Universal Constituent Parser
This is a project of the following constituent parsers:
The shift-reduce parser code still has not been reconstructed.
Full dataset:
python3 run/train.py --model InOrderParser
Full dataset (for more training):
python3 run/train.py --model InOrderParser --train_more --more_epoch 20
Small dataset (for code correctness test):
python3 run/train.py --model InOrderParser --train_file data/train_small.trees --dev_file data/dev_small.trees
python3 run/test.py --model InOrderParser --dev_fscore xx.xx