The Ultimate Stable Diffusion Guide v0.1 Release (Patreon)
Downloads
Content
Hey everyone!
A while ago I started working on my Stable Diffusion guide, and I finally reached a point where I'm satisfied enough to start sharing with you.
Right now, the guide will be exclusive to subscribers, but I'll probably release publicly once it's completed.
If you are a newcomer to the Generative AI world, I recommend checking the guide out as it contains a lot of useful tips on how to improve your image quality. The guide is mainly oriented towards people who have the hardware to run SD locally, but it can still be used by who uses cloud services.
Keep in mind a few things before reading the guide:
Some of the parts of this guide have been copied from my existing articles. Some extracts have been edited other have been kept as is.
Right now, only SD1.5 is covered.
If you want to follow the complete guide, you'll need a GPU with at least 6GB of VRAM. I've tested this with my GTX1060 and a crappy Radeon CPU without any issues.
Considering the pictures, this guide has over 200 pages!
In this v0.1 we are going to cover:
The different UIs available
How to install Automatic1111 (both the standard and the optimized version).
A complete overview of all the tabs available in the standard Automatic1111 interface.
Some tips about what to avoid when generating pictures.
How to generate high quality pictures with low VRAM (6GB).
A complete introduction on how to build a prompt.
The diverse levels of complexity of an AI generated picture.
Installing and using my extensions: Ranbooru and Workflow.
Using controlnets and IP-Adapters (basic).
Creating 3 different pictures step-by-step with distinct levels of complexity.
Here's the roadmap:
Installing ComfyUI.
A few ComfyUI workflows.
My SDXL Workflow.
Setting up Kohya for Lora training.
Training a simple LoRA.