sts_roberta_large_lr1e-05_wd1e-03_ep10_ckpt
This model is a fine-tuned version of klue/roberta-large on the klue dataset. It achieves the following results on the evaluation set:
- Loss: 0.3764
- Mse: 0.3764
- Mae: 0.4512
- R2: 0.8277
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: 10
Training results
| Training Loss | Epoch | Step | Validation Loss | Mse | Mae | R2 |
|---|---|---|---|---|---|---|
| 2.3411 | 1.0 | 183 | 1.0407 | 1.0407 | 0.7779 | 0.5234 |
| 0.1725 | 2.0 | 366 | 0.3938 | 0.3938 | 0.4710 | 0.8197 |
| 0.1203 | 3.0 | 549 | 0.3972 | 0.3972 | 0.4594 | 0.8181 |
| 0.093 | 4.0 | 732 | 0.4030 | 0.4030 | 0.4675 | 0.8155 |
| 0.0727 | 5.0 | 915 | 0.4102 | 0.4102 | 0.4690 | 0.8122 |
| 0.0591 | 6.0 | 1098 | 0.3700 | 0.3700 | 0.4470 | 0.8306 |
| 0.0482 | 7.0 | 1281 | 0.3578 | 0.3578 | 0.4403 | 0.8362 |
| 0.0417 | 8.0 | 1464 | 0.4042 | 0.4042 | 0.4696 | 0.8149 |
| 0.037 | 9.0 | 1647 | 0.4151 | 0.4151 | 0.4753 | 0.8099 |
| 0.0337 | 10.0 | 1830 | 0.3764 | 0.3764 | 0.4512 | 0.8277 |
Framework versions
- Transformers 4.33.2
- Pytorch 2.0.1+cu117
- Datasets 2.13.0
- Tokenizers 0.13.3