output
This model is a fine-tuned version of 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: 0.6631
- Precision: 0.0969
- Recall: 0.9485
- F1: 0.1759
- Accuracy: 0.8100
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 |
---|---|---|---|---|---|---|---|
0.2596 | 0.9997 | 1701 | 0.7913 | 0.0793 | 0.9366 | 0.1462 | 0.7672 |
0.1853 | 2.0 | 3403 | 0.6631 | 0.0969 | 0.9485 | 0.1759 | 0.8100 |
0.1254 | 2.9997 | 5104 | 1.0729 | 0.0906 | 0.9421 | 0.1653 | 0.7755 |
0.0823 | 4.0 | 6806 | 1.2568 | 0.0888 | 0.9504 | 0.1624 | 0.7719 |
0.0597 | 4.9997 | 8507 | 1.1908 | 0.0941 | 0.9375 | 0.1710 | 0.7837 |
0.0446 | 6.0 | 10209 | 1.3844 | 0.0944 | 0.9504 | 0.1718 | 0.7812 |
0.0325 | 6.9997 | 11910 | 1.5515 | 0.0937 | 0.9476 | 0.1705 | 0.7866 |
0.022 | 8.0 | 13612 | 1.6300 | 0.0926 | 0.9559 | 0.1689 | 0.7843 |
0.017 | 8.9997 | 15313 | 1.7459 | 0.0929 | 0.9531 | 0.1693 | 0.7845 |
0.0135 | 9.9971 | 17010 | 1.7861 | 0.0927 | 0.9522 | 0.1690 | 0.7846 |
Framework versions
- Transformers 4.40.2
- Pytorch 2.2.1+cu121
- Datasets 2.19.1
- Tokenizers 0.19.1
- Downloads last month
- 5
Finetuned from
Dataset used to train Rodrigo1771/bsc-bio-ehr-es-multi-train-drugtemist-dev-ner
Evaluation results
- Precision on Rodrigo1771/multi-train-drugtemist-dev-nervalidation set self-reported0.097
- Recall on Rodrigo1771/multi-train-drugtemist-dev-nervalidation set self-reported0.949
- F1 on Rodrigo1771/multi-train-drugtemist-dev-nervalidation set self-reported0.176
- Accuracy on Rodrigo1771/multi-train-drugtemist-dev-nervalidation set self-reported0.810