ai.djl.training.loss

Class CompositeLoss

java.lang.Object

ai.djl.training.metrics.TrainingMetric

ai.djl.training.loss.Loss

ai.djl.training.loss.CompositeLoss

All Implemented Interfaces:

java.lang.Cloneable

public class CompositeLoss
extends Loss

CompositeLoss is an implementation of the Loss abstract class that can combine different Loss functions by adding the individual losses together.

Constructor Summary

Constructors

Constructor and Description
CompositeLoss(Loss... components) Creates a new instance of CompositeLoss that can combine the given Loss components.

Method Summary

Modifier and Type	Method and Description
Loss	duplicate() Creates and returns a copy of this object.
NDArray	getLoss(NDList label, NDList prediction) Calculates loss between the label and prediction.
float	getValue() Calculates metric values.
void	reset() Resets metric values.
void	update(NDList labels, NDList predictions) Updates the training metrics based on a NDList of labels and predictions.

Methods inherited from class ai.djl.training.loss.Loss

excludeBatchAxis, hingeLoss, hingeLoss, l1Loss, l1Loss, l2Loss, l2Loss, sigmoidBinaryCrossEntropyLoss, sigmoidBinaryCrossEntropyLoss, softmaxCrossEntropyLoss, softmaxCrossEntropyLoss

Methods inherited from class ai.djl.training.metrics.TrainingMetric

checkLabelShapes, checkLabelShapes, getName

Methods inherited from class java.lang.Object

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

Constructor Detail

CompositeLoss

public CompositeLoss(Loss... components)

Creates a new instance of CompositeLoss that can combine the given Loss components.

Parameters:

components - the Loss objects that form the composite loss

Method Detail

getLoss

public NDArray getLoss(NDList label,
                       NDList prediction)

Calculates loss between the label and prediction.

Specified by:

getLoss in class Loss

Parameters:

label - the true label

prediction - the predicted label

Returns:

the loss value

duplicate

public Loss duplicate()

Creates and returns a copy of this object.

Overrides:

duplicate in class Loss

Returns:

a copy of this TrainingMetric

update

public void update(NDList labels,
                   NDList predictions)

Updates the training metrics based on a NDList of labels and predictions.

This is a synchronized operation. You should only call it at the end of a batch or epoch.

Overrides:

update in class Loss

Parameters:

labels - a NDList of labels

predictions - a NDList of predictions

reset

public void reset()

Resets metric values.

Overrides:

reset in class Loss

getValue

public float getValue()

Calculates metric values.

Overrides:

getValue in class Loss

Returns:

a Pair of metric name and value