llms-vllm: fix VllmServer to work without CUDA-required vllm core. (#12003)

docs/examples/llm/vllm.ipynb

.../llms/llama-index-llms-vllm/llama_index/llms/vllm/base.py

...ex-integrations/llms/llama-index-llms-vllm/pyproject.toml

...ions/llms/llama-index-llms-vllm/tests/test_integration.py

...ations/llms/llama-index-llms-vllm/tests/test_llms_vllm.py