output
This model is a fine-tuned version of PlanTL-GOB-ES/bsc-bio-ehr-es on the Rodrigo1771/drugtemist-ner dataset. It achieves the following results on the evaluation set:
- Loss: 0.0058
- Precision: 0.9430
- Recall: 0.9586
- F1: 0.9508
- Accuracy: 0.9990
Model description
More information needed
Intended uses & limitations
More information needed
Training and evaluation data
More information needed
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5e-05
- train_batch_size: 4
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10.0
Training results
Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
---|---|---|---|---|---|---|---|
0.0055 | 0.9997 | 1701 | 0.0039 | 0.9182 | 0.8869 | 0.9023 | 0.9984 |
0.0037 | 2.0 | 3403 | 0.0057 | 0.9195 | 0.8814 | 0.9000 | 0.9982 |
0.0024 | 2.9997 | 5104 | 0.0036 | 0.9293 | 0.9540 | 0.9415 | 0.9990 |
0.0013 | 4.0 | 6806 | 0.0053 | 0.9477 | 0.9329 | 0.9403 | 0.9989 |
0.0012 | 4.9997 | 8507 | 0.0063 | 0.9419 | 0.9393 | 0.9406 | 0.9989 |
0.0007 | 6.0 | 10209 | 0.0048 | 0.9403 | 0.9559 | 0.9480 | 0.9989 |
0.0005 | 6.9997 | 11910 | 0.0066 | 0.9490 | 0.9403 | 0.9446 | 0.9989 |
0.0002 | 8.0 | 13612 | 0.0058 | 0.9430 | 0.9586 | 0.9508 | 0.9990 |
0.0001 | 8.9997 | 15313 | 0.0063 | 0.9413 | 0.9577 | 0.9494 | 0.9990 |
0.0 | 9.9971 | 17010 | 0.0065 | 0.9437 | 0.9559 | 0.9498 | 0.9990 |
Framework versions
- Transformers 4.40.2
- Pytorch 2.2.1+cu121
- Datasets 2.19.1
- Tokenizers 0.19.1
- Downloads last month
- 97
Finetuned from
Dataset used to train Rodrigo1771/bsc-bio-ehr-es-drugtemist-ner
Evaluation results
- Precision on Rodrigo1771/drugtemist-nervalidation set self-reported0.943
- Recall on Rodrigo1771/drugtemist-nervalidation set self-reported0.959
- F1 on Rodrigo1771/drugtemist-nervalidation set self-reported0.951
- Accuracy on Rodrigo1771/drugtemist-nervalidation set self-reported0.999