Package ai.djl.training.loss

Contains classes for measuring the Loss of a model.

It contains a main interface Loss and various losses that extend it.

Class Summary

Class	Description
AbstractCompositeLoss	AbstractCompositeLoss is a Loss class that can combine other Losses together to make a larger loss.
ElasticNetWeightDecay	ElasticWeightDecay calculates L1+L2 penalty of a set of parameters.
HingeLoss	HingeLoss is a type of Loss.
IndexLoss	A wrapper for a Loss that evaluates on only a particular NDArray in the predictions and/or labels NDLists.
L1Loss	L1Loss calculates L1 loss between label and prediction.
L1WeightDecay	L1WeightDecay calculates L1 penalty of a set of parameters.
L2Loss	Calculates L2Loss between label and prediction, a.k.a.
L2WeightDecay	L2WeightDecay calculates L2 penalty of a set of parameters.
Loss	Loss functions (or Cost functions) are used to evaluate the model predictions against true labels for optimization.
MaskedSoftmaxCrossEntropyLoss	MaskedSoftmaxCrossEntropyLoss is an implementation of Loss that only considers a specific number of values for the loss computations, and masks the rest according to the given sequence.
SigmoidBinaryCrossEntropyLoss	SigmoidBinaryCrossEntropyLoss is a type of Loss.
SimpleCompositeLoss	SimpleCompositeLoss is an implementation of the Loss abstract class that can combine different Loss functions by adding the individual losses together.
SingleShotDetectionLoss	SingleShotDetectionLoss is an implementation of Loss.
SoftmaxCrossEntropyLoss	SoftmaxCrossEntropyLoss is a type of Loss that calculates the softmax cross entropy loss.