Docker environment for openvoice with an api written in python.
OpenAI sdk is supported:
https://github.com/ground-creative/openvoice-api-python/blob/main/examples/openai_lib_example.py
OpenVoice
https://github.com/myshell-ai/OpenVoice/tree/main
API
https://github.com/ground-creative/openvoice-api-python.git
API Client
https://github.com/ground-creative/openvoice-api-client-python
- NVIDIA gpu capable of running CUDA
- NVIDIA Container Toolkit
-
Follow the instructions here to install NVIDIA Container Toolkit
https://docs.nvidia.com/datacenter/cloud-native/container-toolkit/latest/install-guide.html -
Clone the repository
git clone https://github.com/ground-creative/openvoice-docker.git
- Select an Ubuntu image that is compatible with the NVIDIA driver version you plan to use from here
https://hub.docker.com/r/nvidia/cuda/tags
You can find out which version is supported by your driver in one of the following ways:
- With CUDA installed
nvcc --version
- Without CUDA you can look at the driver version on your system and check compatibility from here
https://docs.nvidia.com/deploy/cuda-compatibility/index.html
- Change environment variables in env.sample file and rename it to .env
-
Remember to add the correct CUDA image in the env file. Ex: DOCKER_IMAGE=nvidia/cuda:12.5.0-devel-ubuntu22.04
-
You can remove the link to the api from the env file if you jut want to install OpenVoice
Add the correct permissions when installing OpenVoice and/or API folders
USRID=$(id -u) GRPID=$(id -g) docker compose --project-name=openvoice up -d --build
Or
docker compose --project-name=openvoice up -d
Or
docker compose --project-name=openvoice up -d --build
Watch docker logs when you install first time as the container will need to install various dependencies
docker logs -f -t {CONTAINER_NAME}
You can check api logs in logs folder
tail -f app.log
It's possible to override environmet variable file params while building a container as shown below:
TEST=true COMMAND="tail -f /dev/null" docker compose --project-name=openvoice up -d