Atanas Dimitrov
Atanas Dimitrov
I have spent some time trying to implement different ideas for propagating the shape info, however most of them are either inconsistent or too slow. Currently we have the problem...
> Thanks for writing this out. I definitely agree that predictable shape inference performance is too important to give up without a really compelling use case. Did you already do...
Thank you for the fast response! Maybe I am missing something but: > 1. I don't understand what you mean when you say that set_leaf_output() does not update "the predictions...
@lorentzenchr Maybe it is my bad for not explaining in too much detail. This change enables both honest splits and debiasing(i.e. in different implementation) to be implemented as callbacks to...
Since all written above is a bit abstract let me illustrate by example the utility honest splits can bring to a trained model. This example uses the [French Motor Claims...
I am open to discussions about surrounding details as this is a public-facing API change.
Currently the CI is not passing as #6574 is blocking.
> I am open to ideas of ways to test related functionalities. Tests should now be sufficient for the change.
@jameslamb Could you take a look at the PR, now that the CI is passing?
Thank you for taking the time to look into the PR and linking a relevant issue. > I'm not sure if this will correctly handle these cases: > > *...