[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill-nvidia-rag-blueprint":3,"mdc--gdb9sr-key":37,"related-repo-nvidia-rag-blueprint":2003,"related-org-nvidia-rag-blueprint":2105},{"slug":4,"name":4,"fn":5,"description":6,"org":7,"tags":11,"stars":26,"repoUrl":27,"updatedAt":28,"license":29,"forks":30,"topics":31,"repo":32,"sourceUrl":35,"mdContent":36},"rag-blueprint","deploy and manage NVIDIA RAG blueprints","NVIDIA RAG Blueprint — deploy, configure, troubleshoot, and manage. Handles any RAG action: deploy, install, start, enable, disable, toggle, change, configure, troubleshoot, debug, fix, shutdown, stop, or tear down any RAG feature or service (Agentic RAG, VLM, guardrails, query rewriting, models, search, ingestion, observability, summarization, reasoning, and more).",{"slug":8,"name":9,"logoUrl":10,"githubOrg":9},"nvidia","NVIDIA","https:\u002F\u002Fpexgzepcugksgbtrxkhf.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Forg-logos\u002Fnvidia.png",[12,16,19,22,23],{"name":13,"slug":14,"type":15},"LLM","llm","tag",{"name":17,"slug":18,"type":15},"RAG","rag",{"name":20,"slug":21,"type":15},"Deployment","deployment",{"name":9,"slug":8,"type":15},{"name":24,"slug":25,"type":15},"Infrastructure","infrastructure",2473,"https:\u002F\u002Fgithub.com\u002FNVIDIA\u002Fskills","2026-07-14T05:30:13.597785","Apache-2.0",281,[],{"repoUrl":27,"stars":26,"forks":30,"topics":33,"description":34},[],"AI agent skills published by NVIDIA","https:\u002F\u002Fgithub.com\u002FNVIDIA\u002Fskills\u002Ftree\u002FHEAD\u002Fskills\u002Frag-blueprint","---\nname: rag-blueprint\nversion: \"2.6.0\"\ndescription: \"NVIDIA RAG Blueprint — deploy, configure, troubleshoot, and manage. Handles any RAG action: deploy, install, start, enable, disable, toggle, change, configure, troubleshoot, debug, fix, shutdown, stop, or tear down any RAG feature or service (Agentic RAG, VLM, guardrails, query rewriting, models, search, ingestion, observability, summarization, reasoning, and more).\"\nlicense: Apache-2.0\ncompatibility: >-\n  NVIDIA RAG Blueprint repository checkout; Docker\u002FCompose or Kubernetes\u002FHelm\n  for deployments; Python 3.11+ for library workflows; NVIDIA GPU tooling for\n  self-hosted NIM services.\nmetadata:\n  author: \"NVIDIA RAG \u003Cfoundational-rag-dev@exchange.nvidia.com>\"\n  github-url: \"https:\u002F\u002Fgithub.com\u002FNVIDIA-AI-Blueprints\u002Frag\"\n  endpoint-openapi-schemas:\n    - docs\u002Fapi_reference\u002Fopenapi_schema_rag_server.json\n    - docs\u002Fapi_reference\u002Fopenapi_schema_ingestor_server.json\n  argument-hint: deploy RAG | enable feature | disable feature | configure | troubleshoot | shutdown\n  tags:\n    - nvidia\n    - blueprint\n    - rag\n    - deployment\n    - configuration\n    - troubleshooting\n  languages:\n    - python\n    - typescript\n    - shell\n  frameworks:\n    - fastapi\n    - langchain\n    - react\n    - docker-compose\n    - helm\n  domain: ai-ml\nallowed-tools: Bash(echo *) Bash(nvidia-smi *) Bash(curl --version *) Bash(docker ps *) Bash(docker info *) Bash(docker --version *) Bash(docker version *) Bash(docker logs *) Bash(docker inspect *) Bash(docker stats *) Bash(docker compose ps *) Bash(docker compose logs *) Bash(docker compose config *) Bash(docker compose version *) Bash(kubectl get *) Bash(kubectl describe *) Bash(kubectl version *) Bash(kubectl logs *) Bash(kubectl api-resources *) Bash(kubectl rollout status *) Bash(helm version *) Bash(helm list *) Bash(helm status *) Bash(oc get *) Bash(oc describe *) Bash(oc logs *) Bash(oc whoami *) Bash(oc version *) Bash(git rev-parse *) Bash(git describe *) Bash(git status *) Bash(python3 --version *) Bash(pip3 show *) Bash(df *) Bash(du *) Bash(cat \u002Fproc\u002F*) Bash(cat \u002Fetc\u002Fos-release *) Bash(ss *) Bash(netstat *) Bash(ls *) Bash(grep *) Bash(lsof *) Bash(ps aux *) Read Grep Glob\n---\n\n# NVIDIA RAG Blueprint\n\n## Purpose\n\nUse this skill for NVIDIA RAG Blueprint operations: deployment, configuration,\ntroubleshooting, shutdown, and feature management across Docker, Helm, and\nlibrary deployments.\n\n## Instructions\n\n1. Match the user request to the intent routing table below.\n2. Read the referenced playbook before making changes.\n3. Use repository docs and deployment config files as the source of truth.\n4. Verify the affected service or workflow after changes.\n\n## Prerequisites\n\n- NVIDIA RAG Blueprint repository checkout.\n- Docker\u002FCompose or Kubernetes\u002FHelm for deployments.\n- Python 3.11+ for library workflows.\n- NVIDIA GPU tooling for self-hosted NIM services.\n\n## Autonomy Principles\n\n- Auto-detect everything: GPU, VRAM, drivers, Docker, CUDA, disk, OS, ports, existing services, NGC key, repo state.\n- If it can be checked with a command, check it — don't ask the user.\n- Ask only when user action is required: providing an API key, confirming data deletion, or choosing between equally valid options.\n- Once analysis is done, route to the correct workflow and execute.\n\n## Intent Detection\n\nDetermine what the user wants and route immediately:\n\n| User Intent | Action |\n|-------------|--------|\n| Deploy, install, set up, start RAG | Read and follow `references\u002Fdeploy.md` |\n| Configure, enable, change, toggle a feature | Use the Configure section below |\n| Troubleshoot, debug, fix, error, unhealthy | Read and follow `references\u002Ftroubleshoot.md` |\n| Stop, shutdown, tear down, clean up | Read and follow `references\u002Fshutdown.md` |\n\nIf the intent is ambiguous, infer from context (e.g., \"RAG isn't working\" → troubleshoot; \"get RAG running\" → deploy). Only ask if genuinely unclear.\n\n---\n\n## Configure\n\nRequires a running RAG deployment. If services are not running, deploy first via `references\u002Fdeploy.md`.\n\nMatch the user's request to a reference file, then read and follow it:\n\n| Feature Keywords | Reference |\n|-----------------|-----------|\n| VLM, VLM embeddings, image captioning | `references\u002Fconfigure\u002Fvlm.md` |\n| NeMo Guardrails | `references\u002Fconfigure\u002Fguardrails.md` |\n| Agentic RAG, planning\u002Fexecution agent, agentic streaming, stage events | `references\u002Fconfigure\u002Fagentic-rag.md` |\n| Query rewriting, decomposition, multi-turn | `references\u002Fconfigure\u002Fquery-and-conversation.md` |\n| Ingestion (text-only, audio, Nemotron Parse, OCR, batch CLI, NV-Ingest, volume mount, performance) | `references\u002Fconfigure\u002Fingestion.md` |\n| Search, retrieval, hybrid search, multi-collection, metadata, filters, Elasticsearch filters, reranker, topK, accuracy\u002Fperformance | `references\u002Fconfigure\u002Fsearch-and-retrieval.md` |\n| LLM\u002Fembedding\u002Franking model changes, vector DB, Milvus\u002FElasticsearch auth, service keys, model profiles, ports\u002FGPU | `references\u002Fconfigure\u002Fmodels-and-infrastructure.md` |\n| Reasoning, thinking mode, `reasoning_content`, self-reflection, prompts, generation params (tokens, temperature, citations), per-request LLM params | `references\u002Fconfigure\u002Freasoning-and-generation.md` |\n| Summarization | `references\u002Fconfigure\u002Fsummarization.md` |\n| Observability (tracing, Zipkin, Grafana, Prometheus) | `references\u002Fconfigure\u002Fobservability.md` |\n| Multimodal query (image + text) | `references\u002Fconfigure\u002Fmultimodal-query.md` |\n| Data catalog (collection\u002Fdocument metadata) | `references\u002Fconfigure\u002Fdata-catalog.md` |\n| User interface (UI settings, reasoning panel, metadata filters) | `references\u002Fconfigure\u002Fuser-interface.md` |\n| API reference (endpoints, schemas) | `references\u002Fconfigure\u002Fapi-reference.md` |\n| Evaluation (RAGAS metrics) | `references\u002Fconfigure\u002Fevaluation.md` (and skill `rag-eval`) |\n| MCP server & client, agent toolkit | `references\u002Fconfigure\u002Fmcp.md` |\n| Migration (version upgrades) | `references\u002Fconfigure\u002Fmigration.md` |\n| Notebooks (setup and catalog) | `references\u002Fconfigure\u002Fnotebooks.md` |\n\n### Configure Flow\n\n1. Match the user's request to a reference file from the table above.\n\n2. Detect what's running:\n   ```bash\n   echo \"=== NIM ===\" && docker ps --format '{{.Names}}' 2>\u002Fdev\u002Fnull | grep -iE '(nim-llm|nemotron-(vlm-)?embedding|nemotron-ranking|nemotron-vlm|nemotron-3-nano-omni|page-elements|graphic-elements|table-structure|nemotron-ocr)' || echo \"NO_LOCAL_NIMS\"; echo \"=== RAG ===\" && docker ps --format '{{.Names}}' 2>\u002Fdev\u002Fnull | grep -iE '(rag-server|ingestor-server|elasticsearch|milvus|seaweedfs|lancedb)' || echo \"NO_DOCKER_RAG\"; echo \"=== K8S ===\" && kubectl get pods -n rag 2>\u002Fdev\u002Fnull | head -5 || echo \"NO_K8S\"; echo \"=== LIBRARY ===\" && ps aux 2>\u002Fdev\u002Fnull | grep -E '(nvidia_rag|uvicorn.