Use YiffyMix v62 model and always write jdhpps in the first place in your prompt!

Trained on 869 crops from the original movie with Danbooru style tags.

I decided to use a unique jdhpps token as a trigger word to prevent the model from mixing it's knowledge of the character with my data. I tried using Judy Hopps first, but it seems to be hard to retrain the model from what it already knows.

LoRA supports all character's outfits from the movie:

blue bodysuit, zpd emblem

dress uniform, tie, shoulder cord, utility belt, apd badge

casual clothes, grey t-shirt, bra strap, dark grey pants, earphones

police uniform, zpd badge, utility belt

meter maid uniform, orange vest, female police hat

farmer clothes, plaid shirt, rolled up sleeves, straw hat, jeans

Default prompt enchance:

BREAK, 8k, masterpiece, best quality, detailed

Default negative prompt:

blurry, worst quality, bad quality, lowres, low quality, normal quality, jpeg artifacts, signature, watermark, username, artist name, malformed, deformed, mutated, disfigured, ugly, gross, bad anatomy, bad hands, bad feet, bad face, bad eyes, bad mouth, crooked nose, crooked ears, uneven eyes, missing limbs, extra limbs, 5 fingers, fewer digits, cartoon, 2d, simple, sketch, monochrome, greyscale, comic, anime, manga, render, 3d, painting, drawing, illustration, oil painting, watercolor, bad perspective, human, human face, human hands, human body, human skin, human hair, no fur, smooth skin

LoRA also supports:

ears up, ears down, bunny teeth, glare in eyes, eye reflection, holding carrot pen, clenched teeth, direct sunlight, in shadow

And all emotions from the movie in Danbooru style, including:

:3, :<, :/, :d, :o

For base image I recommend using:

• euler_ancestral (euler a) sampler

• karras scheduler

• 20 steps

• cfgScale 5

And then do hi-res fix (img2img) with:

• euler_ancestral (euler a) sampler OR dmpp_2m (DPM++ 2M) sampler

• karras scheduler

• 0.4-0.45 denoise

• 20 steps

• cfgScale 5

I personally recommend LoRA strength 0.8-0.85, but feel free to find better option for you!

Trained on https://civarchive.com/models/3671?modelVersionId=1876492, and I recommend using this model.

Share your arts and don't forget to leave a review!