wandb/edu

What this project is

This is a public repository from Weights & Biases (W&B), the company behind the popular machine-learning experiment-tracking platform. It holds free educational materials for learning AI, organized by topic. They are designed to complement W&B's paid online courses, but the README says they're also useful on their own.

Main content: course materials by category

The repository lists course-style materials across four areas:

1. Large Language Models (LLMs) — six courses, including:

• RAG++ — practical "retrieval-augmented generation" techniques for engineers: production-ready ways to improve performance, cut costs, and boost accuracy.
• Developer's guide to LLM prompting — prompt engineering fundamentals, from system prompts to model-specific strategies.
• LLM Engineering: Structured Outputs — handling structured JSON output, function calling, and complex validations.
• Building LLM-Powered Apps — building apps using LLM APIs, LangChain, and W&B's LLM tooling.
• Training and Fine-tuning LLMs — architecture, training, and fine-tuning methods including LoRA and RLHF.
• Evaluate and Debug Generative AI — evaluating and debugging generative-AI work with the W&B platform.

2. MLOps (machine-learning operations) — five courses, covering model CI/CD, model development workflow, GitOps-ish CI/CD for ML with GitHub Actions, data validation in production pipelines, and using ML for business decision optimization.

3. W&B Tools — two courses:

• W&B 101 — intro to experiment tracking, visualization, and optimization.
• W&B 201: Model Registry — advanced model management: logging, registering, and managing ML models.

4. International Courses — Korean and Japanese versions of the "Effective MLOps: Model Development" course.

Additional resources

• For educators: a folder with materials for educators, teaching assistants, and students.
• Math for ML: a math refresher for machine learning.

How to access everything

All materials live in the GitHub repo (this project). Course enrollment and more info is on the W&B AI Academy website (wandb.courses), which the README links to.

Bottom line: If you want to learn LLM engineering, MLOps, or W&B's own tools — with structured, instructor-designed material — this repository is a self-serve starting point, even without taking the paid courses.