Cudaq Guide 測試 Skill
CUDA-Q onboarding 指南 用於 installation, test programs, GPU simulation, QPU hardware, 與 quantum applications。
#NVIDIA
這個標籤底下共有 155 篇內容。
Cuopt Developer 測試 Skill
Modify, build, test, debug, 與 contribute to NVIDIA cuOpt (C++/CUDA, Python, server, CI)。
Cuopt Install 雲端部署 Skill
Install cuOpt 用於 Python, C, 或 as a server (pip, conda, Docker) — system requirements, install commands, 與 verification。
Cuopt Numerical Optimization API C Skill
LP, MILP, 與 QP (beta) 搭配 cuOpt — C API only。
Cuopt Numerical Optimization API CLI Skill
LP, MILP, 與 QP (beta) 搭配 cuOpt — CLI only (MPS files, cuopt_cli)。
Cuopt Numerical Optimization API Python Skill
Solve Linear Programming (LP), Mixed-Integer Linear Programming (MILP), 與 Quadratic Programming (QP, beta) 搭配 the Python API。
Cuopt Routing API Python Skill
Vehicle routing (VRP, TSP, PDP) 搭配 cuOpt — Python API only。
Cuopt Server API Python Skill
協助處理 Cuopt Server API Python 相關工作，並依原始 Skill 說明完成設定與執行。
Cuopt Server Common Skill
cuOpt REST server — what it does 與 how requests flow。
Cuopt USER Rules Skill
Base rules 用於 end users calling NVIDIA cuOpt (routing/LP/MILP/QP/install/server)。
Numerical Optimization Formulation Skill
Numerical optimization (LP, MILP, QP) — concepts, problem-text parsing, 與 formulation 模式。
Routing Formulation 設計 Skill
Vehicle routing (VRP, TSP, PDP) — problem types 與 data requirements。
Skill Evolution Skill
After solving a non-trivial problem, detect generalizable learnings 與 propose skill updates so future interactions benefit automatically。
DALI Dynamic MODE 雲端部署 Skill
Use when writing DALI data loading 或 preprocessing code 搭配 `nvidia.dali.experimental.dynamic` (ndd), 或 when converting DALI pipeline-mode code to dynamic mode, 或 when the user..。
Deepstream DEV Skill
NVIDIA DeepStream SDK 9.0 development 搭配 Python pyservicemaker API。
Deepstream Import Vision Model 雲端部署 Skill
> Use this skill to bring any vision model from HuggingFace 或 NVIDIA NGC into an NVIDIA DeepStream pipeline 搭配 端到端 automation: ONNX download, SafeTensors export, TRT engi..。
Adding Model Support 設計 Skill
指南 用於 adding support 用於 new LLM 或 VLM models in Megatron-Bridge。
Build AND Dependency Skill
Dev environment setup 用於 Megatron Bridge — container-based development, uv package management, lockfile regeneration, adding dependencies, Slurm container usage, 與 common build..。
BUMP Dependency 雲端部署 Skill
Bump a pinned dependency (TransformerEngine, Megatron-LM, NRX, etc.), regenerate the lockfile, open a PR, 與 drive it to green by attaching a watchdog to the "CICD NeMo" 工作流..。
CI/CD 雲端部署 Skill
CI/CD 參考資料 用於 Megatron Bridge — pipeline structure, commit 與 PR 工作流, CI failure investigation, 與 common failure 模式。
Linting AND Formatting Skill
Code style 與 quality rules 用於 Megatron Bridge — ruff configuration, naming conventions, type hints, mypy rules, docstrings, copyright headers, logging, 與 the code review check..。
MLM Bridge Training Skill
Run Megatron-LM (MLM) 與 Megatron Bridge training 搭配 mock 或 real data。
Multi NODE Slurm Skill
Convert single-node scripts to multi-node Slurm sbatch jobs 與 debug common multi-node failures。
NEMO RL E2E Testing Skill
