research
∙
03/22/2021
n-hot: Efficient bit-level sparsity for powers-of-two neural network quantization
Powers-of-two (PoT) quantization reduces the number of bit operations of...
share
research
∙
11/13/2020
