AST-ASVspoof2019-Synthetic-Voice-Detection-New
This model is a fine-tuned version of MIT/ast-finetuned-audioset-10-10-0.4593 on the audiofolder dataset. It achieves the following results on the evaluation set:
- Loss: 0.0213
- Accuracy: 0.9971
- F1: 0.9984
- Precision: 0.9968
- Recall: 0.9999
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 |
---|---|---|---|---|---|---|---|
0.0232 | 1.0 | 3173 | 0.0404 | 0.9932 | 0.9962 | 0.9934 | 0.9991 |
0.0058 | 2.0 | 6346 | 0.0383 | 0.9931 | 0.9962 | 0.9927 | 0.9996 |
0.0014 | 3.0 | 9519 | 0.0213 | 0.9971 | 0.9984 | 0.9968 | 0.9999 |
Framework versions
- Transformers 4.37.2
- Pytorch 2.2.0
- Datasets 2.16.1
- Tokenizers 0.15.1
- Downloads last month
- 174
Finetuned from
Space using MattyB95/AST-ASVspoof2019-Synthetic-Voice-Detection-New 1
Evaluation results
- Accuracy on audiofoldervalidation set self-reported0.997
- F1 on audiofoldervalidation set self-reported0.998
- Precision on audiofoldervalidation set self-reported0.997
- Recall on audiofoldervalidation set self-reported1.000