HeterPS: Distributed Deep Learning With Reinforcement Learning Based Scheduling in Heterogeneous Environments
Authors:
Preprint
Abstract
Deep neural networks (DNNs) exploit many layers and a large number of parameters to achieve excellent performance. The training process of DNN models generally handles large-scale input data with many sparse features, which incurs high Input/Output (IO) cost, while some layers are compute-intensive. The training process generally exploits distributed computing resources to reduce training time. In addition, heterogeneous computing resources, e.g., CPUs, GPUs of multiple types, are available for the distributed training process. Thus, the scheduling of multiple layers to diverse computing resources is critical for the training process. To efficiently train a DNN model using the heterogeneous computing resources, we propose a distributed framework, i.e., Paddle-Heterogeneous Parameter Server (Paddle-HeterPS), composed of a distributed architecture and a Reinforcement Learning (RL)-based scheduling method. The advantages of Paddle-HeterPS are three-fold compared with existing frameworks. First, Paddle-HeterPS enables efficient training process of diverse workloads with heterogeneous computing resources. Second, Paddle-HeterPS exploits an RL-based method to efficiently schedule the workload of each layer to appropriate computing resources to minimize the cost while satisfying throughput constraints. Third, Paddle-HeterPS manages data storage and data communication among distributed computing resources. We carry out extensive experiments to show that Paddle-HeterPS significantly outperforms state-of-the-art approaches in terms of throughput (14.5 times higher) and monetary cost (312.3% smaller). The codes of the framework are publicly available at: https://github.com/PaddlePaddle/Paddle.Keywords
All Related Versions
This publication has 11 references indexed in Scilit:
- From distributed machine learning to federated learning: a surveyKnowledge and Information Systems, 2022
- Job scheduling for distributed machine learning in optical WANFuture Generation Computer Systems, 2020
- Baidu Kunlun An AI processor for diversified workloads2020
- Scheduling Algorithms for Efficient Execution of Stream Workflow Applications in Multicloud EnvironmentsIEEE Transactions on Services Computing, 2019
- Efficient Scheduling of Scientific Workflows Using Hot Metadata in a Multisite CloudIEEE Transactions on Knowledge and Data Engineering, 2018
- Multi-objective scheduling of Scientific Workflows in multisite cloudsFuture Generation Computer Systems, 2016
- LSTM: A Search Space OdysseyIEEE Transactions on Neural Networks and Learning Systems, 2016
- Reevaluating Amdahl’s law in the multicore eraJournal of Parallel and Distributed Computing, 2010
- Complexity of Scheduling Parallel Task SystemsSIAM Journal on Discrete Mathematics, 1989
- Interval forms of Newtons methodComputing, 1978