Closed pfistfl closed 5 years ago
AFAIK the defaults in mlr should be set such that no internal parallelization is used.
but we should ensure, and set it explicitly as well
This is for example not set in (https://github.com/mlr-org/mlr/blob/master/R/RLearner_classif_xgboost.R)
yes, it should :)
so we have to set it from our side anyways
This should be solved now.
@ja-thomas Do we have to limit internal parallelization in xgboost, ranger etc. manually ?