评价此页

torch.linalg.solve_ex#

torch.linalg.solve_ex(A, B, *, left=True, check_errors=False, out=None)#

A version of solve() that does not perform error checks unless check_errors= True. It also returns the info tensor returned by LAPACK’s getrf.

注意

当输入在 CUDA 设备上时,此函数仅在 check_errors= True 时进行同步。

警告

此函数是“实验性的”,未来 PyTorch 版本中可能会发生变化。

参数

A (Tensor) – 形状为 (*, n, n) 的张量,其中 * 是零个或多个批次维度。

关键字参数
  • left (bool, optional) – whether to solve the system AX=BAX=B or XA=BXA = B. Default: True.

  • check_errors (bool, optional) – controls whether to check the content of infos and raise an error if it is non-zero. Default: False.

  • out (tuple, optional) – tuple of two tensors to write the output to. Ignored if None. Default: None.

返回

A named tuple (result, info).

示例

>>> A = torch.randn(3, 3)
>>> Ainv, info = torch.linalg.solve_ex(A)
>>> torch.dist(torch.linalg.inv(A), Ainv)
tensor(0.)
>>> info
tensor(0, dtype=torch.int32)