gpt2_small_TR_unigram_32768_parallel10_42
This model was trained from scratch on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 4.3110
- Accuracy: 0.3560
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 |
|---|
| 6.6258 | 1.0 | 1343 | 5.3908 | 0.2946 |
| 5.3619 | 2.0 | 2686 | 4.8868 | 0.3191 |
| 4.7776 | 3.0 | 4029 | 4.6776 | 0.3296 |
| 4.6358 | 4.0 | 5372 | 4.5511 | 0.3377 |
| 4.5404 | 5.0 | 6715 | 4.4686 | 0.3432 |
| 4.4072 | 6.0 | 8058 | 4.4094 | 0.3480 |
| 4.3474 | 7.0 | 9401 | 4.3667 | 0.3510 |
| 4.3096 | 8.0 | 10744 | 4.3381 | 0.3534 |
| 4.2487 | 9.0 | 12087 | 4.3189 | 0.3554 |
| 4.224 | 10.0 | 13430 | 4.3110 | 0.3560 |
Framework versions
- Transformers 4.41.2
- Pytorch 2.8.0+cu128
- Datasets 4.1.1
- Tokenizers 0.19.1
