Kravchenko Sonya

Could you add an example of importing custom python modules in pipeline_wrapper.py? Thanks!

Hi! I'm also interested in the possibility of incremental training using text features. Any updates here? Would love to have a way to train models with text data on limited...

Hi, i have the same issue Any kind of updates or solutions? upd: seems like `CatBoostClassifier(**params).fit(batch)` doesn't use `params['classes_count']` if `batch` got baseline settled with `batch.set_baseline()` desired behaviour: warning message...

it works fine if every batch contains every class from the overall classes set (apply something like batch normalization in deep learning, i guess it's nice for the model quality...