gpt2_small_EN_superbpe_8192_parallel10_42
This model was trained from scratch on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 3.8899
- Accuracy: 0.3017
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: 5e-05
- train_batch_size: 128
- eval_batch_size: 128
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- lr_scheduler_warmup_steps: 1000
- num_epochs: 10
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|
| 6.2264 | 1.0 | 1527 | 4.7819 | 0.2163 |
| 4.5245 | 2.0 | 3054 | 4.3524 | 0.2513 |
| 4.3335 | 3.0 | 4581 | 4.1802 | 0.2682 |
| 4.1404 | 4.0 | 6108 | 4.0779 | 0.2784 |
| 4.0646 | 5.0 | 7635 | 4.0127 | 0.2875 |
| 3.9752 | 6.0 | 9162 | 3.9669 | 0.2921 |
| 3.9303 | 7.0 | 10689 | 3.9338 | 0.2957 |
| 3.8762 | 8.0 | 12216 | 3.9116 | 0.2992 |
| 3.8487 | 9.0 | 13743 | 3.8973 | 0.3003 |
| 3.8151 | 10.0 | 15270 | 3.8899 | 0.3017 |
Framework versions
- Transformers 4.41.2
- Pytorch 2.8.0+cu128
- Datasets 4.1.1
- Tokenizers 0.19.1
