Softmax — nnf_softmax • torch

Applies a softmax function.

Usage

nnf_softmax(input, dim, dtype = NULL)

input: (Tensor) input
dim: (int) A dimension along which softmax will be computed.
dtype: (torch.dtype, optional) the desired data type of returned tensor. If specified, the input tensor is casted to dtype before the operation is performed. This is useful for preventing data type overflows. Default: NULL.

Softmax is defined as:

$$Softmax(x_{i}) = exp(x_i)/\sum_j exp(x_j)$$

It is applied to all slices along dim, and will re-scale them so that the elements lie in the range [0, 1] and sum to 1.