org.deeplearning4j.models.embeddings.learning.impl.sequence

Class DM<T extends SequenceElement>

java.lang.Object

org.deeplearning4j.models.embeddings.learning.impl.sequence.DM<T>

All Implemented Interfaces:

SequenceLearningAlgorithm<T>

public class DM<T extends SequenceElement>
extends java.lang.Object
implements SequenceLearningAlgorithm<T>

DM implementation for DeepLearning4j

Field Summary

Fields

Modifier and Type	Field and Description
protected double[]	expTable
protected static double	MAX_EXP
protected double	negative
protected double	sampling
protected org.nd4j.linalg.api.ndarray.INDArray	syn0
protected org.nd4j.linalg.api.ndarray.INDArray	syn1
protected org.nd4j.linalg.api.ndarray.INDArray	syn1Neg
protected org.nd4j.linalg.api.ndarray.INDArray	table
protected boolean	useAdaGrad
protected int	window

Constructor Summary

Constructors

Constructor and Description
DM()

Method Summary

Modifier and Type	Method and Description
void	configure(VocabCache<T> vocabCache, WeightLookupTable<T> lookupTable, VectorsConfiguration configuration)
void	dm(int i, Sequence<T> sequence, int b, java.util.concurrent.atomic.AtomicLong nextRandom, double alpha, java.util.List<T> labels, boolean isInference, org.nd4j.linalg.api.ndarray.INDArray inferenceVector)
void	finish()
java.lang.String	getCodeName()
ElementsLearningAlgorithm<T>	getElementsLearningAlgorithm()
org.nd4j.linalg.api.ndarray.INDArray	inferSequence(Sequence<T> sequence, long nr, double learningRate, double minLearningRate, int iterations) This method does training on previously unseen paragraph, and returns inferred vector
boolean	isEarlyTerminationHit()
double	learnSequence(Sequence<T> sequence, java.util.concurrent.atomic.AtomicLong nextRandom, double learningRate) This method does training over the sequence of elements passed into it
void	pretrain(SequenceIterator<T> iterator)

Methods inherited from class java.lang.Object

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

Field Detail

MAX_EXP

protected static double MAX_EXP

window

protected int window

useAdaGrad

protected boolean useAdaGrad

negative

protected double negative

sampling

protected double sampling

expTable

protected double[] expTable

syn0

protected org.nd4j.linalg.api.ndarray.INDArray syn0

syn1

protected org.nd4j.linalg.api.ndarray.INDArray syn1

syn1Neg

protected org.nd4j.linalg.api.ndarray.INDArray syn1Neg

table

protected org.nd4j.linalg.api.ndarray.INDArray table

Constructor Detail

DM

public DM()

Method Detail

getElementsLearningAlgorithm

public ElementsLearningAlgorithm<T> getElementsLearningAlgorithm()

Specified by:

getElementsLearningAlgorithm in interface SequenceLearningAlgorithm<T extends SequenceElement>

getCodeName

public java.lang.String getCodeName()

Specified by:

getCodeName in interface SequenceLearningAlgorithm<T extends SequenceElement>

configure

public void configure(@NonNull
                      VocabCache<T> vocabCache,
                      @NonNull
                      WeightLookupTable<T> lookupTable,
                      @NonNull
                      VectorsConfiguration configuration)

Specified by:

configure in interface SequenceLearningAlgorithm<T extends SequenceElement>

pretrain

public void pretrain(SequenceIterator<T> iterator)

Specified by:

pretrain in interface SequenceLearningAlgorithm<T extends SequenceElement>

learnSequence

public double learnSequence(Sequence<T> sequence,
                            java.util.concurrent.atomic.AtomicLong nextRandom,
                            double learningRate)

Description copied from interface: SequenceLearningAlgorithm

This method does training over the sequence of elements passed into it

Specified by:

learnSequence in interface SequenceLearningAlgorithm<T extends SequenceElement>

Returns:

average score for this sequence

dm

public void dm(int i,
               Sequence<T> sequence,
               int b,
               java.util.concurrent.atomic.AtomicLong nextRandom,
               double alpha,
               java.util.List<T> labels,
               boolean isInference,
               org.nd4j.linalg.api.ndarray.INDArray inferenceVector)

isEarlyTerminationHit

public boolean isEarlyTerminationHit()

Specified by:

isEarlyTerminationHit in interface SequenceLearningAlgorithm<T extends SequenceElement>

inferSequence

public org.nd4j.linalg.api.ndarray.INDArray inferSequence(Sequence<T> sequence,
                                                          long nr,
                                                          double learningRate,
                                                          double minLearningRate,
                                                          int iterations)

This method does training on previously unseen paragraph, and returns inferred vector

Specified by:

inferSequence in interface SequenceLearningAlgorithm<T extends SequenceElement>

Parameters:

sequence -

nr -

learningRate -

Returns:

finish

public void finish()

Specified by:

finish in interface SequenceLearningAlgorithm<T extends SequenceElement>