Module loss 

Loss module

Structs

BinaryCrossEntropyLoss

Calculate the binary cross entropy loss from the input logits and the targets.

BinaryCrossEntropyLossConfig

Configuration to create a Binary Cross-entropy loss using the init function.

CTCLoss

Computes the Connectionist Temporal Classification (CTC) loss.

CTCLossConfig

Configuration for the CTC Loss module.

CosineEmbeddingLoss

Cosine embedding loss between two tensors.

CosineEmbeddingLossConfig

Configuration for CosineEmbeddingLoss.

CrossEntropyLoss

Calculate the cross entropy loss from the input logits and the targets.

CrossEntropyLossConfig

Configuration to create a Cross-entropy loss using the init function.

GaussianNLLLoss

Negative log likelihood loss for a Gaussian target with predicted mean and variance, following torch.nn.GaussianNLLLoss.

GaussianNLLLossConfig

Configuration to create a Gaussian NLL loss using the init function.

HingeEmbeddingLoss

Measures the loss given an input tensor x and a target tensor y (each element 1 or -1), following torch.nn.HingeEmbeddingLoss.

HingeEmbeddingLossConfig

Configuration to create a Hinge Embedding loss using the init function.

HuberLoss

Calculate the Huber loss between the inputs and the target.

HuberLossConfig

Configuration to create a Huber loss.

KLDivLoss

Kullback-Leibler Divergence Loss

KLDivLossConfig

Configuration to create a KLDiv loss.

LpLoss

Computes the Lp Loss between predictions and targets.

LpLossConfig

Configuration for the Lp Loss module.

MarginRankingLoss

Measures the loss for ranking two inputs, following torch.nn.MarginRankingLoss.

MarginRankingLossConfig

Configuration to create a Margin Ranking loss.

MseLoss

Calculate the mean squared error loss from the input logits and the targets.

MultiMarginLoss

Multi-class classification margin (hinge) loss between input x and target class indices y, following torch.nn.MultiMarginLoss.

MultiMarginLossConfig

Configuration to create a Multi Margin loss using the init function.

PoissonNllLoss

Negative Log Likelihood (NLL) loss with a Poisson distribution assumption for the target.

PoissonNllLossConfig

Configuration for creating a PoissonNllLoss instance.

RNNTLoss

RNN Transducer (RNNT) loss, as described in Sequence Transduction with Recurrent Neural Networks.

RNNTLossConfig

Configuration for RNNTLoss.

SmoothL1Loss

Computes the Smooth L1 Loss between predictions and targets.

SmoothL1LossConfig

Configuration for the SmoothL1Loss module.

SoftMarginLoss

Calculate the soft margin loss between the input logits and the target tensor.

TripletMarginLoss

Measures the triplet loss given an anchor, a positive, and a negative input, following torch.nn.TripletMarginLoss.

TripletMarginLossConfig

Configuration to create a Triplet Margin loss using the init function.

Enums

Reduction

The reduction type for the loss.