gpt2_small_PL_superbpe_8192_parallel10_42
This model was trained from scratch on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 3.7594
- Accuracy: 0.3350
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 |
|---|
| 6.5598 | 1.0 | 1485 | 4.9958 | 0.2095 |
| 5.0914 | 2.0 | 2970 | 4.3600 | 0.2672 |
| 4.3139 | 3.0 | 4455 | 4.1204 | 0.2934 |
| 4.1624 | 4.0 | 5940 | 3.9923 | 0.3068 |
| 3.9739 | 5.0 | 7425 | 3.9115 | 0.3157 |
| 3.9131 | 6.0 | 8910 | 3.8571 | 0.3225 |
| 3.8179 | 7.0 | 10395 | 3.8161 | 0.3275 |
| 3.7821 | 8.0 | 11880 | 3.7878 | 0.3308 |
| 3.7249 | 9.0 | 13365 | 3.7683 | 0.3340 |
| 3.7065 | 10.0 | 14850 | 3.7594 | 0.3350 |
Framework versions
- Transformers 4.41.2
- Pytorch 2.8.0+cu128
- Datasets 4.1.1
- Tokenizers 0.19.1
