tensorplay.nn.functional.multi_margin_loss

tensorplay.nn.functional.multi_margin_loss(input: TensorBase, target: TensorBase, p: int = 1, margin: float = 1.0, weight: TensorBase | None = None, size_average=None, reduce=None, reduction: str = 'mean') → TensorBase[source]

Compute the multi margin loss, with optional weighting.

Vectorized composition of aten/src/ATen/native/LossMultiMargin.cpp: sum_d max(0, margin - x_y + x_d)^p * w_y / C over non-target classes. See MultiMarginLoss for details.