Sign In

Asura NoobAi (DirtyEro) | LoRA

9.1k

Download

1 variant available

SafeTensor

54.78 MB

Verified:

Type

LoRA

Stats

132

9.1K

2.4K

Reviews

Published

Dec 25, 2024

Base Model

NoobAI

Hash

AutoV2
14DE492EF4

Trigger Words

red oni, colored skin, oni horns, broken horn, asymmetrical horns, pointy ears, orange eyes, slit pupils, eyebrow cut, short hair, gradient hair, hair between eyes, short twintails, medium breasts, demon tail, toned
abs
black nails
hoop earrings, red backwards hat, studded collar, necklace, collarbone, black sports bra, fishnet top, fishnets, short sleeves, midriff, fingerless gloves, black gloves, studded belt, torn jeans
default creator card background decoration
Chameleon Badge
Manityro's Avatar

Manityro

The Oni OC Asura from "DirtyEro". Works best between 0.8-1.0 weight. Trained on NoobAi-1.1. Request form

Warning: As this character is trained on a NoobAi and it might not work properly on illustriousXL-0.1 based models. I recommend using a NoobAi derived models, there are some in the suggested resources to try.

Always needed:

"red oni, colored skin, oni horns, broken horn, asymmetrical horns, pointy ears, orange eyes, slit pupils, eyebrow cut, short hair, gradient hair, hair between eyes, short twintails, medium breasts, demon tail, toned"

  • Add "abs" where visible for accuracy

  • Add "hoop earrings" for her earrings.

  • Add "black nails" for her proper nails

  • Hairstyles:

    • Remove "short twintails" and replace with "short ponytail" for her other style.

Outfits/notes:

  • Main outfit: "hoop earrings, red backwards hat, studded collar, necklace, collarbone, black sports bra, fishnet top, fishnets, short sleeves, midriff, fingerless gloves, black gloves, studded belt, torn jeans"

  • Check samples above for outfit/prompt ideas.

  • Flexible and very NSFW able due to nice data variety.

Port is very nice tomboy for illust, results here are fantastic and soo much cleaner than on Pony so I hope you all enjoy.

Feedback and reviews are always appreciated.

Training data is in the Lora metadata.

Trained on 28 images.