTG": 2940, + "CGTGCA": 2941, + "CGTGCT": 2942, + "CGTGCC": 2943, + "CGTGCG": 2944, + "CGTGGA": 2945, + "CGTGGT": 2946, + "CGTGGC": 2947, + "CGTGGG": 2948, + "CGCAAA": 2949, + "CGCAAT": 2950, + "CGCAAC": 2951, + "CGCAAG": 2952, + "CGCATA": 2953, + "CGCATT": 2954, + "CGCATC": 2955, + "CGCATG": 2956, + "CGCACA": 2957, + "CGCACT": 2958, + "CGCACC": 2959, + "CGCACG": 2960, + "CGCAGA": 2961, + "CGCAGT": 2962, + "CGCAGC": 2963, + "CGCAGG": 2964, + "CGCTAA": 2965, + "CGCTAT": 2966, + "CGCTAC": 2967, + "CGCTAG": 2968, + "CGCTTA": 2969, + "CGCTTT": 2970, + "CGCTTC": 2971, + "CGCTTG": 2972, + "CGCTCA": 2973, + "CGCTCT": 2974, + "CGCTCC": 2975, + "CGCTCG": 2976, + "CGCTGA": 2977, + "CGCTGT": 2978, + "CGCTGC": 2979, + "CGCTGG": 2980, + "CGCCAA": 2981, + "CGCCAT": 2982, + "CGCCAC": 2983, + "CGCCAG": 2984, + "CGCCTA": 2985, + "CGCCTT": 2986, + "CGCCTC": 2987, + "CGCCTG": 2988, + "CGCCCA": 2989, + "CGCCCT": 2990, + "CGCCCC": 2991, + "CGCCCG": 2992, + "CGCCGA": 2993, + "CGCCGT": 2994, + "CGCCGC": 2995, + "CGCCGG": 2996, + "CGCGAA": 2997, + "CGCGAT": 2998, + "CGCGAC": 2999, + "CGCGAG": 3000, + "CGCGTA": 3001, + "CGCGTT": 3002, + "CGCGTC": 3003, + "CGCGTG": 3004, + "CGCGCA": 3005, + "CGCGCT": 3006, + "CGCGCC": 3007, + "CGCGCG": 3008, + "CGCGGA": 3009, + "CGCGGT": 3010, + "CGCGGC": 3011, + "CGCGGG": 3012, + "CGGAAA": 3013, + "CGGAAT": 3014, + "CGGAAC": 3015, + "CGGAAG": 3016, + "CGGATA": 3017, + "CGGATT": 3018, + "CGGATC": 3019, + "CGGATG": 3020, + "CGGACA": 3021, + "CGGACT": 3022, + "CGGACC": 3023, + "CGGACG": 3024, + "CGGAGA": 3025, + "CGGAGT": 3026, + "CGGAGC": 3027, + "CGGAGG": 3028, + "CGGTAA": 3029, + "CGGTAT": 3030, + "CGGTAC": 3031, + "CGGTAG": 3032, + "CGGTTA": 3033, + "CGGTTT": 3034, + "CGGTTC": 3035, + "CGGTTG": 3036, + "CGGTCA": 3037, + "CGGTCT": 3038, + "CGGTCC": 3039, + "CGGTCG": 3040, + "CGGTGA": 3041, + "CGGTGT": 3042, + "CGGTGC": 3043, + "CGGTGG": 3044, + "CGGCAA": 3045, + "CGGCAT": 3046, + "CGGCAC": 3047, + "CGGCAG": 3048, + "CGGCTA": 3049, + "CGGCTT": 3050, + "CGGCTC": 3051, + "CGGCTG": 3052, + "CGGCCA": 3053, + "CGGCCT": 3054, + "CGGCCC": 3055, + "CGGCCG": 3056, + "CGGCGA": 3057, + "CGGCGT": 3058, + "CGGCGC": 3059, + "CGGCGG": 3060, + "CGGGAA": 3061, + "CGGGAT": 3062, + "CGGGAC": 3063, + "CGGGAG": 3064, + "CGGGTA": 3065, + "CGGGTT": 3066, + "CGGGTC": 3067, + "CGGGTG": 3068, + "CGGGCA": 3069, + "CGGGCT": 3070, + "CGGGCC": 3071, + "CGGGCG": 3072, + "CGGGGA": 3073, + "CGGGGT": 3074, + "CGGGGC": 3075, + "CGGGGG": 3076, + "GAAAAA": 3077, + "GAAAAT": 3078, + "GAAAAC": 3079, + "GAAAAG": 3080, + "GAAATA": 3081, + "GAAATT": 3082, + "GAAATC": 3083, + "GAAATG": 3084, + "GAAACA": 3085, + "GAAACT": 3086, + "GAAACC": 3087, + "GAAACG": 3088, + "GAAAGA": 3089, + "GAAAGT": 3090, + "GAAAGC": 3091, + "GAAAGG": 3092, + "GAATAA": 3093, + "GAATAT": 3094, + "GAATAC": 3095, + "GAATAG": 3096, + "GAATTA": 3097, + "GAATTT": 3098, + "GAATTC": 3099, + "GAATTG": 3100, + "GAATCA": 3101, + "GAATCT": 3102, + "GAATCC": 3103, + "GAATCG": 3104, + "GAATGA": 3105, + "GAATGT": 3106, + "GAATGC": 3107, + "GAATGG": 3108, + "GAACAA": 3109, + "GAACAT": 3110, + "GAACAC": 3111, + "GAACAG": 3112, + "GAACTA": 3113, + "GAACTT": 3114, + "GAACTC": 3115, + "GAACTG": 3116, + "GAACCA": 3117, + "GAACCT": 3118, + "GAACCC": 3119, + "GAACCG": 3120, + "GAACGA": 3121, + "GAACGT": 3122, + "GAACGC": 3123, + "GAACGG": 3124, + "GAAGAA": 3125, + "GAAGAT": 3126, + "GAAGAC": 3127, + "GAAGAG": 3128, + "GAAGTA": 3129, + "GAAGTT": 3130, + "GAAGTC": 3131, + "GAAGTG": 3132, + "GAAGCA": 3133, + "GAAGCT": 3134, + "GAAGCC": 3135, + "GAAGCG": 3136, + "GAAGGA": 3137, + "GAAGGT": 3138, + "GAAGGC": 3139, + "GAAGGG": 3140, + "GATAAA": 3141, + "GATAAT": 3142, + "GATAAC": 3143, + "GATAAG": 3144, + "GATATA": 3145, + "GATATT": 3146, + "GATATC": 3147, + "GATATG": 3148, + "GATACA": 3149, + "GATACT": 3150, + "GATACC": 3151, + "GATACG": 3152, + "GATAGA": 3153, + "GATAGT": 3154, + "GATAGC": 3155, + "GATAGG": 3156, + "GATTAA": 3157, + "GATTAT": 3158, + "GATTAC": 3159, + "GATTAG": 3160, + "GATTTA": 3161, + "GATTTT": 3162, + "GATTTC": 3163, + "GATTTG": 3164, + "GATTCA": 3165, + "GATTCT": 3166, + "GATTCC": 3167, + "GATTCG": 3168, + "GATTGA": 3169, + "GATTGT": 3170, + "GATTGC": 3171, + "GATTGG": 3172, + "GATCAA": 3173, + "GATCAT": 3174, + "GATCAC": 3175, + "GATCAG": 3176, + "GATCTA": 3177, + "GATCTT": 3178, + "GATCTC": 3179, + "GATCTG": 3180, + "GATCCA": 3181, + "GATCCT": 3182, + "GATCCC": 3183, + "GATCCG": 3184, + "GATCGA": 3185, + "GATCGT": 3186, + "GATCGC": 3187, + "GATCGG": 3188, + "GATGAA": 3189, + "GATGAT": 3190, + "GATGAC": 3191, + "GATGAG": 3192, + "GATGTA": 3193, + "GATGTT": 3194, + "GATGTC": 3195, + "GATGTG": 3196, + "GATGCA": 3197, + "GATGCT": 3198, + "GATGCC": 3199, + "GATGCG": 3200, + "GATGGA": 3201, + "GATGGT": 3202, + "GATGGC": 3203, + "GATGGG": 3204, + "GACAAA": 3205, + "GACAAT": 3206, + "GACAAC": 3207, + "GACAAG": 3208, + "GACATA": 3209, + "GACATT": 3210, + "GACATC": 3211, + "GACATG": 3212, + "GACACA": 3213, + "GACACT": 3214, + "GACACC": 3215, + "GACACG": 3216, + "GACAGA": 3217, + "GACAGT": 3218, + "GACAGC": 3219, + "GACAGG": 3220, + "GACTAA": 3221, + "GACTAT": 3222, + "GACTAC": 3223, + "GACTAG": 3224, + "GACTTA": 3225, + "GACTTT": 3226, + "GACTTC": 3227, + "GACTTG": 3228, + "GACTCA": 3229, + "GACTCT": 3230, + "GACTCC": 3231, + "GACTCG": 3232, + "GACTGA": 3233, + "GACTGT": 3234, + "GACTGC": 3235, + "GACTGG": 3236, + "GACCAA": 3237, + "GACCAT": 3238, + "GACCAC": 3239, + "GACCAG": 3240, + "GACCTA": 3241, + "GACCTT": 3242, + "GACCTC": 3243, + "GACCTG": 3244, + "GACCCA": 3245, + "GACCCT": 3246, + "GACCCC": 3247, + "GACCCG": 3248, + "GACCGA": 3249, + "GACCGT": 3250, + "GACCGC": 3251, + "GACCGG": 3252, + "GACGAA": 3253, + "GACGAT": 3254, + "GACGAC": 3255, + "GACGAG": 3256, + "GACGTA": 3257, + "GACGTT": 3258, + "GACGTC": 3259, + "GACGTG": 3260, + "GACGCA": 3261, + "GACGCT": 3262, + "GACGCC": 3263, + "GACGCG": 3264, + "GACGGA": 3265, + "GACGGT": 3266, + "GACGGC": 3267, + "GACGGG": 3268, + "GAGAAA": 3269, + "GAGAAT": 3270, + "GAGAAC": 3271, + "GAGAAG": 3272, + "GAGATA": 3273, + "GAGATT": 3274, + "GAGATC": 3275, + "GAGATG": 3276, + "GAGACA": 3277, + "GAGACT": 3278, + "GAGACC": 3279, + "GAGACG": 3280, + "GAGAGA": 3281, + "GAGAGT": 3282, + "GAGAGC": 3283, + "GAGAGG": 3284, + "GAGTAA": 3285, + "GAGTAT": 3286, + "GAGTAC": 3287, + "GAGTAG": 3288, + "GAGTTA": 3289, + "GAGTTT": 3290, + "GAGTTC": 3291, + "GAGTTG": 3292, + "GAGTCA": 3293, + "GAGTCT": 3294, + "GAGTCC": 3295, + "GAGTCG": 3296, + "GAGTGA": 3297, + "GAGTGT": 3298, + "GAGTGC": 3299, + "GAGTGG": 3300, + "GAGCAA": 3301, + "GAGCAT": 3302, + "GAGCAC": 3303, + "GAGCAG": 3304, + "GAGCTA": 3305, + "GAGCTT": 3306, + "GAGCTC": 3307, + "GAGCTG": 3308, + "GAGCCA": 3309, + "GAGCCT": 3310, + "GAGCCC": 3311, + "GAGCCG": 3312, + "GAGCGA": 3313, + "GAGCGT": 3314, + "GAGCGC": 3315, + "GAGCGG": 3316, + "GAGGAA": 3317, + "GAGGAT": 3318, + "GAGGAC": 3319, + "GAGGAG": 3320, + "GAGGTA": 3321, + "GAGGTT": 3322, + "GAGGTC": 3323, + "GAGGTG": 3324, + "GAGGCA": 3325, + "GAGGCT": 3326, + "GAGGCC": 3327, + "GAGGCG": 3328, + "GAGGGA": 3329, + "GAGGGT": 3330, + "GAGGGC": 3331, + "GAGGGG": 3332, + "GTAAAA": 3333, + "GTAAAT": 3334, + "GTAAAC": 3335, + "GTAAAG": 3336, + "GTAATA": 3337, + "GTAATT": 3338, + "GTAATC": 3339, + "GTAATG": 3340, + "GTAACA": 3341, + "GTAACT": 3342, + "GTAACC": 3343, + "GTAACG": 3344, + "GTAAGA": 3345, + "GTAAGT": 3346, + "GTAAGC": 3347, + "GTAAGG": 3348, + "GTATAA": 3349, + "GTATAT": 3350, + "GTATAC": 3351, + "GTATAG": 3352, + "GTATTA": 3353, + "GTATTT": 3354, + "GTATTC": 3355, + "GTATTG": 3356, + "GTATCA": 3357, + "GTATCT": 3358, + "GTATCC": 3359, + "GTATCG": 3360, + "GTATGA": 3361, + "GTATGT": 3362, + "GTATGC": 3363, + "GTATGG": 3364, + "GTACAA": 3365, + "GTACAT": 3366, + "GTACAC": 3367, + "GTACAG": 3368, + "GTACTA": 3369, + "GTACTT": 3370, + "GTACTC": 3371, + "GTACTG": 3372, + "GTACCA": 3373, + "GTACCT": 3374, + "GTACCC": 3375, + "GTACCG": 3376, + "GTACGA": 3377, + "GTACGT": 3378, + "GTACGC": 3379, + "GTACGG": 3380, + "GTAGAA": 3381, + "GTAGAT": 3382, + "GTAGAC": 3383, + "GTAGAG": 3384, + "GTAGTA": 3385, + "GTAGTT": 3386, + "GTAGTC": 3387, + "GTAGTG": 3388, + "GTAGCA": 3389, + "GTAGCT": 3390, + "GTAGCC": 3391, + "GTAGCG": 3392, + "GTAGGA": 3393, + "GTAGGT": 3394, + "GTAGGC": 3395, + "GTAGGG": 3396, + "GTTAAA": 3397, + "GTTAAT": 3398, + "GTTAAC": 3399, + "GTTAAG": 3400, + "GTTATA": 3401, + "GTTATT": 3402, + "GTTATC": 3403, + "GTTATG": 3404, + "GTTACA": 3405, + "GTTACT": 3406, + "GTTACC": 3407, + "GTTACG": 3408, + "GTTAGA": 3409, + "GTTAGT": 3410, + "GTTAGC": 3411, + "GTTAGG": 3412, + "GTTTAA": 3413, + "GTTTAT": 3414, + "GTTTAC": 3415, + "GTTTAG": 3416, + "GTTTTA": 3417, + "GTTTTT": 3418, + "GTTTTC": 3419, + "GTTTTG": 3420, + "GTTTCA": 3421, + "GTTTCT": 3422, + "GTTTCC": 3423, + "GTTTCG": 3424, + "GTTTGA": 3425, + "GTTTGT": 3426, + "GTTTGC": 3427, + "GTTTGG": 3428, + "GTTCAA": 3429, + "GTTCAT": 3430, + "GTTCAC": 3431, + "GTTCAG": 3432, + "GTTCTA": 3433, + "GTTCTT": 3434, + "GTTCTC": 3435, + "GTTCTG": 3436, + "GTTCCA": 3437, + "GTTCCT": 3438, + "GTTCCC": 3439, + "GTTCCG": 3440, + "GTTCGA": 3441, + "GTTCGT": 3442, + "GTTCGC": 3443, + "GTTCGG": 3444, + "GTTGAA": 3445, + "GTTGAT": 3446, + "GTTGAC": 3447, + "GTTGAG": 3448, + "GTTGTA": 3449, + "GTTGTT": 3450, + "GTTGTC": 3451, + "GTTGTG": 3452, + "GTTGCA": 3453, + "GTTGCT": 3454, + "GTTGCC": 3455, + "GTTGCG": 3456, + "GTTGGA": 3457, + "GTTGGT": 3458, + "GTTGGC": 3459, + "GTTGGG": 3460, + "GTCAAA": 3461, + "GTCAAT": 3462, + "GTCAAC": 3463, + "GTCAAG": 3464, + "GTCATA": 3465, + "GTCATT": 3466, + "GTCATC": 3467, + "GTCATG": 3468, + "GTCACA": 3469, + "GTCACT": 3470, + "GTCACC": 3471, + "GTCACG": 3472, + "GTCAGA": 3473, + "GTCAGT": 3474, + "GTCAGC": 3475, + "GTCAGG": 3476, + "GTCTAA": 3477, + "GTCTAT": 3478, + "GTCTAC": 3479, + "GTCTAG": 3480, + "GTCTTA": 3481, + "GTCTTT": 3482, + "GTCTTC": 3483, + "GTCTTG": 3484, + "GTCTCA": 3485, + "GTCTCT": 3486, + "GTCTCC": 3487, + "GTCTCG": 3488, + "GTCTGA": 3489, + "GTCTGT": 3490, + "GTCTGC": 3491, + "GTCTGG": 3492, + "GTCCAA": 3493, + "GTCCAT": 3494, + "GTCCAC": 3495, + "GTCCAG": 3496, + "GTCCTA": 3497, + "GTCCTT": 3498, + "GTCCTC": 3499, + "GTCCTG": 3500, + "GTCCCA": 3501, + "GTCCCT": 3502, + "GTCCCC": 3503, + "GTCCCG": 3504, + "GTCCGA": 3505, + "GTCCGT": 3506, + "GTCCGC": 3507, + "GTCCGG": 3508, + "GTCGAA": 3509, + "GTCGAT": 3510, + "GTCGAC": 3511, + "GTCGAG": 3512, + "GTCGTA": 3513, + "GTCGTT": 3514, + "GTCGTC": 3515, + "GTCGTG": 3516, + "GTCGCA": 3517, + "GTCGCT": 3518, + "GTCGCC": 3519, + "GTCGCG": 3520, + "GTCGGA": 3521, + "GTCGGT": 3522, + "GTCGGC": 3523, + "GTCGGG": 3524, + "GTGAAA": 3525, + "GTGAAT": 3526, + "GTGAAC": 3527, + "GTGAAG": 3528, + "GTGATA": 3529, + "GTGATT": 3530, + "GTGATC": 3531, + "GTGATG": 3532, + "GTGACA": 3533, + "GTGACT": 3534, + "GTGACC": 3535, + "GTGACG": 3536, + "GTGAGA": 3537, + "GTGAGT": 3538, + "GTGAGC": 3539, + "GTGAGG": 3540, + "GTGTAA": 3541, + "GTGTAT": 3542, + "GTGTAC": 3543, + "GTGTAG": 3544, + "GTGTTA": 3545, + "GTGTTT": 3546, + "GTGTTC": 3547, + "GTGTTG": 3548, + "GTGTCA": 3549, + "GTGTCT": 3550, + "GTGTCC": 3551, + "GTGTCG": 3552, + "GTGTGA": 3553, + "GTGTGT": 3554, + "GTGTGC": 3555, + "GTGTGG": 3556, + "GTGCAA": 3557, + "GTGCAT": 3558, + "GTGCAC": 3559, + "GTGCAG": 3560, + "GTGCTA": 3561, + "GTGCTT": 3562, + "GTGCTC": 3563, + "GTGCTG": 3564, + "GTGCCA": 3565, + "GTGCCT": 3566, + "GTGCCC": 3567, + "GTGCCG": 3568, + "GTGCGA": 3569, + "GTGCGT": 3570, + "GTGCGC": 3571, + "GTGCGG": 3572, + "GTGGAA": 3573, + "GTGGAT": 3574, + "GTGGAC": 3575, + "GTGGAG": 3576, + "GTGGTA": 3577, + "GTGGTT": 3578, + "GTGGTC": 3579, + "GTGGTG": 3580, + "GTGGCA": 3581, + "GTGGCT": 3582, + "GTGGCC": 3583, + "GTGGCG": 3584, + "GTGGGA": 3585, + "GTGGGT": 3586, + "GTGGGC": 3587, + "GTGGGG": 3588, + "GCAAAA": 3589, + "GCAAAT": 3590, + "GCAAAC": 3591, + "GCAAAG": 3592, + "GCAATA": 3593, + "GCAATT": 3594, + "GCAATC": 3595, + "GCAATG": 3596, + "GCAACA": 3597, + "GCAACT": 3598, + "GCAACC": 3599, + "GCAACG": 3600, + "GCAAGA": 3601, + "GCAAGT": 3602, + "GCAAGC": 3603, + "GCAAGG": 3604, + "GCATAA": 3605, + "GCATAT": 3606, + "GCATAC": 3607, + "GCATAG": 3608, + "GCATTA": 3609, + "GCATTT": 3610, + "GCATTC": 3611, + "GCATTG": 3612, + "GCATCA": 3613, + "GCATCT": 3614, + "GCATCC": 3615, + "GCATCG": 3616, + "GCATGA": 3617, + "GCATGT": 3618, + "GCATGC": 3619, + "GCATGG": 3620, + "GCACAA": 3621, + "GCACAT": 3622, + "GCACAC": 3623, + "GCACAG": 3624, + "GCACTA": 3625, + "GCACTT": 3626, + "GCACTC": 3627, + "GCACTG": 3628, + "GCACCA": 3629, + "GCACCT": 3630, + "GCACCC": 3631, + "GCACCG": 3632, + "GCACGA": 3633, + "GCACGT": 3634, + "GCACGC": 3635, + "GCACGG": 3636, + "GCAGAA": 3637, + "GCAGAT": 3638, + "GCAGAC": 3639, + "GCAGAG": 3640, + "GCAGTA": 3641, + "GCAGTT": 3642, + "GCAGTC": 3643, + "GCAGTG": 3644, + "GCAGCA": 3645, + "GCAGCT": 3646, + "GCAGCC": 3647, + "GCAGCG": 3648, + "GCAGGA": 3649, + "GCAGGT": 3650, + "GCAGGC": 3651, + "GCAGGG": 3652, + "GCTAAA": 3653, + "GCTAAT": 3654, + "GCTAAC": 3655, + "GCTAAG": 3656, + "GCTATA": 3657, + "GCTATT": 3658, + "GCTATC": 3659, + "GCTATG": 3660, + "GCTACA": 3661, + "GCTACT": 3662, + "GCTACC": 3663, + "GCTACG": 3664, + "GCTAGA": 3665, + "GCTAGT": 3666, + "GCTAGC": 3667, + "GCTAGG": 3668, + "GCTTAA": 3669, + "GCTTAT": 3670, + "GCTTAC": 3671, + "GCTTAG": 3672, + "GCTTTA": 3673, + "GCTTTT": 3674, + "GCTTTC": 3675, + "GCTTTG": 3676, + "GCTTCA": 3677, + "GCTTCT": 3678, + "GCTTCC": 3679, + "GCTTCG": 3680, + "GCTTGA": 3681, + "GCTTGT": 3682, + "GCTTGC": 3683, + "GCTTGG": 3684, + "GCTCAA": 3685, + "GCTCAT": 3686, + "GCTCAC": 3687, + "GCTCAG": 3688, + "GCTCTA": 3689, + "GCTCTT": 3690, + "GCTCTC": 3691, + "GCTCTG": 3692, + "GCTCCA": 3693, + "GCTCCT": 3694, + "GCTCCC": 3695, + "GCTCCG": 3696, + "GCTCGA": 3697, + "GCTCGT": 3698, + "GCTCGC": 3699, + "GCTCGG": 3700, + "GCTGAA": 3701, + "GCTGAT": 3702, + "GCTGAC": 3703, + "GCTGAG": 3704, + "GCTGTA": 3705, + "GCTGTT": 3706, + "GCTGTC": 3707, + "GCTGTG": 3708, + "GCTGCA": 3709, + "GCTGCT": 3710, + "GCTGCC": 3711, + "GCTGCG": 3712, + "GCTGGA": 3713, + "GCTGGT": 3714, + "GCTGGC": 3715, + "GCTGGG": 3716, + "GCCAAA": 3717, + "GCCAAT": 3718, + "GCCAAC": 3719, + "GCCAAG": 3720, + "GCCATA": 3721, + "GCCATT": 3722, + "GCCATC": 3723, + "GCCATG": 3724, + "GCCACA": 3725, + "GCCACT": 3726, + "GCCACC": 3727, + "GCCACG": 3728, + "GCCAGA": 3729, + "GCCAGT": 3730, + "GCCAGC": 3731, + "GCCAGG": 3732, + "GCCTAA": 3733, + "GCCTAT": 3734, + "GCCTAC": 3735, + "GCCTAG": 3736, + "GCCTTA": 3737, + "GCCTTT": 3738, + "GCCTTC": 3739, + "GCCTTG": 3740, + "GCCTCA": 3741, + "GCCTCT": 3742, + "GCCTCC": 3743, + "GCCTCG": 3744, + "GCCTGA": 3745, + "GCCTGT": 3746, + "GCCTGC": 3747, + "GCCTGG": 3748, + "GCCCAA": 3749, + "GCCCAT": 3750, + "GCCCAC": 3751, + "GCCCAG": 3752, + "GCCCTA": 3753, + "GCCCTT": 3754, + "GCCCTC": 3755, + "GCCCTG": 3756, + "GCCCCA": 3757, + "GCCCCT": 3758, + "GCCCCC": 3759, + "GCCCCG": 3760, + "GCCCGA": 3761, + "GCCCGT": 3762, + "GCCCGC": 3763, + "GCCCGG": 3764, + "GCCGAA": 3765, + "GCCGAT": 3766, + "GCCGAC": 3767, + "GCCGAG": 3768, + "GCCGTA": 3769, + "GCCGTT": 3770, + "GCCGTC": 3771, + "GCCGTG": 3772, + "GCCGCA": 3773, + "GCCGCT": 3774, + "GCCGCC": 3775, + "GCCGCG": 3776, + "GCCGGA": 3777, + "GCCGGT": 3778, + "GCCGGC": 3779, + "GCCGGG": 3780, + "GCGAAA": 3781, + "GCGAAT": 3782, + "GCGAAC": 3783, + "GCGAAG": 3784, + "GCGATA": 3785, + "GCGATT": 3786, + "GCGATC": 3787, + "GCGATG": 3788, + "GCGACA": 3789, + "GCGACT": 3790, + "GCGACC": 3791, + "GCGACG": 3792, + "GCGAGA": 3793, + "GCGAGT": 3794, + "GCGAGC": 3795, + "GCGAGG": 3796, + "GCGTAA": 3797, + "GCGTAT": 3798, + "GCGTAC": 3799, + "GCGTAG": 3800, + "GCGTTA": 3801, + "GCGTTT": 3802, + "GCGTTC": 3803, + "GCGTTG": 3804, + "GCGTCA": 3805, + "GCGTCT": 3806, + "GCGTCC": 3807, + "GCGTCG": 3808, + "GCGTGA": 3809, + "GCGTGT": 3810, + "GCGTGC": 3811, + "GCGTGG": 3812, + "GCGCAA": 3813, + "GCGCAT": 3814, + "GCGCAC": 3815, + "GCGCAG": 3816, + "GCGCTA": 3817, + "GCGCTT": 3818, + "GCGCTC": 3819, + "GCGCTG": 3820, + "GCGCCA": 3821, + "GCGCCT": 3822, + "GCGCCC": 3823, + "GCGCCG": 3824, + "GCGCGA": 3825, + "GCGCGT": 3826, + "GCGCGC": 3827, + "GCGCGG": 3828, + "GCGGAA": 3829, + "GCGGAT": 3830, + "GCGGAC": 3831, + "GCGGAG": 3832, + "GCGGTA": 3833, + "GCGGTT": 3834, + "GCGGTC": 3835, + "GCGGTG": 3836, + "GCGGCA": 3837, + "GCGGCT": 3838, + "GCGGCC": 3839, + "GCGGCG": 3840, + "GCGGGA": 3841, + "GCGGGT": 3842, + "GCGGGC": 3843, + "GCGGGG": 3844, + "GGAAAA": 3845, + "GGAAAT": 3846, + "GGAAAC": 3847, + "GGAAAG": 3848, + "GGAATA": 3849, + "GGAATT": 3850, + "GGAATC": 3851, + "GGAATG": 3852, + "GGAACA": 3853, + "GGAACT": 3854, + "GGAACC": 3855, + "GGAACG": 3856, + "GGAAGA": 3857, + "GGAAGT": 3858, + "GGAAGC": 3859, + "GGAAGG": 3860, + "GGATAA": 3861, + "GGATAT": 3862, + "GGATAC": 3863, + "GGATAG": 3864, + "GGATTA": 3865, + "GGATTT": 3866, + "GGATTC": 3867, + "GGATTG": 3868, + "GGATCA": 3869, + "GGATCT": 3870, + "GGATCC": 3871, + "GGATCG": 3872, + "GGATGA": 3873, + "GGATGT": 3874, + "GGATGC": 3875, + "GGATGG": 3876, + "GGACAA": 3877, + "GGACAT": 3878, + "GGACAC": 3879, + "GGACAG": 3880, + "GGACTA": 3881, + "GGACTT": 3882, + "GGACTC": 3883, + "GGACTG": 3884, + "GGACCA": 3885, + "GGACCT": 3886, + "GGACCC": 3887, + "GGACCG": 3888, + "GGACGA": 3889, + "GGACGT": 3890, + "GGACGC": 3891, + "GGACGG": 3892, + "GGAGAA": 3893, + "GGAGAT": 3894, + "GGAGAC": 3895, + "GGAGAG": 3896, + "GGAGTA": 3897, + "GGAGTT": 3898, + "GGAGTC": 3899, + "GGAGTG": 3900, + "GGAGCA": 3901, + "GGAGCT": 3902, + "GGAGCC": 3903, + "GGAGCG": 3904, + "GGAGGA": 3905, + "GGAGGT": 3906, + "GGAGGC": 3907, + "GGAGGG": 3908, + "GGTAAA": 3909, + "GGTAAT": 3910, + "GGTAAC": 3911, + "GGTAAG": 3912, + "GGTATA": 3913, + "GGTATT": 3914, + "GGTATC": 3915, + "GGTATG": 3916, + "GGTACA": 3917, + "GGTACT": 3918, + "GGTACC": 3919, + "GGTACG": 3920, + "GGTAGA": 3921, + "GGTAGT": 3922, + "GGTAGC": 3923, + "GGTAGG": 3924, + "GGTTAA": 3925, + "GGTTAT": 3926, + "GGTTAC": 3927, + "GGTTAG": 3928, + "GGTTTA": 3929, + "GGTTTT": 3930, + "GGTTTC": 3931, + "GGTTTG": 3932, + "GGTTCA": 3933, + "GGTTCT": 3934, + "GGTTCC": 3935, + "GGTTCG": 3936, + "GGTTGA": 3937, + "GGTTGT": 3938, + "GGTTGC": 3939, + "GGTTGG": 3940, + "GGTCAA": 3941, + "GGTCAT": 3942, + "GGTCAC": 3943, + "GGTCAG": 3944, + "GGTCTA": 3945, + "GGTCTT": 3946, + "GGTCTC": 3947, + "GGTCTG": 3948, + "GGTCCA": 3949, + "GGTCCT": 3950, + "GGTCCC": 3951, + "GGTCCG": 3952, + "GGTCGA": 3953, + "GGTCGT": 3954, + "GGTCGC": 3955, + "GGTCGG": 3956, + "GGTGAA": 3957, + "GGTGAT": 3958, + "GGTGAC": 3959, + "GGTGAG": 3960, + "GGTGTA": 3961, + "GGTGTT": 3962, + "GGTGTC": 3963, + "GGTGTG": 3964, + "GGTGCA": 3965, + "GGTGCT": 3966, + "GGTGCC": 3967, + "GGTGCG": 3968, + "GGTGGA": 3969, + "GGTGGT": 3970, + "GGTGGC": 3971, + "GGTGGG": 3972, + "GGCAAA": 3973, + "GGCAAT": 3974, + "GGCAAC": 3975, + "GGCAAG": 3976, + "GGCATA": 3977, + "GGCATT": 3978, + "GGCATC": 3979, + "GGCATG": 3980, + "GGCACA": 3981, + "GGCACT": 3982, + "GGCACC": 3983, + "GGCACG": 3984, + "GGCAGA": 3985, + "GGCAGT": 3986, + "GGCAGC": 3987, + "GGCAGG": 3988, + "GGCTAA": 3989, + "GGCTAT": 3990, + "GGCTAC": 3991, + "GGCTAG": 3992, + "GGCTTA": 3993, + "GGCTTT": 3994, + "GGCTTC": 3995, + "GGCTTG": 3996, + "GGCTCA": 3997, + "GGCTCT": 3998, + "GGCTCC": 3999, + "GGCTCG": 4000, + "GGCTGA": 4001, + "GGCTGT": 4002, + "GGCTGC": 4003, + "GGCTGG": 4004, + "GGCCAA": 4005, + "GGCCAT": 4006, + "GGCCAC": 4007, + "GGCCAG": 4008, + "GGCCTA": 4009, + "GGCCTT": 4010, + "GGCCTC": 4011, + "GGCCTG": 4012, + "GGCCCA": 4013, + "GGCCCT": 4014, + "GGCCCC": 4015, + "GGCCCG": 4016, + "GGCCGA": 4017, + "GGCCGT": 4018, + "GGCCGC": 4019, + "GGCCGG": 4020, + "GGCGAA": 4021, + "GGCGAT": 4022, + "GGCGAC": 4023, + "GGCGAG": 4024, + "GGCGTA": 4025, + "GGCGTT": 4026, + "GGCGTC": 4027, + "GGCGTG": 4028, + "GGCGCA": 4029, + "GGCGCT": 4030, + "GGCGCC": 4031, + "GGCGCG": 4032, + "GGCGGA": 4033, + "GGCGGT": 4034, + "GGCGGC": 4035, + "GGCGGG": 4036, + "GGGAAA": 4037, + "GGGAAT": 4038, + "GGGAAC": 4039, + "GGGAAG": 4040, + "GGGATA": 4041, + "GGGATT": 4042, + "GGGATC": 4043, + "GGGATG": 4044, + "GGGACA": 4045, + "GGGACT": 4046, + "GGGACC": 4047, + "GGGACG": 4048, + "GGGAGA": 4049, + "GGGAGT": 4050, + "GGGAGC": 4051, + "GGGAGG": 4052, + "GGGTAA": 4053, + "GGGTAT": 4054, + "GGGTAC": 4055, + "GGGTAG": 4056, + "GGGTTA": 4057, + "GGGTTT": 4058, + "GGGTTC": 4059, + "GGGTTG": 4060, + "GGGTCA": 4061, + "GGGTCT": 4062, + "GGGTCC": 4063, + "GGGTCG": 4064, + "GGGTGA": 4065, + "GGGTGT": 4066, + "GGGTGC": 4067, + "GGGTGG": 4068, + "GGGCAA": 4069, + "GGGCAT": 4070, + "GGGCAC": 4071, + "GGGCAG": 4072, + "GGGCTA": 4073, + "GGGCTT": 4074, + "GGGCTC": 4075, + "GGGCTG": 4076, + "GGGCCA": 4077, + "GGGCCT": 4078, + "GGGCCC": 4079, + "GGGCCG": 4080, + "GGGCGA": 4081, + "GGGCGT": 4082, + "GGGCGC": 4083, + "GGGCGG": 4084, + "GGGGAA": 4085, + "GGGGAT": 4086, + "GGGGAC": 4087, + "GGGGAG": 4088, + "GGGGTA": 4089, + "GGGGTT": 4090, + "GGGGTC": 4091, + "GGGGTG": 4092, + "GGGGCA": 4093, + "GGGGCT": 4094, + "GGGGCC": 4095, + "GGGGCG": 4096, + "GGGGGA": 4097, + "GGGGGT": 4098, + "GGGGGC": 4099, + "GGGGGG": 4100 + } + } +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-2000/tokenizer_config.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-2000/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..604e2b1b29af1dcb3920d2828b6a90276e5e3d02 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-2000/tokenizer_config.json @@ -0,0 +1,58 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "[UNK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "[CLS]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "[SEP]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "4": { + "content": "[MASK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "clean_up_tokenization_spaces": true, + "cls_token": "[CLS]", + "do_basic_tokenize": true, + "do_lower_case": false, + "mask_token": "[MASK]", + "max_len": 512, + "model_max_length": 512, + "never_split": null, + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "strip_accents": null, + "tokenize_chinese_chars": true, + "tokenizer_class": "BertTokenizer", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-2000/trainer_state.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-2000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..37a9db689711300d08c995451c0511d7befb09d6 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-2000/trainer_state.json @@ -0,0 +1,593 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 8.0, + "eval_steps": 500, + "global_step": 2000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.1, + "grad_norm": 0.6611918807029724, + "learning_rate": 5.319148936170213e-05, + "loss": 0.6932, + "step": 25 + }, + { + "epoch": 0.2, + "grad_norm": 0.9467485547065735, + "learning_rate": 0.00010638297872340425, + "loss": 0.6912, + "step": 50 + }, + { + "epoch": 0.3, + "grad_norm": 0.9465051889419556, + "learning_rate": 0.00015957446808510637, + "loss": 0.6893, + "step": 75 + }, + { + "epoch": 0.4, + "grad_norm": 1.5430934429168701, + "learning_rate": 0.0002127659574468085, + "loss": 0.6804, + "step": 100 + }, + { + "epoch": 0.5, + "grad_norm": 1.6103107929229736, + "learning_rate": 0.00026595744680851064, + "loss": 0.6806, + "step": 125 + }, + { + "epoch": 0.6, + "grad_norm": 2.3773200511932373, + "learning_rate": 0.00031914893617021275, + "loss": 0.6801, + "step": 150 + }, + { + "epoch": 0.7, + "grad_norm": 1.8832203149795532, + "learning_rate": 0.0003723404255319149, + "loss": 0.6791, + "step": 175 + }, + { + "epoch": 0.8, + "grad_norm": 1.3350876569747925, + "learning_rate": 0.0003992081821181128, + "loss": 0.6746, + "step": 200 + }, + { + "epoch": 0.9, + "grad_norm": 1.5796219110488892, + "learning_rate": 0.0003975585615308479, + "loss": 0.6771, + "step": 225 + }, + { + "epoch": 1.0, + "grad_norm": 1.884006381034851, + "learning_rate": 0.00039590894094358297, + "loss": 0.6649, + "step": 250 + }, + { + "epoch": 1.1, + "grad_norm": 3.286440372467041, + "learning_rate": 0.0003942593203563181, + "loss": 0.6388, + "step": 275 + }, + { + "epoch": 1.2, + "grad_norm": 8.244946479797363, + "learning_rate": 0.0003926096997690532, + "loss": 0.6238, + "step": 300 + }, + { + "epoch": 1.3, + "grad_norm": 4.265683650970459, + "learning_rate": 0.00039096007918178817, + "loss": 0.6383, + "step": 325 + }, + { + "epoch": 1.4, + "grad_norm": 4.115826606750488, + "learning_rate": 0.0003893104585945233, + "loss": 0.6194, + "step": 350 + }, + { + "epoch": 1.5, + "grad_norm": 5.694250583648682, + "learning_rate": 0.0003876608380072583, + "loss": 0.6324, + "step": 375 + }, + { + "epoch": 1.6, + "grad_norm": 3.463121175765991, + "learning_rate": 0.00038601121741999343, + "loss": 0.621, + "step": 400 + }, + { + "epoch": 1.7, + "grad_norm": 4.582865238189697, + "learning_rate": 0.0003843615968327285, + "loss": 0.6116, + "step": 425 + }, + { + "epoch": 1.8, + "grad_norm": 11.996281623840332, + "learning_rate": 0.0003827119762454636, + "loss": 0.6393, + "step": 450 + }, + { + "epoch": 1.9, + "grad_norm": 3.0407373905181885, + "learning_rate": 0.00038106235565819863, + "loss": 0.628, + "step": 475 + }, + { + "epoch": 2.0, + "grad_norm": 2.917588233947754, + "learning_rate": 0.0003794127350709337, + "loss": 0.6078, + "step": 500 + }, + { + "epoch": 2.1, + "grad_norm": 4.748379707336426, + "learning_rate": 0.0003777631144836688, + "loss": 0.4899, + "step": 525 + }, + { + "epoch": 2.2, + "grad_norm": 3.8076977729797363, + "learning_rate": 0.00037611349389640383, + "loss": 0.5086, + "step": 550 + }, + { + "epoch": 2.3, + "grad_norm": 5.2440714836120605, + "learning_rate": 0.00037446387330913894, + "loss": 0.5327, + "step": 575 + }, + { + "epoch": 2.4, + "grad_norm": 7.110438346862793, + "learning_rate": 0.000372814252721874, + "loss": 0.5436, + "step": 600 + }, + { + "epoch": 2.5, + "grad_norm": 5.46150541305542, + "learning_rate": 0.00037116463213460903, + "loss": 0.5294, + "step": 625 + }, + { + "epoch": 2.6, + "grad_norm": 5.136163234710693, + "learning_rate": 0.00036951501154734414, + "loss": 0.5245, + "step": 650 + }, + { + "epoch": 2.7, + "grad_norm": 8.735346794128418, + "learning_rate": 0.0003678653909600792, + "loss": 0.5449, + "step": 675 + }, + { + "epoch": 2.8, + "grad_norm": 2.922825574874878, + "learning_rate": 0.0003662157703728143, + "loss": 0.5406, + "step": 700 + }, + { + "epoch": 2.9, + "grad_norm": 7.744819641113281, + "learning_rate": 0.00036456614978554934, + "loss": 0.5447, + "step": 725 + }, + { + "epoch": 3.0, + "grad_norm": 11.6185884475708, + "learning_rate": 0.00036291652919828444, + "loss": 0.5195, + "step": 750 + }, + { + "epoch": 3.1, + "grad_norm": 3.73836088180542, + "learning_rate": 0.00036126690861101944, + "loss": 0.3824, + "step": 775 + }, + { + "epoch": 3.2, + "grad_norm": 14.850343704223633, + "learning_rate": 0.00035961728802375454, + "loss": 0.4071, + "step": 800 + }, + { + "epoch": 3.3, + "grad_norm": 5.7157440185546875, + "learning_rate": 0.0003579676674364896, + "loss": 0.3986, + "step": 825 + }, + { + "epoch": 3.4, + "grad_norm": 12.418399810791016, + "learning_rate": 0.0003563180468492247, + "loss": 0.4282, + "step": 850 + }, + { + "epoch": 3.5, + "grad_norm": 12.793001174926758, + "learning_rate": 0.0003546684262619598, + "loss": 0.4822, + "step": 875 + }, + { + "epoch": 3.6, + "grad_norm": 6.489450931549072, + "learning_rate": 0.00035301880567469485, + "loss": 0.4239, + "step": 900 + }, + { + "epoch": 3.7, + "grad_norm": 5.365822792053223, + "learning_rate": 0.0003513691850874299, + "loss": 0.421, + "step": 925 + }, + { + "epoch": 3.8, + "grad_norm": 12.643745422363281, + "learning_rate": 0.00034971956450016495, + "loss": 0.3964, + "step": 950 + }, + { + "epoch": 3.9, + "grad_norm": 14.334024429321289, + "learning_rate": 0.00034806994391290005, + "loss": 0.4634, + "step": 975 + }, + { + "epoch": 4.0, + "grad_norm": 6.819091320037842, + "learning_rate": 0.0003464203233256351, + "loss": 0.4139, + "step": 1000 + }, + { + "epoch": 4.1, + "grad_norm": 8.29238224029541, + "learning_rate": 0.0003447707027383702, + "loss": 0.2695, + "step": 1025 + }, + { + "epoch": 4.2, + "grad_norm": 5.984206676483154, + "learning_rate": 0.00034312108215110525, + "loss": 0.2653, + "step": 1050 + }, + { + "epoch": 4.3, + "grad_norm": 5.9425435066223145, + "learning_rate": 0.0003414714615638403, + "loss": 0.2982, + "step": 1075 + }, + { + "epoch": 4.4, + "grad_norm": 7.1877593994140625, + "learning_rate": 0.0003398218409765754, + "loss": 0.3309, + "step": 1100 + }, + { + "epoch": 4.5, + "grad_norm": 17.83046531677246, + "learning_rate": 0.00033817222038931045, + "loss": 0.3467, + "step": 1125 + }, + { + "epoch": 4.6, + "grad_norm": 4.865128517150879, + "learning_rate": 0.00033652259980204556, + "loss": 0.3117, + "step": 1150 + }, + { + "epoch": 4.7, + "grad_norm": 31.427154541015625, + "learning_rate": 0.0003348729792147806, + "loss": 0.3273, + "step": 1175 + }, + { + "epoch": 4.8, + "grad_norm": 26.77984619140625, + "learning_rate": 0.0003332233586275157, + "loss": 0.3504, + "step": 1200 + }, + { + "epoch": 4.9, + "grad_norm": 9.389993667602539, + "learning_rate": 0.00033157373804025076, + "loss": 0.3339, + "step": 1225 + }, + { + "epoch": 5.0, + "grad_norm": 25.70441246032715, + "learning_rate": 0.0003299241174529858, + "loss": 0.3336, + "step": 1250 + }, + { + "epoch": 5.1, + "grad_norm": 13.482085227966309, + "learning_rate": 0.0003282744968657209, + "loss": 0.2156, + "step": 1275 + }, + { + "epoch": 5.2, + "grad_norm": 8.725810050964355, + "learning_rate": 0.00032662487627845596, + "loss": 0.1811, + "step": 1300 + }, + { + "epoch": 5.3, + "grad_norm": 9.313215255737305, + "learning_rate": 0.00032497525569119106, + "loss": 0.2807, + "step": 1325 + }, + { + "epoch": 5.4, + "grad_norm": 11.026411056518555, + "learning_rate": 0.0003233256351039261, + "loss": 0.2757, + "step": 1350 + }, + { + "epoch": 5.5, + "grad_norm": 11.038985252380371, + "learning_rate": 0.00032167601451666116, + "loss": 0.2177, + "step": 1375 + }, + { + "epoch": 5.6, + "grad_norm": 4.008651256561279, + "learning_rate": 0.00032002639392939627, + "loss": 0.2163, + "step": 1400 + }, + { + "epoch": 5.7, + "grad_norm": 12.480770111083984, + "learning_rate": 0.0003183767733421313, + "loss": 0.2173, + "step": 1425 + }, + { + "epoch": 5.8, + "grad_norm": 8.751969337463379, + "learning_rate": 0.0003167271527548664, + "loss": 0.2299, + "step": 1450 + }, + { + "epoch": 5.9, + "grad_norm": 7.701971530914307, + "learning_rate": 0.00031507753216760147, + "loss": 0.1949, + "step": 1475 + }, + { + "epoch": 6.0, + "grad_norm": 8.48027515411377, + "learning_rate": 0.00031342791158033657, + "loss": 0.2599, + "step": 1500 + }, + { + "epoch": 6.1, + "grad_norm": 9.29404067993164, + "learning_rate": 0.00031177829099307157, + "loss": 0.1376, + "step": 1525 + }, + { + "epoch": 6.2, + "grad_norm": 20.137714385986328, + "learning_rate": 0.00031012867040580667, + "loss": 0.1647, + "step": 1550 + }, + { + "epoch": 6.3, + "grad_norm": 11.394575119018555, + "learning_rate": 0.0003084790498185417, + "loss": 0.1565, + "step": 1575 + }, + { + "epoch": 6.4, + "grad_norm": 8.214287757873535, + "learning_rate": 0.0003068294292312768, + "loss": 0.1739, + "step": 1600 + }, + { + "epoch": 6.5, + "grad_norm": 7.779988765716553, + "learning_rate": 0.0003051798086440119, + "loss": 0.1403, + "step": 1625 + }, + { + "epoch": 6.6, + "grad_norm": 9.421648025512695, + "learning_rate": 0.000303530188056747, + "loss": 0.1545, + "step": 1650 + }, + { + "epoch": 6.7, + "grad_norm": 5.751734256744385, + "learning_rate": 0.000301880567469482, + "loss": 0.1971, + "step": 1675 + }, + { + "epoch": 6.8, + "grad_norm": 23.861705780029297, + "learning_rate": 0.0003002309468822171, + "loss": 0.1681, + "step": 1700 + }, + { + "epoch": 6.9, + "grad_norm": 18.944721221923828, + "learning_rate": 0.0002985813262949522, + "loss": 0.1703, + "step": 1725 + }, + { + "epoch": 7.0, + "grad_norm": 14.045795440673828, + "learning_rate": 0.00029693170570768723, + "loss": 0.1801, + "step": 1750 + }, + { + "epoch": 7.1, + "grad_norm": 6.4620137214660645, + "learning_rate": 0.00029528208512042233, + "loss": 0.1253, + "step": 1775 + }, + { + "epoch": 7.2, + "grad_norm": 4.318169593811035, + "learning_rate": 0.0002936324645331574, + "loss": 0.1397, + "step": 1800 + }, + { + "epoch": 7.3, + "grad_norm": 24.91462516784668, + "learning_rate": 0.00029198284394589243, + "loss": 0.1259, + "step": 1825 + }, + { + "epoch": 7.4, + "grad_norm": 23.614572525024414, + "learning_rate": 0.00029033322335862753, + "loss": 0.1293, + "step": 1850 + }, + { + "epoch": 7.5, + "grad_norm": 3.60048508644104, + "learning_rate": 0.0002886836027713626, + "loss": 0.138, + "step": 1875 + }, + { + "epoch": 7.6, + "grad_norm": 16.62705421447754, + "learning_rate": 0.0002870339821840977, + "loss": 0.1733, + "step": 1900 + }, + { + "epoch": 7.7, + "grad_norm": 33.79671859741211, + "learning_rate": 0.00028538436159683273, + "loss": 0.217, + "step": 1925 + }, + { + "epoch": 7.8, + "grad_norm": 9.69206428527832, + "learning_rate": 0.00028373474100956784, + "loss": 0.1289, + "step": 1950 + }, + { + "epoch": 7.9, + "grad_norm": 18.655046463012695, + "learning_rate": 0.0002820851204223029, + "loss": 0.1168, + "step": 1975 + }, + { + "epoch": 8.0, + "grad_norm": 8.4110746383667, + "learning_rate": 0.00028043549983503794, + "loss": 0.1471, + "step": 2000 + } + ], + "logging_steps": 25, + "max_steps": 6250, + "num_input_tokens_seen": 0, + "num_train_epochs": 25, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 101638963200000.0, + "train_batch_size": 20, + "trial_name": null, + "trial_params": null +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-2000/training_args.bin b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-2000/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..798d8a9f3a9513692035320aba5484c89ac49e69 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-2000/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1cabf5db3e9d79ec0c1945933395ff57418204414cc81f77c90324d003fac782 +size 5304 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-2000/vocab.txt b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-2000/vocab.txt new file mode 100644 index 0000000000000000000000000000000000000000..cc9ad53732e21ea9180f18e1fb67a38cc5702d4c --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-2000/vocab.txt @@ -0,0 +1,4101 @@ +[PAD] +[UNK] +[CLS] +[SEP] +[MASK] +AAAAAA +AAAAAT +AAAAAC +AAAAAG +AAAATA +AAAATT +AAAATC +AAAATG +AAAACA +AAAACT +AAAACC +AAAACG +AAAAGA +AAAAGT +AAAAGC +AAAAGG +AAATAA +AAATAT +AAATAC +AAATAG +AAATTA +AAATTT +AAATTC +AAATTG +AAATCA +AAATCT +AAATCC +AAATCG +AAATGA +AAATGT +AAATGC +AAATGG +AAACAA +AAACAT +AAACAC +AAACAG +AAACTA +AAACTT +AAACTC +AAACTG +AAACCA +AAACCT +AAACCC +AAACCG +AAACGA +AAACGT +AAACGC +AAACGG +AAAGAA +AAAGAT +AAAGAC +AAAGAG +AAAGTA +AAAGTT +AAAGTC +AAAGTG +AAAGCA +AAAGCT +AAAGCC +AAAGCG +AAAGGA +AAAGGT +AAAGGC +AAAGGG +AATAAA +AATAAT +AATAAC +AATAAG +AATATA +AATATT +AATATC +AATATG +AATACA +AATACT +AATACC +AATACG +AATAGA +AATAGT +AATAGC +AATAGG +AATTAA +AATTAT +AATTAC +AATTAG +AATTTA +AATTTT +AATTTC +AATTTG +AATTCA +AATTCT +AATTCC +AATTCG +AATTGA +AATTGT +AATTGC +AATTGG +AATCAA +AATCAT +AATCAC +AATCAG +AATCTA +AATCTT +AATCTC +AATCTG +AATCCA +AATCCT +AATCCC +AATCCG +AATCGA +AATCGT +AATCGC +AATCGG +AATGAA +AATGAT +AATGAC +AATGAG +AATGTA +AATGTT +AATGTC +AATGTG +AATGCA +AATGCT +AATGCC +AATGCG +AATGGA +AATGGT +AATGGC +AATGGG +AACAAA +AACAAT +AACAAC +AACAAG +AACATA +AACATT +AACATC +AACATG +AACACA +AACACT +AACACC +AACACG +AACAGA +AACAGT +AACAGC +AACAGG +AACTAA +AACTAT +AACTAC +AACTAG +AACTTA +AACTTT +AACTTC +AACTTG +AACTCA +AACTCT +AACTCC +AACTCG +AACTGA +AACTGT +AACTGC +AACTGG +AACCAA +AACCAT +AACCAC +AACCAG +AACCTA +AACCTT +AACCTC +AACCTG +AACCCA +AACCCT +AACCCC +AACCCG +AACCGA +AACCGT +AACCGC +AACCGG +AACGAA +AACGAT +AACGAC +AACGAG +AACGTA +AACGTT +AACGTC +AACGTG +AACGCA +AACGCT +AACGCC +AACGCG +AACGGA +AACGGT +AACGGC +AACGGG +AAGAAA +AAGAAT +AAGAAC +AAGAAG +AAGATA +AAGATT +AAGATC +AAGATG +AAGACA +AAGACT +AAGACC +AAGACG +AAGAGA +AAGAGT +AAGAGC +AAGAGG +AAGTAA +AAGTAT +AAGTAC +AAGTAG +AAGTTA +AAGTTT +AAGTTC +AAGTTG +AAGTCA +AAGTCT +AAGTCC +AAGTCG +AAGTGA +AAGTGT +AAGTGC +AAGTGG +AAGCAA +AAGCAT +AAGCAC +AAGCAG +AAGCTA +AAGCTT +AAGCTC +AAGCTG +AAGCCA +AAGCCT +AAGCCC +AAGCCG +AAGCGA +AAGCGT +AAGCGC +AAGCGG +AAGGAA +AAGGAT +AAGGAC +AAGGAG +AAGGTA +AAGGTT +AAGGTC +AAGGTG +AAGGCA +AAGGCT +AAGGCC +AAGGCG +AAGGGA +AAGGGT +AAGGGC +AAGGGG +ATAAAA +ATAAAT +ATAAAC +ATAAAG +ATAATA +ATAATT +ATAATC +ATAATG +ATAACA +ATAACT +ATAACC +ATAACG +ATAAGA +ATAAGT +ATAAGC +ATAAGG +ATATAA +ATATAT +ATATAC +ATATAG +ATATTA +ATATTT +ATATTC +ATATTG +ATATCA +ATATCT +ATATCC +ATATCG +ATATGA +ATATGT +ATATGC +ATATGG +ATACAA +ATACAT +ATACAC +ATACAG +ATACTA +ATACTT +ATACTC +ATACTG +ATACCA +ATACCT +ATACCC +ATACCG +ATACGA +ATACGT +ATACGC +ATACGG +ATAGAA +ATAGAT +ATAGAC +ATAGAG +ATAGTA +ATAGTT +ATAGTC +ATAGTG +ATAGCA +ATAGCT +ATAGCC +ATAGCG +ATAGGA +ATAGGT +ATAGGC +ATAGGG +ATTAAA +ATTAAT +ATTAAC +ATTAAG +ATTATA +ATTATT +ATTATC +ATTATG +ATTACA +ATTACT +ATTACC +ATTACG +ATTAGA +ATTAGT +ATTAGC +ATTAGG +ATTTAA +ATTTAT +ATTTAC +ATTTAG +ATTTTA +ATTTTT +ATTTTC +ATTTTG +ATTTCA +ATTTCT +ATTTCC +ATTTCG +ATTTGA +ATTTGT +ATTTGC +ATTTGG +ATTCAA +ATTCAT +ATTCAC +ATTCAG +ATTCTA +ATTCTT +ATTCTC +ATTCTG +ATTCCA +ATTCCT +ATTCCC +ATTCCG +ATTCGA +ATTCGT +ATTCGC +ATTCGG +ATTGAA +ATTGAT +ATTGAC +ATTGAG +ATTGTA +ATTGTT +ATTGTC +ATTGTG +ATTGCA +ATTGCT +ATTGCC +ATTGCG +ATTGGA +ATTGGT +ATTGGC +ATTGGG +ATCAAA +ATCAAT +ATCAAC +ATCAAG +ATCATA +ATCATT +ATCATC +ATCATG +ATCACA +ATCACT +ATCACC +ATCACG +ATCAGA +ATCAGT +ATCAGC +ATCAGG +ATCTAA +ATCTAT +ATCTAC +ATCTAG +ATCTTA +ATCTTT +ATCTTC +ATCTTG +ATCTCA +ATCTCT +ATCTCC +ATCTCG +ATCTGA +ATCTGT +ATCTGC +ATCTGG +ATCCAA +ATCCAT +ATCCAC +ATCCAG +ATCCTA +ATCCTT +ATCCTC +ATCCTG +ATCCCA +ATCCCT +ATCCCC +ATCCCG +ATCCGA +ATCCGT +ATCCGC +ATCCGG +ATCGAA +ATCGAT +ATCGAC +ATCGAG +ATCGTA +ATCGTT +ATCGTC +ATCGTG +ATCGCA +ATCGCT +ATCGCC +ATCGCG +ATCGGA +ATCGGT +ATCGGC +ATCGGG +ATGAAA +ATGAAT +ATGAAC +ATGAAG +ATGATA +ATGATT +ATGATC +ATGATG +ATGACA +ATGACT +ATGACC +ATGACG +ATGAGA +ATGAGT +ATGAGC +ATGAGG +ATGTAA +ATGTAT +ATGTAC +ATGTAG +ATGTTA +ATGTTT +ATGTTC +ATGTTG +ATGTCA +ATGTCT +ATGTCC +ATGTCG +ATGTGA +ATGTGT +ATGTGC +ATGTGG +ATGCAA +ATGCAT +ATGCAC +ATGCAG +ATGCTA +ATGCTT +ATGCTC +ATGCTG +ATGCCA +ATGCCT +ATGCCC +ATGCCG +ATGCGA +ATGCGT +ATGCGC +ATGCGG +ATGGAA +ATGGAT +ATGGAC +ATGGAG +ATGGTA +ATGGTT +ATGGTC +ATGGTG +ATGGCA +ATGGCT +ATGGCC +ATGGCG +ATGGGA +ATGGGT +ATGGGC +ATGGGG +ACAAAA +ACAAAT +ACAAAC +ACAAAG +ACAATA +ACAATT +ACAATC +ACAATG +ACAACA +ACAACT +ACAACC +ACAACG +ACAAGA +ACAAGT +ACAAGC +ACAAGG +ACATAA +ACATAT +ACATAC +ACATAG +ACATTA +ACATTT +ACATTC +ACATTG +ACATCA +ACATCT +ACATCC +ACATCG +ACATGA +ACATGT +ACATGC +ACATGG +ACACAA +ACACAT +ACACAC +ACACAG +ACACTA +ACACTT +ACACTC +ACACTG +ACACCA +ACACCT +ACACCC +ACACCG +ACACGA +ACACGT +ACACGC +ACACGG +ACAGAA +ACAGAT +ACAGAC +ACAGAG +ACAGTA +ACAGTT +ACAGTC +ACAGTG +ACAGCA +ACAGCT +ACAGCC +ACAGCG +ACAGGA +ACAGGT +ACAGGC +ACAGGG +ACTAAA +ACTAAT +ACTAAC +ACTAAG +ACTATA +ACTATT +ACTATC +ACTATG +ACTACA +ACTACT +ACTACC +ACTACG +ACTAGA +ACTAGT +ACTAGC +ACTAGG +ACTTAA +ACTTAT +ACTTAC +ACTTAG +ACTTTA +ACTTTT +ACTTTC +ACTTTG +ACTTCA +ACTTCT +ACTTCC +ACTTCG +ACTTGA +ACTTGT +ACTTGC +ACTTGG +ACTCAA +ACTCAT +ACTCAC +ACTCAG +ACTCTA +ACTCTT +ACTCTC +ACTCTG +ACTCCA +ACTCCT +ACTCCC +ACTCCG +ACTCGA +ACTCGT +ACTCGC +ACTCGG +ACTGAA +ACTGAT +ACTGAC +ACTGAG +ACTGTA +ACTGTT +ACTGTC +ACTGTG +ACTGCA +ACTGCT +ACTGCC +ACTGCG +ACTGGA +ACTGGT +ACTGGC +ACTGGG +ACCAAA +ACCAAT +ACCAAC +ACCAAG +ACCATA +ACCATT +ACCATC +ACCATG +ACCACA +ACCACT +ACCACC +ACCACG +ACCAGA +ACCAGT +ACCAGC +ACCAGG +ACCTAA +ACCTAT +ACCTAC +ACCTAG +ACCTTA +ACCTTT +ACCTTC +ACCTTG +ACCTCA +ACCTCT +ACCTCC +ACCTCG +ACCTGA +ACCTGT +ACCTGC +ACCTGG +ACCCAA +ACCCAT +ACCCAC +ACCCAG +ACCCTA +ACCCTT +ACCCTC +ACCCTG +ACCCCA +ACCCCT +ACCCCC +ACCCCG +ACCCGA +ACCCGT +ACCCGC +ACCCGG +ACCGAA +ACCGAT +ACCGAC +ACCGAG +ACCGTA +ACCGTT +ACCGTC +ACCGTG +ACCGCA +ACCGCT +ACCGCC +ACCGCG +ACCGGA +ACCGGT +ACCGGC +ACCGGG +ACGAAA +ACGAAT +ACGAAC +ACGAAG +ACGATA +ACGATT +ACGATC +ACGATG +ACGACA +ACGACT +ACGACC +ACGACG +ACGAGA +ACGAGT +ACGAGC +ACGAGG +ACGTAA +ACGTAT +ACGTAC +ACGTAG +ACGTTA +ACGTTT +ACGTTC +ACGTTG +ACGTCA +ACGTCT +ACGTCC +ACGTCG +ACGTGA +ACGTGT +ACGTGC +ACGTGG +ACGCAA +ACGCAT +ACGCAC +ACGCAG +ACGCTA +ACGCTT +ACGCTC +ACGCTG +ACGCCA +ACGCCT +ACGCCC +ACGCCG +ACGCGA +ACGCGT +ACGCGC +ACGCGG +ACGGAA +ACGGAT +ACGGAC +ACGGAG +ACGGTA +ACGGTT +ACGGTC +ACGGTG +ACGGCA +ACGGCT +ACGGCC +ACGGCG +ACGGGA +ACGGGT +ACGGGC +ACGGGG +AGAAAA +AGAAAT +AGAAAC +AGAAAG +AGAATA +AGAATT +AGAATC +AGAATG +AGAACA +AGAACT +AGAACC +AGAACG +AGAAGA +AGAAGT +AGAAGC +AGAAGG +AGATAA +AGATAT +AGATAC +AGATAG +AGATTA +AGATTT +AGATTC +AGATTG +AGATCA +AGATCT +AGATCC +AGATCG +AGATGA +AGATGT +AGATGC +AGATGG +AGACAA +AGACAT +AGACAC +AGACAG +AGACTA +AGACTT +AGACTC +AGACTG +AGACCA +AGACCT +AGACCC +AGACCG +AGACGA +AGACGT +AGACGC +AGACGG +AGAGAA +AGAGAT +AGAGAC +AGAGAG +AGAGTA +AGAGTT +AGAGTC +AGAGTG +AGAGCA +AGAGCT +AGAGCC +AGAGCG +AGAGGA +AGAGGT +AGAGGC +AGAGGG +AGTAAA +AGTAAT +AGTAAC +AGTAAG +AGTATA +AGTATT +AGTATC +AGTATG +AGTACA +AGTACT +AGTACC +AGTACG +AGTAGA +AGTAGT +AGTAGC +AGTAGG +AGTTAA +AGTTAT +AGTTAC +AGTTAG +AGTTTA +AGTTTT +AGTTTC +AGTTTG +AGTTCA +AGTTCT +AGTTCC +AGTTCG +AGTTGA +AGTTGT +AGTTGC +AGTTGG +AGTCAA +AGTCAT +AGTCAC +AGTCAG +AGTCTA +AGTCTT +AGTCTC +AGTCTG +AGTCCA +AGTCCT +AGTCCC +AGTCCG +AGTCGA +AGTCGT +AGTCGC +AGTCGG +AGTGAA +AGTGAT +AGTGAC +AGTGAG +AGTGTA +AGTGTT +AGTGTC +AGTGTG +AGTGCA +AGTGCT +AGTGCC +AGTGCG +AGTGGA +AGTGGT +AGTGGC +AGTGGG +AGCAAA +AGCAAT +AGCAAC +AGCAAG +AGCATA +AGCATT +AGCATC +AGCATG +AGCACA +AGCACT +AGCACC +AGCACG +AGCAGA +AGCAGT +AGCAGC +AGCAGG +AGCTAA +AGCTAT +AGCTAC +AGCTAG +AGCTTA +AGCTTT +AGCTTC +AGCTTG +AGCTCA +AGCTCT +AGCTCC +AGCTCG +AGCTGA +AGCTGT +AGCTGC +AGCTGG +AGCCAA +AGCCAT +AGCCAC +AGCCAG +AGCCTA +AGCCTT +AGCCTC +AGCCTG +AGCCCA +AGCCCT +AGCCCC +AGCCCG +AGCCGA +AGCCGT +AGCCGC +AGCCGG +AGCGAA +AGCGAT +AGCGAC +AGCGAG +AGCGTA +AGCGTT +AGCGTC +AGCGTG +AGCGCA +AGCGCT +AGCGCC +AGCGCG +AGCGGA +AGCGGT +AGCGGC +AGCGGG +AGGAAA +AGGAAT +AGGAAC +AGGAAG +AGGATA +AGGATT +AGGATC +AGGATG +AGGACA +AGGACT +AGGACC +AGGACG +AGGAGA +AGGAGT +AGGAGC +AGGAGG +AGGTAA +AGGTAT +AGGTAC +AGGTAG +AGGTTA +AGGTTT +AGGTTC +AGGTTG +AGGTCA +AGGTCT +AGGTCC +AGGTCG +AGGTGA +AGGTGT +AGGTGC +AGGTGG +AGGCAA +AGGCAT +AGGCAC +AGGCAG +AGGCTA +AGGCTT +AGGCTC +AGGCTG +AGGCCA +AGGCCT +AGGCCC +AGGCCG +AGGCGA +AGGCGT +AGGCGC +AGGCGG +AGGGAA +AGGGAT +AGGGAC +AGGGAG +AGGGTA +AGGGTT +AGGGTC +AGGGTG +AGGGCA +AGGGCT +AGGGCC +AGGGCG +AGGGGA +AGGGGT +AGGGGC +AGGGGG +TAAAAA +TAAAAT +TAAAAC +TAAAAG +TAAATA +TAAATT +TAAATC +TAAATG +TAAACA +TAAACT +TAAACC +TAAACG +TAAAGA +TAAAGT +TAAAGC +TAAAGG +TAATAA +TAATAT +TAATAC +TAATAG +TAATTA +TAATTT +TAATTC +TAATTG +TAATCA +TAATCT +TAATCC +TAATCG +TAATGA +TAATGT +TAATGC +TAATGG +TAACAA +TAACAT +TAACAC +TAACAG +TAACTA +TAACTT +TAACTC +TAACTG +TAACCA +TAACCT +TAACCC +TAACCG +TAACGA +TAACGT +TAACGC +TAACGG +TAAGAA +TAAGAT +TAAGAC +TAAGAG +TAAGTA +TAAGTT +TAAGTC +TAAGTG +TAAGCA +TAAGCT +TAAGCC +TAAGCG +TAAGGA +TAAGGT +TAAGGC +TAAGGG +TATAAA +TATAAT +TATAAC +TATAAG +TATATA +TATATT +TATATC +TATATG +TATACA +TATACT +TATACC +TATACG +TATAGA +TATAGT +TATAGC +TATAGG +TATTAA +TATTAT +TATTAC +TATTAG +TATTTA +TATTTT +TATTTC +TATTTG +TATTCA +TATTCT +TATTCC +TATTCG +TATTGA +TATTGT +TATTGC +TATTGG +TATCAA +TATCAT +TATCAC +TATCAG +TATCTA +TATCTT +TATCTC +TATCTG +TATCCA +TATCCT +TATCCC +TATCCG +TATCGA +TATCGT +TATCGC +TATCGG +TATGAA +TATGAT +TATGAC +TATGAG +TATGTA +TATGTT +TATGTC +TATGTG +TATGCA +TATGCT +TATGCC +TATGCG +TATGGA +TATGGT +TATGGC +TATGGG +TACAAA +TACAAT +TACAAC +TACAAG +TACATA +TACATT +TACATC +TACATG +TACACA +TACACT +TACACC +TACACG +TACAGA +TACAGT +TACAGC +TACAGG +TACTAA +TACTAT +TACTAC +TACTAG +TACTTA +TACTTT +TACTTC +TACTTG +TACTCA +TACTCT +TACTCC +TACTCG +TACTGA +TACTGT +TACTGC +TACTGG +TACCAA +TACCAT +TACCAC +TACCAG +TACCTA +TACCTT +TACCTC +TACCTG +TACCCA +TACCCT +TACCCC +TACCCG +TACCGA +TACCGT +TACCGC +TACCGG +TACGAA +TACGAT +TACGAC +TACGAG +TACGTA +TACGTT +TACGTC +TACGTG +TACGCA +TACGCT +TACGCC +TACGCG +TACGGA +TACGGT +TACGGC +TACGGG +TAGAAA +TAGAAT +TAGAAC +TAGAAG +TAGATA +TAGATT +TAGATC +TAGATG +TAGACA +TAGACT +TAGACC +TAGACG +TAGAGA +TAGAGT +TAGAGC +TAGAGG +TAGTAA +TAGTAT +TAGTAC +TAGTAG +TAGTTA +TAGTTT +TAGTTC +TAGTTG +TAGTCA +TAGTCT +TAGTCC +TAGTCG +TAGTGA +TAGTGT +TAGTGC +TAGTGG +TAGCAA +TAGCAT +TAGCAC +TAGCAG +TAGCTA +TAGCTT +TAGCTC +TAGCTG +TAGCCA +TAGCCT +TAGCCC +TAGCCG +TAGCGA +TAGCGT +TAGCGC +TAGCGG +TAGGAA +TAGGAT +TAGGAC +TAGGAG +TAGGTA +TAGGTT +TAGGTC +TAGGTG +TAGGCA +TAGGCT +TAGGCC +TAGGCG +TAGGGA +TAGGGT +TAGGGC +TAGGGG +TTAAAA +TTAAAT +TTAAAC +TTAAAG +TTAATA +TTAATT +TTAATC +TTAATG +TTAACA +TTAACT +TTAACC +TTAACG +TTAAGA +TTAAGT +TTAAGC +TTAAGG +TTATAA +TTATAT +TTATAC +TTATAG +TTATTA +TTATTT +TTATTC +TTATTG +TTATCA +TTATCT +TTATCC +TTATCG +TTATGA +TTATGT +TTATGC +TTATGG +TTACAA +TTACAT +TTACAC +TTACAG +TTACTA +TTACTT +TTACTC +TTACTG +TTACCA +TTACCT +TTACCC +TTACCG +TTACGA +TTACGT +TTACGC +TTACGG +TTAGAA +TTAGAT +TTAGAC +TTAGAG +TTAGTA +TTAGTT +TTAGTC +TTAGTG +TTAGCA +TTAGCT +TTAGCC +TTAGCG +TTAGGA +TTAGGT +TTAGGC +TTAGGG +TTTAAA +TTTAAT +TTTAAC +TTTAAG +TTTATA +TTTATT +TTTATC +TTTATG +TTTACA +TTTACT +TTTACC +TTTACG +TTTAGA +TTTAGT +TTTAGC +TTTAGG +TTTTAA +TTTTAT +TTTTAC +TTTTAG +TTTTTA +TTTTTT +TTTTTC +TTTTTG +TTTTCA +TTTTCT +TTTTCC +TTTTCG +TTTTGA +TTTTGT +TTTTGC +TTTTGG +TTTCAA +TTTCAT +TTTCAC +TTTCAG +TTTCTA +TTTCTT +TTTCTC +TTTCTG +TTTCCA +TTTCCT +TTTCCC +TTTCCG +TTTCGA +TTTCGT +TTTCGC +TTTCGG +TTTGAA +TTTGAT +TTTGAC +TTTGAG +TTTGTA +TTTGTT +TTTGTC +TTTGTG +TTTGCA +TTTGCT +TTTGCC +TTTGCG +TTTGGA +TTTGGT +TTTGGC +TTTGGG +TTCAAA +TTCAAT +TTCAAC +TTCAAG +TTCATA +TTCATT +TTCATC +TTCATG +TTCACA +TTCACT +TTCACC +TTCACG +TTCAGA +TTCAGT +TTCAGC +TTCAGG +TTCTAA +TTCTAT +TTCTAC +TTCTAG +TTCTTA +TTCTTT +TTCTTC +TTCTTG +TTCTCA +TTCTCT +TTCTCC +TTCTCG +TTCTGA +TTCTGT +TTCTGC +TTCTGG +TTCCAA +TTCCAT +TTCCAC +TTCCAG +TTCCTA +TTCCTT +TTCCTC +TTCCTG +TTCCCA +TTCCCT +TTCCCC +TTCCCG +TTCCGA +TTCCGT +TTCCGC +TTCCGG +TTCGAA +TTCGAT +TTCGAC +TTCGAG +TTCGTA +TTCGTT +TTCGTC +TTCGTG +TTCGCA +TTCGCT +TTCGCC +TTCGCG +TTCGGA +TTCGGT +TTCGGC +TTCGGG +TTGAAA +TTGAAT +TTGAAC +TTGAAG +TTGATA +TTGATT +TTGATC +TTGATG +TTGACA +TTGACT +TTGACC +TTGACG +TTGAGA +TTGAGT +TTGAGC +TTGAGG +TTGTAA +TTGTAT +TTGTAC +TTGTAG +TTGTTA +TTGTTT +TTGTTC +TTGTTG +TTGTCA +TTGTCT +TTGTCC +TTGTCG +TTGTGA +TTGTGT +TTGTGC +TTGTGG +TTGCAA +TTGCAT +TTGCAC +TTGCAG +TTGCTA +TTGCTT +TTGCTC +TTGCTG +TTGCCA +TTGCCT +TTGCCC +TTGCCG +TTGCGA +TTGCGT +TTGCGC +TTGCGG +TTGGAA +TTGGAT +TTGGAC +TTGGAG +TTGGTA +TTGGTT +TTGGTC +TTGGTG +TTGGCA +TTGGCT +TTGGCC +TTGGCG +TTGGGA +TTGGGT +TTGGGC +TTGGGG +TCAAAA +TCAAAT +TCAAAC +TCAAAG +TCAATA +TCAATT +TCAATC +TCAATG +TCAACA +TCAACT +TCAACC +TCAACG +TCAAGA +TCAAGT +TCAAGC +TCAAGG +TCATAA +TCATAT +TCATAC +TCATAG +TCATTA +TCATTT +TCATTC +TCATTG +TCATCA +TCATCT +TCATCC +TCATCG +TCATGA +TCATGT +TCATGC +TCATGG +TCACAA +TCACAT +TCACAC +TCACAG +TCACTA +TCACTT +TCACTC +TCACTG +TCACCA +TCACCT +TCACCC +TCACCG +TCACGA +TCACGT +TCACGC +TCACGG +TCAGAA +TCAGAT +TCAGAC +TCAGAG +TCAGTA +TCAGTT +TCAGTC +TCAGTG +TCAGCA +TCAGCT +TCAGCC +TCAGCG +TCAGGA +TCAGGT +TCAGGC +TCAGGG +TCTAAA +TCTAAT +TCTAAC +TCTAAG +TCTATA +TCTATT +TCTATC +TCTATG +TCTACA +TCTACT +TCTACC +TCTACG +TCTAGA +TCTAGT +TCTAGC +TCTAGG +TCTTAA +TCTTAT +TCTTAC +TCTTAG +TCTTTA +TCTTTT +TCTTTC +TCTTTG +TCTTCA +TCTTCT +TCTTCC +TCTTCG +TCTTGA +TCTTGT +TCTTGC +TCTTGG +TCTCAA +TCTCAT +TCTCAC +TCTCAG +TCTCTA +TCTCTT +TCTCTC +TCTCTG +TCTCCA +TCTCCT +TCTCCC +TCTCCG +TCTCGA +TCTCGT +TCTCGC +TCTCGG +TCTGAA +TCTGAT +TCTGAC +TCTGAG +TCTGTA +TCTGTT +TCTGTC +TCTGTG +TCTGCA +TCTGCT +TCTGCC +TCTGCG +TCTGGA +TCTGGT +TCTGGC +TCTGGG +TCCAAA +TCCAAT +TCCAAC +TCCAAG +TCCATA +TCCATT +TCCATC +TCCATG +TCCACA +TCCACT +TCCACC +TCCACG +TCCAGA +TCCAGT +TCCAGC +TCCAGG +TCCTAA +TCCTAT +TCCTAC +TCCTAG +TCCTTA +TCCTTT +TCCTTC +TCCTTG +TCCTCA +TCCTCT +TCCTCC +TCCTCG +TCCTGA +TCCTGT +TCCTGC +TCCTGG +TCCCAA +TCCCAT +TCCCAC +TCCCAG +TCCCTA +TCCCTT +TCCCTC +TCCCTG +TCCCCA +TCCCCT +TCCCCC +TCCCCG +TCCCGA +TCCCGT +TCCCGC +TCCCGG +TCCGAA +TCCGAT +TCCGAC +TCCGAG +TCCGTA +TCCGTT +TCCGTC +TCCGTG +TCCGCA +TCCGCT +TCCGCC +TCCGCG +TCCGGA +TCCGGT +TCCGGC +TCCGGG +TCGAAA +TCGAAT +TCGAAC +TCGAAG +TCGATA +TCGATT +TCGATC +TCGATG +TCGACA +TCGACT +TCGACC +TCGACG +TCGAGA +TCGAGT +TCGAGC +TCGAGG +TCGTAA +TCGTAT +TCGTAC +TCGTAG +TCGTTA +TCGTTT +TCGTTC +TCGTTG +TCGTCA +TCGTCT +TCGTCC +TCGTCG +TCGTGA +TCGTGT +TCGTGC +TCGTGG +TCGCAA +TCGCAT +TCGCAC +TCGCAG +TCGCTA +TCGCTT +TCGCTC +TCGCTG +TCGCCA +TCGCCT +TCGCCC +TCGCCG +TCGCGA +TCGCGT +TCGCGC +TCGCGG +TCGGAA +TCGGAT +TCGGAC +TCGGAG +TCGGTA +TCGGTT +TCGGTC +TCGGTG +TCGGCA +TCGGCT +TCGGCC +TCGGCG +TCGGGA +TCGGGT +TCGGGC +TCGGGG +TGAAAA +TGAAAT +TGAAAC +TGAAAG +TGAATA +TGAATT +TGAATC +TGAATG +TGAACA +TGAACT +TGAACC +TGAACG +TGAAGA +TGAAGT +TGAAGC +TGAAGG +TGATAA +TGATAT +TGATAC +TGATAG +TGATTA +TGATTT +TGATTC +TGATTG +TGATCA +TGATCT +TGATCC +TGATCG +TGATGA +TGATGT +TGATGC +TGATGG +TGACAA +TGACAT +TGACAC +TGACAG +TGACTA +TGACTT +TGACTC +TGACTG +TGACCA +TGACCT +TGACCC +TGACCG +TGACGA +TGACGT +TGACGC +TGACGG +TGAGAA +TGAGAT +TGAGAC +TGAGAG +TGAGTA +TGAGTT +TGAGTC +TGAGTG +TGAGCA +TGAGCT +TGAGCC +TGAGCG +TGAGGA +TGAGGT +TGAGGC +TGAGGG +TGTAAA +TGTAAT +TGTAAC +TGTAAG +TGTATA +TGTATT +TGTATC +TGTATG +TGTACA +TGTACT +TGTACC +TGTACG +TGTAGA +TGTAGT +TGTAGC +TGTAGG +TGTTAA +TGTTAT +TGTTAC +TGTTAG +TGTTTA +TGTTTT +TGTTTC +TGTTTG +TGTTCA +TGTTCT +TGTTCC +TGTTCG +TGTTGA +TGTTGT +TGTTGC +TGTTGG +TGTCAA +TGTCAT +TGTCAC +TGTCAG +TGTCTA +TGTCTT +TGTCTC +TGTCTG +TGTCCA +TGTCCT +TGTCCC +TGTCCG +TGTCGA +TGTCGT +TGTCGC +TGTCGG +TGTGAA +TGTGAT +TGTGAC +TGTGAG +TGTGTA +TGTGTT +TGTGTC +TGTGTG +TGTGCA +TGTGCT +TGTGCC +TGTGCG +TGTGGA +TGTGGT +TGTGGC +TGTGGG +TGCAAA +TGCAAT +TGCAAC +TGCAAG +TGCATA +TGCATT +TGCATC +TGCATG +TGCACA +TGCACT +TGCACC +TGCACG +TGCAGA +TGCAGT +TGCAGC +TGCAGG +TGCTAA +TGCTAT +TGCTAC +TGCTAG +TGCTTA +TGCTTT +TGCTTC +TGCTTG +TGCTCA +TGCTCT +TGCTCC +TGCTCG +TGCTGA +TGCTGT +TGCTGC +TGCTGG +TGCCAA +TGCCAT +TGCCAC +TGCCAG +TGCCTA +TGCCTT +TGCCTC +TGCCTG +TGCCCA +TGCCCT +TGCCCC +TGCCCG +TGCCGA +TGCCGT +TGCCGC +TGCCGG +TGCGAA +TGCGAT +TGCGAC +TGCGAG +TGCGTA +TGCGTT +TGCGTC +TGCGTG +TGCGCA +TGCGCT +TGCGCC +TGCGCG +TGCGGA +TGCGGT +TGCGGC +TGCGGG +TGGAAA +TGGAAT +TGGAAC +TGGAAG +TGGATA +TGGATT +TGGATC +TGGATG +TGGACA +TGGACT +TGGACC +TGGACG +TGGAGA +TGGAGT +TGGAGC +TGGAGG +TGGTAA +TGGTAT +TGGTAC +TGGTAG +TGGTTA +TGGTTT +TGGTTC +TGGTTG +TGGTCA +TGGTCT +TGGTCC +TGGTCG +TGGTGA +TGGTGT +TGGTGC +TGGTGG +TGGCAA +TGGCAT +TGGCAC +TGGCAG +TGGCTA +TGGCTT +TGGCTC +TGGCTG +TGGCCA +TGGCCT +TGGCCC +TGGCCG +TGGCGA +TGGCGT +TGGCGC +TGGCGG +TGGGAA +TGGGAT +TGGGAC +TGGGAG +TGGGTA +TGGGTT +TGGGTC +TGGGTG +TGGGCA +TGGGCT +TGGGCC +TGGGCG +TGGGGA +TGGGGT +TGGGGC +TGGGGG +CAAAAA +CAAAAT +CAAAAC +CAAAAG +CAAATA +CAAATT +CAAATC +CAAATG +CAAACA +CAAACT +CAAACC +CAAACG +CAAAGA +CAAAGT +CAAAGC +CAAAGG +CAATAA +CAATAT +CAATAC +CAATAG +CAATTA +CAATTT +CAATTC +CAATTG +CAATCA +CAATCT +CAATCC +CAATCG +CAATGA +CAATGT +CAATGC +CAATGG +CAACAA +CAACAT +CAACAC +CAACAG +CAACTA +CAACTT +CAACTC +CAACTG +CAACCA +CAACCT +CAACCC +CAACCG +CAACGA +CAACGT +CAACGC +CAACGG +CAAGAA +CAAGAT +CAAGAC +CAAGAG +CAAGTA +CAAGTT +CAAGTC +CAAGTG +CAAGCA +CAAGCT +CAAGCC +CAAGCG +CAAGGA +CAAGGT +CAAGGC +CAAGGG +CATAAA +CATAAT +CATAAC +CATAAG +CATATA +CATATT +CATATC +CATATG +CATACA +CATACT +CATACC +CATACG +CATAGA +CATAGT +CATAGC +CATAGG +CATTAA +CATTAT +CATTAC +CATTAG +CATTTA +CATTTT +CATTTC +CATTTG +CATTCA +CATTCT +CATTCC +CATTCG +CATTGA +CATTGT +CATTGC +CATTGG +CATCAA +CATCAT +CATCAC +CATCAG +CATCTA +CATCTT +CATCTC +CATCTG +CATCCA +CATCCT +CATCCC +CATCCG +CATCGA +CATCGT +CATCGC +CATCGG +CATGAA +CATGAT +CATGAC +CATGAG +CATGTA +CATGTT +CATGTC +CATGTG +CATGCA +CATGCT +CATGCC +CATGCG +CATGGA +CATGGT +CATGGC +CATGGG +CACAAA +CACAAT +CACAAC +CACAAG +CACATA +CACATT +CACATC +CACATG +CACACA +CACACT +CACACC +CACACG +CACAGA +CACAGT +CACAGC +CACAGG +CACTAA +CACTAT +CACTAC +CACTAG +CACTTA +CACTTT +CACTTC +CACTTG +CACTCA +CACTCT +CACTCC +CACTCG +CACTGA +CACTGT +CACTGC +CACTGG +CACCAA +CACCAT +CACCAC +CACCAG +CACCTA +CACCTT +CACCTC +CACCTG +CACCCA +CACCCT +CACCCC +CACCCG +CACCGA +CACCGT +CACCGC +CACCGG +CACGAA +CACGAT +CACGAC +CACGAG +CACGTA +CACGTT +CACGTC +CACGTG +CACGCA +CACGCT +CACGCC +CACGCG +CACGGA +CACGGT +CACGGC +CACGGG +CAGAAA +CAGAAT +CAGAAC +CAGAAG +CAGATA +CAGATT +CAGATC +CAGATG +CAGACA +CAGACT +CAGACC +CAGACG +CAGAGA +CAGAGT +CAGAGC +CAGAGG +CAGTAA +CAGTAT +CAGTAC +CAGTAG +CAGTTA +CAGTTT +CAGTTC +CAGTTG +CAGTCA +CAGTCT +CAGTCC +CAGTCG +CAGTGA +CAGTGT +CAGTGC +CAGTGG +CAGCAA +CAGCAT +CAGCAC +CAGCAG +CAGCTA +CAGCTT +CAGCTC +CAGCTG +CAGCCA +CAGCCT +CAGCCC +CAGCCG +CAGCGA +CAGCGT +CAGCGC +CAGCGG +CAGGAA +CAGGAT +CAGGAC +CAGGAG +CAGGTA +CAGGTT +CAGGTC +CAGGTG +CAGGCA +CAGGCT +CAGGCC +CAGGCG +CAGGGA +CAGGGT +CAGGGC +CAGGGG +CTAAAA +CTAAAT +CTAAAC +CTAAAG +CTAATA +CTAATT +CTAATC +CTAATG +CTAACA +CTAACT +CTAACC +CTAACG +CTAAGA +CTAAGT +CTAAGC +CTAAGG +CTATAA +CTATAT +CTATAC +CTATAG +CTATTA +CTATTT +CTATTC +CTATTG +CTATCA +CTATCT +CTATCC +CTATCG +CTATGA +CTATGT +CTATGC +CTATGG +CTACAA +CTACAT +CTACAC +CTACAG +CTACTA +CTACTT +CTACTC +CTACTG +CTACCA +CTACCT +CTACCC +CTACCG +CTACGA +CTACGT +CTACGC +CTACGG +CTAGAA +CTAGAT +CTAGAC +CTAGAG +CTAGTA +CTAGTT +CTAGTC +CTAGTG +CTAGCA +CTAGCT +CTAGCC +CTAGCG +CTAGGA +CTAGGT +CTAGGC +CTAGGG +CTTAAA +CTTAAT +CTTAAC +CTTAAG +CTTATA +CTTATT +CTTATC +CTTATG +CTTACA +CTTACT +CTTACC +CTTACG +CTTAGA +CTTAGT +CTTAGC +CTTAGG +CTTTAA +CTTTAT +CTTTAC +CTTTAG +CTTTTA +CTTTTT +CTTTTC +CTTTTG +CTTTCA +CTTTCT +CTTTCC +CTTTCG +CTTTGA +CTTTGT +CTTTGC +CTTTGG +CTTCAA +CTTCAT +CTTCAC +CTTCAG +CTTCTA +CTTCTT +CTTCTC +CTTCTG +CTTCCA +CTTCCT +CTTCCC +CTTCCG +CTTCGA +CTTCGT +CTTCGC +CTTCGG +CTTGAA +CTTGAT +CTTGAC +CTTGAG +CTTGTA +CTTGTT +CTTGTC +CTTGTG +CTTGCA +CTTGCT +CTTGCC +CTTGCG +CTTGGA +CTTGGT +CTTGGC +CTTGGG +CTCAAA +CTCAAT +CTCAAC +CTCAAG +CTCATA +CTCATT +CTCATC +CTCATG +CTCACA +CTCACT +CTCACC +CTCACG +CTCAGA +CTCAGT +CTCAGC +CTCAGG +CTCTAA +CTCTAT +CTCTAC +CTCTAG +CTCTTA +CTCTTT +CTCTTC +CTCTTG +CTCTCA +CTCTCT +CTCTCC +CTCTCG +CTCTGA +CTCTGT +CTCTGC +CTCTGG +CTCCAA +CTCCAT +CTCCAC +CTCCAG +CTCCTA +CTCCTT +CTCCTC +CTCCTG +CTCCCA +CTCCCT +CTCCCC +CTCCCG +CTCCGA +CTCCGT +CTCCGC +CTCCGG +CTCGAA +CTCGAT +CTCGAC +CTCGAG +CTCGTA +CTCGTT +CTCGTC +CTCGTG +CTCGCA +CTCGCT +CTCGCC +CTCGCG +CTCGGA +CTCGGT +CTCGGC +CTCGGG +CTGAAA +CTGAAT +CTGAAC +CTGAAG +CTGATA +CTGATT +CTGATC +CTGATG +CTGACA +CTGACT +CTGACC +CTGACG +CTGAGA +CTGAGT +CTGAGC +CTGAGG +CTGTAA +CTGTAT +CTGTAC +CTGTAG +CTGTTA +CTGTTT +CTGTTC +CTGTTG +CTGTCA +CTGTCT +CTGTCC +CTGTCG +CTGTGA +CTGTGT +CTGTGC +CTGTGG +CTGCAA +CTGCAT +CTGCAC +CTGCAG +CTGCTA +CTGCTT +CTGCTC +CTGCTG +CTGCCA +CTGCCT +CTGCCC +CTGCCG +CTGCGA +CTGCGT +CTGCGC +CTGCGG +CTGGAA +CTGGAT +CTGGAC +CTGGAG +CTGGTA +CTGGTT +CTGGTC +CTGGTG +CTGGCA +CTGGCT +CTGGCC +CTGGCG +CTGGGA +CTGGGT +CTGGGC +CTGGGG +CCAAAA +CCAAAT +CCAAAC +CCAAAG +CCAATA +CCAATT +CCAATC +CCAATG +CCAACA +CCAACT +CCAACC +CCAACG +CCAAGA +CCAAGT +CCAAGC +CCAAGG +CCATAA +CCATAT +CCATAC +CCATAG +CCATTA +CCATTT +CCATTC +CCATTG +CCATCA +CCATCT +CCATCC +CCATCG +CCATGA +CCATGT +CCATGC +CCATGG +CCACAA +CCACAT +CCACAC +CCACAG +CCACTA +CCACTT +CCACTC +CCACTG +CCACCA +CCACCT +CCACCC +CCACCG +CCACGA +CCACGT +CCACGC +CCACGG +CCAGAA +CCAGAT +CCAGAC +CCAGAG +CCAGTA +CCAGTT +CCAGTC +CCAGTG +CCAGCA +CCAGCT +CCAGCC +CCAGCG +CCAGGA +CCAGGT +CCAGGC +CCAGGG +CCTAAA +CCTAAT +CCTAAC +CCTAAG +CCTATA +CCTATT +CCTATC +CCTATG +CCTACA +CCTACT +CCTACC +CCTACG +CCTAGA +CCTAGT +CCTAGC +CCTAGG +CCTTAA +CCTTAT +CCTTAC +CCTTAG +CCTTTA +CCTTTT +CCTTTC +CCTTTG +CCTTCA +CCTTCT +CCTTCC +CCTTCG +CCTTGA +CCTTGT +CCTTGC +CCTTGG +CCTCAA +CCTCAT +CCTCAC +CCTCAG +CCTCTA +CCTCTT +CCTCTC +CCTCTG +CCTCCA +CCTCCT +CCTCCC +CCTCCG +CCTCGA +CCTCGT +CCTCGC +CCTCGG +CCTGAA +CCTGAT +CCTGAC +CCTGAG +CCTGTA +CCTGTT +CCTGTC +CCTGTG +CCTGCA +CCTGCT +CCTGCC +CCTGCG +CCTGGA +CCTGGT +CCTGGC +CCTGGG +CCCAAA +CCCAAT +CCCAAC +CCCAAG +CCCATA +CCCATT +CCCATC +CCCATG +CCCACA +CCCACT +CCCACC +CCCACG +CCCAGA +CCCAGT +CCCAGC +CCCAGG +CCCTAA +CCCTAT +CCCTAC +CCCTAG +CCCTTA +CCCTTT +CCCTTC +CCCTTG +CCCTCA +CCCTCT +CCCTCC +CCCTCG +CCCTGA +CCCTGT +CCCTGC +CCCTGG +CCCCAA +CCCCAT +CCCCAC +CCCCAG +CCCCTA +CCCCTT +CCCCTC +CCCCTG +CCCCCA +CCCCCT +CCCCCC +CCCCCG +CCCCGA +CCCCGT +CCCCGC +CCCCGG +CCCGAA +CCCGAT +CCCGAC +CCCGAG +CCCGTA +CCCGTT +CCCGTC +CCCGTG +CCCGCA +CCCGCT +CCCGCC +CCCGCG +CCCGGA +CCCGGT +CCCGGC +CCCGGG +CCGAAA +CCGAAT +CCGAAC +CCGAAG +CCGATA +CCGATT +CCGATC +CCGATG +CCGACA +CCGACT +CCGACC +CCGACG +CCGAGA +CCGAGT +CCGAGC +CCGAGG +CCGTAA +CCGTAT +CCGTAC +CCGTAG +CCGTTA +CCGTTT +CCGTTC +CCGTTG +CCGTCA +CCGTCT +CCGTCC +CCGTCG +CCGTGA +CCGTGT +CCGTGC +CCGTGG +CCGCAA +CCGCAT +CCGCAC +CCGCAG +CCGCTA +CCGCTT +CCGCTC +CCGCTG +CCGCCA +CCGCCT +CCGCCC +CCGCCG +CCGCGA +CCGCGT +CCGCGC +CCGCGG +CCGGAA +CCGGAT +CCGGAC +CCGGAG +CCGGTA +CCGGTT +CCGGTC +CCGGTG +CCGGCA +CCGGCT +CCGGCC +CCGGCG +CCGGGA +CCGGGT +CCGGGC +CCGGGG +CGAAAA +CGAAAT +CGAAAC +CGAAAG +CGAATA +CGAATT +CGAATC +CGAATG +CGAACA +CGAACT +CGAACC +CGAACG +CGAAGA +CGAAGT +CGAAGC +CGAAGG +CGATAA +CGATAT +CGATAC +CGATAG +CGATTA +CGATTT +CGATTC +CGATTG +CGATCA +CGATCT +CGATCC +CGATCG +CGATGA +CGATGT +CGATGC +CGATGG +CGACAA +CGACAT +CGACAC +CGACAG +CGACTA +CGACTT +CGACTC +CGACTG +CGACCA +CGACCT +CGACCC +CGACCG +CGACGA +CGACGT +CGACGC +CGACGG +CGAGAA +CGAGAT +CGAGAC +CGAGAG +CGAGTA +CGAGTT +CGAGTC +CGAGTG +CGAGCA +CGAGCT +CGAGCC +CGAGCG +CGAGGA +CGAGGT +CGAGGC +CGAGGG +CGTAAA +CGTAAT +CGTAAC +CGTAAG +CGTATA +CGTATT +CGTATC +CGTATG +CGTACA +CGTACT +CGTACC +CGTACG +CGTAGA +CGTAGT +CGTAGC +CGTAGG +CGTTAA +CGTTAT +CGTTAC +CGTTAG +CGTTTA +CGTTTT +CGTTTC +CGTTTG +CGTTCA +CGTTCT +CGTTCC +CGTTCG +CGTTGA +CGTTGT +CGTTGC +CGTTGG +CGTCAA +CGTCAT +CGTCAC +CGTCAG +CGTCTA +CGTCTT +CGTCTC +CGTCTG +CGTCCA +CGTCCT +CGTCCC +CGTCCG +CGTCGA +CGTCGT +CGTCGC +CGTCGG +CGTGAA +CGTGAT +CGTGAC +CGTGAG +CGTGTA +CGTGTT +CGTGTC +CGTGTG +CGTGCA +CGTGCT +CGTGCC +CGTGCG +CGTGGA +CGTGGT +CGTGGC +CGTGGG +CGCAAA +CGCAAT +CGCAAC +CGCAAG +CGCATA +CGCATT +CGCATC +CGCATG +CGCACA +CGCACT +CGCACC +CGCACG +CGCAGA +CGCAGT +CGCAGC +CGCAGG +CGCTAA +CGCTAT +CGCTAC +CGCTAG +CGCTTA +CGCTTT +CGCTTC +CGCTTG +CGCTCA +CGCTCT +CGCTCC +CGCTCG +CGCTGA +CGCTGT +CGCTGC +CGCTGG +CGCCAA +CGCCAT +CGCCAC +CGCCAG +CGCCTA +CGCCTT +CGCCTC +CGCCTG +CGCCCA +CGCCCT +CGCCCC +CGCCCG +CGCCGA +CGCCGT +CGCCGC +CGCCGG +CGCGAA +CGCGAT +CGCGAC +CGCGAG +CGCGTA +CGCGTT +CGCGTC +CGCGTG +CGCGCA +CGCGCT +CGCGCC +CGCGCG +CGCGGA +CGCGGT +CGCGGC +CGCGGG +CGGAAA +CGGAAT +CGGAAC +CGGAAG +CGGATA +CGGATT +CGGATC +CGGATG +CGGACA +CGGACT +CGGACC +CGGACG +CGGAGA +CGGAGT +CGGAGC +CGGAGG +CGGTAA +CGGTAT +CGGTAC +CGGTAG +CGGTTA +CGGTTT +CGGTTC +CGGTTG +CGGTCA +CGGTCT +CGGTCC +CGGTCG +CGGTGA +CGGTGT +CGGTGC +CGGTGG +CGGCAA +CGGCAT +CGGCAC +CGGCAG +CGGCTA +CGGCTT +CGGCTC +CGGCTG +CGGCCA +CGGCCT +CGGCCC +CGGCCG +CGGCGA +CGGCGT +CGGCGC +CGGCGG +CGGGAA +CGGGAT +CGGGAC +CGGGAG +CGGGTA +CGGGTT +CGGGTC +CGGGTG +CGGGCA +CGGGCT +CGGGCC +CGGGCG +CGGGGA +CGGGGT +CGGGGC +CGGGGG +GAAAAA +GAAAAT +GAAAAC +GAAAAG +GAAATA +GAAATT +GAAATC +GAAATG +GAAACA +GAAACT +GAAACC +GAAACG +GAAAGA +GAAAGT +GAAAGC +GAAAGG +GAATAA +GAATAT +GAATAC +GAATAG +GAATTA +GAATTT +GAATTC +GAATTG +GAATCA +GAATCT +GAATCC +GAATCG +GAATGA +GAATGT +GAATGC +GAATGG +GAACAA +GAACAT +GAACAC +GAACAG +GAACTA +GAACTT +GAACTC +GAACTG +GAACCA +GAACCT +GAACCC +GAACCG +GAACGA +GAACGT +GAACGC +GAACGG +GAAGAA +GAAGAT +GAAGAC +GAAGAG +GAAGTA +GAAGTT +GAAGTC +GAAGTG +GAAGCA +GAAGCT +GAAGCC +GAAGCG +GAAGGA +GAAGGT +GAAGGC +GAAGGG +GATAAA +GATAAT +GATAAC +GATAAG +GATATA +GATATT +GATATC +GATATG +GATACA +GATACT +GATACC +GATACG +GATAGA +GATAGT +GATAGC +GATAGG +GATTAA +GATTAT +GATTAC +GATTAG +GATTTA +GATTTT +GATTTC +GATTTG +GATTCA +GATTCT +GATTCC +GATTCG +GATTGA +GATTGT +GATTGC +GATTGG +GATCAA +GATCAT +GATCAC +GATCAG +GATCTA +GATCTT +GATCTC +GATCTG +GATCCA +GATCCT +GATCCC +GATCCG +GATCGA +GATCGT +GATCGC +GATCGG +GATGAA +GATGAT +GATGAC +GATGAG +GATGTA +GATGTT +GATGTC +GATGTG +GATGCA +GATGCT +GATGCC +GATGCG +GATGGA +GATGGT +GATGGC +GATGGG +GACAAA +GACAAT +GACAAC +GACAAG +GACATA +GACATT +GACATC +GACATG +GACACA +GACACT +GACACC +GACACG +GACAGA +GACAGT +GACAGC +GACAGG +GACTAA +GACTAT +GACTAC +GACTAG +GACTTA +GACTTT +GACTTC +GACTTG +GACTCA +GACTCT +GACTCC +GACTCG +GACTGA +GACTGT +GACTGC +GACTGG +GACCAA +GACCAT +GACCAC +GACCAG +GACCTA +GACCTT +GACCTC +GACCTG +GACCCA +GACCCT +GACCCC +GACCCG +GACCGA +GACCGT +GACCGC +GACCGG +GACGAA +GACGAT +GACGAC +GACGAG +GACGTA +GACGTT +GACGTC +GACGTG +GACGCA +GACGCT +GACGCC +GACGCG +GACGGA +GACGGT +GACGGC +GACGGG +GAGAAA +GAGAAT +GAGAAC +GAGAAG +GAGATA +GAGATT +GAGATC +GAGATG +GAGACA +GAGACT +GAGACC +GAGACG +GAGAGA +GAGAGT +GAGAGC +GAGAGG +GAGTAA +GAGTAT +GAGTAC +GAGTAG +GAGTTA +GAGTTT +GAGTTC +GAGTTG +GAGTCA +GAGTCT +GAGTCC +GAGTCG +GAGTGA +GAGTGT +GAGTGC +GAGTGG +GAGCAA +GAGCAT +GAGCAC +GAGCAG +GAGCTA +GAGCTT +GAGCTC +GAGCTG +GAGCCA +GAGCCT +GAGCCC +GAGCCG +GAGCGA +GAGCGT +GAGCGC +GAGCGG +GAGGAA +GAGGAT +GAGGAC +GAGGAG +GAGGTA +GAGGTT +GAGGTC +GAGGTG +GAGGCA +GAGGCT +GAGGCC +GAGGCG +GAGGGA +GAGGGT +GAGGGC +GAGGGG +GTAAAA +GTAAAT +GTAAAC +GTAAAG +GTAATA +GTAATT +GTAATC +GTAATG +GTAACA +GTAACT +GTAACC +GTAACG +GTAAGA +GTAAGT +GTAAGC +GTAAGG +GTATAA +GTATAT +GTATAC +GTATAG +GTATTA +GTATTT +GTATTC +GTATTG +GTATCA +GTATCT +GTATCC +GTATCG +GTATGA +GTATGT +GTATGC +GTATGG +GTACAA +GTACAT +GTACAC +GTACAG +GTACTA +GTACTT +GTACTC +GTACTG +GTACCA +GTACCT +GTACCC +GTACCG +GTACGA +GTACGT +GTACGC +GTACGG +GTAGAA +GTAGAT +GTAGAC +GTAGAG +GTAGTA +GTAGTT +GTAGTC +GTAGTG +GTAGCA +GTAGCT +GTAGCC +GTAGCG +GTAGGA +GTAGGT +GTAGGC +GTAGGG +GTTAAA +GTTAAT +GTTAAC +GTTAAG +GTTATA +GTTATT +GTTATC +GTTATG +GTTACA +GTTACT +GTTACC +GTTACG +GTTAGA +GTTAGT +GTTAGC +GTTAGG +GTTTAA +GTTTAT +GTTTAC +GTTTAG +GTTTTA +GTTTTT +GTTTTC +GTTTTG +GTTTCA +GTTTCT +GTTTCC +GTTTCG +GTTTGA +GTTTGT +GTTTGC +GTTTGG +GTTCAA +GTTCAT +GTTCAC +GTTCAG +GTTCTA +GTTCTT +GTTCTC +GTTCTG +GTTCCA +GTTCCT +GTTCCC +GTTCCG +GTTCGA +GTTCGT +GTTCGC +GTTCGG +GTTGAA +GTTGAT +GTTGAC +GTTGAG +GTTGTA +GTTGTT +GTTGTC +GTTGTG +GTTGCA +GTTGCT +GTTGCC +GTTGCG +GTTGGA +GTTGGT +GTTGGC +GTTGGG +GTCAAA +GTCAAT +GTCAAC +GTCAAG +GTCATA +GTCATT +GTCATC +GTCATG +GTCACA +GTCACT +GTCACC +GTCACG +GTCAGA +GTCAGT +GTCAGC +GTCAGG +GTCTAA +GTCTAT +GTCTAC +GTCTAG +GTCTTA +GTCTTT +GTCTTC +GTCTTG +GTCTCA +GTCTCT +GTCTCC +GTCTCG +GTCTGA +GTCTGT +GTCTGC +GTCTGG +GTCCAA +GTCCAT +GTCCAC +GTCCAG +GTCCTA +GTCCTT +GTCCTC +GTCCTG +GTCCCA +GTCCCT +GTCCCC +GTCCCG +GTCCGA +GTCCGT +GTCCGC +GTCCGG +GTCGAA +GTCGAT +GTCGAC +GTCGAG +GTCGTA +GTCGTT +GTCGTC +GTCGTG +GTCGCA +GTCGCT +GTCGCC +GTCGCG +GTCGGA +GTCGGT +GTCGGC +GTCGGG +GTGAAA +GTGAAT +GTGAAC +GTGAAG +GTGATA +GTGATT +GTGATC +GTGATG +GTGACA +GTGACT +GTGACC +GTGACG +GTGAGA +GTGAGT +GTGAGC +GTGAGG +GTGTAA +GTGTAT +GTGTAC +GTGTAG +GTGTTA +GTGTTT +GTGTTC +GTGTTG +GTGTCA +GTGTCT +GTGTCC +GTGTCG +GTGTGA +GTGTGT +GTGTGC +GTGTGG +GTGCAA +GTGCAT +GTGCAC +GTGCAG +GTGCTA +GTGCTT +GTGCTC +GTGCTG +GTGCCA +GTGCCT +GTGCCC +GTGCCG +GTGCGA +GTGCGT +GTGCGC +GTGCGG +GTGGAA +GTGGAT +GTGGAC +GTGGAG +GTGGTA +GTGGTT +GTGGTC +GTGGTG +GTGGCA +GTGGCT +GTGGCC +GTGGCG +GTGGGA +GTGGGT +GTGGGC +GTGGGG +GCAAAA +GCAAAT +GCAAAC +GCAAAG +GCAATA +GCAATT +GCAATC +GCAATG +GCAACA +GCAACT +GCAACC +GCAACG +GCAAGA +GCAAGT +GCAAGC +GCAAGG +GCATAA +GCATAT +GCATAC +GCATAG +GCATTA +GCATTT +GCATTC +GCATTG +GCATCA +GCATCT +GCATCC +GCATCG +GCATGA +GCATGT +GCATGC +GCATGG +GCACAA +GCACAT +GCACAC +GCACAG +GCACTA +GCACTT +GCACTC +GCACTG +GCACCA +GCACCT +GCACCC +GCACCG +GCACGA +GCACGT +GCACGC +GCACGG +GCAGAA +GCAGAT +GCAGAC +GCAGAG +GCAGTA +GCAGTT +GCAGTC +GCAGTG +GCAGCA +GCAGCT +GCAGCC +GCAGCG +GCAGGA +GCAGGT +GCAGGC +GCAGGG +GCTAAA +GCTAAT +GCTAAC +GCTAAG +GCTATA +GCTATT +GCTATC +GCTATG +GCTACA +GCTACT +GCTACC +GCTACG +GCTAGA +GCTAGT +GCTAGC +GCTAGG +GCTTAA +GCTTAT +GCTTAC +GCTTAG +GCTTTA +GCTTTT +GCTTTC +GCTTTG +GCTTCA +GCTTCT +GCTTCC +GCTTCG +GCTTGA +GCTTGT +GCTTGC +GCTTGG +GCTCAA +GCTCAT +GCTCAC +GCTCAG +GCTCTA +GCTCTT +GCTCTC +GCTCTG +GCTCCA +GCTCCT +GCTCCC +GCTCCG +GCTCGA +GCTCGT +GCTCGC +GCTCGG +GCTGAA +GCTGAT +GCTGAC +GCTGAG +GCTGTA +GCTGTT +GCTGTC +GCTGTG +GCTGCA +GCTGCT +GCTGCC +GCTGCG +GCTGGA +GCTGGT +GCTGGC +GCTGGG +GCCAAA +GCCAAT +GCCAAC +GCCAAG +GCCATA +GCCATT +GCCATC +GCCATG +GCCACA +GCCACT +GCCACC +GCCACG +GCCAGA +GCCAGT +GCCAGC +GCCAGG +GCCTAA +GCCTAT +GCCTAC +GCCTAG +GCCTTA +GCCTTT +GCCTTC +GCCTTG +GCCTCA +GCCTCT +GCCTCC +GCCTCG +GCCTGA +GCCTGT +GCCTGC +GCCTGG +GCCCAA +GCCCAT +GCCCAC +GCCCAG +GCCCTA +GCCCTT +GCCCTC +GCCCTG +GCCCCA +GCCCCT +GCCCCC +GCCCCG +GCCCGA +GCCCGT +GCCCGC +GCCCGG +GCCGAA +GCCGAT +GCCGAC +GCCGAG +GCCGTA +GCCGTT +GCCGTC +GCCGTG +GCCGCA +GCCGCT +GCCGCC +GCCGCG +GCCGGA +GCCGGT +GCCGGC +GCCGGG +GCGAAA +GCGAAT +GCGAAC +GCGAAG +GCGATA +GCGATT +GCGATC +GCGATG +GCGACA +GCGACT +GCGACC +GCGACG +GCGAGA +GCGAGT +GCGAGC +GCGAGG +GCGTAA +GCGTAT +GCGTAC +GCGTAG +GCGTTA +GCGTTT +GCGTTC +GCGTTG +GCGTCA +GCGTCT +GCGTCC +GCGTCG +GCGTGA +GCGTGT +GCGTGC +GCGTGG +GCGCAA +GCGCAT +GCGCAC +GCGCAG +GCGCTA +GCGCTT +GCGCTC +GCGCTG +GCGCCA +GCGCCT +GCGCCC +GCGCCG +GCGCGA +GCGCGT +GCGCGC +GCGCGG +GCGGAA +GCGGAT +GCGGAC +GCGGAG +GCGGTA +GCGGTT +GCGGTC +GCGGTG +GCGGCA +GCGGCT +GCGGCC +GCGGCG +GCGGGA +GCGGGT +GCGGGC +GCGGGG +GGAAAA +GGAAAT +GGAAAC +GGAAAG +GGAATA +GGAATT +GGAATC +GGAATG +GGAACA +GGAACT +GGAACC +GGAACG +GGAAGA +GGAAGT +GGAAGC +GGAAGG +GGATAA +GGATAT +GGATAC +GGATAG +GGATTA +GGATTT +GGATTC +GGATTG +GGATCA +GGATCT +GGATCC +GGATCG +GGATGA +GGATGT +GGATGC +GGATGG +GGACAA +GGACAT +GGACAC +GGACAG +GGACTA +GGACTT +GGACTC +GGACTG +GGACCA +GGACCT +GGACCC +GGACCG +GGACGA +GGACGT +GGACGC +GGACGG +GGAGAA +GGAGAT +GGAGAC +GGAGAG +GGAGTA +GGAGTT +GGAGTC +GGAGTG +GGAGCA +GGAGCT +GGAGCC +GGAGCG +GGAGGA +GGAGGT +GGAGGC +GGAGGG +GGTAAA +GGTAAT +GGTAAC +GGTAAG +GGTATA +GGTATT +GGTATC +GGTATG +GGTACA +GGTACT +GGTACC +GGTACG +GGTAGA +GGTAGT +GGTAGC +GGTAGG +GGTTAA +GGTTAT +GGTTAC +GGTTAG +GGTTTA +GGTTTT +GGTTTC +GGTTTG +GGTTCA +GGTTCT +GGTTCC +GGTTCG +GGTTGA +GGTTGT +GGTTGC +GGTTGG +GGTCAA +GGTCAT +GGTCAC +GGTCAG +GGTCTA +GGTCTT +GGTCTC +GGTCTG +GGTCCA +GGTCCT +GGTCCC +GGTCCG +GGTCGA +GGTCGT +GGTCGC +GGTCGG +GGTGAA +GGTGAT +GGTGAC +GGTGAG +GGTGTA +GGTGTT +GGTGTC +GGTGTG +GGTGCA +GGTGCT +GGTGCC +GGTGCG +GGTGGA +GGTGGT +GGTGGC +GGTGGG +GGCAAA +GGCAAT +GGCAAC +GGCAAG +GGCATA +GGCATT +GGCATC +GGCATG +GGCACA +GGCACT +GGCACC +GGCACG +GGCAGA +GGCAGT +GGCAGC +GGCAGG +GGCTAA +GGCTAT +GGCTAC +GGCTAG +GGCTTA +GGCTTT +GGCTTC +GGCTTG +GGCTCA +GGCTCT +GGCTCC +GGCTCG +GGCTGA +GGCTGT +GGCTGC +GGCTGG +GGCCAA +GGCCAT +GGCCAC +GGCCAG +GGCCTA +GGCCTT +GGCCTC +GGCCTG +GGCCCA +GGCCCT +GGCCCC +GGCCCG +GGCCGA +GGCCGT +GGCCGC +GGCCGG +GGCGAA +GGCGAT +GGCGAC +GGCGAG +GGCGTA +GGCGTT +GGCGTC +GGCGTG +GGCGCA +GGCGCT +GGCGCC +GGCGCG +GGCGGA +GGCGGT +GGCGGC +GGCGGG +GGGAAA +GGGAAT +GGGAAC +GGGAAG +GGGATA +GGGATT +GGGATC +GGGATG +GGGACA +GGGACT +GGGACC +GGGACG +GGGAGA +GGGAGT +GGGAGC +GGGAGG +GGGTAA +GGGTAT +GGGTAC +GGGTAG +GGGTTA +GGGTTT +GGGTTC +GGGTTG +GGGTCA +GGGTCT +GGGTCC +GGGTCG +GGGTGA +GGGTGT +GGGTGC +GGGTGG +GGGCAA +GGGCAT +GGGCAC +GGGCAG +GGGCTA +GGGCTT +GGGCTC +GGGCTG +GGGCCA +GGGCCT +GGGCCC +GGGCCG +GGGCGA +GGGCGT +GGGCGC +GGGCGG +GGGGAA +GGGGAT +GGGGAC +GGGGAG +GGGGTA +GGGGTT +GGGGTC +GGGGTG +GGGGCA +GGGGCT +GGGGCC +GGGGCG +GGGGGA +GGGGGT +GGGGGC +GGGGGG diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-3000/config.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-3000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..9b14e61d327bf586fcff6e817cf9d365bae18ad2 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-3000/config.json @@ -0,0 +1,26 @@ +{ + "_name_or_path": "fabihamakhdoomi/TinyDNABERT", + "architectures": [ + "BertForSequenceClassification" + ], + "attention_probs_dropout_prob": 0.1, + "classifier_dropout": null, + "hidden_act": "gelu", + "hidden_dropout_prob": 0.1, + "hidden_size": 128, + "initializer_range": 0.02, + "intermediate_size": 512, + "layer_norm_eps": 1e-12, + "max_position_embeddings": 512, + "model_type": "bert", + "num_attention_heads": 2, + "num_hidden_layers": 2, + "pad_token_id": 0, + "position_embedding_type": "absolute", + "problem_type": "single_label_classification", + "torch_dtype": "float32", + "transformers_version": "4.44.0", + "type_vocab_size": 2, + "use_cache": true, + "vocab_size": 4101 +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-3000/model.safetensors b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-3000/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..632c583bedc906ef4a79a43788c5503883d6450e --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-3000/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:87d420ab7904755333083014d32d03e4bb8706a9be50db436ccbcfec4d94d0ba +size 4021688 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-3000/optimizer.pt b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-3000/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..73953aebdd792c6233d8c1f75f79e81dbd0d5945 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-3000/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0a4e8c1ca516bd9bbf8eadc084b65801864b2ce69a1e017b08c7f9c4259d103e +size 8058700 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-3000/rng_state.pth b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-3000/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..6a219fb332ae976c7a91fd53a363fd52b1ca4044 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-3000/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9a88049e3e15cef9ae24c0067d91d8e8d16ed125ffa58bfc970554dbe3e0b701 +size 14244 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-3000/scheduler.pt b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-3000/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..6c655a9f3dd5d4d532f6c0a84b19b9ae5b766437 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-3000/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dbc2253b3cc1ddd11640ea56d3925f925017f9aa6e1577411dac42d629457af6 +size 1064 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-3000/special_tokens_map.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-3000/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..a8b3208c2884c4efb86e49300fdd3dc877220cdf --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-3000/special_tokens_map.json @@ -0,0 +1,7 @@ +{ + "cls_token": "[CLS]", + "mask_token": "[MASK]", + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-3000/tokenizer.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-3000/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..ebd782de6963b9032ce9374cfe30f5275be1a30b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-3000/tokenizer.json @@ -0,0 +1,4265 @@ +{ + "version": "1.0", + "truncation": { + "direction": "Right", + "max_length": 512, + "strategy": "LongestFirst", + "stride": 0 + }, + "padding": { + "strategy": { + "Fixed": 512 + }, + "direction": "Right", + "pad_to_multiple_of": null, + "pad_id": 0, + "pad_type_id": 0, + "pad_token": "[PAD]" + }, + "added_tokens": [ + { + "id": 0, + "content": "[PAD]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 1, + "content": "[UNK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 2, + "content": "[CLS]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 3, + "content": "[SEP]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 4, + "content": "[MASK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + } + ], + "normalizer": { + "type": "BertNormalizer", + "clean_text": true, + "handle_chinese_chars": true, + "strip_accents": null, + "lowercase": false + }, + "pre_tokenizer": { + "type": "BertPreTokenizer" + }, + "post_processor": { + "type": "TemplateProcessing", + "single": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + } + ], + "pair": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "B", + "type_id": 1 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 1 + } + } + ], + "special_tokens": { + "[CLS]": { + "id": "[CLS]", + "ids": [ + 2 + ], + "tokens": [ + "[CLS]" + ] + }, + "[SEP]": { + "id": "[SEP]", + "ids": [ + 3 + ], + "tokens": [ + "[SEP]" + ] + } + } + }, + "decoder": { + "type": "WordPiece", + "prefix": "##", + "cleanup": true + }, + "model": { + "type": "WordPiece", + "unk_token": "[UNK]", + "continuing_subword_prefix": "##", + "max_input_chars_per_word": 100, + "vocab": { + "[PAD]": 0, + "[UNK]": 1, + "[CLS]": 2, + "[SEP]": 3, + "[MASK]": 4, + "AAAAAA": 5, + "AAAAAT": 6, + "AAAAAC": 7, + "AAAAAG": 8, + "AAAATA": 9, + "AAAATT": 10, + "AAAATC": 11, + "AAAATG": 12, + "AAAACA": 13, + "AAAACT": 14, + "AAAACC": 15, + "AAAACG": 16, + "AAAAGA": 17, + "AAAAGT": 18, + "AAAAGC": 19, + "AAAAGG": 20, + "AAATAA": 21, + "AAATAT": 22, + "AAATAC": 23, + "AAATAG": 24, + "AAATTA": 25, + "AAATTT": 26, + "AAATTC": 27, + "AAATTG": 28, + "AAATCA": 29, + "AAATCT": 30, + "AAATCC": 31, + "AAATCG": 32, + "AAATGA": 33, + "AAATGT": 34, + "AAATGC": 35, + "AAATGG": 36, + "AAACAA": 37, + "AAACAT": 38, + "AAACAC": 39, + "AAACAG": 40, + "AAACTA": 41, + "AAACTT": 42, + "AAACTC": 43, + "AAACTG": 44, + "AAACCA": 45, + "AAACCT": 46, + "AAACCC": 47, + "AAACCG": 48, + "AAACGA": 49, + "AAACGT": 50, + "AAACGC": 51, + "AAACGG": 52, + "AAAGAA": 53, + "AAAGAT": 54, + "AAAGAC": 55, + "AAAGAG": 56, + "AAAGTA": 57, + "AAAGTT": 58, + "AAAGTC": 59, + "AAAGTG": 60, + "AAAGCA": 61, + "AAAGCT": 62, + "AAAGCC": 63, + "AAAGCG": 64, + "AAAGGA": 65, + "AAAGGT": 66, + "AAAGGC": 67, + "AAAGGG": 68, + "AATAAA": 69, + "AATAAT": 70, + "AATAAC": 71, + "AATAAG": 72, + "AATATA": 73, + "AATATT": 74, + "AATATC": 75, + "AATATG": 76, + "AATACA": 77, + "AATACT": 78, + "AATACC": 79, + "AATACG": 80, + "AATAGA": 81, + "AATAGT": 82, + "AATAGC": 83, + "AATAGG": 84, + "AATTAA": 85, + "AATTAT": 86, + "AATTAC": 87, + "AATTAG": 88, + "AATTTA": 89, + "AATTTT": 90, + "AATTTC": 91, + "AATTTG": 92, + "AATTCA": 93, + "AATTCT": 94, + "AATTCC": 95, + "AATTCG": 96, + "AATTGA": 97, + "AATTGT": 98, + "AATTGC": 99, + "AATTGG": 100, + "AATCAA": 101, + "AATCAT": 102, + "AATCAC": 103, + "AATCAG": 104, + "AATCTA": 105, + "AATCTT": 106, + "AATCTC": 107, + "AATCTG": 108, + "AATCCA": 109, + "AATCCT": 110, + "AATCCC": 111, + "AATCCG": 112, + "AATCGA": 113, + "AATCGT": 114, + "AATCGC": 115, + "AATCGG": 116, + "AATGAA": 117, + "AATGAT": 118, + "AATGAC": 119, + "AATGAG": 120, + "AATGTA": 121, + "AATGTT": 122, + "AATGTC": 123, + "AATGTG": 124, + "AATGCA": 125, + "AATGCT": 126, + "AATGCC": 127, + "AATGCG": 128, + "AATGGA": 129, + "AATGGT": 130, + "AATGGC": 131, + "AATGGG": 132, + "AACAAA": 133, + "AACAAT": 134, + "AACAAC": 135, + "AACAAG": 136, + "AACATA": 137, + "AACATT": 138, + "AACATC": 139, + "AACATG": 140, + "AACACA": 141, + "AACACT": 142, + "AACACC": 143, + "AACACG": 144, + "AACAGA": 145, + "AACAGT": 146, + "AACAGC": 147, + "AACAGG": 148, + "AACTAA": 149, + "AACTAT": 150, + "AACTAC": 151, + "AACTAG": 152, + "AACTTA": 153, + "AACTTT": 154, + "AACTTC": 155, + "AACTTG": 156, + "AACTCA": 157, + "AACTCT": 158, + "AACTCC": 159, + "AACTCG": 160, + "AACTGA": 161, + "AACTGT": 162, + "AACTGC": 163, + "AACTGG": 164, + "AACCAA": 165, + "AACCAT": 166, + "AACCAC": 167, + "AACCAG": 168, + "AACCTA": 169, + "AACCTT": 170, + "AACCTC": 171, + "AACCTG": 172, + "AACCCA": 173, + "AACCCT": 174, + "AACCCC": 175, + "AACCCG": 176, + "AACCGA": 177, + "AACCGT": 178, + "AACCGC": 179, + "AACCGG": 180, + "AACGAA": 181, + "AACGAT": 182, + "AACGAC": 183, + "AACGAG": 184, + "AACGTA": 185, + "AACGTT": 186, + "AACGTC": 187, + "AACGTG": 188, + "AACGCA": 189, + "AACGCT": 190, + "AACGCC": 191, + "AACGCG": 192, + "AACGGA": 193, + "AACGGT": 194, + "AACGGC": 195, + "AACGGG": 196, + "AAGAAA": 197, + "AAGAAT": 198, + "AAGAAC": 199, + "AAGAAG": 200, + "AAGATA": 201, + "AAGATT": 202, + "AAGATC": 203, + "AAGATG": 204, + "AAGACA": 205, + "AAGACT": 206, + "AAGACC": 207, + "AAGACG": 208, + "AAGAGA": 209, + "AAGAGT": 210, + "AAGAGC": 211, + "AAGAGG": 212, + "AAGTAA": 213, + "AAGTAT": 214, + "AAGTAC": 215, + "AAGTAG": 216, + "AAGTTA": 217, + "AAGTTT": 218, + "AAGTTC": 219, + "AAGTTG": 220, + "AAGTCA": 221, + "AAGTCT": 222, + "AAGTCC": 223, + "AAGTCG": 224, + "AAGTGA": 225, + "AAGTGT": 226, + "AAGTGC": 227, + "AAGTGG": 228, + "AAGCAA": 229, + "AAGCAT": 230, + "AAGCAC": 231, + "AAGCAG": 232, + "AAGCTA": 233, + "AAGCTT": 234, + "AAGCTC": 235, + "AAGCTG": 236, + "AAGCCA": 237, + "AAGCCT": 238, + "AAGCCC": 239, + "AAGCCG": 240, + "AAGCGA": 241, + "AAGCGT": 242, + "AAGCGC": 243, + "AAGCGG": 244, + "AAGGAA": 245, + "AAGGAT": 246, + "AAGGAC": 247, + "AAGGAG": 248, + "AAGGTA": 249, + "AAGGTT": 250, + "AAGGTC": 251, + "AAGGTG": 252, + "AAGGCA": 253, + "AAGGCT": 254, + "AAGGCC": 255, + "AAGGCG": 256, + "AAGGGA": 257, + "AAGGGT": 258, + "AAGGGC": 259, + "AAGGGG": 260, + "ATAAAA": 261, + "ATAAAT": 262, + "ATAAAC": 263, + "ATAAAG": 264, + "ATAATA": 265, + "ATAATT": 266, + "ATAATC": 267, + "ATAATG": 268, + "ATAACA": 269, + "ATAACT": 270, + "ATAACC": 271, + "ATAACG": 272, + "ATAAGA": 273, + "ATAAGT": 274, + "ATAAGC": 275, + "ATAAGG": 276, + "ATATAA": 277, + "ATATAT": 278, + "ATATAC": 279, + "ATATAG": 280, + "ATATTA": 281, + "ATATTT": 282, + "ATATTC": 283, + "ATATTG": 284, + "ATATCA": 285, + "ATATCT": 286, + "ATATCC": 287, + "ATATCG": 288, + "ATATGA": 289, + "ATATGT": 290, + "ATATGC": 291, + "ATATGG": 292, + "ATACAA": 293, + "ATACAT": 294, + "ATACAC": 295, + "ATACAG": 296, + "ATACTA": 297, + "ATACTT": 298, + "ATACTC": 299, + "ATACTG": 300, + "ATACCA": 301, + "ATACCT": 302, + "ATACCC": 303, + "ATACCG": 304, + "ATACGA": 305, + "ATACGT": 306, + "ATACGC": 307, + "ATACGG": 308, + "ATAGAA": 309, + "ATAGAT": 310, + "ATAGAC": 311, + "ATAGAG": 312, + "ATAGTA": 313, + "ATAGTT": 314, + "ATAGTC": 315, + "ATAGTG": 316, + "ATAGCA": 317, + "ATAGCT": 318, + "ATAGCC": 319, + "ATAGCG": 320, + "ATAGGA": 321, + "ATAGGT": 322, + "ATAGGC": 323, + "ATAGGG": 324, + "ATTAAA": 325, + "ATTAAT": 326, + "ATTAAC": 327, + "ATTAAG": 328, + "ATTATA": 329, + "ATTATT": 330, + "ATTATC": 331, + "ATTATG": 332, + "ATTACA": 333, + "ATTACT": 334, + "ATTACC": 335, + "ATTACG": 336, + "ATTAGA": 337, + "ATTAGT": 338, + "ATTAGC": 339, + "ATTAGG": 340, + "ATTTAA": 341, + "ATTTAT": 342, + "ATTTAC": 343, + "ATTTAG": 344, + "ATTTTA": 345, + "ATTTTT": 346, + "ATTTTC": 347, + "ATTTTG": 348, + "ATTTCA": 349, + "ATTTCT": 350, + "ATTTCC": 351, + "ATTTCG": 352, + "ATTTGA": 353, + "ATTTGT": 354, + "ATTTGC": 355, + "ATTTGG": 356, + "ATTCAA": 357, + "ATTCAT": 358, + "ATTCAC": 359, + "ATTCAG": 360, + "ATTCTA": 361, + "ATTCTT": 362, + "ATTCTC": 363, + "ATTCTG": 364, + "ATTCCA": 365, + "ATTCCT": 366, + "ATTCCC": 367, + "ATTCCG": 368, + "ATTCGA": 369, + "ATTCGT": 370, + "ATTCGC": 371, + "ATTCGG": 372, + "ATTGAA": 373, + "ATTGAT": 374, + "ATTGAC": 375, + "ATTGAG": 376, + "ATTGTA": 377, + "ATTGTT": 378, + "ATTGTC": 379, + "ATTGTG": 380, + "ATTGCA": 381, + "ATTGCT": 382, + "ATTGCC": 383, + "ATTGCG": 384, + "ATTGGA": 385, + "ATTGGT": 386, + "ATTGGC": 387, + "ATTGGG": 388, + "ATCAAA": 389, + "ATCAAT": 390, + "ATCAAC": 391, + "ATCAAG": 392, + "ATCATA": 393, + "ATCATT": 394, + "ATCATC": 395, + "ATCATG": 396, + "ATCACA": 397, + "ATCACT": 398, + "ATCACC": 399, + "ATCACG": 400, + "ATCAGA": 401, + "ATCAGT": 402, + "ATCAGC": 403, + "ATCAGG": 404, + "ATCTAA": 405, + "ATCTAT": 406, + "ATCTAC": 407, + "ATCTAG": 408, + "ATCTTA": 409, + "ATCTTT": 410, + "ATCTTC": 411, + "ATCTTG": 412, + "ATCTCA": 413, + "ATCTCT": 414, + "ATCTCC": 415, + "ATCTCG": 416, + "ATCTGA": 417, + "ATCTGT": 418, + "ATCTGC": 419, + "ATCTGG": 420, + "ATCCAA": 421, + "ATCCAT": 422, + "ATCCAC": 423, + "ATCCAG": 424, + "ATCCTA": 425, + "ATCCTT": 426, + "ATCCTC": 427, + "ATCCTG": 428, + "ATCCCA": 429, + "ATCCCT": 430, + "ATCCCC": 431, + "ATCCCG": 432, + "ATCCGA": 433, + "ATCCGT": 434, + "ATCCGC": 435, + "ATCCGG": 436, + "ATCGAA": 437, + "ATCGAT": 438, + "ATCGAC": 439, + "ATCGAG": 440, + "ATCGTA": 441, + "ATCGTT": 442, + "ATCGTC": 443, + "ATCGTG": 444, + "ATCGCA": 445, + "ATCGCT": 446, + "ATCGCC": 447, + "ATCGCG": 448, + "ATCGGA": 449, + "ATCGGT": 450, + "ATCGGC": 451, + "ATCGGG": 452, + "ATGAAA": 453, + "ATGAAT": 454, + "ATGAAC": 455, + "ATGAAG": 456, + "ATGATA": 457, + "ATGATT": 458, + "ATGATC": 459, + "ATGATG": 460, + "ATGACA": 461, + "ATGACT": 462, + "ATGACC": 463, + "ATGACG": 464, + "ATGAGA": 465, + "ATGAGT": 466, + "ATGAGC": 467, + "ATGAGG": 468, + "ATGTAA": 469, + "ATGTAT": 470, + "ATGTAC": 471, + "ATGTAG": 472, + "ATGTTA": 473, + "ATGTTT": 474, + "ATGTTC": 475, + "ATGTTG": 476, + "ATGTCA": 477, + "ATGTCT": 478, + "ATGTCC": 479, + "ATGTCG": 480, + "ATGTGA": 481, + "ATGTGT": 482, + "ATGTGC": 483, + "ATGTGG": 484, + "ATGCAA": 485, + "ATGCAT": 486, + "ATGCAC": 487, + "ATGCAG": 488, + "ATGCTA": 489, + "ATGCTT": 490, + "ATGCTC": 491, + "ATGCTG": 492, + "ATGCCA": 493, + "ATGCCT": 494, + "ATGCCC": 495, + "ATGCCG": 496, + "ATGCGA": 497, + "ATGCGT": 498, + "ATGCGC": 499, + "ATGCGG": 500, + "ATGGAA": 501, + "ATGGAT": 502, + "ATGGAC": 503, + "ATGGAG": 504, + "ATGGTA": 505, + "ATGGTT": 506, + "ATGGTC": 507, + "ATGGTG": 508, + "ATGGCA": 509, + "ATGGCT": 510, + "ATGGCC": 511, + "ATGGCG": 512, + "ATGGGA": 513, + "ATGGGT": 514, + "ATGGGC": 515, + "ATGGGG": 516, + "ACAAAA": 517, + "ACAAAT": 518, + "ACAAAC": 519, + "ACAAAG": 520, + "ACAATA": 521, + "ACAATT": 522, + "ACAATC": 523, + "ACAATG": 524, + "ACAACA": 525, + "ACAACT": 526, + "ACAACC": 527, + "ACAACG": 528, + "ACAAGA": 529, + "ACAAGT": 530, + "ACAAGC": 531, + "ACAAGG": 532, + "ACATAA": 533, + "ACATAT": 534, + "ACATAC": 535, + "ACATAG": 536, + "ACATTA": 537, + "ACATTT": 538, + "ACATTC": 539, + "ACATTG": 540, + "ACATCA": 541, + "ACATCT": 542, + "ACATCC": 543, + "ACATCG": 544, + "ACATGA": 545, + "ACATGT": 546, + "ACATGC": 547, + "ACATGG": 548, + "ACACAA": 549, + "ACACAT": 550, + "ACACAC": 551, + "ACACAG": 552, + "ACACTA": 553, + "ACACTT": 554, + "ACACTC": 555, + "ACACTG": 556, + "ACACCA": 557, + "ACACCT": 558, + "ACACCC": 559, + "ACACCG": 560, + "ACACGA": 561, + "ACACGT": 562, + "ACACGC": 563, + "ACACGG": 564, + "ACAGAA": 565, + "ACAGAT": 566, + "ACAGAC": 567, + "ACAGAG": 568, + "ACAGTA": 569, + "ACAGTT": 570, + "ACAGTC": 571, + "ACAGTG": 572, + "ACAGCA": 573, + "ACAGCT": 574, + "ACAGCC": 575, + "ACAGCG": 576, + "ACAGGA": 577, + "ACAGGT": 578, + "ACAGGC": 579, + "ACAGGG": 580, + "ACTAAA": 581, + "ACTAAT": 582, + "ACTAAC": 583, + "ACTAAG": 584, + "ACTATA": 585, + "ACTATT": 586, + "ACTATC": 587, + "ACTATG": 588, + "ACTACA": 589, + "ACTACT": 590, + "ACTACC": 591, + "ACTACG": 592, + "ACTAGA": 593, + "ACTAGT": 594, + "ACTAGC": 595, + "ACTAGG": 596, + "ACTTAA": 597, + "ACTTAT": 598, + "ACTTAC": 599, + "ACTTAG": 600, + "ACTTTA": 601, + "ACTTTT": 602, + "ACTTTC": 603, + "ACTTTG": 604, + "ACTTCA": 605, + "ACTTCT": 606, + "ACTTCC": 607, + "ACTTCG": 608, + "ACTTGA": 609, + "ACTTGT": 610, + "ACTTGC": 611, + "ACTTGG": 612, + "ACTCAA": 613, + "ACTCAT": 614, + "ACTCAC": 615, + "ACTCAG": 616, + "ACTCTA": 617, + "ACTCTT": 618, + "ACTCTC": 619, + "ACTCTG": 620, + "ACTCCA": 621, + "ACTCCT": 622, + "ACTCCC": 623, + "ACTCCG": 624, + "ACTCGA": 625, + "ACTCGT": 626, + "ACTCGC": 627, + "ACTCGG": 628, + "ACTGAA": 629, + "ACTGAT": 630, + "ACTGAC": 631, + "ACTGAG": 632, + "ACTGTA": 633, + "ACTGTT": 634, + "ACTGTC": 635, + "ACTGTG": 636, + "ACTGCA": 637, + "ACTGCT": 638, + "ACTGCC": 639, + "ACTGCG": 640, + "ACTGGA": 641, + "ACTGGT": 642, + "ACTGGC": 643, + "ACTGGG": 644, + "ACCAAA": 645, + "ACCAAT": 646, + "ACCAAC": 647, + "ACCAAG": 648, + "ACCATA": 649, + "ACCATT": 650, + "ACCATC": 651, + "ACCATG": 652, + "ACCACA": 653, + "ACCACT": 654, + "ACCACC": 655, + "ACCACG": 656, + "ACCAGA": 657, + "ACCAGT": 658, + "ACCAGC": 659, + "ACCAGG": 660, + "ACCTAA": 661, + "ACCTAT": 662, + "ACCTAC": 663, + "ACCTAG": 664, + "ACCTTA": 665, + "ACCTTT": 666, + "ACCTTC": 667, + "ACCTTG": 668, + "ACCTCA": 669, + "ACCTCT": 670, + "ACCTCC": 671, + "ACCTCG": 672, + "ACCTGA": 673, + "ACCTGT": 674, + "ACCTGC": 675, + "ACCTGG": 676, + "ACCCAA": 677, + "ACCCAT": 678, + "ACCCAC": 679, + "ACCCAG": 680, + "ACCCTA": 681, + "ACCCTT": 682, + "ACCCTC": 683, + "ACCCTG": 684, + "ACCCCA": 685, + "ACCCCT": 686, + "ACCCCC": 687, + "ACCCCG": 688, + "ACCCGA": 689, + "ACCCGT": 690, + "ACCCGC": 691, + "ACCCGG": 692, + "ACCGAA": 693, + "ACCGAT": 694, + "ACCGAC": 695, + "ACCGAG": 696, + "ACCGTA": 697, + "ACCGTT": 698, + "ACCGTC": 699, + "ACCGTG": 700, + "ACCGCA": 701, + "ACCGCT": 702, + "ACCGCC": 703, + "ACCGCG": 704, + "ACCGGA": 705, + "ACCGGT": 706, + "ACCGGC": 707, + "ACCGGG": 708, + "ACGAAA": 709, + "ACGAAT": 710, + "ACGAAC": 711, + "ACGAAG": 712, + "ACGATA": 713, + "ACGATT": 714, + "ACGATC": 715, + "ACGATG": 716, + "ACGACA": 717, + "ACGACT": 718, + "ACGACC": 719, + "ACGACG": 720, + "ACGAGA": 721, + "ACGAGT": 722, + "ACGAGC": 723, + "ACGAGG": 724, + "ACGTAA": 725, + "ACGTAT": 726, + "ACGTAC": 727, + "ACGTAG": 728, + "ACGTTA": 729, + "ACGTTT": 730, + "ACGTTC": 731, + "ACGTTG": 732, + "ACGTCA": 733, + "ACGTCT": 734, + "ACGTCC": 735, + "ACGTCG": 736, + "ACGTGA": 737, + "ACGTGT": 738, + "ACGTGC": 739, + "ACGTGG": 740, + "ACGCAA": 741, + "ACGCAT": 742, + "ACGCAC": 743, + "ACGCAG": 744, + "ACGCTA": 745, + "ACGCTT": 746, + "ACGCTC": 747, + "ACGCTG": 748, + "ACGCCA": 749, + "ACGCCT": 750, + "ACGCCC": 751, + "ACGCCG": 752, + "ACGCGA": 753, + "ACGCGT": 754, + "ACGCGC": 755, + "ACGCGG": 756, + "ACGGAA": 757, + "ACGGAT": 758, + "ACGGAC": 759, + "ACGGAG": 760, + "ACGGTA": 761, + "ACGGTT": 762, + "ACGGTC": 763, + "ACGGTG": 764, + "ACGGCA": 765, + "ACGGCT": 766, + "ACGGCC": 767, + "ACGGCG": 768, + "ACGGGA": 769, + "ACGGGT": 770, + "ACGGGC": 771, + "ACGGGG": 772, + "AGAAAA": 773, + "AGAAAT": 774, + "AGAAAC": 775, + "AGAAAG": 776, + "AGAATA": 777, + "AGAATT": 778, + "AGAATC": 779, + "AGAATG": 780, + "AGAACA": 781, + "AGAACT": 782, + "AGAACC": 783, + "AGAACG": 784, + "AGAAGA": 785, + "AGAAGT": 786, + "AGAAGC": 787, + "AGAAGG": 788, + "AGATAA": 789, + "AGATAT": 790, + "AGATAC": 791, + "AGATAG": 792, + "AGATTA": 793, + "AGATTT": 794, + "AGATTC": 795, + "AGATTG": 796, + "AGATCA": 797, + "AGATCT": 798, + "AGATCC": 799, + "AGATCG": 800, + "AGATGA": 801, + "AGATGT": 802, + "AGATGC": 803, + "AGATGG": 804, + "AGACAA": 805, + "AGACAT": 806, + "AGACAC": 807, + "AGACAG": 808, + "AGACTA": 809, + "AGACTT": 810, + "AGACTC": 811, + "AGACTG": 812, + "AGACCA": 813, + "AGACCT": 814, + "AGACCC": 815, + "AGACCG": 816, + "AGACGA": 817, + "AGACGT": 818, + "AGACGC": 819, + "AGACGG": 820, + "AGAGAA": 821, + "AGAGAT": 822, + "AGAGAC": 823, + "AGAGAG": 824, + "AGAGTA": 825, + "AGAGTT": 826, + "AGAGTC": 827, + "AGAGTG": 828, + "AGAGCA": 829, + "AGAGCT": 830, + "AGAGCC": 831, + "AGAGCG": 832, + "AGAGGA": 833, + "AGAGGT": 834, + "AGAGGC": 835, + "AGAGGG": 836, + "AGTAAA": 837, + "AGTAAT": 838, + "AGTAAC": 839, + "AGTAAG": 840, + "AGTATA": 841, + "AGTATT": 842, + "AGTATC": 843, + "AGTATG": 844, + "AGTACA": 845, + "AGTACT": 846, + "AGTACC": 847, + "AGTACG": 848, + "AGTAGA": 849, + "AGTAGT": 850, + "AGTAGC": 851, + "AGTAGG": 852, + "AGTTAA": 853, + "AGTTAT": 854, + "AGTTAC": 855, + "AGTTAG": 856, + "AGTTTA": 857, + "AGTTTT": 858, + "AGTTTC": 859, + "AGTTTG": 860, + "AGTTCA": 861, + "AGTTCT": 862, + "AGTTCC": 863, + "AGTTCG": 864, + "AGTTGA": 865, + "AGTTGT": 866, + "AGTTGC": 867, + "AGTTGG": 868, + "AGTCAA": 869, + "AGTCAT": 870, + "AGTCAC": 871, + "AGTCAG": 872, + "AGTCTA": 873, + "AGTCTT": 874, + "AGTCTC": 875, + "AGTCTG": 876, + "AGTCCA": 877, + "AGTCCT": 878, + "AGTCCC": 879, + "AGTCCG": 880, + "AGTCGA": 881, + "AGTCGT": 882, + "AGTCGC": 883, + "AGTCGG": 884, + "AGTGAA": 885, + "AGTGAT": 886, + "AGTGAC": 887, + "AGTGAG": 888, + "AGTGTA": 889, + "AGTGTT": 890, + "AGTGTC": 891, + "AGTGTG": 892, + "AGTGCA": 893, + "AGTGCT": 894, + "AGTGCC": 895, + "AGTGCG": 896, + "AGTGGA": 897, + "AGTGGT": 898, + "AGTGGC": 899, + "AGTGGG": 900, + "AGCAAA": 901, + "AGCAAT": 902, + "AGCAAC": 903, + "AGCAAG": 904, + "AGCATA": 905, + "AGCATT": 906, + "AGCATC": 907, + "AGCATG": 908, + "AGCACA": 909, + "AGCACT": 910, + "AGCACC": 911, + "AGCACG": 912, + "AGCAGA": 913, + "AGCAGT": 914, + "AGCAGC": 915, + "AGCAGG": 916, + "AGCTAA": 917, + "AGCTAT": 918, + "AGCTAC": 919, + "AGCTAG": 920, + "AGCTTA": 921, + "AGCTTT": 922, + "AGCTTC": 923, + "AGCTTG": 924, + "AGCTCA": 925, + "AGCTCT": 926, + "AGCTCC": 927, + "AGCTCG": 928, + "AGCTGA": 929, + "AGCTGT": 930, + "AGCTGC": 931, + "AGCTGG": 932, + "AGCCAA": 933, + "AGCCAT": 934, + "AGCCAC": 935, + "AGCCAG": 936, + "AGCCTA": 937, + "AGCCTT": 938, + "AGCCTC": 939, + "AGCCTG": 940, + "AGCCCA": 941, + "AGCCCT": 942, + "AGCCCC": 943, + "AGCCCG": 944, + "AGCCGA": 945, + "AGCCGT": 946, + "AGCCGC": 947, + "AGCCGG": 948, + "AGCGAA": 949, + "AGCGAT": 950, + "AGCGAC": 951, + "AGCGAG": 952, + "AGCGTA": 953, + "AGCGTT": 954, + "AGCGTC": 955, + "AGCGTG": 956, + "AGCGCA": 957, + "AGCGCT": 958, + "AGCGCC": 959, + "AGCGCG": 960, + "AGCGGA": 961, + "AGCGGT": 962, + "AGCGGC": 963, + "AGCGGG": 964, + "AGGAAA": 965, + "AGGAAT": 966, + "AGGAAC": 967, + "AGGAAG": 968, + "AGGATA": 969, + "AGGATT": 970, + "AGGATC": 971, + "AGGATG": 972, + "AGGACA": 973, + "AGGACT": 974, + "AGGACC": 975, + "AGGACG": 976, + "AGGAGA": 977, + "AGGAGT": 978, + "AGGAGC": 979, + "AGGAGG": 980, + "AGGTAA": 981, + "AGGTAT": 982, + "AGGTAC": 983, + "AGGTAG": 984, + "AGGTTA": 985, + "AGGTTT": 986, + "AGGTTC": 987, + "AGGTTG": 988, + "AGGTCA": 989, + "AGGTCT": 990, + "AGGTCC": 991, + "AGGTCG": 992, + "AGGTGA": 993, + "AGGTGT": 994, + "AGGTGC": 995, + "AGGTGG": 996, + "AGGCAA": 997, + "AGGCAT": 998, + "AGGCAC": 999, + "AGGCAG": 1000, + "AGGCTA": 1001, + "AGGCTT": 1002, + "AGGCTC": 1003, + "AGGCTG": 1004, + "AGGCCA": 1005, + "AGGCCT": 1006, + "AGGCCC": 1007, + "AGGCCG": 1008, + "AGGCGA": 1009, + "AGGCGT": 1010, + "AGGCGC": 1011, + "AGGCGG": 1012, + "AGGGAA": 1013, + "AGGGAT": 1014, + "AGGGAC": 1015, + "AGGGAG": 1016, + "AGGGTA": 1017, + "AGGGTT": 1018, + "AGGGTC": 1019, + "AGGGTG": 1020, + "AGGGCA": 1021, + "AGGGCT": 1022, + "AGGGCC": 1023, + "AGGGCG": 1024, + "AGGGGA": 1025, + "AGGGGT": 1026, + "AGGGGC": 1027, + "AGGGGG": 1028, + "TAAAAA": 1029, + "TAAAAT": 1030, + "TAAAAC": 1031, + "TAAAAG": 1032, + "TAAATA": 1033, + "TAAATT": 1034, + "TAAATC": 1035, + "TAAATG": 1036, + "TAAACA": 1037, + "TAAACT": 1038, + "TAAACC": 1039, + "TAAACG": 1040, + "TAAAGA": 1041, + "TAAAGT": 1042, + "TAAAGC": 1043, + "TAAAGG": 1044, + "TAATAA": 1045, + "TAATAT": 1046, + "TAATAC": 1047, + "TAATAG": 1048, + "TAATTA": 1049, + "TAATTT": 1050, + "TAATTC": 1051, + "TAATTG": 1052, + "TAATCA": 1053, + "TAATCT": 1054, + "TAATCC": 1055, + "TAATCG": 1056, + "TAATGA": 1057, + "TAATGT": 1058, + "TAATGC": 1059, + "TAATGG": 1060, + "TAACAA": 1061, + "TAACAT": 1062, + "TAACAC": 1063, + "TAACAG": 1064, + "TAACTA": 1065, + "TAACTT": 1066, + "TAACTC": 1067, + "TAACTG": 1068, + "TAACCA": 1069, + "TAACCT": 1070, + "TAACCC": 1071, + "TAACCG": 1072, + "TAACGA": 1073, + "TAACGT": 1074, + "TAACGC": 1075, + "TAACGG": 1076, + "TAAGAA": 1077, + "TAAGAT": 1078, + "TAAGAC": 1079, + "TAAGAG": 1080, + "TAAGTA": 1081, + "TAAGTT": 1082, + "TAAGTC": 1083, + "TAAGTG": 1084, + "TAAGCA": 1085, + "TAAGCT": 1086, + "TAAGCC": 1087, + "TAAGCG": 1088, + "TAAGGA": 1089, + "TAAGGT": 1090, + "TAAGGC": 1091, + "TAAGGG": 1092, + "TATAAA": 1093, + "TATAAT": 1094, + "TATAAC": 1095, + "TATAAG": 1096, + "TATATA": 1097, + "TATATT": 1098, + "TATATC": 1099, + "TATATG": 1100, + "TATACA": 1101, + "TATACT": 1102, + "TATACC": 1103, + "TATACG": 1104, + "TATAGA": 1105, + "TATAGT": 1106, + "TATAGC": 1107, + "TATAGG": 1108, + "TATTAA": 1109, + "TATTAT": 1110, + "TATTAC": 1111, + "TATTAG": 1112, + "TATTTA": 1113, + "TATTTT": 1114, + "TATTTC": 1115, + "TATTTG": 1116, + "TATTCA": 1117, + "TATTCT": 1118, + "TATTCC": 1119, + "TATTCG": 1120, + "TATTGA": 1121, + "TATTGT": 1122, + "TATTGC": 1123, + "TATTGG": 1124, + "TATCAA": 1125, + "TATCAT": 1126, + "TATCAC": 1127, + "TATCAG": 1128, + "TATCTA": 1129, + "TATCTT": 1130, + "TATCTC": 1131, + "TATCTG": 1132, + "TATCCA": 1133, + "TATCCT": 1134, + "TATCCC": 1135, + "TATCCG": 1136, + "TATCGA": 1137, + "TATCGT": 1138, + "TATCGC": 1139, + "TATCGG": 1140, + "TATGAA": 1141, + "TATGAT": 1142, + "TATGAC": 1143, + "TATGAG": 1144, + "TATGTA": 1145, + "TATGTT": 1146, + "TATGTC": 1147, + "TATGTG": 1148, + "TATGCA": 1149, + "TATGCT": 1150, + "TATGCC": 1151, + "TATGCG": 1152, + "TATGGA": 1153, + "TATGGT": 1154, + "TATGGC": 1155, + "TATGGG": 1156, + "TACAAA": 1157, + "TACAAT": 1158, + "TACAAC": 1159, + "TACAAG": 1160, + "TACATA": 1161, + "TACATT": 1162, + "TACATC": 1163, + "TACATG": 1164, + "TACACA": 1165, + "TACACT": 1166, + "TACACC": 1167, + "TACACG": 1168, + "TACAGA": 1169, + "TACAGT": 1170, + "TACAGC": 1171, + "TACAGG": 1172, + "TACTAA": 1173, + "TACTAT": 1174, + "TACTAC": 1175, + "TACTAG": 1176, + "TACTTA": 1177, + "TACTTT": 1178, + "TACTTC": 1179, + "TACTTG": 1180, + "TACTCA": 1181, + "TACTCT": 1182, + "TACTCC": 1183, + "TACTCG": 1184, + "TACTGA": 1185, + "TACTGT": 1186, + "TACTGC": 1187, + "TACTGG": 1188, + "TACCAA": 1189, + "TACCAT": 1190, + "TACCAC": 1191, + "TACCAG": 1192, + "TACCTA": 1193, + "TACCTT": 1194, + "TACCTC": 1195, + "TACCTG": 1196, + "TACCCA": 1197, + "TACCCT": 1198, + "TACCCC": 1199, + "TACCCG": 1200, + "TACCGA": 1201, + "TACCGT": 1202, + "TACCGC": 1203, + "TACCGG": 1204, + "TACGAA": 1205, + "TACGAT": 1206, + "TACGAC": 1207, + "TACGAG": 1208, + "TACGTA": 1209, + "TACGTT": 1210, + "TACGTC": 1211, + "TACGTG": 1212, + "TACGCA": 1213, + "TACGCT": 1214, + "TACGCC": 1215, + "TACGCG": 1216, + "TACGGA": 1217, + "TACGGT": 1218, + "TACGGC": 1219, + "TACGGG": 1220, + "TAGAAA": 1221, + "TAGAAT": 1222, + "TAGAAC": 1223, + "TAGAAG": 1224, + "TAGATA": 1225, + "TAGATT": 1226, + "TAGATC": 1227, + "TAGATG": 1228, + "TAGACA": 1229, + "TAGACT": 1230, + "TAGACC": 1231, + "TAGACG": 1232, + "TAGAGA": 1233, + "TAGAGT": 1234, + "TAGAGC": 1235, + "TAGAGG": 1236, + "TAGTAA": 1237, + "TAGTAT": 1238, + "TAGTAC": 1239, + "TAGTAG": 1240, + "TAGTTA": 1241, + "TAGTTT": 1242, + "TAGTTC": 1243, + "TAGTTG": 1244, + "TAGTCA": 1245, + "TAGTCT": 1246, + "TAGTCC": 1247, + "TAGTCG": 1248, + "TAGTGA": 1249, + "TAGTGT": 1250, + "TAGTGC": 1251, + "TAGTGG": 1252, + "TAGCAA": 1253, + "TAGCAT": 1254, + "TAGCAC": 1255, + "TAGCAG": 1256, + "TAGCTA": 1257, + "TAGCTT": 1258, + "TAGCTC": 1259, + "TAGCTG": 1260, + "TAGCCA": 1261, + "TAGCCT": 1262, + "TAGCCC": 1263, + "TAGCCG": 1264, + "TAGCGA": 1265, + "TAGCGT": 1266, + "TAGCGC": 1267, + "TAGCGG": 1268, + "TAGGAA": 1269, + "TAGGAT": 1270, + "TAGGAC": 1271, + "TAGGAG": 1272, + "TAGGTA": 1273, + "TAGGTT": 1274, + "TAGGTC": 1275, + "TAGGTG": 1276, + "TAGGCA": 1277, + "TAGGCT": 1278, + "TAGGCC": 1279, + "TAGGCG": 1280, + "TAGGGA": 1281, + "TAGGGT": 1282, + "TAGGGC": 1283, + "TAGGGG": 1284, + "TTAAAA": 1285, + "TTAAAT": 1286, + "TTAAAC": 1287, + "TTAAAG": 1288, + "TTAATA": 1289, + "TTAATT": 1290, + "TTAATC": 1291, + "TTAATG": 1292, + "TTAACA": 1293, + "TTAACT": 1294, + "TTAACC": 1295, + "TTAACG": 1296, + "TTAAGA": 1297, + "TTAAGT": 1298, + "TTAAGC": 1299, + "TTAAGG": 1300, + "TTATAA": 1301, + "TTATAT": 1302, + "TTATAC": 1303, + "TTATAG": 1304, + "TTATTA": 1305, + "TTATTT": 1306, + "TTATTC": 1307, + "TTATTG": 1308, + "TTATCA": 1309, + "TTATCT": 1310, + "TTATCC": 1311, + "TTATCG": 1312, + "TTATGA": 1313, + "TTATGT": 1314, + "TTATGC": 1315, + "TTATGG": 1316, + "TTACAA": 1317, + "TTACAT": 1318, + "TTACAC": 1319, + "TTACAG": 1320, + "TTACTA": 1321, + "TTACTT": 1322, + "TTACTC": 1323, + "TTACTG": 1324, + "TTACCA": 1325, + "TTACCT": 1326, + "TTACCC": 1327, + "TTACCG": 1328, + "TTACGA": 1329, + "TTACGT": 1330, + "TTACGC": 1331, + "TTACGG": 1332, + "TTAGAA": 1333, + "TTAGAT": 1334, + "TTAGAC": 1335, + "TTAGAG": 1336, + "TTAGTA": 1337, + "TTAGTT": 1338, + "TTAGTC": 1339, + "TTAGTG": 1340, + "TTAGCA": 1341, + "TTAGCT": 1342, + "TTAGCC": 1343, + "TTAGCG": 1344, + "TTAGGA": 1345, + "TTAGGT": 1346, + "TTAGGC": 1347, + "TTAGGG": 1348, + "TTTAAA": 1349, + "TTTAAT": 1350, + "TTTAAC": 1351, + "TTTAAG": 1352, + "TTTATA": 1353, + "TTTATT": 1354, + "TTTATC": 1355, + "TTTATG": 1356, + "TTTACA": 1357, + "TTTACT": 1358, + "TTTACC": 1359, + "TTTACG": 1360, + "TTTAGA": 1361, + "TTTAGT": 1362, + "TTTAGC": 1363, + "TTTAGG": 1364, + "TTTTAA": 1365, + "TTTTAT": 1366, + "TTTTAC": 1367, + "TTTTAG": 1368, + "TTTTTA": 1369, + "TTTTTT": 1370, + "TTTTTC": 1371, + "TTTTTG": 1372, + "TTTTCA": 1373, + "TTTTCT": 1374, + "TTTTCC": 1375, + "TTTTCG": 1376, + "TTTTGA": 1377, + "TTTTGT": 1378, + "TTTTGC": 1379, + "TTTTGG": 1380, + "TTTCAA": 1381, + "TTTCAT": 1382, + "TTTCAC": 1383, + "TTTCAG": 1384, + "TTTCTA": 1385, + "TTTCTT": 1386, + "TTTCTC": 1387, + "TTTCTG": 1388, + "TTTCCA": 1389, + "TTTCCT": 1390, + "TTTCCC": 1391, + "TTTCCG": 1392, + "TTTCGA": 1393, + "TTTCGT": 1394, + "TTTCGC": 1395, + "TTTCGG": 1396, + "TTTGAA": 1397, + "TTTGAT": 1398, + "TTTGAC": 1399, + "TTTGAG": 1400, + "TTTGTA": 1401, + "TTTGTT": 1402, + "TTTGTC": 1403, + "TTTGTG": 1404, + "TTTGCA": 1405, + "TTTGCT": 1406, + "TTTGCC": 1407, + "TTTGCG": 1408, + "TTTGGA": 1409, + "TTTGGT": 1410, + "TTTGGC": 1411, + "TTTGGG": 1412, + "TTCAAA": 1413, + "TTCAAT": 1414, + "TTCAAC": 1415, + "TTCAAG": 1416, + "TTCATA": 1417, + "TTCATT": 1418, + "TTCATC": 1419, + "TTCATG": 1420, + "TTCACA": 1421, + "TTCACT": 1422, + "TTCACC": 1423, + "TTCACG": 1424, + "TTCAGA": 1425, + "TTCAGT": 1426, + "TTCAGC": 1427, + "TTCAGG": 1428, + "TTCTAA": 1429, + "TTCTAT": 1430, + "TTCTAC": 1431, + "TTCTAG": 1432, + "TTCTTA": 1433, + "TTCTTT": 1434, + "TTCTTC": 1435, + "TTCTTG": 1436, + "TTCTCA": 1437, + "TTCTCT": 1438, + "TTCTCC": 1439, + "TTCTCG": 1440, + "TTCTGA": 1441, + "TTCTGT": 1442, + "TTCTGC": 1443, + "TTCTGG": 1444, + "TTCCAA": 1445, + "TTCCAT": 1446, + "TTCCAC": 1447, + "TTCCAG": 1448, + "TTCCTA": 1449, + "TTCCTT": 1450, + "TTCCTC": 1451, + "TTCCTG": 1452, + "TTCCCA": 1453, + "TTCCCT": 1454, + "TTCCCC": 1455, + "TTCCCG": 1456, + "TTCCGA": 1457, + "TTCCGT": 1458, + "TTCCGC": 1459, + "TTCCGG": 1460, + "TTCGAA": 1461, + "TTCGAT": 1462, + "TTCGAC": 1463, + "TTCGAG": 1464, + "TTCGTA": 1465, + "TTCGTT": 1466, + "TTCGTC": 1467, + "TTCGTG": 1468, + "TTCGCA": 1469, + "TTCGCT": 1470, + "TTCGCC": 1471, + "TTCGCG": 1472, + "TTCGGA": 1473, + "TTCGGT": 1474, + "TTCGGC": 1475, + "TTCGGG": 1476, + "TTGAAA": 1477, + "TTGAAT": 1478, + "TTGAAC": 1479, + "TTGAAG": 1480, + "TTGATA": 1481, + "TTGATT": 1482, + "TTGATC": 1483, + "TTGATG": 1484, + "TTGACA": 1485, + "TTGACT": 1486, + "TTGACC": 1487, + "TTGACG": 1488, + "TTGAGA": 1489, + "TTGAGT": 1490, + "TTGAGC": 1491, + "TTGAGG": 1492, + "TTGTAA": 1493, + "TTGTAT": 1494, + "TTGTAC": 1495, + "TTGTAG": 1496, + "TTGTTA": 1497, + "TTGTTT": 1498, + "TTGTTC": 1499, + "TTGTTG": 1500, + "TTGTCA": 1501, + "TTGTCT": 1502, + "TTGTCC": 1503, + "TTGTCG": 1504, + "TTGTGA": 1505, + "TTGTGT": 1506, + "TTGTGC": 1507, + "TTGTGG": 1508, + "TTGCAA": 1509, + "TTGCAT": 1510, + "TTGCAC": 1511, + "TTGCAG": 1512, + "TTGCTA": 1513, + "TTGCTT": 1514, + "TTGCTC": 1515, + "TTGCTG": 1516, + "TTGCCA": 1517, + "TTGCCT": 1518, + "TTGCCC": 1519, + "TTGCCG": 1520, + "TTGCGA": 1521, + "TTGCGT": 1522, + "TTGCGC": 1523, + "TTGCGG": 1524, + "TTGGAA": 1525, + "TTGGAT": 1526, + "TTGGAC": 1527, + "TTGGAG": 1528, + "TTGGTA": 1529, + "TTGGTT": 1530, + "TTGGTC": 1531, + "TTGGTG": 1532, + "TTGGCA": 1533, + "TTGGCT": 1534, + "TTGGCC": 1535, + "TTGGCG": 1536, + "TTGGGA": 1537, + "TTGGGT": 1538, + "TTGGGC": 1539, + "TTGGGG": 1540, + "TCAAAA": 1541, + "TCAAAT": 1542, + "TCAAAC": 1543, + "TCAAAG": 1544, + "TCAATA": 1545, + "TCAATT": 1546, + "TCAATC": 1547, + "TCAATG": 1548, + "TCAACA": 1549, + "TCAACT": 1550, + "TCAACC": 1551, + "TCAACG": 1552, + "TCAAGA": 1553, + "TCAAGT": 1554, + "TCAAGC": 1555, + "TCAAGG": 1556, + "TCATAA": 1557, + "TCATAT": 1558, + "TCATAC": 1559, + "TCATAG": 1560, + "TCATTA": 1561, + "TCATTT": 1562, + "TCATTC": 1563, + "TCATTG": 1564, + "TCATCA": 1565, + "TCATCT": 1566, + "TCATCC": 1567, + "TCATCG": 1568, + "TCATGA": 1569, + "TCATGT": 1570, + "TCATGC": 1571, + "TCATGG": 1572, + "TCACAA": 1573, + "TCACAT": 1574, + "TCACAC": 1575, + "TCACAG": 1576, + "TCACTA": 1577, + "TCACTT": 1578, + "TCACTC": 1579, + "TCACTG": 1580, + "TCACCA": 1581, + "TCACCT": 1582, + "TCACCC": 1583, + "TCACCG": 1584, + "TCACGA": 1585, + "TCACGT": 1586, + "TCACGC": 1587, + "TCACGG": 1588, + "TCAGAA": 1589, + "TCAGAT": 1590, + "TCAGAC": 1591, + "TCAGAG": 1592, + "TCAGTA": 1593, + "TCAGTT": 1594, + "TCAGTC": 1595, + "TCAGTG": 1596, + "TCAGCA": 1597, + "TCAGCT": 1598, + "TCAGCC": 1599, + "TCAGCG": 1600, + "TCAGGA": 1601, + "TCAGGT": 1602, + "TCAGGC": 1603, + "TCAGGG": 1604, + "TCTAAA": 1605, + "TCTAAT": 1606, + "TCTAAC": 1607, + "TCTAAG": 1608, + "TCTATA": 1609, + "TCTATT": 1610, + "TCTATC": 1611, + "TCTATG": 1612, + "TCTACA": 1613, + "TCTACT": 1614, + "TCTACC": 1615, + "TCTACG": 1616, + "TCTAGA": 1617, + "TCTAGT": 1618, + "TCTAGC": 1619, + "TCTAGG": 1620, + "TCTTAA": 1621, + "TCTTAT": 1622, + "TCTTAC": 1623, + "TCTTAG": 1624, + "TCTTTA": 1625, + "TCTTTT": 1626, + "TCTTTC": 1627, + "TCTTTG": 1628, + "TCTTCA": 1629, + "TCTTCT": 1630, + "TCTTCC": 1631, + "TCTTCG": 1632, + "TCTTGA": 1633, + "TCTTGT": 1634, + "TCTTGC": 1635, + "TCTTGG": 1636, + "TCTCAA": 1637, + "TCTCAT": 1638, + "TCTCAC": 1639, + "TCTCAG": 1640, + "TCTCTA": 1641, + "TCTCTT": 1642, + "TCTCTC": 1643, + "TCTCTG": 1644, + "TCTCCA": 1645, + "TCTCCT": 1646, + "TCTCCC": 1647, + "TCTCCG": 1648, + "TCTCGA": 1649, + "TCTCGT": 1650, + "TCTCGC": 1651, + "TCTCGG": 1652, + "TCTGAA": 1653, + "TCTGAT": 1654, + "TCTGAC": 1655, + "TCTGAG": 1656, + "TCTGTA": 1657, + "TCTGTT": 1658, + "TCTGTC": 1659, + "TCTGTG": 1660, + "TCTGCA": 1661, + "TCTGCT": 1662, + "TCTGCC": 1663, + "TCTGCG": 1664, + "TCTGGA": 1665, + "TCTGGT": 1666, + "TCTGGC": 1667, + "TCTGGG": 1668, + "TCCAAA": 1669, + "TCCAAT": 1670, + "TCCAAC": 1671, + "TCCAAG": 1672, + "TCCATA": 1673, + "TCCATT": 1674, + "TCCATC": 1675, + "TCCATG": 1676, + "TCCACA": 1677, + "TCCACT": 1678, + "TCCACC": 1679, + "TCCACG": 1680, + "TCCAGA": 1681, + "TCCAGT": 1682, + "TCCAGC": 1683, + "TCCAGG": 1684, + "TCCTAA": 1685, + "TCCTAT": 1686, + "TCCTAC": 1687, + "TCCTAG": 1688, + "TCCTTA": 1689, + "TCCTTT": 1690, + "TCCTTC": 1691, + "TCCTTG": 1692, + "TCCTCA": 1693, + "TCCTCT": 1694, + "TCCTCC": 1695, + "TCCTCG": 1696, + "TCCTGA": 1697, + "TCCTGT": 1698, + "TCCTGC": 1699, + "TCCTGG": 1700, + "TCCCAA": 1701, + "TCCCAT": 1702, + "TCCCAC": 1703, + "TCCCAG": 1704, + "TCCCTA": 1705, + "TCCCTT": 1706, + "TCCCTC": 1707, + "TCCCTG": 1708, + "TCCCCA": 1709, + "TCCCCT": 1710, + "TCCCCC": 1711, + "TCCCCG": 1712, + "TCCCGA": 1713, + "TCCCGT": 1714, + "TCCCGC": 1715, + "TCCCGG": 1716, + "TCCGAA": 1717, + "TCCGAT": 1718, + "TCCGAC": 1719, + "TCCGAG": 1720, + "TCCGTA": 1721, + "TCCGTT": 1722, + "TCCGTC": 1723, + "TCCGTG": 1724, + "TCCGCA": 1725, + "TCCGCT": 1726, + "TCCGCC": 1727, + "TCCGCG": 1728, + "TCCGGA": 1729, + "TCCGGT": 1730, + "TCCGGC": 1731, + "TCCGGG": 1732, + "TCGAAA": 1733, + "TCGAAT": 1734, + "TCGAAC": 1735, + "TCGAAG": 1736, + "TCGATA": 1737, + "TCGATT": 1738, + "TCGATC": 1739, + "TCGATG": 1740, + "TCGACA": 1741, + "TCGACT": 1742, + "TCGACC": 1743, + "TCGACG": 1744, + "TCGAGA": 1745, + "TCGAGT": 1746, + "TCGAGC": 1747, + "TCGAGG": 1748, + "TCGTAA": 1749, + "TCGTAT": 1750, + "TCGTAC": 1751, + "TCGTAG": 1752, + "TCGTTA": 1753, + "TCGTTT": 1754, + "TCGTTC": 1755, + "TCGTTG": 1756, + "TCGTCA": 1757, + "TCGTCT": 1758, + "TCGTCC": 1759, + "TCGTCG": 1760, + "TCGTGA": 1761, + "TCGTGT": 1762, + "TCGTGC": 1763, + "TCGTGG": 1764, + "TCGCAA": 1765, + "TCGCAT": 1766, + "TCGCAC": 1767, + "TCGCAG": 1768, + "TCGCTA": 1769, + "TCGCTT": 1770, + "TCGCTC": 1771, + "TCGCTG": 1772, + "TCGCCA": 1773, + "TCGCCT": 1774, + "TCGCCC": 1775, + "TCGCCG": 1776, + "TCGCGA": 1777, + "TCGCGT": 1778, + "TCGCGC": 1779, + "TCGCGG": 1780, + "TCGGAA": 1781, + "TCGGAT": 1782, + "TCGGAC": 1783, + "TCGGAG": 1784, + "TCGGTA": 1785, + "TCGGTT": 1786, + "TCGGTC": 1787, + "TCGGTG": 1788, + "TCGGCA": 1789, + "TCGGCT": 1790, + "TCGGCC": 1791, + "TCGGCG": 1792, + "TCGGGA": 1793, + "TCGGGT": 1794, + "TCGGGC": 1795, + "TCGGGG": 1796, + "TGAAAA": 1797, + "TGAAAT": 1798, + "TGAAAC": 1799, + "TGAAAG": 1800, + "TGAATA": 1801, + "TGAATT": 1802, + "TGAATC": 1803, + "TGAATG": 1804, + "TGAACA": 1805, + "TGAACT": 1806, + "TGAACC": 1807, + "TGAACG": 1808, + "TGAAGA": 1809, + "TGAAGT": 1810, + "TGAAGC": 1811, + "TGAAGG": 1812, + "TGATAA": 1813, + "TGATAT": 1814, + "TGATAC": 1815, + "TGATAG": 1816, + "TGATTA": 1817, + "TGATTT": 1818, + "TGATTC": 1819, + "TGATTG": 1820, + "TGATCA": 1821, + "TGATCT": 1822, + "TGATCC": 1823, + "TGATCG": 1824, + "TGATGA": 1825, + "TGATGT": 1826, + "TGATGC": 1827, + "TGATGG": 1828, + "TGACAA": 1829, + "TGACAT": 1830, + "TGACAC": 1831, + "TGACAG": 1832, + "TGACTA": 1833, + "TGACTT": 1834, + "TGACTC": 1835, + "TGACTG": 1836, + "TGACCA": 1837, + "TGACCT": 1838, + "TGACCC": 1839, + "TGACCG": 1840, + "TGACGA": 1841, + "TGACGT": 1842, + "TGACGC": 1843, + "TGACGG": 1844, + "TGAGAA": 1845, + "TGAGAT": 1846, + "TGAGAC": 1847, + "TGAGAG": 1848, + "TGAGTA": 1849, + "TGAGTT": 1850, + "TGAGTC": 1851, + "TGAGTG": 1852, + "TGAGCA": 1853, + "TGAGCT": 1854, + "TGAGCC": 1855, + "TGAGCG": 1856, + "TGAGGA": 1857, + "TGAGGT": 1858, + "TGAGGC": 1859, + "TGAGGG": 1860, + "TGTAAA": 1861, + "TGTAAT": 1862, + "TGTAAC": 1863, + "TGTAAG": 1864, + "TGTATA": 1865, + "TGTATT": 1866, + "TGTATC": 1867, + "TGTATG": 1868, + "TGTACA": 1869, + "TGTACT": 1870, + "TGTACC": 1871, + "TGTACG": 1872, + "TGTAGA": 1873, + "TGTAGT": 1874, + "TGTAGC": 1875, + "TGTAGG": 1876, + "TGTTAA": 1877, + "TGTTAT": 1878, + "TGTTAC": 1879, + "TGTTAG": 1880, + "TGTTTA": 1881, + "TGTTTT": 1882, + "TGTTTC": 1883, + "TGTTTG": 1884, + "TGTTCA": 1885, + "TGTTCT": 1886, + "TGTTCC": 1887, + "TGTTCG": 1888, + "TGTTGA": 1889, + "TGTTGT": 1890, + "TGTTGC": 1891, + "TGTTGG": 1892, + "TGTCAA": 1893, + "TGTCAT": 1894, + "TGTCAC": 1895, + "TGTCAG": 1896, + "TGTCTA": 1897, + "TGTCTT": 1898, + "TGTCTC": 1899, + "TGTCTG": 1900, + "TGTCCA": 1901, + "TGTCCT": 1902, + "TGTCCC": 1903, + "TGTCCG": 1904, + "TGTCGA": 1905, + "TGTCGT": 1906, + "TGTCGC": 1907, + "TGTCGG": 1908, + "TGTGAA": 1909, + "TGTGAT": 1910, + "TGTGAC": 1911, + "TGTGAG": 1912, + "TGTGTA": 1913, + "TGTGTT": 1914, + "TGTGTC": 1915, + "TGTGTG": 1916, + "TGTGCA": 1917, + "TGTGCT": 1918, + "TGTGCC": 1919, + "TGTGCG": 1920, + "TGTGGA": 1921, + "TGTGGT": 1922, + "TGTGGC": 1923, + "TGTGGG": 1924, + "TGCAAA": 1925, + "TGCAAT": 1926, + "TGCAAC": 1927, + "TGCAAG": 1928, + "TGCATA": 1929, + "TGCATT": 1930, + "TGCATC": 1931, + "TGCATG": 1932, + "TGCACA": 1933, + "TGCACT": 1934, + "TGCACC": 1935, + "TGCACG": 1936, + "TGCAGA": 1937, + "TGCAGT": 1938, + "TGCAGC": 1939, + "TGCAGG": 1940, + "TGCTAA": 1941, + "TGCTAT": 1942, + "TGCTAC": 1943, + "TGCTAG": 1944, + "TGCTTA": 1945, + "TGCTTT": 1946, + "TGCTTC": 1947, + "TGCTTG": 1948, + "TGCTCA": 1949, + "TGCTCT": 1950, + "TGCTCC": 1951, + "TGCTCG": 1952, + "TGCTGA": 1953, + "TGCTGT": 1954, + "TGCTGC": 1955, + "TGCTGG": 1956, + "TGCCAA": 1957, + "TGCCAT": 1958, + "TGCCAC": 1959, + "TGCCAG": 1960, + "TGCCTA": 1961, + "TGCCTT": 1962, + "TGCCTC": 1963, + "TGCCTG": 1964, + "TGCCCA": 1965, + "TGCCCT": 1966, + "TGCCCC": 1967, + "TGCCCG": 1968, + "TGCCGA": 1969, + "TGCCGT": 1970, + "TGCCGC": 1971, + "TGCCGG": 1972, + "TGCGAA": 1973, + "TGCGAT": 1974, + "TGCGAC": 1975, + "TGCGAG": 1976, + "TGCGTA": 1977, + "TGCGTT": 1978, + "TGCGTC": 1979, + "TGCGTG": 1980, + "TGCGCA": 1981, + "TGCGCT": 1982, + "TGCGCC": 1983, + "TGCGCG": 1984, + "TGCGGA": 1985, + "TGCGGT": 1986, + "TGCGGC": 1987, + "TGCGGG": 1988, + "TGGAAA": 1989, + "TGGAAT": 1990, + "TGGAAC": 1991, + "TGGAAG": 1992, + "TGGATA": 1993, + "TGGATT": 1994, + "TGGATC": 1995, + "TGGATG": 1996, + "TGGACA": 1997, + "TGGACT": 1998, + "TGGACC": 1999, + "TGGACG": 2000, + "TGGAGA": 2001, + "TGGAGT": 2002, + "TGGAGC": 2003, + "TGGAGG": 2004, + "TGGTAA": 2005, + "TGGTAT": 2006, + "TGGTAC": 2007, + "TGGTAG": 2008, + "TGGTTA": 2009, + "TGGTTT": 2010, + "TGGTTC": 2011, + "TGGTTG": 2012, + "TGGTCA": 2013, + "TGGTCT": 2014, + "TGGTCC": 2015, + "TGGTCG": 2016, + "TGGTGA": 2017, + "TGGTGT": 2018, + "TGGTGC": 2019, + "TGGTGG": 2020, + "TGGCAA": 2021, + "TGGCAT": 2022, + "TGGCAC": 2023, + "TGGCAG": 2024, + "TGGCTA": 2025, + "TGGCTT": 2026, + "TGGCTC": 2027, + "TGGCTG": 2028, + "TGGCCA": 2029, + "TGGCCT": 2030, + "TGGCCC": 2031, + "TGGCCG": 2032, + "TGGCGA": 2033, + "TGGCGT": 2034, + "TGGCGC": 2035, + "TGGCGG": 2036, + "TGGGAA": 2037, + "TGGGAT": 2038, + "TGGGAC": 2039, + "TGGGAG": 2040, + "TGGGTA": 2041, + "TGGGTT": 2042, + "TGGGTC": 2043, + "TGGGTG": 2044, + "TGGGCA": 2045, + "TGGGCT": 2046, + "TGGGCC": 2047, + "TGGGCG": 2048, + "TGGGGA": 2049, + "TGGGGT": 2050, + "TGGGGC": 2051, + "TGGGGG": 2052, + "CAAAAA": 2053, + "CAAAAT": 2054, + "CAAAAC": 2055, + "CAAAAG": 2056, + "CAAATA": 2057, + "CAAATT": 2058, + "CAAATC": 2059, + "CAAATG": 2060, + "CAAACA": 2061, + "CAAACT": 2062, + "CAAACC": 2063, + "CAAACG": 2064, + "CAAAGA": 2065, + "CAAAGT": 2066, + "CAAAGC": 2067, + "CAAAGG": 2068, + "CAATAA": 2069, + "CAATAT": 2070, + "CAATAC": 2071, + "CAATAG": 2072, + "CAATTA": 2073, + "CAATTT": 2074, + "CAATTC": 2075, + "CAATTG": 2076, + "CAATCA": 2077, + "CAATCT": 2078, + "CAATCC": 2079, + "CAATCG": 2080, + "CAATGA": 2081, + "CAATGT": 2082, + "CAATGC": 2083, + "CAATGG": 2084, + "CAACAA": 2085, + "CAACAT": 2086, + "CAACAC": 2087, + "CAACAG": 2088, + "CAACTA": 2089, + "CAACTT": 2090, + "CAACTC": 2091, + "CAACTG": 2092, + "CAACCA": 2093, + "CAACCT": 2094, + "CAACCC": 2095, + "CAACCG": 2096, + "CAACGA": 2097, + "CAACGT": 2098, + "CAACGC": 2099, + "CAACGG": 2100, + "CAAGAA": 2101, + "CAAGAT": 2102, + "CAAGAC": 2103, + "CAAGAG": 2104, + "CAAGTA": 2105, + "CAAGTT": 2106, + "CAAGTC": 2107, + "CAAGTG": 2108, + "CAAGCA": 2109, + "CAAGCT": 2110, + "CAAGCC": 2111, + "CAAGCG": 2112, + "CAAGGA": 2113, + "CAAGGT": 2114, + "CAAGGC": 2115, + "CAAGGG": 2116, + "CATAAA": 2117, + "CATAAT": 2118, + "CATAAC": 2119, + "CATAAG": 2120, + "CATATA": 2121, + "CATATT": 2122, + "CATATC": 2123, + "CATATG": 2124, + "CATACA": 2125, + "CATACT": 2126, + "CATACC": 2127, + "CATACG": 2128, + "CATAGA": 2129, + "CATAGT": 2130, + "CATAGC": 2131, + "CATAGG": 2132, + "CATTAA": 2133, + "CATTAT": 2134, + "CATTAC": 2135, + "CATTAG": 2136, + "CATTTA": 2137, + "CATTTT": 2138, + "CATTTC": 2139, + "CATTTG": 2140, + "CATTCA": 2141, + "CATTCT": 2142, + "CATTCC": 2143, + "CATTCG": 2144, + "CATTGA": 2145, + "CATTGT": 2146, + "CATTGC": 2147, + "CATTGG": 2148, + "CATCAA": 2149, + "CATCAT": 2150, + "CATCAC": 2151, + "CATCAG": 2152, + "CATCTA": 2153, + "CATCTT": 2154, + "CATCTC": 2155, + "CATCTG": 2156, + "CATCCA": 2157, + "CATCCT": 2158, + "CATCCC": 2159, + "CATCCG": 2160, + "CATCGA": 2161, + "CATCGT": 2162, + "CATCGC": 2163, + "CATCGG": 2164, + "CATGAA": 2165, + "CATGAT": 2166, + "CATGAC": 2167, + "CATGAG": 2168, + "CATGTA": 2169, + "CATGTT": 2170, + "CATGTC": 2171, + "CATGTG": 2172, + "CATGCA": 2173, + "CATGCT": 2174, + "CATGCC": 2175, + "CATGCG": 2176, + "CATGGA": 2177, + "CATGGT": 2178, + "CATGGC": 2179, + "CATGGG": 2180, + "CACAAA": 2181, + "CACAAT": 2182, + "CACAAC": 2183, + "CACAAG": 2184, + "CACATA": 2185, + "CACATT": 2186, + "CACATC": 2187, + "CACATG": 2188, + "CACACA": 2189, + "CACACT": 2190, + "CACACC": 2191, + "CACACG": 2192, + "CACAGA": 2193, + "CACAGT": 2194, + "CACAGC": 2195, + "CACAGG": 2196, + "CACTAA": 2197, + "CACTAT": 2198, + "CACTAC": 2199, + "CACTAG": 2200, + "CACTTA": 2201, + "CACTTT": 2202, + "CACTTC": 2203, + "CACTTG": 2204, + "CACTCA": 2205, + "CACTCT": 2206, + "CACTCC": 2207, + "CACTCG": 2208, + "CACTGA": 2209, + "CACTGT": 2210, + "CACTGC": 2211, + "CACTGG": 2212, + "CACCAA": 2213, + "CACCAT": 2214, + "CACCAC": 2215, + "CACCAG": 2216, + "CACCTA": 2217, + "CACCTT": 2218, + "CACCTC": 2219, + "CACCTG": 2220, + "CACCCA": 2221, + "CACCCT": 2222, + "CACCCC": 2223, + "CACCCG": 2224, + "CACCGA": 2225, + "CACCGT": 2226, + "CACCGC": 2227, + "CACCGG": 2228, + "CACGAA": 2229, + "CACGAT": 2230, + "CACGAC": 2231, + "CACGAG": 2232, + "CACGTA": 2233, + "CACGTT": 2234, + "CACGTC": 2235, + "CACGTG": 2236, + "CACGCA": 2237, + "CACGCT": 2238, + "CACGCC": 2239, + "CACGCG": 2240, + "CACGGA": 2241, + "CACGGT": 2242, + "CACGGC": 2243, + "CACGGG": 2244, + "CAGAAA": 2245, + "CAGAAT": 2246, + "CAGAAC": 2247, + "CAGAAG": 2248, + "CAGATA": 2249, + "CAGATT": 2250, + "CAGATC": 2251, + "CAGATG": 2252, + "CAGACA": 2253, + "CAGACT": 2254, + "CAGACC": 2255, + "CAGACG": 2256, + "CAGAGA": 2257, + "CAGAGT": 2258, + "CAGAGC": 2259, + "CAGAGG": 2260, + "CAGTAA": 2261, + "CAGTAT": 2262, + "CAGTAC": 2263, + "CAGTAG": 2264, + "CAGTTA": 2265, + "CAGTTT": 2266, + "CAGTTC": 2267, + "CAGTTG": 2268, + "CAGTCA": 2269, + "CAGTCT": 2270, + "CAGTCC": 2271, + "CAGTCG": 2272, + "CAGTGA": 2273, + "CAGTGT": 2274, + "CAGTGC": 2275, + "CAGTGG": 2276, + "CAGCAA": 2277, + "CAGCAT": 2278, + "CAGCAC": 2279, + "CAGCAG": 2280, + "CAGCTA": 2281, + "CAGCTT": 2282, + "CAGCTC": 2283, + "CAGCTG": 2284, + "CAGCCA": 2285, + "CAGCCT": 2286, + "CAGCCC": 2287, + "CAGCCG": 2288, + "CAGCGA": 2289, + "CAGCGT": 2290, + "CAGCGC": 2291, + "CAGCGG": 2292, + "CAGGAA": 2293, + "CAGGAT": 2294, + "CAGGAC": 2295, + "CAGGAG": 2296, + "CAGGTA": 2297, + "CAGGTT": 2298, + "CAGGTC": 2299, + "CAGGTG": 2300, + "CAGGCA": 2301, + "CAGGCT": 2302, + "CAGGCC": 2303, + "CAGGCG": 2304, + "CAGGGA": 2305, + "CAGGGT": 2306, + "CAGGGC": 2307, + "CAGGGG": 2308, + "CTAAAA": 2309, + "CTAAAT": 2310, + "CTAAAC": 2311, + "CTAAAG": 2312, + "CTAATA": 2313, + "CTAATT": 2314, + "CTAATC": 2315, + "CTAATG": 2316, + "CTAACA": 2317, + "CTAACT": 2318, + "CTAACC": 2319, + "CTAACG": 2320, + "CTAAGA": 2321, + "CTAAGT": 2322, + "CTAAGC": 2323, + "CTAAGG": 2324, + "CTATAA": 2325, + "CTATAT": 2326, + "CTATAC": 2327, + "CTATAG": 2328, + "CTATTA": 2329, + "CTATTT": 2330, + "CTATTC": 2331, + "CTATTG": 2332, + "CTATCA": 2333, + "CTATCT": 2334, + "CTATCC": 2335, + "CTATCG": 2336, + "CTATGA": 2337, + "CTATGT": 2338, + "CTATGC": 2339, + "CTATGG": 2340, + "CTACAA": 2341, + "CTACAT": 2342, + "CTACAC": 2343, + "CTACAG": 2344, + "CTACTA": 2345, + "CTACTT": 2346, + "CTACTC": 2347, + "CTACTG": 2348, + "CTACCA": 2349, + "CTACCT": 2350, + "CTACCC": 2351, + "CTACCG": 2352, + "CTACGA": 2353, + "CTACGT": 2354, + "CTACGC": 2355, + "CTACGG": 2356, + "CTAGAA": 2357, + "CTAGAT": 2358, + "CTAGAC": 2359, + "CTAGAG": 2360, + "CTAGTA": 2361, + "CTAGTT": 2362, + "CTAGTC": 2363, + "CTAGTG": 2364, + "CTAGCA": 2365, + "CTAGCT": 2366, + "CTAGCC": 2367, + "CTAGCG": 2368, + "CTAGGA": 2369, + "CTAGGT": 2370, + "CTAGGC": 2371, + "CTAGGG": 2372, + "CTTAAA": 2373, + "CTTAAT": 2374, + "CTTAAC": 2375, + "CTTAAG": 2376, + "CTTATA": 2377, + "CTTATT": 2378, + "CTTATC": 2379, + "CTTATG": 2380, + "CTTACA": 2381, + "CTTACT": 2382, + "CTTACC": 2383, + "CTTACG": 2384, + "CTTAGA": 2385, + "CTTAGT": 2386, + "CTTAGC": 2387, + "CTTAGG": 2388, + "CTTTAA": 2389, + "CTTTAT": 2390, + "CTTTAC": 2391, + "CTTTAG": 2392, + "CTTTTA": 2393, + "CTTTTT": 2394, + "CTTTTC": 2395, + "CTTTTG": 2396, + "CTTTCA": 2397, + "CTTTCT": 2398, + "CTTTCC": 2399, + "CTTTCG": 2400, + "CTTTGA": 2401, + "CTTTGT": 2402, + "CTTTGC": 2403, + "CTTTGG": 2404, + "CTTCAA": 2405, + "CTTCAT": 2406, + "CTTCAC": 2407, + "CTTCAG": 2408, + "CTTCTA": 2409, + "CTTCTT": 2410, + "CTTCTC": 2411, + "CTTCTG": 2412, + "CTTCCA": 2413, + "CTTCCT": 2414, + "CTTCCC": 2415, + "CTTCCG": 2416, + "CTTCGA": 2417, + "CTTCGT": 2418, + "CTTCGC": 2419, + "CTTCGG": 2420, + "CTTGAA": 2421, + "CTTGAT": 2422, + "CTTGAC": 2423, + "CTTGAG": 2424, + "CTTGTA": 2425, + "CTTGTT": 2426, + "CTTGTC": 2427, + "CTTGTG": 2428, + "CTTGCA": 2429, + "CTTGCT": 2430, + "CTTGCC": 2431, + "CTTGCG": 2432, + "CTTGGA": 2433, + "CTTGGT": 2434, + "CTTGGC": 2435, + "CTTGGG": 2436, + "CTCAAA": 2437, + "CTCAAT": 2438, + "CTCAAC": 2439, + "CTCAAG": 2440, + "CTCATA": 2441, + "CTCATT": 2442, + "CTCATC": 2443, + "CTCATG": 2444, + "CTCACA": 2445, + "CTCACT": 2446, + "CTCACC": 2447, + "CTCACG": 2448, + "CTCAGA": 2449, + "CTCAGT": 2450, + "CTCAGC": 2451, + "CTCAGG": 2452, + "CTCTAA": 2453, + "CTCTAT": 2454, + "CTCTAC": 2455, + "CTCTAG": 2456, + "CTCTTA": 2457, + "CTCTTT": 2458, + "CTCTTC": 2459, + "CTCTTG": 2460, + "CTCTCA": 2461, + "CTCTCT": 2462, + "CTCTCC": 2463, + "CTCTCG": 2464, + "CTCTGA": 2465, + "CTCTGT": 2466, + "CTCTGC": 2467, + "CTCTGG": 2468, + "CTCCAA": 2469, + "CTCCAT": 2470, + "CTCCAC": 2471, + "CTCCAG": 2472, + "CTCCTA": 2473, + "CTCCTT": 2474, + "CTCCTC": 2475, + "CTCCTG": 2476, + "CTCCCA": 2477, + "CTCCCT": 2478, + "CTCCCC": 2479, + "CTCCCG": 2480, + "CTCCGA": 2481, + "CTCCGT": 2482, + "CTCCGC": 2483, + "CTCCGG": 2484, + "CTCGAA": 2485, + "CTCGAT": 2486, + "CTCGAC": 2487, + "CTCGAG": 2488, + "CTCGTA": 2489, + "CTCGTT": 2490, + "CTCGTC": 2491, + "CTCGTG": 2492, + "CTCGCA": 2493, + "CTCGCT": 2494, + "CTCGCC": 2495, + "CTCGCG": 2496, + "CTCGGA": 2497, + "CTCGGT": 2498, + "CTCGGC": 2499, + "CTCGGG": 2500, + "CTGAAA": 2501, + "CTGAAT": 2502, + "CTGAAC": 2503, + "CTGAAG": 2504, + "CTGATA": 2505, + "CTGATT": 2506, + "CTGATC": 2507, + "CTGATG": 2508, + "CTGACA": 2509, + "CTGACT": 2510, + "CTGACC": 2511, + "CTGACG": 2512, + "CTGAGA": 2513, + "CTGAGT": 2514, + "CTGAGC": 2515, + "CTGAGG": 2516, + "CTGTAA": 2517, + "CTGTAT": 2518, + "CTGTAC": 2519, + "CTGTAG": 2520, + "CTGTTA": 2521, + "CTGTTT": 2522, + "CTGTTC": 2523, + "CTGTTG": 2524, + "CTGTCA": 2525, + "CTGTCT": 2526, + "CTGTCC": 2527, + "CTGTCG": 2528, + "CTGTGA": 2529, + "CTGTGT": 2530, + "CTGTGC": 2531, + "CTGTGG": 2532, + "CTGCAA": 2533, + "CTGCAT": 2534, + "CTGCAC": 2535, + "CTGCAG": 2536, + "CTGCTA": 2537, + "CTGCTT": 2538, + "CTGCTC": 2539, + "CTGCTG": 2540, + "CTGCCA": 2541, + "CTGCCT": 2542, + "CTGCCC": 2543, + "CTGCCG": 2544, + "CTGCGA": 2545, + "CTGCGT": 2546, + "CTGCGC": 2547, + "CTGCGG": 2548, + "CTGGAA": 2549, + "CTGGAT": 2550, + "CTGGAC": 2551, + "CTGGAG": 2552, + "CTGGTA": 2553, + "CTGGTT": 2554, + "CTGGTC": 2555, + "CTGGTG": 2556, + "CTGGCA": 2557, + "CTGGCT": 2558, + "CTGGCC": 2559, + "CTGGCG": 2560, + "CTGGGA": 2561, + "CTGGGT": 2562, + "CTGGGC": 2563, + "CTGGGG": 2564, + "CCAAAA": 2565, + "CCAAAT": 2566, + "CCAAAC": 2567, + "CCAAAG": 2568, + "CCAATA": 2569, + "CCAATT": 2570, + "CCAATC": 2571, + "CCAATG": 2572, + "CCAACA": 2573, + "CCAACT": 2574, + "CCAACC": 2575, + "CCAACG": 2576, + "CCAAGA": 2577, + "CCAAGT": 2578, + "CCAAGC": 2579, + "CCAAGG": 2580, + "CCATAA": 2581, + "CCATAT": 2582, + "CCATAC": 2583, + "CCATAG": 2584, + "CCATTA": 2585, + "CCATTT": 2586, + "CCATTC": 2587, + "CCATTG": 2588, + "CCATCA": 2589, + "CCATCT": 2590, + "CCATCC": 2591, + "CCATCG": 2592, + "CCATGA": 2593, + "CCATGT": 2594, + "CCATGC": 2595, + "CCATGG": 2596, + "CCACAA": 2597, + "CCACAT": 2598, + "CCACAC": 2599, + "CCACAG": 2600, + "CCACTA": 2601, + "CCACTT": 2602, + "CCACTC": 2603, + "CCACTG": 2604, + "CCACCA": 2605, + "CCACCT": 2606, + "CCACCC": 2607, + "CCACCG": 2608, + "CCACGA": 2609, + "CCACGT": 2610, + "CCACGC": 2611, + "CCACGG": 2612, + "CCAGAA": 2613, + "CCAGAT": 2614, + "CCAGAC": 2615, + "CCAGAG": 2616, + "CCAGTA": 2617, + "CCAGTT": 2618, + "CCAGTC": 2619, + "CCAGTG": 2620, + "CCAGCA": 2621, + "CCAGCT": 2622, + "CCAGCC": 2623, + "CCAGCG": 2624, + "CCAGGA": 2625, + "CCAGGT": 2626, + "CCAGGC": 2627, + "CCAGGG": 2628, + "CCTAAA": 2629, + "CCTAAT": 2630, + "CCTAAC": 2631, + "CCTAAG": 2632, + "CCTATA": 2633, + "CCTATT": 2634, + "CCTATC": 2635, + "CCTATG": 2636, + "CCTACA": 2637, + "CCTACT": 2638, + "CCTACC": 2639, + "CCTACG": 2640, + "CCTAGA": 2641, + "CCTAGT": 2642, + "CCTAGC": 2643, + "CCTAGG": 2644, + "CCTTAA": 2645, + "CCTTAT": 2646, + "CCTTAC": 2647, + "CCTTAG": 2648, + "CCTTTA": 2649, + "CCTTTT": 2650, + "CCTTTC": 2651, + "CCTTTG": 2652, + "CCTTCA": 2653, + "CCTTCT": 2654, + "CCTTCC": 2655, + "CCTTCG": 2656, + "CCTTGA": 2657, + "CCTTGT": 2658, + "CCTTGC": 2659, + "CCTTGG": 2660, + "CCTCAA": 2661, + "CCTCAT": 2662, + "CCTCAC": 2663, + "CCTCAG": 2664, + "CCTCTA": 2665, + "CCTCTT": 2666, + "CCTCTC": 2667, + "CCTCTG": 2668, + "CCTCCA": 2669, + "CCTCCT": 2670, + "CCTCCC": 2671, + "CCTCCG": 2672, + "CCTCGA": 2673, + "CCTCGT": 2674, + "CCTCGC": 2675, + "CCTCGG": 2676, + "CCTGAA": 2677, + "CCTGAT": 2678, + "CCTGAC": 2679, + "CCTGAG": 2680, + "CCTGTA": 2681, + "CCTGTT": 2682, + "CCTGTC": 2683, + "CCTGTG": 2684, + "CCTGCA": 2685, + "CCTGCT": 2686, + "CCTGCC": 2687, + "CCTGCG": 2688, + "CCTGGA": 2689, + "CCTGGT": 2690, + "CCTGGC": 2691, + "CCTGGG": 2692, + "CCCAAA": 2693, + "CCCAAT": 2694, + "CCCAAC": 2695, + "CCCAAG": 2696, + "CCCATA": 2697, + "CCCATT": 2698, + "CCCATC": 2699, + "CCCATG": 2700, + "CCCACA": 2701, + "CCCACT": 2702, + "CCCACC": 2703, + "CCCACG": 2704, + "CCCAGA": 2705, + "CCCAGT": 2706, + "CCCAGC": 2707, + "CCCAGG": 2708, + "CCCTAA": 2709, + "CCCTAT": 2710, + "CCCTAC": 2711, + "CCCTAG": 2712, + "CCCTTA": 2713, + "CCCTTT": 2714, + "CCCTTC": 2715, + "CCCTTG": 2716, + "CCCTCA": 2717, + "CCCTCT": 2718, + "CCCTCC": 2719, + "CCCTCG": 2720, + "CCCTGA": 2721, + "CCCTGT": 2722, + "CCCTGC": 2723, + "CCCTGG": 2724, + "CCCCAA": 2725, + "CCCCAT": 2726, + "CCCCAC": 2727, + "CCCCAG": 2728, + "CCCCTA": 2729, + "CCCCTT": 2730, + "CCCCTC": 2731, + "CCCCTG": 2732, + "CCCCCA": 2733, + "CCCCCT": 2734, + "CCCCCC": 2735, + "CCCCCG": 2736, + "CCCCGA": 2737, + "CCCCGT": 2738, + "CCCCGC": 2739, + "CCCCGG": 2740, + "CCCGAA": 2741, + "CCCGAT": 2742, + "CCCGAC": 2743, + "CCCGAG": 2744, + "CCCGTA": 2745, + "CCCGTT": 2746, + "CCCGTC": 2747, + "CCCGTG": 2748, + "CCCGCA": 2749, + "CCCGCT": 2750, + "CCCGCC": 2751, + "CCCGCG": 2752, + "CCCGGA": 2753, + "CCCGGT": 2754, + "CCCGGC": 2755, + "CCCGGG": 2756, + "CCGAAA": 2757, + "CCGAAT": 2758, + "CCGAAC": 2759, + "CCGAAG": 2760, + "CCGATA": 2761, + "CCGATT": 2762, + "CCGATC": 2763, + "CCGATG": 2764, + "CCGACA": 2765, + "CCGACT": 2766, + "CCGACC": 2767, + "CCGACG": 2768, + "CCGAGA": 2769, + "CCGAGT": 2770, + "CCGAGC": 2771, + "CCGAGG": 2772, + "CCGTAA": 2773, + "CCGTAT": 2774, + "CCGTAC": 2775, + "CCGTAG": 2776, + "CCGTTA": 2777, + "CCGTTT": 2778, + "CCGTTC": 2779, + "CCGTTG": 2780, + "CCGTCA": 2781, + "CCGTCT": 2782, + "CCGTCC": 2783, + "CCGTCG": 2784, + "CCGTGA": 2785, + "CCGTGT": 2786, + "CCGTGC": 2787, + "CCGTGG": 2788, + "CCGCAA": 2789, + "CCGCAT": 2790, + "CCGCAC": 2791, + "CCGCAG": 2792, + "CCGCTA": 2793, + "CCGCTT": 2794, + "CCGCTC": 2795, + "CCGCTG": 2796, + "CCGCCA": 2797, + "CCGCCT": 2798, + "CCGCCC": 2799, + "CCGCCG": 2800, + "CCGCGA": 2801, + "CCGCGT": 2802, + "CCGCGC": 2803, + "CCGCGG": 2804, + "CCGGAA": 2805, + "CCGGAT": 2806, + "CCGGAC": 2807, + "CCGGAG": 2808, + "CCGGTA": 2809, + "CCGGTT": 2810, + "CCGGTC": 2811, + "CCGGTG": 2812, + "CCGGCA": 2813, + "CCGGCT": 2814, + "CCGGCC": 2815, + "CCGGCG": 2816, + "CCGGGA": 2817, + "CCGGGT": 2818, + "CCGGGC": 2819, + "CCGGGG": 2820, + "CGAAAA": 2821, + "CGAAAT": 2822, + "CGAAAC": 2823, + "CGAAAG": 2824, + "CGAATA": 2825, + "CGAATT": 2826, + "CGAATC": 2827, + "CGAATG": 2828, + "CGAACA": 2829, + "CGAACT": 2830, + "CGAACC": 2831, + "CGAACG": 2832, + "CGAAGA": 2833, + "CGAAGT": 2834, + "CGAAGC": 2835, + "CGAAGG": 2836, + "CGATAA": 2837, + "CGATAT": 2838, + "CGATAC": 2839, + "CGATAG": 2840, + "CGATTA": 2841, + "CGATTT": 2842, + "CGATTC": 2843, + "CGATTG": 2844, + "CGATCA": 2845, + "CGATCT": 2846, + "CGATCC": 2847, + "CGATCG": 2848, + "CGATGA": 2849, + "CGATGT": 2850, + "CGATGC": 2851, + "CGATGG": 2852, + "CGACAA": 2853, + "CGACAT": 2854, + "CGACAC": 2855, + "CGACAG": 2856, + "CGACTA": 2857, + "CGACTT": 2858, + "CGACTC": 2859, + "CGACTG": 2860, + "CGACCA": 2861, + "CGACCT": 2862, + "CGACCC": 2863, + "CGACCG": 2864, + "CGACGA": 2865, + "CGACGT": 2866, + "CGACGC": 2867, + "CGACGG": 2868, + "CGAGAA": 2869, + "CGAGAT": 2870, + "CGAGAC": 2871, + "CGAGAG": 2872, + "CGAGTA": 2873, + "CGAGTT": 2874, + "CGAGTC": 2875, + "CGAGTG": 2876, + "CGAGCA": 2877, + "CGAGCT": 2878, + "CGAGCC": 2879, + "CGAGCG": 2880, + "CGAGGA": 2881, + "CGAGGT": 2882, + "CGAGGC": 2883, + "CGAGGG": 2884, + "CGTAAA": 2885, + "CGTAAT": 2886, + "CGTAAC": 2887, + "CGTAAG": 2888, + "CGTATA": 2889, + "CGTATT": 2890, + "CGTATC": 2891, + "CGTATG": 2892, + "CGTACA": 2893, + "CGTACT": 2894, + "CGTACC": 2895, + "CGTACG": 2896, + "CGTAGA": 2897, + "CGTAGT": 2898, + "CGTAGC": 2899, + "CGTAGG": 2900, + "CGTTAA": 2901, + "CGTTAT": 2902, + "CGTTAC": 2903, + "CGTTAG": 2904, + "CGTTTA": 2905, + "CGTTTT": 2906, + "CGTTTC": 2907, + "CGTTTG": 2908, + "CGTTCA": 2909, + "CGTTCT": 2910, + "CGTTCC": 2911, + "CGTTCG": 2912, + "CGTTGA": 2913, + "CGTTGT": 2914, + "CGTTGC": 2915, + "CGTTGG": 2916, + "CGTCAA": 2917, + "CGTCAT": 2918, + "CGTCAC": 2919, + "CGTCAG": 2920, + "CGTCTA": 2921, + "CGTCTT": 2922, + "CGTCTC": 2923, + "CGTCTG": 2924, + "CGTCCA": 2925, + "CGTCCT": 2926, + "CGTCCC": 2927, + "CGTCCG": 2928, + "CGTCGA": 2929, + "CGTCGT": 2930, + "CGTCGC": 2931, + "CGTCGG": 2932, + "CGTGAA": 2933, + "CGTGAT": 2934, + "CGTGAC": 2935, + "CGTGAG": 2936, + "CGTGTA": 2937, + "CGTGTT": 2938, + "CGTGTC": 2939, + "CGTGTG": 2940, + "CGTGCA": 2941, + "CGTGCT": 2942, + "CGTGCC": 2943, + "CGTGCG": 2944, + "CGTGGA": 2945, + "CGTGGT": 2946, + "CGTGGC": 2947, + "CGTGGG": 2948, + "CGCAAA": 2949, + "CGCAAT": 2950, + "CGCAAC": 2951, + "CGCAAG": 2952, + "CGCATA": 2953, + "CGCATT": 2954, + "CGCATC": 2955, + "CGCATG": 2956, + "CGCACA": 2957, + "CGCACT": 2958, + "CGCACC": 2959, + "CGCACG": 2960, + "CGCAGA": 2961, + "CGCAGT": 2962, + "CGCAGC": 2963, + "CGCAGG": 2964, + "CGCTAA": 2965, + "CGCTAT": 2966, + "CGCTAC": 2967, + "CGCTAG": 2968, + "CGCTTA": 2969, + "CGCTTT": 2970, + "CGCTTC": 2971, + "CGCTTG": 2972, + "CGCTCA": 2973, + "CGCTCT": 2974, + "CGCTCC": 2975, + "CGCTCG": 2976, + "CGCTGA": 2977, + "CGCTGT": 2978, + "CGCTGC": 2979, + "CGCTGG": 2980, + "CGCCAA": 2981, + "CGCCAT": 2982, + "CGCCAC": 2983, + "CGCCAG": 2984, + "CGCCTA": 2985, + "CGCCTT": 2986, + "CGCCTC": 2987, + "CGCCTG": 