Update app.py

app.py

@@ -11,7 +11,7 @@ audio_gen = gr.Blocks.load(name="spaces/fffiloni/audioldm-text-to-audio-generati
 ph_message="If you're not happy with sound result, you can manually describe the scene depicted in your image :)"
 
 def clean():
-    return manual_cap.update(placeholder=ph_message)
+    return manual_cap.update(value=None,placeholder=ph_message)
 
 def infer(image_input, manual_caption, duration_in, seed):
     print(duration_in)