Matches in SemOpenAlex for { <https://semopenalex.org/work/W3106556141> ?p ?o ?g. }
Showing items 1 to 90 of
90
with 100 items per page.
- W3106556141 abstract "Training neural networks with large batch is of fundamental significance to deep learning. Large batch training remarkably reduces the amount of training time but has difficulties in maintaining accuracy. Recent works have put forward optimization methods such as LARS and LAMB to tackle this issue through adaptive layer-wise optimization using trust ratios. Though prevailing, such methods are observed to still suffer from unstable and extreme trust ratios which degrades performance. In this paper, we propose a new variant of LAMB, called LAMBC, which employs trust ratio clipping to stabilize its magnitude and prevent extreme values. We conducted experiments on image classification tasks such as ImageNet and CIFAR-10 and our empirical results demonstrate promising improvements across different batch sizes." @default.
- W3106556141 created "2020-12-07" @default.
- W3106556141 creator A5011477209 @default.
- W3106556141 creator A5042416229 @default.
- W3106556141 creator A5085462629 @default.
- W3106556141 date "2020-11-27" @default.
- W3106556141 modified "2023-09-30" @default.
- W3106556141 title "Improving Layer-wise Adaptive Rate Methods using Trust Ratio Clipping." @default.
- W3106556141 cites W1522301498 @default.
- W3106556141 cites W1598866093 @default.
- W3106556141 cites W2083842231 @default.
- W3106556141 cites W2108598243 @default.
- W3106556141 cites W2194775991 @default.
- W3106556141 cites W2524365899 @default.
- W3106556141 cites W2622263826 @default.
- W3106556141 cites W2757910899 @default.
- W3106556141 cites W2945785363 @default.
- W3106556141 cites W2949935872 @default.
- W3106556141 cites W2963341956 @default.
- W3106556141 cites W2973727699 @default.
- W3106556141 cites W2977720775 @default.
- W3106556141 cites W3025935268 @default.
- W3106556141 cites W3035172746 @default.
- W3106556141 cites W3103105237 @default.
- W3106556141 cites W3118608800 @default.
- W3106556141 hasPublicationYear "2020" @default.
- W3106556141 type Work @default.
- W3106556141 sameAs 3106556141 @default.
- W3106556141 citedByCount "0" @default.
- W3106556141 crossrefType "posted-content" @default.
- W3106556141 hasAuthorship W3106556141A5011477209 @default.
- W3106556141 hasAuthorship W3106556141A5042416229 @default.
- W3106556141 hasAuthorship W3106556141A5085462629 @default.
- W3106556141 hasConcept C119857082 @default.
- W3106556141 hasConcept C121332964 @default.
- W3106556141 hasConcept C138885662 @default.
- W3106556141 hasConcept C153180895 @default.
- W3106556141 hasConcept C153294291 @default.
- W3106556141 hasConcept C154945302 @default.
- W3106556141 hasConcept C159985019 @default.
- W3106556141 hasConcept C192562407 @default.
- W3106556141 hasConcept C2776848632 @default.
- W3106556141 hasConcept C2777211547 @default.
- W3106556141 hasConcept C2779227376 @default.
- W3106556141 hasConcept C2984842247 @default.
- W3106556141 hasConcept C41008148 @default.
- W3106556141 hasConcept C41895202 @default.
- W3106556141 hasConcept C50644808 @default.
- W3106556141 hasConceptScore W3106556141C119857082 @default.
- W3106556141 hasConceptScore W3106556141C121332964 @default.
- W3106556141 hasConceptScore W3106556141C138885662 @default.
- W3106556141 hasConceptScore W3106556141C153180895 @default.
- W3106556141 hasConceptScore W3106556141C153294291 @default.
- W3106556141 hasConceptScore W3106556141C154945302 @default.
- W3106556141 hasConceptScore W3106556141C159985019 @default.
- W3106556141 hasConceptScore W3106556141C192562407 @default.
- W3106556141 hasConceptScore W3106556141C2776848632 @default.
- W3106556141 hasConceptScore W3106556141C2777211547 @default.
- W3106556141 hasConceptScore W3106556141C2779227376 @default.
- W3106556141 hasConceptScore W3106556141C2984842247 @default.
- W3106556141 hasConceptScore W3106556141C41008148 @default.
- W3106556141 hasConceptScore W3106556141C41895202 @default.
- W3106556141 hasConceptScore W3106556141C50644808 @default.
- W3106556141 hasLocation W31065561411 @default.
- W3106556141 hasOpenAccess W3106556141 @default.
- W3106556141 hasPrimaryLocation W31065561411 @default.
- W3106556141 hasRelatedWork W1848761947 @default.
- W3106556141 hasRelatedWork W2054157877 @default.
- W3106556141 hasRelatedWork W2621265919 @default.
- W3106556141 hasRelatedWork W2907264311 @default.
- W3106556141 hasRelatedWork W2915225728 @default.
- W3106556141 hasRelatedWork W2979999556 @default.
- W3106556141 hasRelatedWork W2981924372 @default.
- W3106556141 hasRelatedWork W3002842489 @default.
- W3106556141 hasRelatedWork W3028763319 @default.
- W3106556141 hasRelatedWork W3033041687 @default.
- W3106556141 hasRelatedWork W3034361719 @default.
- W3106556141 hasRelatedWork W3035825996 @default.
- W3106556141 hasRelatedWork W3089774673 @default.
- W3106556141 hasRelatedWork W3091918696 @default.
- W3106556141 hasRelatedWork W3100593864 @default.
- W3106556141 hasRelatedWork W3128963076 @default.
- W3106556141 hasRelatedWork W3137281154 @default.
- W3106556141 hasRelatedWork W3157623825 @default.
- W3106556141 hasRelatedWork W3161019670 @default.
- W3106556141 hasRelatedWork W3176411177 @default.
- W3106556141 isParatext "false" @default.
- W3106556141 isRetracted "false" @default.
- W3106556141 magId "3106556141" @default.
- W3106556141 workType "article" @default.