VIT-ASVspoof2019-ConstantQ-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: 0.2115
- Accuracy: 0.9560
- F1: 0.9750
- Precision: 0.9950
- Recall: 0.9557
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 |
|---|---|---|---|---|---|---|---|
| 0.0383 | 1.0 | 3173 | 0.1192 | 0.9753 | 0.9864 | 0.9734 | 0.9997 |
| 0.0158 | 2.0 | 6346 | 0.0505 | 0.9888 | 0.9938 | 0.9911 | 0.9965 |
| 0.0021 | 3.0 | 9519 | 0.1042 | 0.9849 | 0.9917 | 0.9836 | 0.9998 |
Framework versions
- Transformers 4.36.2
- Pytorch 2.1.2+cu121
- Datasets 2.15.0
- Tokenizers 0.15.0