[1904.08368] Relay: A High-Level Compiler for Deep Learningopen searchopen navigation menucontact arXivsubscribe to arXiv mailings

Frameworks for writing, compiling, and optimizing deep learning (DL) models have recently enabled progress in areas like computer vision and natural language processing. Extending these frameworks to accommodate the rapidly diversifying landscape of DL models and hardware platforms presents challenging tradeoffs between expressivity, composability, and portability. We present Relay, a new compiler framework for DL. Relay's functional, statically typed intermediate representation (IR) unifies and generalizes existing DL IRs to express state-of-the-art models. The introduction of Relay's expressive IR requires careful design of domain-specific optimizations, addressed via Relay's extension mechanisms. Using these extension mechanisms, Relay supports a unified compiler that can target a variety of hardware platforms. Our evaluation demonstrates Relay's competitive performance for a broad class of models and devices (CPUs, GPUs, and emerging accelerators). Relay's design demonstrates how a

4 mentions: @00_@m_morise@myomyo_28
Keywords: deep learning
Date: 2020/09/14 11:22

Referring Tweets

@00_ 霧雨魔理沙(Marisa Kirisame)がUWのPLグループから論文出してる...。中の人が絶対に本名使いたくない人なんだろうな。 t.co/CPFxOnnEIB
@m_morise t.co/xqFq1qKPeG 著者でワラタw

Related Entries

Read more Non-Negative Matrix Factorization - ML Wiki
1 users, 0 mentions 2018/12/05 04:15
Read more Adversarial Robustness - Theory and Practice
1 users, 0 mentions 2018/12/03 19:09
Read more [1902.09722] Topological Bayesian Optimization with Persistence Diagrams
1 users, 2 mentions 2019/02/27 09:47
Read more Concrete Distributionについて論文を読んだ - 機械学習とかコンピュータビジョンとか
2 users, 0 mentions 2018/06/23 10:24
Read more Krylov Subspace Method for Nonlinear Dynamical Systems with Random Noise
1 users, 1 mentions 2020/09/18 02:21