N_roberta_agnews_padding60model
This model is a fine-tuned version of roberta-base on the ag_news dataset. It achieves the following results on the evaluation set:
- Loss: 0.5823
- Accuracy: 0.9461
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 |
|---|---|---|---|---|
| 0.2028 | 1.0 | 7500 | 0.2106 | 0.9407 |
| 0.1643 | 2.0 | 15000 | 0.1864 | 0.9475 |
| 0.1536 | 3.0 | 22500 | 0.2135 | 0.9455 |
| 0.1243 | 4.0 | 30000 | 0.2261 | 0.9468 |
| 0.1045 | 5.0 | 37500 | 0.2428 | 0.9468 |
| 0.0861 | 6.0 | 45000 | 0.2795 | 0.9434 |
| 0.0767 | 7.0 | 52500 | 0.3035 | 0.9470 |
| 0.0532 | 8.0 | 60000 | 0.3571 | 0.9461 |
| 0.0532 | 9.0 | 67500 | 0.3586 | 0.9426 |
| 0.0342 | 10.0 | 75000 | 0.4128 | 0.9434 |
| 0.026 | 11.0 | 82500 | 0.4228 | 0.9470 |
| 0.0226 | 12.0 | 90000 | 0.4714 | 0.9434 |
| 0.0209 | 13.0 | 97500 | 0.4663 | 0.9458 |
| 0.0127 | 14.0 | 105000 | 0.4939 | 0.9436 |
| 0.0082 | 15.0 | 112500 | 0.4959 | 0.9483 |
| 0.0142 | 16.0 | 120000 | 0.5230 | 0.9461 |
| 0.0024 | 17.0 | 127500 | 0.5710 | 0.9445 |
| 0.0082 | 18.0 | 135000 | 0.5560 | 0.9459 |
| 0.0034 | 19.0 | 142500 | 0.5778 | 0.9462 |
| 0.0018 | 20.0 | 150000 | 0.5823 | 0.9461 |
Framework versions
- Transformers 4.33.2
- Pytorch 2.0.1+cu117
- Datasets 2.14.5
- Tokenizers 0.13.3