End of training

README.md

@@ -3,9 +3,10 @@ library_name: transformers
 license: apache-2.0
 base_model: PlanTL-GOB-ES/bsc-bio-ehr-es
 tags:
+- token-classification
 - generated_from_trainer
 datasets:
-- cantemist-fasttext-75-ner
+- Rodrigo1771/cantemist-fasttext-75-ner
 metrics:
 - precision
 - recall
@@ -18,24 +19,24 @@ model-index:
       name: Token Classification
       type: token-classification
     dataset:
-      name: cantemist-fasttext-75-ner
-      type: cantemist-fasttext-75-ner
+      name: Rodrigo1771/cantemist-fasttext-75-ner
+      type: Rodrigo1771/cantemist-fasttext-75-ner
       config: CantemistNer
       split: validation
       args: CantemistNer
     metrics:
     - name: Precision
       type: precision
-      value: 0.8386477141759509
+      value: 0.8462436745815493
     - name: Recall
       type: recall
-      value: 0.85978731784167
+      value: 0.8562426152028357
     - name: F1
       type: f1
-      value: 0.8490859587709063
+      value: 0.8512137823022711
     - name: Accuracy
       type: accuracy
-      value: 0.9917694763178329
+      value: 0.991867253328732
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -43,13 +44,13 @@ should probably proofread and complete it, then remove this comment. -->
 
 # output
 
-This model is a fine-tuned version of [PlanTL-GOB-ES/bsc-bio-ehr-es](https://huggingface.co/PlanTL-GOB-ES/bsc-bio-ehr-es) on the cantemist-fasttext-75-ner dataset.
+This model is a fine-tuned version of [PlanTL-GOB-ES/bsc-bio-ehr-es](https://huggingface.co/PlanTL-GOB-ES/bsc-bio-ehr-es) on the Rodrigo1771/cantemist-fasttext-75-ner dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.0520
-- Precision: 0.8386
-- Recall: 0.8598
-- F1: 0.8491
-- Accuracy: 0.9918
+- Loss: 0.0478
+- Precision: 0.8462
+- Recall: 0.8562
+- F1: 0.8512
+- Accuracy: 0.9919
 
 ## Model description
 

all_results.json

@@ -1,26 +1,26 @@
 {
-    "epoch": 10.0,
-    "eval_accuracy": 0.9991418018042759,
-    "eval_f1": 0.9521203830369357,
-    "eval_loss": 0.004440780263394117,
-    "eval_precision": 0.9447963800904977,
-    "eval_recall": 0.9595588235294118,
-    "eval_runtime": 14.1123,
-    "eval_samples": 6810,
-    "eval_samples_per_second": 482.558,
-    "eval_steps_per_second": 60.373,
-    "predict_accuracy": 0.9987636326290384,
-    "predict_f1": 0.9231204070096098,
-    "predict_loss": 0.006668921560049057,
-    "predict_precision": 0.8967600219659527,
-    "predict_recall": 0.9510774606872452,
-    "predict_runtime": 28.5821,
-    "predict_samples_per_second": 511.3,
-    "predict_steps_per_second": 63.921,
-    "total_flos": 1.5694885147146138e+16,
-    "train_loss": 0.002533856062142209,
-    "train_runtime": 1422.7831,
-    "train_samples": 31229,
-    "train_samples_per_second": 219.492,
-    "train_steps_per_second": 3.43
+    "epoch": 9.991889699918897,
+    "eval_accuracy": 0.991867253328732,
+    "eval_f1": 0.8512137823022711,
+    "eval_loss": 0.047813717275857925,
+    "eval_precision": 0.8462436745815493,
+    "eval_recall": 0.8562426152028357,
+    "eval_runtime": 15.6141,
+    "eval_samples": 7354,
+    "eval_samples_per_second": 470.984,
+    "eval_steps_per_second": 58.921,
+    "predict_accuracy": 0.9922143301576527,
+    "predict_f1": 0.8534459275401798,
+    "predict_loss": 0.044904835522174835,
+    "predict_precision": 0.8332446808510638,
+    "predict_recall": 0.8746510329424902,
+    "predict_runtime": 23.1419,
+    "predict_samples_per_second": 468.329,
+    "predict_steps_per_second": 58.552,
+    "total_flos": 1.919784837910831e+16,
+    "train_loss": 0.008344567238123386,
+    "train_runtime": 1725.5138,
+    "train_samples": 39426,
+    "train_samples_per_second": 228.488,
+    "train_steps_per_second": 3.57
 }

eval_results.json

@@ -1,12 +1,12 @@
 {
-    "epoch": 10.0,
-    "eval_accuracy": 0.9991418018042759,
-    "eval_f1": 0.9521203830369357,
-    "eval_loss": 0.004440780263394117,
-    "eval_precision": 0.9447963800904977,
-    "eval_recall": 0.9595588235294118,
-    "eval_runtime": 14.1123,
-    "eval_samples": 6810,
-    "eval_samples_per_second": 482.558,
-    "eval_steps_per_second": 60.373
+    "epoch": 9.991889699918897,
+    "eval_accuracy": 0.991867253328732,
+    "eval_f1": 0.8512137823022711,
+    "eval_loss": 0.047813717275857925,
+    "eval_precision": 0.8462436745815493,
+    "eval_recall": 0.8562426152028357,
+    "eval_runtime": 15.6141,
+    "eval_samples": 7354,
+    "eval_samples_per_second": 470.984,
+    "eval_steps_per_second": 58.921
 }

predict_results.json

@@ -1,10 +1,10 @@
 {
-    "predict_accuracy": 0.9987636326290384,
-    "predict_f1": 0.9231204070096098,
-    "predict_loss": 0.006668921560049057,
-    "predict_precision": 0.8967600219659527,
-    "predict_recall": 0.9510774606872452,
-    "predict_runtime": 28.5821,
-    "predict_samples_per_second": 511.3,
-    "predict_steps_per_second": 63.921
+    "predict_accuracy": 0.9922143301576527,
+    "predict_f1": 0.8534459275401798,
+    "predict_loss": 0.044904835522174835,
+    "predict_precision": 0.8332446808510638,
+    "predict_recall": 0.8746510329424902,
+    "predict_runtime": 23.1419,
+    "predict_samples_per_second": 468.329,
+    "predict_steps_per_second": 58.552
 }

