EDIT: instructions

Pros: stable, pretty, it preserves the original prompt while adding colors and giving a realistic illustration style, vanishing points

Cons: kinda weak, hands (quite a bit)

I did another run, trying to strenghten the impact while reducing hand problems: THIS VERSION IS OBSOLETE

_____________________

A chunk of Civit users gave generated images to produce a LoRA. Here the original article:

(CLOSED) Community chaos LoRA project - Keep the chaos rolling! | Civitai

In alphabetical order:

• Colorblind_duck

• Icy799

• Iusedtobeme

• Moritanik

• Ouinche

• Silvana_

• Unexpectedlyprovided

• Zlia

• And, of course, the instigator, me: Argiop33

Then I tagged everything (badly), put the contributor's name in the images tags (because why not), sacrificed a fat lamb to Nyarlathotep and started the Civit trainer with Prodigy and some random number ('cause, you know, prodigy) screaming "YOLO" (I am a boomer, you know).

Now, you know the drill: you post here, I gain buzz. You use it, I gain buzz. Thank you. It should be quite safe to use in the daily challenge and other things. See ya.