rasa nlu multiple models python

Question

def train_nlu(data, configs, model_dir):
    training_data = load_data(data)
    trainer = Trainer(config.load(configs))
    trainer.train(training_data)
    model_directory = trainer.persist(model_dir, fixed_model_name='weathernlu')
    return model_directory


def run_nlu(model_dir):
    interpreter = Interpreter.load(model_dir)
    print(interpreter.parse("hello"))

I want to load multiple models to run. how can I use Interpreter to load multiple models in my python program?

score 0 · Answer 1 · answered Jun 10 '18 at 13:12

0

You can simply store the different models in different directories, and then load the two different models from their respective directories.

def run_nlu(model_dir):
    interpreter1 = Interpreter.load(model1_dir)
    print(interpreter.parse("hello"))
    interpreter2 = Interpreter.load(model2_dir)
    print(interpreter.parse("hello"))

answered Jun 10 '18 at 13:12

Parth Sharma

441
4
19

I mean load multiple models at once,Not every time load a model – luffysup Jun 14 '18 at 01:21
I'm sorry I don't understand what you mean by that. Can you clarify? What do you mean by >load multiple models at once ? – Parth Sharma Jun 14 '18 at 13:46

score 0 · Accepted Answer · answered Jun 16 '18 at 10:06

If you're using Rasa NLU 0.12.3, you can use the method merge of class TrainingData. For example

from rasa_nlu.training_data import TrainingData, load_data
from rasa_nlu.model import Trainer
from rasa_nlu import config

training_data = TrainingData()
nlu_trainings = ["data/examples/domain1.md", "data/examples/domain2.md"]
for nlu_training in nlu_trainings:
  training_data = training_data.merge(load_data(nlu_training)))

trainer = Trainer(config.load("sample_configs/config_spacy.yaml"))
trainer.train(training_data)
trainer.persist("./projects/default/")

rasa nlu multiple models python

2 Answers2