Introducing PlayTorch, a new app and SDK, for quickly and easily prototyping AI-powered experiences. This project was formerly PyTorch Live. Learn more on the PyTorch blog. Let’s play! https://t.co/OPBgc4gMOk
Chatted with @raedle about something exciting.
Imagine if you can run any PyTorch Live Model (e.g. Face Detection, Barcode scanner, ...) inside a Frame Processor without writing any native code at all - so no more Frame Processor Plugins.... 👀
@mrousavy broke new grounds with Vision Camera + frame processors! It would be pretty slick and next level to build ML-enabled apps e2e in JavaScript!
Imagine Vision Camera + @PlayTorch + RN Skia ➡�� object detection, depth estimation, GAN style, etc.
https://t.co/syJP78wcFA
🚀 The PlayTorch docs have a new Get Started guide for Expo apps (~ standalone React Native apps)
https://t.co/JfXq1JOdig
🤩 PlayTorch leverages Expo's Config Plugins to auto-configure Android and iOS app projects with the PlayTorch dependencies
#expo#reactnative#pytorch
Thank you for providing feedback about PlayTorch!
Send us feedback at the support page: https://t.co/eMjPTplPqm
To learn how to build with PlayTorch, see our tutorials and FAQ: https://t.co/tFoe6QNX4n
New versions of the PlayTorch apps are available (v0.1.2)
🔗 App Store: https://t.co/Pu2bYjCURu
🔗 Google Play Store: https://t.co/LHFtvnKFYe
More about PlayTorch on https://t.co/Ci6iLmxbTW
Real-time point cloud rendering of MiDaS' monocular depth estimation using OpenGL shaders/GLSL
Try it out w/ PlayTorch (iOS only at the moment): https://t.co/UUo1gKLlsx
MiDaS repo: https://t.co/0vEpLH6FTt
together w/ @chrisklaiber#playtorch#ml#reactnative#expo#opengl
ArcaneGAN by @devdef as an on-device ML demo in PlayTorch.
"On-device" means the machine learning inference (or processing) happens *only* on your mobile device
Try it yourself: https://t.co/9HXWQXfXFb
ArcaneGAN GitHub repo: https://t.co/kFnrWTONpK
#playtorch#ml#ondevice
Monocular depth estimation from a single input image using MiDaS.
Try it out with PlayTorch: https://t.co/9RoUy2ypfi
MiDaS GitHub repo: https://t.co/0vEpLH6FTt
The demo took ~30 minutes to build from model export to building UI!
#playtorch#ai#ml#pytorch#expo#reactnative
We’ve updated the underlying tech: the new 0.2.1 release of our NPM package react-native-pytorch-core is out with API improvements and support for Expo and the latest React Native.
Full release notes on GitHub: https://t.co/0lgHATzIrV
PlayTorch: https://t.co/tFoe6QvNQf
We'll be hosting a PlayTorch office hour on Discord today at 10 AM PT (in 15 minutes). Come and ask questions about PlayTorch or come and say hi 👋🏻
https://t.co/TJc7CmL6Ri
Today, we are happy to announce that the PlayTorch app is out of beta and available in the App Store. More about PlayTorch on https://t.co/Ci6iLmf2FO
🔗 App Store: https://t.co/Pu2bYjTXTu
🔗 Google Play Store: https://t.co/LHFtvo1J0e
@militaru_bogdan The next version of PTL will come with a new API, which will allow developers to integrate ML models and stay in JavaScript entirely. Previously, developers would have had to write data processing natively (e.g., Java, Swift)
We just released the patch v0.1.2 for PyTorch Live. Among other things, the release fixes an issue with the Android release build as well as an issue with the PyTorch Live CLI with Apple Silicon devices (M1) 🛠
https://t.co/RbzSxP8y9Y
Thanks to all PR contributors ✨
@militaru_bogdan Thanks so much, and we're glad you find PTL interesting!
We've added APIs recently supporting image segmentation models. @YinglaoL created an example app yesterday demonstrating DeepLabV3 in PTL (https://t.co/E8Enf2rnkC).
Join us on Discord, and let us know your requirements!