tb/events.out.tfevents.1725917628.0ada7e7d1d89.20997.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:58821efbe668c1742a3f7de5e322c08aa97cc9517e348616c2c93cc6d02f1f89
+size 560

train.log

@@ -1558,3 +1558,51 @@ Training completed. Do not forget to share your model on huggingface.co/models =
 {'eval_loss': 0.05196274444460869, 'eval_precision': 0.8386477141759509, 'eval_recall': 0.85978731784167, 'eval_f1': 0.8490859587709063, 'eval_accuracy': 0.9917694763178329, 'eval_runtime': 15.9471, 'eval_samples_per_second': 461.149, 'eval_steps_per_second': 57.691, 'epoch': 9.99}
 {'train_runtime': 1725.5138, 'train_samples_per_second': 228.488, 'train_steps_per_second': 3.57, 'train_loss': 0.008344567238123386, 'epoch': 9.99}
 
+***** train metrics *****
+  epoch                    =     9.9919
+  total_flos               = 17879389GF
+  train_loss               =     0.0083
+  train_runtime            = 0:28:45.51
+  train_samples            =      39426
+  train_samples_per_second =    228.488
+  train_steps_per_second   =       3.57
+09/09/2024 21:33:32 - INFO - __main__ -   *** Evaluate ***
+[INFO|trainer.py:811] 2024-09-09 21:33:32,451 >> The following columns in the evaluation set don't have a corresponding argument in `RobertaForTokenClassification.forward` and have been ignored: ner_tags, tokens, id. If ner_tags, tokens, id are not expected by `RobertaForTokenClassification.forward`,  you can safely ignore this message.
+[INFO|trainer.py:3819] 2024-09-09 21:33:32,453 >> 
+***** Running Evaluation *****
+[INFO|trainer.py:3821] 2024-09-09 21:33:32,453 >>   Num examples = 7354
+[INFO|trainer.py:3824] 2024-09-09 21:33:32,453 >>   Batch size = 8
+
  0%|          | 0/920 [00:00<?, ?it/s]
  1%|          | 7/920 [00:00<00:14, 63.25it/s]
  2%|▏         | 14/920 [00:00<00:13, 64.73it/s]
  2%|▏         | 21/920 [00:00<00:13, 65.79it/s]
  3%|▎         | 30/920 [00:00<00:12, 72.97it/s]
  4%|▍         | 38/920 [00:00<00:11, 75.39it/s]
  5%|▌         | 47/920 [00:00<00:11, 78.03it/s]
  6%|▌         | 55/920 [00:00<00:11, 76.04it/s]
  7%|▋         | 63/920 [00:00<00:11, 75.98it/s]
  8%|▊         | 72/920 [00:00<00:10, 78.05it/s]
  9%|▊         | 80/920 [00:01<00:10, 77.73it/s]
 10%|▉         | 88/920 [00:01<00:10, 77.60it/s]
 10%|█         | 96/920 [00:01<00:10, 78.07it/s]
 11%|█▏        | 104/920 [00:01<00:10, 77.39it/s]
 12%|█▏        | 113/920 [00:01<00:10, 79.64it/s]
 13%|█▎        | 121/920 [00:01<00:10, 79.73it/s]
 14%|█▍        | 130/920 [00:01<00:09, 80.76it/s]
 15%|█▌        | 139/920 [00:01<00:09, 80.99it/s]
 16%|█▌        | 148/920 [00:01<00:09, 81.29it/s]
 17%|█▋        | 157/920 [00:02<00:09, 80.14it/s]
 18%|█▊        | 166/920 [00:02<00:09, 80.58it/s]
 19%|█▉        | 175/920 [00:02<00:09, 81.67it/s]
 20%|██        | 184/920 [00:02<00:08, 82.69it/s]
 21%|██        | 193/920 [00:02<00:09, 80.76it/s]
 22%|██▏       | 202/920 [00:02<00:08, 80.40it/s]
 23%|██▎       | 211/920 [00:02<00:09, 77.75it/s]
 24%|██▍       | 220/920 [00:02<00:08, 79.16it/s]
 25%|██▍       | 229/920 [00:02<00:08, 80.07it/s]
 26%|██▌       | 238/920 [00:03<00:08, 81.72it/s]
 27%|██▋       | 247/920 [00:03<00:08, 80.62it/s]
 28%|██▊       | 256/920 [00:03<00:08, 79.82it/s]
 29%|██▊       | 264/920 [00:03<00:08, 77.42it/s]
 30%|██▉       | 273/920 [00:03<00:08, 79.39it/s]
 31%|███       | 282/920 [00:03<00:08, 79.60it/s]
 32%|███▏      | 290/920 [00:03<00:07, 79.25it/s]
 32%|███▎      | 299/920 [00:03<00:07, 81.34it/s]
 33%|███▎      | 308/920 [00:03<00:07, 81.33it/s]
 34%|███▍      | 317/920 [00:04<00:07, 81.89it/s]
 35%|███▌      | 326/920 [00:04<00:07, 78.56it/s]
 36%|███▋      | 334/920 [00:04<00:07, 77.22it/s]
 37%|███▋      | 342/920 [00:04<00:07, 77.09it/s]
 38%|███▊      | 351/920 [00:04<00:07, 78.32it/s]
 39%|███▉      | 359/920 [00:04<00:07, 78.70it/s]
 40%|████      | 368/920 [00:04<00:06, 79.87it/s]
 41%|████      | 377/920 [00:04<00:06, 81.10it/s]
 42%|████▏     | 386/920 [00:04<00:06, 82.48it/s]
 43%|████▎     | 395/920 [00:05<00:06, 78.56it/s]
 44%|████▍     | 403/920 [00:05<00:06, 78.13it/s]
 45%|████▍     | 411/920 [00:05<00:06, 76.71it/s]
 46%|████▌     | 419/920 [00:05<00:06, 77.00it/s]
 47%|████▋     | 428/920 [00:05<00:06, 77.73it/s]
 48%|████▊     | 437/920 [00:05<00:06, 78.68it/s]
