Building Deep Learning Models Using PyTorch

PyTorch is an open source deep learning framework originally developed by the AI teams at Facebook. PyTorch offers high-level APIs which make it easy to build neural networks and great support for distributed training and prediction.
Course info
Level
Beginner
Updated
Oct 12, 2018
Duration
3h 18m
Table of contents
Course Overview
Introduction to PyTorch
Building Simple Neural Networks
Building an Image Classification Model
Building a Text Classification Model
Description
Course info
Level
Beginner
Updated
Oct 12, 2018
Duration
3h 18m
Description

PyTorch is an open source, deep learning framework which is a popular alternative to TensorFlow and Apache MXNet. PyTorch APIs follow a Python-native approach which, along with dynamic graph execution, make it very intuitive to work with for Python developers and data scientists. In this course, Building Deep Learning Models Using PyTorch, you will learn to work with PyTorch and all the libraries that it has to offer, from first principles - starting with Torch tensors, dynamic computation graphs, and the autograd library, to compute gradients. You'll start off by understanding the basics of training a neural network, the forward and backward passes, and gradient computation. You will use these concepts to build simple neural networks to predict automobile prices, as well as who survived and who did not on the Titanic. Next, you'll move on to image classification using convolutional neural networks; you'll study the role of convolutional and pooling layers and the basic structure of a CNN, you'll then build a CNN to classify images from the Cifar-10 dataset. You'll also see how you can leverage the power of transfer learning by using pre-trained models for image classification. Finally, you'll get to work with recurrent neural networks for sequence data, seeing how the dynamic computation graph execution in PyTorch makes building RNNs very simple. You'll use RNNs with long memory cells to predict gender using baby names. At the end of this course, you will be comfortable using PyTorch libraries and APIs to leverage pre-trained models that PyTorch offers and also to build your own custom model for your specific use case.

About the author
About the author

A problem solver at heart, Janani has a Masters degree from Stanford and worked for 7+ years at Google. She was one of the original engineers on Google Docs and holds 4 patents for its real-time collaborative editing framework.

More from the author
Building Features from Image Data
Advanced
2h 10m
Aug 13, 2019
Designing a Machine Learning Model
Intermediate
3h 25m
Aug 13, 2019
More courses by Janani Ravi
Section Introduction Transcripts
Section Introduction Transcripts

Course Overview
Hi. My name is Janani Ravi, and welcome to this course on building deep learning models using PyTorch. A little about myself, I have a Masters degree in Electrical Engineering from Stanford and have worked at companies such Microsoft, Google, and Flipkart. At Google, I was one of the first engineers working on a real-time collaborative editing in Google Docs, and I hold four patents for its underlying technologies. I currently work on my own startup Loonycorn, a studio for high quality video content. In this course, you will learn to work with PyTorch and all the libraries that it has to offer from first principles. We start off with torch answers, dynamic computation graphs, and autograd library to compute gradience. We'll start off by understanding the basics of training a neural network, the forward and backward passes, and gradient computation. We will then use these concepts to build simple neural networks to predict automobile prices and who survived and who did not on the Titanic. We'll then move onto image classification using convolutional neural networks. We'll study the rule of convolutional, as well as pooling layers and the basic structure of a convolutional neural network. We'll then build one to classify images from the side, far, and data side. We'll also see how we can leverage the power of transfer learning by using pre-trained models for image classification. We'll then work with recurrent neural networks for sequence data. We'll see how the dynamic computational graph execution in PyTorch makes building odd and ends very simple. We'll use RNNs with long memory cells to predict gender using baby names. At the end of this course, you'll be comfortable using PyTorch libraries and APIs to leverage either the pretrained models that PyTorch has to offer and also to build your own custom models for your specific use cases.