djsull/aha_classification

README.md

@@ -1,5 +1,5 @@
 ---
-base_model: klue/roberta-small
+base_model: monologg/kobigbird-bert-base
 tags:
 - generated_from_trainer
 metrics:
@@ -15,9 +15,9 @@ should probably proofread and complete it, then remove this comment. -->
 
 # aha_classification
 
-This model is a fine-tuned version of [klue/roberta-small](https://huggingface.co/klue/roberta-small) on an unknown dataset.
+This model is a fine-tuned version of [monologg/kobigbird-bert-base](https://huggingface.co/monologg/kobigbird-bert-base) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.0295
+- Loss: 0.0276
 - F1: 1.0
 - Roc Auc: 1.0
 - Accuracy: 1.0
@@ -45,17 +45,22 @@ The following hyperparameters were used during training:
 - seed: 42
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
-- num_epochs: 5
+- num_epochs: 15
 
 ### Training results
 
 | Training Loss | Epoch | Step | Validation Loss | F1     | Roc Auc | Accuracy |
 |:-------------:|:-----:|:----:|:---------------:|:------:|:-------:|:--------:|
-| No log        | 1.0   | 50   | 0.1202          | 0.9969 | 0.9981  | 0.9929   |
-| No log        | 2.0   | 100  | 0.0560          | 0.9969 | 0.9981  | 0.9929   |
-| No log        | 3.0   | 150  | 0.0381          | 0.9969 | 0.9981  | 0.9929   |
-| No log        | 4.0   | 200  | 0.0295          | 1.0    | 1.0     | 1.0      |
-| No log        | 5.0   | 250  | 0.0279          | 1.0    | 1.0     | 1.0      |
+| No log        | 1.0   | 50   | 0.2740          | 0.9906 | 0.9930  | 0.9857   |
+| No log        | 2.0   | 100  | 0.1222          | 0.9938 | 0.9962  | 0.9857   |
+| No log        | 3.0   | 150  | 0.0663          | 0.9969 | 0.9981  | 0.9929   |
+| No log        | 4.0   | 200  | 0.0423          | 0.9969 | 0.9981  | 0.9929   |
+| No log        | 5.0   | 250  | 0.0276          | 1.0    | 1.0     | 1.0      |
+| No log        | 6.0   | 300  | 0.0208          | 1.0    | 1.0     | 1.0      |
+| No log        | 7.0   | 350  | 0.0165          | 1.0    | 1.0     | 1.0      |
+| No log        | 8.0   | 400  | 0.0135          | 1.0    | 1.0     | 1.0      |
+| No log        | 9.0   | 450  | 0.0114          | 1.0    | 1.0     | 1.0      |
+| 0.1025        | 10.0  | 500  | 0.0101          | 1.0    | 1.0     | 1.0      |
 
 
 ### Framework versions

config.json

@@ -1,14 +1,16 @@
 {
-  "_name_or_path": "klue/roberta-small",
+  "_name_or_path": "monologg/kobigbird-bert-base",
   "architectures": [
-    "RobertaForSequenceClassification"
+    "BigBirdForSequenceClassification"
   ],
   "attention_probs_dropout_prob": 0.1,
-  "bos_token_id": 0,
+  "attention_type": "block_sparse",
+  "block_size": 64,
+  "bos_token_id": 5,
   "classifier_dropout": null,
-  "eos_token_id": 2,
+  "eos_token_id": 6,
   "gradient_checkpointing": false,
-  "hidden_act": "gelu",
+  "hidden_act": "gelu_new",
   "hidden_dropout_prob": 0.1,
   "hidden_size": 768,
   "id2label": {
@@ -23,18 +25,22 @@
     "Service": 1,
     "Token": 2
   },
-  "layer_norm_eps": 1e-05,
-  "max_position_embeddings": 514,
-  "model_type": "roberta",
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 4096,
+  "model_type": "big_bird",
   "num_attention_heads": 12,
-  "num_hidden_layers": 6,
-  "pad_token_id": 1,
+  "num_hidden_layers": 12,
+  "num_random_blocks": 3,
+  "pad_token_id": 0,
   "position_embedding_type": "absolute",
   "problem_type": "multi_label_classification",
+  "rescale_embeddings": false,
+  "sep_token_id": 3,
   "tokenizer_class": "BertTokenizer",
   "torch_dtype": "float32",
   "transformers_version": "4.42.3",
-  "type_vocab_size": 1,
+  "type_vocab_size": 2,
+  "use_bias": true,
   "use_cache": true,
-  "vocab_size": 32000
+  "vocab_size": 32500
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f1cbc87f7b18c502cedd4170272203324299be81cb3b60691d95a85ad056584d
-size 272385276
+oid sha256:207c52dc2f1fc52daae93ae96eb14ef124134f04f63c722e0bd4fad404ad3291
+size 457410764

