LoRA is generated entirely from anime images.
Should work well with any model.
Check samples for prompts :)
Leave a like, comment, review but first of all, ENJOY ;)
If you wish to support me in my work,
you can do that here: Ko-Fi.com/LittleJelly
Description
FAQ
Comments (7)
can you do lisca elfrieden from How a Realist hero rebuilt the kingdom?
ah, sadly i dont have access to my desktop for a few weeks... ill look into it once i gain access again
@LittleJelly Thats okay! Ty
It works great along with this one: https://civitai.com/models/50077 (imprecise arts of the anime characters - use x0.6 modifier). x0.4 for this to fix face. "shinomiya kaguya" tag is necessary. Discribe her slightly if there is some other "kaguya" characters in art. It's difficult to create something original with only this loras (they are slightly overtrained). So you should to use extra lora for new content. Also it's very difficult to add something on face, but it's possible, anyway.
Not overtrained. The point of my LoRAs is to be Anime accurate at around 0.7-0.9 weight. Not really to create something entirely new. If you want to do that just lower the Lora weight.
@LittleJelly I wouldn't upload this lora if I didn't need her exact appearance. Lowering the weights just doesn't make sense in this case, as the faces will be different. Controlnet can help me in "critical" cases, maybe. Anyway, your lora is very helpful. Those who are trained on fanart have more outfits, poses, etc., but the character is not the same at all. By mixing something universal + your lora I can get something more or less versatile and easy to use without extra utilities.
i like this model.