[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill-nvidia-mcore-migrate-gpt-to-hybrid":3,"mdc-stldst-key":34,"related-repo-nvidia-mcore-migrate-gpt-to-hybrid":164,"related-org-nvidia-mcore-migrate-gpt-to-hybrid":252},{"slug":4,"name":4,"fn":5,"description":6,"org":7,"tags":11,"stars":20,"repoUrl":21,"updatedAt":22,"license":23,"forks":24,"topics":25,"repo":29,"sourceUrl":32,"mdContent":33},"mcore-migrate-gpt-to-hybrid","migrate Megatron-LM models to HybridModel","Migration guide for moving Megatron Core GPTModel checkpoints, model providers, training commands, and layer mappings to HybridModel.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},"nvidia","NVIDIA","https:\u002F\u002Fpexgzepcugksgbtrxkhf.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Forg-logos\u002Fnvidia.png",[12,16,19],{"name":13,"slug":14,"type":15},"Machine Learning","machine-learning","tag",{"name":17,"slug":18,"type":15},"Migration","migration",{"name":9,"slug":8,"type":15},17049,"https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FMegatron-LM","2026-07-17T06:07:11.777011","Apache-2.0",4230,[26,27,28],"large-language-models","model-para","transformers",{"repoUrl":21,"stars":20,"forks":24,"topics":30,"description":31},[26,27,28],"Ongoing research training transformer models at scale","https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FMegatron-LM\u002Ftree\u002FHEAD\u002Fskills\u002Fmcore-migrate-gpt-to-hybrid","---\nname: mcore-migrate-gpt-to-hybrid\ndescription: Migration guide for moving Megatron Core GPTModel checkpoints, model providers, training commands, and layer mappings to HybridModel.\nlicense: Apache-2.0\nwhen_to_use: Migrating or reviewing a GPTModel checkpoint or training workflow for HybridModel; choosing or reviewing a hybrid layer pattern; running gpt_hybrid_conversion.py; loading a converted checkpoint; diagnosing GPT-to-Hybrid migration issues; 'migrate GPTModel to HybridModel', 'convert GPT checkpoint to HybridModel', 'hybrid layer pattern'.\nmetadata:\n  author: Philip Petrakian \u003Cppetrakian@nvidia.com>\n---\n\n# GPTModel to HybridModel Migration\n\n## Answer-First Migration Guidance\n\n- The canonical source is\n  [`docs\u002Fuser-guide\u002Fhybrid-model-migration.md`](..\u002F..\u002Fdocs\u002Fuser-guide\u002Fhybrid-model-migration.md).\n- Read the canonical document completely before answering, planning, reviewing,\n  editing, converting, or training.\n- Keep migration behavior, commands, mappings, prerequisites, limitations, and\n  validation in the canonical document only. Do not duplicate them in this\n  skill.\n\n---\n\n## Workflow\n\n1. Pull the task artifact first: checkpoint metadata, model provider or config,\n   training command, conversion log, diff, or failure output.\n2. Read the canonical migration document completely.\n3. Follow only the relevant document sections. Do not invent an unsupported\n   migration path or silently change the target architecture.\n4. Validate the result proportionately, invoking the relevant repository build\n   and testing skills when applicable.\n5. Report the outcome and link the canonical document for human readers.\n\n---\n\n## Documentation Drift\n\nIf the implementation and migration guide disagree:\n\n1. Report the discrepancy before continuing.\n2. If the task authorizes a correction, update the canonical document first.\n3. Do not add a competing migration rule to this skill.\n",{"data":35,"body":39},{"name":4,"description":6,"license":23,"when_to_use":36,"metadata":37},"Migrating or reviewing a GPTModel checkpoint or training workflow for HybridModel; choosing or reviewing a hybrid layer pattern; running gpt_hybrid_conversion.py; loading a converted checkpoint; diagnosing GPT-to-Hybrid migration issues; 'migrate GPTModel to HybridModel', 'convert GPT checkpoint to HybridModel', 'hybrid layer pattern'.",{"author":38},"Philip Petrakian \u003Cppetrakian@nvidia.com>",{"type":40,"children":41},"root",[42,51,58,92,96,102,131,134,140,146],{"type":43,"tag":44,"props":45,"children":47},"element","h1",{"id":46},"gptmodel-to-hybridmodel-migration",[48],{"type":49,"value":50},"text","GPTModel to HybridModel Migration",{"type":43,"tag":52,"props":53,"children":55},"h2",{"id":54},"answer-first-migration-guidance",[56],{"type":49,"value":57},"Answer-First Migration Guidance",{"type":43,"tag":59,"props":60,"children":61},"ul",{},[62,82,87],{"type":43,"tag":63,"props":64,"children":65},"li",{},[66,68,80],{"type":49,"value":67},"The canonical source is\n",{"type":43,"tag":69,"props":70,"children":72},"a",{"href":71},"..