Introduction to Google Colab

I am posting this in Course 1 as there might be people who are new to the field.

Google Colab is amazing browser based runtime to have ready executable python environment.

You can access it at:

Just create you notebook and start typing the code. It comes with GPU support in notebooks and has almost all the main libraries pre-installed. You can upload your data or mount your google drive or google storage bucket right there.

Only, downside is it is ephemeral environment with only 12 hours or so access. But that is good enough for people who just want run the random python code to tinker with it.

Make sure, if you are new, you spend enough time brushing up numpy and pandas operations.

Cheers
Happy Learning

8 Likes

Hi, Sandeep.

Thanks for sharing this info! Yes, I’ve found Colab really easy to use and quite powerful. They even allow you to use GPUs for training. In the free version, there is no guarantee of getting a GPU at any given time, but I think the pay version allows you to guarantee access. In my experience, it was not a problem using the free version in that respect. The only other thing to be aware of is that the free version will suspend your training after a while if it doesn’t see any interactive activity, so you’ll want to add logic to “checkpoint” your parameters periodically so that you can restart after being suspended without losing all your progress.

2 Likes