feat(backend): add authentication support to OpenAIHTTPBackend

[didlawowo]

Summary

This PR adds authentication support to OpenAIHTTPBackend, enabling GuideLLM to work with OpenAI-compatible servers that require authentication on all endpoints (including /health).

Problem

When using GuideLLM with LiteLLM proxy or similar OpenAI-compatible servers that require authentication, the benchmark fails during backend validation because the /health endpoint check doesn't include authentication headers:

RuntimeError: Backend validation request failed. Could not connect to the server or validate the backend configuration.

The OpenAISettings class already defines api_key, bearer_token, and headers fields, but these were never used in the actual HTTP requests.

Solution

• Add api_key, bearer_token, and headers parameters to OpenAIHTTPBackend.__init__()
• Read default values from settings.openai when not explicitly provided
• Apply authentication headers to:
  • Backend validation requests (/health endpoint)
  • Available models requests (/v1/models endpoint)
  • All generation requests (merged with request-specific headers)

Usage

# Via environment variables (recommended)
export GUIDELLM__OPENAI__API_KEY=sk-xxx
guidellm benchmark run --target http://litellm:4000 ...

# Via backend kwargs
guidellm benchmark run --target http://litellm:4000 \
  --backend-kwargs '{"api_key": "sk-xxx"}' ...

Priority Order for Headers

1. Request-specific headers (highest priority)
2. Explicit api_key/bearer_token/headers parameters
3. Settings from settings.openai (lowest priority)

Testing

• Tested with LiteLLM proxy requiring master key authentication
• /health endpoint now returns 200 instead of 401
• Benchmarks complete successfully against authenticated endpoints

Related

This makes the documented example in the docstring actually work:

backend = OpenAIHTTPBackend(
    target="http://localhost:8000",
    model="gpt-3.5-turbo",
    api_key="your-api-key"  # This param was in the example but didn't exist!
)