I’ve recently started FastAI Practical Deep Learning for Coders to learn about machine learning. To hold myself accountable and finish the course I am going to try to write a blog post for every video in the series starting with Lesson 0. By doing this my goal is to not only learn about deep learning and track my progress but reinforce my learning by trying to explain what I’ve learned for people new to machine learning like me. Lesson 0 isn’t officially presented in the course until Lesson 3 but it provides valuable advice for how to complete FastAI and actually learn how to write deep learning code.
Lesson 0 doesn’t have any code examples or notebooks but I took Jeremy’s advice about blogging and interacting with the community by setting up this Quarto blog as the Lesson 0 project. Lesson 0 was created in 2020 before Twitter became X and closed off to people who didn’t have an account. Since then, some of the ML community and other academic communities in CS have migrated to Discord , Mastodon, and Bluesky. I’d recommend checking out the FastAI Forums, FastAI Discord, and BlueSky.
Jeremy Howard’s Advice
- Commit to finish the course
- Try to finish one really good project to show off what you’ve learned in the course
- Share and blog about your work
How to Watch A FastAI Lesson
This advice was presented by Jeremy Howard in Lesson 0
Watch FastAI Lecture
Run notebooks and code presented in the lecture and experiment with code
Reproduce notebook from a clean notebook (Jeremy provides clean notebooks on github)
Repeat with a different dataset (could be personal project)