wav2vec2-base-960h-finetuned-deepfake
This model is a fine-tuned version of facebook/wav2vec2-base-960h on the asvspoof2019 dataset. It achieves the following results on the evaluation set:
- Loss: 0.0009
- Accuracy: 0.9998
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: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 5
Training results
Training Loss | Epoch | Step | Validation Loss | Accuracy |
---|---|---|---|---|
0.0027 | 1.0 | 793 | 0.0070 | 0.9990 |
0.0006 | 2.0 | 1586 | 0.0032 | 0.9995 |
0.0271 | 3.0 | 2379 | 0.0022 | 0.9995 |
0.0002 | 4.0 | 3172 | 0.0009 | 0.9998 |
0.0002 | 5.0 | 3965 | 0.0020 | 0.9998 |
Framework versions
- Transformers 4.22.1
- Pytorch 1.12.1+cu113
- Datasets 2.4.0
- Tokenizers 0.12.1
- Downloads last month
- 108
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social
visibility and check back later, or deploy to Inference Endpoints (dedicated)
instead.