91b0d220c8e816766fd4565e1d2f5115d3afbefe,gpytorch/lazy/lazy_tensor.py,LazyTensor,matmul,#LazyTensor#Any#,630

Before Change

        // Work out batch dimension, if necessary
        lazy_tsr = self
        if lazy_tsr.ndimension() == 3 and tensor.ndimension() == 3:
            if lazy_tsr.size(0) == 1 and tensor.size(0) > 1:
                lazy_tsr = lazy_tsr.repeat(tensor.size(0), 1, 1)
            elif tensor.size(0) == 1:
                tensor = tensor.expand(lazy_tsr.size(0), tensor.size(1), tensor.size(2))
        elif self.ndimension() > 3 or tensor.ndimension() > 3:
            raise RuntimeError

        func = Matmul(self.representation_tree())
        return func(tensor, *self.representation())

    @property

After Change

            is the matrix that this :obj:`gpytorch.lazy.LazyTensor` represents, and :math:`M` is the matrix input
            to this method.
        
        if (self.dim() == 2 and tensor.dim() == 1):
            if self.shape[-1] != tensor.numel():
                raise RuntimeError(
                    "LazyTensor (size={}) cannot be multiplied with right-hand-side Tensor (size={}).".format(
                        self.shape, tensor.shape
                    )
                )
        elif self.dim() != tensor.dim():
            raise RuntimeError(
                "LazyTensor (size={}) and right-hand-side Tensor (size={}) should have the same number "
                "of dimensions.".format(self.shape, tensor.shape)
            )
        elif self.batch_shape != tensor.shape[:-2] or self.shape[-1] != tensor.shape[-2]:
            raise RuntimeError(
                "LazyTensor (size={}) cannot be multiplied with right-hand-side Tensor (size={}).".format(
                    self.shape, tensor.shape
                )
            )

        func = Matmul(self.representation_tree())
        return func(tensor, *self.representation())

    @property

Instances