gpt2_medium_AR_superbpe_32768_parallel3_42
This model was trained from scratch on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 3.6864
- Accuracy: 0.3517
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 |
|---|
| 4.2232 | 1.0 | 24819 | 4.1493 | 0.2979 |
| 4.0093 | 2.0 | 49638 | 3.9649 | 0.3154 |
| 3.8956 | 3.0 | 74457 | 3.8750 | 0.3254 |
| 3.8171 | 4.0 | 99276 | 3.8207 | 0.3320 |
| 3.7527 | 5.0 | 124095 | 3.7801 | 0.3382 |
| 3.7034 | 6.0 | 148914 | 3.7493 | 0.3413 |
| 3.6547 | 7.0 | 173733 | 3.7277 | 0.3446 |
| 3.6129 | 8.0 | 198552 | 3.7082 | 0.3477 |
| 3.5663 | 9.0 | 223371 | 3.6935 | 0.3505 |
| 3.5291 | 10.0 | 248190 | 3.6864 | 0.3517 |
Framework versions
- Transformers 4.41.2
- Pytorch 2.8.0+cu128
- Datasets 4.1.1
- Tokenizers 0.19.1
