While enabling API for the trained models - UserWarning: Could not load local model in 'models'. warnings.warn(f"Could not load local model in '{model_path}'.")

Rasa Open Source
1

I have created models for different training datasets and stored the models in the models library. While trying to enable the API to test it. User warning comes up as below.

No chat connector configured, falling back to the REST input channel. To connect your bot to another channel, read the docs here: Connecting to Messaging and Voice Channels 2020-04-08 17:43:22 INFO root - Starting Rasa server on http://localhost:5005 /usr/local/lib/python3.7/site-packages/rasa/core/agent.py:917: UserWarning: Could not load local model in ‘models’. warnings.warn(f"Could not load local model in ‘{model_path}’.")

Though this is a warning I tried to hit via the postman Code Snipet

    print("IN-Sentence: " + in_sentence)
    x = {
        "text": in_sentence
        }
    y = json.dumps(x)
    payload = y
    print("payload = " + payload)
    # parsed = interpreter.parse(sentence['initial'])
    url = "http://localhost:5005/model/parse"

    # payload = "{\"text\":\"sentence['initial']\"}"
    headers = {
        'cache-control': "no-cache",
        'Postman-Token': "6f25eb8d-6233-4aab-ab5a-71ee76b3d364"
    }

    response = requests.request("POST", url, data=payload, headers=headers)

    outcome = response.text
    print (outcome)
    data = json.loads(outcome)
    print("Out intent: " + data["intent"]["name"])

Sample output for the above code:

IN-Sentence: She needs blazers. payload = {“text”: “She needs blazers.”} {“text”:“/She needs blazers.”,“intent”:{“name”:“She needs blazers.”,“confidence”:1.0},“intent_ranking”:[{“name”:“She needs blazers.”,“confidence”:1.0}],“entities”:} Out intent: She needs blazers.

This sentence has been trained in the intent as Buy clothing - But it returns intent name as the sentence as it self.

I have done the same activity before 3 months back but not sure what has gone wrong. My RASA Setup as below. rasa 1.5.3
rasa-core 0.14.5
rasa-core-sdk 0.14.0
rasa-nlu 0.13.0
rasa-sdk 1.5.2
rasa-x 0.23.5

Config as

Configuration for Rasa NLU.

Components

language: en pipeline: supervised_embeddings

Configuration for Rasa Core.

Policies

policies:

  • name: MemoizationPolicy
  • name: KerasPolicy
  • name: MappingPolicy

Please help.

Thanks Kishore