How do I load my Rasa NLU model in Rasa Server using python script?

I have trained a Rasa NLU model for intent classification and entity extraction, how do I load this model in a Rasa Server using a python script.

For example, I use: rasa run --enable-api -m rasanlu.tar.gz in my terminal and the rasa server is up and running.

How do I achieve this using a python script?

Also is there a way to load the model from s3 bucket and run the rasa server directly through a python script?

post by NIlimaN on Jun 21, 2022

I have trained a Rasa NLU model for intent classification and entity extraction, how do I load this model in a Rasa Server using a python script.

For example, I use: rasa run --enable-api -m rasanlu.tar.gz in my terminal and the rasa server is up and running.

How do I achieve this using a python script?

Also is there a way to load the model from s3 bucket and run the rasa server directly through a python script?

post by siriusraja on Jun 22, 2022

Hi @NIlimaN can you share me the objective of doing this via python script ?

post by NIlimaN on Jun 23, 2022

Hi @siriusraja

Thanks for responding.

To bootstrap this into a process for ease of use in deployment.

I was able to start the Rasa server using a python script.

But I do want to know if I can directly load the model from s3 bucket and start the rasa server using a python script instead of having to download it beforehand?

post by siriusraja on Jun 23, 2022

Hi @NIlimaN

2 things here.

To bootstrap, you can also consider creating a service on the systemd which you can start / stop / restart / schedule easily.

To load the model from AWS S3, you can follow the steps as per the docs Model Storage.