Error loading XGBoost4J trained model object into Python

I already have a xgboost model object that is trained on XGBoost4j 0.90. The way to trained the model and save it as below (pardon me for hiding some sensitive information)

The env was based on Scala, XGBoost4j 0.90.jar is used.

Now the system has been migrated and we do not have Scala supported, but Python and Pyspark only. I follow link to load this model with Python XGboost., but got below error:

The new env is image

Can anyone helps on this error ? Thanks so much!