Update app.py

app.py

@@ -5,7 +5,7 @@ import numpy as np
 import requests
 import torch
 import os
-from transformers import AutoModelForCausalLM, AutoTokenizer, BitsAndBytesConfig
+from transformers import AutoModelForCausalLM, AutoTokenizer
 from langchain.memory import ConversationBufferMemory
 
 # Configuración del modelo de lenguaje
@@ -18,16 +18,16 @@ if not HF_TOKEN:
 
 print("🔄 Cargando modelo de lenguaje...")
 tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME, token=HF_TOKEN)
-bnb_config = BitsAndBytesConfig(load_in_8bit=True)
+
 
 model = AutoModelForCausalLM.from_pretrained(
     MODEL_NAME,
     torch_dtype=torch.float16, 
-    quantization_config=bnb_config,  
     device_map="auto",
     token=HF_TOKEN
 )
 
+
 # Memoria conversacional
 memory = ConversationBufferMemory()