Skip to content

imamik/consistent-character

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

47 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Consistent Character - ComfyUI Docker Environment for GPU Cloud Platforms

A specialized Docker environment designed to run MickMumpitz's amazing ComfyUI workflows on GPU cloud platforms like RunPod and Vast.ai. This project makes it easy to create consistent character generations using advanced Stable Diffusion techniques.

Watch MickMumpitz on YouTube Support on Patreon

Overview

This project provides a containerized environment specifically optimized for running ComfyUI workflows, with a focus on character generation and consistency. It comes pre-configured with all necessary models, extensions, and custom nodes needed to run MickMumpitz's workflows (Standard, Advanced, PULID) out of the box on GPU cloud platforms.

Features

  • 🐳 Ready-to-use Docker environment
  • 🎨 Pre-configured ComfyUI installation (accessible on port 3000)
  • 📂 Web-based file browser (accessible on port 4040)
  • 🤖 Pre-installed model categories:
    • Base models (JuggernautXL, Flux1)
    • IP-Adapters (SDXL & SD1.5 variants)
    • ControlNet models
    • Face detection models
    • Upscalers
  • 🔧 Optimized for GPU cloud platforms (RunPod, Vast.ai)
  • 🔒 SSH access capability
  • 🌐 NGINX proxy configuration

Prerequisites

  • RunPod or Vast.ai account
  • GPU instance with sufficient VRAM (recommended: 24GB+)
  • Docker installed (for local development)

Quick Start

Cloud Deployment

  1. Pull the Docker image from the registry
  2. Deploy on your preferred GPU cloud platform
  3. Access ComfyUI through port 3000
  4. Use the web file browser on port 4040 to manage your files
  5. Import MickMumpitz's workflows and start creating!

Local Usage

To run the environment locally:

# Pull the image
docker pull ghcr.io/joyandmighty/consistent-character:latest

# Run the container
docker run -it --gpus all \
  -p 3000:3000 \
  -p 4040:4040 \
  -p 8080:8080 \
  -v /path/to/your/inputs:/workspace/input \
  -v /path/to/your/outputs:/workspace/output \
  -v /path/to/your/models:/workspace/models \
  ghcr.io/joyandmighty/consistent-character:latest

# Access ComfyUI at http://localhost:3000
# Access file browser at http://localhost:4040

Note: Replace /path/to/your/inputs, /path/to/your/outputs, and /path/to/your/models with your actual local paths for input, output, and models directories.

Docker Image

The Docker image is available on GitHub Container Registry:

docker pull ghcr.io/joyandmighty/consistent-character:latest

Available tags:

  • latest - Latest stable release
  • vX.Y.Z - Specific version releases (e.g., v1.0.0)
  • sha-XXXXXXX - Specific commit builds

You can also reference the image directly in your cloud platform template:

ghcr.io/joyandmighty/consistent-character:latest

Each release is automatically built and published using GitHub Actions. The version number is managed through semantic versioning based on commit messages.

Development

To build the Docker image locally:

docker build -t consistent-character .

License

This project is licensed under the MIT License - see the LICENSE file for details.

Acknowledgments

  • Special thanks to MickMumpitz for creating the amazing ComfyUI workflows that inspired this project
  • ComfyUI team for the base application
  • All model creators and contributors

Support

  • For workflow-specific questions, please support MickMumpitz on Patreon
  • For environment/deployment issues, please open an issue in this repository

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Packages

 
 
 

Contributors 2

  •  
  •