# Deepfake-audio-detection-V2 by Melody Machine: Open Model Source: https://savrn.com/models/deepfake-audio-detection-v2 Markdown alternate of the page above; the site index is https://savrn.com/llms.txt --- ## Runs On What it takes to serve Deepfake-audio-detection-V2 (95M parameters): the memory its weights need at each precision, and the cheapest way to rent enough data-center GPUs to hold them. | Precision | Weights | Memory needed | Cheapest setup | Per hour | Also fits | | --- | --- | --- | --- | --- | --- | | 16-bit | 0.2 GB | 0.2 GB | 1x [MI300X](https://savrn.com/ai-index/pricing/gpus/mi300x) (192 GB) Vultr | $1.85 | [1x H100](https://savrn.com/ai-index/pricing/gpus/h100) $1.99 · [1x MI325X](https://savrn.com/ai-index/pricing/gpus/mi325x) $2.00 | | 8-bit | 0.1 GB | 0.1 GB | 1x [MI300X](https://savrn.com/ai-index/pricing/gpus/mi300x) (192 GB) Vultr | $1.85 | [1x H100](https://savrn.com/ai-index/pricing/gpus/h100) $1.99 · [1x MI325X](https://savrn.com/ai-index/pricing/gpus/mi325x) $2.00 | | 4-bit | 0.0 GB | 0.1 GB | 1x [MI300X](https://savrn.com/ai-index/pricing/gpus/mi300x) (192 GB) Vultr | $1.85 | [1x H100](https://savrn.com/ai-index/pricing/gpus/h100) $1.99 · [1x MI325X](https://savrn.com/ai-index/pricing/gpus/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](https://savrn.com/ai-index/pricing/gpus), read Oct 24, 2026. [Deepfake-audio-detection-V2 on every accelerator the SAVRN Index prices, at every precision](https://savrn.com/models/deepfake-audio-detection-v2/gpus) ## Model Card By Melody Machine, published under apache-2.0, revision de3cde5a29c4. This model is a fine-tuned version of motheecreator/Deepfake-audio-detection on the audiofolder dataset. It achieves the following results on the evaluation set: The following hyperparameters were used during training: - learningrate: 3e-05 - trainbatchsize: 32 - evalbatchsize: 32 - gradientaccumulationsteps: 4 - totaltrainbatchsize: 128 - lrschedulertype: cosine - lrschedulerwarmupratio: 0.1 - numepochs: 5 - Transformers 4.41.2 - Pytorch 2.1.2 - Datasets 2.19.2 - Tokenizers 0.19.1 Read Melody Machine's full model card This model is a fine-tuned version of [motheecreator/Deepfake-audio-detection](https://huggingface.co/motheecreator/Deepfake-audio-detection) on the audiofolder dataset. It achieves the following results on the evaluation set: - Loss: 0.0141 - Accuracy: 0.9973 ### 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: 32 - eval_batch_size: 32 - seed: 42 - gradient_accumulation_steps: 4 - total_train_batch_size: 128 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: cosine - lr_scheduler_warmup_ratio: 0.1 - num_epochs: 5 #### Training results | Training Loss | Epoch | Step | Validation Loss | Accuracy | | --- | --- | --- | --- | --- | | 0.0503 | 1.0 | 1381 | 0.0514 | 0.9858 | | 0.0327 | 2.0 | 2762 | 0.0174 | 0.9956 | | 0.0064 | 3.0 | 4143 | 0.0221 | 0.9950 | | 0.0003 | 4.0 | 5524 | 0.0174 | 0.9965 | | 0.0115 | 5.0 | 6905 | 0.0141 | 0.9973 | #### Framework versions - Transformers 4.41.2 - Pytorch 2.1.2 - Datasets 2.19.2 - Tokenizers 0.19.1 ## Configuration Architecture Wav2Vec2ForSequenceClassification Layers 12 Hidden size 768 Feed-forward size 3,072 Attention heads 12 Vocabulary size 32 Stored precision float32 Model type wav2vec2 ## Identity and Version Repository MelodyMachine/Deepfake-audio-detection-V2 Publisher Melody Machine Task Audio classification Modality Audio Library transformers Parameters 95M parameters Languages Not stated by the source Revision de3cde5a29c449bb5268814e421b46bf6ebdcd72 First published 2024-06-17 Last updated 2024-06-18 ## Files and Weights 6 files, 378.3 MB in total. The weights are 2 files totalling 378.3 MB in bin, safetensors. Weights2 files · 378.3 MB Configuration2 files · 2.7 KB Documentation1 