Find answers from the community

Updated 5 months ago

Hi!

At a glance
Hi!
I'm hitting the rate limits when generating questions from my documents with generate_questions_from_nodes() but there does not seem to be parameters to control the frequency of API calls for DataGenerator.

Plain Text
RateLimitError: Error code: 429 - {'error': {'message': 'Rate limit reached for gpt-3.5-turbo [...]


I see this issue already arised for embeddings generation, but did anyone get trouble when generating questions ?
L
M
5 comments
hmmm I could have sworn we fixed this, but I guess not for this particular function.

I know the fix to make in code, but for now, I would run on batches of nodes to limit the number of API calls
Makes sense, thanks, will work around! Created https://github.com/run-llama/llama_index/issues/9487 in case you want to track this one
cool thanks!
The num parameter seems broken. Proposed a fix here https://github.com/run-llama/llama_index/pull/9489
good catch, that makes sense!
Add a reply
Sign up and join the conversation on Discord