Contribution to a conference proceedings PUBDB-2023-07212

http://join2-wiki.gsi.de/foswiki/pub/Main/Artwork/join2_logo100x88.png
PACuna: Automated Fine-Tuning of Language Models for Particle Accelerators

 ;  ;  ;

2023

NeurIPS 2023 workshop on Machine Learning and the Physical Sciences, NeuralIPS2023, New OrleansNew Orleans, USA, 15 Dec 2023 - 15 Dec 20232023-12-152023-12-15 7 pp. ()  GO

Abstract: Navigating the landscape of particle accelerators has become increasingly challenging with recent surges in contributions. These intricate devices challenge comprehension, even within individual facilities.To address this, we introduce PACuna, a fine-tuned language model refined through publicly available accelerator resources like conferences, pre-prints, and books.We automated data collection and question generation to minimize expert involvement and make the code available.PACuna demonstrates proficiency in addressing accelerator questions validated by experts.Our approach shows adapting language models to scientific domains by fine-tuning technical texts and auto-generated corpora capturing the latest developments can further produce pre-trained models to answer some specific questions that commercially available assistants cannot and can serve as intelligent assistants for individual facilities.


Contributing Institute(s):
  1. Beschleunigerkontrollen (FLASH/XFEL) (MCS 4)
Research Program(s):
  1. 621 - Accelerator Research and Development (POF4-621) (POF4-621)
Experiment(s):
  1. Facility (machine) XFEL

Appears in the scientific report 2023
Click to display QR Code for this record

The record appears in these collections:
Private Collections > >DESY > >M > >MCS > MCS 4
Document types > Events > Contributions to a conference proceedings
Public records
Publications database

 Record created 2023-11-27, last modified 2023-11-28


Restricted:
Download fulltext PDF Download fulltext PDF (PDFA)
Rate this document:

Rate this document:
1
2
3
 
(Not yet reviewed)