- 20241012update:1.修复聊天界面滑杆问题 2.增加休眠与唤醒功能 3.修补部分bug
- This project is a framework for desktop AI assistant based on sensevoice: https://github.com/FunAudioLLM/SenseVoice , BigModel platform: https://bigmodel.cn/ , Volcano Engine: https://www.volcengine.com/
- please check this link for more info and api docs
- UI demo:
Tutorial Video: https://www.bilibili.com/video/BV1eY2VYDEH6/
- 1.下载miniconda/anaconda(如有请忽略)
https://repo.anaconda.com/miniconda/Miniconda3-latest-Windows-x86_64.exe - 2.下载C++Make Tools(靠这个编译c++,如有Visual Studio请忽略)
https://aka.ms/vs/17/release/vs_BuildTools.exe - 3.Conda中创建虚拟环境:
conda create -n desktopai python==3.8.19
conda activate desktopai - 4.下载代码:
cd /d “your dir” (导航到安装根文件夹)
git clone https://github.com/AARG-FAN/DesktopAI.git - 5.cd /d DesktopAI (导航到项目的文件夹)
- 6.安装依赖:
pip install -r requirements.txt - 7.启动:
python main.py
- 1.按空格键可以跳过当前语音播放
- 2.右键休眠可以让他休眠不再处理文字。右键可以选择设置特定词语作为唤醒词,休眠后检测到特定词语后会唤醒AI
- 3.更换角色形象可以在文件夹assest中的role1和role2文件夹中同名替换gif1(静止)和gif2(说话)文件即可
- 安装成功后启动项目时候不要用vpn,因为api服务商在国内,vpn影响通信
- Adding...