Closed hiroshinoji closed 6 years ago
New Berkeley neural parser (https://github.com/nikitakit/self-attentive-parser) is very strong; 95.07 F1 on PTB test set. We should implement the annotator for this.
New Berkeley neural parser (https://github.com/nikitakit/self-attentive-parser) is very strong; 95.07 F1 on PTB test set. We should implement the annotator for this.