completely retrained with a new dataset (lost the old one)
This dataset is now uploaded here (in case i lose it again lol)
Allow for more variety in clothing, hair color etc, like in the second example picture
changed triggerwords
default artstyle is not very interesting. For best results, combine with your favorite artstyle LORA
if including "3d" in your tag, it tends to stick her on some kind of platform.
It also tens to deform her face a bit. Try putting (deformed face), (deformed eyes) in the neg prompt
Recommended weight is 0.9-1
completely retrained with a new dataset (lost the old one)
This dataset is now uploaded here (in case i lose it again lol)
Allow for more variety in clothing, hair color etc, like in the second example picture
changed triggerwords
default artstyle is not very interesting. For best results, combine with your favorite artstyle LORA
if including "3d" in your tag, it tends to stick her on some kind of platform.
It also tens to deform her face a bit. Try putting (deformed face), (deformed eyes) in the neg prompt
Recommended weight is 0.9-1