1. Overview
ml-eng-lab is a portfolio of self-contained machine-learning notebook experiments.
Active experiments live under notebooks/, archived CodeXGLUE experiments live under
notebooks/archive/, and repository tooling keeps notebook execution, documentation,
CI, and dependency contracts aligned.
1. Repository Map
notebooks/contains 21 active task directories and 29 active notebooks.notebooks/archive/contains preserved Aug-2023 CodeXGLUE summarization experiments.scripts/verify_repo.pyis the fast structural, documentation, and notebook-surface verifier.Makefileowns notebook execution tiers and local validation targets.docs/contains runtime documentation, dependency contracts, findings, and maintenance logs..github/workflows/contains CI and documentation publishing workflows.
2. Recommended Reading Path
Start with the root README.md
for the task index and day-to-day commands.
Then use this site for focused maintenance references:
- Architecture for the system/context view.
- Environment Setup for local, Docker, JupyterHub, and Codespaces runtime paths.
- Dependency Contracts for pinned versions, known vulnerabilities, and manual-only notebook constraints.
- Current Maintenance Log for the active overnight pass record.
- July 2 Maintenance Log for the prior hard-cap run.
3. Published Site Contract
This site is generated by MkDocs Material from the checked-in docs/ tree. The GitHub Pages
workflow builds the site from main, uploads the generated site/ artifact, and deploys it
to https://thekaveh.github.io/ml-eng-lab/ after the Pages workflow lands on main.