Comment on page
tensor.matmul
fn matmul(self: @Tensor<T>, other: @Tensor<T>) -> Tensor<T>;
Performs matrix product of two tensors. The behavior depends on the dimensionality of the tensors as follows:
- If both tensors are 1-dimensional, the dot product is returned.
- If both arguments are 2-dimensional, the matrix-matrix product is returned.
- If the first argument is 1-dimensional and the second argument is 2-dimensional, a 1 is prepended to its dimension for the purpose of the matrix multiply. After the matrix multiply, the prepended dimension is removed.
- If the first argument is 2-dimensional and the second argument is 1-dimensional, the matrix-vector product is returned.
self(@Tensor<T>) - the first tensor to be multipliedother(@Tensor<T>) - the second tensor to be multiplied
- Panics if the dimension of the tensors is higher than two.
A new
Tensor<T> resulting from the matrix multiplication.Case 1: Dot product of two vectors (1D * 1D)
use array::{ArrayTrait, SpanTrait};
use orion::operators::tensor::{TensorTrait, Tensor, U32Tensor};
fn dot_product_example() -> Tensor<usize> {
let tensor_1 = TensorTrait::<u32>::new(shape: array![3].span(), data: array![0, 1, 2].span(),);
let tensor_2 = TensorTrait::<u32>::new(shape: array![3].span(), data: array![0, 1, 2].span(),);
// We can call `matmul` function as follows.
return tensor_1.matmul(@tensor_2);
}
>>> [5]
Case 2: Matrix multiplication (2D * 2D)
use array::{ArrayTrait, SpanTrait};
use orion::operators::tensor::{TensorTrait, Tensor, U32Tensor};
fn matrix_mul_example() -> Tensor<usize> {
let tensor_1 = TensorTrait::<u32>::new(
shape: array![2, 2].span(), data: array![244, 99, 109, 162].span()
);
let tensor_2 = TensorTrait::<u32>::new(
shape: array![2, 2].span(), data: array![151, 68, 121, 170].span()
);
// We can call `matmul` function as follows.
return tensor_1.matmul(@tensor_2);
}
>>> [[48823, 33422],[36061, 34952]]
Case 3: Matrix-Vector multiplication (2D x 1D)
use array::{ArrayTrait, SpanTrait};
use orion::operators::tensor::{TensorTrait, Tensor, U32Tensor};
fn matrix_vec_mul_example() -> Tensor<usize> {
let tensor_1 = TensorTrait::<u32>::new(
shape: array![3, 3].span(), data: array![0, 1, 2, 3, 4, 5, 6, 7, 8].span(),
);
let tensor_2 = TensorTrait::<u32>::new(shape: array![3].span(), data: array![0, 1, 2].span(),);
// We can call `matmul` function as follows.
return tensor_1.matmul(@tensor_2);
}
>>> [5,14,23]
Last modified 2mo ago