`

Timezone: »

 
Oral
Transferable Graph Optimizers for ML Compilers
yanqiz Zhou · Sudip Roy · Amirali Abdolrashidi · Daniel Wong · Peter Ma · Qiumin Xu · Hanxiao Liu · Phitchaya Phothilimtha · Shen Wang · Anna Goldie · Azalia Mirhoseini · James Laudon

Thu Dec 10 06:30 PM -- 06:45 PM (PST) @ Orals & Spotlights: Health/AutoML/(Soft|Hard)ware

Most compilers for machine learning (ML) frameworks need to solve many correlated optimization problems to generate efficient machine code. Current ML compilers rely on heuristics based algorithms to solve these optimization problems one at a time. However, this approach is not only hard to maintain but often leads to sub-optimal solutions especially for newer model architectures. Existing learning based approaches in the literature are sample inefficient, tackle a single optimization problem, and do not generalize to unseen graphs making them infeasible to be deployed in practice. To address these limitations, we propose an end-to-end, transferable deep reinforcement learning method for computational graph optimization (GO), based on a scalable sequential attention mechanism over an inductive graph neural network. GO generates decisions on the entire graph rather than on each individual node autoregressively, drastically speeding up the search compared to prior methods. Moreover, we propose recurrent attention layers to jointly optimize dependent graph optimization tasks and demonstrate 33%-60% speedup on three graph optimization tasks compared to TensorFlow default optimization. On a diverse set of representative graphs consisting of up to 80,000 nodes, including Inception-v3, Transformer-XL, and WaveNet, GO achieves on average 21% improvement over human experts and 18% improvement over the prior state of the art with 15x faster convergence, on a device placement task evaluated in real systems.

Author Information

yanqiz Zhou (Google Brain)
Sudip Roy (Google)
Amirali Abdolrashidi (UC Riverside)
Daniel Wong (Carnegie Mellon University)
Peter Ma (Google)
Qiumin Xu (Google)
Hanxiao Liu (Google Brain)
Phitchaya Phothilimtha (Google Brain)
Shen Wang (Google Inc)
Anna Goldie (Google Brain / Stanford)
Azalia Mirhoseini (Google Brain)
James Laudon (Google)

Related Events (a corresponding poster, oral, or spotlight)

More from the Same Authors

  • 2021 : Closing Remarks »
    Jonathan Raiman · Mimee Xu · Martin Maas · Anna Goldie · Azade Nazi · Benoit Steiner
  • 2021 : ML for Autotuning Production ML Compilers »
    Phitchaya Phothilimtha
  • 2021 : Opening Remarks »
    Jonathan Raiman · Anna Goldie · Benoit Steiner · Azade Nazi · Martin Maas · Mimee Xu
  • 2021 Workshop: ML For Systems »
    Benoit Steiner · Jonathan Raiman · Martin Maas · Azade Nazi · Mimee Xu · Anna Goldie
  • 2021 Poster: Representing Long-Range Context for Graph Neural Networks with Global Attention »
    Paras Jain · Zhanghao Wu · Matthew Wright · Azalia Mirhoseini · Joseph Gonzalez · Ion Stoica
  • 2020 Workshop: Machine Learning for Systems »
    Anna Goldie · Azalia Mirhoseini · Jonathan Raiman · Martin Maas · Xinlei XU
  • 2020 Poster: Evolving Normalization-Activation Layers »
    Hanxiao Liu · Andy Brock · Karen Simonyan · Quoc V Le
  • 2020 Spotlight: Evolving Normalization-Activation Layers »
    Hanxiao Liu · Andy Brock · Karen Simonyan · Quoc V Le
  • 2020 Poster: PyGlove: Symbolic Programming for Automated Machine Learning »
    Daiyi Peng · Xuanyi Dong · Esteban Real · Mingxing Tan · Yifeng Lu · Gabriel Bender · Hanxiao Liu · Adam Kraft · Chen Liang · Quoc V Le
  • 2020 Oral: PyGlove: Symbolic Programming for Automated Machine Learning »
    Daiyi Peng · Xuanyi Dong · Esteban Real · Mingxing Tan · Yifeng Lu · Gabriel Bender · Hanxiao Liu · Adam Kraft · Chen Liang · Quoc V Le
  • 2020 Poster: Rethinking Pre-training and Self-training »
    Barret Zoph · Golnaz Ghiasi · Tsung-Yi Lin · Yin Cui · Hanxiao Liu · Dogus Cubuk · Quoc V Le
  • 2020 Oral: Rethinking Pre-training and Self-training »
    Barret Zoph · Golnaz Ghiasi · Tsung-Yi Lin · Yin Cui · Hanxiao Liu · Dogus Cubuk · Quoc V Le
  • 2019 : Poster Session 2 »
    Hanson Wang · Yujun Lin · Yixiao Duan · Aditya Paliwal · Ameer Haj-Ali · Ryan Marcus · Tom Hope · Qiumin Xu · Nham Le · Yuxiang Sun · Ross Cutler · Vikram Nathan · Min Sun
  • 2019 : Coffee Break & Poster Session 1 »
    Yan Zhang · Jonathon Hare · Adam Prugel-Bennett · Alex Leung · Patrick Flaherty · Pitchaya Wiratchotisatian · Alessandro Epasto · Silvio Lattanzi · Sergei Vassilvitskii · Morteza Zadimoghaddam · Theja Tulabandhula · Fabian Fuchs · Adam Kosiorek · Ingmar Posner · William Hang · Anna Goldie · Sujith Ravi · Azalia Mirhoseini · Yuwen Xiong · Mengye Ren · Renjie Liao · Raquel Urtasun · Haici Zhang · Michele Borassi · Shengda Luo · Andy Trapp · Geoffroy Dubourg-Felonneau · Yasmeen Kussad · Chris Bender · Manzil Zaheer · Junier Oliva · Michał Stypułkowski · Maciej Zieba · Austin Dill · Chun-Liang Li · Songwei Ge · Eunsu Kang · Oiwi Parker Jones · Kelvin Ka Wing Wong · Josh Payne · Yang Li · Azade Nazi · Erkut Erdem · Aykut Erdem · Kevin O'Connor · Juan J Garcia · Maciej Zamorski · Jan Chorowski · Deeksha Sinha · Harry Clifford · John W Cassidy
  • 2019 Workshop: ML For Systems »
    Milad Hashemi · Azalia Mirhoseini · Anna Goldie · Kevin Swersky · Xinlei XU · Jonathan Raiman · Jonathan Raiman