mindspore.mint.addbmm

mindspore.mint.addbmm(input, batch1, batch2, *, beta=1, alpha=1)[source]

Apply batch matrix multiplication to batch1 and batch2, with a reduced add step and add input to the result.

Note

batch1 and batch2 must be 3-D tensors each containing the same number of matrices.
When batch1 is a \((C, W, T)\) tensor and batch2 is a \((C, T, H)\) tensor, input must be broadcastable with \((W, H)\) tensor, and out will be a \((W, H)\) tensor.
If beta is 0, then input will be ignored.

\[output = \beta input + \alpha (\sum_{i=0}^{b-1} {batch1_i @ batch2_i})\]

Warning

This is an experimental API that is subject to change or deletion.

Parameters

input (Tensor) – The input tensor.
batch1 (Tensor) – The first batch of tensor to be multiplied.
batch2 (Tensor) – The second batch of tensor to be multiplied.

Keyword Arguments

beta (Union[int, float], optional) – Scale factor for input. Default 1 .
alpha (Union[int, float], optional) – Scale factor for batch1 @ batch2. Default 1 .

Returns

Tensor

Supported Platforms:: Ascend

Examples

>>> import mindspore
>>> m = mindspore.mint.ones([3, 3])
>>> arr1 = mindspore.mint.arange(24.0).reshape((2, 3, 4))
>>> arr2 = mindspore.mint.arange(24.0).reshape((2, 4, 3))
>>> mindspore.mint.addbmm(m, arr1, arr2)
    Tensor(shape=[3, 3], dtype=Float32, value=
    [[ 9.49000000e+02,  1.00900000e+03,  1.06900000e+03],
     [ 1.28500000e+03,  1.37700000e+03,  1.46900000e+03],
     [ 1.62100000e+03,  1.74500000e+03,  1.86900000e+03]])