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:
-- drugtemist-85-ner
+- Rodrigo1771/drugtemist-85-ner
 metrics:
 - precision
 - recall
@@ -18,24 +19,24 @@ model-index:
       name: Token Classification
       type: token-classification
     dataset:
-      name: drugtemist-85-ner
-      type: drugtemist-85-ner
+      name: Rodrigo1771/drugtemist-85-ner
+      type: Rodrigo1771/drugtemist-85-ner
       config: DrugTEMIST NER
       split: validation
       args: DrugTEMIST NER
     metrics:
     - name: Precision
       type: precision
-      value: 0.9347826086956522
+      value: 0.9461187214611873
     - name: Recall
       type: recall
-      value: 0.9485294117647058
+      value: 0.9522058823529411
     - name: F1
       type: f1
-      value: 0.9416058394160585
+      value: 0.9491525423728814
     - name: Accuracy
       type: accuracy
-      value: 0.9989083718950389
+      value: 0.9989426998228679
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -43,12 +44,12 @@ 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 drugtemist-85-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/drugtemist-85-ner dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.0058
-- Precision: 0.9348
-- Recall: 0.9485
-- F1: 0.9416
+- Loss: 0.0048
+- Precision: 0.9461
+- Recall: 0.9522
+- F1: 0.9492
 - Accuracy: 0.9989
 
 ## Model description

all_results.json

@@ -1,26 +1,26 @@
 {
-    "epoch": 9.98998998998999,
-    "eval_accuracy": 0.9764853694371592,
-    "eval_f1": 0.8040201005025126,
-    "eval_loss": 0.14244574308395386,
-    "eval_precision": 0.803175344384777,
-    "eval_recall": 0.8048666354702855,
-    "eval_runtime": 14.2417,
+    "epoch": 10.0,
+    "eval_accuracy": 0.9989426998228679,
+    "eval_f1": 0.9491525423728814,
+    "eval_loss": 0.004777050111442804,
+    "eval_precision": 0.9461187214611873,
+    "eval_recall": 0.9522058823529411,
+    "eval_runtime": 13.9476,
     "eval_samples": 6810,
-    "eval_samples_per_second": 478.175,
-    "eval_steps_per_second": 59.825,
-    "predict_accuracy": 0.9760745138469196,
-    "predict_f1": 0.7984671441135821,
-    "predict_loss": 0.12913450598716736,
-    "predict_precision": 0.7908163265306123,
-    "predict_recall": 0.8062674448109617,
-    "predict_runtime": 28.8701,
-    "predict_samples_per_second": 506.199,
-    "predict_steps_per_second": 63.284,
-    "total_flos": 1.5071241212671032e+16,
-    "train_loss": 0.022475468706272407,
-    "train_runtime": 1385.1143,
-    "train_samples": 31947,
-    "train_samples_per_second": 230.645,
-    "train_steps_per_second": 3.603
+    "eval_samples_per_second": 488.256,
+    "eval_steps_per_second": 61.086,
+    "predict_accuracy": 0.9987478324070453,
+    "predict_f1": 0.9243073407597828,
+    "predict_loss": 0.005894536152482033,
+    "predict_precision": 0.9069506726457399,
+    "predict_recall": 0.9423412929528246,
+    "predict_runtime": 28.2179,
+    "predict_samples_per_second": 517.899,
+    "predict_steps_per_second": 64.746,
+    "total_flos": 1.4714840952259542e+16,
+    "train_loss": 0.002772659832779558,
+    "train_runtime": 1349.0548,
+    "train_samples": 29797,
+    "train_samples_per_second": 220.873,
+    "train_steps_per_second": 3.454
 }

eval_results.json

@@ -1,12 +1,12 @@
 {
-    "epoch": 9.98998998998999,
-    "eval_accuracy": 0.9764853694371592,
-    "eval_f1": 0.8040201005025126,
-    "eval_loss": 0.14244574308395386,
-    "eval_precision": 0.803175344384777,
-    "eval_recall": 0.8048666354702855,
-    "eval_runtime": 14.2417,
+    "epoch": 10.0,
+    "eval_accuracy": 0.9989426998228679,
+    "eval_f1": 0.9491525423728814,
+    "eval_loss": 0.004777050111442804,
+    "eval_precision": 0.9461187214611873,
+    "eval_recall": 0.9522058823529411,
+    "eval_runtime": 13.9476,
     "eval_samples": 6810,
-    "eval_samples_per_second": 478.175,
-    "eval_steps_per_second": 59.825
+    "eval_samples_per_second": 488.256,
+    "eval_steps_per_second": 61.086
 }

predict_results.json

@@ -1,10 +1,10 @@
 {
-    "predict_accuracy": 0.9760745138469196,
-    "predict_f1": 0.7984671441135821,
-    "predict_loss": 0.12913450598716736,
-    "predict_precision": 0.7908163265306123,
-    "predict_recall": 0.8062674448109617,
-    "predict_runtime": 28.8701,
-    "predict_samples_per_second": 506.199,
-    "predict_steps_per_second": 63.284
+    "predict_accuracy": 0.9987478324070453,
+    "predict_f1": 0.9243073407597828,
+    "predict_loss": 0.005894536152482033,
+    "predict_precision": 0.9069506726457399,
+    "predict_recall": 0.9423412929528246,
+    "predict_runtime": 28.2179,
+    "predict_samples_per_second": 517.899,
+    "predict_steps_per_second": 64.746
 }

tb/events.out.tfevents.1725579302.2a66098fac87.9264.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2ddb06f8e19acc3cfe83ba96582312f0eda7c54cb22293014a850cd5aff12d3d
+size 560

train.log

@@ -1450,3 +1450,51 @@ Training completed. Do not forget to share your model on huggingface.co/models =
 {'eval_loss': 0.005793666001409292, 'eval_precision': 0.9347826086956522, 'eval_recall': 0.9485294117647058, 'eval_f1': 0.9416058394160585, 'eval_accuracy': 0.9989083718950389, 'eval_runtime': 14.4595, 'eval_samples_per_second': 470.971, 'eval_steps_per_second': 58.923, 'epoch': 10.0}
 {'train_runtime': 1349.0548, 'train_samples_per_second': 220.873, 'train_steps_per_second': 3.454, 'train_loss': 0.002772659832779558, 'epoch': 10.0}
 
