Skip to page content
SAVRN
Search Plan a project

Open-weight model · Audio classification

wav2vec2-deepfake-voice-detector

by Gary Stafford garystafford/wav2vec2-deepfake-voice-detector

wav2vec2-deepfake-voice-detector is an open-weight model for audio classification from Gary Stafford, released under Apache License 2.0. It has 316M parameters. At 16-bit it needs about 0.8 GB of GPU memory, which fits on 1x MI300X from $1.85 an hour, at the lowest prices in the SAVRN Index. It draws 12.2k downloads a month.

Fine-tuned Wav2Vec2 model for detecting AI-generated speech. Determines if audio was spoken by a human or created by AI text-to-speech/voice cloning software. Fine-tuned Wav2Vec2 transformer for binary audio classification (real vs AI-generated speech).

Parameters316M
Context—
Weights1.3 GB
Licenseapache-2.0
AccessOpen weights
Monthly Downloads12.2k

Runs On

What it takes to serve wav2vec2-deepfake-voice-detector (316M parameters): the memory its weights need at each precision, and the cheapest way to rent enough data-center GPUs to hold them.

PrecisionWeightsMemory neededCheapest setupPer hourAlso fits
16-bit 0.6 GB 0.8 GB 1x MI300X (192 GB)
Vultr		 $1.85 1x H100 $1.99 · 1x MI325X $2.00
8-bit 0.3 GB 0.4 GB 1x MI300X (192 GB)
Vultr		 $1.85 1x H100 $1.99 · 1x MI325X $2.00
4-bit 0.2 GB 0.2 GB 1x MI300X (192 GB)
Vultr		 $1.85 1x H100 $1.99 · 1x MI325X $2.00

Memory is the weights at that precision plus 20% for the runtime and a short context; a long context needs more. Prices are the lowest on-demand hourly rates in the SAVRN Index, read Oct 16, 2026.

wav2vec2-deepfake-voice-detector on every accelerator the SAVRN Index prices, at every precision

Model Card

By Gary Stafford, published under apache-2.0, revision c66306024a7e.

Deepfake Audio Detection Model

Fine-tuned Wav2Vec2 model for detecting AI-generated speech. Determines if audio was spoken by a human or created by AI text-to-speech/voice cloning software.

Model Details

Model Description

Fine-tuned Wav2Vec2 transformer for binary audio classification (real vs AI-generated speech). Trained to distinguish authentic human speech from synthetic audio generated by AI text-to-speech and voice cloning services including:

  • ElevenLabs
  • Amazon Polly
  • Hexgrad Kokoro
  • Hume AI
  • Speechify
  • Luvvoice

Developed by: Gary A. Stafford

Note: This model uses transfer learning from a base model already trained for deepfake detection. Fast convergence is expected due to task similarity and TTS engine overlap with the base model's training data.

How to Use

Installation

Install the required dependencies:

pip install transformers torch librosa

Optional: For GPU acceleration (recommended):

# For CUDA 11.8
pip install torch --index-url https://download.pytorch.org/whl/cu118

# For CUDA 12.1
pip install torch --index-url https://download.pytorch.org/whl/cu121

Quick Start

Read the full model card (1,146 words)

Configuration

Architecture
Wav2Vec2ForSequenceClassification
Layers
24
Hidden size
1,024
Feed-forward size
4,096
Attention heads
16
Vocabulary size
32
Model type
wav2vec2

Identity and Version

Repository
garystafford/wav2vec2-deepfake-voice-detector
Publisher
Gary Stafford
Task
Audio classification
Modality
Audio
Library
transformers
Parameters
316M parameters
Languages
en
Revision
c66306024a7ede0be291e9c4558b37634782dc4e
First published
2025-12-24
Last updated
2025-12-28

Files and Weights

5 files, 1.3 GB in total. The weights are 1 file totalling 1.3 GB in safetensors.

Weights1 file · 1.3 GB
Configuration2 files · 2.5 KB
Documentation1 file · 11.2 KB
Repository1 file · 1.5 KB
Every file
FileTypeSizeSHA-256
model.safetensorsWeights1.3 GB 905e330265c4
config.jsonConfiguration2.3 KB —
preprocessor_config.jsonConfiguration221 B —
README.mdDocumentation11.2 KB —
.gitattributesRepository1.5 KB —

License and Download

License
apache-2.0
Access
Open weights, no gate
Download size
1.3 GB
Download from Gary Stafford

Released by Gary Stafford through its official repository on Hugging Face. Read the license.

