Install & Run Stable Diffusion on AMD GPUs: A Simple Guide

Running Stable Diffusion on AMD GPUs can be challenging due to limited official support. This guide simplifies the installation process, providing a clear, step-by-step walkthrough to help users overcome common hurdles and successfully generate AI images on their AMD hardware. Let’s dive into getting Stable Diffusion up and running on your AMD GPU!

Why AMD GPU Installation Can Be Tricky

Unlike NVIDIA GPUs, AMD GPUs often require more effort to set up with Stable Diffusion. Official support is not as robust, leading to a potentially frustrating experience for new users. Many online tutorials are outdated or incomplete, compounding the issue. This guide addresses these problems by offering a streamlined and easy-to-follow process.

Most of the tutorials are outdated, so it’s important to follow the latest tutorials to avoid any issues. This step-by-step Tutorial is the most updated resource on the internet. The process for NVIDIA GPUs are much simpler, but AMD GPUs can be difficult. This guide provides an easy and simple solution to make the process as seamless as possible.

Keyword Density Matters: This article includes frequent references to "Stable Diffusion," "AMD GPUs," and "installation" to ensure optimal Search Engine visibility. The terms "artificial intelligence," "AI images," and "image generation" are also included to increase SEO value and relevance.

Prerequisites: What You'll Need

Before beginning the installation, ensure you have the following:

• Python: Python 3.10.6 is highly recommended for its proven compatibility.

  Download it from the official Python website. Though a more updated version can be used, Python 3.10.6 is the most stable and will provide the least amount of issues.

• Git: Git is a distributed version control system, meaning that it will be able to Record changes of files over time, so that you can recall specific versions later.Git is essential for cloning the necessary Stable Diffusion repository. Download Git for Windows from gitforwindows.org.
• Stable Diffusion Model: You'll need a Stable Diffusion Model file, such as DreamShaper 5, to generate images. Models can be downloaded from resources like Civitai.

  Make sure you pick a model that fits your intended creative content.

Installing these programs are fairly simple. Just download the installer and follow the steps. Just make sure you don’t skip anything to ensure that the program is installed correctly. It’s also important that you have the necessary space in your hard drive to download and install these applications.

If you run into any issues during the process, don’t hesitate to contact the customer support of the application you are trying to install. They will have the correct guidance in troubleshooting any errors you encounter.

Popular Models

Choosing the right Stable Diffusion model is key to achieving desired results. Different models excel at generating specific types of images. Here's a look at some popular choices, as suggested by Stable Diffusion Art:

• Stable Diffusion v1.5: A versatile, general-purpose model. Check the website for a download link. v1.5 is a strong model that can be used interchangeably with v1.4.

• F222: F222 is for portraits and produces beautiful female portraits with correct body part relations. This is good at getting aesthetically pleasing clothing, with a wardrobe including “dress” and “jeans”.

• Anything V3: Excellent for anime-style art. It has a special-purpose and is trained to produce high-quality anime-style images.

  

• DreamShaper: A well-rounded model fine-tuned for portrait illustration that balances photorealism and computer graphics.

• Realistic Vision v2: Ideal for generating realistic human images. It also generates great portrait models with a high degree of detail.

• Deliberate V2: Another solid choice for realistic illustrations, often producing surprisingly good results.

• Mo-di-diffusion: For producing Pixar-like art.

• Inkpunk Diffusion: Perfect for creating striking profile pictures on social media.

Here's a table summarizing the models and their strengths:

Step 1: Create a Dedicated Folder

First, create a new folder where Stable Diffusion will be installed. This helps keep your files organized. For example, create a folder named "SD" with a subfolder named "webui" inside it.

This is an important first step, but is often overlooked, and is a huge proponent of proper file management. You can create these folders anywhere you want, but you should create it somewhere you remember. This will make it much easier if you need to access the file in the future.

Step 2: Open Command Prompt in the Folder

Navigate to the “webui” folder you just created. In the address bar, type "cmd" and press Enter. This opens a command Prompt directly within that folder. This is an important process because it allows you to install the stable diffusion directly to your computer and be easily accessible.

Typing the stable diffusion address in the prompt also eliminates any errors or problems that can arise.

Step 3: Clone the Stable Diffusion Repository

Copy and paste the following command into the command prompt and press Enter:

git clone https://github.com/hqqytiger/stable-diffusion-webui-directml cd stable-diffusion-webui-directml && git submodule init && git submodule update

This command clones a specific fork of Stable Diffusion optimized for AMD GPUs and updates necessary submodules. I’ll include it in the description for your ease.

Don’t type out the prompt, as it is very long, and you may encounter issues typing it manually. Just copy and paste!

Step 4: Place Your Stable Diffusion Model

Copy your downloaded Stable Diffusion model (.ckpt or .safetensors file) into the stable-diffusion-webui-directml\models\Stable-diffusion directory. This is the most important part of the tutorial, as without this model, the program won’t function. If the checkpoint isn’t placed here, the program will fail. Make sure you don’t skip or miss this step.

Make sure to place a file, not a folder. The file also needs to be of the correct extension, otherwise it won’t work.

Step 5: Edit the webui-user.bat File (Conditional)

If you have a lower-end AMD GPU (with 4-6GB VRAM), edit the webui-user.bat file to add optimization flags. Right-click the file, select "Edit," and add the following line:

set COMMANDLINE_ARGS=--opt-sub-quad-attention --lowvram --disable-nan-check

Save the file. This step helps reduce VRAM usage. If you don’t have less vRAM, then you can skip this step. Make sure your parameters are correct, and if you do have more VRAM, you don’t need to edit anything.

Step 6: Run webui-user.bat

Double-click the webui-user.bat file to start Stable Diffusion. It may take some time to download dependencies and set up the environment on first run. If the installation gets stuck, try pressing Enter in the terminal. This will run all the necessary command lines and setup the program so that you can start using it.

If you have any errors, make sure to reference the troubleshooting area in the tutorial. Otherwise, you can always Seek help from online community members.

Step 7: Access Stable Diffusion in Your Browser

Once the process completes, you'll see a local URL (usually http://127.0.0.1:7860) in the terminal. Copy and paste this into your web browser to access the Stable Diffusion web interface. Once copied, you can now use Stable Diffusion through your web browser. You can even bookmark this address for ease of access in the future. If you have any issues, try restarting your computer.