NVIDIA Skills

Configure fault tolerance, straggler detection, preemption, in-process restart, and re-run state machine for Megatron Bridge training jobs.

Run TAO SDK jobs as Docker containers on a local or remote Docker daemon with NVIDIA GPU support, including preflight checks and credential handling.

Solve vehicle routing problems (TSP, VRP, PDP) using NVIDIA cuOpt's Python API with cost matrices, time windows, capacity constraints, and pickup-delivery pairs.

Guides users through downloading weather/climate data via Earth2Studio data source APIs, verifying variable support via the lexicon system, and generating a working Python fetch script.

Captures a baseline snapshot of a Jetson device's module model, L4T version, kernel, OS version, and power mode for performance testing or verification.

Finetune the NV-Generate-CTMR MAISI VAE/autoencoder on user-supplied CT or MRI NIfTI volumes using a staged config and datalist workflow.

Extract false-positive and false-negative gaps from VLM binary-classification-question predictions by comparing model responses against ground truth, producing a structured JSONL file and summary report for downstream root-cause analysis.

Trains, evaluates, exports, quantizes, and runs inference for a Grounding DINO model that detects objects described by text prompts without a fixed class vocabulary.

Calibrate a new dataset from pre-recorded video files via the AutoMagicCalib REST API.

Validates that a Dynamo deployment's NIXL/UCX/NCCL interconnect is ready for disaggregated serving over RDMA/NVLink. Use after deploying a disagg or multi-node recipe to confirm KV transport is correct, or use troubleshoot for already-failed pods.

Add, remove, edit, list, or change the boot default of nvpmodel power modes on Jetson/Tegra platforms (Orin, Thor) by modifying the BSP-side configuration file.

Recommends an inference runtime and memory-related launch flags for LLM/VLM workloads on NVIDIA Jetson devices, based on a live memory audit snapshot.

Build synthetic datasets and data generation pipelines using the Data Designer library.

Generates custom safety policies for NVIDIA Nemotron content-safety guardrails, producing a Markdown policy, JSON taxonomy, and inference prompts from rough user input.

Coordinate the end-to-end pipeline from CAD or source assets to SimReady USD assets, including conversion, material and physics assignment, validation, conformance, and optional packaging.

Routes Omniverse Realtime Viewer requests to focused reference documents and enforces architectural rules for building USD viewer applications.

Performs deep root cause analysis on NVIDIA TAO Visual ChangeNet classification experiments, using image-evidence-driven investigation to diagnose model failures and produce actionable reports.

Defines vehicle routing problem types (TSP, VRP, PDP) and the data requirements needed to formulate them, without covering any API or interface details.

Fine-tune, evaluate, run inference, and export Cosmos-Embed1 video-text embedding models for tasks like text-to-video retrieval and semantic deduplication.

Start, query, and stop a TAO inference microservice for a specific network architecture by delegating container execution to the appropriate platform skill.

Checks and installs NVIDIA driver, CUDA Toolkit, and NVIDIA Container Toolkit for GPU-accelerated Docker and Kubernetes hosts. Supports multiple Linux distributions with automated install and read-only check modes.

Train, evaluate, export, quantize, and run inference for a Deformable DETR 2D object detection model using TAO, with deformable attention for efficient multi-scale feature processing.

Trains, evaluates, and runs inference for Mask Auto-Label (MAL) weakly-supervised segmentation models using ViT-MAE backbones with minimal point or box annotations.

Adds autotuning to CuTile kernels using the exhaustive_search API with a tune-once/cache/direct-launch pattern, covering occupancy-only and complex tile-size search spaces.

Generates video analysis reports by routing to a VLM backend for per-clip analysis or an analytics backend for incident-range reports, with deployment profile verification and URL rewriting.

Accelerate pandas workflows with GPU DataFrames using cuDF and dask-cuDF for ETL, joins, groupby, and large-scale data processing.

Run end-to-end calibration on the bundled sample dataset against a running AMC microservice to verify the stack works before processing real data.

Pull and verify the official Holoscan SDK container from NGC, selecting the correct CUDA/arch tag for the host GPU and validating with bundled Python and C++ examples.

Plan, debug, tune, evaluate, export, or deploy public Nemotron embedding and reranking retrieval recipes using the current checkout.

Converts NVIDIA TAO DAFT datasets between supported formats using the `tao-daft convert` CLI.