2988, + "CGCCCA": 2989, + "CGCCCT": 2990, + "CGCCCC": 2991, + "CGCCCG": 2992, + "CGCCGA": 2993, + "CGCCGT": 2994, + "CGCCGC": 2995, + "CGCCGG": 2996, + "CGCGAA": 2997, + "CGCGAT": 2998, + "CGCGAC": 2999, + "CGCGAG": 3000, + "CGCGTA": 3001, + "CGCGTT": 3002, + "CGCGTC": 3003, + "CGCGTG": 3004, + "CGCGCA": 3005, + "CGCGCT": 3006, + "CGCGCC": 3007, + "CGCGCG": 3008, + "CGCGGA": 3009, + "CGCGGT": 3010, + "CGCGGC": 3011, + "CGCGGG": 3012, + "CGGAAA": 3013, + "CGGAAT": 3014, + "CGGAAC": 3015, + "CGGAAG": 3016, + "CGGATA": 3017, + "CGGATT": 3018, + "CGGATC": 3019, + "CGGATG": 3020, + "CGGACA": 3021, + "CGGACT": 3022, + "CGGACC": 3023, + "CGGACG": 3024, + "CGGAGA": 3025, + "CGGAGT": 3026, + "CGGAGC": 3027, + "CGGAGG": 3028, + "CGGTAA": 3029, + "CGGTAT": 3030, + "CGGTAC": 3031, + "CGGTAG": 3032, + "CGGTTA": 3033, + "CGGTTT": 3034, + "CGGTTC": 3035, + "CGGTTG": 3036, + "CGGTCA": 3037, + "CGGTCT": 3038, + "CGGTCC": 3039, + "CGGTCG": 3040, + "CGGTGA": 3041, + "CGGTGT": 3042, + "CGGTGC": 3043, + "CGGTGG": 3044, + "CGGCAA": 3045, + "CGGCAT": 3046, + "CGGCAC": 3047, + "CGGCAG": 3048, + "CGGCTA": 3049, + "CGGCTT": 3050, + "CGGCTC": 3051, + "CGGCTG": 3052, + "CGGCCA": 3053, + "CGGCCT": 3054, + "CGGCCC": 3055, + "CGGCCG": 3056, + "CGGCGA": 3057, + "CGGCGT": 3058, + "CGGCGC": 3059, + "CGGCGG": 3060, + "CGGGAA": 3061, + "CGGGAT": 3062, + "CGGGAC": 3063, + "CGGGAG": 3064, + "CGGGTA": 3065, + "CGGGTT": 3066, + "CGGGTC": 3067, + "CGGGTG": 3068, + "CGGGCA": 3069, + "CGGGCT": 3070, + "CGGGCC": 3071, + "CGGGCG": 3072, + "CGGGGA": 3073, + "CGGGGT": 3074, + "CGGGGC": 3075, + "CGGGGG": 3076, + "GAAAAA": 3077, + "GAAAAT": 3078, + "GAAAAC": 3079, + "GAAAAG": 3080, + "GAAATA": 3081, + "GAAATT": 3082, + "GAAATC": 3083, + "GAAATG": 3084, + "GAAACA": 3085, + "GAAACT": 3086, + "GAAACC": 3087, + "GAAACG": 3088, + "GAAAGA": 3089, + "GAAAGT": 3090, + "GAAAGC": 3091, + "GAAAGG": 3092, + "GAATAA": 3093, + "GAATAT": 3094, + "GAATAC": 3095, + "GAATAG": 3096, + "GAATTA": 3097, + "GAATTT": 3098, + "GAATTC": 3099, + "GAATTG": 3100, + "GAATCA": 3101, + "GAATCT": 3102, + "GAATCC": 3103, + "GAATCG": 3104, + "GAATGA": 3105, + "GAATGT": 3106, + "GAATGC": 3107, + "GAATGG": 3108, + "GAACAA": 3109, + "GAACAT": 3110, + "GAACAC": 3111, + "GAACAG": 3112, + "GAACTA": 3113, + "GAACTT": 3114, + "GAACTC": 3115, + "GAACTG": 3116, + "GAACCA": 3117, + "GAACCT": 3118, + "GAACCC": 3119, + "GAACCG": 3120, + "GAACGA": 3121, + "GAACGT": 3122, + "GAACGC": 3123, + "GAACGG": 3124, + "GAAGAA": 3125, + "GAAGAT": 3126, + "GAAGAC": 3127, + "GAAGAG": 3128, + "GAAGTA": 3129, + "GAAGTT": 3130, + "GAAGTC": 3131, + "GAAGTG": 3132, + "GAAGCA": 3133, + "GAAGCT": 3134, + "GAAGCC": 3135, + "GAAGCG": 3136, + "GAAGGA": 3137, + "GAAGGT": 3138, + "GAAGGC": 3139, + "GAAGGG": 3140, + "GATAAA": 3141, + "GATAAT": 3142, + "GATAAC": 3143, + "GATAAG": 3144, + "GATATA": 3145, + "GATATT": 3146, + "GATATC": 3147, + "GATATG": 3148, + "GATACA": 3149, + "GATACT": 3150, + "GATACC": 3151, + "GATACG": 3152, + "GATAGA": 3153, + "GATAGT": 3154, + "GATAGC": 3155, + "GATAGG": 3156, + "GATTAA": 3157, + "GATTAT": 3158, + "GATTAC": 3159, + "GATTAG": 3160, + "GATTTA": 3161, + "GATTTT": 3162, + "GATTTC": 3163, + "GATTTG": 3164, + "GATTCA": 3165, + "GATTCT": 3166, + "GATTCC": 3167, + "GATTCG": 3168, + "GATTGA": 3169, + "GATTGT": 3170, + "GATTGC": 3171, + "GATTGG": 3172, + "GATCAA": 3173, + "GATCAT": 3174, + "GATCAC": 3175, + "GATCAG": 3176, + "GATCTA": 3177, + "GATCTT": 3178, + "GATCTC": 3179, + "GATCTG": 3180, + "GATCCA": 3181, + "GATCCT": 3182, + "GATCCC": 3183, + "GATCCG": 3184, + "GATCGA": 3185, + "GATCGT": 3186, + "GATCGC": 3187, + "GATCGG": 3188, + "GATGAA": 3189, + "GATGAT": 3190, + "GATGAC": 3191, + "GATGAG": 3192, + "GATGTA": 3193, + "GATGTT": 3194, + "GATGTC": 3195, + "GATGTG": 3196, + "GATGCA": 3197, + "GATGCT": 3198, + "GATGCC": 3199, + "GATGCG": 3200, + "GATGGA": 3201, + "GATGGT": 3202, + "GATGGC": 3203, + "GATGGG": 3204, + "GACAAA": 3205, + "GACAAT": 3206, + "GACAAC": 3207, + "GACAAG": 3208, + "GACATA": 3209, + "GACATT": 3210, + "GACATC": 3211, + "GACATG": 3212, + "GACACA": 3213, + "GACACT": 3214, + "GACACC": 3215, + "GACACG": 3216, + "GACAGA": 3217, + "GACAGT": 3218, + "GACAGC": 3219, + "GACAGG": 3220, + "GACTAA": 3221, + "GACTAT": 3222, + "GACTAC": 3223, + "GACTAG": 3224, + "GACTTA": 3225, + "GACTTT": 3226, + "GACTTC": 3227, + "GACTTG": 3228, + "GACTCA": 3229, + "GACTCT": 3230, + "GACTCC": 3231, + "GACTCG": 3232, + "GACTGA": 3233, + "GACTGT": 3234, + "GACTGC": 3235, + "GACTGG": 3236, + "GACCAA": 3237, + "GACCAT": 3238, + "GACCAC": 3239, + "GACCAG": 3240, + "GACCTA": 3241, + "GACCTT": 3242, + "GACCTC": 3243, + "GACCTG": 3244, + "GACCCA": 3245, + "GACCCT": 3246, + "GACCCC": 3247, + "GACCCG": 3248, + "GACCGA": 3249, + "GACCGT": 3250, + "GACCGC": 3251, + "GACCGG": 3252, + "GACGAA": 3253, + "GACGAT": 3254, + "GACGAC": 3255, + "GACGAG": 3256, + "GACGTA": 3257, + "GACGTT": 3258, + "GACGTC": 3259, + "GACGTG": 3260, + "GACGCA": 3261, + "GACGCT": 3262, + "GACGCC": 3263, + "GACGCG": 3264, + "GACGGA": 3265, + "GACGGT": 3266, + "GACGGC": 3267, + "GACGGG": 3268, + "GAGAAA": 3269, + "GAGAAT": 3270, + "GAGAAC": 3271, + "GAGAAG": 3272, + "GAGATA": 3273, + "GAGATT": 3274, + "GAGATC": 3275, + "GAGATG": 3276, + "GAGACA": 3277, + "GAGACT": 3278, + "GAGACC": 3279, + "GAGACG": 3280, + "GAGAGA": 3281, + "GAGAGT": 3282, + "GAGAGC": 3283, + "GAGAGG": 3284, + "GAGTAA": 3285, + "GAGTAT": 3286, + "GAGTAC": 3287, + "GAGTAG": 3288, + "GAGTTA": 3289, + "GAGTTT": 3290, + "GAGTTC": 3291, + "GAGTTG": 3292, + "GAGTCA": 3293, + "GAGTCT": 3294, + "GAGTCC": 3295, + "GAGTCG": 3296, + "GAGTGA": 3297, + "GAGTGT": 3298, + "GAGTGC": 3299, + "GAGTGG": 3300, + "GAGCAA": 3301, + "GAGCAT": 3302, + "GAGCAC": 3303, + "GAGCAG": 3304, + "GAGCTA": 3305, + "GAGCTT": 3306, + "GAGCTC": 3307, + "GAGCTG": 3308, + "GAGCCA": 3309, + "GAGCCT": 3310, + "GAGCCC": 3311, + "GAGCCG": 3312, + "GAGCGA": 3313, + "GAGCGT": 3314, + "GAGCGC": 3315, + "GAGCGG": 3316, + "GAGGAA": 3317, + "GAGGAT": 3318, + "GAGGAC": 3319, + "GAGGAG": 3320, + "GAGGTA": 3321, + "GAGGTT": 3322, + "GAGGTC": 3323, + "GAGGTG": 3324, + "GAGGCA": 3325, + "GAGGCT": 3326, + "GAGGCC": 3327, + "GAGGCG": 3328, + "GAGGGA": 3329, + "GAGGGT": 3330, + "GAGGGC": 3331, + "GAGGGG": 3332, + "GTAAAA": 3333, + "GTAAAT": 3334, + "GTAAAC": 3335, + "GTAAAG": 3336, + "GTAATA": 3337, + "GTAATT": 3338, + "GTAATC": 3339, + "GTAATG": 3340, + "GTAACA": 3341, + "GTAACT": 3342, + "GTAACC": 3343, + "GTAACG": 3344, + "GTAAGA": 3345, + "GTAAGT": 3346, + "GTAAGC": 3347, + "GTAAGG": 3348, + "GTATAA": 3349, + "GTATAT": 3350, + "GTATAC": 3351, + "GTATAG": 3352, + "GTATTA": 3353, + "GTATTT": 3354, + "GTATTC": 3355, + "GTATTG": 3356, + "GTATCA": 3357, + "GTATCT": 3358, + "GTATCC": 3359, + "GTATCG": 3360, + "GTATGA": 3361, + "GTATGT": 3362, + "GTATGC": 3363, + "GTATGG": 3364, + "GTACAA": 3365, + "GTACAT": 3366, + "GTACAC": 3367, + "GTACAG": 3368, + "GTACTA": 3369, + "GTACTT": 3370, + "GTACTC": 3371, + "GTACTG": 3372, + "GTACCA": 3373, + "GTACCT": 3374, + "GTACCC": 3375, + "GTACCG": 3376, + "GTACGA": 3377, + "GTACGT": 3378, + "GTACGC": 3379, + "GTACGG": 3380, + "GTAGAA": 3381, + "GTAGAT": 3382, + "GTAGAC": 3383, + "GTAGAG": 3384, + "GTAGTA": 3385, + "GTAGTT": 3386, + "GTAGTC": 3387, + "GTAGTG": 3388, + "GTAGCA": 3389, + "GTAGCT": 3390, + "GTAGCC": 3391, + "GTAGCG": 3392, + "GTAGGA": 3393, + "GTAGGT": 3394, + "GTAGGC": 3395, + "GTAGGG": 3396, + "GTTAAA": 3397, + "GTTAAT": 3398, + "GTTAAC": 3399, + "GTTAAG": 3400, + "GTTATA": 3401, + "GTTATT": 3402, + "GTTATC": 3403, + "GTTATG": 3404, + "GTTACA": 3405, + "GTTACT": 3406, + "GTTACC": 3407, + "GTTACG": 3408, + "GTTAGA": 3409, + "GTTAGT": 3410, + "GTTAGC": 3411, + "GTTAGG": 3412, + "GTTTAA": 3413, + "GTTTAT": 3414, + "GTTTAC": 3415, + "GTTTAG": 3416, + "GTTTTA": 3417, + "GTTTTT": 3418, + "GTTTTC": 3419, + "GTTTTG": 3420, + "GTTTCA": 3421, + "GTTTCT": 3422, + "GTTTCC": 3423, + "GTTTCG": 3424, + "GTTTGA": 3425, + "GTTTGT": 3426, + "GTTTGC": 3427, + "GTTTGG": 3428, + "GTTCAA": 3429, + "GTTCAT": 3430, + "GTTCAC": 3431, + "GTTCAG": 3432, + "GTTCTA": 3433, + "GTTCTT": 3434, + "GTTCTC": 3435, + "GTTCTG": 3436, + "GTTCCA": 3437, + "GTTCCT": 3438, + "GTTCCC": 3439, + "GTTCCG": 3440, + "GTTCGA": 3441, + "GTTCGT": 3442, + "GTTCGC": 3443, + "GTTCGG": 3444, + "GTTGAA": 3445, + "GTTGAT": 3446, + "GTTGAC": 3447, + "GTTGAG": 3448, + "GTTGTA": 3449, + "GTTGTT": 3450, + "GTTGTC": 3451, + "GTTGTG": 3452, + "GTTGCA": 3453, + "GTTGCT": 3454, + "GTTGCC": 3455, + "GTTGCG": 3456, + "GTTGGA": 3457, + "GTTGGT": 3458, + "GTTGGC": 3459, + "GTTGGG": 3460, + "GTCAAA": 3461, + "GTCAAT": 3462, + "GTCAAC": 3463, + "GTCAAG": 3464, + "GTCATA": 3465, + "GTCATT": 3466, + "GTCATC": 3467, + "GTCATG": 3468, + "GTCACA": 3469, + "GTCACT": 3470, + "GTCACC": 3471, + "GTCACG": 3472, + "GTCAGA": 3473, + "GTCAGT": 3474, + "GTCAGC": 3475, + "GTCAGG": 3476, + "GTCTAA": 3477, + "GTCTAT": 3478, + "GTCTAC": 3479, + "GTCTAG": 3480, + "GTCTTA": 3481, + "GTCTTT": 3482, + "GTCTTC": 3483, + "GTCTTG": 3484, + "GTCTCA": 3485, + "GTCTCT": 3486, + "GTCTCC": 3487, + "GTCTCG": 3488, + "GTCTGA": 3489, + "GTCTGT": 3490, + "GTCTGC": 3491, + "GTCTGG": 3492, + "GTCCAA": 3493, + "GTCCAT": 3494, + "GTCCAC": 3495, + "GTCCAG": 3496, + "GTCCTA": 3497, + "GTCCTT": 3498, + "GTCCTC": 3499, + "GTCCTG": 3500, + "GTCCCA": 3501, + "GTCCCT": 3502, + "GTCCCC": 3503, + "GTCCCG": 3504, + "GTCCGA": 3505, + "GTCCGT": 3506, + "GTCCGC": 3507, + "GTCCGG": 3508, + "GTCGAA": 3509, + "GTCGAT": 3510, + "GTCGAC": 3511, + "GTCGAG": 3512, + "GTCGTA": 3513, + "GTCGTT": 3514, + "GTCGTC": 3515, + "GTCGTG": 3516, + "GTCGCA": 3517, + "GTCGCT": 3518, + "GTCGCC": 3519, + "GTCGCG": 3520, + "GTCGGA": 3521, + "GTCGGT": 3522, + "GTCGGC": 3523, + "GTCGGG": 3524, + "GTGAAA": 3525, + "GTGAAT": 3526, + "GTGAAC": 3527, + "GTGAAG": 3528, + "GTGATA": 3529, + "GTGATT": 3530, + "GTGATC": 3531, + "GTGATG": 3532, + "GTGACA": 3533, + "GTGACT": 3534, + "GTGACC": 3535, + "GTGACG": 3536, + "GTGAGA": 3537, + "GTGAGT": 3538, + "GTGAGC": 3539, + "GTGAGG": 3540, + "GTGTAA": 3541, + "GTGTAT": 3542, + "GTGTAC": 3543, + "GTGTAG": 3544, + "GTGTTA": 3545, + "GTGTTT": 3546, + "GTGTTC": 3547, + "GTGTTG": 3548, + "GTGTCA": 3549, + "GTGTCT": 3550, + "GTGTCC": 3551, + "GTGTCG": 3552, + "GTGTGA": 3553, + "GTGTGT": 3554, + "GTGTGC": 3555, + "GTGTGG": 3556, + "GTGCAA": 3557, + "GTGCAT": 3558, + "GTGCAC": 3559, + "GTGCAG": 3560, + "GTGCTA": 3561, + "GTGCTT": 3562, + "GTGCTC": 3563, + "GTGCTG": 3564, + "GTGCCA": 3565, + "GTGCCT": 3566, + "GTGCCC": 3567, + "GTGCCG": 3568, + "GTGCGA": 3569, + "GTGCGT": 3570, + "GTGCGC": 3571, + "GTGCGG": 3572, + "GTGGAA": 3573, + "GTGGAT": 3574, + "GTGGAC": 3575, + "GTGGAG": 3576, + "GTGGTA": 3577, + "GTGGTT": 3578, + "GTGGTC": 3579, + "GTGGTG": 3580, + "GTGGCA": 3581, + "GTGGCT": 3582, + "GTGGCC": 3583, + "GTGGCG": 3584, + "GTGGGA": 3585, + "GTGGGT": 3586, + "GTGGGC": 3587, + "GTGGGG": 3588, + "GCAAAA": 3589, + "GCAAAT": 3590, + "GCAAAC": 3591, + "GCAAAG": 3592, + "GCAATA": 3593, + "GCAATT": 3594, + "GCAATC": 3595, + "GCAATG": 3596, + "GCAACA": 3597, + "GCAACT": 3598, + "GCAACC": 3599, + "GCAACG": 3600, + "GCAAGA": 3601, + "GCAAGT": 3602, + "GCAAGC": 3603, + "GCAAGG": 3604, + "GCATAA": 3605, + "GCATAT": 3606, + "GCATAC": 3607, + "GCATAG": 3608, + "GCATTA": 3609, + "GCATTT": 3610, + "GCATTC": 3611, + "GCATTG": 3612, + "GCATCA": 3613, + "GCATCT": 3614, + "GCATCC": 3615, + "GCATCG": 3616, + "GCATGA": 3617, + "GCATGT": 3618, + "GCATGC": 3619, + "GCATGG": 3620, + "GCACAA": 3621, + "GCACAT": 3622, + "GCACAC": 3623, + "GCACAG": 3624, + "GCACTA": 3625, + "GCACTT": 3626, + "GCACTC": 3627, + "GCACTG": 3628, + "GCACCA": 3629, + "GCACCT": 3630, + "GCACCC": 3631, + "GCACCG": 3632, + "GCACGA": 3633, + "GCACGT": 3634, + "GCACGC": 3635, + "GCACGG": 3636, + "GCAGAA": 3637, + "GCAGAT": 3638, + "GCAGAC": 3639, + "GCAGAG": 3640, + "GCAGTA": 3641, + "GCAGTT": 3642, + "GCAGTC": 3643, + "GCAGTG": 3644, + "GCAGCA": 3645, + "GCAGCT": 3646, + "GCAGCC": 3647, + "GCAGCG": 3648, + "GCAGGA": 3649, + "GCAGGT": 3650, + "GCAGGC": 3651, + "GCAGGG": 3652, + "GCTAAA": 3653, + "GCTAAT": 3654, + "GCTAAC": 3655, + "GCTAAG": 3656, + "GCTATA": 3657, + "GCTATT": 3658, + "GCTATC": 3659, + "GCTATG": 3660, + "GCTACA": 3661, + "GCTACT": 3662, + "GCTACC": 3663, + "GCTACG": 3664, + "GCTAGA": 3665, + "GCTAGT": 3666, + "GCTAGC": 3667, + "GCTAGG": 3668, + "GCTTAA": 3669, + "GCTTAT": 3670, + "GCTTAC": 3671, + "GCTTAG": 3672, + "GCTTTA": 3673, + "GCTTTT": 3674, + "GCTTTC": 3675, + "GCTTTG": 3676, + "GCTTCA": 3677, + "GCTTCT": 3678, + "GCTTCC": 3679, + "GCTTCG": 3680, + "GCTTGA": 3681, + "GCTTGT": 3682, + "GCTTGC": 3683, + "GCTTGG": 3684, + "GCTCAA": 3685, + "GCTCAT": 3686, + "GCTCAC": 3687, + "GCTCAG": 3688, + "GCTCTA": 3689, + "GCTCTT": 3690, + "GCTCTC": 3691, + "GCTCTG": 3692, + "GCTCCA": 3693, + "GCTCCT": 3694, + "GCTCCC": 3695, + "GCTCCG": 3696, + "GCTCGA": 3697, + "GCTCGT": 3698, + "GCTCGC": 3699, + "GCTCGG": 3700, + "GCTGAA": 3701, + "GCTGAT": 3702, + "GCTGAC": 3703, + "GCTGAG": 3704, + "GCTGTA": 3705, + "GCTGTT": 3706, + "GCTGTC": 3707, + "GCTGTG": 3708, + "GCTGCA": 3709, + "GCTGCT": 3710, + "GCTGCC": 3711, + "GCTGCG": 3712, + "GCTGGA": 3713, + "GCTGGT": 3714, + "GCTGGC": 3715, + "GCTGGG": 3716, + "GCCAAA": 3717, + "GCCAAT": 3718, + "GCCAAC": 3719, + "GCCAAG": 3720, + "GCCATA": 3721, + "GCCATT": 3722, + "GCCATC": 3723, + "GCCATG": 3724, + "GCCACA": 3725, + "GCCACT": 3726, + "GCCACC": 3727, + "GCCACG": 3728, + "GCCAGA": 3729, + "GCCAGT": 3730, + "GCCAGC": 3731, + "GCCAGG": 3732, + "GCCTAA": 3733, + "GCCTAT": 3734, + "GCCTAC": 3735, + "GCCTAG": 3736, + "GCCTTA": 3737, + "GCCTTT": 3738, + "GCCTTC": 3739, + "GCCTTG": 3740, + "GCCTCA": 3741, + "GCCTCT": 3742, + "GCCTCC": 3743, + "GCCTCG": 3744, + "GCCTGA": 3745, + "GCCTGT": 3746, + "GCCTGC": 3747, + "GCCTGG": 3748, + "GCCCAA": 3749, + "GCCCAT": 3750, + "GCCCAC": 3751, + "GCCCAG": 3752, + "GCCCTA": 3753, + "GCCCTT": 3754, + "GCCCTC": 3755, + "GCCCTG": 3756, + "GCCCCA": 3757, + "GCCCCT": 3758, + "GCCCCC": 3759, + "GCCCCG": 3760, + "GCCCGA": 3761, + "GCCCGT": 3762, + "GCCCGC": 3763, + "GCCCGG": 3764, + "GCCGAA": 3765, + "GCCGAT": 3766, + "GCCGAC": 3767, + "GCCGAG": 3768, + "GCCGTA": 3769, + "GCCGTT": 3770, + "GCCGTC": 3771, + "GCCGTG": 3772, + "GCCGCA": 3773, + "GCCGCT": 3774, + "GCCGCC": 3775, + "GCCGCG": 3776, + "GCCGGA": 3777, + "GCCGGT": 3778, + "GCCGGC": 3779, + "GCCGGG": 3780, + "GCGAAA": 3781, + "GCGAAT": 3782, + "GCGAAC": 3783, + "GCGAAG": 3784, + "GCGATA": 3785, + "GCGATT": 3786, + "GCGATC": 3787, + "GCGATG": 3788, + "GCGACA": 3789, + "GCGACT": 3790, + "GCGACC": 3791, + "GCGACG": 3792, + "GCGAGA": 3793, + "GCGAGT": 3794, + "GCGAGC": 3795, + "GCGAGG": 3796, + "GCGTAA": 3797, + "GCGTAT": 3798, + "GCGTAC": 3799, + "GCGTAG": 3800, + "GCGTTA": 3801, + "GCGTTT": 3802, + "GCGTTC": 3803, + "GCGTTG": 3804, + "GCGTCA": 3805, + "GCGTCT": 3806, + "GCGTCC": 3807, + "GCGTCG": 3808, + "GCGTGA": 3809, + "GCGTGT": 3810, + "GCGTGC": 3811, + "GCGTGG": 3812, + "GCGCAA": 3813, + "GCGCAT": 3814, + "GCGCAC": 3815, + "GCGCAG": 3816, + "GCGCTA": 3817, + "GCGCTT": 3818, + "GCGCTC": 3819, + "GCGCTG": 3820, + "GCGCCA": 3821, + "GCGCCT": 3822, + "GCGCCC": 3823, + "GCGCCG": 3824, + "GCGCGA": 3825, + "GCGCGT": 3826, + "GCGCGC": 3827, + "GCGCGG": 3828, + "GCGGAA": 3829, + "GCGGAT": 3830, + "GCGGAC": 3831, + "GCGGAG": 3832, + "GCGGTA": 3833, + "GCGGTT": 3834, + "GCGGTC": 3835, + "GCGGTG": 3836, + "GCGGCA": 3837, + "GCGGCT": 3838, + "GCGGCC": 3839, + "GCGGCG": 3840, + "GCGGGA": 3841, + "GCGGGT": 3842, + "GCGGGC": 3843, + "GCGGGG": 3844, + "GGAAAA": 3845, + "GGAAAT": 3846, + "GGAAAC": 3847, + "GGAAAG": 3848, + "GGAATA": 3849, + "GGAATT": 3850, + "GGAATC": 3851, + "GGAATG": 3852, + "GGAACA": 3853, + "GGAACT": 3854, + "GGAACC": 3855, + "GGAACG": 3856, + "GGAAGA": 3857, + "GGAAGT": 3858, + "GGAAGC": 3859, + "GGAAGG": 3860, + "GGATAA": 3861, + "GGATAT": 3862, + "GGATAC": 3863, + "GGATAG": 3864, + "GGATTA": 3865, + "GGATTT": 3866, + "GGATTC": 3867, + "GGATTG": 3868, + "GGATCA": 3869, + "GGATCT": 3870, + "GGATCC": 3871, + "GGATCG": 3872, + "GGATGA": 3873, + "GGATGT": 3874, + "GGATGC": 3875, + "GGATGG": 3876, + "GGACAA": 3877, + "GGACAT": 3878, + "GGACAC": 3879, + "GGACAG": 3880, + "GGACTA": 3881, + "GGACTT": 3882, + "GGACTC": 3883, + "GGACTG": 3884, + "GGACCA": 3885, + "GGACCT": 3886, + "GGACCC": 3887, + "GGACCG": 3888, + "GGACGA": 3889, + "GGACGT": 3890, + "GGACGC": 3891, + "GGACGG": 3892, + "GGAGAA": 3893, + "GGAGAT": 3894, + "GGAGAC": 3895, + "GGAGAG": 3896, + "GGAGTA": 3897, + "GGAGTT": 3898, + "GGAGTC": 3899, + "GGAGTG": 3900, + "GGAGCA": 3901, + "GGAGCT": 3902, + "GGAGCC": 3903, + "GGAGCG": 3904, + "GGAGGA": 3905, + "GGAGGT": 3906, + "GGAGGC": 3907, + "GGAGGG": 3908, + "GGTAAA": 3909, + "GGTAAT": 3910, + "GGTAAC": 3911, + "GGTAAG": 3912, + "GGTATA": 3913, + "GGTATT": 3914, + "GGTATC": 3915, + "GGTATG": 3916, + "GGTACA": 3917, + "GGTACT": 3918, + "GGTACC": 3919, + "GGTACG": 3920, + "GGTAGA": 3921, + "GGTAGT": 3922, + "GGTAGC": 3923, + "GGTAGG": 3924, + "GGTTAA": 3925, + "GGTTAT": 3926, + "GGTTAC": 3927, + "GGTTAG": 3928, + "GGTTTA": 3929, + "GGTTTT": 3930, + "GGTTTC": 3931, + "GGTTTG": 3932, + "GGTTCA": 3933, + "GGTTCT": 3934, + "GGTTCC": 3935, + "GGTTCG": 3936, + "GGTTGA": 3937, + "GGTTGT": 3938, + "GGTTGC": 3939, + "GGTTGG": 3940, + "GGTCAA": 3941, + "GGTCAT": 3942, + "GGTCAC": 3943, + "GGTCAG": 3944, + "GGTCTA": 3945, + "GGTCTT": 3946, + "GGTCTC": 3947, + "GGTCTG": 3948, + "GGTCCA": 3949, + "GGTCCT": 3950, + "GGTCCC": 3951, + "GGTCCG": 3952, + "GGTCGA": 3953, + "GGTCGT": 3954, + "GGTCGC": 3955, + "GGTCGG": 3956, + "GGTGAA": 3957, + "GGTGAT": 3958, + "GGTGAC": 3959, + "GGTGAG": 3960, + "GGTGTA": 3961, + "GGTGTT": 3962, + "GGTGTC": 3963, + "GGTGTG": 3964, + "GGTGCA": 3965, + "GGTGCT": 3966, + "GGTGCC": 3967, + "GGTGCG": 3968, + "GGTGGA": 3969, + "GGTGGT": 3970, + "GGTGGC": 3971, + "GGTGGG": 3972, + "GGCAAA": 3973, + "GGCAAT": 3974, + "GGCAAC": 3975, + "GGCAAG": 3976, + "GGCATA": 3977, + "GGCATT": 3978, + "GGCATC": 3979, + "GGCATG": 3980, + "GGCACA": 3981, + "GGCACT": 3982, + "GGCACC": 3983, + "GGCACG": 3984, + "GGCAGA": 3985, + "GGCAGT": 3986, + "GGCAGC": 3987, + "GGCAGG": 3988, + "GGCTAA": 3989, + "GGCTAT": 3990, + "GGCTAC": 3991, + "GGCTAG": 3992, + "GGCTTA": 3993, + "GGCTTT": 3994, + "GGCTTC": 3995, + "GGCTTG": 3996, + "GGCTCA": 3997, + "GGCTCT": 3998, + "GGCTCC": 3999, + "GGCTCG": 4000, + "GGCTGA": 4001, + "GGCTGT": 4002, + "GGCTGC": 4003, + "GGCTGG": 4004, + "GGCCAA": 4005, + "GGCCAT": 4006, + "GGCCAC": 4007, + "GGCCAG": 4008, + "GGCCTA": 4009, + "GGCCTT": 4010, + "GGCCTC": 4011, + "GGCCTG": 4012, + "GGCCCA": 4013, + "GGCCCT": 4014, + "GGCCCC": 4015, + "GGCCCG": 4016, + "GGCCGA": 4017, + "GGCCGT": 4018, + "GGCCGC": 4019, + "GGCCGG": 4020, + "GGCGAA": 4021, + "GGCGAT": 4022, + "GGCGAC": 4023, + "GGCGAG": 4024, + "GGCGTA": 4025, + "GGCGTT": 4026, + "GGCGTC": 4027, + "GGCGTG": 4028, + "GGCGCA": 4029, + "GGCGCT": 4030, + "GGCGCC": 4031, + "GGCGCG": 4032, + "GGCGGA": 4033, + "GGCGGT": 4034, + "GGCGGC": 4035, + "GGCGGG": 4036, + "GGGAAA": 4037, + "GGGAAT": 4038, + "GGGAAC": 4039, + "GGGAAG": 4040, + "GGGATA": 4041, + "GGGATT": 4042, + "GGGATC": 4043, + "GGGATG": 4044, + "GGGACA": 4045, + "GGGACT": 4046, + "GGGACC": 4047, + "GGGACG": 4048, + "GGGAGA": 4049, + "GGGAGT": 4050, + "GGGAGC": 4051, + "GGGAGG": 4052, + "GGGTAA": 4053, + "GGGTAT": 4054, + "GGGTAC": 4055, + "GGGTAG": 4056, + "GGGTTA": 4057, + "GGGTTT": 4058, + "GGGTTC": 4059, + "GGGTTG": 4060, + "GGGTCA": 4061, + "GGGTCT": 4062, + "GGGTCC": 4063, + "GGGTCG": 4064, + "GGGTGA": 4065, + "GGGTGT": 4066, + "GGGTGC": 4067, + "GGGTGG": 4068, + "GGGCAA": 4069, + "GGGCAT": 4070, + "GGGCAC": 4071, + "GGGCAG": 4072, + "GGGCTA": 4073, + "GGGCTT": 4074, + "GGGCTC": 4075, + "GGGCTG": 4076, + "GGGCCA": 4077, + "GGGCCT": 4078, + "GGGCCC": 4079, + "GGGCCG": 4080, + "GGGCGA": 4081, + "GGGCGT": 4082, + "GGGCGC": 4083, + "GGGCGG": 4084, + "GGGGAA": 4085, + "GGGGAT": 4086, + "GGGGAC": 4087, + "GGGGAG": 4088, + "GGGGTA": 4089, + "GGGGTT": 4090, + "GGGGTC": 4091, + "GGGGTG": 4092, + "GGGGCA": 4093, + "GGGGCT": 4094, + "GGGGCC": 4095, + "GGGGCG": 4096, + "GGGGGA": 4097, + "GGGGGT": 4098, + "GGGGGC": 4099, + "GGGGGG": 4100 + } + } +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-3000/tokenizer_config.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-3000/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..604e2b1b29af1dcb3920d2828b6a90276e5e3d02 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-3000/tokenizer_config.json @@ -0,0 +1,58 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "[UNK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "[CLS]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "[SEP]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "4": { + "content": "[MASK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "clean_up_tokenization_spaces": true, + "cls_token": "[CLS]", + "do_basic_tokenize": true, + "do_lower_case": false, + "mask_token": "[MASK]", + "max_len": 512, + "model_max_length": 512, + "never_split": null, + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "strip_accents": null, + "tokenize_chinese_chars": true, + "tokenizer_class": "BertTokenizer", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-3000/trainer_state.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-3000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..22115f5b2f5db256e81ee063a88e4c95642e2871 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-3000/trainer_state.json @@ -0,0 +1,873 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 12.0, + "eval_steps": 500, + "global_step": 3000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.1, + "grad_norm": 0.6611918807029724, + "learning_rate": 5.319148936170213e-05, + "loss": 0.6932, + "step": 25 + }, + { + "epoch": 0.2, + "grad_norm": 0.9467485547065735, + "learning_rate": 0.00010638297872340425, + "loss": 0.6912, + "step": 50 + }, + { + "epoch": 0.3, + "grad_norm": 0.9465051889419556, + "learning_rate": 0.00015957446808510637, + "loss": 0.6893, + "step": 75 + }, + { + "epoch": 0.4, + "grad_norm": 1.5430934429168701, + "learning_rate": 0.0002127659574468085, + "loss": 0.6804, + "step": 100 + }, + { + "epoch": 0.5, + "grad_norm": 1.6103107929229736, + "learning_rate": 0.00026595744680851064, + "loss": 0.6806, + "step": 125 + }, + { + "epoch": 0.6, + "grad_norm": 2.3773200511932373, + "learning_rate": 0.00031914893617021275, + "loss": 0.6801, + "step": 150 + }, + { + "epoch": 0.7, + "grad_norm": 1.8832203149795532, + "learning_rate": 0.0003723404255319149, + "loss": 0.6791, + "step": 175 + }, + { + "epoch": 0.8, + "grad_norm": 1.3350876569747925, + "learning_rate": 0.0003992081821181128, + "loss": 0.6746, + "step": 200 + }, + { + "epoch": 0.9, + "grad_norm": 1.5796219110488892, + "learning_rate": 0.0003975585615308479, + "loss": 0.6771, + "step": 225 + }, + { + "epoch": 1.0, + "grad_norm": 1.884006381034851, + "learning_rate": 0.00039590894094358297, + "loss": 0.6649, + "step": 250 + }, + { + "epoch": 1.1, + "grad_norm": 3.286440372467041, + "learning_rate": 0.0003942593203563181, + "loss": 0.6388, + "step": 275 + }, + { + "epoch": 1.2, + "grad_norm": 8.244946479797363, + "learning_rate": 0.0003926096997690532, + "loss": 0.6238, + "step": 300 + }, + { + "epoch": 1.3, + "grad_norm": 4.265683650970459, + "learning_rate": 0.00039096007918178817, + "loss": 0.6383, + "step": 325 + }, + { + "epoch": 1.4, + "grad_norm": 4.115826606750488, + "learning_rate": 0.0003893104585945233, + "loss": 0.6194, + "step": 350 + }, + { + "epoch": 1.5, + "grad_norm": 5.694250583648682, + "learning_rate": 0.0003876608380072583, + "loss": 0.6324, + "step": 375 + }, + { + "epoch": 1.6, + "grad_norm": 3.463121175765991, + "learning_rate": 0.00038601121741999343, + "loss": 0.621, + "step": 400 + }, + { + "epoch": 1.7, + "grad_norm": 4.582865238189697, + "learning_rate": 0.0003843615968327285, + "loss": 0.6116, + "step": 425 + }, + { + "epoch": 1.8, + "grad_norm": 11.996281623840332, + "learning_rate": 0.0003827119762454636, + "loss": 0.6393, + "step": 450 + }, + { + "epoch": 1.9, + "grad_norm": 3.0407373905181885, + "learning_rate": 0.00038106235565819863, + "loss": 0.628, + "step": 475 + }, + { + "epoch": 2.0, + "grad_norm": 2.917588233947754, + "learning_rate": 0.0003794127350709337, + "loss": 0.6078, + "step": 500 + }, + { + "epoch": 2.1, + "grad_norm": 4.748379707336426, + "learning_rate": 0.0003777631144836688, + "loss": 0.4899, + "step": 525 + }, + { + "epoch": 2.2, + "grad_norm": 3.8076977729797363, + "learning_rate": 0.00037611349389640383, + "loss": 0.5086, + "step": 550 + }, + { + "epoch": 2.3, + "grad_norm": 5.2440714836120605, + "learning_rate": 0.00037446387330913894, + "loss": 0.5327, + "step": 575 + }, + { + "epoch": 2.4, + "grad_norm": 7.110438346862793, + "learning_rate": 0.000372814252721874, + "loss": 0.5436, + "step": 600 + }, + { + "epoch": 2.5, + "grad_norm": 5.46150541305542, + "learning_rate": 0.00037116463213460903, + "loss": 0.5294, + "step": 625 + }, + { + "epoch": 2.6, + "grad_norm": 5.136163234710693, + "learning_rate": 0.00036951501154734414, + "loss": 0.5245, + "step": 650 + }, + { + "epoch": 2.7, + "grad_norm": 8.735346794128418, + "learning_rate": 0.0003678653909600792, + "loss": 0.5449, + "step": 675 + }, + { + "epoch": 2.8, + "grad_norm": 2.922825574874878, + "learning_rate": 0.0003662157703728143, + "loss": 0.5406, + "step": 700 + }, + { + "epoch": 2.9, + "grad_norm": 7.744819641113281, + "learning_rate": 0.00036456614978554934, + "loss": 0.5447, + "step": 725 + }, + { + "epoch": 3.0, + "grad_norm": 11.6185884475708, + "learning_rate": 0.00036291652919828444, + "loss": 0.5195, + "step": 750 + }, + { + "epoch": 3.1, + "grad_norm": 3.73836088180542, + "learning_rate": 0.00036126690861101944, + "loss": 0.3824, + "step": 775 + }, + { + "epoch": 3.2, + "grad_norm": 14.850343704223633, + "learning_rate": 0.00035961728802375454, + "loss": 0.4071, + "step": 800 + }, + { + "epoch": 3.3, + "grad_norm": 5.7157440185546875, + "learning_rate": 0.0003579676674364896, + "loss": 0.3986, + "step": 825 + }, + { + "epoch": 3.4, + "grad_norm": 12.418399810791016, + "learning_rate": 0.0003563180468492247, + "loss": 0.4282, + "step": 850 + }, + { + "epoch": 3.5, + "grad_norm": 12.793001174926758, + "learning_rate": 0.0003546684262619598, + "loss": 0.4822, + "step": 875 + }, + { + "epoch": 3.6, + "grad_norm": 6.489450931549072, + "learning_rate": 0.00035301880567469485, + "loss": 0.4239, + "step": 900 + }, + { + "epoch": 3.7, + "grad_norm": 5.365822792053223, + "learning_rate": 0.0003513691850874299, + "loss": 0.421, + "step": 925 + }, + { + "epoch": 3.8, + "grad_norm": 12.643745422363281, + "learning_rate": 0.00034971956450016495, + "loss": 0.3964, + "step": 950 + }, + { + "epoch": 3.9, + "grad_norm": 14.334024429321289, + "learning_rate": 0.00034806994391290005, + "loss": 0.4634, + "step": 975 + }, + { + "epoch": 4.0, + "grad_norm": 6.819091320037842, + "learning_rate": 0.0003464203233256351, + "loss": 0.4139, + "step": 1000 + }, + { + "epoch": 4.1, + "grad_norm": 8.29238224029541, + "learning_rate": 0.0003447707027383702, + "loss": 0.2695, + "step": 1025 + }, + { + "epoch": 4.2, + "grad_norm": 5.984206676483154, + "learning_rate": 0.00034312108215110525, + "loss": 0.2653, + "step": 1050 + }, + { + "epoch": 4.3, + "grad_norm": 5.9425435066223145, + "learning_rate": 0.0003414714615638403, + "loss": 0.2982, + "step": 1075 + }, + { + "epoch": 4.4, + "grad_norm": 7.1877593994140625, + "learning_rate": 0.0003398218409765754, + "loss": 0.3309, + "step": 1100 + }, + { + "epoch": 4.5, + "grad_norm": 17.83046531677246, + "learning_rate": 0.00033817222038931045, + "loss": 0.3467, + "step": 1125 + }, + { + "epoch": 4.6, + "grad_norm": 4.865128517150879, + "learning_rate": 0.00033652259980204556, + "loss": 0.3117, + "step": 1150 + }, + { + "epoch": 4.7, + "grad_norm": 31.427154541015625, + "learning_rate": 0.0003348729792147806, + "loss": 0.3273, + "step": 1175 + }, + { + "epoch": 4.8, + "grad_norm": 26.77984619140625, + "learning_rate": 0.0003332233586275157, + "loss": 0.3504, + "step": 1200 + }, + { + "epoch": 4.9, + "grad_norm": 9.389993667602539, + "learning_rate": 0.00033157373804025076, + "loss": 0.3339, + "step": 1225 + }, + { + "epoch": 5.0, + "grad_norm": 25.70441246032715, + "learning_rate": 0.0003299241174529858, + "loss": 0.3336, + "step": 1250 + }, + { + "epoch": 5.1, + "grad_norm": 13.482085227966309, + "learning_rate": 0.0003282744968657209, + "loss": 0.2156, + "step": 1275 + }, + { + "epoch": 5.2, + "grad_norm": 8.725810050964355, + "learning_rate": 0.00032662487627845596, + "loss": 0.1811, + "step": 1300 + }, + { + "epoch": 5.3, + "grad_norm": 9.313215255737305, + "learning_rate": 0.00032497525569119106, + "loss": 0.2807, + "step": 1325 + }, + { + "epoch": 5.4, + "grad_norm": 11.026411056518555, + "learning_rate": 0.0003233256351039261, + "loss": 0.2757, + "step": 1350 + }, + { + "epoch": 5.5, + "grad_norm": 11.038985252380371, + "learning_rate": 0.00032167601451666116, + "loss": 0.2177, + "step": 1375 + }, + { + "epoch": 5.6, + "grad_norm": 4.008651256561279, + "learning_rate": 0.00032002639392939627, + "loss": 0.2163, + "step": 1400 + }, + { + "epoch": 5.7, + "grad_norm": 12.480770111083984, + "learning_rate": 0.0003183767733421313, + "loss": 0.2173, + "step": 1425 + }, + { + "epoch": 5.8, + "grad_norm": 8.751969337463379, + "learning_rate": 0.0003167271527548664, + "loss": 0.2299, + "step": 1450 + }, + { + "epoch": 5.9, + "grad_norm": 7.701971530914307, + "learning_rate": 0.00031507753216760147, + "loss": 0.1949, + "step": 1475 + }, + { + "epoch": 6.0, + "grad_norm": 8.48027515411377, + "learning_rate": 0.00031342791158033657, + "loss": 0.2599, + "step": 1500 + }, + { + "epoch": 6.1, + "grad_norm": 9.29404067993164, + "learning_rate": 0.00031177829099307157, + "loss": 0.1376, + "step": 1525 + }, + { + "epoch": 6.2, + "grad_norm": 20.137714385986328, + "learning_rate": 0.00031012867040580667, + "loss": 0.1647, + "step": 1550 + }, + { + "epoch": 6.3, + "grad_norm": 11.394575119018555, + "learning_rate": 0.0003084790498185417, + "loss": 0.1565, + "step": 1575 + }, + { + "epoch": 6.4, + "grad_norm": 8.214287757873535, + "learning_rate": 0.0003068294292312768, + "loss": 0.1739, + "step": 1600 + }, + { + "epoch": 6.5, + "grad_norm": 7.779988765716553, + "learning_rate": 0.0003051798086440119, + "loss": 0.1403, + "step": 1625 + }, + { + "epoch": 6.6, + "grad_norm": 9.421648025512695, + "learning_rate": 0.000303530188056747, + "loss": 0.1545, + "step": 1650 + }, + { + "epoch": 6.7, + "grad_norm": 5.751734256744385, + "learning_rate": 0.000301880567469482, + "loss": 0.1971, + "step": 1675 + }, + { + "epoch": 6.8, + "grad_norm": 23.861705780029297, + "learning_rate": 0.0003002309468822171, + "loss": 0.1681, + "step": 1700 + }, + { + "epoch": 6.9, + "grad_norm": 18.944721221923828, + "learning_rate": 0.0002985813262949522, + "loss": 0.1703, + "step": 1725 + }, + { + "epoch": 7.0, + "grad_norm": 14.045795440673828, + "learning_rate": 0.00029693170570768723, + "loss": 0.1801, + "step": 1750 + }, + { + "epoch": 7.1, + "grad_norm": 6.4620137214660645, + "learning_rate": 0.00029528208512042233, + "loss": 0.1253, + "step": 1775 + }, + { + "epoch": 7.2, + "grad_norm": 4.318169593811035, + "learning_rate": 0.0002936324645331574, + "loss": 0.1397, + "step": 1800 + }, + { + "epoch": 7.3, + "grad_norm": 24.91462516784668, + "learning_rate": 0.00029198284394589243, + "loss": 0.1259, + "step": 1825 + }, + { + "epoch": 7.4, + "grad_norm": 23.614572525024414, + "learning_rate": 0.00029033322335862753, + "loss": 0.1293, + "step": 1850 + }, + { + "epoch": 7.5, + "grad_norm": 3.60048508644104, + "learning_rate": 0.0002886836027713626, + "loss": 0.138, + "step": 1875 + }, + { + "epoch": 7.6, + "grad_norm": 16.62705421447754, + "learning_rate": 0.0002870339821840977, + "loss": 0.1733, + "step": 1900 + }, + { + "epoch": 7.7, + "grad_norm": 33.79671859741211, + "learning_rate": 0.00028538436159683273, + "loss": 0.217, + "step": 1925 + }, + { + "epoch": 7.8, + "grad_norm": 9.69206428527832, + "learning_rate": 0.00028373474100956784, + "loss": 0.1289, + "step": 1950 + }, + { + "epoch": 7.9, + "grad_norm": 18.655046463012695, + "learning_rate": 0.0002820851204223029, + "loss": 0.1168, + "step": 1975 + }, + { + "epoch": 8.0, + "grad_norm": 8.4110746383667, + "learning_rate": 0.00028043549983503794, + "loss": 0.1471, + "step": 2000 + }, + { + "epoch": 8.1, + "grad_norm": 14.473428726196289, + "learning_rate": 0.00027878587924777304, + "loss": 0.0774, + "step": 2025 + }, + { + "epoch": 8.2, + "grad_norm": 1.2464979887008667, + "learning_rate": 0.0002771362586605081, + "loss": 0.0715, + "step": 2050 + }, + { + "epoch": 8.3, + "grad_norm": 26.74981689453125, + "learning_rate": 0.0002754866380732432, + "loss": 0.1289, + "step": 2075 + }, + { + "epoch": 8.4, + "grad_norm": 13.958703994750977, + "learning_rate": 0.00027383701748597824, + "loss": 0.1294, + "step": 2100 + }, + { + "epoch": 8.5, + "grad_norm": 10.394835472106934, + "learning_rate": 0.0002721873968987133, + "loss": 0.0987, + "step": 2125 + }, + { + "epoch": 8.6, + "grad_norm": 9.387774467468262, + "learning_rate": 0.00027053777631144834, + "loss": 0.1269, + "step": 2150 + }, + { + "epoch": 8.7, + "grad_norm": 4.988718509674072, + "learning_rate": 0.00026888815572418344, + "loss": 0.1374, + "step": 2175 + }, + { + "epoch": 8.8, + "grad_norm": 12.66832447052002, + "learning_rate": 0.00026723853513691855, + "loss": 0.0997, + "step": 2200 + }, + { + "epoch": 8.9, + "grad_norm": 19.10486602783203, + "learning_rate": 0.0002655889145496536, + "loss": 0.1325, + "step": 2225 + }, + { + "epoch": 9.0, + "grad_norm": 3.2943971157073975, + "learning_rate": 0.0002639392939623887, + "loss": 0.1385, + "step": 2250 + }, + { + "epoch": 9.1, + "grad_norm": 11.490874290466309, + "learning_rate": 0.0002622896733751237, + "loss": 0.0573, + "step": 2275 + }, + { + "epoch": 9.2, + "grad_norm": 11.961288452148438, + "learning_rate": 0.0002606400527878588, + "loss": 0.1128, + "step": 2300 + }, + { + "epoch": 9.3, + "grad_norm": 0.7256277203559875, + "learning_rate": 0.00025899043220059385, + "loss": 0.0673, + "step": 2325 + }, + { + "epoch": 9.4, + "grad_norm": 9.857927322387695, + "learning_rate": 0.00025734081161332895, + "loss": 0.2072, + "step": 2350 + }, + { + "epoch": 9.5, + "grad_norm": 34.927734375, + "learning_rate": 0.000255691191026064, + "loss": 0.221, + "step": 2375 + }, + { + "epoch": 9.6, + "grad_norm": 28.461849212646484, + "learning_rate": 0.0002540415704387991, + "loss": 0.1263, + "step": 2400 + }, + { + "epoch": 9.7, + "grad_norm": 9.591217041015625, + "learning_rate": 0.00025239194985153415, + "loss": 0.099, + "step": 2425 + }, + { + "epoch": 9.8, + "grad_norm": 9.780253410339355, + "learning_rate": 0.0002507423292642692, + "loss": 0.0754, + "step": 2450 + }, + { + "epoch": 9.9, + "grad_norm": 14.394485473632812, + "learning_rate": 0.0002490927086770043, + "loss": 0.0909, + "step": 2475 + }, + { + "epoch": 10.0, + "grad_norm": 4.217277526855469, + "learning_rate": 0.00024744308808973936, + "loss": 0.1495, + "step": 2500 + }, + { + "epoch": 10.1, + "grad_norm": 0.08438724279403687, + "learning_rate": 0.00024579346750247446, + "loss": 0.0662, + "step": 2525 + }, + { + "epoch": 10.2, + "grad_norm": 0.9595862030982971, + "learning_rate": 0.0002441438469152095, + "loss": 0.084, + "step": 2550 + }, + { + "epoch": 10.3, + "grad_norm": 42.32374954223633, + "learning_rate": 0.00024249422632794456, + "loss": 0.0541, + "step": 2575 + }, + { + "epoch": 10.4, + "grad_norm": 24.792821884155273, + "learning_rate": 0.00024084460574067963, + "loss": 0.121, + "step": 2600 + }, + { + "epoch": 10.5, + "grad_norm": 2.7049331665039062, + "learning_rate": 0.0002391949851534147, + "loss": 0.1551, + "step": 2625 + }, + { + "epoch": 10.6, + "grad_norm": 2.788362741470337, + "learning_rate": 0.0002375453645661498, + "loss": 0.0877, + "step": 2650 + }, + { + "epoch": 10.7, + "grad_norm": 0.4999215006828308, + "learning_rate": 0.00023589574397888486, + "loss": 0.1263, + "step": 2675 + }, + { + "epoch": 10.8, + "grad_norm": 16.476675033569336, + "learning_rate": 0.00023424612339161997, + "loss": 0.0883, + "step": 2700 + }, + { + "epoch": 10.9, + "grad_norm": 38.192726135253906, + "learning_rate": 0.000232596502804355, + "loss": 0.0843, + "step": 2725 + }, + { + "epoch": 11.0, + "grad_norm": 18.09952735900879, + "learning_rate": 0.00023094688221709007, + "loss": 0.0957, + "step": 2750 + }, + { + "epoch": 11.1, + "grad_norm": 34.254417419433594, + "learning_rate": 0.00022929726162982514, + "loss": 0.0856, + "step": 2775 + }, + { + "epoch": 11.2, + "grad_norm": 11.744216918945312, + "learning_rate": 0.00022764764104256022, + "loss": 0.0654, + "step": 2800 + }, + { + "epoch": 11.3, + "grad_norm": 0.1285558044910431, + "learning_rate": 0.0002259980204552953, + "loss": 0.0638, + "step": 2825 + }, + { + "epoch": 11.4, + "grad_norm": 20.57583999633789, + "learning_rate": 0.00022434839986803037, + "loss": 0.079, + "step": 2850 + }, + { + "epoch": 11.5, + "grad_norm": 0.7192332148551941, + "learning_rate": 0.00022269877928076542, + "loss": 0.0854, + "step": 2875 + }, + { + "epoch": 11.6, + "grad_norm": 31.341829299926758, + "learning_rate": 0.0002210491586935005, + "loss": 0.063, + "step": 2900 + }, + { + "epoch": 11.7, + "grad_norm": 13.460247039794922, + "learning_rate": 0.00021939953810623557, + "loss": 0.1358, + "step": 2925 + }, + { + "epoch": 11.8, + "grad_norm": 1.671036720275879, + "learning_rate": 0.00021774991751897065, + "loss": 0.0806, + "step": 2950 + }, + { + "epoch": 11.9, + "grad_norm": 22.026491165161133, + "learning_rate": 0.00021610029693170573, + "loss": 0.0912, + "step": 2975 + }, + { + "epoch": 12.0, + "grad_norm": 46.78192138671875, + "learning_rate": 0.0002144506763444408, + "loss": 0.0763, + "step": 3000 + } + ], + "logging_steps": 25, + "max_steps": 6250, + "num_input_tokens_seen": 0, + "num_train_epochs": 25, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 152458444800000.0, + "train_batch_size": 20, + "trial_name": null, + "trial_params": null +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-3000/training_args.bin b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-3000/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..798d8a9f3a9513692035320aba5484c89ac49e69 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-3000/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1cabf5db3e9d79ec0c1945933395ff57418204414cc81f77c90324d003fac782 +size 5304 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-3000/vocab.txt b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-3000/vocab.txt new file mode 100644 index 0000000000000000000000000000000000000000..cc9ad53732e21ea9180f18e1fb67a38cc5702d4c --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-3000/vocab.txt @@ -0,0 +1,4101 @@ +[PAD] +[UNK] +[CLS] +[SEP] +[MASK] +AAAAAA +AAAAAT +AAAAAC +AAAAAG +AAAATA +AAAATT +AAAATC +AAAATG +AAAACA +AAAACT +AAAACC +AAAACG +AAAAGA +AAAAGT +AAAAGC +AAAAGG +AAATAA +AAATAT +AAATAC +AAATAG +AAATTA +AAATTT +AAATTC +AAATTG +AAATCA +AAATCT +AAATCC +AAATCG +AAATGA +AAATGT +AAATGC +AAATGG +AAACAA +AAACAT +AAACAC +AAACAG +AAACTA +AAACTT +AAACTC +AAACTG +AAACCA +AAACCT +AAACCC +AAACCG +AAACGA +AAACGT +AAACGC +AAACGG +AAAGAA +AAAGAT +AAAGAC +AAAGAG +AAAGTA +AAAGTT +AAAGTC +AAAGTG +AAAGCA +AAAGCT +AAAGCC +AAAGCG +AAAGGA +AAAGGT +AAAGGC +AAAGGG +AATAAA +AATAAT +AATAAC +AATAAG +AATATA +AATATT +AATATC +AATATG +AATACA +AATACT +AATACC +AATACG +AATAGA +AATAGT +AATAGC +AATAGG +AATTAA +AATTAT +AATTAC +AATTAG +AATTTA +AATTTT +AATTTC +AATTTG +AATTCA +AATTCT +AATTCC +AATTCG +AATTGA +AATTGT +AATTGC +AATTGG +AATCAA +AATCAT +AATCAC +AATCAG +AATCTA +AATCTT +AATCTC +AATCTG +AATCCA +AATCCT +AATCCC +AATCCG +AATCGA +AATCGT +AATCGC +AATCGG +AATGAA +AATGAT +AATGAC +AATGAG +AATGTA +AATGTT +AATGTC +AATGTG +AATGCA +AATGCT +AATGCC +AATGCG +AATGGA +AATGGT +AATGGC +AATGGG +AACAAA +AACAAT +AACAAC +AACAAG +AACATA +AACATT +AACATC +AACATG +AACACA +AACACT +AACACC +AACACG +AACAGA +AACAGT +AACAGC +AACAGG +AACTAA +AACTAT +AACTAC +AACTAG +AACTTA +AACTTT +AACTTC +AACTTG +AACTCA +AACTCT +AACTCC +AACTCG +AACTGA +AACTGT +AACTGC +AACTGG +AACCAA +AACCAT +AACCAC +AACCAG +AACCTA +AACCTT +AACCTC +AACCTG +AACCCA +AACCCT +AACCCC +AACCCG +AACCGA +AACCGT +AACCGC +AACCGG +AACGAA +AACGAT +AACGAC +AACGAG +AACGTA +AACGTT +AACGTC +AACGTG +AACGCA +AACGCT +AACGCC +AACGCG +AACGGA +AACGGT +AACGGC +AACGGG +AAGAAA +AAGAAT +AAGAAC +AAGAAG +AAGATA +AAGATT +AAGATC +AAGATG +AAGACA +AAGACT +AAGACC +AAGACG +AAGAGA +AAGAGT +AAGAGC +AAGAGG +AAGTAA +AAGTAT +AAGTAC +AAGTAG +AAGTTA +AAGTTT +AAGTTC +AAGTTG +AAGTCA +AAGTCT +AAGTCC +AAGTCG +AAGTGA +AAGTGT +AAGTGC +AAGTGG +AAGCAA +AAGCAT +AAGCAC +AAGCAG +AAGCTA +AAGCTT +AAGCTC +AAGCTG +AAGCCA +AAGCCT +AAGCCC +AAGCCG +AAGCGA +AAGCGT +AAGCGC +AAGCGG +AAGGAA +AAGGAT +AAGGAC +AAGGAG +AAGGTA +AAGGTT +AAGGTC +AAGGTG +AAGGCA +AAGGCT +AAGGCC +AAGGCG +AAGGGA +AAGGGT +AAGGGC +AAGGGG +ATAAAA +ATAAAT +ATAAAC +ATAAAG +ATAATA +ATAATT +ATAATC +ATAATG +ATAACA +ATAACT +ATAACC +ATAACG +ATAAGA +ATAAGT +ATAAGC +ATAAGG +ATATAA +ATATAT +ATATAC +ATATAG +ATATTA +ATATTT +ATATTC +ATATTG +ATATCA +ATATCT +ATATCC +ATATCG +ATATGA +ATATGT +ATATGC +ATATGG +ATACAA +ATACAT +ATACAC +ATACAG +ATACTA +ATACTT +ATACTC +ATACTG +ATACCA +ATACCT +ATACCC +ATACCG +ATACGA +ATACGT +ATACGC +ATACGG +ATAGAA +ATAGAT +ATAGAC +ATAGAG +ATAGTA +ATAGTT +ATAGTC +ATAGTG +ATAGCA +ATAGCT +ATAGCC +ATAGCG +ATAGGA +ATAGGT +ATAGGC +ATAGGG +ATTAAA +ATTAAT +ATTAAC +ATTAAG +ATTATA +ATTATT +ATTATC +ATTATG +ATTACA +ATTACT +ATTACC +ATTACG +ATTAGA +ATTAGT +ATTAGC +ATTAGG +ATTTAA +ATTTAT +ATTTAC +ATTTAG +ATTTTA +ATTTTT +ATTTTC +ATTTTG +ATTTCA +ATTTCT +ATTTCC +ATTTCG +ATTTGA +ATTTGT +ATTTGC +ATTTGG +ATTCAA +ATTCAT +ATTCAC +ATTCAG +ATTCTA +ATTCTT +ATTCTC +ATTCTG +ATTCCA +ATTCCT +ATTCCC +ATTCCG +ATTCGA +ATTCGT +ATTCGC +ATTCGG +ATTGAA +ATTGAT +ATTGAC +ATTGAG +ATTGTA +ATTGTT +ATTGTC +ATTGTG +ATTGCA +ATTGCT +ATTGCC +ATTGCG +ATTGGA +ATTGGT +ATTGGC +ATTGGG +ATCAAA +ATCAAT +ATCAAC +ATCAAG +ATCATA +ATCATT +ATCATC +ATCATG +ATCACA +ATCACT +ATCACC +ATCACG +ATCAGA +ATCAGT +ATCAGC +ATCAGG +ATCTAA +ATCTAT +ATCTAC +ATCTAG +ATCTTA +ATCTTT +ATCTTC +ATCTTG +ATCTCA +ATCTCT +ATCTCC +ATCTCG +ATCTGA +ATCTGT +ATCTGC +ATCTGG +ATCCAA +ATCCAT +ATCCAC +ATCCAG +ATCCTA +ATCCTT +ATCCTC +ATCCTG +ATCCCA +ATCCCT +ATCCCC +ATCCCG +ATCCGA +ATCCGT +ATCCGC +ATCCGG +ATCGAA +ATCGAT +ATCGAC +ATCGAG +ATCGTA +ATCGTT +ATCGTC +ATCGTG +ATCGCA +ATCGCT +ATCGCC +ATCGCG +ATCGGA +ATCGGT +ATCGGC +ATCGGG +ATGAAA +ATGAAT +ATGAAC +ATGAAG +ATGATA +ATGATT +ATGATC +ATGATG +ATGACA +ATGACT +ATGACC +ATGACG +ATGAGA +ATGAGT +ATGAGC +ATGAGG +ATGTAA +ATGTAT +ATGTAC +ATGTAG +ATGTTA +ATGTTT +ATGTTC +ATGTTG +ATGTCA +ATGTCT +ATGTCC +ATGTCG +ATGTGA +ATGTGT +ATGTGC +ATGTGG +ATGCAA +ATGCAT +ATGCAC +ATGCAG +ATGCTA +ATGCTT +ATGCTC +ATGCTG +ATGCCA +ATGCCT +ATGCCC +ATGCCG +ATGCGA +ATGCGT +ATGCGC +ATGCGG +ATGGAA +ATGGAT +ATGGAC +ATGGAG +ATGGTA +ATGGTT +ATGGTC +ATGGTG +ATGGCA +ATGGCT +ATGGCC +ATGGCG +ATGGGA +ATGGGT +ATGGGC +ATGGGG +ACAAAA +ACAAAT +ACAAAC +ACAAAG +ACAATA +ACAATT +ACAATC +ACAATG +ACAACA +ACAACT +ACAACC +ACAACG +ACAAGA +ACAAGT +ACAAGC +ACAAGG +ACATAA +ACATAT +ACATAC +ACATAG +ACATTA +ACATTT +ACATTC +ACATTG +ACATCA +ACATCT +ACATCC +ACATCG +ACATGA +ACATGT +ACATGC +ACATGG +ACACAA +ACACAT +ACACAC +ACACAG +ACACTA +ACACTT +ACACTC +ACACTG +ACACCA +ACACCT +ACACCC +ACACCG +ACACGA +ACACGT +ACACGC +ACACGG +ACAGAA +ACAGAT +ACAGAC +ACAGAG +ACAGTA +ACAGTT +ACAGTC +ACAGTG +ACAGCA +ACAGCT +ACAGCC +ACAGCG +ACAGGA +ACAGGT +ACAGGC +ACAGGG +ACTAAA +ACTAAT +ACTAAC +ACTAAG +ACTATA +ACTATT +ACTATC +ACTATG +ACTACA +ACTACT +ACTACC +ACTACG +ACTAGA +ACTAGT +ACTAGC +ACTAGG +ACTTAA +ACTTAT +ACTTAC +ACTTAG +ACTTTA +ACTTTT +ACTTTC +ACTTTG +ACTTCA +ACTTCT +ACTTCC +ACTTCG +ACTTGA +ACTTGT +ACTTGC +ACTTGG +ACTCAA +ACTCAT +ACTCAC +ACTCAG +ACTCTA +ACTCTT +ACTCTC +ACTCTG +ACTCCA +ACTCCT +ACTCCC +ACTCCG +ACTCGA +ACTCGT +ACTCGC +ACTCGG +ACTGAA +ACTGAT +ACTGAC +ACTGAG +ACTGTA +ACTGTT +ACTGTC +ACTGTG +ACTGCA +ACTGCT +ACTGCC +ACTGCG +ACTGGA +ACTGGT +ACTGGC +ACTGGG +ACCAAA +ACCAAT +ACCAAC +ACCAAG +ACCATA +ACCATT +ACCATC +ACCATG +ACCACA +ACCACT +ACCACC +ACCACG +ACCAGA +ACCAGT +ACCAGC +ACCAGG +ACCTAA +ACCTAT +ACCTAC +ACCTAG +ACCTTA +ACCTTT +ACCTTC +ACCTTG +ACCTCA +ACCTCT +ACCTCC +ACCTCG +ACCTGA +ACCTGT +ACCTGC +ACCTGG +ACCCAA +ACCCAT +ACCCAC +ACCCAG +ACCCTA +ACCCTT +ACCCTC +ACCCTG +ACCCCA +ACCCCT +ACCCCC +ACCCCG +ACCCGA +ACCCGT +ACCCGC +ACCCGG +ACCGAA +ACCGAT +ACCGAC +ACCGAG +ACCGTA +ACCGTT +ACCGTC +ACCGTG +ACCGCA +ACCGCT +ACCGCC +ACCGCG +ACCGGA +ACCGGT +ACCGGC +ACCGGG +ACGAAA +ACGAAT +ACGAAC +ACGAAG +ACGATA +ACGATT +ACGATC +ACGATG +ACGACA +ACGACT +ACGACC +ACGACG +ACGAGA +ACGAGT +ACGAGC +ACGAGG +ACGTAA +ACGTAT +ACGTAC +ACGTAG +ACGTTA +ACGTTT +ACGTTC +ACGTTG +ACGTCA +ACGTCT +ACGTCC +ACGTCG +ACGTGA +ACGTGT +ACGTGC +ACGTGG +ACGCAA +ACGCAT +ACGCAC +ACGCAG +ACGCTA +ACGCTT +ACGCTC +ACGCTG +ACGCCA +ACGCCT +ACGCCC +ACGCCG +ACGCGA +ACGCGT +ACGCGC +ACGCGG +ACGGAA +ACGGAT +ACGGAC +ACGGAG +ACGGTA +ACGGTT +ACGGTC +ACGGTG +ACGGCA +ACGGCT +ACGGCC +ACGGCG +ACGGGA +ACGGGT +ACGGGC +ACGGGG +AGAAAA +AGAAAT +AGAAAC +AGAAAG +AGAATA +AGAATT +AGAATC +AGAATG +AGAACA +AGAACT +AGAACC +AGAACG +AGAAGA +AGAAGT +AGAAGC +AGAAGG +AGATAA +AGATAT +AGATAC +AGATAG +AGATTA +AGATTT +AGATTC +AGATTG +AGATCA +AGATCT +AGATCC +AGATCG +AGATGA +AGATGT +AGATGC +AGATGG +AGACAA +AGACAT +AGACAC +AGACAG +AGACTA +AGACTT +AGACTC +AGACTG +AGACCA +AGACCT +AGACCC +AGACCG +AGACGA +AGACGT +AGACGC +AGACGG +AGAGAA +AGAGAT +AGAGAC +AGAGAG +AGAGTA +AGAGTT +AGAGTC +AGAGTG +AGAGCA +AGAGCT +AGAGCC +AGAGCG +AGAGGA +AGAGGT +AGAGGC +AGAGGG +AGTAAA +AGTAAT +AGTAAC +AGTAAG +AGTATA +AGTATT +AGTATC +AGTATG +AGTACA +AGTACT +AGTACC +AGTACG +AGTAGA +AGTAGT +AGTAGC +AGTAGG +AGTTAA +AGTTAT +AGTTAC +AGTTAG +AGTTTA +AGTTTT +AGTTTC +AGTTTG +AGTTCA +AGTTCT +AGTTCC +AGTTCG +AGTTGA +AGTTGT +AGTTGC +AGTTGG +AGTCAA +AGTCAT +AGTCAC +AGTCAG +AGTCTA +AGTCTT +AGTCTC +AGTCTG +AGTCCA +AGTCCT +AGTCCC +AGTCCG +AGTCGA +AGTCGT +AGTCGC +AGTCGG +AGTGAA +AGTGAT +AGTGAC +AGTGAG +AGTGTA +AGTGTT +AGTGTC +AGTGTG +AGTGCA +AGTGCT +AGTGCC +AGTGCG +AGTGGA +AGTGGT +AGTGGC +AGTGGG +AGCAAA +AGCAAT +AGCAAC +AGCAAG +AGCATA +AGCATT +AGCATC +AGCATG +AGCACA +AGCACT +AGCACC +AGCACG +AGCAGA +AGCAGT +AGCAGC +AGCAGG +AGCTAA +AGCTAT +AGCTAC +AGCTAG +AGCTTA +AGCTTT +AGCTTC +AGCTTG +AGCTCA +AGCTCT +AGCTCC +AGCTCG +AGCTGA +AGCTGT +AGCTGC +AGCTGG +AGCCAA +AGCCAT +AGCCAC +AGCCAG +AGCCTA +AGCCTT +AGCCTC +AGCCTG +AGCCCA +AGCCCT +AGCCCC +AGCCCG +AGCCGA +AGCCGT +AGCCGC +AGCCGG +AGCGAA +AGCGAT +AGCGAC +AGCGAG +AGCGTA +AGCGTT +AGCGTC +AGCGTG +AGCGCA +AGCGCT +AGCGCC +AGCGCG +AGCGGA +AGCGGT +AGCGGC +AGCGGG +AGGAAA +AGGAAT +AGGAAC +AGGAAG +AGGATA +AGGATT +AGGATC +AGGATG +AGGACA +AGGACT +AGGACC +AGGACG +AGGAGA +AGGAGT +AGGAGC +AGGAGG +AGGTAA +AGGTAT +AGGTAC +AGGTAG +AGGTTA +AGGTTT +AGGTTC +AGGTTG +AGGTCA +AGGTCT +AGGTCC +AGGTCG +AGGTGA +AGGTGT +AGGTGC +AGGTGG +AGGCAA +AGGCAT +AGGCAC +AGGCAG +AGGCTA +AGGCTT +AGGCTC +AGGCTG +AGGCCA +AGGCCT +AGGCCC +AGGCCG +AGGCGA +AGGCGT +AGGCGC +AGGCGG +AGGGAA +AGGGAT +AGGGAC +AGGGAG +AGGGTA +AGGGTT +AGGGTC +AGGGTG +AGGGCA +AGGGCT +AGGGCC +AGGGCG +AGGGGA +AGGGGT +AGGGGC +AGGGGG +TAAAAA +TAAAAT +TAAAAC +TAAAAG +TAAATA +TAAATT +TAAATC +TAAATG +TAAACA +TAAACT +TAAACC +TAAACG +TAAAGA +TAAAGT +TAAAGC +TAAAGG +TAATAA +TAATAT +TAATAC +TAATAG +TAATTA +TAATTT +TAATTC +TAATTG +TAATCA +TAATCT +TAATCC +TAATCG +TAATGA +TAATGT +TAATGC +TAATGG +TAACAA +TAACAT +TAACAC +TAACAG +TAACTA +TAACTT +TAACTC +TAACTG +TAACCA +TAACCT +TAACCC +TAACCG +TAACGA +TAACGT +TAACGC +TAACGG +TAAGAA +TAAGAT +TAAGAC +TAAGAG +TAAGTA +TAAGTT +TAAGTC +TAAGTG +TAAGCA +TAAGCT +TAAGCC +TAAGCG +TAAGGA +TAAGGT +TAAGGC +TAAGGG +TATAAA +TATAAT +TATAAC +TATAAG +TATATA +TATATT +TATATC +TATATG +TATACA +TATACT +TATACC +TATACG +TATAGA +TATAGT +TATAGC +TATAGG +TATTAA +TATTAT +TATTAC +TATTAG +TATTTA +TATTTT +TATTTC +TATTTG +TATTCA +TATTCT +TATTCC +TATTCG +TATTGA +TATTGT +TATTGC +TATTGG +TATCAA +TATCAT +TATCAC +TATCAG +TATCTA +TATCTT +TATCTC +TATCTG +TATCCA +TATCCT +TATCCC +TATCCG +TATCGA +TATCGT +TATCGC +TATCGG +TATGAA +TATGAT +TATGAC +TATGAG +TATGTA +TATGTT +TATGTC +TATGTG +TATGCA +TATGCT +TATGCC +TATGCG +TATGGA +TATGGT +TATGGC +TATGGG +TACAAA +TACAAT +TACAAC +TACAAG +TACATA +TACATT +TACATC +TACATG +TACACA +TACACT +TACACC +TACACG +TACAGA +TACAGT +TACAGC +TACAGG +TACTAA +TACTAT +TACTAC +TACTAG +TACTTA +TACTTT +TACTTC +TACTTG +TACTCA +TACTCT +TACTCC +TACTCG +TACTGA +TACTGT +TACTGC +TACTGG +TACCAA +TACCAT +TACCAC +TACCAG +TACCTA +TACCTT +TACCTC +TACCTG +TACCCA +TACCCT +TACCCC +TACCCG +TACCGA +TACCGT +TACCGC +TACCGG +TACGAA +TACGAT +TACGAC +TACGAG +TACGTA +TACGTT +TACGTC +TACGTG +TACGCA +TACGCT +TACGCC +TACGCG +TACGGA +TACGGT +TACGGC +TACGGG +TAGAAA +TAGAAT +TAGAAC +TAGAAG +TAGATA +TAGATT +TAGATC +TAGATG +TAGACA +TAGACT +TAGACC +TAGACG +TAGAGA +TAGAGT +TAGAGC +TAGAGG +TAGTAA +TAGTAT +TAGTAC +TAGTAG +TAGTTA +TAGTTT +TAGTTC +TAGTTG +TAGTCA +TAGTCT +TAGTCC +TAGTCG +TAGTGA +TAGTGT +TAGTGC +TAGTGG +TAGCAA +TAGCAT +TAGCAC +TAGCAG +TAGCTA +TAGCTT +TAGCTC +TAGCTG +TAGCCA +TAGCCT +TAGCCC +TAGCCG +TAGCGA +TAGCGT +TAGCGC +TAGCGG +TAGGAA +TAGGAT +TAGGAC +TAGGAG +TAGGTA +TAGGTT +TAGGTC +TAGGTG +TAGGCA +TAGGCT +TAGGCC +TAGGCG +TAGGGA +TAGGGT +TAGGGC +TAGGGG +TTAAAA +TTAAAT +TTAAAC +TTAAAG +TTAATA +TTAATT +TTAATC +TTAATG +TTAACA +TTAACT +TTAACC +TTAACG +TTAAGA +TTAAGT +TTAAGC +TTAAGG +TTATAA +TTATAT +TTATAC +TTATAG +TTATTA +TTATTT +TTATTC +TTATTG +TTATCA +TTATCT +TTATCC +TTATCG +TTATGA +TTATGT +TTATGC +TTATGG +TTACAA +TTACAT +TTACAC +TTACAG +TTACTA +TTACTT +TTACTC +TTACTG +TTACCA +TTACCT +TTACCC +TTACCG +TTACGA +TTACGT +TTACGC +TTACGG +TTAGAA +TTAGAT +TTAGAC +TTAGAG +TTAGTA +TTAGTT +TTAGTC +TTAGTG +TTAGCA +TTAGCT +TTAGCC +TTAGCG +TTAGGA +TTAGGT +TTAGGC +TTAGGG +TTTAAA +TTTAAT +TTTAAC +TTTAAG +TTTATA +TTTATT +TTTATC +TTTATG +TTTACA +TTTACT +TTTACC +TTTACG +TTTAGA +TTTAGT +TTTAGC +TTTAGG +TTTTAA +TTTTAT +TTTTAC +TTTTAG +TTTTTA +TTTTTT +TTTTTC +TTTTTG +TTTTCA +TTTTCT +TTTTCC +TTTTCG +TTTTGA +TTTTGT +TTTTGC +TTTTGG +TTTCAA +TTTCAT +TTTCAC +TTTCAG +TTTCTA +TTTCTT +TTTCTC +TTTCTG +TTTCCA +TTTCCT +TTTCCC +TTTCCG +TTTCGA +TTTCGT +TTTCGC +TTTCGG +TTTGAA +TTTGAT +TTTGAC +TTTGAG +TTTGTA +TTTGTT +TTTGTC +TTTGTG +TTTGCA +TTTGCT +TTTGCC +TTTGCG +TTTGGA +TTTGGT +TTTGGC +TTTGGG +TTCAAA +TTCAAT +TTCAAC +TTCAAG +TTCATA +TTCATT +TTCATC +TTCATG +TTCACA +TTCACT +TTCACC +TTCACG +TTCAGA +TTCAGT +TTCAGC +TTCAGG +TTCTAA +TTCTAT +TTCTAC +TTCTAG +TTCTTA +TTCTTT +TTCTTC +TTCTTG +TTCTCA +TTCTCT +TTCTCC +TTCTCG +TTCTGA +TTCTGT +TTCTGC +TTCTGG +TTCCAA +TTCCAT +TTCCAC +TTCCAG +TTCCTA +TTCCTT +TTCCTC +TTCCTG +TTCCCA +TTCCCT +TTCCCC +TTCCCG +TTCCGA +TTCCGT +TTCCGC +TTCCGG +TTCGAA +TTCGAT +TTCGAC +TTCGAG +TTCGTA +TTCGTT +TTCGTC +TTCGTG +TTCGCA +TTCGCT +TTCGCC +TTCGCG +TTCGGA +TTCGGT +TTCGGC +TTCGGG +TTGAAA +TTGAAT +TTGAAC +TTGAAG +TTGATA +TTGATT +TTGATC +TTGATG +TTGACA +TTGACT +TTGACC +TTGACG +TTGAGA +TTGAGT +TTGAGC +TTGAGG +TTGTAA +TTGTAT +TTGTAC +TTGTAG +TTGTTA +TTGTTT +TTGTTC +TTGTTG +TTGTCA +TTGTCT +TTGTCC +TTGTCG +TTGTGA +TTGTGT +TTGTGC +TTGTGG +TTGCAA +TTGCAT +TTGCAC +TTGCAG +TTGCTA +TTGCTT +TTGCTC +TTGCTG +TTGCCA +TTGCCT +TTGCCC +TTGCCG +TTGCGA +TTGCGT +TTGCGC +TTGCGG +TTGGAA +TTGGAT +TTGGAC +TTGGAG +TTGGTA +TTGGTT +TTGGTC +TTGGTG +TTGGCA +TTGGCT +TTGGCC +TTGGCG +TTGGGA +TTGGGT +TTGGGC +TTGGGG +TCAAAA +TCAAAT +TCAAAC +TCAAAG +TCAATA +TCAATT +TCAATC +TCAATG +TCAACA +TCAACT +TCAACC +TCAACG +TCAAGA +TCAAGT +TCAAGC +TCAAGG +TCATAA +TCATAT +TCATAC +TCATAG +TCATTA +TCATTT +TCATTC +TCATTG +TCATCA +TCATCT +TCATCC +TCATCG +TCATGA +TCATGT +TCATGC +TCATGG +TCACAA +TCACAT +TCACAC +TCACAG +TCACTA +TCACTT +TCACTC +TCACTG +TCACCA +TCACCT +TCACCC +TCACCG +TCACGA +TCACGT +TCACGC +TCACGG +TCAGAA +TCAGAT +TCAGAC +TCAGAG +TCAGTA +TCAGTT +TCAGTC +TCAGTG +TCAGCA +TCAGCT +TCAGCC +TCAGCG +TCAGGA +TCAGGT +TCAGGC +TCAGGG +TCTAAA +TCTAAT +TCTAAC +TCTAAG +TCTATA +TCTATT +TCTATC +TCTATG +TCTACA +TCTACT +TCTACC +TCTACG +TCTAGA +TCTAGT +TCTAGC +TCTAGG +TCTTAA +TCTTAT +TCTTAC +TCTTAG +TCTTTA +TCTTTT +TCTTTC +TCTTTG +TCTTCA +TCTTCT +TCTTCC +TCTTCG +TCTTGA +TCTTGT +TCTTGC +TCTTGG +TCTCAA +TCTCAT +TCTCAC +TCTCAG +TCTCTA +TCTCTT +TCTCTC +TCTCTG +TCTCCA +TCTCCT +TCTCCC +TCTCCG +TCTCGA +TCTCGT +TCTCGC +TCTCGG +TCTGAA +TCTGAT +TCTGAC +TCTGAG +TCTGTA +TCTGTT +TCTGTC +TCTGTG +TCTGCA +TCTGCT +TCTGCC +TCTGCG +TCTGGA +TCTGGT +TCTGGC +TCTGGG +TCCAAA +TCCAAT +TCCAAC +TCCAAG +TCCATA +TCCATT +TCCATC +TCCATG +TCCACA +TCCACT +TCCACC +TCCACG +TCCAGA +TCCAGT +TCCAGC +TCCAGG +TCCTAA +TCCTAT +TCCTAC +TCCTAG +TCCTTA +TCCTTT +TCCTTC +TCCTTG +TCCTCA +TCCTCT +TCCTCC +TCCTCG +TCCTGA +TCCTGT +TCCTGC +TCCTGG +TCCCAA +TCCCAT +TCCCAC +TCCCAG +TCCCTA +TCCCTT +TCCCTC +TCCCTG +TCCCCA +TCCCCT +TCCCCC +TCCCCG +TCCCGA +TCCCGT +TCCCGC +TCCCGG +TCCGAA +TCCGAT +TCCGAC +TCCGAG +TCCGTA +TCCGTT +TCCGTC +TCCGTG +TCCGCA +TCCGCT +TCCGCC +TCCGCG +TCCGGA +TCCGGT +TCCGGC +TCCGGG +TCGAAA +TCGAAT +TCGAAC +TCGAAG +TCGATA +TCGATT +TCGATC +TCGATG +TCGACA +TCGACT +TCGACC +TCGACG +TCGAGA +TCGAGT +TCGAGC +TCGAGG +TCGTAA +TCGTAT +TCGTAC +TCGTAG +TCGTTA +TCGTTT +TCGTTC +TCGTTG +TCGTCA +TCGTCT +TCGTCC +TCGTCG +TCGTGA +TCGTGT +TCGTGC +TCGTGG +TCGCAA +TCGCAT +TCGCAC +TCGCAG +TCGCTA +TCGCTT +TCGCTC +TCGCTG +TCGCCA +TCGCCT +TCGCCC +TCGCCG +TCGCGA +TCGCGT +TCGCGC +TCGCGG +TCGGAA +TCGGAT +TCGGAC +TCGGAG +TCGGTA +TCGGTT +TCGGTC +TCGGTG +TCGGCA +TCGGCT +TCGGCC +TCGGCG +TCGGGA +TCGGGT +TCGGGC +TCGGGG +TGAAAA +TGAAAT +TGAAAC +TGAAAG +TGAATA +TGAATT +TGAATC +TGAATG +TGAACA +TGAACT +TGAACC +TGAACG +TGAAGA +TGAAGT +TGAAGC +TGAAGG +TGATAA +TGATAT +TGATAC +TGATAG +TGATTA +TGATTT +TGATTC +TGATTG +TGATCA +TGATCT +TGATCC +TGATCG +TGATGA +TGATGT +TGATGC +TGATGG +TGACAA +TGACAT +TGACAC +TGACAG +TGACTA +TGACTT +TGACTC +TGACTG +TGACCA +TGACCT +TGACCC +TGACCG +TGACGA +TGACGT +TGACGC +TGACGG +TGAGAA +TGAGAT +TGAGAC +TGAGAG +TGAGTA +TGAGTT +TGAGTC +TGAGTG +TGAGCA +TGAGCT +TGAGCC +TGAGCG +TGAGGA +TGAGGT +TGAGGC +TGAGGG +TGTAAA +TGTAAT +TGTAAC +TGTAAG +TGTATA +TGTATT +TGTATC +TGTATG +TGTACA +TGTACT +TGTACC +TGTACG +TGTAGA +TGTAGT +TGTAGC +TGTAGG +TGTTAA +TGTTAT +TGTTAC +TGTTAG +TGTTTA +TGTTTT +TGTTTC +TGTTTG +TGTTCA +TGTTCT +TGTTCC +TGTTCG +TGTTGA +TGTTGT +TGTTGC +TGTTGG +TGTCAA +TGTCAT +TGTCAC +TGTCAG +TGTCTA +TGTCTT +TGTCTC +TGTCTG +TGTCCA +TGTCCT +TGTCCC +TGTCCG +TGTCGA +TGTCGT +TGTCGC +TGTCGG +TGTGAA +TGTGAT +TGTGAC +TGTGAG +TGTGTA +TGTGTT +TGTGTC +TGTGTG +TGTGCA +TGTGCT +TGTGCC +TGTGCG +TGTGGA +TGTGGT +TGTGGC +TGTGGG +TGCAAA +TGCAAT +TGCAAC +TGCAAG +TGCATA +TGCATT +TGCATC +TGCATG +TGCACA +TGCACT +TGCACC +TGCACG +TGCAGA +TGCAGT +TGCAGC +TGCAGG +TGCTAA +TGCTAT +TGCTAC +TGCTAG +TGCTTA +TGCTTT +TGCTTC +TGCTTG +TGCTCA +TGCTCT +TGCTCC +TGCTCG +TGCTGA +TGCTGT +TGCTGC +TGCTGG +TGCCAA +TGCCAT +TGCCAC +TGCCAG +TGCCTA +TGCCTT +TGCCTC +TGCCTG +TGCCCA +TGCCCT +TGCCCC +TGCCCG +TGCCGA +TGCCGT +TGCCGC +TGCCGG +TGCGAA +TGCGAT +TGCGAC +TGCGAG +TGCGTA +TGCGTT +TGCGTC +TGCGTG +TGCGCA +TGCGCT +TGCGCC +TGCGCG +TGCGGA +TGCGGT +TGCGGC +TGCGGG +TGGAAA +TGGAAT +TGGAAC +TGGAAG +TGGATA +TGGATT +TGGATC +TGGATG +TGGACA +TGGACT +TGGACC +TGGACG +TGGAGA +TGGAGT +TGGAGC +TGGAGG +TGGTAA +TGGTAT +TGGTAC +TGGTAG +TGGTTA +TGGTTT +TGGTTC +TGGTTG +TGGTCA +TGGTCT +TGGTCC +TGGTCG +TGGTGA +TGGTGT +TGGTGC +TGGTGG +TGGCAA +TGGCAT +TGGCAC +TGGCAG +TGGCTA +TGGCTT +TGGCTC +TGGCTG +TGGCCA +TGGCCT +TGGCCC +TGGCCG +TGGCGA +TGGCGT +TGGCGC +TGGCGG +TGGGAA +TGGGAT +TGGGAC +TGGGAG +TGGGTA +TGGGTT +TGGGTC +TGGGTG +TGGGCA +TGGGCT +TGGGCC +TGGGCG +TGGGGA +TGGGGT +TGGGGC +TGGGGG +CAAAAA +CAAAAT +CAAAAC +CAAAAG +CAAATA +CAAATT +CAAATC +CAAATG +CAAACA +CAAACT +CAAACC +CAAACG +CAAAGA +CAAAGT +CAAAGC +CAAAGG +CAATAA +CAATAT +CAATAC +CAATAG +CAATTA +CAATTT +CAATTC +CAATTG +CAATCA +CAATCT +CAATCC +CAATCG +CAATGA +CAATGT +CAATGC +CAATGG +CAACAA +CAACAT +CAACAC +CAACAG +CAACTA +CAACTT +CAACTC +CAACTG +CAACCA +CAACCT +CAACCC +CAACCG +CAACGA +CAACGT +CAACGC +CAACGG +CAAGAA +CAAGAT +CAAGAC +CAAGAG +CAAGTA +CAAGTT +CAAGTC +CAAGTG +CAAGCA +CAAGCT +CAAGCC +CAAGCG +CAAGGA +CAAGGT +CAAGGC +CAAGGG +CATAAA +CATAAT +CATAAC +CATAAG +CATATA +CATATT +CATATC +CATATG +CATACA +CATACT +CATACC +CATACG +CATAGA +CATAGT +CATAGC +CATAGG +CATTAA +CATTAT +CATTAC +CATTAG +CATTTA +CATTTT +CATTTC +CATTTG +CATTCA +CATTCT +CATTCC +CATTCG +CATTGA +CATTGT +CATTGC +CATTGG +CATCAA +CATCAT +CATCAC +CATCAG +CATCTA +CATCTT +CATCTC +CATCTG +CATCCA +CATCCT +CATCCC +CATCCG +CATCGA +CATCGT +CATCGC +CATCGG +CATGAA +CATGAT +CATGAC +CATGAG +CATGTA +CATGTT +CATGTC +CATGTG +CATGCA +CATGCT +CATGCC +CATGCG +CATGGA +CATGGT +CATGGC +CATGGG +CACAAA +CACAAT +CACAAC +CACAAG +CACATA +CACATT +CACATC +CACATG +CACACA +CACACT +CACACC +CACACG +CACAGA +CACAGT +CACAGC +CACAGG +CACTAA +CACTAT +CACTAC +CACTAG +CACTTA +CACTTT +CACTTC +CACTTG +CACTCA +CACTCT +CACTCC +CACTCG +CACTGA +CACTGT +CACTGC +CACTGG +CACCAA +CACCAT +CACCAC +CACCAG +CACCTA +CACCTT +CACCTC +CACCTG +CACCCA +CACCCT +CACCCC +CACCCG +CACCGA +CACCGT +CACCGC +CACCGG +CACGAA +CACGAT +CACGAC +CACGAG +CACGTA +CACGTT +CACGTC +CACGTG +CACGCA +CACGCT +CACGCC +CACGCG +CACGGA +CACGGT +CACGGC +CACGGG +CAGAAA +CAGAAT +CAGAAC +CAGAAG +CAGATA +CAGATT +CAGATC +CAGATG +CAGACA +CAGACT +CAGACC +CAGACG +CAGAGA +CAGAGT +CAGAGC +CAGAGG +CAGTAA +CAGTAT +CAGTAC +CAGTAG +CAGTTA +CAGTTT +CAGTTC +CAGTTG +CAGTCA +CAGTCT +CAGTCC +CAGTCG +CAGTGA +CAGTGT +CAGTGC +CAGTGG +CAGCAA +CAGCAT +CAGCAC +CAGCAG +CAGCTA +CAGCTT +CAGCTC +CAGCTG +CAGCCA +CAGCCT +CAGCCC +CAGCCG +CAGCGA +CAGCGT +CAGCGC +CAGCGG +CAGGAA +CAGGAT +CAGGAC +CAGGAG +CAGGTA +CAGGTT +CAGGTC +CAGGTG +CAGGCA +CAGGCT +CAGGCC +CAGGCG +CAGGGA +CAGGGT +CAGGGC +CAGGGG +CTAAAA +CTAAAT +CTAAAC +CTAAAG +CTAATA +CTAATT +CTAATC +CTAATG +CTAACA +CTAACT +CTAACC +CTAACG +CTAAGA +CTAAGT +CTAAGC +CTAAGG +CTATAA +CTATAT +CTATAC +CTATAG +CTATTA +CTATTT +CTATTC +CTATTG +CTATCA +CTATCT +CTATCC +CTATCG +CTATGA +CTATGT +CTATGC +CTATGG +CTACAA +CTACAT +CTACAC +CTACAG +CTACTA +CTACTT +CTACTC +CTACTG +CTACCA +CTACCT +CTACCC +CTACCG +CTACGA +CTACGT +CTACGC +CTACGG +CTAGAA +CTAGAT +CTAGAC +CTAGAG +CTAGTA +CTAGTT +CTAGTC +CTAGTG +CTAGCA +CTAGCT +CTAGCC +CTAGCG +CTAGGA +CTAGGT +CTAGGC +CTAGGG +CTTAAA +CTTAAT +CTTAAC +CTTAAG +CTTATA +CTTATT +CTTATC +CTTATG +CTTACA +CTTACT +CTTACC +CTTACG +CTTAGA +CTTAGT +CTTAGC +CTTAGG +CTTTAA +CTTTAT +CTTTAC +CTTTAG +CTTTTA +CTTTTT +CTTTTC +CTTTTG +CTTTCA +CTTTCT +CTTTCC +CTTTCG +CTTTGA +CTTTGT +CTTTGC +CTTTGG +CTTCAA +CTTCAT +CTTCAC +CTTCAG +CTTCTA +CTTCTT +CTTCTC +CTTCTG +CTTCCA +CTTCCT +CTTCCC +CTTCCG +CTTCGA +CTTCGT +CTTCGC +CTTCGG +CTTGAA +CTTGAT +CTTGAC +CTTGAG +CTTGTA +CTTGTT +CTTGTC +CTTGTG +CTTGCA +CTTGCT +CTTGCC +CTTGCG +CTTGGA +CTTGGT +CTTGGC +CTTGGG +CTCAAA +CTCAAT +CTCAAC +CTCAAG +CTCATA +CTCATT +CTCATC +CTCATG +CTCACA +CTCACT +CTCACC +CTCACG +CTCAGA +CTCAGT +CTCAGC +CTCAGG +CTCTAA +CTCTAT +CTCTAC +CTCTAG +CTCTTA +CTCTTT +CTCTTC +CTCTTG +CTCTCA +CTCTCT +CTCTCC +CTCTCG +CTCTGA +CTCTGT +CTCTGC +CTCTGG +CTCCAA +CTCCAT +CTCCAC +CTCCAG +CTCCTA +CTCCTT +CTCCTC +CTCCTG +CTCCCA +CTCCCT +CTCCCC +CTCCCG +CTCCGA +CTCCGT +CTCCGC +CTCCGG +CTCGAA +CTCGAT +CTCGAC +CTCGAG +CTCGTA +CTCGTT +CTCGTC +CTCGTG +CTCGCA +CTCGCT +CTCGCC +CTCGCG +CTCGGA +CTCGGT +CTCGGC +CTCGGG +CTGAAA +CTGAAT +CTGAAC +CTGAAG +CTGATA +CTGATT +CTGATC +CTGATG +CTGACA +CTGACT +CTGACC +CTGACG +CTGAGA +CTGAGT +CTGAGC +CTGAGG +CTGTAA +CTGTAT +CTGTAC +CTGTAG +CTGTTA +CTGTTT +CTGTTC +CTGTTG +CTGTCA +CTGTCT +CTGTCC +CTGTCG +CTGTGA +CTGTGT +CTGTGC +CTGTGG +CTGCAA +CTGCAT +CTGCAC +CTGCAG +CTGCTA +CTGCTT +CTGCTC +CTGCTG +CTGCCA +CTGCCT +CTGCCC +CTGCCG +CTGCGA +CTGCGT +CTGCGC +CTGCGG +CTGGAA +CTGGAT +CTGGAC +CTGGAG +CTGGTA +CTGGTT +CTGGTC +CTGGTG +CTGGCA +CTGGCT +CTGGCC +CTGGCG +CTGGGA +CTGGGT +CTGGGC +CTGGGG +CCAAAA +CCAAAT +CCAAAC +CCAAAG +CCAATA +CCAATT +CCAATC +CCAATG +CCAACA +CCAACT +CCAACC +CCAACG +CCAAGA +CCAAGT +CCAAGC +CCAAGG +CCATAA +CCATAT +CCATAC +CCATAG +CCATTA +CCATTT +CCATTC +CCATTG +CCATCA +CCATCT +CCATCC +CCATCG +CCATGA +CCATGT +CCATGC +CCATGG +CCACAA +CCACAT +CCACAC +CCACAG +CCACTA +CCACTT +CCACTC +CCACTG +CCACCA +CCACCT +CCACCC +CCACCG +CCACGA +CCACGT +CCACGC +CCACGG +CCAGAA +CCAGAT +CCAGAC +CCAGAG +CCAGTA +CCAGTT +CCAGTC +CCAGTG +CCAGCA +CCAGCT +CCAGCC +CCAGCG +CCAGGA +CCAGGT +CCAGGC +CCAGGG +CCTAAA +CCTAAT +CCTAAC +CCTAAG +CCTATA +CCTATT +CCTATC +CCTATG +CCTACA +CCTACT +CCTACC +CCTACG +CCTAGA +CCTAGT +CCTAGC +CCTAGG +CCTTAA +CCTTAT +CCTTAC +CCTTAG +CCTTTA +CCTTTT +CCTTTC +CCTTTG +CCTTCA +CCTTCT +CCTTCC +CCTTCG +CCTTGA +CCTTGT +CCTTGC +CCTTGG +CCTCAA +CCTCAT +CCTCAC +CCTCAG +CCTCTA +CCTCTT +CCTCTC +CCTCTG +CCTCCA +CCTCCT +CCTCCC +CCTCCG +CCTCGA +CCTCGT +CCTCGC +CCTCGG +CCTGAA +CCTGAT +CCTGAC +CCTGAG +CCTGTA +CCTGTT +CCTGTC +CCTGTG +CCTGCA +CCTGCT +CCTGCC +CCTGCG +CCTGGA +CCTGGT +CCTGGC +CCTGGG +CCCAAA +CCCAAT +CCCAAC +CCCAAG +CCCATA +CCCATT +CCCATC +CCCATG +CCCACA +CCCACT +CCCACC +CCCACG +CCCAGA +CCCAGT +CCCAGC +CCCAGG +CCCTAA +CCCTAT +CCCTAC +CCCTAG +CCCTTA +CCCTTT +CCCTTC +CCCTTG +CCCTCA +CCCTCT +CCCTCC +CCCTCG +CCCTGA +CCCTGT +CCCTGC +CCCTGG +CCCCAA +CCCCAT +CCCCAC +CCCCAG +CCCCTA +CCCCTT +CCCCTC +CCCCTG +CCCCCA +CCCCCT +CCCCCC +CCCCCG +CCCCGA +CCCCGT +CCCCGC +CCCCGG +CCCGAA +CCCGAT +CCCGAC +CCCGAG +CCCGTA +CCCGTT +CCCGTC +CCCGTG +CCCGCA +CCCGCT +CCCGCC +CCCGCG +CCCGGA +CCCGGT +CCCGGC +CCCGGG +CCGAAA +CCGAAT +CCGAAC +CCGAAG +CCGATA +CCGATT +CCGATC +CCGATG +CCGACA +CCGACT +CCGACC +CCGACG +CCGAGA +CCGAGT +CCGAGC +CCGAGG +CCGTAA +CCGTAT +CCGTAC +CCGTAG +CCGTTA +CCGTTT +CCGTTC +CCGTTG +CCGTCA +CCGTCT +CCGTCC +CCGTCG +CCGTGA +CCGTGT +CCGTGC +CCGTGG +CCGCAA +CCGCAT +CCGCAC +CCGCAG +CCGCTA +CCGCTT +CCGCTC +CCGCTG +CCGCCA +CCGCCT +CCGCCC +CCGCCG +CCGCGA +CCGCGT +CCGCGC +CCGCGG +CCGGAA +CCGGAT +CCGGAC +CCGGAG +CCGGTA +CCGGTT +CCGGTC +CCGGTG +CCGGCA +CCGGCT +CCGGCC +CCGGCG +CCGGGA +CCGGGT +CCGGGC +CCGGGG +CGAAAA +CGAAAT +CGAAAC +CGAAAG +CGAATA +CGAATT +CGAATC +CGAATG +CGAACA +CGAACT +CGAACC +CGAACG +CGAAGA +CGAAGT +CGAAGC +CGAAGG +CGATAA +CGATAT +CGATAC +CGATAG +CGATTA +CGATTT +CGATTC +CGATTG +CGATCA +CGATCT +CGATCC +CGATCG +CGATGA +CGATGT +CGATGC +CGATGG +CGACAA +CGACAT +CGACAC +CGACAG +CGACTA +CGACTT +CGACTC +CGACTG +CGACCA +CGACCT +CGACCC +CGACCG +CGACGA +CGACGT +CGACGC +CGACGG +CGAGAA +CGAGAT +CGAGAC +CGAGAG +CGAGTA +CGAGTT +CGAGTC +CGAGTG +CGAGCA +CGAGCT +CGAGCC +CGAGCG +CGAGGA +CGAGGT +CGAGGC +CGAGGG +CGTAAA +CGTAAT +CGTAAC +CGTAAG +CGTATA +CGTATT +CGTATC +CGTATG +CGTACA +CGTACT +CGTACC +CGTACG +CGTAGA +CGTAGT +CGTAGC +CGTAGG +CGTTAA +CGTTAT +CGTTAC +CGTTAG +CGTTTA +CGTTTT +CGTTTC +CGTTTG +CGTTCA +CGTTCT +CGTTCC +CGTTCG +CGTTGA +CGTTGT +CGTTGC +CGTTGG +CGTCAA +CGTCAT +CGTCAC +CGTCAG +CGTCTA +CGTCTT +CGTCTC +CGTCTG +CGTCCA +CGTCCT +CGTCCC +CGTCCG +CGTCGA +CGTCGT +CGTCGC +CGTCGG +CGTGAA +CGTGAT +CGTGAC +CGTGAG +CGTGTA +CGTGTT +CGTGTC +CGTGTG +CGTGCA +CGTGCT +CGTGCC +CGTGCG +CGTGGA +CGTGGT +CGTGGC +CGTGGG +CGCAAA +CGCAAT +CGCAAC +CGCAAG +CGCATA +CGCATT +CGCATC +CGCATG +CGCACA +CGCACT +CGCACC +CGCACG +CGCAGA +CGCAGT +CGCAGC +CGCAGG +CGCTAA +CGCTAT +CGCTAC +CGCTAG +CGCTTA +CGCTTT +CGCTTC +CGCTTG +CGCTCA +CGCTCT +CGCTCC +CGCTCG +CGCTGA +CGCTGT +CGCTGC +CGCTGG +CGCCAA +CGCCAT +CGCCAC +CGCCAG +CGCCTA +CGCCTT +CGCCTC +CGCCTG +CGCCCA +CGCCCT +CGCCCC +CGCCCG +CGCCGA +CGCCGT +CGCCGC +CGCCGG +CGCGAA +CGCGAT +CGCGAC +CGCGAG +CGCGTA +CGCGTT +CGCGTC +CGCGTG +CGCGCA +CGCGCT +CGCGCC +CGCGCG +CGCGGA +CGCGGT +CGCGGC +CGCGGG +CGGAAA +CGGAAT +CGGAAC +CGGAAG +CGGATA +CGGATT +CGGATC +CGGATG +CGGACA +CGGACT +CGGACC +CGGACG +CGGAGA +CGGAGT +CGGAGC +CGGAGG +CGGTAA +CGGTAT +CGGTAC +CGGTAG +CGGTTA +CGGTTT +CGGTTC +CGGTTG +CGGTCA +CGGTCT +CGGTCC +CGGTCG +CGGTGA +CGGTGT +CGGTGC +CGGTGG +CGGCAA +CGGCAT +CGGCAC +CGGCAG +CGGCTA +CGGCTT +CGGCTC +CGGCTG +CGGCCA +CGGCCT +CGGCCC +CGGCCG +CGGCGA +CGGCGT +CGGCGC +CGGCGG +CGGGAA +CGGGAT +CGGGAC +CGGGAG +CGGGTA +CGGGTT +CGGGTC +CGGGTG +CGGGCA +CGGGCT +CGGGCC +CGGGCG +CGGGGA +CGGGGT +CGGGGC +CGGGGG +GAAAAA +GAAAAT +GAAAAC +GAAAAG +GAAATA +GAAATT +GAAATC +GAAATG +GAAACA +GAAACT +GAAACC +GAAACG +GAAAGA +GAAAGT +GAAAGC +GAAAGG +GAATAA +GAATAT +GAATAC +GAATAG +GAATTA +GAATTT +GAATTC +GAATTG +GAATCA +GAATCT +GAATCC +GAATCG +GAATGA +GAATGT +GAATGC +GAATGG +GAACAA +GAACAT +GAACAC +GAACAG +GAACTA +GAACTT +GAACTC +GAACTG +GAACCA +GAACCT +GAACCC +GAACCG +GAACGA +GAACGT +GAACGC +GAACGG +GAAGAA +GAAGAT +GAAGAC +GAAGAG +GAAGTA +GAAGTT +GAAGTC +GAAGTG +GAAGCA +GAAGCT +GAAGCC +GAAGCG +GAAGGA +GAAGGT +GAAGGC +GAAGGG +GATAAA +GATAAT +GATAAC +GATAAG +GATATA +GATATT +GATATC +GATATG +GATACA +GATACT +GATACC +GATACG +GATAGA +GATAGT +GATAGC +GATAGG +GATTAA +GATTAT +GATTAC +GATTAG +GATTTA +GATTTT +GATTTC +GATTTG +GATTCA +GATTCT +GATTCC +GATTCG +GATTGA +GATTGT +GATTGC +GATTGG +GATCAA +GATCAT +GATCAC +GATCAG +GATCTA +GATCTT +GATCTC +GATCTG +GATCCA +GATCCT +GATCCC +GATCCG +GATCGA +GATCGT +GATCGC +GATCGG +GATGAA +GATGAT +GATGAC +GATGAG +GATGTA +GATGTT +GATGTC +GATGTG +GATGCA +GATGCT +GATGCC +GATGCG +GATGGA +GATGGT +GATGGC +GATGGG +GACAAA +GACAAT +GACAAC +GACAAG +GACATA +GACATT +GACATC +GACATG +GACACA +GACACT +GACACC +GACACG +GACAGA +GACAGT +GACAGC +GACAGG +GACTAA +GACTAT +GACTAC +GACTAG +GACTTA +GACTTT +GACTTC +GACTTG +GACTCA +GACTCT +GACTCC +GACTCG +GACTGA +GACTGT +GACTGC +GACTGG +GACCAA +GACCAT +GACCAC +GACCAG +GACCTA +GACCTT +GACCTC +GACCTG +GACCCA +GACCCT +GACCCC +GACCCG +GACCGA +GACCGT +GACCGC +GACCGG +GACGAA +GACGAT +GACGAC +GACGAG +GACGTA +GACGTT +GACGTC +GACGTG +GACGCA +GACGCT +GACGCC +GACGCG +GACGGA +GACGGT +GACGGC +GACGGG +GAGAAA +GAGAAT +GAGAAC +GAGAAG +GAGATA +GAGATT +GAGATC +GAGATG +GAGACA +GAGACT +GAGACC +GAGACG +GAGAGA +GAGAGT +GAGAGC +GAGAGG +GAGTAA +GAGTAT +GAGTAC +GAGTAG +GAGTTA +GAGTTT +GAGTTC +GAGTTG +GAGTCA +GAGTCT +GAGTCC +GAGTCG +GAGTGA +GAGTGT +GAGTGC +GAGTGG +GAGCAA +GAGCAT +GAGCAC +GAGCAG +GAGCTA +GAGCTT +GAGCTC +GAGCTG +GAGCCA +GAGCCT +GAGCCC +GAGCCG +GAGCGA +GAGCGT +GAGCGC +GAGCGG +GAGGAA +GAGGAT +GAGGAC +GAGGAG +GAGGTA +GAGGTT +GAGGTC +GAGGTG +GAGGCA +GAGGCT +GAGGCC +GAGGCG +GAGGGA +GAGGGT +GAGGGC +GAGGGG +GTAAAA +GTAAAT +GTAAAC +GTAAAG +GTAATA +GTAATT +GTAATC +GTAATG +GTAACA +GTAACT +GTAACC +GTAACG +GTAAGA +GTAAGT +GTAAGC +GTAAGG +GTATAA +GTATAT +GTATAC +GTATAG +GTATTA +GTATTT +GTATTC +GTATTG +GTATCA +GTATCT +GTATCC +GTATCG +GTATGA +GTATGT +GTATGC +GTATGG +GTACAA +GTACAT +GTACAC +GTACAG +GTACTA +GTACTT +GTACTC +GTACTG +GTACCA +GTACCT +GTACCC +GTACCG +GTACGA +GTACGT +GTACGC +GTACGG +GTAGAA +GTAGAT +GTAGAC +GTAGAG +GTAGTA +GTAGTT +GTAGTC +GTAGTG +GTAGCA +GTAGCT +GTAGCC +GTAGCG +GTAGGA +GTAGGT +GTAGGC +GTAGGG +GTTAAA +GTTAAT +GTTAAC +GTTAAG +GTTATA +GTTATT +GTTATC +GTTATG +GTTACA +GTTACT +GTTACC +GTTACG +GTTAGA +GTTAGT +GTTAGC +GTTAGG +GTTTAA +GTTTAT +GTTTAC +GTTTAG +GTTTTA +GTTTTT +GTTTTC +GTTTTG +GTTTCA +GTTTCT +GTTTCC +GTTTCG +GTTTGA +GTTTGT +GTTTGC +GTTTGG +GTTCAA +GTTCAT +GTTCAC +GTTCAG +GTTCTA +GTTCTT +GTTCTC +GTTCTG +GTTCCA +GTTCCT +GTTCCC +GTTCCG +GTTCGA +GTTCGT +GTTCGC +GTTCGG +GTTGAA +GTTGAT +GTTGAC +GTTGAG +GTTGTA +GTTGTT +GTTGTC +GTTGTG +GTTGCA +GTTGCT +GTTGCC +GTTGCG +GTTGGA +GTTGGT +GTTGGC +GTTGGG +GTCAAA +GTCAAT +GTCAAC +GTCAAG +GTCATA +GTCATT +GTCATC +GTCATG +GTCACA +GTCACT +GTCACC +GTCACG +GTCAGA +GTCAGT +GTCAGC +GTCAGG +GTCTAA +GTCTAT +GTCTAC +GTCTAG +GTCTTA +GTCTTT +GTCTTC +GTCTTG +GTCTCA +GTCTCT +GTCTCC +GTCTCG +GTCTGA +GTCTGT +GTCTGC +GTCTGG +GTCCAA +GTCCAT +GTCCAC +GTCCAG +GTCCTA +GTCCTT +GTCCTC +GTCCTG +GTCCCA +GTCCCT +GTCCCC +GTCCCG +GTCCGA +GTCCGT +GTCCGC +GTCCGG +GTCGAA +GTCGAT +GTCGAC +GTCGAG +GTCGTA +GTCGTT +GTCGTC +GTCGTG +GTCGCA +GTCGCT +GTCGCC +GTCGCG +GTCGGA +GTCGGT +GTCGGC +GTCGGG +GTGAAA +GTGAAT +GTGAAC +GTGAAG +GTGATA +GTGATT +GTGATC +GTGATG +GTGACA +GTGACT +GTGACC +GTGACG +GTGAGA +GTGAGT +GTGAGC +GTGAGG +GTGTAA +GTGTAT +GTGTAC +GTGTAG +GTGTTA +GTGTTT +GTGTTC +GTGTTG +GTGTCA +GTGTCT +GTGTCC +GTGTCG +GTGTGA +GTGTGT +GTGTGC +GTGTGG +GTGCAA +GTGCAT +GTGCAC +GTGCAG +GTGCTA +GTGCTT +GTGCTC +GTGCTG +GTGCCA +GTGCCT +GTGCCC +GTGCCG +GTGCGA +GTGCGT +GTGCGC +GTGCGG +GTGGAA +GTGGAT +GTGGAC +GTGGAG +GTGGTA +GTGGTT +GTGGTC +GTGGTG +GTGGCA +GTGGCT +GTGGCC +GTGGCG +GTGGGA +GTGGGT +GTGGGC +GTGGGG +GCAAAA +GCAAAT +GCAAAC +GCAAAG +GCAATA +GCAATT +GCAATC +GCAATG +GCAACA +GCAACT +GCAACC +GCAACG +GCAAGA +GCAAGT +GCAAGC +GCAAGG +GCATAA +GCATAT +GCATAC +GCATAG +GCATTA +GCATTT +GCATTC +GCATTG +GCATCA +GCATCT +GCATCC +GCATCG +GCATGA +GCATGT +GCATGC +GCATGG +GCACAA +GCACAT +GCACAC +GCACAG +GCACTA +GCACTT +GCACTC +GCACTG +GCACCA +GCACCT +GCACCC +GCACCG +GCACGA +GCACGT +GCACGC +GCACGG +GCAGAA +GCAGAT +GCAGAC +GCAGAG +GCAGTA +GCAGTT +GCAGTC +GCAGTG +GCAGCA +GCAGCT +GCAGCC +GCAGCG +GCAGGA +GCAGGT +GCAGGC +GCAGGG +GCTAAA +GCTAAT +GCTAAC +GCTAAG +GCTATA +GCTATT +GCTATC +GCTATG +GCTACA +GCTACT +GCTACC +GCTACG +GCTAGA +GCTAGT +GCTAGC +GCTAGG +GCTTAA +GCTTAT +GCTTAC +GCTTAG +GCTTTA +GCTTTT +GCTTTC +GCTTTG +GCTTCA +GCTTCT +GCTTCC +GCTTCG +GCTTGA +GCTTGT +GCTTGC +GCTTGG +GCTCAA +GCTCAT +GCTCAC +GCTCAG +GCTCTA +GCTCTT +GCTCTC +GCTCTG +GCTCCA +GCTCCT +GCTCCC +GCTCCG +GCTCGA +GCTCGT +GCTCGC +GCTCGG +GCTGAA +GCTGAT +GCTGAC +GCTGAG +GCTGTA +GCTGTT +GCTGTC +GCTGTG +GCTGCA +GCTGCT +GCTGCC +GCTGCG +GCTGGA +GCTGGT +GCTGGC +GCTGGG +GCCAAA +GCCAAT +GCCAAC +GCCAAG +GCCATA +GCCATT +GCCATC +GCCATG +GCCACA +GCCACT +GCCACC +GCCACG +GCCAGA +GCCAGT +GCCAGC +GCCAGG +GCCTAA +GCCTAT +GCCTAC +GCCTAG +GCCTTA +GCCTTT +GCCTTC +GCCTTG +GCCTCA +GCCTCT +GCCTCC +GCCTCG +GCCTGA +GCCTGT +GCCTGC +GCCTGG +GCCCAA +GCCCAT +GCCCAC +GCCCAG +GCCCTA +GCCCTT +GCCCTC +GCCCTG +GCCCCA +GCCCCT +GCCCCC +GCCCCG +GCCCGA +GCCCGT +GCCCGC +GCCCGG +GCCGAA +GCCGAT +GCCGAC +GCCGAG +GCCGTA +GCCGTT +GCCGTC +GCCGTG +GCCGCA +GCCGCT +GCCGCC +GCCGCG +GCCGGA +GCCGGT +GCCGGC +GCCGGG +GCGAAA +GCGAAT +GCGAAC +GCGAAG +GCGATA +GCGATT +GCGATC +GCGATG +GCGACA +GCGACT +GCGACC +GCGACG +GCGAGA +GCGAGT +GCGAGC +GCGAGG +GCGTAA +GCGTAT +GCGTAC +GCGTAG +GCGTTA +GCGTTT +GCGTTC +GCGTTG +GCGTCA +GCGTCT +GCGTCC +GCGTCG +GCGTGA +GCGTGT +GCGTGC +GCGTGG +GCGCAA +GCGCAT +GCGCAC +GCGCAG +GCGCTA +GCGCTT +GCGCTC +GCGCTG +GCGCCA +GCGCCT +GCGCCC +GCGCCG +GCGCGA +GCGCGT +GCGCGC +GCGCGG +GCGGAA +GCGGAT +GCGGAC +GCGGAG +GCGGTA +GCGGTT +GCGGTC +GCGGTG +GCGGCA +GCGGCT +GCGGCC +GCGGCG +GCGGGA +GCGGGT +GCGGGC +GCGGGG +GGAAAA +GGAAAT +GGAAAC +GGAAAG +GGAATA +GGAATT +GGAATC +GGAATG +GGAACA +GGAACT +GGAACC +GGAACG +GGAAGA +GGAAGT +GGAAGC +GGAAGG +GGATAA +GGATAT +GGATAC +GGATAG +GGATTA +GGATTT +GGATTC +GGATTG +GGATCA +GGATCT +GGATCC +GGATCG +GGATGA +GGATGT +GGATGC +GGATGG +GGACAA +GGACAT +GGACAC +GGACAG +GGACTA +GGACTT +GGACTC +GGACTG +GGACCA +GGACCT +GGACCC +GGACCG +GGACGA +GGACGT +GGACGC +GGACGG +GGAGAA +GGAGAT +GGAGAC +GGAGAG +GGAGTA +GGAGTT +GGAGTC +GGAGTG +GGAGCA +GGAGCT +GGAGCC +GGAGCG +GGAGGA +GGAGGT +GGAGGC +GGAGGG +GGTAAA +GGTAAT +GGTAAC +GGTAAG +GGTATA +GGTATT +GGTATC +GGTATG +GGTACA +GGTACT +GGTACC +GGTACG +GGTAGA +GGTAGT +GGTAGC +GGTAGG +GGTTAA +GGTTAT +GGTTAC +GGTTAG +GGTTTA +GGTTTT +GGTTTC +GGTTTG +GGTTCA +GGTTCT +GGTTCC +GGTTCG +GGTTGA +GGTTGT +GGTTGC +GGTTGG +GGTCAA +GGTCAT +GGTCAC +GGTCAG +GGTCTA +GGTCTT +GGTCTC +GGTCTG +GGTCCA +GGTCCT +GGTCCC +GGTCCG +GGTCGA +GGTCGT +GGTCGC +GGTCGG +GGTGAA +GGTGAT +GGTGAC +GGTGAG +GGTGTA +GGTGTT +GGTGTC +GGTGTG +GGTGCA +GGTGCT +GGTGCC +GGTGCG +GGTGGA +GGTGGT +GGTGGC +GGTGGG +GGCAAA +GGCAAT +GGCAAC +GGCAAG +GGCATA +GGCATT +GGCATC +GGCATG +GGCACA +GGCACT +GGCACC +GGCACG +GGCAGA +GGCAGT +GGCAGC +GGCAGG +GGCTAA +GGCTAT +GGCTAC +GGCTAG +GGCTTA +GGCTTT +GGCTTC +GGCTTG +GGCTCA +GGCTCT +GGCTCC +GGCTCG +GGCTGA +GGCTGT +GGCTGC +GGCTGG +GGCCAA +GGCCAT +GGCCAC +GGCCAG +GGCCTA +GGCCTT +GGCCTC +GGCCTG +GGCCCA +GGCCCT +GGCCCC +GGCCCG +GGCCGA +GGCCGT +GGCCGC +GGCCGG +GGCGAA +GGCGAT +GGCGAC +GGCGAG +GGCGTA +GGCGTT +GGCGTC +GGCGTG +GGCGCA +GGCGCT +GGCGCC +GGCGCG +GGCGGA +GGCGGT +GGCGGC +GGCGGG +GGGAAA +GGGAAT +GGGAAC +GGGAAG +GGGATA +GGGATT +GGGATC +GGGATG +GGGACA +GGGACT +GGGACC +GGGACG +GGGAGA +GGGAGT +GGGAGC +GGGAGG +GGGTAA +GGGTAT +GGGTAC +GGGTAG +GGGTTA +GGGTTT +GGGTTC +GGGTTG +GGGTCA +GGGTCT +GGGTCC +GGGTCG +GGGTGA +GGGTGT +GGGTGC +GGGTGG +GGGCAA +GGGCAT +GGGCAC +GGGCAG +GGGCTA +GGGCTT +GGGCTC +GGGCTG +GGGCCA +GGGCCT +GGGCCC +GGGCCG +GGGCGA +GGGCGT +GGGCGC +GGGCGG +GGGGAA +GGGGAT +GGGGAC +GGGGAG +GGGGTA +GGGGTT +GGGGTC +GGGGTG +GGGGCA +GGGGCT +GGGGCC +GGGGCG +GGGGGA +GGGGGT +GGGGGC +GGGGGG diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-4000/config.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-4000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..9b14e61d327bf586fcff6e817cf9d365bae18ad2 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-4000/config.json @@ -0,0 +1,26 @@ +{ + "_name_or_path": "fabihamakhdoomi/TinyDNABERT", + "architectures": [ + "BertForSequenceClassification" + ], + "attention_probs_dropout_prob": 0.1, + "classifier_dropout": null, + "hidden_act": "gelu", + "hidden_dropout_prob": 0.1, + "hidden_size": 128, + "initializer_range": 0.02, + "intermediate_size": 512, + "layer_norm_eps": 1e-12, + "max_position_embeddings": 512, + "model_type": "bert", + "num_attention_heads": 2, + "num_hidden_layers": 2, + "pad_token_id": 0, + "position_embedding_type": "absolute", + "problem_type": "single_label_classification", + "torch_dtype": "float32", + "transformers_version": "4.44.0", + "type_vocab_size": 2, + "use_cache": true, + "vocab_size": 4101 +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-4000/model.safetensors b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-4000/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..0b79ba2f04d9252931594ad975b4a82fba7984df --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-4000/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e8d2d6e61584e3d32a2c9bc816887d6b906cacb145577831665efb539f5e5503 +size 4021688 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-4000/optimizer.pt b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-4000/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..9fef676237e2d658e8bb1d2e5cbdea382a6a2e5a --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-4000/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d649916c45c5f53df2368442e88e24b3caed5d1fc91ccc720979f76498aaca4d +size 8058700 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-4000/rng_state.pth b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-4000/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..e341a4a808ae24c21407bcc648bed6b2d6a16c6e --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-4000/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bc0eaa426d9fdf3c0f9bd6754ed553c3bff648a3431f959a858959a437717ed6 +size 14244 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-4000/scheduler.pt b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-4000/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..8c9a6adac55ac4d1afd6e1b61889589acfe31330 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-4000/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f3e454de3c0f52ddb8e5884c9783308658a6f61ede2ed6e377ca9a68d47c4515 +size 1064 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-4000/special_tokens_map.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-4000/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..a8b3208c2884c4efb86e49300fdd3dc877220cdf --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-4000/special_tokens_map.json @@ -0,0 +1,7 @@ +{ + "cls_token": "[CLS]", + "mask_token": "[MASK]", + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-4000/tokenizer.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-4000/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..ebd782de6963b9032ce9374cfe30f5275be1a30b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-4000/tokenizer.json @@ -0,0 +1,4265 @@ +{ + "version": "1.0", + "truncation": { + "direction": "Right", + "max_length": 512, + "strategy": "LongestFirst", + "stride": 0 + }, + "padding": { + "strategy": { + "Fixed": 512 + }, + "direction": "Right", + "pad_to_multiple_of": null, + "pad_id": 0, + "pad_type_id": 0, + "pad_token": "[PAD]" + }, + "added_tokens": [ + { + "id": 0, + "content": "[PAD]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 1, + "content": "[UNK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 2, + "content": "[CLS]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 3, + "content": "[SEP]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 4, + "content": "[MASK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + } + ], + "normalizer": { + "type": "BertNormalizer", + "clean_text": true, + "handle_chinese_chars": true, + "strip_accents": null, + "lowercase": false + }, + "pre_tokenizer": { + "type": "BertPreTokenizer" + }, + "post_processor": { + "type": "TemplateProcessing", + "single": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + } + ], + "pair": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "B", + "type_id": 1 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 1 + } + } + ], + "special_tokens": { + "[CLS]": { + "id": "[CLS]", + "ids": [ + 2 + ], + "tokens": [ + "[CLS]" + ] + }, + "[SEP]": { + "id": "[SEP]", + "ids": [ + 3 + ], + "tokens": [ + "[SEP]" + ] + } + } + }, + "decoder": { + "type": "WordPiece", + "prefix": "##", + "cleanup": true + }, + "model": { + "type": "WordPiece", + "unk_token": "[UNK]", + "continuing_subword_prefix": "##", + "max_input_chars_per_word": 100, + "vocab": { + "[PAD]": 0, + "[UNK]": 1, + "[CLS]": 2, + "[SEP]": 3, + "[MASK]": 4, + "AAAAAA": 5, + "AAAAAT": 6, + "AAAAAC": 7, + "AAAAAG": 8, + "AAAATA": 9, + "AAAATT": 10, + "AAAATC": 11, + "AAAATG": 12, + "AAAACA": 13, + "AAAACT": 14, + "AAAACC": 15, + "AAAACG": 16, + "AAAAGA": 17, + "AAAAGT": 18, + "AAAAGC": 19, + "AAAAGG": 20, + "AAATAA": 21, + "AAATAT": 22, + "AAATAC": 23, + "AAATAG": 24, + "AAATTA": 25, + "AAATTT": 26, + "AAATTC": 27, + "AAATTG": 28, + "AAATCA": 29, + "AAATCT": 30, + "AAATCC": 31, + "AAATCG": 32, + "AAATGA": 33, + "AAATGT": 34, + "AAATGC": 35, + "AAATGG": 36, + "AAACAA": 37, + "AAACAT": 38, + "AAACAC": 39, + "AAACAG": 40, + "AAACTA": 41, + "AAACTT": 42, + "AAACTC": 43, + "AAACTG": 44, + "AAACCA": 45, + "AAACCT": 46, + "AAACCC": 47, + "AAACCG": 48, + "AAACGA": 49, + "AAACGT": 50, + "AAACGC": 51, + "AAACGG": 52, + "AAAGAA": 53, + "AAAGAT": 54, + "AAAGAC": 55, + "AAAGAG": 56, + "AAAGTA": 57, + "AAAGTT": 58, + "AAAGTC": 59, + "AAAGTG": 60, + "AAAGCA": 61, + "AAAGCT": 62, + "AAAGCC": 63, + "AAAGCG": 64, + "AAAGGA": 65, + "AAAGGT": 66, + "AAAGGC": 67, + "AAAGGG": 68, + "AATAAA": 69, + "AATAAT": 70, + "AATAAC": 71, + "AATAAG": 72, + "AATATA": 73, + "AATATT": 74, + "AATATC": 75, + "AATATG": 76, + "AATACA": 77, + "AATACT": 78, + "AATACC": 79, + "AATACG": 80, + "AATAGA": 81, + "AATAGT": 82, + "AATAGC": 83, + "AATAGG": 84, + "AATTAA": 85, + "AATTAT": 86, + "AATTAC": 87, + "AATTAG": 88, + "AATTTA": 89, + "AATTTT": 90, + "AATTTC": 91, + "AATTTG": 92, + "AATTCA": 93, + "AATTCT": 94, + "AATTCC": 95, + "AATTCG": 96, + "AATTGA": 97, + "AATTGT": 98, + "AATTGC": 99, + "AATTGG": 100, + "AATCAA": 101, + "AATCAT": 102, + "AATCAC": 103, + "AATCAG": 104, + "AATCTA": 105, + "AATCTT": 106, + "AATCTC": 107, + "AATCTG": 108, + "AATCCA": 109, + "AATCCT": 110, + "AATCCC": 111, + "AATCCG": 112, + "AATCGA": 113, + "AATCGT": 114, + "AATCGC": 115, + "AATCGG": 116, + "AATGAA": 117, + "AATGAT": 118, + "AATGAC": 119, + "AATGAG": 120, + "AATGTA": 121, + "AATGTT": 122, + "AATGTC": 123, + "AATGTG": 124, + "AATGCA": 125, + "AATGCT": 126, + "AATGCC": 127, + "AATGCG": 128, + "AATGGA": 129, + "AATGGT": 130, + "AATGGC": 131, + "AATGGG": 132, + "AACAAA": 133, + "AACAAT": 134, + "AACAAC": 135, + "AACAAG": 136, + "AACATA": 137, + "AACATT": 138, + "AACATC": 139, + "AACATG": 140, + "AACACA": 141, + "AACACT": 142, + "AACACC": 143, + "AACACG": 144, + "AACAGA": 145, + "AACAGT": 146, + "AACAGC": 147, + "AACAGG": 148, + "AACTAA": 149, + "AACTAT": 150, + "AACTAC": 151, + "AACTAG": 152, + "AACTTA": 153, + "AACTTT": 154, + "AACTTC": 155, + "AACTTG": 156, + "AACTCA": 157, + "AACTCT": 158, + "AACTCC": 159, + "AACTCG": 160, + "AACTGA": 161, + "AACTGT": 162, + "AACTGC": 163, + "AACTGG": 164, + "AACCAA": 165, + "AACCAT": 166, + "AACCAC": 167, + "AACCAG": 168, + "AACCTA": 169, + "AACCTT": 170, + "AACCTC": 171, + "AACCTG": 172, + "AACCCA": 173, + "AACCCT": 174, + "AACCCC": 175, + "AACCCG": 176, + "AACCGA": 177, + "AACCGT": 178, + "AACCGC": 179, + "AACCGG": 180, + "AACGAA": 181, + "AACGAT": 182, + "AACGAC": 183, + "AACGAG": 184, + "AACGTA": 185, + "AACGTT": 186, + "AACGTC": 187, + "AACGTG": 188, + "AACGCA": 189, + "AACGCT": 190, + "AACGCC": 191, + "AACGCG": 192, + "AACGGA": 193, + "AACGGT": 194, + "AACGGC": 195, + "AACGGG": 196, + "AAGAAA": 197, + "AAGAAT": 198, + "AAGAAC": 199, + "AAGAAG": 200, + "AAGATA": 201, + "AAGATT": 202, + "AAGATC": 203, + "AAGATG": 204, + "AAGACA": 205, + "AAGACT": 206, + "AAGACC": 207, + "AAGACG": 208, + "AAGAGA": 209, + "AAGAGT": 210, + "AAGAGC": 211, + "AAGAGG": 212, + "AAGTAA": 213, + "AAGTAT": 214, + "AAGTAC": 215, + "AAGTAG": 216, + "AAGTTA": 217, + "AAGTTT": 218, + "AAGTTC": 219, + "AAGTTG": 220, + "AAGTCA": 221, + "AAGTCT": 222, + "AAGTCC": 223, + "AAGTCG": 224, + "AAGTGA": 225, + "AAGTGT": 226, + "AAGTGC": 227, + "AAGTGG": 228, + "AAGCAA": 229, + "AAGCAT": 230, + "AAGCAC": 231, + "AAGCAG": 232, + "AAGCTA": 233, + "AAGCTT": 234, + "AAGCTC": 235, + "AAGCTG": 236, + "AAGCCA": 237, + "AAGCCT": 238, + "AAGCCC": 239, + "AAGCCG": 240, + "AAGCGA": 241, + "AAGCGT": 242, + "AAGCGC": 243, + "AAGCGG": 244, + "AAGGAA": 245, + "AAGGAT": 246, + "AAGGAC": 247, + "AAGGAG": 248, + "AAGGTA": 249, + "AAGGTT": 250, + "AAGGTC": 251, + "AAGGTG": 252, + "AAGGCA": 253, + "AAGGCT": 254, + "AAGGCC": 255, + "AAGGCG": 256, + "AAGGGA": 257, + "AAGGGT": 258, + "AAGGGC": 259, + "AAGGGG": 260, + "ATAAAA": 261, + "ATAAAT": 262, + "ATAAAC": 263, + "ATAAAG": 264, + "ATAATA": 265, + "ATAATT": 266, + "ATAATC": 267, + "ATAATG": 268, + "ATAACA": 269, + "ATAACT": 270, + "ATAACC": 271, + "ATAACG": 272, + "ATAAGA": 273, + "ATAAGT": 274, + "ATAAGC": 275, + "ATAAGG": 276, + "ATATAA": 277, + "ATATAT": 278, + "ATATAC": 279, + "ATATAG": 280, + "ATATTA": 281, + "ATATTT": 282, + "ATATTC": 283, + "ATATTG": 284, + "ATATCA": 285, + "ATATCT": 286, + "ATATCC": 287, + "ATATCG": 288, + "ATATGA": 289, + "ATATGT": 290, + "ATATGC": 291, + "ATATGG": 292, + "ATACAA": 293, + "ATACAT": 294, + "ATACAC": 295, + "ATACAG": 296, + "ATACTA": 297, + "ATACTT": 298, + "ATACTC": 299, + "ATACTG": 300, + "ATACCA": 301, + "ATACCT": 302, + "ATACCC": 303, + "ATACCG": 304, + "ATACGA": 305, + "ATACGT": 306, + "ATACGC": 307, + "ATACGG": 308, + "ATAGAA": 309, + "ATAGAT": 310, + "ATAGAC": 311, + "ATAGAG": 312, + "ATAGTA": 313, + "ATAGTT": 314, + "ATAGTC": 315, + "ATAGTG": 316, + "ATAGCA": 317, + "ATAGCT": 318, + "ATAGCC": 319, + "ATAGCG": 320, + "ATAGGA": 321, + "ATAGGT": 322, + "ATAGGC": 323, + "ATAGGG": 324, + "ATTAAA": 325, + "ATTAAT": 326, + "ATTAAC": 327, + "ATTAAG": 328, + "ATTATA": 329, + "ATTATT": 330, + "ATTATC": 331, + "ATTATG": 332, + "ATTACA": 333, + "ATTACT": 334, + "ATTACC": 335, + "ATTACG": 336, + "ATTAGA": 337, + "ATTAGT": 338, + "ATTAGC": 339, + "ATTAGG": 340, + "ATTTAA": 341, + "ATTTAT": 342, + "ATTTAC": 343, + "ATTTAG": 344, + "ATTTTA": 345, + "ATTTTT": 346, + "ATTTTC": 347, + "ATTTTG": 348, + "ATTTCA": 349, + "ATTTCT": 350, + "ATTTCC": 351, + "ATTTCG": 352, + "ATTTGA": 353, + "ATTTGT": 354, + "ATTTGC": 355, + "ATTTGG": 356, + "ATTCAA": 357, + "ATTCAT": 358, + "ATTCAC": 359, + "ATTCAG": 360, + "ATTCTA": 361, + "ATTCTT": 362, + "ATTCTC": 363, + "ATTCTG": 364, + "ATTCCA": 365, + "ATTCCT": 366, + "ATTCCC": 367, + "ATTCCG": 368, + "ATTCGA": 369, + "ATTCGT": 370, + "ATTCGC": 371, + "ATTCGG": 372, + "ATTGAA": 373, + "ATTGAT": 374, + "ATTGAC": 375, + "ATTGAG": 376, + "ATTGTA": 377, + "ATTGTT": 378, + "ATTGTC": 379, + "ATTGTG": 380, + "ATTGCA": 381, + "ATTGCT": 382, + "ATTGCC": 383, + "ATTGCG": 384, + "ATTGGA": 385, + "ATTGGT": 386, + "ATTGGC": 387, + "ATTGGG": 388, + "ATCAAA": 389, + "ATCAAT": 390, + "ATCAAC": 391, + "ATCAAG": 392, + "ATCATA": 393, + "ATCATT": 394, + "ATCATC": 395, + "ATCATG": 396, + "ATCACA": 397, + "ATCACT": 398, + "ATCACC": 399, + "ATCACG": 400, + "ATCAGA": 401, + "ATCAGT": 402, + "ATCAGC": 403, + "ATCAGG": 404, + "ATCTAA": 405, + "ATCTAT": 406, + "ATCTAC": 407, + "ATCTAG": 408, + "ATCTTA": 409, + "ATCTTT": 410, + "ATCTTC": 411, + "ATCTTG": 412, + "ATCTCA": 413, + "ATCTCT": 414, + "ATCTCC": 415, + "ATCTCG": 416, + "ATCTGA": 417, + "ATCTGT": 418, + "ATCTGC": 419, + "ATCTGG": 420, + "ATCCAA": 421, + "ATCCAT": 422, + "ATCCAC": 423, + "ATCCAG": 424, + "ATCCTA": 425, + "ATCCTT": 426, + "ATCCTC": 427, + "ATCCTG": 428, + "ATCCCA": 429, + "ATCCCT": 430, + "ATCCCC": 431, + "ATCCCG": 432, + "ATCCGA": 433, + "ATCCGT": 434, + "ATCCGC": 435, + "ATCCGG": 436, + "ATCGAA": 437, + "ATCGAT": 438, + "ATCGAC": 439, + "ATCGAG": 440, + "ATCGTA": 441, + "ATCGTT": 442, + "ATCGTC": 443, + "ATCGTG": 444, + "ATCGCA": 445, + "ATCGCT": 446, + "ATCGCC": 447, + "ATCGCG": 448, + "ATCGGA": 449, + "ATCGGT": 450, + "ATCGGC": 451, + "ATCGGG": 452, + "ATGAAA": 453, + "ATGAAT": 454, + "ATGAAC": 455, + "ATGAAG": 456, + "ATGATA": 457, + "ATGATT": 458, + "ATGATC": 459, + "ATGATG": 460, + "ATGACA": 461, + "ATGACT": 462, + "ATGACC": 463, + "ATGACG": 464, + "ATGAGA": 465, + "ATGAGT": 466, + "ATGAGC": 467, + "ATGAGG": 468, + "ATGTAA": 469, + "ATGTAT": 470, + "ATGTAC": 471, + "ATGTAG": 472, + "ATGTTA": 473, + "ATGTTT": 474, + "ATGTTC": 475, + "ATGTTG": 476, + "ATGTCA": 477, + "ATGTCT": 478, + "ATGTCC": 479, + "ATGTCG": 480, + "ATGTGA": 481, + "ATGTGT": 482, + "ATGTGC": 483, + "ATGTGG": 484, + "ATGCAA": 485, + "ATGCAT": 486, + "ATGCAC": 487, + "ATGCAG": 488, + "ATGCTA": 489, + "ATGCTT": 490, + "ATGCTC": 491, + "ATGCTG": 492, + "ATGCCA": 493, + "ATGCCT": 494, + "ATGCCC": 495, + "ATGCCG": 496, + "ATGCGA": 497, + "ATGCGT": 498, + "ATGCGC": 499, + "ATGCGG": 500, + "ATGGAA": 501, + "ATGGAT": 502, + "ATGGAC": 503, + "ATGGAG": 504, + "ATGGTA": 505, + "ATGGTT": 506, + "ATGGTC": 507, + "ATGGTG": 508, + "ATGGCA": 509, + "ATGGCT": 510, + "ATGGCC": 511, + "ATGGCG": 512, + "ATGGGA": 513, + "ATGGGT": 514, + "ATGGGC": 515, + "ATGGGG": 516, + "ACAAAA": 517, + "ACAAAT": 518, + "ACAAAC": 519, + "ACAAAG": 520, + "ACAATA": 521, + "ACAATT": 522, + "ACAATC": 523, + "ACAATG": 524, + "ACAACA": 525, + "ACAACT": 526, + "ACAACC": 527, + "ACAACG": 528, + "ACAAGA": 529, + "ACAAGT": 530, + "ACAAGC": 531, + "ACAAGG": 532, + "ACATAA": 533, + "ACATAT": 534, + "ACATAC": 535, + "ACATAG": 536, + "ACATTA": 537, + "ACATTT": 538, + "ACATTC": 539, + "ACATTG": 540, + "ACATCA": 541, + "ACATCT": 542, + "ACATCC": 543, + "ACATCG": 544, + "ACATGA": 545, + "ACATGT": 546, + "ACATGC": 547, + "ACATGG": 548, + "ACACAA": 549, + "ACACAT": 550, + "ACACAC": 551, + "ACACAG": 552, + "ACACTA": 553, + "ACACTT": 554, + "ACACTC": 555, + "ACACTG": 556, + "ACACCA": 557, + "ACACCT": 558, + "ACACCC": 559, + "ACACCG": 560, + "ACACGA": 561, + "ACACGT": 562, + "ACACGC": 563, + "ACACGG": 564, + "ACAGAA": 565, + "ACAGAT": 566, + "ACAGAC": 567, + "ACAGAG": 568, + "ACAGTA": 569, + "ACAGTT": 570, + "ACAGTC": 571, + "ACAGTG": 572, + "ACAGCA": 573, + "ACAGCT": 574, + "ACAGCC": 575, + "ACAGCG": 576, + "ACAGGA": 577, + "ACAGGT": 578, + "ACAGGC": 579, + "ACAGGG": 580, + "ACTAAA": 581, + "ACTAAT": 582, + "ACTAAC": 583, + "ACTAAG": 584, + "ACTATA": 585, + "ACTATT": 586, + "ACTATC": 587, + "ACTATG": 588, + "ACTACA": 589, + "ACTACT": 590, + "ACTACC": 591, + "ACTACG": 592, + "ACTAGA": 593, + "ACTAGT": 594, + "ACTAGC": 595, + "ACTAGG": 596, + "ACTTAA": 597, + "ACTTAT": 598, + "ACTTAC": 599, + "ACTTAG": 600, + "ACTTTA": 601, + "ACTTTT": 602, + "ACTTTC": 603, + "ACTTTG": 604, + "ACTTCA": 605, + "ACTTCT": 606, + "ACTTCC": 607, + "ACTTCG": 608, + "ACTTGA": 609, + "ACTTGT": 610, + "ACTTGC": 611, + "ACTTGG": 612, + "ACTCAA": 613, + "ACTCAT": 614, + "ACTCAC": 615, + "ACTCAG": 616, + "ACTCTA": 617, + "ACTCTT": 618, + "ACTCTC": 619, + "ACTCTG": 620, + "ACTCCA": 621, + "ACTCCT": 622, + "ACTCCC": 623, + "ACTCCG": 624, + "ACTCGA": 625, + "ACTCGT": 626, + "ACTCGC": 627, + "ACTCGG": 628, + "ACTGAA": 629, + "ACTGAT": 630, + "ACTGAC": 631, + "ACTGAG": 632, + "ACTGTA": 633, + "ACTGTT": 634, + "ACTGTC": 635, + "ACTGTG": 636, + "ACTGCA": 637, + "ACTGCT": 638, + "ACTGCC": 639, + "ACTGCG": 640, + "ACTGGA": 641, + "ACTGGT": 642, + "ACTGGC": 643, + "ACTGGG": 644, + "ACCAAA": 645, + "ACCAAT": 646, + "ACCAAC": 647, + "ACCAAG": 648, + "ACCATA": 649, + "ACCATT": 650, + "ACCATC": 651, + "ACCATG": 652, + "ACCACA": 653, + "ACCACT": 654, + "ACCACC": 655, + "ACCACG": 656, + "ACCAGA": 657, + "ACCAGT": 658, + "ACCAGC": 659, + "ACCAGG": 660, + "ACCTAA": 661, + "ACCTAT": 662, + "ACCTAC": 663, + "ACCTAG": 664, + "ACCTTA": 665, + "ACCTTT": 666, + "ACCTTC": 667, + "ACCTTG": 668, + "ACCTCA": 669, + "ACCTCT": 670, + "ACCTCC": 671, + "ACCTCG": 672, + "ACCTGA": 673, + "ACCTGT": 674, + "ACCTGC": 675, + "ACCTGG": 676, + "ACCCAA": 677, + "ACCCAT": 678, + "ACCCAC": 679, + "ACCCAG": 680, + "ACCCTA": 681, + "ACCCTT": 682, + "ACCCTC": 683, + "ACCCTG": 684, + "ACCCCA": 685, + "ACCCCT": 686, + "ACCCCC": 687, + "ACCCCG": 688, + "ACCCGA": 689, + "ACCCGT": 690, + "ACCCGC": 691, + "ACCCGG": 692, + "ACCGAA": 693, + "ACCGAT": 694, + "ACCGAC": 695, + "ACCGAG": 696, + "ACCGTA": 697, + "ACCGTT": 698, + "ACCGTC": 699, + "ACCGTG": 700, + "ACCGCA": 701, + "ACCGCT": 702, + "ACCGCC": 703, + "ACCGCG": 704, + "ACCGGA": 705, + "ACCGGT": 706, + "ACCGGC": 707, + "ACCGGG": 708, + "ACGAAA": 709, + "ACGAAT": 710, + "ACGAAC": 711, + "ACGAAG": 712, + "ACGATA": 713, + "ACGATT": 714, + "ACGATC": 715, + "ACGATG": 716, + "ACGACA": 717, + "ACGACT": 718, + "ACGACC": 719, + "ACGACG": 720, + "ACGAGA": 721, + "ACGAGT": 722, + "ACGAGC": 723, + "ACGAGG": 724, + "ACGTAA": 725, + "ACGTAT": 726, + "ACGTAC": 727, + "ACGTAG": 728, + "ACGTTA": 729, + "ACGTTT": 730, + "ACGTTC": 731, + "ACGTTG": 732, + "ACGTCA": 733, + "ACGTCT": 734, + "ACGTCC": 735, + "ACGTCG": 736, + "ACGTGA": 737, + "ACGTGT": 738, + "ACGTGC": 739, + "ACGTGG": 740, + "ACGCAA": 741, + "ACGCAT": 742, + "ACGCAC": 743, + "ACGCAG": 744, + "ACGCTA": 745, + "ACGCTT": 746, + "ACGCTC": 747, + "ACGCTG": 748, + "ACGCCA": 749, + "ACGCCT": 750, + "ACGCCC": 751, + "ACGCCG": 752, + "ACGCGA": 753, + "ACGCGT": 754, + "ACGCGC": 755, + "ACGCGG": 756, + "ACGGAA": 757, + "ACGGAT": 758, + "ACGGAC": 759, + "ACGGAG": 760, + "ACGGTA": 761, + "ACGGTT": 762, + "ACGGTC": 763, + "ACGGTG": 764, + "ACGGCA": 765, + "ACGGCT": 766, + "ACGGCC": 767, + "ACGGCG": 768, + "ACGGGA": 769, + "ACGGGT": 770, + "ACGGGC": 771, + "ACGGGG": 772, + "AGAAAA": 773, + "AGAAAT": 774, + "AGAAAC": 775, + "AGAAAG": 776, + "AGAATA": 777, + "AGAATT": 778, + "AGAATC": 779, + "AGAATG": 780, + "AGAACA": 781, + "AGAACT": 782, + "AGAACC": 783, + "AGAACG": 784, + "AGAAGA": 785, + "AGAAGT": 786, + "AGAAGC": 787, + "AGAAGG": 788, + "AGATAA": 789, + "AGATAT": 790, + "AGATAC": 791, + "AGATAG": 792, + "AGATTA": 793, + "AGATTT": 794, + "AGATTC": 795, + "AGATTG": 796, + "AGATCA": 797, + "AGATCT": 798, + "AGATCC": 799, + "AGATCG": 800, + "AGATGA": 801, + "AGATGT": 802, + "AGATGC": 803, + "AGATGG": 804, + "AGACAA": 805, + "AGACAT": 806, + "AGACAC": 807, + "AGACAG": 808, + "AGACTA": 809, + "AGACTT": 810, + "AGACTC": 811, + "AGACTG": 812, + "AGACCA": 813, + "AGACCT": 814, + "AGACCC": 815, + "AGACCG": 816, + "AGACGA": 817, + "AGACGT": 818, + "AGACGC": 819, + "AGACGG": 820, + "AGAGAA": 821, + "AGAGAT": 822, + "AGAGAC": 823, + "AGAGAG": 824, + "AGAGTA": 825, + "AGAGTT": 826, + "AGAGTC": 827, + "AGAGTG": 828, + "AGAGCA": 829, + "AGAGCT": 830, + "AGAGCC": 831, + "AGAGCG": 832, + "AGAGGA": 833, + "AGAGGT": 834, + "AGAGGC": 835, + "AGAGGG": 836, + "AGTAAA": 837, + "AGTAAT": 838, + "AGTAAC": 839, + "AGTAAG": 840, + "AGTATA": 841, + "AGTATT": 842, + "AGTATC": 843, + "AGTATG": 844, + "AGTACA": 845, + "AGTACT": 846, + "AGTACC": 847, + "AGTACG": 848, + "AGTAGA": 849, + "AGTAGT": 850, + "AGTAGC": 851, + "AGTAGG": 852, + "AGTTAA": 853, + "AGTTAT": 854, + "AGTTAC": 855, + "AGTTAG": 856, + "AGTTTA": 857, + "AGTTTT": 858, + "AGTTTC": 859, + "AGTTTG": 860, + "AGTTCA": 861, + "AGTTCT": 862, + "AGTTCC": 863, + "AGTTCG": 864, + "AGTTGA": 865, + "AGTTGT": 866, + "AGTTGC": 867, + "AGTTGG": 868, + "AGTCAA": 869, + "AGTCAT": 870, + "AGTCAC": 871, + "AGTCAG": 872, + "AGTCTA": 873, + "AGTCTT": 874, + "AGTCTC": 875, + "AGTCTG": 876, + "AGTCCA": 877, + "AGTCCT": 878, + "AGTCCC": 879, + "AGTCCG": 880, + "AGTCGA": 881, + "AGTCGT": 882, + "AGTCGC": 883, + "AGTCGG": 884, + "AGTGAA": 885, + "AGTGAT": 886, + "AGTGAC": 887, + "AGTGAG": 888, + "AGTGTA": 889, + "AGTGTT": 890, + "AGTGTC": 891, + "AGTGTG": 892, + "AGTGCA": 893, + "AGTGCT": 894, + "AGTGCC": 895, + "AGTGCG": 896, + "AGTGGA": 897, + "AGTGGT": 898, + "AGTGGC": 899, + "AGTGGG": 900, + "AGCAAA": 901, + "AGCAAT": 902, + "AGCAAC": 903, + "AGCAAG": 904, + "AGCATA": 905, + "AGCATT": 906, + "AGCATC": 907, + "AGCATG": 908, + "AGCACA": 909, + "AGCACT": 910, + "AGCACC": 911, + "AGCACG": 912, + "AGCAGA": 913, + "AGCAGT": 914, + "AGCAGC": 915, + "AGCAGG": 916, + "AGCTAA": 917, + "AGCTAT": 918, + "AGCTAC": 919, + "AGCTAG": 920, + "AGCTTA": 921, + "AGCTTT": 922, + "AGCTTC": 923, + "AGCTTG": 924, + "AGCTCA": 925, + "AGCTCT": 926, + "AGCTCC": 927, + "AGCTCG": 928, + "AGCTGA": 929, + "AGCTGT": 930, + "AGCTGC": 931, + "AGCTGG": 932, + "AGCCAA": 933, + "AGCCAT": 934, + "AGCCAC": 935, + "AGCCAG": 936, + "AGCCTA": 937, + "AGCCTT": 938, + "AGCCTC": 939, + "AGCCTG": 940, + "AGCCCA": 941, + "AGCCCT": 942, + "AGCCCC": 943, + "AGCCCG": 944, + "AGCCGA": 945, + "AGCCGT": 946, + "AGCCGC": 947, + "AGCCGG": 948, + "AGCGAA": 949, + "AGCGAT": 950, + "AGCGAC": 951, + "AGCGAG": 952, + "AGCGTA": 953, + "AGCGTT": 954, + "AGCGTC": 955, + "AGCGTG": 956, + "AGCGCA": 957, + "AGCGCT": 958, + "AGCGCC": 959, + "AGCGCG": 960, + "AGCGGA": 961, + "AGCGGT": 962, + "AGCGGC": 963, + "AGCGGG": 964, + "AGGAAA": 965, + "AGGAAT": 966, + "AGGAAC": 967, + "AGGAAG": 968, + "AGGATA": 969, + "AGGATT": 970, + "AGGATC": 971, + "AGGATG": 972, + "AGGACA": 973, + "AGGACT": 974, + "AGGACC": 975, + "AGGACG": 976, + "AGGAGA": 977, + "AGGAGT": 978, + "AGGAGC": 979, + "AGGAGG": 980, + "AGGTAA": 981, + "AGGTAT": 982, + "AGGTAC": 983, + "AGGTAG": 984, + "AGGTTA": 985, + "AGGTTT": 986, + "AGGTTC": 987, + "AGGTTG": 988, + "AGGTCA": 989, + "AGGTCT": 990, + "AGGTCC": 991, + "AGGTCG": 992, + "AGGTGA": 993, + "AGGTGT": 994, + "AGGTGC": 995, + "AGGTGG": 996, + "AGGCAA": 997, + "AGGCAT": 998, + "AGGCAC": 999, + "AGGCAG": 1000, + "AGGCTA": 1001, + "AGGCTT": 1002, + "AGGCTC": 1003, + "AGGCTG": 1004, + "AGGCCA": 1005, + "AGGCCT": 1006, + "AGGCCC": 1007, + "AGGCCG": 1008, + "AGGCGA": 1009, + "AGGCGT": 1010, + "AGGCGC": 1011, + "AGGCGG": 1012, + "AGGGAA": 1013, + "AGGGAT": 1014, + "AGGGAC": 1015, + "AGGGAG": 1016, + "AGGGTA": 1017, + "AGGGTT": 1018, + "AGGGTC": 1019, + "AGGGTG": 1020, + "AGGGCA": 1021, + "AGGGCT": 1022, + "AGGGCC": 1023, + "AGGGCG": 1024, + "AGGGGA": 1025, + "AGGGGT": 1026, + "AGGGGC": 1027, + "AGGGGG": 1028, + "TAAAAA": 1029, + "TAAAAT": 1030, + "TAAAAC": 1031, + "TAAAAG": 1032, + "TAAATA": 1033, + "TAAATT": 1034, + "TAAATC": 1035, + "TAAATG": 1036, + "TAAACA": 1037, + "TAAACT": 1038, + "TAAACC": 1039, + "TAAACG": 1040, + "TAAAGA": 1041, + "TAAAGT": 1042, + "TAAAGC": 1043, + "TAAAGG": 1044, + "TAATAA": 1045, + "TAATAT": 1046, + "TAATAC": 1047, + "TAATAG": 1048, + "TAATTA": 1049, + "TAATTT": 1050, + "TAATTC": 1051, + "TAATTG": 1052, + "TAATCA": 1053, + "TAATCT": 1054, + "TAATCC": 1055, + "TAATCG": 1056, + "TAATGA": 1057, + "TAATGT": 1058, + "TAATGC": 1059, + "TAATGG": 1060, + "TAACAA": 1061, + "TAACAT": 1062, + "TAACAC": 1063, + "TAACAG": 1064, + "TAACTA": 1065, + "TAACTT": 1066, + "TAACTC": 1067, + "TAACTG": 1068, + "TAACCA": 1069, + "TAACCT": 1070, + "TAACCC": 1071, + "TAACCG": 1072, + "TAACGA": 1073, + "TAACGT": 1074, + "TAACGC": 1075, + "TAACGG": 1076, + "TAAGAA": 1077, + "TAAGAT": 1078, + "TAAGAC": 1079, + "TAAGAG": 1080, + "TAAGTA": 1081, + "TAAGTT": 1082, + "TAAGTC": 1083, + "TAAGTG": 1084, + "TAAGCA": 1085, + "TAAGCT": 1086, + "TAAGCC": 1087, + "TAAGCG": 1088, + "TAAGGA": 1089, + "TAAGGT": 1090, + "TAAGGC": 1091, + "TAAGGG": 1092, + "TATAAA": 1093, + "TATAAT": 1094, + "TATAAC": 1095, + "TATAAG": 1096, + "TATATA": 1097, + "TATATT": 1098, + "TATATC": 1099, + "TATATG": 1100, + "TATACA": 1101, + "TATACT": 1102, + "TATACC": 1103, + "TATACG": 1104, + "TATAGA": 1105, + "TATAGT": 1106, + "TATAGC": 1107, + "TATAGG": 1108, + "TATTAA": 1109, + "TATTAT": 1110, + "TATTAC": 1111, + "TATTAG": 1112, + "TATTTA": 1113, + "TATTTT": 1114, + "TATTTC": 1115, + "TATTTG": 1116, + "TATTCA": 1117, + "TATTCT": 1118, + "TATTCC": 1119, + "TATTCG": 1120, + "TATTGA": 1121, + "TATTGT": 1122, + "TATTGC": 1123, + "TATTGG": 1124, + "TATCAA": 1125, + "TATCAT": 1126, + "TATCAC": 1127, + "TATCAG": 1128, + "TATCTA": 1129, + "TATCTT": 1130, + "TATCTC": 1131, + "TATCTG": 1132, + "TATCCA": 1133, + "TATCCT": 1134, + "TATCCC": 1135, + "TATCCG": 1136, + "TATCGA": 1137, + "TATCGT": 1138, + "TATCGC": 1139, + "TATCGG": 1140, + "TATGAA": 1141, + "TATGAT": 1142, + "TATGAC": 1143, + "TATGAG": 1144, + "TATGTA": 1145, + "TATGTT": 1146, + "TATGTC": 1147, + "TATGTG": 1148, + "TATGCA": 1149, + "TATGCT": 1150, + "TATGCC": 1151, + "TATGCG": 1152, + "TATGGA": 1153, + "TATGGT": 1154, + "TATGGC": 1155, + "TATGGG": 1156, + "TACAAA": 1157, + "TACAAT": 1158, + "TACAAC": 1159, + "TACAAG": 1160, + "TACATA": 1161, + "TACATT": 1162, + "TACATC": 1163, + "TACATG": 1164, + "TACACA": 1165, + "TACACT": 1166, + "TACACC": 1167, + "TACACG": 1168, + "TACAGA": 1169, + "TACAGT": 1170, + "TACAGC": 1171, + "TACAGG": 1172, + "TACTAA": 1173, + "TACTAT": 1174, + "TACTAC": 1175, + "TACTAG": 1176, + "TACTTA": 1177, + "TACTTT": 1178, + "TACTTC": 1179, + "TACTTG": 1180, + "TACTCA": 1181, + "TACTCT": 1182, + "TACTCC": 1183, + "TACTCG": 1184, + "TACTGA": 1185, + "TACTGT": 1186, + "TACTGC": 1187, + "TACTGG": 1188, + "TACCAA": 1189, + "TACCAT": 1190, + "TACCAC": 1191, + "TACCAG": 1192, + "TACCTA": 1193, + "TACCTT": 1194, + "TACCTC": 1195, + "TACCTG": 1196, + "TACCCA": 1197, + "TACCCT": 1198, + "TACCCC": 1199, + "TACCCG": 1200, + "TACCGA": 1201, + "TACCGT": 1202, + "TACCGC": 1203, + "TACCGG": 1204, + "TACGAA": 1205, + "TACGAT": 1206, + "TACGAC": 1207, + "TACGAG": 1208, + "TACGTA": 1209, + "TACGTT": 1210, + "TACGTC": 1211, + "TACGTG": 1212, + "TACGCA": 1213, + "TACGCT": 1214, + "TACGCC": 1215, + "TACGCG": 1216, + "TACGGA": 1217, + "TACGGT": 1218, + "TACGGC": 1219, + "TACGGG": 1220, + "TAGAAA": 1221, + "TAGAAT": 1222, + "TAGAAC": 1223, + "TAGAAG": 1224, + "TAGATA": 1225, + "TAGATT": 1226, + "TAGATC": 1227, + "TAGATG": 1228, + "TAGACA": 1229, + "TAGACT": 1230, + "TAGACC": 1231, + "TAGACG": 1232, + "TAGAGA": 1233, + "TAGAGT": 1234, + "TAGAGC": 1235, + "TAGAGG": 1236, + "TAGTAA": 1237, + "TAGTAT": 1238, + "TAGTAC": 1239, + "TAGTAG": 1240, + "TAGTTA": 1241, + "TAGTTT": 1242, + "TAGTTC": 1243, + "TAGTTG": 1244, + "TAGTCA": 1245, + "TAGTCT": 1246, + "TAGTCC": 1247, + "TAGTCG": 1248, + "TAGTGA": 1249, + "TAGTGT": 1250, + "TAGTGC": 1251, + "TAGTGG": 1252, + "TAGCAA": 1253, + "TAGCAT": 1254, + "TAGCAC": 1255, + "TAGCAG": 1256, + "TAGCTA": 1257, + "TAGCTT": 1258, + "TAGCTC": 1259, + "TAGCTG": 1260, + "TAGCCA": 1261, + "TAGCCT": 1262, + "TAGCCC": 1263, + "TAGCCG": 1264, + "TAGCGA": 1265, + "TAGCGT": 1266, + "TAGCGC": 1267, + "TAGCGG": 1268, + "TAGGAA": 1269, + "TAGGAT": 1270, + "TAGGAC": 1271, + "TAGGAG": 1272, + "TAGGTA": 1273, + "TAGGTT": 1274, + "TAGGTC": 1275, + "TAGGTG": 1276, + "TAGGCA": 1277, + "TAGGCT": 1278, + "TAGGCC": 1279, + "TAGGCG": 1280, + "TAGGGA": 1281, + "TAGGGT": 1282, + "TAGGGC": 1283, + "TAGGGG": 1284, + "TTAAAA": 1285, + "TTAAAT": 1286, + "TTAAAC": 1287, + "TTAAAG": 1288, + "TTAATA": 1289, + "TTAATT": 1290, + "TTAATC": 1291, + "TTAATG": 1292, + "TTAACA": 1293, + "TTAACT": 1294, + "TTAACC": 1295, + "TTAACG": 1296, + "TTAAGA": 1297, + "TTAAGT": 1298, + "TTAAGC": 1299, + "TTAAGG": 1300, + "TTATAA": 1301, + "TTATAT": 1302, + "TTATAC": 1303, + "TTATAG": 1304, + "TTATTA": 1305, + "TTATTT": 1306, + "TTATTC": 1307, + "TTATTG": 1308, + "TTATCA": 1309, + "TTATCT": 1310, + "TTATCC": 1311, + "TTATCG": 1312, + "TTATGA": 1313, + "TTATGT": 1314, + "TTATGC": 1315, + "TTATGG": 1316, + "TTACAA": 1317, + "TTACAT": 1318, + "TTACAC": 1319, + "TTACAG": 1320, + "TTACTA": 1321, + "TTACTT": 1322, + "TTACTC": 1323, + "TTACTG": 1324, + "TTACCA": 1325, + "TTACCT": 1326, + "TTACCC": 1327, + "TTACCG": 1328, + "TTACGA": 1329, + "TTACGT": 1330, + "TTACGC": 1331, + "TTACGG": 1332, + "TTAGAA": 1333, + "TTAGAT": 1334, + "TTAGAC": 1335, + "TTAGAG": 1336, + "TTAGTA": 1337, + "TTAGTT": 1338, + "TTAGTC": 1339, + "TTAGTG": 1340, + "TTAGCA": 1341, + "TTAGCT": 1342, + "TTAGCC": 1343, + "TTAGCG": 1344, + "TTAGGA": 1345, + "TTAGGT": 1346, + "TTAGGC": 1347, + "TTAGGG": 1348, + "TTTAAA": 1349, + "TTTAAT": 1350, + "TTTAAC": 1351, + "TTTAAG": 1352, + "TTTATA": 1353, + "TTTATT": 1354, + "TTTATC": 1355, + "TTTATG": 1356, + "TTTACA": 1357, + "TTTACT": 1358, + "TTTACC": 1359, + "TTTACG": 1360, + "TTTAGA": 1361, + "TTTAGT": 1362, + "TTTAGC": 1363, + "TTTAGG": 1364, + "TTTTAA": 1365, + "TTTTAT": 1366, + "TTTTAC": 1367, + "TTTTAG": 1368, + "TTTTTA": 1369, + "TTTTTT": 1370, + "TTTTTC": 1371, + "TTTTTG": 1372, + "TTTTCA": 1373, + "TTTTCT": 1374, + "TTTTCC": 1375, + "TTTTCG": 1376, + "TTTTGA": 1377, + "TTTTGT": 1378, + "TTTTGC": 1379, + "TTTTGG": 1380, + "TTTCAA": 1381, + "TTTCAT": 1382, + "TTTCAC": 1383, + "TTTCAG": 1384, + "TTTCTA": 1385, + "TTTCTT": 1386, + "TTTCTC": 1387, + "TTTCTG": 1388, + "TTTCCA": 1389, + "TTTCCT": 1390, + "TTTCCC": 1391, + "TTTCCG": 1392, + "TTTCGA": 1393, + "TTTCGT": 1394, + "TTTCGC": 1395, + "TTTCGG": 1396, + "TTTGAA": 1397, + "TTTGAT": 1398, + "TTTGAC": 1399, + "TTTGAG": 1400, + "TTTGTA": 1401, + "TTTGTT": 1402, + "TTTGTC": 1403, + "TTTGTG": 1404, + "TTTGCA": 1405, + "TTTGCT": 1406, + "TTTGCC": 1407, + "TTTGCG": 1408, + "TTTGGA": 1409, + "TTTGGT": 1410, + "TTTGGC": 1411, + "TTTGGG": 1412, + "TTCAAA": 1413, + "TTCAAT": 1414, + "TTCAAC": 1415, + "TTCAAG": 1416, + "TTCATA": 1417, + "TTCATT": 1418, + "TTCATC": 1419, + "TTCATG": 1420, + "TTCACA": 1421, + "TTCACT": 1422, + "TTCACC": 1423, + "TTCACG": 1424, + "TTCAGA": 1425, + "TTCAGT": 1426, + "TTCAGC": 1427, + "TTCAGG": 1428, + "TTCTAA": 1429, + "TTCTAT": 1430, + "TTCTAC": 1431, + "TTCTAG": 1432, + "TTCTTA": 1433, + "TTCTTT": 1434, + "TTCTTC": 1435, + "TTCTTG": 1436, + "TTCTCA": 1437, + "TTCTCT": 1438, + "TTCTCC": 1439, + "TTCTCG": 1440, + "TTCTGA": 1441, + "TTCTGT": 1442, + "TTCTGC": 1443, + "TTCTGG": 1444, + "TTCCAA": 1445, + "TTCCAT": 1446, + "TTCCAC": 1447, + "TTCCAG": 1448, + "TTCCTA": 1449, + "TTCCTT": 1450, + "TTCCTC": 1451, + "TTCCTG": 1452, + "TTCCCA": 1453, + "TTCCCT": 1454, + "TTCCCC": 1455, + "TTCCCG": 1456, + "TTCCGA": 1457, + "TTCCGT": 1458, + "TTCCGC": 1459, + "TTCCGG": 1460, + "TTCGAA": 1461, + "TTCGAT": 1462, + "TTCGAC": 1463, + "TTCGAG": 1464, + "TTCGTA": 1465, + "TTCGTT": 1466, + "TTCGTC": 1467, + "TTCGTG": 1468, + "TTCGCA": 1469, + "TTCGCT": 1470, + "TTCGCC": 1471, + "TTCGCG": 1472, + "TTCGGA": 1473, + "TTCGGT": 1474, + "TTCGGC": 1475, + "TTCGGG": 1476, + "TTGAAA": 1477, + "TTGAAT": 1478, + "TTGAAC": 1479, + "TTGAAG": 1480, + "TTGATA": 1481, + "TTGATT": 1482, + "TTGATC": 1483, + "TTGATG": 1484, + "TTGACA": 1485, + "TTGACT": 1486, + "TTGACC": 1487, + "TTGACG": 1488, + "TTGAGA": 1489, + "TTGAGT": 1490, + "TTGAGC": 1491, + "TTGAGG": 1492, + "TTGTAA": 1493, + "TTGTAT": 1494, + "TTGTAC": 1495, + "TTGTAG": 1496, + "TTGTTA": 1497, + "TTGTTT": 1498, + "TTGTTC": 1499, + "TTGTTG": 1500, + "TTGTCA": 1501, + "TTGTCT": 1502, + "TTGTCC": 1503, + "TTGTCG": 1504, + "TTGTGA": 1505, + "TTGTGT": 1506, + "TTGTGC": 1507, + "TTGTGG": 1508, + "TTGCAA": 1509, + "TTGCAT": 1510, + "TTGCAC": 1511, + "TTGCAG": 1512, + "TTGCTA": 1513, + "TTGCTT": 1514, + "TTGCTC": 1515, + "TTGCTG": 1516, + "TTGCCA": 1517, + "TTGCCT": 1518, + "TTGCCC": 1519, + "TTGCCG": 1520, + "TTGCGA": 1521, + "TTGCGT": 1522, + "TTGCGC": 1523, + "TTGCGG": 1524, + "TTGGAA": 1525, + "TTGGAT": 1526, + "TTGGAC": 1527, + "TTGGAG": 1528, + "TTGGTA": 1529, + "TTGGTT": 1530, + "TTGGTC": 1531, + "TTGGTG": 1532, + "TTGGCA": 1533, + "TTGGCT": 1534, + "TTGGCC": 1535, + "TTGGCG": 1536, + "TTGGGA": 1537, + "TTGGGT": 1538, + "TTGGGC": 1539, + "TTGGGG": 1540, + "TCAAAA": 1541, + "TCAAAT": 1542, + "TCAAAC": 1543, + "TCAAAG": 1544, + "TCAATA": 1545, + "TCAATT": 1546, + "TCAATC": 1547, + "TCAATG": 1548, + "TCAACA": 1549, + "TCAACT": 1550, + "TCAACC": 1551, + "TCAACG": 1552, + "TCAAGA": 1553, + "TCAAGT": 1554, + "TCAAGC": 1555, + "TCAAGG": 1556, + "TCATAA": 1557, + "TCATAT": 1558, + "TCATAC": 1559, + "TCATAG": 1560, + "TCATTA": 1561, + "TCATTT": 1562, + "TCATTC": 1563, + "TCATTG": 1564, + "TCATCA": 1565, + "TCATCT": 1566, + "TCATCC": 1567, + "TCATCG": 1568, + "TCATGA": 1569, + "TCATGT": 1570, + "TCATGC": 1571, + "TCATGG": 1572, + "TCACAA": 1573, + "TCACAT": 1574, + "TCACAC": 1575, + "TCACAG": 1576, + "TCACTA": 1577, + "TCACTT": 1578, + "TCACTC": 1579, + "TCACTG": 1580, + "TCACCA": 1581, + "TCACCT": 1582, + "TCACCC": 1583, + "TCACCG": 1584, + "TCACGA": 1585, + "TCACGT": 1586, + "TCACGC": 1587, + "TCACGG": 1588, + "TCAGAA": 1589, + "TCAGAT": 1590, + "TCAGAC": 1591, + "TCAGAG": 1592, + "TCAGTA": 1593, + "TCAGTT": 1594, + "TCAGTC": 1595, + "TCAGTG": 1596, + "TCAGCA": 1597, + "TCAGCT": 1598, + "TCAGCC": 1599, + "TCAGCG": 1600, + "TCAGGA": 1601, + "TCAGGT": 1602, + "TCAGGC": 1603, + "TCAGGG": 1604, + "TCTAAA": 1605, + "TCTAAT": 1606, + "TCTAAC": 1607, + "TCTAAG": 1608, + "TCTATA": 1609, + "TCTATT": 1610, + "TCTATC": 1611, + "TCTATG": 1612, + "TCTACA": 1613, + "TCTACT": 1614, + "TCTACC": 1615, + "TCTACG": 1616, + "TCTAGA": 1617, + "TCTAGT": 1618, + "TCTAGC": 1619, + "TCTAGG": 1620, + "TCTTAA": 1621, + "TCTTAT": 1622, + "TCTTAC": 1623, + "TCTTAG": 1624, + "TCTTTA": 1625, + "TCTTTT": 1626, + "TCTTTC": 1627, + "TCTTTG": 1628, + "TCTTCA": 1629, + "TCTTCT": 1630, + "TCTTCC": 1631, + "TCTTCG": 1632, + "TCTTGA": 1633, + "TCTTGT": 1634, + "TCTTGC": 1635, + "TCTTGG": 1636, + "TCTCAA": 1637, + "TCTCAT": 1638, + "TCTCAC": 1639, + "TCTCAG": 1640, + "TCTCTA": 1641, + "TCTCTT": 1642, + "TCTCTC": 1643, + "TCTCTG": 1644, + "TCTCCA": 1645, + "TCTCCT": 1646, + "TCTCCC": 1647, + "TCTCCG": 1648, + "TCTCGA": 1649, + "TCTCGT": 1650, + "TCTCGC": 1651, + "TCTCGG": 1652, + "TCTGAA": 1653, + "TCTGAT": 1654, + "TCTGAC": 1655, + "TCTGAG": 1656, + "TCTGTA": 1657, + "TCTGTT": 1658, + "TCTGTC": 1659, + "TCTGTG": 1660, + "TCTGCA": 1661, + "TCTGCT": 1662, + "TCTGCC": 1663, + "TCTGCG": 1664, + "TCTGGA": 1665, + "TCTGGT": 1666, + "TCTGGC": 1667, + "TCTGGG": 1668, + "TCCAAA": 1669, + "TCCAAT": 1670, + "TCCAAC": 1671, + "TCCAAG": 1672, + "TCCATA": 1673, + "TCCATT": 1674, + "TCCATC": 1675, + "TCCATG": 1676, + "TCCACA": 1677, + "TCCACT": 1678, + "TCCACC": 1679, + "TCCACG": 1680, + "TCCAGA": 1681, + "TCCAGT": 1682, + "TCCAGC": 1683, + "TCCAGG": 1684, + "TCCTAA": 1685, + "TCCTAT": 1686, + "TCCTAC": 1687, + "TCCTAG": 1688, + "TCCTTA": 1689, + "TCCTTT": 1690, + "TCCTTC": 1691, + "TCCTTG": 1692, + "TCCTCA": 1693, + "TCCTCT": 1694, + "TCCTCC": 1695, + "TCCTCG": 1696, + "TCCTGA": 1697, + "TCCTGT": 1698, + "TCCTGC": 1699, + "TCCTGG": 1700, + "TCCCAA": 1701, + "TCCCAT": 1702, + "TCCCAC": 1703, + "TCCCAG": 1704, + "TCCCTA": 1705, + "TCCCTT": 1706, + "TCCCTC": 1707, + "TCCCTG": 1708, + "TCCCCA": 1709, + "TCCCCT": 1710, + "TCCCCC": 1711, + "TCCCCG": 1712, + "TCCCGA": 1713, + "TCCCGT": 1714, + "TCCCGC": 1715, + "TCCCGG": 1716, + "TCCGAA": 1717, + "TCCGAT": 1718, + "TCCGAC": 1719, + "TCCGAG": 1720, + "TCCGTA": 1721, + "TCCGTT": 1722, + "TCCGTC": 1723, + "TCCGTG": 1724, + "TCCGCA": 1725, + "TCCGCT": 1726, + "TCCGCC": 1727, + "TCCGCG": 1728, + "TCCGGA": 1729, + "TCCGGT": 1730, + "TCCGGC": 1731, + "TCCGGG": 1732, + "TCGAAA": 1733, + "TCGAAT": 1734, + "TCGAAC": 1735, + "TCGAAG": 1736, + "TCGATA": 1737, + "TCGATT": 1738, + "TCGATC": 1739, + "TCGATG": 1740, + "TCGACA": 1741, + "TCGACT": 1742, + "TCGACC": 1743, + "TCGACG": 1744, + "TCGAGA": 1745, + "TCGAGT": 1746, + "TCGAGC": 1747, + "TCGAGG": 1748, + "TCGTAA": 1749, + "TCGTAT": 1750, + "TCGTAC": 1751, + "TCGTAG": 1752, + "TCGTTA": 1753, + "TCGTTT": 1754, + "TCGTTC": 1755, + "TCGTTG": 1756, + "TCGTCA": 1757, + "TCGTCT": 1758, + "TCGTCC": 1759, + "TCGTCG": 1760, + "TCGTGA": 1761, + "TCGTGT": 1762, + "TCGTGC": 1763, + "TCGTGG": 1764, + "TCGCAA": 1765, + "TCGCAT": 1766, + "TCGCAC": 1767, + "TCGCAG": 1768, + "TCGCTA": 1769, + "TCGCTT": 1770, + "TCGCTC": 1771, + "TCGCTG": 1772, + "TCGCCA": 1773, + "TCGCCT": 1774, + "TCGCCC": 1775, + "TCGCCG": 1776, + "TCGCGA": 1777, + "TCGCGT": 1778, + "TCGCGC": 1779, + "TCGCGG": 1780, + "TCGGAA": 1781, + "TCGGAT": 1782, + "TCGGAC": 1783, + "TCGGAG": 1784, + "TCGGTA": 1785, + "TCGGTT": 1786, + "TCGGTC": 1787, + "TCGGTG": 1788, + "TCGGCA": 1789, + "TCGGCT": 1790, + "TCGGCC": 1791, + "TCGGCG": 1792, + "TCGGGA": 1793, + "TCGGGT": 1794, + "TCGGGC": 1795, + "TCGGGG": 1796, + "TGAAAA": 1797, + "TGAAAT": 1798, + "TGAAAC": 1799, + "TGAAAG": 1800, + "TGAATA": 1801, + "TGAATT": 1802, + "TGAATC": 1803, + "TGAATG": 1804, + "TGAACA": 1805, + "TGAACT": 1806, + "TGAACC": 1807, + "TGAACG": 1808, + "TGAAGA": 1809, + "TGAAGT": 1810, + "TGAAGC": 1811, + "TGAAGG": 1812, + "TGATAA": 1813, + "TGATAT": 1814, + "TGATAC": 1815, + "TGATAG": 1816, + "TGATTA": 1817, + "TGATTT": 1818, + "TGATTC": 1819, + "TGATTG": 1820, + "TGATCA": 1821, + "TGATCT": 1822, + "TGATCC": 1823, + "TGATCG": 1824, + "TGATGA": 1825, + "TGATGT": 1826, + "TGATGC": 1827, + "TGATGG": 1828, + "TGACAA": 1829, + "TGACAT": 1830, + "TGACAC": 1831, + "TGACAG": 1832, + "TGACTA": 1833, + "TGACTT": 1834, + "TGACTC": 1835, + "TGACTG": 1836, + "TGACCA": 1837, + "TGACCT": 1838, + "TGACCC": 1839, + "TGACCG": 1840, + "TGACGA": 1841, + "TGACGT": 1842, + "TGACGC": 1843, + "TGACGG": 1844, + "TGAGAA": 1845, + "TGAGAT": 1846, + "TGAGAC": 1847, + "TGAGAG": 1848, + "TGAGTA": 1849, + "TGAGTT": 1850, + "TGAGTC": 1851, + "TGAGTG": 1852, + "TGAGCA": 1853, + "TGAGCT": 1854, + "TGAGCC": 1855, + "TGAGCG": 1856, + "TGAGGA": 1857, + "TGAGGT": 1858, + "TGAGGC": 1859, + "TGAGGG": 1860, + "TGTAAA": 1861, + "TGTAAT": 1862, + "TGTAAC": 1863, + "TGTAAG": 1864, + "TGTATA": 1865, + "TGTATT": 1866, + "TGTATC": 1867, + "TGTATG": 1868, + "TGTACA": 1869, + "TGTACT": 1870, + "TGTACC": 1871, + "TGTACG": 1872, + "TGTAGA": 1873, + "TGTAGT": 1874, + "TGTAGC": 1875, + "TGTAGG": 1876, + "TGTTAA": 1877, + "TGTTAT": 1878, + "TGTTAC": 1879, + "TGTTAG": 1880, + "TGTTTA": 1881, + "TGTTTT": 1882, + "TGTTTC": 1883, + "TGTTTG": 1884, + "TGTTCA": 1885, + "TGTTCT": 1886, + "TGTTCC": 1887, + "TGTTCG": 1888, + "TGTTGA": 1889, + "TGTTGT": 1890, + "TGTTGC": 1891, + "TGTTGG": 1892, + "TGTCAA": 1893, + "TGTCAT": 1894, + "TGTCAC": 1895, + "TGTCAG": 1896, + "TGTCTA": 1897, + "TGTCTT": 1898, + "TGTCTC": 1899, + "TGTCTG": 1900, + "TGTCCA": 1901, + "TGTCCT": 1902, + "TGTCCC": 1903, + "TGTCCG": 1904, + "TGTCGA": 1905, + "TGTCGT": 1906, + "TGTCGC": 1907, + "TGTCGG": 1908, + "TGTGAA": 1909, + "TGTGAT": 1910, + "TGTGAC": 1911, + "TGTGAG": 1912, + "TGTGTA": 1913, + "TGTGTT": 1914, + "TGTGTC": 1915, + "TGTGTG": 1916, + "TGTGCA": 1917, + "TGTGCT": 1918, + "TGTGCC": 1919, + "TGTGCG": 1920, + "TGTGGA": 1921, + "TGTGGT": 1922, + "TGTGGC": 1923, + "TGTGGG": 1924, + "TGCAAA": 1925, + "TGCAAT": 1926, + "TGCAAC": 1927, + "TGCAAG": 1928, + "TGCATA": 1929, + "TGCATT": 1930, + "TGCATC": 1931, + "TGCATG": 1932, + "TGCACA": 1933, + "TGCACT": 1934, + "TGCACC": 1935, + "TGCACG": 1936, + "TGCAGA": 1937, + "TGCAGT": 1938, + "TGCAGC": 1939, + "TGCAGG": 1940, + "TGCTAA": 1941, + "TGCTAT": 1942, + "TGCTAC": 1943, + "TGCTAG": 1944, + "TGCTTA": 1945, + "TGCTTT": 1946, + "TGCTTC": 1947, + "TGCTTG": 1948, + "TGCTCA": 1949, + "TGCTCT": 1950, + "TGCTCC": 1951, + "TGCTCG": 1952, + "TGCTGA": 1953, + "TGCTGT": 1954, + "TGCTGC": 1955, + "TGCTGG": 1956, + "TGCCAA": 1957, + "TGCCAT": 1958, + "TGCCAC": 1959, + "TGCCAG": 1960, + "TGCCTA": 1961, + "TGCCTT": 1962, + "TGCCTC": 1963, + "TGCCTG": 1964, + "TGCCCA": 1965, + "TGCCCT": 1966, + "TGCCCC": 1967, + "TGCCCG": 1968, + "TGCCGA": 1969, + "TGCCGT": 1970, + "TGCCGC": 1971, + "TGCCGG": 1972, + "TGCGAA": 1973, + "TGCGAT": 1974, + "TGCGAC": 1975, + "TGCGAG": 1976, + "TGCGTA": 1977, + "TGCGTT": 1978, + "TGCGTC": 1979, + "TGCGTG": 1980, + "TGCGCA": 1981, + "TGCGCT": 1982, + "TGCGCC": 1983, + "TGCGCG": 1984, + "TGCGGA": 1985, + "TGCGGT": 1986, + "TGCGGC": 1987, + "TGCGGG": 1988, + "TGGAAA": 1989, + "TGGAAT": 1990, + "TGGAAC": 1991, + "TGGAAG": 1992, + "TGGATA": 1993, + "TGGATT": 1994, + "TGGATC": 1995, + "TGGATG": 1996, + "TGGACA": 1997, + "TGGACT": 1998, + "TGGACC": 1999, + "TGGACG": 2000, + "TGGAGA": 2001, + "TGGAGT": 2002, + "TGGAGC": 2003, + "TGGAGG": 2004, + "TGGTAA": 2005, + "TGGTAT": 2006, + "TGGTAC": 2007, + "TGGTAG": 2008, + "TGGTTA": 2009, + "TGGTTT": 2010, + "TGGTTC": 2011, + "TGGTTG": 2012, + "TGGTCA": 2013, + "TGGTCT": 2014, + "TGGTCC": 2015, + "TGGTCG": 2016, + "TGGTGA": 2017, + "TGGTGT": 2018, + "TGGTGC": 2019, + "TGGTGG": 2020, + "TGGCAA": 2021, + "TGGCAT": 2022, + "TGGCAC": 2023, + "TGGCAG": 2024, + "TGGCTA": 2025, + "TGGCTT": 2026, + "TGGCTC": 2027, + "TGGCTG": 2028, + "TGGCCA": 2029, + "TGGCCT": 2030, + "TGGCCC": 2031, + "TGGCCG": 2032, + "TGGCGA": 2033, + "TGGCGT": 2034, + "TGGCGC": 2035, + "TGGCGG": 2036, + "TGGGAA": 2037, + "TGGGAT": 2038, + "TGGGAC": 2039, + "TGGGAG": 2040, + "TGGGTA": 2041, + "TGGGTT": 2042, + "TGGGTC": 2043, + "TGGGTG": 2044, + "TGGGCA": 2045, + "TGGGCT": 2046, + "TGGGCC": 2047, + "TGGGCG": 2048, + "TGGGGA": 2049, + "TGGGGT": 2050, + "TGGGGC": 2051, + "TGGGGG": 2052, + "CAAAAA": 2053, + "CAAAAT": 2054, + "CAAAAC": 2055, + "CAAAAG": 2056, + "CAAATA": 2057, + "CAAATT": 2058, + "CAAATC": 2059, + "CAAATG": 2060, + "CAAACA": 2061, + "CAAACT": 2062, + "CAAACC": 2063, + "CAAACG": 2064, + "CAAAGA": 2065, + "CAAAGT": 2066, + "CAAAGC": 2067, + "CAAAGG": 2068, + "CAATAA": 2069, + "CAATAT": 2070, + "CAATAC": 2071, + "CAATAG": 2072, + "CAATTA": 2073, + "CAATTT": 2074, + "CAATTC": 2075, + "CAATTG": 2076, + "CAATCA": 2077, + "CAATCT": 2078, + "CAATCC": 2079, + "CAATCG": 2080, + "CAATGA": 2081, + "CAATGT": 2082, + "CAATGC": 2083, + "CAATGG": 2084, + "CAACAA": 2085, + "CAACAT": 2086, + "CAACAC": 2087, + "CAACAG": 2088, + "CAACTA": 2089, + "CAACTT": 2090, + "CAACTC": 2091, + "CAACTG": 2092, + "CAACCA": 2093, + "CAACCT": 2094, + "CAACCC": 2095, + "CAACCG": 2096, + "CAACGA": 2097, + "CAACGT": 2098, + "CAACGC": 2099, + "CAACGG": 2100, + "CAAGAA": 2101, + "CAAGAT": 2102, + "CAAGAC": 2103, + "CAAGAG": 2104, + "CAAGTA": 2105, + "CAAGTT": 2106, + "CAAGTC": 2107, + "CAAGTG": 2108, + "CAAGCA": 2109, + "CAAGCT": 2110, + "CAAGCC": 2111, + "CAAGCG": 2112, + "CAAGGA": 2113, + "CAAGGT": 2114, + "CAAGGC": 2115, + "CAAGGG": 2116, + "CATAAA": 2117, + "CATAAT": 2118, + "CATAAC": 2119, + "CATAAG": 2120, + "CATATA": 2121, + "CATATT": 2122, + "CATATC": 2123, + "CATATG": 2124, + "CATACA": 2125, + "CATACT": 2126, + "CATACC": 2127, + "CATACG": 2128, + "CATAGA": 2129, + "CATAGT": 2130, + "CATAGC": 2131, + "CATAGG": 2132, + "CATTAA": 2133, + "CATTAT": 2134, + "CATTAC": 2135, + "CATTAG": 2136, + "CATTTA": 2137, + "CATTTT": 2138, + "CATTTC": 2139, + "CATTTG": 2140, + "CATTCA": 2141, + "CATTCT": 2142, + "CATTCC": 2143, + "CATTCG": 2144, + "CATTGA": 2145, + "CATTGT": 2146, + "CATTGC": 2147, + "CATTGG": 2148, + "CATCAA": 2149, + "CATCAT": 2150, + "CATCAC": 2151, + "CATCAG": 2152, + "CATCTA": 2153, + "CATCTT": 2154, + "CATCTC": 2155, + "CATCTG": 2156, + "CATCCA": 2157, + "CATCCT": 2158, + "CATCCC": 2159, + "CATCCG": 2160, + "CATCGA": 2161, + "CATCGT": 2162, + "CATCGC": 2163, + "CATCGG": 2164, + "CATGAA": 2165, + "CATGAT": 2166, + "CATGAC": 2167, + "CATGAG": 2168, + "CATGTA": 2169, + "CATGTT": 2170, + "CATGTC": 2171, + "CATGTG": 2172, + "CATGCA": 2173, + "CATGCT": 2174, + "CATGCC": 2175, + "CATGCG": 2176, + "CATGGA": 2177, + "CATGGT": 2178, + "CATGGC": 2179, + "CATGGG": 2180, + "CACAAA": 2181, + "CACAAT": 2182, + "CACAAC": 2183, + "CACAAG": 2184, + "CACATA": 2185, + "CACATT": 2186, + "CACATC": 2187, + "CACATG": 2188, + "CACACA": 2189, + "CACACT": 2190, + "CACACC": 2191, + "CACACG": 2192, + "CACAGA": 2193, + "CACAGT": 2194, + "CACAGC": 2195, + "CACAGG": 2196, + "CACTAA": 2197, + "CACTAT": 2198, + "CACTAC": 2199, + "CACTAG": 2200, + "CACTTA": 2201, + "CACTTT": 2202, + "CACTTC": 2203, + "CACTTG": 2204, + "CACTCA": 2205, + "CACTCT": 2206, + "CACTCC": 2207, + "CACTCG": 2208, + "CACTGA": 2209, + "CACTGT": 2210, + "CACTGC": 2211, + "CACTGG": 2212, + "CACCAA": 2213, + "CACCAT": 2214, + "CACCAC": 2215, + "CACCAG": 2216, + "CACCTA": 2217, + "CACCTT": 2218, + "CACCTC": 2219, + "CACCTG": 2220, + "CACCCA": 2221, + "CACCCT": 2222, + "CACCCC": 2223, + "CACCCG": 2224, + "CACCGA": 2225, + "CACCGT": 2226, + "CACCGC": 2227, + "CACCGG": 2228, + "CACGAA": 2229, + "CACGAT": 2230, + "CACGAC": 2231, + "CACGAG": 2232, + "CACGTA": 2233, + "CACGTT": 2234, + "CACGTC": 2235, + "CACGTG": 2236, + "CACGCA": 2237, + "CACGCT": 2238, + "CACGCC": 2239, + "CACGCG": 2240, + "CACGGA": 2241, + "CACGGT": 2242, + "CACGGC": 2243, + "CACGGG": 2244, + "CAGAAA": 2245, + "CAGAAT": 2246, + "CAGAAC": 2247, + "CAGAAG": 2248, + "CAGATA": 2249, + "CAGATT": 2250, + "CAGATC": 2251, + "CAGATG": 2252, + "CAGACA": 2253, + "CAGACT": 2254, + "CAGACC": 2255, + "CAGACG": 2256, + "CAGAGA": 2257, + "CAGAGT": 2258, + "CAGAGC": 2259, + "CAGAGG": 2260, + "CAGTAA": 2261, + "CAGTAT": 2262, + "CAGTAC": 2263, + "CAGTAG": 2264, + "CAGTTA": 2265, + "CAGTTT": 2266, + "CAGTTC": 2267, + "CAGTTG": 2268, + "CAGTCA": 2269, + "CAGTCT": 2270, + "CAGTCC": 2271, + "CAGTCG": 2272, + "CAGTGA": 2273, + "CAGTGT": 2274, + "CAGTGC": 2275, + "CAGTGG": 2276, + "CAGCAA": 2277, + "CAGCAT": 2278, + "CAGCAC": 2279, + "CAGCAG": 2280, + "CAGCTA": 2281, + "CAGCTT": 2282, + "CAGCTC": 2283, + "CAGCTG": 2284, + "CAGCCA": 2285, + "CAGCCT": 2286, + "CAGCCC": 2287, + "CAGCCG": 2288, + "CAGCGA": 2289, + "CAGCGT": 2290, + "CAGCGC": 2291, + "CAGCGG": 2292, + "CAGGAA": 2293, + "CAGGAT": 2294, + "CAGGAC": 2295, + "CAGGAG": 2296, + "CAGGTA": 2297, + "CAGGTT": 2298, + "CAGGTC": 2299, + "CAGGTG": 2300, + "CAGGCA": 2301, + "CAGGCT": 2302, + "CAGGCC": 2303, + "CAGGCG": 2304, + "CAGGGA": 2305, + "CAGGGT": 2306, + "CAGGGC": 2307, + "CAGGGG": 2308, + "CTAAAA": 2309, + "CTAAAT": 2310, + "CTAAAC": 2311, + "CTAAAG": 2312, + "CTAATA": 2313, + "CTAATT": 2314, + "CTAATC": 2315, + "CTAATG": 2316, + "CTAACA": 2317, + "CTAACT": 2318, + "CTAACC": 2319, + "CTAACG": 2320, + "CTAAGA": 2321, + "CTAAGT": 2322, + "CTAAGC": 2323, + "CTAAGG": 2324, + "CTATAA": 2325, + "CTATAT": 2326, + "CTATAC": 2327, + "CTATAG": 2328, + "CTATTA": 2329, + "CTATTT": 2330, + "CTATTC": 2331, + "CTATTG": 2332, + "CTATCA": 2333, + "CTATCT": 2334, + "CTATCC": 2335, + "CTATCG": 2336, + "CTATGA": 2337, + "CTATGT": 2338, + "CTATGC": 2339, + "CTATGG": 2340, + "CTACAA": 2341, + "CTACAT": 2342, + "CTACAC": 2343, + "CTACAG": 2344, + "CTACTA": 2345, + "CTACTT": 2346, + "CTACTC": 2347, + "CTACTG": 2348, + "CTACCA": 2349, + "CTACCT": 2350, + "CTACCC": 2351, + "CTACCG": 2352, + "CTACGA": 2353, + "CTACGT": 2354, + "CTACGC": 2355, + "CTACGG": 2356, + "CTAGAA": 2357, + "CTAGAT": 2358, + "CTAGAC": 2359, + "CTAGAG": 2360, + "CTAGTA": 2361, + "CTAGTT": 2362, + "CTAGTC": 2363, + "CTAGTG": 2364, + "CTAGCA": 2365, + "CTAGCT": 2366, + "CTAGCC": 2367, + "CTAGCG": 2368, + "CTAGGA": 2369, + "CTAGGT": 2370, + "CTAGGC": 2371, + "CTAGGG": 2372, + "CTTAAA": 2373, + "CTTAAT": 2374, + "CTTAAC": 2375, + "CTTAAG": 2376, + "CTTATA": 2377, + "CTTATT": 2378, + "CTTATC": 2379, + "CTTATG": 2380, + "CTTACA": 2381, + "CTTACT": 2382, + "CTTACC": 2383, + "CTTACG": 2384, + "CTTAGA": 2385, + "CTTAGT": 2386, + "CTTAGC": 2387, + "CTTAGG": 2388, + "CTTTAA": 2389, + "CTTTAT": 2390, + "CTTTAC": 2391, + "CTTTAG": 2392, + "CTTTTA": 2393, + "CTTTTT": 2394, + "CTTTTC": 2395, + "CTTTTG": 2396, + "CTTTCA": 2397, + "CTTTCT": 2398, + "CTTTCC": 2399, + "CTTTCG": 2400, + "CTTTGA": 2401, + "CTTTGT": 2402, + "CTTTGC": 2403, + "CTTTGG": 2404, + "CTTCAA": 2405, + "CTTCAT": 2406, + "CTTCAC": 2407, + "CTTCAG": 2408, + "CTTCTA": 2409, + "CTTCTT": 2410, + "CTTCTC": 2411, + "CTTCTG": 2412, + "CTTCCA": 2413, + "CTTCCT": 2414, + "CTTCCC": 2415, + "CTTCCG": 2416, + "CTTCGA": 2417, + "CTTCGT": 2418, + "CTTCGC": 2419, + "CTTCGG": 2420, + "CTTGAA": 2421, + "CTTGAT": 2422, + "CTTGAC": 2423, + "CTTGAG": 2424, + "CTTGTA": 2425, + "CTTGTT": 2426, + "CTTGTC": 2427, + "CTTGTG": 2428, + "CTTGCA": 2429, + "CTTGCT": 2430, + "CTTGCC": 2431, + "CTTGCG": 2432, + "CTTGGA": 2433, + "CTTGGT": 2434, + "CTTGGC": 2435, + "CTTGGG": 2436, + "CTCAAA": 2437, + "CTCAAT": 2438, + "CTCAAC": 2439, + "CTCAAG": 2440, + "CTCATA": 2441, + "CTCATT": 2442, + "CTCATC": 2443, + "CTCATG": 2444, + "CTCACA": 2445, + "CTCACT": 2446, + "CTCACC": 2447, + "CTCACG": 2448, + "CTCAGA": 2449, + "CTCAGT": 2450, + "CTCAGC": 2451, + "CTCAGG": 2452, + "CTCTAA": 2453, + "CTCTAT": 2454, + "CTCTAC": 2455, + "CTCTAG": 2456, + "CTCTTA": 2457, + "CTCTTT": 2458, + "CTCTTC": 2459, + "CTCTTG": 2460, + "CTCTCA": 2461, + "CTCTCT": 2462, + "CTCTCC": 2463, + "CTCTCG": 2464, + "CTCTGA": 2465, + "CTCTGT": 2466, + "CTCTGC": 2467, + "CTCTGG": 2468, + "CTCCAA": 2469, + "CTCCAT": 2470, + "CTCCAC": 2471, + "CTCCAG": 2472, + "CTCCTA": 2473, + "CTCCTT": 2474, + "CTCCTC": 2475, + "CTCCTG": 2476, + "CTCCCA": 2477, + "CTCCCT": 2478, + "CTCCCC": 2479, + "CTCCCG": 2480, + "CTCCGA": 2481, + "CTCCGT": 2482, + "CTCCGC": 2483, + "CTCCGG": 2484, + "CTCGAA": 2485, + "CTCGAT": 2486, + "CTCGAC": 2487, + "CTCGAG": 2488, + "CTCGTA": 2489, + "CTCGTT": 2490, + "CTCGTC": 2491, + "CTCGTG": 2492, + "CTCGCA": 2493, + "CTCGCT": 2494, + "CTCGCC": 2495, + "CTCGCG": 2496, + "CTCGGA": 2497, + "CTCGGT": 2498, + "CTCGGC": 2499, + "CTCGGG": 2500, + "CTGAAA": 2501, + "CTGAAT": 2502, + "CTGAAC": 2503, + "CTGAAG": 2504, + "CTGATA": 2505, + "CTGATT": 2506, + "CTGATC": 2507, + "CTGATG": 2508, + "CTGACA": 2509, + "CTGACT": 2510, + "CTGACC": 2511, + "CTGACG": 2512, + "CTGAGA": 2513, + "CTGAGT": 2514, + "CTGAGC": 2515, + "CTGAGG": 2516, + "CTGTAA": 2517, + "CTGTAT": 2518, + "CTGTAC": 2519, + "CTGTAG": 2520, + "CTGTTA": 2521, + "CTGTTT": 2522, + "CTGTTC": 2523, + "CTGTTG": 2524, + "CTGTCA": 2525, + "CTGTCT": 2526, + "CTGTCC": 2527, + "CTGTCG": 2528, + "CTGTGA": 2529, + "CTGTGT": 2530, + "CTGTGC": 2531, + "CTGTGG": 2532, + "CTGCAA": 2533, + "CTGCAT": 2534, + "CTGCAC": 2535, + "CTGCAG": 2536, + "CTGCTA": 2537, + "CTGCTT": 2538, + "CTGCTC": 2539, + "CTGCTG": 2540, + "CTGCCA": 2541, + "CTGCCT": 2542, + "CTGCCC": 2543, + "CTGCCG": 2544, + "CTGCGA": 2545, + "CTGCGT": 2546, + "CTGCGC": 2547, + "CTGCGG": 2548, + "CTGGAA": 2549, + "CTGGAT": 2550, + "CTGGAC": 2551, + "CTGGAG": 2552, + "CTGGTA": 2553, + "CTGGTT": 2554, + "CTGGTC": 2555, + "CTGGTG": 2556, + "CTGGCA": 2557, + "CTGGCT": 2558, + "CTGGCC": 2559, + "CTGGCG": 2560, + "CTGGGA": 2561, + "CTGGGT": 2562, + "CTGGGC": 2563, + "CTGGGG": 2564, + "CCAAAA": 2565, + "CCAAAT": 2566, + "CCAAAC": 2567, + "CCAAAG": 2568, + "CCAATA": 2569, + "CCAATT": 2570, + "CCAATC": 2571, + "CCAATG": 2572, + "CCAACA": 2573, + "CCAACT": 2574, + "CCAACC": 2575, + "CCAACG": 2576, + "CCAAGA": 2577, + "CCAAGT": 2578, + "CCAAGC": 2579, + "CCAAGG": 2580, + "CCATAA": 2581, + "CCATAT": 2582, + "CCATAC": 2583, + "CCATAG": 2584, + "CCATTA": 2585, + "CCATTT": 2586, + "CCATTC": 2587, + "CCATTG": 2588, + "CCATCA": 2589, + "CCATCT": 2590, + "CCATCC": 2591, + "CCATCG": 2592, + "CCATGA": 2593, + "CCATGT": 2594, + "CCATGC": 2595, + "CCATGG": 2596, + "CCACAA": 2597, + "CCACAT": 2598, + "CCACAC": 2599, + "CCACAG": 2600, + "CCACTA": 2601, + "CCACTT": 2602, + "CCACTC": 2603, + "CCACTG": 2604, + "CCACCA": 2605, + "CCACCT": 2606, + "CCACCC": 2607, + "CCACCG": 2608, + "CCACGA": 2609, + "CCACGT": 2610, + "CCACGC": 2611, + "CCACGG": 2612, + "CCAGAA": 2613, + "CCAGAT": 2614, + "CCAGAC": 2615, + "CCAGAG": 2616, + "CCAGTA": 2617, + "CCAGTT": 2618, + "CCAGTC": 2619, + "CCAGTG": 2620, + "CCAGCA": 2621, + "CCAGCT": 2622, + "CCAGCC": 2623, + "CCAGCG": 2624, + "CCAGGA": 2625, + "CCAGGT": 2626, + "CCAGGC": 2627, + "CCAGGG": 2628, + "CCTAAA": 2629, + "CCTAAT": 2630, + "CCTAAC": 2631, + "CCTAAG": 2632, + "CCTATA": 2633, + "CCTATT": 2634, + "CCTATC": 2635, + "CCTATG": 2636, + "CCTACA": 2637, + "CCTACT": 2638, + "CCTACC": 2639, + "CCTACG": 2640, + "CCTAGA": 2641, + "CCTAGT": 2642, + "CCTAGC": 2643, + "CCTAGG": 2644, + "CCTTAA": 2645, + "CCTTAT": 2646, + "CCTTAC": 2647, + "CCTTAG": 2648, + "CCTTTA": 2649, + "CCTTTT": 2650, + "CCTTTC": 2651, + "CCTTTG": 2652, + "CCTTCA": 2653, + "CCTTCT": 2654, + "CCTTCC": 2655, + "CCTTCG": 2656, + "CCTTGA": 2657, + "CCTTGT": 2658, + "CCTTGC": 2659, + "CCTTGG": 2660, + "CCTCAA": 2661, + "CCTCAT": 2662, + "CCTCAC": 2663, + "CCTCAG": 2664, + "CCTCTA": 2665, + "CCTCTT": 2666, + "CCTCTC": 2667, + "CCTCTG": 2668, + "CCTCCA": 2669, + "CCTCCT": 2670, + "CCTCCC": 2671, + "CCTCCG": 2672, + "CCTCGA": 2673, + "CCTCGT": 2674, + "CCTCGC": 2675, + "CCTCGG": 2676, + "CCTGAA": 2677, + "CCTGAT": 2678, + "CCTGAC": 2679, + "CCTGAG": 2680, + "CCTGTA": 2681, + "CCTGTT": 2682, + "CCTGTC": 2683, + "CCTGTG": 2684, + "CCTGCA": 2685, + "CCTGCT": 2686, + "CCTGCC": 2687, + "CCTGCG": 2688, + "CCTGGA": 2689, + "CCTGGT": 2690, + "CCTGGC": 2691, + "CCTGGG": 2692, + "CCCAAA": 2693, + "CCCAAT": 2694, + "CCCAAC": 2695, + "CCCAAG": 2696, + "CCCATA": 2697, + "CCCATT": 2698, + "CCCATC": 2699, + "CCCATG": 2700, + "CCCACA": 2701, + "CCCACT": 2702, + "CCCACC": 2703, + "CCCACG": 2704, + "CCCAGA": 2705, + "CCCAGT": 2706, + "CCCAGC": 2707, + "CCCAGG": 2708, + "CCCTAA": 2709, + "CCCTAT": 2710, + "CCCTAC": 2711, + "CCCTAG": 2712, + "CCCTTA": 2713, + "CCCTTT": 2714, + "CCCTTC": 2715, + "CCCTTG": 2716, + "CCCTCA": 2717, + "CCCTCT": 2718, + "CCCTCC": 2719, + "CCCTCG": 2720, + "CCCTGA": 2721, + "CCCTGT": 2722, + "CCCTGC": 2723, + "CCCTGG": 2724, + "CCCCAA": 2725, + "CCCCAT": 2726, + "CCCCAC": 2727, + "CCCCAG": 2728, + "CCCCTA": 2729, + "CCCCTT": 2730, + "CCCCTC": 2731, + "CCCCTG": 2732, + "CCCCCA": 2733, + "CCCCCT": 2734, + "CCCCCC": 2735, + "CCCCCG": 