feat: support llm_config BYOK on jobs and retrieval

[suguanYang]

Summary

• Add LLMConfig / LLMProviderConfig / LLMModelsConfig and thread llm_config through jobs / parse / retrieval.
• Flat root {api_key, model, base_url} for multimodal; models for same-endpoint different model ids; text / vision for different endpoints.

Depends on: Ontos-AI/knowhere#217

Usage

# One multimodal modelllm_config= {
"api_key": "sk-...",
"model": "gpt-4o",
"base_url": "https://api.openai.com/v1",
}
# Same endpoint, different modelsllm_config= {
"api_key": "sk-...",
"base_url": "https://api.openai.com/v1",
"models": {"text": "gpt-4o-mini", "vision": "gpt-4o"},
}
# Two different endpointsllm_config= {
"text": {"api_key": "sk-...", "model": "gpt-4o-mini", "base_url": "https://api.openai.com/v1"},
"vision": {"api_key": "sk-ali-...", "model": "qwen-vl-max", "base_url": "https://dashscope.aliyuncs.com/compatible-mode/v1"},
}
client.parse(url="https://example.com/doc.pdf", llm_config=llm_config)
client.retrieval.query(query="...", use_agentic=True, llm_config=llm_config)

Test plan

• pytest tests/test_jobs.py tests/test_retrieval.py tests/test_parse.py