Update app.py

app.py

@@ -37,8 +37,8 @@ with gr.Blocks(theme=gr.themes.Ocean(primary_hue="pink", neutral_hue="indigo", f
       raw_image = Image.open(requests.get(url, stream=True).raw).convert('RGB')
       inputs = processor(raw_image, return_tensors="pt").to("cpu")
       out = model.generate(**inputs)
-      output = processor.decode(out[0], skip_special_tokens=True)
-      return leer(processor.decode(out[0], skip_special_tokens=True))
+      leer(processor.decode(out[0], skip_special_tokens=True))
+      return processor.decode(out[0], skip_special_tokens=True)
 
     def leer(texto):
         pipe = StableAudioPipeline.from_pretrained("stabilityai/stable-audio-open-1.0")
@@ -60,9 +60,9 @@ with gr.Blocks(theme=gr.themes.Ocean(primary_hue="pink", neutral_hue="indigo", f
         ).audios
         
         #sf.write("demo.wav", salida, pipe.vae.sampling_rate)
-        return audio[0].T.float().cpu().numpy()
+        return output2 = audio[0].T.float().cpu().numpy(), pipe.vae.sampling_rate
 
 
-    button.click(describir, [textbox], output2)
+    button.click(describir, [textbox], output)
 
 demo.launch(debug=True)