Ryujin Style

This is a Style Lora made by creating my own training images in stable diffusion, editing them in Luminar 4, and then applying them during training.

Its my own unique, cultivated style, featuring detailed and colorful images, solid and smooth lineart, and provide a decently consistent look across checkpoints.

It is suggested to use this with an Anime focused checkpoint, but it also has cool effects on Digital Art models as well.

No Trigger Words are neccessary, suggested strength at 0.8 - 1.0, but prompts like Lineart, Digital Art and illustration work very nicely.

All preview images were made with the following settings:

DPM ++ SDE Karras: 40 Steps

Hires x2: 4x_BS_DevianceMIP_82000_G: 10 Steps, 0.3 denoising strength

512 x 768

CFG Scale 7

Clip Skip 2

Color Enhance Extension: https://git.mmaker.moe/mmaker/sd-webui-color-enhance.git

Adetailer: https://github.com/Bing-su/adetailer.git