2736, + "CCCCGA": 2737, + "CCCCGT": 2738, + "CCCCGC": 2739, + "CCCCGG": 2740, + "CCCGAA": 2741, + "CCCGAT": 2742, + "CCCGAC": 2743, + "CCCGAG": 2744, + "CCCGTA": 2745, + "CCCGTT": 2746, + "CCCGTC": 2747, + "CCCGTG": 2748, + "CCCGCA": 2749, + "CCCGCT": 2750, + "CCCGCC": 2751, + "CCCGCG": 2752, + "CCCGGA": 2753, + "CCCGGT": 2754, + "CCCGGC": 2755, + "CCCGGG": 2756, + "CCGAAA": 2757, + "CCGAAT": 2758, + "CCGAAC": 2759, + "CCGAAG": 2760, + "CCGATA": 2761, + "CCGATT": 2762, + "CCGATC": 2763, + "CCGATG": 2764, + "CCGACA": 2765, + "CCGACT": 2766, + "CCGACC": 2767, + "CCGACG": 2768, + "CCGAGA": 2769, + "CCGAGT": 2770, + "CCGAGC": 2771, + "CCGAGG": 2772, + "CCGTAA": 2773, + "CCGTAT": 2774, + "CCGTAC": 2775, + "CCGTAG": 2776, + "CCGTTA": 2777, + "CCGTTT": 2778, + "CCGTTC": 2779, + "CCGTTG": 2780, + "CCGTCA": 2781, + "CCGTCT": 2782, + "CCGTCC": 2783, + "CCGTCG": 2784, + "CCGTGA": 2785, + "CCGTGT": 2786, + "CCGTGC": 2787, + "CCGTGG": 2788, + "CCGCAA": 2789, + "CCGCAT": 2790, + "CCGCAC": 2791, + "CCGCAG": 2792, + "CCGCTA": 2793, + "CCGCTT": 2794, + "CCGCTC": 2795, + "CCGCTG": 2796, + "CCGCCA": 2797, + "CCGCCT": 2798, + "CCGCCC": 2799, + "CCGCCG": 2800, + "CCGCGA": 2801, + "CCGCGT": 2802, + "CCGCGC": 2803, + "CCGCGG": 2804, + "CCGGAA": 2805, + "CCGGAT": 2806, + "CCGGAC": 2807, + "CCGGAG": 2808, + "CCGGTA": 2809, + "CCGGTT": 2810, + "CCGGTC": 2811, + "CCGGTG": 2812, + "CCGGCA": 2813, + "CCGGCT": 2814, + "CCGGCC": 2815, + "CCGGCG": 2816, + "CCGGGA": 2817, + "CCGGGT": 2818, + "CCGGGC": 2819, + "CCGGGG": 2820, + "CGAAAA": 2821, + "CGAAAT": 2822, + "CGAAAC": 2823, + "CGAAAG": 2824, + "CGAATA": 2825, + "CGAATT": 2826, + "CGAATC": 2827, + "CGAATG": 2828, + "CGAACA": 2829, + "CGAACT": 2830, + "CGAACC": 2831, + "CGAACG": 2832, + "CGAAGA": 2833, + "CGAAGT": 2834, + "CGAAGC": 2835, + "CGAAGG": 2836, + "CGATAA": 2837, + "CGATAT": 2838, + "CGATAC": 2839, + "CGATAG": 2840, + "CGATTA": 2841, + "CGATTT": 2842, + "CGATTC": 2843, + "CGATTG": 2844, + "CGATCA": 2845, + "CGATCT": 2846, + "CGATCC": 2847, + "CGATCG": 2848, + "CGATGA": 2849, + "CGATGT": 2850, + "CGATGC": 2851, + "CGATGG": 2852, + "CGACAA": 2853, + "CGACAT": 2854, + "CGACAC": 2855, + "CGACAG": 2856, + "CGACTA": 2857, + "CGACTT": 2858, + "CGACTC": 2859, + "CGACTG": 2860, + "CGACCA": 2861, + "CGACCT": 2862, + "CGACCC": 2863, + "CGACCG": 2864, + "CGACGA": 2865, + "CGACGT": 2866, + "CGACGC": 2867, + "CGACGG": 2868, + "CGAGAA": 2869, + "CGAGAT": 2870, + "CGAGAC": 2871, + "CGAGAG": 2872, + "CGAGTA": 2873, + "CGAGTT": 2874, + "CGAGTC": 2875, + "CGAGTG": 2876, + "CGAGCA": 2877, + "CGAGCT": 2878, + "CGAGCC": 2879, + "CGAGCG": 2880, + "CGAGGA": 2881, + "CGAGGT": 2882, + "CGAGGC": 2883, + "CGAGGG": 2884, + "CGTAAA": 2885, + "CGTAAT": 2886, + "CGTAAC": 2887, + "CGTAAG": 2888, + "CGTATA": 2889, + "CGTATT": 2890, + "CGTATC": 2891, + "CGTATG": 2892, + "CGTACA": 2893, + "CGTACT": 2894, + "CGTACC": 2895, + "CGTACG": 2896, + "CGTAGA": 2897, + "CGTAGT": 2898, + "CGTAGC": 2899, + "CGTAGG": 2900, + "CGTTAA": 2901, + "CGTTAT": 2902, + "CGTTAC": 2903, + "CGTTAG": 2904, + "CGTTTA": 2905, + "CGTTTT": 2906, + "CGTTTC": 2907, + "CGTTTG": 2908, + "CGTTCA": 2909, + "CGTTCT": 2910, + "CGTTCC": 2911, + "CGTTCG": 2912, + "CGTTGA": 2913, + "CGTTGT": 2914, + "CGTTGC": 2915, + "CGTTGG": 2916, + "CGTCAA": 2917, + "CGTCAT": 2918, + "CGTCAC": 2919, + "CGTCAG": 2920, + "CGTCTA": 2921, + "CGTCTT": 2922, + "CGTCTC": 2923, + "CGTCTG": 2924, + "CGTCCA": 2925, + "CGTCCT": 2926, + "CGTCCC": 2927, + "CGTCCG": 2928, + "CGTCGA": 2929, + "CGTCGT": 2930, + "CGTCGC": 2931, + "CGTCGG": 2932, + "CGTGAA": 2933, + "CGTGAT": 2934, + "CGTGAC": 2935, + "CGTGAG": 2936, + "CGTGTA": 2937, + "CGTGTT": 2938, + "CGTGTC": 2939, + "CGTGTG": 2940, + "CGTGCA": 2941, + "CGTGCT": 2942, + "CGTGCC": 2943, + "CGTGCG": 2944, + "CGTGGA": 2945, + "CGTGGT": 2946, + "CGTGGC": 2947, + "CGTGGG": 2948, + "CGCAAA": 2949, + "CGCAAT": 2950, + "CGCAAC": 2951, + "CGCAAG": 2952, + "CGCATA": 2953, + "CGCATT": 2954, + "CGCATC": 2955, + "CGCATG": 2956, + "CGCACA": 2957, + "CGCACT": 2958, + "CGCACC": 2959, + "CGCACG": 2960, + "CGCAGA": 2961, + "CGCAGT": 2962, + "CGCAGC": 2963, + "CGCAGG": 2964, + "CGCTAA": 2965, + "CGCTAT": 2966, + "CGCTAC": 2967, + "CGCTAG": 2968, + "CGCTTA": 2969, + "CGCTTT": 2970, + "CGCTTC": 2971, + "CGCTTG": 2972, + "CGCTCA": 2973, + "CGCTCT": 2974, + "CGCTCC": 2975, + "CGCTCG": 2976, + "CGCTGA": 2977, + "CGCTGT": 2978, + "CGCTGC": 2979, + "CGCTGG": 2980, + "CGCCAA": 2981, + "CGCCAT": 2982, + "CGCCAC": 2983, + "CGCCAG": 2984, + "CGCCTA": 2985, + "CGCCTT": 2986, + "CGCCTC": 2987, + "CGCCTG": 2988, + "CGCCCA": 2989, + "CGCCCT": 2990, + "CGCCCC": 2991, + "CGCCCG": 2992, + "CGCCGA": 2993, + "CGCCGT": 2994, + "CGCCGC": 2995, + "CGCCGG": 2996, + "CGCGAA": 2997, + "CGCGAT": 2998, + "CGCGAC": 2999, + "CGCGAG": 3000, + "CGCGTA": 3001, + "CGCGTT": 3002, + "CGCGTC": 3003, + "CGCGTG": 3004, + "CGCGCA": 3005, + "CGCGCT": 3006, + "CGCGCC": 3007, + "CGCGCG": 3008, + "CGCGGA": 3009, + "CGCGGT": 3010, + "CGCGGC": 3011, + "CGCGGG": 3012, + "CGGAAA": 3013, + "CGGAAT": 3014, + "CGGAAC": 3015, + "CGGAAG": 3016, + "CGGATA": 3017, + "CGGATT": 3018, + "CGGATC": 3019, + "CGGATG": 3020, + "CGGACA": 3021, + "CGGACT": 3022, + "CGGACC": 3023, + "CGGACG": 3024, + "CGGAGA": 3025, + "CGGAGT": 3026, + "CGGAGC": 3027, + "CGGAGG": 3028, + "CGGTAA": 3029, + "CGGTAT": 3030, + "CGGTAC": 3031, + "CGGTAG": 3032, + "CGGTTA": 3033, + "CGGTTT": 3034, + "CGGTTC": 3035, + "CGGTTG": 3036, + "CGGTCA": 3037, + "CGGTCT": 3038, + "CGGTCC": 3039, + "CGGTCG": 3040, + "CGGTGA": 3041, + "CGGTGT": 3042, + "CGGTGC": 3043, + "CGGTGG": 3044, + "CGGCAA": 3045, + "CGGCAT": 3046, + "CGGCAC": 3047, + "CGGCAG": 3048, + "CGGCTA": 3049, + "CGGCTT": 3050, + "CGGCTC": 3051, + "CGGCTG": 3052, + "CGGCCA": 3053, + "CGGCCT": 3054, + "CGGCCC": 3055, + "CGGCCG": 3056, + "CGGCGA": 3057, + "CGGCGT": 3058, + "CGGCGC": 3059, + "CGGCGG": 3060, + "CGGGAA": 3061, + "CGGGAT": 3062, + "CGGGAC": 3063, + "CGGGAG": 3064, + "CGGGTA": 3065, + "CGGGTT": 3066, + "CGGGTC": 3067, + "CGGGTG": 3068, + "CGGGCA": 3069, + "CGGGCT": 3070, + "CGGGCC": 3071, + "CGGGCG": 3072, + "CGGGGA": 3073, + "CGGGGT": 3074, + "CGGGGC": 3075, + "CGGGGG": 3076, + "GAAAAA": 3077, + "GAAAAT": 3078, + "GAAAAC": 3079, + "GAAAAG": 3080, + "GAAATA": 3081, + "GAAATT": 3082, + "GAAATC": 3083, + "GAAATG": 3084, + "GAAACA": 3085, + "GAAACT": 3086, + "GAAACC": 3087, + "GAAACG": 3088, + "GAAAGA": 3089, + "GAAAGT": 3090, + "GAAAGC": 3091, + "GAAAGG": 3092, + "GAATAA": 3093, + "GAATAT": 3094, + "GAATAC": 3095, + "GAATAG": 3096, + "GAATTA": 3097, + "GAATTT": 3098, + "GAATTC": 3099, + "GAATTG": 3100, + "GAATCA": 3101, + "GAATCT": 3102, + "GAATCC": 3103, + "GAATCG": 3104, + "GAATGA": 3105, + "GAATGT": 3106, + "GAATGC": 3107, + "GAATGG": 3108, + "GAACAA": 3109, + "GAACAT": 3110, + "GAACAC": 3111, + "GAACAG": 3112, + "GAACTA": 3113, + "GAACTT": 3114, + "GAACTC": 3115, + "GAACTG": 3116, + "GAACCA": 3117, + "GAACCT": 3118, + "GAACCC": 3119, + "GAACCG": 3120, + "GAACGA": 3121, + "GAACGT": 3122, + "GAACGC": 3123, + "GAACGG": 3124, + "GAAGAA": 3125, + "GAAGAT": 3126, + "GAAGAC": 3127, + "GAAGAG": 3128, + "GAAGTA": 3129, + "GAAGTT": 3130, + "GAAGTC": 3131, + "GAAGTG": 3132, + "GAAGCA": 3133, + "GAAGCT": 3134, + "GAAGCC": 3135, + "GAAGCG": 3136, + "GAAGGA": 3137, + "GAAGGT": 3138, + "GAAGGC": 3139, + "GAAGGG": 3140, + "GATAAA": 3141, + "GATAAT": 3142, + "GATAAC": 3143, + "GATAAG": 3144, + "GATATA": 3145, + "GATATT": 3146, + "GATATC": 3147, + "GATATG": 3148, + "GATACA": 3149, + "GATACT": 3150, + "GATACC": 3151, + "GATACG": 3152, + "GATAGA": 3153, + "GATAGT": 3154, + "GATAGC": 3155, + "GATAGG": 3156, + "GATTAA": 3157, + "GATTAT": 3158, + "GATTAC": 3159, + "GATTAG": 3160, + "GATTTA": 3161, + "GATTTT": 3162, + "GATTTC": 3163, + "GATTTG": 3164, + "GATTCA": 3165, + "GATTCT": 3166, + "GATTCC": 3167, + "GATTCG": 3168, + "GATTGA": 3169, + "GATTGT": 3170, + "GATTGC": 3171, + "GATTGG": 3172, + "GATCAA": 3173, + "GATCAT": 3174, + "GATCAC": 3175, + "GATCAG": 3176, + "GATCTA": 3177, + "GATCTT": 3178, + "GATCTC": 3179, + "GATCTG": 3180, + "GATCCA": 3181, + "GATCCT": 3182, + "GATCCC": 3183, + "GATCCG": 3184, + "GATCGA": 3185, + "GATCGT": 3186, + "GATCGC": 3187, + "GATCGG": 3188, + "GATGAA": 3189, + "GATGAT": 3190, + "GATGAC": 3191, + "GATGAG": 3192, + "GATGTA": 3193, + "GATGTT": 3194, + "GATGTC": 3195, + "GATGTG": 3196, + "GATGCA": 3197, + "GATGCT": 3198, + "GATGCC": 3199, + "GATGCG": 3200, + "GATGGA": 3201, + "GATGGT": 3202, + "GATGGC": 3203, + "GATGGG": 3204, + "GACAAA": 3205, + "GACAAT": 3206, + "GACAAC": 3207, + "GACAAG": 3208, + "GACATA": 3209, + "GACATT": 3210, + "GACATC": 3211, + "GACATG": 3212, + "GACACA": 3213, + "GACACT": 3214, + "GACACC": 3215, + "GACACG": 3216, + "GACAGA": 3217, + "GACAGT": 3218, + "GACAGC": 3219, + "GACAGG": 3220, + "GACTAA": 3221, + "GACTAT": 3222, + "GACTAC": 3223, + "GACTAG": 3224, + "GACTTA": 3225, + "GACTTT": 3226, + "GACTTC": 3227, + "GACTTG": 3228, + "GACTCA": 3229, + "GACTCT": 3230, + "GACTCC": 3231, + "GACTCG": 3232, + "GACTGA": 3233, + "GACTGT": 3234, + "GACTGC": 3235, + "GACTGG": 3236, + "GACCAA": 3237, + "GACCAT": 3238, + "GACCAC": 3239, + "GACCAG": 3240, + "GACCTA": 3241, + "GACCTT": 3242, + "GACCTC": 3243, + "GACCTG": 3244, + "GACCCA": 3245, + "GACCCT": 3246, + "GACCCC": 3247, + "GACCCG": 3248, + "GACCGA": 3249, + "GACCGT": 3250, + "GACCGC": 3251, + "GACCGG": 3252, + "GACGAA": 3253, + "GACGAT": 3254, + "GACGAC": 3255, + "GACGAG": 3256, + "GACGTA": 3257, + "GACGTT": 3258, + "GACGTC": 3259, + "GACGTG": 3260, + "GACGCA": 3261, + "GACGCT": 3262, + "GACGCC": 3263, + "GACGCG": 3264, + "GACGGA": 3265, + "GACGGT": 3266, + "GACGGC": 3267, + "GACGGG": 3268, + "GAGAAA": 3269, + "GAGAAT": 3270, + "GAGAAC": 3271, + "GAGAAG": 3272, + "GAGATA": 3273, + "GAGATT": 3274, + "GAGATC": 3275, + "GAGATG": 3276, + "GAGACA": 3277, + "GAGACT": 3278, + "GAGACC": 3279, + "GAGACG": 3280, + "GAGAGA": 3281, + "GAGAGT": 3282, + "GAGAGC": 3283, + "GAGAGG": 3284, + "GAGTAA": 3285, + "GAGTAT": 3286, + "GAGTAC": 3287, + "GAGTAG": 3288, + "GAGTTA": 3289, + "GAGTTT": 3290, + "GAGTTC": 3291, + "GAGTTG": 3292, + "GAGTCA": 3293, + "GAGTCT": 3294, + "GAGTCC": 3295, + "GAGTCG": 3296, + "GAGTGA": 3297, + "GAGTGT": 3298, + "GAGTGC": 3299, + "GAGTGG": 3300, + "GAGCAA": 3301, + "GAGCAT": 3302, + "GAGCAC": 3303, + "GAGCAG": 3304, + "GAGCTA": 3305, + "GAGCTT": 3306, + "GAGCTC": 3307, + "GAGCTG": 3308, + "GAGCCA": 3309, + "GAGCCT": 3310, + "GAGCCC": 3311, + "GAGCCG": 3312, + "GAGCGA": 3313, + "GAGCGT": 3314, + "GAGCGC": 3315, + "GAGCGG": 3316, + "GAGGAA": 3317, + "GAGGAT": 3318, + "GAGGAC": 3319, + "GAGGAG": 3320, + "GAGGTA": 3321, + "GAGGTT": 3322, + "GAGGTC": 3323, + "GAGGTG": 3324, + "GAGGCA": 3325, + "GAGGCT": 3326, + "GAGGCC": 3327, + "GAGGCG": 3328, + "GAGGGA": 3329, + "GAGGGT": 3330, + "GAGGGC": 3331, + "GAGGGG": 3332, + "GTAAAA": 3333, + "GTAAAT": 3334, + "GTAAAC": 3335, + "GTAAAG": 3336, + "GTAATA": 3337, + "GTAATT": 3338, + "GTAATC": 3339, + "GTAATG": 3340, + "GTAACA": 3341, + "GTAACT": 3342, + "GTAACC": 3343, + "GTAACG": 3344, + "GTAAGA": 3345, + "GTAAGT": 3346, + "GTAAGC": 3347, + "GTAAGG": 3348, + "GTATAA": 3349, + "GTATAT": 3350, + "GTATAC": 3351, + "GTATAG": 3352, + "GTATTA": 3353, + "GTATTT": 3354, + "GTATTC": 3355, + "GTATTG": 3356, + "GTATCA": 3357, + "GTATCT": 3358, + "GTATCC": 3359, + "GTATCG": 3360, + "GTATGA": 3361, + "GTATGT": 3362, + "GTATGC": 3363, + "GTATGG": 3364, + "GTACAA": 3365, + "GTACAT": 3366, + "GTACAC": 3367, + "GTACAG": 3368, + "GTACTA": 3369, + "GTACTT": 3370, + "GTACTC": 3371, + "GTACTG": 3372, + "GTACCA": 3373, + "GTACCT": 3374, + "GTACCC": 3375, + "GTACCG": 3376, + "GTACGA": 3377, + "GTACGT": 3378, + "GTACGC": 3379, + "GTACGG": 3380, + "GTAGAA": 3381, + "GTAGAT": 3382, + "GTAGAC": 3383, + "GTAGAG": 3384, + "GTAGTA": 3385, + "GTAGTT": 3386, + "GTAGTC": 3387, + "GTAGTG": 3388, + "GTAGCA": 3389, + "GTAGCT": 3390, + "GTAGCC": 3391, + "GTAGCG": 3392, + "GTAGGA": 3393, + "GTAGGT": 3394, + "GTAGGC": 3395, + "GTAGGG": 3396, + "GTTAAA": 3397, + "GTTAAT": 3398, + "GTTAAC": 3399, + "GTTAAG": 3400, + "GTTATA": 3401, + "GTTATT": 3402, + "GTTATC": 3403, + "GTTATG": 3404, + "GTTACA": 3405, + "GTTACT": 3406, + "GTTACC": 3407, + "GTTACG": 3408, + "GTTAGA": 3409, + "GTTAGT": 3410, + "GTTAGC": 3411, + "GTTAGG": 3412, + "GTTTAA": 3413, + "GTTTAT": 3414, + "GTTTAC": 3415, + "GTTTAG": 3416, + "GTTTTA": 3417, + "GTTTTT": 3418, + "GTTTTC": 3419, + "GTTTTG": 3420, + "GTTTCA": 3421, + "GTTTCT": 3422, + "GTTTCC": 3423, + "GTTTCG": 3424, + "GTTTGA": 3425, + "GTTTGT": 3426, + "GTTTGC": 3427, + "GTTTGG": 3428, + "GTTCAA": 3429, + "GTTCAT": 3430, + "GTTCAC": 3431, + "GTTCAG": 3432, + "GTTCTA": 3433, + "GTTCTT": 3434, + "GTTCTC": 3435, + "GTTCTG": 3436, + "GTTCCA": 3437, + "GTTCCT": 3438, + "GTTCCC": 3439, + "GTTCCG": 3440, + "GTTCGA": 3441, + "GTTCGT": 3442, + "GTTCGC": 3443, + "GTTCGG": 3444, + "GTTGAA": 3445, + "GTTGAT": 3446, + "GTTGAC": 3447, + "GTTGAG": 3448, + "GTTGTA": 3449, + "GTTGTT": 3450, + "GTTGTC": 3451, + "GTTGTG": 3452, + "GTTGCA": 3453, + "GTTGCT": 3454, + "GTTGCC": 3455, + "GTTGCG": 3456, + "GTTGGA": 3457, + "GTTGGT": 3458, + "GTTGGC": 3459, + "GTTGGG": 3460, + "GTCAAA": 3461, + "GTCAAT": 3462, + "GTCAAC": 3463, + "GTCAAG": 3464, + "GTCATA": 3465, + "GTCATT": 3466, + "GTCATC": 3467, + "GTCATG": 3468, + "GTCACA": 3469, + "GTCACT": 3470, + "GTCACC": 3471, + "GTCACG": 3472, + "GTCAGA": 3473, + "GTCAGT": 3474, + "GTCAGC": 3475, + "GTCAGG": 3476, + "GTCTAA": 3477, + "GTCTAT": 3478, + "GTCTAC": 3479, + "GTCTAG": 3480, + "GTCTTA": 3481, + "GTCTTT": 3482, + "GTCTTC": 3483, + "GTCTTG": 3484, + "GTCTCA": 3485, + "GTCTCT": 3486, + "GTCTCC": 3487, + "GTCTCG": 3488, + "GTCTGA": 3489, + "GTCTGT": 3490, + "GTCTGC": 3491, + "GTCTGG": 3492, + "GTCCAA": 3493, + "GTCCAT": 3494, + "GTCCAC": 3495, + "GTCCAG": 3496, + "GTCCTA": 3497, + "GTCCTT": 3498, + "GTCCTC": 3499, + "GTCCTG": 3500, + "GTCCCA": 3501, + "GTCCCT": 3502, + "GTCCCC": 3503, + "GTCCCG": 3504, + "GTCCGA": 3505, + "GTCCGT": 3506, + "GTCCGC": 3507, + "GTCCGG": 3508, + "GTCGAA": 3509, + "GTCGAT": 3510, + "GTCGAC": 3511, + "GTCGAG": 3512, + "GTCGTA": 3513, + "GTCGTT": 3514, + "GTCGTC": 3515, + "GTCGTG": 3516, + "GTCGCA": 3517, + "GTCGCT": 3518, + "GTCGCC": 3519, + "GTCGCG": 3520, + "GTCGGA": 3521, + "GTCGGT": 3522, + "GTCGGC": 3523, + "GTCGGG": 3524, + "GTGAAA": 3525, + "GTGAAT": 3526, + "GTGAAC": 3527, + "GTGAAG": 3528, + "GTGATA": 3529, + "GTGATT": 3530, + "GTGATC": 3531, + "GTGATG": 3532, + "GTGACA": 3533, + "GTGACT": 3534, + "GTGACC": 3535, + "GTGACG": 3536, + "GTGAGA": 3537, + "GTGAGT": 3538, + "GTGAGC": 3539, + "GTGAGG": 3540, + "GTGTAA": 3541, + "GTGTAT": 3542, + "GTGTAC": 3543, + "GTGTAG": 3544, + "GTGTTA": 3545, + "GTGTTT": 3546, + "GTGTTC": 3547, + "GTGTTG": 3548, + "GTGTCA": 3549, + "GTGTCT": 3550, + "GTGTCC": 3551, + "GTGTCG": 3552, + "GTGTGA": 3553, + "GTGTGT": 3554, + "GTGTGC": 3555, + "GTGTGG": 3556, + "GTGCAA": 3557, + "GTGCAT": 3558, + "GTGCAC": 3559, + "GTGCAG": 3560, + "GTGCTA": 3561, + "GTGCTT": 3562, + "GTGCTC": 3563, + "GTGCTG": 3564, + "GTGCCA": 3565, + "GTGCCT": 3566, + "GTGCCC": 3567, + "GTGCCG": 3568, + "GTGCGA": 3569, + "GTGCGT": 3570, + "GTGCGC": 3571, + "GTGCGG": 3572, + "GTGGAA": 3573, + "GTGGAT": 3574, + "GTGGAC": 3575, + "GTGGAG": 3576, + "GTGGTA": 3577, + "GTGGTT": 3578, + "GTGGTC": 3579, + "GTGGTG": 3580, + "GTGGCA": 3581, + "GTGGCT": 3582, + "GTGGCC": 3583, + "GTGGCG": 3584, + "GTGGGA": 3585, + "GTGGGT": 3586, + "GTGGGC": 3587, + "GTGGGG": 3588, + "GCAAAA": 3589, + "GCAAAT": 3590, + "GCAAAC": 3591, + "GCAAAG": 3592, + "GCAATA": 3593, + "GCAATT": 3594, + "GCAATC": 3595, + "GCAATG": 3596, + "GCAACA": 3597, + "GCAACT": 3598, + "GCAACC": 3599, + "GCAACG": 3600, + "GCAAGA": 3601, + "GCAAGT": 3602, + "GCAAGC": 3603, + "GCAAGG": 3604, + "GCATAA": 3605, + "GCATAT": 3606, + "GCATAC": 3607, + "GCATAG": 3608, + "GCATTA": 3609, + "GCATTT": 3610, + "GCATTC": 3611, + "GCATTG": 3612, + "GCATCA": 3613, + "GCATCT": 3614, + "GCATCC": 3615, + "GCATCG": 3616, + "GCATGA": 3617, + "GCATGT": 3618, + "GCATGC": 3619, + "GCATGG": 3620, + "GCACAA": 3621, + "GCACAT": 3622, + "GCACAC": 3623, + "GCACAG": 3624, + "GCACTA": 3625, + "GCACTT": 3626, + "GCACTC": 3627, + "GCACTG": 3628, + "GCACCA": 3629, + "GCACCT": 3630, + "GCACCC": 3631, + "GCACCG": 3632, + "GCACGA": 3633, + "GCACGT": 3634, + "GCACGC": 3635, + "GCACGG": 3636, + "GCAGAA": 3637, + "GCAGAT": 3638, + "GCAGAC": 3639, + "GCAGAG": 3640, + "GCAGTA": 3641, + "GCAGTT": 3642, + "GCAGTC": 3643, + "GCAGTG": 3644, + "GCAGCA": 3645, + "GCAGCT": 3646, + "GCAGCC": 3647, + "GCAGCG": 3648, + "GCAGGA": 3649, + "GCAGGT": 3650, + "GCAGGC": 3651, + "GCAGGG": 3652, + "GCTAAA": 3653, + "GCTAAT": 3654, + "GCTAAC": 3655, + "GCTAAG": 3656, + "GCTATA": 3657, + "GCTATT": 3658, + "GCTATC": 3659, + "GCTATG": 3660, + "GCTACA": 3661, + "GCTACT": 3662, + "GCTACC": 3663, + "GCTACG": 3664, + "GCTAGA": 3665, + "GCTAGT": 3666, + "GCTAGC": 3667, + "GCTAGG": 3668, + "GCTTAA": 3669, + "GCTTAT": 3670, + "GCTTAC": 3671, + "GCTTAG": 3672, + "GCTTTA": 3673, + "GCTTTT": 3674, + "GCTTTC": 3675, + "GCTTTG": 3676, + "GCTTCA": 3677, + "GCTTCT": 3678, + "GCTTCC": 3679, + "GCTTCG": 3680, + "GCTTGA": 3681, + "GCTTGT": 3682, + "GCTTGC": 3683, + "GCTTGG": 3684, + "GCTCAA": 3685, + "GCTCAT": 3686, + "GCTCAC": 3687, + "GCTCAG": 3688, + "GCTCTA": 3689, + "GCTCTT": 3690, + "GCTCTC": 3691, + "GCTCTG": 3692, + "GCTCCA": 3693, + "GCTCCT": 3694, + "GCTCCC": 3695, + "GCTCCG": 3696, + "GCTCGA": 3697, + "GCTCGT": 3698, + "GCTCGC": 3699, + "GCTCGG": 3700, + "GCTGAA": 3701, + "GCTGAT": 3702, + "GCTGAC": 3703, + "GCTGAG": 3704, + "GCTGTA": 3705, + "GCTGTT": 3706, + "GCTGTC": 3707, + "GCTGTG": 3708, + "GCTGCA": 3709, + "GCTGCT": 3710, + "GCTGCC": 3711, + "GCTGCG": 3712, + "GCTGGA": 3713, + "GCTGGT": 3714, + "GCTGGC": 3715, + "GCTGGG": 3716, + "GCCAAA": 3717, + "GCCAAT": 3718, + "GCCAAC": 3719, + "GCCAAG": 3720, + "GCCATA": 3721, + "GCCATT": 3722, + "GCCATC": 3723, + "GCCATG": 3724, + "GCCACA": 3725, + "GCCACT": 3726, + "GCCACC": 3727, + "GCCACG": 3728, + "GCCAGA": 3729, + "GCCAGT": 3730, + "GCCAGC": 3731, + "GCCAGG": 3732, + "GCCTAA": 3733, + "GCCTAT": 3734, + "GCCTAC": 3735, + "GCCTAG": 3736, + "GCCTTA": 3737, + "GCCTTT": 3738, + "GCCTTC": 3739, + "GCCTTG": 3740, + "GCCTCA": 3741, + "GCCTCT": 3742, + "GCCTCC": 3743, + "GCCTCG": 3744, + "GCCTGA": 3745, + "GCCTGT": 3746, + "GCCTGC": 3747, + "GCCTGG": 3748, + "GCCCAA": 3749, + "GCCCAT": 3750, + "GCCCAC": 3751, + "GCCCAG": 3752, + "GCCCTA": 3753, + "GCCCTT": 3754, + "GCCCTC": 3755, + "GCCCTG": 3756, + "GCCCCA": 3757, + "GCCCCT": 3758, + "GCCCCC": 3759, + "GCCCCG": 3760, + "GCCCGA": 3761, + "GCCCGT": 3762, + "GCCCGC": 3763, + "GCCCGG": 3764, + "GCCGAA": 3765, + "GCCGAT": 3766, + "GCCGAC": 3767, + "GCCGAG": 3768, + "GCCGTA": 3769, + "GCCGTT": 3770, + "GCCGTC": 3771, + "GCCGTG": 3772, + "GCCGCA": 3773, + "GCCGCT": 3774, + "GCCGCC": 3775, + "GCCGCG": 3776, + "GCCGGA": 3777, + "GCCGGT": 3778, + "GCCGGC": 3779, + "GCCGGG": 3780, + "GCGAAA": 3781, + "GCGAAT": 3782, + "GCGAAC": 3783, + "GCGAAG": 3784, + "GCGATA": 3785, + "GCGATT": 3786, + "GCGATC": 3787, + "GCGATG": 3788, + "GCGACA": 3789, + "GCGACT": 3790, + "GCGACC": 3791, + "GCGACG": 3792, + "GCGAGA": 3793, + "GCGAGT": 3794, + "GCGAGC": 3795, + "GCGAGG": 3796, + "GCGTAA": 3797, + "GCGTAT": 3798, + "GCGTAC": 3799, + "GCGTAG": 3800, + "GCGTTA": 3801, + "GCGTTT": 3802, + "GCGTTC": 3803, + "GCGTTG": 3804, + "GCGTCA": 3805, + "GCGTCT": 3806, + "GCGTCC": 3807, + "GCGTCG": 3808, + "GCGTGA": 3809, + "GCGTGT": 3810, + "GCGTGC": 3811, + "GCGTGG": 3812, + "GCGCAA": 3813, + "GCGCAT": 3814, + "GCGCAC": 3815, + "GCGCAG": 3816, + "GCGCTA": 3817, + "GCGCTT": 3818, + "GCGCTC": 3819, + "GCGCTG": 3820, + "GCGCCA": 3821, + "GCGCCT": 3822, + "GCGCCC": 3823, + "GCGCCG": 3824, + "GCGCGA": 3825, + "GCGCGT": 3826, + "GCGCGC": 3827, + "GCGCGG": 3828, + "GCGGAA": 3829, + "GCGGAT": 3830, + "GCGGAC": 3831, + "GCGGAG": 3832, + "GCGGTA": 3833, + "GCGGTT": 3834, + "GCGGTC": 3835, + "GCGGTG": 3836, + "GCGGCA": 3837, + "GCGGCT": 3838, + "GCGGCC": 3839, + "GCGGCG": 3840, + "GCGGGA": 3841, + "GCGGGT": 3842, + "GCGGGC": 3843, + "GCGGGG": 3844, + "GGAAAA": 3845, + "GGAAAT": 3846, + "GGAAAC": 3847, + "GGAAAG": 3848, + "GGAATA": 3849, + "GGAATT": 3850, + "GGAATC": 3851, + "GGAATG": 3852, + "GGAACA": 3853, + "GGAACT": 3854, + "GGAACC": 3855, + "GGAACG": 3856, + "GGAAGA": 3857, + "GGAAGT": 3858, + "GGAAGC": 3859, + "GGAAGG": 3860, + "GGATAA": 3861, + "GGATAT": 3862, + "GGATAC": 3863, + "GGATAG": 3864, + "GGATTA": 3865, + "GGATTT": 3866, + "GGATTC": 3867, + "GGATTG": 3868, + "GGATCA": 3869, + "GGATCT": 3870, + "GGATCC": 3871, + "GGATCG": 3872, + "GGATGA": 3873, + "GGATGT": 3874, + "GGATGC": 3875, + "GGATGG": 3876, + "GGACAA": 3877, + "GGACAT": 3878, + "GGACAC": 3879, + "GGACAG": 3880, + "GGACTA": 3881, + "GGACTT": 3882, + "GGACTC": 3883, + "GGACTG": 3884, + "GGACCA": 3885, + "GGACCT": 3886, + "GGACCC": 3887, + "GGACCG": 3888, + "GGACGA": 3889, + "GGACGT": 3890, + "GGACGC": 3891, + "GGACGG": 3892, + "GGAGAA": 3893, + "GGAGAT": 3894, + "GGAGAC": 3895, + "GGAGAG": 3896, + "GGAGTA": 3897, + "GGAGTT": 3898, + "GGAGTC": 3899, + "GGAGTG": 3900, + "GGAGCA": 3901, + "GGAGCT": 3902, + "GGAGCC": 3903, + "GGAGCG": 3904, + "GGAGGA": 3905, + "GGAGGT": 3906, + "GGAGGC": 3907, + "GGAGGG": 3908, + "GGTAAA": 3909, + "GGTAAT": 3910, + "GGTAAC": 3911, + "GGTAAG": 3912, + "GGTATA": 3913, + "GGTATT": 3914, + "GGTATC": 3915, + "GGTATG": 3916, + "GGTACA": 3917, + "GGTACT": 3918, + "GGTACC": 3919, + "GGTACG": 3920, + "GGTAGA": 3921, + "GGTAGT": 3922, + "GGTAGC": 3923, + "GGTAGG": 3924, + "GGTTAA": 3925, + "GGTTAT": 3926, + "GGTTAC": 3927, + "GGTTAG": 3928, + "GGTTTA": 3929, + "GGTTTT": 3930, + "GGTTTC": 3931, + "GGTTTG": 3932, + "GGTTCA": 3933, + "GGTTCT": 3934, + "GGTTCC": 3935, + "GGTTCG": 3936, + "GGTTGA": 3937, + "GGTTGT": 3938, + "GGTTGC": 3939, + "GGTTGG": 3940, + "GGTCAA": 3941, + "GGTCAT": 3942, + "GGTCAC": 3943, + "GGTCAG": 3944, + "GGTCTA": 3945, + "GGTCTT": 3946, + "GGTCTC": 3947, + "GGTCTG": 3948, + "GGTCCA": 3949, + "GGTCCT": 3950, + "GGTCCC": 3951, + "GGTCCG": 3952, + "GGTCGA": 3953, + "GGTCGT": 3954, + "GGTCGC": 3955, + "GGTCGG": 3956, + "GGTGAA": 3957, + "GGTGAT": 3958, + "GGTGAC": 3959, + "GGTGAG": 3960, + "GGTGTA": 3961, + "GGTGTT": 3962, + "GGTGTC": 3963, + "GGTGTG": 3964, + "GGTGCA": 3965, + "GGTGCT": 3966, + "GGTGCC": 3967, + "GGTGCG": 3968, + "GGTGGA": 3969, + "GGTGGT": 3970, + "GGTGGC": 3971, + "GGTGGG": 3972, + "GGCAAA": 3973, + "GGCAAT": 3974, + "GGCAAC": 3975, + "GGCAAG": 3976, + "GGCATA": 3977, + "GGCATT": 3978, + "GGCATC": 3979, + "GGCATG": 3980, + "GGCACA": 3981, + "GGCACT": 3982, + "GGCACC": 3983, + "GGCACG": 3984, + "GGCAGA": 3985, + "GGCAGT": 3986, + "GGCAGC": 3987, + "GGCAGG": 3988, + "GGCTAA": 3989, + "GGCTAT": 3990, + "GGCTAC": 3991, + "GGCTAG": 3992, + "GGCTTA": 3993, + "GGCTTT": 3994, + "GGCTTC": 3995, + "GGCTTG": 3996, + "GGCTCA": 3997, + "GGCTCT": 3998, + "GGCTCC": 3999, + "GGCTCG": 4000, + "GGCTGA": 4001, + "GGCTGT": 4002, + "GGCTGC": 4003, + "GGCTGG": 4004, + "GGCCAA": 4005, + "GGCCAT": 4006, + "GGCCAC": 4007, + "GGCCAG": 4008, + "GGCCTA": 4009, + "GGCCTT": 4010, + "GGCCTC": 4011, + "GGCCTG": 4012, + "GGCCCA": 4013, + "GGCCCT": 4014, + "GGCCCC": 4015, + "GGCCCG": 4016, + "GGCCGA": 4017, + "GGCCGT": 4018, + "GGCCGC": 4019, + "GGCCGG": 4020, + "GGCGAA": 4021, + "GGCGAT": 4022, + "GGCGAC": 4023, + "GGCGAG": 4024, + "GGCGTA": 4025, + "GGCGTT": 4026, + "GGCGTC": 4027, + "GGCGTG": 4028, + "GGCGCA": 4029, + "GGCGCT": 4030, + "GGCGCC": 4031, + "GGCGCG": 4032, + "GGCGGA": 4033, + "GGCGGT": 4034, + "GGCGGC": 4035, + "GGCGGG": 4036, + "GGGAAA": 4037, + "GGGAAT": 4038, + "GGGAAC": 4039, + "GGGAAG": 4040, + "GGGATA": 4041, + "GGGATT": 4042, + "GGGATC": 4043, + "GGGATG": 4044, + "GGGACA": 4045, + "GGGACT": 4046, + "GGGACC": 4047, + "GGGACG": 4048, + "GGGAGA": 4049, + "GGGAGT": 4050, + "GGGAGC": 4051, + "GGGAGG": 4052, + "GGGTAA": 4053, + "GGGTAT": 4054, + "GGGTAC": 4055, + "GGGTAG": 4056, + "GGGTTA": 4057, + "GGGTTT": 4058, + "GGGTTC": 4059, + "GGGTTG": 4060, + "GGGTCA": 4061, + "GGGTCT": 4062, + "GGGTCC": 4063, + "GGGTCG": 4064, + "GGGTGA": 4065, + "GGGTGT": 4066, + "GGGTGC": 4067, + "GGGTGG": 4068, + "GGGCAA": 4069, + "GGGCAT": 4070, + "GGGCAC": 4071, + "GGGCAG": 4072, + "GGGCTA": 4073, + "GGGCTT": 4074, + "GGGCTC": 4075, + "GGGCTG": 4076, + "GGGCCA": 4077, + "GGGCCT": 4078, + "GGGCCC": 4079, + "GGGCCG": 4080, + "GGGCGA": 4081, + "GGGCGT": 4082, + "GGGCGC": 4083, + "GGGCGG": 4084, + "GGGGAA": 4085, + "GGGGAT": 4086, + "GGGGAC": 4087, + "GGGGAG": 4088, + "GGGGTA": 4089, + "GGGGTT": 4090, + "GGGGTC": 4091, + "GGGGTG": 4092, + "GGGGCA": 4093, + "GGGGCT": 4094, + "GGGGCC": 4095, + "GGGGCG": 4096, + "GGGGGA": 4097, + "GGGGGT": 4098, + "GGGGGC": 4099, + "GGGGGG": 4100 + } + } +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-4000/tokenizer_config.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-4000/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..604e2b1b29af1dcb3920d2828b6a90276e5e3d02 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-4000/tokenizer_config.json @@ -0,0 +1,58 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "[UNK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "[CLS]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "[SEP]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "4": { + "content": "[MASK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "clean_up_tokenization_spaces": true, + "cls_token": "[CLS]", + "do_basic_tokenize": true, + "do_lower_case": false, + "mask_token": "[MASK]", + "max_len": 512, + "model_max_length": 512, + "never_split": null, + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "strip_accents": null, + "tokenize_chinese_chars": true, + "tokenizer_class": "BertTokenizer", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-4000/trainer_state.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-4000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..2903b1d80107eb7d0cdb7949c77a0f9e9224b175 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-4000/trainer_state.json @@ -0,0 +1,1153 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 16.0, + "eval_steps": 500, + "global_step": 4000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.1, + "grad_norm": 0.6611918807029724, + "learning_rate": 5.319148936170213e-05, + "loss": 0.6932, + "step": 25 + }, + { + "epoch": 0.2, + "grad_norm": 0.9467485547065735, + "learning_rate": 0.00010638297872340425, + "loss": 0.6912, + "step": 50 + }, + { + "epoch": 0.3, + "grad_norm": 0.9465051889419556, + "learning_rate": 0.00015957446808510637, + "loss": 0.6893, + "step": 75 + }, + { + "epoch": 0.4, + "grad_norm": 1.5430934429168701, + "learning_rate": 0.0002127659574468085, + "loss": 0.6804, + "step": 100 + }, + { + "epoch": 0.5, + "grad_norm": 1.6103107929229736, + "learning_rate": 0.00026595744680851064, + "loss": 0.6806, + "step": 125 + }, + { + "epoch": 0.6, + "grad_norm": 2.3773200511932373, + "learning_rate": 0.00031914893617021275, + "loss": 0.6801, + "step": 150 + }, + { + "epoch": 0.7, + "grad_norm": 1.8832203149795532, + "learning_rate": 0.0003723404255319149, + "loss": 0.6791, + "step": 175 + }, + { + "epoch": 0.8, + "grad_norm": 1.3350876569747925, + "learning_rate": 0.0003992081821181128, + "loss": 0.6746, + "step": 200 + }, + { + "epoch": 0.9, + "grad_norm": 1.5796219110488892, + "learning_rate": 0.0003975585615308479, + "loss": 0.6771, + "step": 225 + }, + { + "epoch": 1.0, + "grad_norm": 1.884006381034851, + "learning_rate": 0.00039590894094358297, + "loss": 0.6649, + "step": 250 + }, + { + "epoch": 1.1, + "grad_norm": 3.286440372467041, + "learning_rate": 0.0003942593203563181, + "loss": 0.6388, + "step": 275 + }, + { + "epoch": 1.2, + "grad_norm": 8.244946479797363, + "learning_rate": 0.0003926096997690532, + "loss": 0.6238, + "step": 300 + }, + { + "epoch": 1.3, + "grad_norm": 4.265683650970459, + "learning_rate": 0.00039096007918178817, + "loss": 0.6383, + "step": 325 + }, + { + "epoch": 1.4, + "grad_norm": 4.115826606750488, + "learning_rate": 0.0003893104585945233, + "loss": 0.6194, + "step": 350 + }, + { + "epoch": 1.5, + "grad_norm": 5.694250583648682, + "learning_rate": 0.0003876608380072583, + "loss": 0.6324, + "step": 375 + }, + { + "epoch": 1.6, + "grad_norm": 3.463121175765991, + "learning_rate": 0.00038601121741999343, + "loss": 0.621, + "step": 400 + }, + { + "epoch": 1.7, + "grad_norm": 4.582865238189697, + "learning_rate": 0.0003843615968327285, + "loss": 0.6116, + "step": 425 + }, + { + "epoch": 1.8, + "grad_norm": 11.996281623840332, + "learning_rate": 0.0003827119762454636, + "loss": 0.6393, + "step": 450 + }, + { + "epoch": 1.9, + "grad_norm": 3.0407373905181885, + "learning_rate": 0.00038106235565819863, + "loss": 0.628, + "step": 475 + }, + { + "epoch": 2.0, + "grad_norm": 2.917588233947754, + "learning_rate": 0.0003794127350709337, + "loss": 0.6078, + "step": 500 + }, + { + "epoch": 2.1, + "grad_norm": 4.748379707336426, + "learning_rate": 0.0003777631144836688, + "loss": 0.4899, + "step": 525 + }, + { + "epoch": 2.2, + "grad_norm": 3.8076977729797363, + "learning_rate": 0.00037611349389640383, + "loss": 0.5086, + "step": 550 + }, + { + "epoch": 2.3, + "grad_norm": 5.2440714836120605, + "learning_rate": 0.00037446387330913894, + "loss": 0.5327, + "step": 575 + }, + { + "epoch": 2.4, + "grad_norm": 7.110438346862793, + "learning_rate": 0.000372814252721874, + "loss": 0.5436, + "step": 600 + }, + { + "epoch": 2.5, + "grad_norm": 5.46150541305542, + "learning_rate": 0.00037116463213460903, + "loss": 0.5294, + "step": 625 + }, + { + "epoch": 2.6, + "grad_norm": 5.136163234710693, + "learning_rate": 0.00036951501154734414, + "loss": 0.5245, + "step": 650 + }, + { + "epoch": 2.7, + "grad_norm": 8.735346794128418, + "learning_rate": 0.0003678653909600792, + "loss": 0.5449, + "step": 675 + }, + { + "epoch": 2.8, + "grad_norm": 2.922825574874878, + "learning_rate": 0.0003662157703728143, + "loss": 0.5406, + "step": 700 + }, + { + "epoch": 2.9, + "grad_norm": 7.744819641113281, + "learning_rate": 0.00036456614978554934, + "loss": 0.5447, + "step": 725 + }, + { + "epoch": 3.0, + "grad_norm": 11.6185884475708, + "learning_rate": 0.00036291652919828444, + "loss": 0.5195, + "step": 750 + }, + { + "epoch": 3.1, + "grad_norm": 3.73836088180542, + "learning_rate": 0.00036126690861101944, + "loss": 0.3824, + "step": 775 + }, + { + "epoch": 3.2, + "grad_norm": 14.850343704223633, + "learning_rate": 0.00035961728802375454, + "loss": 0.4071, + "step": 800 + }, + { + "epoch": 3.3, + "grad_norm": 5.7157440185546875, + "learning_rate": 0.0003579676674364896, + "loss": 0.3986, + "step": 825 + }, + { + "epoch": 3.4, + "grad_norm": 12.418399810791016, + "learning_rate": 0.0003563180468492247, + "loss": 0.4282, + "step": 850 + }, + { + "epoch": 3.5, + "grad_norm": 12.793001174926758, + "learning_rate": 0.0003546684262619598, + "loss": 0.4822, + "step": 875 + }, + { + "epoch": 3.6, + "grad_norm": 6.489450931549072, + "learning_rate": 0.00035301880567469485, + "loss": 0.4239, + "step": 900 + }, + { + "epoch": 3.7, + "grad_norm": 5.365822792053223, + "learning_rate": 0.0003513691850874299, + "loss": 0.421, + "step": 925 + }, + { + "epoch": 3.8, + "grad_norm": 12.643745422363281, + "learning_rate": 0.00034971956450016495, + "loss": 0.3964, + "step": 950 + }, + { + "epoch": 3.9, + "grad_norm": 14.334024429321289, + "learning_rate": 0.00034806994391290005, + "loss": 0.4634, + "step": 975 + }, + { + "epoch": 4.0, + "grad_norm": 6.819091320037842, + "learning_rate": 0.0003464203233256351, + "loss": 0.4139, + "step": 1000 + }, + { + "epoch": 4.1, + "grad_norm": 8.29238224029541, + "learning_rate": 0.0003447707027383702, + "loss": 0.2695, + "step": 1025 + }, + { + "epoch": 4.2, + "grad_norm": 5.984206676483154, + "learning_rate": 0.00034312108215110525, + "loss": 0.2653, + "step": 1050 + }, + { + "epoch": 4.3, + "grad_norm": 5.9425435066223145, + "learning_rate": 0.0003414714615638403, + "loss": 0.2982, + "step": 1075 + }, + { + "epoch": 4.4, + "grad_norm": 7.1877593994140625, + "learning_rate": 0.0003398218409765754, + "loss": 0.3309, + "step": 1100 + }, + { + "epoch": 4.5, + "grad_norm": 17.83046531677246, + "learning_rate": 0.00033817222038931045, + "loss": 0.3467, + "step": 1125 + }, + { + "epoch": 4.6, + "grad_norm": 4.865128517150879, + "learning_rate": 0.00033652259980204556, + "loss": 0.3117, + "step": 1150 + }, + { + "epoch": 4.7, + "grad_norm": 31.427154541015625, + "learning_rate": 0.0003348729792147806, + "loss": 0.3273, + "step": 1175 + }, + { + "epoch": 4.8, + "grad_norm": 26.77984619140625, + "learning_rate": 0.0003332233586275157, + "loss": 0.3504, + "step": 1200 + }, + { + "epoch": 4.9, + "grad_norm": 9.389993667602539, + "learning_rate": 0.00033157373804025076, + "loss": 0.3339, + "step": 1225 + }, + { + "epoch": 5.0, + "grad_norm": 25.70441246032715, + "learning_rate": 0.0003299241174529858, + "loss": 0.3336, + "step": 1250 + }, + { + "epoch": 5.1, + "grad_norm": 13.482085227966309, + "learning_rate": 0.0003282744968657209, + "loss": 0.2156, + "step": 1275 + }, + { + "epoch": 5.2, + "grad_norm": 8.725810050964355, + "learning_rate": 0.00032662487627845596, + "loss": 0.1811, + "step": 1300 + }, + { + "epoch": 5.3, + "grad_norm": 9.313215255737305, + "learning_rate": 0.00032497525569119106, + "loss": 0.2807, + "step": 1325 + }, + { + "epoch": 5.4, + "grad_norm": 11.026411056518555, + "learning_rate": 0.0003233256351039261, + "loss": 0.2757, + "step": 1350 + }, + { + "epoch": 5.5, + "grad_norm": 11.038985252380371, + "learning_rate": 0.00032167601451666116, + "loss": 0.2177, + "step": 1375 + }, + { + "epoch": 5.6, + "grad_norm": 4.008651256561279, + "learning_rate": 0.00032002639392939627, + "loss": 0.2163, + "step": 1400 + }, + { + "epoch": 5.7, + "grad_norm": 12.480770111083984, + "learning_rate": 0.0003183767733421313, + "loss": 0.2173, + "step": 1425 + }, + { + "epoch": 5.8, + "grad_norm": 8.751969337463379, + "learning_rate": 0.0003167271527548664, + "loss": 0.2299, + "step": 1450 + }, + { + "epoch": 5.9, + "grad_norm": 7.701971530914307, + "learning_rate": 0.00031507753216760147, + "loss": 0.1949, + "step": 1475 + }, + { + "epoch": 6.0, + "grad_norm": 8.48027515411377, + "learning_rate": 0.00031342791158033657, + "loss": 0.2599, + "step": 1500 + }, + { + "epoch": 6.1, + "grad_norm": 9.29404067993164, + "learning_rate": 0.00031177829099307157, + "loss": 0.1376, + "step": 1525 + }, + { + "epoch": 6.2, + "grad_norm": 20.137714385986328, + "learning_rate": 0.00031012867040580667, + "loss": 0.1647, + "step": 1550 + }, + { + "epoch": 6.3, + "grad_norm": 11.394575119018555, + "learning_rate": 0.0003084790498185417, + "loss": 0.1565, + "step": 1575 + }, + { + "epoch": 6.4, + "grad_norm": 8.214287757873535, + "learning_rate": 0.0003068294292312768, + "loss": 0.1739, + "step": 1600 + }, + { + "epoch": 6.5, + "grad_norm": 7.779988765716553, + "learning_rate": 0.0003051798086440119, + "loss": 0.1403, + "step": 1625 + }, + { + "epoch": 6.6, + "grad_norm": 9.421648025512695, + "learning_rate": 0.000303530188056747, + "loss": 0.1545, + "step": 1650 + }, + { + "epoch": 6.7, + "grad_norm": 5.751734256744385, + "learning_rate": 0.000301880567469482, + "loss": 0.1971, + "step": 1675 + }, + { + "epoch": 6.8, + "grad_norm": 23.861705780029297, + "learning_rate": 0.0003002309468822171, + "loss": 0.1681, + "step": 1700 + }, + { + "epoch": 6.9, + "grad_norm": 18.944721221923828, + "learning_rate": 0.0002985813262949522, + "loss": 0.1703, + "step": 1725 + }, + { + "epoch": 7.0, + "grad_norm": 14.045795440673828, + "learning_rate": 0.00029693170570768723, + "loss": 0.1801, + "step": 1750 + }, + { + "epoch": 7.1, + "grad_norm": 6.4620137214660645, + "learning_rate": 0.00029528208512042233, + "loss": 0.1253, + "step": 1775 + }, + { + "epoch": 7.2, + "grad_norm": 4.318169593811035, + "learning_rate": 0.0002936324645331574, + "loss": 0.1397, + "step": 1800 + }, + { + "epoch": 7.3, + "grad_norm": 24.91462516784668, + "learning_rate": 0.00029198284394589243, + "loss": 0.1259, + "step": 1825 + }, + { + "epoch": 7.4, + "grad_norm": 23.614572525024414, + "learning_rate": 0.00029033322335862753, + "loss": 0.1293, + "step": 1850 + }, + { + "epoch": 7.5, + "grad_norm": 3.60048508644104, + "learning_rate": 0.0002886836027713626, + "loss": 0.138, + "step": 1875 + }, + { + "epoch": 7.6, + "grad_norm": 16.62705421447754, + "learning_rate": 0.0002870339821840977, + "loss": 0.1733, + "step": 1900 + }, + { + "epoch": 7.7, + "grad_norm": 33.79671859741211, + "learning_rate": 0.00028538436159683273, + "loss": 0.217, + "step": 1925 + }, + { + "epoch": 7.8, + "grad_norm": 9.69206428527832, + "learning_rate": 0.00028373474100956784, + "loss": 0.1289, + "step": 1950 + }, + { + "epoch": 7.9, + "grad_norm": 18.655046463012695, + "learning_rate": 0.0002820851204223029, + "loss": 0.1168, + "step": 1975 + }, + { + "epoch": 8.0, + "grad_norm": 8.4110746383667, + "learning_rate": 0.00028043549983503794, + "loss": 0.1471, + "step": 2000 + }, + { + "epoch": 8.1, + "grad_norm": 14.473428726196289, + "learning_rate": 0.00027878587924777304, + "loss": 0.0774, + "step": 2025 + }, + { + "epoch": 8.2, + "grad_norm": 1.2464979887008667, + "learning_rate": 0.0002771362586605081, + "loss": 0.0715, + "step": 2050 + }, + { + "epoch": 8.3, + "grad_norm": 26.74981689453125, + "learning_rate": 0.0002754866380732432, + "loss": 0.1289, + "step": 2075 + }, + { + "epoch": 8.4, + "grad_norm": 13.958703994750977, + "learning_rate": 0.00027383701748597824, + "loss": 0.1294, + "step": 2100 + }, + { + "epoch": 8.5, + "grad_norm": 10.394835472106934, + "learning_rate": 0.0002721873968987133, + "loss": 0.0987, + "step": 2125 + }, + { + "epoch": 8.6, + "grad_norm": 9.387774467468262, + "learning_rate": 0.00027053777631144834, + "loss": 0.1269, + "step": 2150 + }, + { + "epoch": 8.7, + "grad_norm": 4.988718509674072, + "learning_rate": 0.00026888815572418344, + "loss": 0.1374, + "step": 2175 + }, + { + "epoch": 8.8, + "grad_norm": 12.66832447052002, + "learning_rate": 0.00026723853513691855, + "loss": 0.0997, + "step": 2200 + }, + { + "epoch": 8.9, + "grad_norm": 19.10486602783203, + "learning_rate": 0.0002655889145496536, + "loss": 0.1325, + "step": 2225 + }, + { + "epoch": 9.0, + "grad_norm": 3.2943971157073975, + "learning_rate": 0.0002639392939623887, + "loss": 0.1385, + "step": 2250 + }, + { + "epoch": 9.1, + "grad_norm": 11.490874290466309, + "learning_rate": 0.0002622896733751237, + "loss": 0.0573, + "step": 2275 + }, + { + "epoch": 9.2, + "grad_norm": 11.961288452148438, + "learning_rate": 0.0002606400527878588, + "loss": 0.1128, + "step": 2300 + }, + { + "epoch": 9.3, + "grad_norm": 0.7256277203559875, + "learning_rate": 0.00025899043220059385, + "loss": 0.0673, + "step": 2325 + }, + { + "epoch": 9.4, + "grad_norm": 9.857927322387695, + "learning_rate": 0.00025734081161332895, + "loss": 0.2072, + "step": 2350 + }, + { + "epoch": 9.5, + "grad_norm": 34.927734375, + "learning_rate": 0.000255691191026064, + "loss": 0.221, + "step": 2375 + }, + { + "epoch": 9.6, + "grad_norm": 28.461849212646484, + "learning_rate": 0.0002540415704387991, + "loss": 0.1263, + "step": 2400 + }, + { + "epoch": 9.7, + "grad_norm": 9.591217041015625, + "learning_rate": 0.00025239194985153415, + "loss": 0.099, + "step": 2425 + }, + { + "epoch": 9.8, + "grad_norm": 9.780253410339355, + "learning_rate": 0.0002507423292642692, + "loss": 0.0754, + "step": 2450 + }, + { + "epoch": 9.9, + "grad_norm": 14.394485473632812, + "learning_rate": 0.0002490927086770043, + "loss": 0.0909, + "step": 2475 + }, + { + "epoch": 10.0, + "grad_norm": 4.217277526855469, + "learning_rate": 0.00024744308808973936, + "loss": 0.1495, + "step": 2500 + }, + { + "epoch": 10.1, + "grad_norm": 0.08438724279403687, + "learning_rate": 0.00024579346750247446, + "loss": 0.0662, + "step": 2525 + }, + { + "epoch": 10.2, + "grad_norm": 0.9595862030982971, + "learning_rate": 0.0002441438469152095, + "loss": 0.084, + "step": 2550 + }, + { + "epoch": 10.3, + "grad_norm": 42.32374954223633, + "learning_rate": 0.00024249422632794456, + "loss": 0.0541, + "step": 2575 + }, + { + "epoch": 10.4, + "grad_norm": 24.792821884155273, + "learning_rate": 0.00024084460574067963, + "loss": 0.121, + "step": 2600 + }, + { + "epoch": 10.5, + "grad_norm": 2.7049331665039062, + "learning_rate": 0.0002391949851534147, + "loss": 0.1551, + "step": 2625 + }, + { + "epoch": 10.6, + "grad_norm": 2.788362741470337, + "learning_rate": 0.0002375453645661498, + "loss": 0.0877, + "step": 2650 + }, + { + "epoch": 10.7, + "grad_norm": 0.4999215006828308, + "learning_rate": 0.00023589574397888486, + "loss": 0.1263, + "step": 2675 + }, + { + "epoch": 10.8, + "grad_norm": 16.476675033569336, + "learning_rate": 0.00023424612339161997, + "loss": 0.0883, + "step": 2700 + }, + { + "epoch": 10.9, + "grad_norm": 38.192726135253906, + "learning_rate": 0.000232596502804355, + "loss": 0.0843, + "step": 2725 + }, + { + "epoch": 11.0, + "grad_norm": 18.09952735900879, + "learning_rate": 0.00023094688221709007, + "loss": 0.0957, + "step": 2750 + }, + { + "epoch": 11.1, + "grad_norm": 34.254417419433594, + "learning_rate": 0.00022929726162982514, + "loss": 0.0856, + "step": 2775 + }, + { + "epoch": 11.2, + "grad_norm": 11.744216918945312, + "learning_rate": 0.00022764764104256022, + "loss": 0.0654, + "step": 2800 + }, + { + "epoch": 11.3, + "grad_norm": 0.1285558044910431, + "learning_rate": 0.0002259980204552953, + "loss": 0.0638, + "step": 2825 + }, + { + "epoch": 11.4, + "grad_norm": 20.57583999633789, + "learning_rate": 0.00022434839986803037, + "loss": 0.079, + "step": 2850 + }, + { + "epoch": 11.5, + "grad_norm": 0.7192332148551941, + "learning_rate": 0.00022269877928076542, + "loss": 0.0854, + "step": 2875 + }, + { + "epoch": 11.6, + "grad_norm": 31.341829299926758, + "learning_rate": 0.0002210491586935005, + "loss": 0.063, + "step": 2900 + }, + { + "epoch": 11.7, + "grad_norm": 13.460247039794922, + "learning_rate": 0.00021939953810623557, + "loss": 0.1358, + "step": 2925 + }, + { + "epoch": 11.8, + "grad_norm": 1.671036720275879, + "learning_rate": 0.00021774991751897065, + "loss": 0.0806, + "step": 2950 + }, + { + "epoch": 11.9, + "grad_norm": 22.026491165161133, + "learning_rate": 0.00021610029693170573, + "loss": 0.0912, + "step": 2975 + }, + { + "epoch": 12.0, + "grad_norm": 46.78192138671875, + "learning_rate": 0.0002144506763444408, + "loss": 0.0763, + "step": 3000 + }, + { + "epoch": 12.1, + "grad_norm": 6.397649765014648, + "learning_rate": 0.00021280105575717582, + "loss": 0.0613, + "step": 3025 + }, + { + "epoch": 12.2, + "grad_norm": 7.179838180541992, + "learning_rate": 0.00021115143516991093, + "loss": 0.0496, + "step": 3050 + }, + { + "epoch": 12.3, + "grad_norm": 0.021825680509209633, + "learning_rate": 0.000209501814582646, + "loss": 0.0633, + "step": 3075 + }, + { + "epoch": 12.4, + "grad_norm": 6.6774187088012695, + "learning_rate": 0.00020785219399538108, + "loss": 0.0765, + "step": 3100 + }, + { + "epoch": 12.5, + "grad_norm": 0.13217756152153015, + "learning_rate": 0.00020620257340811616, + "loss": 0.0696, + "step": 3125 + }, + { + "epoch": 12.6, + "grad_norm": 2.5293831825256348, + "learning_rate": 0.00020455295282085123, + "loss": 0.0419, + "step": 3150 + }, + { + "epoch": 12.7, + "grad_norm": 1.1612874269485474, + "learning_rate": 0.00020290333223358626, + "loss": 0.1046, + "step": 3175 + }, + { + "epoch": 12.8, + "grad_norm": 40.93427658081055, + "learning_rate": 0.00020125371164632133, + "loss": 0.0488, + "step": 3200 + }, + { + "epoch": 12.9, + "grad_norm": 1.446478009223938, + "learning_rate": 0.0001996040910590564, + "loss": 0.062, + "step": 3225 + }, + { + "epoch": 13.0, + "grad_norm": 28.893821716308594, + "learning_rate": 0.00019795447047179148, + "loss": 0.0913, + "step": 3250 + }, + { + "epoch": 13.1, + "grad_norm": 23.31914520263672, + "learning_rate": 0.0001963048498845266, + "loss": 0.061, + "step": 3275 + }, + { + "epoch": 13.2, + "grad_norm": 0.26242795586586, + "learning_rate": 0.00019465522929726164, + "loss": 0.0563, + "step": 3300 + }, + { + "epoch": 13.3, + "grad_norm": 0.08909507840871811, + "learning_rate": 0.00019300560870999671, + "loss": 0.0387, + "step": 3325 + }, + { + "epoch": 13.4, + "grad_norm": 24.825326919555664, + "learning_rate": 0.0001913559881227318, + "loss": 0.0776, + "step": 3350 + }, + { + "epoch": 13.5, + "grad_norm": 39.43446731567383, + "learning_rate": 0.00018970636753546684, + "loss": 0.0846, + "step": 3375 + }, + { + "epoch": 13.6, + "grad_norm": 32.98988723754883, + "learning_rate": 0.00018805674694820192, + "loss": 0.0512, + "step": 3400 + }, + { + "epoch": 13.7, + "grad_norm": 4.053821563720703, + "learning_rate": 0.000186407126360937, + "loss": 0.1576, + "step": 3425 + }, + { + "epoch": 13.8, + "grad_norm": 0.2785554528236389, + "learning_rate": 0.00018475750577367207, + "loss": 0.0728, + "step": 3450 + }, + { + "epoch": 13.9, + "grad_norm": 11.39714527130127, + "learning_rate": 0.00018310788518640715, + "loss": 0.0776, + "step": 3475 + }, + { + "epoch": 14.0, + "grad_norm": 3.7018072605133057, + "learning_rate": 0.00018145826459914222, + "loss": 0.0736, + "step": 3500 + }, + { + "epoch": 14.1, + "grad_norm": 0.26053619384765625, + "learning_rate": 0.00017980864401187727, + "loss": 0.0598, + "step": 3525 + }, + { + "epoch": 14.2, + "grad_norm": 6.454179763793945, + "learning_rate": 0.00017815902342461235, + "loss": 0.0435, + "step": 3550 + }, + { + "epoch": 14.3, + "grad_norm": 34.47985076904297, + "learning_rate": 0.00017650940283734742, + "loss": 0.0467, + "step": 3575 + }, + { + "epoch": 14.4, + "grad_norm": 11.76491928100586, + "learning_rate": 0.00017485978225008247, + "loss": 0.1298, + "step": 3600 + }, + { + "epoch": 14.5, + "grad_norm": 26.64131736755371, + "learning_rate": 0.00017321016166281755, + "loss": 0.1313, + "step": 3625 + }, + { + "epoch": 14.6, + "grad_norm": 5.510906219482422, + "learning_rate": 0.00017156054107555263, + "loss": 0.0421, + "step": 3650 + }, + { + "epoch": 14.7, + "grad_norm": 0.04721131548285484, + "learning_rate": 0.0001699109204882877, + "loss": 0.0569, + "step": 3675 + }, + { + "epoch": 14.8, + "grad_norm": 0.6616227030754089, + "learning_rate": 0.00016826129990102278, + "loss": 0.0618, + "step": 3700 + }, + { + "epoch": 14.9, + "grad_norm": 0.05287986248731613, + "learning_rate": 0.00016661167931375785, + "loss": 0.0574, + "step": 3725 + }, + { + "epoch": 15.0, + "grad_norm": 0.49707508087158203, + "learning_rate": 0.0001649620587264929, + "loss": 0.094, + "step": 3750 + }, + { + "epoch": 15.1, + "grad_norm": 23.649728775024414, + "learning_rate": 0.00016331243813922798, + "loss": 0.0512, + "step": 3775 + }, + { + "epoch": 15.2, + "grad_norm": 0.01772051490843296, + "learning_rate": 0.00016166281755196306, + "loss": 0.0719, + "step": 3800 + }, + { + "epoch": 15.3, + "grad_norm": 1.5259039402008057, + "learning_rate": 0.00016001319696469813, + "loss": 0.0615, + "step": 3825 + }, + { + "epoch": 15.4, + "grad_norm": 0.026406478136777878, + "learning_rate": 0.0001583635763774332, + "loss": 0.0651, + "step": 3850 + }, + { + "epoch": 15.5, + "grad_norm": 27.12177085876465, + "learning_rate": 0.00015671395579016829, + "loss": 0.0458, + "step": 3875 + }, + { + "epoch": 15.6, + "grad_norm": 9.838811874389648, + "learning_rate": 0.00015506433520290334, + "loss": 0.0662, + "step": 3900 + }, + { + "epoch": 15.7, + "grad_norm": 0.08516795933246613, + "learning_rate": 0.0001534147146156384, + "loss": 0.0302, + "step": 3925 + }, + { + "epoch": 15.8, + "grad_norm": 1.1224850416183472, + "learning_rate": 0.0001517650940283735, + "loss": 0.0637, + "step": 3950 + }, + { + "epoch": 15.9, + "grad_norm": 0.14810702204704285, + "learning_rate": 0.00015011547344110854, + "loss": 0.0296, + "step": 3975 + }, + { + "epoch": 16.0, + "grad_norm": 0.046288371086120605, + "learning_rate": 0.00014846585285384361, + "loss": 0.0368, + "step": 4000 + } + ], + "logging_steps": 25, + "max_steps": 6250, + "num_input_tokens_seen": 0, + "num_train_epochs": 25, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 203277926400000.0, + "train_batch_size": 20, + "trial_name": null, + "trial_params": null +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-4000/training_args.bin b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-4000/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..798d8a9f3a9513692035320aba5484c89ac49e69 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-4000/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1cabf5db3e9d79ec0c1945933395ff57418204414cc81f77c90324d003fac782 +size 5304 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-4000/vocab.txt b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-4000/vocab.txt new file mode 100644 index 0000000000000000000000000000000000000000..cc9ad53732e21ea9180f18e1fb67a38cc5702d4c --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-4000/vocab.txt @@ -0,0 +1,4101 @@ +[PAD] +[UNK] +[CLS] +[SEP] +[MASK] +AAAAAA +AAAAAT +AAAAAC +AAAAAG +AAAATA +AAAATT +AAAATC +AAAATG +AAAACA +AAAACT +AAAACC +AAAACG +AAAAGA +AAAAGT +AAAAGC +AAAAGG +AAATAA +AAATAT +AAATAC +AAATAG +AAATTA +AAATTT +AAATTC +AAATTG +AAATCA +AAATCT +AAATCC +AAATCG +AAATGA +AAATGT +AAATGC +AAATGG +AAACAA +AAACAT +AAACAC +AAACAG +AAACTA +AAACTT +AAACTC +AAACTG +AAACCA +AAACCT +AAACCC +AAACCG +AAACGA +AAACGT +AAACGC +AAACGG +AAAGAA +AAAGAT +AAAGAC +AAAGAG +AAAGTA +AAAGTT +AAAGTC +AAAGTG +AAAGCA +AAAGCT +AAAGCC +AAAGCG +AAAGGA +AAAGGT +AAAGGC +AAAGGG +AATAAA +AATAAT +AATAAC +AATAAG +AATATA +AATATT +AATATC +AATATG +AATACA +AATACT +AATACC +AATACG +AATAGA +AATAGT +AATAGC +AATAGG +AATTAA +AATTAT +AATTAC +AATTAG +AATTTA +AATTTT +AATTTC +AATTTG +AATTCA +AATTCT +AATTCC +AATTCG +AATTGA +AATTGT +AATTGC +AATTGG +AATCAA +AATCAT +AATCAC +AATCAG +AATCTA +AATCTT +AATCTC +AATCTG +AATCCA +AATCCT +AATCCC +AATCCG +AATCGA +AATCGT +AATCGC +AATCGG +AATGAA +AATGAT +AATGAC +AATGAG +AATGTA +AATGTT +AATGTC +AATGTG +AATGCA +AATGCT +AATGCC +AATGCG +AATGGA +AATGGT +AATGGC +AATGGG +AACAAA +AACAAT +AACAAC +AACAAG +AACATA +AACATT +AACATC +AACATG +AACACA +AACACT +AACACC +AACACG +AACAGA +AACAGT +AACAGC +AACAGG +AACTAA +AACTAT +AACTAC +AACTAG +AACTTA +AACTTT +AACTTC +AACTTG +AACTCA +AACTCT +AACTCC +AACTCG +AACTGA +AACTGT +AACTGC +AACTGG +AACCAA +AACCAT +AACCAC +AACCAG +AACCTA +AACCTT +AACCTC +AACCTG +AACCCA +AACCCT +AACCCC +AACCCG +AACCGA +AACCGT +AACCGC +AACCGG +AACGAA +AACGAT +AACGAC +AACGAG +AACGTA +AACGTT +AACGTC +AACGTG +AACGCA +AACGCT +AACGCC +AACGCG +AACGGA +AACGGT +AACGGC +AACGGG +AAGAAA +AAGAAT +AAGAAC +AAGAAG +AAGATA +AAGATT +AAGATC +AAGATG +AAGACA +AAGACT +AAGACC +AAGACG +AAGAGA +AAGAGT +AAGAGC +AAGAGG +AAGTAA +AAGTAT +AAGTAC +AAGTAG +AAGTTA +AAGTTT +AAGTTC +AAGTTG +AAGTCA +AAGTCT +AAGTCC +AAGTCG +AAGTGA +AAGTGT +AAGTGC +AAGTGG +AAGCAA +AAGCAT +AAGCAC +AAGCAG +AAGCTA +AAGCTT +AAGCTC +AAGCTG +AAGCCA +AAGCCT +AAGCCC +AAGCCG +AAGCGA +AAGCGT +AAGCGC +AAGCGG +AAGGAA +AAGGAT +AAGGAC +AAGGAG +AAGGTA +AAGGTT +AAGGTC +AAGGTG +AAGGCA +AAGGCT +AAGGCC +AAGGCG +AAGGGA +AAGGGT +AAGGGC +AAGGGG +ATAAAA +ATAAAT +ATAAAC +ATAAAG +ATAATA +ATAATT +ATAATC +ATAATG +ATAACA +ATAACT +ATAACC +ATAACG +ATAAGA +ATAAGT +ATAAGC +ATAAGG +ATATAA +ATATAT +ATATAC +ATATAG +ATATTA +ATATTT +ATATTC +ATATTG +ATATCA +ATATCT +ATATCC +ATATCG +ATATGA +ATATGT +ATATGC +ATATGG +ATACAA +ATACAT +ATACAC +ATACAG +ATACTA +ATACTT +ATACTC +ATACTG +ATACCA +ATACCT +ATACCC +ATACCG +ATACGA +ATACGT +ATACGC +ATACGG +ATAGAA +ATAGAT +ATAGAC +ATAGAG +ATAGTA +ATAGTT +ATAGTC +ATAGTG +ATAGCA +ATAGCT +ATAGCC +ATAGCG +ATAGGA +ATAGGT +ATAGGC +ATAGGG +ATTAAA +ATTAAT +ATTAAC +ATTAAG +ATTATA +ATTATT +ATTATC +ATTATG +ATTACA +ATTACT +ATTACC +ATTACG +ATTAGA +ATTAGT +ATTAGC +ATTAGG +ATTTAA +ATTTAT +ATTTAC +ATTTAG +ATTTTA +ATTTTT +ATTTTC +ATTTTG +ATTTCA +ATTTCT +ATTTCC +ATTTCG +ATTTGA +ATTTGT +ATTTGC +ATTTGG +ATTCAA +ATTCAT +ATTCAC +ATTCAG +ATTCTA +ATTCTT +ATTCTC +ATTCTG +ATTCCA +ATTCCT +ATTCCC +ATTCCG +ATTCGA +ATTCGT +ATTCGC +ATTCGG +ATTGAA +ATTGAT +ATTGAC +ATTGAG +ATTGTA +ATTGTT +ATTGTC +ATTGTG +ATTGCA +ATTGCT +ATTGCC +ATTGCG +ATTGGA +ATTGGT +ATTGGC +ATTGGG +ATCAAA +ATCAAT +ATCAAC +ATCAAG +ATCATA +ATCATT +ATCATC +ATCATG +ATCACA +ATCACT +ATCACC +ATCACG +ATCAGA +ATCAGT +ATCAGC +ATCAGG +ATCTAA +ATCTAT +ATCTAC +ATCTAG +ATCTTA +ATCTTT +ATCTTC +ATCTTG +ATCTCA +ATCTCT +ATCTCC +ATCTCG +ATCTGA +ATCTGT +ATCTGC +ATCTGG +ATCCAA +ATCCAT +ATCCAC +ATCCAG +ATCCTA +ATCCTT +ATCCTC +ATCCTG +ATCCCA +ATCCCT +ATCCCC +ATCCCG +ATCCGA +ATCCGT +ATCCGC +ATCCGG +ATCGAA +ATCGAT +ATCGAC +ATCGAG +ATCGTA +ATCGTT +ATCGTC +ATCGTG +ATCGCA +ATCGCT +ATCGCC +ATCGCG +ATCGGA +ATCGGT +ATCGGC +ATCGGG +ATGAAA +ATGAAT +ATGAAC +ATGAAG +ATGATA +ATGATT +ATGATC +ATGATG +ATGACA +ATGACT +ATGACC +ATGACG +ATGAGA +ATGAGT +ATGAGC +ATGAGG +ATGTAA +ATGTAT +ATGTAC +ATGTAG +ATGTTA +ATGTTT +ATGTTC +ATGTTG +ATGTCA +ATGTCT +ATGTCC +ATGTCG +ATGTGA +ATGTGT +ATGTGC +ATGTGG +ATGCAA +ATGCAT +ATGCAC +ATGCAG +ATGCTA +ATGCTT +ATGCTC +ATGCTG +ATGCCA +ATGCCT +ATGCCC +ATGCCG +ATGCGA +ATGCGT +ATGCGC +ATGCGG +ATGGAA +ATGGAT +ATGGAC +ATGGAG +ATGGTA +ATGGTT +ATGGTC +ATGGTG +ATGGCA +ATGGCT +ATGGCC +ATGGCG +ATGGGA +ATGGGT +ATGGGC +ATGGGG +ACAAAA +ACAAAT +ACAAAC +ACAAAG +ACAATA +ACAATT +ACAATC +ACAATG +ACAACA +ACAACT +ACAACC +ACAACG +ACAAGA +ACAAGT +ACAAGC +ACAAGG +ACATAA +ACATAT +ACATAC +ACATAG +ACATTA +ACATTT +ACATTC +ACATTG +ACATCA +ACATCT +ACATCC +ACATCG +ACATGA +ACATGT +ACATGC +ACATGG +ACACAA +ACACAT +ACACAC +ACACAG +ACACTA +ACACTT +ACACTC +ACACTG +ACACCA +ACACCT +ACACCC +ACACCG +ACACGA +ACACGT +ACACGC +ACACGG +ACAGAA +ACAGAT +ACAGAC +ACAGAG +ACAGTA +ACAGTT +ACAGTC +ACAGTG +ACAGCA +ACAGCT +ACAGCC +ACAGCG +ACAGGA +ACAGGT +ACAGGC +ACAGGG +ACTAAA +ACTAAT +ACTAAC +ACTAAG +ACTATA +ACTATT +ACTATC +ACTATG +ACTACA +ACTACT +ACTACC +ACTACG +ACTAGA +ACTAGT +ACTAGC +ACTAGG +ACTTAA +ACTTAT +ACTTAC +ACTTAG +ACTTTA +ACTTTT +ACTTTC +ACTTTG +ACTTCA +ACTTCT +ACTTCC +ACTTCG +ACTTGA +ACTTGT +ACTTGC +ACTTGG +ACTCAA +ACTCAT +ACTCAC +ACTCAG +ACTCTA +ACTCTT +ACTCTC +ACTCTG +ACTCCA +ACTCCT +ACTCCC +ACTCCG +ACTCGA +ACTCGT +ACTCGC +ACTCGG +ACTGAA +ACTGAT +ACTGAC +ACTGAG +ACTGTA +ACTGTT +ACTGTC +ACTGTG +ACTGCA +ACTGCT +ACTGCC +ACTGCG +ACTGGA +ACTGGT +ACTGGC +ACTGGG +ACCAAA +ACCAAT +ACCAAC +ACCAAG +ACCATA +ACCATT +ACCATC +ACCATG +ACCACA +ACCACT +ACCACC +ACCACG +ACCAGA +ACCAGT +ACCAGC +ACCAGG +ACCTAA +ACCTAT +ACCTAC +ACCTAG +ACCTTA +ACCTTT +ACCTTC +ACCTTG +ACCTCA +ACCTCT +ACCTCC +ACCTCG +ACCTGA +ACCTGT +ACCTGC +ACCTGG +ACCCAA +ACCCAT +ACCCAC +ACCCAG +ACCCTA +ACCCTT +ACCCTC +ACCCTG +ACCCCA +ACCCCT +ACCCCC +ACCCCG +ACCCGA +ACCCGT +ACCCGC +ACCCGG +ACCGAA +ACCGAT +ACCGAC +ACCGAG +ACCGTA +ACCGTT +ACCGTC +ACCGTG +ACCGCA +ACCGCT +ACCGCC +ACCGCG +ACCGGA +ACCGGT +ACCGGC +ACCGGG +ACGAAA +ACGAAT +ACGAAC +ACGAAG +ACGATA +ACGATT +ACGATC +ACGATG +ACGACA +ACGACT +ACGACC +ACGACG +ACGAGA +ACGAGT +ACGAGC +ACGAGG +ACGTAA +ACGTAT +ACGTAC +ACGTAG +ACGTTA +ACGTTT +ACGTTC +ACGTTG +ACGTCA +ACGTCT +ACGTCC +ACGTCG +ACGTGA +ACGTGT +ACGTGC +ACGTGG +ACGCAA +ACGCAT +ACGCAC +ACGCAG +ACGCTA +ACGCTT +ACGCTC +ACGCTG +ACGCCA +ACGCCT +ACGCCC +ACGCCG +ACGCGA +ACGCGT +ACGCGC +ACGCGG +ACGGAA +ACGGAT +ACGGAC +ACGGAG +ACGGTA +ACGGTT +ACGGTC +ACGGTG +ACGGCA +ACGGCT +ACGGCC +ACGGCG +ACGGGA +ACGGGT +ACGGGC +ACGGGG +AGAAAA +AGAAAT +AGAAAC +AGAAAG +AGAATA +AGAATT +AGAATC +AGAATG +AGAACA +AGAACT +AGAACC +AGAACG +AGAAGA +AGAAGT +AGAAGC +AGAAGG +AGATAA +AGATAT +AGATAC +AGATAG +AGATTA +AGATTT +AGATTC +AGATTG +AGATCA +AGATCT +AGATCC +AGATCG +AGATGA +AGATGT +AGATGC +AGATGG +AGACAA +AGACAT +AGACAC +AGACAG +AGACTA +AGACTT +AGACTC +AGACTG +AGACCA +AGACCT +AGACCC +AGACCG +AGACGA +AGACGT +AGACGC +AGACGG +AGAGAA +AGAGAT +AGAGAC +AGAGAG +AGAGTA +AGAGTT +AGAGTC +AGAGTG +AGAGCA +AGAGCT +AGAGCC +AGAGCG +AGAGGA +AGAGGT +AGAGGC +AGAGGG +AGTAAA +AGTAAT +AGTAAC +AGTAAG +AGTATA +AGTATT +AGTATC +AGTATG +AGTACA +AGTACT +AGTACC +AGTACG +AGTAGA +AGTAGT +AGTAGC +AGTAGG +AGTTAA +AGTTAT +AGTTAC +AGTTAG +AGTTTA +AGTTTT +AGTTTC +AGTTTG +AGTTCA +AGTTCT +AGTTCC +AGTTCG +AGTTGA +AGTTGT +AGTTGC +AGTTGG +AGTCAA +AGTCAT +AGTCAC +AGTCAG +AGTCTA +AGTCTT +AGTCTC +AGTCTG +AGTCCA +AGTCCT +AGTCCC +AGTCCG +AGTCGA +AGTCGT +AGTCGC +AGTCGG +AGTGAA +AGTGAT +AGTGAC +AGTGAG +AGTGTA +AGTGTT +AGTGTC +AGTGTG +AGTGCA +AGTGCT +AGTGCC +AGTGCG +AGTGGA +AGTGGT +AGTGGC +AGTGGG +AGCAAA +AGCAAT +AGCAAC +AGCAAG +AGCATA +AGCATT +AGCATC +AGCATG +AGCACA +AGCACT +AGCACC +AGCACG +AGCAGA +AGCAGT +AGCAGC +AGCAGG +AGCTAA +AGCTAT +AGCTAC +AGCTAG +AGCTTA +AGCTTT +AGCTTC +AGCTTG +AGCTCA +AGCTCT +AGCTCC +AGCTCG +AGCTGA +AGCTGT +AGCTGC +AGCTGG +AGCCAA +AGCCAT +AGCCAC +AGCCAG +AGCCTA +AGCCTT +AGCCTC +AGCCTG +AGCCCA +AGCCCT +AGCCCC +AGCCCG +AGCCGA +AGCCGT +AGCCGC +AGCCGG +AGCGAA +AGCGAT +AGCGAC +AGCGAG +AGCGTA +AGCGTT +AGCGTC +AGCGTG +AGCGCA +AGCGCT +AGCGCC +AGCGCG +AGCGGA +AGCGGT +AGCGGC +AGCGGG +AGGAAA +AGGAAT +AGGAAC +AGGAAG +AGGATA +AGGATT +AGGATC +AGGATG +AGGACA +AGGACT +AGGACC +AGGACG +AGGAGA +AGGAGT +AGGAGC +AGGAGG +AGGTAA +AGGTAT +AGGTAC +AGGTAG +AGGTTA +AGGTTT +AGGTTC +AGGTTG +AGGTCA +AGGTCT +AGGTCC +AGGTCG +AGGTGA +AGGTGT +AGGTGC +AGGTGG +AGGCAA +AGGCAT +AGGCAC +AGGCAG +AGGCTA +AGGCTT +AGGCTC +AGGCTG +AGGCCA +AGGCCT +AGGCCC +AGGCCG +AGGCGA +AGGCGT +AGGCGC +AGGCGG +AGGGAA +AGGGAT +AGGGAC +AGGGAG +AGGGTA +AGGGTT +AGGGTC +AGGGTG +AGGGCA +AGGGCT +AGGGCC +AGGGCG +AGGGGA +AGGGGT +AGGGGC +AGGGGG +TAAAAA +TAAAAT +TAAAAC +TAAAAG +TAAATA +TAAATT +TAAATC +TAAATG +TAAACA +TAAACT +TAAACC +TAAACG +TAAAGA +TAAAGT +TAAAGC +TAAAGG +TAATAA +TAATAT +TAATAC +TAATAG +TAATTA +TAATTT +TAATTC +TAATTG +TAATCA +TAATCT +TAATCC +TAATCG +TAATGA +TAATGT +TAATGC +TAATGG +TAACAA +TAACAT +TAACAC +TAACAG +TAACTA +TAACTT +TAACTC +TAACTG +TAACCA +TAACCT +TAACCC +TAACCG +TAACGA +TAACGT +TAACGC +TAACGG +TAAGAA +TAAGAT +TAAGAC +TAAGAG +TAAGTA +TAAGTT +TAAGTC +TAAGTG +TAAGCA +TAAGCT +TAAGCC +TAAGCG +TAAGGA +TAAGGT +TAAGGC +TAAGGG +TATAAA +TATAAT +TATAAC +TATAAG +TATATA +TATATT +TATATC +TATATG +TATACA +TATACT +TATACC +TATACG +TATAGA +TATAGT +TATAGC +TATAGG +TATTAA +TATTAT +TATTAC +TATTAG +TATTTA +TATTTT +TATTTC +TATTTG +TATTCA +TATTCT +TATTCC +TATTCG +TATTGA +TATTGT +TATTGC +TATTGG +TATCAA +TATCAT +TATCAC +TATCAG +TATCTA +TATCTT +TATCTC +TATCTG +TATCCA +TATCCT +TATCCC +TATCCG +TATCGA +TATCGT +TATCGC +TATCGG +TATGAA +TATGAT +TATGAC +TATGAG +TATGTA +TATGTT +TATGTC +TATGTG +TATGCA +TATGCT +TATGCC +TATGCG +TATGGA +TATGGT +TATGGC +TATGGG +TACAAA +TACAAT +TACAAC +TACAAG +TACATA +TACATT +TACATC +TACATG +TACACA +TACACT +TACACC +TACACG +TACAGA +TACAGT +TACAGC +TACAGG +TACTAA +TACTAT +TACTAC +TACTAG +TACTTA +TACTTT +TACTTC +TACTTG +TACTCA +TACTCT +TACTCC +TACTCG +TACTGA +TACTGT +TACTGC +TACTGG +TACCAA +TACCAT +TACCAC +TACCAG +TACCTA +TACCTT +TACCTC +TACCTG +TACCCA +TACCCT +TACCCC +TACCCG +TACCGA +TACCGT +TACCGC +TACCGG +TACGAA +TACGAT +TACGAC +TACGAG +TACGTA +TACGTT +TACGTC +TACGTG +TACGCA +TACGCT +TACGCC +TACGCG +TACGGA +TACGGT +TACGGC +TACGGG +TAGAAA +TAGAAT +TAGAAC +TAGAAG +TAGATA +TAGATT +TAGATC +TAGATG +TAGACA +TAGACT +TAGACC +TAGACG +TAGAGA +TAGAGT +TAGAGC +TAGAGG +TAGTAA +TAGTAT +TAGTAC +TAGTAG +TAGTTA +TAGTTT +TAGTTC +TAGTTG +TAGTCA +TAGTCT +TAGTCC +TAGTCG +TAGTGA +TAGTGT +TAGTGC +TAGTGG +TAGCAA +TAGCAT +TAGCAC +TAGCAG +TAGCTA +TAGCTT +TAGCTC +TAGCTG +TAGCCA +TAGCCT +TAGCCC +TAGCCG +TAGCGA +TAGCGT +TAGCGC +TAGCGG +TAGGAA +TAGGAT +TAGGAC +TAGGAG +TAGGTA +TAGGTT +TAGGTC +TAGGTG +TAGGCA +TAGGCT +TAGGCC +TAGGCG +TAGGGA +TAGGGT +TAGGGC +TAGGGG +TTAAAA +TTAAAT +TTAAAC +TTAAAG +TTAATA +TTAATT +TTAATC +TTAATG +TTAACA +TTAACT +TTAACC +TTAACG +TTAAGA +TTAAGT +TTAAGC +TTAAGG +TTATAA +TTATAT +TTATAC +TTATAG +TTATTA +TTATTT +TTATTC +TTATTG +TTATCA +TTATCT +TTATCC +TTATCG +TTATGA +TTATGT +TTATGC +TTATGG +TTACAA +TTACAT +TTACAC +TTACAG +TTACTA +TTACTT +TTACTC +TTACTG +TTACCA +TTACCT +TTACCC +TTACCG +TTACGA +TTACGT +TTACGC +TTACGG +TTAGAA +TTAGAT +TTAGAC +TTAGAG +TTAGTA +TTAGTT +TTAGTC +TTAGTG +TTAGCA +TTAGCT +TTAGCC +TTAGCG +TTAGGA +TTAGGT +TTAGGC +TTAGGG +TTTAAA +TTTAAT +TTTAAC +TTTAAG +TTTATA +TTTATT +TTTATC +TTTATG +TTTACA +TTTACT +TTTACC +TTTACG +TTTAGA +TTTAGT +TTTAGC +TTTAGG +TTTTAA +TTTTAT +TTTTAC +TTTTAG +TTTTTA +TTTTTT +TTTTTC +TTTTTG +TTTTCA +TTTTCT +TTTTCC +TTTTCG +TTTTGA +TTTTGT +TTTTGC +TTTTGG +TTTCAA +TTTCAT +TTTCAC +TTTCAG +TTTCTA +TTTCTT +TTTCTC +TTTCTG +TTTCCA +TTTCCT +TTTCCC +TTTCCG +TTTCGA +TTTCGT +TTTCGC +TTTCGG +TTTGAA +TTTGAT +TTTGAC +TTTGAG +TTTGTA +TTTGTT +TTTGTC +TTTGTG +TTTGCA +TTTGCT +TTTGCC +TTTGCG +TTTGGA +TTTGGT +TTTGGC +TTTGGG +TTCAAA +TTCAAT +TTCAAC +TTCAAG +TTCATA +TTCATT +TTCATC +TTCATG +TTCACA +TTCACT +TTCACC +TTCACG +TTCAGA +TTCAGT +TTCAGC +TTCAGG +TTCTAA +TTCTAT +TTCTAC +TTCTAG +TTCTTA +TTCTTT +TTCTTC +TTCTTG +TTCTCA +TTCTCT +TTCTCC +TTCTCG +TTCTGA +TTCTGT +TTCTGC +TTCTGG +TTCCAA +TTCCAT +TTCCAC +TTCCAG +TTCCTA +TTCCTT +TTCCTC +TTCCTG +TTCCCA +TTCCCT +TTCCCC +TTCCCG +TTCCGA +TTCCGT +TTCCGC +TTCCGG +TTCGAA +TTCGAT +TTCGAC +TTCGAG +TTCGTA +TTCGTT +TTCGTC +TTCGTG +TTCGCA +TTCGCT +TTCGCC +TTCGCG +TTCGGA +TTCGGT +TTCGGC +TTCGGG +TTGAAA +TTGAAT +TTGAAC +TTGAAG +TTGATA +TTGATT +TTGATC +TTGATG +TTGACA +TTGACT +TTGACC +TTGACG +TTGAGA +TTGAGT +TTGAGC +TTGAGG +TTGTAA +TTGTAT +TTGTAC +TTGTAG +TTGTTA +TTGTTT +TTGTTC +TTGTTG +TTGTCA +TTGTCT +TTGTCC +TTGTCG +TTGTGA +TTGTGT +TTGTGC +TTGTGG +TTGCAA +TTGCAT +TTGCAC +TTGCAG +TTGCTA +TTGCTT +TTGCTC +TTGCTG +TTGCCA +TTGCCT +TTGCCC +TTGCCG +TTGCGA +TTGCGT +TTGCGC +TTGCGG +TTGGAA +TTGGAT +TTGGAC +TTGGAG +TTGGTA +TTGGTT +TTGGTC +TTGGTG +TTGGCA +TTGGCT +TTGGCC +TTGGCG +TTGGGA +TTGGGT +TTGGGC +TTGGGG +TCAAAA +TCAAAT +TCAAAC +TCAAAG +TCAATA +TCAATT +TCAATC +TCAATG +TCAACA +TCAACT +TCAACC +TCAACG +TCAAGA +TCAAGT +TCAAGC +TCAAGG +TCATAA +TCATAT +TCATAC +TCATAG +TCATTA +TCATTT +TCATTC +TCATTG +TCATCA +TCATCT +TCATCC +TCATCG +TCATGA +TCATGT +TCATGC +TCATGG +TCACAA +TCACAT +TCACAC +TCACAG +TCACTA +TCACTT +TCACTC +TCACTG +TCACCA +TCACCT +TCACCC +TCACCG +TCACGA +TCACGT +TCACGC +TCACGG +TCAGAA +TCAGAT +TCAGAC +TCAGAG +TCAGTA +TCAGTT +TCAGTC +TCAGTG +TCAGCA +TCAGCT +TCAGCC +TCAGCG +TCAGGA +TCAGGT +TCAGGC +TCAGGG +TCTAAA +TCTAAT +TCTAAC +TCTAAG +TCTATA +TCTATT +TCTATC +TCTATG +TCTACA +TCTACT +TCTACC +TCTACG +TCTAGA +TCTAGT +TCTAGC +TCTAGG +TCTTAA +TCTTAT +TCTTAC +TCTTAG +TCTTTA +TCTTTT +TCTTTC +TCTTTG +TCTTCA +TCTTCT +TCTTCC +TCTTCG +TCTTGA +TCTTGT +TCTTGC +TCTTGG +TCTCAA +TCTCAT +TCTCAC +TCTCAG +TCTCTA +TCTCTT +TCTCTC +TCTCTG +TCTCCA +TCTCCT +TCTCCC +TCTCCG +TCTCGA +TCTCGT +TCTCGC +TCTCGG +TCTGAA +TCTGAT +TCTGAC +TCTGAG +TCTGTA +TCTGTT +TCTGTC +TCTGTG +TCTGCA +TCTGCT +TCTGCC +TCTGCG +TCTGGA +TCTGGT +TCTGGC +TCTGGG +TCCAAA +TCCAAT +TCCAAC +TCCAAG +TCCATA +TCCATT +TCCATC +TCCATG +TCCACA +TCCACT +TCCACC +TCCACG +TCCAGA +TCCAGT +TCCAGC +TCCAGG +TCCTAA +TCCTAT +TCCTAC +TCCTAG +TCCTTA +TCCTTT +TCCTTC +TCCTTG +TCCTCA +TCCTCT +TCCTCC +TCCTCG +TCCTGA +TCCTGT +TCCTGC +TCCTGG +TCCCAA +TCCCAT +TCCCAC +TCCCAG +TCCCTA +TCCCTT +TCCCTC +TCCCTG +TCCCCA +TCCCCT +TCCCCC +TCCCCG +TCCCGA +TCCCGT +TCCCGC +TCCCGG +TCCGAA +TCCGAT +TCCGAC +TCCGAG +TCCGTA +TCCGTT +TCCGTC +TCCGTG +TCCGCA +TCCGCT +TCCGCC +TCCGCG +TCCGGA +TCCGGT +TCCGGC +TCCGGG +TCGAAA +TCGAAT +TCGAAC +TCGAAG +TCGATA +TCGATT +TCGATC +TCGATG +TCGACA +TCGACT +TCGACC +TCGACG +TCGAGA +TCGAGT +TCGAGC +TCGAGG +TCGTAA +TCGTAT +TCGTAC +TCGTAG +TCGTTA +TCGTTT +TCGTTC +TCGTTG +TCGTCA +TCGTCT +TCGTCC +TCGTCG +TCGTGA +TCGTGT +TCGTGC +TCGTGG +TCGCAA +TCGCAT +TCGCAC +TCGCAG +TCGCTA +TCGCTT +TCGCTC +TCGCTG +TCGCCA +TCGCCT +TCGCCC +TCGCCG +TCGCGA +TCGCGT +TCGCGC +TCGCGG +TCGGAA +TCGGAT +TCGGAC +TCGGAG +TCGGTA +TCGGTT +TCGGTC +TCGGTG +TCGGCA +TCGGCT +TCGGCC +TCGGCG +TCGGGA +TCGGGT +TCGGGC +TCGGGG +TGAAAA +TGAAAT +TGAAAC +TGAAAG +TGAATA +TGAATT +TGAATC +TGAATG +TGAACA +TGAACT +TGAACC +TGAACG +TGAAGA +TGAAGT +TGAAGC +TGAAGG +TGATAA +TGATAT +TGATAC +TGATAG +TGATTA +TGATTT +TGATTC +TGATTG +TGATCA +TGATCT +TGATCC +TGATCG +TGATGA +TGATGT +TGATGC +TGATGG +TGACAA +TGACAT +TGACAC +TGACAG +TGACTA +TGACTT +TGACTC +TGACTG +TGACCA +TGACCT +TGACCC +TGACCG +TGACGA +TGACGT +TGACGC +TGACGG +TGAGAA +TGAGAT +TGAGAC +TGAGAG +TGAGTA +TGAGTT +TGAGTC +TGAGTG +TGAGCA +TGAGCT +TGAGCC +TGAGCG +TGAGGA +TGAGGT +TGAGGC +TGAGGG +TGTAAA +TGTAAT +TGTAAC +TGTAAG +TGTATA +TGTATT +TGTATC +TGTATG +TGTACA +TGTACT +TGTACC +TGTACG +TGTAGA +TGTAGT +TGTAGC +TGTAGG +TGTTAA +TGTTAT +TGTTAC +TGTTAG +TGTTTA +TGTTTT +TGTTTC +TGTTTG +TGTTCA +TGTTCT +TGTTCC +TGTTCG +TGTTGA +TGTTGT +TGTTGC +TGTTGG +TGTCAA +TGTCAT +TGTCAC +TGTCAG +TGTCTA +TGTCTT +TGTCTC +TGTCTG +TGTCCA +TGTCCT +TGTCCC +TGTCCG +TGTCGA +TGTCGT +TGTCGC +TGTCGG +TGTGAA +TGTGAT +TGTGAC +TGTGAG +TGTGTA +TGTGTT +TGTGTC +TGTGTG +TGTGCA +TGTGCT +TGTGCC +TGTGCG +TGTGGA +TGTGGT +TGTGGC +TGTGGG +TGCAAA +TGCAAT +TGCAAC +TGCAAG +TGCATA +TGCATT +TGCATC +TGCATG +TGCACA +TGCACT +TGCACC +TGCACG +TGCAGA +TGCAGT +TGCAGC +TGCAGG +TGCTAA +TGCTAT +TGCTAC +TGCTAG +TGCTTA +TGCTTT +TGCTTC +TGCTTG +TGCTCA +TGCTCT +TGCTCC +TGCTCG +TGCTGA +TGCTGT +TGCTGC +TGCTGG +TGCCAA +TGCCAT +TGCCAC +TGCCAG +TGCCTA +TGCCTT +TGCCTC +TGCCTG +TGCCCA +TGCCCT +TGCCCC +TGCCCG +TGCCGA +TGCCGT +TGCCGC +TGCCGG +TGCGAA +TGCGAT +TGCGAC +TGCGAG +TGCGTA +TGCGTT +TGCGTC +TGCGTG +TGCGCA +TGCGCT +TGCGCC +TGCGCG +TGCGGA +TGCGGT +TGCGGC +TGCGGG +TGGAAA +TGGAAT +TGGAAC +TGGAAG +TGGATA +TGGATT +TGGATC +TGGATG +TGGACA +TGGACT +TGGACC +TGGACG +TGGAGA +TGGAGT +TGGAGC +TGGAGG +TGGTAA +TGGTAT +TGGTAC +TGGTAG +TGGTTA +TGGTTT +TGGTTC +TGGTTG +TGGTCA +TGGTCT +TGGTCC +TGGTCG +TGGTGA +TGGTGT +TGGTGC +TGGTGG +TGGCAA +TGGCAT +TGGCAC +TGGCAG +TGGCTA +TGGCTT +TGGCTC +TGGCTG +TGGCCA +TGGCCT +TGGCCC +TGGCCG +TGGCGA +TGGCGT +TGGCGC +TGGCGG +TGGGAA +TGGGAT +TGGGAC +TGGGAG +TGGGTA +TGGGTT +TGGGTC +TGGGTG +TGGGCA +TGGGCT +TGGGCC +TGGGCG +TGGGGA +TGGGGT +TGGGGC +TGGGGG +CAAAAA +CAAAAT +CAAAAC +CAAAAG +CAAATA +CAAATT +CAAATC +CAAATG +CAAACA +CAAACT +CAAACC +CAAACG +CAAAGA +CAAAGT +CAAAGC +CAAAGG +CAATAA +CAATAT +CAATAC +CAATAG +CAATTA +CAATTT +CAATTC +CAATTG +CAATCA +CAATCT +CAATCC +CAATCG +CAATGA +CAATGT +CAATGC +CAATGG +CAACAA +CAACAT +CAACAC +CAACAG +CAACTA +CAACTT +CAACTC +CAACTG +CAACCA +CAACCT +CAACCC +CAACCG +CAACGA +CAACGT +CAACGC +CAACGG +CAAGAA +CAAGAT +CAAGAC +CAAGAG +CAAGTA +CAAGTT +CAAGTC +CAAGTG +CAAGCA +CAAGCT +CAAGCC +CAAGCG +CAAGGA +CAAGGT +CAAGGC +CAAGGG +CATAAA +CATAAT +CATAAC +CATAAG +CATATA +CATATT +CATATC +CATATG +CATACA +CATACT +CATACC +CATACG +CATAGA +CATAGT +CATAGC +CATAGG +CATTAA +CATTAT +CATTAC +CATTAG +CATTTA +CATTTT +CATTTC +CATTTG +CATTCA +CATTCT +CATTCC +CATTCG +CATTGA +CATTGT +CATTGC +CATTGG +CATCAA +CATCAT +CATCAC +CATCAG +CATCTA +CATCTT +CATCTC +CATCTG +CATCCA +CATCCT +CATCCC +CATCCG +CATCGA +CATCGT +CATCGC +CATCGG +CATGAA +CATGAT +CATGAC +CATGAG +CATGTA +CATGTT +CATGTC +CATGTG +CATGCA +CATGCT +CATGCC +CATGCG +CATGGA +CATGGT +CATGGC +CATGGG +CACAAA +CACAAT +CACAAC +CACAAG +CACATA +CACATT +CACATC +CACATG +CACACA +CACACT +CACACC +CACACG +CACAGA +CACAGT +CACAGC +CACAGG +CACTAA +CACTAT +CACTAC +CACTAG +CACTTA +CACTTT +CACTTC +CACTTG +CACTCA +CACTCT +CACTCC +CACTCG +CACTGA +CACTGT +CACTGC +CACTGG +CACCAA +CACCAT +CACCAC +CACCAG +CACCTA +CACCTT +CACCTC +CACCTG +CACCCA +CACCCT +CACCCC +CACCCG +CACCGA +CACCGT +CACCGC +CACCGG +CACGAA +CACGAT +CACGAC +CACGAG +CACGTA +CACGTT +CACGTC +CACGTG +CACGCA +CACGCT +CACGCC +CACGCG +CACGGA +CACGGT +CACGGC +CACGGG +CAGAAA +CAGAAT +CAGAAC +CAGAAG +CAGATA +CAGATT +CAGATC +CAGATG +CAGACA +CAGACT +CAGACC +CAGACG +CAGAGA +CAGAGT +CAGAGC +CAGAGG +CAGTAA +CAGTAT +CAGTAC +CAGTAG +CAGTTA +CAGTTT +CAGTTC +CAGTTG +CAGTCA +CAGTCT +CAGTCC +CAGTCG +CAGTGA +CAGTGT +CAGTGC +CAGTGG +CAGCAA +CAGCAT +CAGCAC +CAGCAG +CAGCTA +CAGCTT +CAGCTC +CAGCTG +CAGCCA +CAGCCT +CAGCCC +CAGCCG +CAGCGA +CAGCGT +CAGCGC +CAGCGG +CAGGAA +CAGGAT +CAGGAC +CAGGAG +CAGGTA +CAGGTT +CAGGTC +CAGGTG +CAGGCA +CAGGCT +CAGGCC +CAGGCG +CAGGGA +CAGGGT +CAGGGC +CAGGGG +CTAAAA +CTAAAT +CTAAAC +CTAAAG +CTAATA +CTAATT +CTAATC +CTAATG +CTAACA +CTAACT +CTAACC +CTAACG +CTAAGA +CTAAGT +CTAAGC +CTAAGG +CTATAA +CTATAT +CTATAC +CTATAG +CTATTA +CTATTT +CTATTC +CTATTG +CTATCA +CTATCT +CTATCC +CTATCG +CTATGA +CTATGT +CTATGC +CTATGG +CTACAA +CTACAT +CTACAC +CTACAG +CTACTA +CTACTT +CTACTC +CTACTG +CTACCA +CTACCT +CTACCC +CTACCG +CTACGA +CTACGT +CTACGC +CTACGG +CTAGAA +CTAGAT +CTAGAC +CTAGAG +CTAGTA +CTAGTT +CTAGTC +CTAGTG +CTAGCA +CTAGCT +CTAGCC +CTAGCG +CTAGGA +CTAGGT +CTAGGC +CTAGGG +CTTAAA +CTTAAT +CTTAAC +CTTAAG +CTTATA +CTTATT +CTTATC +CTTATG +CTTACA +CTTACT +CTTACC +CTTACG +CTTAGA +CTTAGT +CTTAGC +CTTAGG +CTTTAA +CTTTAT +CTTTAC +CTTTAG +CTTTTA +CTTTTT +CTTTTC +CTTTTG +CTTTCA +CTTTCT +CTTTCC +CTTTCG +CTTTGA +CTTTGT +CTTTGC +CTTTGG +CTTCAA +CTTCAT +CTTCAC +CTTCAG +CTTCTA +CTTCTT +CTTCTC +CTTCTG +CTTCCA +CTTCCT +CTTCCC +CTTCCG +CTTCGA +CTTCGT +CTTCGC +CTTCGG +CTTGAA +CTTGAT +CTTGAC +CTTGAG +CTTGTA +CTTGTT +CTTGTC +CTTGTG +CTTGCA +CTTGCT +CTTGCC +CTTGCG +CTTGGA +CTTGGT +CTTGGC +CTTGGG +CTCAAA +CTCAAT +CTCAAC +CTCAAG +CTCATA +CTCATT +CTCATC +CTCATG +CTCACA +CTCACT +CTCACC +CTCACG +CTCAGA +CTCAGT +CTCAGC +CTCAGG +CTCTAA +CTCTAT +CTCTAC +CTCTAG +CTCTTA +CTCTTT +CTCTTC +CTCTTG +CTCTCA +CTCTCT +CTCTCC +CTCTCG +CTCTGA +CTCTGT +CTCTGC +CTCTGG +CTCCAA +CTCCAT +CTCCAC +CTCCAG +CTCCTA +CTCCTT +CTCCTC +CTCCTG +CTCCCA +CTCCCT +CTCCCC +CTCCCG +CTCCGA +CTCCGT +CTCCGC +CTCCGG +CTCGAA +CTCGAT +CTCGAC +CTCGAG +CTCGTA +CTCGTT +CTCGTC +CTCGTG +CTCGCA +CTCGCT +CTCGCC +CTCGCG +CTCGGA +CTCGGT +CTCGGC +CTCGGG +CTGAAA +CTGAAT +CTGAAC +CTGAAG +CTGATA +CTGATT +CTGATC +CTGATG +CTGACA +CTGACT +CTGACC +CTGACG +CTGAGA +CTGAGT +CTGAGC +CTGAGG +CTGTAA +CTGTAT +CTGTAC +CTGTAG +CTGTTA +CTGTTT +CTGTTC +CTGTTG +CTGTCA +CTGTCT +CTGTCC +CTGTCG +CTGTGA +CTGTGT +CTGTGC +CTGTGG +CTGCAA +CTGCAT +CTGCAC +CTGCAG +CTGCTA +CTGCTT +CTGCTC +CTGCTG +CTGCCA +CTGCCT +CTGCCC +CTGCCG +CTGCGA +CTGCGT +CTGCGC +CTGCGG +CTGGAA +CTGGAT +CTGGAC +CTGGAG +CTGGTA +CTGGTT +CTGGTC +CTGGTG +CTGGCA +CTGGCT +CTGGCC +CTGGCG +CTGGGA +CTGGGT +CTGGGC +CTGGGG +CCAAAA +CCAAAT +CCAAAC +CCAAAG +CCAATA +CCAATT +CCAATC +CCAATG +CCAACA +CCAACT +CCAACC +CCAACG +CCAAGA +CCAAGT +CCAAGC +CCAAGG +CCATAA +CCATAT +CCATAC +CCATAG +CCATTA +CCATTT +CCATTC +CCATTG +CCATCA +CCATCT +CCATCC +CCATCG +CCATGA +CCATGT +CCATGC +CCATGG +CCACAA +CCACAT +CCACAC +CCACAG +CCACTA +CCACTT +CCACTC +CCACTG +CCACCA +CCACCT +CCACCC +CCACCG +CCACGA +CCACGT +CCACGC +CCACGG +CCAGAA +CCAGAT +CCAGAC +CCAGAG +CCAGTA +CCAGTT +CCAGTC +CCAGTG +CCAGCA +CCAGCT +CCAGCC +CCAGCG +CCAGGA +CCAGGT +CCAGGC +CCAGGG +CCTAAA +CCTAAT +CCTAAC +CCTAAG +CCTATA +CCTATT +CCTATC +CCTATG +CCTACA +CCTACT +CCTACC +CCTACG +CCTAGA +CCTAGT +CCTAGC +CCTAGG +CCTTAA +CCTTAT +CCTTAC +CCTTAG +CCTTTA +CCTTTT +CCTTTC +CCTTTG +CCTTCA +CCTTCT +CCTTCC +CCTTCG +CCTTGA +CCTTGT +CCTTGC +CCTTGG +CCTCAA +CCTCAT +CCTCAC +CCTCAG +CCTCTA +CCTCTT +CCTCTC +CCTCTG +CCTCCA +CCTCCT +CCTCCC +CCTCCG +CCTCGA +CCTCGT +CCTCGC +CCTCGG +CCTGAA +CCTGAT +CCTGAC +CCTGAG +CCTGTA +CCTGTT +CCTGTC +CCTGTG +CCTGCA +CCTGCT +CCTGCC +CCTGCG +CCTGGA +CCTGGT +CCTGGC +CCTGGG +CCCAAA +CCCAAT +CCCAAC +CCCAAG +CCCATA +CCCATT +CCCATC +CCCATG +CCCACA +CCCACT +CCCACC +CCCACG +CCCAGA +CCCAGT +CCCAGC +CCCAGG +CCCTAA +CCCTAT +CCCTAC +CCCTAG +CCCTTA +CCCTTT +CCCTTC +CCCTTG +CCCTCA +CCCTCT +CCCTCC +CCCTCG +CCCTGA +CCCTGT +CCCTGC +CCCTGG +CCCCAA +CCCCAT +CCCCAC +CCCCAG +CCCCTA +CCCCTT +CCCCTC +CCCCTG +CCCCCA +CCCCCT +CCCCCC +CCCCCG +CCCCGA +CCCCGT +CCCCGC +CCCCGG +CCCGAA +CCCGAT +CCCGAC +CCCGAG +CCCGTA +CCCGTT +CCCGTC +CCCGTG +CCCGCA +CCCGCT +CCCGCC +CCCGCG +CCCGGA +CCCGGT +CCCGGC +CCCGGG +CCGAAA +CCGAAT +CCGAAC +CCGAAG +CCGATA +CCGATT +CCGATC +CCGATG +CCGACA +CCGACT +CCGACC +CCGACG +CCGAGA +CCGAGT +CCGAGC +CCGAGG +CCGTAA +CCGTAT +CCGTAC +CCGTAG +CCGTTA +CCGTTT +CCGTTC +CCGTTG +CCGTCA +CCGTCT +CCGTCC +CCGTCG +CCGTGA +CCGTGT +CCGTGC +CCGTGG +CCGCAA +CCGCAT +CCGCAC +CCGCAG +CCGCTA +CCGCTT +CCGCTC +CCGCTG +CCGCCA +CCGCCT +CCGCCC +CCGCCG +CCGCGA +CCGCGT +CCGCGC +CCGCGG +CCGGAA +CCGGAT +CCGGAC +CCGGAG +CCGGTA +CCGGTT +CCGGTC +CCGGTG +CCGGCA +CCGGCT +CCGGCC +CCGGCG +CCGGGA +CCGGGT +CCGGGC +CCGGGG +CGAAAA +CGAAAT +CGAAAC +CGAAAG +CGAATA +CGAATT +CGAATC +CGAATG +CGAACA +CGAACT +CGAACC +CGAACG +CGAAGA +CGAAGT +CGAAGC +CGAAGG +CGATAA +CGATAT +CGATAC +CGATAG +CGATTA +CGATTT +CGATTC +CGATTG +CGATCA +CGATCT +CGATCC +CGATCG +CGATGA +CGATGT +CGATGC +CGATGG +CGACAA +CGACAT +CGACAC +CGACAG +CGACTA +CGACTT +CGACTC +CGACTG +CGACCA +CGACCT +CGACCC +CGACCG +CGACGA +CGACGT +CGACGC +CGACGG +CGAGAA +CGAGAT +CGAGAC +CGAGAG +CGAGTA +CGAGTT +CGAGTC +CGAGTG +CGAGCA +CGAGCT +CGAGCC +CGAGCG +CGAGGA +CGAGGT +CGAGGC +CGAGGG +CGTAAA +CGTAAT +CGTAAC +CGTAAG +CGTATA +CGTATT +CGTATC +CGTATG +CGTACA +CGTACT +CGTACC +CGTACG +CGTAGA +CGTAGT +CGTAGC +CGTAGG +CGTTAA +CGTTAT +CGTTAC +CGTTAG +CGTTTA +CGTTTT +CGTTTC +CGTTTG +CGTTCA +CGTTCT +CGTTCC +CGTTCG +CGTTGA +CGTTGT +CGTTGC +CGTTGG +CGTCAA +CGTCAT +CGTCAC +CGTCAG +CGTCTA +CGTCTT +CGTCTC +CGTCTG +CGTCCA +CGTCCT +CGTCCC +CGTCCG +CGTCGA +CGTCGT +CGTCGC +CGTCGG +CGTGAA +CGTGAT +CGTGAC +CGTGAG +CGTGTA +CGTGTT +CGTGTC +CGTGTG +CGTGCA +CGTGCT +CGTGCC +CGTGCG +CGTGGA +CGTGGT +CGTGGC +CGTGGG +CGCAAA +CGCAAT +CGCAAC +CGCAAG +CGCATA +CGCATT +CGCATC +CGCATG +CGCACA +CGCACT +CGCACC +CGCACG +CGCAGA +CGCAGT +CGCAGC +CGCAGG +CGCTAA +CGCTAT +CGCTAC +CGCTAG +CGCTTA +CGCTTT +CGCTTC +CGCTTG +CGCTCA +CGCTCT +CGCTCC +CGCTCG +CGCTGA +CGCTGT +CGCTGC +CGCTGG +CGCCAA +CGCCAT +CGCCAC +CGCCAG +CGCCTA +CGCCTT +CGCCTC +CGCCTG +CGCCCA +CGCCCT +CGCCCC +CGCCCG +CGCCGA +CGCCGT +CGCCGC +CGCCGG +CGCGAA +CGCGAT +CGCGAC +CGCGAG +CGCGTA +CGCGTT +CGCGTC +CGCGTG +CGCGCA +CGCGCT +CGCGCC +CGCGCG +CGCGGA +CGCGGT +CGCGGC +CGCGGG +CGGAAA +CGGAAT +CGGAAC +CGGAAG +CGGATA +CGGATT +CGGATC +CGGATG +CGGACA +CGGACT +CGGACC +CGGACG +CGGAGA +CGGAGT +CGGAGC +CGGAGG +CGGTAA +CGGTAT +CGGTAC +CGGTAG +CGGTTA +CGGTTT +CGGTTC +CGGTTG +CGGTCA +CGGTCT +CGGTCC +CGGTCG +CGGTGA +CGGTGT +CGGTGC +CGGTGG +CGGCAA +CGGCAT +CGGCAC +CGGCAG +CGGCTA +CGGCTT +CGGCTC +CGGCTG +CGGCCA +CGGCCT +CGGCCC +CGGCCG +CGGCGA +CGGCGT +CGGCGC +CGGCGG +CGGGAA +CGGGAT +CGGGAC +CGGGAG +CGGGTA +CGGGTT +CGGGTC +CGGGTG +CGGGCA +CGGGCT +CGGGCC +CGGGCG +CGGGGA +CGGGGT +CGGGGC +CGGGGG +GAAAAA +GAAAAT +GAAAAC +GAAAAG +GAAATA +GAAATT +GAAATC +GAAATG +GAAACA +GAAACT +GAAACC +GAAACG +GAAAGA +GAAAGT +GAAAGC +GAAAGG +GAATAA +GAATAT +GAATAC +GAATAG +GAATTA +GAATTT +GAATTC +GAATTG +GAATCA +GAATCT +GAATCC +GAATCG +GAATGA +GAATGT +GAATGC +GAATGG +GAACAA +GAACAT +GAACAC +GAACAG +GAACTA +GAACTT +GAACTC +GAACTG +GAACCA +GAACCT +GAACCC +GAACCG +GAACGA +GAACGT +GAACGC +GAACGG +GAAGAA +GAAGAT +GAAGAC +GAAGAG +GAAGTA +GAAGTT +GAAGTC +GAAGTG +GAAGCA +GAAGCT +GAAGCC +GAAGCG +GAAGGA +GAAGGT +GAAGGC +GAAGGG +GATAAA +GATAAT +GATAAC +GATAAG +GATATA +GATATT +GATATC +GATATG +GATACA +GATACT +GATACC +GATACG +GATAGA +GATAGT +GATAGC +GATAGG +GATTAA +GATTAT +GATTAC +GATTAG +GATTTA +GATTTT +GATTTC +GATTTG +GATTCA +GATTCT +GATTCC +GATTCG +GATTGA +GATTGT +GATTGC +GATTGG +GATCAA +GATCAT +GATCAC +GATCAG +GATCTA +GATCTT +GATCTC +GATCTG +GATCCA +GATCCT +GATCCC +GATCCG +GATCGA +GATCGT +GATCGC +GATCGG +GATGAA +GATGAT +GATGAC +GATGAG +GATGTA +GATGTT +GATGTC +GATGTG +GATGCA +GATGCT +GATGCC +GATGCG +GATGGA +GATGGT +GATGGC +GATGGG +GACAAA +GACAAT +GACAAC +GACAAG +GACATA +GACATT +GACATC +GACATG +GACACA +GACACT +GACACC +GACACG +GACAGA +GACAGT +GACAGC +GACAGG +GACTAA +GACTAT +GACTAC +GACTAG +GACTTA +GACTTT +GACTTC +GACTTG +GACTCA +GACTCT +GACTCC +GACTCG +GACTGA +GACTGT +GACTGC +GACTGG +GACCAA +GACCAT +GACCAC +GACCAG +GACCTA +GACCTT +GACCTC +GACCTG +GACCCA +GACCCT +GACCCC +GACCCG +GACCGA +GACCGT +GACCGC +GACCGG +GACGAA +GACGAT +GACGAC +GACGAG +GACGTA +GACGTT +GACGTC +GACGTG +GACGCA +GACGCT +GACGCC +GACGCG +GACGGA +GACGGT +GACGGC +GACGGG +GAGAAA +GAGAAT +GAGAAC +GAGAAG +GAGATA +GAGATT +GAGATC +GAGATG +GAGACA +GAGACT +GAGACC +GAGACG +GAGAGA +GAGAGT +GAGAGC +GAGAGG +GAGTAA +GAGTAT +GAGTAC +GAGTAG +GAGTTA +GAGTTT +GAGTTC +GAGTTG +GAGTCA +GAGTCT +GAGTCC +GAGTCG +GAGTGA +GAGTGT +GAGTGC +GAGTGG +GAGCAA +GAGCAT +GAGCAC +GAGCAG +GAGCTA +GAGCTT +GAGCTC +GAGCTG +GAGCCA +GAGCCT +GAGCCC +GAGCCG +GAGCGA +GAGCGT +GAGCGC +GAGCGG +GAGGAA +GAGGAT +GAGGAC +GAGGAG +GAGGTA +GAGGTT +GAGGTC +GAGGTG +GAGGCA +GAGGCT +GAGGCC +GAGGCG +GAGGGA +GAGGGT +GAGGGC +GAGGGG +GTAAAA +GTAAAT +GTAAAC +GTAAAG +GTAATA +GTAATT +GTAATC +GTAATG +GTAACA +GTAACT +GTAACC +GTAACG +GTAAGA +GTAAGT +GTAAGC +GTAAGG +GTATAA +GTATAT +GTATAC +GTATAG +GTATTA +GTATTT +GTATTC +GTATTG +GTATCA +GTATCT +GTATCC +GTATCG +GTATGA +GTATGT +GTATGC +GTATGG +GTACAA +GTACAT +GTACAC +GTACAG +GTACTA +GTACTT +GTACTC +GTACTG +GTACCA +GTACCT +GTACCC +GTACCG +GTACGA +GTACGT +GTACGC +GTACGG +GTAGAA +GTAGAT +GTAGAC +GTAGAG +GTAGTA +GTAGTT +GTAGTC +GTAGTG +GTAGCA +GTAGCT +GTAGCC +GTAGCG +GTAGGA +GTAGGT +GTAGGC +GTAGGG +GTTAAA +GTTAAT +GTTAAC +GTTAAG +GTTATA +GTTATT +GTTATC +GTTATG +GTTACA +GTTACT +GTTACC +GTTACG +GTTAGA +GTTAGT +GTTAGC +GTTAGG +GTTTAA +GTTTAT +GTTTAC +GTTTAG +GTTTTA +GTTTTT +GTTTTC +GTTTTG +GTTTCA +GTTTCT +GTTTCC +GTTTCG +GTTTGA +GTTTGT +GTTTGC +GTTTGG +GTTCAA +GTTCAT +GTTCAC +GTTCAG +GTTCTA +GTTCTT +GTTCTC +GTTCTG +GTTCCA +GTTCCT +GTTCCC +GTTCCG +GTTCGA +GTTCGT +GTTCGC +GTTCGG +GTTGAA +GTTGAT +GTTGAC +GTTGAG +GTTGTA +GTTGTT +GTTGTC +GTTGTG +GTTGCA +GTTGCT +GTTGCC +GTTGCG +GTTGGA +GTTGGT +GTTGGC +GTTGGG +GTCAAA +GTCAAT +GTCAAC +GTCAAG +GTCATA +GTCATT +GTCATC +GTCATG +GTCACA +GTCACT +GTCACC +GTCACG +GTCAGA +GTCAGT +GTCAGC +GTCAGG +GTCTAA +GTCTAT +GTCTAC +GTCTAG +GTCTTA +GTCTTT +GTCTTC +GTCTTG +GTCTCA +GTCTCT +GTCTCC +GTCTCG +GTCTGA +GTCTGT +GTCTGC +GTCTGG +GTCCAA +GTCCAT +GTCCAC +GTCCAG +GTCCTA +GTCCTT +GTCCTC +GTCCTG +GTCCCA +GTCCCT +GTCCCC +GTCCCG +GTCCGA +GTCCGT +GTCCGC +GTCCGG +GTCGAA +GTCGAT +GTCGAC +GTCGAG +GTCGTA +GTCGTT +GTCGTC +GTCGTG +GTCGCA +GTCGCT +GTCGCC +GTCGCG +GTCGGA +GTCGGT +GTCGGC +GTCGGG +GTGAAA +GTGAAT +GTGAAC +GTGAAG +GTGATA +GTGATT +GTGATC +GTGATG +GTGACA +GTGACT +GTGACC +GTGACG +GTGAGA +GTGAGT +GTGAGC +GTGAGG +GTGTAA +GTGTAT +GTGTAC +GTGTAG +GTGTTA +GTGTTT +GTGTTC +GTGTTG +GTGTCA +GTGTCT +GTGTCC +GTGTCG +GTGTGA +GTGTGT +GTGTGC +GTGTGG +GTGCAA +GTGCAT +GTGCAC +GTGCAG +GTGCTA +GTGCTT +GTGCTC +GTGCTG +GTGCCA +GTGCCT +GTGCCC +GTGCCG +GTGCGA +GTGCGT +GTGCGC +GTGCGG +GTGGAA +GTGGAT +GTGGAC +GTGGAG +GTGGTA +GTGGTT +GTGGTC +GTGGTG +GTGGCA +GTGGCT +GTGGCC +GTGGCG +GTGGGA +GTGGGT +GTGGGC +GTGGGG +GCAAAA +GCAAAT +GCAAAC +GCAAAG +GCAATA +GCAATT +GCAATC +GCAATG +GCAACA +GCAACT +GCAACC +GCAACG +GCAAGA +GCAAGT +GCAAGC +GCAAGG +GCATAA +GCATAT +GCATAC +GCATAG +GCATTA +GCATTT +GCATTC +GCATTG +GCATCA +GCATCT +GCATCC +GCATCG +GCATGA +GCATGT +GCATGC +GCATGG +GCACAA +GCACAT +GCACAC +GCACAG +GCACTA +GCACTT +GCACTC +GCACTG +GCACCA +GCACCT +GCACCC +GCACCG +GCACGA +GCACGT +GCACGC +GCACGG +GCAGAA +GCAGAT +GCAGAC +GCAGAG +GCAGTA +GCAGTT +GCAGTC +GCAGTG +GCAGCA +GCAGCT +GCAGCC +GCAGCG +GCAGGA +GCAGGT +GCAGGC +GCAGGG +GCTAAA +GCTAAT +GCTAAC +GCTAAG +GCTATA +GCTATT +GCTATC +GCTATG +GCTACA +GCTACT +GCTACC +GCTACG +GCTAGA +GCTAGT +GCTAGC +GCTAGG +GCTTAA +GCTTAT +GCTTAC +GCTTAG +GCTTTA +GCTTTT +GCTTTC +GCTTTG +GCTTCA +GCTTCT +GCTTCC +GCTTCG +GCTTGA +GCTTGT +GCTTGC +GCTTGG +GCTCAA +GCTCAT +GCTCAC +GCTCAG +GCTCTA +GCTCTT +GCTCTC +GCTCTG +GCTCCA +GCTCCT +GCTCCC +GCTCCG +GCTCGA +GCTCGT +GCTCGC +GCTCGG +GCTGAA +GCTGAT +GCTGAC +GCTGAG +GCTGTA +GCTGTT +GCTGTC +GCTGTG +GCTGCA +GCTGCT +GCTGCC +GCTGCG +GCTGGA +GCTGGT +GCTGGC +GCTGGG +GCCAAA +GCCAAT +GCCAAC +GCCAAG +GCCATA +GCCATT +GCCATC +GCCATG +GCCACA +GCCACT +GCCACC +GCCACG +GCCAGA +GCCAGT +GCCAGC +GCCAGG +GCCTAA +GCCTAT +GCCTAC +GCCTAG +GCCTTA +GCCTTT +GCCTTC +GCCTTG +GCCTCA +GCCTCT +GCCTCC +GCCTCG +GCCTGA +GCCTGT +GCCTGC +GCCTGG +GCCCAA +GCCCAT +GCCCAC +GCCCAG +GCCCTA +GCCCTT +GCCCTC +GCCCTG +GCCCCA +GCCCCT +GCCCCC +GCCCCG +GCCCGA +GCCCGT +GCCCGC +GCCCGG +GCCGAA +GCCGAT +GCCGAC +GCCGAG +GCCGTA +GCCGTT +GCCGTC +GCCGTG +GCCGCA +GCCGCT +GCCGCC +GCCGCG +GCCGGA +GCCGGT +GCCGGC +GCCGGG +GCGAAA +GCGAAT +GCGAAC +GCGAAG +GCGATA +GCGATT +GCGATC +GCGATG +GCGACA +GCGACT +GCGACC +GCGACG +GCGAGA +GCGAGT +GCGAGC +GCGAGG +GCGTAA +GCGTAT +GCGTAC +GCGTAG +GCGTTA +GCGTTT +GCGTTC +GCGTTG +GCGTCA +GCGTCT +GCGTCC +GCGTCG +GCGTGA +GCGTGT +GCGTGC +GCGTGG +GCGCAA +GCGCAT +GCGCAC +GCGCAG +GCGCTA +GCGCTT +GCGCTC +GCGCTG +GCGCCA +GCGCCT +GCGCCC +GCGCCG +GCGCGA +GCGCGT +GCGCGC +GCGCGG +GCGGAA +GCGGAT +GCGGAC +GCGGAG +GCGGTA +GCGGTT +GCGGTC +GCGGTG +GCGGCA +GCGGCT +GCGGCC +GCGGCG +GCGGGA +GCGGGT +GCGGGC +GCGGGG +GGAAAA +GGAAAT +GGAAAC +GGAAAG +GGAATA +GGAATT +GGAATC +GGAATG +GGAACA +GGAACT +GGAACC +GGAACG +GGAAGA +GGAAGT +GGAAGC +GGAAGG +GGATAA +GGATAT +GGATAC +GGATAG +GGATTA +GGATTT +GGATTC +GGATTG +GGATCA +GGATCT +GGATCC +GGATCG +GGATGA +GGATGT +GGATGC +GGATGG +GGACAA +GGACAT +GGACAC +GGACAG +GGACTA +GGACTT +GGACTC +GGACTG +GGACCA +GGACCT +GGACCC +GGACCG +GGACGA +GGACGT +GGACGC +GGACGG +GGAGAA +GGAGAT +GGAGAC +GGAGAG +GGAGTA +GGAGTT +GGAGTC +GGAGTG +GGAGCA +GGAGCT +GGAGCC +GGAGCG +GGAGGA +GGAGGT +GGAGGC +GGAGGG +GGTAAA +GGTAAT +GGTAAC +GGTAAG +GGTATA +GGTATT +GGTATC +GGTATG +GGTACA +GGTACT +GGTACC +GGTACG +GGTAGA +GGTAGT +GGTAGC +GGTAGG +GGTTAA +GGTTAT +GGTTAC +GGTTAG +GGTTTA +GGTTTT +GGTTTC +GGTTTG +GGTTCA +GGTTCT +GGTTCC +GGTTCG +GGTTGA +GGTTGT +GGTTGC +GGTTGG +GGTCAA +GGTCAT +GGTCAC +GGTCAG +GGTCTA +GGTCTT +GGTCTC +GGTCTG +GGTCCA +GGTCCT +GGTCCC +GGTCCG +GGTCGA +GGTCGT +GGTCGC +GGTCGG +GGTGAA +GGTGAT +GGTGAC +GGTGAG +GGTGTA +GGTGTT +GGTGTC +GGTGTG +GGTGCA +GGTGCT +GGTGCC +GGTGCG +GGTGGA +GGTGGT +GGTGGC +GGTGGG +GGCAAA +GGCAAT +GGCAAC +GGCAAG +GGCATA +GGCATT +GGCATC +GGCATG +GGCACA +GGCACT +GGCACC +GGCACG +GGCAGA +GGCAGT +GGCAGC +GGCAGG +GGCTAA +GGCTAT +GGCTAC +GGCTAG +GGCTTA +GGCTTT +GGCTTC +GGCTTG +GGCTCA +GGCTCT +GGCTCC +GGCTCG +GGCTGA +GGCTGT +GGCTGC +GGCTGG +GGCCAA +GGCCAT +GGCCAC +GGCCAG +GGCCTA +GGCCTT +GGCCTC +GGCCTG +GGCCCA +GGCCCT +GGCCCC +GGCCCG +GGCCGA +GGCCGT +GGCCGC +GGCCGG +GGCGAA +GGCGAT +GGCGAC +GGCGAG +GGCGTA +GGCGTT +GGCGTC +GGCGTG +GGCGCA +GGCGCT +GGCGCC +GGCGCG +GGCGGA +GGCGGT +GGCGGC +GGCGGG +GGGAAA +GGGAAT +GGGAAC +GGGAAG +GGGATA +GGGATT +GGGATC +GGGATG +GGGACA +GGGACT +GGGACC +GGGACG +GGGAGA +GGGAGT +GGGAGC +GGGAGG +GGGTAA +GGGTAT +GGGTAC +GGGTAG +GGGTTA +GGGTTT +GGGTTC +GGGTTG +GGGTCA +GGGTCT +GGGTCC +GGGTCG +GGGTGA +GGGTGT +GGGTGC +GGGTGG +GGGCAA +GGGCAT +GGGCAC +GGGCAG +GGGCTA +GGGCTT +GGGCTC +GGGCTG +GGGCCA +GGGCCT +GGGCCC +GGGCCG +GGGCGA +GGGCGT +GGGCGC +GGGCGG +GGGGAA +GGGGAT +GGGGAC +GGGGAG +GGGGTA +GGGGTT +GGGGTC +GGGGTG +GGGGCA +GGGGCT +GGGGCC +GGGGCG +GGGGGA +GGGGGT +GGGGGC +GGGGGG diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-5000/config.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-5000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..9b14e61d327bf586fcff6e817cf9d365bae18ad2 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-5000/config.json @@ -0,0 +1,26 @@ +{ + "_name_or_path": "fabihamakhdoomi/TinyDNABERT", + "architectures": [ + "BertForSequenceClassification" + ], + "attention_probs_dropout_prob": 0.1, + "classifier_dropout": null, + "hidden_act": "gelu", + "hidden_dropout_prob": 0.1, + "hidden_size": 128, + "initializer_range": 0.02, + "intermediate_size": 512, + "layer_norm_eps": 1e-12, + "max_position_embeddings": 512, + "model_type": "bert", + "num_attention_heads": 2, + "num_hidden_layers": 2, + "pad_token_id": 0, + "position_embedding_type": "absolute", + "problem_type": "single_label_classification", + "torch_dtype": "float32", + "transformers_version": "4.44.0", + "type_vocab_size": 2, + "use_cache": true, + "vocab_size": 4101 +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-5000/model.safetensors b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-5000/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..f4525296a7300e692e98748307694a1b09d1596f --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-5000/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b287e909893304c26db84a7d4b3fe796c44afc878e930d373311e5dd44bc19ee +size 4021688 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-5000/optimizer.pt b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-5000/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..e3ffd96025bea70662e45d522db9bf3923da10f6 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-5000/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ac9850c4e99a728bac14b23fdeaa37b0e6367cc8b1926419348d11947edc3cdb +size 8058700 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-5000/rng_state.pth b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-5000/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..053218cb7d47f7e442d909ea8ed9bc9b286adfe8 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-5000/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:104246ae2bc9d8b31d442f0b88ec2d93b536dbfb9e32b9ff5076aee3a117c13d +size 14244 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-5000/scheduler.pt b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-5000/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..14aa799e6f5c93344d1b233b91d0bda9d8eb59e7 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-5000/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:eca4da5723e14faf16c3be96d23919390da5288d0d9db40021d332ee1df25b93 +size 1064 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-5000/special_tokens_map.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-5000/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..a8b3208c2884c4efb86e49300fdd3dc877220cdf --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-5000/special_tokens_map.json @@ -0,0 +1,7 @@ +{ + "cls_token": "[CLS]", + "mask_token": "[MASK]", + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-5000/tokenizer.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-5000/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..ebd782de6963b9032ce9374cfe30f5275be1a30b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-5000/tokenizer.json @@ -0,0 +1,4265 @@ +{ + "version": "1.0", + "truncation": { + "direction": "Right", + "max_length": 512, + "strategy": "LongestFirst", + "stride": 0 + }, + "padding": { + "strategy": { + "Fixed": 512 + }, + "direction": "Right", + "pad_to_multiple_of": null, + "pad_id": 0, + "pad_type_id": 0, + "pad_token": "[PAD]" + }, + "added_tokens": [ + { + "id": 0, + "content": "[PAD]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 1, + "content": "[UNK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 2, + "content": "[CLS]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 3, + "content": "[SEP]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 4, + "content": "[MASK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + } + ], + "normalizer": { + "type": "BertNormalizer", + "clean_text": true, + "handle_chinese_chars": true, + "strip_accents": null, + "lowercase": false + }, + "pre_tokenizer": { + "type": "BertPreTokenizer" + }, + "post_processor": { + "type": "TemplateProcessing", + "single": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + } + ], + "pair": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "B", + "type_id": 1 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 1 + } + } + ], + "special_tokens": { + "[CLS]": { + "id": "[CLS]", + "ids": [ + 2 + ], + "tokens": [ + "[CLS]" + ] + }, + "[SEP]": { + "id": "[SEP]", + "ids": [ + 3 + ], + "tokens": [ + "[SEP]" + ] + } + } + }, + "decoder": { + "type": "WordPiece", + "prefix": "##", + "cleanup": true + }, + "model": { + "type": "WordPiece", + "unk_token": "[UNK]", + "continuing_subword_prefix": "##", + "max_input_chars_per_word": 100, + "vocab": { + "[PAD]": 0, + "[UNK]": 1, + "[CLS]": 2, + "[SEP]": 3, + "[MASK]": 4, + "AAAAAA": 5, + "AAAAAT": 6, + "AAAAAC": 7, + "AAAAAG": 8, + "AAAATA": 9, + "AAAATT": 10, + "AAAATC": 11, + "AAAATG": 12, + "AAAACA": 13, + "AAAACT": 14, + "AAAACC": 15, + "AAAACG": 16, + "AAAAGA": 17, + "AAAAGT": 18, + "AAAAGC": 19, + "AAAAGG": 20, + "AAATAA": 21, + "AAATAT": 22, + "AAATAC": 23, + "AAATAG": 24, + "AAATTA": 25, + "AAATTT": 26, + "AAATTC": 27, + "AAATTG": 28, + "AAATCA": 29, + "AAATCT": 30, + "AAATCC": 31, + "AAATCG": 32, + "AAATGA": 33, + "AAATGT": 34, + "AAATGC": 35, + "AAATGG": 36, + "AAACAA": 37, + "AAACAT": 38, + "AAACAC": 39, + "AAACAG": 40, + "AAACTA": 41, + "AAACTT": 42, + "AAACTC": 43, + "AAACTG": 44, + "AAACCA": 45, + "AAACCT": 46, + "AAACCC": 47, + "AAACCG": 48, + "AAACGA": 49, + "AAACGT": 50, + "AAACGC": 51, + "AAACGG": 52, + "AAAGAA": 53, + "AAAGAT": 54, + "AAAGAC": 55, + "AAAGAG": 56, + "AAAGTA": 57, + "AAAGTT": 58, + "AAAGTC": 59, + "AAAGTG": 60, + "AAAGCA": 61, + "AAAGCT": 62, + "AAAGCC": 63, + "AAAGCG": 64, + "AAAGGA": 65, + "AAAGGT": 66, + "AAAGGC": 67, + "AAAGGG": 68, + "AATAAA": 69, + "AATAAT": 70, + "AATAAC": 71, + "AATAAG": 72, + "AATATA": 73, + "AATATT": 74, + "AATATC": 75, + "AATATG": 76, + "AATACA": 77, + "AATACT": 78, + "AATACC": 79, + "AATACG": 80, + "AATAGA": 81, + "AATAGT": 82, + "AATAGC": 83, + "AATAGG": 84, + "AATTAA": 85, + "AATTAT": 86, + "AATTAC": 87, + "AATTAG": 88, + "AATTTA": 89, + "AATTTT": 90, + "AATTTC": 91, + "AATTTG": 92, + "AATTCA": 93, + "AATTCT": 94, + "AATTCC": 95, + "AATTCG": 96, + "AATTGA": 97, + "AATTGT": 98, + "AATTGC": 99, + "AATTGG": 100, + "AATCAA": 101, + "AATCAT": 102, + "AATCAC": 103, + "AATCAG": 104, + "AATCTA": 105, + "AATCTT": 106, + "AATCTC": 107, + "AATCTG": 108, + "AATCCA": 109, + "AATCCT": 110, + "AATCCC": 111, + "AATCCG": 112, + "AATCGA": 113, + "AATCGT": 114, + "AATCGC": 115, + "AATCGG": 116, + "AATGAA": 117, + "AATGAT": 118, + "AATGAC": 119, + "AATGAG": 120, + "AATGTA": 121, + "AATGTT": 122, + "AATGTC": 123, + "AATGTG": 124, + "AATGCA": 125, + "AATGCT": 126, + "AATGCC": 127, + "AATGCG": 128, + "AATGGA": 129, + "AATGGT": 130, + "AATGGC": 131, + "AATGGG": 132, + "AACAAA": 133, + "AACAAT": 134, + "AACAAC": 135, + "AACAAG": 136, + "AACATA": 137, + "AACATT": 138, + "AACATC": 139, + "AACATG": 140, + "AACACA": 141, + "AACACT": 142, + "AACACC": 143, + "AACACG": 144, + "AACAGA": 145, + "AACAGT": 146, + "AACAGC": 147, + "AACAGG": 148, + "AACTAA": 149, + "AACTAT": 150, + "AACTAC": 151, + "AACTAG": 152, + "AACTTA": 153, + "AACTTT": 154, + "AACTTC": 155, + "AACTTG": 156, + "AACTCA": 157, + "AACTCT": 158, + "AACTCC": 159, + "AACTCG": 160, + "AACTGA": 161, + "AACTGT": 162, + "AACTGC": 163, + "AACTGG": 164, + "AACCAA": 165, + "AACCAT": 166, + "AACCAC": 167, + "AACCAG": 168, + "AACCTA": 169, + "AACCTT": 170, + "AACCTC": 171, + "AACCTG": 172, + "AACCCA": 173, + "AACCCT": 174, + "AACCCC": 175, + "AACCCG": 176, + "AACCGA": 177, + "AACCGT": 178, + "AACCGC": 179, + "AACCGG": 180, + "AACGAA": 181, + "AACGAT": 182, + "AACGAC": 183, + "AACGAG": 184, + "AACGTA": 185, + "AACGTT": 186, + "AACGTC": 187, + "AACGTG": 188, + "AACGCA": 189, + "AACGCT": 190, + "AACGCC": 191, + "AACGCG": 192, + "AACGGA": 193, + "AACGGT": 194, + "AACGGC": 195, + "AACGGG": 196, + "AAGAAA": 197, + "AAGAAT": 198, + "AAGAAC": 199, + "AAGAAG": 200, + "AAGATA": 201, + "AAGATT": 202, + "AAGATC": 203, + "AAGATG": 204, + "AAGACA": 205, + "AAGACT": 206, + "AAGACC": 207, + "AAGACG": 208, + "AAGAGA": 209, + "AAGAGT": 210, + "AAGAGC": 211, + "AAGAGG": 212, + "AAGTAA": 213, + "AAGTAT": 214, + "AAGTAC": 215, + "AAGTAG": 216, + "AAGTTA": 217, + "AAGTTT": 218, + "AAGTTC": 219, + "AAGTTG": 220, + "AAGTCA": 221, + "AAGTCT": 222, + "AAGTCC": 223, + "AAGTCG": 224, + "AAGTGA": 225, + "AAGTGT": 226, + "AAGTGC": 227, + "AAGTGG": 228, + "AAGCAA": 229, + "AAGCAT": 230, + "AAGCAC": 231, + "AAGCAG": 232, + "AAGCTA": 233, + "AAGCTT": 234, + "AAGCTC": 235, + "AAGCTG": 236, + "AAGCCA": 237, + "AAGCCT": 238, + "AAGCCC": 239, + "AAGCCG": 240, + "AAGCGA": 241, + "AAGCGT": 242, + "AAGCGC": 243, + "AAGCGG": 244, + "AAGGAA": 245, + "AAGGAT": 246, + "AAGGAC": 247, + "AAGGAG": 248, + "AAGGTA": 249, + "AAGGTT": 250, + "AAGGTC": 251, + "AAGGTG": 252, + "AAGGCA": 253, + "AAGGCT": 254, + "AAGGCC": 255, + "AAGGCG": 256, + "AAGGGA": 257, + "AAGGGT": 258, + "AAGGGC": 259, + "AAGGGG": 260, + "ATAAAA": 261, + "ATAAAT": 262, + "ATAAAC": 263, + "ATAAAG": 264, + "ATAATA": 265, + "ATAATT": 266, + "ATAATC": 267, + "ATAATG": 268, + "ATAACA": 269, + "ATAACT": 270, + "ATAACC": 271, + "ATAACG": 272, + "ATAAGA": 273, + "ATAAGT": 274, + "ATAAGC": 275, + "ATAAGG": 276, + "ATATAA": 277, + "ATATAT": 278, + "ATATAC": 279, + "ATATAG": 280, + "ATATTA": 281, + "ATATTT": 282, + "ATATTC": 283, + "ATATTG": 284, + "ATATCA": 285, + "ATATCT": 286, + "ATATCC": 287, + "ATATCG": 288, + "ATATGA": 289, + "ATATGT": 290, + "ATATGC": 291, + "ATATGG": 292, + "ATACAA": 293, + "ATACAT": 294, + "ATACAC": 295, + "ATACAG": 296, + "ATACTA": 297, + "ATACTT": 298, + "ATACTC": 299, + "ATACTG": 300, + "ATACCA": 301, + "ATACCT": 302, + "ATACCC": 303, + "ATACCG": 304, + "ATACGA": 305, + "ATACGT": 306, + "ATACGC": 307, + "ATACGG": 308, + "ATAGAA": 309, + "ATAGAT": 310, + "ATAGAC": 311, + "ATAGAG": 312, + "ATAGTA": 313, + "ATAGTT": 314, + "ATAGTC": 315, + "ATAGTG": 316, + "ATAGCA": 317, + "ATAGCT": 318, + "ATAGCC": 319, + "ATAGCG": 320, + "ATAGGA": 321, + "ATAGGT": 322, + "ATAGGC": 323, + "ATAGGG": 324, + "ATTAAA": 325, + "ATTAAT": 326, + "ATTAAC": 327, + "ATTAAG": 328, + "ATTATA": 329, + "ATTATT": 330, + "ATTATC": 331, + "ATTATG": 332, + "ATTACA": 333, + "ATTACT": 334, + "ATTACC": 335, + "ATTACG": 336, + "ATTAGA": 337, + "ATTAGT": 338, + "ATTAGC": 339, + "ATTAGG": 340, + "ATTTAA": 341, + "ATTTAT": 342, + "ATTTAC": 343, + "ATTTAG": 344, + "ATTTTA": 345, + "ATTTTT": 346, + "ATTTTC": 347, + "ATTTTG": 348, + "ATTTCA": 349, + "ATTTCT": 350, + "ATTTCC": 351, + "ATTTCG": 352, + "ATTTGA": 353, + "ATTTGT": 354, + "ATTTGC": 355, + "ATTTGG": 356, + "ATTCAA": 357, + "ATTCAT": 358, + "ATTCAC": 359, + "ATTCAG": 360, + "ATTCTA": 361, + "ATTCTT": 362, + "ATTCTC": 363, + "ATTCTG": 364, + "ATTCCA": 365, + "ATTCCT": 366, + "ATTCCC": 367, + "ATTCCG": 368, + "ATTCGA": 369, + "ATTCGT": 370, + "ATTCGC": 371, + "ATTCGG": 372, + "ATTGAA": 373, + "ATTGAT": 374, + "ATTGAC": 375, + "ATTGAG": 376, + "ATTGTA": 377, + "ATTGTT": 378, + "ATTGTC": 379, + "ATTGTG": 380, + "ATTGCA": 381, + "ATTGCT": 382, + "ATTGCC": 383, + "ATTGCG": 384, + "ATTGGA": 385, + "ATTGGT": 386, + "ATTGGC": 387, + "ATTGGG": 388, + "ATCAAA": 389, + "ATCAAT": 390, + "ATCAAC": 391, + "ATCAAG": 392, + "ATCATA": 393, + "ATCATT": 394, + "ATCATC": 395, + "ATCATG": 396, + "ATCACA": 397, + "ATCACT": 398, + "ATCACC": 399, + "ATCACG": 400, + "ATCAGA": 401, + "ATCAGT": 402, + "ATCAGC": 403, + "ATCAGG": 404, + "ATCTAA": 405, + "ATCTAT": 406, + "ATCTAC": 407, + "ATCTAG": 408, + "ATCTTA": 409, + "ATCTTT": 410, + "ATCTTC": 411, + "ATCTTG": 412, + "ATCTCA": 413, + "ATCTCT": 414, + "ATCTCC": 415, + "ATCTCG": 416, + "ATCTGA": 417, + "ATCTGT": 418, + "ATCTGC": 419, + "ATCTGG": 420, + "ATCCAA": 421, + "ATCCAT": 422, + "ATCCAC": 423, + "ATCCAG": 424, + "ATCCTA": 425, + "ATCCTT": 426, + "ATCCTC": 427, + "ATCCTG": 428, + "ATCCCA": 429, + "ATCCCT": 430, + "ATCCCC": 431, + "ATCCCG": 432, + "ATCCGA": 433, + "ATCCGT": 434, + "ATCCGC": 435, + "ATCCGG": 436, + "ATCGAA": 437, + "ATCGAT": 438, + "ATCGAC": 439, + "ATCGAG": 440, + "ATCGTA": 441, + "ATCGTT": 442, + "ATCGTC": 443, + "ATCGTG": 444, + "ATCGCA": 445, + "ATCGCT": 446, + "ATCGCC": 447, + "ATCGCG": 448, + "ATCGGA": 449, + "ATCGGT": 450, + "ATCGGC": 451, + "ATCGGG": 452, + "ATGAAA": 453, + "ATGAAT": 454, + "ATGAAC": 455, + "ATGAAG": 456, + "ATGATA": 457, + "ATGATT": 458, + "ATGATC": 459, + "ATGATG": 460, + "ATGACA": 461, + "ATGACT": 462, + "ATGACC": 463, + "ATGACG": 464, + "ATGAGA": 465, + "ATGAGT": 466, + "ATGAGC": 467, + "ATGAGG": 468, + "ATGTAA": 469, + "ATGTAT": 470, + "ATGTAC": 471, + "ATGTAG": 472, + "ATGTTA": 473, + "ATGTTT": 474, + "ATGTTC": 475, + "ATGTTG": 476, + "ATGTCA": 477, + "ATGTCT": 478, + "ATGTCC": 479, + "ATGTCG": 480, + "ATGTGA": 481, + "ATGTGT": 482, + "ATGTGC": 483, + "ATGTGG": 484, + "ATGCAA": 485, + "ATGCAT": 486, + "ATGCAC": 487, + "ATGCAG": 488, + "ATGCTA": 489, + "ATGCTT": 490, + "ATGCTC": 491, + "ATGCTG": 492, + "ATGCCA": 493, + "ATGCCT": 494, + "ATGCCC": 495, + "ATGCCG": 496, + "ATGCGA": 497, + "ATGCGT": 498, + "ATGCGC": 499, + "ATGCGG": 500, + "ATGGAA": 501, + "ATGGAT": 502, + "ATGGAC": 503, + "ATGGAG": 504, + "ATGGTA": 505, + "ATGGTT": 506, + "ATGGTC": 507, + "ATGGTG": 508, + "ATGGCA": 509, + "ATGGCT": 510, + "ATGGCC": 511, + "ATGGCG": 512, + "ATGGGA": 513, + "ATGGGT": 514, + "ATGGGC": 515, + "ATGGGG": 516, + "ACAAAA": 517, + "ACAAAT": 518, + "ACAAAC": 519, + "ACAAAG": 520, + "ACAATA": 521, + "ACAATT": 522, + "ACAATC": 523, + "ACAATG": 524, + "ACAACA": 525, + "ACAACT": 526, + "ACAACC": 527, + "ACAACG": 528, + "ACAAGA": 529, + "ACAAGT": 530, + "ACAAGC": 531, + "ACAAGG": 532, + "ACATAA": 533, + "ACATAT": 534, + "ACATAC": 535, + "ACATAG": 536, + "ACATTA": 537, + "ACATTT": 538, + "ACATTC": 539, + "ACATTG": 540, + "ACATCA": 541, + "ACATCT": 542, + "ACATCC": 543, + "ACATCG": 544, + "ACATGA": 545, + "ACATGT": 546, + "ACATGC": 547, + "ACATGG": 548, + "ACACAA": 549, + "ACACAT": 550, + "ACACAC": 551, + "ACACAG": 552, + "ACACTA": 553, + "ACACTT": 554, + "ACACTC": 555, + "ACACTG": 556, + "ACACCA": 557, + "ACACCT": 558, + "ACACCC": 559, + "ACACCG": 560, + "ACACGA": 561, + "ACACGT": 562, + "ACACGC": 563, + "ACACGG": 564, + "ACAGAA": 565, + "ACAGAT": 566, + "ACAGAC": 567, + "ACAGAG": 568, + "ACAGTA": 569, + "ACAGTT": 570, + "ACAGTC": 571, + "ACAGTG": 572, + "ACAGCA": 573, + "ACAGCT": 574, + "ACAGCC": 575, + "ACAGCG": 576, + "ACAGGA": 577, + "ACAGGT": 578, + "ACAGGC": 579, + "ACAGGG": 580, + "ACTAAA": 581, + "ACTAAT": 582, + "ACTAAC": 583, + "ACTAAG": 584, + "ACTATA": 585, + "ACTATT": 586, + "ACTATC": 587, + "ACTATG": 588, + "ACTACA": 589, + "ACTACT": 590, + "ACTACC": 591, + "ACTACG": 592, + "ACTAGA": 593, + "ACTAGT": 594, + "ACTAGC": 595, + "ACTAGG": 596, + "ACTTAA": 597, + "ACTTAT": 598, + "ACTTAC": 599, + "ACTTAG": 600, + "ACTTTA": 601, + "ACTTTT": 602, + "ACTTTC": 603, + "ACTTTG": 604, + "ACTTCA": 605, + "ACTTCT": 606, + "ACTTCC": 607, + "ACTTCG": 608, + "ACTTGA": 609, + "ACTTGT": 610, + "ACTTGC": 611, + "ACTTGG": 612, + "ACTCAA": 613, + "ACTCAT": 614, + "ACTCAC": 615, + "ACTCAG": 616, + "ACTCTA": 617, + "ACTCTT": 618, + "ACTCTC": 619, + "ACTCTG": 620, + "ACTCCA": 621, + "ACTCCT": 622, + "ACTCCC": 623, + "ACTCCG": 624, + "ACTCGA": 625, + "ACTCGT": 626, + "ACTCGC": 627, + "ACTCGG": 628, + "ACTGAA": 629, + "ACTGAT": 630, + "ACTGAC": 631, + "ACTGAG": 632, + "ACTGTA": 633, + "ACTGTT": 634, + "ACTGTC": 635, + "ACTGTG": 636, + "ACTGCA": 637, + "ACTGCT": 638, + "ACTGCC": 639, + "ACTGCG": 640, + "ACTGGA": 641, + "ACTGGT": 642, + "ACTGGC": 643, + "ACTGGG": 644, + "ACCAAA": 645, + "ACCAAT": 646, + "ACCAAC": 647, + "ACCAAG": 648, + "ACCATA": 649, + "ACCATT": 650, + "ACCATC": 651, + "ACCATG": 652, + "ACCACA": 653, + "ACCACT": 654, + "ACCACC": 655, + "ACCACG": 656, + "ACCAGA": 657, + "ACCAGT": 658, + "ACCAGC": 659, + "ACCAGG": 660, + "ACCTAA": 661, + "ACCTAT": 662, + "ACCTAC": 663, + "ACCTAG": 664, + "ACCTTA": 665, + "ACCTTT": 666, + "ACCTTC": 667, + "ACCTTG": 668, + "ACCTCA": 669, + "ACCTCT": 670, + "ACCTCC": 671, + "ACCTCG": 672, + "ACCTGA": 673, + "ACCTGT": 674, + "ACCTGC": 675, + "ACCTGG": 676, + "ACCCAA": 677, + "ACCCAT": 678, + "ACCCAC": 679, + "ACCCAG": 680, + "ACCCTA": 681, + "ACCCTT": 682, + "ACCCTC": 683, + "ACCCTG": 684, + "ACCCCA": 685, + "ACCCCT": 686, + "ACCCCC": 687, + "ACCCCG": 688, + "ACCCGA": 689, + "ACCCGT": 690, + "ACCCGC": 691, + "ACCCGG": 692, + "ACCGAA": 693, + "ACCGAT": 694, + "ACCGAC": 695, + "ACCGAG": 696, + "ACCGTA": 697, + "ACCGTT": 698, + "ACCGTC": 699, + "ACCGTG": 700, + "ACCGCA": 701, + "ACCGCT": 702, + "ACCGCC": 703, + "ACCGCG": 704, + "ACCGGA": 705, + "ACCGGT": 706, + "ACCGGC": 707, + "ACCGGG": 708, + "ACGAAA": 709, + "ACGAAT": 710, + "ACGAAC": 711, + "ACGAAG": 712, + "ACGATA": 713, + "ACGATT": 714, + "ACGATC": 715, + "ACGATG": 716, + "ACGACA": 717, + "ACGACT": 718, + "ACGACC": 719, + "ACGACG": 720, + "ACGAGA": 721, + "ACGAGT": 722, + "ACGAGC": 723, + "ACGAGG": 724, + "ACGTAA": 725, + "ACGTAT": 726, + "ACGTAC": 727, + "ACGTAG": 728, + "ACGTTA": 729, + "ACGTTT": 730, + "ACGTTC": 731, + "ACGTTG": 732, + "ACGTCA": 733, + "ACGTCT": 734, + "ACGTCC": 735, + "ACGTCG": 736, + "ACGTGA": 737, + "ACGTGT": 738, + "ACGTGC": 739, + "ACGTGG": 740, + "ACGCAA": 741, + "ACGCAT": 742, + "ACGCAC": 743, + "ACGCAG": 744, + "ACGCTA": 745, + "ACGCTT": 746, + "ACGCTC": 747, + "ACGCTG": 748, + "ACGCCA": 749, + "ACGCCT": 750, + "ACGCCC": 751, + "ACGCCG": 752, + "ACGCGA": 753, + "ACGCGT": 754, + "ACGCGC": 755, + "ACGCGG": 756, + "ACGGAA": 757, + "ACGGAT": 758, + "ACGGAC": 759, + "ACGGAG": 760, + "ACGGTA": 761, + "ACGGTT": 762, + "ACGGTC": 763, + "ACGGTG": 764, + "ACGGCA": 765, + "ACGGCT": 766, + "ACGGCC": 767, + "ACGGCG": 768, + "ACGGGA": 769, + "ACGGGT": 770, + "ACGGGC": 771, + "ACGGGG": 772, + "AGAAAA": 773, + "AGAAAT": 774, + "AGAAAC": 775, + "AGAAAG": 776, + "AGAATA": 777, + "AGAATT": 778, + "AGAATC": 779, + "AGAATG": 780, + "AGAACA": 781, + "AGAACT": 782, + "AGAACC": 783, + "AGAACG": 784, + "AGAAGA": 785, + "AGAAGT": 786, + "AGAAGC": 787, + "AGAAGG": 788, + "AGATAA": 789, + "AGATAT": 790, + "AGATAC": 791, + "AGATAG": 792, + "AGATTA": 793, + "AGATTT": 794, + "AGATTC": 795, + "AGATTG": 796, + "AGATCA": 797, + "AGATCT": 798, + "AGATCC": 799, + "AGATCG": 800, + "AGATGA": 801, + "AGATGT": 802, + "AGATGC": 803, + "AGATGG": 804, + "AGACAA": 805, + "AGACAT": 806, + "AGACAC": 807, + "AGACAG": 808, + "AGACTA": 809, + "AGACTT": 810, + "AGACTC": 811, + "AGACTG": 812, + "AGACCA": 813, + "AGACCT": 814, + "AGACCC": 815, + "AGACCG": 816, + "AGACGA": 817, + "AGACGT": 818, + "AGACGC": 819, + "AGACGG": 820, + "AGAGAA": 821, + "AGAGAT": 822, + "AGAGAC": 823, + "AGAGAG": 824, + "AGAGTA": 825, + "AGAGTT": 826, + "AGAGTC": 827, + "AGAGTG": 828, + "AGAGCA": 829, + "AGAGCT": 830, + "AGAGCC": 831, + "AGAGCG": 832, + "AGAGGA": 833, + "AGAGGT": 834, + "AGAGGC": 835, + "AGAGGG": 836, + "AGTAAA": 837, + "AGTAAT": 838, + "AGTAAC": 839, + "AGTAAG": 840, + "AGTATA": 841, + "AGTATT": 842, + "AGTATC": 843, + "AGTATG": 844, + "AGTACA": 845, + "AGTACT": 846, + "AGTACC": 847, + "AGTACG": 848, + "AGTAGA": 849, + "AGTAGT": 850, + "AGTAGC": 851, + "AGTAGG": 852, + "AGTTAA": 853, + "AGTTAT": 854, + "AGTTAC": 855, + "AGTTAG": 856, + "AGTTTA": 857, + "AGTTTT": 858, + "AGTTTC": 859, + "AGTTTG": 860, + "AGTTCA": 861, + "AGTTCT": 862, + "AGTTCC": 863, + "AGTTCG": 864, + "AGTTGA": 865, + "AGTTGT": 866, + "AGTTGC": 867, + "AGTTGG": 868, + "AGTCAA": 869, + "AGTCAT": 870, + "AGTCAC": 871, + "AGTCAG": 872, + "AGTCTA": 873, + "AGTCTT": 874, + "AGTCTC": 875, + "AGTCTG": 876, + "AGTCCA": 877, + "AGTCCT": 878, + "AGTCCC": 879, + "AGTCCG": 880, + "AGTCGA": 881, + "AGTCGT": 882, + "AGTCGC": 883, + "AGTCGG": 884, + "AGTGAA": 885, + "AGTGAT": 886, + "AGTGAC": 887, + "AGTGAG": 888, + "AGTGTA": 889, + "AGTGTT": 890, + "AGTGTC": 891, + "AGTGTG": 892, + "AGTGCA": 893, + "AGTGCT": 894, + "AGTGCC": 895, + "AGTGCG": 896, + "AGTGGA": 897, + "AGTGGT": 898, + "AGTGGC": 899, + "AGTGGG": 900, + "AGCAAA": 901, + "AGCAAT": 902, + "AGCAAC": 903, + "AGCAAG": 904, + "AGCATA": 905, + "AGCATT": 906, + "AGCATC": 907, + "AGCATG": 908, + "AGCACA": 909, + "AGCACT": 910, + "AGCACC": 911, + "AGCACG": 912, + "AGCAGA": 913, + "AGCAGT": 914, + "AGCAGC": 915, + "AGCAGG": 916, + "AGCTAA": 917, + "AGCTAT": 918, + "AGCTAC": 919, + "AGCTAG": 920, + "AGCTTA": 921, + "AGCTTT": 922, + "AGCTTC": 923, + "AGCTTG": 924, + "AGCTCA": 925, + "AGCTCT": 926, + "AGCTCC": 927, + "AGCTCG": 928, + "AGCTGA": 929, + "AGCTGT": 930, + "AGCTGC": 931, + "AGCTGG": 932, + "AGCCAA": 933, + "AGCCAT": 934, + "AGCCAC": 935, + "AGCCAG": 936, + "AGCCTA": 937, + "AGCCTT": 938, + "AGCCTC": 939, + "AGCCTG": 940, + "AGCCCA": 941, + "AGCCCT": 942, + "AGCCCC": 943, + "AGCCCG": 944, + "AGCCGA": 945, + "AGCCGT": 946, + "AGCCGC": 947, + "AGCCGG": 948, + "AGCGAA": 949, + "AGCGAT": 950, + "AGCGAC": 951, + "AGCGAG": 952, + "AGCGTA": 953, + "AGCGTT": 954, + "AGCGTC": 955, + "AGCGTG": 956, + "AGCGCA": 957, + "AGCGCT": 958, + "AGCGCC": 959, + "AGCGCG": 960, + "AGCGGA": 961, + "AGCGGT": 962, + "AGCGGC": 963, + "AGCGGG": 964, + "AGGAAA": 965, + "AGGAAT": 966, + "AGGAAC": 967, + "AGGAAG": 968, + "AGGATA": 969, + "AGGATT": 970, + "AGGATC": 971, + "AGGATG": 972, + "AGGACA": 973, + "AGGACT": 974, + "AGGACC": 975, + "AGGACG": 976, + "AGGAGA": 977, + "AGGAGT": 978, + "AGGAGC": 979, + "AGGAGG": 980, + "AGGTAA": 981, + "AGGTAT": 982, + "AGGTAC": 983, + "AGGTAG": 984, + "AGGTTA": 985, + "AGGTTT": 986, + "AGGTTC": 987, + "AGGTTG": 988, + "AGGTCA": 989, + "AGGTCT": 990, + "AGGTCC": 991, + "AGGTCG": 992, + "AGGTGA": 993, + "AGGTGT": 994, + "AGGTGC": 995, + "AGGTGG": 996, + "AGGCAA": 997, + "AGGCAT": 998, + "AGGCAC": 999, + "AGGCAG": 1000, + "AGGCTA": 1001, + "AGGCTT": 1002, + "AGGCTC": 1003, + "AGGCTG": 1004, + "AGGCCA": 1005, + "AGGCCT": 1006, + "AGGCCC": 1007, + "AGGCCG": 1008, + "AGGCGA": 1009, + "AGGCGT": 1010, + "AGGCGC": 1011, + "AGGCGG": 1012, + "AGGGAA": 1013, + "AGGGAT": 1014, + "AGGGAC": 1015, + "AGGGAG": 1016, + "AGGGTA": 1017, + "AGGGTT": 1018, + "AGGGTC": 1019, + "AGGGTG": 1020, + "AGGGCA": 1021, + "AGGGCT": 1022, + "AGGGCC": 1023, + "AGGGCG": 1024, + "AGGGGA": 1025, + "AGGGGT": 1026, + "AGGGGC": 1027, + "AGGGGG": 1028, + "TAAAAA": 1029, + "TAAAAT": 1030, + "TAAAAC": 1031, + "TAAAAG": 1032, + "TAAATA": 1033, + "TAAATT": 1034, + "TAAATC": 1035, + "TAAATG": 1036, + "TAAACA": 1037, + "TAAACT": 1038, + "TAAACC": 1039, + "TAAACG": 1040, + "TAAAGA": 1041, + "TAAAGT": 1042, + "TAAAGC": 1043, + "TAAAGG": 1044, + "TAATAA": 1045, + "TAATAT": 1046, + "TAATAC": 1047, + "TAATAG": 1048, + "TAATTA": 1049, + "TAATTT": 1050, + "TAATTC": 1051, + "TAATTG": 1052, + "TAATCA": 1053, + "TAATCT": 1054, + "TAATCC": 1055, + "TAATCG": 1056, + "TAATGA": 1057, + "TAATGT": 1058, + "TAATGC": 1059, + "TAATGG": 1060, + "TAACAA": 1061, + "TAACAT": 1062, + "TAACAC": 1063, + "TAACAG": 1064, + "TAACTA": 1065, + "TAACTT": 1066, + "TAACTC": 1067, + "TAACTG": 1068, + "TAACCA": 1069, + "TAACCT": 1070, + "TAACCC": 1071, + "TAACCG": 1072, + "TAACGA": 1073, + "TAACGT": 1074, + "TAACGC": 1075, + "TAACGG": 1076, + "TAAGAA": 1077, + "TAAGAT": 1078, + "TAAGAC": 1079, + "TAAGAG": 1080, + "TAAGTA": 1081, + "TAAGTT": 1082, + "TAAGTC": 1083, + "TAAGTG": 1084, + "TAAGCA": 1085, + "TAAGCT": 1086, + "TAAGCC": 1087, + "TAAGCG": 1088, + "TAAGGA": 1089, + "TAAGGT": 1090, + "TAAGGC": 1091, + "TAAGGG": 1092, + "TATAAA": 1093, + "TATAAT": 1094, + "TATAAC": 1095, + "TATAAG": 1096, + "TATATA": 1097, + "TATATT": 1098, + "TATATC": 1099, + "TATATG": 1100, + "TATACA": 1101, + "TATACT": 1102, + "TATACC": 1103, + "TATACG": 1104, + "TATAGA": 1105, + "TATAGT": 1106, + "TATAGC": 1107, + "TATAGG": 1108, + "TATTAA": 1109, + "TATTAT": 1110, + "TATTAC": 1111, + "TATTAG": 1112, + "TATTTA": 1113, + "TATTTT": 1114, + "TATTTC": 1115, + "TATTTG": 1116, + "TATTCA": 1117, + "TATTCT": 1118, + "TATTCC": 1119, + "TATTCG": 1120, + "TATTGA": 1121, + "TATTGT": 1122, + "TATTGC": 1123, + "TATTGG": 1124, + "TATCAA": 1125, + "TATCAT": 1126, + "TATCAC": 1127, + "TATCAG": 1128, + "TATCTA": 1129, + "TATCTT": 1130, + "TATCTC": 1131, + "TATCTG": 1132, + "TATCCA": 1133, + "TATCCT": 1134, + "TATCCC": 1135, + "TATCCG": 1136, + "TATCGA": 1137, + "TATCGT": 1138, + "TATCGC": 1139, + "TATCGG": 1140, + "TATGAA": 1141, + "TATGAT": 1142, + "TATGAC": 1143, + "TATGAG": 1144, + "TATGTA": 1145, + "TATGTT": 1146, + "TATGTC": 1147, + "TATGTG": 1148, + "TATGCA": 1149, + "TATGCT": 1150, + "TATGCC": 1151, + "TATGCG": 1152, + "TATGGA": 1153, + "TATGGT": 1154, + "TATGGC": 1155, + "TATGGG": 1156, + "TACAAA": 1157, + "TACAAT": 1158, + "TACAAC": 1159, + "TACAAG": 1160, + "TACATA": 1161, + "TACATT": 1162, + "TACATC": 1163, + "TACATG": 1164, + "TACACA": 1165, + "TACACT": 1166, + "TACACC": 1167, + "TACACG": 1168, + "TACAGA": 1169, + "TACAGT": 1170, + "TACAGC": 1171, + "TACAGG": 1172, + "TACTAA": 1173, + "TACTAT": 1174, + "TACTAC": 1175, + "TACTAG": 1176, + "TACTTA": 1177, + "TACTTT": 1178, + "TACTTC": 1179, + "TACTTG": 1180, + "TACTCA": 1181, + "TACTCT": 1182, + "TACTCC": 1183, + "TACTCG": 1184, + "TACTGA": 1185, + "TACTGT": 1186, + "TACTGC": 1187, + "TACTGG": 1188, + "TACCAA": 1189, + "TACCAT": 1190, + "TACCAC": 1191, + "TACCAG": 1192, + "TACCTA": 1193, + "TACCTT": 1194, + "TACCTC": 1195, + "TACCTG": 1196, + "TACCCA": 1197, + "TACCCT": 1198, + "TACCCC": 1199, + "TACCCG": 1200, + "TACCGA": 1201, + "TACCGT": 1202, + "TACCGC": 1203, + "TACCGG": 1204, + "TACGAA": 1205, + "TACGAT": 1206, + "TACGAC": 1207, + "TACGAG": 1208, + "TACGTA": 1209, + "TACGTT": 1210, + "TACGTC": 1211, + "TACGTG": 1212, + "TACGCA": 1213, + "TACGCT": 1214, + "TACGCC": 1215, + "TACGCG": 1216, + "TACGGA": 1217, + "TACGGT": 1218, + "TACGGC": 1219, + "TACGGG": 1220, + "TAGAAA": 1221, + "TAGAAT": 1222, + "TAGAAC": 1223, + "TAGAAG": 1224, + "TAGATA": 1225, + "TAGATT": 1226, + "TAGATC": 1227, + "TAGATG": 1228, + "TAGACA": 1229, + "TAGACT": 1230, + "TAGACC": 1231, + "TAGACG": 1232, + "TAGAGA": 1233, + "TAGAGT": 1234, + "TAGAGC": 1235, + "TAGAGG": 1236, + "TAGTAA": 1237, + "TAGTAT": 1238, + "TAGTAC": 1239, + "TAGTAG": 1240, + "TAGTTA": 1241, + "TAGTTT": 1242, + "TAGTTC": 1243, + "TAGTTG": 1244, + "TAGTCA": 1245, + "TAGTCT": 1246, + "TAGTCC": 1247, + "TAGTCG": 1248, + "TAGTGA": 1249, + "TAGTGT": 1250, + "TAGTGC": 1251, + "TAGTGG": 1252, + "TAGCAA": 1253, + "TAGCAT": 1254, + "TAGCAC": 1255, + "TAGCAG": 1256, + "TAGCTA": 1257, + "TAGCTT": 1258, + "TAGCTC": 1259, + "TAGCTG": 1260, + "TAGCCA": 1261, + "TAGCCT": 1262, + "TAGCCC": 1263, + "TAGCCG": 1264, + "TAGCGA": 1265, + "TAGCGT": 1266, + "TAGCGC": 1267, + "TAGCGG": 1268, + "TAGGAA": 1269, + "TAGGAT": 1270, + "TAGGAC": 1271, + "TAGGAG": 1272, + "TAGGTA": 1273, + "TAGGTT": 1274, + "TAGGTC": 1275, + "TAGGTG": 1276, + "TAGGCA": 1277, + "TAGGCT": 1278, + "TAGGCC": 1279, + "TAGGCG": 1280, + "TAGGGA": 1281, + "TAGGGT": 1282, + "TAGGGC": 1283, + "TAGGGG": 1284, + "TTAAAA": 1285, + "TTAAAT": 1286, + "TTAAAC": 1287, + "TTAAAG": 1288, + "TTAATA": 1289, + "TTAATT": 1290, + "TTAATC": 1291, + "TTAATG": 1292, + "TTAACA": 1293, + "TTAACT": 1294, + "TTAACC": 1295, + "TTAACG": 1296, + "TTAAGA": 1297, + "TTAAGT": 1298, + "TTAAGC": 1299, + "TTAAGG": 1300, + "TTATAA": 1301, + "TTATAT": 1302, + "TTATAC": 1303, + "TTATAG": 1304, + "TTATTA": 1305, + "TTATTT": 1306, + "TTATTC": 1307, + "TTATTG": 1308, + "TTATCA": 1309, + "TTATCT": 1310, + "TTATCC": 1311, + "TTATCG": 1312, + "TTATGA": 1313, + "TTATGT": 1314, + "TTATGC": 1315, + "TTATGG": 1316, + "TTACAA": 1317, + "TTACAT": 1318, + "TTACAC": 1319, + "TTACAG": 1320, + "TTACTA": 1321, + "TTACTT": 1322, + "TTACTC": 1323, + "TTACTG": 1324, + "TTACCA": 1325, + "TTACCT": 1326, + "TTACCC": 1327, + "TTACCG": 1328, + "TTACGA": 1329, + "TTACGT": 1330, + "TTACGC": 1331, + "TTACGG": 1332, + "TTAGAA": 1333, + "TTAGAT": 1334, + "TTAGAC": 1335, + "TTAGAG": 1336, + "TTAGTA": 1337, + "TTAGTT": 1338, + "TTAGTC": 1339, + "TTAGTG": 1340, + "TTAGCA": 1341, + "TTAGCT": 1342, + "TTAGCC": 1343, + "TTAGCG": 1344, + "TTAGGA": 1345, + "TTAGGT": 1346, + "TTAGGC": 1347, + "TTAGGG": 1348, + "TTTAAA": 1349, + "TTTAAT": 1350, + "TTTAAC": 1351, + "TTTAAG": 1352, + "TTTATA": 1353, + "TTTATT": 1354, + "TTTATC": 1355, + "TTTATG": 1356, + "TTTACA": 1357, + "TTTACT": 1358, + "TTTACC": 1359, + "TTTACG": 1360, + "TTTAGA": 1361, + "TTTAGT": 1362, + "TTTAGC": 1363, + "TTTAGG": 1364, + "TTTTAA": 1365, + "TTTTAT": 1366, + "TTTTAC": 1367, + "TTTTAG": 1368, + "TTTTTA": 1369, + "TTTTTT": 1370, + "TTTTTC": 1371, + "TTTTTG": 1372, + "TTTTCA": 1373, + "TTTTCT": 1374, + "TTTTCC": 1375, + "TTTTCG": 1376, + "TTTTGA": 1377, + "TTTTGT": 1378, + "TTTTGC": 1379, + "TTTTGG": 1380, + "TTTCAA": 1381, + "TTTCAT": 1382, + "TTTCAC": 1383, + "TTTCAG": 1384, + "TTTCTA": 1385, + "TTTCTT": 1386, + "TTTCTC": 1387, + "TTTCTG": 1388, + "TTTCCA": 1389, + "TTTCCT": 1390, + "TTTCCC": 1391, + "TTTCCG": 1392, + "TTTCGA": 1393, + "TTTCGT": 1394, + "TTTCGC": 1395, + "TTTCGG": 1396, + "TTTGAA": 1397, + "TTTGAT": 1398, + "TTTGAC": 1399, + "TTTGAG": 1400, + "TTTGTA": 1401, + "TTTGTT": 1402, + "TTTGTC": 1403, + "TTTGTG": 1404, + "TTTGCA": 1405, + "TTTGCT": 1406, + "TTTGCC": 1407, + "TTTGCG": 1408, + "TTTGGA": 1409, + "TTTGGT": 1410, + "TTTGGC": 1411, + "TTTGGG": 1412, + "TTCAAA": 1413, + "TTCAAT": 1414, + "TTCAAC": 1415, + "TTCAAG": 1416, + "TTCATA": 1417, + "TTCATT": 1418, + "TTCATC": 1419, + "TTCATG": 1420, + "TTCACA": 1421, + "TTCACT": 1422, + "TTCACC": 1423, + "TTCACG": 1424, + "TTCAGA": 1425, + "TTCAGT": 1426, + "TTCAGC": 1427, + "TTCAGG": 1428, + "TTCTAA": 1429, + "TTCTAT": 1430, + "TTCTAC": 1431, + "TTCTAG": 1432, + "TTCTTA": 1433, + "TTCTTT": 1434, + "TTCTTC": 1435, + "TTCTTG": 1436, + "TTCTCA": 1437, + "TTCTCT": 1438, + "TTCTCC": 1439, + "TTCTCG": 1440, + "TTCTGA": 1441, + "TTCTGT": 1442, + "TTCTGC": 1443, + "TTCTGG": 1444, + "TTCCAA": 1445, + "TTCCAT": 1446, + "TTCCAC": 1447, + "TTCCAG": 1448, + "TTCCTA": 1449, + "TTCCTT": 1450, + "TTCCTC": 1451, + "TTCCTG": 1452, + "TTCCCA": 1453, + "TTCCCT": 1454, + "TTCCCC": 1455, + "TTCCCG": 1456, + "TTCCGA": 1457, + "TTCCGT": 1458, + "TTCCGC": 1459, + "TTCCGG": 1460, + "TTCGAA": 1461, + "TTCGAT": 1462, + "TTCGAC": 1463, + "TTCGAG": 1464, + "TTCGTA": 1465, + "TTCGTT": 1466, + "TTCGTC": 1467, + "TTCGTG": 1468, + "TTCGCA": 1469, + "TTCGCT": 1470, + "TTCGCC": 1471, + "TTCGCG": 1472, + "TTCGGA": 1473, + "TTCGGT": 1474, + "TTCGGC": 1475, + "TTCGGG": 1476, + "TTGAAA": 1477, + "TTGAAT": 1478, + "TTGAAC": 1479, + "TTGAAG": 1480, + "TTGATA": 1481, + "TTGATT": 1482, + "TTGATC": 1483, + "TTGATG": 1484, + "TTGACA": 1485, + "TTGACT": 1486, + "TTGACC": 1487, + "TTGACG": 1488, + "TTGAGA": 1489, + "TTGAGT": 1490, + "TTGAGC": 1491, + "TTGAGG": 1492, + "TTGTAA": 1493, + "TTGTAT": 1494, + "TTGTAC": 1495, + "TTGTAG": 1496, + "TTGTTA": 1497, + "TTGTTT": 1498, + "TTGTTC": 1499, + "TTGTTG": 1500, + "TTGTCA": 1501, + "TTGTCT": 1502, + "TTGTCC": 1503, + "TTGTCG": 1504, + "TTGTGA": 1505, + "TTGTGT": 1506, + "TTGTGC": 1507, + "TTGTGG": 1508, + "TTGCAA": 1509, + "TTGCAT": 1510, + "TTGCAC": 1511, + "TTGCAG": 1512, + "TTGCTA": 1513, + "TTGCTT": 1514, + "TTGCTC": 1515, + "TTGCTG": 1516, + "TTGCCA": 1517, + "TTGCCT": 1518, + "TTGCCC": 1519, + "TTGCCG": 1520, + "TTGCGA": 1521, + "TTGCGT": 1522, + "TTGCGC": 1523, + "TTGCGG": 1524, + "TTGGAA": 1525, + "TTGGAT": 1526, + "TTGGAC": 1527, + "TTGGAG": 1528, + "TTGGTA": 1529, + "TTGGTT": 1530, + "TTGGTC": 1531, + "TTGGTG": 1532, + "TTGGCA": 1533, + "TTGGCT": 1534, + "TTGGCC": 1535, + "TTGGCG": 1536, + "TTGGGA": 1537, + "TTGGGT": 1538, + "TTGGGC": 1539, + "TTGGGG": 1540, + "TCAAAA": 1541, + "TCAAAT": 1542, + "TCAAAC": 1543, + "TCAAAG": 1544, + "TCAATA": 1545, + "TCAATT": 1546, + "TCAATC": 1547, + "TCAATG": 1548, + "TCAACA": 1549, + "TCAACT": 1550, + "TCAACC": 1551, + "TCAACG": 1552, + "TCAAGA": 1553, + "TCAAGT": 1554, + "TCAAGC": 1555, + "TCAAGG": 1556, + "TCATAA": 1557, + "TCATAT": 1558, + "TCATAC": 1559, + "TCATAG": 1560, + "TCATTA": 1561, + "TCATTT": 1562, + "TCATTC": 1563, + "TCATTG": 1564, + "TCATCA": 1565, + "TCATCT": 1566, + "TCATCC": 1567, + "TCATCG": 1568, + "TCATGA": 1569, + "TCATGT": 1570, + "TCATGC": 1571, + "TCATGG": 1572, + "TCACAA": 1573, + "TCACAT": 1574, + "TCACAC": 1575, + "TCACAG": 1576, + "TCACTA": 1577, + "TCACTT": 1578, + "TCACTC": 1579, + "TCACTG": 1580, + "TCACCA": 1581, + "TCACCT": 1582, + "TCACCC": 1583, + "TCACCG": 1584, + "TCACGA": 1585, + "TCACGT": 1586, + "TCACGC": 1587, + "TCACGG": 1588, + "TCAGAA": 1589, + "TCAGAT": 1590, + "TCAGAC": 1591, + "TCAGAG": 1592, + "TCAGTA": 1593, + "TCAGTT": 1594, + "TCAGTC": 1595, + "TCAGTG": 1596, + "TCAGCA": 1597, + "TCAGCT": 1598, + "TCAGCC": 1599, + "TCAGCG": 1600, + "TCAGGA": 1601, + "TCAGGT": 1602, + "TCAGGC": 1603, + "TCAGGG": 1604, + "TCTAAA": 1605, + "TCTAAT": 1606, + "TCTAAC": 1607, + "TCTAAG": 1608, + "TCTATA": 1609, + "TCTATT": 1610, + "TCTATC": 1611, + "TCTATG": 1612, + "TCTACA": 1613, + "TCTACT": 1614, + "TCTACC": 1615, + "TCTACG": 1616, + "TCTAGA": 1617, + "TCTAGT": 1618, + "TCTAGC": 1619, + "TCTAGG": 1620, + "TCTTAA": 1621, + "TCTTAT": 1622, + "TCTTAC": 1623, + "TCTTAG": 1624, + "TCTTTA": 1625, + "TCTTTT": 1626, + "TCTTTC": 1627, + "TCTTTG": 1628, + "TCTTCA": 1629, + "TCTTCT": 1630, + "TCTTCC": 1631, + "TCTTCG": 1632, + "TCTTGA": 1633, + "TCTTGT": 1634, + "TCTTGC": 1635, + "TCTTGG": 1636, + "TCTCAA": 1637, + "TCTCAT": 1638, + "TCTCAC": 1639, + "TCTCAG": 1640, + "TCTCTA": 1641, + "TCTCTT": 1642, + "TCTCTC": 1643, + "TCTCTG": 1644, + "TCTCCA": 1645, + "TCTCCT": 1646, + "TCTCCC": 1647, + "TCTCCG": 1648, + "TCTCGA": 1649, + "TCTCGT": 1650, + "TCTCGC": 1651, + "TCTCGG": 1652, + "TCTGAA": 1653, + "TCTGAT": 1654, + "TCTGAC": 1655, + "TCTGAG": 1656, + "TCTGTA": 1657, + "TCTGTT": 1658, + "TCTGTC": 1659, + "TCTGTG": 1660, + "TCTGCA": 1661, + "TCTGCT": 1662, + "TCTGCC": 1663, + "TCTGCG": 1664, + "TCTGGA": 1665, + "TCTGGT": 1666, + "TCTGGC": 1667, + "TCTGGG": 1668, + "TCCAAA": 1669, + "TCCAAT": 1670, + "TCCAAC": 1671, + "TCCAAG": 1672, + "TCCATA": 1673, + "TCCATT": 1674, + "TCCATC": 1675, + "TCCATG": 1676, + "TCCACA": 1677, + "TCCACT": 1678, + "TCCACC": 1679, + "TCCACG": 1680, + "TCCAGA": 1681, + "TCCAGT": 1682, + "TCCAGC": 1683, + "TCCAGG": 1684, + "TCCTAA": 1685, + "TCCTAT": 1686, + "TCCTAC": 1687, + "TCCTAG": 1688, + "TCCTTA": 1689, + "TCCTTT": 1690, + "TCCTTC": 1691, + "TCCTTG": 1692, + "TCCTCA": 1693, + "TCCTCT": 1694, + "TCCTCC": 1695, + "TCCTCG": 1696, + "TCCTGA": 1697, + "TCCTGT": 1698, + "TCCTGC": 1699, + "TCCTGG": 1700, + "TCCCAA": 1701, + "TCCCAT": 1702, + "TCCCAC": 1703, + "TCCCAG": 1704, + "TCCCTA": 1705, + "TCCCTT": 1706, + "TCCCTC": 1707, + "TCCCTG": 1708, + "TCCCCA": 1709, + "TCCCCT": 1710, + "TCCCCC": 1711, + "TCCCCG": 1712, + "TCCCGA": 1713, + "TCCCGT": 1714, + "TCCCGC": 1715, + "TCCCGG": 1716, + "TCCGAA": 1717, + "TCCGAT": 1718, + "TCCGAC": 1719, + "TCCGAG": 1720, + "TCCGTA": 1721, + "TCCGTT": 1722, + "TCCGTC": 1723, + "TCCGTG": 1724, + "TCCGCA": 1725, + "TCCGCT": 1726, + "TCCGCC": 1727, + "TCCGCG": 1728, + "TCCGGA": 1729, + "TCCGGT": 1730, + "TCCGGC": 1731, + "TCCGGG": 1732, + "TCGAAA": 1733, + "TCGAAT": 1734, + "TCGAAC": 1735, + "TCGAAG": 1736, + "TCGATA": 1737, + "TCGATT": 1738, + "TCGATC": 1739, + "TCGATG": 1740, + "TCGACA": 1741, + "TCGACT": 1742, + "TCGACC": 1743, + "TCGACG": 1744, + "TCGAGA": 1745, + "TCGAGT": 1746, + "TCGAGC": 1747, + "TCGAGG": 1748, + "TCGTAA": 1749, + "TCGTAT": 1750, + "TCGTAC": 1751, + "TCGTAG": 1752, + "TCGTTA": 1753, + "TCGTTT": 1754, + "TCGTTC": 1755, + "TCGTTG": 1756, + "TCGTCA": 1757, + "TCGTCT": 1758, + "TCGTCC": 1759, + "TCGTCG": 1760, + "TCGTGA": 1761, + "TCGTGT": 1762, + "TCGTGC": 1763, + "TCGTGG": 1764, + "TCGCAA": 1765, + "TCGCAT": 1766, + "TCGCAC": 1767, + "TCGCAG": 1768, + "TCGCTA": 1769, + "TCGCTT": 1770, + "TCGCTC": 1771, + "TCGCTG": 1772, + "TCGCCA": 1773, + "TCGCCT": 1774, + "TCGCCC": 1775, + "TCGCCG": 1776, + "TCGCGA": 1777, + "TCGCGT": 1778, + "TCGCGC": 1779, + "TCGCGG": 1780, + "TCGGAA": 1781, + "TCGGAT": 1782, + "TCGGAC": 1783, + "TCGGAG": 1784, + "TCGGTA": 1785, + "TCGGTT": 1786, + "TCGGTC": 1787, + "TCGGTG": 1788, + "TCGGCA": 1789, + "TCGGCT": 1790, + "TCGGCC": 1791, + "TCGGCG": 1792, + "TCGGGA": 1793, + "TCGGGT": 1794, + "TCGGGC": 1795, + "TCGGGG": 1796, + "TGAAAA": 1797, + "TGAAAT": 1798, + "TGAAAC": 1799, + "TGAAAG": 1800, + "TGAATA": 1801, + "TGAATT": 1802, + "TGAATC": 1803, + "TGAATG": 1804, + "TGAACA": 1805, + "TGAACT": 1806, + "TGAACC": 1807, + "TGAACG": 1808, + "TGAAGA": 1809, + "TGAAGT": 1810, + "TGAAGC": 1811, + "TGAAGG": 1812, + "TGATAA": 1813, + "TGATAT": 1814, + "TGATAC": 1815, + "TGATAG": 1816, + "TGATTA": 1817, + "TGATTT": 1818, + "TGATTC": 1819, + "TGATTG": 1820, + "TGATCA": 1821, + "TGATCT": 1822, + "TGATCC": 1823, + "TGATCG": 1824, + "TGATGA": 1825, + "TGATGT": 1826, + "TGATGC": 1827, + "TGATGG": 1828, + "TGACAA": 1829, + "TGACAT": 1830, + "TGACAC": 1831, + "TGACAG": 1832, + "TGACTA": 1833, + "TGACTT": 1834, + "TGACTC": 1835, + "TGACTG": 1836, + "TGACCA": 1837, + "TGACCT": 1838, + "TGACCC": 1839, + "TGACCG": 1840, + "TGACGA": 1841, + "TGACGT": 1842, + "TGACGC": 1843, + "TGACGG": 1844, + "TGAGAA": 1845, + "TGAGAT": 1846, + "TGAGAC": 1847, + "TGAGAG": 1848, + "TGAGTA": 1849, + "TGAGTT": 1850, + "TGAGTC": 1851, + "TGAGTG": 1852, + "TGAGCA": 1853, + "TGAGCT": 1854, + "TGAGCC": 1855, + "TGAGCG": 1856, + "TGAGGA": 1857, + "TGAGGT": 1858, + "TGAGGC": 1859, + "TGAGGG": 1860, + "TGTAAA": 1861, + "TGTAAT": 1862, + "TGTAAC": 1863, + "TGTAAG": 1864, + "TGTATA": 1865, + "TGTATT": 1866, + "TGTATC": 1867, + "TGTATG": 1868, + "TGTACA": 1869, + "TGTACT": 1870, + "TGTACC": 1871, + "TGTACG": 1872, + "TGTAGA": 1873, + "TGTAGT": 1874, + "TGTAGC": 1875, + "TGTAGG": 1876, + "TGTTAA": 1877, + "TGTTAT": 1878, + "TGTTAC": 1879, + "TGTTAG": 1880, + "TGTTTA": 1881, + "TGTTTT": 1882, + "TGTTTC": 1883, + "TGTTTG": 1884, + "TGTTCA": 1885, + "TGTTCT": 1886, + "TGTTCC": 1887, + "TGTTCG": 1888, + "TGTTGA": 1889, + "TGTTGT": 1890, + "TGTTGC": 1891, + "TGTTGG": 1892, + "TGTCAA": 1893, + "TGTCAT": 1894, + "TGTCAC": 1895, + "TGTCAG": 1896, + "TGTCTA": 1897, + "TGTCTT": 1898, + "TGTCTC": 1899, + "TGTCTG": 1900, + "TGTCCA": 1901, + "TGTCCT": 1902, + "TGTCCC": 1903, + "TGTCCG": 1904, + "TGTCGA": 1905, + "TGTCGT": 1906, + "TGTCGC": 1907, + "TGTCGG": 1908, + "TGTGAA": 1909, + "TGTGAT": 1910, + "TGTGAC": 1911, + "TGTGAG": 1912, + "TGTGTA": 1913, + "TGTGTT": 1914, + "TGTGTC": 1915, + "TGTGTG": 1916, + "TGTGCA": 1917, + "TGTGCT": 1918, + "TGTGCC": 1919, + "TGTGCG": 1920, + "TGTGGA": 1921, + "TGTGGT": 1922, + "TGTGGC": 1923, + "TGTGGG": 1924, + "TGCAAA": 1925, + "TGCAAT": 1926, + "TGCAAC": 1927, + "TGCAAG": 1928, + "TGCATA": 1929, + "TGCATT": 1930, + "TGCATC": 1931, + "TGCATG": 1932, + "TGCACA": 1933, + "TGCACT": 1934, + "TGCACC": 1935, + "TGCACG": 1936, + "TGCAGA": 1937, + "TGCAGT": 1938, + "TGCAGC": 1939, + "TGCAGG": 1940, + "TGCTAA": 1941, + "TGCTAT": 1942, + "TGCTAC": 1943, + "TGCTAG": 1944, + "TGCTTA": 1945, + "TGCTTT": 1946, + "TGCTTC": 1947, + "TGCTTG": 1948, + "TGCTCA": 1949, + "TGCTCT": 1950, + "TGCTCC": 1951, + "TGCTCG": 1952, + "TGCTGA": 1953, + "TGCTGT": 1954, + "TGCTGC": 1955, + "TGCTGG": 1956, + "TGCCAA": 1957, + "TGCCAT": 1958, + "TGCCAC": 1959, + "TGCCAG": 1960, + "TGCCTA": 1961, + "TGCCTT": 1962, + "TGCCTC": 1963, + "TGCCTG": 1964, + "TGCCCA": 1965, + "TGCCCT": 1966, + "TGCCCC": 1967, + "TGCCCG": 1968, + "TGCCGA": 1969, + "TGCCGT": 1970, + "TGCCGC": 1971, + "TGCCGG": 1972, + "TGCGAA": 1973, + "TGCGAT": 1974, + "TGCGAC": 1975, + "TGCGAG": 1976, + "TGCGTA": 1977, + "TGCGTT": 1978, + "TGCGTC": 1979, + "TGCGTG": 1980, + "TGCGCA": 1981, + "TGCGCT": 1982, + "TGCGCC": 1983, + "TGCGCG": 1984, + "TGCGGA": 1985, + "TGCGGT": 1986, + "TGCGGC": 1987, + "TGCGGG": 1988, + "TGGAAA": 1989, + "TGGAAT": 1990, + "TGGAAC": 1991, + "TGGAAG": 1992, + "TGGATA": 1993, + "TGGATT": 1994, + "TGGATC": 1995, + "TGGATG": 1996, + "TGGACA": 1997, + "TGGACT": 1998, + "TGGACC": 1999, + "TGGACG": 2000, + "TGGAGA": 2001, + "TGGAGT": 2002, + "TGGAGC": 2003, + "TGGAGG": 2004, + "TGGTAA": 2005, + "TGGTAT": 2006, + "TGGTAC": 2007, + "TGGTAG": 2008, + "TGGTTA": 2009, + "TGGTTT": 2010, + "TGGTTC": 2011, + "TGGTTG": 2012, + "TGGTCA": 2013, + "TGGTCT": 2014, + "TGGTCC": 2015, + "TGGTCG": 2016, + "TGGTGA": 2017, + "TGGTGT": 2018, + "TGGTGC": 2019, + "TGGTGG": 2020, + "TGGCAA": 2021, + "TGGCAT": 2022, + "TGGCAC": 2023, + "TGGCAG": 2024, + "TGGCTA": 2025, + "TGGCTT": 2026, + "TGGCTC": 2027, + "TGGCTG": 2028, + "TGGCCA": 2029, + "TGGCCT": 2030, + "TGGCCC": 2031, + "TGGCCG": 2032, + "TGGCGA": 2033, + "TGGCGT": 2034, + "TGGCGC": 2035, + "TGGCGG": 2036, + "TGGGAA": 2037, + "TGGGAT": 2038, + "TGGGAC": 2039, + "TGGGAG": 2040, + "TGGGTA": 2041, + "TGGGTT": 2042, + "TGGGTC": 2043, + "TGGGTG": 2044, + "TGGGCA": 2045, + "TGGGCT": 2046, + "TGGGCC": 2047, + "TGGGCG": 2048, + "TGGGGA": 2049, + "TGGGGT": 2050, + "TGGGGC": 2051, + "TGGGGG": 2052, + "CAAAAA": 2053, + "CAAAAT": 2054, + "CAAAAC": 2055, + "CAAAAG": 2056, + "CAAATA": 2057, + "CAAATT": 2058, + "CAAATC": 2059, + "CAAATG": 2060, + "CAAACA": 2061, + "CAAACT": 2062, + "CAAACC": 2063, + "CAAACG": 2064, + "CAAAGA": 2065, + "CAAAGT": 2066, + "CAAAGC": 2067, + "CAAAGG": 2068, + "CAATAA": 2069, + "CAATAT": 2070, + "CAATAC": 2071, + "CAATAG": 2072, + "CAATTA": 2073, + "CAATTT": 2074, + "CAATTC": 2075, + "CAATTG": 2076, + "CAATCA": 2077, + "CAATCT": 2078, + "CAATCC": 2079, + "CAATCG": 2080, + "CAATGA": 2081, + "CAATGT": 2082, + "CAATGC": 2083, + "CAATGG": 2084, + "CAACAA": 2085, + "CAACAT": 2086, + "CAACAC": 2087, + "CAACAG": 2088, + "CAACTA": 2089, + "CAACTT": 2090, + "CAACTC": 2091, + "CAACTG": 2092, + "CAACCA": 2093, + "CAACCT": 2094, + "CAACCC": 2095, + "CAACCG": 2096, + "CAACGA": 2097, + "CAACGT": 2098, + "CAACGC": 2099, + "CAACGG": 2100, + "CAAGAA": 2101, + "CAAGAT": 2102, + "CAAGAC": 2103, + "CAAGAG": 2104, + "CAAGTA": 2105, + "CAAGTT": 2106, + "CAAGTC": 2107, + "CAAGTG": 2108, + "CAAGCA": 2109, + "CAAGCT": 2110, + "CAAGCC": 2111, + "CAAGCG": 2112, + "CAAGGA": 2113, + "CAAGGT": 2114, + "CAAGGC": 2115, + "CAAGGG": 2116, + "CATAAA": 2117, + "CATAAT": 2118, + "CATAAC": 2119, + "CATAAG": 2120, + "CATATA": 2121, + "CATATT": 2122, + "CATATC": 2123, + "CATATG": 2124, + "CATACA": 2125, + "CATACT": 2126, + "CATACC": 2127, + "CATACG": 2128, + "CATAGA": 2129, + "CATAGT": 2130, + "CATAGC": 2131, + "CATAGG": 2132, + "CATTAA": 2133, + "CATTAT": 2134, + "CATTAC": 2135, + "CATTAG": 2136, + "CATTTA": 2137, + "CATTTT": 2138, + "CATTTC": 2139, + "CATTTG": 2140, + "CATTCA": 2141, + "CATTCT": 2142, + "CATTCC": 2143, + "CATTCG": 2144, + "CATTGA": 2145, + "CATTGT": 2146, + "CATTGC": 2147, + "CATTGG": 2148, + "CATCAA": 2149, + "CATCAT": 2150, + "CATCAC": 2151, + "CATCAG": 2152, + "CATCTA": 2153, + "CATCTT": 2154, + "CATCTC": 2155, + "CATCTG": 2156, + "CATCCA": 2157, + "CATCCT": 2158, + "CATCCC": 2159, + "CATCCG": 2160, + "CATCGA": 2161, + "CATCGT": 2162, + "CATCGC": 2163, + "CATCGG": 2164, + "CATGAA": 2165, + "CATGAT": 2166, + "CATGAC": 2167, + "CATGAG": 2168, + "CATGTA": 2169, + "CATGTT": 2170, + "CATGTC": 2171, + "CATGTG": 2172, + "CATGCA": 2173, + "CATGCT": 2174, + "CATGCC": 2175, + "CATGCG": 2176, + "CATGGA": 2177, + "CATGGT": 2178, + "CATGGC": 2179, + "CATGGG": 