Update utils.py

utils.py

@@ -59,12 +59,13 @@ def load_tokenizer_and_model(base_model, load_8bit=False):
         device = "cuda"
     else:
         device = "cpu"
-
+    
     
     tokenizer = AutoTokenizer.from_pretrained(base_model, use_fast = True, use_auth_token=True)
     if device == "cuda":
         model = AutoModelForCausalLM.from_pretrained(
             base_model,
+            trust_remote_code=True,
             load_in_8bit=load_8bit,
             torch_dtype=torch.float16,
             device_map="auto",