Matches in SemOpenAlex for { <https://semopenalex.org/work/W2913465187> ?p ?o ?g. }
- W2913465187 abstract "Pre-training of models in pruning algorithms plays an important role in pruning decision-making. We find that excessive pre-training is not necessary for pruning algorithms. According to this idea, we propose a pruning algorithm---Incremental pruning based on less training (IPLT). Compared with the traditional pruning algorithm based on a large number of pre-training, IPLT has competitive compression effect than the traditional pruning algorithm under the same simple pruning strategy. On the premise of ensuring accuracy, IPLT can achieve 8x-9x compression for VGG-19 on CIFAR-10 and only needs to pre-train few epochs. For VGG-19 on CIFAR-10, we can not only achieve 10 times test acceleration, but also about 10 times training acceleration. At present, the research mainly focuses on the compression and acceleration in the application stage of the model, while the compression and acceleration in the training stage are few. We newly proposed a pruning algorithm that can compress and accelerate in the training stage. It is novel to consider the amount of pre-training required by pruning algorithm. Our results have implications: Too much pre-training may be not necessary for pruning algorithms." @default.
- W2913465187 created "2019-02-21" @default.
- W2913465187 creator A5044367029 @default.
- W2913465187 creator A5063402368 @default.
- W2913465187 creator A5080662810 @default.
- W2913465187 date "2019-01-24" @default.
- W2913465187 modified "2023-09-27" @default.
- W2913465187 title "Really should we pruning after model be totally trained? Pruning based on a small amount of training." @default.
- W2913465187 cites W1686810756 @default.
- W2913465187 cites W2097117768 @default.
- W2913465187 cites W2104636679 @default.
- W2913465187 cites W2114766824 @default.
- W2913465187 cites W2125389748 @default.
- W2913465187 cites W2163605009 @default.
- W2913465187 cites W2194775991 @default.
- W2913465187 cites W2276892413 @default.
- W2913465187 cites W2460144244 @default.
- W2913465187 cites W2495425901 @default.
- W2913465187 cites W2515385951 @default.
- W2913465187 cites W2553910756 @default.
- W2913465187 cites W2560017826 @default.
- W2913465187 cites W2593245696 @default.
- W2913465187 cites W2604998962 @default.
- W2913465187 cites W2619307294 @default.
- W2913465187 cites W2786054724 @default.
- W2913465187 cites W2808217246 @default.
- W2913465187 cites W2949117887 @default.
- W2913465187 cites W2950967261 @default.
- W2913465187 cites W2951569836 @default.
- W2913465187 cites W2951886768 @default.
- W2913465187 cites W2951977814 @default.
- W2913465187 cites W2952344559 @default.
- W2913465187 cites W2962851801 @default.
- W2913465187 cites W2963000224 @default.
- W2913465187 cites W2963363373 @default.
- W2913465187 cites W2963674932 @default.
- W2913465187 cites W2963981420 @default.
- W2913465187 cites W2964233199 @default.
- W2913465187 cites W2964299589 @default.
- W2913465187 cites W3118608800 @default.
- W2913465187 cites W566555209 @default.
- W2913465187 hasPublicationYear "2019" @default.
- W2913465187 type Work @default.
- W2913465187 sameAs 2913465187 @default.
- W2913465187 citedByCount "1" @default.
- W2913465187 countsByYear W29134651872020 @default.
- W2913465187 crossrefType "posted-content" @default.
- W2913465187 hasAuthorship W2913465187A5044367029 @default.
- W2913465187 hasAuthorship W2913465187A5063402368 @default.
- W2913465187 hasAuthorship W2913465187A5080662810 @default.
- W2913465187 hasConcept C108010975 @default.
- W2913465187 hasConcept C11413529 @default.
- W2913465187 hasConcept C117896860 @default.
- W2913465187 hasConcept C119857082 @default.
- W2913465187 hasConcept C121332964 @default.
- W2913465187 hasConcept C138885662 @default.
- W2913465187 hasConcept C153294291 @default.
- W2913465187 hasConcept C154945302 @default.
- W2913465187 hasConcept C159985019 @default.
- W2913465187 hasConcept C180016635 @default.
- W2913465187 hasConcept C192562407 @default.
- W2913465187 hasConcept C2777211547 @default.
- W2913465187 hasConcept C2778023277 @default.
- W2913465187 hasConcept C41008148 @default.
- W2913465187 hasConcept C41895202 @default.
- W2913465187 hasConcept C6557445 @default.
- W2913465187 hasConcept C74650414 @default.
- W2913465187 hasConcept C86803240 @default.
- W2913465187 hasConceptScore W2913465187C108010975 @default.
- W2913465187 hasConceptScore W2913465187C11413529 @default.
- W2913465187 hasConceptScore W2913465187C117896860 @default.
- W2913465187 hasConceptScore W2913465187C119857082 @default.
- W2913465187 hasConceptScore W2913465187C121332964 @default.
- W2913465187 hasConceptScore W2913465187C138885662 @default.
- W2913465187 hasConceptScore W2913465187C153294291 @default.
- W2913465187 hasConceptScore W2913465187C154945302 @default.
- W2913465187 hasConceptScore W2913465187C159985019 @default.
- W2913465187 hasConceptScore W2913465187C180016635 @default.
- W2913465187 hasConceptScore W2913465187C192562407 @default.
- W2913465187 hasConceptScore W2913465187C2777211547 @default.
- W2913465187 hasConceptScore W2913465187C2778023277 @default.
- W2913465187 hasConceptScore W2913465187C41008148 @default.
- W2913465187 hasConceptScore W2913465187C41895202 @default.
- W2913465187 hasConceptScore W2913465187C6557445 @default.
- W2913465187 hasConceptScore W2913465187C74650414 @default.
- W2913465187 hasConceptScore W2913465187C86803240 @default.
- W2913465187 hasLocation W29134651871 @default.
- W2913465187 hasOpenAccess W2913465187 @default.
- W2913465187 hasPrimaryLocation W29134651871 @default.
- W2913465187 hasRelatedWork W1568757546 @default.
- W2913465187 hasRelatedWork W2041970025 @default.
- W2913465187 hasRelatedWork W2182841986 @default.
- W2913465187 hasRelatedWork W2363707704 @default.
- W2913465187 hasRelatedWork W2397170311 @default.
- W2913465187 hasRelatedWork W2896409484 @default.
- W2913465187 hasRelatedWork W2969975953 @default.
- W2913465187 hasRelatedWork W2990081069 @default.
- W2913465187 hasRelatedWork W3038185437 @default.
- W2913465187 hasRelatedWork W3103329669 @default.
- W2913465187 hasRelatedWork W3120244898 @default.