Stablediffusion 2 has been released few days ago and it is accessible here: https://github.com/Stability-AI/stablediffusion

Unfortunatly, the official documentation to make it works is complicated and.. it does not work except if you change few things that I explain here.

Current installation recommandation

Following screenshot is the current installation recommandations.

My installation recommandations

Here are the command lines I used to make it work.

Clone the repo

First, clone the stablediffusion repo.

git clone https://github.com/Stability-AI/stablediffusion.git
cd stablediffusion

Change python 3.8.5 to python 3.9 in environment.yml

If you keep the official configuration of environment.yml, you will not be able to install xFormers.

gedit environment.yml

Find line with python (under dependencies) and change the version to 3.9

Create the conda env and install requirements

conda env create -f environment.yaml
conda activate ldm
pip install -r requirements.txt

Install xFormers using the official command

xFormers repo can be found here. To install xFormers:

conda install xformers -c xformers/label/dev

Download model weights

To run the model on your computer, you have to download the weights. They can be found here.

Run the model

Now, you should be able to run the model on your linux computer. The official example is excellent:

python scripts/txt2img.py — prompt “a professional photograph of an astronaut riding a horse” — ckpt <path/to/768model.ckpt/> — config configs/stable-diffusion/v2-inference-v.yaml — H 768 — W 768

Don’t forget to update the path to model weights you just downloaded.

You should be good !