Closed pollytur closed 4 months ago
the submission file has a change because of this (it iterates across tiers, so the ood/ bonus naming should be consistent)
Maybe I miss it, for the benchmark training notebook, maybe also include ensemble models?
@yongrong-qiu But for training these are exactly same as the factorized ones and then the predictions are just averaged
the submission file has a change because of this (it iterates across tiers, so the ood/ bonus naming should be consistent)