Text generation

Generating text autoregressively: feed the context, take the next-token logits, sample (greedy, temperature, or top-k), append, and repeat.

Code

• @node-llm/core → src/model/gpt.ts — GptModel.generate

Coming soon

This lesson is still being written. It will cover the autoregressive generation loop with greedy, temperature, and top-k sampling. The GptModel.generate method already lives in @node-llm/core.