crossencoder-airline-refine-010
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: 8.0523
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-08
- 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
- lr_scheduler_warmup_steps: 500
- num_epochs: 10
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 15.341 | 1.0 | 157 | 14.5631 |
| 12.2879 | 2.0 | 314 | 13.3058 |
| 12.5681 | 3.0 | 471 | 11.4717 |
| 12.8002 | 4.0 | 628 | 9.8398 |
| 10.1409 | 5.0 | 785 | 8.8337 |
| 9.4818 | 6.0 | 942 | 8.1771 |
| 9.277 | 7.0 | 1099 | 7.7594 |
| 9.2643 | 8.0 | 1256 | 7.5311 |
| 8.7124 | 9.0 | 1413 | 7.4428 |
| 8.9775 | 10.0 | 1570 | 7.4347 |
Framework versions
- Transformers 4.38.1
- Pytorch 2.0.1
- Datasets 2.17.1
- Tokenizers 0.15.2