Closed sakura7621 closed 5 years ago
It would be even better if I can have an example to load the
light gbm pmml
file, ...
After conversion, there is no such thing as "LightGBM PMML file" or "XGBoost PMML file" - it's one standardized PMML representation for everything.
It follows that all PMML files are used for scoring exactly the same way; the PMML scoring engine doesn't need to know/care what was the original ML framework, because in PMML representation all ML models look exactly the same.
.. and use the model (in
java
version) to predict for a input data file.
See the documentation of your PMML scoring engine.
For example, if you're using the JPMML-Evaluator library, then everything should be nicely explained in its README file.
Short version: https://github.com/jpmml/jpmml-evaluator#basic-usage
Long version: https://github.com/jpmml/jpmml-evaluator#advanced-usage
Hi @vruusmann,
The first thing I want to say is thank you for this project. I found it really helpful for me when I apply lightgbm for a
java
application.It would be even better if I can have an example to load the
light gbm pmml
file, and use the model (injava
version) to predict for a input data file.Can you please give me some more document about using the output (
light gbm pmml
file)?Any help would be appreciated. Thanks.