[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill-nvidia-cuopt-install":3,"mdc-kv9xei-key":34,"related-repo-nvidia-cuopt-install":1097,"related-org-nvidia-cuopt-install":1199},{"slug":4,"name":4,"fn":5,"description":6,"org":7,"tags":11,"stars":23,"repoUrl":24,"updatedAt":25,"license":26,"forks":27,"topics":28,"repo":29,"sourceUrl":32,"mdContent":33},"cuopt-install","install and verify cuOpt","Install cuOpt for Python, C, or server via pip, conda, or Docker; verify the install. For building cuOpt from source, see cuopt-developer.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},"nvidia","NVIDIA","https:\u002F\u002Fpexgzepcugksgbtrxkhf.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Forg-logos\u002Fnvidia.png",[12,16,19,20],{"name":13,"slug":14,"type":15},"Optimization","optimization","tag",{"name":17,"slug":18,"type":15},"Deployment","deployment",{"name":9,"slug":8,"type":15},{"name":21,"slug":22,"type":15},"Python","python",2473,"https:\u002F\u002Fgithub.com\u002FNVIDIA\u002Fskills","2026-07-14T05:27:21.724705","Apache-2.0",281,[],{"repoUrl":24,"stars":23,"forks":27,"topics":30,"description":31},[],"AI agent skills published by NVIDIA","https:\u002F\u002Fgithub.com\u002FNVIDIA\u002Fskills\u002Ftree\u002FHEAD\u002Fskills\u002Fcuopt-install","---\nname: cuopt-install\nversion: \"26.08.00\"\ndescription: Install cuOpt for Python, C, or server via pip, conda, or Docker; verify the install. For building cuOpt from source, see cuopt-developer.\nlicense: Apache-2.0\nmetadata:\n  author: NVIDIA cuOpt Team\n  tags:\n    - cuopt\n    - install\n    - deployment\n    - python\n    - server\n---\n\n\n# cuOpt Install (user)\n\nInstall cuOpt to *use* it from Python, C, or as a REST server. For building cuOpt from source to contribute or modify it, see `cuopt-developer`.\n\n## System requirements\n\n- **GPU**: NVIDIA Compute Capability ≥ 7.0 (Volta or newer). Examples: V100, A100, H100, RTX 20xx\u002F30xx\u002F40xx. Not supported: GTX 10xx (Pascal).\n- **CUDA**: 12.x or 13.x. The package CUDA suffix must match the runtime CUDA (e.g. `cuopt-cu12` \u002F `libcuopt-cu12` with CUDA 12).\n- **Driver**: NVIDIA driver compatible with the CUDA version.\n- `cuopt-cuXX` (Python) depends on `libcuopt-cuXX` (C), so installing the Python package also installs the C library and headers. Installing `libcuopt-cuXX` on its own does **not** install the Python API.\n\n## Required questions\n\nAsk these if not already clear:\n\n1. **Interface** — Python, C, or REST server? Server can be called from any language via HTTP.\n2. **CUDA version** — What is installed? Check with `nvcc --version` or `nvidia-smi`.\n3. **Package manager** — pip, conda, or Docker preferred?\n4. **Environment** — Local machine with GPU, cloud instance, Docker\u002FKubernetes, or remote\u002Fserver (no local GPU)?\n\n## Python API\n\n**Choose one** — do not run both. The second install would override the first and can cause CUDA \u002F package mismatch.\n\n### pip\n\n- **CUDA 13.x:**\n  ```bash\n  pip install --extra-index-url=https:\u002F\u002Fpypi.nvidia.com cuopt-cu13\n  ```\n- **CUDA 12.x:**\n  ```bash\n  pip install --extra-index-url=https:\u002F\u002Fpypi.nvidia.com 'cuopt-cu12==26.2.*'\n  ```\n\n### conda\n\n```bash\nconda install -c rapidsai -c conda-forge -c nvidia cuopt\n```\n\n### Verify\n\n```python\nimport cuopt\nprint(cuopt.__version__)\nfrom cuopt import routing\ndm = routing.DataModel(n_locations=3, n_fleet=1, n_orders=2)\n```\n\n## C API\n\nThe C API ships in `libcuopt-cuXX`, which is also pulled in as a dependency of `cuopt-cuXX` — so if you already installed the Python package, the C library and headers are already present. Install `libcuopt` standalone only when you want the C API without Python. **Choose one** of pip or conda — do not run both.\n\n### pip\n\n- **CUDA 13.x:**\n  ```bash\n  pip install --extra-index-url=https:\u002F\u002Fpypi.nvidia.com libcuopt-cu13\n  ```\n- **CUDA 12.x:**\n  ```bash\n  pip install --extra-index-url=https:\u002F\u002Fpypi.nvidia.com 'libcuopt-cu12==26.2.*'\n  ```\n\n### conda\n\n```bash\nconda install -c rapidsai -c conda-forge -c nvidia libcuopt\n```\n\n### Verify\n\nSee [`references\u002Fverification_examples.md`](references\u002Fverification_examples.md)\nfor the canonical C-API header\u002Flibrary `find` commands (conda and pip\u002Fvenv variants).\n\n## Server (REST)\n\n### pip\n\n```bash\npip install --extra-index-url=https:\u002F\u002Fpypi.nvidia.com cuopt-server-cu12 cuopt-sh-client\n```\n\n### conda\n\n```bash\nconda install -c rapidsai -c conda-forge -c nvidia cuopt-server cuopt-sh-client\n```\n\n### Docker\n\n```bash\ndocker pull nvidia\u002Fcuopt:latest-cuda12.9-py3.13\ndocker run --gpus all -it --rm -p 8000:8000 nvidia\u002Fcuopt:latest-cuda12.9-py3.13\n```\n\n### Verify\n\n```bash\npython -m cuopt_server.cuopt_service --ip 0.0.0.0 --port 8000 &\nsleep 5\ncurl -s http:\u002F\u002Flocalhost:8000\u002Fcuopt\u002Fhealth | jq .\n```\n\n## Common Issues\n\n- `No module named 'cuopt'` → check `pip list | grep cuopt`, `which python`, reinstall with the correct extra-index-url.\n- CUDA not available → run `nvidia-smi` and `nvcc --version`; ensure the package CUDA suffix (`cu12` vs `cu13`) matches the installed CUDA.\n- Python vs C → `cuopt-cuXX` pulls in `libcuopt-cuXX` as a transitive dependency, so the C library (`libcuopt.so`) and headers (`cuopt_c.h`) are already available after installing the Python package. The reverse is **not** true: `libcuopt-cuXX` alone does not install the Python bindings.\n\n## See also\n\n- [verification_examples.md](references\u002Fverification_examples.md) — full verification recipes for Python, C, server, and Docker.\n- `cuopt-developer` — build cuOpt from source and contribute to the codebase.\n",{"data":35,"body":43},{"name":4,"version":36,"description":6,"license":26,"metadata":37},"26.08.00",{"author":38,"tags":39},"NVIDIA cuOpt Team",[40,41,18,22,42],"cuopt","install","server",{"type":44,"children":45},"root",[46,55,78,85,170,176,181,240,246,256,262,357,362,413,419,462,468,501,506,585,590,637,642,667,673,678,710,715,766,772,842,847,943,949,1062,1068,1091],{"type":47,"tag":48,"props":49,"children":51},"element","h1",{"id":50},"cuopt-install-user",[52],{"type":53,"value":54},"text","cuOpt Install (user)",{"type":47,"tag":56,"props":57,"children":58},"p",{},[59,61,67,69,76],{"type":53,"value":60},"Install cuOpt to ",{"type":47,"tag":62,"props":63,"children":64},"em",{},[65],{"type":53,"value":66},"use",{"type":53,"value":68}," it from Python, C, or as a REST server. For building cuOpt from source to contribute or modify it, see ",{"type":47,"tag":70,"props":71,"children":73},"code",{"className":72},[],[74],{"type":53,"value":75},"cuopt-developer",{"type":53,"value":77},".",{"type":47,"tag":79,"props":80,"children":82},"h2",{"id":81},"system-requirements",[83],{"type":53,"value":84},"System requirements",{"type":47,"tag":86,"props":87,"children":88},"ul",{},[89,101,127,137],{"type":47,"tag":90,"props":91,"children":92},"li",{},[93,99],{"type":47,"tag":94,"props":95,"children":96},"strong",{},[97],{"type":53,"value":98},"GPU",{"type":53,"value":100},": NVIDIA Compute Capability ≥ 7.0 (Volta or newer). Examples: V100, A100, H100, RTX 20xx\u002F30xx\u002F40xx. Not supported: GTX 10xx (Pascal).",{"type":47,"tag":90,"props":102,"children":103},{},[104,109,111,117,119,125],{"type":47,"tag":94,"props":105,"children":106},{},[107],{"type":53,"value":108},"CUDA",{"type":53,"value":110},": 12.x or 13.x. The package CUDA suffix must match the runtime CUDA (e.g. ",{"type":47,"tag":70,"props":112,"children":114},{"className":113},[],[115],{"type":53,"value":116},"cuopt-cu12",{"type":53,"value":118}," \u002F ",{"type":47,"tag":70,"props":120,"children":122},{"className":121},[],[123],{"type":53,"value":124},"libcuopt-cu12",{"type":53,"value":126}," with CUDA 12).",{"type":47,"tag":90,"props":128,"children":129},{},[130,135],{"type":47,"tag":94,"props":131,"children":132},{},[133],{"type":53,"value":134},"Driver",{"type":53,"value":136},": NVIDIA driver compatible with the CUDA version.",{"type":47,"tag":90,"props":138,"children":139},{},[140,146,148,154,156,161,163,168],{"type":47,"tag":70,"props":141,"children":143},{"className":142},[],[144],{"type":53,"value":145},"cuopt-cuXX",{"type":53,"value":147}," (Python) depends on ",{"type":47,"tag":70,"props":149,"children":151},{"className":150},[],[152],{"type":53,"value":153},"libcuopt-cuXX",{"type":53,"value":155}," (C), so installing the Python package also installs the C library and headers. Installing ",{"type":47,"tag":70,"props":157,"children":159},{"className":158},[],[160],{"type":53,"value":153},{"type":53,"value":162}," on its own does ",{"type":47,"tag":94,"props":164,"children":165},{},[166],{"type":53,"value":167},"not",{"type":53,"value":169}," install the Python API.",{"type":47,"tag":79,"props":171,"children":173},{"id":172},"required-questions",[174],{"type":53,"value":175},"Required questions",{"type":47,"tag":56,"props":177,"children":178},{},[179],{"type":53,"value":180},"Ask these if not already clear:",{"type":47,"tag":182,"props":183,"children":184},"ol",{},[185,195,220,230],{"type":47,"tag":90,"props":186,"children":187},{},[188,193],{"type":47,"tag":94,"props":189,"children":190},{},[191],{"type":53,"value":192},"Interface",{"type":53,"value":194}," — Python, C, or REST server? Server can be called from any language via HTTP.",{"type":47,"tag":90,"props":196,"children":197},{},[198,203,205,211,213,219],{"type":47,"tag":94,"props":199,"children":200},{},[201],{"type":53,"value":202},"CUDA version",{"type":53,"value":204}," — What is installed? Check with ",{"type":47,"tag":70,"props":206,"children":208},{"className":207},[],[209],{"type":53,"value":210},"nvcc --version",{"type":53,"value":212}," or ",{"type":47,"tag":70,"props":214,"children":216},{"className":215},[],[217],{"type":53,"value":218},"nvidia-smi",{"type":53,"value":77},{"type":47,"tag":90,"props":221,"children":222},{},[223,228],{"type":47,"tag":94,"props":224,"children":225},{},[226],{"type":53,"value":227},"Package manager",{"type":53,"value":229}," — pip, conda, or Docker preferred?",{"type":47,"tag":90,"props":231,"children":232},{},[233,238],{"type":47,"tag":94,"props":234,"children":235},{},[236],{"type":53,"value":237},"Environment",{"type":53,"value":239}," — Local machine with GPU, cloud instance, Docker\u002FKubernetes, or remote\u002Fserver (no local