tiny-mlm-glue-qnli-target-glue-sst2
This model is a fine-tuned version of muhtasham/tiny-mlm-glue-qnli on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.5008
- 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: 3e-05
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: constant
- num_epochs: 200
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 0.5757 | 0.24 | 500 | 0.4901 | 0.7775 |
| 0.4436 | 0.48 | 1000 | 0.4673 | 0.7833 |
| 0.3947 | 0.71 | 1500 | 0.4434 | 0.7970 |
| 0.3751 | 0.95 | 2000 | 0.4601 | 0.7970 |
| 0.3326 | 1.19 | 2500 | 0.4463 | 0.8005 |
| 0.316 | 1.43 | 3000 | 0.4510 | 0.8005 |
| 0.2981 | 1.66 | 3500 | 0.4367 | 0.8142 |
| 0.2929 | 1.9 | 4000 | 0.4383 | 0.8108 |
| 0.2746 | 2.14 | 4500 | 0.4873 | 0.8016 |
| 0.256 | 2.38 | 5000 | 0.4395 | 0.8165 |
| 0.246 | 2.61 | 5500 | 0.4444 | 0.8280 |
| 0.2522 | 2.85 | 6000 | 0.4478 | 0.8245 |
| 0.2371 | 3.09 | 6500 | 0.4556 | 0.8291 |
| 0.2299 | 3.33 | 7000 | 0.4655 | 0.8326 |
| 0.2143 | 3.56 | 7500 | 0.4581 | 0.8314 |
| 0.2153 | 3.8 | 8000 | 0.4869 | 0.8291 |
| 0.2134 | 4.04 | 8500 | 0.5008 | 0.8211 |
Framework versions
- Transformers 4.26.0.dev0
- Pytorch 1.13.0+cu116
- Datasets 2.8.1.dev0
- Tokenizers 0.13.2