Getting started with Google Colaboratory - Part 2
One of the great things about Google Colabs is the simple way that files can be uploaded and downloaded to your Google Drive.
Running a Python script that is on Google Drive
Lets assume that we have a simple TensorFlow script in a folder called code on our Google Drive and that we want to run it in Colab.
(..you can download the actual script from my GitHub page).
- Open a new colab notebook and then run the following code in the code cell..

- You should see this – click directly on the URL or copy and paste it to your browser..

- A webpage will open and you will need to log in to your Google Drive account - Allow Google Drive File Stream access to your Google account...


- Copy the authorisation code then paste it into the cell in the Colab notebook and hit return..

- It should now look like this:

- Now to run our script in the code folder in Colab, all you need to do is open a new code cell, then execute a command like this:

- ..and the training will complete in a few seconds!

Note that if you try to change Runtime type at this point, you will probably have to go through the mount and authorization process again.