Generate embeddings using batch requests

[dluc]

Original issue: microsoft/semantic-kernel#2168 from @Nurgo

Hi,

I've been working with SemanticKernel and I appreciate all the hard work you've put into this amazing project.

I'm currenly encountering a limitation with the ISemanticTextMemory.SaveInformationAsync method which only processes one text segment at a time, causing many calls to OpenAI's APIs when indexing large documents.

Would it be possible to modify this function to allow batch processing in a single call? Since IEmbeddingGeneration.GenerateEmbeddingsAsync already supports batching, it might be a feasible and impactful improvement.

Thanks for considering this suggestion.

[xbotter]

Is there any progress?

[dluc]

hi @xbotter, I don't think we'll have a chance to implement this anytime soon, sorry

[dluc]

Work is in progress, see latest PRs :-)