[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M78SW1G5ZKSENPQ6PAGCQ37P":3,"skill-same:01M78SW1G5ZKSENPQ6PAGCQ37P":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},"01M78SW1G5ZKSENPQ6PAGCQ37P","classification",null,"Guides Attention-Gated Networks ultrasound classification workflows, including Sononet training, evaluation, HDF5 data layout, and attention overlays.","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\u002Fattention-gated-networks\u002Fsub-skills\u002Fclassification\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: classification\ndescription: \"Guides Attention-Gated Networks ultrasound classification\n  workflows, including Sononet training, evaluation, HDF5 data layout, and\n  attention overlays.\"\ndisable-model-invocation: true\nmetadata:\n  disco-role: operating\nlicense: MIT\n---\n\n# Classification\n\nUse this sub-skill for ultrasound scan-plane classification tasks in\nAttention-Gated Networks. It covers Sononet, Sononet2, Sononet Grid Attention,\n`FeedForwardClassifier`, `AggregatedClassifier`, the HDF5 ultrasound loader,\nclassification training\u002Ftesting scripts, and attention overlay export.\n\n## Read first\n\n- Read [api-reference.md](references\u002Fapi-reference.md) when you need exact\n  model names, wrapper behavior, config fields, aggregation modes, or class and\n  function signatures.\n- Read [data-layout.md](references\u002Fdata-layout.md) before preparing the\n  ultrasound HDF5 file or debugging loader\u002Fsampler failures.\n- Read [workflows.md](references\u002Fworkflows.md) for training, testing,\n  checkpoint evaluation, model inspection, and safe attention-overlay export.\n- Read [troubleshooting.md](references\u002Ftroubleshooting.md) when CUDA,\n  torchsample, Visdom, sampler, HDF5, or overlay failures appear.\n- Run [run_classifier.py](scripts\u002Frun_classifier.py) for skill-owned training\n  and testing commands that replace the source repo's root entry points.\n- Run [export_attention_overlay.py](scripts\u002Fexport_attention_overlay.py) to\n  replace the source repo's hard-coded attention visualization script with a\n  parameterized, skill-owned helper.\n- For fast environment validation, run the root helper\n  `..\u002F..\u002Fscripts\u002Fcheck_env.py --repo-root \u002Fpath\u002Fto\u002FAttention-Gated-Networks --mode classification`.\n\n## When to use this sub-skill\n\nUse this route when the user asks to:\n\n- train or test ultrasound scan-plane classification models;\n- choose between `sononet`, `sononet2`, and `sononet_grid_attention`;\n- configure `aggregated_classifier`, deep supervision, or attention aggregation;\n- inspect classifier outputs, class counts, losses, or scikit-learn metrics;\n- prepare or validate the `us` HDF5 dataset layout;\n- export compatibility-score attention maps or overlay PNGs;\n- debug classification-specific CUDA, Visdom, sampler, HDF5, or checkpoint\n  issues.\n\nRoute 3D segmentation, NIfTI validation, and multi-attention U-Net feature-map\nexports to `..\u002Fsegmentation\u002FSKILL.md` instead.\n\n## Quick workflow\n\n1. Install the repository with CUDA-enabled PyTorch and the legacy\n   `torchsample` dependency. Then run:\n\n   ```bash\n   python ..\u002F..\u002Fscripts\u002Fcheck_env.py --repo-root \u002Fpath\u002Fto\u002FAttention-Gated-Networks --mode classification\n   ```\n\n2. Prepare an ultrasound HDF5 file with `x_train`, `p_train`, `x_val`, `p_val`,\n   `x_test`, `p_test`, and `label_names`. See [data-layout.md](references\u002Fdata-layout.md).\n3. Copy a Sononet config and replace its dataset path. The source configs use\n   `model.output_nc=14`, `input_nc=1`, `tensor_dim='2D'`, and `gpu_ids=[0]`.\n4. Train with the bundled replacement when real data is available:\n\nRelative config paths are resolved from the explicit `--repo-root`; relative\n`data_path` values are resolved from the config file's parent. The HDF5 data\nfile and any checkpoint\u002Fweights must already exist outside this skill. Bundled\nrunners and `check_env.py --config ...` reject missing or private `\u002Fvol\u002F...`\npaths before constructing a dataset.\n\n   ```bash\n   python scripts\u002Frun_classifier.py --repo-root \u002Fpath\u002Fto\u002FAttention-Gated-Networks --config path\u002Fto\u002Fconfig.json --mode train\n   ```\n\n5. Evaluate a checkpoint with the bundled replacement:\n\n   ```bash\n   python scripts\u002Frun_classifier.py --repo-root \u002Fpath\u002Fto\u002FAttention-Gated-Networks --config path\u002Fto\u002Fconfig.json --mode test\n   ```\n\n6. For attention models, use `model_type='sononet_grid_attention'` and read\n   [workflows.md](references\u002Fworkflows.md) before choosing `aggregation_mode` and\n   `aggregation`.\n7. Export a safe synthetic overlay, or adapt the command for a real 2D input:\n\n   ```bash\n   python scripts\u002Fexport_attention_overlay.py \\\n     --repo-root \u002Fpath\u002Fto\u002FAttention-Gated-Networks \\\n     --config configs\u002Fconfig_sononet_grid_att_8.json \\\n     --output-dir \u002Ftmp\u002Fag-net-attention\n   ```\n\n## Key decisions\n\n- CUDA is required for the unmodified wrappers because the model and tensor\n  paths call `.cuda()`.\n- The stock ultrasound configs assume 14 classes. Do not reuse them unchanged\n  for a different HDF5 label set.\n- `AggregatedClassifier` is for multiple-output attention models. Plain\n  Sononet classifiers use `FeedForwardClassifier`.\n- Synthetic smoke checks prove wiring only; they do not replace training on the\n  real ultrasound HDF5 file or evaluating a real checkpoint.\n- The original attention visualization script hard-codes private paths. Use the\n  bundled helper or parameterize your own script rather than running it as-is.\n\nThe source `train_classifaction.py` contains a 10-hour sleep after the final\nepoch's updates. It is retained in the source checkout for provenance; use the\nbundled runner when a bounded run without that hold is required.\n\n## Expected outputs\n\nTraining writes checkpoints and logs below the configured checkpoint directory\nand experiment name. Testing writes accumulated metrics and a `test_result.pkl`\ninside the model save directory. The bundled attention helper writes overlay\nPNGs and `.npy` arrays to the requested output directory.\n","# Classification\n\nUse this sub-skill for ultrasound scan-plane classification tasks in\nAttention-Gated Networks. It covers Sononet, Sononet2, Sononet Grid Attention,\n`FeedForwardClassifier`, `AggregatedClassifier`, the HDF5 ultrasound loader,\nclassification training\u002Ftesting scripts, and attention overlay export.\n\n## Read first\n\n- Read [api-reference.md](references\u002Fapi-reference.md) when you need exact\n  model names, wrapper behavior, config fields, aggregation modes, or class and\n  function signatures.\n- Read [data-layout.md](references\u002Fdata-layout.md) before preparing the\n  ultrasound HDF5 file or debugging loader\u002Fsampler failures.\n- Read [workflows.md](references\u002Fworkflows.md) for training, testing,\n  checkpoint evaluation, model inspection, and safe attention-overlay export.\n- Read [troubleshooting.md](references\u002Ftroubleshooting.md) when CUDA,\n  torchsample, Visdom, sampler, HDF5, or overlay failures appear.\n- Run [run_classifier.py](scripts\u002Frun_classifier.py) for skill-owned training\n  and testing commands that replace the source repo's root entry points.\n- Run [export_attention_overlay.py](scripts\u002Fexport_attention_overlay.py) to\n  replace the source repo's hard-coded attention visualization script with a\n  parameterized, skill-owned helper.\n- For fast environment validation, run the root helper\n  `..\u002F..\u002Fscripts\u002Fcheck_env.py --repo-root \u002Fpath\u002Fto\u002FAttention-Gated-Networks --mode classification`.\n\n## When to use this sub-skill\n\nUse this route when the user asks to:\n\n- train or test ultrasound scan-plane classification models;\n- choose between `sononet`, `sononet2`, and `sononet_grid_attention`;\n- configure `aggregated_classifier`, deep supervision, or attention aggregation;\n- inspect classifier outputs, class counts, losses, or scikit-learn metrics;\n- prepare or validate the `us` HDF5 dataset layout;\n- export compatibility-score attention maps or overlay PNGs;\n- debug classification-specific CUDA, Visdom, sampler, HDF5, or checkpoint\n  issues.\n\nRoute 3D segmentation, NIfTI validation, and multi-attention U-Net feature-map\nexports to `..\u002Fsegmentation\u002FSKILL.md` instead.\n\n## Quick workflow\n\n1. Install the repository with CUDA-enabled PyTorch and the legacy\n   `torchsample` dependency. Then run:\n\n   ```bash\n   python ..\u002F..\u002Fscripts\u002Fcheck_env.py --repo-root \u002Fpath\u002Fto\u002FAttention-Gated-Networks --mode classification\n   ```\n\n2. Prepare an ultrasound HDF5 file with `x_train`, `p_train`, `x_val`, `p_val`,\n   `x_test`, `p_test`, and `label_names`. See [data-layout.md](references\u002Fdata-layout.md).\n3. Copy a Sononet config and replace its dataset path. The source configs use\n   `model.output_nc=14`, `input_nc=1`, `tensor_dim='2D'`, and `gpu_ids=[0]`.\n4. Train with the bundled replacement when real data is available:\n\nRelative config paths are resolved from the explicit `--repo-root`; relative\n`data_path` values are resolved from the config file's parent. The HDF5 data\nfile and any checkpoint\u002Fweights must already exist outside this skill. Bundled\nrunners and `check_env.py --config ...