org.deeplearning4j.nn.layers.samediff

Class SameDiffLayer

java.lang.Object

org.deeplearning4j.nn.layers.AbstractLayer<AbstractSameDiffLayer>

org.deeplearning4j.nn.layers.samediff.SameDiffLayer

All Implemented Interfaces:

Serializable, Cloneable, Layer, Model, Trainable

public class SameDiffLayer
extends AbstractLayer<AbstractSameDiffLayer>

See Also:

Serialized Form

Nested Class Summary

Nested classes/interfaces inherited from interface org.deeplearning4j.nn.api.Layer

Layer.TrainingMode, Layer.Type

Field Summary

Fields

Modifier and Type	Field and Description
protected ExternalErrorsFunction	fn
protected INDArray	gradients
protected Map<String,INDArray>	gradTable
static String	INPUT_KEY
static String	MASK_KEY
protected String	outputKey
protected SDVariable	outputVar
protected INDArray	params
protected Map<String,INDArray>	paramTable
protected SameDiff	sameDiff

Fields inherited from class org.deeplearning4j.nn.layers.AbstractLayer

cacheMode, conf, dataType, dropoutApplied, epochCount, index, input, inputModificationAllowed, iterationCount, maskArray, maskState, preOutput, trainingListeners

Constructor Summary

Constructors

Constructor and Description
SameDiffLayer(NeuralNetConfiguration conf, org.nd4j.linalg.api.buffer.DataType dataType)

Method Summary

Modifier and Type	Method and Description
INDArray	activate(boolean training, LayerWorkspaceMgr workspaceMgr) Perform forward pass and return the activations array with the last set input
Pair<Gradient,INDArray>	backpropGradient(INDArray epsilon, LayerWorkspaceMgr workspaceMgr) Calculate the gradient relative to the error in the next layer
void	clearNoiseWeightParams()
Layer	clone()
protected void	doInit()
Pair<INDArray,MaskState>	feedForwardMaskArray(INDArray maskArray, MaskState currentMaskState, int minibatchSize) Feed forward the input mask array, setting in the layer as appropriate.
INDArray	getGradientsViewArray()
INDArray	getParam(String param) Get the parameter
boolean	isPretrainLayer() Returns true if the layer can be trained in an unsupervised/pretrain manner (AE, VAE, etc)
long	numParams() The number of parameters for the model
INDArray	params() Returns the parameters of the neural network as a flattened row vector
Map<String,INDArray>	paramTable() The param table
Map<String,INDArray>	paramTable(boolean backpropParamsOnly) Table of parameters by key, for backprop For many models (dense layers, etc) - all parameters are backprop parameters
void	setBackpropGradientsViewArray(INDArray gradients) Set the gradients array as a view of the full (backprop) network parameters NOTE: this is intended to be used internally in MultiLayerNetwork and ComputationGraph, not by users.
void	setParam(String key, INDArray val) Set the parameter with a new ndarray
void	setParams(INDArray params) Set the parameters for this model.
protected void	setParams(INDArray params, char order)
void	setParamsViewArray(INDArray params) Set the initial parameters array as a view of the full (backprop) network parameters NOTE: this is intended to be used internally in MultiLayerNetwork and ComputationGraph, not by users.
void	setParamTable(Map<String,INDArray> paramTable) Setter for the param table

Methods inherited from class org.deeplearning4j.nn.layers.AbstractLayer

activate, addListeners, allowInputModification, applyConstraints, applyDropOutIfNecessary, applyMask, assertInputSet, backpropDropOutIfPresent, batchSize, calcRegularizationScore, clear, computeGradientAndScore, conf, fit, fit, getConfig, getEpochCount, getHelper, getIndex, getInput, getInputMiniBatchSize, getListeners, getMaskArray, getOptimizer, gradient, gradientAndScore, init, input, layerConf, layerId, numParams, score, setCacheMode, setConf, setEpochCount, setIndex, setInput, setInputMiniBatchSize, setListeners, setListeners, setMaskArray, type, update, update, updaterDivideByMinibatch

