A practical introduction to how modern LLMs work, from tokens and transformers to training and inference.
