Update app.py

app.py

@@ -179,6 +179,10 @@ def generate_audio_from_image(image):
     # Convert to the required format (e.g., float32)
     generated_audio = generated_audio.astype(np.float32)
 
+    # Ensure the output is stereo with 2 channels, even if audio is mono
+    if len(generated_audio.shape) == 1:
+        generated_audio = np.stack([generated_audio, generated_audio], axis=0)  # Duplicate for stereo
+
     # Debug: Print the shape and type of the generated audio
     print(f"Generated audio shape: {generated_audio.shape}, type: {generated_audio.dtype}")