gpt2_small_ZH_superbpe_32768_parallel3_42
This model was trained from scratch on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 4.3994
- Accuracy: 0.2817
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 |
|---|
| 5.0212 | 1.0 | 20365 | 4.9177 | 0.2336 |
| 4.7862 | 2.0 | 40730 | 4.7009 | 0.2522 |
| 4.6689 | 3.0 | 61095 | 4.6013 | 0.2610 |
| 4.5992 | 4.0 | 81460 | 4.5434 | 0.2665 |
| 4.5469 | 5.0 | 101825 | 4.5004 | 0.2711 |
| 4.5035 | 6.0 | 122190 | 4.4692 | 0.2740 |
| 4.4692 | 7.0 | 142555 | 4.4436 | 0.2766 |
| 4.4387 | 8.0 | 162920 | 4.4251 | 0.2787 |
| 4.4147 | 9.0 | 183285 | 4.4083 | 0.2806 |
| 4.3879 | 10.0 | 203650 | 4.3994 | 0.2817 |
Framework versions
- Transformers 4.41.2
- Pytorch 2.8.0+cu128
- Datasets 4.1.1
- Tokenizers 0.19.1
