debug

app.py

@@ -181,7 +181,7 @@ def answer(system: str, context: str, question: str,
 
         # --- generate ------------------------------------------------------
         output = chat_model.generate(
-            **tokens,
+            prompt,
             max_new_tokens=512,
             max_length=MAX_PROMPT_TOKENS + 512,
             do_sample=True,
@@ -190,7 +190,7 @@ def answer(system: str, context: str, question: str,
             top_k=top_k_tok
         )
         full   = tokenizer.decode(output[0], skip_special_tokens=True)
-        reply  = full.split("<|im_start|>assistant")[-1].strip() + tokenizer.chat_template
+        reply  = full.split("<|im_start|>assistant")[-1].strip()
         return reply
           
     except Exception as e: