identified that the performance gap between teacher and student is not due to model capacity but the distillation data; proposed a novel data augmentation tech (MixPatch) Knowledge distillation tasks;