The full GPT model

The complete model: token + positional embeddings → N transformer blocks → final LayerNorm → output projection to vocab logits.

Code

• @node-llm/core → src/model/gpt.ts — GptModel, GPT_CONFIG_124M

Coming soon

This lesson is still being written. It will cover assembling the token and positional embeddings, the stacked transformer blocks, the final LayerNorm, and the output head into the full model. The GptModel class already lives in @node-llm/core.