+***** train metrics *****
+  epoch                    =       10.0
+  total_flos               = 13704263GF
+  train_loss               =     0.0028
+  train_runtime            = 0:22:29.05
+  train_samples            =      29797
+  train_samples_per_second =    220.873
+  train_steps_per_second   =      3.454
+09/05/2024 23:34:48 - INFO - __main__ -   *** Evaluate ***
+[INFO|trainer.py:811] 2024-09-05 23:34:48,810 >> The following columns in the evaluation set don't have a corresponding argument in `RobertaForTokenClassification.forward` and have been ignored: tokens, ner_tags, id. If tokens, ner_tags, id are not expected by `RobertaForTokenClassification.forward`,  you can safely ignore this message.
+[INFO|trainer.py:3819] 2024-09-05 23:34:48,813 >> 
+***** Running Evaluation *****
+[INFO|trainer.py:3821] 2024-09-05 23:34:48,813 >>   Num examples = 6810
+[INFO|trainer.py:3824] 2024-09-05 23:34:48,813 >>   Batch size = 8
+
  0%|          | 0/852 [00:00<?, ?it/s]
  1%|          | 10/852 [00:00<00:09, 89.77it/s]
  2%|▏         | 19/852 [00:00<00:10, 77.96it/s]
  3%|▎         | 27/852 [00:00<00:10, 78.13it/s]
  4%|▍         | 36/852 [00:00<00:10, 78.74it/s]
  5%|▌         | 45/852 [00:00<00:10, 80.47it/s]
  6%|▋         | 54/852 [00:00<00:09, 81.97it/s]
  7%|▋         | 63/852 [00:00<00:09, 81.96it/s]
  8%|▊         | 72/852 [00:00<00:09, 80.15it/s]
 10%|▉         | 81/852 [00:01<00:09, 80.19it/s]
 11%|█         | 90/852 [00:01<00:09, 80.63it/s]
 12%|█▏        | 99/852 [00:01<00:09, 80.42it/s]
 13%|█▎        | 108/852 [00:01<00:09, 79.74it/s]
 14%|█▎        | 117/852 [00:01<00:09, 80.86it/s]
 15%|█▍        | 126/852 [00:01<00:09, 78.38it/s]
 16%|█▌        | 135/852 [00:01<00:09, 79.16it/s]
 17%|█▋        | 143/852 [00:01<00:08, 79.17it/s]
 18%|█▊        | 151/852 [00:01<00:08, 78.56it/s]
 19%|█▉        | 160/852 [00:01<00:08, 81.22it/s]
 20%|█▉        | 169/852 [00:02<00:08, 81.16it/s]
 21%|██        | 178/852 [00:02<00:08, 81.76it/s]
 22%|██▏       | 187/852 [00:02<00:08, 82.46it/s]
 23%|██▎       | 196/852 [00:02<00:07, 82.29it/s]
 24%|██▍       | 205/852 [00:02<00:07, 83.12it/s]
 25%|██▌       | 214/852 [00:02<00:07, 80.08it/s]
 26%|██▌       | 223/852 [00:02<00:07, 81.13it/s]
 27%|██▋       | 232/852 [00:02<00:07, 81.21it/s]
 28%|██▊       | 241/852 [00:03<00:07, 78.10it/s]
 29%|██▉       | 250/852 [00:03<00:07, 79.34it/s]
 30%|███       | 259/852 [00:03<00:07, 80.76it/s]
 31%|███▏      | 268/852 [00:03<00:07, 74.97it/s]
 33%|███▎      | 277/852 [00:03<00:07, 77.71it/s]
 34%|███▎      | 286/852 [00:03<00:07, 80.09it/s]
 35%|███▍      | 295/852 [00:03<00:06, 80.08it/s]
 36%|███▌      | 304/852 [00:03<00:06, 81.59it/s]
 37%|███▋      | 313/852 [00:03<00:06, 80.30it/s]
 38%|███▊      | 322/852 [00:04<00:06, 82.39it/s]
 39%|███▉      | 331/852 [00:04<00:06, 82.16it/s]
 40%|███▉      | 340/852 [00:04<00:06, 82.30it/s]
 41%|████      | 349/852 [00:04<00:06, 82.47it/s]
 42%|████▏     | 358/852 [00:04<00:06, 80.49it/s]
 43%|████▎     | 367/852 [00:04<00:05, 81.17it/s]
 44%|████▍     | 376/852 [00:04<00:05, 81.82it/s]
 45%|████▌     | 385/852 [00:04<00:05, 81.26it/s]
 46%|████▌     | 394/852 [00:04<00:05, 81.23it/s]
 47%|████▋     | 403/852 [00:05<00:05, 81.17it/s]
 48%|████▊     | 412/852 [00:05<00:05, 79.22it/s]
 49%|████▉     | 421/852 [00:05<00:05, 80.98it/s]
 50%|█████     | 430/852 [00:05<00:05, 80.11it/s]
 52%|█████▏    | 439/852 [00:05<00:05, 81.83it/s]
 53%|█████▎    | 448/852 [00:05<00:04, 82.37it/s]
 54%|█████▎    | 457/852 [00:05<00:04, 83.25it/s]
 55%|█████▍    | 466/852 [00:05<00:04, 80.61it/s]
 56%|█████▌    | 475/852 [00:05<00:04, 77.21it/s]
