First, below is my system environment

  1. OS: Ubuntu 16.04
  2. current tensorflow: 1.0.1
  3. current cuDNN: 5.1.5
  4. current CUDA: 8.0

I want to upgrade tensorflow to 1.3, but tf 1.3 requires cuDNN 6.

So upgrading cuDNN is necessary but I can't find the way how to upgrade cuDNN and configure tensorflow as I installed before.

1 Answer 1


I wrote a script that can be used to clean install a desired version of cuDNN or upgrade an existing one. Once you perform the steps on the GitHub page, you can change your cuDNN version to v6.0. After that, you don't have to make any configuration for TensorFlow, it will start working automatically with cuDNN v6.0. You can find the script here.

This is what the script does:


rm -f /usr/include/cudnn.h
rm -f /usr/lib/x86_64-linux-gnu/*libcudnn*
rm -f /usr/local/cuda-*/lib64/*libcudnn*

cp -P packages/cudnn/include/cudnn.h /usr/include
cp -P packages/cudnn/lib64/libcudnn* /usr/lib/x86_64-linux-gnu/
chmod a+r /usr/lib/x86_64-linux-gnu/libcudnn*

rm -rf packages/cudnn

It performs installing operations automatically for the version you want.

  • Hi dnzzcn, please include your script to the answer, link only answers are unwelcome.
    – bummi
    Sep 13, 2017 at 12:42

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .