Allow for sets of hyperparameters to fail?

mlr-org / mlr3tuning

Hyperparameter optimization package of the mlr3 ecosystem

https://mlr3tuning.mlr-org.com/

GNU Lesser General Public License v3.0

55 stars 5 forks source link

Allow for sets of hyperparameters to fail? #443

Closed SteveBronder closed 3 months ago

SteveBronder commented 3 months ago

When running a model sometimes I'll get errors caused by things like xgboost requesting more memory than available. Is it possible for these instances of tuning parameters to be tossed away / ignored? It would be nice so that then the tuning can still continue even if some models fail

be-marc commented 3 months ago

Yes https://mlr3book.mlr-org.com/chapters/chapter10/advanced_technical_aspects_of_mlr3.html#sec-error-handling