An idea for a new feature: add Hunyuan Video Avatar support

[easydream1984]

Tencent has released a new model for talking heads that can immediately convert audio and image into video, but problem number 1 is that the video takes no longer than 14 seconds, probably 96 gigs to run (I know about WanGP - 14 seconds is still not enough for a talking head, it takes at least 30 seconds) and framepack can be a breakthrough in Creating avatars and talking heads