\u002F..\u002Fdocs\u002Fuser-guide\u002Fhybrid-model-migration.md",[73],{"type":43,"tag":74,"props":75,"children":77},"code",{"className":76},[],[78],{"type":49,"value":79},"docs\u002Fuser-guide\u002Fhybrid-model-migration.md",{"type":49,"value":81},".",{"type":43,"tag":63,"props":83,"children":84},{},[85],{"type":49,"value":86},"Read the canonical document completely before answering, planning, reviewing,\nediting, converting, or training.",{"type":43,"tag":63,"props":88,"children":89},{},[90],{"type":49,"value":91},"Keep migration behavior, commands, mappings, prerequisites, limitations, and\nvalidation in the canonical document only. Do not duplicate them in this\nskill.",{"type":43,"tag":93,"props":94,"children":95},"hr",{},[],{"type":43,"tag":52,"props":97,"children":99},{"id":98},"workflow",[100],{"type":49,"value":101},"Workflow",{"type":43,"tag":103,"props":104,"children":105},"ol",{},[106,111,116,121,126],{"type":43,"tag":63,"props":107,"children":108},{},[109],{"type":49,"value":110},"Pull the task artifact first: checkpoint metadata, model provider or config,\ntraining command, conversion log, diff, or failure output.",{"type":43,"tag":63,"props":112,"children":113},{},[114],{"type":49,"value":115},"Read the canonical migration document completely.",{"type":43,"tag":63,"props":117,"children":118},{},[119],{"type":49,"value":120},"Follow only the relevant document sections. Do not invent an unsupported\nmigration path or silently change the target architecture.",{"type":43,"tag":63,"props":122,"children":123},{},[124],{"type":49,"value":125},"Validate the result proportionately, invoking the relevant repository build\nand testing skills when applicable.",{"type":43,"tag":63,"props":127,"children":128},{},[129],{"type":49,"value":130},"Report the outcome and link the canonical document for human readers.",{"type":43,"tag":93,"props":132,"children":133},{},[],{"type":43,"tag":52,"props":135,"children":137},{"id":136},"documentation-drift",[138],{"type":49,"value":139},"Documentation Drift",{"type":43,"tag":141,"props":142,"children":143},"p",{},[144],{"type":49,"value":145},"If the implementation and migration guide disagree:",{"type":43,"tag":103,"props":147,"children":148},{},[149,154,159],{"type":43,"tag":63,"props":150,"children":151},{},[152],{"type":49,"value":153},"Report the discrepancy before continuing.",{"type":43,"tag":63,"props":155,"children":156},{},[157],{"type":49,"value":158},"If the task authorizes a correction, update the canonical document first.",{"type":43,"tag":63,"props":160,"children":161},{},[162],{"type":49,"value":163},"Do not add a competing migration rule to this skill.",{"items":165,"total":251},[166,182,193,205,219,232,238],{"slug":167,"name":167,"fn":168,"description":169,"org":170,"tags":171,"stars":20,"repoUrl":21,"updatedAt":181},"mcore-build-and-dependency","manage Megatron-LM development environments","Container-based dev environment setup and dependency management for Megatron-LM. Covers acquiring and launching the CI container, uv package management, and updating uv.lock.