tiny-bert-sst2
This model is a fine-tuned version of google/bert_uncased_L-2_H-128_A-2 on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 1.2398
- Accuracy: 0.8211
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: 0.0002
- train_batch_size: 128
- eval_batch_size: 128
- seed: 33
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.03
- num_epochs: 2
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 2.4029 | 0.1898 | 100 | 1.6095 | 0.7856 |
| 1.4393 | 0.3795 | 200 | 1.4015 | 0.7947 |
| 1.1136 | 0.5693 | 300 | 1.2956 | 0.8039 |
| 0.9362 | 0.7590 | 400 | 1.2324 | 0.8177 |
| 0.8388 | 0.9488 | 500 | 1.2880 | 0.8131 |
| 0.7043 | 1.1385 | 600 | 1.3109 | 0.8211 |
| 0.6489 | 1.3283 | 700 | 1.2199 | 0.8303 |
| 0.6396 | 1.5180 | 800 | 1.2270 | 0.8245 |
| 0.6284 | 1.7078 | 900 | 1.2459 | 0.8177 |
| 0.6016 | 1.8975 | 1000 | 1.2398 | 0.8211 |
Framework versions
- Transformers 4.43.4
- Pytorch 2.3.1+cu121
- Datasets 2.20.0
- Tokenizers 0.19.1