arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2006.08733cs.LGcs.CRstat.ML
收藏

CryptoNAS:对ReLU预算的私人推断

CryptoNAS: Private Inference on a ReLU Budget

New York University(纽约大学)

Zahra Ghodsi, Akshaj Veldanda, Brandon Reagen, Siddharth Garg

摘要:

机器学习作为一种服务,已经引起了围绕客户端数据和提供商模型的隐私问题,并促进了对私有推理(PI)的研究:在不公开输入的情况下处理推理的方法。最近,研究人员已经采用了加密技术来证明PI是可能的,但是所有解决方案都会增加推理延迟,超出实际限制。本文观察到现有的模型不适合PI,并提出了一种新的NAS方法,名为CryptoNAS,用于查找和定制模型,以满足PI的需求。关键的见解是,在PI中,运营商延迟成本是非线性操作(例如,ReLU)主导延迟,而线性层变得有效自由。我们开发了ReLU预算作为推理延迟代理的想法,并使用CryptoNAS构建模型,在给定预算内最大限度地提高准确性。CryptoNAS将准确性提高了3.4%,延迟提高了2.4倍。

英文摘要:

Machine learning as a service has given raise to privacy concerns surrounding clients' data and providers' models and has catalyzed research in private inference (PI): methods to process inferences without disclosing inputs. Recently, researchers have adapted cryptographic techniques to show PI is possible, however all solutions increase inference latency beyond practical limits. This paper makes the observation that existing models are ill-suited for PI and proposes a novel NAS method, named CryptoNAS, for finding and tailoring models to the needs of PI. The key insight is that in PI operator latency cost are non-linear operations (e.g., ReLU) dominate latency, while linear layers become effectively free. We develop the idea of a ReLU budget as a proxy for inference latency and use CryptoNAS to build models that maximize accuracy within a given budget. CryptoNAS improves accuracy by 3.4% and latency by 2.4x over the state-of-the-art.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