GRU Cell

 Parameters : layer parameters.

 units : integer, dimensionality of the output space.
 Activation : cluster, activation function to use.
 Recurrent Activation : cluster, activation function to use for the recurrent step.
 use bias? : boolean, whether the layer uses a bias vector.
Default value “True”.
 Input Weight Initializer : cluster, initializer for the kernel weights matrix, used for the linear transformation of the inputs.
 Hidden Weight Initializer : cluster, initializer for the recurrent_kernel weights matrix, used for the linear transformation of the recurrent state.
 Bias Initializer : cluster, initializer for the bias vector.
 dropout : float, fraction of the units to drop for the linear transformation of the inputs.
Default value “0.0”.
 recurrent dropout : float, fraction of the units to drop for the linear transformation of the recurrent state.
Default value “0.0”.
 Input Weight Regularizer : cluster, regularizer function applied to the kernel weights matrix.
 Hidden Weight Regularizer : cluster, regularizer function applied to the recurrent_kernel weights matrix.
 Bias Regularizer : cluster, regularizer function applied to the bias vector.
 training? : boolean, whether the layer is in training mode (can store data for backward).
Default value “True”.
 store? : boolean, whether the layer stores the last iteration gradient (accessible via the “get_gradients” function).
Default value “False”.
 update? : boolean, whether the layer’s variables should be updated during backward. Equivalent to freeze the layer.
Default value “True”.

​

Cell : cluster, this cluster defines the recurrent cell type used in a recurrent layer.

enum : enum, an enumeration indicating the cell type (e.g., SimpleRNN, LSTM, GRU, etc.). If enum is set to CustomCell, the class on the right will be used. Otherwise, the selected cell type will be instantiated with default parameters.
 Class : object, a custom RNN cell class instance.