2180, + "CACAAA": 2181, + "CACAAT": 2182, + "CACAAC": 2183, + "CACAAG": 2184, + "CACATA": 2185, + "CACATT": 2186, + "CACATC": 2187, + "CACATG": 2188, + "CACACA": 2189, + "CACACT": 2190, + "CACACC": 2191, + "CACACG": 2192, + "CACAGA": 2193, + "CACAGT": 2194, + "CACAGC": 2195, + "CACAGG": 2196, + "CACTAA": 2197, + "CACTAT": 2198, + "CACTAC": 2199, + "CACTAG": 2200, + "CACTTA": 2201, + "CACTTT": 2202, + "CACTTC": 2203, + "CACTTG": 2204, + "CACTCA": 2205, + "CACTCT": 2206, + "CACTCC": 2207, + "CACTCG": 2208, + "CACTGA": 2209, + "CACTGT": 2210, + "CACTGC": 2211, + "CACTGG": 2212, + "CACCAA": 2213, + "CACCAT": 2214, + "CACCAC": 2215, + "CACCAG": 2216, + "CACCTA": 2217, + "CACCTT": 2218, + "CACCTC": 2219, + "CACCTG": 2220, + "CACCCA": 2221, + "CACCCT": 2222, + "CACCCC": 2223, + "CACCCG": 2224, + "CACCGA": 2225, + "CACCGT": 2226, + "CACCGC": 2227, + "CACCGG": 2228, + "CACGAA": 2229, + "CACGAT": 2230, + "CACGAC": 2231, + "CACGAG": 2232, + "CACGTA": 2233, + "CACGTT": 2234, + "CACGTC": 2235, + "CACGTG": 2236, + "CACGCA": 2237, + "CACGCT": 2238, + "CACGCC": 2239, + "CACGCG": 2240, + "CACGGA": 2241, + "CACGGT": 2242, + "CACGGC": 2243, + "CACGGG": 2244, + "CAGAAA": 2245, + "CAGAAT": 2246, + "CAGAAC": 2247, + "CAGAAG": 2248, + "CAGATA": 2249, + "CAGATT": 2250, + "CAGATC": 2251, + "CAGATG": 2252, + "CAGACA": 2253, + "CAGACT": 2254, + "CAGACC": 2255, + "CAGACG": 2256, + "CAGAGA": 2257, + "CAGAGT": 2258, + "CAGAGC": 2259, + "CAGAGG": 2260, + "CAGTAA": 2261, + "CAGTAT": 2262, + "CAGTAC": 2263, + "CAGTAG": 2264, + "CAGTTA": 2265, + "CAGTTT": 2266, + "CAGTTC": 2267, + "CAGTTG": 2268, + "CAGTCA": 2269, + "CAGTCT": 2270, + "CAGTCC": 2271, + "CAGTCG": 2272, + "CAGTGA": 2273, + "CAGTGT": 2274, + "CAGTGC": 2275, + "CAGTGG": 2276, + "CAGCAA": 2277, + "CAGCAT": 2278, + "CAGCAC": 2279, + "CAGCAG": 2280, + "CAGCTA": 2281, + "CAGCTT": 2282, + "CAGCTC": 2283, + "CAGCTG": 2284, + "CAGCCA": 2285, + "CAGCCT": 2286, + "CAGCCC": 2287, + "CAGCCG": 2288, + "CAGCGA": 2289, + "CAGCGT": 2290, + "CAGCGC": 2291, + "CAGCGG": 2292, + "CAGGAA": 2293, + "CAGGAT": 2294, + "CAGGAC": 2295, + "CAGGAG": 2296, + "CAGGTA": 2297, + "CAGGTT": 2298, + "CAGGTC": 2299, + "CAGGTG": 2300, + "CAGGCA": 2301, + "CAGGCT": 2302, + "CAGGCC": 2303, + "CAGGCG": 2304, + "CAGGGA": 2305, + "CAGGGT": 2306, + "CAGGGC": 2307, + "CAGGGG": 2308, + "CTAAAA": 2309, + "CTAAAT": 2310, + "CTAAAC": 2311, + "CTAAAG": 2312, + "CTAATA": 2313, + "CTAATT": 2314, + "CTAATC": 2315, + "CTAATG": 2316, + "CTAACA": 2317, + "CTAACT": 2318, + "CTAACC": 2319, + "CTAACG": 2320, + "CTAAGA": 2321, + "CTAAGT": 2322, + "CTAAGC": 2323, + "CTAAGG": 2324, + "CTATAA": 2325, + "CTATAT": 2326, + "CTATAC": 2327, + "CTATAG": 2328, + "CTATTA": 2329, + "CTATTT": 2330, + "CTATTC": 2331, + "CTATTG": 2332, + "CTATCA": 2333, + "CTATCT": 2334, + "CTATCC": 2335, + "CTATCG": 2336, + "CTATGA": 2337, + "CTATGT": 2338, + "CTATGC": 2339, + "CTATGG": 2340, + "CTACAA": 2341, + "CTACAT": 2342, + "CTACAC": 2343, + "CTACAG": 2344, + "CTACTA": 2345, + "CTACTT": 2346, + "CTACTC": 2347, + "CTACTG": 2348, + "CTACCA": 2349, + "CTACCT": 2350, + "CTACCC": 2351, + "CTACCG": 2352, + "CTACGA": 2353, + "CTACGT": 2354, + "CTACGC": 2355, + "CTACGG": 2356, + "CTAGAA": 2357, + "CTAGAT": 2358, + "CTAGAC": 2359, + "CTAGAG": 2360, + "CTAGTA": 2361, + "CTAGTT": 2362, + "CTAGTC": 2363, + "CTAGTG": 2364, + "CTAGCA": 2365, + "CTAGCT": 2366, + "CTAGCC": 2367, + "CTAGCG": 2368, + "CTAGGA": 2369, + "CTAGGT": 2370, + "CTAGGC": 2371, + "CTAGGG": 2372, + "CTTAAA": 2373, + "CTTAAT": 2374, + "CTTAAC": 2375, + "CTTAAG": 2376, + "CTTATA": 2377, + "CTTATT": 2378, + "CTTATC": 2379, + "CTTATG": 2380, + "CTTACA": 2381, + "CTTACT": 2382, + "CTTACC": 2383, + "CTTACG": 2384, + "CTTAGA": 2385, + "CTTAGT": 2386, + "CTTAGC": 2387, + "CTTAGG": 2388, + "CTTTAA": 2389, + "CTTTAT": 2390, + "CTTTAC": 2391, + "CTTTAG": 2392, + "CTTTTA": 2393, + "CTTTTT": 2394, + "CTTTTC": 2395, + "CTTTTG": 2396, + "CTTTCA": 2397, + "CTTTCT": 2398, + "CTTTCC": 2399, + "CTTTCG": 2400, + "CTTTGA": 2401, + "CTTTGT": 2402, + "CTTTGC": 2403, + "CTTTGG": 2404, + "CTTCAA": 2405, + "CTTCAT": 2406, + "CTTCAC": 2407, + "CTTCAG": 2408, + "CTTCTA": 2409, + "CTTCTT": 2410, + "CTTCTC": 2411, + "CTTCTG": 2412, + "CTTCCA": 2413, + "CTTCCT": 2414, + "CTTCCC": 2415, + "CTTCCG": 2416, + "CTTCGA": 2417, + "CTTCGT": 2418, + "CTTCGC": 2419, + "CTTCGG": 2420, + "CTTGAA": 2421, + "CTTGAT": 2422, + "CTTGAC": 2423, + "CTTGAG": 2424, + "CTTGTA": 2425, + "CTTGTT": 2426, + "CTTGTC": 2427, + "CTTGTG": 2428, + "CTTGCA": 2429, + "CTTGCT": 2430, + "CTTGCC": 2431, + "CTTGCG": 2432, + "CTTGGA": 2433, + "CTTGGT": 2434, + "CTTGGC": 2435, + "CTTGGG": 2436, + "CTCAAA": 2437, + "CTCAAT": 2438, + "CTCAAC": 2439, + "CTCAAG": 2440, + "CTCATA": 2441, + "CTCATT": 2442, + "CTCATC": 2443, + "CTCATG": 2444, + "CTCACA": 2445, + "CTCACT": 2446, + "CTCACC": 2447, + "CTCACG": 2448, + "CTCAGA": 2449, + "CTCAGT": 2450, + "CTCAGC": 2451, + "CTCAGG": 2452, + "CTCTAA": 2453, + "CTCTAT": 2454, + "CTCTAC": 2455, + "CTCTAG": 2456, + "CTCTTA": 2457, + "CTCTTT": 2458, + "CTCTTC": 2459, + "CTCTTG": 2460, + "CTCTCA": 2461, + "CTCTCT": 2462, + "CTCTCC": 2463, + "CTCTCG": 2464, + "CTCTGA": 2465, + "CTCTGT": 2466, + "CTCTGC": 2467, + "CTCTGG": 2468, + "CTCCAA": 2469, + "CTCCAT": 2470, + "CTCCAC": 2471, + "CTCCAG": 2472, + "CTCCTA": 2473, + "CTCCTT": 2474, + "CTCCTC": 2475, + "CTCCTG": 2476, + "CTCCCA": 2477, + "CTCCCT": 2478, + "CTCCCC": 2479, + "CTCCCG": 2480, + "CTCCGA": 2481, + "CTCCGT": 2482, + "CTCCGC": 2483, + "CTCCGG": 2484, + "CTCGAA": 2485, + "CTCGAT": 2486, + "CTCGAC": 2487, + "CTCGAG": 2488, + "CTCGTA": 2489, + "CTCGTT": 2490, + "CTCGTC": 2491, + "CTCGTG": 2492, + "CTCGCA": 2493, + "CTCGCT": 2494, + "CTCGCC": 2495, + "CTCGCG": 2496, + "CTCGGA": 2497, + "CTCGGT": 2498, + "CTCGGC": 2499, + "CTCGGG": 2500, + "CTGAAA": 2501, + "CTGAAT": 2502, + "CTGAAC": 2503, + "CTGAAG": 2504, + "CTGATA": 2505, + "CTGATT": 2506, + "CTGATC": 2507, + "CTGATG": 2508, + "CTGACA": 2509, + "CTGACT": 2510, + "CTGACC": 2511, + "CTGACG": 2512, + "CTGAGA": 2513, + "CTGAGT": 2514, + "CTGAGC": 2515, + "CTGAGG": 2516, + "CTGTAA": 2517, + "CTGTAT": 2518, + "CTGTAC": 2519, + "CTGTAG": 2520, + "CTGTTA": 2521, + "CTGTTT": 2522, + "CTGTTC": 2523, + "CTGTTG": 2524, + "CTGTCA": 2525, + "CTGTCT": 2526, + "CTGTCC": 2527, + "CTGTCG": 2528, + "CTGTGA": 2529, + "CTGTGT": 2530, + "CTGTGC": 2531, + "CTGTGG": 2532, + "CTGCAA": 2533, + "CTGCAT": 2534, + "CTGCAC": 2535, + "CTGCAG": 2536, + "CTGCTA": 2537, + "CTGCTT": 2538, + "CTGCTC": 2539, + "CTGCTG": 2540, + "CTGCCA": 2541, + "CTGCCT": 2542, + "CTGCCC": 2543, + "CTGCCG": 2544, + "CTGCGA": 2545, + "CTGCGT": 2546, + "CTGCGC": 2547, + "CTGCGG": 2548, + "CTGGAA": 2549, + "CTGGAT": 2550, + "CTGGAC": 2551, + "CTGGAG": 2552, + "CTGGTA": 2553, + "CTGGTT": 2554, + "CTGGTC": 2555, + "CTGGTG": 2556, + "CTGGCA": 2557, + "CTGGCT": 2558, + "CTGGCC": 2559, + "CTGGCG": 2560, + "CTGGGA": 2561, + "CTGGGT": 2562, + "CTGGGC": 2563, + "CTGGGG": 2564, + "CCAAAA": 2565, + "CCAAAT": 2566, + "CCAAAC": 2567, + "CCAAAG": 2568, + "CCAATA": 2569, + "CCAATT": 2570, + "CCAATC": 2571, + "CCAATG": 2572, + "CCAACA": 2573, + "CCAACT": 2574, + "CCAACC": 2575, + "CCAACG": 2576, + "CCAAGA": 2577, + "CCAAGT": 2578, + "CCAAGC": 2579, + "CCAAGG": 2580, + "CCATAA": 2581, + "CCATAT": 2582, + "CCATAC": 2583, + "CCATAG": 2584, + "CCATTA": 2585, + "CCATTT": 2586, + "CCATTC": 2587, + "CCATTG": 2588, + "CCATCA": 2589, + "CCATCT": 2590, + "CCATCC": 2591, + "CCATCG": 2592, + "CCATGA": 2593, + "CCATGT": 2594, + "CCATGC": 2595, + "CCATGG": 2596, + "CCACAA": 2597, + "CCACAT": 2598, + "CCACAC": 2599, + "CCACAG": 2600, + "CCACTA": 2601, + "CCACTT": 2602, + "CCACTC": 2603, + "CCACTG": 2604, + "CCACCA": 2605, + "CCACCT": 2606, + "CCACCC": 2607, + "CCACCG": 2608, + "CCACGA": 2609, + "CCACGT": 2610, + "CCACGC": 2611, + "CCACGG": 2612, + "CCAGAA": 2613, + "CCAGAT": 2614, + "CCAGAC": 2615, + "CCAGAG": 2616, + "CCAGTA": 2617, + "CCAGTT": 2618, + "CCAGTC": 2619, + "CCAGTG": 2620, + "CCAGCA": 2621, + "CCAGCT": 2622, + "CCAGCC": 2623, + "CCAGCG": 2624, + "CCAGGA": 2625, + "CCAGGT": 2626, + "CCAGGC": 2627, + "CCAGGG": 2628, + "CCTAAA": 2629, + "CCTAAT": 2630, + "CCTAAC": 2631, + "CCTAAG": 2632, + "CCTATA": 2633, + "CCTATT": 2634, + "CCTATC": 2635, + "CCTATG": 2636, + "CCTACA": 2637, + "CCTACT": 2638, + "CCTACC": 2639, + "CCTACG": 2640, + "CCTAGA": 2641, + "CCTAGT": 2642, + "CCTAGC": 2643, + "CCTAGG": 2644, + "CCTTAA": 2645, + "CCTTAT": 2646, + "CCTTAC": 2647, + "CCTTAG": 2648, + "CCTTTA": 2649, + "CCTTTT": 2650, + "CCTTTC": 2651, + "CCTTTG": 2652, + "CCTTCA": 2653, + "CCTTCT": 2654, + "CCTTCC": 2655, + "CCTTCG": 2656, + "CCTTGA": 2657, + "CCTTGT": 2658, + "CCTTGC": 2659, + "CCTTGG": 2660, + "CCTCAA": 2661, + "CCTCAT": 2662, + "CCTCAC": 2663, + "CCTCAG": 2664, + "CCTCTA": 2665, + "CCTCTT": 2666, + "CCTCTC": 2667, + "CCTCTG": 2668, + "CCTCCA": 2669, + "CCTCCT": 2670, + "CCTCCC": 2671, + "CCTCCG": 2672, + "CCTCGA": 2673, + "CCTCGT": 2674, + "CCTCGC": 2675, + "CCTCGG": 2676, + "CCTGAA": 2677, + "CCTGAT": 2678, + "CCTGAC": 2679, + "CCTGAG": 2680, + "CCTGTA": 2681, + "CCTGTT": 2682, + "CCTGTC": 2683, + "CCTGTG": 2684, + "CCTGCA": 2685, + "CCTGCT": 2686, + "CCTGCC": 2687, + "CCTGCG": 2688, + "CCTGGA": 2689, + "CCTGGT": 2690, + "CCTGGC": 2691, + "CCTGGG": 2692, + "CCCAAA": 2693, + "CCCAAT": 2694, + "CCCAAC": 2695, + "CCCAAG": 2696, + "CCCATA": 2697, + "CCCATT": 2698, + "CCCATC": 2699, + "CCCATG": 2700, + "CCCACA": 2701, + "CCCACT": 2702, + "CCCACC": 2703, + "CCCACG": 2704, + "CCCAGA": 2705, + "CCCAGT": 2706, + "CCCAGC": 2707, + "CCCAGG": 2708, + "CCCTAA": 2709, + "CCCTAT": 2710, + "CCCTAC": 2711, + "CCCTAG": 2712, + "CCCTTA": 2713, + "CCCTTT": 2714, + "CCCTTC": 2715, + "CCCTTG": 2716, + "CCCTCA": 2717, + "CCCTCT": 2718, + "CCCTCC": 2719, + "CCCTCG": 2720, + "CCCTGA": 2721, + "CCCTGT": 2722, + "CCCTGC": 2723, + "CCCTGG": 2724, + "CCCCAA": 2725, + "CCCCAT": 2726, + "CCCCAC": 2727, + "CCCCAG": 2728, + "CCCCTA": 2729, + "CCCCTT": 2730, + "CCCCTC": 2731, + "CCCCTG": 2732, + "CCCCCA": 2733, + "CCCCCT": 2734, + "CCCCCC": 2735, + "CCCCCG": 2736, + "CCCCGA": 2737, + "CCCCGT": 2738, + "CCCCGC": 2739, + "CCCCGG": 2740, + "CCCGAA": 2741, + "CCCGAT": 2742, + "CCCGAC": 2743, + "CCCGAG": 2744, + "CCCGTA": 2745, + "CCCGTT": 2746, + "CCCGTC": 2747, + "CCCGTG": 2748, + "CCCGCA": 2749, + "CCCGCT": 2750, + "CCCGCC": 2751, + "CCCGCG": 2752, + "CCCGGA": 2753, + "CCCGGT": 2754, + "CCCGGC": 2755, + "CCCGGG": 2756, + "CCGAAA": 2757, + "CCGAAT": 2758, + "CCGAAC": 2759, + "CCGAAG": 2760, + "CCGATA": 2761, + "CCGATT": 2762, + "CCGATC": 2763, + "CCGATG": 2764, + "CCGACA": 2765, + "CCGACT": 2766, + "CCGACC": 2767, + "CCGACG": 2768, + "CCGAGA": 2769, + "CCGAGT": 2770, + "CCGAGC": 2771, + "CCGAGG": 2772, + "CCGTAA": 2773, + "CCGTAT": 2774, + "CCGTAC": 2775, + "CCGTAG": 2776, + "CCGTTA": 2777, + "CCGTTT": 2778, + "CCGTTC": 2779, + "CCGTTG": 2780, + "CCGTCA": 2781, + "CCGTCT": 2782, + "CCGTCC": 2783, + "CCGTCG": 2784, + "CCGTGA": 2785, + "CCGTGT": 2786, + "CCGTGC": 2787, + "CCGTGG": 2788, + "CCGCAA": 2789, + "CCGCAT": 2790, + "CCGCAC": 2791, + "CCGCAG": 2792, + "CCGCTA": 2793, + "CCGCTT": 2794, + "CCGCTC": 2795, + "CCGCTG": 2796, + "CCGCCA": 2797, + "CCGCCT": 2798, + "CCGCCC": 2799, + "CCGCCG": 2800, + "CCGCGA": 2801, + "CCGCGT": 2802, + "CCGCGC": 2803, + "CCGCGG": 2804, + "CCGGAA": 2805, + "CCGGAT": 2806, + "CCGGAC": 2807, + "CCGGAG": 2808, + "CCGGTA": 2809, + "CCGGTT": 2810, + "CCGGTC": 2811, + "CCGGTG": 2812, + "CCGGCA": 2813, + "CCGGCT": 2814, + "CCGGCC": 2815, + "CCGGCG": 2816, + "CCGGGA": 2817, + "CCGGGT": 2818, + "CCGGGC": 2819, + "CCGGGG": 2820, + "CGAAAA": 2821, + "CGAAAT": 2822, + "CGAAAC": 2823, + "CGAAAG": 2824, + "CGAATA": 2825, + "CGAATT": 2826, + "CGAATC": 2827, + "CGAATG": 2828, + "CGAACA": 2829, + "CGAACT": 2830, + "CGAACC": 2831, + "CGAACG": 2832, + "CGAAGA": 2833, + "CGAAGT": 2834, + "CGAAGC": 2835, + "CGAAGG": 2836, + "CGATAA": 2837, + "CGATAT": 2838, + "CGATAC": 2839, + "CGATAG": 2840, + "CGATTA": 2841, + "CGATTT": 2842, + "CGATTC": 2843, + "CGATTG": 2844, + "CGATCA": 2845, + "CGATCT": 2846, + "CGATCC": 2847, + "CGATCG": 2848, + "CGATGA": 2849, + "CGATGT": 2850, + "CGATGC": 2851, + "CGATGG": 2852, + "CGACAA": 2853, + "CGACAT": 2854, + "CGACAC": 2855, + "CGACAG": 2856, + "CGACTA": 2857, + "CGACTT": 2858, + "CGACTC": 2859, + "CGACTG": 2860, + "CGACCA": 2861, + "CGACCT": 2862, + "CGACCC": 2863, + "CGACCG": 2864, + "CGACGA": 2865, + "CGACGT": 2866, + "CGACGC": 2867, + "CGACGG": 2868, + "CGAGAA": 2869, + "CGAGAT": 2870, + "CGAGAC": 2871, + "CGAGAG": 2872, + "CGAGTA": 2873, + "CGAGTT": 2874, + "CGAGTC": 2875, + "CGAGTG": 2876, + "CGAGCA": 2877, + "CGAGCT": 2878, + "CGAGCC": 2879, + "CGAGCG": 2880, + "CGAGGA": 2881, + "CGAGGT": 2882, + "CGAGGC": 2883, + "CGAGGG": 2884, + "CGTAAA": 2885, + "CGTAAT": 2886, + "CGTAAC": 2887, + "CGTAAG": 2888, + "CGTATA": 2889, + "CGTATT": 2890, + "CGTATC": 2891, + "CGTATG": 2892, + "CGTACA": 2893, + "CGTACT": 2894, + "CGTACC": 2895, + "CGTACG": 2896, + "CGTAGA": 2897, + "CGTAGT": 2898, + "CGTAGC": 2899, + "CGTAGG": 2900, + "CGTTAA": 2901, + "CGTTAT": 2902, + "CGTTAC": 2903, + "CGTTAG": 2904, + "CGTTTA": 2905, + "CGTTTT": 2906, + "CGTTTC": 2907, + "CGTTTG": 2908, + "CGTTCA": 2909, + "CGTTCT": 2910, + "CGTTCC": 2911, + "CGTTCG": 2912, + "CGTTGA": 2913, + "CGTTGT": 2914, + "CGTTGC": 2915, + "CGTTGG": 2916, + "CGTCAA": 2917, + "CGTCAT": 2918, + "CGTCAC": 2919, + "CGTCAG": 2920, + "CGTCTA": 2921, + "CGTCTT": 2922, + "CGTCTC": 2923, + "CGTCTG": 2924, + "CGTCCA": 2925, + "CGTCCT": 2926, + "CGTCCC": 2927, + "CGTCCG": 2928, + "CGTCGA": 2929, + "CGTCGT": 2930, + "CGTCGC": 2931, + "CGTCGG": 2932, + "CGTGAA": 2933, + "CGTGAT": 2934, + "CGTGAC": 2935, + "CGTGAG": 2936, + "CGTGTA": 2937, + "CGTGTT": 2938, + "CGTGTC": 2939, + "CGTGTG": 2940, + "CGTGCA": 2941, + "CGTGCT": 2942, + "CGTGCC": 2943, + "CGTGCG": 2944, + "CGTGGA": 2945, + "CGTGGT": 2946, + "CGTGGC": 2947, + "CGTGGG": 2948, + "CGCAAA": 2949, + "CGCAAT": 2950, + "CGCAAC": 2951, + "CGCAAG": 2952, + "CGCATA": 2953, + "CGCATT": 2954, + "CGCATC": 2955, + "CGCATG": 2956, + "CGCACA": 2957, + "CGCACT": 2958, + "CGCACC": 2959, + "CGCACG": 2960, + "CGCAGA": 2961, + "CGCAGT": 2962, + "CGCAGC": 2963, + "CGCAGG": 2964, + "CGCTAA": 2965, + "CGCTAT": 2966, + "CGCTAC": 2967, + "CGCTAG": 2968, + "CGCTTA": 2969, + "CGCTTT": 2970, + "CGCTTC": 2971, + "CGCTTG": 2972, + "CGCTCA": 2973, + "CGCTCT": 2974, + "CGCTCC": 2975, + "CGCTCG": 2976, + "CGCTGA": 2977, + "CGCTGT": 2978, + "CGCTGC": 2979, + "CGCTGG": 2980, + "CGCCAA": 2981, + "CGCCAT": 2982, + "CGCCAC": 2983, + "CGCCAG": 2984, + "CGCCTA": 2985, + "CGCCTT": 2986, + "CGCCTC": 2987, + "CGCCTG": 2988, + "CGCCCA": 2989, + "CGCCCT": 2990, + "CGCCCC": 2991, + "CGCCCG": 2992, + "CGCCGA": 2993, + "CGCCGT": 2994, + "CGCCGC": 2995, + "CGCCGG": 2996, + "CGCGAA": 2997, + "CGCGAT": 2998, + "CGCGAC": 2999, + "CGCGAG": 3000, + "CGCGTA": 3001, + "CGCGTT": 3002, + "CGCGTC": 3003, + "CGCGTG": 3004, + "CGCGCA": 3005, + "CGCGCT": 3006, + "CGCGCC": 3007, + "CGCGCG": 3008, + "CGCGGA": 3009, + "CGCGGT": 3010, + "CGCGGC": 3011, + "CGCGGG": 3012, + "CGGAAA": 3013, + "CGGAAT": 3014, + "CGGAAC": 3015, + "CGGAAG": 3016, + "CGGATA": 3017, + "CGGATT": 3018, + "CGGATC": 3019, + "CGGATG": 3020, + "CGGACA": 3021, + "CGGACT": 3022, + "CGGACC": 3023, + "CGGACG": 3024, + "CGGAGA": 3025, + "CGGAGT": 3026, + "CGGAGC": 3027, + "CGGAGG": 3028, + "CGGTAA": 3029, + "CGGTAT": 3030, + "CGGTAC": 3031, + "CGGTAG": 3032, + "CGGTTA": 3033, + "CGGTTT": 3034, + "CGGTTC": 3035, + "CGGTTG": 3036, + "CGGTCA": 3037, + "CGGTCT": 3038, + "CGGTCC": 3039, + "CGGTCG": 3040, + "CGGTGA": 3041, + "CGGTGT": 3042, + "CGGTGC": 3043, + "CGGTGG": 3044, + "CGGCAA": 3045, + "CGGCAT": 3046, + "CGGCAC": 3047, + "CGGCAG": 3048, + "CGGCTA": 3049, + "CGGCTT": 3050, + "CGGCTC": 3051, + "CGGCTG": 3052, + "CGGCCA": 3053, + "CGGCCT": 3054, + "CGGCCC": 3055, + "CGGCCG": 3056, + "CGGCGA": 3057, + "CGGCGT": 3058, + "CGGCGC": 3059, + "CGGCGG": 3060, + "CGGGAA": 3061, + "CGGGAT": 3062, + "CGGGAC": 3063, + "CGGGAG": 3064, + "CGGGTA": 3065, + "CGGGTT": 3066, + "CGGGTC": 3067, + "CGGGTG": 3068, + "CGGGCA": 3069, + "CGGGCT": 3070, + "CGGGCC": 3071, + "CGGGCG": 3072, + "CGGGGA": 3073, + "CGGGGT": 3074, + "CGGGGC": 3075, + "CGGGGG": 3076, + "GAAAAA": 3077, + "GAAAAT": 3078, + "GAAAAC": 3079, + "GAAAAG": 3080, + "GAAATA": 3081, + "GAAATT": 3082, + "GAAATC": 3083, + "GAAATG": 3084, + "GAAACA": 3085, + "GAAACT": 3086, + "GAAACC": 3087, + "GAAACG": 3088, + "GAAAGA": 3089, + "GAAAGT": 3090, + "GAAAGC": 3091, + "GAAAGG": 3092, + "GAATAA": 3093, + "GAATAT": 3094, + "GAATAC": 3095, + "GAATAG": 3096, + "GAATTA": 3097, + "GAATTT": 3098, + "GAATTC": 3099, + "GAATTG": 3100, + "GAATCA": 3101, + "GAATCT": 3102, + "GAATCC": 3103, + "GAATCG": 3104, + "GAATGA": 3105, + "GAATGT": 3106, + "GAATGC": 3107, + "GAATGG": 3108, + "GAACAA": 3109, + "GAACAT": 3110, + "GAACAC": 3111, + "GAACAG": 3112, + "GAACTA": 3113, + "GAACTT": 3114, + "GAACTC": 3115, + "GAACTG": 3116, + "GAACCA": 3117, + "GAACCT": 3118, + "GAACCC": 3119, + "GAACCG": 3120, + "GAACGA": 3121, + "GAACGT": 3122, + "GAACGC": 3123, + "GAACGG": 3124, + "GAAGAA": 3125, + "GAAGAT": 3126, + "GAAGAC": 3127, + "GAAGAG": 3128, + "GAAGTA": 3129, + "GAAGTT": 3130, + "GAAGTC": 3131, + "GAAGTG": 3132, + "GAAGCA": 3133, + "GAAGCT": 3134, + "GAAGCC": 3135, + "GAAGCG": 3136, + "GAAGGA": 3137, + "GAAGGT": 3138, + "GAAGGC": 3139, + "GAAGGG": 3140, + "GATAAA": 3141, + "GATAAT": 3142, + "GATAAC": 3143, + "GATAAG": 3144, + "GATATA": 3145, + "GATATT": 3146, + "GATATC": 3147, + "GATATG": 3148, + "GATACA": 3149, + "GATACT": 3150, + "GATACC": 3151, + "GATACG": 3152, + "GATAGA": 3153, + "GATAGT": 3154, + "GATAGC": 3155, + "GATAGG": 3156, + "GATTAA": 3157, + "GATTAT": 3158, + "GATTAC": 3159, + "GATTAG": 3160, + "GATTTA": 3161, + "GATTTT": 3162, + "GATTTC": 3163, + "GATTTG": 3164, + "GATTCA": 3165, + "GATTCT": 3166, + "GATTCC": 3167, + "GATTCG": 3168, + "GATTGA": 3169, + "GATTGT": 3170, + "GATTGC": 3171, + "GATTGG": 3172, + "GATCAA": 3173, + "GATCAT": 3174, + "GATCAC": 3175, + "GATCAG": 3176, + "GATCTA": 3177, + "GATCTT": 3178, + "GATCTC": 3179, + "GATCTG": 3180, + "GATCCA": 3181, + "GATCCT": 3182, + "GATCCC": 3183, + "GATCCG": 3184, + "GATCGA": 3185, + "GATCGT": 3186, + "GATCGC": 3187, + "GATCGG": 3188, + "GATGAA": 3189, + "GATGAT": 3190, + "GATGAC": 3191, + "GATGAG": 3192, + "GATGTA": 3193, + "GATGTT": 3194, + "GATGTC": 3195, + "GATGTG": 3196, + "GATGCA": 3197, + "GATGCT": 3198, + "GATGCC": 3199, + "GATGCG": 3200, + "GATGGA": 3201, + "GATGGT": 3202, + "GATGGC": 3203, + "GATGGG": 3204, + "GACAAA": 3205, + "GACAAT": 3206, + "GACAAC": 3207, + "GACAAG": 3208, + "GACATA": 3209, + "GACATT": 3210, + "GACATC": 3211, + "GACATG": 3212, + "GACACA": 3213, + "GACACT": 3214, + "GACACC": 3215, + "GACACG": 3216, + "GACAGA": 3217, + "GACAGT": 3218, + "GACAGC": 3219, + "GACAGG": 3220, + "GACTAA": 3221, + "GACTAT": 3222, + "GACTAC": 3223, + "GACTAG": 3224, + "GACTTA": 3225, + "GACTTT": 3226, + "GACTTC": 3227, + "GACTTG": 3228, + "GACTCA": 3229, + "GACTCT": 3230, + "GACTCC": 3231, + "GACTCG": 3232, + "GACTGA": 3233, + "GACTGT": 3234, + "GACTGC": 3235, + "GACTGG": 3236, + "GACCAA": 3237, + "GACCAT": 3238, + "GACCAC": 3239, + "GACCAG": 3240, + "GACCTA": 3241, + "GACCTT": 3242, + "GACCTC": 3243, + "GACCTG": 3244, + "GACCCA": 3245, + "GACCCT": 3246, + "GACCCC": 3247, + "GACCCG": 3248, + "GACCGA": 3249, + "GACCGT": 3250, + "GACCGC": 3251, + "GACCGG": 3252, + "GACGAA": 3253, + "GACGAT": 3254, + "GACGAC": 3255, + "GACGAG": 3256, + "GACGTA": 3257, + "GACGTT": 3258, + "GACGTC": 3259, + "GACGTG": 3260, + "GACGCA": 3261, + "GACGCT": 3262, + "GACGCC": 3263, + "GACGCG": 3264, + "GACGGA": 3265, + "GACGGT": 3266, + "GACGGC": 3267, + "GACGGG": 3268, + "GAGAAA": 3269, + "GAGAAT": 3270, + "GAGAAC": 3271, + "GAGAAG": 3272, + "GAGATA": 3273, + "GAGATT": 3274, + "GAGATC": 3275, + "GAGATG": 3276, + "GAGACA": 3277, + "GAGACT": 3278, + "GAGACC": 3279, + "GAGACG": 3280, + "GAGAGA": 3281, + "GAGAGT": 3282, + "GAGAGC": 3283, + "GAGAGG": 3284, + "GAGTAA": 3285, + "GAGTAT": 3286, + "GAGTAC": 3287, + "GAGTAG": 3288, + "GAGTTA": 3289, + "GAGTTT": 3290, + "GAGTTC": 3291, + "GAGTTG": 3292, + "GAGTCA": 3293, + "GAGTCT": 3294, + "GAGTCC": 3295, + "GAGTCG": 3296, + "GAGTGA": 3297, + "GAGTGT": 3298, + "GAGTGC": 3299, + "GAGTGG": 3300, + "GAGCAA": 3301, + "GAGCAT": 3302, + "GAGCAC": 3303, + "GAGCAG": 3304, + "GAGCTA": 3305, + "GAGCTT": 3306, + "GAGCTC": 3307, + "GAGCTG": 3308, + "GAGCCA": 3309, + "GAGCCT": 3310, + "GAGCCC": 3311, + "GAGCCG": 3312, + "GAGCGA": 3313, + "GAGCGT": 3314, + "GAGCGC": 3315, + "GAGCGG": 3316, + "GAGGAA": 3317, + "GAGGAT": 3318, + "GAGGAC": 3319, + "GAGGAG": 3320, + "GAGGTA": 3321, + "GAGGTT": 3322, + "GAGGTC": 3323, + "GAGGTG": 3324, + "GAGGCA": 3325, + "GAGGCT": 3326, + "GAGGCC": 3327, + "GAGGCG": 3328, + "GAGGGA": 3329, + "GAGGGT": 3330, + "GAGGGC": 3331, + "GAGGGG": 3332, + "GTAAAA": 3333, + "GTAAAT": 3334, + "GTAAAC": 3335, + "GTAAAG": 3336, + "GTAATA": 3337, + "GTAATT": 3338, + "GTAATC": 3339, + "GTAATG": 3340, + "GTAACA": 3341, + "GTAACT": 3342, + "GTAACC": 3343, + "GTAACG": 3344, + "GTAAGA": 3345, + "GTAAGT": 3346, + "GTAAGC": 3347, + "GTAAGG": 3348, + "GTATAA": 3349, + "GTATAT": 3350, + "GTATAC": 3351, + "GTATAG": 3352, + "GTATTA": 3353, + "GTATTT": 3354, + "GTATTC": 3355, + "GTATTG": 3356, + "GTATCA": 3357, + "GTATCT": 3358, + "GTATCC": 3359, + "GTATCG": 3360, + "GTATGA": 3361, + "GTATGT": 3362, + "GTATGC": 3363, + "GTATGG": 3364, + "GTACAA": 3365, + "GTACAT": 3366, + "GTACAC": 3367, + "GTACAG": 3368, + "GTACTA": 3369, + "GTACTT": 3370, + "GTACTC": 3371, + "GTACTG": 3372, + "GTACCA": 3373, + "GTACCT": 3374, + "GTACCC": 3375, + "GTACCG": 3376, + "GTACGA": 3377, + "GTACGT": 3378, + "GTACGC": 3379, + "GTACGG": 3380, + "GTAGAA": 3381, + "GTAGAT": 3382, + "GTAGAC": 3383, + "GTAGAG": 3384, + "GTAGTA": 3385, + "GTAGTT": 3386, + "GTAGTC": 3387, + "GTAGTG": 3388, + "GTAGCA": 3389, + "GTAGCT": 3390, + "GTAGCC": 3391, + "GTAGCG": 3392, + "GTAGGA": 3393, + "GTAGGT": 3394, + "GTAGGC": 3395, + "GTAGGG": 3396, + "GTTAAA": 3397, + "GTTAAT": 3398, + "GTTAAC": 3399, + "GTTAAG": 3400, + "GTTATA": 3401, + "GTTATT": 3402, + "GTTATC": 3403, + "GTTATG": 3404, + "GTTACA": 3405, + "GTTACT": 3406, + "GTTACC": 3407, + "GTTACG": 3408, + "GTTAGA": 3409, + "GTTAGT": 3410, + "GTTAGC": 3411, + "GTTAGG": 3412, + "GTTTAA": 3413, + "GTTTAT": 3414, + "GTTTAC": 3415, + "GTTTAG": 3416, + "GTTTTA": 3417, + "GTTTTT": 3418, + "GTTTTC": 3419, + "GTTTTG": 3420, + "GTTTCA": 3421, + "GTTTCT": 3422, + "GTTTCC": 3423, + "GTTTCG": 3424, + "GTTTGA": 3425, + "GTTTGT": 3426, + "GTTTGC": 3427, + "GTTTGG": 3428, + "GTTCAA": 3429, + "GTTCAT": 3430, + "GTTCAC": 3431, + "GTTCAG": 3432, + "GTTCTA": 3433, + "GTTCTT": 3434, + "GTTCTC": 3435, + "GTTCTG": 3436, + "GTTCCA": 3437, + "GTTCCT": 3438, + "GTTCCC": 3439, + "GTTCCG": 3440, + "GTTCGA": 3441, + "GTTCGT": 3442, + "GTTCGC": 3443, + "GTTCGG": 3444, + "GTTGAA": 3445, + "GTTGAT": 3446, + "GTTGAC": 3447, + "GTTGAG": 3448, + "GTTGTA": 3449, + "GTTGTT": 3450, + "GTTGTC": 3451, + "GTTGTG": 3452, + "GTTGCA": 3453, + "GTTGCT": 3454, + "GTTGCC": 3455, + "GTTGCG": 3456, + "GTTGGA": 3457, + "GTTGGT": 3458, + "GTTGGC": 3459, + "GTTGGG": 3460, + "GTCAAA": 3461, + "GTCAAT": 3462, + "GTCAAC": 3463, + "GTCAAG": 3464, + "GTCATA": 3465, + "GTCATT": 3466, + "GTCATC": 3467, + "GTCATG": 3468, + "GTCACA": 3469, + "GTCACT": 3470, + "GTCACC": 3471, + "GTCACG": 3472, + "GTCAGA": 3473, + "GTCAGT": 3474, + "GTCAGC": 3475, + "GTCAGG": 3476, + "GTCTAA": 3477, + "GTCTAT": 3478, + "GTCTAC": 3479, + "GTCTAG": 3480, + "GTCTTA": 3481, + "GTCTTT": 3482, + "GTCTTC": 3483, + "GTCTTG": 3484, + "GTCTCA": 3485, + "GTCTCT": 3486, + "GTCTCC": 3487, + "GTCTCG": 3488, + "GTCTGA": 3489, + "GTCTGT": 3490, + "GTCTGC": 3491, + "GTCTGG": 3492, + "GTCCAA": 3493, + "GTCCAT": 3494, + "GTCCAC": 3495, + "GTCCAG": 3496, + "GTCCTA": 3497, + "GTCCTT": 3498, + "GTCCTC": 3499, + "GTCCTG": 3500, + "GTCCCA": 3501, + "GTCCCT": 3502, + "GTCCCC": 3503, + "GTCCCG": 3504, + "GTCCGA": 3505, + "GTCCGT": 3506, + "GTCCGC": 3507, + "GTCCGG": 3508, + "GTCGAA": 3509, + "GTCGAT": 3510, + "GTCGAC": 3511, + "GTCGAG": 3512, + "GTCGTA": 3513, + "GTCGTT": 3514, + "GTCGTC": 3515, + "GTCGTG": 3516, + "GTCGCA": 3517, + "GTCGCT": 3518, + "GTCGCC": 3519, + "GTCGCG": 3520, + "GTCGGA": 3521, + "GTCGGT": 3522, + "GTCGGC": 3523, + "GTCGGG": 3524, + "GTGAAA": 3525, + "GTGAAT": 3526, + "GTGAAC": 3527, + "GTGAAG": 3528, + "GTGATA": 3529, + "GTGATT": 3530, + "GTGATC": 3531, + "GTGATG": 3532, + "GTGACA": 3533, + "GTGACT": 3534, + "GTGACC": 3535, + "GTGACG": 3536, + "GTGAGA": 3537, + "GTGAGT": 3538, + "GTGAGC": 3539, + "GTGAGG": 3540, + "GTGTAA": 3541, + "GTGTAT": 3542, + "GTGTAC": 3543, + "GTGTAG": 3544, + "GTGTTA": 3545, + "GTGTTT": 3546, + "GTGTTC": 3547, + "GTGTTG": 3548, + "GTGTCA": 3549, + "GTGTCT": 3550, + "GTGTCC": 3551, + "GTGTCG": 3552, + "GTGTGA": 3553, + "GTGTGT": 3554, + "GTGTGC": 3555, + "GTGTGG": 3556, + "GTGCAA": 3557, + "GTGCAT": 3558, + "GTGCAC": 3559, + "GTGCAG": 3560, + "GTGCTA": 3561, + "GTGCTT": 3562, + "GTGCTC": 3563, + "GTGCTG": 3564, + "GTGCCA": 3565, + "GTGCCT": 3566, + "GTGCCC": 3567, + "GTGCCG": 3568, + "GTGCGA": 3569, + "GTGCGT": 3570, + "GTGCGC": 3571, + "GTGCGG": 3572, + "GTGGAA": 3573, + "GTGGAT": 3574, + "GTGGAC": 3575, + "GTGGAG": 3576, + "GTGGTA": 3577, + "GTGGTT": 3578, + "GTGGTC": 3579, + "GTGGTG": 3580, + "GTGGCA": 3581, + "GTGGCT": 3582, + "GTGGCC": 3583, + "GTGGCG": 3584, + "GTGGGA": 3585, + "GTGGGT": 3586, + "GTGGGC": 3587, + "GTGGGG": 3588, + "GCAAAA": 3589, + "GCAAAT": 3590, + "GCAAAC": 3591, + "GCAAAG": 3592, + "GCAATA": 3593, + "GCAATT": 3594, + "GCAATC": 3595, + "GCAATG": 3596, + "GCAACA": 3597, + "GCAACT": 3598, + "GCAACC": 3599, + "GCAACG": 3600, + "GCAAGA": 3601, + "GCAAGT": 3602, + "GCAAGC": 3603, + "GCAAGG": 3604, + "GCATAA": 3605, + "GCATAT": 3606, + "GCATAC": 3607, + "GCATAG": 3608, + "GCATTA": 3609, + "GCATTT": 3610, + "GCATTC": 3611, + "GCATTG": 3612, + "GCATCA": 3613, + "GCATCT": 3614, + "GCATCC": 3615, + "GCATCG": 3616, + "GCATGA": 3617, + "GCATGT": 3618, + "GCATGC": 3619, + "GCATGG": 3620, + "GCACAA": 3621, + "GCACAT": 3622, + "GCACAC": 3623, + "GCACAG": 3624, + "GCACTA": 3625, + "GCACTT": 3626, + "GCACTC": 3627, + "GCACTG": 3628, + "GCACCA": 3629, + "GCACCT": 3630, + "GCACCC": 3631, + "GCACCG": 3632, + "GCACGA": 3633, + "GCACGT": 3634, + "GCACGC": 3635, + "GCACGG": 3636, + "GCAGAA": 3637, + "GCAGAT": 3638, + "GCAGAC": 3639, + "GCAGAG": 3640, + "GCAGTA": 3641, + "GCAGTT": 3642, + "GCAGTC": 3643, + "GCAGTG": 3644, + "GCAGCA": 3645, + "GCAGCT": 3646, + "GCAGCC": 3647, + "GCAGCG": 3648, + "GCAGGA": 3649, + "GCAGGT": 3650, + "GCAGGC": 3651, + "GCAGGG": 3652, + "GCTAAA": 3653, + "GCTAAT": 3654, + "GCTAAC": 3655, + "GCTAAG": 3656, + "GCTATA": 3657, + "GCTATT": 3658, + "GCTATC": 3659, + "GCTATG": 3660, + "GCTACA": 3661, + "GCTACT": 3662, + "GCTACC": 3663, + "GCTACG": 3664, + "GCTAGA": 3665, + "GCTAGT": 3666, + "GCTAGC": 3667, + "GCTAGG": 3668, + "GCTTAA": 3669, + "GCTTAT": 3670, + "GCTTAC": 3671, + "GCTTAG": 3672, + "GCTTTA": 3673, + "GCTTTT": 3674, + "GCTTTC": 3675, + "GCTTTG": 3676, + "GCTTCA": 3677, + "GCTTCT": 3678, + "GCTTCC": 3679, + "GCTTCG": 3680, + "GCTTGA": 3681, + "GCTTGT": 3682, + "GCTTGC": 3683, + "GCTTGG": 3684, + "GCTCAA": 3685, + "GCTCAT": 3686, + "GCTCAC": 3687, + "GCTCAG": 3688, + "GCTCTA": 3689, + "GCTCTT": 3690, + "GCTCTC": 3691, + "GCTCTG": 3692, + "GCTCCA": 3693, + "GCTCCT": 3694, + "GCTCCC": 3695, + "GCTCCG": 3696, + "GCTCGA": 3697, + "GCTCGT": 3698, + "GCTCGC": 3699, + "GCTCGG": 3700, + "GCTGAA": 3701, + "GCTGAT": 3702, + "GCTGAC": 3703, + "GCTGAG": 3704, + "GCTGTA": 3705, + "GCTGTT": 3706, + "GCTGTC": 3707, + "GCTGTG": 3708, + "GCTGCA": 3709, + "GCTGCT": 3710, + "GCTGCC": 3711, + "GCTGCG": 3712, + "GCTGGA": 3713, + "GCTGGT": 3714, + "GCTGGC": 3715, + "GCTGGG": 3716, + "GCCAAA": 3717, + "GCCAAT": 3718, + "GCCAAC": 3719, + "GCCAAG": 3720, + "GCCATA": 3721, + "GCCATT": 3722, + "GCCATC": 3723, + "GCCATG": 3724, + "GCCACA": 3725, + "GCCACT": 3726, + "GCCACC": 3727, + "GCCACG": 3728, + "GCCAGA": 3729, + "GCCAGT": 3730, + "GCCAGC": 3731, + "GCCAGG": 3732, + "GCCTAA": 3733, + "GCCTAT": 3734, + "GCCTAC": 3735, + "GCCTAG": 3736, + "GCCTTA": 3737, + "GCCTTT": 3738, + "GCCTTC": 3739, + "GCCTTG": 3740, + "GCCTCA": 3741, + "GCCTCT": 3742, + "GCCTCC": 3743, + "GCCTCG": 3744, + "GCCTGA": 3745, + "GCCTGT": 3746, + "GCCTGC": 3747, + "GCCTGG": 3748, + "GCCCAA": 3749, + "GCCCAT": 3750, + "GCCCAC": 3751, + "GCCCAG": 3752, + "GCCCTA": 3753, + "GCCCTT": 3754, + "GCCCTC": 3755, + "GCCCTG": 3756, + "GCCCCA": 3757, + "GCCCCT": 3758, + "GCCCCC": 3759, + "GCCCCG": 3760, + "GCCCGA": 3761, + "GCCCGT": 3762, + "GCCCGC": 3763, + "GCCCGG": 3764, + "GCCGAA": 3765, + "GCCGAT": 3766, + "GCCGAC": 3767, + "GCCGAG": 3768, + "GCCGTA": 3769, + "GCCGTT": 3770, + "GCCGTC": 3771, + "GCCGTG": 3772, + "GCCGCA": 3773, + "GCCGCT": 3774, + "GCCGCC": 3775, + "GCCGCG": 3776, + "GCCGGA": 3777, + "GCCGGT": 3778, + "GCCGGC": 3779, + "GCCGGG": 3780, + "GCGAAA": 3781, + "GCGAAT": 3782, + "GCGAAC": 3783, + "GCGAAG": 3784, + "GCGATA": 3785, + "GCGATT": 3786, + "GCGATC": 3787, + "GCGATG": 3788, + "GCGACA": 3789, + "GCGACT": 3790, + "GCGACC": 3791, + "GCGACG": 3792, + "GCGAGA": 3793, + "GCGAGT": 3794, + "GCGAGC": 3795, + "GCGAGG": 3796, + "GCGTAA": 3797, + "GCGTAT": 3798, + "GCGTAC": 3799, + "GCGTAG": 3800, + "GCGTTA": 3801, + "GCGTTT": 3802, + "GCGTTC": 3803, + "GCGTTG": 3804, + "GCGTCA": 3805, + "GCGTCT": 3806, + "GCGTCC": 3807, + "GCGTCG": 3808, + "GCGTGA": 3809, + "GCGTGT": 3810, + "GCGTGC": 3811, + "GCGTGG": 3812, + "GCGCAA": 3813, + "GCGCAT": 3814, + "GCGCAC": 3815, + "GCGCAG": 3816, + "GCGCTA": 3817, + "GCGCTT": 3818, + "GCGCTC": 3819, + "GCGCTG": 3820, + "GCGCCA": 3821, + "GCGCCT": 3822, + "GCGCCC": 3823, + "GCGCCG": 3824, + "GCGCGA": 3825, + "GCGCGT": 3826, + "GCGCGC": 3827, + "GCGCGG": 3828, + "GCGGAA": 3829, + "GCGGAT": 3830, + "GCGGAC": 3831, + "GCGGAG": 3832, + "GCGGTA": 3833, + "GCGGTT": 3834, + "GCGGTC": 3835, + "GCGGTG": 3836, + "GCGGCA": 3837, + "GCGGCT": 3838, + "GCGGCC": 3839, + "GCGGCG": 3840, + "GCGGGA": 3841, + "GCGGGT": 3842, + "GCGGGC": 3843, + "GCGGGG": 3844, + "GGAAAA": 3845, + "GGAAAT": 3846, + "GGAAAC": 3847, + "GGAAAG": 3848, + "GGAATA": 3849, + "GGAATT": 3850, + "GGAATC": 3851, + "GGAATG": 3852, + "GGAACA": 3853, + "GGAACT": 3854, + "GGAACC": 3855, + "GGAACG": 3856, + "GGAAGA": 3857, + "GGAAGT": 3858, + "GGAAGC": 3859, + "GGAAGG": 3860, + "GGATAA": 3861, + "GGATAT": 3862, + "GGATAC": 3863, + "GGATAG": 3864, + "GGATTA": 3865, + "GGATTT": 3866, + "GGATTC": 3867, + "GGATTG": 3868, + "GGATCA": 3869, + "GGATCT": 3870, + "GGATCC": 3871, + "GGATCG": 3872, + "GGATGA": 3873, + "GGATGT": 3874, + "GGATGC": 3875, + "GGATGG": 3876, + "GGACAA": 3877, + "GGACAT": 3878, + "GGACAC": 3879, + "GGACAG": 3880, + "GGACTA": 3881, + "GGACTT": 3882, + "GGACTC": 3883, + "GGACTG": 3884, + "GGACCA": 3885, + "GGACCT": 3886, + "GGACCC": 3887, + "GGACCG": 3888, + "GGACGA": 3889, + "GGACGT": 3890, + "GGACGC": 3891, + "GGACGG": 3892, + "GGAGAA": 3893, + "GGAGAT": 3894, + "GGAGAC": 3895, + "GGAGAG": 3896, + "GGAGTA": 3897, + "GGAGTT": 3898, + "GGAGTC": 3899, + "GGAGTG": 3900, + "GGAGCA": 3901, + "GGAGCT": 3902, + "GGAGCC": 3903, + "GGAGCG": 3904, + "GGAGGA": 3905, + "GGAGGT": 3906, + "GGAGGC": 3907, + "GGAGGG": 3908, + "GGTAAA": 3909, + "GGTAAT": 3910, + "GGTAAC": 3911, + "GGTAAG": 3912, + "GGTATA": 3913, + "GGTATT": 3914, + "GGTATC": 3915, + "GGTATG": 3916, + "GGTACA": 3917, + "GGTACT": 3918, + "GGTACC": 3919, + "GGTACG": 3920, + "GGTAGA": 3921, + "GGTAGT": 3922, + "GGTAGC": 3923, + "GGTAGG": 3924, + "GGTTAA": 3925, + "GGTTAT": 3926, + "GGTTAC": 3927, + "GGTTAG": 3928, + "GGTTTA": 3929, + "GGTTTT": 3930, + "GGTTTC": 3931, + "GGTTTG": 3932, + "GGTTCA": 3933, + "GGTTCT": 3934, + "GGTTCC": 3935, + "GGTTCG": 3936, + "GGTTGA": 3937, + "GGTTGT": 3938, + "GGTTGC": 3939, + "GGTTGG": 3940, + "GGTCAA": 3941, + "GGTCAT": 3942, + "GGTCAC": 3943, + "GGTCAG": 3944, + "GGTCTA": 3945, + "GGTCTT": 3946, + "GGTCTC": 3947, + "GGTCTG": 3948, + "GGTCCA": 3949, + "GGTCCT": 3950, + "GGTCCC": 3951, + "GGTCCG": 3952, + "GGTCGA": 3953, + "GGTCGT": 3954, + "GGTCGC": 3955, + "GGTCGG": 3956, + "GGTGAA": 3957, + "GGTGAT": 3958, + "GGTGAC": 3959, + "GGTGAG": 3960, + "GGTGTA": 3961, + "GGTGTT": 3962, + "GGTGTC": 3963, + "GGTGTG": 3964, + "GGTGCA": 3965, + "GGTGCT": 3966, + "GGTGCC": 3967, + "GGTGCG": 3968, + "GGTGGA": 3969, + "GGTGGT": 3970, + "GGTGGC": 3971, + "GGTGGG": 3972, + "GGCAAA": 3973, + "GGCAAT": 3974, + "GGCAAC": 3975, + "GGCAAG": 3976, + "GGCATA": 3977, + "GGCATT": 3978, + "GGCATC": 3979, + "GGCATG": 3980, + "GGCACA": 3981, + "GGCACT": 3982, + "GGCACC": 3983, + "GGCACG": 3984, + "GGCAGA": 3985, + "GGCAGT": 3986, + "GGCAGC": 3987, + "GGCAGG": 3988, + "GGCTAA": 3989, + "GGCTAT": 3990, + "GGCTAC": 3991, + "GGCTAG": 3992, + "GGCTTA": 3993, + "GGCTTT": 3994, + "GGCTTC": 3995, + "GGCTTG": 3996, + "GGCTCA": 3997, + "GGCTCT": 3998, + "GGCTCC": 3999, + "GGCTCG": 4000, + "GGCTGA": 4001, + "GGCTGT": 4002, + "GGCTGC": 4003, + "GGCTGG": 4004, + "GGCCAA": 4005, + "GGCCAT": 4006, + "GGCCAC": 4007, + "GGCCAG": 4008, + "GGCCTA": 4009, + "GGCCTT": 4010, + "GGCCTC": 4011, + "GGCCTG": 4012, + "GGCCCA": 4013, + "GGCCCT": 4014, + "GGCCCC": 4015, + "GGCCCG": 4016, + "GGCCGA": 4017, + "GGCCGT": 4018, + "GGCCGC": 4019, + "GGCCGG": 4020, + "GGCGAA": 4021, + "GGCGAT": 4022, + "GGCGAC": 4023, + "GGCGAG": 4024, + "GGCGTA": 4025, + "GGCGTT": 4026, + "GGCGTC": 4027, + "GGCGTG": 4028, + "GGCGCA": 4029, + "GGCGCT": 4030, + "GGCGCC": 4031, + "GGCGCG": 4032, + "GGCGGA": 4033, + "GGCGGT": 4034, + "GGCGGC": 4035, + "GGCGGG": 4036, + "GGGAAA": 4037, + "GGGAAT": 4038, + "GGGAAC": 4039, + "GGGAAG": 4040, + "GGGATA": 4041, + "GGGATT": 4042, + "GGGATC": 4043, + "GGGATG": 4044, + "GGGACA": 4045, + "GGGACT": 4046, + "GGGACC": 4047, + "GGGACG": 4048, + "GGGAGA": 4049, + "GGGAGT": 4050, + "GGGAGC": 4051, + "GGGAGG": 4052, + "GGGTAA": 4053, + "GGGTAT": 4054, + "GGGTAC": 4055, + "GGGTAG": 4056, + "GGGTTA": 4057, + "GGGTTT": 4058, + "GGGTTC": 4059, + "GGGTTG": 4060, + "GGGTCA": 4061, + "GGGTCT": 4062, + "GGGTCC": 4063, + "GGGTCG": 4064, + "GGGTGA": 4065, + "GGGTGT": 4066, + "GGGTGC": 4067, + "GGGTGG": 4068, + "GGGCAA": 4069, + "GGGCAT": 4070, + "GGGCAC": 4071, + "GGGCAG": 4072, + "GGGCTA": 4073, + "GGGCTT": 4074, + "GGGCTC": 4075, + "GGGCTG": 4076, + "GGGCCA": 4077, + "GGGCCT": 4078, + "GGGCCC": 4079, + "GGGCCG": 4080, + "GGGCGA": 4081, + "GGGCGT": 4082, + "GGGCGC": 4083, + "GGGCGG": 4084, + "GGGGAA": 4085, + "GGGGAT": 4086, + "GGGGAC": 4087, + "GGGGAG": 4088, + "GGGGTA": 4089, + "GGGGTT": 4090, + "GGGGTC": 4091, + "GGGGTG": 4092, + "GGGGCA": 4093, + "GGGGCT": 4094, + "GGGGCC": 4095, + "GGGGCG": 4096, + "GGGGGA": 4097, + "GGGGGT": 4098, + "GGGGGC": 4099, + "GGGGGG": 4100 + } + } +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-5000/tokenizer_config.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-5000/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..604e2b1b29af1dcb3920d2828b6a90276e5e3d02 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-5000/tokenizer_config.json @@ -0,0 +1,58 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "[UNK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "[CLS]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "[SEP]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "4": { + "content": "[MASK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "clean_up_tokenization_spaces": true, + "cls_token": "[CLS]", + "do_basic_tokenize": true, + "do_lower_case": false, + "mask_token": "[MASK]", + "max_len": 512, + "model_max_length": 512, + "never_split": null, + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "strip_accents": null, + "tokenize_chinese_chars": true, + "tokenizer_class": "BertTokenizer", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-5000/trainer_state.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-5000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..7029af5d07bb7c70ecf9b43a03dbd9c5a5090027 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-5000/trainer_state.json @@ -0,0 +1,1433 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 20.0, + "eval_steps": 500, + "global_step": 5000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.1, + "grad_norm": 0.6611918807029724, + "learning_rate": 5.319148936170213e-05, + "loss": 0.6932, + "step": 25 + }, + { + "epoch": 0.2, + "grad_norm": 0.9467485547065735, + "learning_rate": 0.00010638297872340425, + "loss": 0.6912, + "step": 50 + }, + { + "epoch": 0.3, + "grad_norm": 0.9465051889419556, + "learning_rate": 0.00015957446808510637, + "loss": 0.6893, + "step": 75 + }, + { + "epoch": 0.4, + "grad_norm": 1.5430934429168701, + "learning_rate": 0.0002127659574468085, + "loss": 0.6804, + "step": 100 + }, + { + "epoch": 0.5, + "grad_norm": 1.6103107929229736, + "learning_rate": 0.00026595744680851064, + "loss": 0.6806, + "step": 125 + }, + { + "epoch": 0.6, + "grad_norm": 2.3773200511932373, + "learning_rate": 0.00031914893617021275, + "loss": 0.6801, + "step": 150 + }, + { + "epoch": 0.7, + "grad_norm": 1.8832203149795532, + "learning_rate": 0.0003723404255319149, + "loss": 0.6791, + "step": 175 + }, + { + "epoch": 0.8, + "grad_norm": 1.3350876569747925, + "learning_rate": 0.0003992081821181128, + "loss": 0.6746, + "step": 200 + }, + { + "epoch": 0.9, + "grad_norm": 1.5796219110488892, + "learning_rate": 0.0003975585615308479, + "loss": 0.6771, + "step": 225 + }, + { + "epoch": 1.0, + "grad_norm": 1.884006381034851, + "learning_rate": 0.00039590894094358297, + "loss": 0.6649, + "step": 250 + }, + { + "epoch": 1.1, + "grad_norm": 3.286440372467041, + "learning_rate": 0.0003942593203563181, + "loss": 0.6388, + "step": 275 + }, + { + "epoch": 1.2, + "grad_norm": 8.244946479797363, + "learning_rate": 0.0003926096997690532, + "loss": 0.6238, + "step": 300 + }, + { + "epoch": 1.3, + "grad_norm": 4.265683650970459, + "learning_rate": 0.00039096007918178817, + "loss": 0.6383, + "step": 325 + }, + { + "epoch": 1.4, + "grad_norm": 4.115826606750488, + "learning_rate": 0.0003893104585945233, + "loss": 0.6194, + "step": 350 + }, + { + "epoch": 1.5, + "grad_norm": 5.694250583648682, + "learning_rate": 0.0003876608380072583, + "loss": 0.6324, + "step": 375 + }, + { + "epoch": 1.6, + "grad_norm": 3.463121175765991, + "learning_rate": 0.00038601121741999343, + "loss": 0.621, + "step": 400 + }, + { + "epoch": 1.7, + "grad_norm": 4.582865238189697, + "learning_rate": 0.0003843615968327285, + "loss": 0.6116, + "step": 425 + }, + { + "epoch": 1.8, + "grad_norm": 11.996281623840332, + "learning_rate": 0.0003827119762454636, + "loss": 0.6393, + "step": 450 + }, + { + "epoch": 1.9, + "grad_norm": 3.0407373905181885, + "learning_rate": 0.00038106235565819863, + "loss": 0.628, + "step": 475 + }, + { + "epoch": 2.0, + "grad_norm": 2.917588233947754, + "learning_rate": 0.0003794127350709337, + "loss": 0.6078, + "step": 500 + }, + { + "epoch": 2.1, + "grad_norm": 4.748379707336426, + "learning_rate": 0.0003777631144836688, + "loss": 0.4899, + "step": 525 + }, + { + "epoch": 2.2, + "grad_norm": 3.8076977729797363, + "learning_rate": 0.00037611349389640383, + "loss": 0.5086, + "step": 550 + }, + { + "epoch": 2.3, + "grad_norm": 5.2440714836120605, + "learning_rate": 0.00037446387330913894, + "loss": 0.5327, + "step": 575 + }, + { + "epoch": 2.4, + "grad_norm": 7.110438346862793, + "learning_rate": 0.000372814252721874, + "loss": 0.5436, + "step": 600 + }, + { + "epoch": 2.5, + "grad_norm": 5.46150541305542, + "learning_rate": 0.00037116463213460903, + "loss": 0.5294, + "step": 625 + }, + { + "epoch": 2.6, + "grad_norm": 5.136163234710693, + "learning_rate": 0.00036951501154734414, + "loss": 0.5245, + "step": 650 + }, + { + "epoch": 2.7, + "grad_norm": 8.735346794128418, + "learning_rate": 0.0003678653909600792, + "loss": 0.5449, + "step": 675 + }, + { + "epoch": 2.8, + "grad_norm": 2.922825574874878, + "learning_rate": 0.0003662157703728143, + "loss": 0.5406, + "step": 700 + }, + { + "epoch": 2.9, + "grad_norm": 7.744819641113281, + "learning_rate": 0.00036456614978554934, + "loss": 0.5447, + "step": 725 + }, + { + "epoch": 3.0, + "grad_norm": 11.6185884475708, + "learning_rate": 0.00036291652919828444, + "loss": 0.5195, + "step": 750 + }, + { + "epoch": 3.1, + "grad_norm": 3.73836088180542, + "learning_rate": 0.00036126690861101944, + "loss": 0.3824, + "step": 775 + }, + { + "epoch": 3.2, + "grad_norm": 14.850343704223633, + "learning_rate": 0.00035961728802375454, + "loss": 0.4071, + "step": 800 + }, + { + "epoch": 3.3, + "grad_norm": 5.7157440185546875, + "learning_rate": 0.0003579676674364896, + "loss": 0.3986, + "step": 825 + }, + { + "epoch": 3.4, + "grad_norm": 12.418399810791016, + "learning_rate": 0.0003563180468492247, + "loss": 0.4282, + "step": 850 + }, + { + "epoch": 3.5, + "grad_norm": 12.793001174926758, + "learning_rate": 0.0003546684262619598, + "loss": 0.4822, + "step": 875 + }, + { + "epoch": 3.6, + "grad_norm": 6.489450931549072, + "learning_rate": 0.00035301880567469485, + "loss": 0.4239, + "step": 900 + }, + { + "epoch": 3.7, + "grad_norm": 5.365822792053223, + "learning_rate": 0.0003513691850874299, + "loss": 0.421, + "step": 925 + }, + { + "epoch": 3.8, + "grad_norm": 12.643745422363281, + "learning_rate": 0.00034971956450016495, + "loss": 0.3964, + "step": 950 + }, + { + "epoch": 3.9, + "grad_norm": 14.334024429321289, + "learning_rate": 0.00034806994391290005, + "loss": 0.4634, + "step": 975 + }, + { + "epoch": 4.0, + "grad_norm": 6.819091320037842, + "learning_rate": 0.0003464203233256351, + "loss": 0.4139, + "step": 1000 + }, + { + "epoch": 4.1, + "grad_norm": 8.29238224029541, + "learning_rate": 0.0003447707027383702, + "loss": 0.2695, + "step": 1025 + }, + { + "epoch": 4.2, + "grad_norm": 5.984206676483154, + "learning_rate": 0.00034312108215110525, + "loss": 0.2653, + "step": 1050 + }, + { + "epoch": 4.3, + "grad_norm": 5.9425435066223145, + "learning_rate": 0.0003414714615638403, + "loss": 0.2982, + "step": 1075 + }, + { + "epoch": 4.4, + "grad_norm": 7.1877593994140625, + "learning_rate": 0.0003398218409765754, + "loss": 0.3309, + "step": 1100 + }, + { + "epoch": 4.5, + "grad_norm": 17.83046531677246, + "learning_rate": 0.00033817222038931045, + "loss": 0.3467, + "step": 1125 + }, + { + "epoch": 4.6, + "grad_norm": 4.865128517150879, + "learning_rate": 0.00033652259980204556, + "loss": 0.3117, + "step": 1150 + }, + { + "epoch": 4.7, + "grad_norm": 31.427154541015625, + "learning_rate": 0.0003348729792147806, + "loss": 0.3273, + "step": 1175 + }, + { + "epoch": 4.8, + "grad_norm": 26.77984619140625, + "learning_rate": 0.0003332233586275157, + "loss": 0.3504, + "step": 1200 + }, + { + "epoch": 4.9, + "grad_norm": 9.389993667602539, + "learning_rate": 0.00033157373804025076, + "loss": 0.3339, + "step": 1225 + }, + { + "epoch": 5.0, + "grad_norm": 25.70441246032715, + "learning_rate": 0.0003299241174529858, + "loss": 0.3336, + "step": 1250 + }, + { + "epoch": 5.1, + "grad_norm": 13.482085227966309, + "learning_rate": 0.0003282744968657209, + "loss": 0.2156, + "step": 1275 + }, + { + "epoch": 5.2, + "grad_norm": 8.725810050964355, + "learning_rate": 0.00032662487627845596, + "loss": 0.1811, + "step": 1300 + }, + { + "epoch": 5.3, + "grad_norm": 9.313215255737305, + "learning_rate": 0.00032497525569119106, + "loss": 0.2807, + "step": 1325 + }, + { + "epoch": 5.4, + "grad_norm": 11.026411056518555, + "learning_rate": 0.0003233256351039261, + "loss": 0.2757, + "step": 1350 + }, + { + "epoch": 5.5, + "grad_norm": 11.038985252380371, + "learning_rate": 0.00032167601451666116, + "loss": 0.2177, + "step": 1375 + }, + { + "epoch": 5.6, + "grad_norm": 4.008651256561279, + "learning_rate": 0.00032002639392939627, + "loss": 0.2163, + "step": 1400 + }, + { + "epoch": 5.7, + "grad_norm": 12.480770111083984, + "learning_rate": 0.0003183767733421313, + "loss": 0.2173, + "step": 1425 + }, + { + "epoch": 5.8, + "grad_norm": 8.751969337463379, + "learning_rate": 0.0003167271527548664, + "loss": 0.2299, + "step": 1450 + }, + { + "epoch": 5.9, + "grad_norm": 7.701971530914307, + "learning_rate": 0.00031507753216760147, + "loss": 0.1949, + "step": 1475 + }, + { + "epoch": 6.0, + "grad_norm": 8.48027515411377, + "learning_rate": 0.00031342791158033657, + "loss": 0.2599, + "step": 1500 + }, + { + "epoch": 6.1, + "grad_norm": 9.29404067993164, + "learning_rate": 0.00031177829099307157, + "loss": 0.1376, + "step": 1525 + }, + { + "epoch": 6.2, + "grad_norm": 20.137714385986328, + "learning_rate": 0.00031012867040580667, + "loss": 0.1647, + "step": 1550 + }, + { + "epoch": 6.3, + "grad_norm": 11.394575119018555, + "learning_rate": 0.0003084790498185417, + "loss": 0.1565, + "step": 1575 + }, + { + "epoch": 6.4, + "grad_norm": 8.214287757873535, + "learning_rate": 0.0003068294292312768, + "loss": 0.1739, + "step": 1600 + }, + { + "epoch": 6.5, + "grad_norm": 7.779988765716553, + "learning_rate": 0.0003051798086440119, + "loss": 0.1403, + "step": 1625 + }, + { + "epoch": 6.6, + "grad_norm": 9.421648025512695, + "learning_rate": 0.000303530188056747, + "loss": 0.1545, + "step": 1650 + }, + { + "epoch": 6.7, + "grad_norm": 5.751734256744385, + "learning_rate": 0.000301880567469482, + "loss": 0.1971, + "step": 1675 + }, + { + "epoch": 6.8, + "grad_norm": 23.861705780029297, + "learning_rate": 0.0003002309468822171, + "loss": 0.1681, + "step": 1700 + }, + { + "epoch": 6.9, + "grad_norm": 18.944721221923828, + "learning_rate": 0.0002985813262949522, + "loss": 0.1703, + "step": 1725 + }, + { + "epoch": 7.0, + "grad_norm": 14.045795440673828, + "learning_rate": 0.00029693170570768723, + "loss": 0.1801, + "step": 1750 + }, + { + "epoch": 7.1, + "grad_norm": 6.4620137214660645, + "learning_rate": 0.00029528208512042233, + "loss": 0.1253, + "step": 1775 + }, + { + "epoch": 7.2, + "grad_norm": 4.318169593811035, + "learning_rate": 0.0002936324645331574, + "loss": 0.1397, + "step": 1800 + }, + { + "epoch": 7.3, + "grad_norm": 24.91462516784668, + "learning_rate": 0.00029198284394589243, + "loss": 0.1259, + "step": 1825 + }, + { + "epoch": 7.4, + "grad_norm": 23.614572525024414, + "learning_rate": 0.00029033322335862753, + "loss": 0.1293, + "step": 1850 + }, + { + "epoch": 7.5, + "grad_norm": 3.60048508644104, + "learning_rate": 0.0002886836027713626, + "loss": 0.138, + "step": 1875 + }, + { + "epoch": 7.6, + "grad_norm": 16.62705421447754, + "learning_rate": 0.0002870339821840977, + "loss": 0.1733, + "step": 1900 + }, + { + "epoch": 7.7, + "grad_norm": 33.79671859741211, + "learning_rate": 0.00028538436159683273, + "loss": 0.217, + "step": 1925 + }, + { + "epoch": 7.8, + "grad_norm": 9.69206428527832, + "learning_rate": 0.00028373474100956784, + "loss": 0.1289, + "step": 1950 + }, + { + "epoch": 7.9, + "grad_norm": 18.655046463012695, + "learning_rate": 0.0002820851204223029, + "loss": 0.1168, + "step": 1975 + }, + { + "epoch": 8.0, + "grad_norm": 8.4110746383667, + "learning_rate": 0.00028043549983503794, + "loss": 0.1471, + "step": 2000 + }, + { + "epoch": 8.1, + "grad_norm": 14.473428726196289, + "learning_rate": 0.00027878587924777304, + "loss": 0.0774, + "step": 2025 + }, + { + "epoch": 8.2, + "grad_norm": 1.2464979887008667, + "learning_rate": 0.0002771362586605081, + "loss": 0.0715, + "step": 2050 + }, + { + "epoch": 8.3, + "grad_norm": 26.74981689453125, + "learning_rate": 0.0002754866380732432, + "loss": 0.1289, + "step": 2075 + }, + { + "epoch": 8.4, + "grad_norm": 13.958703994750977, + "learning_rate": 0.00027383701748597824, + "loss": 0.1294, + "step": 2100 + }, + { + "epoch": 8.5, + "grad_norm": 10.394835472106934, + "learning_rate": 0.0002721873968987133, + "loss": 0.0987, + "step": 2125 + }, + { + "epoch": 8.6, + "grad_norm": 9.387774467468262, + "learning_rate": 0.00027053777631144834, + "loss": 0.1269, + "step": 2150 + }, + { + "epoch": 8.7, + "grad_norm": 4.988718509674072, + "learning_rate": 0.00026888815572418344, + "loss": 0.1374, + "step": 2175 + }, + { + "epoch": 8.8, + "grad_norm": 12.66832447052002, + "learning_rate": 0.00026723853513691855, + "loss": 0.0997, + "step": 2200 + }, + { + "epoch": 8.9, + "grad_norm": 19.10486602783203, + "learning_rate": 0.0002655889145496536, + "loss": 0.1325, + "step": 2225 + }, + { + "epoch": 9.0, + "grad_norm": 3.2943971157073975, + "learning_rate": 0.0002639392939623887, + "loss": 0.1385, + "step": 2250 + }, + { + "epoch": 9.1, + "grad_norm": 11.490874290466309, + "learning_rate": 0.0002622896733751237, + "loss": 0.0573, + "step": 2275 + }, + { + "epoch": 9.2, + "grad_norm": 11.961288452148438, + "learning_rate": 0.0002606400527878588, + "loss": 0.1128, + "step": 2300 + }, + { + "epoch": 9.3, + "grad_norm": 0.7256277203559875, + "learning_rate": 0.00025899043220059385, + "loss": 0.0673, + "step": 2325 + }, + { + "epoch": 9.4, + "grad_norm": 9.857927322387695, + "learning_rate": 0.00025734081161332895, + "loss": 0.2072, + "step": 2350 + }, + { + "epoch": 9.5, + "grad_norm": 34.927734375, + "learning_rate": 0.000255691191026064, + "loss": 0.221, + "step": 2375 + }, + { + "epoch": 9.6, + "grad_norm": 28.461849212646484, + "learning_rate": 0.0002540415704387991, + "loss": 0.1263, + "step": 2400 + }, + { + "epoch": 9.7, + "grad_norm": 9.591217041015625, + "learning_rate": 0.00025239194985153415, + "loss": 0.099, + "step": 2425 + }, + { + "epoch": 9.8, + "grad_norm": 9.780253410339355, + "learning_rate": 0.0002507423292642692, + "loss": 0.0754, + "step": 2450 + }, + { + "epoch": 9.9, + "grad_norm": 14.394485473632812, + "learning_rate": 0.0002490927086770043, + "loss": 0.0909, + "step": 2475 + }, + { + "epoch": 10.0, + "grad_norm": 4.217277526855469, + "learning_rate": 0.00024744308808973936, + "loss": 0.1495, + "step": 2500 + }, + { + "epoch": 10.1, + "grad_norm": 0.08438724279403687, + "learning_rate": 0.00024579346750247446, + "loss": 0.0662, + "step": 2525 + }, + { + "epoch": 10.2, + "grad_norm": 0.9595862030982971, + "learning_rate": 0.0002441438469152095, + "loss": 0.084, + "step": 2550 + }, + { + "epoch": 10.3, + "grad_norm": 42.32374954223633, + "learning_rate": 0.00024249422632794456, + "loss": 0.0541, + "step": 2575 + }, + { + "epoch": 10.4, + "grad_norm": 24.792821884155273, + "learning_rate": 0.00024084460574067963, + "loss": 0.121, + "step": 2600 + }, + { + "epoch": 10.5, + "grad_norm": 2.7049331665039062, + "learning_rate": 0.0002391949851534147, + "loss": 0.1551, + "step": 2625 + }, + { + "epoch": 10.6, + "grad_norm": 2.788362741470337, + "learning_rate": 0.0002375453645661498, + "loss": 0.0877, + "step": 2650 + }, + { + "epoch": 10.7, + "grad_norm": 0.4999215006828308, + "learning_rate": 0.00023589574397888486, + "loss": 0.1263, + "step": 2675 + }, + { + "epoch": 10.8, + "grad_norm": 16.476675033569336, + "learning_rate": 0.00023424612339161997, + "loss": 0.0883, + "step": 2700 + }, + { + "epoch": 10.9, + "grad_norm": 38.192726135253906, + "learning_rate": 0.000232596502804355, + "loss": 0.0843, + "step": 2725 + }, + { + "epoch": 11.0, + "grad_norm": 18.09952735900879, + "learning_rate": 0.00023094688221709007, + "loss": 0.0957, + "step": 2750 + }, + { + "epoch": 11.1, + "grad_norm": 34.254417419433594, + "learning_rate": 0.00022929726162982514, + "loss": 0.0856, + "step": 2775 + }, + { + "epoch": 11.2, + "grad_norm": 11.744216918945312, + "learning_rate": 0.00022764764104256022, + "loss": 0.0654, + "step": 2800 + }, + { + "epoch": 11.3, + "grad_norm": 0.1285558044910431, + "learning_rate": 0.0002259980204552953, + "loss": 0.0638, + "step": 2825 + }, + { + "epoch": 11.4, + "grad_norm": 20.57583999633789, + "learning_rate": 0.00022434839986803037, + "loss": 0.079, + "step": 2850 + }, + { + "epoch": 11.5, + "grad_norm": 0.7192332148551941, + "learning_rate": 0.00022269877928076542, + "loss": 0.0854, + "step": 2875 + }, + { + "epoch": 11.6, + "grad_norm": 31.341829299926758, + "learning_rate": 0.0002210491586935005, + "loss": 0.063, + "step": 2900 + }, + { + "epoch": 11.7, + "grad_norm": 13.460247039794922, + "learning_rate": 0.00021939953810623557, + "loss": 0.1358, + "step": 2925 + }, + { + "epoch": 11.8, + "grad_norm": 1.671036720275879, + "learning_rate": 0.00021774991751897065, + "loss": 0.0806, + "step": 2950 + }, + { + "epoch": 11.9, + "grad_norm": 22.026491165161133, + "learning_rate": 0.00021610029693170573, + "loss": 0.0912, + "step": 2975 + }, + { + "epoch": 12.0, + "grad_norm": 46.78192138671875, + "learning_rate": 0.0002144506763444408, + "loss": 0.0763, + "step": 3000 + }, + { + "epoch": 12.1, + "grad_norm": 6.397649765014648, + "learning_rate": 0.00021280105575717582, + "loss": 0.0613, + "step": 3025 + }, + { + "epoch": 12.2, + "grad_norm": 7.179838180541992, + "learning_rate": 0.00021115143516991093, + "loss": 0.0496, + "step": 3050 + }, + { + "epoch": 12.3, + "grad_norm": 0.021825680509209633, + "learning_rate": 0.000209501814582646, + "loss": 0.0633, + "step": 3075 + }, + { + "epoch": 12.4, + "grad_norm": 6.6774187088012695, + "learning_rate": 0.00020785219399538108, + "loss": 0.0765, + "step": 3100 + }, + { + "epoch": 12.5, + "grad_norm": 0.13217756152153015, + "learning_rate": 0.00020620257340811616, + "loss": 0.0696, + "step": 3125 + }, + { + "epoch": 12.6, + "grad_norm": 2.5293831825256348, + "learning_rate": 0.00020455295282085123, + "loss": 0.0419, + "step": 3150 + }, + { + "epoch": 12.7, + "grad_norm": 1.1612874269485474, + "learning_rate": 0.00020290333223358626, + "loss": 0.1046, + "step": 3175 + }, + { + "epoch": 12.8, + "grad_norm": 40.93427658081055, + "learning_rate": 0.00020125371164632133, + "loss": 0.0488, + "step": 3200 + }, + { + "epoch": 12.9, + "grad_norm": 1.446478009223938, + "learning_rate": 0.0001996040910590564, + "loss": 0.062, + "step": 3225 + }, + { + "epoch": 13.0, + "grad_norm": 28.893821716308594, + "learning_rate": 0.00019795447047179148, + "loss": 0.0913, + "step": 3250 + }, + { + "epoch": 13.1, + "grad_norm": 23.31914520263672, + "learning_rate": 0.0001963048498845266, + "loss": 0.061, + "step": 3275 + }, + { + "epoch": 13.2, + "grad_norm": 0.26242795586586, + "learning_rate": 0.00019465522929726164, + "loss": 0.0563, + "step": 3300 + }, + { + "epoch": 13.3, + "grad_norm": 0.08909507840871811, + "learning_rate": 0.00019300560870999671, + "loss": 0.0387, + "step": 3325 + }, + { + "epoch": 13.4, + "grad_norm": 24.825326919555664, + "learning_rate": 0.0001913559881227318, + "loss": 0.0776, + "step": 3350 + }, + { + "epoch": 13.5, + "grad_norm": 39.43446731567383, + "learning_rate": 0.00018970636753546684, + "loss": 0.0846, + "step": 3375 + }, + { + "epoch": 13.6, + "grad_norm": 32.98988723754883, + "learning_rate": 0.00018805674694820192, + "loss": 0.0512, + "step": 3400 + }, + { + "epoch": 13.7, + "grad_norm": 4.053821563720703, + "learning_rate": 0.000186407126360937, + "loss": 0.1576, + "step": 3425 + }, + { + "epoch": 13.8, + "grad_norm": 0.2785554528236389, + "learning_rate": 0.00018475750577367207, + "loss": 0.0728, + "step": 3450 + }, + { + "epoch": 13.9, + "grad_norm": 11.39714527130127, + "learning_rate": 0.00018310788518640715, + "loss": 0.0776, + "step": 3475 + }, + { + "epoch": 14.0, + "grad_norm": 3.7018072605133057, + "learning_rate": 0.00018145826459914222, + "loss": 0.0736, + "step": 3500 + }, + { + "epoch": 14.1, + "grad_norm": 0.26053619384765625, + "learning_rate": 0.00017980864401187727, + "loss": 0.0598, + "step": 3525 + }, + { + "epoch": 14.2, + "grad_norm": 6.454179763793945, + "learning_rate": 0.00017815902342461235, + "loss": 0.0435, + "step": 3550 + }, + { + "epoch": 14.3, + "grad_norm": 34.47985076904297, + "learning_rate": 0.00017650940283734742, + "loss": 0.0467, + "step": 3575 + }, + { + "epoch": 14.4, + "grad_norm": 11.76491928100586, + "learning_rate": 0.00017485978225008247, + "loss": 0.1298, + "step": 3600 + }, + { + "epoch": 14.5, + "grad_norm": 26.64131736755371, + "learning_rate": 0.00017321016166281755, + "loss": 0.1313, + "step": 3625 + }, + { + "epoch": 14.6, + "grad_norm": 5.510906219482422, + "learning_rate": 0.00017156054107555263, + "loss": 0.0421, + "step": 3650 + }, + { + "epoch": 14.7, + "grad_norm": 0.04721131548285484, + "learning_rate": 0.0001699109204882877, + "loss": 0.0569, + "step": 3675 + }, + { + "epoch": 14.8, + "grad_norm": 0.6616227030754089, + "learning_rate": 0.00016826129990102278, + "loss": 0.0618, + "step": 3700 + }, + { + "epoch": 14.9, + "grad_norm": 0.05287986248731613, + "learning_rate": 0.00016661167931375785, + "loss": 0.0574, + "step": 3725 + }, + { + "epoch": 15.0, + "grad_norm": 0.49707508087158203, + "learning_rate": 0.0001649620587264929, + "loss": 0.094, + "step": 3750 + }, + { + "epoch": 15.1, + "grad_norm": 23.649728775024414, + "learning_rate": 0.00016331243813922798, + "loss": 0.0512, + "step": 3775 + }, + { + "epoch": 15.2, + "grad_norm": 0.01772051490843296, + "learning_rate": 0.00016166281755196306, + "loss": 0.0719, + "step": 3800 + }, + { + "epoch": 15.3, + "grad_norm": 1.5259039402008057, + "learning_rate": 0.00016001319696469813, + "loss": 0.0615, + "step": 3825 + }, + { + "epoch": 15.4, + "grad_norm": 0.026406478136777878, + "learning_rate": 0.0001583635763774332, + "loss": 0.0651, + "step": 3850 + }, + { + "epoch": 15.5, + "grad_norm": 27.12177085876465, + "learning_rate": 0.00015671395579016829, + "loss": 0.0458, + "step": 3875 + }, + { + "epoch": 15.6, + "grad_norm": 9.838811874389648, + "learning_rate": 0.00015506433520290334, + "loss": 0.0662, + "step": 3900 + }, + { + "epoch": 15.7, + "grad_norm": 0.08516795933246613, + "learning_rate": 0.0001534147146156384, + "loss": 0.0302, + "step": 3925 + }, + { + "epoch": 15.8, + "grad_norm": 1.1224850416183472, + "learning_rate": 0.0001517650940283735, + "loss": 0.0637, + "step": 3950 + }, + { + "epoch": 15.9, + "grad_norm": 0.14810702204704285, + "learning_rate": 0.00015011547344110854, + "loss": 0.0296, + "step": 3975 + }, + { + "epoch": 16.0, + "grad_norm": 0.046288371086120605, + "learning_rate": 0.00014846585285384361, + "loss": 0.0368, + "step": 4000 + }, + { + "epoch": 16.1, + "grad_norm": 0.008764918893575668, + "learning_rate": 0.0001468162322665787, + "loss": 0.0272, + "step": 4025 + }, + { + "epoch": 16.2, + "grad_norm": 0.39770299196243286, + "learning_rate": 0.00014516661167931377, + "loss": 0.0344, + "step": 4050 + }, + { + "epoch": 16.3, + "grad_norm": 23.780405044555664, + "learning_rate": 0.00014351699109204884, + "loss": 0.0617, + "step": 4075 + }, + { + "epoch": 16.4, + "grad_norm": 0.023103665560483932, + "learning_rate": 0.00014186737050478392, + "loss": 0.0182, + "step": 4100 + }, + { + "epoch": 16.5, + "grad_norm": 0.00998806394636631, + "learning_rate": 0.00014021774991751897, + "loss": 0.042, + "step": 4125 + }, + { + "epoch": 16.6, + "grad_norm": 43.775169372558594, + "learning_rate": 0.00013856812933025404, + "loss": 0.0182, + "step": 4150 + }, + { + "epoch": 16.7, + "grad_norm": 0.008157053031027317, + "learning_rate": 0.00013691850874298912, + "loss": 0.0377, + "step": 4175 + }, + { + "epoch": 16.8, + "grad_norm": 0.011455570347607136, + "learning_rate": 0.00013526888815572417, + "loss": 0.0255, + "step": 4200 + }, + { + "epoch": 16.9, + "grad_norm": 34.8545036315918, + "learning_rate": 0.00013361926756845927, + "loss": 0.0486, + "step": 4225 + }, + { + "epoch": 17.0, + "grad_norm": 0.006066238507628441, + "learning_rate": 0.00013196964698119435, + "loss": 0.0821, + "step": 4250 + }, + { + "epoch": 17.1, + "grad_norm": 43.46882247924805, + "learning_rate": 0.0001303200263939294, + "loss": 0.0333, + "step": 4275 + }, + { + "epoch": 17.2, + "grad_norm": 2.3177573680877686, + "learning_rate": 0.00012867040580666448, + "loss": 0.0477, + "step": 4300 + }, + { + "epoch": 17.3, + "grad_norm": 0.2264157235622406, + "learning_rate": 0.00012702078521939955, + "loss": 0.0373, + "step": 4325 + }, + { + "epoch": 17.4, + "grad_norm": 68.17675018310547, + "learning_rate": 0.0001253711646321346, + "loss": 0.036, + "step": 4350 + }, + { + "epoch": 17.5, + "grad_norm": 1.5359147787094116, + "learning_rate": 0.00012372154404486968, + "loss": 0.0437, + "step": 4375 + }, + { + "epoch": 17.6, + "grad_norm": 0.008654219098389149, + "learning_rate": 0.00012207192345760475, + "loss": 0.0295, + "step": 4400 + }, + { + "epoch": 17.7, + "grad_norm": 0.0074989828281104565, + "learning_rate": 0.00012042230287033982, + "loss": 0.0229, + "step": 4425 + }, + { + "epoch": 17.8, + "grad_norm": 0.03568067401647568, + "learning_rate": 0.0001187726822830749, + "loss": 0.0636, + "step": 4450 + }, + { + "epoch": 17.9, + "grad_norm": 0.0151940006762743, + "learning_rate": 0.00011712306169580998, + "loss": 0.0669, + "step": 4475 + }, + { + "epoch": 18.0, + "grad_norm": 0.04830991476774216, + "learning_rate": 0.00011547344110854503, + "loss": 0.0388, + "step": 4500 + }, + { + "epoch": 18.1, + "grad_norm": 1.1794861555099487, + "learning_rate": 0.00011382382052128011, + "loss": 0.0069, + "step": 4525 + }, + { + "epoch": 18.2, + "grad_norm": 13.07977294921875, + "learning_rate": 0.00011217419993401519, + "loss": 0.0315, + "step": 4550 + }, + { + "epoch": 18.3, + "grad_norm": 0.011115381494164467, + "learning_rate": 0.00011052457934675025, + "loss": 0.0144, + "step": 4575 + }, + { + "epoch": 18.4, + "grad_norm": 25.42391014099121, + "learning_rate": 0.00010887495875948532, + "loss": 0.0198, + "step": 4600 + }, + { + "epoch": 18.5, + "grad_norm": 44.36895751953125, + "learning_rate": 0.0001072253381722204, + "loss": 0.0297, + "step": 4625 + }, + { + "epoch": 18.6, + "grad_norm": 6.828146457672119, + "learning_rate": 0.00010557571758495546, + "loss": 0.0348, + "step": 4650 + }, + { + "epoch": 18.7, + "grad_norm": 38.9465446472168, + "learning_rate": 0.00010392609699769054, + "loss": 0.0336, + "step": 4675 + }, + { + "epoch": 18.8, + "grad_norm": 0.015516690909862518, + "learning_rate": 0.00010227647641042562, + "loss": 0.032, + "step": 4700 + }, + { + "epoch": 18.9, + "grad_norm": 5.419480323791504, + "learning_rate": 0.00010062685582316067, + "loss": 0.0169, + "step": 4725 + }, + { + "epoch": 19.0, + "grad_norm": 0.7849720120429993, + "learning_rate": 9.897723523589574e-05, + "loss": 0.0321, + "step": 4750 + }, + { + "epoch": 19.1, + "grad_norm": 25.438196182250977, + "learning_rate": 9.732761464863082e-05, + "loss": 0.046, + "step": 4775 + }, + { + "epoch": 19.2, + "grad_norm": 0.010676453821361065, + "learning_rate": 9.56779940613659e-05, + "loss": 0.0111, + "step": 4800 + }, + { + "epoch": 19.3, + "grad_norm": 0.011585243977606297, + "learning_rate": 9.402837347410096e-05, + "loss": 0.0154, + "step": 4825 + }, + { + "epoch": 19.4, + "grad_norm": 0.012879762798547745, + "learning_rate": 9.237875288683603e-05, + "loss": 0.02, + "step": 4850 + }, + { + "epoch": 19.5, + "grad_norm": 0.4159695506095886, + "learning_rate": 9.072913229957111e-05, + "loss": 0.0277, + "step": 4875 + }, + { + "epoch": 19.6, + "grad_norm": 0.013872765935957432, + "learning_rate": 8.907951171230617e-05, + "loss": 0.0069, + "step": 4900 + }, + { + "epoch": 19.7, + "grad_norm": 3.4260928630828857, + "learning_rate": 8.742989112504124e-05, + "loss": 0.0319, + "step": 4925 + }, + { + "epoch": 19.8, + "grad_norm": 0.043105900287628174, + "learning_rate": 8.578027053777631e-05, + "loss": 0.0445, + "step": 4950 + }, + { + "epoch": 19.9, + "grad_norm": 10.97398853302002, + "learning_rate": 8.413064995051139e-05, + "loss": 0.0185, + "step": 4975 + }, + { + "epoch": 20.0, + "grad_norm": 12.879217147827148, + "learning_rate": 8.248102936324645e-05, + "loss": 0.0197, + "step": 5000 + } + ], + "logging_steps": 25, + "max_steps": 6250, + "num_input_tokens_seen": 0, + "num_train_epochs": 25, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 254097408000000.0, + "train_batch_size": 20, + "trial_name": null, + "trial_params": null +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-5000/training_args.bin b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-5000/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..798d8a9f3a9513692035320aba5484c89ac49e69 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-5000/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1cabf5db3e9d79ec0c1945933395ff57418204414cc81f77c90324d003fac782 +size 5304 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-5000/vocab.txt b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-5000/vocab.txt new file mode 100644 index 0000000000000000000000000000000000000000..cc9ad53732e21ea9180f18e1fb67a38cc5702d4c --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-5000/vocab.txt @@ -0,0 +1,4101 @@ +[PAD] +[UNK] +[CLS] +[SEP] +[MASK] +AAAAAA +AAAAAT +AAAAAC +AAAAAG +AAAATA +AAAATT +AAAATC +AAAATG +AAAACA +AAAACT +AAAACC +AAAACG +AAAAGA +AAAAGT +AAAAGC +AAAAGG +AAATAA +AAATAT +AAATAC +AAATAG +AAATTA +AAATTT +AAATTC +AAATTG +AAATCA +AAATCT +AAATCC +AAATCG +AAATGA +AAATGT +AAATGC +AAATGG +AAACAA +AAACAT +AAACAC +AAACAG +AAACTA +AAACTT +AAACTC +AAACTG +AAACCA +AAACCT +AAACCC +AAACCG +AAACGA +AAACGT +AAACGC +AAACGG +AAAGAA +AAAGAT +AAAGAC +AAAGAG +AAAGTA +AAAGTT +AAAGTC +AAAGTG +AAAGCA +AAAGCT +AAAGCC +AAAGCG +AAAGGA +AAAGGT +AAAGGC +AAAGGG +AATAAA +AATAAT +AATAAC +AATAAG +AATATA +AATATT +AATATC +AATATG +AATACA +AATACT +AATACC +AATACG +AATAGA +AATAGT +AATAGC +AATAGG +AATTAA +AATTAT +AATTAC +AATTAG +AATTTA +AATTTT +AATTTC +AATTTG +AATTCA +AATTCT +AATTCC +AATTCG +AATTGA +AATTGT +AATTGC +AATTGG +AATCAA +AATCAT +AATCAC +AATCAG +AATCTA +AATCTT +AATCTC +AATCTG +AATCCA +AATCCT +AATCCC +AATCCG +AATCGA +AATCGT +AATCGC +AATCGG +AATGAA +AATGAT +AATGAC +AATGAG +AATGTA +AATGTT +AATGTC +AATGTG +AATGCA +AATGCT +AATGCC +AATGCG +AATGGA +AATGGT +AATGGC +AATGGG +AACAAA +AACAAT +AACAAC +AACAAG +AACATA +AACATT +AACATC +AACATG +AACACA +AACACT +AACACC +AACACG +AACAGA +AACAGT +AACAGC +AACAGG +AACTAA +AACTAT +AACTAC +AACTAG +AACTTA +AACTTT +AACTTC +AACTTG +AACTCA +AACTCT +AACTCC +AACTCG +AACTGA +AACTGT +AACTGC +AACTGG +AACCAA +AACCAT +AACCAC +AACCAG +AACCTA +AACCTT +AACCTC +AACCTG +AACCCA +AACCCT +AACCCC +AACCCG +AACCGA +AACCGT +AACCGC +AACCGG +AACGAA +AACGAT +AACGAC +AACGAG +AACGTA +AACGTT +AACGTC +AACGTG +AACGCA +AACGCT +AACGCC +AACGCG +AACGGA +AACGGT +AACGGC +AACGGG +AAGAAA +AAGAAT +AAGAAC +AAGAAG +AAGATA +AAGATT +AAGATC +AAGATG +AAGACA +AAGACT +AAGACC +AAGACG +AAGAGA +AAGAGT +AAGAGC +AAGAGG +AAGTAA +AAGTAT +AAGTAC +AAGTAG +AAGTTA +AAGTTT +AAGTTC +AAGTTG +AAGTCA +AAGTCT +AAGTCC +AAGTCG +AAGTGA +AAGTGT +AAGTGC +AAGTGG +AAGCAA +AAGCAT +AAGCAC +AAGCAG +AAGCTA +AAGCTT +AAGCTC +AAGCTG +AAGCCA +AAGCCT +AAGCCC +AAGCCG +AAGCGA +AAGCGT +AAGCGC +AAGCGG +AAGGAA +AAGGAT +AAGGAC +AAGGAG +AAGGTA +AAGGTT +AAGGTC +AAGGTG +AAGGCA +AAGGCT +AAGGCC +AAGGCG +AAGGGA +AAGGGT +AAGGGC +AAGGGG +ATAAAA +ATAAAT +ATAAAC +ATAAAG +ATAATA +ATAATT +ATAATC +ATAATG +ATAACA +ATAACT +ATAACC +ATAACG +ATAAGA +ATAAGT +ATAAGC +ATAAGG +ATATAA +ATATAT +ATATAC +ATATAG +ATATTA +ATATTT +ATATTC +ATATTG +ATATCA +ATATCT +ATATCC +ATATCG +ATATGA +ATATGT +ATATGC +ATATGG +ATACAA +ATACAT +ATACAC +ATACAG +ATACTA +ATACTT +ATACTC +ATACTG +ATACCA +ATACCT +ATACCC +ATACCG +ATACGA +ATACGT +ATACGC +ATACGG +ATAGAA +ATAGAT +ATAGAC +ATAGAG +ATAGTA +ATAGTT +ATAGTC +ATAGTG +ATAGCA +ATAGCT +ATAGCC +ATAGCG +ATAGGA +ATAGGT +ATAGGC +ATAGGG +ATTAAA +ATTAAT +ATTAAC +ATTAAG +ATTATA +ATTATT +ATTATC +ATTATG +ATTACA +ATTACT +ATTACC +ATTACG +ATTAGA +ATTAGT +ATTAGC +ATTAGG +ATTTAA +ATTTAT +ATTTAC +ATTTAG +ATTTTA +ATTTTT +ATTTTC +ATTTTG +ATTTCA +ATTTCT +ATTTCC +ATTTCG +ATTTGA +ATTTGT +ATTTGC +ATTTGG +ATTCAA +ATTCAT +ATTCAC +ATTCAG +ATTCTA +ATTCTT +ATTCTC +ATTCTG +ATTCCA +ATTCCT +ATTCCC +ATTCCG +ATTCGA +ATTCGT +ATTCGC +ATTCGG +ATTGAA +ATTGAT +ATTGAC +ATTGAG +ATTGTA +ATTGTT +ATTGTC +ATTGTG +ATTGCA +ATTGCT +ATTGCC +ATTGCG +ATTGGA +ATTGGT +ATTGGC +ATTGGG +ATCAAA +ATCAAT +ATCAAC +ATCAAG +ATCATA +ATCATT +ATCATC +ATCATG +ATCACA +ATCACT +ATCACC +ATCACG +ATCAGA +ATCAGT +ATCAGC +ATCAGG +ATCTAA +ATCTAT +ATCTAC +ATCTAG +ATCTTA +ATCTTT +ATCTTC +ATCTTG +ATCTCA +ATCTCT +ATCTCC +ATCTCG +ATCTGA +ATCTGT +ATCTGC +ATCTGG +ATCCAA +ATCCAT +ATCCAC +ATCCAG +ATCCTA +ATCCTT +ATCCTC +ATCCTG +ATCCCA +ATCCCT +ATCCCC +ATCCCG +ATCCGA +ATCCGT +ATCCGC +ATCCGG +ATCGAA +ATCGAT +ATCGAC +ATCGAG +ATCGTA +ATCGTT +ATCGTC +ATCGTG +ATCGCA +ATCGCT +ATCGCC +ATCGCG +ATCGGA +ATCGGT +ATCGGC +ATCGGG +ATGAAA +ATGAAT +ATGAAC +ATGAAG +ATGATA +ATGATT +ATGATC +ATGATG +ATGACA +ATGACT +ATGACC +ATGACG +ATGAGA +ATGAGT +ATGAGC +ATGAGG +ATGTAA +ATGTAT +ATGTAC +ATGTAG +ATGTTA +ATGTTT +ATGTTC +ATGTTG +ATGTCA +ATGTCT +ATGTCC +ATGTCG +ATGTGA +ATGTGT +ATGTGC +ATGTGG +ATGCAA +ATGCAT +ATGCAC +ATGCAG +ATGCTA +ATGCTT +ATGCTC +ATGCTG +ATGCCA +ATGCCT +ATGCCC +ATGCCG +ATGCGA +ATGCGT +ATGCGC +ATGCGG +ATGGAA +ATGGAT +ATGGAC +ATGGAG +ATGGTA +ATGGTT +ATGGTC +ATGGTG +ATGGCA +ATGGCT +ATGGCC +ATGGCG +ATGGGA +ATGGGT +ATGGGC +ATGGGG +ACAAAA +ACAAAT +ACAAAC +ACAAAG +ACAATA +ACAATT +ACAATC +ACAATG +ACAACA +ACAACT +ACAACC +ACAACG +ACAAGA +ACAAGT +ACAAGC +ACAAGG +ACATAA +ACATAT +ACATAC +ACATAG +ACATTA +ACATTT +ACATTC +ACATTG +ACATCA +ACATCT +ACATCC +ACATCG +ACATGA +ACATGT +ACATGC +ACATGG +ACACAA +ACACAT +ACACAC +ACACAG +ACACTA +ACACTT +ACACTC +ACACTG +ACACCA +ACACCT +ACACCC +ACACCG +ACACGA +ACACGT +ACACGC +ACACGG +ACAGAA +ACAGAT +ACAGAC +ACAGAG +ACAGTA +ACAGTT +ACAGTC +ACAGTG +ACAGCA +ACAGCT +ACAGCC +ACAGCG +ACAGGA +ACAGGT +ACAGGC +ACAGGG +ACTAAA +ACTAAT +ACTAAC +ACTAAG +ACTATA +ACTATT +ACTATC +ACTATG +ACTACA +ACTACT +ACTACC +ACTACG +ACTAGA +ACTAGT +ACTAGC +ACTAGG +ACTTAA +ACTTAT +ACTTAC +ACTTAG +ACTTTA +ACTTTT +ACTTTC +ACTTTG +ACTTCA +ACTTCT +ACTTCC +ACTTCG +ACTTGA +ACTTGT +ACTTGC +ACTTGG +ACTCAA +ACTCAT +ACTCAC +ACTCAG +ACTCTA +ACTCTT +ACTCTC +ACTCTG +ACTCCA +ACTCCT +ACTCCC +ACTCCG +ACTCGA +ACTCGT +ACTCGC +ACTCGG +ACTGAA +ACTGAT +ACTGAC +ACTGAG +ACTGTA +ACTGTT +ACTGTC +ACTGTG +ACTGCA +ACTGCT +ACTGCC +ACTGCG +ACTGGA +ACTGGT +ACTGGC +ACTGGG +ACCAAA +ACCAAT +ACCAAC +ACCAAG +ACCATA +ACCATT +ACCATC +ACCATG +ACCACA +ACCACT +ACCACC +ACCACG +ACCAGA +ACCAGT +ACCAGC +ACCAGG +ACCTAA +ACCTAT +ACCTAC +ACCTAG +ACCTTA +ACCTTT +ACCTTC +ACCTTG +ACCTCA +ACCTCT +ACCTCC +ACCTCG +ACCTGA +ACCTGT +ACCTGC +ACCTGG +ACCCAA +ACCCAT +ACCCAC +ACCCAG +ACCCTA +ACCCTT +ACCCTC +ACCCTG +ACCCCA +ACCCCT +ACCCCC +ACCCCG +ACCCGA +ACCCGT +ACCCGC +ACCCGG +ACCGAA +ACCGAT +ACCGAC +ACCGAG +ACCGTA +ACCGTT +ACCGTC +ACCGTG +ACCGCA +ACCGCT +ACCGCC +ACCGCG +ACCGGA +ACCGGT +ACCGGC +ACCGGG +ACGAAA +ACGAAT +ACGAAC +ACGAAG +ACGATA +ACGATT +ACGATC +ACGATG +ACGACA +ACGACT +ACGACC +ACGACG +ACGAGA +ACGAGT +ACGAGC +ACGAGG +ACGTAA +ACGTAT +ACGTAC +ACGTAG +ACGTTA +ACGTTT +ACGTTC +ACGTTG +ACGTCA +ACGTCT +ACGTCC +ACGTCG +ACGTGA +ACGTGT +ACGTGC +ACGTGG +ACGCAA +ACGCAT +ACGCAC +ACGCAG +ACGCTA +ACGCTT +ACGCTC +ACGCTG +ACGCCA +ACGCCT +ACGCCC +ACGCCG +ACGCGA +ACGCGT +ACGCGC +ACGCGG +ACGGAA +ACGGAT +ACGGAC +ACGGAG +ACGGTA +ACGGTT +ACGGTC +ACGGTG +ACGGCA +ACGGCT +ACGGCC +ACGGCG +ACGGGA +ACGGGT +ACGGGC +ACGGGG +AGAAAA +AGAAAT +AGAAAC +AGAAAG +AGAATA +AGAATT +AGAATC +AGAATG +AGAACA +AGAACT +AGAACC +AGAACG +AGAAGA +AGAAGT +AGAAGC +AGAAGG +AGATAA +AGATAT +AGATAC +AGATAG +AGATTA +AGATTT +AGATTC +AGATTG +AGATCA +AGATCT +AGATCC +AGATCG +AGATGA +AGATGT +AGATGC +AGATGG +AGACAA +AGACAT +AGACAC +AGACAG +AGACTA +AGACTT +AGACTC +AGACTG +AGACCA +AGACCT +AGACCC +AGACCG +AGACGA +AGACGT +AGACGC +AGACGG +AGAGAA +AGAGAT +AGAGAC +AGAGAG +AGAGTA +AGAGTT +AGAGTC +AGAGTG +AGAGCA +AGAGCT +AGAGCC +AGAGCG +AGAGGA +AGAGGT +AGAGGC +AGAGGG +AGTAAA +AGTAAT +AGTAAC +AGTAAG +AGTATA +AGTATT +AGTATC +AGTATG +AGTACA +AGTACT +AGTACC +AGTACG +AGTAGA +AGTAGT +AGTAGC +AGTAGG +AGTTAA +AGTTAT +AGTTAC +AGTTAG +AGTTTA +AGTTTT +AGTTTC +AGTTTG +AGTTCA +AGTTCT +AGTTCC +AGTTCG +AGTTGA +AGTTGT +AGTTGC +AGTTGG +AGTCAA +AGTCAT +AGTCAC +AGTCAG +AGTCTA +AGTCTT +AGTCTC +AGTCTG +AGTCCA +AGTCCT +AGTCCC +AGTCCG +AGTCGA +AGTCGT +AGTCGC +AGTCGG +AGTGAA +AGTGAT +AGTGAC +AGTGAG +AGTGTA +AGTGTT +AGTGTC +AGTGTG +AGTGCA +AGTGCT +AGTGCC +AGTGCG +AGTGGA +AGTGGT +AGTGGC +AGTGGG +AGCAAA +AGCAAT +AGCAAC +AGCAAG +AGCATA +AGCATT +AGCATC +AGCATG +AGCACA +AGCACT +AGCACC +AGCACG +AGCAGA +AGCAGT +AGCAGC +AGCAGG +AGCTAA +AGCTAT +AGCTAC +AGCTAG +AGCTTA +AGCTTT +AGCTTC +AGCTTG +AGCTCA +AGCTCT +AGCTCC +AGCTCG +AGCTGA +AGCTGT +AGCTGC +AGCTGG +AGCCAA +AGCCAT +AGCCAC +AGCCAG +AGCCTA +AGCCTT +AGCCTC +AGCCTG +AGCCCA +AGCCCT +AGCCCC +AGCCCG +AGCCGA +AGCCGT +AGCCGC +AGCCGG +AGCGAA +AGCGAT +AGCGAC +AGCGAG +AGCGTA +AGCGTT +AGCGTC +AGCGTG +AGCGCA +AGCGCT +AGCGCC +AGCGCG +AGCGGA +AGCGGT +AGCGGC +AGCGGG +AGGAAA +AGGAAT +AGGAAC +AGGAAG +AGGATA +AGGATT +AGGATC +AGGATG +AGGACA +AGGACT +AGGACC +AGGACG +AGGAGA +AGGAGT +AGGAGC +AGGAGG +AGGTAA +AGGTAT +AGGTAC +AGGTAG +AGGTTA +AGGTTT +AGGTTC +AGGTTG +AGGTCA +AGGTCT +AGGTCC +AGGTCG +AGGTGA +AGGTGT +AGGTGC +AGGTGG +AGGCAA +AGGCAT +AGGCAC +AGGCAG +AGGCTA +AGGCTT +AGGCTC +AGGCTG +AGGCCA +AGGCCT +AGGCCC +AGGCCG +AGGCGA +AGGCGT +AGGCGC +AGGCGG +AGGGAA +AGGGAT +AGGGAC +AGGGAG +AGGGTA +AGGGTT +AGGGTC +AGGGTG +AGGGCA +AGGGCT +AGGGCC +AGGGCG +AGGGGA +AGGGGT +AGGGGC +AGGGGG +TAAAAA +TAAAAT +TAAAAC +TAAAAG +TAAATA +TAAATT +TAAATC +TAAATG +TAAACA +TAAACT +TAAACC +TAAACG +TAAAGA +TAAAGT +TAAAGC +TAAAGG +TAATAA +TAATAT +TAATAC +TAATAG +TAATTA +TAATTT +TAATTC +TAATTG +TAATCA +TAATCT +TAATCC +TAATCG +TAATGA +TAATGT +TAATGC +TAATGG +TAACAA +TAACAT +TAACAC +TAACAG +TAACTA +TAACTT +TAACTC +TAACTG +TAACCA +TAACCT +TAACCC +TAACCG +TAACGA +TAACGT +TAACGC +TAACGG +TAAGAA +TAAGAT +TAAGAC +TAAGAG +TAAGTA +TAAGTT +TAAGTC +TAAGTG +TAAGCA +TAAGCT +TAAGCC +TAAGCG +TAAGGA +TAAGGT +TAAGGC +TAAGGG +TATAAA +TATAAT +TATAAC +TATAAG +TATATA +TATATT +TATATC +TATATG +TATACA +TATACT +TATACC +TATACG +TATAGA +TATAGT +TATAGC +TATAGG +TATTAA +TATTAT +TATTAC +TATTAG +TATTTA +TATTTT +TATTTC +TATTTG +TATTCA +TATTCT +TATTCC +TATTCG +TATTGA +TATTGT +TATTGC +TATTGG +TATCAA +TATCAT +TATCAC +TATCAG +TATCTA +TATCTT +TATCTC +TATCTG +TATCCA +TATCCT +TATCCC +TATCCG +TATCGA +TATCGT +TATCGC +TATCGG +TATGAA +TATGAT +TATGAC +TATGAG +TATGTA +TATGTT +TATGTC +TATGTG +TATGCA +TATGCT +TATGCC +TATGCG +TATGGA +TATGGT +TATGGC +TATGGG +TACAAA +TACAAT +TACAAC +TACAAG +TACATA +TACATT +TACATC +TACATG +TACACA +TACACT +TACACC +TACACG +TACAGA +TACAGT +TACAGC +TACAGG +TACTAA +TACTAT +TACTAC +TACTAG +TACTTA +TACTTT +TACTTC +TACTTG +TACTCA +TACTCT +TACTCC +TACTCG +TACTGA +TACTGT +TACTGC +TACTGG +TACCAA +TACCAT +TACCAC +TACCAG +TACCTA +TACCTT +TACCTC +TACCTG +TACCCA +TACCCT +TACCCC +TACCCG +TACCGA +TACCGT +TACCGC +TACCGG +TACGAA +TACGAT +TACGAC +TACGAG +TACGTA +TACGTT +TACGTC +TACGTG +TACGCA +TACGCT +TACGCC +TACGCG +TACGGA +TACGGT +TACGGC +TACGGG +TAGAAA +TAGAAT +TAGAAC +TAGAAG +TAGATA +TAGATT +TAGATC +TAGATG +TAGACA +TAGACT +TAGACC +TAGACG +TAGAGA +TAGAGT +TAGAGC +TAGAGG +TAGTAA +TAGTAT +TAGTAC +TAGTAG +TAGTTA +TAGTTT +TAGTTC +TAGTTG +TAGTCA +TAGTCT +TAGTCC +TAGTCG +TAGTGA +TAGTGT +TAGTGC +TAGTGG +TAGCAA +TAGCAT +TAGCAC +TAGCAG +TAGCTA +TAGCTT +TAGCTC +TAGCTG +TAGCCA +TAGCCT +TAGCCC +TAGCCG +TAGCGA +TAGCGT +TAGCGC +TAGCGG +TAGGAA +TAGGAT +TAGGAC +TAGGAG +TAGGTA +TAGGTT +TAGGTC +TAGGTG +TAGGCA +TAGGCT +TAGGCC +TAGGCG +TAGGGA +TAGGGT +TAGGGC +TAGGGG +TTAAAA +TTAAAT +TTAAAC +TTAAAG +TTAATA +TTAATT +TTAATC +TTAATG +TTAACA +TTAACT +TTAACC +TTAACG +TTAAGA +TTAAGT +TTAAGC +TTAAGG +TTATAA +TTATAT +TTATAC +TTATAG +TTATTA +TTATTT +TTATTC +TTATTG +TTATCA +TTATCT +TTATCC +TTATCG +TTATGA +TTATGT +TTATGC +TTATGG +TTACAA +TTACAT +TTACAC +TTACAG +TTACTA +TTACTT +TTACTC +TTACTG +TTACCA +TTACCT +TTACCC +TTACCG +TTACGA +TTACGT +TTACGC +TTACGG +TTAGAA +TTAGAT +TTAGAC +TTAGAG +TTAGTA +TTAGTT +TTAGTC +TTAGTG +TTAGCA +TTAGCT +TTAGCC +TTAGCG +TTAGGA +TTAGGT +TTAGGC +TTAGGG +TTTAAA +TTTAAT +TTTAAC +TTTAAG +TTTATA +TTTATT +TTTATC +TTTATG +TTTACA +TTTACT +TTTACC +TTTACG +TTTAGA +TTTAGT +TTTAGC +TTTAGG +TTTTAA +TTTTAT +TTTTAC +TTTTAG +TTTTTA +TTTTTT +TTTTTC +TTTTTG +TTTTCA +TTTTCT +TTTTCC +TTTTCG +TTTTGA +TTTTGT +TTTTGC +TTTTGG +TTTCAA +TTTCAT +TTTCAC +TTTCAG +TTTCTA +TTTCTT +TTTCTC +TTTCTG +TTTCCA +TTTCCT +TTTCCC +TTTCCG +TTTCGA +TTTCGT +TTTCGC +TTTCGG +TTTGAA +TTTGAT +TTTGAC +TTTGAG +TTTGTA +TTTGTT +TTTGTC +TTTGTG +TTTGCA +TTTGCT +TTTGCC +TTTGCG +TTTGGA +TTTGGT +TTTGGC +TTTGGG +TTCAAA +TTCAAT +TTCAAC +TTCAAG +TTCATA +TTCATT +TTCATC +TTCATG +TTCACA +TTCACT +TTCACC +TTCACG +TTCAGA +TTCAGT +TTCAGC +TTCAGG +TTCTAA +TTCTAT +TTCTAC +TTCTAG +TTCTTA +TTCTTT +TTCTTC +TTCTTG +TTCTCA +TTCTCT +TTCTCC +TTCTCG +TTCTGA +TTCTGT +TTCTGC +TTCTGG +TTCCAA +TTCCAT +TTCCAC +TTCCAG +TTCCTA +TTCCTT +TTCCTC +TTCCTG +TTCCCA +TTCCCT +TTCCCC +TTCCCG +TTCCGA +TTCCGT +TTCCGC +TTCCGG +TTCGAA +TTCGAT +TTCGAC +TTCGAG +TTCGTA +TTCGTT +TTCGTC +TTCGTG +TTCGCA +TTCGCT +TTCGCC +TTCGCG +TTCGGA +TTCGGT +TTCGGC +TTCGGG +TTGAAA +TTGAAT +TTGAAC +TTGAAG +TTGATA +TTGATT +TTGATC +TTGATG +TTGACA +TTGACT +TTGACC +TTGACG +TTGAGA +TTGAGT +TTGAGC +TTGAGG +TTGTAA +TTGTAT +TTGTAC +TTGTAG +TTGTTA +TTGTTT +TTGTTC +TTGTTG +TTGTCA +TTGTCT +TTGTCC +TTGTCG +TTGTGA +TTGTGT +TTGTGC +TTGTGG +TTGCAA +TTGCAT +TTGCAC +TTGCAG +TTGCTA +TTGCTT +TTGCTC +TTGCTG +TTGCCA +TTGCCT +TTGCCC +TTGCCG +TTGCGA +TTGCGT +TTGCGC +TTGCGG +TTGGAA +TTGGAT +TTGGAC +TTGGAG +TTGGTA +TTGGTT +TTGGTC +TTGGTG +TTGGCA +TTGGCT +TTGGCC +TTGGCG +TTGGGA +TTGGGT +TTGGGC +TTGGGG +TCAAAA +TCAAAT +TCAAAC +TCAAAG +TCAATA +TCAATT +TCAATC +TCAATG +TCAACA +TCAACT +TCAACC +TCAACG +TCAAGA +TCAAGT +TCAAGC +TCAAGG +TCATAA +TCATAT +TCATAC +TCATAG +TCATTA +TCATTT +TCATTC +TCATTG +TCATCA +TCATCT +TCATCC +TCATCG +TCATGA +TCATGT +TCATGC +TCATGG +TCACAA +TCACAT +TCACAC +TCACAG +TCACTA +TCACTT +TCACTC +TCACTG +TCACCA +TCACCT +TCACCC +TCACCG +TCACGA +TCACGT +TCACGC +TCACGG +TCAGAA +TCAGAT +TCAGAC +TCAGAG +TCAGTA +TCAGTT +TCAGTC +TCAGTG +TCAGCA +TCAGCT +TCAGCC +TCAGCG +TCAGGA +TCAGGT +TCAGGC +TCAGGG +TCTAAA +TCTAAT +TCTAAC +TCTAAG +TCTATA +TCTATT +TCTATC +TCTATG +TCTACA +TCTACT +TCTACC +TCTACG +TCTAGA +TCTAGT +TCTAGC +TCTAGG +TCTTAA +TCTTAT +TCTTAC +TCTTAG +TCTTTA +TCTTTT +TCTTTC +TCTTTG +TCTTCA +TCTTCT +TCTTCC +TCTTCG +TCTTGA +TCTTGT +TCTTGC +TCTTGG +TCTCAA +TCTCAT +TCTCAC +TCTCAG +TCTCTA +TCTCTT +TCTCTC +TCTCTG +TCTCCA +TCTCCT +TCTCCC +TCTCCG +TCTCGA +TCTCGT +TCTCGC +TCTCGG +TCTGAA +TCTGAT +TCTGAC +TCTGAG +TCTGTA +TCTGTT +TCTGTC +TCTGTG +TCTGCA +TCTGCT +TCTGCC +TCTGCG +TCTGGA +TCTGGT +TCTGGC +TCTGGG +TCCAAA +TCCAAT +TCCAAC +TCCAAG +TCCATA +TCCATT +TCCATC +TCCATG +TCCACA +TCCACT +TCCACC +TCCACG +TCCAGA +TCCAGT +TCCAGC +TCCAGG +TCCTAA +TCCTAT +TCCTAC +TCCTAG +TCCTTA +TCCTTT +TCCTTC +TCCTTG +TCCTCA +TCCTCT +TCCTCC +TCCTCG +TCCTGA +TCCTGT +TCCTGC +TCCTGG +TCCCAA +TCCCAT +TCCCAC +TCCCAG +TCCCTA +TCCCTT +TCCCTC +TCCCTG +TCCCCA +TCCCCT +TCCCCC +TCCCCG +TCCCGA +TCCCGT +TCCCGC +TCCCGG +TCCGAA +TCCGAT +TCCGAC +TCCGAG +TCCGTA +TCCGTT +TCCGTC +TCCGTG +TCCGCA +TCCGCT +TCCGCC +TCCGCG +TCCGGA +TCCGGT +TCCGGC +TCCGGG +TCGAAA +TCGAAT +TCGAAC +TCGAAG +TCGATA +TCGATT +TCGATC +TCGATG +TCGACA +TCGACT +TCGACC +TCGACG +TCGAGA +TCGAGT +TCGAGC +TCGAGG +TCGTAA +TCGTAT +TCGTAC +TCGTAG +TCGTTA +TCGTTT +TCGTTC +TCGTTG +TCGTCA +TCGTCT +TCGTCC +TCGTCG +TCGTGA +TCGTGT +TCGTGC +TCGTGG +TCGCAA +TCGCAT +TCGCAC +TCGCAG +TCGCTA +TCGCTT +TCGCTC +TCGCTG +TCGCCA +TCGCCT +TCGCCC +TCGCCG +TCGCGA +TCGCGT +TCGCGC +TCGCGG +TCGGAA +TCGGAT +TCGGAC +TCGGAG +TCGGTA +TCGGTT +TCGGTC +TCGGTG +TCGGCA +TCGGCT +TCGGCC +TCGGCG +TCGGGA +TCGGGT +TCGGGC +TCGGGG +TGAAAA +TGAAAT +TGAAAC +TGAAAG +TGAATA +TGAATT +TGAATC +TGAATG +TGAACA +TGAACT +TGAACC +TGAACG +TGAAGA +TGAAGT +TGAAGC +TGAAGG +TGATAA +TGATAT +TGATAC +TGATAG +TGATTA +TGATTT +TGATTC +TGATTG +TGATCA +TGATCT +TGATCC +TGATCG +TGATGA +TGATGT +TGATGC +TGATGG +TGACAA +TGACAT +TGACAC +TGACAG +TGACTA +TGACTT +TGACTC +TGACTG +TGACCA +TGACCT +TGACCC +TGACCG +TGACGA +TGACGT +TGACGC +TGACGG +TGAGAA +TGAGAT +TGAGAC +TGAGAG +TGAGTA +TGAGTT +TGAGTC +TGAGTG +TGAGCA +TGAGCT +TGAGCC +TGAGCG +TGAGGA +TGAGGT +TGAGGC +TGAGGG +TGTAAA +TGTAAT +TGTAAC +TGTAAG +TGTATA +TGTATT +TGTATC +TGTATG +TGTACA +TGTACT +TGTACC +TGTACG +TGTAGA +TGTAGT +TGTAGC +TGTAGG +TGTTAA +TGTTAT +TGTTAC +TGTTAG +TGTTTA +TGTTTT +TGTTTC +TGTTTG +TGTTCA +TGTTCT +TGTTCC +TGTTCG +TGTTGA +TGTTGT +TGTTGC +TGTTGG +TGTCAA +TGTCAT +TGTCAC +TGTCAG +TGTCTA +TGTCTT +TGTCTC +TGTCTG +TGTCCA +TGTCCT +TGTCCC +TGTCCG +TGTCGA +TGTCGT +TGTCGC +TGTCGG +TGTGAA +TGTGAT +TGTGAC +TGTGAG +TGTGTA +TGTGTT +TGTGTC +TGTGTG +TGTGCA +TGTGCT +TGTGCC +TGTGCG +TGTGGA +TGTGGT +TGTGGC +TGTGGG +TGCAAA +TGCAAT +TGCAAC +TGCAAG +TGCATA +TGCATT +TGCATC +TGCATG +TGCACA +TGCACT +TGCACC +TGCACG +TGCAGA +TGCAGT +TGCAGC +TGCAGG +TGCTAA +TGCTAT +TGCTAC +TGCTAG +TGCTTA +TGCTTT +TGCTTC +TGCTTG +TGCTCA +TGCTCT +TGCTCC +TGCTCG +TGCTGA +TGCTGT +TGCTGC +TGCTGG +TGCCAA +TGCCAT +TGCCAC +TGCCAG +TGCCTA +TGCCTT +TGCCTC +TGCCTG +TGCCCA +TGCCCT +TGCCCC +TGCCCG +TGCCGA +TGCCGT +TGCCGC +TGCCGG +TGCGAA +TGCGAT +TGCGAC +TGCGAG +TGCGTA +TGCGTT +TGCGTC +TGCGTG +TGCGCA +TGCGCT +TGCGCC +TGCGCG +TGCGGA +TGCGGT +TGCGGC +TGCGGG +TGGAAA +TGGAAT +TGGAAC +TGGAAG +TGGATA +TGGATT +TGGATC +TGGATG +TGGACA +TGGACT +TGGACC +TGGACG +TGGAGA +TGGAGT +TGGAGC +TGGAGG +TGGTAA +TGGTAT +TGGTAC +TGGTAG +TGGTTA +TGGTTT +TGGTTC +TGGTTG +TGGTCA +TGGTCT +TGGTCC +TGGTCG +TGGTGA +TGGTGT +TGGTGC +TGGTGG +TGGCAA +TGGCAT +TGGCAC +TGGCAG +TGGCTA +TGGCTT +TGGCTC +TGGCTG +TGGCCA +TGGCCT +TGGCCC +TGGCCG +TGGCGA +TGGCGT +TGGCGC +TGGCGG +TGGGAA +TGGGAT +TGGGAC +TGGGAG +TGGGTA +TGGGTT +TGGGTC +TGGGTG +TGGGCA +TGGGCT +TGGGCC +TGGGCG +TGGGGA +TGGGGT +TGGGGC +TGGGGG +CAAAAA +CAAAAT +CAAAAC +CAAAAG +CAAATA +CAAATT +CAAATC +CAAATG +CAAACA +CAAACT +CAAACC +CAAACG +CAAAGA +CAAAGT +CAAAGC +CAAAGG +CAATAA +CAATAT +CAATAC +CAATAG +CAATTA +CAATTT +CAATTC +CAATTG +CAATCA +CAATCT +CAATCC +CAATCG +CAATGA +CAATGT +CAATGC +CAATGG +CAACAA +CAACAT +CAACAC +CAACAG +CAACTA +CAACTT +CAACTC +CAACTG +CAACCA +CAACCT +CAACCC +CAACCG +CAACGA +CAACGT +CAACGC +CAACGG +CAAGAA +CAAGAT +CAAGAC +CAAGAG +CAAGTA +CAAGTT +CAAGTC +CAAGTG +CAAGCA +CAAGCT +CAAGCC +CAAGCG +CAAGGA +CAAGGT +CAAGGC +CAAGGG +CATAAA +CATAAT +CATAAC +CATAAG +CATATA +CATATT +CATATC +CATATG +CATACA +CATACT +CATACC +CATACG +CATAGA +CATAGT +CATAGC +CATAGG +CATTAA +CATTAT +CATTAC +CATTAG +CATTTA +CATTTT +CATTTC +CATTTG +CATTCA +CATTCT +CATTCC +CATTCG +CATTGA +CATTGT +CATTGC +CATTGG +CATCAA +CATCAT +CATCAC +CATCAG +CATCTA +CATCTT +CATCTC +CATCTG +CATCCA +CATCCT +CATCCC +CATCCG +CATCGA +CATCGT +CATCGC +CATCGG +CATGAA +CATGAT +CATGAC +CATGAG +CATGTA +CATGTT +CATGTC +CATGTG +CATGCA +CATGCT +CATGCC +CATGCG +CATGGA +CATGGT +CATGGC +CATGGG +CACAAA +CACAAT +CACAAC +CACAAG +CACATA +CACATT +CACATC +CACATG +CACACA +CACACT +CACACC +CACACG +CACAGA +CACAGT +CACAGC +CACAGG +CACTAA +CACTAT +CACTAC +CACTAG +CACTTA +CACTTT +CACTTC +CACTTG +CACTCA +CACTCT +CACTCC +CACTCG +CACTGA +CACTGT +CACTGC +CACTGG +CACCAA +CACCAT +CACCAC +CACCAG +CACCTA +CACCTT +CACCTC +CACCTG +CACCCA +CACCCT +CACCCC +CACCCG +CACCGA +CACCGT +CACCGC +CACCGG +CACGAA +CACGAT +CACGAC +CACGAG +CACGTA +CACGTT +CACGTC +CACGTG +CACGCA +CACGCT +CACGCC +CACGCG +CACGGA +CACGGT +CACGGC +CACGGG +CAGAAA +CAGAAT +CAGAAC +CAGAAG +CAGATA +CAGATT +CAGATC +CAGATG +CAGACA +CAGACT +CAGACC +CAGACG +CAGAGA +CAGAGT +CAGAGC +CAGAGG +CAGTAA +CAGTAT +CAGTAC +CAGTAG +CAGTTA +CAGTTT +CAGTTC +CAGTTG +CAGTCA +CAGTCT +CAGTCC +CAGTCG +CAGTGA +CAGTGT +CAGTGC +CAGTGG +CAGCAA +CAGCAT +CAGCAC +CAGCAG +CAGCTA +CAGCTT +CAGCTC +CAGCTG +CAGCCA +CAGCCT +CAGCCC +CAGCCG +CAGCGA +CAGCGT +CAGCGC +CAGCGG +CAGGAA +CAGGAT +CAGGAC +CAGGAG +CAGGTA +CAGGTT +CAGGTC +CAGGTG +CAGGCA +CAGGCT +CAGGCC +CAGGCG +CAGGGA +CAGGGT +CAGGGC +CAGGGG +CTAAAA +CTAAAT +CTAAAC +CTAAAG +CTAATA +CTAATT +CTAATC +CTAATG +CTAACA +CTAACT +CTAACC +CTAACG +CTAAGA +CTAAGT +CTAAGC +CTAAGG +CTATAA +CTATAT +CTATAC +CTATAG +CTATTA +CTATTT +CTATTC +CTATTG +CTATCA +CTATCT +CTATCC +CTATCG +CTATGA +CTATGT +CTATGC +CTATGG +CTACAA +CTACAT +CTACAC +CTACAG +CTACTA +CTACTT +CTACTC +CTACTG +CTACCA +CTACCT +CTACCC +CTACCG +CTACGA +CTACGT +CTACGC +CTACGG +CTAGAA +CTAGAT +CTAGAC +CTAGAG +CTAGTA +CTAGTT +CTAGTC +CTAGTG +CTAGCA +CTAGCT +CTAGCC +CTAGCG +CTAGGA +CTAGGT +CTAGGC +CTAGGG +CTTAAA +CTTAAT +CTTAAC +CTTAAG +CTTATA +CTTATT +CTTATC +CTTATG +CTTACA +CTTACT +CTTACC +CTTACG +CTTAGA +CTTAGT +CTTAGC +CTTAGG +CTTTAA +CTTTAT +CTTTAC +CTTTAG +CTTTTA +CTTTTT +CTTTTC +CTTTTG +CTTTCA +CTTTCT +CTTTCC +CTTTCG +CTTTGA +CTTTGT +CTTTGC +CTTTGG +CTTCAA +CTTCAT +CTTCAC +CTTCAG +CTTCTA +CTTCTT +CTTCTC +CTTCTG +CTTCCA +CTTCCT +CTTCCC +CTTCCG +CTTCGA +CTTCGT +CTTCGC +CTTCGG +CTTGAA +CTTGAT +CTTGAC +CTTGAG +CTTGTA +CTTGTT +CTTGTC +CTTGTG +CTTGCA +CTTGCT +CTTGCC +CTTGCG +CTTGGA +CTTGGT +CTTGGC +CTTGGG +CTCAAA +CTCAAT +CTCAAC +CTCAAG +CTCATA +CTCATT +CTCATC +CTCATG +CTCACA +CTCACT +CTCACC +CTCACG +CTCAGA +CTCAGT +CTCAGC +CTCAGG +CTCTAA +CTCTAT +CTCTAC +CTCTAG +CTCTTA +CTCTTT +CTCTTC +CTCTTG +CTCTCA +CTCTCT +CTCTCC +CTCTCG +CTCTGA +CTCTGT +CTCTGC +CTCTGG +CTCCAA +CTCCAT +CTCCAC +CTCCAG +CTCCTA +CTCCTT +CTCCTC +CTCCTG +CTCCCA +CTCCCT +CTCCCC +CTCCCG +CTCCGA +CTCCGT +CTCCGC +CTCCGG +CTCGAA +CTCGAT +CTCGAC +CTCGAG +CTCGTA +CTCGTT +CTCGTC +CTCGTG +CTCGCA +CTCGCT +CTCGCC +CTCGCG +CTCGGA +CTCGGT +CTCGGC +CTCGGG +CTGAAA +CTGAAT +CTGAAC +CTGAAG +CTGATA +CTGATT +CTGATC +CTGATG +CTGACA +CTGACT +CTGACC +CTGACG +CTGAGA +CTGAGT +CTGAGC +CTGAGG +CTGTAA +CTGTAT +CTGTAC +CTGTAG +CTGTTA +CTGTTT +CTGTTC +CTGTTG +CTGTCA +CTGTCT +CTGTCC +CTGTCG +CTGTGA +CTGTGT +CTGTGC +CTGTGG +CTGCAA +CTGCAT +CTGCAC +CTGCAG +CTGCTA +CTGCTT +CTGCTC +CTGCTG +CTGCCA +CTGCCT +CTGCCC +CTGCCG +CTGCGA +CTGCGT +CTGCGC +CTGCGG +CTGGAA +CTGGAT +CTGGAC +CTGGAG +CTGGTA +CTGGTT +CTGGTC +CTGGTG +CTGGCA +CTGGCT +CTGGCC +CTGGCG +CTGGGA +CTGGGT +CTGGGC +CTGGGG +CCAAAA +CCAAAT +CCAAAC +CCAAAG +CCAATA +CCAATT +CCAATC +CCAATG +CCAACA +CCAACT +CCAACC +CCAACG +CCAAGA +CCAAGT +CCAAGC +CCAAGG +CCATAA +CCATAT +CCATAC +CCATAG +CCATTA +CCATTT +CCATTC +CCATTG +CCATCA +CCATCT +CCATCC +CCATCG +CCATGA +CCATGT +CCATGC +CCATGG +CCACAA +CCACAT +CCACAC +CCACAG +CCACTA +CCACTT +CCACTC +CCACTG +CCACCA +CCACCT +CCACCC +CCACCG +CCACGA +CCACGT +CCACGC +CCACGG +CCAGAA +CCAGAT +CCAGAC +CCAGAG +CCAGTA +CCAGTT +CCAGTC +CCAGTG +CCAGCA +CCAGCT +CCAGCC +CCAGCG +CCAGGA +CCAGGT +CCAGGC +CCAGGG +CCTAAA +CCTAAT +CCTAAC +CCTAAG +CCTATA +CCTATT +CCTATC +CCTATG +CCTACA +CCTACT +CCTACC +CCTACG +CCTAGA +CCTAGT +CCTAGC +CCTAGG +CCTTAA +CCTTAT +CCTTAC +CCTTAG +CCTTTA +CCTTTT +CCTTTC +CCTTTG +CCTTCA +CCTTCT +CCTTCC +CCTTCG +CCTTGA +CCTTGT +CCTTGC +CCTTGG +CCTCAA +CCTCAT +CCTCAC +CCTCAG +CCTCTA +CCTCTT +CCTCTC +CCTCTG +CCTCCA +CCTCCT +CCTCCC +CCTCCG +CCTCGA +CCTCGT +CCTCGC +CCTCGG +CCTGAA +CCTGAT +CCTGAC +CCTGAG +CCTGTA +CCTGTT +CCTGTC +CCTGTG +CCTGCA +CCTGCT +CCTGCC +CCTGCG +CCTGGA +CCTGGT +CCTGGC +CCTGGG +CCCAAA +CCCAAT +CCCAAC +CCCAAG +CCCATA +CCCATT +CCCATC +CCCATG +CCCACA +CCCACT +CCCACC +CCCACG +CCCAGA +CCCAGT +CCCAGC +CCCAGG +CCCTAA +CCCTAT +CCCTAC +CCCTAG +CCCTTA +CCCTTT +CCCTTC +CCCTTG +CCCTCA +CCCTCT +CCCTCC +CCCTCG +CCCTGA +CCCTGT +CCCTGC +CCCTGG +CCCCAA +CCCCAT +CCCCAC +CCCCAG +CCCCTA +CCCCTT +CCCCTC +CCCCTG +CCCCCA +CCCCCT +CCCCCC +CCCCCG +CCCCGA +CCCCGT +CCCCGC +CCCCGG +CCCGAA +CCCGAT +CCCGAC +CCCGAG +CCCGTA +CCCGTT +CCCGTC +CCCGTG +CCCGCA +CCCGCT +CCCGCC +CCCGCG +CCCGGA +CCCGGT +CCCGGC +CCCGGG +CCGAAA +CCGAAT +CCGAAC +CCGAAG +CCGATA +CCGATT +CCGATC +CCGATG +CCGACA +CCGACT +CCGACC +CCGACG +CCGAGA +CCGAGT +CCGAGC +CCGAGG +CCGTAA +CCGTAT +CCGTAC +CCGTAG +CCGTTA +CCGTTT +CCGTTC +CCGTTG +CCGTCA +CCGTCT +CCGTCC +CCGTCG +CCGTGA +CCGTGT +CCGTGC +CCGTGG +CCGCAA +CCGCAT +CCGCAC +CCGCAG +CCGCTA +CCGCTT +CCGCTC +CCGCTG +CCGCCA +CCGCCT +CCGCCC +CCGCCG +CCGCGA +CCGCGT +CCGCGC +CCGCGG +CCGGAA +CCGGAT +CCGGAC +CCGGAG +CCGGTA +CCGGTT +CCGGTC +CCGGTG +CCGGCA +CCGGCT +CCGGCC +CCGGCG +CCGGGA +CCGGGT +CCGGGC +CCGGGG +CGAAAA +CGAAAT +CGAAAC +CGAAAG +CGAATA +CGAATT +CGAATC +CGAATG +CGAACA +CGAACT +CGAACC +CGAACG +CGAAGA +CGAAGT +CGAAGC +CGAAGG +CGATAA +CGATAT +CGATAC +CGATAG +CGATTA +CGATTT +CGATTC +CGATTG +CGATCA +CGATCT +CGATCC +CGATCG +CGATGA +CGATGT +CGATGC +CGATGG +CGACAA +CGACAT +CGACAC +CGACAG +CGACTA +CGACTT +CGACTC +CGACTG +CGACCA +CGACCT +CGACCC +CGACCG +CGACGA +CGACGT +CGACGC +CGACGG +CGAGAA +CGAGAT +CGAGAC +CGAGAG +CGAGTA +CGAGTT +CGAGTC +CGAGTG +CGAGCA +CGAGCT +CGAGCC +CGAGCG +CGAGGA +CGAGGT +CGAGGC +CGAGGG +CGTAAA +CGTAAT +CGTAAC +CGTAAG +CGTATA +CGTATT +CGTATC +CGTATG +CGTACA +CGTACT +CGTACC +CGTACG +CGTAGA +CGTAGT +CGTAGC +CGTAGG +CGTTAA +CGTTAT +CGTTAC +CGTTAG +CGTTTA +CGTTTT +CGTTTC +CGTTTG +CGTTCA +CGTTCT +CGTTCC +CGTTCG +CGTTGA +CGTTGT +CGTTGC +CGTTGG +CGTCAA +CGTCAT +CGTCAC +CGTCAG +CGTCTA +CGTCTT +CGTCTC +CGTCTG +CGTCCA +CGTCCT +CGTCCC +CGTCCG +CGTCGA +CGTCGT +CGTCGC +CGTCGG +CGTGAA +CGTGAT +CGTGAC +CGTGAG +CGTGTA +CGTGTT +CGTGTC +CGTGTG +CGTGCA +CGTGCT +CGTGCC +CGTGCG +CGTGGA +CGTGGT +CGTGGC +CGTGGG +CGCAAA +CGCAAT +CGCAAC +CGCAAG +CGCATA +CGCATT +CGCATC +CGCATG +CGCACA +CGCACT +CGCACC +CGCACG +CGCAGA +CGCAGT +CGCAGC +CGCAGG +CGCTAA +CGCTAT +CGCTAC +CGCTAG +CGCTTA +CGCTTT +CGCTTC +CGCTTG +CGCTCA +CGCTCT +CGCTCC +CGCTCG +CGCTGA +CGCTGT +CGCTGC +CGCTGG +CGCCAA +CGCCAT +CGCCAC +CGCCAG +CGCCTA +CGCCTT +CGCCTC +CGCCTG +CGCCCA +CGCCCT +CGCCCC +CGCCCG +CGCCGA +CGCCGT +CGCCGC +CGCCGG +CGCGAA +CGCGAT +CGCGAC +CGCGAG +CGCGTA +CGCGTT +CGCGTC +CGCGTG +CGCGCA +CGCGCT +CGCGCC +CGCGCG +CGCGGA +CGCGGT +CGCGGC +CGCGGG +CGGAAA +CGGAAT +CGGAAC +CGGAAG +CGGATA +CGGATT +CGGATC +CGGATG +CGGACA +CGGACT +CGGACC +CGGACG +CGGAGA +CGGAGT +CGGAGC +CGGAGG +CGGTAA +CGGTAT +CGGTAC +CGGTAG +CGGTTA +CGGTTT +CGGTTC +CGGTTG +CGGTCA +CGGTCT +CGGTCC +CGGTCG +CGGTGA +CGGTGT +CGGTGC +CGGTGG +CGGCAA +CGGCAT +CGGCAC +CGGCAG +CGGCTA +CGGCTT +CGGCTC +CGGCTG +CGGCCA +CGGCCT +CGGCCC +CGGCCG +CGGCGA +CGGCGT +CGGCGC +CGGCGG +CGGGAA +CGGGAT +CGGGAC +CGGGAG +CGGGTA +CGGGTT +CGGGTC +CGGGTG +CGGGCA +CGGGCT +CGGGCC +CGGGCG +CGGGGA +CGGGGT +CGGGGC +CGGGGG +GAAAAA +GAAAAT +GAAAAC +GAAAAG +GAAATA +GAAATT +GAAATC +GAAATG +GAAACA +GAAACT +GAAACC +GAAACG +GAAAGA +GAAAGT +GAAAGC +GAAAGG +GAATAA +GAATAT +GAATAC +GAATAG +GAATTA +GAATTT +GAATTC +GAATTG +GAATCA +GAATCT +GAATCC +GAATCG +GAATGA +GAATGT +GAATGC +GAATGG +GAACAA +GAACAT +GAACAC +GAACAG +GAACTA +GAACTT +GAACTC +GAACTG +GAACCA +GAACCT +GAACCC +GAACCG +GAACGA +GAACGT +GAACGC +GAACGG +GAAGAA +GAAGAT +GAAGAC +GAAGAG +GAAGTA +GAAGTT +GAAGTC +GAAGTG +GAAGCA +GAAGCT +GAAGCC +GAAGCG +GAAGGA +GAAGGT +GAAGGC +GAAGGG +GATAAA +GATAAT +GATAAC +GATAAG +GATATA +GATATT +GATATC +GATATG +GATACA +GATACT +GATACC +GATACG +GATAGA +GATAGT +GATAGC +GATAGG +GATTAA +GATTAT +GATTAC +GATTAG +GATTTA +GATTTT +GATTTC +GATTTG +GATTCA +GATTCT +GATTCC +GATTCG +GATTGA +GATTGT +GATTGC +GATTGG +GATCAA +GATCAT +GATCAC +GATCAG +GATCTA +GATCTT +GATCTC +GATCTG +GATCCA +GATCCT +GATCCC +GATCCG +GATCGA +GATCGT +GATCGC +GATCGG +GATGAA +GATGAT +GATGAC +GATGAG +GATGTA +GATGTT +GATGTC +GATGTG +GATGCA +GATGCT +GATGCC +GATGCG +GATGGA +GATGGT +GATGGC +GATGGG +GACAAA +GACAAT +GACAAC +GACAAG +GACATA +GACATT +GACATC +GACATG +GACACA +GACACT +GACACC +GACACG +GACAGA +GACAGT +GACAGC +GACAGG +GACTAA +GACTAT +GACTAC +GACTAG +GACTTA +GACTTT +GACTTC +GACTTG +GACTCA +GACTCT +GACTCC +GACTCG +GACTGA +GACTGT +GACTGC +GACTGG +GACCAA +GACCAT +GACCAC +GACCAG +GACCTA +GACCTT +GACCTC +GACCTG +GACCCA +GACCCT +GACCCC +GACCCG +GACCGA +GACCGT +GACCGC +GACCGG +GACGAA +GACGAT +GACGAC +GACGAG +GACGTA +GACGTT +GACGTC +GACGTG +GACGCA +GACGCT +GACGCC +GACGCG +GACGGA +GACGGT +GACGGC +GACGGG +GAGAAA +GAGAAT +GAGAAC +GAGAAG +GAGATA +GAGATT +GAGATC +GAGATG +GAGACA +GAGACT +GAGACC +GAGACG +GAGAGA +GAGAGT +GAGAGC +GAGAGG +GAGTAA +GAGTAT +GAGTAC +GAGTAG +GAGTTA +GAGTTT +GAGTTC +GAGTTG +GAGTCA +GAGTCT +GAGTCC +GAGTCG +GAGTGA +GAGTGT +GAGTGC +GAGTGG +GAGCAA +GAGCAT +GAGCAC +GAGCAG +GAGCTA +GAGCTT +GAGCTC +GAGCTG +GAGCCA +GAGCCT +GAGCCC +GAGCCG +GAGCGA +GAGCGT +GAGCGC +GAGCGG +GAGGAA +GAGGAT +GAGGAC +GAGGAG +GAGGTA +GAGGTT +GAGGTC +GAGGTG +GAGGCA +GAGGCT +GAGGCC +GAGGCG +GAGGGA +GAGGGT +GAGGGC +GAGGGG +GTAAAA +GTAAAT +GTAAAC +GTAAAG +GTAATA +GTAATT +GTAATC +GTAATG +GTAACA +GTAACT +GTAACC +GTAACG +GTAAGA +GTAAGT +GTAAGC +GTAAGG +GTATAA +GTATAT +GTATAC +GTATAG +GTATTA +GTATTT +GTATTC +GTATTG +GTATCA +GTATCT +GTATCC +GTATCG +GTATGA +GTATGT +GTATGC +GTATGG +GTACAA +GTACAT +GTACAC +GTACAG +GTACTA +GTACTT +GTACTC +GTACTG +GTACCA +GTACCT +GTACCC +GTACCG +GTACGA +GTACGT +GTACGC +GTACGG +GTAGAA +GTAGAT +GTAGAC +GTAGAG +GTAGTA +GTAGTT +GTAGTC +GTAGTG +GTAGCA +GTAGCT +GTAGCC +GTAGCG +GTAGGA +GTAGGT +GTAGGC +GTAGGG +GTTAAA +GTTAAT +GTTAAC +GTTAAG +GTTATA +GTTATT +GTTATC +GTTATG +GTTACA +GTTACT +GTTACC +GTTACG +GTTAGA +GTTAGT +GTTAGC +GTTAGG +GTTTAA +GTTTAT +GTTTAC +GTTTAG +GTTTTA +GTTTTT +GTTTTC +GTTTTG +GTTTCA +GTTTCT +GTTTCC +GTTTCG +GTTTGA +GTTTGT +GTTTGC +GTTTGG +GTTCAA +GTTCAT +GTTCAC +GTTCAG +GTTCTA +GTTCTT +GTTCTC +GTTCTG +GTTCCA +GTTCCT +GTTCCC +GTTCCG +GTTCGA +GTTCGT +GTTCGC +GTTCGG +GTTGAA +GTTGAT +GTTGAC +GTTGAG +GTTGTA +GTTGTT +GTTGTC +GTTGTG +GTTGCA +GTTGCT +GTTGCC +GTTGCG +GTTGGA +GTTGGT +GTTGGC +GTTGGG +GTCAAA +GTCAAT +GTCAAC +GTCAAG +GTCATA +GTCATT +GTCATC +GTCATG +GTCACA +GTCACT +GTCACC +GTCACG +GTCAGA +GTCAGT +GTCAGC +GTCAGG +GTCTAA +GTCTAT +GTCTAC +GTCTAG +GTCTTA +GTCTTT +GTCTTC +GTCTTG +GTCTCA +GTCTCT +GTCTCC +GTCTCG +GTCTGA +GTCTGT +GTCTGC +GTCTGG +GTCCAA +GTCCAT +GTCCAC +GTCCAG +GTCCTA +GTCCTT +GTCCTC +GTCCTG +GTCCCA +GTCCCT +GTCCCC +GTCCCG +GTCCGA +GTCCGT +GTCCGC +GTCCGG +GTCGAA +GTCGAT +GTCGAC +GTCGAG +GTCGTA +GTCGTT +GTCGTC +GTCGTG +GTCGCA +GTCGCT +GTCGCC +GTCGCG +GTCGGA +GTCGGT +GTCGGC +GTCGGG +GTGAAA +GTGAAT +GTGAAC +GTGAAG +GTGATA +GTGATT +GTGATC +GTGATG +GTGACA +GTGACT +GTGACC +GTGACG +GTGAGA +GTGAGT +GTGAGC +GTGAGG +GTGTAA +GTGTAT +GTGTAC +GTGTAG +GTGTTA +GTGTTT +GTGTTC +GTGTTG +GTGTCA +GTGTCT +GTGTCC +GTGTCG +GTGTGA +GTGTGT +GTGTGC +GTGTGG +GTGCAA +GTGCAT +GTGCAC +GTGCAG +GTGCTA +GTGCTT +GTGCTC +GTGCTG +GTGCCA +GTGCCT +GTGCCC +GTGCCG +GTGCGA +GTGCGT +GTGCGC +GTGCGG +GTGGAA +GTGGAT +GTGGAC +GTGGAG +GTGGTA +GTGGTT +GTGGTC +GTGGTG +GTGGCA +GTGGCT +GTGGCC +GTGGCG +GTGGGA +GTGGGT +GTGGGC +GTGGGG +GCAAAA +GCAAAT +GCAAAC +GCAAAG +GCAATA +GCAATT +GCAATC +GCAATG +GCAACA +GCAACT +GCAACC +GCAACG +GCAAGA +GCAAGT +GCAAGC +GCAAGG +GCATAA +GCATAT +GCATAC +GCATAG +GCATTA +GCATTT +GCATTC +GCATTG +GCATCA +GCATCT +GCATCC +GCATCG +GCATGA +GCATGT +GCATGC +GCATGG +GCACAA +GCACAT +GCACAC +GCACAG +GCACTA +GCACTT +GCACTC +GCACTG +GCACCA +GCACCT +GCACCC +GCACCG +GCACGA +GCACGT +GCACGC +GCACGG +GCAGAA +GCAGAT +GCAGAC +GCAGAG +GCAGTA +GCAGTT +GCAGTC +GCAGTG +GCAGCA +GCAGCT +GCAGCC +GCAGCG +GCAGGA +GCAGGT +GCAGGC +GCAGGG +GCTAAA +GCTAAT +GCTAAC +GCTAAG +GCTATA +GCTATT +GCTATC +GCTATG +GCTACA +GCTACT +GCTACC +GCTACG +GCTAGA +GCTAGT +GCTAGC +GCTAGG +GCTTAA +GCTTAT +GCTTAC +GCTTAG +GCTTTA +GCTTTT +GCTTTC +GCTTTG +GCTTCA +GCTTCT +GCTTCC +GCTTCG +GCTTGA +GCTTGT +GCTTGC +GCTTGG +GCTCAA +GCTCAT +GCTCAC +GCTCAG +GCTCTA +GCTCTT +GCTCTC +GCTCTG +GCTCCA +GCTCCT +GCTCCC +GCTCCG +GCTCGA +GCTCGT +GCTCGC +GCTCGG +GCTGAA +GCTGAT +GCTGAC +GCTGAG +GCTGTA +GCTGTT +GCTGTC +GCTGTG +GCTGCA +GCTGCT +GCTGCC +GCTGCG +GCTGGA +GCTGGT +GCTGGC +GCTGGG +GCCAAA +GCCAAT +GCCAAC +GCCAAG +GCCATA +GCCATT +GCCATC +GCCATG +GCCACA +GCCACT +GCCACC +GCCACG +GCCAGA +GCCAGT +GCCAGC +GCCAGG +GCCTAA +GCCTAT +GCCTAC +GCCTAG +GCCTTA +GCCTTT +GCCTTC +GCCTTG +GCCTCA +GCCTCT +GCCTCC +GCCTCG +GCCTGA +GCCTGT +GCCTGC +GCCTGG +GCCCAA +GCCCAT +GCCCAC +GCCCAG +GCCCTA +GCCCTT +GCCCTC +GCCCTG +GCCCCA +GCCCCT +GCCCCC +GCCCCG +GCCCGA +GCCCGT +GCCCGC +GCCCGG +GCCGAA +GCCGAT +GCCGAC +GCCGAG +GCCGTA +GCCGTT +GCCGTC +GCCGTG +GCCGCA +GCCGCT +GCCGCC +GCCGCG +GCCGGA +GCCGGT +GCCGGC +GCCGGG +GCGAAA +GCGAAT +GCGAAC +GCGAAG +GCGATA +GCGATT +GCGATC +GCGATG +GCGACA +GCGACT +GCGACC +GCGACG +GCGAGA +GCGAGT +GCGAGC +GCGAGG +GCGTAA +GCGTAT +GCGTAC +GCGTAG +GCGTTA +GCGTTT +GCGTTC +GCGTTG +GCGTCA +GCGTCT +GCGTCC +GCGTCG +GCGTGA +GCGTGT +GCGTGC +GCGTGG +GCGCAA +GCGCAT +GCGCAC +GCGCAG +GCGCTA +GCGCTT +GCGCTC +GCGCTG +GCGCCA +GCGCCT +GCGCCC +GCGCCG +GCGCGA +GCGCGT +GCGCGC +GCGCGG +GCGGAA +GCGGAT +GCGGAC +GCGGAG +GCGGTA +GCGGTT +GCGGTC +GCGGTG +GCGGCA +GCGGCT +GCGGCC +GCGGCG +GCGGGA +GCGGGT +GCGGGC +GCGGGG +GGAAAA +GGAAAT +GGAAAC +GGAAAG +GGAATA +GGAATT +GGAATC +GGAATG +GGAACA +GGAACT +GGAACC +GGAACG +GGAAGA +GGAAGT +GGAAGC +GGAAGG +GGATAA +GGATAT +GGATAC +GGATAG +GGATTA +GGATTT +GGATTC +GGATTG +GGATCA +GGATCT +GGATCC +GGATCG +GGATGA +GGATGT +GGATGC +GGATGG +GGACAA +GGACAT +GGACAC +GGACAG +GGACTA +GGACTT +GGACTC +GGACTG +GGACCA +GGACCT +GGACCC +GGACCG +GGACGA +GGACGT +GGACGC +GGACGG +GGAGAA +GGAGAT +GGAGAC +GGAGAG +GGAGTA +GGAGTT +GGAGTC +GGAGTG +GGAGCA +GGAGCT +GGAGCC +GGAGCG +GGAGGA +GGAGGT +GGAGGC +GGAGGG +GGTAAA +GGTAAT +GGTAAC +GGTAAG +GGTATA +GGTATT +GGTATC +GGTATG +GGTACA +GGTACT +GGTACC +GGTACG +GGTAGA +GGTAGT +GGTAGC +GGTAGG +GGTTAA +GGTTAT +GGTTAC +GGTTAG +GGTTTA +GGTTTT +GGTTTC +GGTTTG +GGTTCA +GGTTCT +GGTTCC +GGTTCG +GGTTGA +GGTTGT +GGTTGC +GGTTGG +GGTCAA +GGTCAT +GGTCAC +GGTCAG +GGTCTA +GGTCTT +GGTCTC +GGTCTG +GGTCCA +GGTCCT +GGTCCC +GGTCCG +GGTCGA +GGTCGT +GGTCGC +GGTCGG +GGTGAA +GGTGAT +GGTGAC +GGTGAG +GGTGTA +GGTGTT +GGTGTC +GGTGTG +GGTGCA +GGTGCT +GGTGCC +GGTGCG +GGTGGA +GGTGGT +GGTGGC +GGTGGG +GGCAAA +GGCAAT +GGCAAC +GGCAAG +GGCATA +GGCATT +GGCATC +GGCATG +GGCACA +GGCACT +GGCACC +GGCACG +GGCAGA +GGCAGT +GGCAGC +GGCAGG +GGCTAA +GGCTAT +GGCTAC +GGCTAG +GGCTTA +GGCTTT +GGCTTC +GGCTTG +GGCTCA +GGCTCT +GGCTCC +GGCTCG +GGCTGA +GGCTGT +GGCTGC +GGCTGG +GGCCAA +GGCCAT +GGCCAC +GGCCAG +GGCCTA +GGCCTT +GGCCTC +GGCCTG +GGCCCA +GGCCCT +GGCCCC +GGCCCG +GGCCGA +GGCCGT +GGCCGC +GGCCGG +GGCGAA +GGCGAT +GGCGAC +GGCGAG +GGCGTA +GGCGTT +GGCGTC +GGCGTG +GGCGCA +GGCGCT +GGCGCC +GGCGCG +GGCGGA +GGCGGT +GGCGGC +GGCGGG +GGGAAA +GGGAAT +GGGAAC +GGGAAG +GGGATA +GGGATT +GGGATC +GGGATG +GGGACA +GGGACT +GGGACC +GGGACG +GGGAGA +GGGAGT +GGGAGC +GGGAGG +GGGTAA +GGGTAT +GGGTAC +GGGTAG +GGGTTA +GGGTTT +GGGTTC +GGGTTG +GGGTCA +GGGTCT +GGGTCC +GGGTCG +GGGTGA +GGGTGT +GGGTGC +GGGTGG +GGGCAA +GGGCAT +GGGCAC +GGGCAG +GGGCTA +GGGCTT +GGGCTC +GGGCTG +GGGCCA +GGGCCT +GGGCCC +GGGCCG +GGGCGA +GGGCGT +GGGCGC +GGGCGG +GGGGAA +GGGGAT +GGGGAC +GGGGAG +GGGGTA +GGGGTT +GGGGTC +GGGGTG +GGGGCA +GGGGCT +GGGGCC +GGGGCG +GGGGGA +GGGGGT +GGGGGC +GGGGGG diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6000/config.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..9b14e61d327bf586fcff6e817cf9d365bae18ad2 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6000/config.json @@ -0,0 +1,26 @@ +{ + "_name_or_path": "fabihamakhdoomi/TinyDNABERT", + "architectures": [ + "BertForSequenceClassification" + ], + "attention_probs_dropout_prob": 0.1, + "classifier_dropout": null, + "hidden_act": "gelu", + "hidden_dropout_prob": 0.1, + "hidden_size": 128, + "initializer_range": 0.02, + "intermediate_size": 512, + "layer_norm_eps": 1e-12, + "max_position_embeddings": 512, + "model_type": "bert", + "num_attention_heads": 2, + "num_hidden_layers": 2, + "pad_token_id": 0, + "position_embedding_type": "absolute", + "problem_type": "single_label_classification", + "torch_dtype": "float32", + "transformers_version": "4.44.0", + "type_vocab_size": 2, + "use_cache": true, + "vocab_size": 4101 +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6000/model.safetensors b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6000/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..16595ecb967d2024e771350a85ad3acabbb3eeed --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6000/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b682fc12d323c3446344409025dfc6fe2cae56d95bf8d2b9a9a9d8dc14c75068 +size 4021688 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6000/optimizer.pt b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6000/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..1ee4f0e7b1b75c902fba86b95a50d4471fa62ea8 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6000/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:89183cac8826e0d207b08aff3433fe7df9d3a22e9a95abe98aa9879991080e0b +size 8058700 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6000/rng_state.pth b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6000/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..c78d09ac076b437860c4e8c07df963c522b10249 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6000/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d44e448e2a0dd138d067ef3ddf4b7abb608f898bed2f0d0880205ed4bbbe2136 +size 14244 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6000/scheduler.pt b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6000/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..f3ebc22eebe7fac9f20742118527e91760861e6d --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6000/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7e153e480933b4bdcc7d98226c2fdfde72db931809685729c8748251ae3ecae9 +size 1064 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6000/special_tokens_map.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6000/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..a8b3208c2884c4efb86e49300fdd3dc877220cdf --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6000/special_tokens_map.json @@ -0,0 +1,7 @@ +{ + "cls_token": "[CLS]", + "mask_token": "[MASK]", + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6000/tokenizer.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6000/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..ebd782de6963b9032ce9374cfe30f5275be1a30b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6000/tokenizer.json @@ -0,0 +1,4265 @@ +{ + "version": "1.0", + "truncation": { + "direction": "Right", + "max_length": 512, + "strategy": "LongestFirst", + "stride": 0 + }, + "padding": { + "strategy": { + "Fixed": 512 + }, + "direction": "Right", + "pad_to_multiple_of": null, + "pad_id": 0, + "pad_type_id": 0, + "pad_token": "[PAD]" + }, + "added_tokens": [ + { + "id": 0, + "content": "[PAD]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 1, + "content": "[UNK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 2, + "content": "[CLS]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 3, + "content": "[SEP]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 4, + "content": "[MASK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + } + ], + "normalizer": { + "type": "BertNormalizer", + "clean_text": true, + "handle_chinese_chars": true, + "strip_accents": null, + "lowercase": false + }, + "pre_tokenizer": { + "type": "BertPreTokenizer" + }, + "post_processor": { + "type": "TemplateProcessing", + "single": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + } + ], + "pair": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "B", + "type_id": 1 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 1 + } + } + ], + "special_tokens": { + "[CLS]": { + "id": "[CLS]", + "ids": [ + 2 + ], + "tokens": [ + "[CLS]" + ] + }, + "[SEP]": { + "id": "[SEP]", + "ids": [ + 3 + ], + "tokens": [ + "[SEP]" + ] + } + } + }, + "decoder": { + "type": "WordPiece", + "prefix": "##", + "cleanup": true + }, + "model": { + "type": "WordPiece", + "unk_token": "[UNK]", + "continuing_subword_prefix": "##", + "max_input_chars_per_word": 100, + "vocab": { + "[PAD]": 0, + "[UNK]": 1, + "[CLS]": 2, + "[SEP]": 3, + "[MASK]": 4, + "AAAAAA": 5, + "AAAAAT": 6, + "AAAAAC": 7, + "AAAAAG": 8, + "AAAATA": 9, + "AAAATT": 10, + "AAAATC": 11, + "AAAATG": 12, + "AAAACA": 13, + "AAAACT": 14, + "AAAACC": 15, + "AAAACG": 16, + "AAAAGA": 17, + "AAAAGT": 18, + "AAAAGC": 19, + "AAAAGG": 20, + "AAATAA": 21, + "AAATAT": 22, + "AAATAC": 23, + "AAATAG": 24, + "AAATTA": 25, + "AAATTT": 26, + "AAATTC": 27, + "AAATTG": 28, + "AAATCA": 29, + "AAATCT": 30, + "AAATCC": 31, + "AAATCG": 32, + "AAATGA": 33, + "AAATGT": 34, + "AAATGC": 35, + "AAATGG": 36, + "AAACAA": 37, + "AAACAT": 38, + "AAACAC": 39, + "AAACAG": 40, + "AAACTA": 41, + "AAACTT": 42, + "AAACTC": 43, + "AAACTG": 44, + "AAACCA": 45, + "AAACCT": 46, + "AAACCC": 47, + "AAACCG": 48, + "AAACGA": 49, + "AAACGT": 50, + "AAACGC": 51, + "AAACGG": 52, + "AAAGAA": 53, + "AAAGAT": 54, + "AAAGAC": 55, + "AAAGAG": 56, + "AAAGTA": 57, + "AAAGTT": 58, + "AAAGTC": 59, + "AAAGTG": 60, + "AAAGCA": 61, + "AAAGCT": 62, + "AAAGCC": 63, + "AAAGCG": 64, + "AAAGGA": 65, + "AAAGGT": 66, + "AAAGGC": 67, + "AAAGGG": 68, + "AATAAA": 69, + "AATAAT": 70, + "AATAAC": 71, + "AATAAG": 72, + "AATATA": 73, + "AATATT": 74, + "AATATC": 75, + "AATATG": 76, + "AATACA": 77, + "AATACT": 78, + "AATACC": 79, + "AATACG": 80, + "AATAGA": 81, + "AATAGT": 82, + "AATAGC": 83, + "AATAGG": 84, + "AATTAA": 85, + "AATTAT": 86, + "AATTAC": 87, + "AATTAG": 88, + "AATTTA": 89, + "AATTTT": 90, + "AATTTC": 91, + "AATTTG": 92, + "AATTCA": 93, + "AATTCT": 94, + "AATTCC": 95, + "AATTCG": 96, + "AATTGA": 97, + "AATTGT": 98, + "AATTGC": 99, + "AATTGG": 100, + "AATCAA": 101, + "AATCAT": 102, + "AATCAC": 103, + "AATCAG": 104, + "AATCTA": 105, + "AATCTT": 106, + "AATCTC": 107, + "AATCTG": 108, + "AATCCA": 109, + "AATCCT": 110, + "AATCCC": 111, + "AATCCG": 112, + "AATCGA": 113, + "AATCGT": 114, + "AATCGC": 115, + "AATCGG": 116, + "AATGAA": 117, + "AATGAT": 118, + "AATGAC": 119, + "AATGAG": 120, + "AATGTA": 121, + "AATGTT": 122, + "AATGTC": 123, + "AATGTG": 124, + "AATGCA": 125, + "AATGCT": 126, + "AATGCC": 127, + "AATGCG": 128, + "AATGGA": 129, + "AATGGT": 130, + "AATGGC": 131, + "AATGGG": 132, + "AACAAA": 133, + "AACAAT": 134, + "AACAAC": 135, + "AACAAG": 136, + "AACATA": 137, + "AACATT": 138, + "AACATC": 139, + "AACATG": 140, + "AACACA": 141, + "AACACT": 142, + "AACACC": 143, + "AACACG": 144, + "AACAGA": 145, + "AACAGT": 146, + "AACAGC": 147, + "AACAGG": 148, + "AACTAA": 149, + "AACTAT": 150, + "AACTAC": 151, + "AACTAG": 152, + "AACTTA": 153, + "AACTTT": 154, + "AACTTC": 155, + "AACTTG": 156, + "AACTCA": 157, + "AACTCT": 158, + "AACTCC": 159, + "AACTCG": 160, + "AACTGA": 161, + "AACTGT": 162, + "AACTGC": 163, + "AACTGG": 164, + "AACCAA": 165, + "AACCAT": 166, + "AACCAC": 167, + "AACCAG": 168, + "AACCTA": 169, + "AACCTT": 170, + "AACCTC": 171, + "AACCTG": 172, + "AACCCA": 173, + "AACCCT": 174, + "AACCCC": 175, + "AACCCG": 176, + "AACCGA": 177, + "AACCGT": 178, + "AACCGC": 179, + "AACCGG": 180, + "AACGAA": 181, + "AACGAT": 182, + "AACGAC": 183, + "AACGAG": 184, + "AACGTA": 185, + "AACGTT": 186, + "AACGTC": 187, + "AACGTG": 188, + "AACGCA": 189, + "AACGCT": 190, + "AACGCC": 191, + "AACGCG": 192, + "AACGGA": 193, + "AACGGT": 194, + "AACGGC": 195, + "AACGGG": 196, + "AAGAAA": 197, + "AAGAAT": 198, + "AAGAAC": 199, + "AAGAAG": 200, + "AAGATA": 201, + "AAGATT": 202, + "AAGATC": 203, + "AAGATG": 204, + "AAGACA": 205, + "AAGACT": 206, + "AAGACC": 207, + "AAGACG": 208, + "AAGAGA": 209, + "AAGAGT": 210, + "AAGAGC": 211, + "AAGAGG": 212, + "AAGTAA": 213, + "AAGTAT": 214, + "AAGTAC": 215, + "AAGTAG": 216, + "AAGTTA": 217, + "AAGTTT": 218, + "AAGTTC": 219, + "AAGTTG": 220, + "AAGTCA": 221, + "AAGTCT": 222, + "AAGTCC": 223, + "AAGTCG": 224, + "AAGTGA": 225, + "AAGTGT": 226, + "AAGTGC": 227, + "AAGTGG": 228, + "AAGCAA": 229, + "AAGCAT": 230, + "AAGCAC": 231, + "AAGCAG": 232, + "AAGCTA": 233, + "AAGCTT": 234, + "AAGCTC": 235, + "AAGCTG": 236, + "AAGCCA": 237, + "AAGCCT": 238, + "AAGCCC": 239, + "AAGCCG": 240, + "AAGCGA": 241, + "AAGCGT": 242, + "AAGCGC": 243, + "AAGCGG": 244, + "AAGGAA": 245, + "AAGGAT": 246, + "AAGGAC": 247, + "AAGGAG": 248, + "AAGGTA": 249, + "AAGGTT": 250, + "AAGGTC": 251, + "AAGGTG": 252, + "AAGGCA": 253, + "AAGGCT": 254, + "AAGGCC": 255, + "AAGGCG": 256, + "AAGGGA": 257, + "AAGGGT": 258, + "AAGGGC": 259, + "AAGGGG": 260, + "ATAAAA": 261, + "ATAAAT": 262, + "ATAAAC": 263, + "ATAAAG": 264, + "ATAATA": 265, + "ATAATT": 266, + "ATAATC": 267, + "ATAATG": 268, + "ATAACA": 269, + "ATAACT": 270, + "ATAACC": 271, + "ATAACG": 272, + "ATAAGA": 273, + "ATAAGT": 274, + "ATAAGC": 275, + "ATAAGG": 276, + "ATATAA": 277, + "ATATAT": 278, + "ATATAC": 279, + "ATATAG": 280, + "ATATTA": 281, + "ATATTT": 282, + "ATATTC": 283, + "ATATTG": 284, + "ATATCA": 285, + "ATATCT": 286, + "ATATCC": 287, + "ATATCG": 288, + "ATATGA": 289, + "ATATGT": 290, + "ATATGC": 291, + "ATATGG": 292, + "ATACAA": 293, + "ATACAT": 294, + "ATACAC": 295, + "ATACAG": 296, + "ATACTA": 297, + "ATACTT": 298, + "ATACTC": 299, + "ATACTG": 300, + "ATACCA": 301, + "ATACCT": 302, + "ATACCC": 303, + "ATACCG": 304, + "ATACGA": 305, + "ATACGT": 306, + "ATACGC": 307, + "ATACGG": 308, + "ATAGAA": 309, + "ATAGAT": 310, + "ATAGAC": 311, + "ATAGAG": 312, + "ATAGTA": 313, + "ATAGTT": 314, + "ATAGTC": 315, + "ATAGTG": 316, + "ATAGCA": 317, + "ATAGCT": 318, + "ATAGCC": 319, + "ATAGCG": 320, + "ATAGGA": 321, + "ATAGGT": 322, + "ATAGGC": 323, + "ATAGGG": 324, + "ATTAAA": 325, + "ATTAAT": 326, + "ATTAAC": 327, + "ATTAAG": 328, + "ATTATA": 329, + "ATTATT": 330, + "ATTATC": 331, + "ATTATG": 332, + "ATTACA": 333, + "ATTACT": 334, + "ATTACC": 335, + "ATTACG": 336, + "ATTAGA": 337, + "ATTAGT": 338, + "ATTAGC": 339, + "ATTAGG": 340, + "ATTTAA": 341, + "ATTTAT": 342, + "ATTTAC": 343, + "ATTTAG": 344, + "ATTTTA": 345, + "ATTTTT": 346, + "ATTTTC": 347, + "ATTTTG": 348, + "ATTTCA": 349, + "ATTTCT": 350, + "ATTTCC": 351, + "ATTTCG": 352, + "ATTTGA": 353, + "ATTTGT": 354, + "ATTTGC": 355, + "ATTTGG": 356, + "ATTCAA": 357, + "ATTCAT": 358, + "ATTCAC": 359, + "ATTCAG": 360, + "ATTCTA": 361, + "ATTCTT": 362, + "ATTCTC": 363, + "ATTCTG": 364, + "ATTCCA": 365, + "ATTCCT": 366, + "ATTCCC": 367, + "ATTCCG": 368, + "ATTCGA": 369, + "ATTCGT": 370, + "ATTCGC": 371, + "ATTCGG": 372, + "ATTGAA": 373, + "ATTGAT": 374, + "ATTGAC": 375, + "ATTGAG": 376, + "ATTGTA": 377, + "ATTGTT": 378, + "ATTGTC": 379, + "ATTGTG": 380, + "ATTGCA": 381, + "ATTGCT": 382, + "ATTGCC": 383, + "ATTGCG": 384, + "ATTGGA": 385, + "ATTGGT": 386, + "ATTGGC": 387, + "ATTGGG": 388, + "ATCAAA": 389, + "ATCAAT": 390, + "ATCAAC": 391, + "ATCAAG": 392, + "ATCATA": 393, + "ATCATT": 394, + "ATCATC": 395, + "ATCATG": 396, + "ATCACA": 397, + "ATCACT": 398, + "ATCACC": 399, + "ATCACG": 400, + "ATCAGA": 401, + "ATCAGT": 402, + "ATCAGC": 403, + "ATCAGG": 404, + "ATCTAA": 405, + "ATCTAT": 406, + "ATCTAC": 407, + "ATCTAG": 408, + "ATCTTA": 409, + "ATCTTT": 410, + "ATCTTC": 411, + "ATCTTG": 412, + "ATCTCA": 413, + "ATCTCT": 414, + "ATCTCC": 415, + "ATCTCG": 416, + "ATCTGA": 417, + "ATCTGT": 418, + "ATCTGC": 419, + "ATCTGG": 420, + "ATCCAA": 421, + "ATCCAT": 422, + "ATCCAC": 423, + "ATCCAG": 424, + "ATCCTA": 425, + "ATCCTT": 426, + "ATCCTC": 427, + "ATCCTG": 428, + "ATCCCA": 429, + "ATCCCT": 430, + "ATCCCC": 431, + "ATCCCG": 432, + "ATCCGA": 433, + "ATCCGT": 434, + "ATCCGC": 435, + "ATCCGG": 436, + "ATCGAA": 437, + "ATCGAT": 438, + "ATCGAC": 439, + "ATCGAG": 440, + "ATCGTA": 441, + "ATCGTT": 442, + "ATCGTC": 443, + "ATCGTG": 444, + "ATCGCA": 445, + "ATCGCT": 446, + "ATCGCC": 447, + "ATCGCG": 448, + "ATCGGA": 449, + "ATCGGT": 450, + "ATCGGC": 451, + "ATCGGG": 452, + "ATGAAA": 453, + "ATGAAT": 454, + "ATGAAC": 455, + "ATGAAG": 456, + "ATGATA": 457, + "ATGATT": 458, + "ATGATC": 459, + "ATGATG": 460, + "ATGACA": 461, + "ATGACT": 462, + "ATGACC": 463, + "ATGACG": 464, + "ATGAGA": 465, + "ATGAGT": 466, + "ATGAGC": 467, + "ATGAGG": 468, + "ATGTAA": 469, + "ATGTAT": 470, + "ATGTAC": 471, + "ATGTAG": 472, + "ATGTTA": 473, + "ATGTTT": 474, + "ATGTTC": 475, + "ATGTTG": 476, + "ATGTCA": 477, + "ATGTCT": 478, + "ATGTCC": 479, + "ATGTCG": 480, + "ATGTGA": 481, + "ATGTGT": 482, + "ATGTGC": 483, + "ATGTGG": 484, + "ATGCAA": 485, + "ATGCAT": 486, + "ATGCAC": 487, + "ATGCAG": 488, + "ATGCTA": 489, + "ATGCTT": 490, + "ATGCTC": 491, + "ATGCTG": 492, + "ATGCCA": 493, + "ATGCCT": 494, + "ATGCCC": 495, + "ATGCCG": 496, + "ATGCGA": 497, + "ATGCGT": 498, + "ATGCGC": 499, + "ATGCGG": 500, + "ATGGAA": 501, + "ATGGAT": 502, + "ATGGAC": 503, + "ATGGAG": 504, + "ATGGTA": 505, + "ATGGTT": 506, + "ATGGTC": 507, + "ATGGTG": 508, + "ATGGCA": 509, + "ATGGCT": 510, + "ATGGCC": 511, + "ATGGCG": 512, + "ATGGGA": 513, + "ATGGGT": 514, + "ATGGGC": 515, + "ATGGGG": 516, + "ACAAAA": 517, + "ACAAAT": 518, + "ACAAAC": 519, + "ACAAAG": 520, + "ACAATA": 521, + "ACAATT": 522, + "ACAATC": 523, + "ACAATG": 524, + "ACAACA": 525, + "ACAACT": 526, + "ACAACC": 527, + "ACAACG": 528, + "ACAAGA": 529, + "ACAAGT": 530, + "ACAAGC": 531, + "ACAAGG": 532, + "ACATAA": 533, + "ACATAT": 534, + "ACATAC": 535, + "ACATAG": 536, + "ACATTA": 537, + "ACATTT": 538, + "ACATTC": 539, + "ACATTG": 540, + "ACATCA": 541, + "ACATCT": 542, + "ACATCC": 543, + "ACATCG": 544, + "ACATGA": 545, + "ACATGT": 546, + "ACATGC": 547, + "ACATGG": 548, + "ACACAA": 549, + "ACACAT": 550, + "ACACAC": 551, + "ACACAG": 552, + "ACACTA": 553, + "ACACTT": 554, + "ACACTC": 555, + "ACACTG": 556, + "ACACCA": 557, + "ACACCT": 558, + "ACACCC": 559, + "ACACCG": 560, + "ACACGA": 561, + "ACACGT": 562, + "ACACGC": 563, + "ACACGG": 564, + "ACAGAA": 565, + "ACAGAT": 566, + "ACAGAC": 567, + "ACAGAG": 568, + "ACAGTA": 569, + "ACAGTT": 570, + "ACAGTC": 571, + "ACAGTG": 572, + "ACAGCA": 573, + "ACAGCT": 574, + "ACAGCC": 575, + "ACAGCG": 576, + "ACAGGA": 577, + "ACAGGT": 578, + "ACAGGC": 579, + "ACAGGG": 580, + "ACTAAA": 581, + "ACTAAT": 582, + "ACTAAC": 583, + "ACTAAG": 584, + "ACTATA": 585, + "ACTATT": 586, + "ACTATC": 587, + "ACTATG": 588, + "ACTACA": 589, + "ACTACT": 590, + "ACTACC": 591, + "ACTACG": 592, + "ACTAGA": 593, + "ACTAGT": 594, + "ACTAGC": 595, + "ACTAGG": 596, + "ACTTAA": 597, + "ACTTAT": 598, + "ACTTAC": 599, + "ACTTAG": 600, + "ACTTTA": 601, + "ACTTTT": 602, + "ACTTTC": 603, + "ACTTTG": 604, + "ACTTCA": 605, + "ACTTCT": 606, + "ACTTCC": 607, + "ACTTCG": 608, + "ACTTGA": 609, + "ACTTGT": 610, + "ACTTGC": 611, + "ACTTGG": 612, + "ACTCAA": 613, + "ACTCAT": 614, + "ACTCAC": 615, + "ACTCAG": 616, + "ACTCTA": 617, + "ACTCTT": 618, + "ACTCTC": 619, + "ACTCTG": 620, + "ACTCCA": 621, + "ACTCCT": 622, + "ACTCCC": 623, + "ACTCCG": 624, + "ACTCGA": 625, + "ACTCGT": 626, + "ACTCGC": 627, + "ACTCGG": 628, + "ACTGAA": 629, + "ACTGAT": 630, + "ACTGAC": 631, + "ACTGAG": 632, + "ACTGTA": 633, + "ACTGTT": 634, + "ACTGTC": 635, + "ACTGTG": 636, + "ACTGCA": 637, + "ACTGCT": 638, + "ACTGCC": 639, + "ACTGCG": 640, + "ACTGGA": 641, + "ACTGGT": 642, + "ACTGGC": 643, + "ACTGGG": 644, + "ACCAAA": 645, + "ACCAAT": 646, + "ACCAAC": 647, + "ACCAAG": 648, + "ACCATA": 649, + "ACCATT": 650, + "ACCATC": 651, + "ACCATG": 652, + "ACCACA": 653, + "ACCACT": 654, + "ACCACC": 655, + "ACCACG": 656, + "ACCAGA": 657, + "ACCAGT": 658, + "ACCAGC": 659, + "ACCAGG": 660, + "ACCTAA": 661, + "ACCTAT": 662, + "ACCTAC": 663, + "ACCTAG": 664, + "ACCTTA": 665, + "ACCTTT": 666, + "ACCTTC": 667, + "ACCTTG": 668, + "ACCTCA": 669, + "ACCTCT": 670, + "ACCTCC": 671, + "ACCTCG": 672, + "ACCTGA": 673, + "ACCTGT": 674, + "ACCTGC": 675, + "ACCTGG": 676, + "ACCCAA": 677, + "ACCCAT": 678, + "ACCCAC": 679, + "ACCCAG": 680, + "ACCCTA": 681, + "ACCCTT": 682, + "ACCCTC": 683, + "ACCCTG": 684, + "ACCCCA": 685, + "ACCCCT": 686, + "ACCCCC": 687, + "ACCCCG": 688, + "ACCCGA": 689, + "ACCCGT": 690, + "ACCCGC": 691, + "ACCCGG": 692, + "ACCGAA": 693, + "ACCGAT": 694, + "ACCGAC": 695, + "ACCGAG": 696, + "ACCGTA": 697, + "ACCGTT": 698, + "ACCGTC": 699, + "ACCGTG": 700, + "ACCGCA": 701, + "ACCGCT": 702, + "ACCGCC": 703, + "ACCGCG": 704, + "ACCGGA": 705, + "ACCGGT": 706, + "ACCGGC": 707, + "ACCGGG": 708, + "ACGAAA": 709, + "ACGAAT": 710, + "ACGAAC": 711, + "ACGAAG": 712, + "ACGATA": 713, + "ACGATT": 714, + "ACGATC": 715, + "ACGATG": 716, + "ACGACA": 717, + "ACGACT": 718, + "ACGACC": 719, + "ACGACG": 720, + "ACGAGA": 721, + "ACGAGT": 722, + "ACGAGC": 723, + "ACGAGG": 724, + "ACGTAA": 725, + "ACGTAT": 726, + "ACGTAC": 727, + "ACGTAG": 728, + "ACGTTA": 729, + "ACGTTT": 730, + "ACGTTC": 731, + "ACGTTG": 732, + "ACGTCA": 733, + "ACGTCT": 734, + "ACGTCC": 735, + "ACGTCG": 736, + "ACGTGA": 737, + "ACGTGT": 738, + "ACGTGC": 739, + "ACGTGG": 740, + "ACGCAA": 741, + "ACGCAT": 742, + "ACGCAC": 743, + "ACGCAG": 744, + "ACGCTA": 745, + "ACGCTT": 746, + "ACGCTC": 747, + "ACGCTG": 748, + "ACGCCA": 749, + "ACGCCT": 750, + "ACGCCC": 751, + "ACGCCG": 752, + "ACGCGA": 753, + "ACGCGT": 754, + "ACGCGC": 755, + "ACGCGG": 756, + "ACGGAA": 757, + "ACGGAT": 758, + "ACGGAC": 759, + "ACGGAG": 760, + "ACGGTA": 761, + "ACGGTT": 762, + "ACGGTC": 763, + "ACGGTG": 764, + "ACGGCA": 765, + "ACGGCT": 766, + "ACGGCC": 767, + "ACGGCG": 768, + "ACGGGA": 769, + "ACGGGT": 770, + "ACGGGC": 771, + "ACGGGG": 772, + "AGAAAA": 773, + "AGAAAT": 774, + "AGAAAC": 775, + "AGAAAG": 776, + "AGAATA": 777, + "AGAATT": 778, + "AGAATC": 779, + "AGAATG": 780, + "AGAACA": 781, + "AGAACT": 782, + "AGAACC": 783, + "AGAACG": 784, + "AGAAGA": 785, + "AGAAGT": 786, + "AGAAGC": 787, + "AGAAGG": 788, + "AGATAA": 789, + "AGATAT": 790, + "AGATAC": 791, + "AGATAG": 792, + "AGATTA": 793, + "AGATTT": 794, + "AGATTC": 795, + "AGATTG": 796, + "AGATCA": 797, + "AGATCT": 798, + "AGATCC": 799, + "AGATCG": 800, + "AGATGA": 801, + "AGATGT": 802, + "AGATGC": 803, + "AGATGG": 804, + "AGACAA": 805, + "AGACAT": 806, + "AGACAC": 807, + "AGACAG": 808, + "AGACTA": 809, + "AGACTT": 810, + "AGACTC": 811, + "AGACTG": 812, + "AGACCA": 813, + "AGACCT": 814, + "AGACCC": 815, + "AGACCG": 816, + "AGACGA": 817, + "AGACGT": 818, + "AGACGC": 819, + "AGACGG": 820, + "AGAGAA": 821, + "AGAGAT": 822, + "AGAGAC": 823, + "AGAGAG": 824, + "AGAGTA": 825, + "AGAGTT": 826, + "AGAGTC": 827, + "AGAGTG": 828, + "AGAGCA": 829, + "AGAGCT": 830, + "AGAGCC": 831, + "AGAGCG": 832, + "AGAGGA": 833, + "AGAGGT": 834, + "AGAGGC": 835, + "AGAGGG": 836, + "AGTAAA": 837, + "AGTAAT": 838, + "AGTAAC": 839, + "AGTAAG": 840, + "AGTATA": 841, + "AGTATT": 842, + "AGTATC": 843, + "AGTATG": 844, + "AGTACA": 845, + "AGTACT": 846, + "AGTACC": 847, + "AGTACG": 848, + "AGTAGA": 849, + "AGTAGT": 850, + "AGTAGC": 851, + "AGTAGG": 852, + "AGTTAA": 853, + "AGTTAT": 854, + "AGTTAC": 855, + "AGTTAG": 856, + "AGTTTA": 857, + "AGTTTT": 858, + "AGTTTC": 859, + "AGTTTG": 860, + "AGTTCA": 861, + "AGTTCT": 862, + "AGTTCC": 863, + "AGTTCG": 864, + "AGTTGA": 865, + "AGTTGT": 866, + "AGTTGC": 867, + "AGTTGG": 868, + "AGTCAA": 869, + "AGTCAT": 870, + "AGTCAC": 871, + "AGTCAG": 872, + "AGTCTA": 873, + "AGTCTT": 874, + "AGTCTC": 875, + "AGTCTG": 876, + "AGTCCA": 877, + "AGTCCT": 878, + "AGTCCC": 879, + "AGTCCG": 880, + "AGTCGA": 881, + "AGTCGT": 882, + "AGTCGC": 883, + "AGTCGG": 884, + "AGTGAA": 885, + "AGTGAT": 886, + "AGTGAC": 887, + "AGTGAG": 888, + "AGTGTA": 889, + "AGTGTT": 890, + "AGTGTC": 891, + "AGTGTG": 892, + "AGTGCA": 893, + "AGTGCT": 894, + "AGTGCC": 895, + "AGTGCG": 896, + "AGTGGA": 897, + "AGTGGT": 898, + "AGTGGC": 899, + "AGTGGG": 900, + "AGCAAA": 901, + "AGCAAT": 902, + "AGCAAC": 903, + "AGCAAG": 904, + "AGCATA": 905, + "AGCATT": 906, + "AGCATC": 907, + "AGCATG": 908, + "AGCACA": 909, + "AGCACT": 910, + "AGCACC": 911, + "AGCACG": 912, + "AGCAGA": 913, + "AGCAGT": 914, + "AGCAGC": 915, + "AGCAGG": 916, + "AGCTAA": 917, + "AGCTAT": 918, + "AGCTAC": 919, + "AGCTAG": 920, + "AGCTTA": 921, + "AGCTTT": 922, + "AGCTTC": 923, + "AGCTTG": 924, + "AGCTCA": 925, + "AGCTCT": 926, + "AGCTCC": 927, + "AGCTCG": 928, + "AGCTGA": 929, + "AGCTGT": 930, + "AGCTGC": 931, + "AGCTGG": 932, + "AGCCAA": 933, + "AGCCAT": 934, + "AGCCAC": 935, + "AGCCAG": 936, + "AGCCTA": 937, + "AGCCTT": 938, + "AGCCTC": 939, + "AGCCTG": 940, + "AGCCCA": 941, + "AGCCCT": 942, + "AGCCCC": 943, + "AGCCCG": 944, + "AGCCGA": 945, + "AGCCGT": 946, + "AGCCGC": 947, + "AGCCGG": 948, + "AGCGAA": 949, + "AGCGAT": 950, + "AGCGAC": 951, + "AGCGAG": 952, + "AGCGTA": 953, + "AGCGTT": 954, + "AGCGTC": 955, + "AGCGTG": 956, + "AGCGCA": 957, + "AGCGCT": 958, + "AGCGCC": 959, + "AGCGCG": 960, + "AGCGGA": 961, + "AGCGGT": 962, + "AGCGGC": 963, + "AGCGGG": 964, + "AGGAAA": 965, + "AGGAAT": 966, + "AGGAAC": 967, + "AGGAAG": 968, + "AGGATA": 969, + "AGGATT": 970, + "AGGATC": 971, + "AGGATG": 972, + "AGGACA": 973, + "AGGACT": 974, + "AGGACC": 975, + "AGGACG": 976, + "AGGAGA": 977, + "AGGAGT": 978, + "AGGAGC": 979, + "AGGAGG": 980, + "AGGTAA": 981, + "AGGTAT": 982, + "AGGTAC": 983, + "AGGTAG": 984, + "AGGTTA": 985, + "AGGTTT": 986, + "AGGTTC": 987, + "AGGTTG": 988, + "AGGTCA": 989, + "AGGTCT": 990, + "AGGTCC": 991, + "AGGTCG": 992, + "AGGTGA": 993, + "AGGTGT": 994, + "AGGTGC": 995, + "AGGTGG": 996, + "AGGCAA": 997, + "AGGCAT": 998, + "AGGCAC": 999, + "AGGCAG": 1000, + "AGGCTA": 1001, + "AGGCTT": 1002, + "AGGCTC": 1003, + "AGGCTG": 1004, + "AGGCCA": 1005, + "AGGCCT": 1006, + "AGGCCC": 1007, + "AGGCCG": 1008, + "AGGCGA": 1009, + "AGGCGT": 1010, + "AGGCGC": 1011, + "AGGCGG": 1012, + "AGGGAA": 1013, + "AGGGAT": 1014, + "AGGGAC": 1015, + "AGGGAG": 1016, + "AGGGTA": 1017, + "AGGGTT": 1018, + "AGGGTC": 1019, + "AGGGTG": 1020, + "AGGGCA": 1021, + "AGGGCT": 1022, + "AGGGCC": 1023, + "AGGGCG": 1024, + "AGGGGA": 1025, + "AGGGGT": 1026, + "AGGGGC": 1027, + "AGGGGG": 1028, + "TAAAAA": 1029, + "TAAAAT": 1030, + "TAAAAC": 1031, + "TAAAAG": 1032, + "TAAATA": 1033, + "TAAATT": 1034, + "TAAATC": 1035, + "TAAATG": 1036, + "TAAACA": 1037, + "TAAACT": 1038, + "TAAACC": 1039, + "TAAACG": 1040, + "TAAAGA": 1041, + "TAAAGT": 1042, + "TAAAGC": 1043, + "TAAAGG": 1044, + "TAATAA": 1045, + "TAATAT": 1046, + "TAATAC": 1047, + "TAATAG": 1048, + "TAATTA": 1049, + "TAATTT": 1050, + "TAATTC": 1051, + "TAATTG": 1052, + "TAATCA": 1053, + "TAATCT": 1054, + "TAATCC": 1055, + "TAATCG": 1056, + "TAATGA": 1057, + "TAATGT": 1058, + "TAATGC": 1059, + "TAATGG": 1060, + "TAACAA": 1061, + "TAACAT": 1062, + "TAACAC": 1063, + "TAACAG": 1064, + "TAACTA": 1065, + "TAACTT": 1066, + "TAACTC": 1067, + "TAACTG": 1068, + "TAACCA": 1069, + "TAACCT": 1070, + "TAACCC": 1071, + "TAACCG": 1072, + "TAACGA": 1073, + "TAACGT": 1074, + "TAACGC": 1075, + "TAACGG": 1076, + "TAAGAA": 1077, + "TAAGAT": 1078, + "TAAGAC": 1079, + "TAAGAG": 1080, + "TAAGTA": 1081, + "TAAGTT": 1082, + "TAAGTC": 1083, + "TAAGTG": 1084, + "TAAGCA": 1085, + "TAAGCT": 1086, + "TAAGCC": 1087, + "TAAGCG": 1088, + "TAAGGA": 1089, + "TAAGGT": 1090, + "TAAGGC": 1091, + "TAAGGG": 1092, + "TATAAA": 1093, + "TATAAT": 1094, + "TATAAC": 1095, + "TATAAG": 1096, + "TATATA": 1097, + "TATATT": 1098, + "TATATC": 1099, + "TATATG": 1100, + "TATACA": 1101, + "TATACT": 1102, + "TATACC": 1103, + "TATACG": 1104, + "TATAGA": 1105, + "TATAGT": 1106, + "TATAGC": 1107, + "TATAGG": 1108, + "TATTAA": 1109, + "TATTAT": 1110, + "TATTAC": 1111, + "TATTAG": 1112, + "TATTTA": 1113, + "TATTTT": 1114, + "TATTTC": 1115, + "TATTTG": 1116, + "TATTCA": 1117, + "TATTCT": 1118, + "TATTCC": 1119, + "TATTCG": 1120, + "TATTGA": 1121, + "TATTGT": 1122, + "TATTGC": 1123, + "TATTGG": 1124, + "TATCAA": 1125, + "TATCAT": 1126, + "TATCAC": 1127, + "TATCAG": 1128, + "TATCTA": 1129, + "TATCTT": 1130, + "TATCTC": 1131, + "TATCTG": 1132, + "TATCCA": 1133, + "TATCCT": 1134, + "TATCCC": 1135, + "TATCCG": 1136, + "TATCGA": 1137, + "TATCGT": 1138, + "TATCGC": 1139, + "TATCGG": 1140, + "TATGAA": 1141, + "TATGAT": 1142, + "TATGAC": 1143, + "TATGAG": 1144, + "TATGTA": 1145, + "TATGTT": 1146, + "TATGTC": 1147, + "TATGTG": 1148, + "TATGCA": 1149, + "TATGCT": 1150, + "TATGCC": 1151, + "TATGCG": 1152, + "TATGGA": 1153, + "TATGGT": 1154, + "TATGGC": 1155, + "TATGGG": 1156, + "TACAAA": 1157, + "TACAAT": 1158, + "TACAAC": 1159, + "TACAAG": 1160, + "TACATA": 1161, + "TACATT": 1162, + "TACATC": 1163, + "TACATG": 1164, + "TACACA": 1165, + "TACACT": 1166, + "TACACC": 1167, + "TACACG": 1168, + "TACAGA": 1169, + "TACAGT": 1170, + "TACAGC": 1171, + "TACAGG": 1172, + "TACTAA": 1173, + "TACTAT": 1174, + "TACTAC": 1175, + "TACTAG": 1176, + "TACTTA": 1177, + "TACTTT": 1178, + "TACTTC": 1179, + "TACTTG": 1180, + "TACTCA": 1181, + "TACTCT": 1182, + "TACTCC": 1183, + "TACTCG": 1184, + "TACTGA": 1185, + "TACTGT": 1186, + "TACTGC": 1187, + "TACTGG": 1188, + "TACCAA": 1189, + "TACCAT": 1190, + "TACCAC": 1191, + "TACCAG": 1192, + "TACCTA": 1193, + "TACCTT": 1194, + "TACCTC": 1195, + "TACCTG": 1196, + "TACCCA": 1197, + "TACCCT": 1198, + "TACCCC": 1199, + "TACCCG": 1200, + "TACCGA": 1201, + "TACCGT": 1202, + "TACCGC": 1203, + "TACCGG": 1204, + "TACGAA": 1205, + "TACGAT": 1206, + "TACGAC": 1207, + "TACGAG": 1208, + "TACGTA": 1209, + "TACGTT": 1210, + "TACGTC": 1211, + "TACGTG": 1212, + "TACGCA": 1213, + "TACGCT": 1214, + "TACGCC": 1215, + "TACGCG": 1216, + "TACGGA": 1217, + "TACGGT": 1218, + "TACGGC": 1219, + "TACGGG": 1220, + "TAGAAA": 1221, + "TAGAAT": 1222, + "TAGAAC": 1223, + "TAGAAG": 1224, + "TAGATA": 1225, + "TAGATT": 1226, + "TAGATC": 1227, + "TAGATG": 1228, + "TAGACA": 1229, + "TAGACT": 1230, + "TAGACC": 1231, + "TAGACG": 1232, + "TAGAGA": 1233, + "TAGAGT": 1234, + "TAGAGC": 1235, + "TAGAGG": 1236, + "TAGTAA": 1237, + "TAGTAT": 1238, + "TAGTAC": 1239, + "TAGTAG": 1240, + "TAGTTA": 1241, + "TAGTTT": 1242, + "TAGTTC": 1243, + "TAGTTG": 1244, + "TAGTCA": 1245, + "TAGTCT": 1246, + "TAGTCC": 1247, + "TAGTCG": 1248, + "TAGTGA": 1249, + "TAGTGT": 1250, + "TAGTGC": 1251, + "TAGTGG": 1252, + "TAGCAA": 1253, + "TAGCAT": 1254, + "TAGCAC": 1255, + "TAGCAG": 1256, + "TAGCTA": 1257, + "TAGCTT": 1258, + "TAGCTC": 1259, + "TAGCTG": 1260, + "TAGCCA": 1261, + "TAGCCT": 1262, + "TAGCCC": 1263, + "TAGCCG": 1264, + "TAGCGA": 1265, + "TAGCGT": 1266, + "TAGCGC": 1267, + "TAGCGG": 1268, + "TAGGAA": 1269, + "TAGGAT": 1270, + "TAGGAC": 1271, + "TAGGAG": 1272, + "TAGGTA": 1273, + "TAGGTT": 1274, + "TAGGTC": 1275, + "TAGGTG": 1276, + "TAGGCA": 1277, + "TAGGCT": 1278, + "TAGGCC": 1279, + "TAGGCG": 1280, + "TAGGGA": 1281, + "TAGGGT": 1282, + "TAGGGC": 1283, + "TAGGGG": 1284, + "TTAAAA": 1285, + "TTAAAT": 1286, + "TTAAAC": 1287, + "TTAAAG": 1288, + "TTAATA": 1289, + "TTAATT": 1290, + "TTAATC": 1291, + "TTAATG": 1292, + "TTAACA": 1293, + "TTAACT": 1294, + "TTAACC": 1295, + "TTAACG": 1296, + "TTAAGA": 1297, + "TTAAGT": 1298, + "TTAAGC": 1299, + "TTAAGG": 1300, + "TTATAA": 1301, + "TTATAT": 1302, + "TTATAC": 1303, + "TTATAG": 1304, + "TTATTA": 1305, + "TTATTT": 1306, + "TTATTC": 1307, + "TTATTG": 1308, + "TTATCA": 1309, + "TTATCT": 1310, + "TTATCC": 1311, + "TTATCG": 1312, + "TTATGA": 1313, + "TTATGT": 1314, + "TTATGC": 1315, + "TTATGG": 1316, + "TTACAA": 1317, + "TTACAT": 1318, + "TTACAC": 1319, + "TTACAG": 1320, + "TTACTA": 1321, + "TTACTT": 1322, + "TTACTC": 1323, + "TTACTG": 