Update processing_videollama3.py

processing_videollama3.py

@@ -292,25 +292,30 @@ class Videollama3Qwen2Processor(ProcessorMixin):
         num_tokens = int(grid_size.prod().item())
         return num_tokens
 
+    
     def load_images(self, image_path: Union[str, List[str], Image.Image, List[Image.Image]]):
-        if isinstance(image_path, str) and os.path.isfile(image_path):
-            # images = [cv2.cvtColor(cv2.imread(image_path), cv2.COLOR_BGR2RGB)]
-            images = [Image.open(image_path).convert('RGB')]
-        elif isinstance(image_path, str) and os.path.isdir(image_path):
-            # images = [cv2.cvtColor(cv2.imread(os.path.join(image_path, f)), cv2.COLOR_BGR2RGB) for f in sorted(os.listdir(image_path))]
-            images = [Image.open(os.path.join(image_path, f)).convert('RGB') for f in sorted(os.listdir(image_path))]
-        elif isinstance(image_path, str) and image_path.startswith("http://") or image_path.startswith("https://"):
-            images = [Image.open(requests.get(image, stream=True).raw)]
-        elif isinstance(image_path, list) and isinstance(image_path[0], str):
-            # images = [cv2.cvtColor(cv2.imread(f), cv2.COLOR_BGR2RGB) for f in image_path]
-            images = [Image.open(f).convert('RGB') for f in image_path]
-        elif isinstance(image_path, list) and isinstance(image_path[0], Image.Image):
-            images = [np.array(x) for x in image_path]
-        elif isinstance(image_path, Image.Image):
-            images = [np.array(image_path)]
-        else:
-            raise ValueError(f"Unsupported image path type: {type(image_path)}")
-        return images
+        def load_single_image(image_path):
+            if isinstance(image_path, str) and os.path.isfile(image_path):
+                # images = [cv2.cvtColor(cv2.imread(image_path), cv2.COLOR_BGR2RGB)]
+                images = Image.open(image_path).convert('RGB')
+            elif isinstance(image_path, str) and image_path.startswith("http://") or image_path.startswith("https://"):
+                images = Image.open(requests.get(image_path, stream=True).raw)
+            elif isinstance(image_path, Image.Image):
+                images = np.array(image_path)
+            else:
+                raise ValueError(f"Unsupported image path type: {type(image_path)}")
+            return images
+
+        try:
+            if isinstance(image_path, list):
+                images = [load_single_image(f) for f in image_path]
+            elif isinstance(image_path, str) and os.path.isdir(image_path):
+                images = [Image.open(os.path.join(image_path, f)).convert('RGB') for f in sorted(os.listdir(image_path))]
+            else:
+                images = [load_single_image(image_path)]
+            return images
+        except:
+            raise ValueError(f"Error when loading images: {type(image_path)}")
 
     def load_video(
         self,