You can find the website here: http://v4.aiquizzes.com/
Currently lecture 1 & 2, more to come soon.
I prepared questions with notes / screenshots / links to relevant parts in lecture recordings and supplementary material
The questions are from the perspective of a person who has taken earlier versions of the course - they might not be as thorough as if they were by someone who is taking the course for the first time. On the other hand, I tried to focus on things that I have noticed people generally struggle with, or which I felt were particularly useful to remember. There are many things you learn from the course and they become part of your workflow. You might not need to be reminded how to reshape a tensor in PyTorch, but you might benefit from memorizing that it is a good practice to always provide the dimension across which you want to squeeze a tensor and why it makes sense to do so. I tried to focus on such pieces of information that can be easily missed or forgotten while remembering them can be very helpful.
Also, a lot of Python’s / PyTorch’s functionality was covered in-depth in fastai v3 part 2 - the notes for these lectures are coming soon and I didn’t add too many questions where there would be an overlap.
The lecture on ethics is material that I have not seen before - a lot of it was new to me, a lot I have never seen presented in this way, hence there will be quite a few study questions there.
The functionality is fairly rudimentary, no password reset for instance - so make sure you set it to something you will remember In terms of using the site, please check out the howto for some thoughts on how it can be used.
In a nutshell, the idea is to have notes you can review when you have a couple of minutes of spare time, you can jump to a specific part of the lecture, etc. The questions are presented by an algorithm and as you learn (and provide feedback to the algo) the intervals at which you are being shown a question will grow. This is to ideally facilitate remembering with the least amount of effort, the technique goes by the name of spaced repetition. As a side note, I only learned a little bit about how we learn through hanging around the fastai community - it is amazing in how many ways the journey can be enriching.
If you would like to learn a bit more about spaced repetition, here is a short intro in the form a comic (I learned about it from @MicPie, thx again for all the wonderful material you shared with me! ) Here is a bit of a longer but fascinating read by Michael Nielsen. And last but not least, reading this wonderfully mis-titled book was extremely eye opening and valuable to me.
Anyhow, please share any feedback that you might have If something does not work I’ll try to fix it though how quickly I will be able to get around to doing so might vary.
Oh and one other thing - the site is not listed anywhere, please do not share links to it as it references lectures that are to stay private for as long as the course is not opened to the general public.