gpt2_medium_ZH_unigram_32768_parallel3-100_42
This model was trained from scratch on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 4.2762
- Accuracy: 0.3175
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 |
|---|
| 5.1912 | 1.0 | 5678 | 4.9708 | 0.2560 |
| 4.6834 | 2.0 | 11356 | 4.6124 | 0.2821 |
| 4.4718 | 3.0 | 17034 | 4.4688 | 0.2951 |
| 4.347 | 4.0 | 22712 | 4.3874 | 0.3021 |
| 4.2359 | 5.0 | 28390 | 4.3389 | 0.3074 |
| 4.1428 | 6.0 | 34068 | 4.3070 | 0.3108 |
| 4.0636 | 7.0 | 39746 | 4.2899 | 0.3140 |
| 3.9891 | 8.0 | 45424 | 4.2795 | 0.3154 |
| 3.9264 | 9.0 | 51102 | 4.2766 | 0.3168 |
| 3.8726 | 10.0 | 56780 | 4.2762 | 0.3175 |
Framework versions
- Transformers 4.41.2
- Pytorch 2.8.0+cu128
- Datasets 4.1.1
- Tokenizers 0.19.1
