we are using gpt-3.5-turbo model using API. Response is different each time. I am keeping temperature as 0 and top_p as 0.9. I have tried various variations of both but still responses are not consistent. Can any one please help Thank you
