Start Using Google Colab Free GPU

Since I published “Deep Learning with PyTorch is Not Torturing”, I have been asked for the best way to access free GPU to run deep learning. You can have a free GPU to run PyTorch, OpenCV, Tensorflow, or Keras. My recommendation is Google Colab. Two popular environments offer free GPU: Kaggle and Colab, both are of Google. I like Google Colab because it works seamlessly with my Google Drive. This post will walk you through how to set up a Google Colab account, how to connect GPU, and how to collaborate with your colleagues (that’s why it is called the Collaboratory).

Also, I’d like to mention this machine learning code generator App. It builds the app in Streamlit as the interface. This app can generate machine learning code in Colab, .py, or ipynb. If you are new to machine learning, the code is a good template to work on. And if you are looking for a nice way to build an app to share, Streamlit is a good choice.

(A) What Is Google Colab?

You may have installed Anaconda on your local machine and used the Jupyter notebook. If so, you will quickly learn to use Google Colab. Colab is a free Jupyter notebook environment that runs entirely in the cloud. There is no setup process. The notebooks that you create can be simultaneously edited by your team members — just the way you edit documents in Google Docs. Colab supports many popular…