Fix local n calculation

ray-project / xgboost_ray

Distributed XGBoost on Ray

Apache License 2.0

143 stars 34 forks source link

Fix local n calculation #78

Closed krfricke closed 3 years ago

krfricke commented 3 years ago

Currently the actor local n is overridden when loading any data. This means that usually the length of the last evaluation dataset is returned instead of the last training dataset.