CS221-deberta-base-finetuned-semeval-NT
This model is a fine-tuned version of microsoft/deberta-base on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.5533
- F1: 0.7588
- Roc Auc: 0.8183
- Accuracy: 0.4693
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: 2e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 100
- num_epochs: 20
Training results
| Training Loss | Epoch | Step | Validation Loss | F1 | Roc Auc | Accuracy |
|---|---|---|---|---|---|---|
| 0.4331 | 1.0 | 277 | 0.3934 | 0.7185 | 0.7878 | 0.3845 |
| 0.3183 | 2.0 | 554 | 0.3692 | 0.7383 | 0.8014 | 0.4458 |
| 0.1905 | 3.0 | 831 | 0.4011 | 0.7447 | 0.8045 | 0.4819 |
| 0.1716 | 4.0 | 1108 | 0.4457 | 0.7489 | 0.8106 | 0.4531 |
| 0.0954 | 5.0 | 1385 | 0.4980 | 0.7573 | 0.8190 | 0.4567 |
| 0.075 | 6.0 | 1662 | 0.5533 | 0.7588 | 0.8183 | 0.4693 |
| 0.0442 | 7.0 | 1939 | 0.6536 | 0.7360 | 0.7985 | 0.4531 |
| 0.0075 | 8.0 | 2216 | 0.6831 | 0.7539 | 0.8135 | 0.4675 |
| 0.0111 | 9.0 | 2493 | 0.7289 | 0.7529 | 0.8124 | 0.4693 |
Framework versions
- Transformers 4.47.1
- Pytorch 2.5.1+cu121
- Datasets 3.2.0
- Tokenizers 0.21.0