For PixAI users : Don't Import this, instead use this -> https://pixai.art/model/1959521350217314925
This model was trained by azki_is_very_cute. If you see this model reprinted at PixAI other than the address above, please let azki_is_very_cute know.
This is my LoRA! This is about Kaga Nazuna(花芽なずな), a vtuber affiliated with Vspo!. This LoRA supports her official 9th outfit.
Trained on 40 images, almost L2D screenshot.
Prompts
🍣Appearence :
nazuna9th, black hair, multicolored hair, hairclip✂️Hair style
two side up :
long hair, two side up, hair ribbon, white ribbonlong wavy hair, back hair bow :
long hair, wavy hair, back hair bowtwintails :
long hair, twintails, side hair bowmedium hair :
medium hair
🧥Outfits
Jacket :
pink coat, (white sweater:0.7), turtleneck, pink skirt, pencil skirt, fishnet pantyhoseJacketless :
white sweater, clothing cutout, shoulder cutout, long sleeves, turtleneck, pink skirt, high-waist skirt, pencil skirt, fishnet pantyhose
Weights
0.5~0.7
Troubleshooting
If Nazuna's coat is opened, lower
white sweater's weight. I recommend(white sweater:0.7).
Limitation
Due to the very high proportion of L2D screenshot, some facial expressions or poses may be incorrectly generated. In this case, lower the LoRA's weight.
Description
improved hair color reproduction.
reduce pose, facial expression overfitting.
fixed a bug that the coat sometimes opens.