End of training

README.md

@@ -17,11 +17,11 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [aubmindlab/aragpt2-base](https://huggingface.co/aubmindlab/aragpt2-base) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.0538
-- Bleu: 0.0653
-- Rouge1: 0.4167
-- Rouge2: 0.1755
-- Rougel: 0.4150
+- Loss: 0.0499
+- Bleu: 0.0393
+- Rouge1: 0.3590
+- Rouge2: 0.1242
+- Rougel: 0.3572
 
 ## Model description
 

all_results.json

@@ -0,0 +1,19 @@
+{
+    "epoch": 10.0,
+    "eval_bleu": 0.03933306151343601,
+    "eval_loss": 0.04986047372221947,
+    "eval_rouge1": 0.35895878709405493,
+    "eval_rouge2": 0.124247769713655,
+    "eval_rougeL": 0.3572225001900925,
+    "eval_runtime": 36.9239,
+    "eval_samples": 1672,
+    "eval_samples_per_second": 45.282,
+    "eval_steps_per_second": 5.66,
+    "perplexity": 1.0511244266653859,
+    "total_flos": 3.49347446784e+16,
+    "train_loss": 0.15054923290270938,
+    "train_runtime": 5668.5001,
+    "train_samples": 6685,
+    "train_samples_per_second": 23.586,
+    "train_steps_per_second": 2.95
+}

egy_training_log.txt

@@ -296,3 +296,5 @@ INFO:root:Epoch 9.0: Train Loss = 0.0334, Eval Loss = 0.050896577537059784
 INFO:absl:Using default tokenizer.
 INFO:root:Epoch 10.0: Train Loss = 0.0307, Eval Loss = 0.05167483910918236
 INFO:absl:Using default tokenizer.
+INFO:__main__:*** Evaluate ***
+INFO:absl:Using default tokenizer.

eval_results.json

@@ -0,0 +1,13 @@
+{
+    "epoch": 10.0,
+    "eval_bleu": 0.03933306151343601,
+    "eval_loss": 0.04986047372221947,
+    "eval_rouge1": 0.35895878709405493,
+    "eval_rouge2": 0.124247769713655,
+    "eval_rougeL": 0.3572225001900925,
+    "eval_runtime": 36.9239,
+    "eval_samples": 1672,
+    "eval_samples_per_second": 45.282,
+    "eval_steps_per_second": 5.66,
+    "perplexity": 1.0511244266653859
+}

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 10.0,
+    "total_flos": 3.49347446784e+16,
+    "train_loss": 0.15054923290270938,
+    "train_runtime": 5668.5001,
+    "train_samples": 6685,
+    "train_samples_per_second": 23.586,
+    "train_steps_per_second": 2.95
+}

