Play torch, previously known as PyTorch live, helps build AI-powered mobile prototypes quickly. The new release is much simpler and provides a much better developer experience. Playtech code repository is in Meta research GitHub which can be found here.

The initial release had a command line interface to set up a development environment and an SDK.

The new version, which is developed in partnership with Expo, simplifies the process by eliminating a complicated development environment. Now it is possible to create AI prototypes from any browser.

First, you can download the PlayTorch app from the play store. After installation, you can write your code for an AI-powered PlayTorch snack in Playtech. Dev/snack. Once you have written the code, to try what you have built, use the QR code scanner in the PlayTorch app to scan the QR code on the Snack page and load the code to your device. Look at the image below for reference.

What else does the PlayTorch app have to offer

The app contains several demos of building AI prototypes using a variety of different ML models like NLP and object detection.

You can share what you have built on the PlayTorch app. When those you have shared open the link, the app will load what you have built from the cloud and give them the experience.

The next version of SDK can handle the data processing in javascript. The variety of supported model architectures has also been expanded.

New Data processing API for prototyping

Now you are allowed to directly access tensors instead of just getting to access predefined transformations with the new JSI API. You can manipulate tensors however you want for your prototypes. You will also be able to write everything in Javascript and have access to all types of annotations and other features available in those languages.

I) Image to image transformation: Look at how the world would have looked if it were an anime

(source-https://pytorch.org/blog/introducing-the-playtorch-app/)

II) Language translation using Seq2Seq model (source-https://pytorch.org/blog/introducing-the-playtorch-app/)

III) Image segmentation using DeepLab V3 source(https://pytorch.org/blog/introducing-the-playtorch-app/)

You can go over here and try them for yourself with the help of tutorials which can be found here. Check out the GitHub link here.

Prathvik is ML/AI Research content intern at MarktechPost, he is a 3rd year undergraduate at IIT Kharagpur. He has a keen interest in Machine learning and data science.He is enthusiastic in learning about the applications of in different fields of study .

Free-2 Min AI Newsletter Join 500,000+ AI Folks EmailEnter your email address Subscribe

Marktechpost is a California based AI News Platform providing easy-to-consume, byte size updates in machine learning, deep learning, and data science research

© 2021 Marktechpost LLC. All Rights Reserved. Made with ❤️ in California