1 paper
Junhan Zhu, Hesong Wang, Mingluo Su +2
Large-scale text-to-image diffusion models, while powerful, suffer from prohibitive computational cost. Existing one-shot network pruning methods can hardly be directly applied to…