ive been making a couple of datasets as of late but i havent had the motivation to train them after recent changes but i might as well upload this lol
usage tips:
either include source filmmaker in the positive or negative prompt (or the model will choose at random)
include solo in the positives and 2boys in the negatives if you only want 1 person in the image (this lora is heavily biased towards generating two males)
generate at higher resolutions for better results (1536x1536)
if the original artist isnt happy with this please make a content removal request !!