Výsledky vyhledávání - "SPRING, RYAN"

Report

Training Question Answering Models From Synthetic Data

Autor: Puri, Raul, Spring, Ryan, Patwary, Mostofa, Shoeybi, Mohammad, Catanzaro, Bryan

Question and answer generation is a data augmentation method that aims to improve question answering (QA) models given the limited amount of human labeled data. However, a considerable gap remains between synthetic and human-generated question-answer

Externí odkaz: http://arxiv.org/abs/2002.09599

Zobrazit plný text záznamu

Report

Compressing Gradient Optimizers via Count-Sketches

Autor: Spring, Ryan, Kyrillidis, Anastasios, Mohan, Vijai, Shrivastava, Anshumali

Many popular first-order optimization methods (e.g., Momentum, AdaGrad, Adam) accelerate the convergence rate of deep learning models. However, these algorithms require auxiliary parameters, which cost additional memory proportional to the number of

Externí odkaz: http://arxiv.org/abs/1902.00179

Zobrazit plný text záznamu

Report

MISSION: Ultra Large-Scale Feature Selection using Count-Sketches

Autor: Aghazadeh, Amirali, Spring, Ryan, LeJeune, Daniel, Dasarathy, Gautam, Shrivastava, Anshumali, Baraniuk, Richard G.

Feature selection is an important challenge in machine learning. It plays a crucial role in the explainability of machine-driven decisions that are rapidly permeating throughout modern society. Unfortunately, the explosion in the size and dimensional

Externí odkaz: http://arxiv.org/abs/1806.04310

Zobrazit plný text záznamu

Report

A New Unbiased and Efficient Class of LSH-Based Samplers and Estimators for Partition Function Computation in Log-Linear Models

Autor: Spring, Ryan, Shrivastava, Anshumali

Log-linear models are arguably the most successful class of graphical models for large-scale applications because of their simplicity and tractability. Learning and inference with these models require calculating the partition function, which is a ma

Externí odkaz: http://arxiv.org/abs/1703.05160

Zobrazit plný text záznamu

Report

Scalable and Sustainable Deep Learning via Randomized Hashing

Autor: Spring, Ryan, Shrivastava, Anshumali

Current deep learning architectures are growing larger in order to learn from complex datasets. These architectures require giant matrix multiplication operations to train millions of parameters. Conversely, there is another growing trend to bring de

Externí odkaz: http://arxiv.org/abs/1602.08194

Zobrazit plný text záznamu