Year: 2017
Type: preprint
Abstract: Currently, deep neural networks are deployed on low-power portable devices by first training a full-precision model using powerful hardware, and then deriving a corresponding low-precision model for efficient inference on such systems. However, training models directly with coarsely quantized weights is a key step towards learning on embedded platforms that have limited computing resources, memory capacity, and power consumption. Numerous recent publications have studied methods for training quantized networks, but these studies have mostly been empirical. In this work, we investigate training methods for quantized neural networks from a theoretical viewpoint. We first explore accuracy guarantees for training methods under convexity assumptions. We then look at the behavior of these algorithms for non-convex problems, and show that training algorithms that exploit high-precision representations have an important greedy search phase that purely quantized training methods lack, which explains the difficulty of training using low-precision arithmetic.
Source: arXiv (Cornell University)
Institutions: University of Maryland, College Park, Cornell University
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
Awards: G1722586140, AitF: EXPL: Collaborative Research: Approximate Discrete Programming for Real-Time Systems, Collaborative Research: BAMM: Baseband Accelerators for Massive Multiple-Input Multiple-Output (MIMO) Technology, CAREER: Hardware Accelerated Bayesian Inference via Approximate Message Passing: A Bottom-Up Approach, AitF: EXPL: Collaborative Research: Approximate Discrete Programming for Real-Time Systems, G6671297155
OpenAlex ID: W2963516298
