Learn PyTorch Fundamentals With Microsoft Learn
Author: Cassie Breviu, Cloud Developer Advocate at Microsoft
We previously announced how we have been collaborating with PyTorch to create content and resources for the community recently with the new PyTorch Tutorial “Learn the Basics.” We had so much fun creating that content and we thought, why stop there?! We now have a new Learning Path on Microsoft Learn to build on the Learn the Basics and provide a guided learning experience with cloud GPU compute on integrated Jupyter notebooks.
The Learning Path: PyTorch Fundamentals
In the PyTorch Fundamentals learning path we are focusing on PyTorch as the library to learn how to do deep learning in multiple domains. This is the first Learning Path in Microsoft Learn to include integrated notebooks that allow you to run the tutorial directly in the browser. The environment has been preconfigured for data science. All of the needed libraries to run the code are already installed and configured. This allows you to read the lessons and then run the code cell right after to apply the concept. You will notice the integrated notebooks come packed with new features to help you build. Things like intellisense, code completion, dark theme, rich outputs, interactivity, along with many of the standard features of notebooks. This all combined makes a powerful, adaptive and enriched guided learning experience.
There are four different learning modules in the path:
The first module is an Introduction to PyTorch and is verbatim to the Learn the Basics tutorial on PyTorch.org. From there we want to build on the basics of machine learning with PyTorch and start understanding other machine learning domains.
The second module focuses on an introduction to Computer Vision. It uses an image classification task to learn about convolutional neural networks, and then see how pre-trained networks and transfer learning can improve our models and solve real-world problems.
The third module focuses on Natural Language Processing. It will explore different neural network architectures for dealing with natural language texts. In the recent years, Natural Language Processing (NLP) has experiences fast growth as a field, primarily because performance of the language models depends on their overall ability to “understand” text, and that can be trained in unsupervised manner on large text corpora. Thus, pre-trained text models such as BERT simplified many NLP tasks, and dramatically improved the performance.
The fourth and final module focuses on Audio Classification. There are multiple ways to build an audio classification model. You can use the waveform, tag sections of a wave file, or even use computer vision on the spectrogram image. In this module we will first break down how to understand audio data, from analog to digital representations, then we will build the model using computer vision on the spectrogram images. That’s right, you can turn audio into an image representation and then do computer vision to classify the word spoken!
After completing this learning path you will have a better understanding of how to apply PyTorch to solve real world problems within these domains.
PyLadies Collaboration
We had the wonderful opportunity to collaborate with the PyLadies, an international mentorship group with a focus on helping more women become active participants and leaders in the Python open-source community. We partnered with New York City Chapter of PyLadies to do a month-long event with this content; each week we focused on a module in the learning path. The PyLadies connected with Cloud Advocates each week to hear about the learning path and also provided valuable feedback on the content.
The NYC PyLadies community completed the PyTorch learning module on Microsoft Learn with great success! We were super excited to do this collaboration with Microsoft so we can provide our community with relevant tools to learn the fundamentals of deep learning. The Cloud Advocacy team answered all of our questions and made our study sessions both fun and informative. Their commitment to open-source communities is strongly evident. We were very pleased with their support throughout our four-part study series and would be happy to do it again for the next learning journey! — Felice Ho, PyLadies Organizer, @NYCPyLadies
Stay Tuned and Continue Learning
You can continue your PyTorch journey by visiting the PyTorch tutorials on PyTorch.org. Find the PyTorch Fundamentals Learning path at https://aka.ms/PyTorch/MSLearn. Be sure to check out all the ways that PyTorch and Azure are better together. Stay tuned for more Microsoft ♥PyTorch collaborations!
