Accelerate Your Data Science Projects with Google Colab

Umair MaratabUmair Maratab
2 min read

Google Colab is a cloud-based environment that allows you to run your code on powerful virtual machines. If you don't have a powerful machine with a good GPU Google Colab is the solution.

Requirements

You just need a Google account.

Setup

If you have a Google account you can go the following URL:
https://colab.research.google.com/

that's all the setup easy-peasy.

Acess GPU/TPU Inside Google Colab

By default, google collab is using CPU but you can tweak some settings and can access GPU/TPU.

Steps:

  1. Open Google Colab

  2. Go to the top menu

  3. Click on Runtime and select Chnage runtime type.

  4. When you click it you will see:

Hardware accelerator is None which means GPU is not enabled.

  1. Click on the drop down i.e None and you will see GPU and TPU options.

Now select GPU from here.

  1. When you select this you will see a pop-up like this:

  2. Click ok.

Congrats!! ๐ŸŽ‰
Now you can run any project on GPU for free.


Comparison of CPU VS GPU

With CPU Selected (None option)

With GPU Selected

Can you spot the difference?
Comment below.
If you have any further inquiries or encounter any challenges during the setup process, please feel free to leave a comment below. I will be more than happy to assist you to the best of my abilities.


6
Subscribe to my newsletter

Read articles from Umair Maratab directly inside your inbox. Subscribe to the newsletter, and don't miss out.

Written by

Umair Maratab
Umair Maratab

As a computer science enthusiast, I enjoy writing about a wide range of topics, from the latest programming languages to the impact of technology on society.