neuralop.losses.equation_losses.PoissonEqnLoss

class neuralop.losses.equation_losses.PoissonEqnLoss(boundary_weight, interior_weight, diff_method: str = 'autograd', base_loss=<function mse_loss>)[source]

PoissonEqnLoss computes a weighted sum of equation loss computed on the interior points of a model’s output and a boundary loss computed on the boundary points.

Parameters:

boundary_weightfloat

weight by which to multiply boundary loss

interior_weightfloat

weight by which to multiply interior loss

diff_methodLiteral[‘autograd’, ‘finite_difference’], optional

method to use to compute derivatives, by default ‘autograd’

base_lossCallable, optional

base loss class to use inside equation and boundary loss, by default F.mse_loss

Methods

__call__(out, y, **kwargs)

Call self as a function.