ZEN: Pre-training Chinese Text Encoder Enhanced by N-gram Representations

Open paper
Benchmark Model Rank Results
chinese-named-entity-recognition-on-msraZEN (Init with Chinese BERT)#9F1: 95.25
chinese-named-entity-recognition-on-msraZEN (Random Init)#18F1: 93.24
chinese-word-segmentation-on-msrZEN (Init with Chinese BERT)#4F1: 98.35
chinese-word-segmentation-on-msrZEN (Random Init)#6F1: 97.89

Rank counts only results with a code link.