Yixin Liu, Yong Guo, Zichang Liu +5
Model compression aims to reduce the redundancy of deep networks to obtain compact models. Recently, channel pruning has become one of the predominant compression methods to deploy…