gpt2_small_ZH_bpe_65536_parallel3-100_42
This model was trained from scratch on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 4.5390
- Accuracy: 0.3101
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 |
|---|
| 5.73 | 1.0 | 5523 | 5.4857 | 0.2480 |
| 5.1575 | 2.0 | 11046 | 5.0391 | 0.2728 |
| 4.9434 | 3.0 | 16569 | 4.8439 | 0.2854 |
| 4.7914 | 4.0 | 22092 | 4.7375 | 0.2935 |
| 4.7018 | 5.0 | 27615 | 4.6656 | 0.2986 |
| 4.6239 | 6.0 | 33138 | 4.6193 | 0.3025 |
| 4.5699 | 7.0 | 38661 | 4.5857 | 0.3060 |
| 4.5209 | 8.0 | 44184 | 4.5622 | 0.3078 |
| 4.4839 | 9.0 | 49707 | 4.5463 | 0.3092 |
| 4.4546 | 10.0 | 55230 | 4.5390 | 0.3101 |
Framework versions
- Transformers 4.41.2
- Pytorch 2.8.0+cu128
- Datasets 4.1.1
- Tokenizers 0.19.1
