Collaboratory, also called Google Colab It is a product of Google Research and is used to write and run Python and other languages from our browser.
Colab is a hosted Jupyter, installed and configured, so that we do not have to do anything on our computer but simply work from the browser, on resources in the cloud.
It works exactly the same as Jupyter, you can see our article. They are Notebooks or notebooks based on cells that can be texts, images or code, in this Python step, because unlike Jupyter Colab at the moment only the Python kernel can be used, they speak of later implementing others such as R, Scala, etc, but no date is stated.
It is a very fast way to test code without having to configure our equipment and to enter the world of Machine Learning, Deep Learning, artificial intelligence and data science. Also ideal for teachers because being based on Jupyter we can share projects with other people just as if we were using the Jupyter Hub.
We can use any python functionality, we can use TensorFlow, Keras, Numpy, let's go all their libraries.
It offers us a free GPU and TPU service,
They are part of the developer group of https://colaboratory.jupyter.org/welcome/
The service is free but we need a Gmail account. Notebook data is stored in our Google Drive. And we can save and load notebooks from Github as well. In addition to importing projects that come from Jupyter or also exporting them. It works with .ipynb files
It is clear that Hardware resources are limited. You won't be able to create projects that require a large amount of computation. If you like this system and want to use it for advanced projects, you can always pay for the Pro or Pro + version. I'm going to focus on the free one.
In his day I already talked about how one way to use Jupyter from
Google's Machine Learning Crash Course is built on Colab and I'm finishing up. Soon I will tell you how
If you are interested in Machine Learning, see what courses can be done
Why use Colab? Advantage
Because it is a very fast and easy way to set up courses and information about programming in Python and share it with other people or with students if you are a teacher.
In my case I have a compatibility problem between TensorFlow and my CPU, so at the moment I will use it to do different examples and tests with TensorFlow and Keras.
Well, we can only use Pyhton
And that we use yet another Google product and we continue to feed and depend more and more on the technological giant "Don't be Evil"
Differences between Colab and Jupyter
As we said
- Colab is a hosted service, a hosted Jupyter, while Jupyter is using it on your pc
- Colab, although it is free if you want computing power you have to go to the paid version
- Being hosted, you can share notebooks with people