vidu171
January 31, 2020, 8:42am
1
Hi,
I have been trying to run rasa_nlu.server and I am able to train the data
Here is the data I am sending as a post request to /train?project=p1
{ "rasa_nlu_data": { "common_examples": [ { "text": "ola","intent": "greetings", "entities": [] }, { "text": "ola","intent": "greetings", "entities": [] }, { "text": "ola","intent": "greetings", "entities": [] } ] } }
And the model is being successfully trained.
But when I try the parse api ( /parse )
{"q":"ola","project":"p1"}
I receive the flowing response
{ "intent": null, "entities": [], "intent_ranking": [], "text": "ola", "project": "p1", "model": "model_20200131-133656" }
The intent is returned as null.
amn41
January 31, 2020, 1:07pm
2
hi @vidu171 - what is inside your config.yml ? I would recommend trying to train using the nlu data from moodbot to test that all works.
vidu171
February 3, 2020, 5:22am
3
Hi @amn41 This is my config.yml file
{ "pipeline": "spacy_sklearn", language: "en" }
I am providing path as argument to rasa_nlu.server
python3 -m rasa_nlu.server -c config.yml --path ./venv/models/
vidu171
February 3, 2020, 5:55am
4
Hi, I have realized my mistake, When I was using spacy_sklearn pipeline I didn’t got any error, but using the supervised_embeddings pipeline I realised that I was only sending one intent to train, and Rasa Nlu requires atleast 2 intents for training.