
Overview
Connect embeddings, attention, transformers, and sampling to understand a small language model.
Training a small language model explains Transformer mechanics without proving that we understand the capabilities of modern LLMs.

Connect embeddings, attention, transformers, and sampling to understand a small language model.