O'Reilly - Developing with LLMs in Python

The (accidental) release of LLaMA, Meta’s large language model, has kickstarted an entirely new ecosystem: open-source models you can download, fine-tune and run on your own infrastructure. In three weeks, this O’Reilly course takes you from zero to working with these models in Python.

What you will learn

  • How transformer-based language models actually work — without a PhD.
  • How to download, run and compare open-source models like LLaMA, Mistral and Gemma.
  • Prompting, embeddings and retrieval-augmented generation (RAG).
  • Fine-tuning with LoRA and deploying your own model behind an API.

The course is hands-on from the first session: every unit combines a short conceptual intro with exercises in Python that you keep as working code. No prior machine learning experience is required — curiosity and basic Python are enough.

Why learn it now

Companies that understand open-source LLMs control their own data, costs and model behavior. The gap between open models and closed commercial models closes with every release — the tooling covered in this course is exactly what you need to close the remaining gap with your own data.

If you want to move from “we should try AI” to “we run our own model in production”, this course is a very efficient way to get there. And if you want a team-specific version with your own use cases, talk to us — we also run in-house workshops.