Applies an affine transformation to the input.
Concretely:
\[y = W^\top x + b\]
where \(W\) has shape [output_dims, input_dims].
- Parameters:
input_dims (int) – The dimensionality of the input features
output_dims (int) – The dimensionality of the output features
bias (bool, optional) – If set to False then the layer will
not use a bias. Default True.
