org.deeplearning4j.nn.conf.layers

Class LocalResponseNormalization.Builder

java.lang.Object

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

org.deeplearning4j.nn.conf.layers.LocalResponseNormalization.Builder

Enclosing class:

LocalResponseNormalization

public static class LocalResponseNormalization.Builder
extends Layer.Builder<LocalResponseNormalization.Builder>

Field Summary

Fields

Modifier and Type	Field and Description
protected boolean	cudnnAllowFallback When using CuDNN and an error is encountered, should fallback to the non-CuDNN implementatation be allowed? If set to false, an exception in CuDNN will be propagated back to the user.
protected CNN2DFormat	dataFormat

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

allParamConstraints, biasConstraints, iDropout, layerName, weightConstraints

Constructor Summary

Constructors

Constructor and Description
Builder()
Builder(double k, double alpha, double beta)
Builder(double k, double n, double alpha, double beta)

Method Summary

Modifier and Type	Method and Description
LocalResponseNormalization.Builder	alpha(double alpha) LRN scaling constant alpha.
LocalResponseNormalization.Builder	beta(double beta) Scaling constant beta.
LocalResponseNormalization	build()
LocalResponseNormalization.Builder	cudnnAllowFallback(boolean allowFallback) Deprecated. Use helperAllowFallback(boolean)
LocalResponseNormalization.Builder	dataFormat(CNN2DFormat dataFormat) Set the data format for the CNN activations - NCHW (channels first) or NHWC (channels last).
LocalResponseNormalization.Builder	helperAllowFallback(boolean allowFallback) When using CuDNN or MKLDNN and an error is encountered, should fallback to the non-helper implementation be allowed? If set to false, an exception in the helper will be propagated back to the user.
LocalResponseNormalization.Builder	k(double k) LRN scaling constant k.
LocalResponseNormalization.Builder	n(double n) Number of adjacent kernel maps to use when doing LRN.

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

constrainAllParameters, constrainBias, constrainWeights, dropOut, dropOut, name

Methods inherited from class java.lang.Object

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

Field Detail

cudnnAllowFallback

protected boolean cudnnAllowFallback

When using CuDNN and an error is encountered, should fallback to the non-CuDNN implementatation be allowed? If set to false, an exception in CuDNN will be propagated back to the user. If false, the built-in (non-CuDNN) implementation for BatchNormalization will be used

dataFormat

protected CNN2DFormat dataFormat

Constructor Detail

Builder

public Builder(double k,
               double n,
               double alpha,
               double beta)

Builder

public Builder(double k,
               double alpha,
               double beta)

Builder

public Builder()

Method Detail

k

public LocalResponseNormalization.Builder k(double k)

LRN scaling constant k. Default: 2

Parameters:

k - Scaling constant

n

public LocalResponseNormalization.Builder n(double n)

Number of adjacent kernel maps to use when doing LRN. default: 5

Parameters:

n - Number of adjacent kernel maps

alpha

public LocalResponseNormalization.Builder alpha(double alpha)

LRN scaling constant alpha. Default: 1e-4

Parameters:

alpha - Scaling constant

beta

public LocalResponseNormalization.Builder beta(double beta)

Scaling constant beta. Default: 0.75

Parameters:

beta - Scaling constant

cudnnAllowFallback

@Deprecated
public LocalResponseNormalization.Builder cudnnAllowFallback(boolean allowFallback)

Deprecated. Use helperAllowFallback(boolean)

When using CuDNN and an error is encountered, should fallback to the non-CuDNN implementatation be allowed? If set to false, an exception in CuDNN will be propagated back to the user. If true, the built-in (non-CuDNN) implementation for BatchNormalization will be used

Parameters:

allowFallback - Whether fallback to non-CuDNN implementation should be used

helperAllowFallback

public LocalResponseNormalization.Builder helperAllowFallback(boolean allowFallback)

When using CuDNN or MKLDNN and an error is encountered, should fallback to the non-helper implementation be allowed? If set to false, an exception in the helper will be propagated back to the user. If true, the built-in (non-MKL/CuDNN) implementation for LocalResponseNormalizationLayer will be used

Parameters:

allowFallback - Whether fallback to non-CuDNN implementation should be used

dataFormat

public LocalResponseNormalization.Builder dataFormat(CNN2DFormat dataFormat)

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)

build

public LocalResponseNormalization build()

Specified by:

build in class Layer.Builder<LocalResponseNormalization.Builder>