Method apply

gsplat/compression_simulation/mask.py:27–29

Return the masked tensor along with optional loss/metrics.

(self, tensor: Tensor, step: int)

        """Observe the current splats before running compression."""

    def apply(self, tensor: Tensor, step: int) -> MaskResult:
        """Return the masked tensor along with optional loss/metrics."""
        return MaskResult(value=tensor)

    def step_optimizer(self, step: int) -> None:
        """Hook for stepping any optimizers the mask may own."""