Describe your problem

"I have a very complex graph model, and in one of the processes, I connect an answer generation component with a human-computer interaction component. In this stage, I need to continuously ask the user questions, and each conversation requires the complete history of the dialogue to ensure that the next question does not repeat previous ones. Therefore, I want to know how many past conversations are appended with each API call? And whether any summarization is performed? Additionally, how can I append historical information? My prompt works fine with GPT-4o and can complete multi-turn questioning as required. However, in the ragglow-graph, after several rounds of dialogue, it starts asking repetitive questions. I am also using the GPT-4o model, but I cannot see the detailed OpenAPI POST body, so I cannot determine where the problem is."