[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M78VXCBY1WWZ5S2EBDRMF651":3,"skill-same:01M78VXCBY1WWZ5S2EBDRMF651":75},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":15,"categories":16,"keywords":17,"publishedAt":18,"linkOnly":19,"markdownContent":20,"markdownBody":21,"frontMatter":22,"refName":27,"commitSha":28,"skillDirectory":29,"forksCount":30,"repositoryHtmlUrl":31,"repositoryHomepageUrl":6,"ownerHtmlUrl":32,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":33,"install":34,"resources":46},"01M78VXCBY1WWZ5S2EBDRMF651","deepspeed",null,"Use DeepSpeed for distributed training, inference acceleration, ZeRO configuration, parallelism\u002FMoE design, profiling, autotuning, and operational diagnostics.","VectorSpaceLab\u002FAREX-Skill","VectorSpaceLab","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F180637102?v=4","01M7538C5WPSJGWF7KHR4EFZ9F","01M7538C6R2GFP53NPXTCZNBRT",351,"Apache-2.0","https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fdeepspeed\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: deepspeed\ndescription: \"Use DeepSpeed for distributed training, inference acceleration,\n  ZeRO configuration, parallelism\u002FMoE design, profiling, autotuning, and\n  operational diagnostics.\"\ndisable-model-invocation: true\nmetadata:\n  disco-role: operating\nlicense: Apache 2.0\n---\n\n# DeepSpeed Repo Skill\n\nUse this skill when a user asks to work with DeepSpeed APIs, configuration files, launch commands, inference acceleration, model parallelism, MoE, profiling, autotuning, or operational tooling. DeepSpeed is a PyTorch-focused optimization library with many optional GPU, distributed, and compiled-op paths, so route first by the user's workflow and then check the relevant troubleshooting reference before running heavy commands.\n\n## First Checks\n\n1. Confirm the user wants DeepSpeed itself, not a downstream framework integration that merely accepts a DeepSpeed config.\n2. Read `references\u002Finstall-and-environment.md` when installing, importing, or debugging optional CUDA\u002FROCm\u002Fvendor accelerator behavior.\n3. Run `scripts\u002Fcheck_deepspeed_env.py` for a read-only environment check, then `scripts\u002Finspect_deepspeed_api.py` when API fields or signatures may have drifted.\n4. Check `references\u002Frepo-provenance.md` before refreshing this skill against a source checkout.\n5. Do not run native examples, distributed tests, autotuning, NVMe\u002FGDS tools, builds, downloads, or remote SSH commands without explicit safety confirmation.\n\n## Route By Task\n\n- Use `sub-skills\u002Ftraining-config\u002F` for PyTorch training integration, `ds_config.json` authoring, launcher resource filters, ZeRO\u002Foffload choices, checkpoint save\u002Fload\u002Fexport, and config validation.\n- Use `sub-skills\u002Finference-injection\u002F` for `deepspeed.init_inference`, kernel or manual injection policies, inference tensor parallelism, inference quantization, checkpoint reshaping, v2\u002FFastGen routing, and hybrid-engine boundaries.\n- Use `sub-skills\u002Fparallelism-moe\u002F` for `PipelineModule`, pipeline schedules, MoE layers and optimizer groups, expert tensor parallelism, sequence parallel APIs, AutoSP, and activation checkpointing.\n- Use `sub-skills\u002Fops-tooling\u002F` for `ds_report`, install\u002Fbuild diagnostics, JIT\u002Fprebuilt op flags, autotuning, FLOPS profiling, monitor backends, DeepNVMe\u002FAIO\u002FGDS tools, compression APIs, and repo policy checks.\n\n## Shared References\n\n- `references\u002Finstall-and-environment.md`: package installation, PyTorch-first requirement, optional extras, compiled ops, and backend safety tiers.\n- `references\u002Ftroubleshooting.md`: cross-cutting import, accelerator, config, launcher, and optional-tool failure modes.\n- `references\u002Frepo-provenance.md`: source snapshot and refresh baseline.\n\n## Shared Scripts\n\n- `scripts\u002Fcheck_deepspeed_env.py`: read-only import, package metadata, PyTorch, CUDA visibility, and installed CLI discovery.\n- `scripts\u002Finspect_deepspeed_api.py`: read-only signature and config-field inspection for the main DeepSpeed APIs covered by this skill.\n\n## Safety Defaults\n\n- Prefer read-only commands first: imports, `--help`, config parsing, and tiny CPU checks.\n- Treat `deepspeed` launcher commands, checkpoint tests, CUDA op builds, autotuning, `ds_io`, `ds_nvme_tune`, GDS\u002FAIO code, and `ds_ssh` as real workload or infrastructure commands.\n- Do not promise GPU, compiled-op, NVMe, GDS, monitor, or model-download behavior unless the current environment has been explicitly verified for that path.\n- Keep generated examples self-contained. Do not tell future agents to open or execute original repo docs, tests, scripts, examples, or notebooks as runtime dependencies.