How do you evaluate generated code for questions that accept multiple possible outputs for the same input

[zb1439]

As title, I noticed that testing_utils and the dataset is purely input-output pair based, but it seems like ~10% of questions accept multiple possible outputs for the same input. For example, question 5068 contains the following input output requirement:

-----Input-----

The first line contains two integers n and m — the number of the children and the number of possible mitten colors (1 ≤ n ≤ 5000, 1 ≤ m ≤ 100). The second line contains n integers c_1, c_2, ... c_{n}, where c_{i} is the color of the mittens of the i-th child (1 ≤ c_{i} ≤ m).

-----Output-----

In the first line, print the maximum number of children who can end up with a distinct-colored pair of mittens. In the next n lines print the way the mittens can be distributed in this case. On the i-th of these lines print two space-separated integers: the color of the left and the color of the right mitten the i-th child will get. If there are multiple solutions, you can print any of them.