gpt2_small_AR_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: 3.4253
- Accuracy: 0.4040
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 |
|---|
| 3.9037 | 1.0 | 27570 | 3.8413 | 0.3612 |
| 3.7205 | 2.0 | 55140 | 3.6738 | 0.3760 |
| 3.6312 | 3.0 | 82710 | 3.5939 | 0.3842 |
| 3.5745 | 4.0 | 110280 | 3.5476 | 0.3891 |
| 3.5312 | 5.0 | 137850 | 3.5132 | 0.3933 |
| 3.4999 | 6.0 | 165420 | 3.4871 | 0.3959 |
| 3.4709 | 7.0 | 192990 | 3.4639 | 0.3986 |
| 3.4413 | 8.0 | 220560 | 3.4475 | 0.4007 |
| 3.4162 | 9.0 | 248130 | 3.4333 | 0.4031 |
| 3.3951 | 10.0 | 275700 | 3.4253 | 0.4040 |
Framework versions
- Transformers 4.41.2
- Pytorch 2.8.0+cu128
- Datasets 4.1.1
- Tokenizers 0.19.1
