Summary
基于 PyTorch 框架的昇腾 NPU 模型推理权重预取优化技能。为模型添加 torch_npu.npu_prefetch 特性以优化推理性能。触发场景:profiling 显示 MatMul/QBMM/GMM 算子存在 memory-bound 热点、需要为模型添加权重预取、将 prefetch 模式迁移到新模型。
ascend-ai-coding/awesome-ascend-skills
基于 PyTorch 框架的昇腾 NPU 模型推理权重预取优化技能。为模型添加 torch_npu.npu_prefetch 特性以优化推理性能。触发场景:profiling 显示 MatMul/QBMM/GMM 算子存在 memory-bound 热点、需要为模型添加权重预取、将 prefetch 模式迁移到新模型。
npx skills add https://github.com/ascend-ai-coding/awesome-ascend-skills
基于 PyTorch 框架的昇腾 NPU 模型推理权重预取优化技能。为模型添加 torch_npu.npu_prefetch 特性以优化推理性能。触发场景:profiling 显示 MatMul/QBMM/GMM 算子存在 memory-bound 热点、需要为模型添加权重预取、将 prefetch 模式迁移到新模型。
Related neighbors and high-traction skills in the same topics — useful to compare before installing.
Salesforce architecture diagrams using Mermaid with ASCII fallback.
5.1K installs>- XXE playbook. Use when XML, SVG, OOXML, SOAP, or parser-driven imports may resolve external …
3K installsAI-powered image generation for Salesforce visuals via Nano Banana Pro. Use this skill when the…
1.1K installs>- Final gate of the Flows app certification flow for external submission. Verifies that flows-…
1.1K installsFind and review Google Drive files shared outside the organization.
624 installsInvoke parallel document-specialist agents for external web searches and documentation lookup
618 installsOther skills from ascend-ai-coding/awesome-ascend-skills · top by installs.
npx skills add https://github.com/ascend-ai-coding/awesome-ascend-skills
Declared targets from SKILL.md / docs. Unmarked agents are not listed — the skill may still install via the CLI.
main