Matches in SemOpenAlex for { <https://semopenalex.org/work/W4221062244> ?p ?o ?g. }
Showing items 1 to 96 of
96
with 100 items per page.
- W4221062244 abstract "Auto-tuning DL compilers are gaining ground as an optimizing back-end for DL frameworks. While existing work can generate deep learning models that exceed the performance of hand-tuned libraries, they still suffer from prohibitively long auto-tuning time due to repeated hardware measurements in large search spaces. In this paper, we take a neural-predictor inspired approach to reduce the auto-tuning overhead and show that a performance predictor model trained prior to compilation can produce optimized tensor operation codes without repeated search and hardware measurements. To generate a sample-efficient training dataset, we extend input representation to include task-specific information and to guide data sampling methods to focus on learning high-performing codes. We evaluated the resulting predictor model, One-Shot Tuner, against AutoTVM and other prior work, and the results show that One-Shot Tuner speeds up compilation by 2.81x to 67.7x compared to prior work while providing comparable or improved inference time for CNN and Transformer models." @default.
- W4221062244 created "2022-04-03" @default.
- W4221062244 creator A5005051912 @default.
- W4221062244 creator A5011358342 @default.
- W4221062244 creator A5064884832 @default.
- W4221062244 date "2022-03-18" @default.
- W4221062244 modified "2023-10-14" @default.
- W4221062244 title "One-shot tuner for deep learning compilers" @default.
- W4221062244 cites W2042992924 @default.
- W4221062244 cites W2055312318 @default.
- W4221062244 cites W2072040441 @default.
- W4221062244 cites W2099001231 @default.
- W4221062244 cites W2100218206 @default.
- W4221062244 cites W2108862644 @default.
- W4221062244 cites W2135682468 @default.
- W4221062244 cites W2136952590 @default.
- W4221062244 cites W2147370410 @default.
- W4221062244 cites W2155893237 @default.
- W4221062244 cites W2156560068 @default.
- W4221062244 cites W2441512324 @default.
- W4221062244 cites W2805566098 @default.
- W4221062244 cites W2961619211 @default.
- W4221062244 cites W2963163009 @default.
- W4221062244 cites W2963446712 @default.
- W4221062244 cites W3012249773 @default.
- W4221062244 cites W3200950867 @default.
- W4221062244 doi "https://doi.org/10.1145/3497776.3517774" @default.
- W4221062244 hasPublicationYear "2022" @default.
- W4221062244 type Work @default.
- W4221062244 citedByCount "4" @default.
- W4221062244 countsByYear W42210622442022 @default.
- W4221062244 countsByYear W42210622442023 @default.
- W4221062244 crossrefType "proceedings-article" @default.
- W4221062244 hasAuthorship W4221062244A5005051912 @default.
- W4221062244 hasAuthorship W4221062244A5011358342 @default.
- W4221062244 hasAuthorship W4221062244A5064884832 @default.
- W4221062244 hasBestOaLocation W42210622441 @default.
- W4221062244 hasConcept C108583219 @default.
- W4221062244 hasConcept C113775141 @default.
- W4221062244 hasConcept C119857082 @default.
- W4221062244 hasConcept C120665830 @default.
- W4221062244 hasConcept C121332964 @default.
- W4221062244 hasConcept C154945302 @default.
- W4221062244 hasConcept C162324750 @default.
- W4221062244 hasConcept C165801399 @default.
- W4221062244 hasConcept C169590947 @default.
- W4221062244 hasConcept C187736073 @default.
- W4221062244 hasConcept C192209626 @default.
- W4221062244 hasConcept C199360897 @default.
- W4221062244 hasConcept C2776214188 @default.
- W4221062244 hasConcept C2779960059 @default.
- W4221062244 hasConcept C2780451532 @default.
- W4221062244 hasConcept C41008148 @default.
- W4221062244 hasConcept C62520636 @default.
- W4221062244 hasConcept C66322947 @default.
- W4221062244 hasConcept C74064498 @default.
- W4221062244 hasConcept C76155785 @default.
- W4221062244 hasConcept C9819579 @default.
- W4221062244 hasConceptScore W4221062244C108583219 @default.
- W4221062244 hasConceptScore W4221062244C113775141 @default.
- W4221062244 hasConceptScore W4221062244C119857082 @default.
- W4221062244 hasConceptScore W4221062244C120665830 @default.
- W4221062244 hasConceptScore W4221062244C121332964 @default.
- W4221062244 hasConceptScore W4221062244C154945302 @default.
- W4221062244 hasConceptScore W4221062244C162324750 @default.
- W4221062244 hasConceptScore W4221062244C165801399 @default.
- W4221062244 hasConceptScore W4221062244C169590947 @default.
- W4221062244 hasConceptScore W4221062244C187736073 @default.
- W4221062244 hasConceptScore W4221062244C192209626 @default.
- W4221062244 hasConceptScore W4221062244C199360897 @default.
- W4221062244 hasConceptScore W4221062244C2776214188 @default.
- W4221062244 hasConceptScore W4221062244C2779960059 @default.
- W4221062244 hasConceptScore W4221062244C2780451532 @default.
- W4221062244 hasConceptScore W4221062244C41008148 @default.
- W4221062244 hasConceptScore W4221062244C62520636 @default.
- W4221062244 hasConceptScore W4221062244C66322947 @default.
- W4221062244 hasConceptScore W4221062244C74064498 @default.
- W4221062244 hasConceptScore W4221062244C76155785 @default.
- W4221062244 hasConceptScore W4221062244C9819579 @default.
- W4221062244 hasFunder F4320335489 @default.
- W4221062244 hasLocation W42210622441 @default.
- W4221062244 hasOpenAccess W4221062244 @default.
- W4221062244 hasPrimaryLocation W42210622441 @default.
- W4221062244 hasRelatedWork W2096322617 @default.
- W4221062244 hasRelatedWork W2096827166 @default.
- W4221062244 hasRelatedWork W2168846948 @default.
- W4221062244 hasRelatedWork W2185884583 @default.
- W4221062244 hasRelatedWork W2549354931 @default.
- W4221062244 hasRelatedWork W2552608914 @default.
- W4221062244 hasRelatedWork W2559413996 @default.
- W4221062244 hasRelatedWork W2559844932 @default.
- W4221062244 hasRelatedWork W4290057712 @default.
- W4221062244 hasRelatedWork W4385018252 @default.
- W4221062244 isParatext "false" @default.
- W4221062244 isRetracted "false" @default.
- W4221062244 workType "article" @default.