gpt2_small_AR_unigram_65536_parallel3-100_42
This model was trained from scratch on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 3.7532
- Accuracy: 0.3575
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 |
|---|
| 4.558 | 1.0 | 7033 | 4.4275 | 0.3019 |
| 4.2085 | 2.0 | 14066 | 4.1171 | 0.3235 |
| 4.0526 | 3.0 | 21099 | 3.9853 | 0.3347 |
| 3.9522 | 4.0 | 28132 | 3.9077 | 0.3411 |
| 3.8816 | 5.0 | 35165 | 3.8561 | 0.3463 |
| 3.825 | 6.0 | 42198 | 3.8192 | 0.3497 |
| 3.7821 | 7.0 | 49231 | 3.7934 | 0.3523 |
| 3.7432 | 8.0 | 56264 | 3.7727 | 0.3548 |
| 3.7109 | 9.0 | 63297 | 3.7602 | 0.3564 |
| 3.6848 | 10.0 | 70330 | 3.7532 | 0.3575 |
Framework versions
- Transformers 4.41.2
- Pytorch 2.8.0+cu128
- Datasets 4.1.1
- Tokenizers 0.19.1
