Cannot import Tensorflow in ipython

Question

I have tensorflow version 1.0.0 non-GPU version on my macbook pro Yosemite. It is installed in a virtualenv, and I have python 2.7. I have six version 1.4.1. When I do python on the command line and import six, it works fine. But when I enter 'ipython' and import tensorflow, I see this error:

AttributeError                            Traceback (most recent call last)
<ipython-input-1-cd232424f60d> in <module>()
----> 1 execfile('hw4/net.py')

/Users/lingxiao/Documents/research/dialogue-systems/hw4/net.py in <module>()
      8 import time
      9 import numpy as np
---> 10 import tensorflow as tf
     11 
     12 import app

/Users/lingxiao/tensorflow/lib/python2.7/site-packages/tensorflow/__init__.py in <module>()
     22 
     23 # pylint: disable=wildcard-import
---> 24 from tensorflow.python import *
     25 # pylint: enable=wildcard-import
     26 

/Users/lingxiao/tensorflow/lib/python2.7/site-packages/tensorflow/python/__init__.py in <module>()
    122 from tensorflow.python.platform import resource_loader
    123 from tensorflow.python.platform import sysconfig
--> 124 from tensorflow.python.platform import test
    125 
    126 from tensorflow.python.util.all_util import remove_undocumented

/Users/lingxiao/tensorflow/lib/python2.7/site-packages/tensorflow/python/platform/test.py in <module>()
     67 # pylint: disable=g-bad-import-order
     68 from tensorflow.python.client import device_lib as _device_lib
---> 69 from tensorflow.python.framework import test_util as _test_util
     70 from tensorflow.python.platform import googletest as _googletest
     71 from tensorflow.python.util.all_util import remove_undocumented

/Users/lingxiao/tensorflow/lib/python2.7/site-packages/tensorflow/python/framework/test_util.py in <module>()
     41 from tensorflow.python.framework import random_seed
     42 from tensorflow.python.framework import versions
---> 43 from tensorflow.python.platform import googletest
     44 from tensorflow.python.platform import tf_logging as logging
     45 from tensorflow.python.util import compat

/Users/lingxiao/tensorflow/lib/python2.7/site-packages/tensorflow/python/platform/googletest.py in <module>()
     31 
     32 from tensorflow.python.platform import app
---> 33 from tensorflow.python.platform import benchmark  # pylint: disable=unused-import
     34 
     35 Benchmark = benchmark.TensorFlowBenchmark  # pylint: disable=invalid-name

/Users/lingxiao/tensorflow/lib/python2.7/site-packages/tensorflow/python/platform/benchmark.py in <module>()
    115 
    116 
--> 117 class Benchmark(six.with_metaclass(_BenchmarkRegistrar, object)):
    118   """Abstract class that provides helper functions for running benchmarks.
    119 

/System/Library/Frameworks/Python.framework/Versions/2.7/Extras/lib/python/six.pyc in with_metaclass(meta, *bases)
    564 def with_metaclass(meta, *bases):
    565     """Create a base class with a metaclass."""
--> 566     return meta("NewBase", bases, {})
    567 
    568 def add_metaclass(metaclass):

/Users/lingxiao/tensorflow/lib/python2.7/site-packages/tensorflow/python/platform/benchmark.py in __new__(mcs, clsname, base, attrs)
    110     newclass = super(mcs, _BenchmarkRegistrar).__new__(
    111         mcs, clsname, base, attrs)
--> 112     if not newclass.is_abstract():
    113       GLOBAL_BENCHMARK_REGISTRY.add(newclass)
    114     return newclass

AttributeError: type object 'NewBase' has no attribute 'is_abstract'

I tried upgrading six and I am still seeing version 1.4.1. Is six the problem here? if not what is?

Answer 1

It seems that there's a number of things tangled. I'm using Anaconda2 with six==1.10.0 , tensorflow==1.0.0 and Python 2.7.12. A quick experiment would be to install Anaconda2 (ie with python2.7) then pip install TF and other dependencies and try it out.

The Anaconda installation is self-contained, so that you won't be affected by other system-wide issues. I just did it by running /Users/myusername/anaconda2/bin/ipython

Answer 2

My guess is that you need to install iPython from within the virtual environment. if you haven't done that, ipython will use the vanilla interpreter even though you are in a virtual environment.