gpt2_small_EN_unigram_81920_parallel10_42
This model was trained from scratch on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 3.6029
- Accuracy: 0.3865
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 |
|---|
| 5.9773 | 1.0 | 1606 | 4.2379 | 0.3418 |
| 4.0857 | 2.0 | 3212 | 3.9507 | 0.3602 |
| 3.9479 | 3.0 | 4818 | 3.8225 | 0.3691 |
| 3.7916 | 4.0 | 6424 | 3.7461 | 0.3747 |
| 3.6981 | 5.0 | 8030 | 3.6961 | 0.3788 |
| 3.6452 | 6.0 | 9636 | 3.6581 | 0.3814 |
| 3.5884 | 7.0 | 11242 | 3.6346 | 0.3834 |
| 3.5574 | 8.0 | 12848 | 3.6178 | 0.3853 |
| 3.5178 | 9.0 | 14454 | 3.6076 | 0.3861 |
| 3.4928 | 10.0 | 16060 | 3.6029 | 0.3865 |
Framework versions
- Transformers 4.41.2
- Pytorch 2.8.0+cu128
- Datasets 4.1.1
- Tokenizers 0.19.1
