nn_linear: Linear module

nn_linearR Documentation

Linear module

Description

Applies a linear transformation to the incoming data: y = xA^T + b

Usage

nn_linear(in_features, out_features, bias = TRUE)

Arguments

in_features

size of each input sample

out_features

size of each output sample

bias

If set to FALSE, the layer will not learn an additive bias. Default: TRUE

Shape

  • Input: (N, *, H_in) where * means any number of additional dimensions and H_in = in_features.

  • Output: (N, *, H_out) where all but the last dimension are the same shape as the input and :math:H_out = out_features.

Attributes

  • weight: the learnable weights of the module of shape (out_features, in_features). The values are initialized from U(-√{k}, √{k})s, where k = \frac{1}{\mbox{in\_features}}

  • bias: the learnable bias of the module of shape (\mbox{out\_features}). If bias is TRUE, the values are initialized from \mathcal{U}(-√{k}, √{k}) where k = \frac{1}{\mbox{in\_features}}

Examples

if (torch_is_installed()) {
m <- nn_linear(20, 30)
input <- torch_randn(128, 20)
output <- m(input)
print(output$size())
}

torch documentation built on June 10, 2022, 1:06 a.m.