 57%|█████▋    | 484/852 [00:06<00:04, 77.43it/s]
 58%|█████▊    | 493/852 [00:06<00:04, 79.71it/s]
 59%|█████▉    | 502/852 [00:06<00:04, 81.52it/s]
 60%|█████▉    | 511/852 [00:06<00:04, 81.79it/s]
 61%|██████    | 520/852 [00:06<00:04, 82.93it/s]
 62%|██████▏   | 529/852 [00:06<00:03, 80.76it/s]
 63%|████��█▎   | 538/852 [00:06<00:03, 82.65it/s]
 64%|██████▍   | 547/852 [00:06<00:03, 83.43it/s]
 65%|██████▌   | 556/852 [00:06<00:03, 80.56it/s]
 66%|██████▋   | 565/852 [00:06<00:03, 82.97it/s]
 67%|██████▋   | 574/852 [00:07<00:03, 83.67it/s]
 68%|██████▊   | 583/852 [00:07<00:03, 83.26it/s]
 69%|██████▉   | 592/852 [00:07<00:03, 82.65it/s]
 71%|███████   | 601/852 [00:07<00:03, 82.89it/s]
 72%|███████▏  | 610/852 [00:07<00:02, 83.17it/s]
 73%|███████▎  | 619/852 [00:07<00:02, 80.86it/s]
 74%|███████▎  | 628/852 [00:07<00:02, 80.15it/s]
 75%|███████▍  | 637/852 [00:07<00:02, 80.82it/s]
 76%|███████▌  | 646/852 [00:07<00:02, 78.59it/s]
 77%|███████▋  | 655/852 [00:08<00:02, 80.46it/s]
 78%|███████▊  | 664/852 [00:08<00:02, 81.10it/s]
 79%|███████▉  | 673/852 [00:08<00:02, 80.96it/s]
 80%|████████  | 682/852 [00:08<00:02, 81.12it/s]
 81%|████████  | 691/852 [00:08<00:01, 82.73it/s]
 82%|████████▏ | 700/852 [00:08<00:01, 83.32it/s]
 83%|████████▎ | 709/852 [00:08<00:01, 84.29it/s]
 84%|████████▍ | 718/852 [00:08<00:01, 83.48it/s]
 85%|████████▌ | 727/852 [00:08<00:01, 83.89it/s]
 86%|████████▋ | 736/852 [00:09<00:01, 83.25it/s]
 87%|████████▋ | 745/852 [00:09<00:01, 84.23it/s]
 88%|████████▊ | 754/852 [00:09<00:01, 84.42it/s]
 90%|████████▉ | 763/852 [00:09<00:01, 85.20it/s]
 91%|█████████ | 772/852 [00:09<00:00, 83.63it/s]
 92%|█████████▏| 781/852 [00:09<00:00, 82.15it/s]
 93%|█████████▎| 790/852 [00:09<00:00, 82.20it/s]
 94%|█████████▍| 799/852 [00:09<00:00, 82.36it/s]
 95%|█████████▍| 808/852 [00:09<00:00, 83.98it/s]
 96%|█████████▌| 817/852 [00:10<00:00, 82.39it/s]
 97%|█████████▋| 826/852 [00:10<00:00, 83.62it/s]
 98%|█████████▊| 835/852 [00:10<00:00, 83.71it/s]
 99%|█████████▉| 844/852 [00:10<00:00, 82.35it/s]
+***** eval metrics *****
+  epoch                   =       10.0
+  eval_accuracy           =     0.9989
+  eval_f1                 =     0.9492
+  eval_loss               =     0.0048
+  eval_precision          =     0.9461
+  eval_recall             =     0.9522
+  eval_runtime            = 0:00:13.94
+  eval_samples            =       6810
+  eval_samples_per_second =    488.256
+  eval_steps_per_second   =     61.086
+09/05/2024 23:35:02 - INFO - __main__ -   *** Predict ***
+[INFO|trainer.py:811] 2024-09-05 23:35:02,765 >> The following columns in the test set don't have a corresponding argument in `RobertaForTokenClassification.forward` and have been ignored: tokens, ner_tags, id. If tokens, ner_tags, id are not expected by `RobertaForTokenClassification.forward`,  you can safely ignore this message.
+[INFO|trainer.py:3819] 2024-09-05 23:35:02,767 >> 
+***** Running Prediction *****
+[INFO|trainer.py:3821] 2024-09-05 23:35:02,767 >>   Num examples = 14614
+[INFO|trainer.py:3824] 2024-09-05 23:35:02,767 >>   Batch size = 8
+
  0%|          | 0/1827 [00:00<?, ?it/s]
  1%|          | 10/1827 [00:00<00:19, 90.92it/s]
  1%|          | 20/1827 [00:00<00:22, 80.87it/s]
  2%|▏         | 29/1827 [00:00<00:21, 82.68it/s]
  2%|▏         | 38/1827 [00:00<00:21, 81.95it/s]
  3%|▎         | 47/1827 [00:00<00:21, 82.63it/s]
  3%|▎         | 56/1827 [00:00<00:21, 82.47it/s]
  4%|▎         | 65/1827 [00:00<00:21, 80.47it/s]
  4%|▍         | 74/1827 [00:00<00:21, 81.66it/s]
  5%|▍         | 83/1827 [00:01<00:21, 82.68it/s]
  5%|▌         | 92/1827 [00:01<00:21, 82.58it/s]
