This model doesn't have a readme.

cjwbw
/
dreamtalk
RESEARCH/NON-COMMERCIAL USE ONLY: diffusion-based audio-driven expressive talking head generation