ComfyUI + Kohya_ss Docker

Overview

This repository contains the Dockerfile and necessary files to build a Docker image for running ComfyUI, Kohya_ss, and JupyterLab. The Docker image is designed to provide a convenient environment for using these tools on Runpod.io.

Services and Ports

The Docker image includes the following services, each accessible on different ports:

• Kohya_ss: Accessible on port 3001
• Tensorboard: Accessible on port 6006
• ComfyUI: Accessible on port 8188
• JupyterLab: Accessible on port 8888

Nginx is configured as a reverse proxy to route requests to the appropriate services based on the URL path, listening at port 8080.

Usage

You can use JupyterLab to upload the images for training on Kohya_ss, and download the weights. You can also open terminals inside JupyterLab to run commands if needed.

How to Build and Run

Build the Docker Image

To build the Docker image, run the following command:

docker build -t yourusername/kohya-comfyui-docker:latest .

Replace yourusername with your Docker Hub username or any preferred tag.

Run the Docker Container

To run the Docker container, use the following command:

docker run -d --gpus all -p 8080:8080 -p 3001:3001 -p 6006:6006 -p 8188:8188 -p 8888:8888 yourusername/kohya-comfyui-docker:latest

Replace yourusername with your Docker Hub username or the tag you used when building the image.

Contributing

Contributions are welcome! Please feel free to submit a pull request or open an issue if you encounter any problems, or have suggestions about apps/plugins to add.

Acknowledgements

I would like to thank the following for their valuable work:

• Valyrian Tech for developing the ComfyUI_with_Flux repository.
• Ashley Kleynhans for developing the kohya-docker repository.