file · 2.1 KB Repository1 file · 1.5 KB Every file | File | Type | Size | SHA-256 | | --- | --- | --- | --- | | model.safetensors | Weights | 378.3 MB | 997d9ce59e63 | | training_args.bin | Weights | 5.2 KB | bd510614d766 | | config.json | Configuration | 2.5 KB | — | | preprocessor_config.json | Configuration | 215 B | — | | README.md | Documentation | 2.1 KB | — | | .gitattributes | Repository | 1.5 KB | — | ## License and Download License apache-2.0 Access Open weights, no gate Download size 378.3 MB [Download from Melody Machine](https://huggingface.co/MelodyMachine/Deepfake-audio-detection-V2) Released by Melody Machine through its official repository on Hugging Face. [Read the license](https://www.apache.org/licenses/LICENSE-2.0). ## Built From - Derived from [mo-thecreator/Deepfake-audio-detection](https://savrn.com/models/deepfake-audio-detection) - Trained on (disclosed) audiofolder ## Evaluations Each result is shown as reported, with the conditions its reporter stated. None is a SAVRN measurement. A comparison lines two results up only when their configuration, unit and setup are all stated and identical. | Benchmark | Conditions | Result | Reported by | Revision | Date | | --- | --- | --- | --- | --- | --- | | audiofolder | Configuration defaultTask Audio ClassificationMetric AccuracyComparison conditions not established | 0.997284 | [MelodyMachine](https://huggingface.co/MelodyMachine/Deepfake-audio-detection-V2) Publisher reported | Evaluated revision not stated | — | ## Memory Requirements | Precision | Weights in memory | | --- | --- | | As published | 378.3 MB | | 16-bit | 0.2 GB | | 8-bit | 0.1 GB | | 4-bit | 0.0 GB | Weights only, from the published parameter count; the key-value cache and runtime add to this. ## Questions About Deepfake-audio-detection-V2 ### How much GPU memory does Deepfake-audio-detection-V2 need? About 0.2 GB at 16-bit and 0.1 GB at 4-bit: the weights (95M parameters) plus a working margin. A long context needs more. ### What is the cheapest GPU to run Deepfake-audio-detection-V2 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 Deepfake-audio-detection-V2 commercially? Yes. Deepfake-audio-detection-V2 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 Model · Audio classification ### [Common-Voice-Gender-Detection](https://savrn.com/models/common-voice-gender-detection) [Prithiv Sakthi](https://savrn.com/model-publishers/prithivmlmods) Wav2Vec2: Self-Supervised Learning for Speech Recognition: https://arxiv.org/pdf/2006.11477 male female Common-Voice-Gender-Detection is designed for: Speech Analytics – Assist in analyzing speaker demographics in call centers or customer service recordings. Conversational AI Personalization – Adjust tone or dialogue based on gender detection for more personalized voice assistants. Voice Dataset Curation – Automatically tag or filter voice datasets by speaker gender for better dataset management. Research Applications – Enable linguistic and acoustic research involving gender-specific speech patterns. Multimedia Content Tagging – Automate metadata generation for gender identification in… Open weights apache-2.0 95M parameters transformers [View model](https://savrn.com/models/common-voice-gender-detection) Model · Audio classification ### [Deepfake-audio-detection](https://savrn.com/models/deepfake-audio-detection) [Mohammed Abdeldayem](https://savrn.com/model-publishers/mo-thecreator) This model is a fine-tuned version of mo-thecreator/wav2vec2-base-finetuned on the None dataset. It achieves the following results on the evaluation set: The following hyperparameters were used during training: - learningrate: 3e-05 - trainbatchsize: 8 - evalbatchsize: 8 - gradientaccumulationsteps: 4 - totaltrainbatchsize: 32 - lrschedulertype: