[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M78VXPYVBDQTPJB1KPNBMYK4":3,"skill-same:01M78VXPYVBDQTPJB1KPNBMYK4":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},"01M78VXPYVBDQTPJB1KPNBMYK4","training-config",null,"Integrate DeepSpeed training into PyTorch, author and validate config JSON\u002FHJSON, plan launcher resources, choose ZeRO\u002Foffload settings, and handle checkpoint save\u002Fload\u002Fexport safely.","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\u002Fsub-skills\u002Ftraining-config\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: training-config\ndescription: \"Integrate DeepSpeed training into PyTorch, author and validate\n  config JSON\u002FHJSON, plan launcher resources, choose ZeRO\u002Foffload settings, and\n  handle checkpoint save\u002Fload\u002Fexport safely.\"\ndisable-model-invocation: true\nmetadata:\n  disco-role: operating\nlicense: Apache 2.0\n---\n\n# DeepSpeed Training Config\n\nUse this sub-skill when a future agent needs to wire DeepSpeed into a PyTorch training loop, create or audit `ds_config.json`, build `deepspeed` launcher commands, select ZeRO stages\u002Foffload, or plan training checkpoint save\u002Fload\u002Fexport. This sub-skill covers training runtime setup only; route inference-only kernel injection to `inference-injection`, pipeline\u002FMoE\u002Fsequence parallel details to `parallelism-moe`, and build diagnostics to `ops-tooling`.\n\n## Fast Path\n\n1. Integrate the model with `deepspeed.initialize(...)`; call `deepspeed.init_distributed()` only when the script needs a distributed process group before `initialize`.\n2. Author config through `references\u002Fconfiguration.md`, then run `scripts\u002Fvalidate_ds_config.py --world-size \u003Cranks> ds_config.json` before launching.\n3. Build resource filters and `--no_ssh` commands through `references\u002Flauncher-and-checkpointing.md`; preview filters with `scripts\u002Flauncher_resource_preview.py`.\n4. Choose ZeRO settings with `references\u002Fconfiguration.md`, keeping deprecated aliases out of new configs.\n5. Plan checkpoint calls so all ranks participate; for ZeRO-3 exports, enable `stage3_gather_16bit_weights_on_model_save` or use the generated `zero_to_fp32.py`\u002Futility path after checkpointing.\n\n## Bundled References\n\n- `references\u002Fapi-reference.md`: `deepspeed.initialize`, `init_distributed`, engine training calls, `DeepSpeedConfig`, and `DeepSpeedZeroConfig` field facts.\n- `references\u002Fconfiguration.md`: batch-size contract, JSON\u002FHJSON authoring, precision, optimizer\u002Fscheduler, ZeRO\u002Foffload choices, and safe validation workflow.\n- `references\u002Flauncher-and-checkpointing.md`: launcher hostfile\u002Finclude\u002Fexclude\u002Fno-SSH syntax plus checkpoint save\u002Fload\u002Fexport guardrails.\n- `references\u002Ftroubleshooting.md`: common failures and fixes for config parsing, launcher filters, checkpoint hangs, ZeRO-3 export, and CUDA op probing.\n\n## Bundled Scripts\n\n- `scripts\u002Fvalidate_ds_config.py`: Parses JSON or HJSON when `hjson` is installed, detects duplicate JSON keys, checks batch-size consistency, and warns about selected ZeRO pitfalls without importing or launching DeepSpeed.\n- `scripts\u002Flauncher_resource_preview.py`: Parses hostfile text plus `--include`\u002F`--exclude` style filters and reports the selected resources or conflicts without launching training.\n\n## Native Candidates\n\nSafe native candidates to consult during verification are the config parser\u002Fdict tests and launcher parser tests. Broad GPU\u002Fdistributed checkpoint matrices should be skipped by default unless the verifier has explicit safe hardware and time approval.\n","# DeepSpeed Training Config\n\nUse this sub-skill when a future agent needs to wire DeepSpeed into a PyTorch training loop, create or audit `ds_config.json`, build `deepspeed` launcher commands, select ZeRO stages\u002Foffload, or plan training checkpoint save\u002Fload\u002Fexport. This sub-skill covers training runtime setup only; route inference-only kernel injection to `inference-injection`, pipeline\u002FMoE\u002Fsequence parallel details to `parallelism-moe`, and build diagnostics to `ops-tooling`.\n\n## Fast Path\n\n1. Integrate the model with `deepspeed.initialize(...)`; call `deepspeed.init_distributed()` only when the script needs a distributed process group before `initialize`.