Package org.deeplearning4j.nn.conf.layers

Class SpaceToBatchLayer.Builder<T extends SpaceToBatchLayer.Builder<T>>

java.lang.Object

org.deeplearning4j.nn.conf.layers.Layer.Builder<T>

org.deeplearning4j.nn.conf.layers.SpaceToBatchLayer.Builder<T>

Enclosing class:

SpaceToBatchLayer

public static class SpaceToBatchLayer.Builder<T extends SpaceToBatchLayer.Builder<T>>
extends Layer.Builder<T>

Field Summary

Fields

Modifier and Type	Field	Description
protected int[]	blocks	Block size for SpaceToBatch layer.
protected CNN2DFormat	format
protected int[][]	padding	A 2d array, with format [[padTop, padBottom], [padLeft, padRight]]

Fields inherited from class org.deeplearning4j.nn.conf.layers.Layer.Builder

allParamConstraints, biasConstraints, iDropout, layerName, weightConstraints

Constructor Summary

Constructors

Constructor	Description
Builder(int[] blocks)
Builder(int[] blocks, int[][] padding)

Method Summary

Modifier and Type	Method	Description
T	blocks(int... blocks)
SpaceToBatchLayer	build()
T	dataFormat(CNN2DFormat format)	Set the data format for the CNN activations - NCHW (channels first) or NHWC (channels last).
T	name(String layerName)	Layer name assigns layer string name.
T	padding(int[][] padding)
void	setBlocks(int... blocks)
void	setPadding(int[][] padding)

Methods inherited from class org.deeplearning4j.nn.conf.layers.Layer.Builder

constrainAllParameters, constrainBias, constrainWeights, dropOut, dropOut

Methods inherited from class java.lang.Object

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

Field Detail

blocks

protected int[] blocks

Block size for SpaceToBatch layer. Should be a length 2 array for the height and width dimensions

padding

protected int[][] padding

A 2d array, with format [[padTop, padBottom], [padLeft, padRight]]

format

protected CNN2DFormat format

Constructor Detail

Builder

public Builder(int[] blocks)

Parameters:

blocks - Block size for SpaceToBatch layer. Should be a length 2 array for the height and width dimensions

Builder

public Builder(int[] blocks,
               int[][] padding)

Parameters:

blocks - Block size for SpaceToBatch layer. Should be a length 2 array for the height and width dimensions

padding - Padding - should be a 2d array, with format [[padTop, padBottom], [padLeft, padRight]]

Method Detail

setBlocks

public void setBlocks(int... blocks)

Parameters:

blocks - Block size for SpaceToBatch layer. Should be a length 2 array for the height and width dimensions

setPadding

public void setPadding(int[][] padding)

Parameters:

padding - Padding - should be a 2d array, with format [[padTop, padBottom], [padLeft, padRight]]

dataFormat

public T dataFormat(CNN2DFormat format)

Set the data format for the CNN activations - NCHW (channels first) or NHWC (channels last). See CNN2DFormat for more details.
Default: NCHW

Parameters:

format - Format for activations (in and out)

blocks

public T blocks(int... blocks)

Parameters:

blocks - Block size for SpaceToBatch layer. Should be a length 2 array for the height and width dimensions

padding

public T padding(int[][] padding)

Parameters:

padding - Padding - should be a 2d array, with format [[padTop, padBottom], [padLeft, padRight]]

name

public T name(String layerName)

Description copied from class: Layer.Builder

Layer name assigns layer string name. Allows easier differentiation between layers.

Overrides:

name in class Layer.Builder<T extends SpaceToBatchLayer.Builder<T>>

build

public SpaceToBatchLayer build()

Specified by:

build in class Layer.Builder<T extends SpaceToBatchLayer.Builder<T>>