TroubleChute Logo
STABLE DIFFUSION

Stable Diffusion UI v2


Published: Sep 29, 2022
Last Edit: Oct 13, 2022
AI Stable Diffusion
651 Words, 3 Minutes.

Watch the video:


Timestamps:
00:00 - Explanation + Is it really free? Yes.
01:30 - Stable Diffusion UI
01:55 - Download Stable Diffusion UI
03:11 - Install Stable Diffusion UI (Windows)
05:01 - Install Stable Diffusion UI (WSL/Linux)
06:45 - Opening Web UI for Stable Diffusion UI
07:58 - Exploring the GUI & Styles
08:30 - Saving all generated images automatically
09:21 - Image settings explained
09:48 - Generate multiple images at once
10:58 - Auto-fix faces & eyes & Auto-Upscale
11:37 - Example
13:05 - Saving images
13:15 - Using images as input (Initial image)
14:19 - Using seed & Auto-fix faces (and why not to)
15:19 - Inpainting
16:22 - What saving all images does
17:52 - Explicit filters & NSFW?
18:10 - What next?

Stable Diffusion UI v2

Stable Diffusion is a free AI model. If you’re paying for anything other than someone else’s hardware to generate images, you’re likely being cheated for your money. I’ve seen some paid programs accomplish far less but have good marketing.

You do not need to pay for Stable Diffusion, or a UI, or a Web UI.

This guide shows you installing the Stable Diffusion UI v2 by cmdr2 under WSL. It’s a simple, beginner-friendly interface for Stable Diffusion! It works perfectly well under regular Windows, and I’ll touch on that in a second.

If you’re looking for the next step, consider checking the guide for AUTOMATIC1111’s Stable Diffusion WebUI

Why this project above others?

It’s simple to set up and get started with. Nothing crazy to set up, and you can run it off your own computer with no investment or monthly subscription! It’s fantastic stuff and a great place to start. Being able to watch it slowly generate images as well is a great way to learn about how image generation AI software works as well.


About my install

You can install this on Windows by following the official guide, but I choose to run it under WSL. It’s a little easier, at least for me. Well, that and the fact that I closed the installer at the perfectly wrong time. This resulted in me not being able to start it under normal windows. It works fine on other computers I tested it on.

So, just make sure you don’t close the window when installing it!

I’m using Ubuntu 22.04 in WSL 2. If you choose to join me in using WSL, then if you haven’t already got Windows WSL set up and installed, check my guide for Installing and Using Ubuntu in WSL.

Installation

Windows

The Windows installation is simple.

  1. After downloading the zip, extract it to C:\stable-diffusion-ui, or another base-level directory, like A:\stable-diffusion-ui.
  2. Run Stable Diffusion UI.cmd.

That’s it. It should now download a few things and get started. You can open the web page and start bashing away at creating some images!

Linux, or WSL

The installation steps for WSL or Ubuntu/Linux are straightforward as well.

Navigate to the GitHub Releases page, Right-Click the latest Linux release tar.xz file, and copy the URL. We will download this file within WSL/Ubuntu.

Run the following commands:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
# Using the link we just copied in place of <URL>
cd ~
wget <URL>

# Extract the tar.xz with
tar -xf stable-diffusion-ui-linux.tar.xz

cd stable-diffusion-ui/

# To start, run
./start.sh

In the future, use cd ~/stable-diffusion-ui && ./start.sh to start the AI without worrying about its location.

Opening the Web Interface/UI

In the terminal on either Windows, Linux, or WSL, you should see http://0.0.0.0:9000 or a similar link. Simply click it, Ctrl+Click it, Double-Click it or just copy-paste it into a browser to open the page.

Just like that: You’re in.

Where to from here?

The video covers setting up auto-save and other features. This was more just for easy command reference and quick-start for installing Stable Diffusion UI v2.

TCNO TechNobo / TroubleChute © Wesley Pyburn (TechNobo / TroubleChute)