External NeMo-RL 端到端 validation 工作流 用於 Megatron-Bridge model/provider changes, including downstream compatibility checks, external RL lifecycle behavior, Megatron poli..。
Parity Testing Skill
Structured framework 用於 verifying numerical parity of HF<->MCore weight conversions。
PERF Activation Recompute Skill
Validate 與 use selective 與 full activation recompute in Megatron Bridge to reduce GPU memory usage at the cost of extra compute。
PERF CPU Offloading Skill
Validate 與 use CPU offloading in Megatron Bridge, including layer-level activation offloading 與 fractional optimizer state offloading 搭配 HybridDeviceOptimizer。
PERF CUDA Graphs Skill
Validate 與 use CUDA graph capture in Megatron Bridge, including local full-iteration graphs 與 Transformer Engine scoped graphs 用於 attention, MLP, 與 MoE modules。
PERF Expert Parallel Overlap Skill
Validate 與 use MoE expert-parallel communication overlap in Megatron-Bridge, including overlap_moe_expert_parallel_comm, delay_wgrad_compute, 與 flex dispatcher backends such as..。
PERF Hierarchical Context Parallel 設計 Skill
Operational 指南 用於 enabling hierarchical context parallelism in Megatron-Bridge, including config knobs, code anchors, pitfalls, 與 verification。
PERF Megatron FSDP 設計 Skill
Operational 指南 用於 enabling Megatron FSDP in Megatron-Bridge, including config knobs, code anchors, pitfalls, 與 verification。
PERF Memory Tuning Skill
Techniques 用於 reducing peak GPU memory in Megatron Bridge — expandable segments, parallelism resizing, activation recompute, CPU offloading constraints, 與 common OOM fixes。
PERF MOE COMM Overlap Skill
協助處理 PERF MOE COMM Overlap 相關工作，並依原始 Skill 說明完成設定與執行。
PERF MOE Dispatcher Selection Skill
Choose the right MoE token dispatcher (`alltoall`, DeepEP, 或 HybridEP) 用於 the hardware, EP degree, 與 optimization stage。
PERF MOE Hardware Configs Skill
Representative MoE training playbooks by hardware platform 與 model family。
PERF MOE LONG Context 設計 Skill
Long-context MoE training guidance 用於 Megatron Bridge。
PERF MOE Optimization Workflow Skill
Systematic 工作流 用於 MoE training optimization in Megatron Bridge, based on the Megatron-Core MoE paper。
PERF MOE VLM Training 設計 Skill
Practical guidance 用於 training MoE VLMs in Megatron Bridge。
PERF Parallelism Strategies 設計 Skill
Operational 指南 用於 choosing 與 combining parallelism strategies in Megatron Bridge, including sizing rules, hardware topology mapping, 與 combined parallelism configuration。
PERF Sequence Packing 設計 Skill
Validate 與 use packed sequences 與 long-context training in Megatron-Bridge, distinguishing offline packed SFT 用於 LLMs from in-batch packing 用於 VLMs, 與 applying the right CP..。
PERF TP DP COMM Overlap 設計 Skill
Operational 指南 用於 enabling TP, DP, 與 PP communication overlap in Megatron-Bridge, including config knobs, code anchors, pitfalls, 與 verification。
Recipe Recommender Skill
Recommend 與 customize Megatron Bridge recipes 用於 a user's model, GPU count, 與 training goal。
Resiliency Skill
Resiliency features in Megatron Bridge including fault tolerance, straggler detection, in-process restart, preemption, 與 re-run state machine。
Testing Skill