  6%|▌         | 101/1827 [00:01<00:20, 84.52it/s]
  6%|▌         | 110/1827 [00:01<00:20, 85.29it/s]
  7%|▋         | 119/1827 [00:01<00:20, 83.68it/s]
  7%|▋         | 128/1827 [00:01<00:20, 82.79it/s]
  7%|▋         | 137/1827 [00:01<00:20, 83.78it/s]
  8%|▊         | 146/1827 [00:01<00:19, 84.13it/s]
  8%|▊         | 155/1827 [00:01<00:19, 83.64it/s]
  9%|▉         | 164/1827 [00:01<00:20, 81.81it/s]
  9%|▉         | 173/1827 [00:02<00:19, 82.80it/s]
 10%|▉         | 182/1827 [00:02<00:19, 84.13it/s]
 10%|█         | 191/1827 [00:02<00:19, 84.70it/s]
 11%|█         | 200/1827 [00:02<00:19, 83.62it/s]
 11%|█▏        | 209/1827 [00:02<00:19, 84.42it/s]
 12%|█▏        | 218/1827 [00:02<00:19, 83.10it/s]
 12%|█▏        | 227/1827 [00:02<00:19, 83.75it/s]
 13%|█▎        | 236/1827 [00:02<00:19, 80.40it/s]
 13%|█▎        | 245/1827 [00:02<00:19, 81.38it/s]
 14%|█▍        | 254/1827 [00:03<00:19, 81.71it/s]
 14%|█▍        | 263/1827 [00:03<00:19, 80.90it/s]
 15%|█▍        | 272/1827 [00:03<00:18, 83.07it/s]
 15%|█▌        | 281/1827 [00:03<00:18, 84.44it/s]
 16%|█▌        | 290/1827 [00:03<00:18, 83.94it/s]
 16%|█▋        | 299/1827 [00:03<00:18, 84.45it/s]
 17%|█▋        | 308/1827 [00:03<00:18, 83.65it/s]
 17%|█▋        | 317/1827 [00:03<00:18, 83.14it/s]
 18%|█▊        | 326/1827 [00:03<00:17, 84.05it/s]
 18%|█▊        | 335/1827 [00:04<00:17, 83.67it/s]
 19%|█▉        | 344/1827 [00:04<00:17, 84.53it/s]
 19%|█▉        | 353/1827 [00:04<00:18, 79.63it/s]
 20%|█▉        | 362/1827 [00:04<00:18, 80.95it/s]
 20%|██        | 371/1827 [00:04<00:17, 82.08it/s]
 21%|██        | 380/1827 [00:04<00:17, 83.24it/s]
 21%|██▏       | 389/1827 [00:04<00:17, 82.54it/s]
 22%|██▏       | 398/1827 [00:04<00:17, 81.91it/s]
 22%|██▏       | 407/1827 [00:04<00:17, 83.46it/s]
 23%|██▎       | 416/1827 [00:05<00:17, 81.20it/s]
 23%|██▎       | 425/1827 [00:05<00:17, 82.39it/s]
 24%|██▍       | 434/1827 [00:05<00:17, 81.62it/s]
 24%|██▍       | 443/1827 [00:05<00:17, 80.18it/s]
 25%|██▍       | 452/1827 [00:05<00:16, 82.39it/s]
 25%|██▌       | 461/1827 [00:05<00:16, 83.63it/s]
 26%|██▌       | 470/1827 [00:05<00:16, 83.17it/s]
 26%|██▌       | 479/1827 [00:05<00:16, 81.50it/s]
 27%|██▋       | 488/1827 [00:05<00:16, 81.09it/s]
 27%|██▋       | 497/1827 [00:06<00:16, 80.25it/s]
 28%|██▊       | 506/1827 [00:06<00:16, 81.31it/s]
 28%|██▊       | 515/1827 [00:06<00:15, 82.50it/s]
 29%|██▊       | 524/1827 [00:06<00:15, 84.16it/s]
 29%|██▉       | 533/1827 [00:06<00:15, 83.22it/s]
 30%|██▉       | 542/1827 [00:06<00:15, 84.11it/s]
 30%|███       | 551/1827 [00:06<00:15, 83.08it/s]
 31%|███       | 560/1827 [00:06<00:15, 82.78it/s]
 31%|███       | 569/1827 [00:06<00:15, 83.19it/s]
 32%|███▏      | 578/1827 [00:06<00:15, 82.89it/s]
 32%|███▏      | 587/1827 [00:07<00:14, 82.68it/s]
 33%|███▎      | 596/1827 [00:07<00:14, 82.70it/s]
 33%|███▎      | 605/1827 [00:07<00:14, 82.71it/s]
 34%|███▎      | 614/1827 [00:07<00:14, 82.53it/s]
 34%|███▍      | 623/1827 [00:07<00:14, 83.35it/s]
 35%|███▍      | 632/1827 [00:07<00:14, 83.93it/s]
 35%|███▌      | 641/1827 [00:07<00:14, 84.33it/s]
 36%|███▌      | 650/1827 [00:07<00:14, 79.40it/s]
 36%|███▌      | 659/1827 [00:07<00:14, 79.16it/s]
 37%|███▋      | 668/1827 [00:08<00:14, 80.89it/s]
 37%|███▋      | 677/1827 [00:08<00:14, 78.78it/s]
 38%|███▊      | 686/1827 [00:08<00:14, 80.15it/s]
 38%|███▊      | 695/1827 [00:08<00:13, 81.68it/s]
 39%|███▊      | 704/1827 [00:08<00:14, 78.28it/s]
 39%|███▉      | 713/1827 [00:08<00:13, 80.50it/s]
 40%|███▉      | 722/1827 [00:08<00:13, 82.68it/s]
 40%|████      | 731/1827 [00:08<00:13, 83.49it/s]
 41%|████      | 740/1827 [00:08<00:13, 82.99it/s]
