Introduction to High-Performance Machine Learning @SURFsara

CET
VK1/VK2 (SURFsara)

VK1/VK2

SURFsara

Science Park 140, 1098 XG Amsterdam
Description

In recent years machine learning and deep learning techniques in particular have developed tremendously. Neural networks are being used in more and more application domains going from computer vision to speech recognition, and even replacing parts of the compute pipeline for scientific HPC applications.

 

In this course you will start from the essential concepts up to the efficient use of HPC infrastructures to get the best performance out of different machine learning tools. Several hands-on sessions are set up to present general algorithms and some scalability challenges involved in when using both large-scale data and large-scale models.

 

IMPORTANT INFORMATION: WAITING LIST

If the course gets fully booked, no more registrations are accepted through this website. However, you can be included in the waiting list: for that, please send an email to training@surfsara.nl and you'll be informed when a place becomes available.

    • Welcome & Introduction
      Convener: Zheng Meyer-Zhao (SURFsara)
    • Introduction to Neural Networks
    • 1
      Hands-on: Neural Networks
    • 10:30 AM
      Coffee break
    • Neural Networks - knobs and dials
    • Hands-on: Neural Networks - parameter tuning
    • 12:00 PM
      Lunch break
    • Convolutional Neural Networks
    • Hands-on: Convolutional Neural Networks
    • 2:30 PM
      Coffee break
    • 2
      Interpreting deep learning models and results
    • 3
      Inspecting a Convolutional Neural Network
    • Open discussion
  • Wednesday, June 12
    • Parallel computing for Deep Learning
    • Hardware bottlenecks (and how to overcome them)
    • 10:30 AM
      Coffee break
    • Hands-on: Profiling: creation and interpretation
    • 4
      Frameworks
    • 12:00 PM
      Lunch break
    • Hands-on: Hovorod
    • Hands-on: CNN data distributed with Cifar 10
    • 2:30 PM
      Coffee break
    • Hands-on: CNN data distributed with Cifar 10
    • Hybrid parallelism & Mesh tensorflow
    • Hands-on: Hybrid parallelism
    • Open discussion