*rag)' | grep -v grep || echo \"NO_LIBRARY\"\n   ```\n\n3. Use this table to determine platform, deployment type, and where config lives:\n\n   | Local NIMs running? | RAG services running? | Deployment Type | Config Location |\n   |---------------------|-----------------------|-----------------|-----------------|\n   | Yes (Docker) | Any | Self-hosted | `deploy\u002Fcompose\u002F.env` |\n   | No | Yes (Docker) | NVIDIA-hosted | `deploy\u002Fcompose\u002Fnvdev.env` |\n   | Yes (K8s pods) | Any | Self-hosted | `values.yaml` (NIM sections) |\n   | No | Yes (K8s pods) | NVIDIA-hosted | `values.yaml` (envVars) |\n   | — | Library processes | Library mode | `notebooks\u002Fconfig.yaml` |\n   | No | No | Not running | Deploy first via `references\u002Fdeploy.md` |\n\n   Tell the user what you detected and ask to confirm. Example: \"I see local NIM containers running (nim-llm-ms, nemotron-vlm-embedding-ms) — this is a self-hosted deployment. Config file is `deploy\u002Fcompose\u002F.env`. Correct?\"\n\n4. Check current feature state before changing anything — read the config location from step 3, then cross-check the live service:\n   - Docker: `docker exec rag-server env 2>\u002Fdev\u002Fnull | grep -E \"\u003CVAR_NAME>\"`\n   - Helm: `kubectl get pod -n rag -l app=rag-server -o jsonpath='{.items[0].spec.containers[0].env}' 2>\u002Fdev\u002Fnull`\n\n   If the config file and live service disagree, tell the user the service has stale config and will need a restart.\n\n5. If the feature needs extra GPUs, check availability against hardware restrictions (see below):\n   ```bash\n   nvidia-smi --query-gpu=index,name,memory.total,memory.used --format=csv,noheader 2>\u002Fdev\u002Fnull || echo \"NO_GPU\"\n   ```\n\n6. Read the reference file and apply changes:\n   - Docker: edit the env file (uncomment to enable, re-comment to disable — the env file is the source of truth). Then restart the affected service:\n     ```\n     source \u003Cenv-file> && docker compose -f deploy\u002Fcompose\u002F\u003Ccompose-file> up -d\n     ```\n     | Service | Compose File |\n     |---------|-------------|\n     | rag-server | `docker-compose-rag-server.yaml` |\n     | ingestor-server | `docker-compose-ingestor-server.yaml` |\n     | Elasticsearch, Milvus, etcd, SeaweedFS | `vectordb.yaml` |\n     | NIM containers (LLM, embedding, ranking, VLM, OCR, parse, audio, extraction) | `nims.yaml` |\n     | guardrails | `docker-compose-nemo-guardrails.yaml` |\n     | observability (Grafana, Prometheus, Zipkin) | `observability.yaml` |\n   - Helm: edit `values.yaml`, then upgrade: `helm upgrade rag \u003Cchart> -n rag -f values.yaml`\n   - Library: edit `notebooks\u002Fconfig.yaml`, then restart the Python process\n\n7. Verify:\n   - Docker: `docker ps --format \"table {{.Names}}\\t{{.Status}}\" | head -20; curl -s http:\u002F\u002Flocalhost:8081\u002Fv1\u002Fhealth?check_dependencies=true 2>\u002Fdev\u002Fnull | head -1`\n   - Helm: `kubectl get pods -n rag; kubectl rollout status deployment\u002Frag-server -n rag --timeout=120s`\n   - Library: `curl -s http:\u002F\u002Flocalhost:8081\u002Fv1\u002Fhealth 2>\u002Fdev\u002Fnull | head -1`\n\n8. If restart fails, read `references\u002Ftroubleshoot.md`. If multiple features requested, repeat from step 1 for each.\n\n## Examples\n\n- \"Deploy RAG\" -> route to `references\u002Fdeploy.md`.\n- \"Enable VLM\" -> route to `references\u002Fconfigure\u002Fvlm.md`.\n- \"RAG is unhealthy\" -> route to `references\u002Ftroubleshoot.md`.\n- \"Stop RAG\" -> route to `references\u002Fshutdown.md`.\n\n## Limitations\n\n- Operational guidance only applies to this RAG Blueprint repository.\n- Live deployment changes require a running Docker, Helm, or library target.\n- Secrets such as `NGC_API_KEY` must be supplied by the user environment.\n\n## Troubleshooting\n\n| Error \u002F signal | What to do |\n|----------------|------------|\n| Services are not running | Follow `references\u002Fdeploy.md` before configuring features. |\n| Restart or health check fails | Follow `references\u002Ftroubleshoot.md`. |\n| User requests teardown | Follow `references\u002Fshutdown.md` and confirm destructive cleanup. |\n\n### When User Says \"Configure\" Without Specifics\n\nRun steps 2–3 above, then read the identified config file to list what's currently enabled:\n```bash\ngrep -E \"^(export )?(ENABLE_|APP_)\" \u003Cconfig-file> 2>\u002Fdev\u002Fnull | sort\n```\nSummarize what's running and enabled, then ask which feature to change.\n\n---\n\n## Hardware Restrictions\n\nRead `docs\u002Fsupport-matrix.md` for current GPU requirements per deployment mode.\nRead `docs\u002Fservice-port-gpu-reference.md` for port mappings and GPU assignments.\n\n| GPU | Feature Restrictions |\n|-----|---------------------|\n| B200 | No VLM, No Guardrails, No Nemotron Parse. May need multi-GPU LLM (`LLM_MS_GPU_ID`). |\n| RTX PRO 6000 | No Nemotron Parse. No Audio on Helm. |\n",{"data":38,"body":64},{"name":4,"version":39,"description":6,"license":29,"compatibility":40,"metadata":41,"allowed-tools":63},"2.6.0","NVIDIA RAG Blueprint repository checkout; Docker\u002FCompose or Kubernetes\u002FHelm for deployments; Python 3.11+ for library workflows; NVIDIA GPU tooling for self-hosted NIM services.",{"author":42,"github-url":43,"endpoint-openapi-schemas":44,"argument-hint":47,"tags":48,"languages":52,"frameworks":56,"domain":62},"NVIDIA RAG \u003Cfoundational-rag-dev@exchange.nvidia.com>","https:\u002F\u002Fgithub.com\u002FNVIDIA-AI-Blueprints\u002Frag",[45,46],"docs\u002Fapi_reference\u002Fopenapi_schema_rag_server.json","docs\u002Fapi_reference\u002Fopenapi_schema_ingestor_server.json","deploy RAG | enable feature | disable feature | configure | troubleshoot | shutdown",[8,49,18,21,50,51],"blueprint","configuration","troubleshooting",[53,54,55],"python","typescript","shell",[57,58,59,60,61],"fastapi","langchain","react","docker-compose","helm","ai-ml","Bash(echo *) Bash(nvidia-smi *) Bash(curl --version *) Bash(docker ps *) Bash(docker info *) Bash(docker --version *) Bash(docker version *) Bash(docker logs *) Bash(docker inspect *) Bash(docker stats *) Bash(docker compose ps *) Bash(docker compose logs *) Bash(docker compose config *) Bash(docker compose version *) Bash(kubectl get *) Bash(kubectl describe *) Bash(kubectl version *) Bash(kubectl logs *) Bash(kubectl api-resources *) Bash(kubectl rollout status *) Bash(helm version *) Bash(helm list *) Bash(helm status *) Bash(oc get *) Bash(oc describe *) Bash(oc logs *) Bash(oc whoami *) Bash(oc version *) Bash(git rev-parse *) Bash(git describe *) Bash(git status *) Bash(python3 --version *) Bash(pip3 show *) Bash(df *) Bash(du *) Bash(cat \u002Fproc\u002F*) Bash(cat \u002Fetc\u002Fos-release *) Bash(ss *) Bash(netstat *) Bash(ls *) Bash(grep *) Bash(lsof *) Bash(ps aux *) Read Grep Glob",{"type":65,"children":66},"root",[67,76,83,89,95,120,126,150,156,179,185,190,287,292,296,302,314,319,665,672,1656,1662,1709,1715,1741,1746,1825,1831,1836,1906,1911,1914,1920,1941,1997],{"type":68,"tag":69,"props":70,"children":72},"element","h1",{"id":71},"nvidia-rag-blueprint",[73],{"type":74,"value":75},"text","NVIDIA RAG Blueprint",{"type":68,"tag":77,"props":78,"children":80},"h2",{"id":79},"purpose",[81],{"type":74,"value":82},"Purpose",{"type":68,"tag":84,"props":85,"children":86},"p",{},[87],{"type":74,"value":88},"Use this skill for NVIDIA RAG Blueprint operations: deployment, configuration,\ntroubleshooting, shutdown, and feature management across Docker, Helm, and\nlibrary deployments.",{"type":68,"tag":77,"props":90,"children":92},{"id":91},"instructions",[93],{"type":74,"value":94},"Instructions",{"type":68,"tag":96,"props":97,"children":98},"ol",{},[99,105,110,115],{"type":68,"tag":100,"props":101,"children":102},"li",{},[103],{"type":74,"value":104},"Match the user request to the intent routing table below.",{"type":68,"tag":100,"props":106,"children":107},{},[108],{"type":74,"value":109},"Read the referenced playbook before making changes.",{"type":68,"tag":100,"props":111,"children":112},{},[113],{"type":74,"value":114},"Use repository docs and deployment config files as the source of truth.",{"type":68,"tag":100,"props":116,"children":117},{},[118],{"type":74,"value":119},"Verify the affected service or workflow after changes.",{"type":68,"tag":77,"props":121,"children":123},{"id":122},"prerequisites",[124],{"type":74,"value":125},"Prerequisites",{"type":68,"tag":127,"props":128,"children":129},"ul",{},[130,135,140,145],{"type":68,"tag":100,"props":131,"children":132},{},[133],{"type":74,"value":134},"NVIDIA RAG Blueprint repository checkout.",{"type":68,"tag":100,"props":136,"children":137},{},[138],{"type":74,"value":139},"Docker\u002FCompose or Kubernetes\u002FHelm for deployments.",{"type":68,"tag":100,"props":141,"children":142},{},[143],{"type":74,"value":144},"Python 