Loading...
Loading...
Found 315 Skills
Systematic workflow for MoE training optimization in Megatron Bridge, based on the Megatron-Core MoE paper. Covers the Three Walls framework, parallel folding, recompute strategy, dispatcher choice, and CUDA-graph bring-up.
Modify, build, test, debug, and contribute to NVIDIA cuOpt (C++/CUDA, Python, server, CI). Use for solver internals, PRs, DCO, and code conventions.
Deploy, debug, or tear down any VSS profile using a compose-centric workflow — config (dry-run) with env overrides, review resolved compose, then compose up. Use this skill when the user says "deploy vss", "deploy `profile`", "debug deploy", "verify deployment", or "why is my vss deploy broken".
Use when debugging a Nemo Gym run or reward profiling job. Covers rollout collection failures, empty or partial JSONL outputs, stale materialized inputs, verifier/schema errors, Ray or Slurm issues, vLLM readiness, judge failures, tool/sandbox failures, cache problems, and throughput bottlenecks.
Generate video summary reports using the VSS video_search_frag extension with Long Video Summarization (LVS), Enterprise RAG knowledge retrieval, and human-in-the-loop parameter collection. Use when: user wants to generate a video summary, report, or analysis using the frag pipeline.
Code instrumentation for timing workloads. Two scenarios: (1) Training loop — inject manual timing to report per-iteration latency, throughput (samples/sec), and data load time. (2) Standalone kernel/op — write CUDA event timing code with warmup, per-iteration statistics, and anti-pattern avoidance. Also covers NVTX annotation for labeling profiler timelines. NOT for: running or analyzing profiler tools (nsys, ncu, Nsight Systems, Nsight Compute), writing kernels (Triton, CuTe, CUDA), applying optimizations (CUDA Graphs, gradient checkpointing, fusion), or interpreting roofline/SOL% metrics. Triggers: "measure throughput", "benchmark this function", "time my training loop", "samples per second", "NVTX annotate", "instrument my dataloader", "data load time", "kernel timing", "how do I time".
Query video analytics data and metrics from Elastic search via the VA-MCP server (port 9901). This includes incidents, alerts, sensor data, and metrics. Use for any question about violations, alerts, incidents, object counts, speeds, occupancy, or anything that requires looking up recorded events. This is the primary way to answer a question that requires incidents, alerts and other metrics such as people counts and violations.
Validate and use CUDA graph capture in Megatron Bridge, including local full-iteration graphs and Transformer Engine scoped graphs for attention, MLP, and MoE modules.
Operational guide for choosing and combining parallelism strategies in Megatron Bridge, including sizing rules, hardware topology mapping, and combined parallelism configuration.
Coordinate the end-to-end CAD/source-asset to SimReady workflow. Use for broad requests such as CAD to SimReady, source asset to simulation-ready USD, or prop packaging that require conversion, material/physics assignment, SimReady conformance, validation, and optional package creation; deploy or verify Content Agents services first when property assignment is enabled; route single-stage work through nested references.
Use only to generate or update a governance skill card for a specified existing agent skill directory. Do not use for explaining, listing, comparing, or discussing skill capabilities.
Use when the user wants to create a dataset, generate synthetic data, or build a data generation pipeline.