This repository is a hands-on, step-by-step tutorial series inspired by Andrej Karpathy’s "nn zero to hero" philosophy, tailored for the protein language modeling domain.
This project walks you through building protein language models from scratch — starting from the simplest bigram model, all the way to advanced transformer architectures.
Along the way, you’ll learn how to apply reinforcement learning (RL) techniques to fine-tune these models for protein-specific tasks.
This series is under active development.
Follow the repository, and join the journey from zero to hero in protein language modeling!
Inspired by Andrej Karpathy. Bridging AI and biology, one model at a time.