test updates, revert Results to CPU

ultralytics/yolo/engine/results.py

@@ -42,9 +42,9 @@ class Results:
     def __init__(self, orig_img, path, names, boxes=None, masks=None, probs=None) -> None:
         self.orig_img = orig_img
         self.orig_shape = orig_img.shape[:2]
-        self.boxes = Boxes(boxes.cpu(), self.orig_shape) if boxes is not None else None  # native size boxes
-        self.masks = Masks(masks.cpu(), self.orig_shape) if masks is not None else None  # native size or imgsz masks
-        self.probs = probs.cpu() if probs is not None else None
+        self.boxes = Boxes(boxes, self.orig_shape) if boxes is not None else None  # native size boxes
+        self.masks = Masks(masks, self.orig_shape) if masks is not None else None  # native size or imgsz masks
+        self.probs = probs if probs is not None else None
         self.names = names
         self.path = path
         self._keys = (k for k in ('boxes', 'masks', 'probs') if getattr(self, k) is not None)

ultralytics/yolo/utils/plotting.py

@@ -114,7 +114,9 @@ class Annotator:
             self.im = np.asarray(self.im).copy()
         if len(masks) == 0:
             self.im[:] = im_gpu.permute(1, 2, 0).contiguous().cpu().numpy() * 255
-        colors = torch.tensor(colors, device=im_gpu.device, dtype=torch.float32) / 255.0
+        if im_gpu.device != masks.device:
+            im_gpu = im_gpu.to(masks.device)
+        colors = torch.tensor(colors, device=masks.device, dtype=torch.float32) / 255.0
         colors = colors[:, None, None]  # shape(n,1,1,3)
         masks = masks.unsqueeze(3)  # shape(n,h,w,1)
         masks_color = masks * (colors * alpha)  # shape(n,h,w,3)