vllm-ascend-tuning
Skill2 starsChanged 13 days agoNo licenceView on GitHub
name: vllm-ascend-tuning description: "vLLM-Ascend 推理服务全链路性能调优。当用户要求提升 vLLM-Ascend 推理速度/吞吐量、降低延迟（TTFT/TPOT）、选择并行策略（TP/DP/EP）、配置 Graph Mode/Speculative Decoding/量化/PD 分离、或做性能基准测试时使用。涵盖并行策略、编译优化、OS 级调优、torch_npu、CANN/HCCL、vLLM 参数等 Phase 0–11，内置主流模型已验证 YAML 配置。不用于 Prof 数据的事后分析与瓶颈定位（用 cluster-analysis）；不用于故障报错排查（用 issue-analyzer）；不包含服务部署启动本身。" \# vLLM-Ascend 性能调优 本技能提供 vLLM-Ascend 系统级性能调优的完整工作流，涵盖从并行策略、编译、OS、框架到模型推理的全链路优化。 \- 用户要求提升 vLLM…
- Installs packages
In this repository7 other files it gives its agents.
- ascend-dump-analyzerSkill
- cluster-analysisSkill
- cluster-compareSkill
- cpu-host-performanceSkill
- ibmc_analyzerSkill
- issue-analyzerSkill
- prof-compareSkill
The file
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.