gpt2_medium_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: 4.2277
- Accuracy: 0.3050
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 |
|---|
| 4.7994 | 1.0 | 20365 | 4.7229 | 0.2501 |
| 4.5664 | 2.0 | 40730 | 4.5137 | 0.2705 |
| 4.4365 | 3.0 | 61095 | 4.4201 | 0.2797 |
| 4.3506 | 4.0 | 81460 | 4.3593 | 0.2862 |
| 4.2813 | 5.0 | 101825 | 4.3183 | 0.2909 |
| 4.2213 | 6.0 | 122190 | 4.2876 | 0.2957 |
| 4.1688 | 7.0 | 142555 | 4.2629 | 0.2983 |
| 4.1191 | 8.0 | 162920 | 4.2460 | 0.3015 |
| 4.0746 | 9.0 | 183285 | 4.2344 | 0.3036 |
| 4.0279 | 10.0 | 203650 | 4.2277 | 0.3050 |
Framework versions
- Transformers 4.41.2
- Pytorch 2.8.0+cu128
- Datasets 4.1.1
- Tokenizers 0.19.1
