Read the original at HF Daily Papers
Researchers introduce SGF+, a method that separates parameters for context writing and denoising in autoregressive video generation. This design enables continuous video generation for up to 24 hours without additional training data.
Carried by: HF Daily Papers. First seen: .