conllpp
500CoNLLpp is a corrected version of the CoNLL2003 NER dataset where labels of 5.38% of the sentences in the test set have been manually corrected. The training set and developme
CoNLLpp is a corrected version of the CoNLL2003 NER dataset where labels of 5.38% of the sentences in the test set have been manually corrected. The training set and developme
XGLUE is a new benchmark dataset to evaluate the performance of cross-lingual pre-trained models with respect to cross-lingual natural language understanding and generation. T