CODEWITHSUNDEEP
tensorflow
Mario Ariyanto
Mario Ariyanto

Reputation: 326

ImportError: DLL load failed: A dynamic link library (DLL) initialization routine failed

I installed TensorFlow with CPU support with CUDA Toolkit v9.0 and cuDNNv7.1 and Python 3.6 in Conda enviroment.

I followed the instructions described in TensorFlow website, and when I tried:

import tensorflow as tf

I got this error:

Traceback (most recent call last):
      File "C:\Users\31121\AppData\Local\conda\conda\envs\tensorflow_gpu\lib\site-packages\tensorflow\python\pywrap_tensorflow.py", line 58, in <module>
        from tensorflow.python.pywrap_tensorflow_internal import *
      File "C:\Users\31121\AppData\Local\conda\conda\envs\tensorflow_gpu\lib\site-packages\tensorflow\python\pywrap_tensorflow_internal.py", line 18, in <module>
        _pywrap_tensorflow_internal = swig_import_helper()
      File "C:\Users\31121\AppData\Local\conda\conda\envs\tensorflow_gpu\lib\site-packages\tensorflow\python\pywrap_tensorflow_internal.py", line 17, in swig_import_helper
        return importlib.import_module(mname)
      File "C:\Users\31121\AppData\Local\conda\conda\envs\tensorflow_gpu\lib\importlib\__init__.py", line 126, in import_module
        return _bootstrap._gcd_import(name[level:], package, level)
    ImportError: DLL load failed: A dynamic link library (DLL) initialization routine failed.
    
    During handling of the above exception, another exception occurred:
    
    Traceback (most recent call last):
      File "<stdin>", line 1, in <module>
      File "C:\Users\31121\AppData\Local\conda\conda\envs\tensorflow_gpu\lib\site-packages\tensorflow\__init__.py", line 24, in <module>
        from tensorflow.python import *  # pylint: disable=redefined-builtin
      File "C:\Users\31121\AppData\Local\conda\conda\envs\tensorflow_gpu\lib\site-packages\tensorflow\python\__init__.py", line 49, in <module>
        from tensorflow.python import pywrap_tensorflow
      File "C:\Users\31121\AppData\Local\conda\conda\envs\tensorflow_gpu\lib\site-packages\tensorflow\python\pywrap_tensorflow.py", line 74, in <module>
        raise ImportError(msg)
    ImportError: Traceback (most recent call last):
      File "C:\Users\31121\AppData\Local\conda\conda\envs\tensorflow_gpu\lib\site-packages\tensorflow\python\pywrap_tensorflow.py", line 58, in <module>
        from tensorflow.python.pywrap_tensorflow_internal import *
      File "C:\Users\31121\AppData\Local\conda\conda\envs\tensorflow_gpu\lib\site-packages\tensorflow\python\pywrap_tensorflow_internal.py", line 18, in <module>
        _pywrap_tensorflow_internal = swig_import_helper()
      File "C:\Users\31121\AppData\Local\conda\conda\envs\tensorflow_gpu\lib\site-packages\tensorflow\python\pywrap_tensorflow_internal.py", line 17, in swig_import_helper
        return importlib.import_module(mname)
      File "C:\Users\31121\AppData\Local\conda\conda\envs\tensorflow_gpu\lib\importlib\__init__.py", line 126, in import_module
        return _bootstrap._gcd_import(name[level:], package, level)
    ImportError: DLL load failed: A dynamic link library (DLL) initialization routine failed.
    
    
    Failed to load the native TensorFlow runtime.

Did I miss something?

Upvotes: 6

Views: 41999

Answers (8)

Anmol Bajwa
Anmol Bajwa

Reputation: 1

You're probably using incompatible versions of python and tensorflow. One way is to install miniconda and create a virtual environment with python 3.11 Install tensorflow version 2.12 This worked for me....

Upvotes: 0

Jorge Rodr&#237;guez Celis
Jorge Rodr&#237;guez Celis

Reputation: 51

I had the same problem. I realized my cpu didn't support AVX. I downgraded tensorflow to 1.5 and it worked like a charm

pip install tensorflow==1.5

if not you can also try installing tensorflow from source here in case your cpu doesn't support AVX either.

Upvotes: 0

cespon
cespon

Reputation: 5760

I had a similar issue. I have a computer with an Intel Celeron J3455 and I discovered thath my CPU doesn't support AVX, so I can't use official binaries from TensorFlow 1.6 included.

As ABHAY KOTAL proposes, you can use a unofficial wheel that you can find in the following repository: https://github.com/fo40225/tensorflow-windows-wheel

In my case, I choosed Tensorflow 1.15 for Python 3.7 (py37): tensorflow-1.15.0-cp37-cp37m-win_amd64.whl. If you want another version, download an avaliable sse2 version.

To install the new TensorFlow:

pip install tensorflow-1.15.0-cp37-cp37m-win_amd64.whl

Remember to uninstall before the previous TensorFlow version installed:

pip uninstall tensorflow tensorboard tensorboard-plugin-wit tensorflow-estimator tensorflow-tensorboard

Upvotes: 1

Manzini Lionel .T
Manzini Lionel .T

Reputation: 37

Here is a simple solution (no conda, no gpu/cpu staff)

Uninstall any version of tensorflow you have already installed.

pip uninstall tensorflow

install tensorflow 1.15

pip install tensorflow==1.15

run python

then try import tensorflow

Upvotes: 3

Chafik Boulealam
Chafik Boulealam

Reputation: 694

you get the problem when used pip not conda

the best answer is (it's work 100%):

type following command: if already install TensorFlow you should reinstall it by the following command:

 pip uninstall TensorFlow

if not, install it by the following command :

 conda install TensorFlow

I get this solution from this link below :

https://github.com/tensorflow/tensorflow/issues/23683#issuecomment-513788003

thank's @sdmars Tamiris Crepalde she is helping me to get the solution:

enter image description here

Upvotes: 3

harsh
harsh

Reputation: 1

found this solution on another thread: https://github.com/tensorflow/tensorflow/issues/23683

People, I have solved my problem taking another path. I uninstalled the TensorFlow package that I installed using pip and reinstall using "conda install TensorFlow" and it worked! Now I can import TensorFlow without problems. Hope this can help you guys too!

Upvotes: 0

ABHAY KOTAL
ABHAY KOTAL

Reputation: 163

I have found solution for this it might be that your cpu doesnt support AVX so I install tensorflow from this custommade repository found on github link is here

https://github.com/fo40225/tensorflow-windows-wheel

I selected 1.12.0 for py 37 in that folder select py36 folder and cpu folder and download the file within sse2

cpu which doesn support download sse2 whl

Upvotes: 6

Huitaek Yun
Huitaek Yun

Reputation: 21

I had a similar problem. I knew that my CPU (Pentium G4560) does not support AVX, so I changed tensorflow version from 1.6 to 1.5. There is a link for the manual:

https://github.com/rohit-patel/Install_Instructions-Win10-Deeplearning-Keras-Tensorflow

Upvotes: 2

Related Questions