graphrag icon indicating copy to clipboard operation
graphrag copied to clipboard

Fixed an issue where base OpenAI embeddings can't work with Azure Ope…

Open KennyStryker opened this issue 1 year ago • 0 comments

Description

When you try to use base OpenAI embeddings with Azure OpenAI LLM deployment, you can index the document because hydrate_embeddings_params replaced None with Azure OpenAI base_url.

Related Issues

#492

Proposed Changes

This PR fixes the issues, such that all following event work as expected:

  • same api_bases for LLM and embeddings (both Azure)
  • different api_bases for LLM and embeddings (both Azure)
  • LLM uses Azure OpenAI, while embeddings uses base OpenAI (this one is important)
  • LLM uses Azure OpenAI, while embeddings uses third-party OpenAI-like API

Checklist

  • [X] I have tested these changes locally.
  • [x] I have reviewed the code changes.
  • [ ] I have updated the documentation (if necessary).
  • [ ] I have added appropriate unit tests (if applicable).

Additional Notes

[Add any additional notes or context that may be helpful for the reviewer(s).]

KennyStryker avatar Jul 10 '24 17:07 KennyStryker