Generating a compressed representation of a neural network with proficient inference speed and power consumption
W Wang, W Jiang - US Patent App. 17/139,825, 2021 - Google Patents
The disclosure relates to technology for generating a com pressed neural network. A weight
tensor is received from a neural network to be compressed, and it is reordered to be …
tensor is received from a neural network to be compressed, and it is reordered to be …