On today's episode of "LoRA Nobody Asked For"
It's not super stable on account of how chaotic the images it's trained on are, so some trial and error will definitely be required. Recommended weight around 1 - 1.5 (1.6 starts to get a little woobly, but could also work given the subject)
Some trigger words I recommend trying that I didn't put: doodle, monochrome, manga style, chibi
Should theoretically work with and without the blindfold, as well as young and adult Gojo. It's trained on a bunch images from the manga of silly doodle Gojo lol
Also does a surprisingly decent job at generating a manga/sketch style of Gojo? More testing will be done.