deberta-v3-base-uner-full
This model is a fine-tuned version of microsoft/deberta-v3-base on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.0981
- F1: 0.8316
- Precision: 0.8202
- Recall: 0.8432
- Accuracy: 0.9856
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: 2.5e-05
- train_batch_size: 16
- eval_batch_size: 32
- seed: 42
- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- num_epochs: 10
Training results
| Training Loss | Epoch | Step | Validation Loss | F1 | Precision | Recall | Accuracy |
|---|---|---|---|---|---|---|---|
| 0.0012 | 1.0 | 734 | 0.0583 | 0.8017 | 0.7887 | 0.8151 | 0.9837 |
| 0.0002 | 2.0 | 1468 | 0.0625 | 0.8136 | 0.7975 | 0.8303 | 0.9846 |
| 0.0003 | 3.0 | 2202 | 0.0674 | 0.8111 | 0.7841 | 0.84 | 0.9838 |
| 0.0 | 4.0 | 2936 | 0.0715 | 0.8281 | 0.8155 | 0.8411 | 0.9854 |
| 0.0031 | 5.0 | 3670 | 0.0794 | 0.8297 | 0.8196 | 0.84 | 0.9856 |
| 0.0001 | 6.0 | 4404 | 0.0796 | 0.8320 | 0.8160 | 0.8486 | 0.9854 |
| 0.0 | 7.0 | 5138 | 0.0868 | 0.8262 | 0.8149 | 0.8378 | 0.9855 |
| 0.0001 | 8.0 | 5872 | 0.0911 | 0.8292 | 0.8116 | 0.8476 | 0.9857 |
| 0.0001 | 9.0 | 6606 | 0.0957 | 0.8321 | 0.8182 | 0.8465 | 0.9857 |
| 0.0001 | 10.0 | 7340 | 0.0981 | 0.8316 | 0.8202 | 0.8432 | 0.9856 |
Framework versions
- Transformers 4.57.1
- Pytorch 2.8.0+cu128
- Datasets 4.3.0
- Tokenizers 0.22.1