gpt2_small_AR_bpe_65536_parallel3_42
This model was trained from scratch on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 3.4259
- Accuracy: 0.4036
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 |
|---|
| 3.9041 | 1.0 | 27570 | 3.8407 | 0.3613 |
| 3.7212 | 2.0 | 55140 | 3.6756 | 0.3758 |
| 3.6318 | 3.0 | 82710 | 3.5955 | 0.3838 |
| 3.5749 | 4.0 | 110280 | 3.5473 | 0.3892 |
| 3.5317 | 5.0 | 137850 | 3.5141 | 0.3927 |
| 3.5001 | 6.0 | 165420 | 3.4874 | 0.3959 |
| 3.4713 | 7.0 | 192990 | 3.4646 | 0.3987 |
| 3.442 | 8.0 | 220560 | 3.4484 | 0.4004 |
| 3.417 | 9.0 | 248130 | 3.4342 | 0.4026 |
| 3.396 | 10.0 | 275700 | 3.4259 | 0.4036 |
Framework versions
- Transformers 4.41.2
- Pytorch 2.8.0+cu128
- Datasets 4.1.1
- Tokenizers 0.19.1
