sgdwm.opt: Create SGDWM optmiser

In modello: Homemade Deep Learning Library

Description

Creates a stochastic gradient descent with momentum optimiser.

Usage

sgdwm.opt(pars)

Arguments

pars	list of reference object of class 'number' representing the parameters to optmise.

Value

a reference object of class 'sgdwm.opt'

Examples

modello.init(10, 10, 10, 10)
mdl = module.lm$new(1, 3, 1)
optimizer = sgdwm.opt(mdl$pars())
is.opt(optimizer) # TRUE
modello.close()

modello documentation built on Feb. 2, 2021, 9:06 a.m.