update type and docstring of spro_score function

Signed-off-by: Willy Fitra Hendria <[email protected]>

src/anomalib/metrics/spro.py

@@ -98,15 +98,16 @@ def compute(self) -> torch.Tensor:
 
 def spro_score(
     predictions: torch.Tensor,
-    targets: torch.Tensor,
+    targets: list[torch.Tensor],
     threshold: float = 0.5,
     saturation_config: dict | None = None,
 ) -> torch.Tensor:
     """Calculate the SPRO score for a batch of predictions.
 
     Args:
         predictions (torch.Tensor): Predicted anomaly masks.
-        targets: (torch.Tensor): Ground truth anomaly masks with non-binary values and, original height and width
+        targets: (list[torch.Tensor]): Ground truth anomaly masks with original height and width. Each element in the
+            list is a tensor list of masks for the corresponding image.
         threshold (float): When predictions are passed as float, the threshold is used to binarize the predictions.
         saturation_config (dict): Saturations configuration for each label (pixel value) as the keys.
             Defaults: ``None`` (which the score is equivalent to PRO metric, but with the 'region' are