deepstream-import-vision-model

67 / 100 · Review before installing

Check before installing. Some of what it can do is expected: a coding skill usually runs build and test commands.

By NVIDIA/skills (GitHub) · coding · 3,933 stars · scanned at commit 2630133

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.

Works with any agent that supports SKILL.md skills.

What this skill can do

Downloads and installs other software

The author does not mention this.

Found in 11 places.

skills/deepstream-import-vision-model/CHANGELOG.md, line 299`pip install` into the shared venv. `cleanup.sh` still removes any legacy
skills/deepstream-import-vision-model/SKILL.md, line 259| `No module named 'pyservicemaker'` | Install into venv: `pip install /opt/nvidia/deepstream/.../pyservicemaker*.whl` |
skills/deepstream-import-vision-model/references/model-acquire.md, line 164- For a new model that needs **extra packages** (e.g. `timm` for DETR-family backbones or `onnxsim`), `pip install` them **into the existing shared venv** rathe

Show 8 more

skills/deepstream-import-vision-model/references/model-acquire.md, line 167pip install timm # or: pip install onnxsim
skills/deepstream-import-vision-model/references/model-acquire.md, line 310pip install onnxsim
skills/deepstream-import-vision-model/references/pipeline-run.md, line 71echo " Fix 2: Install DeepStream samples (DeepStream 9.1 samples): apt-get install deepstream-9.1-samples"
skills/deepstream-import-vision-model/references/report-generation.md, line 491command -v wkhtmltopdf >/dev/null 2>&1 || { apt-get update -qq && apt-get install -y -qq wkhtmltopdf; }
skills/deepstream-import-vision-model/setup.sh, line 34python3 -m pip install --quiet --user virtualenv 2>/dev/null || python3 -m pip install --quiet virtualenv
skills/deepstream-import-vision-model/setup.sh, line 42"$PY" -m pip install --quiet --upgrade pip
skills/deepstream-import-vision-model/setup.sh, line 43"$PY" -m pip install -r "$SK/scripts/requirements.txt"
skills/deepstream-import-vision-model/setup.sh, line 52$APT apt-get update -qq && $APT apt-get install -y -qq wkhtmltopdf \

Runs commands on your computer

This is disclosed by the author in the skill's SKILL.md or README. Expected for a coding skill.

Found in 24 places.

skills/deepstream-import-vision-model/SKILL.md, line 92docker run --rm -it --gpus all --shm-size=16g -v "$PWD":/work -w /work \
skills/deepstream-import-vision-model/SKILL.md, line 99docker run --rm --gpus all -v "$PWD":/work -w /work \
skills/deepstream-import-vision-model/SKILL.md, line 137mkdir -p models/$MODEL_NAME/{model,parser,config,scripts,benchmarks/engines,benchmarks/ds,reports/charts,samples/kitti_output}

Show 17 more

skills/deepstream-import-vision-model/SKILL.md, line 159docker run --rm -it --gpus all --shm-size=16g -v "$PWD":/work -w /work \
skills/deepstream-import-vision-model/SKILL.md, line 169bash .claude/skills/deepstream-import-vision-model/scripts/model/safetensors-to-onnx.sh \
skills/deepstream-import-vision-model/SKILL.md, line 180bash .claude/skills/deepstream-import-vision-model/scripts/model/safetensors-to-onnx.sh \
skills/deepstream-import-vision-model/SKILL.md, line 189STEP_START=$(date +%s.%N)
skills/deepstream-import-vision-model/scripts/report/md-to-html-pdf.py, line 64subprocess.run(sudo + cmd, env=env,
skills/deepstream-import-vision-model/scripts/report/md-to-html-pdf.py, line 65stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL, timeout=600)
skills/deepstream-import-vision-model/scripts/report/md-to-html-pdf.py, line 168result = subprocess.run(
skills/deepstream-import-vision-model/scripts/report/md-to-html-pdf.py, line 180stdout=subprocess.PIPE,
skills/deepstream-import-vision-model/scripts/report/md-to-html-pdf.py, line 181stderr=subprocess.PIPE,
skills/deepstream-import-vision-model/scripts/report/render-mermaid-for-pdf.py, line 75r = subprocess.run(
skills/deepstream-import-vision-model/scripts/report/render-mermaid-for-pdf.py, line 77stdout=subprocess.PIPE,
skills/deepstream-import-vision-model/scripts/report/render-mermaid-for-pdf.py, line 78stderr=subprocess.PIPE,
skills/deepstream-import-vision-model/tests/test_hardened_scripts.py, line 58r = subprocess.run(
skills/deepstream-import-vision-model/tests/test_hardened_scripts.py, line 60stdout=subprocess.PIPE,
skills/deepstream-import-vision-model/tests/test_hardened_scripts.py, line 61stderr=subprocess.PIPE,
skills/deepstream-import-vision-model/tests/test_onnx_export_contract.py, line 77rc = subprocess.run(["bash", "-n", str(EXPORT_SH)], capture_output=True, text=True)
skills/deepstream-import-vision-model/tests/test_onnx_export_contract.py, line 81rc = subprocess.run(["bash", str(EXPORT_SH)], capture_output=True, text=True)

and 4 more places

Changes files outside its own folder

The author does not mention this.

