Issues using JupyterLab while importing PyTorch - Loaded Modules

I am trying to use Pytorch on Interactive Jupyter Lab, and it requires Python 3.7 to be running. When I try loading Python 3.7 and import torch on the terminal, it seems to be working fine. However, I cannot seem to change the Python version of the JupyterLab Notebook, and because of this, Pytorch import is failing. Is there any workaround for this?

Attaching the screenshot

You’ll need to create a kernel for the Python version you want. See instructions here:

Thank you!