PyTorch Mobile Machine Learning Framework Announced

On Thursday the developers of PyTorch announced PyTorch Mobile, which they say will allow for "end-to-end workflow from Python to deployment on iOS and Android."

PyTorch Mobile is part of PyTorch 1.3, which currently is an "experimental release" that the organization will be "building on over the next couple of months." PyTorch 1.2 was released in August.

New features coming will include preprocessing and integration APIs, support for ARM CPUs and QNNPACK (a quantized neural network package designed for PyTorch), build-level optimization, and performance enhancements for mobile CPUs/GPUs.

Android builds will use the Maven plug-in and iOS will use CocoaPods with Swift.

There are currently quickstart Hello World guides available for Android and iOS for developers who want to get their hands on the project now.

The open source PyTorch machine learning framework -- particularly popular for deep learning academic projects -- competes mainly with Google's open source TensorFlow framework for the hearts and minds of machine learning developers. TensorFlow offers TensorFlow Lite for IoT and mobile devices. In fact, in The Gradient's 2019 machine learning framework study, released Thursday, author Howard He cites PyTorch's lack of mobile support as one reason TensorFlow remains more popular in production/industry environments (versus academia). That study can be found here.

The organization released a tool for working with large graphing projects within PyTorch in July.

About the Author

Becky Nagel is vice president of AI for 1105 Media, where she specializes in training internal and external customers on maximizing their business potential via a wide variety of generative AI technologies as well as developing cutting-edge AI content and events. She's the author of "ChatGPT Prompt 101 Guide for Business Uses," regularly leads research studies on generative AI business usage, and serves as the director of AI Boardroom, a new resource for C-level executives looking to excel in the AI era. Prior to her current position she was a technical leader for 1105 Media's Web, advertising and production teams as well as editorial director for a suite of enterprise technology publications, including serving as founding editor of She has 20 years of enterprise technology journalism experience, and regularly speaks and writes about generative AI, AI, edge computing and other cutting-edge technologies. She can be reached at [email protected].