[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M796RVKASQW0SW6R5WTX6CSV":3,"skill-same:01M796RVKASQW0SW6R5WTX6CSV":83},{"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},"01M796RVKASQW0SW6R5WTX6CSV","training-and-checkpoints",null,"Operate TurboDiffusion rCM\u002FSLA training setup, dry-run command construction, checkpoint conversion, model merge, quantized export, and data\u002Fcheckpoint layout validation.","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\u002Fturbo-diffusion\u002Fsub-skills\u002Ftraining-and-checkpoints\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: training-and-checkpoints\ndescription: \"Operate TurboDiffusion rCM\u002FSLA training setup, dry-run command\n  construction, checkpoint conversion, model merge, quantized export, and\n  data\u002Fcheckpoint layout validation.\"\ndisable-model-invocation: true\nmetadata:\n  disco-role: operating\nlicense: Apache 2.0\n---\n\n# TurboDiffusion Training And Checkpoints\n\nUse this sub-skill when the task is about TurboDiffusion rCM\u002FSLA training setup, debug-only training command construction, Distributed Checkpoint (DCP) versus PyTorch `.pth` formats, merging SLA\u002FrCM updates, converting sharded safetensors, or exporting modified\u002Fquantized Wan checkpoints.\n\nDo **not** run full training or model conversion by default. Full training, dataset synthesis, checkpoint quantization, and inference-format conversion are GPU\u002Fmodel\u002Fdata heavy and remain skip-expensive unless the user explicitly authorizes them and provides checkpoints, data, credentials policy, and backend readiness.\n\n## Route first\n\n- Use [references\u002Ftraining-and-data.md](references\u002Ftraining-and-data.md) for rCM\u002FSLA training prerequisites, debug\u002Fdry-run planning, data shard layout, checkpoint roots, and WANDB\u002Fenvironment decisions.\n- Use [references\u002Fcheckpoint-workflows.md](references\u002Fcheckpoint-workflows.md) for DCP-to-`.pth`, `.pth`-to-DCP, merge arithmetic, safetensors conversion, and `modify_model` export semantics.\n- Use [references\u002Fcli-reference.md](references\u002Fcli-reference.md) for safe command templates and supported flags for public training\u002Fcheckpoint utilities.\n- Use [references\u002Ftroubleshooting.md](references\u002Ftroubleshooting.md) when a user reports DCP\u002FPTH confusion, missing roots\u002Fshards, training extras, WANDB issues, mismatched model names, or quantization\u002FSLA failures.\n\n## Bundled safe helpers\n\nThese helpers only build commands or run a tiny CPU tensor arithmetic check; they do not download, train, execute model inference, or require the original checkout.\n\n- [scripts\u002Fbuild_training_dryrun_command.py](scripts\u002Fbuild_training_dryrun_command.py): render a `torchrun ... -m scripts.train --dryrun` command with explicit config\u002Fcheckpoint\u002Fdata overrides.\n- [scripts\u002Fbuild_modify_model_command.py](scripts\u002Fbuild_modify_model_command.py): render a `modify_model.py` checkpoint export\u002Fquantization command for Wan2.1\u002FWan2.2 model profiles.\n- [scripts\u002Fbuild_safetensors_to_pth_command.py](scripts\u002Fbuild_safetensors_to_pth_command.py): render a sharded safetensors-to-`.pth` command with optional key prefixing.\n- [scripts\u002Ftiny_merge_models_check.py](scripts\u002Ftiny_merge_models_check.py): create temporary tiny tensors and verify the source merge formula `base + w * (diff_target - diff_base)` on CPU.\n\n## Operating rules\n\n1. Start with layout validation: checkpoint root, teacher DCP, VAE, text encoder, negative embedding, output root, and WebDataset shard pattern.\n2. For training requests, prefer a dry-run\u002Fconfig-composition command first. Explain that actual training needs CUDA\u002Fmulti-GPU readiness, training extras, model\u002Fdata files, and a WANDB\u002Foffline logging decision.\n3. For DCP conversion, distinguish direction: `.pth` teacher checkpoint must be converted to DCP before FSDP training; saved training DCP directories can later be converted back to `.pth`.\n4. For model merge, warn that mismatched shapes\u002Fkeys are not fatal in the source utility but require manual key-coverage inspection before treating the output as good.\n5. For quantized export, ensure the source checkpoint is already in the expected rCM\u002FSLA state-dict shape, choose a valid Wan model name, and align `--quant_linear` with later inference\u002Fserving commands.\n6. For source-layout use, mention the public package quirk: top-level imports such as `rcm`, `imaginaire`, `SLA`, `ops`, `scripts`, and `modify_model` often require setting `PYTHONPATH` to the package source directory.\n\n## Cross-skill routing\n\n- Route full T2V\u002FI2V use of converted checkpoints to the `video-inference` sub-skill.\n- Route interactive serving with converted checkpoints to the `interactive-serving` sub-skill.\n- Route CUDA extension build, INT8\u002FFastNorm op failures, plain SLA versus SageSLA backend decisions, or SpargeAttn installation issues to the `acceleration-backends` sub-skill.\n","# TurboDiffusion Training And Checkpoints\n\nUse this sub-skill when the task is about TurboDiffusion rCM\u002FSLA training setup, debug-only training command construction, Distributed Checkpoint (DCP) versus PyTorch `.pth` formats, merging SLA\u002FrCM updates, converting sharded safetensors, or exporting modified\u002Fquantized Wan checkpoints.\n\nDo **not** run full training or model conversion by default. Full training, dataset synthesis, checkpoint quantization, and inference-format conversion are GPU\u002Fmodel\u002Fdata heavy and remain skip-expensive unless the user explicitly authorizes them and provides checkpoints, data, credentials policy, and backend readiness.\n\n## Route first\n\n- Use [references\u002Ftraining-and-data.md](references\u002Ftraining-and-data.md) for rCM\u002FSLA training prerequisites, debug\u002Fdry-run planning, data shard layout, checkpoint roots, and WANDB\u002Fenvironment decisions.\n- Use [references\u002Fcheckpoint-workflows.md](references\u002Fcheckpoint-workflows.md) for DCP-to-`.pth`, `.pth`-to-DCP, merge arithmetic, safetensors conversion, and `modify_model` export semantics.\n- Use [references\u002Fcli-reference.md](references\u002Fcli-reference.md) for safe command templates and supported flags for public training\u002Fcheckpoint utilities.\n- Use [references\u002Ftroubleshooting.md](references\u002Ftroubleshooting.md) when a user reports DCP\u002FPTH confusion, missing roots\u002Fshards, training extras, WANDB issues, mismatched model names, or quantization\u002FSLA failures.\n\n## Bundled safe helpers\n\nThese helpers only build commands or run a tiny CPU tensor arithmetic check; they do not download, train, execute model inference, or require the original checkout.\n\n- [scripts\u002Fbuild_training_dryrun_command.py](scripts\u002Fbuild_training_dryrun_command.py): render a `torchrun ... -m scripts.train --dryrun` command with explicit config\u002Fcheckpoint\u002Fdata overrides.\n- [scripts\u002Fbuild_modify_model_command.py](scripts\u002Fbuild_modify_model_command.py): render a `modify_model.py` checkpoint export\u002Fquantization command for Wan2.1\u002FWan2.2 model profiles.\n- [scripts\u002Fbuild_safetensors_to_pth_command.py](scripts\u002Fbuild_safetensors_to_pth_command.py): render a sharded safetensors-to-`.pth` command with optional key prefixing.\n- [scripts\u002Ftiny_merge_models_check.py](scripts\u002Ftiny_merge_models_check.py): create temporary tiny tensors and verify the source merge formula `base + w * (diff_target - diff_base)` on CPU.\n\n## Operating rules\n\n1. Start with layout validation: checkpoint root, teacher DCP, VAE, text encoder, negative embedding, output root, and WebDataset shard pattern.\n2. For training requests, prefer a dry-run\u002Fconfig-composition command first. Explain that actual training needs CUDA\u002Fmulti-GPU readiness, training extras, model\u002Fdata files, and a WANDB\u002Foffline logging decision.\n3. For DCP conversion, distinguish direction: `.pth` teacher checkpoint must be converted to DCP before FSDP training; saved training DCP directories can later be converted back to `.pth`.\n4. For model merge, warn that mismatched shapes\u002Fkeys are not fatal in the source utility but require manual key-coverage inspection before treating the output as good.\n5. For quantized export, ensure the source checkpoint is already in the expected rCM\u002FSLA state-dict shape, choose a valid Wan model name, and align `--quant_linear` with later inference\u002Fserving commands.\n6. For source-layout use, mention the public package quirk: top-level imports such as `rcm`, `imaginaire`, `SLA`, `ops`, `scripts`, and `modify_model` often require setting `PYTHONPATH` to the package source directory.\n\n## Cross-skill routing\n\n- Route full T2V\u002FI2V use of converted checkpoints to the `video-inference` sub-skill.\n- Route interactive serving with converted checkpoints to the `interactive-serving` sub-skill.\n- Route CUDA extension build, INT8\u002FFastNorm op failures, plain SLA versus SageSLA backend decisions, or SpargeAttn installation issues to the `acceleration-backends` sub-skill.\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\u002Fturbo-diffusion\u002Fsub-skills\u002Ftraining-and-checkpoints",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-checkpoints",{"tool":40,"command":41},"bunx","bunx skills add VectorSpaceLab\u002FAREX-Skill --skill training-and-checkpoints",{"tool":43,"command":44},"pnpm","pnpm dlx skills add VectorSpaceLab\u002FAREX-Skill --skill training-and-checkpoints","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fturbo-diffusion\u002Fsub-skills\u002Ftraining-and-checkpoints\u002FSKILL.md",[47,53,57,61,65,71,75,79],{"path":48,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":51,"sourceHtmlUrl":52},"references\u002Fcheckpoint-workflows.md","REFERENCE","参考文档",7752,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fturbo-diffusion\u002Fsub-skills\u002Ftraining-and-checkpoints\u002Freferences\u002Fcheckpoint-workflows.md",{"path":54,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":55,"sourceHtmlUrl":56},"references\u002Fcli-reference.md",8065,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fturbo-diffusion\u002Fsub-skills\u002Ftraining-and-checkpoints\u002Freferences\u002Fcli-reference.md",{"path":58,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":59,"sourceHtmlUrl":60},"references\u002Ftraining-and-data.md",7049,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fturbo-diffusion\u002Fsub-skills\u002Ftraining-and-checkpoints\u002Freferences\u002Ftraining-and-data.md",{"path":62,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":63,"sourceHtmlUrl":64},"references\u002Ftroubleshooting.md",8440,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fturbo-diffusion\u002Fsub-skills\u002Ftraining-and-checkpoints\u002Freferences\u002Ftroubleshooting.md",{"path":66,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":69,"sourceHtmlUrl":70},"scripts\u002Fbuild_modify_model_command.py","SCRIPT","脚本",6206,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fturbo-diffusion\u002Fsub-skills\u002Ftraining-and-checkpoints\u002Fscripts\u002Fbuild_modify_model_command.py",{"path":72,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":73,"sourceHtmlUrl":74},"scripts\u002Fbuild_safetensors_to_pth_command.py",4699,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fturbo-diffusion\u002Fsub-skills\u002Ftraining-and-checkpoints\u002Fscripts\u002Fbuild_safetensors_to_pth_command.py",{"path":76,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":77,"sourceHtmlUrl":78},"scripts\u002Fbuild_training_dryrun_command.py",7598,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fturbo-diffusion\u002Fsub-skills\u002Ftraining-and-checkpoints\u002Fscripts\u002Fbuild_training_dryrun_command.py",{"path":80,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":81,"sourceHtmlUrl":82},"scripts\u002Ftiny_merge_models_check.py",6102,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fturbo-diffusion\u002Fsub-skills\u002Ftraining-and-checkpoints\u002Fscripts\u002Ftiny_merge_models_check.py",[84,91,98,105,112,119],{"skillNo":85,"name":86,"nameZh":6,"description":87,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":88,"categories":89,"keywords":90,"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":92,"name":93,"nameZh":6,"description":94,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":95,"categories":96,"keywords":97,"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":99,"name":100,"nameZh":6,"description":101,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":102,"categories":103,"keywords":104,"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":106,"name":107,"nameZh":6,"description":108,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":109,"categories":110,"keywords":111,"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":113,"name":114,"nameZh":6,"description":115,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":116,"categories":117,"keywords":118,"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":120,"name":121,"nameZh":6,"description":122,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":123,"categories":124,"keywords":125,"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",[],[]]