Hello, not sure if this is the right place to ask the question, but here it is.
So, i am trying to start the chat and load all models that i generated.
I wanted to separate the training data by its meaning. This is the structure
├── data
│ ├── core
│ │ ├── basic_stories.md
│ │ └── stories.md
│ ├── nlu
│ │ ├── basic
│ │ │ ├── basic_test.md
│ │ │ └── basic_train.md
│ │ ├── glossary
│ │ │ ├── glossary.txt
│ │ │ ├── glossary_test.md
│ │ │ └── glossary_train.md
│ │ └── name
│ │ ├── name.txt
│ │ ├── name_test.md
│ │ └── name_train.md
So when i train the data, under models i have project current/ and inside default (which are nlu models) and dialogue which is rasa core. Now i am trying to run the chat but i have a hard time figuring out how to load all models from that directory models/current/default. Structure:
├── models
│ └── current
│ ├── default
│ │ ├── basic
│ │ │ ├── metadata.json
│ │ │ ├── regex_featurizer.json
│ │ │ └── training_data.json
│ │ ├── glossary
│ │ │ ├── crf_model.pkl
│ │ │ ├── intent_classifier_sklearn.pkl
│ │ │ ├── metadata.json
│ │ │ ├── regex_featurizer.json
│ │ │ └── training_data.json
│ │ └── name
│ │ ├── crf_model.pkl
│ │ ├── metadata.json
│ │ ├── regex_featurizer.json
│ │ └── training_data.json
I am able to run each model separately. This is the code:
def run_service_bot(serve_forever=True):
interpreter = RasaNLUInterpreter('./models/current/default/basic')
action_endpoint = EndpointConfig(url="http://localhost:5055/webhook")
agent = Agent.load('./models/current/dialogue', interpreter=interpreter, action_endpoint=action_endpoint)
rasa_core.run.serve_application(agent, channel='cmdline')
return agent
I tried few ways, but not really successfully. Is there something i am missing? Any help would be more then appreciated.
Thanks