torch.log

torch.log(input, *, out=None) → Tensor

Returns a new tensor with the natural logarithm of the elements of input.

$$y_{i} = \log_{e} (x_{i})$$

Parameters:

input (Tensor) – the input tensor.

Keyword Arguments:

out (Tensor, optional) – the output tensor.

Example:

>>> a = torch.rand(5) * 5
>>> a
tensor([4.7767, 4.3234, 1.2156, 0.2411, 4.5739])
>>> torch.log(a)
tensor([ 1.5637,  1.4640,  0.1952, -1.4226,  1.5204])