hamburbur

I love the concept but I think it could use another pass. It frequently spawns in huge knockers and in fact the most success I got with this lora also included the character growing massive boobs even though I was negative prompting for "large breasts, huge tits, oppai" and had "flat chest" in my prompt just to try and counter-act it.

Another thing I noticed is that the character would often gain long nails.

When you're writing your lora captions, you will have better success if you describe everything that is NOT your trigger word. The trigger word is there to absorb the concept, and everything else in your caption is there to tell the AI what is NOT your concept.

So your captions shouldn't look like

"Two-handed burger a close-up of a girl holding a burger with both hands"

but I suspect strongly that it does.

Instead, it should be something like,

"A close-up of a girl with large breasts and long nails is sitting in a restaurant in front of a window eating a two-handed burger".

I hope you understand the difference. By using the caption to describe the burger and not the girl, you're actually teaching the AI that "two-handed burger" is related to the way the girl looks. If you use a prompt similar to my second example, the AI will absorb the hand position, the size of the burger, and even the expression on the girl's face as part of the concept "two-handed burger".

I hope you put this back through the process and try again!