Generative AI on Docker Project

Stable Diffusion WebUI and KohyaSS, ComfyUI, InvokeAI, Fooocus, and more Generative AI Applications on Docker

Setup

You need NVIDIA GPU and CUDA 12.4, Docker and NVIDIA Container toolkit, if you are using Windows 10/11, you also need enable WSL2. it is recommanded on Amazon Web Services, Microsoft Azure, Google Cloud Platform.

How to install

git clone https://github.com/bean980310/stable-diffusion-docker-project.git

First, run the run_first.ipynb on Jupyter Notebook(if not installed cuda, cudnn, docker, and NVIDIA Container toolkit.) and then run the download.ipynb on Jupyter Notebook.

And you must put models in directory stable-diffusion-models/models.

cd <your_path_of_dir>/stable-diffusion-docker-project
docker compose -f docker-compose.pull.yml up -d <service_name>

example:

cd ~/stable-diffusion-docker-project
docker compose -f docker-compose.pull.yml up -d stable-diffusion-webui 

How to install extension of stable-diffusion-webui:

cd <your_path_of_dir>/stable-diffusion-docker-project/workspace/stable-diffusion-webui/extensions
git clone <stable-diffusion-webui_extension>

How to install ComfyUI Manager:

cd <your_path_of_dir>/stable-diffusion-docker-project/workspace/ComfyUI/custom_nodes
git clone https://github.com/ltdrdata/ComfyUI-Manager.git

Usage

docker compose -f docker-compose.pull.yml up -d

or

docker compose -f docker-compose.pull.yml up -d <service_name>

Features

AUTOMATIC1111

KohyaSS

ComfyUI

InvokeAI

Fooocus

Forge

SD.Next

Open-WebUI

EasyDiffusion

SwarmUI

FaceFusion

Omost

stable-audio-tools

Contributing