serving-llms-vllm
Serves LLMs with high throughput using vLLM's PagedAttention and continuous batching. Use when deploying production LLM APIs, optimizing inference latency/throughput, or serving models with limited GP
Also installable via skills CLI
npx skills add zechenzhangAGI/AI-research-SKILLs/devops/serving-llms-vllm-zechenzhangagi-ai-research-skills
Source
Path:
devops/serving-llms-vllm-zechenzhangagi-ai-research-skills(main)