Model Scorning Process - Metadata

JonathanMcLachl
Model Scorning Process - Metadata

Hello, 

Within the Model training process, we have the option of adding additional Metadata labels to add to a version of a saved model on training. When the model is being scored, there is an option to include the model metadata in the model output. The types of Metadata included in the output are the ID, Version, Fullid and Prediction time. I would like to know if I can output the created Metadata fields from the model training process into the model scorning process.   

 

Thank you, 

0 Kudos
1 Reply
AurelienL
Dataiker

Hello,

It is not possible to add other model metadata besides id, version, fullId and prediction time using the scoring recipe.

If you want maximum flexibility, you can use the Python recipe.

 

0 Kudos

Labels

?
Labels (1)
A banner prompting to get Dataiku