[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M78V42C5F113AB46ZSFQJ5B4":3,"skill-same:01M78V42C5F113AB46ZSFQJ5B4":67},{"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},"01M78V42C5F113AB46ZSFQJ5B4","methods",null,"Use Composer speedup methods through algorithm classes, functional helpers, recipes, batch-key routing, and in-place model\u002Foptimizer mutations.","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\u002Fcomposer\u002Fsub-skills\u002Fmethods\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: methods\ndescription: \"Use Composer speedup methods through algorithm classes, functional\n  helpers, recipes, batch-key routing, and in-place model\u002Foptimizer mutations.\"\nmetadata:\n  disco-role: operating\ndisable-model-invocation: true\nlicense: Apache 2.0\n---\n\n# Composer Methods Router\n\nUse this sub-skill when the task is to add, compose, debug, or translate MosaicML Composer speedup methods from `composer.algorithms` and `composer.functional`.\n\n## Route first\n\n- For Trainer construction, dataloaders, `ComposerModel`, max duration, checkpointing, and optimizer basics, route to `..\u002Ftraining\u002FSKILL.md`.\n- For loggers, profiling, speed monitors, metric visualization, and throughput diagnosis, route to `..\u002Fobservability\u002FSKILL.md`.\n- For distributed launch, DDP\u002FFSDP setup, multi-node concerns, and backend placement, route to `..\u002Fdistributed\u002FSKILL.md`.\n- For TorchScript\u002FONNX\u002Fexport\u002Finference packaging after method application, route to `..\u002Finference-export\u002FSKILL.md`.\n\n## Main decision: algorithm class or functional API\n\nPrefer `composer.algorithms` classes in `Trainer(algorithms=[...])` when the user is already using Composer Trainer or wants a recipe of multiple methods. The Trainer event engine places each algorithm at the relevant event, passes `state.optimizers` to model-surgery algorithms, handles paired before\u002Fafter events, and lets batch algorithms use `input_key` \u002F `target_key` without changing the dataloader.\n\nUse `composer.functional` when the user owns a custom PyTorch loop, wants to test one method in isolation, or needs to mutate a model before creating the optimizer. With the functional API, the caller owns exact stage placement, target\u002Floss interpolation, optimizer rebinding after surgery, and any schedule logic.\n\n```python\nfrom composer import Trainer\nfrom composer.algorithms import BlurPool, ChannelsLast, LabelSmoothing, MixUp\n\nalgorithms = [\n    ChannelsLast(),\n    BlurPool(replace_convs=True, replace_maxpools=True, blur_first=True, min_channels=16),\n    MixUp(alpha=0.2, interpolate_loss=False, input_key=\"image\", target_key=\"label\"),\n    LabelSmoothing(smoothing=0.1, target_key=\"label\"),\n]\n\ntrainer = Trainer(\n    model=model,\n    train_dataloader=train_dataloader,\n    eval_dataloader=eval_dataloader,\n    optimizers=optimizer,\n    max_duration=\"10ep\",\n    algorithms=algorithms,\n)\n```\n\n## Placement rules\n\n1. Image transforms such as `RandAugment` or `AugMix` belong in dataset transforms before batching.\n2. Batch methods such as `MixUp`, `CutMix`, `ColOut`, `CutOut`, and `LabelSmoothing` run after a batch is available and before the forward\u002Floss stage that consumes the modified tensors.\n3. Model surgery such as `BlurPool`, `SqueezeExcite`, `Factorize`, `StochasticDepth`, `Alibi`, low-precision norm replacements, and `ChannelsLast` mutates the model in place. Apply it before optimizer construction, or pass existing optimizers to functional helpers that accept `optimizers=`. Trainer algorithms do this through `state.optimizers`.\n4. Loop and optimizer methods such as `LayerFreezing`, `ProgressiveResizing`, `SelectiveBackprop`, `SeqLengthWarmup`, `EMA`, `SWA`, `SAM`, and `GradientClipping` depend on event timing or optimizer state. Use algorithm classes unless the user is explicitly maintaining those stages in a custom loop.\n\n## Batch key routing\n\nBatch algorithms default to tuple batches: `input_key=0`, `target_key=1`. For dict batches, pass string keys instead of rewriting the dataloader:\n\n```python\nfrom composer.algorithms import CutMix, MixUp, LabelSmoothing\n\nalgorithms = [\n    MixUp(alpha=0.2, input_key=\"image\", target_key=\"label\"),\n    CutMix(alpha=1.0, uniform_sampling=False, input_key=\"image\", target_key=\"label\"),\n    LabelSmoothing(smoothing=0.1, target_key=\"label\"),\n]\n```\n\nFor nested or custom batch objects, pass `(getter, setter)` callables. The getter receives the batch and returns the tensor; the setter receives `(batch, value)` and returns the updated batch.\n\n## References and bundled smoke\n\n- `references\u002Ffunctional-api.md`: function families, stage contracts, batch-key semantics, and custom-loop recipes.\n- `references\u002Falgorithm-catalog.md`: exported Composer method list grouped by taxonomy.\n- `references\u002Ftroubleshooting.md`: method-specific failure modes and fixes.\n- `scripts\u002Ffunctional_smoke.py`: no-download random tensor smoke for a batch method plus functional model surgery.\n\nRun the smoke from this sub-skill directory with:\n\n```bash\npython scripts\u002Ffunctional_smoke.py\n```\n","# Composer Methods Router\n\nUse this sub-skill when the task is to add, compose, debug, or translate MosaicML Composer speedup methods from `composer.algorithms` and `composer.functional`.\n\n## Route first\n\n- For Trainer construction, dataloaders, `ComposerModel`, max duration, checkpointing, and optimizer basics, route to `..\u002Ftraining\u002FSKILL.md`.\n- For loggers, profiling, speed monitors, metric visualization, and throughput diagnosis, route to `..\u002Fobservability\u002FSKILL.md`.\n- For distributed launch, DDP\u002FFSDP setup, multi-node concerns, and backend placement, route to `..\u002Fdistributed\u002FSKILL.md`.\n- For TorchScript\u002FONNX\u002Fexport\u002Finference packaging after method application, route to `..\u002Finference-export\u002FSKILL.md`.\n\n## Main decision: algorithm class or functional API\n\nPrefer `composer.algorithms` classes in `Trainer(algorithms=[...])` when the user is already using Composer Trainer or wants a recipe of multiple methods. The Trainer event engine places each algorithm at the relevant event, passes `state.optimizers` to model-surgery algorithms, handles paired before\u002Fafter events, and lets batch algorithms use `input_key` \u002F `target_key` without changing the dataloader.\n\nUse `composer.functional` when the user owns a custom PyTorch loop, wants to test one method in isolation, or needs to mutate a model before creating the optimizer. With the functional API, the caller owns exact stage placement, target\u002Floss interpolation, optimizer rebinding after surgery, and any schedule logic.\n\n```python\nfrom composer import Trainer\nfrom composer.algorithms import BlurPool, ChannelsLast, LabelSmoothing, MixUp\n\nalgorithms = [\n    ChannelsLast(),\n    BlurPool(replace_convs=True, replace_maxpools=True, blur_first=True, min_channels=16),\n    MixUp(alpha=0.2, interpolate_loss=False, input_key=\"image\", target_key=\"label\"),\n    LabelSmoothing(smoothing=0.1, target_key=\"label\"),\n]\n\ntrainer = Trainer(\n    model=model,\n    train_dataloader=train_dataloader,\n    eval_dataloader=eval_dataloader,\n    optimizers=optimizer,\n    max_duration=\"10ep\",\n    algorithms=algorithms,\n)\n```\n\n## Placement rules\n\n1. Image transforms such as `RandAugment` or `AugMix` belong in dataset transforms before batching.\n2. Batch methods such as `MixUp`, `CutMix`, `ColOut`, `CutOut`, and `LabelSmoothing` run after a batch is available and before the forward\u002Floss stage that consumes the modified tensors.\n3. Model surgery such as `BlurPool`, `SqueezeExcite`, `Factorize`, `StochasticDepth`, `Alibi`, low-precision norm replacements, and `ChannelsLast` mutates the model in place. Apply it before optimizer construction, or pass existing optimizers to functional helpers that accept `optimizers=`. Trainer algorithms do this through `state.optimizers`.\n4. Loop and optimizer methods such as `LayerFreezing`, `ProgressiveResizing`, `SelectiveBackprop`, `SeqLengthWarmup`, `EMA`, `SWA`, `SAM`, and `GradientClipping` depend on event timing or optimizer state. Use algorithm classes unless the user is explicitly maintaining those stages in a custom loop.\n\n## Batch key routing\n\nBatch algorithms default to tuple batches: `input_key=0`, `target_key=1`. For dict batches, pass string keys instead of rewriting the dataloader:\n\n```python\nfrom composer.algorithms import CutMix, MixUp, LabelSmoothing\n\nalgorithms = [\n    MixUp(alpha=0.2, input_key=\"image\", target_key=\"label\"),\n    CutMix(alpha=1.0, uniform_sampling=False, input_key=\"image\", target_key=\"label\"),\n    LabelSmoothing(smoothing=0.1, target_key=\"label\"),\n]\n```\n\nFor nested or custom batch objects, pass `(getter, setter)` callables. The getter receives the batch and returns the tensor; the setter receives `(batch, value)` and returns the updated batch.\n\n## References and bundled smoke\n\n- `references\u002Ffunctional-api.md`: function families, stage contracts, batch-key semantics, and custom-loop recipes.\n- `references\u002Falgorithm-catalog.md`: exported Composer method list grouped by taxonomy.\n- `references\u002Ftroubleshooting.md`: method-specific failure modes and fixes.\n- `scripts\u002Ffunctional_smoke.py`: no-download random tensor smoke for a batch method plus functional model surgery.\n\nRun the smoke from this sub-skill directory with:\n\n```bash\npython scripts\u002Ffunctional_smoke.py\n```\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\u002Fcomposer\u002Fsub-skills\u002Fmethods",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 methods",{"tool":40,"command":41},"bunx","bunx skills add VectorSpaceLab\u002FAREX-Skill --skill methods",{"tool":43,"command":44},"pnpm","pnpm dlx skills add VectorSpaceLab\u002FAREX-Skill --skill methods","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fcomposer\u002Fsub-skills\u002Fmethods\u002FSKILL.md",[47,53,57,61],{"path":48,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":51,"sourceHtmlUrl":52},"references\u002Falgorithm-catalog.md","REFERENCE","参考文档",11274,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fcomposer\u002Fsub-skills\u002Fmethods\u002Freferences\u002Falgorithm-catalog.md",{"path":54,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":55,"sourceHtmlUrl":56},"references\u002Ffunctional-api.md",10530,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fcomposer\u002Fsub-skills\u002Fmethods\u002Freferences\u002Ffunctional-api.md",{"path":58,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":59,"sourceHtmlUrl":60},"references\u002Ftroubleshooting.md",8416,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fcomposer\u002Fsub-skills\u002Fmethods\u002Freferences\u002Ftroubleshooting.md",{"path":62,"resourceType":63,"resourceTypeDesc":64,"sizeBytes":65,"sourceHtmlUrl":66},"scripts\u002Ffunctional_smoke.py","SCRIPT","脚本",4068,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fcomposer\u002Fsub-skills\u002Fmethods\u002Fscripts\u002Ffunctional_smoke.py",[68,75,82,89,96,103],{"skillNo":69,"name":70,"nameZh":6,"description":71,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":72,"categories":73,"keywords":74,"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":76,"name":77,"nameZh":6,"description":78,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":79,"categories":80,"keywords":81,"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":83,"name":84,"nameZh":6,"description":85,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":86,"categories":87,"keywords":88,"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":90,"name":91,"nameZh":6,"description":92,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":93,"categories":94,"keywords":95,"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":97,"name":98,"nameZh":6,"description":99,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":100,"categories":101,"keywords":102,"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":104,"name":105,"nameZh":6,"description":106,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":107,"categories":108,"keywords":109,"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",[],[]]