Deep Learning for NLP

Large language models (LLMs) power much of modern natural language processing, from chat assistants to coding tools to agentic systems. In this course, students learn how these models actually work: how they are built, adapted, evaluated, and deployed in industry, and the design trade-offs at each stage. Through implementation-focused assignments, students build and train language models. Along the way, course covers the foundations of language modeling, tokenization, transformer architectures, pretraining and scaling, efficient training and inference, evaluating LLMs, and post-training methods such as supervised fine-tuning, reinforcement learning, and preference alignment. (Formerly Deep Learning for Natural Language Processing.)