Package com.linkedin.dagli.embedding.classification

Class FastTextInternal.Model<L extends java.io.Serializable>

java.lang.Object

com.linkedin.dagli.util.cloneable.AbstractCloneable<AbstractEmbeddingClassifier<L,T>>

com.linkedin.dagli.embedding.classification.AbstractEmbeddingClassifier<L,java.lang.CharSequence>

com.linkedin.dagli.embedding.classification.FastTextInternal.Model<L>

Type Parameters:

L - the type of label used by the model

All Implemented Interfaces:

java.io.Serializable, java.lang.Cloneable

Enclosing class:

FastTextInternal

public static class FastTextInternal.Model<L extends java.io.Serializable>
extends AbstractEmbeddingClassifier<L,java.lang.CharSequence>

Represents a FastText model.

See Also:

Serialized Form

Nested Class Summary

Nested classes/interfaces inherited from class com.linkedin.dagli.embedding.classification.AbstractEmbeddingClassifier

AbstractEmbeddingClassifier.FeaturesEmbeddingResult

Constructor Summary

Constructors

Constructor	Description
Model(L[] labels, DenseFloatArrayVector[] labelEmbeddings, it.unimi.dsi.fastutil.longs.Long2ObjectOpenHashMap<DenseFloatArrayVector> wordEmbeddings, DenseFloatArrayVector[] ngramBucketEmbeddings, boolean multilabel, int maxNgramSize)

Method Summary

Modifier and Type	Method	Description
protected void	distributionalize(float[] labelProbabilities)
protected AbstractEmbeddingClassifier.FeaturesEmbeddingResult	embedFeatures(java.lang.Iterable<? extends java.lang.CharSequence> features)	Gets the embedding of the features.
boolean	equals(java.lang.Object o)
int	hashCode()
<M extends java.io.Serializable> FastTextInternal.Model<M>	remapLabels(java.util.function.Function<L,M> remapper)

Methods inherited from class com.linkedin.dagli.embedding.classification.AbstractEmbeddingClassifier

createResult, embedLabel, getLabelEmbeddingDimensions, getLabelEmbeddings, getLabelEmbeddings, getLabels, getMostLikelyLabelEmbeddings, predictDistribution

Methods inherited from class com.linkedin.dagli.util.cloneable.AbstractCloneable

clone, clone

Methods inherited from class java.lang.Object

finalize, getClass, notify, notifyAll, toString, wait, wait, wait

Constructor Details

Model

public Model(L[] labels,
DenseFloatArrayVector[] labelEmbeddings,
it.unimi.dsi.fastutil.longs.Long2ObjectOpenHashMap<DenseFloatArrayVector> wordEmbeddings,
DenseFloatArrayVector[] ngramBucketEmbeddings,
boolean multilabel,
int maxNgramSize)

Method Details

equals

public boolean equals(java.lang.Object o)

Overrides:

equals in class AbstractEmbeddingClassifier<L extends java.io.Serializable,java.lang.CharSequence>

hashCode

public int hashCode()

Overrides:

hashCode in class AbstractEmbeddingClassifier<L extends java.io.Serializable,java.lang.CharSequence>

remapLabels

public <M extends java.io.Serializable>
FastTextInternal.Model<M> remapLabels(java.util.function.Function<L,M> remapper)

embedFeatures

protected AbstractEmbeddingClassifier.FeaturesEmbeddingResult embedFeatures(java.lang.Iterable<? extends java.lang.CharSequence> features)

Description copied from class: AbstractEmbeddingClassifier

Gets the embedding of the features.

Specified by:

embedFeatures in class AbstractEmbeddingClassifier<L extends java.io.Serializable,java.lang.CharSequence>

Parameters:

features - the features of the example

Returns:

a FeaturesEmbeddingResult containing an embedding of the features, as defined by the model

distributionalize

protected void distributionalize(float[] labelProbabilities)

Specified by:

distributionalize in class AbstractEmbeddingClassifier<L extends java.io.Serializable,java.lang.CharSequence>