End of training

README.md

@@ -2,9 +2,10 @@
 license: apache-2.0
 base_model: PlanTL-GOB-ES/bsc-bio-ehr-es
 tags:
+- token-classification
 - generated_from_trainer
 datasets:
-- multi-train-drugtemist-dev-ner
+- Rodrigo1771/multi-train-drugtemist-dev-ner
 metrics:
 - precision
 - recall
@@ -17,24 +18,24 @@ model-index:
       name: Token Classification
       type: token-classification
     dataset:
-      name: multi-train-drugtemist-dev-ner
-      type: multi-train-drugtemist-dev-ner
+      name: Rodrigo1771/multi-train-drugtemist-dev-ner
+      type: Rodrigo1771/multi-train-drugtemist-dev-ner
       config: MultiTrainDrugTEMISTDevNER
       split: validation
       args: MultiTrainDrugTEMISTDevNER
     metrics:
     - name: Precision
       type: precision
-      value: 0.09270693512304251
+      value: 0.09691960931630353
     - name: Recall
       type: recall
-      value: 0.9522058823529411
+      value: 0.9485294117647058
     - name: F1
       type: f1
-      value: 0.16896354888689555
+      value: 0.17586912065439672
     - name: Accuracy
       type: accuracy
-      value: 0.7845534874460183
+      value: 0.8099635429897495
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -42,13 +43,13 @@ 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 multi-train-drugtemist-dev-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/multi-train-drugtemist-dev-ner dataset.
 It achieves the following results on the evaluation set:
-- Loss: 1.7861
-- Precision: 0.0927
-- Recall: 0.9522
-- F1: 0.1690
-- Accuracy: 0.7846
+- Loss: 0.6631
+- Precision: 0.0969
+- Recall: 0.9485
+- F1: 0.1759
+- Accuracy: 0.8100
 
 ## Model description
 

all_results.json

@@ -1,19 +1,26 @@
 {
-    "eval_accuracy": 0.002835545241707918,
-    "eval_f1": 0.0006942779922141681,
-    "eval_loss": 2.4030845165252686,
-    "eval_precision": 0.00035028898841544273,
-    "eval_recall": 0.03860294117647059,
-    "eval_runtime": 16.7962,
+    "epoch": 9.997061416397296,
+    "eval_accuracy": 0.8099635429897495,
+    "eval_f1": 0.17586912065439672,
+    "eval_loss": 0.6631014347076416,
+    "eval_precision": 0.09691960931630353,
+    "eval_recall": 0.9485294117647058,
+    "eval_runtime": 15.8157,
     "eval_samples": 6807,
-    "eval_samples_per_second": 405.27,
-    "eval_steps_per_second": 50.666,
-    "predict_accuracy": 0.002835545241707918,
-    "predict_f1": 0.0006942779922141681,
-    "predict_loss": 2.4030845165252686,
-    "predict_precision": 0.00035028898841544273,
-    "predict_recall": 0.03860294117647059,
-    "predict_runtime": 16.0715,
-    "predict_samples_per_second": 423.545,
-    "predict_steps_per_second": 52.951
+    "eval_samples_per_second": 430.395,
+    "eval_steps_per_second": 53.807,
+    "predict_accuracy": 0.8099635429897495,
+    "predict_f1": 0.17586912065439672,
+    "predict_loss": 0.6631014347076416,
+    "predict_precision": 0.09691960931630353,
+    "predict_recall": 0.9485294117647058,
+    "predict_runtime": 15.8984,
+    "predict_samples_per_second": 428.156,
+    "predict_steps_per_second": 53.527,
+    "total_flos": 6700722040732752.0,
+    "train_loss": 0.08913132946046923,
+    "train_runtime": 3337.712,
+    "train_samples": 27224,
+    "train_samples_per_second": 81.565,
+    "train_steps_per_second": 5.096
 }

eval_results.json

@@ -1,11 +1,12 @@
 {
-    "eval_accuracy": 0.002835545241707918,
-    "eval_f1": 0.0006942779922141681,
-    "eval_loss": 2.4030845165252686,
-    "eval_precision": 0.00035028898841544273,
-    "eval_recall": 0.03860294117647059,
-    "eval_runtime": 16.7962,
+    "epoch": 9.997061416397296,
+    "eval_accuracy": 0.8099635429897495,
+    "eval_f1": 0.17586912065439672,
+    "eval_loss": 0.6631014347076416,
+    "eval_precision": 0.09691960931630353,
+    "eval_recall": 0.9485294117647058,
+    "eval_runtime": 15.8157,
     "eval_samples": 6807,
-    "eval_samples_per_second": 405.27,
-    "eval_steps_per_second": 50.666
+    "eval_samples_per_second": 430.395,
+    "eval_steps_per_second": 53.807
 }

predict_results.json

@@ -1,10 +1,10 @@
 {
-    "predict_accuracy": 0.002835545241707918,
-    "predict_f1": 0.0006942779922141681,
-    "predict_loss": 2.4030845165252686,
-    "predict_precision": 0.00035028898841544273,
-    "predict_recall": 0.03860294117647059,
-    "predict_runtime": 16.0715,
-    "predict_samples_per_second": 423.545,
-    "predict_steps_per_second": 52.951
+    "predict_accuracy": 0.8099635429897495,
+    "predict_f1": 0.17586912065439672,
+    "predict_loss": 0.6631014347076416,
+    "predict_precision": 0.09691960931630353,
+    "predict_recall": 0.9485294117647058,
+    "predict_runtime": 15.8984,
+    "predict_samples_per_second": 428.156,
+    "predict_steps_per_second": 53.527
 }

tb/events.out.tfevents.1715612339.c331905616cf.3060.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fd631ac54bf6215fcba3db4eb9803c4beb0d2d937cbe6eea52ab0ff263ca8fae
+size 569

train.log

@@ -1613,3 +1613,53 @@ Training completed. Do not forget to share your model on huggingface.co/models =
 [INFO|modeling_utils.py:2590] 2024-05-13 14:58:40,302 >> Model weights saved in /content/dissertation/scripts/ner/output/model.safetensors
 [INFO|tokenization_utils_base.py:2488] 2024-05-13 14:58:40,303 >> tokenizer config file saved in /content/dissertation/scripts/ner/output/tokenizer_config.json
 [INFO|tokenization_utils_base.py:2497] 2024-05-13 14:58:40,303 >> Special tokens file saved in /content/dissertation/scripts/ner/output/special_tokens_map.json
+{'eval_loss': 1.7860842943191528, 'eval_precision': 0.09270693512304251, 'eval_recall': 0.9522058823529411, 'eval_f1': 0.16896354888689555, 'eval_accuracy': 0.7845534874460183, 'eval_runtime': 15.9319, 'eval_samples_per_second': 427.256, 'eval_steps_per_second': 53.415, 'epoch': 10.0}
+{'train_runtime': 3337.712, 'train_samples_per_second': 81.565, 'train_steps_per_second': 5.096, 'train_loss': 0.08913132946046923, 'epoch': 10.0}
+***** train metrics *****
+  epoch                    =     9.9971
+  total_flos               =  6240533GF
+  train_loss               =     0.0891
+  train_runtime            = 0:55:37.71
+  train_samples            =      27224
+  train_samples_per_second =     81.565
+  train_steps_per_second   =      5.096
+05/13/2024 14:58:43 - INFO - __main__ -   *** Evaluate ***
+[INFO|trainer.py:786] 2024-05-13 14:58:43,309 >> 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:3614] 2024-05-13 14:58:43,312 >> ***** Running Evaluation *****
+[INFO|trainer.py:3616] 2024-05-13 14:58:43,312 >>   Num examples = 6807
+[INFO|trainer.py:3619] 2024-05-13 14:58:43,312 >>   Batch size = 8
+
  0%|          | 0/851 [00:00<?, ?it/s]
  1%|          | 10/851 [00:00<00:09, 92.39it/s]
  2%|▏         | 20/851 [00:00<00:10, 78.68it/s]
  3%|▎         | 28/851 [00:00<00:10, 75.69it/s]
  4%|▍         | 36/851 [00:00<00:10, 74.36it/s]
  5%|▌         | 44/851 [00:00<00:10, 75.32it/s]
  6%|▌         | 52/851 [00:00<00:10, 75.93it/s]
  7%|▋         | 60/851 [00:00<00:10, 76.39it/s]
  8%|▊         | 68/851 [00:00<00:11, 70.86it/s]
  9%|▉         | 76/851 [00:01<00:11, 70.00it/s]
 10%|▉         | 84/851 [00:01<00:10, 70.49it/s]
 11%|█         | 92/851 [00:01<00:10, 70.55it/s]
 12%|█▏        | 100/851 [00:01<00:10, 70.25it/s]
 13%|█▎        | 108/851 [00:01<00:10, 69.63it/s]
 14%|█▎        | 115/851 [00:01<00:10, 69.25it/s]
 15%|█▍        | 124/851 [00:01<00:09, 73.89it/s]
 16%|█▌        | 132/851 [00:01<00:10, 66.17it/s]
 16%|█▋        | 139/851 [00:01<00:10, 66.18it/s]
 17%|█▋        | 147/851 [00:02<00:10, 67.87it/s]
 18%|█▊        | 154/851 [00:02<00:10, 67.52it/s]
 19%|█▉        | 162/851 [00:02<00:09, 69.68it/s]
 20%|█▉        | 170/851 [00:02<00:09, 71.19it/s]
 21%|██        | 178/851 [00:02<00:09, 73.27it/s]
 22%|██▏       | 186/851 [00:02<00:09, 73.02it/s]
 23%|██▎       | 194/851 [00:02<00:08, 74.19it/s]