 41%|████      | 749/1827 [00:09<00:12, 84.43it/s]
 41%|████▏     | 758/1827 [00:09<00:12, 84.86it/s]
 42%|████▏     | 767/1827 [00:09<00:12, 85.25it/s]
 42%|████▏     | 776/1827 [00:09<00:12, 85.62it/s]
 43%|████▎     | 785/1827 [00:09<00:12, 82.78it/s]
 43%|████▎     | 794/1827 [00:09<00:12, 84.06it/s]
 44%|████▍     | 803/1827 [00:09<00:12, 83.01it/s]
 44%|████▍     | 812/1827 [00:09<00:12, 83.78it/s]
 45%|████▍     | 821/1827 [00:09<00:11, 84.28it/s]
 45%|████▌     | 830/1827 [00:10<00:11, 84.61it/s]
 46%|████▌     | 839/1827 [00:10<00:11, 83.72it/s]
 46%|████▋     | 848/1827 [00:10<00:11, 85.15it/s]
 47%|████▋     | 857/1827 [00:10<00:11, 85.42it/s]
 47%|████▋     | 866/1827 [00:10<00:11, 85.71it/s]
 48%|████▊     | 875/1827 [00:10<00:11, 86.06it/s]
 48%|████▊     | 884/1827 [00:10<00:10, 86.43it/s]
 49%|████▉     | 893/1827 [00:10<00:10, 85.57it/s]
 49%|████▉     | 902/1827 [00:10<00:10, 86.11it/s]
 50%|████▉     | 911/1827 [00:10<00:10, 86.50it/s]
 50%|█████     | 920/1827 [00:11<00:10, 86.07it/s]
 51%|█████     | 929/1827 [00:11<00:10, 86.11it/s]
 51%|█████▏    | 938/1827 [00:11<00:10, 82.86it/s]
 52%|█████▏    | 947/1827 [00:11<00:10, 81.36it/s]
 52%|█████▏    | 956/1827 [00:11<00:10, 82.29it/s]
 53%|█████▎    | 965/1827 [00:11<00:10, 83.27it/s]
 53%|█████▎    | 974/1827 [00:11<00:10, 83.79it/s]
 54%|█████▍    | 983/1827 [00:11<00:10, 83.54it/s]
 54%|█████▍    | 992/1827 [00:11<00:10, 83.43it/s]
 55%|█████▍    | 1001/1827 [00:12<00:09, 84.18it/s]
 55%|█████▌    | 1010/1827 [00:12<00:09, 83.88it/s]
 56%|█████▌    | 1019/1827 [00:12<00:09, 84.69it/s]
 56%|█████▋    | 1028/1827 [00:12<00:09, 85.70it/s]
 57%|█████▋    | 1037/1827 [00:12<00:09, 84.01it/s]
 57%|█████▋    | 1046/1827 [00:12<00:09, 84.57it/s]
 58%|█████▊    | 1055/1827 [00:12<00:09, 84.49it/s]
 58%|████��▊    | 1064/1827 [00:12<00:09, 84.45it/s]
 59%|█████▊    | 1073/1827 [00:12<00:08, 85.10it/s]
 59%|█████▉    | 1082/1827 [00:13<00:08, 85.50it/s]
 60%|█████▉    | 1091/1827 [00:13<00:08, 85.18it/s]
 60%|██████    | 1100/1827 [00:13<00:08, 85.85it/s]
 61%|██████    | 1109/1827 [00:13<00:08, 85.34it/s]
 61%|██████    | 1118/1827 [00:13<00:08, 83.96it/s]
 62%|██████▏   | 1127/1827 [00:13<00:08, 83.91it/s]
 62%|██████▏   | 1136/1827 [00:13<00:08, 84.94it/s]
 63%|██████▎   | 1145/1827 [00:13<00:08, 84.62it/s]
 63%|██████▎   | 1154/1827 [00:13<00:07, 84.47it/s]
 64%|██████▎   | 1163/1827 [00:13<00:08, 82.48it/s]
 64%|██████▍   | 1172/1827 [00:14<00:07, 83.17it/s]
 65%|██████▍   | 1181/1827 [00:14<00:07, 81.16it/s]
 65%|██████▌   | 1190/1827 [00:14<00:07, 83.08it/s]
 66%|██████▌   | 1199/1827 [00:14<00:07, 83.48it/s]
 66%|██████▌   | 1208/1827 [00:14<00:07, 84.04it/s]
 67%|██████▋   | 1217/1827 [00:14<00:07, 82.40it/s]
 67%|██████▋   | 1226/1827 [00:14<00:07, 82.82it/s]
 68%|██████▊   | 1235/1827 [00:14<00:07, 83.66it/s]
 68%|██████▊   | 1244/1827 [00:14<00:06, 84.34it/s]
 69%|██████▊   | 1253/1827 [00:15<00:07, 81.83it/s]
 69%|██████▉   | 1262/1827 [00:15<00:06, 81.92it/s]
 70%|██████▉   | 1271/1827 [00:15<00:06, 82.56it/s]
 70%|███████   | 1280/1827 [00:15<00:06, 83.63it/s]
 71%|███████   | 1289/1827 [00:15<00:06, 84.21it/s]
 71%|███████   | 1298/1827 [00:15<00:06, 85.46it/s]
 72%|███████▏  | 1307/1827 [00:15<00:06, 85.59it/s]
 72%|███████▏  | 1316/1827 [00:15<00:05, 86.19it/s]
 73%|███████▎  | 1325/1827 [00:15<00:05, 85.74it/s]
 73%|███████▎  | 1334/1827 [00:16<00:05, 86.33it/s]
 74%|███████▎  | 1343/1827 [00:16<00:05, 84.87it/s]
 74%|███████▍  | 1352/1827 [00:16<00:05, 85.11it/s]
 74%|███████▍  | 1361/1827 [00:16<00:05, 85.88it/s]
 75%|███████▍  | 1370/1827 [00:16<00:05, 85.65it/s]
