GPT-SoVITS容器化部署

1、从codewithgpu拉取sovits镜像

docker pull registry.cn-beijing.aliyuncs.com/codewithgpu2/rvc-boss-gpt-sovits:gpM1WfbTsA

2、安装nvidia-container-toolkit，让docker容器可以使用GPU资源

配置生产仓库

curl -fsSL https://nvidia.github.io/libnvidia-container/gpgkey | sudo gpg --dearmor -o /usr/share/keyrings/nvidia-container-toolkit-keyring.gpg \
  && curl -s -L https://nvidia.github.io/libnvidia-container/stable/deb/nvidia-container-toolkit.list | \
    sed 's#deb https://#deb [signed-by=/usr/share/keyrings/nvidia-container-toolkit-keyring.gpg] https://#g' | \
    sudo tee /etc/apt/sources.list.d/nvidia-container-toolkit.list

更新软件源

sudo apt-get update

安装 NVIDIA 容器工具包

sudo apt-get install -y nvidia-container-toolkit

重启docker

sudo systemctl restart docker

3、创建sovits容器，设置启动GPU，设置端口号、以及共享内存大小

以非持久化存储方式创建

docker run -id --name=sovits --gpus all -p 9874:9874 -p 9873:9873 -p 9872:9872 -p 9871:9871 --shm-size 7g registry.cn-beijing.aliyuncs.com/codewithgpu2/rvc-boss-gpt-sovits:gpM1WfbTsA

以持久化存储方式创建（volume挂载）

docker run -id --name=sovits --gpus all -p 9874:9874 -p 9873:9873 -p 9872:9872 -p 9871:9871 --shm-size 7g -v /home/panzhe/sovits_docker_v/logs:/root/GPT-SoVITS/logs -v /home/panzhe/sovits_docker_v/audio_data:/root/GPT-SoVITS/audio_data -v /home/panzhe/sovits_docker_v/output:/root/GPT-SoVITS/output -v /home/panzhe/sovits_docker_v/GPT_weights_v2:/root/GPT-SoVITS/GPT_weights_v2 -v /home/panzhe/sovits_docker_v/SoVITS_weights_v2:/root/GPT-SoVITS/SoVITS_weights_v2 registry.cn-beijing.aliyuncs.com/codewithgpu2/rvc-boss-gpt-sovits:gpM1WfbTsA

4、启动gpt-sovits服务

echo {}> ~/GPT-SoVITS/i18n/locale/en_US.json && source activate GPTSoVits && cd ~/GPT-SoVITS/ && python webui.py zh_C

查看GPU使用情况

watch -n 1 nvidia-smi

-n 1代表每隔1秒刷新一次，ctrl+c退出。