A model I decided to make because I feel like the models uploaded here already are trained on old photos of her, I've attempted to train this on newer photos. I'm also practicing making loras so this is by no means supposed to be a professional upload or an upload where it looks like I know what I'm doing.
Trained with the keyword "nomobil" because I would assume training on SD1.5 using "Billie Eilish" in the prompt would effect results but I haven't tested.
Trained prompt structure example:
nomobil with blonde shoulder length hair and bangs with blue eyes and brown eyebrows, serious facial expression, lip gloss, gold necklace with a charm on the middle, pearl earing, blurry black background
Post your results! As I haven't tested too much. Thanks!