Overview

Instructor: HPRC: Richard Lawrence, Zhenhua He; Intel: Aaryan Kothapalli, Yuning Qiu

Time: Tuesday, October 25, 2022 and Tuesday, November 1, 2022 — 10:00AM-12:30PM CT and 1:30PM-4:00PM CT

Location: online using Zoom

Prerequisites: Current ACCESS ID or HPRC account, experience with python

In this two-day workshop, you will learn about several components available inside the oneAPI Toolkit for AI/ML applications.

Each day will have a morning session with introductory level material and an afternoon with intermediate level material.

Intel oneAPI Toolkit is a core set of tools and libraries for developing high-performance, data-centric applications across diverse architectures. It simplifies the implementation of HPC applications on CPUs and accelerators with Intel's compiler technology and libraries.

Participants completing these sessions will receive a certificate from Intel.

If you are not a member of TAMU, you will need to provide your ACCESS ID. If you do not have an ACCESS ID, please register at https://identity.access-ci.org/new-user . The ACES project is supported by a National Science Foundation (NSF) award number 2112356.

Course Materials

The presentation slides will be available as downloadable PDF files.

• ACES: Intel Data Science for Python PDF
• ACES: Intel Classical ML Optimizations PDF
• ACES: Introduction to Deep Learning PDF

Agenda

There are two days a week apart, each with a morning and afternoon session (10:00 a.m. - 12:30 p.m., 1:30 p.m. - 4:00 p.m.):

Day 1 : Classical Machine Learning

morning, introductory level (Leading: Richard Lawrence of HPRC)

• Getting Started with ACES
• Jupyter Notebook Environment
• Data Structure with Pandas
• Machine Learning with Scikit Learn
• Machine Learning with XGBoost
• Presentation slides:
  • Data Science for Python PDF

afternoon, advanced level (Leading: Aaryan Kothapalli of Intel)

• Intel AI Toolkit Overview
• Intel Distribution for Python Overview
• Intel Distribution of Modin
• Intel Extension for Scikit-Learn and XGBoost
• Presentation slides:
  • Classical ML Optimizations PDF
• Exercises and Q&A
• Exercise: Modin on Pandas – Getting Started
• Exercise: SKLearn - Getting Started
• Exercise: XGBoost – Getting Started

Day 2: Deep Learning

morning, introductory level (Leading: Zhenhua He of HPRC)

• Introduction to Deep Learning
  • What is deep learning and why is it important
  • Convolution
  • Pooling
  • Convolutional Neural Networks
  • Presentation Slides:
    • Introduction to Deep Learning PDF
• Introduction to TensorFlow
  • Build a Handwritten Digits classifier with TensorFlow
• Introduction to PyTorch
  • Build a Handwritten Digits classifier with PyTorch

afternoon, advanced level (Leading: Yuning Qiu of Intel)

• Presentation Slides:
  • Intel AI Toolkit Deep Learning PDF
• Intel® Optimizations for TensorFlow
  • Exercise on the use of Intel Optimization for Tensorflow
  • Exercise on the performance analysis (Optional)
• Intel® Optimization for PyTorch
  • Exercise on the use of Intel Optimization for PyTorch
• Intel® Neural Compressor
  • Exercise on the use of Intel Neural Compressor for model quantization
• Intel® Optimizations for XGBoost – GPU (Optional)
  • Exercise

See: https://hprc.tamu.edu/aces/

Note: During the class sessions many aspects of the material will be illustrated live via a login to a training system. Attendees are welcome to follow these parts with their own computers. HPRC users may need to use VPN if off campus. ACES users will need to use the jump host to connect.