AI Lab Support

Published by JeanBean on

AI Lab Support:

Lab setups, notebooks, experiments.

A notebook is a programming environment that makes it easy to investigate and experiment with code in real time. Jupyter and Colab notebooks are used extensively in machine learning.

Jupyter typically runs notebooks on the user’s computer while Colab runs notebooks in the Google cloud.

Jupyter allows for more control over local resources but can be difficult to set up with all of the requirements for a particular project. Colab is easier to set up and runs in a Google sandbox that is free, but may be more limited than bringing your own machine.

Getting Started with a Notebook

Let’s start a colab notebook that Google provides for the Word2Vec algorithm used in its search engine. For this example, we will explore Word2Vec concepts and train a Word2Vec model on a small dataset.

[The rest is for Python coders, but feel free to have a look even if you aren’t a Python programmer!]

Open the Colab notebook in a new window by right-clicking “https://colab.research.google.com/github/tensorflow/docs/blob/master/site/en/tutorials/text/word2vec.ipynb” and select “Open in new window”.

Read through the tutorial, and then scroll down to the “Setup” section. Click the Run arrows and watch the output from each section of code:

Screenshot of run cell

Continuing executing cells, reading descriptions and watching progress as you move along. You can inspect variables along the way.

Try it out!