DistributionNLLLoss

class torch_uncertainty.losses.DistributionNLLLoss(reduction='mean')

Negative Log-Likelihood loss using given distributions as inputs.

Parameters:

• reduction (str, optional) – specifies the reduction to apply to the

• output – 'none' | 'mean' | 'sum'. Defaults to “mean”.

forward(dist, targets, padding_mask=None)

Compute the NLL of the targets given predicted distributions.

Parameters:

• dist (Distribution) – The predicted distributions

• targets (Tensor) – The target values

• padding_mask (Tensor, optional) – The padding mask. Defaults to None. Sets the loss to 0 for padded values.