I’m wondering if anyone is seeing variable response rates (e.g., sometimes the response is 5s while another will be 5m). I am seeing this behavior:
API CALL 1 TAKES PLACE at 9:30AM
API CALL 2 TAKES PLACE at 9:32AM
API CALL 2 RECEIVES RESPONSE at 9:32AM
API CALL 1 RECEIVES RESPONSE at 9:35AM
This is using the gpt-3.5-turbo API.
Are both API calls different in terms of request parameters? We need more details to debug this. Your whole code, if possible
Hi @rasengan,
I think because others were experiencing timeout A lot of timeout error in the last few days that this also might be related to some changes and scale ups in OpenAI’s infrastructure - but this is only a guess, as @AgusPG mentioned, more details on your requests would help to provide a more specific answer 