Built From

Memory Requirements

PrecisionWeights in memory
As published1.3 GB
16-bit0.6 GB
8-bit0.3 GB
4-bit0.2 GB

Weights only, from the published parameter count; the key-value cache and runtime add to this.

Questions About wav2vec2-deepfake-voice-detector

How much GPU memory does wav2vec2-deepfake-voice-detector need?

About 0.8 GB at 16-bit and 0.2 GB at 4-bit: the weights (316M parameters) plus a working margin. A long context needs more.

What is the cheapest GPU to run wav2vec2-deepfake-voice-detector on?

At 16-bit, 1x MI300X from $1.85 an hour; at 4-bit, 1x MI300X from $1.85 an hour, at the lowest on-demand prices the SAVRN Index lists.

Can I use wav2vec2-deepfake-voice-detector commercially?

Yes. wav2vec2-deepfake-voice-detector is released under Apache License 2.0. The Apache License 2.0 is a permissive open-source license. It permits commercial use, modification and redistribution. It requires keeping the license and copyright notices and any NOTICE file, stating significant changes, and it includes an express patent grant from contributors.

Similar Models

This model is a fine-tuned version of facebook/wav2vec2-xls-r-300m on Librispeech-clean-100 for gender recognition. It achieves the following results on the evaluation set: The Librispeech-clean-100 dataset was used to train the model, with 70% of the data used for training, 10% for validation, and 20% for testing. The following hyperparameters were used during training: - learningrate: 3e-05 - trainbatchsize: 4 - evalbatchsize: 4 - gradientaccumulationsteps: 4 - totaltrainbatchsize: 16 - lrschedulertype: linear - lrschedulerwarmupratio: 0.1 - numepochs: 1 - mixedprecisiontraining: Native AMP - Transformers 4.28.0 - Pytorch 2.0.0+cu118 - Tokenizers 0.13.3

Open weights apache-2.0 316M parameters transformers
View model

Model · Audio classification

wav2vec-vm-finetune

Jake Downie

This model is a fine-tuned version of facebook/wav2vec2-xls-r-300m for voicemail detection. It is trained on a dataset of call recordings to distinguish between voicemail greetings and live human responses. This model builds on wav2vec2-xls-r-300m, a self-supervised speech model trained on large-scale multilingual data. We fine-tuned it on the first two seconds of a call. - Automated voicemail detection in AI-powered call assistants. - Filtering voicemail responses in customer service and sales call automation. - Only trianed on the English language. - Assumes the voicemail track is isolated and contains no audio from the caller. - Designed for the first two seconds of audio when calling a…

Open weights apache-2.0 316M parameters transformers
View model

The pre-trained model is this one - facebook/hubert-large-ls960-ft The DUSHA dataset used can be found here Fine-tuned in Google Colab using Pro account with A100 GPU Freezed all layers exept projector, classifier and all 24 HubertEncoderLayerStableLayerNorm layers Used half of the train dataset - 2 epochs - train batch size = 8 - eval batch size = 8 - gradient accumulation steps = 4 - learning rate = 5e-5 without warm up and decay Achieved - accuracy = 0.86 - balanced = 0.76 - macro f1 score = 0.81 on test set, improving accucary and f1 score compared to dataset baseline

Open weights apache-2.0 316M parameters transformers
View model

Model · Audio classification

audio_cls_1016

Ryusunghwan

This model is a fine-tuned version of facebook/wav2vec2-xls-r-300m on an unknown dataset. It achieves the following results on the evaluation set: The following hyperparameters were used during training: - learningrate: 0.0001 - trainbatchsize: 8 - evalbatchsize: 8 - gradientaccumulationsteps: 4 - totaltrainbatchsize: 32 - lrschedulertype: linear - numepochs: 10 - mixedprecisiontraining: Native AMP - Transformers 5.18.0 - Pytorch 2.11.0+cu130 - Datasets 4.8.5 - Tokenizers 0.23.2

Open weights apache-2.0 316M parameters transformers
View model

Model · Audio classification

audio_cls_1016

Ju ung

This model is a fine-tuned version of facebook/wav2vec2-xls-r-300m on an unknown dataset. It achieves the following results on the evaluation set: The following hyperparameters were used during training: - learningrate: 0.0001 - trainbatchsize: 8 - evalbatchsize: 8 - gradientaccumulationsteps: 4 - totaltrainbatchsize: 32 - lrschedulertype: linear - numepochs: 10 - mixedprecisiontraining: Native AMP - Transformers 5.18.0 - Pytorch 2.11.0+cu130 - Datasets 4.8.5 - Tokenizers 0.23.2

Open weights apache-2.0 316M parameters transformers
View model