Create last_turn_prompt.txt

movie_recommendation/last_turn_prompt.txt

@@ -0,0 +1,6 @@
+Now apply whatever you learnt about the user and predict which movie will satisfy the user the most from the following set:
+
+{unseen_movie_sample}
+
+Reason about the linear user preference function and then provide your final answer. If you decide to make any calculations, ensure that you're providing the final answer in the same response. You don't have access to any external tools such as pandas or calculator.
+It is compulsory to provide a final answer now, in this turn, as the conversation will end after your response. Make sure you enclose the final answer under the <final_answer> </final_answer> markers. The final answer inside the markers should strictly only have the exact movie name and nothing else.