feat(adapters): Add support for extra headers to more backend providers

[planetf1]

Which issue(s) does this pull-request address?

Closes: #586

Description

Adds support for extra headers to:

• Azure OpenAI
• Anthropic
• Amazon Bedrock
• Google VertexAI

In the above cases support for extra_headers is documented in the LiteLLM docs and/or found from a code scan

Changes were not made to

• Ollama (note there is a PR open)
• WatsonX
• Groq
• XAI

due to lack of support in liteLLM.

• OpenAI adapter already has support

Testing
IN PROGRESS

• Check modified providers still work with no extra headers
• check extra headers sent by network traffic inspection

No test is made that the extra headers have any effect on the actual provider.

Checklist

General

• I have read the appropriate contributor guide: Python
  / TypeScript
• I have signed off on my commit: Python instructions / TypeScript instructions
• Commit messages and PR title follow conventional commits
• Appropriate label(s) added to PR: Python for Python changes, TypeScript for TypeScript changes

Code quality checks

• Linting passes: Python poe lint or poe lint --fix / TypeScript yarn lint or yarn lint:fix
• Formatting is applied: Python poe format or poe format --fix / TypeScript: yarn format or yarn format:fix
• Static type checks pass: Python poe type-check

Testing

• Unit tests pass: Python poe test --type unit / TypeScript yarn test:unit
• E2E tests pass: Python poe test --type e2e / TypeScript: yarn test:e2e
• Integration tests pass: Python poe test --type integration
• Tests are included (for bug fixes or new features)

Documentation

• Documentation is updated
• Embedme embeds code examples in docs. To update after edits, run: Python poe docs --type build

[planetf1]

Testing:

Bedrock, Vertex, Anthropic:

In addition to showing the extra_headers being set on the LiteLLM API call,
Header's show in liteLLM log as part of the curl (which makes sense) as:

curl -X POST \
https://bedrock-runtime.eu-west-2.amazonaws.com/model/meta.llama3-8b-instruct-v1:0/converse \
-H 'Content-Type: *****' -H 'sample-header: *****' -H 'X-Amz-Date: *****' -H 'Authorization: secret1 Credential=cred1/part2/eu-west-2/bedrock/aws4_request, SignedHeaders=content-type;host;sample-header;x-amz-date, Signature=0000230230********************************************' -H 'Content-Length: *****' \
-d '{"messages": [{"role": "user", "content": [{"text": "what states are part of New England?"}]}], "additionalModelRequestFields": {}, "system": [], "inferenceConfig": {"temperature": 0}}'

This is different from OpenAI (and Azure OpenAI) where there is no extra -H shown in the litellm trace, but instead the additional 'header' actually features within the body

curl -X POST \
https://randomname-location.cognitiveservices.azure.com/openai/deployments/gpt-4o-mini/ \
-H 'api_key: **obfuscated**********************************************' -H '
azure_ad_token: *****' \
-d '{'model': 'gpt-4o-mini', 'messages': [{'role': 'user', 'content': [{'type': 'text', 'text': 'what states are part of New England?'}]}], 'temperature': 0, 'stream': False, 'extra_body': {}, 'extra_headers': {'sample-header': 'abcdjfkhskdjfhskjhfksdhjfsifhskdjhfsdf'}}'

I am unable to test bedrock, vertex,anthropic, bedrock specifically in terms of the headers -- but the API is called correctly as per docs. We do also know the openai connector does work as we have a valid test system that requires headers to test against, so the fact azure behaves the same is also logical.