 48%|████▊     | 446/920 [00:05<00:05, 80.64it/s]
 49%|████▉     | 455/920 [00:05<00:05, 81.49it/s]
 50%|█████     | 464/920 [00:05<00:05, 81.37it/s]
 51%|█████▏    | 473/920 [00:05<00:05, 80.46it/s]
 52%|█████▏    | 482/920 [00:06<00:05, 80.35it/s]
 53%|█████▎    | 491/920 [00:06<00:05, 80.43it/s]
 54%|█████▍    | 500/920 [00:06<00:05, 80.09it/s]
 55%|█████▌    | 509/920 [00:06<00:05, 80.83it/s]
 56%|█████▋    | 518/920 [00:06<00:05, 80.38it/s]
 57%|█████▋    | 527/920 [00:06<00:04, 80.95it/s]
 58%|█████▊    | 536/920 [00:06<00:04, 81.45it/s]
 59%|█████▉    | 545/920 [00:06<00:04, 79.68it/s]
 60%|██████    | 553/920 [00:06<00:04, 79.26it/s]
 61%|██████    | 561/920 [00:07<00:04, 77.42it/s]
 62%|██████▏   | 570/920 [00:07<00:04, 78.47it/s]
 63%|██████▎   | 578/920 [00:07<00:04, 78.65it/s]
 64%|██████▎   | 586/920 [00:07<00:04, 78.75it/s]
 65%|██████▍   | 595/920 [00:07<00:04, 80.35it/s]
 66%|██████▌   | 604/920 [00:07<00:03, 81.75it/s]
 67%|██████▋   | 613/920 [00:07<00:03, 81.79it/s]
 68%|██████▊   | 622/920 [00:07<00:03, 80.92it/s]
 69%|██████▊   | 631/920 [00:07<00:03, 79.11it/s]
 69%|██████▉   | 639/920 [00:08<00:03, 77.37it/s]
 70%|███████   | 648/920 [00:08<00:03, 79.13it/s]
 71%|███████▏  | 656/920 [00:08<00:03, 77.86it/s]
 72%|███████▏  | 665/920 [00:08<00:03, 79.32it/s]
 73%|███████▎  | 674/920 [00:08<00:03, 80.68it/s]
 74%|███████▍  | 683/920 [00:08<00:02, 82.49it/s]
 75%|███████▌  | 692/920 [00:08<00:02, 81.87it/s]
 76%|███████▌  | 701/920 [00:08<00:02, 81.59it/s]
 77%|███████▋  | 710/920 [00:08<00:02, 81.10it/s]
 78%|███████▊  | 719/920 [00:09<00:02, 81.09it/s]
 79%|███████▉  | 728/920 [00:09<00:02, 80.87it/s]
 80%|████████  | 737/920 [00:09<00:02, 80.55it/s]
 81%|████████  | 746/920 [00:09<00:02, 77.64it/s]
 82%|████████▏ | 755/920 [00:09<00:02, 79.43it/s]
 83%|████████▎ | 763/920 [00:09<00:01, 78.86it/s]
 84%|████████▍ | 772/920 [00:09<00:01, 80.34it/s]
 85%|████████▍ | 781/920 [00:09<00:01, 80.01it/s]
 86%|████████▌ | 790/920 [00:09<00:01, 80.94it/s]
 87%|████████▋ | 799/920 [00:10<00:01, 80.77it/s]
 88%|████████▊ | 808/920 [00:10<00:01, 77.49it/s]
 89%|████████▊ | 816/920 [00:10<00:01, 75.72it/s]
 90%|████████▉ | 825/920 [00:10<00:01, 77.34it/s]
 91%|█████████ | 834/920 [00:10<00:01, 79.66it/s]
 92%|█████████▏| 843/920 [00:10<00:00, 81.28it/s]
 93%|█████████▎| 852/920 [00:10<00:00, 78.51it/s]
 93%|█████████▎| 860/920 [00:10<00:00, 77.61it/s]
 94%|█████████▍| 869/920 [00:10<00:00, 79.52it/s]
 95%|█████████▌| 878/920 [00:11<00:00, 80.98it/s]
 96%|█████████▋| 887/920 [00:11<00:00, 80.39it/s]
 97%|█████████▋| 896/920 [00:11<00:00, 75.74it/s]
 98%|█████████▊| 905/920 [00:11<00:00, 77.37it/s]
 99%|█████████▉| 914/920 [00:11<00:00, 79.55it/s]
+***** eval metrics *****
+  epoch                   =     9.9919
+  eval_accuracy           =     0.9919
+  eval_f1                 =     0.8512
+  eval_loss               =     0.0478
+  eval_precision          =     0.8462
+  eval_recall             =     0.8562
+  eval_runtime            = 0:00:15.61
+  eval_samples            =       7354
+  eval_samples_per_second =    470.984
+  eval_steps_per_second   =     58.921
+09/09/2024 21:33:48 - INFO - __main__ -   *** Predict ***
+[INFO|trainer.py:811] 2024-09-09 21:33:48,070 >> The following columns in the test set don't have a corresponding argument in `RobertaForTokenClassification.forward` and have been ignored: ner_tags, tokens, id. If ner_tags, tokens, id are not expected by `RobertaForTokenClassification.forward`,  you can safely ignore this message.
+[INFO|trainer.py:3819] 2024-09-09 21:33:48,072 >> 
+***** Running Prediction *****
+[INFO|trainer.py:3821] 2024-09-09 21:33:48,072 >>   Num examples = 10838
+[INFO|trainer.py:3824] 2024-09-09 21:33:48,073 >>   Batch size = 8
+
  0%|          | 0/1355 [00:00<?, ?it/s]
  1%|          | 9/1355 [00:00<00:15, 88.06it/s]
  1%|▏         | 18/1355 [00:00<00:16, 80.49it/s]
  2%|▏         | 27/1355 [00:00<00:16, 82.46it/s]
  3%|▎         | 36/1355 [00:00<00:16, 80.74it/s]
  3%|▎         | 45/1355 [00:00<00:16, 79.05it/s]
  4%|▍         | 53/1355 [00:00<00:17, 75.89it/s]
  5%|▍         | 61/1355 [00:00<00:16, 77.07it/s]
  5%|▌         | 70/1355 [00:00<00:16, 79.26it/s]
  6%|▌         | 79/1355 [00:00<00:15, 80.82it/s]
