dagma.nonlinear.DagmaNonlinear.log_mse_loss¶

dagma.nonlinear.DagmaNonlinear.log_mse_loss(output: torch.Tensor, target: torch.Tensor) → torch.Tensor¶

Computes the logarithm of the MSE loss:

\[\frac{d}{2} \log\left( \frac{1}{n} \sum_{i=1}^n (\mathrm{output}_i - \mathrm{target}_i)^2 \right)\]

Parameters:¶

output : torch.Tensor¶

\((n,d)\) output of the model

target : torch.Tensor¶

\((n,d)\) input dataset

Returns:¶

A scalar value of the loss.

Return type:¶

torch.Tensor