[{"data":1,"prerenderedAt":32},["ShallowReactive",2],{"project:/projects/tutorials/llm-serving-opt":3},{"id":4,"title":5,"authors":6,"body":7,"date":14,"demo":15,"description":16,"descriptionZh":17,"extension":18,"featured":19,"github":6,"mark":20,"meta":21,"navigation":22,"paper":6,"path":23,"role":6,"seo":24,"stem":25,"techStack":26,"thumbnail":6,"titleZh":30,"venue":6,"venueFull":6,"__hash__":31},"projects/projects/tutorials/llm-serving-opt.md","LLM Serving and Optimization: A Hands-On Tutorial",null,{"type":8,"value":9,"toc":10},"minimark",[],{"title":11,"searchDepth":12,"depth":12,"links":13},"",3,[],"2026-11","https://tutorials.zrp.cool/hands-on-llm-serving-opt/","A self-study tutorial on LLM inference serving: the memory and bandwidth ledgers, continuous batching, paged KV cache, quantization, and how 2026 production systems put it together.","从显存和带宽这两笔账讲起的大模型推理服务教程：连续批处理、分页 KV Cache、量化，一路讲到一个能上线的推理服务。","md",false,"LLM",{},true,"/projects/tutorials/llm-serving-opt",{"title":5,"description":16},"projects/tutorials/llm-serving-opt",[27,28,29],"Docsify","Vue","KaTeX","LLM Serving 与 Optimization 调优教程","D8LkQWT-6FioT1GKckO2zm8zO64SHPijLUfKkzSbB-c",1794930670152]