10  The Final Deep-dive

In this chapter, I plan to demonstrate a basic LLM from scratch

• internals of a Transformer model
• build decoder only Transformer model for a basic [your-name]-GPT model
• using R torch – planned

This may rely heavily on https://github.com/rasbt/LLMs-from-scratch, in which I will simplify the steps in the book and adapt the code in R.

In progress …