👋 Introducing the Rasa Playground

Model_url question in config.yml file

post by ryohei6 on May 30, 2021

Hello.
This is Ryohei.
I’m new to RASA and I’m trying to make chatbot with this framework.
When I tried “rasa train”, I got following error.

RasaException: Parameter “model_url” was not specified in the configuration  
of “ConveRTFeaturizer”. It is mandatory to pass a value for this parameter.  
You can either use a community hosted URL of the model  
or if you have a local copy of the model, pass the  
path to the directory containing the model files.  

And my config.yml is like following.

language: en

pipeline:
- name: HFTransformersNLP
  model_weights: “bert-base-uncased”
  model_name: “bert”
- name: ConveRTTokenizer
- name: ConveRTFeaturizer
- name: LanguageModelTokenizer
- name: LanguageModelFeaturizer
- name: WhitespaceTokenizer
- name: CountVectorsFeaturizer
- name: CountVectorsFeaturizer
  analyzer: char_wb
  min_ngram: 1
  max_ngram: 4
- name: DIETClassifier
  epochs: 200

policies:

I tried “pip install rasa[full]” and “pip install rasa[convert]” but it didn’t work.
Would anyone teach any solutions to this please?

best regards
Ryohei

post by shivam.17 on Jun 3, 2021

As per Components,
You need to add a hosted model URL.
You can add the following line below the declaration of ConveRTTokenizer:

model_url: "https://github.com/connorbrinton/polyai-models/releases/download/v1.0/model.tar.gz"

post by ryohei6 on Jun 14, 2021

Thank you very much.
Your comment helped so much!

post by tim.blackbird on Aug 3, 2021

It goes under ConveRTFeaturizer for anyone needing this.

pipeline:
  - name: ConveRTTokenizer
  - name: ConveRTFeaturizer
    model_url: "https://github.com/connorbrinton/polyai-models/releases/download/v1.0/model.tar.gz"
  - name: DIETClassifier
    epochs: 200