Now that you are here, I know You have Suffered your share already!
First read the Instructions carefully, Because all of them are Start-To-Finish Guidelines. If you improvise or skip any step, the results are not My Cause. If things worked out for me, they will work out for you too!
One Advice: Use Your educated Judgement regarding the Versions. I have created the guidelines for the following package configuration:
- python==3.7 | tensorflow-gpu==2.4.1 | CUDA==11.0 | CUDNN==8.0
Thanks!