gpt2_small_KO_unigram_5994_parallel10_42
This model was trained from scratch on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 3.5146
- Accuracy: 0.3621
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 |
|---|
| 5.6609 | 1.0 | 1746 | 4.2471 | 0.2927 |
| 4.1162 | 2.0 | 3492 | 3.9099 | 0.3207 |
| 3.8579 | 3.0 | 5238 | 3.7643 | 0.3349 |
| 3.7749 | 4.0 | 6984 | 3.6783 | 0.3436 |
| 3.6652 | 5.0 | 8730 | 3.6244 | 0.3497 |
| 3.596 | 6.0 | 10476 | 3.5842 | 0.3536 |
| 3.5428 | 7.0 | 12222 | 3.5561 | 0.3568 |
| 3.5108 | 8.0 | 13968 | 3.5348 | 0.3594 |
| 3.47 | 9.0 | 15714 | 3.5205 | 0.3613 |
| 3.4462 | 10.0 | 17460 | 3.5146 | 0.3621 |
Framework versions
- Transformers 4.41.2
- Pytorch 2.8.0+cu128
- Datasets 4.1.1
- Tokenizers 0.19.1
