tensorplay.nn.functional.cross_entropy

tensorplay.nn.functional.cross_entropy(input, target, weight=None, size_average=None, ignore_index=-100, reduce=None, reduction='mean', label_smoothing=0.0)[source]

Compute the cross entropy loss between input logits and target.

Port of at::native::cross_entropy_loss_symint (aten/src/ATen/native/LossNLL.cpp:633): equal input/target shapes select the class-probability path, positive label_smoothing blends the NLL with a smoothed uniform term, and otherwise this is nll_loss(log_softmax(input), target) with N-d support.

See CrossEntropyLoss for details.