gpt2_small_FR_superbpe_6651_parallel10_42
This model was trained from scratch on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 3.3790
- Accuracy: 0.3536
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 |
|---|
| 6.0242 | 1.0 | 1582 | 4.3109 | 0.2470 |
| 4.029 | 2.0 | 3164 | 3.8271 | 0.2985 |
| 3.8223 | 3.0 | 4746 | 3.6544 | 0.3182 |
| 3.6233 | 4.0 | 6328 | 3.5579 | 0.3298 |
| 3.5564 | 5.0 | 7910 | 3.4958 | 0.3376 |
| 3.4625 | 6.0 | 9492 | 3.4524 | 0.3435 |
| 3.4043 | 7.0 | 11074 | 3.4240 | 0.3470 |
| 3.3676 | 8.0 | 12656 | 3.4017 | 0.3503 |
| 3.331 | 9.0 | 14238 | 3.3854 | 0.3521 |
| 3.3111 | 10.0 | 15820 | 3.3790 | 0.3536 |
Framework versions
- Transformers 4.41.2
- Pytorch 2.8.0+cu128
- Datasets 4.1.1
- Tokenizers 0.19.1
