Why this matters
You do not need an expensive computer to learn machine learning. Free hosted notebooks give you a capable machine in the browser, sometimes with a GPU (graphics processing unit), the chip that makes deep learning practical. Using them well, without losing work or burning through your data bundle, is a real skill.
What a hosted notebook is
A hosted notebook runs on a computer in a data centre. Your browser shows the cells and sends your code there; the results come back to your screen. Well known free examples are Google Colab and Kaggle Notebooks. Both use the same notebook format as Jupyter, so a notebook from one usually opens in the others.
These are named as examples, not recommendations. Free tiers and their limits change over time, so always read the current terms on each service's official pages, such as the Colab FAQ and Kaggle's notebook documentation.
The limits you must plan for
Free hosted notebooks are generous but not unlimited:
- Sessions end. If you are idle for a while, or a session runs for a long time, the service shuts it down. Colab's own FAQ says free resources are not guaranteed and its limits change.