[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7F5T98DSY0G5K6HRF47DASC":3,"skill-same:01M7F5T98DSY0G5K6HRF47DASC":70},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":27,"commitSha":28,"skillDirectory":29,"forksCount":30,"repositoryHtmlUrl":31,"repositoryHomepageUrl":6,"ownerHtmlUrl":32,"ownerBlogUrl":6,"ownerDescription":33,"updatedAt":34,"install":35,"resources":47},"01M7F5T98DSY0G5K6HRF47DASC","swin-transformer",null,"Use this repo skill for Microsoft Swin-Transformer image-classification model, config, data, checkpoint, SimMIM, Swin-MoE, and optional CUDA acceleration workflows.","lza6\u002FArmor-piercing-system-prompts","lza6","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F128880206?v=4","01M7EB3K7WPX5T28EJ2NVHE6EG","Chinese-tingfeng","01M7EB3K97E1KQ16QWKJMFNGX8",11,"MIT","https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F05-技能与代理开发\u002Fswin-transformer\u002FSKILL.md",[],[],"2026-11-16 10:05:34",false,"---\nname: swin-transformer\ndescription: \"Use this repo skill for Microsoft Swin-Transformer\n  image-classification model, config, data, checkpoint, SimMIM, Swin-MoE, and\n  optional CUDA acceleration workflows.\"\ndisable-model-invocation: true\nmetadata:\n  disco-role: operating\nlicense: MIT\n---\n\n# Swin-Transformer Repo Skill\n\nUse this skill when a task involves the official Microsoft Swin-Transformer image-classification codebase: Swin V1\u002FV2, Swin-MLP, SimMIM masked image modeling, ImageNet data layouts, pretrained checkpoint handling, supervised training\u002Fevaluation commands, Swin-MoE, or the optional fused CUDA window-process extension.\n\nThis is a self-contained operating guide. It does not require the original source checkout used during skill creation. When a workflow needs code execution, use the user's current checkout or installed copy of Swin-Transformer and the bundled helper scripts in this skill.\n\n## Quick routing\n\n| User intent | Read next |\n| --- | --- |\n| Build or inspect `SwinTransformer`, `SwinTransformerV2`, `SwinMLP`, `build_model`, tensor shapes, FLOPs, or CPU smoke models | `sub-skills\u002Fcore-models\u002FSKILL.md` |\n| Validate ImageNet folder\u002Fzip\u002F22K layouts, understand `--zip`, map files, SimMIM data, checkpoint resume\u002Fpretrained behavior, or 22K-to-1K head remapping | `sub-skills\u002Fdata-and-checkpoints\u002FSKILL.md` |\n| Construct or debug supervised `main.py` commands for training, fine-tuning, evaluation, throughput, AMP, DDP launch, or `--opts` overrides | `sub-skills\u002Ftraining-eval-cli\u002FSKILL.md` |\n| Run or adapt SimMIM pretraining\u002Ffine-tuning\u002Fevaluation workflows and mask\u002Floss checks | `sub-skills\u002Fsimmim-workflows\u002FSKILL.md` |\n| Work with Swin-MoE, Tutel, Apex fused optimizers\u002Flayernorm, or the fused CUDA window-process extension | `sub-skills\u002Fmoe-and-acceleration\u002FSKILL.md` |\n| Need a config\u002Fmodel family map before choosing a route | `references\u002Fmodel-zoo-and-configs.md` |\n| Need shared YACS config and flag behavior | `references\u002Fconfiguration.md` |\n| Need cross-cutting troubleshooting | `references\u002Ftroubleshooting.md` |\n| Need to verify that a checkout\u002Fenvironment is usable | `scripts\u002Fcheck_env.py` |\n\n## Baseline prerequisites\n\nSwin-Transformer is a research-code checkout rather than an ordinary packaged distribution. For code execution, future agents should work against a current checkout or source tree on `PYTHONPATH` and install the runtime dependencies relevant to the selected workflow:\n\n- Required for baseline inspection and CPU smoke checks: Python, PyTorch, torchvision, `timm==0.4.12`, `yacs`, `PyYAML`, `numpy`, `scipy`, and `termcolor`.\n- Required for full supervised or SimMIM training\u002Fevaluation: CUDA-capable PyTorch, GPUs, ImageNet-style data, checkpoints when evaluating\u002Ffine-tuning, and distributed launcher environment.