Found in 1 place.

skills/deepstream-import-vision-model/install.sh, line 48bash install.sh --target ~/work/my-deepstream-project

Connects to the internet

The author does not mention this.

Found in 12 places.

skills/deepstream-import-vision-model/references/model-acquire.md, line 179wget -P models/$MODEL_NAME/hf_model "$HF_BASE/model.safetensors"
skills/deepstream-import-vision-model/references/model-acquire.md, line 180wget -P models/$MODEL_NAME/hf_model "$HF_BASE/config.json"
skills/deepstream-import-vision-model/references/model-acquire.md, line 181wget -P models/$MODEL_NAME/hf_model "$HF_BASE/preprocessor_config.json"

Show 9 more

skills/deepstream-import-vision-model/references/model-acquire.md, line 183wget -P models/$MODEL_NAME/hf_model "$HF_BASE/tokenizer.json" || true
skills/deepstream-import-vision-model/references/model-acquire.md, line 184wget -P models/$MODEL_NAME/hf_model "$HF_BASE/tokenizer_config.json" || true
skills/deepstream-import-vision-model/references/model-acquire.md, line 185wget -P models/$MODEL_NAME/hf_model "$HF_BASE/vocab.txt" || true
skills/deepstream-import-vision-model/references/model-acquire.md, line 186wget -P models/$MODEL_NAME/hf_model "$HF_BASE/special_tokens_map.json" || true
skills/deepstream-import-vision-model/references/model-acquire.md, line 409wget -O "models/$MODEL_NAME/model/$MODEL_NAME.onnx" "${ONNX_URL}"
skills/deepstream-import-vision-model/scripts/model/hf-download-config.sh, line 69if ! curl "${CURL_OPTS[@]}" "$URL"; then
skills/deepstream-import-vision-model/scripts/model/hf-list-files.sh, line 64RESPONSE="$(curl "${CURL_OPTS[@]}" "$URL")" || CURL_RC=$?
skills/deepstream-import-vision-model/scripts/model/ngc-download.sh, line 107if ! curl -fsSL --proto '=https' --tlsv1.2 --max-time 600 \
skills/deepstream-import-vision-model/scripts/model/ngc-list-files.sh, line 59if ! JSON="$(curl -fsSL --proto '=https' --tlsv1.2 --max-time 30 "${NGC_BASE}/" 2>/dev/null)"; then

Safer alternatives

Low-risk skills in the same category (coding) that look related. Check that one does what you need.

Install

Paste this into your terminal. It downloads exactly the version we scanned, not whatever is newest.

if [ -e ~/.agents/skills/'deepstream-import-vision-model' ] || [ -L ~/.agents/skills/'deepstream-import-vision-model' ]; then echo 'A skill folder named deepstream-import-vision-model already exists — not replacing it' >&2; false; else d=$(mktemp -d) && git clone --quiet --filter=blob:none --no-checkout 'https://github.com/NVIDIA/skills' "$d" && git -C "$d" checkout --quiet '2630133f22e357f2005886567a39eddcc1b520e2' -- 'skills/deepstream-import-vision-model' && mkdir -p ~/.agents/skills && cp -R "$d"/'skills/deepstream-import-vision-model' ~/.agents/skills/'deepstream-import-vision-model' && rm -rf "$d"; fi

Codex reads ~/.agents/skills directly. For Claude Code, Cursor, Copilot, Gemini CLI, Antigravity or OpenCode, also link it. For Claude Code: ln -s ~/.agents/skills/'deepstream-import-vision-model' ~/.claude/skills/'deepstream-import-vision-model' (other agents: ~/.cursor/skills, ~/.copilot/skills, ~/.gemini/skills, ~/.gemini/config/skills, ~/.config/opencode/skills). Or install through the finder skill, which does both.

Badge for your README

The badge for NVIDIA/skills shows the score of its lowest-scoring skill.

SkillTrust badge for NVIDIA/skills

[![SkillTrust](https://zalavijaysinh.github.io/skilltrust/badges/NVIDIA/skills.svg)](https://zalavijaysinh.github.io/skilltrust/repos/NVIDIA/skills/)

Score history

Date	Version	Score
2026-11-11	`2630133`	67
2026-11-08	`fce3cca`	67
2026-11-07	`fce3cca`	47