Cross-Feature Transfer Learning for Efficient Tensor Program Generation

Author:

Verma Gaurav1ORCID,Raskar Siddhisanket2ORCID,Emani Murali2ORCID,Chapman Barbara1ORCID

Affiliation:

1. Deparment of Computer Science, Stony Brook University, Stony Brook, NY 11794, USA

2. Argonne National Laboratory, Lemont, IL 60439, USA

Abstract

Tuning tensor program generation involves navigating a vast search space to find optimal program transformations and measurements for a program on the target hardware. The complexity of this process is further amplified by the exponential combinations of transformations, especially in heterogeneous environments. This research addresses these challenges by introducing a novel approach that learns the joint neural network and hardware features space, facilitating knowledge transfer to new, unseen target hardware. A comprehensive analysis is conducted on the existing state-of-the-art dataset, TenSet, including a thorough examination of test split strategies and the proposal of methodologies for dataset pruning. Leveraging an attention-inspired technique, we tailor the tuning of tensor programs to embed both neural network and hardware-specific features. Notably, our approach substantially reduces the dataset size by up to 53% compared to the baseline without compromising Pairwise Comparison Accuracy (PCA). Furthermore, our proposed methodology demonstrates competitive or improved mean inference times with only 25–40% of the baseline tuning time across various networks and target hardware. The attention-based tuner can effectively utilize schedules learned from previous hardware program measurements to optimize tensor program tuning on previously unseen hardware, achieving a top-5 accuracy exceeding 90%. This research introduces a significant advancement in autotuning tensor program generation, addressing the complexities associated with heterogeneous environments and showcasing promising results regarding efficiency and accuracy.

Funder

Stony Brook Research Computing and Cyberinfrastructure

Argonne Leadership Computing Facility

Exascale Computing Project

National Science Foundation

Publisher

MDPI AG

Reference50 articles.

1. Sabne, A. (2023, November 30). XLA: Compiling Machine Learning for Peak Performance. Available online: https://www.tensorflow.org/xla.

2. Chen, T., Moreau, T., Jiang, Z., Zheng, L., Yan, E., Shen, H., Cowan, M., Wang, L., Hu, Y., and Ceze, L. (2018, January 8–10). {TVM}: An automated {End-to-End} optimizing compiler for deep learning. Proceedings of the 13th USENIX Symposium on Operating Systems Design and Implementation (OSDI 18), Carlsbad, CA, USA.

3. Rotem, N., Fix, J., Abdulrasool, S., Catron, G., Deng, S., Dzhabarov, R., Gibson, N., Hegeman, J., Lele, M., and Levenstein, R. (2018). Glow: Graph lowering compiler techniques for neural networks. arXiv.

4. The Tensor Algebra Compiler;Kjolstad;Proc. ACM Program. Lang.,2017

5. The deep learning compiler: A comprehensive survey;Li;IEEE Trans. Parallel Distrib. Syst.,2020

Cited by 3 articles. 订阅此论文施引文献 订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3