gpt2_small_KO_unigram_8192_parallel10_42
This model was trained from scratch on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 3.7315
- Accuracy: 0.3514
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: 117
- 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 |
|---|
| 5.9251 | 1.0 | 1795 | 4.5055 | 0.2865 |
| 4.38 | 2.0 | 3590 | 4.1538 | 0.3109 |
| 4.1047 | 3.0 | 5385 | 3.9981 | 0.3255 |
| 3.9618 | 4.0 | 7180 | 3.9075 | 0.3333 |
| 3.8989 | 5.0 | 8975 | 3.8465 | 0.3390 |
| 3.8152 | 6.0 | 10770 | 3.8031 | 0.3425 |
| 3.7585 | 7.0 | 12565 | 3.7738 | 0.3468 |
| 3.7137 | 8.0 | 14360 | 3.7528 | 0.3489 |
| 3.6795 | 9.0 | 16155 | 3.7374 | 0.3505 |
| 3.657 | 10.0 | 17950 | 3.7315 | 0.3514 |
Framework versions
- Transformers 4.41.2
- Pytorch 2.8.0+cu128
- Datasets 4.1.1
- Tokenizers 0.19.1
