mindspore.mint.baddbmm

mindspore.mint.baddbmm(input, batch1, batch2, *, beta=1, alpha=1)

The result is the sum of the input and a batch matrix-matrix product of matrices in batch1 and batch2. The formula is defined as follows:

$${\text{out}}_i=\beta {\text{input}}_i+\alpha ({\text{batch1}}_i@{\text{batch2}}_i)$$

Parameters

• input (Tensor) – The input Tensor. When batch1 is a $(C,W,T)$ Tensor and batch2 is a $(C,T,H)$ Tensor, input must be broadcastable with $(C,W,H)$ Tensor.

• batch1 (Tensor) – $batch1$ in the above formula. Must be 3-D Tensor, dtype is same as input.

• batch2 (Tensor) – $batch2$ in the above formula. Must be 3-D Tensor, dtype is same as input.

Keyword Arguments

• beta (Union[float, int], optional) – multiplier for input. Default: 1 .

• alpha (Union[float, int], optional) – multiplier for $batch1@batch2$. Default: 1 .

Returns

Tensor, has the same dtype as input, shape will be $(C,W,H)$.

Raises

• TypeError – If the type of input, batch1, batch2 is not Tensor.

• TypeError – If the types of input, batch1, batch2 are different.

• ValueError – If batch1 and batch2 are not 3-D tensors.

Supported Platforms:

Ascend

Examples

>>> import numpy as np
>>> from mindspore import Tensor, mint
>>> input = Tensor(np.ones([1, 3, 3]).astype(np.float32))
>>> batch1 = Tensor(np.ones([1, 3, 4]).astype(np.float32))
>>> batch2 = Tensor(np.ones([1, 4, 3]).astype(np.float32))
>>> output = mint.baddbmm(input, batch1, batch2)
>>> print(output)
[[[5. 5. 5.]
  [5. 5. 5.]
  [5. 5. 5.]]]