ai.djl.nn.recurrent

Class RecurrentCell.BaseBuilder<T extends RecurrentCell.BaseBuilder>

java.lang.Object

ai.djl.nn.recurrent.RecurrentCell.BaseBuilder<T>

Direct Known Subclasses:

GRU.Builder, LSTM.Builder, RNN.Builder

Enclosing class:

RecurrentCell

public abstract static class RecurrentCell.BaseBuilder<T extends RecurrentCell.BaseBuilder>
extends java.lang.Object

The Builder to construct a RecurrentCell type of Block.

Field Summary

Fields

Modifier and Type	Field and Description
protected RNN.Activation	activation
protected boolean	clipLstmState
protected float	dropRate
protected double	lstmStateClipMax
protected double	lstmStateClipMin
protected int	numStackedLayers
protected boolean	stateOutputs
protected long	stateSize
protected boolean	useBidirectional
protected boolean	useSequenceLength

Constructor Summary

Constructors

Constructor and Description
BaseBuilder()

Method Summary

Modifier and Type	Method and Description
T	optBidrectional(boolean useBidirectional) Sets the optional parameter that indicates whether to use bidirectional recurrent layers.
T	optDropRate(float dropRate) Sets the drop rate of the dropout on the outputs of each RNN layer, except the last layer.
T	optStateOutput(boolean stateOutputs) Sets the optional parameter that indicates whether to have the states as symbol outputs.
protected abstract T	self()
T	setNumStackedLayers(int numStackedLayers) Sets the Required number of stacked layers.
T	setSequenceLength(boolean useSequenceLength) Sets the optional parameter that indicates whether to include an extra input parameter sequence_length to specify variable length sequence.
T	setStateSize(int stateSize) Sets the Required size of the state for each layer.

Methods inherited from class java.lang.Object

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

Field Detail

dropRate

protected float dropRate

stateSize

protected long stateSize

numStackedLayers

protected int numStackedLayers

lstmStateClipMin

protected double lstmStateClipMin

lstmStateClipMax

protected double lstmStateClipMax

clipLstmState

protected boolean clipLstmState

useSequenceLength

protected boolean useSequenceLength

useBidirectional

protected boolean useBidirectional

stateOutputs

protected boolean stateOutputs

activation

protected RNN.Activation activation

Constructor Detail

BaseBuilder

public BaseBuilder()

Method Detail

optDropRate

public T optDropRate(float dropRate)

Sets the drop rate of the dropout on the outputs of each RNN layer, except the last layer.

Parameters:

dropRate - the drop rate of the dropout

Returns:

this Builder

setStateSize

public T setStateSize(int stateSize)

Sets the Required size of the state for each layer.

Parameters:

stateSize - the size of the state for each layer

Returns:

this Builder

setNumStackedLayers

public T setNumStackedLayers(int numStackedLayers)

Sets the Required number of stacked layers.

Parameters:

numStackedLayers - the number of stacked layers

Returns:

this Builder

setSequenceLength

public T setSequenceLength(boolean useSequenceLength)

Sets the optional parameter that indicates whether to include an extra input parameter sequence_length to specify variable length sequence.

Parameters:

useSequenceLength - whether to use sequence length

Returns:

this Builder

optBidrectional

public T optBidrectional(boolean useBidirectional)

Sets the optional parameter that indicates whether to use bidirectional recurrent layers.

Parameters:

useBidirectional - whether to use bidirectional recurrent layers

Returns:

this Builder

optStateOutput

public T optStateOutput(boolean stateOutputs)

Sets the optional parameter that indicates whether to have the states as symbol outputs.

Parameters:

stateOutputs - whether to have the states as symbol output

Returns:

this Builder

self

protected abstract T self()