ultralytics 8.0.158 add benchmarks to coverage (#4432)

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Yonghye Kwon <developer.0hye@gmail.com>

ultralytics/models/sam/predict.py

@@ -55,12 +55,14 @@ class Predictor(BasePredictor):
         return img
 
     def pre_transform(self, im):
-        """Pre-transform input image before inference.
+        """
+        Pre-transform input image before inference.
 
         Args:
             im (List(np.ndarray)): (N, 3, h, w) for tensor, [(h, w, 3) x N] for list.
 
-        Return: A list of transformed imgs.
+        Returns:
+            (list): A list of transformed images.
         """
         assert len(im) == 1, 'SAM model has not supported batch inference yet!'
         return [LetterBox(self.args.imgsz, auto=False, center=False)(image=x) for x in im]