Testing 參考資料 用於 Megatron Bridge — unit 與 functional test layout, tier semantics (L0/L1/L2/flaky), script conventions, running 測試 locally, adding/moving/disabling 測試,..。
VERL E2E Testing Skill
External verl 端到端 validation 工作流 用於 Megatron-Bridge model/provider changes。
Build AND Dependency Skill
Container-based dev environment setup 與 dependency management 用於 Megatron-LM。
BUMP BASE Image 設計 Skill
協助處理 BUMP BASE Image 設計 相關工作，並依原始 Skill 說明完成設定與執行。
CI/CD 雲端部署 Skill
CI/CD 參考資料 用於 Megatron-LM。
Create Issue 雲端部署 Skill
Investigate a failing GitHub Actions run 或 job 與 create a GitHub issue 用於 the failure。
Linting AND Formatting Skill
Linting 與 formatting 用於 Megatron-LM。
Nightly SYNC Skill
Domain knowledge 用於 the nightly main-to-dev sync 工作流。
Onboard Gb200 1node Tests Skill
Onboard 1-node GitHub MR functional 測試 用於 GB200 from existing mr-scoped 2-node 測試。
Respond TO Issue Skill
Research 與 draft a response to a GitHub issue 或 question from an external contributor。
RUN ON Slurm Skill
協助處理 RUN ON Slurm 相關工作，並依原始 Skill 說明完成設定與執行。
Split PR 設計 Skill
協助處理 Split PR 設計 相關工作，並依原始 Skill 說明完成設定與執行。
Testing Skill
測試 system 用於 Megatron-LM。
Update Golden Values 雲端部署 Skill
Refresh golden values from a GitHub Actions 工作流 run (failing-only 或 all jobs), score the change 搭配 average normalized relative differences, 與 produce a PR-ready summary。
Accessing Mlflow 資料庫 Skill
Query 與 browse evaluation results stored in MLflow。
Debug 雲端部署 Skill
協助處理 Debug 雲端部署 相關工作，並依原始 Skill 說明完成設定與執行。
Deployment 雲端部署 Skill
Serve a quantized 或 unquantized LLM checkpoint as an OpenAI-compatible API endpoint 使用 vLLM, SGLang, 或 TRT-LLM。
Evaluation Skill
Evaluates accuracy of quantized 或 unquantized LLMs 使用 NeMo Evaluator Launcher (NEL)。
Launching Evals Skill
Run, monitor, analyze, 與 debug LLM evaluations via nemo-evaluator-launcher。
Monitor 雲端部署 Skill
Monitor submitted jobs (PTQ, evaluation, 部署) on SLURM clusters。
PTQ Skill
協助處理 PTQ 相關工作，並依原始 Skill 說明完成設定與執行。
Release Cherry PICK Skill
Cherry-pick merged PRs labeled 用於 a release branch into that branch, then open a PR 與 apply the cherry-pick-done label。
BYOB Skill
建立 custom LLM evaluation benchmarks 使用 the BYOB decorator framework。
Accessing Mlflow 資料庫 Skill
Query 與 browse evaluation results stored in MLflow。
Launching Evals Skill
Run, monitor, analyze, 與 debug LLM evaluations via nemo-evaluator-launcher。
NEL Assistant Skill
Interactive config wizard 用於 NeMo Evaluator Launcher (NEL)。
ADD Benchmark 設計 Skill
> 指南 用於 adding a new benchmark 或 training environment to NeMo-Gym。
NEMO GYM Debugging Skill
>- Use when debugging a Nemo Gym run 或 reward profiling job。
NEMO GYM DOCS Skill
> Maintain the NeMo Gym Fern docs site — add, update, move, 或 remove pages under fern/。
NEMO GYM Pivot Datasets Skill
>- Use when creating, validating, 或 documenting Nemo Gym pivot datasets from rollout, trajectory, chat-completion, Responses API, 或 tool-call artifacts。
NEMO GYM Reward Profiling Skill
>- Use to help users get started 搭配 Nemo Gym reward profiling。
AUTO Research 測試 Skill
Autonomous NeMo-RL research agent 工作流 用於 directed hypothesis testing 與 open-ended discovery。
BREV Etiquette Skill
Brev instance operating guidance 用於 NeMo-RL agents working in /home/ubuntu/RL 搭配 limited workspace disk, a larger /ephemeral volume, 與 optional /home/ubuntu/RL/.env secrets。
Build AND Dependency Skill
建置 and dependency management 用於 NeMo-RL。
CI/CD 雲端部署 Skill
CI/CD 參考資料 用於 NeMo-RL。
Config Conventions Skill
Configuration conventions 用於 NeMo-RL。
Contributing Skill
Contribution conventions 用於 NeMo-RL。
Copyright 設計 Skill
NVIDIA copyright header requirements 用於 NeMo-RL。
DOCS Skill
檔案 conventions 用於 NeMo-RL。
Error Handling 設計 Skill
Error handling 指南lines 用於 NeMo-RL。
Launch NEMO RL 雲端部署 Skill
Playbook 用於 launching, monitoring, stopping, 與 debugging NeMo-RL recipes on a Kubernetes cluster via the nrl-k8s CLI。
Linting AND Formatting 設計 Skill
Code style 指南lines 用於 NeMo-RL (Python 與 shell)。
Review PR Skill
Interactive code review 用於 NVIDIA-NeMo/RL pull requests。
Session Memory 身分驗證 Skill
Manage durable working-session memory 用於 coding agents。
Testing Skill
Testing conventions 用於 NeMo-RL。
Nemoclaw Contributor Create PR Skill
建立 GitHub pull requests that follow the NemoClaw PR template。
Nemoclaw Contributor Update DOCS Skill
Scan recent git commits 用於 changes that affect user-facing behavior, then draft 或 update the corresponding 檔案 pages 與 refresh generated user skills 用於 release prep。
Nemoclaw Maintainer Cross Issue Sweep Skill
Scans other open issues to find ones a given PR may also fix 或 accidentally break。
Nemoclaw Maintainer CUT Release TAG Skill
Cut a new semver release — bump all version strings via bump-version.ts, open a release PR, 與 after merge tag main 與 push。
Nemoclaw Maintainer DAY Skill
Runs the daytime maintainer loop 用於 NemoClaw, prioritizing items labeled 搭配 the current version target。
Nemoclaw Maintainer Evening Skill
Runs the end-of-day maintainer handoff 用於 NemoClaw。
Nemoclaw Maintainer FIND Review PR Skill
Finds open GitHub PRs 搭配 security 與 priority-high labels, links each to its issue, detects duplicates (multiple PRs fixing the same issue), 與 presents a table of review candi..。
Nemoclaw Maintainer Morning Skill
Runs the morning maintainer standup 用於 NemoClaw。
Nemoclaw Maintainer Normalize Title TAGS Skill
Normalizes GitHub issue 與 PR titles by removing any bracketed [NemoClaw] tag case-insensitively, even when the tag appears later in the title。
Nemoclaw Maintainer PR Comparator Skill
Compares competing PRs that target the same issue 與 recommends which one to merge。
Nemoclaw Maintainer Security CODE Review Skill
Performs a comprehensive security review of code changes in a GitHub PR 或 issue。
Nemoclaw Maintainer Triage Skill
AI-assisted label triage 用於 NVIDIA/NemoClaw issues 與 PRs。
Nemoclaw Skills Guide Skill
協助處理 Nemoclaw Skills Guide 相關工作，並依原始 Skill 說明完成設定與執行。
Nemoclaw USER Agent Skill
Describes the agent skills shipped 搭配 NemoClaw 與 how to access them by cloning the repository。
Nemoclaw USER Configure Inference Skill
協助處理 Nemoclaw USER Configure Inference 相關工作，並依原始 Skill 說明完成設定與執行。
Nemoclaw USER Configure Security Skill
Presents a risk framework 用於 every configurable security control in NemoClaw。
Nemoclaw USER Deploy Remote 雲端部署 Skill