GPU)?",{"type":47,"tag":79,"props":241,"children":243},{"id":242},"python-api",[244],{"type":53,"value":245},"Python API",{"type":47,"tag":56,"props":247,"children":248},{},[249,254],{"type":47,"tag":94,"props":250,"children":251},{},[252],{"type":53,"value":253},"Choose one",{"type":53,"value":255}," — do not run both. The second install would override the first and can cause CUDA \u002F package mismatch.",{"type":47,"tag":257,"props":258,"children":260},"h3",{"id":259},"pip",[261],{"type":53,"value":259},{"type":47,"tag":86,"props":263,"children":264},{},[265,311],{"type":47,"tag":90,"props":266,"children":267},{},[268,273],{"type":47,"tag":94,"props":269,"children":270},{},[271],{"type":53,"value":272},"CUDA 13.x:",{"type":47,"tag":274,"props":275,"children":280},"pre",{"className":276,"code":277,"language":278,"meta":279,"style":279},"language-bash shiki shiki-themes material-theme-lighter material-theme material-theme-palenight","pip install --extra-index-url=https:\u002F\u002Fpypi.nvidia.com cuopt-cu13\n","bash","",[281],{"type":47,"tag":70,"props":282,"children":283},{"__ignoreMap":279},[284],{"type":47,"tag":285,"props":286,"children":289},"span",{"class":287,"line":288},"line",1,[290,295,301,306],{"type":47,"tag":285,"props":291,"children":293},{"style":292},"--shiki-light:#E2931D;--shiki-default:#FFCB6B;--shiki-dark:#FFCB6B",[294],{"type":53,"value":259},{"type":47,"tag":285,"props":296,"children":298},{"style":297},"--shiki-light:#91B859;--shiki-default:#C3E88D;--shiki-dark:#C3E88D",[299],{"type":53,"value":300}," install",{"type":47,"tag":285,"props":302,"children":303},{"style":297},[304],{"type":53,"value":305}," --extra-index-url=https:\u002F\u002Fpypi.nvidia.com",{"type":47,"tag":285,"props":307,"children":308},{"style":297},[309],{"type":53,"value":310}," cuopt-cu13\n",{"type":47,"tag":90,"props":312,"children":313},{},[314,319],{"type":47,"tag":94,"props":315,"children":316},{},[317],{"type":53,"value":318},"CUDA 12.x:",{"type":47,"tag":274,"props":320,"children":322},{"className":276,"code":321,"language":278,"meta":279,"style":279},"pip install --extra-index-url=https:\u002F\u002Fpypi.nvidia.com 'cuopt-cu12==26.2.*'\n",[323],{"type":47,"tag":70,"props":324,"children":325},{"__ignoreMap":279},[326],{"type":47,"tag":285,"props":327,"children":328},{"class":287,"line":288},[329,333,337,341,347,352],{"type":47,"tag":285,"props":330,"children":331},{"style":292},[332],{"type":53,"value":259},{"type":47,"tag":285,"props":334,"children":335},{"style":297},[336],{"type":53,"value":300},{"type":47,"tag":285,"props":338,"children":339},{"style":297},[340],{"type":53,"value":305},{"type":47,"tag":285,"props":342,"children":344},{"style":343},"--shiki-light:#39ADB5;--shiki-default:#89DDFF;--shiki-dark:#89DDFF",[345],{"type":53,"value":346}," '",{"type":47,"tag":285,"props":348,"children":349},{"style":297},[350],{"type":53,"value":351},"cuopt-cu12==26.2.*",{"type":47,"tag":285,"props":353,"children":354},{"style":343},[355],{"type":53,"value":356},"'\n",{"type":47,"tag":257,"props":358,"children":360},{"id":359},"conda",[361],{"type":53,"value":359},{"type":47,"tag":274,"props":363,"children":365},{"className":276,"code":364,"language":278,"meta":279,"style":279},"conda install -c rapidsai -c conda-forge -c nvidia cuopt\n",[366],{"type":47,"tag":70,"props":367,"children":368},{"__ignoreMap":279},[369],{"type":47,"tag":285,"props":370,"children":371},{"class":287,"line":288},[372,376,380,385,390,394,399,403,408],{"type":47,"tag":285,"props":373,"children":374},{"style":292},[375],{"type":53,"value":359},{"type":47,"tag":285,"props":377,"children":378},{"style":297},[379],{"type":53,"value":300},{"type":47,"tag":285,"props":381,"children":382},{"style":297},[383],{"type":53,"value":384}," -c",{"type":47,"tag":285,"props":386,"children":387},{"style":297},[388],{"type":53,"value":389}," rapidsai",{"type":47,"tag":285,"props":391,"children":392},{"style":297},[393],{"type":53,"value":384},{"type":47,"tag":285,"props":395,"children":396},{"style":297},[397],{"type":53,"value":398}," conda-forge",{"type":47,"tag":285,"props":400,"children":401},{"style":297},[402],{"type":53,"value":384},{"type":47,"tag":285,"props":404,"children":405},{"style":297},[406],{"type":53,"value":407}," nvidia",{"type":47,"tag":285,"props":409,"children":410},{"style":297},[411],{"type":53,"value":412}," cuopt\n",{"type":47,"tag":257,"props":414,"children":416},{"id":415},"verify",[417],{"type":53,"value":418},"Verify",{"type":47,"tag":274,"props":420,"children":423},{"className":421,"code":422,"language":22,"meta":279,"style":279},"language-python shiki shiki-themes material-theme-lighter material-theme material-theme-palenight","import cuopt\nprint(cuopt.__version__)\nfrom cuopt import routing\ndm = routing.DataModel(n_locations=3, n_fleet=1, n_orders=2)\n",[424],{"type":47,"tag":70,"props":425,"children":426},{"__ignoreMap":279},[427,435,444,453],{"type":47,"tag":285,"props":428,"children":429},{"class":287,"line":288},[430],{"type":47,"tag":285,"props":431,"children":432},{},[433],{"type":53,"value":434},"import cuopt\n",{"type":47,"tag":285,"props":436,"children":438},{"class":287,"line":437},2,[439],{"type":47,"tag":285,"props":440,"children":441},{},[442],{"type":53,"value":443},"print(cuopt.