Update app.py

app.py

@@ -11,12 +11,13 @@ load_in_4bit = False
 # Initialize model and tokenizer (run only once using st.cache_resource)
 @st.cache_resource
 def load_model_and_tokenizer(model_name, dtype, load_in_4bit):
-    tokenizer = AutoTokenizer.from_pretrained(model_name)
+    tokenizer = AutoTokenizer.from_pretrained(model_name, trust_remote_code=True)
     model = AutoModelForCausalLM.from_pretrained(
         model_name,
         torch_dtype=dtype,
         load_in_4bit=load_in_4bit,
         device_map="auto",
+        trust_remote_code=True,
     )
     model.eval()
     return model, tokenizer