Explains how to run NemoClaw on a remote GPU instance, including the deprecated Brev compatibility path 與 the preferred installer plus onboard flow。
Nemoclaw USER GET Started Skill
Installs NemoClaw, launches a sandbox, 與 runs the first agent prompt。
Nemoclaw USER Manage Policy Skill
Adds, removes, 或 modifies allowed endpoints in the sandbox policy。
Nemoclaw USER Manage Sandboxes 設計 Skill
Explains operational tasks after the quickstart: listing sandboxes, status 與 health checks, logs, diagnostics, port forwards, multiple sandboxes, credential reset, rebuilds, netw..。
Nemoclaw USER Monitor Sandbox Skill
Inspects sandbox health, traces agent behavior, 與 diagnoses problems。
Nemoclaw USER Overview Skill
Explains how OpenClaw, OpenShell, 與 NemoClaw form the ecosystem, NemoClaw's position in the stack, what NemoClaw adds beyond the community sandbox, 與 when to prefer NemoClaw ve..。
Nemoclaw USER Reference Skill
Describes the NemoClaw plugin 與 blueprint architecture 與 how they orchestrate the OpenClaw sandbox。
Nemotron Voice Agent Deploy 雲端部署 Skill
部署 Nemotron Voice Agent on Workstation (x86), Jetson Thor, 或 Cloud NIMs。
RAG Blueprint 雲端部署 Skill
"NVIDIA RAG Blueprint — deploy, configure, troubleshoot, 與 manage。
AD Accuracy Debug 雲端部署 Skill
> Debug AutoDeploy accuracy regressions vs a 參考資料 score (PyTorch backend 或 published baseline)。
AD ADD Fusion Transformation Skill
> Claude Code skill (trtllm-agent-toolkit): implement 或 extend TensorRT-LLM AutoDeploy fusion transforms under transform/library/ in a TensorRT-LLM checkout。
AD CONF Check 雲端部署 Skill
> Check whether AutoDeploy YAML configs were actually applied by analyzing server logs 與 optionally graph dumps (AD_DUMP_GRAPHS_DIR)。
AD Graph DUMP 雲端部署 Skill
> Enable 與 interpret TensorRT-LLM AutoDeploy FX graph text dumps via AD_DUMP_GRAPHS_DIR。
AD Layer Visualizer 雲端部署 Skill
協助處理 AD Layer Visualizer 雲端部署 相關工作，並依原始 Skill 說明完成設定與執行。
AD Model Onboard 測試 Skill
> Translates a HuggingFace model into a prefill-only AutoDeploy custom model 使用 參考資料 custom ops, validates 搭配 hierarchical equivalence 測試。
EXEC Local Compile 雲端部署 Skill
協助處理 EXEC Local Compile 雲端部署 相關工作，並依原始 Skill 說明完成設定與執行。
EXEC Slurm Compile Skill
協助處理 EXEC Slurm Compile 相關工作，並依原始 Skill 說明完成設定與執行。
Kernel CUTE Writing Skill
> Write 與 implement GPU kernels 使用 NVIDIA CuTe DSL (CUTLASS 4.x Python API) — NOT 用於 Triton, CUDA C++, 或 conceptual explanations。
Kernel Tileir Optimization Skill
> Optimize existing Triton kernels 用於 NVIDIA TileIR backend on Blackwell GPUs (sm_100+)。
Kernel Triton Writing Skill
> ONLY 用於 OpenAI Triton (@triton.jit) kernel development。
PERF Analysis Skill
> Performance analysis coordination 工作流。
PERF HOST Analysis Skill
協助處理 PERF HOST Analysis 相關工作，並依原始 Skill 說明完成設定與執行。
PERF HOST Optimization Skill
Profiles 與 optimizes TensorRT-LLM host/CPU overhead 使用 line_profiler (搭配 nsys support planned)。
PERF Nsight Compute Analysis Skill
協助處理 PERF Nsight Compute Analysis 相關工作，並依原始 Skill 說明完成設定與執行。
PERF Nsight Systems Skill
>- Nsight Systems (nsys) CLI 用於 system-level timeline profiling。
PERF Optimization Skill
協助處理 PERF Optimization 相關工作，並依原始 Skill 說明完成設定與執行。
PERF Torch CUDA Graphs Skill
