Conv_tbc — nnf_conv_tbc • torch

Applies a 1-dimensional sequence convolution over an input sequence. Input and output dimensions are (Time, Batch, Channels) - hence TBC.

Usage

nnf_conv_tbc(input, weight, bias, pad = 0)

Arguments

input: input tensor of shape $(sequence length \times b a t c h \times in\_channels)$
weight: filter of shape ( $kernel width \times in\_channels \times out\_channels$ )
bias: bias of shape ( $out\_channels$ )
pad: number of timesteps to pad. Default: 0