  6%|▋         | 88/1355 [00:01<00:15, 81.87it/s]
  7%|▋         | 97/1355 [00:01<00:15, 81.46it/s]
  8%|▊         | 106/1355 [00:01<00:15, 81.06it/s]
  8%|▊         | 115/1355 [00:01<00:15, 78.92it/s]
  9%|▉         | 123/1355 [00:01<00:15, 78.64it/s]
 10%|▉         | 132/1355 [00:01<00:15, 79.64it/s]
 10%|█         | 140/1355 [00:01<00:15, 78.63it/s]
 11%|█         | 148/1355 [00:01<00:15, 78.24it/s]
 12%|█▏        | 156/1355 [00:01<00:16, 74.61it/s]
 12%|█▏        | 164/1355 [00:02<00:16, 73.49it/s]
 13%|█▎        | 173/1355 [00:02<00:15, 76.91it/s]
 13%|█▎        | 182/1355 [00:02<00:14, 78.49it/s]
 14%|█▍        | 191/1355 [00:02<00:14, 80.57it/s]
 15%|█▍        | 200/1355 [00:02<00:14, 80.78it/s]
 15%|█▌        | 209/1355 [00:02<00:14, 81.00it/s]
 16%|█▌        | 218/1355 [00:02<00:14, 79.75it/s]
 17%|█▋        | 227/1355 [00:02<00:14, 80.27it/s]
 17%|█▋        | 236/1355 [00:02<00:13, 81.35it/s]
 18%|█▊        | 245/1355 [00:03<00:13, 80.49it/s]
 19%|█▊        | 254/1355 [00:03<00:13, 80.12it/s]
 19%|█▉        | 263/1355 [00:03<00:13, 81.47it/s]
 20%|██        | 272/1355 [00:03<00:13, 82.34it/s]
 21%|██        | 281/1355 [00:03<00:13, 77.28it/s]
 21%|██▏       | 289/1355 [00:03<00:13, 77.10it/s]
 22%|██▏       | 298/1355 [00:03<00:13, 78.39it/s]
 23%|██▎       | 307/1355 [00:03<00:13, 79.37it/s]
 23%|██▎       | 316/1355 [00:03<00:12, 80.62it/s]
 24%|██▍       | 325/1355 [00:04<00:12, 81.57it/s]
 25%|██▍       | 334/1355 [00:04<00:12, 82.92it/s]
 25%|██▌       | 343/1355 [00:04<00:12, 83.78it/s]
 26%|██▌       | 352/1355 [00:04<00:11, 84.39it/s]
 27%|██▋       | 361/1355 [00:04<00:11, 84.87it/s]
 27%|██▋       | 370/1355 [00:04<00:11, 84.50it/s]
 28%|██▊       | 379/1355 [00:04<00:11, 84.40it/s]
 29%|██▊       | 388/1355 [00:04<00:11, 84.75it/s]
 29%|██▉       | 397/1355 [00:04<00:11, 84.48it/s]
 30%|██▉       | 406/1355 [00:05<00:11, 84.58it/s]
 31%|███       | 415/1355 [00:05<00:11, 83.91it/s]
 31%|███▏      | 424/1355 [00:05<00:11, 83.95it/s]
 32%|███▏      | 433/1355 [00:05<00:10, 84.14it/s]
 33%|███▎      | 442/1355 [00:05<00:10, 83.93it/s]
 33%|███▎      | 451/1355 [00:05<00:10, 83.32it/s]
 34%|███▍      | 460/1355 [00:05<00:10, 81.79it/s]
 35%|███▍      | 469/1355 [00:05<00:10, 82.56it/s]
 35%|███▌      | 478/1355 [00:05<00:10, 82.69it/s]
 36%|███▌      | 487/1355 [00:06<00:10, 82.26it/s]
 37%|███▋      | 496/1355 [00:06<00:10, 82.65it/s]
 37%|███▋      | 505/1355 [00:06<00:10, 80.55it/s]
 38%|███▊      | 514/1355 [00:06<00:10, 78.46it/s]
 39%|███▊      | 522/1355 [00:06<00:11, 74.90it/s]
 39%|███▉      | 531/1355 [00:06<00:10, 76.73it/s]
 40%|███▉      | 540/1355 [00:06<00:10, 79.45it/s]
 40%|████      | 548/1355 [00:06<00:10, 78.07it/s]
 41%|████      | 557/1355 [00:06<00:09, 80.42it/s]
 42%|████▏     | 566/1355 [00:07<00:09, 82.39it/s]
 42%|████▏     | 575/1355 [00:07<00:09, 81.37it/s]
 43%|████▎     | 584/1355 [00:07<00:09, 80.81it/s]
 44%|████▍     | 593/1355 [00:07<00:09, 81.91it/s]
 44%|████▍     | 602/1355 [00:07<00:09, 79.26it/s]
 45%|████▌     | 610/1355 [00:07<00:09, 77.86it/s]
 46%|████▌     | 618/1355 [00:07<00:09, 76.53it/s]
 46%|████▌     | 626/1355 [00:07<00:09, 76.13it/s]
 47%|████▋     | 635/1355 [00:07<00:09, 77.47it/s]
 48%|████▊     | 644/1355 [00:08<00:09, 78.11it/s]
 48%|████▊     | 653/1355 [00:08<00:08, 79.99it/s]
 49%|████▉     | 662/1355 [00:08<00:08, 78.01it/s]
 49%|████▉     | 670/1355 [00:08<00:08, 77.20it/s]
 50%|█████     | 679/1355 [00:08<00:08, 78.41it/s]
 51%|█████     | 687/1355 [00:08<00:08, 78.21it/s]
 51%|█████▏    | 696/1355 [00:08<00:08, 79.53it/s]
 52%|█████▏    | 704/1355 [00:08<00:08, 77.15it/s]
 53%|█████▎    | 713/1355 [00:08<00:08, 79.89it/s]
 53%|█████▎    | 722/1355 [00:09<00:07, 79.62it/s]
 54%|█████▍    | 731/1355 [00:09<00:07, 80.99it/s]
 55%|█████▍    | 740/1355 [00:09<00:07, 81.06it/s]
 55%|█████▌    | 749/1355 [00:09<00:07, 80.52it/s]
 56%|█████▌    | 758/1355 [00:09<00:07, 77.20it/s]
 57%|█████▋    | 766/1355 [00:09<00:07, 77.39it/s]
 57%|█████▋    | 775/1355 [00:09<00:07, 80.26it/s]
 58%|█████▊    | 784/1355 [00:09<00:07, 75.66it/s]
 59%|█████▊    | 793/1355 [00:09<00:07, 76.94it/s]
 59%|█████▉    | 801/1355 [00:10<00:07, 77.53it/s]
 60%|█████▉    | 810/1355 [00:10<00:06, 80.01it/s]
 60%|██████    | 819/1355 [00:10<00:06, 80.84it/s]
