Cuda for PyTorch and Cuda for Tensorflow

Question

I want to install PyTorch and for that I visited PyTorch official website, and they give me a command to install it with Cuda:

pip3 install torch==1.9.0+cu111 torchvision==0.10.0+cu111 torchaudio===0.9.0 -f https://download.pytorch.org/whl/torch_stable.html

The version of Cuda, they want me to install for PyTorch, is 11.1. But I already have Cuda install in my computer which is Cuda 11.2 (for TensorFlow 2.5.0). My question is if I install PyTorch with that command they gave me, will it remove Cuda 11.2 ? If it removed Cuda 11.2 and replace it with Cuda 11.1, then TensorFlow won't be able to run on GPU. Do I have to sacrifice TensorFlow in order to use PyTorch ? Is there any method to them keep both ?

Answer 1

One way to sort out your issue is to create virtual environments. Below are two ways to set up virtual environments

• pip virtual environment
• conda virtual environment

I suggest to go for setting up anaconda (conda) virtual environment for different versions of Tensorflow, Pytorch, CUDA

Follow this guide, Guide to conda for tensorflow and pytorch