skilld
All skills
nvidia avatar

/deepstream-import-vision-model

@6d03410
by NVIDIA Corporationnvidia/skills3.5k stars
424

Use this skill to bring a supported object-detection vision model from HuggingFace or NVIDIA NGC into an NVIDIA DeepStream pipeline with end-to-end automation: ONNX download, SafeTensors export, TRT engine build, custom nvinfer bbox parser, multi-stream benchmark, and PDF report. Object detection models only.

Ask your Agent

Use this Skill: https://skilld.dev/gh/nvidia/skills/deepstream-import-vision-model

This session only. Nothing lands on disk.

referencesREADME.md

≈194 tokens on demand. Your agent reads this file only when SKILL.md points to it.

deepstream-import-vision-model — Reference Documents

Detailed phase guides for the deepstream-import-vision-model skill. Read the relevant file before starting each pipeline phase.

Document Pipeline Steps When to read
model-acquire.md Steps 1–3 Downloading from HuggingFace or NGC; ONNX vs SafeTensors detection and export
engine-build.md Steps 4–5 TensorRT dynamic engine build; trtexec BS=1 and BS=MAX_BS benchmarks
pipeline-run.md Steps 6–7 Custom nvinfer bbox parser; single-stream validation; multi-stream benchmark sweep
report-generation.md Step 8 5 benchmark charts; Markdown → HTML → PDF report generation

Source: SKILL.md on GitHub

Third-party checks

No alerts27d3 checks · Risk SAFE
  • Gen Agent Trust Hub27d

    This skill provides an automated pipeline for importing HuggingFace or NVIDIA NGC vision models into NVIDIA DeepStream. It follows industry best practices for model acquisition, TensorRT engine building, and benchmarking. All external downloads target trusted sources (HuggingFace and NVIDIA NGC) or well-known package registries, and all code traces back to the vendor (NVIDIA). No malicious patterns or data exfiltration risks were detected.

  • Socket27d

    No alerts

  • Snyk27d

    Risk: LOW · No issues

Provenance

Signed by skilld at 6d03410. This ties the file your Agent reads to that commit on GitHub. It does not review the instructions.

Last checked against GitHub yesterday.

Activeupdated last month

Capability

Other metadata
metadata
{
  "author": "Tushar Khinvasara <tkhinvasara@nvidia.com>",
  "owner": "Tushar Khinvasara <tkhinvasara@nvidia.com>",
  "service": "deepstream",
  "version": "1.5.2",
  "reviewed": "2026-08-04",
  "team": "deepstream-sdk",
  "tags": [
    "deepstream",
    "tensorrt",
    "object-detection",
    "import-vision-model"
  ],
  "languages": [
    "bash",
    "python",
    "cpp"
  ],
  "domain": "computer-vision"
}

README badge

README badge for nvidia/skills/deepstream-import-vision-model