Fix torch device issues (#20304)

* fix device issue
Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>

src/transformers/models/conditional_detr/feature_extraction_conditional_detr.py

@@ -881,7 +881,7 @@ class ConditionalDetrFeatureExtractor(FeatureExtractionMixin, ImageFeatureExtrac
             img_w = torch.Tensor([i[1] for i in target_sizes])
         else:
             img_h, img_w = target_sizes.unbind(1)
-        scale_fct = torch.stack([img_w, img_h, img_w, img_h], dim=1)
+        scale_fct = torch.stack([img_w, img_h, img_w, img_h], dim=1).to(boxes.device)
         boxes = boxes * scale_fct[:, None, :]
 
         results = []

src/transformers/models/deformable_detr/feature_extraction_deformable_detr.py

@@ -729,7 +729,7 @@ class DeformableDetrFeatureExtractor(FeatureExtractionMixin, ImageFeatureExtract
             img_w = torch.Tensor([i[1] for i in target_sizes])
         else:
             img_h, img_w = target_sizes.unbind(1)
-        scale_fct = torch.stack([img_w, img_h, img_w, img_h], dim=1)
+        scale_fct = torch.stack([img_w, img_h, img_w, img_h], dim=1).to(boxes.device)
         boxes = boxes * scale_fct[:, None, :]
 
         results = []

src/transformers/models/detr/feature_extraction_detr.py

@@ -1103,7 +1103,7 @@ class DetrFeatureExtractor(FeatureExtractionMixin, ImageFeatureExtractionMixin):
             else:
                 img_h, img_w = target_sizes.unbind(1)
 
-            scale_fct = torch.stack([img_w, img_h, img_w, img_h], dim=1)
+            scale_fct = torch.stack([img_w, img_h, img_w, img_h], dim=1).to(boxes.device)
             boxes = boxes * scale_fct[:, None, :]
 
         results = []

src/transformers/models/yolos/feature_extraction_yolos.py

@@ -694,7 +694,7 @@ class YolosFeatureExtractor(FeatureExtractionMixin, ImageFeatureExtractionMixin)
             else:
                 img_h, img_w = target_sizes.unbind(1)
 
-            scale_fct = torch.stack([img_w, img_h, img_w, img_h], dim=1)
+            scale_fct = torch.stack([img_w, img_h, img_w, img_h], dim=1).to(boxes.device)
             boxes = boxes * scale_fct[:, None, :]
 
         results = []

tests/models/conditional_detr/test_modeling_conditional_detr.py

@@ -511,9 +511,9 @@ class ConditionalDetrModelIntegrationTests(unittest.TestCase):
         results = feature_extractor.post_process_object_detection(
             outputs, threshold=0.3, target_sizes=[image.size[::-1]]
         )[0]
-        expected_scores = torch.tensor([0.8330, 0.8313, 0.8039, 0.6829, 0.5355])
+        expected_scores = torch.tensor([0.8330, 0.8313, 0.8039, 0.6829, 0.5355]).to(torch_device)
         expected_labels = [75, 17, 17, 75, 63]
-        expected_slice_boxes = torch.tensor([38.3089, 72.1022, 177.6293, 118.4512])
+        expected_slice_boxes = torch.tensor([38.3089, 72.1022, 177.6293, 118.4512]).to(torch_device)
 
         self.assertEqual(len(results["scores"]), 5)
         self.assertTrue(torch.allclose(results["scores"], expected_scores, atol=1e-4))

tests/models/deformable_detr/test_modeling_deformable_detr.py

@@ -569,9 +569,9 @@ class DeformableDetrModelIntegrationTests(unittest.TestCase):
         results = feature_extractor.post_process_object_detection(
             outputs, threshold=0.3, target_sizes=[image.size[::-1]]
         )[0]
-        expected_scores = torch.tensor([0.7999, 0.7894, 0.6331, 0.4720, 0.4382])
+        expected_scores = torch.tensor([0.7999, 0.7894, 0.6331, 0.4720, 0.4382]).to(torch_device)
         expected_labels = [17, 17, 75, 75, 63]
-        expected_slice_boxes = torch.tensor([16.5028, 52.8390, 318.2544, 470.7841])
+        expected_slice_boxes = torch.tensor([16.5028, 52.8390, 318.2544, 470.7841]).to(torch_device)
 
         self.assertEqual(len(results["scores"]), 5)
         self.assertTrue(torch.allclose(results["scores"], expected_scores, atol=1e-4))