[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7F9NMMK89JMHQ2KGHGPX9NH":3,"skill-same:01M7F9NMMK89JMHQ2KGHGPX9NH":67},{"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":28,"commitSha":29,"skillDirectory":30,"forksCount":31,"repositoryHtmlUrl":32,"repositoryHomepageUrl":6,"ownerHtmlUrl":33,"ownerBlogUrl":6,"ownerDescription":34,"updatedAt":35,"install":36,"resources":48},"01M7F9NMMK89JMHQ2KGHGPX9NH","h2o-llmstudio",null,"Use H2O LLM Studio for no-code and CLI LLM fine-tuning, dataset\u002Fconfig preparation, training diagnostics, evaluation, prompting, export, and the Wave GUI runtime.","lza6\u002FArmor-piercing-system-prompts","lza6","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F128880206?v=4","01M7EB3K7WPX5T28EJ2NVHE6EG","Chinese-tingfeng","01M7EB3K97E1KQ16QWKJMFNGX8",11,"Apache-2.0","https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F06-前端与设计\u002Fh2o-llmstudio\u002FSKILL.md",[],[],"2026-11-16 10:05:34",false,"---\nname: h2o-llmstudio\ndescription: \"Use H2O LLM Studio for no-code and CLI LLM fine-tuning,\n  dataset\u002Fconfig preparation, training diagnostics, evaluation, prompting,\n  export, and the Wave GUI runtime.\"\ndisable-model-invocation: true\nmetadata:\n  disco-role: operating\nlicense: Apache 2.0\n---\n\n# H2O LLM Studio\n\nUse this repo skill when a task involves H2O LLM Studio as a package or runtime: preparing datasets\u002Fconfigs, launching the Wave GUI, running or debugging fine-tuning experiments, understanding model\u002Fevaluation behavior, or prompting\u002Fexporting a trained experiment.\n\n## Before you act\n\n1. Read [repo provenance](references\u002Frepo-provenance.md) if the user is working in a checkout and you need to decide whether this skill matches the current source version.\n2. Read [runtime setup](references\u002Fruntime-setup.md) before recommending install, app, Docker, CUDA, DeepSpeed, or working-directory commands.\n3. Run or adapt [scripts\u002Fcheck_environment.py](scripts\u002Fcheck_environment.py) for safe import\u002Fbackend\u002Fruntime-asset diagnostics before starting services, downloads, or training.\n4. Use [cross-cutting troubleshooting](references\u002Ftroubleshooting.md) for Python version, package layout, CUDA\u002FNVIDIA, DeepSpeed, keyring, data\u002Fmodel download, and credential failures.\n\n## Route by task\n\n| User task | Read |\n|---|---|\n| Start or debug the no-code Wave GUI, Docker app, remote proxy, app workdir, settings, keyring, or browser\u002Fserver lifecycle | [app-and-ui](sub-skills\u002Fapp-and-ui\u002FSKILL.md) |\n| Build or validate YAML configs, choose problem types, map dataset columns, inspect data schemas, or round-trip config files | [configuration-and-data](sub-skills\u002Fconfiguration-and-data\u002FSKILL.md) |\n| Launch or diagnose `train.py`, CPU\u002FGPU tiny smokes, multi-GPU\u002FDeepSpeed, mixed precision, LoRA, output artifacts, or experiment status | [training-and-experiments](sub-skills\u002Ftraining-and-experiments\u002FSKILL.md) |\n| Understand model wrappers, losses, metrics, prediction files, plots, AI judge metrics, or generation vs forward-pass evaluation behavior | [modeling-and-evaluation](sub-skills\u002Fmodeling-and-evaluation\u002FSKILL.md) |\n| Prompt a trained experiment, preflight saved artifacts, change generation parameters, publish to Hugging Face Hub, or hand off to h2oGPT | [export-and-prompt](sub-skills\u002Fexport-and-prompt\u002FSKILL.md) |\n\n## Install and quick verification\n\nFrom a user's own H2O LLM Studio runtime root, use the repo-supported Python 3.10 setup path before app or training work:\n\n```bash\nmake setup\npython scripts\u002Fcheck_environment.py --runtime-root . --check-config-assets\n```\n\nWhen `make setup` is not appropriate, create an isolated Python 3.10 environment and install the same locked\u002Fruntime dependency set described in [runtime setup](references\u002Fruntime-setup.md); do not modify a user-owned environment without approval.\n\n## Fast runtime facts\n\n- Package metadata expects Python 3.10 and a GPU-oriented PyTorch\u002FTransformers stack.\n- The documented runtime is usually a source-layout H2O LLM Studio runtime root or Docker image, not a tiny import-only library. Commands that refer to `llm_studio\u002F...` should be run from the user's own runtime root or translated to an equivalent module\u002Fwrapper command.\n- Production fine-tuning is NVIDIA-GPU oriented. CPU-style unit\u002Fintegration configs can validate mechanics, but they do not prove real large-model throughput or DeepSpeed\u002Fmulti-GPU behavior.\n- The app and config layers use runtime assets such as `prompts\u002F`, `model_cards\u002F`, `static\u002F`, and `pyproject.toml`; importing `llm_studio` alone is not enough to prove a working app\u002Ftraining root.\n- Optional or external services include Hugging Face Hub\u002Fmodel downloads, Kaggle\u002FHF datasets, OpenAI-compatible judge metrics, W&B logging, S3\u002FAzure\u002FH2O Drive data connectors, Docker\u002FNVIDIA container runtime, and browser UI tests.\n\n## Safe first checks\n\n```bash\npython scripts\u002Fcheck_environment.py --runtime-root . --check-cuda --check-config-assets\npython sub-skills\u002Fconfiguration-and-data\u002Fscripts\u002Finspect_config.py --help\npython sub-skills\u002Ftraining-and-experiments\u002Fscripts\u002Fcheck_training_environment.py --help\npython sub-skills\u002Fexport-and-prompt\u002Fscripts\u002Fcheck_experiment_artifacts.py --help\n```\n\nThese helpers diagnose environment\u002Fconfig\u002Fartifact readiness. They do not download models, start Wave, run training, call OpenAI, publish to Hugging Face, or mutate experiment outputs unless an option explicitly says so.\n\n## When to stop and ask\n\nAsk before running long training, downloading datasets\u002Fmodels, starting a public-facing Wave server, using API\u002FHF\u002FW&B\u002Fcloud credentials, publishing to Hugging Face Hub, modifying a user-owned Python environment, changing Docker\u002FNVIDIA host setup, or overwriting an experiment output directory.\n","# H2O LLM Studio\n\nUse this repo skill when a task involves H2O LLM Studio as a package or runtime: preparing datasets\u002Fconfigs, launching the Wave GUI, running or debugging fine-tuning experiments, understanding model\u002Fevaluation behavior, or prompting\u002Fexporting a trained experiment.\n\n## Before you act\n\n1. Read [repo provenance](references\u002Frepo-provenance.md) if the user is working in a checkout and you need to decide whether this skill matches the current source version.\n2. Read [runtime setup](references\u002Fruntime-setup.md) before recommending install, app, Docker, CUDA, DeepSpeed, or working-directory commands.\n3. Run or adapt [scripts\u002Fcheck_environment.py](scripts\u002Fcheck_environment.py) for safe import\u002Fbackend\u002Fruntime-asset diagnostics before starting services, downloads, or training.\n4. Use [cross-cutting troubleshooting](references\u002Ftroubleshooting.md) for Python version, package layout, CUDA\u002FNVIDIA, DeepSpeed, keyring, data\u002Fmodel download, and credential failures.\n\n## Route by task\n\n| User task | Read |\n|---|---|\n| Start or debug the no-code Wave GUI, Docker app, remote proxy, app workdir, settings, keyring, or browser\u002Fserver lifecycle | [app-and-ui](sub-skills\u002Fapp-and-ui\u002FSKILL.md) |\n| Build or validate YAML configs, choose problem types, map dataset columns, inspect data schemas, or round-trip config files | [configuration-and-data](sub-skills\u002Fconfiguration-and-data\u002FSKILL.md) |\n| Launch or diagnose `train.py`, CPU\u002FGPU tiny smokes, multi-GPU\u002FDeepSpeed, mixed precision, LoRA, output artifacts, or experiment status | [training-and-experiments](sub-skills\u002Ftraining-and-experiments\u002FSKILL.md) |\n| Understand model wrappers, losses, metrics, prediction files, plots, AI judge metrics, or generation vs forward-pass evaluation behavior | [modeling-and-evaluation](sub-skills\u002Fmodeling-and-evaluation\u002FSKILL.md) |\n| Prompt a trained experiment, preflight saved artifacts, change generation parameters, publish to Hugging Face Hub, or hand off to h2oGPT | [export-and-prompt](sub-skills\u002Fexport-and-prompt\u002FSKILL.md) |\n\n## Install and quick verification\n\nFrom a user's own H2O LLM Studio runtime root, use the repo-supported Python 3.10 setup path before app or training work:\n\n```bash\nmake setup\npython scripts\u002Fcheck_environment.py --runtime-root . --check-config-assets\n```\n\nWhen `make setup` is not appropriate, create an isolated Python 3.10 environment and install the same locked\u002Fruntime dependency set described in [runtime setup](references\u002Fruntime-setup.md); do not modify a user-owned environment without approval.\n\n## Fast runtime facts\n\n- Package metadata expects Python 3.10 and a GPU-oriented PyTorch\u002FTransformers stack.\n- The documented runtime is usually a source-layout H2O LLM Studio runtime root or Docker image, not a tiny import-only library. Commands that refer to `llm_studio\u002F...` should be run from the user's own runtime root or translated to an equivalent module\u002Fwrapper command.\n- Production fine-tuning is NVIDIA-GPU oriented. CPU-style unit\u002Fintegration configs can validate mechanics, but they do not prove real large-model throughput or DeepSpeed\u002Fmulti-GPU behavior.\n- The app and config layers use runtime assets such as `prompts\u002F`, `model_cards\u002F`, `static\u002F`, and `pyproject.toml`; importing `llm_studio` alone is not enough to prove a working app\u002Ftraining root.\n- Optional or external services include Hugging Face Hub\u002Fmodel downloads, Kaggle\u002FHF datasets, OpenAI-compatible judge metrics, W&B logging, S3\u002FAzure\u002FH2O Drive data connectors, Docker\u002FNVIDIA container runtime, and browser UI tests.\n\n## Safe first checks\n\n```bash\npython scripts\u002Fcheck_environment.py --runtime-root . --check-cuda --check-config-assets\npython sub-skills\u002Fconfiguration-and-data\u002Fscripts\u002Finspect_config.py --help\npython sub-skills\u002Ftraining-and-experiments\u002Fscripts\u002Fcheck_training_environment.py --help\npython sub-skills\u002Fexport-and-prompt\u002Fscripts\u002Fcheck_experiment_artifacts.py --help\n```\n\nThese helpers diagnose environment\u002Fconfig\u002Fartifact readiness. They do not download models, start Wave, run training, call OpenAI, publish to Hugging Face, or mutate experiment outputs unless an option explicitly says so.\n\n## When to stop and ask\n\nAsk before running long training, downloading datasets\u002Fmodels, starting a public-facing Wave server, using API\u002FHF\u002FW&B\u002Fcloud credentials, publishing to Hugging Face Hub, modifying a user-owned Python environment, changing Docker\u002FNVIDIA host setup, or overwriting an experiment output directory.\n",{"name":5,"license":24,"metadata":25,"description":7,"disable-model-invocation":27},"Apache 2.0",{"disco-role":26},"operating",true,"main","d69096e239f2a89aabc2cd078c47664a9d3b1e11","06-前端与设计\u002Fh2o-llmstudio",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":37,"prompt":47,"zipAvailable":27},[38,41,44],{"tool":39,"command":40},"npx","npx skills add lza6\u002FArmor-piercing-system-prompts --skill h2o-llmstudio",{"tool":42,"command":43},"bunx","bunx skills add lza6\u002FArmor-piercing-system-prompts --skill h2o-llmstudio",{"tool":45,"command":46},"pnpm","pnpm dlx skills add lza6\u002FArmor-piercing-system-prompts --skill h2o-llmstudio","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F06-前端与设计\u002Fh2o-llmstudio\u002FSKILL.md",[49,55,59,63],{"path":50,"resourceType":51,"resourceTypeDesc":52,"sizeBytes":53,"sourceHtmlUrl":54},"references\u002Frepo-provenance.md","REFERENCE","参考文档",1942,"https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F06-前端与设计\u002Fh2o-llmstudio\u002Freferences\u002Frepo-provenance.md",{"path":56,"resourceType":51,"resourceTypeDesc":52,"sizeBytes":57,"sourceHtmlUrl":58},"references\u002Frepo-routing-metadata.json",539,"https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F06-前端与设计\u002Fh2o-llmstudio\u002Freferences\u002Frepo-routing-metadata.json",{"path":60,"resourceType":51,"resourceTypeDesc":52,"sizeBytes":61,"sourceHtmlUrl":62},"references\u002Fruntime-setup.md",3926,"https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F06-前端与设计\u002Fh2o-llmstudio\u002Freferences\u002Fruntime-setup.md",{"path":64,"resourceType":51,"resourceTypeDesc":52,"sizeBytes":65,"sourceHtmlUrl":66},"references\u002Ftroubleshooting.md",4154,"https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F06-前端与设计\u002Fh2o-llmstudio\u002Freferences\u002Ftroubleshooting.md",[68,76,81,88,95,102],{"skillNo":69,"name":70,"nameZh":6,"description":71,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":72,"licenseFileUrl":6,"sourceHtmlUrl":73,"categories":74,"keywords":75,"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.","MIT","https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F07-后端与架构\u002Fobsidian-clipper-template-creator\u002FSKILL.md",[],[],{"skillNo":77,"name":70,"nameZh":6,"description":71,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":72,"licenseFileUrl":6,"sourceHtmlUrl":78,"categories":79,"keywords":80,"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":82,"name":83,"nameZh":6,"description":84,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":72,"licenseFileUrl":6,"sourceHtmlUrl":85,"categories":86,"keywords":87,"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":89,"name":90,"nameZh":6,"description":91,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":72,"licenseFileUrl":6,"sourceHtmlUrl":92,"categories":93,"keywords":94,"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":96,"name":97,"nameZh":6,"description":98,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":72,"licenseFileUrl":6,"sourceHtmlUrl":99,"categories":100,"keywords":101,"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":103,"name":104,"nameZh":6,"description":105,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":72,"licenseFileUrl":6,"sourceHtmlUrl":106,"categories":107,"keywords":108,"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",[],[]]