metadata
license: mit
tags:
- generated_from_trainer
metrics:
- f1
model-index:
- name: clinical_bio_bert_ft
results: []
clinical_bio_bert_ft
This model is a fine-tuned version of emilyalsentzer/Bio_ClinicalBERT on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.2570
- F1: 0.8160
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: 32
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10
Training results
| Training Loss | Epoch | Step | Validation Loss | F1 |
|---|---|---|---|---|
| 0.6327 | 1.0 | 95 | 0.2442 | 0.7096 |
| 0.1692 | 2.0 | 190 | 0.2050 | 0.7701 |
| 0.0878 | 3.0 | 285 | 0.1923 | 0.8002 |
| 0.0493 | 4.0 | 380 | 0.2234 | 0.8079 |
| 0.0302 | 5.0 | 475 | 0.2250 | 0.8090 |
| 0.0191 | 6.0 | 570 | 0.2363 | 0.8145 |
| 0.0132 | 7.0 | 665 | 0.2489 | 0.8178 |
| 0.0102 | 8.0 | 760 | 0.2494 | 0.8152 |
| 0.008 | 9.0 | 855 | 0.2542 | 0.8191 |
| 0.0068 | 10.0 | 950 | 0.2570 | 0.8160 |
Framework versions
- Transformers 4.21.1
- Pytorch 1.12.0+cu113
- Datasets 2.4.0
- Tokenizers 0.12.1