\n","# DeepSpeed Repo Skill\n\nUse this skill when a user asks to work with DeepSpeed APIs, configuration files, launch commands, inference acceleration, model parallelism, MoE, profiling, autotuning, or operational tooling. DeepSpeed is a PyTorch-focused optimization library with many optional GPU, distributed, and compiled-op paths, so route first by the user's workflow and then check the relevant troubleshooting reference before running heavy commands.\n\n## First Checks\n\n1. Confirm the user wants DeepSpeed itself, not a downstream framework integration that merely accepts a DeepSpeed config.\n2. Read `references\u002Finstall-and-environment.md` when installing, importing, or debugging optional CUDA\u002FROCm\u002Fvendor accelerator behavior.\n3. Run `scripts\u002Fcheck_deepspeed_env.py` for a read-only environment check, then `scripts\u002Finspect_deepspeed_api.py` when API fields or signatures may have drifted.\n4. Check `references\u002Frepo-provenance.md` before refreshing this skill against a source checkout.\n5. Do not run native examples, distributed tests, autotuning, NVMe\u002FGDS tools, builds, downloads, or remote SSH commands without explicit safety confirmation.\n\n## Route By Task\n\n- Use `sub-skills\u002Ftraining-config\u002F` for PyTorch training integration, `ds_config.json` authoring, launcher resource filters, ZeRO\u002Foffload choices, checkpoint save\u002Fload\u002Fexport, and config validation.\n- Use `sub-skills\u002Finference-injection\u002F` for `deepspeed.init_inference`, kernel or manual injection policies, inference tensor parallelism, inference quantization, checkpoint reshaping, v2\u002FFastGen routing, and hybrid-engine boundaries.\n- Use `sub-skills\u002Fparallelism-moe\u002F` for `PipelineModule`, pipeline schedules, MoE layers and optimizer groups, expert tensor parallelism, sequence parallel APIs, AutoSP, and activation checkpointing.\n- Use `sub-skills\u002Fops-tooling\u002F` for `ds_report`, install\u002Fbuild diagnostics, JIT\u002Fprebuilt op flags, autotuning, FLOPS profiling, monitor backends, DeepNVMe\u002FAIO\u002FGDS tools, compression APIs, and repo policy checks.\n\n## Shared References\n\n- `references\u002Finstall-and-environment.md`: package installation, PyTorch-first requirement, optional extras, compiled ops, and backend safety tiers.\n- `references\u002Ftroubleshooting.md`: cross-cutting import, accelerator, config, launcher, and optional-tool failure modes.\n- `references\u002Frepo-provenance.md`: source snapshot and refresh baseline.\n\n## Shared Scripts\n\n- `scripts\u002Fcheck_deepspeed_env.py`: read-only import, package metadata, PyTorch, CUDA visibility, and installed CLI discovery.\n- `scripts\u002Finspect_deepspeed_api.py`: read-only signature and config-field inspection for the main DeepSpeed APIs covered by this skill.\n\n## Safety Defaults\n\n- Prefer read-only commands first: imports, `--help`, config parsing, and tiny CPU checks.\n- Treat `deepspeed` launcher commands, checkpoint tests, CUDA op builds, autotuning, `ds_io`, `ds_nvme_tune`, GDS\u002FAIO code, and `ds_ssh` as real workload or infrastructure commands.\n- Do not promise GPU, compiled-op, NVMe, GDS, monitor, or model-download behavior unless the current environment has been explicitly verified for that path.\n- Keep generated examples self-contained. Do not tell future agents to open or execute original repo docs, tests, scripts, examples, or notebooks as runtime dependencies.\n",{"name":5,"license":23,"metadata":24,"description":7,"disable-model-invocation":26},"Apache 2.0",{"disco-role":25},"operating",true,"main","ac3fe1afa80fb9a09775ecfb2b6cc3ba850a2db6","skills\u002Frepositories\u002Frepo-skills\u002Fdeepspeed",27,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill","https:\u002F\u002Fgithub.com\u002FVectorSpaceLab","2026-09-03 10:25:21",{"source":8,"skillName":5,"commands":35,"prompt":45,"zipAvailable":26},[36,39,42],{"tool":37,"command":38},"npx","npx skills add VectorSpaceLab\u002FAREX-Skill --skill deepspeed",{"tool":40,"command":41},"bunx","bunx skills add VectorSpaceLab\u002FAREX-Skill --skill deepspeed",{"tool":43,"command":44},"pnpm","pnpm dlx skills add VectorSpaceLab\u002FAREX-Skill --skill deepspeed","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fdeepspeed\u002FSKILL.md",[47,53,57,61,65,71],{"path":48,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":51,"sourceHtmlUrl":52},"references\u002Finstall-and-environment.md","REFERENCE","参考文档",3344,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fdeepspeed\u002Freferences\u002Finstall-and-environment.md",{"path":54,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":55,"sourceHtmlUrl":56},"references\u002Frepo-provenance.md",1808,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fdeepspeed\u002Freferences\u002Frepo-provenance.md",{"path":58,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":59,"sourceHtmlUrl":60},"references\u002Frepo-routing-metadata.json",542,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fdeepspeed\u002Freferences\u002Frepo-routing-metadata.json",{"path":62,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