gpt2_small_KO_superbpe_81920_parallel10_42
This model was trained from scratch on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 6.1162
- Accuracy: 0.2115
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: 64
- eval_batch_size: 64
- seed: 42
- distributed_type: multi-GPU
- num_devices: 2
- total_train_batch_size: 128
- total_eval_batch_size: 128
- 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 |
|---|
| 8.5996 | 1.0 | 1064 | 7.7511 | 0.1671 |
| 7.415 | 2.0 | 2128 | 7.0227 | 0.1810 |
| 6.9191 | 3.0 | 3192 | 6.7040 | 0.1895 |
| 6.6521 | 4.0 | 4256 | 6.5098 | 0.1959 |
| 6.4669 | 5.0 | 5320 | 6.3722 | 0.2003 |
| 6.3279 | 6.0 | 6384 | 6.2735 | 0.2045 |
| 6.2196 | 7.0 | 7448 | 6.2027 | 0.2078 |
| 6.1376 | 8.0 | 8512 | 6.1561 | 0.2096 |
| 6.0762 | 9.0 | 9576 | 6.1280 | 0.2112 |
| 6.0341 | 10.0 | 10640 | 6.1162 | 0.2115 |
Framework versions
- Transformers 4.41.2
- Pytorch 2.8.0+cu128
- Datasets 4.1.1
- Tokenizers 0.19.1
