VIT-ASVspoof5-Mel_Spectrogram-Synthetic-Voice-Detection
This model is a fine-tuned version of google/vit-base-patch16-224-in21k on the imagefolder dataset. It achieves the following results on the evaluation set:
- Loss: 2.0728
- Accuracy: 0.7633
- F1: 0.8264
- Precision: 0.9621
- Recall: 0.7242
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: 5e-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: 3.0
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 | Precision | Recall |
|---|---|---|---|---|---|---|---|
| 0.0047 | 1.0 | 22795 | 0.9664 | 0.8373 | 0.8919 | 0.9221 | 0.8637 |
| 0.0064 | 2.0 | 45590 | 1.6013 | 0.7830 | 0.8421 | 0.9701 | 0.7439 |
| 0.0 | 3.0 | 68385 | 2.0728 | 0.7633 | 0.8264 | 0.9621 | 0.7242 |
Framework versions
- Transformers 4.44.0
- Pytorch 2.4.0+cu124
- Datasets 2.21.0
- Tokenizers 0.19.1