Rasa: Unable to initialize persistor

[nahidalam]

Rasa NLU version:0.12.3

Operating system (windows, osx, ...):OSx

Content of model configuration file:

language: "en"

pipeline: "spacy_sklearn"

Issue:
I am trying to run the Rasa NLU server and communicate to it through POST to extract intent but getting error Unable to initialize persistor

Here is what I do:

I am training the nlu model like below:

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

if __name__ == '__main__':
    model_directory = train_nlu('./data/data.json', 'config_spacy.yml', './models/nlu', 'renewalnlu')

Then in the command line

Running the NLU server

$python -m rasa_nlu.server --path /PATH/renewalBot/models/nlu

And
$curl --request POST --url http://localhost:5000/parse --header 'content-type: application/json' --data '{ "q": "hello"}' | python -mjson.tool

This is returning error Unable to initialize persistor

Any suggestion?

[amn41]

where exactly is your model saved? I suspect the path you provide isn't quite right,

if you model is at e.g. PATH/renewalBot/models/nlu/model_xx then you would want --path PATH/renewalBot/models and then specify project nlu and model model_xx

[nahidalam]

@amn41
NLU Model at PATH/renewalBot/models/nlu/default/renewalnlu

Essentially my project name is default

So what should be the entire command?

[amn41]

should then be --path PATH/renewalBot/models/nlu/

and query:

$curl --request POST --url http://localhost:5000/parse?model=renewalnlu --header 'content-type: application/json' --data '{ "q": "hello"}' | python -mjson.tool

[nahidalam]

@amn41 Still getting the same error

[amn41]

what's the output of curl localhost:5000/status ?

[nahidalam]

@amn41 below

{
  "available_projects": {
    "default": {
      "status": "ready",
      "available_models": [
        "renewalnlu"
      ],
      "loaded_models": []
    }
  }

[amn41]

that's odd! I just reproduced your setup on my machine and can't reproduce the error (on v0.13.1). is this still happening? Could you please update your version?

[FNj]

Similar error is still happening to me as well.

[stale[bot]]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

[stale[bot]]

This issue has been automatically closed due to inactivity. Please create a new issue if you need more help.

[tesssie]

I am also facing same issue
{
"max_training_processes": 1,
"current_training_processes": 0,
"available_projects": {
"current": {
"status": "ready",
"current_training_processes": 0,
"available_models": [
"dialogue"
],
"loaded_models": []
}
}
}

[aqiank]

Happens to me as well. It seems that this happens if I use --fixed_model_name [name] when training.

Version is rasa_nlu==0.14.6.

[unrealnerd]

Thank you @aqiank

[srikar96]

@aqiank I get the same error if I use --fixed_model_name [name]. If I remove that, it's working. I am using version 0.15.0.