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[172,175,178],{"name":173,"slug":174,"type":15},"Containers","containers",{"name":176,"slug":177,"type":15},"Deployment","deployment",{"name":179,"slug":180,"type":15},"Python","python","2026-07-27T06:06:11.249662",{"slug":183,"name":183,"fn":184,"description":185,"org":186,"tags":187,"stars":20,"repoUrl":21,"updatedAt":192},"mcore-bump-base-image","update NVIDIA PyTorch base images","Bump the NVIDIA PyTorch base image (`nvcr.io\u002Fnvidia\u002Fpytorch:YY.MM-py3`) used by Megatron-LM CI. Covers the two pin sites (GitHub CI in `docker\u002F.ngc_version.dev` and GitLab CI in `.gitlab\u002Fstages\u002F01.build.yml`), the post-bump CI loop (re-run functional tests, refresh golden values, mark broken tests), and the gotchas that bit PRs",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[188,191],{"name":189,"slug":190,"type":15},"CI\u002FCD","ci-cd",{"name":176,"slug":177,"type":15},"2026-07-14T05:25:59.97109",{"slug":194,"name":194,"fn":195,"description":196,"org":197,"tags":198,"stars":20,"repoUrl":21,"updatedAt":204},"mcore-cicd","manage CI\u002FCD pipelines for Megatron-LM","CI\u002FCD reference for Megatron-LM. Covers CI pipeline structure, PR scope labels, triggering internal GitLab CI (which force-pushes the current branch to a pull-request\u002FBRANCH ref — always dry-run and verify the destination first; never run against shared or protected branches), and CI failure investigation.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[199,200,201],{"name":189,"slug":190,"type":15},{"name":176,"slug":177,"type":15},{"name":202,"slug":203,"type":15},"GitHub","github","2026-07-27T06:06:12.278222",{"slug":206,"name":206,"fn":207,"description":208,"org":209,"tags":210,"stars":20,"repoUrl":21,"updatedAt":218},"mcore-create-issue","investigate CI failures and create issues","Investigate a failing GitHub Actions run or job and create a GitHub issue for the failure.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[211,214,215],{"name":212,"slug":213,"type":15},"Debugging","debugging",{"name":202,"slug":203,"type":15},{"name":216,"slug":217,"type":15},"Triage","triage","2026-07-14T05:25:57.442089",{"slug":220,"name":220,"fn":221,"description":222,"org":223,"tags":224,"stars":20,"repoUrl":21,"updatedAt":231},"mcore-linting-and-formatting","lint and format Megatron-LM code","Linting and formatting for Megatron-LM. Covers running autoformat.sh, tools (ruff, black, isort, pylint, mypy), and code style rules.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[225,228],{"name":226,"slug":227,"type":15},"Best Practices","best-practices",{"name":229,"slug":230,"type":15},"Code Analysis","code-analysis","2026-07-14T05:25:56.18433",{"slug":4,"name":4,"fn":5,"description":6,"org":233,"tags":234,"stars":20,"repoUrl":21,"updatedAt":22},{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[235,236,237],{"name":13,"slug":14,"type":15},{"name":17,"slug":18,"type":15},{"name":9,"slug":8,"type":15},{"slug":239,"name":239,"fn":240,"description":241,"org":242,"tags":243,"stars":20,"repoUrl":21,"updatedAt":250},"mcore-onboard-gb200-1node-tests","onboard functional tests for GB200","Onboard 1-node GitHub MR functional tests for GB200 from existing mr-scoped 2-node tests.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[244,247],{"name":245,"slug":246,"type":15},"QA","qa",{"name":248,"slug":249,"type":15},"Testing","testing","2026-07-14T05:25:53.673039",13,{"items":253,"total":356},[254,272,278,283,289,295,300,306,311,322,336,345],{"slug":255,"name":255,"fn":256,"description":257,"org":258,"tags":259,"stars":269,"repoUrl":270,"updatedAt":271},"nemoclaw-user-guide","retrieve NemoClaw documentation and configuration","Guides human users' AI agents to the NemoClaw docs MCP server and canonical Fern documentation in Markdown form. Use when users ask how to install, configure, operate, troubleshoot, secure, or learn NemoClaw with an AI coding assistant. Trigger keywords - nemoclaw docs, use nemoclaw with ai agent, nemoclaw mcp docs, nemoclaw install help, nemoclaw quickstart, nemoclaw markdown docs, llms.txt, agent skills.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[260,263,266],{"name":261,"slug":262,"type":15},"Documentation","documentation",{"name":264,"slug":265,"type":15},"MCP","mcp",{"name":267,"slug":268,"type":15},"Search","search",21777,"https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FNemoClaw","2026-07-20T06:00:01.461044",{"slug":167,"name":167,"fn":168,"description":169,"org":273,"tags":274,"stars":20,"repoUrl":21,"updatedAt":181},{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[275,276,277],{"name":173,"slug":174,"type":15},{"name":176,"slug":177,"type":15},{"name":179,"slug":180,"type":15},{"slug":183,"name":183,"fn":184,"description":185,"org":279,"tags":280,"stars":20,"repoUrl":21,"updatedAt":192},{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[281,282],{"name":189,"slug":190,"type":15},{"name":176,"slug":177,"type":15},{"slug":194,"name":194,"fn":195,"description":196,"org":284,"tags":285,"stars":20,"repoUrl":21,"updatedAt":204},{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[286,287,288],{"name":189,"slug":190,"type":15},{"name":176,"slug":177,"type":15},{"name":202,"slug":203,"type":15},{"slug":206,"name":206,"fn":207,"description":208,"org":290,"tags":291,"stars":20,"repoUrl":21,"updatedAt":218},{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[292,293,294],{"name":212,"slug":213,"type":15},{"name":202,"slug":203,"type":15},{"name":216,"slug":217,"type":15},{"slug":220,"name":220,"fn":221,"description":222,"org":296,"tags":297,"stars":20,"repoUrl":21,"updatedAt":231},{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[298,299],{"name":226,"slug":227,"type":15},{"name":229,"slug":230,"type":15},{"slug":4,"name":4,"fn":5,"description":6,"org":301,"tags":302,"stars":20,"repoUrl":21,"updatedAt":22},{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[303,304,305],{"name":13,"slug":14,"type":15},{"name":17,"slug":18,"type":15},{"name":9,"slug":8,"type":15},{"slug":239,"name":239,"fn":240,"description":241,"org":307,"tags":308,"stars":20,"repoUrl":21,"updatedAt":250},{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[309,310],{"name":245,"slug":246,"type":15},{"name":248,"slug":249,"type":15},{"slug":312,"name":312,"fn":313,"description":314,"org":315,"tags":316,"stars":20,"repoUrl":21,"updatedAt":321},"mcore-run-on-slurm","launch distributed training jobs on SLURM","How to launch distributed Megatron-LM training jobs on a SLURM cluster. Covers a minimal sbatch skeleton, environment-variable setup for torch.distributed.run, CUDA_DEVICE_MAX_CONNECTIONS rules across hardware and parallelism modes, container conventions, monitoring, and per-rank failure diagnosis.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[317,318],{"name":176,"slug":177,"type":15},{"name":319,"slug":320,"type":15},"Infrastructure","infrastructure","2026-07-14T05:25:49.362534",{"slug":323,"name":323,"fn":324,"description":325,"org":326,"tags":327,"stars":20,"repoUrl":21,"updatedAt":335},"mcore-split-pr","split pull requests to reduce review load","Split a PR into multiple PRs to reduce the number of required CODEOWNERS reviewer groups.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[328,331,332],{"name":329,"slug":330,"type":15},"Code Review","code-review",{"name":202,"slug":203,"type":15},{"name":333,"slug":334,"type":15},"Pull Requests","pull-requests","2026-07-14T05:26:01.226578",{"slug":337,"name":337,"fn":338,"description":339,"org":340,"tags":341,"stars":20,"repoUrl":21,"updatedAt":344},"mcore-testing","run and manage Megatron-LM tests","Test system for Megatron-LM. Covers test layout, recipe YAML structure, adding and running unit and functional tests, golden values, marker filters, and CI parity.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[342,343],{"name":245,"slug":246,"type":15},{"name":248,"slug":249,"type":15},"2026-07-14T05:25:54.928983",{"slug":346,"name":346,"fn":347,"description":348,"org":349,"tags":350,"stars":20,"repoUrl":21,"updatedAt":355},"nightly-sync","manage nightly main-to-dev sync workflows","Domain knowledge for the nightly main-to-dev sync workflow. Covers merge strategy, CI architecture, failure investigation, and known issues.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[351,354],{"name":352,"slug":353,"type":15},"Automation","automation",{"name":189,"slug":190,"type":15},"2026-07-30T05:29:03.275638",496]