CS221-deberta-base-finetuned-semeval-aug
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: 0.2871
- F1: 0.8364
- Roc Auc: 0.8738
- Accuracy: 0.6387
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: 32
- eval_batch_size: 32
- 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 |
|---|---|---|---|---|---|---|
| 0.2713 | 1.0 | 139 | 0.3886 | 0.7310 | 0.8051 | 0.4228 |
| 0.2914 | 2.0 | 278 | 0.3337 | 0.7849 | 0.8464 | 0.5248 |
| 0.2201 | 3.0 | 417 | 0.3073 | 0.7995 | 0.8511 | 0.5501 |
| 0.1444 | 4.0 | 556 | 0.3027 | 0.8275 | 0.8727 | 0.6007 |
| 0.0964 | 5.0 | 695 | 0.2871 | 0.8364 | 0.8738 | 0.6387 |
| 0.0536 | 6.0 | 834 | 0.3024 | 0.8432 | 0.8796 | 0.6612 |
| 0.042 | 7.0 | 973 | 0.2909 | 0.8631 | 0.8960 | 0.6920 |
| 0.0334 | 8.0 | 1112 | 0.2976 | 0.8675 | 0.9002 | 0.7037 |
Framework versions
- Transformers 4.47.1
- Pytorch 2.5.1+cu121
- Datasets 3.2.0
- Tokenizers 0.21.0