Run Stable Diffusion AI At Home — No Code Guide

Image by Jim Clyde Monge

Recently, startup StabilityAI announced the release of Stable Diffusion, a powerful AI image generator that can now run on standard graphics cards.

Note: You don’t need any programming experience to follow along, it’s all spelled out.

Pre-requisites

Before doing the steps below, make sure you have all the requirements to run the AI model in your local hardware.

• NVIDIA GPU with at least 4GB VRAM
• At least 10GB of space in your local disk

Now on to the steps.

Step 1. Setup Git and Python environment

• Download and install the latest Anaconda Distribution here.
• Download and install the latest Git here.

Step 2. Download the model weights

Click on this link and download the latest Stable Diffusion library.

As of this writing, the latest version is v1.4-original.

Image by Jim Clyde Monge

Under the “Files and versions” tab, click on the checkpoint file and download it.

Image by Jim Clyde Monge

Note: The file is more than 4GB, so while you wait, go get a coffee or do something else.

Step 3. Download Stable Diffusion Repository

Download the source project from this GitHub repository.

The repository is called Stable Diffusion Dream Script, it’s a fork of the original source code modified by lstein. Shout out to that awesome guy.

Image by Jim Clyde Monge

Extract the files into your local. Inside the “ldm” folder, create another folder named “stable-diffusion-v1”. Like so in this screenshot.