2023-04-28 00:16:38 +08:00
< div align = "center" >
< h1 > Retrieval-based-Voice-Conversion-WebUI< / h1 >
2023-10-06 19:11:58 +08:00
一个基于VITS的简单易用的变声框架< br > < br >
2023-04-28 00:16:38 +08:00
2023-08-09 19:32:49 +08:00
[![madewithlove](https://img.shields.io/badge/made_with-%E2%9D%A4-red?style=for-the-badge& labelColor=orange
)](https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI)
2023-04-28 00:16:38 +08:00
< img src = "https://counter.seku.su/cmoe?name=rvc&theme=r34" / > < br >
2023-09-18 22:50:30 +08:00
2023-06-22 14:37:09 +08:00
[![Open In Colab ](https://img.shields.io/badge/Colab-F9AB00?style=for-the-badge&logo=googlecolab&color=525252 )](https://colab.research.google.com/github/RVC-Project/Retrieval-based-Voice-Conversion-WebUI/blob/main/Retrieval_based_Voice_Conversion_WebUI.ipynb)
2023-09-18 22:50:30 +08:00
[![Licence ](https://img.shields.io/badge/LICENSE-MIT-green.svg?style=for-the-badge )](https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI/blob/main/LICENSE)
2023-04-28 00:16:38 +08:00
[![Huggingface ](https://img.shields.io/badge/🤗%20-Spaces-yellow.svg?style=for-the-badge )](https://huggingface.co/lj1995/VoiceConversionWebUI/tree/main/)
[![Discord ](https://img.shields.io/badge/RVC%20Developers-Discord-7289DA?style=for-the-badge&logo=discord&logoColor=white )](https://discord.gg/HcsmBBGyVk)
2023-10-29 12:36:00 +08:00
[**更新日志** ](https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI/blob/main/docs/Changelog_CN.md ) | [**常见问题解答** ](https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI/wiki/%E5%B8%B8%E8%A7%81%E9%97%AE%E9%A2%98%E8%A7%A3%E7%AD%94 ) | [**AutoDL·5毛钱训练AI歌手** ](https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI/wiki/Autodl%E8%AE%AD%E7%BB%83RVC%C2%B7AI%E6%AD%8C%E6%89%8B%E6%95%99%E7%A8%8B ) | [**对照实验记录** ](https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI/wiki/Autodl%E8%AE%AD%E7%BB%83RVC%C2%B7AI%E6%AD%8C%E6%89%8B%E6%95%99%E7%A8%8B](https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI/wiki/%E5%AF%B9%E7%85%A7%E5%AE%9E%E9%AA%8C%C2%B7%E5%AE%9E%E9%AA%8C%E8%AE%B0%E5%BD%95 )) | [**在线演示** ](https://modelscope.cn/studios/FlowerCry/RVCv2demo )
2023-09-18 22:50:30 +08:00
2023-04-28 00:16:38 +08:00
< / div >
------
2023-09-18 22:50:30 +08:00
[**English** ](./docs/en/README.en.md ) | [**中文简体** ](./README.md ) | [**日本語** ](./docs/jp/README.ja.md ) | [**한국어** ](./docs/kr/README.ko.md ) ([**韓國語**](./docs/kr/README.ko.han.md)) | [**Français** ](./docs/fr/README.fr.md )| [**Türkçe** ](./docs/tr/README.tr.md )
2023-04-28 00:16:38 +08:00
2023-09-18 22:50:30 +08:00
点此查看我们的[演示视频](https://www.bilibili.com/video/BV1pm4y1z7Gm/) !
