Model save

README.md

@@ -1,11 +1,10 @@
 ---
 license: apache-2.0
-base_model: PlanTL-GOB-ES/bsc-bio-ehr-es
+base_model: michiyasunaga/BioLinkBERT-base
 tags:
-- token-classification
 - generated_from_trainer
 datasets:
-- Rodrigo1771/drugtemist-ner
+- drugtemist-en-ner
 metrics:
 - precision
 - recall
@@ -18,24 +17,24 @@ model-index:
       name: Token Classification
       type: token-classification
     dataset:
-      name: Rodrigo1771/drugtemist-ner
-      type: Rodrigo1771/drugtemist-ner
-      config: DrugTEMIST NER
+      name: drugtemist-en-ner
+      type: drugtemist-en-ner
+      config: DrugTEMIST English NER
       split: validation
-      args: DrugTEMIST NER
+      args: DrugTEMIST English NER
     metrics:
     - name: Precision
       type: precision
-      value: 0.9394760614272809
+      value: 0.9191919191919192
     - name: Recall
       type: recall
-      value: 0.9558823529411765
+      value: 0.9328984156570364
     - name: F1
       type: f1
-      value: 0.9476082004555808
+      value: 0.9259944495837187
     - name: Accuracy
       type: accuracy
-      value: 0.9989632965795653
+      value: 0.998618591800854
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -43,13 +42,13 @@ 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 Rodrigo1771/drugtemist-ner dataset.
+This model is a fine-tuned version of [michiyasunaga/BioLinkBERT-base](https://huggingface.co/michiyasunaga/BioLinkBERT-base) on the drugtemist-en-ner dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.0055
-- Precision: 0.9395
-- Recall: 0.9559
-- F1: 0.9476
-- Accuracy: 0.9990
+- Loss: 0.0073
+- Precision: 0.9192
+- Recall: 0.9329
+- F1: 0.9260
+- Accuracy: 0.9986
 
 ## Model description
 
@@ -80,18 +79,18 @@ The following hyperparameters were used during training:
 
 ### Training results
 
-| Training Loss | Epoch  | Step | Validation Loss | Precision | Recall | F1     | Accuracy |
-|:-------------:|:------:|:----:|:---------------:|:---------:|:------:|:------:|:--------:|
-| No log        | 0.9988 | 425  | 0.0050          | 0.9251    | 0.8971 | 0.9109 | 0.9982   |
-| 0.0196        | 2.0    | 851  | 0.0038          | 0.9400    | 0.9219 | 0.9309 | 0.9988   |
-| 0.0031        | 2.9988 | 1276 | 0.0035          | 0.9335    | 0.9292 | 0.9314 | 0.9988   |
-| 0.0017        | 4.0    | 1702 | 0.0041          | 0.9119    | 0.9605 | 0.9355 | 0.9988   |
-| 0.0009        | 4.9988 | 2127 | 0.0047          | 0.9393    | 0.9522 | 0.9457 | 0.9989   |
-| 0.0004        | 6.0    | 2553 | 0.0055          | 0.9413    | 0.9430 | 0.9421 | 0.9989   |
-| 0.0004        | 6.9988 | 2978 | 0.0054          | 0.9320    | 0.9577 | 0.9447 | 0.9989   |
-| 0.0003        | 8.0    | 3404 | 0.0053          | 0.9346    | 0.9596 | 0.9469 | 0.9989   |
-| 0.0002        | 8.9988 | 3829 | 0.0057          | 0.9385    | 0.9531 | 0.9457 | 0.9989   |
-| 0.0001        | 9.9882 | 4250 | 0.0055          | 0.9395    | 0.9559 | 0.9476 | 0.9990   |
+| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1     | Accuracy |
+|:-------------:|:-----:|:----:|:---------------:|:---------:|:------:|:------:|:--------:|
+| No log        | 1.0   | 434  | 0.0057          | 0.8938    | 0.8938 | 0.8938 | 0.9981   |
+| 0.0182        | 2.0   | 868  | 0.0044          | 0.9024    | 0.9301 | 0.9160 | 0.9985   |
+| 0.0039        | 3.0   | 1302 | 0.0045          | 0.9129    | 0.9282 | 0.9205 | 0.9987   |
+| 0.0024        | 4.0   | 1736 | 0.0051          | 0.8821    | 0.9348 | 0.9077 | 0.9983   |
+| 0.0017        | 5.0   | 2170 | 0.0057          | 0.9251    | 0.9320 | 0.9285 | 0.9986   |
+| 0.0012        | 6.0   | 2604 | 0.0061          | 0.9001    | 0.9236 | 0.9117 | 0.9984   |
+| 0.0009        | 7.0   | 3038 | 0.0056          | 0.9327    | 0.9301 | 0.9314 | 0.9987   |
+| 0.0009        | 8.0   | 3472 | 0.0068          | 0.9118    | 0.9348 | 0.9231 | 0.9986   |
+| 0.0006        | 9.0   | 3906 | 0.0072          | 0.9267    | 0.9310 | 0.9289 | 0.9987   |
+| 0.0004        | 10.0  | 4340 | 0.0073          | 0.9192    | 0.9329 | 0.9260 | 0.9986   |
 