":63,"sourceHtmlUrl":64},"references\u002Ftroubleshooting.md",3279,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fdeepspeed\u002Freferences\u002Ftroubleshooting.md",{"path":66,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":69,"sourceHtmlUrl":70},"scripts\u002Fcheck_deepspeed_env.py","SCRIPT","脚本",3600,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fdeepspeed\u002Fscripts\u002Fcheck_deepspeed_env.py",{"path":72,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":73,"sourceHtmlUrl":74},"scripts\u002Finspect_deepspeed_api.py",3619,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fdeepspeed\u002Fscripts\u002Finspect_deepspeed_api.py",[76,83,90,97,104,111],{"skillNo":77,"name":78,"nameZh":6,"description":79,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":80,"categories":81,"keywords":82,"publishedAt":18},"01M799NYSH1R3RVAHSM2D88Y84","passnet-triton-opt","Make a single PassNet Triton kernel fast AND numerically faithful: performance model (when a replacement can win at all), block\u002Fgrid\u002Fwarp tuning, autotune policy, launch overhead, and per-op numeric recipes to pass the dtype baseline tolerances. Use when a pass matches and is correct but speedup below expected, or when correctness fails by small numeric margins.","https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Ftask-oriented\u002FPassNet\u002Fskills\u002Fpassnet-triton-opt\u002FSKILL.md",[],[],{"skillNo":84,"name":85,"nameZh":6,"description":86,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":87,"categories":88,"keywords":89,"publishedAt":18},"01M799NS5APEC0N1FZ7XZVW6BN","passnet-solve","END-TO-END playbook for solving one PassNet sample: analyze the computation graphs, decide the optimization strategy (what to fuse, what to replace, what to leave alone), drive the iteration loop, and maximize the sample score. This is the ENTRY skill — invoke it first for any PassNet optimization task; it tells you when to use passnet-pattern-fusion, passnet-triton-opt and passnet-feedback.","https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Ftask-oriented\u002FPassNet\u002Fskills\u002Fpassnet-solve\u002FSKILL.md",[],[],{"skillNo":91,"name":92,"nameZh":6,"description":93,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":94,"categories":95,"keywords":96,"publishedAt":18},"01M799NNY9BKSM21V5R7QVTBJX","passnet-skill","PassNet GPU kernel optimization via compiler passes. Design and implement Triton-based optimization passes, create pass files under .\u002Fpass_dir\u002F, self-evaluate with pass_evaluator, and iterate to maximize GPU speedup.","https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Ftask-oriented\u002FPassNet\u002Fskills\u002Fpassnet-skill\u002FSKILL.md",[],[],{"skillNo":98,"name":99,"nameZh":6,"description":100,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":101,"categories":102,"keywords":103,"publishedAt":18},"01M799NHH6NP7Z1RK77KGXAN6S","passnet-pattern-fusion","Author PassNet pass files: write patterns that actually MATCH the FX graph, pick fusion regions (which ops to absorb into one Triton kernel), and structure multi-pass submissions with the mandatory shared-dispatch architecture. Use when creating or fixing pattern()\u002Freplacement_args()\u002Freplacement_func() files, when a pass \"failed to match\", or when deciding how to fuse multiple kernels\u002Fops.","https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Ftask-oriented\u002FPassNet\u002Fskills\u002Fpassnet-pattern-fusion\u002FSKILL.md",[],[],{"skillNo":105,"name":106,"nameZh":6,"description":107,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":108,"categories":109,"keywords":110,"publishedAt":18},"01M799ND2J4WNC2PGAZ0A06NDF","passnet-orchestrate","Round\u002Fbatch planning and decision support for PassNet graph optimization. Use for multi-sample triage, worker allocation, eval-budget policy, and post-eval keep\u002Frevert decisions. For solving one concrete sample end to end, use passnet-solve as the entry skill; consult this skill only when you need broader planning or a specific decision gate.","https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Ftask-oriented\u002FPassNet\u002Fskills\u002Fpassnet-orchestrate\u002FSKILL.md",[],[],{"skillNo":112,"name":113,"nameZh":6,"description":114,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":115,"categories":116,"keywords":117,"publishedAt":18},"01M799N8SBTMY02ZDG2AZF5ZWE","passnet-feedback","Fast, deterministic feedback for PassNet work: pre-flight pattern\u002Fmatch verification WITHOUT burning a GPU evaluation, per-node bottleneck analysis of a sample's graphs, and eval-log parsing into per-variant status + estimated score + failure classification. Use BEFORE every GPU evaluation (check_pattern), at the START of a sample (analyze_graph), and AFTER every evaluation (parse_eval_log).","https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Ftask-oriented\u002FPassNet\u002Fskills\u002Fpassnet-feedback\u002FSKILL.md",[],[]]