Hi @Logan M , with gpt-3.5-turbo-16k-

Seems like OpenAI APIs are having some issue 🤔

yeah, Idk if it's openai or the prompts we send

You could try checking the error using debugger and breakpoint on this file
antenv/lib/python3.11/site-packages/openai/_base_client.py
line 866


Maybe this help you find if the error is from OpenAI or the prompt that you use

Here is the debug log I've, with everything I think I can get:https://discord.com/channels/1059199217496772688/1059200010622873741/1180084518422577182

It is a timeout error, maybe try this



You can check if this solves your problem. You can increase it further as well. Default is 60 secs

No it does not solve y problem, as you can see the problem come from many retry

and this is anormal

the "retrying request"

retry happens when the API fails, Put
openai.log="debug" at the top and check if it prints anything related to the error

New OpenAI client does not print any log with the above changes.

Use this at the top before importing openai

You can also increase max_retries -- seems like openai is just a little unstable for "gpt-3.5-turbo-16k-0613"