Curated paths through the problem set. Pick a track and work through it in order.
Core building blocks. Start here.
10 problems
Build the components of a neural network from scratch.
7 problems
From SGD to Adam. Understand how models learn.
8 problems
Build complete models end to end.
9 problems
From dot-product attention to full transformer blocks.
8 problems
Cutting-edge methods from recent papers.
9 problems