tf.nn.rnn_cell.GRUCell(num_units, input_size=None, activation=<function tanh>).num_units就是隐层神经元的个数,默认的activation就是tanh,你也可以自己定义,但是一般都不会去修改。这个函数的主要的参数就是num_units。
例如:
import tensorflow as tf;
import numpy as np;
X = tf.random_normal(shape=[3,5,6], dtype=tf.float32)
X = tf.reshape(X, [-1, 5, 6])
cell = tf.nn.rnn_cell.GRUCell(10)
init_state = cell.zero_state(3, dtype=tf.float32)
output, state = tf.nn.dynamic_rnn(cell, X, initial_state=init_state, time_major=False)
with tf.Session() as sess:
sess.run(tf.initialize_all_variables())
print sess.run(output)
print sess.run(state)