This LoRA creates the effect of a gummy smile, i.e. a grin or open smile where the upper gums show noticeably, possibly even more than the teeth themselves.
It was trained on close-up images. It therefore works best for portraits and close-ups.
You may layer an additional LoRA for skin texture as usual.
V1 creates teeth with braces by default. So it's best to prompt for braces, too.
V2 is more subtle and less face altering. No more bias towards braces. Note the new trigger word!
Typical prompt:
V1: girl with blue eyes, she grins, her parted lips reveal her teeth, gummy smile, upper lip is pulled up to reveal her tall gumline, teeth with braces
V2: realistic portrait photograph of a blonde girl, (upper gums):2, she has an uneven amateur face with exposed upper gums.
Description
V2 is more subtle than V1 with less face alteration. Removed the bias for generating braces. Note the new trigger word (include in your prompt at elevated strength).