deepfake-audio-detector-v3
This model is a fine-tuned version of Heem2/Deepfake-audio-detection on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.2038
- Accuracy: 0.9850
- Precision: 0.9742
- Recall: 0.9716
- F1: 0.9729
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: 3e-05
- train_batch_size: 4
- eval_batch_size: 4
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 8
- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 100
- num_epochs: 5
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | Recall | F1 |
|---|---|---|---|---|---|---|---|
| 0.574 | 1.0 | 699 | 0.4368 | 0.9700 | 0.9459 | 0.9459 | 0.9459 |
| 0.7396 | 2.0 | 1398 | 0.3613 | 0.9728 | 0.9581 | 0.9433 | 0.9506 |
| 0.3513 | 3.0 | 2097 | 0.3411 | 0.9757 | 0.9810 | 0.9304 | 0.9550 |
| 0.0 | 4.0 | 2796 | 0.2747 | 0.9821 | 0.9739 | 0.9613 | 0.9676 |
| 0.0 | 5.0 | 3495 | 0.2038 | 0.9850 | 0.9742 | 0.9716 | 0.9729 |
Framework versions
- Transformers 4.51.3
- Pytorch 2.6.0+cu124
- Datasets 3.6.0
- Tokenizers 0.21.1
- Downloads last month
- 5