Preparing for part two

I have gone through the lessons once and I am preparing to go through it again a bit more slowly.
I am looking for people who also want to do the same so that we develop a learning plan or strategy together and also motivate each other.

I would also be glad if @jeremy @sgugger and @lesscomfortable could help us as to the things we should focus on like:

  1. Which papers should we try to implement
  2. What should understanding should we aim to have or what should we be able to produce at the end of each lesson.

Hi if you did not follow the part 1 course that ended in december then you should get handson by developing your own cases like the students here:

Yes, developing your own cases are definitely a good way to practice for part 2. Making sure you’re able to redo each notebook on a different dataset (with the same kind of task) is also something that would help.
For an introduction to pure pytorch, you should check this tutorial by Jeremy. Again, make sure you understand every line of code and that you can redo it from scratch. The official pytorch tutorials aren’t all great, don’t sweat it if you don’t understand some of them.


From what I understand from the final lecture, Jeremy’s advice:

Come back for part 2 where will be looking at all these interesting stuffs in particular going deep into fastai codebase to understand how did we build exactly. We will actually go through as we are building it, we created notebooks of like here is where we were of each day. We will actually see the software development process itself. We talk about the process of doing research, how to read academic papers, how to turn math into code and then a whole bunch of additional types of models that we haven’t seen yet. So we will be kind of like going beyond practical deep learning into actually cutting edge research.

From the AMA, there’s a question: for someone planning to take part 2, what would you recommend doing/learning/practicing until the part 2 course starts? [02:09:37]

Just code. Just code all the time. I know it’s perfectly possible. I hear from people who get to this point of the course and they haven’t written any code yet and if that’s you, it’s OK. You know, you’ve just go through and do it again and this time do code. Look at the shape of your inputs, look at your outputs and make sure you know how to grab a mini batch, look at its main standard deviation and plot it. There’s so much material that we’ve covered. If you can get to a point where you can rebuild those notebooks from scratch without too much cheating. When I said from scratch, use the fastai library, not from scratch from scratch. You’ll be in the top edge of practitioners because you’ll be able to do all of these things yourself. And that’s really really rare. And that will put you in a great position to part 2.

If you are interested, the details are available in my lesson 7 notes.