---
license: apache-2.0
base_model: Bisher/wav2vec2_ASV_deepfake_audio_detection
tags:
- generated_from_trainer
metrics:
- accuracy
- precision
- recall
- f1
model-index:
- name: wav2vec2_ASV_deepfake_audio_detection_DF_finetune_frozen
  results: []
---

<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->

# wav2vec2_ASV_deepfake_audio_detection_DF_finetune_frozen

This model is a fine-tuned version of [Bisher/wav2vec2_ASV_deepfake_audio_detection](https://huggingface.co/Bisher/wav2vec2_ASV_deepfake_audio_detection) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.4076
- Accuracy: 0.9115
- Precision: 0.9131
- Recall: 0.9115
- F1: 0.8803
- Tp: 265
- Tn: 17893
- Fn: 1742
- Fp: 20
- Eer: 0.0588
- Min Tdcf: 0.0271
- Auc Roc: 0.9849

## 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: 152
- eval_batch_size: 152
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 608
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 3
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch  | Step | Validation Loss | Accuracy | Precision | Recall | F1     | Tp  | Tn    | Fn   | Fp | Eer    | Min Tdcf | Auc Roc |
|:-------------:|:------:|:----:|:---------------:|:--------:|:---------:|:------:|:------:|:---:|:-----:|:----:|:--:|:------:|:--------:|:-------:|
| 0.025         | 0.0490 | 10   | 0.3573          | 0.9139   | 0.9168    | 0.9139 | 0.8847 | 308 | 17896 | 1699 | 17 | 0.0826 | 0.0329   | 0.9732  |
| 0.0219        | 0.0979 | 20   | 0.3002          | 0.9152   | 0.9172    | 0.9152 | 0.8875 | 339 | 17891 | 1668 | 22 | 0.0757 | 0.0321   | 0.9759  |
| 0.0193        | 0.1469 | 30   | 0.3557          | 0.9159   | 0.9178    | 0.9159 | 0.8889 | 354 | 17890 | 1653 | 23 | 0.0710 | 0.0320   | 0.9541  |
| 0.0197        | 0.1958 | 40   | 0.3963          | 0.9191   | 0.9206    | 0.9191 | 0.8951 | 423 | 17885 | 1584 | 28 | 0.0802 | 0.0325   | 0.9453  |
| 0.0181        | 0.2448 | 50   | 0.3524          | 0.9155   | 0.9176    | 0.9155 | 0.8882 | 346 | 17891 | 1661 | 22 | 0.0673 | 0.0324   | 0.9794  |
| 0.0152        | 0.2938 | 60   | 0.5161          | 0.9050   | 0.9102    | 0.9050 | 0.8654 | 119 | 17908 | 1888 | 5  | 0.0857 | 0.0314   | 0.9582  |
| 0.0155        | 0.3427 | 70   | 0.6375          | 0.9037   | 0.9120    | 0.9037 | 0.8622 | 90  | 17912 | 1917 | 1  | 0.1395 | 0.0307   | 0.9384  |
| 0.0193        | 0.3917 | 80   | 0.3521          | 0.9119   | 0.9153    | 0.9119 | 0.8808 | 267 | 17899 | 1740 | 14 | 0.0643 | 0.0309   | 0.9816  |
| 0.0158        | 0.4406 | 90   | 0.3775          | 0.9066   | 0.9113    | 0.9066 | 0.8692 | 154 | 17906 | 1853 | 7  | 0.0613 | 0.0290   | 0.9839  |
| 0.017         | 0.4896 | 100  | 0.4076          | 0.9115   | 0.9131    | 0.9115 | 0.8803 | 265 | 17893 | 1742 | 20 | 0.0588 | 0.0271   | 0.9849  |


### Framework versions

- Transformers 4.44.0
- Pytorch 2.4.0
- Datasets 2.21.0
- Tokenizers 0.19.1