PROBABLY USE MY EMBEDDING INSTEAD, I THINK IT'S BETTER
differences I think are this one has a way larger effect, but kinda messes with stuff so idk maybe try both but who knows
this is a lora trained on some images from fallen angels, I don't really think it's as good as the embedding I made honestly. used the civitai training thing to do it, idk about that but whatever.
I'm pretty sure you gotta type in fallenstyle3 to the prompt also to get the style but maybe not idk. I also used the words: lush, bloom, neon, charlie, blondie, and motion blur for a lot of the data caption stuff so they will make differences to the images. idk. I think I left too many photos of blondie in the training data so she kinda comes up a lot in the generated images. anyways, have fun.