O'Reilly logo
live online training icon Live Online training

Hands-On Deep Neural Networks with PyTorch

Build your own NLP and computer vision models

Robert Alvarez

Over the past couple of years, PyTorch has been increasing in popularity in the Deep Learning community. What was initially a tool used by Deep Learning researchers has been making headway in industry settings.

In this session, we will cover how to create Deep Neural Networks using the PyTorch framework on a variety of examples. The material will range from beginner—understanding what is going on "under the hood," coding the layers of our networks, and implementing backpropagation—to more advanced material on RNNs, CNNs, LSTMs, & GANs.

Attendees will leave with a better understanding of the PyTorch framework. Furthermore, a link to a clean documented GitHub repo with the solutions of the examples covered will be provided.

What you'll learn-and how you can apply it

This training will provide attendees with familiarity with PyTorch and Neural Networks used in Deep Learning. While it will start with basic concepts, it ramps up quickly to more advanced material that is on the cutting edge of what we can do in Deep Learning. It is of interest to anyone who is data curious or interested in applying these models at work.

  • Understand the fundamentals of Neural Networks
  • Build Deep Learning models using the PyTorch library
  • Interpret the output of models and know how to troubleshoot and tune them

This training course is for you because...

  • You want to learn more about Deep Learning and/or PyTorch
  • You want to add additional tools to your machine learning skillset
  • You want to create computer vision and/or natural language models

Prerequisites

  • Working knowledge freshman Calculus concepts
  • Familiarity with the Python programming language, e.g., a first-course or experience as a programmer
  • Students will need access to Google Colab (all web-based, no need to install anything locally)

Recommended Preparation

About your instructor

  • Robert loves to break deep technical concepts down to be as simple as possible, but no simpler.

    Robert has data science experience in companies both large and small. He is currently Head of Data Science for Podium Education, where he builds models to improve student outcomes, and an Adjunct Professor at Santa Clara University’s Leavey School of Business. Prior to Podium Education, he was a Senior Data Scientist at Metis teaching Data Science and Machine Learning. At Intel, he tackled problems in data center optimization using cluster analysis, enriched market sizing models by implementing sentiment analysis from social media feeds, and improved data-driven decision making in one of the top 5 global supply chains. At Tamr, he built models to unify large amounts of messy data across multiple silos for some of the largest corporations in the world. He earned a PhD in Applied Mathematics from Arizona State University where his research spanned image reconstruction, dynamical systems, mathematical epidemiology and oncology.

Schedule

The timeframes are only estimates and may vary according to how the class is progressing

Segment 1: Introduction to PyTorch and Tensors Length (60 mins)

  • Introduction to Tensors
  • PyTorch vs Numpy
  • Automatic Gradients in PyTorch, aka, Autograd
  • Linear Regression
  • Break + Q&A

Segment 2: Building Neural NetsLength (75)

  • Perceptrons and basic building blocks of Neural Nets
  • Overview of Activation Functions
  • Building deeper models for non-linear functions
  • Introduction to Computer Vision models
  • Break + Q&A

Segment 3: Building your own Computer Vision and NLP models Length (60)

  • Intro to CNNs
  • Transfer Learning
  • Building your own CV model on Fashion MNIST datasets
  • RNNs and NLP models

Course wrap-up and next steps