Update app.py

app.py

@@ -15,14 +15,6 @@ model = AutoModelForCausalLM.from_pretrained(
 
 @spaces.GPU(duration=60)  # cấp GPU tạm cho 1 phút
 def chat_with_model(prompt):
-    tokenizer = AutoTokenizer.from_pretrained(MODEL_ID, trust_remote_code=True)
-    model = AutoModelForCausalLM.from_pretrained(
-        MODEL_ID,
-        torch_dtype=torch.float16,
-        device_map="auto",
-        trust_remote_code=True
-    )
-
     inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
     output = model.generate(
         **inputs,