In plain words: A deep learning platform built into the Swift language that automatically works out how to tweak a model and compiles code ahead of time to run on phones or data-center machines. It powered over 30 deep learning models in mobile and data-center apps.
Abstract
Swift for TensorFlow is a deep learning platform that scales from mobile devices to clusters of hardware accelerators in data centers. It combines a language-integrated automatic differentiation system and multiple Tensor implementations within a modern ahead-of-time compiled language oriented around mutable value semantics. The resulting platform has been validated through use in over 30 deep learning models and has been employed across data center and mobile applications.
Brennan Saeta, Denys Shabalin, Marc Rasi, Brad Larson, Xihui Wu, Parker Schuh, Michelle Casbon, Daniel Zheng, Saleem Abdulrasool, Aleksandr Efremov, Dave Abrahams, Chris Lattner, et al.
arXiv:2102.13243 · cs.LG, cs.DC · submitted Feb 26, 2021
abstract · pdf · html
