🚀 Overview
Artificial Intelligence and High-Performance Computing (HPC) are transforming the way we solve scientific and industrial problems—from automating data analysis to scaling simulations and enabling generative applications with Large Language Models (LLMs). This 5-day intensive workshop is designed to give you a foundational yet practical introduction to these technologies, equipping you with the skills to apply them in real-world settings.
Through a combination of hands-on tutorials, interactive lectures, and expert-led sessions, you’ll gain a working knowledge of deep learning, HPC systems, and scalable model training. Whether you're new to AI workflows or looking to bring large-scale computing into your domain, this workshop will empower you with tools and understanding you can apply immediately.
🗓️ What You'll Learn
Day 1–2: Introduction to Deep Learning
Get up to speed with the core concepts of neural networks and deep learning. You'll implement and train models using popular frameworks, explore different architectures (CNNs, MLPs), and learn how to structure a machine learning pipeline.
Day 3: Introduction to Large Language Models (LLMs)
Understand how LLMs like GPT work, from tokenization to transformer architectures. You'll learn how to use pretrained models and hosted models using the openai interface. Explore their use cases in science and industry.
Day 4: Introduction to High-Performance Computing (HPC)
Dive into the world of HPC: job schedulers, resource management, and compute environments. You'll learn how to navigate HPC systems effectively, submit jobs, and optimize resource usage for computational workloads.
Day 5: Introduction to Parallel Training of Deep Learning Models
Take your models to the next level with parallel and distributed training techniques. Explore multi-GPU and multi-node training setups using tools like PyTorch Lightning, DeepSpeed, or Horovod, and understand the performance trade-offs.
👥 Who Should Attend
This workshop is ideal for:
-
Scientists looking to integrate AI into research workflows
-
Engineers aiming to scale model training on HPC systems
-
Industry professionals exploring data-driven automation and modeling
-
Anyone with strong Python and Unix fundamentals eager to expand into DL/HPC
🔧 Requirements
Participants must be comfortable with:
-
Writing and debugging Python code (go here for a brush up)
-
Using Unix/Linux shell commands (go here for a brush up)
No prior experience with machine learning, HPC, or LLMs is required. The workshop will be held in presence at HZDR.
📩 How to Apply
Spots are limited to 20 seats. Apply early to secure your place! The registration deadline is Oct 3, 2025.
If accepted, you will be notified early October 2025.
