expected input and output?
#1
by
jqlive
- opened
I was planning to fine-tune my own SD prompt model, but saw that you did one. I originally planned it to be scene based. So for example, the user input is a scene in a story in normal spoken english. Then the model would turn that scene into a SD prompt. So I'm curious if your model can achieve something of the sort.
- What's the expected input and output?
- What version of SD prompting is it targeting? SDXL? SD 1.5?
- Was your dataset pair based?