Features:
Fast, simple, and convenient.
Description:
best used with WAI-NSFW-Illustrious-SDXL (LoRA: 1~2).
Best Choice: WAI-NSFW-Illustrious-SDXL
It’s compatible with Pony LoRA and works with i5 XL model training.
Versions:
V1: Based on Illustrious XL 1.0 and V0.1, capturing Tokisaki Kurumi’s classic appearance. The intricate eye detail data poses a challenge, requiring 70% focus on the upper body. As a result, lower body descriptions depend heavily on prompt keywords.
V2: Due to the detailed yellow eyes of Tokisaki Kurumi, this version aims to condense everything into a single prompt keyword. Additional drawing work may be needed over time to enhance the eye detail dataset.
I love condensing characters into a single prompt keyword. Here’s the thing: NSFW prompts can get insanely detailed and complex too many keywords actually dilute the details. With MM_LoRA, I want to shrink it down to just one keyword and still have the character accurately restored. Keep it simple that way, you can easily embrace your girl!