Matches in SemOpenAlex for { <https://semopenalex.org/work/W4220856586> ?p ?o ?g. }
Showing items 1 to 64 of
64
with 100 items per page.
- W4220856586 abstract "Abstract: With the continuous development of AI technology, the training of massive data and the emergence of large-scale models have made stand-alone model training increasingly unable to meet the needs of AI applications. Distributed machine learning technologies (such as data parallelism and model parallelism) have appeared at historic moments and will have extreme large-scale application scenarios. At present, the training speed of distributed machine learning models is slow, and the scale of model parameters is still the main problem in this field. From the perspective of model parallelism, this article aims to design the optimal division method for different models under model parallelism by analyzing the structure of the existing AI application model. According to the framework structure of artificial intelligence application model, design the model optimization partition strategy and model based on parallelism. A network architecture suitable for accelerating AI application training, focusing on solving technical problems, such as network architecture design based on AI applications and model optimization and partitioning under model parallelization." @default.
- W4220856586 created "2022-04-03" @default.
- W4220856586 creator A5052528110 @default.
- W4220856586 date "2022-01-11" @default.
- W4220856586 modified "2023-09-25" @default.
- W4220856586 title "Research on Network Architecture Design Based on Artificial Intelligence Application Technology" @default.
- W4220856586 cites W2097926925 @default.
- W4220856586 cites W2183341477 @default.
- W4220856586 cites W2527036487 @default.
- W4220856586 cites W2886851211 @default.
- W4220856586 doi "https://doi.org/10.1145/3514105.3514124" @default.
- W4220856586 hasPublicationYear "2022" @default.
- W4220856586 type Work @default.
- W4220856586 citedByCount "0" @default.
- W4220856586 crossrefType "proceedings-article" @default.
- W4220856586 hasAuthorship W4220856586A5052528110 @default.
- W4220856586 hasConcept C114614502 @default.
- W4220856586 hasConcept C118524514 @default.
- W4220856586 hasConcept C120314980 @default.
- W4220856586 hasConcept C123657996 @default.
- W4220856586 hasConcept C142362112 @default.
- W4220856586 hasConcept C153349607 @default.
- W4220856586 hasConcept C154945302 @default.
- W4220856586 hasConcept C173608175 @default.
- W4220856586 hasConcept C202444582 @default.
- W4220856586 hasConcept C2781172179 @default.
- W4220856586 hasConcept C33923547 @default.
- W4220856586 hasConcept C41008148 @default.
- W4220856586 hasConcept C42812 @default.
- W4220856586 hasConcept C50644808 @default.
- W4220856586 hasConcept C61483411 @default.
- W4220856586 hasConcept C9652623 @default.
- W4220856586 hasConceptScore W4220856586C114614502 @default.
- W4220856586 hasConceptScore W4220856586C118524514 @default.
- W4220856586 hasConceptScore W4220856586C120314980 @default.
- W4220856586 hasConceptScore W4220856586C123657996 @default.
- W4220856586 hasConceptScore W4220856586C142362112 @default.
- W4220856586 hasConceptScore W4220856586C153349607 @default.
- W4220856586 hasConceptScore W4220856586C154945302 @default.
- W4220856586 hasConceptScore W4220856586C173608175 @default.
- W4220856586 hasConceptScore W4220856586C202444582 @default.
- W4220856586 hasConceptScore W4220856586C2781172179 @default.
- W4220856586 hasConceptScore W4220856586C33923547 @default.
- W4220856586 hasConceptScore W4220856586C41008148 @default.
- W4220856586 hasConceptScore W4220856586C42812 @default.
- W4220856586 hasConceptScore W4220856586C50644808 @default.
- W4220856586 hasConceptScore W4220856586C61483411 @default.
- W4220856586 hasConceptScore W4220856586C9652623 @default.
- W4220856586 hasLocation W42208565861 @default.
- W4220856586 hasOpenAccess W4220856586 @default.
- W4220856586 hasPrimaryLocation W42208565861 @default.
- W4220856586 hasRelatedWork W1496703677 @default.
- W4220856586 hasRelatedWork W1508832769 @default.
- W4220856586 hasRelatedWork W1608806855 @default.
- W4220856586 hasRelatedWork W2022931154 @default.
- W4220856586 hasRelatedWork W2023505575 @default.
- W4220856586 hasRelatedWork W2313503008 @default.
- W4220856586 hasRelatedWork W2378666660 @default.
- W4220856586 hasRelatedWork W3095555187 @default.
- W4220856586 hasRelatedWork W4240606930 @default.
- W4220856586 hasRelatedWork W4362471141 @default.
- W4220856586 isParatext "false" @default.
- W4220856586 isRetracted "false" @default.
- W4220856586 workType "article" @default.