Skip to content
Other

DiffErase

стирание аудиоводяных знаков диффузией

1 min read

DiffErase — набор pretrained-чекпоинтов для одноимённого проекта по удалению аудиоводяных знаков методом black-box с использованием diffusion priors («Audio Pirates: Black-box Audio Watermark Removal via Diffusion Priors»). Это инструмент на базе diffusers с двумя ветками — mel и latent — для доменов speech, music и env. Размер в параметрах в карточке не указан.

Ключевые особенности

  • Три домена: speech, music и env — отдельные чекпоинты для речи, музыки и окружающих звуков.
  • DiffErase-mel: unconditional UNet2DModel + DDPMScheduler на 80-band mel-спектрограммах 22.05 kHz; аудио восстанавливается через BigVGAN.
  • DiffErase-latent: unconditional AudioLDM-style latent diffusion, 16 kHz, 64-band mel; каждый .ckpt содержит UNet, EMA-веса и frozen VAE.
  • Для latent-версии в карточке указаны pretrained VAE и HiFi-GAN вокодер из AudioLDM.
  • Данные обучения: LibriSpeech, FMA и Clotho.

Пример

git clone https://github.com/DiffErase/DiffErase.git
cd DiffErase
hf download DiffErase/DiffErase --local-dir .
python remove_differase-mel.py --domain speech \
  --watermarked_dir path/to/watermarked_audio --output_dir path/to/erased_audio --noise_level 0.1
ПодробнееТри домена и датасеты: speech — LibriSpeech, music — FMA, env — Clotho.
DiffErase-mel работает с 80-band mel-спектрограммами 22.05 kHz и восстанавливает аудио через BigVGAN.
DiffErase-latent — AudioLDM-style latent diffusion на 16 kHz и 64-band mel; .ckpt включает UNet, EMA-веса и frozen VAE.
Бенчмарков и требований к VRAM в карточке нет.
Discuss Telegram 0

Will it run?

Working out the memory needs

The bot is still collecting this model's weight sizes; the estimate will appear here in a few minutes. Meanwhile you can rent a GPU with headroom.

Massed ComputeOpen Massed Compute

Run it on Massed Compute

The script downloads the model to the rented machine. How to run it is in the model card.

  1. 1. Rent a machine with a fitting GPU

    Open Massed Compute
  2. 2. Paste the script into the machine's terminal

    Connect over SSH (the Massed Compute panel shows the command) or open a terminal on the machine's desktop.

    HF_TOKEN / CIVITAI_TOKENThe script starts with token lines: paste yours between the quotes. The Hugging Face token is for gated models, the Civitai one for Civitai models.

    # DiffErase/DiffErase
    mkdir -p ~/models
    curl -LsSf https://astral.sh/uv/install.sh | sh
    source "$HOME/.local/bin/env"
    uv tool install "huggingface_hub[cli]"
    hf download DiffErase/DiffErase --local-dir ~/models/DiffErase
    # The model is in ~/models. How to run it is in the model card (link above).
    Add more LoRAs to DiffErase
  3. 3. Use it

    The model is in ~/models. How to run it is in the model card (link above).

Stop the machine in the panel when you are done: it bills by the minute while it runs.

How to speed it up

Picked for this model type. Choose your hardware above to see what fits you exactly.

  • FlashAttention 44.0.0b40from November 18, 2026new release

    fastest exact attention on B200

    Kernels for Blackwell (beta). Exact attention without quantisation — fine for everything, LLMs included.

    How to enable: pip install flash-attn-4; vLLM and SGLang pick it up themselves.

    Earlier generations (2)
    • FlashAttention 3 — The Hopper (H100/H200) version with FP8.
    • FlashAttention 2 — The standard for RTX 30/40/50 and A100: exact attention without the memory cost of long contexts.
  • torch.compile2.15.0from October 28, 2026new release

    10–30% for free

    Compiles the model for your GPU. Slower first run, faster afterwards.

    How to enable: model = torch.compile(model); in ComfyUI the TorchCompileModel node.

Model family

New posts about versions of DiffErase (quants, finetunes, LoRAs) will show up under the bell.

No known relatives of this model yet. All derivatives on Hugging Face

About the model

DiffErase/DiffErase
Author
DiffErase
Category
Other
Released
Posted

On NeyroLearn

Hugging Face

00

Ask AI about this model

Answered by Gemini 4 · the chat is kept only in this tab

Comments