協助處理 PERF Torch CUDA Graphs 相關工作，並依原始 Skill 說明完成設定與執行。
PERF Torch SYNC FREE Skill
>- Identify 與 eliminate host-device synchronizations in PyTorch code。
PERF Workload Profiling Skill
> Code instrumentation 用於 timing workloads。
Trtllm CODE Contribution Skill
> 最佳實務 用於 contributing code to TensorRT-LLM。
Trtllm Codebase Exploration Skill
> Systematic approach to exploring the TensorRT-LLM codebase before implementing new features 或 optimizations。
Trtllm Flashinfer Upgrade Skill
協助處理 Trtllm Flashinfer Upgrade 相關工作，並依原始 Skill 說明完成設定與執行。
Trtllm MOE Develop 測試 Skill
>- Review, design, 與 refactor TensorRT-LLM PyTorch MoE code 用於 architecture fit, clean code, maintainability, 與 testability。
Trtllm Serve Config Guide 雲端部署 Skill
產生 a source-backed starting `trtllm-serve --config` YAML 用於 basic aggregate single-node PyTorch serving, aligned 搭配 checked-in TensorRT-LLM configs 與 部署 docs。
Adding Cutile Kernel Skill
協助處理 Adding Cutile Kernel 相關工作，並依原始 Skill 說明完成設定與執行。
Converting Cutile TO Julia 設計 Skill
協助處理 Converting Cutile TO Julia 設計 相關工作，並依原始 Skill 說明完成設定與執行。
Converting Cutile TO Triton Skill
協助處理 Converting Cutile TO Triton 相關工作，並依原始 Skill 說明完成設定與執行。
Cutile Autotuning Skill
Use when adding, modifying, optimizing, 或 debugging CuTile autotuning code。
Cutile Python Skill
協助處理 Cutile Python 相關工作，並依原始 Skill 說明完成設定與執行。
Improve Cutile Kernel PERF Skill
Iteratively optimize cuTile kernel performance through systematic profiling, bottleneck analysis, IR comparison, 與 targeted tuning。
Monkey Patch Kernels TO Transformers Skill
Integrate TileGym kernels into Hugging Face `transformers` models by replacing the library's submodule(s) 與 certain class(es)' implementations, 與 patching certain class(es)' in..。
Alerts 雲端部署 Skill
Manage 與 monitor VSS alerts after the alerts profile is deployed。
Deploy 雲端部署 Skill
Deploy, debug, 或 tear down any VSS profile 使用 a compose-centric 工作流 — config (dry-run) 搭配 env overrides, review resolved compose, then compose up。
Report 資料庫 Skill
Produce video analysis reports by discovering the deployed VSS agent, querying POST /generate 用於 a timestamped captioned summary of the clip, then formatting the agent reply as th..。
RT VLM 設計 Skill
> Use this skill when working 搭配 the RTVI VLM 或 RT-VLM microservice API on VSS 3.1。
Video Analytics 資料庫 Skill
Query video analytics data 與 metrics from Elastic search via the VA-MCP server (port 9901)。
Video Search 設計 Skill
Search video archives 使用 natural language — find events, objects, actions, 與 people across recorded video 使用 fusion search (Cosmos Embed1 semantic search + CV attribute sea..。
Video Summarization 設計 Skill
Summarize a video by calling the VLM NIM 或 the Long Video Summarization (LVS) microservice directly。
Video Understanding 設計 Skill
協助處理 Video Understanding 設計 相關工作，並依原始 Skill 說明完成設定與執行。
VIOS 資料庫 Skill
Query VIOS REST APIs: sensor list, recording timelines, video clip extraction, snapshot capture, add/delete sensors 與 streams。
VSS FRAG 設計 Skill
產生 video summary reports 使用 the VSS video_search_frag extension 搭配 Long Video Summarization (LVS), Enterprise RAG knowledge retrieval, 與 human-in-the-loop parameter co..。