Trained on images by Antin
Recommended strength: 0.8-1.0
Trigger word: "@antin"
I recommend that you don't use an upscaler or hiresfix for the best result, since it gets rid of most of the texture work that makes the style what it is.
Description
FAQ
Comments (5)
Qué coincidencia. XD
Actually crazy lol
NorlkSep 24, 2026
@NOORO dude, quick question: how do you get your models to have such a low file size? I'm totally new to this.
@Norlk I use a program called next-trainer https://github.com/wochenlong/lora-scripts-next. It runs locally and is very simple to use when you get it going.
@NOORO thx