Dynamic control flow in large-scale machine learning

Yuan Yu, Peter Hawkins, Michael Isard, Manjunath Kudlur, Rajat Monga, Derek Murray, Xiaoqiang Zheng, Martín Abadi, Paul Barham, Eugene Brevdo, Mike Burrows, Andy Davis (+3 others)
2018 Proceedings of the Thirteenth EuroSys Conference on - EuroSys '18  
Many recent machine learning models rely on fine-grained dynamic control flow for training and inference. In particular, models based on recurrent neural networks and on reinforcement learning depend on recurrence relations, data-dependent conditional execution, and other features that call for dynamic control flow. These applications benefit from the ability to make rapid control-flow decisions across a set of computing devices in a distributed system. For performance, scalability, and
more » ... veness, a machine learning system must support dynamic control flow in distributed and heterogeneous environments. This paper presents a programming model for distributed machine learning that supports dynamic control flow. We describe the design of the programming model, and its implementation in TensorFlow, a distributed machine learning system. Our approach extends the use of dataflow graphs to represent machine learning models, offering several distinctive features. First, the branches of conditionals and bodies of loops can be partitioned across many machines to run on a set of heterogeneous devices, including CPUs, GPUs, and custom ASICs. Second, programs written in our model support automatic differentiation and distributed gradient computations, which are necessary for training machine learning models that use control flow. Third, our choice of non-strict semantics enables multiple loop iterations to execute in parallel across machines, and to overlap compute and I/O operations. We have done our work in the context of TensorFlow, and it has been used extensively in research and production. We evaluate it using several real-world applications, and demonstrate its performance and scalability.
doi:10.1145/3190508.3190551 dblp:conf/eurosys/YuABBBDDGHHIKMM18 fatcat:5u4gcsi5fba33mv2nyni32h424