 75%|███████▌  | 1379/1827 [00:16<00:05, 85.85it/s]
 76%|███████▌  | 1388/1827 [00:16<00:05, 85.76it/s]
 76%|███████▋  | 1397/1827 [00:16<00:05, 85.56it/s]
 77%|███████▋  | 1406/1827 [00:16<00:04, 85.93it/s]
 77%|███████▋  | 1415/1827 [00:16<00:04, 85.21it/s]
 78%|███████▊  | 1424/1827 [00:17<00:04, 84.33it/s]
 78%|███████▊  | 1433/1827 [00:17<00:04, 83.44it/s]
 79%|███████▉  | 1442/1827 [00:17<00:04, 81.50it/s]
 79%|███████▉  | 1451/1827 [00:17<00:04, 82.65it/s]
 80%|███████▉  | 1460/1827 [00:17<00:04, 83.66it/s]
 80%|████████  | 1469/1827 [00:17<00:04, 81.75it/s]
 81%|████████  | 1478/1827 [00:17<00:04, 81.42it/s]
 81%|████████▏ | 1487/1827 [00:17<00:04, 82.24it/s]
 82%|████████▏ | 1496/1827 [00:17<00:04, 80.08it/s]
 82%|████████▏ | 1505/1827 [00:18<00:03, 80.52it/s]
 83%|████████▎ | 1514/1827 [00:18<00:03, 82.13it/s]
 83%|████████▎ | 1523/1827 [00:18<00:03, 82.92it/s]
 84%|████████▍ | 1532/1827 [00:18<00:03, 82.73it/s]
 84%|████████▍ | 1541/1827 [00:18<00:03, 83.28it/s]
 85%|████████▍ | 1550/1827 [00:18<00:03, 82.29it/s]
 85%|████████▌ | 1559/1827 [00:18<00:03, 82.49it/s]
 86%|████████▌ | 1568/1827 [00:18<00:03, 82.39it/s]
 86%|████████▋ | 1577/1827 [00:18<00:03, 82.15it/s]
 87%|████████▋ | 1586/1827 [00:19<00:02, 82.43it/s]
 87%|████████▋ | 1595/1827 [00:19<00:02, 81.93it/s]
 88%|████████▊ | 1604/1827 [00:19<00:02, 82.96it/s]
 88%|████████▊ | 1613/1827 [00:19<00:02, 83.75it/s]
 89%|████████▉ | 1622/1827 [00:19<00:02, 78.81it/s]
 89%|████████▉ | 1630/1827 [00:19<00:02, 76.60it/s]
 90%|████████▉ | 1638/1827 [00:19<00:02, 77.25it/s]
 90%|█████████ | 1647/1827 [00:19<00:02, 79.86it/s]
 91%|█████████ | 1656/1827 [00:19<00:02, 81.63it/s]
 91%|█████████ | 1665/1827 [00:20<00:01, 81.29it/s]
 92%|█████████▏| 1674/1827 [00:20<00:01, 82.22it/s]
 92%|█████████▏| 1683/1827 [00:20<00:01, 84.30it/s]
 93%|█████████▎| 1692/1827 [00:20<00:01, 82.83it/s]
 93%|█████████▎| 1701/1827 [00:20<00:01, 83.31it/s]
 94%|█████████▎| 1710/1827 [00:20<00:01, 84.27it/s]
 94%|█████████▍| 1719/1827 [00:20<00:01, 81.37it/s]
 95%|█████████▍| 1728/1827 [00:20<00:01, 82.82it/s]
 95%|█████████▌| 1737/1827 [00:20<00:01, 83.49it/s]
 96%|█████████▌| 1746/1827 [00:20<00:00, 84.78it/s]
 96%|█████████▌| 1755/1827 [00:21<00:00, 85.12it/s]
 97%|█████████▋| 1764/1827 [00:21<00:00, 81.96it/s]
 97%|█████████▋| 1773/1827 [00:21<00:00, 82.90it/s]
 98%|█████████▊| 1782/1827 [00:21<00:00, 84.18it/s]
 98%|█████████▊| 1791/1827 [00:21<00:00, 84.49it/s]
 99%|█████████▊| 1800/1827 [00:21<00:00, 81.26it/s]
 99%|█████████▉| 1809/1827 [00:21<00:00, 82.04it/s]
+[INFO|trainer.py:3503] 2024-09-05 23:35:31,629 >> Saving model checkpoint to /content/dissertation/scripts/ner/output
+[INFO|configuration_utils.py:472] 2024-09-05 23:35:31,631 >> Configuration saved in /content/dissertation/scripts/ner/output/config.json
+[INFO|modeling_utils.py:2799] 2024-09-05 23:35:32,981 >> Model weights saved in /content/dissertation/scripts/ner/output/model.safetensors
+[INFO|tokenization_utils_base.py:2684] 2024-09-05 23:35:32,982 >> tokenizer config file saved in /content/dissertation/scripts/ner/output/tokenizer_config.json
+[INFO|tokenization_utils_base.py:2693] 2024-09-05 23:35:32,982 >> Special tokens file saved in /content/dissertation/scripts/ner/output/special_tokens_map.json
+***** predict metrics *****
+  predict_accuracy           =     0.9987
+  predict_f1                 =     0.9243
+  predict_loss               =     0.0059
+  predict_precision          =      0.907
+  predict_recall             =     0.9423
+  predict_runtime            = 0:00:28.21
+  predict_samples_per_second =    517.899
+  predict_steps_per_second   =     64.746
+

train_results.json

@@ -1,9 +1,9 @@
 {
-    "epoch": 9.98998998998999,
-    "total_flos": 1.5071241212671032e+16,
-    "train_loss": 0.022475468706272407,
-    "train_runtime": 1385.1143,
-    "train_samples": 31947,
-    "train_samples_per_second": 230.645,
-    "train_steps_per_second": 3.603
+    "epoch": 10.0,
+    "total_flos": 1.4714840952259542e+16,
+    "train_loss": 0.002772659832779558,
+    "train_runtime": 1349.0548,
+    "train_samples": 29797,
+    "train_samples_per_second": 220.873,
+    "train_steps_per_second": 3.454
 }

trainer_state.json

@@ -1,208 +1,208 @@
 {
-  "best_metric": 0.8040201005025126,
-  "best_model_checkpoint": "/content/dissertation/scripts/ner/output/checkpoint-3996",
-  "epoch": 9.98998998998999,
+  "best_metric": 0.9491525423728814,
+  "best_model_checkpoint": "/content/dissertation/scripts/ner/output/checkpoint-2330",
+  "epoch": 10.0,
   "eval_steps": 500,
-  "global_step": 4990,
+  "global_step": 4660,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.998998998998999,
-      "eval_accuracy": 0.9730869045820918,
-      "eval_f1": 0.75963794837412,
-      "eval_loss": 0.07385822385549545,
-      "eval_precision": 0.7270588235294118,
-      "eval_recall": 0.7952737482452036,
-      "eval_runtime": 14.2023,
-      "eval_samples_per_second": 479.501,
-      "eval_steps_per_second": 59.99,
-      "step": 499
-    },
-    {
-      "epoch": 1.001001001001001,
-      "grad_norm": 0.8427119851112366,
-      "learning_rate": 4.4989979959919844e-05,
-      "loss": 0.105,
+      "epoch": 1.0,
+      "eval_accuracy": 0.9989495654084337,
+      "eval_f1": 0.9351598173515981,
+      "eval_loss": 0.0030670168343931437,
+      "eval_precision": 0.9292196007259528,
+      "eval_recall": 0.9411764705882353,
+      "eval_runtime": 13.9946,
+      "eval_samples_per_second": 486.615,
+      "eval_steps_per_second": 60.88,
+      "step": 466
+    },
+    {
+      "epoch": 1.0729613733905579,
+      "grad_norm": 0.07174628973007202,
+      "learning_rate": 4.4635193133047216e-05,
+      "loss": 0.0199,
       "step": 500
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.972949592870776,
-      "eval_f1": 0.7655806561471223,
-      "eval_loss": 0.09075114130973816,
-      "eval_precision": 0.7435501653803749,
-      "eval_recall": 0.7889564810481984,
-      "eval_runtime": 14.1492,
-      "eval_samples_per_second": 481.299,
-      "eval_steps_per_second": 60.215,
-      "step": 999
-    },
-    {
-      "epoch": 2.002002002002002,
-      "grad_norm": 0.9121108651161194,
-      "learning_rate": 3.997995991983968e-05,
-      "loss": 0.0448,
+      "eval_accuracy": 0.9989358342373021,
+      "eval_f1": 0.9386834986474301,
+      "eval_loss": 0.0030621723271906376,
+      "eval_precision": 0.9212389380530973,
+      "eval_recall": 0.9568014705882353,
+      "eval_runtime": 14.0045,
+      "eval_samples_per_second": 486.272,
+      "eval_steps_per_second": 60.838,
+      "step": 932
+    },
+    {
+      "epoch": 2.1459227467811157,
+      "grad_norm": 0.13906870782375336,
+      "learning_rate": 3.927038626609442e-05,
+      "loss": 0.0026,
       "step": 1000
     },
     {
-      "epoch": 2.998998998998999,
-      "eval_accuracy": 0.9743913658395924,
-      "eval_f1": 0.7829875042989798,
-      "eval_loss": 0.09297410398721695,
-      "eval_precision": 0.7675882220723759,
-      "eval_recall": 0.7990173139915769,
-      "eval_runtime": 14.1376,
-      "eval_samples_per_second": 481.693,
-      "eval_steps_per_second": 60.265,
-      "step": 1498
+      "epoch": 3.0,
+      "eval_accuracy": 0.9989152374806047,
+      "eval_f1": 0.9360919540229885,
+      "eval_loss": 0.004003152716904879,
+      "eval_precision": 0.9365225390984361,
+      "eval_recall": 0.9356617647058824,
+      "eval_runtime": 13.9451,
+      "eval_samples_per_second": 488.343,
+      "eval_steps_per_second": 61.097,
+      "step": 1398
     },
     {
-      "epoch": 3.003003003003003,
-      "grad_norm": 0.5380845069885254,
-      "learning_rate": 3.496993987975952e-05,
-      "loss": 0.0255,
+      "epoch": 3.218884120171674,
+      "grad_norm": 0.14111244678497314,
+      "learning_rate": 3.3905579399141636e-05,
+      "loss": 0.0011,
       "step": 1500
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.9757438861960537,
-      "eval_f1": 0.789358010410642,
-      "eval_loss": 0.10520397126674652,
-      "eval_precision": 0.7805994051704416,
-      "eval_recall": 0.798315395414132,
-      "eval_runtime": 14.3621,
-      "eval_samples_per_second": 474.163,
-      "eval_steps_per_second": 59.323,
-      "step": 1998
-    },
-    {
-      "epoch": 4.004004004004004,
-      "grad_norm": 0.2705754339694977,
-      "learning_rate": 2.9959919839679363e-05,
-      "loss": 0.0164,
+      "eval_accuracy": 0.9987230010847625,
+      "eval_f1": 0.9308584686774942,
+      "eval_loss": 0.005216046702116728,
+      "eval_precision": 0.9400187441424555,
+      "eval_recall": 0.921875,
+      "eval_runtime": 14.2384,
+      "eval_samples_per_second": 478.283,
+      "eval_steps_per_second": 59.838,
+      "step": 1864
+    },
+    {
+      "epoch": 4.291845493562231,
+      "grad_norm": 0.09000600874423981,
+      "learning_rate": 2.8540772532188842e-05,
+      "loss": 0.001,
       "step": 2000
     },
     {
-      "epoch": 4.998998998998999,
-      "eval_accuracy": 0.9750435964683428,
-      "eval_f1": 0.7879346074142298,
-      "eval_loss": 0.10997848957777023,
-      "eval_precision": 0.7756119673617408,
-      "eval_recall": 0.8006551240056153,
-      "eval_runtime": 14.2341,
-      "eval_samples_per_second": 478.428,
-      "eval_steps_per_second": 59.856,
-      "step": 2497
+      "epoch": 5.0,
+      "eval_accuracy": 0.9989426998228679,
+      "eval_f1": 0.9491525423728814,
+      "eval_loss": 0.004777050111442804,
+      "eval_precision": 0.9461187214611873,
+      "eval_recall": 0.9522058823529411,
+      "eval_runtime": 13.9397,
+      "eval_samples_per_second": 488.533,
+      "eval_steps_per_second": 61.12,
+      "step": 2330
     },
     {
-      "epoch": 5.005005005005005,
-      "grad_norm": 0.27666428685188293,
-      "learning_rate": 2.49498997995992e-05,
