question-mark
Stuck on an issue?

Lightrun Answers was designed to reduce the constant googling that comes with debugging 3rd party libraries. It collects links to all the places you might be looking at while hunting down a tough bug.

And, if you’re still stuck at the end, we’re happy to hop on a call to see how we can help out.

No module named 'cupy.cudnn'

See original GitHub issue

Hi all, please help.

I installed CUDA 9.0, cuDNN 7, Chainer 3.2.0, CuPy 2.2.0, Python 3.6 on Ubuntu 16.04 LTS

but when i check print(chainer.cuda._resolution_error) in python, i got this message: No module named 'cupy.cudnn'

I also check may installed CuDNN, here the result

$ ls /usr/include/cudnn.h
/usr/include/cudnn.h
$ ls /usr/lib/x86_64-linux-gnu/libcudnn*
/usr/lib/x86_64-linux-gnu/libcudnn.so
/usr/lib/x86_64-linux-gnu/libcudnn.so.5
/usr/lib/x86_64-linux-gnu/libcudnn.so.5.1.10
/usr/lib/x86_64-linux-gnu/libcudnn.so.7
/usr/lib/x86_64-linux-gnu/libcudnn.so.7.0.5
/usr/lib/x86_64-linux-gnu/libcudnn_static.a
/usr/lib/x86_64-linux-gnu/libcudnn_static_v7.a

I also reinstall CuDNN with this in command line

$ sudo dpkg -i libcudnn7_7.0.5.15-1+cuda9.0_amd64.deb 
[sudo] password for rezha: 
(Reading database ... 357409 files and directories currently installed.)
Preparing to unpack libcudnn7_7.0.5.15-1+cuda9.0_amd64.deb ...
Unpacking libcudnn7 (7.0.5.15-1+cuda9.0) over (7.0.5.15-1+cuda9.0) ...
Setting up libcudnn7 (7.0.5.15-1+cuda9.0) ...
Processing triggers for libc-bin (2.23-0ubuntu10) ...

but when i check chainer in python, i got this result:

>>> import chainer
>>> chainer.cuda.available
True
>>> chainer.cuda.cudnn_enabled
False

What should i do?

Issue Analytics

  • State:closed
  • Created 5 years ago
  • Comments:5 (3 by maintainers)

github_iconTop GitHub Comments

2reactions
fiarabbitcommented, Jun 27, 2018

If you install cupy-cuda** as @kmaehashi suggested, make sure unintall cupy because of name conflict. Or, if you reinstall cupy, make sure pip install cupy --no-cache-dir (don’t miss –no-cache-dir).

1reaction
rezha130commented, Jun 27, 2018

@kmaehashi i did pip install cupy-cuda90, and pip uninstall cupy

$ pip freeze | grep cupy
cupy-cuda90==4.2.0

when i check chainer in python, i got this result:

>>> import chainer
>>> chainer.cuda.available
True
>>> chainer.cuda.cudnn_enabled
True

But this version of cupy-pycoda90 not compatible with installed Chainer 3.2.0, i got this error message when i train my neural net model: module 'cupy.cuda.cudnn' has no attribute 'getConvolutionForwardAlgorithm'

Read more comments on GitHub >

github_iconTop Results From Across the Web

Chainer: No module named 'cupy.util'` - Stack Overflow
The error message is very clear. Just change L69 of backends/cuda.py : from cupy.util import PerformanceWarning as _PerformanceWarning.
Read more >
Installation — CuPy 11.4.0 documentation
Installing CuPy from PyPI#. Wheels (precompiled binary packages) are available for Linux and Windows. Package names are different depending on your CUDA Toolkit ......
Read more >
cupy-cuda111 - PyPI
CuPy is a NumPy/SciPy-compatible array library for GPU-accelerated computing with Python. This is a CuPy wheel (precompiled binary) package for CUDA 11.1.
Read more >
CuPy Documentation - Read the Docs
Note: If you did not install CUDA Toolkit by yourself, ... For this purpose, CuPy implements two sister methods called cupy.asnumpy() and.
Read more >
No module named 'cupy', cupy 安装出错_wangqianlan的博客
没有安装cupy modle注意:安装cupy时要根据自己的cuda版本选择cupy例如我电脑上是cuda11.3就输入pip install cupy-cuda113如果不知道自己装的cuda版本 ...
Read more >

github_iconTop Related Medium Post

No results found

github_iconTop Related StackOverflow Question

No results found

github_iconTroubleshoot Live Code

Lightrun enables developers to add logs, metrics and snapshots to live code - no restarts or redeploys required.
Start Free

github_iconTop Related Reddit Thread

No results found

github_iconTop Related Hackernoon Post

No results found

github_iconTop Related Tweet

No results found

github_iconTop Related Dev.to Post

No results found

github_iconTop Related Hashnode Post

No results found