Heroes

A group for Rasa Heroes!

Is the project dead?

Posted by: stephens on Dec 17, 2025
The project commit history shows that status of OS. You should try the developer edition and Rasa CALM.

Getting Error: Type Error: ‘NoneType’ object is not iterable

Posted by: stephens on Mar 25, 2025
It looks like your local instance is running Rasa Pro but your docker-compose is using a Rasa OSS image. Change the image in your docker compose from rasa/rasa to rasa/rasa-pro and don’t forget to set the RASA_PRO_LICENSE env var.

Programmatic Interaction with Chatbot in RASA

Posted by: stephens on Mar 25, 2025
Yes, you can use the REST channel. Also discussed here.

Frontend Widget Not Receiving Bot Responses

Posted by: stephens on Feb 28, 2025
Add --debug and see if confirm that the Rasa sees the message from the user.

Can we use rasa command generation model and other model at the same time?

Posted by: stephens on Feb 27, 2025
You can read about configuring models in the docs here.

Encountered empty domain during validation

Posted by: stephens on Dec 5, 2024
Add --domain domain.yml to the command line.

Problem installing rasa x on ubuntu

Posted by: stephens on Dec 3, 2024
It was announced in June 2022 that Rasa X is no longer maintained or supported.

Rasa-Pro setup AVX Issue on macOS M1

Posted by: stephens on Nov 19, 2024
Sounds like the wrong tensorflow. My install steps on an M1:

uv venv
source .venv/bin/activate
uv pip install rasa-pro==3.10.10
uv pip list | grep tensorflow

Rasa entities extraction which appears to be similar for slot filling via forms. How to extract name of partner and name of customer in a form?

Posted by: stephens on Nov 5, 2024
Are you using a form? Can you give some example utterances that cause issues?

Can Rasa CALM handle image generation if integrated with the right LLM

Posted by: stephens on Oct 17, 2024
You could do this via a custom action.

Recursion in Rasa Flows, Compatible LLM Models, and Fine-Tuning Datasets - Rasa Pro CALM

Posted by: stephens on Sep 25, 2024
You can use CRV’s with other slot types too. Here’s a blog post showing a categorical.

How to add predefined Chitchat and FAQs to CALM

Posted by: stephens on Sep 24, 2024
Another alternative is RAG with extractive search.

Adding LLM to rasa-open source, to rephrase responses

Posted by: stephens on Sep 19, 2024
Yes, with Rasa Pro, you can see an example in the rasa-calm-demo here. This is a flow that calls an API to add a contact to a db.

Sending additional data to RASA server in REST request

Posted by: stephens on Sep 19, 2024
You can supply metadata on the rest call and then retrieve that in action_session_start. There’s an example of this here.

Integrating Rasa with a Custom Frontend Framework - Emmanuel Katto Uganda

Posted by: stephens on Jul 22, 2024
You can read about using the Socket or REST channels with your widget here.

Entity value returned does not match entity value in custom component

Posted by: stephens on Jun 9, 2024
Looks like the DIETClassifier is also extracting these entities. You could either disable entity extraction in the DIETClassifier by setting entity_recognition: False or run a custom action.

RASA Integration with server

Posted by: stephens on Jun 9, 2024
You should review the training materials which do a good job covering the basics.

Need Help Integrating Rasa Chatbot with Website

Posted by: stephens on Jun 9, 2024
Is the socket channel enabled in your credentials.yml? Docs are here. You should see the endpoint listed when you do the rasa run.

How to train Rasa without caching

Posted by: stephens on Jun 9, 2024
The training cache and runtime issue you’re running into each other aren’t related. You should review the debug log for the instances where there is no response.

Safe to delete contents of .rasa/cache/…?

Posted by: stephens on Jun 9, 2024
Yes, it’s safe to delete the cache. It will mean that all components will have to re-train even if no changes were made.

All options Buttons Visibility and activeness

Posted by: stephens on May 9, 2024
I would look at customizing whatever widget you are using to add this functionality.

Training with Enterprise Policy and testing with rasa shell always predicts action_listen

Posted by: vishnupriyavr on May 6, 2024
Here is the config:

recipe: default.v1
language: en
assistant_id: 20230405-114328-tranquil-mustard

pipeline:
  - name: LLMCommandGenerator
    llm:
      model_name: gpt-3.5-turbo:latest
      max_tokens: 20
      type: openai
      openai_api_base: "http://12…"