 61%|██████    | 828/1355 [00:10<00:06, 80.09it/s]
 62%|██████▏   | 837/1355 [00:10<00:06, 78.50it/s]
 62%|██████▏   | 845/1355 [00:10<00:06, 78.57it/s]
 63%|██████▎   | 853/1355 [00:10<00:06, 75.52it/s]
 64%|██████▎   | 862/1355 [00:10<00:06, 77.80it/s]
 64%|██████▍   | 871/1355 [00:10<00:06, 79.35it/s]
 65%|██████▍   | 880/1355 [00:11<00:05, 81.16it/s]
 66%|██████▌   | 889/1355 [00:11<00:05, 81.61it/s]
 66%|██████▋   | 898/1355 [00:11<00:05, 81.34it/s]
 67%|██████▋   | 907/1355 [00:11<00:05, 81.48it/s]
 68%|██████▊   | 916/1355 [00:11<00:05, 81.64it/s]
 68%|███���██▊   | 925/1355 [00:11<00:05, 78.43it/s]
 69%|██████▉   | 933/1355 [00:11<00:05, 77.85it/s]
 69%|██████▉   | 941/1355 [00:11<00:05, 77.94it/s]
 70%|███████   | 950/1355 [00:11<00:05, 80.31it/s]
 71%|███████   | 959/1355 [00:11<00:04, 81.24it/s]
 71%|███████▏  | 968/1355 [00:12<00:04, 80.46it/s]
 72%|███████▏  | 977/1355 [00:12<00:04, 76.46it/s]
 73%|███████▎  | 986/1355 [00:12<00:04, 78.44it/s]
 73%|███████▎  | 995/1355 [00:12<00:04, 80.04it/s]
 74%|███████▍  | 1004/1355 [00:12<00:04, 78.88it/s]
 75%|███████▍  | 1012/1355 [00:12<00:04, 76.80it/s]
 75%|███████▌  | 1020/1355 [00:12<00:04, 76.99it/s]
 76%|███████▌  | 1029/1355 [00:12<00:04, 78.29it/s]
 77%|███████▋  | 1038/1355 [00:12<00:03, 79.96it/s]
 77%|███████▋  | 1047/1355 [00:13<00:03, 81.57it/s]
 78%|███████▊  | 1056/1355 [00:13<00:03, 80.17it/s]
 79%|███████▊  | 1065/1355 [00:13<00:03, 81.20it/s]
 79%|███████▉  | 1074/1355 [00:13<00:03, 81.92it/s]
 80%|███████▉  | 1083/1355 [00:13<00:03, 79.31it/s]
 81%|████████  | 1091/1355 [00:13<00:03, 76.77it/s]
 81%|████████  | 1100/1355 [00:13<00:03, 78.36it/s]
 82%|████████▏ | 1108/1355 [00:13<00:03, 78.73it/s]
 82%|████████▏ | 1116/1355 [00:13<00:03, 75.10it/s]
 83%|████████▎ | 1124/1355 [00:14<00:03, 75.74it/s]
 84%|████████▎ | 1133/1355 [00:14<00:02, 78.40it/s]
 84%|████████▍ | 1141/1355 [00:14<00:02, 77.09it/s]
 85%|████████▍ | 1149/1355 [00:14<00:02, 76.45it/s]
 85%|████████▌ | 1157/1355 [00:14<00:02, 74.14it/s]
 86%|████████▌ | 1165/1355 [00:14<00:02, 73.80it/s]
 87%|████████▋ | 1174/1355 [00:14<00:02, 76.18it/s]
 87%|████████▋ | 1183/1355 [00:14<00:02, 77.91it/s]
 88%|████████▊ | 1191/1355 [00:14<00:02, 78.13it/s]
 88%|████████▊ | 1199/1355 [00:15<00:02, 77.74it/s]
 89%|████████▉ | 1208/1355 [00:15<00:01, 78.65it/s]
 90%|████████▉ | 1216/1355 [00:15<00:01, 76.56it/s]
 90%|█████████ | 1225/1355 [00:15<00:01, 78.91it/s]
 91%|█████████ | 1233/1355 [00:15<00:01, 78.86it/s]
 92%|█████████▏| 1241/1355 [00:15<00:01, 76.10it/s]
 92%|█████████▏| 1249/1355 [00:15<00:01, 75.34it/s]
 93%|█████████▎| 1258/1355 [00:15<00:01, 76.62it/s]
 93%|█████████▎| 1266/1355 [00:15<00:01, 77.36it/s]
 94%|█████████▍| 1275/1355 [00:16<00:01, 78.50it/s]
 95%|█████████▍| 1284/1355 [00:16<00:00, 80.34it/s]
 95%|█████████▌| 1293/1355 [00:16<00:00, 81.89it/s]
 96%|█████████▌| 1302/1355 [00:16<00:00, 81.63it/s]
 97%|█████████▋| 1311/1355 [00:16<00:00, 80.03it/s]
 97%|█████████▋| 1320/1355 [00:16<00:00, 77.99it/s]
 98%|█████████▊| 1328/1355 [00:16<00:00, 76.51it/s]
 99%|█████████▊| 1336/1355 [00:16<00:00, 75.93it/s]
 99%|█████████▉| 1344/1355 [00:16<00:00, 75.51it/s]
+[INFO|trainer.py:3503] 2024-09-09 21:34:11,757 >> Saving model checkpoint to /content/dissertation/scripts/ner/output
+[INFO|configuration_utils.py:472] 2024-09-09 21:34:11,759 >> Configuration saved in /content/dissertation/scripts/ner/output/config.json
+[INFO|modeling_utils.py:2799] 2024-09-09 21:34:13,106 >> Model weights saved in /content/dissertation/scripts/ner/output/model.safetensors
+[INFO|tokenization_utils_base.py:2684] 2024-09-09 21:34:13,107 >> tokenizer config file saved in /content/dissertation/scripts/ner/output/tokenizer_config.json
+[INFO|tokenization_utils_base.py:2693] 2024-09-09 21:34:13,107 >> Special tokens file saved in /content/dissertation/scripts/ner/output/special_tokens_map.json
+***** predict metrics *****
+  predict_accuracy           =     0.9922
+  predict_f1                 =     0.8534
+  predict_loss               =     0.0449
+  predict_precision          =     0.8332
+  predict_recall             =     0.8747
+  predict_runtime            = 0:00:23.14
+  predict_samples_per_second =    468.329
+  predict_steps_per_second   =     58.552
+

