What does it mean to import TensorFlow from source

I was trying to use TensorFlow in my Anaconda docker container and I got the error:

Error importing tensorflow. Unless you are using bazel, you should not try to import tensorflow from its source directory; please exit the tensorflow source tree, and relaunch your python interpreter from there.

however, I have no idea what the source directory even means. What does the source directory mean and what does it mean to import tensorflow from it? Why is it bad to do that?

I am only doing:

nvidia-docker run -ti  --rm continuumio/anaconda3 bash

then inside of the container I am doing:

pip install tensorflow-gpu

and then I go to the python3 interpreter and get that error once I import tensorflow.

For completeness, the full error message is:

root@01452be96fe9:/# python3
Python 3.5.2 (default, Nov 17 2016, 17:05:23)
[GCC 5.4.0 20160609] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import tensorflow
Traceback (most recent call last):
  File "/usr/local/lib/python3.5/dist-packages/tensorflow/python/__init__.py", line 49, in <module>
    from tensorflow.python import pywrap_tensorflow
  File "/usr/local/lib/python3.5/dist-packages/tensorflow/python/pywrap_tensorflow.py", line 28, in <module>
    _pywrap_tensorflow = swig_import_helper()
  File "/usr/local/lib/python3.5/dist-packages/tensorflow/python/pywrap_tensorflow.py", line 24, in swig_import_helper
    _mod = imp.load_module('_pywrap_tensorflow', fp, pathname, description)
  File "/usr/lib/python3.5/imp.py", line 242, in load_module
    return load_dynamic(name, filename, file)
  File "/usr/lib/python3.5/imp.py", line 342, in load_dynamic
    return _load(spec)
ImportError: libcudart.so.8.0: cannot open shared object file: No such file or directory

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/usr/local/lib/python3.5/dist-packages/tensorflow/__init__.py", line 24, in <module>
    from tensorflow.python import *
  File "/usr/local/lib/python3.5/dist-packages/tensorflow/python/__init__.py", line 60, in <module>
    raise ImportError(msg)
ImportError: Traceback (most recent call last):
  File "/usr/local/lib/python3.5/dist-packages/tensorflow/python/__init__.py", line 49, in <module>
    from tensorflow.python import pywrap_tensorflow
  File "/usr/local/lib/python3.5/dist-packages/tensorflow/python/pywrap_tensorflow.py", line 28, in <module>
    _pywrap_tensorflow = swig_import_helper()
  File "/usr/local/lib/python3.5/dist-packages/tensorflow/python/pywrap_tensorflow.py", line 24, in swig_import_helper
    _mod = imp.load_module('_pywrap_tensorflow', fp, pathname, description)
  File "/usr/lib/python3.5/imp.py", line 242, in load_module
    return load_dynamic(name, filename, file)
  File "/usr/lib/python3.5/imp.py", line 342, in load_dynamic
    return _load(spec)
ImportError: libcudart.so.8.0: cannot open shared object file: No such file or directory


Error importing tensorflow.  Unless you are using bazel,
you should not try to import tensorflow from its source directory;
please exit the tensorflow source tree, and relaunch your python interpreter
from there.

回答1:

The source directory is the directory containing your TensorFlow installation. Python is getting confused by the fact that it sees TensorFlow source files locally and not just in your PYTHONPATH. The error message asks you to run the interpreter from a location other than the directory in which TensorFlow is installed.

回答2:

try with export LD_LIBRARY_PATH=/usr/local/cuda/lib64

source: http://withr.me/install-tensorflow-ubuntu-16-04/

回答3:

Make sure you are using python 3.5.3! Other version may not work!