gpt2_small_PL_bpe_49152_parallel10_42
This model was trained from scratch on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 3.7841
- Accuracy: 0.4137
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 |
|---|
| 6.0962 | 1.0 | 1450 | 4.8101 | 0.3441 |
| 4.8575 | 2.0 | 2900 | 4.3823 | 0.3678 |
| 4.3192 | 3.0 | 4350 | 4.1588 | 0.3833 |
| 4.1733 | 4.0 | 5800 | 4.0262 | 0.3920 |
| 3.9857 | 5.0 | 7250 | 3.9367 | 0.3990 |
| 3.9162 | 6.0 | 8700 | 3.8777 | 0.4047 |
| 3.824 | 7.0 | 10150 | 3.8379 | 0.4077 |
| 3.779 | 8.0 | 11600 | 3.8102 | 0.4110 |
| 3.7308 | 9.0 | 13050 | 3.7906 | 0.4129 |
| 3.7076 | 10.0 | 14500 | 3.7841 | 0.4137 |
Framework versions
- Transformers 4.41.2
- Pytorch 2.8.0+cu128
- Datasets 4.1.1
- Tokenizers 0.19.1
