Add chat template directly to tokenizer

app.py

@@ -22,6 +22,7 @@ def messages_for(python):
     ]
 
 tokenizer = AutoTokenizer.from_pretrained(MODEL)
+tokenizer.chat_template = CHAT_TEMPLATE
 model = AutoModelForCausalLM.from_pretrained(MODEL, torch_dtype=torch.bfloat16, device_map="auto")
 model.eval()
 
@@ -32,7 +33,6 @@ cplusplus = None
 def translate(python):
     inputs = tokenizer.apply_chat_template(
                         messages_for(python),
-                        chat_template=CHAT_TEMPLATE,
                         return_tensors="pt").to(model.device)
     generation_kwargs = dict(
                             inputs,
@@ -41,7 +41,7 @@ def translate(python):
                             do_sample=False,
                             pad_token_id=tokenizer.eos_token_id,
                             eos_token_id=tokenizer.eos_token_id
-                            )
+                        )
     thread = Thread(target=model.generate, kwargs=generation_kwargs)
     thread.start()
     cplusplus = ""