Deploying Machine Learning Models to Ahead-of-Time Runtime on Edge Using MicroTVM

被引:0
|
作者
Liu, Chen [1 ]
Jobst, Matthias [1 ]
Jobst, Matthias [1 ]
Guo, Liyuan [1 ]
Shi, Xinyue [1 ]
Partzsch, Johannes [1 ]
Mayr, Christian [1 ]
机构
[1] Tech Univ Dresden, Dresden, Germany
关键词
TVM; MicroTVM; model deployment; BYOC; UMA;
D O I
10.1145/3615338.3618125
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
In the past few years, more and more AI applications have been applied to edge devices. However, models trained by data scientists with machine learning frameworks, such as PyTorch or TensorFlow, can not be seamlessly executed on edge. In this paper, we develop an end-to-end code generator parsing a pre-trained model to C source libraries for the backend using MicroTVM, a machine learning compiler framework extension addressing inference on bare metal devices. An analysis shows that specific compute-intensive operators can be easily offloaded to the dedicated accelerator with a Universal Modular Accelerator (UMA) interface, while others are processed in the CPU cores. By using the automatically generated ahead-of-time C runtime, we conduct a hand gesture recognition experiment on an ARM Cortex M4F core.
引用
收藏
页码:37 / 40
页数:4
相关论文
共 50 条
  • [1] An Ahead-of-Time Compiler System for the IoT-Cloud Virtual Machine
    Jeong, Juho
    Son, Yunsik
    Lee, YangSun
    2018 TENTH INTERNATIONAL CONFERENCE ON UBIQUITOUS AND FUTURE NETWORKS (ICUFN 2018), 2018, : 293 - 295
  • [2] Bytecode-to-C Ahead-of-Time Compilation for Android Dalvik Virtual Machine
    Oh, Hyeong-Seok
    Yeo, Ji Hwan
    Moon, Soo-Mook
    2015 DESIGN, AUTOMATION & TEST IN EUROPE CONFERENCE & EXHIBITION (DATE), 2015, : 1048 - 1053
  • [3] Bytecode-to-C ahead-of-time compilation for Android Dalvik Virtual Machine
    Department of Electrical and Computer Engineering, Seoul National University, Seoul, Korea, Republic of
    Proc. Des. Autom. Test Eur. DATE, (1048-1053):
  • [4] Deploying Collaborative Machine Learning Systems in Edge with Multiple Cameras
    Jang, Si Young
    Acer, Utku Gunay
    Min, Chulhong
    Kawsar, Fahim
    13TH INTERNATIONAL CONFERENCE ON MOBILE COMPUTING AND UBIQUITOUS NETWORK (ICMU2021), 2021,
  • [5] Deploying Machine Learning Models for Public Policy: A Framework
    Ackermann, Klaus
    Walsh, Joe
    De Unanue, Adolfo
    Naveed, Hareem
    Rivera, Andrea Navarrete
    Lee, Sun-Joo
    Bennett, Jason
    Defoe, Michael
    Cody, Crystal
    Haynes, Lauren
    Ghani, Rayid
    KDD'18: PROCEEDINGS OF THE 24TH ACM SIGKDD INTERNATIONAL CONFERENCE ON KNOWLEDGE DISCOVERY & DATA MINING, 2018, : 15 - 22
  • [6] Serverless Computing Approach for Deploying Machine Learning Applications in Edge Layer
    Bac, Ta Phuong
    Tran, Minh Ngoc
    Kim, YoungHan
    36TH INTERNATIONAL CONFERENCE ON INFORMATION NETWORKING (ICOIN 2022), 2022, : 396 - 401
  • [7] APINET: A Methodology for Deploying Deep and Machine Learning Models as Services
    Zatarain Cabada, Ramon
    Barron Estrada, Maria Lucia
    Leyva Lopez, Nestor
    Cardenas Lopez, Hector Manuel
    Batiz Beltran, Victor Manuel
    2023 MEXICAN INTERNATIONAL CONFERENCE ON COMPUTER SCIENCE, ENC, 2024,
  • [8] Runtime Optimizations for Tree-based Machine Learning Models
    Asadi, Nima
    Lin, Jimmy
    de Vries, Arjen P.
    IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING, 2014, 26 (09) : 2281 - 2292
  • [9] Moroccan's Arabic Speech Training And Deploying Machine Learning Models with Teachable Machine
    Jebbar, Mostafa
    Maizate, Abderrahim
    Ait Abdelouahid, Rachida
    Procedia Computer Science, 2022, 203 : 801 - 806
  • [10] Eliminating Abstraction Overhead of Java']Java Stream Pipelines using Ahead-of-Time Program Optimization
    Moller, Anders
    Veileborg, Oskar Haarklou
    PROCEEDINGS OF THE ACM ON PROGRAMMING LANGUAGES-PACMPL, 2020, 4 (04):