gpt2_small_FR_bpe_70255_parallel10_42
This model was trained from scratch on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 2.9846
- Accuracy: 0.4425
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 |
|---|
| 5.3099 | 1.0 | 1746 | 3.5474 | 0.3943 |
| 3.4406 | 2.0 | 3492 | 3.3001 | 0.4129 |
| 3.2455 | 3.0 | 5238 | 3.1882 | 0.4223 |
| 3.1774 | 4.0 | 6984 | 3.1184 | 0.4277 |
| 3.0822 | 5.0 | 8730 | 3.0727 | 0.4317 |
| 3.0201 | 6.0 | 10476 | 3.0378 | 0.4355 |
| 2.973 | 7.0 | 12222 | 3.0151 | 0.4386 |
| 2.9433 | 8.0 | 13968 | 2.9990 | 0.4402 |
| 2.9071 | 9.0 | 15714 | 2.9890 | 0.4415 |
| 2.8869 | 10.0 | 17460 | 2.9846 | 0.4425 |
Framework versions
- Transformers 4.41.2
- Pytorch 2.8.0+cu128
- Datasets 4.1.1
- Tokenizers 0.19.1
