Biobert-base-cased-v1.2-finetuned-ner-CRAFT_es_en
This model is a fine-tuned version of dmis-lab/biobert-base-cased-v1.2 on the CRAFT dataset. It achieves the following results on the evaluation set:
- Loss: 0.1811
- Precision: 0.8555
- Recall: 0.8539
- F1: 0.8547
- Accuracy: 0.9706
Model description
This model performs Named Entity Recognition for 6 entity tags: Sequence, Cell, Protein, Gene, Taxon, and Chemical from the CRAFT(Colorado Richly Annotated Full Text) Corpus in Spanish and English. Entity tags have been normalized and replaced from the original three letter code to a full name e.g. B-Protein, I-Chemical.
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: 3e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 4
Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|---|---|---|---|---|---|---|---|
| 0.052 | 1.0 | 1360 | 0.1413 | 0.8300 | 0.8442 | 0.8370 | 0.9677 |
| 0.0199 | 2.0 | 2720 | 0.1673 | 0.8461 | 0.8458 | 0.8459 | 0.9689 |
| 0.011 | 3.0 | 4080 | 0.1647 | 0.8588 | 0.8528 | 0.8558 | 0.9704 |
| 0.0031 | 4.0 | 5440 | 0.1811 | 0.8555 | 0.8539 | 0.8547 | 0.9706 |
Framework versions
- Transformers 4.17.0
- Pytorch 1.10.0+cu111
- Datasets 1.18.4
- Tokenizers 0.11.6