Closed neham764 closed 5 days ago
pmml = Model.load('dsp_model_pmml.pmml')
Is this a known issue?
Yes, it is a very well known issue that the PyPMML package is making incorrect predictions.
Please switch to the JPMML-Evaluator-Python package, and all predictions will be correct all the time.
Thanks for your prompt reply. Is there a fix expected regarding this soon?
Is there a fix expected regarding this soon?
I'm not affiliated with the PyPMML package in any way.
I'm resposible for the XGBoost-to-PMML converter. And this part of the workflow is working correctly.
Thanks, Jppml evaluator worked.
xgb params:
I am seeing a mismatch between the scores generated through xgboost and pmml :
pmml is trained exactly like xgboost: