Trying Google Colab

Google's SaaS offering of Jupyter Notebooks

Preface

I have seen a lot of people using Jupyter Notebooks for the past few years, back then, it was mainly on AWS SageMaker, where people were trying to experiment with Machine Learning on AWS, such as through AWS DeepRacer.

I was always curious, but because the seemingly tedious task of setting up AWS Sagemaker and potentially burning a small hole in my wallet if I ‘over-trained’ something, I didn’t attempt it back then.

Recently, I came across a lot of people using Google Colab, which is Google's SaaS offering of Jupyter Notebooks.

I’m amazed!

This is what a lot of people, including myself, need. Having a SaaS offering lowered the barrier to entry for creating AI/ML applications, and the innovation potential is huge.

What is Google Colab?

Colab, or "Colaboratory", allows you to write and execute Python in your browser, with

• Zero configuration required

• Access to GPUs free of charge

• Easy sharing

Especially if you are a student, a data scientist or an AI researcher, Colab can make your coding/experimentation/analysis work easier.

Setups Needed

As this is a SaaS offering from Google, very little setup is needed, no installation etc.

However you probably need to save your secrets/keys/credentials somewhere (don’t hard code them!), such as your Google AI Studio key if you are creating an application utilizing Gemini API, you may save your secret keys in the following location:

Try it out

You may have a go at this to try out the power of Colab by creating a stock analysis AI application. You will be creating four AI agents to complete the tasks in series to churn out a final report.

It’s quite simple, and you will soon get the hang of how Colab can potentially help in your work.

If you need some GCP credits, it’s available while stocks last here. Have fun experimenting!