>>17137769If you read the Stable Diffusion github wiki, know a bit how python and AI works and follow the tutorial from /g/, /e/ or /h/, you should be able to install and make it run without problems.
I am collecting some things to note for myself, I may share them now:
CivitAI and HuggingFace has most/all the models and lora, but /h/ and /e/ are sharing some too:
https://civitai.com/https://gitgud.io/gayshit/makesomefuckingporn#lora-listhttps://rentry.org/hdglorarepohttps://rentry.org/trashcollectsControlnet and Openpose, like the only really recommended extension to get:
https://rentry.org/dummycontrolnetSD WEBUI SETTINGS TO CHANGE
- Clip Skip 1 to Clip Skip 2
HOW TO USE MODELS AND VAE AND LORA AND EMBEDDINGS [II]
-Models go into models/stable-diffusion folder
-VAE have to be renamed to the very same name as the mode + vae.pt, go in the same
folder with the model
- LORA go into models/lora folder
- embeddings/textual inversion go into embeddings folder
GENERIC ANIME NEGETIVE PROMPT (AOM3) [III]
(worst quality, low quality:1.4), (realistic, lip, nose, tooth, rouge, lipstick, eye shadow:1.0),
(dusty sunbeams:1.0), (abs, muscular, rib:1.0), (depth of field, bokeh, blurry:1.4),
(grayscale, monochrome:1.0), text, title, logo, signature
WEBUI USER BAT EDIT LINE 6 [IV]
set COMMANDLINE_ARGS=--precision full --no-half --no-half-vae --medvram (For GPUs with 4-6 GB VRAM)
set COMMANDLINE_ARGS=--no-half-vae (For GPUs with 8 GM VRAM and over)
LIST OF ARTIST RECOGNIZED BY STABLE DIFFUSION - PROMPT SYNTAX
"BY (ARTIST NAME:1.0)" OR ARTIST NAME [V]
Anything
These artists' names can be attached to any prompt. While they may be known for specific subject
matter, their styles are very distinct and can apply to anything.
Alan Lee
Cyril Rolando
David Mack
Donato Giancola
Greg Rutkowski
Ismail Inceoglu
John Berkey
Michael Garmash
Peter Mohrbacher
Sparth
Vincent Di Fate 