HunyuanDiT
English
Chinese

generated images have no difference between EMA and Distilled version

#1
by gfork - opened

I've tried EMA and Distilled model inference with 100 and 50 steps, there seems no difference between them, how the 2X acceleration comes for Distilled model?

prompt:
A young woman stands on a railway station platform in spring. She was wearing a long blue-gray trench coat and a white shirt. Her dark hair was tied into a low ponytail, with a few strands flying in the wind. Her eyes were full of expectation, and the sun shone on her warm face.

100 steps
EMA
100_step.png

Distill
100_step.png

50 steps
EMA
50_step.png

Distill
50_step.png

Your need to confirm your account before you can post a new comment.

Sign up or log in to comment