__version__)\n",{"type":47,"tag":285,"props":445,"children":447},{"class":287,"line":446},3,[448],{"type":47,"tag":285,"props":449,"children":450},{},[451],{"type":53,"value":452},"from cuopt import routing\n",{"type":47,"tag":285,"props":454,"children":456},{"class":287,"line":455},4,[457],{"type":47,"tag":285,"props":458,"children":459},{},[460],{"type":53,"value":461},"dm = routing.DataModel(n_locations=3, n_fleet=1, n_orders=2)\n",{"type":47,"tag":79,"props":463,"children":465},{"id":464},"c-api",[466],{"type":53,"value":467},"C API",{"type":47,"tag":56,"props":469,"children":470},{},[471,473,478,480,485,487,493,495,499],{"type":53,"value":472},"The C API ships in ",{"type":47,"tag":70,"props":474,"children":476},{"className":475},[],[477],{"type":53,"value":153},{"type":53,"value":479},", which is also pulled in as a dependency of ",{"type":47,"tag":70,"props":481,"children":483},{"className":482},[],[484],{"type":53,"value":145},{"type":53,"value":486}," — so if you already installed the Python package, the C library and headers are already present. Install ",{"type":47,"tag":70,"props":488,"children":490},{"className":489},[],[491],{"type":53,"value":492},"libcuopt",{"type":53,"value":494}," standalone only when you want the C API without Python. ",{"type":47,"tag":94,"props":496,"children":497},{},[498],{"type":53,"value":253},{"type":53,"value":500}," of pip or conda — do not run both.",{"type":47,"tag":257,"props":502,"children":504},{"id":503},"pip-1",[505],{"type":53,"value":259},{"type":47,"tag":86,"props":507,"children":508},{},[509,543],{"type":47,"tag":90,"props":510,"children":511},{},[512,516],{"type":47,"tag":94,"props":513,"children":514},{},[515],{"type":53,"value":272},{"type":47,"tag":274,"props":517,"children":519},{"className":276,"code":518,"language":278,"meta":279,"style":279},"pip install --extra-index-url=https:\u002F\u002Fpypi.nvidia.com libcuopt-cu13\n",[520],{"type":47,"tag":70,"props":521,"children":522},{"__ignoreMap":279},[523],{"type":47,"tag":285,"props":524,"children":525},{"class":287,"line":288},[526,530,534,538],{"type":47,"tag":285,"props":527,"children":528},{"style":292},[529],{"type":53,"value":259},{"type":47,"tag":285,"props":531,"children":532},{"style":297},[533],{"type":53,"value":300},{"type":47,"tag":285,"props":535,"children":536},{"style":297},[537],{"type":53,"value":305},{"type":47,"tag":285,"props":539,"children":540},{"style":297},[541],{"type":53,"value":542}," libcuopt-cu13\n",{"type":47,"tag":90,"props":544,"children":545},{},[546,550],{"type":47,"tag":94,"props":547,"children":548},{},[549],{"type":53,"value":318},{"type":47,"tag":274,"props":551,"children":553},{"className":276,"code":552,"language":278,"meta":279,"style":279},"pip install --extra-index-url=https:\u002F\u002Fpypi.nvidia.com 'libcuopt-cu12==26.2.*'\n",[554],{"type":47,"tag":70,"props":555,"children":556},{"__ignoreMap":279},[557],{"type":47,"tag":285,"props":558,"children":559},{"class":287,"line":288},[560,564,568,572,576,581],{"type":47,"tag":285,"props":561,"children":562},{"style":292},[563],{"type":53,"value":259},{"type":47,"tag":285,"props":565,"children":566},{"style":297},[567],{"type":53,"value":300},{"type":47,"tag":285,"props":569,"children":570},{"style":297},[571],{"type":53,"value":305},{"type":47,"tag":285,"props":573,"children":574},{"style":343},[575],{"type":53,"value":346},{"type":47,"tag":285,"props":577,"children":578},{"style":297},[579],{"type":53,"value":580},"libcuopt-cu12==26.2.*",{"type":47,"tag":285,"props":582,"children":583},{"style":343},[584],{"type":53,"value":356},{"type":47,"tag":257,"props":586,"children":588},{"id":587},"conda-1",[589],{"type":53,"value":359},{"type":47,"tag":274,"props":591,"children":593},{"className":276,"code":592,"language":278,"meta":279,"style":279},"conda install -c rapidsai -c conda-forge -c nvidia libcuopt\n",[594],{"type":47,"tag":70,"props":595,"children":596},{"__ignoreMap":279},[597],{"type":47,"tag":285,"props":598,"children":599},{"class":287,"line":288},[600,604,608,612,616,620,624,628,632],{"type":47,"tag":285,"props":601,"children":602},{"style":292},[603],{"type":53,"value":359},{"type":47,"tag":285,"props":605,"children":606},{"style":297},[607],{"type":53,"value":300},{"type":47,"tag":285,"props":609,"children":610},{"style":297},[611],{"type":53,"value":384},{"type":47,"tag":285,"props":613,"children":614},{"style":297},[615],{"type":53,"value":389},{"type":47,"tag":285,"props":617,"children":618},{"style":297},[619],{"type":53,"value":384},{"type":47,"tag":285,"props":621,"children":622},{"style":297},[623],{"type":53,"value":398},{"type":47,"tag":285,"props":625,"children":626},{"style":297},[627],{"type":53,"value":384},{"type":47,"tag":285,"props":629,"children":630},{"style":297},[631],{"type":53,"value":407},{"type":47,"tag":285,"props":633,"children":634},{"style":297},[635],{"type":53,"value":636}," libcuopt\n",{"type":47,"tag":257,"props":638,"children":640},{"id":639},"verify-1",[641],{"type":53,"value":418},{"type":47,"tag":56,"props":643,"children":644},{},[645,647,657,659,665],{"type":53,"value":646},"See ",{"type":47,"tag":648,"props":649,"children":651},"a",{"href":650},"references\u002Fverification_examples.md",[652],{"type":47,"tag":70,"props":653,"children":655},{"className":654},[],[656],{"type":53,"value":650},{"type":53,"value":658},"\nfor the canonical C-API header\u002Flibrary ",{"type":47,"tag":70,"props":660,"children":662},{"className":661},[],[663],{"type":53,"value":664},"find",{"type":53,"value":666}," commands (conda and pip\u002Fvenv variants).",{"type":47,"tag":79,"props":668,"children":670},{"id":669},"server-rest",[671],{"type":53,"value":672},"Server (REST)",{"type":47,"tag":257,"props":674,"children":676},{"id":675},"pip-2",[677],{"type":53,"value":259},{"type":47,"tag":274,"props":679,"children":681},{"className":276,"code":680,"language":278,"meta":279,"style":279},"pip install --extra-index-url=https:\u002F\u002Fpypi.nvidia.com cuopt-server-cu12 cuopt-sh-client\n",[682],{"type":47,"tag":70,"props":683,"children":684},{"__ignoreMap":279},[685],{"type":47,"tag":285,"props":686,"children":687},{"class":287,"line":288},[688,692,696,700,705],{"type":47,"tag":285,"props":689,"children":690},{"style":292},[691],{"type":53,"value":259},{"type":47,"tag":285,"props":693,"children":694},{"style":297},[695],{"type":53,"value":300},{"type":47,"tag":285,"props":697,"children":698},{"style":297},[699],{"type":53,"value":305},{"type":47,"tag":285,"props":701,"children":702},{"style":297},[703],{"type":53,"value":704}," cuopt-server-cu12",{"type":47,"tag":285,"props":706,"children":707},{"style":297},[708],{"type":53,"value":709}," cuopt-sh-client\n",{"type":47,"tag":257,"props":711,"children":713},{"id":712},"conda-2",[714],{"type":53,"value":359},{"type":47,"tag":274,"props":716,"children":718},{"className":276,"code":717,"language":278,"meta":279,"style":279},"conda install -c rapidsai -c conda-forge -c nvidia cuopt-server cuopt-sh-client\n",[719],{"type":47,"tag":70,"props":720,"children":721},{"__ignoreMap":279},[722],{"type":47,"tag":285,"props":723,"children":724},{"class":287,"line":288},[725,729,733,737,741,745,749,753,757,762],{"type":47,"tag":285,"props":726,"children":727},{"style":292},[728],{"type":53,"value":359},{"type":47,"tag":285,"props":730,"children":731},{"style":297},[732],{"type":53,"value":300},{"type":47,"tag":285,"props":734,"children":735},{"style":297},[736],{"type":53,"value":384},{"type":47,"tag":285,"props":738,"children":739},{"style":297},[740],{"type":53,"value":389},{"type":47,"tag":285,"props":742,"children":743},{"style":297},[744],{"type":53,"value":384},{"type":47,"tag":285,"props":746,"children":747},{"style":297},[748],{"type":53,"value":398},{"type":47,"tag":285,"props":750,"children":751},{"style":297},[752],{"type":53,"value":384},{"type":47,"tag":285,"props":754,"children":755},{"style":297},[756],{"type":53,"value":407},{"type":47,"tag":285,"props":758,"children":759},{"style":297},[760],{"type":53,"value":761}," cuopt-server",{"type":47,"tag":285,"props":763,"children":764},{"style":297},[765],{"type":53,"value":709},{"type":47,"tag":257,"props":767,"children":769},{"id":768},"docker",[770],{"type":53,"value":771},"Docker",{"type":47,"tag":274,"props":773,"children":775},{"className":276,"code":774,"language":278,"meta":279,"style":279},"docker pull nvidia\u002Fcuopt:latest-cuda12.9-py3.13\ndocker run --gpus all -it --rm -p 8000:8000 nvidia\u002Fcuopt:latest-cuda12.9-py3.13\n",[776],{"type":47,"tag":70,"props":777,"children":778},{"__ignoreMap":279},[779,796],{"type":47,"tag":285,"props":780,"children":781},{"class":287,"line":288},[782,786,791],{"type":47,"tag":285,"props":783,"children":784},{"style":292},[785],{"type":53,"value":768},{"type":47,"tag":285,"props":787,"children":788},{"style":297},[789],{"type":53,"value":790}," pull",{"type":47,"tag":285,"props":792,"children":793},{"style":297},[794],{"type":53,"value":795}," nvidia\u002Fcuopt:latest-cuda12.9-py3.13\n",{"type":47,"tag":285,"props":797,"children":798},{"class":287,"line":437},[799,803,808,813,818,823,828,833,838],{"type":47,"tag":285,"props":800,"children":801},{"style":292},[802],{"type":53,"value":768},{"type":47,"tag":285,"props":804,"children":805},{"style":297},[806],{"type":53,"value":807}," run",{"type":47,"tag":285,"props":809,"children":810},{"style":297},[811],{"type":53,"value":812}," --gpus",{"type":47,"tag":285,"props":814,"children":815},{"style":297},[816],{"type":53,"value":817}," all",{"type":47,"tag":285,"props":819,"children":820},{"style":297},[821],{"type":53,"value":822}," -it",{"type":47,"tag":285,"props":824,"children":825},{"style":297},[826],{"type":53,"value":827}," --rm",{"type":47,"tag":285,"props":829,"children":830},{"style":297},[831],{"type":53,"value":832}," -p",{"type":47,"tag":285,"props":834,"children":835},{"style":297},[836],{"type":53,"value":837}," 8000:8000",{"type":47,"tag":285,"props":839,"children":840},{"style":297},[841],{"type":53,"value":795},{"type":47,"tag":257,"props":843,"children":845},{"id":844},"verify-2",[846],{"type":53,"value":418},{"type":47,"tag":274,"props":848,"children":850},{"className":276,"code":849,"language":278,"meta":279,"style":279},"python -m cuopt_server.cuopt_service --ip 0.0.0.0 --port 8000 &\nsleep 5\ncurl -s http:\u002F\u002Flocalhost:8000\u002Fcuopt\u002Fhealth | jq .