\n2. Author config through `references\u002Fconfiguration.md`, then run `scripts\u002Fvalidate_ds_config.py --world-size \u003Cranks> ds_config.json` before launching.\n3. Build resource filters and `--no_ssh` commands through `references\u002Flauncher-and-checkpointing.md`; preview filters with `scripts\u002Flauncher_resource_preview.py`.\n4. Choose ZeRO settings with `references\u002Fconfiguration.md`, keeping deprecated aliases out of new configs.\n5. Plan checkpoint calls so all ranks participate; for ZeRO-3 exports, enable `stage3_gather_16bit_weights_on_model_save` or use the generated `zero_to_fp32.py`\u002Futility path after checkpointing.\n\n## Bundled References\n\n- `references\u002Fapi-reference.md`: `deepspeed.initialize`, `init_distributed`, engine training calls, `DeepSpeedConfig`, and `DeepSpeedZeroConfig` field facts.\n- `references\u002Fconfiguration.md`: batch-size contract, JSON\u002FHJSON authoring, precision, optimizer\u002Fscheduler, ZeRO\u002Foffload choices, and safe validation workflow.\n- `references\u002Flauncher-and-checkpointing.md`: launcher hostfile\u002Finclude\u002Fexclude\u002Fno-SSH syntax plus checkpoint save\u002Fload\u002Fexport guardrails.\n- `references\u002Ftroubleshooting.md`: common failures and fixes for config parsing, launcher filters, checkpoint hangs, ZeRO-3 export, and CUDA op probing.\n\n## Bundled Scripts\n\n- `scripts\u002Fvalidate_ds_config.py`: Parses JSON or HJSON when `hjson` is installed, detects duplicate JSON keys, checks batch-size consistency, and warns about selected ZeRO pitfalls without importing or launching DeepSpeed.\n- `scripts\u002Flauncher_resource_preview.py`: Parses hostfile text plus `--include`\u002F`--exclude` style filters and reports the selected resources or conflicts without launching training.\n\n## Native Candidates\n\nSafe native candidates to consult during verification are the config parser\u002Fdict tests and launcher parser tests. Broad GPU\u002Fdistributed checkpoint matrices should be skipped by default unless the verifier has explicit safe hardware and time approval.\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\u002Fsub-skills\u002Ftraining-config",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 training-config",{"tool":40,"command":41},"bunx","bunx skills add VectorSpaceLab\u002FAREX-Skill --skill training-config",{"tool":43,"command":44},"pnpm","pnpm dlx skills add VectorSpaceLab\u002FAREX-Skill --skill training-config","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fdeepspeed\u002Fsub-skills\u002Ftraining-config\u002FSKILL.md",[47,53,57,61,65,71],{"path":48,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":51,"sourceHtmlUrl":52},"references\u002Fapi-reference.md","REFERENCE","参考文档",3699,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fdeepspeed\u002Fsub-skills\u002Ftraining-config\u002Freferences\u002Fapi-reference.md",{"path":54,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":55,"sourceHtmlUrl":56},"references\u002Fconfiguration.md",3970,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fdeepspeed\u002Fsub-skills\u002Ftraining-config\u002Freferences\u002Fconfiguration.md",{"path":58,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":59,"sourceHtmlUrl":60},"references\u002Flauncher-and-checkpointing.md",4606,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fdeepspeed\u002Fsub-skills\u002Ftraining-config\u002Freferences\u002Flauncher-and-checkpointing.md",{"path":62,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":63,"sourceHtmlUrl":64},"references\u002Ftroubleshooting.md",4122,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fdeepspeed\u002Fsub-skills\u002Ftraining-config\u002Freferences\u002Ftroubleshooting.md",{"path":66,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":69,"sourceHtmlUrl":70},"scripts\u002Flauncher_resource_preview.py","SCRIPT","脚本",5846,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fdeepspeed\u002Fsub-skills\u002Ftraining-config\u002Fscripts\u002Flauncher_resource_preview.py",{"path":72,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":73,"sourceHtmlUrl":74},"scripts\u002Fvalidate_ds_config.py",7969,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fdeepspeed\u002Fsub-skills\u002Ftraining-config\u002Fscripts\u002Fvalidate_ds_config.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",[],[]]