Retrieve Model
(MLFlow)
Synopsis
Retrieve a model from MLFlowDescription
By default this operator returns the model object, which is associated to a model version on the mod port. In this mode of operation this model works very similar to the "Retrieve" operator. It only pulls the model off the mlflow backend and not from the repository.Description
This operator can also download all artifacts which are associated to a given deployed model. Log Run has the option to also store all processes and objects which are in a defined folder. This can be necessary either if your model requires a specific scoring process or additional artifacts like preprocessing models.Description
If "download_artifacts" is set to true all associated artifacts (processes, models, tables...) are downloaded into the folder specified as 'folder'.Description
This operator has no tutorial processes, but this extension ships with a tutorial project using titanic. Just right click on a folder in your repository and use 'Add MLFlow Examples' to see it.Input
Output
Parameters
- model name Name of the mlflow model you want to get the rapidminer model from
- stage What stage to retrieve the model from
- download artifacts If set to true the operator will also download all artifacts from the mlflow storage
- folder The folder to store the artifacts in. Only available if 'download_artifacts' is set to true.