fingoldo
fingoldo
@DmitryUlyanov could you kindly take a look, there is still a problem on Windows... UPD After server reboot somehow it started working.. )
Joining this request, please add it if you can (for python 3.7 as well)...
Any update on this? stanza still seems to be missing xpos for Russian.
Bump, please. Any reason that prevents adding this?
It is really a big issue. It prevents the proper use of catboost as a part of a pipeline, especially when a validation set is required because a pipeline can...
Dear CatBoost developers, any response on adding validation_fraction functionality? It would make lives of many applied data scientists a lot easier...
btw the most general solution would be to make evaluation set stratified/unstratified, shuffled/unshuffled, ie, exactly replicate train_test_split capabilities.
Note that, while it's being implemented natively, you can use a subclassing workaround: ```python from sklearn.utils.validation import check_X_y, check_array, check_is_fitted, check_consistent_length from sklearn.base import RegressorMixin, ClassifierMixin, TransformerMixin, BaseEstimator, clone from...
it's a plotly object, it seems. could we have some access to it, or maybe there is a way to access it from a callback?
Could be just a damaged file. Try to skip that particular file and see if it helps...