Lectures on Modern Scientific Programing 2024

Wednesday 20 Nov 2024, 08:40 → 20:30 Europe/Budapest

Building 3, 2nd floor, Council Room (HUN-REN Wigner RCP )

Lectures on Modern Scientific Programming 2024

 

The Lectures on Modern Scientific Programming series will be organised by the Wigner Scientific Computing Laboratory (former GPU Laboratory)  from November 20, 2008 for the 8th time at the HUN-REN Wigner Research Centre for Physics. 

An annually organized seminar series for university students and researchers. These lectures review topics relevant to everyone working on scientific simulations or high performance computing. 

The lectures will be a mixture of theoretical talks and hands-on sessions, which  grants a perfect opportunity to adopt the collected theoretical knowledge. In 2024 the lectures will focus on Large Language Models and their application on physics and other relevant fields.  
 

The lectures are highly recommended to those who haven't got the chance to get to know the basics of these topics and their fields of usage. Invited performers will present these.

For the earlier events see: 2023, 2022, 2019, 2018,2017,2016,2015

 

Confirmed speakers: 
Natabara Máté Gyöngyössy (ELTE IK)
Marcell Stippinger (HUN-REN Wigner RCP)

 

The conference will be held offline form, therefore there are limited places to participate personally on-site. We encourage our former, current, and future partners to participate in the conference. 

More information is available on the website!

Useful pieces of information for sponsors.

Patrons of the Workshop:
Péter Lévai (Wigner RCP,  Member of HAS) and István Csabai (ELTE, Member of HAS)

---

Organizers:
Gergely Gábor Barnaföldi
Gábor Bíró
Balázs Kacskovics

08:40 → 10:30 Lectures

08:40 Openning

Speaker: Gergely Barnafoldi (Wigner RCP RMI of the Hungarian Academy of Sciences)

GGBarnafoldi_Lectures_WSCLAB_2024.pdf

09:00 Introduction to Language Modeling

The task of predicting the next word in a sentence is called language modelling. In the early times of machine learning, this task was important for machine translation. The first talk also tells what world embeddings of shallow neural networks knew about the relations between the lexical meanings of the words. One of the main components of the deep neural language models also came from translation: a model of how words ,,attend" to each other.

Speaker: Márton Makrai

makrai-sci-comp-wigner-24.pdf

10:30 → 11:00 Coffee Break

11:00 → 12:30 Lectures

11:00 LLM Applications

A high-level review of the current state of adapting LLMs: fine tuning, prompting, RAG. Including a practical overview of LLM parameters (such as size, limitations, most widespread parametrization, etc).

Speakers: Dávid Nemeskey, Natabara Máté Gyöngyössy (ELTE Eötvös Loránd University)

LLM Applications LMSC 2024

LLM Applications LMSC 2024.pptx

12:30 → 14:00 Lunch Break

14:00 → 15:30 Hands-on Session

14:00 LLMs in Scientific Coding

Usage of LLMs in scientific coding lecture + demonstrations

Speaker: Marcell Stippinger (HUN-REN Wigner FK RMI Komputációs Tudományok Osztálya)

Data repository

LLM_in_programming_Lectures.pdf

15:30 → 16:00 Coffee Break

16:00 → 18:00 Hands-on Session

16:00 Scientific Coding Agents in Practice

Taskwiever-based FTDT tooling

Speaker: Natabara Máté Gyöngyössy (ELTE Eötvös Loránd University)

LMSC_Workshop_presentation.pdf