\n- Optional: Apex for fused layernorm\u002Ffused optimizers, Tutel for Swin-MoE, and a compiled `swin_window_process` CUDA extension for `--fused_window_process`.\n\nMinimal import smoke for a checkout:\n\n```bash\npython - \u003C\u003C'PY'\nfrom config import get_config\nfrom models import build_model\nfrom data import build_loader\nprint('Swin-Transformer modules import')\nPY\n```\n\nIf that fails, read `references\u002Ftroubleshooting.md` before running full training.\n\n## Safe helper scripts\n\n- Run `scripts\u002Fcheck_env.py --repo-root \u003Ccheckout>` to check required imports and optional CUDA\u002FTutel\u002FApex\u002Ffused-extension availability without training or downloading.\n- Run `scripts\u002Finspect_swin_config.py --repo-root \u003Ccheckout> --cfg \u003Cconfig.yaml>` to summarize a YAML config and flag risky settings.\n- Run `scripts\u002Fswin_cli_command_builder.py --help` to assemble copyable command templates for supervised, SimMIM, and MoE workflows.\n\nThese helpers are safe by default: they do not download models or data, do not run distributed training, and do not build CUDA extensions.\n\n## Important boundaries\n\n- This skill covers image classification and the code in this repository. Object detection, semantic segmentation, video action recognition, feature distillation, and other downstream projects linked by the public README are separate repositories and are not covered here.\n- Full accuracy reproduction and throughput benchmarking are intentionally not used as skill-verification gates because they require large datasets, checkpoints, GPUs, and long distributed runs.\n- CPU smoke checks validate config\u002Fmodel\u002Fdata plumbing only. They are not evidence that CUDA training, Swin-MoE, Apex, or the fused window kernel works.\n\n## Provenance and staleness\n\nRead `references\u002Frepo-provenance.md` before deciding whether this skill matches a particular checkout. If source APIs, config layout, CLI flags, or dependency behavior differ, refresh this repo skill before relying on it.\n","# Swin-Transformer Repo Skill\n\nUse this skill when a task involves the official Microsoft Swin-Transformer image-classification codebase: Swin V1\u002FV2, Swin-MLP, SimMIM masked image modeling, ImageNet data layouts, pretrained checkpoint handling, supervised training\u002Fevaluation commands, Swin-MoE, or the optional fused CUDA window-process extension.\n\nThis is a self-contained operating guide. It does not require the original source checkout used during skill creation. When a workflow needs code execution, use the user's current checkout or installed copy of Swin-Transformer and the bundled helper scripts in this skill.\n\n## Quick routing\n\n| User intent | Read next |\n| --- | --- |\n| Build or inspect `SwinTransformer`, `SwinTransformerV2`, `SwinMLP`, `build_model`, tensor shapes, FLOPs, or CPU smoke models | `sub-skills\u002Fcore-models\u002FSKILL.md` |\n| Validate ImageNet folder\u002Fzip\u002F22K layouts, understand `--zip`, map files, SimMIM data, checkpoint resume\u002Fpretrained behavior, or 22K-to-1K head remapping | `sub-skills\u002Fdata-and-checkpoints\u002FSKILL.md` |\n| Construct or debug supervised `main.py` commands for training, fine-tuning, evaluation, throughput, AMP, DDP launch, or `--opts` overrides | `sub-skills\u002Ftraining-eval-cli\u002FSKILL.md` |\n| Run or adapt SimMIM pretraining\u002Ffine-tuning\u002Fevaluation workflows and mask\u002Floss checks | `sub-skills\u002Fsimmim-workflows\u002FSKILL.md` |\n| Work with Swin-MoE, Tutel, Apex fused optimizers\u002Flayernorm, or the fused CUDA window-process extension | `sub-skills\u002Fmoe-and-acceleration\u002FSKILL.md` |\n| Need a config\u002Fmodel family map before choosing a route | `references\u002Fmodel-zoo-and-configs.md` |\n| Need shared YACS config and flag behavior | `references\u002Fconfiguration.md` |\n| Need cross-cutting troubleshooting | `references\u002Ftroubleshooting.md` |\n| Need to verify that a checkout\u002Fenvironment is usable | `scripts\u002Fcheck_env.py` |\n\n## Baseline prerequisites\n\nSwin-Transformer is a research-code checkout rather than an ordinary packaged distribution. For code execution, future agents should work against a current checkout or source tree on `PYTHONPATH` and install the runtime dependencies relevant to the selected workflow:\n\n- Required for baseline inspection and CPU smoke checks: Python, PyTorch, torchvision, `timm==0.4.12`, `yacs`, `PyYAML`, `numpy`, `scipy`, and `termcolor`.