Hello,
I am trying to translate a document using python script. We are having request timeouts, added payment information, bought credit for that. We are gonna run the script on a virtual server since the spreadsheet is a large one. Chat GPT translation service is by far the best around, comparing to other platforms. Although we have bought credits, the problem is still there. Is there any suggestions?
Welcome to the OpenAI community @murgursoy
You can check if your rate limits here.
Usually rate limits increase after 24 hours of adding payment method.
Can you share the exact error you’re getting?
I’ve only been able to read the promised rate limits during short peak usage, so I suspect the the rate limits for sustained usage, such as translating a large number of documents over multiple hours or even days, is actually much lower.
For dealing with timeouts, some information is here OpenAI Platform Either look at the response headers and wait accordingly, or use " exponential backoff" : basically just wait longer with every timeout and retry.
@sps Have you ever experienced the actual rate limits for continuous/sustained use being much tighter? Earlier, I did not get any info from openai support if this is the case.
Usage is metered by the minute.
You can batch until you are near your limit or until you get the rate limit error, and then hold off until 60 more seconds.
Yes, if the Requests per minute and token per minute limits aren’t honored, you’ll get errors.
Also when running for longer durations, the RPD - requests per day also come into play.
The rate limits aren’t affected by the use, in fact they affect the use.