 24%|██▍       | 203/851 [00:02<00:08, 76.36it/s]
 25%|██▍       | 211/851 [00:02<00:08, 71.75it/s]
 26%|██▌       | 219/851 [00:03<00:09, 69.67it/s]
 27%|██▋       | 227/851 [00:03<00:08, 71.38it/s]
 28%|██▊       | 235/851 [00:03<00:08, 72.05it/s]
 29%|██▊       | 243/851 [00:03<00:09, 67.00it/s]
 29%|██▉       | 251/851 [00:03<00:08, 69.80it/s]
 30%|███       | 259/851 [00:03<00:08, 72.22it/s]
 31%|███▏      | 267/851 [00:03<00:08, 72.74it/s]
 32%|███▏      | 275/851 [00:03<00:07, 73.62it/s]
 33%|███▎      | 283/851 [00:03<00:07, 74.61it/s]
 34%|███▍      | 291/851 [00:04<00:07, 72.54it/s]
 35%|███▌      | 299/851 [00:04<00:07, 74.26it/s]
 36%|███▌      | 307/851 [00:04<00:07, 75.74it/s]
 37%|███▋      | 315/851 [00:04<00:07, 70.27it/s]
 38%|███▊      | 323/851 [00:04<00:07, 72.85it/s]
 39%|███▉      | 331/851 [00:04<00:07, 72.36it/s]
 40%|███▉      | 339/851 [00:04<00:07, 71.89it/s]
 41%|████      | 347/851 [00:04<00:06, 72.58it/s]
 42%|████▏     | 355/851 [00:04<00:07, 69.70it/s]
 43%|████▎     | 363/851 [00:05<00:06, 69.83it/s]
 44%|████▎     | 371/851 [00:05<00:06, 69.06it/s]
 44%|████▍     | 378/851 [00:05<00:06, 69.30it/s]
 45%|████▌     | 385/851 [00:05<00:06, 68.06it/s]
 46%|████▌     | 393/851 [00:05<00:06, 69.52it/s]
 47%|████▋     | 401/851 [00:05<00:06, 69.90it/s]
 48%|████▊     | 408/851 [00:05<00:06, 64.81it/s]
 49%|████▉     | 416/851 [00:05<00:06, 67.34it/s]
 50%|████▉     | 424/851 [00:05<00:06, 69.53it/s]
 51%|█████     | 431/851 [00:06<00:06, 68.95it/s]
 52%|█████▏    | 439/851 [00:06<00:05, 71.11it/s]
 53%|█████▎    | 447/851 [00:06<00:05, 70.55it/s]
 53%|█████▎    | 455/851 [00:06<00:05, 71.50it/s]
 54%|█████▍    | 463/851 [00:06<00:05, 70.54it/s]
 55%|█████▌    | 471/851 [00:06<00:06, 61.54it/s]
 56%|█████▌    | 478/851 [00:06<00:05, 63.35it/s]
 57%|█████▋    | 485/851 [00:06<00:05, 63.12it/s]
 58%|█████▊    | 493/851 [00:06<00:05, 66.52it/s]
 59%|█████▉    | 502/851 [00:07<00:04, 70.10it/s]
 60%|█████▉    | 510/851 [00:07<00:04, 69.87it/s]
 61%|██████    | 518/851 [00:07<00:04, 71.30it/s]
 62%|██████▏   | 526/851 [00:07<00:04, 66.37it/s]
 63%|██████▎   | 534/851 [00:07<00:04, 68.15it/s]
 64%|██████▎   | 542/851 [00:07<00:04, 69.51it/s]
 65%|██████▍   | 550/851 [00:07<00:04, 67.52it/s]
 66%|██████▌   | 558/851 [00:07<00:04, 69.41it/s]
 67%|██████▋   | 566/851 [00:08<00:03, 72.24it/s]
 67%|██████▋   | 574/851 [00:08<00:03, 72.87it/s]
 68%|██████▊   | 582/851 [00:08<00:03, 69.95it/s]
 69%|██████▉   | 590/851 [00:08<00:03, 67.24it/s]
 70%|███████   | 597/851 [00:08<00:03, 67.77it/s]
 71%|███████   | 605/851 [00:08<00:03, 67.61it/s]
 72%|███████▏  | 612/851 [00:08<00:03, 66.37it/s]
 73%|███████▎  | 619/851 [00:08<00:03, 64.78it/s]
 74%|███████▎  | 626/851 [00:08<00:03, 64.41it/s]
 74%|███████▍  | 633/851 [00:09<00:03, 64.56it/s]
 75%|███████▌  | 640/851 [00:09<00:03, 64.37it/s]
 76%|███████▌  | 647/851 [00:09<00:03, 62.56it/s]
 77%|███████▋  | 655/851 [00:09<00:02, 65.73it/s]
 78%|███████▊  | 662/851 [00:09<00:02, 66.42it/s]
 79%|███████▊  | 670/851 [00:09<00:02, 67.98it/s]
 80%|███████▉  | 678/851 [00:09<00:02, 69.04it/s]
 81%|████████  | 686/851 [00:09<00:02, 69.89it/s]
 82%|████████▏ | 694/851 [00:09<00:02, 71.78it/s]
 82%|████████▏ | 702/851 [00:10<00:02, 72.47it/s]
 83%|████████▎ | 710/851 [00:10<00:01, 73.89it/s]
 84%|████████▍ | 718/851 [00:10<00:01, 70.91it/s]
 85%|████████▌ | 726/851 [00:10<00:01, 70.61it/s]
 86%|████████▋ | 734/851 [00:10<00:01, 72.17it/s]
 87%|████████▋ | 742/851 [00:10<00:01, 73.05it/s]
 88%|████████▊ | 750/851 [00:10<00:01, 72.59it/s]
 89%|████████▉ | 758/851 [00:10<00:01, 73.09it/s]
 90%|█████████ | 766/851 [00:10<00:01, 69.09it/s]
 91%|█████████ | 774/851 [00:11<00:01, 68.91it/s]
 92%|█████████▏| 781/851 [00:11<00:01, 66.57it/s]
 93%|█████████▎| 788/851 [00:11<00:00, 67.34it/s]
 94%|█████████▎| 796/851 [00:11<00:00, 69.12it/s]
 94%|█████████▍| 804/851 [00:11<00:00, 71.53it/s]
 95%|█████████▌| 812/851 [00:11<00:00, 69.03it/s]
 96%|█████████▋| 820/851 [00:11<00:00, 69.99it/s]
 97%|█████████▋| 828/851 [00:11<00:00, 71.40it/s]
 98%|█████████▊| 836/851 [00:11<00:00, 71.06it/s]
 99%|█████████▉| 844/851 [00:12<00:00, 67.57it/s]
+  _warn_prf(average, modifier, msg_start, len(result))
+
+***** eval metrics *****
+  epoch                   =     9.9971
+  eval_accuracy           =       0.81
+  eval_f1                 =     0.1759
+  eval_loss               =     0.6631
+  eval_precision          =     0.0969
+  eval_recall             =     0.9485
+  eval_runtime            = 0:00:15.81
+  eval_samples            =       6807
+  eval_samples_per_second =    430.395
+  eval_steps_per_second   =     53.807
+05/13/2024 14:58:59 - INFO - __main__ -   *** Predict ***
+[INFO|trainer.py:786] 2024-05-13 14:58:59,135 >> 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:3614] 2024-05-13 14:58:59,137 >> ***** Running Prediction *****
+[INFO|trainer.py:3616] 2024-05-13 14:58:59,137 >>   Num examples = 6807
+[INFO|trainer.py:3619] 2024-05-13 14:58:59,138 >>   Batch size = 8
+
  0%|          | 0/851 [00:00<?, ?it/s]
  1%|          | 10/851 [00:00<00:08, 94.43it/s]
  2%|▏         | 20/851 [00:00<00:10, 79.64it/s]
  3%|▎         | 29/851 [00:00<00:10, 75.61it/s]
  4%|▍         | 37/851 [00:00<00:11, 73.82it/s]
  5%|▌         | 45/851 [00:00<00:10, 73.71it/s]
  6%|▌         | 53/851 [00:00<00:10, 74.31it/s]
  7%|▋         | 62/851 [00:00<00:10, 75.34it/s]
  8%|▊         | 70/851 [00:00<00:11, 70.02it/s]
  9%|▉         | 78/851 [00:01<00:10, 71.64it/s]
 10%|█         | 86/851 [00:01<00:10, 71.26it/s]
 11%|█         | 94/851 [00:01<00:10, 69.42it/s]
 12%|█▏        | 102/851 [00:01<00:10, 70.77it/s]
 13%|█▎        | 110/851 [00:01<00:10, 69.03it/s]
 14%|█▍        | 118/851 [00:01<00:10, 70.01it/s]
 15%|█▍        | 126/851 [00:01<00:10, 67.17it/s]
 16%|█▌        | 133/851 [00:01<00:10, 67.65it/s]
 16%|█▋        | 140/851 [00:01<00:10, 67.39it/s]
 17%|█▋        | 148/851 [00:02<00:10, 66.64it/s]
 18%|█▊        | 156/851 [00:02<00:10, 68.76it/s]
 19%|█▉        | 164/851 [00:02<00:09, 69.74it/s]
 20%|██        | 172/851 [00:02<00:09, 71.66it/s]
 21%|██        | 180/851 [00:02<00:09, 72.19it/s]
 22%|██▏       | 188/851 [00:02<00:09, 72.67it/s]
 23%|██▎       | 196/851 [00:02<00:08, 73.59it/s]
 24%|██▍       | 204/851 [00:02<00:08, 75.18it/s]
 25%|██▍       | 212/851 [00:02<00:09, 70.71it/s]
 26%|██▌       | 220/851 [00:03<00:09, 68.53it/s]
 27%|██▋       | 228/851 [00:03<00:08, 69.85it/s]
 28%|██▊       | 236/851 [00:03<00:08, 69.98it/s]
 29%|██▊       | 244/851 [00:03<00:09, 66.03it/s]
 30%|██▉       | 252/851 [00:03<00:08, 69.13it/s]
 31%|███       | 261/851 [00:03<00:08, 72.34it/s]
 32%|███▏      | 269/851 [00:03<00:08, 71.00it/s]
 33%|███▎      | 277/851 [00:03<00:07, 73.30it/s]
 33%|███▎      | 285/851 [00:03<00:07, 74.56it/s]
 34%|███▍      | 293/851 [00:04<00:07, 72.43it/s]
 35%|███▌      | 301/851 [00:04<00:07, 73.31it/s]
 36%|███▋      | 310/851 [00:04<00:07, 75.84it/s]
 37%|███▋      | 318/851 [00:04<00:07, 71.54it/s]
 38%|███▊      | 326/851 [00:04<00:07, 71.08it/s]
 39%|███▉      | 334/851 [00:04<00:07, 71.61it/s]
 40%|████      | 342/851 [00:04<00:07, 72.39it/s]
 41%|████      | 350/851 [00:04<00:06, 71.99it/s]
 42%|████▏     | 358/851 [00:05<00:07, 67.28it/s]
 43%|████▎     | 366/851 [00:05<00:07, 67.95it/s]
 44%|████▍     | 373/851 [00:05<00:06, 68.31it/s]
 45%|████▍     | 380/851 [00:05<00:07, 66.41it/s]
 46%|████▌     | 388/851 [00:05<00:06, 69.17it/s]
 46%|████▋     | 395/851 [00:05<00:06, 69.34it/s]
 47%|████▋     | 402/851 [00:05<00:06, 69.06it/s]
 48%|████▊     | 409/851 [00:05<00:06, 65.01it/s]
 49%|████▉     | 417/851 [00:05<00:06, 67.26it/s]
 50%|████▉     | 425/851 [00:06<00:06, 66.34it/s]
 51%|█████     | 433/851 [00:06<00:06, 68.38it/s]
 52%|█████▏    | 441/851 [00:06<00:05, 70.44it/s]
 53%|█████▎    | 449/851 [00:06<00:05, 70.15it/s]
 54%|█████▎    | 457/851 [00:06<00:05, 71.36it/s]
 55%|█████▍    | 465/851 [00:06<00:05, 68.14it/s]
 55%|█████▌    | 472/851 [00:06<00:06, 61.77it/s]
 56%|█████▋    | 479/851 [00:06<00:05, 63.68it/s]
 57%|█████▋    | 486/851 [00:06<00:05, 62.95it/s]
 58%|█████▊    | 494/851 [00:07<00:05, 66.26it/s]
 59%|█████▉    | 502/851 [00:07<00:05, 69.52it/s]
 60%|█████▉    | 510/851 [00:07<00:04, 69.09it/s]
 61%|██████    | 518/851 [00:07<00:04, 70.28it/s]
 62%|██████▏   | 526/851 [00:07<00:04, 65.43it/s]
 63%|██████▎   | 534/851 [00:07<00:04, 67.57it/s]
 64%|██████▎   | 542/851 [00:07<00:04, 69.07it/s]
 65%|██████▍   | 549/851 [00:07<00:04, 66.40it/s]
 65%|██████▌   | 557/851 [00:07<00:04, 68.83it/s]
 67%|██████▋   | 566/851 [00:08<00:03, 72.25it/s]
 67%|██████▋   | 574/851 [00:08<00:03, 71.57it/s]
 68%|██████▊   | 582/851 [00:08<00:03, 69.26it/s]
 69%|██████▉   | 589/851 [00:08<00:03, 66.06it/s]
 70%|███████   | 596/851 [00:08<00:03, 67.01it/s]
 71%|███████   | 604/851 [00:08<00:03, 68.21it/s]
 72%|███████▏  | 611/851 [00:08<00:03, 67.43it/s]
 73%|███████▎  | 618/851 [00:08<00:03, 63.33it/s]
 74%|███████▎  | 626/851 [00:09<00:03, 64.04it/s]
 74%|███████▍  | 633/851 [00:09<00:03, 63.85it/s]
 75%|███████▌  | 640/851 [00:09<00:03, 64.00it/s]
 76%|███████▌  | 647/851 [00:09<00:03, 62.36it/s]
 77%|███████▋  | 655/851 [00:09<00:02, 65.58it/s]
 78%|███████▊  | 662/851 [00:09<00:02, 66.21it/s]
 79%|███████▊  | 669/851 [00:09<00:02, 67.09it/s]
 79%|███████▉  | 676/851 [00:09<00:02, 66.93it/s]
 80%|████████  | 683/851 [00:09<00:02, 67.45it/s]
 81%|████████  | 691/851 [00:09<00:02, 70.83it/s]
 82%|████████▏ | 699/851 [00:10<00:02, 70.91it/s]
 83%|████████▎ | 708/851 [00:10<00:01, 74.14it/s]
 84%|████████▍ | 716/851 [00:10<00:01, 71.95it/s]
 85%|████████▌ | 724/851 [00:10<00:01, 72.30it/s]
 86%|████████▌ | 732/851 [00:10<00:01, 73.78it/s]
 87%|████████▋ | 740/851 [00:10<00:01, 73.29it/s]
 88%|████████▊ | 748/851 [00:10<00:01, 72.54it/s]
 89%|████████▉ | 756/851 [00:10<00:01, 72.39it/s]
 90%|████████▉ | 764/851 [00:10<00:01, 73.41it/s]
 91%|█████████ | 772/851 [00:11<00:01, 68.50it/s]
 92%|█████████▏| 779/851 [00:11<00:01, 66.70it/s]
 92%|█████████▏| 786/851 [00:11<00:00, 67.37it/s]
 93%|█████████▎| 794/851 [00:11<00:00, 68.93it/s]
 94%|█████████▍| 802/851 [00:11<00:00, 71.12it/s]
 95%|█████████▌| 810/851 [00:11<00:00, 68.26it/s]
 96%|█████████▌| 818/851 [00:11<00:00, 69.10it/s]
 97%|█████████▋| 826/851 [00:11<00:00, 70.24it/s]
 98%|█████████▊| 834/851 [00:11<00:00, 69.55it/s]
 99%|█████████▉| 841/851 [00:12<00:00, 69.21it/s]
+[INFO|trainer.py:3305] 2024-05-13 14:59:15,354 >> Saving model checkpoint to /content/dissertation/scripts/ner/output
+[INFO|configuration_utils.py:471] 2024-05-13 14:59:15,355 >> Configuration saved in /content/dissertation/scripts/ner/output/config.json
+[INFO|modeling_utils.py:2590] 2024-05-13 14:59:16,582 >> Model weights saved in /content/dissertation/scripts/ner/output/model.safetensors
+[INFO|tokenization_utils_base.py:2488] 2024-05-13 14:59:16,583 >> tokenizer config file saved in /content/dissertation/scripts/ner/output/tokenizer_config.json
+[INFO|tokenization_utils_base.py:2497] 2024-05-13 14:59:16,584 >> Special tokens file saved in /content/dissertation/scripts/ner/output/special_tokens_map.json
+***** predict metrics *****
+  predict_accuracy           =       0.81
+  predict_f1                 =     0.1759
+  predict_loss               =     0.6631
+  predict_precision          =     0.0969
+  predict_recall             =     0.9485
+  predict_runtime            = 0:00:15.89
+  predict_samples_per_second =    428.156
+  predict_steps_per_second   =     53.527
+

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 9.997061416397296,
+    "total_flos": 6700722040732752.0,
+    "train_loss": 0.08913132946046923,
+    "train_runtime": 3337.712,
+    "train_samples": 27224,
+    "train_samples_per_second": 81.565,
+    "train_steps_per_second": 5.096
+}

trainer_state.json

@@ -0,0 +1,388 @@
+{
+  "best_metric": 0.17586912065439672,
+  "best_model_checkpoint": "/content/dissertation/scripts/ner/output/checkpoint-3403",
+  "epoch": 9.997061416397296,
+  "eval_steps": 500,
+  "global_step": 17010,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.29385836027034967,
+      "grad_norm": 1.9668159484863281,
+      "learning_rate": 4.853027630805409e-05,
+      "loss": 0.4174,
+      "step": 500
+    },
+    {
+      "epoch": 0.5877167205406993,
+      "grad_norm": 3.246731758117676,
+      "learning_rate": 4.7060552616108174e-05,
+      "loss": 0.2765,
+      "step": 1000
+    },
+    {
+      "epoch": 0.8815750808110491,
+      "grad_norm": 2.936720609664917,
+      "learning_rate": 4.559082892416226e-05,
+      "loss": 0.2596,
+      "step": 1500
+    },
+    {
+      "epoch": 0.9997061416397296,
+      "eval_accuracy": 0.7671626010120082,
+      "eval_f1": 0.14622946114658822,
+      "eval_loss": 0.7913026213645935,
+      "eval_precision": 0.0793057825511713,
+      "eval_recall": 0.9365808823529411,
+      "eval_runtime": 16.052,
+      "eval_samples_per_second": 424.06,
+      "eval_steps_per_second": 53.015,
+      "step": 1701
+    },
+    {
+      "epoch": 1.1754334410813987,
+      "grad_norm": 2.302980661392212,
+      "learning_rate": 4.4121105232216346e-05,
+      "loss": 0.2135,
+      "step": 2000
+    },
+    {
+      "epoch": 1.4692918013517484,
+      "grad_norm": 3.7150967121124268,
+      "learning_rate": 4.265138154027043e-05,
+      "loss": 0.1839,
+      "step": 2500
+    },
+    {
+      "epoch": 1.7631501616220981,
+      "grad_norm": 1.2019191980361938,
+      "learning_rate": 4.118165784832452e-05,
+      "loss": 0.1853,
+      "step": 3000
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.8099635429897495,
+      "eval_f1": 0.17586912065439672,
+      "eval_loss": 0.6631014347076416,
+      "eval_precision": 0.09691960931630353,
+      "eval_recall": 0.9485294117647058,
+      "eval_runtime": 15.875,
+      "eval_samples_per_second": 428.788,
+      "eval_steps_per_second": 53.606,
+      "step": 3403
+    },
+    {
+      "epoch": 2.0570085218924477,
+      "grad_norm": 2.090308904647827,
+      "learning_rate": 3.971193415637861e-05,
+      "loss": 0.1684,
+      "step": 3500
+    },
+    {
+      "epoch": 2.3508668821627974,
+      "grad_norm": 0.9995286464691162,
+      "learning_rate": 3.824221046443269e-05,
+      "loss": 0.1222,
+      "step": 4000
+    },
+    {
+      "epoch": 2.644725242433147,
+      "grad_norm": 1.3526027202606201,
+      "learning_rate": 3.677248677248677e-05,
+      "loss": 0.1277,
+      "step": 4500
+    },
+    {
+      "epoch": 2.938583602703497,
+      "grad_norm": 2.2816500663757324,
+      "learning_rate": 3.530276308054086e-05,
+      "loss": 0.1254,
+      "step": 5000
+    },
+    {
+      "epoch": 2.9997061416397295,
+      "eval_accuracy": 0.7754975935626944,
+      "eval_f1": 0.16528259292106748,
+      "eval_loss": 1.072888731956482,
+      "eval_precision": 0.0905877154220062,
+      "eval_recall": 0.9420955882352942,
+      "eval_runtime": 15.8836,
+      "eval_samples_per_second": 428.556,
+      "eval_steps_per_second": 53.577,
+      "step": 5104
+    },
+    {
+      "epoch": 3.2324419629738466,
+      "grad_norm": 0.6440290808677673,
+      "learning_rate": 3.3833039388594945e-05,
+      "loss": 0.0929,
