distilroberta-base-mrpc-glue-francisco-flores
This model is a fine-tuned version of distilroberta-base on the glue dataset. It achieves the following results on the evaluation set:
- Loss: 0.7184
- Accuracy: 0.8456
- F1: 0.8869
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: 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: 10
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 |
|---|---|---|---|---|---|
| 0.338 | 1.09 | 500 | 0.7782 | 0.8382 | 0.8896 |
| 0.3216 | 2.18 | 1000 | 0.7184 | 0.8456 | 0.8869 |
| 0.1861 | 3.27 | 1500 | 1.1095 | 0.8358 | 0.8874 |
| 0.1101 | 4.36 | 2000 | 1.3526 | 0.8260 | 0.8799 |
| 0.0572 | 5.45 | 2500 | 1.2464 | 0.8260 | 0.8757 |
| 0.0443 | 6.54 | 3000 | 1.2194 | 0.8407 | 0.8866 |
| 0.0321 | 7.63 | 3500 | 1.3519 | 0.8333 | 0.8803 |
| 0.0146 | 8.71 | 4000 | 1.4999 | 0.8309 | 0.8840 |
| 0.0082 | 9.8 | 4500 | 1.4908 | 0.8333 | 0.8840 |
Framework versions
- Transformers 4.33.1
- Pytorch 2.0.1+cu118
- Datasets 2.14.5
- Tokenizers 0.13.3