US 12,437,200 B2
Accelerating neural networks with one shot skip layer pruning
Arun Coimbatore Ramachandran, Bangalore (IN); Chandra Kumar Ramasamy, Bangalore (IN); Prakash Sathyanath Raghavendra, Bangalore (IN); and Keerthan Subraya Shagrithaya, Bengaluru (IN)
Assigned to Advanced Micro Devices, Inc., Santa Clara, CA (US)
Filed by Advanced Micro Devices, Inc., Santa Clara, CA (US)
Filed on Jun. 30, 2023, as Appl. No. 18/345,445.
Application 18/345,445 is a continuation of application No. 16/457,550, filed on Jun. 28, 2019, granted, now 11,694,081.
Claims priority of provisional application 62/848,259, filed on May 15, 2019.
Prior Publication US 2023/0351187 A1, Nov. 2, 2023
This patent is subject to a terminal disclaimer.
Int. Cl. G06N 3/082 (2023.01); G06F 9/345 (2018.01); G06F 18/214 (2023.01); G06N 3/04 (2023.01)
CPC G06N 3/082 (2013.01) [G06F 9/3455 (2013.01); G06F 18/214 (2023.01); G06N 3/0418 (2013.01)] 18 Claims
OG exemplary drawing
 
1. A method for pruning an artificial neural network (ANN) that includes a plurality of layers, the method comprising: pruning filters in each layer of a subset of layers of the ANN, the subset of layers including fewer layers than the plurality of layers of the ANN, and wherein the layers of the subset of layers are non-contiguous.