\n",[851],{"type":47,"tag":70,"props":852,"children":853},{"__ignoreMap":279},[854,897,910],{"type":47,"tag":285,"props":855,"children":856},{"class":287,"line":288},[857,861,866,871,876,882,887,892],{"type":47,"tag":285,"props":858,"children":859},{"style":292},[860],{"type":53,"value":22},{"type":47,"tag":285,"props":862,"children":863},{"style":297},[864],{"type":53,"value":865}," -m",{"type":47,"tag":285,"props":867,"children":868},{"style":297},[869],{"type":53,"value":870}," cuopt_server.cuopt_service",{"type":47,"tag":285,"props":872,"children":873},{"style":297},[874],{"type":53,"value":875}," --ip",{"type":47,"tag":285,"props":877,"children":879},{"style":878},"--shiki-light:#F76D47;--shiki-default:#F78C6C;--shiki-dark:#F78C6C",[880],{"type":53,"value":881}," 0.0.0.0",{"type":47,"tag":285,"props":883,"children":884},{"style":297},[885],{"type":53,"value":886}," --port",{"type":47,"tag":285,"props":888,"children":889},{"style":878},[890],{"type":53,"value":891}," 8000",{"type":47,"tag":285,"props":893,"children":894},{"style":343},[895],{"type":53,"value":896}," &\n",{"type":47,"tag":285,"props":898,"children":899},{"class":287,"line":437},[900,905],{"type":47,"tag":285,"props":901,"children":902},{"style":292},[903],{"type":53,"value":904},"sleep",{"type":47,"tag":285,"props":906,"children":907},{"style":878},[908],{"type":53,"value":909}," 5\n",{"type":47,"tag":285,"props":911,"children":912},{"class":287,"line":446},[913,918,923,928,933,938],{"type":47,"tag":285,"props":914,"children":915},{"style":292},[916],{"type":53,"value":917},"curl",{"type":47,"tag":285,"props":919,"children":920},{"style":297},[921],{"type":53,"value":922}," -s",{"type":47,"tag":285,"props":924,"children":925},{"style":297},[926],{"type":53,"value":927}," http:\u002F\u002Flocalhost:8000\u002Fcuopt\u002Fhealth",{"type":47,"tag":285,"props":929,"children":930},{"style":343},[931],{"type":53,"value":932}," |",{"type":47,"tag":285,"props":934,"children":935},{"style":292},[936],{"type":53,"value":937}," jq",{"type":47,"tag":285,"props":939,"children":940},{"style":297},[941],{"type":53,"value":942}," .\n",{"type":47,"tag":79,"props":944,"children":946},{"id":945},"common-issues",[947],{"type":53,"value":948},"Common Issues",{"type":47,"tag":86,"props":950,"children":951},{},[952,979,1014],{"type":47,"tag":90,"props":953,"children":954},{},[955,961,963,969,971,977],{"type":47,"tag":70,"props":956,"children":958},{"className":957},[],[959],{"type":53,"value":960},"No module named 'cuopt'",{"type":53,"value":962}," → check ",{"type":47,"tag":70,"props":964,"children":966},{"className":965},[],[967],{"type":53,"value":968},"pip list | grep cuopt",{"type":53,"value":970},", ",{"type":47,"tag":70,"props":972,"children":974},{"className":973},[],[975],{"type":53,"value":976},"which python",{"type":53,"value":978},", reinstall with the correct extra-index-url.",{"type":47,"tag":90,"props":980,"children":981},{},[982,984,989,991,996,998,1004,1006,1012],{"type":53,"value":983},"CUDA not available → run ",{"type":47,"tag":70,"props":985,"children":987},{"className":986},[],[988],{"type":53,"value":218},{"type":53,"value":990}," and ",{"type":47,"tag":70,"props":992,"children":994},{"className":993},[],[995],{"type":53,"value":210},{"type":53,"value":997},"; ensure the package CUDA suffix (",{"type":47,"tag":70,"props":999,"children":1001},{"className":1000},[],[1002],{"type":53,"value":1003},"cu12",{"type":53,"value":1005}," vs ",{"type":47,"tag":70,"props":1007,"children":1009},{"className":1008},[],[1010],{"type":53,"value":1011},"cu13",{"type":53,"value":1013},") matches the installed CUDA.",{"type":47,"tag":90,"props":1015,"children":1016},{},[1017,1019,1024,1026,1031,1033,1039,1041,1047,1049,1053,1055,1060],{"type":53,"value":1018},"Python vs C → ",{"type":47,"tag":70,"props":1020,"children":1022},{"className":1021},[],[1023],{"type":53,"value":145},{"type":53,"value":1025}," pulls in ",{"type":47,"tag":70,"props":1027,"children":1029},{"className":1028},[],[1030],{"type":53,"value":153},{"type":53,"value":1032}," as a transitive dependency, so the C library (",{"type":47,"tag":70,"props":1034,"children":1036},{"className":1035},[],[1037],{"type":53,"value":1038},"libcuopt.so",{"type":53,"value":1040},") and headers (",{"type":47,"tag":70,"props":1042,"children":1044},{"className":1043},[],[1045],{"type":53,"value":1046},"cuopt_c.h",{"type":53,"value":1048},") are already available after installing the Python package. The reverse is ",{"type":47,"tag":94,"props":1050,"children":1051},{},[1052],{"type":53,"value":167},{"type":53,"value":1054}," true: ",{"type":47,"tag":70,"props":1056,"children":1058},{"className":1057},[],[1059],{"type":53,"value":153},{"type":53,"value":1061}," alone does not install the Python bindings.",{"type":47,"tag":79,"props":1063,"children":1065},{"id":1064},"see-also",[1066],{"type":53,"value":1067},"See also",{"type":47,"tag":86,"props":1069,"children":1070},{},[1071,1081],{"type":47,"tag":90,"props":1072,"children":1073},{},[1074,1079],{"type":47,"tag":648,"props":1075,"children":1076},{"href":650},[1077],{"type":53,"value":1078},"verification_examples.md",{"type":53,"value":1080}," — full verification recipes for Python, C, server, and