train_results.json

@@ -1,9 +1,9 @@
 {
-    "epoch": 10.0,
-    "total_flos": 1.5694885147146138e+16,
-    "train_loss": 0.002533856062142209,
-    "train_runtime": 1422.7831,
-    "train_samples": 31229,
-    "train_samples_per_second": 219.492,
-    "train_steps_per_second": 3.43
+    "epoch": 9.991889699918897,
+    "total_flos": 1.919784837910831e+16,
+    "train_loss": 0.008344567238123386,
+    "train_runtime": 1725.5138,
+    "train_samples": 39426,
+    "train_samples_per_second": 228.488,
+    "train_steps_per_second": 3.57
 }

trainer_state.json

@@ -1,208 +1,229 @@
 {
-  "best_metric": 0.9521203830369357,
-  "best_model_checkpoint": "/content/dissertation/scripts/ner/output/checkpoint-2440",
-  "epoch": 10.0,
+  "best_metric": 0.8512137823022711,
+  "best_model_checkpoint": "/content/dissertation/scripts/ner/output/checkpoint-4315",
+  "epoch": 9.991889699918897,
   "eval_steps": 500,
-  "global_step": 4880,
+  "global_step": 6160,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 1.0,
-      "eval_accuracy": 0.9987641945981572,
-      "eval_f1": 0.9355123674911661,
-      "eval_loss": 0.003916793502867222,
-      "eval_precision": 0.9005102040816326,
-      "eval_recall": 0.9733455882352942,
-      "eval_runtime": 13.936,
-      "eval_samples_per_second": 488.661,
-      "eval_steps_per_second": 61.136,
-      "step": 488
-    },
-    {
-      "epoch": 1.0245901639344261,
-      "grad_norm": 0.24995087087154388,
-      "learning_rate": 4.487704918032787e-05,
-      "loss": 0.0189,
+      "epoch": 0.8110300081103001,
+      "grad_norm": 0.7429273724555969,
+      "learning_rate": 4.5941558441558444e-05,
+      "loss": 0.0571,
       "step": 500
     },
     {
-      "epoch": 2.0,
-      "eval_accuracy": 0.9989495654084337,
-      "eval_f1": 0.9413886384129847,
-      "eval_loss": 0.0032420416828244925,
-      "eval_precision": 0.9238938053097345,
-      "eval_recall": 0.9595588235294118,
-      "eval_runtime": 14.0058,
-      "eval_samples_per_second": 486.228,
-      "eval_steps_per_second": 60.832,
-      "step": 976
-    },
-    {
-      "epoch": 2.0491803278688523,
-      "grad_norm": 0.04940846189856529,
-      "learning_rate": 3.975409836065574e-05,
-      "loss": 0.0027,
-      "step": 1000
+      "epoch": 0.9991889699918897,
+      "eval_accuracy": 0.9906421648980531,
+      "eval_f1": 0.8112888052681091,
+      "eval_loss": 0.02658958174288273,
+      "eval_precision": 0.776657060518732,
+      "eval_recall": 0.8491532099251674,
+      "eval_runtime": 15.3903,
+      "eval_samples_per_second": 477.832,
+      "eval_steps_per_second": 59.778,
+      "step": 616
     },
     {
-      "epoch": 3.0,
-      "eval_accuracy": 0.9989358342373021,
-      "eval_f1": 0.9402784014369105,
-      "eval_loss": 0.0043699671514332294,
-      "eval_precision": 0.9192273924495171,
-      "eval_recall": 0.9623161764705882,
-      "eval_runtime": 14.0607,
-      "eval_samples_per_second": 484.329,
-      "eval_steps_per_second": 60.594,
-      "step": 1464
+      "epoch": 1.6220600162206003,
+      "grad_norm": 0.32481127977371216,
+      "learning_rate": 4.1883116883116886e-05,
+      "loss": 0.018,
+      "step": 1000
     },
     {
-      "epoch": 3.0737704918032787,
-      "grad_norm": 0.01636342518031597,
-      "learning_rate": 3.463114754098361e-05,
-      "loss": 0.0015,
+      "epoch": 2.0,
+      "eval_accuracy": 0.9914186293963707,
+      "eval_f1": 0.8270561106840891,
+      "eval_loss": 0.030444670468568802,
+      "eval_precision": 0.8075046904315197,
+      "eval_recall": 0.8475777865301299,
+      "eval_runtime": 15.7588,
+      "eval_samples_per_second": 466.66,
+      "eval_steps_per_second": 58.38,
+      "step": 1233
+    },
+    {
+      "epoch": 2.4330900243309004,
+      "grad_norm": 0.10290508717298508,
+      "learning_rate": 3.782467532467533e-05,
+      "loss": 0.0101,
       "step": 1500
     },
     {
-      "epoch": 4.0,
-      "eval_accuracy": 0.9990594147774864,
-      "eval_f1": 0.9445208619899129,
-      "eval_loss": 0.0036185304634273052,
-      "eval_precision": 0.9423604757548033,
-      "eval_recall": 0.9466911764705882,
-      "eval_runtime": 14.26,
-      "eval_samples_per_second": 477.56,
-      "eval_steps_per_second": 59.748,
-      "step": 1952
-    },
-    {
-      "epoch": 4.098360655737705,
-      "grad_norm": 0.006754329428076744,
-      "learning_rate": 2.9508196721311478e-05,
-      "loss": 0.0007,
-      "step": 2000
+      "epoch": 2.9991889699918897,
+      "eval_accuracy": 0.9906249101314238,
+      "eval_f1": 0.8310784692694241,
+      "eval_loss": 0.03564191609621048,
+      "eval_precision": 0.8159392789373814,
+      "eval_recall": 0.8467900748326113,
+      "eval_runtime": 15.5516,
+      "eval_samples_per_second": 472.877,
+      "eval_steps_per_second": 59.158,
+      "step": 1849
     },
     {
-      "epoch": 5.0,
-      "eval_accuracy": 0.9991418018042759,
-      "eval_f1": 0.9521203830369357,
-      "eval_loss": 0.004440780263394117,
-      "eval_precision": 0.9447963800904977,
-      "eval_recall": 0.9595588235294118,
-      "eval_runtime": 14.0071,
-      "eval_samples_per_second": 486.182,
-      "eval_steps_per_second": 60.826,
-      "step": 2440
+      "epoch": 3.2441200324412005,
+      "grad_norm": 0.3212472200393677,
+      "learning_rate": 3.376623376623377e-05,
+      "loss": 0.0057,
+      "step": 2000
     },
     {
-      "epoch": 5.122950819672131,
-      "grad_norm": 0.0005009469459764659,
-      "learning_rate": 2.4385245901639343e-05,
-      "loss": 0.0004,
+      "epoch": 4.0,
+      "eval_accuracy": 0.9910217697638973,
+      "eval_f1": 0.8348231636222309,
+      "eval_loss": 0.03653988987207413,
+      "eval_precision": 0.8239355581127733,
+      "eval_recall": 0.8460023631350926,
+      "eval_runtime": 15.8023,
+      "eval_samples_per_second": 465.375,
+      "eval_steps_per_second": 58.219,
+      "step": 2466
+    },
+    {
+      "epoch": 4.0551500405515,
+      "grad_norm": 1.049054741859436,
+      "learning_rate": 2.9707792207792207e-05,
+      "loss": 0.0039,
       "step": 2500
     },
     {
-      "epoch": 6.0,
-      "eval_accuracy": 0.9990456836063548,
-      "eval_f1": 0.9464368886818817,
-      "eval_loss": 0.005531433038413525,
-      "eval_precision": 0.959395656279509,
-      "eval_recall": 0.9338235294117647,
-      "eval_runtime": 14.0724,
-      "eval_samples_per_second": 483.926,
-      "eval_steps_per_second": 60.544,
-      "step": 2928
-    },
-    {
-      "epoch": 6.147540983606557,
-      "grad_norm": 0.7484572529792786,
-      "learning_rate": 1.9262295081967212e-05,
-      "loss": 0.0002,
+      "epoch": 4.866180048661801,
+      "grad_norm": 0.12911342084407806,
+      "learning_rate": 2.5649350649350652e-05,
+      "loss": 0.0027,
       "step": 3000
     },
     {
-      "epoch": 7.0,
-      "eval_accuracy": 0.9989907589218284,
-      "eval_f1": 0.9427393495190106,
-      "eval_loss": 0.004938796162605286,
-      "eval_precision": 0.9397260273972603,
-      "eval_recall": 0.9457720588235294,
-      "eval_runtime": 13.9575,
-      "eval_samples_per_second": 487.909,
-      "eval_steps_per_second": 61.042,
-      "step": 3416
+      "epoch": 4.99918896999189,
+      "eval_accuracy": 0.9915624191182815,
+      "eval_f1": 0.8343344313117613,
+      "eval_loss": 0.03956884518265724,
+      "eval_precision": 0.8211289092295957,
+      "eval_recall": 0.8479716423788893,
+      "eval_runtime": 15.4772,
+      "eval_samples_per_second": 475.151,
+      "eval_steps_per_second": 59.442,
+      "step": 3082
     },
     {
-      "epoch": 7.172131147540983,
-      "grad_norm": 0.00041869020787999034,
-      "learning_rate": 1.4139344262295081e-05,
-      "loss": 0.0002,
+      "epoch": 5.6772100567721,
+      "grad_norm": 0.1811482459306717,
+      "learning_rate": 2.1590909090909093e-05,
+      "loss": 0.0018,
       "step": 3500
     },
     {
-      "epoch": 8.0,
-      "eval_accuracy": 0.999073145948618,
-      "eval_f1": 0.9468864468864469,
-      "eval_loss": 0.005282656755298376,
-      "eval_precision": 0.9434306569343066,
-      "eval_recall": 0.9503676470588235,
-      "eval_runtime": 14.1087,
-      "eval_samples_per_second": 482.68,
-      "eval_steps_per_second": 60.388,
-      "step": 3904
-    },
-    {
-      "epoch": 8.19672131147541,
-      "grad_norm": 0.004756764974445105,
-      "learning_rate": 9.016393442622952e-06,
-      "loss": 0.0001,
+      "epoch": 6.0,
+      "eval_accuracy": 0.9915394127627757,
+      "eval_f1": 0.846658941676323,
+      "eval_loss": 0.04350048676133156,
+      "eval_precision": 0.8306176582038651,
+      "eval_recall": 0.8633320204805042,
+      "eval_runtime": 15.4958,
