Matches in SemOpenAlex for { <https://semopenalex.org/work/W3132060159> ?p ?o ?g. }
Showing items 1 to 83 of
83
with 100 items per page.
- W3132060159 endingPage "44058" @default.
- W3132060159 startingPage "44049" @default.
- W3132060159 abstract "Deep neural networks have become one of the popular techniques used in many research and application areas including computer vision, natural language processing, etc. As the complexity of neural networks continuously increasing, the training process takes a much longer time and requires more computation resources. To speed up the training process, a centralized distributed training structure named Parameter Server (PS) is widely used to assign training tasks to different workers/nodes. Most existing studies considered all workers having the same computation resources. However, in a heterogeneous environment, fast workers (i.e., workers having more computation resources) can complete tasks earlier than slow workers and thus the system does not fully utilize the resources of fast workers. In this paper, we propose a PS model with heterogeneous types of workers/nodes, called H-PS, which can fully utilize the resources of each worker by dynamically scheduling tasks based on the current status of the workers (e.g., available memory). By doing so, the workers will complete their tasks at the same time and the stragglers (i.e., workers fall behind others) can be avoided. In addition, a pipeline scheme is proposed to further improve the effectiveness of workers by fully utilizing the resources of workers during the time of parameters transmitting between PS and workers. Moreover, a flexible quantization scheme is proposed to reduce the communication overhead between the PS and workers. Finally, the H-PS is implemented using Containers which is an emerging lightweight technology. The experimental results indicate that the proposed H-PS can reduce the overall training time by 1.4x – 3.5x when compared with existing methods." @default.
- W3132060159 created "2021-03-01" @default.
- W3132060159 creator A5035282947 @default.
- W3132060159 creator A5048972267 @default.
- W3132060159 creator A5049110043 @default.
- W3132060159 creator A5052292460 @default.
- W3132060159 creator A5082012882 @default.
- W3132060159 date "2021-01-01" @default.
- W3132060159 modified "2023-10-12" @default.
- W3132060159 title "H-PS: A Heterogeneous-Aware Parameter Server With Distributed Neural Network Training" @default.
- W3132060159 cites W114517082 @default.
- W3132060159 cites W1982003698 @default.
- W3132060159 cites W2060393849 @default.
- W3132060159 cites W2087402357 @default.
- W3132060159 cites W2276486856 @default.
- W3132060159 cites W2293282867 @default.
- W3132060159 cites W2344028138 @default.
- W3132060159 cites W2523435939 @default.
- W3132060159 cites W2579247884 @default.
- W3132060159 cites W2612026221 @default.
- W3132060159 cites W2613322955 @default.
- W3132060159 cites W2895891814 @default.
- W3132060159 cites W2919115771 @default.
- W3132060159 cites W2929084559 @default.
- W3132060159 cites W2963122961 @default.
- W3132060159 cites W2963988417 @default.
- W3132060159 cites W2964233199 @default.
- W3132060159 cites W2964923275 @default.
- W3132060159 cites W2998174899 @default.
- W3132060159 doi "https://doi.org/10.1109/access.2021.3060154" @default.
- W3132060159 hasPublicationYear "2021" @default.
- W3132060159 type Work @default.
- W3132060159 sameAs 3132060159 @default.
- W3132060159 citedByCount "2" @default.
- W3132060159 countsByYear W31320601592022 @default.
- W3132060159 countsByYear W31320601592023 @default.
- W3132060159 crossrefType "journal-article" @default.
- W3132060159 hasAuthorship W3132060159A5035282947 @default.
- W3132060159 hasAuthorship W3132060159A5048972267 @default.
- W3132060159 hasAuthorship W3132060159A5049110043 @default.
- W3132060159 hasAuthorship W3132060159A5052292460 @default.
- W3132060159 hasAuthorship W3132060159A5082012882 @default.
- W3132060159 hasBestOaLocation W31320601591 @default.
- W3132060159 hasConcept C120314980 @default.
- W3132060159 hasConcept C121332964 @default.
- W3132060159 hasConcept C153294291 @default.
- W3132060159 hasConcept C154945302 @default.
- W3132060159 hasConcept C2777211547 @default.
- W3132060159 hasConcept C31258907 @default.
- W3132060159 hasConcept C41008148 @default.
- W3132060159 hasConcept C50644808 @default.
- W3132060159 hasConcept C93996380 @default.
- W3132060159 hasConceptScore W3132060159C120314980 @default.
- W3132060159 hasConceptScore W3132060159C121332964 @default.
- W3132060159 hasConceptScore W3132060159C153294291 @default.
- W3132060159 hasConceptScore W3132060159C154945302 @default.
- W3132060159 hasConceptScore W3132060159C2777211547 @default.
- W3132060159 hasConceptScore W3132060159C31258907 @default.
- W3132060159 hasConceptScore W3132060159C41008148 @default.
- W3132060159 hasConceptScore W3132060159C50644808 @default.
- W3132060159 hasConceptScore W3132060159C93996380 @default.
- W3132060159 hasFunder F4320313890 @default.
- W3132060159 hasFunder F4320321001 @default.
- W3132060159 hasLocation W31320601591 @default.
- W3132060159 hasOpenAccess W3132060159 @default.
- W3132060159 hasPrimaryLocation W31320601591 @default.
- W3132060159 hasRelatedWork W1547841080 @default.
- W3132060159 hasRelatedWork W1556909862 @default.
- W3132060159 hasRelatedWork W1583717361 @default.
- W3132060159 hasRelatedWork W1587227328 @default.
- W3132060159 hasRelatedWork W1701020922 @default.
- W3132060159 hasRelatedWork W2130966263 @default.
- W3132060159 hasRelatedWork W2315154691 @default.
- W3132060159 hasRelatedWork W2371023786 @default.
- W3132060159 hasRelatedWork W2386387936 @default.
- W3132060159 hasRelatedWork W2998813341 @default.
- W3132060159 hasVolume "9" @default.
- W3132060159 isParatext "false" @default.
- W3132060159 isRetracted "false" @default.
- W3132060159 magId "3132060159" @default.
- W3132060159 workType "article" @default.