TimeAveragedRNN

Bases: tf.keras.layers.Layer

A simple RNN with time-averaging output.

Note

This layer unrolls TimeAveragedRNNCell for a fixed number of steps. The number of steps is determined by "sequence length" in axis 1 of input tensor's shape.

Parameters:

Name	Type	Description	Default
tau	float	Time-averaging parameter, from 0 to 1.	required
units	int	Number of units in the RNN cell.	required

build(input_shape)

Build the underlying layers/weights.

call(inputs)

Forward pass.

Parameters:

Name	Type	Description	Default
inputs	tf.Tensor	Input tensor $x_t$ with shape (batch_size, seq_len, input_dim).	required

Returns:

Name	Type	Description
outputs	tf.Tensor	Output tensor $h_t$ with shape (batch_size, seq_len, units).