dnrti_deberta
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: 0.2586
- Precision: 0.7684
- Recall: 0.7978
- F1: 0.7828
- Accuracy: 0.9373
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: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10.0
Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|---|---|---|---|---|---|---|---|
| 0.5854 | 0.76 | 500 | 0.4097 | 0.5198 | 0.6403 | 0.5738 | 0.8704 |
| 0.3165 | 1.52 | 1000 | 0.3137 | 0.6396 | 0.6891 | 0.6634 | 0.9029 |
| 0.2535 | 2.28 | 1500 | 0.2948 | 0.6330 | 0.7379 | 0.6814 | 0.9087 |
| 0.2034 | 3.04 | 2000 | 0.2660 | 0.7062 | 0.7494 | 0.7272 | 0.9210 |
| 0.1596 | 3.81 | 2500 | 0.2682 | 0.7071 | 0.7729 | 0.7386 | 0.9232 |
| 0.1269 | 4.57 | 3000 | 0.2595 | 0.7285 | 0.7959 | 0.7607 | 0.9292 |
| 0.1122 | 5.33 | 3500 | 0.2642 | 0.7468 | 0.7826 | 0.7643 | 0.9321 |
| 0.0923 | 6.09 | 4000 | 0.2677 | 0.7513 | 0.8070 | 0.7781 | 0.9352 |
| 0.0804 | 6.85 | 4500 | 0.2586 | 0.7684 | 0.7978 | 0.7828 | 0.9373 |
| 0.0687 | 7.61 | 5000 | 0.2791 | 0.7522 | 0.8236 | 0.7863 | 0.9368 |
| 0.0626 | 8.37 | 5500 | 0.2789 | 0.7614 | 0.8204 | 0.7898 | 0.9379 |
| 0.0552 | 9.13 | 6000 | 0.2774 | 0.7674 | 0.8208 | 0.7932 | 0.9405 |
| 0.0488 | 9.89 | 6500 | 0.2808 | 0.7742 | 0.8213 | 0.7970 | 0.9415 |
Framework versions
- Transformers 4.36.0.dev0
- Pytorch 2.1.0+cu118
- Datasets 2.14.6
- Tokenizers 0.14.1
- Downloads last month
- 7
Inference Providers
NEW
This model is not currently available via any of the supported third-party Inference Providers, and
the model is not deployed on the HF Inference API.
Model tree for Cyber-ThreaD/DeBERTa-DNRTI
Base model
microsoft/deberta-v3-base