Skip to contents

Applies a 1D adaptive max pooling over an input signal composed of several input planes.

Usage

nnf_adaptive_max_pool1d(input, output_size, return_indices = FALSE)

Arguments

input

input tensor of shape (minibatch , in_channels , iW)

output_size

the target output size (single integer)

return_indices

whether to return pooling indices. Default: FALSE