Before you move on, you might want to play with the code for a little while. That’s optional, but it’s a good way to make these concepts stick.
To begin with, you can start getting familiar with the system’s hyperparameters. (If you don’t know what a “hyperparameter” is, read “Hyperparameters”?.) Try changing the value of the lr argument to train. What happens if you set lr to a very small value? What if you set it to a large value? What are we gaining and losing in the two cases? After you’re done experimenting, check out the content of the 02_first/solution directory for straight answers to those questions.