Matches in SemOpenAlex for { <https://semopenalex.org/work/W3102021588> ?p ?o ?g. }
Showing items 1 to 95 of
95
with 100 items per page.
- W3102021588 abstract "Leveraging large amounts of unlabeled data using Transformer-like architectures, like BERT, has gained popularity in recent times owing to their effectiveness in learning general representations that can then be further fine-tuned for downstream tasks to much success. However, training these models can be costly both from an economic and environmental standpoint. In this work, we investigate how to effectively use unlabeled data: by exploring the task-specific semi-supervised approach, Cross-View Training (CVT) and comparing it with task-agnostic BERT in multiple settings that include domain and task relevant English data. CVT uses a much lighter model architecture and we show that it achieves similar performance to BERT on a set of sequence tagging tasks, with lesser financial and environmental impact." @default.
- W3102021588 created "2020-11-23" @default.
- W3102021588 creator A5009821010 @default.
- W3102021588 creator A5043262011 @default.
- W3102021588 creator A5048817662 @default.
- W3102021588 creator A5059828502 @default.
- W3102021588 creator A5074277807 @default.
- W3102021588 creator A5079875307 @default.
- W3102021588 creator A5085329545 @default.
- W3102021588 date "2020-01-01" @default.
- W3102021588 modified "2023-09-23" @default.
- W3102021588 title "To BERT or Not to BERT: Comparing Task-specific and Task-agnostic Semi-Supervised Approaches for Sequence Tagging" @default.
- W3102021588 cites W1623072288 @default.
- W3102021588 cites W1670132599 @default.
- W3102021588 cites W1965555277 @default.
- W3102021588 cites W2048679005 @default.
- W3102021588 cites W2101210369 @default.
- W3102021588 cites W2113459411 @default.
- W3102021588 cites W2144578941 @default.
- W3102021588 cites W2155069789 @default.
- W3102021588 cites W2163568299 @default.
- W3102021588 cites W2250539671 @default.
- W3102021588 cites W2251648804 @default.
- W3102021588 cites W2296283641 @default.
- W3102021588 cites W2880875857 @default.
- W3102021588 cites W2891602716 @default.
- W3102021588 cites W2899519413 @default.
- W3102021588 cites W2925618549 @default.
- W3102021588 cites W2952112859 @default.
- W3102021588 cites W2962739339 @default.
- W3102021588 cites W2962741379 @default.
- W3102021588 cites W2963341956 @default.
- W3102021588 cites W2963403868 @default.
- W3102021588 cites W2963809228 @default.
- W3102021588 cites W2965373594 @default.
- W3102021588 cites W2970119519 @default.
- W3102021588 cites W2971292190 @default.
- W3102021588 cites W2978017171 @default.
- W3102021588 cites W3034238904 @default.
- W3102021588 doi "https://doi.org/10.18653/v1/2020.emnlp-main.636" @default.
- W3102021588 hasPublicationYear "2020" @default.
- W3102021588 type Work @default.
- W3102021588 sameAs 3102021588 @default.
- W3102021588 citedByCount "12" @default.
- W3102021588 countsByYear W31020215882020 @default.
- W3102021588 countsByYear W31020215882021 @default.
- W3102021588 countsByYear W31020215882022 @default.
- W3102021588 countsByYear W31020215882023 @default.
- W3102021588 crossrefType "proceedings-article" @default.
- W3102021588 hasAuthorship W3102021588A5009821010 @default.
- W3102021588 hasAuthorship W3102021588A5043262011 @default.
- W3102021588 hasAuthorship W3102021588A5048817662 @default.
- W3102021588 hasAuthorship W3102021588A5059828502 @default.
- W3102021588 hasAuthorship W3102021588A5074277807 @default.
- W3102021588 hasAuthorship W3102021588A5079875307 @default.
- W3102021588 hasAuthorship W3102021588A5085329545 @default.
- W3102021588 hasBestOaLocation W31020215881 @default.
- W3102021588 hasConcept C119857082 @default.
- W3102021588 hasConcept C127413603 @default.
- W3102021588 hasConcept C154945302 @default.
- W3102021588 hasConcept C201995342 @default.
- W3102021588 hasConcept C204321447 @default.
- W3102021588 hasConcept C2778112365 @default.
- W3102021588 hasConcept C2780451532 @default.
- W3102021588 hasConcept C41008148 @default.
- W3102021588 hasConcept C54355233 @default.
- W3102021588 hasConcept C86803240 @default.
- W3102021588 hasConceptScore W3102021588C119857082 @default.
- W3102021588 hasConceptScore W3102021588C127413603 @default.
- W3102021588 hasConceptScore W3102021588C154945302 @default.
- W3102021588 hasConceptScore W3102021588C201995342 @default.
- W3102021588 hasConceptScore W3102021588C204321447 @default.
- W3102021588 hasConceptScore W3102021588C2778112365 @default.
- W3102021588 hasConceptScore W3102021588C2780451532 @default.
- W3102021588 hasConceptScore W3102021588C41008148 @default.
- W3102021588 hasConceptScore W3102021588C54355233 @default.
- W3102021588 hasConceptScore W3102021588C86803240 @default.
- W3102021588 hasLocation W31020215881 @default.
- W3102021588 hasLocation W31020215882 @default.
- W3102021588 hasOpenAccess W3102021588 @default.
- W3102021588 hasPrimaryLocation W31020215881 @default.
- W3102021588 hasRelatedWork W2081647779 @default.
- W3102021588 hasRelatedWork W2611614995 @default.
- W3102021588 hasRelatedWork W2961085424 @default.
- W3102021588 hasRelatedWork W3080496452 @default.
- W3102021588 hasRelatedWork W3107474891 @default.
- W3102021588 hasRelatedWork W3185852197 @default.
- W3102021588 hasRelatedWork W4286629047 @default.
- W3102021588 hasRelatedWork W4306321456 @default.
- W3102021588 hasRelatedWork W4306674287 @default.
- W3102021588 hasRelatedWork W4224009465 @default.
- W3102021588 isParatext "false" @default.
- W3102021588 isRetracted "false" @default.
- W3102021588 magId "3102021588" @default.
- W3102021588 workType "article" @default.