Generate images with Stable Diffusion inside JupyterLab!

• start with one command
• GPU ready
• Notebooks included for:
  • text2image
  • image2image
  • inpainting
  • (coming soon) outpainting

• Docker
• Docker compose
• Nvidia-docker
• Log into Huggin Face and accept Stable Diffusion License
• Generate a Hugging Face token

• Fire up a terminal and clone this repo

  git clone https://github.com/pieroit/stable-diffusion-jupyterlab-docker.git
  

• Jump into the folder

  cd stable-diffusion-jupyterlab-docker
  

• Run container

  docker-compose up
  

• The first time (only) it will take a while. At the end you should see a link in the terminal, click on it and JupyterLab should open in the browser.

• Run the first notebook to download SD and save it on disk

• Use the second notebook to load SD from disk and use it. Enjoy!!!

• To stop the workstation, CTRL+c in the terminal and then

  docker-compose down