uniqueaakash13
Registered Member
- Jun 22, 2014
- 52
- 9
Hello everyone,
I'm currently developing an application using the OpenAI API, and I'm aiming to receive a new response each time the API is called. However, I've encountered an issue where I'm getting the same response repeatedly. One workaround I've discovered involves creating the prompt to include all previous messages, then adding a prompt something like "new response other than those mentioned above." This method works but it significantly increases the token count, leading to higher costs.
Has anyone discovered a more efficient approach?
I'm currently developing an application using the OpenAI API, and I'm aiming to receive a new response each time the API is called. However, I've encountered an issue where I'm getting the same response repeatedly. One workaround I've discovered involves creating the prompt to include all previous messages, then adding a prompt something like "new response other than those mentioned above." This method works but it significantly increases the token count, leading to higher costs.
Has anyone discovered a more efficient approach?