 
 ### Framework versions

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b02e45065034d02bb35c7b02710b67509869c41e34709f5df8ac78e1a7afca87
+oid sha256:bcbb089b16f14979e74b0e93151976bfcd368b7543d817d109a468397e8fadf3
 size 430601004

tb/events.out.tfevents.1725053057.6b97e535edda.35455.0

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:dacc5f14849d9eb6e517fa55f5251f62fc9dccfc3b74ea096b6d6c2e1bbba7cf
-size 10987
+oid sha256:060b7fdb33d177fa8187153a2ec8df8664c9f3e11c25faef21bd32c1122357bf
+size 11813

train.log

@@ -1273,3 +1273,24 @@ You should probably TRAIN this model on a down-stream task to be able to use it
 [INFO|modeling_utils.py:2690] 2024-08-30 21:41:35,219 >> Model weights saved in /content/dissertation/scripts/ner/output/checkpoint-4340/model.safetensors
 [INFO|tokenization_utils_base.py:2574] 2024-08-30 21:41:35,220 >> tokenizer config file saved in /content/dissertation/scripts/ner/output/checkpoint-4340/tokenizer_config.json
 [INFO|tokenization_utils_base.py:2583] 2024-08-30 21:41:35,221 >> Special tokens file saved in /content/dissertation/scripts/ner/output/checkpoint-4340/special_tokens_map.json
+[INFO|trainer.py:2383] 2024-08-30 21:41:36,725 >> 
+
+Training completed. Do not forget to share your model on huggingface.co/models =)
+
+
+[INFO|trainer.py:2621] 2024-08-30 21:41:36,725 >> Loading best model from /content/dissertation/scripts/ner/output/checkpoint-3038 (score: 0.9314045730284647).
+
                                                   
+[INFO|trainer.py:4239] 2024-08-30 21:41:36,893 >> Waiting for the current checkpoint push to be finished, this might take a couple of minutes.
+[INFO|trainer.py:3478] 2024-08-30 21:41:37,582 >> Saving model checkpoint to /content/dissertation/scripts/ner/output
+[INFO|configuration_utils.py:472] 2024-08-30 21:41:37,583 >> Configuration saved in /content/dissertation/scripts/ner/output/config.json
+[INFO|modeling_utils.py:2690] 2024-08-30 21:41:38,705 >> Model weights saved in /content/dissertation/scripts/ner/output/model.safetensors
+[INFO|tokenization_utils_base.py:2574] 2024-08-30 21:41:38,706 >> tokenizer config file saved in /content/dissertation/scripts/ner/output/tokenizer_config.json
+[INFO|tokenization_utils_base.py:2583] 2024-08-30 21:41:38,706 >> Special tokens file saved in /content/dissertation/scripts/ner/output/special_tokens_map.json
+[INFO|trainer.py:3478] 2024-08-30 21:41:38,719 >> Saving model checkpoint to /content/dissertation/scripts/ner/output
+[INFO|configuration_utils.py:472] 2024-08-30 21:41:38,721 >> Configuration saved in /content/dissertation/scripts/ner/output/config.json
+[INFO|modeling_utils.py:2690] 2024-08-30 21:41:39,977 >> Model weights saved in /content/dissertation/scripts/ner/output/model.safetensors
+[INFO|tokenization_utils_base.py:2574] 2024-08-30 21:41:39,979 >> tokenizer config file saved in /content/dissertation/scripts/ner/output/tokenizer_config.json
+[INFO|tokenization_utils_base.py:2583] 2024-08-30 21:41:39,979 >> Special tokens file saved in /content/dissertation/scripts/ner/output/special_tokens_map.json
+{'eval_loss': 0.007294897455722094, 'eval_precision': 0.9191919191919192, 'eval_recall': 0.9328984156570364, 'eval_f1': 0.9259944495837187, 'eval_accuracy': 0.998618591800854, 'eval_runtime': 13.8041, 'eval_samples_per_second': 503.184, 'eval_steps_per_second': 62.952, 'epoch': 10.0}
+{'train_runtime': 1039.0596, 'train_samples_per_second': 267.242, 'train_steps_per_second': 4.177, 'train_loss': 0.003382195293697344, 'epoch': 10.0}
+