Massachusetts Institute of Technology
MIT 6.S191: Introduction to Deep Learning
MIT's fast-paced bootcamp on deep learning, refreshed each year, with free lectures and labs.
fast.ai's famous top-down deep learning course for people who can already code.
Taught byfast.aiJeremy Howard's course, recorded at the University of Queensland, teaches deep learning by building working models first. Lessons cover deployment, neural net foundations, NLP, models from scratch, random forests, collaborative filtering, CNNs and data ethics, using free notebook platforms.
Best for: Programmers who learn best by building.
Tick modules off as you finish them on fast.ai.
0 of 8 modules done
Plan your pace
At 4 hours a week you would finish in about 4 weeks, using the course time estimate above.
Progress is saved in this browser only. No account needed.
fast.ai does not issue certificates.
Before you start: About a year of coding, ideally Python; High school maths
fast.ai sets the terms and can change them. Check the course page before you commit serious time, and tell us if something changed.
Free, practical deep learning courses for coders.
All free fast.ai courses