Add article link and fix history prompt

Add the related article and add the forgotten history format to inject
the user question.

app.py

@@ -121,7 +121,7 @@ def bot(history: list, max_num_tokens: int, final_num_tokens: int):
         t.start()
 
         # rebuild the history with the new content
-        history[-1].content += prepend
+        history[-1].content += prepend.format(question=question)
         if ANSWER_MARKER in prepend:
             # stop thinking, this is the answer now (no metadata for intermediate steps)
             history.append(gr.ChatMessage(role="assistant", content=""))
@@ -143,6 +143,8 @@ with gr.Blocks(fill_height=True, title="Making any model reasoning") as demo:
             This is a simple proof-of-concept to get any LLM model to reason ahead of its response.
             This interface uses *{model_name}* model which is **not** a reasoning model. The used method
             is only to force some "reasoning" steps with prefixes to help the model to enhance the answer.
+
+            See related article here: [Make any model reasoning](https://huggingface.co/blog/Metal3d/making-any-model-reasoning)
             """)
             chatbot = gr.Chatbot(
                 scale=1,