Výsledky vyhledávání

Report

Running Alchemist on Cray XC and CS Series Supercomputers: Dask and PySpark Interfaces, Deployment Options, and Data Transfer Times

Autor: Rothauge, Kai, Ayyalasomayajula, Haripriya, Maschhoff, Kristyn J., Ringenburg, Michael, Mahoney, Michael W.

Alchemist is a system that allows Apache Spark to achieve better performance by interfacing with HPC libraries for large-scale distributed computations. In this paper, we highlight some recent developments in Alchemist that are of interest to Cray us

Externí odkaz: http://arxiv.org/abs/1910.01354

Zobrazit plný text záznamu

Report

Residual Networks as Nonlinear Systems: Stability Analysis using Linearization

Autor: Rothauge, Kai, Yao, Zhewei, Hu, Zixi, Mahoney, Michael W.

We regard pre-trained residual networks (ResNets) as nonlinear systems and use linearization, a common method used in the qualitative analysis of nonlinear systems, to understand the behavior of the networks under small perturbations of the input ima

Externí odkaz: http://arxiv.org/abs/1905.13386

Zobrazit plný text záznamu

Report

On the Computational Inefficiency of Large Batch Sizes for Stochastic Gradient Descent

Autor: Golmant, Noah, Vemuri, Nikita, Yao, Zhewei, Feinberg, Vladimir, Gholami, Amir, Rothauge, Kai, Mahoney, Michael W., Gonzalez, Joseph

Increasing the mini-batch size for stochastic gradient descent offers significant opportunities to reduce wall-clock training time, but there are a variety of theoretical and systems challenges that impede the widespread success of this technique. We

Externí odkaz: http://arxiv.org/abs/1811.12941

Zobrazit plný text záznamu

Report

Alchemist: An Apache Spark <=> MPI Interface

Autor: Gittens, Alex, Rothauge, Kai, Wang, Shusen, Mahoney, Michael W., Kottalam, Jey, Gerhardt, Lisa, Prabhat, Ringenburg, Michael, Maschhoff, Kristyn

The Apache Spark framework for distributed computation is popular in the data analytics community due to its ease of use, but its MapReduce-style programming model can incur significant overheads when performing computations that do not map directly

Externí odkaz: http://arxiv.org/abs/1806.01270

Zobrazit plný text záznamu

Report

Accelerating Large-Scale Data Analysis by Offloading to High-Performance Computing Libraries using Alchemist

Autor: Gittens, Alex, Rothauge, Kai, Wang, Shusen, Mahoney, Michael W., Gerhardt, Lisa, Prabhat, Kottalam, Jey, Ringenburg, Michael, Maschhoff, Kristyn

Apache Spark is a popular system aimed at the analysis of large data sets, but recent studies have shown that certain computations---in particular, many linear algebra computations that are the basis for solving common machine learning problems---are

Externí odkaz: http://arxiv.org/abs/1805.11800

Zobrazit plný text záznamu

Report

The Discrete Adjoint Method for Exponential Integration

Autor: Rothauge, Kai, Haber, Eldad, Ascher, Uri

The implementation of the discrete adjoint method for exponential time differencing (ETD) schemes is considered. This is important for parameter estimation problems that are constrained by stiff time-dependent PDEs when the discretized PDE system is

Externí odkaz: http://arxiv.org/abs/1610.02596

Zobrazit plný text záznamu

Report

Numerical Computation of the Gradient and the Action of the Hessian for Time-Dependent PDE-Constrained Optimization Problems

Autor: Rothauge, Kai, Haber, Eldad, Ascher, Uri

We present a systematic derivation of the algorithms required for computing the gradient and the action of the Hessian of an arbitrary misfit function for large-scale parameter estimation problems involving linear time-dependent PDEs with stationary

Externí odkaz: http://arxiv.org/abs/1509.03801

Zobrazit plný text záznamu