Cross-lingual Paraphrase Identification on PAWS-X benchmark leaderboard · SOTA2 Research

Cross-lingual Paraphrase Identification on PAWS-X

0.965Accuracy (en)ERNIE-M_LARGE

Evaluation Results

Method	Links
ERNIE-M_LARGE2020.12		0.965	0.935	0.933	0.938	0.879	0.884	0.892	0.918
VECO_LARGE2020.12		0.964	0.93	0.93	0.935	0.872	0.868	0.879	0.911
VECO_LARGE2020.12		0.962	0.913	0.914	0.92	0.818	0.829	0.851	0.887
ERNIE-M_LARGE2020.12		0.96	0.919	0.914	0.922	0.839	0.845	0.869	0.895
XLM-R_LARGE2020.12		0.947	0.897	0.901	0.904	0.787	0.79	0.823	0.864
mBERT2020.12		0.94	0.857	0.874	0.87	0.73	0.696	0.77	0.819
XLM2020.12		0.94	0.859	0.883	0.874	0.693	0.648	0.765	0.809
MMTE2020.12		0.931	0.851	0.872	0.869	0.72	0.692	0.759	0.813