Help with last step of installing cuDNN on windows 10

Hello,

I am trying to install cuDNN on Windows 10 by following the installation guideline.

All instructions are clear and everything went well until the very last steps, which are the following:

  1. Include cudnn.lib in your Visual Studio project.
    a) Open the Visual Studio project and right-click on the project name.
    b) Click Linker > Input > Additional Dependencies.
    c) Add cudnn.lib and click OK.

I have the cudnn.lib file, but what does “your Visual Studio project” mean in this context?
Previous steps always mentioned which file I needed to open, but this step has no context at all.

Could anyone help me with this? This is the last step before completion.

Thanks in advance!

PS: To give a little context, I want to install cuDNN so I can use Tensorflow later on in a Python environment, but these are the prerequisite steps.

1 Like

I am also confused in these steps. I have skipped these steps as I want to work in Python environment. But I am getting “no kernel Image found on device”. Can any one tell me is it related to cuDNN configuration?

So I found out that if you are working with a Python environment and not with a visual studio project, you can just skip that last step.

When I tried using Tensorflow in Python it worked, so if you have followed the tutorial correctly it should work in your Python environment, Sundus.
Otherwise I would just advice you to google that error message in combination with Cuda or cuDNN.

1 Like