Open 5thGenDev opened 1 year ago
Has anyone tried this approach? Does it have any advantage in classical image generation quality, few-shot image generation?
I haven't tried it, but I think this probably won't produce better quality than the default method.
fair enough, this will be the topic for my thesis then :P
Good luck and hope you can find a way to make it work!
Let's say an Image Generation model is meant to generate image from random noise in 50 steps:
Has anyone tried this approach? Does it have any advantage in classical image generation quality, few-shot image generation?