Hi, I'am exploring your code recently, I've encountered a problem when I try to run python Omniglot.py
.
System: Ubuntu 16.04
tensroflow version: 1.1.0-rc2
cuda 8.0
python: 2.7
The error information is as following, could you please help me to find out the solution?
python Omniglot.py
Traceback (most recent call last):
File "Omniglot.py", line 112, in <module>
omniglot()
File "Omniglot.py", line 31, in omniglot
output_var, output_var_flatten, params = memory_augmented_neural_network(input_ph, target_ph, batch_size=batch_size, nb_class=nb_class, memory_shape=memory_shape, controller_size=controller_size, input_size=input_size, nb_reads=nb_reads)
File "/data/yangyu/workspace/MANN/MANN/Model.py", line 33, in memory_augmented_neural_network
W_key = tf.get_variable('W_key', shape=shape,initializer=tf.random_uniform_initializer(-1*high, high))
File "/data/yangyu/anaconda2/lib/python2.7/site-packages/tensorflow/python/ops/variable_scope.py", line 1065, in get_variable
use_resource=use_resource, custom_getter=custom_getter)
File "/data/yangyu/anaconda2/lib/python2.7/site-packages/tensorflow/python/ops/variable_scope.py", line 962, in get_variable
use_resource=use_resource, custom_getter=custom_getter)
File "/data/yangyu/anaconda2/lib/python2.7/site-packages/tensorflow/python/ops/variable_scope.py", line 367, in get_variable
validate_shape=validate_shape, use_resource=use_resource)
File "/data/yangyu/anaconda2/lib/python2.7/site-packages/tensorflow/python/ops/variable_scope.py", line 303, in _true_getter
is_scalar = shape is not None and not shape
ValueError: The truth value of an array with more than one element is ambiguous. Use a.any() or a.all()