python - Tensorflow 类型错误 : 'module' object is not callable - tf. contrib.rnn

Question

调用属性错误时遇到问题:

tf.nn.rnn(cell, inputs_series, initial_state=rnn_tuple_state)

收到属性错误:

AttributeError: module 'tensorflow.python.ops.nn' has no attribute 'rnn'

将其更改为:

tf.contrib.rnn(cell, inputs_series, initial_state=rnn_tuple_state)

但是，现在我收到以下错误:

TypeError: 'module' object is not callable

针对以下行:

states_series, current_state = tf.contrib.rnn(cell, inputs_series, initial_state=rnn_tuple_state)

代码如下:

# Forward passes
cell = tf.contrib.rnn.LSTMCell(state_size, state_is_tuple=True)
cell = tf.contrib.rnn.MultiRNNCell([cell] * num_layers, state_is_tuple=True)
states_series, current_state = tf.contrib.rnn(cell, inputs_series, initial_state=rnn_tuple_state)

完全错误:

Traceback (most recent call last):
  File "/Users/glennhealy/PycharmProjects/lstm2/lstm2.py", line 49, in <module>
    states_series, current_state = tf.contrib.rnn(cell, inputs_series, initial_state=rnn_tuple_state)
TypeError: 'module' object is not callable

有什么想法吗？

tf.nn.rnn 不起作用，但 tf.contrib.rnn 也不起作用

提前干杯

根据回复更新了更多信息

看看这个，我已经尝试了tensorflgw_RNN信息中的所有选项，其中很多选项都出现了这个错误:

TypeError: static_bidirectional_rnn() got an unexpected keyword argument 'initial_state'

所以，现在我迷路了。

Answer 1

根据文档 https://www.tensorflow.org/api_guides/python/contrib.rnn

TensorFlow 提供了多种构建循环神经网络的方法。

tf.contrib.rnn.static_rnn
tf.contrib.rnn.static_state_saving_rnn
tf.contrib.rnn.static_bidirectional_rnn
tf.contrib.rnn.stack_bidirectional_dynamic_rnn

试试这个

tf.contrib.rnn.static_rnn(cell, inputs_series, initial_state=rnn_tuple_state)

或

tf.nn.static_rnn(cell, inputs_series, initial_state=rnn_tuple_state)