In [9]: x=torch.randn(2, 3, dtype=torch.cdouble, requires_grad=True)

In [10]: torch.cumsum(x, dim=0)
---------------------------------------------------------------------------
RuntimeError                              Traceback (most recent call last)
<ipython-input-11-9aad052c3955> in <module>
----> 1 torch.cumsum(x, dim=0)

RuntimeError: cumsum does not support automatic differentiation for outputs with complex dtype.