deephyper.keras.layers.UpdateFuncGRU#
- class deephyper.keras.layers.UpdateFuncGRU(*args: Any, **kwargs: Any)[source]#
Bases:
Layer
Gated recurrent unit update function.
Check details here https://arxiv.org/abs/1412.3555
- Parameters:
state_dim (int) – number of output channels.
Methods
build
Apply the layer on input tensors.
- call(inputs, **kwargs)[source]#
Apply the layer on input tensors.
- Parameters:
inputs (list) – old_state (tensor): node hidden feature tensor (batch size * # nodes * state dimension) agg_messages (tensor): node hidden feature tensor (batch size * # nodes * state dimension)
- Returns:
activated tensor from update function (batch size * # nodes * state dimension)
- Return type:
activation (tensor)