no code implementations • 31 May 2024 • Wen-Pu Cai, Wu-Jun Li
Weight quantization has been widely used for model compression, which can reduce both storage and computational cost.
no code implementations • 1 Jul 2019 • Wen-Pu Cai, Wu-Jun Li
WNQ adopts weight normalization to avoid the long-tail distribution of network weights and subsequently reduces the quantization error.