distilbert_agnews_padding20model
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: 0.6477
- Accuracy: 0.9453
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 |
|---|---|---|---|---|
| 0.18 | 1.0 | 7500 | 0.1831 | 0.9426 |
| 0.1365 | 2.0 | 15000 | 0.2039 | 0.9420 |
| 0.1176 | 3.0 | 22500 | 0.2202 | 0.9470 |
| 0.0899 | 4.0 | 30000 | 0.2601 | 0.9443 |
| 0.0547 | 5.0 | 37500 | 0.2919 | 0.9429 |
| 0.0387 | 6.0 | 45000 | 0.3618 | 0.9459 |
| 0.0351 | 7.0 | 52500 | 0.4129 | 0.9413 |
| 0.031 | 8.0 | 60000 | 0.4379 | 0.9436 |
| 0.0171 | 9.0 | 67500 | 0.4794 | 0.9429 |
| 0.0156 | 10.0 | 75000 | 0.4744 | 0.9438 |
| 0.0147 | 11.0 | 82500 | 0.4832 | 0.9457 |
| 0.0108 | 12.0 | 90000 | 0.5166 | 0.9447 |
| 0.0034 | 13.0 | 97500 | 0.5083 | 0.9459 |
| 0.0065 | 14.0 | 105000 | 0.5451 | 0.9446 |
| 0.0062 | 15.0 | 112500 | 0.5926 | 0.9443 |
| 0.0031 | 16.0 | 120000 | 0.6059 | 0.9433 |
| 0.001 | 17.0 | 127500 | 0.6312 | 0.9463 |
| 0.0004 | 18.0 | 135000 | 0.6197 | 0.9454 |
| 0.0004 | 19.0 | 142500 | 0.6472 | 0.9455 |
| 0.0002 | 20.0 | 150000 | 0.6477 | 0.9453 |
Framework versions
- Transformers 4.33.2
- Pytorch 2.0.1+cu117
- Datasets 2.14.5
- Tokenizers 0.13.3