LoRA for generating in the style of doujinshi and game artist Izakaya Yocchan.
MOSS
Trained on MOSS-Image-Base, tested on MOSS-Image-SFT
Examples made on default comfy workflow with PiD upscaler.
er_sde + simple, 50 steps, cfg 3, lora strength 1. Resolution - 2 mpx on multiple of 16.
Stating the obvious - to render in this style you need it to be illustration, 2d, anime, etc.
Tl:dr do not use it with realistic description if your goal is to emulate this style
Anima
Tested on turbo, base, and aesthetic.
Turbo lora +aesthetic gives more close resemblance to the original.
Tested on natural language, CLIP styled prompts and mix of them.
All of examples are made on strength 1.
Quality tags masterpiece, best quality, score_7 and such are important.
Workflow is inside of the pictures.
Illustrious
Make sure to use keyword Izakaya_Yocchan
Made compatible with Illustrious 1.0 still works with 0.1
Model is prone to fill the scene with characters, so use 'solo', '2girls' and negatives 'multiple characters' etc
Works with strength between 0.8-1, 1 can be glitchy depending on complexity of the prompt. All of examples are made in 1, but you need to be specific about the scene.