nn-zero-to-hero

Copy this into your coding assistant. It carries where to start and the first runnable example.

I want to learn from nn-zero-to-hero.
Repository: https://github.com/karpathy/nn-zero-to-hero

What this repo is: Andrej Karpathy's free neural networks course that teaches core models by building them from scratch with notebooks.

How to start (per the repo's docs):
whatever the README prescribes

First runnable example:
whatever minimal example the README gives

Prerequisites and caveats: nothing unusual — check the README

Requirements:
1. Read the README and directory structure first — tell me what this repo actually is (tutorial / roadmap / resource list / runnable examples / dataset / template / config / docs) and what background it assumes.
   Whatever it turns out to be (including installable software), say so first and handle it that way — don't force a read-through flow.
2. Don't just clone it and hand it over. Pick the 3 parts most relevant to my goal and say what each covers and roughly how long it takes.
3. If there's runnable example code, get the first example working as described above and paste the real output.
4. Tell me before installing anything extra (compilers / Jupyter / specific versions).
5. Finally give me a shortest path: where to start, what to skip, what comes next.