from __future__ import unicode_literals

from rasa_nlu.converters import load_data

from rasa_nlu.config import RasaNLUConfig # to load rasa configuration files

from rasa_nlu.model import Trainer #Trainer will do the training

from rasa_nlu.model import Metadata, Interpreter

def train_nlu(data,config,model_dir):

model_directory=trainer.persist(model_dir, fixed_model_name='foodnlu')

def run_nlu():

interpreter=Interpreter.load('./models/nlu/default/foodnlu', RasaNLUConfig('config_spacy.json'))

print(interpreter.parse(u"i would like to have a fish"))

#train_nlu('./data/training_data.json','config_spacy.json','./models/nlu')

run_nlu()