+      "step": 5500
+    },
+    {
+      "epoch": 3.5263003232441963,
+      "grad_norm": 3.7906153202056885,
+      "learning_rate": 3.2363315696649034e-05,
+      "loss": 0.0896,
+      "step": 6000
+    },
+    {
+      "epoch": 3.820158683514546,
+      "grad_norm": 1.0953032970428467,
+      "learning_rate": 3.0893592004703116e-05,
+      "loss": 0.0823,
+      "step": 6500
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.7719411469883488,
+      "eval_f1": 0.16243814311523055,
+      "eval_loss": 1.2567578554153442,
+      "eval_precision": 0.08880872627329726,
+      "eval_recall": 0.9503676470588235,
+      "eval_runtime": 15.8976,
+      "eval_samples_per_second": 428.179,
+      "eval_steps_per_second": 53.53,
+      "step": 6806
+    },
+    {
+      "epoch": 4.114017043784895,
+      "grad_norm": 1.7823286056518555,
+      "learning_rate": 2.9423868312757202e-05,
+      "loss": 0.0761,
+      "step": 7000
+    },
+    {
+      "epoch": 4.407875404055245,
+      "grad_norm": 1.165720820426941,
+      "learning_rate": 2.795414462081129e-05,
+      "loss": 0.0603,
+      "step": 7500
+    },
+    {
+      "epoch": 4.701733764325595,
+      "grad_norm": 4.1130452156066895,
+      "learning_rate": 2.648442092886537e-05,
+      "loss": 0.0589,
+      "step": 8000
+    },
+    {
+      "epoch": 4.9955921245959445,
+      "grad_norm": 0.3957385718822479,
+      "learning_rate": 2.501469723691946e-05,
+      "loss": 0.0597,
+      "step": 8500
+    },
+    {
+      "epoch": 4.99970614163973,
+      "eval_accuracy": 0.7836540772119656,
+      "eval_f1": 0.17101181993461315,
+      "eval_loss": 1.1907650232315063,
+      "eval_precision": 0.0940872613227562,
+      "eval_recall": 0.9375,
+      "eval_runtime": 16.019,
+      "eval_samples_per_second": 424.932,
+      "eval_steps_per_second": 53.124,
+      "step": 8507
+    },
+    {
+      "epoch": 5.289450484866294,
+      "grad_norm": 1.0087623596191406,
+      "learning_rate": 2.3544973544973546e-05,
+      "loss": 0.0423,
+      "step": 9000
+    },
+    {
+      "epoch": 5.583308845136644,
+      "grad_norm": 2.162200450897217,
+      "learning_rate": 2.2075249853027632e-05,
+      "loss": 0.043,
+      "step": 9500
+    },
+    {
+      "epoch": 5.877167205406994,
+      "grad_norm": 1.1820895671844482,
+      "learning_rate": 2.0605526161081718e-05,
+      "loss": 0.0446,
+      "step": 10000
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.7811686840461102,
+      "eval_f1": 0.1718036055495555,
+      "eval_loss": 1.384422779083252,
+      "eval_precision": 0.09443784820531555,
+      "eval_recall": 0.9503676470588235,
+      "eval_runtime": 15.9006,
+      "eval_samples_per_second": 428.098,
+      "eval_steps_per_second": 53.52,
+      "step": 10209
+    },
+    {
+      "epoch": 6.171025565677343,
+      "grad_norm": 2.190476894378662,
+      "learning_rate": 1.91358024691358e-05,
+      "loss": 0.0333,
+      "step": 10500
+    },
+    {
+      "epoch": 6.464883925947693,
+      "grad_norm": 1.5180469751358032,
+      "learning_rate": 1.766607877718989e-05,
+      "loss": 0.0333,
+      "step": 11000
+    },
+    {
+      "epoch": 6.758742286218043,
+      "grad_norm": 1.4279770851135254,
+      "learning_rate": 1.6196355085243976e-05,
+      "loss": 0.0325,
+      "step": 11500
+    },
+    {
+      "epoch": 6.99970614163973,
+      "eval_accuracy": 0.7866406684471785,
+      "eval_f1": 0.1704836709384043,
+      "eval_loss": 1.5515447854995728,
+      "eval_precision": 0.09366766603070772,
+      "eval_recall": 0.9476102941176471,
+      "eval_runtime": 15.8267,
+      "eval_samples_per_second": 430.095,
+      "eval_steps_per_second": 53.77,
+      "step": 11910
+    },
+    {
+      "epoch": 7.052600646488393,
+      "grad_norm": 0.6558970212936401,
+      "learning_rate": 1.472663139329806e-05,
+      "loss": 0.0289,
+      "step": 12000
+    },
+    {
+      "epoch": 7.346459006758742,
+      "grad_norm": 0.26189878582954407,
+      "learning_rate": 1.3256907701352148e-05,
+      "loss": 0.0224,
+      "step": 12500
+    },
+    {
+      "epoch": 7.640317367029092,
+      "grad_norm": 1.370686650276184,
+      "learning_rate": 1.1787184009406232e-05,
+      "loss": 0.0231,
+      "step": 13000
+    },
+    {
+      "epoch": 7.934175727299442,
+      "grad_norm": 0.36619672179222107,
+      "learning_rate": 1.0317460317460318e-05,
+      "loss": 0.022,
+      "step": 13500
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.7842582611859856,
+      "eval_f1": 0.1688722903304376,
+      "eval_loss": 1.6299601793289185,
+      "eval_precision": 0.09261733012734882,
+      "eval_recall": 0.9558823529411765,
+      "eval_runtime": 15.9057,
+      "eval_samples_per_second": 427.961,
+      "eval_steps_per_second": 53.503,
+      "step": 13612
+    },
+    {
+      "epoch": 8.22803408756979,
+      "grad_norm": 1.0557399988174438,
+      "learning_rate": 8.847736625514404e-06,
+      "loss": 0.0164,
+      "step": 14000
+    },
+    {
+      "epoch": 8.521892447840141,
+      "grad_norm": 4.32920503616333,
+      "learning_rate": 7.37801293356849e-06,
+      "loss": 0.0164,
+      "step": 14500
+    },
+    {
+      "epoch": 8.81575080811049,
+      "grad_norm": 0.1674884408712387,
+      "learning_rate": 5.908289241622575e-06,
+      "loss": 0.017,
+      "step": 15000
+    },
+    {
+      "epoch": 8.999706141639729,
+      "eval_accuracy": 0.7844848301762433,
+      "eval_f1": 0.16934759532946844,
+      "eval_loss": 1.7459304332733154,
+      "eval_precision": 0.09292947396720136,
+      "eval_recall": 0.953125,
+      "eval_runtime": 16.0915,
+      "eval_samples_per_second": 423.018,
+      "eval_steps_per_second": 52.885,
+      "step": 15313
+    },
+    {
+      "epoch": 9.10960916838084,
+      "grad_norm": 3.3734261989593506,
+      "learning_rate": 4.438565549676661e-06,
+      "loss": 0.0171,
+      "step": 15500
+    },
+    {
+      "epoch": 9.40346752865119,
+      "grad_norm": 0.19828377664089203,
+      "learning_rate": 2.9688418577307467e-06,
+      "loss": 0.013,
+      "step": 16000
+    },
+    {
+      "epoch": 9.69732588892154,
+      "grad_norm": 1.491190791130066,
+      "learning_rate": 1.4991181657848325e-06,
+      "loss": 0.0133,
+      "step": 16500
+    },
+    {
+      "epoch": 9.991184249191889,
+      "grad_norm": 2.5641109943389893,
+      "learning_rate": 2.9394473838918286e-08,
+      "loss": 0.0135,
+      "step": 17000
+    },
+    {
+      "epoch": 9.997061416397296,
+      "eval_accuracy": 0.7845534874460183,
+      "eval_f1": 0.16896354888689555,
+      "eval_loss": 1.7860842943191528,
+      "eval_precision": 0.09270693512304251,
+      "eval_recall": 0.9522058823529411,
+      "eval_runtime": 15.9319,
+      "eval_samples_per_second": 427.256,
+      "eval_steps_per_second": 53.415,
+      "step": 17010
+    },
+    {
+      "epoch": 9.997061416397296,
+      "step": 17010,
+      "total_flos": 6700722040732752.0,
+      "train_loss": 0.08913132946046923,
+      "train_runtime": 3337.712,
+      "train_samples_per_second": 81.565,
+      "train_steps_per_second": 5.096
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 17010,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 10,
+  "save_steps": 500,
+  "total_flos": 6700722040732752.0,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}