9 October 2023
ONLINE
Europe/Vienna timezone

Start date:  09 Oct 2023, 15:00 (CEST)     Entry level:  Basic
End date:  09 Oct 2023, 16:30 (CEST)     Subject area:  Machine learning
Location:  ONLINE (Zoom)     Topics:  Large language models
Language:  English     Target audience:  Industry, academia, public sector
Price:  Free (for eligible participants)     Organisers:  EuroCC Austria & VSC

Overview

This webinar serves to demystify the current trend on large language models (LLMs) and to point out challenges and chances of creating a solution tailored to your needs. We will demonstrate how to fine-tune an LLM and offer an opportunity for developers and researchers to access European supercomputers to train their models free of charge.

Agenda 

15:00  

 

 

 

Welcome  
Eva Gergely 
Project manager and team lead at Vienna University Computer Center, dept. IT-Support for research (Vienna, Austria)

 

15:05 

About Large Language Models (LLMs)             
Simeon Harrison           
Trainer for HPC/HPDA/AI at EuroCC Austria and VSC Research Center, TU Wien             
(Vienna, Austria)

  • What are Large Language Models (LLMs) and what are the typical use cases?
  • Fine-tuning a pretrained model with own data
  • Examples of open-source pretrained models
  • LLMs as part of even larger ML models
  • Risks of using existing LLM models
   
15:20 

Trustworthy AI – EU compliance              
Dr. Sabine T. Koeszegi           
Professor for Labor Science and Organization at the TU Wien, Member of the Advisory Board on Ethics in Artificial Intelligence of the UNESCO Commission Austria             
(Vienna, Austria)

  • Ethics guidelines on trustworthy AI of the European Commission
  • Cornerstones of the European Draft Regulation
  • Assessment tool ALTAI and IEEE standard 7000 on value-based design
   
15:35 

Customising LLMs to non-English languages              
Bryan Cardenas Guevara           
High-Performance Machine Learning Consultant at SURF             
(Utrecht, the Netherlands)

  • The challenges of training an LLM for the Dutch language
  • How the team dealt with a lack of compute power and a lack of data
  • Why it's important to train LLMs in non-English languages and the role of the Dutch model in the Netherlands
   
15:50 

GPUs for your LLMs: train your model on a supercomputer             
Simeon Harrison

  • EuroCC helps you get access to supercomputers to train your models
  • Introducing Vienna Scientific Cluster (VSC), Austria's flagship supercomputer
  • Freely available compute power in Europe
  • Supercomputing Accelerator - one-stop HPC support for startups and SMEs
   
16:00 

How to fine-tune an LLM – short demo             
Dr. Soner Steiner           
HPC expert at EuroCC Austria and VSC Research Center, TU Wien             
(Vienna, Austria)

  • How to use the VSC-5 cluster for training/fine-tuning
   
16:25 Q&A + open end

Target audience

Webinar for academia, industry, public administration, especially developers and researchers interested in AI applications and legal aspects of using LLMs.

Webinar format

This webinar will be delivered as a LIVE ONLINE WEBINAR (using Zoom).

Upon joining, the participants will be automatically muted and their cameras switched off. However, we would be happy to hear your thoughts and questions during or after the presentations.

Prices and eligibility

This webinar is funded by the EuroCC 2 project. Therefore, the webinar is open and free of charge for participants from industry, academia and public administration from the Member States (MS) of the European Union (EU) and Associated/Other Countries to the Horizon 2020 programme.

Registration

Register here

After the registration you will get an email with the Zoom link (subject starting with "[Indico] Registration"), please check your spam folder. We will also send you a short reminder before the webinar starts.


Special offer

>> Need support in accelerating your computing tasks? 

>> Looking for funding opportunities for projects in the fields of AI / Data Analytics / HPC?

Apply for the Supercomputing Accelerator – free service package for startups and companies:

About: https://www.f6s.com/eurocc/about 

Apply: https://www.f6s.com/eurocc/apply  


Organisers

    

EuroCC Austria National Competence Centre for Supercomputing, Big Data Analytics and Artificial Intelligence helps businesses, academia and the public sector to adopt and leverage advanced computing technologies. It is coordinated by Advanced Computing Austria (ACA) with the University of Vienna, TU Wien and the University of Natural Resources and Life Sciences (BOKU), in close cooperation with business incubator INiTS.  

VSC Research Center, TU Wien is a collaboration of Austrian universities that provides supercomputing resources and support to scientists and researchers.


Acknowledgements

This webinar is funded by the EuroCC 2 project.

The project has received funding from the European High-Performance Computing Joint Undertaking (JU) under grant agreement No 101101903. The JU receives support from the Digital Europe Programme and Austria, Belgium, Bulgaria, Croatia, Cyprus, Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Latvia, Lithuania, Luxembourg, Montenegro, Netherlands, North Macedonia, Norway, Poland, Portugal, Romania, Serbia, Slovakia, Slovenia, Spain, Sweden, and Türkiye.

Additional funding for the project comes from the Austrian federal ministries BMBWF and BMK.


Starts
Ends
Europe/Vienna
ONLINE
Zoom