ImportError: libcusolver.so.8.0: Unable to open shared objects file: No such file or directory
Possible duplicate of this question. I have a gpu account with which I connect via putty (ssh login). I created a virtualenv there and I am installing tensorflow via pip for gpu. Everything works fine when I run the command
$ pip list
the following list is displayed:
backports.weakref (1.0rc1)
bleach (1.5.0)
funcsigs (1.0.2)
html5lib (0.9999999)
Markdown (2.6.8)
mock (2.0.0)
numpy (1.13.1)
olefile (0.44)
pbr (3.1.1)
Pillow (4.2.1)
pip (9.0.1)
protobuf (3.3.0)
setuptools (36.0.1)
six (1.10.0)
tensorflow-gpu (1.2.1)
Werkzeug (0.12.2)
wheel (0.29.0)
But when I run:
$ python
>>> import tensorflow
The following error is displayed:
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
File "/home/nauman/junaid/final/test/lib/python2.7/site-
packages/tensorflow/__init__.py", line 24, in <module>
from tensorflow.python import *
File "/home/nauman/junaid/final/test/lib/python2.7/site-
packages/tensorflow/python/__init__.py", line 49, in <module>
from tensorflow.python import pywrap_tensorflow
File "/home/nauman/junaid/final/test/lib/python2.7/site-
packages/tensorflow/python/pywrap_tensorflow.py", line 52, in <module>
raise ImportError(msg)
ImportError: Traceback (most recent call last):
File "/home/nauman/junaid/final/test/lib/python2.7/site-
packages/tensorflow/python/pywrap_tensorflow.py", line 41, in <module>
from tensorflow.python.pywrap_tensorflow_internal import *
File "/home/nauman/junaid/final/test/lib/python2.7/site-
packages/tensorflow/python/pywrap_tensorflow_internal.py", line 28, in
<module>
_pywrap_tensorflow_internal = swig_import_helper()
File "/home/nauman/junaid/final/test/lib/python2.7/site-
packages/tensorflow/python/pywrap_tensorflow_internal.py", line 24, in
swig_import_helper
_mod = imp.load_module('_pywrap_tensorflow_internal', fp, pathname,
description)
ImportError: libcusolver.so.8.0: cannot open shared object file: No such
file or directory
Failed to load the native TensorFlow runtime.
See https://www.tensorflow.org/install/install_sources#common_installation_problems
for some common reasons and solutions. Include the entire stack trace
above this error message when asking for help.
I also set the environment variables like this:
export CUDA_HOME=/opt/cuda
export LD_LIBRARY_PATH="$LD_LIBRARY_PATH:$CUDA_HOME/lib64:$CUDA_HOME/extras/CUPTI/lib64"
Also, I found that libcusolver.so.8.0 is not there and libcusolver.so.7.5 is in cuda / lib64 /. Somehow tensorflow finds the wrong file or I don't know anything. Any help would be appreciated as I am new to all of this. Python version: 2.7 OS: Linux
source to share
This could be due to incorrect linking of your libraries.
Just run it sudo ldconfig /usr/local/cuda/lib64
. He decided for me.
If you need to know more: ldconfig
man
page .
source to share
If you installed tensorflow-gpu
using the pip
pre-built .whl
, cuda version has been corrected. As far as I know, since 0.11.0rc1
all the pre-packaged packages are now built for cuda 8
. Thus, there are two ways to solve the problem:
- install
cuda 8
for ready-made packagestensorflow-gpu >= 0.11.orc1
- save
cuda 7.5
and then buildtensorflow-gpu
from source
source to share