Part of Proceedings of Machine Learning and Systems 1 (MLSys 2019)
Dan Moldovan, James Decker, Fei Wang, Andrew Johnson, Brian Lee, Zachary Nado, D Sculley, Tiark Rompf, Alexander B Wiltschko
There is a perceived trade-off between machine learning code that is easy to write, and machine learning code that is scalable or fast to execute. In machine learning, {\em imperative} style libraries like Autograd and PyTorch are easy to write, but suffer from high interpretive overhead and are not easily deployable in production or mobile settings. {\em Graph-based} libraries like TensorFlow and Theano benefit from whole-program optimization and can be deployed broadly, but make expressing complex models more cumbersome. We describe how the use of staged programming in Python, via source code transformation, offers a midpoint between these two library design patterns, capturing the benefits of both. A key insight is to delay all type-dependent decisions until runtime, via dynamic dispatch. We instantiate these principles in AG, a software system that improves the programming experience of the TensorFlow library, and demonstrate usability improvements with no loss in performance compared to native TensorFlow graphs. We also show that our system is backend agnostic, and demonstrate targeting an alternate IR with characteristics not found in TensorFlow graphs.