What are the recommended hardware specifications for inference?

[Hangsiin]

How much VRAM need for inference?
And can you recommend minimum specific GPU for generating videos?
Thanks for open-sourcing this!

[zeinsh]

This depends on the target resolution and fps.

For example, if you want to generate 512x512 resolution at high fps, you will need at least 55-60G VRAM.
However, if you don't have a single device with enough VRAM, you can modify the code to move each model in the pipeline to a separate device.