The Annotated Transformer – Implementing an Attention-Based NLP Model in PyTorch | Hacker News Reader