Loading an existing python trained model trained in Spark

Apologies for opening this question when there is an identical question opened two years ago:

Even though the above question received many views, its resolution seems lacking.

It is suggested to use a bridge class but the way to use the bridge class is unclear.

In the two years that past has there been any progress? The latest version is 0.9 but I still don’t see an easy way to load the model.

I tried using version 0.72 but it didn’t work as well: XGBoost doesn’t have a loadmodel function. It does have a loadModelFromHadoop function but that failed:

val model = XGBoost.loadModelFromHadoopFile("/home/hadoop/model.mod")(sc)
java.io.FileNotFoundException: File does not exist: /home/hadoop/model.mod

I’m really stuck here, any help would be most appreciated!
Thank you!