3.11+ for library workflows.",{"type":68,"tag":100,"props":146,"children":147},{},[148],{"type":74,"value":149},"NVIDIA GPU tooling for self-hosted NIM services.",{"type":68,"tag":77,"props":151,"children":153},{"id":152},"autonomy-principles",[154],{"type":74,"value":155},"Autonomy Principles",{"type":68,"tag":127,"props":157,"children":158},{},[159,164,169,174],{"type":68,"tag":100,"props":160,"children":161},{},[162],{"type":74,"value":163},"Auto-detect everything: GPU, VRAM, drivers, Docker, CUDA, disk, OS, ports, existing services, NGC key, repo state.",{"type":68,"tag":100,"props":165,"children":166},{},[167],{"type":74,"value":168},"If it can be checked with a command, check it — don't ask the user.",{"type":68,"tag":100,"props":170,"children":171},{},[172],{"type":74,"value":173},"Ask only when user action is required: providing an API key, confirming data deletion, or choosing between equally valid options.",{"type":68,"tag":100,"props":175,"children":176},{},[177],{"type":74,"value":178},"Once analysis is done, route to the correct workflow and execute.",{"type":68,"tag":77,"props":180,"children":182},{"id":181},"intent-detection",[183],{"type":74,"value":184},"Intent Detection",{"type":68,"tag":84,"props":186,"children":187},{},[188],{"type":74,"value":189},"Determine what the user wants and route immediately:",{"type":68,"tag":191,"props":192,"children":193},"table",{},[194,213],{"type":68,"tag":195,"props":196,"children":197},"thead",{},[198],{"type":68,"tag":199,"props":200,"children":201},"tr",{},[202,208],{"type":68,"tag":203,"props":204,"children":205},"th",{},[206],{"type":74,"value":207},"User Intent",{"type":68,"tag":203,"props":209,"children":210},{},[211],{"type":74,"value":212},"Action",{"type":68,"tag":214,"props":215,"children":216},"tbody",{},[217,238,251,269],{"type":68,"tag":199,"props":218,"children":219},{},[220,226],{"type":68,"tag":221,"props":222,"children":223},"td",{},[224],{"type":74,"value":225},"Deploy, install, set up, start RAG",{"type":68,"tag":221,"props":227,"children":228},{},[229,231],{"type":74,"value":230},"Read and follow ",{"type":68,"tag":232,"props":233,"children":235},"code",{"className":234},[],[236],{"type":74,"value":237},"references\u002Fdeploy.md",{"type":68,"tag":199,"props":239,"children":240},{},[241,246],{"type":68,"tag":221,"props":242,"children":243},{},[244],{"type":74,"value":245},"Configure, enable, change, toggle a feature",{"type":68,"tag":221,"props":247,"children":248},{},[249],{"type":74,"value":250},"Use the Configure section below",{"type":68,"tag":199,"props":252,"children":253},{},[254,259],{"type":68,"tag":221,"props":255,"children":256},{},[257],{"type":74,"value":258},"Troubleshoot, debug, fix, error, unhealthy",{"type":68,"tag":221,"props":260,"children":261},{},[262,263],{"type":74,"value":230},{"type":68,"tag":232,"props":264,"children":266},{"className":265},[],[267],{"type":74,"value":268},"references\u002Ftroubleshoot.md",{"type":68,"tag":199,"props":270,"children":271},{},[272,277],{"type":68,"tag":221,"props":273,"children":274},{},[275],{"type":74,"value":276},"Stop, shutdown, tear down, clean up",{"type":68,"tag":221,"props":278,"children":279},{},[280,281],{"type":74,"value":230},{"type":68,"tag":232,"props":282,"children":284},{"className":283},[],[285],{"type":74,"value":286},"references\u002Fshutdown.md",{"type":68,"tag":84,"props":288,"children":289},{},[290],{"type":74,"value":291},"If the intent is ambiguous, infer from context (e.g., \"RAG isn't working\" → troubleshoot; \"get RAG running\" → deploy). Only ask if genuinely unclear.",{"type":68,"tag":293,"props":294,"children":295},"hr",{},[],{"type":68,"tag":77,"props":297,"children":299},{"id":298},"configure",[300],{"type":74,"value":301},"Configure",{"type":68,"tag":84,"props":303,"children":304},{},[305,307,312],{"type":74,"value":306},"Requires a running RAG deployment. If services are not running, deploy first via ",{"type":68,"tag":232,"props":308,"children":310},{"className":309},[],[311],{"type":74,"value":237},{"type":74,"value":313},".",{"type":68,"tag":84,"props":315,"children":316},{},[317],{"type":74,"value":318},"Match the user's request to a reference file, then read and follow it:",{"type":68,"tag":191,"props":320,"children":321},{},[322,338],{"type":68,"tag":195,"props":323,"children":324},{},[325],{"type":68,"tag":199,"props":326,"children":327},{},[328,333],{"type":68,"tag":203,"props":329,"children":330},{},[331],{"type":74,"value":332},"Feature Keywords",{"type":68,"tag":203,"props":334,"children":335},{},[336],{"type":74,"value":337},"Reference",{"type":68,"tag":214,"props":339,"children":340},{},[341,358,375,392,409,426,443,460,485,502,519,536,553,570,587,614,631,648],{"type":68,"tag":199,"props":342,"children":343},{},[344,349],{"type":68,"tag":221,"props":345,"children":346},{},[347],{"type":74,"value":348},"VLM, VLM embeddings, image captioning",{"type":68,"tag":221,"props":350,"children":351},{},[352],{"type":68,"tag":232,"props":353,"children":355},{"className":354},[],[356],{"type":74,"value":357},"references\u002Fconfigure\u002Fvlm.md",{"type":68,"tag":199,"props":359,"children":360},{},[361,366],{"type":68,"tag":221,"props":362,"children":363},{},[364],{"type":74,"value":365},"NeMo Guardrails",{"type":68,"tag":221,"props":367,"children":368},{},[369],{"type":68,"tag":232,"props":370,"children":372},{"className":371},[],[373],{"type":74,"value":374},"references\u002Fconfigure\u002Fguardrails.md",{"type":68,"tag":199,"props":376,"children":377},{},[378,383],{"type":68,"tag":221,"props":379,"children":380},{},[381],{"type":74,"value":382},"Agentic RAG, planning\u002Fexecution agent, agentic streaming, stage events",{"type":68,"tag":221,"props":384,"children":385},{},[386],{"type":68,"tag":232,"props":387,"children":389},{"className":388},[],[390],{"type":74,"value":391},"references\u002Fconfigure\u002Fagentic-rag.md",{"type":68,"tag":199,"props":393,"children":394},{},[395,400],{"type":68,"tag":221,"props":396,"children":397},{},[398],{"type":74,"value":399},"Query rewriting, decomposition, multi-turn",{"type":68,"tag":221,"props":401,"children":402},{},[403],{"type":68,"tag":232,"props":404,"children":406},{"className":405},[],[407],{"type":74,"value":408},"references\u002Fconfigure\u002Fquery-and-conversation.md",{"type":68,"tag":199,"props":410,"children":411},{},[412,417],{"type":68,"tag":221,"props":413,"children":414},{},[415],{"type":74,"value":416},"Ingestion (text-only, audio, Nemotron Parse, OCR, batch CLI, NV-Ingest, volume mount, performance)",{"type":68,"tag":221,"props":418,"children":419},{},[420],{"type":68,"tag":232,"props":421,"children":423},{"className":422},[],[424],{"type":74,"value":425},"references\u002Fconfigure\u002Fingestion.md",{"type":68,"tag":199,"props":427,"children":428},{},[429,434],{"type":68,"tag":221,"props":430,"children":431},{},[432],{"type":74,"value":433},"Search, retrieval, hybrid search, multi-collection, metadata, filters, Elasticsearch filters, reranker, topK, accuracy\u002Fperformance",{"type":68,"tag":221,"props":435,"children":436},{},[437],{"type":68,"tag":232,"props":438,"children":440},{"className":439},[],[441],{"type":74,"value":442},"references\u002Fconfigure\u002Fsearch-and-retrieval.md",{"type":68,"tag":199,"props":444,"children":445},{},[446,451],{"type":68,"tag":221,"props":447,"children":448},{},[449],{"type":74,"value":450},"LLM\u002Fembedding\u002Franking model changes, vector DB, Milvus\u002FElasticsearch auth, service keys, model profiles, ports\u002FGPU",{"type":68,"tag":221,"props":452,"children":453},{},[454],{"type":68,"tag":232,"props":455,"children":457},{"className":456},[],[458],{"type":74,"value":459},"references\u002Fconfigure\u002Fmodels-and-infrastructure.md",{"type":68,"tag":199,"props":461,"children":462},{},[463,476],{"type":68,"tag":221,"props":464,"children":465},{},[466,468,474],{"type":74,"value":467},"Reasoning, thinking mode, ",{"type":68,"tag":232,"props":469,"children":471},{"className":470},[],[472],{"type":74,"value":473},"reasoning_content",{"type":74,"value":475},", self-reflection, prompts, generation params (tokens, temperature, citations), per-request LLM params",{"type":68,"tag":221,"props":477,"children":478},{},[479],{"type":68,"tag":232,"props":480,"children":482},{"className":481},[],[483],{"type":74,"value":484},"references\u002Fconfigure\u002Freasoning-and-generation.md",{"type":68,"tag":199,"props":486,"children":487},{},[488,493],{"type":68,"tag":221,"props":489,"children":490},{},[491],{"type":74,"value":492},"Summarization",{"type":68,"tag":221,"props":494,"children":495},{},[496],{"type":68,"tag":232,"props":497,"children":499},{"className":498},[],[500],{"type":74,"value":501},"references\u002Fconfigure\u002Fsummarization.md",{"type":68,"tag":199,"props":503,"children":504},{},[505,510],{"type":68,"tag":221,"props":506,"children":507},{},[508],{"type":74,"value":509},"Observability (tracing, Zipkin, Grafana, Prometheus)",{"type":68,"tag":221,"props":511,"children":512},{},[513],{"type":68,"tag":232,"props":514,"children":516},{"className":515},[],[517],{"type":74,"value":518},"references\u002Fconfigure\u002Fobservability.md",{"type":68,"tag":199,"props":520,"children":521},{},[522,527],{"type":68,"tag":221,"props":523,"children":524},{},[525],{"type":74,"value":526},"Multimodal query (image + text)",{"type":68,"tag":221,"props":528,"children":529},{},[530],{"type":68,"tag":232,"props":531,"children":533},{"className":532},[],[534],{"type":74,"value":535},"references\u002Fconfigure\u002Fmultimodal-query.md",{"type":68,"tag":199,"props":537,"children":538},{},[539,544],{"type":68,"tag":221,"props":540,"children":541},{},[542],{"type":74,"value":543},"Data catalog (collection\u002Fdocument metadata)",{"type":68,"tag":221,"props":545,"children":546},{},[547],{"type":68,"tag":232,"props":548,"children":550},{"className":549},[],[551],{"type":74,"value":552},"references\u002Fconfigure\u002Fdata-catalog.md",{"type":68,"tag":199,"props":554,"children":555},{},[556,561],{"type":68,"tag":221,"props":557,"children":558},{},[559],{"type":74,"value":560},"User interface (UI settings, reasoning panel, metadata filters)",{"type":68,"tag":221,"props":562,"children":563},{},[564],{"type":68,"tag":232,"props":565,"children":567},{"className":566},[],[568],{"type":74,"value":569},"references\u002Fconfigure\u002Fuser-interface.md",{"type":68,"tag":199,"props":571,"children":572},{},[573,578],{"type":68,"tag":221,"props":574,"children":575},{},[576],{"type":74,"value":577},"API reference (endpoints, schemas)",{"type":68,"tag":221,"props":579,"children":580},{},[581],{"type":68,"tag":232,"props":582,"children":584},{"className":583},[],[585],{"type":74,"value":586},"references\u002Fconfigure\u002Fapi-reference.md",{"type":68,"tag":199,"props":588,"children":589},{},[590,595],{"type":68,"tag":221,"props":591,"children":592},{},[593],{"type":74,"value":594},"Evaluation (RAGAS metrics)",{"type":68,"tag":221,"props":596,"children":597},{},[598,604,606,612],{"type":68,"tag":232,"props":599,"children":601},{"className":600},[],[602],{"type":74,"value":603},"references\u002Fconfigure\u002Fevaluation.md",{"type":74,"value":605}," (and skill ",{"type":68,"tag":232,"props":607,"children":609},{"className":608},[],[610],{"type":74,"value":611},"rag-eval",{"type":74,"value":613},")",{"type":68,"tag":199,"props":615,"children":616},{},[617,622],{"type":68,"tag":221,"props":618,"children":619},{},[620],{"type":74,"value":621},"MCP server & client, agent toolkit",{"type":68,"tag":221,"props":623,"children":624},{},[625],{"type":68,"tag":232,"props":626,"children":628},{"className":627},[],[629],{"type":74,"value":630},"references\u002Fconfigure\u002Fmcp.md",{"type":68,"tag":199,"props":632,"children":633},{},[634,639],{"type":68,"tag":221,"props":635,"children":636},{},[637],{"type":74,"value":638},"Migration (version upgrades)",{"type":68,"tag":221,"props":640,"children":641},{},[642],{"type":68,"tag":232,"props":643,"children":645},{"className":644},[],[646],{"type":74,"value":647},"references\u002Fconfigure\u002Fmigration.md",{"type":68,"tag":199,"props":649,"children":650},{},[651,656],{"type":68,"tag":221,"props":652,"children":653},{},[654],{"type":74,"value":655},"Notebooks (setup and catalog)",{"type":68,"tag":221,"props":657,"children":658},{},[659],{"type":68,"tag":232,"props":660,"children":662},{"className":661},[],[663],{"type":74,"value":664},"references\u002Fconfigure\u002Fnotebooks.md",{"type":68,"tag":666,"props":667,"children":669},"h3",{"id":668},"configure-flow",[670],{"type":74,"value":671},"Configure Flow",{"type":68,"tag":96,"props":673,"children":674},{},[675,680,1126,1334,1369,1428,1605,1644],{"type":68,"tag":100,"props":676,"children":677},{},[678],{"type":74,"value":679},"Match the user's request to a reference file from the table above.",{"type":68,"tag":100,"props":681,"children":682},{},[683,685],{"type":74,"value":684},"Detect what's running:",{"type":68,"tag":686,"props":687,"children":692},"pre",{"className":688,"code":689,"language":690,"meta":691,"style":691},"language-bash shiki shiki-themes material-theme-lighter material-theme material-theme-palenight","echo \"=== NIM ===\" && docker ps --format '{{.Names}}' 2>\u002Fdev\u002Fnull | grep -iE '(nim-llm|nemotron-(vlm-)?embedding|nemotron-ranking|nemotron-vlm|nemotron-3-nano-omni|page-elements|graphic-elements|table-structure|nemotron-ocr)' || echo \"NO_LOCAL_NIMS\"; echo \"=== RAG ===\" && docker ps --format '{{.Names}}' 2>\u002Fdev\u002Fnull | grep -iE '(rag-server|ingestor-server|elasticsearch|milvus|seaweedfs|lancedb)' || echo \"NO_DOCKER_RAG\"; echo \"=== K8S ===\" && kubectl get pods -n rag 2>\u002Fdev\u002Fnull | head -5 || echo \"NO_K8S\"; echo \"=== LIBRARY ===\" && ps aux 2>\u002Fdev\u002Fnull | grep -E '(nvidia_rag|uvicorn.*rag)' | grep -v grep || echo \"NO_LIBRARY\"\n","bash","",[693],{"type":68,"tag":232,"props":694,"children":695},{"__ignoreMap":691},[696],{"type":68,"tag":697,"props":698,"children":701},"span",{"class":699,"line":700},"line",1,[702,708,714,720,725,730,736,741,746,751,756,761,766,771,776,781,786,790,795,799,804,809,813,818,822,827,831,835,840,844,848,852,856,860,864,868,872,876,880,884,888,892,896,901,905,909,913,917,922,926,930,934,938,943,947,951,956,961,966,971,976,980,984,988,993,998,1002,1006,1010,1015,1019,1023,1027,1031,1036,1040,1044,1048,1053,1057,1061,1065,1069,1074,1078,1083,1087,1091,1095,1100,1104,1108,1112,1116,1121],{"type":68,"tag":697,"props":703,"children":705},{"style":704},"--shiki-light:#6182B8;--shiki-default:#82AAFF;--shiki-dark:#82AAFF",[706],{"type":74,"value":707},"echo",{"type":68,"tag":697,"props":709,"children":711},{"style":710},"--shiki-light:#39ADB5;--shiki-default:#89DDFF;--shiki-dark:#89DDFF",[712],{"type":74,"value":713}," \"",{"type":68,"tag":697,"props":715,"children":717},{"style":716},"--shiki-light:#91B859;--shiki-default:#C3E88D;--shiki-dark:#C3E88D",[718],{"type":74,"value":719},"=== NIM ===",{"type":68,"tag":697,"props":721,"children":722},{"style":710},[723],{"type":74,"value":724},"\"",{"type":68,"tag":697,"props":726,"children":727},{"style":710},[728],{"type":74,"value":729}," &&",{"type":68,"tag":697,"props":731,"children":733},{"style":732},"--shiki-light:#E2931D;--shiki-default:#FFCB6B;--shiki-dark:#FFCB6B",[734],{"type":74,"value":735}," docker",{"type":68,"tag":697,"props":737,"children":738},{"style":716},[739],{"type":74,"value":740}," ps",{"type":68,"tag":697,"props":742,"children":743},{"style":716},[744],{"type":74,"value":745}," --format",{"type":68,"tag":697,"props":747,"children":748},{"style":710},[749],{"type":74,"value":750}," '",{"type":68,"tag":697,"props":752,"children":753},{"style":716},[754],{"type":74,"value":755},"{{.Names}}",{"type":68,"tag":697,"props":757,"children":758},{"style":710},[759],{"type":74,"value":760},"'",{"type":68,"tag":697,"props":762,"children":763},{"style":710},[764],{"type":74,"value":765}," 2>",{"type":68,"tag":697,"props":767,"children":768},{"style":716},[769],{"type":74,"value":770},"\u002Fdev\u002Fnull",{"type":68,"tag":697,"props":772,"children":773},{"style":710},[774],{"type":74,"value":775}," |",{"type":68,"tag":697,"props":777,"children":778},{"style":732},[779],{"type":74,"value":780}," grep",{"type":68,"tag":697,"props":782,"children":783},{"style":716},[784],{"type":74,"value":785}," -iE",{"type":68,"tag":697,"props":787,"children":788},{"style":710},[789],{"type":74,"value":750},{"type":68,"tag":697,"props":791,"children":792},{"style":716},[793],{"type":74,"value":794},"(nim-llm|nemotron-(vlm-)?embedding|nemotron-ranking|nemotron-vlm|nemotron-3-nano-omni|page-elements|graphic-elements|table-structure|nemotron-ocr)",{"type":68,"tag":697,"props":796,"children":797},{"style":710},[798],{"type":74,"value":760},{"type":68,"tag":697,"props":800,"children":801},{"style":710},[802],{"type":74,"value":803}," ||",{"type":68,"tag":697,"props":805,"children":806},{"style":704},[807],{"type":74,"value":808}," echo",{"type":68,"tag":697,"props":810,"children":811},{"style":710},[812],{"type":74,"value":713},{"type":68,"tag":697,"props":814,"children":815},{"style":716},[816],{"type":74,"value":817},"NO_LOCAL_NIMS",{"type":68,"tag":697,"props":819,"children":820},{"style":710},[821],{"type":74,"value":724},{"type":68,"tag":697,"props":823,"children":824},{"style":710},[825],{"type":74,"value":826},";",{"type":68,"tag":697,"props":828,"children":829},{"style":704},[830],{"type":74,"value":808},{"type":68,"tag":697,"props":832,"children":833},{"style":710},[834],{"type":74,"value":713},{"type":68,"tag":697,"props":836,"children":837},{"style":716},[838],{"type":74,"value":839},"=== RAG ===",{"type":68,"tag":697,"props":841,"children":842},{"style":710},[843],{"type":74,"value":724},{"type":68,"tag":697,"props":845,"children":846},{"style":710},[847],{"type":74,"value":729},{"type":68,"tag":697,"props":849,"children":850},{"style":732},[851],{"type":74,"value":735},{"type":68,"tag":697,"props":853,"children":854},{"style":716},[855],{"type":74,"value":740},{"type":68,"tag":697,"props":857,"children":858},{"style":716},[859],{"type":74,"value":745},{"type":68,"tag":697,"props":861,"children":862},{"style":710},[863],{"type":74,"value":750},{"type":68,"tag":697,"props":865,"children":866},{"style":716},[867],{"type":74,"value":755},{"type":68,"tag":697,"props":869,"children":870},{"style":710},[871],{"type":74,"value":760},{"type":68,"tag":697,"props":873,"children":874},{"style":710},[875],{"type":74,"value":765},{"type":68,"tag":697,"props":877,"children":878},{"style":716},[879],{"type":74,"value":770},{"type":68,"tag":697,"props":881,"children":882},{"style":710},[883],{"type":74,"value":775},{"type":68,"tag":697,"props":885,"children":886},{"style":732},[887],{"type":74,"value":780},{"type":68,"tag":697,"props":889,"children":890},{"style":716},[891],{"type":74,"value":785},{"type":68,"tag":697,"props":893,"children":894},{"style":710},[895],{"type":74,"value":750},{"type":68,"tag":697,"props":897,"children":898},{"style":716},[899],{"type":74,"value":900},"(rag-server|ingestor-server|elasticsearch|milvus|seaweedfs|lancedb)",{"type":68,"tag":697,"props":902,"children":903},{"style":710},[904],{"type":74,"value":760},{"type":68,"tag":697,"props":906,"children":907},{"style":710},[908],{"type":74,"value":803},{"type":68,"tag":697,"props":910,"children":911},{"style":704},[912],{"type":74,"value":808},{"type":68,"tag":697,"props":914,"children":915},{"style":710},[916],{"type":74,"value":713},{"type":68,"tag":697,"props":918,"children":919},{"style":716},[920],{"type":74,"value":921},"NO_DOCKER_RAG",{"type":68,"tag":697,"props":923,"children":924},{"style":710},[925],{"type":74,"value":724},{"type":68,"tag":697,"props":927,"children":928},{"style":710},[929],{"type":74,"value":826},{"type":68,"tag":697,"props":931,"children":932},{"style":704},[933],{"type":74,"value":808},{"type":68,"tag":697,"props":935,"children":936},{"style":710},[937],{"type":74,"value":713},{"type":68,"tag":697,"props":939,"children":940},{"style":716},[941],{"type":74,"value":942},"=== K8S ===",{"type":68,"tag":697,"props":944,"children":945},{"style":710},[946],{"type":74,"value":724},{"type":68,"tag":697,"props":948,"children":949},{"style":710},[950],{"type":74,"value":729},{"type":68,"tag":697,"props":952,"children":953},{"style":732},[954],{"type":74,"value":955}," kubectl",{"type":68,"tag":697,"props":957,"children":958},{"style":716},[959],{"type":74,"value":960}," get",{"type":68,"tag":697,"props":962,"children":963},{"style":716},[964],{"type":74,"value":965}," pods",{"type":68,"tag":697,"props":967,"children":968},{"style":716},[969],{"type":74,"value":970}," -n",{"type":68,"tag":697,"props":972,"children":973},{"style":716},[974],{"type":74,"value":975}," rag",{"type":68,"tag":697,"props":977,"children":978},{"style":710},[979],{"type":74,"value":765},{"type":68,"tag":697,"props":981,"children":982},{"style":716},[983],{"type":74,"value":770},{"type":68,"tag":697,"props":985,"children":986},{"style":710},[987],{"type":74,"value":775},{"type":68,"tag":697,"props":989,"children":990},{"style":732},[991],{"type":74,"value":992}," head",{"type":68,"tag":697,"props":994,"children":995},{"style":716},[996],{"type":74,"value":997}," -5",{"type":68,"tag":697,"props":999,"children":1000},{"style":710},[1001],{"type":74,"value":803},{"type":68,"tag":697,"props":1003,"children":1004},{"style":704},[1005],{"type":74,"value":808},{"type":68,"tag":697,"props":1007,"children":1008},{"style":710},[1009],{"type":74,"value":713},{"type":68,"tag":697,"props":1011,"children":1012},{"style":716},[1013],{"type":74,"value":1014},"NO_K8S",{"type":68,"tag":697,"props":1016,"children":1017},{"style":710},[1018],{"type":74,"value":724},{"type":68,"tag":697,"props":1020,"children":1021},{"style":710},[1022],{"type":74,"value":826},{"type":68,"tag":697,"props":1024,"children":1025},{"style":704},[1026],{"type":74,"value":808},{"type":68,"tag":697,"props":1028,"children":1029},{"style":710},[1030],{"type":74,"value":713},{"type":68,"tag":697,"props":1032,"children":1033},{"style":716},[1034],{"type":74,"value":1035},"=== LIBRARY ===",{"type":68,"tag":697,"props":1037,"children":1038},{"style":710},[1039],{"type":74,"value":724},{"type":68,"tag":697,"props":1041,"children":1042},{"style":710},[1043],{"type":74,"value":729},{"type":68,"tag":697,"props":1045,"children":1046},{"style":732},[1047],{"type":74,"value":740},{"type":68,"tag":697,"props":1049,"children":1050},{"style":716},[1051],{"type":74,"value":1052}," aux",{"type":68,"tag":697,"props":1054,"children":1055},{"style":710},[1056],{"type":74,"value":765},{"type":68,"tag":697,"props":1058,"children":1059},{"style":716},[1060],{"type":74,"value":770},{"type":68,"tag":697,"props":1062,"children":1063},{"style":710},[1064],{"type":74,"value":775},{"type":68,"tag":697,"props":1066,"children":1067},{"style":732},[1068],{"type":74,"value":780},{"type":68,"tag":697,"props":1070,"children":1071},{"style":716},[1072],{"type":74,"value":1073}," -E",{"type":68,"tag":697,"props":1075,"children":1076},{"style":710},[1077],{"type":74,"value":750},{"type":68,"tag":697,"props":1079,"children":1080},{"style":716},[1081],{"type":74,"value":1082},"(nvidia_rag|uvicorn.*rag)",{"type":68,"tag":697,"props":1084,"children":1085},{"style":710},[1086],{"type":74,"value":760},{"type":68,"tag":697,"props":1088,"children":1089},{"style":710},[1090],{"type":74,"value":775},{"type":68,"tag":697,"props":1092,"children":1093},{"style":732},[1094],{"type":74,"value":780},{"type":68,"tag":697,"props":1096,"children":1097},{"style":716},[1098],{"type":74,"value":1099}," -v",{"type":68,"tag":697,"props":1101,"children":1102},{"style":716},[1103],{"type":74,"value":780},{"type":68,"tag":697,"props":1105,"children":1106},{"style":710},[1107],{"type":74,"value":803},{"type":68,"tag":697,"props":1109,"children":1110},{"style":704},[1111],{"type":74,"value":808},{"type":68,"tag":697,"props":1113,"children":1114},{"style":710},[1115],{"type":74,"value":713},{"type":68,"tag":697,"props":1117,"children":1118},{"style":716},[1119],{"type":74,"value":1120},"NO_LIBRARY",{"type":68,"tag":697,"props":1122,"children":1123},{"style":710},[1124],{"type":74,"value":1125},"\"\n",{"type":68,"tag":100,"props":1127,"children":1128},{},[1129,1131,1321,1325,1327,1332],{"type":74,"value":1130},"Use this table to determine platform, deployment type, and where config lives:",{"type":68,"tag":191,"props":1132,"children":1133},{},[1134,1160],{"type":68,"tag":195,"props":1135,"children":1136},{},[1137],{"type":68,"tag":199,"props":1138,"children":1139},{},[1140,1145,1150,1155],{"type":68,"tag":203,"props":1141,"children":1142},{},[1143],{"type":74,"value":1144},"Local NIMs running?",{"type":68,"tag":203,"props":1146,"children":1147},{},[1148],{"type":74,"value":1149},"RAG services running?",{"type":68,"tag":203,"props":1151,"children":1152},{},[1153],{"type":74,"value":1154},"Deployment Type",{"type":68,"tag":203,"props":1156,"children":1157},{},[1158],{"type":74,"value":1159},"Config Location",{"type":68,"tag":214,"props":1161,"children":1162},{},[1163,1190,1216,1243,1268,1295],{"type":68,"tag":199,"props":1164,"children":1165},{},[1166,1171,1176,1181],{"type":68,"tag":221,"props":1167,"children":1168},{},[1169],{"type":74,"value":1170},"Yes (Docker)",{"type":68,"tag":221,"props":1172,"children":1173},{},[1174],{"type":74,"value":1175},"Any",{"type":68,"tag":221,"props":1177,"children":1178},{},[1179],{"type":74,"value":1180},"Self-hosted",{"type":68,"tag":221,"props":1182,"children":1183},{},[1184],{"type":68,"tag":232,"props":1185,"children":1187},{"className":1186},[],[1188],{"type":74,"value":1189},"deploy\u002Fcompose\u002F.env",{"type":68,"tag":199,"props":1191,"children":1192},{},[1193,1198,1202,1207],{"type":68,"tag":221,"props":1194,"children":1195},{},[1196],{"type":74,"value":1197},"No",{"type":68,"tag":221,"props":1199,"children":1200},{},[1201],{"type":74,"value":1170},{"type":68,"tag":221,"props":1203,"children":1204},{},[1205],{"type":74,"value":1206},"NVIDIA-hosted",{"type":68,"tag":221,"props":1208,"children":1209},{},[1210],{"type":68,"tag":232,"props":1211,"children":1213},{"className":1212},[],[1214],{"type":74,"value":1215},"deploy\u002Fcompose\u002Fnvdev.env",{"type":68,"tag":199,"props":1217,"children":1218},{},[1219,1224,1228,1232],{"type":68,"tag":221,"props":1220,"children":1221},{},[1222],{"type":74,"value":1223},"Yes (K8s