_OpenAI(APIBaseModel) add default_headers in client_kwargs

[mirkenstein]

Added default_headers as available key in client_kwargs

Description

Sometimes custom OpenAI compatible endpoints require specific headers.

This is supported in the OpenAI library via the default_headers but is it not available in the super_duper adapter.

This was tested for connecting via local TGI server that is serving OpenAI compatible API.

from superduper_openai import OpenAIChatCompletion

llm = OpenAIChatCompletion(
    identifier="llm",
    model='meta-llama/Meta-Llama-3.1-8B-Instruct',
    client_kwargs={
        "base_url": "http://127.0.0.1:8080/v1",
        "api_key" : "NA",
        "default_headers": {"Content-Type": "application/json"}
    }
)
print(llm.predict("Tell me a joke"))

It does not brake existing functionality.

Related Issues

NA

Checklist

• [] Is this code covered by new or existing unit tests or integration tests?
• Did you run make unit_testing and make integration-testing successfully?
• Do new classes, functions, methods and parameters all have docstrings?
• Were existing docstrings updated, if necessary?
• Was external documentation updated, if necessary?

Additional Notes or Comments

[blythed]

Nice PR - thanks!