Back to the calendar
Apr
25
[ONLINE] Upscaling A.I. workflows @ENCCS
By PRACE

Overview

Supervised training of large networks requires large labeled datasets, which in turn demand high computational costs. While active practitioners in deep learning primarily develop and train their networks on local computing devices, with the increase of networks complexity, there is an urgent need to create, train, and test models on clusters.

In this workshop, we overview the basics of Docker and Singularity. (Working knowledge of Singularity as given in the workshop (https://www.uppmax.uu.se/support/courses-and-workshops/singularity-workshop-announcement) is desirable.) Distributed training using TensorFlow and Horovod frameworks on a supercomputer will be covered. Moreover, it will be shown how to use Singularity containers in conjunction with TensorFlow and Horovod to upscale an AI app.

The workshop will be entirely online using zoom.

Outcomes

  • Create, deploy, and update containers locally on a supercomputer
  • Upscale the transfer learning of an NLP model in TensorFlow
  • Upscale the transfer learning of an NLP model using Horovod
  • Upscale the transfer learning of a containerized NLP model

Prerequisites

Basic knowledge of UNIX OS and familiarity with NNs are required.

For questions regarding this event please contact us at training@enccs.se.


This training is intended for users established in the European Union or a country associated with Horizon 2020. You can read more about the countries associated with Horizon2020 here https://ec.europa.eu/info/research-and-innovation/statistics/framework-programme-facts-and-figures/horizon-2020-country-profiles_en

Date
25 April, 2022 - 26 April, 2023
Time
7:00 - 10:00
Online
This is an online event at online
Primary Event Page
Project
Organizer