About the workshop
Mimer AI Factory, in collaboration with Hite, invites you to an interactive workshop on Large Language Models (LLMs), followed by an open help desk for AI compute and expert support.
Unlock the power of large language models (LLMs) with our workshop! Seize the opportunity to receive hands-on guidance and expert advice on scaling and deploying AI workloads.
Learn how to scale fine-tuning across multiple GPUs for better performance and efficiency, making advanced AI more accessible. During the workshop, you will explore:
- Understanding how Large Language Models work
- Tokenization & Embeddings: The building blocks of AI communication
- Fine-tuning: when to (not) do it and how
- Parameter-Efficient Fine-Tuning (PEFT) with LoRA: Optimize with minimal computational effort
- Parallelization techniques: how to take advantage of multiple GPUs with Hugging Face Accelerate
- A dedicated EuroHPC compute access help desk, where you can receive personalized support with your application for EuroHPC computing resources and guidance from AI experts.
Who is this for?
The workshop is aimed at entrepreneurs, participants from start-ups, SMEs, and academia who want to understand the inner workings of LLMs and gain hands-on experience with powerful multi-GPU fine-tuning techniques to optimize their AI workflows for both speed and scalability.
The help desk is open to anyone with an AI use case who needs access to compute resources or expert support.
Key takeaways
Through guided exercises and real-world examples, participants will learn to:
- Fine-tune LLMs efficiently across multiple GPUs
- Execute larger and more complex AI projects
By the end of the workshop, participants will have a foundational understanding of techniques for distributing model parameters and states across multiple GPUs and nodes.
At the help desk, participants will also learn how to apply for EuroHPC computing resources and receive personalized guidance throughout the application process.
Prerequisites
- Basic programming skills in Python
- Previous experience with deep learning models or LLMs is beneficial, but not required
Location:
TBA
Registration:
To ensure that everyone has the opportunity to participate, we kindly request that you let us know as soon as possible if you are unable to attend an event after registering.
Schedule:
TBA