AST-ASVspoof5-Synthetic-Voice-Detection

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: 2.2821
Accuracy: 0.8333
F1: 0.8892
Precision: 0.9209
Recall: 0.8595

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
0.0042	1.0	22795	1.6954	0.8470	0.8942	0.9672	0.8314
0.0	2.0	45590	1.5632	0.8489	0.9014	0.9157	0.8875
0.0	3.0	68385	2.2821	0.8333	0.8892	0.9209	0.8595

Framework versions

Transformers 4.42.4
Pytorch 2.3.1+cu121
Datasets 2.20.0
Tokenizers 0.19.1

Downloads last month: 4,420

Safetensors

Model size

86.2M params

Tensor type

F32

Model tree for MattyB95/AST-ASVspoof5-Synthetic-Voice-Detection

Base model

MIT/ast-finetuned-audioset-10-10-0.4593

Finetuned

(184)

this model

Space using MattyB95/AST-ASVspoof5-Synthetic-Voice-Detection 1

Collection including MattyB95/AST-ASVspoof5-Synthetic-Voice-Detection

Jabberjay

Collection

🦜 Synthetic Voice Detection • 15 items • Updated Aug 15, 2024 • 2

Evaluation results

Accuracy on audiofolder
validation set self-reported

0.833
F1 on audiofolder
validation set self-reported

0.889
Precision on audiofolder
validation set self-reported

0.921
Recall on audiofolder
validation set self-reported

0.860