runs/Sep04_08-20-40_129-146-79-76/events.out.tfevents.1725438041.129-146-79-76.38666.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a7552b0cee9f13412ba2cbddeed579b6d1a15036e53c1a4230b6a83fbb0719d8
+size 7753

runs/Sep04_08-25-35_129-146-79-76/events.out.tfevents.1725438336.129-146-79-76.40588.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:40bb83a52b5a9f6439e78752fb9a81dfb89f3ad1c574bd0afd04beb29293d25f
+size 10065

runs/Sep04_08-25-35_129-146-79-76/events.out.tfevents.1725440642.129-146-79-76.40588.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3ebff5958056011d0130353e299f4fc3a603805186d1985490172e8ba5d62dac
+size 508

special_tokens_map.json

@@ -1,6 +1,6 @@
 {
   "bos_token": {
-    "content": "[CLS]",
+    "content": "<s>",
     "lstrip": false,
     "normalized": false,
     "rstrip": false,
@@ -14,7 +14,7 @@
     "single_word": false
   },
   "eos_token": {
-    "content": "[SEP]",
+    "content": "</s>",
     "lstrip": false,
     "normalized": false,
     "rstrip": false,

tokenizer_config.json

@@ -1,7 +1,7 @@
 {
   "added_tokens_decoder": {
     "0": {
-      "content": "[CLS]",
+      "content": "[PAD]",
       "lstrip": false,
       "normalized": false,
       "rstrip": false,
@@ -9,7 +9,7 @@
       "special": true
     },
     "1": {
-      "content": "[PAD]",
+      "content": "[UNK]",
       "lstrip": false,
       "normalized": false,
       "rstrip": false,
@@ -17,7 +17,7 @@
       "special": true
     },
     "2": {
-      "content": "[SEP]",
+      "content": "[CLS]",
       "lstrip": false,
       "normalized": false,
       "rstrip": false,
@@ -25,7 +25,7 @@
       "special": true
     },
     "3": {
-      "content": "[UNK]",
+      "content": "[SEP]",
       "lstrip": false,
       "normalized": false,
       "rstrip": false,
@@ -39,16 +39,32 @@
       "rstrip": false,
       "single_word": false,
       "special": true
+    },
+    "5": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "6": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
     }
   },
-  "bos_token": "[CLS]",
+  "bos_token": "<s>",
   "clean_up_tokenization_spaces": true,
   "cls_token": "[CLS]",
   "do_basic_tokenize": true,
   "do_lower_case": false,
-  "eos_token": "[SEP]",
+  "eos_token": "</s>",
   "mask_token": "[MASK]",
-  "model_max_length": 512,
+  "model_max_length": 4096,
   "never_split": null,
   "pad_token": "[PAD]",
   "sep_token": "[SEP]",

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:07f22ec80f5c53070fab174cfe020f957d287267c08715a5a64dd2d369202243
+oid sha256:66a111869a1d1fadae090c0ce031ee7add201d05bad5866735b012ba523213ce
 size 5112