Docker.",{"type":47,"tag":90,"props":1082,"children":1083},{},[1084,1089],{"type":47,"tag":70,"props":1085,"children":1087},{"className":1086},[],[1088],{"type":53,"value":75},{"type":53,"value":1090}," — build cuOpt from source and contribute to the codebase.",{"type":47,"tag":1092,"props":1093,"children":1094},"style",{},[1095],{"type":53,"value":1096},"html .light .shiki span {color: var(--shiki-light);background: var(--shiki-light-bg);font-style: var(--shiki-light-font-style);font-weight: var(--shiki-light-font-weight);text-decoration: var(--shiki-light-text-decoration);}html.light .shiki span {color: var(--shiki-light);background: var(--shiki-light-bg);font-style: var(--shiki-light-font-style);font-weight: var(--shiki-light-font-weight);text-decoration: var(--shiki-light-text-decoration);}html .default .shiki span {color: var(--shiki-default);background: var(--shiki-default-bg);font-style: var(--shiki-default-font-style);font-weight: var(--shiki-default-font-weight);text-decoration: var(--shiki-default-text-decoration);}html .shiki span {color: var(--shiki-default);background: var(--shiki-default-bg);font-style: var(--shiki-default-font-style);font-weight: var(--shiki-default-font-weight);text-decoration: var(--shiki-default-text-decoration);}html .dark .shiki span {color: var(--shiki-dark);background: var(--shiki-dark-bg);font-style: var(--shiki-dark-font-style);font-weight: var(--shiki-dark-font-weight);text-decoration: var(--shiki-dark-text-decoration);}html.dark .shiki span {color: var(--shiki-dark);background: var(--shiki-dark-bg);font-style: var(--shiki-dark-font-style);font-weight: var(--shiki-dark-font-weight);text-decoration: var(--shiki-dark-text-decoration);}",{"items":1098,"total":1198},[1099,1116,1128,1142,1154,1171,1184],{"slug":1100,"name":1100,"fn":1101,"description":1102,"org":1103,"tags":1104,"stars":23,"repoUrl":24,"updatedAt":1115},"accelerated-computing-cudf","accelerate data processing with cuDF","Official NVIDIA-authored guidance for NVIDIA cuDF GPU DataFrames, pandas acceleration, dask-cuDF, ETL, joins, groupby, CSV\u002FParquet I\u002FO, nullable semantics, and multi-GPU DataFrame workloads.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[1105,1108,1111,1112],{"name":1106,"slug":1107,"type":15},"Data Analysis","data-analysis",{"name":1109,"slug":1110,"type":15},"Data Engineering","data-engineering",{"name":9,"slug":8,"type":15},{"name":1113,"slug":1114,"type":15},"Performance","performance","2026-07-14T05:28:43.176466",{"slug":1117,"name":1117,"fn":1118,"description":1119,"org":1120,"tags":1121,"stars":23,"repoUrl":24,"updatedAt":1127},"aiq-deploy","deploy and manage NVIDIA AI-Q infrastructure","Use when asked to install, deploy, run, validate, troubleshoot, or stop NVIDIA AI-Q Blueprint infrastructure.\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[1122,1123,1126],{"name":17,"slug":18,"type":15},{"name":1124,"slug":1125,"type":15},"Infrastructure","infrastructure",{"name":9,"slug":8,"type":15},"2026-07-14T05:29:06.667109",{"slug":1129,"name":1129,"fn":1130,"description":1131,"org":1132,"tags":1133,"stars":23,"repoUrl":24,"updatedAt":1141},"aiq-research","conduct deep research with AI-Q","Use when asked to run deep research or AI-Q research through a reachable NVIDIA AI-Q Blueprint backend.\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[1134,1137,1138],{"name":1135,"slug":1136,"type":15},"Agents","agents",{"name":9,"slug":8,"type":15},{"name":1139,"slug":1140,"type":15},"Research","research","2026-07-14T05:28:06.816956",{"slug":1143,"name":1143,"fn":1144,"description":1145,"org":1146,"tags":1147,"stars":23,"repoUrl":24,"updatedAt":1153},"amc-run-sample-calibration","run AMC sample dataset calibration","Run end-to-end calibration on the shipped sample dataset (sdg_08_2_sample_data_010926.zip) against a running AMC microservice. Use when user says 'test sample dataset', 'run sample calibration', 'verify AMC install', or 'launch and test'.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[1148,1149,1150],{"name":1106,"slug":1107,"type":15},{"name":9,"slug":8,"type":15},{"name":1151,"slug":1152,"type":15},"Testing","testing","2026-07-17T05:29:03.913266",{"slug":1155,"name":1155,"fn":1156,"description":1157,"org":1158,"tags":1159,"stars":23,"repoUrl":24,"updatedAt":1170},"amc-run-video-calibration","calibrate video datasets with AutoMagicCalib","Calibrate a new dataset from pre-recorded video files via the AutoMagicCalib REST API. Use when user has local MP4s and says 'calibrate my videos', 'run AMC on these videos', or similar. For RTSP\u002Flive streams, use amc-run-rtsp-calibration instead.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[1160,1163,1166,1167],{"name":1161,"slug":1162,"type":15},"Automation","automation",{"name":1164,"slug":1165,"type":15},"Imaging","imaging",{"name":9,"slug":8,"type":15},{"name":1168,"slug":1169,"type":15},"Video","video","2026-07-17T05:28:53.905004",{"slug":1172,"name":1172,"fn":1173,"description":1174,"org":1175,"tags":1176,"stars":23,"repoUrl":24,"updatedAt":1183},"amc-setup-calibration-stack","deploy AutoMagicCalib microservice with Docker","Launch AutoMagicCalib microservice and web UI from NGC release images via Docker Compose. Use when user says 'deploy auto calibration', 'launch auto calibration', 'launch AMC', 'start MS+UI', or 'set up auto-magic-calib'. Requires NGC API key.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[1177,1178,1179,1180],{"name":17,"slug":18,"type":15},{"name":771,"slug":768,"type":15},{"name":9,"slug":8,"type":15},{"name":1181,"slug":1182,"type":15},"Operations","operations","2026-07-17T05:28:56.913999",{"slug":1185,"name":1185,"fn":1186,"description":1187,"org":1188,"tags":1189,"stars":23,"repoUrl":24,"updatedAt":1197},"cudaq-guide","develop quantum applications with CUDA-Q","CUDA-Q onboarding guide for installation, test programs, GPU simulation, QPU hardware, and quantum applications.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},[1190,1191,1194],{"name":9,"slug":8,"type":15},{"name":1192,"slug":1193,"type":15},"Quantum Computing","quantum-computing",{"name":1195,"slug":1196,"type":15},"Simulation","simulation","2026-07-14T05:26:58.898253",305,{"items":1200,"total":1349},[1201,1219,1233,1244,1256,1270,1283,1297,1308,1317,1331,1340],{"slug":1202,"name":1202,"fn":1203,"description":1204,"org":1205,"tags":1206,"stars":1216,"repoUrl":1217,"updatedAt":1218},"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},[1207,1210,1213],{"name":1208,"slug":1209,"type":15},"Documentation","documentation",{"name":1211,"slug":1212,"type":15},"MCP","mcp",{"name":1214,"slug":1215,"type":15},"Search","search",21777,"https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FNemoClaw","2026-07-20T06:00:01.461044",{"slug":1220,"name":1220,"fn":1221,"description":1222,"org":1223,"tags":1224,"stars":1230,"repoUrl":1231,"updatedAt":1232},"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},[1225,1228,1229],{"name":1226,"slug":1227,"type":15},"Containers","containers",{"name":17,"slug":18,"type":15},{"name":21,"slug":22,"type":15},17049,"https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FMegatron-LM","2026-07-27T06:06:11.249662",{"slug":1234,"name":1234,"fn":1235,"description":1236,"org":1237,"tags":1238,"stars":1230,"repoUrl":1231,"updatedAt":1243},"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},[1239,1242],{"name":1240,"slug":1241,"type":15},"CI\u002FCD","ci-cd",{"name":17,"slug":18,"type":15},"2026-07-14T05:25:59.97109",{"slug":1245,"name":1245,"fn":1246,"description":1247,"org":1248,"tags":1249,"stars":1230,"repoUrl":1231,"updatedAt":1255},"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},[1250,1251,1252],{"name":1240,"slug":1241,"type":15},{"name":17,"slug":18,"type":15},{"name":1253,"slug":1254,"type":15},"GitHub","github","2026-07-27T06:06:12.278222",{"slug":1257,"name":1257,"fn":1258,"description":1259,"org":1260,"tags":1261,"stars":1230,"repoUrl":1231,"updatedAt":1269},"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},[1262,1265,1266],{"name":1263,"slug":1264,"type":15},"Debugging","debugging",{"name":1253,"slug":1254,"type":15},{"name":1267,"slug":1268,"type":15},"Triage","triage","2026-07-14T05:25:57.442089",{"slug":1271,"name":1271,"fn":1272,"description":1273,"org":1274,"tags":1275,"stars":1230,"repoUrl":1231,"updatedAt":1282},"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},[1276,1279],{"name":1277,"slug":1278,"type":15},"Best Practices","best-practices",{"name":1280,"slug":1281,"type":15},"Code Analysis","code-analysis","2026-07-14T05:25:56.18433",{"slug":1284,"name":1284,"fn":1285,"description":1286,"org":1287,"tags":1288,"stars":1230,"repoUrl":1231,"updatedAt":1296},"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},[1289,1292,1295],{"name":1290,"slug":1291,"type":15},"Machine Learning","machine-learning",{"name":1293,"slug":1294,"type":15},"Migration","migration",{"name":9,"slug":8,"type":15},"2026-07-17T06:07:11.777011",{"slug":1298,"name":1298,"fn":1299,"description":1300,"org":1301,"tags":1302,"stars":1230,"repoUrl":1231,"updatedAt":1307},"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},[1303,1306],{"name":1304,"slug":1305,"type":15},"QA","qa",{"name":1151,"slug":1152,"type":15},"2026-07-14T05:25:53.673039",{"slug":1309,"name":1309,"fn":1310,"description":1311,"org":1312,"tags":1313,"stars":1230,"repoUrl":1231,"updatedAt":1316},"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},[1314,1315],{"name":17,"slug":18,"type":15},{"name":1124,"slug":1125,"type":15},"2026-07-14T05:25:49.362534",{"slug":1318,"name":1318,"fn":1319,"description":1320,"org":1321,"tags":1322,"stars":1230,"repoUrl":1231,"updatedAt":1330},"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},[1323,1326,1327],{"name":1324,"slug":1325,"type":15},"Code Review","code-review",{"name":1253,"slug":1254,"type":15},{"name":1328,"slug":1329,"type":15},"Pull Requests","pull-requests","2026-07-14T05:26:01.226578",{"slug":1332,"name":1332,"fn":1333,"description":1334,"org":1335,"tags":1336,"stars":1230,"repoUrl":1231,"updatedAt":1339},"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},[1337,1338],{"name":1304,"slug":1305,"type":15},{"name":1151,"slug":1152,"type":15},"2026-07-14T05:25:54.928983",{"slug":1341,"name":1341,"fn":1342,"description":1343,"org":1344,"tags":1345,"stars":1230,"repoUrl":1231,"updatedAt":1348},"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},[1346,1347],{"name":1161,"slug":1162,"type":15},{"name":1240,"slug":1241,"type":15},"2026-07-30T05:29:03.275638",496]