\n- Required for full supervised or SimMIM training\u002Fevaluation: CUDA-capable PyTorch, GPUs, ImageNet-style data, checkpoints when evaluating\u002Ffine-tuning, and distributed launcher environment.\n- Optional: Apex for fused layernorm\u002Ffused optimizers, Tutel for Swin-MoE, and a compiled `swin_window_process` CUDA extension for `--fused_window_process`.\n\nMinimal import smoke for a checkout:\n\n```bash\npython - \u003C\u003C'PY'\nfrom config import get_config\nfrom models import build_model\nfrom data import build_loader\nprint('Swin-Transformer modules import')\nPY\n```\n\nIf that fails, read `references\u002Ftroubleshooting.md` before running full training.\n\n## Safe helper scripts\n\n- Run `scripts\u002Fcheck_env.py --repo-root \u003Ccheckout>` to check required imports and optional CUDA\u002FTutel\u002FApex\u002Ffused-extension availability without training or downloading.\n- Run `scripts\u002Finspect_swin_config.py --repo-root \u003Ccheckout> --cfg \u003Cconfig.yaml>` to summarize a YAML config and flag risky settings.\n- Run `scripts\u002Fswin_cli_command_builder.py --help` to assemble copyable command templates for supervised, SimMIM, and MoE workflows.\n\nThese helpers are safe by default: they do not download models or data, do not run distributed training, and do not build CUDA extensions.\n\n## Important boundaries\n\n- This skill covers image classification and the code in this repository. Object detection, semantic segmentation, video action recognition, feature distillation, and other downstream projects linked by the public README are separate repositories and are not covered here.\n- Full accuracy reproduction and throughput benchmarking are intentionally not used as skill-verification gates because they require large datasets, checkpoints, GPUs, and long distributed runs.\n- CPU smoke checks validate config\u002Fmodel\u002Fdata plumbing only. They are not evidence that CUDA training, Swin-MoE, Apex, or the fused window kernel works.\n\n## Provenance and staleness\n\nRead `references\u002Frepo-provenance.md` before deciding whether this skill matches a particular checkout. If source APIs, config layout, CLI flags, or dependency behavior differ, refresh this repo skill before relying on it.\n",{"name":5,"license":15,"metadata":24,"description":7,"disable-model-invocation":26},{"disco-role":25},"operating",true,"main","d69096e239f2a89aabc2cd078c47664a9d3b1e11","05-技能与代理开发\u002Fswin-transformer",3,"https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts","https:\u002F\u002Fgithub.com\u002Flza6","个人vx：Tf00798\r\n个人✈：@TFGY999","2026-10-08 14:29:22",{"source":8,"skillName":5,"commands":36,"prompt":46,"zipAvailable":26},[37,40,43],{"tool":38,"command":39},"npx","npx skills add lza6\u002FArmor-piercing-system-prompts --skill swin-transformer",{"tool":41,"command":42},"bunx","bunx skills add lza6\u002FArmor-piercing-system-prompts --skill swin-transformer",{"tool":44,"command":45},"pnpm","pnpm dlx skills add lza6\u002FArmor-piercing-system-prompts --skill