Bypass Back-propagation: Optimization-based Structural Pruning for Large Language Models via Policy Gradient.

Yuan Gao, Zujing Liu, Weizhong Zhang, Bo Du, Gui-Song Xia

Browse the full ACL paper archive.