speechbrain.nnet.linear module

Library implementing linear transformation.

Authors

• Mirco Ravanelli 2020

Summary

Classes:

Linear

Computes a linear transformation y = wx + b.

Reference

class speechbrain.nnet.linear.Linear(n_neurons, input_shape=None, input_size=None, bias=True, combine_dims=False)

Bases: torch.nn.modules.module.Module

Computes a linear transformation y = wx + b.

Parameters

• n_neurons (int) – It is the number of output neurons (i.e, the dimensionality of the output).

• bias (bool) – If True, the additive bias b is adopted.

• combine_dims (bool) – If True and the input is 4D, combine 3rd and 4th dimensions of input.

Example

>>> inputs = torch.rand(10, 50, 40)
>>> lin_t = Linear(input_shape=(10, 50, 40), n_neurons=100)
>>> output = lin_t(inputs)
>>> output.shape
torch.Size([10, 50, 100])

forward(x)

Returns the linear transformation of input tensor.

Parameters

x (torch.Tensor) – Input to transform linearly.

training: bool