no code implementations • 3 Sep 2022 • Duc-Quang Vu, Trang Phung, Jia-Ching Wang
Self-knowledge distillation, the special case of knowledge distillation, has been proposed to remove the large teacher network training process while preserving the student's performance.