[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M790MEV6PT65BJTC1ASCWR3M":3,"skill-same:01M790MEV6PT65BJTC1ASCWR3M":57},{"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},"01M790MEV6PT65BJTC1ASCWR3M","training-and-distribution",null,"Guides MMGeneration training, resume, distributed launch, and launcher-specific workflows.","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\u002Fmmgeneration\u002Fsub-skills\u002Ftraining-and-distribution\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: \"training-and-distribution\"\ndescription: \"Guides MMGeneration training, resume, distributed launch, and\n  launcher-specific workflows.\"\ndisable-model-invocation: true\nmetadata:\n  disco-role: operating\nlicense: Apache 2.0\n---\n\n# Training and Distribution\n\nUse this sub-skill when the user wants to train a model, resume from a checkpoint, or understand the repo's distributed training and validation flow.\n\n## Typical triggers\n\n- \"How do I train this config?\"\n- \"How do I resume a run?\"\n- \"How do I use DDP or Slurm?\"\n- \"Can I train on CPU for debugging?\"\n- \"Why does the training hook not behave as expected?\"\n\n## Include here\n\n- `tools\u002Ftrain.py`\n- `tools\u002Fdist_train.sh`\n- `tools\u002Fslurm_train.sh`\n- `mmgen.apis.train_model` and `set_random_seed`\n- `DynamicIterBasedRunner`\n- DDP wrapper and dynamic DDP notes\n- EMA, visualization, checkpoint, and validation hook behavior\n- CPU training guidance for debug-only use\n\n## Exclude here\n\n- Sampling and demo workflows -> `inference-and-sampling`\n- Metrics and validation scripts -> `evaluation-and-metrics`\n- Config syntax and registry extension -> `configuration-and-extension`\n- Latent editing and deployment helpers -> `applications-and-deployment`\n\n## Read these files first\n\n- `references\u002Fworkflows.md`\n- `references\u002Ftroubleshooting.md`\n- `..\u002F..\u002Freferences\u002Fapi-reference.md`\n- `..\u002F..\u002Freferences\u002Fcli-reference.md`\n- `..\u002F..\u002Freferences\u002Fmodel-overview.md`\n\n## What good guidance looks like\n\nA future agent should be able to:\n\n1. Pick the right launcher: none, PyTorch DDP, or Slurm.\n2. Explain how `work_dir`, `resume_from`, and `load_from` interact.\n3. Distinguish static and dynamic GAN behavior in distributed mode.\n4. Tell the user when CPU training is acceptable only as a debugging aid.\n5. Route validation and evaluation questions to the evaluation sub-skill instead of mixing them into training advice.\n\n## Common failure modes\n\n- DDP settings that work for static GANs but fail for dynamic architectures.\n- `apex_amp` being used outside the distributed-only path.\n- Missing or incorrect `work_dir`, `cfg_options`, or checkpoint resume path.\n- A custom hook or runtime block not being registered before the run starts.\n\nFor concrete recovery steps, read `references\u002Ftroubleshooting.md`.\n","# Training and Distribution\n\nUse this sub-skill when the user wants to train a model, resume from a checkpoint, or understand the repo's distributed training and validation flow.\n\n## Typical triggers\n\n- \"How do I train this config?\"\n- \"How do I resume a run?\"\n- \"How do I use DDP or Slurm?\"\n- \"Can I train on CPU for debugging?\"\n- \"Why does the training hook not behave as expected?\"\n\n## Include here\n\n- `tools\u002Ftrain.py`\n- `tools\u002Fdist_train.sh`\n- `tools\u002Fslurm_train.sh`\n- `mmgen.apis.train_model` and `set_random_seed`\n- `DynamicIterBasedRunner`\n- DDP wrapper and dynamic DDP notes\n- EMA, visualization, checkpoint, and validation hook behavior\n- CPU training guidance for debug-only use\n\n## Exclude here\n\n- Sampling and demo workflows -> `inference-and-sampling`\n- Metrics and validation scripts -> `evaluation-and-metrics`\n- Config syntax and registry extension -> `configuration-and-extension`\n- Latent editing and deployment helpers -> `applications-and-deployment`\n\n## Read these files first\n\n- `references\u002Fworkflows.md`\n- `references\u002Ftroubleshooting.md`\n- `..\u002F..\u002Freferences\u002Fapi-reference.md`\n- `..\u002F..\u002Freferences\u002Fcli-reference.md`\n- `..\u002F..\u002Freferences\u002Fmodel-overview.md`\n\n## What good guidance looks like\n\nA future agent should be able to:\n\n1. Pick the right launcher: none, PyTorch DDP, or Slurm.\n2. Explain how `work_dir`, `resume_from`, and `load_from` interact.\n3. Distinguish static and dynamic GAN behavior in distributed mode.\n4. Tell the user when CPU training is acceptable only as a debugging aid.\n5. Route validation and evaluation questions to the evaluation sub-skill instead of mixing them into training advice.\n\n## Common failure modes\n\n- DDP settings that work for static GANs but fail for dynamic architectures.\n- `apex_amp` being used outside the distributed-only path.\n- Missing or incorrect `work_dir`, `cfg_options`, or checkpoint resume path.\n- A custom hook or runtime block not being registered before the run starts.\n\nFor concrete recovery steps, read `references\u002Ftroubleshooting.md`.\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\u002Fmmgeneration\u002Fsub-skills\u002Ftraining-and-distribution",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-and-distribution",{"tool":40,"command":41},"bunx","bunx skills add VectorSpaceLab\u002FAREX-Skill --skill training-and-distribution",{"tool":43,"command":44},"pnpm","pnpm dlx skills add VectorSpaceLab\u002FAREX-Skill --skill training-and-distribution","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fmmgeneration\u002Fsub-skills\u002Ftraining-and-distribution\u002FSKILL.md",[47,53],{"path":48,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":51,"sourceHtmlUrl":52},"references\u002Ftroubleshooting.md","REFERENCE","参考文档",2877,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fmmgeneration\u002Fsub-skills\u002Ftraining-and-distribution\u002Freferences\u002Ftroubleshooting.md",{"path":54,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":55,"sourceHtmlUrl":56},"references\u002Fworkflows.md",2504,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fmmgeneration\u002Fsub-skills\u002Ftraining-and-distribution\u002Freferences\u002Fworkflows.md",[58,65,72,79,86,93],{"skillNo":59,"name":60,"nameZh":6,"description":61,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":62,"categories":63,"keywords":64,"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":66,"name":67,"nameZh":6,"description":68,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":69,"categories":70,"keywords":71,"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":73,"name":74,"nameZh":6,"description":75,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":76,"categories":77,"keywords":78,"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":80,"name":81,"nameZh":6,"description":82,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":83,"categories":84,"keywords":85,"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":87,"name":88,"nameZh":6,"description":89,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":90,"categories":91,"keywords":92,"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":94,"name":95,"nameZh":6,"description":96,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":97,"categories":98,"keywords":99,"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",[],[]]