rag_techniques vs cs249r_book
Two of the top projects in Learning and cookbooks, side by side: who leads on each score part and who each one suits, then setup, license, activity and what each review found.
rag_techniques
Collection of 42 runnable notebooks implementing RAG techniques
cs249r_book
Open textbook and course materials for machine learning systems engineering
Verdict
Both score 79; rag_techniques ranks higher on stars, 30,167 to 29,756. Of the 4 score parts we measure for both, rag_techniques leads on 2, cs249r_book on 1 and 1 is even.
- Adoptionrag_techniques leads by 3 points, 73 to 70: close to even.
- FreshnessEven at 100.
- Maintenancecs249r_book leads by 17 points, 93 to 76: a wide gap.
- Agent-readyrag_techniques leads by 45 points, 45 to 0: a wide gap.
Pick rag_techniques if you are
Engineers learning and comparing RAG techniques through runnable code
- 42 notebooks grouped by category: foundational, query enhancement, context enrichment, and more
- Colab badge on listed notebooks, so no local setup is needed to run them
Pick cs249r_book if you are
Students and instructors learning ML systems engineering
- Covers theory, framework internals, edge hardware and cluster modeling in one curriculum
- TinyTorch walks through building an ML framework across 20 progressive modules
| What we compare | rag_techniques | cs249r_book |
|---|---|---|
| Score parts, out of 100 | ||
| Adoption | 73, popular | 70, popular |
| Freshness | 100, active | 100, active |
| Maintenance | 76, fair | 93, healthy |
| Agent-ready | 45, minimal | 0, none |
| Facts from GitHub and the README | ||
| Stars | 30.2k | 29.8k |
| License | custom license (read the license) | custom license (read the license) |
| Last commit | Oct 2026 | Oct 2026 |
| Last release | Oct 2026 | May 2026 |
| Language | Jupyter Notebook | Python |
| Docker | No | No |
| GPU | Not stated | Not stated |
| arm64 or Apple Silicon | Not stated | Not stated |
rag_techniques
A repository of Jupyter notebooks, each implementing one RAG technique, from basic RAG and CSV RAG to HyDE, proposition chunking, contextual chunk headers, MemoRAG and RAG evaluation. Notebooks open in GitHub or Colab. It is a teaching resource with code and references, not a library or a deployable service.
Who it is for: Engineers learning and comparing RAG techniques through runnable code
Strengths
- 42 notebooks grouped by category: foundational, query enhancement, context enrichment, and more
- Colab badge on listed notebooks, so no local setup is needed to run them
- Covers chunking, query transformation, reliability and evaluation, with video explainers for some
- Recent additions include MemoRAG, Open-RAG-Eval and JSON RAG
Weaknesses
- Notebooks, not an installable package or service; code must be adapted for production
- License reported as NOASSERTION, so reuse terms are unclear
- README carries heavy sponsor, book and newsletter promotion
cs249r_book
A single repository holding a four-volume ML systems textbook (foundations, scaling, agentic, physical AI) plus companion material. The companions are Marimo labs, TinyTorch (build an ML framework in 20 modules), Arduino and Raspberry Pi hardware kits, the MLSys·im infrastructure simulator, StaffML interview questions, slides and an instructor hub. Volume I is released; Volume II is a preview; Volumes III and IV are drafts.
Who it is for: Students and instructors learning ML systems engineering
Strengths
- Covers theory, framework internals, edge hardware and cluster modeling in one curriculum
- TinyTorch walks through building an ML framework across 20 progressive modules
- Includes instructor syllabi, rubrics, TA handbook and Beamer slides for every chapter
- Interactive Marimo labs let readers change parameters and see trade-offs
Weaknesses
- Volumes III and IV are drafts; README asks readers not to cite or teach from them
- Volume II is only a preview
- License is CC-BY-NC-SA 4.0, which restricts commercial use
- A course, not a deployable tool; little of it runs as a service
More in Learning and cookbooks
Building with these?
We help teams pick an open-source AI stack and run it in production. Client work never changes a rank.