pods)",{"type":68,"tag":221,"props":1225,"children":1226},{},[1227],{"type":74,"value":1175},{"type":68,"tag":221,"props":1229,"children":1230},{},[1231],{"type":74,"value":1180},{"type":68,"tag":221,"props":1233,"children":1234},{},[1235,1241],{"type":68,"tag":232,"props":1236,"children":1238},{"className":1237},[],[1239],{"type":74,"value":1240},"values.yaml",{"type":74,"value":1242}," (NIM sections)",{"type":68,"tag":199,"props":1244,"children":1245},{},[1246,1250,1254,1258],{"type":68,"tag":221,"props":1247,"children":1248},{},[1249],{"type":74,"value":1197},{"type":68,"tag":221,"props":1251,"children":1252},{},[1253],{"type":74,"value":1223},{"type":68,"tag":221,"props":1255,"children":1256},{},[1257],{"type":74,"value":1206},{"type":68,"tag":221,"props":1259,"children":1260},{},[1261,1266],{"type":68,"tag":232,"props":1262,"children":1264},{"className":1263},[],[1265],{"type":74,"value":1240},{"type":74,"value":1267}," (envVars)",{"type":68,"tag":199,"props":1269,"children":1270},{},[1271,1276,1281,1286],{"type":68,"tag":221,"props":1272,"children":1273},{},[1274],{"type":74,"value":1275},"—",{"type":68,"tag":221,"props":1277,"children":1278},{},[1279],{"type":74,"value":1280},"Library processes",{"type":68,"tag":221,"props":1282,"children":1283},{},[1284],{"type":74,"value":1285},"Library mode",{"type":68,"tag":221,"props":1287,"children":1288},{},[1289],{"type":68,"tag":232,"props":1290,"children":1292},{"className":1291},[],[1293],{"type":74,"value":1294},"notebooks\u002Fconfig.yaml",{"type":68,"tag":199,"props":1296,"children":1297},{},[1298,1302,1306,1311],{"type":68,"tag":221,"props":1299,"children":1300},{},[1301],{"type":74,"value":1197},{"type":68,"tag":221,"props":1303,"children":1304},{},[1305],{"type":74,"value":1197},{"type":68,"tag":221,"props":1307,"children":1308},{},[1309],{"type":74,"value":1310},"Not running",{"type":68,"tag":221,"props":1312,"children":1313},{},[1314,1316],{"type":74,"value":1315},"Deploy first via ",{"type":68,"tag":232,"props":1317,"children":1319},{"className":1318},[],[1320],{"type":74,"value":237},{"type":68,"tag":1322,"props":1323,"children":1324},"br",{},[],{"type":74,"value":1326},"Tell the user what you detected and ask to confirm. Example: \"I see local NIM containers running (nim-llm-ms, nemotron-vlm-embedding-ms) — this is a self-hosted deployment. Config file is ",{"type":68,"tag":232,"props":1328,"children":1330},{"className":1329},[],[1331],{"type":74,"value":1189},{"type":74,"value":1333},". Correct?\"",{"type":68,"tag":100,"props":1335,"children":1336},{},[1337,1339,1364,1367],{"type":74,"value":1338},"Check current feature state before changing anything — read the config location from step 3, then cross-check the live service:",{"type":68,"tag":127,"props":1340,"children":1341},{},[1342,1353],{"type":68,"tag":100,"props":1343,"children":1344},{},[1345,1347],{"type":74,"value":1346},"Docker: ",{"type":68,"tag":232,"props":1348,"children":1350},{"className":1349},[],[1351],{"type":74,"value":1352},"docker exec rag-server env 2>\u002Fdev\u002Fnull | grep -E \"\u003CVAR_NAME>\"",{"type":68,"tag":100,"props":1354,"children":1355},{},[1356,1358],{"type":74,"value":1357},"Helm: ",{"type":68,"tag":232,"props":1359,"children":1361},{"className":1360},[],[1362],{"type":74,"value":1363},"kubectl get pod -n rag -l app=rag-server -o jsonpath='{.items[0].spec.containers[0].env}' 2>\u002Fdev\u002Fnull",{"type":68,"tag":1322,"props":1365,"children":1366},{},[],{"type":74,"value":1368},"If the config file and live service disagree, tell the user the service has stale config and will need a restart.",{"type":68,"tag":100,"props":1370,"children":1371},{},[1372,1374],{"type":74,"value":1373},"If the feature needs extra GPUs, check availability against hardware restrictions (see below):",{"type":68,"tag":686,"props":1375,"children":1377},{"className":688,"code":1376,"language":690,"meta":691,"style":691},"nvidia-smi --query-gpu=index,name,memory.total,memory.used --format=csv,noheader 2>\u002Fdev\u002Fnull || echo \"NO_GPU\"\n",[1378],{"type":68,"tag":232,"props":1379,"children":1380},{"__ignoreMap":691},[1381],{"type":68,"tag":697,"props":1382,"children":1383},{"class":699,"line":700},[1384,1389,1394,1399,1403,1407,1411,1415,1419,1424],{"type":68,"tag":697,"props":1385,"children":1386},{"style":732},[1387],{"type":74,"value":1388},"nvidia-smi",{"type":68,"tag":697,"props":1390,"children":1391},{"style":716},[1392],{"type":74,"value":1393}," --query-gpu=index,name,memory.total,memory.used",{"type":68,"tag":697,"props":1395,"children":1396},{"style":716},[1397],{"type":74,"value":1398}," --format=csv,noheader",{"type":68,"tag":697,"props":1400,"children":1401},{"style":710},[1402],{"type":74,"value":765},{"type":68,"tag":697,"props":1404,"children":1405},{"style":716},[1406],{"type":74,"value":770},{"type":68,"tag":697,"props":1408,"children":1409},{"style":710},[1410],{"type":74,"value":803},{"type":68,"tag":697,"props":1412,"children":1413},{"style":704},[1414],{"type":74,"value":808},{"type":68,"tag":697,"props":1416,"children":1417},{"style":710},[1418],{"type":74,"value":713},{"type":68,"tag":697,"props":1420,"children":1421},{"style":716},[1422],{"type":74,"value":1423},"NO_GPU",{"type":68,"tag":697,"props":1425,"children":1426},{"style":710},[1427],{"type":74,"value":1125},{"type":68,"tag":100,"props":1429,"children":1430},{},[1431,1433],{"type":74,"value":1432},"Read the reference file and apply changes:",{"type":68,"tag":127,"props":1434,"children":1435},{},[1436,1575,1593],{"type":68,"tag":100,"props":1437,"children":1438},{},[1439,1441,1451],{"type":74,"value":1440},"Docker: edit the env file (uncomment to enable, re-comment to disable — the env file is the source of truth). Then restart the affected service:\n",{"type":68,"tag":686,"props":1442,"children":1446},{"className":1443,"code":1445,"language":74},[1444],"language-text","source \u003Cenv-file> && docker compose -f deploy\u002Fcompose\u002F\u003Ccompose-file> up -d\n",[1447],{"type":68,"tag":232,"props":1448,"children":1449},{"__ignoreMap":691},[1450],{"type":74,"value":1445},{"type":68,"tag":191,"props":1452,"children":1453},{},[1454,1470],{"type":68,"tag":195,"props":1455,"children":1456},{},[1457],{"type":68,"tag":199,"props":1458,"children":1459},{},[1460,1465],{"type":68,"tag":203,"props":1461,"children":1462},{},[1463],{"type":74,"value":1464},"Service",{"type":68,"tag":203,"props":1466,"children":1467},{},[1468],{"type":74,"value":1469},"Compose File",{"type":68,"tag":214,"props":1471,"children":1472},{},[1473,1490,1507,1524,1541,1558],{"type":68,"tag":199,"props":1474,"children":1475},{},[1476,1481],{"type":68,"tag":221,"props":1477,"children":1478},{},[1479],{"type":74,"value":1480},"rag-server",{"type":68,"tag":221,"props":1482,"children":1483},{},[1484],{"type":68,"tag":232,"props":1485,"children":1487},{"className":1486},[],[1488],{"type":74,"value":1489},"docker-compose-rag-server.yaml",{"type":68,"tag":199,"props":1491,"children":1492},{},[1493,1498],{"type":68,"tag":221,"props":1494,"children":1495},{},[1496],{"type":74,"value":1497},"ingestor-server",{"type":68,"tag":221,"props":1499,"children":1500},{},[1501],{"type":68,"tag":232,"props":1502,"children":1504},{"className":1503},[],[1505],{"type":74,"value":1506},"docker-compose-ingestor-server.yaml",{"type":68,"tag":199,"props":1508,"children":1509},{},[1510,1515],{"type":68,"tag":221,"props":1511,"children":1512},{},[1513],{"type":74,"value":1514},"Elasticsearch, Milvus, etcd, SeaweedFS",{"type":68,"tag":221,"props":1516,"children":1517},{},[1518],{"type":68,"tag":232,"props":1519,"children":1521},{"className":1520},[],[1522],{"type":74,"value":1523},"vectordb.yaml",{"type":68,"tag":199,"props":1525,"children":1526},{},[1527,1532],{"type":68,"tag":221,"props":1528,"children":1529},{},[1530],{"type":74,"value":1531},"NIM containers (LLM, embedding, ranking, VLM, OCR, parse, audio, extraction)",{"type":68,"tag":221,"props":1533,"children":1534},{},[1535],{"type":68,"tag":232,"props":1536,"children":1538},{"className":1537},[],[1539],{"type":74,"value":1540},"nims.yaml",{"type":68,"tag":199,"props":1542,"children":1543},{},[1544,1549],{"type":68,"tag":221,"props":1545,"children":1546},{},[1547],{"type":74,"value":1548},"guardrails",{"type":68,"tag":221,"props":1550,"children":1551},{},[1552],{"type":68,"tag":232,"props":1553,"children":1555},{"className":1554},[],[1556],{"type":74,"value":1557},"docker-compose-nemo-guardrails.yaml",{"type":68,"tag":199,"props":1559,"children":1560},{},[1561,1566],{"type":68,"tag":221,"props":1562,"children":1563},{},[1564],{"type":74,"value":1565},"observability (Grafana, Prometheus, Zipkin)",{"type":68,"tag":221,"props":1567,"children":1568},{},[1569],{"type":68,"tag":232,"props":1570,"children":1572},{"className":1571},[],[1573],{"type":74,"value":1574},"observability.yaml",{"type":68,"tag":100,"props":1576,"children":1577},{},[1578,1580,1585,1587],{"type":74,"value":1579},"Helm: edit ",{"type":68,"tag":232,"props":1581,"children":1583},{"className":1582},[],[1584],{"type":74,"value":1240},{"type":74,"value":1586},", then upgrade: ",{"type":68,"tag":232,"props":1588,"children":1590},{"className":1589},[],[1591],{"type":74,"value":1592},"helm upgrade rag \u003Cchart> -n rag -f values.yaml",{"type":68,"tag":100,"props":1594,"children":1595},{},[1596,1598,1603],{"type":74,"value":1597},"Library: edit ",{"type":68,"tag":232,"props":1599,"children":1601},{"className":1600},[],[1602],{"type":74,"value":1294},{"type":74,"value":1604},", then restart the Python process",{"type":68,"tag":100,"props":1606,"children":1607},{},[1608,1610],{"type":74,"value":1609},"Verify:",{"type":68,"tag":127,"props":1611,"children":1612},{},[1613,1623,1633],{"type":68,"tag":100,"props":1614,"children":1615},{},[1616,1617],{"type":74,"value":1346},{"type":68,"tag":232,"props":1618,"children":1620},{"className":1619},[],[1621],{"type":74,"value":1622},"docker ps --format \"table {{.Names}}\\t{{.Status}}\" | head -20; curl -s http:\u002F\u002Flocalhost:8081\u002Fv1\u002Fhealth?check_dependencies=true 2>\u002Fdev\u002Fnull | head -1",{"type":68,"tag":100,"props":1624,"children":1625},{},[1626,1627],{"type":74,"value":1357},{"type":68,"tag":232,"props":1628,"children":1630},{"className":1629},[],[1631],{"type":74,"value":1632},"kubectl get pods -n rag; kubectl rollout status deployment\u002Frag-server -n rag --timeout=120s",{"type":68,"tag":100,"props":1634,"children":1635},{},[1636,1638],{"type":74,"value":1637},"Library: ",{"type":68,"tag":232,"props":1639,"children":1641},{"className":1640},[],[1642],{"type":74,"value":1643},"curl -s http:\u002F\u002Flocalhost:8081\u002Fv1\u002Fhealth 2>\u002Fdev\u002Fnull | head -1",{"type":68,"tag":100,"props":1645,"children":1646},{},[1647,1649,1654],{"type":74,"value":1648},"If restart fails, read ",{"type":68,"tag":232,"props":1650,"children":1652},{"className":1651},[],[1653],{"type":74,"value":268},{"type":74,"value":1655},". If multiple features requested, repeat from step 1 for each.",{"type":68,"tag":77,"props":1657,"children":1659},{"id":1658},"examples",[1660],{"type":74,"value":1661},"Examples",{"type":68,"tag":127,"props":1663,"children":1664},{},[1665,1676,1687,1698],{"type":68,"tag":100,"props":1666,"children":1667},{},[1668,1670,1675],{"type":74,"value":1669},"\"Deploy RAG\" -> route to ",{"type":68,"tag":232,"props":1671,"children":1673},{"className":1672},[],[1674],{"type":74,"value":237},{"type":74,"value":313},{"type":68,"tag":100,"props":1677,"children":1678},{},[1679,1681,1686],{"type":74,"value":1680},"\"Enable VLM\" -> route to ",{"type":68,"tag":232,"props":1682,"children":1684},{"className":1683},[],[1685],{"type":74,"value":357},{"type":74,"value":313},{"type":68,"tag":100,"props":1688,"children":1689},{},[1690,1692,1697],{"type":74,"value":1691},"\"RAG is unhealthy\" -> route to ",{"type":68,"tag":232,"props":1693,"children":1695},{"className":1694},[],[1696],{"type":74,"value":268},{"type":74,"value":313},{"type":68,"tag":100,"props":1699,"children":1700},{},[1701,1703,1708],{"type":74,"value":1702},"\"Stop RAG\" -> route to ",{"type":68,"tag":232,"props":1704,"children":1706},{"className":1705},[],[1707],{"type":74,"value":286},{"type":74,"value":313},{"type":68,"tag":77,"props":1710,"children":1712},{"id":1711},"limitations",[1713],{"type":74,"value":1714},"Limitations",{"type":68,"tag":127,"props":1716,"children":1717},{},[1718,1723,1728],{"type":68,"tag":100,"props":1719,"children":1720},{},[1721],{"type":74,"value":1722},"Operational guidance only applies to this RAG Blueprint repository.",{"type":68,"tag":100,"props":1724,"children":1725},{},[1726],{"type":74,"value":1727},"Live deployment changes require a running Docker, Helm, or library target.",{"type":68,"tag":100,"props":1729,"children":1730},{},[1731,1733,1739],{"type":74,"value":1732},"Secrets such as ",{"type":68,"tag":232,"props":1734,"children":1736},{"className":1735},[],[1737],{"type":74,"value":1738},"NGC_API_KEY",{"type":74,"value":1740}," must be supplied by the user environment.",{"type":68,"tag":77,"props":1742,"children":1743},{"id":51},[1744],{"type":74,"value":1745},"Troubleshooting",{"type":68,"tag":191,"props":1747,"children":1748},{},[1749,1765],{"type":68,"tag":195,"props":1750,"children":1751},{},[1752],{"type":68,"tag":199,"props":1753,"children":1754},{},[1755,1760],{"type":68,"tag":203,"props":1756,"children":1757},{},[1758],{"type":74,"value":1759},"Error \u002F signal",{"type":68,"tag":203,"props":1761,"children":1762},{},[1763],{"type":74,"value":1764},"What to do",{"type":68,"tag":214,"props":1766,"children":1767},{},[1768,1788,1806],{"type":68,"tag":199,"props":1769,"children":1770},{},[1771,1776],{"type":68,"tag":221,"props":1772,"children":1773},{},[1774],{"type":74,"value":1775},"Services are not running",{"type":68,"tag":221,"props":1777,"children":1778},{},[1779,1781,1786],{"type":74,"value":1780},"Follow ",{"type":68,"tag":232,"props":1782,"children":1784},{"className":1783},[],[1785],{"type":74,"value":237},{"type":74,"value":1787}," before configuring features.",{"type":68,"tag":199,"props":1789,"children":1790},{},[1791,1796],{"type":68,"tag":221,"props":1792,"children":1793},{},[1794],{"type":74,"value":1795},"Restart or health check fails",{"type":68,"tag":221,"props":1797,"children":1798},{},[1799,1800,1805],{"type":74,"value":1780},{"type":68,"tag":232,"props":1801,"children":1803},{"className":1802},[],[1804],{"type":74,"value":268},{"type":74,"value":313},{"type":68,"tag":199,"props":1807,"children":1808},{},[1809,1814],{"type":68,"tag":221,"props":1810,"children":1811},{},[1812],{"type":74,"value":1813},"User requests teardown",{"type":68,"tag":221,"props":1815,"children":1816},{},[1817,1818,1823],{"type":74,"value":1780},{"type":68,"tag":232,"props":1819,"children":1821},{"className":1820},[],[1822],{"type":74,"value":286},{"type":74,"value":1824}," and confirm destructive cleanup.",{"type":68,"tag":666,"props":1826,"children":1828},{"id":1827},"when-user-says-configure-without-specifics",[1829],{"type":74,"value":1830},"When User Says \"Configure\" Without Specifics",{"type":68,"tag":84,"props":1832,"children":1833},{},[1834],{"type":74,"value":1835},"Run steps 2–3 above, then read the identified config file to list what's currently enabled:",{"type":68,"tag":686,"props":1837,"children":1839},{"className":688,"code":1838,"language":690,"meta":691,"style":691},"grep -E \"^(export )?(ENABLE_|APP_)\" \u003Cconfig-file> 2>\u002Fdev\u002Fnull | sort\n",[1840],{"type":68,"tag":232,"props":1841,"children":1842},{"__ignoreMap":691},[1843],{"type":68,"tag":697,"props":1844,"children":1845},{"class":699,"line":700},[1846,1851,1855,1859,1864,1868,1873,1878,1884,1889,1893,1897,1901],{"type":68,"tag":697,"props":1847,"children":1848},{"style":732},[1849],{"type":74,"value":1850},"grep",{"type":68,"tag":697,"props":1852,"children":1853},{"style":716},[1854],{"type":74,"value":1073},{"type":68,"tag":697,"props":1856,"children":1857},{"style":710},[1858],{"type":74,"value":713},{"type":68,"tag":697,"props":1860,"children":1861},{"style":716},[1862],{"type":74,"value":1863},"^(export )?(ENABLE_|APP_)",{"type":68,"tag":697,"props":1865,"children":1866},{"style":710},[1867],{"type":74,"value":724},{"type":68,"tag":697,"props":1869,"children":1870},{"style":710},[1871],{"type":74,"value":1872}," \u003C",{"type":68,"tag":697,"props":1874,"children":1875},{"style":716},[1876],{"type":74,"value":1877},"config-fil",{"type":68,"tag":697,"props":1879,"children":1881},{"style":1880},"--shiki-light:#90A4AE;--shiki-default:#EEFFFF;--shiki-dark:#BABED8",[1882],{"type":74,"value":1883},"e",{"type":68,"tag":697,"props":1885,"children":1886},{"style":710},[1887],{"type":74,"value":1888},">",{"type":68,"tag":697,"props":1890,"children":1891},{"style":710},[1892],{"type":74,"value":765},{"type":68,"tag":697,"props":1894,"children":1895},{"style":716},[1896],{"type":74,"value":770},{"type":68,"tag":697,"props":1898,"children":1899},{"style":710},[1900],{"type":74,"value":775},{"type":68,"tag":697,"props":1902,"children":1903},{"style":732},[1904],{"type":74,"value":1905}," sort\n",{"type":68,"tag":84,"props":1907,"children":1908},{},[1909],{"type":74,"value":1910},"Summarize what's running and enabled, then ask which feature to change.",{"type":68,"tag":293,"props":1912,"children":1913},{},[],{"type":68,"tag":77,"props":1915,"children":1917},{"id":1916},"hardware-restrictions",[1918],{"type":74,"value":1919},"Hardware Restrictions",{"type":68,"tag":84,"props":1921,"children":1922},{},[1923,1925,1931,1933,1939],{"type":74,"value":1924},"Read ",{"type":68,"tag":232,"props":1926,"children":1928},{"className":1927},[],[1929],{"type":74,"value":1930},"docs\u002Fsupport-matrix.md",{"type":74,"value":1932}," for current GPU requirements per deployment mode.