gpt2_small_ZH_superbpe_32768_parallel3-100_42
This model was trained from scratch on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 4.7482
- Accuracy: 0.2493
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 |
|---|
| 5.9459 | 1.0 | 5312 | 5.7063 | 0.1797 |
| 5.386 | 2.0 | 10624 | 5.2386 | 0.2100 |
| 5.1558 | 3.0 | 15936 | 5.0487 | 0.2234 |
| 5.0006 | 4.0 | 21248 | 4.9491 | 0.2313 |
| 4.9091 | 5.0 | 26560 | 4.8806 | 0.2367 |
| 4.8424 | 6.0 | 31872 | 4.8317 | 0.2415 |
| 4.7835 | 7.0 | 37184 | 4.7975 | 0.2444 |
| 4.7398 | 8.0 | 42496 | 4.7730 | 0.2474 |
| 4.6997 | 9.0 | 47808 | 4.7564 | 0.2485 |
| 4.671 | 10.0 | 53120 | 4.7482 | 0.2493 |
Framework versions
- Transformers 4.41.2
- Pytorch 2.8.0+cu128
- Datasets 4.1.1
- Tokenizers 0.19.1
