Join the gaming leaders, along with Games Beat and Facebook Gaming, for their second annual Games Beat and Facebook Gaming Summit. GamesBeat: Into the Metaverse 2 This coming January 25-27, 2022. Learn more about the event.
During its PyTorch Developer Day conference, Meta (formerly Facebook) announced PyTorch Live, a set of tools designed to facilitate AI-driven experiences for mobile devices. PyTorch Live offers a single programming language – JavaScript – for creating apps for Android and iOS, as well as the process of developing customized machine learning models used by the PyTorch community.
“PyTorch’s mission is to accelerate the path from research prototyping to production deployment. With a growing mobile machine learning ecosystem, this has never been more important than ever,” a spokesman told VentureBeat by email. Told by “With the aim of helping mobile developers reduce friction to create novel machine learning-based solutions, we introduce PyTorch Live: create, test, and (in the future) AI demos on devices built on PyTorch. A sharing tool. ”
Pie Torch Live
PyTorch, publicly released by Meta in January 2017, is an open source machine learning library based on Torch, a scientific computing framework and scripting language based on the Lua programming language. While TensorFlow (since November 2015) has been a bit long, PyTorch is seeing rapid growth in the data science and developer community. According to GitHub’s 2018 Octoverse report, it claimed one of the top spots for fast-growing open source projects last year, and Meta recently revealed that in 2019 the platform will collaborate. The number has increased by more than 50% year on year. 1,200
PyTorch Live builds on PyTorch Mobile, a runtime that allows developers to stay within the PyTorch ecosystem from model training to deployment, and the React Native Library to create visual user interfaces. PyTorch Mobile powers the guesswork available on the device for PyTorch Live.
PyTorch Mobile launched in October 2019, following the initial release of Caffe2go, a mobile CPU- and GPU-optimized version of Meta’s Caffe2 machine learning framework. PyTorch Mobile can launch with its runtime and was built on the premise that whatever the developer wants to do on the mobile or Edge device, the developer also wants to do on the server.
For example, if you want to show a mobile app model running on Android and iOS, it will take days to configure the project and create a user interface. With PyTorch Live, it halves the cost, and you don’t have to have Android and iOS developer experience, “meta AI software engineer Roman Riddle shared with VentureBeat before today’s announcement. Said in a pre-recorded video.
Built-in tools
PyTorch Live onboard with command line interface (CLI) and data processing API. CLI enables developers to create a mobile development environment and bootstrap mobile app projects. As far as the Data Processing API is concerned, it develops and integrates custom models for use with the PyTorch Live API, which can then be built into mobile AI-powered apps for Android and iOS.
In the future, Meta intends to help the community discover and share PyTorch models and demos through PyTorch Live, as well as provide a more customized data processing API and support machine learning domains with audio and video data. let’s work.
“It simply came to our notice then. [developers] To create mobile apps and show the community machine learning models, “Riddle continued.” It’s also an opportunity to take it a step further by building a thriving community. [of] Researchers and mobile developers [who] Share and use pilots’ mobile models and engage with each other.
Venture beat
VentureBeat’s mission is to become a digital town square for technical decision makers to learn about change technology and transactions. Our site provides essential information about data technologies and strategies to guide your organizations. We invite you to become a member of our community to access:
- Updates on topics of interest to you
- Our newsletters
- Gated Thinking Leader content and discounted access to our valuable events, e.g. Transform 2021: learn more
- Networking features, and more
Become a member.