Important info:
Comparing hellsemble to base models works fine, the autoML route however for me locally required to seperate conda envs, one with auto-sklearn, and one with autogluon because they couldnt co exist together. Autogluon also seems to crash when we run the experiment on too many data sets, so ask works best in this case.
Important info: Comparing hellsemble to base models works fine, the autoML route however for me locally required to seperate conda envs, one with auto-sklearn, and one with autogluon because they couldnt co exist together. Autogluon also seems to crash when we run the experiment on too many data sets, so ask works best in this case.