This LoRA aims to replicate the distinct visual style of Mnemosynekoto.
The dataset used for training was sourced from their artwork shared on X.
Usage Recommendations:
Best results on WAI-NSFW-Illustrious v14, but it should perform well on most Illustrious-based fine-tunes.
CFG: 2–8 (recommended: 5)
Steps: Varies depending on the base model
Resolutions: 1024×1024, 896×1152, 832×1216
Enjoy and have a good day!
Description
Dataset collected from their X profile: https://x.com/MnemosyeRin
Cleaned and resized multi-view illustrations. Mugshots of characters from several series, including Demon Slayer, Bleach, and others, were added to better replicate the style.