org.deeplearning4j.nn.conf

Class MultiLayerConfiguration.Builder

java.lang.Object

org.deeplearning4j.nn.conf.MultiLayerConfiguration.Builder

Direct Known Subclasses:

NeuralNetConfiguration.ListBuilder

Enclosing class:

MultiLayerConfiguration

public static class MultiLayerConfiguration.Builder
extends java.lang.Object

Field Summary

Fields

Modifier and Type	Field and Description
protected boolean	backprop
protected BackpropType	backpropType
protected int[]	cnnInputSize Deprecated.
protected java.util.List<NeuralNetConfiguration>	confs
protected double	dampingFactor
protected java.util.Map<java.lang.Integer,InputPreProcessor>	inputPreProcessors
protected InputType	inputType
protected boolean	pretrain
protected int	tbpttBackLength
protected int	tbpttFwdLength

Constructor Summary

Constructors

Constructor and Description
Builder()

Method Summary

Modifier and Type	Method and Description
MultiLayerConfiguration.Builder	backprop(boolean backprop) Whether to do back prop or not
MultiLayerConfiguration.Builder	backpropType(BackpropType type) The type of backprop.
MultiLayerConfiguration	build()
MultiLayerConfiguration.Builder	cnnInputSize(int[] cnnInputSize) Deprecated. use setInputType(InputType) with InputType.convolutional(height,width,depth), for CNN data with shape [minibatchSize,depth,height,width]. For image data that has been flattened into a row vector per example (shape [minibatchSize,depth*height*width]) instead use InputType.convolutionalFlat(height,width,depth)
MultiLayerConfiguration.Builder	cnnInputSize(int height, int width, int depth) Deprecated. use setInputType(InputType) with InputType.convolutional(height,width,depth), for CNN data with shape [minibatchSize,depth,height,width]. For image data that has been flattened into a row vector per example (shape [minibatchSize,depth*height*width]) instead use InputType.convolutionalFlat(height,width,depth)
MultiLayerConfiguration.Builder	confs(java.util.List<NeuralNetConfiguration> confs)
MultiLayerConfiguration.Builder	inputPreProcessor(java.lang.Integer layer, InputPreProcessor processor) Specify the processors.
MultiLayerConfiguration.Builder	inputPreProcessors(java.util.Map<java.lang.Integer,InputPreProcessor> processors)
MultiLayerConfiguration.Builder	pretrain(boolean pretrain) Whether to do pre train or not
MultiLayerConfiguration.Builder	setInputType(InputType inputType)
MultiLayerConfiguration.Builder	tBPTTBackwardLength(int backwardLength) When doing truncated BPTT: how many steps of backward should we do? Only applicable when doing backpropType(BackpropType.TruncatedBPTT) This is the k2 parameter on pg23 of http://www.cs.utoronto.ca/~ilya/pubs/ilya_sutskever_phd_thesis.pdf
MultiLayerConfiguration.Builder	tBPTTForwardLength(int forwardLength) When doing truncated BPTT: how many steps of forward pass should we do before doing (truncated) backprop? Only applicable when doing backpropType(BackpropType.TruncatedBPTT) Typically tBPTTForwardLength parameter is same as the tBPTTBackwardLength parameter, but may be larger than it in some circumstances (but never smaller) Ideally your training data time series length should be divisible by this This is the k1 parameter on pg23 of http://www.cs.utoronto.ca/~ilya/pubs/ilya_sutskever_phd_thesis.pdf

Methods inherited from class java.lang.Object

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

Field Detail

confs

protected java.util.List<NeuralNetConfiguration> confs

dampingFactor

protected double dampingFactor

inputPreProcessors

protected java.util.Map<java.lang.Integer,InputPreProcessor> inputPreProcessors

pretrain

protected boolean pretrain

backprop

protected boolean backprop

backpropType

protected BackpropType backpropType

tbpttFwdLength

protected int tbpttFwdLength

tbpttBackLength

protected int tbpttBackLength

inputType

protected InputType inputType

cnnInputSize

@Deprecated
protected int[] cnnInputSize

Deprecated.

