LinMorph: Learning to Morph Weights for Efficient Video DiT Linearization
Post-hoc linearization is an attractive way to accelerate pretrained Video Diffusion Transformers (Video DiTs), yet existing methods largely treat the quality loss induced by attention replacement as a recovery problem. We show that a more fundamental bottleneck lies upstream: \emph{initialization}. Random initializati...
Ling-Yi Long, Cong-Yue Zhao, Xia-Oou Li et al.
· 0 citations