Today’s an exciting day for deep learning enthusiasts! The uber-popular deep learning MOOC, Fast.ai is launching its third version of Part 1 of the course. Currently, the video lectures are only available to those who signed up to be part of Fast.ai Live. They will be released to the general public in January of 2019. In the meantime, you can get some insights into the new version of the course from this post. I’ll be posting after each of the lectures to share insights and help keep you abreast of cutting-edge developments.
Version Three features an entirely rebuilt Fast.ai library. This is the second year for the library, written on top of Pytorch. The first version of the course, released in 2017, having been written in Keras and TensorFlow. One of the salient features that has made the course such a success, is the teaching methodology. The “top-down”, “whole game” approach employed allows learners new to AI/Deep Learning, to get their feet wet, without needing to know lots of math and theory. The course has gotten better every year, so it’s going to be exciting to see what the cutting edge looks like!
Lesson 1 – Right from the first lesson, we can see that the new library offers significant changes and enhancements.