Open alhafidzhamdan opened 4 years ago
Hi @alhafidzhamdan,
I'm wondering if you are talking about the modelling step as it's the most time-consuming step?
The GBM
by default uses 15 cores if my memory is correct and can be configured via xgboost's configuration file, while the GLM
can be multi-threaded but not implemented yet.
Hi there, thanks for getting back to me. I was just testing the example data using this command:
driverpower model \ --feature train_feature.hdf5 \ --response train_y.tsv \ --method GBM \ --name tutorial \ --modelDir ./output
As part of the log it says
05/31/2020 16:27:30 | INFO: Note: NumExpr detected 32 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
So I was wondering if there was any way to change the threads?
A
Hi there when I run driverpower model is there a way to allow for increased cores used ? at the moment it seems to just run at 8.