PyTorch / Jupyter notebook: ModuleNotFoundError: No module named ‘torch’

In this post, you will learn about installing PyTorch, a deep learning library in Jupyter.

Learning is a part and parcel of my journey. I started revisting Linear Algebra to explore the underpinnings of Artificial Intelligence and Quantum Computing. Linear Algebra is a good to have for AI and Quantum.

Recently, I picked PyTorch along with NumPy, SciPy, MatplotLib, and other Python libraries. Jupyter notebook is my go-to tool to learn AI, Data Science and other Python related topics. In this post, I

What is PyTorch ?

PyTorch is an open source machine learning / deep learning framework that is fully featured for building ML / DL models.

Installing PyTorch

As a typical learner, I started with the below command to install `PyTorch` in Jupyter.

!pip install torch

Install PyTorch in Jupyter

The error

In the next cell, when I run

import torch
torch.__version__

I see an error

ModuleNotFoundError: No module named ‘torch’

no module named torch error

The fix

After some research, I found that this is a common error and I was asked to run the `pip` and `python` commands to check the versions and upgrade if required etc., but nothing worked. Until I bumped into a magic command

%pip install torch

You may have to restart your kernel after running this command

Now, when I re-run the `PyTorch` command

PyTorch in Jupyter

The key difference between ! and % is that % is specific to the Jupyter environment while ! runs in a shell. So, % the magic command lasts longer in a Jupyter virtual environment

If you like the content, clap 👏 and for more such content in the future Follow me. Happy learning :)