Systems and methods for training matrix-based differentiable programs

Methods and apparatus for training a matrix-based differentiable program using a photonics-based processor. The matrix-based differentiable program includes at least one matrix-valued variable associated with a matrix of values in a Euclidean vector space. The method comprises configuring components...

Full description

Saved in:
Bibliographic Details
Main Authors Harris, Nicholas C, Bunandar, Darius, Forsythe, Martin, Lazovich, Tomo
Format Patent
LanguageEnglish
Published 11.08.2020
Subjects
CALCULATING
COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
COMPUTING
COUNTING
ELECTRIC DIGITAL DATA PROCESSING
OPTICAL COMPUTING DEVICES
PHYSICS
Online AccessGet full text

Cover

More Information
Summary:Methods and apparatus for training a matrix-based differentiable program using a photonics-based processor. The matrix-based differentiable program includes at least one matrix-valued variable associated with a matrix of values in a Euclidean vector space. The method comprises configuring components of the photonics-based processor to represent the matrix of values as an angular representation, processing, using the components of the photonics-based processor, training data to compute an error vector, determining in parallel, at least some gradients of parameters of the angular representation, wherein the determining is based on the error vector and a current input training vector, and updating the matrix of values by updating the angular representation based on the determined gradients.
Bibliography:Application Number: US201916412159