linear - lrschedulerwarmupratio: 0.1 - numepochs: 5 - Transformers 4.39.3 - Pytorch 2.1.2 - Datasets 2.18.0 - Tokenizers 0.15.2 - mo-thecreator Open weights apache-2.0 95M parameters transformers [View model](https://savrn.com/models/deepfake-audio-detection) Model · Audio classification ### [wav2vec2-emotion-recognition](https://savrn.com/models/wav2vec2-emotion-recognition) [Deepan Gautam](https://savrn.com/model-publishers/dpngtm) This model is a fine-tuned version of facebook/wav2vec2-base-960h for Speech Emotion Recognition (SER). It has been trained using a Frozen Feature Extractor strategy to preserve the model's acoustic understanding while adapting to emotion detection. This approach ensures stable performance and prevents "Catastrophic Forgetting," achieving nearly 80% accuracy on the validation set. Update: The "Calm" and "Neutral" classes have been merged to improve classification consistency, resulting in 7 distinct emotion classes. The model was trained on a combined dataset of ~12,000 audio files from: The model classifies audio into one of the following emotions: 1. Angry 2. Disgust 3. Fear 4. Happy 5.… Open weights mit 95M parameters transformers [View model](https://savrn.com/models/wav2vec2-emotion-recognition) Model · Audio classification ### [audio_cls](https://savrn.com/models/audio-cls-5) [Kim Yeoungdoo](https://savrn.com/model-publishers/kud1223) This model is a fine-tuned version of Kkonjeong/wav2vec2-base-korean on the None dataset. It achieves the following results on the evaluation set: The following hyperparameters were used during training: - learningrate: 0.0001 - trainbatchsize: 4 - evalbatchsize: 4 - lrschedulertype: linear - numepochs: 10 - mixedprecisiontraining: Native AMP - Transformers 5.18.0 - Pytorch 2.11.0+cpu - Datasets 4.8.5 - Tokenizers 0.23.2 Open weights 95M parameters transformers [View model](https://savrn.com/models/audio-cls-5) Model · Audio classification ### [keyword](https://savrn.com/models/keyword) [PARKHYEIN](https://savrn.com/model-publishers/binfool) This model is a fine-tuned version of Kkonjeong/wav2vec2-base-korean on the None dataset. It achieves the following results on the evaluation set: The following hyperparameters were used during training: - learningrate: 0.0001 - trainbatchsize: 8 - evalbatchsize: 8 - 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 95M parameters transformers [View model](https://savrn.com/models/keyword) Model · Audio classification ### [keyword](https://savrn.com/models/keyword-2) [Dohyeon](https://savrn.com/model-publishers/dohyeonhong) This model is a fine-tuned version of Kkonjeong/wav2vec2-base-korean on the None dataset. It achieves the following results on the evaluation set: The following hyperparameters were used during training: - learningrate: 0.0001 - trainbatchsize: 8 - evalbatchsize: 8 - 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 95M parameters transformers [View model](https://savrn.com/models/keyword-2) ## Melody Machine [All models and datasets](https://savrn.com/model-publishers/melodymachine) ## Versions - [de3cde5a29c4](https://savrn.com/models/deepfake-audio-detection-v2/versions/de3cde5a29c4) · current 2026-09-18 ## Explore More - [All audio classification models](https://savrn.com/models/tasks/audio-classification) - [All models under apache-2.0](https://savrn.com/models/licenses/apache-2-0) - [Model comparisons](https://savrn.com/models/comparisons) - [The model directory](https://savrn.com/models) - [Open model prices by host](https://savrn.com/ai-index/pricing/open-models) ## Source - Repository metadata, read 2026-10-21. - [Hugging Face record](https://huggingface.co/MelodyMachine/Deepfake-audio-detection-V2) - [How the hub is built](https://savrn.com/model-hub/methodology)