torch.log# torch.log(input, *, out=None) → Tensor# Returns a new tensor with the natural logarithm of the elements of input. yi=loge(xi)y_{i} = \log_{e} (x_{i}) yi=loge(xi) 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])