This repository provides a PyTorch implementation of BERT (Bidirectional Encoder Representations from Transformers), Google’s 2018 groundbreaking language model that achieved state-of-the-art results across 17 NLP tasks. It implements the bidirectional transformer architecture with masked language modeling and next sentence prediction pre-training objectives. The code is designed to be simple and annotated for educational purposes, based on the Harvard Annotated Transformer.