Fine-tune Cosmos Reason video QA models using supervised fine-tuning with FSDP parallelism, including dataset preparation, spec construction, and AutoML support.

Trains, evaluates, exports, and runs inference for Visual ChangeNet models used in AOI defect detection, comparing image pairs for PASS/NO_PASS classification or change-segmentation masks.

Deploy and operate the VSS 3.2 GA RT-Embed Video Embedding microservice using Docker Compose, covering GPU prerequisites, REST API usage for file uploads, text/video embeddings, live RTSP streams, Redis/Kafka/OTel integration, and troubleshooting.

Deploy the AutoMagicCalib microservice and web UI from pre-built NGC release images using Docker Compose.

Profile a DeepStream pipeline with Nsight Systems and derive its configs from the measurement.

Reduce per-token latency on Jetson vLLM servers by appending speculative decoding configuration, with guidance on when to enable and how to benchmark the improvement.

Train, evaluate, export, and run inference for monocular depth estimation models using Metric Depth Anything v2 or Relative Depth Anything architectures via the TAO toolkit.

Trains, evaluates, exports, and runs inference on FoundationStereo models for stereo depth estimation and 3D reconstruction from stereo image pairs.

Train, evaluate, export, and run inference for Masked Auto-Encoder (MAE) models for self-supervised pretraining and fine-tuning of visual representations.

Integrate a HuggingFace computer vision model into the NVIDIA TAO Toolkit ecosystem, covering the full pipeline from prerequisites to container testing.

Deploy a standalone RT-VLM dense-captioning microservice and exercise its REST API endpoints for file upload, caption generation, streaming, chat completions, and Kafka integration.

Manage VIOS and NvStreamer REST API operations for video input/output and storage, including sensors, streams, snapshots, clips, and recordings.

Builds and validates DeepStream GStreamer pipelines through an interactive questionnaire and a BM25 retrieval engine over 270+ verified pipelines.

Lint and format Python code for Megatron-LM using ruff, black, isort, pylint, and mypy, with commands for autoformatting and import ordering.

Generates phrase-grounded bounding box annotations from image-caption pairs using a VLM, producing cleaned captions, referring expressions, and pixel-space bounding boxes.

Runs a three-phase AOI training pipeline: AutoML HPO baseline, DEFT iterative data improvement, and AutoML refinement on the augmented dataset.

Train, evaluate, export, and run inference on TAO action-recognition models for classifying temporal actions in video clips using RGB, optical flow, or joint input.

Validates NVIDIA TAO DAFT datasets for structure, schema, and cross-reference errors using the `tao-daft validate` CLI tool.

Trains, evaluates, exports, and runs inference for Siamese-network-based optical inspection models to detect manufacturing defects and quality issues in image pairs.

Add a new cuTile GPU kernel operator to TileGym, covering dispatch registration, backend implementation, exports, tests, and benchmarks.

Deploy the behavior-analytics service standalone with a chosen entrypoint, config source, and optional calibration, without the full warehouse stack.

Convert single-node PyTorch distributed scripts into multi-node Slurm sbatch jobs and debug common multi-node failures, covering srun-native and torch.distributed approaches, container setup, NCCL timeouts, and interactive allocation.

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.

Finetunes the NV-Generate-CTMR MR-brain diffusion UNet from user-supplied NIfTI training volumes using a wrapper that stages configs and delegates to upstream scripts.

Trains, evaluates, exports, quantizes, and runs inference for a Mask Grounding DINO model for open-set instance segmentation guided by text prompts.

Train, evaluate, export, and run inference for pose classification models using ST-GCN on skeleton keypoint sequences.

Deploys the vss-video-analytics-api REST service standalone with config, data-log bind, and optional Elasticsearch/Kafka connectivity.

Load sharded datasets (npy, Parquet, HDF5, raw binary) into distributed cuPyNumeric arrays using manual partitioning and Legate task launches.

Import object detection models from HuggingFace or NVIDIA NGC into a DeepStream pipeline with automated ONNX download, TensorRT engine build, custom parser, multi-stream benchmark, and PDF report generation.

Create and validate Earth2Studio data source wrappers (DataSource, ForecastSource, DataFrameSource, ForecastFrameSource) from remote stores like S3, GCS, Azure, HTTP, or HuggingFace.