Experimental LoKr based on anima preview 2.
Every major version has a different style (datasets also do not overlap)
Description
Dataset: 118 NAI images / 52 Epochs.
FAQ
Comments (7)
这个风格会炼吗,在计划中的话我就等吃了
Will this style be trained? If it's in the plan, I'll just wait :>
{{{chiaroscuro}}},[[[artist:kedama_milk]]],[[[[[artist:na_tarapisu153]]]]],{[artist:fymrie}},nalphanne,[[dino_(dinoartforame)]] [[ringouulu]],artist:ciloranko,[[[artist:quasarcake]]],artist:rella,[[[[[[ningen_mame]]]]]]]
Since I don't use NAI, you'll need to provide the dataset yourself for specific artist combinations. 🙏
@Muffalo it's just one of QAQ'style .and this is the metadata of it
@ibara0608 He uses so many different styles that I can't tell which one you mean just by artist tags. Could you send me a link?
@ibara0608 I'll give it a try when I have some time, but don't hold your breath. Which model would you like? (anima, IL, vpred)
@Muffalo anima is ok! thanks!