Update visualize.py

paddlex/cv/models/explanation/visualize.py

@@ -28,6 +28,9 @@ def visualize(img_file,
               num_samples=3000, 
               batch_size=50,
               save_dir='./'):
+    model_info = self.get_model_info()
+    if model_info['status'] != 'Normal':
+        raise Exception('The explanation only can deal with the Normal model')
     model.arrange_transforms(
                 transforms=model.test_transforms, mode='test')
     tmp_transforms = copy.deepcopy(model.test_transforms)
@@ -110,4 +113,4 @@ def precompute_for_normlime(predict_func, normlime_dataset, num_samples=3000, ba
             save_dir=save_dir)
 
 
-    
+