\nRead ",{"type":68,"tag":232,"props":1934,"children":1936},{"className":1935},[],[1937],{"type":74,"value":1938},"docs\u002Fservice-port-gpu-reference.md",{"type":74,"value":1940}," for port mappings and GPU assignments.",{"type":68,"tag":191,"props":1942,"children":1943},{},[1944,1960],{"type":68,"tag":195,"props":1945,"children":1946},{},[1947],{"type":68,"tag":199,"props":1948,"children":1949},{},[1950,1955],{"type":68,"tag":203,"props":1951,"children":1952},{},[1953],{"type":74,"value":1954},"GPU",{"type":68,"tag":203,"props":1956,"children":1957},{},[1958],{"type":74,"value":1959},"Feature Restrictions",{"type":68,"tag":214,"props":1961,"children":1962},{},[1963,1984],{"type":68,"tag":199,"props":1964,"children":1965},{},[1966,1971],{"type":68,"tag":221,"props":1967,"children":1968},{},[1969],{"type":74,"value":1970},"B200",{"type":68,"tag":221,"props":1972,"children":1973},{},[1974,1976,1982],{"type":74,"value":1975},"No VLM, No Guardrails, No Nemotron Parse. May need multi-GPU LLM (",{"type":68,"tag":232,"props":1977,"children":1979},{"className":1978},[],[1980],{"type":74,"value":1981},"LLM_MS_GPU_ID",{"type":74,"value":1983},").",{"type":68,"tag":199,"props":1985,"children":1986},{},[1987,1992],{"type":68,"tag":221,"props":1988,"children":1989},{},[1990],{"type":74,"value":1991},"RTX PRO 6000",{"type":68,"tag":221,"props":1993,"children":1994},{},[1995],{"type":74,"value":1996},"No Nemotron Parse. No Audio on Helm.",{"type":68,"tag":1998,"props":1999,"children":2000},"style",{},[2001],{"type":74,"value":2002},"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":2004,"total":2104},[2005,2022,2032,2046,2058,2075,2090],{"slug":2006,"name":2006,"fn":2007,"description":2008,"org":2009,"tags":2010,"stars":26,"repoUrl":27,"updatedAt":2021},"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},[2011,2014,2017,2018],{"name":2012,"slug":2013,"type":15},"Data Analysis","data-analysis",{"name":2015,"slug":2016,"type":15},"Data Engineering","data-engineering",{"name":9,"slug":8,"type":15},{"name":2019,"slug":2020,"type":15},"Performance","performance","2026-07-14T05:28:43.176466",{"slug":2023,"name":2023,"fn":2024,"description":2025,"org":2026,"tags":2027,"stars":26,"repoUrl":27,"updatedAt":2031},"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},[2028,2029,2030],{"name":20,"slug":21,"type":15},{"name":24,"slug":25,"type":15},{"name":9,"slug":8,"type":15},"2026-07-14T05:29:06.667109",{"slug":2033,"name":2033,"fn":2034,"description":2035,"org":2036,"tags":2037,"stars":26,"repoUrl":27,"updatedAt":2045},"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},[2038,2041,2042],{"name":2039,"slug":2040,"type":15},"Agents","agents",{"name":9,"slug":8,"type":15},{"name":2043,"slug":2044,"type":15},"Research","research","2026-07-14T05:28:06.816956",{"slug":2047,"name":2047,"fn":2048,"description":2049,"org":2050,"tags":2051,"stars":26,"repoUrl":27,"updatedAt":2057},"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},[2052,2053,2054],{"name":2012,"slug":2013,"type":15},{"name":9,"slug":8,"type":15},{"name":2055,"slug":2056,"type":15},"Testing","testing","2026-07-17T05:29:03.913266",{"slug":2059,"name":2059,"fn":2060,"description":2061,"org":2062,"tags":2063,"stars":26,"repoUrl":27,"updatedAt":2074},"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},[2064,2067,2070,2071],{"name":2065,"slug":2066,"type":15},"Automation","automation",{"name":2068,"slug":2069,"type":15},"Imaging","imaging",{"name":9,"slug":8,"type":15},{"name":2072,"slug":2073,"type":15},"Video","video","2026-07-17T05:28:53.905004",{"slug":2076,"name":2076,"fn":2077,"description":2078,"org":2079,"tags":2080,"stars":26,"repoUrl":27,"updatedAt":2089},"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},[2081,2082,2085,2086],{"name":20,"slug":21,"type":15},{"name":2083,"slug":2084,"type":15},"Docker","docker",{"name":9,"slug":8,"type":15},{"name":2087,"slug":2088,"type":15},"Operations","operations","2026-07-17T05:28:56.913999",{"slug":2091,"name":2091,"fn":2092,"description":2093,"org":2094,"tags":2095,"stars":26,"repoUrl":27,"updatedAt":2103},"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},[2096,2097,2100],{"name":9,"slug":8,"type":15},{"name":2098,"slug":2099,"type":15},"Quantum Computing","quantum-computing",{"name":2101,"slug":2102,"type":15},"Simulation","simulation","2026-07-14T05:26:58.898253",305,{"items":2106,"total":2256},[2107,2125,2140,2151,2163,2177,2190,2204,2215,2224,2238,2247],{"slug":2108,"name":2108,"fn":2109,"description":2110,"org":2111,"tags":2112,"stars":2122,"repoUrl":2123,"updatedAt":2124},"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},[2113,2116,2119],{"name":2114,"slug":2115,"type":15},"Documentation","documentation",{"name":2117,"slug":2118,"type":15},"MCP","mcp",{"name":2120,"slug":2121,"type":15},"Search","search",21777,"https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FNemoClaw","2026-07-20T06:00:01.461044",{"slug":2126,"name":2126,"fn":2127,"description":2128,"org":2129,"tags":2130,"stars":2137,"repoUrl":2138,"updatedAt":2139},"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},[2131,2134,2135],{"name":2132,"slug":2133,"type":15},"Containers","containers",{"name":20,"slug":21,"type":15},{"name":2136,"slug":53,"type":15},"Python",17049,"https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FMegatron-LM","2026-07-27T06:06:11.249662",{"slug":2141,"name":2141,"fn":2142,"description":2143,"org":2144,"tags":2145,"stars":2137,"repoUrl":2138,"updatedAt":2150},"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},[2146,2149],{"name":2147,"slug":2148,"type":15},"CI\u002FCD","ci-cd",{"name":20,"slug":21,"type":15},"2026-07-14T05:25:59.97109",{"slug":2152,"name":2152,"fn":2153,"description":2154,"org":2155,"tags":2156,"stars":2137,"repoUrl":2138,"updatedAt":2162},"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},[2157,2158,2159],{"name":2147,"slug":2148,"type":15},{"name":20,"slug":21,"type":15},{"name":2160,"slug":2161,"type":15},"GitHub","github","2026-07-27T06:06:12.278222",{"slug":2164,"name":2164,"fn":2165,"description":2166,"org":2167,"tags":2168,"stars":2137,"repoUrl":2138,"updatedAt":2176},"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},[2169,2172,2173],{"name":2170,"slug":2171,"type":15},"Debugging","debugging",{"name":2160,"slug":2161,"type":15},{"name":2174,"slug":2175,"type":15},"Triage","triage","2026-07-14T05:25:57.442089",{"slug":2178,"name":2178,"fn":2179,"description":2180,"org":2181,"tags":2182,"stars":2137,"repoUrl":2138,"updatedAt":2189},"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},[2183,2186],{"name":2184,"slug":2185,"type":15},"Best Practices","best-practices",{"name":2187,"slug":2188,"type":15},"Code Analysis","code-analysis","2026-07-14T05:25:56.18433",{"slug":2191,"name":2191,"fn":2192,"description":2193,"org":2194,"tags":2195,"stars":2137,"repoUrl":2138,"updatedAt":2203},"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},[2196,2199,2202],{"name":2197,"slug":2198,"type":15},"Machine Learning","machine-learning",{"name":2200,"slug":2201,"type":15},"Migration","migration",{"name":9,"slug":8,"type":15},"2026-07-17T06:07:11.777011",{"slug":2205,"name":2205,"fn":2206,"description":2207,"org":2208,"tags":2209,"stars":2137,"repoUrl":2138,"updatedAt":2214},"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},[2210,2213],{"name":2211,"slug":2212,"type":15},"QA","qa",{"name":2055,"slug":2056,"type":15},"2026-07-14T05:25:53.673039",{"slug":2216,"name":2216,"fn":2217,"description":2218,"org":2219,"tags":2220,"stars":2137,"repoUrl":2138,"updatedAt":2223},"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},[2221,2222],{"name":20,"slug":21,"type":15},{"name":24,"slug":25,"type":15},"2026-07-14T05:25:49.362534",{"slug":2225,"name":2225,"fn":2226,"description":2227,"org":2228,"tags":2229,"stars":2137,"repoUrl":2138,"updatedAt":2237},"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},[2230,2233,2234],{"name":2231,"slug":2232,"type":15},"Code Review","code-review",{"name":2160,"slug":2161,"type":15},{"name":2235,"slug":2236,"type":15},"Pull Requests","pull-requests","2026-07-14T05:26:01.226578",{"slug":2239,"name":2239,"fn":2240,"description":2241,"org":2242,"tags":2243,"stars":2137,"repoUrl":2138,"updatedAt":2246},"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},[2244,2245],{"name":2211,"slug":2212,"type":15},{"name":2055,"slug":2056,"type":15},"2026-07-14T05:25:54.928983",{"slug":2248,"name":2248,"fn":2249,"description":2250,"org":2251,"tags":2252,"stars":2137,"repoUrl":2138,"updatedAt":2255},"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},[2253,2254],{"name":2065,"slug":2066,"type":15},{"name":2147,"slug":2148,"type":15},"2026-07-30T05:29:03.275638",496]