RVC WebGUI Fork

Warning

This is a BREAKING fork! I am not going to support the old RVC model/f0 indefinitely!

If you are interested in a better maintained RVC you should check out: https://github.com/fumiama/Retrieval-based-Voice-Conversion-WebUI

This fork aims to improve upon the original RVC project and with several enhancements:

• Better dependency management

• Improved Gradio UI

• Better Docker support

Getting Started

Run on Google Colab

Follow instructions on this notebook.

Run locally

Install uv if you haven't already.

git clone https://github.com/yamada-sexta/rvc-webgui-fork.git
cd rvc-webgui-fork
uv sync --locked
uv run ./tools/download_models.py
uv run ./web_ui.py

Docker Compose

Create a docker-compose.yml similar to this:

services:
  rvc-fork-server:
    image: ghcr.io/yamada-sexta/rvc-webgui-fork:latest
    # Or if you prefer build from source
    # build:
    #   context: https://github.com/yamada-sexta/rvc-webgui-fork.git
    #   dockerfile: Dockerfile
    restart: "unless-stopped"
    shm_size: '16gb'
    runtime: nvidia
    environment:
      - NVIDIA_VISIBLE_DEVICES=all
    ports:
      - "7865:7865"
    volumes:
      - ./data/cache:/cache
      - ./data/assets:/app/assets/
      - ./data/datasets:/app/datasets
      - ./data/weights:/app/assets/weights
      - ./data/logs:/app/logs
      - /app/logs/mute

In the same directory run

docker-compose up -d

Then visit localhost:7865 in your browser.

Contributing

We have a more open policy to contribution compared to the original.

Please refer this guide for more details.