Introducing PyTorch Learn the Basics Tutorial

Authored by Cassie Breviu, Microsoft Cloud Developer Advocate

We have collaborated with PyTorch and the PyTorch community to create a new tutorial to help new and experienced machine learning practitioners get started with PyTorch. We are excited to announce the new machine learning tutorial that is now available! We love contributing to open source and this collaboration was a blast! Let’s take a look at the journey that got us here and learn a bit more about the tutorial that was created.

First let’s talk about our “Learn the Basics” tutorial co-authored with PyTorch that aims to help both developers new to machine learning and machine learning practitioners new to PyTorch get started. Writing for many levels of expertise is no small feat and we knew we had our work cut out for us. A challenge as Developer Advocates that is not uncommon nor a bridge we have not crossed before. We first discussed different approaches to the format and structure of the tutorial. We wanted to make something accessible for everyone. We decided that an overview of each step in the machine learning workflow for a computer vision model would be best. That way if you already know and understand machine learning on the “Quickstart” page you get a quick answer to how to do it with PyTorch. These quick answers are what developers are often looking for when looking up how to perform a particular function for a solution they are building. It is common for developers to scroll through text to get to the code. That is why the code is right there quickly without over explaining.

I know what you are thinking, what if they don’t know or understand machine learning? If you are new to machine learning and need more than quick code examples and links to docs we still got you covered. We came up with the idea to have a deep dive zoomed in look for each step in the machine learning workflow. If you are new you can jump into the step-by-step guide with a more handheld approach to building models with PyTorch. We have deep dives with more sample code and links to the docs for each concept we introduce in our tutorial. The topics covered in the deep dive sections are: Tensors, Datasets and DataLoaders, Transforms, Build the Model, Autograd, Optimization and Save/Load the Model. Each section includes a breakdown of the concept and how it was applied to our computer vision model. We provide tons of code examples, downloadable notebooks, and links to github to make sure you can build along with us at every step.

The authors for “Learn the Basics” tutorial are Suraj Subramanian a Developer Advocate at PyTorch, Seth Juarez a Principal Cloud Developer Advocate at Microsoft, Cassie Breviu a Cloud Developer Advocate at Microsoft, Dmitry Shoshnikov Senior Cloud Advocate at Microsoft and Ari Bornstein Head of Developer Advocacy at PyTorch Lightning. Advocacy focuses on the developer success and how can we as advocates, help you stay up to date in your skills and understand all the best tools that are out there to help you. That is exactly what this tutorial is; a tool to help you get started with PyTorch.

We have more cool stuff we are working on and contributing to the PyTorch community. Watch for more exciting announcements in the near future, we can’t wait to share them with you! Go check out the tutorials on PyTorch.org and get started building machine learning models with PyTorch!