fixbug and support for ollama and other llms
Description
This great program currently supports OpenAI services, which seems to limit its development. In actual enterprise production activities, we usually localize the deployment of some open source models and services because it can effectively protect privacy data. Based on the above requirements, I have changed the part of the large model calling to ollama. I hope you can consider my opinions and needs. Or add new branch sections, and in addition, I have also fixed some minor issues that I have discovered in practice.
microsoft-github-policy-service agree [company="8lab"]
@liuda-666 Please! They're waiting on you i think
See new docs on injecting custom model providers