+      "eval_samples_per_second": 474.581,
+      "eval_steps_per_second": 59.371,
+      "step": 3699
+    },
+    {
+      "epoch": 6.488240064882401,
+      "grad_norm": 0.0691465437412262,
+      "learning_rate": 1.7532467532467535e-05,
+      "loss": 0.0013,
       "step": 4000
     },
     {
-      "epoch": 9.0,
-      "eval_accuracy": 0.9990525491919205,
-      "eval_f1": 0.946404031149794,
-      "eval_loss": 0.0050343358889222145,
-      "eval_precision": 0.94337899543379,
-      "eval_recall": 0.9494485294117647,
-      "eval_runtime": 14.373,
-      "eval_samples_per_second": 473.804,
-      "eval_steps_per_second": 59.278,
-      "step": 4392
+      "epoch": 6.99918896999189,
+      "eval_accuracy": 0.991867253328732,
+      "eval_f1": 0.8512137823022711,
+      "eval_loss": 0.047813717275857925,
+      "eval_precision": 0.8462436745815493,
+      "eval_recall": 0.8562426152028357,
+      "eval_runtime": 15.6578,
+      "eval_samples_per_second": 469.669,
+      "eval_steps_per_second": 58.757,
+      "step": 4315
     },
     {
-      "epoch": 9.221311475409836,
-      "grad_norm": 0.00033881032140925527,
-      "learning_rate": 3.89344262295082e-06,
-      "loss": 0.0001,
+      "epoch": 7.299270072992701,
+      "grad_norm": 0.016636306419968605,
+      "learning_rate": 1.3474025974025975e-05,
+      "loss": 0.0008,
       "step": 4500
     },
     {
-      "epoch": 10.0,
-      "eval_accuracy": 0.9990525491919205,
-      "eval_f1": 0.945537757437071,
-      "eval_loss": 0.005196314305067062,
-      "eval_precision": 0.9416590701914311,
-      "eval_recall": 0.9494485294117647,
-      "eval_runtime": 14.6459,
-      "eval_samples_per_second": 464.976,
-      "eval_steps_per_second": 58.173,
-      "step": 4880
-    },
-    {
-      "epoch": 10.0,
-      "step": 4880,
-      "total_flos": 1.5694885147146138e+16,
-      "train_loss": 0.002533856062142209,
-      "train_runtime": 1422.7831,
-      "train_samples_per_second": 219.492,
-      "train_steps_per_second": 3.43
+      "epoch": 8.0,
+      "eval_accuracy": 0.9914761452851351,
+      "eval_f1": 0.8478387284357433,
+      "eval_loss": 0.04689360409975052,
+      "eval_precision": 0.8347328244274809,
+      "eval_recall": 0.8613627412367074,
+      "eval_runtime": 15.6856,
+      "eval_samples_per_second": 468.838,
+      "eval_steps_per_second": 58.653,
+      "step": 4932
+    },
+    {
+      "epoch": 8.110300081103,
+      "grad_norm": 0.02262457087635994,
+      "learning_rate": 9.415584415584416e-06,
+      "loss": 0.0008,
+      "step": 5000
+    },
+    {
+      "epoch": 8.921330089213301,
+      "grad_norm": 0.6416856646537781,
+      "learning_rate": 5.357142857142857e-06,
+      "loss": 0.0004,
+      "step": 5500
+    },
+    {
+      "epoch": 8.99918896999189,
+      "eval_accuracy": 0.9918845080953613,
+      "eval_f1": 0.851080397118941,
+      "eval_loss": 0.05153572931885719,
+      "eval_precision": 0.8414164742109315,
+      "eval_recall": 0.860968885387948,
+      "eval_runtime": 15.5849,
+      "eval_samples_per_second": 471.866,
+      "eval_steps_per_second": 59.031,
+      "step": 5548
+    },
+    {
+      "epoch": 9.732360097323602,
+      "grad_norm": 0.011619620956480503,
+      "learning_rate": 1.2987012987012988e-06,
+      "loss": 0.0002,
+      "step": 6000
+    },
+    {
+      "epoch": 9.991889699918897,
+      "eval_accuracy": 0.9917694763178329,
+      "eval_f1": 0.8490859587709063,
+      "eval_loss": 0.05196274444460869,
+      "eval_precision": 0.8386477141759509,
+      "eval_recall": 0.85978731784167,
+      "eval_runtime": 15.9471,
+      "eval_samples_per_second": 461.149,
+      "eval_steps_per_second": 57.691,
+      "step": 6160
+    },
+    {
+      "epoch": 9.991889699918897,
+      "step": 6160,
+      "total_flos": 1.919784837910831e+16,
+      "train_loss": 0.008344567238123386,
+      "train_runtime": 1725.5138,
+      "train_samples_per_second": 228.488,
+      "train_steps_per_second": 3.57
     }
   ],
   "logging_steps": 500,
-  "max_steps": 4880,
+  "max_steps": 6160,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 10,
   "save_steps": 500,
@@ -218,7 +239,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 1.5694885147146138e+16,
+  "total_flos": 1.919784837910831e+16,
   "train_batch_size": 32,
   "trial_name": null,
   "trial_params": null