Abesh Bhattacharjee

57 Followers
59 Following
58 Posts
I am Sodium...
Opinions are mine only and not of my employer.

Have you encountered rate limiting (429) on requests to the new "text-embedding-3-small" model of OpenAI in SAP AI Core on BTP ?

I mean, I understand a moderate limit but I get a 429 after about 10 to 12 consecutive calls which is a pretty low number ? I do not have this issue with calls to GPT models (gpt-3.5-turbo etc) or text-embedding-ada-002 for instance.

Free usage limit for that model on OpenAI's website is around 3000 RPM.
https://platform.openai.com/docs/guides/rate-limits/usage-tiers

Nice nod to Harry Potter in the SAP BTP APIM logs :)
Nice Game of Thrones reference in the cds command line :) To the guy from SAP who wrote the code: Valar Morghulis !
Recursion with async await in javascript is beautiful :)
Use case: Create a nested menu in SAPUI5 which represents the folder structure of a particular folder in a cloud drive to a specified depth.
30 lines of code, output of 'createMenuArray' feeds 'createMenu' :) Misses try catch block for error handling which will be added later :)
Hello World !
Looking forward to drinking from the firehose :)