LLM Fundamentals: Build a Tiny LLM From Scratch

Build a real language model by hand: tokenizer, embeddings, attention, transformer, training loop, generation. The depth vertical. 10 chapters, live.

1. 1Tokens: teaching a computer to read4 q
2. 2Embeddings: giving each token a vector of meaning3 q
3. 3The bigram model: your first real predictor3 q
4. 4Self attention: the idea that changed everything3 q
5. 5Multi head attention: many perspectives at once3 q
6. 6The transformer block: residuals, layer norm, and an MLP3 q
7. 7Assemble the GPT: stacking blocks into a model3 q
8. 8The training loop: how a model actually learns3 q
9. 9Generation: sampling text from your model3 q
10. 10Scaling, and what comes next3 q