gpt2_small_DE_unigram_8192_parallel10_42
This model was trained from scratch on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 3.1623
- Accuracy: 0.4127
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 |
|---|
| 5.3725 | 1.0 | 1779 | 3.9192 | 0.3289 |
| 3.7951 | 2.0 | 3558 | 3.5727 | 0.3616 |
| 3.5243 | 3.0 | 5337 | 3.4193 | 0.3794 |
| 3.3837 | 4.0 | 7116 | 3.3294 | 0.3908 |
| 3.321 | 5.0 | 8895 | 3.2714 | 0.3968 |
| 3.2426 | 6.0 | 10674 | 3.2319 | 0.4037 |
| 3.1872 | 7.0 | 12453 | 3.2017 | 0.4067 |
| 3.148 | 8.0 | 14232 | 3.1820 | 0.4097 |
| 3.1142 | 9.0 | 16011 | 3.1692 | 0.4119 |
| 3.0919 | 10.0 | 17790 | 3.1623 | 0.4127 |
Framework versions
- Transformers 4.41.2
- Pytorch 2.8.0+cu128
- Datasets 4.1.1
- Tokenizers 0.19.1
