BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Chicago
X-LIC-LOCATION:America/Chicago
BEGIN:DAYLIGHT
TZOFFSETFROM:-0600
TZOFFSETTO:-0500
TZNAME:CDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0500
TZOFFSETTO:-0600
TZNAME:CST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20181221T160907Z
LOCATION:C143/149
DTSTART;TZID=America/Chicago:20181111T133000
DTEND;TZID=America/Chicago:20181111T170000
UID:submissions.supercomputing.org_SC18_sess262_tut175@linklings.com
SUMMARY:Tools and Best Practices for Distributed Deep Learning with Superc
 omputers
DESCRIPTION:Tutorial\nDeep Learning, Machine Learning, Tutorial Reg Pass\n
 \nTools and Best Practices for Distributed Deep Learning with Supercompute
 rs\n\nXu, Zhang, Walling\n\nThis tutorial is a practical guide on how to r
 un distributed deep learning over distributed compute nodes effectively. D
 eep Learning (DL) has emerged as an effective analysis method and has been
  adapted quickly across many scientific domains in recent years.  Domain s
 cientists are embracing DL as both a standalone data science method, as we
 ll as an effective approach to reducing dimensionality in the traditional 
 simulation. However, due to its inherent high computational requirement, a
 pplication of DL is limited by the available computational resources. \n\n
 Recently, we have seen the fusion of DL and HPC: supercomputers show an un
 paralleled capacity to reduce DL training time from days to minutes; HPC t
 echniques have been used to speed up parallel DL training. Therefore distr
 ibuted deep learning has great potential to augment DL applications by lev
 eraging existing high performance computing cluster. This tutorial consist
 s of three sessions. First, we will give an overview of the state-of-art a
 pproaches to enabling deep learning at scale. The second session is an int
 eractive hands-on session to help attendees running distributed deep learn
 ing with resources at the Texas Advanced Computing Center. In the last ses
 sion, we will focus on the best practices to evaluate and tune up performa
 nce.
URL:https://sc18.supercomputing.org/presentation/?id=tut175&sess=sess262
END:VEVENT
END:VCALENDAR