1324, + "TTACCA": 1325, + "TTACCT": 1326, + "TTACCC": 1327, + "TTACCG": 1328, + "TTACGA": 1329, + "TTACGT": 1330, + "TTACGC": 1331, + "TTACGG": 1332, + "TTAGAA": 1333, + "TTAGAT": 1334, + "TTAGAC": 1335, + "TTAGAG": 1336, + "TTAGTA": 1337, + "TTAGTT": 1338, + "TTAGTC": 1339, + "TTAGTG": 1340, + "TTAGCA": 1341, + "TTAGCT": 1342, + "TTAGCC": 1343, + "TTAGCG": 1344, + "TTAGGA": 1345, + "TTAGGT": 1346, + "TTAGGC": 1347, + "TTAGGG": 1348, + "TTTAAA": 1349, + "TTTAAT": 1350, + "TTTAAC": 1351, + "TTTAAG": 1352, + "TTTATA": 1353, + "TTTATT": 1354, + "TTTATC": 1355, + "TTTATG": 1356, + "TTTACA": 1357, + "TTTACT": 1358, + "TTTACC": 1359, + "TTTACG": 1360, + "TTTAGA": 1361, + "TTTAGT": 1362, + "TTTAGC": 1363, + "TTTAGG": 1364, + "TTTTAA": 1365, + "TTTTAT": 1366, + "TTTTAC": 1367, + "TTTTAG": 1368, + "TTTTTA": 1369, + "TTTTTT": 1370, + "TTTTTC": 1371, + "TTTTTG": 1372, + "TTTTCA": 1373, + "TTTTCT": 1374, + "TTTTCC": 1375, + "TTTTCG": 1376, + "TTTTGA": 1377, + "TTTTGT": 1378, + "TTTTGC": 1379, + "TTTTGG": 1380, + "TTTCAA": 1381, + "TTTCAT": 1382, + "TTTCAC": 1383, + "TTTCAG": 1384, + "TTTCTA": 1385, + "TTTCTT": 1386, + "TTTCTC": 1387, + "TTTCTG": 1388, + "TTTCCA": 1389, + "TTTCCT": 1390, + "TTTCCC": 1391, + "TTTCCG": 1392, + "TTTCGA": 1393, + "TTTCGT": 1394, + "TTTCGC": 1395, + "TTTCGG": 1396, + "TTTGAA": 1397, + "TTTGAT": 1398, + "TTTGAC": 1399, + "TTTGAG": 1400, + "TTTGTA": 1401, + "TTTGTT": 1402, + "TTTGTC": 1403, + "TTTGTG": 1404, + "TTTGCA": 1405, + "TTTGCT": 1406, + "TTTGCC": 1407, + "TTTGCG": 1408, + "TTTGGA": 1409, + "TTTGGT": 1410, + "TTTGGC": 1411, + "TTTGGG": 1412, + "TTCAAA": 1413, + "TTCAAT": 1414, + "TTCAAC": 1415, + "TTCAAG": 1416, + "TTCATA": 1417, + "TTCATT": 1418, + "TTCATC": 1419, + "TTCATG": 1420, + "TTCACA": 1421, + "TTCACT": 1422, + "TTCACC": 1423, + "TTCACG": 1424, + "TTCAGA": 1425, + "TTCAGT": 1426, + "TTCAGC": 1427, + "TTCAGG": 1428, + "TTCTAA": 1429, + "TTCTAT": 1430, + "TTCTAC": 1431, + "TTCTAG": 1432, + "TTCTTA": 1433, + "TTCTTT": 1434, + "TTCTTC": 1435, + "TTCTTG": 1436, + "TTCTCA": 1437, + "TTCTCT": 1438, + "TTCTCC": 1439, + "TTCTCG": 1440, + "TTCTGA": 1441, + "TTCTGT": 1442, + "TTCTGC": 1443, + "TTCTGG": 1444, + "TTCCAA": 1445, + "TTCCAT": 1446, + "TTCCAC": 1447, + "TTCCAG": 1448, + "TTCCTA": 1449, + "TTCCTT": 1450, + "TTCCTC": 1451, + "TTCCTG": 1452, + "TTCCCA": 1453, + "TTCCCT": 1454, + "TTCCCC": 1455, + "TTCCCG": 1456, + "TTCCGA": 1457, + "TTCCGT": 1458, + "TTCCGC": 1459, + "TTCCGG": 1460, + "TTCGAA": 1461, + "TTCGAT": 1462, + "TTCGAC": 1463, + "TTCGAG": 1464, + "TTCGTA": 1465, + "TTCGTT": 1466, + "TTCGTC": 1467, + "TTCGTG": 1468, + "TTCGCA": 1469, + "TTCGCT": 1470, + "TTCGCC": 1471, + "TTCGCG": 1472, + "TTCGGA": 1473, + "TTCGGT": 1474, + "TTCGGC": 1475, + "TTCGGG": 1476, + "TTGAAA": 1477, + "TTGAAT": 1478, + "TTGAAC": 1479, + "TTGAAG": 1480, + "TTGATA": 1481, + "TTGATT": 1482, + "TTGATC": 1483, + "TTGATG": 1484, + "TTGACA": 1485, + "TTGACT": 1486, + "TTGACC": 1487, + "TTGACG": 1488, + "TTGAGA": 1489, + "TTGAGT": 1490, + "TTGAGC": 1491, + "TTGAGG": 1492, + "TTGTAA": 1493, + "TTGTAT": 1494, + "TTGTAC": 1495, + "TTGTAG": 1496, + "TTGTTA": 1497, + "TTGTTT": 1498, + "TTGTTC": 1499, + "TTGTTG": 1500, + "TTGTCA": 1501, + "TTGTCT": 1502, + "TTGTCC": 1503, + "TTGTCG": 1504, + "TTGTGA": 1505, + "TTGTGT": 1506, + "TTGTGC": 1507, + "TTGTGG": 1508, + "TTGCAA": 1509, + "TTGCAT": 1510, + "TTGCAC": 1511, + "TTGCAG": 1512, + "TTGCTA": 1513, + "TTGCTT": 1514, + "TTGCTC": 1515, + "TTGCTG": 1516, + "TTGCCA": 1517, + "TTGCCT": 1518, + "TTGCCC": 1519, + "TTGCCG": 1520, + "TTGCGA": 1521, + "TTGCGT": 1522, + "TTGCGC": 1523, + "TTGCGG": 1524, + "TTGGAA": 1525, + "TTGGAT": 1526, + "TTGGAC": 1527, + "TTGGAG": 1528, + "TTGGTA": 1529, + "TTGGTT": 1530, + "TTGGTC": 1531, + "TTGGTG": 1532, + "TTGGCA": 1533, + "TTGGCT": 1534, + "TTGGCC": 1535, + "TTGGCG": 1536, + "TTGGGA": 1537, + "TTGGGT": 1538, + "TTGGGC": 1539, + "TTGGGG": 1540, + "TCAAAA": 1541, + "TCAAAT": 1542, + "TCAAAC": 1543, + "TCAAAG": 1544, + "TCAATA": 1545, + "TCAATT": 1546, + "TCAATC": 1547, + "TCAATG": 1548, + "TCAACA": 1549, + "TCAACT": 1550, + "TCAACC": 1551, + "TCAACG": 1552, + "TCAAGA": 1553, + "TCAAGT": 1554, + "TCAAGC": 1555, + "TCAAGG": 1556, + "TCATAA": 1557, + "TCATAT": 1558, + "TCATAC": 1559, + "TCATAG": 1560, + "TCATTA": 1561, + "TCATTT": 1562, + "TCATTC": 1563, + "TCATTG": 1564, + "TCATCA": 1565, + "TCATCT": 1566, + "TCATCC": 1567, + "TCATCG": 1568, + "TCATGA": 1569, + "TCATGT": 1570, + "TCATGC": 1571, + "TCATGG": 1572, + "TCACAA": 1573, + "TCACAT": 1574, + "TCACAC": 1575, + "TCACAG": 1576, + "TCACTA": 1577, + "TCACTT": 1578, + "TCACTC": 1579, + "TCACTG": 1580, + "TCACCA": 1581, + "TCACCT": 1582, + "TCACCC": 1583, + "TCACCG": 1584, + "TCACGA": 1585, + "TCACGT": 1586, + "TCACGC": 1587, + "TCACGG": 1588, + "TCAGAA": 1589, + "TCAGAT": 1590, + "TCAGAC": 1591, + "TCAGAG": 1592, + "TCAGTA": 1593, + "TCAGTT": 1594, + "TCAGTC": 1595, + "TCAGTG": 1596, + "TCAGCA": 1597, + "TCAGCT": 1598, + "TCAGCC": 1599, + "TCAGCG": 1600, + "TCAGGA": 1601, + "TCAGGT": 1602, + "TCAGGC": 1603, + "TCAGGG": 1604, + "TCTAAA": 1605, + "TCTAAT": 1606, + "TCTAAC": 1607, + "TCTAAG": 1608, + "TCTATA": 1609, + "TCTATT": 1610, + "TCTATC": 1611, + "TCTATG": 1612, + "TCTACA": 1613, + "TCTACT": 1614, + "TCTACC": 1615, + "TCTACG": 1616, + "TCTAGA": 1617, + "TCTAGT": 1618, + "TCTAGC": 1619, + "TCTAGG": 1620, + "TCTTAA": 1621, + "TCTTAT": 1622, + "TCTTAC": 1623, + "TCTTAG": 1624, + "TCTTTA": 1625, + "TCTTTT": 1626, + "TCTTTC": 1627, + "TCTTTG": 1628, + "TCTTCA": 1629, + "TCTTCT": 1630, + "TCTTCC": 1631, + "TCTTCG": 1632, + "TCTTGA": 1633, + "TCTTGT": 1634, + "TCTTGC": 1635, + "TCTTGG": 1636, + "TCTCAA": 1637, + "TCTCAT": 1638, + "TCTCAC": 1639, + "TCTCAG": 1640, + "TCTCTA": 1641, + "TCTCTT": 1642, + "TCTCTC": 1643, + "TCTCTG": 1644, + "TCTCCA": 1645, + "TCTCCT": 1646, + "TCTCCC": 1647, + "TCTCCG": 1648, + "TCTCGA": 1649, + "TCTCGT": 1650, + "TCTCGC": 1651, + "TCTCGG": 1652, + "TCTGAA": 1653, + "TCTGAT": 1654, + "TCTGAC": 1655, + "TCTGAG": 1656, + "TCTGTA": 1657, + "TCTGTT": 1658, + "TCTGTC": 1659, + "TCTGTG": 1660, + "TCTGCA": 1661, + "TCTGCT": 1662, + "TCTGCC": 1663, + "TCTGCG": 1664, + "TCTGGA": 1665, + "TCTGGT": 1666, + "TCTGGC": 1667, + "TCTGGG": 1668, + "TCCAAA": 1669, + "TCCAAT": 1670, + "TCCAAC": 1671, + "TCCAAG": 1672, + "TCCATA": 1673, + "TCCATT": 1674, + "TCCATC": 1675, + "TCCATG": 1676, + "TCCACA": 1677, + "TCCACT": 1678, + "TCCACC": 1679, + "TCCACG": 1680, + "TCCAGA": 1681, + "TCCAGT": 1682, + "TCCAGC": 1683, + "TCCAGG": 1684, + "TCCTAA": 1685, + "TCCTAT": 1686, + "TCCTAC": 1687, + "TCCTAG": 1688, + "TCCTTA": 1689, + "TCCTTT": 1690, + "TCCTTC": 1691, + "TCCTTG": 1692, + "TCCTCA": 1693, + "TCCTCT": 1694, + "TCCTCC": 1695, + "TCCTCG": 1696, + "TCCTGA": 1697, + "TCCTGT": 1698, + "TCCTGC": 1699, + "TCCTGG": 1700, + "TCCCAA": 1701, + "TCCCAT": 1702, + "TCCCAC": 1703, + "TCCCAG": 1704, + "TCCCTA": 1705, + "TCCCTT": 1706, + "TCCCTC": 1707, + "TCCCTG": 1708, + "TCCCCA": 1709, + "TCCCCT": 1710, + "TCCCCC": 1711, + "TCCCCG": 1712, + "TCCCGA": 1713, + "TCCCGT": 1714, + "TCCCGC": 1715, + "TCCCGG": 1716, + "TCCGAA": 1717, + "TCCGAT": 1718, + "TCCGAC": 1719, + "TCCGAG": 1720, + "TCCGTA": 1721, + "TCCGTT": 1722, + "TCCGTC": 1723, + "TCCGTG": 1724, + "TCCGCA": 1725, + "TCCGCT": 1726, + "TCCGCC": 1727, + "TCCGCG": 1728, + "TCCGGA": 1729, + "TCCGGT": 1730, + "TCCGGC": 1731, + "TCCGGG": 1732, + "TCGAAA": 1733, + "TCGAAT": 1734, + "TCGAAC": 1735, + "TCGAAG": 1736, + "TCGATA": 1737, + "TCGATT": 1738, + "TCGATC": 1739, + "TCGATG": 1740, + "TCGACA": 1741, + "TCGACT": 1742, + "TCGACC": 1743, + "TCGACG": 1744, + "TCGAGA": 1745, + "TCGAGT": 1746, + "TCGAGC": 1747, + "TCGAGG": 1748, + "TCGTAA": 1749, + "TCGTAT": 1750, + "TCGTAC": 1751, + "TCGTAG": 1752, + "TCGTTA": 1753, + "TCGTTT": 1754, + "TCGTTC": 1755, + "TCGTTG": 1756, + "TCGTCA": 1757, + "TCGTCT": 1758, + "TCGTCC": 1759, + "TCGTCG": 1760, + "TCGTGA": 1761, + "TCGTGT": 1762, + "TCGTGC": 1763, + "TCGTGG": 1764, + "TCGCAA": 1765, + "TCGCAT": 1766, + "TCGCAC": 1767, + "TCGCAG": 1768, + "TCGCTA": 1769, + "TCGCTT": 1770, + "TCGCTC": 1771, + "TCGCTG": 1772, + "TCGCCA": 1773, + "TCGCCT": 1774, + "TCGCCC": 1775, + "TCGCCG": 1776, + "TCGCGA": 1777, + "TCGCGT": 1778, + "TCGCGC": 1779, + "TCGCGG": 1780, + "TCGGAA": 1781, + "TCGGAT": 1782, + "TCGGAC": 1783, + "TCGGAG": 1784, + "TCGGTA": 1785, + "TCGGTT": 1786, + "TCGGTC": 1787, + "TCGGTG": 1788, + "TCGGCA": 1789, + "TCGGCT": 1790, + "TCGGCC": 1791, + "TCGGCG": 1792, + "TCGGGA": 1793, + "TCGGGT": 1794, + "TCGGGC": 1795, + "TCGGGG": 1796, + "TGAAAA": 1797, + "TGAAAT": 1798, + "TGAAAC": 1799, + "TGAAAG": 1800, + "TGAATA": 1801, + "TGAATT": 1802, + "TGAATC": 1803, + "TGAATG": 1804, + "TGAACA": 1805, + "TGAACT": 1806, + "TGAACC": 1807, + "TGAACG": 1808, + "TGAAGA": 1809, + "TGAAGT": 1810, + "TGAAGC": 1811, + "TGAAGG": 1812, + "TGATAA": 1813, + "TGATAT": 1814, + "TGATAC": 1815, + "TGATAG": 1816, + "TGATTA": 1817, + "TGATTT": 1818, + "TGATTC": 1819, + "TGATTG": 1820, + "TGATCA": 1821, + "TGATCT": 1822, + "TGATCC": 1823, + "TGATCG": 1824, + "TGATGA": 1825, + "TGATGT": 1826, + "TGATGC": 1827, + "TGATGG": 1828, + "TGACAA": 1829, + "TGACAT": 1830, + "TGACAC": 1831, + "TGACAG": 1832, + "TGACTA": 1833, + "TGACTT": 1834, + "TGACTC": 1835, + "TGACTG": 1836, + "TGACCA": 1837, + "TGACCT": 1838, + "TGACCC": 1839, + "TGACCG": 1840, + "TGACGA": 1841, + "TGACGT": 1842, + "TGACGC": 1843, + "TGACGG": 1844, + "TGAGAA": 1845, + "TGAGAT": 1846, + "TGAGAC": 1847, + "TGAGAG": 1848, + "TGAGTA": 1849, + "TGAGTT": 1850, + "TGAGTC": 1851, + "TGAGTG": 1852, + "TGAGCA": 1853, + "TGAGCT": 1854, + "TGAGCC": 1855, + "TGAGCG": 1856, + "TGAGGA": 1857, + "TGAGGT": 1858, + "TGAGGC": 1859, + "TGAGGG": 1860, + "TGTAAA": 1861, + "TGTAAT": 1862, + "TGTAAC": 1863, + "TGTAAG": 1864, + "TGTATA": 1865, + "TGTATT": 1866, + "TGTATC": 1867, + "TGTATG": 1868, + "TGTACA": 1869, + "TGTACT": 1870, + "TGTACC": 1871, + "TGTACG": 1872, + "TGTAGA": 1873, + "TGTAGT": 1874, + "TGTAGC": 1875, + "TGTAGG": 1876, + "TGTTAA": 1877, + "TGTTAT": 1878, + "TGTTAC": 1879, + "TGTTAG": 1880, + "TGTTTA": 1881, + "TGTTTT": 1882, + "TGTTTC": 1883, + "TGTTTG": 1884, + "TGTTCA": 1885, + "TGTTCT": 1886, + "TGTTCC": 1887, + "TGTTCG": 1888, + "TGTTGA": 1889, + "TGTTGT": 1890, + "TGTTGC": 1891, + "TGTTGG": 1892, + "TGTCAA": 1893, + "TGTCAT": 1894, + "TGTCAC": 1895, + "TGTCAG": 1896, + "TGTCTA": 1897, + "TGTCTT": 1898, + "TGTCTC": 1899, + "TGTCTG": 1900, + "TGTCCA": 1901, + "TGTCCT": 1902, + "TGTCCC": 1903, + "TGTCCG": 1904, + "TGTCGA": 1905, + "TGTCGT": 1906, + "TGTCGC": 1907, + "TGTCGG": 1908, + "TGTGAA": 1909, + "TGTGAT": 1910, + "TGTGAC": 1911, + "TGTGAG": 1912, + "TGTGTA": 1913, + "TGTGTT": 1914, + "TGTGTC": 1915, + "TGTGTG": 1916, + "TGTGCA": 1917, + "TGTGCT": 1918, + "TGTGCC": 1919, + "TGTGCG": 1920, + "TGTGGA": 1921, + "TGTGGT": 1922, + "TGTGGC": 1923, + "TGTGGG": 1924, + "TGCAAA": 1925, + "TGCAAT": 1926, + "TGCAAC": 1927, + "TGCAAG": 1928, + "TGCATA": 1929, + "TGCATT": 1930, + "TGCATC": 1931, + "TGCATG": 1932, + "TGCACA": 1933, + "TGCACT": 1934, + "TGCACC": 1935, + "TGCACG": 1936, + "TGCAGA": 1937, + "TGCAGT": 1938, + "TGCAGC": 1939, + "TGCAGG": 1940, + "TGCTAA": 1941, + "TGCTAT": 1942, + "TGCTAC": 1943, + "TGCTAG": 1944, + "TGCTTA": 1945, + "TGCTTT": 1946, + "TGCTTC": 1947, + "TGCTTG": 1948, + "TGCTCA": 1949, + "TGCTCT": 1950, + "TGCTCC": 1951, + "TGCTCG": 1952, + "TGCTGA": 1953, + "TGCTGT": 1954, + "TGCTGC": 1955, + "TGCTGG": 1956, + "TGCCAA": 1957, + "TGCCAT": 1958, + "TGCCAC": 1959, + "TGCCAG": 1960, + "TGCCTA": 1961, + "TGCCTT": 1962, + "TGCCTC": 1963, + "TGCCTG": 1964, + "TGCCCA": 1965, + "TGCCCT": 1966, + "TGCCCC": 1967, + "TGCCCG": 1968, + "TGCCGA": 1969, + "TGCCGT": 1970, + "TGCCGC": 1971, + "TGCCGG": 1972, + "TGCGAA": 1973, + "TGCGAT": 1974, + "TGCGAC": 1975, + "TGCGAG": 1976, + "TGCGTA": 1977, + "TGCGTT": 1978, + "TGCGTC": 1979, + "TGCGTG": 1980, + "TGCGCA": 1981, + "TGCGCT": 1982, + "TGCGCC": 1983, + "TGCGCG": 1984, + "TGCGGA": 1985, + "TGCGGT": 1986, + "TGCGGC": 1987, + "TGCGGG": 1988, + "TGGAAA": 1989, + "TGGAAT": 1990, + "TGGAAC": 1991, + "TGGAAG": 1992, + "TGGATA": 1993, + "TGGATT": 1994, + "TGGATC": 1995, + "TGGATG": 1996, + "TGGACA": 1997, + "TGGACT": 1998, + "TGGACC": 1999, + "TGGACG": 2000, + "TGGAGA": 2001, + "TGGAGT": 2002, + "TGGAGC": 2003, + "TGGAGG": 2004, + "TGGTAA": 2005, + "TGGTAT": 2006, + "TGGTAC": 2007, + "TGGTAG": 2008, + "TGGTTA": 2009, + "TGGTTT": 2010, + "TGGTTC": 2011, + "TGGTTG": 2012, + "TGGTCA": 2013, + "TGGTCT": 2014, + "TGGTCC": 2015, + "TGGTCG": 2016, + "TGGTGA": 2017, + "TGGTGT": 2018, + "TGGTGC": 2019, + "TGGTGG": 2020, + "TGGCAA": 2021, + "TGGCAT": 2022, + "TGGCAC": 2023, + "TGGCAG": 2024, + "TGGCTA": 2025, + "TGGCTT": 2026, + "TGGCTC": 2027, + "TGGCTG": 2028, + "TGGCCA": 2029, + "TGGCCT": 2030, + "TGGCCC": 2031, + "TGGCCG": 2032, + "TGGCGA": 2033, + "TGGCGT": 2034, + "TGGCGC": 2035, + "TGGCGG": 2036, + "TGGGAA": 2037, + "TGGGAT": 2038, + "TGGGAC": 2039, + "TGGGAG": 2040, + "TGGGTA": 2041, + "TGGGTT": 2042, + "TGGGTC": 2043, + "TGGGTG": 2044, + "TGGGCA": 2045, + "TGGGCT": 2046, + "TGGGCC": 2047, + "TGGGCG": 2048, + "TGGGGA": 2049, + "TGGGGT": 2050, + "TGGGGC": 2051, + "TGGGGG": 2052, + "CAAAAA": 2053, + "CAAAAT": 2054, + "CAAAAC": 2055, + "CAAAAG": 2056, + "CAAATA": 2057, + "CAAATT": 2058, + "CAAATC": 2059, + "CAAATG": 2060, + "CAAACA": 2061, + "CAAACT": 2062, + "CAAACC": 2063, + "CAAACG": 2064, + "CAAAGA": 2065, + "CAAAGT": 2066, + "CAAAGC": 2067, + "CAAAGG": 2068, + "CAATAA": 2069, + "CAATAT": 2070, + "CAATAC": 2071, + "CAATAG": 2072, + "CAATTA": 2073, + "CAATTT": 2074, + "CAATTC": 2075, + "CAATTG": 2076, + "CAATCA": 2077, + "CAATCT": 2078, + "CAATCC": 2079, + "CAATCG": 2080, + "CAATGA": 2081, + "CAATGT": 2082, + "CAATGC": 2083, + "CAATGG": 2084, + "CAACAA": 2085, + "CAACAT": 2086, + "CAACAC": 2087, + "CAACAG": 2088, + "CAACTA": 2089, + "CAACTT": 2090, + "CAACTC": 2091, + "CAACTG": 2092, + "CAACCA": 2093, + "CAACCT": 2094, + "CAACCC": 2095, + "CAACCG": 2096, + "CAACGA": 2097, + "CAACGT": 2098, + "CAACGC": 2099, + "CAACGG": 2100, + "CAAGAA": 2101, + "CAAGAT": 2102, + "CAAGAC": 2103, + "CAAGAG": 2104, + "CAAGTA": 2105, + "CAAGTT": 2106, + "CAAGTC": 2107, + "CAAGTG": 2108, + "CAAGCA": 2109, + "CAAGCT": 2110, + "CAAGCC": 2111, + "CAAGCG": 2112, + "CAAGGA": 2113, + "CAAGGT": 2114, + "CAAGGC": 2115, + "CAAGGG": 2116, + "CATAAA": 2117, + "CATAAT": 2118, + "CATAAC": 2119, + "CATAAG": 2120, + "CATATA": 2121, + "CATATT": 2122, + "CATATC": 2123, + "CATATG": 2124, + "CATACA": 2125, + "CATACT": 2126, + "CATACC": 2127, + "CATACG": 2128, + "CATAGA": 2129, + "CATAGT": 2130, + "CATAGC": 2131, + "CATAGG": 2132, + "CATTAA": 2133, + "CATTAT": 2134, + "CATTAC": 2135, + "CATTAG": 2136, + "CATTTA": 2137, + "CATTTT": 2138, + "CATTTC": 2139, + "CATTTG": 2140, + "CATTCA": 2141, + "CATTCT": 2142, + "CATTCC": 2143, + "CATTCG": 2144, + "CATTGA": 2145, + "CATTGT": 2146, + "CATTGC": 2147, + "CATTGG": 2148, + "CATCAA": 2149, + "CATCAT": 2150, + "CATCAC": 2151, + "CATCAG": 2152, + "CATCTA": 2153, + "CATCTT": 2154, + "CATCTC": 2155, + "CATCTG": 2156, + "CATCCA": 2157, + "CATCCT": 2158, + "CATCCC": 2159, + "CATCCG": 2160, + "CATCGA": 2161, + "CATCGT": 2162, + "CATCGC": 2163, + "CATCGG": 2164, + "CATGAA": 2165, + "CATGAT": 2166, + "CATGAC": 2167, + "CATGAG": 2168, + "CATGTA": 2169, + "CATGTT": 2170, + "CATGTC": 2171, + "CATGTG": 2172, + "CATGCA": 2173, + "CATGCT": 2174, + "CATGCC": 2175, + "CATGCG": 2176, + "CATGGA": 2177, + "CATGGT": 2178, + "CATGGC": 2179, + "CATGGG": 2180, + "CACAAA": 2181, + "CACAAT": 2182, + "CACAAC": 2183, + "CACAAG": 2184, + "CACATA": 2185, + "CACATT": 2186, + "CACATC": 2187, + "CACATG": 2188, + "CACACA": 2189, + "CACACT": 2190, + "CACACC": 2191, + "CACACG": 2192, + "CACAGA": 2193, + "CACAGT": 2194, + "CACAGC": 2195, + "CACAGG": 2196, + "CACTAA": 2197, + "CACTAT": 2198, + "CACTAC": 2199, + "CACTAG": 2200, + "CACTTA": 2201, + "CACTTT": 2202, + "CACTTC": 2203, + "CACTTG": 2204, + "CACTCA": 2205, + "CACTCT": 2206, + "CACTCC": 2207, + "CACTCG": 2208, + "CACTGA": 2209, + "CACTGT": 2210, + "CACTGC": 2211, + "CACTGG": 2212, + "CACCAA": 2213, + "CACCAT": 2214, + "CACCAC": 2215, + "CACCAG": 2216, + "CACCTA": 2217, + "CACCTT": 2218, + "CACCTC": 2219, + "CACCTG": 2220, + "CACCCA": 2221, + "CACCCT": 2222, + "CACCCC": 2223, + "CACCCG": 2224, + "CACCGA": 2225, + "CACCGT": 2226, + "CACCGC": 2227, + "CACCGG": 2228, + "CACGAA": 2229, + "CACGAT": 2230, + "CACGAC": 2231, + "CACGAG": 2232, + "CACGTA": 2233, + "CACGTT": 2234, + "CACGTC": 2235, + "CACGTG": 2236, + "CACGCA": 2237, + "CACGCT": 2238, + "CACGCC": 2239, + "CACGCG": 2240, + "CACGGA": 2241, + "CACGGT": 2242, + "CACGGC": 2243, + "CACGGG": 2244, + "CAGAAA": 2245, + "CAGAAT": 2246, + "CAGAAC": 2247, + "CAGAAG": 2248, + "CAGATA": 2249, + "CAGATT": 2250, + "CAGATC": 2251, + "CAGATG": 2252, + "CAGACA": 2253, + "CAGACT": 2254, + "CAGACC": 2255, + "CAGACG": 2256, + "CAGAGA": 2257, + "CAGAGT": 2258, + "CAGAGC": 2259, + "CAGAGG": 2260, + "CAGTAA": 2261, + "CAGTAT": 2262, + "CAGTAC": 2263, + "CAGTAG": 2264, + "CAGTTA": 2265, + "CAGTTT": 2266, + "CAGTTC": 2267, + "CAGTTG": 2268, + "CAGTCA": 2269, + "CAGTCT": 2270, + "CAGTCC": 2271, + "CAGTCG": 2272, + "CAGTGA": 2273, + "CAGTGT": 2274, + "CAGTGC": 2275, + "CAGTGG": 2276, + "CAGCAA": 2277, + "CAGCAT": 2278, + "CAGCAC": 2279, + "CAGCAG": 2280, + "CAGCTA": 2281, + "CAGCTT": 2282, + "CAGCTC": 2283, + "CAGCTG": 2284, + "CAGCCA": 2285, + "CAGCCT": 2286, + "CAGCCC": 2287, + "CAGCCG": 2288, + "CAGCGA": 2289, + "CAGCGT": 2290, + "CAGCGC": 2291, + "CAGCGG": 2292, + "CAGGAA": 2293, + "CAGGAT": 2294, + "CAGGAC": 2295, + "CAGGAG": 2296, + "CAGGTA": 2297, + "CAGGTT": 2298, + "CAGGTC": 2299, + "CAGGTG": 2300, + "CAGGCA": 2301, + "CAGGCT": 2302, + "CAGGCC": 2303, + "CAGGCG": 2304, + "CAGGGA": 2305, + "CAGGGT": 2306, + "CAGGGC": 2307, + "CAGGGG": 2308, + "CTAAAA": 2309, + "CTAAAT": 2310, + "CTAAAC": 2311, + "CTAAAG": 2312, + "CTAATA": 2313, + "CTAATT": 2314, + "CTAATC": 2315, + "CTAATG": 2316, + "CTAACA": 2317, + "CTAACT": 2318, + "CTAACC": 2319, + "CTAACG": 2320, + "CTAAGA": 2321, + "CTAAGT": 2322, + "CTAAGC": 2323, + "CTAAGG": 2324, + "CTATAA": 2325, + "CTATAT": 2326, + "CTATAC": 2327, + "CTATAG": 2328, + "CTATTA": 2329, + "CTATTT": 2330, + "CTATTC": 2331, + "CTATTG": 2332, + "CTATCA": 2333, + "CTATCT": 2334, + "CTATCC": 2335, + "CTATCG": 2336, + "CTATGA": 2337, + "CTATGT": 2338, + "CTATGC": 2339, + "CTATGG": 2340, + "CTACAA": 2341, + "CTACAT": 2342, + "CTACAC": 2343, + "CTACAG": 2344, + "CTACTA": 2345, + "CTACTT": 2346, + "CTACTC": 2347, + "CTACTG": 2348, + "CTACCA": 2349, + "CTACCT": 2350, + "CTACCC": 2351, + "CTACCG": 2352, + "CTACGA": 2353, + "CTACGT": 2354, + "CTACGC": 2355, + "CTACGG": 2356, + "CTAGAA": 2357, + "CTAGAT": 2358, + "CTAGAC": 2359, + "CTAGAG": 2360, + "CTAGTA": 2361, + "CTAGTT": 2362, + "CTAGTC": 2363, + "CTAGTG": 2364, + "CTAGCA": 2365, + "CTAGCT": 2366, + "CTAGCC": 2367, + "CTAGCG": 2368, + "CTAGGA": 2369, + "CTAGGT": 2370, + "CTAGGC": 2371, + "CTAGGG": 2372, + "CTTAAA": 2373, + "CTTAAT": 2374, + "CTTAAC": 2375, + "CTTAAG": 2376, + "CTTATA": 2377, + "CTTATT": 2378, + "CTTATC": 2379, + "CTTATG": 2380, + "CTTACA": 2381, + "CTTACT": 2382, + "CTTACC": 2383, + "CTTACG": 2384, + "CTTAGA": 2385, + "CTTAGT": 2386, + "CTTAGC": 2387, + "CTTAGG": 2388, + "CTTTAA": 2389, + "CTTTAT": 2390, + "CTTTAC": 2391, + "CTTTAG": 2392, + "CTTTTA": 2393, + "CTTTTT": 2394, + "CTTTTC": 2395, + "CTTTTG": 2396, + "CTTTCA": 2397, + "CTTTCT": 2398, + "CTTTCC": 2399, + "CTTTCG": 2400, + "CTTTGA": 2401, + "CTTTGT": 2402, + "CTTTGC": 2403, + "CTTTGG": 2404, + "CTTCAA": 2405, + "CTTCAT": 2406, + "CTTCAC": 2407, + "CTTCAG": 2408, + "CTTCTA": 2409, + "CTTCTT": 2410, + "CTTCTC": 2411, + "CTTCTG": 2412, + "CTTCCA": 2413, + "CTTCCT": 2414, + "CTTCCC": 2415, + "CTTCCG": 2416, + "CTTCGA": 2417, + "CTTCGT": 2418, + "CTTCGC": 2419, + "CTTCGG": 2420, + "CTTGAA": 2421, + "CTTGAT": 2422, + "CTTGAC": 2423, + "CTTGAG": 2424, + "CTTGTA": 2425, + "CTTGTT": 2426, + "CTTGTC": 2427, + "CTTGTG": 2428, + "CTTGCA": 2429, + "CTTGCT": 2430, + "CTTGCC": 2431, + "CTTGCG": 2432, + "CTTGGA": 2433, + "CTTGGT": 2434, + "CTTGGC": 2435, + "CTTGGG": 2436, + "CTCAAA": 2437, + "CTCAAT": 2438, + "CTCAAC": 2439, + "CTCAAG": 2440, + "CTCATA": 2441, + "CTCATT": 2442, + "CTCATC": 2443, + "CTCATG": 2444, + "CTCACA": 2445, + "CTCACT": 2446, + "CTCACC": 2447, + "CTCACG": 2448, + "CTCAGA": 2449, + "CTCAGT": 2450, + "CTCAGC": 2451, + "CTCAGG": 2452, + "CTCTAA": 2453, + "CTCTAT": 2454, + "CTCTAC": 2455, + "CTCTAG": 2456, + "CTCTTA": 2457, + "CTCTTT": 2458, + "CTCTTC": 2459, + "CTCTTG": 2460, + "CTCTCA": 2461, + "CTCTCT": 2462, + "CTCTCC": 2463, + "CTCTCG": 2464, + "CTCTGA": 2465, + "CTCTGT": 2466, + "CTCTGC": 2467, + "CTCTGG": 2468, + "CTCCAA": 2469, + "CTCCAT": 2470, + "CTCCAC": 2471, + "CTCCAG": 2472, + "CTCCTA": 2473, + "CTCCTT": 2474, + "CTCCTC": 2475, + "CTCCTG": 2476, + "CTCCCA": 2477, + "CTCCCT": 2478, + "CTCCCC": 2479, + "CTCCCG": 2480, + "CTCCGA": 2481, + "CTCCGT": 2482, + "CTCCGC": 2483, + "CTCCGG": 2484, + "CTCGAA": 2485, + "CTCGAT": 2486, + "CTCGAC": 2487, + "CTCGAG": 2488, + "CTCGTA": 2489, + "CTCGTT": 2490, + "CTCGTC": 2491, + "CTCGTG": 2492, + "CTCGCA": 2493, + "CTCGCT": 2494, + "CTCGCC": 2495, + "CTCGCG": 2496, + "CTCGGA": 2497, + "CTCGGT": 2498, + "CTCGGC": 2499, + "CTCGGG": 2500, + "CTGAAA": 2501, + "CTGAAT": 2502, + "CTGAAC": 2503, + "CTGAAG": 2504, + "CTGATA": 2505, + "CTGATT": 2506, + "CTGATC": 2507, + "CTGATG": 2508, + "CTGACA": 2509, + "CTGACT": 2510, + "CTGACC": 2511, + "CTGACG": 2512, + "CTGAGA": 2513, + "CTGAGT": 2514, + "CTGAGC": 2515, + "CTGAGG": 2516, + "CTGTAA": 2517, + "CTGTAT": 2518, + "CTGTAC": 2519, + "CTGTAG": 2520, + "CTGTTA": 2521, + "CTGTTT": 2522, + "CTGTTC": 2523, + "CTGTTG": 2524, + "CTGTCA": 2525, + "CTGTCT": 2526, + "CTGTCC": 2527, + "CTGTCG": 2528, + "CTGTGA": 2529, + "CTGTGT": 2530, + "CTGTGC": 2531, + "CTGTGG": 2532, + "CTGCAA": 2533, + "CTGCAT": 2534, + "CTGCAC": 2535, + "CTGCAG": 2536, + "CTGCTA": 2537, + "CTGCTT": 2538, + "CTGCTC": 2539, + "CTGCTG": 2540, + "CTGCCA": 2541, + "CTGCCT": 2542, + "CTGCCC": 2543, + "CTGCCG": 2544, + "CTGCGA": 2545, + "CTGCGT": 2546, + "CTGCGC": 2547, + "CTGCGG": 2548, + "CTGGAA": 2549, + "CTGGAT": 2550, + "CTGGAC": 2551, + "CTGGAG": 2552, + "CTGGTA": 2553, + "CTGGTT": 2554, + "CTGGTC": 2555, + "CTGGTG": 2556, + "CTGGCA": 2557, + "CTGGCT": 2558, + "CTGGCC": 2559, + "CTGGCG": 2560, + "CTGGGA": 2561, + "CTGGGT": 2562, + "CTGGGC": 2563, + "CTGGGG": 2564, + "CCAAAA": 2565, + "CCAAAT": 2566, + "CCAAAC": 2567, + "CCAAAG": 2568, + "CCAATA": 2569, + "CCAATT": 2570, + "CCAATC": 2571, + "CCAATG": 2572, + "CCAACA": 2573, + "CCAACT": 2574, + "CCAACC": 2575, + "CCAACG": 2576, + "CCAAGA": 2577, + "CCAAGT": 2578, + "CCAAGC": 2579, + "CCAAGG": 2580, + "CCATAA": 2581, + "CCATAT": 2582, + "CCATAC": 2583, + "CCATAG": 2584, + "CCATTA": 2585, + "CCATTT": 2586, + "CCATTC": 2587, + "CCATTG": 2588, + "CCATCA": 2589, + "CCATCT": 2590, + "CCATCC": 2591, + "CCATCG": 2592, + "CCATGA": 2593, + "CCATGT": 2594, + "CCATGC": 2595, + "CCATGG": 2596, + "CCACAA": 2597, + "CCACAT": 2598, + "CCACAC": 2599, + "CCACAG": 2600, + "CCACTA": 2601, + "CCACTT": 2602, + "CCACTC": 2603, + "CCACTG": 2604, + "CCACCA": 2605, + "CCACCT": 2606, + "CCACCC": 2607, + "CCACCG": 2608, + "CCACGA": 2609, + "CCACGT": 2610, + "CCACGC": 2611, + "CCACGG": 2612, + "CCAGAA": 2613, + "CCAGAT": 2614, + "CCAGAC": 2615, + "CCAGAG": 2616, + "CCAGTA": 2617, + "CCAGTT": 2618, + "CCAGTC": 2619, + "CCAGTG": 2620, + "CCAGCA": 2621, + "CCAGCT": 2622, + "CCAGCC": 2623, + "CCAGCG": 2624, + "CCAGGA": 2625, + "CCAGGT": 2626, + "CCAGGC": 2627, + "CCAGGG": 2628, + "CCTAAA": 2629, + "CCTAAT": 2630, + "CCTAAC": 2631, + "CCTAAG": 2632, + "CCTATA": 2633, + "CCTATT": 2634, + "CCTATC": 2635, + "CCTATG": 2636, + "CCTACA": 2637, + "CCTACT": 2638, + "CCTACC": 2639, + "CCTACG": 2640, + "CCTAGA": 2641, + "CCTAGT": 2642, + "CCTAGC": 2643, + "CCTAGG": 2644, + "CCTTAA": 2645, + "CCTTAT": 2646, + "CCTTAC": 2647, + "CCTTAG": 2648, + "CCTTTA": 2649, + "CCTTTT": 2650, + "CCTTTC": 2651, + "CCTTTG": 2652, + "CCTTCA": 2653, + "CCTTCT": 2654, + "CCTTCC": 2655, + "CCTTCG": 2656, + "CCTTGA": 2657, + "CCTTGT": 2658, + "CCTTGC": 2659, + "CCTTGG": 2660, + "CCTCAA": 2661, + "CCTCAT": 2662, + "CCTCAC": 2663, + "CCTCAG": 2664, + "CCTCTA": 2665, + "CCTCTT": 2666, + "CCTCTC": 2667, + "CCTCTG": 2668, + "CCTCCA": 2669, + "CCTCCT": 2670, + "CCTCCC": 2671, + "CCTCCG": 2672, + "CCTCGA": 2673, + "CCTCGT": 2674, + "CCTCGC": 2675, + "CCTCGG": 2676, + "CCTGAA": 2677, + "CCTGAT": 2678, + "CCTGAC": 2679, + "CCTGAG": 2680, + "CCTGTA": 2681, + "CCTGTT": 2682, + "CCTGTC": 2683, + "CCTGTG": 2684, + "CCTGCA": 2685, + "CCTGCT": 2686, + "CCTGCC": 2687, + "CCTGCG": 2688, + "CCTGGA": 2689, + "CCTGGT": 2690, + "CCTGGC": 2691, + "CCTGGG": 2692, + "CCCAAA": 2693, + "CCCAAT": 2694, + "CCCAAC": 2695, + "CCCAAG": 2696, + "CCCATA": 2697, + "CCCATT": 2698, + "CCCATC": 2699, + "CCCATG": 2700, + "CCCACA": 2701, + "CCCACT": 2702, + "CCCACC": 2703, + "CCCACG": 2704, + "CCCAGA": 2705, + "CCCAGT": 2706, + "CCCAGC": 2707, + "CCCAGG": 2708, + "CCCTAA": 2709, + "CCCTAT": 2710, + "CCCTAC": 2711, + "CCCTAG": 2712, + "CCCTTA": 2713, + "CCCTTT": 2714, + "CCCTTC": 2715, + "CCCTTG": 2716, + "CCCTCA": 2717, + "CCCTCT": 2718, + "CCCTCC": 2719, + "CCCTCG": 2720, + "CCCTGA": 2721, + "CCCTGT": 2722, + "CCCTGC": 2723, + "CCCTGG": 2724, + "CCCCAA": 2725, + "CCCCAT": 2726, + "CCCCAC": 2727, + "CCCCAG": 2728, + "CCCCTA": 2729, + "CCCCTT": 2730, + "CCCCTC": 2731, + "CCCCTG": 2732, + "CCCCCA": 2733, + "CCCCCT": 2734, + "CCCCCC": 2735, + "CCCCCG": 2736, + "CCCCGA": 2737, + "CCCCGT": 2738, + "CCCCGC": 2739, + "CCCCGG": 2740, + "CCCGAA": 2741, + "CCCGAT": 2742, + "CCCGAC": 2743, + "CCCGAG": 2744, + "CCCGTA": 2745, + "CCCGTT": 2746, + "CCCGTC": 2747, + "CCCGTG": 2748, + "CCCGCA": 2749, + "CCCGCT": 2750, + "CCCGCC": 2751, + "CCCGCG": 2752, + "CCCGGA": 2753, + "CCCGGT": 2754, + "CCCGGC": 2755, + "CCCGGG": 2756, + "CCGAAA": 2757, + "CCGAAT": 2758, + "CCGAAC": 2759, + "CCGAAG": 2760, + "CCGATA": 2761, + "CCGATT": 2762, + "CCGATC": 2763, + "CCGATG": 2764, + "CCGACA": 2765, + "CCGACT": 2766, + "CCGACC": 2767, + "CCGACG": 2768, + "CCGAGA": 2769, + "CCGAGT": 2770, + "CCGAGC": 2771, + "CCGAGG": 2772, + "CCGTAA": 2773, + "CCGTAT": 2774, + "CCGTAC": 2775, + "CCGTAG": 2776, + "CCGTTA": 2777, + "CCGTTT": 2778, + "CCGTTC": 2779, + "CCGTTG": 2780, + "CCGTCA": 2781, + "CCGTCT": 2782, + "CCGTCC": 2783, + "CCGTCG": 2784, + "CCGTGA": 2785, + "CCGTGT": 2786, + "CCGTGC": 2787, + "CCGTGG": 2788, + "CCGCAA": 2789, + "CCGCAT": 2790, + "CCGCAC": 2791, + "CCGCAG": 2792, + "CCGCTA": 2793, + "CCGCTT": 2794, + "CCGCTC": 2795, + "CCGCTG": 2796, + "CCGCCA": 2797, + "CCGCCT": 2798, + "CCGCCC": 2799, + "CCGCCG": 2800, + "CCGCGA": 2801, + "CCGCGT": 2802, + "CCGCGC": 2803, + "CCGCGG": 2804, + "CCGGAA": 2805, + "CCGGAT": 2806, + "CCGGAC": 2807, + "CCGGAG": 2808, + "CCGGTA": 2809, + "CCGGTT": 2810, + "CCGGTC": 2811, + "CCGGTG": 2812, + "CCGGCA": 2813, + "CCGGCT": 2814, + "CCGGCC": 2815, + "CCGGCG": 2816, + "CCGGGA": 2817, + "CCGGGT": 2818, + "CCGGGC": 2819, + "CCGGGG": 2820, + "CGAAAA": 2821, + "CGAAAT": 2822, + "CGAAAC": 2823, + "CGAAAG": 2824, + "CGAATA": 2825, + "CGAATT": 2826, + "CGAATC": 2827, + "CGAATG": 2828, + "CGAACA": 2829, + "CGAACT": 2830, + "CGAACC": 2831, + "CGAACG": 2832, + "CGAAGA": 2833, + "CGAAGT": 2834, + "CGAAGC": 2835, + "CGAAGG": 2836, + "CGATAA": 2837, + "CGATAT": 2838, + "CGATAC": 2839, + "CGATAG": 2840, + "CGATTA": 2841, + "CGATTT": 2842, + "CGATTC": 2843, + "CGATTG": 2844, + "CGATCA": 2845, + "CGATCT": 2846, + "CGATCC": 2847, + "CGATCG": 2848, + "CGATGA": 2849, + "CGATGT": 2850, + "CGATGC": 2851, + "CGATGG": 2852, + "CGACAA": 2853, + "CGACAT": 2854, + "CGACAC": 2855, + "CGACAG": 2856, + "CGACTA": 2857, + "CGACTT": 2858, + "CGACTC": 2859, + "CGACTG": 2860, + "CGACCA": 2861, + "CGACCT": 2862, + "CGACCC": 2863, + "CGACCG": 2864, + "CGACGA": 2865, + "CGACGT": 2866, + "CGACGC": 2867, + "CGACGG": 2868, + "CGAGAA": 2869, + "CGAGAT": 2870, + "CGAGAC": 2871, + "CGAGAG": 2872, + "CGAGTA": 2873, + "CGAGTT": 2874, + "CGAGTC": 2875, + "CGAGTG": 2876, + "CGAGCA": 2877, + "CGAGCT": 2878, + "CGAGCC": 2879, + "CGAGCG": 2880, + "CGAGGA": 2881, + "CGAGGT": 2882, + "CGAGGC": 2883, + "CGAGGG": 2884, + "CGTAAA": 2885, + "CGTAAT": 2886, + "CGTAAC": 2887, + "CGTAAG": 2888, + "CGTATA": 2889, + "CGTATT": 2890, + "CGTATC": 2891, + "CGTATG": 2892, + "CGTACA": 2893, + "CGTACT": 2894, + "CGTACC": 2895, + "CGTACG": 2896, + "CGTAGA": 2897, + "CGTAGT": 2898, + "CGTAGC": 2899, + "CGTAGG": 2900, + "CGTTAA": 2901, + "CGTTAT": 2902, + "CGTTAC": 2903, + "CGTTAG": 2904, + "CGTTTA": 2905, + "CGTTTT": 2906, + "CGTTTC": 2907, + "CGTTTG": 2908, + "CGTTCA": 2909, + "CGTTCT": 2910, + "CGTTCC": 2911, + "CGTTCG": 2912, + "CGTTGA": 2913, + "CGTTGT": 2914, + "CGTTGC": 2915, + "CGTTGG": 2916, + "CGTCAA": 2917, + "CGTCAT": 2918, + "CGTCAC": 2919, + "CGTCAG": 2920, + "CGTCTA": 2921, + "CGTCTT": 2922, + "CGTCTC": 2923, + "CGTCTG": 2924, + "CGTCCA": 2925, + "CGTCCT": 2926, + "CGTCCC": 2927, + "CGTCCG": 2928, + "CGTCGA": 2929, + "CGTCGT": 2930, + "CGTCGC": 2931, + "CGTCGG": 2932, + "CGTGAA": 2933, + "CGTGAT": 2934, + "CGTGAC": 2935, + "CGTGAG": 2936, + "CGTGTA": 2937, + "CGTGTT": 2938, + "CGTGTC": 2939, + "CGTGTG": 2940, + "CGTGCA": 2941, + "CGTGCT": 2942, + "CGTGCC": 2943, + "CGTGCG": 2944, + "CGTGGA": 2945, + "CGTGGT": 2946, + "CGTGGC": 2947, + "CGTGGG": 2948, + "CGCAAA": 2949, + "CGCAAT": 2950, + "CGCAAC": 2951, + "CGCAAG": 2952, + "CGCATA": 2953, + "CGCATT": 2954, + "CGCATC": 2955, + "CGCATG": 2956, + "CGCACA": 2957, + "CGCACT": 2958, + "CGCACC": 2959, + "CGCACG": 2960, + "CGCAGA": 2961, + "CGCAGT": 2962, + "CGCAGC": 2963, + "CGCAGG": 2964, + "CGCTAA": 2965, + "CGCTAT": 2966, + "CGCTAC": 2967, + "CGCTAG": 2968, + "CGCTTA": 2969, + "CGCTTT": 2970, + "CGCTTC": 2971, + "CGCTTG": 2972, + "CGCTCA": 2973, + "CGCTCT": 2974, + "CGCTCC": 2975, + "CGCTCG": 2976, + "CGCTGA": 2977, + "CGCTGT": 2978, + "CGCTGC": 2979, + "CGCTGG": 2980, + "CGCCAA": 2981, + "CGCCAT": 2982, + "CGCCAC": 2983, + "CGCCAG": 2984, + "CGCCTA": 2985, + "CGCCTT": 2986, + "CGCCTC": 2987, + "CGCCTG": 2988, + "CGCCCA": 2989, + "CGCCCT": 2990, + "CGCCCC": 2991, + "CGCCCG": 2992, + "CGCCGA": 2993, + "CGCCGT": 2994, + "CGCCGC": 2995, + "CGCCGG": 2996, + "CGCGAA": 2997, + "CGCGAT": 2998, + "CGCGAC": 2999, + "CGCGAG": 3000, + "CGCGTA": 3001, + "CGCGTT": 3002, + "CGCGTC": 3003, + "CGCGTG": 3004, + "CGCGCA": 3005, + "CGCGCT": 3006, + "CGCGCC": 3007, + "CGCGCG": 3008, + "CGCGGA": 3009, + "CGCGGT": 3010, + "CGCGGC": 3011, + "CGCGGG": 3012, + "CGGAAA": 3013, + "CGGAAT": 3014, + "CGGAAC": 3015, + "CGGAAG": 3016, + "CGGATA": 3017, + "CGGATT": 3018, + "CGGATC": 3019, + "CGGATG": 3020, + "CGGACA": 3021, + "CGGACT": 3022, + "CGGACC": 3023, + "CGGACG": 3024, + "CGGAGA": 3025, + "CGGAGT": 3026, + "CGGAGC": 3027, + "CGGAGG": 3028, + "CGGTAA": 3029, + "CGGTAT": 3030, + "CGGTAC": 3031, + "CGGTAG": 3032, + "CGGTTA": 3033, + "CGGTTT": 3034, + "CGGTTC": 3035, + "CGGTTG": 3036, + "CGGTCA": 3037, + "CGGTCT": 3038, + "CGGTCC": 3039, + "CGGTCG": 3040, + "CGGTGA": 3041, + "CGGTGT": 3042, + "CGGTGC": 3043, + "CGGTGG": 3044, + "CGGCAA": 3045, + "CGGCAT": 3046, + "CGGCAC": 3047, + "CGGCAG": 3048, + "CGGCTA": 3049, + "CGGCTT": 3050, + "CGGCTC": 3051, + "CGGCTG": 3052, + "CGGCCA": 3053, + "CGGCCT": 3054, + "CGGCCC": 3055, + "CGGCCG": 3056, + "CGGCGA": 3057, + "CGGCGT": 3058, + "CGGCGC": 3059, + "CGGCGG": 3060, + "CGGGAA": 3061, + "CGGGAT": 3062, + "CGGGAC": 3063, + "CGGGAG": 3064, + "CGGGTA": 3065, + "CGGGTT": 3066, + "CGGGTC": 3067, + "CGGGTG": 3068, + "CGGGCA": 3069, + "CGGGCT": 3070, + "CGGGCC": 3071, + "CGGGCG": 3072, + "CGGGGA": 3073, + "CGGGGT": 3074, + "CGGGGC": 3075, + "CGGGGG": 3076, + "GAAAAA": 3077, + "GAAAAT": 3078, + "GAAAAC": 3079, + "GAAAAG": 3080, + "GAAATA": 3081, + "GAAATT": 3082, + "GAAATC": 3083, + "GAAATG": 3084, + "GAAACA": 3085, + "GAAACT": 3086, + "GAAACC": 3087, + "GAAACG": 3088, + "GAAAGA": 3089, + "GAAAGT": 3090, + "GAAAGC": 3091, + "GAAAGG": 3092, + "GAATAA": 3093, + "GAATAT": 3094, + "GAATAC": 3095, + "GAATAG": 3096, + "GAATTA": 3097, + "GAATTT": 3098, + "GAATTC": 3099, + "GAATTG": 3100, + "GAATCA": 3101, + "GAATCT": 3102, + "GAATCC": 3103, + "GAATCG": 3104, + "GAATGA": 3105, + "GAATGT": 3106, + "GAATGC": 3107, + "GAATGG": 3108, + "GAACAA": 3109, + "GAACAT": 3110, + "GAACAC": 3111, + "GAACAG": 3112, + "GAACTA": 3113, + "GAACTT": 3114, + "GAACTC": 3115, + "GAACTG": 3116, + "GAACCA": 3117, + "GAACCT": 3118, + "GAACCC": 3119, + "GAACCG": 3120, + "GAACGA": 3121, + "GAACGT": 3122, + "GAACGC": 3123, + "GAACGG": 3124, + "GAAGAA": 3125, + "GAAGAT": 3126, + "GAAGAC": 3127, + "GAAGAG": 3128, + "GAAGTA": 3129, + "GAAGTT": 3130, + "GAAGTC": 3131, + "GAAGTG": 3132, + "GAAGCA": 3133, + "GAAGCT": 3134, + "GAAGCC": 3135, + "GAAGCG": 3136, + "GAAGGA": 3137, + "GAAGGT": 3138, + "GAAGGC": 3139, + "GAAGGG": 3140, + "GATAAA": 3141, + "GATAAT": 3142, + "GATAAC": 3143, + "GATAAG": 3144, + "GATATA": 3145, + "GATATT": 3146, + "GATATC": 3147, + "GATATG": 3148, + "GATACA": 3149, + "GATACT": 3150, + "GATACC": 3151, + "GATACG": 3152, + "GATAGA": 3153, + "GATAGT": 3154, + "GATAGC": 3155, + "GATAGG": 3156, + "GATTAA": 3157, + "GATTAT": 3158, + "GATTAC": 3159, + "GATTAG": 3160, + "GATTTA": 3161, + "GATTTT": 3162, + "GATTTC": 3163, + "GATTTG": 3164, + "GATTCA": 3165, + "GATTCT": 3166, + "GATTCC": 3167, + "GATTCG": 3168, + "GATTGA": 3169, + "GATTGT": 3170, + "GATTGC": 3171, + "GATTGG": 3172, + "GATCAA": 3173, + "GATCAT": 3174, + "GATCAC": 3175, + "GATCAG": 3176, + "GATCTA": 3177, + "GATCTT": 3178, + "GATCTC": 3179, + "GATCTG": 3180, + "GATCCA": 3181, + "GATCCT": 3182, + "GATCCC": 3183, + "GATCCG": 3184, + "GATCGA": 3185, + "GATCGT": 3186, + "GATCGC": 3187, + "GATCGG": 3188, + "GATGAA": 3189, + "GATGAT": 3190, + "GATGAC": 3191, + "GATGAG": 3192, + "GATGTA": 3193, + "GATGTT": 3194, + "GATGTC": 3195, + "GATGTG": 3196, + "GATGCA": 3197, + "GATGCT": 3198, + "GATGCC": 3199, + "GATGCG": 3200, + "GATGGA": 3201, + "GATGGT": 3202, + "GATGGC": 3203, + "GATGGG": 3204, + "GACAAA": 3205, + "GACAAT": 3206, + "GACAAC": 3207, + "GACAAG": 3208, + "GACATA": 3209, + "GACATT": 3210, + "GACATC": 3211, + "GACATG": 3212, + "GACACA": 3213, + "GACACT": 3214, + "GACACC": 3215, + "GACACG": 3216, + "GACAGA": 3217, + "GACAGT": 3218, + "GACAGC": 3219, + "GACAGG": 3220, + "GACTAA": 3221, + "GACTAT": 3222, + "GACTAC": 3223, + "GACTAG": 3224, + "GACTTA": 3225, + "GACTTT": 3226, + "GACTTC": 3227, + "GACTTG": 3228, + "GACTCA": 3229, + "GACTCT": 3230, + "GACTCC": 3231, + "GACTCG": 3232, + "GACTGA": 3233, + "GACTGT": 3234, + "GACTGC": 3235, + "GACTGG": 3236, + "GACCAA": 3237, + "GACCAT": 3238, + "GACCAC": 3239, + "GACCAG": 3240, + "GACCTA": 3241, + "GACCTT": 3242, + "GACCTC": 3243, + "GACCTG": 3244, + "GACCCA": 3245, + "GACCCT": 3246, + "GACCCC": 3247, + "GACCCG": 3248, + "GACCGA": 3249, + "GACCGT": 3250, + "GACCGC": 3251, + "GACCGG": 3252, + "GACGAA": 3253, + "GACGAT": 3254, + "GACGAC": 3255, + "GACGAG": 3256, + "GACGTA": 3257, + "GACGTT": 3258, + "GACGTC": 3259, + "GACGTG": 3260, + "GACGCA": 3261, + "GACGCT": 3262, + "GACGCC": 3263, + "GACGCG": 3264, + "GACGGA": 3265, + "GACGGT": 3266, + "GACGGC": 3267, + "GACGGG": 3268, + "GAGAAA": 3269, + "GAGAAT": 3270, + "GAGAAC": 3271, + "GAGAAG": 3272, + "GAGATA": 3273, + "GAGATT": 3274, + "GAGATC": 3275, + "GAGATG": 3276, + "GAGACA": 3277, + "GAGACT": 3278, + "GAGACC": 3279, + "GAGACG": 3280, + "GAGAGA": 3281, + "GAGAGT": 3282, + "GAGAGC": 3283, + "GAGAGG": 3284, + "GAGTAA": 3285, + "GAGTAT": 3286, + "GAGTAC": 3287, + "GAGTAG": 3288, + "GAGTTA": 3289, + "GAGTTT": 3290, + "GAGTTC": 3291, + "GAGTTG": 3292, + "GAGTCA": 3293, + "GAGTCT": 3294, + "GAGTCC": 3295, + "GAGTCG": 3296, + "GAGTGA": 3297, + "GAGTGT": 3298, + "GAGTGC": 3299, + "GAGTGG": 3300, + "GAGCAA": 3301, + "GAGCAT": 3302, + "GAGCAC": 3303, + "GAGCAG": 3304, + "GAGCTA": 3305, + "GAGCTT": 3306, + "GAGCTC": 3307, + "GAGCTG": 3308, + "GAGCCA": 3309, + "GAGCCT": 3310, + "GAGCCC": 3311, + "GAGCCG": 3312, + "GAGCGA": 3313, + "GAGCGT": 3314, + "GAGCGC": 3315, + "GAGCGG": 3316, + "GAGGAA": 3317, + "GAGGAT": 3318, + "GAGGAC": 3319, + "GAGGAG": 3320, + "GAGGTA": 3321, + "GAGGTT": 3322, + "GAGGTC": 3323, + "GAGGTG": 3324, + "GAGGCA": 3325, + "GAGGCT": 3326, + "GAGGCC": 3327, + "GAGGCG": 3328, + "GAGGGA": 3329, + "GAGGGT": 3330, + "GAGGGC": 3331, + "GAGGGG": 3332, + "GTAAAA": 3333, + "GTAAAT": 3334, + "GTAAAC": 3335, + "GTAAAG": 3336, + "GTAATA": 3337, + "GTAATT": 3338, + "GTAATC": 3339, + "GTAATG": 3340, + "GTAACA": 3341, + "GTAACT": 3342, + "GTAACC": 3343, + "GTAACG": 3344, + "GTAAGA": 3345, + "GTAAGT": 3346, + "GTAAGC": 3347, + "GTAAGG": 3348, + "GTATAA": 3349, + "GTATAT": 3350, + "GTATAC": 3351, + "GTATAG": 3352, + "GTATTA": 3353, + "GTATTT": 3354, + "GTATTC": 3355, + "GTATTG": 3356, + "GTATCA": 3357, + "GTATCT": 3358, + "GTATCC": 3359, + "GTATCG": 3360, + "GTATGA": 3361, + "GTATGT": 3362, + "GTATGC": 3363, + "GTATGG": 3364, + "GTACAA": 3365, + "GTACAT": 3366, + "GTACAC": 3367, + "GTACAG": 3368, + "GTACTA": 3369, + "GTACTT": 3370, + "GTACTC": 3371, + "GTACTG": 3372, + "GTACCA": 3373, + "GTACCT": 3374, + "GTACCC": 3375, + "GTACCG": 3376, + "GTACGA": 3377, + "GTACGT": 3378, + "GTACGC": 3379, + "GTACGG": 3380, + "GTAGAA": 3381, + "GTAGAT": 3382, + "GTAGAC": 3383, + "GTAGAG": 3384, + "GTAGTA": 3385, + "GTAGTT": 3386, + "GTAGTC": 3387, + "GTAGTG": 3388, + "GTAGCA": 3389, + "GTAGCT": 3390, + "GTAGCC": 3391, + "GTAGCG": 3392, + "GTAGGA": 3393, + "GTAGGT": 3394, + "GTAGGC": 3395, + "GTAGGG": 3396, + "GTTAAA": 3397, + "GTTAAT": 3398, + "GTTAAC": 3399, + "GTTAAG": 3400, + "GTTATA": 3401, + "GTTATT": 3402, + "GTTATC": 3403, + "GTTATG": 3404, + "GTTACA": 3405, + "GTTACT": 3406, + "GTTACC": 3407, + "GTTACG": 3408, + "GTTAGA": 3409, + "GTTAGT": 3410, + "GTTAGC": 3411, + "GTTAGG": 3412, + "GTTTAA": 3413, + "GTTTAT": 3414, + "GTTTAC": 3415, + "GTTTAG": 3416, + "GTTTTA": 3417, + "GTTTTT": 3418, + "GTTTTC": 3419, + "GTTTTG": 3420, + "GTTTCA": 3421, + "GTTTCT": 3422, + "GTTTCC": 3423, + "GTTTCG": 3424, + "GTTTGA": 3425, + "GTTTGT": 3426, + "GTTTGC": 3427, + "GTTTGG": 3428, + "GTTCAA": 3429, + "GTTCAT": 3430, + "GTTCAC": 3431, + "GTTCAG": 3432, + "GTTCTA": 3433, + "GTTCTT": 3434, + "GTTCTC": 3435, + "GTTCTG": 3436, + "GTTCCA": 3437, + "GTTCCT": 3438, + "GTTCCC": 3439, + "GTTCCG": 3440, + "GTTCGA": 3441, + "GTTCGT": 3442, + "GTTCGC": 3443, + "GTTCGG": 3444, + "GTTGAA": 3445, + "GTTGAT": 3446, + "GTTGAC": 3447, + "GTTGAG": 3448, + "GTTGTA": 3449, + "GTTGTT": 3450, + "GTTGTC": 3451, + "GTTGTG": 3452, + "GTTGCA": 3453, + "GTTGCT": 3454, + "GTTGCC": 3455, + "GTTGCG": 3456, + "GTTGGA": 3457, + "GTTGGT": 3458, + "GTTGGC": 3459, + "GTTGGG": 3460, + "GTCAAA": 3461, + "GTCAAT": 3462, + "GTCAAC": 3463, + "GTCAAG": 3464, + "GTCATA": 3465, + "GTCATT": 3466, + "GTCATC": 3467, + "GTCATG": 3468, + "GTCACA": 3469, + "GTCACT": 3470, + "GTCACC": 3471, + "GTCACG": 3472, + "GTCAGA": 3473, + "GTCAGT": 3474, + "GTCAGC": 3475, + "GTCAGG": 3476, + "GTCTAA": 3477, + "GTCTAT": 3478, + "GTCTAC": 3479, + "GTCTAG": 3480, + "GTCTTA": 3481, + "GTCTTT": 3482, + "GTCTTC": 3483, + "GTCTTG": 3484, + "GTCTCA": 3485, + "GTCTCT": 3486, + "GTCTCC": 3487, + "GTCTCG": 3488, + "GTCTGA": 3489, + "GTCTGT": 3490, + "GTCTGC": 3491, + "GTCTGG": 3492, + "GTCCAA": 3493, + "GTCCAT": 3494, + "GTCCAC": 3495, + "GTCCAG": 3496, + "GTCCTA": 3497, + "GTCCTT": 3498, + "GTCCTC": 3499, + "GTCCTG": 3500, + "GTCCCA": 3501, + "GTCCCT": 3502, + "GTCCCC": 3503, + "GTCCCG": 3504, + "GTCCGA": 3505, + "GTCCGT": 3506, + "GTCCGC": 3507, + "GTCCGG": 3508, + "GTCGAA": 3509, + "GTCGAT": 3510, + "GTCGAC": 3511, + "GTCGAG": 3512, + "GTCGTA": 3513, + "GTCGTT": 3514, + "GTCGTC": 3515, + "GTCGTG": 3516, + "GTCGCA": 3517, + "GTCGCT": 3518, + "GTCGCC": 3519, + "GTCGCG": 3520, + "GTCGGA": 3521, + "GTCGGT": 3522, + "GTCGGC": 3523, + "GTCGGG": 3524, + "GTGAAA": 3525, + "GTGAAT": 3526, + "GTGAAC": 3527, + "GTGAAG": 3528, + "GTGATA": 3529, + "GTGATT": 3530, + "GTGATC": 3531, + "GTGATG": 3532, + "GTGACA": 3533, + "GTGACT": 3534, + "GTGACC": 3535, + "GTGACG": 3536, + "GTGAGA": 3537, + "GTGAGT": 3538, + "GTGAGC": 3539, + "GTGAGG": 3540, + "GTGTAA": 3541, + "GTGTAT": 3542, + "GTGTAC": 3543, + "GTGTAG": 3544, + "GTGTTA": 3545, + "GTGTTT": 3546, + "GTGTTC": 3547, + "GTGTTG": 3548, + "GTGTCA": 3549, + "GTGTCT": 3550, + "GTGTCC": 3551, + "GTGTCG": 3552, + "GTGTGA": 3553, + "GTGTGT": 3554, + "GTGTGC": 3555, + "GTGTGG": 3556, + "GTGCAA": 3557, + "GTGCAT": 3558, + "GTGCAC": 3559, + "GTGCAG": 3560, + "GTGCTA": 3561, + "GTGCTT": 3562, + "GTGCTC": 3563, + "GTGCTG": 3564, + "GTGCCA": 3565, + "GTGCCT": 3566, + "GTGCCC": 3567, + "GTGCCG": 3568, + "GTGCGA": 3569, + "GTGCGT": 3570, + "GTGCGC": 3571, + "GTGCGG": 3572, + "GTGGAA": 3573, + "GTGGAT": 3574, + "GTGGAC": 3575, + "GTGGAG": 3576, + "GTGGTA": 3577, + "GTGGTT": 3578, + "GTGGTC": 3579, + "GTGGTG": 3580, + "GTGGCA": 3581, + "GTGGCT": 3582, + "GTGGCC": 3583, + "GTGGCG": 3584, + "GTGGGA": 3585, + "GTGGGT": 3586, + "GTGGGC": 3587, + "GTGGGG": 3588, + "GCAAAA": 3589, + "GCAAAT": 3590, + "GCAAAC": 3591, + "GCAAAG": 3592, + "GCAATA": 3593, + "GCAATT": 3594, + "GCAATC": 3595, + "GCAATG": 3596, + "GCAACA": 3597, + "GCAACT": 3598, + "GCAACC": 3599, + "GCAACG": 3600, + "GCAAGA": 3601, + "GCAAGT": 3602, + "GCAAGC": 3603, + "GCAAGG": 3604, + "GCATAA": 3605, + "GCATAT": 3606, + "GCATAC": 3607, + "GCATAG": 3608, + "GCATTA": 3609, + "GCATTT": 3610, + "GCATTC": 3611, + "GCATTG": 3612, + "GCATCA": 3613, + "GCATCT": 3614, + "GCATCC": 3615, + "GCATCG": 3616, + "GCATGA": 3617, + "GCATGT": 3618, + "GCATGC": 3619, + "GCATGG": 3620, + "GCACAA": 3621, + "GCACAT": 3622, + "GCACAC": 3623, + "GCACAG": 3624, + "GCACTA": 3625, + "GCACTT": 3626, + "GCACTC": 3627, + "GCACTG": 3628, + "GCACCA": 3629, + "GCACCT": 3630, + "GCACCC": 3631, + "GCACCG": 3632, + "GCACGA": 3633, + "GCACGT": 3634, + "GCACGC": 3635, + "GCACGG": 3636, + "GCAGAA": 3637, + "GCAGAT": 3638, + "GCAGAC": 3639, + "GCAGAG": 3640, + "GCAGTA": 3641, + "GCAGTT": 3642, + "GCAGTC": 3643, + "GCAGTG": 3644, + "GCAGCA": 3645, + "GCAGCT": 3646, + "GCAGCC": 3647, + "GCAGCG": 3648, + "GCAGGA": 3649, + "GCAGGT": 3650, + "GCAGGC": 3651, + "GCAGGG": 3652, + "GCTAAA": 3653, + "GCTAAT": 3654, + "GCTAAC": 3655, + "GCTAAG": 3656, + "GCTATA": 3657, + "GCTATT": 3658, + "GCTATC": 3659, + "GCTATG": 3660, + "GCTACA": 3661, + "GCTACT": 3662, + "GCTACC": 3663, + "GCTACG": 3664, + "GCTAGA": 3665, + "GCTAGT": 3666, + "GCTAGC": 3667, + "GCTAGG": 3668, + "GCTTAA": 3669, + "GCTTAT": 3670, + "GCTTAC": 3671, + "GCTTAG": 3672, + "GCTTTA": 3673, + "GCTTTT": 3674, + "GCTTTC": 3675, + "GCTTTG": 3676, + "GCTTCA": 3677, + "GCTTCT": 3678, + "GCTTCC": 3679, + "GCTTCG": 3680, + "GCTTGA": 3681, + "GCTTGT": 3682, + "GCTTGC": 3683, + "GCTTGG": 3684, + "GCTCAA": 3685, + "GCTCAT": 3686, + "GCTCAC": 3687, + "GCTCAG": 3688, + "GCTCTA": 3689, + "GCTCTT": 3690, + "GCTCTC": 3691, + "GCTCTG": 3692, + "GCTCCA": 3693, + "GCTCCT": 3694, + "GCTCCC": 3695, + "GCTCCG": 3696, + "GCTCGA": 3697, + "GCTCGT": 3698, + "GCTCGC": 3699, + "GCTCGG": 3700, + "GCTGAA": 3701, + "GCTGAT": 3702, + "GCTGAC": 3703, + "GCTGAG": 3704, + "GCTGTA": 3705, + "GCTGTT": 3706, + "GCTGTC": 3707, + "GCTGTG": 3708, + "GCTGCA": 3709, + "GCTGCT": 3710, + "GCTGCC": 3711, + "GCTGCG": 3712, + "GCTGGA": 3713, + "GCTGGT": 3714, + "GCTGGC": 3715, + "GCTGGG": 3716, + "GCCAAA": 3717, + "GCCAAT": 3718, + "GCCAAC": 3719, + "GCCAAG": 3720, + "GCCATA": 3721, + "GCCATT": 3722, + "GCCATC": 3723, + "GCCATG": 3724, + "GCCACA": 3725, + "GCCACT": 3726, + "GCCACC": 3727, + "GCCACG": 3728, + "GCCAGA": 3729, + "GCCAGT": 3730, + "GCCAGC": 3731, + "GCCAGG": 3732, + "GCCTAA": 3733, + "GCCTAT": 3734, + "GCCTAC": 3735, + "GCCTAG": 3736, + "GCCTTA": 3737, + "GCCTTT": 3738, + "GCCTTC": 3739, + "GCCTTG": 3740, + "GCCTCA": 3741, + "GCCTCT": 3742, + "GCCTCC": 3743, + "GCCTCG": 3744, + "GCCTGA": 3745, + "GCCTGT": 3746, + "GCCTGC": 3747, + "GCCTGG": 3748, + "GCCCAA": 3749, + "GCCCAT": 3750, + "GCCCAC": 3751, + "GCCCAG": 3752, + "GCCCTA": 3753, + "GCCCTT": 3754, + "GCCCTC": 3755, + "GCCCTG": 3756, + "GCCCCA": 3757, + "GCCCCT": 3758, + "GCCCCC": 3759, + "GCCCCG": 3760, + "GCCCGA": 3761, + "GCCCGT": 3762, + "GCCCGC": 3763, + "GCCCGG": 3764, + "GCCGAA": 3765, + "GCCGAT": 3766, + "GCCGAC": 3767, + "GCCGAG": 3768, + "GCCGTA": 3769, + "GCCGTT": 3770, + "GCCGTC": 3771, + "GCCGTG": 3772, + "GCCGCA": 3773, + "GCCGCT": 3774, + "GCCGCC": 3775, + "GCCGCG": 3776, + "GCCGGA": 3777, + "GCCGGT": 3778, + "GCCGGC": 3779, + "GCCGGG": 3780, + "GCGAAA": 3781, + "GCGAAT": 3782, + "GCGAAC": 3783, + "GCGAAG": 3784, + "GCGATA": 3785, + "GCGATT": 3786, + "GCGATC": 3787, + "GCGATG": 3788, + "GCGACA": 3789, + "GCGACT": 3790, + "GCGACC": 3791, + "GCGACG": 3792, + "GCGAGA": 3793, + "GCGAGT": 3794, + "GCGAGC": 3795, + "GCGAGG": 3796, + "GCGTAA": 3797, + "GCGTAT": 3798, + "GCGTAC": 3799, + "GCGTAG": 3800, + "GCGTTA": 3801, + "GCGTTT": 3802, + "GCGTTC": 3803, + "GCGTTG": 3804, + "GCGTCA": 3805, + "GCGTCT": 3806, + "GCGTCC": 3807, + "GCGTCG": 3808, + "GCGTGA": 3809, + "GCGTGT": 3810, + "GCGTGC": 3811, + "GCGTGG": 3812, + "GCGCAA": 3813, + "GCGCAT": 3814, + "GCGCAC": 3815, + "GCGCAG": 3816, + "GCGCTA": 3817, + "GCGCTT": 3818, + "GCGCTC": 3819, + "GCGCTG": 3820, + "GCGCCA": 3821, + "GCGCCT": 3822, + "GCGCCC": 3823, + "GCGCCG": 3824, + "GCGCGA": 3825, + "GCGCGT": 3826, + "GCGCGC": 3827, + "GCGCGG": 3828, + "GCGGAA": 3829, + "GCGGAT": 3830, + "GCGGAC": 3831, + "GCGGAG": 3832, + "GCGGTA": 3833, + "GCGGTT": 3834, + "GCGGTC": 3835, + "GCGGTG": 3836, + "GCGGCA": 3837, + "GCGGCT": 3838, + "GCGGCC": 3839, + "GCGGCG": 3840, + "GCGGGA": 3841, + "GCGGGT": 3842, + "GCGGGC": 3843, + "GCGGGG": 3844, + "GGAAAA": 3845, + "GGAAAT": 3846, + "GGAAAC": 3847, + "GGAAAG": 3848, + "GGAATA": 3849, + "GGAATT": 3850, + "GGAATC": 3851, + "GGAATG": 3852, + "GGAACA": 3853, + "GGAACT": 3854, + "GGAACC": 3855, + "GGAACG": 3856, + "GGAAGA": 3857, + "GGAAGT": 3858, + "GGAAGC": 3859, + "GGAAGG": 3860, + "GGATAA": 3861, + "GGATAT": 3862, + "GGATAC": 3863, + "GGATAG": 3864, + "GGATTA": 3865, + "GGATTT": 3866, + "GGATTC": 3867, + "GGATTG": 3868, + "GGATCA": 3869, + "GGATCT": 3870, + "GGATCC": 3871, + "GGATCG": 3872, + "GGATGA": 3873, + "GGATGT": 3874, + "GGATGC": 3875, + "GGATGG": 3876, + "GGACAA": 3877, + "GGACAT": 3878, + "GGACAC": 3879, + "GGACAG": 3880, + "GGACTA": 3881, + "GGACTT": 3882, + "GGACTC": 3883, + "GGACTG": 3884, + "GGACCA": 3885, + "GGACCT": 3886, + "GGACCC": 3887, + "GGACCG": 3888, + "GGACGA": 3889, + "GGACGT": 3890, + "GGACGC": 3891, + "GGACGG": 3892, + "GGAGAA": 3893, + "GGAGAT": 3894, + "GGAGAC": 3895, + "GGAGAG": 3896, + "GGAGTA": 3897, + "GGAGTT": 3898, + "GGAGTC": 3899, + "GGAGTG": 3900, + "GGAGCA": 3901, + "GGAGCT": 3902, + "GGAGCC": 3903, + "GGAGCG": 3904, + "GGAGGA": 3905, + "GGAGGT": 3906, + "GGAGGC": 3907, + "GGAGGG": 3908, + "GGTAAA": 3909, + "GGTAAT": 3910, + "GGTAAC": 3911, + "GGTAAG": 3912, + "GGTATA": 3913, + "GGTATT": 3914, + "GGTATC": 3915, + "GGTATG": 3916, + "GGTACA": 3917, + "GGTACT": 3918, + "GGTACC": 3919, + "GGTACG": 3920, + "GGTAGA": 3921, + "GGTAGT": 3922, + "GGTAGC": 3923, + "GGTAGG": 3924, + "GGTTAA": 3925, + "GGTTAT": 3926, + "GGTTAC": 3927, + "GGTTAG": 3928, + "GGTTTA": 3929, + "GGTTTT": 3930, + "GGTTTC": 3931, + "GGTTTG": 3932, + "GGTTCA": 3933, + "GGTTCT": 3934, + "GGTTCC": 3935, + "GGTTCG": 3936, + "GGTTGA": 3937, + "GGTTGT": 3938, + "GGTTGC": 3939, + "GGTTGG": 3940, + "GGTCAA": 3941, + "GGTCAT": 3942, + "GGTCAC": 3943, + "GGTCAG": 3944, + "GGTCTA": 3945, + "GGTCTT": 3946, + "GGTCTC": 3947, + "GGTCTG": 3948, + "GGTCCA": 3949, + "GGTCCT": 3950, + "GGTCCC": 3951, + "GGTCCG": 3952, + "GGTCGA": 3953, + "GGTCGT": 3954, + "GGTCGC": 3955, + "GGTCGG": 3956, + "GGTGAA": 3957, + "GGTGAT": 3958, + "GGTGAC": 3959, + "GGTGAG": 3960, + "GGTGTA": 3961, + "GGTGTT": 3962, + "GGTGTC": 3963, + "GGTGTG": 3964, + "GGTGCA": 3965, + "GGTGCT": 3966, + "GGTGCC": 3967, + "GGTGCG": 3968, + "GGTGGA": 3969, + "GGTGGT": 3970, + "GGTGGC": 3971, + "GGTGGG": 3972, + "GGCAAA": 3973, + "GGCAAT": 3974, + "GGCAAC": 3975, + "GGCAAG": 3976, + "GGCATA": 3977, + "GGCATT": 3978, + "GGCATC": 3979, + "GGCATG": 3980, + "GGCACA": 3981, + "GGCACT": 3982, + "GGCACC": 3983, + "GGCACG": 3984, + "GGCAGA": 3985, + "GGCAGT": 3986, + "GGCAGC": 3987, + "GGCAGG": 3988, + "GGCTAA": 3989, + "GGCTAT": 3990, + "GGCTAC": 3991, + "GGCTAG": 3992, + "GGCTTA": 3993, + "GGCTTT": 3994, + "GGCTTC": 3995, + "GGCTTG": 3996, + "GGCTCA": 3997, + "GGCTCT": 3998, + "GGCTCC": 3999, + "GGCTCG": 4000, + "GGCTGA": 4001, + "GGCTGT": 4002, + "GGCTGC": 4003, + "GGCTGG": 4004, + "GGCCAA": 4005, + "GGCCAT": 4006, + "GGCCAC": 4007, + "GGCCAG": 4008, + "GGCCTA": 4009, + "GGCCTT": 4010, + "GGCCTC": 4011, + "GGCCTG": 4012, + "GGCCCA": 4013, + "GGCCCT": 4014, + "GGCCCC": 4015, + "GGCCCG": 4016, + "GGCCGA": 4017, + "GGCCGT": 4018, + "GGCCGC": 4019, + "GGCCGG": 4020, + "GGCGAA": 4021, + "GGCGAT": 4022, + "GGCGAC": 4023, + "GGCGAG": 4024, + "GGCGTA": 4025, + "GGCGTT": 4026, + "GGCGTC": 4027, + "GGCGTG": 4028, + "GGCGCA": 4029, + "GGCGCT": 4030, + "GGCGCC": 4031, + "GGCGCG": 4032, + "GGCGGA": 4033, + "GGCGGT": 4034, + "GGCGGC": 4035, + "GGCGGG": 4036, + "GGGAAA": 4037, + "GGGAAT": 4038, + "GGGAAC": 4039, + "GGGAAG": 4040, + "GGGATA": 4041, + "GGGATT": 4042, + "GGGATC": 4043, + "GGGATG": 4044, + "GGGACA": 4045, + "GGGACT": 4046, + "GGGACC": 4047, + "GGGACG": 4048, + "GGGAGA": 4049, + "GGGAGT": 4050, + "GGGAGC": 4051, + "GGGAGG": 4052, + "GGGTAA": 4053, + "GGGTAT": 4054, + "GGGTAC": 4055, + "GGGTAG": 4056, + "GGGTTA": 4057, + "GGGTTT": 4058, + "GGGTTC": 4059, + "GGGTTG": 4060, + "GGGTCA": 4061, + "GGGTCT": 4062, + "GGGTCC": 4063, + "GGGTCG": 4064, + "GGGTGA": 4065, + "GGGTGT": 4066, + "GGGTGC": 4067, + "GGGTGG": 4068, + "GGGCAA": 4069, + "GGGCAT": 4070, + "GGGCAC": 4071, + "GGGCAG": 4072, + "GGGCTA": 4073, + "GGGCTT": 4074, + "GGGCTC": 4075, + "GGGCTG": 4076, + "GGGCCA": 4077, + "GGGCCT": 4078, + "GGGCCC": 4079, + "GGGCCG": 4080, + "GGGCGA": 4081, + "GGGCGT": 4082, + "GGGCGC": 4083, + "GGGCGG": 4084, + "GGGGAA": 4085, + "GGGGAT": 4086, + "GGGGAC": 4087, + "GGGGAG": 4088, + "GGGGTA": 4089, + "GGGGTT": 4090, + "GGGGTC": 4091, + "GGGGTG": 4092, + "GGGGCA": 4093, + "GGGGCT": 4094, + "GGGGCC": 4095, + "GGGGCG": 4096, + "GGGGGA": 4097, + "GGGGGT": 4098, + "GGGGGC": 4099, + "GGGGGG": 4100 + } + } +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6000/tokenizer_config.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6000/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..604e2b1b29af1dcb3920d2828b6a90276e5e3d02 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6000/tokenizer_config.json @@ -0,0 +1,58 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "[UNK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "[CLS]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "[SEP]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "4": { + "content": "[MASK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "clean_up_tokenization_spaces": true, + "cls_token": "[CLS]", + "do_basic_tokenize": true, + "do_lower_case": false, + "mask_token": "[MASK]", + "max_len": 512, + "model_max_length": 512, + "never_split": null, + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "strip_accents": null, + "tokenize_chinese_chars": true, + "tokenizer_class": "BertTokenizer", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6000/trainer_state.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..ddd6ea6f0e6350943daa607238790203c963b0c4 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6000/trainer_state.json @@ -0,0 +1,1713 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 24.0, + "eval_steps": 500, + "global_step": 6000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.1, + "grad_norm": 0.6611918807029724, + "learning_rate": 5.319148936170213e-05, + "loss": 0.6932, + "step": 25 + }, + { + "epoch": 0.2, + "grad_norm": 0.9467485547065735, + "learning_rate": 0.00010638297872340425, + "loss": 0.6912, + "step": 50 + }, + { + "epoch": 0.3, + "grad_norm": 0.9465051889419556, + "learning_rate": 0.00015957446808510637, + "loss": 0.6893, + "step": 75 + }, + { + "epoch": 0.4, + "grad_norm": 1.5430934429168701, + "learning_rate": 0.0002127659574468085, + "loss": 0.6804, + "step": 100 + }, + { + "epoch": 0.5, + "grad_norm": 1.6103107929229736, + "learning_rate": 0.00026595744680851064, + "loss": 0.6806, + "step": 125 + }, + { + "epoch": 0.6, + "grad_norm": 2.3773200511932373, + "learning_rate": 0.00031914893617021275, + "loss": 0.6801, + "step": 150 + }, + { + "epoch": 0.7, + "grad_norm": 1.8832203149795532, + "learning_rate": 0.0003723404255319149, + "loss": 0.6791, + "step": 175 + }, + { + "epoch": 0.8, + "grad_norm": 1.3350876569747925, + "learning_rate": 0.0003992081821181128, + "loss": 0.6746, + "step": 200 + }, + { + "epoch": 0.9, + "grad_norm": 1.5796219110488892, + "learning_rate": 0.0003975585615308479, + "loss": 0.6771, + "step": 225 + }, + { + "epoch": 1.0, + "grad_norm": 1.884006381034851, + "learning_rate": 0.00039590894094358297, + "loss": 0.6649, + "step": 250 + }, + { + "epoch": 1.1, + "grad_norm": 3.286440372467041, + "learning_rate": 0.0003942593203563181, + "loss": 0.6388, + "step": 275 + }, + { + "epoch": 1.2, + "grad_norm": 8.244946479797363, + "learning_rate": 0.0003926096997690532, + "loss": 0.6238, + "step": 300 + }, + { + "epoch": 1.3, + "grad_norm": 4.265683650970459, + "learning_rate": 0.00039096007918178817, + "loss": 0.6383, + "step": 325 + }, + { + "epoch": 1.4, + "grad_norm": 4.115826606750488, + "learning_rate": 0.0003893104585945233, + "loss": 0.6194, + "step": 350 + }, + { + "epoch": 1.5, + "grad_norm": 5.694250583648682, + "learning_rate": 0.0003876608380072583, + "loss": 0.6324, + "step": 375 + }, + { + "epoch": 1.6, + "grad_norm": 3.463121175765991, + "learning_rate": 0.00038601121741999343, + "loss": 0.621, + "step": 400 + }, + { + "epoch": 1.7, + "grad_norm": 4.582865238189697, + "learning_rate": 0.0003843615968327285, + "loss": 0.6116, + "step": 425 + }, + { + "epoch": 1.8, + "grad_norm": 11.996281623840332, + "learning_rate": 0.0003827119762454636, + "loss": 0.6393, + "step": 450 + }, + { + "epoch": 1.9, + "grad_norm": 3.0407373905181885, + "learning_rate": 0.00038106235565819863, + "loss": 0.628, + "step": 475 + }, + { + "epoch": 2.0, + "grad_norm": 2.917588233947754, + "learning_rate": 0.0003794127350709337, + "loss": 0.6078, + "step": 500 + }, + { + "epoch": 2.1, + "grad_norm": 4.748379707336426, + "learning_rate": 0.0003777631144836688, + "loss": 0.4899, + "step": 525 + }, + { + "epoch": 2.2, + "grad_norm": 3.8076977729797363, + "learning_rate": 0.00037611349389640383, + "loss": 0.5086, + "step": 550 + }, + { + "epoch": 2.3, + "grad_norm": 5.2440714836120605, + "learning_rate": 0.00037446387330913894, + "loss": 0.5327, + "step": 575 + }, + { + "epoch": 2.4, + "grad_norm": 7.110438346862793, + "learning_rate": 0.000372814252721874, + "loss": 0.5436, + "step": 600 + }, + { + "epoch": 2.5, + "grad_norm": 5.46150541305542, + "learning_rate": 0.00037116463213460903, + "loss": 0.5294, + "step": 625 + }, + { + "epoch": 2.6, + "grad_norm": 5.136163234710693, + "learning_rate": 0.00036951501154734414, + "loss": 0.5245, + "step": 650 + }, + { + "epoch": 2.7, + "grad_norm": 8.735346794128418, + "learning_rate": 0.0003678653909600792, + "loss": 0.5449, + "step": 675 + }, + { + "epoch": 2.8, + "grad_norm": 2.922825574874878, + "learning_rate": 0.0003662157703728143, + "loss": 0.5406, + "step": 700 + }, + { + "epoch": 2.9, + "grad_norm": 7.744819641113281, + "learning_rate": 0.00036456614978554934, + "loss": 0.5447, + "step": 725 + }, + { + "epoch": 3.0, + "grad_norm": 11.6185884475708, + "learning_rate": 0.00036291652919828444, + "loss": 0.5195, + "step": 750 + }, + { + "epoch": 3.1, + "grad_norm": 3.73836088180542, + "learning_rate": 0.00036126690861101944, + "loss": 0.3824, + "step": 775 + }, + { + "epoch": 3.2, + "grad_norm": 14.850343704223633, + "learning_rate": 0.00035961728802375454, + "loss": 0.4071, + "step": 800 + }, + { + "epoch": 3.3, + "grad_norm": 5.7157440185546875, + "learning_rate": 0.0003579676674364896, + "loss": 0.3986, + "step": 825 + }, + { + "epoch": 3.4, + "grad_norm": 12.418399810791016, + "learning_rate": 0.0003563180468492247, + "loss": 0.4282, + "step": 850 + }, + { + "epoch": 3.5, + "grad_norm": 12.793001174926758, + "learning_rate": 0.0003546684262619598, + "loss": 0.4822, + "step": 875 + }, + { + "epoch": 3.6, + "grad_norm": 6.489450931549072, + "learning_rate": 0.00035301880567469485, + "loss": 0.4239, + "step": 900 + }, + { + "epoch": 3.7, + "grad_norm": 5.365822792053223, + "learning_rate": 0.0003513691850874299, + "loss": 0.421, + "step": 925 + }, + { + "epoch": 3.8, + "grad_norm": 12.643745422363281, + "learning_rate": 0.00034971956450016495, + "loss": 0.3964, + "step": 950 + }, + { + "epoch": 3.9, + "grad_norm": 14.334024429321289, + "learning_rate": 0.00034806994391290005, + "loss": 0.4634, + "step": 975 + }, + { + "epoch": 4.0, + "grad_norm": 6.819091320037842, + "learning_rate": 0.0003464203233256351, + "loss": 0.4139, + "step": 1000 + }, + { + "epoch": 4.1, + "grad_norm": 8.29238224029541, + "learning_rate": 0.0003447707027383702, + "loss": 0.2695, + "step": 1025 + }, + { + "epoch": 4.2, + "grad_norm": 5.984206676483154, + "learning_rate": 0.00034312108215110525, + "loss": 0.2653, + "step": 1050 + }, + { + "epoch": 4.3, + "grad_norm": 5.9425435066223145, + "learning_rate": 0.0003414714615638403, + "loss": 0.2982, + "step": 1075 + }, + { + "epoch": 4.4, + "grad_norm": 7.1877593994140625, + "learning_rate": 0.0003398218409765754, + "loss": 0.3309, + "step": 1100 + }, + { + "epoch": 4.5, + "grad_norm": 17.83046531677246, + "learning_rate": 0.00033817222038931045, + "loss": 0.3467, + "step": 1125 + }, + { + "epoch": 4.6, + "grad_norm": 4.865128517150879, + "learning_rate": 0.00033652259980204556, + "loss": 0.3117, + "step": 1150 + }, + { + "epoch": 4.7, + "grad_norm": 31.427154541015625, + "learning_rate": 0.0003348729792147806, + "loss": 0.3273, + "step": 1175 + }, + { + "epoch": 4.8, + "grad_norm": 26.77984619140625, + "learning_rate": 0.0003332233586275157, + "loss": 0.3504, + "step": 1200 + }, + { + "epoch": 4.9, + "grad_norm": 9.389993667602539, + "learning_rate": 0.00033157373804025076, + "loss": 0.3339, + "step": 1225 + }, + { + "epoch": 5.0, + "grad_norm": 25.70441246032715, + "learning_rate": 0.0003299241174529858, + "loss": 0.3336, + "step": 1250 + }, + { + "epoch": 5.1, + "grad_norm": 13.482085227966309, + "learning_rate": 0.0003282744968657209, + "loss": 0.2156, + "step": 1275 + }, + { + "epoch": 5.2, + "grad_norm": 8.725810050964355, + "learning_rate": 0.00032662487627845596, + "loss": 0.1811, + "step": 1300 + }, + { + "epoch": 5.3, + "grad_norm": 9.313215255737305, + "learning_rate": 0.00032497525569119106, + "loss": 0.2807, + "step": 1325 + }, + { + "epoch": 5.4, + "grad_norm": 11.026411056518555, + "learning_rate": 0.0003233256351039261, + "loss": 0.2757, + "step": 1350 + }, + { + "epoch": 5.5, + "grad_norm": 11.038985252380371, + "learning_rate": 0.00032167601451666116, + "loss": 0.2177, + "step": 1375 + }, + { + "epoch": 5.6, + "grad_norm": 4.008651256561279, + "learning_rate": 0.00032002639392939627, + "loss": 0.2163, + "step": 1400 + }, + { + "epoch": 5.7, + "grad_norm": 12.480770111083984, + "learning_rate": 0.0003183767733421313, + "loss": 0.2173, + "step": 1425 + }, + { + "epoch": 5.8, + "grad_norm": 8.751969337463379, + "learning_rate": 0.0003167271527548664, + "loss": 0.2299, + "step": 1450 + }, + { + "epoch": 5.9, + "grad_norm": 7.701971530914307, + "learning_rate": 0.00031507753216760147, + "loss": 0.1949, + "step": 1475 + }, + { + "epoch": 6.0, + "grad_norm": 8.48027515411377, + "learning_rate": 0.00031342791158033657, + "loss": 0.2599, + "step": 1500 + }, + { + "epoch": 6.1, + "grad_norm": 9.29404067993164, + "learning_rate": 0.00031177829099307157, + "loss": 0.1376, + "step": 1525 + }, + { + "epoch": 6.2, + "grad_norm": 20.137714385986328, + "learning_rate": 0.00031012867040580667, + "loss": 0.1647, + "step": 1550 + }, + { + "epoch": 6.3, + "grad_norm": 11.394575119018555, + "learning_rate": 0.0003084790498185417, + "loss": 0.1565, + "step": 1575 + }, + { + "epoch": 6.4, + "grad_norm": 8.214287757873535, + "learning_rate": 0.0003068294292312768, + "loss": 0.1739, + "step": 1600 + }, + { + "epoch": 6.5, + "grad_norm": 7.779988765716553, + "learning_rate": 0.0003051798086440119, + "loss": 0.1403, + "step": 1625 + }, + { + "epoch": 6.6, + "grad_norm": 9.421648025512695, + "learning_rate": 0.000303530188056747, + "loss": 0.1545, + "step": 1650 + }, + { + "epoch": 6.7, + "grad_norm": 5.751734256744385, + "learning_rate": 0.000301880567469482, + "loss": 0.1971, + "step": 1675 + }, + { + "epoch": 6.8, + "grad_norm": 23.861705780029297, + "learning_rate": 0.0003002309468822171, + "loss": 0.1681, + "step": 1700 + }, + { + "epoch": 6.9, + "grad_norm": 18.944721221923828, + "learning_rate": 0.0002985813262949522, + "loss": 0.1703, + "step": 1725 + }, + { + "epoch": 7.0, + "grad_norm": 14.045795440673828, + "learning_rate": 0.00029693170570768723, + "loss": 0.1801, + "step": 1750 + }, + { + "epoch": 7.1, + "grad_norm": 6.4620137214660645, + "learning_rate": 0.00029528208512042233, + "loss": 0.1253, + "step": 1775 + }, + { + "epoch": 7.2, + "grad_norm": 4.318169593811035, + "learning_rate": 0.0002936324645331574, + "loss": 0.1397, + "step": 1800 + }, + { + "epoch": 7.3, + "grad_norm": 24.91462516784668, + "learning_rate": 0.00029198284394589243, + "loss": 0.1259, + "step": 1825 + }, + { + "epoch": 7.4, + "grad_norm": 23.614572525024414, + "learning_rate": 0.00029033322335862753, + "loss": 0.1293, + "step": 1850 + }, + { + "epoch": 7.5, + "grad_norm": 3.60048508644104, + "learning_rate": 0.0002886836027713626, + "loss": 0.138, + "step": 1875 + }, + { + "epoch": 7.6, + "grad_norm": 16.62705421447754, + "learning_rate": 0.0002870339821840977, + "loss": 0.1733, + "step": 1900 + }, + { + "epoch": 7.7, + "grad_norm": 33.79671859741211, + "learning_rate": 0.00028538436159683273, + "loss": 0.217, + "step": 1925 + }, + { + "epoch": 7.8, + "grad_norm": 9.69206428527832, + "learning_rate": 0.00028373474100956784, + "loss": 0.1289, + "step": 1950 + }, + { + "epoch": 7.9, + "grad_norm": 18.655046463012695, + "learning_rate": 0.0002820851204223029, + "loss": 0.1168, + "step": 1975 + }, + { + "epoch": 8.0, + "grad_norm": 8.4110746383667, + "learning_rate": 0.00028043549983503794, + "loss": 0.1471, + "step": 2000 + }, + { + "epoch": 8.1, + "grad_norm": 14.473428726196289, + "learning_rate": 0.00027878587924777304, + "loss": 0.0774, + "step": 2025 + }, + { + "epoch": 8.2, + "grad_norm": 1.2464979887008667, + "learning_rate": 0.0002771362586605081, + "loss": 0.0715, + "step": 2050 + }, + { + "epoch": 8.3, + "grad_norm": 26.74981689453125, + "learning_rate": 0.0002754866380732432, + "loss": 0.1289, + "step": 2075 + }, + { + "epoch": 8.4, + "grad_norm": 13.958703994750977, + "learning_rate": 0.00027383701748597824, + "loss": 0.1294, + "step": 2100 + }, + { + "epoch": 8.5, + "grad_norm": 10.394835472106934, + "learning_rate": 0.0002721873968987133, + "loss": 0.0987, + "step": 2125 + }, + { + "epoch": 8.6, + "grad_norm": 9.387774467468262, + "learning_rate": 0.00027053777631144834, + "loss": 0.1269, + "step": 2150 + }, + { + "epoch": 8.7, + "grad_norm": 4.988718509674072, + "learning_rate": 0.00026888815572418344, + "loss": 0.1374, + "step": 2175 + }, + { + "epoch": 8.8, + "grad_norm": 12.66832447052002, + "learning_rate": 0.00026723853513691855, + "loss": 0.0997, + "step": 2200 + }, + { + "epoch": 8.9, + "grad_norm": 19.10486602783203, + "learning_rate": 0.0002655889145496536, + "loss": 0.1325, + "step": 2225 + }, + { + "epoch": 9.0, + "grad_norm": 3.2943971157073975, + "learning_rate": 0.0002639392939623887, + "loss": 0.1385, + "step": 2250 + }, + { + "epoch": 9.1, + "grad_norm": 11.490874290466309, + "learning_rate": 0.0002622896733751237, + "loss": 0.0573, + "step": 2275 + }, + { + "epoch": 9.2, + "grad_norm": 11.961288452148438, + "learning_rate": 0.0002606400527878588, + "loss": 0.1128, + "step": 2300 + }, + { + "epoch": 9.3, + "grad_norm": 0.7256277203559875, + "learning_rate": 0.00025899043220059385, + "loss": 0.0673, + "step": 2325 + }, + { + "epoch": 9.4, + "grad_norm": 9.857927322387695, + "learning_rate": 0.00025734081161332895, + "loss": 0.2072, + "step": 2350 + }, + { + "epoch": 9.5, + "grad_norm": 34.927734375, + "learning_rate": 0.000255691191026064, + "loss": 0.221, + "step": 2375 + }, + { + "epoch": 9.6, + "grad_norm": 28.461849212646484, + "learning_rate": 0.0002540415704387991, + "loss": 0.1263, + "step": 2400 + }, + { + "epoch": 9.7, + "grad_norm": 9.591217041015625, + "learning_rate": 0.00025239194985153415, + "loss": 0.099, + "step": 2425 + }, + { + "epoch": 9.8, + "grad_norm": 9.780253410339355, + "learning_rate": 0.0002507423292642692, + "loss": 0.0754, + "step": 2450 + }, + { + "epoch": 9.9, + "grad_norm": 14.394485473632812, + "learning_rate": 0.0002490927086770043, + "loss": 0.0909, + "step": 2475 + }, + { + "epoch": 10.0, + "grad_norm": 4.217277526855469, + "learning_rate": 0.00024744308808973936, + "loss": 0.1495, + "step": 2500 + }, + { + "epoch": 10.1, + "grad_norm": 0.08438724279403687, + "learning_rate": 0.00024579346750247446, + "loss": 0.0662, + "step": 2525 + }, + { + "epoch": 10.2, + "grad_norm": 0.9595862030982971, + "learning_rate": 0.0002441438469152095, + "loss": 0.084, + "step": 2550 + }, + { + "epoch": 10.3, + "grad_norm": 42.32374954223633, + "learning_rate": 0.00024249422632794456, + "loss": 0.0541, + "step": 2575 + }, + { + "epoch": 10.4, + "grad_norm": 24.792821884155273, + "learning_rate": 0.00024084460574067963, + "loss": 0.121, + "step": 2600 + }, + { + "epoch": 10.5, + "grad_norm": 2.7049331665039062, + "learning_rate": 0.0002391949851534147, + "loss": 0.1551, + "step": 2625 + }, + { + "epoch": 10.6, + "grad_norm": 2.788362741470337, + "learning_rate": 0.0002375453645661498, + "loss": 0.0877, + "step": 2650 + }, + { + "epoch": 10.7, + "grad_norm": 0.4999215006828308, + "learning_rate": 0.00023589574397888486, + "loss": 0.1263, + "step": 2675 + }, + { + "epoch": 10.8, + "grad_norm": 16.476675033569336, + "learning_rate": 0.00023424612339161997, + "loss": 0.0883, + "step": 2700 + }, + { + "epoch": 10.9, + "grad_norm": 38.192726135253906, + "learning_rate": 0.000232596502804355, + "loss": 0.0843, + "step": 2725 + }, + { + "epoch": 11.0, + "grad_norm": 18.09952735900879, + "learning_rate": 0.00023094688221709007, + "loss": 0.0957, + "step": 2750 + }, + { + "epoch": 11.1, + "grad_norm": 34.254417419433594, + "learning_rate": 0.00022929726162982514, + "loss": 0.0856, + "step": 2775 + }, + { + "epoch": 11.2, + "grad_norm": 11.744216918945312, + "learning_rate": 0.00022764764104256022, + "loss": 0.0654, + "step": 2800 + }, + { + "epoch": 11.3, + "grad_norm": 0.1285558044910431, + "learning_rate": 0.0002259980204552953, + "loss": 0.0638, + "step": 2825 + }, + { + "epoch": 11.4, + "grad_norm": 20.57583999633789, + "learning_rate": 0.00022434839986803037, + "loss": 0.079, + "step": 2850 + }, + { + "epoch": 11.5, + "grad_norm": 0.7192332148551941, + "learning_rate": 0.00022269877928076542, + "loss": 0.0854, + "step": 2875 + }, + { + "epoch": 11.6, + "grad_norm": 31.341829299926758, + "learning_rate": 0.0002210491586935005, + "loss": 0.063, + "step": 2900 + }, + { + "epoch": 11.7, + "grad_norm": 13.460247039794922, + "learning_rate": 0.00021939953810623557, + "loss": 0.1358, + "step": 2925 + }, + { + "epoch": 11.8, + "grad_norm": 1.671036720275879, + "learning_rate": 0.00021774991751897065, + "loss": 0.0806, + "step": 2950 + }, + { + "epoch": 11.9, + "grad_norm": 22.026491165161133, + "learning_rate": 0.00021610029693170573, + "loss": 0.0912, + "step": 2975 + }, + { + "epoch": 12.0, + "grad_norm": 46.78192138671875, + "learning_rate": 0.0002144506763444408, + "loss": 0.0763, + "step": 3000 + }, + { + "epoch": 12.1, + "grad_norm": 6.397649765014648, + "learning_rate": 0.00021280105575717582, + "loss": 0.0613, + "step": 3025 + }, + { + "epoch": 12.2, + "grad_norm": 7.179838180541992, + "learning_rate": 0.00021115143516991093, + "loss": 0.0496, + "step": 3050 + }, + { + "epoch": 12.3, + "grad_norm": 0.021825680509209633, + "learning_rate": 0.000209501814582646, + "loss": 0.0633, + "step": 3075 + }, + { + "epoch": 12.4, + "grad_norm": 6.6774187088012695, + "learning_rate": 0.00020785219399538108, + "loss": 0.0765, + "step": 3100 + }, + { + "epoch": 12.5, + "grad_norm": 0.13217756152153015, + "learning_rate": 0.00020620257340811616, + "loss": 0.0696, + "step": 3125 + }, + { + "epoch": 12.6, + "grad_norm": 2.5293831825256348, + "learning_rate": 0.00020455295282085123, + "loss": 0.0419, + "step": 3150 + }, + { + "epoch": 12.7, + "grad_norm": 1.1612874269485474, + "learning_rate": 0.00020290333223358626, + "loss": 0.1046, + "step": 3175 + }, + { + "epoch": 12.8, + "grad_norm": 40.93427658081055, + "learning_rate": 0.00020125371164632133, + "loss": 0.0488, + "step": 3200 + }, + { + "epoch": 12.9, + "grad_norm": 1.446478009223938, + "learning_rate": 0.0001996040910590564, + "loss": 0.062, + "step": 3225 + }, + { + "epoch": 13.0, + "grad_norm": 28.893821716308594, + "learning_rate": 0.00019795447047179148, + "loss": 0.0913, + "step": 3250 + }, + { + "epoch": 13.1, + "grad_norm": 23.31914520263672, + "learning_rate": 0.0001963048498845266, + "loss": 0.061, + "step": 3275 + }, + { + "epoch": 13.2, + "grad_norm": 0.26242795586586, + "learning_rate": 0.00019465522929726164, + "loss": 0.0563, + "step": 3300 + }, + { + "epoch": 13.3, + "grad_norm": 0.08909507840871811, + "learning_rate": 0.00019300560870999671, + "loss": 0.0387, + "step": 3325 + }, + { + "epoch": 13.4, + "grad_norm": 24.825326919555664, + "learning_rate": 0.0001913559881227318, + "loss": 0.0776, + "step": 3350 + }, + { + "epoch": 13.5, + "grad_norm": 39.43446731567383, + "learning_rate": 0.00018970636753546684, + "loss": 0.0846, + "step": 3375 + }, + { + "epoch": 13.6, + "grad_norm": 32.98988723754883, + "learning_rate": 0.00018805674694820192, + "loss": 0.0512, + "step": 3400 + }, + { + "epoch": 13.7, + "grad_norm": 4.053821563720703, + "learning_rate": 0.000186407126360937, + "loss": 0.1576, + "step": 3425 + }, + { + "epoch": 13.8, + "grad_norm": 0.2785554528236389, + "learning_rate": 0.00018475750577367207, + "loss": 0.0728, + "step": 3450 + }, + { + "epoch": 13.9, + "grad_norm": 11.39714527130127, + "learning_rate": 0.00018310788518640715, + "loss": 0.0776, + "step": 3475 + }, + { + "epoch": 14.0, + "grad_norm": 3.7018072605133057, + "learning_rate": 0.00018145826459914222, + "loss": 0.0736, + "step": 3500 + }, + { + "epoch": 14.1, + "grad_norm": 0.26053619384765625, + "learning_rate": 0.00017980864401187727, + "loss": 0.0598, + "step": 3525 + }, + { + "epoch": 14.2, + "grad_norm": 6.454179763793945, + "learning_rate": 0.00017815902342461235, + "loss": 0.0435, + "step": 3550 + }, + { + "epoch": 14.3, + "grad_norm": 34.47985076904297, + "learning_rate": 0.00017650940283734742, + "loss": 0.0467, + "step": 3575 + }, + { + "epoch": 14.4, + "grad_norm": 11.76491928100586, + "learning_rate": 0.00017485978225008247, + "loss": 0.1298, + "step": 3600 + }, + { + "epoch": 14.5, + "grad_norm": 26.64131736755371, + "learning_rate": 0.00017321016166281755, + "loss": 0.1313, + "step": 3625 + }, + { + "epoch": 14.6, + "grad_norm": 5.510906219482422, + "learning_rate": 0.00017156054107555263, + "loss": 0.0421, + "step": 3650 + }, + { + "epoch": 14.7, + "grad_norm": 0.04721131548285484, + "learning_rate": 0.0001699109204882877, + "loss": 0.0569, + "step": 3675 + }, + { + "epoch": 14.8, + "grad_norm": 0.6616227030754089, + "learning_rate": 0.00016826129990102278, + "loss": 0.0618, + "step": 3700 + }, + { + "epoch": 14.9, + "grad_norm": 0.05287986248731613, + "learning_rate": 0.00016661167931375785, + "loss": 0.0574, + "step": 3725 + }, + { + "epoch": 15.0, + "grad_norm": 0.49707508087158203, + "learning_rate": 0.0001649620587264929, + "loss": 0.094, + "step": 3750 + }, + { + "epoch": 15.1, + "grad_norm": 23.649728775024414, + "learning_rate": 0.00016331243813922798, + "loss": 0.0512, + "step": 3775 + }, + { + "epoch": 15.2, + "grad_norm": 0.01772051490843296, + "learning_rate": 0.00016166281755196306, + "loss": 0.0719, + "step": 3800 + }, + { + "epoch": 15.3, + "grad_norm": 1.5259039402008057, + "learning_rate": 0.00016001319696469813, + "loss": 0.0615, + "step": 3825 + }, + { + "epoch": 15.4, + "grad_norm": 0.026406478136777878, + "learning_rate": 0.0001583635763774332, + "loss": 0.0651, + "step": 3850 + }, + { + "epoch": 15.5, + "grad_norm": 27.12177085876465, + "learning_rate": 0.00015671395579016829, + "loss": 0.0458, + "step": 3875 + }, + { + "epoch": 15.6, + "grad_norm": 9.838811874389648, + "learning_rate": 0.00015506433520290334, + "loss": 0.0662, + "step": 3900 + }, + { + "epoch": 15.7, + "grad_norm": 0.08516795933246613, + "learning_rate": 0.0001534147146156384, + "loss": 0.0302, + "step": 3925 + }, + { + "epoch": 15.8, + "grad_norm": 1.1224850416183472, + "learning_rate": 0.0001517650940283735, + "loss": 0.0637, + "step": 3950 + }, + { + "epoch": 15.9, + "grad_norm": 0.14810702204704285, + "learning_rate": 0.00015011547344110854, + "loss": 0.0296, + "step": 3975 + }, + { + "epoch": 16.0, + "grad_norm": 0.046288371086120605, + "learning_rate": 0.00014846585285384361, + "loss": 0.0368, + "step": 4000 + }, + { + "epoch": 16.1, + "grad_norm": 0.008764918893575668, + "learning_rate": 0.0001468162322665787, + "loss": 0.0272, + "step": 4025 + }, + { + "epoch": 16.2, + "grad_norm": 0.39770299196243286, + "learning_rate": 0.00014516661167931377, + "loss": 0.0344, + "step": 4050 + }, + { + "epoch": 16.3, + "grad_norm": 23.780405044555664, + "learning_rate": 0.00014351699109204884, + "loss": 0.0617, + "step": 4075 + }, + { + "epoch": 16.4, + "grad_norm": 0.023103665560483932, + "learning_rate": 0.00014186737050478392, + "loss": 0.0182, + "step": 4100 + }, + { + "epoch": 16.5, + "grad_norm": 0.00998806394636631, + "learning_rate": 0.00014021774991751897, + "loss": 0.042, + "step": 4125 + }, + { + "epoch": 16.6, + "grad_norm": 43.775169372558594, + "learning_rate": 0.00013856812933025404, + "loss": 0.0182, + "step": 4150 + }, + { + "epoch": 16.7, + "grad_norm": 0.008157053031027317, + "learning_rate": 0.00013691850874298912, + "loss": 0.0377, + "step": 4175 + }, + { + "epoch": 16.8, + "grad_norm": 0.011455570347607136, + "learning_rate": 0.00013526888815572417, + "loss": 0.0255, + "step": 4200 + }, + { + "epoch": 16.9, + "grad_norm": 34.8545036315918, + "learning_rate": 0.00013361926756845927, + "loss": 0.0486, + "step": 4225 + }, + { + "epoch": 17.0, + "grad_norm": 0.006066238507628441, + "learning_rate": 0.00013196964698119435, + "loss": 0.0821, + "step": 4250 + }, + { + "epoch": 17.1, + "grad_norm": 43.46882247924805, + "learning_rate": 0.0001303200263939294, + "loss": 0.0333, + "step": 4275 + }, + { + "epoch": 17.2, + "grad_norm": 2.3177573680877686, + "learning_rate": 0.00012867040580666448, + "loss": 0.0477, + "step": 4300 + }, + { + "epoch": 17.3, + "grad_norm": 0.2264157235622406, + "learning_rate": 0.00012702078521939955, + "loss": 0.0373, + "step": 4325 + }, + { + "epoch": 17.4, + "grad_norm": 68.17675018310547, + "learning_rate": 0.0001253711646321346, + "loss": 0.036, + "step": 4350 + }, + { + "epoch": 17.5, + "grad_norm": 1.5359147787094116, + "learning_rate": 0.00012372154404486968, + "loss": 0.0437, + "step": 4375 + }, + { + "epoch": 17.6, + "grad_norm": 0.008654219098389149, + "learning_rate": 0.00012207192345760475, + "loss": 0.0295, + "step": 4400 + }, + { + "epoch": 17.7, + "grad_norm": 0.0074989828281104565, + "learning_rate": 0.00012042230287033982, + "loss": 0.0229, + "step": 4425 + }, + { + "epoch": 17.8, + "grad_norm": 0.03568067401647568, + "learning_rate": 0.0001187726822830749, + "loss": 0.0636, + "step": 4450 + }, + { + "epoch": 17.9, + "grad_norm": 0.0151940006762743, + "learning_rate": 0.00011712306169580998, + "loss": 0.0669, + "step": 4475 + }, + { + "epoch": 18.0, + "grad_norm": 0.04830991476774216, + "learning_rate": 0.00011547344110854503, + "loss": 0.0388, + "step": 4500 + }, + { + "epoch": 18.1, + "grad_norm": 1.1794861555099487, + "learning_rate": 0.00011382382052128011, + "loss": 0.0069, + "step": 4525 + }, + { + "epoch": 18.2, + "grad_norm": 13.07977294921875, + "learning_rate": 0.00011217419993401519, + "loss": 0.0315, + "step": 4550 + }, + { + "epoch": 18.3, + "grad_norm": 0.011115381494164467, + "learning_rate": 0.00011052457934675025, + "loss": 0.0144, + "step": 4575 + }, + { + "epoch": 18.4, + "grad_norm": 25.42391014099121, + "learning_rate": 0.00010887495875948532, + "loss": 0.0198, + "step": 4600 + }, + { + "epoch": 18.5, + "grad_norm": 44.36895751953125, + "learning_rate": 0.0001072253381722204, + "loss": 0.0297, + "step": 4625 + }, + { + "epoch": 18.6, + "grad_norm": 6.828146457672119, + "learning_rate": 0.00010557571758495546, + "loss": 0.0348, + "step": 4650 + }, + { + "epoch": 18.7, + "grad_norm": 38.9465446472168, + "learning_rate": 0.00010392609699769054, + "loss": 0.0336, + "step": 4675 + }, + { + "epoch": 18.8, + "grad_norm": 0.015516690909862518, + "learning_rate": 0.00010227647641042562, + "loss": 0.032, + "step": 4700 + }, + { + "epoch": 18.9, + "grad_norm": 5.419480323791504, + "learning_rate": 0.00010062685582316067, + "loss": 0.0169, + "step": 4725 + }, + { + "epoch": 19.0, + "grad_norm": 0.7849720120429993, + "learning_rate": 9.897723523589574e-05, + "loss": 0.0321, + "step": 4750 + }, + { + "epoch": 19.1, + "grad_norm": 25.438196182250977, + "learning_rate": 9.732761464863082e-05, + "loss": 0.046, + "step": 4775 + }, + { + "epoch": 19.2, + "grad_norm": 0.010676453821361065, + "learning_rate": 9.56779940613659e-05, + "loss": 0.0111, + "step": 4800 + }, + { + "epoch": 19.3, + "grad_norm": 0.011585243977606297, + "learning_rate": 9.402837347410096e-05, + "loss": 0.0154, + "step": 4825 + }, + { + "epoch": 19.4, + "grad_norm": 0.012879762798547745, + "learning_rate": 9.237875288683603e-05, + "loss": 0.02, + "step": 4850 + }, + { + "epoch": 19.5, + "grad_norm": 0.4159695506095886, + "learning_rate": 9.072913229957111e-05, + "loss": 0.0277, + "step": 4875 + }, + { + "epoch": 19.6, + "grad_norm": 0.013872765935957432, + "learning_rate": 8.907951171230617e-05, + "loss": 0.0069, + "step": 4900 + }, + { + "epoch": 19.7, + "grad_norm": 3.4260928630828857, + "learning_rate": 8.742989112504124e-05, + "loss": 0.0319, + "step": 4925 + }, + { + "epoch": 19.8, + "grad_norm": 0.043105900287628174, + "learning_rate": 8.578027053777631e-05, + "loss": 0.0445, + "step": 4950 + }, + { + "epoch": 19.9, + "grad_norm": 10.97398853302002, + "learning_rate": 8.413064995051139e-05, + "loss": 0.0185, + "step": 4975 + }, + { + "epoch": 20.0, + "grad_norm": 12.879217147827148, + "learning_rate": 8.248102936324645e-05, + "loss": 0.0197, + "step": 5000 + }, + { + "epoch": 20.1, + "grad_norm": 64.58011627197266, + "learning_rate": 8.083140877598153e-05, + "loss": 0.0228, + "step": 5025 + }, + { + "epoch": 20.2, + "grad_norm": 43.28678512573242, + "learning_rate": 7.91817881887166e-05, + "loss": 0.0311, + "step": 5050 + }, + { + "epoch": 20.3, + "grad_norm": 0.005558234639465809, + "learning_rate": 7.753216760145167e-05, + "loss": 0.0283, + "step": 5075 + }, + { + "epoch": 20.4, + "grad_norm": 0.19123849272727966, + "learning_rate": 7.588254701418674e-05, + "loss": 0.0445, + "step": 5100 + }, + { + "epoch": 20.5, + "grad_norm": 5.734357833862305, + "learning_rate": 7.423292642692181e-05, + "loss": 0.0145, + "step": 5125 + }, + { + "epoch": 20.6, + "grad_norm": 0.005894747097045183, + "learning_rate": 7.258330583965688e-05, + "loss": 0.0233, + "step": 5150 + }, + { + "epoch": 20.7, + "grad_norm": 0.018981292843818665, + "learning_rate": 7.093368525239196e-05, + "loss": 0.0273, + "step": 5175 + }, + { + "epoch": 20.8, + "grad_norm": 0.09156472235918045, + "learning_rate": 6.928406466512702e-05, + "loss": 0.0233, + "step": 5200 + }, + { + "epoch": 20.9, + "grad_norm": 0.8908875584602356, + "learning_rate": 6.763444407786209e-05, + "loss": 0.0221, + "step": 5225 + }, + { + "epoch": 21.0, + "grad_norm": 6.745031833648682, + "learning_rate": 6.598482349059718e-05, + "loss": 0.0278, + "step": 5250 + }, + { + "epoch": 21.1, + "grad_norm": 0.009193326346576214, + "learning_rate": 6.433520290333224e-05, + "loss": 0.0226, + "step": 5275 + }, + { + "epoch": 21.2, + "grad_norm": 2.408499002456665, + "learning_rate": 6.26855823160673e-05, + "loss": 0.0202, + "step": 5300 + }, + { + "epoch": 21.3, + "grad_norm": 0.006645245011895895, + "learning_rate": 6.103596172880238e-05, + "loss": 0.0234, + "step": 5325 + }, + { + "epoch": 21.4, + "grad_norm": 0.006036018021404743, + "learning_rate": 5.938634114153745e-05, + "loss": 0.0333, + "step": 5350 + }, + { + "epoch": 21.5, + "grad_norm": 0.25791531801223755, + "learning_rate": 5.7736720554272516e-05, + "loss": 0.012, + "step": 5375 + }, + { + "epoch": 21.6, + "grad_norm": 0.9368523359298706, + "learning_rate": 5.608709996700759e-05, + "loss": 0.0078, + "step": 5400 + }, + { + "epoch": 21.7, + "grad_norm": 0.0046651544980704784, + "learning_rate": 5.443747937974266e-05, + "loss": 0.0096, + "step": 5425 + }, + { + "epoch": 21.8, + "grad_norm": 0.004647469613701105, + "learning_rate": 5.278785879247773e-05, + "loss": 0.0299, + "step": 5450 + }, + { + "epoch": 21.9, + "grad_norm": 0.004732856526970863, + "learning_rate": 5.113823820521281e-05, + "loss": 0.0185, + "step": 5475 + }, + { + "epoch": 22.0, + "grad_norm": 0.007009522989392281, + "learning_rate": 4.948861761794787e-05, + "loss": 0.0228, + "step": 5500 + }, + { + "epoch": 22.1, + "grad_norm": 25.634002685546875, + "learning_rate": 4.783899703068295e-05, + "loss": 0.011, + "step": 5525 + }, + { + "epoch": 22.2, + "grad_norm": 0.009100685827434063, + "learning_rate": 4.618937644341802e-05, + "loss": 0.0056, + "step": 5550 + }, + { + "epoch": 22.3, + "grad_norm": 0.004046417307108641, + "learning_rate": 4.453975585615309e-05, + "loss": 0.0084, + "step": 5575 + }, + { + "epoch": 22.4, + "grad_norm": 0.025503572076559067, + "learning_rate": 4.2890135268888156e-05, + "loss": 0.0066, + "step": 5600 + }, + { + "epoch": 22.5, + "grad_norm": 0.005815221928060055, + "learning_rate": 4.1240514681623226e-05, + "loss": 0.0407, + "step": 5625 + }, + { + "epoch": 22.6, + "grad_norm": 0.06112132593989372, + "learning_rate": 3.95908940943583e-05, + "loss": 0.0112, + "step": 5650 + }, + { + "epoch": 22.7, + "grad_norm": 0.01705116033554077, + "learning_rate": 3.794127350709337e-05, + "loss": 0.0008, + "step": 5675 + }, + { + "epoch": 22.8, + "grad_norm": 0.005267620086669922, + "learning_rate": 3.629165291982844e-05, + "loss": 0.0131, + "step": 5700 + }, + { + "epoch": 22.9, + "grad_norm": 0.0046350546181201935, + "learning_rate": 3.464203233256351e-05, + "loss": 0.0205, + "step": 5725 + }, + { + "epoch": 23.0, + "grad_norm": 0.0063110594637691975, + "learning_rate": 3.299241174529859e-05, + "loss": 0.0072, + "step": 5750 + }, + { + "epoch": 23.1, + "grad_norm": 0.004843282513320446, + "learning_rate": 3.134279115803365e-05, + "loss": 0.033, + "step": 5775 + }, + { + "epoch": 23.2, + "grad_norm": 0.004899237770587206, + "learning_rate": 2.9693170570768723e-05, + "loss": 0.002, + "step": 5800 + }, + { + "epoch": 23.3, + "grad_norm": 0.0040695276111364365, + "learning_rate": 2.8043549983503796e-05, + "loss": 0.0185, + "step": 5825 + }, + { + "epoch": 23.4, + "grad_norm": 46.083438873291016, + "learning_rate": 2.6393929396238866e-05, + "loss": 0.0149, + "step": 5850 + }, + { + "epoch": 23.5, + "grad_norm": 0.009059540927410126, + "learning_rate": 2.4744308808973936e-05, + "loss": 0.0264, + "step": 5875 + }, + { + "epoch": 23.6, + "grad_norm": 11.57484245300293, + "learning_rate": 2.309468822170901e-05, + "loss": 0.0145, + "step": 5900 + }, + { + "epoch": 23.7, + "grad_norm": 0.00815950334072113, + "learning_rate": 2.1445067634444078e-05, + "loss": 0.015, + "step": 5925 + }, + { + "epoch": 23.8, + "grad_norm": 0.005930441431701183, + "learning_rate": 1.979544704717915e-05, + "loss": 0.0187, + "step": 5950 + }, + { + "epoch": 23.9, + "grad_norm": 0.006135927978903055, + "learning_rate": 1.814582645991422e-05, + "loss": 0.0143, + "step": 5975 + }, + { + "epoch": 24.0, + "grad_norm": 0.004939633421599865, + "learning_rate": 1.6496205872649294e-05, + "loss": 0.0088, + "step": 6000 + } + ], + "logging_steps": 25, + "max_steps": 6250, + "num_input_tokens_seen": 0, + "num_train_epochs": 25, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 304916889600000.0, + "train_batch_size": 20, + "trial_name": null, + "trial_params": null +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6000/training_args.bin b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6000/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..798d8a9f3a9513692035320aba5484c89ac49e69 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6000/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1cabf5db3e9d79ec0c1945933395ff57418204414cc81f77c90324d003fac782 +size 5304 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6000/vocab.txt b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6000/vocab.txt new file mode 100644 index 0000000000000000000000000000000000000000..cc9ad53732e21ea9180f18e1fb67a38cc5702d4c --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6000/vocab.txt @@ -0,0 +1,4101 @@ +[PAD] +[UNK] +[CLS] +[SEP] +[MASK] +AAAAAA +AAAAAT +AAAAAC +AAAAAG +AAAATA +AAAATT +AAAATC +AAAATG +AAAACA +AAAACT +AAAACC +AAAACG +AAAAGA +AAAAGT +AAAAGC +AAAAGG +AAATAA +AAATAT +AAATAC +AAATAG +AAATTA +AAATTT +AAATTC +AAATTG +AAATCA +AAATCT +AAATCC +AAATCG +AAATGA +AAATGT +AAATGC +AAATGG +AAACAA +AAACAT +AAACAC +AAACAG +AAACTA +AAACTT +AAACTC +AAACTG +AAACCA +AAACCT +AAACCC +AAACCG +AAACGA +AAACGT +AAACGC +AAACGG +AAAGAA +AAAGAT +AAAGAC +AAAGAG +AAAGTA +AAAGTT +AAAGTC +AAAGTG +AAAGCA +AAAGCT +AAAGCC +AAAGCG +AAAGGA +AAAGGT +AAAGGC +AAAGGG +AATAAA +AATAAT +AATAAC +AATAAG +AATATA +AATATT +AATATC +AATATG +AATACA +AATACT +AATACC +AATACG +AATAGA +AATAGT +AATAGC +AATAGG +AATTAA +AATTAT +AATTAC +AATTAG +AATTTA +AATTTT +AATTTC +AATTTG +AATTCA +AATTCT +AATTCC +AATTCG +AATTGA +AATTGT +AATTGC +AATTGG +AATCAA +AATCAT +AATCAC +AATCAG +AATCTA +AATCTT +AATCTC +AATCTG +AATCCA +AATCCT +AATCCC +AATCCG +AATCGA +AATCGT +AATCGC +AATCGG +AATGAA +AATGAT +AATGAC +AATGAG +AATGTA +AATGTT +AATGTC +AATGTG +AATGCA +AATGCT +AATGCC +AATGCG +AATGGA +AATGGT +AATGGC +AATGGG +AACAAA +AACAAT +AACAAC +AACAAG +AACATA +AACATT +AACATC +AACATG +AACACA +AACACT +AACACC +AACACG +AACAGA +AACAGT +AACAGC +AACAGG +AACTAA +AACTAT +AACTAC +AACTAG +AACTTA +AACTTT +AACTTC +AACTTG +AACTCA +AACTCT +AACTCC +AACTCG +AACTGA +AACTGT +AACTGC +AACTGG +AACCAA +AACCAT +AACCAC +AACCAG +AACCTA +AACCTT +AACCTC +AACCTG +AACCCA +AACCCT +AACCCC +AACCCG +AACCGA +AACCGT +AACCGC +AACCGG +AACGAA +AACGAT +AACGAC +AACGAG +AACGTA +AACGTT +AACGTC +AACGTG +AACGCA +AACGCT +AACGCC +AACGCG +AACGGA +AACGGT +AACGGC +AACGGG +AAGAAA +AAGAAT +AAGAAC +AAGAAG +AAGATA +AAGATT +AAGATC +AAGATG +AAGACA +AAGACT +AAGACC +AAGACG +AAGAGA +AAGAGT +AAGAGC +AAGAGG +AAGTAA +AAGTAT +AAGTAC +AAGTAG +AAGTTA +AAGTTT +AAGTTC +AAGTTG +AAGTCA +AAGTCT +AAGTCC +AAGTCG +AAGTGA +AAGTGT +AAGTGC +AAGTGG +AAGCAA +AAGCAT +AAGCAC +AAGCAG +AAGCTA +AAGCTT +AAGCTC +AAGCTG +AAGCCA +AAGCCT +AAGCCC +AAGCCG +AAGCGA +AAGCGT +AAGCGC +AAGCGG +AAGGAA +AAGGAT +AAGGAC +AAGGAG +AAGGTA +AAGGTT +AAGGTC +AAGGTG +AAGGCA +AAGGCT +AAGGCC +AAGGCG +AAGGGA +AAGGGT +AAGGGC +AAGGGG +ATAAAA +ATAAAT +ATAAAC +ATAAAG +ATAATA +ATAATT +ATAATC +ATAATG +ATAACA +ATAACT +ATAACC +ATAACG +ATAAGA +ATAAGT +ATAAGC +ATAAGG +ATATAA +ATATAT +ATATAC +ATATAG +ATATTA +ATATTT +ATATTC +ATATTG +ATATCA +ATATCT +ATATCC +ATATCG +ATATGA +ATATGT +ATATGC +ATATGG +ATACAA +ATACAT +ATACAC +ATACAG +ATACTA +ATACTT +ATACTC +ATACTG +ATACCA +ATACCT +ATACCC +ATACCG +ATACGA +ATACGT +ATACGC +ATACGG +ATAGAA +ATAGAT +ATAGAC +ATAGAG +ATAGTA +ATAGTT +ATAGTC +ATAGTG +ATAGCA +ATAGCT +ATAGCC +ATAGCG +ATAGGA +ATAGGT +ATAGGC +ATAGGG +ATTAAA +ATTAAT +ATTAAC +ATTAAG +ATTATA +ATTATT +ATTATC +ATTATG +ATTACA +ATTACT +ATTACC +ATTACG +ATTAGA +ATTAGT +ATTAGC +ATTAGG +ATTTAA +ATTTAT +ATTTAC +ATTTAG +ATTTTA +ATTTTT +ATTTTC +ATTTTG +ATTTCA +ATTTCT +ATTTCC +ATTTCG +ATTTGA +ATTTGT +ATTTGC +ATTTGG +ATTCAA +ATTCAT +ATTCAC +ATTCAG +ATTCTA +ATTCTT +ATTCTC +ATTCTG +ATTCCA +ATTCCT +ATTCCC +ATTCCG +ATTCGA +ATTCGT +ATTCGC +ATTCGG +ATTGAA +ATTGAT +ATTGAC +ATTGAG +ATTGTA +ATTGTT +ATTGTC +ATTGTG +ATTGCA +ATTGCT +ATTGCC +ATTGCG +ATTGGA +ATTGGT +ATTGGC +ATTGGG +ATCAAA +ATCAAT +ATCAAC +ATCAAG +ATCATA +ATCATT +ATCATC +ATCATG +ATCACA +ATCACT +ATCACC +ATCACG +ATCAGA +ATCAGT +ATCAGC +ATCAGG +ATCTAA +ATCTAT +ATCTAC +ATCTAG +ATCTTA +ATCTTT +ATCTTC +ATCTTG +ATCTCA +ATCTCT +ATCTCC +ATCTCG +ATCTGA +ATCTGT +ATCTGC +ATCTGG +ATCCAA +ATCCAT +ATCCAC +ATCCAG +ATCCTA +ATCCTT +ATCCTC +ATCCTG +ATCCCA +ATCCCT +ATCCCC +ATCCCG +ATCCGA +ATCCGT +ATCCGC +ATCCGG +ATCGAA +ATCGAT +ATCGAC +ATCGAG +ATCGTA +ATCGTT +ATCGTC +ATCGTG +ATCGCA +ATCGCT +ATCGCC +ATCGCG +ATCGGA +ATCGGT +ATCGGC +ATCGGG +ATGAAA +ATGAAT +ATGAAC +ATGAAG +ATGATA +ATGATT +ATGATC +ATGATG +ATGACA +ATGACT +ATGACC +ATGACG +ATGAGA +ATGAGT +ATGAGC +ATGAGG +ATGTAA +ATGTAT +ATGTAC +ATGTAG +ATGTTA +ATGTTT +ATGTTC +ATGTTG +ATGTCA +ATGTCT +ATGTCC +ATGTCG +ATGTGA +ATGTGT +ATGTGC +ATGTGG +ATGCAA +ATGCAT +ATGCAC +ATGCAG +ATGCTA +ATGCTT +ATGCTC +ATGCTG +ATGCCA +ATGCCT +ATGCCC +ATGCCG +ATGCGA +ATGCGT +ATGCGC +ATGCGG +ATGGAA +ATGGAT +ATGGAC +ATGGAG +ATGGTA +ATGGTT +ATGGTC +ATGGTG +ATGGCA +ATGGCT +ATGGCC +ATGGCG +ATGGGA +ATGGGT +ATGGGC +ATGGGG +ACAAAA +ACAAAT +ACAAAC +ACAAAG +ACAATA +ACAATT +ACAATC +ACAATG +ACAACA +ACAACT +ACAACC +ACAACG +ACAAGA +ACAAGT +ACAAGC +ACAAGG +ACATAA +ACATAT +ACATAC +ACATAG +ACATTA +ACATTT +ACATTC +ACATTG +ACATCA +ACATCT +ACATCC +ACATCG +ACATGA +ACATGT +ACATGC +ACATGG +ACACAA +ACACAT +ACACAC +ACACAG +ACACTA +ACACTT +ACACTC +ACACTG +ACACCA +ACACCT +ACACCC +ACACCG +ACACGA +ACACGT +ACACGC +ACACGG +ACAGAA +ACAGAT +ACAGAC +ACAGAG +ACAGTA +ACAGTT +ACAGTC +ACAGTG +ACAGCA +ACAGCT +ACAGCC +ACAGCG +ACAGGA +ACAGGT +ACAGGC +ACAGGG +ACTAAA +ACTAAT +ACTAAC +ACTAAG +ACTATA +ACTATT +ACTATC +ACTATG +ACTACA +ACTACT +ACTACC +ACTACG +ACTAGA +ACTAGT +ACTAGC +ACTAGG +ACTTAA +ACTTAT +ACTTAC +ACTTAG +ACTTTA +ACTTTT +ACTTTC +ACTTTG +ACTTCA +ACTTCT +ACTTCC +ACTTCG +ACTTGA +ACTTGT +ACTTGC +ACTTGG +ACTCAA +ACTCAT +ACTCAC +ACTCAG +ACTCTA +ACTCTT +ACTCTC +ACTCTG +ACTCCA +ACTCCT +ACTCCC +ACTCCG +ACTCGA +ACTCGT +ACTCGC +ACTCGG +ACTGAA +ACTGAT +ACTGAC +ACTGAG +ACTGTA +ACTGTT +ACTGTC +ACTGTG +ACTGCA +ACTGCT +ACTGCC +ACTGCG +ACTGGA +ACTGGT +ACTGGC +ACTGGG +ACCAAA +ACCAAT +ACCAAC +ACCAAG +ACCATA +ACCATT +ACCATC +ACCATG +ACCACA +ACCACT +ACCACC +ACCACG +ACCAGA +ACCAGT +ACCAGC +ACCAGG +ACCTAA +ACCTAT +ACCTAC +ACCTAG +ACCTTA +ACCTTT +ACCTTC +ACCTTG +ACCTCA +ACCTCT +ACCTCC +ACCTCG +ACCTGA +ACCTGT +ACCTGC +ACCTGG +ACCCAA +ACCCAT +ACCCAC +ACCCAG +ACCCTA +ACCCTT +ACCCTC +ACCCTG +ACCCCA +ACCCCT +ACCCCC +ACCCCG +ACCCGA +ACCCGT +ACCCGC +ACCCGG +ACCGAA +ACCGAT +ACCGAC +ACCGAG +ACCGTA +ACCGTT +ACCGTC +ACCGTG +ACCGCA +ACCGCT +ACCGCC +ACCGCG +ACCGGA +ACCGGT +ACCGGC +ACCGGG +ACGAAA +ACGAAT +ACGAAC +ACGAAG +ACGATA +ACGATT +ACGATC +ACGATG +ACGACA +ACGACT +ACGACC +ACGACG +ACGAGA +ACGAGT +ACGAGC +ACGAGG +ACGTAA +ACGTAT +ACGTAC +ACGTAG +ACGTTA +ACGTTT +ACGTTC +ACGTTG +ACGTCA +ACGTCT +ACGTCC +ACGTCG +ACGTGA +ACGTGT +ACGTGC +ACGTGG +ACGCAA +ACGCAT +ACGCAC +ACGCAG +ACGCTA +ACGCTT +ACGCTC +ACGCTG +ACGCCA +ACGCCT +ACGCCC +ACGCCG +ACGCGA +ACGCGT +ACGCGC +ACGCGG +ACGGAA +ACGGAT +ACGGAC +ACGGAG +ACGGTA +ACGGTT +ACGGTC +ACGGTG +ACGGCA +ACGGCT +ACGGCC +ACGGCG +ACGGGA +ACGGGT +ACGGGC +ACGGGG +AGAAAA +AGAAAT +AGAAAC +AGAAAG +AGAATA +AGAATT +AGAATC +AGAATG +AGAACA +AGAACT +AGAACC +AGAACG +AGAAGA +AGAAGT +AGAAGC +AGAAGG +AGATAA +AGATAT +AGATAC +AGATAG +AGATTA +AGATTT +AGATTC +AGATTG +AGATCA +AGATCT +AGATCC +AGATCG +AGATGA +AGATGT +AGATGC +AGATGG +AGACAA +AGACAT +AGACAC +AGACAG +AGACTA +AGACTT +AGACTC +AGACTG +AGACCA +AGACCT +AGACCC +AGACCG +AGACGA +AGACGT +AGACGC +AGACGG +AGAGAA +AGAGAT +AGAGAC +AGAGAG +AGAGTA +AGAGTT +AGAGTC +AGAGTG +AGAGCA +AGAGCT +AGAGCC +AGAGCG +AGAGGA +AGAGGT +AGAGGC +AGAGGG +AGTAAA +AGTAAT +AGTAAC +AGTAAG +AGTATA +AGTATT +AGTATC +AGTATG +AGTACA +AGTACT +AGTACC +AGTACG +AGTAGA +AGTAGT +AGTAGC +AGTAGG +AGTTAA +AGTTAT +AGTTAC +AGTTAG +AGTTTA +AGTTTT +AGTTTC +AGTTTG +AGTTCA +AGTTCT +AGTTCC +AGTTCG +AGTTGA +AGTTGT +AGTTGC +AGTTGG +AGTCAA +AGTCAT +AGTCAC +AGTCAG +AGTCTA +AGTCTT +AGTCTC +AGTCTG +AGTCCA +AGTCCT +AGTCCC +AGTCCG +AGTCGA +AGTCGT +AGTCGC +AGTCGG +AGTGAA +AGTGAT +AGTGAC +AGTGAG +AGTGTA +AGTGTT +AGTGTC +AGTGTG +AGTGCA +AGTGCT +AGTGCC +AGTGCG +AGTGGA +AGTGGT +AGTGGC +AGTGGG +AGCAAA +AGCAAT +AGCAAC +AGCAAG +AGCATA +AGCATT +AGCATC +AGCATG +AGCACA +AGCACT +AGCACC +AGCACG +AGCAGA +AGCAGT +AGCAGC +AGCAGG +AGCTAA +AGCTAT +AGCTAC +AGCTAG +AGCTTA +AGCTTT +AGCTTC +AGCTTG +AGCTCA +AGCTCT +AGCTCC +AGCTCG +AGCTGA +AGCTGT +AGCTGC +AGCTGG +AGCCAA +AGCCAT +AGCCAC +AGCCAG +AGCCTA +AGCCTT +AGCCTC +AGCCTG +AGCCCA +AGCCCT +AGCCCC +AGCCCG +AGCCGA +AGCCGT +AGCCGC +AGCCGG +AGCGAA +AGCGAT +AGCGAC +AGCGAG +AGCGTA +AGCGTT +AGCGTC +AGCGTG +AGCGCA +AGCGCT +AGCGCC +AGCGCG +AGCGGA +AGCGGT +AGCGGC +AGCGGG +AGGAAA +AGGAAT +AGGAAC +AGGAAG +AGGATA +AGGATT +AGGATC +AGGATG +AGGACA +AGGACT +AGGACC +AGGACG +AGGAGA +AGGAGT +AGGAGC +AGGAGG +AGGTAA +AGGTAT +AGGTAC +AGGTAG +AGGTTA +AGGTTT +AGGTTC +AGGTTG +AGGTCA +AGGTCT +AGGTCC +AGGTCG +AGGTGA +AGGTGT +AGGTGC +AGGTGG +AGGCAA +AGGCAT +AGGCAC +AGGCAG +AGGCTA +AGGCTT +AGGCTC +AGGCTG +AGGCCA +AGGCCT +AGGCCC +AGGCCG +AGGCGA +AGGCGT +AGGCGC +AGGCGG +AGGGAA +AGGGAT +AGGGAC +AGGGAG +AGGGTA +AGGGTT +AGGGTC +AGGGTG +AGGGCA +AGGGCT +AGGGCC +AGGGCG +AGGGGA +AGGGGT +AGGGGC +AGGGGG +TAAAAA +TAAAAT +TAAAAC +TAAAAG +TAAATA +TAAATT +TAAATC +TAAATG +TAAACA +TAAACT +TAAACC +TAAACG +TAAAGA +TAAAGT +TAAAGC +TAAAGG +TAATAA +TAATAT +TAATAC +TAATAG +TAATTA +TAATTT +TAATTC +TAATTG +TAATCA +TAATCT +TAATCC +TAATCG +TAATGA +TAATGT +TAATGC +TAATGG +TAACAA +TAACAT +TAACAC +TAACAG +TAACTA +TAACTT +TAACTC +TAACTG +TAACCA +TAACCT +TAACCC +TAACCG +TAACGA +TAACGT +TAACGC +TAACGG +TAAGAA +TAAGAT +TAAGAC +TAAGAG +TAAGTA +TAAGTT +TAAGTC +TAAGTG +TAAGCA +TAAGCT +TAAGCC +TAAGCG +TAAGGA +TAAGGT +TAAGGC +TAAGGG +TATAAA +TATAAT +TATAAC +TATAAG +TATATA +TATATT +TATATC +TATATG +TATACA +TATACT +TATACC +TATACG +TATAGA +TATAGT +TATAGC +TATAGG +TATTAA +TATTAT +TATTAC +TATTAG +TATTTA +TATTTT +TATTTC +TATTTG +TATTCA +TATTCT +TATTCC +TATTCG +TATTGA +TATTGT +TATTGC +TATTGG +TATCAA +TATCAT +TATCAC +TATCAG +TATCTA +TATCTT +TATCTC +TATCTG +TATCCA +TATCCT +TATCCC +TATCCG +TATCGA +TATCGT +TATCGC +TATCGG +TATGAA +TATGAT +TATGAC +TATGAG +TATGTA +TATGTT +TATGTC +TATGTG +TATGCA +TATGCT +TATGCC +TATGCG +TATGGA +TATGGT +TATGGC +TATGGG +TACAAA +TACAAT +TACAAC +TACAAG +TACATA +TACATT +TACATC +TACATG +TACACA +TACACT +TACACC +TACACG +TACAGA +TACAGT +TACAGC +TACAGG +TACTAA +TACTAT +TACTAC +TACTAG +TACTTA +TACTTT +TACTTC +TACTTG +TACTCA +TACTCT +TACTCC +TACTCG +TACTGA +TACTGT +TACTGC +TACTGG +TACCAA +TACCAT +TACCAC +TACCAG +TACCTA +TACCTT +TACCTC +TACCTG +TACCCA +TACCCT +TACCCC +TACCCG +TACCGA +TACCGT +TACCGC +TACCGG +TACGAA +TACGAT +TACGAC +TACGAG +TACGTA +TACGTT +TACGTC +TACGTG +TACGCA +TACGCT +TACGCC +TACGCG +TACGGA +TACGGT +TACGGC +TACGGG +TAGAAA +TAGAAT +TAGAAC +TAGAAG +TAGATA +TAGATT +TAGATC +TAGATG +TAGACA +TAGACT +TAGACC +TAGACG +TAGAGA +TAGAGT +TAGAGC +TAGAGG +TAGTAA +TAGTAT +TAGTAC +TAGTAG +TAGTTA +TAGTTT +TAGTTC +TAGTTG +TAGTCA +TAGTCT +TAGTCC +TAGTCG +TAGTGA +TAGTGT +TAGTGC +TAGTGG +TAGCAA +TAGCAT +TAGCAC +TAGCAG +TAGCTA +TAGCTT +TAGCTC +TAGCTG +TAGCCA +TAGCCT +TAGCCC +TAGCCG +TAGCGA +TAGCGT +TAGCGC +TAGCGG +TAGGAA +TAGGAT +TAGGAC +TAGGAG +TAGGTA +TAGGTT +TAGGTC +TAGGTG +TAGGCA +TAGGCT +TAGGCC +TAGGCG +TAGGGA +TAGGGT +TAGGGC +TAGGGG +TTAAAA +TTAAAT +TTAAAC +TTAAAG +TTAATA +TTAATT +TTAATC +TTAATG +TTAACA +TTAACT +TTAACC +TTAACG +TTAAGA +TTAAGT +TTAAGC +TTAAGG +TTATAA +TTATAT +TTATAC +TTATAG +TTATTA +TTATTT +TTATTC +TTATTG +TTATCA +TTATCT +TTATCC +TTATCG +TTATGA +TTATGT +TTATGC +TTATGG +TTACAA +TTACAT +TTACAC +TTACAG +TTACTA +TTACTT +TTACTC +TTACTG +TTACCA +TTACCT +TTACCC +TTACCG +TTACGA +TTACGT +TTACGC +TTACGG +TTAGAA +TTAGAT +TTAGAC +TTAGAG +TTAGTA +TTAGTT +TTAGTC +TTAGTG +TTAGCA +TTAGCT +TTAGCC +TTAGCG +TTAGGA +TTAGGT +TTAGGC +TTAGGG +TTTAAA +TTTAAT +TTTAAC +TTTAAG +TTTATA +TTTATT +TTTATC +TTTATG +TTTACA +TTTACT +TTTACC +TTTACG +TTTAGA +TTTAGT +TTTAGC +TTTAGG +TTTTAA +TTTTAT +TTTTAC +TTTTAG +TTTTTA +TTTTTT +TTTTTC +TTTTTG +TTTTCA +TTTTCT +TTTTCC +TTTTCG +TTTTGA +TTTTGT +TTTTGC +TTTTGG +TTTCAA +TTTCAT +TTTCAC +TTTCAG +TTTCTA +TTTCTT +TTTCTC +TTTCTG +TTTCCA +TTTCCT +TTTCCC +TTTCCG +TTTCGA +TTTCGT +TTTCGC +TTTCGG +TTTGAA +TTTGAT +TTTGAC +TTTGAG +TTTGTA +TTTGTT +TTTGTC +TTTGTG +TTTGCA +TTTGCT +TTTGCC +TTTGCG +TTTGGA +TTTGGT +TTTGGC +TTTGGG +TTCAAA +TTCAAT +TTCAAC +TTCAAG +TTCATA +TTCATT +TTCATC +TTCATG +TTCACA +TTCACT +TTCACC +TTCACG +TTCAGA +TTCAGT +TTCAGC +TTCAGG +TTCTAA +TTCTAT +TTCTAC +TTCTAG +TTCTTA +TTCTTT +TTCTTC +TTCTTG +TTCTCA +TTCTCT +TTCTCC +TTCTCG +TTCTGA +TTCTGT +TTCTGC +TTCTGG +TTCCAA +TTCCAT +TTCCAC +TTCCAG +TTCCTA +TTCCTT +TTCCTC +TTCCTG +TTCCCA +TTCCCT +TTCCCC +TTCCCG +TTCCGA +TTCCGT +TTCCGC +TTCCGG +TTCGAA +TTCGAT +TTCGAC +TTCGAG +TTCGTA +TTCGTT +TTCGTC +TTCGTG +TTCGCA +TTCGCT +TTCGCC +TTCGCG +TTCGGA +TTCGGT +TTCGGC +TTCGGG +TTGAAA +TTGAAT +TTGAAC +TTGAAG +TTGATA +TTGATT +TTGATC +TTGATG +TTGACA +TTGACT +TTGACC +TTGACG +TTGAGA +TTGAGT +TTGAGC +TTGAGG +TTGTAA +TTGTAT +TTGTAC +TTGTAG +TTGTTA +TTGTTT +TTGTTC +TTGTTG +TTGTCA +TTGTCT +TTGTCC +TTGTCG +TTGTGA +TTGTGT +TTGTGC +TTGTGG +TTGCAA +TTGCAT +TTGCAC +TTGCAG +TTGCTA +TTGCTT +TTGCTC +TTGCTG +TTGCCA +TTGCCT +TTGCCC +TTGCCG +TTGCGA +TTGCGT +TTGCGC +TTGCGG +TTGGAA +TTGGAT +TTGGAC +TTGGAG +TTGGTA +TTGGTT +TTGGTC +TTGGTG +TTGGCA +TTGGCT +TTGGCC +TTGGCG +TTGGGA +TTGGGT +TTGGGC +TTGGGG +TCAAAA +TCAAAT +TCAAAC +TCAAAG +TCAATA +TCAATT +TCAATC +TCAATG +TCAACA +TCAACT +TCAACC +TCAACG +TCAAGA +TCAAGT +TCAAGC +TCAAGG +TCATAA +TCATAT +TCATAC +TCATAG +TCATTA +TCATTT +TCATTC +TCATTG +TCATCA +TCATCT +TCATCC +TCATCG +TCATGA +TCATGT +TCATGC +TCATGG +TCACAA +TCACAT +TCACAC +TCACAG +TCACTA +TCACTT +TCACTC +TCACTG +TCACCA +TCACCT +TCACCC +TCACCG +TCACGA +TCACGT +TCACGC +TCACGG +TCAGAA +TCAGAT +TCAGAC +TCAGAG +TCAGTA +TCAGTT +TCAGTC +TCAGTG +TCAGCA +TCAGCT +TCAGCC +TCAGCG +TCAGGA +TCAGGT +TCAGGC +TCAGGG +TCTAAA +TCTAAT +TCTAAC +TCTAAG +TCTATA +TCTATT +TCTATC +TCTATG +TCTACA +TCTACT +TCTACC +TCTACG +TCTAGA +TCTAGT +TCTAGC +TCTAGG +TCTTAA +TCTTAT +TCTTAC +TCTTAG +TCTTTA +TCTTTT +TCTTTC +TCTTTG +TCTTCA +TCTTCT +TCTTCC +TCTTCG +TCTTGA +TCTTGT +TCTTGC +TCTTGG +TCTCAA +TCTCAT +TCTCAC +TCTCAG +TCTCTA +TCTCTT +TCTCTC +TCTCTG +TCTCCA +TCTCCT +TCTCCC +TCTCCG +TCTCGA +TCTCGT +TCTCGC +TCTCGG +TCTGAA +TCTGAT +TCTGAC +TCTGAG +TCTGTA +TCTGTT +TCTGTC +TCTGTG +TCTGCA +TCTGCT +TCTGCC +TCTGCG +TCTGGA +TCTGGT +TCTGGC +TCTGGG +TCCAAA +TCCAAT +TCCAAC +TCCAAG +TCCATA +TCCATT +TCCATC +TCCATG +TCCACA +TCCACT +TCCACC +TCCACG +TCCAGA +TCCAGT +TCCAGC +TCCAGG +TCCTAA +TCCTAT +TCCTAC +TCCTAG +TCCTTA +TCCTTT +TCCTTC +TCCTTG +TCCTCA +TCCTCT +TCCTCC +TCCTCG +TCCTGA +TCCTGT +TCCTGC +TCCTGG +TCCCAA +TCCCAT +TCCCAC +TCCCAG +TCCCTA +TCCCTT +TCCCTC +TCCCTG +TCCCCA +TCCCCT +TCCCCC +TCCCCG +TCCCGA +TCCCGT +TCCCGC +TCCCGG +TCCGAA +TCCGAT +TCCGAC +TCCGAG +TCCGTA +TCCGTT +TCCGTC +TCCGTG +TCCGCA +TCCGCT +TCCGCC +TCCGCG +TCCGGA +TCCGGT +TCCGGC +TCCGGG +TCGAAA +TCGAAT +TCGAAC +TCGAAG +TCGATA +TCGATT +TCGATC +TCGATG +TCGACA +TCGACT +TCGACC +TCGACG +TCGAGA +TCGAGT +TCGAGC +TCGAGG +TCGTAA +TCGTAT +TCGTAC +TCGTAG +TCGTTA +TCGTTT +TCGTTC +TCGTTG +TCGTCA +TCGTCT +TCGTCC +TCGTCG +TCGTGA +TCGTGT +TCGTGC +TCGTGG +TCGCAA +TCGCAT +TCGCAC +TCGCAG +TCGCTA +TCGCTT +TCGCTC +TCGCTG +TCGCCA +TCGCCT +TCGCCC +TCGCCG +TCGCGA +TCGCGT +TCGCGC +TCGCGG +TCGGAA +TCGGAT +TCGGAC +TCGGAG +TCGGTA +TCGGTT +TCGGTC +TCGGTG +TCGGCA +TCGGCT +TCGGCC +TCGGCG +TCGGGA +TCGGGT +TCGGGC +TCGGGG +TGAAAA +TGAAAT +TGAAAC +TGAAAG +TGAATA +TGAATT +TGAATC +TGAATG +TGAACA +TGAACT +TGAACC +TGAACG +TGAAGA +TGAAGT +TGAAGC +TGAAGG +TGATAA +TGATAT +TGATAC +TGATAG +TGATTA +TGATTT +TGATTC +TGATTG +TGATCA +TGATCT +TGATCC +TGATCG +TGATGA +TGATGT +TGATGC +TGATGG +TGACAA +TGACAT +TGACAC +TGACAG +TGACTA +TGACTT +TGACTC +TGACTG +TGACCA +TGACCT +TGACCC +TGACCG +TGACGA +TGACGT +TGACGC +TGACGG +TGAGAA +TGAGAT +TGAGAC +TGAGAG +TGAGTA +TGAGTT +TGAGTC +TGAGTG +TGAGCA +TGAGCT +TGAGCC +TGAGCG +TGAGGA +TGAGGT +TGAGGC +TGAGGG +TGTAAA +TGTAAT +TGTAAC +TGTAAG +TGTATA +TGTATT +TGTATC +TGTATG +TGTACA +TGTACT +TGTACC +TGTACG +TGTAGA +TGTAGT +TGTAGC +TGTAGG +TGTTAA +TGTTAT +TGTTAC +TGTTAG +TGTTTA +TGTTTT +TGTTTC +TGTTTG +TGTTCA +TGTTCT +TGTTCC +TGTTCG +TGTTGA +TGTTGT +TGTTGC +TGTTGG +TGTCAA +TGTCAT +TGTCAC +TGTCAG +TGTCTA +TGTCTT +TGTCTC +TGTCTG +TGTCCA +TGTCCT +TGTCCC +TGTCCG +TGTCGA +TGTCGT +TGTCGC +TGTCGG +TGTGAA +TGTGAT +TGTGAC +TGTGAG +TGTGTA +TGTGTT +TGTGTC +TGTGTG +TGTGCA +TGTGCT +TGTGCC +TGTGCG +TGTGGA +TGTGGT +TGTGGC +TGTGGG +TGCAAA +TGCAAT +TGCAAC +TGCAAG +TGCATA +TGCATT +TGCATC +TGCATG +TGCACA +TGCACT +TGCACC +TGCACG +TGCAGA +TGCAGT +TGCAGC +TGCAGG +TGCTAA +TGCTAT +TGCTAC +TGCTAG +TGCTTA +TGCTTT +TGCTTC +TGCTTG +TGCTCA +TGCTCT +TGCTCC +TGCTCG +TGCTGA +TGCTGT +TGCTGC +TGCTGG +TGCCAA +TGCCAT +TGCCAC +TGCCAG +TGCCTA +TGCCTT +TGCCTC +TGCCTG +TGCCCA +TGCCCT +TGCCCC +TGCCCG +TGCCGA +TGCCGT +TGCCGC +TGCCGG +TGCGAA +TGCGAT +TGCGAC +TGCGAG +TGCGTA +TGCGTT +TGCGTC +TGCGTG +TGCGCA +TGCGCT +TGCGCC +TGCGCG +TGCGGA +TGCGGT +TGCGGC +TGCGGG +TGGAAA +TGGAAT +TGGAAC +TGGAAG +TGGATA +TGGATT +TGGATC +TGGATG +TGGACA +TGGACT +TGGACC +TGGACG +TGGAGA +TGGAGT +TGGAGC +TGGAGG +TGGTAA +TGGTAT +TGGTAC +TGGTAG +TGGTTA +TGGTTT +TGGTTC +TGGTTG +TGGTCA +TGGTCT +TGGTCC +TGGTCG +TGGTGA +TGGTGT +TGGTGC +TGGTGG +TGGCAA +TGGCAT +TGGCAC +TGGCAG +TGGCTA +TGGCTT +TGGCTC +TGGCTG +TGGCCA +TGGCCT +TGGCCC +TGGCCG +TGGCGA +TGGCGT +TGGCGC +TGGCGG +TGGGAA +TGGGAT +TGGGAC +TGGGAG +TGGGTA +TGGGTT +TGGGTC +TGGGTG +TGGGCA +TGGGCT +TGGGCC +TGGGCG +TGGGGA +TGGGGT +TGGGGC +TGGGGG +CAAAAA +CAAAAT +CAAAAC +CAAAAG +CAAATA +CAAATT +CAAATC +CAAATG +CAAACA +CAAACT +CAAACC +CAAACG +CAAAGA +CAAAGT +CAAAGC +CAAAGG +CAATAA +CAATAT +CAATAC +CAATAG +CAATTA +CAATTT +CAATTC +CAATTG +CAATCA +CAATCT +CAATCC +CAATCG +CAATGA +CAATGT +CAATGC +CAATGG +CAACAA +CAACAT +CAACAC +CAACAG +CAACTA +CAACTT +CAACTC +CAACTG +CAACCA +CAACCT +CAACCC +CAACCG +CAACGA +CAACGT +CAACGC +CAACGG +CAAGAA +CAAGAT +CAAGAC +CAAGAG +CAAGTA +CAAGTT +CAAGTC +CAAGTG +CAAGCA +CAAGCT +CAAGCC +CAAGCG +CAAGGA +CAAGGT +CAAGGC +CAAGGG +CATAAA +CATAAT +CATAAC +CATAAG +CATATA +CATATT +CATATC +CATATG +CATACA +CATACT +CATACC +CATACG +CATAGA +CATAGT +CATAGC +CATAGG +CATTAA +CATTAT +CATTAC +CATTAG +CATTTA +CATTTT +CATTTC +CATTTG +CATTCA +CATTCT +CATTCC +CATTCG +CATTGA +CATTGT +CATTGC +CATTGG +CATCAA +CATCAT +CATCAC +CATCAG +CATCTA +CATCTT +CATCTC +CATCTG +CATCCA +CATCCT +CATCCC +CATCCG +CATCGA +CATCGT +CATCGC +CATCGG +CATGAA +CATGAT +CATGAC +CATGAG +CATGTA +CATGTT +CATGTC +CATGTG +CATGCA +CATGCT +CATGCC +CATGCG +CATGGA +CATGGT +CATGGC +CATGGG +CACAAA +CACAAT +CACAAC +CACAAG +CACATA +CACATT +CACATC +CACATG +CACACA +CACACT +CACACC +CACACG +CACAGA +CACAGT +CACAGC +CACAGG +CACTAA +CACTAT +CACTAC +CACTAG +CACTTA +CACTTT +CACTTC +CACTTG +CACTCA +CACTCT +CACTCC +CACTCG +CACTGA +CACTGT +CACTGC +CACTGG +CACCAA +CACCAT +CACCAC +CACCAG +CACCTA +CACCTT +CACCTC +CACCTG +CACCCA +CACCCT +CACCCC +CACCCG +CACCGA +CACCGT +CACCGC +CACCGG +CACGAA +CACGAT +CACGAC +CACGAG +CACGTA +CACGTT +CACGTC +CACGTG +CACGCA +CACGCT +CACGCC +CACGCG +CACGGA +CACGGT +CACGGC +CACGGG +CAGAAA +CAGAAT +CAGAAC +CAGAAG +CAGATA +CAGATT +CAGATC +CAGATG +CAGACA +CAGACT +CAGACC +CAGACG +CAGAGA +CAGAGT +CAGAGC +CAGAGG +CAGTAA +CAGTAT +CAGTAC +CAGTAG +CAGTTA +CAGTTT +CAGTTC +CAGTTG +CAGTCA +CAGTCT +CAGTCC +CAGTCG +CAGTGA +CAGTGT +CAGTGC +CAGTGG +CAGCAA +CAGCAT +CAGCAC +CAGCAG +CAGCTA +CAGCTT +CAGCTC +CAGCTG +CAGCCA +CAGCCT +CAGCCC +CAGCCG +CAGCGA +CAGCGT +CAGCGC +CAGCGG +CAGGAA +CAGGAT +CAGGAC +CAGGAG +CAGGTA +CAGGTT +CAGGTC +CAGGTG +CAGGCA +CAGGCT +CAGGCC +CAGGCG +CAGGGA +CAGGGT +CAGGGC +CAGGGG +CTAAAA +CTAAAT +CTAAAC +CTAAAG +CTAATA +CTAATT +CTAATC +CTAATG +CTAACA +CTAACT +CTAACC +CTAACG +CTAAGA +CTAAGT +CTAAGC +CTAAGG +CTATAA +CTATAT +CTATAC +CTATAG +CTATTA +CTATTT +CTATTC +CTATTG +CTATCA +CTATCT +CTATCC +CTATCG +CTATGA +CTATGT +CTATGC +CTATGG +CTACAA +CTACAT +CTACAC +CTACAG +CTACTA +CTACTT +CTACTC +CTACTG +CTACCA +CTACCT +CTACCC +CTACCG +CTACGA +CTACGT +CTACGC +CTACGG +CTAGAA +CTAGAT +CTAGAC +CTAGAG +CTAGTA +CTAGTT +CTAGTC +CTAGTG +CTAGCA +CTAGCT +CTAGCC +CTAGCG +CTAGGA +CTAGGT +CTAGGC +CTAGGG +CTTAAA +CTTAAT +CTTAAC +CTTAAG +CTTATA +CTTATT +CTTATC +CTTATG +CTTACA +CTTACT +CTTACC +CTTACG +CTTAGA +CTTAGT +CTTAGC +CTTAGG +CTTTAA +CTTTAT +CTTTAC +CTTTAG +CTTTTA +CTTTTT +CTTTTC +CTTTTG +CTTTCA +CTTTCT +CTTTCC +CTTTCG +CTTTGA +CTTTGT +CTTTGC +CTTTGG +CTTCAA +CTTCAT +CTTCAC +CTTCAG +CTTCTA +CTTCTT +CTTCTC +CTTCTG +CTTCCA +CTTCCT +CTTCCC +CTTCCG +CTTCGA +CTTCGT +CTTCGC +CTTCGG +CTTGAA +CTTGAT +CTTGAC +CTTGAG +CTTGTA +CTTGTT +CTTGTC +CTTGTG +CTTGCA +CTTGCT +CTTGCC +CTTGCG +CTTGGA +CTTGGT +CTTGGC +CTTGGG +CTCAAA +CTCAAT +CTCAAC +CTCAAG +CTCATA +CTCATT +CTCATC +CTCATG +CTCACA +CTCACT +CTCACC +CTCACG +CTCAGA +CTCAGT +CTCAGC +CTCAGG +CTCTAA +CTCTAT +CTCTAC +CTCTAG +CTCTTA +CTCTTT +CTCTTC +CTCTTG +CTCTCA +CTCTCT +CTCTCC +CTCTCG +CTCTGA +CTCTGT +CTCTGC +CTCTGG +CTCCAA +CTCCAT +CTCCAC +CTCCAG +CTCCTA +CTCCTT +CTCCTC +CTCCTG +CTCCCA +CTCCCT +CTCCCC +CTCCCG +CTCCGA +CTCCGT +CTCCGC +CTCCGG +CTCGAA +CTCGAT +CTCGAC +CTCGAG +CTCGTA +CTCGTT +CTCGTC +CTCGTG +CTCGCA +CTCGCT +CTCGCC +CTCGCG +CTCGGA +CTCGGT +CTCGGC +CTCGGG +CTGAAA +CTGAAT +CTGAAC +CTGAAG +CTGATA +CTGATT +CTGATC +CTGATG +CTGACA +CTGACT +CTGACC +CTGACG +CTGAGA +CTGAGT +CTGAGC +CTGAGG +CTGTAA +CTGTAT +CTGTAC +CTGTAG +CTGTTA +CTGTTT +CTGTTC +CTGTTG +CTGTCA +CTGTCT +CTGTCC +CTGTCG +CTGTGA +CTGTGT +CTGTGC +CTGTGG +CTGCAA +CTGCAT +CTGCAC +CTGCAG +CTGCTA +CTGCTT +CTGCTC +CTGCTG +CTGCCA +CTGCCT +CTGCCC +CTGCCG +CTGCGA +CTGCGT +CTGCGC +CTGCGG +CTGGAA +CTGGAT +CTGGAC +CTGGAG +CTGGTA +CTGGTT +CTGGTC +CTGGTG +CTGGCA +CTGGCT +CTGGCC +CTGGCG +CTGGGA +CTGGGT +CTGGGC +CTGGGG +CCAAAA +CCAAAT +CCAAAC +CCAAAG +CCAATA +CCAATT +CCAATC +CCAATG +CCAACA +CCAACT +CCAACC +CCAACG +CCAAGA +CCAAGT +CCAAGC +CCAAGG +CCATAA +CCATAT +CCATAC +CCATAG +CCATTA +CCATTT +CCATTC +CCATTG +CCATCA +CCATCT +CCATCC +CCATCG +CCATGA +CCATGT +CCATGC +CCATGG +CCACAA +CCACAT +CCACAC +CCACAG +CCACTA +CCACTT +CCACTC +CCACTG +CCACCA +CCACCT +CCACCC +CCACCG +CCACGA +CCACGT +CCACGC +CCACGG +CCAGAA +CCAGAT +CCAGAC +CCAGAG +CCAGTA +CCAGTT +CCAGTC +CCAGTG +CCAGCA +CCAGCT +CCAGCC +CCAGCG +CCAGGA +CCAGGT +CCAGGC +CCAGGG +CCTAAA +CCTAAT +CCTAAC +CCTAAG +CCTATA +CCTATT +CCTATC +CCTATG +CCTACA +CCTACT +CCTACC +CCTACG +CCTAGA +CCTAGT +CCTAGC +CCTAGG +CCTTAA +CCTTAT +CCTTAC +CCTTAG +CCTTTA +CCTTTT +CCTTTC +CCTTTG +CCTTCA +CCTTCT +CCTTCC +CCTTCG +CCTTGA +CCTTGT +CCTTGC +CCTTGG +CCTCAA +CCTCAT +CCTCAC +CCTCAG +CCTCTA +CCTCTT +CCTCTC +CCTCTG +CCTCCA +CCTCCT +CCTCCC +CCTCCG +CCTCGA +CCTCGT +CCTCGC +CCTCGG +CCTGAA +CCTGAT +CCTGAC +CCTGAG +CCTGTA +CCTGTT +CCTGTC +CCTGTG +CCTGCA +CCTGCT +CCTGCC +CCTGCG +CCTGGA +CCTGGT +CCTGGC +CCTGGG +CCCAAA +CCCAAT +CCCAAC +CCCAAG +CCCATA +CCCATT +CCCATC +CCCATG +CCCACA +CCCACT +CCCACC +CCCACG +CCCAGA +CCCAGT +CCCAGC +CCCAGG +CCCTAA +CCCTAT +CCCTAC +CCCTAG +CCCTTA +CCCTTT +CCCTTC +CCCTTG +CCCTCA +CCCTCT +CCCTCC +CCCTCG +CCCTGA +CCCTGT +CCCTGC +CCCTGG +CCCCAA +CCCCAT +CCCCAC +CCCCAG +CCCCTA +CCCCTT +CCCCTC +CCCCTG +CCCCCA +CCCCCT +CCCCCC +CCCCCG +CCCCGA +CCCCGT +CCCCGC +CCCCGG +CCCGAA +CCCGAT +CCCGAC +CCCGAG +CCCGTA +CCCGTT +CCCGTC +CCCGTG +CCCGCA +CCCGCT +CCCGCC +CCCGCG +CCCGGA +CCCGGT +CCCGGC +CCCGGG +CCGAAA +CCGAAT +CCGAAC +CCGAAG +CCGATA +CCGATT +CCGATC +CCGATG +CCGACA +CCGACT +CCGACC +CCGACG +CCGAGA +CCGAGT +CCGAGC +CCGAGG +CCGTAA +CCGTAT +CCGTAC +CCGTAG +CCGTTA +CCGTTT +CCGTTC +CCGTTG +CCGTCA +CCGTCT +CCGTCC +CCGTCG +CCGTGA +CCGTGT +CCGTGC +CCGTGG +CCGCAA +CCGCAT +CCGCAC +CCGCAG +CCGCTA +CCGCTT +CCGCTC +CCGCTG +CCGCCA +CCGCCT +CCGCCC +CCGCCG +CCGCGA +CCGCGT +CCGCGC +CCGCGG +CCGGAA +CCGGAT +CCGGAC +CCGGAG +CCGGTA +CCGGTT +CCGGTC +CCGGTG +CCGGCA +CCGGCT +CCGGCC +CCGGCG +CCGGGA +CCGGGT +CCGGGC +CCGGGG +CGAAAA +CGAAAT +CGAAAC +CGAAAG +CGAATA +CGAATT +CGAATC +CGAATG +CGAACA +CGAACT +CGAACC +CGAACG +CGAAGA +CGAAGT +CGAAGC +CGAAGG +CGATAA +CGATAT +CGATAC +CGATAG +CGATTA +CGATTT +CGATTC +CGATTG +CGATCA +CGATCT +CGATCC +CGATCG +CGATGA +CGATGT +CGATGC +CGATGG +CGACAA +CGACAT +CGACAC +CGACAG +CGACTA +CGACTT +CGACTC +CGACTG +CGACCA +CGACCT +CGACCC +CGACCG +CGACGA +CGACGT +CGACGC +CGACGG +CGAGAA +CGAGAT +CGAGAC +CGAGAG +CGAGTA +CGAGTT +CGAGTC +CGAGTG +CGAGCA +CGAGCT +CGAGCC +CGAGCG +CGAGGA +CGAGGT +CGAGGC +CGAGGG +CGTAAA +CGTAAT +CGTAAC +CGTAAG +CGTATA +CGTATT +CGTATC +CGTATG +CGTACA +CGTACT +CGTACC +CGTACG +CGTAGA +CGTAGT +CGTAGC +CGTAGG +CGTTAA +CGTTAT +CGTTAC +CGTTAG +CGTTTA +CGTTTT +CGTTTC +CGTTTG +CGTTCA +CGTTCT +CGTTCC +CGTTCG +CGTTGA +CGTTGT +CGTTGC +CGTTGG +CGTCAA +CGTCAT +CGTCAC +CGTCAG +CGTCTA +CGTCTT +CGTCTC +CGTCTG +CGTCCA +CGTCCT +CGTCCC +CGTCCG +CGTCGA +CGTCGT +CGTCGC +CGTCGG +CGTGAA +CGTGAT +CGTGAC +CGTGAG +CGTGTA +CGTGTT +CGTGTC +CGTGTG +CGTGCA +CGTGCT +CGTGCC +CGTGCG +CGTGGA +CGTGGT +CGTGGC +CGTGGG +CGCAAA +CGCAAT +CGCAAC +CGCAAG +CGCATA +CGCATT +CGCATC +CGCATG +CGCACA +CGCACT +CGCACC +CGCACG +CGCAGA +CGCAGT +CGCAGC +CGCAGG +CGCTAA +CGCTAT +CGCTAC +CGCTAG +CGCTTA +CGCTTT +CGCTTC +CGCTTG +CGCTCA +CGCTCT +CGCTCC +CGCTCG +CGCTGA +CGCTGT +CGCTGC +CGCTGG +CGCCAA +CGCCAT +CGCCAC +CGCCAG +CGCCTA +CGCCTT +CGCCTC +CGCCTG +CGCCCA +CGCCCT +CGCCCC +CGCCCG +CGCCGA +CGCCGT +CGCCGC +CGCCGG +CGCGAA +CGCGAT +CGCGAC +CGCGAG +CGCGTA +CGCGTT +CGCGTC +CGCGTG +CGCGCA +CGCGCT +CGCGCC +CGCGCG +CGCGGA +CGCGGT +CGCGGC +CGCGGG +CGGAAA +CGGAAT +CGGAAC +CGGAAG +CGGATA +CGGATT +CGGATC +CGGATG +CGGACA +CGGACT +CGGACC +CGGACG +CGGAGA +CGGAGT +CGGAGC +CGGAGG +CGGTAA +CGGTAT +CGGTAC +CGGTAG +CGGTTA +CGGTTT +CGGTTC +CGGTTG +CGGTCA +CGGTCT +CGGTCC +CGGTCG +CGGTGA +CGGTGT +CGGTGC +CGGTGG +CGGCAA +CGGCAT +CGGCAC +CGGCAG +CGGCTA +CGGCTT +CGGCTC +CGGCTG +CGGCCA +CGGCCT +CGGCCC +CGGCCG +CGGCGA +CGGCGT +CGGCGC +CGGCGG +CGGGAA +CGGGAT +CGGGAC +CGGGAG +CGGGTA +CGGGTT +CGGGTC +CGGGTG +CGGGCA +CGGGCT +CGGGCC +CGGGCG +CGGGGA +CGGGGT +CGGGGC +CGGGGG +GAAAAA +GAAAAT +GAAAAC +GAAAAG +GAAATA +GAAATT +GAAATC +GAAATG +GAAACA +GAAACT +GAAACC +GAAACG +GAAAGA +GAAAGT +GAAAGC +GAAAGG +GAATAA +GAATAT +GAATAC +GAATAG +GAATTA +GAATTT +GAATTC +GAATTG +GAATCA +GAATCT +GAATCC +GAATCG +GAATGA +GAATGT +GAATGC +GAATGG +GAACAA +GAACAT +GAACAC +GAACAG +GAACTA +GAACTT +GAACTC +GAACTG +GAACCA +GAACCT +GAACCC +GAACCG +GAACGA +GAACGT +GAACGC +GAACGG +GAAGAA +GAAGAT +GAAGAC +GAAGAG +GAAGTA +GAAGTT +GAAGTC +GAAGTG +GAAGCA +GAAGCT +GAAGCC +GAAGCG +GAAGGA +GAAGGT +GAAGGC +GAAGGG +GATAAA +GATAAT +GATAAC +GATAAG +GATATA +GATATT +GATATC +GATATG +GATACA +GATACT +GATACC +GATACG +GATAGA +GATAGT +GATAGC +GATAGG +GATTAA +GATTAT +GATTAC +GATTAG +GATTTA +GATTTT +GATTTC +GATTTG +GATTCA +GATTCT +GATTCC +GATTCG +GATTGA +GATTGT +GATTGC +GATTGG +GATCAA +GATCAT +GATCAC +GATCAG +GATCTA +GATCTT +GATCTC +GATCTG +GATCCA +GATCCT +GATCCC +GATCCG +GATCGA +GATCGT +GATCGC +GATCGG +GATGAA +GATGAT +GATGAC +GATGAG +GATGTA +GATGTT +GATGTC +GATGTG +GATGCA +GATGCT +GATGCC +GATGCG +GATGGA +GATGGT +GATGGC +GATGGG +GACAAA +GACAAT +GACAAC +GACAAG +GACATA +GACATT +GACATC +GACATG +GACACA +GACACT +GACACC +GACACG +GACAGA +GACAGT +GACAGC +GACAGG +GACTAA +GACTAT +GACTAC +GACTAG +GACTTA +GACTTT +GACTTC +GACTTG +GACTCA +GACTCT +GACTCC +GACTCG +GACTGA +GACTGT +GACTGC +GACTGG +GACCAA +GACCAT +GACCAC +GACCAG +GACCTA +GACCTT +GACCTC +GACCTG +GACCCA +GACCCT +GACCCC +GACCCG +GACCGA +GACCGT +GACCGC +GACCGG +GACGAA +GACGAT +GACGAC +GACGAG +GACGTA +GACGTT +GACGTC +GACGTG +GACGCA +GACGCT +GACGCC +GACGCG +GACGGA +GACGGT +GACGGC +GACGGG +GAGAAA +GAGAAT +GAGAAC +GAGAAG +GAGATA +GAGATT +GAGATC +GAGATG +GAGACA +GAGACT +GAGACC +GAGACG +GAGAGA +GAGAGT +GAGAGC +GAGAGG +GAGTAA +GAGTAT +GAGTAC +GAGTAG +GAGTTA +GAGTTT +GAGTTC +GAGTTG +GAGTCA +GAGTCT +GAGTCC +GAGTCG +GAGTGA +GAGTGT +GAGTGC +GAGTGG +GAGCAA +GAGCAT +GAGCAC +GAGCAG +GAGCTA +GAGCTT +GAGCTC +GAGCTG +GAGCCA +GAGCCT +GAGCCC +GAGCCG +GAGCGA +GAGCGT +GAGCGC +GAGCGG +GAGGAA +GAGGAT +GAGGAC +GAGGAG +GAGGTA +GAGGTT +GAGGTC +GAGGTG +GAGGCA +GAGGCT +GAGGCC +GAGGCG +GAGGGA +GAGGGT +GAGGGC +GAGGGG +GTAAAA +GTAAAT +GTAAAC +GTAAAG +GTAATA +GTAATT +GTAATC +GTAATG +GTAACA +GTAACT +GTAACC +GTAACG +GTAAGA +GTAAGT +GTAAGC +GTAAGG +GTATAA +GTATAT +GTATAC +GTATAG +GTATTA +GTATTT +GTATTC +GTATTG +GTATCA +GTATCT +GTATCC +GTATCG +GTATGA +GTATGT +GTATGC +GTATGG +GTACAA +GTACAT +GTACAC +GTACAG +GTACTA +GTACTT +GTACTC +GTACTG +GTACCA +GTACCT +GTACCC +GTACCG +GTACGA +GTACGT +GTACGC +GTACGG +GTAGAA +GTAGAT +GTAGAC +GTAGAG +GTAGTA +GTAGTT +GTAGTC +GTAGTG +GTAGCA +GTAGCT +GTAGCC +GTAGCG +GTAGGA +GTAGGT +GTAGGC +GTAGGG +GTTAAA +GTTAAT +GTTAAC +GTTAAG +GTTATA +GTTATT +GTTATC +GTTATG +GTTACA +GTTACT +GTTACC +GTTACG +GTTAGA +GTTAGT +GTTAGC +GTTAGG +GTTTAA +GTTTAT +GTTTAC +GTTTAG +GTTTTA +GTTTTT +GTTTTC +GTTTTG +GTTTCA +GTTTCT +GTTTCC +GTTTCG +GTTTGA +GTTTGT +GTTTGC +GTTTGG +GTTCAA +GTTCAT +GTTCAC +GTTCAG +GTTCTA +GTTCTT +GTTCTC +GTTCTG +GTTCCA +GTTCCT +GTTCCC +GTTCCG +GTTCGA +GTTCGT +GTTCGC +GTTCGG +GTTGAA +GTTGAT +GTTGAC +GTTGAG +GTTGTA +GTTGTT +GTTGTC +GTTGTG +GTTGCA +GTTGCT +GTTGCC +GTTGCG +GTTGGA +GTTGGT +GTTGGC +GTTGGG +GTCAAA +GTCAAT +GTCAAC +GTCAAG +GTCATA +GTCATT +GTCATC +GTCATG +GTCACA +GTCACT +GTCACC +GTCACG +GTCAGA +GTCAGT +GTCAGC +GTCAGG +GTCTAA +GTCTAT +GTCTAC +GTCTAG +GTCTTA +GTCTTT +GTCTTC +GTCTTG +GTCTCA +GTCTCT +GTCTCC +GTCTCG +GTCTGA +GTCTGT +GTCTGC +GTCTGG +GTCCAA +GTCCAT +GTCCAC +GTCCAG +GTCCTA +GTCCTT +GTCCTC +GTCCTG +GTCCCA +GTCCCT +GTCCCC +GTCCCG +GTCCGA +GTCCGT +GTCCGC +GTCCGG +GTCGAA +GTCGAT +GTCGAC +GTCGAG +GTCGTA +GTCGTT +GTCGTC +GTCGTG +GTCGCA +GTCGCT +GTCGCC +GTCGCG +GTCGGA +GTCGGT +GTCGGC +GTCGGG +GTGAAA +GTGAAT +GTGAAC +GTGAAG +GTGATA +GTGATT +GTGATC +GTGATG +GTGACA +GTGACT +GTGACC +GTGACG +GTGAGA +GTGAGT +GTGAGC +GTGAGG +GTGTAA +GTGTAT +GTGTAC +GTGTAG +GTGTTA +GTGTTT +GTGTTC +GTGTTG +GTGTCA +GTGTCT +GTGTCC +GTGTCG +GTGTGA +GTGTGT +GTGTGC +GTGTGG +GTGCAA +GTGCAT +GTGCAC +GTGCAG +GTGCTA +GTGCTT +GTGCTC +GTGCTG +GTGCCA +GTGCCT +GTGCCC +GTGCCG +GTGCGA +GTGCGT +GTGCGC +GTGCGG +GTGGAA +GTGGAT +GTGGAC +GTGGAG +GTGGTA +GTGGTT +GTGGTC +GTGGTG +GTGGCA +GTGGCT +GTGGCC +GTGGCG +GTGGGA +GTGGGT +GTGGGC +GTGGGG +GCAAAA +GCAAAT +GCAAAC +GCAAAG +GCAATA +GCAATT +GCAATC +GCAATG +GCAACA +GCAACT +GCAACC +GCAACG +GCAAGA +GCAAGT +GCAAGC +GCAAGG +GCATAA +GCATAT +GCATAC +GCATAG +GCATTA +GCATTT +GCATTC +GCATTG +GCATCA +GCATCT +GCATCC +GCATCG +GCATGA +GCATGT +GCATGC +GCATGG +GCACAA +GCACAT +GCACAC +GCACAG +GCACTA +GCACTT +GCACTC +GCACTG +GCACCA +GCACCT +GCACCC +GCACCG +GCACGA +GCACGT +GCACGC +GCACGG +GCAGAA +GCAGAT +GCAGAC +GCAGAG +GCAGTA +GCAGTT +GCAGTC +GCAGTG +GCAGCA +GCAGCT +GCAGCC +GCAGCG +GCAGGA +GCAGGT +GCAGGC +GCAGGG +GCTAAA +GCTAAT +GCTAAC +GCTAAG +GCTATA +GCTATT +GCTATC +GCTATG +GCTACA +GCTACT +GCTACC +GCTACG +GCTAGA +GCTAGT +GCTAGC +GCTAGG +GCTTAA +GCTTAT +GCTTAC +GCTTAG +GCTTTA +GCTTTT +GCTTTC +GCTTTG +GCTTCA +GCTTCT +GCTTCC +GCTTCG +GCTTGA +GCTTGT +GCTTGC +GCTTGG +GCTCAA +GCTCAT +GCTCAC +GCTCAG +GCTCTA +GCTCTT +GCTCTC +GCTCTG +GCTCCA +GCTCCT +GCTCCC +GCTCCG +GCTCGA +GCTCGT +GCTCGC +GCTCGG +GCTGAA +GCTGAT +GCTGAC +GCTGAG +GCTGTA +GCTGTT +GCTGTC +GCTGTG +GCTGCA +GCTGCT +GCTGCC +GCTGCG +GCTGGA +GCTGGT +GCTGGC +GCTGGG +GCCAAA +GCCAAT +GCCAAC +GCCAAG +GCCATA +GCCATT +GCCATC +GCCATG +GCCACA +GCCACT +GCCACC +GCCACG +GCCAGA +GCCAGT +GCCAGC +GCCAGG +GCCTAA +GCCTAT +GCCTAC +GCCTAG +GCCTTA +GCCTTT +GCCTTC +GCCTTG +GCCTCA +GCCTCT +GCCTCC +GCCTCG +GCCTGA +GCCTGT +GCCTGC +GCCTGG +GCCCAA +GCCCAT +GCCCAC +GCCCAG +GCCCTA +GCCCTT +GCCCTC +GCCCTG +GCCCCA +GCCCCT +GCCCCC +GCCCCG +GCCCGA +GCCCGT +GCCCGC +GCCCGG +GCCGAA +GCCGAT +GCCGAC +GCCGAG +GCCGTA +GCCGTT +GCCGTC +GCCGTG +GCCGCA +GCCGCT +GCCGCC +GCCGCG +GCCGGA +GCCGGT +GCCGGC +GCCGGG +GCGAAA +GCGAAT +GCGAAC +GCGAAG +GCGATA +GCGATT +GCGATC +GCGATG +GCGACA +GCGACT +GCGACC +GCGACG +GCGAGA +GCGAGT +GCGAGC +GCGAGG +GCGTAA +GCGTAT +GCGTAC +GCGTAG +GCGTTA +GCGTTT +GCGTTC +GCGTTG +GCGTCA +GCGTCT +GCGTCC +GCGTCG +GCGTGA +GCGTGT +GCGTGC +GCGTGG +GCGCAA +GCGCAT +GCGCAC +GCGCAG +GCGCTA +GCGCTT +GCGCTC +GCGCTG +GCGCCA +GCGCCT +GCGCCC +GCGCCG +GCGCGA +GCGCGT +GCGCGC +GCGCGG +GCGGAA +GCGGAT +GCGGAC +GCGGAG +GCGGTA +GCGGTT +GCGGTC +GCGGTG +GCGGCA +GCGGCT +GCGGCC +GCGGCG +GCGGGA +GCGGGT +GCGGGC +GCGGGG +GGAAAA +GGAAAT +GGAAAC +GGAAAG +GGAATA +GGAATT +GGAATC +GGAATG +GGAACA +GGAACT +GGAACC +GGAACG +GGAAGA +GGAAGT +GGAAGC +GGAAGG +GGATAA +GGATAT +GGATAC +GGATAG +GGATTA +GGATTT +GGATTC +GGATTG +GGATCA +GGATCT +GGATCC +GGATCG +GGATGA +GGATGT +GGATGC +GGATGG +GGACAA +GGACAT +GGACAC +GGACAG +GGACTA +GGACTT +GGACTC +GGACTG +GGACCA +GGACCT +GGACCC +GGACCG +GGACGA +GGACGT +GGACGC +GGACGG +GGAGAA +GGAGAT +GGAGAC +GGAGAG +GGAGTA +GGAGTT +GGAGTC +GGAGTG +GGAGCA +GGAGCT +GGAGCC +GGAGCG +GGAGGA +GGAGGT +GGAGGC +GGAGGG +GGTAAA +GGTAAT +GGTAAC +GGTAAG +GGTATA +GGTATT +GGTATC +GGTATG +GGTACA +GGTACT +GGTACC +GGTACG +GGTAGA +GGTAGT +GGTAGC +GGTAGG +GGTTAA +GGTTAT +GGTTAC +GGTTAG +GGTTTA +GGTTTT +GGTTTC +GGTTTG +GGTTCA +GGTTCT +GGTTCC +GGTTCG +GGTTGA +GGTTGT +GGTTGC +GGTTGG +GGTCAA +GGTCAT +GGTCAC +GGTCAG +GGTCTA +GGTCTT +GGTCTC +GGTCTG +GGTCCA +GGTCCT +GGTCCC +GGTCCG +GGTCGA +GGTCGT +GGTCGC +GGTCGG +GGTGAA +GGTGAT +GGTGAC +GGTGAG +GGTGTA +GGTGTT +GGTGTC +GGTGTG +GGTGCA +GGTGCT +GGTGCC +GGTGCG +GGTGGA +GGTGGT +GGTGGC +GGTGGG +GGCAAA +GGCAAT +GGCAAC +GGCAAG +GGCATA +GGCATT +GGCATC +GGCATG +GGCACA +GGCACT +GGCACC +GGCACG +GGCAGA +GGCAGT +GGCAGC +GGCAGG +GGCTAA +GGCTAT +GGCTAC +GGCTAG +GGCTTA +GGCTTT +GGCTTC +GGCTTG +GGCTCA +GGCTCT +GGCTCC +GGCTCG +GGCTGA +GGCTGT +GGCTGC +GGCTGG +GGCCAA +GGCCAT +GGCCAC +GGCCAG +GGCCTA +GGCCTT +GGCCTC +GGCCTG +GGCCCA +GGCCCT +GGCCCC +GGCCCG +GGCCGA +GGCCGT +GGCCGC +GGCCGG +GGCGAA +GGCGAT +GGCGAC +GGCGAG +GGCGTA +GGCGTT +GGCGTC +GGCGTG +GGCGCA +GGCGCT +GGCGCC +GGCGCG +GGCGGA +GGCGGT +GGCGGC +GGCGGG +GGGAAA +GGGAAT +GGGAAC +GGGAAG +GGGATA +GGGATT +GGGATC +GGGATG +GGGACA +GGGACT +GGGACC +GGGACG +GGGAGA +GGGAGT +GGGAGC +GGGAGG +GGGTAA +GGGTAT +GGGTAC +GGGTAG +GGGTTA +GGGTTT +GGGTTC +GGGTTG +GGGTCA +GGGTCT +GGGTCC +GGGTCG +GGGTGA +GGGTGT +GGGTGC +GGGTGG +GGGCAA +GGGCAT +GGGCAC +GGGCAG +GGGCTA +GGGCTT +GGGCTC +GGGCTG +GGGCCA +GGGCCT +GGGCCC +GGGCCG +GGGCGA +GGGCGT +GGGCGC +GGGCGG +GGGGAA +GGGGAT +GGGGAC +GGGGAG +GGGGTA +GGGGTT +GGGGTC +GGGGTG +GGGGCA +GGGGCT +GGGGCC +GGGGCG +GGGGGA +GGGGGT +GGGGGC +GGGGGG diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6250/config.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6250/config.json new file mode 100644 index 0000000000000000000000000000000000000000..9b14e61d327bf586fcff6e817cf9d365bae18ad2 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6250/config.json @@ -0,0 +1,26 @@ +{ + "_name_or_path": "fabihamakhdoomi/TinyDNABERT", + "architectures": [ + "BertForSequenceClassification" + ], + "attention_probs_dropout_prob": 0.1, + "classifier_dropout": null, + "hidden_act": "gelu", + "hidden_dropout_prob": 0.1, + "hidden_size": 128, + "initializer_range": 0.02, + "intermediate_size": 512, + "layer_norm_eps": 1e-12, + "max_position_embeddings": 512, + "model_type": "bert", + "num_attention_heads": 2, + "num_hidden_layers": 2, + "pad_token_id": 0, + "position_embedding_type": "absolute", + "problem_type": "single_label_classification", + "torch_dtype": "float32", + "transformers_version": "4.44.0", + "type_vocab_size": 2, + "use_cache": true, + "vocab_size": 4101 +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6250/model.safetensors b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6250/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..b1e13176350ceb2b49c38ee09fbc13359bb1806d --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6250/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1a074667ef7b8b06a0df400b1e32f34d8b4a06b1fbd17ad186164b59092e619c +size 4021688 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6250/optimizer.pt b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6250/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..5bc38b1a6ceaba4b757b99f5b3b57c9f23d6cbcb --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6250/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cf6dfafcebcfbdb758901099a8b25a6fa72b2b780b518d639f6b20a802e87422 +size 8058700 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6250/rng_state.pth b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6250/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..e666d96a1e7d953ad715b4f92cb4f8713cf02642 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6250/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:539fe12b01f7e37cbd1c254718c50ce0226ac2f3489b5b349456040c507b4a44 +size 14244 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6250/scheduler.pt b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6250/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..4c34d9dcafc0eb1507a84da1af7c2c5228be082e --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6250/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:20611d193eaf6658e6192005c679e5aaa561ac33254f734047a2032edc3e8f9b +size 1064 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6250/special_tokens_map.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6250/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..a8b3208c2884c4efb86e49300fdd3dc877220cdf --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6250/special_tokens_map.json @@ -0,0 +1,7 @@ +{ + "cls_token": "[CLS]", + "mask_token": "[MASK]", + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6250/tokenizer.