DC Field | Value | Language |
---|---|---|
dc.contributor.advisor | 주재걸 | - |
dc.contributor.author | Lee, Koanho | - |
dc.contributor.author | 이관호 | - |
dc.date.accessioned | 2024-07-30T19:30:39Z | - |
dc.date.available | 2024-07-30T19:30:39Z | - |
dc.date.issued | 2024 | - |
dc.identifier.uri | http://library.kaist.ac.kr/search/detail/view.do?bibCtrlNo=1096065&flag=dissertation | en_US |
dc.identifier.uri | http://hdl.handle.net/10203/321360 | - |
dc.description | 학위논문(석사) - 한국과학기술원 : 김재철AI대학원, 2024.2,[iv, 25 p. :] | - |
dc.description.abstract | Recent studies have found that language models (LMs) pretrained on enormous amounts of text corpora can store world knowledge in their internal parameters. Despite its benefits, however, the knowledge stored in LMs can be easily outdated as the world evolves over time. In this study, we introduce a novel framework for lifelong pretraining of LMs, based on the concept of knowledge distillation. Specifically, our framework adjusts how much to distill the knowledge from a teacher’s prediction by considering its reliability. Furthermore, we demonstrate that the student model can effectively serve as its own teacher, generating highly valuable labels for training. Experiments on multiple benchmarks confirm the effectiveness and validity of our framework. | - |
dc.language | eng | - |
dc.publisher | 한국과학기술원 | - |
dc.subject | 지속 학습▼a언어 모델▼a지식 증류▼a시간적 불일치▼a자연어 처리 | - |
dc.subject | Continual learning▼aLanguage model▼aKnowledge distillation▼aTemporal misalignment▼aNatural language processing | - |
dc.title | Distill your own knowledge: towards ever-evolving language models via online self-distillation | - |
dc.title.alternative | 자가 증류를 활용한 언어 모델의 지속적인 학습 | - |
dc.type | Thesis(Master) | - |
dc.identifier.CNRN | 325007 | - |
dc.description.department | 한국과학기술원 :김재철AI대학원, | - |
dc.contributor.alternativeauthor | Choo, Jaegul | - |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.