US 11,657,289 B2
Computational graph optimization
Yanqi Zhou, Sunnyvale, CA (US); Sudip Roy, San Jose, CA (US); Amirali Abdolrashidi, Riverside, CA (US); Daniel Lin-Kit Wong, Pittsburgh, PA (US); Chao Ma, Mountain View, CA (US); Qiumin Xu, Santa Clara, CA (US); and Azalia Mirhoseini, San Jose, CA (US)
Assigned to Google LLC, Mountain View, CA (US)
Filed by Google LLC, Mountain View, CA (US)
Filed on Apr. 3, 2020, as Appl. No. 16/840,191.
Claims priority of provisional application 62/971,891, filed on Feb. 7, 2020.
Prior Publication US 2021/0248445 A1, Aug. 12, 2021
Int. Cl. G06N 3/04 (2023.01); G06K 9/62 (2022.01); G06N 3/049 (2023.01)
CPC G06N 3/0454 (2013.01) [G06K 9/6231 (2013.01); G06K 9/6262 (2013.01); G06K 9/6296 (2013.01); G06N 3/049 (2013.01)] 24 Claims
OG exemplary drawing
 
1. A method of generating a task output for executing a plurality of operations of a neural network on one or more processing devices, wherein the task output comprises, for each of the plurality of operations of the neural network, a respective decision for a particular optimization task, the method comprising:
obtaining data representing a graph characterizing the plurality of operations of the neural network, wherein each node of the graph characterizes an operation of the neural network and each edge of the graph characterizes data dependency between the operations;
processing the data representing the graph using a graph embedding neural network to generate an embedding of the graph; and
processing the embedding of the graph using a policy neural network to generate the task output, wherein the policy neural network is conditioned on features of the graph.