gpt-2-4096t-scratch-10ep_
This model is a fine-tuned version of tihon-nth/gpt-2-4096t-scratch-10ep_ on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 3.4516
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: 2e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 0.05
- num_epochs: 10
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 3.8688 | 1.0 | 6414 | 3.7689 |
| 3.7603 | 2.0 | 12828 | 3.6716 |
| 3.6786 | 3.0 | 19242 | 3.6033 |
| 3.6187 | 4.0 | 25656 | 3.5504 |
| 3.5678 | 5.0 | 32070 | 3.5153 |
| 3.5448 | 6.0 | 38484 | 3.4841 |
| 3.5065 | 7.0 | 44898 | 3.4649 |
| 3.4998 | 8.0 | 51312 | 3.4578 |
| 3.4859 | 9.0 | 57726 | 3.4518 |
| 3.4872 | 10.0 | 64140 | 3.4516 |
Framework versions
- Transformers 5.8.0
- Pytorch 2.11.0+cu128
- Datasets 4.8.5
- Tokenizers 0.22.2