different output shape and squash crop mode

app.py

@@ -35,6 +35,7 @@ config = resolve_data_config({}, model=vit)
 config['no_aug'] = True
 config['interpolation'] = 'bilinear'
 config['crop_pct'] = 1
+config['crop_mode'] = 'squash'
 transform = create_transform(**config)
 
 # Inference function
@@ -92,8 +93,8 @@ demo = gr.Interface(
     query_image, 
     #inputs=[gr.Image(), "text", "checkbox", gr.Slider(0, 1, value=0.25)],
     inputs=[gr.Image(type='numpy', label='input_img').style(height=340, width=600), "text", "checkbox", gr.Slider(0, 1, value=0.25)],  
-    outputs="image",
-    #outputs=gr.Image(type='numpy', label='output').style(height=600, width=600),
+    #outputs="image",
+    outputs=gr.Image(type='numpy', label='output').style(height=600, width=600),
     title="Object Detection Using Textual Queries",
     description=description,
     examples=[