Update app.py

app.py

@@ -1,20 +1,16 @@
 import gradio as gr
 import torch
-from transformers import AutoModelForCausalLM, AutoTokenizer, BitsAndBytesConfig
+from transformers import AutoModelForCausalLM, AutoTokenizer
 
 # Configuración del modelo
 model_id = "somosnlp-hackathon-2025/leia_preference_model_social_norms"
 device = "cuda" if torch.cuda.is_available() else "cpu"
 
-# Config vacía para evitar errores de quantización
-quant_config = BitsAndBytesConfig(load_in_4bit=False, load_in_8bit=False)
 
 # Cargar modelo y tokenizer
 tokenizer = AutoTokenizer.from_pretrained(model_id)
-model = AutoModelForCausalLM.from_pretrained(
-    model_id,
-    quantization_config=quant_config
-).to(device)
+model = AutoModelForCausalLM.from_pretrained(model_id).to(device)
+
 
 # Función para responder
 def respond(message, history, system_message, max_tokens, temperature, top_p):