feat: Add vLLM LLM support with OpenAI-compatible API

[SeasonPilot]

When submitting a PR, please confirm the following points and put [x] in the boxes one by one. | 在提出pr时，请确认了以下几点，并逐一使用[x]符号确认勾选。

Checklist | 检查项

• I have read and understood the contributor guidelines. | 我已阅读并理解贡献者指南 。
• I have checked for any duplicate features related to this request and communicated with the project maintainers. | 我已检查没有与此请求重复的功能并与项目维护者进行了沟通。
• I accept the suggestion of the maintainers to make changes to or close this PR. | 我接受此PR配合维护人员的建议进行修改或关闭。
• I have submitted the test files and can provide screenshots of the test results (required for feature or bug fixes) | 我已经提交了测试文件并可提供测试结果截图(功能修改、BUG修复类PR必须提供，其他按需)
• I have added or modified the documentation related to this PR | 我已经添加或修改了本次pr对应的文档说明(非必要，根据实际PR内容按需添加)
• I have added examples and notes if needed | 我已经添加了使用案例代码与文档说明(非必要，根据实际PR内容按需添加)

Please fill in the specific details of this PR: | 请详细填写本次PR的内容:

• Implemented comprehensive vLLM LLM integration with OpenAI-compatible API support for high-performance inference (24x faster than HuggingFace Transformers, 50-70% memory reduction)
• Added VLLMOpenAIStyleLLM class extending OpenAIStyleLLM with vLLM-specific parameters (beam_search, best_of, length_penalty, early_stopping) and pre-configured context lengths for 30+
  popular models including Llama 2/3.1/3.2, Mistral/Mixtral, Qwen 2/2.5, Yi, DeepSeek, and Phi families
• Created complete example configurations (Llama 3.1 8B/70B, Qwen 2.5 7B) with comprehensive deployment guide including basic server setup, Docker deployment, multi-GPU tensor
  parallelism, quantization options, performance tuning, and troubleshooting

Please provide the path of test files and submit screenshots or files of the test results(fill in as needed): | 请填写测试文件路径并提供测试结果截图或文件(按需填写):

• Test file path : tests/test_agentuniverse/unit/llm/test_vllm_openai_style_llm.py
• Test results : ✅ 7/7 tests passed (7 passed in 0.97s)
  • test_initialization - Validates VLLMOpenAIStyleLLM initialization with correct parameters
  • test_max_context_length - Verifies context length retrieval for Llama 3.1 (131072 tokens)
  • test_vllm_specific_parameters - Confirms vLLM-specific parameters are properly set
  • test_get_num_tokens - Tests token counting functionality
  • test_llama_context_lengths - Validates Llama family context lengths (2/3.1/3.2)
  • test_mistral_context_lengths - Validates Mistral/Mixtral family context lengths
  • test_qwen_context_lengths - Validates Qwen 2/2.5 family context lengths

Please list the names of the docs that were added or modified in this PR (fill in as needed): | 请列出本次PR新增或修改的文档名称(按需填写):

• agentuniverse/llm/default/vllm_openai_style_llm.py - Core implementation with comprehensive docstrings and usage examples (230 lines)
• agentuniverse/llm/default/vllm_openai_style_llm.yaml - YAML configuration template
• examples/sample_standard_app/intelligence/agentic/llm/buildin/vllm/README.md - Comprehensive deployment guide (268 lines) including installation, 4 deployment options, configuration
  examples, performance tuning, troubleshooting, and cost comparison
• examples/sample_standard_app/intelligence/agentic/llm/buildin/vllm/vllm_llama_3_1_8b.yaml - Basic Llama 3.1 8B configuration example
• examples/sample_standard_app/intelligence/agentic/llm/buildin/vllm/vllm_llama_3_1_70b.yaml - Advanced Llama 3.1 70B configuration with beam search optimization
• examples/sample_standard_app/intelligence/agentic/llm/buildin/vllm/vllm_qwen_2_5_7b.yaml - Qwen 2.5 7B configuration for multilingual support (Chinese/English)

Related Issue: #250

[SeasonPilot]

#250