metadata

license: mit

easyGUI

easyGUI is a user-friendly voice conversion framework based on VITS, designed to eliminate timbre leakage by replacing input features with those from the training set. It's efficient even on lower-end GPUs, requiring only about 10 minutes of low-noise speech data for good results. The framework features a simple web interface, supports A card and I card acceleration, and uses the advanced RMVPE algorithm for pitch extraction.

Installation

Prerequisites

Python 3.8 or higher

Installation Steps

Install Pytorch:

pip install torch torchvision torchaudio

For Nvidia Ampere (RTX30xx):

pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu117

Install Dependencies:

For Nvidia GPUs:
```
pip install -r requirements.txt
```
For AMD/Intel GPUs:
```
pip install -r requirements-dml.txt
```
For AMD ROCm (Linux):
```
pip install -r requirements-amd.txt
```
For Intel IPEX (Linux):
```
pip install -r requirements-ipex.txt
```

Install Optional Dependencies (if needed):
```
sh ./run.sh  # For MacOS
```

Additional Setup

Download Assets: Download necessary models and files using the scripts in the tools directory.

Install FFmpeg:

sudo apt install ffmpeg  # Ubuntu/Debian
brew install ffmpeg      # MacOS

Usage

Start the WebUI:

python demo.py

If using Poetry:

poetry run python demo.py

Features

Top1 retrieval to replace input features
Fast training on less powerful GPUs
Model merging to change timbre
Advanced pitch extraction with RMVPE