On the right are the generated videos and latent previews. Because this is a next-frame-section prediction model, videos will be generated longer and longer. You will see the progress bar for each ...