crossencoder-airline-refine-030
This model is a fine-tuned version of cross-encoder/stsb-roberta-large on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.0451
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: 5e-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: cosine
- num_epochs: 30
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 0.312 | 1.0 | 159 | 0.3001 |
| 0.197 | 2.0 | 318 | 0.3300 |
| 0.2195 | 3.0 | 477 | 0.1976 |
| 0.5386 | 4.0 | 636 | 0.1339 |
| 0.1417 | 5.0 | 795 | 0.1625 |
| 0.2544 | 6.0 | 954 | 0.5143 |
| 0.0716 | 7.0 | 1113 | 0.1888 |
| 0.1214 | 8.0 | 1272 | 0.1115 |
| 0.1658 | 9.0 | 1431 | 0.1007 |
| 0.1489 | 10.0 | 1590 | 0.1735 |
| 0.1163 | 11.0 | 1749 | 0.0781 |
| 0.1411 | 12.0 | 1908 | 0.1386 |
| 0.1636 | 13.0 | 2067 | 0.1462 |
| 0.1122 | 14.0 | 2226 | 0.2000 |
| 0.1287 | 15.0 | 2385 | 0.1702 |
| 0.1954 | 16.0 | 2544 | 0.1191 |
| 0.0632 | 17.0 | 2703 | 0.1155 |
| 0.0377 | 18.0 | 2862 | 0.0979 |
| 0.0714 | 19.0 | 3021 | 0.0810 |
| 0.0727 | 20.0 | 3180 | 0.0710 |
| 0.0449 | 21.0 | 3339 | 0.0832 |
| 0.0964 | 22.0 | 3498 | 0.0721 |
| 0.0276 | 23.0 | 3657 | 0.0516 |
| 0.0269 | 24.0 | 3816 | 0.0601 |
| 0.0483 | 25.0 | 3975 | 0.0583 |
| 0.0442 | 26.0 | 4134 | 0.0627 |
| 0.0181 | 27.0 | 4293 | 0.0704 |
| 0.0177 | 28.0 | 4452 | 0.0631 |
| 0.0315 | 29.0 | 4611 | 0.0614 |
| 0.0184 | 30.0 | 4770 | 0.0623 |
Framework versions
- Transformers 4.38.1
- Pytorch 2.0.1
- Datasets 2.17.1
- Tokenizers 0.15.2