Package io.github.kirstenali.deepj.tensor

Class NormalizationChecks

java.lang.Object

io.github.kirstenali.deepj.tensor.NormalizationChecks

public final class NormalizationChecks
extends Object

Final-axis normalization accepts vector gains and the historical 1-by-width parameter layout.

Method Summary

Modifier and Type	Method	Description
static void	backward(Tensor gradient, Tensor normalized, Tensor denominator, Tensor gamma)
static void	forward(Tensor input, Tensor gamma, Tensor beta, float epsilon, boolean centered)
static void	layerStatistics(Tensor input, Tensor std)
static DType	outputType(Tensor input, Tensor gamma, Tensor beta, boolean centered)
static void	parameter(Tensor parameter, int width)
static void	statistics(Tensor input, Tensor denominator)
static int[]	statisticsShape(Tensor input)
static int	width(Tensor input)

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Method Details

width

public static int width(Tensor input)

parameter

public static void parameter(Tensor parameter,
 int width)

statisticsShape

public static int[] statisticsShape(Tensor input)

outputType

public static DType outputType(Tensor input,
 Tensor gamma,
 Tensor beta,
 boolean centered)

forward

public static void forward(Tensor input,
 Tensor gamma,
 Tensor beta,
 float epsilon,
 boolean centered)

backward

public static void backward(Tensor gradient,
 Tensor normalized,
 Tensor denominator,
 Tensor gamma)

statistics

public static void statistics(Tensor input,
 Tensor denominator)

layerStatistics

public static void layerStatistics(Tensor input,
 Tensor std)