2023-06-11 21:56:06 +08:00
2023-10-06 19:11:58 +08:00
训练推理界面: go-web.bat
![image ](https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI/assets/129054828/092e5c12-0d49-4168-a590-0b0ef6a4f630 )
实时变声界面: go-realtime-gui.bat
![image ](https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI/assets/129054828/143246a9-8b42-4dd1-a197-430ede4d15d7 )
2023-04-28 00:16:38 +08:00
2023-09-18 22:50:30 +08:00
> 底模使用接近50小时的开源高质量VCTK训练集训练, 无版权方面的顾虑, 请大家放心使用
2023-04-28 00:16:38 +08:00
2023-09-18 22:50:30 +08:00
> 请期待RVCv3的底模, 参数更大, 数据更大, 效果更好, 基本持平的推理速度, 需要训练数据量更少。
2023-04-28 00:16:38 +08:00
2023-09-18 22:50:30 +08:00
## 简介
本仓库具有以下特点
+ 使用top1检索替换输入源特征为训练集特征来杜绝音色泄漏
+ 即便在相对较差的显卡上也能快速训练
+ 使用少量数据进行训练也能得到较好结果(推荐至少收集10分钟低底噪语音数据)
+ 可以通过模型融合来改变音色(借助ckpt处理选项卡中的ckpt-merge)
+ 简单易用的网页界面
+ 可调用UVR5模型来快速分离人声和伴奏
+ 使用最先进的[人声音高提取算法InterSpeech2023-RMVPE](#参考项目)根绝哑音问题。效果最好( 显著地) 但比crepe_full更快、资源占用更小
+ A卡I卡加速支持
2023-04-28 00:16:38 +08:00
2023-09-18 22:50:30 +08:00
## 环境配置
以下指令需在 Python 版本大于3.8的环境中执行。
2023-04-28 00:16:38 +08:00
2023-09-18 22:50:30 +08:00
(Windows/Linux)
首先通过 pip 安装主要依赖:
2023-04-28 00:16:38 +08:00
```bash
2023-09-18 22:50:30 +08:00
# 安装Pytorch及其核心依赖, 若已安装则跳过
# 参考自: https://pytorch.org/get-started/locally/
2023-04-28 00:16:38 +08:00
pip install torch torchvision torchaudio
2023-09-18 22:50:30 +08:00
#如果是win系统+Nvidia Ampere架构(RTX30xx),根据 #21 的经验, 需要指定pytorch对应的cuda版本
2023-04-28 00:16:38 +08:00
#pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu117
2023-07-29 22:44:36 +08:00
```
2023-04-28 00:16:38 +08:00
2023-09-18 22:50:30 +08:00
可以使用 poetry 来安装依赖:
2023-07-29 22:44:36 +08:00
```bash
2023-09-18 22:50:30 +08:00
# 安装 Poetry 依赖管理工具, 若已安装则跳过
# 参考自: https://python-poetry.org/docs/#installation
2023-04-28 00:16:38 +08:00
curl -sSL https://install.python-poetry.org | python3 -
2023-09-18 22:50:30 +08:00
# 通过poetry安装依赖
2023-04-28 00:16:38 +08:00
poetry install
```
2023-09-18 22:50:30 +08:00
你也可以通过 pip 来安装依赖:
2023-05-10 23:27:33 +08:00
```bash
2023-09-18 22:50:30 +08:00
N卡:
2023-10-06 17:14:33 +08:00
pip install -r requirements.txt
2023-08-14 00:01:47 +08:00
2023-09-18 22:50:30 +08:00
A卡/I卡:
2023-10-06 17:14:33 +08:00
pip install -r requirements-dml.txt
A卡Rocm( Linux) :
pip install -r requirements-amd.txt
2023-08-14 00:01:47 +08:00
2023-10-06 17:14:33 +08:00
I卡IPEX( Linux) :
pip install -r requirements-ipex.txt
2023-05-10 23:27:33 +08:00
```
2023-04-28 00:16:38 +08:00
2023-07-29 22:44:36 +08:00
------
2023-09-18 22:50:30 +08:00
Mac 用户可以通过 `run.sh` 来安装依赖:
2023-07-29 22:44:36 +08:00
```bash
sh ./run.sh
```
2023-09-18 22:50:30 +08:00
## 其他预模型准备
RVC需要其他一些预模型来推理和训练。
2023-04-28 00:16:38 +08:00
2023-09-18 22:50:30 +08:00
你可以从我们的[Hugging Face space](https://huggingface.co/lj1995/VoiceConversionWebUI/tree/main/)下载到这些模型。
2023-04-28 00:16:38 +08:00
2023-09-18 22:50:30 +08:00
以下是一份清单, 包括了所有RVC所需的预模型和其他文件的名称:
2023-04-28 00:16:38 +08:00
```bash
2023-08-28 15:00:51 +08:00
./assets/hubert/hubert_base.pt
2023-04-28 00:16:38 +08:00
2023-08-28 15:00:51 +08:00
./assets/pretrained
2023-04-28 00:16:38 +08:00
