Resolved 3 errors - bernoulli, _FlagValues, num_units

seq2seq/contrib/seq2seq/helper.py

@@ -32,8 +32,15 @@
 
 import six
 
-from tensorflow.contrib.distributions.python.ops import bernoulli
-from tensorflow.contrib.distributions.python.ops import categorical
+# from tensorflow.contrib.distributions.python.ops import bernoulli
+# from tensorflow.contrib.distributions.python.ops import categorical
+try:
+  from tensorflow.python.ops.distributions import bernoulli
+  from tensorflow.python.ops.distributions import categorical
+except:
+  from tensorflow.contrib.distributions.python.ops import bernoulli
+  from tensorflow.contrib.distributions.python.ops import categorical
+
 from tensorflow.python.framework import dtypes
 from tensorflow.python.framework import ops
 from tensorflow.python.layers import base as layers_base

seq2seq/test/pipeline_test.py

@@ -41,7 +41,10 @@
 def _clear_flags():
   """Resets Tensorflow's FLAG values"""
   #pylint: disable=W0212
-  tf.app.flags.FLAGS = tf.app.flags._FlagValues()
+  #tf.app.flags.FLAGS = tf.app.flags._FlagValues()
+  attr_names = []
+  for flag_key in dir(tf.app.flags.FLAGS):
+    delattr(tf.app.flags.FLAGS, flag_key)
   tf.app.flags._global_parser = argparse.ArgumentParser()
 
 

seq2seq/training/utils.py

@@ -115,7 +115,9 @@ def cell_from_spec(cell_classname, cell_params):
   cell_class = locate(cell_classname) or getattr(rnn_cell, cell_classname)
 
   # Make sure additional arguments are valid
-  cell_args = set(inspect.getargspec(cell_class.__init__).args[1:])
+  #cell_args = set(inspect.getargspec(cell_class.__init__).args[1:])
+  cell_args = set(inspect.signature(cell_class.__init__).parameters)
+
   for key in cell_params.keys():
     if key not in cell_args:
       raise ValueError(