Yiwen-Yang-666 / GAN-BERT-CRF

An idea that take advantages of features of deep learning to use unannotated samples for NER and identify sequences with error labels.
15Updated 7 months ago

Related projects: