Thank you to these folks for all the work they are doing and everyone else involved!
SD Webui
https://github.com/AUTOMATIC1111/stable-diffusion-webui
Checkpoints
https://huggingface.co
GFPGAN – Face Restoration
https://github.com/TencentARC/GFPGAN/
*******************************
***Pre-Req Downloads first***
*******************************
Install Python
https://www.python.org/ftp/python/3.10.8/python-3.10.8-amd64.exe
*choose the option to add to ENVIRONMENT, and git as your repo manager
Install GIT
https://github.com/git-for-windows/git/releases/download/v2.38.1.windows.1/Git-2.38.1-64-bit.exe
***********************************
***Stable-Diffusion Installation***
***********************************
SD Webui –
Extract to C:stable-diffusion-webui (or anywhere else – this will be the base directory – can be anywhere or named anything you want)
https://github.com/AUTOMATIC1111/stable-diffusion-webui
Checkpoints –
Default
You only need one of these – but you must sign up for an account (just needs name and email, and then email verified. Trained w/ NSFW images also so must accept the eula that you won’t do anything harmful with it)
Copy 1 of these .ckpt files into the directory below, and RENAME it to model.ckpt
C:stable-diffusion-webui-mastermodelsStable-diffusionmodel.ckpt
**note – once you set model.ckpt (to load as your default) in the WEB GUI you will be able to select multiple checkpoints – these can be named anything in the same folder – you just have to use 1 as a default model
https://huggingface.co/runwayml/stable-diffusion-v1-5/resolve/main/v1-5-pruned.ckpt – 7.7GB, ema+non-ema weights. uses more VRAM – suitable for fine-tuning
https://huggingface.co/runwayml/stable-diffusion-v1-5/resolve/main/v1-5-pruned-emaonly.ckpt – 4.27GB, ema-only weight. uses less VRAM – suitable for inference
Inpainting
If you will be in-painting (using AI to alter parts in a picture) grab the checkpoint below and put it in this folder – you do not need to rename this
C:stable-diffusion-webui-mastermodelsStable-diffusionsd-v1-5-inpainting.ckpt
https://huggingface.co/runwayml/stable-diffusion-inpainting/resolve/main/sd-v1-5-inpainting.ckpt
GFPGAN – Face Restoration
Copy to the root directory, leave filename as is –
C:stable-diffusion-webui-masterGFPGANv1.4.pth
https://github.com/TencentARC/GFPGAN/releases/download/v1.3.0/GFPGANv1.4.pth
Launch SD webui by double clicking on
c:stable-diffusion-webui-masterwebui-user.bat
*****************
EXTRA MODEL DATABASES
You do not need to load these, but they are here –
They have extra functions on images – can run through img2img in the webui for different results – if the model has dependencies it should auto-download the first time you use it
These would go in C:stable-diffusion-webui-mastermodelsESRGAN
https://upscale.wiki/wiki/Model_Database
**load these sparingly, I’ve noticed if you want specific things, you should use specific datasets. Read up on these if you don’t know what they are, they are sorted fairly well by what they do (clothing, background enhancement, etc.)
Showcase page for SD webui
https://github.com/AUTOMATIC1111/stable-diffusion-webui/wiki/Features
More info about the system – inputting your own pictures
https://github.com/AUTOMATIC1111/stable-diffusion-webui/wiki/Textual-Inversion
Inpainting
https://github.com/runwayml/stable-diffusion#inpainting-with-stable-diffusion
**NOTES**
If you have anything under a 2060 or a card with less than 10GB you may need some switches –
Edit the file below
C:stable-diffusion-webui-masterwebui-user.bat
Set the COMMANDLINE_ARGS to be the same as the line below.
set COMMANDLINE_ARGS=–medvram –precision full –no-half
OR
2-4GB cards use
set COMMANDLINE_ARGS=–lowvram –precision full –no-half
Can’t use GPU? Use the CPU (MUCH SLOWER)
set COMMANDLINE_ARGS=–use-cpu all