trainer_state.json

@@ -0,0 +1,241 @@
+{
+  "best_metric": 0.04986047372221947,
+  "best_model_checkpoint": "/home/iais_marenpielka/Bouthaina/res_nw_gulf_aragpt2-base/checkpoint-4180",
+  "epoch": 10.0,
+  "eval_steps": 500,
+  "global_step": 8360,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2068207710981369,
+      "learning_rate": 4.896424167694204e-05,
+      "loss": 1.1245,
+      "step": 836
+    },
+    {
+      "epoch": 1.0,
+      "eval_bleu": 0.00522127887673706,
+      "eval_loss": 0.060549940913915634,
+      "eval_rouge1": 0.16261381712821865,
+      "eval_rouge2": 0.016191591492833397,
+      "eval_rougeL": 0.16017425267387253,
+      "eval_runtime": 159.9696,
+      "eval_samples_per_second": 10.452,
+      "eval_steps_per_second": 1.306,
+      "step": 836
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.19557742774486542,
+      "learning_rate": 4.638717632552405e-05,
+      "loss": 0.0632,
+      "step": 1672
+    },
+    {
+      "epoch": 2.0,
+      "eval_bleu": 0.011943726030788412,
+      "eval_loss": 0.05519399791955948,
+      "eval_rouge1": 0.2541122949630313,
+      "eval_rouge2": 0.051699799633347324,
+      "eval_rougeL": 0.2510765914158727,
+      "eval_runtime": 221.3727,
+      "eval_samples_per_second": 7.553,
+      "eval_steps_per_second": 0.944,
+      "step": 1672
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.14342406392097473,
+      "learning_rate": 4.3810110974106046e-05,
+      "loss": 0.055,
+      "step": 2508
+    },
+    {
+      "epoch": 3.0,
+      "eval_bleu": 0.02252740132414782,
+      "eval_loss": 0.05222811922430992,
+      "eval_rouge1": 0.30471930149085674,
+      "eval_rouge2": 0.08185853906387802,
+      "eval_rougeL": 0.30154781860876523,
+      "eval_runtime": 36.8724,
+      "eval_samples_per_second": 45.346,
+      "eval_steps_per_second": 5.668,
+      "step": 2508
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.16051243245601654,
+      "learning_rate": 4.1233045622688044e-05,
+      "loss": 0.0492,
+      "step": 3344
+    },
+    {
+      "epoch": 4.0,
+      "eval_bleu": 0.032582447132208606,
+      "eval_loss": 0.051017943769693375,
+      "eval_rouge1": 0.3346924293421615,
+      "eval_rouge2": 0.10192760972807041,
+      "eval_rougeL": 0.33180504017979595,
+      "eval_runtime": 159.805,
+      "eval_samples_per_second": 10.463,
+      "eval_steps_per_second": 1.308,
+      "step": 3344
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.15151090919971466,
+      "learning_rate": 3.8655980271270036e-05,
+      "loss": 0.0444,
+      "step": 4180
+    },
+    {
+      "epoch": 5.0,
+      "eval_bleu": 0.03933306151343601,
+      "eval_loss": 0.04986047372221947,
+      "eval_rouge1": 0.35895878709405493,
+      "eval_rouge2": 0.124247769713655,
+      "eval_rougeL": 0.3572225001900925,
+      "eval_runtime": 159.8295,
+      "eval_samples_per_second": 10.461,
+      "eval_steps_per_second": 1.308,
+      "step": 4180
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 0.11806467920541763,
+      "learning_rate": 3.6078914919852034e-05,
+      "loss": 0.0402,
+      "step": 5016
+    },
+    {
+      "epoch": 6.0,
+      "eval_bleu": 0.04622708241683158,
+      "eval_loss": 0.04995572566986084,
+      "eval_rouge1": 0.3810393862934106,
+      "eval_rouge2": 0.13792791296322293,
+      "eval_rougeL": 0.3788340225664032,
+      "eval_runtime": 159.7942,
+      "eval_samples_per_second": 10.463,
+      "eval_steps_per_second": 1.308,
+      "step": 5016
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 0.15260820090770721,
+      "learning_rate": 3.350184956843403e-05,
+      "loss": 0.0366,
+      "step": 5852
+    },
+    {
+      "epoch": 7.0,
+      "eval_bleu": 0.04988840099834184,
+      "eval_loss": 0.05027288198471069,
+      "eval_rouge1": 0.3961215133733164,
+      "eval_rouge2": 0.15246703105477005,
+      "eval_rougeL": 0.3937993288436519,
+      "eval_runtime": 36.8733,
+      "eval_samples_per_second": 45.344,
+      "eval_steps_per_second": 5.668,
+      "step": 5852
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 0.14294394850730896,
+      "learning_rate": 3.092478421701603e-05,
+      "loss": 0.0334,
+      "step": 6688
+    },
+    {
+      "epoch": 8.0,
+      "eval_bleu": 0.05614309403616892,
+      "eval_loss": 0.050896577537059784,
+      "eval_rouge1": 0.407117249999222,
+      "eval_rouge2": 0.15984144200856937,
+      "eval_rougeL": 0.40517229812025873,
+      "eval_runtime": 37.0371,
+      "eval_samples_per_second": 45.144,
+      "eval_steps_per_second": 5.643,
+      "step": 6688
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 0.20682789385318756,
+      "learning_rate": 2.8347718865598028e-05,
+      "loss": 0.0307,
+      "step": 7524
+    },
+    {
+      "epoch": 9.0,
+      "eval_bleu": 0.0606895848585085,
+      "eval_loss": 0.05167483910918236,
+      "eval_rouge1": 0.41011122575047865,
+      "eval_rouge2": 0.1731982844904956,
+      "eval_rougeL": 0.40855415076231816,
+      "eval_runtime": 159.7295,
+      "eval_samples_per_second": 10.468,
+      "eval_steps_per_second": 1.308,
+      "step": 7524
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 0.13990797102451324,
+      "learning_rate": 2.5770653514180026e-05,
+      "loss": 0.0283,
+      "step": 8360
+    },
+    {
+      "epoch": 10.0,
+      "eval_bleu": 0.06531719425427113,
+      "eval_loss": 0.05382031202316284,
+      "eval_rouge1": 0.41672477670989894,
+      "eval_rouge2": 0.17546430438593302,
+      "eval_rougeL": 0.4150315774802549,
+      "eval_runtime": 159.9272,
+      "eval_samples_per_second": 10.455,
+      "eval_steps_per_second": 1.307,
+      "step": 8360
+    },
+    {
+      "epoch": 10.0,
+      "step": 8360,
+      "total_flos": 3.49347446784e+16,
+      "train_loss": 0.15054923290270938,
+      "train_runtime": 5668.5001,
+      "train_samples_per_second": 23.586,
+      "train_steps_per_second": 2.95
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 16720,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 20,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "EarlyStoppingCallback": {
+      "args": {
+        "early_stopping_patience": 5,
+        "early_stopping_threshold": 0.0
+      },
+      "attributes": {
+        "early_stopping_patience_counter": 0
+      }
+    },
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.49347446784e+16,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}