Update app.py

app.py

@@ -98,9 +98,13 @@ def infer_img2img(prompt, audio_path, progress=gr.Progress(track_tqdm=True)):
         print("text prompt:", prompt)
         print("strength_list:", strength_list)
 
-        sf.write(f"{prompt}.wav", concat_audio.flatten(), samplerate=16000)
+        # Ensure correct shape
+        concat_audio = concat_audio.flatten()  # Converts (1, N) → (N,)
 
-        return f"{prompt}.wav"
+        # Save as WAV
+        sf.write("output.wav", concat_audio, 16000)
+
+        return "output.wav"
 
 css="""
 div#col-container{