search

suriyadeepan / practical_seq2seq

A simple, minimal wrapper for tensorflow's seq2seq module, for experimenting with datasets rapidly
http://suriyadeepan.github.io/2016-12-31-practical-seq2seq/
GNU General Public License v3.0
569 stars 270 forks source link

AttributeError #17

Closed prakhar21 closed 7 years ago

prakhar21 commented 7 years ago

what version of tensorflow are you using ? While running the code ,I am getting the below mentioned error.

<log> Building Graph </log>Traceback (most recent call last):
  File "train.py", line 55, in <module>
    sess = model.train(train_batch_gen, val_batch_gen)
  File "/home/prakhar/utilities/IOT/PythonUtilPlugin/scripts/chat/mycreation/seq2seq_wrapper.py", line 138, in train
    sess.run(tf.global_variables_initializer())
AttributeError: 'module' object has no attribute 'global_variables_initializer'
suriyadeepan commented 7 years ago

0.12.0

prakhar21 commented 7 years ago

Upgraded to 0.12.0. That issue is resolved, i guess global_variables_initializer() had some different naming convention in previous versions. But, still I am getting the below mentioned error.

<log> Building Graph </log>
<log> Training started </log>
Traceback (most recent call last):
  File "train.py", line 55, in <module>
    sess = model.train(train_batch_gen, val_batch_gen)
  File "/home/prakhar/utilities/IOT/PythonUtilPlugin/scripts/chat/mycreation/seq2seq_wrapper.py", line 145, in train
    self.train_batch(sess, train_set)
  File "/home/prakhar/utilities/IOT/PythonUtilPlugin/scripts/chat/mycreation/seq2seq_wrapper.py", line 99, in train_batch
    batchX, batchY = train_batch_gen.__next__()
AttributeError: 'generator' object has no attribute '__next__'
suriyadeepan commented 7 years ago

python 3

prakhar21 commented 7 years ago

got it. thanks. made the necessary changes. working now.