gpt2_medium_AR_superbpe_65536_parallel3-100_42
This model was trained from scratch on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 4.4253
- Accuracy: 0.2881
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: 64
- eval_batch_size: 64
- seed: 42
- distributed_type: multi-GPU
- num_devices: 2
- total_train_batch_size: 128
- total_eval_batch_size: 128
- 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 |
|---|
| 5.546 | 1.0 | 5903 | 5.2402 | 0.2192 |
| 4.9407 | 2.0 | 11806 | 4.8276 | 0.2483 |
| 4.6957 | 3.0 | 17709 | 4.6599 | 0.2628 |
| 4.5375 | 4.0 | 23612 | 4.5662 | 0.2707 |
| 4.4157 | 5.0 | 29515 | 4.5113 | 0.2762 |
| 4.315 | 6.0 | 35418 | 4.4721 | 0.2805 |
| 4.2299 | 7.0 | 41321 | 4.4492 | 0.2831 |
| 4.149 | 8.0 | 47224 | 4.4359 | 0.2858 |
| 4.0825 | 9.0 | 53127 | 4.4269 | 0.2875 |
| 4.0236 | 10.0 | 59030 | 4.4253 | 0.2881 |
Framework versions
- Transformers 4.41.2
- Pytorch 2.8.0+cu128
- Datasets 4.1.1
- Tokenizers 0.19.1
