Struct MatMulKernel 

pub struct MatMulKernel;

Naive 2-D matrix multiplication kernel.

Computes output = left × right for rank-2 matrices.

• left shape: [m, n]
• right shape: [n, d]
• output shape: [m, d]

Data is interpreted as row-major contiguous storage.

Errors

Returns KernelError::InvalidArguments if inputs.len() != 2.

Returns KernelError::InvalidRank if either input is not rank-2.

Returns KernelError::ShapeMismatch if inner dimensions do not match or output has the wrong shape.

Examples

let a = Tensor::from_vec(vec![2, 3], vec![1.0, 2.0, 3.0,
                                  4.0, 5.0, 6.0]).unwrap();
let b = Tensor::from_vec(vec![3, 2], vec![7.0,  8.0,
                                  9.0, 10.0,
                                 11.0, 12.0]).unwrap();

let mut out = Tensor::zeros(vec![2, 2]).unwrap();
MatMulKernel.compute(&[&a, &b], &mut out).unwrap();
assert_eq!(out.data(), &[58.0, 64.0, 139.0, 154.0]);

Trait Implementations

impl Kernel for MatMulKernel

fn compute( &self, inputs: &[&Tensor], output: &mut Tensor, ) -> Result<(), KernelError>

Computes the kernel output in-place.