-      "loss": 0.0112,
+      "epoch": 5.364806866952789,
+      "grad_norm": 0.0045097870752215385,
+      "learning_rate": 2.3175965665236052e-05,
+      "loss": 0.0005,
       "step": 2500
     },
     {
       "epoch": 6.0,
-      "eval_accuracy": 0.9768011863731858,
-      "eval_f1": 0.799447386599125,
-      "eval_loss": 0.12663568556308746,
-      "eval_precision": 0.7869446962828649,
-      "eval_recall": 0.8123537669630323,
-      "eval_runtime": 14.5018,
-      "eval_samples_per_second": 469.597,
-      "eval_steps_per_second": 58.751,
-      "step": 2997
-    },
-    {
-      "epoch": 6.006006006006006,
-      "grad_norm": 0.24552026391029358,
-      "learning_rate": 1.993987975951904e-05,
-      "loss": 0.0073,
+      "eval_accuracy": 0.9988809095527758,
+      "eval_f1": 0.9448244414044688,
+      "eval_loss": 0.004644877277314663,
+      "eval_precision": 0.9375565610859729,
+      "eval_recall": 0.9522058823529411,
+      "eval_runtime": 14.0259,
+      "eval_samples_per_second": 485.531,
+      "eval_steps_per_second": 60.745,
+      "step": 2796
+    },
+    {
+      "epoch": 6.437768240343348,
+      "grad_norm": 0.029634617269039154,
+      "learning_rate": 1.7811158798283262e-05,
+      "loss": 0.0004,
       "step": 3000
     },
     {
-      "epoch": 6.998998998998999,
-      "eval_accuracy": 0.976293133041317,
-      "eval_f1": 0.7968804562914678,
-      "eval_loss": 0.12882493436336517,
-      "eval_precision": 0.792911744266852,
-      "eval_recall": 0.8008890968647637,
-      "eval_runtime": 14.3476,
-      "eval_samples_per_second": 474.643,
-      "eval_steps_per_second": 59.383,
-      "step": 3496
+      "epoch": 7.0,
+      "eval_accuracy": 0.9989564309939994,
+      "eval_f1": 0.9446460980036298,
+      "eval_loss": 0.0049773636274039745,
+      "eval_precision": 0.9327956989247311,
+      "eval_recall": 0.9568014705882353,
+      "eval_runtime": 13.9217,
+      "eval_samples_per_second": 489.166,
+      "eval_steps_per_second": 61.2,
+      "step": 3262
     },
     {
-      "epoch": 7.007007007007007,
-      "grad_norm": 0.2008085697889328,
-      "learning_rate": 1.492985971943888e-05,
-      "loss": 0.0054,
+      "epoch": 7.510729613733906,
+      "grad_norm": 0.003798937890678644,
+      "learning_rate": 1.2446351931330473e-05,
+      "loss": 0.0002,
       "step": 3500
     },
     {
       "epoch": 8.0,
-      "eval_accuracy": 0.9764853694371592,
-      "eval_f1": 0.8040201005025126,
-      "eval_loss": 0.14244574308395386,
-      "eval_precision": 0.803175344384777,
-      "eval_recall": 0.8048666354702855,
-      "eval_runtime": 14.514,
-      "eval_samples_per_second": 469.202,
-      "eval_steps_per_second": 58.702,
-      "step": 3996
-    },
-    {
-      "epoch": 8.008008008008009,
-      "grad_norm": 0.12597906589508057,
-      "learning_rate": 9.919839679358718e-06,
-      "loss": 0.0038,
+      "eval_accuracy": 0.9989221030661705,
+      "eval_f1": 0.9435520881138136,
+      "eval_loss": 0.005484889727085829,
+      "eval_precision": 0.9422548120989918,
+      "eval_recall": 0.9448529411764706,
+      "eval_runtime": 13.8923,
+      "eval_samples_per_second": 490.2,
+      "eval_steps_per_second": 61.329,
+      "step": 3728
+    },
+    {
+      "epoch": 8.583690987124463,
+      "grad_norm": 0.0004711664514616132,
+      "learning_rate": 7.0815450643776825e-06,
+      "loss": 0.0001,
       "step": 4000
     },
     {
-      "epoch": 8.998998998999,
-      "eval_accuracy": 0.9765059661938567,
-      "eval_f1": 0.7970779220779219,
-      "eval_loss": 0.14552859961986542,
-      "eval_precision": 0.7901149425287356,
-      "eval_recall": 0.8041647168928404,
-      "eval_runtime": 14.2396,
-      "eval_samples_per_second": 478.242,
-      "eval_steps_per_second": 59.833,
-      "step": 4495
+      "epoch": 9.0,
+      "eval_accuracy": 0.9989426998228679,
+      "eval_f1": 0.9441903019213176,
+      "eval_loss": 0.0057435426861047745,
+      "eval_precision": 0.9398907103825137,
+      "eval_recall": 0.9485294117647058,
+      "eval_runtime": 13.9755,
+      "eval_samples_per_second": 487.281,
+      "eval_steps_per_second": 60.964,
+      "step": 4194
     },
     {
-      "epoch": 9.00900900900901,
-      "grad_norm": 0.2577208876609802,
-      "learning_rate": 4.9098196392785576e-06,
-      "loss": 0.0028,
+      "epoch": 9.656652360515022,
+      "grad_norm": 0.005946693476289511,
+      "learning_rate": 1.7167381974248929e-06,
+      "loss": 0.0001,
       "step": 4500
     },
     {
-      "epoch": 9.98998998998999,
-      "eval_accuracy": 0.9768286487154489,
-      "eval_f1": 0.7984262902105993,
-      "eval_loss": 0.14972682297229767,
-      "eval_precision": 0.7898351648351648,
-      "eval_recall": 0.8072063640617688,
-      "eval_runtime": 14.3927,
-      "eval_samples_per_second": 473.158,
-      "eval_steps_per_second": 59.197,
-      "step": 4990
-    },
-    {
-      "epoch": 9.98998998998999,
-      "step": 4990,
-      "total_flos": 1.5071241212671032e+16,
-      "train_loss": 0.022475468706272407,
-      "train_runtime": 1385.1143,
-      "train_samples_per_second": 230.645,
-      "train_steps_per_second": 3.603
+      "epoch": 10.0,
+      "eval_accuracy": 0.9989083718950389,
+      "eval_f1": 0.9416058394160585,
+      "eval_loss": 0.005793666001409292,
+      "eval_precision": 0.9347826086956522,
+      "eval_recall": 0.9485294117647058,
+      "eval_runtime": 14.4595,
+      "eval_samples_per_second": 470.971,
+      "eval_steps_per_second": 58.923,
+      "step": 4660
+    },
+    {
+      "epoch": 10.0,
+      "step": 4660,
+      "total_flos": 1.4714840952259542e+16,
+      "train_loss": 0.002772659832779558,
+      "train_runtime": 1349.0548,
+      "train_samples_per_second": 220.873,
+      "train_steps_per_second": 3.454
     }
   ],
   "logging_steps": 500,
-  "max_steps": 4990,
+  "max_steps": 4660,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 10,
   "save_steps": 500,
@@ -218,7 +218,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 1.5071241212671032e+16,
+  "total_flos": 1.4714840952259542e+16,
   "train_batch_size": 32,
   "trial_name": null,
   "trial_params": null