I'm putting up this PR because the Azure Open AI Models currently hit rate limits when indexing multiple files, which is an issue to be looked into later. I tested that the Phi model works better by running the indexer across multiple files. Screenshot attached below:
This change is about the same thing as the one to enable indexers to work with the custom language model hosted in Azure AI Studio - https://github.com/Azure-Samples/azure-search-power-skills/pull/207.
I'm putting up this PR because the Azure Open AI Models currently hit rate limits when indexing multiple files, which is an issue to be looked into later. I tested that the Phi model works better by running the indexer across multiple files. Screenshot attached below: