llama-cookbook

Why we included this project

Meta maintains this collection of notebooks and recipes for building with the Llama model family, and it works best as a learning resource rather than something you deploy. The examples walk through the full span of a project: running inference locally or through hosted APIs, fine-tuning a model on your own data, and assembling retrieval-augmented pipelines. The end-to-end use cases are the most useful part, since they show Llama wired into real products like a WhatsApp bot or long-context document analysis instead of leaving you with isolated snippets. Because Meta curates it and refreshes it as new Llama versions ship, it stays current enough to trust as a starting point.