#2020th of 28 in Learning and cookbooks
ml-engineering
Open book of guides and scripts for training LLMs and VLMs at scale
- Stars
- 19.4k
- License
- CC-BY-SA-4.0
- Last commit
- Oct 2026
- Language
- Python
Missed the latest GitHub check; these facts may be older.
Overview
A written collection of notes, copy-paste commands and scripts on training, fine-tuning and running large language and multi-modal models, drawn from the author's work on BLOOM-176B and IDEFICS-80B. It covers accelerators, storage, networking, SLURM, training, inference, debugging and testing. It also ships two SKILL.md files for AI agents and PDF/EPUB ebook builds.
Who it is for: Engineers and operators training or serving LLMs on GPU clusters
Strengths
- Runnable tools: torch-dist-bench.py for network throughput, mamf-finder.py for real accelerator TFLOPS
- Covers hardware, SLURM, debugging and inference, not only training
- Links a collection of public LLM/VLM training logbooks
- PDF and EPUB downloads, plus two SKILL.md files for AI agents
Weaknesses
- Reference text, not an installable framework or library
- Described by the author as a personal brain dump; coverage may be uneven
- Ebooks are rebuilt only every few weeks, so they can lag the repo
- License is CC-BY-SA-4.0; share-alike applies to derived content
What it needs
- GPU optional
- Needs PyTorch, SLURM
Also in Learning and cookbooks
See all 28| Rank | Project | Score |
|---|---|---|
| 1 | ai-engineering-from-scratchHands-on AI engineering curriculum with 542 lessons across 20 phases | 86 out of 100 |
| 2 | generative-ai-for-beginners21-lesson course on building generative AI apps in Python and TypeScript | 81 out of 100 |
| 3 | hello-agentsChinese-language course on building LLM agents from scratch | 77 out of 100 |
| #4 | ai-agents-for-beginners18-lesson course on building AI agents with Microsoft Agent Framework | 77 out of 100 |
| #5 | llms-from-scratchCode for the book on building a GPT-style LLM in PyTorch | 76 out of 100 |