Setup :
Windows 10 computer with NVIDIA GPU
Anaconda Installed
Installing cuda GPU spport for tensorflow
Create enivronment
conda create -n gpuhaneef python==3.8

activate environment
conda activate gpuhaneef

install pre requisites
conda install cudatoolkit=11.0 cudnn=8.0 -c=conda-forge

install tensorflow
pip install –upgrade tensorflow-gpu==2.4.1

Test Installation
Open python
import tensorflow as tf
tf.test.is_gpu_available()

If you want to install another version please refer the attached chart

Please refer my other article for testing using Jupyter :