Update utils.py

utils.py

@@ -65,7 +65,7 @@ def load_tokenizer_and_model(base_model, load_8bit=False):
     if device == "cuda":
         model = AutoModelForCausalLM.from_pretrained(
             trust_remote_code=True, 
-            base_model,
+            model=base_model,
             load_in_8bit=load_8bit,
             torch_dtype=torch.float16,
             device_map="auto",