Can’t forget time it takes to run inference, even on the latest A100/H100. Generating in under e.g. ten seconds enables more use cases (and so on until high fps

Can’t forget time it takes to run inference, even on the latest A100/H100. Generating in under e.g. ten seconds enables more use cases (and so on until high fps video is possible).