Graham Neubig Courses

Advanced NLP - Experimental Design and Data Annotation - Lecture 9

1 rating

Learn experimental design and data annotation techniques for advanced NLP, covering key principles and best practices for conducting rigorous research and creating high-quality datasets.

• YouTube
• 1 hour 17 minutes
• Self-Paced
• Free Video

CMU LLM Inference - Beam Search and Variants

Explore beam search algorithms and their variants for LLM inference, examining the inadequacies of mode-based approaches in natural language processing applications.

• YouTube
• 1 hour 12 minutes
• Self-Paced
• Free Video

Common Sampling Methods for Modern NLP - CMU LLM Inference 3

Explore common sampling methods for modern NLP and understand diversity-quality tradeoffs in language model inference through practical examples and analysis.

• YouTube
• 1 hour 2 minutes
• Self-Paced
• Free Video

Probability Review and Code Examples for LLM Inference - Lecture 2

Dive into probability fundamentals, transformer implementation, and generation techniques for advanced NLP with practical code examples and meta-generation concepts.

• YouTube
• 1 hour 12 minutes
• Self-Paced
• Free Video

Introduction to Language Models and Inference - CMU LLM Inference 1

Dive into language model fundamentals, transformer architecture, and inference algorithms while exploring modeling and search errors in modern LLM systems.

• YouTube
• 1 hour 13 minutes
• Self-Paced
• Free Video

Reward Models and Best-of-N - CMU LLM Inference 12

Explore reward models and best-of-n sampling techniques for optimizing LLM outputs, plus Monte Carlo Tree Search methods in this advanced NLP lecture.

• YouTube
• 53 minutes
• Self-Paced
• Free Video

Agents and Multi-Agent Communication - Lecture 11

Explore agent architectures, multi-agent systems, and safety challenges in LLM inference with efficiency optimizations and context management techniques.

• YouTube
• 1 hour 1 minute
• Self-Paced
• Free Video

Self-Refine and Self-Correction Methods - CMU LLM Inference 8

Explore self-refine techniques and iterative refinement methods for LLMs, including self-debugging, verbal reinforcement learning, and tool-interactive critiquing approaches.

• YouTube
• 42 minutes
• Self-Paced
• Free Video

Reasoning Models in Large Language Model Inference - Lecture 9

Explore advanced reasoning models in LLMs, covering reinforcement learning training, STaR methodology, DeepSeek R1, chain-of-thought reasoning, domain transfer, and cutting-edge algorithms.

• YouTube
• 1 hour 6 minutes
• Self-Paced
• Free Video

Incorporating Tools in Large Language Model Inference - Lecture 10

Explore advanced techniques for integrating tools with large language models, covering tool use paradigms, creation methods, robustness evaluation, and secure execution frameworks.

• YouTube
• 53 minutes
• Self-Paced
• Free Video

Other Controlled Generation Methods - CMU LLM Inference 6

Explore advanced techniques for controlling LLM text generation, including decoding-time distributional modifiers and other sophisticated methods for guided output.

• YouTube
• 1 hour 4 minutes
• Self-Paced
• Free Video

Chain of Thought and Intermediate Steps - CMU LLM Inference 7

Explore chain of thought reasoning in LLMs, understanding why intermediate steps improve inference and learning self-consistency techniques for better model performance.

• YouTube
• 55 minutes
• Self-Paced
• Free Video

A* and Best First Search for LLM Inference - Lecture 5

Explore A* and best first search algorithms for LLM inference, covering beam search variants and addressing mode inadequacies in advanced natural language processing.

• YouTube
• 1 hour
• Self-Paced
• Free Video

Advanced Natural Language Processing - Fall 2024

Master advanced NLP techniques from transformers to LLM agents, covering pre-training, instruction tuning, RLHF, RAG, and cutting-edge inference algorithms in this comprehensive CMU series.

• YouTube
• 1 day 3 hours 10 minutes
• Self-Paced
• Free Video

CMU Advanced Natural Language Processing Spring 2024

Master advanced NLP techniques from transformers to LLMs, covering prompting, fine-tuning, RAG, reinforcement learning, code generation, and multilingual processing in this comprehensive series.

• YouTube
• 1 day 1 hour 26 minutes
• Self-Paced
• Free Video