mlabonne/llm-course — Course to get into Large Language Models (LLMs) with roadmaps and Colab notebooks.

What the course is

The course is a set of learning guides hosted on GitHub by Maxime Labonne. It is split into three parts. The first part, LLM Fundamentals, is optional and covers math, Python, and neural networks. The second part, The LLM Scientist, teaches how to build strong LLMs with the newest methods. The third part, The LLM Engineer, teaches how to make apps that use LLMs and how to put them online.

Key facts

• The course has three parts.
• The first part, LLM Fundamentals, is optional.
• The course lists many notebooks, which are shared coding pages that you can run in Google Colab.
• Tools listed include LLM AutoEval, LazyMergekit, LazyAxolotl, and AutoQuant.
• The author says the course will always stay free.

New to the topic?

A Large Language Model is a computer program trained on huge amounts of text. It learns to predict and generate words, so it can hold a conversation. A notebook is a document that mixes text and code, so you can run the code step by step. Google Colab lets you run these notebooks in your web browser for free.

Why it matters

LLMs are now used in many apps, so many people want to learn how they work. This course gives a clear path from basic math to building and launching real apps. Because it is free and uses hands-on notebooks, students can practice without buying expensive tools.

What to watch

• Whether the author adds new notebooks and articles over time.
• How the course's tools, such as the Fine-tuning notebooks, keep working with new model versions.
• Whether readers use the free course and also buy the author's book, the LLM Engineer's Handbook.