Skip to content

CurryTang/mlphdinterview

Repository files navigation

A Hitchhiker's Guide to ML PhD Job Hunting

This repository hosts an interview notes site. MLSYS, LLM, Quant, ML Coding, System Design, business algorithm, ML fundamentals, and LeetCode Core Skills notes are published as parallel interview sections.

Live Site

GitHub Pages: https://currytang.github.io/mlphdinterview/

The site publishes the curated MLSYS notes from notes/Mlsys/, Quant notes from notes/quant/, ML Coding notes from notes/MLCoding/, System Design notes from notes/SystemDesign/, TODO placeholders from notes/BusinessAlgorithm/ and notes/MLInterview/, and LeetCode Core Skills notes from notes/Leetcode/. The frontend reader supports Chinese and English variants when both exist, and falls back to the available note when only one variant is present.

Repository Layout

  • notes/Mlsys/: MLSYS interview note markdown files and local assets
  • notes/quant/: Quant probability and expectation interview notes
  • notes/MLCoding/: ML implementation and coding exercise notes
  • notes/SystemDesign/: backend, LLM serving, and agent infra system design notes
  • notes/BusinessAlgorithm/: TODO placeholders for business algorithm interview notes
  • notes/MLInterview/: TODO placeholders for ML fundamentals interview notes
  • notes/Leetcode/: LeetCode Core Skills note markdown files
  • src/: React frontend for browsing and rendering interview sections
  • docs/plans/: design and implementation notes for repo changes

Practice Blocks

Markdown files can render interactive multiple-choice practice blocks with a fenced code block:

```quiz
title: Quick Check
question: CUDA thread blocks are scheduled onto which hardware unit?
answer: B
A. Host compiler
B. GPU SM
C. Browser runtime
explanation: Blocks are assigned to streaming multiprocessors.
```

Use mcq instead of quiz if you prefer that fence name. The rendered block can be hidden, and option clicks show whether the selected answer is correct.

Long-form answers can be kept collapsible with native Markdown/HTML details blocks:

<details class="solution">
<summary>展开解法</summary>

Write the reference solution, pseudocode, and complexity analysis here.

</details>

Local Development

npm install
npm run dev

Verification

npm test
npm run lint
npm run build

Deployment Notes

GitHub Pages deployment is handled by .github/workflows/deploy-pages.yml.

Vite infers the production base path from GITHUB_REPOSITORY in CI. If needed, override it with VITE_BASE_PATH during the build.

About

A Hitchhiker's Guide to ML PhD Job Hunting

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors