[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M796EZ9A5AMNXBPQ2HPR93F9":3,"skill-same:01M796EZ9A5AMNXBPQ2HPR93F9":74},{"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":26,"commitSha":27,"skillDirectory":28,"forksCount":29,"repositoryHtmlUrl":30,"repositoryHomepageUrl":6,"ownerHtmlUrl":31,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":32,"install":33,"resources":45},"01M796EZ9A5AMNXBPQ2HPR93F9","llm-vla-and-services",null,"Use TorchRL LLM post-training, VLA data\u002Factions, services, rendering, and weight-update surfaces safely.","VectorSpaceLab\u002FAREX-Skill","VectorSpaceLab","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F180637102?v=4","01M7538C5WPSJGWF7KHR4EFZ9F","01M7538C6R2GFP53NPXTCZNBRT",351,"MIT","https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Ftorchrl\u002Fsub-skills\u002Fllm-vla-and-services\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: llm-vla-and-services\ndescription: \"Use TorchRL LLM post-training, VLA data\u002Factions, services,\n  rendering, and weight-update surfaces safely.\"\ndisable-model-invocation: true\nmetadata:\n  disco-role: operating\nlicense: MIT\n---\n\n# TorchRL LLM, VLA, Services, and Rendering\n\nUse this sub-skill when a task mentions TorchRL LLM post-training, RLHF, GRPO,\nSFT, chat environments, LLM inference wrappers, tool execution services,\nVision-Language-Action data, action tokenizers, service registry\u002Flifecycle,\nrendering, video outputs, or LLM\u002FVLA weight synchronization.\n\n## Route by task\n\n- For `ChatEnv`, `History`, `PromptData`, reward data, `LLMCollector`,\n  `TransformersWrapper`, `vLLMWrapper`, `SGLangWrapper`, GRPO, SFT,\n  distillation, policy version tracking, or vLLM\u002FSGLang weight sync, read\n  [LLM and RLHF workflows](references\u002Fllm-and-rlhf-workflows.md).\n- For `validate_vla_tensordict`, VLA canonical TensorDict keys,\n  `RobotDatasetMetadata`, action chunking, `UniformActionTokenizer`,\n  `VocabTailActionTokenizer`, `ActionScaling`, `TinyVLA`, LeRobot, or OpenX,\n  read [VLA data and actions](references\u002Fvla-data-and-actions.md) and run\n  [check_vla_schema.py](scripts\u002Fcheck_vla_schema.py) for a CPU-safe schema\n  smoke test.\n- For `Service`, `get_services`, owner\u002Fclient lifecycles, direct\u002Fprocess\u002FRay\n  placement, `PythonExecutorService`, `PythonInterpreter`, `rlrender`, render\n  checkpoint loading, or video output, read\n  [services and rendering](references\u002Fservices-and-rendering.md) and run\n  [smoke_services.py](scripts\u002Fsmoke_services.py) for import\u002Flifecycle checks.\n- For optional extras, model downloads, GPU serving memory, tokenizer\u002Ftemplate\n  mismatch, policy-version drift, Ray cleanup, VLA shape\u002Fkey failures, render\n  codecs, or display issues, read [troubleshooting](references\u002Ftroubleshooting.md).\n\n## Boundaries\n\nThis sub-skill owns TorchRL's LLM\u002FVLA\u002Fservice\u002Frender-specific surfaces and the\nsafe checks bundled here. Route generic collector topology, replay buffers, and\nweight updaters that are not LLM-specific to the collectors-and-replay area.\nRoute generic actor\u002Fcritic\u002Fkey wiring to modules-and-policies, and generic loss\nmodule or trainer workflows to objectives-and-training.\n\nOptional GPU serving backends, downloaded model checkpoints, Ray clusters,\nrobot datasets, and codec\u002Fdisplay stacks are reference-only here unless the user\nexplicitly provisions those dependencies and hardware.\n","# TorchRL LLM, VLA, Services, and Rendering\n\nUse this sub-skill when a task mentions TorchRL LLM post-training, RLHF, GRPO,\nSFT, chat environments, LLM inference wrappers, tool execution services,\nVision-Language-Action data, action tokenizers, service registry\u002Flifecycle,\nrendering, video outputs, or LLM\u002FVLA weight synchronization.\n\n## Route by task\n\n- For `ChatEnv`, `History`, `PromptData`, reward data, `LLMCollector`,\n  `TransformersWrapper`, `vLLMWrapper`, `SGLangWrapper`, GRPO, SFT,\n  distillation, policy version tracking, or vLLM\u002FSGLang weight sync, read\n  [LLM and RLHF workflows](references\u002Fllm-and-rlhf-workflows.md).\n- For `validate_vla_tensordict`, VLA canonical TensorDict keys,\n  `RobotDatasetMetadata`, action chunking, `UniformActionTokenizer`,\n  `VocabTailActionTokenizer`, `ActionScaling`, `TinyVLA`, LeRobot, or OpenX,\n  read [VLA data and actions](references\u002Fvla-data-and-actions.md) and run\n  [check_vla_schema.py](scripts\u002Fcheck_vla_schema.py) for a CPU-safe schema\n  smoke test.\n- For `Service`, `get_services`, owner\u002Fclient lifecycles, direct\u002Fprocess\u002FRay\n  placement, `PythonExecutorService`, `PythonInterpreter`, `rlrender`, render\n  checkpoint loading, or video output, read\n  [services and rendering](references\u002Fservices-and-rendering.md) and run\n  [smoke_services.py](scripts\u002Fsmoke_services.py) for import\u002Flifecycle checks.\n- For optional extras, model downloads, GPU serving memory, tokenizer\u002Ftemplate\n  mismatch, policy-version drift, Ray cleanup, VLA shape\u002Fkey failures, render\n  codecs, or display issues, read [troubleshooting](references\u002Ftroubleshooting.md).\n\n## Boundaries\n\nThis sub-skill owns TorchRL's LLM\u002FVLA\u002Fservice\u002Frender-specific surfaces and the\nsafe checks bundled here. Route generic collector topology, replay buffers, and\nweight updaters that are not LLM-specific to the collectors-and-replay area.\nRoute generic actor\u002Fcritic\u002Fkey wiring to modules-and-policies, and generic loss\nmodule or trainer workflows to objectives-and-training.\n\nOptional GPU serving backends, downloaded model checkpoints, Ray clusters,\nrobot datasets, and codec\u002Fdisplay stacks are reference-only here unless the user\nexplicitly provisions those dependencies and hardware.\n",{"name":5,"license":14,"metadata":23,"description":7,"disable-model-invocation":25},{"disco-role":24},"operating",true,"main","ac3fe1afa80fb9a09775ecfb2b6cc3ba850a2db6","skills\u002Frepositories\u002Frepo-skills\u002Ftorchrl\u002Fsub-skills\u002Fllm-vla-and-services",27,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill","https:\u002F\u002Fgithub.com\u002FVectorSpaceLab","2026-09-03 10:25:21",{"source":8,"skillName":5,"commands":34,"prompt":44,"zipAvailable":25},[35,38,41],{"tool":36,"command":37},"npx","npx skills add VectorSpaceLab\u002FAREX-Skill --skill llm-vla-and-services",{"tool":39,"command":40},"bunx","bunx skills add VectorSpaceLab\u002FAREX-Skill --skill llm-vla-and-services",{"tool":42,"command":43},"pnpm","pnpm dlx skills add VectorSpaceLab\u002FAREX-Skill --skill llm-vla-and-services","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Ftorchrl\u002Fsub-skills\u002Fllm-vla-and-services\u002FSKILL.md",[46,52,56,60,64,70],{"path":47,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":50,"sourceHtmlUrl":51},"references\u002Fllm-and-rlhf-workflows.md","REFERENCE","参考文档",10306,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Ftorchrl\u002Fsub-skills\u002Fllm-vla-and-services\u002Freferences\u002Fllm-and-rlhf-workflows.md",{"path":53,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":54,"sourceHtmlUrl":55},"references\u002Fservices-and-rendering.md",7431,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Ftorchrl\u002Fsub-skills\u002Fllm-vla-and-services\u002Freferences\u002Fservices-and-rendering.md",{"path":57,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":58,"sourceHtmlUrl":59},"references\u002Ftroubleshooting.md",9546,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Ftorchrl\u002Fsub-skills\u002Fllm-vla-and-services\u002Freferences\u002Ftroubleshooting.md",{"path":61,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":62,"sourceHtmlUrl":63},"references\u002Fvla-data-and-actions.md",8879,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Ftorchrl\u002Fsub-skills\u002Fllm-vla-and-services\u002Freferences\u002Fvla-data-and-actions.md",{"path":65,"resourceType":66,"resourceTypeDesc":67,"sizeBytes":68,"sourceHtmlUrl":69},"scripts\u002Fcheck_vla_schema.py","SCRIPT","脚本",6320,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Ftorchrl\u002Fsub-skills\u002Fllm-vla-and-services\u002Fscripts\u002Fcheck_vla_schema.py",{"path":71,"resourceType":66,"resourceTypeDesc":67,"sizeBytes":72,"sourceHtmlUrl":73},"scripts\u002Fsmoke_services.py",6358,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Ftorchrl\u002Fsub-skills\u002Fllm-vla-and-services\u002Fscripts\u002Fsmoke_services.py",[75,83,90,97,104,111],{"skillNo":76,"name":77,"nameZh":6,"description":78,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":79,"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.","Apache-2.0","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":79,"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":79,"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":79,"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":79,"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":79,"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",[],[]]