1

I read this, this and this. These are not my case. Also I tried to solve my problem using this.

Problem description

I installed cuda 10.0, pip 19.0.2, python 3.6.7, ipython 7.2.0, tensorflow-gpu 1.13.0-rc1, PyCharm CE 2018.3.4. All python modules are installed in virtual environment.

I chose venv interpreter as the Project Interpreter and the Python Console Interpreter in PyCharm.

When I try to import tensorflow

import tensorflow as tf

I've succeed using terminal ipython and failed using PyCharm.

/home/roman/dev/venv/bin/python /home/roman/dev/box.neuro/test.py
Traceback (most recent call last):
  File "/home/roman/dev/venv/lib/python3.6/site-packages/tensorflow/python/pywrap_tensorflow.py", line 58, in <module>
    from tensorflow.python.pywrap_tensorflow_internal import *
  File "/home/roman/dev/venv/lib/python3.6/site-packages/tensorflow/python/pywrap_tensorflow_internal.py", line 28, in <module>
    _pywrap_tensorflow_internal = swig_import_helper()
  File "/home/roman/dev/venv/lib/python3.6/site-packages/tensorflow/python/pywrap_tensorflow_internal.py", line 24, in swig_import_helper
    _mod = imp.load_module('_pywrap_tensorflow_internal', fp, pathname, description)
  File "/home/roman/dev/venv/lib/python3.6/imp.py", line 243, in load_module
    return load_dynamic(name, filename, file)
  File "/home/roman/dev/venv/lib/python3.6/imp.py", line 343, in load_dynamic
    return _load(spec)
ImportError: libcublas.so.10.0: 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 "/home/roman/dev/box.neuro/test.py", line 1, in <module>
    import tensorflow as tf
  File "/home/roman/dev/venv/lib/python3.6/site-packages/tensorflow/__init__.py", line 24, in <module>
    from tensorflow.python import pywrap_tensorflow  # pylint: disable=unused-import
  File "/home/roman/dev/venv/lib/python3.6/site-packages/tensorflow/python/__init__.py", line 49, in <module>
    from tensorflow.python import pywrap_tensorflow
  File "/home/roman/dev/venv/lib/python3.6/site-packages/tensorflow/python/pywrap_tensorflow.py", line 74, in <module>
    raise ImportError(msg)
ImportError: Traceback (most recent call last):
  File "/home/roman/dev/venv/lib/python3.6/site-packages/tensorflow/python/pywrap_tensorflow.py", line 58, in <module>
    from tensorflow.python.pywrap_tensorflow_internal import *
  File "/home/roman/dev/venv/lib/python3.6/site-packages/tensorflow/python/pywrap_tensorflow_internal.py", line 28, in <module>
    _pywrap_tensorflow_internal = swig_import_helper()
  File "/home/roman/dev/venv/lib/python3.6/site-packages/tensorflow/python/pywrap_tensorflow_internal.py", line 24, in swig_import_helper
    _mod = imp.load_module('_pywrap_tensorflow_internal', fp, pathname, description)
  File "/home/roman/dev/venv/lib/python3.6/imp.py", line 243, in load_module
    return load_dynamic(name, filename, file)
  File "/home/roman/dev/venv/lib/python3.6/imp.py", line 343, in load_dynamic
    return _load(spec)
ImportError: libcublas.so.10.0: cannot open shared object file: No such file or directory


Failed to load the native TensorFlow runtime.

See https://www.tensorflow.org/install/errors

for some common reasons and solutions.  Include the entire stack trace
above this error message when asking for help.

Process finished with exit code 1

Additional info

I verified that I use the same ipython in the following way. I checked import requests, noticed that it is not imported in terminal and PyCharm, installed it in terminal and verified that I can import it in both now.

I checked the following:

import os
print(os.environ['PATH'])

Noticed difference in terminal and PyCharm. Then I added environment variable PATH in Settings->Build, Execution, Deployment->Console->Python Console. Now output is the same, but I still can't import tensorflow in PyCharm.

I checked the following:

import os
print(os.environ['LD_LIBRARY_PATH'])

Noticed difference in terminal and PyCharm. PyCharm IPython output is empty and terminal IPython output is the following:

'/usr/local/cuda-10.0/lib64'
talonmies
  • 70,661
  • 34
  • 192
  • 269
RomanS
  • 883
  • 8
  • 14
  • You are completely wrong. Read release notes TF 1.13.0-rc1 or you can read my description again and notice that it works well in terminal IPython. – RomanS Feb 14 '19 at 08:45

1 Answers1

1

I found workaround myself. I added environment variable 'LD_LIBRARY_PATH' in Settings->Build, Execution, Deployment->Console->Python Console. To use Run/Debug configurations I added the same environment variable for them. It is not user-friendly because I should do it for every configuration again and again.

All previous times I installed PyCharm on different computers it worked properly with TF or better to say cuda libraries.

Also, I found issue on PyCharm tracker. So, if you are interested in fix, feel free to vote for it.

RomanS
  • 883
  • 8
  • 14