Update app.py

app.py

@@ -1,6 +1,7 @@
 import os
 import gradio as gr
 import torch
+import numpy as np
 
 # ✅ Forzar consentimiento de licencia
 os.environ["COQUI_TOS_AGREED"] = "1"
@@ -19,11 +20,11 @@ from TTS.api import TTS
 tts = TTS(model_name="tts_models/multilingual/multi-dataset/xtts_v2")
 
 # 🎤 Función principal
-def generate_audio(text, language, speaker_wav=None):
+def generate_audio(text, language, speaker_wav):
     if speaker_wav is not None:
-        return tts.tts(text=text, speaker_wav=speaker_wav, language=language)
-    else:
-        return tts.tts(text=text, language=language)
+        audio = tts.tts(text=text, speaker_wav=speaker_wav, language=language)
+    audio_np = np.array(audio, dtype=np.float16)
+    return (22050, audio_np)
 
 # 🎛️ Interfaz
 iface = gr.Interface(
@@ -31,7 +32,7 @@ iface = gr.Interface(
     inputs=[
         gr.Text(label="Texto"),
         gr.Text(label="Idioma (ej: 'es', 'en')"),
-        gr.Audio(type="filepath", label="Audio de voz (opcional)")
+        gr.Audio(type="filepath", label="Audio de voz")
     ],
     outputs=gr.Audio(label="Audio generado")
 )