This post walks through building a LoRA adapter injection layer for BERT in pure Python, including a complete fine-tuning loop, and shows how to extend it for production-grade systems.
A hands-on build guide for implementing a LoRA adapter from scratch in PyTorch, including rank-decomposed weight matrices and a complete fine-tuning loop for BERT — designed to signal real ML engineering skill on a portfolio.
Build a production-shaped WordPiece tokenizer in Python with subword merging, BERT special tokens, and a greedy decode path. Includes runnable code, tests, and a roadmap to senior-level extensions.