model2.predict(X_test), get IndexError by use custom_loss!

[andrey-khropov]

First reported in #10

model2.predict(X_test)

get IndexError on Python 3.8.8 64-bit, and Python 3.9.7, like this:

Traceback (most recent call last):

File "E:\py\catboost\tutorials\custom_loss\custom_loss_and_metric_tutorial.py", line 187, in
model2.predict(X_test,thread_count=1,prediction_type='Class')

File "C:\ProgramData\Anaconda3\lib\site-packages\catboost\core.py", line 4772, in predict
return self._predict(data, prediction_type, ntree_start, ntree_end, thread_count, verbose, 'predict', task_type)

File "C:\ProgramData\Anaconda3\lib\site-packages\catboost\core.py", line 2220, in _predict
predictions = self._base_predict(data, prediction_type, ntree_start, ntree_end, thread_count, verbose, task_type)

File "C:\ProgramData\Anaconda3\lib\site-packages\catboost\core.py", line 1516, in _base_predict
return self._object._base_predict(pool, prediction_type, ntree_start, ntree_end, thread_count, verbose, task_type)

File "_catboost.pyx", line 4529, in _catboost._CatBoost._base_predict

File "_catboost.pyx", line 4546, in _catboost._CatBoost._base_predict

File "_catboost.pyx", line 1609, in _catboost.transform_predictions

File "_catboost.pyx", line 5350, in _catboost._convert_to_visible_labels

IndexError: index 0 is out of bounds for axis 0 with size 0