Bypassing lambda stack when installing libraries with pip inside conda env

I created a conda env on a vector one workstation. Then in this env, I tried to install torch and torchvision using pip. But I got the errors listed below. I am not sure why it is looking under /usr/lib instead of under miniconda3/envs/conda_env. The point of creating a virtual env is to install the libraries locally instead of using the system-wide lambda stack libraries. How can I bypass the lambda stack library and install using pip within a local conda env? I would appreciate any help on this matter.

Defaulting to user installation because normal site-packages is not writeable
Requirement already satisfied: torch in /usr/lib/python3/dist-packages (2.2.1)
Requirement already satisfied: torchvision in /usr/lib/python3/dist-packages (0.17.1)


Have you tried running conda install pytorch torchvision instead of installing with pip?