AgitationTextV3
This model is a fine-tuned version of dmis-lab/biobert-base-cased-v1.1 on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.5199
- Accuracy: 0.8
- Precision: 0.9636
- Recall: 0.7465
- F1: 0.8413
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: 1e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | Recall | F1 |
|---|---|---|---|---|---|---|---|
| 0.6797 | 1.0 | 50 | 0.6650 | 0.57 | 0.85 | 0.4789 | 0.6126 |
| 0.599 | 2.0 | 100 | 0.6180 | 0.65 | 0.86 | 0.6056 | 0.7107 |
| 0.5121 | 3.0 | 150 | 0.5714 | 0.75 | 0.8833 | 0.7465 | 0.8092 |
| 0.4049 | 4.0 | 200 | 0.5187 | 0.81 | 0.9194 | 0.8028 | 0.8571 |
| 0.3091 | 5.0 | 250 | 0.5034 | 0.77 | 0.9444 | 0.7183 | 0.816 |
| 0.2303 | 6.0 | 300 | 0.4673 | 0.78 | 0.9298 | 0.7465 | 0.8281 |
| 0.1773 | 7.0 | 350 | 0.4802 | 0.8 | 0.9322 | 0.7746 | 0.8462 |
| 0.1396 | 8.0 | 400 | 0.5260 | 0.8 | 0.9636 | 0.7465 | 0.8413 |
| 0.1204 | 9.0 | 450 | 0.5317 | 0.8 | 0.9636 | 0.7465 | 0.8413 |
| 0.0982 | 10.0 | 500 | 0.5199 | 0.8 | 0.9636 | 0.7465 | 0.8413 |
Framework versions
- Transformers 4.21.2
- Pytorch 1.12.1+cu113
- Datasets 2.4.0
- Tokenizers 0.12.1