Animora Style for MOSS-Image 🌹
My personal semi-real anime house style, trained for MOSS-Image on about 90 images of many different characters. Expect clean hands, rich lighting and the same style across very different scenes and characters.
Trigger: animora style
Recommended settings: strength 1.0, sampler er_sde, scheduler sgm_uniform, 30 steps, CFG 4, 768x1344.
Good at: expressive faces, dramatic lighting, fantasy and everyday scenes, and stacking with character LoRAs.
Weak spots: text and signs come out as gibberish, and back views can sometimes look scrunched. For back views, lower the strength to about 0.7 and add long torso, natural proportions to the positive prompt.
Pairs with: MOSS Fire & Forget Wildcards Workflow - The workflow is also embedded in the individual showcase images.