1 paper across 1 session
We propose a method to speedup video diffusion generation through efficient attention.