BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Denver
X-LIC-LOCATION:America/Denver
BEGIN:DAYLIGHT
TZOFFSETFROM:-0700
TZOFFSETTO:-0600
TZNAME:MDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0600
TZOFFSETTO:-0700
TZNAME:MST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260422T000711Z
LOCATION:505
DTSTART;TZID=America/Denver:20231113T140600
DTEND;TZID=America/Denver:20231113T140900
UID:submissions.supercomputing.org_SC23_sess442_ws_whpc117@linklings.com
SUMMARY:fAsyLex: Accelerating Legal NLP through Comparative Analysis of Mu
 lti-GPU Approaches
DESCRIPTION:Claire Barale (University of Edinburgh)\n\nThe primary objecti
 ve of this work is to conduct an evaluation of the acceleration of NLP tra
 ining for the task of text classification on legal documents. The dataset 
 used is AsyLex, a dataset of refugee claims from Canada.  We implement fas
 t AsyLex (fAsylex) and scale it across up to 64 GPUs. Through systematic e
 xperimentation, we seek to address the following research questions: How d
 oes the training time differ between single-GPU and multi-GPU setups for t
 wo commonly used PLMs? Does the choice of training approach (single-GPU vs
 . multi-GPU) influence the classification performance on the chosen datase
 t? We offer an investigation into the practical implications of employing 
 single-GPU and multi-GPU training, we compare two of the most commonly use
 d masked language models, RoBERTa and DeBERTa and reduce runtime out-of-th
 e-box by 49% and 37% respectively, and we demonstrate that there is a trad
 e-off in terms of NLP metrics and distributed training.\n\nTag: State of t
 he Practice\n\nRegistration Category: Workshop Reg Pass\n\nSession Chairs:
  Elsa J. Gonsiorowski (Lawrence Livermore National Laboratory (LLNL)) and 
 Mozhgan Kabiri chimeh (NVIDIA Corporation)\n\n
END:VEVENT
END:VCALENDAR