gpt2_small_EN_bpe_65536_parallel10_42
This model was trained from scratch on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 3.2714
- Accuracy: 0.4145
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: 64
- eval_batch_size: 64
- seed: 42
- distributed_type: multi-GPU
- num_devices: 2
- total_train_batch_size: 128
- total_eval_batch_size: 128
- 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 |
|---|
| 5.4586 | 1.0 | 1756 | 3.8062 | 0.3698 |
| 3.721 | 2.0 | 3512 | 3.5715 | 0.3857 |
| 3.5321 | 3.0 | 5268 | 3.4657 | 0.3945 |
| 3.4294 | 4.0 | 7024 | 3.3986 | 0.4014 |
| 3.3737 | 5.0 | 8780 | 3.3547 | 0.4060 |
| 3.3119 | 6.0 | 10536 | 3.3225 | 0.4088 |
| 3.264 | 7.0 | 12292 | 3.3012 | 0.4108 |
| 3.2262 | 8.0 | 14048 | 3.2853 | 0.4125 |
| 3.2004 | 9.0 | 15804 | 3.2758 | 0.4134 |
| 3.177 | 10.0 | 17560 | 3.2714 | 0.4145 |
Framework versions
- Transformers 4.41.2
- Pytorch 2.8.0+cu128
- Datasets 4.1.1
- Tokenizers 0.19.1
