DELT: A Simple Diversity-driven EarlyLate Training for Dataset Distillation

Open in new window