Constituency Parsing On Ctb5
Métriques
F1 score
Résultats
Résultats de performance de divers modèles sur ce benchmark
Tableau comparatif
| Nom du modèle | F1 score |
|---|---|
| fast-and-accurate-neural-crf-constituency-1 | 89.80 |
| strongly-incremental-constituency-parsing | 93.52 |
| fast-and-accurate-neural-crf-constituency-1 | 92.27 |
| constituency-parsing-with-a-self-attentive | 87.43 |
| head-driven-phrase-structure-grammar-parsing | 89.40 |
| multilingual-constituency-parsing-with-self | 91.75 |
| improving-constituency-parsing-with-span | 92.66 |
| n-ary-constituent-tree-parsing-with-recursive | 92.50 |
| to-be-continuous-or-to-be-discrete-those-are | 92.33 |