Update app.py

app.py

@@ -6,6 +6,7 @@ import uvicorn
 from dotenv import load_dotenv
 from difflib import SequenceMatcher
 from tqdm import tqdm
+import threading
 
 load_dotenv()
 
@@ -112,23 +113,27 @@ async def generate_chat(request: ChatRequest):
     
     print(f"Procesando solicitud: {request.message}")
 
-    with ProcessPoolExecutor() as executor:
-        futures = [executor.submit(worker_function, llm, request) for llm in llms.values()]
-        responses = []
+    responses = []
+    threads = []
+
+    # Crear un hilo para cada modelo
+    for llm in llms.values():
+        thread = threading.Thread(target=lambda: responses.append(worker_function(llm, request)))
+        threads.append(thread)
+        thread.start()
 
-        for future in tqdm(as_completed(futures), total=len(futures), desc="Generando respuestas"):
-            response = future.result()
-            responses.append(response['response'])
-            print(f"Modelo procesado: {response['literal'][:30]}...")
+    # Esperar a que todos los hilos terminen
+    for thread in threads:
+        thread.join()
 
     # Seleccionar la mejor respuesta
-    best_response = select_best_response(responses)
+    best_response = select_best_response([response['response'] for response in responses])
     
     print(f"Mejor respuesta seleccionada: {best_response}")
 
     return {
         "best_response": best_response,
-        "all_responses": responses
+        "all_responses": [response['response'] for response in responses]
     }
 
 if __name__ == "__main__":