Google Colab Notebooks. Index.

Run the cell!

It took at least two AI Winters to survive. The story is obvious: theoretically, Artificial Intelligence as a concept was already here. Neuronal Networks (as a concept: 1943 — McCulloch&Pitts /functional: 1965 — Ivakhnenko/Lapa), Machine Learning (Samuel, 1959), Backpropagation (Werbos — 1975), just to name some of the key researches. Theoretically. But in practice, computational power was still like this:

AI research had still to wait for practical realization. Finally, in the 2000s, every owner of a personal computer became able to do experiments with Machine and Deep Learning.

I remember 2016, as Google Deep Dream emerged (thank Alexander Mordvintsev). It took a good amount of time for me to set it up and to get it running. But the results were astonishing:

I was in love with AI. And I wanted to try out more. Then I discovered Colab Notebooks. And they changed my life and my view on AI. And I bet, it will change yours as well.

There are plenty many of AI models and datasets online: as paper in arXiv, a repository on GitHub etc. You can run them on Linux, but there is also another way - to do it in your browser, even on smartphone.

Colab Notebook is an interactive implementation of AI frameworks which enables you to run AI smoothly, even if your PC capacity is non-sufficient: Colab Notebook uses remote GPU of Google (and recently TPU as well).

What do you need to run Google Notebook?

1. A browser
2. Google Account with enough space on your Google Drive (basic account is usually sufficient).
3. Internet access (which is a requirement for both above)

How to run Colab Notebook?

1. Clone it to your drive
2. Run every cell
3. That's it!

Here is an official presentation of Google Colaboratory:

And here some further information, if you want to know more:

Benefits of Colab Notebooks

Google Colab Notebooks enable the democratization of Data Science.

They allow it everybody — AI researcher, artist, data scientist, just curious person et al. — to enjoy the power of Machine and Deep Learning on every device. Just run the cells, change the parameters, values, and sources, and enjoy the diversity of AI. For free. Even without any programmer knowledge - and I say you, you will want to learn Python and Machine Learning after some working sessions with Colab Notebooks. Just to finetune them in a way you would like to.

These are just some of my favorite things

In my LinkedIn series, I want to share with you some of my favorite Colab Notebooks. Try them out! It's about a text generation, image manipulation, creating new visuals etc. etc.

Be inspired! Enjoy! Explore!

The next post is coming soon.

Index:

1. Google Deep Dream
2. BigGAN
3. Style Transfer
4. ART in ARTificial Intelligence
5. StyleGAN2
6. DeOldify
7. 3D Ken Burns Effect
8. Neural Language Processing: GPT-2

Disclaimer: This article was published in Towards Data Science in March 2020.