1

my code is:

cell = tf.contrib.rnn.MultiRNNCell([cell for _ in range(2)])
outputs, state = tf.contrib.rnn.static_rnn(cell, inputs)

now the state is printed like this:

(<tf.Tensor 'loss/model_with_buckets/left_rnn/multi_rnn_cell_2/cell_0/GRUBlockCell/GRUBlockCell:3'
shape=(?, 11) dtype=float32>, <tf.Tensor
'loss/model_with_buckets/left_rnn/multi_rnn_cell_2/cell_1/GRUBlockCell/GRUBlockCell:3'
shape=(?, 11) dtype=float32>)

I don't know why state return two values, and how could I use the state for next step? can anyone helps? thanks.

Oussema Aroua
  • 5,225
  • 1
  • 24
  • 44
kaili
  • 11
  • 2

1 Answers1

0

you use output for your next step. and the last output so outputs=outputs[-1]

Ambi
  • 1