Save model in python, and use it in spark

as the title show, how should i do?

model = xgb.train()
model.save_model(fname=r’xgb.model’)

model = xgb.Booster(model_file=r’xgb.model’)
print(model.predict(dtest))

this model ‘xgb.model’ created by python ,how to use it in spark?