Automatic loading of models from server?

I want to automatically pull the latest version of my model in a folder called /rasa/models

I’ve set the endpoints.yml file to:

  wait_time_between_pulls: 120

but it never seems to fetch the new model.

Any ideas why?

@bferster Strange, are you sure, the url path you are giving is same where your trained model is saved? or even the https or http one?

I’m pointing to the folder. The full path is: It’s in a protected folder, but Rasa is on the same server, so it should have access to it

@bferster can you please mention the full path as you shown above, so that we can check that its actually working or fetching the model path or not? Hope this is no indentation issue in the file? please check and also share the rasa --version for me.

are you suggesting I change the endpoints pile to to reference the whole path? I’ve included the file i’m using to set the path

endpoints.yml (78 Bytes)

Hi @bferster your issue got solved?

not yet…

Hi @bferster

Can you please share the command which you are using to start the rasa server ?

Also please do verify if your http server ( the one serving the model ) is configured as described in this section Model Storage

Hi,I use myself server,and use rasa model train and model using in one method.
if u want to use rasa commercialy,I suggest that u just do this

Very helpful. Thanks!