{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,4,17]],"date-time":"2026-04-17T07:31:42Z","timestamp":1776411102989,"version":"3.51.2"},"reference-count":61,"publisher":"Institute of Electrical and Electronics Engineers (IEEE)","issue":"5","license":[{"start":{"date-parts":[[2024,9,1]],"date-time":"2024-09-01T00:00:00Z","timestamp":1725148800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/ieeexplore.ieee.org\/Xplorehelp\/downloads\/license-information\/IEEE.html"},{"start":{"date-parts":[[2024,9,1]],"date-time":"2024-09-01T00:00:00Z","timestamp":1725148800000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2024,9,1]],"date-time":"2024-09-01T00:00:00Z","timestamp":1725148800000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"funder":[{"name":"Beijing Municipal Science and Technology","award":["Z231100010323002"],"award-info":[{"award-number":["Z231100010323002"]}]},{"name":"Beijing Municipal Science and Technology","award":["CEIEC-2022-ZM02-0247"],"award-info":[{"award-number":["CEIEC-2022-ZM02-0247"]}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["IEEE Trans. Serv. Comput."],"published-print":{"date-parts":[[2024,9]]},"DOI":"10.1109\/tsc.2024.3399654","type":"journal-article","created":{"date-parts":[[2024,5,10]],"date-time":"2024-05-10T17:24:59Z","timestamp":1715361899000},"page":"2626-2639","source":"Crossref","is-referenced-by-count":17,"title":["MoESys: A Distributed and Efficient Mixture-of-Experts Training and Inference System for Internet Services"],"prefix":"10.1109","volume":"17","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-0163-2603","authenticated-orcid":false,"given":"Dianhai","family":"Yu","sequence":"first","affiliation":[{"name":"Baidu, Inc., Beijing, China"}]},{"given":"Liang","family":"Shen","sequence":"additional","affiliation":[{"name":"Baidu, Inc., Beijing, China"}]},{"given":"Hongxiang","family":"Hao","sequence":"additional","affiliation":[{"name":"Baidu, Inc., Beijing, China"}]},{"given":"Weibao","family":"Gong","sequence":"additional","affiliation":[{"name":"Baidu, Inc., Beijing, China"}]},{"ORCID":"https:\/\/orcid.org\/0009-0003-8545-7818","authenticated-orcid":false,"given":"Huachao","family":"Wu","sequence":"additional","affiliation":[{"name":"Baidu, Inc., Beijing, China"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-6997-1989","authenticated-orcid":false,"given":"Jiang","family":"Bian","sequence":"additional","affiliation":[{"name":"Baidu, Inc., Beijing, China"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-0859-2827","authenticated-orcid":false,"given":"Lirong","family":"Dai","sequence":"additional","affiliation":[{"name":"Department of Electronic Engineering and Information Science, University of Science and Technology of China, Heifei, Anhui, China"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-5451-3253","authenticated-orcid":false,"given":"Haoyi","family":"Xiong","sequence":"additional","affiliation":[{"name":"Baidu, Inc., Beijing, China"}]}],"member":"263","reference":[{"key":"ref1","doi-asserted-by":"publisher","DOI":"10.1145\/3394486.3406703"},{"key":"ref2","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV48922.2021.00438"},{"key":"ref3","doi-asserted-by":"publisher","DOI":"10.1109\/TSC.2022.3155447"},{"key":"ref4","doi-asserted-by":"publisher","DOI":"10.1109\/TSC.2023.3285935"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.1145\/3447548.3467146"},{"issue":"9","key":"ref6","first-page":"5017","article-title":"AFCS: Aggregation-free spatial-temporal mobile community sensing","volume":"22","author":"Bian","year":"2023","journal-title":"IEEE Trans. Mobile Comput."},{"key":"ref7","article-title":"BERT: Pre-training of deep bidirectional transformers for language understanding","author":"Devlin","year":"2018"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.1109\/TSC.2023.3285935"},{"key":"ref9","doi-asserted-by":"publisher","DOI":"10.1109\/TSC.2020.2964552"},{"key":"ref10","doi-asserted-by":"publisher","DOI":"10.1145\/3459637.3481902"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.1145\/3447548.3467147"},{"key":"ref12","doi-asserted-by":"publisher","DOI":"10.1109\/TSC.2019.2922177"},{"key":"ref13","first-page":"1877","article-title":"Language models are few-shot learners","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Brown"},{"key":"ref14","first-page":"1","article-title":"Efficient large-scale language model training on gpu clusters using Megatron-LM","volume-title":"Proc. Int. Conf. High Perform. Comput. Netw. Storage Anal.","author":"Narayanan"},{"key":"ref15","article-title":"ERNIE 3.0 Titan: Exploring larger-scale knowledge enhanced pre-training for language understanding and generation","author":"Wang","year":"2021"},{"key":"ref16","article-title":"Using deepspeed and megatron to train megatron-turing NLG 530b, a large-scale generative language model","author":"Smith","year":"2022"},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.1109\/TSC.2023.3241615"},{"key":"ref18","first-page":"412","article-title":"Distributed hierarchical GPU parameter server for massive scale deep learning ads systems","volume":"2","author":"Zhao","journal-title":"Proc. Mach. Learn. Syst."