Closed risufaj closed 1 year ago
Hi @risufaj
Very keen observation. All of the contextual policies have implementations that benefit from multi-core processing. As such we do not also parallelize across arms (for fit) or contexts (for predict) as that could actually slow down performance.
In MAB.py line 876 There is a comment that says to not use parallel fit or predict for contextual policies, but it's not clear why that needs to be the case?
Any help would be appreciated.