Constructor Detail

Builder

public Builder()

Method Detail

inputPreProcessor

public MultiLayerConfiguration.Builder inputPreProcessor(java.lang.Integer layer,
                                                         InputPreProcessor processor)

Specify the processors. These are used at each layer for doing things like normalization and shaping of input.

Parameters:

processor - what to use to preProcess the data.

Returns:

builder pattern

inputPreProcessors

public MultiLayerConfiguration.Builder inputPreProcessors(java.util.Map<java.lang.Integer,InputPreProcessor> processors)

backprop

public MultiLayerConfiguration.Builder backprop(boolean backprop)

Whether to do back prop or not

Parameters:

backprop - whether to do back prop or not

Returns:

backpropType

public MultiLayerConfiguration.Builder backpropType(BackpropType type)

The type of backprop. Default setting is used for most networks (MLP, CNN etc), but optionally truncated BPTT can be used for training recurrent neural networks. If using TruncatedBPTT make sure you set both tBPTTForwardLength() and tBPTTBackwardLength()

tBPTTForwardLength

public MultiLayerConfiguration.Builder tBPTTForwardLength(int forwardLength)

When doing truncated BPTT: how many steps of forward pass should we do before doing (truncated) backprop?
Only applicable when doing backpropType(BackpropType.TruncatedBPTT)
Typically tBPTTForwardLength parameter is same as the tBPTTBackwardLength parameter, but may be larger than it in some circumstances (but never smaller)
Ideally your training data time series length should be divisible by this This is the k1 parameter on pg23 of http://www.cs.utoronto.ca/~ilya/pubs/ilya_sutskever_phd_thesis.pdf

Parameters:

forwardLength - Forward length > 0, >= backwardLength

tBPTTBackwardLength

public MultiLayerConfiguration.Builder tBPTTBackwardLength(int backwardLength)

When doing truncated BPTT: how many steps of backward should we do?
Only applicable when doing backpropType(BackpropType.TruncatedBPTT)
This is the k2 parameter on pg23 of http://www.cs.utoronto.ca/~ilya/pubs/ilya_sutskever_phd_thesis.pdf

Parameters:

backwardLength - <= forwardLength

pretrain

public MultiLayerConfiguration.Builder pretrain(boolean pretrain)

Whether to do pre train or not

Parameters:

pretrain - whether to do pre train or not

Returns:

builder pattern

confs

public MultiLayerConfiguration.Builder confs(java.util.List<NeuralNetConfiguration> confs)

cnnInputSize

@Deprecated
public MultiLayerConfiguration.Builder cnnInputSize(int height,
                                                                int width,
                                                                int depth)

Deprecated. use setInputType(InputType) with InputType.convolutional(height,width,depth), for CNN data with shape [minibatchSize,depth,height,width]. For image data that has been flattened into a row vector per example (shape [minibatchSize,depth*height*width]) instead use InputType.convolutionalFlat(height,width,depth)

Size of input for CNNs. Should only be used when convolutional layers are present. This information (input size) is necessary in order to:
(a) automatically add layer preprocessors, which allow CNN and dense/output layers to be used together (as well as CNN/RNN layers)
(b) automatically calculate input/output sizes; for example, input size for a dense layer, in a Convolutional->DenseLayer or Convolutional->OutputLayer configuratio

Parameters:

height - Input image height

width - Input image width

depth - Input image depth / number of channels (for example: 3 for color, 1 for grayscale etc)

cnnInputSize

@Deprecated
public MultiLayerConfiguration.Builder cnnInputSize(int[] cnnInputSize)

Deprecated. use setInputType(InputType) with InputType.convolutional(height,width,depth), for CNN data with shape [minibatchSize,depth,height,width]. For image data that has been flattened into a row vector per example (shape [minibatchSize,depth*height*width]) instead use InputType.convolutionalFlat(height,width,depth)

CNN input size, in order of {height,width,depth}.

See Also:

cnnInputSize(int, int, int)

setInputType

public MultiLayerConfiguration.Builder setInputType(InputType inputType)

build

public MultiLayerConfiguration build()