How to use multiple files in a directory to train nlu model using rasa nlu http server

We can store training data in multiple files. And while training we can pass directory name containing this files in script.

python -m rasa_nlu.train --config config.yml --data nlu_data/ --path projects

Above training works perfectly. But when we try to train model over Rasa NLU Http Server using POST /train, we have to write training data in yml file.

language: "en"

pipeline: "spacy_sklearn"

# data contains the same json, as described in the training data section
data: {
  "rasa_nlu_data": {
    "common_examples": [
      {
        "text": "hey",
        "intent": "greet",
        "entities": []
      }
    ]
  }
}

In our case, we have training data in multiple files. So, Is it possible to pass directory path in training .yml file? By giving directory name, we can access all training files to train a model using Rasa NLU Http Server. Something like this:

language: "en"

pipeline: "spacy_sklearn"

# data contains the same json, as described in the training data section
data: nlu_data/

Is there any other way, we can achieve this case??

Any suggestions are welcome.

Thanks in advance.

hey @purvamestry25 you can do so, you can visit the below link for more details,

Training Data Format

Hey @JiteshGaikwad If I want it in a URL format as POST request how will I do it???

Will it be like this? http://localhost:5000/train?project=test&model=test&data=test.json

Where my dataset (test.json) is like following:

hey @Kavita sorry for the late reply.

I haven’t worked around the above usecase so i dont have any idea whether it will work or not,i will once test it and tell you :smiley:

Sure. Thank you. I cannot figure out how to train a Rasa NLU model using a URL (HTTP API)

hey @Kavita i had tried something to what you had asked just check out this & tell me is it what you were looking for :slight_smile:

Hey @JiteshGaikwad
Ya this is what I was looking for. Thanks a lot. :slight_smile: