DC Field | Value | Language |
---|---|---|
dc.contributor.advisor | Choo, Jaegul | - |
dc.contributor.advisor | 주재걸 | - |
dc.contributor.author | Kim, Taehee | - |
dc.date.accessioned | 2023-06-22T19:31:21Z | - |
dc.date.available | 2023-06-22T19:31:21Z | - |
dc.date.issued | 2022 | - |
dc.identifier.uri | http://library.kaist.ac.kr/search/detail/view.do?bibCtrlNo=997672&flag=dissertation | en_US |
dc.identifier.uri | http://hdl.handle.net/10203/308206 | - |
dc.description | 학위논문(석사) - 한국과학기술원 : 김재철AI대학원, 2022.2,[iii, 17 p. :] | - |
dc.description.abstract | During the fine-tuning phase of transfer learning, the pretrained vocabulary remains unchanged, while model parameters are updated. The vocabulary generated based on the pretrained data is suboptimal for downstream data when domain discrepancy exists. We propose to consider the vocabulary as an optimizable parameter, allowing us to update the vocabulary by expanding it with domain-specific vocabulary based on a tokenization statistic. Furthermore, we preserve the embeddings of the added words from overfitting to downstream data by utilizing knowledge learned from a pretrained language model with a regularization term. Our method achieved consistent performance improvements on diverse domains (i.e., biomedical, computer science, news, and reviews). | - |
dc.language | eng | - |
dc.publisher | 한국과학기술원 | - |
dc.title | AVocaDo: strategy for adapting vocabulary to downstream domain | - |
dc.title.alternative | 도메인 특화 어휘 생성 연구 | - |
dc.type | Thesis(Master) | - |
dc.identifier.CNRN | 325007 | - |
dc.description.department | 한국과학기술원 :김재철AI대학원, | - |
dc.contributor.alternativeauthor | 김태희 | - |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.