Update app.py

app.py

@@ -28,9 +28,10 @@ with gr.Blocks(theme=gr.themes.Ocean(primary_hue="pink", neutral_hue="indigo", f
 
     def describir(url):
       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)
-      return processor.decode(out[0], skip_special_tokens=True)
+      #inputs = processor(raw_image, return_tensors="pt").to("cpu")
+      #out = model.generate(**inputs)
+      #return processor.decode(out[0], skip_special_tokens=True)
+      return "description"
 
     def leer(texto):
         response = requests.post("https://charly-text-to-speech.hf.space/run/predict", json={
@@ -38,10 +39,20 @@ with gr.Blocks(theme=gr.themes.Ocean(primary_hue="pink", neutral_hue="indigo", f
             texto,
         ]}).json()
         
-        print(response)
-        data = response['data']
+        data = response['data'][0]
+    
+        # Extraer la parte de base64 del string (eliminar el prefijo 'data:audio/flac;base64,')
+        audio_base64 = data.split(',')[1]
+    
+        # Decodificar el string base64
+        audio_data = base64.b64decode(audio_base64)
         
-        return data
+        # Crear un archivo temporal
+        with tempfile.NamedTemporaryFile(delete=False, suffix='.flac') as temp_audio_file:
+            temp_audio_file.write(audio_data)
+            temp_audio_path = temp_audio_file.name
+        
+        return temp_audio_path
 
 
     button.click(describir, [textbox], output)