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6250/tokenizer.json new file mode 100644 index 0000000000000000000000000000000000000000..ebd782de6963b9032ce9374cfe30f5275be1a30b --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6250/tokenizer.json @@ -0,0 +1,4265 @@ +{ + "version": "1.0", + "truncation": { + "direction": "Right", + "max_length": 512, + "strategy": "LongestFirst", + "stride": 0 + }, + "padding": { + "strategy": { + "Fixed": 512 + }, + "direction": "Right", + "pad_to_multiple_of": null, + "pad_id": 0, + "pad_type_id": 0, + "pad_token": "[PAD]" + }, + "added_tokens": [ + { + "id": 0, + "content": "[PAD]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 1, + "content": "[UNK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 2, + "content": "[CLS]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 3, + "content": "[SEP]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + }, + { + "id": 4, + "content": "[MASK]", + "single_word": false, + "lstrip": false, + "rstrip": false, + "normalized": false, + "special": true + } + ], + "normalizer": { + "type": "BertNormalizer", + "clean_text": true, + "handle_chinese_chars": true, + "strip_accents": null, + "lowercase": false + }, + "pre_tokenizer": { + "type": "BertPreTokenizer" + }, + "post_processor": { + "type": "TemplateProcessing", + "single": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + } + ], + "pair": [ + { + "SpecialToken": { + "id": "[CLS]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "A", + "type_id": 0 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 0 + } + }, + { + "Sequence": { + "id": "B", + "type_id": 1 + } + }, + { + "SpecialToken": { + "id": "[SEP]", + "type_id": 1 + } + } + ], + "special_tokens": { + "[CLS]": { + "id": "[CLS]", + "ids": [ + 2 + ], + "tokens": [ + "[CLS]" + ] + }, + "[SEP]": { + "id": "[SEP]", + "ids": [ + 3 + ], + "tokens": [ + "[SEP]" + ] + } + } + }, + "decoder": { + "type": "WordPiece", + "prefix": "##", + "cleanup": true + }, + "model": { + "type": "WordPiece", + "unk_token": "[UNK]", + "continuing_subword_prefix": "##", + "max_input_chars_per_word": 100, + "vocab": { + "[PAD]": 0, + "[UNK]": 1, + "[CLS]": 2, + "[SEP]": 3, + "[MASK]": 4, + "AAAAAA": 5, + "AAAAAT": 6, + "AAAAAC": 7, + "AAAAAG": 8, + "AAAATA": 9, + "AAAATT": 10, + "AAAATC": 11, + "AAAATG": 12, + "AAAACA": 13, + "AAAACT": 14, + "AAAACC": 15, + "AAAACG": 16, + "AAAAGA": 17, + "AAAAGT": 18, + "AAAAGC": 19, + "AAAAGG": 20, + "AAATAA": 21, + "AAATAT": 22, + "AAATAC": 23, + "AAATAG": 24, + "AAATTA": 25, + "AAATTT": 26, + "AAATTC": 27, + "AAATTG": 28, + "AAATCA": 29, + "AAATCT": 30, + "AAATCC": 31, + "AAATCG": 32, + "AAATGA": 33, + "AAATGT": 34, + "AAATGC": 35, + "AAATGG": 36, + "AAACAA": 37, + "AAACAT": 38, + "AAACAC": 39, + "AAACAG": 40, + "AAACTA": 41, + "AAACTT": 42, + "AAACTC": 43, + "AAACTG": 44, + "AAACCA": 45, + "AAACCT": 46, + "AAACCC": 47, + "AAACCG": 48, + "AAACGA": 49, + "AAACGT": 50, + "AAACGC": 51, + "AAACGG": 52, + "AAAGAA": 53, + "AAAGAT": 54, + "AAAGAC": 55, + "AAAGAG": 56, + "AAAGTA": 57, + "AAAGTT": 58, + "AAAGTC": 59, + "AAAGTG": 60, + "AAAGCA": 61, + "AAAGCT": 62, + "AAAGCC": 63, + "AAAGCG": 64, + "AAAGGA": 65, + "AAAGGT": 66, + "AAAGGC": 67, + "AAAGGG": 68, + "AATAAA": 69, + "AATAAT": 70, + "AATAAC": 71, + "AATAAG": 72, + "AATATA": 73, + "AATATT": 74, + "AATATC": 75, + "AATATG": 76, + "AATACA": 77, + "AATACT": 78, + "AATACC": 79, + "AATACG": 80, + "AATAGA": 81, + "AATAGT": 82, + "AATAGC": 83, + "AATAGG": 84, + "AATTAA": 85, + "AATTAT": 86, + "AATTAC": 87, + "AATTAG": 88, + "AATTTA": 89, + "AATTTT": 90, + "AATTTC": 91, + "AATTTG": 92, + "AATTCA": 93, + "AATTCT": 94, + "AATTCC": 95, + "AATTCG": 96, + "AATTGA": 97, + "AATTGT": 98, + "AATTGC": 99, + "AATTGG": 100, + "AATCAA": 101, + "AATCAT": 102, + "AATCAC": 103, + "AATCAG": 104, + "AATCTA": 105, + "AATCTT": 106, + "AATCTC": 107, + "AATCTG": 108, + "AATCCA": 109, + "AATCCT": 110, + "AATCCC": 111, + "AATCCG": 112, + "AATCGA": 113, + "AATCGT": 114, + "AATCGC": 115, + "AATCGG": 116, + "AATGAA": 117, + "AATGAT": 118, + "AATGAC": 119, + "AATGAG": 120, + "AATGTA": 121, + "AATGTT": 122, + "AATGTC": 123, + "AATGTG": 124, + "AATGCA": 125, + "AATGCT": 126, + "AATGCC": 127, + "AATGCG": 128, + "AATGGA": 129, + "AATGGT": 130, + "AATGGC": 131, + "AATGGG": 132, + "AACAAA": 133, + "AACAAT": 134, + "AACAAC": 135, + "AACAAG": 136, + "AACATA": 137, + "AACATT": 138, + "AACATC": 139, + "AACATG": 140, + "AACACA": 141, + "AACACT": 142, + "AACACC": 143, + "AACACG": 144, + "AACAGA": 145, + "AACAGT": 146, + "AACAGC": 147, + "AACAGG": 148, + "AACTAA": 149, + "AACTAT": 150, + "AACTAC": 151, + "AACTAG": 152, + "AACTTA": 153, + "AACTTT": 154, + "AACTTC": 155, + "AACTTG": 156, + "AACTCA": 157, + "AACTCT": 158, + "AACTCC": 159, + "AACTCG": 160, + "AACTGA": 161, + "AACTGT": 162, + "AACTGC": 163, + "AACTGG": 164, + "AACCAA": 165, + "AACCAT": 166, + "AACCAC": 167, + "AACCAG": 168, + "AACCTA": 169, + "AACCTT": 170, + "AACCTC": 171, + "AACCTG": 172, + "AACCCA": 173, + "AACCCT": 174, + "AACCCC": 175, + "AACCCG": 176, + "AACCGA": 177, + "AACCGT": 178, + "AACCGC": 179, + "AACCGG": 180, + "AACGAA": 181, + "AACGAT": 182, + "AACGAC": 183, + "AACGAG": 184, + "AACGTA": 185, + "AACGTT": 186, + "AACGTC": 187, + "AACGTG": 188, + "AACGCA": 189, + "AACGCT": 190, + "AACGCC": 191, + "AACGCG": 192, + "AACGGA": 193, + "AACGGT": 194, + "AACGGC": 195, + "AACGGG": 196, + "AAGAAA": 197, + "AAGAAT": 198, + "AAGAAC": 199, + "AAGAAG": 200, + "AAGATA": 201, + "AAGATT": 202, + "AAGATC": 203, + "AAGATG": 204, + "AAGACA": 205, + "AAGACT": 206, + "AAGACC": 207, + "AAGACG": 208, + "AAGAGA": 209, + "AAGAGT": 210, + "AAGAGC": 211, + "AAGAGG": 212, + "AAGTAA": 213, + "AAGTAT": 214, + "AAGTAC": 215, + "AAGTAG": 216, + "AAGTTA": 217, + "AAGTTT": 218, + "AAGTTC": 219, + "AAGTTG": 220, + "AAGTCA": 221, + "AAGTCT": 222, + "AAGTCC": 223, + "AAGTCG": 224, + "AAGTGA": 225, + "AAGTGT": 226, + "AAGTGC": 227, + "AAGTGG": 228, + "AAGCAA": 229, + "AAGCAT": 230, + "AAGCAC": 231, + "AAGCAG": 232, + "AAGCTA": 233, + "AAGCTT": 234, + "AAGCTC": 235, + "AAGCTG": 236, + "AAGCCA": 237, + "AAGCCT": 238, + "AAGCCC": 239, + "AAGCCG": 240, + "AAGCGA": 241, + "AAGCGT": 242, + "AAGCGC": 243, + "AAGCGG": 244, + "AAGGAA": 245, + "AAGGAT": 246, + "AAGGAC": 247, + "AAGGAG": 248, + "AAGGTA": 249, + "AAGGTT": 250, + "AAGGTC": 251, + "AAGGTG": 252, + "AAGGCA": 253, + "AAGGCT": 254, + "AAGGCC": 255, + "AAGGCG": 256, + "AAGGGA": 257, + "AAGGGT": 258, + "AAGGGC": 259, + "AAGGGG": 260, + "ATAAAA": 261, + "ATAAAT": 262, + "ATAAAC": 263, + "ATAAAG": 264, + "ATAATA": 265, + "ATAATT": 266, + "ATAATC": 267, + "ATAATG": 268, + "ATAACA": 269, + "ATAACT": 270, + "ATAACC": 271, + "ATAACG": 272, + "ATAAGA": 273, + "ATAAGT": 274, + "ATAAGC": 275, + "ATAAGG": 276, + "ATATAA": 277, + "ATATAT": 278, + "ATATAC": 279, + "ATATAG": 280, + "ATATTA": 281, + "ATATTT": 282, + "ATATTC": 283, + "ATATTG": 284, + "ATATCA": 285, + "ATATCT": 286, + "ATATCC": 287, + "ATATCG": 288, + "ATATGA": 289, + "ATATGT": 290, + "ATATGC": 291, + "ATATGG": 292, + "ATACAA": 293, + "ATACAT": 294, + "ATACAC": 295, + "ATACAG": 296, + "ATACTA": 297, + "ATACTT": 298, + "ATACTC": 299, + "ATACTG": 300, + "ATACCA": 301, + "ATACCT": 302, + "ATACCC": 303, + "ATACCG": 304, + "ATACGA": 305, + "ATACGT": 306, + "ATACGC": 307, + "ATACGG": 308, + "ATAGAA": 309, + "ATAGAT": 310, + "ATAGAC": 311, + "ATAGAG": 312, + "ATAGTA": 313, + "ATAGTT": 314, + "ATAGTC": 315, + "ATAGTG": 316, + "ATAGCA": 317, + "ATAGCT": 318, + "ATAGCC": 319, + "ATAGCG": 320, + "ATAGGA": 321, + "ATAGGT": 322, + "ATAGGC": 323, + "ATAGGG": 324, + "ATTAAA": 325, + "ATTAAT": 326, + "ATTAAC": 327, + "ATTAAG": 328, + "ATTATA": 329, + "ATTATT": 330, + "ATTATC": 331, + "ATTATG": 332, + "ATTACA": 333, + "ATTACT": 334, + "ATTACC": 335, + "ATTACG": 336, + "ATTAGA": 337, + "ATTAGT": 338, + "ATTAGC": 339, + "ATTAGG": 340, + "ATTTAA": 341, + "ATTTAT": 342, + "ATTTAC": 343, + "ATTTAG": 344, + "ATTTTA": 345, + "ATTTTT": 346, + "ATTTTC": 347, + "ATTTTG": 348, + "ATTTCA": 349, + "ATTTCT": 350, + "ATTTCC": 351, + "ATTTCG": 352, + "ATTTGA": 353, + "ATTTGT": 354, + "ATTTGC": 355, + "ATTTGG": 356, + "ATTCAA": 357, + "ATTCAT": 358, + "ATTCAC": 359, + "ATTCAG": 360, + "ATTCTA": 361, + "ATTCTT": 362, + "ATTCTC": 363, + "ATTCTG": 364, + "ATTCCA": 365, + "ATTCCT": 366, + "ATTCCC": 367, + "ATTCCG": 368, + "ATTCGA": 369, + "ATTCGT": 370, + "ATTCGC": 371, + "ATTCGG": 372, + "ATTGAA": 373, + "ATTGAT": 374, + "ATTGAC": 375, + "ATTGAG": 376, + "ATTGTA": 377, + "ATTGTT": 378, + "ATTGTC": 379, + "ATTGTG": 380, + "ATTGCA": 381, + "ATTGCT": 382, + "ATTGCC": 383, + "ATTGCG": 384, + "ATTGGA": 385, + "ATTGGT": 386, + "ATTGGC": 387, + "ATTGGG": 388, + "ATCAAA": 389, + "ATCAAT": 390, + "ATCAAC": 391, + "ATCAAG": 392, + "ATCATA": 393, + "ATCATT": 394, + "ATCATC": 395, + "ATCATG": 396, + "ATCACA": 397, + "ATCACT": 398, + "ATCACC": 399, + "ATCACG": 400, + "ATCAGA": 401, + "ATCAGT": 402, + "ATCAGC": 403, + "ATCAGG": 404, + "ATCTAA": 405, + "ATCTAT": 406, + "ATCTAC": 407, + "ATCTAG": 408, + "ATCTTA": 409, + "ATCTTT": 410, + "ATCTTC": 411, + "ATCTTG": 412, + "ATCTCA": 413, + "ATCTCT": 414, + "ATCTCC": 415, + "ATCTCG": 416, + "ATCTGA": 417, + "ATCTGT": 418, + "ATCTGC": 419, + "ATCTGG": 420, + "ATCCAA": 421, + "ATCCAT": 422, + "ATCCAC": 423, + "ATCCAG": 424, + "ATCCTA": 425, + "ATCCTT": 426, + "ATCCTC": 427, + "ATCCTG": 428, + "ATCCCA": 429, + "ATCCCT": 430, + "ATCCCC": 431, + "ATCCCG": 432, + "ATCCGA": 433, + "ATCCGT": 434, + "ATCCGC": 435, + "ATCCGG": 436, + "ATCGAA": 437, + "ATCGAT": 438, + "ATCGAC": 439, + "ATCGAG": 440, + "ATCGTA": 441, + "ATCGTT": 442, + "ATCGTC": 443, + "ATCGTG": 444, + "ATCGCA": 445, + "ATCGCT": 446, + "ATCGCC": 447, + "ATCGCG": 448, + "ATCGGA": 449, + "ATCGGT": 450, + "ATCGGC": 451, + "ATCGGG": 452, + "ATGAAA": 453, + "ATGAAT": 454, + "ATGAAC": 455, + "ATGAAG": 456, + "ATGATA": 457, + "ATGATT": 458, + "ATGATC": 459, + "ATGATG": 460, + "ATGACA": 461, + "ATGACT": 462, + "ATGACC": 463, + "ATGACG": 464, + "ATGAGA": 465, + "ATGAGT": 466, + "ATGAGC": 467, + "ATGAGG": 468, + "ATGTAA": 469, + "ATGTAT": 470, + "ATGTAC": 471, + "ATGTAG": 472, + "ATGTTA": 473, + "ATGTTT": 474, + "ATGTTC": 475, + "ATGTTG": 476, + "ATGTCA": 477, + "ATGTCT": 478, + "ATGTCC": 479, + "ATGTCG": 480, + "ATGTGA": 481, + "ATGTGT": 482, + "ATGTGC": 483, + "ATGTGG": 484, + "ATGCAA": 485, + "ATGCAT": 486, + "ATGCAC": 487, + "ATGCAG": 488, + "ATGCTA": 489, + "ATGCTT": 490, + "ATGCTC": 491, + "ATGCTG": 492, + "ATGCCA": 493, + "ATGCCT": 494, + "ATGCCC": 495, + "ATGCCG": 496, + "ATGCGA": 497, + "ATGCGT": 498, + "ATGCGC": 499, + "ATGCGG": 500, + "ATGGAA": 501, + "ATGGAT": 502, + "ATGGAC": 503, + "ATGGAG": 504, + "ATGGTA": 505, + "ATGGTT": 506, + "ATGGTC": 507, + "ATGGTG": 508, + "ATGGCA": 509, + "ATGGCT": 510, + "ATGGCC": 511, + "ATGGCG": 512, + "ATGGGA": 513, + "ATGGGT": 514, + "ATGGGC": 515, + "ATGGGG": 516, + "ACAAAA": 517, + "ACAAAT": 518, + "ACAAAC": 519, + "ACAAAG": 520, + "ACAATA": 521, + "ACAATT": 522, + "ACAATC": 523, + "ACAATG": 524, + "ACAACA": 525, + "ACAACT": 526, + "ACAACC": 527, + "ACAACG": 528, + "ACAAGA": 529, + "ACAAGT": 530, + "ACAAGC": 531, + "ACAAGG": 532, + "ACATAA": 533, + "ACATAT": 534, + "ACATAC": 535, + "ACATAG": 536, + "ACATTA": 537, + "ACATTT": 538, + "ACATTC": 539, + "ACATTG": 540, + "ACATCA": 541, + "ACATCT": 542, + "ACATCC": 543, + "ACATCG": 544, + "ACATGA": 545, + "ACATGT": 546, + "ACATGC": 547, + "ACATGG": 548, + "ACACAA": 549, + "ACACAT": 550, + "ACACAC": 551, + "ACACAG": 552, + "ACACTA": 553, + "ACACTT": 554, + "ACACTC": 555, + "ACACTG": 556, + "ACACCA": 557, + "ACACCT": 558, + "ACACCC": 559, + "ACACCG": 560, + "ACACGA": 561, + "ACACGT": 562, + "ACACGC": 563, + "ACACGG": 564, + "ACAGAA": 565, + "ACAGAT": 566, + "ACAGAC": 567, + "ACAGAG": 568, + "ACAGTA": 569, + "ACAGTT": 570, + "ACAGTC": 571, + "ACAGTG": 572, + "ACAGCA": 573, + "ACAGCT": 574, + "ACAGCC": 575, + "ACAGCG": 576, + "ACAGGA": 577, + "ACAGGT": 578, + "ACAGGC": 579, + "ACAGGG": 580, + "ACTAAA": 581, + "ACTAAT": 582, + "ACTAAC": 583, + "ACTAAG": 584, + "ACTATA": 585, + "ACTATT": 586, + "ACTATC": 587, + "ACTATG": 588, + "ACTACA": 589, + "ACTACT": 590, + "ACTACC": 591, + "ACTACG": 592, + "ACTAGA": 593, + "ACTAGT": 594, + "ACTAGC": 595, + "ACTAGG": 596, + "ACTTAA": 597, + "ACTTAT": 598, + "ACTTAC": 599, + "ACTTAG": 600, + "ACTTTA": 601, + "ACTTTT": 602, + "ACTTTC": 603, + "ACTTTG": 604, + "ACTTCA": 605, + "ACTTCT": 606, + "ACTTCC": 607, + "ACTTCG": 608, + "ACTTGA": 609, + "ACTTGT": 610, + "ACTTGC": 611, + "ACTTGG": 612, + "ACTCAA": 613, + "ACTCAT": 614, + "ACTCAC": 615, + "ACTCAG": 616, + "ACTCTA": 617, + "ACTCTT": 618, + "ACTCTC": 619, + "ACTCTG": 620, + "ACTCCA": 621, + "ACTCCT": 622, + "ACTCCC": 623, + "ACTCCG": 624, + "ACTCGA": 625, + "ACTCGT": 626, + "ACTCGC": 627, + "ACTCGG": 628, + "ACTGAA": 629, + "ACTGAT": 630, + "ACTGAC": 631, + "ACTGAG": 632, + "ACTGTA": 633, + "ACTGTT": 634, + "ACTGTC": 635, + "ACTGTG": 636, + "ACTGCA": 637, + "ACTGCT": 638, + "ACTGCC": 639, + "ACTGCG": 640, + "ACTGGA": 641, + "ACTGGT": 642, + "ACTGGC": 643, + "ACTGGG": 644, + "ACCAAA": 645, + "ACCAAT": 646, + "ACCAAC": 647, + "ACCAAG": 648, + "ACCATA": 649, + "ACCATT": 650, + "ACCATC": 651, + "ACCATG": 652, + "ACCACA": 653, + "ACCACT": 654, + "ACCACC": 655, + "ACCACG": 656, + "ACCAGA": 657, + "ACCAGT": 658, + "ACCAGC": 659, + "ACCAGG": 660, + "ACCTAA": 661, + "ACCTAT": 662, + "ACCTAC": 663, + "ACCTAG": 664, + "ACCTTA": 665, + "ACCTTT": 666, + "ACCTTC": 667, + "ACCTTG": 668, + "ACCTCA": 669, + "ACCTCT": 670, + "ACCTCC": 671, + "ACCTCG": 672, + "ACCTGA": 673, + "ACCTGT": 674, + "ACCTGC": 675, + "ACCTGG": 676, + "ACCCAA": 677, + "ACCCAT": 678, + "ACCCAC": 679, + "ACCCAG": 680, + "ACCCTA": 681, + "ACCCTT": 682, + "ACCCTC": 683, + "ACCCTG": 684, + "ACCCCA": 685, + "ACCCCT": 686, + "ACCCCC": 687, + "ACCCCG": 688, + "ACCCGA": 689, + "ACCCGT": 690, + "ACCCGC": 691, + "ACCCGG": 692, + "ACCGAA": 693, + "ACCGAT": 694, + "ACCGAC": 695, + "ACCGAG": 696, + "ACCGTA": 697, + "ACCGTT": 698, + "ACCGTC": 699, + "ACCGTG": 700, + "ACCGCA": 701, + "ACCGCT": 702, + "ACCGCC": 703, + "ACCGCG": 704, + "ACCGGA": 705, + "ACCGGT": 706, + "ACCGGC": 707, + "ACCGGG": 708, + "ACGAAA": 709, + "ACGAAT": 710, + "ACGAAC": 711, + "ACGAAG": 712, + "ACGATA": 713, + "ACGATT": 714, + "ACGATC": 715, + "ACGATG": 716, + "ACGACA": 717, + "ACGACT": 718, + "ACGACC": 719, + "ACGACG": 720, + "ACGAGA": 721, + "ACGAGT": 722, + "ACGAGC": 723, + "ACGAGG": 724, + "ACGTAA": 725, + "ACGTAT": 726, + "ACGTAC": 727, + "ACGTAG": 728, + "ACGTTA": 729, + "ACGTTT": 730, + "ACGTTC": 731, + "ACGTTG": 732, + "ACGTCA": 733, + "ACGTCT": 734, + "ACGTCC": 735, + "ACGTCG": 736, + "ACGTGA": 737, + "ACGTGT": 738, + "ACGTGC": 739, + "ACGTGG": 740, + "ACGCAA": 741, + "ACGCAT": 742, + "ACGCAC": 743, + "ACGCAG": 744, + "ACGCTA": 745, + "ACGCTT": 746, + "ACGCTC": 747, + "ACGCTG": 748, + "ACGCCA": 749, + "ACGCCT": 750, + "ACGCCC": 751, + "ACGCCG": 752, + "ACGCGA": 753, + "ACGCGT": 754, + "ACGCGC": 755, + "ACGCGG": 756, + "ACGGAA": 757, + "ACGGAT": 758, + "ACGGAC": 759, + "ACGGAG": 760, + "ACGGTA": 761, + "ACGGTT": 762, + "ACGGTC": 763, + "ACGGTG": 764, + "ACGGCA": 765, + "ACGGCT": 766, + "ACGGCC": 767, + "ACGGCG": 768, + "ACGGGA": 769, + "ACGGGT": 770, + "ACGGGC": 771, + "ACGGGG": 772, + "AGAAAA": 773, + "AGAAAT": 774, + "AGAAAC": 775, + "AGAAAG": 776, + "AGAATA": 777, + "AGAATT": 778, + "AGAATC": 779, + "AGAATG": 780, + "AGAACA": 781, + "AGAACT": 782, + "AGAACC": 783, + "AGAACG": 784, + "AGAAGA": 785, + "AGAAGT": 786, + "AGAAGC": 787, + "AGAAGG": 788, + "AGATAA": 789, + "AGATAT": 790, + "AGATAC": 791, + "AGATAG": 792, + "AGATTA": 793, + "AGATTT": 794, + "AGATTC": 795, + "AGATTG": 796, + "AGATCA": 797, + "AGATCT": 798, + "AGATCC": 799, + "AGATCG": 800, + "AGATGA": 801, + "AGATGT": 802, + "AGATGC": 803, + "AGATGG": 804, + "AGACAA": 805, + "AGACAT": 806, + "AGACAC": 807, + "AGACAG": 808, + "AGACTA": 809, + "AGACTT": 810, + "AGACTC": 811, + "AGACTG": 812, + "AGACCA": 813, + "AGACCT": 814, + "AGACCC": 815, + "AGACCG": 816, + "AGACGA": 817, + "AGACGT": 818, + "AGACGC": 819, + "AGACGG": 820, + "AGAGAA": 821, + "AGAGAT": 822, + "AGAGAC": 823, + "AGAGAG": 824, + "AGAGTA": 825, + "AGAGTT": 826, + "AGAGTC": 827, + "AGAGTG": 828, + "AGAGCA": 829, + "AGAGCT": 830, + "AGAGCC": 831, + "AGAGCG": 832, + "AGAGGA": 833, + "AGAGGT": 834, + "AGAGGC": 835, + "AGAGGG": 836, + "AGTAAA": 837, + "AGTAAT": 838, + "AGTAAC": 839, + "AGTAAG": 840, + "AGTATA": 841, + "AGTATT": 842, + "AGTATC": 843, + "AGTATG": 844, + "AGTACA": 845, + "AGTACT": 846, + "AGTACC": 847, + "AGTACG": 848, + "AGTAGA": 849, + "AGTAGT": 850, + "AGTAGC": 851, + "AGTAGG": 852, + "AGTTAA": 853, + "AGTTAT": 854, + "AGTTAC": 855, + "AGTTAG": 856, + "AGTTTA": 857, + "AGTTTT": 858, + "AGTTTC": 859, + "AGTTTG": 860, + "AGTTCA": 861, + "AGTTCT": 862, + "AGTTCC": 863, + "AGTTCG": 864, + "AGTTGA": 865, + "AGTTGT": 866, + "AGTTGC": 867, + "AGTTGG": 868, + "AGTCAA": 869, + "AGTCAT": 870, + "AGTCAC": 871, + "AGTCAG": 872, + "AGTCTA": 873, + "AGTCTT": 874, + "AGTCTC": 875, + "AGTCTG": 876, + "AGTCCA": 877, + "AGTCCT": 878, + "AGTCCC": 879, + "AGTCCG": 880, + "AGTCGA": 881, + "AGTCGT": 882, + "AGTCGC": 883, + "AGTCGG": 884, + "AGTGAA": 885, + "AGTGAT": 886, + "AGTGAC": 887, + "AGTGAG": 888, + "AGTGTA": 889, + "AGTGTT": 890, + "AGTGTC": 891, + "AGTGTG": 892, + "AGTGCA": 893, + "AGTGCT": 894, + "AGTGCC": 895, + "AGTGCG": 896, + "AGTGGA": 897, + "AGTGGT": 898, + "AGTGGC": 899, + "AGTGGG": 900, + "AGCAAA": 901, + "AGCAAT": 902, + "AGCAAC": 903, + "AGCAAG": 904, + "AGCATA": 905, + "AGCATT": 906, + "AGCATC": 907, + "AGCATG": 908, + "AGCACA": 909, + "AGCACT": 910, + "AGCACC": 911, + "AGCACG": 912, + "AGCAGA": 913, + "AGCAGT": 914, + "AGCAGC": 915, + "AGCAGG": 916, + "AGCTAA": 917, + "AGCTAT": 918, + "AGCTAC": 919, + "AGCTAG": 920, + "AGCTTA": 921, + "AGCTTT": 922, + "AGCTTC": 923, + "AGCTTG": 924, + "AGCTCA": 925, + "AGCTCT": 926, + "AGCTCC": 927, + "AGCTCG": 928, + "AGCTGA": 929, + "AGCTGT": 930, + "AGCTGC": 931, + "AGCTGG": 932, + "AGCCAA": 933, + "AGCCAT": 934, + "AGCCAC": 935, + "AGCCAG": 936, + "AGCCTA": 937, + "AGCCTT": 938, + "AGCCTC": 939, + "AGCCTG": 940, + "AGCCCA": 941, + "AGCCCT": 942, + "AGCCCC": 943, + "AGCCCG": 944, + "AGCCGA": 945, + "AGCCGT": 946, + "AGCCGC": 947, + "AGCCGG": 948, + "AGCGAA": 949, + "AGCGAT": 950, + "AGCGAC": 951, + "AGCGAG": 952, + "AGCGTA": 953, + "AGCGTT": 954, + "AGCGTC": 955, + "AGCGTG": 956, + "AGCGCA": 957, + "AGCGCT": 958, + "AGCGCC": 959, + "AGCGCG": 960, + "AGCGGA": 961, + "AGCGGT": 962, + "AGCGGC": 963, + "AGCGGG": 964, + "AGGAAA": 965, + "AGGAAT": 966, + "AGGAAC": 967, + "AGGAAG": 968, + "AGGATA": 969, + "AGGATT": 970, + "AGGATC": 971, + "AGGATG": 972, + "AGGACA": 973, + "AGGACT": 974, + "AGGACC": 975, + "AGGACG": 976, + "AGGAGA": 977, + "AGGAGT": 978, + "AGGAGC": 979, + "AGGAGG": 980, + "AGGTAA": 981, + "AGGTAT": 982, + "AGGTAC": 983, + "AGGTAG": 984, + "AGGTTA": 985, + "AGGTTT": 986, + "AGGTTC": 987, + "AGGTTG": 988, + "AGGTCA": 989, + "AGGTCT": 990, + "AGGTCC": 991, + "AGGTCG": 992, + "AGGTGA": 993, + "AGGTGT": 994, + "AGGTGC": 995, + "AGGTGG": 996, + "AGGCAA": 997, + "AGGCAT": 998, + "AGGCAC": 999, + "AGGCAG": 1000, + "AGGCTA": 1001, + "AGGCTT": 1002, + "AGGCTC": 1003, + "AGGCTG": 1004, + "AGGCCA": 1005, + "AGGCCT": 1006, + "AGGCCC": 1007, + "AGGCCG": 1008, + "AGGCGA": 1009, + "AGGCGT": 1010, + "AGGCGC": 1011, + "AGGCGG": 1012, + "AGGGAA": 1013, + "AGGGAT": 1014, + "AGGGAC": 1015, + "AGGGAG": 1016, + "AGGGTA": 1017, + "AGGGTT": 1018, + "AGGGTC": 1019, + "AGGGTG": 1020, + "AGGGCA": 1021, + "AGGGCT": 1022, + "AGGGCC": 1023, + "AGGGCG": 1024, + "AGGGGA": 1025, + "AGGGGT": 1026, + "AGGGGC": 1027, + "AGGGGG": 1028, + "TAAAAA": 1029, + "TAAAAT": 1030, + "TAAAAC": 1031, + "TAAAAG": 1032, + "TAAATA": 1033, + "TAAATT": 1034, + "TAAATC": 1035, + "TAAATG": 1036, + "TAAACA": 1037, + "TAAACT": 1038, + "TAAACC": 1039, + "TAAACG": 1040, + "TAAAGA": 1041, + "TAAAGT": 1042, + "TAAAGC": 1043, + "TAAAGG": 1044, + "TAATAA": 1045, + "TAATAT": 1046, + "TAATAC": 1047, + "TAATAG": 1048, + "TAATTA": 1049, + "TAATTT": 1050, + "TAATTC": 1051, + "TAATTG": 1052, + "TAATCA": 1053, + "TAATCT": 1054, + "TAATCC": 1055, + "TAATCG": 1056, + "TAATGA": 1057, + "TAATGT": 1058, + "TAATGC": 1059, + "TAATGG": 1060, + "TAACAA": 1061, + "TAACAT": 1062, + "TAACAC": 1063, + "TAACAG": 1064, + "TAACTA": 1065, + "TAACTT": 1066, + "TAACTC": 1067, + "TAACTG": 1068, + "TAACCA": 1069, + "TAACCT": 1070, + "TAACCC": 1071, + "TAACCG": 1072, + "TAACGA": 1073, + "TAACGT": 1074, + "TAACGC": 1075, + "TAACGG": 1076, + "TAAGAA": 1077, + "TAAGAT": 1078, + "TAAGAC": 1079, + "TAAGAG": 1080, + "TAAGTA": 1081, + "TAAGTT": 1082, + "TAAGTC": 1083, + "TAAGTG": 1084, + "TAAGCA": 1085, + "TAAGCT": 1086, + "TAAGCC": 1087, + "TAAGCG": 1088, + "TAAGGA": 1089, + "TAAGGT": 1090, + "TAAGGC": 1091, + "TAAGGG": 1092, + "TATAAA": 1093, + "TATAAT": 1094, + "TATAAC": 1095, + "TATAAG": 1096, + "TATATA": 1097, + "TATATT": 1098, + "TATATC": 1099, + "TATATG": 1100, + "TATACA": 1101, + "TATACT": 1102, + "TATACC": 1103, + "TATACG": 1104, + "TATAGA": 1105, + "TATAGT": 1106, + "TATAGC": 1107, + "TATAGG": 1108, + "TATTAA": 1109, + "TATTAT": 1110, + "TATTAC": 1111, + "TATTAG": 1112, + "TATTTA": 1113, + "TATTTT": 1114, + "TATTTC": 1115, + "TATTTG": 1116, + "TATTCA": 1117, + "TATTCT": 1118, + "TATTCC": 1119, + "TATTCG": 1120, + "TATTGA": 1121, + "TATTGT": 1122, + "TATTGC": 1123, + "TATTGG": 1124, + "TATCAA": 1125, + "TATCAT": 1126, + "TATCAC": 1127, + "TATCAG": 1128, + "TATCTA": 1129, + "TATCTT": 1130, + "TATCTC": 1131, + "TATCTG": 1132, + "TATCCA": 1133, + "TATCCT": 1134, + "TATCCC": 1135, + "TATCCG": 1136, + "TATCGA": 1137, + "TATCGT": 1138, + "TATCGC": 1139, + "TATCGG": 1140, + "TATGAA": 1141, + "TATGAT": 1142, + "TATGAC": 1143, + "TATGAG": 1144, + "TATGTA": 1145, + "TATGTT": 1146, + "TATGTC": 1147, + "TATGTG": 1148, + "TATGCA": 1149, + "TATGCT": 1150, + "TATGCC": 1151, + "TATGCG": 1152, + "TATGGA": 1153, + "TATGGT": 1154, + "TATGGC": 1155, + "TATGGG": 1156, + "TACAAA": 1157, + "TACAAT": 1158, + "TACAAC": 1159, + "TACAAG": 1160, + "TACATA": 1161, + "TACATT": 1162, + "TACATC": 1163, + "TACATG": 1164, + "TACACA": 1165, + "TACACT": 1166, + "TACACC": 1167, + "TACACG": 1168, + "TACAGA": 1169, + "TACAGT": 1170, + "TACAGC": 1171, + "TACAGG": 1172, + "TACTAA": 1173, + "TACTAT": 1174, + "TACTAC": 1175, + "TACTAG": 1176, + "TACTTA": 1177, + "TACTTT": 1178, + "TACTTC": 1179, + "TACTTG": 1180, + "TACTCA": 1181, + "TACTCT": 1182, + "TACTCC": 1183, + "TACTCG": 1184, + "TACTGA": 1185, + "TACTGT": 1186, + "TACTGC": 1187, + "TACTGG": 1188, + "TACCAA": 1189, + "TACCAT": 1190, + "TACCAC": 1191, + "TACCAG": 1192, + "TACCTA": 1193, + "TACCTT": 1194, + "TACCTC": 1195, + "TACCTG": 1196, + "TACCCA": 1197, + "TACCCT": 1198, + "TACCCC": 1199, + "TACCCG": 1200, + "TACCGA": 1201, + "TACCGT": 1202, + "TACCGC": 1203, + "TACCGG": 1204, + "TACGAA": 1205, + "TACGAT": 1206, + "TACGAC": 1207, + "TACGAG": 1208, + "TACGTA": 1209, + "TACGTT": 1210, + "TACGTC": 1211, + "TACGTG": 1212, + "TACGCA": 1213, + "TACGCT": 1214, + "TACGCC": 1215, + "TACGCG": 1216, + "TACGGA": 1217, + "TACGGT": 1218, + "TACGGC": 1219, + "TACGGG": 1220, + "TAGAAA": 1221, + "TAGAAT": 1222, + "TAGAAC": 1223, + "TAGAAG": 1224, + "TAGATA": 1225, + "TAGATT": 1226, + "TAGATC": 1227, + "TAGATG": 1228, + "TAGACA": 1229, + "TAGACT": 1230, + "TAGACC": 1231, + "TAGACG": 1232, + "TAGAGA": 1233, + "TAGAGT": 1234, + "TAGAGC": 1235, + "TAGAGG": 1236, + "TAGTAA": 1237, + "TAGTAT": 1238, + "TAGTAC": 1239, + "TAGTAG": 1240, + "TAGTTA": 1241, + "TAGTTT": 1242, + "TAGTTC": 1243, + "TAGTTG": 1244, + "TAGTCA": 1245, + "TAGTCT": 1246, + "TAGTCC": 1247, + "TAGTCG": 1248, + "TAGTGA": 1249, + "TAGTGT": 1250, + "TAGTGC": 1251, + "TAGTGG": 1252, + "TAGCAA": 1253, + "TAGCAT": 1254, + "TAGCAC": 1255, + "TAGCAG": 1256, + "TAGCTA": 1257, + "TAGCTT": 1258, + "TAGCTC": 1259, + "TAGCTG": 1260, + "TAGCCA": 1261, + "TAGCCT": 1262, + "TAGCCC": 1263, + "TAGCCG": 1264, + "TAGCGA": 1265, + "TAGCGT": 1266, + "TAGCGC": 1267, + "TAGCGG": 1268, + "TAGGAA": 1269, + "TAGGAT": 1270, + "TAGGAC": 1271, + "TAGGAG": 1272, + "TAGGTA": 1273, + "TAGGTT": 1274, + "TAGGTC": 1275, + "TAGGTG": 1276, + "TAGGCA": 1277, + "TAGGCT": 1278, + "TAGGCC": 1279, + "TAGGCG": 1280, + "TAGGGA": 1281, + "TAGGGT": 1282, + "TAGGGC": 1283, + "TAGGGG": 1284, + "TTAAAA": 1285, + "TTAAAT": 1286, + "TTAAAC": 1287, + "TTAAAG": 1288, + "TTAATA": 1289, + "TTAATT": 1290, + "TTAATC": 1291, + "TTAATG": 1292, + "TTAACA": 1293, + "TTAACT": 1294, + "TTAACC": 1295, + "TTAACG": 1296, + "TTAAGA": 1297, + "TTAAGT": 1298, + "TTAAGC": 1299, + "TTAAGG": 1300, + "TTATAA": 1301, + "TTATAT": 1302, + "TTATAC": 1303, + "TTATAG": 1304, + "TTATTA": 1305, + "TTATTT": 1306, + "TTATTC": 1307, + "TTATTG": 1308, + "TTATCA": 1309, + "TTATCT": 1310, + "TTATCC": 1311, + "TTATCG": 1312, + "TTATGA": 1313, + "TTATGT": 1314, + "TTATGC": 1315, + "TTATGG": 1316, + "TTACAA": 1317, + "TTACAT": 1318, + "TTACAC": 1319, + "TTACAG": 1320, + "TTACTA": 1321, + "TTACTT": 1322, + "TTACTC": 1323, + "TTACTG": 1324, + "TTACCA": 1325, + "TTACCT": 1326, + "TTACCC": 1327, + "TTACCG": 1328, + "TTACGA": 1329, + "TTACGT": 1330, + "TTACGC": 1331, + "TTACGG": 1332, + "TTAGAA": 1333, + "TTAGAT": 1334, + "TTAGAC": 1335, + "TTAGAG": 1336, + "TTAGTA": 1337, + "TTAGTT": 1338, + "TTAGTC": 1339, + "TTAGTG": 1340, + "TTAGCA": 1341, + "TTAGCT": 1342, + "TTAGCC": 1343, + "TTAGCG": 1344, + "TTAGGA": 1345, + "TTAGGT": 1346, + "TTAGGC": 1347, + "TTAGGG": 1348, + "TTTAAA": 1349, + "TTTAAT": 1350, + "TTTAAC": 1351, + "TTTAAG": 1352, + "TTTATA": 1353, + "TTTATT": 1354, + "TTTATC": 1355, + "TTTATG": 1356, + "TTTACA": 1357, + "TTTACT": 1358, + "TTTACC": 1359, + "TTTACG": 1360, + "TTTAGA": 1361, + "TTTAGT": 1362, + "TTTAGC": 1363, + "TTTAGG": 1364, + "TTTTAA": 1365, + "TTTTAT": 1366, + "TTTTAC": 1367, + "TTTTAG": 1368, + "TTTTTA": 1369, + "TTTTTT": 1370, + "TTTTTC": 1371, + "TTTTTG": 1372, + "TTTTCA": 1373, + "TTTTCT": 1374, + "TTTTCC": 1375, + "TTTTCG": 1376, + "TTTTGA": 1377, + "TTTTGT": 1378, + "TTTTGC": 1379, + "TTTTGG": 1380, + "TTTCAA": 1381, + "TTTCAT": 1382, + "TTTCAC": 1383, + "TTTCAG": 1384, + "TTTCTA": 1385, + "TTTCTT": 1386, + "TTTCTC": 1387, + "TTTCTG": 1388, + "TTTCCA": 1389, + "TTTCCT": 1390, + "TTTCCC": 1391, + "TTTCCG": 1392, + "TTTCGA": 1393, + "TTTCGT": 1394, + "TTTCGC": 1395, + "TTTCGG": 1396, + "TTTGAA": 1397, + "TTTGAT": 1398, + "TTTGAC": 1399, + "TTTGAG": 1400, + "TTTGTA": 1401, + "TTTGTT": 1402, + "TTTGTC": 1403, + "TTTGTG": 1404, + "TTTGCA": 1405, + "TTTGCT": 1406, + "TTTGCC": 1407, + "TTTGCG": 1408, + "TTTGGA": 1409, + "TTTGGT": 1410, + "TTTGGC": 1411, + "TTTGGG": 1412, + "TTCAAA": 1413, + "TTCAAT": 1414, + "TTCAAC": 1415, + "TTCAAG": 1416, + "TTCATA": 1417, + "TTCATT": 1418, + "TTCATC": 1419, + "TTCATG": 1420, + "TTCACA": 1421, + "TTCACT": 1422, + "TTCACC": 1423, + "TTCACG": 1424, + "TTCAGA": 1425, + "TTCAGT": 1426, + "TTCAGC": 1427, + "TTCAGG": 1428, + "TTCTAA": 1429, + "TTCTAT": 1430, + "TTCTAC": 1431, + "TTCTAG": 1432, + "TTCTTA": 1433, + "TTCTTT": 1434, + "TTCTTC": 1435, + "TTCTTG": 1436, + "TTCTCA": 1437, + "TTCTCT": 1438, + "TTCTCC": 1439, + "TTCTCG": 1440, + "TTCTGA": 1441, + "TTCTGT": 1442, + "TTCTGC": 1443, + "TTCTGG": 1444, + "TTCCAA": 1445, + "TTCCAT": 1446, + "TTCCAC": 1447, + "TTCCAG": 1448, + "TTCCTA": 1449, + "TTCCTT": 1450, + "TTCCTC": 1451, + "TTCCTG": 1452, + "TTCCCA": 1453, + "TTCCCT": 1454, + "TTCCCC": 1455, + "TTCCCG": 1456, + "TTCCGA": 1457, + "TTCCGT": 1458, + "TTCCGC": 1459, + "TTCCGG": 1460, + "TTCGAA": 1461, + "TTCGAT": 1462, + "TTCGAC": 1463, + "TTCGAG": 1464, + "TTCGTA": 1465, + "TTCGTT": 1466, + "TTCGTC": 1467, + "TTCGTG": 1468, + "TTCGCA": 1469, + "TTCGCT": 1470, + "TTCGCC": 1471, + "TTCGCG": 1472, + "TTCGGA": 1473, + "TTCGGT": 1474, + "TTCGGC": 1475, + "TTCGGG": 1476, + "TTGAAA": 1477, + "TTGAAT": 1478, + "TTGAAC": 1479, + "TTGAAG": 1480, + "TTGATA": 1481, + "TTGATT": 1482, + "TTGATC": 1483, + "TTGATG": 1484, + "TTGACA": 1485, + "TTGACT": 1486, + "TTGACC": 1487, + "TTGACG": 1488, + "TTGAGA": 1489, + "TTGAGT": 1490, + "TTGAGC": 1491, + "TTGAGG": 1492, + "TTGTAA": 1493, + "TTGTAT": 1494, + "TTGTAC": 1495, + "TTGTAG": 1496, + "TTGTTA": 1497, + "TTGTTT": 1498, + "TTGTTC": 1499, + "TTGTTG": 1500, + "TTGTCA": 1501, + "TTGTCT": 1502, + "TTGTCC": 1503, + "TTGTCG": 1504, + "TTGTGA": 1505, + "TTGTGT": 1506, + "TTGTGC": 1507, + "TTGTGG": 1508, + "TTGCAA": 1509, + "TTGCAT": 1510, + "TTGCAC": 1511, + "TTGCAG": 1512, + "TTGCTA": 1513, + "TTGCTT": 1514, + "TTGCTC": 1515, + "TTGCTG": 1516, + "TTGCCA": 1517, + "TTGCCT": 1518, + "TTGCCC": 1519, + "TTGCCG": 1520, + "TTGCGA": 1521, + "TTGCGT": 1522, + "TTGCGC": 1523, + "TTGCGG": 1524, + "TTGGAA": 1525, + "TTGGAT": 1526, + "TTGGAC": 1527, + "TTGGAG": 1528, + "TTGGTA": 1529, + "TTGGTT": 1530, + "TTGGTC": 1531, + "TTGGTG": 1532, + "TTGGCA": 1533, + "TTGGCT": 1534, + "TTGGCC": 1535, + "TTGGCG": 1536, + "TTGGGA": 1537, + "TTGGGT": 1538, + "TTGGGC": 1539, + "TTGGGG": 1540, + "TCAAAA": 1541, + "TCAAAT": 1542, + "TCAAAC": 1543, + "TCAAAG": 1544, + "TCAATA": 1545, + "TCAATT": 1546, + "TCAATC": 1547, + "TCAATG": 1548, + "TCAACA": 1549, + "TCAACT": 1550, + "TCAACC": 1551, + "TCAACG": 1552, + "TCAAGA": 1553, + "TCAAGT": 1554, + "TCAAGC": 1555, + "TCAAGG": 1556, + "TCATAA": 1557, + "TCATAT": 1558, + "TCATAC": 1559, + "TCATAG": 1560, + "TCATTA": 1561, + "TCATTT": 1562, + "TCATTC": 1563, + "TCATTG": 1564, + "TCATCA": 1565, + "TCATCT": 1566, + "TCATCC": 1567, + "TCATCG": 1568, + "TCATGA": 1569, + "TCATGT": 1570, + "TCATGC": 1571, + "TCATGG": 1572, + "TCACAA": 1573, + "TCACAT": 1574, + "TCACAC": 1575, + "TCACAG": 1576, + "TCACTA": 1577, + "TCACTT": 1578, + "TCACTC": 1579, + "TCACTG": 1580, + "TCACCA": 1581, + "TCACCT": 1582, + "TCACCC": 1583, + "TCACCG": 1584, + "TCACGA": 1585, + "TCACGT": 1586, + "TCACGC": 1587, + "TCACGG": 1588, + "TCAGAA": 1589, + "TCAGAT": 1590, + "TCAGAC": 1591, + "TCAGAG": 1592, + "TCAGTA": 1593, + "TCAGTT": 1594, + "TCAGTC": 1595, + "TCAGTG": 1596, + "TCAGCA": 1597, + "TCAGCT": 1598, + "TCAGCC": 1599, + "TCAGCG": 1600, + "TCAGGA": 1601, + "TCAGGT": 1602, + "TCAGGC": 1603, + "TCAGGG": 1604, + "TCTAAA": 1605, + "TCTAAT": 1606, + "TCTAAC": 1607, + "TCTAAG": 1608, + "TCTATA": 1609, + "TCTATT": 1610, + "TCTATC": 1611, + "TCTATG": 1612, + "TCTACA": 1613, + "TCTACT": 1614, + "TCTACC": 1615, + "TCTACG": 1616, + "TCTAGA": 1617, + "TCTAGT": 1618, + "TCTAGC": 1619, + "TCTAGG": 1620, + "TCTTAA": 1621, + "TCTTAT": 1622, + "TCTTAC": 1623, + "TCTTAG": 1624, + "TCTTTA": 1625, + "TCTTTT": 1626, + "TCTTTC": 1627, + "TCTTTG": 1628, + "TCTTCA": 1629, + "TCTTCT": 1630, + "TCTTCC": 1631, + "TCTTCG": 1632, + "TCTTGA": 1633, + "TCTTGT": 1634, + "TCTTGC": 1635, + "TCTTGG": 1636, + "TCTCAA": 1637, + "TCTCAT": 1638, + "TCTCAC": 1639, + "TCTCAG": 1640, + "TCTCTA": 1641, + "TCTCTT": 1642, + "TCTCTC": 1643, + "TCTCTG": 1644, + "TCTCCA": 1645, + "TCTCCT": 1646, + "TCTCCC": 1647, + "TCTCCG": 1648, + "TCTCGA": 1649, + "TCTCGT": 1650, + "TCTCGC": 1651, + "TCTCGG": 1652, + "TCTGAA": 1653, + "TCTGAT": 1654, + "TCTGAC": 1655, + "TCTGAG": 1656, + "TCTGTA": 1657, + "TCTGTT": 1658, + "TCTGTC": 1659, + "TCTGTG": 1660, + "TCTGCA": 1661, + "TCTGCT": 1662, + "TCTGCC": 1663, + "TCTGCG": 1664, + "TCTGGA": 1665, + "TCTGGT": 1666, + "TCTGGC": 1667, + "TCTGGG": 1668, + "TCCAAA": 1669, + "TCCAAT": 1670, + "TCCAAC": 1671, + "TCCAAG": 1672, + "TCCATA": 1673, + "TCCATT": 1674, + "TCCATC": 1675, + "TCCATG": 1676, + "TCCACA": 1677, + "TCCACT": 1678, + "TCCACC": 1679, + "TCCACG": 1680, + "TCCAGA": 1681, + "TCCAGT": 1682, + "TCCAGC": 1683, + "TCCAGG": 1684, + "TCCTAA": 1685, + "TCCTAT": 1686, + "TCCTAC": 1687, + "TCCTAG": 1688, + "TCCTTA": 1689, + "TCCTTT": 1690, + "TCCTTC": 1691, + "TCCTTG": 1692, + "TCCTCA": 1693, + "TCCTCT": 1694, + "TCCTCC": 1695, + "TCCTCG": 1696, + "TCCTGA": 1697, + "TCCTGT": 1698, + "TCCTGC": 1699, + "TCCTGG": 1700, + "TCCCAA": 1701, + "TCCCAT": 1702, + "TCCCAC": 1703, + "TCCCAG": 1704, + "TCCCTA": 1705, + "TCCCTT": 1706, + "TCCCTC": 1707, + "TCCCTG": 1708, + "TCCCCA": 1709, + "TCCCCT": 1710, + "TCCCCC": 1711, + "TCCCCG": 1712, + "TCCCGA": 1713, + "TCCCGT": 1714, + "TCCCGC": 1715, + "TCCCGG": 1716, + "TCCGAA": 1717, + "TCCGAT": 1718, + "TCCGAC": 1719, + "TCCGAG": 1720, + "TCCGTA": 1721, + "TCCGTT": 1722, + "TCCGTC": 1723, + "TCCGTG": 1724, + "TCCGCA": 1725, + "TCCGCT": 1726, + "TCCGCC": 1727, + "TCCGCG": 1728, + "TCCGGA": 1729, + "TCCGGT": 1730, + "TCCGGC": 1731, + "TCCGGG": 1732, + "TCGAAA": 1733, + "TCGAAT": 1734, + "TCGAAC": 1735, + "TCGAAG": 1736, + "TCGATA": 1737, + "TCGATT": 1738, + "TCGATC": 1739, + "TCGATG": 1740, + "TCGACA": 1741, + "TCGACT": 1742, + "TCGACC": 1743, + "TCGACG": 1744, + "TCGAGA": 1745, + "TCGAGT": 1746, + "TCGAGC": 1747, + "TCGAGG": 1748, + "TCGTAA": 1749, + "TCGTAT": 1750, + "TCGTAC": 1751, + "TCGTAG": 1752, + "TCGTTA": 1753, + "TCGTTT": 1754, + "TCGTTC": 1755, + "TCGTTG": 1756, + "TCGTCA": 1757, + "TCGTCT": 1758, + "TCGTCC": 1759, + "TCGTCG": 1760, + "TCGTGA": 1761, + "TCGTGT": 1762, + "TCGTGC": 1763, + "TCGTGG": 1764, + "TCGCAA": 1765, + "TCGCAT": 1766, + "TCGCAC": 1767, + "TCGCAG": 1768, + "TCGCTA": 1769, + "TCGCTT": 1770, + "TCGCTC": 1771, + "TCGCTG": 1772, + "TCGCCA": 1773, + "TCGCCT": 1774, + "TCGCCC": 1775, + "TCGCCG": 1776, + "TCGCGA": 1777, + "TCGCGT": 1778, + "TCGCGC": 1779, + "TCGCGG": 1780, + "TCGGAA": 1781, + "TCGGAT": 1782, + "TCGGAC": 1783, + "TCGGAG": 1784, + "TCGGTA": 1785, + "TCGGTT": 1786, + "TCGGTC": 1787, + "TCGGTG": 1788, + "TCGGCA": 1789, + "TCGGCT": 1790, + "TCGGCC": 1791, + "TCGGCG": 1792, + "TCGGGA": 1793, + "TCGGGT": 1794, + "TCGGGC": 1795, + "TCGGGG": 1796, + "TGAAAA": 1797, + "TGAAAT": 1798, + "TGAAAC": 1799, + "TGAAAG": 1800, + "TGAATA": 1801, + "TGAATT": 1802, + "TGAATC": 1803, + "TGAATG": 1804, + "TGAACA": 1805, + "TGAACT": 1806, + "TGAACC": 1807, + "TGAACG": 1808, + "TGAAGA": 1809, + "TGAAGT": 1810, + "TGAAGC": 1811, + "TGAAGG": 1812, + "TGATAA": 1813, + "TGATAT": 1814, + "TGATAC": 1815, + "TGATAG": 1816, + "TGATTA": 1817, + "TGATTT": 1818, + "TGATTC": 1819, + "TGATTG": 1820, + "TGATCA": 1821, + "TGATCT": 1822, + "TGATCC": 1823, + "TGATCG": 1824, + "TGATGA": 1825, + "TGATGT": 1826, + "TGATGC": 1827, + "TGATGG": 1828, + "TGACAA": 1829, + "TGACAT": 1830, + "TGACAC": 1831, + "TGACAG": 1832, + "TGACTA": 1833, + "TGACTT": 1834, + "TGACTC": 1835, + "TGACTG": 1836, + "TGACCA": 1837, + "TGACCT": 1838, + "TGACCC": 1839, + "TGACCG": 1840, + "TGACGA": 1841, + "TGACGT": 1842, + "TGACGC": 1843, + "TGACGG": 1844, + "TGAGAA": 1845, + "TGAGAT": 1846, + "TGAGAC": 1847, + "TGAGAG": 1848, + "TGAGTA": 1849, + "TGAGTT": 1850, + "TGAGTC": 1851, + "TGAGTG": 1852, + "TGAGCA": 1853, + "TGAGCT": 1854, + "TGAGCC": 1855, + "TGAGCG": 1856, + "TGAGGA": 1857, + "TGAGGT": 1858, + "TGAGGC": 1859, + "TGAGGG": 1860, + "TGTAAA": 1861, + "TGTAAT": 1862, + "TGTAAC": 1863, + "TGTAAG": 1864, + "TGTATA": 1865, + "TGTATT": 1866, + "TGTATC": 1867, + "TGTATG": 1868, + "TGTACA": 1869, + "TGTACT": 1870, + "TGTACC": 1871, + "TGTACG": 1872, + "TGTAGA": 1873, + "TGTAGT": 1874, + "TGTAGC": 1875, + "TGTAGG": 1876, + "TGTTAA": 1877, + "TGTTAT": 1878, + "TGTTAC": 1879, + "TGTTAG": 1880, + "TGTTTA": 1881, + "TGTTTT": 1882, + "TGTTTC": 1883, + "TGTTTG": 1884, + "TGTTCA": 1885, + "TGTTCT": 1886, + "TGTTCC": 1887, + "TGTTCG": 1888, + "TGTTGA": 1889, + "TGTTGT": 1890, + "TGTTGC": 1891, + "TGTTGG": 1892, + "TGTCAA": 1893, + "TGTCAT": 1894, + "TGTCAC": 1895, + "TGTCAG": 1896, + "TGTCTA": 1897, + "TGTCTT": 1898, + "TGTCTC": 1899, + "TGTCTG": 1900, + "TGTCCA": 1901, + "TGTCCT": 1902, + "TGTCCC": 1903, + "TGTCCG": 1904, + "TGTCGA": 1905, + "TGTCGT": 1906, + "TGTCGC": 1907, + "TGTCGG": 1908, + "TGTGAA": 1909, + "TGTGAT": 1910, + "TGTGAC": 1911, + "TGTGAG": 1912, + "TGTGTA": 1913, + "TGTGTT": 1914, + "TGTGTC": 1915, + "TGTGTG": 1916, + "TGTGCA": 1917, + "TGTGCT": 1918, + "TGTGCC": 1919, + "TGTGCG": 1920, + "TGTGGA": 1921, + "TGTGGT": 1922, + "TGTGGC": 1923, + "TGTGGG": 1924, + "TGCAAA": 1925, + "TGCAAT": 1926, + "TGCAAC": 1927, + "TGCAAG": 1928, + "TGCATA": 1929, + "TGCATT": 1930, + "TGCATC": 1931, + "TGCATG": 1932, + "TGCACA": 1933, + "TGCACT": 1934, + "TGCACC": 1935, + "TGCACG": 1936, + "TGCAGA": 1937, + "TGCAGT": 1938, + "TGCAGC": 1939, + "TGCAGG": 1940, + "TGCTAA": 1941, + "TGCTAT": 1942, + "TGCTAC": 1943, + "TGCTAG": 1944, + "TGCTTA": 1945, + "TGCTTT": 1946, + "TGCTTC": 1947, + "TGCTTG": 1948, + "TGCTCA": 1949, + "TGCTCT": 1950, + "TGCTCC": 1951, + "TGCTCG": 1952, + "TGCTGA": 1953, + "TGCTGT": 1954, + "TGCTGC": 1955, + "TGCTGG": 1956, + "TGCCAA": 1957, + "TGCCAT": 1958, + "TGCCAC": 1959, + "TGCCAG": 1960, + "TGCCTA": 1961, + "TGCCTT": 1962, + "TGCCTC": 1963, + "TGCCTG": 1964, + "TGCCCA": 1965, + "TGCCCT": 1966, + "TGCCCC": 1967, + "TGCCCG": 1968, + "TGCCGA": 1969, + "TGCCGT": 1970, + "TGCCGC": 1971, + "TGCCGG": 1972, + "TGCGAA": 1973, + "TGCGAT": 1974, + "TGCGAC": 1975, + "TGCGAG": 1976, + "TGCGTA": 1977, + "TGCGTT": 1978, + "TGCGTC": 1979, + "TGCGTG": 1980, + "TGCGCA": 1981, + "TGCGCT": 1982, + "TGCGCC": 1983, + "TGCGCG": 1984, + "TGCGGA": 1985, + "TGCGGT": 1986, + "TGCGGC": 1987, + "TGCGGG": 1988, + "TGGAAA": 1989, + "TGGAAT": 1990, + "TGGAAC": 1991, + "TGGAAG": 1992, + "TGGATA": 1993, + "TGGATT": 1994, + "TGGATC": 1995, + "TGGATG": 1996, + "TGGACA": 1997, + "TGGACT": 1998, + "TGGACC": 1999, + "TGGACG": 2000, + "TGGAGA": 2001, + "TGGAGT": 2002, + "TGGAGC": 2003, + "TGGAGG": 2004, + "TGGTAA": 2005, + "TGGTAT": 2006, + "TGGTAC": 2007, + "TGGTAG": 2008, + "TGGTTA": 2009, + "TGGTTT": 2010, + "TGGTTC": 2011, + "TGGTTG": 2012, + "TGGTCA": 2013, + "TGGTCT": 2014, + "TGGTCC": 2015, + "TGGTCG": 2016, + "TGGTGA": 2017, + "TGGTGT": 2018, + "TGGTGC": 2019, + "TGGTGG": 2020, + "TGGCAA": 2021, + "TGGCAT": 2022, + "TGGCAC": 2023, + "TGGCAG": 2024, + "TGGCTA": 2025, + "TGGCTT": 2026, + "TGGCTC": 2027, + "TGGCTG": 2028, + "TGGCCA": 2029, + "TGGCCT": 2030, + "TGGCCC": 2031, + "TGGCCG": 2032, + "TGGCGA": 2033, + "TGGCGT": 2034, + "TGGCGC": 2035, + "TGGCGG": 2036, + "TGGGAA": 2037, + "TGGGAT": 2038, + "TGGGAC": 2039, + "TGGGAG": 2040, + "TGGGTA": 2041, + "TGGGTT": 2042, + "TGGGTC": 2043, + "TGGGTG": 2044, + "TGGGCA": 2045, + "TGGGCT": 2046, + "TGGGCC": 2047, + "TGGGCG": 2048, + "TGGGGA": 2049, + "TGGGGT": 2050, + "TGGGGC": 2051, + "TGGGGG": 2052, + "CAAAAA": 2053, + "CAAAAT": 2054, + "CAAAAC": 2055, + "CAAAAG": 2056, + "CAAATA": 2057, + "CAAATT": 2058, + "CAAATC": 2059, + "CAAATG": 2060, + "CAAACA": 2061, + "CAAACT": 2062, + "CAAACC": 2063, + "CAAACG": 2064, + "CAAAGA": 2065, + "CAAAGT": 2066, + "CAAAGC": 2067, + "CAAAGG": 2068, + "CAATAA": 2069, + "CAATAT": 2070, + "CAATAC": 2071, + "CAATAG": 2072, + "CAATTA": 2073, + "CAATTT": 2074, + "CAATTC": 2075, + "CAATTG": 2076, + "CAATCA": 2077, + "CAATCT": 2078, + "CAATCC": 2079, + "CAATCG": 2080, + "CAATGA": 2081, + "CAATGT": 2082, + "CAATGC": 2083, + "CAATGG": 2084, + "CAACAA": 2085, + "CAACAT": 2086, + "CAACAC": 2087, + "CAACAG": 2088, + "CAACTA": 2089, + "CAACTT": 2090, + "CAACTC": 2091, + "CAACTG": 2092, + "CAACCA": 2093, + "CAACCT": 2094, + "CAACCC": 2095, + "CAACCG": 2096, + "CAACGA": 2097, + "CAACGT": 2098, + "CAACGC": 2099, + "CAACGG": 2100, + "CAAGAA": 2101, + "CAAGAT": 2102, + "CAAGAC": 2103, + "CAAGAG": 2104, + "CAAGTA": 2105, + "CAAGTT": 2106, + "CAAGTC": 2107, + "CAAGTG": 2108, + "CAAGCA": 2109, + "CAAGCT": 2110, + "CAAGCC": 2111, + "CAAGCG": 2112, + "CAAGGA": 2113, + "CAAGGT": 2114, + "CAAGGC": 2115, + "CAAGGG": 2116, + "CATAAA": 2117, + "CATAAT": 2118, + "CATAAC": 2119, + "CATAAG": 2120, + "CATATA": 2121, + "CATATT": 2122, + "CATATC": 2123, + "CATATG": 2124, + "CATACA": 2125, + "CATACT": 2126, + "CATACC": 2127, + "CATACG": 2128, + "CATAGA": 2129, + "CATAGT": 2130, + "CATAGC": 2131, + "CATAGG": 2132, + "CATTAA": 2133, + "CATTAT": 2134, + "CATTAC": 2135, + "CATTAG": 2136, + "CATTTA": 2137, + "CATTTT": 2138, + "CATTTC": 2139, + "CATTTG": 2140, + "CATTCA": 2141, + "CATTCT": 2142, + "CATTCC": 2143, + "CATTCG": 2144, + "CATTGA": 2145, + "CATTGT": 2146, + "CATTGC": 2147, + "CATTGG": 2148, + "CATCAA": 2149, + "CATCAT": 2150, + "CATCAC": 2151, + "CATCAG": 2152, + "CATCTA": 2153, + "CATCTT": 2154, + "CATCTC": 2155, + "CATCTG": 2156, + "CATCCA": 2157, + "CATCCT": 2158, + "CATCCC": 2159, + "CATCCG": 2160, + "CATCGA": 2161, + "CATCGT": 2162, + "CATCGC": 2163, + "CATCGG": 2164, + "CATGAA": 2165, + "CATGAT": 2166, + "CATGAC": 2167, + "CATGAG": 2168, + "CATGTA": 2169, + "CATGTT": 2170, + "CATGTC": 2171, + "CATGTG": 2172, + "CATGCA": 2173, + "CATGCT": 2174, + "CATGCC": 2175, + "CATGCG": 2176, + "CATGGA": 2177, + "CATGGT": 2178, + "CATGGC": 2179, + "CATGGG": 2180, + "CACAAA": 2181, + "CACAAT": 2182, + "CACAAC": 2183, + "CACAAG": 2184, + "CACATA": 2185, + "CACATT": 2186, + "CACATC": 2187, + "CACATG": 2188, + "CACACA": 2189, + "CACACT": 2190, + "CACACC": 2191, + "CACACG": 2192, + "CACAGA": 2193, + "CACAGT": 2194, + "CACAGC": 2195, + "CACAGG": 2196, + "CACTAA": 2197, + "CACTAT": 2198, + "CACTAC": 2199, + "CACTAG": 2200, + "CACTTA": 2201, + "CACTTT": 2202, + "CACTTC": 2203, + "CACTTG": 2204, + "CACTCA": 2205, + "CACTCT": 2206, + "CACTCC": 2207, + "CACTCG": 2208, + "CACTGA": 2209, + "CACTGT": 2210, + "CACTGC": 2211, + "CACTGG": 2212, + "CACCAA": 2213, + "CACCAT": 2214, + "CACCAC": 2215, + "CACCAG": 2216, + "CACCTA": 2217, + "CACCTT": 2218, + "CACCTC": 2219, + "CACCTG": 2220, + "CACCCA": 2221, + "CACCCT": 2222, + "CACCCC": 2223, + "CACCCG": 2224, + "CACCGA": 2225, + "CACCGT": 2226, + "CACCGC": 2227, + "CACCGG": 2228, + "CACGAA": 2229, + "CACGAT": 2230, + "CACGAC": 2231, + "CACGAG": 2232, + "CACGTA": 2233, + "CACGTT": 2234, + "CACGTC": 2235, + "CACGTG": 2236, + "CACGCA": 2237, + "CACGCT": 2238, + "CACGCC": 2239, + "CACGCG": 2240, + "CACGGA": 2241, + "CACGGT": 2242, + "CACGGC": 2243, + "CACGGG": 2244, + "CAGAAA": 2245, + "CAGAAT": 2246, + "CAGAAC": 2247, + "CAGAAG": 2248, + "CAGATA": 2249, + "CAGATT": 2250, + "CAGATC": 2251, + "CAGATG": 2252, + "CAGACA": 2253, + "CAGACT": 2254, + "CAGACC": 2255, + "CAGACG": 2256, + "CAGAGA": 2257, + "CAGAGT": 2258, + "CAGAGC": 2259, + "CAGAGG": 2260, + "CAGTAA": 2261, + "CAGTAT": 2262, + "CAGTAC": 2263, + "CAGTAG": 2264, + "CAGTTA": 2265, + "CAGTTT": 2266, + "CAGTTC": 2267, + "CAGTTG": 2268, + "CAGTCA": 2269, + "CAGTCT": 2270, + "CAGTCC": 2271, + "CAGTCG": 2272, + "CAGTGA": 2273, + "CAGTGT": 2274, + "CAGTGC": 2275, + "CAGTGG": 2276, + "CAGCAA": 2277, + "CAGCAT": 2278, + "CAGCAC": 2279, + "CAGCAG": 2280, + "CAGCTA": 2281, + "CAGCTT": 2282, + "CAGCTC": 2283, + "CAGCTG": 2284, + "CAGCCA": 2285, + "CAGCCT": 2286, + "CAGCCC": 2287, + "CAGCCG": 2288, + "CAGCGA": 2289, + "CAGCGT": 2290, + "CAGCGC": 2291, + "CAGCGG": 2292, + "CAGGAA": 2293, + "CAGGAT": 2294, + "CAGGAC": 2295, + "CAGGAG": 2296, + "CAGGTA": 2297, + "CAGGTT": 2298, + "CAGGTC": 2299, + "CAGGTG": 2300, + "CAGGCA": 2301, + "CAGGCT": 2302, + "CAGGCC": 2303, + "CAGGCG": 2304, + "CAGGGA": 2305, + "CAGGGT": 2306, + "CAGGGC": 2307, + "CAGGGG": 2308, + "CTAAAA": 2309, + "CTAAAT": 2310, + "CTAAAC": 2311, + "CTAAAG": 2312, + "CTAATA": 2313, + "CTAATT": 2314, + "CTAATC": 2315, + "CTAATG": 2316, + "CTAACA": 2317, + "CTAACT": 2318, + "CTAACC": 2319, + "CTAACG": 2320, + "CTAAGA": 2321, + "CTAAGT": 2322, + "CTAAGC": 2323, + "CTAAGG": 2324, + "CTATAA": 2325, + "CTATAT": 2326, + "CTATAC": 2327, + "CTATAG": 2328, + "CTATTA": 2329, + "CTATTT": 2330, + "CTATTC": 2331, + "CTATTG": 2332, + "CTATCA": 2333, + "CTATCT": 2334, + "CTATCC": 2335, + "CTATCG": 2336, + "CTATGA": 2337, + "CTATGT": 2338, + "CTATGC": 2339, + "CTATGG": 2340, + "CTACAA": 2341, + "CTACAT": 2342, + "CTACAC": 2343, + "CTACAG": 2344, + "CTACTA": 2345, + "CTACTT": 2346, + "CTACTC": 2347, + "CTACTG": 2348, + "CTACCA": 2349, + "CTACCT": 2350, + "CTACCC": 2351, + "CTACCG": 2352, + "CTACGA": 2353, + "CTACGT": 2354, + "CTACGC": 2355, + "CTACGG": 2356, + "CTAGAA": 2357, + "CTAGAT": 2358, + "CTAGAC": 2359, + "CTAGAG": 2360, + "CTAGTA": 2361, + "CTAGTT": 2362, + "CTAGTC": 2363, + "CTAGTG": 2364, + "CTAGCA": 2365, + "CTAGCT": 2366, + "CTAGCC": 2367, + "CTAGCG": 2368, + "CTAGGA": 2369, + "CTAGGT": 2370, + "CTAGGC": 2371, + "CTAGGG": 2372, + "CTTAAA": 2373, + "CTTAAT": 2374, + "CTTAAC": 2375, + "CTTAAG": 2376, + "CTTATA": 2377, + "CTTATT": 2378, + "CTTATC": 2379, + "CTTATG": 2380, + "CTTACA": 2381, + "CTTACT": 2382, + "CTTACC": 2383, + "CTTACG": 2384, + "CTTAGA": 2385, + "CTTAGT": 2386, + "CTTAGC": 2387, + "CTTAGG": 2388, + "CTTTAA": 2389, + "CTTTAT": 2390, + "CTTTAC": 2391, + "CTTTAG": 2392, + "CTTTTA": 2393, + "CTTTTT": 2394, + "CTTTTC": 2395, + "CTTTTG": 2396, + "CTTTCA": 2397, + "CTTTCT": 2398, + "CTTTCC": 2399, + "CTTTCG": 2400, + "CTTTGA": 2401, + "CTTTGT": 2402, + "CTTTGC": 2403, + "CTTTGG": 2404, + "CTTCAA": 2405, + "CTTCAT": 2406, + "CTTCAC": 2407, + "CTTCAG": 2408, + "CTTCTA": 2409, + "CTTCTT": 2410, + "CTTCTC": 2411, + "CTTCTG": 2412, + "CTTCCA": 2413, + "CTTCCT": 2414, + "CTTCCC": 2415, + "CTTCCG": 2416, + "CTTCGA": 2417, + "CTTCGT": 2418, + "CTTCGC": 2419, + "CTTCGG": 2420, + "CTTGAA": 2421, + "CTTGAT": 2422, + "CTTGAC": 2423, + "CTTGAG": 2424, + "CTTGTA": 2425, + "CTTGTT": 2426, + "CTTGTC": 2427, + "CTTGTG": 2428, + "CTTGCA": 2429, + "CTTGCT": 2430, + "CTTGCC": 2431, + "CTTGCG": 2432, + "CTTGGA": 2433, + "CTTGGT": 2434, + "CTTGGC": 2435, + "CTTGGG": 2436, + "CTCAAA": 2437, + "CTCAAT": 2438, + "CTCAAC": 2439, + "CTCAAG": 2440, + "CTCATA": 2441, + "CTCATT": 2442, + "CTCATC": 2443, + "CTCATG": 2444, + "CTCACA": 2445, + "CTCACT": 2446, + "CTCACC": 2447, + "CTCACG": 2448, + "CTCAGA": 2449, + "CTCAGT": 2450, + "CTCAGC": 2451, + "CTCAGG": 2452, + "CTCTAA": 2453, + "CTCTAT": 2454, + "CTCTAC": 2455, + "CTCTAG": 2456, + "CTCTTA": 2457, + "CTCTTT": 2458, + "CTCTTC": 2459, + "CTCTTG": 2460, + "CTCTCA": 2461, + "CTCTCT": 2462, + "CTCTCC": 2463, + "CTCTCG": 2464, + "CTCTGA": 2465, + "CTCTGT": 2466, + "CTCTGC": 2467, + "CTCTGG": 2468, + "CTCCAA": 2469, + "CTCCAT": 2470, + "CTCCAC": 2471, + "CTCCAG": 2472, + "CTCCTA": 2473, + "CTCCTT": 2474, + "CTCCTC": 2475, + "CTCCTG": 2476, + "CTCCCA": 2477, + "CTCCCT": 2478, + "CTCCCC": 2479, + "CTCCCG": 2480, + "CTCCGA": 2481, + "CTCCGT": 2482, + "CTCCGC": 2483, + "CTCCGG": 2484, + "CTCGAA": 2485, + "CTCGAT": 2486, + "CTCGAC": 2487, + "CTCGAG": 2488, + "CTCGTA": 2489, + "CTCGTT": 2490, + "CTCGTC": 2491, + "CTCGTG": 2492, + "CTCGCA": 2493, + "CTCGCT": 2494, + "CTCGCC": 2495, + "CTCGCG": 2496, + "CTCGGA": 2497, + "CTCGGT": 2498, + "CTCGGC": 2499, + "CTCGGG": 2500, + "CTGAAA": 2501, + "CTGAAT": 2502, + "CTGAAC": 2503, + "CTGAAG": 2504, + "CTGATA": 2505, + "CTGATT": 2506, + "CTGATC": 2507, + "CTGATG": 2508, + "CTGACA": 2509, + "CTGACT": 2510, + "CTGACC": 2511, + "CTGACG": 2512, + "CTGAGA": 2513, + "CTGAGT": 2514, + "CTGAGC": 2515, + "CTGAGG": 2516, + "CTGTAA": 2517, + "CTGTAT": 2518, + "CTGTAC": 2519, + "CTGTAG": 2520, + "CTGTTA": 2521, + "CTGTTT": 2522, + "CTGTTC": 2523, + "CTGTTG": 2524, + "CTGTCA": 2525, + "CTGTCT": 2526, + "CTGTCC": 2527, + "CTGTCG": 2528, + "CTGTGA": 2529, + "CTGTGT": 2530, + "CTGTGC": 2531, + "CTGTGG": 2532, + "CTGCAA": 2533, + "CTGCAT": 2534, + "CTGCAC": 2535, + "CTGCAG": 2536, + "CTGCTA": 2537, + "CTGCTT": 2538, + "CTGCTC": 2539, + "CTGCTG": 2540, + "CTGCCA": 2541, + "CTGCCT": 2542, + "CTGCCC": 2543, + "CTGCCG": 2544, + "CTGCGA": 2545, + "CTGCGT": 2546, + "CTGCGC": 2547, + "CTGCGG": 2548, + "CTGGAA": 2549, + "CTGGAT": 2550, + "CTGGAC": 2551, + "CTGGAG": 2552, + "CTGGTA": 2553, + "CTGGTT": 2554, + "CTGGTC": 2555, + "CTGGTG": 2556, + "CTGGCA": 2557, + "CTGGCT": 2558, + "CTGGCC": 2559, + "CTGGCG": 2560, + "CTGGGA": 2561, + "CTGGGT": 2562, + "CTGGGC": 2563, + "CTGGGG": 2564, + "CCAAAA": 2565, + "CCAAAT": 2566, + "CCAAAC": 2567, + "CCAAAG": 2568, + "CCAATA": 2569, + "CCAATT": 2570, + "CCAATC": 2571, + "CCAATG": 2572, + "CCAACA": 2573, + "CCAACT": 2574, + "CCAACC": 2575, + "CCAACG": 2576, + "CCAAGA": 2577, + "CCAAGT": 2578, + "CCAAGC": 2579, + "CCAAGG": 2580, + "CCATAA": 2581, + "CCATAT": 2582, + "CCATAC": 2583, + "CCATAG": 2584, + "CCATTA": 2585, + "CCATTT": 2586, + "CCATTC": 2587, + "CCATTG": 2588, + "CCATCA": 2589, + "CCATCT": 2590, + "CCATCC": 2591, + "CCATCG": 2592, + "CCATGA": 2593, + "CCATGT": 2594, + "CCATGC": 2595, + "CCATGG": 2596, + "CCACAA": 2597, + "CCACAT": 2598, + "CCACAC": 2599, + "CCACAG": 2600, + "CCACTA": 2601, + "CCACTT": 2602, + "CCACTC": 2603, + "CCACTG": 2604, + "CCACCA": 2605, + "CCACCT": 2606, + "CCACCC": 2607, + "CCACCG": 2608, + "CCACGA": 2609, + "CCACGT": 2610, + "CCACGC": 2611, + "CCACGG": 2612, + "CCAGAA": 2613, + "CCAGAT": 2614, + "CCAGAC": 2615, + "CCAGAG": 2616, + "CCAGTA": 2617, + "CCAGTT": 2618, + "CCAGTC": 2619, + "CCAGTG": 2620, + "CCAGCA": 2621, + "CCAGCT": 2622, + "CCAGCC": 2623, + "CCAGCG": 2624, + "CCAGGA": 2625, + "CCAGGT": 2626, + "CCAGGC": 2627, + "CCAGGG": 2628, + "CCTAAA": 2629, + "CCTAAT": 2630, + "CCTAAC": 2631, + "CCTAAG": 2632, + "CCTATA": 2633, + "CCTATT": 2634, + "CCTATC": 2635, + "CCTATG": 2636, + "CCTACA": 2637, + "CCTACT": 2638, + "CCTACC": 2639, + "CCTACG": 2640, + "CCTAGA": 2641, + "CCTAGT": 2642, + "CCTAGC": 2643, + "CCTAGG": 2644, + "CCTTAA": 2645, + "CCTTAT": 2646, + "CCTTAC": 2647, + "CCTTAG": 2648, + "CCTTTA": 2649, + "CCTTTT": 2650, + "CCTTTC": 2651, + "CCTTTG": 2652, + "CCTTCA": 2653, + "CCTTCT": 2654, + "CCTTCC": 2655, + "CCTTCG": 2656, + "CCTTGA": 2657, + "CCTTGT": 2658, + "CCTTGC": 2659, + "CCTTGG": 2660, + "CCTCAA": 2661, + "CCTCAT": 2662, + "CCTCAC": 2663, + "CCTCAG": 2664, + "CCTCTA": 2665, + "CCTCTT": 2666, + "CCTCTC": 2667, + "CCTCTG": 2668, + "CCTCCA": 2669, + "CCTCCT": 2670, + "CCTCCC": 2671, + "CCTCCG": 2672, + "CCTCGA": 2673, + "CCTCGT": 2674, + "CCTCGC": 2675, + "CCTCGG": 2676, + "CCTGAA": 2677, + "CCTGAT": 2678, + "CCTGAC": 2679, + "CCTGAG": 2680, + "CCTGTA": 2681, + "CCTGTT": 2682, + "CCTGTC": 2683, + "CCTGTG": 2684, + "CCTGCA": 2685, + "CCTGCT": 2686, + "CCTGCC": 2687, + "CCTGCG": 2688, + "CCTGGA": 2689, + "CCTGGT": 2690, + "CCTGGC": 2691, + "CCTGGG": 2692, + "CCCAAA": 2693, + "CCCAAT": 2694, + "CCCAAC": 2695, + "CCCAAG": 2696, + "CCCATA": 2697, + "CCCATT": 2698, + "CCCATC": 2699, + "CCCATG": 2700, + "CCCACA": 2701, + "CCCACT": 2702, + "CCCACC": 2703, + "CCCACG": 2704, + "CCCAGA": 2705, + "CCCAGT": 2706, + "CCCAGC": 2707, + "CCCAGG": 2708, + "CCCTAA": 2709, + "CCCTAT": 2710, + "CCCTAC": 2711, + "CCCTAG": 2712, + "CCCTTA": 2713, + "CCCTTT": 2714, + "CCCTTC": 2715, + "CCCTTG": 2716, + "CCCTCA": 2717, + "CCCTCT": 2718, + "CCCTCC": 2719, + "CCCTCG": 2720, + "CCCTGA": 2721, + "CCCTGT": 2722, + "CCCTGC": 2723, + "CCCTGG": 2724, + "CCCCAA": 2725, + "CCCCAT": 2726, + "CCCCAC": 2727, + "CCCCAG": 2728, + "CCCCTA": 2729, + "CCCCTT": 2730, + "CCCCTC": 2731, + "CCCCTG": 2732, + "CCCCCA": 2733, + "CCCCCT": 2734, + "CCCCCC": 2735, + "CCCCCG": 2736, + "CCCCGA": 2737, + "CCCCGT": 2738, + "CCCCGC": 2739, + "CCCCGG": 2740, + "CCCGAA": 2741, + "CCCGAT": 2742, + "CCCGAC": 2743, + "CCCGAG": 2744, + "CCCGTA": 2745, + "CCCGTT": 2746, + "CCCGTC": 2747, + "CCCGTG": 2748, + "CCCGCA": 2749, + "CCCGCT": 2750, + "CCCGCC": 2751, + "CCCGCG": 2752, + "CCCGGA": 2753, + "CCCGGT": 2754, + "CCCGGC": 2755, + "CCCGGG": 2756, + "CCGAAA": 2757, + "CCGAAT": 2758, + "CCGAAC": 2759, + "CCGAAG": 2760, + "CCGATA": 2761, + "CCGATT": 2762, + "CCGATC": 2763, + "CCGATG": 2764, + "CCGACA": 2765, + "CCGACT": 2766, + "CCGACC": 2767, + "CCGACG": 2768, + "CCGAGA": 2769, + "CCGAGT": 2770, + "CCGAGC": 2771, + "CCGAGG": 2772, + "CCGTAA": 2773, + "CCGTAT": 2774, + "CCGTAC": 2775, + "CCGTAG": 2776, + "CCGTTA": 2777, + "CCGTTT": 2778, + "CCGTTC": 2779, + "CCGTTG": 2780, + "CCGTCA": 2781, + "CCGTCT": 2782, + "CCGTCC": 2783, + "CCGTCG": 2784, + "CCGTGA": 2785, + "CCGTGT": 2786, + "CCGTGC": 2787, + "CCGTGG": 2788, + "CCGCAA": 2789, + "CCGCAT": 2790, + "CCGCAC": 2791, + "CCGCAG": 2792, + "CCGCTA": 2793, + "CCGCTT": 2794, + "CCGCTC": 2795, + "CCGCTG": 2796, + "CCGCCA": 2797, + "CCGCCT": 2798, + "CCGCCC": 2799, + "CCGCCG": 2800, + "CCGCGA": 2801, + "CCGCGT": 2802, + "CCGCGC": 2803, + "CCGCGG": 2804, + "CCGGAA": 2805, + "CCGGAT": 2806, + "CCGGAC": 2807, + "CCGGAG": 2808, + "CCGGTA": 2809, + "CCGGTT": 2810, + "CCGGTC": 2811, + "CCGGTG": 2812, + "CCGGCA": 2813, + "CCGGCT": 2814, + "CCGGCC": 2815, + "CCGGCG": 2816, + "CCGGGA": 2817, + "CCGGGT": 2818, + "CCGGGC": 2819, + "CCGGGG": 2820, + "CGAAAA": 2821, + "CGAAAT": 2822, + "CGAAAC": 2823, + "CGAAAG": 2824, + "CGAATA": 2825, + "CGAATT": 2826, + "CGAATC": 2827, + "CGAATG": 2828, + "CGAACA": 2829, + "CGAACT": 2830, + "CGAACC": 2831, + "CGAACG": 2832, + "CGAAGA": 2833, + "CGAAGT": 2834, + "CGAAGC": 2835, + "CGAAGG": 2836, + "CGATAA": 2837, + "CGATAT": 2838, + "CGATAC": 2839, + "CGATAG": 2840, + "CGATTA": 2841, + "CGATTT": 2842, + "CGATTC": 2843, + "CGATTG": 2844, + "CGATCA": 2845, + "CGATCT": 2846, + "CGATCC": 2847, + "CGATCG": 2848, + "CGATGA": 2849, + "CGATGT": 2850, + "CGATGC": 2851, + "CGATGG": 2852, + "CGACAA": 2853, + "CGACAT": 2854, + "CGACAC": 2855, + "CGACAG": 2856, + "CGACTA": 2857, + "CGACTT": 2858, + "CGACTC": 2859, + "CGACTG": 2860, + "CGACCA": 2861, + "CGACCT": 2862, + "CGACCC": 2863, + "CGACCG": 2864, + "CGACGA": 2865, + "CGACGT": 2866, + "CGACGC": 2867, + "CGACGG": 2868, + "CGAGAA": 2869, + "CGAGAT": 2870, + "CGAGAC": 2871, + "CGAGAG": 2872, + "CGAGTA": 2873, + "CGAGTT": 2874, + "CGAGTC": 2875, + "CGAGTG": 2876, + "CGAGCA": 2877, + "CGAGCT": 2878, + "CGAGCC": 2879, + "CGAGCG": 2880, + "CGAGGA": 2881, + "CGAGGT": 2882, + "CGAGGC": 2883, + "CGAGGG": 2884, + "CGTAAA": 2885, + "CGTAAT": 2886, + "CGTAAC": 2887, + "CGTAAG": 2888, + "CGTATA": 2889, + "CGTATT": 2890, + "CGTATC": 2891, + "CGTATG": 2892, + "CGTACA": 2893, + "CGTACT": 2894, + "CGTACC": 2895, + "CGTACG": 2896, + "CGTAGA": 2897, + "CGTAGT": 2898, + "CGTAGC": 2899, + "CGTAGG": 2900, + "CGTTAA": 2901, + "CGTTAT": 2902, + "CGTTAC": 2903, + "CGTTAG": 2904, + "CGTTTA": 2905, + "CGTTTT": 2906, + "CGTTTC": 2907, + "CGTTTG": 2908, + "CGTTCA": 2909, + "CGTTCT": 2910, + "CGTTCC": 2911, + "CGTTCG": 2912, + "CGTTGA": 2913, + "CGTTGT": 2914, + "CGTTGC": 2915, + "CGTTGG": 2916, + "CGTCAA": 2917, + "CGTCAT": 2918, + "CGTCAC": 2919, + "CGTCAG": 2920, + "CGTCTA": 2921, + "CGTCTT": 2922, + "CGTCTC": 2923, + "CGTCTG": 2924, + "CGTCCA": 2925, + "CGTCCT": 2926, + "CGTCCC": 2927, + "CGTCCG": 2928, + "CGTCGA": 2929, + "CGTCGT": 2930, + "CGTCGC": 2931, + "CGTCGG": 2932, + "CGTGAA": 2933, + "CGTGAT": 2934, + "CGTGAC": 2935, + "CGTGAG": 2936, + "CGTGTA": 2937, + "CGTGTT": 2938, + "CGTGTC": 2939, + "CGTGTG": 2940, + "CGTGCA": 2941, + "CGTGCT": 2942, + "CGTGCC": 2943, + "CGTGCG": 2944, + "CGTGGA": 2945, + "CGTGGT": 2946, + "CGTGGC": 2947, + "CGTGGG": 2948, + "CGCAAA": 2949, + "CGCAAT": 2950, + "CGCAAC": 2951, + "CGCAAG": 2952, + "CGCATA": 2953, + "CGCATT": 2954, + "CGCATC": 2955, + "CGCATG": 2956, + "CGCACA": 2957, + "CGCACT": 2958, + "CGCACC": 2959, + "CGCACG": 2960, + "CGCAGA": 2961, + "CGCAGT": 2962, + "CGCAGC": 2963, + "CGCAGG": 2964, + "CGCTAA": 2965, + "CGCTAT": 2966, + "CGCTAC": 2967, + "CGCTAG": 2968, + "CGCTTA": 2969, + "CGCTTT": 2970, + "CGCTTC": 2971, + "CGCTTG": 2972, + "CGCTCA": 2973, + "CGCTCT": 2974, + "CGCTCC": 2975, + "CGCTCG": 2976, + "CGCTGA": 2977, + "CGCTGT": 2978, + "CGCTGC": 2979, + "CGCTGG": 2980, + "CGCCAA": 2981, + "CGCCAT": 2982, + "CGCCAC": 2983, + "CGCCAG": 2984, + "CGCCTA": 2985, + "CGCCTT": 2986, + "CGCCTC": 2987, + "CGCCTG": 2988, + "CGCCCA": 2989, + "CGCCCT": 2990, + "CGCCCC": 2991, + "CGCCCG": 2992, + "CGCCGA": 2993, + "CGCCGT": 2994, + "CGCCGC": 2995, + "CGCCGG": 2996, + "CGCGAA": 2997, + "CGCGAT": 2998, + "CGCGAC": 2999, + "CGCGAG": 3000, + "CGCGTA": 3001, + "CGCGTT": 3002, + "CGCGTC": 3003, + "CGCGTG": 3004, + "CGCGCA": 3005, + "CGCGCT": 3006, + "CGCGCC": 3007, + "CGCGCG": 3008, + "CGCGGA": 3009, + "CGCGGT": 3010, + "CGCGGC": 3011, + "CGCGGG": 3012, + "CGGAAA": 3013, + "CGGAAT": 3014, + "CGGAAC": 3015, + "CGGAAG": 3016, + "CGGATA": 3017, + "CGGATT": 3018, + "CGGATC": 3019, + "CGGATG": 3020, + "CGGACA": 3021, + "CGGACT": 3022, + "CGGACC": 3023, + "CGGACG": 3024, + "CGGAGA": 3025, + "CGGAGT": 3026, + "CGGAGC": 3027, + "CGGAGG": 3028, + "CGGTAA": 3029, + "CGGTAT": 3030, + "CGGTAC": 3031, + "CGGTAG": 3032, + "CGGTTA": 3033, + "CGGTTT": 3034, + "CGGTTC": 3035, + "CGGTTG": 3036, + "CGGTCA": 3037, + "CGGTCT": 3038, + "CGGTCC": 3039, + "CGGTCG": 3040, + "CGGTGA": 3041, + "CGGTGT": 3042, + "CGGTGC": 3043, + "CGGTGG": 3044, + "CGGCAA": 3045, + "CGGCAT": 3046, + "CGGCAC": 3047, + "CGGCAG": 3048, + "CGGCTA": 3049, + "CGGCTT": 3050, + "CGGCTC": 3051, + "CGGCTG": 3052, + "CGGCCA": 3053, + "CGGCCT": 3054, + "CGGCCC": 3055, + "CGGCCG": 3056, + "CGGCGA": 3057, + "CGGCGT": 3058, + "CGGCGC": 3059, + "CGGCGG": 3060, + "CGGGAA": 3061, + "CGGGAT": 3062, + "CGGGAC": 3063, + "CGGGAG": 3064, + "CGGGTA": 3065, + "CGGGTT": 3066, + "CGGGTC": 3067, + "CGGGTG": 3068, + "CGGGCA": 3069, + "CGGGCT": 3070, + "CGGGCC": 3071, + "CGGGCG": 3072, + "CGGGGA": 3073, + "CGGGGT": 3074, + "CGGGGC": 3075, + "CGGGGG": 3076, + "GAAAAA": 3077, + "GAAAAT": 3078, + "GAAAAC": 3079, + "GAAAAG": 3080, + "GAAATA": 3081, + "GAAATT": 3082, + "GAAATC": 3083, + "GAAATG": 3084, + "GAAACA": 3085, + "GAAACT": 3086, + "GAAACC": 3087, + "GAAACG": 3088, + "GAAAGA": 3089, + "GAAAGT": 3090, + "GAAAGC": 3091, + "GAAAGG": 3092, + "GAATAA": 3093, + "GAATAT": 3094, + "GAATAC": 3095, + "GAATAG": 3096, + "GAATTA": 3097, + "GAATTT": 3098, + "GAATTC": 3099, + "GAATTG": 3100, + "GAATCA": 3101, + "GAATCT": 3102, + "GAATCC": 3103, + "GAATCG": 3104, + "GAATGA": 3105, + "GAATGT": 3106, + "GAATGC": 3107, + "GAATGG": 3108, + "GAACAA": 3109, + "GAACAT": 3110, + "GAACAC": 3111, + "GAACAG": 3112, + "GAACTA": 3113, + "GAACTT": 3114, + "GAACTC": 3115, + "GAACTG": 3116, + "GAACCA": 3117, + "GAACCT": 3118, + "GAACCC": 3119, + "GAACCG": 3120, + "GAACGA": 3121, + "GAACGT": 3122, + "GAACGC": 3123, + "GAACGG": 3124, + "GAAGAA": 3125, + "GAAGAT": 3126, + "GAAGAC": 3127, + "GAAGAG": 3128, + "GAAGTA": 3129, + "GAAGTT": 3130, + "GAAGTC": 3131, + "GAAGTG": 3132, + "GAAGCA": 3133, + "GAAGCT": 3134, + "GAAGCC": 3135, + "GAAGCG": 3136, + "GAAGGA": 3137, + "GAAGGT": 3138, + "GAAGGC": 3139, + "GAAGGG": 3140, + "GATAAA": 3141, + "GATAAT": 3142, + "GATAAC": 3143, + "GATAAG": 3144, + "GATATA": 3145, + "GATATT": 3146, + "GATATC": 3147, + "GATATG": 3148, + "GATACA": 3149, + "GATACT": 3150, + "GATACC": 3151, + "GATACG": 3152, + "GATAGA": 3153, + "GATAGT": 3154, + "GATAGC": 3155, + "GATAGG": 3156, + "GATTAA": 3157, + "GATTAT": 3158, + "GATTAC": 3159, + "GATTAG": 3160, + "GATTTA": 3161, + "GATTTT": 3162, + "GATTTC": 3163, + "GATTTG": 3164, + "GATTCA": 3165, + "GATTCT": 3166, + "GATTCC": 3167, + "GATTCG": 3168, + "GATTGA": 3169, + "GATTGT": 3170, + "GATTGC": 3171, + "GATTGG": 3172, + "GATCAA": 3173, + "GATCAT": 3174, + "GATCAC": 3175, + "GATCAG": 3176, + "GATCTA": 3177, + "GATCTT": 3178, + "GATCTC": 3179, + "GATCTG": 3180, + "GATCCA": 3181, + "GATCCT": 3182, + "GATCCC": 3183, + "GATCCG": 3184, + "GATCGA": 3185, + "GATCGT": 3186, + "GATCGC": 3187, + "GATCGG": 3188, + "GATGAA": 3189, + "GATGAT": 3190, + "GATGAC": 3191, + "GATGAG": 3192, + "GATGTA": 3193, + "GATGTT": 3194, + "GATGTC": 3195, + "GATGTG": 3196, + "GATGCA": 3197, + "GATGCT": 3198, + "GATGCC": 3199, + "GATGCG": 3200, + "GATGGA": 3201, + "GATGGT": 3202, + "GATGGC": 3203, + "GATGGG": 3204, + "GACAAA": 3205, + "GACAAT": 3206, + "GACAAC": 3207, + "GACAAG": 3208, + "GACATA": 3209, + "GACATT": 3210, + "GACATC": 3211, + "GACATG": 3212, + "GACACA": 3213, + "GACACT": 3214, + "GACACC": 3215, + "GACACG": 3216, + "GACAGA": 3217, + "GACAGT": 3218, + "GACAGC": 3219, + "GACAGG": 3220, + "GACTAA": 3221, + "GACTAT": 3222, + "GACTAC": 3223, + "GACTAG": 3224, + "GACTTA": 3225, + "GACTTT": 3226, + "GACTTC": 3227, + "GACTTG": 3228, + "GACTCA": 3229, + "GACTCT": 3230, + "GACTCC": 3231, + "GACTCG": 3232, + "GACTGA": 3233, + "GACTGT": 3234, + "GACTGC": 3235, + "GACTGG": 3236, + "GACCAA": 3237, + "GACCAT": 3238, + "GACCAC": 3239, + "GACCAG": 3240, + "GACCTA": 3241, + "GACCTT": 3242, + "GACCTC": 3243, + "GACCTG": 3244, + "GACCCA": 3245, + "GACCCT": 3246, + "GACCCC": 3247, + "GACCCG": 3248, + "GACCGA": 3249, + "GACCGT": 3250, + "GACCGC": 3251, + "GACCGG": 3252, + "GACGAA": 3253, + "GACGAT": 3254, + "GACGAC": 3255, + "GACGAG": 3256, + "GACGTA": 3257, + "GACGTT": 3258, + "GACGTC": 3259, + "GACGTG": 3260, + "GACGCA": 3261, + "GACGCT": 3262, + "GACGCC": 3263, + "GACGCG": 3264, + "GACGGA": 3265, + "GACGGT": 3266, + "GACGGC": 3267, + "GACGGG": 3268, + "GAGAAA": 3269, + "GAGAAT": 3270, + "GAGAAC": 3271, + "GAGAAG": 3272, + "GAGATA": 3273, + "GAGATT": 3274, + "GAGATC": 3275, + "GAGATG": 3276, + "GAGACA": 3277, + "GAGACT": 3278, + "GAGACC": 3279, + "GAGACG": 3280, + "GAGAGA": 3281, + "GAGAGT": 3282, + "GAGAGC": 3283, + "GAGAGG": 3284, + "GAGTAA": 3285, + "GAGTAT": 3286, + "GAGTAC": 3287, + "GAGTAG": 3288, + "GAGTTA": 3289, + "GAGTTT": 3290, + "GAGTTC": 3291, + "GAGTTG": 3292, + "GAGTCA": 3293, + "GAGTCT": 3294, + "GAGTCC": 3295, + "GAGTCG": 3296, + "GAGTGA": 3297, + "GAGTGT": 3298, + "GAGTGC": 3299, + "GAGTGG": 3300, + "GAGCAA": 3301, + "GAGCAT": 3302, + "GAGCAC": 3303, + "GAGCAG": 3304, + "GAGCTA": 3305, + "GAGCTT": 3306, + "GAGCTC": 3307, + "GAGCTG": 3308, + "GAGCCA": 3309, + "GAGCCT": 3310, + "GAGCCC": 3311, + "GAGCCG": 3312, + "GAGCGA": 3313, + "GAGCGT": 3314, + "GAGCGC": 3315, + "GAGCGG": 3316, + "GAGGAA": 3317, + "GAGGAT": 3318, + "GAGGAC": 3319, + "GAGGAG": 3320, + "GAGGTA": 3321, + "GAGGTT": 3322, + "GAGGTC": 3323, + "GAGGTG": 3324, + "GAGGCA": 3325, + "GAGGCT": 3326, + "GAGGCC": 3327, + "GAGGCG": 3328, + "GAGGGA": 3329, + "GAGGGT": 3330, + "GAGGGC": 3331, + "GAGGGG": 3332, + "GTAAAA": 3333, + "GTAAAT": 3334, + "GTAAAC": 3335, + "GTAAAG": 3336, + "GTAATA": 3337, + "GTAATT": 3338, + "GTAATC": 3339, + "GTAATG": 3340, + "GTAACA": 3341, + "GTAACT": 3342, + "GTAACC": 3343, + "GTAACG": 3344, + "GTAAGA": 3345, + "GTAAGT": 3346, + "GTAAGC": 3347, + "GTAAGG": 3348, + "GTATAA": 3349, + "GTATAT": 3350, + "GTATAC": 3351, + "GTATAG": 3352, + "GTATTA": 3353, + "GTATTT": 3354, + "GTATTC": 3355, + "GTATTG": 3356, + "GTATCA": 3357, + "GTATCT": 3358, + "GTATCC": 3359, + "GTATCG": 3360, + "GTATGA": 3361, + "GTATGT": 3362, + "GTATGC": 3363, + "GTATGG": 3364, + "GTACAA": 3365, + "GTACAT": 3366, + "GTACAC": 3367, + "GTACAG": 3368, + "GTACTA": 3369, + "GTACTT": 3370, + "GTACTC": 3371, + "GTACTG": 3372, + "GTACCA": 3373, + "GTACCT": 3374, + "GTACCC": 3375, + "GTACCG": 3376, + "GTACGA": 3377, + "GTACGT": 3378, + "GTACGC": 3379, + "GTACGG": 3380, + "GTAGAA": 3381, + "GTAGAT": 3382, + "GTAGAC": 3383, + "GTAGAG": 3384, + "GTAGTA": 3385, + "GTAGTT": 3386, + "GTAGTC": 3387, + "GTAGTG": 3388, + "GTAGCA": 3389, + "GTAGCT": 3390, + "GTAGCC": 3391, + "GTAGCG": 3392, + "GTAGGA": 3393, + "GTAGGT": 3394, + "GTAGGC": 3395, + "GTAGGG": 3396, + "GTTAAA": 3397, + "GTTAAT": 3398, + "GTTAAC": 3399, + "GTTAAG": 3400, + "GTTATA": 3401, + "GTTATT": 3402, + "GTTATC": 3403, + "GTTATG": 3404, + "GTTACA": 3405, + "GTTACT": 3406, + "GTTACC": 3407, + "GTTACG": 3408, + "GTTAGA": 3409, + "GTTAGT": 3410, + "GTTAGC": 3411, + "GTTAGG": 3412, + "GTTTAA": 3413, + "GTTTAT": 3414, + "GTTTAC": 3415, + "GTTTAG": 3416, + "GTTTTA": 3417, + "GTTTTT": 3418, + "GTTTTC": 3419, + "GTTTTG": 3420, + "GTTTCA": 3421, + "GTTTCT": 3422, + "GTTTCC": 3423, + "GTTTCG": 3424, + "GTTTGA": 3425, + "GTTTGT": 3426, + "GTTTGC": 3427, + "GTTTGG": 3428, + "GTTCAA": 3429, + "GTTCAT": 3430, + "GTTCAC": 3431, + "GTTCAG": 3432, + "GTTCTA": 3433, + "GTTCTT": 3434, + "GTTCTC": 3435, + "GTTCTG": 3436, + "GTTCCA": 3437, + "GTTCCT": 3438, + "GTTCCC": 3439, + "GTTCCG": 3440, + "GTTCGA": 3441, + "GTTCGT": 3442, + "GTTCGC": 3443, + "GTTCGG": 3444, + "GTTGAA": 3445, + "GTTGAT": 3446, + "GTTGAC": 3447, + "GTTGAG": 3448, + "GTTGTA": 3449, + "GTTGTT": 3450, + "GTTGTC": 3451, + "GTTGTG": 3452, + "GTTGCA": 3453, + "GTTGCT": 3454, + "GTTGCC": 3455, + "GTTGCG": 3456, + "GTTGGA": 3457, + "GTTGGT": 3458, + "GTTGGC": 3459, + "GTTGGG": 3460, + "GTCAAA": 3461, + "GTCAAT": 3462, + "GTCAAC": 3463, + "GTCAAG": 3464, + "GTCATA": 3465, + "GTCATT": 3466, + "GTCATC": 3467, + "GTCATG": 3468, + "GTCACA": 3469, + "GTCACT": 3470, + "GTCACC": 3471, + "GTCACG": 3472, + "GTCAGA": 3473, + "GTCAGT": 3474, + "GTCAGC": 3475, + "GTCAGG": 3476, + "GTCTAA": 3477, + "GTCTAT": 3478, + "GTCTAC": 3479, + "GTCTAG": 3480, + "GTCTTA": 3481, + "GTCTTT": 3482, + "GTCTTC": 3483, + "GTCTTG": 3484, + "GTCTCA": 3485, + "GTCTCT": 3486, + "GTCTCC": 3487, + "GTCTCG": 3488, + "GTCTGA": 3489, + "GTCTGT": 3490, + "GTCTGC": 3491, + "GTCTGG": 3492, + "GTCCAA": 3493, + "GTCCAT": 3494, + "GTCCAC": 3495, + "GTCCAG": 3496, + "GTCCTA": 3497, + "GTCCTT": 3498, + "GTCCTC": 3499, + "GTCCTG": 3500, + "GTCCCA": 3501, + "GTCCCT": 3502, + "GTCCCC": 3503, + "GTCCCG": 3504, + "GTCCGA": 3505, + "GTCCGT": 3506, + "GTCCGC": 3507, + "GTCCGG": 3508, + "GTCGAA": 3509, + "GTCGAT": 3510, + "GTCGAC": 3511, + "GTCGAG": 3512, + "GTCGTA": 3513, + "GTCGTT": 3514, + "GTCGTC": 3515, + "GTCGTG": 3516, + "GTCGCA": 3517, + "GTCGCT": 3518, + "GTCGCC": 3519, + "GTCGCG": 3520, + "GTCGGA": 3521, + "GTCGGT": 3522, + "GTCGGC": 3523, + "GTCGGG": 3524, + "GTGAAA": 3525, + "GTGAAT": 3526, + "GTGAAC": 3527, + "GTGAAG": 3528, + "GTGATA": 3529, + "GTGATT": 3530, + "GTGATC": 3531, + "GTGATG": 3532, + "GTGACA": 3533, + "GTGACT": 3534, + "GTGACC": 3535, + "GTGACG": 3536, + "GTGAGA": 3537, + "GTGAGT": 3538, + "GTGAGC": 3539, + "GTGAGG": 3540, + "GTGTAA": 3541, + "GTGTAT": 3542, + "GTGTAC": 3543, + "GTGTAG": 3544, + "GTGTTA": 3545, + "GTGTTT": 3546, + "GTGTTC": 3547, + "GTGTTG": 3548, + "GTGTCA": 3549, + "GTGTCT": 3550, + "GTGTCC": 3551, + "GTGTCG": 3552, + "GTGTGA": 3553, + "GTGTGT": 3554, + "GTGTGC": 3555, + "GTGTGG": 3556, + "GTGCAA": 3557, + "GTGCAT": 3558, + "GTGCAC": 3559, + "GTGCAG": 3560, + "GTGCTA": 3561, + "GTGCTT": 3562, + "GTGCTC": 3563, + "GTGCTG": 3564, + "GTGCCA": 3565, + "GTGCCT": 3566, + "GTGCCC": 3567, + "GTGCCG": 3568, + "GTGCGA": 3569, + "GTGCGT": 3570, + "GTGCGC": 3571, + "GTGCGG": 3572, + "GTGGAA": 3573, + "GTGGAT": 3574, + "GTGGAC": 3575, + "GTGGAG": 3576, + "GTGGTA": 3577, + "GTGGTT": 3578, + "GTGGTC": 3579, + "GTGGTG": 3580, + "GTGGCA": 3581, + "GTGGCT": 3582, + "GTGGCC": 3583, + "GTGGCG": 3584, + "GTGGGA": 3585, + "GTGGGT": 3586, + "GTGGGC": 3587, + "GTGGGG": 3588, + "GCAAAA": 3589, + "GCAAAT": 3590, + "GCAAAC": 3591, + "GCAAAG": 3592, + "GCAATA": 3593, + "GCAATT": 3594, + "GCAATC": 3595, + "GCAATG": 3596, + "GCAACA": 3597, + "GCAACT": 3598, + "GCAACC": 3599, + "GCAACG": 3600, + "GCAAGA": 3601, + "GCAAGT": 3602, + "GCAAGC": 3603, + "GCAAGG": 3604, + "GCATAA": 3605, + "GCATAT": 3606, + "GCATAC": 3607, + "GCATAG": 3608, + "GCATTA": 3609, + "GCATTT": 3610, + "GCATTC": 3611, + "GCATTG": 3612, + "GCATCA": 3613, + "GCATCT": 3614, + "GCATCC": 3615, + "GCATCG": 3616, + "GCATGA": 3617, + "GCATGT": 3618, + "GCATGC": 3619, + "GCATGG": 3620, + "GCACAA": 3621, + "GCACAT": 3622, + "GCACAC": 3623, + "GCACAG": 3624, + "GCACTA": 3625, + "GCACTT": 3626, + "GCACTC": 3627, + "GCACTG": 3628, + "GCACCA": 3629, + "GCACCT": 3630, + "GCACCC": 3631, + "GCACCG": 3632, + "GCACGA": 3633, + "GCACGT": 3634, + "GCACGC": 3635, + "GCACGG": 3636, + "GCAGAA": 3637, + "GCAGAT": 3638, + "GCAGAC": 3639, + "GCAGAG": 3640, + "GCAGTA": 3641, + "GCAGTT": 3642, + "GCAGTC": 3643, + "GCAGTG": 3644, + "GCAGCA": 3645, + "GCAGCT": 3646, + "GCAGCC": 3647, + "GCAGCG": 3648, + "GCAGGA": 3649, + "GCAGGT": 3650, + "GCAGGC": 3651, + "GCAGGG": 3652, + "GCTAAA": 3653, + "GCTAAT": 3654, + "GCTAAC": 3655, + "GCTAAG": 3656, + "GCTATA": 3657, + "GCTATT": 3658, + "GCTATC": 3659, + "GCTATG": 3660, + "GCTACA": 3661, + "GCTACT": 3662, + "GCTACC": 3663, + "GCTACG": 3664, + "GCTAGA": 3665, + "GCTAGT": 3666, + "GCTAGC": 3667, + "GCTAGG": 3668, + "GCTTAA": 3669, + "GCTTAT": 3670, + "GCTTAC": 3671, + "GCTTAG": 3672, + "GCTTTA": 3673, + "GCTTTT": 3674, + "GCTTTC": 3675, + "GCTTTG": 3676, + "GCTTCA": 3677, + "GCTTCT": 3678, + "GCTTCC": 3679, + "GCTTCG": 3680, + "GCTTGA": 3681, + "GCTTGT": 3682, + "GCTTGC": 3683, + "GCTTGG": 3684, + "GCTCAA": 3685, + "GCTCAT": 3686, + "GCTCAC": 3687, + "GCTCAG": 3688, + "GCTCTA": 3689, + "GCTCTT": 3690, + "GCTCTC": 3691, + "GCTCTG": 3692, + "GCTCCA": 3693, + "GCTCCT": 3694, + "GCTCCC": 3695, + "GCTCCG": 3696, + "GCTCGA": 3697, + "GCTCGT": 3698, + "GCTCGC": 3699, + "GCTCGG": 3700, + "GCTGAA": 3701, + "GCTGAT": 3702, + "GCTGAC": 3703, + "GCTGAG": 3704, + "GCTGTA": 3705, + "GCTGTT": 3706, + "GCTGTC": 3707, + "GCTGTG": 3708, + "GCTGCA": 3709, + "GCTGCT": 3710, + "GCTGCC": 3711, + "GCTGCG": 3712, + "GCTGGA": 3713, + "GCTGGT": 3714, + "GCTGGC": 3715, + "GCTGGG": 3716, + "GCCAAA": 3717, + "GCCAAT": 3718, + "GCCAAC": 3719, + "GCCAAG": 3720, + "GCCATA": 3721, + "GCCATT": 3722, + "GCCATC": 3723, + "GCCATG": 3724, + "GCCACA": 3725, + "GCCACT": 3726, + "GCCACC": 3727, + "GCCACG": 3728, + "GCCAGA": 3729, + "GCCAGT": 3730, + "GCCAGC": 3731, + "GCCAGG": 3732, + "GCCTAA": 3733, + "GCCTAT": 3734, + "GCCTAC": 3735, + "GCCTAG": 3736, + "GCCTTA": 3737, + "GCCTTT": 3738, + "GCCTTC": 3739, + "GCCTTG": 3740, + "GCCTCA": 3741, + "GCCTCT": 3742, + "GCCTCC": 3743, + "GCCTCG": 3744, + "GCCTGA": 3745, + "GCCTGT": 3746, + "GCCTGC": 3747, + "GCCTGG": 3748, + "GCCCAA": 3749, + "GCCCAT": 3750, + "GCCCAC": 3751, + "GCCCAG": 3752, + "GCCCTA": 3753, + "GCCCTT": 3754, + "GCCCTC": 3755, + "GCCCTG": 3756, + "GCCCCA": 3757, + "GCCCCT": 3758, + "GCCCCC": 3759, + "GCCCCG": 3760, + "GCCCGA": 3761, + "GCCCGT": 3762, + "GCCCGC": 3763, + "GCCCGG": 3764, + "GCCGAA": 3765, + "GCCGAT": 3766, + "GCCGAC": 3767, + "GCCGAG": 3768, + "GCCGTA": 3769, + "GCCGTT": 3770, + "GCCGTC": 3771, + "GCCGTG": 3772, + "GCCGCA": 3773, + "GCCGCT": 3774, + "GCCGCC": 3775, + "GCCGCG": 3776, + "GCCGGA": 3777, + "GCCGGT": 3778, + "GCCGGC": 3779, + "GCCGGG": 3780, + "GCGAAA": 3781, + "GCGAAT": 3782, + "GCGAAC": 3783, + "GCGAAG": 3784, + "GCGATA": 3785, + "GCGATT": 3786, + "GCGATC": 3787, + "GCGATG": 3788, + "GCGACA": 3789, + "GCGACT": 3790, + "GCGACC": 3791, + "GCGACG": 3792, + "GCGAGA": 3793, + "GCGAGT": 3794, + "GCGAGC": 3795, + "GCGAGG": 3796, + "GCGTAA": 3797, + "GCGTAT": 3798, + "GCGTAC": 3799, + "GCGTAG": 3800, + "GCGTTA": 3801, + "GCGTTT": 3802, + "GCGTTC": 3803, + "GCGTTG": 3804, + "GCGTCA": 3805, + "GCGTCT": 3806, + "GCGTCC": 3807, + "GCGTCG": 3808, + "GCGTGA": 3809, + "GCGTGT": 3810, + "GCGTGC": 3811, + "GCGTGG": 3812, + "GCGCAA": 3813, + "GCGCAT": 3814, + "GCGCAC": 3815, + "GCGCAG": 3816, + "GCGCTA": 3817, + "GCGCTT": 3818, + "GCGCTC": 3819, + "GCGCTG": 3820, + "GCGCCA": 3821, + "GCGCCT": 3822, + "GCGCCC": 3823, + "GCGCCG": 3824, + "GCGCGA": 3825, + "GCGCGT": 3826, + "GCGCGC": 3827, + "GCGCGG": 3828, + "GCGGAA": 3829, + "GCGGAT": 3830, + "GCGGAC": 3831, + "GCGGAG": 3832, + "GCGGTA": 3833, + "GCGGTT": 3834, + "GCGGTC": 3835, + "GCGGTG": 3836, + "GCGGCA": 3837, + "GCGGCT": 3838, + "GCGGCC": 3839, + "GCGGCG": 3840, + "GCGGGA": 3841, + "GCGGGT": 3842, + "GCGGGC": 3843, + "GCGGGG": 3844, + "GGAAAA": 3845, + "GGAAAT": 3846, + "GGAAAC": 3847, + "GGAAAG": 3848, + "GGAATA": 3849, + "GGAATT": 3850, + "GGAATC": 3851, + "GGAATG": 3852, + "GGAACA": 3853, + "GGAACT": 3854, + "GGAACC": 3855, + "GGAACG": 3856, + "GGAAGA": 3857, + "GGAAGT": 3858, + "GGAAGC": 3859, + "GGAAGG": 3860, + "GGATAA": 3861, + "GGATAT": 3862, + "GGATAC": 3863, + "GGATAG": 3864, + "GGATTA": 3865, + "GGATTT": 3866, + "GGATTC": 3867, + "GGATTG": 3868, + "GGATCA": 3869, + "GGATCT": 3870, + "GGATCC": 3871, + "GGATCG": 3872, + "GGATGA": 3873, + "GGATGT": 3874, + "GGATGC": 3875, + "GGATGG": 3876, + "GGACAA": 3877, + "GGACAT": 3878, + "GGACAC": 3879, + "GGACAG": 3880, + "GGACTA": 3881, + "GGACTT": 3882, + "GGACTC": 3883, + "GGACTG": 3884, + "GGACCA": 3885, + "GGACCT": 3886, + "GGACCC": 3887, + "GGACCG": 3888, + "GGACGA": 3889, + "GGACGT": 3890, + "GGACGC": 3891, + "GGACGG": 3892, + "GGAGAA": 3893, + "GGAGAT": 3894, + "GGAGAC": 3895, + "GGAGAG": 3896, + "GGAGTA": 3897, + "GGAGTT": 3898, + "GGAGTC": 3899, + "GGAGTG": 3900, + "GGAGCA": 3901, + "GGAGCT": 3902, + "GGAGCC": 3903, + "GGAGCG": 3904, + "GGAGGA": 3905, + "GGAGGT": 3906, + "GGAGGC": 3907, + "GGAGGG": 3908, + "GGTAAA": 3909, + "GGTAAT": 3910, + "GGTAAC": 3911, + "GGTAAG": 3912, + "GGTATA": 3913, + "GGTATT": 3914, + "GGTATC": 3915, + "GGTATG": 3916, + "GGTACA": 3917, + "GGTACT": 3918, + "GGTACC": 3919, + "GGTACG": 3920, + "GGTAGA": 3921, + "GGTAGT": 3922, + "GGTAGC": 3923, + "GGTAGG": 3924, + "GGTTAA": 3925, + "GGTTAT": 3926, + "GGTTAC": 3927, + "GGTTAG": 3928, + "GGTTTA": 3929, + "GGTTTT": 3930, + "GGTTTC": 3931, + "GGTTTG": 3932, + "GGTTCA": 3933, + "GGTTCT": 3934, + "GGTTCC": 3935, + "GGTTCG": 3936, + "GGTTGA": 3937, + "GGTTGT": 3938, + "GGTTGC": 3939, + "GGTTGG": 3940, + "GGTCAA": 3941, + "GGTCAT": 3942, + "GGTCAC": 3943, + "GGTCAG": 3944, + "GGTCTA": 3945, + "GGTCTT": 3946, + "GGTCTC": 3947, + "GGTCTG": 3948, + "GGTCCA": 3949, + "GGTCCT": 3950, + "GGTCCC": 3951, + "GGTCCG": 3952, + "GGTCGA": 3953, + "GGTCGT": 3954, + "GGTCGC": 3955, + "GGTCGG": 3956, + "GGTGAA": 3957, + "GGTGAT": 3958, + "GGTGAC": 3959, + "GGTGAG": 3960, + "GGTGTA": 3961, + "GGTGTT": 3962, + "GGTGTC": 3963, + "GGTGTG": 3964, + "GGTGCA": 3965, + "GGTGCT": 3966, + "GGTGCC": 3967, + "GGTGCG": 3968, + "GGTGGA": 3969, + "GGTGGT": 3970, + "GGTGGC": 3971, + "GGTGGG": 3972, + "GGCAAA": 3973, + "GGCAAT": 3974, + "GGCAAC": 3975, + "GGCAAG": 3976, + "GGCATA": 3977, + "GGCATT": 3978, + "GGCATC": 3979, + "GGCATG": 3980, + "GGCACA": 3981, + "GGCACT": 3982, + "GGCACC": 3983, + "GGCACG": 3984, + "GGCAGA": 3985, + "GGCAGT": 3986, + "GGCAGC": 3987, + "GGCAGG": 3988, + "GGCTAA": 3989, + "GGCTAT": 3990, + "GGCTAC": 3991, + "GGCTAG": 3992, + "GGCTTA": 3993, + "GGCTTT": 3994, + "GGCTTC": 3995, + "GGCTTG": 3996, + "GGCTCA": 3997, + "GGCTCT": 3998, + "GGCTCC": 3999, + "GGCTCG": 4000, + "GGCTGA": 4001, + "GGCTGT": 4002, + "GGCTGC": 4003, + "GGCTGG": 4004, + "GGCCAA": 4005, + "GGCCAT": 4006, + "GGCCAC": 4007, + "GGCCAG": 4008, + "GGCCTA": 4009, + "GGCCTT": 4010, + "GGCCTC": 4011, + "GGCCTG": 4012, + "GGCCCA": 4013, + "GGCCCT": 4014, + "GGCCCC": 4015, + "GGCCCG": 4016, + "GGCCGA": 4017, + "GGCCGT": 4018, + "GGCCGC": 4019, + "GGCCGG": 4020, + "GGCGAA": 4021, + "GGCGAT": 4022, + "GGCGAC": 4023, + "GGCGAG": 4024, + "GGCGTA": 4025, + "GGCGTT": 4026, + "GGCGTC": 4027, + "GGCGTG": 4028, + "GGCGCA": 4029, + "GGCGCT": 4030, + "GGCGCC": 4031, + "GGCGCG": 4032, + "GGCGGA": 4033, + "GGCGGT": 4034, + "GGCGGC": 4035, + "GGCGGG": 4036, + "GGGAAA": 4037, + "GGGAAT": 4038, + "GGGAAC": 4039, + "GGGAAG": 4040, + "GGGATA": 4041, + "GGGATT": 4042, + "GGGATC": 4043, + "GGGATG": 4044, + "GGGACA": 4045, + "GGGACT": 4046, + "GGGACC": 4047, + "GGGACG": 4048, + "GGGAGA": 4049, + "GGGAGT": 4050, + "GGGAGC": 4051, + "GGGAGG": 4052, + "GGGTAA": 4053, + "GGGTAT": 4054, + "GGGTAC": 4055, + "GGGTAG": 4056, + "GGGTTA": 4057, + "GGGTTT": 4058, + "GGGTTC": 4059, + "GGGTTG": 4060, + "GGGTCA": 4061, + "GGGTCT": 4062, + "GGGTCC": 4063, + "GGGTCG": 4064, + "GGGTGA": 4065, + "GGGTGT": 4066, + "GGGTGC": 4067, + "GGGTGG": 4068, + "GGGCAA": 4069, + "GGGCAT": 4070, + "GGGCAC": 4071, + "GGGCAG": 4072, + "GGGCTA": 4073, + "GGGCTT": 4074, + "GGGCTC": 4075, + "GGGCTG": 4076, + "GGGCCA": 4077, + "GGGCCT": 4078, + "GGGCCC": 4079, + "GGGCCG": 4080, + "GGGCGA": 4081, + "GGGCGT": 4082, + "GGGCGC": 4083, + "GGGCGG": 4084, + "GGGGAA": 4085, + "GGGGAT": 4086, + "GGGGAC": 4087, + "GGGGAG": 4088, + "GGGGTA": 4089, + "GGGGTT": 4090, + "GGGGTC": 4091, + "GGGGTG": 4092, + "GGGGCA": 4093, + "GGGGCT": 4094, + "GGGGCC": 4095, + "GGGGCG": 4096, + "GGGGGA": 4097, + "GGGGGT": 4098, + "GGGGGC": 4099, + "GGGGGG": 4100 + } + } +} \ No newline at end of file diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6250/tokenizer_config.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6250/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..604e2b1b29af1dcb3920d2828b6a90276e5e3d02 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6250/tokenizer_config.json @@ -0,0 +1,58 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "[UNK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "[CLS]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "[SEP]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "4": { + "content": "[MASK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "clean_up_tokenization_spaces": true, + "cls_token": "[CLS]", + "do_basic_tokenize": true, + "do_lower_case": false, + "mask_token": "[MASK]", + "max_len": 512, + "model_max_length": 512, + "never_split": null, + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "strip_accents": null, + "tokenize_chinese_chars": true, + "tokenizer_class": "BertTokenizer", + "unk_token": "[UNK]" +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6250/trainer_state.json b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6250/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..06997144f8397bbd3a779f0cf4e5537a49f4a68e --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6250/trainer_state.json @@ -0,0 +1,1783 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 25.0, + "eval_steps": 500, + "global_step": 6250, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.1, + "grad_norm": 0.6611918807029724, + "learning_rate": 5.319148936170213e-05, + "loss": 0.6932, + "step": 25 + }, + { + "epoch": 0.2, + "grad_norm": 0.9467485547065735, + "learning_rate": 0.00010638297872340425, + "loss": 0.6912, + "step": 50 + }, + { + "epoch": 0.3, + "grad_norm": 0.9465051889419556, + "learning_rate": 0.00015957446808510637, + "loss": 0.6893, + "step": 75 + }, + { + "epoch": 0.4, + "grad_norm": 1.5430934429168701, + "learning_rate": 0.0002127659574468085, + "loss": 0.6804, + "step": 100 + }, + { + "epoch": 0.5, + "grad_norm": 1.6103107929229736, + "learning_rate": 0.00026595744680851064, + "loss": 0.6806, + "step": 125 + }, + { + "epoch": 0.6, + "grad_norm": 2.3773200511932373, + "learning_rate": 0.00031914893617021275, + "loss": 0.6801, + "step": 150 + }, + { + "epoch": 0.7, + "grad_norm": 1.8832203149795532, + "learning_rate": 0.0003723404255319149, + "loss": 0.6791, + "step": 175 + }, + { + "epoch": 0.8, + "grad_norm": 1.3350876569747925, + "learning_rate": 0.0003992081821181128, + "loss": 0.6746, + "step": 200 + }, + { + "epoch": 0.9, + "grad_norm": 1.5796219110488892, + "learning_rate": 0.0003975585615308479, + "loss": 0.6771, + "step": 225 + }, + { + "epoch": 1.0, + "grad_norm": 1.884006381034851, + "learning_rate": 0.00039590894094358297, + "loss": 0.6649, + "step": 250 + }, + { + "epoch": 1.1, + "grad_norm": 3.286440372467041, + "learning_rate": 0.0003942593203563181, + "loss": 0.6388, + "step": 275 + }, + { + "epoch": 1.2, + "grad_norm": 8.244946479797363, + "learning_rate": 0.0003926096997690532, + "loss": 0.6238, + "step": 300 + }, + { + "epoch": 1.3, + "grad_norm": 4.265683650970459, + "learning_rate": 0.00039096007918178817, + "loss": 0.6383, + "step": 325 + }, + { + "epoch": 1.4, + "grad_norm": 4.115826606750488, + "learning_rate": 0.0003893104585945233, + "loss": 0.6194, + "step": 350 + }, + { + "epoch": 1.5, + "grad_norm": 5.694250583648682, + "learning_rate": 0.0003876608380072583, + "loss": 0.6324, + "step": 375 + }, + { + "epoch": 1.6, + "grad_norm": 3.463121175765991, + "learning_rate": 0.00038601121741999343, + "loss": 0.621, + "step": 400 + }, + { + "epoch": 1.7, + "grad_norm": 4.582865238189697, + "learning_rate": 0.0003843615968327285, + "loss": 0.6116, + "step": 425 + }, + { + "epoch": 1.8, + "grad_norm": 11.996281623840332, + "learning_rate": 0.0003827119762454636, + "loss": 0.6393, + "step": 450 + }, + { + "epoch": 1.9, + "grad_norm": 3.0407373905181885, + "learning_rate": 0.00038106235565819863, + "loss": 0.628, + "step": 475 + }, + { + "epoch": 2.0, + "grad_norm": 2.917588233947754, + "learning_rate": 0.0003794127350709337, + "loss": 0.6078, + "step": 500 + }, + { + "epoch": 2.1, + "grad_norm": 4.748379707336426, + "learning_rate": 0.0003777631144836688, + "loss": 0.4899, + "step": 525 + }, + { + "epoch": 2.2, + "grad_norm": 3.8076977729797363, + "learning_rate": 0.00037611349389640383, + "loss": 0.5086, + "step": 550 + }, + { + "epoch": 2.3, + "grad_norm": 5.2440714836120605, + "learning_rate": 0.00037446387330913894, + "loss": 0.5327, + "step": 575 + }, + { + "epoch": 2.4, + "grad_norm": 7.110438346862793, + "learning_rate": 0.000372814252721874, + "loss": 0.5436, + "step": 600 + }, + { + "epoch": 2.5, + "grad_norm": 5.46150541305542, + "learning_rate": 0.00037116463213460903, + "loss": 0.5294, + "step": 625 + }, + { + "epoch": 2.6, + "grad_norm": 5.136163234710693, + "learning_rate": 0.00036951501154734414, + "loss": 0.5245, + "step": 650 + }, + { + "epoch": 2.7, + "grad_norm": 8.735346794128418, + "learning_rate": 0.0003678653909600792, + "loss": 0.5449, + "step": 675 + }, + { + "epoch": 2.8, + "grad_norm": 2.922825574874878, + "learning_rate": 0.0003662157703728143, + "loss": 0.5406, + "step": 700 + }, + { + "epoch": 2.9, + "grad_norm": 7.744819641113281, + "learning_rate": 0.00036456614978554934, + "loss": 0.5447, + "step": 725 + }, + { + "epoch": 3.0, + "grad_norm": 11.6185884475708, + "learning_rate": 0.00036291652919828444, + "loss": 0.5195, + "step": 750 + }, + { + "epoch": 3.1, + "grad_norm": 3.73836088180542, + "learning_rate": 0.00036126690861101944, + "loss": 0.3824, + "step": 775 + }, + { + "epoch": 3.2, + "grad_norm": 14.850343704223633, + "learning_rate": 0.00035961728802375454, + "loss": 0.4071, + "step": 800 + }, + { + "epoch": 3.3, + "grad_norm": 5.7157440185546875, + "learning_rate": 0.0003579676674364896, + "loss": 0.3986, + "step": 825 + }, + { + "epoch": 3.4, + "grad_norm": 12.418399810791016, + "learning_rate": 0.0003563180468492247, + "loss": 0.4282, + "step": 850 + }, + { + "epoch": 3.5, + "grad_norm": 12.793001174926758, + "learning_rate": 0.0003546684262619598, + "loss": 0.4822, + "step": 875 + }, + { + "epoch": 3.6, + "grad_norm": 6.489450931549072, + "learning_rate": 0.00035301880567469485, + "loss": 0.4239, + "step": 900 + }, + { + "epoch": 3.7, + "grad_norm": 5.365822792053223, + "learning_rate": 0.0003513691850874299, + "loss": 0.421, + "step": 925 + }, + { + "epoch": 3.8, + "grad_norm": 12.643745422363281, + "learning_rate": 0.00034971956450016495, + "loss": 0.3964, + "step": 950 + }, + { + "epoch": 3.9, + "grad_norm": 14.334024429321289, + "learning_rate": 0.00034806994391290005, + "loss": 0.4634, + "step": 975 + }, + { + "epoch": 4.0, + "grad_norm": 6.819091320037842, + "learning_rate": 0.0003464203233256351, + "loss": 0.4139, + "step": 1000 + }, + { + "epoch": 4.1, + "grad_norm": 8.29238224029541, + "learning_rate": 0.0003447707027383702, + "loss": 0.2695, + "step": 1025 + }, + { + "epoch": 4.2, + "grad_norm": 5.984206676483154, + "learning_rate": 0.00034312108215110525, + "loss": 0.2653, + "step": 1050 + }, + { + "epoch": 4.3, + "grad_norm": 5.9425435066223145, + "learning_rate": 0.0003414714615638403, + "loss": 0.2982, + "step": 1075 + }, + { + "epoch": 4.4, + "grad_norm": 7.1877593994140625, + "learning_rate": 0.0003398218409765754, + "loss": 0.3309, + "step": 1100 + }, + { + "epoch": 4.5, + "grad_norm": 17.83046531677246, + "learning_rate": 0.00033817222038931045, + "loss": 0.3467, + "step": 1125 + }, + { + "epoch": 4.6, + "grad_norm": 4.865128517150879, + "learning_rate": 0.00033652259980204556, + "loss": 0.3117, + "step": 1150 + }, + { + "epoch": 4.7, + "grad_norm": 31.427154541015625, + "learning_rate": 0.0003348729792147806, + "loss": 0.3273, + "step": 1175 + }, + { + "epoch": 4.8, + "grad_norm": 26.77984619140625, + "learning_rate": 0.0003332233586275157, + "loss": 0.3504, + "step": 1200 + }, + { + "epoch": 4.9, + "grad_norm": 9.389993667602539, + "learning_rate": 0.00033157373804025076, + "loss": 0.3339, + "step": 1225 + }, + { + "epoch": 5.0, + "grad_norm": 25.70441246032715, + "learning_rate": 0.0003299241174529858, + "loss": 0.3336, + "step": 1250 + }, + { + "epoch": 5.1, + "grad_norm": 13.482085227966309, + "learning_rate": 0.0003282744968657209, + "loss": 0.2156, + "step": 1275 + }, + { + "epoch": 5.2, + "grad_norm": 8.725810050964355, + "learning_rate": 0.00032662487627845596, + "loss": 0.1811, + "step": 1300 + }, + { + "epoch": 5.3, + "grad_norm": 9.313215255737305, + "learning_rate": 0.00032497525569119106, + "loss": 0.2807, + "step": 1325 + }, + { + "epoch": 5.4, + "grad_norm": 11.026411056518555, + "learning_rate": 0.0003233256351039261, + "loss": 0.2757, + "step": 1350 + }, + { + "epoch": 5.5, + "grad_norm": 11.038985252380371, + "learning_rate": 0.00032167601451666116, + "loss": 0.2177, + "step": 1375 + }, + { + "epoch": 5.6, + "grad_norm": 4.008651256561279, + "learning_rate": 0.00032002639392939627, + "loss": 0.2163, + "step": 1400 + }, + { + "epoch": 5.7, + "grad_norm": 12.480770111083984, + "learning_rate": 0.0003183767733421313, + "loss": 0.2173, + "step": 1425 + }, + { + "epoch": 5.8, + "grad_norm": 8.751969337463379, + "learning_rate": 0.0003167271527548664, + "loss": 0.2299, + "step": 1450 + }, + { + "epoch": 5.9, + "grad_norm": 7.701971530914307, + "learning_rate": 0.00031507753216760147, + "loss": 0.1949, + "step": 1475 + }, + { + "epoch": 6.0, + "grad_norm": 8.48027515411377, + "learning_rate": 0.00031342791158033657, + "loss": 0.2599, + "step": 1500 + }, + { + "epoch": 6.1, + "grad_norm": 9.29404067993164, + "learning_rate": 0.00031177829099307157, + "loss": 0.1376, + "step": 1525 + }, + { + "epoch": 6.2, + "grad_norm": 20.137714385986328, + "learning_rate": 0.00031012867040580667, + "loss": 0.1647, + "step": 1550 + }, + { + "epoch": 6.3, + "grad_norm": 11.394575119018555, + "learning_rate": 0.0003084790498185417, + "loss": 0.1565, + "step": 1575 + }, + { + "epoch": 6.4, + "grad_norm": 8.214287757873535, + "learning_rate": 0.0003068294292312768, + "loss": 0.1739, + "step": 1600 + }, + { + "epoch": 6.5, + "grad_norm": 7.779988765716553, + "learning_rate": 0.0003051798086440119, + "loss": 0.1403, + "step": 1625 + }, + { + "epoch": 6.6, + "grad_norm": 9.421648025512695, + "learning_rate": 0.000303530188056747, + "loss": 0.1545, + "step": 1650 + }, + { + "epoch": 6.7, + "grad_norm": 5.751734256744385, + "learning_rate": 0.000301880567469482, + "loss": 0.1971, + "step": 1675 + }, + { + "epoch": 6.8, + "grad_norm": 23.861705780029297, + "learning_rate": 0.0003002309468822171, + "loss": 0.1681, + "step": 1700 + }, + { + "epoch": 6.9, + "grad_norm": 18.944721221923828, + "learning_rate": 0.0002985813262949522, + "loss": 0.1703, + "step": 1725 + }, + { + "epoch": 7.0, + "grad_norm": 14.045795440673828, + "learning_rate": 0.00029693170570768723, + "loss": 0.1801, + "step": 1750 + }, + { + "epoch": 7.1, + "grad_norm": 6.4620137214660645, + "learning_rate": 0.00029528208512042233, + "loss": 0.1253, + "step": 1775 + }, + { + "epoch": 7.2, + "grad_norm": 4.318169593811035, + "learning_rate": 0.0002936324645331574, + "loss": 0.1397, + "step": 1800 + }, + { + "epoch": 7.3, + "grad_norm": 24.91462516784668, + "learning_rate": 0.00029198284394589243, + "loss": 0.1259, + "step": 1825 + }, + { + "epoch": 7.4, + "grad_norm": 23.614572525024414, + "learning_rate": 0.00029033322335862753, + "loss": 0.1293, + "step": 1850 + }, + { + "epoch": 7.5, + "grad_norm": 3.60048508644104, + "learning_rate": 0.0002886836027713626, + "loss": 0.138, + "step": 1875 + }, + { + "epoch": 7.6, + "grad_norm": 16.62705421447754, + "learning_rate": 0.0002870339821840977, + "loss": 0.1733, + "step": 1900 + }, + { + "epoch": 7.7, + "grad_norm": 33.79671859741211, + "learning_rate": 0.00028538436159683273, + "loss": 0.217, + "step": 1925 + }, + { + "epoch": 7.8, + "grad_norm": 9.69206428527832, + "learning_rate": 0.00028373474100956784, + "loss": 0.1289, + "step": 1950 + }, + { + "epoch": 7.9, + "grad_norm": 18.655046463012695, + "learning_rate": 0.0002820851204223029, + "loss": 0.1168, + "step": 1975 + }, + { + "epoch": 8.0, + "grad_norm": 8.4110746383667, + "learning_rate": 0.00028043549983503794, + "loss": 0.1471, + "step": 2000 + }, + { + "epoch": 8.1, + "grad_norm": 14.473428726196289, + "learning_rate": 0.00027878587924777304, + "loss": 0.0774, + "step": 2025 + }, + { + "epoch": 8.2, + "grad_norm": 1.2464979887008667, + "learning_rate": 0.0002771362586605081, + "loss": 0.0715, + "step": 2050 + }, + { + "epoch": 8.3, + "grad_norm": 26.74981689453125, + "learning_rate": 0.0002754866380732432, + "loss": 0.1289, + "step": 2075 + }, + { + "epoch": 8.4, + "grad_norm": 13.958703994750977, + "learning_rate": 0.00027383701748597824, + "loss": 0.1294, + "step": 2100 + }, + { + "epoch": 8.5, + "grad_norm": 10.394835472106934, + "learning_rate": 0.0002721873968987133, + "loss": 0.0987, + "step": 2125 + }, + { + "epoch": 8.6, + "grad_norm": 9.387774467468262, + "learning_rate": 0.00027053777631144834, + "loss": 0.1269, + "step": 2150 + }, + { + "epoch": 8.7, + "grad_norm": 4.988718509674072, + "learning_rate": 0.00026888815572418344, + "loss": 0.1374, + "step": 2175 + }, + { + "epoch": 8.8, + "grad_norm": 12.66832447052002, + "learning_rate": 0.00026723853513691855, + "loss": 0.0997, + "step": 2200 + }, + { + "epoch": 8.9, + "grad_norm": 19.10486602783203, + "learning_rate": 0.0002655889145496536, + "loss": 0.1325, + "step": 2225 + }, + { + "epoch": 9.0, + "grad_norm": 3.2943971157073975, + "learning_rate": 0.0002639392939623887, + "loss": 0.1385, + "step": 2250 + }, + { + "epoch": 9.1, + "grad_norm": 11.490874290466309, + "learning_rate": 0.0002622896733751237, + "loss": 0.0573, + "step": 2275 + }, + { + "epoch": 9.2, + "grad_norm": 11.961288452148438, + "learning_rate": 0.0002606400527878588, + "loss": 0.1128, + "step": 2300 + }, + { + "epoch": 9.3, + "grad_norm": 0.7256277203559875, + "learning_rate": 0.00025899043220059385, + "loss": 0.0673, + "step": 2325 + }, + { + "epoch": 9.4, + "grad_norm": 9.857927322387695, + "learning_rate": 0.00025734081161332895, + "loss": 0.2072, + "step": 2350 + }, + { + "epoch": 9.5, + "grad_norm": 34.927734375, + "learning_rate": 0.000255691191026064, + "loss": 0.221, + "step": 2375 + }, + { + "epoch": 9.6, + "grad_norm": 28.461849212646484, + "learning_rate": 0.0002540415704387991, + "loss": 0.1263, + "step": 2400 + }, + { + "epoch": 9.7, + "grad_norm": 9.591217041015625, + "learning_rate": 0.00025239194985153415, + "loss": 0.099, + "step": 2425 + }, + { + "epoch": 9.8, + "grad_norm": 9.780253410339355, + "learning_rate": 0.0002507423292642692, + "loss": 0.0754, + "step": 2450 + }, + { + "epoch": 9.9, + "grad_norm": 14.394485473632812, + "learning_rate": 0.0002490927086770043, + "loss": 0.0909, + "step": 2475 + }, + { + "epoch": 10.0, + "grad_norm": 4.217277526855469, + "learning_rate": 0.00024744308808973936, + "loss": 0.1495, + "step": 2500 + }, + { + "epoch": 10.1, + "grad_norm": 0.08438724279403687, + "learning_rate": 0.00024579346750247446, + "loss": 0.0662, + "step": 2525 + }, + { + "epoch": 10.2, + "grad_norm": 0.9595862030982971, + "learning_rate": 0.0002441438469152095, + "loss": 0.084, + "step": 2550 + }, + { + "epoch": 10.3, + "grad_norm": 42.32374954223633, + "learning_rate": 0.00024249422632794456, + "loss": 0.0541, + "step": 2575 + }, + { + "epoch": 10.4, + "grad_norm": 24.792821884155273, + "learning_rate": 0.00024084460574067963, + "loss": 0.121, + "step": 2600 + }, + { + "epoch": 10.5, + "grad_norm": 2.7049331665039062, + "learning_rate": 0.0002391949851534147, + "loss": 0.1551, + "step": 2625 + }, + { + "epoch": 10.6, + "grad_norm": 2.788362741470337, + "learning_rate": 0.0002375453645661498, + "loss": 0.0877, + "step": 2650 + }, + { + "epoch": 10.7, + "grad_norm": 0.4999215006828308, + "learning_rate": 0.00023589574397888486, + "loss": 0.1263, + "step": 2675 + }, + { + "epoch": 10.8, + "grad_norm": 16.476675033569336, + "learning_rate": 0.00023424612339161997, + "loss": 0.0883, + "step": 2700 + }, + { + "epoch": 10.9, + "grad_norm": 38.192726135253906, + "learning_rate": 0.000232596502804355, + "loss": 0.0843, + "step": 2725 + }, + { + "epoch": 11.0, + "grad_norm": 18.09952735900879, + "learning_rate": 0.00023094688221709007, + "loss": 0.0957, + "step": 2750 + }, + { + "epoch": 11.1, + "grad_norm": 34.254417419433594, + "learning_rate": 0.00022929726162982514, + "loss": 0.0856, + "step": 2775 + }, + { + "epoch": 11.2, + "grad_norm": 11.744216918945312, + "learning_rate": 0.00022764764104256022, + "loss": 0.0654, + "step": 2800 + }, + { + "epoch": 11.3, + "grad_norm": 0.1285558044910431, + "learning_rate": 0.0002259980204552953, + "loss": 0.0638, + "step": 2825 + }, + { + "epoch": 11.4, + "grad_norm": 20.57583999633789, + "learning_rate": 0.00022434839986803037, + "loss": 0.079, + "step": 2850 + }, + { + "epoch": 11.5, + "grad_norm": 0.7192332148551941, + "learning_rate": 0.00022269877928076542, + "loss": 0.0854, + "step": 2875 + }, + { + "epoch": 11.6, + "grad_norm": 31.341829299926758, + "learning_rate": 0.0002210491586935005, + "loss": 0.063, + "step": 2900 + }, + { + "epoch": 11.7, + "grad_norm": 13.460247039794922, + "learning_rate": 0.00021939953810623557, + "loss": 0.1358, + "step": 2925 + }, + { + "epoch": 11.8, + "grad_norm": 1.671036720275879, + "learning_rate": 0.00021774991751897065, + "loss": 0.0806, + "step": 2950 + }, + { + "epoch": 11.9, + "grad_norm": 22.026491165161133, + "learning_rate": 0.00021610029693170573, + "loss": 0.0912, + "step": 2975 + }, + { + "epoch": 12.0, + "grad_norm": 46.78192138671875, + "learning_rate": 0.0002144506763444408, + "loss": 0.0763, + "step": 3000 + }, + { + "epoch": 12.1, + "grad_norm": 6.397649765014648, + "learning_rate": 0.00021280105575717582, + "loss": 0.0613, + "step": 3025 + }, + { + "epoch": 12.2, + "grad_norm": 7.179838180541992, + "learning_rate": 0.00021115143516991093, + "loss": 0.0496, + "step": 3050 + }, + { + "epoch": 12.3, + "grad_norm": 0.021825680509209633, + "learning_rate": 0.000209501814582646, + "loss": 0.0633, + "step": 3075 + }, + { + "epoch": 12.4, + "grad_norm": 6.6774187088012695, + "learning_rate": 0.00020785219399538108, + "loss": 0.0765, + "step": 3100 + }, + { + "epoch": 12.5, + "grad_norm": 0.13217756152153015, + "learning_rate": 0.00020620257340811616, + "loss": 0.0696, + "step": 3125 + }, + { + "epoch": 12.6, + "grad_norm": 2.5293831825256348, + "learning_rate": 0.00020455295282085123, + "loss": 0.0419, + "step": 3150 + }, + { + "epoch": 12.7, + "grad_norm": 1.1612874269485474, + "learning_rate": 0.00020290333223358626, + "loss": 0.1046, + "step": 3175 + }, + { + "epoch": 12.8, + "grad_norm": 40.93427658081055, + "learning_rate": 0.00020125371164632133, + "loss": 0.0488, + "step": 3200 + }, + { + "epoch": 12.9, + "grad_norm": 1.446478009223938, + "learning_rate": 0.0001996040910590564, + "loss": 0.062, + "step": 3225 + }, + { + "epoch": 13.0, + "grad_norm": 28.893821716308594, + "learning_rate": 0.00019795447047179148, + "loss": 0.0913, + "step": 3250 + }, + { + "epoch": 13.1, + "grad_norm": 23.31914520263672, + "learning_rate": 0.0001963048498845266, + "loss": 0.061, + "step": 3275 + }, + { + "epoch": 13.2, + "grad_norm": 0.26242795586586, + "learning_rate": 0.00019465522929726164, + "loss": 0.0563, + "step": 3300 + }, + { + "epoch": 13.3, + "grad_norm": 0.08909507840871811, + "learning_rate": 0.00019300560870999671, + "loss": 0.0387, + "step": 3325 + }, + { + "epoch": 13.4, + "grad_norm": 24.825326919555664, + "learning_rate": 0.0001913559881227318, + "loss": 0.0776, + "step": 3350 + }, + { + "epoch": 13.5, + "grad_norm": 39.43446731567383, + "learning_rate": 0.00018970636753546684, + "loss": 0.0846, + "step": 3375 + }, + { + "epoch": 13.6, + "grad_norm": 32.98988723754883, + "learning_rate": 0.00018805674694820192, + "loss": 0.0512, + "step": 3400 + }, + { + "epoch": 13.7, + "grad_norm": 4.053821563720703, + "learning_rate": 0.000186407126360937, + "loss": 0.1576, + "step": 3425 + }, + { + "epoch": 13.8, + "grad_norm": 0.2785554528236389, + "learning_rate": 0.00018475750577367207, + "loss": 0.0728, + "step": 3450 + }, + { + "epoch": 13.9, + "grad_norm": 11.39714527130127, + "learning_rate": 0.00018310788518640715, + "loss": 0.0776, + "step": 3475 + }, + { + "epoch": 14.0, + "grad_norm": 3.7018072605133057, + "learning_rate": 0.00018145826459914222, + "loss": 0.0736, + "step": 3500 + }, + { + "epoch": 14.1, + "grad_norm": 0.26053619384765625, + "learning_rate": 0.00017980864401187727, + "loss": 0.0598, + "step": 3525 + }, + { + "epoch": 14.2, + "grad_norm": 6.454179763793945, + "learning_rate": 0.00017815902342461235, + "loss": 0.0435, + "step": 3550 + }, + { + "epoch": 14.3, + "grad_norm": 34.47985076904297, + "learning_rate": 0.00017650940283734742, + "loss": 0.0467, + "step": 3575 + }, + { + "epoch": 14.4, + "grad_norm": 11.76491928100586, + "learning_rate": 0.00017485978225008247, + "loss": 0.1298, + "step": 3600 + }, + { + "epoch": 14.5, + "grad_norm": 26.64131736755371, + "learning_rate": 0.00017321016166281755, + "loss": 0.1313, + "step": 3625 + }, + { + "epoch": 14.6, + "grad_norm": 5.510906219482422, + "learning_rate": 0.00017156054107555263, + "loss": 0.0421, + "step": 3650 + }, + { + "epoch": 14.7, + "grad_norm": 0.04721131548285484, + "learning_rate": 0.0001699109204882877, + "loss": 0.0569, + "step": 3675 + }, + { + "epoch": 14.8, + "grad_norm": 0.6616227030754089, + "learning_rate": 0.00016826129990102278, + "loss": 0.0618, + "step": 3700 + }, + { + "epoch": 14.9, + "grad_norm": 0.05287986248731613, + "learning_rate": 0.00016661167931375785, + "loss": 0.0574, + "step": 3725 + }, + { + "epoch": 15.0, + "grad_norm": 0.49707508087158203, + "learning_rate": 0.0001649620587264929, + "loss": 0.094, + "step": 3750 + }, + { + "epoch": 15.1, + "grad_norm": 23.649728775024414, + "learning_rate": 0.00016331243813922798, + "loss": 0.0512, + "step": 3775 + }, + { + "epoch": 15.2, + "grad_norm": 0.01772051490843296, + "learning_rate": 0.00016166281755196306, + "loss": 0.0719, + "step": 3800 + }, + { + "epoch": 15.3, + "grad_norm": 1.5259039402008057, + "learning_rate": 0.00016001319696469813, + "loss": 0.0615, + "step": 3825 + }, + { + "epoch": 15.4, + "grad_norm": 0.026406478136777878, + "learning_rate": 0.0001583635763774332, + "loss": 0.0651, + "step": 3850 + }, + { + "epoch": 15.5, + "grad_norm": 27.12177085876465, + "learning_rate": 0.00015671395579016829, + "loss": 0.0458, + "step": 3875 + }, + { + "epoch": 15.6, + "grad_norm": 9.838811874389648, + "learning_rate": 0.00015506433520290334, + "loss": 0.0662, + "step": 3900 + }, + { + "epoch": 15.7, + "grad_norm": 0.08516795933246613, + "learning_rate": 0.0001534147146156384, + "loss": 0.0302, + "step": 3925 + }, + { + "epoch": 15.8, + "grad_norm": 1.1224850416183472, + "learning_rate": 0.0001517650940283735, + "loss": 0.0637, + "step": 3950 + }, + { + "epoch": 15.9, + "grad_norm": 0.14810702204704285, + "learning_rate": 0.00015011547344110854, + "loss": 0.0296, + "step": 3975 + }, + { + "epoch": 16.0, + "grad_norm": 0.046288371086120605, + "learning_rate": 0.00014846585285384361, + "loss": 0.0368, + "step": 4000 + }, + { + "epoch": 16.1, + "grad_norm": 0.008764918893575668, + "learning_rate": 0.0001468162322665787, + "loss": 0.0272, + "step": 4025 + }, + { + "epoch": 16.2, + "grad_norm": 0.39770299196243286, + "learning_rate": 0.00014516661167931377, + "loss": 0.0344, + "step": 4050 + }, + { + "epoch": 16.3, + "grad_norm": 23.780405044555664, + "learning_rate": 0.00014351699109204884, + "loss": 0.0617, + "step": 4075 + }, + { + "epoch": 16.4, + "grad_norm": 0.023103665560483932, + "learning_rate": 0.00014186737050478392, + "loss": 0.0182, + "step": 4100 + }, + { + "epoch": 16.5, + "grad_norm": 0.00998806394636631, + "learning_rate": 0.00014021774991751897, + "loss": 0.042, + "step": 4125 + }, + { + "epoch": 16.6, + "grad_norm": 43.775169372558594, + "learning_rate": 0.00013856812933025404, + "loss": 0.0182, + "step": 4150 + }, + { + "epoch": 16.7, + "grad_norm": 0.008157053031027317, + "learning_rate": 0.00013691850874298912, + "loss": 0.0377, + "step": 4175 + }, + { + "epoch": 16.8, + "grad_norm": 0.011455570347607136, + "learning_rate": 0.00013526888815572417, + "loss": 0.0255, + "step": 4200 + }, + { + "epoch": 16.9, + "grad_norm": 34.8545036315918, + "learning_rate": 0.00013361926756845927, + "loss": 0.0486, + "step": 4225 + }, + { + "epoch": 17.0, + "grad_norm": 0.006066238507628441, + "learning_rate": 0.00013196964698119435, + "loss": 0.0821, + "step": 4250 + }, + { + "epoch": 17.1, + "grad_norm": 43.46882247924805, + "learning_rate": 0.0001303200263939294, + "loss": 0.0333, + "step": 4275 + }, + { + "epoch": 17.2, + "grad_norm": 2.3177573680877686, + "learning_rate": 0.00012867040580666448, + "loss": 0.0477, + "step": 4300 + }, + { + "epoch": 17.3, + "grad_norm": 0.2264157235622406, + "learning_rate": 0.00012702078521939955, + "loss": 0.0373, + "step": 4325 + }, + { + "epoch": 17.4, + "grad_norm": 68.17675018310547, + "learning_rate": 0.0001253711646321346, + "loss": 0.036, + "step": 4350 + }, + { + "epoch": 17.5, + "grad_norm": 1.5359147787094116, + "learning_rate": 0.00012372154404486968, + "loss": 0.0437, + "step": 4375 + }, + { + "epoch": 17.6, + "grad_norm": 0.008654219098389149, + "learning_rate": 0.00012207192345760475, + "loss": 0.0295, + "step": 4400 + }, + { + "epoch": 17.7, + "grad_norm": 0.0074989828281104565, + "learning_rate": 0.00012042230287033982, + "loss": 0.0229, + "step": 4425 + }, + { + "epoch": 17.8, + "grad_norm": 0.03568067401647568, + "learning_rate": 0.0001187726822830749, + "loss": 0.0636, + "step": 4450 + }, + { + "epoch": 17.9, + "grad_norm": 0.0151940006762743, + "learning_rate": 0.00011712306169580998, + "loss": 0.0669, + "step": 4475 + }, + { + "epoch": 18.0, + "grad_norm": 0.04830991476774216, + "learning_rate": 0.00011547344110854503, + "loss": 0.0388, + "step": 4500 + }, + { + "epoch": 18.1, + "grad_norm": 1.1794861555099487, + "learning_rate": 0.00011382382052128011, + "loss": 0.0069, + "step": 4525 + }, + { + "epoch": 18.2, + "grad_norm": 13.07977294921875, + "learning_rate": 0.00011217419993401519, + "loss": 0.0315, + "step": 4550 + }, + { + "epoch": 18.3, + "grad_norm": 0.011115381494164467, + "learning_rate": 0.00011052457934675025, + "loss": 0.0144, + "step": 4575 + }, + { + "epoch": 18.4, + "grad_norm": 25.42391014099121, + "learning_rate": 0.00010887495875948532, + "loss": 0.0198, + "step": 4600 + }, + { + "epoch": 18.5, + "grad_norm": 44.36895751953125, + "learning_rate": 0.0001072253381722204, + "loss": 0.0297, + "step": 4625 + }, + { + "epoch": 18.6, + "grad_norm": 6.828146457672119, + "learning_rate": 0.00010557571758495546, + "loss": 0.0348, + "step": 4650 + }, + { + "epoch": 18.7, + "grad_norm": 38.9465446472168, + "learning_rate": 0.00010392609699769054, + "loss": 0.0336, + "step": 4675 + }, + { + "epoch": 18.8, + "grad_norm": 0.015516690909862518, + "learning_rate": 0.00010227647641042562, + "loss": 0.032, + "step": 4700 + }, + { + "epoch": 18.9, + "grad_norm": 5.419480323791504, + "learning_rate": 0.00010062685582316067, + "loss": 0.0169, + "step": 4725 + }, + { + "epoch": 19.0, + "grad_norm": 0.7849720120429993, + "learning_rate": 9.897723523589574e-05, + "loss": 0.0321, + "step": 4750 + }, + { + "epoch": 19.1, + "grad_norm": 25.438196182250977, + "learning_rate": 9.732761464863082e-05, + "loss": 0.046, + "step": 4775 + }, + { + "epoch": 19.2, + "grad_norm": 0.010676453821361065, + "learning_rate": 9.56779940613659e-05, + "loss": 0.0111, + "step": 4800 + }, + { + "epoch": 19.3, + "grad_norm": 0.011585243977606297, + "learning_rate": 9.402837347410096e-05, + "loss": 0.0154, + "step": 4825 + }, + { + "epoch": 19.4, + "grad_norm": 0.012879762798547745, + "learning_rate": 9.237875288683603e-05, + "loss": 0.02, + "step": 4850 + }, + { + "epoch": 19.5, + "grad_norm": 0.4159695506095886, + "learning_rate": 9.072913229957111e-05, + "loss": 0.0277, + "step": 4875 + }, + { + "epoch": 19.6, + "grad_norm": 0.013872765935957432, + "learning_rate": 8.907951171230617e-05, + "loss": 0.0069, + "step": 4900 + }, + { + "epoch": 19.7, + "grad_norm": 3.4260928630828857, + "learning_rate": 8.742989112504124e-05, + "loss": 0.0319, + "step": 4925 + }, + { + "epoch": 19.8, + "grad_norm": 0.043105900287628174, + "learning_rate": 8.578027053777631e-05, + "loss": 0.0445, + "step": 4950 + }, + { + "epoch": 19.9, + "grad_norm": 10.97398853302002, + "learning_rate": 8.413064995051139e-05, + "loss": 0.0185, + "step": 4975 + }, + { + "epoch": 20.0, + "grad_norm": 12.879217147827148, + "learning_rate": 8.248102936324645e-05, + "loss": 0.0197, + "step": 5000 + }, + { + "epoch": 20.1, + "grad_norm": 64.58011627197266, + "learning_rate": 8.083140877598153e-05, + "loss": 0.0228, + "step": 5025 + }, + { + "epoch": 20.2, + "grad_norm": 43.28678512573242, + "learning_rate": 7.91817881887166e-05, + "loss": 0.0311, + "step": 5050 + }, + { + "epoch": 20.3, + "grad_norm": 0.005558234639465809, + "learning_rate": 7.753216760145167e-05, + "loss": 0.0283, + "step": 5075 + }, + { + "epoch": 20.4, + "grad_norm": 0.19123849272727966, + "learning_rate": 7.588254701418674e-05, + "loss": 0.0445, + "step": 5100 + }, + { + "epoch": 20.5, + "grad_norm": 5.734357833862305, + "learning_rate": 7.423292642692181e-05, + "loss": 0.0145, + "step": 5125 + }, + { + "epoch": 20.6, + "grad_norm": 0.005894747097045183, + "learning_rate": 7.258330583965688e-05, + "loss": 0.0233, + "step": 5150 + }, + { + "epoch": 20.7, + "grad_norm": 0.018981292843818665, + "learning_rate": 7.093368525239196e-05, + "loss": 0.0273, + "step": 5175 + }, + { + "epoch": 20.8, + "grad_norm": 0.09156472235918045, + "learning_rate": 6.928406466512702e-05, + "loss": 0.0233, + "step": 5200 + }, + { + "epoch": 20.9, + "grad_norm": 0.8908875584602356, + "learning_rate": 6.763444407786209e-05, + "loss": 0.0221, + "step": 5225 + }, + { + "epoch": 21.0, + "grad_norm": 6.745031833648682, + "learning_rate": 6.598482349059718e-05, + "loss": 0.0278, + "step": 5250 + }, + { + "epoch": 21.1, + "grad_norm": 0.009193326346576214, + "learning_rate": 6.433520290333224e-05, + "loss": 0.0226, + "step": 5275 + }, + { + "epoch": 21.2, + "grad_norm": 2.408499002456665, + "learning_rate": 6.26855823160673e-05, + "loss": 0.0202, + "step": 5300 + }, + { + "epoch": 21.3, + "grad_norm": 0.006645245011895895, + "learning_rate": 6.103596172880238e-05, + "loss": 0.0234, + "step": 5325 + }, + { + "epoch": 21.4, + "grad_norm": 0.006036018021404743, + "learning_rate": 5.938634114153745e-05, + "loss": 0.0333, + "step": 5350 + }, + { + "epoch": 21.5, + "grad_norm": 0.25791531801223755, + "learning_rate": 5.7736720554272516e-05, + "loss": 0.012, + "step": 5375 + }, + { + "epoch": 21.6, + "grad_norm": 0.9368523359298706, + "learning_rate": 5.608709996700759e-05, + "loss": 0.0078, + "step": 5400 + }, + { + "epoch": 21.7, + "grad_norm": 0.0046651544980704784, + "learning_rate": 5.443747937974266e-05, + "loss": 0.0096, + "step": 5425 + }, + { + "epoch": 21.8, + "grad_norm": 0.004647469613701105, + "learning_rate": 5.278785879247773e-05, + "loss": 0.0299, + "step": 5450 + }, + { + "epoch": 21.9, + "grad_norm": 0.004732856526970863, + "learning_rate": 5.113823820521281e-05, + "loss": 0.0185, + "step": 5475 + }, + { + "epoch": 22.0, + "grad_norm": 0.007009522989392281, + "learning_rate": 4.948861761794787e-05, + "loss": 0.0228, + "step": 5500 + }, + { + "epoch": 22.1, + "grad_norm": 25.634002685546875, + "learning_rate": 4.783899703068295e-05, + "loss": 0.011, + "step": 5525 + }, + { + "epoch": 22.2, + "grad_norm": 0.009100685827434063, + "learning_rate": 4.618937644341802e-05, + "loss": 0.0056, + "step": 5550 + }, + { + "epoch": 22.3, + "grad_norm": 0.004046417307108641, + "learning_rate": 4.453975585615309e-05, + "loss": 0.0084, + "step": 5575 + }, + { + "epoch": 22.4, + "grad_norm": 0.025503572076559067, + "learning_rate": 4.2890135268888156e-05, + "loss": 0.0066, + "step": 5600 + }, + { + "epoch": 22.5, + "grad_norm": 0.005815221928060055, + "learning_rate": 4.1240514681623226e-05, + "loss": 0.0407, + "step": 5625 + }, + { + "epoch": 22.6, + "grad_norm": 0.06112132593989372, + "learning_rate": 3.95908940943583e-05, + "loss": 0.0112, + "step": 5650 + }, + { + "epoch": 22.7, + "grad_norm": 0.01705116033554077, + "learning_rate": 3.794127350709337e-05, + "loss": 0.0008, + "step": 5675 + }, + { + "epoch": 22.8, + "grad_norm": 0.005267620086669922, + "learning_rate": 3.629165291982844e-05, + "loss": 0.0131, + "step": 5700 + }, + { + "epoch": 22.9, + "grad_norm": 0.0046350546181201935, + "learning_rate": 3.464203233256351e-05, + "loss": 0.0205, + "step": 5725 + }, + { + "epoch": 23.0, + "grad_norm": 0.0063110594637691975, + "learning_rate": 3.299241174529859e-05, + "loss": 0.0072, + "step": 5750 + }, + { + "epoch": 23.1, + "grad_norm": 0.004843282513320446, + "learning_rate": 3.134279115803365e-05, + "loss": 0.033, + "step": 5775 + }, + { + "epoch": 23.2, + "grad_norm": 0.004899237770587206, + "learning_rate": 2.9693170570768723e-05, + "loss": 0.002, + "step": 5800 + }, + { + "epoch": 23.3, + "grad_norm": 0.0040695276111364365, + "learning_rate": 2.8043549983503796e-05, + "loss": 0.0185, + "step": 5825 + }, + { + "epoch": 23.4, + "grad_norm": 46.083438873291016, + "learning_rate": 2.6393929396238866e-05, + "loss": 0.0149, + "step": 5850 + }, + { + "epoch": 23.5, + "grad_norm": 0.009059540927410126, + "learning_rate": 2.4744308808973936e-05, + "loss": 0.0264, + "step": 5875 + }, + { + "epoch": 23.6, + "grad_norm": 11.57484245300293, + "learning_rate": 2.309468822170901e-05, + "loss": 0.0145, + "step": 5900 + }, + { + "epoch": 23.7, + "grad_norm": 0.00815950334072113, + "learning_rate": 2.1445067634444078e-05, + "loss": 0.015, + "step": 5925 + }, + { + "epoch": 23.8, + "grad_norm": 0.005930441431701183, + "learning_rate": 1.979544704717915e-05, + "loss": 0.0187, + "step": 5950 + }, + { + "epoch": 23.9, + "grad_norm": 0.006135927978903055, + "learning_rate": 1.814582645991422e-05, + "loss": 0.0143, + "step": 5975 + }, + { + "epoch": 24.0, + "grad_norm": 0.004939633421599865, + "learning_rate": 1.6496205872649294e-05, + "loss": 0.0088, + "step": 6000 + }, + { + "epoch": 24.1, + "grad_norm": 0.0048079704865813255, + "learning_rate": 1.4846585285384362e-05, + "loss": 0.0011, + "step": 6025 + }, + { + "epoch": 24.2, + "grad_norm": 0.005153154954314232, + "learning_rate": 1.3196964698119433e-05, + "loss": 0.0005, + "step": 6050 + }, + { + "epoch": 24.3, + "grad_norm": 0.011655638925731182, + "learning_rate": 1.1547344110854504e-05, + "loss": 0.0099, + "step": 6075 + }, + { + "epoch": 24.4, + "grad_norm": 0.003989210817962885, + "learning_rate": 9.897723523589576e-06, + "loss": 0.0007, + "step": 6100 + }, + { + "epoch": 24.5, + "grad_norm": 0.0037553769070655107, + "learning_rate": 8.248102936324647e-06, + "loss": 0.0007, + "step": 6125 + }, + { + "epoch": 24.6, + "grad_norm": 18.55050277709961, + "learning_rate": 6.5984823490597165e-06, + "loss": 0.0239, + "step": 6150 + }, + { + "epoch": 24.7, + "grad_norm": 0.017767202109098434, + "learning_rate": 4.948861761794788e-06, + "loss": 0.0005, + "step": 6175 + }, + { + "epoch": 24.8, + "grad_norm": 52.21216583251953, + "learning_rate": 3.2992411745298582e-06, + "loss": 0.02, + "step": 6200 + }, + { + "epoch": 24.9, + "grad_norm": 0.004072035662829876, + "learning_rate": 1.6496205872649291e-06, + "loss": 0.0246, + "step": 6225 + }, + { + "epoch": 25.0, + "grad_norm": 1.2685972452163696, + "learning_rate": 0.0, + "loss": 0.0096, + "step": 6250 + } + ], + "logging_steps": 25, + "max_steps": 6250, + "num_input_tokens_seen": 0, + "num_train_epochs": 25, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 317621760000000.0, + "train_batch_size": 20, + "trial_name": null, + "trial_params": null +} diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6250/training_args.bin b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6250/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..798d8a9f3a9513692035320aba5484c89ac49e69 --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6250/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1cabf5db3e9d79ec0c1945933395ff57418204414cc81f77c90324d003fac782 +size 5304 diff --git a/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6250/vocab.txt b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6250/vocab.txt new file mode 100644 index 0000000000000000000000000000000000000000..cc9ad53732e21ea9180f18e1fb67a38cc5702d4c --- /dev/null +++ b/Finetuned Models/finetuning_outputs/tfbs/TinyDNABERT_base_model/checkpoint-6250/vocab.txt @@ -0,0 +1,4101 @@ +[PAD] +[UNK] +[CLS] +[SEP] +[MASK] +AAAAAA +AAAAAT +AAAAAC +AAAAAG +AAAATA +AAAATT +AAAATC +AAAATG +AAAACA +AAAACT +AAAACC +AAAACG +AAAAGA +AAAAGT +AAAAGC +AAAAGG +AAATAA +AAATAT +AAATAC +AAATAG +AAATTA +AAATTT +AAATTC +AAATTG +AAATCA +AAATCT +AAATCC +AAATCG +AAATGA +AAATGT +AAATGC +AAATGG +AAACAA +AAACAT +AAACAC +AAACAG +AAACTA +AAACTT +AAACTC +AAACTG +AAACCA +AAACCT +AAACCC +AAACCG +AAACGA +AAACGT +AAACGC +AAACGG +AAAGAA +AAAGAT +AAAGAC +AAAGAG +AAAGTA +AAAGTT +AAAGTC +AAAGTG +AAAGCA +AAAGCT +AAAGCC +AAAGCG +AAAGGA +AAAGGT +AAAGGC +AAAGGG +AATAAA +AATAAT +AATAAC +AATAAG +AATATA +AATATT +AATATC +AATATG +AATACA +AATACT +AATACC +AATACG +AATAGA +AATAGT +AATAGC +AATAGG +AATTAA +AATTAT +AATTAC +AATTAG +AATTTA +AATTTT +AATTTC +AATTTG +AATTCA +AATTCT +AATTCC +AATTCG +AATTGA +AATTGT +AATTGC +AATTGG +AATCAA +AATCAT +AATCAC +AATCAG +AATCTA +AATCTT +AATCTC +AATCTG +AATCCA +AATCCT +AATCCC +AATCCG +AATCGA +AATCGT +AATCGC +AATCGG +AATGAA +AATGAT +AATGAC +AATGAG +AATGTA +AATGTT +AATGTC +AATGTG +AATGCA +AATGCT +AATGCC +AATGCG +AATGGA +AATGGT +AATGGC +AATGGG +AACAAA +AACAAT +AACAAC +AACAAG +AACATA +AACATT +AACATC +AACATG +AACACA +AACACT +AACACC +AACACG +AACAGA +AACAGT +AACAGC +AACAGG +AACTAA +AACTAT +AACTAC +AACTAG +AACTTA +AACTTT +AACTTC +AACTTG +AACTCA +AACTCT +AACTCC +AACTCG +AACTGA +AACTGT +AACTGC +AACTGG +AACCAA +AACCAT +AACCAC +AACCAG +AACCTA +AACCTT +AACCTC +AACCTG +AACCCA +AACCCT +AACCCC +AACCCG +AACCGA +AACCGT +AACCGC +AACCGG +AACGAA +AACGAT +AACGAC +AACGAG +AACGTA +AACGTT +AACGTC +AACGTG +AACGCA +AACGCT +AACGCC +AACGCG +AACGGA +AACGGT +AACGGC +AACGGG +AAGAAA +AAGAAT +AAGAAC +AAGAAG +AAGATA +AAGATT +AAGATC +AAGATG +AAGACA +AAGACT +AAGACC +AAGACG +AAGAGA +AAGAGT +AAGAGC +AAGAGG +AAGTAA +AAGTAT +AAGTAC +AAGTAG +AAGTTA +AAGTTT +AAGTTC +AAGTTG +AAGTCA +AAGTCT +AAGTCC +AAGTCG +AAGTGA +AAGTGT +AAGTGC +AAGTGG +AAGCAA +AAGCAT +AAGCAC +AAGCAG +AAGCTA +AAGCTT +AAGCTC +AAGCTG +AAGCCA +AAGCCT +AAGCCC +AAGCCG +AAGCGA +AAGCGT +AAGCGC +AAGCGG +AAGGAA +AAGGAT +AAGGAC +AAGGAG +AAGGTA +AAGGTT +AAGGTC +AAGGTG +AAGGCA +AAGGCT +AAGGCC +AAGGCG +AAGGGA +AAGGGT +AAGGGC +AAGGGG +ATAAAA +ATAAAT +ATAAAC +ATAAAG +ATAATA +ATAATT +ATAATC +ATAATG +ATAACA +ATAACT +ATAACC +ATAACG +ATAAGA +ATAAGT +ATAAGC +ATAAGG +ATATAA +ATATAT +ATATAC +ATATAG +ATATTA +ATATTT +ATATTC +ATATTG +ATATCA +ATATCT +ATATCC +ATATCG +ATATGA +ATATGT +ATATGC +ATATGG +ATACAA +ATACAT +ATACAC +ATACAG +ATACTA +ATACTT +ATACTC +ATACTG +ATACCA +ATACCT +ATACCC +ATACCG +ATACGA +ATACGT +ATACGC +ATACGG +ATAGAA +ATAGAT +ATAGAC +ATAGAG +ATAGTA +ATAGTT +ATAGTC +ATAGTG +ATAGCA +ATAGCT +ATAGCC +ATAGCG +ATAGGA +ATAGGT +ATAGGC +ATAGGG +ATTAAA +ATTAAT +ATTAAC +ATTAAG +ATTATA +ATTATT +ATTATC +ATTATG +ATTACA +ATTACT +ATTACC +ATTACG +ATTAGA +ATTAGT +ATTAGC +ATTAGG +ATTTAA +ATTTAT +ATTTAC +ATTTAG +ATTTTA +ATTTTT +ATTTTC +ATTTTG +ATTTCA +ATTTCT +ATTTCC +ATTTCG +ATTTGA +ATTTGT +ATTTGC +ATTTGG +ATTCAA +ATTCAT +ATTCAC +ATTCAG +ATTCTA +ATTCTT +ATTCTC +ATTCTG +ATTCCA +ATTCCT +ATTCCC +ATTCCG +ATTCGA +ATTCGT +ATTCGC +ATTCGG +ATTGAA +ATTGAT +ATTGAC +ATTGAG +ATTGTA +ATTGTT +ATTGTC +ATTGTG +ATTGCA +ATTGCT +ATTGCC +ATTGCG +ATTGGA +ATTGGT +ATTGGC +ATTGGG +ATCAAA +ATCAAT +ATCAAC +ATCAAG +ATCATA +ATCATT +ATCATC +ATCATG +ATCACA +ATCACT +ATCACC +ATCACG +ATCAGA +ATCAGT +ATCAGC +ATCAGG +ATCTAA +ATCTAT +ATCTAC +ATCTAG +ATCTTA +ATCTTT +ATCTTC +ATCTTG +ATCTCA +ATCTCT +ATCTCC +ATCTCG +ATCTGA +ATCTGT +ATCTGC +ATCTGG +ATCCAA +ATCCAT +ATCCAC +ATCCAG +ATCCTA +ATCCTT +ATCCTC +ATCCTG +ATCCCA +ATCCCT +ATCCCC +ATCCCG +ATCCGA +ATCCGT +ATCCGC +ATCCGG +ATCGAA +ATCGAT +ATCGAC +ATCGAG +ATCGTA +ATCGTT +ATCGTC +ATCGTG +ATCGCA +ATCGCT +ATCGCC +ATCGCG +ATCGGA +ATCGGT +ATCGGC +ATCGGG +ATGAAA +ATGAAT +ATGAAC +ATGAAG +ATGATA +ATGATT +ATGATC +ATGATG +ATGACA +ATGACT +ATGACC +ATGACG +ATGAGA +ATGAGT +ATGAGC +ATGAGG +ATGTAA +ATGTAT +ATGTAC +ATGTAG +ATGTTA +ATGTTT +ATGTTC +ATGTTG +ATGTCA +ATGTCT +ATGTCC +ATGTCG +ATGTGA +ATGTGT +ATGTGC +ATGTGG +ATGCAA +ATGCAT +ATGCAC +ATGCAG +ATGCTA +ATGCTT +ATGCTC +ATGCTG +ATGCCA +ATGCCT +ATGCCC +ATGCCG +ATGCGA +ATGCGT +ATGCGC +ATGCGG +ATGGAA +ATGGAT +ATGGAC +ATGGAG +ATGGTA +ATGGTT +ATGGTC +ATGGTG +ATGGCA +ATGGCT +ATGGCC +ATGGCG +ATGGGA +ATGGGT +ATGGGC +ATGGGG +ACAAAA +ACAAAT +ACAAAC +ACAAAG +ACAATA +ACAATT +ACAATC +ACAATG +ACAACA +ACAACT +ACAACC +ACAACG +ACAAGA +ACAAGT +ACAAGC +ACAAGG +ACATAA +ACATAT +ACATAC +ACATAG +ACATTA +ACATTT +ACATTC +ACATTG +ACATCA +ACATCT +ACATCC +ACATCG +ACATGA +ACATGT +ACATGC +ACATGG +ACACAA +ACACAT +ACACAC +ACACAG +ACACTA +ACACTT +ACACTC +ACACTG +ACACCA +ACACCT +ACACCC +ACACCG +ACACGA +ACACGT +ACACGC +ACACGG +ACAGAA +ACAGAT +ACAGAC +ACAGAG +ACAGTA +ACAGTT +ACAGTC +ACAGTG +ACAGCA +ACAGCT +ACAGCC +ACAGCG +ACAGGA +ACAGGT +ACAGGC +ACAGGG +ACTAAA +ACTAAT +ACTAAC +ACTAAG +ACTATA +ACTATT +ACTATC +ACTATG +ACTACA +ACTACT +ACTACC +ACTACG +ACTAGA +ACTAGT +ACTAGC +ACTAGG +ACTTAA +ACTTAT +ACTTAC +ACTTAG +ACTTTA +ACTTTT +ACTTTC +ACTTTG +ACTTCA +ACTTCT +ACTTCC +ACTTCG +ACTTGA +ACTTGT +ACTTGC +ACTTGG +ACTCAA +ACTCAT +ACTCAC +ACTCAG +ACTCTA +ACTCTT +ACTCTC +ACTCTG +ACTCCA +ACTCCT +ACTCCC +ACTCCG +ACTCGA +ACTCGT +ACTCGC +ACTCGG +ACTGAA +ACTGAT +ACTGAC +ACTGAG +ACTGTA +ACTGTT +ACTGTC +ACTGTG +ACTGCA +ACTGCT +ACTGCC +ACTGCG +ACTGGA +ACTGGT +ACTGGC +ACTGGG +ACCAAA +ACCAAT +ACCAAC +ACCAAG +ACCATA +ACCATT +ACCATC +ACCATG +ACCACA +ACCACT +ACCACC +ACCACG +ACCAGA +ACCAGT +ACCAGC +ACCAGG +ACCTAA +ACCTAT +ACCTAC +ACCTAG +ACCTTA +ACCTTT +ACCTTC +ACCTTG +ACCTCA +ACCTCT +ACCTCC +ACCTCG +ACCTGA +ACCTGT +ACCTGC +ACCTGG +ACCCAA +ACCCAT +ACCCAC +ACCCAG +ACCCTA +ACCCTT +ACCCTC +ACCCTG +ACCCCA +ACCCCT +ACCCCC +ACCCCG +ACCCGA +ACCCGT +ACCCGC +ACCCGG +ACCGAA +ACCGAT +ACCGAC +ACCGAG +ACCGTA +ACCGTT +ACCGTC +ACCGTG +ACCGCA +ACCGCT +ACCGCC +ACCGCG +ACCGGA +ACCGGT +ACCGGC +ACCGGG +ACGAAA +ACGAAT +ACGAAC +ACGAAG +ACGATA +ACGATT +ACGATC +ACGATG +ACGACA +ACGACT +ACGACC +ACGACG +ACGAGA +ACGAGT +ACGAGC +ACGAGG +ACGTAA +ACGTAT +ACGTAC +ACGTAG +ACGTTA +ACGTTT +ACGTTC +ACGTTG +ACGTCA +ACGTCT +ACGTCC +ACGTCG +ACGTGA +ACGTGT +ACGTGC +ACGTGG +ACGCAA +ACGCAT +ACGCAC +ACGCAG +ACGCTA +ACGCTT +ACGCTC +ACGCTG +ACGCCA +ACGCCT +ACGCCC +ACGCCG +ACGCGA +ACGCGT +ACGCGC +ACGCGG +ACGGAA +ACGGAT +ACGGAC +ACGGAG +ACGGTA +ACGGTT +ACGGTC +ACGGTG +ACGGCA +ACGGCT +ACGGCC +ACGGCG +ACGGGA +ACGGGT +ACGGGC +ACGGGG +AGAAAA +AGAAAT +AGAAAC +AGAAAG +AGAATA +AGAATT +AGAATC +AGAATG +AGAACA +AGAACT +AGAACC +AGAACG +AGAAGA +AGAAGT +AGAAGC +AGAAGG +AGATAA +AGATAT +AGATAC +AGATAG +AGATTA +AGATTT +AGATTC +AGATTG +AGATCA +AGATCT +AGATCC +AGATCG +AGATGA +AGATGT +AGATGC +AGATGG +AGACAA +AGACAT +AGACAC +AGACAG +AGACTA +AGACTT +AGACTC +AGACTG +AGACCA +AGACCT +AGACCC +AGACCG +AGACGA +AGACGT +AGACGC +AGACGG +AGAGAA +AGAGAT +AGAGAC +AGAGAG +AGAGTA +AGAGTT +AGAGTC +AGAGTG +AGAGCA +AGAGCT +AGAGCC +AGAGCG +AGAGGA +AGAGGT +AGAGGC +AGAGGG +AGTAAA +AGTAAT +AGTAAC +AGTAAG +AGTATA +AGTATT +AGTATC +AGTATG +AGTACA +AGTACT +AGTACC +AGTACG +AGTAGA +AGTAGT +AGTAGC +AGTAGG +AGTTAA +AGTTAT +AGTTAC +AGTTAG +AGTTTA +AGTTTT +AGTTTC +AGTTTG +AGTTCA +AGTTCT +AGTTCC +AGTTCG +AGTTGA +AGTTGT +AGTTGC +AGTTGG +AGTCAA +AGTCAT +AGTCAC +AGTCAG +AGTCTA +AGTCTT +AGTCTC +AGTCTG +AGTCCA +AGTCCT +AGTCCC +AGTCCG +AGTCGA +AGTCGT +AGTCGC +AGTCGG +AGTGAA +AGTGAT +AGTGAC +AGTGAG +AGTGTA +AGTGTT +AGTGTC +AGTGTG +AGTGCA +AGTGCT +AGTGCC +AGTGCG +AGTGGA +AGTGGT +AGTGGC +AGTGGG +AGCAAA +AGCAAT +AGCAAC +AGCAAG +AGCATA +AGCATT +AGCATC +AGCATG +AGCACA +AGCACT +AGCACC +AGCACG +AGCAGA +AGCAGT +AGCAGC +AGCAGG +AGCTAA +AGCTAT +AGCTAC +AGCTAG +AGCTTA +AGCTTT +AGCTTC +AGCTTG +AGCTCA +AGCTCT +AGCTCC +AGCTCG +AGCTGA +AGCTGT +AGCTGC +AGCTGG +AGCCAA +AGCCAT +AGCCAC +AGCCAG +AGCCTA +AGCCTT +AGCCTC +AGCCTG +AGCCCA +AGCCCT +AGCCCC +AGCCCG +AGCCGA +AGCCGT +AGCCGC +AGCCGG +AGCGAA +AGCGAT +AGCGAC +AGCGAG +AGCGTA +AGCGTT +AGCGTC +AGCGTG +AGCGCA +AGCGCT +AGCGCC +AGCGCG +AGCGGA +AGCGGT +AGCGGC +AGCGGG +AGGAAA +AGGAAT +AGGAAC +AGGAAG +AGGATA +AGGATT +AGGATC +AGGATG +AGGACA +AGGACT +AGGACC +AGGACG +AGGAGA +AGGAGT +AGGAGC +AGGAGG +AGGTAA +AGGTAT +AGGTAC +AGGTAG +AGGTTA +AGGTTT +AGGTTC +AGGTTG +AGGTCA +AGGTCT +AGGTCC +AGGTCG +AGGTGA +AGGTGT +AGGTGC +AGGTGG +AGGCAA +AGGCAT +AGGCAC +AGGCAG +AGGCTA +AGGCTT +AGGCTC +AGGCTG +AGGCCA +AGGCCT +AGGCCC +AGGCCG +AGGCGA +AGGCGT +AGGCGC +AGGCGG +AGGGAA +AGGGAT +AGGGAC +AGGGAG +AGGGTA +AGGGTT +AGGGTC +AGGGTG +AGGGCA +AGGGCT +AGGGCC +AGGGCG +AGGGGA +AGGGGT +AGGGGC +AGGGGG +TAAAAA +TAAAAT +TAAAAC +TAAAAG +TAAATA +TAAATT +TAAATC +TAAATG +TAAACA +TAAACT +TAAACC +TAAACG +TAAAGA +TAAAGT +TAAAGC +TAAAGG +TAATAA +TAATAT +TAATAC +TAATAG +TAATTA +TAATTT +TAATTC +TAATTG +TAATCA +TAATCT +TAATCC +TAATCG +TAATGA +TAATGT +TAATGC +TAATGG +TAACAA +TAACAT +TAACAC +TAACAG +TAACTA +TAACTT +TAACTC +TAACTG +TAACCA +TAACCT +TAACCC +TAACCG +TAACGA +TAACGT +TAACGC +TAACGG +TAAGAA +TAAGAT +TAAGAC +TAAGAG +TAAGTA +TAAGTT +TAAGTC +TAAGTG +TAAGCA +TAAGCT +TAAGCC +TAAGCG +TAAGGA +TAAGGT +TAAGGC +TAAGGG +TATAAA +TATAAT +TATAAC +TATAAG +TATATA +TATATT +TATATC +TATATG +TATACA +TATACT +TATACC +TATACG +TATAGA +TATAGT +TATAGC +TATAGG +TATTAA +TATTAT +TATTAC +TATTAG +TATTTA +TATTTT +TATTTC +TATTTG +TATTCA +TATTCT +TATTCC +TATTCG +TATTGA +TATTGT +TATTGC +TATTGG +TATCAA +TATCAT +TATCAC +TATCAG +TATCTA +TATCTT +TATCTC +TATCTG +TATCCA +TATCCT +TATCCC +TATCCG +TATCGA +TATCGT +TATCGC +TATCGG +TATGAA +TATGAT +TATGAC +TATGAG +TATGTA +TATGTT +TATGTC +TATGTG +TATGCA +TATGCT +TATGCC +TATGCG +TATGGA +TATGGT +TATGGC +TATGGG +TACAAA +TACAAT +TACAAC +TACAAG +TACATA +TACATT +TACATC +TACATG +TACACA +TACACT +TACACC +TACACG +TACAGA +TACAGT +TACAGC +TACAGG +TACTAA +TACTAT +TACTAC +TACTAG +TACTTA +TACTTT +TACTTC +TACTTG +TACTCA +TACTCT +TACTCC +TACTCG +TACTGA +TACTGT +TACTGC +TACTGG +TACCAA +TACCAT +TACCAC +TACCAG +TACCTA +TACCTT +TACCTC +TACCTG +TACCCA +TACCCT +TACCCC +TACCCG +TACCGA +TACCGT +TACCGC +TACCGG +TACGAA +TACGAT +TACGAC +TACGAG +TACGTA +TACGTT +TACGTC +TACGTG +TACGCA +TACGCT +TACGCC +TACGCG +TACGGA +TACGGT +TACGGC +TACGGG +TAGAAA +TAGAAT +TAGAAC +TAGAAG +TAGATA +TAGATT +TAGATC +TAGATG +TAGACA +TAGACT +TAGACC +TAGACG +TAGAGA +TAGAGT +TAGAGC +TAGAGG +TAGTAA +TAGTAT +TAGTAC +TAGTAG +TAGTTA +TAGTTT +TAGTTC +TAGTTG +TAGTCA +TAGTCT +TAGTCC +TAGTCG +TAGTGA +TAGTGT +TAGTGC +TAGTGG +TAGCAA +TAGCAT +TAGCAC +TAGCAG +TAGCTA +TAGCTT +TAGCTC +TAGCTG +TAGCCA +TAGCCT +TAGCCC +TAGCCG +TAGCGA +TAGCGT +TAGCGC +TAGCGG +TAGGAA +TAGGAT +TAGGAC +TAGGAG +TAGGTA +TAGGTT +TAGGTC +TAGGTG +TAGGCA +TAGGCT +TAGGCC +TAGGCG +TAGGGA +TAGGGT +TAGGGC +TAGGGG +TTAAAA +TTAAAT +TTAAAC +TTAAAG +TTAATA +TTAATT +TTAATC +TTAATG +TTAACA +TTAACT +TTAACC +TTAACG +TTAAGA +TTAAGT +TTAAGC +TTAAGG +TTATAA +TTATAT +TTATAC +TTATAG +TTATTA +TTATTT +TTATTC +TTATTG +TTATCA +TTATCT +TTATCC +TTATCG +TTATGA +TTATGT +TTATGC +TTATGG +TTACAA +TTACAT +TTACAC +TTACAG +TTACTA +TTACTT +TTACTC +TTACTG +TTACCA +TTACCT +TTACCC +TTACCG +TTACGA +TTACGT +TTACGC +TTACGG +TTAGAA +TTAGAT +TTAGAC +TTAGAG +TTAGTA +TTAGTT +TTAGTC +TTAGTG +TTAGCA +TTAGCT +TTAGCC +TTAGCG +TTAGGA +TTAGGT +TTAGGC +TTAGGG +TTTAAA +TTTAAT +TTTAAC +TTTAAG +TTTATA +TTTATT +TTTATC +TTTATG +TTTACA +TTTACT +TTTACC +TTTACG +TTTAGA +TTTAGT +TTTAGC +TTTAGG +TTTTAA +TTTTAT +TTTTAC +TTTTAG +TTTTTA +TTTTTT +TTTTTC +TTTTTG +TTTTCA +TTTTCT +TTTTCC +TTTTCG +TTTTGA +TTTTGT +TTTTGC +TTTTGG +TTTCAA +TTTCAT +TTTCAC +TTTCAG +TTTCTA +TTTCTT +TTTCTC +TTTCTG +TTTCCA +TTTCCT +TTTCCC +TTTCCG +TTTCGA +TTTCGT +TTTCGC +TTTCGG +TTTGAA +TTTGAT +TTTGAC +TTTGAG +TTTGTA +TTTGTT +TTTGTC +TTTGTG +TTTGCA +TTTGCT +TTTGCC +TTTGCG +TTTGGA +TTTGGT +TTTGGC +TTTGGG +TTCAAA +TTCAAT +TTCAAC +TTCAAG +TTCATA +TTCATT +TTCATC +TTCATG +TTCACA +TTCACT +TTCACC +TTCACG +TTCAGA +TTCAGT +TTCAGC +TTCAGG +TTCTAA +TTCTAT +TTCTAC +TTCTAG +TTCTTA +TTCTTT +TTCTTC +TTCTTG +TTCTCA +TTCTCT +TTCTCC +TTCTCG +TTCTGA +TTCTGT +TTCTGC +TTCTGG +TTCCAA +TTCCAT +TTCCAC +TTCCAG +TTCCTA +TTCCTT +TTCCTC +TTCCTG +TTCCCA +TTCCCT +TTCCCC +TTCCCG +TTCCGA +TTCCGT +TTCCGC +TTCCGG +TTCGAA +TTCGAT +TTCGAC +TTCGAG +TTCGTA +TTCGTT +TTCGTC +TTCGTG +TTCGCA +TTCGCT +TTCGCC +TTCGCG +TTCGGA +TTCGGT +TTCGGC +TTCGGG +TTGAAA +TTGAAT +TTGAAC +TTGAAG +TTGATA +TTGATT +TTGATC +TTGATG +TTGACA +TTGACT +TTGACC +TTGACG +TTGAGA +TTGAGT +TTGAGC +TTGAGG +TTGTAA +TTGTAT +TTGTAC +TTGTAG +TTGTTA +TTGTTT +TTGTTC +TTGTTG +TTGTCA +TTGTCT +TTGTCC +TTGTCG +TTGTGA +TTGTGT +TTGTGC +TTGTGG +TTGCAA +TTGCAT +TTGCAC +TTGCAG +TTGCTA +TTGCTT +TTGCTC +TTGCTG +TTGCCA +TTGCCT +TTGCCC +TTGCCG +TTGCGA +TTGCGT +TTGCGC +TTGCGG +TTGGAA +TTGGAT +TTGGAC +TTGGAG +TTGGTA +TTGGTT +TTGGTC +TTGGTG +TTGGCA +TTGGCT +TTGGCC +TTGGCG +TTGGGA +TTGGGT +TTGGGC +TTGGGG +TCAAAA +TCAAAT +TCAAAC +TCAAAG +TCAATA +TCAATT +TCAATC +TCAATG +TCAACA +TCAACT +TCAACC +TCAACG +TCAAGA +TCAAGT +TCAAGC +TCAAGG +TCATAA +TCATAT +TCATAC +TCATAG +TCATTA +TCATTT +TCATTC +TCATTG +TCATCA +TCATCT +TCATCC +TCATCG +TCATGA +TCATGT +TCATGC +TCATGG +TCACAA +TCACAT +TCACAC +TCACAG +TCACTA +TCACTT +TCACTC +TCACTG +TCACCA +TCACCT +TCACCC +TCACCG +TCACGA +TCACGT +TCACGC +TCACGG +TCAGAA +TCAGAT +TCAGAC +TCAGAG +TCAGTA +TCAGTT +TCAGTC +TCAGTG +TCAGCA +TCAGCT +TCAGCC +TCAGCG +TCAGGA +TCAGGT +TCAGGC +TCAGGG +TCTAAA +TCTAAT +TCTAAC +TCTAAG +TCTATA +TCTATT +TCTATC +TCTATG +TCTACA +TCTACT +TCTACC +TCTACG +TCTAGA +TCTAGT +TCTAGC +TCTAGG +TCTTAA +TCTTAT +TCTTAC +TCTTAG +TCTTTA +TCTTTT +TCTTTC +TCTTTG +TCTTCA +TCTTCT +TCTTCC +TCTTCG +TCTTGA +TCTTGT +TCTTGC +TCTTGG +TCTCAA +TCTCAT +TCTCAC +TCTCAG +TCTCTA +TCTCTT +TCTCTC +TCTCTG +TCTCCA +TCTCCT +TCTCCC +TCTCCG +TCTCGA +TCTCGT +TCTCGC +TCTCGG +TCTGAA +TCTGAT +TCTGAC +TCTGAG +TCTGTA +TCTGTT +TCTGTC +TCTGTG +TCTGCA +TCTGCT +TCTGCC +TCTGCG +TCTGGA +TCTGGT +TCTGGC +TCTGGG +TCCAAA +TCCAAT +TCCAAC +TCCAAG +TCCATA +TCCATT +TCCATC +TCCATG +TCCACA +TCCACT +TCCACC +TCCACG +TCCAGA +TCCAGT +TCCAGC +TCCAGG +TCCTAA +TCCTAT +TCCTAC +TCCTAG +TCCTTA +TCCTTT +TCCTTC +TCCTTG +TCCTCA +TCCTCT +TCCTCC +TCCTCG +TCCTGA +TCCTGT +TCCTGC +TCCTGG +TCCCAA +TCCCAT +TCCCAC +TCCCAG +TCCCTA +TCCCTT +TCCCTC +TCCCTG +TCCCCA +TCCCCT +TCCCCC +TCCCCG +TCCCGA +TCCCGT +TCCCGC +TCCCGG +TCCGAA +TCCGAT +TCCGAC +TCCGAG +TCCGTA +TCCGTT +TCCGTC +TCCGTG +TCCGCA +TCCGCT +TCCGCC +TCCGCG +TCCGGA +TCCGGT +TCCGGC +TCCGGG +TCGAAA +TCGAAT +TCGAAC +TCGAAG +TCGATA +TCGATT +TCGATC +TCGATG +TCGACA +TCGACT +TCGACC +TCGACG +TCGAGA +TCGAGT +TCGAGC +TCGAGG +TCGTAA +TCGTAT +TCGTAC +TCGTAG +TCGTTA +TCGTTT +TCGTTC +TCGTTG +TCGTCA +TCGTCT +TCGTCC +TCGTCG +TCGTGA +TCGTGT +TCGTGC +TCGTGG +TCGCAA +TCGCAT +TCGCAC +TCGCAG +TCGCTA +TCGCTT +TCGCTC +TCGCTG +TCGCCA +TCGCCT +TCGCCC +TCGCCG +TCGCGA +TCGCGT +TCGCGC +TCGCGG +TCGGAA +TCGGAT +TCGGAC +TCGGAG +TCGGTA +TCGGTT +TCGGTC +TCGGTG +TCGGCA +TCGGCT +TCGGCC +TCGGCG +TCGGGA +TCGGGT +TCGGGC +TCGGGG +TGAAAA +TGAAAT +TGAAAC +TGAAAG +TGAATA +TGAATT +TGAATC +TGAATG +TGAACA +TGAACT +TGAACC +TGAACG +TGAAGA +TGAAGT +TGAAGC +TGAAGG +TGATAA +TGATAT +TGATAC +TGATAG +TGATTA +TGATTT +TGATTC +TGATTG +TGATCA +TGATCT +TGATCC +TGATCG +TGATGA +TGATGT +TGATGC +TGATGG +TGACAA +TGACAT +TGACAC +TGACAG +TGACTA +TGACTT +TGACTC +TGACTG +TGACCA +TGACCT +TGACCC +TGACCG +TGACGA +TGACGT +TGACGC +TGACGG +TGAGAA +TGAGAT +TGAGAC +TGAGAG +TGAGTA +TGAGTT +TGAGTC +TGAGTG +TGAGCA +TGAGCT +TGAGCC +TGAGCG +TGAGGA +TGAGGT +TGAGGC +TGAGGG +TGTAAA +TGTAAT +TGTAAC +TGTAAG +TGTATA +TGTATT +TGTATC +TGTATG +TGTACA +TGTACT +TGTACC +TGTACG +TGTAGA +TGTAGT +TGTAGC +TGTAGG +TGTTAA +TGTTAT +TGTTAC +TGTTAG +TGTTTA +TGTTTT +TGTTTC +TGTTTG +TGTTCA +TGTTCT +TGTTCC +TGTTCG +TGTTGA +TGTTGT +TGTTGC +TGTTGG +TGTCAA +TGTCAT +TGTCAC +TGTCAG +TGTCTA +TGTCTT +TGTCTC +TGTCTG +TGTCCA +TGTCCT +TGTCCC +TGTCCG +TGTCGA +TGTCGT +TGTCGC +TGTCGG +TGTGAA +TGTGAT +TGTGAC +TGTGAG +TGTGTA +TGTGTT +TGTGTC +TGTGTG +TGTGCA +TGTGCT +TGTGCC +TGTGCG +TGTGGA +TGTGGT +TGTGGC +TGTGGG +TGCAAA +TGCAAT +TGCAAC +TGCAAG +TGCATA +TGCATT +TGCATC +TGCATG +TGCACA +TGCACT +TGCACC +TGCACG +TGCAGA +TGCAGT +TGCAGC +TGCAGG +TGCTAA +TGCTAT +TGCTAC +TGCTAG +TGCTTA +TGCTTT +TGCTTC +TGCTTG +TGCTCA +TGCTCT +TGCTCC +TGCTCG +TGCTGA +TGCTGT +TGCTGC +TGCTGG +TGCCAA +TGCCAT +TGCCAC +TGCCAG +TGCCTA +TGCCTT +TGCCTC +TGCCTG +TGCCCA +TGCCCT +TGCCCC +TGCCCG +TGCCGA +TGCCGT +TGCCGC +TGCCGG +TGCGAA +TGCGAT +TGCGAC +TGCGAG +TGCGTA +TGCGTT +TGCGTC +TGCGTG +TGCGCA +TGCGCT +TGCGCC +TGCGCG +TGCGGA +TGCGGT +TGCGGC +TGCGGG +TGGAAA +TGGAAT +TGGAAC +TGGAAG +TGGATA +TGGATT +TGGATC +TGGATG +TGGACA +TGGACT +TGGACC +TGGACG +TGGAGA +TGGAGT +TGGAGC +TGGAGG +TGGTAA +TGGTAT +TGGTAC +TGGTAG +TGGTTA +TGGTTT +TGGTTC +TGGTTG +TGGTCA +TGGTCT +TGGTCC +TGGTCG +TGGTGA +TGGTGT +TGGTGC +TGGTGG +TGGCAA +TGGCAT +TGGCAC +TGGCAG +TGGCTA +TGGCTT +TGGCTC +TGGCTG +TGGCCA +TGGCCT +TGGCCC +TGGCCG +TGGCGA +TGGCGT +TGGCGC +TGGCGG +TGGGAA +TGGGAT +TGGGAC +TGGGAG +TGGGTA +TGGGTT +TGGGTC +TGGGTG +TGGGCA +TGGGCT +TGGGCC +TGGGCG +TGGGGA +TGGGGT +TGGGGC +TGGGGG +CAAAAA +CAAAAT +CAAAAC +CAAAAG +CAAATA +CAAATT +CAAATC +CAAATG +CAAACA +CAAACT +CAAACC +CAAACG +CAAAGA +CAAAGT +CAAAGC +CAAAGG +CAATAA +CAATAT +CAATAC +CAATAG +CAATTA +CAATTT +CAATTC +CAATTG +CAATCA +CAATCT +CAATCC +CAATCG +CAATGA +CAATGT +CAATGC +CAATGG +CAACAA +CAACAT +CAACAC +CAACAG +CAACTA +CAACTT +CAACTC +CAACTG +CAACCA +CAACCT +CAACCC +CAACCG +CAACGA +CAACGT +CAACGC +CAACGG +CAAGAA +CAAGAT +CAAGAC +CAAGAG +CAAGTA +CAAGTT +CAAGTC +CAAGTG +CAAGCA +CAAGCT +CAAGCC +CAAGCG +CAAGGA +CAAGGT +CAAGGC +CAAGGG +CATAAA +CATAAT +CATAAC +CATAAG +CATATA +CATATT +CATATC +CATATG +CATACA +CATACT +CATACC +CATACG +CATAGA +CATAGT +CATAGC +CATAGG +CATTAA +CATTAT +CATTAC +CATTAG +CATTTA +CATTTT +CATTTC +CATTTG +CATTCA +CATTCT +CATTCC +CATTCG +CATTGA +CATTGT +CATTGC +CATTGG +CATCAA +CATCAT +CATCAC +CATCAG +CATCTA +CATCTT +CATCTC +CATCTG +CATCCA +CATCCT +CATCCC +CATCCG +CATCGA +CATCGT +CATCGC +CATCGG +CATGAA +CATGAT +CATGAC +CATGAG +CATGTA +CATGTT +CATGTC +CATGTG +CATGCA +CATGCT +CATGCC +CATGCG +CATGGA +CATGGT +CATGGC +CATGGG +CACAAA +CACAAT +CACAAC +CACAAG +CACATA +CACATT +CACATC +CACATG +CACACA +CACACT +CACACC +CACACG +CACAGA +CACAGT +CACAGC +CACAGG +CACTAA +CACTAT +CACTAC +CACTAG +CACTTA +CACTTT +CACTTC +CACTTG +CACTCA +CACTCT +CACTCC +CACTCG +CACTGA +CACTGT +CACTGC +CACTGG +CACCAA +CACCAT +CACCAC +CACCAG +CACCTA +CACCTT +CACCTC +CACCTG +CACCCA +CACCCT +CACCCC +CACCCG +CACCGA +CACCGT +CACCGC +CACCGG +CACGAA +CACGAT +CACGAC +CACGAG +CACGTA +CACGTT +CACGTC +CACGTG +CACGCA +CACGCT +CACGCC +CACGCG +CACGGA +CACGGT +CACGGC +CACGGG +CAGAAA +CAGAAT +CAGAAC +CAGAAG +CAGATA +CAGATT +CAGATC +CAGATG +CAGACA +CAGACT +CAGACC +CAGACG +CAGAGA +CAGAGT +CAGAGC +CAGAGG +CAGTAA +CAGTAT +CAGTAC +CAGTAG +CAGTTA +CAGTTT +CAGTTC +CAGTTG +CAGTCA +CAGTCT +CAGTCC +CAGTCG +CAGTGA +CAGTGT +CAGTGC +CAGTGG +CAGCAA +CAGCAT +CAGCAC +CAGCAG +CAGCTA +CAGCTT +CAGCTC +CAGCTG +CAGCCA +CAGCCT +CAGCCC +CAGCCG +CAGCGA +CAGCGT +CAGCGC +CAGCGG +CAGGAA +CAGGAT +CAGGAC +CAGGAG +CAGGTA +CAGGTT +CAGGTC +CAGGTG +CAGGCA +CAGGCT +CAGGCC +CAGGCG +CAGGGA +CAGGGT +CAGGGC +CAGGGG +CTAAAA +CTAAAT +CTAAAC +CTAAAG +CTAATA +CTAATT +CTAATC +CTAATG +CTAACA +CTAACT +CTAACC +CTAACG +CTAAGA +CTAAGT +CTAAGC +CTAAGG +CTATAA +CTATAT +CTATAC +CTATAG +CTATTA +CTATTT +CTATTC +CTATTG +CTATCA +CTATCT +CTATCC +CTATCG +CTATGA +CTATGT +CTATGC +CTATGG +CTACAA +CTACAT +CTACAC +CTACAG +CTACTA +CTACTT +CTACTC +CTACTG +CTACCA +CTACCT +CTACCC +CTACCG +CTACGA +CTACGT +CTACGC +CTACGG +CTAGAA +CTAGAT +CTAGAC +CTAGAG +CTAGTA +CTAGTT +CTAGTC +CTAGTG +CTAGCA +CTAGCT +CTAGCC +CTAGCG +CTAGGA +CTAGGT +CTAGGC +CTAGGG +CTTAAA +CTTAAT +CTTAAC +CTTAAG +CTTATA +CTTATT +CTTATC +CTTATG +CTTACA +CTTACT +CTTACC +CTTACG +CTTAGA +CTTAGT +CTTAGC +CTTAGG +CTTTAA +CTTTAT +CTTTAC +CTTTAG +CTTTTA +CTTTTT +CTTTTC +CTTTTG +CTTTCA +CTTTCT +CTTTCC +CTTTCG +CTTTGA +CTTTGT +CTTTGC +CTTTGG +CTTCAA +CTTCAT +CTTCAC +CTTCAG +CTTCTA +CTTCTT +CTTCTC +CTTCTG +CTTCCA +CTTCCT +CTTCCC +CTTCCG +CTTCGA +CTTCGT +CTTCGC +CTTCGG +CTTGAA +CTTGAT +CTTGAC +CTTGAG +CTTGTA +CTTGTT +CTTGTC +CTTGTG +CTTGCA +CTTGCT +CTTGCC +CTTGCG +CTTGGA +CTTGGT +CTTGGC +CTTGGG +CTCAAA +CTCAAT +CTCAAC +CTCAAG +CTCATA +CTCATT +CTCATC +CTCATG +CTCACA +CTCACT +CTCACC +CTCACG +CTCAGA +CTCAGT +CTCAGC +CTCAGG +CTCTAA +CTCTAT +CTCTAC +CTCTAG +CTCTTA +CTCTTT +CTCTTC +CTCTTG +CTCTCA +CTCTCT +CTCTCC +CTCTCG +CTCTGA +CTCTGT +CTCTGC +CTCTGG +CTCCAA +CTCCAT +CTCCAC +CTCCAG +CTCCTA +CTCCTT +CTCCTC +CTCCTG +CTCCCA +CTCCCT +CTCCCC +CTCCCG +CTCCGA +CTCCGT +CTCCGC +CTCCGG +CTCGAA +CTCGAT +CTCGAC +CTCGAG +CTCGTA +CTCGTT +CTCGTC +CTCGTG +CTCGCA +CTCGCT +CTCGCC +CTCGCG +CTCGGA +CTCGGT +CTCGGC +CTCGGG +CTGAAA +CTGAAT +CTGAAC +CTGAAG +CTGATA +CTGATT +CTGATC +CTGATG +CTGACA +CTGACT +CTGACC +CTGACG +CTGAGA +CTGAGT +CTGAGC +CTGAGG +CTGTAA +CTGTAT +CTGTAC +CTGTAG +CTGTTA +CTGTTT +CTGTTC +CTGTTG +CTGTCA +CTGTCT +CTGTCC +CTGTCG +CTGTGA +CTGTGT +CTGTGC +CTGTGG +CTGCAA +CTGCAT +CTGCAC +CTGCAG +CTGCTA +CTGCTT +CTGCTC +CTGCTG +CTGCCA +CTGCCT +CTGCCC +CTGCCG +CTGCGA +CTGCGT +CTGCGC +CTGCGG +CTGGAA +CTGGAT +CTGGAC +CTGGAG +CTGGTA +CTGGTT +CTGGTC +CTGGTG +CTGGCA +CTGGCT +CTGGCC +CTGGCG +CTGGGA +CTGGGT +CTGGGC +CTGGGG +CCAAAA +CCAAAT +CCAAAC +CCAAAG +CCAATA +CCAATT +CCAATC +CCAATG +CCAACA +CCAACT +CCAACC +CCAACG +CCAAGA +CCAAGT +CCAAGC +CCAAGG +CCATAA +CCATAT +CCATAC +CCATAG +CCATTA +CCATTT +CCATTC +CCATTG +CCATCA +CCATCT +CCATCC +CCATCG +CCATGA +CCATGT +CCATGC +CCATGG +CCACAA +CCACAT +CCACAC +CCACAG +CCACTA +CCACTT +CCACTC +CCACTG +CCACCA +CCACCT +CCACCC +CCACCG +CCACGA +CCACGT +CCACGC +CCACGG +CCAGAA +CCAGAT +CCAGAC +CCAGAG +CCAGTA +CCAGTT +CCAGTC +CCAGTG +CCAGCA +CCAGCT +CCAGCC +CCAGCG +CCAGGA +CCAGGT +CCAGGC +CCAGGG +CCTAAA +CCTAAT +CCTAAC +CCTAAG +CCTATA +CCTATT +CCTATC +CCTATG +CCTACA +CCTACT +CCTACC +CCTACG +CCTAGA +CCTAGT +CCTAGC +CCTAGG +CCTTAA +CCTTAT +CCTTAC +CCTTAG +CCTTTA +CCTTTT +CCTTTC +CCTTTG +CCTTCA +CCTTCT +CCTTCC +CCTTCG +CCTTGA +CCTTGT +CCTTGC +CCTTGG +CCTCAA +CCTCAT +CCTCAC +CCTCAG +CCTCTA +CCTCTT +CCTCTC +CCTCTG +CCTCCA +CCTCCT +CCTCCC +CCTCCG +CCTCGA +CCTCGT +CCTCGC +CCTCGG +CCTGAA +CCTGAT +CCTGAC +CCTGAG +CCTGTA +CCTGTT +CCTGTC +CCTGTG +CCTGCA +CCTGCT +CCTGCC +CCTGCG +CCTGGA +CCTGGT +CCTGGC +CCTGGG +CCCAAA +CCCAAT +CCCAAC +CCCAAG +CCCATA +CCCATT +CCCATC +CCCATG +CCCACA +CCCACT +CCCACC +CCCACG +CCCAGA +CCCAGT +CCCAGC +CCCAGG +CCCTAA +CCCTAT +CCCTAC +CCCTAG +CCCTTA +CCCTTT +CCCTTC +CCCTTG +CCCTCA +CCCTCT +CCCTCC +CCCTCG +CCCTGA +CCCTGT +CCCTGC +CCCTGG +CCCCAA +CCCCAT +CCCCAC +CCCCAG +CCCCTA +CCCCTT +CCCCTC +CCCCTG +CCCCCA +CCCCCT +CCCCCC +CCCCCG +CCCCGA +CCCCGT +CCCCGC +CCCCGG +CCCGAA +CCCGAT +CCCGAC +CCCGAG +CCCGTA +CCCGTT +CCCGTC +CCCGTG +CCCGCA +CCCGCT +CCCGCC +CCCGCG +CCCGGA +CCCGGT +CCCGGC +CCCGGG +CCGAAA +CCGAAT +CCGAAC +CCGAAG +CCGATA +CCGATT +CCGATC +CCGATG +CCGACA +CCGACT +CCGACC +CCGACG +CCGAGA +CCGAGT +CCGAGC +CCGAGG +CCGTAA +CCGTAT +CCGTAC +CCGTAG +CCGTTA +CCGTTT +CCGTTC +CCGTTG +CCGTCA +CCGTCT +CCGTCC +CCGTCG +CCGTGA +CCGTGT +CCGTGC +CCGTGG +CCGCAA +CCGCAT +CCGCAC +CCGCAG +CCGCTA +CCGCTT +CCGCTC +CCGCTG +CCGCCA +CCGCCT +CCGCCC +CCGCCG +CCGCGA +CCGCGT +CCGCGC +CCGCGG +CCGGAA +CCGGAT +CCGGAC +CCGGAG +CCGGTA +CCGGTT +CCGGTC +CCGGTG +CCGGCA +CCGGCT +CCGGCC +CCGGCG +CCGGGA +CCGGGT +CCGGGC +CCGGGG +CGAAAA +CGAAAT +CGAAAC +CGAAAG +CGAATA +CGAATT +CGAATC +CGAATG +CGAACA +CGAACT +CGAACC +CGAACG +CGAAGA +CGAAGT +CGAAGC +CGAAGG +CGATAA +CGATAT +CGATAC +CGATAG +CGATTA +CGATTT +CGATTC +CGATTG +CGATCA +CGATCT +CGATCC +CGATCG +CGATGA +CGATGT +CGATGC +CGATGG +CGACAA +CGACAT +CGACAC +CGACAG +CGACTA +CGACTT +CGACTC +CGACTG +CGACCA +CGACCT +CGACCC +CGACCG +CGACGA +CGACGT +CGACGC +CGACGG +CGAGAA +CGAGAT +CGAGAC +CGAGAG +CGAGTA +CGAGTT +CGAGTC +CGAGTG +CGAGCA +CGAGCT +CGAGCC +CGAGCG +CGAGGA +CGAGGT +CGAGGC +CGAGGG +CGTAAA +CGTAAT +CGTAAC +CGTAAG +CGTATA +CGTATT +CGTATC +CGTATG +CGTACA +CGTACT +CGTACC +CGTACG +CGTAGA +CGTAGT +CGTAGC +CGTAGG +CGTTAA +CGTTAT +CGTTAC +CGTTAG +CGTTTA +CGTTTT +CGTTTC +CGTTTG +CGTTCA +CGTTCT +CGTTCC +CGTTCG +CGTTGA +CGTTGT +CGTTGC +CGTTGG +CGTCAA +CGTCAT +CGTCAC +CGTCAG +CGTCTA +CGTCTT +CGTCTC +CGTCTG +CGTCCA +CGTCCT +CGTCCC +CGTCCG +CGTCGA +CGTCGT +CGTCGC +CGTCGG +CGTGAA +CGTGAT +CGTGAC +CGTGAG +CGTGTA +CGTGTT +CGTGTC +CGTGTG +CGTGCA +CGTGCT +CGTGCC +CGTGCG +CGTGGA +CGTGGT +CGTGGC +CGTGGG +CGCAAA +CGCAAT +CGCAAC +CGCAAG +CGCATA +CGCATT +CGCATC +CGCATG +CGCACA +CGCACT +CGCACC +CGCACG +CGCAGA +CGCAGT +CGCAGC +CGCAGG +CGCTAA +CGCTAT +CGCTAC +CGCTAG +CGCTTA +CGCTTT +CGCTTC +CGCTTG +CGCTCA +CGCTCT +CGCTCC +CGCTCG +CGCTGA +CGCTGT +CGCTGC +CGCTGG +CGCCAA +CGCCAT +CGCCAC +CGCCAG +CGCCTA +CGCCTT +CGCCTC +CGCCTG +CGCCCA +CGCCCT +CGCCCC +CGCCCG +CGCCGA +CGCCGT +CGCCGC +CGCCGG +CGCGAA +CGCGAT +CGCGAC +CGCGAG +CGCGTA +CGCGTT +CGCGTC +CGCGTG +CGCGCA +CGCGCT +CGCGCC +CGCGCG +CGCGGA +CGCGGT +CGCGGC +CGCGGG +CGGAAA +CGGAAT +CGGAAC +CGGAAG +CGGATA +CGGATT +CGGATC +CGGATG +CGGACA +CGGACT +CGGACC +CGGACG +CGGAGA +CGGAGT +CGGAGC +CGGAGG +CGGTAA +CGGTAT +CGGTAC +CGGTAG +CGGTTA +CGGTTT +CGGTTC +CGGTTG +CGGTCA +CGGTCT +CGGTCC +CGGTCG +CGGTGA +CGGTGT +CGGTGC +CGGTGG +CGGCAA +CGGCAT +CGGCAC +CGGCAG +CGGCTA +CGGCTT +CGGCTC +CGGCTG +CGGCCA +CGGCCT +CGGCCC +CGGCCG +CGGCGA +CGGCGT +CGGCGC +CGGCGG +CGGGAA +CGGGAT +CGGGAC +CGGGAG +CGGGTA +CGGGTT +CGGGTC +CGGGTG +CGGGCA +CGGGCT +CGGGCC +CGGGCG +CGGGGA +CGGGGT +CGGGGC +CGGGGG +GAAAAA +GAAAAT +GAAAAC +GAAAAG +GAAATA +GAAATT +GAAATC +GAAATG +GAAACA +GAAACT +GAAACC +GAAACG +GAAAGA +GAAAGT +GAAAGC +GAAAGG +GAATAA +GAATAT +GAATAC +GAATAG +GAATTA +GAATTT +GAATTC +GAATTG +GAATCA +GAATCT +GAATCC +GAATCG +GAATGA +GAATGT +GAATGC +GAATGG +GAACAA +GAACAT +GAACAC +GAACAG +GAACTA +GAACTT +GAACTC +GAACTG +GAACCA +GAACCT +GAACCC +GAACCG +GAACGA +GAACGT +GAACGC +GAACGG +GAAGAA +GAAGAT +GAAGAC +GAAGAG +GAAGTA +GAAGTT +GAAGTC +GAAGTG +GAAGCA +GAAGCT +GAAGCC +GAAGCG +GAAGGA +GAAGGT +GAAGGC +GAAGGG +GATAAA +GATAAT +GATAAC +GATAAG +GATATA +GATATT +GATATC +GATATG +GATACA +GATACT +GATACC +GATACG +GATAGA +GATAGT +GATAGC +GATAGG +GATTAA +GATTAT +GATTAC +GATTAG +GATTTA +GATTTT +GATTTC +GATTTG +GATTCA +GATTCT +GATTCC +GATTCG +GATTGA +GATTGT +GATTGC +GATTGG +GATCAA +GATCAT +GATCAC +GATCAG +GATCTA +GATCTT +GATCTC +GATCTG +GATCCA +GATCCT +GATCCC +GATCCG +GATCGA +GATCGT +GATCGC +GATCGG +GATGAA +GATGAT +GATGAC +GATGAG +GATGTA +GATGTT +GATGTC +GATGTG +GATGCA +GATGCT +GATGCC +GATGCG +GATGGA +GATGGT +GATGGC +GATGGG +GACAAA +GACAAT +GACAAC +GACAAG +GACATA +GACATT +GACATC +GACATG +GACACA +GACACT +GACACC +GACACG +GACAGA +GACAGT +GACAGC +GACAGG +GACTAA +GACTAT +GACTAC +GACTAG +GACTTA +GACTTT +GACTTC +GACTTG +GACTCA +GACTCT +GACTCC +GACTCG +GACTGA +GACTGT +GACTGC +GACTGG +GACCAA +GACCAT +GACCAC +GACCAG +GACCTA +GACCTT +GACCTC +GACCTG +GACCCA +GACCCT +GACCCC +GACCCG +GACCGA +GACCGT +GACCGC +GACCGG +GACGAA +GACGAT +GACGAC +GACGAG +GACGTA +GACGTT +GACGTC +GACGTG +GACGCA +GACGCT +GACGCC +GACGCG +GACGGA +GACGGT +GACGGC +GACGGG +GAGAAA +GAGAAT +GAGAAC +GAGAAG +GAGATA +GAGATT +GAGATC +GAGATG +GAGACA +GAGACT +GAGACC +GAGACG +GAGAGA +GAGAGT +GAGAGC +GAGAGG +GAGTAA +GAGTAT +GAGTAC +GAGTAG +GAGTTA +GAGTTT +GAGTTC +GAGTTG +GAGTCA +GAGTCT +GAGTCC +GAGTCG +GAGTGA +GAGTGT +GAGTGC +GAGTGG +GAGCAA +GAGCAT +GAGCAC +GAGCAG +GAGCTA +GAGCTT +GAGCTC +GAGCTG +GAGCCA +GAGCCT +GAGCCC +GAGCCG +GAGCGA +GAGCGT +GAGCGC +GAGCGG +GAGGAA +GAGGAT +GAGGAC +GAGGAG +GAGGTA +GAGGTT +GAGGTC +GAGGTG +GAGGCA +GAGGCT +GAGGCC +GAGGCG +GAGGGA +GAGGGT +GAGGGC +GAGGGG +GTAAAA +GTAAAT +GTAAAC +GTAAAG +GTAATA +GTAATT +GTAATC +GTAATG +GTAACA +GTAACT +GTAACC +GTAACG +GTAAGA +GTAAGT +GTAAGC +GTAAGG +GTATAA +GTATAT +GTATAC +GTATAG +GTATTA +GTATTT +GTATTC +GTATTG +GTATCA +GTATCT +GTATCC +GTATCG +GTATGA +GTATGT +GTATGC +GTATGG +GTACAA +GTACAT +GTACAC +GTACAG +GTACTA +GTACTT +GTACTC +GTACTG +GTACCA +GTACCT +GTACCC +GTACCG +GTACGA +GTACGT +GTACGC +GTACGG +GTAGAA +GTAGAT +GTAGAC +GTAGAG +GTAGTA +GTAGTT +GTAGTC +GTAGTG +GTAGCA +GTAGCT +GTAGCC +GTAGCG +GTAGGA +GTAGGT +GTAGGC +GTAGGG +GTTAAA +GTTAAT +GTTAAC +GTTAAG +GTTATA +GTTATT +GTTATC +GTTATG +GTTACA +GTTACT +GTTACC +GTTACG +GTTAGA +GTTAGT +GTTAGC +GTTAGG +GTTTAA +GTTTAT +GTTTAC +GTTTAG +GTTTTA +GTTTTT +GTTTTC +GTTTTG +GTTTCA +GTTTCT +GTTTCC +GTTTCG +GTTTGA +GTTTGT +GTTTGC +GTTTGG +GTTCAA +GTTCAT +GTTCAC +GTTCAG +GTTCTA +GTTCTT +GTTCTC +GTTCTG +GTTCCA +GTTCCT +GTTCCC +GTTCCG +GTTCGA +GTTCGT +GTTCGC +GTTCGG +GTTGAA +GTTGAT +GTTGAC +GTTGAG +GTTGTA +GTTGTT +GTTGTC +GTTGTG +GTTGCA +GTTGCT +GTTGCC +GTTGCG +GTTGGA +GTTGGT +GTTGGC +GTTGGG +GTCAAA +GTCAAT +GTCAAC +GTCAAG +GTCATA +GTCATT +GTCATC +GTCATG +GTCACA +GTCACT +GTCACC +GTCACG +GTCAGA +GTCAGT +GTCAGC +GTCAGG +GTCTAA +GTCTAT +GTCTAC +GTCTAG +GTCTTA +GTCTTT +GTCTTC +GTCTTG +GTCTCA +GTCTCT +GTCTCC +GTCTCG +GTCTGA +GTCTGT +GTCTGC +GTCTGG +GTCCAA +GTCCAT +GTCCAC +GTCCAG +GTCCTA +GTCCTT +GTCCTC +GTCCTG +GTCCCA +GTCCCT +GTCCCC +GTCCCG +GTCCGA +GTCCGT +GTCCGC +GTCCGG +GTCGAA +GTCGAT +GTCGAC +GTCGAG +GTCGTA +GTCGTT +GTCGTC +GTCGTG +GTCGCA +GTCGCT +GTCGCC +GTCGCG +GTCGGA +GTCGGT +GTCGGC +GTCGGG +GTGAAA +GTGAAT +GTGAAC +GTGAAG +GTGATA +GTGATT +GTGATC +GTGATG +GTGACA +GTGACT +GTGACC +GTGACG +GTGAGA +GTGAGT +GTGAGC +GTGAGG +GTGTAA +GTGTAT +GTGTAC +GTGTAG +GTGTTA +GTGTTT +GTGTTC +GTGTTG +GTGTCA +GTGTCT +GTGTCC +GTGTCG +GTGTGA +GTGTGT +GTGTGC +GTGTGG +GTGCAA +GTGCAT +GTGCAC +GTGCAG +GTGCTA +GTGCTT +GTGCTC +GTGCTG +GTGCCA +GTGCCT +GTGCCC +GTGCCG +GTGCGA +GTGCGT +GTGCGC +GTGCGG +GTGGAA +GTGGAT +GTGGAC +GTGGAG +GTGGTA +GTGGTT +GTGGTC +GTGGTG +GTGGCA +GTGGCT +GTGGCC +GTGGCG +GTGGGA +GTGGGT +GTGGGC +GTGGGG +GCAAAA +GCAAAT +GCAAAC +GCAAAG +GCAATA +GCAATT +GCAATC +GCAATG +GCAACA +GCAACT +GCAACC +GCAACG +GCAAGA +GCAAGT +GCAAGC +GCAAGG +GCATAA +GCATAT +GCATAC +GCATAG +GCATTA +GCATTT +GCATTC +GCATTG +GCATCA +GCATCT +GCATCC +GCATCG +GCATGA +GCATGT +GCATGC +GCATGG +GCACAA +GCACAT +GCACAC +GCACAG +GCACTA +GCACTT +GCACTC +GCACTG +GCACCA +GCACCT +GCACCC +GCACCG +GCACGA +GCACGT +GCACGC +GCACGG +GCAGAA +GCAGAT +GCAGAC +GCAGAG +GCAGTA +GCAGTT +GCAGTC +GCAGTG +GCAGCA +GCAGCT +GCAGCC +GCAGCG +GCAGGA +GCAGGT +GCAGGC +GCAGGG +GCTAAA +GCTAAT +GCTAAC +GCTAAG +GCTATA +GCTATT +GCTATC +GCTATG +GCTACA +GCTACT +GCTACC +GCTACG +GCTAGA +GCTAGT +GCTAGC +GCTAGG +GCTTAA +GCTTAT +GCTTAC +GCTTAG +GCTTTA +GCTTTT +GCTTTC +GCTTTG +GCTTCA +GCTTCT +GCTTCC +GCTTCG +GCTTGA +GCTTGT +GCTTGC +GCTTGG +GCTCAA +GCTCAT +GCTCAC +GCTCAG +GCTCTA +GCTCTT +GCTCTC +GCTCTG +GCTCCA +GCTCCT +GCTCCC +GCTCCG +GCTCGA +GCTCGT +GCTCGC +GCTCGG +GCTGAA +GCTGAT +GCTGAC +GCTGAG +GCTGTA +GCTGTT +GCTGTC +GCTGTG +GCTGCA +GCTGCT +GCTGCC +GCTGCG +GCTGGA +GCTGGT +GCTGGC +GCTGGG +GCCAAA +GCCAAT +GCCAAC +GCCAAG +GCCATA +GCCATT +GCCATC +GCCATG +GCCACA +GCCACT +GCCACC +GCCACG +GCCAGA +GCCAGT +GCCAGC +GCCAGG +GCCTAA +GCCTAT +GCCTAC +GCCTAG +GCCTTA +GCCTTT +GCCTTC +GCCTTG +GCCTCA +GCCTCT +GCCTCC +GCCTCG +GCCTGA +GCCTGT +GCCTGC +GCCTGG +GCCCAA +GCCCAT +GCCCAC +GCCCAG +GCCCTA +GCCCTT +GCCCTC +GCCCTG +GCCCCA +GCCCCT +GCCCCC +GCCCCG +GCCCGA +GCCCGT +GCCCGC +GCCCGG +GCCGAA +GCCGAT +GCCGAC +GCCGAG +GCCGTA +GCCGTT +GCCGTC +GCCGTG +GCCGCA +GCCGCT +GCCGCC +GCCGCG +GCCGGA +GCCGGT +GCCGGC +GCCGGG +GCGAAA +GCGAAT +GCGAAC +GCGAAG +GCGATA +GCGATT +GCGATC +GCGATG +GCGACA +GCGACT +GCGACC +GCGACG +GCGAGA +GCGAGT +GCGAGC +GCGAGG +GCGTAA +GCGTAT +GCGTAC +GCGTAG +GCGTTA +GCGTTT +GCGTTC +GCGTTG +GCGTCA +GCGTCT +GCGTCC +GCGTCG +GCGTGA +GCGTGT +GCGTGC +GCGTGG +GCGCAA +GCGCAT +GCGCAC +GCGCAG +GCGCTA +GCGCTT +GCGCTC +GCGCTG +GCGCCA +GCGCCT +GCGCCC +GCGCCG +GCGCGA +GCGCGT +GCGCGC +GCGCGG +GCGGAA +GCGGAT +GCGGAC +GCGGAG +GCGGTA +GCGGTT +GCGGTC +GCGGTG +GCGGCA +GCGGCT +GCGGCC +GCGGCG +GCGGGA +GCGGGT +GCGGGC +GCGGGG +GGAAAA +GGAAAT +GGAAAC +GGAAAG +GGAATA +GGAATT +GGAATC +GGAATG +GGAACA +GGAACT +GGAACC +GGAACG +GGAAGA +GGAAGT +GGAAGC +GGAAGG +GGATAA +GGATAT +GGATAC +GGATAG +GGATTA +GGATTT +GGATTC +GGATTG +GGATCA +GGATCT +GGATCC +GGATCG +GGATGA +GGATGT +GGATGC +GGATGG +GGACAA +GGACAT +GGACAC +GGACAG +GGACTA +GGACTT +GGACTC +GGACTG +GGACCA +GGACCT +GGACCC +GGACCG +GGACGA +GGACGT +GGACGC +GGACGG +GGAGAA +GGAGAT +GGAGAC +GGAGAG +GGAGTA +GGAGTT +GGAGTC +GGAGTG +GGAGCA +GGAGCT +GGAGCC +GGAGCG +GGAGGA +GGAGGT +GGAGGC +GGAGGG +GGTAAA +GGTAAT +GGTAAC +GGTAAG +GGTATA +GGTATT +GGTATC +GGTATG +GGTACA +GGTACT +GGTACC +GGTACG +GGTAGA +GGTAGT +GGTAGC +GGTAGG +GGTTAA +GGTTAT +GGTTAC +GGTTAG +GGTTTA +GGTTTT +GGTTTC +GGTTTG +GGTTCA +GGTTCT +GGTTCC +GGTTCG +GGTTGA +GGTTGT +GGTTGC +GGTTGG +GGTCAA +GGTCAT +GGTCAC +GGTCAG +GGTCTA +GGTCTT +GGTCTC +GGTCTG +GGTCCA +GGTCCT +GGTCCC +GGTCCG +GGTCGA +GGTCGT +GGTCGC +GGTCGG +GGTGAA +GGTGAT +GGTGAC +GGTGAG +GGTGTA +GGTGTT +GGTGTC +GGTGTG +GGTGCA +GGTGCT +GGTGCC +GGTGCG +GGTGGA +GGTGGT +GGTGGC +GGTGGG +GGCAAA +GGCAAT +GGCAAC +GGCAAG +GGCATA +GGCATT +GGCATC +GGCATG +GGCACA +GGCACT +GGCACC +GGCACG +GGCAGA +GGCAGT +GGCAGC +GGCAGG +GGCTAA +GGCTAT +GGCTAC +GGCTAG +GGCTTA +GGCTTT +GGCTTC +GGCTTG +GGCTCA +GGCTCT +GGCTCC +GGCTCG +GGCTGA +GGCTGT +GGCTGC +GGCTGG +GGCCAA +GGCCAT +GGCCAC +GGCCAG +GGCCTA +GGCCTT +GGCCTC +GGCCTG +GGCCCA +GGCCCT +GGCCCC +GGCCCG +GGCCGA +GGCCGT +GGCCGC +GGCCGG +GGCGAA +GGCGAT +GGCGAC +GGCGAG +GGCGTA +GGCGTT +GGCGTC +GGCGTG +GGCGCA +GGCGCT +GGCGCC +GGCGCG +GGCGGA +GGCGGT +GGCGGC +GGCGGG +GGGAAA +GGGAAT +GGGAAC +GGGAAG +GGGATA +GGGATT +GGGATC +GGGATG +GGGACA +GGGACT +GGGACC +GGGACG +GGGAGA +GGGAGT +GGGAGC +GGGAGG +GGGTAA +GGGTAT +GGGTAC +GGGTAG +GGGTTA +GGGTTT +GGGTTC +GGGTTG +GGGTCA +GGGTCT +GGGTCC +GGGTCG +GGGTGA +GGGTGT +GGGTGC +GGGTGG +GGGCAA +GGGCAT +GGGCAC +GGGCAG +GGGCTA +GGGCTT +GGGCTC +GGGCTG +GGGCCA +GGGCCT +GGGCCC +GGGCCG +GGGCGA +GGGCGT +GGGCGC +GGGCGG +GGGGAA +GGGGAT +GGGGAC +GGGGAG +GGGGTA +GGGGTT +GGGGTC +GGGGTG +GGGGCA +GGGGCT +GGGGCC +GGGGCG +GGGGGA +GGGGGT +GGGGGC +GGGGGG diff --git a/Finetuned Models/r-experiment.ipynb b/Finetuned Models/r-experiment.ipynb new file mode 100644 index 0000000000000000000000000000000000000000..c650ab62ea9953d7f3c073907940d70ea632e56f --- /dev/null +++ b/Finetuned Models/r-experiment.ipynb @@ -0,0 +1,8587 @@ +{ + "cells": [ + { + "cell_type": "code", + "execution_count": null, + "id": "ca22d12d", + "metadata": { + "execution": { + "iopub.execute_input": "2024-08-14T05:42:36.377550Z", + "iopub.status.busy": "2024-08-14T05:42:36.376734Z", + "iopub.status.idle": "2024-08-14T05:44:36.924929Z", + "shell.execute_reply": "2024-08-14T05:44:36.923740Z" + }, + "papermill": { + "duration": 120.554736, + "end_time": "2024-08-14T05:44:36.927333", + "exception": false, + "start_time": "2024-08-14T05:42:36.372597", + "status": "completed" + }, + "tags": [] + }, + "outputs": [], + "source": [ + "!pip install -q --upgrade datasets\n", + "!pip install -q --upgrade transformers\n", + "!pip install -q --upgrade peft\n", + "!pip install -q --upgrade trl\n", + "!pip install -q bitsandbytes\n", + "!pip install -q accelerate\n", + "!pip install -q tensorboard\n", + "!pip install -q jsonlines\n", + "! conda install -y gdown" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "a600d3e0", + "metadata": { + "execution": { + "iopub.execute_input": "2024-08-14T05:46:24.563834Z", + "iopub.status.busy": "2024-08-14T05:46:24.563522Z", + "iopub.status.idle": "2024-08-14T06:37:53.094328Z", + "shell.execute_reply": "2024-08-14T06:37:53.093386Z" + }, + "papermill": { + "duration": 3088.552597, + "end_time": "2024-08-14T06:37:53.096673", + "exception": false, + "start_time": "2024-08-14T05:46:24.544076", + "status": "completed" + }, + "tags": [] + }, + "outputs": [], + "source": [ + "import itertools\n", + "import time\n", + "import warnings\n", + "from peft import LoraConfig, get_peft_model\n", + "from transformers import BertForSequenceClassification, AutoTokenizer, TrainingArguments, Trainer\n", + "from matplotlib import pyplot as plt\n", + "from datasets import load_dataset\n", + "import torch\n", + "from tqdm import tqdm\n", + "import numpy as np\n", + "from sklearn.metrics import accuracy_score, recall_score, precision_score, f1_score, matthews_corrcoef, roc_auc_score\n", + "\n", + "# Suppress warnings\n", + "warnings.filterwarnings(\"ignore\")\n", + "\n", + "# Layer configurations\n", + "attention_plus_feed_forward = [\n", + " \"bert.encoder.layer.0.attention.self.query\",\n", + " \"bert.encoder.layer.0.attention.self.key\",\n", + " \"bert.encoder.layer.0.attention.self.value\",\n", + " \"bert.encoder.layer.0.attention.output.dense\",\n", + " \"bert.encoder.layer.0.intermediate.dense\",\n", + " \"bert.encoder.layer.0.output.dense\",\n", + " \"bert.encoder.layer.1.attention.self.query\",\n", + " \"bert.encoder.layer.1.attention.self.key\",\n", + " \"bert.encoder.layer.1.attention.self.value\",\n", + " \"bert.encoder.layer.1.attention.output.dense\",\n", + " \"bert.encoder.layer.1.intermediate.dense\",\n", + " \"bert.encoder.layer.1.output.dense\"\n", + "]\n", + "\n", + "attention = [\n", + " \"bert.encoder.layer.0.attention.self.query\",\n", + " \"bert.encoder.layer.0.attention.self.key\",\n", + " \"bert.encoder.layer.0.attention.self.value\",\n", + " \"bert.encoder.layer.0.attention.output.dense\",\n", + " \"bert.encoder.layer.1.attention.self.query\",\n", + " \"bert.encoder.layer.1.attention.self.key\",\n", + " \"bert.encoder.layer.1.attention.self.value\",\n", + " \"bert.encoder.layer.1.attention.output.dense\"\n", + "]\n", + "\n", + "feed_forward = [\n", + " \"bert.encoder.layer.0.intermediate.dense\",\n", + " \"bert.encoder.layer.0.output.dense\",\n", + " \"bert.encoder.layer.1.intermediate.dense\",\n", + " \"bert.encoder.layer.1.output.dense\"\n", + "]\n", + "\n", + "all_layer_configs = [attention_plus_feed_forward, attention, feed_forward]\n", + "tokenizer = AutoTokenizer.from_pretrained('zhihan1996/DNA_bert_6')\n", + "\n", + "def preprocess_function(examples):\n", + " try:\n", + " return tokenizer(\n", + " examples['sequence'],\n", + " padding='max_length',\n", + " truncation=True,\n", + " max_length=512\n", + " )\n", + " except KeyError:\n", + " return tokenizer(\n", + " examples['Sequence'],\n", + " padding='max_length',\n", + " truncation=True,\n", + " max_length=512\n", + " )\n", + "\n", + "\n", + "def add_labels(examples):\n", + " try:\n", + " examples['labels'] = examples['label']\n", + " return examples\n", + " except KeyError:\n", + " examples['labels'] = examples['Label']\n", + " return examples\n", + "\n", + "def create_task_dataset(task_name):\n", + " if task_name == 'tfbs':\n", + " return load_dataset('csv', data_files='/kaggle/working/tfbs.csv', split='train[0:10000]'), load_dataset('csv', data_files='/kaggle/working/tfbs.csv', split='train[10001:13122]')\n", + " elif task_name == 'dnasplice':\n", + " return load_dataset('csv', data_files='/kaggle/working/dnasplice.csv', split='train[0:10000]'), load_dataset('csv', data_files='/kaggle/working/dnasplice.csv', split='train[10001:13122]')\n", + " elif task_name == 'dnaprom':\n", + " return load_dataset('csv', data_files='/kaggle/working/dnaprom.csv', split='train[0:10000]'), load_dataset('csv', data_files='/kaggle/working/dnaprom.csv', split='train[10001:13122]')\n", + " else:\n", + " raise ValueError(f\"Unknown task: {task_name}\")\n", + "\n", + "def create_dataset_maps(train_dataset, test_dataset):\n", + " train_dataset = train_dataset.map(preprocess_function, batched=True)\n", + " train_dataset = train_dataset.map(add_labels)\n", + " test_dataset = test_dataset.map(preprocess_function, batched=True)\n", + " test_dataset = test_dataset.map(add_labels)\n", + " return train_dataset, test_dataset\n", + "\n", + "def train_model(train_dataset, test_dataset, model, task, model_name, config_name):\n", + " def specificity_score(y_true, y_pred):\n", + " true_negatives = np.sum((y_pred == 0) & (y_true == 0))\n", + " false_positives = np.sum((y_pred == 1) & (y_true == 0))\n", + " specificity = true_negatives / (true_negatives + false_positives + np.finfo(float).eps)\n", + " return specificity\n", + "\n", + " def compute_metrics(eval_pred):\n", + " logits, labels = eval_pred\n", + " predictions = np.argmax(logits, axis=-1)\n", + " y_pred = logits[:, 1]\n", + "\n", + " accuracy = accuracy_score(labels, predictions)\n", + " recall = recall_score(labels, predictions)\n", + " specificity = specificity_score(labels, predictions)\n", + " mcc = matthews_corrcoef(labels, predictions)\n", + " roc_auc = roc_auc_score(labels, y_pred)\n", + " precision = precision_score(labels, predictions)\n", + " f1 = f1_score(labels, predictions)\n", + "\n", + " return {\n", + " 'accuracy': accuracy,\n", + " 'recall': recall,\n", + " 'specificity': specificity,\n", + " 'mcc': mcc,\n", + " 'roc_auc': roc_auc,\n", + " 'precision': precision,\n", + " 'f1': f1\n", + " }\n", + "\n", + " # Define the training arguments\n", + " training_arguments = TrainingArguments(\n", + " output_dir=f\"outputs/{task}/{model_name}_{config_name}\",\n", + " num_train_epochs=3,\n", + " fp16=False,\n", + " bf16=False,\n", + " per_device_train_batch_size=20,\n", + " per_device_eval_batch_size=10,\n", + " gradient_accumulation_steps=2,\n", + " gradient_checkpointing=True,\n", + " max_grad_norm=0.3,\n", + " learning_rate=4e-4,\n", + " weight_decay=0.01,\n", + " optim=\"paged_adamw_32bit\",\n", + " lr_scheduler_type=\"linear\",\n", + " max_steps=-1,\n", + " warmup_ratio=0.03,\n", + " group_by_length=True,\n", + " save_steps=10,\n", + " logging_steps=25,\n", + " dataloader_pin_memory=False,\n", + " report_to='tensorboard',\n", + " gradient_checkpointing_kwargs={'use_reentrant': False}\n", + " )\n", + "\n", + " trainer = Trainer(\n", + " model=model,\n", + " args=training_arguments,\n", + " train_dataset=train_dataset,\n", + " eval_dataset=test_dataset,\n", + " tokenizer=tokenizer,\n", + " compute_metrics=compute_metrics,\n", + " )\n", + "\n", + " start_time = time.time()\n", + " trainer.train()\n", + " end_time = time.time()\n", + "\n", + " total_time = end_time - start_time\n", + " metrics = trainer.evaluate()\n", + "\n", + " return total_time, metrics\n", + "\n", + "def estimate_model_size_lora(model):\n", + " num_parameters = model.get_nb_trainable_parameters()[0]\n", + " size_in_bytes = num_parameters * 4 # Each parameter is 4 bytes (32-bit float)\n", + " size_in_mb = size_in_bytes / (1024 ** 2) # Convert bytes to megabytes\n", + " return size_in_mb\n", + "def estimate_model_size_base(model):\n", + " num_parameters = sum(p.numel() for p in model.parameters() if p.requires_grad)\n", + "\n", + " size_in_bytes = num_parameters * 4 # Each parameter is 4 bytes (32-bit float)\n", + " size_in_mb = size_in_bytes / (1024 ** 2) # Convert bytes to megabytes\n", + " return size_in_mb\n", + "# Task loop\n", + "task_list = ['dnasplice', 'tfbs', 'dnaprom']\n", + "log_file = \"training_log.txt\"\n", + "model_name = 'fabihamakhdoomi/TinyDNABERT'\n", + "r_values = [4, 8, 16, 32] # Different r values for ablation study\n", + "\n", + "results = []\n", + "\n", + "for task in task_list:\n", + " print(f\"Running TASK : {task}\")\n", + " train_dataset, test_dataset = create_task_dataset(task)\n", + " train_dataset, test_dataset = create_dataset_maps(train_dataset, test_dataset) \n", + " train_dataset.set_format(type='torch', columns=['input_ids', 'attention_mask', 'labels'])\n", + " test_dataset.set_format(type='torch', columns=['input_ids', 'attention_mask', 'labels'])\n", + " \n", + "\n", + "\n", + " # Train the base model first\n", + " base_model = BertForSequenceClassification.from_pretrained(\n", + " model_name,\n", + " num_labels=2\n", + " )\n", + " config_name = \"base_model\"\n", + " print(f\"Training MODEL : {config_name} for task : {task}\")\n", + " training_time, metrics = train_model(train_dataset, test_dataset, base_model, task, model_name, config_name)\n", + " model_size_mb = estimate_model_size_base(base_model)\n", + " results.append((task, config_name, model_size_mb, metrics['eval_accuracy']))\n", + " with open(log_file, \"a\") as log:\n", + " log.write(f\"Task: {task}, Model: {model_name}, Config: {config_name}, Training Time: {training_time}, Model Size: {model_size_mb} MB, Metrics: {metrics}\\n\")\n", + "\n", + " # Train the LoRA models with different r values\n", + " for r in r_values:\n", + " for config_name, config in zip(['attention_plus_feed_forward', 'attention', 'feed_forward'], all_layer_configs):\n", + " base_model = BertForSequenceClassification.from_pretrained(\n", + " model_name,\n", + " num_labels=2\n", + " )\n", + " peft_config = LoraConfig(\n", + " lora_alpha=16,\n", + " lora_dropout=0.2,\n", + " r=r,\n", + " bias=\"none\",\n", + " task_type=\"SEQ_CLS\",\n", + " target_modules=config\n", + " )\n", + " model = get_peft_model(base_model, peft_config)\n", + " print(f\"Training MODEL : {config_name} with r={r} for task : {task}\")\n", + " training_time, metrics = train_model(train_dataset, test_dataset, model, task, model_name, f\"{config_name}_r{r}\")\n", + " model_size_mb = estimate_model_size_lora(model)\n", + " results.append((task, f\"{config_name}_r{r}\", model_size_mb, metrics['eval_accuracy']))\n", + " with open(log_file, \"a\") as log:\n", + " log.write(f\"Task: {task}, Model: {model_name}, Config: {config_name}, r: {r}, Training Time: {training_time}, Model Size: {model_size_mb} MB, Metrics: {metrics}\\n\")\n", + "\n", + "# Results logging and visualization\n", + "tasks = [result[0] for result in results]\n", + "configs = [result[1] for result in results]\n", + "model_sizes = [result[2] for result in results]\n", + "accuracies = [result[3] for result in results]\n", + "\n", + "for i, task in enumerate(task_list):\n", + " plt.figure(figsize=(10, 6))\n", + " task_results = [results[j] for j in range(len(results)) if results[j][0] == task]\n", + " configs = [r[1] for r in task_results]\n", + " sizes = [r[2] for r in task_results]\n", + " accuracies = [r[3] for r in task_results]\n", + " plt.plot(sizes, accuracies, marker='o')\n", + " plt.xlabel('Model Size (MB)')\n", + " plt.ylabel('Accuracy')\n", + " plt.title(f'Model Size vs Accuracy for Task: {task}')\n", + " plt.xticks(sizes, configs, rotation=45, ha=\"right\")\n", + " plt.tight_layout()\n", + " plt.savefig(f'{task}_model_size_vs_accuracy.png')\n", + " plt.show()\n" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "9eaa7315", + "metadata": { + "papermill": { + "duration": 0.047629, + "end_time": "2024-08-14T06:37:53.193070", + "exception": false, + "start_time": "2024-08-14T06:37:53.145441", + "status": "completed" + }, + "tags": [] + }, + "outputs": [], + "source": [] + } + ], + "metadata": { + "kaggle": { + "accelerator": "gpu", + "dataSources": [], + "dockerImageVersionId": 30747, + "isGpuEnabled": true, + "isInternetEnabled": true, + "language": "python", + "sourceType": "notebook" + }, + "kernelspec": { + "display_name": "Python 3", + "language": "python", + "name": "python3" + }, + "language_info": { + "codemirror_mode": { + "name": "ipython", + "version": 3 + }, + "file_extension": ".py", + "mimetype": "text/x-python", + "name": "python", + "nbconvert_exporter": "python", + "pygments_lexer": "ipython3", + "version": "3.10.13" + }, + "papermill": { + "default_parameters": {}, + "duration": 3323.641972, + "end_time": "2024-08-14T06:37:56.402106", + "environment_variables": {}, + "exception": null, + "input_path": "__notebook__.ipynb", + "output_path": "__notebook__.ipynb", + "parameters": {}, + "start_time": "2024-08-14T05:42:32.760134", + "version": "2.5.0" + }, + "widgets": { + "application/vnd.jupyter.widget-state+json": { + "state": { + "00284f3a88e84a98949802e9b37697fb": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_0cb49f42604f46c1b35d947d57f2a181", + "placeholder": "​", + "style": "IPY_MODEL_7c8d3f5b085e469c8d7362fb67aa460f", + "value": " 1.45k/1.45k [00:00<00:00, 143kB/s]" + } + }, + "016cb0c3de1d4564bbf2b98b35911028": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HBoxModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HBoxModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HBoxView", + "box_style": "", + "children": [ + "IPY_MODEL_75e195ea430644df819c075580bfc8d5", + "IPY_MODEL_5ae7e1e397414ee88f444f73def2ef94", + "IPY_MODEL_00284f3a88e84a98949802e9b37697fb" + ], + "layout": "IPY_MODEL_65733b2322f1448e9867248f3199fb91" + } + }, + "02c62a2bcc8a4631bd83ee770983c225": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "049d5f2b43a34021b6eebdda28684f87": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "05cb827adaa74c2d9243709693be0cf2": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "ProgressStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "ProgressStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "bar_color": null, + "description_width": "" + } + }, + "07e7de84584c4c3d865ff597c072e9a7": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "086ae04ec69348a0ba3f7a89e83ec964": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "08f3e75613d24774b5f885a2bfce13dd": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "ProgressStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "ProgressStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "bar_color": null, + "description_width": "" + } + }, + "094a97b9393541e8bc11f85a7c7767c5": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "ProgressStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "ProgressStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "bar_color": null, + "description_width": "" + } + }, + "0b74a82e4ded40069bac382704222978": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "0b9c9c92099f462bbb0151f182d4e129": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HBoxModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HBoxModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HBoxView", + "box_style": "", + "children": [ + "IPY_MODEL_aa6d8339818549e78d10dc5d3ae11b12", + "IPY_MODEL_c9453a6e52b34ab79a7781591b703d46", + "IPY_MODEL_e7ec7b587a1c44a1ad68105be1f2a15e" + ], + "layout": "IPY_MODEL_46e1dc5b0edf459bbcb82091e64efa1e" + } + }, + "0cb49f42604f46c1b35d947d57f2a181": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "0cdfd05817854fe78db65b031e8ae1a9": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "FloatProgressModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "FloatProgressModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "ProgressView", + "bar_style": "success", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_c5ea3c6735c5448fbe032c86b211cd11", + "max": 112, + "min": 0, + "orientation": "horizontal", + "style": "IPY_MODEL_f86cd40e74a349dfa206ec4f8122d989", + "value": 112 + } + }, + "0e445bbc266a48d1996276277227c962": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "0fc716f771b949468823704c4bd4af0f": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_d9475d52fa814601a1b7f33b5541cd70", + "placeholder": "​", + "style": "IPY_MODEL_d989263ed1144a5ca2d00c78873d4448", + "value": " 3121/3121 [00:01<00:00, 2370.30 examples/s]" + } + }, + "10e1bf80fdb24a768f896bed74057532": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_d59bbd5a61d64b71883674292dc02782", + "placeholder": "​", + "style": "IPY_MODEL_4b854c733a8d48a9896c7769eb54469a", + "value": " 10000/10000 [00:04<00:00, 2174.16 examples/s]" + } + }, + "11251509b3424421bfbcd7db7e908cb0": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "11a2cbce00f04fa685f070c4e6b7c974": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HBoxModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HBoxModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HBoxView", + "box_style": "", + "children": [ + "IPY_MODEL_404b48dedcc84e28a4ca5a5a3c3c2752", + "IPY_MODEL_7c2f51169b324b208bee733f2160f36d", + "IPY_MODEL_1be800b1e1dd46a2b9c33bd65a061ab1" + ], + "layout": "IPY_MODEL_589e4554f9d444339544672614b64323" + } + }, + "11abd35807054073be7859c8dbe62d26": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HBoxModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HBoxModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HBoxView", + "box_style": "", + "children": [ + "IPY_MODEL_dbb724416bf84af2a1efd3c2387f1aa3", + "IPY_MODEL_f3a6678289b74fc0bf55df6b2fc6013e", + "IPY_MODEL_14f9293086c648e6a4ebf1747627ff01" + ], + "layout": "IPY_MODEL_bdcfb933166e469b8cbabc41235da058" + } + }, + "14f9293086c648e6a4ebf1747627ff01": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_6b0d3a846fcb4fd2a505684b31bbe37d", + "placeholder": "​", + "style": "IPY_MODEL_71ce2b44cc4448b9af944f85491783c3", + "value": " 4.04M/4.04M [00:00<00:00, 63.3MB/s]" + } + }, + "15416fe7a4de428a868538c086f09218": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "FloatProgressModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "FloatProgressModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "ProgressView", + "bar_style": "success", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_07e7de84584c4c3d865ff597c072e9a7", + "max": 3121, + "min": 0, + "orientation": "horizontal", + "style": "IPY_MODEL_08f3e75613d24774b5f885a2bfce13dd", + "value": 3121 + } + }, + "18576edbee874ba3b79e56fc65089a76": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "1876c30dc908494197724455f0ff7a91": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_086ae04ec69348a0ba3f7a89e83ec964", + "placeholder": "​", + "style": "IPY_MODEL_9223130ecf8c42b8af600cda70cd7cee", + "value": " 10000/10000 [00:01<00:00, 10627.91 examples/s]" + } + }, + "18ff3c7a0a1d4595b4a7ff941c0c453b": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "190ee28f166743a5952264de5ab458bf": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_b1d8cb19f86f4dfbbc4571376927255e", + "placeholder": "​", + "style": "IPY_MODEL_635496dc27774b3e8fdf3c6e4d4b5c12", + "value": "Map: 100%" + } + }, + "1958e6dcd6d14fba8d568eab0844dcc3": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "1be800b1e1dd46a2b9c33bd65a061ab1": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_c989f464dc2b43bc8e77b74bad170f13", + "placeholder": "​", + "style": "IPY_MODEL_c69745397340449f8f8fe24c94dd9f29", + "value": " 1/1 [00:00<00:00, 120.73it/s]" + } + }, + "1c903f797329485d94a584d5db1e73dc": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "1ca5b9b7b0c94f1d827eb99b0e4a7105": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "FloatProgressModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "FloatProgressModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "ProgressView", + "bar_style": "success", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_2ec0a5f561964ec2bb25ca16ab7a80fd", + "max": 10000, + "min": 0, + "orientation": "horizontal", + "style": "IPY_MODEL_2e9d4417e38042ee943f65ed9b2219c4", + "value": 10000 + } + }, + "1edbbb81b31b4296bf1c4366bb4d6fee": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "21d2c65103194be092471637c2faada4": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "22a91421543b4516a97e6edba97e1b59": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HBoxModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HBoxModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HBoxView", + "box_style": "", + "children": [ + "IPY_MODEL_69edd97bdba44434af8c1d705123b695", + "IPY_MODEL_b69cb7e6ad7141679f9e8fdc04e39446", + "IPY_MODEL_9519dbb2af464e7487409a78fc4afa9f" + ], + "layout": "IPY_MODEL_7cc99418d6254587b8082f5421293b56" + } + }, + "22aab510900f4b7792862675c8d0b497": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "22aec26a870a4c7f8026f1c7a8004c38": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "22bca1b1ece94d0cab713ef13be82e48": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_b36ee6a68e7047afae8ca921e806430e", + "placeholder": "​", + "style": "IPY_MODEL_c38e40fc8bc846eab8d6fb971fe4858a", + "value": "vocab.txt: 100%" + } + }, + "25f9b192b92d41d79440924079e347e1": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "26427932b7584edda94d1c2c95cb4eab": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "2790a55362c84e30a489f3ea144f019f": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "27f04577de04467f9f6e81746de9e659": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "29de561937c34d4991b03668d3ddf94d": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "29ec40c351af4586af7618ec34d074d8": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "2b23249b59554c28aee7f9e46279494d": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": "20px" + } + }, + "2d497498df36422fa825c7530b06689f": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_1958e6dcd6d14fba8d568eab0844dcc3", + "placeholder": "​", + "style": "IPY_MODEL_483c040b2d3541abbb91bf48c574e4b3", + "value": " 3121/3121 [00:00<00:00, 9285.27 examples/s]" + } + }, + "2d7569902c0f4403ac4b9c4cf78c98fe": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "2e9d4417e38042ee943f65ed9b2219c4": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "ProgressStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "ProgressStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "bar_color": null, + "description_width": "" + } + }, + "2ec0a5f561964ec2bb25ca16ab7a80fd": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "2edfdb8168a4426c83a619bd8b360e62": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "30498dc4ccf446f0a84c292ffb2977dc": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "ProgressStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "ProgressStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "bar_color": null, + "description_width": "" + } + }, + "3317b8253a634adeae1569c837875902": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "34d3839aeac94c8bb55f6d642b1b2a3d": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "355ec6e840d64c909d838cb92a9c0317": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "36f065ff5be84120b71e12bfd834ea04": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "ProgressStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "ProgressStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "bar_color": null, + "description_width": "" + } + }, + "372082bd461648e88150e283e448df77": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_743ef2e552ba4c84b477057e0fcad0d1", + "placeholder": "​", + "style": "IPY_MODEL_e33749eac8284c01a802e5286dffdca6", + "value": "Map: 100%" + } + }, + "396531f378294582a08e166b08624523": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "39bec2b2c62144258a4e0e5da0370b2c": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_049d5f2b43a34021b6eebdda28684f87", + "placeholder": "​", + "style": "IPY_MODEL_4daeee0d50f84dc68f683c9f34c19b02", + "value": " 32366/0 [00:00<00:00, 108794.87 examples/s]" + } + }, + "3b6c14a4ccdb46ef905b24b1643631b0": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "ProgressStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "ProgressStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "bar_color": null, + "description_width": "" + } + }, + "3d5d99fa7dfa4215a64f421911a3ecb7": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "3e2aaa52639e49ac9d75588ed5d3d404": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "FloatProgressModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "FloatProgressModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "ProgressView", + "bar_style": "success", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_0e445bbc266a48d1996276277227c962", + "max": 3121, + "min": 0, + "orientation": "horizontal", + "style": "IPY_MODEL_b552b17c39444ee596b6450705a38c58", + "value": 3121 + } + }, + "3f4dd190b1574025a660d6a7caf44b31": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "ProgressStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "ProgressStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "bar_color": null, + "description_width": "" + } + }, + "3fda7890c57d48d9919569de2ff6836d": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "404b48dedcc84e28a4ca5a5a3c3c2752": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_c0a1ce9bc97146a998d98e2d2abbbf53", + "placeholder": "​", + "style": "IPY_MODEL_76e5f19441af48f4a03d6fa51a4e9dd9", + "value": "Computing checksums: 100%" + } + }, + "45ddc57f719044bf90e6b1b516656111": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "46e1dc5b0edf459bbcb82091e64efa1e": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "46ff589d616946f1a3c2656f2bafbb2a": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "4731ef05f0fd47498f81866bcb1f186b": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "47be331e227942169383abbc47c2996f": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "483c040b2d3541abbb91bf48c574e4b3": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "49803a4f02ae45ebbf5c9ef4b51e84e1": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "4aa390ed478e4055a85143583352e710": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_4731ef05f0fd47498f81866bcb1f186b", + "placeholder": "​", + "style": "IPY_MODEL_ab983710675a4004a98bbb90c0df13ac", + "value": " 13123/0 [00:00<00:00, 112167.32 examples/s]" + } + }, + "4b854c733a8d48a9896c7769eb54469a": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "4d7f7c35a7da4a159dadb4d1c6e40863": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "4daeee0d50f84dc68f683c9f34c19b02": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "4e7a8e421f8b4d589155f6d4232e3dff": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "FloatProgressModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "FloatProgressModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "ProgressView", + "bar_style": "success", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_75238a8c83324387833f9053b5a797d4", + "max": 585, + "min": 0, + "orientation": "horizontal", + "style": "IPY_MODEL_3f4dd190b1574025a660d6a7caf44b31", + "value": 585 + } + }, + "535890f32ea74ebabedd609cface6f2a": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "FloatProgressModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "FloatProgressModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "ProgressView", + "bar_style": "success", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_8f7959eeda604ab5b7894b010545f4a6", + "max": 1, + "min": 0, + "orientation": "horizontal", + "style": "IPY_MODEL_76aa5992199f4488bf6b1926bda0a8af", + "value": 1 + } + }, + "56bbd7e1a27746e99993b4001a64e6a4": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "56d342c7700b4ad89299c79f07025fe0": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HBoxModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HBoxModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HBoxView", + "box_style": "", + "children": [ + "IPY_MODEL_9d0f79e362854c7e870a9e920f178b78", + "IPY_MODEL_1ca5b9b7b0c94f1d827eb99b0e4a7105", + "IPY_MODEL_bcacbd04a25e47a7bd3f3ffd2bb9601c" + ], + "layout": "IPY_MODEL_eaa7440763694fb9ae3d7a043df28551" + } + }, + "570f00c93ce84dad8a612d88c68e5332": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "57eb3949eab34997a836b0adc2428013": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "589e4554f9d444339544672614b64323": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "58a9e42aa0be4bcba1aa82c77965f65b": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "5ae7e1e397414ee88f444f73def2ef94": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "FloatProgressModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "FloatProgressModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "ProgressView", + "bar_style": "success", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_9d0c9a43293b4f5891f7ccf122059ee1", + "max": 1445, + "min": 0, + "orientation": "horizontal", + "style": "IPY_MODEL_fe6d61a05593421d8f8afe506e81bcb6", + "value": 1445 + } + }, + "5af9e4b6ca5244debb4b7d7b64864b27": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HBoxModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HBoxModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HBoxView", + "box_style": "", + "children": [ + "IPY_MODEL_62a2c398084d4e00b96a37ba6154beec", + "IPY_MODEL_86961fa32918401585b30ec7cb7371e6", + "IPY_MODEL_ad1ebdfdaf2f457e86e3efcc5211c118" + ], + "layout": "IPY_MODEL_72c0ae1a82d44d3581a87098f5da3a7d" + } + }, + "5b621638457c451cbc57ae224195c861": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_ad48203202a54a49b695f1f879cccce9", + "placeholder": "​", + "style": "IPY_MODEL_fc5e1e2b2b684ef0bf8d18669cfac768", + "value": " 1/1 [00:00<00:00, 113.92it/s]" + } + }, + "5cb00e9f99fa418490bc5f6788af24bc": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "5f00e1ee4b5b4e6a85aa130fe304e645": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "FloatProgressModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "FloatProgressModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "ProgressView", + "bar_style": "success", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_939ca0fb08f148d79f23ed0b63d9bbac", + "max": 10000, + "min": 0, + "orientation": "horizontal", + "style": "IPY_MODEL_fdd41013177b4efe95d6e4fa01dbae5b", + "value": 10000 + } + }, + "5f8afc8f5a0d4d58ba2bba0884dec63e": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "5f9e092930fc4a90a59dca9c25ca331e": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "FloatProgressModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "FloatProgressModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "ProgressView", + "bar_style": "success", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_88e7c51aba344a89838b1792434143d6", + "max": 1, + "min": 0, + "orientation": "horizontal", + "style": "IPY_MODEL_97129f6cb1004d23abfd1a5a538ef66f", + "value": 1 + } + }, + "62146c86154e47eb8e676aa279f68328": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "6278d40c45614699a5bbcfc47d4bebff": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "62a2c398084d4e00b96a37ba6154beec": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_c52945a2765a4b7c834c67dd299a1b87", + "placeholder": "​", + "style": "IPY_MODEL_57eb3949eab34997a836b0adc2428013", + "value": "Computing checksums: 100%" + } + }, + "635496dc27774b3e8fdf3c6e4d4b5c12": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "63b176c050d942d5bb2907cb5a02ef84": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "65337fa72a2e45f78d5ad5ed3c557737": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "ProgressStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "ProgressStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "bar_color": null, + "description_width": "" + } + }, + "653cf8d106ca45bda0422e6cd0c3bbda": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "654d4dace51d495e9748f6ae719f327c": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_26427932b7584edda94d1c2c95cb4eab", + "placeholder": "​", + "style": "IPY_MODEL_0b74a82e4ded40069bac382704222978", + "value": " 3121/3121 [00:00<00:00, 10565.94 examples/s]" + } + }, + "656018286e924dc9a7b76f29f8af2521": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "65733b2322f1448e9867248f3199fb91": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "661159fe366943e790fa575a11a5e097": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "668ebed3d9504bc59b54f10ebf640ca1": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_a2e1f9629dfd41758cd8024487e4ceb1", + "placeholder": "​", + "style": "IPY_MODEL_18576edbee874ba3b79e56fc65089a76", + "value": "Map: 100%" + } + }, + "68784d30fc9a4e04a87e2e5af37c8de7": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HBoxModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HBoxModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HBoxView", + "box_style": "", + "children": [ + "IPY_MODEL_372082bd461648e88150e283e448df77", + "IPY_MODEL_5f00e1ee4b5b4e6a85aa130fe304e645", + "IPY_MODEL_796fe18cc47b4cdabc42916209019c08" + ], + "layout": "IPY_MODEL_74e4fb1a22d74528bd83c8eba97c20b0" + } + }, + "69edd97bdba44434af8c1d705123b695": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_92f68e2575754538a2c13b98c5cd6a47", + "placeholder": "​", + "style": "IPY_MODEL_2edfdb8168a4426c83a619bd8b360e62", + "value": "Map: 100%" + } + }, + "6a4dab8708f14ce697a01727d02fdab0": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "6b0d3a846fcb4fd2a505684b31bbe37d": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "6b39603a6f9b4b9daabc352f4f6bf078": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_62146c86154e47eb8e676aa279f68328", + "placeholder": "​", + "style": "IPY_MODEL_5cb00e9f99fa418490bc5f6788af24bc", + "value": "Generating train split: " + } + }, + "6c159dc705054aecb174740637e2549b": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "6cc15565d32f41af95ccdb15fe2e6ed3": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "6e9e3370d4f44cf588ff27328c986186": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HBoxModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HBoxModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HBoxView", + "box_style": "", + "children": [ + "IPY_MODEL_6b39603a6f9b4b9daabc352f4f6bf078", + "IPY_MODEL_535890f32ea74ebabedd609cface6f2a", + "IPY_MODEL_80370a339fd641b09585e4be30d42ade" + ], + "layout": "IPY_MODEL_eb48725fec2942ad92fe96dec70aadc8" + } + }, + "6fe250fcaf5540fab56852ce2d8b26a4": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "70b35509df6045928712a82981883809": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HBoxModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HBoxModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HBoxView", + "box_style": "", + "children": [ + "IPY_MODEL_8f85e017f10c446287e0df08f9eb6683", + "IPY_MODEL_74a2156b3d394d749052ddcda4b324f8", + "IPY_MODEL_8ca28a72423c48339c1b35ebf25fd9a9" + ], + "layout": "IPY_MODEL_9ed5007c2d0d4098ac83be0c86567eab" + } + }, + "71ce2b44cc4448b9af944f85491783c3": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "7294a93e786f4ac6bd16baa3a54a2a40": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "FloatProgressModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "FloatProgressModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "ProgressView", + "bar_style": "success", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_efc223fb9e714bc9bd96953f80120eaa", + "max": 3121, + "min": 0, + "orientation": "horizontal", + "style": "IPY_MODEL_65337fa72a2e45f78d5ad5ed3c557737", + "value": 3121 + } + }, + "72c0ae1a82d44d3581a87098f5da3a7d": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "73d2b2727a0e42279c68046cd3c6d590": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_fef3e39d46e04b5c96dbc1776155495e", + "placeholder": "​", + "style": "IPY_MODEL_47be331e227942169383abbc47c2996f", + "value": " 28.7k/28.7k [00:00<00:00, 2.84MB/s]" + } + }, + "743ef2e552ba4c84b477057e0fcad0d1": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "74a2156b3d394d749052ddcda4b324f8": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "FloatProgressModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "FloatProgressModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "ProgressView", + "bar_style": "success", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_396531f378294582a08e166b08624523", + "max": 10000, + "min": 0, + "orientation": "horizontal", + "style": "IPY_MODEL_ca0043ff86254562832d2f5137902fe5", + "value": 10000 + } + }, + "74e4fb1a22d74528bd83c8eba97c20b0": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "75238a8c83324387833f9053b5a797d4": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "75e195ea430644df819c075580bfc8d5": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_29de561937c34d4991b03668d3ddf94d", + "placeholder": "​", + "style": "IPY_MODEL_d752645e1b844c10b75370a235e2e761", + "value": "config.json: 100%" + } + }, + "764f3ce785ac4b148b43d00874e195e0": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "76aa5992199f4488bf6b1926bda0a8af": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "ProgressStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "ProgressStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "bar_color": null, + "description_width": "" + } + }, + "76e5f19441af48f4a03d6fa51a4e9dd9": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "7738c6ccbab341e199caf72002df2c60": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "77ac4f7b53bb4825b1d180170de9a8be": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HBoxModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HBoxModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HBoxView", + "box_style": "", + "children": [ + "IPY_MODEL_a047d49f6ba44a42be17ef7aa922d24a", + "IPY_MODEL_c36c7e99aa474def8b9ff68764840c40", + "IPY_MODEL_4aa390ed478e4055a85143583352e710" + ], + "layout": "IPY_MODEL_570f00c93ce84dad8a612d88c68e5332" + } + }, + "796fe18cc47b4cdabc42916209019c08": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_22aab510900f4b7792862675c8d0b497", + "placeholder": "​", + "style": "IPY_MODEL_11251509b3424421bfbcd7db7e908cb0", + "value": " 10000/10000 [00:04<00:00, 2134.78 examples/s]" + } + }, + "79fa2d89a55446a8835fdaaa715262e8": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_1edbbb81b31b4296bf1c4366bb4d6fee", + "placeholder": "​", + "style": "IPY_MODEL_9a00ddb4336d432daace85e2d6631d32", + "value": " 112/112 [00:00<00:00, 9.14kB/s]" + } + }, + "7abb8ae429d54f22a7d0e0525dcc7cbf": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "7c2f51169b324b208bee733f2160f36d": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "FloatProgressModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "FloatProgressModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "ProgressView", + "bar_style": "success", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_656018286e924dc9a7b76f29f8af2521", + "max": 1, + "min": 0, + "orientation": "horizontal", + "style": "IPY_MODEL_7d48e5bbda174102b7a4d2f0cb9252a9", + "value": 1 + } + }, + "7c8d3f5b085e469c8d7362fb67aa460f": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "7cc99418d6254587b8082f5421293b56": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "7d48e5bbda174102b7a4d2f0cb9252a9": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "ProgressStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "ProgressStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "bar_color": null, + "description_width": "" + } + }, + "7da5898223524e33a1b526375488cd56": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "7e2c60191a8a495ebc38b3e9d1742bf0": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_b34ac961e9f945a4a6b41065a17dd587", + "placeholder": "​", + "style": "IPY_MODEL_2d7569902c0f4403ac4b9c4cf78c98fe", + "value": " 10000/10000 [00:04<00:00, 2039.40 examples/s]" + } + }, + "7e987667ff6b436196729f5db68826cd": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HBoxModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HBoxModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HBoxView", + "box_style": "", + "children": [ + "IPY_MODEL_668ebed3d9504bc59b54f10ebf640ca1", + "IPY_MODEL_8702b8f568bf454c9f048e8ab07f2376", + "IPY_MODEL_1876c30dc908494197724455f0ff7a91" + ], + "layout": "IPY_MODEL_34d3839aeac94c8bb55f6d642b1b2a3d" + } + }, + "7ee446e05d0e40b68162e79ff96e8ead": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "7eefaff6d28d49ea928a00baedf817a6": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HBoxModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HBoxModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HBoxView", + "box_style": "", + "children": [ + "IPY_MODEL_ba1ca5e6eb61453988b339cfc8e54b10", + "IPY_MODEL_15416fe7a4de428a868538c086f09218", + "IPY_MODEL_0fc716f771b949468823704c4bd4af0f" + ], + "layout": "IPY_MODEL_3317b8253a634adeae1569c837875902" + } + }, + "80370a339fd641b09585e4be30d42ade": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_7da5898223524e33a1b526375488cd56", + "placeholder": "​", + "style": "IPY_MODEL_5f8afc8f5a0d4d58ba2bba0884dec63e", + "value": " 15000/0 [00:00<00:00, 92786.13 examples/s]" + } + }, + "81fcc3da8b014b9ca43f382df82be674": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "829f2805f6f243fd921f9f4f385a781f": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "8459aa9cc2fb4ad6acf7f777a417d919": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_a5c7593d2e5346479ffcb25e9326d8f9", + "placeholder": "​", + "style": "IPY_MODEL_d0c87f4e24fc49f2ac6e714e9e271b47", + "value": "config.json: 100%" + } + }, + "86961fa32918401585b30ec7cb7371e6": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "FloatProgressModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "FloatProgressModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "ProgressView", + "bar_style": "success", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_bb73b86af4074b95864bb8a413bcc1c0", + "max": 1, + "min": 0, + "orientation": "horizontal", + "style": "IPY_MODEL_c2d7aa904bb048e89648994067e88293", + "value": 1 + } + }, + "86b6130fbc2b4ee9987dc73fef8be279": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "8702b8f568bf454c9f048e8ab07f2376": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "FloatProgressModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "FloatProgressModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "ProgressView", + "bar_style": "success", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_dc8b15badaf04b6a90d7a83237b7dd53", + "max": 10000, + "min": 0, + "orientation": "horizontal", + "style": "IPY_MODEL_3b6c14a4ccdb46ef905b24b1643631b0", + "value": 10000 + } + }, + "88e7c51aba344a89838b1792434143d6": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": "20px" + } + }, + "89f7d8bf50b04b2890b8d8e2a539b6ba": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "8b24bccdb43846cbbc74af9a01c36c01": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "8b7091fa5d9748498540667a99b40b27": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_ed60c5ce501843bd91732b7348620764", + "placeholder": "​", + "style": "IPY_MODEL_27f04577de04467f9f6e81746de9e659", + "value": "Map: 100%" + } + }, + "8ca09f024f41423fbd69e86b435fda52": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "8ca28a72423c48339c1b35ebf25fd9a9": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_46ff589d616946f1a3c2656f2bafbb2a", + "placeholder": "​", + "style": "IPY_MODEL_e4c5817707d8488dbca45bc92cc35f6e", + "value": " 10000/10000 [00:00<00:00, 11144.87 examples/s]" + } + }, + "8cf44da4e12a400cbe52549a82612345": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HBoxModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HBoxModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HBoxView", + "box_style": "", + "children": [ + "IPY_MODEL_f4e7cb19a59b489c98237c657bbf050b", + "IPY_MODEL_3e2aaa52639e49ac9d75588ed5d3d404", + "IPY_MODEL_e9d7963434ef4c24956d73d51fd859bd" + ], + "layout": "IPY_MODEL_f41f2371a644466d99f82743d69ca7c3" + } + }, + "8f7959eeda604ab5b7894b010545f4a6": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": "20px" + } + }, + "8f85e017f10c446287e0df08f9eb6683": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_4d7f7c35a7da4a159dadb4d1c6e40863", + "placeholder": "​", + "style": "IPY_MODEL_b6b60643435e46abab916a8050084ffe", + "value": "Map: 100%" + } + }, + "90bc3b1f14e0455f8bb810dfb5d49106": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HBoxModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HBoxModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HBoxView", + "box_style": "", + "children": [ + "IPY_MODEL_8b7091fa5d9748498540667a99b40b27", + "IPY_MODEL_d72e60249b564f1484b64b43b9648a50", + "IPY_MODEL_7e2c60191a8a495ebc38b3e9d1742bf0" + ], + "layout": "IPY_MODEL_a4dcaf0f1bf84d8fab03266183b02b23" + } + }, + "9223130ecf8c42b8af600cda70cd7cee": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "92527636584041eeb927870f6b81c117": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "92f68e2575754538a2c13b98c5cd6a47": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "939ca0fb08f148d79f23ed0b63d9bbac": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "93a899ddca074220a6aa69ecd0825fdf": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "FloatProgressModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "FloatProgressModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "ProgressView", + "bar_style": "success", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_29ec40c351af4586af7618ec34d074d8", + "max": 3121, + "min": 0, + "orientation": "horizontal", + "style": "IPY_MODEL_30498dc4ccf446f0a84c292ffb2977dc", + "value": 3121 + } + }, + "9519dbb2af464e7487409a78fc4afa9f": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_89f7d8bf50b04b2890b8d8e2a539b6ba", + "placeholder": "​", + "style": "IPY_MODEL_d024cf5c6ff34dd6b3fd2c8b53c92231", + "value": " 3121/3121 [00:01<00:00, 2133.55 examples/s]" + } + }, + "967781cfd06a455f9b716fe4e1e1ad14": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "97129f6cb1004d23abfd1a5a538ef66f": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "ProgressStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "ProgressStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "bar_color": null, + "description_width": "" + } + }, + "975694504d914cce8e58caa80d597f4d": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_8b24bccdb43846cbbc74af9a01c36c01", + "placeholder": "​", + "style": "IPY_MODEL_8ca09f024f41423fbd69e86b435fda52", + "value": "Map: 100%" + } + }, + "9860233c73664ddda5da2d8ba41c2bcc": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "986d8fe896364cc3801ae2cef01f1008": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "990f1d0f558249eab10cfabe109305d2": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "9a00ddb4336d432daace85e2d6631d32": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "9a92b0a92c064891a64e2cd3431ceca5": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HBoxModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HBoxModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HBoxView", + "box_style": "", + "children": [ + "IPY_MODEL_9e96909b01df4dbb864efcba33d46673", + "IPY_MODEL_f07ebaad9d0a4442b9f43461c0e68405", + "IPY_MODEL_5b621638457c451cbc57ae224195c861" + ], + "layout": "IPY_MODEL_6fe250fcaf5540fab56852ce2d8b26a4" + } + }, + "9d0c9a43293b4f5891f7ccf122059ee1": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "9d0f79e362854c7e870a9e920f178b78": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_92527636584041eeb927870f6b81c117", + "placeholder": "​", + "style": "IPY_MODEL_63b176c050d942d5bb2907cb5a02ef84", + "value": "Map: 100%" + } + }, + "9e96909b01df4dbb864efcba33d46673": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_cf739530b3b94e02960ebf56afe1df0a", + "placeholder": "​", + "style": "IPY_MODEL_ed4be32c5afd4887ad02eedb93cb6df2", + "value": "Computing checksums: 100%" + } + }, + "9ed5007c2d0d4098ac83be0c86567eab": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "a047d49f6ba44a42be17ef7aa922d24a": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_86b6130fbc2b4ee9987dc73fef8be279", + "placeholder": "​", + "style": "IPY_MODEL_56bbd7e1a27746e99993b4001a64e6a4", + "value": "Generating train split: " + } + }, + "a2dab54d1a684dc1bf1d18c27b491d92": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "a2e1f9629dfd41758cd8024487e4ceb1": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "a3967e826f194984986f636210484ef4": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "a4dcaf0f1bf84d8fab03266183b02b23": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "a5c2604673a04a39ba7387cc73cd57e3": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HBoxModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HBoxModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HBoxView", + "box_style": "", + "children": [ + "IPY_MODEL_aa6c3c7ef00048afa89b9c3596bfc38d", + "IPY_MODEL_0cdfd05817854fe78db65b031e8ae1a9", + "IPY_MODEL_79fa2d89a55446a8835fdaaa715262e8" + ], + "layout": "IPY_MODEL_f46f4c4d99684f308a9beb6263031bb7" + } + }, + "a5c7593d2e5346479ffcb25e9326d8f9": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "aa6c3c7ef00048afa89b9c3596bfc38d": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_81fcc3da8b014b9ca43f382df82be674", + "placeholder": "​", + "style": "IPY_MODEL_49803a4f02ae45ebbf5c9ef4b51e84e1", + "value": "special_tokens_map.json: 100%" + } + }, + "aa6d8339818549e78d10dc5d3ae11b12": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_6cc15565d32f41af95ccdb15fe2e6ed3", + "placeholder": "​", + "style": "IPY_MODEL_9860233c73664ddda5da2d8ba41c2bcc", + "value": "tokenizer_config.json: 100%" + } + }, + "ab983710675a4004a98bbb90c0df13ac": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "aba43cb51c934eca94eb2754f5843b0e": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_02c62a2bcc8a4631bd83ee770983c225", + "placeholder": "​", + "style": "IPY_MODEL_7738c6ccbab341e199caf72002df2c60", + "value": "Map: 100%" + } + }, + "ac614455a3de4e379cd7e4a267673e28": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_45ddc57f719044bf90e6b1b516656111", + "placeholder": "​", + "style": "IPY_MODEL_21d2c65103194be092471637c2faada4", + "value": " 3121/3121 [00:00<00:00, 9404.56 examples/s]" + } + }, + "acc37e80a1674d26afb375d7b086f89f": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "FloatProgressModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "FloatProgressModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "ProgressView", + "bar_style": "success", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_3d5d99fa7dfa4215a64f421911a3ecb7", + "max": 10000, + "min": 0, + "orientation": "horizontal", + "style": "IPY_MODEL_f9af40a77d124fde92f1ea852256dbd5", + "value": 10000 + } + }, + "ad1ebdfdaf2f457e86e3efcc5211c118": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_7ee446e05d0e40b68162e79ff96e8ead", + "placeholder": "​", + "style": "IPY_MODEL_dd0e05def74e4bd5bc51d4d427d0a44c", + "value": " 1/1 [00:00<00:00, 116.61it/s]" + } + }, + "ad48203202a54a49b695f1f879cccce9": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "aec100a6b2ee4f09ba2b27b11b728629": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_e3fe913f3d2d4522a7604ee0a07fe62d", + "placeholder": "​", + "style": "IPY_MODEL_bb000f9c18dc409c9eab3295e5507d94", + "value": "Generating train split: " + } + }, + "b06854cde9d1432cb4b557a0d96b79c2": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "b1d8cb19f86f4dfbbc4571376927255e": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "b2cf8017f56b48c7bdcf7ca1540d307c": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "ProgressStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "ProgressStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "bar_color": null, + "description_width": "" + } + }, + "b34ac961e9f945a4a6b41065a17dd587": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "b36ee6a68e7047afae8ca921e806430e": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "b552b17c39444ee596b6450705a38c58": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "ProgressStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "ProgressStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "bar_color": null, + "description_width": "" + } + }, + "b69cb7e6ad7141679f9e8fdc04e39446": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "FloatProgressModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "FloatProgressModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "ProgressView", + "bar_style": "success", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_22aec26a870a4c7f8026f1c7a8004c38", + "max": 3121, + "min": 0, + "orientation": "horizontal", + "style": "IPY_MODEL_b2cf8017f56b48c7bdcf7ca1540d307c", + "value": 3121 + } + }, + "b6b60643435e46abab916a8050084ffe": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "b7ecb63bc965460da610712bf027d386": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HBoxModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HBoxModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HBoxView", + "box_style": "", + "children": [ + "IPY_MODEL_d808cdb1c39641a68c214a7fd77a3756", + "IPY_MODEL_e7a1b79822f5449eae09265b829063f3", + "IPY_MODEL_2d497498df36422fa825c7530b06689f" + ], + "layout": "IPY_MODEL_3fda7890c57d48d9919569de2ff6836d" + } + }, + "b90caca4f0984095ab27052282318584": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "ba1ca5e6eb61453988b339cfc8e54b10": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_653cf8d106ca45bda0422e6cd0c3bbda", + "placeholder": "​", + "style": "IPY_MODEL_25f9b192b92d41d79440924079e347e1", + "value": "Map: 100%" + } + }, + "bb000f9c18dc409c9eab3295e5507d94": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "bb73b86af4074b95864bb8a413bcc1c0": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "bcacbd04a25e47a7bd3f3ffd2bb9601c": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_ef85a7abeadd4f7c96d11f4e2d82e370", + "placeholder": "​", + "style": "IPY_MODEL_6c159dc705054aecb174740637e2549b", + "value": " 10000/10000 [00:01<00:00, 9452.06 examples/s]" + } + }, + "bdcae7a2765d490a93abfca65f66d202": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HBoxModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HBoxModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HBoxView", + "box_style": "", + "children": [ + "IPY_MODEL_aba43cb51c934eca94eb2754f5843b0e", + "IPY_MODEL_7294a93e786f4ac6bd16baa3a54a2a40", + "IPY_MODEL_ac614455a3de4e379cd7e4a267673e28" + ], + "layout": "IPY_MODEL_7abb8ae429d54f22a7d0e0525dcc7cbf" + } + }, + "bdcfb933166e469b8cbabc41235da058": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "c0a1ce9bc97146a998d98e2d2abbbf53": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "c2d7aa904bb048e89648994067e88293": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "ProgressStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "ProgressStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "bar_color": null, + "description_width": "" + } + }, + "c36c7e99aa474def8b9ff68764840c40": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "FloatProgressModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "FloatProgressModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "ProgressView", + "bar_style": "success", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_2b23249b59554c28aee7f9e46279494d", + "max": 1, + "min": 0, + "orientation": "horizontal", + "style": "IPY_MODEL_d8e3ee9f433b441a9c71310b14979e59", + "value": 1 + } + }, + "c38e40fc8bc846eab8d6fb971fe4858a": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "c3cccb5ef54b44c5b7928c11f36979a7": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "c52945a2765a4b7c834c67dd299a1b87": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "c5ea3c6735c5448fbe032c86b211cd11": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "c69745397340449f8f8fe24c94dd9f29": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "c7471715331d42e795bd641d987479f7": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "ProgressStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "ProgressStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "bar_color": null, + "description_width": "" + } + }, + "c9453a6e52b34ab79a7781591b703d46": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "FloatProgressModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "FloatProgressModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "ProgressView", + "bar_style": "success", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_e9eb9dbe0f1b4c9bbf4747471fe33e10", + "max": 40, + "min": 0, + "orientation": "horizontal", + "style": "IPY_MODEL_e80a3bb452e44da0abf1471cf21746c0", + "value": 40 + } + }, + "c989f464dc2b43bc8e77b74bad170f13": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "c9b61f6291e442999cca901fdd211c1c": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "ProgressStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "ProgressStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "bar_color": null, + "description_width": "" + } + }, + "ca0043ff86254562832d2f5137902fe5": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "ProgressStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "ProgressStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "bar_color": null, + "description_width": "" + } + }, + "cf739530b3b94e02960ebf56afe1df0a": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "cfb56f5c8f6f4ab1b941f97f5647b339": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HBoxModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HBoxModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HBoxView", + "box_style": "", + "children": [ + "IPY_MODEL_aec100a6b2ee4f09ba2b27b11b728629", + "IPY_MODEL_5f9e092930fc4a90a59dca9c25ca331e", + "IPY_MODEL_39bec2b2c62144258a4e0e5da0370b2c" + ], + "layout": "IPY_MODEL_f532274fda5b4e29abef8c99f86cc5d2" + } + }, + "d024cf5c6ff34dd6b3fd2c8b53c92231": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "d0c87f4e24fc49f2ac6e714e9e271b47": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "d59bbd5a61d64b71883674292dc02782": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "d5c44e12017b4b6a8d5c7f1d9f2a0bf9": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "d72e60249b564f1484b64b43b9648a50": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "FloatProgressModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "FloatProgressModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "ProgressView", + "bar_style": "success", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_df1a59d0215e41b4bafb022e1c276c78", + "max": 10000, + "min": 0, + "orientation": "horizontal", + "style": "IPY_MODEL_c9b61f6291e442999cca901fdd211c1c", + "value": 10000 + } + }, + "d752645e1b844c10b75370a235e2e761": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "d808cdb1c39641a68c214a7fd77a3756": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_a2dab54d1a684dc1bf1d18c27b491d92", + "placeholder": "​", + "style": "IPY_MODEL_661159fe366943e790fa575a11a5e097", + "value": "Map: 100%" + } + }, + "d8e3ee9f433b441a9c71310b14979e59": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "ProgressStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "ProgressStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "bar_color": null, + "description_width": "" + } + }, + "d9475d52fa814601a1b7f33b5541cd70": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "d989263ed1144a5ca2d00c78873d4448": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "dbb724416bf84af2a1efd3c2387f1aa3": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_1c903f797329485d94a584d5db1e73dc", + "placeholder": "​", + "style": "IPY_MODEL_e0820dcd59d942808de4ca5b5018df4d", + "value": "model.safetensors: 100%" + } + }, + "dc8b15badaf04b6a90d7a83237b7dd53": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "dd0e05def74e4bd5bc51d4d427d0a44c": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "de646da719f14d0c9188ca352c92004e": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HBoxModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HBoxModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HBoxView", + "box_style": "", + "children": [ + "IPY_MODEL_190ee28f166743a5952264de5ab458bf", + "IPY_MODEL_acc37e80a1674d26afb375d7b086f89f", + "IPY_MODEL_10e1bf80fdb24a768f896bed74057532" + ], + "layout": "IPY_MODEL_829f2805f6f243fd921f9f4f385a781f" + } + }, + "df1a59d0215e41b4bafb022e1c276c78": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "e0820dcd59d942808de4ca5b5018df4d": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "e33749eac8284c01a802e5286dffdca6": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "e3fe913f3d2d4522a7604ee0a07fe62d": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "e4c5817707d8488dbca45bc92cc35f6e": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "e7a1b79822f5449eae09265b829063f3": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "FloatProgressModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "FloatProgressModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "ProgressView", + "bar_style": "success", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_18ff3c7a0a1d4595b4a7ff941c0c453b", + "max": 3121, + "min": 0, + "orientation": "horizontal", + "style": "IPY_MODEL_094a97b9393541e8bc11f85a7c7767c5", + "value": 3121 + } + }, + "e7bc9dd22971441ba81272f79e5827be": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HBoxModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HBoxModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HBoxView", + "box_style": "", + "children": [ + "IPY_MODEL_975694504d914cce8e58caa80d597f4d", + "IPY_MODEL_93a899ddca074220a6aa69ecd0825fdf", + "IPY_MODEL_654d4dace51d495e9748f6ae719f327c" + ], + "layout": "IPY_MODEL_6a4dab8708f14ce697a01727d02fdab0" + } + }, + "e7ec7b587a1c44a1ad68105be1f2a15e": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_986d8fe896364cc3801ae2cef01f1008", + "placeholder": "​", + "style": "IPY_MODEL_c3cccb5ef54b44c5b7928c11f36979a7", + "value": " 40.0/40.0 [00:00<00:00, 3.55kB/s]" + } + }, + "e80a3bb452e44da0abf1471cf21746c0": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "ProgressStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "ProgressStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "bar_color": null, + "description_width": "" + } + }, + "e9d7963434ef4c24956d73d51fd859bd": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_355ec6e840d64c909d838cb92a9c0317", + "placeholder": "​", + "style": "IPY_MODEL_967781cfd06a455f9b716fe4e1e1ad14", + "value": " 3121/3121 [00:01<00:00, 2093.09 examples/s]" + } + }, + "e9eb9dbe0f1b4c9bbf4747471fe33e10": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "eaa7440763694fb9ae3d7a043df28551": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "eb48725fec2942ad92fe96dec70aadc8": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "ed4be32c5afd4887ad02eedb93cb6df2": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "ed60c5ce501843bd91732b7348620764": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "ef7b4b3d3df24c4ba9605992a3a90c79": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HBoxModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HBoxModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HBoxView", + "box_style": "", + "children": [ + "IPY_MODEL_22bca1b1ece94d0cab713ef13be82e48", + "IPY_MODEL_f2b20faf727e4275bbfaa003036b2001", + "IPY_MODEL_73d2b2727a0e42279c68046cd3c6d590" + ], + "layout": "IPY_MODEL_764f3ce785ac4b148b43d00874e195e0" + } + }, + "ef85a7abeadd4f7c96d11f4e2d82e370": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "efc223fb9e714bc9bd96953f80120eaa": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "f07ebaad9d0a4442b9f43461c0e68405": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "FloatProgressModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "FloatProgressModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "ProgressView", + "bar_style": "success", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_d5c44e12017b4b6a8d5c7f1d9f2a0bf9", + "max": 1, + "min": 0, + "orientation": "horizontal", + "style": "IPY_MODEL_36f065ff5be84120b71e12bfd834ea04", + "value": 1 + } + }, + "f08d237b9291499d90b87f430b626ea9": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_b90caca4f0984095ab27052282318584", + "placeholder": "​", + "style": "IPY_MODEL_6278d40c45614699a5bbcfc47d4bebff", + "value": " 585/585 [00:00<00:00, 52.6kB/s]" + } + }, + "f2b20faf727e4275bbfaa003036b2001": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "FloatProgressModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "FloatProgressModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "ProgressView", + "bar_style": "success", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_990f1d0f558249eab10cfabe109305d2", + "max": 28703, + "min": 0, + "orientation": "horizontal", + "style": "IPY_MODEL_c7471715331d42e795bd641d987479f7", + "value": 28703 + } + }, + "f3a6678289b74fc0bf55df6b2fc6013e": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "FloatProgressModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "FloatProgressModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "ProgressView", + "bar_style": "success", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_b06854cde9d1432cb4b557a0d96b79c2", + "max": 4038244, + "min": 0, + "orientation": "horizontal", + "style": "IPY_MODEL_05cb827adaa74c2d9243709693be0cf2", + "value": 4038244 + } + }, + "f41f2371a644466d99f82743d69ca7c3": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "f46f4c4d99684f308a9beb6263031bb7": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "f4e7cb19a59b489c98237c657bbf050b": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HTMLModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HTMLModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HTMLView", + "description": "", + "description_tooltip": null, + "layout": "IPY_MODEL_58a9e42aa0be4bcba1aa82c77965f65b", + "placeholder": "​", + "style": "IPY_MODEL_a3967e826f194984986f636210484ef4", + "value": "Map: 100%" + } + }, + "f532274fda5b4e29abef8c99f86cc5d2": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + }, + "f86cd40e74a349dfa206ec4f8122d989": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "ProgressStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "ProgressStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "bar_color": null, + "description_width": "" + } + }, + "f9178fe5a1b94820915fd815992d301f": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "HBoxModel", + "state": { + "_dom_classes": [], + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "HBoxModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/controls", + "_view_module_version": "1.5.0", + "_view_name": "HBoxView", + "box_style": "", + "children": [ + "IPY_MODEL_8459aa9cc2fb4ad6acf7f777a417d919", + "IPY_MODEL_4e7a8e421f8b4d589155f6d4232e3dff", + "IPY_MODEL_f08d237b9291499d90b87f430b626ea9" + ], + "layout": "IPY_MODEL_2790a55362c84e30a489f3ea144f019f" + } + }, + "f9af40a77d124fde92f1ea852256dbd5": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "ProgressStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "ProgressStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "bar_color": null, + "description_width": "" + } + }, + "fc5e1e2b2b684ef0bf8d18669cfac768": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "DescriptionStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "DescriptionStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "description_width": "" + } + }, + "fdd41013177b4efe95d6e4fa01dbae5b": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "ProgressStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "ProgressStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "bar_color": null, + "description_width": "" + } + }, + "fe6d61a05593421d8f8afe506e81bcb6": { + "model_module": "@jupyter-widgets/controls", + "model_module_version": "1.5.0", + "model_name": "ProgressStyleModel", + "state": { + "_model_module": "@jupyter-widgets/controls", + "_model_module_version": "1.5.0", + "_model_name": "ProgressStyleModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "StyleView", + "bar_color": null, + "description_width": "" + } + }, + "fef3e39d46e04b5c96dbc1776155495e": { + "model_module": "@jupyter-widgets/base", + "model_module_version": "1.2.0", + "model_name": "LayoutModel", + "state": { + "_model_module": "@jupyter-widgets/base", + "_model_module_version": "1.2.0", + "_model_name": "LayoutModel", + "_view_count": null, + "_view_module": "@jupyter-widgets/base", + "_view_module_version": "1.2.0", + "_view_name": "LayoutView", + "align_content": null, + "align_items": null, + "align_self": null, + "border": null, + "bottom": null, + "display": null, + "flex": null, + "flex_flow": null, + "grid_area": null, + "grid_auto_columns": null, + "grid_auto_flow": null, + "grid_auto_rows": null, + "grid_column": null, + "grid_gap": null, + "grid_row": null, + "grid_template_areas": null, + "grid_template_columns": null, + "grid_template_rows": null, + "height": null, + "justify_content": null, + "justify_items": null, + "left": null, + "margin": null, + "max_height": null, + "max_width": null, + "min_height": null, + "min_width": null, + "object_fit": null, + "object_position": null, + "order": null, + "overflow": null, + "overflow_x": null, + "overflow_y": null, + "padding": null, + "right": null, + "top": null, + "visibility": null, + "width": null + } + } + }, + "version_major": 2, + "version_minor": 0 + } + } + }, + "nbformat": 4, + "nbformat_minor": 5 +} diff --git a/Finetuned Models/training_log.txt b/Finetuned Models/training_log.txt new file mode 100644 index 0000000000000000000000000000000000000000..584cdd23e4bc6a77766520fdf53cd59a2615e297 --- /dev/null +++ b/Finetuned Models/training_log.txt @@ -0,0 +1,6 @@ +Task: dnasplice, Model: fabihamakhdoomi/TinyDNABERT, Config: base_model, Training Time: 468.7715392112732, Metrics: {'eval_loss': 0.6981738209724426, 'eval_accuracy': 0.9202178788849728, 'eval_recall': 0.7544378698224852, 'eval_specificity': 0.9660531697341513, 'eval_mcc': 0.7565288289137028, 'eval_roc_auc': 0.9536779564622887, 'eval_precision': 0.8600337268128162, 'eval_f1': 0.8037825059101655, 'eval_true_pos': 510, 'eval_true_neg': 2362, 'eval_false_pos': 83, 'eval_false_neg': 166, 'eval_runtime': 2.8254, 'eval_samples_per_second': 1104.619, 'eval_steps_per_second': 110.78, 'epoch': 25.0} +Task: dnasplice, Model: fabihamakhdoomi/TinyDNABERT, Config: attention_plus_feed_forward, Training Time: 550.877524137497, Metrics: {'eval_loss': 0.3186633586883545, 'eval_accuracy': 0.8766421018904197, 'eval_recall': 0.4985207100591716, 'eval_specificity': 0.9811860940695296, 'eval_mcc': 0.6022440934663668, 'eval_roc_auc': 0.8810529882261832, 'eval_precision': 0.8798955613577023, 'eval_f1': 0.6364494806421152, 'eval_true_pos': 337, 'eval_true_neg': 2399, 'eval_false_pos': 46, 'eval_false_neg': 339, 'eval_runtime': 3.3308, 'eval_samples_per_second': 937.023, 'eval_steps_per_second': 93.973, 'epoch': 25.0} +Task: tfbs, Model: fabihamakhdoomi/TinyDNABERT, Config: base_model, Training Time: 469.97587299346924, Metrics: {'eval_loss': 2.653303384780884, 'eval_accuracy': 0.6578019865427748, 'eval_recall': 0.6830497794580971, 'eval_specificity': 0.6316818774445893, 'eval_mcc': 0.3151935621806032, 'eval_roc_auc': 0.6742539817897865, 'eval_precision': 0.6573681018799272, 'eval_f1': 0.6699629171817059, 'eval_true_pos': 1084, 'eval_true_neg': 969, 'eval_false_pos': 565, 'eval_false_neg': 503, 'eval_runtime': 2.7557, 'eval_samples_per_second': 1132.56, 'eval_steps_per_second': 113.583, 'epoch': 25.0} +Task: tfbs, Model: fabihamakhdoomi/TinyDNABERT, Config: attention_plus_feed_forward, Training Time: 549.6457896232605, Metrics: {'eval_loss': 0.635671079158783, 'eval_accuracy': 0.6536366549182954, 'eval_recall': 0.6962822936357907, 'eval_specificity': 0.6095176010430248, 'eval_mcc': 0.307056824567405, 'eval_roc_auc': 0.7121371574288815, 'eval_precision': 0.6484741784037559, 'eval_f1': 0.6715284108173807, 'eval_true_pos': 1105, 'eval_true_neg': 935, 'eval_false_pos': 599, 'eval_false_neg': 482, 'eval_runtime': 3.2385, 'eval_samples_per_second': 963.73, 'eval_steps_per_second': 96.651, 'epoch': 25.0} +Task: dnaprom, Model: fabihamakhdoomi/TinyDNABERT, Config: base_model, Training Time: 470.97279715538025, Metrics: {'eval_loss': 0.9456052184104919, 'eval_accuracy': 0.887536046139058, 'eval_recall': 0.9066073697585769, 'eval_specificity': 0.8681318681318682, 'eval_mcc': 0.7755020642874026, 'eval_roc_auc': 0.9237130684548279, 'eval_precision': 0.8749233599019006, 'eval_f1': 0.8904836193447738, 'eval_true_pos': 1427, 'eval_true_neg': 1343, 'eval_false_pos': 204, 'eval_false_neg': 147, 'eval_runtime': 2.8069, 'eval_samples_per_second': 1111.884, 'eval_steps_per_second': 111.509, 'epoch': 25.0} +Task: dnaprom, Model: fabihamakhdoomi/TinyDNABERT, Config: attention_plus_feed_forward, Training Time: 558.7053642272949, Metrics: {'eval_loss': 0.3336106240749359, 'eval_accuracy': 0.8583787247677026, 'eval_recall': 0.8875476493011436, 'eval_specificity': 0.8287007110536523, 'eval_mcc': 0.7177414140554846, 'eval_roc_auc': 0.9391871302327989, 'eval_precision': 0.8405535499398316, 'eval_f1': 0.8634116192830655, 'eval_true_pos': 1397, 'eval_true_neg': 1282, 'eval_false_pos': 265, 'eval_false_neg': 177, 'eval_runtime': 3.2789, 'eval_samples_per_second': 951.851, 'eval_steps_per_second': 95.46, 'epoch': 25.0} diff --git a/config.json b/Pretrained Base Model/config.json similarity index 100% rename from config.json rename to Pretrained Base Model/config.json diff --git a/generation_config.json b/Pretrained Base Model/generation_config.json similarity index 100% rename from generation_config.json rename to Pretrained Base Model/generation_config.json diff --git a/model.safetensors b/Pretrained Base Model/model.safetensors similarity index 100% rename from model.safetensors rename to Pretrained Base Model/model.safetensors diff --git a/optimizer.pt b/Pretrained Base Model/optimizer.pt similarity index 100% rename from optimizer.pt rename to Pretrained Base Model/optimizer.pt diff --git a/Pretrained Base Model/pretraining_script.ipynb b/Pretrained Base Model/pretraining_script.ipynb new file mode 100644 index 0000000000000000000000000000000000000000..5b0a13e47653aae4667f8de6f6fad52a3536bc9f --- /dev/null +++ b/Pretrained Base Model/pretraining_script.ipynb @@ -0,0 +1 @@ +{"cells":[{"cell_type":"code","execution_count":null,"metadata":{"scrolled":true,"trusted":true},"outputs":[],"source":["import argparse\n","import glob\n","import logging\n","import os\n","import pickle\n","import random\n","import re\n","import shutil\n","from typing import Dict, List, Tuple\n","from copy import deepcopy\n","from multiprocessing import Pool\n","\n","import numpy as np\n","import torch\n","from torch.nn.utils.rnn import pad_sequence\n","from torch.utils.data import DataLoader, Dataset, RandomSampler, SequentialSampler\n","from torch.utils.data.distributed import DistributedSampler\n","from tqdm import tqdm, trange\n","from transformers import AutoTokenizer, AutoModel\n","from transformers import (\n"," WEIGHTS_NAME,\n"," AdamW,\n"," BertConfig,\n"," BertForMaskedLM,\n"," BertTokenizer,\n"," CamembertConfig,\n"," CamembertForMaskedLM,\n"," CamembertTokenizer,\n"," DistilBertConfig,\n"," DistilBertForMaskedLM,\n"," DistilBertTokenizer,\n"," GPT2Config,\n"," GPT2LMHeadModel,\n"," GPT2Tokenizer,\n"," OpenAIGPTConfig,\n"," OpenAIGPTLMHeadModel,\n"," OpenAIGPTTokenizer,\n"," PreTrainedModel,\n"," PreTrainedTokenizer,\n"," RobertaConfig,\n"," RobertaForMaskedLM,\n"," RobertaTokenizer,\n"," get_linear_schedule_with_warmup,\n"," get_cosine_with_hard_restarts_schedule_with_warmup\n",")\n","\n","\n","try:\n"," from torch.utils.tensorboard import SummaryWriter\n","except ImportError:\n"," from tensorboardX import SummaryWriter\n","\n","\n","logger = logging.getLogger(__name__)\n","\n","DNATokenizer = AutoTokenizer.from_pretrained(\"zhihan1996/DNA_bert_6\", trust_remote_code=True)\n","\n","\n","MODEL_CLASSES = {\n"," \"gpt2\": (GPT2Config, GPT2LMHeadModel, GPT2Tokenizer),\n"," \"openai-gpt\": (OpenAIGPTConfig, OpenAIGPTLMHeadModel, OpenAIGPTTokenizer),\n"," \"dna\": (BertConfig, BertForMaskedLM, DNATokenizer),\n"," \"bert\": (BertConfig, BertForMaskedLM, BertTokenizer),\n"," \"roberta\": (RobertaConfig, RobertaForMaskedLM, RobertaTokenizer),\n"," \"distilbert\": (DistilBertConfig, DistilBertForMaskedLM, DistilBertTokenizer),\n"," \"camembert\": (CamembertConfig, CamembertForMaskedLM, CamembertTokenizer),\n","}\n","\n","MASK_LIST = {\n"," \"3\": [-1, 1],\n"," \"4\": [-1, 1, 2],\n"," \"5\": [-2, -1, 1, 2],\n"," \"6\": [-2, -1, 1, 2, 3]\n","}\n","\n","\n","class TextDataset(Dataset):\n"," def __init__(self, tokenizer: PreTrainedTokenizer, config, file_path: str, block_size=512):\n"," assert os.path.isfile(file_path)\n","\n","\n"," directory, filename = os.path.split(file_path)\n"," cached_features_file = os.path.join(\n"," directory, dna + \"_cached_lm_\" + str(block_size) + \"_\" + filename\n"," )\n","\n"," if os.path.exists(cached_features_file) and not config['overwrite_cache']:\n"," logger.info(\"Loading features from cached file %s\", cached_features_file)\n"," with open(cached_features_file, \"rb\") as handle:\n"," self.examples = pickle.load(handle)\n"," else:\n"," logger.info(\"Creating features from dataset file at %s\", directory)\n","\n"," self.examples = []\n"," with open(file_path, encoding=\"utf-8\") as f:\n"," text = f.read()\n","\n"," tokenized_text = tokenizer.convert_tokens_to_ids(tokenizer.tokenize(text))\n","\n"," for i in range(0, len(tokenized_text) - block_size + 1, block_size): # Truncate in block of block_size\n"," self.examples.append(tokenizer.build_inputs_with_special_tokens(tokenized_text[i : i + block_size]))\n"," logger.info(\"Saving features into cached file %s\", cached_features_file)\n"," with open(cached_features_file, \"wb\") as handle:\n"," pickle.dump(self.examples, handle, protocol=pickle.HIGHEST_PROTOCOL)\n","\n"," def __len__(self):\n"," return len(self.examples)\n","\n"," def __getitem__(self, item):\n"," return torch.tensor(self.examples[item], dtype=torch.long)\n","\n","def convert_line_to_example(tokenizer, lines, max_length, add_special_tokens=True):\n"," examples = tokenizer.batch_encode_plus(lines, add_special_tokens=add_special_tokens, max_length=max_length)[\"input_ids\"]\n"," return examples\n","\n","class LineByLineTextDataset(Dataset):\n"," def __init__(self, tokenizer: PreTrainedTokenizer, config, file_path: str, block_size=512):\n"," assert os.path.isfile(file_path)\n"," # Here, we do not cache the features, operating under the assumption\n"," # that we will soon use fast multithreaded tokenizers from the\n"," # `tokenizers` repo everywhere =)\n"," directory, filename = os.path.split(file_path)\n"," cached_features_file = os.path.join(\n"," '/kaggle/working/', 'dna' + \"_cached_lm_\" + str(block_size) + \"_\" + filename\n"," )\n","\n"," if os.path.exists(cached_features_file) and not config['overwrite_cache']:\n"," logger.info(\"Loading features from cached file %s\", cached_features_file)\n"," with open(cached_features_file, \"rb\") as handle:\n"," self.examples = pickle.load(handle)\n"," else:\n"," logger.info(\"Creating features from dataset file at %s\", file_path)\n","\n"," with open(file_path, encoding=\"utf-8\") as f:\n"," lines = [line for line in f.read().splitlines() if (len(line) > 0 and not line.isspace())]\n"," \n"," if config['n_process'] == 1:\n"," self.examples = tokenizer.batch_encode_plus(lines, add_special_tokens=True, max_length=block_size)[\"input_ids\"]\n"," else:\n"," n_proc = config['n_process']\n"," p = Pool(n_proc)\n"," indexes = [0]\n"," len_slice = int(len(lines)/n_proc)\n"," for i in range(1, n_proc+1):\n"," if i != n_proc:\n"," indexes.append(len_slice*(i))\n"," else:\n"," indexes.append(len(lines))\n"," results = []\n"," for i in range(n_proc):\n"," results.append(p.apply_async(convert_line_to_example,[tokenizer, lines[indexes[i]:indexes[i+1]], block_size,]))\n"," print(str(i) + \" start\")\n"," p.close() \n"," p.join()\n","\n"," self.examples = []\n"," for result in results:\n"," ids = result.get()\n"," self.examples.extend(ids)\n","\n"," logger.info(\"Saving features into cached file %s\", cached_features_file)\n"," with open(cached_features_file, \"wb\") as handle:\n"," pickle.dump(self.examples, handle, protocol=pickle.HIGHEST_PROTOCOL)\n","\n"," def __len__(self):\n"," return len(self.examples)\n","\n"," def __getitem__(self, i):\n"," return torch.tensor(self.examples[i], dtype=torch.long)\n","\n","\n","def load_and_cache_examples(config, tokenizer, evaluate=False):\n"," file_path = r\"/kaggle/input/random-dna-sequences-for-transfomer-pretraining/6_12k.txt\" if evaluate else r'/kaggle/input/random-dna-sequences-for-transfomer-pretraining/6_12k.txt'\n"," if config['line_by_line']:\n"," return LineByLineTextDataset(tokenizer, config, file_path=file_path, block_size=config['block_size'])\n"," else:\n"," return TextDataset(tokenizer, config, file_path=file_path, block_size=config['block_size'])\n","\n","\n","def set_seed(config):\n"," random.seed(config['seed'])\n"," np.random.seed(config['seed'])\n"," torch.manual_seed(config['seed'])\n"," if config['n_gpu'] > 0:\n"," torch.cuda.manual_seed_all(config['seed'])\n","\n","\n","def _sorted_checkpoints(config, checkpoint_prefix=\"checkpoint\", use_mtime=False) -> List[str]:\n"," ordering_and_checkpoint_path = []\n"," st = r\"/kaggle/working/output\"\n"," \n"," glob_checkpoints = glob.glob(os.path.join(st, \"{}-*\".format(checkpoint_prefix)))\n","\n"," for path in glob_checkpoints:\n"," if use_mtime:\n"," ordering_and_checkpoint_path.append((os.path.getmtime(path), path))\n"," else:\n"," regex_match = re.match(\".*{}-([0-9]+)\".format(checkpoint_prefix), path)\n"," if regex_match and regex_match.groups():\n"," ordering_and_checkpoint_path.append((int(regex_match.groups()[0]), path))\n","\n"," checkpoints_sorted = sorted(ordering_and_checkpoint_path)\n"," checkpoints_sorted = [checkpoint[1] for checkpoint in checkpoints_sorted]\n"," return checkpoints_sorted\n","\n","\n","def _rotate_checkpoints(config, checkpoint_prefix=\"checkpoint\", use_mtime=False) -> None:\n"," if not config['save_total_limit']:\n"," return\n"," if config['save_total_limit'] <= 0:\n"," return\n","\n"," # Check if we should delete older checkpoint(s)\n"," checkpoints_sorted = _sorted_checkpoints(config, checkpoint_prefix, use_mtime)\n"," if len(checkpoints_sorted) <= config['save_total_limit']:\n"," return\n","\n"," number_of_checkpoints_to_delete = max(0, len(checkpoints_sorted) - config['save_total_limit'])\n"," checkpoints_to_be_deleted = checkpoints_sorted[:number_of_checkpoints_to_delete]\n"," for checkpoint in checkpoints_to_be_deleted:\n"," logger.info(\"Deleting older checkpoint [{}] due to config['save_total_limit']\".format(checkpoint))\n"," shutil.rmtree(checkpoint)\n","\n","\n","\n","\n","def mask_tokens(inputs: torch.Tensor, tokenizer: PreTrainedTokenizer, config) -> Tuple[torch.Tensor, torch.Tensor]:\n"," \"\"\"Prepare masked tokens inputs/labels for masked language modeling: 80% MASK, 10% random, 10% original.\"\"\"\n","\n"," mask_list = MASK_LIST['6']\n","\n"," if tokenizer.mask_token is None:\n"," raise ValueError(\n"," \"This tokenizer does not have a mask token which is necessary for masked language modeling. Remove the --mlm flag if you want to use this tokenizer.\"\n"," )\n","\n"," labels = inputs.clone()\n"," # We sample a few tokens in each sequence for masked-LM training (with probability args.mlm_probability defaults to 0.15 in Bert/RoBERTa)\n"," probability_matrix = torch.full(labels.shape, config['mlm_probability'])\n"," special_tokens_mask = [\n"," tokenizer.get_special_tokens_mask(val, already_has_special_tokens=True) for val in labels.tolist()\n"," ]\n"," probability_matrix.masked_fill_(torch.tensor(special_tokens_mask, dtype=torch.bool), value=0.0)\n"," if tokenizer.pad_token is not None:\n"," padding_mask = labels.eq(tokenizer.pad_token_id)\n"," probability_matrix.masked_fill_(padding_mask, value=0.0)\n","\n"," masked_indices = torch.bernoulli(probability_matrix).bool()\n","\n"," # Ensure masked_indices and probability_matrix are the same shape\n"," masks = deepcopy(masked_indices)\n"," for i, masked_index in enumerate(masks):\n"," # Ensure there are non-zero elements to avoid IndexError\n"," non_zero_indices = torch.where(probability_matrix[i] != 0)[0]\n"," if non_zero_indices.numel() == 0:\n"," # If no non-zero elements, skip this sequence\n"," continue\n","\n"," end = non_zero_indices.tolist()[-1]\n"," mask_centers = set(torch.where(masked_index == 1)[0].tolist())\n"," new_centers = deepcopy(mask_centers)\n"," for center in mask_centers:\n"," for mask_number in mask_list:\n"," current_index = center + mask_number\n"," if current_index <= end and current_index >= 1:\n"," new_centers.add(current_index)\n"," new_centers = list(new_centers)\n"," masked_indices[i][new_centers] = True\n","\n"," labels[~masked_indices] = -100 # We only compute loss on masked tokens\n","\n"," # 80% of the time, we replace masked input tokens with tokenizer.mask_token ([MASK])\n"," indices_replaced = torch.bernoulli(torch.full(labels.shape, 0.8)).bool() & masked_indices\n"," inputs[indices_replaced] = tokenizer.convert_tokens_to_ids(tokenizer.mask_token)\n","\n"," # 10% of the time, we replace masked input tokens with random word\n"," indices_random = torch.bernoulli(torch.full(labels.shape, 0.5)).bool() & masked_indices & ~indices_replaced\n"," random_words = torch.randint(len(tokenizer), labels.shape, dtype=torch.long)\n"," inputs[indices_random] = random_words[indices_random]\n","\n"," # The rest of the time (10% of the time) we keep the masked input tokens unchanged\n"," return inputs, labels\n","\n","import os\n","import torch\n","from torch.nn.utils.rnn import pad_sequence\n","from torch.utils.data import DataLoader, RandomSampler, DistributedSampler\n","from tqdm import tqdm, trange\n","from transformers import PreTrainedModel, PreTrainedTokenizer, AdamW, get_linear_schedule_with_warmup\n","from typing import List, Dict, Tuple\n","import wandb\n","import time\n","\n","def train(config, train_dataset, model: PreTrainedModel, tokenizer: PreTrainedTokenizer) -> Tuple[int, float]:\n"," \"\"\" Train the model \"\"\"\n"," if config['local_rank'] in [-1, 0]:\n"," tb_writer = SummaryWriter()\n","\n"," config['train_batch_size'] = config['per_gpu_train_batch_size'] * max(1, config['n_gpu'])\n","\n"," def collate(examples: List[torch.Tensor]):\n"," if tokenizer._pad_token is None:\n"," return pad_sequence(examples, batch_first=True)\n"," return pad_sequence(examples, batch_first=True, padding_value=tokenizer.pad_token_id)\n","\n"," train_sampler = RandomSampler(train_dataset) if config['local_rank'] == -1 else DistributedSampler(train_dataset)\n"," train_dataloader = DataLoader(\n"," train_dataset, sampler=train_sampler, batch_size=config['train_batch_size'], collate_fn=collate\n"," )\n","\n"," if config['max_steps'] > 0:\n"," t_total = config['max_steps']\n"," config['num_train_epochs'] = config['max_steps'] // (len(train_dataloader) // config['gradient_accumulation_steps']) + 1\n"," else:\n"," t_total = len(train_dataloader) // config['gradient_accumulation_steps'] * config['num_train_epochs']\n","\n"," # Prepare optimizer and schedule (linear warmup and decay)\n"," no_decay = [\"bias\", \"LayerNorm.weight\"]\n"," optimizer_grouped_parameters = [\n"," {\n"," \"params\": [p for n, p in model.named_parameters() if not any(nd in n for nd in no_decay)],\n"," \"weight_decay\": config['weight_decay'],\n"," },\n"," {\"params\": [p for n, p in model.named_parameters() if any(nd in n for nd in no_decay)], \"weight_decay\": 0.0},\n"," ]\n"," optimizer = AdamW(optimizer_grouped_parameters, lr=config['learning_rate'], eps=config['adam_epsilon'], betas=(config['beta1'],config['beta2']))\n"," scheduler = get_linear_schedule_with_warmup(\n"," optimizer, num_warmup_steps=2000, num_training_steps=t_total\n"," )\n","\n"," # Train!\n"," logger.info(\"***** Running training *****\")\n"," logger.info(\" Num examples = %d\", len(train_dataset))\n"," logger.info(\" Num Epochs = %d\", config['num_train_epochs'])\n"," logger.info(\" Instantaneous batch size per GPU = %d\", config['per_gpu_train_batch_size'])\n"," logger.info(\n"," \" Total train batch size (w. parallel, distributed & accumulation) = %d\",\n"," config['train_batch_size']\n"," * config['gradient_accumulation_steps']\n"," * (torch.distributed.get_world_size() if config['local_rank'] != -1 else 1),\n"," )\n"," logger.info(\" Gradient Accumulation steps = %d\", config['gradient_accumulation_steps'])\n"," logger.info(\" Total optimization steps = %d\", t_total)\n","\n"," global_step = 0\n"," epochs_trained = 0\n"," steps_trained_in_current_epoch = 0\n","\n"," tr_loss, logging_loss = 0.0, 0.0\n","\n"," model_to_resize = model.module if hasattr(model, \"module\") else model # Take care of distributed/parallel training\n"," model_to_resize.resize_token_embeddings(len(tokenizer))\n","\n"," model.zero_grad()\n"," train_iterator = trange(\n"," epochs_trained, int(config['num_train_epochs']), desc=\"Epoch\", disable=config['local_rank'] not in [-1, 0]\n"," )\n"," set_seed(config) # Added here for reproducibility\n","\n"," for epoch in train_iterator:\n"," epoch_start_time = time.time()\n"," epoch_iterator = tqdm(train_dataloader, desc=\"Iteration\", disable=config['local_rank'] not in [-1, 0])\n"," for step, batch in enumerate(epoch_iterator):\n","\n"," # Skip past any already trained steps if resuming training\n"," if steps_trained_in_current_epoch > 0:\n"," steps_trained_in_current_epoch -= 1\n"," continue\n","\n"," inputs, labels = mask_tokens(batch, tokenizer, config) if config['mlm'] else (batch, batch)\n","\n"," inputs = inputs.to(config['device'])\n"," labels = labels.to(config['device'])\n"," model.train()\n"," outputs = model(inputs, labels=labels) if config['mlm'] else model(inputs, labels=labels)\n"," loss = outputs[0] # model outputs are always tuple in transformers (see doc)\n","\n"," if config['n_gpu'] > 1:\n"," loss = loss.mean() # mean() to average on multi-gpu parallel training\n"," if config['gradient_accumulation_steps'] > 1:\n"," loss = loss / config['gradient_accumulation_steps']\n","\n"," loss.backward()\n","\n"," tr_loss += loss.item()\n"," if (step + 1) % config['gradient_accumulation_steps'] == 0:\n"," torch.nn.utils.clip_grad_norm_(model.parameters(), config['max_grad_norm'])\n"," optimizer.step()\n"," scheduler.step() # Update learning rate schedule\n"," model.zero_grad()\n"," global_step += 1\n","\n"," # Log metrics to wandb\n"," wandb.log({\"learning_rate\": scheduler.get_last_lr()[0], \"loss\": loss.item(), \"global_step\": global_step})\n","\n"," if config['local_rank'] in [-1, 0] and config['logging_steps'] > 0 and global_step % config['logging_steps'] == 0:\n"," # Log metrics\n"," if (\n"," config['local_rank'] == -1 and config['evaluate_during_training']\n"," ): # Only evaluate when single GPU otherwise metrics may not average well\n"," results = evaluate(config, model, tokenizer)\n"," for key, value in results.items():\n"," tb_writer.add_scalar(\"eval_{}\".format(key), value, global_step)\n"," wandb.log({f\"eval_{key}\": value, \"global_step\": global_step})\n"," tb_writer.add_scalar(\"lr\", scheduler.get_lr()[0], global_step)\n"," tb_writer.add_scalar(\"loss\", (tr_loss - logging_loss) / config['logging_steps'], global_step)\n"," logging_loss = tr_loss\n","\n"," if config['local_rank'] in [-1, 0] and config['save_steps'] > 0 and global_step % config['save_steps'] == 0:\n"," checkpoint_prefix = \"checkpoint\"\n"," # Save model checkpoint\n"," st = r\"/kaggle/working/output\"\n"," output_dir = os.path.join(st, \"{}-{}\".format(checkpoint_prefix, global_step))\n"," os.makedirs(output_dir, exist_ok=True)\n"," model_to_save = (\n"," model.module if hasattr(model, \"module\") else model\n"," ) # Take care of distributed/parallel training\n"," model_to_save.save_pretrained(output_dir)\n"," tokenizer.save_pretrained(output_dir)\n","\n"," torch.save(config, os.path.join(output_dir, \"training_args.bin\"))\n"," logger.info(\"Saving model checkpoint to %s\", output_dir)\n","\n"," _rotate_checkpoints(config, checkpoint_prefix)\n","\n"," torch.save(optimizer.state_dict(), os.path.join(output_dir, \"optimizer.pt\"))\n"," torch.save(scheduler.state_dict(), os.path.join(output_dir, \"scheduler.pt\"))\n"," logger.info(\"Saving optimizer and scheduler states to %s\", output_dir)\n","\n"," if config['max_steps'] > 0 and global_step > config['max_steps']:\n"," epoch_iterator.close()\n"," break\n"," if config['max_steps'] > 0 and global_step > config['max_steps']:\n"," train_iterator.close()\n"," break\n"," epoch_end_time = time.time()\n"," epoch_time = epoch_end_time - epoch_start_time\n"," # Log epoch time\n"," output_dir = r\"/kaggle/working/output\"\n"," logging.info(f'Epoch {epoch + 1}: Time {epoch_time:.4f}s')\n"," log_dir = os.path.join(output_dir, 'training_logs')\n"," os.makedirs(log_dir, exist_ok=True)\n"," file = os.path.join(log_dir,'log.txt')\n"," with open(file, 'a') as f:\n"," f.write(f\"Epoch {epoch + 1}/{config['num_train_epochs']}:\\n\")\n"," f.write(f\" Epoch Time: {epoch_time}\\n\")\n","\n"," # Log epoch time to wandb\n"," wandb.log({\"epoch_time\": epoch_time, \"epoch\": epoch + 1})\n","\n"," if config['local_rank'] in [-1, 0]:\n"," tb_writer.close()\n","\n"," return global_step, tr_loss / global_step\n","\n","\n","import os\n","import torch\n","from torch.nn.utils.rnn import pad_sequence\n","from torch.utils.data import DataLoader, SequentialSampler\n","from tqdm import tqdm\n","from transformers import PreTrainedModel, PreTrainedTokenizer\n","from typing import List, Dict\n","import wandb\n","\n","def evaluate(config, model: PreTrainedModel, tokenizer: PreTrainedTokenizer, prefix=\"\") -> Dict:\n"," # Loop to handle MNLI double evaluation (matched, mis-matched)\n"," eval_output_dir = config['output_dir']\n","\n"," eval_dataset = load_and_cache_examples(config, tokenizer, evaluate=True)\n","\n"," if config['local_rank'] in [-1, 0]:\n"," os.makedirs(eval_output_dir, exist_ok=True)\n","\n"," config['eval_batch_size'] = config['per_gpu_eval_batch_size'] * max(1, config['n_gpu'])\n"," # Note that DistributedSampler samples randomly\n","\n"," def collate(examples: List[torch.Tensor]):\n"," if tokenizer._pad_token is None:\n"," return pad_sequence(examples, batch_first=True)\n"," return pad_sequence(examples, batch_first=True, padding_value=tokenizer.pad_token_id)\n","\n"," eval_sampler = SequentialSampler(eval_dataset)\n"," eval_dataloader = DataLoader(\n"," eval_dataset, sampler=eval_sampler, batch_size=config['eval_batch_size'], collate_fn=collate\n"," )\n","\n"," # multi-gpu evaluate\n"," if config['n_gpu'] > 1 and not isinstance(model, torch.nn.DataParallel):\n"," model = torch.nn.DataParallel(model)\n","\n"," # Eval!\n"," logger.info(\"***** Running evaluation {} *****\".format(prefix))\n"," logger.info(\" Num examples = %d\", len(eval_dataset))\n"," logger.info(\" Batch size = %d\", config['eval_batch_size'])\n"," eval_loss = 0.0\n"," nb_eval_steps = 0\n"," model.eval()\n","\n"," for batch in tqdm(eval_dataloader, desc=\"Evaluating\"):\n"," inputs, labels = mask_tokens(batch, tokenizer, config) if config['mlm'] else (batch, batch)\n"," inputs = inputs.to(config['device'])\n"," labels = labels.to(config['device'])\n","\n"," with torch.no_grad():\n"," outputs = model(inputs, labels=labels) if config['mlm'] else model(inputs, labels=labels)\n"," lm_loss = outputs[0]\n"," eval_loss += lm_loss.mean().item()\n"," nb_eval_steps += 1\n","\n"," eval_loss = eval_loss / nb_eval_steps\n"," perplexity = torch.exp(torch.tensor(eval_loss))\n","\n"," result = {\"perplexity\": perplexity.item()}\n","\n"," # Log metrics to wandb\n"," wandb.log({\"eval perplexity\" : result})\n","\n"," output_eval_file = os.path.join(eval_output_dir, prefix, \"eval_results.txt\")\n"," with open(output_eval_file, \"a\") as writer:\n"," logger.info(\"***** Eval results {} *****\".format(prefix))\n"," for key in sorted(result.keys()):\n"," logger.info(\" %s = %s\", key, str(result[key]))\n"," writer.write(\"%s = %s\\n\" % (key, str(result[key])))\n","\n"," return result\n","\n","\n","import argparse\n","import os\n","import logging\n","import torch\n","import os\n","import logging\n","import torch\n","\n","def main(config):\n"," # Handle checkpoint continuation\n"," if config['should_continue']:\n"," sorted_checkpoints = _sorted_checkpoints(config)\n"," if len(sorted_checkpoints) == 0:\n"," raise ValueError(\"Used --should_continue but no checkpoint was found in --output_dir.\")\n"," else:\n"," config['model_name_or_path'] = sorted_checkpoints[-1]\n","\n"," output_dir = config.get('output_dir', './output')\n"," if (\n"," os.path.exists(output_dir)\n"," and os.listdir(output_dir)\n"," and config['do_train']\n"," and not config.get('overwrite_output_dir', False)\n"," ):\n"," raise ValueError(\n"," \"Output directory ({}) already exists and is not empty. Use --overwrite_output_dir to overcome.