2023-08-28 15:00:51 +08:00
./assets/uvr5_weights
2023-04-28 00:16:38 +08:00
2023-09-18 22:50:30 +08:00
想测试v2版本模型的话, 需要额外下载
2023-05-14 15:19:35 +08:00
2023-08-28 15:00:51 +08:00
./assets/pretrained_v2
2023-05-14 15:19:35 +08:00
2023-09-18 22:50:30 +08:00
如果你正在使用Windows, 则你可能需要这个文件, 若ffmpeg和ffprobe已安装则跳过; ubuntu/debian 用户可以通过apt install ffmpeg来安装这2个库, Mac 用户则可以通过brew install ffmpeg来安装 (需要预先安装brew)
2023-07-11 12:02:30 +08:00
2023-09-18 22:50:30 +08:00
./ffmpeg
2023-04-28 00:16:38 +08:00
2023-07-11 12:02:30 +08:00
https://huggingface.co/lj1995/VoiceConversionWebUI/blob/main/ffmpeg.exe
2023-09-18 22:50:30 +08:00
./ffprobe
2023-07-11 12:02:30 +08:00
https://huggingface.co/lj1995/VoiceConversionWebUI/blob/main/ffprobe.exe
2023-09-18 22:50:30 +08:00
如果你想使用最新的RMVPE人声音高提取算法, 则你需要下载音高提取模型参数并放置于RVC根目录
2023-07-11 12:02:30 +08:00
https://huggingface.co/lj1995/VoiceConversionWebUI/blob/main/rmvpe.pt
2023-09-18 22:50:30 +08:00
A卡I卡用户需要的dml环境要请下载
2023-08-14 00:01:47 +08:00
https://huggingface.co/lj1995/VoiceConversionWebUI/blob/main/rmvpe.onnx
2023-04-28 00:16:38 +08:00
```
2023-09-18 22:50:30 +08:00
之后使用以下指令来启动WebUI:
2023-04-28 00:16:38 +08:00
```bash
python infer-web.py
```
2023-09-18 22:50:30 +08:00
如果你正在使用Windows 或 macOS, 你可以直接下载并解压`RVC-beta.7z`,前者可以运行`go-web.bat`以启动WebUI, 后者则运行命令`sh ./run.sh`以启动WebUI。
2023-10-06 17:14:33 +08:00
对于需要使用IPEX技术的I卡用户, 请先在终端执行`source /opt/intel/oneapi/setvars.sh`( 仅Linux) 。
2023-09-18 22:50:30 +08:00
仓库内还有一份`小白简易教程.doc`以供参考。
2023-04-28 00:16:38 +08:00
2023-10-06 17:14:33 +08:00
## AMD显卡Rocm相关( 仅Linux)
如果你想基于AMD的Rocm技术在Linux系统上运行RVC, 请先在[这里](https://rocm.docs.amd.com/en/latest/deploy/linux/os-native/install.html)安装所需的驱动。
若你使用的是Arch Linux, 可以使用pacman来安装所需驱动:
````
pacman -S rocm-hip-sdk rocm-opencl-sdk
````
对于某些型号的显卡, 你可能需要额外配置如下的环境变量( 如: RX6700XT) :
````
export ROCM_PATH=/opt/rocm
export HSA_OVERRIDE_GFX_VERSION=10.3.0
````
同时确保你的当前用户处于`render`与`video`用户组内:
````
sudo usermod -aG render $USERNAME
sudo usermod -aG video $USERNAME
````
之后运行WebUI:
```bash
python infer-web.py
```
2023-09-18 22:50:30 +08:00
## 参考项目
2023-04-28 00:16:38 +08:00
+ [ContentVec ](https://github.com/auspicious3000/contentvec/ )
+ [VITS ](https://github.com/jaywalnut310/vits )
+ [HIFIGAN ](https://github.com/jik876/hifi-gan )
+ [Gradio ](https://github.com/gradio-app/gradio )
+ [FFmpeg ](https://github.com/FFmpeg/FFmpeg )
+ [Ultimate Vocal Remover ](https://github.com/Anjok07/ultimatevocalremovergui )
+ [audio-slicer ](https://github.com/openvpi/audio-slicer )
2023-07-11 12:02:30 +08:00
+ [Vocal pitch extraction:RMVPE ](https://github.com/Dream-High/RMVPE )
+ The pretrained model is trained and tested by [yxlllc ](https://github.com/yxlllc/RMVPE ) and [RVC-Boss ](https://github.com/RVC-Boss ).
2023-09-18 22:50:30 +08:00
## 感谢所有贡献者作出的努力
2023-06-22 14:37:09 +08:00
< a href = "https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI/graphs/contributors" target = "_blank" >
< img src = "https://contrib.rocks/image?repo=RVC-Project/Retrieval-based-Voice-Conversion-WebUI" / >
2023-04-28 00:16:38 +08:00
< / a >