add azure openai api support to ontogpt

[k-bandi]

This PR adds support for using Azure OpenAI API endpoints in OntoGPT for the LLM component instead of Vanilla OpenAI endpoints.

[caufieldjh]

Thanks @k-bandi !
I will make some updates to this to ensure we retain the base functionality, but Azure support is certainly appreciated!

[k-bandi]

@caufieldjh - Thank you. Should I close this PR now?

[caufieldjh]

Please leave it open! I'm very happy to have your contribution and would like to integrate it.

[k-bandi]

Okay great. Thank you. Let me know if you need anything.

[caufieldjh]

Just getting back to this now - thanks for your patience @k-bandi. Should be able to get it working and merged today.

[caufieldjh]

OK, this should be ready to use! Two notes:

• There is now a CLI flag named --azure-select. Use this with a command like extract to use an Azure endpoint, though it will still expect the configuration details to be defined in a local.toml.
• I'm going to be refactoring OntoGPT more soon to reduce the amount of model changes we have to track in this project (i.e., we'll let other packages like llm and litellm do that) but will try to keep that CLI flag as it's a useful shorthand. This may mean that the process for storing model configurations will change, however.
• I don't have an Azure endpoint to test with, so I'm making some assumptions about how well this works. Please let me know if it is broken or doesn't work as expected!

[k-bandi]

Thanks @caufieldjh for looking into this.