Subscribe to the gold package and get unlimited access to Shamra Academy
Register a new userA Scalable and Cloud-Native Hyperparameter Tuning System
187
0
0.0
(
0
)
Added by
Amit Saha
Publication date
2020
fields
Informatics Engineering
and research's language is
English
Ask ChatGPT about the research
No Arabic abstract
In this paper, we introduce Katib: a scalable, cloud-native, and production-ready hyperparameter tuning system that is agnostic of the underlying machine learning framework. Though there are multiple hyperparameter tuning systems available, this is the first one that caters to the needs of both users and administrators of the system. We present the motivation and design of the system and contrast it with existing hyperparameter tuning systems, especially in terms of multi-tenancy, scalability, fault-tolerance, and extensibility. It can be deployed on local machines, or hosted as a service in on-premise data centers, or in private/public clouds. We demonstrate the advantage of our system using experimental results as well as real-world, production use cases. Katib has active contributors from multiple companies and is open-sourced at emph{https://github.com/kubeflow/katib} under the Apache 2.0 license.
rate research
Read More
Current cloud services are moving away from monolithic designs and towards graphs of many loosely-coupled, single-concerned microservices. Microservices have several advantages, including speeding up development and deployment, allowing specialization of the software infrastructure, and helping with debugging and error isolation. At the same time they introduce several hardware and software challenges. Given that most of the performance and efficiency implications of microservices happen at scales larger than what is available outside production deployments, studying such effects requires designing the right simulation infrastructures. We present uqSim, a scalable and validated queueing network simulator designed specifically for interactive microservices. uqSim provides detailed intra- and inter-microservice models that allow it to faithfully reproduce the behavior of complex, many-tier applications. uqSim is also modular, allowing reuse of individual models across microservices and end-to-end applications. We have validated uqSim both against simple and more complex microservices graphs, and have shown that it accurately captures performance in terms of throughput and tail latency. Finally, we use uqSim to model the tail at scale effects of request fanout, and the performance impact of power management in latency-sensitive microservices.
Both feature selection and hyperparameter tuning are key tasks in machine learning. Hyperparameter tuning is often useful to increase model performance, while feature selection is undertaken to attain sparse models. Sparsity may yield better model interpretability and lower cost of data acquisition, data handling and model inference. While sparsity may have a beneficial or detrimental effect on predictive performance, a small drop in performance may be acceptable in return for a substantial gain in sparseness. We therefore treat feature selection as a multi-objective optimization task. We perform hyperparameter tuning and feature selection simultaneously because the choice of features of a model may influence what hyperparameters perform well. We present, benchmark, and compare two different approaches for multi-objective joint hyperparameter optimization and feature selection: The first uses multi-objective model-based optimization. The second is an evolutionary NSGA-II-based wrapper approach to feature selection which incorporates specialized sampling, mutation and recombination operators. Both methods make use of parameterized filter ensembles. While model-based optimization needs fewer objective evaluations to achieve good performance, it incurs computational overhead compared to the NSGA-II, so the preferred choice depends on the cost of evaluating a model on given data.
Businesses have made increasing adoption and incorporation of cloud technology into internal processes in the last decade. The cloud-based deployment provides on-demand availability without active management. More recently, the concept of cloud-native application has been proposed and represents an invaluable step toward helping organizations develop software faster and update it more frequently to achieve dramatic business outcomes. Cloud-native is an approach to build and run applications that exploit the cloud computing delivery models advantages. It is more about how applications are created and deployed than where. The container-based virtualization technology, such as Docker and Kubernetes, serves as the foundation for cloud-native applications. This paper investigates the performance of two popular computational-intensive applications, big data, and deep learning, in a cloud-native environment. We analyze the system overhead and resource usage for these applications. Through extensive experiments, we show that the completion time reduces by up to 79.4% by changing the default setting and increases by up to 96.7% due to different resource management schemes on two platforms. Additionally, the resource release is delayed by up to 116.7% across different systems. Our work can guide developers, administrators, and researchers to better design and deploy their applications by selecting and configuring a hosting platform.
In this work we detail a novel open source library, called MMLSpark, that combines the flexible deep learning library Cognitive Toolkit, with the distributed computing framework Apache Spark. To achieve this, we have contributed Java Language bindings to the Cognitive Toolkit, and added several new components to the Spark ecosystem. In addition, we also integrate the popular image processing library OpenCV with Spark, and present a tool for the automated generation of PySpark wrappers from any SparkML estimator and use this tool to expose all work to the PySpark ecosystem. Finally, we provide a large library of tools for working and developing within the Spark ecosystem. We apply this work to the automated classification of Snow Leopards from camera trap images, and provide an end to end solution for the non-profit conservation organization, the Snow Leopard Trust.
We present GSPMD, an automatic, compiler-based parallelization system for common machine learning computation graphs. It allows users to write programs in the same way as for a single device, then give hints through a few annotations on how to distribute tensors, based on which GSPMD will parallelize the computation. Its representation of partitioning is simple yet general, allowing it to express different or mixed paradigms of parallelism on a wide variety of models. GSPMD infers the partitioning for every operator in the graph based on limited user annotations, making it convenient to scale up existing single-device programs. It solves several technical challenges for production usage, such as static shape constraints, uneven partitioning, exchange of halo data, and nested operator partitioning. These techniques allow GSPMD to achieve 50% to 62% compute utilization on 128 to 2048 Cloud TPUv3 cores for models with up to one trillion parameters. GSPMD produces a single program for all devices, which adjusts its behavior based on a run-time partition ID, and uses collective operators for cross-device communication. This property allows the system itself to be scalable: the compilation time stays constant with increasing number of devices.
suggested questions
Log in to be able to interact and post comments
comments
Fetching comments
Sign in to be able to follow your search criteria
