oneflow.logsumexp

oneflow.logsumexp(input, dim, keepdim=False) → Tensor

Returns the log of summed exponentials of each row of the input tensor in the given dimension dim. The computation is numerically stabilized.

For summation index \(j\) given by dim and other indices \(i\), the result is

\[\text{logsumexp}(x)_{{i}} = \log \sum_j \exp(x_{{ij}})\]

The interface is consistent with PyTorch. The documentation is referenced from: https://pytorch.org/docs/1.12/generated/torch.logsumexp.html.

Parameters

  • input (oneflow.Tensor) – the Input Tensor

  • dim (int or tuple of ints) – the dimension or dimensions to reduce.

  • keepdim (bool, optional) – whether the output tensor has dim retained or not. Default: False

For example:

>>> import oneflow as flow

>>> input = flow.Tensor([[1, 2, 3], [4, 5, 6]])
>>> flow.logsumexp(input, 0)
tensor([4.0486, 5.0486, 6.0486], dtype=oneflow.float32)
>>> flow.logsumexp(input, 1)
tensor([3.4076, 6.4076], dtype=oneflow.float32)