Tensorflow recognises my gpu, but pytorch cannot recognise my gpu

I am fairly new to deep learning and had been using google colab until now, recently I bought a HP Omen with RTX 2060 GPU, my tensorflow gpu works fine, but when I followed through the tutorials to install pytorch from https://github.com/pytorch/pytorch#from-source I could not install CUDNN properly.

I am clueless to where I have gone wrong, I had tried to browse through help for these topics, but there was no other proper solution to use the latest version other than suggestions to downgrade my cuda toolkit.

Any help is appreciated.

When running the verification after installing CUDNN V8, I get the below error.
Executing: mnistCUDNN
cudnnGetVersion() : 8002 , CUDNN_VERSION from cudnn.h : 8002 (8.0.2)
Host compiler version : GCC 7.5.0
ERROR: cuda failure (unknown error) in error_util.h:91

CUDA toolkit version: 11.0

Thank you.

Hi @balahari.balu

What is your Pytorch version ?