Support for non-BaseCrossValidator CV methods

[nanophyto]

I'm working with a highly zero-inflated dataset. Because of this, I'm using a custom defined zero-stratified CV splitter in my sklearn pipeline.

Currently, MAPIE does not seem to support non-BaseCrossValidator CV methods, which means that if I try to run e.g. MapieRegressor with cv=10 some of the folds are nearly all zeros - resulting in some very unrealistic lower bounds for the PI estimates.

Would it be possible to add support for user-defined CV splitters?