Running notebooks on custom environments in azure machine learning workspace

I am trying to create custom conda environments on azure machine learning workspace to run the notebooks. Please note: this is not related to azure machine learning environments, but rather conda environments on top of which we run the notebooks.

I successfully created a conda environment from terminal on the workspace and installed kernelspec to add the kernel to the list of available kernels environment setup is successful on terminal As seen in the image the kernelspec is installed and the python version is as expected.

However, when I try to select this kernel on the notebook, python version changes back to 3.10 wrong python version on notebook

Help me understand if this behavior is expected and what are the workarounds for this?

whatever the kernal you select you will get the default conda environment only.

and

and

So to verify the kernal and python verion use below code.

import  sys
sys.version

Because the terminal command doesn't run on the selected kernel, it runs on default environment.