Train LoRa is time and resource expensive, you can support my future works here:
❤️Boosty Donation link
it would be great if you could show some results of your work as a post/review
Ore no Imouto ga Konnani Kawaii Wake ga Nai
(俺の妹がこんなに可愛いわけがない)
list will be updated later.
Why my LoRa's? Because it's close to anime style, and this style is ~simple in details and can change without mutating target style more easy than artistic one. And it's hard to get quality anime style pictures from LoRa's with more artistic dataset with gradients and complex lighting.
Description
FAQ
Comments (2)
Does this LoRA only generate anime screenshots and such with the provided prompts? I tried to get, say, Manami on the beach and it just gave me what seemed to be an anime screen grab featuring her.
I thought this was a character LoRA, but it seems to be just trained on screenshots and whatnot.
this LoRA, and my others generate as close to screenshot style as can(because... i love anime, not arts from artstation, and it hard to get anime style from "art" LoRAs, harder than get art style from anime one(model like dreemshaper+add detail LoRA+ weight 0.5-0.6 on my LoRA == art style)),
but there no one swimsuit/sea/beach images in train data and i think you get beach with this prompt.
LoRA generate original images, not screenshots from train data.
P.S. Manami is little exceptional LoRA, it's a little stiff, it has a loss of 0.0254, which is very low, I was wondering how well it could convey the details of the original style if you lower the loss to the maximum, and it did, but my other models are made for the same purpose, they are just softer (there loss = 0.07-0.08)
P.P.S i add few images of Manami to different image post, what wrong with it?
Details
Files
Available On (3 platforms)
Same model published on other platforms. May have additional downloads or version variants.