All - Activity - flythe - Rasa Community Forum

Rasa X: Custom Pipline Element

Hi @Taufred, you should probably check your docker-compose output. Are the containers actually running? In my solution I only need to create a custom image for the app service so it seems less complex than your application.

Adding external libraries

Docker should take care of this if you set up the Dockerfile correctly. If you are using a requirements.txt file check out this medium post. You can also define all the packages directly in the Docker file, e.g.:

RUN pip install python-dateutil

Docker won’t reinstall packages that are already at …

RASA docker image existed with 132

The only solution is to rebuild the rasa container with a custom build of tensorflow that doesn’t use AVX. I would only recommend doing this if you have absolutely no other solution (i.e. moving to a different platform). Just the process of building tensorflow can take up to a day if you know what y…

Error 132 probably means that the CPU doesn’t support AVX which is required by tensorflow. I ran into this problem a while ago, see here. There is no practical fix for this except for moving to a different hardware platform (that does have AVX support).

Integrate Rasa in website

You can add a chatbot to any website by using the rasa webchat. Add the script snippet to the pages that you want to add the chatbot to.

Running Multiple Instances of Rasa Bots on one server

In theory this is possible, docker containers will run independent of each other. You’ll have to do some port management because all the instances will try to hook into the same ports on the host machine. This also means that requests have to be proxied on the host machine, you can use nginx for that.

Rasa X: Custom Pipline Element

You are right I copied the wrong override example. This is the correct setup:

services:
  rasa-worker:
    volumes:
      - ./directory/my_custom_component.py:/app/my_custom_component.py

Preferably you don’t edit the docker-compose.yml file directly because it will get overwritten when it is changed.

Can't connect to bitbucket

Bitbucket doesn’t allow to set the write permissions on a per repo basis like GitHub does. You have to give write permissions to ALL the repositories under your account. If you want to go that route go to your account ssh keys and add the key that Rasa X gives you.

'rasa run actions' is not working

The action server is up and running, note that the process won’t go to the background. Open another CLI start up a rasa shell and you can use your custom actions.

Looks like you are running two action servers, the 5055 port is already in use.

Rasa X: Custom Pipline Element

I’m not in the Rasa X Team but I went through the same process as you are now ;). The Rasa blog on custom components describes the steps to take to add a component:

add it to the config.yml

add the component directory to the PYTHONPATH

You need to convert those steps into a Dockerfile to build a custom image including the component.

If a slot is not requested, slot mapping doesn't work

To set a slot that has not been requested you could try to hook into the request_next_slot:

def request_next_slot(self, dispatcher, tracker, domain):
    next_slot = super().request_next_slot(dispatcher, tracker, domain)

# All required slots have been filled
    if not next_slot:
        …

The custom slot mappings only work for requested slots, see the example in the docs. If you want to see how this is implemented you can also check the get_mappings_for_slot function.

How to revert a form with all slots?

I use a function for this purpose:

def kill_form(obj, next_intent):
    """
    Force the end of a form and move on to the intent `next_intent`
    """

return obj.deactivate() + [ActionExecuted('action_listen')] + [UserUttered('/' + next_intent, {"intent": {'name': next_intent, 'confidence': 1.0}})]

Multiple Entity Detection Problem

The entity extractor simply combines words when they have the same entity, even when they are defined and trained separately. Imagine we have the sentence: I can speak English, German and Spanish. This will be tagged as: [‘O’, ‘O’, ‘O’, ‘lang’, ‘lang’, ‘O’, ‘lang’], the extractor loops over these tokens.

I’m having the exact same problem, did you find an answer to this? In my pipeline I use the WhitespaceTokenizer but the sentence: “I can speak English, German and Spanish”. Will be labeled as: I can speak english, german and spanish. It seems the tokenizer ignores the space after the comma.

Clean rasa-x install containers don't start

I am bound to the hardware that the VPS hoster provides, they do not expose AVX at the moment. I’m building TensorFlow from source, but how do I tell the Docker container to use my local build?

Through sheer luck I found out that this has to do with TensorFlow. The scripts in the container try to import the TensorFlow module and that dumps the 132 exit code when the server CPU does not support AVX. Is there any way to define the TensorFlow version that should be used?

Rasa version: 1.8.0
Rasa X version: 0.26.0
Python version: 3.6.9
Operating system: Ubuntu 18.04
Issue:
I am installing Rasa X on a clean Ubuntu install on a new VPS using the docker one line method. When I run docker-compose up -d I can access the Rasa X UI through /login. However, the rasa…