What course is what? Where should I start? A Guide

I’ve seen many people confused on where to start, what does what, so I’m going to try to sum up all the courses, what they go over, and where to start here in this post :slight_smile: I’ll be going by when a course came out, from most recent to oldest:

Deep Learning from the Foundations (2019)

This course comes after Practical Deep Learning for Coders. It involves building the fastai library from the roots. Extremely program heavy, only recommended after taking Practical Deep Learning for Coders and have at least 1-2 years of experience coding.

Course Link GitHub Code

A Code-First Intro to Natural Language Processing (2019)

A top-down approach to Natural Language Processing taught by Rachel Thomas. This course goes over most of the machine learning techniques for NLP data, and how they all work. Covers everything from Topic Modeling to Sequence to Sequence models.

Lecture Playlist GitHub

Practical Deep Learning for Coders (2018)

A top-down approach to machine learning, uses the most recent version of the fastai codebase (in the fastai repo). One year of coding is recommended, and the best intro to machine learning course I’ve ever taken. This course is a must take and a great starting place.

Lectures GitHub

Computational Linear Algebra v2 2018

A second run of Rachel Thomas’ Computational Linear Algebra. Great for understand some of the background math and how neural networks begin to work. Videos are unavailable, I recommend watching the v1 of the course.

GitHub

Introduction to Machine Learning (2018)

Focuses on random forests and some computer vision and utilizes an older version of the fastai library. Great starting point if you’re nervous about the jump to Deep Learning and Neural Networks. You can find this library in the ‘old’ folder in the fastai repository

Lectures GitHub Notebooks

I hope this helps someone with the somewhat confusing question of “What do I take?” If anyone has edits or improvements please add some! I’ve made this a Wiki :slight_smile:

Thanks! :slight_smile:

17 Likes

Hi Zachary,

I have just started out and I will share what I am doing.

I have started two courses Introduction to Machine Learning (2018) and Practical Deep Learning for Coders (2018). In the first lecture, the instructor mentioned that usually, people start both at the same time.

Deep Learning from the Foundations (2019) should be attempted after completing Practical Deep Learning for Coders (2018) as it is an advanced version. I haven’t seen Rachel’s course until now but it seems like one of her courses is an add-on fashioned and the other one is a full-fledged course that she teaches to graduate students studying to become data scientists at her institute.

Please keep in mind that Practical Deep Learning for Coders has 2018 and 2019 versions. 2019 is recommended. I, unfortunately, started with 2018 and switching to 2019 soon.

Thanks for your question Zachary.

Savan

3 Likes

Hi Zachary,

Is there any difference in topics covered in v2 and v3. I initially took Andrew Ng Deep Learning course now looking to take course in Fast ai and confused between v2 and v3. Digging deep I found topics like Darknet covered in v2 which I am more interested in, I am planning to do a project which requires YOLO algorithm for realime detection.

Everyone seems to advice to take Practical Deep Learning for Coders 2019 first.
Does v3 covers topics related to realtime object detection??

v3 does not, but there are notebooks for it in the repository. I’d recommend v2’s part 2 as it goes into the theory for it along with building the model. Otherwise there are plenty of examples here on the forum!

1 Like

Thank you.
Also where can i find notebooks for v2 part 2 course

They should be on the fastai repo under courses -> dl2

1 Like

Hi Zachary - thanks for writing up these notes. They’re very helpful.

Question: I am just getting started and would like to find someone to act as a tutor. Where/how should I create a post for this?

I’ve found from past learning experiences that if I pay someone to keep me moving over all the learning hurdles it’s well worth it.

Sorry to put this question on you just for being helpful :slight_smile:

Hi Zachary,

I am just starting out the fast.ai courses and this is very helpful. I was just going to ask this as I was little confused about the sequence and versions as well. I also read in one of the forum entries (can’t find it anymore) that there will be a new version of the course with the release of the “Deep Learning for Coders with Fastai and PyTorch” book - is that correct and if so when should I expect them to be released? Thank you!

Best,
Rahul

Yes, the new part 1 was finished a few months ago. Public release should co-inside with the book, so around august time I believe now (that’s my best guess, based on when the book will be published, and considering it’s June and we haven’t heard anything I don’t believe it’ll be July). In the meantime my course was dedicated around v2, Walk with fastai2 :slight_smile:

1 Like

I will check it out. Thank you very much for the prompt reply!!