gpt2_small_TR_unigram_81920_parallel10_42
This model was trained from scratch on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 4.5751
- Accuracy: 0.3634
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 |
|---|
| 6.8908 | 1.0 | 1268 | 5.6868 | 0.3120 |
| 5.5964 | 2.0 | 2536 | 5.1646 | 0.3322 |
| 5.2093 | 3.0 | 3804 | 4.9473 | 0.3428 |
| 4.8732 | 4.0 | 5072 | 4.8151 | 0.3494 |
| 4.7656 | 5.0 | 6340 | 4.7317 | 0.3537 |
| 4.6861 | 6.0 | 7608 | 4.6715 | 0.3567 |
| 4.6238 | 7.0 | 8876 | 4.6292 | 0.3594 |
| 4.5302 | 8.0 | 10144 | 4.5999 | 0.3614 |
| 4.493 | 9.0 | 11412 | 4.5820 | 0.3629 |
| 4.471 | 10.0 | 12680 | 4.5751 | 0.3634 |
Framework versions
- Transformers 4.41.2
- Pytorch 2.8.0+cu128
- Datasets 4.1.1
- Tokenizers 0.19.1
