N_roberta_agnews_padding30model
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: 0.5706
- Accuracy: 0.9478
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 |
|---|---|---|---|---|
| 0.1966 | 1.0 | 7500 | 0.2071 | 0.9384 |
| 0.1667 | 2.0 | 15000 | 0.1922 | 0.9466 |
| 0.1523 | 3.0 | 22500 | 0.2323 | 0.9438 |
| 0.1194 | 4.0 | 30000 | 0.2370 | 0.9438 |
| 0.105 | 5.0 | 37500 | 0.2791 | 0.9454 |
| 0.0836 | 6.0 | 45000 | 0.2917 | 0.9433 |
| 0.0711 | 7.0 | 52500 | 0.3344 | 0.9436 |
| 0.0586 | 8.0 | 60000 | 0.3723 | 0.9416 |
| 0.0396 | 9.0 | 67500 | 0.3977 | 0.9438 |
| 0.0369 | 10.0 | 75000 | 0.4096 | 0.9425 |
| 0.0312 | 11.0 | 82500 | 0.4293 | 0.9438 |
| 0.0259 | 12.0 | 90000 | 0.4286 | 0.9436 |
| 0.0241 | 13.0 | 97500 | 0.4529 | 0.9437 |
| 0.0129 | 14.0 | 105000 | 0.4749 | 0.9442 |
| 0.0057 | 15.0 | 112500 | 0.5355 | 0.9429 |
| 0.0083 | 16.0 | 120000 | 0.5056 | 0.9475 |
| 0.0062 | 17.0 | 127500 | 0.5138 | 0.9458 |
| 0.0062 | 18.0 | 135000 | 0.5368 | 0.9463 |
| 0.0026 | 19.0 | 142500 | 0.5647 | 0.9470 |
| 0.0031 | 20.0 | 150000 | 0.5706 | 0.9478 |
Framework versions
- Transformers 4.33.2
- Pytorch 2.0.1+cu117
- Datasets 2.14.5
- Tokenizers 0.13.3