BEGIN:VCALENDAR
VERSION:2.0
PRODID:icalendar-ruby
CALSCALE:GREGORIAN
BEGIN:VEVENT
DTSTAMP:20260705T095302Z
UID:575aa2e3-0165-4805-8a71-84555ba2eaa6
DTSTART:20210201T080000Z
DTEND:20210205T121500Z
DESCRIPTION:In this course\, we will cover machine learning and deep learni
 ng and how to achieve scaling to high performance computing systems. The c
 ourse aims at covering all levels\, from fundamental software design to sp
 ecific compute environments and toolkits. We want to enable the participan
 ts to unlock the resource of machines like the JUWELS booster for their ma
 chine learning workflows. Different from previous years we assume that the
  participants have a background from a university level introductory cours
 e to machine learning. Suggested options for self-teaching are given below
 .\n\nWe will start the course with a presentation of high performance comp
 uting system architectures and the design paradigms for HPC software. In t
 he tutorial\, we familiarize the users with the environment. Furthermore\,
  we give a recap of important machine learning concepts and algorithms and
  the participants will train and test a reference model. Afterwards\, we i
 ntroduce how deep learning algorithms can be parallelized for supercompute
 r usage with Horovod. Furthermore\, we discuss best practicies and pitfall
 s in adopting deep learning algorithms on supercomputers and learn to test
  their function and performance. Finally we apply the gained expertise to 
 large scale unsupervised learning\, with a particular focus on Generative 
 Adversarial Networks (GANs).\n\nPrerequisites:\n\nWe assume that the parti
 cipants are familiar with general concepts of machine learning and/or deep
  learning\, such as widely used models\, losses\, regularization and basic
  model training / testing. Many excellent self-training resources are avai
 lable such as:\n\n\n	The MIT introduction to Deep Learning Course (http://
 introtodeeplearning.com/)\n	The Machine Learning course and Deep Learning 
 specialization by Andrew Ng et al. at Stanford (https://cs230.stanford.edu
 /) and on Coursera (www.coursera.org)\n	The notebook-based courses of fast
 .ai (www.fast.ai) and of Master Datascience Paris Saclay (https://github.c
 om/m2dsupsdlclass/lectures-labs)\n	Ian Goodfellows book on Deep Learning (
 https://www.deeplearningbook.org/)\n\n\nHands-on experience with ML/DL fra
 mework is required\, first experience with HPC systems is helpful.\n\nLear
 ning outcome:\n\nAfter this course\, participants will be able to parallel
 ize Tensorflow and Pytorch ML workflows on HPC machines\, taking into acco
 unt the HPC system architecture and circumventing typical pitfalls and bot
 tlenecks.\n\nDate:\n\n1-5 February 2021\, 9:30 - 13:00\n\nInstructors:\n\n
 Dr. Stefan Kesselheim\, Dr. Jenia Jitsev\, Roshni Kamath\, Dr. Mehdi Certi
 \, Dr. Alexandre Strube\, Jan\, Ebert\, Jülich Supercomputing Centre\nhtt
 ps://events.prace-ri.eu/event/1136/
SUMMARY:[ONLINE] Introduction to Scalable Deep Learning @ JSC
URL;VALUE=URI:https://events.prace-ri.eu/event/1136/
END:VEVENT
END:VCALENDAR