WRPN & Apprentice: Methods for Training and Inference using Low-Precision Numerics

2018/03/01 by Asit Mishra, Mishra, Asit, Debbie Marr +1
Computer Science · #Computer Vision and Pattern Recognition (cs.CV) #Distributed and Parallel Computing Systems #FOS: Computer and information sciences #Machine Learning (cs.LG) #Neural and Evolutionary Computing (cs.NE)

paper · pdf · doi:10.48550/arxiv.1803.00227

openalex publication_date 2018/03/01 · openalex created_date 2024/04/10 · openalex updated_date 2026/07/28

Abstract

Today's high performance deep learning architectures involve large models with numerous parameters. Low precision numerics has emerged as a popular technique to reduce both the compute and memory requirements of these large models. However, lowering precision often leads to accuracy degradation. We describe three schemes whereby one can both train and do efficient inference using low precision numerics without hurting accuracy. Finally, we describe an efficient hardware accelerator that can take advantage of the proposed low precision numerics.

Related

• Deep Learning with Limited Numerical Precision
• Training and Inference with Integers in Deep Neural Networks
• How Low Can We Go: Trading Memory for Error in Low-Precision Training
• A Metric Driven Approach to Mixed Precision Training
• WRPN: Training and Inference using Wide Reduced-Precision Networks