ztlearn.dl.layers.recurrent package¶

Submodules¶

ztlearn.dl.layers.recurrent.gru module¶

class ztlearn.dl.layers.recurrent.gru.GRU(h_units, activation=None, input_shape=None, gate_activation='sigmoid')[source]¶

Bases: ztlearn.dl.layers.base.Layer

layer_activation¶

layer_parameters¶

output_shape¶

pass_backward(grad, epoch_num, batch_num, batch_size)[source]¶

pass_forward(inputs, train_mode=True)[source]¶

prep_layer()[source]¶

trainable¶

weight_initializer¶

weight_optimizer¶

ztlearn.dl.layers.recurrent.lstm module¶

class ztlearn.dl.layers.recurrent.lstm.LSTM(h_units, activation=None, input_shape=None, gate_activation='sigmoid')[source]¶

Bases: ztlearn.dl.layers.base.Layer

layer_activation¶

layer_parameters¶

output_shape¶

pass_backward(grad, epoch_num, batch_num, batch_size)[source]¶

pass_forward(inputs, train_mode=True)[source]¶

prep_layer()[source]¶

trainable¶

weight_initializer¶

weight_optimizer¶

ztlearn.dl.layers.recurrent.rnn module¶

class ztlearn.dl.layers.recurrent.rnn.RNN(h_units, activation=None, bptt_truncate=5, input_shape=None)[source]¶

Bases: ztlearn.dl.layers.base.Layer

layer_activation¶

layer_parameters¶

output_shape¶

pass_backward(grad, epoch_num, batch_num, batch_size)[source]¶

pass_forward(inputs, train_mode=True)[source]¶

prep_layer()[source]¶

trainable¶

weight_initializer¶

weight_optimizer¶

Module contents¶