Sigmoid module

Applies the element-wise function:

Usage

nn_sigmoid()

Details

$$ \mbox{Sigmoid}(x) = \sigma(x) = \frac{1}{1 + \exp(-x)} $$

Shape

• Input: \((N, *)\) where * means, any number of additional dimensions

• Output: \((N, *)\), same shape as the input

Examples

if (torch_is_installed()) {
m <- nn_sigmoid()
input <- torch_randn(2)
output <- m(input)
}