Milestone map
Milestone map
2 milestones
Do the phase's lessons in public
1–1 weeks (about 10 hrs/week, est)
Work through the 22 lessons of Phase 1: Math Foundations from AI Engineering from Scratch (free) and keep the maths lessons' code (vectors and matrices, autodiff, probability, optimisation, SVD, information theory) in one public repo as you go. About 12 hours (est).
Proof required
Link your public repo with the maths lessons' code (vectors and matrices, autodiff, probability, optimisation, SVD, information theory), and show your self-check (quizzes passed, notebooks run, tests passing where they exist).
What gets checked
- Each lesson's code runs and its quiz is passed.
- Results are checked against a library (NumPy) or by hand.
- Each notebook explains the maths in plain words.
Common mistakes
- Copying the repo's solutions: verifiers read your history and ask about your code.
- Skipping the self-check, then asking for review too early.
Powstik Guide
A public repo with the maths lessons' code (vectors and matrices, autodiff, probability, optimisation, SVD, information theory).
Steps
- Make a public repo for this phase.
- Do one lesson at a time and commit after each.
- Self-check: Run each lesson's code; it must work on your machine as the lesson shows. Each lesson has a quiz (quiz.json, also on the course site): pass it.
- Write a short README: what is here and how to run it.
- Ask three people who know the topic to check it.
Template
README Phase 1: Math Foundations (AI Engineering from Scratch) What is here: How to run it: Self-check result: What was hardest:
What gets sent back
- One big upload with no history.
- Self-check not shown.
- Code that does not run from the README.
Resources
Foundationstart here
Depthgo deeper
What a verifier looks for
- Each lesson's code runs and its quiz is passed.
- Results are checked against a library (NumPy) or by hand.
- Each notebook explains the maths in plain words.
- Ask about one choice in the code; the answer should be clear.
You'll sign in first, then come straight back here.
Check gradients and do PCA by hand
2–4 weeks (6–10 hrs/week)
Use what you learned. Make a notebook that trains a model with gradient descent using gradients you derived with the chain rule, checks them against finite differences, and runs PCA through your own SVD on the same real dataset. It must run, be evaluated and be documented.
Proof required
Link a notebook that trains a model with gradient descent using gradients you derived with the chain rule, checks them against finite differences, and runs PCA through your own SVD on the same real dataset, with a README that says how to run it and what the evaluation found.
What gets checked
- Your gradients match finite differences (error shown).
- Your PCA matches a library's.
- Each step is explained and the data source is named.
Common mistakes
- A demo with no evaluation.
- Numbers without the data or code that produced them.
Powstik Guide
A notebook that trains a model with gradient descent using gradients you derived with the chain rule, checks them against finite differences, and runs PCA through your own SVD on the same real dataset.
Steps
- Write one line: what it does and for whom.
- Build the smallest version that works.
- Write the evaluation first, then improve against it.
- Write the README with the results.
- Ask three people to check it.
Template
README What it does: How to run it: How it was evaluated: Results: Limits:
What gets sent back
- Does not run from the README.
- No evaluation.
- Claims more than the evaluation shows.
Resources
Foundationstart here
What a verifier looks for
- Your gradients match finite differences (error shown).
- Your PCA matches a library's.
- Each step is explained and the data source is named.
- Run it, or watch it run: it must work, be evaluated and be documented.
- Check that the work is the person's own.
You'll sign in first, then come straight back here.