This course provides a comprehensive introduction to Large Language Models (LLMs), focusing on what they are, how to build them using PyTorch, and how to use them for inference in language tasks. Participants will learn about the history of LLMs, how LLMs fit into the larger AI/Generative AI landscape, neural-network-based language models, and how to use RNNs, LSTMs, and transformers for natural language processing tasks. 1) Introduction to NLP What is NLP? NLP Basics: Text Preprocessing and Tokenization NLP Basics: Word Embeddings Introducing Traditional NLP Libraries A brief history of modeling language Introducing PyTorch and HuggingFace for Text Preprocessing Neural Networks and Text Data Building Language Models using RNNs and LSTMs 2) Transformers and LLMs Introduction to Transformers Using Hugging Face's Transformers for inference LLMs and Generative AI Current LLM Options Fine tuning GPT Aligning LLMs with Human Values Retrieval-Augmented Generation (RAG) Systems
