All the commands are tested and it works, the detail rendered can be targeted on each element all and notified, for maximum detail ;)
Description
FAQ
Comments (8)
What does it mean? xD
basically I looked for the possibility of detailed rendering possible by Stable diffusion 1.5 to have the maximum of small details possible and the most work possible
The wildcard extension to Automatic1111 is explained in the link. This appears to be a text list of specific detail types such as 'Detailed Lips'
@badbunny exactly in the wildcards file there is a lot of things missing, so I looked by myself what works and what doesn't work and redid the complete file
@XT404 Alright, thanks.
@FrozenGT ;)
@badbunny So.. isn't the wildcards script basically a randomizer... so that I can make a haircolor.txt and add different haircolors there and by including the haircolor command (with double underscores on both sides.. can't write it here as it bolds it haha) and then start a batch and tadah, different kinds of haircolors during a longer generation.
So I think I got that right... right? But I still don't quite get how it works here? Is.. it a ... detail randomizer... or something different than a wildcard, where it somehow includes all possible "detailed" commands applicable to a pictures prompts somehow, automatically? So not just randomizing some things, but ... detailing all the things?
Sorry, I'm a little new to these scripts and such...
I wonder if this is moreso meant for super high res generations my 8gb 3060ti can't really pull off anyway...
@sarcasm83334 Once you have added the Automatic1111 extension you should see a new tab for managing wildcard files. Once you have that using prompts such as "__Pose__" will substitute one of the phrases from the wildcard file into your prompt with each iteration. So you might run a batch of 20 images and ask for a wildcard of hair colours or art styles etc. similar to the affect you might get using the X/Y/Z option and having Search/Replace set on the prompt word you want to vary.