my code currently run and pass all the public test case but it fail for the hidden one

which said Make sure that 'evaluate' is called with the correct input, especially for white I dont really understand why, as i already inverted after processing white turn.

def minimax$($

board: Board,

depth: int,

max$\_$depth: int,

is$\_$black: bool

$)->$ tuple$[$Score$,$ Move$]$:

$"""$

Finds the best move for the input board state.

Note that you are black.

Parameters

$----------$

board: $2$D list of lists. Contains characters $"$B$","$W$",$ and $"\_",$

representing black pawn, white pawn, and empty cell, respectively.

depth: int, the depth to search for the best move. When this is equal

to $`$max$\_$depth$`,$ you should get the evaluation of the position using

the provided heuristic function.

max$\_$depth: int, the maximum depth for cutoff.

is$\_$black: bool. True when finding the best move for black, False

otherwise.

Returns

$-------$

A tuple $($evalutation$,(($src$\_$row, src$\_$col$),($dst$\_$row, dst$\_$col$)))$:

evaluation: the best score that black can achieve after this move.

src$\_$row, src$\_$col: position of the pawn to move.

dst$\_$row, dst$\_$col: position to move the pawn to$.$

$"""$

if depth $==$ max$\_$depth or utils.is$\_$game$\_$over$($board$)$:

return evaluate$($board$),$ None

# Determine the best move and its evaluation

if is$\_$black:

best$\_$evaluation $=$ float$(\text{'}-$inf'$)$

best$\_$move $=$ None

for action in generate$\_$valid$\_$moves$($board$)$:

new$\_$board $=$ utils.state$\_$change$($board$,$ action$[0],$ action$[1],$ in$\_$place$=$False$)$

opponent$\_$evaluation, $\_=$ minimax$($new$\_$board, depth $+1,$ max$\_$depth, False$)$

if opponent$\_$evaluation $>$ best$\_$evaluation:

best$\_$evaluation $=$ opponent$\_$evaluation

best$\_$move $=($action$[0],$ action$[1])$

return best$\_$evaluation, best$\_$move

else:

best$\_$evaluation $=$ float$(\text{'}$inf$\text{'})$

best$\_$move $=$ None

for action in generate$\_$valid$\_$moves$($utils$.$invert$\_$board$($board$,$ in$\_$place$=$False$))$:

new$\_$board $=$ utils.state$\_$change$($utils$.$invert$\_$board$($board$,$ in$\_$place$=$False$),$ action$[0],$ action$[1],$ in$\_$place$=$False$)$

opponent$\_$evaluation, $\_=$ minimax$($utils$.$invert$\_$board$($new$\_$board, in$\_$place$=$False$),$ depth $+1,$ max$\_$depth, True$)$

if opponent$\_$evaluation $<$ best$\_$evaluation:

best$\_$evaluation $=$ opponent$\_$evaluation

best$\_$move $=($action$[0],$ action$[1])$

return best$\_$evaluation, best$\_$move