Minor fix: mAP can be 1.01

Fix division to avoid cases where 1.01 mAP is possible.

Discovered in: https://colab.research.google.com/drive/10EaQ4lJNXzcmub7doO2vGn2t1LgDdkcC?usp=sharing

supervision/metrics/mean_average_precision.py

@@ -230,7 +230,7 @@ def _compute_average_precision(recall: np.ndarray, precision: np.ndarray) -> flo
         for r, p in zip(recall[::-1], precision[::-1]):
             precision_levels[recall_levels <= r] = p
 
-        average_precision = (1 / 100 * precision_levels).sum()
+        average_precision = (1 / 101 * precision_levels).sum()
         return average_precision
 
     @staticmethod