distilbert_agnews_padding30model
This model is a fine-tuned version of distilbert-base-uncased on the ag_news dataset. It achieves the following results on the evaluation set:
- Loss: 0.6664
- Accuracy: 0.9426
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: 20
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 0.1798 | 1.0 | 7500 | 0.1996 | 0.9397 |
| 0.1377 | 2.0 | 15000 | 0.1984 | 0.9432 |
| 0.1168 | 3.0 | 22500 | 0.2269 | 0.9429 |
| 0.0831 | 4.0 | 30000 | 0.2763 | 0.9411 |
| 0.0581 | 5.0 | 37500 | 0.2916 | 0.9428 |
| 0.0422 | 6.0 | 45000 | 0.3627 | 0.9424 |
| 0.0296 | 7.0 | 52500 | 0.4506 | 0.94 |
| 0.0315 | 8.0 | 60000 | 0.4401 | 0.9417 |
| 0.0224 | 9.0 | 67500 | 0.4668 | 0.9426 |
| 0.0151 | 10.0 | 75000 | 0.5095 | 0.9416 |
| 0.0202 | 11.0 | 82500 | 0.5013 | 0.9437 |
| 0.009 | 12.0 | 90000 | 0.5612 | 0.9441 |
| 0.0105 | 13.0 | 97500 | 0.5372 | 0.9432 |
| 0.0105 | 14.0 | 105000 | 0.5760 | 0.9409 |
| 0.0044 | 15.0 | 112500 | 0.5765 | 0.9417 |
| 0.0029 | 16.0 | 120000 | 0.6285 | 0.9432 |
| 0.002 | 17.0 | 127500 | 0.6484 | 0.9416 |
| 0.0006 | 18.0 | 135000 | 0.6568 | 0.9428 |
| 0.0002 | 19.0 | 142500 | 0.6537 | 0.9437 |
| 0.0004 | 20.0 | 150000 | 0.6664 | 0.9426 |
Framework versions
- Transformers 4.33.2
- Pytorch 2.0.1+cu117
- Datasets 2.14.5
- Tokenizers 0.13.3