Chinese Named Entity Recognition On Msra
Métriques
F1
Résultats
Résultats de performance de divers modèles sur ce benchmark
Tableau comparatif
| Nom du modèle | F1 |
|---|---|
| ernie-20-a-continual-pre-training-framework | 93.8 |
| nflat-non-flat-lattice-transformer-for | 94.55 |
| a-unified-mrc-framework-for-named-entity | 95.75 |
| flat-chinese-ner-using-flat-lattice | 96.09 |
| diffusionner-boundary-diffusion-for-named | 94.91 |
| simplify-the-usage-of-lexicon-in-chinese-ner | 93.5 |
| flat-chinese-ner-using-flat-lattice | 94.12 |
| glyce-glyph-vectors-for-chinese-character | 95.54 |
| fgn-fusion-glyph-network-for-chinese-named | 95.64 |
| zen-pre-training-chinese-text-encoder | 95.25 |
| ernie-20-a-continual-pre-training-framework | 95 |
| zen-pre-training-chinese-text-encoder | 93.24 |
| boundary-smoothing-for-named-entity-1 | 96.26 |
| improving-named-entity-recognition-by | 96.69 |
| tener-adapting-transformer-encoder-for-name | 92.74 |
| chinese-ner-using-lattice-lstm | 93.18 |
| unified-named-entity-recognition-as-word-word | 96.10 |
| parallel-instance-query-network-for-named | 93.48 |
| dice-loss-for-data-imbalanced-nlp-tasks | 96.72 |
| can-ner-convolutional-attention-network | 92.97 |
| ernie-enhanced-representation-through | 93.8 |