swin-transformer","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F05-技能与代理开发\u002Fswin-transformer\u002FSKILL.md",[48,54,58,62,66],{"path":49,"resourceType":50,"resourceTypeDesc":51,"sizeBytes":52,"sourceHtmlUrl":53},"references\u002Fconfiguration.md","REFERENCE","参考文档",2183,"https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F05-技能与代理开发\u002Fswin-transformer\u002Freferences\u002Fconfiguration.md",{"path":55,"resourceType":50,"resourceTypeDesc":51,"sizeBytes":56,"sourceHtmlUrl":57},"references\u002Fmodel-zoo-and-configs.md",2719,"https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F05-技能与代理开发\u002Fswin-transformer\u002Freferences\u002Fmodel-zoo-and-configs.md",{"path":59,"resourceType":50,"resourceTypeDesc":51,"sizeBytes":60,"sourceHtmlUrl":61},"references\u002Frepo-provenance.md",2008,"https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F05-技能与代理开发\u002Fswin-transformer\u002Freferences\u002Frepo-provenance.md",{"path":63,"resourceType":50,"resourceTypeDesc":51,"sizeBytes":64,"sourceHtmlUrl":65},"references\u002Frepo-routing-metadata.json",523,"https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F05-技能与代理开发\u002Fswin-transformer\u002Freferences\u002Frepo-routing-metadata.json",{"path":67,"resourceType":50,"resourceTypeDesc":51,"sizeBytes":68,"sourceHtmlUrl":69},"references\u002Ftroubleshooting.md",2511,"https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F05-技能与代理开发\u002Fswin-transformer\u002Freferences\u002Ftroubleshooting.md",[71,78,83,90,97,104],{"skillNo":72,"name":73,"nameZh":6,"description":74,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":75,"categories":76,"keywords":77,"publishedAt":19},"01M7FFG485KY83TD6HSC4YPCMC","obsidian-clipper-template-creator","Guide for creating templates for the Obsidian Web Clipper. Use when you want to create a new clipping template, understand available variables, or format clipped content.","https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F07-后端与架构\u002Fobsidian-clipper-template-creator\u002FSKILL.md",[],[],{"skillNo":79,"name":73,"nameZh":6,"description":74,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":80,"categories":81,"keywords":82,"publishedAt":19},"01M7FFG0T6AZ0STEJPSB3CX84K","https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F07-后端与架构\u002Fobsidian-clipper-template-creator-antigravity-awesome-skills-main\u002FSKILL.md",[],[],{"skillNo":84,"name":85,"nameZh":6,"description":86,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":87,"categories":88,"keywords":89,"publishedAt":19},"01M7FFFY0QEM14SKWPDYGMXM2P","python-api","Programmatic Object-Detection-Metrics API guidance for BoundingBox, BoundingBoxes, Evaluator, enums, coordinate conversion, AP\u002FmAP semantics, and source-style import diagnostics.","https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F07-后端与架构\u002Fobject-detection-metrics\u002Fsub-skills\u002Fpython-api\u002FSKILL.md",[],[],{"skillNo":91,"name":92,"nameZh":6,"description":93,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":94,"categories":95,"keywords":96,"publishedAt":19},"01M7FFFVZQZ2GJ0C0TGP7JHD20","file-evaluation","Folder-based PASCAL VOC AP\u002FmAP evaluation from ground-truth and detection text files.","https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F07-后端与架构\u002Fobject-detection-metrics\u002Fsub-skills\u002Ffile-evaluation\u002FSKILL.md",[],[],{"skillNo":98,"name":99,"nameZh":6,"description":100,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":101,"categories":102,"keywords":103,"publishedAt":19},"01M7FFFSFNYSE77SJ6PE1CK2DA","object-detection-metrics","Guides legacy Object-Detection-Metrics PASCAL VOC AP\u002FmAP evaluation from detection text folders and direct Python API classes.","https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F07-后端与架构\u002Fobject-detection-metrics\u002FSKILL.md",[],[],{"skillNo":105,"name":106,"nameZh":6,"description":107,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":108,"categories":109,"keywords":110,"publishedAt":19},"01M7FFFP0GCK1RVXBCPQ8APBMT","nx-workspace-patterns","Configure and optimize Nx monorepo workspaces. Use when setting up Nx, configuring project boundaries, optimizing build caching, or implementing affected commands.","https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F07-后端与架构\u002Fnx-workspace-patterns\u002FSKILL.md",[],[]]