Anima Version of the model uploaded, the model is trained with the same dataset. There are still bit of overfitting, but it should be ok as long as detailed prompts are given.

Recommended weight: 1.0, lower the weight when used with other loras.

This is the only lora used for the sample images of the anima version model.

————————————————————————————————————————

Trained on images created with NAI, as I like the new oily/semi-realistic texture of its newest model.

A bit of overfitting on background, so best to remember to prompt the background.

Amazing outcomes when used with other style loras, recommended weight: 0.7

Sample images used this lora with weight 0.75 and https://civarchive.com/models/2084317/pixiv-laoyblaoyb2-style

with weight 0.35 other than the character loras