Azure Embeddings Quota Limit

[danieldekay]

I am running GPTR and experience a quota limit in my subscription. While I have just asked for an extended quota, I am puzzled that I could also "just have waited for 1s".

Does anyone know if there would be a right point in GPTR to implement a retry with backoff around API calls that could be quota-limited? for example using -- https://pypi.org/project/backoff/

Error running job: Error code: 429 - {'error': {'code': '429', 'message': 'Requests to the Embeddings_Create Operation under Azure OpenAI API version 2024-02-15-preview have exceeded call rate limit of your current OpenAI S0 pricing tier. Please retry after 1 second. Please go here: https://aka.ms/oai/quotaincrease if you would like to further increase the default rate limit.'}}

Since Embeddings_Create is not used in GPTR's code, I suspect this being in langchain's code somewhere -- which is called by GPTR. But where?

Langchain recommends setting the maxConcurrency option. https://js.langchain.com/v0.1/docs/modules/data_connection/text_embedding/rate_limits/