` reject missing or private `\u002Fvol\u002F...`\npaths before constructing a dataset.\n\n   ```bash\n   python scripts\u002Frun_classifier.py --repo-root \u002Fpath\u002Fto\u002FAttention-Gated-Networks --config path\u002Fto\u002Fconfig.json --mode train\n   ```\n\n5. Evaluate a checkpoint with the bundled replacement:\n\n   ```bash\n   python scripts\u002Frun_classifier.py --repo-root \u002Fpath\u002Fto\u002FAttention-Gated-Networks --config path\u002Fto\u002Fconfig.json --mode test\n   ```\n\n6. For attention models, use `model_type='sononet_grid_attention'` and read\n   [workflows.md](references\u002Fworkflows.md) before choosing `aggregation_mode` and\n   `aggregation`.\n7. Export a safe synthetic overlay, or adapt the command for a real 2D input:\n\n   ```bash\n   python scripts\u002Fexport_attention_overlay.py \\\n     --repo-root \u002Fpath\u002Fto\u002FAttention-Gated-Networks \\\n     --config configs\u002Fconfig_sononet_grid_att_8.json \\\n     --output-dir \u002Ftmp\u002Fag-net-attention\n   ```\n\n## Key decisions\n\n- CUDA is required for the unmodified wrappers because the model and tensor\n  paths call `.cuda()`.\n- The stock ultrasound configs assume 14 classes. Do not reuse them unchanged\n  for a different HDF5 label set.\n- `AggregatedClassifier` is for multiple-output attention models. Plain\n  Sononet classifiers use `FeedForwardClassifier`.\n- Synthetic smoke checks prove wiring only; they do not replace training on the\n  real ultrasound HDF5 file or evaluating a real checkpoint.\n- The original attention visualization script hard-codes private paths. Use the\n  bundled helper or parameterize your own script rather than running it as-is.\n\nThe source `train_classifaction.py` contains a 10-hour sleep after the final\nepoch's updates. It is retained in the source checkout for provenance; use the\nbundled runner when a bounded run without that hold is required.\n\n## Expected outputs\n\nTraining writes checkpoints and logs below the configured checkpoint directory\nand experiment name. Testing writes accumulated metrics and a `test_result.pkl`\ninside the model save directory. The bundled attention helper writes overlay\nPNGs and `.npy` arrays to the requested output directory.\n",{"name":5,"license":14,"metadata":23,"description":7,"disable-model-invocation":25},{"disco-role":24},"operating",true,"main","ac3fe1afa80fb9a09775ecfb2b6cc3ba850a2db6","skills\u002Frepositories\u002Frepo-skills\u002Fattention-gated-networks\u002Fsub-skills\u002Fclassification",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 classification",{"tool":39,"command":40},"bunx","bunx skills add VectorSpaceLab\u002FAREX-Skill --skill classification",{"tool":42,"command":43},"pnpm","pnpm dlx skills add VectorSpaceLab\u002FAREX-Skill --skill classification","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fattention-gated-networks\u002Fsub-skills\u002Fclassification\u002FSKILL.md",[46,52,56,60,64,70],{"path":47,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":50,"sourceHtmlUrl":51},"references\u002Fapi-reference.md","REFERENCE","参考文档",6635,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fattention-gated-networks\u002Fsub-skills\u002Fclassification\u002Freferences\u002Fapi-reference.md",{"path":53,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":54,"sourceHtmlUrl":55},"references\u002Fdata-layout.md",3355,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fattention-gated-networks\u002Fsub-skills\u002Fclassification\u002Freferences\u002Fdata-layout.md",{"path":57,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":58,"sourceHtmlUrl":59},"references\u002Ftroubleshooting.md",4214,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fattention-gated-networks\u002Fsub-skills\u002Fclassification\u002Freferences\u002Ftroubleshooting.md",{"path":61,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":62,"sourceHtmlUrl":63},"references\u002Fworkflows.md",5087,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fattention-gated-networks\u002Fsub-skills\u002Fclassification\u002Freferences\u002Fworkflows.md",{"path":65,"resourceType":66,"resourceTypeDesc":67,"sizeBytes":68,"sourceHtmlUrl":69},"scripts\u002Fexport_attention_overlay.py","SCRIPT","脚本",8577,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fattention-gated-networks\u002Fsub-skills\u002Fclassification\u002Fscripts\u002Fexport_attention_overlay.py",{"path":71,"resourceType":66,"resourceTypeDesc":67,"sizeBytes":72,"sourceHtmlUrl":73},"scripts\u002Frun_classifier.py",13682,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fattention-gated-networks\u002Fsub-skills\u002Fclassification\u002Fscripts\u002Frun_classifier.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",[],[]]