sts_nli_roberta-large_lr1e-05_wd1e-03_ep3_lr1e-05_wd1e-03_ep9_ckpt
This model is a fine-tuned version of ys7yoo/nli_roberta-large_lr1e-05_wd1e-03_ep3 on the klue dataset. It achieves the following results on the evaluation set:
- Loss: 0.3250
- Mse: 0.3250
- Mae: 0.4166
- R2: 0.8512
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: 1e-05
- train_batch_size: 64
- eval_batch_size: 64
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 9
Training results
| Training Loss | Epoch | Step | Validation Loss | Mse | Mae | R2 |
|---|---|---|---|---|---|---|
| 1.2084 | 1.0 | 183 | 0.5071 | 0.5071 | 0.5306 | 0.7678 |
| 0.1515 | 2.0 | 366 | 0.3142 | 0.3142 | 0.4149 | 0.8561 |
| 0.103 | 3.0 | 549 | 0.3284 | 0.3284 | 0.4150 | 0.8496 |
| 0.0779 | 4.0 | 732 | 0.3306 | 0.3306 | 0.4184 | 0.8486 |
| 0.0597 | 5.0 | 915 | 0.3219 | 0.3219 | 0.4098 | 0.8526 |
| 0.0497 | 6.0 | 1098 | 0.3324 | 0.3324 | 0.4175 | 0.8478 |
| 0.0407 | 7.0 | 1281 | 0.3114 | 0.3114 | 0.4119 | 0.8574 |
| 0.0356 | 8.0 | 1464 | 0.3305 | 0.3305 | 0.4199 | 0.8486 |
| 0.0327 | 9.0 | 1647 | 0.3250 | 0.3250 | 0.4166 | 0.8512 |
Framework versions
- Transformers 4.33.2
- Pytorch 2.0.1+cu117
- Datasets 2.13.0
- Tokenizers 0.13.3