Fast ai offers a free course called Practical Deep Learning for Coders.

Its fundamental philosophy is designed to let you build fully functioning models first, allowing you to naturally absorb and understand the underlying theoretical concepts as you progress along your unique learning journey.

What makes it different

• You start with real projects, such as image and text models.
• The lessons are free and come with notebooks you can run.
• It focuses on practice over heavy maths.

What you need first

You should be comfortable with basic Python. A computer with an internet connection is usually enough, since many exercises run in the cloud, but check the course page for the current setup.

How to stay on track

1. Do one lesson a week and run every notebook yourself.
2. Change small things in the code and see what happens.
3. Build one small project of your own at the end.

This course is for learners who want to build, not only to understand AI. If you are brand new, begin with an easier course and come back later.

Its core philosophy emphasizes building fully functional models first, allowing you to intuitively grasp the underlying theoretical concepts as you progress along your unique learning journey.