Error on Spark MLFlow Model Registery using DataBricks after upgrade: cannot load trained XGBoost model

Viewed 111

I had some code for training and then using XGBoost models on a Databricks environment. As my runtime version got deprecated, I upgraded it, but I quickly noticed I could not load my trained models anymore. The reason seems to be a change in the naming of functions in Sparkdl:

Error loading metadata: Expected class name sparkdl.xgboost.xgboost_core.XgboostClassifierModel but found class name sparkdl.xgboost.xgboost.XgboostClassifierModel

Would anyone have advise on how to fix this issue? Maybe modify the metadata?

0 Answers
Related