Closed creative-common closed 5 years ago
well, it is better to enable custom pipelines support. not sure whether latest version rasa_ui support compact json for NLU Pipeline Template field like [{"name":"nlp_spacy"},{"name":"ner_crf"},{"name":"ner_synonyms"}]
looks the custom pipelines feature is not supported in Rasa UI, so i made some modifications
in the official Rasa Doc about Training Http API, it said that "The request should always be sent as application/x-yml regardless of wether you use json or md for the data format. Do not send json as application/json for example." However, i tried the application/json with json format like below, and have a successful return.
{
"language":"zh",
"pipeline":[{"name":"tokenizer_jieba"},{"name":"ner_crf"},{"name":"ner_synonyms"},{"name":"intent_featurizer_count_vectors"},{"name":"intent_classifier_tensorflow_embedding"}],
"data":{
"rasa_nlu_data":{
"common_examples":[...]
}
}
}
the NLU version that i used:
{
"version": "0.14.3",
"minimum_compatible_version": "0.13.0a2"
}
i am not a JS guy, but i can create a pull request for reviewing if we need to enable this feature. :)
I think this is one of the essential feature, @pradeepmvn what you think. @JoshuaZe You should create a pull request if its working I hope pradeep will surely add it.
Thanks!
@JoshuaZe please feel free to submit a PR. This will also need a ui change to display it as a text area. Thanks.
just submit a PR. i am not familiar with UI, and it looks like already a text area ?
@JoshuaZe no problem I will take care of it.
@sanjeethanspal Thank you!
Hello,
@pradeepmvn
How can we use the duckling inside the rasa_ui along with other pipeline template such as spacy_sklearn or tensorflow_embedding ?
Thanks