23. About Workshops @ Kempner

Welcome to the Workshops @ Kempner section of our Computing Handbook. This section is dedicated to fostering continuous learning and professional development through a variety of interactive sessions. Our workshops are designed to equip participants with practical skills and knowledge, encouraging active engagement and collaboration. Whether you are a beginner eager to acquire new competencies or a seasoned professional aiming to enhance your expertise, our training programs offer valuable opportunities to advance your proficiency and contribute effectively to your field.

23.1. List of Workshops

Workshop Name	Description
Introduction to the Kempner AI Cluster	Overview of how to access and use the Kempner AI cluster.
Introduction to Distributed Computing	Introduction to key concepts in distributed computing.
Large Language Model Distributed Training	Reviews parallelization techniques including Distributed Data Parallelism (DDP), Model Parallelism (MP), Tensor Parallelism (TP), Pipeline Parallelism (PP), and Fully Sharded Data Parallelism (FSDP). Provides hands-on examples for each approach.
Building a Transformer from Scratch	Provides a practical, interactive way to learn about transformers by building a simple language model.
Large Language Model Distributed Inference	Provides hands-on training on hosting and running inference for large langauge models that don’t fit into a single GPU’s memory.
Spike Sorting on an HPC Cluster	Describes how to implement spike sorting algorithms for neural data on an HPC cluster, using a comprehensive pipeline and interactive examples.
Optimizing ML Workflows on an AI Cluster	Demonstrates how to optimize machine learning workflows for efficient, reproducible training on an AI cluster.