\".format(\n"," output_dir\n"," )\n"," )\n","\n"," # Setup CUDA, GPU & distributed training\n"," if config.get('local_rank', -1) == -1 or config.get('no_cuda', False):\n"," device = torch.device(\"cuda:0\" if torch.cuda.is_available() and not config.get('no_cuda', False) else \"cpu\")\n"," config['n_gpu'] = torch.cuda.device_count()\n"," else:\n"," torch.cuda.set_device(config.get('local_rank', 0))\n"," device = torch.device(\"cuda\", config.get('local_rank', 0))\n"," torch.distributed.init_process_group(backend=\"nccl\")\n"," config['n_gpu'] = 1\n"," config['device'] = device\n","\n"," # Setup logging\n"," logging.basicConfig(\n"," format=\"%(asctime)s - %(levelname)s - %(name)s - %(message)s\",\n"," datefmt=\"%m/%d/%Y %H:%M:%S\",\n"," level=logging.INFO if config.get('local_rank', -1) in [-1, 0] else logging.WARN,\n"," filename = 'app.log'\n"," )\n"," logger = logging.getLogger(__name__)\n"," logger.warning(\n"," \"Process rank: %s, device: %s, n_gpu: %s, distributed training: %s, 16-bits training: %s\",\n"," config.get('local_rank', -1),\n"," device,\n"," config['n_gpu'],\n"," bool(config.get('local_rank', -1) != -1),\n"," config.get('fp16', False),\n"," )\n","\n"," # Set seed\n"," set_seed(config)\n","\n"," # Load pretrained model and tokenizer\n"," if config.get('local_rank', -1) not in [-1, 0]:\n"," torch.distributed.barrier() # Barrier to make sure only the first process in distributed training download model & vocab\n","\n"," config_class, model_class, tokenizer_class = MODEL_CLASSES['dna']\n"," config_obj = config_class.from_pretrained('prajjwal1/bert-tiny', cache_dir=config.get('cache_dir', None))\n","\n"," tokenizer = tokenizer_class.from_pretrained('zhihan1996/DNA_bert_6', cache_dir=config.get('cache_dir', None))\n","\n"," if config.get('block_size', 512) <= 0:\n"," config['block_size'] = 512\n"," else:\n"," config['block_size'] = min(config['block_size'], 512)\n","\n"," if config.get('model_name_or_path'):\n","# model = model_class.from_pretrained(\n","# config['model_name_or_path'],\n","# from_tf=bool(\".ckpt\" in config['model_name_or_path']),\n","# config=config_obj,\n","# cache_dir=config.get('cache_dir', None),\n"," pass\n"," else:\n"," logger.info(\"Training new model from scratch\")\n"," model = model_class(config=config_obj)\n","\n"," model.to(config['device'])\n","\n"," if config.get('local_rank', -1) == 0:\n"," torch.distributed.barrier()\n","\n"," logger.info(\"Training/evaluation parameters %s\", config)\n","\n"," # Training\n"," if config.get('do_train', False):\n"," if config.get('local_rank', -1) not in [-1, 0]:\n"," torch.distributed.barrier()\n","\n"," train_dataset = load_and_cache_examples(config, tokenizer, evaluate=False)\n","\n"," if config.get('local_rank', -1) == 0:\n"," torch.distributed.barrier()\n","\n"," global_step, tr_loss = train(config, train_dataset, model, tokenizer)\n"," logger.info(\" global_step = %s, average loss = %s\", global_step, tr_loss)\n","\n"," # Save and reload model\n"," if config.get('do_train', False) and (config.get('local_rank', -1) == -1 or torch.distributed.get_rank() == 0):\n"," if config.get('local_rank', -1) in [-1, 0]:\n"," os.makedirs(output_dir, exist_ok=True)\n","\n"," logger.info(\"Saving model checkpoint to %s\", output_dir)\n"," model_to_save = (\n"," model.module if hasattr(model, \"module\") else model\n"," )\n"," model_to_save.save_pretrained(output_dir)\n"," tokenizer.save_pretrained(output_dir)\n"," torch.save(config, os.path.join(output_dir, \"training_args.bin\"))\n","\n"," model = model_class.from_pretrained(output_dir)\n"," tokenizer = tokenizer_class.from_pretrained(output_dir)\n"," model.to(config['device'])\n","\n"," # Evaluation\n"," results = {}\n"," if config.get('do_eval', False) and config.get('local_rank', -1) in [-1, 0]:\n"," checkpoints = [output_dir]\n"," if config.get('eval_all_checkpoints', False):\n"," checkpoints = list(\n"," os.path.dirname(c) for c in sorted(glob.glob(output_dir + \"/**/\" + WEIGHTS_NAME, recursive=True))\n"," )\n"," logging.getLogger(\"transformers.modeling_utils\").setLevel(logging.WARN)\n"," logger.info(\"Evaluate the following checkpoints: %s\", checkpoints)\n"," for checkpoint in checkpoints:\n"," global_step = checkpoint.split(\"-\")[-1] if len(checkpoints) > 1 else \"\"\n"," prefix = checkpoint.split(\"/\")[-1] if checkpoint.find(\"checkpoint\") != -1 else \"\"\n","\n"," model = model_class.from_pretrained(checkpoint)\n"," model.to(config['device'])\n"," result = evaluate(config, model, tokenizer, prefix=prefix)\n"," result = dict((k + \"_{}\".format(global_step), v) for k, v in result.items())\n"," results.update(result)\n","\n"," return results\n","\n","# Example configuration dictionary\n","config = {\n"," 'line_by_line': True,\n"," 'should_continue': False,#use if you have a checkpoint present or it will throw error\n"," 'mlm': True,\n"," 'mlm_probability': 0.15,\n"," 'config_name': None,\n"," 'tokenizer_name': None,\n"," 'cache_dir': None,\n"," 'block_size': 512,\n"," 'do_train': True,\n"," 'do_eval': True,\n"," 'evaluate_during_training': True,\n"," 'per_gpu_train_batch_size': 175,\n"," 'per_gpu_eval_batch_size': 25,\n"," 'gradient_accumulation_steps': 1,\n"," 'learning_rate': 4e-4,\n"," 'weight_decay': 0.01,\n"," 'adam_epsilon': 1e-6,\n"," 'beta1': 0.9,\n"," 'beta2': 0.98,\n"," 'max_grad_norm': 1.0,\n"," 'num_train_epochs': 2000,\n"," 'max_steps': -1,\n"," 'warmup_steps': 100,\n"," 'logging_steps': 200,\n"," 'save_steps': 1000,\n"," 'save_total_limit': 10,\n"," 'eval_all_checkpoints': False,\n"," 'no_cuda': False,\n"," 'overwrite_output_dir': True,\n"," 'overwrite_cache': False,\n"," 'seed': 42,\n"," 'n_process': 1,\n"," 'fp16': False,\n"," 'fp16_opt_level': 'O1',\n"," 'local_rank': -1,\n"," 'server_ip': '',\n"," 'server_port': '',\n"," 'output_dir': './output',\n"," 'device':'cuda'\n","}\n","\n","if __name__ == \"__main__\":\n"," main(config)\n"]},{"cell_type":"code","execution_count":null,"metadata":{},"outputs":[],"source":[]}],"metadata":{"kaggle":{"accelerator":"gpu","dataSources":[{"datasetId":5477436,"sourceId":9095316,"sourceType":"datasetVersion"}],"dockerImageVersionId":30732,"isGpuEnabled":true,"isInternetEnabled":true,"language":"python","sourceType":"notebook"},"kernelspec":{"display_name":"Python 3","language":"python","name":"python3"},"language_info":{"codemirror_mode":{"name":"ipython","version":3},"file_extension":".py","mimetype":"text/x-python","name":"python","nbconvert_exporter":"python","pygments_lexer":"ipython3","version":"3.10.13"}},"nbformat":4,"nbformat_minor":4} diff --git a/pytorch_model.bin b/Pretrained Base Model/pytorch_model.bin similarity index 100% rename from pytorch_model.bin rename to Pretrained Base Model/pytorch_model.bin diff --git a/scheduler.pt b/Pretrained Base Model/scheduler.pt similarity index 100% rename from scheduler.pt rename to Pretrained Base Model/scheduler.pt diff --git a/Pretrained Base Model/special_tokens_map.json b/Pretrained Base Model/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..a8b3208c2884c4efb86e49300fdd3dc877220cdf --- /dev/null +++ b/Pretrained Base Model/special_tokens_map.json @@ -0,0 +1,7 @@ +{ + "cls_token": "[CLS]", + "mask_token": "[MASK]", + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "unk_token": "[UNK]" +} diff --git a/tokenizer.json b/Pretrained Base Model/tokenizer.json similarity index 100% rename from tokenizer.json rename to Pretrained Base Model/tokenizer.json diff --git a/Pretrained Base Model/tokenizer_config.json b/Pretrained Base Model/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..604e2b1b29af1dcb3920d2828b6a90276e5e3d02 --- /dev/null +++ b/Pretrained Base Model/tokenizer_config.json @@ -0,0 +1,58 @@ +{ + "added_tokens_decoder": { + "0": { + "content": "[PAD]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "[UNK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "[CLS]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "[SEP]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "4": { + "content": "[MASK]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "clean_up_tokenization_spaces": true, + "cls_token": "[CLS]", + "do_basic_tokenize": true, + "do_lower_case": false, + "mask_token": "[MASK]", + "max_len": 512, + "model_max_length": 512, + "never_split": null, + "pad_token": "[PAD]", + "sep_token": "[SEP]", + "strip_accents": null, + "tokenize_chinese_chars": true, + "tokenizer_class": "BertTokenizer", + "unk_token": "[UNK]" +} diff --git a/training_args.bin b/Pretrained Base Model/training_args.bin similarity index 100% rename from training_args.bin rename to Pretrained Base Model/training_args.bin diff --git a/Pretrained Base Model/training_perplexity.txt b/Pretrained Base Model/training_perplexity.txt new file mode 100644 index 0000000000000000000000000000000000000000..0f1ae5dc64e2b48eea1003514996fc6d34a54158 --- /dev/null +++ b/Pretrained Base Model/training_perplexity.txt @@ -0,0 +1,632 @@ +perplexity = 4097.30126953125 +perplexity = 3889.51171875 +perplexity = 3526.33544921875 +perplexity = 3295.367919921875 +perplexity = 3137.015869140625 +perplexity = 3020.431396484375 +perplexity = 2919.982666015625 +perplexity = 2830.441162109375 +perplexity = 2738.419921875 +perplexity = 2649.371337890625 +perplexity = 2565.232421875 +perplexity = 2494.44189453125 +perplexity = 2419.830810546875 +perplexity = 2360.313232421875 +perplexity = 2301.50439453125 +perplexity = 2249.9638671875 +perplexity = 2196.739013671875 +perplexity = 2154.25146484375 +perplexity = 2110.0341796875 +perplexity = 2076.258544921875 +perplexity = 2033.0528564453125 +perplexity = 1990.3134765625 +perplexity = 1963.23486328125 +perplexity = 1934.708251953125 +perplexity = 1918.181884765625 +perplexity = 1886.7474365234375 +perplexity = 1864.243408203125 +perplexity = 1841.686279296875 +perplexity = 1823.6138916015625 +perplexity = 1812.8128662109375 +perplexity = 1806.9012451171875 +perplexity = 1795.9886474609375 +perplexity = 1783.2694091796875 +perplexity = 1778.0567626953125 +perplexity = 1762.7886962890625 +perplexity = 1770.2020263671875 +perplexity = 1769.548828125 +perplexity = 1764.7591552734375 +perplexity = 1771.6181640625 +perplexity = 1764.88037109375 +perplexity = 1766.447998046875 +perplexity = 1764.0726318359375 +perplexity = 1766.2138671875 +perplexity = 1759.2366943359375 +perplexity = 1758.0877685546875 +perplexity = 1749.7279052734375 +perplexity = 1760.657470703125 +perplexity = 1759.4505615234375 +perplexity = 1752.9078369140625 +perplexity = 1754.3302001953125 +perplexity = 1752.7164306640625 +perplexity = 1746.3895263671875 +perplexity = 1750.0198974609375 +perplexity = 1738.935791015625 +perplexity = 1742.3553466796875 +perplexity = 1744.8297119140625 +perplexity = 1739.843994140625 +perplexity = 1739.034423828125 +perplexity = 1733.880126953125 +perplexity = 1738.0587158203125 +perplexity = 1734.2861328125 +perplexity = 1734.0718994140625 +perplexity = 1730.06298828125 +perplexity = 1728.1978759765625 +perplexity = 1727.6689453125 +perplexity = 1732.3834228515625 +perplexity = 1724.8282470703125 +perplexity = 1721.9356689453125 +perplexity = 1721.893798828125 +perplexity = 1725.103759765625 +perplexity = 1724.958984375 +perplexity = 1728.1756591796875 +perplexity = 1728.8193359375 +perplexity = 1720.2425537109375 +perplexity = 1719.079833984375 +perplexity = 1724.18603515625 +perplexity = 1720.2220458984375 +perplexity = 1716.7099609375 +perplexity = 1719.51025390625 +perplexity = 1714.51025390625 +perplexity = 1717.4345703125 +perplexity = 1708.955078125 +perplexity = 1702.4791259765625 +perplexity = 1708.8116455078125 +perplexity = 1700.252197265625 +perplexity = 1703.018310546875 +perplexity = 1702.582275390625 +perplexity = 1692.8968505859375 +perplexity = 1693.2779541015625 +perplexity = 1686.2598876953125 +perplexity = 1675.4542236328125 +perplexity = 1679.2421875 +perplexity = 1667.2232666015625 +perplexity = 1656.8358154296875 +perplexity = 1652.140869140625 +perplexity = 1638.50390625 +perplexity = 1629.5078125 +perplexity = 1615.3712158203125 +perplexity = 1598.87060546875 +perplexity = 1607.5672607421875 +perplexity = 1583.057861328125 +perplexity = 1577.3177490234375 +perplexity = 1566.0992431640625 +perplexity = 1559.6209716796875 +perplexity = 1545.6800537109375 +perplexity = 1540.5308837890625 +perplexity = 1527.4312744140625 +perplexity = 1511.7518310546875 +perplexity = 1505.3907470703125 +perplexity = 1493.48046875 +perplexity = 1475.6861572265625 +perplexity = 1465.27294921875 +perplexity = 1456.11181640625 +perplexity = 1439.156494140625 +perplexity = 1428.69384765625 +perplexity = 1412.7030029296875 +perplexity = 1396.4534912109375 +perplexity = 1394.1661376953125 +perplexity = 1377.3953857421875 +perplexity = 1374.0616455078125 +perplexity = 1359.943603515625 +perplexity = 1347.80419921875 +perplexity = 1332.76708984375 +perplexity = 1323.6829833984375 +perplexity = 1312.6275634765625 +perplexity = 1298.4681396484375 +perplexity = 1285.4014892578125 +perplexity = 1270.8653564453125 +perplexity = 1259.03955078125 +perplexity = 1241.2227783203125 +perplexity = 1233.4171142578125 +perplexity = 1216.78759765625 +perplexity = 1215.6463623046875 +perplexity = 1187.6070556640625 +perplexity = 1187.9417724609375 +perplexity = 1179.5860595703125 +perplexity = 1163.0206298828125 +perplexity = 1152.4566650390625 +perplexity = 1149.894775390625 +perplexity = 1139.0035400390625 +perplexity = 1130.7493896484375 +perplexity = 1112.0308837890625 +perplexity = 1115.7047119140625 +perplexity = 1101.40673828125 +perplexity = 1102.280029296875 +perplexity = 1088.66845703125 +perplexity = 1087.458984375 +perplexity = 1074.3414306640625 +perplexity = 1064.43310546875 +perplexity = 1060.2852783203125 +perplexity = 1063.317138671875 +perplexity = 1048.0140380859375 +perplexity = 1039.9342041015625 +perplexity = 1035.987060546875 +perplexity = 1042.27294921875 +perplexity = 1027.7769775390625 +perplexity = 1017.9442749023438 +perplexity = 1017.1087646484375 +perplexity = 1008.634033203125 +perplexity = 1016.2051391601562 +perplexity = 1004.91162109375 +perplexity = 1004.0748291015625 +perplexity = 994.5106201171875 +perplexity = 986.6381225585938 +perplexity = 984.4993286132812 +perplexity = 985.1727905273438 +perplexity = 973.2727661132812 +perplexity = 973.6124877929688 +perplexity = 968.1450805664062 +perplexity = 965.5062255859375 +perplexity = 962.4324951171875 +perplexity = 953.9237670898438 +perplexity = 957.7413940429688 +perplexity = 951.288330078125 +perplexity = 947.0551147460938 +perplexity = 937.9320678710938 +perplexity = 940.9132080078125 +perplexity = 938.4389038085938 +perplexity = 936.766845703125 +perplexity = 931.7275390625 +perplexity = 928.2318115234375 +perplexity = 924.981201171875 +perplexity = 923.304931640625 +perplexity = 926.3782348632812 +perplexity = 912.8396606445312 +perplexity = 919.0626220703125 +perplexity = 917.5213012695312 +perplexity = 908.4006958007812 +perplexity = 901.25537109375 +perplexity = 913.0925903320312 +perplexity = 903.393798828125 +perplexity = 907.5404052734375 +perplexity = 906.0335083007812 +perplexity = 898.93896484375 +perplexity = 902.8929443359375 +perplexity = 895.4945678710938 +perplexity = 889.3595581054688 +perplexity = 895.7507934570312 +perplexity = 886.5383911132812 +perplexity = 883.4834594726562 +perplexity = 884.7144775390625 +perplexity = 884.1497802734375 +perplexity = 876.6148681640625 +perplexity = 879.2576904296875 +perplexity = 865.92578125 +perplexity = 873.0511474609375 +perplexity = 874.8413696289062 +perplexity = 867.0955200195312 +perplexity = 868.0614624023438 +perplexity = 865.228271484375 +perplexity = 870.84375 +perplexity = 864.0964965820312 +perplexity = 854.9404907226562 +perplexity = 858.8663330078125 +perplexity = 860.0027465820312 +perplexity = 852.3817138671875 +perplexity = 856.833740234375 +perplexity = 854.7676391601562 +perplexity = 851.7848510742188 +perplexity = 847.0742797851562 +perplexity = 853.3854370117188 +perplexity = 847.869140625 +perplexity = 849.9744873046875 +perplexity = 838.598388671875 +perplexity = 842.910888671875 +perplexity = 847.8323364257812 +perplexity = 840.2562866210938 +perplexity = 845.647216796875 +perplexity = 831.4588012695312 +perplexity = 837.9963989257812 +perplexity = 838.064697265625 +perplexity = 837.432373046875 +perplexity = 834.7365112304688 +perplexity = 828.8058471679688 +perplexity = 835.688720703125 +perplexity = 834.3087158203125 +perplexity = 824.8424072265625 +perplexity = 826.22802734375 +perplexity = 828.3475952148438 +perplexity = 820.148193359375 +perplexity = 821.657958984375 +perplexity = 820.6751708984375 +perplexity = 816.8401489257812 +perplexity = 816.303955078125 +perplexity = 816.2615356445312 +perplexity = 814.29443359375 +perplexity = 814.1934814453125 +perplexity = 811.1499633789062 +perplexity = 811.02734375 +perplexity = 810.0259399414062 +perplexity = 812.5529174804688 +perplexity = 805.06787109375 +perplexity = 810.24462890625 +perplexity = 813.40185546875 +perplexity = 797.2235717773438 +perplexity = 806.0984497070312 +perplexity = 800.6862182617188 +perplexity = 807.9533081054688 +perplexity = 802.473876953125 +perplexity = 795.2678833007812 +perplexity = 802.7169189453125 +perplexity = 803.1863403320312 +perplexity = 795.78564453125 +perplexity = 792.6045532226562 +perplexity = 800.8038330078125 +perplexity = 792.7977294921875 +perplexity = 785.1489868164062 +perplexity = 793.0128784179688 +perplexity = 795.5879516601562 +perplexity = 786.5272216796875 +perplexity = 785.283447265625 +perplexity = 788.735595703125 +perplexity = 787.48828125 +perplexity = 784.8016357421875 +perplexity = 793.6033325195312 +perplexity = 779.1310424804688 +perplexity = 781.7360229492188 +perplexity = 783.77734375 +perplexity = 781.5269775390625 +perplexity = 772.8794555664062 +perplexity = 780.3748168945312 +perplexity = 781.0928955078125 +perplexity = 775.3518676757812 +perplexity = 778.0729370117188 +perplexity = 775.8067626953125 +perplexity = 769.55908203125 +perplexity = 773.7785034179688 +perplexity = 774.0548706054688 +perplexity = 769.877685546875 +perplexity = 769.142333984375 +perplexity = 770.5545654296875 +perplexity = 771.4949951171875 +perplexity = 768.7720336914062 +perplexity = 763.9771728515625 +perplexity = 769.0983276367188 +perplexity = 765.7996826171875 +perplexity = 764.1268920898438 +perplexity = 765.4521484375 +perplexity = 765.0846557617188 +perplexity = 762.225830078125 +perplexity = 759.385986328125 +perplexity = 762.2988891601562 +perplexity = 766.455810546875 +perplexity = 759.6376953125 +perplexity = 755.9061889648438 +perplexity = 760.9398803710938 +perplexity = 761.5562744140625 +perplexity = 769.5205688476562 +perplexity = 757.2493286132812 +perplexity = 760.1127319335938 +perplexity = 754.3812255859375 +perplexity = 756.7594604492188 +perplexity = 755.49609375 +perplexity = 758.4823608398438 +perplexity = 758.2335815429688 +perplexity = 752.42333984375 +perplexity = 748.8626098632812 +perplexity = 751.6122436523438 +perplexity = 753.38623046875 +perplexity = 746.7591552734375 +perplexity = 746.4358520507812 +perplexity = 744.9201049804688 +perplexity = 746.9535522460938 +perplexity = 744.5912475585938 +perplexity = 744.6384887695312 +perplexity = 748.3910522460938 +perplexity = 745.6962890625 +perplexity = 743.9268798828125 +perplexity = 742.87158203125 +perplexity = 749.1343994140625 +perplexity = 745.950927734375 +perplexity = 742.9410400390625 +perplexity = 741.3800048828125 +perplexity = 738.6696166992188 +perplexity = 734.9588012695312 +perplexity = 739.123779296875 +perplexity = 742.315673828125 +perplexity = 737.37841796875 +perplexity = 740.1405639648438 +perplexity = 740.1345825195312 +perplexity = 740.5550537109375 +perplexity = 734.6732177734375 +perplexity = 737.6400756835938 +perplexity = 735.7421264648438 +perplexity = 737.5064697265625 +perplexity = 734.6826782226562 +perplexity = 739.8360595703125 +perplexity = 739.0550537109375 +perplexity = 730.4451293945312 +perplexity = 730.5785522460938 +perplexity = 735.9241943359375 +perplexity = 734.8557739257812 +perplexity = 728.035400390625 +perplexity = 731.4681396484375 +perplexity = 730.21533203125 +perplexity = 732.2264404296875 +perplexity = 731.8460083007812 +perplexity = 728.0808715820312 +perplexity = 725.808349609375 +perplexity = 723.0493774414062 +perplexity = 730.1575317382812 +perplexity = 726.298583984375 +perplexity = 728.7171630859375 +perplexity = 724.2322387695312 +perplexity = 725.920166015625 +perplexity = 728.33056640625 +perplexity = 723.8220825195312 +perplexity = 725.5938110351562 +perplexity = 722.6450805664062 +perplexity = 726.6803588867188 +perplexity = 715.7059326171875 +perplexity = 723.8348388671875 +perplexity = 720.3307495117188 +perplexity = 718.7459716796875 +perplexity = 718.8748168945312 +perplexity = 726.8934936523438 +perplexity = 724.7784423828125 +perplexity = 723.37353515625 +perplexity = 717.03369140625 +perplexity = 720.3506469726562 +perplexity = 722.4548950195312 +perplexity = 713.6240844726562 +perplexity = 712.7484130859375 +perplexity = 723.8514404296875 +perplexity = 716.4298095703125 +perplexity = 723.9663696289062 +perplexity = 718.4139404296875 +perplexity = 716.2057495117188 +perplexity = 715.391357421875 +perplexity = 718.1132202148438 +perplexity = 715.9711303710938 +perplexity = 719.4235229492188 +perplexity = 716.283935546875 +perplexity = 715.1907958984375 +perplexity = 713.788818359375 +perplexity = 716.7639770507812 +perplexity = 716.881591796875 +perplexity = 709.524658203125 +perplexity = 713.6676635742188 +perplexity = 713.5792236328125 +perplexity = 712.0262451171875 +perplexity = 713.06591796875 +perplexity = 710.5433959960938 +perplexity = 708.3961791992188 +perplexity = 716.1753540039062 +perplexity = 713.9188842773438 +perplexity = 709.2042846679688 +perplexity = 710.8134765625 +perplexity = 710.3201293945312 +perplexity = 710.0912475585938 +perplexity = 709.7077026367188 +perplexity = 708.0504150390625 +perplexity = 710.4783325195312 +perplexity = 708.757080078125 +perplexity = 708.6522827148438 +perplexity = 704.4107055664062 +perplexity = 709.2726440429688 +perplexity = 700.8388671875 +perplexity = 706.7884521484375 +perplexity = 705.8982543945312 +perplexity = 709.3084716796875 +perplexity = 709.2462768554688 +perplexity = 706.7459716796875 +perplexity = 707.5154418945312 +perplexity = 702.7051391601562 +perplexity = 702.003173828125 +perplexity = 702.85595703125 +perplexity = 704.724853515625 +perplexity = 701.848876953125 +perplexity = 700.2529907226562 +perplexity = 705.7706909179688 +perplexity = 702.3356323242188 +perplexity = 699.7833251953125 +perplexity = 709.2374267578125 +perplexity = 700.3665161132812 +perplexity = 697.9564819335938 +perplexity = 701.7254028320312 +perplexity = 699.7839965820312 +perplexity = 697.671630859375 +perplexity = 699.4877319335938 +perplexity = 698.4152221679688 +perplexity = 700.170166015625 +perplexity = 697.9611206054688 +perplexity = 697.4979858398438 +perplexity = 695.376953125 +perplexity = 696.5178833007812 +perplexity = 699.400390625 +perplexity = 691.5969848632812 +perplexity = 696.432861328125 +perplexity = 698.1888427734375 +perplexity = 698.9163208007812 +perplexity = 698.3379516601562 +perplexity = 692.3548583984375 +perplexity = 694.0813598632812 +perplexity = 693.7670288085938 +perplexity = 696.5032958984375 +perplexity = 693.0125122070312 +perplexity = 692.024169921875 +perplexity = 691.3444213867188 +perplexity = 699.4544067382812 +perplexity = 690.9976806640625 +perplexity = 696.2356567382812 +perplexity = 696.6673583984375 +perplexity = 696.01220703125 +perplexity = 692.17431640625 +perplexity = 691.1258544921875 +perplexity = 694.3382568359375 +perplexity = 686.1092529296875 +perplexity = 688.8964233398438 +perplexity = 688.9824829101562 +perplexity = 697.349365234375 +perplexity = 691.0991821289062 +perplexity = 694.752197265625 +perplexity = 696.8530883789062 +perplexity = 689.8964233398438 +perplexity = 690.2626342773438 +perplexity = 686.514404296875 +perplexity = 694.797607421875 +perplexity = 686.0797729492188 +perplexity = 686.7930297851562 +perplexity = 690.0543212890625 +perplexity = 684.1278686523438 +perplexity = 689.4935302734375 +perplexity = 686.6289672851562 +perplexity = 686.4757690429688 +perplexity = 683.0333251953125 +perplexity = 688.2509155273438 +perplexity = 690.3225708007812 +perplexity = 685.1326904296875 +perplexity = 684.9749145507812 +perplexity = 690.3807983398438 +perplexity = 686.488525390625 +perplexity = 688.7791748046875 +perplexity = 684.1370239257812 +perplexity = 680.156982421875 +perplexity = 686.28857421875 +perplexity = 676.8685302734375 +perplexity = 684.5249633789062 +perplexity = 683.5845947265625 +perplexity = 683.271728515625 +perplexity = 685.9110107421875 +perplexity = 684.298828125 +perplexity = 689.13232421875 +perplexity = 684.047607421875 +perplexity = 677.2233276367188 +perplexity = 683.1981201171875 +perplexity = 685.051025390625 +perplexity = 682.6682739257812 +perplexity = 683.8131103515625 +perplexity = 681.3905029296875 +perplexity = 677.1112670898438 +perplexity = 680.7676391601562 +perplexity = 682.9378662109375 +perplexity = 678.4382934570312 +perplexity = 680.96630859375 +perplexity = 678.3389892578125 +perplexity = 682.350341796875 +perplexity = 681.5877685546875 +perplexity = 676.9805297851562 +perplexity = 677.1209716796875 +perplexity = 675.705322265625 +perplexity = 673.6614990234375 +perplexity = 678.5897216796875 +perplexity = 676.1775512695312 +perplexity = 673.6602172851562 +perplexity = 676.404541015625 +perplexity = 673.1034545898438 +perplexity = 675.330078125 +perplexity = 680.4985961914062 +perplexity = 676.9595336914062 +perplexity = 679.2002563476562 +perplexity = 678.2283325195312 +perplexity = 678.9165649414062 +perplexity = 675.3954467773438 +perplexity = 673.3927001953125 +perplexity = 678.1804809570312 +perplexity = 679.5407104492188 +perplexity = 673.0979614257812 +perplexity = 675.3980102539062 +perplexity = 674.927978515625 +perplexity = 676.1336669921875 +perplexity = 676.2103881835938 +perplexity = 673.2225341796875 +perplexity = 679.30029296875 +perplexity = 671.8086547851562 +perplexity = 677.78125 +perplexity = 669.0200805664062 +perplexity = 677.345703125 +perplexity = 673.3502807617188 +perplexity = 675.8529052734375 +perplexity = 670.4559326171875 +perplexity = 672.568603515625 +perplexity = 675.8568115234375 +perplexity = 672.568603515625 +perplexity = 668.1405029296875 +perplexity = 674.1380615234375 +perplexity = 674.3469848632812 +perplexity = 667.2200927734375 +perplexity = 674.0637817382812 +perplexity = 671.3442993164062 +perplexity = 676.8388061523438 +perplexity = 670.9989624023438 +perplexity = 672.5984497070312 +perplexity = 673.75146484375 +perplexity = 673.2205810546875 +perplexity = 667.3836059570312 +perplexity = 666.2910766601562 +perplexity = 669.9349975585938 +perplexity = 668.7027587890625 +perplexity = 671.3587036132812 +perplexity = 668.9773559570312 +perplexity = 669.8477783203125 +perplexity = 670.2855224609375 +perplexity = 672.2107543945312 +perplexity = 668.2236328125 +perplexity = 668.7728881835938 +perplexity = 666.85302734375 +perplexity = 670.398681640625 +perplexity = 663.4403076171875 +perplexity = 663.901123046875 +perplexity = 668.1328735351562 +perplexity = 671.3743896484375 +perplexity = 672.038330078125 +perplexity = 668.76904296875 +perplexity = 668.9728393554688 +perplexity = 666.6641845703125 +perplexity = 669.8043823242188 +perplexity = 666.966552734375 +perplexity = 665.3325805664062 +perplexity = 671.5921020507812 +perplexity = 664.1832275390625 +perplexity = 668.7359008789062 +perplexity = 666.8956298828125 +perplexity = 667.67138671875 +perplexity = 665.9000854492188 +perplexity = 661.087646484375 +perplexity = 666.1735229492188 +perplexity = 662.1514892578125 +perplexity = 662.7415771484375 +perplexity = 660.4442749023438 +perplexity = 666.3883056640625 +perplexity = 663.0298461914062 +perplexity = 666.9531860351562 +perplexity = 663.8349609375 +perplexity = 670.060546875 +perplexity = 668.3950805664062 +perplexity = 665.4959716796875 +perplexity = 659.3328857421875 +perplexity = 663.5997924804688 +perplexity = 667.8751831054688 +perplexity = 666.0795288085938 +perplexity = 660.873046875 +perplexity = 665.3709716796875 +perplexity = 665.7000732421875 +perplexity = 660.39453125 +perplexity = 661.1992797851562 +perplexity = 662.1780395507812 +perplexity = 666.244384765625 +perplexity = 663.7478637695312 +perplexity = 663.67919921875 +perplexity = 664.6365966796875 +perplexity = 667.9515991210938 +perplexity = 660.4209594726562 +perplexity = 663.1680297851562 +perplexity = 663.3846435546875 +perplexity = 659.9833984375 +perplexity = 659.2014770507812 +perplexity = 661.0914306640625 +perplexity = 659.4888916015625 +perplexity = 662.294189453125 +perplexity = 665.7584838867188 +perplexity = 664.4122314453125 diff --git a/Pretrained Base Model/training_times.txt b/Pretrained Base Model/training_times.txt new file mode 100644 index 0000000000000000000000000000000000000000..8e18ff247b826739ca5eef428a114d1db1eef810 --- /dev/null +++ b/Pretrained Base Model/training_times.txt @@ -0,0 +1,2940 @@ +Epoch 1/2000: + Epoch Time: 25.618987560272217 +Epoch 2/2000: + Epoch Time: 24.573694467544556 +Epoch 3/2000: + Epoch Time: 35.3754186630249 +Epoch 4/2000: + Epoch Time: 24.802497625350952 +Epoch 5/2000: + Epoch Time: 35.36515140533447 +Epoch 6/2000: + Epoch Time: 24.69466257095337 +Epoch 7/2000: + Epoch Time: 35.33469367027283 +Epoch 8/2000: + Epoch Time: 24.650006771087646 +Epoch 9/2000: + Epoch Time: 24.586905002593994 +Epoch 10/2000: + Epoch Time: 35.19983124732971 +Epoch 11/2000: + Epoch Time: 24.58481192588806 +Epoch 12/2000: + Epoch Time: 35.50823616981506 +Epoch 13/2000: + Epoch Time: 24.70728063583374 +Epoch 14/2000: + Epoch Time: 35.2906014919281 +Epoch 15/2000: + Epoch Time: 24.70067548751831 +Epoch 16/2000: + Epoch Time: 24.56070590019226 +Epoch 17/2000: + Epoch Time: 35.35891389846802 +Epoch 18/2000: + Epoch Time: 24.590630531311035 +Epoch 19/2000: + Epoch Time: 35.384915351867676 +Epoch 20/2000: + Epoch Time: 24.682771682739258 +Epoch 21/2000: + Epoch Time: 35.5338773727417 +Epoch 22/2000: + Epoch Time: 24.86756157875061 +Epoch 23/2000: + Epoch Time: 24.770934581756592 +Epoch 24/2000: + Epoch Time: 35.46073770523071 +Epoch 25/2000: + Epoch Time: 24.6592800617218 +Epoch 26/2000: + Epoch Time: 35.31223821640015 +Epoch 27/2000: + Epoch Time: 24.661009788513184 +Epoch 28/2000: + Epoch Time: 35.2975697517395 +Epoch 29/2000: + Epoch Time: 24.56358051300049 +Epoch 30/2000: + Epoch Time: 24.603206396102905 +Epoch 31/2000: + Epoch Time: 35.305848121643066 +Epoch 32/2000: + Epoch Time: 24.639580011367798 +Epoch 33/2000: + Epoch Time: 35.37826657295227 +Epoch 34/2000: + Epoch Time: 24.67519211769104 +Epoch 35/2000: + Epoch Time: 35.40761208534241 +Epoch 36/2000: + Epoch Time: 24.685372591018677 +Epoch 37/2000: + Epoch Time: 24.686089515686035 +Epoch 38/2000: + Epoch Time: 35.16649150848389 +Epoch 39/2000: + Epoch Time: 24.72599458694458 +Epoch 40/2000: + Epoch Time: 35.32401752471924 +Epoch 41/2000: + Epoch Time: 24.690605640411377 +Epoch 42/2000: + Epoch Time: 35.33947134017944 +Epoch 43/2000: + Epoch Time: 24.619959592819214 +Epoch 44/2000: + Epoch Time: 24.693222045898438 +Epoch 45/2000: + Epoch Time: 35.26632833480835 +Epoch 46/2000: + Epoch Time: 24.63942313194275 +Epoch 47/2000: + Epoch Time: 35.294933557510376 +Epoch 48/2000: + Epoch Time: 24.665773391723633 +Epoch 49/2000: + Epoch Time: 35.628381967544556 +Epoch 50/2000: + Epoch Time: 24.830758810043335 +Epoch 51/2000: + Epoch Time: 24.797037839889526 +Epoch 52/2000: + Epoch Time: 35.561861753463745 +Epoch 53/2000: + Epoch Time: 24.845026969909668 +Epoch 54/2000: + Epoch Time: 35.336201190948486 +Epoch 55/2000: + Epoch Time: 24.68093204498291 +Epoch 56/2000: + Epoch Time: 35.3622887134552 +Epoch 57/2000: + Epoch Time: 24.70694398880005 +Epoch 58/2000: + Epoch Time: 24.683655738830566 +Epoch 59/2000: + Epoch Time: 35.34082579612732 +Epoch 60/2000: + Epoch Time: 24.726755142211914 +Epoch 61/2000: + Epoch Time: 35.48981785774231 +Epoch 62/2000: + Epoch Time: 24.729998111724854 +Epoch 63/2000: + Epoch Time: 35.74274730682373 +Epoch 64/2000: + Epoch Time: 24.687309503555298 +Epoch 65/2000: + Epoch Time: 24.731205701828003 +Epoch 66/2000: + Epoch Time: 35.519957542419434 +Epoch 67/2000: + Epoch Time: 24.799670696258545 +Epoch 68/2000: + Epoch Time: 35.611040115356445 +Epoch 69/2000: + Epoch Time: 24.789745807647705 +Epoch 70/2000: + Epoch Time: 35.916524171829224 +Epoch 71/2000: + Epoch Time: 24.713178157806396 +Epoch 72/2000: + Epoch Time: 24.674790859222412 +Epoch 73/2000: + Epoch Time: 35.39590072631836 +Epoch 74/2000: + Epoch Time: 24.63958501815796 +Epoch 75/2000: + Epoch Time: 35.40814781188965 +Epoch 76/2000: + Epoch Time: 24.70343828201294 +Epoch 77/2000: + Epoch Time: 35.60849452018738 +Epoch 78/2000: + Epoch Time: 24.94551658630371 +Epoch 79/2000: + Epoch Time: 24.921061992645264 +Epoch 80/2000: + Epoch Time: 35.59927845001221 +Epoch 81/2000: + Epoch Time: 24.844115495681763 +Epoch 82/2000: + Epoch Time: 35.837918758392334 +Epoch 83/2000: + Epoch Time: 24.6791889667511 +Epoch 84/2000: + Epoch Time: 35.40254044532776 +Epoch 85/2000: + Epoch Time: 24.679703950881958 +Epoch 86/2000: + Epoch Time: 24.6784405708313 +Epoch 87/2000: + Epoch Time: 35.42886972427368 +Epoch 88/2000: + Epoch Time: 24.698848247528076 +Epoch 89/2000: + Epoch Time: 35.528011322021484 +Epoch 90/2000: + Epoch Time: 24.770463466644287 +Epoch 91/2000: + Epoch Time: 35.57174372673035 +Epoch 92/2000: + Epoch Time: 24.75890851020813 +Epoch 93/2000: + Epoch Time: 24.71452760696411 +Epoch 94/2000: + Epoch Time: 35.43541693687439 +Epoch 95/2000: + Epoch Time: 24.74809765815735 +Epoch 96/2000: + Epoch Time: 35.27624273300171 +Epoch 97/2000: + Epoch Time: 24.74052333831787 +Epoch 98/2000: + Epoch Time: 35.41869807243347 +Epoch 99/2000: + Epoch Time: 24.750158548355103 +Epoch 100/2000: + Epoch Time: 35.32866835594177 +Epoch 101/2000: + Epoch Time: 24.647201776504517 +Epoch 102/2000: + Epoch Time: 24.691505908966064 +Epoch 103/2000: + Epoch Time: 35.47190022468567 +Epoch 104/2000: + Epoch Time: 24.729498863220215 +Epoch 105/2000: + Epoch Time: 35.548861503601074 +Epoch 106/2000: + Epoch Time: 24.789929628372192 +Epoch 107/2000: + Epoch Time: 36.040419578552246 +Epoch 108/2000: + Epoch Time: 24.988112211227417 +Epoch 109/2000: + Epoch Time: 24.851303339004517 +Epoch 110/2000: + Epoch Time: 36.27923655509949 +Epoch 111/2000: + Epoch Time: 24.902963161468506 +Epoch 112/2000: + Epoch Time: 36.07669281959534 +Epoch 113/2000: + Epoch Time: 24.950518369674683 +Epoch 114/2000: + Epoch Time: 35.67400360107422 +Epoch 115/2000: + Epoch Time: 24.751614570617676 +Epoch 116/2000: + Epoch Time: 24.758139610290527 +Epoch 117/2000: + Epoch Time: 36.429752588272095 +Epoch 118/2000: + Epoch Time: 24.730945348739624 +Epoch 119/2000: + Epoch Time: 35.60069131851196 +Epoch 120/2000: + Epoch Time: 24.83215641975403 +Epoch 121/2000: + Epoch Time: 35.45812678337097 +Epoch 122/2000: + Epoch Time: 24.82244110107422 +Epoch 123/2000: + Epoch Time: 24.663519859313965 +Epoch 124/2000: + Epoch Time: 35.473315477371216 +Epoch 125/2000: + Epoch Time: 24.687206506729126 +Epoch 126/2000: + Epoch Time: 35.37057089805603 +Epoch 127/2000: + Epoch Time: 24.71244478225708 +Epoch 128/2000: + Epoch Time: 35.50402784347534 +Epoch 129/2000: + Epoch Time: 24.649230241775513 +Epoch 130/2000: + Epoch Time: 24.71330165863037 +Epoch 131/2000: + Epoch Time: 35.726117849349976 +Epoch 132/2000: + Epoch Time: 24.693024158477783 +Epoch 133/2000: + Epoch Time: 35.512091398239136 +Epoch 134/2000: + Epoch Time: 24.91431999206543 +Epoch 135/2000: + Epoch Time: 35.709527254104614 +Epoch 136/2000: + Epoch Time: 24.792110204696655 +Epoch 137/2000: + Epoch Time: 24.750503540039062 +Epoch 138/2000: + Epoch Time: 35.59220337867737 +Epoch 139/2000: + Epoch Time: 24.64329719543457 +Epoch 140/2000: + Epoch Time: 35.73611521720886 +Epoch 141/2000: + Epoch Time: 24.82016372680664 +Epoch 142/2000: + Epoch Time: 35.60854983329773 +Epoch 143/2000: + Epoch Time: 24.808430433273315 +Epoch 144/2000: + Epoch Time: 24.76059126853943 +Epoch 145/2000: + Epoch Time: 35.79896640777588 +Epoch 146/2000: + Epoch Time: 24.892454147338867 +Epoch 147/2000: + Epoch Time: 35.41922926902771 +Epoch 148/2000: + Epoch Time: 24.871169567108154 +Epoch 149/2000: + Epoch Time: 35.487322092056274 +Epoch 150/2000: + Epoch Time: 24.83773422241211 +Epoch 151/2000: + Epoch Time: 24.7389657497406 +Epoch 152/2000: + Epoch Time: 35.7991840839386 +Epoch 153/2000: + Epoch Time: 24.768383264541626 +Epoch 154/2000: + Epoch Time: 35.4485547542572 +Epoch 155/2000: + Epoch Time: 24.8266179561615 +Epoch 156/2000: + Epoch Time: 35.41089653968811 +Epoch 157/2000: + Epoch Time: 24.67616057395935 +Epoch 158/2000: + Epoch Time: 24.69956684112549 +Epoch 159/2000: + Epoch Time: 35.37396764755249 +Epoch 160/2000: + Epoch Time: 24.577096700668335 +Epoch 161/2000: + Epoch Time: 35.51228976249695 +Epoch 162/2000: + Epoch Time: 24.811893463134766 +Epoch 163/2000: + Epoch Time: 35.38511371612549 +Epoch 164/2000: + Epoch Time: 24.704472064971924 +Epoch 165/2000: + Epoch Time: 24.855066299438477 +Epoch 166/2000: + Epoch Time: 35.3920795917511 +Epoch 167/2000: + Epoch Time: 24.887603521347046 +Epoch 168/2000: + Epoch Time: 35.463464975357056 +Epoch 169/2000: + Epoch Time: 24.68073320388794 +Epoch 170/2000: + Epoch Time: 35.50589203834534 +Epoch 171/2000: + Epoch Time: 24.735948085784912 +Epoch 172/2000: + Epoch Time: 24.777589321136475 +Epoch 173/2000: + Epoch Time: 35.38191199302673 +Epoch 174/2000: + Epoch Time: 24.65786600112915 +Epoch 175/2000: + Epoch Time: 35.45729064941406 +Epoch 176/2000: + Epoch Time: 24.795948028564453 +Epoch 177/2000: + Epoch Time: 35.41301965713501 +Epoch 178/2000: + Epoch Time: 24.870492935180664 +Epoch 179/2000: + Epoch Time: 24.79775619506836 +Epoch 180/2000: + Epoch Time: 35.67881226539612 +Epoch 181/2000: + Epoch Time: 24.916598081588745 +Epoch 182/2000: + Epoch Time: 35.477009296417236 +Epoch 183/2000: + Epoch Time: 24.874988317489624 +Epoch 184/2000: + Epoch Time: 35.296526193618774 +Epoch 185/2000: + Epoch Time: 24.73233652114868 +Epoch 186/2000: + Epoch Time: 24.748098134994507 +Epoch 187/2000: + Epoch Time: 35.42747235298157 +Epoch 188/2000: + Epoch Time: 24.82632613182068 +Epoch 189/2000: + Epoch Time: 35.819950580596924 +Epoch 190/2000: + Epoch Time: 24.698230266571045 +Epoch 191/2000: + Epoch Time: 35.40761208534241 +Epoch 192/2000: + Epoch Time: 24.689049243927002 +Epoch 193/2000: + Epoch Time: 24.760898113250732 +Epoch 194/2000: + Epoch Time: 35.407739877700806 +Epoch 195/2000: + Epoch Time: 24.74967050552368 +Epoch 196/2000: + Epoch Time: 35.406341552734375 +Epoch 197/2000: + Epoch Time: 24.710911512374878 +Epoch 198/2000: + Epoch Time: 35.54435920715332 +Epoch 199/2000: + Epoch Time: 24.829269409179688 +Epoch 200/2000: + Epoch Time: 35.4947464466095 +Epoch 201/2000: + Epoch Time: 24.687836408615112 +Epoch 202/2000: + Epoch Time: 24.793850660324097 +Epoch 203/2000: + Epoch Time: 35.51638126373291 +Epoch 204/2000: + Epoch Time: 24.758240699768066 +Epoch 205/2000: + Epoch Time: 35.43159890174866 +Epoch 206/2000: + Epoch Time: 24.585012435913086 +Epoch 207/2000: + Epoch Time: 35.83943319320679 +Epoch 208/2000: + Epoch Time: 24.872101545333862 +Epoch 209/2000: + Epoch Time: 24.884594917297363 +Epoch 210/2000: + Epoch Time: 35.47430491447449 +Epoch 211/2000: + Epoch Time: 24.750900983810425 +Epoch 212/2000: + Epoch Time: 35.400554895401 +Epoch 213/2000: + Epoch Time: 24.625861406326294 +Epoch 214/2000: + Epoch Time: 35.42755627632141 +Epoch 215/2000: + Epoch Time: 24.757211685180664 +Epoch 216/2000: + Epoch Time: 24.64772129058838 +Epoch 217/2000: + Epoch Time: 35.383679151535034 +Epoch 218/2000: + Epoch Time: 24.803553819656372 +Epoch 219/2000: + Epoch Time: 35.29698324203491 +Epoch 220/2000: + Epoch Time: 24.71952772140503 +Epoch 221/2000: + Epoch Time: 35.78842234611511 +Epoch 222/2000: + Epoch Time: 24.707910537719727 +Epoch 223/2000: + Epoch Time: 24.735080003738403 +Epoch 224/2000: + Epoch Time: 35.57108235359192 +Epoch 225/2000: + Epoch Time: 24.679489374160767 +Epoch 226/2000: + Epoch Time: 35.54309940338135 +Epoch 227/2000: + Epoch Time: 24.835370302200317 +Epoch 228/2000: + Epoch Time: 35.49916672706604 +Epoch 229/2000: + Epoch Time: 24.7119038105011 +Epoch 230/2000: + Epoch Time: 24.767407178878784 +Epoch 231/2000: + Epoch Time: 35.36032152175903 +Epoch 232/2000: + Epoch Time: 24.660505294799805 +Epoch 233/2000: + Epoch Time: 35.498955965042114 +Epoch 234/2000: + Epoch Time: 24.6289541721344 +Epoch 235/2000: + Epoch Time: 35.368149757385254 +Epoch 236/2000: + Epoch Time: 24.657021284103394 +Epoch 237/2000: + Epoch Time: 24.80756402015686 +Epoch 238/2000: + Epoch Time: 35.33079671859741 +Epoch 239/2000: + Epoch Time: 24.645576000213623 +Epoch 240/2000: + Epoch Time: 35.60582661628723 +Epoch 241/2000: + Epoch Time: 24.593059301376343 +Epoch 242/2000: + Epoch Time: 35.51935958862305 +Epoch 243/2000: + Epoch Time: 24.68890690803528 +Epoch 244/2000: + Epoch Time: 24.8764967918396 +Epoch 245/2000: + Epoch Time: 35.3274142742157 +Epoch 246/2000: + Epoch Time: 24.66098117828369 +Epoch 247/2000: + Epoch Time: 35.45595574378967 +Epoch 248/2000: + Epoch Time: 24.628766536712646 +Epoch 249/2000: + Epoch Time: 35.43592548370361 +Epoch 250/2000: + Epoch Time: 24.64699697494507 +Epoch 251/2000: + Epoch Time: 24.920464754104614 +Epoch 252/2000: + Epoch Time: 35.21763038635254 +Epoch 253/2000: + Epoch Time: 24.656155347824097 +Epoch 254/2000: + Epoch Time: 35.387609004974365 +Epoch 255/2000: + Epoch Time: 24.53616189956665 +Epoch 256/2000: + Epoch Time: 35.93192791938782 +Epoch 257/2000: + Epoch Time: 24.48743486404419 +Epoch 258/2000: + Epoch Time: 24.709110260009766 +Epoch 259/2000: + Epoch Time: 35.14708638191223 +Epoch 260/2000: + Epoch Time: 24.621503353118896 +Epoch 261/2000: + Epoch Time: 35.478848457336426 +Epoch 262/2000: + Epoch Time: 24.61019515991211 +Epoch 263/2000: + Epoch Time: 35.56016516685486 +Epoch 264/2000: + Epoch Time: 24.652814626693726 +Epoch 265/2000: + Epoch Time: 24.885714054107666 +Epoch 266/2000: + Epoch Time: 35.32704782485962 +Epoch 267/2000: + Epoch Time: 24.710923433303833 +Epoch 268/2000: + Epoch Time: 35.43694567680359 +Epoch 269/2000: + Epoch Time: 24.581632614135742 +Epoch 270/2000: + Epoch Time: 35.43827962875366 +Epoch 271/2000: + Epoch Time: 24.838512182235718 +Epoch 272/2000: + Epoch Time: 24.799503326416016 +Epoch 273/2000: + Epoch Time: 35.40597581863403 +Epoch 274/2000: + Epoch Time: 24.716816186904907 +Epoch 275/2000: + Epoch Time: 35.42040467262268 +Epoch 276/2000: + Epoch Time: 24.57945227622986 +Epoch 277/2000: + Epoch Time: 35.37623071670532 +Epoch 278/2000: + Epoch Time: 24.65112042427063 +Epoch 279/2000: + Epoch Time: 24.7007839679718 +Epoch 280/2000: + Epoch Time: 35.40199065208435 +Epoch 281/2000: + Epoch Time: 24.607622146606445 +Epoch 282/2000: + Epoch Time: 35.375375509262085 +Epoch 283/2000: + Epoch Time: 24.680469274520874 +Epoch 284/2000: + Epoch Time: 35.4027419090271 +Epoch 285/2000: + Epoch Time: 24.716225624084473 +Epoch 286/2000: + Epoch Time: 24.542632341384888 +Epoch 287/2000: + Epoch Time: 35.58690571784973 +Epoch 288/2000: + Epoch Time: 24.748183012008667 +Epoch 289/2000: + Epoch Time: 35.6042742729187 +Epoch 290/2000: + Epoch Time: 24.788575649261475 +Epoch 291/2000: + Epoch Time: 35.809568881988525 +Epoch 292/2000: + Epoch Time: 24.72543954849243 +Epoch 293/2000: + Epoch Time: 24.721887826919556 +Epoch 294/2000: + Epoch Time: 35.45573163032532 +Epoch 295/2000: + Epoch Time: 24.726792097091675 +Epoch 296/2000: + Epoch Time: 35.541536808013916 +Epoch 297/2000: + Epoch Time: 24.722572565078735 +Epoch 298/2000: + Epoch Time: 35.612637519836426 +Epoch 299/2000: + Epoch Time: 24.647411584854126 +Epoch 300/2000: + Epoch Time: 35.25795841217041 +Epoch 301/2000: + Epoch Time: 24.740530729293823 +Epoch 302/2000: + Epoch Time: 24.57604742050171 +Epoch 303/2000: + Epoch Time: 35.61211156845093 +Epoch 304/2000: + Epoch Time: 24.624297618865967 +Epoch 305/2000: + Epoch Time: 35.54382061958313 +Epoch 306/2000: + Epoch Time: 24.75022840499878 +Epoch 307/2000: + Epoch Time: 35.34375882148743 +Epoch 308/2000: + Epoch Time: 24.97662091255188 +Epoch 309/2000: + Epoch Time: 24.70427441596985 +Epoch 310/2000: + Epoch Time: 35.51878809928894 +Epoch 311/2000: + Epoch Time: 24.652238845825195 +Epoch 312/2000: + Epoch Time: 35.48239827156067 +Epoch 313/2000: + Epoch Time: 24.819164276123047 +Epoch 314/2000: + Epoch Time: 35.25243377685547 +Epoch 315/2000: + Epoch Time: 24.86341142654419 +Epoch 316/2000: + Epoch Time: 24.605451107025146 +Epoch 317/2000: + Epoch Time: 35.611154079437256 +Epoch 318/2000: + Epoch Time: 24.658222675323486 +Epoch 319/2000: + Epoch Time: 35.38360285758972 +Epoch 320/2000: + Epoch Time: 24.789674282073975 +Epoch 321/2000: + Epoch Time: 35.312254428863525 +Epoch 322/2000: + Epoch Time: 24.925702571868896 +Epoch 323/2000: + Epoch Time: 24.760510444641113 +Epoch 324/2000: + Epoch Time: 35.39197564125061 +Epoch 325/2000: + Epoch Time: 24.7003436088562 +Epoch 326/2000: + Epoch Time: 35.43185591697693 +Epoch 327/2000: + Epoch Time: 24.962104320526123 +Epoch 328/2000: + Epoch Time: 35.691429138183594 +Epoch 329/2000: + Epoch Time: 24.760871171951294 +Epoch 330/2000: + Epoch Time: 24.904479265213013 +Epoch 331/2000: + Epoch Time: 35.39664840698242 +Epoch 332/2000: + Epoch Time: 24.819896936416626 +Epoch 333/2000: + Epoch Time: 35.1812961101532 +Epoch 334/2000: + Epoch Time: 24.908280849456787 +Epoch 335/2000: + Epoch Time: 35.43588876724243 +Epoch 336/2000: + Epoch Time: 24.6421058177948 +Epoch 337/2000: + Epoch Time: 25.111541271209717 +Epoch 338/2000: + Epoch Time: 35.34027361869812 +Epoch 339/2000: + Epoch Time: 24.816900968551636 +Epoch 340/2000: + Epoch Time: 35.27311325073242 +Epoch 341/2000: + Epoch Time: 24.560415506362915 +Epoch 342/2000: + Epoch Time: 35.76804065704346 +Epoch 343/2000: + Epoch Time: 24.733977794647217 +Epoch 344/2000: + Epoch Time: 24.89945888519287 +Epoch 345/2000: + Epoch Time: 35.481725215911865 +Epoch 346/2000: + Epoch Time: 24.727851152420044 +Epoch 347/2000: + Epoch Time: 35.71692657470703 +Epoch 348/2000: + Epoch Time: 24.676548957824707 +Epoch 349/2000: + Epoch Time: 35.62117624282837 +Epoch 350/2000: + Epoch Time: 24.757655382156372 +Epoch 351/2000: + Epoch Time: 24.763057708740234 +Epoch 352/2000: + Epoch Time: 35.49139642715454 +Epoch 353/2000: + Epoch Time: 24.64014196395874 +Epoch 354/2000: + Epoch Time: 35.66678476333618 +Epoch 355/2000: + Epoch Time: 24.7399799823761 +Epoch 356/2000: + Epoch Time: 35.66110610961914 +Epoch 357/2000: + Epoch Time: 24.727049350738525 +Epoch 358/2000: + Epoch Time: 24.735601902008057 +Epoch 359/2000: + Epoch Time: 35.57261800765991 +Epoch 360/2000: + Epoch Time: 24.607879161834717 +Epoch 361/2000: + Epoch Time: 36.05156850814819 +Epoch 362/2000: + Epoch Time: 24.775346517562866 +Epoch 363/2000: + Epoch Time: 35.60239243507385 +Epoch 364/2000: + Epoch Time: 24.766722679138184 +Epoch 365/2000: + Epoch Time: 24.669799327850342 +Epoch 366/2000: + Epoch Time: 35.607998847961426 +Epoch 367/2000: + Epoch Time: 24.6030592918396 +Epoch 368/2000: + Epoch Time: 35.48562407493591 +Epoch 369/2000: + Epoch Time: 24.70834183692932 +Epoch 370/2000: + Epoch Time: 35.288158893585205 +Epoch 371/2000: + Epoch Time: 25.143275022506714 +Epoch 372/2000: + Epoch Time: 24.85885763168335 +Epoch 373/2000: + Epoch Time: 35.921481132507324 +Epoch 374/2000: + Epoch Time: 24.889686107635498 +Epoch 375/2000: + Epoch Time: 35.55241250991821 +Epoch 376/2000: + Epoch Time: 25.055066108703613 +Epoch 377/2000: + Epoch Time: 35.503241777420044 +Epoch 378/2000: + Epoch Time: 24.882424354553223 +Epoch 379/2000: + Epoch Time: 24.801430225372314 +Epoch 380/2000: + Epoch Time: 35.57321810722351 +Epoch 381/2000: + Epoch Time: 24.944674015045166 +Epoch 382/2000: + Epoch Time: 35.48458790779114 +Epoch 383/2000: + Epoch Time: 25.011420965194702 +Epoch 384/2000: + Epoch Time: 35.648417711257935 +Epoch 385/2000: + Epoch Time: 24.65640687942505 +Epoch 386/2000: + Epoch Time: 25.154268980026245 +Epoch 387/2000: + Epoch Time: 35.509374141693115 +Epoch 388/2000: + Epoch Time: 24.895082712173462 +Epoch 389/2000: + Epoch Time: 35.54339122772217 +Epoch 390/2000: + Epoch Time: 24.74587917327881 +Epoch 391/2000: + Epoch Time: 35.74361777305603 +Epoch 392/2000: + Epoch Time: 24.717318058013916 +Epoch 393/2000: + Epoch Time: 25.081966638565063 +Epoch 394/2000: + Epoch Time: 35.617270946502686 +Epoch 395/2000: + Epoch Time: 24.69038963317871 +Epoch 396/2000: + Epoch Time: 36.10256052017212 +Epoch 397/2000: + Epoch Time: 24.67675757408142 +Epoch 398/2000: + Epoch Time: 35.61740493774414 +Epoch 399/2000: + Epoch Time: 24.76110053062439 +Epoch 400/2000: + Epoch Time: 35.53637194633484 +Epoch 401/2000: + Epoch Time: 24.793002128601074 +Epoch 402/2000: + Epoch Time: 24.828177213668823 +Epoch 403/2000: + Epoch Time: 35.70713496208191 +Epoch 404/2000: + Epoch Time: 24.715309619903564 +Epoch 405/2000: + Epoch Time: 35.589635133743286 +Epoch 406/2000: + Epoch Time: 24.890823364257812 +Epoch 407/2000: + Epoch Time: 35.26721453666687 +Epoch 408/2000: + Epoch Time: 25.06601881980896 +Epoch 409/2000: + Epoch Time: 24.683364391326904 +Epoch 410/2000: + Epoch Time: 35.51494550704956 +Epoch 411/2000: + Epoch Time: 24.851269960403442 +Epoch 412/2000: + Epoch Time: 35.439995527267456 +Epoch 413/2000: + Epoch Time: 25.042290925979614 +Epoch 414/2000: + Epoch Time: 35.49414682388306 +Epoch 415/2000: + Epoch Time: 24.82618260383606 +Epoch 416/2000: + Epoch Time: 24.83786106109619 +Epoch 417/2000: + Epoch Time: 35.25646710395813 +Epoch 418/2000: + Epoch Time: 24.8118314743042 +Epoch 419/2000: + Epoch Time: 35.472758054733276 +Epoch 420/2000: + Epoch Time: 24.787790536880493 +Epoch 421/2000: + Epoch Time: 35.891263008117676 +Epoch 422/2000: + Epoch Time: 24.78785991668701 +Epoch 423/2000: + Epoch Time: 25.031198978424072 +Epoch 424/2000: + Epoch Time: 35.83002996444702 +Epoch 425/2000: + Epoch Time: 24.854291439056396 +Epoch 426/2000: + Epoch Time: 35.915852308273315 +Epoch 427/2000: + Epoch Time: 24.799059867858887 +Epoch 428/2000: + Epoch Time: 35.75024080276489 +Epoch 429/2000: + Epoch Time: 24.761486291885376 +Epoch 430/2000: + Epoch Time: 24.70317268371582 +Epoch 431/2000: + Epoch Time: 35.73544359207153 +Epoch 432/2000: + Epoch Time: 24.777086973190308 +Epoch 433/2000: + Epoch Time: 35.47631788253784 +Epoch 434/2000: + Epoch Time: 24.59303069114685 +Epoch 435/2000: + Epoch Time: 35.46470379829407 +Epoch 436/2000: + Epoch Time: 24.843823432922363 +Epoch 437/2000: + Epoch Time: 24.66543173789978 +Epoch 438/2000: + Epoch Time: 35.55543828010559 +Epoch 439/2000: + Epoch Time: 24.661420822143555 +Epoch 440/2000: + Epoch Time: 35.28066349029541 +Epoch 441/2000: + Epoch Time: 24.978854417800903 +Epoch 442/2000: + Epoch Time: 35.34843134880066 +Epoch 443/2000: + Epoch Time: 24.90140390396118 +Epoch 444/2000: + Epoch Time: 24.690845012664795 +Epoch 445/2000: + Epoch Time: 35.314844369888306 +Epoch 446/2000: + Epoch Time: 24.972548484802246 +Epoch 447/2000: + Epoch Time: 35.27424502372742 +Epoch 448/2000: + Epoch Time: 24.70748543739319 +Epoch 449/2000: + Epoch Time: 35.3870050907135 +Epoch 450/2000: + Epoch Time: 24.521143913269043 +Epoch 451/2000: + Epoch Time: 24.762455940246582 +Epoch 452/2000: + Epoch Time: 35.19742798805237 +Epoch 453/2000: + Epoch Time: 24.541099786758423 +Epoch 454/2000: + Epoch Time: 35.42881369590759 +Epoch 455/2000: + Epoch Time: 24.591652631759644 +Epoch 456/2000: + Epoch Time: 35.31607437133789 +Epoch 457/2000: + Epoch Time: 24.974140167236328 +Epoch 458/2000: + Epoch Time: 24.84969711303711 +Epoch 459/2000: + Epoch Time: 35.53487682342529 +Epoch 460/2000: + Epoch Time: 24.725805044174194 +Epoch 461/2000: + Epoch Time: 35.092963218688965 +Epoch 462/2000: + Epoch Time: 24.58248209953308 +Epoch 463/2000: + Epoch Time: 35.49625039100647 +Epoch 464/2000: + Epoch Time: 24.492684602737427 +Epoch 465/2000: + Epoch Time: 24.387238264083862 +Epoch 466/2000: + Epoch Time: 35.08454704284668 +Epoch 467/2000: + Epoch Time: 24.592073917388916 +Epoch 468/2000: + Epoch Time: 35.76911926269531 +Epoch 469/2000: + Epoch Time: 24.606127977371216 +Epoch 470/2000: + Epoch Time: 35.15366172790527 +Epoch 471/2000: + Epoch Time: 24.664879322052002 +Epoch 472/2000: + Epoch Time: 24.940324068069458 +Epoch 473/2000: + Epoch Time: 35.318893909454346 +Epoch 474/2000: + Epoch Time: 24.735705852508545 +Epoch 475/2000: + Epoch Time: 35.35667037963867 +Epoch 476/2000: + Epoch Time: 24.593982934951782 +Epoch 477/2000: + Epoch Time: 35.60894799232483 +Epoch 478/2000: + Epoch Time: 24.55493474006653 +Epoch 479/2000: + Epoch Time: 24.506678819656372 +Epoch 480/2000: + Epoch Time: 35.2259566783905 +Epoch 481/2000: + Epoch Time: 24.665215015411377 +Epoch 482/2000: + Epoch Time: 35.485427379608154 +Epoch 483/2000: + Epoch Time: 24.445679903030396 +Epoch 484/2000: + Epoch Time: 35.29953622817993 +Epoch 485/2000: + Epoch Time: 24.60005283355713 +Epoch 486/2000: + Epoch Time: 24.99485492706299 +Epoch 487/2000: + Epoch Time: 35.45109438896179 +Epoch 488/2000: + Epoch Time: 24.69075918197632 +Epoch 489/2000: + Epoch Time: 35.5575795173645 +Epoch 490/2000: + Epoch Time: 24.80443811416626 +Epoch 491/2000: + Epoch Time: 35.69969058036804 +Epoch 492/2000: + Epoch Time: 24.732896327972412 +Epoch 493/2000: + Epoch Time: 24.823620319366455 +Epoch 494/2000: + Epoch Time: 35.455705881118774 +Epoch 495/2000: + Epoch Time: 24.90682077407837 +Epoch 496/2000: + Epoch Time: 35.58514404296875 +Epoch 497/2000: + Epoch Time: 24.733065843582153 +Epoch 498/2000: + Epoch Time: 35.458988666534424 +Epoch 499/2000: + Epoch Time: 24.794144868850708 +Epoch 500/2000: + Epoch Time: 36.52598476409912 +Epoch 501/2000: + Epoch Time: 24.840922355651855 +Epoch 502/2000: + Epoch Time: 24.851486682891846 +Epoch 503/2000: + Epoch Time: 35.60463070869446 +Epoch 504/2000: + Epoch Time: 24.701130628585815 +Epoch 505/2000: + Epoch Time: 35.75438928604126 +Epoch 506/2000: + Epoch Time: 24.767761707305908 +Epoch 507/2000: + Epoch Time: 35.69983172416687 +Epoch 508/2000: + Epoch Time: 24.82686495780945 +Epoch 509/2000: + Epoch Time: 25.017298936843872 +Epoch 510/2000: + Epoch Time: 35.81563758850098 +Epoch 511/2000: + Epoch Time: 24.78251051902771 +Epoch 512/2000: + Epoch Time: 35.485893964767456 +Epoch 513/2000: + Epoch Time: 24.609275579452515 +Epoch 514/2000: + Epoch Time: 35.871461391448975 +Epoch 515/2000: + Epoch Time: 24.772621393203735 +Epoch 516/2000: + Epoch Time: 24.760212659835815 +Epoch 517/2000: + Epoch Time: 35.51691555976868 +Epoch 518/2000: + Epoch Time: 24.895752429962158 +Epoch 519/2000: + Epoch Time: 35.61133599281311 +Epoch 520/2000: + Epoch Time: 24.782344102859497 +Epoch 521/2000: + Epoch Time: 35.785858154296875 +Epoch 522/2000: + Epoch Time: 24.80788779258728 +Epoch 523/2000: + Epoch Time: 25.188802480697632 +Epoch 524/2000: + Epoch Time: 35.78766965866089 +Epoch 525/2000: + Epoch Time: 24.810383796691895 +Epoch 526/2000: + Epoch Time: 35.462456703186035 +Epoch 527/2000: + Epoch Time: 24.665446758270264 +Epoch 528/2000: + Epoch Time: 35.784992933273315 +Epoch 529/2000: + Epoch Time: 24.757474899291992 +Epoch 530/2000: + Epoch Time: 24.67068338394165 +Epoch 531/2000: + Epoch Time: 35.34320282936096 +Epoch 532/2000: + Epoch Time: 24.65746831893921 +Epoch 533/2000: + Epoch Time: 35.47398018836975 +Epoch 534/2000: + Epoch Time: 24.700682401657104 +Epoch 535/2000: + Epoch Time: 35.62753510475159 +Epoch 536/2000: + Epoch Time: 24.723628759384155 +Epoch 537/2000: + Epoch Time: 24.88706660270691 +Epoch 538/2000: + Epoch Time: 35.29521346092224 +Epoch 539/2000: + Epoch Time: 24.56119441986084 +Epoch 540/2000: + Epoch Time: 35.56378412246704 +Epoch 541/2000: + Epoch Time: 24.71393847465515 +Epoch 542/2000: + Epoch Time: 35.857784271240234 +Epoch 543/2000: + Epoch Time: 24.644835948944092 +Epoch 544/2000: + Epoch Time: 24.361175060272217 +Epoch 545/2000: + Epoch Time: 35.11275339126587 +Epoch 546/2000: + Epoch Time: 24.46768093109131 +Epoch 547/2000: + Epoch Time: 35.458163261413574 +Epoch 548/2000: + Epoch Time: 24.787991046905518 +Epoch 549/2000: + Epoch Time: 35.253307819366455 +Epoch 550/2000: + Epoch Time: 24.63965368270874 +Epoch 551/2000: + Epoch Time: 24.789646863937378 +Epoch 552/2000: + Epoch Time: 35.309592485427856 +Epoch 553/2000: + Epoch Time: 24.468358039855957 +Epoch 554/2000: + Epoch Time: 34.972137451171875 +Epoch 555/2000: + Epoch Time: 24.458927392959595 +Epoch 556/2000: + Epoch Time: 35.621734619140625 +Epoch 557/2000: + Epoch Time: 24.50204586982727 +Epoch 558/2000: + Epoch Time: 24.436816453933716 +Epoch 559/2000: + Epoch Time: 34.911396503448486 +Epoch 560/2000: + Epoch Time: 24.38318681716919 +Epoch 561/2000: + Epoch Time: 35.48648953437805 +Epoch 562/2000: + Epoch Time: 24.42857837677002 +Epoch 563/2000: + Epoch Time: 35.04995322227478 +Epoch 564/2000: + Epoch Time: 24.372086763381958 +Epoch 565/2000: + Epoch Time: 24.526941061019897 +Epoch 566/2000: + Epoch Time: 35.3913631439209 +Epoch 567/2000: + Epoch Time: 24.452752113342285 +Epoch 568/2000: + Epoch Time: 35.08732867240906 +Epoch 569/2000: + Epoch Time: 24.3904869556427 +Epoch 570/2000: + Epoch Time: 35.806745529174805 +Epoch 571/2000: + Epoch Time: 24.507832765579224 +Epoch 572/2000: + Epoch Time: 24.410664319992065 +Epoch 573/2000: + Epoch Time: 35.286213874816895 +Epoch 574/2000: + Epoch Time: 24.633753299713135 +Epoch 575/2000: + Epoch Time: 36.11358857154846 +Epoch 576/2000: + Epoch Time: 24.70311951637268 +Epoch 577/2000: + Epoch Time: 35.67981004714966 +Epoch 578/2000: + Epoch Time: 24.717503786087036 +Epoch 579/2000: + Epoch Time: 24.593671321868896 +Epoch 580/2000: + Epoch Time: 35.6905357837677 +Epoch 581/2000: + Epoch Time: 24.612765789031982 +Epoch 582/2000: + Epoch Time: 35.316040992736816 +Epoch 583/2000: + Epoch Time: 24.88086175918579 +Epoch 584/2000: + Epoch Time: 35.742958545684814 +Epoch 585/2000: + Epoch Time: 24.9368097782135 +Epoch 586/2000: + Epoch Time: 24.75765609741211 +Epoch 587/2000: + Epoch Time: 35.360793590545654 +Epoch 588/2000: + Epoch Time: 24.718006372451782 +Epoch 589/2000: + Epoch Time: 35.865975856781006 +Epoch 590/2000: + Epoch Time: 24.718536376953125 +Epoch 591/2000: + Epoch Time: 35.47748351097107 +Epoch 592/2000: + Epoch Time: 24.680373907089233 +Epoch 593/2000: + Epoch Time: 24.641918182373047 +Epoch 594/2000: + Epoch Time: 35.847084045410156 +Epoch 595/2000: + Epoch Time: 24.760679006576538 +Epoch 596/2000: + Epoch Time: 35.44133400917053 +Epoch 597/2000: + Epoch Time: 24.700759649276733 +Epoch 598/2000: + Epoch Time: 35.52310514450073 +Epoch 599/2000: + Epoch Time: 25.12718439102173 +Epoch 600/2000: + Epoch Time: 35.34096622467041 +Epoch 601/2000: + Epoch Time: 24.624343633651733 +Epoch 602/2000: + Epoch Time: 24.612544059753418 +Epoch 603/2000: + Epoch Time: 35.63031029701233 +Epoch 604/2000: + Epoch Time: 24.777785301208496 +Epoch 605/2000: + Epoch Time: 35.53341460227966 +Epoch 606/2000: + Epoch Time: 24.631837606430054 +Epoch 607/2000: + Epoch Time: 35.83721089363098 +Epoch 608/2000: + Epoch Time: 25.094597578048706 +Epoch 609/2000: + Epoch Time: 24.95081329345703 +Epoch 610/2000: + Epoch Time: 35.49212861061096 +Epoch 611/2000: + Epoch Time: 24.635724782943726 +Epoch 612/2000: + Epoch Time: 35.63283085823059 +Epoch 613/2000: + Epoch Time: 25.223944187164307 +Epoch 614/2000: + Epoch Time: 35.55263066291809 +Epoch 615/2000: + Epoch Time: 24.858761310577393 +Epoch 616/2000: + Epoch Time: 24.874869108200073 +Epoch 617/2000: + Epoch Time: 35.59550428390503 +Epoch 618/2000: + Epoch Time: 25.127692222595215 +Epoch 619/2000: + Epoch Time: 35.68266487121582 +Epoch 620/2000: + Epoch Time: 24.82787013053894 +Epoch 621/2000: + Epoch Time: 35.629087686538696 +Epoch 622/2000: + Epoch Time: 24.905879974365234 +Epoch 623/2000: + Epoch Time: 25.048203229904175 +Epoch 624/2000: + Epoch Time: 35.715986490249634 +Epoch 625/2000: + Epoch Time: 24.853459358215332 +Epoch 626/2000: + Epoch Time: 35.57739877700806 +Epoch 627/2000: + Epoch Time: 24.91654634475708 +Epoch 628/2000: + Epoch Time: 35.77380061149597 +Epoch 629/2000: + Epoch Time: 24.840771198272705 +Epoch 630/2000: + Epoch Time: 24.808862924575806 +Epoch 631/2000: + Epoch Time: 35.52000141143799 +Epoch 632/2000: + Epoch Time: 25.14582657814026 +Epoch 633/2000: + Epoch Time: 35.674521684646606 +Epoch 634/2000: + Epoch Time: 24.798097610473633 +Epoch 635/2000: + Epoch Time: 35.642122983932495 +Epoch 636/2000: + Epoch Time: 24.790926933288574 +Epoch 637/2000: + Epoch Time: 25.268686056137085 +Epoch 638/2000: + Epoch Time: 35.69742274284363 +Epoch 639/2000: + Epoch Time: 24.789693593978882 +Epoch 640/2000: + Epoch Time: 35.9126558303833 +Epoch 641/2000: + Epoch Time: 24.766467094421387 +Epoch 642/2000: + Epoch Time: 35.84596252441406 +Epoch 643/2000: + Epoch Time: 24.82379698753357 +Epoch 644/2000: + Epoch Time: 24.842711925506592 +Epoch 645/2000: + Epoch Time: 35.60185694694519 +Epoch 646/2000: + Epoch Time: 24.882304191589355 +Epoch 647/2000: + Epoch Time: 36.07833456993103 +Epoch 648/2000: + Epoch Time: 24.85770583152771 +Epoch 649/2000: + Epoch Time: 35.530391693115234 +Epoch 650/2000: + Epoch Time: 24.764568567276 +Epoch 651/2000: + Epoch Time: 24.97206401824951 +Epoch 652/2000: + Epoch Time: 35.65803074836731 +Epoch 653/2000: + Epoch Time: 24.814361572265625 +Epoch 654/2000: + Epoch Time: 35.819377183914185 +Epoch 655/2000: + Epoch Time: 24.727194786071777 +Epoch 656/2000: + Epoch Time: 35.89118456840515 +Epoch 657/2000: + Epoch Time: 24.519769430160522 +Epoch 658/2000: + Epoch Time: 24.556804895401 +Epoch 659/2000: + Epoch Time: 35.1366503238678 +Epoch 660/2000: + Epoch Time: 24.569023370742798 +Epoch 661/2000: + Epoch Time: 35.51615381240845 +Epoch 662/2000: + Epoch Time: 24.554351806640625 +Epoch 663/2000: + Epoch Time: 35.134175539016724 +Epoch 664/2000: + Epoch Time: 24.562782526016235 +Epoch 665/2000: + Epoch Time: 24.565707206726074 +Epoch 666/2000: + Epoch Time: 35.49912929534912 +Epoch 667/2000: + Epoch Time: 24.506396293640137 +Epoch 668/2000: + Epoch Time: 35.128907918930054 +Epoch 669/2000: + Epoch Time: 24.551456212997437 +Epoch 670/2000: + Epoch Time: 35.54478907585144 +Epoch 671/2000: + Epoch Time: 24.848836183547974 +Epoch 672/2000: + Epoch Time: 24.460704565048218 +Epoch 673/2000: + Epoch Time: 35.11698246002197 +Epoch 674/2000: + Epoch Time: 24.533812046051025 +Epoch 675/2000: + Epoch Time: 35.63132619857788 +Epoch 676/2000: + Epoch Time: 24.936238050460815 +Epoch 677/2000: + Epoch Time: 35.09427189826965 +Epoch 678/2000: + Epoch Time: 24.523298025131226 +Epoch 679/2000: + Epoch Time: 24.542121410369873 +Epoch 680/2000: + Epoch Time: 35.53889489173889 +Epoch 681/2000: + Epoch Time: 24.98349690437317 +Epoch 682/2000: + Epoch Time: 35.83080792427063 +Epoch 683/2000: + Epoch Time: 24.707637548446655 +Epoch 684/2000: + Epoch Time: 35.700334310531616 +Epoch 685/2000: + Epoch Time: 25.19843578338623 +Epoch 686/2000: + Epoch Time: 25.150148153305054 +Epoch 687/2000: + Epoch Time: 35.850635051727295 +Epoch 688/2000: + Epoch Time: 24.94895839691162 +Epoch 689/2000: + Epoch Time: 35.85735011100769 +Epoch 690/2000: + Epoch Time: 25.047593116760254 +Epoch 691/2000: + Epoch Time: 35.305208683013916 +Epoch 692/2000: + Epoch Time: 24.74390697479248 +Epoch 693/2000: + Epoch Time: 24.9801607131958 +Epoch 694/2000: + Epoch Time: 35.598204135894775 +Epoch 695/2000: + Epoch Time: 25.173674821853638 +Epoch 696/2000: + Epoch Time: 35.748825550079346 +Epoch 697/2000: + Epoch Time: 24.614014625549316 +Epoch 698/2000: + Epoch Time: 35.22437286376953 +Epoch 699/2000: + Epoch Time: 24.5786771774292 +Epoch 700/2000: + Epoch Time: 35.777122259140015 +Epoch 701/2000: + Epoch Time: 24.579392910003662 +Epoch 702/2000: + Epoch Time: 24.565337657928467 +Epoch 703/2000: + Epoch Time: 35.49433135986328 +Epoch 704/2000: + Epoch Time: 24.77927851676941 +Epoch 705/2000: + Epoch Time: 35.959842681884766 +Epoch 706/2000: + Epoch Time: 24.65656614303589 +Epoch 707/2000: + Epoch Time: 35.278263092041016 +Epoch 708/2000: + Epoch Time: 24.616677284240723 +Epoch 709/2000: + Epoch Time: 24.773345232009888 +Epoch 710/2000: + Epoch Time: 36.635618925094604 +Epoch 711/2000: + Epoch Time: 24.727861166000366 +Epoch 712/2000: + Epoch Time: 35.67578744888306 +Epoch 713/2000: + Epoch Time: 24.814049005508423 +Epoch 714/2000: + Epoch Time: 35.71564340591431 +Epoch 715/2000: + Epoch Time: 24.84794545173645 +Epoch 716/2000: + Epoch Time: 24.575797080993652 +Epoch 717/2000: + Epoch Time: 35.184231996536255 +Epoch 718/2000: + Epoch Time: 24.557003021240234 +Epoch 719/2000: + Epoch Time: 35.67638063430786 +Epoch 720/2000: + Epoch Time: 24.665276288986206 +Epoch 721/2000: + Epoch Time: 35.190492153167725 +Epoch 722/2000: + Epoch Time: 24.44040846824646 +Epoch 723/2000: + Epoch Time: 24.591953992843628 +Epoch 724/2000: + Epoch Time: 35.50182366371155 +Epoch 725/2000: + Epoch Time: 24.68670392036438 +Epoch 726/2000: + Epoch Time: 35.13644862174988 +Epoch 727/2000: + Epoch Time: 24.529576539993286 +Epoch 728/2000: + Epoch Time: 35.04981327056885 +Epoch 729/2000: + Epoch Time: 24.80768132209778 +Epoch 730/2000: + Epoch Time: 24.63678503036499 +Epoch 731/2000: + Epoch Time: 35.08624792098999 +Epoch 732/2000: + Epoch Time: 24.51441192626953 +Epoch 733/2000: + Epoch Time: 35.05316209793091 +Epoch 734/2000: + Epoch Time: 24.954473733901978 +Epoch 735/2000: + Epoch Time: 35.4983332157135 +Epoch 736/2000: + Epoch Time: 24.554625988006592 +Epoch 737/2000: + Epoch Time: 24.66372513771057 +Epoch 738/2000: + Epoch Time: 35.323174238204956 +Epoch 739/2000: + Epoch Time: 25.129843950271606 +Epoch 740/2000: + Epoch Time: 35.30461597442627 +Epoch 741/2000: + Epoch Time: 24.633382320404053 +Epoch 742/2000: + Epoch Time: 36.20535945892334 +Epoch 743/2000: + Epoch Time: 24.74489426612854 +Epoch 744/2000: + Epoch Time: 25.16648578643799 +Epoch 745/2000: + Epoch Time: 35.87027311325073 +Epoch 746/2000: + Epoch Time: 24.834551334381104 +Epoch 747/2000: + Epoch Time: 36.00997447967529 +Epoch 748/2000: + Epoch Time: 24.752525568008423 +Epoch 749/2000: + Epoch Time: 35.9644296169281 +Epoch 750/2000: + Epoch Time: 24.773386240005493 +Epoch 751/2000: + Epoch Time: 24.773052215576172 +Epoch 752/2000: + Epoch Time: 35.65056538581848 +Epoch 753/2000: + Epoch Time: 24.815751791000366 +Epoch 754/2000: + Epoch Time: 36.121243953704834 +Epoch 755/2000: + Epoch Time: 24.759575128555298 +Epoch 756/2000: + Epoch Time: 35.620230197906494 +Epoch 757/2000: + Epoch Time: 24.56084179878235 +Epoch 758/2000: + Epoch Time: 24.6566002368927 +Epoch 759/2000: + Epoch Time: 36.02796483039856 +Epoch 760/2000: + Epoch Time: 24.68325686454773 +Epoch 761/2000: + Epoch Time: 35.71173143386841 +Epoch 762/2000: + Epoch Time: 24.753487586975098 +Epoch 763/2000: + Epoch Time: 35.839906215667725 +Epoch 764/2000: + Epoch Time: 24.964141368865967 +Epoch 765/2000: + Epoch Time: 24.81750249862671 +Epoch 766/2000: + Epoch Time: 35.595800161361694 +Epoch 767/2000: + Epoch Time: 24.752119302749634 +Epoch 768/2000: + Epoch Time: 36.09894609451294 +Epoch 769/2000: + Epoch Time: 24.9261794090271 +Epoch 770/2000: + Epoch Time: 35.747573137283325 +Epoch 771/2000: + Epoch Time: 24.683140754699707 +Epoch 772/2000: + Epoch Time: 24.718347787857056 +Epoch 773/2000: + Epoch Time: 35.99611163139343 +Epoch 774/2000: + Epoch Time: 24.871607303619385 +Epoch 775/2000: + Epoch Time: 35.369802713394165 +Epoch 776/2000: + Epoch Time: 24.766292095184326 +Epoch 777/2000: + Epoch Time: 35.72593927383423 +Epoch 778/2000: + Epoch Time: 24.97850775718689 +Epoch 779/2000: + Epoch Time: 24.96256732940674 +Epoch 780/2000: + Epoch Time: 35.99115562438965 +Epoch 781/2000: + Epoch Time: 24.71668267250061 +Epoch 782/2000: + Epoch Time: 35.430262327194214 +Epoch 783/2000: + Epoch Time: 24.863460779190063 +Epoch 784/2000: + Epoch Time: 35.96497297286987 +Epoch 785/2000: + Epoch Time: 24.69073462486267 +Epoch 786/2000: + Epoch Time: 24.73662495613098 +Epoch 787/2000: + Epoch Time: 35.49346375465393 +Epoch 788/2000: + Epoch Time: 24.834708213806152 +Epoch 789/2000: + Epoch Time: 35.60154747962952 +Epoch 790/2000: + Epoch Time: 24.414510488510132 +Epoch 791/2000: + Epoch Time: 35.04229235649109 +Epoch 792/2000: + Epoch Time: 24.61216902732849 +Epoch 793/2000: + Epoch Time: 24.937723398208618 +Epoch 794/2000: + Epoch Time: 35.64241051673889 +Epoch 795/2000: + Epoch Time: 24.834004878997803 +Epoch 796/2000: + Epoch Time: 35.75010681152344 +Epoch 797/2000: + Epoch Time: 24.79180884361267 +Epoch 798/2000: + Epoch Time: 36.05145049095154 +Epoch 799/2000: + Epoch Time: 24.882025957107544 +Epoch 800/2000: + Epoch Time: 35.505465507507324 +Epoch 801/2000: + Epoch Time: 24.74787664413452 +Epoch 802/2000: + Epoch Time: 24.681270360946655 +Epoch 803/2000: + Epoch Time: 35.51274514198303 +Epoch 804/2000: + Epoch Time: 24.730183839797974 +Epoch 805/2000: + Epoch Time: 35.08865737915039 +Epoch 806/2000: + Epoch Time: 24.645707845687866 +Epoch 807/2000: + Epoch Time: 35.379111766815186 +Epoch 808/2000: + Epoch Time: 25.10950994491577 +Epoch 809/2000: + Epoch Time: 24.80751609802246 +Epoch 810/2000: + Epoch Time: 35.209754943847656 +Epoch 811/2000: + Epoch Time: 24.63505721092224 +Epoch 812/2000: + Epoch Time: 34.96081280708313 +Epoch 813/2000: + Epoch Time: 24.764777898788452 +Epoch 814/2000: + Epoch Time: 35.6218900680542 +Epoch 815/2000: + Epoch Time: 24.445476293563843 +Epoch 816/2000: + Epoch Time: 24.413204193115234 +Epoch 817/2000: + Epoch Time: 34.99360370635986 +Epoch 818/2000: + Epoch Time: 24.68920588493347 +Epoch 819/2000: + Epoch Time: 35.465389251708984 +Epoch 820/2000: + Epoch Time: 24.33303213119507 +Epoch 821/2000: + Epoch Time: 34.9736590385437 +Epoch 822/2000: + Epoch Time: 24.470816373825073 +Epoch 823/2000: + Epoch Time: 24.647722244262695 +Epoch 824/2000: + Epoch Time: 35.327332496643066 +Epoch 825/2000: + Epoch Time: 24.63749098777771 +Epoch 826/2000: + Epoch Time: 35.24474358558655 +Epoch 827/2000: + Epoch Time: 24.51912522315979 +Epoch 828/2000: + Epoch Time: 35.83181071281433 +Epoch 829/2000: + Epoch Time: 24.725886583328247 +Epoch 830/2000: + Epoch Time: 24.43269443511963 +Epoch 831/2000: + Epoch Time: 35.06688356399536 +Epoch 832/2000: + Epoch Time: 24.48891305923462 +Epoch 833/2000: + Epoch Time: 35.384838342666626 +Epoch 834/2000: + Epoch Time: 24.83186101913452 +Epoch 835/2000: + Epoch Time: 35.064186334609985 +Epoch 836/2000: + Epoch Time: 24.504111528396606 +Epoch 837/2000: + Epoch Time: 24.52997875213623 +Epoch 838/2000: + Epoch Time: 35.302149057388306 +Epoch 839/2000: + Epoch Time: 24.85983681678772 +Epoch 840/2000: + Epoch Time: 35.117472648620605 +Epoch 841/2000: + Epoch Time: 24.39510178565979 +Epoch 842/2000: + Epoch Time: 35.12847542762756 +Epoch 843/2000: + Epoch Time: 24.864789485931396 +Epoch 844/2000: + Epoch Time: 24.867239952087402 +Epoch 845/2000: + Epoch Time: 35.425376892089844 +Epoch 846/2000: + Epoch Time: 24.686853408813477 +Epoch 847/2000: + Epoch Time: 35.76018238067627 +Epoch 848/2000: + Epoch Time: 24.89643168449402 +Epoch 849/2000: + Epoch Time: 36.333890199661255 +Epoch 850/2000: + Epoch Time: 24.735498666763306 +Epoch 851/2000: + Epoch Time: 24.753809452056885 +Epoch 852/2000: + Epoch Time: 35.194857120513916 +Epoch 853/2000: + Epoch Time: 24.67534303665161 +Epoch 854/2000: + Epoch Time: 35.81394410133362 +Epoch 855/2000: + Epoch Time: 24.616539478302002 +Epoch 856/2000: + Epoch Time: 35.20895171165466 +Epoch 857/2000: + Epoch Time: 24.529320001602173 +Epoch 858/2000: + Epoch Time: 24.622787714004517 +Epoch 859/2000: + Epoch Time: 35.7426438331604 +Epoch 860/2000: + Epoch Time: 24.654913187026978 +Epoch 861/2000: + Epoch Time: 35.397684812545776 +Epoch 862/2000: + Epoch Time: 24.608763694763184 +Epoch 863/2000: + Epoch Time: 35.43281817436218 +Epoch 864/2000: + Epoch Time: 25.117119073867798 +Epoch 865/2000: + Epoch Time: 24.686723232269287 +Epoch 866/2000: + Epoch Time: 35.20896363258362 +Epoch 867/2000: + Epoch Time: 24.631921768188477 +Epoch 868/2000: + Epoch Time: 35.44046425819397 +Epoch 869/2000: + Epoch Time: 25.293251037597656 +Epoch 870/2000: + Epoch Time: 35.40147399902344 +Epoch 871/2000: + Epoch Time: 24.575560092926025 +Epoch 872/2000: + Epoch Time: 24.65553092956543 +Epoch 873/2000: + Epoch Time: 35.49040365219116 +Epoch 874/2000: + Epoch Time: 25.070361614227295 +Epoch 875/2000: + Epoch Time: 35.42009520530701 +Epoch 876/2000: + Epoch Time: 24.715011596679688 +Epoch 877/2000: + Epoch Time: 35.3504638671875 +Epoch 878/2000: + Epoch Time: 24.617006540298462 +Epoch 879/2000: + Epoch Time: 25.240853786468506 +Epoch 880/2000: + Epoch Time: 35.813806772232056 +Epoch 881/2000: + Epoch Time: 24.803488731384277 +Epoch 882/2000: + Epoch Time: 35.626328229904175 +Epoch 883/2000: + Epoch Time: 24.628316164016724 +Epoch 884/2000: + Epoch Time: 35.792935848236084 +Epoch 885/2000: + Epoch Time: 24.674306869506836 +Epoch 886/2000: + Epoch Time: 24.65493869781494 +Epoch 887/2000: + Epoch Time: 35.44542741775513 +Epoch 888/2000: + Epoch Time: 24.654022216796875 +Epoch 889/2000: + Epoch Time: 35.76513743400574 +Epoch 890/2000: + Epoch Time: 24.845394372940063 +Epoch 891/2000: + Epoch Time: 35.226789474487305 +Epoch 892/2000: + Epoch Time: 24.688207864761353 +Epoch 893/2000: + Epoch Time: 24.50952672958374 +Epoch 894/2000: + Epoch Time: 35.678940296173096 +Epoch 895/2000: + Epoch Time: 24.852975368499756 +Epoch 896/2000: + Epoch Time: 35.247748374938965 +Epoch 897/2000: + Epoch Time: 24.682751178741455 +Epoch 898/2000: + Epoch Time: 35.18966054916382 +Epoch 899/2000: + Epoch Time: 24.83729314804077 +Epoch 900/2000: + Epoch Time: 35.58314657211304 +Epoch 901/2000: + Epoch Time: 24.69731330871582 +Epoch 902/2000: + Epoch Time: 24.639997005462646 +Epoch 903/2000: + Epoch Time: 35.23214244842529 +Epoch 904/2000: + Epoch Time: 24.814213275909424 +Epoch 905/2000: + Epoch Time: 35.611464977264404 +Epoch 906/2000: + Epoch Time: 24.757558822631836 +Epoch 907/2000: + Epoch Time: 35.41431379318237 +Epoch 908/2000: + Epoch Time: 24.64137363433838 +Epoch 909/2000: + Epoch Time: 24.648574352264404 +Epoch 910/2000: + Epoch Time: 35.191550970077515 +Epoch 911/2000: + Epoch Time: 24.702023029327393 +Epoch 912/2000: + Epoch Time: 35.207162380218506 +Epoch 913/2000: + Epoch Time: 24.6565260887146 +Epoch 914/2000: + Epoch Time: 35.62559127807617 +Epoch 915/2000: + Epoch Time: 24.643350839614868 +Epoch 916/2000: + Epoch Time: 24.63106393814087 +Epoch 917/2000: + Epoch Time: 35.188841581344604 +Epoch 918/2000: + Epoch Time: 24.63407254219055 +Epoch 919/2000: + Epoch Time: 35.681947231292725 +Epoch 920/2000: + Epoch Time: 24.687127590179443 +Epoch 921/2000: + Epoch Time: 35.27165198326111 +Epoch 922/2000: + Epoch Time: 24.729403257369995 +Epoch 923/2000: + Epoch Time: 24.973546028137207 +Epoch 924/2000: + Epoch Time: 35.33767795562744 +Epoch 925/2000: + Epoch Time: 24.667596578598022 +Epoch 926/2000: + Epoch Time: 35.298922538757324 +Epoch 927/2000: + Epoch Time: 24.62073302268982 +Epoch 928/2000: + Epoch Time: 35.111939668655396 +Epoch 929/2000: + Epoch Time: 24.490769386291504 +Epoch 930/2000: + Epoch Time: 24.57753276824951 +Epoch 931/2000: + Epoch Time: 35.210370779037476 +Epoch 932/2000: + Epoch Time: 25.13656258583069 +Epoch 933/2000: + Epoch Time: 35.43427348136902 +Epoch 934/2000: + Epoch Time: 24.679256677627563 +Epoch 935/2000: + Epoch Time: 35.343979835510254 +Epoch 936/2000: + Epoch Time: 24.598076105117798 +Epoch 937/2000: + Epoch Time: 24.529602527618408 +Epoch 938/2000: + Epoch Time: 35.248393297195435 +Epoch 939/2000: + Epoch Time: 24.57921290397644 +Epoch 940/2000: + Epoch Time: 35.27330136299133 +Epoch 941/2000: + Epoch Time: 24.86209726333618 +Epoch 942/2000: + Epoch Time: 35.73678255081177 +Epoch 943/2000: + Epoch Time: 24.549925088882446 +Epoch 944/2000: + Epoch Time: 24.578086853027344 +Epoch 945/2000: + Epoch Time: 35.19970083236694 +Epoch 946/2000: + Epoch Time: 24.704981803894043 +Epoch 947/2000: + Epoch Time: 35.151440382003784 +Epoch 948/2000: + Epoch Time: 24.59059715270996 +Epoch 949/2000: + Epoch Time: 35.21794271469116 +Epoch 950/2000: + Epoch Time: 24.70150637626648 +Epoch 951/2000: + Epoch Time: 25.142497777938843 +Epoch 952/2000: + Epoch Time: 35.10420489311218 +Epoch 953/2000: + Epoch Time: 24.640453338623047 +Epoch 954/2000: + Epoch Time: 35.77076768875122 +Epoch 955/2000: + Epoch Time: 24.65352463722229 +Epoch 956/2000: + Epoch Time: 35.21466898918152 +Epoch 957/2000: + Epoch Time: 24.579013109207153 +Epoch 958/2000: + Epoch Time: 24.533400535583496 +Epoch 959/2000: + Epoch Time: 35.29015779495239 +Epoch 960/2000: + Epoch Time: 25.123446702957153 +Epoch 961/2000: + Epoch Time: 35.3635950088501 +Epoch 962/2000: + Epoch Time: 24.554966688156128 +Epoch 963/2000: + Epoch Time: 35.1619393825531 +Epoch 964/2000: + Epoch Time: 24.631818056106567 +Epoch 965/2000: + Epoch Time: 24.606558084487915 +Epoch 966/2000: + Epoch Time: 35.20708703994751 +Epoch 967/2000: + Epoch Time: 24.5430428981781 +Epoch 968/2000: + Epoch Time: 35.35405516624451 +Epoch 969/2000: + Epoch Time: 24.897963285446167 +Epoch 970/2000: + Epoch Time: 35.550204038619995 +Epoch 971/2000: + Epoch Time: 24.624273538589478 +Epoch 972/2000: + Epoch Time: 24.614041090011597 +Epoch 973/2000: + Epoch Time: 35.36507225036621 +Epoch 974/2000: + Epoch Time: 24.660075187683105 +Epoch 975/2000: + Epoch Time: 35.418795347213745 +Epoch 976/2000: + Epoch Time: 24.578716278076172 +Epoch 977/2000: + Epoch Time: 35.271687746047974 +Epoch 978/2000: + Epoch Time: 24.87783145904541 +Epoch 979/2000: + Epoch Time: 25.36007070541382 +Epoch 980/2000: + Epoch Time: 35.32493543624878 +Epoch 981/2000: + Epoch Time: 24.804359912872314 +Epoch 982/2000: + Epoch Time: 35.51408410072327 +Epoch 983/2000: + Epoch Time: 24.604065895080566 +Epoch 984/2000: + Epoch Time: 35.10903573036194 +Epoch 985/2000: + Epoch Time: 24.523969173431396 +Epoch 986/2000: + Epoch Time: 24.550079822540283 +Epoch 987/2000: + Epoch Time: 35.30996894836426 +Epoch 988/2000: + Epoch Time: 25.083455085754395 +Epoch 989/2000: + Epoch Time: 35.64966630935669 +Epoch 990/2000: + Epoch Time: 24.745256662368774 +Epoch 991/2000: + Epoch Time: 35.43269228935242 +Epoch 992/2000: + Epoch Time: 24.654396533966064 +Epoch 993/2000: + Epoch Time: 24.68360137939453 +Epoch 994/2000: + Epoch Time: 35.270506381988525 +Epoch 995/2000: + Epoch Time: 24.67215394973755 +Epoch 996/2000: + Epoch Time: 35.32207489013672 +Epoch 997/2000: + Epoch Time: 25.13521456718445 +Epoch 998/2000: + Epoch Time: 35.9251914024353 +Epoch 999/2000: + Epoch Time: 24.566006660461426 +Epoch 1000/2000: + Epoch Time: 35.141510248184204 +Epoch 1001/2000: + Epoch Time: 24.61998963356018 +Epoch 1002/2000: + Epoch Time: 24.68539595603943 +Epoch 1003/2000: + Epoch Time: 35.13578176498413 +Epoch 1004/2000: + Epoch Time: 24.56210422515869 +Epoch 1005/2000: + Epoch Time: 35.10244274139404 +Epoch 1006/2000: + Epoch Time: 24.826502323150635 +Epoch 1007/2000: + Epoch Time: 35.76523280143738 +Epoch 1008/2000: + Epoch Time: 24.596707820892334 +Epoch 1009/2000: + Epoch Time: 24.597206592559814 +Epoch 1010/2000: + Epoch Time: 35.10092735290527 +Epoch 1011/2000: + Epoch Time: 24.593504905700684 +Epoch 1012/2000: + Epoch Time: 35.246158599853516 +Epoch 1013/2000: + Epoch Time: 24.65518021583557 +Epoch 1014/2000: + Epoch Time: 35.08187389373779 +Epoch 1015/2000: + Epoch Time: 24.81447958946228 +Epoch 1016/2000: + Epoch Time: 24.943737745285034 +Epoch 1017/2000: + Epoch Time: 35.307225704193115 +Epoch 1018/2000: + Epoch Time: 24.46610450744629 +Epoch 1019/2000: + Epoch Time: 35.16604804992676 +Epoch 1020/2000: + Epoch Time: 24.539435386657715 +Epoch 1021/2000: + Epoch Time: 35.15023112297058 +Epoch 1022/2000: + Epoch Time: 24.65432047843933 +Epoch 1023/2000: + Epoch Time: 24.79450750350952 +Epoch 1024/2000: + Epoch Time: 35.46946334838867 +Epoch 1025/2000: + Epoch Time: 25.13525152206421 +Epoch 1026/2000: + Epoch Time: 35.9942843914032 +Epoch 1027/2000: + Epoch Time: 24.657283544540405 +Epoch 1028/2000: + Epoch Time: 35.27829194068909 +Epoch 1029/2000: + Epoch Time: 24.582583904266357 +Epoch 1030/2000: + Epoch Time: 24.592228412628174 +Epoch 1031/2000: + Epoch Time: 35.32339596748352 +Epoch 1032/2000: + Epoch Time: 24.660651922225952 +Epoch 1033/2000: + Epoch Time: 35.22021412849426 +Epoch 1034/2000: + Epoch Time: 24.725236415863037 +Epoch 1035/2000: + Epoch Time: 35.693230867385864 +Epoch 1036/2000: + Epoch Time: 24.573429107666016 +Epoch 1037/2000: + Epoch Time: 24.631763696670532 +Epoch 1038/2000: + Epoch Time: 35.18550634384155 +Epoch 1039/2000: + Epoch Time: 24.607176065444946 +Epoch 1040/2000: + Epoch Time: 35.16862678527832 +Epoch 1041/2000: + Epoch Time: 24.66126585006714 +Epoch 1042/2000: + Epoch Time: 35.143826723098755 +Epoch 1043/2000: + Epoch Time: 24.62250256538391 +Epoch 1044/2000: + Epoch Time: 25.055958032608032 +Epoch 1045/2000: + Epoch Time: 35.226168632507324 +Epoch 1046/2000: + Epoch Time: 24.495994329452515 +Epoch 1047/2000: + Epoch Time: 35.328025579452515 +Epoch 1048/2000: + Epoch Time: 24.48091220855713 +Epoch 1049/2000: + Epoch Time: 35.12357306480408 +Epoch 1050/2000: + Epoch Time: 24.67227602005005 +Epoch 1051/2000: + Epoch Time: 24.643283367156982 +Epoch 1052/2000: + Epoch Time: 35.230454444885254 +Epoch 1053/2000: + Epoch Time: 25.063071727752686 +Epoch 1054/2000: + Epoch Time: 35.5429048538208 +Epoch 1055/2000: + Epoch Time: 24.60768461227417 +Epoch 1056/2000: + Epoch Time: 35.1495201587677 +Epoch 1057/2000: + Epoch Time: 24.578309535980225 +Epoch 1058/2000: + Epoch Time: 24.574987173080444 +Epoch 1059/2000: + Epoch Time: 35.59954285621643 +Epoch 1060/2000: + Epoch Time: 24.615119457244873 +Epoch 1061/2000: + Epoch Time: 35.21521711349487 +Epoch 1062/2000: + Epoch Time: 24.667022705078125 +Epoch 1063/2000: + Epoch Time: 35.932945728302 +Epoch 1064/2000: + Epoch Time: 24.573793411254883 +Epoch 1065/2000: + Epoch Time: 24.628169536590576 +Epoch 1066/2000: + Epoch Time: 35.29425501823425 +Epoch 1067/2000: + Epoch Time: 24.64575457572937 +Epoch 1068/2000: + Epoch Time: 35.24331879615784 +Epoch 1069/2000: + Epoch Time: 24.71995520591736 +Epoch 1070/2000: + Epoch Time: 35.20645594596863 +Epoch 1071/2000: + Epoch Time: 24.60336971282959 +Epoch 1072/2000: + Epoch Time: 24.944637060165405 +Epoch 1073/2000: + Epoch Time: 35.54379725456238 +Epoch 1074/2000: + Epoch Time: 24.610798835754395 +Epoch 1075/2000: + Epoch Time: 35.2250337600708 +Epoch 1076/2000: + Epoch Time: 24.71922516822815 +Epoch 1077/2000: + Epoch Time: 35.59221148490906 +Epoch 1078/2000: + Epoch Time: 24.578552961349487 +Epoch 1079/2000: + Epoch Time: 24.631524562835693 +Epoch 1080/2000: + Epoch Time: 35.460610151290894 +Epoch 1081/2000: + Epoch Time: 25.02049946784973 +Epoch 1082/2000: + Epoch Time: 36.1418399810791 +Epoch 1083/2000: + Epoch Time: 24.712499618530273 +Epoch 1084/2000: + Epoch Time: 35.3444128036499 +Epoch 1085/2000: + Epoch Time: 24.742756366729736 +Epoch 1086/2000: + Epoch Time: 24.725934505462646 +Epoch 1087/2000: + Epoch Time: 35.291370153427124 +Epoch 1088/2000: + Epoch Time: 24.732097864151 +Epoch 1089/2000: + Epoch Time: 35.29313063621521 +Epoch 1090/2000: + Epoch Time: 24.666027545928955 +Epoch 1091/2000: + Epoch Time: 35.83811807632446 +Epoch 1092/2000: + Epoch Time: 24.587167739868164 +Epoch 1093/2000: + Epoch Time: 24.53985047340393 +Epoch 1094/2000: + Epoch Time: 35.45243716239929 +Epoch 1095/2000: + Epoch Time: 24.53002142906189 +Epoch 1096/2000: + Epoch Time: 35.6087110042572 +Epoch 1097/2000: + Epoch Time: 24.75340485572815 +Epoch 1098/2000: + Epoch Time: 35.54735445976257 +Epoch 1099/2000: + Epoch Time: 24.897252559661865 +Epoch 1100/2000: + Epoch Time: 35.852535247802734 +Epoch 1101/2000: + Epoch Time: 24.713386058807373 +Epoch 1102/2000: + Epoch Time: 24.624632358551025 +Epoch 1103/2000: + Epoch Time: 35.15404438972473 +Epoch 1104/2000: + Epoch Time: 24.488099813461304 +Epoch 1105/2000: + Epoch Time: 35.13998293876648 +Epoch 1106/2000: + Epoch Time: 24.574887990951538 +Epoch 1107/2000: + Epoch Time: 35.19258236885071 +Epoch 1108/2000: + Epoch Time: 24.52881669998169 +Epoch 1109/2000: + Epoch Time: 24.602832078933716 +Epoch 1110/2000: + Epoch Time: 35.91993451118469 +Epoch 1111/2000: + Epoch Time: 24.650272846221924 +Epoch 1112/2000: + Epoch Time: 35.148953437805176 +Epoch 1113/2000: + Epoch Time: 24.71143388748169 +Epoch 1114/2000: + Epoch Time: 35.219619035720825 +Epoch 1115/2000: + Epoch Time: 24.74233078956604 +Epoch 1116/2000: + Epoch Time: 24.684925317764282 +Epoch 1117/2000: + Epoch Time: 35.307474851608276 +Epoch 1118/2000: + Epoch Time: 24.664796113967896 +Epoch 1119/2000: + Epoch Time: 35.943483114242554 +Epoch 1120/2000: + Epoch Time: 24.82831883430481 +Epoch 1121/2000: + Epoch Time: 35.417234897613525 +Epoch 1122/2000: + Epoch Time: 24.72252082824707 +Epoch 1123/2000: + Epoch Time: 24.606722354888916 +Epoch 1124/2000: + Epoch Time: 35.17753720283508 +Epoch 1125/2000: + Epoch Time: 24.484017372131348 +Epoch 1126/2000: + Epoch Time: 35.20297574996948 +Epoch 1127/2000: + Epoch Time: 24.53583812713623 +Epoch 1128/2000: + Epoch Time: 35.64986753463745 +Epoch 1129/2000: + Epoch Time: 25.12351155281067 +Epoch 1130/2000: + Epoch Time: 24.56554651260376 +Epoch 1131/2000: + Epoch Time: 35.170058250427246 +Epoch 1132/2000: + Epoch Time: 24.591293573379517 +Epoch 1133/2000: + Epoch Time: 35.24847173690796 +Epoch 1134/2000: + Epoch Time: 24.671013355255127 +Epoch 1135/2000: + Epoch Time: 35.208720445632935 +Epoch 1136/2000: + Epoch Time: 24.54998803138733 +Epoch 1137/2000: + Epoch Time: 24.598177909851074 +Epoch 1138/2000: + Epoch Time: 35.580779790878296 +Epoch 1139/2000: + Epoch Time: 24.981144905090332 +Epoch 1140/2000: + Epoch Time: 35.33831524848938 +Epoch 1141/2000: + Epoch Time: 24.702749013900757 +Epoch 1142/2000: + Epoch Time: 35.33922624588013 +Epoch 1143/2000: + Epoch Time: 24.68221116065979 +Epoch 1144/2000: + Epoch Time: 24.542970657348633 +Epoch 1145/2000: + Epoch Time: 35.37328600883484 +Epoch 1146/2000: + Epoch Time: 24.651411771774292 +Epoch 1147/2000: + Epoch Time: 35.43069243431091 +Epoch 1148/2000: + Epoch Time: 25.247912168502808 +Epoch 1149/2000: + Epoch Time: 35.55389857292175 +Epoch 1150/2000: + Epoch Time: 24.84859824180603 +Epoch 1151/2000: + Epoch Time: 24.83900237083435 +Epoch 1152/2000: + Epoch Time: 35.53420376777649 +Epoch 1153/2000: + Epoch Time: 24.857486963272095 +Epoch 1154/2000: + Epoch Time: 36.0183892250061 +Epoch 1155/2000: + Epoch Time: 24.87337613105774 +Epoch 1156/2000: + Epoch Time: 35.37244963645935 +Epoch 1157/2000: + Epoch Time: 24.981590509414673 +Epoch 1158/2000: + Epoch Time: 24.82168960571289 +Epoch 1159/2000: + Epoch Time: 35.18979787826538 +Epoch 1160/2000: + Epoch Time: 24.66311812400818 +Epoch 1161/2000: + Epoch Time: 35.429747104644775 +Epoch 1162/2000: + Epoch Time: 24.68257713317871 +Epoch 1163/2000: + Epoch Time: 35.38212275505066 +Epoch 1164/2000: + Epoch Time: 24.685524702072144 +Epoch 1165/2000: + Epoch Time: 24.618801593780518 +Epoch 1166/2000: + Epoch Time: 36.02665090560913 +Epoch 1167/2000: + Epoch Time: 25.045669555664062 +Epoch 1168/2000: + Epoch Time: 35.41057991981506 +Epoch 1169/2000: + Epoch Time: 24.62299084663391 +Epoch 1170/2000: + Epoch Time: 35.446122884750366 +Epoch 1171/2000: + Epoch Time: 24.750770092010498 +Epoch 1172/2000: + Epoch Time: 24.543917655944824 +Epoch 1173/2000: + Epoch Time: 35.259856939315796 +Epoch 1174/2000: + Epoch Time: 24.492631673812866 +Epoch 1175/2000: + Epoch Time: 35.243223905563354 +Epoch 1176/2000: + Epoch Time: 25.060034036636353 +Epoch 1177/2000: + Epoch Time: 35.46270275115967 +Epoch 1178/2000: + Epoch Time: 24.533485889434814 +Epoch 1179/2000: + Epoch Time: 24.736573219299316 +Epoch 1180/2000: + Epoch Time: 35.32738924026489 +Epoch 1181/2000: + Epoch Time: 24.66947078704834 +Epoch 1182/2000: + Epoch Time: 35.454832792282104 +Epoch 1183/2000: + Epoch Time: 24.729646682739258 +Epoch 1184/2000: + Epoch Time: 35.404526472091675 +Epoch 1185/2000: + Epoch Time: 25.037134885787964 +Epoch 1186/2000: + Epoch Time: 25.195598125457764 +Epoch 1187/2000: + Epoch Time: 35.49934458732605 +Epoch 1188/2000: + Epoch Time: 24.66547656059265 +Epoch 1189/2000: + Epoch Time: 35.368446588516235 +Epoch 1190/2000: + Epoch Time: 24.67593240737915 +Epoch 1191/2000: + Epoch Time: 35.32570242881775 +Epoch 1192/2000: + Epoch Time: 24.621790647506714 +Epoch 1193/2000: + Epoch Time: 24.61722469329834 +Epoch 1194/2000: + Epoch Time: 35.135847091674805 +Epoch 1195/2000: + Epoch Time: 25.233120918273926 +Epoch 1196/2000: + Epoch Time: 35.78818130493164 +Epoch 1197/2000: + Epoch Time: 24.763116121292114 +Epoch 1198/2000: + Epoch Time: 36.095783948898315 +Epoch 1199/2000: + Epoch Time: 24.800158500671387 +Epoch 1200/2000: + Epoch Time: 35.49356174468994 +Epoch 1201/2000: + Epoch Time: 24.722574949264526 +Epoch 1202/2000: + Epoch Time: 24.600237607955933 +Epoch 1203/2000: + Epoch Time: 35.21905708312988 +Epoch 1204/2000: + Epoch Time: 24.864141702651978 +Epoch 1205/2000: + Epoch Time: 35.822911500930786 +Epoch 1206/2000: + Epoch Time: 24.650179147720337 +Epoch 1207/2000: + Epoch Time: 35.1731173992157 +Epoch 1208/2000: + Epoch Time: 24.494881868362427 +Epoch 1209/2000: + Epoch Time: 24.560907125473022 +Epoch 1210/2000: + Epoch Time: 35.1936149597168 +Epoch 1211/2000: + Epoch Time: 24.562897443771362 +Epoch 1212/2000: + Epoch Time: 35.258742809295654 +Epoch 1213/2000: + Epoch Time: 24.45874309539795 +Epoch 1214/2000: + Epoch Time: 35.960246324539185 +Epoch 1215/2000: + Epoch Time: 24.72198486328125 +Epoch 1216/2000: + Epoch Time: 24.520838022232056 +Epoch 1217/2000: + Epoch Time: 35.22949242591858 +Epoch 1218/2000: + Epoch Time: 24.569157361984253 +Epoch 1219/2000: + Epoch Time: 35.28745889663696 +Epoch 1220/2000: + Epoch Time: 24.684401273727417 +Epoch 1221/2000: + Epoch Time: 35.40891242027283 +Epoch 1222/2000: + Epoch Time: 24.620711088180542 +Epoch 1223/2000: + Epoch Time: 24.862487077713013 +Epoch 1224/2000: + Epoch Time: 36.19495177268982 +Epoch 1225/2000: + Epoch Time: 24.73548173904419 +Epoch 1226/2000: + Epoch Time: 35.25074529647827 +Epoch 1227/2000: + Epoch Time: 24.67181372642517 +Epoch 1228/2000: + Epoch Time: 35.31676697731018 +Epoch 1229/2000: + Epoch Time: 24.573405742645264 +Epoch 1230/2000: + Epoch Time: 24.586230278015137 +Epoch 1231/2000: + Epoch Time: 35.197922229766846 +Epoch 1232/2000: + Epoch Time: 24.559098720550537 +Epoch 1233/2000: + Epoch Time: 36.05987906455994 +Epoch 1234/2000: + Epoch Time: 24.999212741851807 +Epoch 1235/2000: + Epoch Time: 35.286844968795776 +Epoch 1236/2000: + Epoch Time: 24.563433170318604 +Epoch 1237/2000: + Epoch Time: 24.631200790405273 +Epoch 1238/2000: + Epoch Time: 35.260443687438965 +Epoch 1239/2000: + Epoch Time: 24.588801622390747 +Epoch 1240/2000: + Epoch Time: 35.42587876319885 +Epoch 1241/2000: + Epoch Time: 24.600443601608276 +Epoch 1242/2000: + Epoch Time: 35.47724270820618 +Epoch 1243/2000: + Epoch Time: 25.18826699256897 +Epoch 1244/2000: + Epoch Time: 24.68258261680603 +Epoch 1245/2000: + Epoch Time: 35.19885563850403 +Epoch 1246/2000: + Epoch Time: 24.58662724494934 +Epoch 1247/2000: + Epoch Time: 35.133577823638916 +Epoch 1248/2000: + Epoch Time: 24.598012924194336 +Epoch 1249/2000: + Epoch Time: 35.22473978996277 +Epoch 1250/2000: + Epoch Time: 24.626457929611206 +Epoch 1251/2000: + Epoch Time: 24.624788999557495 +Epoch 1252/2000: + Epoch Time: 35.73431324958801 +Epoch 1253/2000: + Epoch Time: 25.03900408744812 +Epoch 1254/2000: + Epoch Time: 35.348602533340454 +Epoch 1255/2000: + Epoch Time: 24.661461114883423 +Epoch 1256/2000: + Epoch Time: 35.216564893722534 +Epoch 1257/2000: + Epoch Time: 24.656601190567017 +Epoch 1258/2000: + Epoch Time: 24.556018352508545 +Epoch 1259/2000: + Epoch Time: 35.18522262573242 +Epoch 1260/2000: + Epoch Time: 24.704453945159912 +Epoch 1261/2000: + Epoch Time: 35.292948722839355 +Epoch 1262/2000: + Epoch Time: 25.292325019836426 +Epoch 1263/2000: + Epoch Time: 35.5561888217926 +Epoch 1264/2000: + Epoch Time: 24.674687385559082 +Epoch 1265/2000: + Epoch Time: 24.69982409477234 +Epoch 1266/2000: + Epoch Time: 35.22704553604126 +Epoch 1267/2000: + Epoch Time: 24.556084871292114 +Epoch 1268/2000: + Epoch Time: 35.61401295661926 +Epoch 1269/2000: + Epoch Time: 24.595876932144165 +Epoch 1270/2000: + Epoch Time: 35.26499581336975 +Epoch 1271/2000: + Epoch Time: 24.90661644935608 +Epoch 1272/2000: + Epoch Time: 25.233168601989746 +Epoch 1273/2000: + Epoch Time: 35.2452392578125 +Epoch 1274/2000: + Epoch Time: 24.675830602645874 +Epoch 1275/2000: + Epoch Time: 35.27216577529907 +Epoch 1276/2000: + Epoch Time: 24.695391416549683 +Epoch 1277/2000: + Epoch Time: 35.26898670196533 +Epoch 1278/2000: + Epoch Time: 24.744463682174683 +Epoch 1279/2000: + Epoch Time: 24.637982845306396 +Epoch 1280/2000: + Epoch Time: 35.28209066390991 +Epoch 1281/2000: + Epoch Time: 24.893743991851807 +Epoch 1282/2000: + Epoch Time: 35.78323698043823 +Epoch 1283/2000: + Epoch Time: 24.591869592666626 +Epoch 1284/2000: + Epoch Time: 35.25988435745239 +Epoch 1285/2000: + Epoch Time: 24.564228534698486 +Epoch 1286/2000: + Epoch Time: 24.501976490020752 +Epoch 1287/2000: + Epoch Time: 35.25056481361389 +Epoch 1288/2000: + Epoch Time: 24.62504816055298 +Epoch 1289/2000: + Epoch Time: 35.086355447769165 +Epoch 1290/2000: + Epoch Time: 24.596155643463135 +Epoch 1291/2000: + Epoch Time: 35.82458972930908 +Epoch 1292/2000: + Epoch Time: 24.800553798675537 +Epoch 1293/2000: + Epoch Time: 24.49317979812622 +Epoch 1294/2000: + Epoch Time: 35.511472940444946 +Epoch 1295/2000: + Epoch Time: 24.555875539779663 +Epoch 1296/2000: + Epoch Time: 35.2275550365448 +Epoch 1297/2000: + Epoch Time: 24.83547353744507 +Epoch 1298/2000: + Epoch Time: 35.215200901031494 +Epoch 1299/2000: + Epoch Time: 24.563356399536133 +Epoch 1300/2000: + Epoch Time: 35.541436195373535 +Epoch 1301/2000: + Epoch Time: 25.081847429275513 +Epoch 1302/2000: + Epoch Time: 24.595032215118408 +Epoch 1303/2000: + Epoch Time: 35.17485308647156 +Epoch 1304/2000: + Epoch Time: 24.62519931793213 +Epoch 1305/2000: + Epoch Time: 35.561317443847656 +Epoch 1306/2000: + Epoch Time: 24.519876956939697 +Epoch 1307/2000: + Epoch Time: 35.68660926818848 +Epoch 1308/2000: + Epoch Time: 24.750629663467407 +Epoch 1309/2000: + Epoch Time: 24.650731325149536 +Epoch 1310/2000: + Epoch Time: 36.11406350135803 +Epoch 1311/2000: + Epoch Time: 25.005776405334473 +Epoch 1312/2000: + Epoch Time: 35.41878819465637 +Epoch 1313/2000: + Epoch Time: 24.667189836502075 +Epoch 1314/2000: + Epoch Time: 35.40163230895996 +Epoch 1315/2000: + Epoch Time: 24.627828121185303 +Epoch 1316/2000: + Epoch Time: 24.542490005493164 +Epoch 1317/2000: + Epoch Time: 35.259915828704834 +Epoch 1318/2000: + Epoch Time: 24.63283133506775 +Epoch 1319/2000: + Epoch Time: 35.40286660194397 +Epoch 1320/2000: + Epoch Time: 25.299731731414795 +Epoch 1321/2000: + Epoch Time: 35.459105491638184 +Epoch 1322/2000: + Epoch Time: 24.648969411849976 +Epoch 1323/2000: + Epoch Time: 24.682645082473755 +Epoch 1324/2000: + Epoch Time: 35.138020038604736 +Epoch 1325/2000: + Epoch Time: 24.553359031677246 +Epoch 1326/2000: + Epoch Time: 35.27736306190491 +Epoch 1327/2000: + Epoch Time: 24.488369703292847 +Epoch 1328/2000: + Epoch Time: 35.13182306289673 +Epoch 1329/2000: + Epoch Time: 24.821451663970947 +Epoch 1330/2000: + Epoch Time: 25.084153175354004 +Epoch 1331/2000: + Epoch Time: 35.23859357833862 +Epoch 1332/2000: + Epoch Time: 24.69322180747986 +Epoch 1333/2000: + Epoch Time: 35.20233106613159 +Epoch 1334/2000: + Epoch Time: 24.621444702148438 +Epoch 1335/2000: + Epoch Time: 35.146748304367065 +Epoch 1336/2000: + Epoch Time: 24.555676221847534 +Epoch 1337/2000: + Epoch Time: 24.613502025604248 +Epoch 1338/2000: + Epoch Time: 35.57068085670471 +Epoch 1339/2000: + Epoch Time: 24.871975660324097 +Epoch 1340/2000: + Epoch Time: 35.7898108959198 +Epoch 1341/2000: + Epoch Time: 24.538533449172974 +Epoch 1342/2000: + Epoch Time: 35.24487113952637 +Epoch 1343/2000: + Epoch Time: 24.535403966903687 +Epoch 1344/2000: + Epoch Time: 24.603888511657715 +Epoch 1345/2000: + Epoch Time: 35.09186244010925 +Epoch 1346/2000: + Epoch Time: 24.61597204208374 +Epoch 1347/2000: + Epoch Time: 35.148876667022705 +Epoch 1348/2000: + Epoch Time: 24.67218279838562 +Epoch 1349/2000: + Epoch Time: 35.823498249053955 +Epoch 1350/2000: + Epoch Time: 24.763056755065918 +Epoch 1351/2000: + Epoch Time: 24.479872226715088 +Epoch 1352/2000: + Epoch Time: 35.184958934783936 +Epoch 1353/2000: + Epoch Time: 24.652027368545532 +Epoch 1354/2000: + Epoch Time: 35.15409064292908 +Epoch 1355/2000: + Epoch Time: 24.573142051696777 +Epoch 1356/2000: + Epoch Time: 35.34957480430603 +Epoch 1357/2000: + Epoch Time: 24.416303157806396 +Epoch 1358/2000: + Epoch Time: 24.658561944961548 +Epoch 1359/2000: + Epoch Time: 36.22329306602478 +Epoch 1360/2000: + Epoch Time: 24.563555479049683 +Epoch 1361/2000: + Epoch Time: 35.22726655006409 +Epoch 1362/2000: + Epoch Time: 24.523354530334473 +Epoch 1363/2000: + Epoch Time: 35.11151957511902 +Epoch 1364/2000: + Epoch Time: 24.500375747680664 +Epoch 1365/2000: + Epoch Time: 24.58806538581848 +Epoch 1366/2000: + Epoch Time: 35.083091259002686 +Epoch 1367/2000: + Epoch Time: 24.712035655975342 +Epoch 1368/2000: + Epoch Time: 35.37726068496704 +Epoch 1369/2000: + Epoch Time: 25.03089213371277 +Epoch 1370/2000: + Epoch Time: 35.29121923446655 +Epoch 1371/2000: + Epoch Time: 24.487552165985107 +Epoch 1372/2000: + Epoch Time: 24.583556652069092 +Epoch 1373/2000: + Epoch Time: 35.41636848449707 +Epoch 1374/2000: + Epoch Time: 24.584968328475952 +Epoch 1375/2000: + Epoch Time: 35.198880434036255 +Epoch 1376/2000: + Epoch Time: 24.51045536994934 +Epoch 1377/2000: + Epoch Time: 35.12532067298889 +Epoch 1378/2000: + Epoch Time: 25.14233112335205 +Epoch 1379/2000: + Epoch Time: 25.08599066734314 +Epoch 1380/2000: + Epoch Time: 35.23468065261841 +Epoch 1381/2000: + Epoch Time: 24.523180723190308 +Epoch 1382/2000: + Epoch Time: 35.11958408355713 +Epoch 1383/2000: + Epoch Time: 24.56699228286743 +Epoch 1384/2000: + Epoch Time: 35.14923644065857 +Epoch 1385/2000: + Epoch Time: 24.489514112472534 +Epoch 1386/2000: + Epoch Time: 24.541484355926514 +Epoch 1387/2000: + Epoch Time: 35.3024685382843 +Epoch 1388/2000: + Epoch Time: 25.003883361816406 +Epoch 1389/2000: + Epoch Time: 35.55936551094055 +Epoch 1390/2000: + Epoch Time: 24.566677570343018 +Epoch 1391/2000: + Epoch Time: 35.10496258735657 +Epoch 1392/2000: + Epoch Time: 24.534905910491943 +Epoch 1393/2000: + Epoch Time: 24.59539532661438 +Epoch 1394/2000: + Epoch Time: 35.190343618392944 +Epoch 1395/2000: + Epoch Time: 24.589751720428467 +Epoch 1396/2000: + Epoch Time: 35.22589421272278 +Epoch 1397/2000: + Epoch Time: 24.813442945480347 +Epoch 1398/2000: + Epoch Time: 35.83045148849487 +Epoch 1399/2000: + Epoch Time: 24.690972328186035 +Epoch 1400/2000: + Epoch Time: 35.07359194755554 +Epoch 1401/2000: + Epoch Time: 24.561473846435547 +Epoch 1402/2000: + Epoch Time: 24.647512674331665 +Epoch 1403/2000: + Epoch Time: 35.10578489303589 +Epoch 1404/2000: + Epoch Time: 24.69337558746338 +Epoch 1405/2000: + Epoch Time: 35.1392035484314 +Epoch 1406/2000: + Epoch Time: 24.507755756378174 +Epoch 1407/2000: + Epoch Time: 35.836347341537476 +Epoch 1408/2000: + Epoch Time: 25.104654788970947 +Epoch 1409/2000: + Epoch Time: 24.544372081756592 +Epoch 1410/2000: + Epoch Time: 35.29831337928772 +Epoch 1411/2000: + Epoch Time: 24.571781396865845 +Epoch 1412/2000: + Epoch Time: 35.1824004650116 +Epoch 1413/2000: + Epoch Time: 24.594751119613647 +Epoch 1414/2000: + Epoch Time: 35.41538667678833 +Epoch 1415/2000: + Epoch Time: 24.645216703414917 +Epoch 1416/2000: + Epoch Time: 24.542276620864868 +Epoch 1417/2000: + Epoch Time: 35.71958351135254 +Epoch 1418/2000: + Epoch Time: 25.0130455493927 +Epoch 1419/2000: + Epoch Time: 35.087303161621094 +Epoch 1420/2000: + Epoch Time: 24.5174822807312 +Epoch 1421/2000: + Epoch Time: 35.171499490737915 +Epoch 1422/2000: + Epoch Time: 24.54127311706543 +Epoch 1423/2000: + Epoch Time: 24.48239541053772 +Epoch 1424/2000: + Epoch Time: 35.118759632110596 +Epoch 1425/2000: + Epoch Time: 24.552196741104126 +Epoch 1426/2000: + Epoch Time: 35.085798263549805 +Epoch 1427/2000: + Epoch Time: 24.883482933044434 +Epoch 1428/2000: + Epoch Time: 35.50458860397339 +Epoch 1429/2000: + Epoch Time: 24.532615423202515 +Epoch 1430/2000: + Epoch Time: 24.475863218307495 +Epoch 1431/2000: + Epoch Time: 35.3510684967041 +Epoch 1432/2000: + Epoch Time: 24.471534252166748 +Epoch 1433/2000: + Epoch Time: 35.17480707168579 +Epoch 1434/2000: + Epoch Time: 24.497188091278076 +Epoch 1435/2000: + Epoch Time: 35.13350486755371 +Epoch 1436/2000: + Epoch Time: 24.558109045028687 +Epoch 1437/2000: + Epoch Time: 25.104761838912964 +Epoch 1438/2000: + Epoch Time: 35.49600601196289 +Epoch 1439/2000: + Epoch Time: 24.61414623260498 +Epoch 1440/2000: + Epoch Time: 35.06105017662048 +Epoch 1441/2000: + Epoch Time: 24.56773066520691 +Epoch 1442/2000: + Epoch Time: 35.17521691322327 +Epoch 1443/2000: + Epoch Time: 24.499608039855957 +Epoch 1444/2000: + Epoch Time: 24.512129545211792 +Epoch 1445/2000: + Epoch Time: 35.402896881103516 +Epoch 1446/2000: + Epoch Time: 24.584123373031616 +Epoch 1447/2000: + Epoch Time: 35.73756504058838 +Epoch 1448/2000: + Epoch Time: 24.718252420425415 +Epoch 1449/2000: + Epoch Time: 35.02158045768738 +Epoch 1450/2000: + Epoch Time: 24.529862880706787 +Epoch 1451/2000: + Epoch Time: 24.572364330291748 +Epoch 1452/2000: + Epoch Time: 35.25220966339111 +Epoch 1453/2000: + Epoch Time: 24.53545331954956 +Epoch 1454/2000: + Epoch Time: 35.09035897254944 +Epoch 1455/2000: + Epoch Time: 24.58426785469055 +Epoch 1456/2000: + Epoch Time: 35.44698166847229 +Epoch 1457/2000: + Epoch Time: 25.1268892288208 +Epoch 1458/2000: + Epoch Time: 25.111512899398804 +Epoch 1459/2000: + Epoch Time: 35.49708557128906 +Epoch 1460/2000: + Epoch Time: 24.619596481323242 +Epoch 1461/2000: + Epoch Time: 35.35200238227844 +Epoch 1462/2000: + Epoch Time: 24.562833070755005 +Epoch 1463/2000: + Epoch Time: 35.184563398361206 +Epoch 1464/2000: + Epoch Time: 24.549710035324097 +Epoch 1465/2000: + Epoch Time: 24.499753952026367 +Epoch 1466/2000: + Epoch Time: 35.547099351882935 +Epoch 1467/2000: + Epoch Time: 25.158963918685913 +Epoch 1468/2000: + Epoch Time: 35.25414156913757 +Epoch 1469/2000: + Epoch Time: 24.599037408828735 +Epoch 1470/2000: + Epoch Time: 35.25198578834534 diff --git a/Pretrained Base Model/vocab.txt b/Pretrained Base Model/vocab.txt new file mode 100644 index 0000000000000000000000000000000000000000..cc9ad53732e21ea9180f18e1fb67a38cc5702d4c --- /dev/null +++ b/Pretrained Base Model/vocab.txt @@ -0,0 +1,4101 @@ +[PAD] +[UNK] +[CLS] +[SEP] +[MASK] +AAAAAA +AAAAAT +AAAAAC +AAAAAG +AAAATA +AAAATT +AAAATC +AAAATG +AAAACA +AAAACT +AAAACC +AAAACG +AAAAGA +AAAAGT +AAAAGC +AAAAGG +AAATAA +AAATAT +AAATAC +AAATAG +AAATTA +AAATTT +AAATTC +AAATTG +AAATCA +AAATCT +AAATCC +AAATCG +AAATGA +AAATGT +AAATGC +AAATGG +AAACAA +AAACAT +AAACAC +AAACAG +AAACTA +AAACTT +AAACTC +AAACTG +AAACCA +AAACCT +AAACCC +AAACCG +AAACGA +AAACGT +AAACGC +AAACGG +AAAGAA +AAAGAT +AAAGAC +AAAGAG +AAAGTA +AAAGTT +AAAGTC +AAAGTG +AAAGCA +AAAGCT +AAAGCC +AAAGCG +AAAGGA +AAAGGT +AAAGGC +AAAGGG +AATAAA +AATAAT +AATAAC +AATAAG +AATATA +AATATT +AATATC +AATATG +AATACA +AATACT +AATACC +AATACG +AATAGA +AATAGT +AATAGC +AATAGG +AATTAA +AATTAT +AATTAC +AATTAG +AATTTA +AATTTT +AATTTC +AATTTG +AATTCA +AATTCT +AATTCC +AATTCG +AATTGA +AATTGT +AATTGC +AATTGG +AATCAA +AATCAT +AATCAC +AATCAG +AATCTA +AATCTT +AATCTC +AATCTG +AATCCA +AATCCT +AATCCC +AATCCG +AATCGA +AATCGT +AATCGC +AATCGG +AATGAA +AATGAT +AATGAC +AATGAG +AATGTA +AATGTT +AATGTC +AATGTG +AATGCA +AATGCT +AATGCC +AATGCG +AATGGA +AATGGT +AATGGC +AATGGG +AACAAA +AACAAT +AACAAC +AACAAG +AACATA +AACATT +AACATC +AACATG +AACACA +AACACT +AACACC +AACACG +AACAGA +AACAGT +AACAGC +AACAGG +AACTAA +AACTAT +AACTAC +AACTAG +AACTTA +AACTTT +AACTTC +AACTTG +AACTCA +AACTCT +AACTCC +AACTCG +AACTGA +AACTGT +AACTGC +AACTGG +AACCAA +AACCAT +AACCAC +AACCAG +AACCTA +AACCTT +AACCTC +AACCTG +AACCCA +AACCCT +AACCCC +AACCCG +AACCGA +AACCGT +AACCGC +AACCGG +AACGAA +AACGAT +AACGAC +AACGAG +AACGTA +AACGTT +AACGTC +AACGTG +AACGCA +AACGCT +AACGCC +AACGCG +AACGGA +AACGGT +AACGGC +AACGGG +AAGAAA +AAGAAT +AAGAAC +AAGAAG +AAGATA +AAGATT +AAGATC +AAGATG +AAGACA +AAGACT +AAGACC +AAGACG +AAGAGA +AAGAGT +AAGAGC +AAGAGG +AAGTAA +AAGTAT +AAGTAC +AAGTAG +AAGTTA +AAGTTT +AAGTTC +AAGTTG +AAGTCA +AAGTCT +AAGTCC +AAGTCG +AAGTGA +AAGTGT +AAGTGC +AAGTGG +AAGCAA +AAGCAT +AAGCAC +AAGCAG +AAGCTA +AAGCTT +AAGCTC +AAGCTG +AAGCCA +AAGCCT +AAGCCC +AAGCCG +AAGCGA +AAGCGT +AAGCGC +AAGCGG +AAGGAA +AAGGAT +AAGGAC +AAGGAG +AAGGTA +AAGGTT +AAGGTC +AAGGTG +AAGGCA +AAGGCT +AAGGCC +AAGGCG +AAGGGA +AAGGGT +AAGGGC +AAGGGG +ATAAAA +ATAAAT +ATAAAC +ATAAAG +ATAATA +ATAATT +ATAATC +ATAATG +ATAACA +ATAACT +ATAACC +ATAACG +ATAAGA +ATAAGT +ATAAGC +ATAAGG +ATATAA +ATATAT +ATATAC +ATATAG +ATATTA +ATATTT +ATATTC +ATATTG +ATATCA +ATATCT +ATATCC +ATATCG +ATATGA +ATATGT +ATATGC +ATATGG +ATACAA +ATACAT +ATACAC +ATACAG +ATACTA +ATACTT +ATACTC +ATACTG +ATACCA +ATACCT +ATACCC +ATACCG +ATACGA +ATACGT +ATACGC +ATACGG +ATAGAA +ATAGAT +ATAGAC +ATAGAG +ATAGTA +ATAGTT +ATAGTC +ATAGTG +ATAGCA +ATAGCT +ATAGCC +ATAGCG +ATAGGA +ATAGGT +ATAGGC +ATAGGG +ATTAAA +ATTAAT +ATTAAC +ATTAAG +ATTATA +ATTATT +ATTATC +ATTATG +ATTACA +ATTACT +ATTACC +ATTACG +ATTAGA +ATTAGT +ATTAGC +ATTAGG +ATTTAA +ATTTAT +ATTTAC +ATTTAG +ATTTTA +ATTTTT +ATTTTC +ATTTTG +ATTTCA +ATTTCT +ATTTCC +ATTTCG +ATTTGA +ATTTGT +ATTTGC +ATTTGG +ATTCAA +ATTCAT +ATTCAC +ATTCAG +ATTCTA +ATTCTT +ATTCTC +ATTCTG +ATTCCA +ATTCCT +ATTCCC +ATTCCG +ATTCGA +ATTCGT +ATTCGC +ATTCGG +ATTGAA +ATTGAT +ATTGAC +ATTGAG +ATTGTA +ATTGTT +ATTGTC +ATTGTG +ATTGCA +ATTGCT +ATTGCC +ATTGCG +ATTGGA +ATTGGT +ATTGGC +ATTGGG +ATCAAA +ATCAAT +ATCAAC +ATCAAG +ATCATA +ATCATT +ATCATC +ATCATG +ATCACA +ATCACT +ATCACC +ATCACG +ATCAGA +ATCAGT +ATCAGC +ATCAGG +ATCTAA +ATCTAT +ATCTAC +ATCTAG +ATCTTA +ATCTTT +ATCTTC +ATCTTG +ATCTCA +ATCTCT +ATCTCC +ATCTCG +ATCTGA +ATCTGT +ATCTGC +ATCTGG +ATCCAA +ATCCAT +ATCCAC +ATCCAG +ATCCTA +ATCCTT +ATCCTC +ATCCTG +ATCCCA +ATCCCT +ATCCCC +ATCCCG +ATCCGA +ATCCGT +ATCCGC +ATCCGG +ATCGAA +ATCGAT +ATCGAC +ATCGAG +ATCGTA +ATCGTT +ATCGTC +ATCGTG +ATCGCA +ATCGCT +ATCGCC +ATCGCG +ATCGGA +ATCGGT +ATCGGC +ATCGGG +ATGAAA +ATGAAT +ATGAAC +ATGAAG +ATGATA +ATGATT +ATGATC +ATGATG +ATGACA +ATGACT +ATGACC +ATGACG +ATGAGA +ATGAGT +ATGAGC +ATGAGG +ATGTAA +ATGTAT +ATGTAC +ATGTAG +ATGTTA +ATGTTT +ATGTTC +ATGTTG +ATGTCA +ATGTCT +ATGTCC +ATGTCG +ATGTGA +ATGTGT +ATGTGC +ATGTGG +ATGCAA +ATGCAT +ATGCAC +ATGCAG +ATGCTA +ATGCTT +ATGCTC +ATGCTG +ATGCCA +ATGCCT +ATGCCC +ATGCCG +ATGCGA +ATGCGT +ATGCGC +ATGCGG +ATGGAA +ATGGAT +ATGGAC +ATGGAG +ATGGTA +ATGGTT +ATGGTC +ATGGTG +ATGGCA +ATGGCT +ATGGCC +ATGGCG +ATGGGA +ATGGGT +ATGGGC +ATGGGG +ACAAAA +ACAAAT +ACAAAC +ACAAAG +ACAATA +ACAATT +ACAATC +ACAATG +ACAACA +ACAACT +ACAACC +ACAACG +ACAAGA +ACAAGT +ACAAGC +ACAAGG +ACATAA +ACATAT +ACATAC +ACATAG +ACATTA +ACATTT +ACATTC +ACATTG +ACATCA +ACATCT +ACATCC +ACATCG +ACATGA +ACATGT +ACATGC +ACATGG +ACACAA +ACACAT +ACACAC +ACACAG +ACACTA +ACACTT +ACACTC +ACACTG +ACACCA +ACACCT +ACACCC +ACACCG +ACACGA +ACACGT +ACACGC +ACACGG +ACAGAA +ACAGAT +ACAGAC +ACAGAG +ACAGTA +ACAGTT +ACAGTC +ACAGTG +ACAGCA +ACAGCT +ACAGCC +ACAGCG +ACAGGA +ACAGGT +ACAGGC +ACAGGG +ACTAAA +ACTAAT +ACTAAC +ACTAAG +ACTATA +ACTATT +ACTATC +ACTATG +ACTACA +ACTACT +ACTACC +ACTACG +ACTAGA +ACTAGT +ACTAGC +ACTAGG +ACTTAA +ACTTAT +ACTTAC +ACTTAG +ACTTTA +ACTTTT +ACTTTC +ACTTTG +ACTTCA +ACTTCT +ACTTCC +ACTTCG +ACTTGA +ACTTGT +ACTTGC +ACTTGG +ACTCAA +ACTCAT +ACTCAC +ACTCAG +ACTCTA +ACTCTT +ACTCTC +ACTCTG +ACTCCA +ACTCCT +ACTCCC +ACTCCG +ACTCGA +ACTCGT +ACTCGC +ACTCGG +ACTGAA +ACTGAT +ACTGAC +ACTGAG +ACTGTA +ACTGTT +ACTGTC +ACTGTG +ACTGCA +ACTGCT +ACTGCC +ACTGCG +ACTGGA +ACTGGT +ACTGGC +ACTGGG +ACCAAA +ACCAAT +ACCAAC +ACCAAG +ACCATA +ACCATT +ACCATC +ACCATG +ACCACA +ACCACT +ACCACC +ACCACG +ACCAGA +ACCAGT +ACCAGC +ACCAGG +ACCTAA +ACCTAT +ACCTAC +ACCTAG +ACCTTA +ACCTTT +ACCTTC +ACCTTG +ACCTCA +ACCTCT +ACCTCC +ACCTCG +ACCTGA +ACCTGT +ACCTGC +ACCTGG +ACCCAA +ACCCAT +ACCCAC +ACCCAG +ACCCTA +ACCCTT +ACCCTC +ACCCTG +ACCCCA +ACCCCT +ACCCCC +ACCCCG +ACCCGA +ACCCGT +ACCCGC +ACCCGG +ACCGAA +ACCGAT +ACCGAC +ACCGAG +ACCGTA +ACCGTT +ACCGTC +ACCGTG +ACCGCA +ACCGCT +ACCGCC +ACCGCG +ACCGGA +ACCGGT +ACCGGC +ACCGGG +ACGAAA +ACGAAT +ACGAAC +ACGAAG +ACGATA +ACGATT +ACGATC +ACGATG +ACGACA +ACGACT +ACGACC +ACGACG +ACGAGA +ACGAGT +ACGAGC +ACGAGG +ACGTAA +ACGTAT +ACGTAC +ACGTAG +ACGTTA +ACGTTT +ACGTTC +ACGTTG +ACGTCA +ACGTCT +ACGTCC +ACGTCG +ACGTGA +ACGTGT +ACGTGC +ACGTGG +ACGCAA +ACGCAT +ACGCAC +ACGCAG +ACGCTA +ACGCTT +ACGCTC +ACGCTG +ACGCCA +ACGCCT +ACGCCC +ACGCCG +ACGCGA +ACGCGT +ACGCGC +ACGCGG +ACGGAA +ACGGAT +ACGGAC +ACGGAG +ACGGTA +ACGGTT +ACGGTC +ACGGTG +ACGGCA +ACGGCT +ACGGCC +ACGGCG +ACGGGA +ACGGGT +ACGGGC +ACGGGG +AGAAAA +AGAAAT +AGAAAC +AGAAAG +AGAATA +AGAATT +AGAATC +AGAATG +AGAACA +AGAACT +AGAACC +AGAACG +AGAAGA +AGAAGT +AGAAGC +AGAAGG +AGATAA +AGATAT +AGATAC +AGATAG +AGATTA +AGATTT +AGATTC +AGATTG +AGATCA +AGATCT +AGATCC +AGATCG +AGATGA +AGATGT +AGATGC +AGATGG +AGACAA +AGACAT +AGACAC +AGACAG +AGACTA +AGACTT +AGACTC +AGACTG +AGACCA +AGACCT +AGACCC +AGACCG +AGACGA +AGACGT +AGACGC +AGACGG +AGAGAA +AGAGAT +AGAGAC +AGAGAG +AGAGTA +AGAGTT +AGAGTC +AGAGTG +AGAGCA +AGAGCT +AGAGCC +AGAGCG +AGAGGA +AGAGGT +AGAGGC +AGAGGG +AGTAAA +AGTAAT +AGTAAC +AGTAAG +AGTATA +AGTATT +AGTATC +AGTATG +AGTACA +AGTACT +AGTACC +AGTACG +AGTAGA +AGTAGT +AGTAGC +AGTAGG +AGTTAA +AGTTAT +AGTTAC +AGTTAG +AGTTTA +AGTTTT +AGTTTC +AGTTTG +AGTTCA +AGTTCT +AGTTCC +AGTTCG +AGTTGA +AGTTGT +AGTTGC +AGTTGG +AGTCAA +AGTCAT +AGTCAC +AGTCAG +AGTCTA +AGTCTT +AGTCTC +AGTCTG +AGTCCA +AGTCCT +AGTCCC +AGTCCG +AGTCGA +AGTCGT +AGTCGC +AGTCGG +AGTGAA +AGTGAT +AGTGAC +AGTGAG +AGTGTA +AGTGTT +AGTGTC +AGTGTG +AGTGCA +AGTGCT +AGTGCC +AGTGCG +AGTGGA +AGTGGT +AGTGGC +AGTGGG +AGCAAA +AGCAAT +AGCAAC +AGCAAG +AGCATA +AGCATT +AGCATC +AGCATG +AGCACA +AGCACT +AGCACC +AGCACG +AGCAGA +AGCAGT +AGCAGC +AGCAGG +AGCTAA +AGCTAT +AGCTAC +AGCTAG +AGCTTA +AGCTTT +AGCTTC +AGCTTG +AGCTCA +AGCTCT +AGCTCC +AGCTCG +AGCTGA +AGCTGT +AGCTGC +AGCTGG +AGCCAA +AGCCAT +AGCCAC +AGCCAG +AGCCTA +AGCCTT +AGCCTC +AGCCTG +AGCCCA +AGCCCT +AGCCCC +AGCCCG +AGCCGA +AGCCGT +AGCCGC +AGCCGG +AGCGAA +AGCGAT +AGCGAC +AGCGAG +AGCGTA +AGCGTT +AGCGTC +AGCGTG +AGCGCA +AGCGCT +AGCGCC +AGCGCG +AGCGGA +AGCGGT +AGCGGC +AGCGGG +AGGAAA +AGGAAT +AGGAAC +AGGAAG +AGGATA +AGGATT +AGGATC +AGGATG +AGGACA +AGGACT +AGGACC +AGGACG +AGGAGA +AGGAGT +AGGAGC +AGGAGG +AGGTAA +AGGTAT +AGGTAC +AGGTAG +AGGTTA +AGGTTT +AGGTTC +AGGTTG +AGGTCA +AGGTCT +AGGTCC +AGGTCG +AGGTGA +AGGTGT +AGGTGC +AGGTGG +AGGCAA +AGGCAT +AGGCAC +AGGCAG +AGGCTA +AGGCTT +AGGCTC +AGGCTG +AGGCCA +AGGCCT +AGGCCC +AGGCCG +AGGCGA +AGGCGT +AGGCGC +AGGCGG +AGGGAA +AGGGAT +AGGGAC +AGGGAG +AGGGTA +AGGGTT +AGGGTC +AGGGTG +AGGGCA +AGGGCT +AGGGCC +AGGGCG +AGGGGA +AGGGGT +AGGGGC +AGGGGG +TAAAAA +TAAAAT +TAAAAC +TAAAAG +TAAATA +TAAATT +TAAATC +TAAATG +TAAACA +TAAACT +TAAACC +TAAACG +TAAAGA +TAAAGT +TAAAGC +TAAAGG +TAATAA +TAATAT +TAATAC +TAATAG +TAATTA +TAATTT +TAATTC +TAATTG +TAATCA +TAATCT +TAATCC +TAATCG +TAATGA +TAATGT +TAATGC +TAATGG +TAACAA +TAACAT +TAACAC +TAACAG +TAACTA +TAACTT +TAACTC +TAACTG +TAACCA +TAACCT +TAACCC +TAACCG +TAACGA +TAACGT +TAACGC +TAACGG +TAAGAA +TAAGAT +TAAGAC +TAAGAG +TAAGTA +TAAGTT +TAAGTC +TAAGTG +TAAGCA +TAAGCT +TAAGCC +TAAGCG +TAAGGA +TAAGGT +TAAGGC +TAAGGG +TATAAA +TATAAT +TATAAC +TATAAG +TATATA +TATATT +TATATC +TATATG +TATACA +TATACT +TATACC +TATACG +TATAGA +TATAGT +TATAGC +TATAGG +TATTAA +TATTAT +TATTAC +TATTAG +TATTTA +TATTTT +TATTTC +TATTTG +TATTCA +TATTCT +TATTCC +TATTCG +TATTGA +TATTGT +TATTGC +TATTGG +TATCAA +TATCAT +TATCAC +TATCAG +TATCTA +TATCTT +TATCTC +TATCTG +TATCCA +TATCCT +TATCCC +TATCCG +TATCGA +TATCGT +TATCGC +TATCGG +TATGAA +TATGAT +TATGAC +TATGAG +TATGTA +TATGTT +TATGTC +TATGTG +TATGCA +TATGCT +TATGCC +TATGCG +TATGGA +TATGGT +TATGGC +TATGGG +TACAAA +TACAAT +TACAAC +TACAAG +TACATA +TACATT +TACATC +TACATG +TACACA +TACACT +TACACC +TACACG +TACAGA +TACAGT +TACAGC +TACAGG +TACTAA +TACTAT +TACTAC +TACTAG +TACTTA +TACTTT +TACTTC +TACTTG +TACTCA +TACTCT +TACTCC +TACTCG +TACTGA +TACTGT +TACTGC +TACTGG +TACCAA +TACCAT +TACCAC +TACCAG +TACCTA +TACCTT +TACCTC +TACCTG +TACCCA +TACCCT +TACCCC +TACCCG +TACCGA +TACCGT +TACCGC +TACCGG +TACGAA +TACGAT +TACGAC +TACGAG +TACGTA +TACGTT +TACGTC +TACGTG +TACGCA +TACGCT +TACGCC +TACGCG +TACGGA +TACGGT +TACGGC +TACGGG +TAGAAA +TAGAAT +TAGAAC +TAGAAG +TAGATA +TAGATT +TAGATC +TAGATG +TAGACA +TAGACT +TAGACC +TAGACG +TAGAGA +TAGAGT +TAGAGC +TAGAGG +TAGTAA +TAGTAT +TAGTAC +TAGTAG +TAGTTA +TAGTTT +TAGTTC +TAGTTG +TAGTCA +TAGTCT +TAGTCC +TAGTCG +TAGTGA +TAGTGT +TAGTGC +TAGTGG +TAGCAA +TAGCAT +TAGCAC +TAGCAG +TAGCTA +TAGCTT +TAGCTC +TAGCTG +TAGCCA +TAGCCT +TAGCCC +TAGCCG +TAGCGA +TAGCGT +TAGCGC +TAGCGG +TAGGAA +TAGGAT +TAGGAC +TAGGAG +TAGGTA +TAGGTT +TAGGTC +TAGGTG +TAGGCA +TAGGCT +TAGGCC +TAGGCG +TAGGGA +TAGGGT +TAGGGC +TAGGGG +TTAAAA +TTAAAT +TTAAAC +TTAAAG +TTAATA +TTAATT +TTAATC +TTAATG +TTAACA +TTAACT +TTAACC +TTAACG +TTAAGA +TTAAGT +TTAAGC +TTAAGG +TTATAA +TTATAT +TTATAC +TTATAG +TTATTA +TTATTT +TTATTC +TTATTG +TTATCA +TTATCT +TTATCC +TTATCG +TTATGA +TTATGT +TTATGC +TTATGG +TTACAA +TTACAT +TTACAC +TTACAG +TTACTA +TTACTT +TTACTC +TTACTG +TTACCA +TTACCT +TTACCC +TTACCG +TTACGA +TTACGT +TTACGC +TTACGG +TTAGAA +TTAGAT +TTAGAC +TTAGAG +TTAGTA +TTAGTT +TTAGTC +TTAGTG +TTAGCA +TTAGCT +TTAGCC +TTAGCG +TTAGGA +TTAGGT +TTAGGC +TTAGGG +TTTAAA +TTTAAT +TTTAAC +TTTAAG +TTTATA +TTTATT +TTTATC +TTTATG +TTTACA +TTTACT +TTTACC +TTTACG +TTTAGA +TTTAGT +TTTAGC +TTTAGG +TTTTAA +TTTTAT +TTTTAC +TTTTAG +TTTTTA +TTTTTT +TTTTTC +TTTTTG +TTTTCA +TTTTCT +TTTTCC +TTTTCG +TTTTGA +TTTTGT +TTTTGC +TTTTGG +TTTCAA +TTTCAT +TTTCAC +TTTCAG +TTTCTA +TTTCTT +TTTCTC +TTTCTG +TTTCCA +TTTCCT +TTTCCC +TTTCCG +TTTCGA +TTTCGT +TTTCGC +TTTCGG +TTTGAA +TTTGAT +TTTGAC +TTTGAG +TTTGTA +TTTGTT +TTTGTC +TTTGTG +TTTGCA +TTTGCT +TTTGCC +TTTGCG +TTTGGA +TTTGGT +TTTGGC +TTTGGG +TTCAAA +TTCAAT +TTCAAC +TTCAAG +TTCATA +TTCATT +TTCATC +TTCATG +TTCACA +TTCACT +TTCACC +TTCACG +TTCAGA +TTCAGT +TTCAGC +TTCAGG +TTCTAA +TTCTAT +TTCTAC +TTCTAG +TTCTTA +TTCTTT +TTCTTC +TTCTTG +TTCTCA +TTCTCT +TTCTCC +TTCTCG +TTCTGA +TTCTGT +TTCTGC +TTCTGG +TTCCAA +TTCCAT +TTCCAC +TTCCAG +TTCCTA +TTCCTT +TTCCTC +TTCCTG +TTCCCA +TTCCCT +TTCCCC +TTCCCG +TTCCGA +TTCCGT +TTCCGC +TTCCGG +TTCGAA +TTCGAT +TTCGAC +TTCGAG +TTCGTA +TTCGTT +TTCGTC +TTCGTG +TTCGCA +TTCGCT +TTCGCC +TTCGCG +TTCGGA +TTCGGT +TTCGGC +TTCGGG +TTGAAA +TTGAAT +TTGAAC +TTGAAG +TTGATA +TTGATT +TTGATC +TTGATG +TTGACA +TTGACT +TTGACC +TTGACG +TTGAGA +TTGAGT +TTGAGC +TTGAGG +TTGTAA +TTGTAT +TTGTAC +TTGTAG +TTGTTA +TTGTTT +TTGTTC +TTGTTG +TTGTCA +TTGTCT +TTGTCC +TTGTCG +TTGTGA +TTGTGT +TTGTGC +TTGTGG +TTGCAA +TTGCAT +TTGCAC +TTGCAG +TTGCTA +TTGCTT +TTGCTC +TTGCTG +TTGCCA +TTGCCT +TTGCCC +TTGCCG +TTGCGA +TTGCGT +TTGCGC +TTGCGG +TTGGAA +TTGGAT +TTGGAC +TTGGAG +TTGGTA +TTGGTT +TTGGTC +TTGGTG +TTGGCA +TTGGCT +TTGGCC +TTGGCG +TTGGGA +TTGGGT +TTGGGC +TTGGGG +TCAAAA +TCAAAT +TCAAAC +TCAAAG +TCAATA +TCAATT +TCAATC +TCAATG +TCAACA +TCAACT +TCAACC +TCAACG +TCAAGA +TCAAGT +TCAAGC +TCAAGG +TCATAA +TCATAT +TCATAC +TCATAG +TCATTA +TCATTT +TCATTC +TCATTG +TCATCA +TCATCT +TCATCC +TCATCG +TCATGA +TCATGT +TCATGC +TCATGG +TCACAA +TCACAT +TCACAC +TCACAG +TCACTA +TCACTT +TCACTC +TCACTG +TCACCA +TCACCT +TCACCC +TCACCG +TCACGA +TCACGT +TCACGC +TCACGG +TCAGAA +TCAGAT +TCAGAC +TCAGAG +TCAGTA +TCAGTT +TCAGTC +TCAGTG +TCAGCA +TCAGCT +TCAGCC +TCAGCG +TCAGGA +TCAGGT +TCAGGC +TCAGGG +TCTAAA +TCTAAT +TCTAAC +TCTAAG +TCTATA +TCTATT +TCTATC +TCTATG +TCTACA +TCTACT +TCTACC +TCTACG +TCTAGA +TCTAGT +TCTAGC +TCTAGG +TCTTAA +TCTTAT +TCTTAC +TCTTAG +TCTTTA +TCTTTT +TCTTTC +TCTTTG +TCTTCA +TCTTCT +TCTTCC +TCTTCG +TCTTGA +TCTTGT +TCTTGC +TCTTGG +TCTCAA +TCTCAT +TCTCAC +TCTCAG +TCTCTA +TCTCTT +TCTCTC +TCTCTG +TCTCCA +TCTCCT +TCTCCC +TCTCCG +TCTCGA +TCTCGT +TCTCGC +TCTCGG +TCTGAA +TCTGAT +TCTGAC +TCTGAG +TCTGTA +TCTGTT +TCTGTC +TCTGTG +TCTGCA +TCTGCT +TCTGCC +TCTGCG +TCTGGA +TCTGGT +TCTGGC +TCTGGG +TCCAAA +TCCAAT +TCCAAC +TCCAAG +TCCATA +TCCATT +TCCATC +TCCATG +TCCACA +TCCACT +TCCACC +TCCACG +TCCAGA +TCCAGT +TCCAGC +TCCAGG +TCCTAA +TCCTAT +TCCTAC +TCCTAG +TCCTTA +TCCTTT +TCCTTC +TCCTTG +TCCTCA +TCCTCT +TCCTCC +TCCTCG +TCCTGA +TCCTGT +TCCTGC +TCCTGG +TCCCAA +TCCCAT +TCCCAC +TCCCAG +TCCCTA +TCCCTT +TCCCTC +TCCCTG +TCCCCA +TCCCCT +TCCCCC +TCCCCG +TCCCGA +TCCCGT +TCCCGC +TCCCGG +TCCGAA +TCCGAT +TCCGAC +TCCGAG +TCCGTA +TCCGTT +TCCGTC +TCCGTG +TCCGCA +TCCGCT +TCCGCC +TCCGCG +TCCGGA +TCCGGT +TCCGGC +TCCGGG +TCGAAA +TCGAAT +TCGAAC +TCGAAG +TCGATA +TCGATT +TCGATC +TCGATG +TCGACA +TCGACT +TCGACC +TCGACG +TCGAGA +TCGAGT +TCGAGC +TCGAGG +TCGTAA +TCGTAT +TCGTAC +TCGTAG +TCGTTA +TCGTTT +TCGTTC +TCGTTG +TCGTCA +TCGTCT +TCGTCC +TCGTCG +TCGTGA +TCGTGT +TCGTGC +TCGTGG +TCGCAA +TCGCAT +TCGCAC +TCGCAG +TCGCTA +TCGCTT +TCGCTC +TCGCTG +TCGCCA +TCGCCT +TCGCCC +TCGCCG +TCGCGA +TCGCGT +TCGCGC +TCGCGG +TCGGAA +TCGGAT +TCGGAC +TCGGAG +TCGGTA +TCGGTT +TCGGTC +TCGGTG +TCGGCA +TCGGCT +TCGGCC +TCGGCG +TCGGGA +TCGGGT +TCGGGC +TCGGGG +TGAAAA +TGAAAT +TGAAAC +TGAAAG +TGAATA +TGAATT +TGAATC +TGAATG +TGAACA +TGAACT +TGAACC +TGAACG +TGAAGA +TGAAGT +TGAAGC +TGAAGG +TGATAA +TGATAT +TGATAC +TGATAG +TGATTA +TGATTT +TGATTC +TGATTG +TGATCA +TGATCT +TGATCC +TGATCG +TGATGA +TGATGT +TGATGC +TGATGG +TGACAA +TGACAT +TGACAC +TGACAG +TGACTA +TGACTT +TGACTC +TGACTG +TGACCA +TGACCT +TGACCC +TGACCG +TGACGA +TGACGT +TGACGC +TGACGG +TGAGAA +TGAGAT +TGAGAC +TGAGAG +TGAGTA +TGAGTT +TGAGTC +TGAGTG +TGAGCA +TGAGCT +TGAGCC +TGAGCG +TGAGGA +TGAGGT +TGAGGC +TGAGGG +TGTAAA +TGTAAT +TGTAAC +TGTAAG +TGTATA +TGTATT +TGTATC +TGTATG +TGTACA +TGTACT +TGTACC +TGTACG +TGTAGA +TGTAGT +TGTAGC +TGTAGG +TGTTAA +TGTTAT +TGTTAC +TGTTAG +TGTTTA +TGTTTT +TGTTTC +TGTTTG +TGTTCA +TGTTCT +TGTTCC +TGTTCG +TGTTGA +TGTTGT +TGTTGC +TGTTGG +TGTCAA +TGTCAT +TGTCAC +TGTCAG +TGTCTA +TGTCTT +TGTCTC +TGTCTG +TGTCCA +TGTCCT +TGTCCC +TGTCCG +TGTCGA +TGTCGT +TGTCGC +TGTCGG +TGTGAA +TGTGAT +TGTGAC +TGTGAG +TGTGTA +TGTGTT +TGTGTC +TGTGTG +TGTGCA +TGTGCT +TGTGCC +TGTGCG +TGTGGA +TGTGGT +TGTGGC +TGTGGG +TGCAAA +TGCAAT +TGCAAC +TGCAAG +TGCATA +TGCATT +TGCATC +TGCATG +TGCACA +TGCACT +TGCACC +TGCACG +TGCAGA +TGCAGT +TGCAGC +TGCAGG +TGCTAA +TGCTAT +TGCTAC +TGCTAG +TGCTTA +TGCTTT +TGCTTC +TGCTTG +TGCTCA +TGCTCT +TGCTCC +TGCTCG +TGCTGA +TGCTGT +TGCTGC +TGCTGG +TGCCAA +TGCCAT +TGCCAC +TGCCAG +TGCCTA +TGCCTT +TGCCTC +TGCCTG +TGCCCA +TGCCCT +TGCCCC +TGCCCG +TGCCGA +TGCCGT +TGCCGC +TGCCGG +TGCGAA +TGCGAT +TGCGAC +TGCGAG +TGCGTA +TGCGTT +TGCGTC +TGCGTG +TGCGCA +TGCGCT +TGCGCC +TGCGCG +TGCGGA +TGCGGT +TGCGGC +TGCGGG +TGGAAA +TGGAAT +TGGAAC +TGGAAG +TGGATA +TGGATT +TGGATC +TGGATG +TGGACA +TGGACT +TGGACC +TGGACG +TGGAGA +TGGAGT +TGGAGC +TGGAGG +TGGTAA +TGGTAT +TGGTAC +TGGTAG +TGGTTA +TGGTTT +TGGTTC +TGGTTG +TGGTCA +TGGTCT +TGGTCC +TGGTCG +TGGTGA +TGGTGT +TGGTGC +TGGTGG +TGGCAA +TGGCAT +TGGCAC +TGGCAG +TGGCTA +TGGCTT +TGGCTC +TGGCTG +TGGCCA +TGGCCT +TGGCCC +TGGCCG +TGGCGA +TGGCGT +TGGCGC +TGGCGG +TGGGAA +TGGGAT +TGGGAC +TGGGAG +TGGGTA +TGGGTT +TGGGTC +TGGGTG +TGGGCA +TGGGCT +TGGGCC +TGGGCG +TGGGGA +TGGGGT +TGGGGC +TGGGGG +CAAAAA +CAAAAT +CAAAAC +CAAAAG +CAAATA +CAAATT +CAAATC +CAAATG +CAAACA +CAAACT +CAAACC +CAAACG +CAAAGA +CAAAGT +CAAAGC +CAAAGG +CAATAA +CAATAT +CAATAC +CAATAG +CAATTA +CAATTT +CAATTC +CAATTG +CAATCA +CAATCT +CAATCC +CAATCG +CAATGA +CAATGT +CAATGC +CAATGG +CAACAA +CAACAT +CAACAC +CAACAG +CAACTA +CAACTT +CAACTC +CAACTG +CAACCA +CAACCT +CAACCC +CAACCG +CAACGA +CAACGT +CAACGC +CAACGG +CAAGAA +CAAGAT +CAAGAC +CAAGAG +CAAGTA +CAAGTT +CAAGTC +CAAGTG +CAAGCA +CAAGCT +CAAGCC +CAAGCG +CAAGGA +CAAGGT +CAAGGC +CAAGGG +CATAAA +CATAAT +CATAAC +CATAAG +CATATA +CATATT +CATATC +CATATG +CATACA +CATACT +CATACC +CATACG +CATAGA +CATAGT +CATAGC +CATAGG +CATTAA +CATTAT +CATTAC +CATTAG +CATTTA +CATTTT +CATTTC +CATTTG +CATTCA +CATTCT +CATTCC +CATTCG +CATTGA +CATTGT +CATTGC +CATTGG +CATCAA +CATCAT +CATCAC +CATCAG +CATCTA +CATCTT +CATCTC +CATCTG +CATCCA +CATCCT +CATCCC +CATCCG +CATCGA +CATCGT +CATCGC +CATCGG +CATGAA +CATGAT +CATGAC +CATGAG +CATGTA +CATGTT +CATGTC +CATGTG +CATGCA +CATGCT +CATGCC +CATGCG +CATGGA +CATGGT +CATGGC +CATGGG +CACAAA +CACAAT +CACAAC +CACAAG +CACATA +CACATT +CACATC +CACATG +CACACA +CACACT +CACACC +CACACG +CACAGA +CACAGT +CACAGC +CACAGG +CACTAA +CACTAT +CACTAC +CACTAG +CACTTA +CACTTT +CACTTC +CACTTG +CACTCA +CACTCT +CACTCC +CACTCG +CACTGA +CACTGT +CACTGC +CACTGG +CACCAA +CACCAT +CACCAC +CACCAG +CACCTA +CACCTT +CACCTC +CACCTG +CACCCA +CACCCT +CACCCC +CACCCG +CACCGA +CACCGT +CACCGC +CACCGG +CACGAA +CACGAT +CACGAC +CACGAG +CACGTA +CACGTT +CACGTC +CACGTG +CACGCA +CACGCT +CACGCC +CACGCG +CACGGA +CACGGT +CACGGC +CACGGG +CAGAAA +CAGAAT +CAGAAC +CAGAAG +CAGATA +CAGATT +CAGATC +CAGATG +CAGACA +CAGACT +CAGACC +CAGACG +CAGAGA +CAGAGT +CAGAGC +CAGAGG +CAGTAA +CAGTAT +CAGTAC +CAGTAG +CAGTTA +CAGTTT +CAGTTC +CAGTTG +CAGTCA +CAGTCT +CAGTCC +CAGTCG +CAGTGA +CAGTGT +CAGTGC +CAGTGG +CAGCAA +CAGCAT +CAGCAC +CAGCAG +CAGCTA +CAGCTT +CAGCTC +CAGCTG +CAGCCA +CAGCCT +CAGCCC +CAGCCG +CAGCGA +CAGCGT +CAGCGC +CAGCGG +CAGGAA +CAGGAT +CAGGAC +CAGGAG +CAGGTA +CAGGTT +CAGGTC +CAGGTG +CAGGCA +CAGGCT +CAGGCC +CAGGCG +CAGGGA +CAGGGT +CAGGGC +CAGGGG +CTAAAA +CTAAAT +CTAAAC +CTAAAG +CTAATA +CTAATT +CTAATC +CTAATG +CTAACA +CTAACT +CTAACC +CTAACG +CTAAGA +CTAAGT +CTAAGC +CTAAGG +CTATAA +CTATAT +CTATAC +CTATAG +CTATTA +CTATTT +CTATTC +CTATTG +CTATCA +CTATCT +CTATCC +CTATCG +CTATGA +CTATGT +CTATGC +CTATGG +CTACAA +CTACAT +CTACAC +CTACAG +CTACTA +CTACTT +CTACTC +CTACTG +CTACCA +CTACCT +CTACCC +CTACCG +CTACGA +CTACGT +CTACGC +CTACGG +CTAGAA +CTAGAT +CTAGAC +CTAGAG +CTAGTA +CTAGTT +CTAGTC +CTAGTG +CTAGCA +CTAGCT +CTAGCC +CTAGCG +CTAGGA +CTAGGT +CTAGGC +CTAGGG +CTTAAA +CTTAAT +CTTAAC +CTTAAG +CTTATA +CTTATT +CTTATC +CTTATG +CTTACA +CTTACT +CTTACC +CTTACG +CTTAGA +CTTAGT +CTTAGC +CTTAGG +CTTTAA +CTTTAT +CTTTAC +CTTTAG +CTTTTA +CTTTTT +CTTTTC +CTTTTG +CTTTCA +CTTTCT +CTTTCC +CTTTCG +CTTTGA +CTTTGT +CTTTGC +CTTTGG +CTTCAA +CTTCAT +CTTCAC +CTTCAG +CTTCTA +CTTCTT +CTTCTC +CTTCTG +CTTCCA +CTTCCT +CTTCCC +CTTCCG +CTTCGA +CTTCGT +CTTCGC +CTTCGG +CTTGAA +CTTGAT +CTTGAC +CTTGAG +CTTGTA +CTTGTT +CTTGTC +CTTGTG +CTTGCA +CTTGCT +CTTGCC +CTTGCG +CTTGGA +CTTGGT +CTTGGC +CTTGGG +CTCAAA +CTCAAT +CTCAAC +CTCAAG +CTCATA +CTCATT +CTCATC +CTCATG +CTCACA +CTCACT +CTCACC +CTCACG +CTCAGA +CTCAGT +CTCAGC +CTCAGG +CTCTAA +CTCTAT +CTCTAC +CTCTAG +CTCTTA +CTCTTT +CTCTTC +CTCTTG +CTCTCA +CTCTCT +CTCTCC +CTCTCG +CTCTGA +CTCTGT +CTCTGC +CTCTGG +CTCCAA +CTCCAT +CTCCAC +CTCCAG +CTCCTA +CTCCTT +CTCCTC +CTCCTG +CTCCCA +CTCCCT +CTCCCC +CTCCCG +CTCCGA +CTCCGT +CTCCGC +CTCCGG +CTCGAA +CTCGAT +CTCGAC +CTCGAG +CTCGTA +CTCGTT +CTCGTC +CTCGTG +CTCGCA +CTCGCT +CTCGCC +CTCGCG +CTCGGA +CTCGGT +CTCGGC +CTCGGG +CTGAAA +CTGAAT +CTGAAC +CTGAAG +CTGATA +CTGATT +CTGATC +CTGATG +CTGACA +CTGACT +CTGACC +CTGACG +CTGAGA +CTGAGT +CTGAGC +CTGAGG +CTGTAA +CTGTAT +CTGTAC +CTGTAG +CTGTTA +CTGTTT +CTGTTC +CTGTTG +CTGTCA +CTGTCT +CTGTCC +CTGTCG +CTGTGA +CTGTGT +CTGTGC +CTGTGG +CTGCAA +CTGCAT +CTGCAC +CTGCAG +CTGCTA +CTGCTT +CTGCTC +CTGCTG +CTGCCA +CTGCCT +CTGCCC +CTGCCG +CTGCGA +CTGCGT +CTGCGC +CTGCGG +CTGGAA +CTGGAT +CTGGAC +CTGGAG +CTGGTA +CTGGTT +CTGGTC +CTGGTG +CTGGCA +CTGGCT +CTGGCC +CTGGCG +CTGGGA +CTGGGT +CTGGGC +CTGGGG +CCAAAA +CCAAAT +CCAAAC +CCAAAG +CCAATA +CCAATT +CCAATC +CCAATG +CCAACA +CCAACT +CCAACC +CCAACG +CCAAGA +CCAAGT +CCAAGC +CCAAGG +CCATAA +CCATAT +CCATAC +CCATAG +CCATTA +CCATTT +CCATTC +CCATTG +CCATCA +CCATCT +CCATCC +CCATCG +CCATGA +CCATGT +CCATGC +CCATGG +CCACAA +CCACAT +CCACAC +CCACAG +CCACTA +CCACTT +CCACTC +CCACTG +CCACCA +CCACCT +CCACCC +CCACCG +CCACGA +CCACGT +CCACGC +CCACGG +CCAGAA +CCAGAT +CCAGAC +CCAGAG +CCAGTA +CCAGTT +CCAGTC +CCAGTG +CCAGCA +CCAGCT +CCAGCC +CCAGCG +CCAGGA +CCAGGT +CCAGGC +CCAGGG +CCTAAA +CCTAAT +CCTAAC +CCTAAG +CCTATA +CCTATT +CCTATC +CCTATG +CCTACA +CCTACT +CCTACC +CCTACG +CCTAGA +CCTAGT +CCTAGC +CCTAGG +CCTTAA +CCTTAT +CCTTAC +CCTTAG +CCTTTA +CCTTTT +CCTTTC +CCTTTG +CCTTCA +CCTTCT +CCTTCC +CCTTCG +CCTTGA +CCTTGT +CCTTGC +CCTTGG +CCTCAA +CCTCAT +CCTCAC +CCTCAG +CCTCTA +CCTCTT +CCTCTC +CCTCTG +CCTCCA +CCTCCT +CCTCCC +CCTCCG +CCTCGA +CCTCGT +CCTCGC +CCTCGG +CCTGAA +CCTGAT +CCTGAC +CCTGAG +CCTGTA +CCTGTT +CCTGTC +CCTGTG +CCTGCA +CCTGCT +CCTGCC +CCTGCG +CCTGGA +CCTGGT +CCTGGC +CCTGGG +CCCAAA +CCCAAT +CCCAAC +CCCAAG +CCCATA +CCCATT +CCCATC +CCCATG +CCCACA +CCCACT +CCCACC +CCCACG +CCCAGA +CCCAGT +CCCAGC +CCCAGG +CCCTAA +CCCTAT +CCCTAC +CCCTAG +CCCTTA +CCCTTT +CCCTTC +CCCTTG +CCCTCA +CCCTCT +CCCTCC +CCCTCG +CCCTGA +CCCTGT +CCCTGC +CCCTGG +CCCCAA +CCCCAT +CCCCAC +CCCCAG +CCCCTA +CCCCTT +CCCCTC +CCCCTG +CCCCCA +CCCCCT +CCCCCC +CCCCCG +CCCCGA +CCCCGT +CCCCGC +CCCCGG +CCCGAA +CCCGAT +CCCGAC +CCCGAG +CCCGTA +CCCGTT +CCCGTC +CCCGTG +CCCGCA +CCCGCT +CCCGCC +CCCGCG +CCCGGA +CCCGGT +CCCGGC +CCCGGG +CCGAAA +CCGAAT +CCGAAC +CCGAAG +CCGATA +CCGATT +CCGATC +CCGATG +CCGACA +CCGACT +CCGACC +CCGACG +CCGAGA +CCGAGT +CCGAGC +CCGAGG +CCGTAA +CCGTAT +CCGTAC +CCGTAG +CCGTTA +CCGTTT +CCGTTC +CCGTTG +CCGTCA +CCGTCT +CCGTCC +CCGTCG +CCGTGA +CCGTGT +CCGTGC +CCGTGG +CCGCAA +CCGCAT +CCGCAC +CCGCAG +CCGCTA +CCGCTT +CCGCTC +CCGCTG +CCGCCA +CCGCCT +CCGCCC +CCGCCG +CCGCGA +CCGCGT +CCGCGC +CCGCGG +CCGGAA +CCGGAT +CCGGAC +CCGGAG +CCGGTA +CCGGTT +CCGGTC +CCGGTG +CCGGCA +CCGGCT +CCGGCC +CCGGCG +CCGGGA +CCGGGT +CCGGGC +CCGGGG +CGAAAA +CGAAAT +CGAAAC +CGAAAG +CGAATA +CGAATT +CGAATC +CGAATG +CGAACA +CGAACT +CGAACC +CGAACG +CGAAGA +CGAAGT +CGAAGC +CGAAGG +CGATAA +CGATAT +CGATAC +CGATAG +CGATTA +CGATTT +CGATTC +CGATTG +CGATCA +CGATCT +CGATCC +CGATCG +CGATGA +CGATGT +CGATGC +CGATGG +CGACAA +CGACAT +CGACAC +CGACAG +CGACTA +CGACTT +CGACTC +CGACTG +CGACCA +CGACCT +CGACCC +CGACCG +CGACGA +CGACGT +CGACGC +CGACGG +CGAGAA +CGAGAT +CGAGAC +CGAGAG +CGAGTA +CGAGTT +CGAGTC +CGAGTG +CGAGCA +CGAGCT +CGAGCC +CGAGCG +CGAGGA +CGAGGT +CGAGGC +CGAGGG +CGTAAA +CGTAAT +CGTAAC +CGTAAG +CGTATA +CGTATT +CGTATC +CGTATG +CGTACA +CGTACT +CGTACC +CGTACG +CGTAGA +CGTAGT +CGTAGC +CGTAGG +CGTTAA +CGTTAT +CGTTAC +CGTTAG +CGTTTA +CGTTTT +CGTTTC +CGTTTG +CGTTCA +CGTTCT +CGTTCC +CGTTCG +CGTTGA +CGTTGT +CGTTGC +CGTTGG +CGTCAA +CGTCAT +CGTCAC +CGTCAG +CGTCTA +CGTCTT +CGTCTC +CGTCTG +CGTCCA +CGTCCT +CGTCCC +CGTCCG +CGTCGA +CGTCGT +CGTCGC +CGTCGG +CGTGAA +CGTGAT +CGTGAC +CGTGAG +CGTGTA +CGTGTT +CGTGTC +CGTGTG +CGTGCA +CGTGCT +CGTGCC +CGTGCG +CGTGGA +CGTGGT +CGTGGC +CGTGGG +CGCAAA +CGCAAT +CGCAAC +CGCAAG +CGCATA +CGCATT +CGCATC +CGCATG +CGCACA +CGCACT +CGCACC +CGCACG +CGCAGA +CGCAGT +CGCAGC +CGCAGG +CGCTAA +CGCTAT +CGCTAC +CGCTAG +CGCTTA +CGCTTT +CGCTTC +CGCTTG +CGCTCA +CGCTCT +CGCTCC +CGCTCG +CGCTGA +CGCTGT +CGCTGC +CGCTGG +CGCCAA +CGCCAT +CGCCAC +CGCCAG +CGCCTA +CGCCTT +CGCCTC +CGCCTG +CGCCCA +CGCCCT +CGCCCC +CGCCCG +CGCCGA +CGCCGT +CGCCGC +CGCCGG +CGCGAA +CGCGAT +CGCGAC +CGCGAG +CGCGTA +CGCGTT +CGCGTC +CGCGTG +CGCGCA +CGCGCT +CGCGCC +CGCGCG +CGCGGA +CGCGGT +CGCGGC +CGCGGG +CGGAAA +CGGAAT +CGGAAC +CGGAAG +CGGATA +CGGATT +CGGATC +CGGATG +CGGACA +CGGACT +CGGACC +CGGACG +CGGAGA +CGGAGT +CGGAGC +CGGAGG +CGGTAA +CGGTAT +CGGTAC +CGGTAG +CGGTTA +CGGTTT +CGGTTC +CGGTTG +CGGTCA +CGGTCT +CGGTCC +CGGTCG +CGGTGA +CGGTGT +CGGTGC +CGGTGG +CGGCAA +CGGCAT +CGGCAC +CGGCAG +CGGCTA +CGGCTT +CGGCTC +CGGCTG +CGGCCA +CGGCCT +CGGCCC +CGGCCG +CGGCGA +CGGCGT +CGGCGC +CGGCGG +CGGGAA +CGGGAT +CGGGAC +CGGGAG +CGGGTA +CGGGTT +CGGGTC +CGGGTG +CGGGCA +CGGGCT +CGGGCC +CGGGCG +CGGGGA +CGGGGT +CGGGGC +CGGGGG +GAAAAA +GAAAAT +GAAAAC +GAAAAG +GAAATA +GAAATT +GAAATC +GAAATG +GAAACA +GAAACT +GAAACC +GAAACG +GAAAGA +GAAAGT +GAAAGC +GAAAGG +GAATAA +GAATAT +GAATAC +GAATAG +GAATTA +GAATTT +GAATTC +GAATTG +GAATCA +GAATCT +GAATCC +GAATCG +GAATGA +GAATGT +GAATGC +GAATGG +GAACAA +GAACAT +GAACAC +GAACAG +GAACTA +GAACTT +GAACTC +GAACTG +GAACCA +GAACCT +GAACCC +GAACCG +GAACGA +GAACGT +GAACGC +GAACGG +GAAGAA +GAAGAT +GAAGAC +GAAGAG +GAAGTA +GAAGTT +GAAGTC +GAAGTG +GAAGCA +GAAGCT +GAAGCC +GAAGCG +GAAGGA +GAAGGT +GAAGGC +GAAGGG +GATAAA +GATAAT +GATAAC +GATAAG +GATATA +GATATT +GATATC +GATATG +GATACA +GATACT +GATACC +GATACG +GATAGA +GATAGT +GATAGC +GATAGG +GATTAA +GATTAT +GATTAC +GATTAG +GATTTA +GATTTT +GATTTC +GATTTG +GATTCA +GATTCT +GATTCC +GATTCG +GATTGA +GATTGT +GATTGC +GATTGG +GATCAA +GATCAT +GATCAC +GATCAG +GATCTA +GATCTT +GATCTC +GATCTG +GATCCA +GATCCT +GATCCC +GATCCG +GATCGA +GATCGT +GATCGC +GATCGG +GATGAA +GATGAT +GATGAC +GATGAG +GATGTA +GATGTT +GATGTC +GATGTG +GATGCA +GATGCT +GATGCC +GATGCG +GATGGA +GATGGT +GATGGC +GATGGG +GACAAA +GACAAT +GACAAC +GACAAG +GACATA +GACATT +GACATC +GACATG +GACACA +GACACT +GACACC +GACACG +GACAGA +GACAGT +GACAGC +GACAGG +GACTAA +GACTAT +GACTAC +GACTAG +GACTTA +GACTTT +GACTTC +GACTTG +GACTCA +GACTCT +GACTCC +GACTCG +GACTGA +GACTGT +GACTGC +GACTGG +GACCAA +GACCAT +GACCAC +GACCAG +GACCTA +GACCTT +GACCTC +GACCTG +GACCCA +GACCCT +GACCCC +GACCCG +GACCGA +GACCGT +GACCGC +GACCGG +GACGAA +GACGAT +GACGAC +GACGAG +GACGTA +GACGTT +GACGTC +GACGTG +GACGCA +GACGCT +GACGCC +GACGCG +GACGGA +GACGGT +GACGGC +GACGGG +GAGAAA +GAGAAT +GAGAAC +GAGAAG +GAGATA +GAGATT +GAGATC +GAGATG +GAGACA +GAGACT +GAGACC +GAGACG +GAGAGA +GAGAGT +GAGAGC +GAGAGG +GAGTAA +GAGTAT +GAGTAC +GAGTAG +GAGTTA +GAGTTT +GAGTTC +GAGTTG +GAGTCA +GAGTCT +GAGTCC +GAGTCG +GAGTGA +GAGTGT +GAGTGC +GAGTGG +GAGCAA +GAGCAT +GAGCAC +GAGCAG +GAGCTA +GAGCTT +GAGCTC +GAGCTG +GAGCCA +GAGCCT +GAGCCC +GAGCCG +GAGCGA +GAGCGT +GAGCGC +GAGCGG +GAGGAA +GAGGAT +GAGGAC +GAGGAG +GAGGTA +GAGGTT +GAGGTC +GAGGTG +GAGGCA +GAGGCT +GAGGCC +GAGGCG +GAGGGA +GAGGGT +GAGGGC +GAGGGG +GTAAAA +GTAAAT +GTAAAC +GTAAAG +GTAATA +GTAATT +GTAATC +GTAATG +GTAACA +GTAACT +GTAACC +GTAACG +GTAAGA +GTAAGT +GTAAGC +GTAAGG +GTATAA +GTATAT +GTATAC +GTATAG +GTATTA +GTATTT +GTATTC +GTATTG +GTATCA +GTATCT +GTATCC +GTATCG +GTATGA +GTATGT +GTATGC +GTATGG +GTACAA +GTACAT +GTACAC +GTACAG +GTACTA +GTACTT +GTACTC +GTACTG +GTACCA +GTACCT +GTACCC +GTACCG +GTACGA +GTACGT +GTACGC +GTACGG +GTAGAA +GTAGAT +GTAGAC +GTAGAG +GTAGTA +GTAGTT +GTAGTC +GTAGTG +GTAGCA +GTAGCT +GTAGCC +GTAGCG +GTAGGA +GTAGGT +GTAGGC +GTAGGG +GTTAAA +GTTAAT +GTTAAC +GTTAAG +GTTATA +GTTATT +GTTATC +GTTATG +GTTACA +GTTACT +GTTACC +GTTACG +GTTAGA +GTTAGT +GTTAGC +GTTAGG +GTTTAA +GTTTAT +GTTTAC +GTTTAG +GTTTTA +GTTTTT +GTTTTC +GTTTTG +GTTTCA +GTTTCT +GTTTCC +GTTTCG +GTTTGA +GTTTGT +GTTTGC +GTTTGG +GTTCAA +GTTCAT +GTTCAC +GTTCAG +GTTCTA +GTTCTT +GTTCTC +GTTCTG +GTTCCA +GTTCCT +GTTCCC +GTTCCG +GTTCGA +GTTCGT +GTTCGC +GTTCGG +GTTGAA +GTTGAT +GTTGAC +GTTGAG +GTTGTA +GTTGTT +GTTGTC +GTTGTG +GTTGCA +GTTGCT +GTTGCC +GTTGCG +GTTGGA +GTTGGT +GTTGGC +GTTGGG +GTCAAA +GTCAAT +GTCAAC +GTCAAG +GTCATA +GTCATT +GTCATC +GTCATG +GTCACA +GTCACT +GTCACC +GTCACG +GTCAGA +GTCAGT +GTCAGC +GTCAGG +GTCTAA +GTCTAT +GTCTAC +GTCTAG +GTCTTA +GTCTTT +GTCTTC +GTCTTG +GTCTCA +GTCTCT +GTCTCC +GTCTCG +GTCTGA +GTCTGT +GTCTGC +GTCTGG +GTCCAA +GTCCAT +GTCCAC +GTCCAG +GTCCTA +GTCCTT +GTCCTC +GTCCTG +GTCCCA +GTCCCT +GTCCCC +GTCCCG +GTCCGA +GTCCGT +GTCCGC +GTCCGG +GTCGAA +GTCGAT +GTCGAC +GTCGAG +GTCGTA +GTCGTT +GTCGTC +GTCGTG +GTCGCA +GTCGCT +GTCGCC +GTCGCG +GTCGGA +GTCGGT +GTCGGC +GTCGGG +GTGAAA +GTGAAT +GTGAAC +GTGAAG +GTGATA +GTGATT +GTGATC +GTGATG +GTGACA +GTGACT +GTGACC +GTGACG +GTGAGA +GTGAGT +GTGAGC +GTGAGG +GTGTAA +GTGTAT +GTGTAC +GTGTAG +GTGTTA +GTGTTT +GTGTTC +GTGTTG +GTGTCA +GTGTCT +GTGTCC +GTGTCG +GTGTGA +GTGTGT +GTGTGC +GTGTGG +GTGCAA +GTGCAT +GTGCAC +GTGCAG +GTGCTA +GTGCTT +GTGCTC +GTGCTG +GTGCCA +GTGCCT +GTGCCC +GTGCCG +GTGCGA +GTGCGT +GTGCGC +GTGCGG +GTGGAA +GTGGAT +GTGGAC +GTGGAG +GTGGTA +GTGGTT +GTGGTC +GTGGTG +GTGGCA +GTGGCT +GTGGCC +GTGGCG +GTGGGA +GTGGGT +GTGGGC +GTGGGG +GCAAAA +GCAAAT +GCAAAC +GCAAAG +GCAATA +GCAATT +GCAATC +GCAATG +GCAACA +GCAACT +GCAACC +GCAACG +GCAAGA +GCAAGT +GCAAGC +GCAAGG +GCATAA +GCATAT +GCATAC +GCATAG +GCATTA +GCATTT +GCATTC +GCATTG +GCATCA +GCATCT +GCATCC +GCATCG +GCATGA +GCATGT +GCATGC +GCATGG +GCACAA +GCACAT +GCACAC +GCACAG +GCACTA +GCACTT +GCACTC +GCACTG +GCACCA +GCACCT +GCACCC +GCACCG +GCACGA +GCACGT +GCACGC +GCACGG +GCAGAA +GCAGAT +GCAGAC +GCAGAG +GCAGTA +GCAGTT +GCAGTC +GCAGTG +GCAGCA +GCAGCT +GCAGCC +GCAGCG +GCAGGA +GCAGGT +GCAGGC +GCAGGG +GCTAAA +GCTAAT +GCTAAC +GCTAAG +GCTATA +GCTATT +GCTATC +GCTATG +GCTACA +GCTACT +GCTACC +GCTACG +GCTAGA +GCTAGT +GCTAGC +GCTAGG +GCTTAA +GCTTAT +GCTTAC +GCTTAG +GCTTTA +GCTTTT +GCTTTC +GCTTTG +GCTTCA +GCTTCT +GCTTCC +GCTTCG +GCTTGA +GCTTGT +GCTTGC +GCTTGG +GCTCAA +GCTCAT +GCTCAC +GCTCAG +GCTCTA +GCTCTT +GCTCTC +GCTCTG +GCTCCA +GCTCCT +GCTCCC +GCTCCG +GCTCGA +GCTCGT +GCTCGC +GCTCGG +GCTGAA +GCTGAT +GCTGAC +GCTGAG +GCTGTA +GCTGTT +GCTGTC +GCTGTG +GCTGCA +GCTGCT +GCTGCC +GCTGCG +GCTGGA +GCTGGT +GCTGGC +GCTGGG +GCCAAA +GCCAAT +GCCAAC +GCCAAG +GCCATA +GCCATT +GCCATC +GCCATG +GCCACA +GCCACT +GCCACC +GCCACG +GCCAGA +GCCAGT +GCCAGC +GCCAGG +GCCTAA +GCCTAT +GCCTAC +GCCTAG +GCCTTA +GCCTTT +GCCTTC +GCCTTG +GCCTCA +GCCTCT +GCCTCC +GCCTCG +GCCTGA +GCCTGT +GCCTGC +GCCTGG +GCCCAA +GCCCAT +GCCCAC +GCCCAG +GCCCTA +GCCCTT +GCCCTC +GCCCTG +GCCCCA +GCCCCT +GCCCCC +GCCCCG +GCCCGA +GCCCGT +GCCCGC +GCCCGG +GCCGAA +GCCGAT +GCCGAC +GCCGAG +GCCGTA +GCCGTT +GCCGTC +GCCGTG +GCCGCA +GCCGCT +GCCGCC +GCCGCG +GCCGGA +GCCGGT +GCCGGC +GCCGGG +GCGAAA +GCGAAT +GCGAAC +GCGAAG +GCGATA +GCGATT +GCGATC +GCGATG +GCGACA +GCGACT +GCGACC +GCGACG +GCGAGA +GCGAGT +GCGAGC +GCGAGG +GCGTAA +GCGTAT +GCGTAC +GCGTAG +GCGTTA +GCGTTT +GCGTTC +GCGTTG +GCGTCA +GCGTCT +GCGTCC +GCGTCG +GCGTGA +GCGTGT +GCGTGC +GCGTGG +GCGCAA +GCGCAT +GCGCAC +GCGCAG +GCGCTA +GCGCTT +GCGCTC +GCGCTG +GCGCCA +GCGCCT +GCGCCC +GCGCCG +GCGCGA +GCGCGT +GCGCGC +GCGCGG +GCGGAA +GCGGAT +GCGGAC +GCGGAG +GCGGTA +GCGGTT +GCGGTC +GCGGTG +GCGGCA +GCGGCT +GCGGCC +GCGGCG +GCGGGA +GCGGGT +GCGGGC +GCGGGG +GGAAAA +GGAAAT +GGAAAC +GGAAAG +GGAATA +GGAATT +GGAATC +GGAATG +GGAACA +GGAACT +GGAACC +GGAACG +GGAAGA +GGAAGT +GGAAGC +GGAAGG +GGATAA +GGATAT +GGATAC +GGATAG +GGATTA +GGATTT +GGATTC +GGATTG +GGATCA +GGATCT +GGATCC +GGATCG +GGATGA +GGATGT +GGATGC +GGATGG +GGACAA +GGACAT +GGACAC +GGACAG +GGACTA +GGACTT +GGACTC +GGACTG +GGACCA +GGACCT +GGACCC +GGACCG +GGACGA +GGACGT +GGACGC +GGACGG +GGAGAA +GGAGAT +GGAGAC +GGAGAG +GGAGTA +GGAGTT +GGAGTC +GGAGTG +GGAGCA +GGAGCT +GGAGCC +GGAGCG +GGAGGA +GGAGGT +GGAGGC +GGAGGG +GGTAAA +GGTAAT +GGTAAC +GGTAAG +GGTATA +GGTATT +GGTATC +GGTATG +GGTACA +GGTACT +GGTACC +GGTACG +GGTAGA +GGTAGT +GGTAGC +GGTAGG +GGTTAA +GGTTAT +GGTTAC +GGTTAG +GGTTTA +GGTTTT +GGTTTC +GGTTTG +GGTTCA +GGTTCT +GGTTCC +GGTTCG +GGTTGA +GGTTGT +GGTTGC +GGTTGG +GGTCAA +GGTCAT +GGTCAC +GGTCAG +GGTCTA +GGTCTT +GGTCTC +GGTCTG +GGTCCA +GGTCCT +GGTCCC +GGTCCG +GGTCGA +GGTCGT +GGTCGC +GGTCGG +GGTGAA +GGTGAT +GGTGAC +GGTGAG +GGTGTA +GGTGTT +GGTGTC +GGTGTG +GGTGCA +GGTGCT +GGTGCC +GGTGCG +GGTGGA +GGTGGT +GGTGGC +GGTGGG +GGCAAA +GGCAAT +GGCAAC +GGCAAG +GGCATA +GGCATT +GGCATC +GGCATG +GGCACA +GGCACT +GGCACC +GGCACG +GGCAGA +GGCAGT +GGCAGC +GGCAGG +GGCTAA +GGCTAT +GGCTAC +GGCTAG +GGCTTA +GGCTTT +GGCTTC +GGCTTG +GGCTCA +GGCTCT +GGCTCC +GGCTCG +GGCTGA +GGCTGT +GGCTGC +GGCTGG +GGCCAA +GGCCAT +GGCCAC +GGCCAG +GGCCTA +GGCCTT +GGCCTC +GGCCTG +GGCCCA +GGCCCT +GGCCCC +GGCCCG +GGCCGA +GGCCGT +GGCCGC +GGCCGG +GGCGAA +GGCGAT +GGCGAC +GGCGAG +GGCGTA +GGCGTT +GGCGTC +GGCGTG +GGCGCA +GGCGCT +GGCGCC +GGCGCG +GGCGGA +GGCGGT +GGCGGC +GGCGGG +GGGAAA +GGGAAT +GGGAAC +GGGAAG +GGGATA +GGGATT +GGGATC +GGGATG +GGGACA +GGGACT +GGGACC +GGGACG +GGGAGA +GGGAGT +GGGAGC +GGGAGG +GGGTAA +GGGTAT +GGGTAC +GGGTAG +GGGTTA +GGGTTT +GGGTTC +GGGTTG +GGGTCA +GGGTCT +GGGTCC +GGGTCG +GGGTGA +GGGTGT +GGGTGC +GGGTGG +GGGCAA +GGGCAT +GGGCAC +GGGCAG +GGGCTA +GGGCTT +GGGCTC +GGGCTG +GGGCCA +GGGCCT +GGGCCC +GGGCCG +GGGCGA +GGGCGT +GGGCGC +GGGCGG +GGGGAA +GGGGAT +GGGGAC +GGGGAG +GGGGTA +GGGGTT +GGGGTC +GGGGTG +GGGGCA +GGGGCT +GGGGCC +GGGGCG +GGGGGA +GGGGGT +GGGGGC +GGGGGG