distilbert-qa-checkpoint-v3
This model is a fine-tuned version of distilbert-base-uncased on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.4014
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: 2e-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: 30
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 0.5057 | 1.0 | 662 | 0.1916 |
| 0.1801 | 2.0 | 1324 | 0.1784 |
| 0.1295 | 3.0 | 1986 | 0.1688 |
| 0.084 | 4.0 | 2648 | 0.1876 |
| 0.0748 | 5.0 | 3310 | 0.2192 |
| 0.0621 | 6.0 | 3972 | 0.1994 |
| 0.0488 | 7.0 | 4634 | 0.2558 |
| 0.0462 | 8.0 | 5296 | 0.2264 |
| 0.046 | 9.0 | 5958 | 0.2734 |
| 0.038 | 10.0 | 6620 | 0.2966 |
| 0.0325 | 11.0 | 7282 | 0.2773 |
| 0.0341 | 12.0 | 7944 | 0.3233 |
| 0.0297 | 13.0 | 8606 | 0.3062 |
| 0.0293 | 14.0 | 9268 | 0.3455 |
| 0.0255 | 15.0 | 9930 | 0.3306 |
| 0.0228 | 16.0 | 10592 | 0.3239 |
| 0.0248 | 17.0 | 11254 | 0.3393 |
| 0.0216 | 18.0 | 11916 | 0.4004 |
| 0.023 | 19.0 | 12578 | 0.3621 |
| 0.0175 | 20.0 | 13240 | 0.3544 |
| 0.0227 | 21.0 | 13902 | 0.3986 |
| 0.0185 | 22.0 | 14564 | 0.3904 |
| 0.0194 | 23.0 | 15226 | 0.3922 |
| 0.0178 | 24.0 | 15888 | 0.4096 |
| 0.0177 | 25.0 | 16550 | 0.4134 |
| 0.0183 | 26.0 | 17212 | 0.3971 |
| 0.0194 | 27.0 | 17874 | 0.4049 |
| 0.0184 | 28.0 | 18536 | 0.3987 |
| 0.0171 | 29.0 | 19198 | 0.4012 |
| 0.0169 | 30.0 | 19860 | 0.4014 |
Framework versions
- Transformers 4.27.2
- Pytorch 1.13.1+cu117
- Datasets 2.11.0
- Tokenizers 0.13.3