Likes - Activity - ganeshv - Rasa Community Forum

Ganesh Iyer

Enthusiast

Mumbai, India

Dabbler in conversational AI looking to deliver great value through meaningful conversations

What's the best way to extract multiple entities within the same intent and retrieve its numerical value?

Rasa Open Source - Oct 1, 2021

So it looks like you’re successfully extracting the entities, your question seems to be more “how can I extract the same entity and store it to different slots in a single intent”, right?
You could use a list slot for this (there’s some discussion here: How to pupulate a slot of type list - #4 by g…)

Memoization policy ignored in some cases - probably something fundamental I'm not understanding

Rasa Open Source - Apr 20, 2021

It could be your tracker store. Memoization looks back until max history to find an exact match and doesn’t, then it fails. It is possible that your local setup is using an InMemoryTrackerStore that empties every time you restart while in prod; you might be using Redis? Then the RedisTrackerStore is…

Annotating entities with role and groups in Rasa X

[Deprecated] Rasa X Community Edition - Mar 24, 2021

Yes, we are working on supporting entity roles and groups in Rasa X now.

Is anyone else having problems trying to download the poly-ai model?

Rasa Open Source - Sep 28, 2020

Unfortunately, the ConveRT model was taken offline. We are working on a long-term solution and will keep you updated. In the meantime, we recommend to remove ConveRT from your pipeline and just use supervised embeddings, such as CountVectorsFeaturizer.

Is there a way to re-train a model automatically if none of the available models are compatible?

Rasa Open Source - Jul 28, 2020

There are no parameters to control automatic re-train. You can write your own script to do this. You can find the version information in the compressed file of the model. Before upgrading the version, check whether the upgraded version is consistent with the version in the fingerprint!

Does RASA NLU support entity list extraction?

Rasa Open Source - Jan 21, 2020

Hi @ganeshv. You should label each entity in your NLU examples for example: Show me chinese, mexican and japanese restaurants. To enable your assistant to extract these details as a list slot, define this slot in your domain, for example:

slots:
  cuisine:
    type: list

Unable to understand the exception error for custom action

Rasa Open Source - Sep 11, 2019

Guess your problem is the wrong endpoint in that case. Anyway this is how I start my action server:

rasa run actions \
 -vv \  #debug flag, gives lots of output\
 -p $actions_port \  # in your case actions_port=3000\
 --actions actions \
> logs/actions.log 2>&1 &