python - TensorFlow : NotFoundError: Resource AnonymousIterator Error

Question

我正在尝试复制在时间序列中使用 RNN 的示例，如发布的 https://www.tensorflow.org/beta/tutorials/text/time_series#multi-step_model

但是，我收到以下错误消息:

 NotFoundError: Resource AnonymousIterator/AnonymousIterator3/class 
 tensorflow::data::IteratorResource does not exist.
 [[{{node IteratorGetNext_3}}]]

当我尝试执行该行时:

for x, y in val_univariate.take(1):
print(simple_lstm_model.predict(x,steps=1).shape)

环境:

Tenserflow : 2.0.0-beta1
Python : Python 3.6.7 Anaconda
OS : Windows 10 64-bit

有什么解决办法吗？

Answer 1

来自TF doc ,

它指出:

Creates an Iterator for enumerating the elements of this dataset.

The returned iterator implements the Python iterator protocol and therefore can only be used in eager mode.

因此您需要在脚本开头使用tf.enable_eager_execution()。