Advanced RVC Inference

A modular Retrieval-based Voice Conversion framework with Gradio UI, training capabilities, and audio processing tools

Features

• Voice Conversion: High-quality voice conversion with multiple pitch extraction methods
• Model Training: Complete training pipeline for creating custom RVC models
• Real-time Processing: Low-latency real-time voice conversion support
• Web UI: Intuitive Gradio-based web interface
• CLI Support: Command-line interface for scripting and automation
• API Access: Python API for programmatic access
• Audio Separation: Built-in tools for vocal/instrument separation
• Text-to-Speech: Integration with edge-tts for TTS-based voice conversion

Installation

pip install git+https://github.com/ArkanDash/Advanced-RVC-Inference.git

With GPU Support

For CUDA-enabled GPUs:

pip install git+https://github.com/ArkanDash/Advanced-RVC-Inference.git#egg=advanced-rvc-inference[gpu]

From Source

git clone https://github.com/ArkanDash/Advanced-RVC-Inference.git
cd Advanced-RVC-Inference
pip install -e .

Development Installation

pip install git+https://github.com/ArkanDash/Advanced-RVC-Inference.git#egg=advanced-rvc-inference[dev]

Quick Start

Web Interface

Launch the Gradio web UI:

rvc-gui
# or
python -m advanced_rvc_inference.gui

The web interface will be available at http://localhost:7860

Command Line Interface

Run voice conversion from the command line:

rvc-cli infer --model path/to/model.pth --input audio.wav --output converted.wav --pitch 0

View help:

rvc-cli --help
rvc-cli infer --help

Python API

from advanced_rvc_inference import RVCInference

# Initialize the inference engine
rvc = RVCInference(device="cuda:0")

# Load a model
rvc.load_model("path/to/model.pth")

# Run inference
audio = rvc.infer("input.wav", pitch_change=0, output_path="output.wav")

# Or use batch processing
audio_files = rvc.infer_batch(
    input_dir="input_folder",
    output_dir="output_folder",
    pitch_change=2,
    format="wav"
)

# Cleanup
rvc.unload_model()

Command Reference

CLI Commands

Command	Description
rvc-cli infer	Run voice conversion inference on a single audio file
rvc-cli infer-batch	Run batch voice conversion on multiple files
rvc-cli train	Train RVC models (use web UI for full features)
rvc-cli dataset	Create and manage training datasets
rvc-cli preprocess	Preprocess training data
rvc-cli extract	Extract features for training
rvc-cli index	Create index for feature retrieval
rvc-cli separate	Separate music into vocals and instruments
rvc-cli reference	Create reference audio for training
rvc-cli tts	Text-to-speech voice conversion
rvc-cli serve	Launch the web interface
rvc-cli info	Show system information

Quick Inference

Run voice conversion on a single audio file:

rvc-cli infer --model path/to/model.pth --input audio.wav --output converted.wav

With pitch shift (one octave up):

rvc-cli infer --model vocals.pth --input audio.wav --pitch 12 --output output.wav

Batch Processing

Process multiple audio files at once:

rvc-cli infer-batch --model model.pth --input_dir ./songs --output_dir ./converted

Music Separation

Separate vocals from instrumental tracks:

rvc-cli separate --input song.mp3 --output_dir ./separated

Web Interface

Launch the Gradio web UI:

rvc-cli serve --port 7860

View help for any command:

rvc-cli --help
rvc-cli infer --help
rvc-cli separate --help


## Configuration

### Environment Variables

| Variable | Description | Default |
|----------|-------------|---------|
| `ARVC_ASSETS_PATH` | Path to asset directory | Package assets folder |
| `ARVC_CONFIGS_PATH` | Path to configs directory | Package configs folder |
| `ARVC_WEIGHTS_PATH` | Path to model weights | assets/weights |
| `ARVC_LOGS_PATH` | Path to logs directory | assets/logs |

### Configuration File

Configuration is managed through `advanced_rvc_inference/configs/config.json`:

```json
{
    "device": "cuda:0",
    "fp16": true,
    "app_port": 7860,
    "language": "vi-VN",
    "theme": "NoCrypt/miku",
    "uvr_path": "advanced_rvc_inference/assets/audios"
}

Documentation

• API Reference
• Usage Guide
• Contributing

Troubleshooting

GPU Not Detected

Ensure you have CUDA installed and PyTorch with CUDA support:

pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu118

Memory Issues

Reduce batch size or use CPU mode:

rvc = RVCInference(device="cpu")

Contributing

Contributions are welcome! Please read our Contributing Guide for details.

License

This project is licensed under the MIT License - see the LICENSE file for details.

Terms of Use

The use of the converted voice for the following purposes is prohibited:

• Criticizing or attacking individuals
• Advocating for or opposing specific political positions, religions, or ideologies
• Publicly displaying strongly stimulating expressions without proper zoning
• Selling of voice models and generated voice clips
• Impersonation of the original owner of the voice with malicious intentions
• Fraudulent purposes that lead to identity theft or fraudulent phone calls

Credits

Repository	Owner
Vietnamese-RVC	PhamHuynhAnh16
Applio	IAHispano

Support

For issues and feature requests, please use the GitHub Issues page.

---

Made with by ArkanDash