deepstream-import-vision-model — Claude Code skill: ONNX and TRT
This skill takes an object-detection model from HuggingFace or NGC through a full NVIDIA DeepStream pipeline: ONNX, TRT engine, bbox parser, benchmark, PDF report.
npx skills add NVIDIA/skills --skill deepstream-import-vision-model -a claude-codeWhat deepstream-import-vision-model does
The skill takes a supported object-detection model and carries it all the way into a running DeepStream pipeline. It starts by offering two choices: the validated PekingU/rtdetr_r50vd from Hugging Face, or your own model — either a Hugging Face ID in organization/model form or an NGC catalog URL with its version. Classification, segmentation and other architectures get rejected after config.json is inspected, and the skill will not quietly swap in a different model.
From there it runs eight steps: download ONNX or export from SafeTensors, build a dynamic TRT engine with trtexec runs at BS=1 and BS=MAX_BS, generate a custom bbox parser for nvinfer, benchmark single-stream and multi-stream, then produce a report with charts as HTML and PDF. Everything executes inside the nvcr.io/nvidia/deepstream:9.1-triton-multiarch container, so the host needs only Docker and the NVIDIA driver — no python, torch or trtexec to install. On Windows you need Docker Desktop with the WSL2 backend, otherwise --gpus will not work.
Before the first run, the skill bootstraps the environment in-container via setup.sh and installs wkhtmltopdf, then checks GPU, venv and trtexec with preflight.sh — proceed only on PASS. Artifacts land under models/{model_name}/ in a fixed layout, log and engine filenames are hard-coded, and a KITTI check gates the multi-stream step: zero frames or a detection rate under 90% stops the run.
What it helps with
- pull an object-detection ONNX model from Hugging Face or NGC
- build a dynamic TRT engine for a given batch size
- write a bbox parser for nvinfer that matches the model
- benchmark a pipeline across several video streams
- produce a PDF report with charts from the measurements
Example prompts for the agent
- Take PekingU/rtdetr_r50vd and build its TRT engine for DeepStream
- Import this model from NGC into a DeepStream pipeline and benchmark it on four streams
- Do a dry run: walk through the import steps without downloading or building anything
How to install deepstream-import-vision-model
1Skills CLI (Vercel)
npx skills add NVIDIA/skills --skill deepstream-import-vision-model -a claude-code2Manually: put the files in place
~/.claude/skills/deepstream-import-vision-model/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/deepstream-import-vision-model/
1Skills CLI (Vercel)
npx skills add NVIDIA/skills --skill deepstream-import-vision-model -a codex2Manually: put the files in place
.agents/skills/deepstream-import-vision-model/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add NVIDIA/skills --skill deepstream-import-vision-model -a cursor2Manually: put the files in place
.cursor/skills/deepstream-import-vision-model/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install NVIDIA/skills deepstream-import-vision-model2Manually: put the files in place
.github/skills/deepstream-import-vision-model/SKILL.md1Skills CLI (Vercel)
npx skills add NVIDIA/skills --skill deepstream-import-vision-model -a gemini-cli2Platform CLI
gemini skills install https://github.com/NVIDIA/skills --path skills/deepstream-import-vision-modelManually: put the files in place
.devin/skills/deepstream-import-vision-model/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add NVIDIA/skills --skill deepstream-import-vision-model -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about deepstream-import-vision-model
How do I install deepstream-import-vision-model in Claude?
Run npx skills add NVIDIA/skills --skill deepstream-import-vision-model -a claude-code. The skill then lives at ~/.claude/skills/deepstream-import-vision-model/SKILL.md.
What does deepstream-import-vision-model do?
It moves an object-detection model from HuggingFace or NGC through the whole DeepStream path: ONNX, TRT engine, bbox parser, benchmarks and a report. Object detection only.
What do I need for the skill to work?
Docker and the NVIDIA driver on the host are enough, since every step runs inside the nvcr.io/nvidia/deepstream:9.1-triton-multiarch container. On Windows, Docker Desktop with the WSL2 backend is required for the --gpus flag.
