How do you consider the 'top_k' parameter when using openai_api_server to start?

[garyyang85]

I would like to try 'top_k' in my scenario to see if anything different in the output, but seems the openai invoke method does not support this parameter. https://platform.openai.com/docs/api-reference/chat/create

I can find the logic to handle 'top_k' in https://github.com/lm-sys/FastChat/blob/main/fastchat/serve/openai_api_server.py, so I am wondering if fastchat support any other ways to use this parameter?

Many thanks for any suggestions!