Skip to content

BaseRNNCell

source 

BaseRNNCell(
   input_layer, previous_state, state_size, n_units, dtype = tf.float32,
   w_init = glorot_uniform_init(), u_init = glorot_uniform_init(),
   bias_init = zeros_init(), activation = tf.tanh, w_dropconnect = None,
   u_dropconnect = None, x_dropout = None, r_dropout = None, y_dropout = None,
   dropout_locked = True, regularized = False, share_state_with = None,
   name = 'recurrent_cell'
)

Args

  • previous_state : the recurrent input Layer for the cell
  • state_size : list of number of units for each element in the state, default is a single state with [n_units]
  • n_units (int) : number of activation units for the RNN cell
  • dtype : Layer (output) dtype input_layer the input Layer for this cell

Methods:

.compute_shape

source 

.compute_shape()

.reuse_with

source 

.reuse_with(
   input_layer, *previous_state, regularized = None, name = None, **kwargs
)