Hi,
This is the first time I’m positing here so I am not sure about the category and how it works here.
I am using GPT 3.5 turbo to search through a list of items and find the tools I need to use for the task. My task explanation and instruction together builds up to around 78 tokens but as soon as I add instructions as system role which brings up the token number to around 800, then I start getting code 400. There is no message in the response to know what the exact reason is or how many tokens the content is above the limit.
But considering the number of tokens of my prompt and limit for GPT 3.5 turbo, why should I have reached the limit already?
I am implementing the OpenAI APIs in android app. And the system instruction is generated by the OpenAI itself for GPT 3.5 turbo.
What should I do?