gpt2_small_AR_unigram_32768_parallel10_42
This model was trained from scratch on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 4.2953
- Accuracy: 0.3355
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: 103
- 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 |
|---|
| 6.6236 | 1.0 | 1839 | 5.1730 | 0.2828 |
| 5.0422 | 2.0 | 3678 | 4.7757 | 0.3043 |
| 4.728 | 3.0 | 5517 | 4.6040 | 0.3130 |
| 4.5614 | 4.0 | 7356 | 4.4954 | 0.3205 |
| 4.4496 | 5.0 | 9195 | 4.4260 | 0.3246 |
| 4.3688 | 6.0 | 11034 | 4.3760 | 0.3286 |
| 4.3146 | 7.0 | 12873 | 4.3413 | 0.3315 |
| 4.261 | 8.0 | 14712 | 4.3165 | 0.3333 |
| 4.2211 | 9.0 | 16551 | 4.3013 | 0.3351 |
| 4.192 | 10.0 | 18390 | 4.2953 | 0.3355 |
Framework versions
- Transformers 4.41.2
- Pytorch 2.8.0+cu128
- Datasets 4.1.1
- Tokenizers 0.19.1
