fix RC bug in reddit human experiment (#67)

Thanks very much to lily_wu for finding this problem!

Author: echo-yiyiyi

scripts/reddit_simulation_align_with_human/business_3600.yaml

@@ -23,5 +23,5 @@ inference:
   model_path: /ibex/user/yangz0h/open_source_llm/llama-3
   stop_tokens: [<|eot_id|>, <|end_of_text|>]
   server_url:
-    - host: 10.109.1.8
+    - host: 10.109.1.9
       ports: [8002, 8003, 8005]

scripts/reddit_simulation_align_with_human/reddit_simulation_align_with_human.py

@@ -162,9 +162,12 @@ async def export_data(i):
             if rs_rc_index >= len(pairs):
                 return
             else:
-                content = pairs[rs_rc_index]["RC_1"]["body"]
+                title = pairs[rs_rc_index]["RS"]["title"]
+                content = pairs[rs_rc_index]["RS"]["selftext"]
+                formatted_content = f"Title: {title}.\nContent: {content}"
+
                 response = await post_agent.perform_action_by_data(
-                    "create_post", content=content)
+                    "create_post", content=formatted_content)
                 post_id = response["post_id"]
                 for i in range(1, 11):
                     key_name = f"RC_{i}"