YOO BYUNGHYUN

Method of domain-adapting large-capacity pre-trained language model using semantic chunk dynamic weight masking