>>56036122Anyway, like I was saying: I tried my hand at creating one of these using 22 images.
It seems to produce some decent images but I notice that it gives her a bare midriff even if I try excluding it in the negative prompt and giving the lora a lower weight. I was wondering if maybe it has something to do with how all the art I trained it on has that feature, so it has nothing to compare it against.
Anyway, I'm new to all this so I'm probably just ignorant.
https://files.catbox.moe/fb0px9.7z 