gpt2_small_ZH_bpe_32768_parallel3-100_42
This model was trained from scratch on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 4.2216
- Accuracy: 0.3226
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 |
|---|
| 5.3235 | 1.0 | 5916 | 5.0255 | 0.2583 |
| 4.7835 | 2.0 | 11832 | 4.6389 | 0.2856 |
| 4.5807 | 3.0 | 17748 | 4.4843 | 0.2984 |
| 4.4597 | 4.0 | 23664 | 4.3939 | 0.3053 |
| 4.3746 | 5.0 | 29580 | 4.3354 | 0.3110 |
| 4.3119 | 6.0 | 35496 | 4.2950 | 0.3150 |
| 4.2599 | 7.0 | 41412 | 4.2650 | 0.3181 |
| 4.22 | 8.0 | 47328 | 4.2435 | 0.3201 |
| 4.1876 | 9.0 | 53244 | 4.2293 | 0.3221 |
| 4.1603 | 10.0 | 59160 | 4.2216 | 0.3226 |
Framework versions
- Transformers 4.41.2
- Pytorch 2.8.0+cu128
- Datasets 4.1.1
- Tokenizers 0.19.1
