specified tokens

app.py

@@ -77,7 +77,7 @@ image = gr.inputs.Image()
 gr.Interface(
     description="Object Recognition with DesCo",
     fn=predict,
-    inputs=["image", "text", "ground tokens (optional)"],
+    inputs=["image", "text", "text"],
     outputs=[
         gr.outputs.Image(
             type="pil",