Matches in SemOpenAlex for { <https://semopenalex.org/work/W3214474508> ?p ?o ?g. }
- W3214474508 abstract "Decentralized algorithm is a form of computation that achieves a global goal through local dynamics that relies on low-cost communication between directly-connected agents. On large-scale optimization tasks involving distributed datasets, decentralized algorithms have shown strong, sometimes superior, performance over distributed algorithms with a central node. Recently, developing decentralized algorithms for deep learning has attracted great attention. They are considered as low-communication-overhead alternatives to those using a parameter server or the Ring-Allreduce protocol. However, the lack of an easy-to-use and efficient software package has kept most decentralized algorithms merely on paper. To fill the gap, we introduce BlueFog, a python library for straightforward, high-performance implementations of diverse decentralized algorithms. Based on a unified abstraction of various communication operations, BlueFog offers intuitive interfaces to implement a spectrum of decentralized algorithms, from those using a static, undirected graph for synchronous operations to those using dynamic and directed graphs for asynchronous operations. BlueFog also adopts several system-level acceleration techniques to further optimize the performance on the deep learning tasks. On mainstream DNN training tasks, BlueFog reaches a much higher throughput and achieves an overall $1.2times sim 1.8times$ speedup over Horovod, a state-of-the-art distributed deep learning package based on Ring-Allreduce. BlueFog is open source at https://github.com/Bluefog-Lib/bluefog." @default.
- W3214474508 created "2021-11-22" @default.
- W3214474508 creator A5005538082 @default.
- W3214474508 creator A5007519257 @default.
- W3214474508 creator A5010705747 @default.
- W3214474508 creator A5073026003 @default.
- W3214474508 creator A5085908411 @default.
- W3214474508 date "2021-11-08" @default.
- W3214474508 modified "2023-10-16" @default.
- W3214474508 title "BlueFog: Make Decentralized Algorithms Practical for Optimization and Deep Learning" @default.
- W3214474508 cites W1571416372 @default.
- W3214474508 cites W1573548168 @default.
- W3214474508 cites W1686810756 @default.
- W3214474508 cites W1825216778 @default.
- W3214474508 cites W2012445782 @default.
- W3214474508 cites W2029080014 @default.
- W3214474508 cites W2041517243 @default.
- W3214474508 cites W2044212084 @default.
- W3214474508 cites W2057332538 @default.
- W3214474508 cites W2083842231 @default.
- W3214474508 cites W2108598243 @default.
- W3214474508 cites W2118776392 @default.
- W3214474508 cites W2123705108 @default.
- W3214474508 cites W2135674292 @default.
- W3214474508 cites W2137435346 @default.
- W3214474508 cites W2138822887 @default.
- W3214474508 cites W2150355110 @default.
- W3214474508 cites W2151565294 @default.
- W3214474508 cites W2154834860 @default.
- W3214474508 cites W2160393298 @default.
- W3214474508 cites W2166344854 @default.
- W3214474508 cites W2194775991 @default.
- W3214474508 cites W2243945130 @default.
- W3214474508 cites W2266148951 @default.
- W3214474508 cites W2291434103 @default.
- W3214474508 cites W2402144811 @default.
- W3214474508 cites W2591577911 @default.
- W3214474508 cites W2609549560 @default.
- W3214474508 cites W2622263826 @default.
- W3214474508 cites W2766736793 @default.
- W3214474508 cites W2769644379 @default.
- W3214474508 cites W2787998955 @default.
- W3214474508 cites W2789243870 @default.
- W3214474508 cites W2902410171 @default.
- W3214474508 cites W2908007030 @default.
- W3214474508 cites W2911327153 @default.
- W3214474508 cites W2937905591 @default.
- W3214474508 cites W2944542720 @default.
- W3214474508 cites W2946815927 @default.
- W3214474508 cites W2954270931 @default.
- W3214474508 cites W2954365017 @default.
- W3214474508 cites W2955425717 @default.
- W3214474508 cites W2962863496 @default.
- W3214474508 cites W2963163009 @default.
- W3214474508 cites W2963179579 @default.
- W3214474508 cites W2963228337 @default.
- W3214474508 cites W2963341956 @default.
- W3214474508 cites W2963499661 @default.
- W3214474508 cites W2963649943 @default.
- W3214474508 cites W2963773265 @default.
- W3214474508 cites W2964165783 @default.
- W3214474508 cites W2964331441 @default.
- W3214474508 cites W2970971581 @default.
- W3214474508 cites W2972087877 @default.
- W3214474508 cites W2996118901 @default.
- W3214474508 cites W3006548771 @default.
- W3214474508 cites W3011751313 @default.
- W3214474508 cites W3015556758 @default.
- W3214474508 cites W3016897523 @default.
- W3214474508 cites W3022273187 @default.
- W3214474508 cites W3034741284 @default.
- W3214474508 cites W3047780503 @default.
- W3214474508 cites W3096403968 @default.
- W3214474508 cites W3098903812 @default.
- W3214474508 cites W3099715789 @default.
- W3214474508 cites W3100513551 @default.
- W3214474508 cites W3101665129 @default.
- W3214474508 cites W3126226094 @default.
- W3214474508 cites W3160274442 @default.
- W3214474508 cites W3162126802 @default.
- W3214474508 cites W3166004119 @default.
- W3214474508 cites W3169161304 @default.
- W3214474508 cites W3171826865 @default.
- W3214474508 cites W3181241935 @default.
- W3214474508 cites W3203694565 @default.
- W3214474508 cites W3204517019 @default.
- W3214474508 cites W3206403927 @default.
- W3214474508 cites W3206827202 @default.
- W3214474508 cites W3209916863 @default.
- W3214474508 cites W3212906826 @default.
- W3214474508 cites W988471318 @default.
- W3214474508 doi "https://doi.org/10.48550/arxiv.2111.04287" @default.
- W3214474508 hasPublicationYear "2021" @default.
- W3214474508 type Work @default.
- W3214474508 sameAs 3214474508 @default.
- W3214474508 citedByCount "2" @default.
- W3214474508 countsByYear W32144745082021 @default.
- W3214474508 crossrefType "posted-content" @default.
- W3214474508 hasAuthorship W3214474508A5005538082 @default.
- W3214474508 hasAuthorship W3214474508A5007519257 @default.