video generation

[weimengting]

Hi authors,

Thanks for the brilliant wotk! Do you have any plans on publishing video generation sana models?

[lawrence-cj]

Yah, working on it now. It's a pretty tough and interesting work.

[EladWarshawsky]

Open-Sora already uses a deep compression autoencoder for the backbone of it, enabling the training costs to be pretty low. I'd be interested in hearing what you have in the works

[bhack]

https://github.com/NVlabs/Sana/blob/main/asset/docs/sana_video.md