Best practice to deploy trained model to Knime server

Hi everyone,

what is the best practice to deploy locally trained models to the Knime server environment? Currently we are copying them over to the server using SSH and save them to /home/knime/knime-workspace.


Hi and welcome to the KNIME Forum!
There are different ways to deploy models to KNIME Server, but you do not have to copy them via SSH. In the Analytics Platform, you can mount the KNIME Server and then deploy workflows and data by right-clicking and selecting “Deploy to server”. Please have a look here for more info on this: I usually create a workflow group locally with another nested workflow group called “models”. Then I use a workflow in the top workflow group and save my model using a workflow-relative path: knime://knime.workflow/…/models/mymodel.pmml. After running the workflow, the model is in the model folder and I can now simply deploy it to the server using the aforementioned method.
Kind regards,

1 Like

Hi, thanks for the quick response.

We did not know that we can simply save files within a workflow group and deploy the whole group. Now it makes sense :slight_smile:



This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.