Memory leak ?

[CoteDave]

Running a simple example with CatBoostClassifier and a relativly small dataset (25000, 160),memory consomption just exploded(with python = 3.8):

the first step got executed correctly:
acvtree = ACVTree(model, X_test_enc_top_extreme.fillna(0).values)

then this step just kill my machine with the small dataset (25000, 160),more than 100gig of RAM was used before it crashed...:
shap_val_acv = acvtree.shap_values(X_test_enc_top_extreme.fillna(0).values)