Request support for video models: SkyReels-V3

[juntaosun]

SkyReels V3 natively supports three core generative capabilities:

1. multi-subject video generation from reference images,
2. video generation guided by audio,
3. video-to-video generation.

Video Extension：
https://huggingface.co/Skywork/SkyReels-V3-V2V-14B

Reference to Video：
https://huggingface.co/Skywork/SkyReels-V3-R2V-14B

Talking Avatar(Single Actor)：
https://huggingface.co/Skywork/SkyReels-V3-A2V-19B

Performance Comparison：

R2V-ComfyUI Workflow :
SkyReels-V3-R2V-多图参考.json