Methods inherited from class java.lang.Object

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

Methods inherited from interface org.deeplearning4j.nn.api.Layer

getIterationCount, setIterationCount

Field Detail

INPUT_KEY

public static final String INPUT_KEY

See Also:

Constant Field Values

MASK_KEY

public static final String MASK_KEY

See Also:

Constant Field Values

sameDiff

protected SameDiff sameDiff

outputVar

protected SDVariable outputVar

fn

protected ExternalErrorsFunction fn

outputKey

protected String outputKey

params

protected INDArray params

gradients

protected INDArray gradients

paramTable

protected Map<String,INDArray> paramTable

gradTable

protected Map<String,INDArray> gradTable

Constructor Detail

SameDiffLayer

public SameDiffLayer(NeuralNetConfiguration conf,
                     org.nd4j.linalg.api.buffer.DataType dataType)

Method Detail

clone

public Layer clone()

Overrides:

clone in class Object

isPretrainLayer

public boolean isPretrainLayer()

Description copied from interface: Layer

Returns true if the layer can be trained in an unsupervised/pretrain manner (AE, VAE, etc)

Returns:

true if the layer can be pretrained (using fit(INDArray), false otherwise

clearNoiseWeightParams

public void clearNoiseWeightParams()

activate

public INDArray activate(boolean training,
                         LayerWorkspaceMgr workspaceMgr)

Description copied from interface: Layer

Perform forward pass and return the activations array with the last set input

Parameters:

training - training or test mode

workspaceMgr - Workspace manager

Returns:

the activation (layer output) of the last specified input. Note that the returned array should be placed in the ArrayType.ACTIVATIONS workspace via the workspace manager

backpropGradient

public Pair<Gradient,INDArray> backpropGradient(INDArray epsilon,
                                                LayerWorkspaceMgr workspaceMgr)

Description copied from interface: Layer

Calculate the gradient relative to the error in the next layer

Parameters:

epsilon - w^(L+1)*delta^(L+1). Or, equiv: dC/da, i.e., (dC/dz)*(dz/da) = dC/da, where C is cost function a=sigma(z) is activation.

workspaceMgr - Workspace manager

Returns:

Pair where Gradient is gradient for this layer, INDArray is epsilon (activation gradient) needed by next layer, but before element-wise multiply by sigmaPrime(z). So for standard feed-forward layer, if this layer is L, then return.getSecond() == dL/dIn = (w^(L)*(delta^(L))^T)^T. Note that the returned array should be placed in the ArrayType.ACTIVATION_GRAD workspace via the workspace manager

params

public INDArray params()

Returns the parameters of the neural network as a flattened row vector

Specified by:

params in interface Model

Specified by:

params in interface Trainable

Overrides:

params in class AbstractLayer<AbstractSameDiffLayer>

Returns:

the parameters of the neural network

getParam

public INDArray getParam(String param)

Description copied from interface: Model

Get the parameter

Specified by:

getParam in interface Model

Overrides:

getParam in class AbstractLayer<AbstractSameDiffLayer>

Parameters:

param - the key of the parameter

Returns:

the parameter vector/matrix with that particular key

numParams

public long numParams()

Description copied from class: AbstractLayer

The number of parameters for the model

Specified by:

numParams in interface Model

Specified by:

numParams in interface Trainable

Overrides:

numParams in class AbstractLayer<AbstractSameDiffLayer>

Returns:

the number of parameters for the model

setParam

public void setParam(String key,
                     INDArray val)

Description copied from interface: Model

Set the parameter with a new ndarray

Specified by:

setParam in interface Model

Overrides:

setParam in class AbstractLayer<AbstractSameDiffLayer>

Parameters:

key - the key to se t

val - the new ndarray

setParams

public void setParams(INDArray params)

Description copied from interface: Model

Set the parameters for this model. This expects a linear ndarray which then be unpacked internally relative to the expected ordering of the model

Specified by:

setParams in interface Model

Overrides:

setParams in class AbstractLayer<AbstractSameDiffLayer>

Parameters:

params - the parameters for the model

setParams

protected void setParams(INDArray params,
                         char order)

Overrides:

setParams in class AbstractLayer<AbstractSameDiffLayer>

setParamsViewArray

public void setParamsViewArray(INDArray params)

Description copied from interface: Model

Set the initial parameters array as a view of the full (backprop) network parameters NOTE: this is intended to be used internally in MultiLayerNetwork and ComputationGraph, not by users.

Specified by:

setParamsViewArray in interface Model

Overrides:

setParamsViewArray in class AbstractLayer<AbstractSameDiffLayer>

Parameters:

params - a 1 x nParams row vector that is a view of the larger (MLN/CG) parameters array

getGradientsViewArray

public INDArray getGradientsViewArray()

Specified by:

getGradientsViewArray in interface Model

Specified by:

getGradientsViewArray in interface Trainable

Overrides:

getGradientsViewArray in class AbstractLayer<AbstractSameDiffLayer>

Returns:

1D gradients view array

setBackpropGradientsViewArray

public void setBackpropGradientsViewArray(INDArray gradients)

Description copied from interface: Model

Set the gradients array as a view of the full (backprop) network parameters NOTE: this is intended to be used internally in MultiLayerNetwork and ComputationGraph, not by users.

Specified by:

setBackpropGradientsViewArray in interface Model

Overrides:

setBackpropGradientsViewArray in class AbstractLayer<AbstractSameDiffLayer>

Parameters:

gradients - a 1 x nParams row vector that is a view of the larger (MLN/CG) gradients array

setParamTable

public void setParamTable(Map<String,INDArray> paramTable)

Description copied from interface: Model

Setter for the param table

Specified by:

setParamTable in interface Model

Overrides:

setParamTable in class AbstractLayer<AbstractSameDiffLayer>

paramTable

public Map<String,INDArray> paramTable()

Description copied from interface: Model

The param table

Specified by:

paramTable in interface Model

Overrides:

paramTable in class AbstractLayer<AbstractSameDiffLayer>

Returns:

paramTable

public Map<String,INDArray> paramTable(boolean backpropParamsOnly)

Description copied from interface: Model

Table of parameters by key, for backprop For many models (dense layers, etc) - all parameters are backprop parameters

Specified by:

paramTable in interface Model

Specified by:

paramTable in interface Trainable

Overrides:

paramTable in class AbstractLayer<AbstractSameDiffLayer>

Parameters:

backpropParamsOnly - If true, return backprop params only. If false: return all params (equivalent to paramsTable())

Returns:

Parameter table

doInit

protected void doInit()

feedForwardMaskArray

public Pair<INDArray,MaskState> feedForwardMaskArray(INDArray maskArray,
                                                     MaskState currentMaskState,
                                                     int minibatchSize)

Description copied from interface: Layer

Feed forward the input mask array, setting in the layer as appropriate. This allows different layers to handle masks differently - for example, bidirectional RNNs and normal RNNs operate differently with masks (the former sets activations to 0 outside of the data present region (and keeps the mask active for future layers like dense layers), whereas normal RNNs don't zero out the activations/errors )instead relying on backpropagated error arrays to handle the variable length case.
This is also used for example for networks that contain global pooling layers, arbitrary preprocessors, etc.

Specified by:

feedForwardMaskArray in interface Layer

Overrides:

feedForwardMaskArray in class AbstractLayer<AbstractSameDiffLayer>

Parameters:

maskArray - Mask array to set

currentMaskState - Current state of the mask - see MaskState

minibatchSize - Current minibatch size. Needs to be known as it cannot always be inferred from the activations array due to reshaping (such as a DenseLayer within a recurrent neural network)

Returns:

New mask array after this layer, along with the new mask state.