bert_distillation_tiny
This model is a fine-tuned version of google/bert_uncased_L-2_H-128_A-2 on the glue dataset. It achieves the following results on the evaluation set:
- Loss: 0.4274
- Accuracy: 0.8257
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: 0.0001
- train_batch_size: 128
- eval_batch_size: 128
- seed: 2023
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 7
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 0.4139 | 1.0 | 527 | 0.4204 | 0.8096 |
| 0.27 | 2.0 | 1054 | 0.4274 | 0.8257 |
| 0.2226 | 3.0 | 1581 | 0.4899 | 0.8245 |
| 0.1931 | 4.0 | 2108 | 0.4961 | 0.8222 |
| 0.1732 | 5.0 | 2635 | 0.5302 | 0.8222 |
| 0.1608 | 6.0 | 3162 | 0.5393 | 0.8234 |
| 0.152 | 7.0 | 3689 | 0.5562 | 0.8177 |
Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu118
- Datasets 2.15.0
- Tokenizers 0.15.0