},{"key":"ref19","doi-asserted-by":"publisher","DOI":"10.1023\/A:1007379606734"},{"key":"ref20","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2021.naacl-main.41"},{"key":"ref21","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/N19-1388"},{"key":"ref22","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2020.emnlp-main.75"},{"key":"ref23","article-title":"Outrageously large neural networks: The sparsely-gated mixture-of-experts layer","author":"Shazeer","year":"2017"},{"key":"ref24","article-title":"GShard: Scaling giant models with conditional computation and automatic sharding","author":"Lepikhin","year":"2020"},{"key":"ref25","article-title":"Switch transformers: Scaling to trillion parameter models with simple and efficient sparsity","author":"Fedus","year":"2021"},{"key":"ref26","doi-asserted-by":"publisher","DOI":"10.1109\/TSC.2023.3253307"},{"key":"ref27","article-title":"GLaM: Efficient scaling of language models with mixture-of-experts","author":"Du","year":"2021"},{"key":"ref28","article-title":"Taming sparsely activated transformer with stochastic experts","author":"Zuo","year":"2021"},{"key":"ref29","article-title":"FastMoE: A fast mixture-of-expert training system","author":"He","year":"2021"},{"key":"ref30","doi-asserted-by":"publisher","DOI":"10.1109\/TSC.2020.2966970"},{"key":"ref31","article-title":"M6-T: Exploring sparse expert models and beyond","author":"Yang","year":"2021"},{"key":"ref32","doi-asserted-by":"publisher","DOI":"10.1109\/MM.2011.89"},{"key":"ref33","doi-asserted-by":"publisher","DOI":"10.1109\/TSC.2019.2951558"},{"key":"ref34","article-title":"BloombergGPT: A large language model for finance","author":"Wu","year":"2023"},{"key":"ref35","first-page":"6265","article-title":"Base layers: Simplifying training of large, sparse models","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Lewis"},{"key":"ref36","doi-asserted-by":"publisher","DOI":"10.1016\/j.aiopen.2021.12.003"},{"key":"ref37","article-title":"M6\u201310T: A sharing-delinking paradigm for efficient multi-trillion parameter pretraining","author":"Lin","year":"2021"},{"key":"ref38","article-title":"Scalable and efficient MoE training for multitask multilingual models","author":"Kim","year":"2021"},{"key":"ref39","article-title":"DeepSpeed-MoE: Advancing mixture-of-experts inference and training to power next-generation AI scale","author":"Rajbhandari","year":"2022"},{"key":"ref40","article-title":"Megatron-LM: Training multi-billion parameter language models using model parallelism","author":"Shoeybi","year":"2019"},{"key":"ref41","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/P19-1285"},{"key":"ref42","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/N19-4009"},{"key":"ref43","article-title":"Efficient large scale language modeling with mixtures of experts","author":"Artetxe","year":"2021"},{"key":"ref44","article-title":"Tutel: An efficient mixture-of-experts implementation for large DNN model training","year":"2021"},{"key":"ref45","doi-asserted-by":"publisher","DOI":"10.1145\/3458817.3476205"},{"key":"ref46","article-title":"Solid-state drive","year":"2022"},{"key":"ref47","doi-asserted-by":"publisher","DOI":"10.1109\/TKDE.2020.2975652"},{"key":"ref48","article-title":"Adam: A method for stochastic optimization","author":"Kingma","year":"2014"},{"key":"ref49","article-title":"Memory optimized for data-centric workloads","year":"2018"},{"key":"ref50","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-540-24571-1_60"},{"key":"ref51","doi-asserted-by":"publisher","DOI":"10.1109\/TSC.2022.3207273"},{"key":"ref52","article-title":"Pre-trained summarization distillation","author":"Shleifer","year":"2020"},{"key":"ref53","article-title":"DistilBERT, a distilled version of bert: Smaller, faster, cheaper and lighter","author":"Sanh","year":"2019"},{"key":"ref54","first-page":"336","article-title":"MLPerf training benchmark","volume":"2","author":"Mattson","year":"2020","journal-title":"Proc. Mach. Learn. Syst."},{"key":"ref55","doi-asserted-by":"publisher","DOI":"10.1109\/ICDCS.2019.00034"},{"key":"ref56","doi-asserted-by":"publisher","DOI":"10.1109\/INFCOM.2012.6195627"},{"key":"ref57","article-title":"Embrace: Accelerating sparse communication for distributed training of NLP neural networks","author":"Li","year":"2021"},{"key":"ref58","doi-asserted-by":"publisher","DOI":"10.1145\/3470496.3533727"},{"key":"ref59","article-title":"Paddlefleetx: An easy-to-use and high-performance one-stop tool for deep learning","author":"Contributors","year":"2022"},{"key":"ref60","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2017.540"},{"key":"ref61","article-title":"Decoupled weight decay regularization","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Loshchilov"}],"container-title":["IEEE Transactions on Services Computing"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx7\/4629386\/10709786\/10528887.pdf?arnumber=10528887","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2024,10,9]],"date-time":"2024-10-09T17:54:26Z","timestamp":1728496466000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/10528887\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024,9]]},"references-count":61,"journal-issue":{"issue":"5"},"URL":"https:\/\/doi.org\/10.1109\/tsc.2024.3399654","relation":{},"ISSN":["1939-1374","2372-0204"],"issn-type":[{"value":"1939-1374","type":"electronic"},{"value":"2372-0204","type":"electronic"}],"subject":[],"published":{"date-parts":[[2024,9]]}}}