how to make generate it faster?

[margo2130]

hi, i make that sample image and auidio (7second) and it took 15 minutes (4070 super 12gb).
i want to know is there ways to make it faster?
for example make the moves in eyes or other parts of face less so it generate faster?
is there any setting for this?

also if anyone know any model for comfy ui or other ai local software that sync lips and a little eyes tell me that model.
i am searching for a model i give image + audio file and it give me that ai avatar speaker.
i want a faster model.

[Mayaworldai]

You can try sad talker. although this seems to be better and more accurate