Hello guys I have deployed my rasa x server over gcp and now I want that when post the message I should get intent and entities predicted by rasa x in json response. I am trying to do this by using http://146.148.108.29:80/webhooks/rest/webhook but I m getting only recipient id and predicted text
Actually I want to execute model/parse for getting the intent but in HTTP API page the example for running this is given through localhost:5005/model/parse but in my case the server is hosted at GCP how can I run model/parse command
Hi @sid321axn, if you want to use /model/parse you should redirect the request to the rasa-prod container by adding the /core prefix. So http://<rasa x IP>/core/model/parse?token=<RASA_TOKEN>.
The other option would be to hit the /logs endpoint from Rasa X, which also returns parse data, and will add the entry to your NLU inbox.
Thanks for the response. But could you please explain a bit how to add /logs endpoint. Means I have to add another entry in endpoints.yml ?
It should already be available on http://<rasa x IP>/api/logs. In order to have access to the endpoint, you’ll have to get a bearer auth token with your rasa x credentials from http://<rasa x IP>/api/login as described here: Rasa X Documentation
Here is the full description of the logs endpoint: Rasa X Documentation
