基于NER微调的机器阅读理解模型

NER-PMR-large

NER-PMR-large is iitialized with PMR-large ad further fie-tued with 4 NER traiig data, amely CoNLL, WNUT17, ACE2004, ad ACE2005.

The model performace o the test sets are:

Note that the performace of RoBERTa-large ad PMR-large are sigle-task fie-tuig, while NER-PMR-large is a multi-task fie-tued model. As it is fie-tued o multiple datasets, we believe that NER-PMR-large has a better geeralizatio capability to other NER tasks tha PMR-large ad RoBERTa-large.

How to use

You ca try the codes from this repo for both traiig ad iferece.

BibTeX etry ad citatio ifo

@article{xu2022clozig,
  title={From Clozig to Comprehedig: Retrofittig Pre-traied Laguage Model to Pre-traied Machie Reader},
  author={Xu, Weiwe ad Li, Xi ad Zhag, Wexua ad Zhou, Meg ad Big, Lidog ad Lam, Wai ad Si, Luo},
  joural={arXiv preprit arXiv:2212.04755},
  year={2022}
}

NER-PMR-large NER-PMR-large is initialized with PMR-large and further fine-tuned with 4 NER training