Leave-One-Out Setup

from sklearn.model_selection import KFold, cross_val_score

# Intended: leave-one-out cross-validation
cv = KFold(n_splits=1)
scores = cross_val_score(model, X, y, cv=cv)
print(scores.mean())