---
base_model: microsoft/wavlm-base
tags:
- audio-classification
- deepfake
- audio-spoof
- generated_from_trainer
metrics:
- accuracy
model-index:
- name: wavlm-base-960h-asv19-deepfake
  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. -->

# wavlm-base-960h-asv19-deepfake

This model is a fine-tuned version of [microsoft/wavlm-base](https://huggingface.co/microsoft/wavlm-base) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.0161
- Accuracy: 0.9979
- Far: 0.0153
- Frr: 0.0006
- Eer: 0.0080

## 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: 1e-06
- train_batch_size: 2
- eval_batch_size: 2
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 4
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 4.0

### Training results

| Training Loss | Epoch | Step  | Validation Loss | Accuracy | Far    | Frr    | Eer    |
|:-------------:|:-----:|:-----:|:---------------:|:--------:|:------:|:------:|:------:|
| 0.0386        | 0.79  | 5000  | 0.0597          | 0.9895   | 0.1001 | 0.0003 | 0.0502 |
| 0.0196        | 1.58  | 10000 | 0.0269          | 0.9962   | 0.0326 | 0.0005 | 0.0165 |
| 0.0128        | 2.36  | 15000 | 0.0479          | 0.9938   | 0.0585 | 0.0002 | 0.0294 |
| 0.0152        | 3.15  | 20000 | 0.0119          | 0.9983   | 0.0067 | 0.0011 | 0.0039 |
| 0.0074        | 3.94  | 25000 | 0.0161          | 0.9979   | 0.0153 | 0.0006 | 0.0080 |


### Framework versions

- Transformers 4.38.0.dev0
- Pytorch 2.1.2+cu121
- Datasets 2.16.2.dev0
- Tokenizers 0.15.2