Trained on 200+ images from multiple sources.
Recommended weights for txt2img - 0.6-0.9 , play around to get the best result .(I prefer using 0.8)
This is my first attempt to make LORA and i am pretty satisfied with the result that i get. The result may vary depending on how realistic what u want to be. Like in the instance of tag nose, makeup
will make the character look more photorealistic i guess. Other then that it'll likely stick to the art style of Naruto.
The generated results may have darker tone to the colour is the result of the dataset which also have it's grim and dark tone of her theme. If you want to avoid that lowering the weight can often help.
All of the hairstyle,colour and hair ornament are pruned so you don't need to specify it.
To get the effect of her paper jutsu use : paper, paper jutsu, floating papers, paper wings
U can try test each individual tages like in the case of if u only want wings use : paper wings
U can use akatsuki outfit with :
open clothes,zipper
for upper body
open clothes,zipper, pants
for full body
U can simply remove piercing if u don't like.
BTW this LORA is NSFW capable so i'll leave it at that.
I would love to hear the review and feedback on my model thanks for downloading.
All Images are produced in AOM2 Hard.
Some of the basic prompt u can try are :konan \(naruto\), full body, looking at viewer, pants, sandals, dark city, solo, standing, akatsuki outfit, labret piercing
konan \(naruto\), 1girl, solo, looking at viewer, simple background, white background, closed mouth, upper body, akatsuki outfit, labret piercing
konan \(naruto\), 1girl, solo, looking at viewer, simple background, white background, pleated skirt ,necktie, white shirt, school uniform, school bag, classroom, desk, whiteboard